id	sid	tid	token	lemma	pos
bjmsr-315	1	1	copyright	copyright	NOUN
bjmsr-315	1	2	©	©	PROPN
bjmsr-315	1	3	cc	cc	PROPN
bjmsr-315	1	4	-	-	PUNCT
bjmsr-315	1	5	by	by	ADP
bjmsr-315	1	6	-	-	PUNCT
bjmsr-315	1	7	nc	nc	PROPN
bjmsr-315	1	8	2019	2019	NUM
bjmsr-315	1	9	,	,	PUNCT
bjmsr-315	1	10	bjmsr	bjmsr	PROPN
bjmsr-315	1	11	bangladesh	bangladesh	PROPN
bjmsr-315	1	12	journal	journal	PROPN
bjmsr-315	1	13	of	of	ADP
bjmsr-315	1	14	multidisciplinary	multidisciplinary	ADJ
bjmsr-315	1	15	scientific	scientific	ADJ
bjmsr-315	1	16	research	research	NOUN
bjmsr-315	1	17	vol	vol	NOUN
bjmsr-315	1	18	.	.	PROPN
bjmsr-315	1	19	1	1	NUM
bjmsr-315	1	20	,	,	PUNCT
bjmsr-315	1	21	no	no	INTJ
bjmsr-315	1	22	.	.	NOUN
bjmsr-315	1	23	1	1	NUM
bjmsr-315	1	24	april	april	PROPN
bjmsr-315	1	25	-	-	PUNCT
bjmsr-315	1	26	june	june	PROPN
bjmsr-315	1	27	;	;	PUNCT
bjmsr-315	1	28	2019	2019	NUM
bjmsr-315	1	29	published	publish	VERB
bjmsr-315	1	30	by	by	ADP
bjmsr-315	1	31	centre	centre	NOUN
bjmsr-315	1	32	for	for	ADP
bjmsr-315	1	33	research	research	NOUN
bjmsr-315	1	34	on	on	ADP
bjmsr-315	1	35	islamic	islamic	ADJ
bjmsr-315	1	36	banking	banking	PROPN
bjmsr-315	1	37	&	&	CCONJ
bjmsr-315	1	38	finance	finance	PROPN
bjmsr-315	1	39	and	and	CCONJ
bjmsr-315	1	40	business	business	NOUN
bjmsr-315	1	41	50	50	NUM
bjmsr-315	1	42	l1	l1	PROPN
bjmsr-315	1	43	norm	norm	NOUN
bjmsr-315	1	44	based	base	VERB
bjmsr-315	1	45	computational	computational	ADJ
bjmsr-315	1	46	algorithms	algorithm	NOUN
bjmsr-315	1	47	bijan	bijan	PROPN
bjmsr-315	1	48	bidabad	bidabad	PROPN
bjmsr-315	1	49	b.a	b.a	PROPN
bjmsr-315	1	50	.	.	PROPN
bjmsr-315	1	51	,	,	PUNCT
bjmsr-315	1	52	m.sc	m.sc	PROPN
bjmsr-315	1	53	.	.	PROPN
bjmsr-315	1	54	,	,	PUNCT
bjmsr-315	1	55	ph.d	ph.d	PROPN
bjmsr-315	1	56	.	.	PROPN
bjmsr-315	1	57	,	,	PUNCT
bjmsr-315	1	58	post	post	PROPN
bjmsr-315	1	59	-	-	ADJ
bjmsr-315	1	60	doc	doc	ADJ
bjmsr-315	1	61	.	.	PROPN
bjmsr-315	2	1	professor	professor	NOUN
bjmsr-315	2	2	economics	economic	NOUN
bjmsr-315	2	3	and	and	CCONJ
bjmsr-315	2	4	chief	chief	ADJ
bjmsr-315	2	5	islamic	islamic	PROPN
bjmsr-315	2	6	banking	banking	PROPN
bjmsr-315	2	7	advisor	advisor	PROPN
bjmsr-315	2	8	bank	bank	PROPN
bjmsr-315	2	9	melli	melli	PROPN
bjmsr-315	2	10	,	,	PUNCT
bjmsr-315	2	11	iran	iran	PROPN
bjmsr-315	2	12	e-mail:bijan@bidabad.com	e-mail:bijan@bidabad.com	X
bjmsr-315	3	1	abstract	abstract	ADJ
bjmsr-315	3	2	this	this	DET
bjmsr-315	3	3	paper	paper	NOUN
bjmsr-315	3	4	gives	give	VERB
bjmsr-315	3	5	a	a	DET
bjmsr-315	3	6	rather	rather	ADV
bjmsr-315	3	7	general	general	ADJ
bjmsr-315	3	8	review	review	NOUN
bjmsr-315	3	9	of	of	ADP
bjmsr-315	3	10	the	the	DET
bjmsr-315	3	11	l1	l1	PROPN
bjmsr-315	3	12	norm	norm	NOUN
bjmsr-315	3	13	algorithms	algorithm	NOUN
bjmsr-315	3	14	.	.	PUNCT
bjmsr-315	4	1	the	the	DET
bjmsr-315	4	2	chronology	chronology	NOUN
bjmsr-315	4	3	and	and	CCONJ
bjmsr-315	4	4	historical	historical	ADJ
bjmsr-315	4	5	development	development	NOUN
bjmsr-315	4	6	of	of	ADP
bjmsr-315	4	7	the	the	DET
bjmsr-315	4	8	l1	l1	PROPN
bjmsr-315	4	9	norm	norm	PROPN
bjmsr-315	4	10	estimation	estimation	NOUN
bjmsr-315	4	11	theory	theory	NOUN
bjmsr-315	4	12	for	for	ADP
bjmsr-315	4	13	the	the	DET
bjmsr-315	4	14	period	period	NOUN
bjmsr-315	4	15	of	of	ADP
bjmsr-315	4	16	1632	1632	NUM
bjmsr-315	4	17	-	-	SYM
bjmsr-315	4	18	1928	1928	NUM
bjmsr-315	4	19	will	will	AUX
bjmsr-315	4	20	be	be	AUX
bjmsr-315	4	21	surveyed	survey	VERB
bjmsr-315	4	22	and	and	CCONJ
bjmsr-315	4	23	the	the	DET
bjmsr-315	4	24	algorithms	algorithm	NOUN
bjmsr-315	4	25	belonging	belong	VERB
bjmsr-315	4	26	to	to	ADP
bjmsr-315	4	27	the	the	DET
bjmsr-315	4	28	after	after	ADP
bjmsr-315	4	29	1928	1928	NUM
bjmsr-315	4	30	period	period	NOUN
bjmsr-315	4	31	will	will	AUX
bjmsr-315	4	32	be	be	AUX
bjmsr-315	4	33	categorized	categorize	VERB
bjmsr-315	4	34	into	into	ADP
bjmsr-315	4	35	three	three	NUM
bjmsr-315	4	36	main	main	ADJ
bjmsr-315	4	37	classes	class	NOUN
bjmsr-315	4	38	of	of	ADP
bjmsr-315	4	39	direct	direct	ADJ
bjmsr-315	4	40	descent	descent	NOUN
bjmsr-315	4	41	,	,	PUNCT
bjmsr-315	4	42	simplex	simplex	NOUN
bjmsr-315	4	43	type	type	NOUN
bjmsr-315	4	44	,	,	PUNCT
bjmsr-315	4	45	and	and	CCONJ
bjmsr-315	4	46	other	other	ADJ
bjmsr-315	4	47	algorithms	algorithm	NOUN
bjmsr-315	4	48	.	.	PUNCT
bjmsr-315	5	1	keywords	keyword	NOUN
bjmsr-315	5	2	:	:	PUNCT
bjmsr-315	5	3	l1	l1	PROPN
bjmsr-315	5	4	norm	norm	NOUN
bjmsr-315	5	5	,	,	PUNCT
bjmsr-315	5	6	regression	regression	NOUN
bjmsr-315	5	7	,	,	PUNCT
bjmsr-315	5	8	algorithm	algorithm	NOUN
bjmsr-315	5	9	,	,	PUNCT
bjmsr-315	5	10	computer	computer	NOUN
bjmsr-315	5	11	historical	historical	ADJ
bjmsr-315	5	12	development	development	NOUN
bjmsr-315	5	13	(	(	PUNCT
bjmsr-315	5	14	1632	1632	NUM
bjmsr-315	5	15	-	-	SYM
bjmsr-315	5	16	1928	1928	NUM
bjmsr-315	5	17	)	)	PUNCT
bjmsr-315	5	18	the	the	DET
bjmsr-315	5	19	origin	origin	NOUN
bjmsr-315	5	20	of	of	ADP
bjmsr-315	5	21	l1	l1	PROPN
bjmsr-315	5	22	norm	norm	NOUN
bjmsr-315	5	23	estimation	estimation	NOUN
bjmsr-315	5	24	may	may	AUX
bjmsr-315	5	25	be	be	AUX
bjmsr-315	5	26	traced	trace	VERB
bjmsr-315	5	27	back	back	ADV
bjmsr-315	5	28	to	to	ADP
bjmsr-315	5	29	galilei	galilei	NOUN
bjmsr-315	5	30	(	(	PUNCT
bjmsr-315	5	31	1632	1632	NUM
bjmsr-315	5	32	)	)	PUNCT
bjmsr-315	5	33	.	.	PUNCT
bjmsr-315	6	1	in	in	ADP
bjmsr-315	6	2	determining	determine	VERB
bjmsr-315	6	3	the	the	DET
bjmsr-315	6	4	position	position	NOUN
bjmsr-315	6	5	of	of	ADP
bjmsr-315	6	6	a	a	DET
bjmsr-315	6	7	newly	newly	ADV
bjmsr-315	6	8	discovered	discover	VERB
bjmsr-315	6	9	star	star	NOUN
bjmsr-315	6	10	,	,	PUNCT
bjmsr-315	6	11	he	he	PRON
bjmsr-315	6	12	proposed	propose	VERB
bjmsr-315	6	13	the	the	DET
bjmsr-315	6	14	least	least	ADV
bjmsr-315	6	15	possible	possible	ADJ
bjmsr-315	6	16	correction	correction	NOUN
bjmsr-315	6	17	in	in	ADP
bjmsr-315	6	18	order	order	NOUN
bjmsr-315	6	19	to	to	PART
bjmsr-315	6	20	obtain	obtain	VERB
bjmsr-315	6	21	a	a	DET
bjmsr-315	6	22	reliable	reliable	ADJ
bjmsr-315	6	23	result	result	NOUN
bjmsr-315	6	24	(	(	PUNCT
bjmsr-315	6	25	see	see	VERB
bjmsr-315	6	26	,	,	PUNCT
bjmsr-315	6	27	ronchetti	ronchetti	X
bjmsr-315	6	28	(	(	PUNCT
bjmsr-315	6	29	1987	1987	NUM
bjmsr-315	6	30	)	)	PUNCT
bjmsr-315	6	31	for	for	ADP
bjmsr-315	6	32	some	some	DET
bjmsr-315	6	33	direct	direct	ADJ
bjmsr-315	6	34	quotations	quotation	NOUN
bjmsr-315	6	35	)	)	PUNCT
bjmsr-315	6	36	.	.	PUNCT
bjmsr-315	7	1	boscovich	boscovich	PROPN
bjmsr-315	7	2	(	(	PUNCT
bjmsr-315	7	3	1757	1757	NUM
bjmsr-315	7	4	)	)	PUNCT
bjmsr-315	7	5	for	for	ADP
bjmsr-315	7	6	the	the	DET
bjmsr-315	7	7	first	first	ADJ
bjmsr-315	7	8	time	time	NOUN
bjmsr-315	7	9	,	,	PUNCT
bjmsr-315	7	10	formulated	formulated	ADJ
bjmsr-315	7	11	and	and	CCONJ
bjmsr-315	7	12	applied	apply	VERB
bjmsr-315	7	13	the	the	DET
bjmsr-315	7	14	minimum	minimum	ADJ
bjmsr-315	7	15	sum	sum	NOUN
bjmsr-315	7	16	of	of	ADP
bjmsr-315	7	17	absolute	absolute	ADJ
bjmsr-315	7	18	errors	error	NOUN
bjmsr-315	7	19	for	for	ADP
bjmsr-315	7	20	obtaining	obtain	VERB
bjmsr-315	7	21	the	the	DET
bjmsr-315	7	22	best	good	ADJ
bjmsr-315	7	23	fitting	fitting	ADJ
bjmsr-315	7	24	line	line	NOUN
bjmsr-315	7	25	given	give	VERB
bjmsr-315	7	26	three	three	NUM
bjmsr-315	7	27	or	or	CCONJ
bjmsr-315	7	28	more	more	ADJ
bjmsr-315	7	29	pairs	pair	NOUN
bjmsr-315	7	30	of	of	ADP
bjmsr-315	7	31	observations	observation	NOUN
bjmsr-315	7	32	for	for	ADP
bjmsr-315	7	33	a	a	DET
bjmsr-315	7	34	simple	simple	ADJ
bjmsr-315	7	35	two	two	NUM
bjmsr-315	7	36	-	-	PUNCT
bjmsr-315	7	37	variable	variable	NOUN
bjmsr-315	7	38	regression	regression	NOUN
bjmsr-315	7	39	model	model	NOUN
bjmsr-315	7	40	.	.	PUNCT
bjmsr-315	8	1	he	he	PRON
bjmsr-315	8	2	also	also	ADV
bjmsr-315	8	3	restricts	restrict	VERB
bjmsr-315	8	4	the	the	DET
bjmsr-315	8	5	line	line	NOUN
bjmsr-315	8	6	to	to	PART
bjmsr-315	8	7	pass	pass	VERB
bjmsr-315	8	8	through	through	ADP
bjmsr-315	8	9	the	the	DET
bjmsr-315	8	10	means	mean	NOUN
bjmsr-315	8	11	of	of	ADP
bjmsr-315	8	12	the	the	DET
bjmsr-315	8	13	observation	observation	NOUN
bjmsr-315	8	14	points	point	NOUN
bjmsr-315	8	15	.	.	PUNCT
bjmsr-315	9	1	that	that	PRON
bjmsr-315	9	2	is	be	AUX
bjmsr-315	9	3	,	,	PUNCT
bjmsr-315	9	4	n	n	PRON
bjmsr-315	9	5	min	min	NOUN
bjmsr-315	9	6	:	:	PUNCT
bjmsr-315	10	1	σ	σ	PROPN
bjmsr-315	10	2	│	│	NOUN
bjmsr-315	10	3	yi	yi	PROPN
bjmsr-315	10	4	-	-	PUNCT
bjmsr-315	10	5	ß0	ß0	NOUN
bjmsr-315	10	6	-	-	PUNCT
bjmsr-315	10	7	ß1xi1	ß1xi1	NOUN
bjmsr-315	10	8	│	│	NOUN
bjmsr-315	10	9	ß0,ß1	ß0,ß1	PROPN
bjmsr-315	10	10	i=1	i=1	PROPN
bjmsr-315	10	11	n	n	PROPN
bjmsr-315	10	12	(	(	PUNCT
bjmsr-315	10	13	1	1	X
bjmsr-315	10	14	)	)	PUNCT
bjmsr-315	10	15	s.to	s.to	ADV
bjmsr-315	10	16	:	:	PUNCT
bjmsr-315	10	17	σ	σ	PROPN
bjmsr-315	10	18	(	(	PUNCT
bjmsr-315	10	19	yi	yi	NOUN
bjmsr-315	10	20	-	-	PUNCT
bjmsr-315	10	21	ß0	ß0	NOUN
bjmsr-315	10	22	-	-	PUNCT
bjmsr-315	10	23	ß1xi1)=0	ß1xi1)=0	ADJ
bjmsr-315	10	24	i=1	i=1	X
bjmsr-315	10	25	boscovich	boscovich	PROPN
bjmsr-315	10	26	(	(	PUNCT
bjmsr-315	10	27	1760	1760	NUM
bjmsr-315	10	28	)	)	PUNCT
bjmsr-315	10	29	gives	give	VERB
bjmsr-315	10	30	a	a	DET
bjmsr-315	10	31	simple	simple	ADJ
bjmsr-315	10	32	geometrical	geometrical	ADJ
bjmsr-315	10	33	solution	solution	NOUN
bjmsr-315	10	34	to	to	ADP
bjmsr-315	10	35	his	his	PRON
bjmsr-315	10	36	previous	previous	ADJ
bjmsr-315	10	37	suggestion	suggestion	NOUN
bjmsr-315	10	38	.	.	PUNCT
bjmsr-315	11	1	this	this	DET
bjmsr-315	11	2	paper	paper	NOUN
bjmsr-315	11	3	has	have	AUX
bjmsr-315	11	4	been	be	AUX
bjmsr-315	11	5	discussed	discuss	VERB
bjmsr-315	11	6	by	by	ADP
bjmsr-315	11	7	eisenhart	eisenhart	NOUN
bjmsr-315	11	8	(	(	PUNCT
bjmsr-315	11	9	1961	1961	NUM
bjmsr-315	11	10	)	)	PUNCT
bjmsr-315	11	11	and	and	CCONJ
bjmsr-315	11	12	sheynin	sheynin	NOUN
bjmsr-315	11	13	(	(	PUNCT
bjmsr-315	11	14	1973	1973	NUM
bjmsr-315	11	15	)	)	PUNCT
bjmsr-315	11	16	.	.	PUNCT
bjmsr-315	12	1	in	in	ADP
bjmsr-315	12	2	a	a	DET
bjmsr-315	12	3	manuscript	manuscript	NOUN
bjmsr-315	12	4	,	,	PUNCT
bjmsr-315	12	5	boscovich	boscovich	PROPN
bjmsr-315	12	6	poses	pose	VERB
bjmsr-315	12	7	the	the	DET
bjmsr-315	12	8	problem	problem	NOUN
bjmsr-315	12	9	to	to	ADP
bjmsr-315	12	10	simpson	simpson	PROPN
bjmsr-315	12	11	and	and	CCONJ
bjmsr-315	12	12	simpson	simpson	PROPN
bjmsr-315	12	13	gives	give	VERB
bjmsr-315	12	14	an	an	DET
bjmsr-315	12	15	analytical	analytical	ADJ
bjmsr-315	12	16	solution	solution	NOUN
bjmsr-315	12	17	to	to	ADP
bjmsr-315	12	18	the	the	DET
bjmsr-315	12	19	problem	problem	NOUN
bjmsr-315	12	20	(	(	PUNCT
bjmsr-315	12	21	see	see	VERB
bjmsr-315	12	22	,	,	PUNCT
bjmsr-315	12	23	stigler	stigler	NOUN
bjmsr-315	12	24	(	(	PUNCT
bjmsr-315	12	25	1984	1984	NUM
bjmsr-315	12	26	)	)	PUNCT
bjmsr-315	12	27	)	)	PUNCT
bjmsr-315	12	28	.	.	PUNCT
bjmsr-315	13	1	laplace	laplace	NOUN
bjmsr-315	13	2	(	(	PUNCT
bjmsr-315	13	3	1793	1793	NUM
bjmsr-315	13	4	)	)	PUNCT
bjmsr-315	13	5	provides	provide	VERB
bjmsr-315	13	6	an	an	DET
bjmsr-315	13	7	algebraic	algebraic	ADJ
bjmsr-315	13	8	formulation	formulation	NOUN
bjmsr-315	13	9	of	of	ADP
bjmsr-315	13	10	an	an	DET
bjmsr-315	13	11	algorithm	algorithm	NOUN
bjmsr-315	13	12	for	for	ADP
bjmsr-315	13	13	the	the	DET
bjmsr-315	13	14	l1	l1	PROPN
bjmsr-315	13	15	norm	norm	PROPN
bjmsr-315	13	16	regression	regression	PROPN
bjmsr-315	13	17	line	line	NOUN
bjmsr-315	13	18	,	,	PUNCT
bjmsr-315	13	19	which	which	PRON
bjmsr-315	13	20	passes	pass	VERB
bjmsr-315	13	21	through	through	ADP
bjmsr-315	13	22	the	the	DET
bjmsr-315	13	23	centroid	centroid	NOUN
bjmsr-315	13	24	of	of	ADP
bjmsr-315	13	25	observations	observation	NOUN
bjmsr-315	13	26	.	.	PUNCT
bjmsr-315	14	1	in	in	ADP
bjmsr-315	14	2	laplace	laplace	NOUN
bjmsr-315	14	3	(	(	PUNCT
bjmsr-315	14	4	1799	1799	NUM
bjmsr-315	14	5	)	)	PUNCT
bjmsr-315	14	6	,	,	PUNCT
bjmsr-315	14	7	an	an	DET
bjmsr-315	14	8	extension	extension	NOUN
bjmsr-315	14	9	of	of	ADP
bjmsr-315	14	10	l1	l1	PROPN
bjmsr-315	14	11	norm	norm	PROPN
bjmsr-315	14	12	regression	regression	NOUN
bjmsr-315	14	13	to	to	ADP
bjmsr-315	14	14	observations	observation	NOUN
bjmsr-315	14	15	with	with	ADP
bjmsr-315	14	16	different	different	ADJ
bjmsr-315	14	17	weights	weight	NOUN
bjmsr-315	14	18	has	have	AUX
bjmsr-315	14	19	also	also	ADV
bjmsr-315	14	20	been	be	AUX
bjmsr-315	14	21	discussed	discuss	VERB
bjmsr-315	14	22	.	.	PUNCT
bjmsr-315	15	1	prony	prony	ADJ
bjmsr-315	15	2	(	(	PUNCT
bjmsr-315	15	3	1804	1804	NUM
bjmsr-315	15	4	)	)	PUNCT
bjmsr-315	15	5	gives	give	VERB
bjmsr-315	15	6	a	a	DET
bjmsr-315	15	7	geometric	geometric	ADJ
bjmsr-315	15	8	interpretation	interpretation	NOUN
bjmsr-315	15	9	of	of	ADP
bjmsr-315	15	10	laplace	laplace	NOUN
bjmsr-315	15	11	's	's	PART
bjmsr-315	15	12	(	(	PUNCT
bjmsr-315	15	13	1799	1799	NUM
bjmsr-315	15	14	)	)	PUNCT
bjmsr-315	15	15	method	method	NOUN
bjmsr-315	15	16	and	and	CCONJ
bjmsr-315	15	17	compares	compare	VERB
bjmsr-315	15	18	it	it	PRON
bjmsr-315	15	19	with	with	ADP
bjmsr-315	15	20	other	other	ADJ
bjmsr-315	15	21	methods	method	NOUN
bjmsr-315	15	22	through	through	ADP
bjmsr-315	15	23	an	an	DET
bjmsr-315	15	24	example	example	NOUN
bjmsr-315	15	25	.	.	PUNCT
bjmsr-315	16	1	svanberg	svanberg	PROPN
bjmsr-315	16	2	(	(	PUNCT
bjmsr-315	16	3	1805	1805	NUM
bjmsr-315	16	4	)	)	PUNCT
bjmsr-315	16	5	applies	apply	VERB
bjmsr-315	16	6	laplace	laplace	NOUN
bjmsr-315	16	7	's	's	PART
bjmsr-315	16	8	method	method	NOUN
bjmsr-315	16	9	in	in	ADP
bjmsr-315	16	10	determining	determine	VERB
bjmsr-315	16	11	a	a	DET
bjmsr-315	16	12	meridian	meridian	ADJ
bjmsr-315	16	13	arc	arc	NOUN
bjmsr-315	16	14	,	,	PUNCT
bjmsr-315	16	15	and	and	CCONJ
bjmsr-315	16	16	von	von	PROPN
bjmsr-315	16	17	lindenau	lindenau	PROPN
bjmsr-315	16	18	(	(	PUNCT
bjmsr-315	16	19	1806	1806	NUM
bjmsr-315	16	20	)	)	PUNCT
bjmsr-315	16	21	uses	use	VERB
bjmsr-315	16	22	this	this	DET
bjmsr-315	16	23	method	method	NOUN
bjmsr-315	16	24	in	in	ADP
bjmsr-315	16	25	determination	determination	NOUN
bjmsr-315	16	26	of	of	ADP
bjmsr-315	16	27	the	the	DET
bjmsr-315	16	28	elliptic	elliptic	ADJ
bjmsr-315	16	29	meridian	meridian	PROPN
bjmsr-315	16	30	.	.	PUNCT
bjmsr-315	17	1	gauss	gauss	PROPN
bjmsr-315	17	2	(	(	PUNCT
bjmsr-315	17	3	1809	1809	NUM
bjmsr-315	17	4	)	)	PUNCT
bjmsr-315	17	5	suggests	suggest	VERB
bjmsr-315	17	6	the	the	DET
bjmsr-315	17	7	minimization	minimization	NOUN
bjmsr-315	17	8	of	of	ADP
bjmsr-315	17	9	the	the	DET
bjmsr-315	17	10	sum	sum	NOUN
bjmsr-315	17	11	of	of	ADP
bjmsr-315	17	12	absolute	absolute	ADJ
bjmsr-315	17	13	errors	error	NOUN
bjmsr-315	17	14	without	without	ADP
bjmsr-315	17	15	constraint	constraint	NOUN
bjmsr-315	17	16	.	.	PUNCT
bjmsr-315	18	1	he	he	PRON
bjmsr-315	18	2	concludes	conclude	VERB
bjmsr-315	18	3	that	that	SCONJ
bjmsr-315	18	4	this	this	DET
bjmsr-315	18	5	criterion	criterion	NOUN
bjmsr-315	18	6	necessarily	necessarily	ADV
bjmsr-315	18	7	sets	set	VERB
bjmsr-315	18	8	m	m	PRON
bjmsr-315	18	9	of	of	ADP
bjmsr-315	18	10	the	the	DET
bjmsr-315	18	11	residuals	residual	NOUN
bjmsr-315	18	12	equal	equal	ADJ
bjmsr-315	18	13	to	to	ADP
bjmsr-315	18	14	zero	zero	NUM
bjmsr-315	18	15	,	,	PUNCT
bjmsr-315	18	16	where	where	SCONJ
bjmsr-315	18	17	m	m	NOUN
bjmsr-315	18	18	is	be	AUX
bjmsr-315	18	19	the	the	DET
bjmsr-315	18	20	number	number	NOUN
bjmsr-315	18	21	of	of	ADP
bjmsr-315	18	22	parameters	parameter	NOUN
bjmsr-315	18	23	,	,	PUNCT
bjmsr-315	18	24	and	and	CCONJ
bjmsr-315	18	25	further	far	ADV
bjmsr-315	18	26	,	,	PUNCT
bjmsr-315	18	27	the	the	DET
bjmsr-315	18	28	solution	solution	NOUN
bjmsr-315	18	29	obtained	obtain	VERB
bjmsr-315	18	30	by	by	ADP
bjmsr-315	18	31	this	this	DET
bjmsr-315	18	32	method	method	NOUN
bjmsr-315	18	33	is	be	AUX
bjmsr-315	18	34	not	not	PART
bjmsr-315	18	35	changed	change	VERB
bjmsr-315	18	36	if	if	SCONJ
bjmsr-315	18	37	the	the	DET
bjmsr-315	18	38	value	value	NOUN
bjmsr-315	18	39	of	of	ADP
bjmsr-315	18	40	the	the	DET
bjmsr-315	18	41	dependent	dependent	ADJ
bjmsr-315	18	42	variable	variable	NOUN
bjmsr-315	18	43	is	be	AUX
bjmsr-315	18	44	increased	increase	VERB
bjmsr-315	18	45	or	or	CCONJ
bjmsr-315	18	46	decreased	decrease	VERB
bjmsr-315	18	47	without	without	ADP
bjmsr-315	18	48	changing	change	VERB
bjmsr-315	18	49	the	the	DET
bjmsr-315	18	50	sign	sign	NOUN
bjmsr-315	18	51	of	of	ADP
bjmsr-315	18	52	the	the	DET
bjmsr-315	18	53	residual	residual	ADJ
bjmsr-315	18	54	.	.	PUNCT
bjmsr-315	19	1	this	this	DET
bjmsr-315	19	2	conclusion	conclusion	NOUN
bjmsr-315	19	3	is	be	AUX
bjmsr-315	19	4	recently	recently	ADV
bjmsr-315	19	5	discussed	discuss	VERB
bjmsr-315	19	6	by	by	ADP
bjmsr-315	19	7	narula	narula	NOUN
bjmsr-315	19	8	and	and	CCONJ
bjmsr-315	19	9	wellington	wellington	PROPN
bjmsr-315	19	10	(	(	PUNCT
bjmsr-315	19	11	1985	1985	NUM
bjmsr-315	19	12	)	)	PUNCT
bjmsr-315	19	13	.	.	PUNCT
bjmsr-315	20	1	he	he	PRON
bjmsr-315	20	2	also	also	ADV
bjmsr-315	20	3	noted	note	VERB
bjmsr-315	20	4	that	that	SCONJ
bjmsr-315	20	5	boscovich	boscovich	PROPN
bjmsr-315	20	6	or	or	CCONJ
bjmsr-315	20	7	laplace	laplace	NOUN
bjmsr-315	20	8	estimators	estimator	NOUN
bjmsr-315	20	9	which	which	PRON
bjmsr-315	20	10	minimize	minimize	VERB
bjmsr-315	20	11	the	the	DET
bjmsr-315	20	12	sum	sum	NOUN
bjmsr-315	20	13	of	of	ADP
bjmsr-315	20	14	absolute	absolute	ADJ
bjmsr-315	20	15	residuals	residual	NOUN
bjmsr-315	20	16	with	with	ADP
bjmsr-315	20	17	zero	zero	NUM
bjmsr-315	20	18	-	-	PUNCT
bjmsr-315	20	19	sum	sum	NOUN
bjmsr-315	20	20	of	of	ADP
bjmsr-315	20	21	residuals	residual	NOUN
bjmsr-315	20	22	constraint	constraint	NOUN
bjmsr-315	20	23	,	,	PUNCT
bjmsr-315	20	24	necessarily	necessarily	ADV
bjmsr-315	20	25	set	set	VERB
bjmsr-315	20	26	m-1	m-1	NUM
bjmsr-315	20	27	of	of	ADP
bjmsr-315	20	28	the	the	DET
bjmsr-315	20	29	residuals	residual	NOUN
bjmsr-315	20	30	equal	equal	ADJ
bjmsr-315	20	31	to	to	ADP
bjmsr-315	20	32	zero	zero	NUM
bjmsr-315	20	33	(	(	PUNCT
bjmsr-315	20	34	see	see	NOUN
bjmsr-315	20	35	,	,	PUNCT
bjmsr-315	20	36	stigler	stigler	NOUN
bjmsr-315	20	37	(	(	PUNCT
bjmsr-315	20	38	1981	1981	NUM
bjmsr-315	20	39	)	)	PUNCT
bjmsr-315	20	40	,	,	PUNCT
bjmsr-315	20	41	farebrother	farebrother	ADV
bjmsr-315	20	42	(	(	PUNCT
bjmsr-315	20	43	1987b	1987b	NUM
bjmsr-315	20	44	)	)	PUNCT
bjmsr-315	20	45	)	)	PUNCT
bjmsr-315	20	46	.	.	PUNCT
bjmsr-315	21	1	mathieu	mathieu	PROPN
bjmsr-315	21	2	(	(	PUNCT
bjmsr-315	21	3	1816	1816	NUM
bjmsr-315	21	4	)	)	PUNCT
bjmsr-315	21	5	used	use	VERB
bjmsr-315	21	6	laplace	laplace	NOUN
bjmsr-315	21	7	's	's	PART
bjmsr-315	21	8	method	method	NOUN
bjmsr-315	21	9	to	to	PART
bjmsr-315	21	10	compute	compute	VERB
bjmsr-315	21	11	the	the	DET
bjmsr-315	21	12	eccentricity	eccentricity	NOUN
bjmsr-315	21	13	of	of	ADP
bjmsr-315	21	14	the	the	DET
bjmsr-315	21	15	earth	earth	NOUN
bjmsr-315	21	16	.	.	PUNCT
bjmsr-315	22	1	van	van	PROPN
bjmsr-315	22	2	beeck	beeck	NOUN
bjmsr-315	22	3	-	-	PUNCT
bjmsr-315	22	4	calkoen	calkoen	NOUN
bjmsr-315	22	5	(	(	PUNCT
bjmsr-315	22	6	1816	1816	NUM
bjmsr-315	22	7	)	)	PUNCT
bjmsr-315	22	8	advocates	advocate	VERB
bjmsr-315	22	9	the	the	DET
bjmsr-315	22	10	using	using	NOUN
bjmsr-315	22	11	of	of	ADP
bjmsr-315	22	12	the	the	DET
bjmsr-315	22	13	least	least	ADJ
bjmsr-315	22	14	absolute	absolute	ADJ
bjmsr-315	22	15	values	value	NOUN
bjmsr-315	22	16	criterion	criterion	NOUN
bjmsr-315	22	17	in	in	ADP
bjmsr-315	22	18	fitting	fitting	ADJ
bjmsr-315	22	19	curvilinear	curvilinear	ADJ
bjmsr-315	22	20	equation	equation	NOUN
bjmsr-315	22	21	obtained	obtain	VERB
bjmsr-315	22	22	by	by	ADP
bjmsr-315	22	23	using	use	VERB
bjmsr-315	22	24	powers	power	NOUN
bjmsr-315	22	25	of	of	ADP
bjmsr-315	22	26	the	the	DET
bjmsr-315	22	27	independent	independent	ADJ
bjmsr-315	22	28	variable	variable	NOUN
bjmsr-315	22	29	.	.	PUNCT
bjmsr-315	23	1	laplace	laplace	NOUN
bjmsr-315	23	2	(	(	PUNCT
bjmsr-315	23	3	1818	1818	NUM
bjmsr-315	23	4	)	)	PUNCT
bjmsr-315	23	5	adapted	adapt	VERB
bjmsr-315	23	6	boscovich	boscovich	NOUN
bjmsr-315	23	7	's	's	PART
bjmsr-315	23	8	criterion	criterion	NOUN
bjmsr-315	23	9	again	again	ADV
bjmsr-315	23	10	and	and	CCONJ
bjmsr-315	23	11	gave	give	VERB
bjmsr-315	23	12	an	an	DET
bjmsr-315	23	13	algebraic	algebraic	ADJ
bjmsr-315	23	14	procedure	procedure	NOUN
bjmsr-315	23	15	(	(	PUNCT
bjmsr-315	23	16	see	see	VERB
bjmsr-315	23	17	,	,	PUNCT
bjmsr-315	23	18	farebrother	farebrother	ADV
bjmsr-315	23	19	(	(	PUNCT
bjmsr-315	23	20	1987b	1987b	NUM
bjmsr-315	23	21	)	)	PUNCT
bjmsr-315	23	22	)	)	PUNCT
bjmsr-315	23	23	.	.	PUNCT
bjmsr-315	24	1	let	let	VERB
bjmsr-315	24	2	1x	1x	PRON
bjmsr-315	24	3	and	and	CCONJ
bjmsr-315	24	4	y	y	PROPN
bjmsr-315	24	5	be	be	AUX
bjmsr-315	24	6	the	the	DET
bjmsr-315	24	7	means	mean	NOUN
bjmsr-315	24	8	of	of	ADP
bjmsr-315	24	9	xi1	xi1	PROPN
bjmsr-315	24	10	and	and	CCONJ
bjmsr-315	24	11	yi	yi	PROPN
bjmsr-315	24	12	then	then	ADV
bjmsr-315	24	13	,	,	PUNCT
bjmsr-315	24	14	ß0	ß0	NOUN
bjmsr-315	24	15	=	=	SYM
bjmsr-315	24	16	y	y	PROPN
bjmsr-315	24	17	ß1	ß1	PROPN
bjmsr-315	24	18	1x	1x	NUM
bjmsr-315	24	19	(	(	PUNCT
bjmsr-315	24	20	2	2	NUM
bjmsr-315	24	21	)	)	PUNCT
bjmsr-315	24	22	value	value	NOUN
bjmsr-315	24	23	of	of	ADP
bjmsr-315	24	24	ß1	ß1	PROPN
bjmsr-315	24	25	is	be	AUX
bjmsr-315	24	26	found	find	VERB
bjmsr-315	24	27	by	by	ADP
bjmsr-315	24	28	,	,	PUNCT
bjmsr-315	25	1	n	n	X
bjmsr-315	25	2	mailto:bijan@bidabad.com	mailto:bijan@bidabad.com	X
bjmsr-315	25	3	copyright	copyright	PROPN
bjmsr-315	25	4	©	©	PROPN
bjmsr-315	25	5	cc	cc	PROPN
bjmsr-315	25	6	-	-	PUNCT
bjmsr-315	25	7	by	by	ADP
bjmsr-315	25	8	-	-	PUNCT
bjmsr-315	25	9	nc	nc	PROPN
bjmsr-315	25	10	2019	2019	NUM
bjmsr-315	25	11	,	,	PUNCT
bjmsr-315	25	12	bjmsr	bjmsr	PROPN
bjmsr-315	25	13	www.cribfb.com/journal/index.php/bjmsr	www.cribfb.com/journal/index.php/bjmsr	X
bjmsr-315	26	1	bangladesh	bangladesh	PROPN
bjmsr-315	26	2	journal	journal	PROPN
bjmsr-315	26	3	of	of	ADP
bjmsr-315	26	4	multidisciplinary	multidisciplinary	ADJ
bjmsr-315	26	5	scientific	scientific	ADJ
bjmsr-315	26	6	research	research	NOUN
bjmsr-315	26	7	vol	vol	NOUN
bjmsr-315	26	8	.	.	PROPN
bjmsr-315	26	9	1	1	NUM
bjmsr-315	26	10	,	,	PUNCT
bjmsr-315	26	11	no	no	INTJ
bjmsr-315	26	12	.	.	NOUN
bjmsr-315	26	13	1	1	NUM
bjmsr-315	26	14	;	;	PUNCT
bjmsr-315	26	15	2019	2019	NUM
bjmsr-315	26	16	51	51	NUM
bjmsr-315	26	17	min	min	NOUN
bjmsr-315	26	18	:	:	PUNCT
bjmsr-315	26	19	s	s	X
bjmsr-315	26	20	=	=	SYM
bjmsr-315	26	21	σ	σ	X
bjmsr-315	26	22	│	│	ADJ
bjmsr-315	26	23	yi~	yi~	PROPN
bjmsr-315	26	24	ß1xi1~	ß1xi1~	NOUN
bjmsr-315	26	25	│	│	X
bjmsr-315	26	26	(	(	PUNCT
bjmsr-315	26	27	3	3	X
bjmsr-315	26	28	)	)	PUNCT
bjmsr-315	26	29	ß1	ß1	NOUN
bjmsr-315	26	30	i=1	i=1	PROPN
bjmsr-315	26	31	where	where	SCONJ
bjmsr-315	26	32	xi1~	xi1~	PROPN
bjmsr-315	26	33	and	and	CCONJ
bjmsr-315	26	34	yi~	yi~	PROPN
bjmsr-315	26	35	are	be	AUX
bjmsr-315	26	36	deviations	deviation	NOUN
bjmsr-315	26	37	of	of	ADP
bjmsr-315	26	38	xi1	xi1	PROPN
bjmsr-315	26	39	and	and	CCONJ
bjmsr-315	26	40	yi	yi	PROPN
bjmsr-315	26	41	from	from	ADP
bjmsr-315	26	42	their	their	PRON
bjmsr-315	26	43	means	mean	NOUN
bjmsr-315	26	44	respectively	respectively	ADV
bjmsr-315	26	45	.	.	PUNCT
bjmsr-315	27	1	by	by	ADP
bjmsr-315	27	2	rearranging	rearrange	VERB
bjmsr-315	27	3	the	the	DET
bjmsr-315	27	4	observations	observation	NOUN
bjmsr-315	27	5	in	in	ADP
bjmsr-315	27	6	descending	descend	VERB
bjmsr-315	27	7	order	order	NOUN
bjmsr-315	27	8	of	of	ADP
bjmsr-315	27	9	yi~/xi1~	yi~/xi1~	NOUN
bjmsr-315	27	10	values	value	NOUN
bjmsr-315	27	11	,	,	PUNCT
bjmsr-315	27	12	laplace	laplace	NOUN
bjmsr-315	27	13	notes	note	VERB
bjmsr-315	27	14	that	that	SCONJ
bjmsr-315	27	15	s	s	VERB
bjmsr-315	27	16	is	be	AUX
bjmsr-315	27	17	infinite	infinite	ADJ
bjmsr-315	27	18	when	when	SCONJ
bjmsr-315	27	19	ß1	ß1	PROPN
bjmsr-315	27	20	is	be	AUX
bjmsr-315	27	21	infinite	infinite	ADJ
bjmsr-315	27	22	and	and	CCONJ
bjmsr-315	27	23	decreases	decrease	NOUN
bjmsr-315	27	24	as	as	SCONJ
bjmsr-315	27	25	ß1	ß1	PROPN
bjmsr-315	27	26	is	be	AUX
bjmsr-315	27	27	reduced	reduce	VERB
bjmsr-315	27	28	.	.	PUNCT
bjmsr-315	28	1	ß1	ß1	PROPN
bjmsr-315	28	2	reaches	reach	VERB
bjmsr-315	28	3	the	the	DET
bjmsr-315	28	4	critical	critical	ADJ
bjmsr-315	28	5	value	value	NOUN
bjmsr-315	28	6	yt~/xt1~	yt~/xt1~	NOUN
bjmsr-315	28	7	when	when	SCONJ
bjmsr-315	28	8	it	it	PRON
bjmsr-315	28	9	again	again	ADV
bjmsr-315	28	10	begins	begin	VERB
bjmsr-315	28	11	to	to	PART
bjmsr-315	28	12	increase	increase	VERB
bjmsr-315	28	13	.	.	PUNCT
bjmsr-315	29	1	this	this	DET
bjmsr-315	29	2	critical	critical	ADJ
bjmsr-315	29	3	value	value	NOUN
bjmsr-315	29	4	of	of	ADP
bjmsr-315	29	5	ß1	ß1	PROPN
bjmsr-315	29	6	is	be	AUX
bjmsr-315	29	7	determined	determine	VERB
bjmsr-315	29	8	when	when	SCONJ
bjmsr-315	29	9	,	,	PUNCT
bjmsr-315	29	10	t-1	t-1	PROPN
bjmsr-315	29	11	n	n	PROPN
bjmsr-315	29	12	t	t	PROPN
bjmsr-315	29	13	σ	σ	PROPN
bjmsr-315	29	14	│	│	NOUN
bjmsr-315	29	15	xi1~	xi1~	PROPN
bjmsr-315	29	16	│	│	X
bjmsr-315	29	17	<	<	X
bjmsr-315	29	18	½	½	PROPN
bjmsr-315	29	19	σ	σ	PROPN
bjmsr-315	29	20	│	│	NOUN
bjmsr-315	29	21	xi1~	xi1~	PROPN
bjmsr-315	29	22	│	│	PUNCT
bjmsr-315	29	23	≤	≤	NOUN
bjmsr-315	29	24	σ	σ	NUM
bjmsr-315	29	25	│	│	NOUN
bjmsr-315	29	26	xi1~	xi1~	PROPN
bjmsr-315	29	27	│	│	X
bjmsr-315	29	28	(	(	PUNCT
bjmsr-315	29	29	4	4	X
bjmsr-315	29	30	)	)	PUNCT
bjmsr-315	29	31	i=1	i=1	VERB
bjmsr-315	29	32	i=1	i=1	X
bjmsr-315	30	1	i=1	i=1	ADP
bjmsr-315	30	2	this	this	DET
bjmsr-315	30	3	procedure	procedure	NOUN
bjmsr-315	30	4	to	to	PART
bjmsr-315	30	5	find	find	VERB
bjmsr-315	30	6	ß1	ß1	PROPN
bjmsr-315	30	7	is	be	AUX
bjmsr-315	30	8	called	call	VERB
bjmsr-315	30	9	weighted	weight	VERB
bjmsr-315	30	10	median	median	NOUN
bjmsr-315	30	11	and	and	CCONJ
bjmsr-315	30	12	has	have	AUX
bjmsr-315	30	13	been	be	AUX
bjmsr-315	30	14	used	use	VERB
bjmsr-315	30	15	in	in	ADP
bjmsr-315	30	16	many	many	ADJ
bjmsr-315	30	17	other	other	ADJ
bjmsr-315	30	18	algorithms	algorithm	NOUN
bjmsr-315	30	19	such	such	ADJ
bjmsr-315	30	20	as	as	ADP
bjmsr-315	30	21	rhodes	rhode	NOUN
bjmsr-315	30	22	(	(	PUNCT
bjmsr-315	30	23	1930	1930	NUM
bjmsr-315	30	24	)	)	PUNCT
bjmsr-315	30	25	,	,	PUNCT
bjmsr-315	30	26	singleton	singleton	PROPN
bjmsr-315	30	27	(	(	PUNCT
bjmsr-315	30	28	1940	1940	NUM
bjmsr-315	30	29	)	)	PUNCT
bjmsr-315	30	30	,	,	PUNCT
bjmsr-315	30	31	karst	karst	PROPN
bjmsr-315	30	32	(	(	PUNCT
bjmsr-315	30	33	1958	1958	NUM
bjmsr-315	30	34	)	)	PUNCT
bjmsr-315	30	35	,	,	PUNCT
bjmsr-315	30	36	bloomfield	bloomfield	PROPN
bjmsr-315	30	37	and	and	CCONJ
bjmsr-315	30	38	steiger	steiger	PROPN
bjmsr-315	30	39	(	(	PUNCT
bjmsr-315	30	40	1980	1980	NUM
bjmsr-315	30	41	)	)	PUNCT
bjmsr-315	30	42	,	,	PUNCT
bjmsr-315	30	43	bidabad	bidabad	NOUN
bjmsr-315	30	44	(	(	PUNCT
bjmsr-315	30	45	1987a	1987a	NUM
bjmsr-315	30	46	,	,	PUNCT
bjmsr-315	30	47	b,88a	b,88a	ADJ
bjmsr-315	30	48	,	,	PUNCT
bjmsr-315	30	49	b,89a	b,89a	NOUN
bjmsr-315	30	50	,	,	PUNCT
bjmsr-315	30	51	b	b	NOUN
bjmsr-315	30	52	)	)	PUNCT
bjmsr-315	30	53	later	later	ADV
bjmsr-315	30	54	.	.	PUNCT
bjmsr-315	31	1	bidabad	bidabad	PROPN
bjmsr-315	31	2	(	(	PUNCT
bjmsr-315	31	3	1987a,88a,89a	1987a,88a,89a	NUM
bjmsr-315	31	4	,	,	PUNCT
bjmsr-315	31	5	b	b	X
bjmsr-315	31	6	)	)	PUNCT
bjmsr-315	31	7	derives	derive	VERB
bjmsr-315	31	8	the	the	DET
bjmsr-315	31	9	condition	condition	NOUN
bjmsr-315	31	10	(	(	PUNCT
bjmsr-315	31	11	4	4	NUM
bjmsr-315	31	12	)	)	PUNCT
bjmsr-315	31	13	via	via	ADP
bjmsr-315	31	14	discrete	discrete	ADJ
bjmsr-315	31	15	differentiation	differentiation	NOUN
bjmsr-315	31	16	method	method	NOUN
bjmsr-315	31	17	.	.	PUNCT
bjmsr-315	32	1	fourier	fourier	NOUN
bjmsr-315	32	2	(	(	PUNCT
bjmsr-315	32	3	1824	1824	NUM
bjmsr-315	32	4	)	)	PUNCT
bjmsr-315	32	5	formulates	formulate	VERB
bjmsr-315	32	6	least	least	ADJ
bjmsr-315	32	7	absolute	absolute	ADJ
bjmsr-315	32	8	residuals	residual	NOUN
bjmsr-315	32	9	regression	regression	NOUN
bjmsr-315	32	10	as	as	ADP
bjmsr-315	32	11	what	what	PRON
bjmsr-315	32	12	we	we	PRON
bjmsr-315	32	13	would	would	AUX
bjmsr-315	32	14	now	now	ADV
bjmsr-315	32	15	call	call	VERB
bjmsr-315	32	16	linear	linear	ADJ
bjmsr-315	32	17	programming	programming	NOUN
bjmsr-315	32	18	;	;	PUNCT
bjmsr-315	32	19	that	that	PRON
bjmsr-315	32	20	is	be	AUX
bjmsr-315	32	21	the	the	DET
bjmsr-315	32	22	minimization	minimization	NOUN
bjmsr-315	32	23	of	of	ADP
bjmsr-315	32	24	a	a	DET
bjmsr-315	32	25	linear	linear	ADJ
bjmsr-315	32	26	objective	objective	ADJ
bjmsr-315	32	27	function	function	NOUN
bjmsr-315	32	28	subject	subject	ADJ
bjmsr-315	32	29	to	to	ADP
bjmsr-315	32	30	linear	linear	ADJ
bjmsr-315	32	31	inequality	inequality	NOUN
bjmsr-315	32	32	constraints	constraint	NOUN
bjmsr-315	32	33	.	.	PUNCT
bjmsr-315	33	1	edgeworth	edgeworth	PROPN
bjmsr-315	33	2	(	(	PUNCT
bjmsr-315	33	3	1883	1883	NUM
bjmsr-315	33	4	)	)	PUNCT
bjmsr-315	33	5	presents	present	VERB
bjmsr-315	33	6	a	a	DET
bjmsr-315	33	7	philosophical	philosophical	ADJ
bjmsr-315	33	8	discussion	discussion	NOUN
bjmsr-315	33	9	on	on	ADP
bjmsr-315	33	10	differences	difference	NOUN
bjmsr-315	33	11	between	between	ADP
bjmsr-315	33	12	minimizing	minimize	VERB
bjmsr-315	33	13	mean	mean	ADJ
bjmsr-315	33	14	square	square	ADJ
bjmsr-315	33	15	errors	error	NOUN
bjmsr-315	33	16	and	and	CCONJ
bjmsr-315	33	17	mean	mean	VERB
bjmsr-315	33	18	absolute	absolute	ADJ
bjmsr-315	33	19	errors	error	NOUN
bjmsr-315	33	20	.	.	PUNCT
bjmsr-315	34	1	edgeworth	edgeworth	PROPN
bjmsr-315	34	2	(	(	PUNCT
bjmsr-315	34	3	1887a	1887a	NUM
bjmsr-315	34	4	,	,	PUNCT
bjmsr-315	34	5	b	b	X
bjmsr-315	34	6	)	)	PUNCT
bjmsr-315	34	7	proposed	propose	VERB
bjmsr-315	34	8	a	a	DET
bjmsr-315	34	9	simple	simple	ADJ
bjmsr-315	34	10	method	method	NOUN
bjmsr-315	34	11	for	for	ADP
bjmsr-315	34	12	choosing	choose	VERB
bjmsr-315	34	13	the	the	DET
bjmsr-315	34	14	regression	regression	NOUN
bjmsr-315	34	15	parameters	parameter	NOUN
bjmsr-315	34	16	.	.	PUNCT
bjmsr-315	35	1	by	by	ADP
bjmsr-315	35	2	fixing	fix	VERB
bjmsr-315	35	3	m-1	m-1	NUM
bjmsr-315	35	4	of	of	ADP
bjmsr-315	35	5	the	the	DET
bjmsr-315	35	6	parameters	parameter	NOUN
bjmsr-315	35	7	,	,	PUNCT
bjmsr-315	35	8	he	he	PRON
bjmsr-315	35	9	used	use	VERB
bjmsr-315	35	10	laplace	laplace	NOUN
bjmsr-315	35	11	's	's	PART
bjmsr-315	35	12	procedure	procedure	NOUN
bjmsr-315	35	13	to	to	PART
bjmsr-315	35	14	determine	determine	VERB
bjmsr-315	35	15	the	the	DET
bjmsr-315	35	16	optimal	optimal	ADJ
bjmsr-315	35	17	value	value	NOUN
bjmsr-315	35	18	of	of	ADP
bjmsr-315	35	19	the	the	DET
bjmsr-315	35	20	remaining	remain	VERB
bjmsr-315	35	21	parameter	parameter	NOUN
bjmsr-315	35	22	.	.	PUNCT
bjmsr-315	36	1	repeating	repeat	VERB
bjmsr-315	36	2	this	this	DET
bjmsr-315	36	3	operation	operation	NOUN
bjmsr-315	36	4	for	for	ADP
bjmsr-315	36	5	a	a	DET
bjmsr-315	36	6	range	range	NOUN
bjmsr-315	36	7	of	of	ADP
bjmsr-315	36	8	values	value	NOUN
bjmsr-315	36	9	for	for	ADP
bjmsr-315	36	10	m-1	m-1	NUM
bjmsr-315	36	11	fixed	fix	VERB
bjmsr-315	36	12	parameters	parameter	NOUN
bjmsr-315	36	13	,	,	PUNCT
bjmsr-315	36	14	he	he	PRON
bjmsr-315	36	15	obtained	obtain	VERB
bjmsr-315	36	16	a	a	DET
bjmsr-315	36	17	set	set	NOUN
bjmsr-315	36	18	of	of	ADP
bjmsr-315	36	19	results	result	NOUN
bjmsr-315	36	20	for	for	ADP
bjmsr-315	36	21	each	each	PRON
bjmsr-315	36	22	of	of	ADP
bjmsr-315	36	23	m	m	NOUN
bjmsr-315	36	24	possible	possible	ADJ
bjmsr-315	36	25	choices	choice	NOUN
bjmsr-315	36	26	of	of	ADP
bjmsr-315	36	27	the	the	DET
bjmsr-315	36	28	free	free	ADJ
bjmsr-315	36	29	parameters	parameter	NOUN
bjmsr-315	36	30	.	.	PUNCT
bjmsr-315	37	1	edgeworth	edgeworth	PROPN
bjmsr-315	37	2	drops	drop	VERB
bjmsr-315	37	3	the	the	DET
bjmsr-315	37	4	restriction	restriction	NOUN
bjmsr-315	37	5	of	of	ADP
bjmsr-315	37	6	passing	pass	VERB
bjmsr-315	37	7	through	through	ADP
bjmsr-315	37	8	the	the	DET
bjmsr-315	37	9	centroid	centroid	NOUN
bjmsr-315	37	10	of	of	ADP
bjmsr-315	37	11	data	data	PROPN
bjmsr-315	37	12	.	.	PUNCT
bjmsr-315	38	1	turner	turner	PROPN
bjmsr-315	38	2	(	(	PUNCT
bjmsr-315	38	3	1887	1887	NUM
bjmsr-315	38	4	)	)	PUNCT
bjmsr-315	38	5	discusses	discuss	VERB
bjmsr-315	38	6	the	the	DET
bjmsr-315	38	7	problem	problem	NOUN
bjmsr-315	38	8	of	of	ADP
bjmsr-315	38	9	nonunique	nonunique	ADJ
bjmsr-315	38	10	solutions	solution	NOUN
bjmsr-315	38	11	under	under	ADP
bjmsr-315	38	12	the	the	DET
bjmsr-315	38	13	least	least	ADJ
bjmsr-315	38	14	absolute	absolute	ADJ
bjmsr-315	38	15	error	error	NOUN
bjmsr-315	38	16	criterion	criterion	NOUN
bjmsr-315	38	17	as	as	ADP
bjmsr-315	38	18	a	a	DET
bjmsr-315	38	19	graphical	graphical	ADJ
bjmsr-315	38	20	variant	variant	NOUN
bjmsr-315	38	21	of	of	ADP
bjmsr-315	38	22	edgeworth	edgeworth	PROPN
bjmsr-315	38	23	(	(	PUNCT
bjmsr-315	38	24	1887a	1887a	NUM
bjmsr-315	38	25	)	)	PUNCT
bjmsr-315	38	26	as	as	ADP
bjmsr-315	38	27	a	a	DET
bjmsr-315	38	28	possible	possible	ADJ
bjmsr-315	38	29	drawback	drawback	NOUN
bjmsr-315	38	30	to	to	ADP
bjmsr-315	38	31	the	the	DET
bjmsr-315	38	32	method	method	NOUN
bjmsr-315	38	33	.	.	PUNCT
bjmsr-315	39	1	edgeworth	edgeworth	PROPN
bjmsr-315	39	2	(	(	PUNCT
bjmsr-315	39	3	1888	1888	NUM
bjmsr-315	39	4	)	)	PUNCT
bjmsr-315	39	5	replies	reply	NOUN
bjmsr-315	39	6	to	to	ADP
bjmsr-315	39	7	turner	turner	PROPN
bjmsr-315	39	8	's	's	PART
bjmsr-315	39	9	criticism	criticism	NOUN
bjmsr-315	39	10	by	by	ADP
bjmsr-315	39	11	proposing	propose	VERB
bjmsr-315	39	12	a	a	DET
bjmsr-315	39	13	second	second	ADJ
bjmsr-315	39	14	method	method	NOUN
bjmsr-315	39	15	for	for	ADP
bjmsr-315	39	16	choosing	choose	VERB
bjmsr-315	39	17	the	the	DET
bjmsr-315	39	18	two	two	NUM
bjmsr-315	39	19	parameters	parameter	NOUN
bjmsr-315	39	20	of	of	ADP
bjmsr-315	39	21	least	least	ADJ
bjmsr-315	39	22	absolute	absolute	ADJ
bjmsr-315	39	23	error	error	NOUN
bjmsr-315	39	24	regression	regression	NOUN
bjmsr-315	39	25	of	of	ADP
bjmsr-315	39	26	a	a	DET
bjmsr-315	39	27	simple	simple	ADJ
bjmsr-315	39	28	linear	linear	NOUN
bjmsr-315	39	29	model	model	NOUN
bjmsr-315	39	30	which	which	PRON
bjmsr-315	39	31	makes	make	VERB
bjmsr-315	39	32	no	no	DET
bjmsr-315	39	33	use	use	NOUN
bjmsr-315	39	34	of	of	ADP
bjmsr-315	39	35	the	the	DET
bjmsr-315	39	36	median	median	ADJ
bjmsr-315	39	37	loci	loci	NOUN
bjmsr-315	39	38	of	of	ADP
bjmsr-315	39	39	his	his	PRON
bjmsr-315	39	40	first	first	ADJ
bjmsr-315	39	41	method	method	NOUN
bjmsr-315	39	42	.	.	PUNCT
bjmsr-315	40	1	edgeworth	edgeworth	PROPN
bjmsr-315	40	2	,	,	PUNCT
bjmsr-315	40	3	in	in	ADP
bjmsr-315	40	4	this	this	DET
bjmsr-315	40	5	paper	paper	NOUN
bjmsr-315	40	6	,	,	PUNCT
bjmsr-315	40	7	followed	follow	VERB
bjmsr-315	40	8	turner	turner	NOUN
bjmsr-315	40	9	's	's	PART
bjmsr-315	40	10	suggestion	suggestion	NOUN
bjmsr-315	40	11	for	for	ADP
bjmsr-315	40	12	graphical	graphical	ADJ
bjmsr-315	40	13	analysis	analysis	NOUN
bjmsr-315	40	14	of	of	ADP
bjmsr-315	40	15	steps	step	NOUN
bjmsr-315	40	16	to	to	PART
bjmsr-315	40	17	reach	reach	VERB
bjmsr-315	40	18	the	the	DET
bjmsr-315	40	19	minimum	minimum	ADJ
bjmsr-315	40	20	solution	solution	NOUN
bjmsr-315	40	21	.	.	PUNCT
bjmsr-315	41	1	before	before	ADP
bjmsr-315	41	2	referring	refer	VERB
bjmsr-315	41	3	to	to	ADP
bjmsr-315	41	4	double	double	ADJ
bjmsr-315	41	5	median	median	ADJ
bjmsr-315	41	6	method	method	NOUN
bjmsr-315	41	7	of	of	ADP
bjmsr-315	41	8	edgeworth	edgeworth	PROPN
bjmsr-315	41	9	(	(	PUNCT
bjmsr-315	41	10	1923	1923	NUM
bjmsr-315	41	11	)	)	PUNCT
bjmsr-315	41	12	,	,	PUNCT
bjmsr-315	41	13	it	it	PRON
bjmsr-315	41	14	should	should	AUX
bjmsr-315	41	15	be	be	AUX
bjmsr-315	41	16	noted	note	VERB
bjmsr-315	41	17	that	that	SCONJ
bjmsr-315	41	18	bowley	bowley	NOUN
bjmsr-315	41	19	(	(	PUNCT
bjmsr-315	41	20	1902	1902	NUM
bjmsr-315	41	21	)	)	PUNCT
bjmsr-315	41	22	completes	complete	VERB
bjmsr-315	41	23	the	the	DET
bjmsr-315	41	24	edgeworth	edgeworth	PROPN
bjmsr-315	41	25	's	's	PART
bjmsr-315	41	26	(	(	PUNCT
bjmsr-315	41	27	1902	1902	NUM
bjmsr-315	41	28	)	)	PUNCT
bjmsr-315	41	29	paper	paper	NOUN
bjmsr-315	41	30	by	by	ADP
bjmsr-315	41	31	a	a	DET
bjmsr-315	41	32	variant	variant	NOUN
bjmsr-315	41	33	of	of	ADP
bjmsr-315	41	34	double	double	ADJ
bjmsr-315	41	35	median	median	ADJ
bjmsr-315	41	36	method	method	NOUN
bjmsr-315	41	37	which	which	PRON
bjmsr-315	41	38	presented	present	VERB
bjmsr-315	41	39	after	after	ADP
bjmsr-315	41	40	him	he	PRON
bjmsr-315	41	41	by	by	ADP
bjmsr-315	41	42	edgeworth	edgeworth	PROPN
bjmsr-315	41	43	(	(	PUNCT
bjmsr-315	41	44	1923	1923	NUM
bjmsr-315	41	45	)	)	PUNCT
bjmsr-315	41	46	.	.	PUNCT
bjmsr-315	42	1	this	this	DET
bjmsr-315	42	2	variant	variant	NOUN
bjmsr-315	42	3	ignores	ignore	VERB
bjmsr-315	42	4	the	the	DET
bjmsr-315	42	5	weights	weight	NOUN
bjmsr-315	42	6	attached	attach	VERB
bjmsr-315	42	7	to	to	ADP
bjmsr-315	42	8	errors	error	NOUN
bjmsr-315	42	9	.	.	PUNCT
bjmsr-315	43	1	edgeworth	edgeworth	PROPN
bjmsr-315	43	2	(	(	PUNCT
bjmsr-315	43	3	1923	1923	NUM
bjmsr-315	43	4	)	)	PUNCT
bjmsr-315	43	5	discussed	discuss	VERB
bjmsr-315	43	6	the	the	DET
bjmsr-315	43	7	more	more	ADV
bjmsr-315	43	8	general	general	ADJ
bjmsr-315	43	9	problem	problem	NOUN
bjmsr-315	43	10	of	of	ADP
bjmsr-315	43	11	estimating	estimate	VERB
bjmsr-315	43	12	the	the	DET
bjmsr-315	43	13	simple	simple	ADJ
bjmsr-315	43	14	linear	linear	ADJ
bjmsr-315	43	15	regression	regression	NOUN
bjmsr-315	43	16	parameters	parameter	NOUN
bjmsr-315	43	17	by	by	ADP
bjmsr-315	43	18	minimizing	minimize	VERB
bjmsr-315	43	19	the	the	DET
bjmsr-315	43	20	weighted	weighted	ADJ
bjmsr-315	43	21	sum	sum	NOUN
bjmsr-315	43	22	of	of	ADP
bjmsr-315	43	23	the	the	DET
bjmsr-315	43	24	absolute	absolute	ADJ
bjmsr-315	43	25	residuals	residual	NOUN
bjmsr-315	43	26	.	.	PUNCT
bjmsr-315	44	1	he	he	PRON
bjmsr-315	44	2	restates	restate	VERB
bjmsr-315	44	3	the	the	DET
bjmsr-315	44	4	rationale	rationale	NOUN
bjmsr-315	44	5	for	for	ADP
bjmsr-315	44	6	the	the	DET
bjmsr-315	44	7	method	method	NOUN
bjmsr-315	44	8	and	and	CCONJ
bjmsr-315	44	9	illustrates	illustrate	VERB
bjmsr-315	44	10	its	its	PRON
bjmsr-315	44	11	usage	usage	NOUN
bjmsr-315	44	12	through	through	ADP
bjmsr-315	44	13	several	several	ADJ
bjmsr-315	44	14	examples	example	NOUN
bjmsr-315	44	15	.	.	PUNCT
bjmsr-315	45	1	he	he	PRON
bjmsr-315	45	2	also	also	ADV
bjmsr-315	45	3	considers	consider	VERB
bjmsr-315	45	4	the	the	DET
bjmsr-315	45	5	nonunique	nonunique	ADJ
bjmsr-315	45	6	solution	solution	NOUN
bjmsr-315	45	7	problem	problem	NOUN
bjmsr-315	45	8	.	.	PUNCT
bjmsr-315	46	1	his	his	PRON
bjmsr-315	46	2	contribution	contribution	NOUN
bjmsr-315	46	3	is	be	AUX
bjmsr-315	46	4	called	call	VERB
bjmsr-315	46	5	double	double	ADJ
bjmsr-315	46	6	median	median	ADJ
bjmsr-315	46	7	method	method	NOUN
bjmsr-315	46	8	.	.	PUNCT
bjmsr-315	47	1	estienne	estienne	NOUN
bjmsr-315	47	2	(	(	PUNCT
bjmsr-315	47	3	1926	1926	NUM
bjmsr-315	47	4	-	-	SYM
bjmsr-315	47	5	28	28	NUM
bjmsr-315	47	6	)	)	PUNCT
bjmsr-315	47	7	proposes	propose	VERB
bjmsr-315	47	8	replacing	replace	VERB
bjmsr-315	47	9	the	the	DET
bjmsr-315	47	10	classical	classical	ADJ
bjmsr-315	47	11	theory	theory	NOUN
bjmsr-315	47	12	of	of	ADP
bjmsr-315	47	13	errors	error	NOUN
bjmsr-315	47	14	of	of	ADP
bjmsr-315	47	15	data	datum	NOUN
bjmsr-315	47	16	based	base	VERB
bjmsr-315	47	17	on	on	ADP
bjmsr-315	47	18	least	least	ADJ
bjmsr-315	47	19	squares	square	NOUN
bjmsr-315	47	20	with	with	ADP
bjmsr-315	47	21	what	what	PRON
bjmsr-315	47	22	he	he	PRON
bjmsr-315	47	23	calls	call	VERB
bjmsr-315	47	24	a	a	DET
bjmsr-315	47	25	rational	rational	ADJ
bjmsr-315	47	26	theory	theory	NOUN
bjmsr-315	47	27	based	base	VERB
bjmsr-315	47	28	on	on	ADP
bjmsr-315	47	29	the	the	DET
bjmsr-315	47	30	least	least	ADJ
bjmsr-315	47	31	absolute	absolute	ADJ
bjmsr-315	47	32	residual	residual	ADJ
bjmsr-315	47	33	procedure	procedure	NOUN
bjmsr-315	47	34	.	.	PUNCT
bjmsr-315	48	1	bowley	bowley	PROPN
bjmsr-315	48	2	(	(	PUNCT
bjmsr-315	48	3	1928	1928	NUM
bjmsr-315	48	4	)	)	PUNCT
bjmsr-315	48	5	summarizes	summarize	VERB
bjmsr-315	48	6	the	the	DET
bjmsr-315	48	7	edgeworth	edgeworth	NOUN
bjmsr-315	48	8	's	's	PART
bjmsr-315	48	9	contributions	contribution	NOUN
bjmsr-315	48	10	to	to	ADP
bjmsr-315	48	11	mathematical	mathematical	ADJ
bjmsr-315	48	12	statistics	statistic	NOUN
bjmsr-315	48	13	,	,	PUNCT
bjmsr-315	48	14	which	which	PRON
bjmsr-315	48	15	includes	include	VERB
bjmsr-315	48	16	his	his	PRON
bjmsr-315	48	17	work	work	NOUN
bjmsr-315	48	18	on	on	ADP
bjmsr-315	48	19	l1	l1	PROPN
bjmsr-315	48	20	norm	norm	PROPN
bjmsr-315	48	21	regression	regression	PROPN
bjmsr-315	48	22	.	.	PUNCT
bjmsr-315	49	1	dufton	dufton	NOUN
bjmsr-315	49	2	(	(	PUNCT
bjmsr-315	49	3	1928	1928	NUM
bjmsr-315	49	4	)	)	PUNCT
bjmsr-315	49	5	also	also	ADV
bjmsr-315	49	6	gives	give	VERB
bjmsr-315	49	7	a	a	DET
bjmsr-315	49	8	graphical	graphical	ADJ
bjmsr-315	49	9	method	method	NOUN
bjmsr-315	49	10	of	of	ADP
bjmsr-315	49	11	fitting	fit	VERB
bjmsr-315	49	12	a	a	DET
bjmsr-315	49	13	regression	regression	NOUN
bjmsr-315	49	14	line	line	NOUN
bjmsr-315	49	15	.	.	PUNCT
bjmsr-315	50	1	farebrother	farebrother	ADV
bjmsr-315	50	2	(	(	PUNCT
bjmsr-315	50	3	1987b	1987b	NUM
bjmsr-315	50	4	)	)	PUNCT
bjmsr-315	50	5	summarizes	summarize	VERB
bjmsr-315	50	6	the	the	DET
bjmsr-315	50	7	important	important	ADJ
bjmsr-315	50	8	contributions	contribution	NOUN
bjmsr-315	50	9	to	to	ADP
bjmsr-315	50	10	l1	l1	PROPN
bjmsr-315	50	11	norm	norm	NOUN
bjmsr-315	50	12	regression	regression	NOUN
bjmsr-315	50	13	for	for	ADP
bjmsr-315	50	14	the	the	DET
bjmsr-315	50	15	period	period	NOUN
bjmsr-315	50	16	of	of	ADP
bjmsr-315	50	17	1793	1793	NUM
bjmsr-315	50	18	-	-	SYM
bjmsr-315	50	19	1930	1930	NUM
bjmsr-315	50	20	.	.	PUNCT
bjmsr-315	51	1	for	for	ADP
bjmsr-315	51	2	more	more	ADJ
bjmsr-315	51	3	references	reference	NOUN
bjmsr-315	51	4	see	see	VERB
bjmsr-315	51	5	also	also	ADV
bjmsr-315	51	6	crocker	crocker	PROPN
bjmsr-315	51	7	(	(	PUNCT
bjmsr-315	51	8	1969	1969	NUM
bjmsr-315	51	9	)	)	PUNCT
bjmsr-315	51	10	,	,	PUNCT
bjmsr-315	51	11	harter	harter	X
bjmsr-315	51	12	(	(	PUNCT
bjmsr-315	51	13	1974a	1974a	NUM
bjmsr-315	51	14	,	,	PUNCT
bjmsr-315	51	15	b,75a	b,75a	NOUN
bjmsr-315	51	16	,	,	PUNCT
bjmsr-315	51	17	b	b	NOUN
bjmsr-315	51	18	,	,	PUNCT
bjmsr-315	51	19	c,76	c,76	NUM
bjmsr-315	51	20	)	)	PUNCT
bjmsr-315	51	21	,	,	PUNCT
bjmsr-315	51	22	dielman	dielman	NOUN
bjmsr-315	51	23	(	(	PUNCT
bjmsr-315	51	24	1984	1984	NUM
bjmsr-315	51	25	)	)	PUNCT
bjmsr-315	51	26	.	.	PUNCT
bjmsr-315	52	1	up	up	ADP
bjmsr-315	52	2	to	to	ADP
bjmsr-315	52	3	1928	1928	NUM
bjmsr-315	52	4	,	,	PUNCT
bjmsr-315	52	5	all	all	DET
bjmsr-315	52	6	algorithms	algorithm	NOUN
bjmsr-315	52	7	had	have	AUX
bjmsr-315	52	8	been	be	AUX
bjmsr-315	52	9	proposed	propose	VERB
bjmsr-315	52	10	for	for	ADP
bjmsr-315	52	11	simple	simple	ADJ
bjmsr-315	52	12	linear	linear	ADJ
bjmsr-315	52	13	regression	regression	NOUN
bjmsr-315	52	14	.	.	PUNCT
bjmsr-315	53	1	though	though	SCONJ
bjmsr-315	53	2	some	some	PRON
bjmsr-315	53	3	of	of	ADP
bjmsr-315	53	4	them	they	PRON
bjmsr-315	53	5	use	use	VERB
bjmsr-315	53	6	algebraic	algebraic	ADJ
bjmsr-315	53	7	propositions	proposition	NOUN
bjmsr-315	53	8	,	,	PUNCT
bjmsr-315	53	9	are	be	AUX
bjmsr-315	53	10	not	not	PART
bjmsr-315	53	11	so	so	ADV
bjmsr-315	53	12	organized	organized	ADJ
bjmsr-315	53	13	to	to	PART
bjmsr-315	53	14	handle	handle	VERB
bjmsr-315	53	15	multiple	multiple	ADJ
bjmsr-315	53	16	l1	l1	PROPN
bjmsr-315	53	17	norm	norm	NOUN
bjmsr-315	53	18	regression	regression	NOUN
bjmsr-315	53	19	problem	problem	NOUN
bjmsr-315	53	20	.	.	PUNCT
bjmsr-315	54	1	in	in	ADP
bjmsr-315	54	2	the	the	DET
bjmsr-315	54	3	next	next	ADJ
bjmsr-315	54	4	section	section	NOUN
bjmsr-315	54	5	,	,	PUNCT
bjmsr-315	54	6	we	we	PRON
bjmsr-315	54	7	will	will	AUX
bjmsr-315	54	8	discuss	discuss	VERB
bjmsr-315	54	9	the	the	DET
bjmsr-315	54	10	more	more	ADV
bjmsr-315	54	11	elaborated	elaborated	ADJ
bjmsr-315	54	12	computational	computational	ADJ
bjmsr-315	54	13	methods	method	NOUN
bjmsr-315	54	14	for	for	ADP
bjmsr-315	54	15	simple	simple	ADJ
bjmsr-315	54	16	and	and	CCONJ
bjmsr-315	54	17	multiple	multiple	ADJ
bjmsr-315	54	18	l1	l1	PROPN
bjmsr-315	54	19	norm	norm	NOUN
bjmsr-315	54	20	regressions	regression	NOUN
bjmsr-315	54	21	not	not	PART
bjmsr-315	54	22	in	in	ADP
bjmsr-315	54	23	a	a	DET
bjmsr-315	54	24	chronological	chronological	ADJ
bjmsr-315	54	25	sense	sense	NOUN
bjmsr-315	54	26	;	;	PUNCT
bjmsr-315	54	27	because	because	SCONJ
bjmsr-315	54	28	many	many	ADJ
bjmsr-315	54	29	digressions	digression	NOUN
bjmsr-315	54	30	have	have	AUX
bjmsr-315	54	31	occurred	occur	VERB
bjmsr-315	54	32	.	.	PUNCT
bjmsr-315	55	1	we	we	PRON
bjmsr-315	55	2	may	may	AUX
bjmsr-315	55	3	denote	denote	VERB
bjmsr-315	55	4	the	the	DET
bjmsr-315	55	5	period	period	NOUN
bjmsr-315	55	6	of	of	ADP
bjmsr-315	55	7	after	after	ADP
bjmsr-315	55	8	1928	1928	NUM
bjmsr-315	55	9	the	the	DET
bjmsr-315	55	10	time	time	NOUN
bjmsr-315	55	11	of	of	ADP
bjmsr-315	55	12	modern	modern	ADJ
bjmsr-315	55	13	algorithms	algorithm	NOUN
bjmsr-315	55	14	in	in	ADP
bjmsr-315	55	15	the	the	DET
bjmsr-315	55	16	subject	subject	NOUN
bjmsr-315	55	17	of	of	ADP
bjmsr-315	55	18	l1	l1	PROPN
bjmsr-315	55	19	norm	norm	PROPN
bjmsr-315	55	20	regression	regression	NOUN
bjmsr-315	55	21	.	.	PUNCT
bjmsr-315	56	1	computational	computational	ADJ
bjmsr-315	56	2	algorithms	algorithm	NOUN
bjmsr-315	56	3	although	although	SCONJ
bjmsr-315	56	4	a	a	DET
bjmsr-315	56	5	closed	closed	ADJ
bjmsr-315	56	6	form	form	NOUN
bjmsr-315	56	7	of	of	ADP
bjmsr-315	56	8	the	the	DET
bjmsr-315	56	9	solution	solution	NOUN
bjmsr-315	56	10	of	of	ADP
bjmsr-315	56	11	l1	l1	PROPN
bjmsr-315	56	12	norm	norm	NOUN
bjmsr-315	56	13	regression	regression	NOUN
bjmsr-315	56	14	has	have	AUX
bjmsr-315	56	15	not	not	PART
bjmsr-315	56	16	been	be	AUX
bjmsr-315	56	17	derived	derive	VERB
bjmsr-315	56	18	yet	yet	ADV
bjmsr-315	56	19	,	,	PUNCT
bjmsr-315	56	20	many	many	ADJ
bjmsr-315	56	21	algorithms	algorithm	NOUN
bjmsr-315	56	22	have	have	AUX
bjmsr-315	56	23	been	be	AUX
bjmsr-315	56	24	proposed	propose	VERB
bjmsr-315	56	25	to	to	PART
bjmsr-315	56	26	minimize	minimize	VERB
bjmsr-315	56	27	its	its	PRON
bjmsr-315	56	28	objective	objective	ADJ
bjmsr-315	56	29	function	function	NOUN
bjmsr-315	56	30	(	(	PUNCT
bjmsr-315	56	31	see	see	VERB
bjmsr-315	56	32	,	,	PUNCT
bjmsr-315	56	33	cheney	cheney	NOUN
bjmsr-315	56	34	(	(	PUNCT
bjmsr-315	56	35	1966	1966	NUM
bjmsr-315	56	36	)	)	PUNCT
bjmsr-315	56	37	,	,	PUNCT
bjmsr-315	56	38	chambers	chamber	NOUN
bjmsr-315	56	39	(	(	PUNCT
bjmsr-315	56	40	1977	1977	NUM
bjmsr-315	56	41	)	)	PUNCT
bjmsr-315	56	42	,	,	PUNCT
bjmsr-315	56	43	dielman	dielman	NOUN
bjmsr-315	56	44	and	and	CCONJ
bjmsr-315	56	45	pfaffenberger	pfaffenberger	ADV
bjmsr-315	56	46	(	(	PUNCT
bjmsr-315	56	47	1982,84	1982,84	NUM
bjmsr-315	56	48	)	)	PUNCT
bjmsr-315	56	49	)	)	PUNCT
bjmsr-315	56	50	.	.	PUNCT
bjmsr-315	57	1	generally	generally	ADV
bjmsr-315	57	2	,	,	PUNCT
bjmsr-315	57	3	we	we	PRON
bjmsr-315	57	4	can	can	AUX
bjmsr-315	57	5	classify	classify	VERB
bjmsr-315	57	6	all	all	DET
bjmsr-315	57	7	l1	l1	PROPN
bjmsr-315	57	8	norm	norm	NOUN
bjmsr-315	57	9	algorithms	algorithm	NOUN
bjmsr-315	57	10	in	in	ADP
bjmsr-315	57	11	three	three	NUM
bjmsr-315	57	12	major	major	ADJ
bjmsr-315	57	13	categories	category	NOUN
bjmsr-315	57	14	as	as	ADP
bjmsr-315	57	15	,	,	PUNCT
bjmsr-315	57	16	i	i	NOUN
bjmsr-315	57	17	)	)	PUNCT
bjmsr-315	57	18	direct	direct	ADJ
bjmsr-315	57	19	descent	descent	NOUN
bjmsr-315	57	20	algorithms	algorithms	PROPN
bjmsr-315	57	21	ii	ii	PROPN
bjmsr-315	57	22	)	)	PUNCT
bjmsr-315	57	23	simplex	simplex	NOUN
bjmsr-315	57	24	type	type	NOUN
bjmsr-315	57	25	algorithms	algorithms	PROPN
bjmsr-315	57	26	iii	iii	NOUN
bjmsr-315	57	27	)	)	PUNCT
bjmsr-315	57	28	other	other	ADJ
bjmsr-315	57	29	algorithms	algorithm	NOUN
bjmsr-315	57	30	which	which	PRON
bjmsr-315	57	31	will	will	AUX
bjmsr-315	57	32	be	be	AUX
bjmsr-315	57	33	discussed	discuss	VERB
bjmsr-315	57	34	in	in	ADP
bjmsr-315	57	35	the	the	DET
bjmsr-315	57	36	following	follow	VERB
bjmsr-315	57	37	sections	section	NOUN
bjmsr-315	57	38	sequentially	sequentially	ADV
bjmsr-315	57	39	.	.	PUNCT
bjmsr-315	58	1	direct	direct	ADJ
bjmsr-315	58	2	descent	descent	NOUN
bjmsr-315	58	3	algorithms	algorithm	VERB
bjmsr-315	58	4	the	the	DET
bjmsr-315	58	5	essence	essence	NOUN
bjmsr-315	58	6	of	of	ADP
bjmsr-315	58	7	the	the	DET
bjmsr-315	58	8	algorithms	algorithm	NOUN
bjmsr-315	58	9	which	which	PRON
bjmsr-315	58	10	fall	fall	VERB
bjmsr-315	58	11	within	within	ADP
bjmsr-315	58	12	this	this	DET
bjmsr-315	58	13	category	category	NOUN
bjmsr-315	58	14	is	be	AUX
bjmsr-315	58	15	finding	find	VERB
bjmsr-315	58	16	a	a	DET
bjmsr-315	58	17	steep	steep	ADJ
bjmsr-315	58	18	path	path	NOUN
bjmsr-315	58	19	to	to	PART
bjmsr-315	58	20	descend	descend	VERB
bjmsr-315	58	21	down	down	ADP
bjmsr-315	58	22	the	the	DET
bjmsr-315	58	23	polyhedron	polyhedron	NOUN
bjmsr-315	58	24	of	of	ADP
bjmsr-315	58	25	the	the	DET
bjmsr-315	58	26	l1	l1	PROPN
bjmsr-315	58	27	norm	norm	PROPN
bjmsr-315	58	28	regression	regression	VERB
bjmsr-315	58	29	objective	objective	ADJ
bjmsr-315	58	30	function	function	NOUN
bjmsr-315	58	31	.	.	PUNCT
bjmsr-315	59	1	although	although	SCONJ
bjmsr-315	59	2	the	the	DET
bjmsr-315	59	3	laplace	laplace	NOUN
bjmsr-315	59	4	's	's	PART
bjmsr-315	59	5	method	method	NOUN
bjmsr-315	59	6	(	(	PUNCT
bjmsr-315	59	7	explained	explain	VERB
bjmsr-315	59	8	hereinbefore	hereinbefore	NOUN
bjmsr-315	59	9	)	)	PUNCT
bjmsr-315	59	10	is	be	AUX
bjmsr-315	59	11	a	a	DET
bjmsr-315	59	12	special	special	ADJ
bjmsr-315	59	13	type	type	NOUN
bjmsr-315	59	14	of	of	ADP
bjmsr-315	59	15	direct	direct	ADJ
bjmsr-315	59	16	descent	descent	NOUN
bjmsr-315	59	17	algorithms	algorithm	NOUN
bjmsr-315	59	18	;	;	PUNCT
bjmsr-315	59	19	the	the	DET
bjmsr-315	59	20	origin	origin	NOUN
bjmsr-315	59	21	of	of	ADP
bjmsr-315	59	22	this	this	DET
bjmsr-315	59	23	procedure	procedure	NOUN
bjmsr-315	59	24	in	in	ADP
bjmsr-315	59	25	the	the	DET
bjmsr-315	59	26	area	area	NOUN
bjmsr-315	59	27	of	of	ADP
bjmsr-315	59	28	l1	l1	PROPN
bjmsr-315	59	29	norm	norm	NOUN
bjmsr-315	59	30	can	can	AUX
bjmsr-315	59	31	be	be	AUX
bjmsr-315	59	32	traced	trace	VERB
bjmsr-315	59	33	back	back	ADV
bjmsr-315	59	34	to	to	ADP
bjmsr-315	59	35	the	the	DET
bjmsr-315	59	36	algorithms	algorithm	NOUN
bjmsr-315	59	37	of	of	ADP
bjmsr-315	59	38	edgeworth	edgeworth	PROPN
bjmsr-315	59	39	which	which	PRON
bjmsr-315	59	40	were	be	AUX
bjmsr-315	59	41	explained	explain	VERB
bjmsr-315	59	42	in	in	ADP
bjmsr-315	59	43	the	the	DET
bjmsr-315	59	44	previous	previous	ADJ
bjmsr-315	59	45	section	section	NOUN
bjmsr-315	59	46	.	.	PUNCT
bjmsr-315	60	1	rhodes	rhode	NOUN
bjmsr-315	60	2	(	(	PUNCT
bjmsr-315	60	3	1930	1930	NUM
bjmsr-315	60	4	)	)	PUNCT
bjmsr-315	60	5	found	find	VERB
bjmsr-315	60	6	edgeworth	edgeworth	PROPN
bjmsr-315	60	7	's	's	PART
bjmsr-315	60	8	graphical	graphical	ADJ
bjmsr-315	60	9	solution	solution	NOUN
bjmsr-315	60	10	laborious	laborious	ADJ
bjmsr-315	60	11	;	;	PUNCT
bjmsr-315	60	12	therefore	therefore	ADV
bjmsr-315	60	13	,	,	PUNCT
bjmsr-315	60	14	he	he	PRON
bjmsr-315	60	15	suggested	suggest	VERB
bjmsr-315	60	16	an	an	DET
bjmsr-315	60	17	alternative	alternative	ADJ
bjmsr-315	60	18	method	method	NOUN
bjmsr-315	60	19	for	for	ADP
bjmsr-315	60	20	a	a	DET
bjmsr-315	60	21	general	general	ADJ
bjmsr-315	60	22	linear	linear	PROPN
bjmsr-315	60	23	model	model	NOUN
bjmsr-315	60	24	,	,	PUNCT
bjmsr-315	60	25	which	which	PRON
bjmsr-315	60	26	may	may	AUX
bjmsr-315	60	27	be	be	AUX
bjmsr-315	60	28	summarized	summarize	VERB
bjmsr-315	60	29	as	as	SCONJ
bjmsr-315	60	30	follows	follow	VERB
bjmsr-315	60	31	(	(	PUNCT
bjmsr-315	60	32	see	see	VERB
bjmsr-315	60	33	,	,	PUNCT
bjmsr-315	60	34	farebrother	farebrother	ADV
bjmsr-315	60	35	(	(	PUNCT
bjmsr-315	60	36	1987b	1987b	NUM
bjmsr-315	60	37	)	)	PUNCT
bjmsr-315	60	38	)	)	PUNCT
bjmsr-315	60	39	.	.	PUNCT
bjmsr-315	61	1	suppose	suppose	VERB
bjmsr-315	61	2	,	,	PUNCT
bjmsr-315	61	3	we	we	PRON
bjmsr-315	61	4	have	have	VERB
bjmsr-315	61	5	n	n	NUM
bjmsr-315	61	6	equations	equation	NOUN
bjmsr-315	61	7	with	with	ADP
bjmsr-315	61	8	m	m	PROPN
bjmsr-315	61	9	<	<	NOUN
bjmsr-315	61	10	n	n	PRON
bjmsr-315	61	11	unknown	unknown	ADJ
bjmsr-315	61	12	parameters	parameter	NOUN
bjmsr-315	61	13	.	.	PUNCT
bjmsr-315	62	1	to	to	PART
bjmsr-315	62	2	find	find	VERB
bjmsr-315	62	3	l1	l1	PROPN
bjmsr-315	62	4	norm	norm	NOUN
bjmsr-315	62	5	solution	solution	NOUN
bjmsr-315	62	6	of	of	ADP
bjmsr-315	62	7	this	this	DET
bjmsr-315	62	8	overdetermined	overdetermine	VERB
bjmsr-315	62	9	system	system	NOUN
bjmsr-315	62	10	of	of	ADP
bjmsr-315	62	11	equations	equation	NOUN
bjmsr-315	62	12	;	;	PUNCT
bjmsr-315	62	13	copyright	copyright	NOUN
bjmsr-315	62	14	©	©	PROPN
bjmsr-315	62	15	cc	cc	PROPN
bjmsr-315	62	16	-	-	PUNCT
bjmsr-315	62	17	by	by	ADP
bjmsr-315	62	18	-	-	PUNCT
bjmsr-315	62	19	nc	nc	PROPN
bjmsr-315	62	20	2019	2019	NUM
bjmsr-315	62	21	,	,	PUNCT
bjmsr-315	62	22	bjmsr	bjmsr	PROPN
bjmsr-315	62	23	www.cribfb.com/journal/index.php/bjmsr	www.cribfb.com/journal/index.php/bjmsr	X
bjmsr-315	62	24	bangladesh	bangladesh	PROPN
bjmsr-315	62	25	journal	journal	PROPN
bjmsr-315	62	26	of	of	ADP
bjmsr-315	62	27	multidisciplinary	multidisciplinary	ADJ
bjmsr-315	62	28	scientific	scientific	ADJ
bjmsr-315	62	29	research	research	NOUN
bjmsr-315	62	30	vol	vol	NOUN
bjmsr-315	62	31	.	.	PROPN
bjmsr-315	63	1	1	1	NUM
bjmsr-315	63	2	,	,	PUNCT
bjmsr-315	63	3	no	no	INTJ
bjmsr-315	63	4	.	.	NOUN
bjmsr-315	63	5	1	1	NUM
bjmsr-315	63	6	;	;	PUNCT
bjmsr-315	63	7	2019	2019	NUM
bjmsr-315	63	8	52	52	NUM
bjmsr-315	63	9	step	step	NOUN
bjmsr-315	63	10	1	1	NUM
bjmsr-315	63	11	)	)	PUNCT
bjmsr-315	63	12	select	select	ADJ
bjmsr-315	63	13	m-1	m-1	PROPN
bjmsr-315	63	14	equations	equation	NOUN
bjmsr-315	63	15	arbitrarily	arbitrarily	ADV
bjmsr-315	63	16	.	.	PUNCT
bjmsr-315	64	1	step	step	NOUN
bjmsr-315	64	2	2	2	NUM
bjmsr-315	64	3	)	)	PUNCT
bjmsr-315	64	4	solve	solve	VERB
bjmsr-315	64	5	these	these	DET
bjmsr-315	64	6	equations	equation	NOUN
bjmsr-315	64	7	for	for	ADP
bjmsr-315	64	8	m-1	m-1	PROPN
bjmsr-315	64	9	parameters	parameter	NOUN
bjmsr-315	64	10	.	.	PUNCT
bjmsr-315	65	1	step	step	NOUN
bjmsr-315	65	2	3	3	NUM
bjmsr-315	65	3	)	)	PUNCT
bjmsr-315	65	4	estimate	estimate	NOUN
bjmsr-315	65	5	the	the	DET
bjmsr-315	65	6	remaining	remain	VERB
bjmsr-315	65	7	mth	mth	NOUN
bjmsr-315	65	8	parameter	parameter	NOUN
bjmsr-315	65	9	by	by	ADP
bjmsr-315	65	10	laplace	laplace	NOUN
bjmsr-315	65	11	's	's	PART
bjmsr-315	65	12	method	method	NOUN
bjmsr-315	65	13	.	.	PUNCT
bjmsr-315	66	1	step	step	NOUN
bjmsr-315	66	2	4	4	NUM
bjmsr-315	66	3	)	)	PUNCT
bjmsr-315	66	4	recognize	recognize	VERB
bjmsr-315	66	5	the	the	DET
bjmsr-315	66	6	resulting	result	VERB
bjmsr-315	66	7	equation	equation	NOUN
bjmsr-315	66	8	in	in	ADP
bjmsr-315	66	9	step	step	NOUN
bjmsr-315	66	10	3	3	NUM
bjmsr-315	66	11	and	and	CCONJ
bjmsr-315	66	12	add	add	VERB
bjmsr-315	66	13	it	it	PRON
bjmsr-315	66	14	to	to	ADP
bjmsr-315	66	15	the	the	DET
bjmsr-315	66	16	m-1	m-1	PROPN
bjmsr-315	66	17	equations	equation	NOUN
bjmsr-315	66	18	in	in	ADP
bjmsr-315	66	19	step	step	NOUN
bjmsr-315	66	20	1	1	NUM
bjmsr-315	66	21	.	.	PUNCT
bjmsr-315	66	22	step	step	NOUN
bjmsr-315	66	23	5	5	NUM
bjmsr-315	66	24	)	)	PUNCT
bjmsr-315	66	25	if	if	SCONJ
bjmsr-315	66	26	the	the	DET
bjmsr-315	66	27	set	set	NOUN
bjmsr-315	66	28	of	of	ADP
bjmsr-315	66	29	m	m	PROPN
bjmsr-315	66	30	equations	equation	NOUN
bjmsr-315	66	31	has	have	AUX
bjmsr-315	66	32	recurred	recur	VERB
bjmsr-315	66	33	m	m	PROPN
bjmsr-315	66	34	times	time	NOUN
bjmsr-315	66	35	then	then	ADV
bjmsr-315	66	36	stop	stop	VERB
bjmsr-315	66	37	;	;	PUNCT
bjmsr-315	66	38	otherwise	otherwise	ADV
bjmsr-315	66	39	discard	discard	VERB
bjmsr-315	66	40	the	the	DET
bjmsr-315	66	41	oldest	old	ADJ
bjmsr-315	66	42	equation	equation	NOUN
bjmsr-315	66	43	and	and	CCONJ
bjmsr-315	66	44	go	go	VERB
bjmsr-315	66	45	to	to	PART
bjmsr-315	66	46	step	step	VERB
bjmsr-315	66	47	2	2	NUM
bjmsr-315	66	48	.	.	PUNCT
bjmsr-315	67	1	rhodes	rhode	NOUN
bjmsr-315	67	2	(	(	PUNCT
bjmsr-315	67	3	1930	1930	NUM
bjmsr-315	67	4	)	)	PUNCT
bjmsr-315	67	5	explained	explain	VERB
bjmsr-315	67	6	his	his	PRON
bjmsr-315	67	7	algorithm	algorithm	NOUN
bjmsr-315	67	8	by	by	ADP
bjmsr-315	67	9	an	an	DET
bjmsr-315	67	10	example	example	NOUN
bjmsr-315	67	11	and	and	CCONJ
bjmsr-315	67	12	did	do	AUX
bjmsr-315	67	13	not	not	PART
bjmsr-315	67	14	give	give	VERB
bjmsr-315	67	15	any	any	DET
bjmsr-315	67	16	proof	proof	NOUN
bjmsr-315	67	17	for	for	ADP
bjmsr-315	67	18	convergence	convergence	NOUN
bjmsr-315	67	19	.	.	PUNCT
bjmsr-315	68	1	bruen	bruen	NOUN
bjmsr-315	68	2	(	(	PUNCT
bjmsr-315	68	3	1938	1938	NUM
bjmsr-315	68	4	)	)	PUNCT
bjmsr-315	68	5	reviews	review	VERB
bjmsr-315	68	6	the	the	DET
bjmsr-315	68	7	l	l	NOUN
bjmsr-315	68	8	1	1	NUM
bjmsr-315	68	9	norm	norm	NOUN
bjmsr-315	68	10	regression	regression	NOUN
bjmsr-315	68	11	methods	method	NOUN
bjmsr-315	68	12	presented	present	VERB
bjmsr-315	68	13	by	by	ADP
bjmsr-315	68	14	earlier	early	ADJ
bjmsr-315	68	15	authors	author	NOUN
bjmsr-315	68	16	.	.	PUNCT
bjmsr-315	69	1	he	he	PRON
bjmsr-315	69	2	also	also	ADV
bjmsr-315	69	3	compares	compare	VERB
bjmsr-315	69	4	l1	l1	PROPN
bjmsr-315	69	5	,	,	PUNCT
bjmsr-315	69	6	l2	l2	NOUN
bjmsr-315	69	7	,	,	PUNCT
bjmsr-315	69	8	and	and	CCONJ
bjmsr-315	69	9	l∞	l∞	NOUN
bjmsr-315	69	10	norms	norm	NOUN
bjmsr-315	69	11	regressions	regression	NOUN
bjmsr-315	69	12	.	.	PUNCT
bjmsr-315	70	1	singleton	singleton	PROPN
bjmsr-315	70	2	(	(	PUNCT
bjmsr-315	70	3	1940	1940	NUM
bjmsr-315	70	4	)	)	PUNCT
bjmsr-315	70	5	applied	apply	VERB
bjmsr-315	70	6	cauchy	cauchy	PROPN
bjmsr-315	70	7	's	's	PART
bjmsr-315	70	8	steepest	steep	ADJ
bjmsr-315	70	9	descent	descent	NOUN
bjmsr-315	70	10	method	method	NOUN
bjmsr-315	70	11	(	(	PUNCT
bjmsr-315	70	12	see	see	VERB
bjmsr-315	70	13	,	,	PUNCT
bjmsr-315	70	14	panik	panik	X
bjmsr-315	70	15	(	(	PUNCT
bjmsr-315	70	16	1976	1976	NUM
bjmsr-315	70	17	)	)	PUNCT
bjmsr-315	70	18	)	)	PUNCT
bjmsr-315	70	19	for	for	ADP
bjmsr-315	70	20	the	the	DET
bjmsr-315	70	21	general	general	ADJ
bjmsr-315	70	22	linear	linear	PROPN
bjmsr-315	70	23	l1	l1	PROPN
bjmsr-315	70	24	norm	norm	PROPN
bjmsr-315	70	25	regression	regression	NOUN
bjmsr-315	70	26	.	.	PUNCT
bjmsr-315	71	1	in	in	ADP
bjmsr-315	71	2	this	this	DET
bjmsr-315	71	3	paper	paper	NOUN
bjmsr-315	71	4	,	,	PUNCT
bjmsr-315	71	5	a	a	DET
bjmsr-315	71	6	geometrical	geometrical	ADJ
bjmsr-315	71	7	interpretation	interpretation	NOUN
bjmsr-315	71	8	of	of	ADP
bjmsr-315	71	9	gradient	gradient	NOUN
bjmsr-315	71	10	on	on	ADP
bjmsr-315	71	11	l1	l1	PROPN
bjmsr-315	71	12	norm	norm	NOUN
bjmsr-315	71	13	polyhedron	polyhedron	NOUN
bjmsr-315	71	14	and	and	CCONJ
bjmsr-315	71	15	some	some	DET
bjmsr-315	71	16	theorems	theorem	NOUN
bjmsr-315	71	17	about	about	ADP
bjmsr-315	71	18	existence	existence	NOUN
bjmsr-315	71	19	and	and	CCONJ
bjmsr-315	71	20	uniqueness	uniqueness	NOUN
bjmsr-315	71	21	of	of	ADP
bjmsr-315	71	22	solution	solution	NOUN
bjmsr-315	71	23	and	and	CCONJ
bjmsr-315	71	24	convexity	convexity	NOUN
bjmsr-315	71	25	property	property	NOUN
bjmsr-315	71	26	all	all	PRON
bjmsr-315	71	27	were	be	AUX
bjmsr-315	71	28	given	give	VERB
bjmsr-315	71	29	.	.	PUNCT
bjmsr-315	72	1	although	although	SCONJ
bjmsr-315	72	2	the	the	DET
bjmsr-315	72	3	paper	paper	NOUN
bjmsr-315	72	4	has	have	AUX
bjmsr-315	72	5	not	not	PART
bjmsr-315	72	6	been	be	AUX
bjmsr-315	72	7	clearly	clearly	ADV
bjmsr-315	72	8	written	write	VERB
bjmsr-315	72	9	,	,	PUNCT
bjmsr-315	72	10	the	the	DET
bjmsr-315	72	11	following	follow	VERB
bjmsr-315	72	12	steps	step	NOUN
bjmsr-315	72	13	summarize	summarize	VERB
bjmsr-315	72	14	his	his	PRON
bjmsr-315	72	15	algorithm	algorithm	NOUN
bjmsr-315	72	16	.	.	PUNCT
bjmsr-315	73	1	step	step	NOUN
bjmsr-315	73	2	1	1	NUM
bjmsr-315	73	3	)	)	PUNCT
bjmsr-315	73	4	select	select	VERB
bjmsr-315	73	5	a	a	DET
bjmsr-315	73	6	point	point	NOUN
bjmsr-315	73	7	ßj(0	ßj(0	NOUN
bjmsr-315	73	8	)	)	PUNCT
bjmsr-315	73	9	,	,	PUNCT
bjmsr-315	73	10	j=1,	j=1,	NOUN
bjmsr-315	73	11	...	...	PUNCT
bjmsr-315	73	12	,m	,m	PUNCT
bjmsr-315	73	13	arbitrarily	arbitrarily	ADV
bjmsr-315	73	14	.	.	PUNCT
bjmsr-315	74	1	step	step	NOUN
bjmsr-315	74	2	2	2	NUM
bjmsr-315	74	3	)	)	PUNCT
bjmsr-315	74	4	determine	determine	VERB
bjmsr-315	74	5	the	the	DET
bjmsr-315	74	6	gradient	gradient	NOUN
bjmsr-315	74	7	,	,	PUNCT
bjmsr-315	74	8	n	n	PRON
bjmsr-315	74	9	gj	gj	NOUN
bjmsr-315	74	10	(	(	PUNCT
bjmsr-315	74	11	0)=-σ	0)=-σ	NOUN
bjmsr-315	74	12	sgn(ui^	sgn(ui^	X
bjmsr-315	74	13	│	│	X
bjmsr-315	74	14	ßj	ßj	PROPN
bjmsr-315	74	15	(	(	PUNCT
bjmsr-315	74	16	0),j=1,	0),j=1,	NOUN
bjmsr-315	74	17	...	...	PUNCT
bjmsr-315	74	18	,m)xij	,m)xij	X
bjmsr-315	74	19	.	.	PUNCT
bjmsr-315	75	1	i=1	i=1	PROPN
bjmsr-315	75	2	m	m	VERB
bjmsr-315	75	3	m	m	VERB
bjmsr-315	75	4	step	step	NOUN
bjmsr-315	75	5	3	3	NUM
bjmsr-315	75	6	)	)	PUNCT
bjmsr-315	75	7	compute	compute	NOUN
bjmsr-315	75	8	,	,	PUNCT
bjmsr-315	75	9	wi	wi	PROPN
bjmsr-315	75	10	(	(	PUNCT
bjmsr-315	75	11	0)=	0)=	PROPN
bjmsr-315	75	12	σ	σ	PROPN
bjmsr-315	75	13	xijgj	xijgj	X
bjmsr-315	75	14	(	(	PUNCT
bjmsr-315	75	15	0	0	NUM
bjmsr-315	75	16	)	)	PUNCT
bjmsr-315	75	17	,	,	PUNCT
bjmsr-315	75	18	zi	zi	PROPN
bjmsr-315	75	19	(	(	PUNCT
bjmsr-315	75	20	0)=yiσ	0)=yiσ	NUM
bjmsr-315	75	21	xijßj	xijßj	PROPN
bjmsr-315	75	22	(	(	PUNCT
bjmsr-315	75	23	0	0	NUM
bjmsr-315	75	24	)	)	PUNCT
bjmsr-315	75	25	.	.	PUNCT
bjmsr-315	76	1	j=1	j=1	NOUN
bjmsr-315	76	2	j=1	j=1	PROPN
bjmsr-315	76	3	n	n	CCONJ
bjmsr-315	76	4	step	step	VERB
bjmsr-315	76	5	4	4	NUM
bjmsr-315	76	6	)	)	PUNCT
bjmsr-315	76	7	determine	determine	VERB
bjmsr-315	76	8	the	the	DET
bjmsr-315	76	9	value	value	NOUN
bjmsr-315	76	10	of	of	ADP
bjmsr-315	76	11	t(0	t(0	PROPN
bjmsr-315	76	12	)	)	PUNCT
bjmsr-315	76	13	as	as	ADP
bjmsr-315	76	14	weighted	weight	VERB
bjmsr-315	76	15	median	median	NOUN
bjmsr-315	76	16	of	of	ADP
bjmsr-315	76	17	σ	σ	PROPN
bjmsr-315	76	18	│	│	ADJ
bjmsr-315	76	19	wit	wit	NOUN
bjmsr-315	76	20	-	-	PUNCT
bjmsr-315	76	21	zi	zi	NOUN
bjmsr-315	76	22	│	│	PUNCT
bjmsr-315	76	23	by	by	ADP
bjmsr-315	76	24	laplace	laplace	NOUN
bjmsr-315	76	25	's	's	PART
bjmsr-315	76	26	method	method	NOUN
bjmsr-315	76	27	.	.	PUNCT
bjmsr-315	77	1	i=1	i=1	PROPN
bjmsr-315	77	2	the	the	DET
bjmsr-315	77	3	t	t	PROPN
bjmsr-315	77	4	value	value	NOUN
bjmsr-315	77	5	is	be	AUX
bjmsr-315	77	6	the	the	DET
bjmsr-315	77	7	length	length	NOUN
bjmsr-315	77	8	of	of	ADP
bjmsr-315	77	9	movement	movement	NOUN
bjmsr-315	77	10	along	along	ADP
bjmsr-315	77	11	the	the	DET
bjmsr-315	77	12	direction	direction	NOUN
bjmsr-315	77	13	of	of	ADP
bjmsr-315	77	14	the	the	DET
bjmsr-315	77	15	gradient	gradient	NOUN
bjmsr-315	77	16	.	.	PUNCT
bjmsr-315	78	1	step	step	NOUN
bjmsr-315	78	2	5	5	NUM
bjmsr-315	78	3	)	)	PUNCT
bjmsr-315	78	4	compute	compute	NOUN
bjmsr-315	78	5	ßj	ßj	INTJ
bjmsr-315	78	6	(	(	PUNCT
bjmsr-315	78	7	1)=ßj	1)=ßj	NUM
bjmsr-315	78	8	(	(	PUNCT
bjmsr-315	78	9	0)+gj	0)+gj	PROPN
bjmsr-315	78	10	(	(	PUNCT
bjmsr-315	78	11	0)t(0	0)t(0	PROPN
bjmsr-315	78	12	)	)	PUNCT
bjmsr-315	78	13	.	.	PUNCT
bjmsr-315	79	1	step	step	NOUN
bjmsr-315	79	2	6	6	NUM
bjmsr-315	79	3	)	)	PUNCT
bjmsr-315	79	4	test	test	NOUN
bjmsr-315	79	5	the	the	DET
bjmsr-315	79	6	optimality	optimality	NOUN
bjmsr-315	79	7	condition	condition	NOUN
bjmsr-315	79	8	.	.	PUNCT
bjmsr-315	80	1	singleton	singleton	PROPN
bjmsr-315	80	2	gives	give	VERB
bjmsr-315	80	3	a	a	DET
bjmsr-315	80	4	condition	condition	NOUN
bjmsr-315	80	5	to	to	PART
bjmsr-315	80	6	stop	stop	VERB
bjmsr-315	80	7	,	,	PUNCT
bjmsr-315	80	8	but	but	CCONJ
bjmsr-315	80	9	it	it	PRON
bjmsr-315	80	10	is	be	AUX
bjmsr-315	80	11	not	not	PART
bjmsr-315	80	12	quite	quite	ADV
bjmsr-315	80	13	clear	clear	ADJ
bjmsr-315	80	14	.	.	PUNCT
bjmsr-315	81	1	in	in	ADP
bjmsr-315	81	2	this	this	DET
bjmsr-315	81	3	step	step	NOUN
bjmsr-315	81	4	,	,	PUNCT
bjmsr-315	81	5	any	any	DET
bjmsr-315	81	6	other	other	ADJ
bjmsr-315	81	7	criterion	criterion	NOUN
bjmsr-315	81	8	relevant	relevant	ADJ
bjmsr-315	81	9	to	to	ADP
bjmsr-315	81	10	l1	l1	PROPN
bjmsr-315	81	11	norm	norm	NOUN
bjmsr-315	81	12	function	function	NOUN
bjmsr-315	81	13	may	may	AUX
bjmsr-315	81	14	be	be	AUX
bjmsr-315	81	15	displaced	displace	VERB
bjmsr-315	81	16	.	.	PUNCT
bjmsr-315	82	1	step	step	NOUN
bjmsr-315	82	2	7	7	NUM
bjmsr-315	82	3	)	)	PUNCT
bjmsr-315	82	4	this	this	DET
bjmsr-315	82	5	step	step	NOUN
bjmsr-315	82	6	is	be	AUX
bjmsr-315	82	7	not	not	PART
bjmsr-315	82	8	well	well	ADV
bjmsr-315	82	9	defined	define	VERB
bjmsr-315	82	10	by	by	ADP
bjmsr-315	82	11	singleton	singleton	NOUN
bjmsr-315	82	12	.	.	PUNCT
bjmsr-315	83	1	in	in	ADP
bjmsr-315	83	2	this	this	DET
bjmsr-315	83	3	phase	phase	NOUN
bjmsr-315	83	4	,	,	PUNCT
bjmsr-315	83	5	he	he	PRON
bjmsr-315	83	6	tries	try	VERB
bjmsr-315	83	7	to	to	PART
bjmsr-315	83	8	choose	choose	VERB
bjmsr-315	83	9	the	the	DET
bjmsr-315	83	10	best	good	ADJ
bjmsr-315	83	11	gradient	gradient	NOUN
bjmsr-315	83	12	among	among	ADP
bjmsr-315	83	13	usable	usable	ADJ
bjmsr-315	83	14	gradients	gradient	NOUN
bjmsr-315	83	15	.	.	PUNCT
bjmsr-315	84	1	without	without	ADP
bjmsr-315	84	2	this	this	DET
bjmsr-315	84	3	step	step	NOUN
bjmsr-315	84	4	,	,	PUNCT
bjmsr-315	84	5	algorithm	algorithm	NOUN
bjmsr-315	84	6	is	be	AUX
bjmsr-315	84	7	still	still	ADV
bjmsr-315	84	8	operational	operational	ADJ
bjmsr-315	84	9	,	,	PUNCT
bjmsr-315	84	10	because	because	SCONJ
bjmsr-315	84	11	,	,	PUNCT
bjmsr-315	84	12	the	the	DET
bjmsr-315	84	13	steps	step	NOUN
bjmsr-315	84	14	are	be	AUX
bjmsr-315	84	15	all	all	PRON
bjmsr-315	84	16	standards	standard	NOUN
bjmsr-315	84	17	of	of	ADP
bjmsr-315	84	18	cauchy	cauchy	ADJ
bjmsr-315	84	19	steepest	steep	ADJ
bjmsr-315	84	20	descent	descent	NOUN
bjmsr-315	84	21	method	method	NOUN
bjmsr-315	84	22	and	and	CCONJ
bjmsr-315	84	23	instead	instead	ADV
bjmsr-315	84	24	of	of	ADP
bjmsr-315	84	25	choosing	choose	VERB
bjmsr-315	84	26	the	the	DET
bjmsr-315	84	27	best	good	ADJ
bjmsr-315	84	28	gradient	gradient	NOUN
bjmsr-315	84	29	,	,	PUNCT
bjmsr-315	84	30	we	we	PRON
bjmsr-315	84	31	can	can	AUX
bjmsr-315	84	32	proceed	proceed	VERB
bjmsr-315	84	33	by	by	ADP
bjmsr-315	84	34	going	go	VERB
bjmsr-315	84	35	to	to	PART
bjmsr-315	84	36	step	step	VERB
bjmsr-315	84	37	2	2	NUM
bjmsr-315	84	38	.	.	PUNCT
bjmsr-315	84	39	bejar	bejar	NOUN
bjmsr-315	84	40	(	(	PUNCT
bjmsr-315	84	41	1956,57	1956,57	NOUN
bjmsr-315	84	42	)	)	PUNCT
bjmsr-315	84	43	focuses	focus	VERB
bjmsr-315	84	44	on	on	ADP
bjmsr-315	84	45	consideration	consideration	NOUN
bjmsr-315	84	46	of	of	ADP
bjmsr-315	84	47	residuals	residual	NOUN
bjmsr-315	84	48	rather	rather	ADV
bjmsr-315	84	49	than	than	ADP
bjmsr-315	84	50	on	on	ADP
bjmsr-315	84	51	the	the	DET
bjmsr-315	84	52	vector	vector	NOUN
bjmsr-315	84	53	of	of	ADP
bjmsr-315	84	54	parameters	parameter	NOUN
bjmsr-315	84	55	.	.	PUNCT
bjmsr-315	85	1	he	he	PRON
bjmsr-315	85	2	puts	put	VERB
bjmsr-315	85	3	forth	forth	ADP
bjmsr-315	85	4	a	a	DET
bjmsr-315	85	5	procedure	procedure	NOUN
bjmsr-315	85	6	with	with	ADP
bjmsr-315	85	7	the	the	DET
bjmsr-315	85	8	essence	essence	NOUN
bjmsr-315	85	9	of	of	ADP
bjmsr-315	85	10	rhodes	rhode	NOUN
bjmsr-315	85	11	(	(	PUNCT
bjmsr-315	85	12	1930	1930	NUM
bjmsr-315	85	13	)	)	PUNCT
bjmsr-315	85	14	.	.	PUNCT
bjmsr-315	86	1	however	however	ADV
bjmsr-315	86	2	,	,	PUNCT
bjmsr-315	86	3	he	he	PRON
bjmsr-315	86	4	is	be	AUX
bjmsr-315	86	5	concerned	concern	VERB
bjmsr-315	86	6	with	with	ADP
bjmsr-315	86	7	two	two	NUM
bjmsr-315	86	8	and	and	CCONJ
bjmsr-315	86	9	three	three	NUM
bjmsr-315	86	10	parameter	parameter	NOUN
bjmsr-315	86	11	linear	linear	PROPN
bjmsr-315	86	12	models	model	NOUN
bjmsr-315	86	13	.	.	PUNCT
bjmsr-315	87	1	karst	karst	PROPN
bjmsr-315	87	2	(	(	PUNCT
bjmsr-315	87	3	1958	1958	NUM
bjmsr-315	87	4	)	)	PUNCT
bjmsr-315	87	5	gives	give	VERB
bjmsr-315	87	6	an	an	DET
bjmsr-315	87	7	expository	expository	ADJ
bjmsr-315	87	8	paper	paper	NOUN
bjmsr-315	87	9	for	for	ADP
bjmsr-315	87	10	one	one	NUM
bjmsr-315	87	11	and	and	CCONJ
bjmsr-315	87	12	two	two	NUM
bjmsr-315	87	13	parameter	parameter	NOUN
bjmsr-315	87	14	regression	regression	NOUN
bjmsr-315	87	15	models	model	NOUN
bjmsr-315	87	16	.	.	PUNCT
bjmsr-315	88	1	in	in	ADP
bjmsr-315	88	2	his	his	PRON
bjmsr-315	88	3	paper	paper	NOUN
bjmsr-315	88	4	,	,	PUNCT
bjmsr-315	88	5	karst	karst	ADV
bjmsr-315	88	6	without	without	ADP
bjmsr-315	88	7	referring	refer	VERB
bjmsr-315	88	8	to	to	ADP
bjmsr-315	88	9	previous	previous	ADJ
bjmsr-315	88	10	literature	literature	NOUN
bjmsr-315	88	11	actually	actually	ADV
bjmsr-315	88	12	reaches	reach	VERB
bjmsr-315	88	13	to	to	ADP
bjmsr-315	88	14	the	the	DET
bjmsr-315	88	15	laplace	laplace	NOUN
bjmsr-315	88	16	proposition	proposition	NOUN
bjmsr-315	88	17	to	to	PART
bjmsr-315	88	18	solve	solve	VERB
bjmsr-315	88	19	the	the	DET
bjmsr-315	88	20	one	one	NUM
bjmsr-315	88	21	parameter	parameter	NOUN
bjmsr-315	88	22	restricted	restrict	VERB
bjmsr-315	88	23	linear	linear	NOUN
bjmsr-315	88	24	model	model	NOUN
bjmsr-315	88	25	,	,	PUNCT
bjmsr-315	88	26	and	and	CCONJ
bjmsr-315	88	27	for	for	ADP
bjmsr-315	88	28	the	the	DET
bjmsr-315	88	29	two	two	NUM
bjmsr-315	88	30	-	-	PUNCT
bjmsr-315	88	31	parameter	parameter	NOUN
bjmsr-315	88	32	model	model	NOUN
bjmsr-315	88	33	,	,	PUNCT
bjmsr-315	88	34	he	he	PRON
bjmsr-315	88	35	proposed	propose	VERB
bjmsr-315	88	36	an	an	DET
bjmsr-315	88	37	algorithm	algorithm	NOUN
bjmsr-315	88	38	similar	similar	ADJ
bjmsr-315	88	39	to	to	ADP
bjmsr-315	88	40	that	that	PRON
bjmsr-315	88	41	of	of	ADP
bjmsr-315	88	42	rhodes	rhode	NOUN
bjmsr-315	88	43	(	(	PUNCT
bjmsr-315	88	44	1930	1930	NUM
bjmsr-315	88	45	)	)	PUNCT
bjmsr-315	88	46	.	.	PUNCT
bjmsr-315	89	1	his	his	PRON
bjmsr-315	89	2	viewpoint	viewpoint	NOUN
bjmsr-315	89	3	is	be	AUX
bjmsr-315	89	4	both	both	CCONJ
bjmsr-315	89	5	geometrical	geometrical	ADJ
bjmsr-315	89	6	and	and	CCONJ
bjmsr-315	89	7	algebraic	algebraic	ADJ
bjmsr-315	89	8	,	,	PUNCT
bjmsr-315	89	9	and	and	CCONJ
bjmsr-315	89	10	no	no	DET
bjmsr-315	89	11	proof	proof	NOUN
bjmsr-315	89	12	of	of	ADP
bjmsr-315	89	13	convergence	convergence	NOUN
bjmsr-315	89	14	for	for	ADP
bjmsr-315	89	15	his	his	PRON
bjmsr-315	89	16	iterative	iterative	NOUN
bjmsr-315	89	17	method	method	NOUN
bjmsr-315	89	18	is	be	AUX
bjmsr-315	89	19	offered	offer	VERB
bjmsr-315	89	20	.	.	PUNCT
bjmsr-315	90	1	sadovski	sadovski	PROPN
bjmsr-315	90	2	(	(	PUNCT
bjmsr-315	90	3	1974	1974	NUM
bjmsr-315	90	4	)	)	PUNCT
bjmsr-315	90	5	uses	use	VERB
bjmsr-315	90	6	a	a	DET
bjmsr-315	90	7	simple	simple	ADJ
bjmsr-315	90	8	"	"	PUNCT
bjmsr-315	90	9	bubble	bubble	NOUN
bjmsr-315	90	10	sort	sort	NOUN
bjmsr-315	90	11	"	"	PUNCT
bjmsr-315	90	12	procedure	procedure	NOUN
bjmsr-315	90	13	and	and	CCONJ
bjmsr-315	90	14	implements	implement	VERB
bjmsr-315	90	15	karst	karst	ADJ
bjmsr-315	90	16	algorithm	algorithm	PROPN
bjmsr-315	90	17	in	in	ADP
bjmsr-315	90	18	fortran	fortran	NOUN
bjmsr-315	90	19	.	.	PUNCT
bjmsr-315	90	20	sposito	sposito	PROPN
bjmsr-315	90	21	(	(	PUNCT
bjmsr-315	90	22	1976	1976	NUM
bjmsr-315	90	23	)	)	PUNCT
bjmsr-315	90	24	pointed	point	VERB
bjmsr-315	90	25	out	out	ADP
bjmsr-315	90	26	that	that	SCONJ
bjmsr-315	90	27	the	the	DET
bjmsr-315	90	28	sadovski	sadovski	PROPN
bjmsr-315	90	29	's	's	PART
bjmsr-315	90	30	program	program	NOUN
bjmsr-315	90	31	may	may	AUX
bjmsr-315	90	32	not	not	PART
bjmsr-315	90	33	converge	converge	VERB
bjmsr-315	90	34	in	in	ADP
bjmsr-315	90	35	general	general	ADJ
bjmsr-315	90	36	.	.	PUNCT
bjmsr-315	91	1	sposito	sposito	PROPN
bjmsr-315	91	2	and	and	CCONJ
bjmsr-315	91	3	smith	smith	PROPN
bjmsr-315	91	4	(	(	PUNCT
bjmsr-315	91	5	1976	1976	NUM
bjmsr-315	91	6	)	)	PUNCT
bjmsr-315	91	7	offered	offer	VERB
bjmsr-315	91	8	another	another	DET
bjmsr-315	91	9	algorithm	algorithm	NOUN
bjmsr-315	91	10	to	to	PART
bjmsr-315	91	11	remove	remove	VERB
bjmsr-315	91	12	this	this	DET
bjmsr-315	91	13	problem	problem	NOUN
bjmsr-315	91	14	.	.	PUNCT
bjmsr-315	92	1	farebrother	farebrother	ADV
bjmsr-315	92	2	(	(	PUNCT
bjmsr-315	92	3	1987c	1987c	NUM
bjmsr-315	92	4	)	)	PUNCT
bjmsr-315	92	5	recodes	recode	VERB
bjmsr-315	92	6	sadovski	sadovski	PROPN
bjmsr-315	92	7	's	's	PART
bjmsr-315	92	8	implementation	implementation	NOUN
bjmsr-315	92	9	in	in	ADP
bjmsr-315	92	10	pascal	pascal	ADJ
bjmsr-315	92	11	language	language	NOUN
bjmsr-315	92	12	with	with	ADP
bjmsr-315	92	13	some	some	DET
bjmsr-315	92	14	improvement	improvement	NOUN
bjmsr-315	92	15	such	such	ADJ
bjmsr-315	92	16	as	as	ADP
bjmsr-315	92	17	applying	apply	VERB
bjmsr-315	92	18	"	"	PUNCT
bjmsr-315	92	19	straight	straight	ADJ
bjmsr-315	92	20	insert	insert	ADJ
bjmsr-315	92	21	sort	sort	NOUN
bjmsr-315	92	22	"	"	PUNCT
bjmsr-315	92	23	.	.	PUNCT
bjmsr-315	93	1	usow	usow	PROPN
bjmsr-315	93	2	(	(	PUNCT
bjmsr-315	93	3	1967b	1967b	NUM
bjmsr-315	93	4	)	)	PUNCT
bjmsr-315	93	5	presents	present	VERB
bjmsr-315	93	6	an	an	DET
bjmsr-315	93	7	algorithm	algorithm	NOUN
bjmsr-315	93	8	for	for	ADP
bjmsr-315	93	9	l1	l1	PROPN
bjmsr-315	93	10	norm	norm	NOUN
bjmsr-315	93	11	approximation	approximation	NOUN
bjmsr-315	93	12	for	for	ADP
bjmsr-315	93	13	discrete	discrete	ADJ
bjmsr-315	93	14	data	datum	NOUN
bjmsr-315	93	15	and	and	CCONJ
bjmsr-315	93	16	proves	prove	VERB
bjmsr-315	93	17	that	that	SCONJ
bjmsr-315	93	18	it	it	PRON
bjmsr-315	93	19	converges	converge	VERB
bjmsr-315	93	20	in	in	ADP
bjmsr-315	93	21	a	a	DET
bjmsr-315	93	22	finite	finite	ADJ
bjmsr-315	93	23	number	number	NOUN
bjmsr-315	93	24	of	of	ADP
bjmsr-315	93	25	steps	step	NOUN
bjmsr-315	93	26	.	.	PUNCT
bjmsr-315	94	1	a	a	DET
bjmsr-315	94	2	similar	similar	ADJ
bjmsr-315	94	3	algorithm	algorithm	NOUN
bjmsr-315	94	4	on	on	ADP
bjmsr-315	94	5	l1	l1	PROPN
bjmsr-315	94	6	norm	norm	NOUN
bjmsr-315	94	7	approximation	approximation	NOUN
bjmsr-315	94	8	for	for	ADP
bjmsr-315	94	9	continuous	continuous	ADJ
bjmsr-315	94	10	data	datum	NOUN
bjmsr-315	94	11	is	be	AUX
bjmsr-315	94	12	given	give	VERB
bjmsr-315	94	13	by	by	ADP
bjmsr-315	94	14	usow	usow	NOUN
bjmsr-315	94	15	(	(	PUNCT
bjmsr-315	94	16	1967a	1967a	NUM
bjmsr-315	94	17	)	)	PUNCT
bjmsr-315	94	18	.	.	PUNCT
bjmsr-315	95	1	given	give	VERB
bjmsr-315	95	2	the	the	DET
bjmsr-315	95	3	function	function	NOUN
bjmsr-315	95	4	f(x	f(x	PROPN
bjmsr-315	95	5	)	)	PUNCT
bjmsr-315	95	6	defined	define	VERB
bjmsr-315	95	7	on	on	ADP
bjmsr-315	95	8	a	a	DET
bjmsr-315	95	9	finite	finite	NOUN
bjmsr-315	95	10	subset	subset	VERB
bjmsr-315	96	1	x={x1,	x={x1,	PROPN
bjmsr-315	96	2	...	...	PUNCT
bjmsr-315	96	3	,xn	,xn	PUNCT
bjmsr-315	96	4	}	}	PUNCT
bjmsr-315	96	5	of	of	ADP
bjmsr-315	96	6	an	an	DET
bjmsr-315	96	7	interval	interval	NOUN
bjmsr-315	96	8	on	on	ADP
bjmsr-315	96	9	the	the	DET
bjmsr-315	96	10	real	real	ADJ
bjmsr-315	96	11	line	line	NOUN
bjmsr-315	96	12	and	and	CCONJ
bjmsr-315	96	13	also	also	ADV
bjmsr-315	96	14	linearly	linearly	ADV
bjmsr-315	96	15	independent	independent	ADJ
bjmsr-315	96	16	continuous	continuous	ADJ
bjmsr-315	96	17	functions	function	NOUN
bjmsr-315	96	18	φ1(x),	φ1(x),	NOUN
bjmsr-315	96	19	...	...	PUNCT
bjmsr-315	96	20	,φm(x	,φm(x	PUNCT
bjmsr-315	96	21	)	)	PUNCT
bjmsr-315	96	22	where	where	SCONJ
bjmsr-315	96	23	m	m	AUX
bjmsr-315	96	24	<	<	X
bjmsr-315	96	25	n	n	CCONJ
bjmsr-315	96	26	;	;	PUNCT
bjmsr-315	96	27	consider	consider	VERB
bjmsr-315	96	28	the	the	DET
bjmsr-315	96	29	"	"	PUNCT
bjmsr-315	96	30	polynomial	polynomial	ADJ
bjmsr-315	96	31	"	"	PUNCT
bjmsr-315	96	32	m	m	VERB
bjmsr-315	96	33	l(ß	l(ß	PROPN
bjmsr-315	96	34	,	,	PUNCT
bjmsr-315	96	35	x)=σßjφj(x	x)=σßjφj(x	PROPN
bjmsr-315	96	36	)	)	PUNCT
bjmsr-315	96	37	.	.	PUNCT
bjmsr-315	97	1	in	in	ADP
bjmsr-315	97	2	usow	usow	NOUN
bjmsr-315	97	3	(	(	PUNCT
bjmsr-315	97	4	1967b	1967b	NUM
bjmsr-315	97	5	)	)	PUNCT
bjmsr-315	97	6	the	the	DET
bjmsr-315	97	7	following	follow	VERB
bjmsr-315	97	8	function	function	NOUN
bjmsr-315	97	9	is	be	AUX
bjmsr-315	97	10	to	to	PART
bjmsr-315	97	11	be	be	AUX
bjmsr-315	97	12	minimized	minimize	VERB
bjmsr-315	97	13	:	:	PUNCT
bjmsr-315	97	14	j=1	j=1	PROPN
bjmsr-315	97	15	n	n	CCONJ
bjmsr-315	97	16	n	n	CCONJ
bjmsr-315	97	17	m	m	NOUN
bjmsr-315	97	18	min	min	NOUN
bjmsr-315	97	19	:	:	PUNCT
bjmsr-315	97	20	s(ß)=	s(ß)=	PROPN
bjmsr-315	97	21	σ	σ	NOUN
bjmsr-315	97	22	│	│	PUNCT
bjmsr-315	97	23	l(ß	l(ß	PROPN
bjmsr-315	97	24	,	,	PUNCT
bjmsr-315	97	25	xi)-f(xi)	xi)-f(xi)	NOUN
bjmsr-315	97	26	│	│	ADJ
bjmsr-315	97	27	=	=	SYM
bjmsr-315	97	28	σ	σ	PROPN
bjmsr-315	97	29	│	│	NOUN
bjmsr-315	97	30	σ	σ	X
bjmsr-315	97	31	ßjφj(xi)-f(xi	ßjφj(xi)-f(xi	NUM
bjmsr-315	97	32	)	)	PUNCT
bjmsr-315	97	33	│	│	X
bjmsr-315	97	34	(	(	PUNCT
bjmsr-315	97	35	5	5	X
bjmsr-315	97	36	)	)	PUNCT
bjmsr-315	97	37	ß	ß	NOUN
bjmsr-315	97	38	i=1	i=1	X
bjmsr-315	97	39	i=1	i=1	X
bjmsr-315	97	40	j=1	j=1	NOUN
bjmsr-315	97	41	where	where	SCONJ
bjmsr-315	97	42	ß	ß	NOUN
bjmsr-315	97	43	is	be	AUX
bjmsr-315	97	44	a	a	DET
bjmsr-315	97	45	vector	vector	NOUN
bjmsr-315	97	46	of	of	ADP
bjmsr-315	97	47	size	size	NOUN
bjmsr-315	97	48	m.	m.	NOUN
bjmsr-315	97	49	the	the	DET
bjmsr-315	97	50	above	above	ADJ
bjmsr-315	97	51	form	form	NOUN
bjmsr-315	97	52	is	be	AUX
bjmsr-315	97	53	general	general	ADJ
bjmsr-315	97	54	;	;	PUNCT
bjmsr-315	97	55	if	if	SCONJ
bjmsr-315	97	56	we	we	PRON
bjmsr-315	97	57	let	let	VERB
bjmsr-315	97	58	f(xi)≡yi	f(xi)≡yi	PRON
bjmsr-315	97	59	and	and	CCONJ
bjmsr-315	97	60	φj(xi)≡xij	φj(xi)≡xij	PROPN
bjmsr-315	98	1	the	the	DET
bjmsr-315	98	2	function	function	NOUN
bjmsr-315	98	3	s	s	PART
bjmsr-315	98	4	becomes	become	VERB
bjmsr-315	98	5	the	the	DET
bjmsr-315	98	6	standard	standard	ADJ
bjmsr-315	98	7	linear	linear	ADJ
bjmsr-315	98	8	regression	regression	NOUN
bjmsr-315	98	9	objective	objective	ADJ
bjmsr-315	98	10	function	function	NOUN
bjmsr-315	98	11	.	.	PUNCT
bjmsr-315	99	1	now	now	ADV
bjmsr-315	99	2	let	let	VERB
bjmsr-315	99	3	the	the	DET
bjmsr-315	99	4	set	set	NOUN
bjmsr-315	99	5	k	k	PROPN
bjmsr-315	99	6	be	be	AUX
bjmsr-315	99	7	:	:	PUNCT
bjmsr-315	99	8	k={(ß	k={(ß	PROPN
bjmsr-315	99	9	,	,	PUNCT
bjmsr-315	99	10	d)	d)	NOUN
bjmsr-315	99	11	│	│	ADJ
bjmsr-315	99	12	(ß	(ß	NOUN
bjmsr-315	99	13	,	,	PUNCT
bjmsr-315	99	14	d)εem+1,s≤d	d)εem+1,s≤d	NOUN
bjmsr-315	99	15	}	}	PUNCT
bjmsr-315	99	16	(	(	PUNCT
bjmsr-315	99	17	6	6	NUM
bjmsr-315	99	18	)	)	PUNCT
bjmsr-315	99	19	then	then	ADV
bjmsr-315	99	20	k	k	PROPN
bjmsr-315	99	21	is	be	AUX
bjmsr-315	99	22	a	a	DET
bjmsr-315	99	23	convex	convex	ADJ
bjmsr-315	99	24	polytope	polytope	NOUN
bjmsr-315	99	25	,	,	PUNCT
bjmsr-315	99	26	the	the	DET
bjmsr-315	99	27	vertices	vertex	NOUN
bjmsr-315	99	28	of	of	ADP
bjmsr-315	99	29	which	which	PRON
bjmsr-315	99	30	occur	occur	VERB
bjmsr-315	99	31	only	only	ADV
bjmsr-315	99	32	when	when	SCONJ
bjmsr-315	99	33	l(ß	l(ß	PROPN
bjmsr-315	99	34	,	,	PUNCT
bjmsr-315	99	35	x)-f(x	x)-f(x	PUNCT
bjmsr-315	99	36	)	)	PUNCT
bjmsr-315	99	37	is	be	AUX
bjmsr-315	99	38	zero	zero	NUM
bjmsr-315	99	39	at	at	ADP
bjmsr-315	99	40	m	m	PROPN
bjmsr-315	99	41	or	or	CCONJ
bjmsr-315	99	42	more	more	ADJ
bjmsr-315	99	43	points	point	NOUN
bjmsr-315	99	44	of	of	ADP
bjmsr-315	99	45	x.	x.	NOUN
bjmsr-315	99	46	the	the	DET
bjmsr-315	99	47	usow	usow	NOUN
bjmsr-315	99	48	's	's	PART
bjmsr-315	99	49	algorithm	algorithm	NOUN
bjmsr-315	99	50	is	be	AUX
bjmsr-315	99	51	to	to	PART
bjmsr-315	99	52	descend	descend	VERB
bjmsr-315	99	53	on	on	ADP
bjmsr-315	99	54	k	k	PROPN
bjmsr-315	99	55	from	from	ADP
bjmsr-315	99	56	vertex	vertex	NOUN
bjmsr-315	99	57	to	to	ADP
bjmsr-315	99	58	vertex	vertex	NOUN
bjmsr-315	99	59	along	along	ADP
bjmsr-315	99	60	connecting	connect	VERB
bjmsr-315	99	61	edges	edge	NOUN
bjmsr-315	99	62	of	of	ADP
bjmsr-315	99	63	the	the	DET
bjmsr-315	99	64	polytope	polytope	NOUN
bjmsr-315	99	65	in	in	ADP
bjmsr-315	99	66	such	such	DET
bjmsr-315	99	67	a	a	DET
bjmsr-315	99	68	way	way	NOUN
bjmsr-315	99	69	that	that	PRON
bjmsr-315	99	70	certain	certain	ADJ
bjmsr-315	99	71	intermediate	intermediate	ADJ
bjmsr-315	99	72	vertices	vertex	NOUN
bjmsr-315	99	73	are	be	AUX
bjmsr-315	99	74	by	by	ADP
bjmsr-315	99	75	-	-	PUNCT
bjmsr-315	99	76	passed	pass	VERB
bjmsr-315	99	77	.	.	PUNCT
bjmsr-315	100	1	this	this	DET
bjmsr-315	100	2	descent	descent	NOUN
bjmsr-315	100	3	continues	continue	VERB
bjmsr-315	100	4	until	until	SCONJ
bjmsr-315	100	5	the	the	DET
bjmsr-315	100	6	lowest	low	ADJ
bjmsr-315	100	7	vertex	vertex	NOUN
bjmsr-315	100	8	(	(	PUNCT
bjmsr-315	100	9	ß*,d	ß*,d	NOUN
bjmsr-315	100	10	*	*	PUNCT
bjmsr-315	100	11	)	)	PUNCT
bjmsr-315	100	12	is	be	AUX
bjmsr-315	100	13	reached	reach	VERB
bjmsr-315	100	14	.	.	PUNCT
bjmsr-315	101	1	to	to	PART
bjmsr-315	101	2	clarify	clarify	VERB
bjmsr-315	101	3	the	the	DET
bjmsr-315	101	4	algorithm	algorithm	NOUN
bjmsr-315	101	5	assume	assume	VERB
bjmsr-315	101	6	that	that	SCONJ
bjmsr-315	101	7	we	we	PRON
bjmsr-315	101	8	are	be	AUX
bjmsr-315	101	9	at	at	ADP
bjmsr-315	101	10	the	the	DET
bjmsr-315	101	11	vertex	vertex	NOUN
bjmsr-315	101	12	(	(	PUNCT
bjmsr-315	101	13	ßk	ßk	PROPN
bjmsr-315	101	14	,	,	PUNCT
bjmsr-315	101	15	dk	dk	NOUN
bjmsr-315	101	16	)	)	PUNCT
bjmsr-315	101	17	on	on	ADP
bjmsr-315	101	18	k	k	PROPN
bjmsr-315	101	19	and	and	CCONJ
bjmsr-315	101	20	the	the	DET
bjmsr-315	101	21	polynomial	polynomial	ADJ
bjmsr-315	101	22	l(ßk	l(ßk	PROPN
bjmsr-315	101	23	,	,	PUNCT
bjmsr-315	101	24	x	x	X
bjmsr-315	101	25	)	)	PUNCT
bjmsr-315	101	26	interpolates	interpolate	VERB
bjmsr-315	101	27	m	m	NOUN
bjmsr-315	101	28	points	point	NOUN
bjmsr-315	101	29	of	of	ADP
bjmsr-315	101	30	x	x	PUNCT
bjmsr-315	101	31	denoted	denote	VERB
bjmsr-315	101	32	by	by	ADP
bjmsr-315	101	33	uk={u1	uk={u1	PROPN
bjmsr-315	101	34	k,	k,	PROPN
bjmsr-315	101	35	...	...	PUNCT
bjmsr-315	101	36	,um	,um	PUNCT
bjmsr-315	101	37	k	k	NOUN
bjmsr-315	101	38	}	}	PUNCT
bjmsr-315	101	39	.	.	PUNCT
bjmsr-315	102	1	thus	thus	ADV
bjmsr-315	102	2	,	,	PUNCT
bjmsr-315	102	3	m	m	VERB
bjmsr-315	102	4	copyright	copyright	NOUN
bjmsr-315	102	5	©	©	PROPN
bjmsr-315	102	6	cc	cc	NOUN
bjmsr-315	102	7	-	-	PUNCT
bjmsr-315	102	8	by	by	ADP
bjmsr-315	102	9	-	-	PUNCT
bjmsr-315	102	10	nc	nc	PROPN
bjmsr-315	102	11	2019	2019	NUM
bjmsr-315	102	12	,	,	PUNCT
bjmsr-315	102	13	bjmsr	bjmsr	PROPN
bjmsr-315	102	14	www.cribfb.com/journal/index.php/bjmsr	www.cribfb.com/journal/index.php/bjmsr	X
bjmsr-315	102	15	bangladesh	bangladesh	PROPN
bjmsr-315	102	16	journal	journal	PROPN
bjmsr-315	102	17	of	of	ADP
bjmsr-315	102	18	multidisciplinary	multidisciplinary	ADJ
bjmsr-315	102	19	scientific	scientific	ADJ
bjmsr-315	102	20	research	research	NOUN
bjmsr-315	102	21	vol	vol	NOUN
bjmsr-315	102	22	.	.	PROPN
bjmsr-315	103	1	1	1	NUM
bjmsr-315	103	2	,	,	PUNCT
bjmsr-315	103	3	no	no	INTJ
bjmsr-315	103	4	.	.	NOUN
bjmsr-315	103	5	1	1	NUM
bjmsr-315	103	6	;	;	PUNCT
bjmsr-315	103	7	2019	2019	NUM
bjmsr-315	103	8	53	53	NUM
bjmsr-315	103	9	l(fk	l(fk	NOUN
bjmsr-315	103	10	,	,	PUNCT
bjmsr-315	103	11	x)=	x)=	PROPN
bjmsr-315	103	12	σ	σ	PROPN
bjmsr-315	103	13	f(ui	f(ui	PROPN
bjmsr-315	103	14	k)πi(x	k)πi(x	NOUN
bjmsr-315	103	15	)	)	PUNCT
bjmsr-315	103	16	(	(	PUNCT
bjmsr-315	103	17	7	7	X
bjmsr-315	103	18	)	)	PUNCT
bjmsr-315	103	19	i=1	i=1	VERB
bjmsr-315	104	1	where	where	SCONJ
bjmsr-315	104	2	,	,	PUNCT
bjmsr-315	104	3	fk=(f(u1	fk=(f(u1	PUNCT
bjmsr-315	104	4	k),	k),	NOUN
bjmsr-315	104	5	...	...	PUNCT
bjmsr-315	104	6	,f(um	,f(um	PUNCT
bjmsr-315	104	7	k	k	NOUN
bjmsr-315	104	8	)	)	PUNCT
bjmsr-315	104	9	)	)	PUNCT
bjmsr-315	105	1	(	(	PUNCT
bjmsr-315	105	2	8)	8)	NUM
bjmsr-315	105	3	m	m	PROPN
bjmsr-315	105	4	πi(x)=	πi(x)=	PROPN
bjmsr-315	105	5	σ	σ	PROPN
bjmsr-315	105	6	aj	aj	PROPN
bjmsr-315	105	7	iφj(x	iφj(x	PROPN
bjmsr-315	105	8	)	)	PUNCT
bjmsr-315	105	9	i=1,	i=1,	NOUN
bjmsr-315	105	10	...	...	PUNCT
bjmsr-315	105	11	,m	,m	PUNCT
bjmsr-315	105	12	(	(	PUNCT
bjmsr-315	105	13	9	9	X
bjmsr-315	105	14	)	)	PUNCT
bjmsr-315	105	15	j=1	j=1	NOUN
bjmsr-315	106	1	the	the	DET
bjmsr-315	106	2	m	m	PROPN
bjmsr-315	106	3	coefficients	coefficient	NOUN
bjmsr-315	106	4	aji	aji	PROPN
bjmsr-315	106	5	are	be	AUX
bjmsr-315	106	6	calculated	calculate	VERB
bjmsr-315	106	7	as	as	ADP
bjmsr-315	106	8	follows	follow	NOUN
bjmsr-315	106	9	.	.	PUNCT
bjmsr-315	107	1	form	form	VERB
bjmsr-315	107	2	the	the	DET
bjmsr-315	107	3	matrix	matrix	NOUN
bjmsr-315	107	4	(	(	PUNCT
bjmsr-315	107	5	φj(ui	φj(ui	X
bjmsr-315	107	6	k	k	NOUN
bjmsr-315	107	7	)	)	PUNCT
bjmsr-315	107	8	)	)	PUNCT
bjmsr-315	107	9	for	for	ADP
bjmsr-315	107	10	i	i	PRON
bjmsr-315	107	11	,	,	PUNCT
bjmsr-315	107	12	j=1,	j=1,	NOUN
bjmsr-315	107	13	...	...	PUNCT
bjmsr-315	107	14	,m	,m	PUNCT
bjmsr-315	107	15	.	.	PUNCT
bjmsr-315	108	1	let	let	VERB
bjmsr-315	108	2	π(x	π(x	ADP
bjmsr-315	108	3	)	)	PUNCT
bjmsr-315	108	4	and	and	CCONJ
bjmsr-315	108	5	φ(x	φ(x	NOUN
bjmsr-315	108	6	)	)	PUNCT
bjmsr-315	108	7	be	be	VERB
bjmsr-315	108	8	two	two	NUM
bjmsr-315	108	9	mx1	mx1	PROPN
bjmsr-315	108	10	vectors	vector	NOUN
bjmsr-315	108	11	for	for	ADP
bjmsr-315	108	12	any	any	DET
bjmsr-315	108	13	xεx	xεx	NOUN
bjmsr-315	108	14	whose	whose	DET
bjmsr-315	108	15	elements	element	NOUN
bjmsr-315	108	16	are	be	AUX
bjmsr-315	108	17	πi(x	πi(x	ADJ
bjmsr-315	108	18	)	)	PUNCT
bjmsr-315	108	19	and	and	CCONJ
bjmsr-315	109	1	φi(x	φi(x	NUM
bjmsr-315	109	2	)	)	PUNCT
bjmsr-315	109	3	for	for	ADP
bjmsr-315	109	4	i=1,	i=1,	PROPN
bjmsr-315	109	5	...	...	PUNCT
bjmsr-315	109	6	,m	,m	PUNCT
bjmsr-315	109	7	respectively	respectively	ADV
bjmsr-315	109	8	.	.	PUNCT
bjmsr-315	110	1	hence	hence	ADV
bjmsr-315	110	2	,	,	PUNCT
bjmsr-315	110	3	we	we	PRON
bjmsr-315	110	4	can	can	AUX
bjmsr-315	110	5	derive	derive	VERB
bjmsr-315	110	6	,	,	PUNCT
bjmsr-315	110	7	π(x)=[(φj(ui	π(x)=[(φj(ui	PROPN
bjmsr-315	110	8	k))t]-1φ(x	k))t]-1φ(x	NOUN
bjmsr-315	110	9	)	)	PUNCT
bjmsr-315	110	10	(	(	PUNCT
bjmsr-315	110	11	10	10	NUM
bjmsr-315	110	12	)	)	PUNCT
bjmsr-315	110	13	aj	aj	PROPN
bjmsr-315	111	1	i	i	PRON
bjmsr-315	111	2	's	be	AUX
bjmsr-315	111	3	are	be	AUX
bjmsr-315	111	4	the	the	DET
bjmsr-315	111	5	elements	element	NOUN
bjmsr-315	111	6	of	of	ADP
bjmsr-315	111	7	the	the	DET
bjmsr-315	111	8	ith	ith	PROPN
bjmsr-315	111	9	row	row	NOUN
bjmsr-315	111	10	of	of	ADP
bjmsr-315	111	11	the	the	DET
bjmsr-315	111	12	matrix	matrix	NOUN
bjmsr-315	111	13	[	[	X
bjmsr-315	111	14	(	(	PUNCT
bjmsr-315	111	15	φj(ui	φj(ui	X
bjmsr-315	111	16	k))t]-1	k))t]-1	PROPN
bjmsr-315	111	17	,	,	PUNCT
bjmsr-315	111	18	where	where	SCONJ
bjmsr-315	111	19	the	the	DET
bjmsr-315	111	20	superscript	superscript	PROPN
bjmsr-315	111	21	t	t	PROPN
bjmsr-315	111	22	denotes	denote	VERB
bjmsr-315	111	23	transposition	transposition	NOUN
bjmsr-315	111	24	.	.	PUNCT
bjmsr-315	112	1	let	let	VERB
bjmsr-315	112	2	ei	ei	PART
bjmsr-315	112	3	be	be	AUX
bjmsr-315	112	4	a	a	DET
bjmsr-315	112	5	zero	zero	NUM
bjmsr-315	112	6	vector	vector	NOUN
bjmsr-315	112	7	of	of	ADP
bjmsr-315	112	8	size	size	NOUN
bjmsr-315	112	9	m	m	PROPN
bjmsr-315	112	10	where	where	SCONJ
bjmsr-315	112	11	its	its	PRON
bjmsr-315	112	12	ith	ith	PROPN
bjmsr-315	112	13	element	element	NOUN
bjmsr-315	112	14	is	be	AUX
bjmsr-315	112	15	equal	equal	ADJ
bjmsr-315	112	16	to	to	ADP
bjmsr-315	112	17	one	one	NUM
bjmsr-315	112	18	.	.	PUNCT
bjmsr-315	113	1	then	then	ADV
bjmsr-315	113	2	if	if	SCONJ
bjmsr-315	113	3	for	for	ADP
bjmsr-315	113	4	some	some	DET
bjmsr-315	113	5	δ	δ	PROPN
bjmsr-315	113	6	,	,	PUNCT
bjmsr-315	113	7	s(fk	s(fk	PROPN
bjmsr-315	113	8	-	-	PROPN
bjmsr-315	113	9	δei)<s(fk	δei)<s(fk	NOUN
bjmsr-315	113	10	)	)	PUNCT
bjmsr-315	113	11	,	,	PUNCT
bjmsr-315	113	12	there	there	PRON
bjmsr-315	113	13	is	be	VERB
bjmsr-315	113	14	a	a	DET
bjmsr-315	113	15	tj	tj	NOUN
bjmsr-315	113	16	such	such	ADJ
bjmsr-315	113	17	that	that	SCONJ
bjmsr-315	113	18	tjδ>0	tjδ>0	PROPN
bjmsr-315	113	19	and	and	CCONJ
bjmsr-315	113	20	s(fk	s(fk	PROPN
bjmsr-315	113	21	-	-	PROPN
bjmsr-315	113	22	tjei)<s(fk	tjei)<s(fk	NUM
bjmsr-315	113	23	)	)	PUNCT
bjmsr-315	113	24	.	.	PUNCT
bjmsr-315	114	1	also	also	ADV
bjmsr-315	114	2	,	,	PUNCT
bjmsr-315	114	3	s(fk	s(fk	PROPN
bjmsr-315	114	4	tjei	tjei	NOUN
bjmsr-315	114	5	)	)	PUNCT
bjmsr-315	114	6	=	=	SYM
bjmsr-315	114	7	min	min	PROPN
bjmsr-315	114	8	{	{	PUNCT
bjmsr-315	114	9	s(fk	s(fk	PROPN
bjmsr-315	114	10	tei	tei	NOUN
bjmsr-315	114	11	)	)	PUNCT
bjmsr-315	114	12	}	}	PUNCT
bjmsr-315	114	13	(	(	PUNCT
bjmsr-315	114	14	11	11	NUM
bjmsr-315	114	15	)	)	PUNCT
bjmsr-315	114	16	t	t	NOUN
bjmsr-315	114	17	and	and	CCONJ
bjmsr-315	114	18	(	(	PUNCT
bjmsr-315	114	19	(	(	PUNCT
bjmsr-315	114	20	fk	fk	INTJ
bjmsr-315	114	21	-	-	PUNCT
bjmsr-315	114	22	tjei),s(fk	tjei),s(fk	NOUN
bjmsr-315	114	23	-	-	PUNCT
bjmsr-315	114	24	tjei	tjei	NOUN
bjmsr-315	114	25	)	)	PUNCT
bjmsr-315	114	26	)	)	PUNCT
bjmsr-315	114	27	is	be	AUX
bjmsr-315	114	28	a	a	DET
bjmsr-315	114	29	vertex	vertex	NOUN
bjmsr-315	114	30	.	.	PUNCT
bjmsr-315	115	1	on	on	ADP
bjmsr-315	115	2	the	the	DET
bjmsr-315	115	3	other	other	ADJ
bjmsr-315	115	4	hand	hand	NOUN
bjmsr-315	115	5	,	,	PUNCT
bjmsr-315	115	6	a	a	DET
bjmsr-315	115	7	point	point	NOUN
bjmsr-315	115	8	ui	ui	PROPN
bjmsr-315	115	9	kεuk	kεuk	PROPN
bjmsr-315	115	10	may	may	AUX
bjmsr-315	115	11	be	be	AUX
bjmsr-315	115	12	replaced	replace	VERB
bjmsr-315	115	13	by	by	ADP
bjmsr-315	115	14	a	a	DET
bjmsr-315	115	15	point	point	NOUN
bjmsr-315	115	16	ui	ui	PROPN
bjmsr-315	115	17	k+1ε{x	k+1ε{x	PROPN
bjmsr-315	115	18	-	-	PUNCT
bjmsr-315	115	19	uk	uk	NOUN
bjmsr-315	115	20	}	}	PUNCT
bjmsr-315	115	21	such	such	ADJ
bjmsr-315	115	22	that	that	SCONJ
bjmsr-315	115	23	the	the	DET
bjmsr-315	115	24	polynomial	polynomial	PROPN
bjmsr-315	115	25	l(ßki	l(ßki	PROPN
bjmsr-315	115	26	,	,	PUNCT
bjmsr-315	115	27	x	x	X
bjmsr-315	115	28	)	)	PUNCT
bjmsr-315	115	29	interpolating	interpolate	VERB
bjmsr-315	115	30	ui	ui	PROPN
bjmsr-315	115	31	k={u1	k={u1	PROPN
bjmsr-315	115	32	k,	k,	PROPN
bjmsr-315	115	33	...	...	PUNCT
bjmsr-315	115	34	,ui	,ui	PUNCT
bjmsr-315	115	35	k+1,	k+1,	NOUN
bjmsr-315	115	36	...	...	PUNCT
bjmsr-315	115	37	,um	,um	PUNCT
bjmsr-315	115	38	k	k	NOUN
bjmsr-315	115	39	}	}	PUNCT
bjmsr-315	115	40	and	and	CCONJ
bjmsr-315	115	41	s(ßki)<s(ßk	s(ßki)<s(ßk	PROPN
bjmsr-315	115	42	)	)	PUNCT
bjmsr-315	115	43	.	.	PUNCT
bjmsr-315	116	1	s(ßki	s(ßki	PROPN
bjmsr-315	116	2	)	)	PUNCT
bjmsr-315	116	3	is	be	AUX
bjmsr-315	116	4	the	the	DET
bjmsr-315	116	5	minimum	minimum	NOUN
bjmsr-315	116	6	of	of	ADP
bjmsr-315	116	7	all	all	DET
bjmsr-315	116	8	norms	norm	NOUN
bjmsr-315	116	9	obtained	obtain	VERB
bjmsr-315	116	10	if	if	SCONJ
bjmsr-315	116	11	ui	ui	PROPN
bjmsr-315	116	12	k	k	PROPN
bjmsr-315	116	13	were	be	AUX
bjmsr-315	116	14	replaced	replace	VERB
bjmsr-315	116	15	by	by	ADP
bjmsr-315	116	16	the	the	DET
bjmsr-315	116	17	different	different	ADJ
bjmsr-315	116	18	points	point	NOUN
bjmsr-315	116	19	of	of	ADP
bjmsr-315	116	20	the	the	DET
bjmsr-315	116	21	set	set	NOUN
bjmsr-315	116	22	{	{	PUNCT
bjmsr-315	116	23	x	x	NOUN
bjmsr-315	116	24	-	-	PUNCT
bjmsr-315	116	25	uk	uk	NOUN
bjmsr-315	116	26	}	}	PUNCT
bjmsr-315	116	27	,	,	PUNCT
bjmsr-315	116	28	as	as	SCONJ
bjmsr-315	116	29	indicated	indicate	VERB
bjmsr-315	116	30	by	by	ADP
bjmsr-315	116	31	the	the	DET
bjmsr-315	116	32	above	above	ADJ
bjmsr-315	116	33	relation	relation	NOUN
bjmsr-315	116	34	.	.	PUNCT
bjmsr-315	117	1	in	in	ADP
bjmsr-315	117	2	going	go	VERB
bjmsr-315	117	3	from	from	ADP
bjmsr-315	117	4	vertex	vertex	NOUN
bjmsr-315	117	5	(	(	PUNCT
bjmsr-315	117	6	ßk	ßk	PROPN
bjmsr-315	117	7	,	,	PUNCT
bjmsr-315	117	8	s(ßk	s(ßk	NOUN
bjmsr-315	117	9	)	)	PUNCT
bjmsr-315	117	10	)	)	PUNCT
bjmsr-315	117	11	to	to	ADP
bjmsr-315	117	12	(	(	PUNCT
bjmsr-315	117	13	ßki	ßki	PROPN
bjmsr-315	117	14	,	,	PUNCT
bjmsr-315	117	15	s(ßki	s(ßki	PROPN
bjmsr-315	117	16	)	)	PUNCT
bjmsr-315	117	17	)	)	PUNCT
bjmsr-315	117	18	,	,	PUNCT
bjmsr-315	117	19	one	one	NUM
bjmsr-315	117	20	or	or	CCONJ
bjmsr-315	117	21	more	more	ADJ
bjmsr-315	117	22	vertices	vertex	NOUN
bjmsr-315	117	23	on	on	ADP
bjmsr-315	117	24	k	k	PROPN
bjmsr-315	117	25	might	might	AUX
bjmsr-315	117	26	have	have	AUX
bjmsr-315	117	27	been	be	AUX
bjmsr-315	117	28	by	by	ADP
bjmsr-315	117	29	-	-	PUNCT
bjmsr-315	117	30	passed	pass	VERB
bjmsr-315	117	31	.	.	PUNCT
bjmsr-315	118	1	the	the	DET
bjmsr-315	118	2	nearest	near	ADJ
bjmsr-315	118	3	vertex	vertex	NOUN
bjmsr-315	118	4	to	to	ADP
bjmsr-315	118	5	(	(	PUNCT
bjmsr-315	118	6	fk	fk	INTJ
bjmsr-315	118	7	,	,	PUNCT
bjmsr-315	118	8	s(fk	s(fk	PROPN
bjmsr-315	118	9	)	)	PUNCT
bjmsr-315	118	10	)	)	PUNCT
bjmsr-315	118	11	and	and	CCONJ
bjmsr-315	118	12	below	below	ADP
bjmsr-315	118	13	it	it	PRON
bjmsr-315	118	14	on	on	ADP
bjmsr-315	118	15	the	the	DET
bjmsr-315	118	16	edge	edge	NOUN
bjmsr-315	118	17	parallel	parallel	ADJ
bjmsr-315	118	18	to	to	ADP
bjmsr-315	118	19	the	the	DET
bjmsr-315	118	20	ith	ith	PROPN
bjmsr-315	118	21	parameter	parameter	PROPN
bjmsr-315	118	22	space	space	NOUN
bjmsr-315	118	23	coordinate	coordinate	NOUN
bjmsr-315	118	24	axis	axis	NOUN
bjmsr-315	118	25	,	,	PUNCT
bjmsr-315	118	26	say	say	VERB
bjmsr-315	118	27	the	the	DET
bjmsr-315	118	28	vertex	vertex	NOUN
bjmsr-315	118	29	(	(	PUNCT
bjmsr-315	118	30	(	(	PUNCT
bjmsr-315	118	31	fk	fk	INTJ
bjmsr-315	118	32	-	-	PUNCT
bjmsr-315	118	33	trei),s(fk	trei),s(fk	NOUN
bjmsr-315	118	34	-	-	PUNCT
bjmsr-315	118	35	trei	trei	NOUN
bjmsr-315	118	36	)	)	PUNCT
bjmsr-315	118	37	)	)	PUNCT
bjmsr-315	118	38	,	,	PUNCT
bjmsr-315	118	39	is	be	AUX
bjmsr-315	118	40	obtained	obtain	VERB
bjmsr-315	118	41	from	from	ADP
bjmsr-315	118	42	:	:	PUNCT
bjmsr-315	118	43	│	│	NOUN
bjmsr-315	118	44	l(fk	l(fk	NOUN
bjmsr-315	118	45	,	,	PUNCT
bjmsr-315	118	46	xs)-f(xs	xs)-f(xs	NUM
bjmsr-315	118	47	)	)	PUNCT
bjmsr-315	118	48	│	│	X
bjmsr-315	118	49	│	│	VERB
bjmsr-315	118	50	tr	tr	VERB
bjmsr-315	118	51	│	│	ADJ
bjmsr-315	118	52	=	=	SYM
bjmsr-315	118	53	min	min	NOUN
bjmsr-315	118	54	{	{	PUNCT
bjmsr-315	118	55	─	─	PROPN
bjmsr-315	118	56	─	─	ADJ
bjmsr-315	118	57	─	─	ADJ
bjmsr-315	118	58	─	─	ADJ
bjmsr-315	118	59	─	─	ADJ
bjmsr-315	118	60	─	─	ADJ
bjmsr-315	118	61	─	─	ADJ
bjmsr-315	118	62	─	─	ADJ
bjmsr-315	118	63	─	─	ADJ
bjmsr-315	118	64	─	─	ADJ
bjmsr-315	118	65	}	}	PUNCT
bjmsr-315	118	66	xsε{x	xsε{x	PROPN
bjmsr-315	118	67	-	-	PUNCT
bjmsr-315	118	68	uk	uk	NOUN
bjmsr-315	118	69	}	}	PUNCT
bjmsr-315	118	70	(	(	PUNCT
bjmsr-315	118	71	12	12	NUM
bjmsr-315	118	72	)	)	PUNCT
bjmsr-315	118	73	s	s	PART
bjmsr-315	118	74	│	│	NOUN
bjmsr-315	118	75	πi(xs	πi(x	NOUN
bjmsr-315	118	76	)	)	PUNCT
bjmsr-315	118	77	│	│	NOUN
bjmsr-315	118	78	the	the	DET
bjmsr-315	118	79	point	point	NOUN
bjmsr-315	118	80	xr	xr	PROPN
bjmsr-315	118	81	is	be	AUX
bjmsr-315	118	82	characterized	characterize	VERB
bjmsr-315	118	83	by	by	ADP
bjmsr-315	118	84	,	,	PUNCT
bjmsr-315	118	85	sgn[l(fk	sgn[l(fk	PROPN
bjmsr-315	118	86	,	,	PUNCT
bjmsr-315	118	87	xl	xl	PROPN
bjmsr-315	118	88	)	)	PUNCT
bjmsr-315	118	89	f(xl	f(xl	NOUN
bjmsr-315	118	90	)	)	PUNCT
bjmsr-315	118	91	]	]	PUNCT
bjmsr-315	119	1	=	=	SYM
bjmsr-315	119	2	sgn[l(fki	sgn[l(fki	PROPN
bjmsr-315	119	3	,	,	PUNCT
bjmsr-315	119	4	xl	xl	PROPN
bjmsr-315	119	5	)	)	PUNCT
bjmsr-315	119	6	f(xl	f(xl	NOUN
bjmsr-315	119	7	)	)	PUNCT
bjmsr-315	119	8	]	]	PUNCT
bjmsr-315	119	9	,	,	PUNCT
bjmsr-315	119	10	xlε{x	xlε{x	PROPN
bjmsr-315	119	11	-	-	PUNCT
bjmsr-315	119	12	uk	uk	PROPN
bjmsr-315	119	13	-	-	PUNCT
bjmsr-315	119	14	xr	xr	NOUN
bjmsr-315	119	15	}	}	PUNCT
bjmsr-315	119	16	(	(	PUNCT
bjmsr-315	119	17	13	13	NUM
bjmsr-315	119	18	)	)	PUNCT
bjmsr-315	119	19	now	now	ADV
bjmsr-315	119	20	,	,	PUNCT
bjmsr-315	119	21	if	if	SCONJ
bjmsr-315	119	22	there	there	PRON
bjmsr-315	119	23	is	be	VERB
bjmsr-315	119	24	not	not	PART
bjmsr-315	119	25	any	any	DET
bjmsr-315	119	26	δ	δ	NOUN
bjmsr-315	119	27	such	such	ADJ
bjmsr-315	119	28	that	that	SCONJ
bjmsr-315	119	29	s(fk	s(fk	PROPN
bjmsr-315	119	30	-	-	PROPN
bjmsr-315	119	31	δei)<s(fk	δei)<s(fk	NOUN
bjmsr-315	119	32	)	)	PUNCT
bjmsr-315	119	33	,	,	PUNCT
bjmsr-315	119	34	then	then	ADV
bjmsr-315	119	35	s(fk	s(fk	NUM
bjmsr-315	119	36	)	)	PUNCT
bjmsr-315	119	37	could	could	AUX
bjmsr-315	119	38	not	not	PART
bjmsr-315	119	39	be	be	AUX
bjmsr-315	119	40	reduced	reduce	VERB
bjmsr-315	119	41	by	by	ADP
bjmsr-315	119	42	moving	move	VERB
bjmsr-315	119	43	on	on	ADP
bjmsr-315	119	44	k	k	X
bjmsr-315	119	45	along	along	ADP
bjmsr-315	119	46	the	the	DET
bjmsr-315	119	47	edge	edge	NOUN
bjmsr-315	119	48	parallel	parallel	ADJ
bjmsr-315	119	49	to	to	ADP
bjmsr-315	119	50	the	the	DET
bjmsr-315	119	51	ith	ith	PROPN
bjmsr-315	119	52	parameter	parameter	PROPN
bjmsr-315	119	53	space	space	NOUN
bjmsr-315	119	54	coordinate	coordinate	NOUN
bjmsr-315	119	55	axis	axis	NOUN
bjmsr-315	119	56	and	and	CCONJ
bjmsr-315	119	57	ui	ui	NOUN
bjmsr-315	119	58	k	k	PROPN
bjmsr-315	119	59	should	should	AUX
bjmsr-315	119	60	not	not	PART
bjmsr-315	119	61	be	be	AUX
bjmsr-315	119	62	replaced	replace	VERB
bjmsr-315	119	63	by	by	ADP
bjmsr-315	119	64	another	another	DET
bjmsr-315	119	65	point	point	NOUN
bjmsr-315	119	66	from	from	ADP
bjmsr-315	119	67	the	the	DET
bjmsr-315	119	68	set	set	NOUN
bjmsr-315	119	69	{	{	PUNCT
bjmsr-315	119	70	x	x	NOUN
bjmsr-315	119	71	-	-	PUNCT
bjmsr-315	119	72	uk	uk	NOUN
bjmsr-315	119	73	}	}	PUNCT
bjmsr-315	119	74	.	.	PUNCT
bjmsr-315	120	1	this	this	DET
bjmsr-315	120	2	iteration	iteration	NOUN
bjmsr-315	120	3	is	be	AUX
bjmsr-315	120	4	repeated	repeat	VERB
bjmsr-315	120	5	m	m	NOUN
bjmsr-315	120	6	times	time	NOUN
bjmsr-315	120	7	,	,	PUNCT
bjmsr-315	120	8	once	once	ADV
bjmsr-315	120	9	for	for	ADP
bjmsr-315	120	10	each	each	DET
bjmsr-315	120	11	point	point	NOUN
bjmsr-315	120	12	in	in	ADP
bjmsr-315	120	13	uk	uk	PROPN
bjmsr-315	120	14	in	in	ADP
bjmsr-315	120	15	succession	succession	NOUN
bjmsr-315	120	16	.	.	PUNCT
bjmsr-315	121	1	the	the	DET
bjmsr-315	121	2	whole	whole	ADJ
bjmsr-315	121	3	cycle	cycle	NOUN
bjmsr-315	121	4	is	be	AUX
bjmsr-315	121	5	then	then	ADV
bjmsr-315	121	6	repeated	repeat	VERB
bjmsr-315	121	7	a	a	DET
bjmsr-315	121	8	finite	finite	ADJ
bjmsr-315	121	9	number	number	NOUN
bjmsr-315	121	10	of	of	ADP
bjmsr-315	121	11	times	time	NOUN
bjmsr-315	121	12	until	until	SCONJ
bjmsr-315	121	13	the	the	DET
bjmsr-315	121	14	solution	solution	NOUN
bjmsr-315	121	15	(	(	PUNCT
bjmsr-315	121	16	ß*,d	ß*,d	NOUN
bjmsr-315	121	17	*	*	PUNCT
bjmsr-315	121	18	)	)	PUNCT
bjmsr-315	121	19	is	be	AUX
bjmsr-315	121	20	reached	reach	VERB
bjmsr-315	121	21	(	(	PUNCT
bjmsr-315	121	22	see	see	VERB
bjmsr-315	121	23	also	also	ADV
bjmsr-315	121	24	,	,	PUNCT
bjmsr-315	121	25	abdelmalek	abdelmalek	PROPN
bjmsr-315	121	26	(	(	PUNCT
bjmsr-315	121	27	1974	1974	NUM
bjmsr-315	121	28	)	)	PUNCT
bjmsr-315	121	29	)	)	PUNCT
bjmsr-315	121	30	.	.	PUNCT
bjmsr-315	122	1	relation	relation	NOUN
bjmsr-315	122	2	of	of	ADP
bjmsr-315	122	3	this	this	DET
bjmsr-315	122	4	algorithm	algorithm	NOUN
bjmsr-315	122	5	with	with	ADP
bjmsr-315	122	6	the	the	DET
bjmsr-315	122	7	simplex	simplex	NOUN
bjmsr-315	122	8	method	method	NOUN
bjmsr-315	122	9	has	have	AUX
bjmsr-315	122	10	been	be	AUX
bjmsr-315	122	11	discussed	discuss	VERB
bjmsr-315	122	12	by	by	ADP
bjmsr-315	122	13	abdelmalek	abdelmalek	PROPN
bjmsr-315	122	14	(	(	PUNCT
bjmsr-315	122	15	1974	1974	NUM
bjmsr-315	122	16	)	)	PUNCT
bjmsr-315	122	17	.	.	PUNCT
bjmsr-315	123	1	he	he	PRON
bjmsr-315	123	2	shows	show	VERB
bjmsr-315	123	3	that	that	SCONJ
bjmsr-315	123	4	usow	usow	NOUN
bjmsr-315	123	5	's	's	PART
bjmsr-315	123	6	algorithm	algorithm	NOUN
bjmsr-315	123	7	is	be	AUX
bjmsr-315	123	8	completely	completely	ADV
bjmsr-315	123	9	equivalent	equivalent	ADJ
bjmsr-315	123	10	to	to	ADP
bjmsr-315	123	11	a	a	DET
bjmsr-315	123	12	dual	dual	ADJ
bjmsr-315	123	13	simplex	simplex	NOUN
bjmsr-315	123	14	algorithm	algorithm	NOUN
bjmsr-315	123	15	applied	apply	VERB
bjmsr-315	123	16	to	to	ADP
bjmsr-315	123	17	a	a	DET
bjmsr-315	123	18	linear	linear	ADJ
bjmsr-315	123	19	programming	programming	NOUN
bjmsr-315	123	20	model	model	NOUN
bjmsr-315	123	21	with	with	ADP
bjmsr-315	123	22	nonnegative	nonnegative	ADJ
bjmsr-315	123	23	bounded	bounded	ADJ
bjmsr-315	123	24	variables	variable	NOUN
bjmsr-315	123	25	,	,	PUNCT
bjmsr-315	123	26	and	and	CCONJ
bjmsr-315	123	27	one	one	NUM
bjmsr-315	123	28	iteration	iteration	NOUN
bjmsr-315	123	29	in	in	ADP
bjmsr-315	123	30	the	the	DET
bjmsr-315	123	31	former	former	NOUN
bjmsr-315	123	32	is	be	AUX
bjmsr-315	123	33	equivalent	equivalent	ADJ
bjmsr-315	123	34	to	to	ADP
bjmsr-315	123	35	one	one	NUM
bjmsr-315	123	36	or	or	CCONJ
bjmsr-315	123	37	more	more	ADJ
bjmsr-315	123	38	iterations	iteration	NOUN
bjmsr-315	123	39	in	in	ADP
bjmsr-315	123	40	the	the	DET
bjmsr-315	123	41	latter	latter	ADJ
bjmsr-315	123	42	.	.	PUNCT
bjmsr-315	124	1	bloomfield	bloomfield	PROPN
bjmsr-315	124	2	and	and	CCONJ
bjmsr-315	124	3	steiger	steiger	PROPN
bjmsr-315	124	4	(	(	PUNCT
bjmsr-315	124	5	1980	1980	NUM
bjmsr-315	124	6	)	)	PUNCT
bjmsr-315	124	7	devise	devise	VERB
bjmsr-315	124	8	an	an	DET
bjmsr-315	124	9	efficient	efficient	ADJ
bjmsr-315	124	10	algorithm	algorithm	NOUN
bjmsr-315	124	11	based	base	VERB
bjmsr-315	124	12	on	on	ADP
bjmsr-315	124	13	the	the	DET
bjmsr-315	124	14	proposition	proposition	NOUN
bjmsr-315	124	15	of	of	ADP
bjmsr-315	124	16	usow	usow	NOUN
bjmsr-315	124	17	explained	explain	VERB
bjmsr-315	124	18	above	above	ADV
bjmsr-315	124	19	.	.	PUNCT
bjmsr-315	125	1	sharpe	sharpe	PROPN
bjmsr-315	125	2	(	(	PUNCT
bjmsr-315	125	3	1971	1971	NUM
bjmsr-315	125	4	)	)	PUNCT
bjmsr-315	125	5	by	by	ADP
bjmsr-315	125	6	applying	apply	VERB
bjmsr-315	125	7	the	the	DET
bjmsr-315	125	8	l1	l1	PROPN
bjmsr-315	125	9	norm	norm	NOUN
bjmsr-315	125	10	regression	regression	NOUN
bjmsr-315	125	11	to	to	ADP
bjmsr-315	125	12	the	the	DET
bjmsr-315	125	13	portfolio	portfolio	NOUN
bjmsr-315	125	14	and	and	CCONJ
bjmsr-315	125	15	its	its	PRON
bjmsr-315	125	16	rate	rate	NOUN
bjmsr-315	125	17	of	of	ADP
bjmsr-315	125	18	return	return	NOUN
bjmsr-315	125	19	,	,	PUNCT
bjmsr-315	125	20	gives	give	VERB
bjmsr-315	125	21	an	an	DET
bjmsr-315	125	22	algorithm	algorithm	NOUN
bjmsr-315	125	23	for	for	ADP
bjmsr-315	125	24	the	the	DET
bjmsr-315	125	25	two	two	NUM
bjmsr-315	125	26	parameters	parameter	NOUN
bjmsr-315	125	27	linear	linear	VERB
bjmsr-315	125	28	regression	regression	NOUN
bjmsr-315	125	29	model	model	NOUN
bjmsr-315	125	30	.	.	PUNCT
bjmsr-315	126	1	he	he	PRON
bjmsr-315	126	2	argues	argue	VERB
bjmsr-315	126	3	that	that	SCONJ
bjmsr-315	126	4	for	for	ADP
bjmsr-315	126	5	the	the	DET
bjmsr-315	126	6	simple	simple	ADJ
bjmsr-315	126	7	model	model	NOUN
bjmsr-315	126	8	with	with	ADP
bjmsr-315	126	9	the	the	DET
bjmsr-315	126	10	objective	objective	ADJ
bjmsr-315	126	11	function	function	NOUN
bjmsr-315	126	12	,	,	PUNCT
bjmsr-315	126	13	n	n	PROPN
bjmsr-315	126	14	s	s	NOUN
bjmsr-315	126	15	=	=	SYM
bjmsr-315	126	16	σ	σ	PROPN
bjmsr-315	126	17	│	│	NOUN
bjmsr-315	126	18	yi	yi	PROPN
bjmsr-315	126	19	(	(	PUNCT
bjmsr-315	126	20	ß0	ß0	NOUN
bjmsr-315	126	21	+	+	CCONJ
bjmsr-315	126	22	ß1xi1	ß1xi1	NOUN
bjmsr-315	126	23	)	)	PUNCT
bjmsr-315	126	24	│	│	X
bjmsr-315	126	25	(	(	PUNCT
bjmsr-315	126	26	14	14	NUM
bjmsr-315	126	27	)	)	PUNCT
bjmsr-315	126	28	i=1	i=1	PROPN
bjmsr-315	127	1	it	it	PRON
bjmsr-315	127	2	must	must	AUX
bjmsr-315	127	3	be	be	AUX
bjmsr-315	127	4	possible	possible	ADJ
bjmsr-315	127	5	to	to	PART
bjmsr-315	127	6	assign	assign	VERB
bjmsr-315	127	7	half	half	NOUN
bjmsr-315	127	8	of	of	ADP
bjmsr-315	127	9	the	the	DET
bjmsr-315	127	10	points	point	NOUN
bjmsr-315	127	11	above	above	ADV
bjmsr-315	127	12	and	and	CCONJ
bjmsr-315	127	13	half	half	NOUN
bjmsr-315	127	14	below	below	ADP
bjmsr-315	127	15	the	the	DET
bjmsr-315	127	16	regression	regression	NOUN
bjmsr-315	127	17	line	line	NOUN
bjmsr-315	127	18	.	.	PUNCT
bjmsr-315	128	1	with	with	ADP
bjmsr-315	128	2	any	any	DET
bjmsr-315	128	3	given	give	VERB
bjmsr-315	128	4	value	value	NOUN
bjmsr-315	128	5	of	of	ADP
bjmsr-315	128	6	ß~	ß~	PROPN
bjmsr-315	128	7	1	1	NUM
bjmsr-315	128	8	,	,	PUNCT
bjmsr-315	128	9	we	we	PRON
bjmsr-315	128	10	can	can	AUX
bjmsr-315	128	11	derive	derive	VERB
bjmsr-315	128	12	ß0	ß0	NOUN
bjmsr-315	128	13	as	as	ADP
bjmsr-315	128	14	the	the	DET
bjmsr-315	128	15	median	median	NOUN
bjmsr-315	128	16	of	of	ADP
bjmsr-315	128	17	ß0i	ß0i	NOUN
bjmsr-315	128	18	=	=	SYM
bjmsr-315	128	19	yi	yi	PROPN
bjmsr-315	128	20	-	-	PROPN
bjmsr-315	128	21	ß~	ß~	PROPN
bjmsr-315	128	22	1xi1	1xi1	NUM
bjmsr-315	128	23	.	.	PUNCT
bjmsr-315	129	1	now	now	ADV
bjmsr-315	129	2	segregate	segregate	VERB
bjmsr-315	129	3	the	the	DET
bjmsr-315	129	4	points	point	NOUN
bjmsr-315	129	5	,	,	PUNCT
bjmsr-315	129	6	such	such	ADJ
bjmsr-315	129	7	that	that	SCONJ
bjmsr-315	129	8	:	:	PUNCT
bjmsr-315	129	9	nabove	nabove	PROPN
bjmsr-315	129	10	nbelow	nbelow	PROPN
bjmsr-315	129	11	s=	s=	NOUN
bjmsr-315	129	12	σ	σ	PROPN
bjmsr-315	130	1	[	[	X
bjmsr-315	130	2	yi-(ß0+ß1xi1	yi-(ß0+ß1xi1	NOUN
bjmsr-315	130	3	)	)	PUNCT
bjmsr-315	130	4	]	]	PUNCT
bjmsr-315	131	1	σ	σ	X
bjmsr-315	132	1	[	[	X
bjmsr-315	132	2	yi-(ß0+ß1xi1	yi-(ß0+ß1xi1	NOUN
bjmsr-315	132	3	)	)	PUNCT
bjmsr-315	132	4	]	]	PUNCT
bjmsr-315	132	5	(	(	PUNCT
bjmsr-315	132	6	15	15	NUM
bjmsr-315	132	7	)	)	PUNCT
bjmsr-315	132	8	iabove	iabove	NOUN
bjmsr-315	132	9	ibelow	ibelow	ADV
bjmsr-315	132	10	rearranging	rearrange	VERB
bjmsr-315	132	11	the	the	DET
bjmsr-315	132	12	terms	term	NOUN
bjmsr-315	132	13	and	and	CCONJ
bjmsr-315	132	14	note	note	VERB
bjmsr-315	132	15	that	that	SCONJ
bjmsr-315	132	16	nabove	nabove	NOUN
bjmsr-315	132	17	=	=	SYM
bjmsr-315	132	18	nbelow	nbelow	NOUN
bjmsr-315	132	19	gives	give	VERB
bjmsr-315	132	20	,	,	PUNCT
bjmsr-315	132	21	s	s	PART
bjmsr-315	132	22	=	=	SYM
bjmsr-315	132	23	k1	k1	PROPN
bjmsr-315	132	24	+	+	CCONJ
bjmsr-315	132	25	k2ß1	k2ß1	PROPN
bjmsr-315	132	26	(	(	PUNCT
bjmsr-315	132	27	16	16	NUM
bjmsr-315	132	28	)	)	PUNCT
bjmsr-315	132	29	where	where	SCONJ
bjmsr-315	132	30	,	,	PUNCT
bjmsr-315	132	31	nabove	nabove	VERB
bjmsr-315	132	32	nbelow	nbelow	ADJ
bjmsr-315	132	33	k1	k1	NOUN
bjmsr-315	132	34	=	=	SYM
bjmsr-315	132	35	σ	σ	PROPN
bjmsr-315	132	36	yi	yi	PROPN
bjmsr-315	132	37	σ	σ	PROPN
bjmsr-315	132	38	yi	yi	PROPN
bjmsr-315	132	39	(	(	PUNCT
bjmsr-315	132	40	17	17	NUM
bjmsr-315	132	41	)	)	PUNCT
bjmsr-315	132	42	iabove	iabove	NOUN
bjmsr-315	132	43	ibelow	ibelow	ADJ
bjmsr-315	132	44	nabove	nabove	PROPN
bjmsr-315	132	45	nbelow	nbelow	ADJ
bjmsr-315	132	46	k2	k2	NOUN
bjmsr-315	132	47	=	=	PUNCT
bjmsr-315	133	1	-σ	-σ	PUNCT
bjmsr-315	133	2	xi1	xi1	PROPN
bjmsr-315	134	1	+	+	PROPN
bjmsr-315	134	2	σ	σ	PROPN
bjmsr-315	134	3	xi1	xi1	PROPN
bjmsr-315	134	4	(	(	PUNCT
bjmsr-315	134	5	18	18	NUM
bjmsr-315	134	6	)	)	PUNCT
bjmsr-315	134	7	iabove	iabove	NOUN
bjmsr-315	134	8	ibelow	ibelow	ADJ
bjmsr-315	134	9	copyright	copyright	PROPN
bjmsr-315	135	1	©	©	PROPN
bjmsr-315	135	2	cc	cc	PROPN
bjmsr-315	135	3	-	-	PUNCT
bjmsr-315	135	4	by	by	ADP
bjmsr-315	135	5	-	-	PUNCT
bjmsr-315	135	6	nc	nc	PROPN
bjmsr-315	135	7	2019	2019	NUM
bjmsr-315	135	8	,	,	PUNCT
bjmsr-315	135	9	bjmsr	bjmsr	PROPN
bjmsr-315	135	10	www.cribfb.com/journal/index.php/bjmsr	www.cribfb.com/journal/index.php/bjmsr	X
bjmsr-315	136	1	bangladesh	bangladesh	PROPN
bjmsr-315	136	2	journal	journal	PROPN
bjmsr-315	136	3	of	of	ADP
bjmsr-315	136	4	multidisciplinary	multidisciplinary	ADJ
bjmsr-315	136	5	scientific	scientific	ADJ
bjmsr-315	136	6	research	research	NOUN
bjmsr-315	136	7	vol	vol	NOUN
bjmsr-315	136	8	.	.	PROPN
bjmsr-315	136	9	1	1	NUM
bjmsr-315	136	10	,	,	PUNCT
bjmsr-315	136	11	no	no	INTJ
bjmsr-315	136	12	.	.	NOUN
bjmsr-315	136	13	1	1	NUM
bjmsr-315	136	14	;	;	PUNCT
bjmsr-315	136	15	2019	2019	NUM
bjmsr-315	136	16	54	54	NUM
bjmsr-315	136	17	the	the	DET
bjmsr-315	136	18	overall	overall	ADJ
bjmsr-315	136	19	solution	solution	NOUN
bjmsr-315	136	20	strategy	strategy	NOUN
bjmsr-315	136	21	may	may	AUX
bjmsr-315	136	22	now	now	ADV
bjmsr-315	136	23	be	be	AUX
bjmsr-315	136	24	formulated	formulate	VERB
bjmsr-315	136	25	as	as	SCONJ
bjmsr-315	136	26	follows	follow	VERB
bjmsr-315	136	27	.	.	PUNCT
bjmsr-315	137	1	any	any	DET
bjmsr-315	137	2	value	value	NOUN
bjmsr-315	137	3	of	of	ADP
bjmsr-315	137	4	ß1	ß1	PROPN
bjmsr-315	137	5	may	may	AUX
bjmsr-315	137	6	be	be	AUX
bjmsr-315	137	7	chosen	choose	VERB
bjmsr-315	137	8	at	at	ADP
bjmsr-315	137	9	the	the	DET
bjmsr-315	137	10	outset	outset	NOUN
bjmsr-315	137	11	.	.	PUNCT
bjmsr-315	138	1	by	by	ADP
bjmsr-315	138	2	computing	compute	VERB
bjmsr-315	138	3	ß0i	ß0i	NOUN
bjmsr-315	138	4	,	,	PUNCT
bjmsr-315	138	5	segregate	segregate	VERB
bjmsr-315	138	6	the	the	DET
bjmsr-315	138	7	points	point	NOUN
bjmsr-315	138	8	above	above	ADP
bjmsr-315	138	9	and	and	CCONJ
bjmsr-315	138	10	below	below	ADP
bjmsr-315	138	11	the	the	DET
bjmsr-315	138	12	line	line	NOUN
bjmsr-315	138	13	.	.	PUNCT
bjmsr-315	139	1	by	by	ADP
bjmsr-315	139	2	using	use	VERB
bjmsr-315	139	3	equations	equation	NOUN
bjmsr-315	139	4	(	(	PUNCT
bjmsr-315	139	5	16	16	NUM
bjmsr-315	139	6	)	)	PUNCT
bjmsr-315	139	7	through	through	ADP
bjmsr-315	139	8	(	(	PUNCT
bjmsr-315	139	9	18	18	NUM
bjmsr-315	139	10	)	)	PUNCT
bjmsr-315	139	11	,	,	PUNCT
bjmsr-315	139	12	the	the	DET
bjmsr-315	139	13	corresponding	corresponding	ADJ
bjmsr-315	139	14	segment	segment	NOUN
bjmsr-315	139	15	of	of	ADP
bjmsr-315	139	16	the	the	DET
bjmsr-315	139	17	s	s	NOUN
bjmsr-315	139	18	and	and	CCONJ
bjmsr-315	139	19	ß	ß	DET
bjmsr-315	139	20	1	1	NUM
bjmsr-315	139	21	relation	relation	NOUN
bjmsr-315	139	22	is	be	AUX
bjmsr-315	139	23	calculated	calculate	VERB
bjmsr-315	139	24	.	.	PUNCT
bjmsr-315	140	1	sign	sign	NOUN
bjmsr-315	140	2	of	of	ADP
bjmsr-315	140	3	k2	k2	PROPN
bjmsr-315	140	4	indicates	indicate	VERB
bjmsr-315	140	5	the	the	DET
bjmsr-315	140	6	appropriate	appropriate	ADJ
bjmsr-315	140	7	direction	direction	NOUN
bjmsr-315	140	8	for	for	ADP
bjmsr-315	140	9	the	the	DET
bjmsr-315	140	10	next	next	ADJ
bjmsr-315	140	11	iteration	iteration	NOUN
bjmsr-315	140	12	.	.	PUNCT
bjmsr-315	141	1	if	if	SCONJ
bjmsr-315	141	2	k2	k2	PROPN
bjmsr-315	141	3	is	be	AUX
bjmsr-315	141	4	positive	positive	ADJ
bjmsr-315	141	5	,	,	PUNCT
bjmsr-315	141	6	only	only	ADV
bjmsr-315	141	7	smaller	small	ADJ
bjmsr-315	141	8	values	value	NOUN
bjmsr-315	141	9	of	of	ADP
bjmsr-315	141	10	ß1	ß1	NOUN
bjmsr-315	141	11	need	need	AUX
bjmsr-315	141	12	be	be	AUX
bjmsr-315	141	13	considered	consider	VERB
bjmsr-315	141	14	.	.	PUNCT
bjmsr-315	142	1	if	if	SCONJ
bjmsr-315	142	2	k2	k2	PROPN
bjmsr-315	142	3	is	be	AUX
bjmsr-315	142	4	negative	negative	ADJ
bjmsr-315	142	5	,	,	PUNCT
bjmsr-315	142	6	only	only	ADV
bjmsr-315	142	7	larger	large	ADJ
bjmsr-315	142	8	values	value	NOUN
bjmsr-315	142	9	of	of	ADP
bjmsr-315	142	10	ß1	ß1	NOUN
bjmsr-315	142	11	need	need	AUX
bjmsr-315	142	12	be	be	AUX
bjmsr-315	142	13	considered	consider	VERB
bjmsr-315	142	14	.	.	PUNCT
bjmsr-315	143	1	if	if	SCONJ
bjmsr-315	143	2	k2	k2	PROPN
bjmsr-315	143	3	is	be	AUX
bjmsr-315	143	4	zero	zero	NUM
bjmsr-315	143	5	,	,	PUNCT
bjmsr-315	143	6	the	the	DET
bjmsr-315	143	7	initial	initial	ADJ
bjmsr-315	143	8	value	value	NOUN
bjmsr-315	143	9	of	of	ADP
bjmsr-315	143	10	ß1	ß1	PROPN
bjmsr-315	143	11	is	be	AUX
bjmsr-315	143	12	a	a	DET
bjmsr-315	143	13	solution	solution	NOUN
bjmsr-315	143	14	.	.	PUNCT
bjmsr-315	144	1	when	when	SCONJ
bjmsr-315	144	2	a	a	DET
bjmsr-315	144	3	borderline	borderline	NOUN
bjmsr-315	144	4	(	(	PUNCT
bjmsr-315	144	5	median	median	NOUN
bjmsr-315	144	6	of	of	ADP
bjmsr-315	144	7	ß0i	ß0i	NOUN
bjmsr-315	144	8	)	)	PUNCT
bjmsr-315	144	9	and	and	CCONJ
bjmsr-315	144	10	the	the	DET
bjmsr-315	144	11	direction	direction	NOUN
bjmsr-315	144	12	of	of	ADP
bjmsr-315	144	13	search	search	NOUN
bjmsr-315	144	14	is	be	AUX
bjmsr-315	144	15	determined	determine	VERB
bjmsr-315	144	16	,	,	PUNCT
bjmsr-315	144	17	the	the	DET
bjmsr-315	144	18	nearest	near	ADJ
bjmsr-315	144	19	intersection	intersection	NOUN
bjmsr-315	144	20	of	of	ADP
bjmsr-315	144	21	the	the	DET
bjmsr-315	144	22	present	present	ADJ
bjmsr-315	144	23	borderline	borderline	NOUN
bjmsr-315	144	24	with	with	ADP
bjmsr-315	144	25	another	another	PRON
bjmsr-315	144	26	should	should	AUX
bjmsr-315	144	27	be	be	AUX
bjmsr-315	144	28	found	find	VERB
bjmsr-315	144	29	.	.	PUNCT
bjmsr-315	145	1	the	the	DET
bjmsr-315	145	2	calculations	calculation	NOUN
bjmsr-315	145	3	can	can	AUX
bjmsr-315	145	4	be	be	AUX
bjmsr-315	145	5	reduced	reduce	VERB
bjmsr-315	145	6	by	by	ADP
bjmsr-315	145	7	comparing	compare	VERB
bjmsr-315	145	8	slopes	slope	NOUN
bjmsr-315	145	9	(	(	PUNCT
bjmsr-315	145	10	x	x	PROPN
bjmsr-315	145	11	i1	i1	PROPN
bjmsr-315	145	12	values	value	NOUN
bjmsr-315	145	13	)	)	PUNCT
bjmsr-315	145	14	to	to	PART
bjmsr-315	145	15	determine	determine	VERB
bjmsr-315	145	16	whether	whether	SCONJ
bjmsr-315	145	17	or	or	CCONJ
bjmsr-315	145	18	not	not	PART
bjmsr-315	145	19	two	two	NUM
bjmsr-315	145	20	lines	line	NOUN
bjmsr-315	145	21	intersect	intersect	ADJ
bjmsr-315	145	22	in	in	ADP
bjmsr-315	145	23	the	the	DET
bjmsr-315	145	24	region	region	NOUN
bjmsr-315	145	25	of	of	ADP
bjmsr-315	145	26	interest	interest	NOUN
bjmsr-315	145	27	(	(	PUNCT
bjmsr-315	145	28	thus	thus	ADV
bjmsr-315	145	29	avoiding	avoid	VERB
bjmsr-315	145	30	needless	needless	ADJ
bjmsr-315	145	31	division	division	NOUN
bjmsr-315	145	32	operations	operation	NOUN
bjmsr-315	145	33	)	)	PUNCT
bjmsr-315	145	34	.	.	PUNCT
bjmsr-315	146	1	the	the	DET
bjmsr-315	146	2	new	new	ADJ
bjmsr-315	146	3	values	value	NOUN
bjmsr-315	146	4	k1	k1	NOUN
bjmsr-315	146	5	and	and	CCONJ
bjmsr-315	146	6	k2	k2	PROPN
bjmsr-315	146	7	must	must	AUX
bjmsr-315	146	8	be	be	AUX
bjmsr-315	146	9	computed	compute	VERB
bjmsr-315	146	10	.	.	PUNCT
bjmsr-315	147	1	if	if	SCONJ
bjmsr-315	147	2	this	this	DET
bjmsr-315	147	3	alteration	alteration	NOUN
bjmsr-315	147	4	causes	cause	VERB
bjmsr-315	147	5	k2	k2	PROPN
bjmsr-315	147	6	to	to	PART
bjmsr-315	147	7	change	change	VERB
bjmsr-315	147	8	sign	sign	NOUN
bjmsr-315	147	9	,	,	PUNCT
bjmsr-315	147	10	the	the	DET
bjmsr-315	147	11	solution	solution	NOUN
bjmsr-315	147	12	has	have	AUX
bjmsr-315	147	13	been	be	AUX
bjmsr-315	147	14	obtained	obtain	VERB
bjmsr-315	147	15	,	,	PUNCT
bjmsr-315	147	16	and	and	CCONJ
bjmsr-315	147	17	the	the	DET
bjmsr-315	147	18	algorithm	algorithm	NOUN
bjmsr-315	147	19	stops	stop	VERB
bjmsr-315	147	20	.	.	PUNCT
bjmsr-315	148	1	rao	rao	NOUN
bjmsr-315	148	2	and	and	CCONJ
bjmsr-315	148	3	srinivasan	srinivasan	PROPN
bjmsr-315	148	4	(	(	PUNCT
bjmsr-315	148	5	1972	1972	NUM
bjmsr-315	148	6	)	)	PUNCT
bjmsr-315	148	7	interpret	interpret	VERB
bjmsr-315	148	8	sharpe	sharpe	PROPN
bjmsr-315	148	9	's	's	PART
bjmsr-315	148	10	procedure	procedure	NOUN
bjmsr-315	148	11	as	as	ADP
bjmsr-315	148	12	the	the	DET
bjmsr-315	148	13	solution	solution	NOUN
bjmsr-315	148	14	of	of	ADP
bjmsr-315	148	15	parametric	parametric	ADJ
bjmsr-315	148	16	dual	dual	ADJ
bjmsr-315	148	17	linear	linear	NOUN
bjmsr-315	148	18	programming	programming	NOUN
bjmsr-315	148	19	formulation	formulation	NOUN
bjmsr-315	148	20	of	of	ADP
bjmsr-315	148	21	the	the	DET
bjmsr-315	148	22	problem	problem	NOUN
bjmsr-315	148	23	.	.	PUNCT
bjmsr-315	149	1	they	they	PRON
bjmsr-315	149	2	give	give	VERB
bjmsr-315	149	3	an	an	DET
bjmsr-315	149	4	alternate	alternate	NOUN
bjmsr-315	149	5	and	and	CCONJ
bjmsr-315	149	6	about	about	ADP
bjmsr-315	149	7	equally	equally	ADV
bjmsr-315	149	8	efficient	efficient	ADJ
bjmsr-315	149	9	procedure	procedure	NOUN
bjmsr-315	149	10	for	for	ADP
bjmsr-315	149	11	solving	solve	VERB
bjmsr-315	149	12	the	the	DET
bjmsr-315	149	13	same	same	ADJ
bjmsr-315	149	14	problem	problem	NOUN
bjmsr-315	149	15	.	.	PUNCT
bjmsr-315	150	1	brown	brown	PROPN
bjmsr-315	150	2	(	(	PUNCT
bjmsr-315	150	3	1980	1980	NUM
bjmsr-315	150	4	)	)	PUNCT
bjmsr-315	150	5	gives	give	VERB
bjmsr-315	150	6	a	a	DET
bjmsr-315	150	7	distinct	distinct	ADJ
bjmsr-315	150	8	but	but	CCONJ
bjmsr-315	150	9	similar	similar	ADJ
bjmsr-315	150	10	approach	approach	NOUN
bjmsr-315	150	11	to	to	ADP
bjmsr-315	150	12	those	those	PRON
bjmsr-315	150	13	of	of	ADP
bjmsr-315	150	14	edgeworth	edgeworth	PROPN
bjmsr-315	150	15	(	(	PUNCT
bjmsr-315	150	16	1923	1923	NUM
bjmsr-315	150	17	)	)	PUNCT
bjmsr-315	150	18	and	and	CCONJ
bjmsr-315	150	19	sharpe	sharpe	PROPN
bjmsr-315	150	20	(	(	PUNCT
bjmsr-315	150	21	1971	1971	NUM
bjmsr-315	150	22	)	)	PUNCT
bjmsr-315	150	23	.	.	PUNCT
bjmsr-315	151	1	he	he	PRON
bjmsr-315	151	2	emphasizes	emphasize	VERB
bjmsr-315	151	3	on	on	ADP
bjmsr-315	151	4	the	the	DET
bjmsr-315	151	5	median	median	ADJ
bjmsr-315	151	6	properties	property	NOUN
bjmsr-315	151	7	of	of	ADP
bjmsr-315	151	8	the	the	DET
bjmsr-315	151	9	estimator	estimator	NOUN
bjmsr-315	151	10	.	.	PUNCT
bjmsr-315	152	1	the	the	DET
bjmsr-315	152	2	similarity	similarity	NOUN
bjmsr-315	152	3	comes	come	VERB
bjmsr-315	152	4	from	from	ADP
bjmsr-315	152	5	the	the	DET
bjmsr-315	152	6	graphical	graphical	ADJ
bjmsr-315	152	7	approach	approach	NOUN
bjmsr-315	152	8	of	of	ADP
bjmsr-315	152	9	the	the	DET
bjmsr-315	152	10	three	three	NUM
bjmsr-315	152	11	authors	author	NOUN
bjmsr-315	152	12	.	.	PUNCT
bjmsr-315	153	1	kawara	kawara	NOUN
bjmsr-315	153	2	(	(	PUNCT
bjmsr-315	153	3	1979	1979	NUM
bjmsr-315	153	4	)	)	PUNCT
bjmsr-315	153	5	also	also	ADV
bjmsr-315	153	6	develops	develop	VERB
bjmsr-315	153	7	a	a	DET
bjmsr-315	153	8	graphical	graphical	ADJ
bjmsr-315	153	9	method	method	NOUN
bjmsr-315	153	10	for	for	ADP
bjmsr-315	153	11	the	the	DET
bjmsr-315	153	12	simple	simple	ADJ
bjmsr-315	153	13	regression	regression	NOUN
bjmsr-315	153	14	model	model	NOUN
bjmsr-315	153	15	.	.	PUNCT
bjmsr-315	154	1	bartels	bartel	NOUN
bjmsr-315	154	2	and	and	CCONJ
bjmsr-315	154	3	conn	conn	PROPN
bjmsr-315	154	4	and	and	CCONJ
bjmsr-315	154	5	sinclair	sinclair	PROPN
bjmsr-315	154	6	(	(	PUNCT
bjmsr-315	154	7	1978	1978	NUM
bjmsr-315	154	8	)	)	PUNCT
bjmsr-315	154	9	apply	apply	VERB
bjmsr-315	154	10	the	the	DET
bjmsr-315	154	11	method	method	NOUN
bjmsr-315	154	12	of	of	ADP
bjmsr-315	154	13	conn	conn	PROPN
bjmsr-315	154	14	(	(	PUNCT
bjmsr-315	154	15	1976	1976	NUM
bjmsr-315	154	16	)	)	PUNCT
bjmsr-315	154	17	to	to	ADP
bjmsr-315	154	18	the	the	DET
bjmsr-315	154	19	l1	l1	PROPN
bjmsr-315	154	20	norm	norm	NOUN
bjmsr-315	154	21	solution	solution	NOUN
bjmsr-315	154	22	of	of	ADP
bjmsr-315	154	23	an	an	DET
bjmsr-315	154	24	overdetermined	overdetermine	VERB
bjmsr-315	154	25	linear	linear	NOUN
bjmsr-315	154	26	system	system	NOUN
bjmsr-315	154	27	.	.	PUNCT
bjmsr-315	155	1	their	their	PRON
bjmsr-315	155	2	approach	approach	NOUN
bjmsr-315	155	3	is	be	AUX
bjmsr-315	155	4	a	a	DET
bjmsr-315	155	5	minimization	minimization	NOUN
bjmsr-315	155	6	technique	technique	NOUN
bjmsr-315	155	7	for	for	ADP
bjmsr-315	155	8	piecewise	piecewise	NOUN
bjmsr-315	155	9	differentiable	differentiable	ADJ
bjmsr-315	155	10	functions	function	NOUN
bjmsr-315	155	11	.	.	PUNCT
bjmsr-315	156	1	the	the	DET
bjmsr-315	156	2	algorithm	algorithm	NOUN
bjmsr-315	156	3	may	may	AUX
bjmsr-315	156	4	be	be	AUX
bjmsr-315	156	5	reduced	reduce	VERB
bjmsr-315	156	6	as	as	SCONJ
bjmsr-315	156	7	follows	follow	VERB
bjmsr-315	156	8	.	.	PUNCT
bjmsr-315	157	1	step	step	NOUN
bjmsr-315	157	2	0	0	NUM
bjmsr-315	157	3	)	)	PUNCT
bjmsr-315	157	4	select	select	VERB
bjmsr-315	157	5	an	an	DET
bjmsr-315	157	6	arbitrary	arbitrary	ADJ
bjmsr-315	157	7	point	point	NOUN
bjmsr-315	157	8	ß	ß	X
bjmsr-315	157	9	.	.	NOUN
bjmsr-315	158	1	step	step	NOUN
bjmsr-315	158	2	1	1	NUM
bjmsr-315	158	3	)	)	PUNCT
bjmsr-315	158	4	i	i	PRON
bjmsr-315	158	5	)	)	PUNCT
bjmsr-315	158	6	identify	identify	VERB
bjmsr-315	158	7	the	the	DET
bjmsr-315	158	8	index	index	NOUN
bjmsr-315	158	9	set	set	NOUN
bjmsr-315	158	10	,	,	PUNCT
bjmsr-315	158	11	i	i	PRON
bjmsr-315	158	12	=	=	X
bjmsr-315	158	13	{	{	PUNCT
bjmsr-315	158	14	i1,	i1,	INTJ
bjmsr-315	158	15	...	...	PUNCT
bjmsr-315	158	16	,im	,im	NOUN
bjmsr-315	158	17	}	}	PUNCT
bjmsr-315	158	18	=	=	SYM
bjmsr-315	158	19	{	{	PUNCT
bjmsr-315	158	20	i	i	PRON
bjmsr-315	158	21	│	│	NOUN
bjmsr-315	158	22	xi	xi	PROPN
bjmsr-315	158	23	tß	tß	NOUN
bjmsr-315	158	24	-	-	PUNCT
bjmsr-315	158	25	yi=0	yi=0	NOUN
bjmsr-315	158	26	}	}	PUNCT
bjmsr-315	158	27	.	.	PUNCT
bjmsr-315	159	1	let	let	VERB
bjmsr-315	159	2	ith	ith	NOUN
bjmsr-315	159	3	row	row	NOUN
bjmsr-315	159	4	of	of	ADP
bjmsr-315	159	5	x	x	PUNCT
bjmsr-315	159	6	be	be	AUX
bjmsr-315	159	7	xi=[xi	xi=[xi	NOUN
bjmsr-315	159	8	,	,	PUNCT
bjmsr-315	159	9	...	...	PUNCT
bjmsr-315	159	10	,	,	PUNCT
bjmsr-315	159	11	xi	xi	X
bjmsr-315	159	12	]	]	PUNCT
bjmsr-315	159	13	where	where	SCONJ
bjmsr-315	159	14	ijεi	ijεi	NOUN
bjmsr-315	159	15	and	and	CCONJ
bjmsr-315	159	16	the	the	DET
bjmsr-315	159	17	nullspace	nullspace	NOUN
bjmsr-315	159	18	n	n	CCONJ
bjmsr-315	159	19	=	=	SYM
bjmsr-315	159	20	n(xi	n(xi	PROPN
bjmsr-315	159	21	t)={δ	t)={δ	PROPN
bjmsr-315	159	22	│	│	X
bjmsr-315	159	23	xiδ=0,iεi	xiδ=0,iεi	NUM
bjmsr-315	159	24	}	}	PUNCT
bjmsr-315	159	25	.	.	PUNCT
bjmsr-315	160	1	1	1	NUM
bjmsr-315	160	2	m	m	NOUN
bjmsr-315	160	3	let	let	VERB
bjmsr-315	160	4	the	the	DET
bjmsr-315	160	5	orthogonal	orthogonal	ADJ
bjmsr-315	160	6	projector	projector	NOUN
bjmsr-315	160	7	onto	onto	ADP
bjmsr-315	160	8	n	n	ADV
bjmsr-315	160	9	be	be	AUX
bjmsr-315	160	10	denoted	denote	VERB
bjmsr-315	160	11	by	by	ADP
bjmsr-315	160	12	pn	pn	PROPN
bjmsr-315	160	13	.	.	PROPN
bjmsr-315	160	14	ii	ii	PROPN
bjmsr-315	160	15	)	)	PUNCT
bjmsr-315	160	16	let	let	VERB
bjmsr-315	160	17	ic	ic	PRON
bjmsr-315	160	18	be	be	AUX
bjmsr-315	160	19	the	the	DET
bjmsr-315	160	20	complement	complement	NOUN
bjmsr-315	160	21	of	of	ADP
bjmsr-315	160	22	the	the	DET
bjmsr-315	160	23	set	set	NOUN
bjmsr-315	160	24	i.	i.	PROPN
bjmsr-315	160	25	compute	compute	VERB
bjmsr-315	161	1	the	the	DET
bjmsr-315	161	2	vector	vector	NOUN
bjmsr-315	161	3	h	h	NOUN
bjmsr-315	161	4	=	=	SYM
bjmsr-315	161	5	σ	σ	PROPN
bjmsr-315	161	6	sgn(xi	sgn(xi	VERB
bjmsr-315	161	7	tß	tß	PROPN
bjmsr-315	161	8	-	-	PUNCT
bjmsr-315	161	9	yi)xi	yi)xi	NOUN
bjmsr-315	161	10	.	.	PUNCT
bjmsr-315	162	1	iεic	iεic	PROPN
bjmsr-315	162	2	iii	iii	PROPN
bjmsr-315	162	3	)	)	PUNCT
bjmsr-315	162	4	compute	compute	NOUN
bjmsr-315	162	5	p=-pnh	p=-pnh	NOUN
bjmsr-315	162	6	that	that	PRON
bjmsr-315	162	7	is	be	AUX
bjmsr-315	162	8	the	the	DET
bjmsr-315	162	9	projection	projection	NOUN
bjmsr-315	162	10	of	of	ADP
bjmsr-315	162	11	-h	-h	PUNCT
bjmsr-315	162	12	onto	onto	ADP
bjmsr-315	162	13	the	the	DET
bjmsr-315	162	14	nullspace	nullspace	NOUN
bjmsr-315	162	15	of	of	ADP
bjmsr-315	162	16	xi	xi	PROPN
bjmsr-315	162	17	,	,	PUNCT
bjmsr-315	162	18	as	as	ADV
bjmsr-315	162	19	long	long	ADV
bjmsr-315	162	20	as	as	SCONJ
bjmsr-315	162	21	this	this	DET
bjmsr-315	162	22	projection	projection	NOUN
bjmsr-315	162	23	is	be	AUX
bjmsr-315	162	24	non	non	ADJ
bjmsr-315	162	25	zero	zero	NUM
bjmsr-315	162	26	.	.	PUNCT
bjmsr-315	163	1	if	if	SCONJ
bjmsr-315	163	2	p≠0	p≠0	NOUN
bjmsr-315	163	3	,	,	PUNCT
bjmsr-315	163	4	let	let	VERB
bjmsr-315	163	5	g	g	PRON
bjmsr-315	163	6	=	=	NOUN
bjmsr-315	163	7	h	h	NOUN
bjmsr-315	163	8	and	and	CCONJ
bjmsr-315	163	9	go	go	VERB
bjmsr-315	163	10	to	to	PART
bjmsr-315	163	11	step	step	VERB
bjmsr-315	163	12	2	2	NUM
bjmsr-315	163	13	.	.	PUNCT
bjmsr-315	163	14	m	m	VERB
bjmsr-315	163	15	iv	iv	X
bjmsr-315	163	16	)	)	PUNCT
bjmsr-315	163	17	compute	compute	PROPN
bjmsr-315	163	18	w	w	NOUN
bjmsr-315	163	19	according	accord	VERB
bjmsr-315	163	20	to	to	ADP
bjmsr-315	163	21	h	h	NOUN
bjmsr-315	163	22	=	=	VERB
bjmsr-315	163	23	xiw=	xiw=	PROPN
bjmsr-315	163	24	σ	σ	PROPN
bjmsr-315	163	25	wjxi	wjxi	PROPN
bjmsr-315	163	26	,	,	PUNCT
bjmsr-315	163	27	ijεi	ijεi	NOUN
bjmsr-315	163	28	.	.	PUNCT
bjmsr-315	164	1	j=1	j=1	PROPN
bjmsr-315	164	2	j	j	PROPN
bjmsr-315	164	3	v	v	NOUN
bjmsr-315	164	4	)	)	PUNCT
bjmsr-315	164	5	if	if	SCONJ
bjmsr-315	164	6	│	│	VERB
bjmsr-315	164	7	wj	wj	NOUN
bjmsr-315	164	8	│	│	VERB
bjmsr-315	164	9	≤1	≤1	PROPN
bjmsr-315	164	10	for	for	ADP
bjmsr-315	164	11	all	all	DET
bjmsr-315	164	12	j=1,	j=1,	NOUN
bjmsr-315	164	13	...	...	PUNCT
bjmsr-315	164	14	,m	,m	PUNCT
bjmsr-315	164	15	stop	stop	VERB
bjmsr-315	164	16	.	.	PUNCT
bjmsr-315	165	1	in	in	ADP
bjmsr-315	165	2	this	this	DET
bjmsr-315	165	3	case	case	NOUN
bjmsr-315	165	4	,	,	PUNCT
bjmsr-315	165	5	ß	ß	PRON
bjmsr-315	165	6	is	be	AUX
bjmsr-315	165	7	optimal	optimal	ADJ
bjmsr-315	165	8	.	.	PUNCT
bjmsr-315	166	1	vi	vi	X
bjmsr-315	166	2	)	)	PUNCT
bjmsr-315	166	3	find	find	VERB
bjmsr-315	166	4	ij0εi	ij0εi	ADJ
bjmsr-315	166	5	such	such	ADJ
bjmsr-315	166	6	that	that	SCONJ
bjmsr-315	166	7	│	│	ADJ
bjmsr-315	166	8	wj0	wj0	NOUN
bjmsr-315	166	9	│	│	NOUN
bjmsr-315	166	10	>1	>1	NOUN
bjmsr-315	166	11	.	.	PUNCT
bjmsr-315	167	1	vii	vii	PROPN
bjmsr-315	167	2	)	)	PUNCT
bjmsr-315	167	3	change	change	NOUN
bjmsr-315	167	4	i	i	PRON
bjmsr-315	167	5	to	to	ADP
bjmsr-315	167	6	i-{ij0	i-{ij0	ADJ
bjmsr-315	167	7	}	}	PUNCT
bjmsr-315	167	8	and	and	CCONJ
bjmsr-315	167	9	make	make	VERB
bjmsr-315	167	10	corresponding	corresponding	ADJ
bjmsr-315	167	11	changes	change	NOUN
bjmsr-315	167	12	to	to	ADP
bjmsr-315	167	13	xi	xi	PROPN
bjmsr-315	167	14	and	and	CCONJ
bjmsr-315	167	15	n.	n.	PROPN
bjmsr-315	167	16	compute	compute	PROPN
bjmsr-315	167	17	p=-sgn(wj0)pnxi	p=-sgn(wj0)pnxi	PROPN
bjmsr-315	167	18	and	and	CCONJ
bjmsr-315	167	19	let	let	VERB
bjmsr-315	167	20	g	g	PRON
bjmsr-315	167	21	=	=	NOUN
bjmsr-315	167	22	h	h	NOUN
bjmsr-315	167	23	-	-	PUNCT
bjmsr-315	167	24	sgn(wj0)xi	sgn(wj0)xi	NOUN
bjmsr-315	167	25	.	.	PUNCT
bjmsr-315	168	1	j0	j0	PROPN
bjmsr-315	168	2	j0	j0	PROPN
bjmsr-315	168	3	step	step	NOUN
bjmsr-315	168	4	2	2	NUM
bjmsr-315	168	5	)	)	PUNCT
bjmsr-315	168	6	determine	determine	VERB
bjmsr-315	168	7	a={αl	a={αl	NOUN
bjmsr-315	168	8	│	│	NOUN
bjmsr-315	168	9	lεic	lεic	NOUN
bjmsr-315	168	10	,	,	PUNCT
bjmsr-315	168	11	αl=(xl	αl=(xl	ADJ
bjmsr-315	168	12	tß	tß	ADP
bjmsr-315	168	13	-	-	PUNCT
bjmsr-315	168	14	yl)/xl	yl)/xl	ADJ
bjmsr-315	168	15	tp	tp	NOUN
bjmsr-315	168	16	,	,	PUNCT
bjmsr-315	168	17	αl>0	αl>0	NOUN
bjmsr-315	168	18	}	}	PUNCT
bjmsr-315	168	19	elements	element	NOUN
bjmsr-315	168	20	and	and	CCONJ
bjmsr-315	168	21	order	order	VERB
bjmsr-315	168	22	them	they	PRON
bjmsr-315	168	23	such	such	ADJ
bjmsr-315	168	24	that	that	SCONJ
bjmsr-315	168	25	0	0	NUM
bjmsr-315	168	26	<	<	X
bjmsr-315	168	27	αl	αl	ADP
bjmsr-315	168	28	<	<	X
bjmsr-315	168	29	αl	αl	ADP
bjmsr-315	168	30	<	<	X
bjmsr-315	168	31	...	...	PUNCT
bjmsr-315	168	32	<	<	X
bjmsr-315	168	33	αl	αl	X
bjmsr-315	168	34	.	.	PUNCT
bjmsr-315	169	1	let	let	VERB
bjmsr-315	169	2	τ=1	τ=1	PROPN
bjmsr-315	169	3	.	.	PROPN
bjmsr-315	169	4	1	1	NUM
bjmsr-315	169	5	2	2	NUM
bjmsr-315	169	6	t	t	NOUN
bjmsr-315	169	7	step	step	NOUN
bjmsr-315	169	8	3	3	NUM
bjmsr-315	169	9	)	)	PUNCT
bjmsr-315	169	10	if	if	SCONJ
bjmsr-315	169	11	ptg≥2sgn(xl	ptg≥2sgn(xl	ADJ
bjmsr-315	169	12	tß	tß	NOUN
bjmsr-315	169	13	-	-	PUNCT
bjmsr-315	169	14	yl	yl	NOUN
bjmsr-315	169	15	)	)	PUNCT
bjmsr-315	169	16	ptxl	ptxl	NOUN
bjmsr-315	169	17	,	,	PUNCT
bjmsr-315	169	18	then	then	ADV
bjmsr-315	169	19	go	go	VERB
bjmsr-315	169	20	to	to	PART
bjmsr-315	169	21	step	step	VERB
bjmsr-315	169	22	5	5	NUM
bjmsr-315	169	23	.	.	PUNCT
bjmsr-315	170	1	τ	τ	PROPN
bjmsr-315	170	2	τ	τ	PROPN
bjmsr-315	170	3	τ	τ	PROPN
bjmsr-315	170	4	step	step	NOUN
bjmsr-315	170	5	4	4	NUM
bjmsr-315	170	6	)	)	PUNCT
bjmsr-315	170	7	change	change	VERB
bjmsr-315	170	8	g	g	NOUN
bjmsr-315	170	9	to	to	ADP
bjmsr-315	170	10	g-2sgn(xl	g-2sgn(xl	ADJ
bjmsr-315	170	11	ß	ß	NOUN
bjmsr-315	170	12	-	-	NOUN
bjmsr-315	170	13	yl	yl	NOUN
bjmsr-315	170	14	)	)	PUNCT
bjmsr-315	170	15	xl	xl	PROPN
bjmsr-315	170	16	and	and	CCONJ
bjmsr-315	170	17	τ	τ	X
bjmsr-315	170	18	to	to	ADP
bjmsr-315	170	19	τ+1	τ+1	PROPN
bjmsr-315	170	20	and	and	CCONJ
bjmsr-315	170	21	go	go	VERB
bjmsr-315	170	22	to	to	PART
bjmsr-315	170	23	step	step	VERB
bjmsr-315	170	24	3	3	NUM
bjmsr-315	170	25	.	.	PUNCT
bjmsr-315	171	1	τ	τ	PROPN
bjmsr-315	171	2	τ	τ	PROPN
bjmsr-315	171	3	τ	τ	PROPN
bjmsr-315	171	4	step	step	NOUN
bjmsr-315	171	5	5	5	NUM
bjmsr-315	171	6	)	)	PUNCT
bjmsr-315	171	7	replace	replace	VERB
bjmsr-315	171	8	ß	ß	NOUN
bjmsr-315	171	9	by	by	ADP
bjmsr-315	171	10	ß+αl	ß+αl	NOUN
bjmsr-315	171	11	p	p	NOUN
bjmsr-315	171	12	and	and	CCONJ
bjmsr-315	171	13	go	go	VERB
bjmsr-315	171	14	to	to	PART
bjmsr-315	171	15	step	step	VERB
bjmsr-315	171	16	1	1	NUM
bjmsr-315	171	17	.	.	PUNCT
bjmsr-315	172	1	τ	τ	PROPN
bjmsr-315	172	2	this	this	DET
bjmsr-315	172	3	algorithm	algorithm	NOUN
bjmsr-315	172	4	has	have	AUX
bjmsr-315	172	5	also	also	ADV
bjmsr-315	172	6	been	be	AUX
bjmsr-315	172	7	modified	modify	VERB
bjmsr-315	172	8	for	for	ADP
bjmsr-315	172	9	the	the	DET
bjmsr-315	172	10	case	case	NOUN
bjmsr-315	172	11	of	of	ADP
bjmsr-315	172	12	degeneracy	degeneracy	NOUN
bjmsr-315	172	13	(	(	PUNCT
bjmsr-315	172	14	see	see	VERB
bjmsr-315	172	15	also	also	ADV
bjmsr-315	172	16	,	,	PUNCT
bjmsr-315	172	17	bartels	bartel	NOUN
bjmsr-315	172	18	and	and	CCONJ
bjmsr-315	172	19	conn	conn	PROPN
bjmsr-315	172	20	and	and	CCONJ
bjmsr-315	172	21	sinclair	sinclair	PROPN
bjmsr-315	172	22	(	(	PUNCT
bjmsr-315	172	23	1976	1976	NUM
bjmsr-315	172	24	)	)	PUNCT
bjmsr-315	172	25	)	)	PUNCT
bjmsr-315	172	26	.	.	PUNCT
bjmsr-315	173	1	bartels	bartel	NOUN
bjmsr-315	173	2	and	and	CCONJ
bjmsr-315	173	3	conn	conn	PROPN
bjmsr-315	173	4	(	(	PUNCT
bjmsr-315	173	5	1977	1977	NUM
bjmsr-315	173	6	)	)	PUNCT
bjmsr-315	173	7	showed	show	VERB
bjmsr-315	173	8	that	that	SCONJ
bjmsr-315	173	9	how	how	SCONJ
bjmsr-315	173	10	l1	l1	PROPN
bjmsr-315	173	11	norm	norm	NOUN
bjmsr-315	173	12	,	,	PUNCT
bjmsr-315	173	13	restricted	restrict	VERB
bjmsr-315	173	14	l1	l1	PROPN
bjmsr-315	173	15	norm	norm	NOUN
bjmsr-315	173	16	,	,	PUNCT
bjmsr-315	173	17	l∞	l∞	NOUN
bjmsr-315	173	18	norm	norm	NOUN
bjmsr-315	173	19	regressions	regression	NOUN
bjmsr-315	173	20	,	,	PUNCT
bjmsr-315	173	21	and	and	CCONJ
bjmsr-315	173	22	general	general	ADJ
bjmsr-315	173	23	linear	linear	NOUN
bjmsr-315	173	24	programming	programming	NOUN
bjmsr-315	173	25	can	can	AUX
bjmsr-315	173	26	all	all	PRON
bjmsr-315	173	27	be	be	AUX
bjmsr-315	173	28	easily	easily	ADV
bjmsr-315	173	29	expressed	express	VERB
bjmsr-315	173	30	as	as	ADP
bjmsr-315	173	31	a	a	DET
bjmsr-315	173	32	piecewise	piecewise	NOUN
bjmsr-315	173	33	linear	linear	PROPN
bjmsr-315	173	34	minimization	minimization	NOUN
bjmsr-315	173	35	problem	problem	NOUN
bjmsr-315	173	36	.	.	PUNCT
bjmsr-315	174	1	let	let	VERB
bjmsr-315	174	2	u	u	PRON
bjmsr-315	174	3	and	and	CCONJ
bjmsr-315	174	4	v	v	NOUN
bjmsr-315	174	5	be	be	AUX
bjmsr-315	174	6	of	of	ADP
bjmsr-315	174	7	sizes	size	NOUN
bjmsr-315	174	8	pxm	pxm	VERB
bjmsr-315	174	9	and	and	CCONJ
bjmsr-315	174	10	px1	px1	NOUN
bjmsr-315	174	11	,	,	PUNCT
bjmsr-315	174	12	respectively	respectively	ADV
bjmsr-315	174	13	.	.	PUNCT
bjmsr-315	175	1	consider	consider	VERB
bjmsr-315	175	2	the	the	DET
bjmsr-315	175	3	function	function	NOUN
bjmsr-315	175	4	:	:	PUNCT
bjmsr-315	175	5	φ(ß	φ(ß	VERB
bjmsr-315	175	6	)	)	PUNCT
bjmsr-315	176	1	=	=	SYM
bjmsr-315	176	2	htß	htß	NOUN
bjmsr-315	176	3	+	+	X
bjmsr-315	176	4	σi	σi	X
bjmsr-315	176	5	│	│	ADJ
bjmsr-315	176	6	yi	yi	PROPN
bjmsr-315	176	7	-	-	PUNCT
bjmsr-315	176	8	xi	xi	NOUN
bjmsr-315	176	9	tß	tß	ADP
bjmsr-315	176	10	│	│	ADJ
bjmsr-315	176	11	+	+	CCONJ
bjmsr-315	176	12	σjmax(0,vj	σjmax(0,vj	PROPN
bjmsr-315	176	13	-	-	PUNCT
bjmsr-315	176	14	uj	uj	NUM
bjmsr-315	176	15	tß	tß	NOUN
bjmsr-315	176	16	)	)	PUNCT
bjmsr-315	176	17	(	(	PUNCT
bjmsr-315	176	18	19	19	NUM
bjmsr-315	176	19	)	)	PUNCT
bjmsr-315	176	20	where	where	SCONJ
bjmsr-315	176	21	yi	yi	PROPN
bjmsr-315	176	22	-	-	PUNCT
bjmsr-315	176	23	xi	xi	NOUN
bjmsr-315	176	24	tß	tß	PROPN
bjmsr-315	176	25	represents	represent	VERB
bjmsr-315	176	26	the	the	DET
bjmsr-315	176	27	ith	ith	PROPN
bjmsr-315	176	28	element	element	NOUN
bjmsr-315	176	29	of	of	ADP
bjmsr-315	176	30	the	the	DET
bjmsr-315	176	31	residual	residual	ADJ
bjmsr-315	176	32	vector	vector	NOUN
bjmsr-315	176	33	y	y	PROPN
bjmsr-315	176	34	-	-	PUNCT
bjmsr-315	176	35	xß	xß	PROPN
bjmsr-315	176	36	and	and	CCONJ
bjmsr-315	176	37	vj	vj	PROPN
bjmsr-315	176	38	-	-	PUNCT
bjmsr-315	176	39	uj	uj	ADJ
bjmsr-315	176	40	tß	tß	NOUN
bjmsr-315	176	41	represents	represent	VERB
bjmsr-315	176	42	the	the	DET
bjmsr-315	176	43	jth	jth	PROPN
bjmsr-315	176	44	element	element	NOUN
bjmsr-315	176	45	of	of	ADP
bjmsr-315	176	46	the	the	DET
bjmsr-315	176	47	residual	residual	ADJ
bjmsr-315	176	48	vector	vector	NOUN
bjmsr-315	176	49	vuß	vuß	NOUN
bjmsr-315	176	50	.	.	PUNCT
bjmsr-315	177	1	to	to	PART
bjmsr-315	177	2	minimize	minimize	VERB
bjmsr-315	177	3	φ	φ	PROPN
bjmsr-315	177	4	with	with	ADP
bjmsr-315	177	5	respect	respect	NOUN
bjmsr-315	177	6	to	to	ADP
bjmsr-315	177	7	ß	ß	PRON
bjmsr-315	177	8	the	the	DET
bjmsr-315	177	9	following	follow	VERB
bjmsr-315	177	10	steps	step	NOUN
bjmsr-315	177	11	should	should	AUX
bjmsr-315	177	12	be	be	AUX
bjmsr-315	177	13	taken	take	VERB
bjmsr-315	177	14	.	.	PUNCT
bjmsr-315	178	1	step	step	NOUN
bjmsr-315	178	2	0	0	NUM
bjmsr-315	178	3	)	)	PUNCT
bjmsr-315	178	4	start	start	VERB
bjmsr-315	178	5	with	with	ADP
bjmsr-315	178	6	arbitrary	arbitrary	ADJ
bjmsr-315	178	7	ß(k	ß(k	NUM
bjmsr-315	178	8	)	)	PUNCT
bjmsr-315	178	9	.	.	PUNCT
bjmsr-315	179	1	step	step	NOUN
bjmsr-315	179	2	1	1	NUM
bjmsr-315	179	3	)	)	PUNCT
bjmsr-315	179	4	find	find	VERB
bjmsr-315	179	5	δ(k	δ(k	NOUN
bjmsr-315	179	6	)	)	PUNCT
bjmsr-315	179	7	so	so	SCONJ
bjmsr-315	179	8	that	that	SCONJ
bjmsr-315	179	9	φ(ß(k)+θδ(k))≤φ(ß(k	φ(ß(k)+θδ(k))≤φ(ß(k	NOUN
bjmsr-315	179	10	)	)	PUNCT
bjmsr-315	179	11	)	)	PUNCT
bjmsr-315	179	12	for	for	ADP
bjmsr-315	179	13	all	all	DET
bjmsr-315	179	14	θ>0	θ>0	ADJ
bjmsr-315	179	15	small	small	ADJ
bjmsr-315	179	16	enough	enough	ADV
bjmsr-315	179	17	.	.	PUNCT
bjmsr-315	180	1	step	step	NOUN
bjmsr-315	180	2	2	2	NUM
bjmsr-315	180	3	)	)	PUNCT
bjmsr-315	180	4	choose	choose	VERB
bjmsr-315	180	5	θ(k)≥0	θ(k)≥0	NOUN
bjmsr-315	180	6	to	to	PART
bjmsr-315	180	7	obtain	obtain	VERB
bjmsr-315	180	8	the	the	DET
bjmsr-315	180	9	largest	large	ADJ
bjmsr-315	180	10	possible	possible	ADJ
bjmsr-315	180	11	decrease	decrease	NOUN
bjmsr-315	180	12	in	in	ADP
bjmsr-315	180	13	φ	φ	PROPN
bjmsr-315	180	14	.	.	PUNCT
bjmsr-315	181	1	copyright	copyright	NOUN
bjmsr-315	181	2	©	©	PROPN
bjmsr-315	181	3	cc	cc	PROPN
bjmsr-315	181	4	-	-	PUNCT
bjmsr-315	181	5	by	by	ADP
bjmsr-315	181	6	-	-	PUNCT
bjmsr-315	181	7	nc	nc	PROPN
bjmsr-315	181	8	2019	2019	NUM
bjmsr-315	181	9	,	,	PUNCT
bjmsr-315	181	10	bjmsr	bjmsr	PROPN
bjmsr-315	181	11	www.cribfb.com/journal/index.php/bjmsr	www.cribfb.com/journal/index.php/bjmsr	X
bjmsr-315	182	1	bangladesh	bangladesh	PROPN
bjmsr-315	182	2	journal	journal	PROPN
bjmsr-315	182	3	of	of	ADP
bjmsr-315	182	4	multidisciplinary	multidisciplinary	ADJ
bjmsr-315	182	5	scientific	scientific	ADJ
bjmsr-315	182	6	research	research	NOUN
bjmsr-315	182	7	vol	vol	NOUN
bjmsr-315	182	8	.	.	PROPN
bjmsr-315	182	9	1	1	NUM
bjmsr-315	182	10	,	,	PUNCT
bjmsr-315	182	11	no	no	INTJ
bjmsr-315	182	12	.	.	NOUN
bjmsr-315	182	13	1	1	NUM
bjmsr-315	182	14	;	;	PUNCT
bjmsr-315	182	15	2019	2019	NUM
bjmsr-315	182	16	55	55	NUM
bjmsr-315	182	17	step	step	NOUN
bjmsr-315	182	18	3	3	NUM
bjmsr-315	182	19	)	)	PUNCT
bjmsr-315	182	20	let	let	VERB
bjmsr-315	182	21	ß(k+1)=ß(k)+θ(k)δ(k	ß(k+1)=ß(k)+θ(k)δ(k	NUM
bjmsr-315	182	22	)	)	PUNCT
bjmsr-315	182	23	.	.	PUNCT
bjmsr-315	183	1	when	when	SCONJ
bjmsr-315	183	2	h=0	h=0	PROPN
bjmsr-315	183	3	and	and	CCONJ
bjmsr-315	183	4	the	the	DET
bjmsr-315	183	5	sum	sum	NOUN
bjmsr-315	183	6	on	on	ADP
bjmsr-315	183	7	j	j	PROPN
bjmsr-315	183	8	is	be	AUX
bjmsr-315	183	9	vacuous	vacuous	ADJ
bjmsr-315	183	10	in	in	ADP
bjmsr-315	183	11	the	the	DET
bjmsr-315	183	12	φ(ß	φ(ß	NOUN
bjmsr-315	183	13	)	)	PUNCT
bjmsr-315	183	14	function	function	NOUN
bjmsr-315	183	15	(	(	PUNCT
bjmsr-315	183	16	19	19	NUM
bjmsr-315	183	17	)	)	PUNCT
bjmsr-315	183	18	,	,	PUNCT
bjmsr-315	183	19	the	the	DET
bjmsr-315	183	20	resulting	result	VERB
bjmsr-315	183	21	simplification	simplification	NOUN
bjmsr-315	183	22	of	of	ADP
bjmsr-315	183	23	the	the	DET
bjmsr-315	183	24	above	above	ADJ
bjmsr-315	183	25	steps	step	NOUN
bjmsr-315	183	26	corresponds	correspond	VERB
bjmsr-315	183	27	precisely	precisely	ADV
bjmsr-315	183	28	to	to	ADP
bjmsr-315	183	29	the	the	DET
bjmsr-315	183	30	algorithm	algorithm	NOUN
bjmsr-315	183	31	proposed	propose	VERB
bjmsr-315	183	32	by	by	ADP
bjmsr-315	183	33	bartels	bartel	NOUN
bjmsr-315	183	34	and	and	CCONJ
bjmsr-315	183	35	conn	conn	PROPN
bjmsr-315	183	36	and	and	CCONJ
bjmsr-315	183	37	sinclair	sinclair	PROPN
bjmsr-315	183	38	(	(	PUNCT
bjmsr-315	183	39	1978	1978	NUM
bjmsr-315	183	40	)	)	PUNCT
bjmsr-315	183	41	.	.	PUNCT
bjmsr-315	184	1	other	other	ADJ
bjmsr-315	184	2	related	related	ADJ
bjmsr-315	184	3	problems	problem	NOUN
bjmsr-315	184	4	can	can	AUX
bjmsr-315	184	5	be	be	AUX
bjmsr-315	184	6	solved	solve	VERB
bjmsr-315	184	7	by	by	ADP
bjmsr-315	184	8	modification	modification	NOUN
bjmsr-315	184	9	of	of	ADP
bjmsr-315	184	10	φ(ß	φ(ß	PROPN
bjmsr-315	184	11	)	)	PUNCT
bjmsr-315	184	12	by	by	ADP
bjmsr-315	184	13	a	a	DET
bjmsr-315	184	14	quantity	quantity	NOUN
bjmsr-315	184	15	µ	µ	NOUN
bjmsr-315	184	16	to	to	PART
bjmsr-315	184	17	obtain	obtain	VERB
bjmsr-315	184	18	a	a	DET
bjmsr-315	184	19	parameterized	parameterized	ADJ
bjmsr-315	184	20	family	family	NOUN
bjmsr-315	184	21	of	of	ADP
bjmsr-315	184	22	piecewise	piecewise	NOUN
bjmsr-315	184	23	linear	linear	NOUN
bjmsr-315	184	24	functions	function	NOUN
bjmsr-315	184	25	φµ(ß	φµ(ß	ADJ
bjmsr-315	184	26	)	)	PUNCT
bjmsr-315	184	27	,	,	PUNCT
bjmsr-315	184	28	and	and	CCONJ
bjmsr-315	184	29	take	take	VERB
bjmsr-315	184	30	following	follow	VERB
bjmsr-315	184	31	steps	step	NOUN
bjmsr-315	184	32	to	to	PART
bjmsr-315	184	33	find	find	VERB
bjmsr-315	184	34	the	the	DET
bjmsr-315	184	35	minimum	minimum	ADJ
bjmsr-315	184	36	solution	solution	NOUN
bjmsr-315	184	37	.	.	PUNCT
bjmsr-315	185	1	step	step	NOUN
bjmsr-315	185	2	0	0	NUM
bjmsr-315	185	3	)	)	PUNCT
bjmsr-315	185	4	set	set	NOUN
bjmsr-315	185	5	µ>0	µ>0	NOUN
bjmsr-315	185	6	;	;	PUNCT
bjmsr-315	185	7	select	select	VERB
bjmsr-315	185	8	any	any	DET
bjmsr-315	185	9	ß	ß	DET
bjmsr-315	185	10	=	=	NOUN
bjmsr-315	185	11	ß(0	ß(0	NOUN
bjmsr-315	185	12	)	)	PUNCT
bjmsr-315	185	13	.	.	PUNCT
bjmsr-315	186	1	step	step	NOUN
bjmsr-315	186	2	1	1	NUM
bjmsr-315	186	3	)	)	PUNCT
bjmsr-315	186	4	minimize	minimize	VERB
bjmsr-315	186	5	φµ(ß	φµ(ß	NOUN
bjmsr-315	186	6	)	)	PUNCT
bjmsr-315	186	7	with	with	ADP
bjmsr-315	186	8	respect	respect	NOUN
bjmsr-315	186	9	to	to	ADP
bjmsr-315	186	10	ß	ß	NOUN
bjmsr-315	186	11	according	accord	VERB
bjmsr-315	186	12	to	to	ADP
bjmsr-315	186	13	the	the	DET
bjmsr-315	186	14	above	above	ADJ
bjmsr-315	186	15	procedure	procedure	NOUN
bjmsr-315	186	16	.	.	PUNCT
bjmsr-315	187	1	step	step	NOUN
bjmsr-315	187	2	2	2	NUM
bjmsr-315	187	3	)	)	PUNCT
bjmsr-315	187	4	stop	stop	VERB
bjmsr-315	187	5	if	if	SCONJ
bjmsr-315	187	6	a	a	DET
bjmsr-315	187	7	prescribed	prescribed	ADJ
bjmsr-315	187	8	terminating	terminate	VERB
bjmsr-315	187	9	condition	condition	NOUN
bjmsr-315	187	10	on	on	ADP
bjmsr-315	187	11	ß	ß	PRON
bjmsr-315	187	12	is	be	AUX
bjmsr-315	187	13	met	meet	VERB
bjmsr-315	187	14	;	;	PUNCT
bjmsr-315	187	15	otherwise	otherwise	ADV
bjmsr-315	187	16	set	set	VERB
bjmsr-315	187	17	µ=µ/10	µ=µ/10	ADJ
bjmsr-315	187	18	and	and	CCONJ
bjmsr-315	187	19	go	go	VERB
bjmsr-315	187	20	to	to	PART
bjmsr-315	187	21	step	step	VERB
bjmsr-315	187	22	1	1	NUM
bjmsr-315	187	23	.	.	PUNCT
bjmsr-315	188	1	the	the	DET
bjmsr-315	188	2	contribution	contribution	NOUN
bjmsr-315	188	3	of	of	ADP
bjmsr-315	188	4	this	this	DET
bjmsr-315	188	5	paper	paper	NOUN
bjmsr-315	188	6	is	be	AUX
bjmsr-315	188	7	putting	put	VERB
bjmsr-315	188	8	a	a	DET
bjmsr-315	188	9	wide	wide	ADJ
bjmsr-315	188	10	class	class	NOUN
bjmsr-315	188	11	of	of	ADP
bjmsr-315	188	12	problems	problem	NOUN
bjmsr-315	188	13	in	in	ADP
bjmsr-315	188	14	the	the	DET
bjmsr-315	188	15	mould	mould	NOUN
bjmsr-315	188	16	of	of	ADP
bjmsr-315	188	17	two	two	NUM
bjmsr-315	188	18	algorithms	algorithm	NOUN
bjmsr-315	188	19	mentioned	mention	VERB
bjmsr-315	188	20	above	above	ADV
bjmsr-315	188	21	.	.	PUNCT
bjmsr-315	189	1	the	the	DET
bjmsr-315	189	2	techniques	technique	NOUN
bjmsr-315	189	3	are	be	AUX
bjmsr-315	189	4	easily	easily	ADV
bjmsr-315	189	5	extended	extend	VERB
bjmsr-315	189	6	to	to	ADP
bjmsr-315	189	7	the	the	DET
bjmsr-315	189	8	models	model	NOUN
bjmsr-315	189	9	with	with	ADP
bjmsr-315	189	10	norm	norm	NOUN
bjmsr-315	189	11	restrictions	restriction	NOUN
bjmsr-315	189	12	.	.	PUNCT
bjmsr-315	190	1	bloomfield	bloomfield	PROPN
bjmsr-315	190	2	and	and	CCONJ
bjmsr-315	190	3	steiger	steiger	PROPN
bjmsr-315	190	4	(	(	PUNCT
bjmsr-315	190	5	1980	1980	NUM
bjmsr-315	190	6	)	)	PUNCT
bjmsr-315	190	7	proposed	propose	VERB
bjmsr-315	190	8	a	a	DET
bjmsr-315	190	9	descent	descent	NOUN
bjmsr-315	190	10	method	method	NOUN
bjmsr-315	190	11	for	for	SCONJ
bjmsr-315	190	12	the	the	DET
bjmsr-315	190	13	l1	l1	PROPN
bjmsr-315	190	14	norm	norm	NOUN
bjmsr-315	190	15	multiple	multiple	ADJ
bjmsr-315	190	16	regression	regression	NOUN
bjmsr-315	190	17	.	.	PUNCT
bjmsr-315	191	1	their	their	PRON
bjmsr-315	191	2	algorithm	algorithm	NOUN
bjmsr-315	191	3	is	be	AUX
bjmsr-315	191	4	also	also	ADV
bjmsr-315	191	5	explained	explain	VERB
bjmsr-315	191	6	in	in	ADP
bjmsr-315	191	7	bloomfield	bloomfield	PROPN
bjmsr-315	191	8	and	and	CCONJ
bjmsr-315	191	9	steiger	steiger	PROPN
bjmsr-315	191	10	(	(	PUNCT
bjmsr-315	191	11	1983	1983	NUM
bjmsr-315	191	12	)	)	PUNCT
bjmsr-315	191	13	.	.	PUNCT
bjmsr-315	192	1	in	in	ADP
bjmsr-315	192	2	some	some	DET
bjmsr-315	192	3	steps	step	NOUN
bjmsr-315	192	4	,	,	PUNCT
bjmsr-315	192	5	this	this	DET
bjmsr-315	192	6	algorithm	algorithm	NOUN
bjmsr-315	192	7	is	be	AUX
bjmsr-315	192	8	related	relate	VERB
bjmsr-315	192	9	to	to	ADP
bjmsr-315	192	10	that	that	PRON
bjmsr-315	192	11	of	of	ADP
bjmsr-315	192	12	singleton	singleton	PROPN
bjmsr-315	192	13	(	(	PUNCT
bjmsr-315	192	14	1940	1940	NUM
bjmsr-315	192	15	)	)	PUNCT
bjmsr-315	192	16	and	and	CCONJ
bjmsr-315	192	17	usow	usow	NOUN
bjmsr-315	192	18	(	(	PUNCT
bjmsr-315	192	19	1967b	1967b	NUM
bjmsr-315	192	20	)	)	PUNCT
bjmsr-315	192	21	.	.	PUNCT
bjmsr-315	193	1	the	the	DET
bjmsr-315	193	2	basis	basis	NOUN
bjmsr-315	193	3	of	of	ADP
bjmsr-315	193	4	this	this	DET
bjmsr-315	193	5	method	method	NOUN
bjmsr-315	193	6	is	be	AUX
bjmsr-315	193	7	to	to	PART
bjmsr-315	193	8	search	search	VERB
bjmsr-315	193	9	for	for	ADP
bjmsr-315	193	10	a	a	DET
bjmsr-315	193	11	set	set	NOUN
bjmsr-315	193	12	of	of	ADP
bjmsr-315	193	13	m	m	PROPN
bjmsr-315	193	14	observations	observation	NOUN
bjmsr-315	193	15	which	which	PRON
bjmsr-315	193	16	locate	locate	VERB
bjmsr-315	193	17	on	on	ADP
bjmsr-315	193	18	the	the	DET
bjmsr-315	193	19	optimal	optimal	ADJ
bjmsr-315	193	20	l1	l1	PROPN
bjmsr-315	193	21	norm	norm	NOUN
bjmsr-315	193	22	regression	regression	NOUN
bjmsr-315	193	23	.	.	PUNCT
bjmsr-315	194	1	this	this	DET
bjmsr-315	194	2	set	set	NOUN
bjmsr-315	194	3	is	be	AUX
bjmsr-315	194	4	found	find	VERB
bjmsr-315	194	5	iteratively	iteratively	ADV
bjmsr-315	194	6	by	by	ADP
bjmsr-315	194	7	successive	successive	ADJ
bjmsr-315	194	8	improvement	improvement	NOUN
bjmsr-315	194	9	.	.	PUNCT
bjmsr-315	195	1	in	in	ADP
bjmsr-315	195	2	each	each	DET
bjmsr-315	195	3	iteration	iteration	NOUN
bjmsr-315	195	4	,	,	PUNCT
bjmsr-315	195	5	one	one	NUM
bjmsr-315	195	6	point	point	NOUN
bjmsr-315	195	7	from	from	ADP
bjmsr-315	195	8	the	the	DET
bjmsr-315	195	9	current	current	ADJ
bjmsr-315	195	10	set	set	NOUN
bjmsr-315	195	11	is	be	AUX
bjmsr-315	195	12	identified	identify	VERB
bjmsr-315	195	13	as	as	ADP
bjmsr-315	195	14	a	a	DET
bjmsr-315	195	15	good	good	ADJ
bjmsr-315	195	16	prospect	prospect	NOUN
bjmsr-315	195	17	for	for	ADP
bjmsr-315	195	18	deletion	deletion	NOUN
bjmsr-315	195	19	.	.	PUNCT
bjmsr-315	196	1	this	this	DET
bjmsr-315	196	2	point	point	NOUN
bjmsr-315	196	3	is	be	AUX
bjmsr-315	196	4	then	then	ADV
bjmsr-315	196	5	replaced	replace	VERB
bjmsr-315	196	6	by	by	ADP
bjmsr-315	196	7	the	the	DET
bjmsr-315	196	8	best	good	ADJ
bjmsr-315	196	9	alternative	alternative	NOUN
bjmsr-315	196	10	.	.	PUNCT
bjmsr-315	197	1	the	the	DET
bjmsr-315	197	2	novel	novel	ADJ
bjmsr-315	197	3	features	feature	NOUN
bjmsr-315	197	4	of	of	ADP
bjmsr-315	197	5	this	this	DET
bjmsr-315	197	6	method	method	NOUN
bjmsr-315	197	7	are	be	AUX
bjmsr-315	197	8	in	in	ADP
bjmsr-315	197	9	an	an	DET
bjmsr-315	197	10	efficient	efficient	ADJ
bjmsr-315	197	11	procedure	procedure	NOUN
bjmsr-315	197	12	for	for	ADP
bjmsr-315	197	13	finding	find	VERB
bjmsr-315	197	14	the	the	DET
bjmsr-315	197	15	optimal	optimal	ADJ
bjmsr-315	197	16	replacement	replacement	NOUN
bjmsr-315	197	17	and	and	CCONJ
bjmsr-315	197	18	a	a	DET
bjmsr-315	197	19	heuristic	heuristic	ADJ
bjmsr-315	197	20	method	method	NOUN
bjmsr-315	197	21	for	for	ADP
bjmsr-315	197	22	identifying	identify	VERB
bjmsr-315	197	23	the	the	DET
bjmsr-315	197	24	point	point	NOUN
bjmsr-315	197	25	to	to	PART
bjmsr-315	197	26	be	be	AUX
bjmsr-315	197	27	deleted	delete	VERB
bjmsr-315	197	28	from	from	ADP
bjmsr-315	197	29	the	the	DET
bjmsr-315	197	30	pivot	pivot	NOUN
bjmsr-315	197	31	.	.	PUNCT
bjmsr-315	198	1	denote	denote	VERB
bjmsr-315	198	2	the	the	DET
bjmsr-315	198	3	x1	x1	PROPN
bjmsr-315	198	4	t,	t,	NUM
bjmsr-315	198	5	...	...	PUNCT
bjmsr-315	198	6	,xm	,xm	PUNCT
bjmsr-315	198	7	t	t	PROPN
bjmsr-315	198	8	as	as	ADP
bjmsr-315	198	9	rows	row	NOUN
bjmsr-315	198	10	of	of	ADP
bjmsr-315	198	11	the	the	DET
bjmsr-315	198	12	independent	independent	ADJ
bjmsr-315	198	13	variables	variable	NOUN
bjmsr-315	198	14	design	design	NOUN
bjmsr-315	198	15	matrix	matrix	NOUN
bjmsr-315	198	16	which	which	PRON
bjmsr-315	198	17	correspond	correspond	VERB
bjmsr-315	198	18	to	to	ADP
bjmsr-315	198	19	the	the	DET
bjmsr-315	198	20	current	current	ADJ
bjmsr-315	198	21	set	set	NOUN
bjmsr-315	198	22	of	of	ADP
bjmsr-315	198	23	points	point	NOUN
bjmsr-315	198	24	1,	1,	NUM
bjmsr-315	198	25	...	...	PUNCT
bjmsr-315	198	26	,m	,m	PUNCT
bjmsr-315	198	27	;	;	PUNCT
bjmsr-315	198	28	and	and	CCONJ
bjmsr-315	198	29	xm	xm	PROPN
bjmsr-315	198	30	t	t	PROPN
bjmsr-315	198	31	for	for	ADP
bjmsr-315	198	32	replacement	replacement	NOUN
bjmsr-315	198	33	.	.	PUNCT
bjmsr-315	199	1	set	set	NOUN
bjmsr-315	199	2	,	,	PUNCT
bjmsr-315	199	3	yi	yi	PROPN
bjmsr-315	200	1	=	=	PRON
bjmsr-315	200	2	xi	xi	PROPN
bjmsr-315	200	3	tß	tß	PROPN
bjmsr-315	200	4	i=1,	i=1,	PROPN
bjmsr-315	200	5	...	...	PUNCT
bjmsr-315	200	6	,m-1	,m-1	PUNCT
bjmsr-315	200	7	(	(	PUNCT
bjmsr-315	200	8	20	20	NUM
bjmsr-315	200	9	)	)	PUNCT
bjmsr-315	200	10	redefine	redefine	VERB
bjmsr-315	200	11	ß	ß	PRON
bjmsr-315	200	12	as	as	ADP
bjmsr-315	200	13	,	,	PUNCT
bjmsr-315	200	14	ß	ß	NOUN
bjmsr-315	200	15	=	=	SYM
bjmsr-315	200	16	ß0	ß0	NOUN
bjmsr-315	201	1	+	+	CCONJ
bjmsr-315	201	2	tδ	tδ	PROPN
bjmsr-315	201	3	(	(	PUNCT
bjmsr-315	201	4	21	21	NUM
bjmsr-315	201	5	)	)	PUNCT
bjmsr-315	201	6	where	where	SCONJ
bjmsr-315	201	7	ß0	ß0	NOUN
bjmsr-315	201	8	is	be	AUX
bjmsr-315	201	9	an	an	DET
bjmsr-315	201	10	arbitrary	arbitrary	ADJ
bjmsr-315	201	11	member	member	NOUN
bjmsr-315	201	12	of	of	ADP
bjmsr-315	201	13	the	the	DET
bjmsr-315	201	14	set	set	NOUN
bjmsr-315	201	15	and	and	CCONJ
bjmsr-315	201	16	δ	δ	PROPN
bjmsr-315	201	17	vector	vector	NOUN
bjmsr-315	201	18	obeys	obey	VERB
bjmsr-315	201	19	the	the	DET
bjmsr-315	201	20	following	follow	VERB
bjmsr-315	201	21	system	system	NOUN
bjmsr-315	201	22	,	,	PUNCT
bjmsr-315	201	23	xi	xi	X
bjmsr-315	201	24	tδ	tδ	NOUN
bjmsr-315	201	25	=	=	SYM
bjmsr-315	201	26	0	0	NUM
bjmsr-315	201	27	i=1,	i=1,	NOUN
bjmsr-315	201	28	...	...	PUNCT
bjmsr-315	201	29	,m-1	,m-1	PUNCT
bjmsr-315	201	30	(	(	PUNCT
bjmsr-315	201	31	22	22	NUM
bjmsr-315	201	32	)	)	PUNCT
bjmsr-315	201	33	given	give	VERB
bjmsr-315	201	34	this	this	DET
bjmsr-315	201	35	set	set	NOUN
bjmsr-315	201	36	of	of	ADP
bjmsr-315	201	37	points	point	NOUN
bjmsr-315	201	38	,	,	PUNCT
bjmsr-315	201	39	the	the	DET
bjmsr-315	201	40	optimum	optimum	ADJ
bjmsr-315	201	41	value	value	NOUN
bjmsr-315	201	42	of	of	ADP
bjmsr-315	201	43	s	s	PRON
bjmsr-315	201	44	may	may	AUX
bjmsr-315	201	45	be	be	AUX
bjmsr-315	201	46	found	find	VERB
bjmsr-315	201	47	by	by	ADP
bjmsr-315	201	48	minimizing	minimize	VERB
bjmsr-315	201	49	the	the	DET
bjmsr-315	201	50	following	follow	VERB
bjmsr-315	201	51	expression	expression	NOUN
bjmsr-315	201	52	with	with	ADP
bjmsr-315	201	53	respect	respect	NOUN
bjmsr-315	201	54	to	to	ADP
bjmsr-315	201	55	the	the	DET
bjmsr-315	201	56	scalar	scalar	ADJ
bjmsr-315	201	57	t.	t.	PROPN
bjmsr-315	201	58	n	n	PROPN
bjmsr-315	201	59	σ	σ	PROPN
bjmsr-315	201	60	│	│	NOUN
bjmsr-315	201	61	yi	yi	PROPN
bjmsr-315	201	62	xi	xi	NUM
bjmsr-315	201	63	t(ß0	t(ß0	PROPN
bjmsr-315	201	64	+	+	CCONJ
bjmsr-315	201	65	tδ	tδ	NOUN
bjmsr-315	201	66	)	)	PUNCT
bjmsr-315	201	67	│	│	X
bjmsr-315	201	68	(	(	PUNCT
bjmsr-315	201	69	23	23	NUM
bjmsr-315	201	70	)	)	PUNCT
bjmsr-315	201	71	i=1	i=1	VERB
bjmsr-315	201	72	rearranging	rearrange	VERB
bjmsr-315	201	73	the	the	DET
bjmsr-315	201	74	terms	term	NOUN
bjmsr-315	201	75	leads	lead	VERB
bjmsr-315	201	76	to	to	ADP
bjmsr-315	201	77	:	:	PUNCT
bjmsr-315	201	78	n	n	PROPN
bjmsr-315	201	79	σ	σ	PROPN
bjmsr-315	201	80	│	│	NOUN
bjmsr-315	201	81	wi	wi	PROPN
bjmsr-315	201	82	│	│	VERB
bjmsr-315	201	83	│	│	NUM
bjmsr-315	201	84	ri	ri	PROPN
bjmsr-315	201	85	t	t	PROPN
bjmsr-315	201	86	│	│	X
bjmsr-315	201	87	(	(	PUNCT
bjmsr-315	201	88	24	24	NUM
bjmsr-315	201	89	)	)	PUNCT
bjmsr-315	201	90	i=1	i=1	PROPN
bjmsr-315	201	91	where	where	SCONJ
bjmsr-315	201	92	wi	wi	PROPN
bjmsr-315	201	93	=	=	PROPN
bjmsr-315	201	94	xi	xi	ADP
bjmsr-315	201	95	tδ	tδ	PROPN
bjmsr-315	201	96	and	and	CCONJ
bjmsr-315	201	97	ri=(yi	ri=(yi	PROPN
bjmsr-315	201	98	-	-	PUNCT
bjmsr-315	201	99	xi	xi	ADP
bjmsr-315	201	100	tß0)/(xi	tß0)/(xi	NUM
bjmsr-315	201	101	tδ	tδ	PROPN
bjmsr-315	201	102	)	)	PUNCT
bjmsr-315	201	103	.	.	PUNCT
bjmsr-315	202	1	value	value	NOUN
bjmsr-315	202	2	of	of	ADP
bjmsr-315	202	3	t	t	PROPN
bjmsr-315	202	4	may	may	AUX
bjmsr-315	202	5	be	be	AUX
bjmsr-315	202	6	found	find	VERB
bjmsr-315	202	7	by	by	ADP
bjmsr-315	202	8	laplace	laplace	NOUN
bjmsr-315	202	9	's	's	PART
bjmsr-315	202	10	weighted	weight	VERB
bjmsr-315	202	11	median	median	ADJ
bjmsr-315	202	12	method	method	NOUN
bjmsr-315	202	13	.	.	PUNCT
bjmsr-315	203	1	bloomfield	bloomfield	PROPN
bjmsr-315	203	2	and	and	CCONJ
bjmsr-315	203	3	steiger	steiger	PROPN
bjmsr-315	203	4	propose	propose	VERB
bjmsr-315	203	5	a	a	DET
bjmsr-315	203	6	weighted	weighted	ADJ
bjmsr-315	203	7	modification	modification	NOUN
bjmsr-315	203	8	of	of	ADP
bjmsr-315	203	9	partial	partial	ADJ
bjmsr-315	203	10	quicksort	quicksort	NOUN
bjmsr-315	203	11	procedure	procedure	NOUN
bjmsr-315	203	12	of	of	ADP
bjmsr-315	203	13	chamber	chamber	PROPN
bjmsr-315	203	14	(	(	PUNCT
bjmsr-315	203	15	1971	1971	NUM
bjmsr-315	203	16	)	)	PUNCT
bjmsr-315	203	17	to	to	PART
bjmsr-315	203	18	find	find	VERB
bjmsr-315	203	19	the	the	DET
bjmsr-315	203	20	t	t	NOUN
bjmsr-315	203	21	value	value	NOUN
bjmsr-315	203	22	efficiently	efficiently	ADV
bjmsr-315	203	23	.	.	PUNCT
bjmsr-315	204	1	the	the	DET
bjmsr-315	204	2	data	data	NOUN
bjmsr-315	204	3	point	point	NOUN
bjmsr-315	204	4	corresponding	correspond	VERB
bjmsr-315	204	5	to	to	ADP
bjmsr-315	204	6	the	the	DET
bjmsr-315	204	7	weighted	weight	VERB
bjmsr-315	204	8	median	median	NOUN
bjmsr-315	204	9	replaces	replace	VERB
bjmsr-315	204	10	xk	xk	PROPN
bjmsr-315	204	11	t.	t.	PROPN
bjmsr-315	204	12	by	by	ADP
bjmsr-315	204	13	using	use	VERB
bjmsr-315	204	14	the	the	DET
bjmsr-315	204	15	yi	yi	NOUN
bjmsr-315	204	16	from	from	ADP
bjmsr-315	204	17	(	(	PUNCT
bjmsr-315	204	18	20	20	NUM
bjmsr-315	204	19	)	)	PUNCT
bjmsr-315	204	20	and	and	CCONJ
bjmsr-315	204	21	the	the	DET
bjmsr-315	204	22	additional	additional	ADJ
bjmsr-315	204	23	mth	mth	NOUN
bjmsr-315	204	24	equation	equation	NOUN
bjmsr-315	204	25	,	,	PUNCT
bjmsr-315	204	26	the	the	DET
bjmsr-315	204	27	value	value	NOUN
bjmsr-315	204	28	of	of	ADP
bjmsr-315	204	29	ß0	ß0	NOUN
bjmsr-315	204	30	is	be	AUX
bjmsr-315	204	31	computed	compute	VERB
bjmsr-315	204	32	.	.	PUNCT
bjmsr-315	205	1	vector	vector	NOUN
bjmsr-315	205	2	δ	δ	PROPN
bjmsr-315	205	3	is	be	AUX
bjmsr-315	205	4	determined	determine	VERB
bjmsr-315	205	5	up	up	ADP
bjmsr-315	205	6	to	to	ADP
bjmsr-315	205	7	scalar	scalar	ADJ
bjmsr-315	205	8	multiples	multiple	NOUN
bjmsr-315	205	9	of	of	ADP
bjmsr-315	205	10	(	(	PUNCT
bjmsr-315	205	11	22	22	NUM
bjmsr-315	205	12	)	)	PUNCT
bjmsr-315	205	13	.	.	PUNCT
bjmsr-315	206	1	the	the	DET
bjmsr-315	206	2	new	new	ADJ
bjmsr-315	206	3	set	set	NOUN
bjmsr-315	206	4	of	of	ADP
bjmsr-315	206	5	parameters	parameter	NOUN
bjmsr-315	206	6	values	value	NOUN
bjmsr-315	206	7	are	be	AUX
bjmsr-315	206	8	computed	compute	VERB
bjmsr-315	206	9	by	by	ADP
bjmsr-315	206	10	(	(	PUNCT
bjmsr-315	206	11	21	21	NUM
bjmsr-315	206	12	)	)	PUNCT
bjmsr-315	206	13	.	.	PUNCT
bjmsr-315	207	1	now	now	ADV
bjmsr-315	207	2	a	a	DET
bjmsr-315	207	3	point	point	NOUN
bjmsr-315	207	4	should	should	AUX
bjmsr-315	207	5	be	be	AUX
bjmsr-315	207	6	deleted	delete	VERB
bjmsr-315	207	7	.	.	PUNCT
bjmsr-315	208	1	bloomfield	bloomfield	PROPN
bjmsr-315	208	2	and	and	CCONJ
bjmsr-315	208	3	steiger	steiger	PROPN
bjmsr-315	208	4	do	do	AUX
bjmsr-315	208	5	not	not	PART
bjmsr-315	208	6	give	give	VERB
bjmsr-315	208	7	an	an	DET
bjmsr-315	208	8	assured	assured	ADJ
bjmsr-315	208	9	way	way	NOUN
bjmsr-315	208	10	to	to	PART
bjmsr-315	208	11	identify	identify	VERB
bjmsr-315	208	12	this	this	DET
bjmsr-315	208	13	point	point	NOUN
bjmsr-315	208	14	.	.	PUNCT
bjmsr-315	209	1	they	they	PRON
bjmsr-315	209	2	propose	propose	VERB
bjmsr-315	209	3	a	a	DET
bjmsr-315	209	4	heuristic	heuristic	ADJ
bjmsr-315	209	5	method	method	NOUN
bjmsr-315	209	6	based	base	VERB
bjmsr-315	209	7	on	on	ADP
bjmsr-315	209	8	gradients	gradient	NOUN
bjmsr-315	209	9	and	and	CCONJ
bjmsr-315	209	10	use	use	VERB
bjmsr-315	209	11	the	the	DET
bjmsr-315	209	12	following	follow	VERB
bjmsr-315	209	13	quantity	quantity	NOUN
bjmsr-315	209	14	:	:	PUNCT
bjmsr-315	209	15	│	│	PUNCT
bjmsr-315	209	16	σ	σ	PROPN
bjmsr-315	209	17	wi	wi	PROPN
bjmsr-315	209	18	σ	σ	PROPN
bjmsr-315	209	19	wi	wi	PROPN
bjmsr-315	209	20	│	│	PROPN
bjmsr-315	209	21	σ	σ	PROPN
bjmsr-315	209	22	wi	wi	PROPN
bjmsr-315	209	23	│	│	PROPN
bjmsr-315	209	24	i	i	PRON
bjmsr-315	209	25	:	:	PUNCT
bjmsr-315	209	26	ri<0	ri<0	PROPN
bjmsr-315	210	1	i	i	PRON
bjmsr-315	210	2	:	:	PUNCT
bjmsr-315	210	3	ri>0	ri>0	ADJ
bjmsr-315	210	4	│	│	X
bjmsr-315	210	5	i	i	NOUN
bjmsr-315	210	6	:	:	PUNCT
bjmsr-315	210	7	ri=0	ri=0	PROPN
bjmsr-315	210	8	ρ	ρ	PROPN
bjmsr-315	210	9	=	=	PUNCT
bjmsr-315	210	10	─	─	PROPN
bjmsr-315	210	11	─	─	ADJ
bjmsr-315	210	12	─	─	ADJ
bjmsr-315	210	13	─	─	ADJ
bjmsr-315	210	14	─	─	ADJ
bjmsr-315	210	15	─	─	ADJ
bjmsr-315	210	16	─	─	ADJ
bjmsr-315	210	17	─	─	ADJ
bjmsr-315	210	18	─	─	ADJ
bjmsr-315	210	19	─	─	ADJ
bjmsr-315	210	20	─	─	ADJ
bjmsr-315	210	21	─	─	PROPN
bjmsr-315	210	22	─	─	ADJ
bjmsr-315	210	23	─	─	X
bjmsr-315	210	24	(	(	PUNCT
bjmsr-315	210	25	25	25	NUM
bjmsr-315	210	26	)	)	PUNCT
bjmsr-315	210	27	n	n	PROPN
bjmsr-315	210	28	σ	σ	PROPN
bjmsr-315	210	29	wi	wi	PROPN
bjmsr-315	210	30	i=1	i=1	PROPN
bjmsr-315	211	1	once	once	SCONJ
bjmsr-315	211	2	ρ	ρ	PROPN
bjmsr-315	211	3	is	be	AUX
bjmsr-315	211	4	calculated	calculate	VERB
bjmsr-315	211	5	for	for	ADP
bjmsr-315	211	6	each	each	DET
bjmsr-315	211	7	candidate	candidate	NOUN
bjmsr-315	211	8	point	point	NOUN
bjmsr-315	211	9	for	for	ADP
bjmsr-315	211	10	deletion	deletion	NOUN
bjmsr-315	211	11	,	,	PUNCT
bjmsr-315	211	12	and	and	CCONJ
bjmsr-315	211	13	delete	delete	VERB
bjmsr-315	211	14	the	the	DET
bjmsr-315	211	15	point	point	NOUN
bjmsr-315	211	16	for	for	ADP
bjmsr-315	211	17	which	which	PRON
bjmsr-315	211	18	ρ	ρ	NOUN
bjmsr-315	211	19	is	be	AUX
bjmsr-315	211	20	largest	large	ADJ
bjmsr-315	211	21	.	.	PUNCT
bjmsr-315	212	1	to	to	PART
bjmsr-315	212	2	start	start	VERB
bjmsr-315	212	3	the	the	DET
bjmsr-315	212	4	algorithm	algorithm	NOUN
bjmsr-315	212	5	,	,	PUNCT
bjmsr-315	212	6	any	any	DET
bjmsr-315	212	7	set	set	NOUN
bjmsr-315	212	8	of	of	ADP
bjmsr-315	212	9	m	m	PROPN
bjmsr-315	212	10	rows	row	NOUN
bjmsr-315	212	11	of	of	ADP
bjmsr-315	212	12	x	x	PUNCT
bjmsr-315	212	13	may	may	AUX
bjmsr-315	212	14	be	be	AUX
bjmsr-315	212	15	chosen	choose	VERB
bjmsr-315	212	16	,	,	PUNCT
bjmsr-315	212	17	with	with	ADP
bjmsr-315	212	18	the	the	DET
bjmsr-315	212	19	appropriate	appropriate	ADJ
bjmsr-315	212	20	ß0	ß0	NOUN
bjmsr-315	212	21	.	.	PUNCT
bjmsr-315	213	1	add	add	VERB
bjmsr-315	213	2	variables	variable	NOUN
bjmsr-315	213	3	stepwise	stepwise	NOUN
bjmsr-315	213	4	until	until	SCONJ
bjmsr-315	213	5	a	a	DET
bjmsr-315	213	6	fit	fit	NOUN
bjmsr-315	213	7	for	for	ADP
bjmsr-315	213	8	ß0	ß0	NOUN
bjmsr-315	213	9	and	and	CCONJ
bjmsr-315	213	10	the	the	DET
bjmsr-315	213	11	corresponding	corresponding	ADJ
bjmsr-315	213	12	set	set	NOUN
bjmsr-315	213	13	of	of	ADP
bjmsr-315	213	14	m	m	PROPN
bjmsr-315	213	15	points	point	NOUN
bjmsr-315	213	16	are	be	AUX
bjmsr-315	213	17	derived	derive	VERB
bjmsr-315	213	18	.	.	PUNCT
bjmsr-315	214	1	at	at	ADP
bjmsr-315	214	2	each	each	DET
bjmsr-315	214	3	intermediate	intermediate	ADJ
bjmsr-315	214	4	step	step	NOUN
bjmsr-315	214	5	,	,	PUNCT
bjmsr-315	214	6	the	the	DET
bjmsr-315	214	7	fit	fit	NOUN
bjmsr-315	214	8	involves	involve	VERB
bjmsr-315	214	9	k	k	PROPN
bjmsr-315	214	10	variables	variable	NOUN
bjmsr-315	214	11	where	where	SCONJ
bjmsr-315	214	12	0≤k	0≤k	ADJ
bjmsr-315	214	13	<	<	X
bjmsr-315	214	14	m	m	VERB
bjmsr-315	214	15	and	and	CCONJ
bjmsr-315	214	16	the	the	DET
bjmsr-315	214	17	corresponding	corresponding	ADJ
bjmsr-315	214	18	set	set	NOUN
bjmsr-315	214	19	of	of	ADP
bjmsr-315	214	20	k	k	PROPN
bjmsr-315	214	21	data	data	PROPN
bjmsr-315	214	22	points	point	NOUN
bjmsr-315	214	23	with	with	ADP
bjmsr-315	214	24	zero	zero	NUM
bjmsr-315	214	25	residuals	residual	NOUN
bjmsr-315	214	26	.	.	PUNCT
bjmsr-315	215	1	improving	improve	VERB
bjmsr-315	215	2	the	the	DET
bjmsr-315	215	3	fit	fit	NOUN
bjmsr-315	215	4	by	by	ADP
bjmsr-315	215	5	entering	enter	VERB
bjmsr-315	215	6	a	a	DET
bjmsr-315	215	7	new	new	ADJ
bjmsr-315	215	8	variable	variable	NOUN
bjmsr-315	215	9	,	,	PUNCT
bjmsr-315	215	10	thus	thus	ADV
bjmsr-315	215	11	increasing	increase	VERB
bjmsr-315	215	12	k	k	PROPN
bjmsr-315	215	13	to	to	ADP
bjmsr-315	215	14	k+1	k+1	VERB
bjmsr-315	215	15	.	.	PUNCT
bjmsr-315	216	1	at	at	ADP
bjmsr-315	216	2	each	each	DET
bjmsr-315	216	3	stage	stage	NOUN
bjmsr-315	216	4	,	,	PUNCT
bjmsr-315	216	5	it	it	PRON
bjmsr-315	216	6	is	be	AUX
bjmsr-315	216	7	the	the	DET
bjmsr-315	216	8	measure	measure	NOUN
bjmsr-315	216	9	ρ	ρ	NOUN
bjmsr-315	216	10	that	that	PRON
bjmsr-315	216	11	determines	determine	VERB
bjmsr-315	216	12	whether	whether	SCONJ
bjmsr-315	216	13	we	we	PRON
bjmsr-315	216	14	set	set	VERB
bjmsr-315	216	15	up	up	ADP
bjmsr-315	216	16	to	to	ADP
bjmsr-315	216	17	a	a	DET
bjmsr-315	216	18	larger	large	ADJ
bjmsr-315	216	19	model	model	NOUN
bjmsr-315	216	20	or	or	CCONJ
bjmsr-315	216	21	improve	improve	VERB
bjmsr-315	216	22	the	the	DET
bjmsr-315	216	23	current	current	ADJ
bjmsr-315	216	24	one	one	NUM
bjmsr-315	216	25	without	without	ADP
bjmsr-315	216	26	setting	set	VERB
bjmsr-315	216	27	up	up	ADP
bjmsr-315	216	28	.	.	PUNCT
bjmsr-315	217	1	suppose	suppose	VERB
bjmsr-315	217	2	at	at	ADP
bjmsr-315	217	3	the	the	DET
bjmsr-315	217	4	current	current	ADJ
bjmsr-315	217	5	stage	stage	NOUN
bjmsr-315	217	6	we	we	PRON
bjmsr-315	217	7	are	be	AUX
bjmsr-315	217	8	dealing	deal	VERB
bjmsr-315	217	9	with	with	ADP
bjmsr-315	217	10	k	k	PROPN
bjmsr-315	217	11	variables	variable	NOUN
bjmsr-315	217	12	.	.	PUNCT
bjmsr-315	218	1	if	if	SCONJ
bjmsr-315	218	2	the	the	DET
bjmsr-315	218	3	value	value	NOUN
bjmsr-315	218	4	of	of	ADP
bjmsr-315	218	5	ρ	ρ	PROPN
bjmsr-315	218	6	is	be	AUX
bjmsr-315	218	7	largest	large	ADJ
bjmsr-315	218	8	at	at	ADP
bjmsr-315	218	9	the	the	DET
bjmsr-315	218	10	variable	variable	ADJ
bjmsr-315	218	11	p	p	NOUN
bjmsr-315	218	12	,	,	PUNCT
bjmsr-315	218	13	which	which	PRON
bjmsr-315	218	14	p	p	NOUN
bjmsr-315	218	15	is	be	AUX
bjmsr-315	218	16	not	not	PART
bjmsr-315	218	17	in	in	ADP
bjmsr-315	218	18	the	the	DET
bjmsr-315	218	19	set	set	NOUN
bjmsr-315	218	20	of	of	ADP
bjmsr-315	218	21	k	k	PROPN
bjmsr-315	218	22	variables	variable	NOUN
bjmsr-315	218	23	above	above	ADV
bjmsr-315	218	24	,	,	PUNCT
bjmsr-315	218	25	k	k	PROPN
bjmsr-315	218	26	is	be	AUX
bjmsr-315	218	27	increased	increase	VERB
bjmsr-315	218	28	to	to	ADP
bjmsr-315	218	29	k+1	k+1	X
bjmsr-315	218	30	by	by	ADP
bjmsr-315	218	31	entering	enter	VERB
bjmsr-315	218	32	the	the	DET
bjmsr-315	218	33	variable	variable	ADJ
bjmsr-315	218	34	p	p	NOUN
bjmsr-315	218	35	and	and	CCONJ
bjmsr-315	218	36	improving	improve	VERB
bjmsr-315	218	37	the	the	DET
bjmsr-315	218	38	fit	fit	NOUN
bjmsr-315	218	39	with	with	ADP
bjmsr-315	218	40	k+1	k+1	NOUN
bjmsr-315	218	41	variables	variable	NOUN
bjmsr-315	218	42	.	.	PUNCT
bjmsr-315	219	1	if	if	SCONJ
bjmsr-315	219	2	p	p	NOUN
bjmsr-315	219	3	is	be	AUX
bjmsr-315	219	4	in	in	ADP
bjmsr-315	219	5	the	the	DET
bjmsr-315	219	6	current	current	ADJ
bjmsr-315	219	7	set	set	NOUN
bjmsr-315	219	8	of	of	ADP
bjmsr-315	219	9	k	k	PROPN
bjmsr-315	219	10	variables	variable	NOUN
bjmsr-315	219	11	and	and	CCONJ
bjmsr-315	219	12	ρ	ρ	NOUN
bjmsr-315	219	13	is	be	AUX
bjmsr-315	219	14	in	in	ADP
bjmsr-315	219	15	the	the	DET
bjmsr-315	219	16	largest	large	ADJ
bjmsr-315	219	17	value	value	NOUN
bjmsr-315	219	18	,	,	PUNCT
bjmsr-315	219	19	the	the	DET
bjmsr-315	219	20	corresponding	corresponding	ADJ
bjmsr-315	219	21	point	point	NOUN
bjmsr-315	219	22	to	to	ADP
bjmsr-315	219	23	p	p	NOUN
bjmsr-315	219	24	is	be	AUX
bjmsr-315	219	25	deleted	delete	VERB
bjmsr-315	219	26	and	and	CCONJ
bjmsr-315	219	27	replaced	replace	VERB
bjmsr-315	219	28	in	in	ADP
bjmsr-315	219	29	a	a	DET
bjmsr-315	219	30	manner	manner	NOUN
bjmsr-315	219	31	described	describe	VERB
bjmsr-315	219	32	before	before	ADV
bjmsr-315	219	33	.	.	PUNCT
bjmsr-315	220	1	in	in	ADP
bjmsr-315	220	2	this	this	DET
bjmsr-315	220	3	paper	paper	NOUN
bjmsr-315	220	4	relationship	relationship	NOUN
bjmsr-315	220	5	of	of	ADP
bjmsr-315	220	6	this	this	DET
bjmsr-315	220	7	algorithm	algorithm	NOUN
bjmsr-315	220	8	to	to	ADP
bjmsr-315	220	9	linear	linear	ADJ
bjmsr-315	220	10	programming	programming	NOUN
bjmsr-315	220	11	is	be	AUX
bjmsr-315	220	12	also	also	ADV
bjmsr-315	220	13	discussed	discuss	VERB
bjmsr-315	220	14	.	.	PUNCT
bjmsr-315	221	1	copyright	copyright	NOUN
bjmsr-315	221	2	©	©	PROPN
bjmsr-315	221	3	cc	cc	PROPN
bjmsr-315	221	4	-	-	PUNCT
bjmsr-315	221	5	by	by	ADP
bjmsr-315	221	6	-	-	PUNCT
bjmsr-315	221	7	nc	nc	PROPN
bjmsr-315	221	8	2019	2019	NUM
bjmsr-315	221	9	,	,	PUNCT
bjmsr-315	221	10	bjmsr	bjmsr	PROPN
bjmsr-315	221	11	www.cribfb.com/journal/index.php/bjmsr	www.cribfb.com/journal/index.php/bjmsr	X
bjmsr-315	222	1	bangladesh	bangladesh	PROPN
bjmsr-315	222	2	journal	journal	PROPN
bjmsr-315	222	3	of	of	ADP
bjmsr-315	222	4	multidisciplinary	multidisciplinary	ADJ
bjmsr-315	222	5	scientific	scientific	ADJ
bjmsr-315	222	6	research	research	NOUN
bjmsr-315	222	7	vol	vol	NOUN
bjmsr-315	222	8	.	.	PROPN
bjmsr-315	222	9	1	1	NUM
bjmsr-315	222	10	,	,	PUNCT
bjmsr-315	222	11	no	no	INTJ
bjmsr-315	222	12	.	.	NOUN
bjmsr-315	222	13	1	1	NUM
bjmsr-315	222	14	;	;	PUNCT
bjmsr-315	222	15	2019	2019	NUM
bjmsr-315	222	16	56	56	NUM
bjmsr-315	222	17	seneta	seneta	NOUN
bjmsr-315	222	18	and	and	CCONJ
bjmsr-315	222	19	steiger	steiger	PROPN
bjmsr-315	222	20	(	(	PUNCT
bjmsr-315	222	21	1984	1984	NUM
bjmsr-315	222	22	)	)	PUNCT
bjmsr-315	222	23	proposed	propose	VERB
bjmsr-315	222	24	an	an	DET
bjmsr-315	222	25	algorithm	algorithm	NOUN
bjmsr-315	222	26	for	for	ADP
bjmsr-315	222	27	the	the	DET
bjmsr-315	222	28	l1	l1	PROPN
bjmsr-315	222	29	norm	norm	NOUN
bjmsr-315	222	30	solution	solution	NOUN
bjmsr-315	222	31	of	of	ADP
bjmsr-315	222	32	a	a	DET
bjmsr-315	222	33	slightly	slightly	ADV
bjmsr-315	222	34	overdetermined	overdetermine	VERB
bjmsr-315	222	35	system	system	NOUN
bjmsr-315	222	36	of	of	ADP
bjmsr-315	222	37	equations	equation	NOUN
bjmsr-315	222	38	.	.	PUNCT
bjmsr-315	223	1	their	their	PRON
bjmsr-315	223	2	proposition	proposition	NOUN
bjmsr-315	223	3	is	be	AUX
bjmsr-315	223	4	based	base	VERB
bjmsr-315	223	5	on	on	ADP
bjmsr-315	223	6	the	the	DET
bjmsr-315	223	7	above	above	ADJ
bjmsr-315	223	8	algorithm	algorithm	NOUN
bjmsr-315	223	9	of	of	ADP
bjmsr-315	223	10	bloomfield	bloomfield	PROPN
bjmsr-315	223	11	and	and	CCONJ
bjmsr-315	223	12	steiger	steiger	PROPN
bjmsr-315	223	13	.	.	PUNCT
bjmsr-315	224	1	it	it	PRON
bjmsr-315	224	2	is	be	AUX
bjmsr-315	224	3	more	more	ADV
bjmsr-315	224	4	efficient	efficient	ADJ
bjmsr-315	224	5	than	than	ADP
bjmsr-315	224	6	the	the	DET
bjmsr-315	224	7	former	former	ADJ
bjmsr-315	224	8	if	if	SCONJ
bjmsr-315	224	9	m	m	NOUN
bjmsr-315	224	10	is	be	AUX
bjmsr-315	224	11	near	near	ADP
bjmsr-315	224	12	n.	n.	NOUN
bjmsr-315	224	13	given	give	VERB
bjmsr-315	224	14	(	(	PUNCT
bjmsr-315	224	15	xi	xi	PROPN
bjmsr-315	224	16	,	,	PUNCT
bjmsr-315	224	17	yi)εrm+1	yi)εrm+1	PROPN
bjmsr-315	224	18	,	,	PUNCT
bjmsr-315	224	19	i=1,	i=1,	PROPN
bjmsr-315	224	20	...	...	PUNCT
bjmsr-315	224	21	,n	,n	PUNCT
bjmsr-315	224	22	and	and	CCONJ
bjmsr-315	224	23	k	k	X
bjmsr-315	224	24	=	=	NOUN
bjmsr-315	224	25	n	n	CCONJ
bjmsr-315	224	26	-	-	PUNCT
bjmsr-315	224	27	m	m	NOUN
bjmsr-315	224	28	,	,	PUNCT
bjmsr-315	224	29	x=(x1,	x=(x1,	PROPN
bjmsr-315	224	30	...	...	PUNCT
bjmsr-315	225	1	,xn)t	,xn)t	CCONJ
bjmsr-315	225	2	,	,	PUNCT
bjmsr-315	225	3	their	their	PRON
bjmsr-315	225	4	algorithm	algorithm	NOUN
bjmsr-315	225	5	may	may	AUX
bjmsr-315	225	6	be	be	AUX
bjmsr-315	225	7	described	describe	VERB
bjmsr-315	225	8	as	as	ADP
bjmsr-315	225	9	follows	follow	VERB
bjmsr-315	225	10	:	:	PUNCT
bjmsr-315	225	11	step	step	NOUN
bjmsr-315	225	12	1	1	NUM
bjmsr-315	225	13	)	)	PUNCT
bjmsr-315	225	14	renumber	renumber	NOUN
bjmsr-315	225	15	the	the	DET
bjmsr-315	225	16	rows	row	NOUN
bjmsr-315	225	17	of	of	ADP
bjmsr-315	225	18	(	(	PUNCT
bjmsr-315	225	19	x|y	x|y	X
bjmsr-315	225	20	)	)	PUNCT
bjmsr-315	226	1	such	such	ADJ
bjmsr-315	226	2	that	that	SCONJ
bjmsr-315	226	3	xn	xn	PROPN
bjmsr-315	226	4	the	the	DET
bjmsr-315	226	5	bottom	bottom	NOUN
bjmsr-315	226	6	m	m	VERB
bjmsr-315	226	7	rows	row	NOUN
bjmsr-315	226	8	of	of	ADP
bjmsr-315	226	9	x	x	X
bjmsr-315	226	10	,	,	PUNCT
bjmsr-315	226	11	is	be	AUX
bjmsr-315	226	12	invertible	invertible	ADJ
bjmsr-315	226	13	.	.	PUNCT
bjmsr-315	227	1	solve	solve	VERB
bjmsr-315	227	2	the	the	DET
bjmsr-315	227	3	k	k	PROPN
bjmsr-315	227	4	linear	linear	PROPN
bjmsr-315	227	5	equations	equation	NOUN
bjmsr-315	227	6	system	system	NOUN
bjmsr-315	227	7	xn	xn	PROPN
bjmsr-315	227	8	tn=-xt	tn=-xt	PROPN
bjmsr-315	227	9	t	t	PROPN
bjmsr-315	227	10	for	for	ADP
bjmsr-315	227	11	n	n	CCONJ
bjmsr-315	227	12	,	,	PUNCT
bjmsr-315	227	13	where	where	SCONJ
bjmsr-315	227	14	xt	xt	PROPN
bjmsr-315	227	15	denotes	denote	VERB
bjmsr-315	227	16	the	the	DET
bjmsr-315	227	17	top	top	ADJ
bjmsr-315	227	18	k	k	PROPN
bjmsr-315	227	19	rows	row	NOUN
bjmsr-315	227	20	of	of	ADP
bjmsr-315	227	21	x.	x.	NOUN
bjmsr-315	227	22	step	step	NOUN
bjmsr-315	227	23	2	2	NUM
bjmsr-315	227	24	)	)	PUNCT
bjmsr-315	227	25	let	let	VERB
bjmsr-315	227	26	d=(a|c	d=(a|c	NOUN
bjmsr-315	227	27	)	)	PUNCT
bjmsr-315	227	28	where	where	SCONJ
bjmsr-315	227	29	a=(i|n	a=(i|n	NUM
bjmsr-315	227	30	)	)	PUNCT
bjmsr-315	227	31	is	be	AUX
bjmsr-315	227	32	of	of	ADP
bjmsr-315	227	33	size	size	NOUN
bjmsr-315	227	34	kxn	kxn	NOUN
bjmsr-315	227	35	and	and	CCONJ
bjmsr-315	227	36	c	c	X
bjmsr-315	227	37	=	=	NOUN
bjmsr-315	227	38	ay	ay	NOUN
bjmsr-315	227	39	.	.	PUNCT
bjmsr-315	227	40	step	step	NOUN
bjmsr-315	227	41	3	3	NUM
bjmsr-315	227	42	)	)	PUNCT
bjmsr-315	227	43	let	let	VERB
bjmsr-315	227	44	σ	σ	PROPN
bjmsr-315	227	45	=(	=(	PROPN
bjmsr-315	227	46	1,	1,	NUM
bjmsr-315	227	47	...	...	PUNCT
bjmsr-315	227	48	,k	,k	PUNCT
bjmsr-315	227	49	)	)	PUNCT
bjmsr-315	227	50	,	,	PUNCT
bjmsr-315	227	51	σc=(k+1,	σc=(k+1,	NOUN
bjmsr-315	227	52	...	...	PUNCT
bjmsr-315	227	53	,n	,n	NOUN
bjmsr-315	227	54	)	)	PUNCT
bjmsr-315	227	55	.	.	PUNCT
bjmsr-315	228	1	step	step	NOUN
bjmsr-315	228	2	4	4	NUM
bjmsr-315	228	3	)	)	PUNCT
bjmsr-315	228	4	set	set	VERB
bjmsr-315	228	5	rσ(i)=bi	rσ(i)=bi	NOUN
bjmsr-315	228	6	for	for	ADP
bjmsr-315	228	7	i=1,	i=1,	NOUN
bjmsr-315	228	8	...	...	PUNCT
bjmsr-315	228	9	,k	,k	PUNCT
bjmsr-315	228	10	;	;	PUNCT
bjmsr-315	228	11	rσc(i)=0	rσc(i)=0	NOUN
bjmsr-315	228	12	for	for	ADP
bjmsr-315	228	13	i	i	PROPN
bjmsr-315	228	14	=	=	NOUN
bjmsr-315	228	15	k+1,	k+1,	NOUN
bjmsr-315	228	16	...	...	PUNCT
bjmsr-315	228	17	,n	,n	PUNCT
bjmsr-315	228	18	.	.	PUNCT
bjmsr-315	229	1	step	step	NOUN
bjmsr-315	229	2	5	5	NUM
bjmsr-315	229	3	)	)	PUNCT
bjmsr-315	229	4	let	let	VERB
bjmsr-315	229	5	i={i	i={i	NOUN
bjmsr-315	229	6	│	│	ADJ
bjmsr-315	229	7	1≤i≤k	1≤i≤k	NUM
bjmsr-315	229	8	,	,	PUNCT
bjmsr-315	229	9	ci=0	ci=0	ADJ
bjmsr-315	229	10	}	}	PUNCT
bjmsr-315	229	11	.	.	PUNCT
bjmsr-315	230	1	step	step	NOUN
bjmsr-315	230	2	6	6	NUM
bjmsr-315	230	3	)	)	PUNCT
bjmsr-315	230	4	do	do	AUX
bjmsr-315	230	5	loop	loop	VERB
bjmsr-315	230	6	for	for	ADP
bjmsr-315	230	7	j=1	j=1	PROPN
bjmsr-315	230	8	to	to	PART
bjmsr-315	230	9	m	m	PRON
bjmsr-315	230	10	:	:	PUNCT
bjmsr-315	230	11	let	let	VERB
bjmsr-315	230	12	vi	vi	NOUN
bjmsr-315	230	13	=	=	SYM
bjmsr-315	230	14	diσc(j	diσc(j	NOUN
bjmsr-315	230	15	)	)	PUNCT
bjmsr-315	230	16	for	for	ADP
bjmsr-315	230	17	j=1,	j=1,	NOUN
bjmsr-315	230	18	...	...	PUNCT
bjmsr-315	230	19	,k	,k	PUNCT
bjmsr-315	230	20	.	.	PUNCT
bjmsr-315	231	1	let	let	VERB
bjmsr-315	231	2	m={i	m={i	NOUN
bjmsr-315	231	3	│	│	NOUN
bjmsr-315	231	4	sgn(ci)≠sgn(vi	sgn(ci)≠sgn(vi	NUM
bjmsr-315	231	5	)	)	PUNCT
bjmsr-315	231	6	}	}	PUNCT
bjmsr-315	231	7	and	and	CCONJ
bjmsr-315	231	8	j={1,	j={1,	PROPN
bjmsr-315	231	9	...	...	PUNCT
bjmsr-315	231	10	,k}\n\i	,k}\n\i	PUNCT
bjmsr-315	231	11	.	.	PUNCT
bjmsr-315	232	1	│	│	ADJ
bjmsr-315	232	2	σm	σm	NOUN
bjmsr-315	232	3	│	│	VERB
bjmsr-315	232	4	vi	vi	NOUN
bjmsr-315	232	5	│	│	ADJ
bjmsr-315	232	6	-σj	-σj	ADJ
bjmsr-315	232	7	│	│	NUM
bjmsr-315	232	8	vi	vi	NOUN
bjmsr-315	232	9	│	│	ADJ
bjmsr-315	232	10	│	│	ADJ
bjmsr-315	232	11	-1	-1	NOUN
bjmsr-315	232	12	-	-	PUNCT
bjmsr-315	232	13	σi	σi	NOUN
bjmsr-315	232	14	│	│	NOUN
bjmsr-315	232	15	vi	vi	PRON
bjmsr-315	232	16	│	│	NOUN
bjmsr-315	232	17	let	let	VERB
bjmsr-315	232	18	ßj	ßj	NOUN
bjmsr-315	232	19	=	=	PUNCT
bjmsr-315	232	20	─	─	VERB
bjmsr-315	232	21	─	─	ADJ
bjmsr-315	232	22	─	─	ADJ
bjmsr-315	232	23	─	─	ADJ
bjmsr-315	232	24	─	─	ADJ
bjmsr-315	232	25	─	─	ADJ
bjmsr-315	232	26	─	─	ADJ
bjmsr-315	232	27	─	─	ADJ
bjmsr-315	232	28	─	─	ADJ
bjmsr-315	232	29	─	─	ADJ
bjmsr-315	232	30	─	─	ADJ
bjmsr-315	232	31	─	─	ADJ
bjmsr-315	232	32	─	─	ADJ
bjmsr-315	232	33	─	─	ADJ
bjmsr-315	232	34	─	─	ADJ
bjmsr-315	232	35	─	─	PROPN
bjmsr-315	232	36	─	─	PROPN
bjmsr-315	232	37	─	─	PROPN
bjmsr-315	232	38	k	k	PROPN
bjmsr-315	233	1	1	1	NUM
bjmsr-315	233	2	+	+	NUM
bjmsr-315	233	3	σ	σ	PROPN
bjmsr-315	233	4	│	│	SYM
bjmsr-315	233	5	vi	vi	NOUN
bjmsr-315	233	6	│	│	X
bjmsr-315	233	7	i=1	i=1	ADP
bjmsr-315	233	8	end	end	NOUN
bjmsr-315	233	9	loop	loop	NOUN
bjmsr-315	233	10	.	.	PUNCT
bjmsr-315	234	1	step	step	NOUN
bjmsr-315	234	2	7	7	NUM
bjmsr-315	234	3	)	)	PUNCT
bjmsr-315	234	4	set	set	VERB
bjmsr-315	234	5	s={1,	s={1,	NOUN
bjmsr-315	234	6	...	...	PUNCT
bjmsr-315	234	7	,m	,m	PUNCT
bjmsr-315	234	8	}	}	PUNCT
bjmsr-315	234	9	.	.	PUNCT
bjmsr-315	235	1	step	step	NOUN
bjmsr-315	235	2	8)	8)	NUM
bjmsr-315	235	3	choose	choose	VERB
bjmsr-315	235	4	ßq	ßq	ADV
bjmsr-315	235	5	as	as	ADP
bjmsr-315	235	6	max{ßj	max{ßj	X
bjmsr-315	235	7	}	}	PUNCT
bjmsr-315	235	8	.	.	PUNCT
bjmsr-315	236	1	s	s	PART
bjmsr-315	236	2	step	step	NOUN
bjmsr-315	236	3	9	9	NUM
bjmsr-315	236	4	)	)	PUNCT
bjmsr-315	236	5	if	if	SCONJ
bjmsr-315	236	6	ßq>0	ßq>0	NOUN
bjmsr-315	236	7	go	go	VERB
bjmsr-315	236	8	to	to	PART
bjmsr-315	236	9	step	step	VERB
bjmsr-315	236	10	11	11	NUM
bjmsr-315	236	11	.	.	PUNCT
bjmsr-315	237	1	k	k	NOUN
bjmsr-315	237	2	step	step	NOUN
bjmsr-315	237	3	10	10	NUM
bjmsr-315	237	4	)	)	PUNCT
bjmsr-315	237	5	if	if	SCONJ
bjmsr-315	237	6	п	п	PROPN
bjmsr-315	237	7	rσ(i)=0	rσ(i)=0	PROPN
bjmsr-315	237	8	,	,	PUNCT
bjmsr-315	237	9	the	the	DET
bjmsr-315	237	10	problem	problem	NOUN
bjmsr-315	237	11	is	be	AUX
bjmsr-315	237	12	degenerate	degenerate	ADJ
bjmsr-315	237	13	;	;	PUNCT
bjmsr-315	237	14	and	and	CCONJ
bjmsr-315	237	15	stop	stop	VERB
bjmsr-315	237	16	.	.	PUNCT
bjmsr-315	238	1	i=1	i=1	PROPN
bjmsr-315	238	2	m	m	VERB
bjmsr-315	238	3	otherwise	otherwise	ADV
bjmsr-315	238	4	,	,	PUNCT
bjmsr-315	238	5	solve	solve	VERB
bjmsr-315	238	6	,	,	PUNCT
bjmsr-315	238	7	yσc(i)=	yσc(i)=	NUM
bjmsr-315	238	8	σ	σ	PROPN
bjmsr-315	238	9	xσc(i)jθj	xσc(i)jθj	PROPN
bjmsr-315	238	10	for	for	ADP
bjmsr-315	238	11	θ	θ	PROPN
bjmsr-315	238	12	,	,	PUNCT
bjmsr-315	238	13	stop	stop	NOUN
bjmsr-315	238	14	.	.	PUNCT
bjmsr-315	239	1	j=1	j=1	PROPN
bjmsr-315	239	2	step	step	NOUN
bjmsr-315	239	3	11	11	NUM
bjmsr-315	239	4	)	)	PUNCT
bjmsr-315	239	5	let	let	VERB
bjmsr-315	239	6	vi	vi	NOUN
bjmsr-315	239	7	=	=	NOUN
bjmsr-315	239	8	di	di	NOUN
bjmsr-315	239	9	,	,	PUNCT
bjmsr-315	239	10	σc(q	σc(q	NUM
bjmsr-315	239	11	)	)	PUNCT
bjmsr-315	239	12	for	for	ADP
bjmsr-315	239	13	i=1,	i=1,	NOUN
bjmsr-315	239	14	...	...	PUNCT
bjmsr-315	239	15	,k	,k	PROPN
bjmsr-315	239	16	.	.	PUNCT
bjmsr-315	240	1	step	step	NOUN
bjmsr-315	240	2	12	12	NUM
bjmsr-315	240	3	)	)	PUNCT
bjmsr-315	240	4	compute	compute	NOUN
bjmsr-315	240	5	t^=weighted	t^=weighte	VERB
bjmsr-315	240	6	median	median	NOUN
bjmsr-315	240	7	of	of	ADP
bjmsr-315	240	8	c1	c1	PROPN
bjmsr-315	240	9	/	/	SYM
bjmsr-315	240	10	v1,	v1,	NOUN
bjmsr-315	240	11	...	...	PUNCT
bjmsr-315	240	12	,ck	,ck	PUNCT
bjmsr-315	240	13	/	/	SYM
bjmsr-315	240	14	vk,0	vk,0	PROPN
bjmsr-315	240	15	with	with	ADP
bjmsr-315	240	16	weights	weight	NOUN
bjmsr-315	240	17	│	│	VERB
bjmsr-315	240	18	v1	v1	NOUN
bjmsr-315	240	19	│	│	VERB
bjmsr-315	240	20	,	,	PUNCT
bjmsr-315	240	21	...	...	PUNCT
bjmsr-315	240	22	,	,	PUNCT
bjmsr-315	240	23	│	│	X
bjmsr-315	240	24	vk	vk	PRON
bjmsr-315	240	25	│	│	VERB
bjmsr-315	240	26	,1	,1	NOUN
bjmsr-315	240	27	.	.	PUNCT
bjmsr-315	241	1	step	step	NOUN
bjmsr-315	241	2	13	13	NUM
bjmsr-315	241	3	)	)	PUNCT
bjmsr-315	241	4	if	if	SCONJ
bjmsr-315	241	5	t^=cp	t^=cp	NOUN
bjmsr-315	241	6	/	/	SYM
bjmsr-315	241	7	vp≠0	vp≠0	NUM
bjmsr-315	241	8	go	go	VERB
bjmsr-315	241	9	to	to	PART
bjmsr-315	241	10	step	step	VERB
bjmsr-315	241	11	16	16	NUM
bjmsr-315	241	12	.	.	PUNCT
bjmsr-315	242	1	step	step	NOUN
bjmsr-315	242	2	14	14	NUM
bjmsr-315	242	3	)	)	PUNCT
bjmsr-315	242	4	let	let	VERB
bjmsr-315	242	5	s	s	NOUN
bjmsr-315	242	6	=	=	NOUN
bjmsr-315	242	7	s\{q	s\{q	X
bjmsr-315	242	8	}	}	PUNCT
bjmsr-315	242	9	;	;	PUNCT
bjmsr-315	242	10	if	if	SCONJ
bjmsr-315	242	11	s	s	NOUN
bjmsr-315	242	12	=	=	NOUN
bjmsr-315	242	13	ø	ø	NOUN
bjmsr-315	242	14	go	go	VERB
bjmsr-315	242	15	to	to	PART
bjmsr-315	242	16	step	step	VERB
bjmsr-315	242	17	10	10	NUM
bjmsr-315	242	18	.	.	PUNCT
bjmsr-315	243	1	step	step	NOUN
bjmsr-315	243	2	15	15	NUM
bjmsr-315	243	3	)	)	PUNCT
bjmsr-315	243	4	go	go	VERB
bjmsr-315	243	5	to	to	PART
bjmsr-315	243	6	step	step	VERB
bjmsr-315	243	7	8	8	NUM
bjmsr-315	243	8	.	.	PUNCT
bjmsr-315	244	1	step	step	NOUN
bjmsr-315	244	2	16	16	NUM
bjmsr-315	244	3	)	)	PUNCT
bjmsr-315	244	4	divide	divide	NOUN
bjmsr-315	244	5	row	row	NOUN
bjmsr-315	244	6	p	p	NOUN
bjmsr-315	244	7	of	of	ADP
bjmsr-315	244	8	d	d	PROPN
bjmsr-315	244	9	by	by	ADP
bjmsr-315	244	10	dpσc(q	dpσc(q	PROPN
bjmsr-315	244	11	)	)	PUNCT
bjmsr-315	244	12	.	.	PUNCT
bjmsr-315	245	1	step	step	NOUN
bjmsr-315	245	2	17	17	NUM
bjmsr-315	245	3	)	)	PUNCT
bjmsr-315	245	4	for	for	ADP
bjmsr-315	245	5	i≠p	i≠p	NOUN
bjmsr-315	245	6	in	in	ADP
bjmsr-315	245	7	d	d	PROPN
bjmsr-315	245	8	,	,	PUNCT
bjmsr-315	245	9	let	let	VERB
bjmsr-315	245	10	(	(	PUNCT
bjmsr-315	245	11	row	row	VERB
bjmsr-315	245	12	i)=(row	i)=(row	PROPN
bjmsr-315	245	13	i)-(row	i)-(row	NOUN
bjmsr-315	245	14	p)*diσc(q	p)*diσc(q	ADJ
bjmsr-315	245	15	)	)	PUNCT
bjmsr-315	245	16	.	.	PUNCT
bjmsr-315	246	1	step	step	NOUN
bjmsr-315	246	2	18	18	NUM
bjmsr-315	246	3	)	)	PUNCT
bjmsr-315	246	4	commute	commute	NOUN
bjmsr-315	246	5	for	for	ADP
bjmsr-315	246	6	pair	pair	NOUN
bjmsr-315	246	7	σ(p	σ(p	PROPN
bjmsr-315	246	8	)	)	PUNCT
bjmsr-315	246	9	and	and	CCONJ
bjmsr-315	246	10	σc(q	σc(q	NUM
bjmsr-315	246	11	)	)	PUNCT
bjmsr-315	246	12	.	.	PUNCT
bjmsr-315	247	1	step	step	NOUN
bjmsr-315	247	2	19	19	NUM
bjmsr-315	247	3	)	)	PUNCT
bjmsr-315	247	4	let	let	VERB
bjmsr-315	247	5	rσ(i)=bi	rσ(i)=bi	NOUN
bjmsr-315	247	6	for	for	ADP
bjmsr-315	247	7	i=1,	i=1,	NOUN
bjmsr-315	247	8	...	...	PUNCT
bjmsr-315	247	9	,k	,k	PUNCT
bjmsr-315	247	10	;	;	PUNCT
bjmsr-315	247	11	and	and	CCONJ
bjmsr-315	247	12	set	set	VERB
bjmsr-315	247	13	rσc(q)=0	rσc(q)=0	NOUN
bjmsr-315	247	14	.	.	PUNCT
bjmsr-315	248	1	step	step	NOUN
bjmsr-315	248	2	20	20	NUM
bjmsr-315	248	3	)	)	PUNCT
bjmsr-315	248	4	go	go	VERB
bjmsr-315	248	5	to	to	PART
bjmsr-315	248	6	step	step	VERB
bjmsr-315	248	7	5	5	NUM
bjmsr-315	248	8	.	.	PUNCT
bjmsr-315	249	1	seneta	seneta	PROPN
bjmsr-315	249	2	(	(	PUNCT
bjmsr-315	249	3	1983	1983	NUM
bjmsr-315	249	4	)	)	PUNCT
bjmsr-315	249	5	reviews	review	VERB
bjmsr-315	249	6	the	the	DET
bjmsr-315	249	7	iterative	iterative	NOUN
bjmsr-315	249	8	use	use	NOUN
bjmsr-315	249	9	of	of	ADP
bjmsr-315	249	10	weighted	weight	VERB
bjmsr-315	249	11	median	median	NOUN
bjmsr-315	249	12	to	to	PART
bjmsr-315	249	13	estimate	estimate	VERB
bjmsr-315	249	14	the	the	DET
bjmsr-315	249	15	parameters	parameter	NOUN
bjmsr-315	249	16	vector	vector	VERB
bjmsr-315	249	17	in	in	ADP
bjmsr-315	249	18	the	the	DET
bjmsr-315	249	19	classical	classical	ADJ
bjmsr-315	249	20	linear	linear	NOUN
bjmsr-315	249	21	model	model	NOUN
bjmsr-315	249	22	when	when	SCONJ
bjmsr-315	249	23	the	the	DET
bjmsr-315	249	24	fitting	fitting	ADJ
bjmsr-315	249	25	criterion	criterion	NOUN
bjmsr-315	249	26	is	be	AUX
bjmsr-315	249	27	l1	l1	PROPN
bjmsr-315	249	28	norm	norm	NOUN
bjmsr-315	249	29	and	and	CCONJ
bjmsr-315	249	30	also	also	ADV
bjmsr-315	249	31	cauchy	cauchy	ADJ
bjmsr-315	249	32	criterion	criterion	NOUN
bjmsr-315	249	33	.	.	PUNCT
bjmsr-315	250	1	wesolowsky	wesolowsky	PROPN
bjmsr-315	250	2	(	(	PUNCT
bjmsr-315	250	3	1981	1981	NUM
bjmsr-315	250	4	)	)	PUNCT
bjmsr-315	250	5	presents	present	VERB
bjmsr-315	250	6	an	an	DET
bjmsr-315	250	7	algorithm	algorithm	NOUN
bjmsr-315	250	8	for	for	ADP
bjmsr-315	250	9	multiple	multiple	ADJ
bjmsr-315	250	10	l1	l1	PROPN
bjmsr-315	250	11	norm	norm	NOUN
bjmsr-315	250	12	regression	regression	NOUN
bjmsr-315	250	13	based	base	VERB
bjmsr-315	250	14	on	on	ADP
bjmsr-315	250	15	the	the	DET
bjmsr-315	250	16	notion	notion	NOUN
bjmsr-315	250	17	of	of	ADP
bjmsr-315	250	18	edge	edge	ADJ
bjmsr-315	250	19	descent	descent	NOUN
bjmsr-315	250	20	along	along	ADP
bjmsr-315	250	21	the	the	DET
bjmsr-315	250	22	polyhedron	polyhedron	NOUN
bjmsr-315	250	23	of	of	ADP
bjmsr-315	250	24	the	the	DET
bjmsr-315	250	25	objective	objective	ADJ
bjmsr-315	250	26	function	function	NOUN
bjmsr-315	250	27	.	.	PUNCT
bjmsr-315	251	1	this	this	DET
bjmsr-315	251	2	algorithm	algorithm	NOUN
bjmsr-315	251	3	is	be	AUX
bjmsr-315	251	4	closely	closely	ADV
bjmsr-315	251	5	related	relate	VERB
bjmsr-315	251	6	to	to	ADP
bjmsr-315	251	7	those	those	PRON
bjmsr-315	251	8	of	of	ADP
bjmsr-315	251	9	rhodes	rhode	NOUN
bjmsr-315	251	10	(	(	PUNCT
bjmsr-315	251	11	1930	1930	NUM
bjmsr-315	251	12	)	)	PUNCT
bjmsr-315	251	13	and	and	CCONJ
bjmsr-315	251	14	bartels	bartel	NOUN
bjmsr-315	251	15	and	and	CCONJ
bjmsr-315	251	16	conn	conn	PROPN
bjmsr-315	251	17	and	and	CCONJ
bjmsr-315	251	18	sinclair	sinclair	PROPN
bjmsr-315	251	19	(	(	PUNCT
bjmsr-315	251	20	1978	1978	NUM
bjmsr-315	251	21	)	)	PUNCT
bjmsr-315	251	22	which	which	PRON
bjmsr-315	251	23	explained	explain	VERB
bjmsr-315	251	24	before	before	ADV
bjmsr-315	251	25	.	.	PUNCT
bjmsr-315	252	1	consider	consider	VERB
bjmsr-315	252	2	the	the	DET
bjmsr-315	252	3	multiple	multiple	ADJ
bjmsr-315	252	4	linear	linear	ADJ
bjmsr-315	252	5	regression	regression	NOUN
bjmsr-315	252	6	as	as	ADP
bjmsr-315	252	7	before	before	ADV
bjmsr-315	252	8	.	.	PUNCT
bjmsr-315	253	1	select	select	VERB
bjmsr-315	253	2	a	a	DET
bjmsr-315	253	3	set	set	NOUN
bjmsr-315	253	4	of	of	ADP
bjmsr-315	253	5	m	m	PROPN
bjmsr-315	253	6	points	point	NOUN
bjmsr-315	253	7	(	(	PUNCT
bjmsr-315	253	8	xj1	xj1	NOUN
bjmsr-315	253	9	i,	i,	NUM
bjmsr-315	253	10	...	...	PUNCT
bjmsr-315	253	11	,xjm	,xjm	PUNCT
bjmsr-315	253	12	i	i	PRON
bjmsr-315	253	13	,	,	PUNCT
bjmsr-315	253	14	yj	yj	PROPN
bjmsr-315	253	15	i	i	PROPN
bjmsr-315	253	16	)	)	PUNCT
bjmsr-315	253	17	.	.	PUNCT
bjmsr-315	254	1	the	the	DET
bjmsr-315	254	2	following	follow	VERB
bjmsr-315	254	3	system	system	NOUN
bjmsr-315	254	4	of	of	ADP
bjmsr-315	254	5	equations	equation	NOUN
bjmsr-315	254	6	can	can	AUX
bjmsr-315	254	7	be	be	AUX
bjmsr-315	254	8	solved	solve	VERB
bjmsr-315	254	9	for	for	ADP
bjmsr-315	254	10	a	a	DET
bjmsr-315	254	11	unique	unique	ADJ
bjmsr-315	254	12	set	set	NOUN
bjmsr-315	254	13	of	of	ADP
bjmsr-315	254	14	coefficients	coefficient	NOUN
bjmsr-315	254	15	.	.	PUNCT
bjmsr-315	255	1	m	m	PROPN
bjmsr-315	255	2	yj	yj	INTJ
bjmsr-315	255	3	i	i	PROPN
bjmsr-315	255	4	σ	σ	PROPN
bjmsr-315	255	5	ßhxjh	ßhxjh	NOUN
bjmsr-315	256	1	i	i	NOUN
bjmsr-315	256	2	=	=	SYM
bjmsr-315	256	3	0	0	NUM
bjmsr-315	256	4	,	,	PUNCT
bjmsr-315	256	5	j=1,	j=1,	NOUN
bjmsr-315	256	6	...	...	PUNCT
bjmsr-315	256	7	,m	,m	PUNCT
bjmsr-315	256	8	(	(	PUNCT
bjmsr-315	256	9	26	26	NUM
bjmsr-315	256	10	)	)	PUNCT
bjmsr-315	256	11	h=1	h=1	NOUN
bjmsr-315	256	12	an	an	DET
bjmsr-315	256	13	edge	edge	NOUN
bjmsr-315	256	14	is	be	AUX
bjmsr-315	256	15	formed	form	VERB
bjmsr-315	256	16	of	of	ADP
bjmsr-315	256	17	any	any	DET
bjmsr-315	256	18	subset	subset	NOUN
bjmsr-315	256	19	j	j	PROPN
bjmsr-315	256	20	consisting	consisting	NOUN
bjmsr-315	256	21	of	of	ADP
bjmsr-315	256	22	m-1	m-1	PROPN
bjmsr-315	256	23	equations	equation	NOUN
bjmsr-315	256	24	.	.	PUNCT
bjmsr-315	257	1	to	to	PART
bjmsr-315	257	2	minimize	minimize	VERB
bjmsr-315	257	3	along	along	ADP
bjmsr-315	257	4	an	an	DET
bjmsr-315	257	5	edge	edge	NOUN
bjmsr-315	257	6	,	,	PUNCT
bjmsr-315	257	7	set	set	VERB
bjmsr-315	257	8	:	:	PUNCT
bjmsr-315	257	9	1	1	NUM
bjmsr-315	257	10	11	11	NUM
bjmsr-315	257	11	1	1	NUM
bjmsr-315	257	12	1	1	NUM
bjmsr-315	257	13	11	11	NUM
bjmsr-315	257	14	1	1	NUM
bjmsr-315	257	15	1	1	NUM
bjmsr-315	257	16	1	1	NUM
bjmsr-315	257	17	1,1	1,1	NUM
bjmsr-315	257	18	1	1	NUM
bjmsr-315	257	19	,	,	PUNCT
bjmsr-315	257	20	,	,	PUNCT
bjmsr-315	257	21	,	,	PUNCT
bjmsr-315	257	22	,	,	PUNCT
bjmsr-315	257	23	i	i	PRON
bjmsr-315	257	24	i	i	PRON
bjmsr-315	258	1	i	i	VERB
bjmsr-315	258	2	j	j	PROPN
bjmsr-315	259	1	j	j	PROPN
bjmsr-315	259	2	j	j	PROPN
bjmsr-315	259	3	m	m	VERB
bjmsr-315	259	4	m	m	VERB
bjmsr-315	260	1	i	i	PRON
bjmsr-315	260	2	i	i	PRON
bjmsr-315	261	1	i	i	PRON
bjmsr-315	261	2	j	j	PROPN
bjmsr-315	262	1	j	j	PROPN
bjmsr-315	262	2	j	j	PROPN
bjmsr-315	262	3	m	m	VERB
bjmsr-315	262	4	m	m	VERB
bjmsr-315	262	5	mm	mm	INTJ
bjmsr-315	262	6	m	m	NOUN
bjmsr-315	262	7	m	m	VERB
bjmsr-315	262	8	m	m	VERB
bjmsr-315	262	9	m	m	VERB
bjmsr-315	262	10	y	y	NOUN
bjmsr-315	262	11	x	x	PUNCT
bjmsr-315	262	12	x	x	PUNCT
bjmsr-315	262	13	y	y	NOUN
bjmsr-315	262	14	x	x	PUNCT
bjmsr-315	262	15	x	x	PUNCT
bjmsr-315	262	16	y	y	NOUN
bjmsr-315	262	17	x	x	PUNCT
bjmsr-315	262	18	x	x	VERB
bjmsr-315	262	19	y	y	PROPN
bjmsr-315	262	20	x	x	SYM
bjmsr-315	262	21	x	x	PROPN
bjmsr-315	262	22			PROPN
bjmsr-315	262	23			PROPN
bjmsr-315	262	24			PROPN
bjmsr-315	262	25			NOUN
bjmsr-315	262	26			NOUN
bjmsr-315	262	27			NOUN
bjmsr-315	262	28			PROPN
bjmsr-315	262	29			NOUN
bjmsr-315	263	1			NOUN
bjmsr-315	263	2			PROPN
bjmsr-315	263	3			PROPN
bjmsr-315	263	4			PROPN
bjmsr-315	264	1			PROPN
bjmsr-315	264	2			PROPN
bjmsr-315	264	3			PROPN
bjmsr-315	264	4			NOUN
bjmsr-315	264	5			NOUN
bjmsr-315	265	1			NUM
bjmsr-315	266	1			NUM
bjmsr-315	266	2			NOUN
bjmsr-315	266	3			PUNCT
bjmsr-315	266	4			PROPN
bjmsr-315	267	1			PROPN
bjmsr-315	267	2			PROPN
bjmsr-315	267	3			PROPN
bjmsr-315	267	4			NOUN
bjmsr-315	267	5			PROPN
bjmsr-315	268	1			PROPN
bjmsr-315	269	1			PROPN
bjmsr-315	270	1			PROPN
bjmsr-315	270	2			PROPN
bjmsr-315	270	3			PROPN
bjmsr-315	270	4			NOUN
bjmsr-315	270	5			NOUN
bjmsr-315	270	6			NOUN
bjmsr-315	270	7			PROPN
bjmsr-315	270	8			PROPN
bjmsr-315	270	9			PROPN
bjmsr-315	270	10			VERB
bjmsr-315	270	11			PROPN
bjmsr-315	270	12			PROPN
bjmsr-315	270	13			NOUN
bjmsr-315	271	1	i	i	PRON
bjmsr-315	271	2	i	i	PRON
bjmsr-315	271	3	j	j	VERB
bjmsr-315	272	1	jy	jy	PROPN
bjmsr-315	272	2	x	x	SYM
bjmsr-315	272	3	y	y	PROPN
bjmsr-315	272	4	x	x	X
bjmsr-315	272	5	(	(	PUNCT
bjmsr-315	272	6	27	27	NUM
bjmsr-315	272	7	)	)	PUNCT
bjmsr-315	272	8	copyright	copyright	NOUN
bjmsr-315	272	9	©	©	PROPN
bjmsr-315	272	10	cc	cc	PROPN
bjmsr-315	272	11	-	-	PUNCT
bjmsr-315	272	12	by	by	ADP
bjmsr-315	272	13	-	-	PUNCT
bjmsr-315	272	14	nc	nc	PROPN
bjmsr-315	272	15	2019	2019	NUM
bjmsr-315	272	16	,	,	PUNCT
bjmsr-315	272	17	bjmsr	bjmsr	PROPN
bjmsr-315	272	18	www.cribfb.com/journal/index.php/bjmsr	www.cribfb.com/journal/index.php/bjmsr	X
bjmsr-315	272	19	bangladesh	bangladesh	PROPN
bjmsr-315	272	20	journal	journal	PROPN
bjmsr-315	272	21	of	of	ADP
bjmsr-315	272	22	multidisciplinary	multidisciplinary	ADJ
bjmsr-315	272	23	scientific	scientific	ADJ
bjmsr-315	272	24	research	research	NOUN
bjmsr-315	272	25	vol	vol	NOUN
bjmsr-315	272	26	.	.	PROPN
bjmsr-315	273	1	1	1	NUM
bjmsr-315	273	2	,	,	PUNCT
bjmsr-315	273	3	no	no	INTJ
bjmsr-315	273	4	.	.	NOUN
bjmsr-315	273	5	1	1	NUM
bjmsr-315	273	6	;	;	PUNCT
bjmsr-315	273	7	2019	2019	NUM
bjmsr-315	273	8	57	57	NUM
bjmsr-315	273	9	let	let	VERB
bjmsr-315	273	10	ßp	ßp	PRON
bjmsr-315	273	11	be	be	AUX
bjmsr-315	273	12	formed	form	VERB
bjmsr-315	273	13	from	from	ADP
bjmsr-315	273	14	ß	ß	PRON
bjmsr-315	273	15	by	by	ADP
bjmsr-315	273	16	deleting	delete	VERB
bjmsr-315	273	17	ßp	ßp	ADP
bjmsr-315	273	18	and	and	CCONJ
bjmsr-315	273	19	let	let	VERB
bjmsr-315	273	20	xp	xp	INTJ
bjmsr-315	273	21	be	be	AUX
bjmsr-315	273	22	the	the	DET
bjmsr-315	273	23	pth	pth	NOUN
bjmsr-315	273	24	column	column	NOUN
bjmsr-315	273	25	in	in	ADP
bjmsr-315	273	26	xj	xj	PROPN
bjmsr-315	273	27	and	and	CCONJ
bjmsr-315	273	28	let	let	VERB
bjmsr-315	273	29	xpj	xpj	PROPN
bjmsr-315	273	30	be	be	AUX
bjmsr-315	273	31	formed	form	VERB
bjmsr-315	273	32	by	by	ADP
bjmsr-315	273	33	removing	remove	VERB
bjmsr-315	273	34	xp	xp	INTJ
bjmsr-315	273	35	from	from	ADP
bjmsr-315	273	36	xj	xj	PROPN
bjmsr-315	273	37	.	.	PUNCT
bjmsr-315	274	1	then	then	ADV
bjmsr-315	274	2	for	for	ADP
bjmsr-315	274	3	a	a	DET
bjmsr-315	274	4	given	give	VERB
bjmsr-315	274	5	ßp	ßp	PRON
bjmsr-315	274	6	it	it	PRON
bjmsr-315	274	7	can	can	AUX
bjmsr-315	274	8	be	be	AUX
bjmsr-315	274	9	shown	show	VERB
bjmsr-315	274	10	that	that	SCONJ
bjmsr-315	274	11	,	,	PUNCT
bjmsr-315	274	12	ßp	ßp	PRON
bjmsr-315	274	13	=	=	PUNCT
bjmsr-315	274	14	xpj	xpj	PROPN
bjmsr-315	274	15	-1yj	-1yj	PROPN
bjmsr-315	274	16	ßpxpj	ßpxpj	VERB
bjmsr-315	274	17	-1xp	-1xp	NUM
bjmsr-315	274	18	(	(	PUNCT
bjmsr-315	274	19	28	28	NUM
bjmsr-315	274	20	)	)	PUNCT
bjmsr-315	274	21	let	let	VERB
bjmsr-315	274	22	the	the	DET
bjmsr-315	274	23	elements	element	NOUN
bjmsr-315	274	24	of	of	ADP
bjmsr-315	274	25	ßp	ßp	INTJ
bjmsr-315	274	26	be	be	AUX
bjmsr-315	275	1	ßq	ßq	ADV
bjmsr-315	275	2	=	=	PUNCT
bjmsr-315	275	3	rq	rq	NOUN
bjmsr-315	275	4	sqßp	sqßp	NOUN
bjmsr-315	275	5	for	for	ADP
bjmsr-315	275	6	q	q	PROPN
bjmsr-315	275	7	≠	≠	PROPN
bjmsr-315	275	8	m	m	PROPN
bjmsr-315	275	9	(	(	PUNCT
bjmsr-315	275	10	29	29	NUM
bjmsr-315	275	11	)	)	PUNCT
bjmsr-315	275	12	substitution	substitution	NOUN
bjmsr-315	275	13	in	in	ADP
bjmsr-315	275	14	the	the	DET
bjmsr-315	275	15	l1	l1	PROPN
bjmsr-315	275	16	norm	norm	NOUN
bjmsr-315	275	17	objective	objective	ADJ
bjmsr-315	275	18	function	function	NOUN
bjmsr-315	275	19	gives	give	VERB
bjmsr-315	275	20	,	,	PUNCT
bjmsr-315	275	21	1	1	NUM
bjmsr-315	275	22	1	1	NUM
bjmsr-315	275	23	1	1	NUM
bjmsr-315	275	24	2	2	NUM
bjmsr-315	275	25	min	min	NOUN
bjmsr-315	275	26	:	:	PUNCT
bjmsr-315	276	1	m	m	AUX
bjmsr-315	276	2	q	q	NOUN
bjmsr-315	276	3	iq	iq	INTJ
bjmsr-315	276	4	ip	ip	VERB
bjmsr-315	276	5	q	q	NOUN
bjmsr-315	277	1	p	p	NOUN
bjmsr-315	278	1	m	m	VERB
bjmsr-315	279	1	i	i	PRON
bjmsr-315	279	2	q	q	NOUN
bjmsr-315	279	3	iqn	iqn	VERB
bjmsr-315	279	4	m	m	VERB
bjmsr-315	279	5	q	q	X
bjmsr-315	279	6	p	p	NOUN
bjmsr-315	279	7	ip	ip	PROPN
bjmsr-315	279	8	q	q	NOUN
bjmsr-315	279	9	iq	iq	INTJ
bjmsr-315	279	10	m	m	INTJ
bjmsr-315	280	1	i	i	PRON
bjmsr-315	280	2	q	q	PROPN
bjmsr-315	280	3	p	p	PROPN
bjmsr-315	280	4	ip	ip	PROPN
bjmsr-315	280	5	q	q	NOUN
bjmsr-315	280	6	iq	iq	NOUN
bjmsr-315	280	7	qs	qs	PROPN
bjmsr-315	280	8	x	x	PUNCT
bjmsr-315	280	9	x	x	PUNCT
bjmsr-315	280	10	y	y	NOUN
bjmsr-315	280	11	r	r	NOUN
bjmsr-315	280	12	x	x	PUNCT
bjmsr-315	281	1	x	x	X
bjmsr-315	281	2	s	s	AUX
bjmsr-315	281	3	x	x	X
bjmsr-315	281	4	x	x	X
bjmsr-315	281	5	s	s	PROPN
bjmsr-315	281	6	x	x	PROPN
bjmsr-315	281	7			PROPN
bjmsr-315	281	8			PROPN
bjmsr-315	281	9			VERB
bjmsr-315	281	10			PROPN
bjmsr-315	281	11			NOUN
bjmsr-315	281	12			PUNCT
bjmsr-315	281	13			VERB
bjmsr-315	281	14			X
bjmsr-315	281	15			PROPN
bjmsr-315	281	16			SYM
bjmsr-315	281	17			PUNCT
bjmsr-315	281	18			VERB
bjmsr-315	281	19			X
bjmsr-315	281	20			X
bjmsr-315	281	21	(	(	PUNCT
bjmsr-315	281	22	30	30	NUM
bjmsr-315	281	23	)	)	PUNCT
bjmsr-315	281	24	now	now	ADV
bjmsr-315	281	25	the	the	DET
bjmsr-315	281	26	following	follow	VERB
bjmsr-315	281	27	steps	step	NOUN
bjmsr-315	281	28	should	should	AUX
bjmsr-315	281	29	be	be	AUX
bjmsr-315	281	30	taken	take	VERB
bjmsr-315	281	31	.	.	PUNCT
bjmsr-315	282	1	step	step	NOUN
bjmsr-315	282	2	1	1	NUM
bjmsr-315	282	3	)	)	PUNCT
bjmsr-315	282	4	set	set	VERB
bjmsr-315	282	5	k=1	k=1	PROPN
bjmsr-315	282	6	,	,	PUNCT
bjmsr-315	282	7	l=0	l=0	PROPN
bjmsr-315	282	8	,	,	PUNCT
bjmsr-315	282	9	choose	choose	VERB
bjmsr-315	282	10	initial	initial	ADJ
bjmsr-315	282	11	values	value	NOUN
bjmsr-315	282	12	for	for	ADP
bjmsr-315	282	13	ß1,	ß1,	NOUN
bjmsr-315	282	14	...	...	PUNCT
bjmsr-315	282	15	,ßm	,ßm	PROPN
bjmsr-315	282	16	.	.	PUNCT
bjmsr-315	283	1	least	least	ADJ
bjmsr-315	283	2	squares	square	NOUN
bjmsr-315	283	3	values	value	NOUN
bjmsr-315	283	4	are	be	AUX
bjmsr-315	283	5	one	one	NUM
bjmsr-315	283	6	possibility	possibility	NOUN
bjmsr-315	283	7	.	.	PUNCT
bjmsr-315	284	1	let	let	VERB
bjmsr-315	284	2	i(1)={j1	i(1)={j1	PROPN
bjmsr-315	284	3	(	(	PUNCT
bjmsr-315	284	4	1),	1),	NUM
bjmsr-315	284	5	...	...	PUNCT
bjmsr-315	284	6	,jm	,jm	PUNCT
bjmsr-315	284	7	(	(	PUNCT
bjmsr-315	284	8	1	1	NUM
bjmsr-315	284	9	)	)	PUNCT
bjmsr-315	284	10	}	}	PUNCT
bjmsr-315	284	11	be	be	AUX
bjmsr-315	284	12	a	a	DET
bjmsr-315	284	13	set	set	NOUN
bjmsr-315	284	14	of	of	ADP
bjmsr-315	284	15	m	m	PROPN
bjmsr-315	284	16	data	data	NOUN
bjmsr-315	284	17	points	point	NOUN
bjmsr-315	284	18	chosen	choose	VERB
bjmsr-315	284	19	in	in	ADP
bjmsr-315	284	20	sequence	sequence	NOUN
bjmsr-315	284	21	as	as	SCONJ
bjmsr-315	284	22	follows	follow	VERB
bjmsr-315	284	23	.	.	PUNCT
bjmsr-315	285	1	the	the	DET
bjmsr-315	285	2	point	point	NOUN
bjmsr-315	285	3	with	with	ADP
bjmsr-315	285	4	the	the	DET
bjmsr-315	285	5	smallest	small	ADJ
bjmsr-315	285	6	squared	square	VERB
bjmsr-315	285	7	residual	residual	ADJ
bjmsr-315	285	8	is	be	AUX
bjmsr-315	285	9	chosen	choose	VERB
bjmsr-315	285	10	each	each	DET
bjmsr-315	285	11	time	time	NOUN
bjmsr-315	285	12	subject	subject	ADJ
bjmsr-315	285	13	to	to	ADP
bjmsr-315	285	14	the	the	DET
bjmsr-315	285	15	condition	condition	NOUN
bjmsr-315	285	16	that	that	SCONJ
bjmsr-315	285	17	xi	xi	X
bjmsr-315	285	18	(	(	PUNCT
bjmsr-315	285	19	1	1	X
bjmsr-315	285	20	)	)	PUNCT
bjmsr-315	285	21	is	be	AUX
bjmsr-315	285	22	nonsingular	nonsingular	ADJ
bjmsr-315	285	23	.	.	PUNCT
bjmsr-315	286	1	find	find	VERB
bjmsr-315	286	2	ßq	ßq	INTJ
bjmsr-315	286	3	(	(	PUNCT
bjmsr-315	286	4	1	1	NUM
bjmsr-315	286	5	)	)	PUNCT
bjmsr-315	286	6	for	for	ADP
bjmsr-315	286	7	q=1,	q=1,	NOUN
bjmsr-315	286	8	...	...	PUNCT
bjmsr-315	286	9	,m	,m	PUNCT
bjmsr-315	286	10	,	,	PUNCT
bjmsr-315	286	11	by	by	ADP
bjmsr-315	286	12	solving	solve	VERB
bjmsr-315	286	13	yi	yi	NOUN
bjmsr-315	286	14	(	(	PUNCT
bjmsr-315	286	15	1)=xi	1)=xi	NUM
bjmsr-315	286	16	(	(	PUNCT
bjmsr-315	286	17	1)ß	1)ß	NUM
bjmsr-315	286	18	.	.	PUNCT
bjmsr-315	287	1	set	set	VERB
bjmsr-315	287	2	i(k)=(j1	i(k)=(j1	PROPN
bjmsr-315	287	3	(	(	PUNCT
bjmsr-315	287	4	k),	k),	PROPN
bjmsr-315	287	5	...	...	PUNCT
bjmsr-315	287	6	,jm	,jm	PUNCT
bjmsr-315	287	7	(	(	PUNCT
bjmsr-315	287	8	k	k	NOUN
bjmsr-315	287	9	)	)	PUNCT
bjmsr-315	287	10	)	)	PUNCT
bjmsr-315	287	11	;	;	PUNCT
bjmsr-315	287	12	j=(j2	j=(j2	PROPN
bjmsr-315	287	13	(	(	PUNCT
bjmsr-315	287	14	k),	k),	PROPN
bjmsr-315	287	15	...	...	PUNCT
bjmsr-315	287	16	,jm	,jm	PUNCT
bjmsr-315	287	17	(	(	PUNCT
bjmsr-315	287	18	k	k	NOUN
bjmsr-315	287	19	)	)	PUNCT
bjmsr-315	287	20	)	)	PUNCT
bjmsr-315	287	21	.	.	PUNCT
bjmsr-315	288	1	step	step	NOUN
bjmsr-315	288	2	2	2	NUM
bjmsr-315	288	3	)	)	PUNCT
bjmsr-315	288	4	set	set	VERB
bjmsr-315	288	5	k	k	X
bjmsr-315	288	6	=	=	VERB
bjmsr-315	288	7	k+1	k+1	X
bjmsr-315	288	8	.	.	X
bjmsr-315	288	9	obtain	obtain	VERB
bjmsr-315	288	10	ßp	ßp	NOUN
bjmsr-315	288	11	for	for	ADP
bjmsr-315	288	12	the	the	DET
bjmsr-315	288	13	smallest	small	ADJ
bjmsr-315	288	14	p	p	NOUN
bjmsr-315	288	15	from	from	ADP
bjmsr-315	288	16	(	(	PUNCT
bjmsr-315	288	17	30	30	NUM
bjmsr-315	288	18	)	)	PUNCT
bjmsr-315	288	19	by	by	ADP
bjmsr-315	288	20	using	use	VERB
bjmsr-315	288	21	the	the	DET
bjmsr-315	288	22	weighted	weight	VERB
bjmsr-315	288	23	median	median	ADJ
bjmsr-315	288	24	procedure	procedure	NOUN
bjmsr-315	288	25	.	.	PUNCT
bjmsr-315	289	1	set	set	VERB
bjmsr-315	289	2	ßp	ßp	INTJ
bjmsr-315	289	3	(	(	PUNCT
bjmsr-315	289	4	k)=ßp	k)=ßp	PROPN
bjmsr-315	289	5	and	and	CCONJ
bjmsr-315	289	6	let	let	VERB
bjmsr-315	289	7	i	i	PRON
bjmsr-315	289	8	be	be	AUX
bjmsr-315	289	9	the	the	DET
bjmsr-315	289	10	index	index	NOUN
bjmsr-315	289	11	which	which	PRON
bjmsr-315	289	12	defines	define	VERB
bjmsr-315	289	13	the	the	DET
bjmsr-315	289	14	lower	low	ADJ
bjmsr-315	289	15	weighted	weight	VERB
bjmsr-315	289	16	median	median	NOUN
bjmsr-315	289	17	ßp	ßp	PRON
bjmsr-315	289	18	in	in	ADP
bjmsr-315	289	19	(	(	PUNCT
bjmsr-315	289	20	30	30	NUM
bjmsr-315	289	21	)	)	PUNCT
bjmsr-315	289	22	for	for	ADP
bjmsr-315	289	23	the	the	DET
bjmsr-315	289	24	lowest	low	ADJ
bjmsr-315	289	25	possible	possible	ADJ
bjmsr-315	289	26	p.	p.	NOUN
bjmsr-315	289	27	step	step	NOUN
bjmsr-315	289	28	3	3	NUM
bjmsr-315	289	29	)	)	PUNCT
bjmsr-315	289	30	a	a	X
bjmsr-315	289	31	)	)	PUNCT
bjmsr-315	289	32	if	if	SCONJ
bjmsr-315	289	33	ßp	ßp	PRON
bjmsr-315	289	34	(	(	PUNCT
bjmsr-315	289	35	k)-ßp	k)-ßp	PROPN
bjmsr-315	289	36	(	(	PUNCT
bjmsr-315	289	37	k-1	k-1	PROPN
bjmsr-315	289	38	)	)	PUNCT
bjmsr-315	290	1	=	=	SYM
bjmsr-315	290	2	0	0	PUNCT
bjmsr-315	291	1	and	and	CCONJ
bjmsr-315	291	2	if	if	SCONJ
bjmsr-315	291	3	l	l	PROPN
bjmsr-315	291	4	>	>	X
bjmsr-315	291	5	m	m	PROPN
bjmsr-315	291	6	,	,	PUNCT
bjmsr-315	291	7	go	go	VERB
bjmsr-315	291	8	to	to	PART
bjmsr-315	291	9	step	step	VERB
bjmsr-315	291	10	4	4	NUM
bjmsr-315	291	11	.	.	PUNCT
bjmsr-315	292	1	otherwise	otherwise	ADV
bjmsr-315	292	2	,	,	PUNCT
bjmsr-315	292	3	set	set	VERB
bjmsr-315	292	4	i=(j2	i=(j2	NOUN
bjmsr-315	292	5	(	(	PUNCT
bjmsr-315	292	6	k-1),	k-1),	ADJ
bjmsr-315	292	7	...	...	PUNCT
bjmsr-315	292	8	,jm	,jm	PUNCT
bjmsr-315	292	9	(	(	PUNCT
bjmsr-315	292	10	k-1),i	k-1),i	PROPN
bjmsr-315	292	11	)	)	PUNCT
bjmsr-315	292	12	and	and	CCONJ
bjmsr-315	292	13	l	l	NOUN
bjmsr-315	292	14	=	=	NOUN
bjmsr-315	292	15	l+1	l+1	PROPN
bjmsr-315	292	16	,	,	PUNCT
bjmsr-315	292	17	ßq	ßq	INTJ
bjmsr-315	292	18	(	(	PUNCT
bjmsr-315	292	19	k)=ßq	k)=ßq	PROPN
bjmsr-315	292	20	(	(	PUNCT
bjmsr-315	292	21	k-1	k-1	PROPN
bjmsr-315	292	22	)	)	PUNCT
bjmsr-315	292	23	for	for	ADP
bjmsr-315	292	24	all	all	DET
bjmsr-315	292	25	q	q	PUNCT
bjmsr-315	292	26	and	and	CCONJ
bjmsr-315	292	27	go	go	VERB
bjmsr-315	292	28	to	to	PART
bjmsr-315	292	29	step	step	VERB
bjmsr-315	292	30	2	2	NUM
bjmsr-315	292	31	.	.	PUNCT
bjmsr-315	293	1	b	b	X
bjmsr-315	293	2	)	)	PUNCT
bjmsr-315	293	3	if	if	SCONJ
bjmsr-315	293	4	ßp	ßp	PRON
bjmsr-315	293	5	(	(	PUNCT
bjmsr-315	293	6	k)-ßp	k)-ßp	PROPN
bjmsr-315	293	7	(	(	PUNCT
bjmsr-315	293	8	k-1)≠0	k-1)≠0	PROPN
bjmsr-315	293	9	,	,	PUNCT
bjmsr-315	293	10	set	set	VERB
bjmsr-315	293	11	l=0	l=0	PROPN
bjmsr-315	293	12	.	.	PUNCT
bjmsr-315	294	1	calculate	calculate	NOUN
bjmsr-315	294	2	ßq	ßq	INTJ
bjmsr-315	294	3	(	(	PUNCT
bjmsr-315	294	4	k	k	NOUN
bjmsr-315	294	5	)	)	PUNCT
bjmsr-315	294	6	for	for	ADP
bjmsr-315	294	7	q≠p	q≠p	NOUN
bjmsr-315	294	8	from	from	ADP
bjmsr-315	294	9	(	(	PUNCT
bjmsr-315	294	10	29	29	NUM
bjmsr-315	294	11	)	)	PUNCT
bjmsr-315	294	12	.	.	PUNCT
bjmsr-315	295	1	set	set	VERB
bjmsr-315	295	2	i(k)=(j2	i(k)=(j2	NOUN
bjmsr-315	295	3	(	(	PUNCT
bjmsr-315	295	4	k-1),	k-1),	ADJ
bjmsr-315	295	5	...	...	PUNCT
bjmsr-315	295	6	,jm	,jm	PUNCT
bjmsr-315	295	7	(	(	PUNCT
bjmsr-315	295	8	k-1),i	k-1),i	X
bjmsr-315	295	9	)	)	PUNCT
bjmsr-315	295	10	and	and	CCONJ
bjmsr-315	295	11	go	go	VERB
bjmsr-315	295	12	to	to	PART
bjmsr-315	295	13	step	step	VERB
bjmsr-315	295	14	2	2	NUM
bjmsr-315	295	15	.	.	PUNCT
bjmsr-315	295	16	step	step	NOUN
bjmsr-315	295	17	4	4	NUM
bjmsr-315	295	18	)	)	PUNCT
bjmsr-315	295	19	calculate	calculate	NOUN
bjmsr-315	295	20	ßq	ßq	INTJ
bjmsr-315	295	21	(	(	PUNCT
bjmsr-315	295	22	k	k	NOUN
bjmsr-315	295	23	)	)	PUNCT
bjmsr-315	295	24	for	for	ADP
bjmsr-315	295	25	q≠p	q≠p	NOUN
bjmsr-315	295	26	from	from	ADP
bjmsr-315	295	27	(	(	PUNCT
bjmsr-315	295	28	29	29	NUM
bjmsr-315	295	29	)	)	PUNCT
bjmsr-315	295	30	;	;	PUNCT
bjmsr-315	295	31	set	set	VERB
bjmsr-315	295	32	ß*=ß(k	ß*=ß(k	NOUN
bjmsr-315	295	33	)	)	PUNCT
bjmsr-315	295	34	and	and	CCONJ
bjmsr-315	295	35	stop	stop	VERB
bjmsr-315	295	36	.	.	PUNCT
bjmsr-315	296	1	in	in	ADP
bjmsr-315	296	2	this	this	DET
bjmsr-315	296	3	paper	paper	NOUN
bjmsr-315	296	4	,	,	PUNCT
bjmsr-315	296	5	wesolowsky	wesolowsky	PROPN
bjmsr-315	296	6	also	also	ADV
bjmsr-315	296	7	discusses	discuss	VERB
bjmsr-315	296	8	the	the	DET
bjmsr-315	296	9	problem	problem	NOUN
bjmsr-315	296	10	of	of	ADP
bjmsr-315	296	11	multicolinearity	multicolinearity	NOUN
bjmsr-315	296	12	and	and	CCONJ
bjmsr-315	296	13	gives	give	VERB
bjmsr-315	296	14	an	an	DET
bjmsr-315	296	15	appropriate	appropriate	ADJ
bjmsr-315	296	16	solution	solution	NOUN
bjmsr-315	296	17	.	.	PUNCT
bjmsr-315	297	1	josvanger	josvanger	NOUN
bjmsr-315	297	2	and	and	CCONJ
bjmsr-315	297	3	sposito	sposito	X
bjmsr-315	297	4	(	(	PUNCT
bjmsr-315	297	5	1983	1983	NUM
bjmsr-315	297	6	)	)	PUNCT
bjmsr-315	297	7	modify	modify	VERB
bjmsr-315	297	8	wesolowsky	wesolowsky	PROPN
bjmsr-315	297	9	's	's	PART
bjmsr-315	297	10	algorithm	algorithm	NOUN
bjmsr-315	297	11	for	for	ADP
bjmsr-315	297	12	the	the	DET
bjmsr-315	297	13	two	two	NUM
bjmsr-315	297	14	-	-	PUNCT
bjmsr-315	297	15	parameter	parameter	NOUN
bjmsr-315	297	16	simple	simple	ADJ
bjmsr-315	297	17	linear	linear	PROPN
bjmsr-315	297	18	regression	regression	NOUN
bjmsr-315	297	19	model	model	NOUN
bjmsr-315	297	20	.	.	PUNCT
bjmsr-315	298	1	the	the	DET
bjmsr-315	298	2	modification	modification	NOUN
bjmsr-315	298	3	is	be	AUX
bjmsr-315	298	4	an	an	DET
bjmsr-315	298	5	alternative	alternative	ADJ
bjmsr-315	298	6	way	way	NOUN
bjmsr-315	298	7	to	to	PART
bjmsr-315	298	8	order	order	VERB
bjmsr-315	298	9	observations	observation	NOUN
bjmsr-315	298	10	instead	instead	ADV
bjmsr-315	298	11	of	of	ADP
bjmsr-315	298	12	sorting	sort	VERB
bjmsr-315	298	13	all	all	PRON
bjmsr-315	298	14	of	of	ADP
bjmsr-315	298	15	them	they	PRON
bjmsr-315	298	16	to	to	PART
bjmsr-315	298	17	find	find	VERB
bjmsr-315	298	18	the	the	DET
bjmsr-315	298	19	necessary	necessary	ADJ
bjmsr-315	298	20	weighted	weight	VERB
bjmsr-315	298	21	median	median	ADJ
bjmsr-315	298	22	value	value	NOUN
bjmsr-315	298	23	.	.	PUNCT
bjmsr-315	299	1	suppose	suppose	VERB
bjmsr-315	299	2	the	the	DET
bjmsr-315	299	3	problem	problem	NOUN
bjmsr-315	299	4	has	have	AUX
bjmsr-315	299	5	been	be	AUX
bjmsr-315	299	6	reduced	reduce	VERB
bjmsr-315	299	7	to	to	ADP
bjmsr-315	299	8	a	a	DET
bjmsr-315	299	9	weighted	weight	VERB
bjmsr-315	299	10	median	median	ADJ
bjmsr-315	299	11	problem	problem	NOUN
bjmsr-315	299	12	.	.	PUNCT
bjmsr-315	300	1	they	they	PRON
bjmsr-315	300	2	place	place	VERB
bjmsr-315	300	3	smaller	small	ADJ
bjmsr-315	300	4	values	value	NOUN
bjmsr-315	300	5	of	of	ADP
bjmsr-315	300	6	factors	factor	NOUN
bjmsr-315	300	7	to	to	PART
bjmsr-315	300	8	be	be	AUX
bjmsr-315	300	9	sorted	sort	VERB
bjmsr-315	300	10	with	with	ADP
bjmsr-315	300	11	corresponding	correspond	VERB
bjmsr-315	300	12	weights	weight	NOUN
bjmsr-315	300	13	below	below	ADP
bjmsr-315	300	14	ß1	ß1	PROPN
bjmsr-315	300	15	(	(	PUNCT
bjmsr-315	300	16	k-1	k-1	PROPN
bjmsr-315	300	17	)	)	PUNCT
bjmsr-315	300	18	and	and	CCONJ
bjmsr-315	300	19	larger	large	ADJ
bjmsr-315	300	20	or	or	CCONJ
bjmsr-315	300	21	equal	equal	ADJ
bjmsr-315	300	22	values	value	NOUN
bjmsr-315	300	23	above	above	ADP
bjmsr-315	300	24	it	it	PRON
bjmsr-315	300	25	,	,	PUNCT
bjmsr-315	300	26	then	then	ADV
bjmsr-315	300	27	recheck	recheck	VERB
bjmsr-315	300	28	the	the	DET
bjmsr-315	300	29	inequalities	inequality	NOUN
bjmsr-315	300	30	(	(	PUNCT
bjmsr-315	300	31	4	4	NUM
bjmsr-315	300	32	)	)	PUNCT
bjmsr-315	300	33	of	of	ADP
bjmsr-315	300	34	the	the	DET
bjmsr-315	300	35	weighted	weight	VERB
bjmsr-315	300	36	median	median	NOUN
bjmsr-315	300	37	.	.	PUNCT
bjmsr-315	301	1	if	if	SCONJ
bjmsr-315	301	2	the	the	DET
bjmsr-315	301	3	inequalities	inequality	NOUN
bjmsr-315	301	4	do	do	AUX
bjmsr-315	301	5	not	not	PART
bjmsr-315	301	6	satisfy	satisfy	VERB
bjmsr-315	301	7	,	,	PUNCT
bjmsr-315	301	8	then	then	ADV
bjmsr-315	301	9	an	an	DET
bjmsr-315	301	10	appropriate	appropriate	ADJ
bjmsr-315	301	11	adjustment	adjustment	NOUN
bjmsr-315	301	12	is	be	AUX
bjmsr-315	301	13	made	make	VERB
bjmsr-315	301	14	.	.	PUNCT
bjmsr-315	302	1	in	in	ADP
bjmsr-315	302	2	particular	particular	ADJ
bjmsr-315	302	3	,	,	PUNCT
bjmsr-315	302	4	if	if	SCONJ
bjmsr-315	302	5	the	the	DET
bjmsr-315	302	6	right	right	ADJ
bjmsr-315	302	7	-	-	PUNCT
bjmsr-315	302	8	hand	hand	NOUN
bjmsr-315	302	9	side	side	NOUN
bjmsr-315	302	10	is	be	AUX
bjmsr-315	302	11	overly	overly	ADV
bjmsr-315	302	12	weighted	weight	VERB
bjmsr-315	302	13	,	,	PUNCT
bjmsr-315	302	14	then	then	ADV
bjmsr-315	302	15	the	the	DET
bjmsr-315	302	16	weight	weight	NOUN
bjmsr-315	302	17	corresponding	correspond	VERB
bjmsr-315	302	18	to	to	ADP
bjmsr-315	302	19	the	the	DET
bjmsr-315	302	20	smallest	small	ADJ
bjmsr-315	302	21	sorting	sort	VERB
bjmsr-315	302	22	factor	factor	NOUN
bjmsr-315	302	23	is	be	AUX
bjmsr-315	302	24	transferred	transfer	VERB
bjmsr-315	302	25	to	to	ADP
bjmsr-315	302	26	the	the	DET
bjmsr-315	302	27	left	left	ADJ
bjmsr-315	302	28	-	-	PUNCT
bjmsr-315	302	29	hand	hand	NOUN
bjmsr-315	302	30	side	side	NOUN
bjmsr-315	302	31	,	,	PUNCT
bjmsr-315	302	32	and	and	CCONJ
bjmsr-315	302	33	the	the	DET
bjmsr-315	302	34	check	check	NOUN
bjmsr-315	302	35	is	be	AUX
bjmsr-315	302	36	made	make	VERB
bjmsr-315	302	37	again	again	ADV
bjmsr-315	302	38	.	.	PUNCT
bjmsr-315	303	1	a	a	DET
bjmsr-315	303	2	computer	computer	NOUN
bjmsr-315	303	3	program	program	NOUN
bjmsr-315	303	4	for	for	ADP
bjmsr-315	303	5	this	this	DET
bjmsr-315	303	6	algorithm	algorithm	NOUN
bjmsr-315	303	7	is	be	AUX
bjmsr-315	303	8	also	also	ADV
bjmsr-315	303	9	given	give	VERB
bjmsr-315	303	10	by	by	ADP
bjmsr-315	303	11	the	the	DET
bjmsr-315	303	12	authors	author	NOUN
bjmsr-315	303	13	.	.	PUNCT
bjmsr-315	304	1	"	"	PUNCT
bjmsr-315	304	2	generalized	generalized	ADJ
bjmsr-315	304	3	gradient	gradient	NOUN
bjmsr-315	304	4	"	"	PUNCT
bjmsr-315	304	5	method	method	NOUN
bjmsr-315	304	6	introduced	introduce	VERB
bjmsr-315	304	7	by	by	ADP
bjmsr-315	304	8	clarke	clarke	PROPN
bjmsr-315	304	9	(	(	PUNCT
bjmsr-315	304	10	see	see	PROPN
bjmsr-315	304	11	,	,	PUNCT
bjmsr-315	304	12	clarke	clarke	PROPN
bjmsr-315	304	13	(	(	PUNCT
bjmsr-315	304	14	1983	1983	NUM
bjmsr-315	304	15	)	)	PUNCT
bjmsr-315	304	16	)	)	PUNCT
bjmsr-315	304	17	is	be	AUX
bjmsr-315	304	18	a	a	DET
bjmsr-315	304	19	general	general	ADJ
bjmsr-315	304	20	procedure	procedure	NOUN
bjmsr-315	304	21	for	for	ADP
bjmsr-315	304	22	nonsmooth	nonsmooth	NOUN
bjmsr-315	304	23	optimization	optimization	NOUN
bjmsr-315	304	24	functions	function	NOUN
bjmsr-315	304	25	and	and	CCONJ
bjmsr-315	304	26	problems	problem	NOUN
bjmsr-315	304	27	(	(	PUNCT
bjmsr-315	304	28	see	see	VERB
bjmsr-315	304	29	,	,	PUNCT
bjmsr-315	304	30	osborne	osborne	NOUN
bjmsr-315	304	31	and	and	CCONJ
bjmsr-315	304	32	pruess	pruess	NOUN
bjmsr-315	304	33	and	and	CCONJ
bjmsr-315	304	34	womersley	womersley	PROPN
bjmsr-315	304	35	(	(	PUNCT
bjmsr-315	304	36	1986	1986	NUM
bjmsr-315	304	37	)	)	PUNCT
bjmsr-315	304	38	)	)	PUNCT
bjmsr-315	304	39	.	.	PUNCT
bjmsr-315	305	1	a	a	DET
bjmsr-315	305	2	subclass	subclass	NOUN
bjmsr-315	305	3	of	of	ADP
bjmsr-315	305	4	this	this	DET
bjmsr-315	305	5	method	method	NOUN
bjmsr-315	305	6	is	be	AUX
bjmsr-315	305	7	called	call	VERB
bjmsr-315	305	8	"	"	PUNCT
bjmsr-315	305	9	reduced	reduced	ADJ
bjmsr-315	305	10	gradient	gradient	NOUN
bjmsr-315	305	11	"	"	PUNCT
bjmsr-315	305	12	explained	explain	VERB
bjmsr-315	305	13	by	by	ADP
bjmsr-315	305	14	osborne	osborne	PROPN
bjmsr-315	305	15	(	(	PUNCT
bjmsr-315	305	16	1985	1985	NUM
bjmsr-315	305	17	)	)	PUNCT
bjmsr-315	305	18	is	be	AUX
bjmsr-315	305	19	a	a	DET
bjmsr-315	305	20	general	general	ADJ
bjmsr-315	305	21	algorithm	algorithm	NOUN
bjmsr-315	305	22	which	which	PRON
bjmsr-315	305	23	contains	contain	VERB
bjmsr-315	305	24	linear	linear	PROPN
bjmsr-315	305	25	programming	programming	NOUN
bjmsr-315	305	26	,	,	PUNCT
bjmsr-315	305	27	piecewise	piecewise	NOUN
bjmsr-315	305	28	linear	linear	ADJ
bjmsr-315	305	29	optimization	optimization	NOUN
bjmsr-315	305	30	problems	problem	NOUN
bjmsr-315	305	31	,	,	PUNCT
bjmsr-315	305	32	and	and	CCONJ
bjmsr-315	305	33	polyhedral	polyhedral	ADJ
bjmsr-315	305	34	convex	convex	NOUN
bjmsr-315	305	35	function	function	NOUN
bjmsr-315	305	36	optimization	optimization	NOUN
bjmsr-315	305	37	algorithms	algorithm	NOUN
bjmsr-315	305	38	inside	inside	ADV
bjmsr-315	305	39	.	.	PUNCT
bjmsr-315	306	1	the	the	DET
bjmsr-315	306	2	reduced	reduce	VERB
bjmsr-315	306	3	gradient	gradient	ADJ
bjmsr-315	306	4	algorithm	algorithm	NOUN
bjmsr-315	306	5	is	be	AUX
bjmsr-315	306	6	a	a	DET
bjmsr-315	306	7	copyright	copyright	NOUN
bjmsr-315	306	8	©	©	PROPN
bjmsr-315	306	9	cc	cc	NOUN
bjmsr-315	306	10	-	-	PUNCT
bjmsr-315	306	11	by	by	ADP
bjmsr-315	306	12	-	-	PUNCT
bjmsr-315	306	13	nc	nc	PROPN
bjmsr-315	306	14	2019	2019	NUM
bjmsr-315	306	15	,	,	PUNCT
bjmsr-315	306	16	bjmsr	bjmsr	PROPN
bjmsr-315	306	17	www.cribfb.com/journal/index.php/bjmsr	www.cribfb.com/journal/index.php/bjmsr	X
bjmsr-315	306	18	bangladesh	bangladesh	PROPN
bjmsr-315	306	19	journal	journal	PROPN
bjmsr-315	306	20	of	of	ADP
bjmsr-315	306	21	multidisciplinary	multidisciplinary	ADJ
bjmsr-315	306	22	scientific	scientific	ADJ
bjmsr-315	306	23	research	research	NOUN
bjmsr-315	306	24	vol	vol	NOUN
bjmsr-315	306	25	.	.	PROPN
bjmsr-315	307	1	1	1	NUM
bjmsr-315	307	2	,	,	PUNCT
bjmsr-315	307	3	no	no	INTJ
bjmsr-315	307	4	.	.	NOUN
bjmsr-315	307	5	1	1	NUM
bjmsr-315	307	6	;	;	PUNCT
bjmsr-315	307	7	2019	2019	NUM
bjmsr-315	307	8	58	58	NUM
bjmsr-315	307	9	special	special	ADJ
bjmsr-315	307	10	case	case	NOUN
bjmsr-315	307	11	of	of	ADP
bjmsr-315	307	12	descent	descent	NOUN
bjmsr-315	307	13	method	method	NOUN
bjmsr-315	307	14	,	,	PUNCT
bjmsr-315	307	15	which	which	PRON
bjmsr-315	307	16	possesses	possess	VERB
bjmsr-315	307	17	two	two	NUM
bjmsr-315	307	18	important	important	ADJ
bjmsr-315	307	19	characteristics	characteristic	NOUN
bjmsr-315	307	20	.	.	PUNCT
bjmsr-315	308	1	identify	identify	VERB
bjmsr-315	308	2	direction	direction	NOUN
bjmsr-315	308	3	and	and	CCONJ
bjmsr-315	308	4	taking	take	VERB
bjmsr-315	308	5	a	a	DET
bjmsr-315	308	6	step	step	NOUN
bjmsr-315	308	7	in	in	ADP
bjmsr-315	308	8	this	this	DET
bjmsr-315	308	9	direction	direction	NOUN
bjmsr-315	308	10	to	to	PART
bjmsr-315	308	11	reduce	reduce	VERB
bjmsr-315	308	12	the	the	DET
bjmsr-315	308	13	function	function	NOUN
bjmsr-315	308	14	value	value	NOUN
bjmsr-315	308	15	(	(	PUNCT
bjmsr-315	308	16	see	see	VERB
bjmsr-315	308	17	also	also	ADV
bjmsr-315	308	18	,	,	PUNCT
bjmsr-315	308	19	anderson	anderson	PROPN
bjmsr-315	308	20	and	and	CCONJ
bjmsr-315	308	21	osborne	osborne	PROPN
bjmsr-315	308	22	(	(	PUNCT
bjmsr-315	308	23	1976	1976	NUM
bjmsr-315	308	24	)	)	PUNCT
bjmsr-315	308	25	,	,	PUNCT
bjmsr-315	308	26	osborne	osborne	NOUN
bjmsr-315	308	27	and	and	CCONJ
bjmsr-315	308	28	watson	watson	PROPN
bjmsr-315	308	29	(	(	PUNCT
bjmsr-315	308	30	1985	1985	NUM
bjmsr-315	308	31	)	)	PUNCT
bjmsr-315	308	32	osborne	osborne	NOUN
bjmsr-315	308	33	(	(	PUNCT
bjmsr-315	308	34	1985,87	1985,87	NUM
bjmsr-315	308	35	)	)	PUNCT
bjmsr-315	308	36	)	)	PUNCT
bjmsr-315	308	37	.	.	PUNCT
bjmsr-315	309	1	the	the	DET
bjmsr-315	309	2	algorithms	algorithm	NOUN
bjmsr-315	309	3	of	of	ADP
bjmsr-315	309	4	bartels	bartel	NOUN
bjmsr-315	309	5	and	and	CCONJ
bjmsr-315	309	6	conn	conn	PROPN
bjmsr-315	309	7	and	and	CCONJ
bjmsr-315	309	8	sinclair	sinclair	PROPN
bjmsr-315	309	9	(	(	PUNCT
bjmsr-315	309	10	1978	1978	NUM
bjmsr-315	309	11	)	)	PUNCT
bjmsr-315	309	12	,	,	PUNCT
bjmsr-315	309	13	armstrong	armstrong	PROPN
bjmsr-315	309	14	and	and	CCONJ
bjmsr-315	309	15	frome	frome	PROPN
bjmsr-315	309	16	and	and	CCONJ
bjmsr-315	309	17	kung	kung	PROPN
bjmsr-315	309	18	(	(	PUNCT
bjmsr-315	309	19	1979	1979	NUM
bjmsr-315	309	20	)	)	PUNCT
bjmsr-315	309	21	,	,	PUNCT
bjmsr-315	309	22	bloomfield	bloomfield	PROPN
bjmsr-315	309	23	and	and	CCONJ
bjmsr-315	309	24	steiger	steiger	PROPN
bjmsr-315	309	25	(	(	PUNCT
bjmsr-315	309	26	1980	1980	NUM
bjmsr-315	309	27	)	)	PUNCT
bjmsr-315	309	28	are	be	AUX
bjmsr-315	309	29	all	all	PRON
bjmsr-315	309	30	special	special	ADJ
bjmsr-315	309	31	cases	case	NOUN
bjmsr-315	309	32	of	of	ADP
bjmsr-315	309	33	reduced	reduce	VERB
bjmsr-315	309	34	gradient	gradient	ADJ
bjmsr-315	309	35	method	method	NOUN
bjmsr-315	309	36	.	.	PUNCT
bjmsr-315	310	1	imai	imai	PROPN
bjmsr-315	310	2	and	and	CCONJ
bjmsr-315	310	3	kato	kato	PROPN
bjmsr-315	310	4	and	and	CCONJ
bjmsr-315	310	5	yamamoto	yamamoto	PROPN
bjmsr-315	310	6	(	(	PUNCT
bjmsr-315	310	7	1987	1987	NUM
bjmsr-315	310	8	)	)	PUNCT
bjmsr-315	310	9	present	present	VERB
bjmsr-315	310	10	a	a	DET
bjmsr-315	310	11	linear	linear	ADJ
bjmsr-315	310	12	time	time	NOUN
bjmsr-315	310	13	algorithm	algorithm	NOUN
bjmsr-315	310	14	for	for	ADP
bjmsr-315	310	15	computing	compute	VERB
bjmsr-315	310	16	the	the	DET
bjmsr-315	310	17	two	two	NUM
bjmsr-315	310	18	-	-	PUNCT
bjmsr-315	310	19	parameter	parameter	NOUN
bjmsr-315	310	20	l1	l1	PROPN
bjmsr-315	310	21	norm	norm	PROPN
bjmsr-315	310	22	linear	linear	PROPN
bjmsr-315	310	23	regression	regression	NOUN
bjmsr-315	310	24	by	by	ADP
bjmsr-315	310	25	applying	apply	VERB
bjmsr-315	310	26	the	the	DET
bjmsr-315	310	27	pruning	prune	VERB
bjmsr-315	310	28	technique	technique	NOUN
bjmsr-315	310	29	.	.	PUNCT
bjmsr-315	311	1	since	since	SCONJ
bjmsr-315	311	2	the	the	DET
bjmsr-315	311	3	optimal	optimal	ADJ
bjmsr-315	311	4	solution	solution	NOUN
bjmsr-315	311	5	in	in	ADP
bjmsr-315	311	6	the	the	DET
bjmsr-315	311	7	ß0xß1	ß0xß1	NOUN
bjmsr-315	311	8	plane	plane	NOUN
bjmsr-315	311	9	lies	lie	VERB
bjmsr-315	311	10	at	at	ADP
bjmsr-315	311	11	the	the	DET
bjmsr-315	311	12	intersection	intersection	NOUN
bjmsr-315	311	13	of	of	ADP
bjmsr-315	311	14	data	datum	NOUN
bjmsr-315	311	15	lines	line	NOUN
bjmsr-315	311	16	,	,	PUNCT
bjmsr-315	311	17	so	so	ADV
bjmsr-315	311	18	,	,	PUNCT
bjmsr-315	311	19	at	at	ADP
bjmsr-315	311	20	each	each	DET
bjmsr-315	311	21	step	step	NOUN
bjmsr-315	311	22	,	,	PUNCT
bjmsr-315	311	23	a	a	DET
bjmsr-315	311	24	set	set	NOUN
bjmsr-315	311	25	of	of	ADP
bjmsr-315	311	26	data	datum	NOUN
bjmsr-315	311	27	lines	line	NOUN
bjmsr-315	311	28	which	which	PRON
bjmsr-315	311	29	does	do	AUX
bjmsr-315	311	30	not	not	PART
bjmsr-315	311	31	determine	determine	VERB
bjmsr-315	311	32	the	the	DET
bjmsr-315	311	33	optimum	optimum	ADJ
bjmsr-315	311	34	solution	solution	NOUN
bjmsr-315	311	35	are	be	AUX
bjmsr-315	311	36	discarded	discard	VERB
bjmsr-315	311	37	.	.	PUNCT
bjmsr-315	312	1	in	in	ADP
bjmsr-315	312	2	this	this	DET
bjmsr-315	312	3	paper	paper	NOUN
bjmsr-315	312	4	algebraic	algebraic	ADJ
bjmsr-315	312	5	explanation	explanation	NOUN
bjmsr-315	312	6	of	of	ADP
bjmsr-315	312	7	the	the	DET
bjmsr-315	312	8	problem	problem	NOUN
bjmsr-315	312	9	is	be	AUX
bjmsr-315	312	10	also	also	ADV
bjmsr-315	312	11	offered	offer	VERB
bjmsr-315	312	12	.	.	PUNCT
bjmsr-315	313	1	pilibossian	pilibossian	NOUN
bjmsr-315	313	2	(	(	PUNCT
bjmsr-315	313	3	1987	1987	NUM
bjmsr-315	313	4	)	)	PUNCT
bjmsr-315	313	5	also	also	ADV
bjmsr-315	313	6	gives	give	VERB
bjmsr-315	313	7	an	an	DET
bjmsr-315	313	8	algorithm	algorithm	NOUN
bjmsr-315	313	9	similar	similar	ADJ
bjmsr-315	313	10	to	to	ADP
bjmsr-315	313	11	karst	karst	PROPN
bjmsr-315	313	12	(	(	PUNCT
bjmsr-315	313	13	1958	1958	NUM
bjmsr-315	313	14	)	)	PUNCT
bjmsr-315	313	15	for	for	ADP
bjmsr-315	313	16	the	the	DET
bjmsr-315	313	17	simple	simple	ADJ
bjmsr-315	313	18	two	two	NUM
bjmsr-315	313	19	-	-	PUNCT
bjmsr-315	313	20	parameter	parameter	NOUN
bjmsr-315	313	21	linear	linear	PROPN
bjmsr-315	313	22	l1	l1	PROPN
bjmsr-315	313	23	norm	norm	PROPN
bjmsr-315	313	24	regression	regression	PROPN
bjmsr-315	313	25	.	.	PUNCT
bjmsr-315	314	1	bidabad	bidabad	PROPN
bjmsr-315	314	2	(	(	PUNCT
bjmsr-315	314	3	1987a	1987a	NUM
bjmsr-315	314	4	,	,	PUNCT
bjmsr-315	314	5	b,88a	b,88a	ADJ
bjmsr-315	314	6	,	,	PUNCT
bjmsr-315	314	7	b	b	NOUN
bjmsr-315	314	8	)	)	PUNCT
bjmsr-315	314	9	proposed	propose	VERB
bjmsr-315	314	10	descent	descent	NOUN
bjmsr-315	314	11	methods	method	NOUN
bjmsr-315	314	12	for	for	ADP
bjmsr-315	314	13	the	the	DET
bjmsr-315	314	14	simple	simple	ADJ
bjmsr-315	314	15	and	and	CCONJ
bjmsr-315	314	16	multiple	multiple	ADJ
bjmsr-315	314	17	l1	l1	PROPN
bjmsr-315	314	18	norm	norm	NOUN
bjmsr-315	314	19	regressions	regression	NOUN
bjmsr-315	314	20	.	.	PUNCT
bjmsr-315	315	1	these	these	DET
bjmsr-315	315	2	algorithms	algorithm	NOUN
bjmsr-315	315	3	,	,	PUNCT
bjmsr-315	315	4	with	with	ADP
bjmsr-315	315	5	many	many	ADJ
bjmsr-315	315	6	improvements	improvement	NOUN
bjmsr-315	315	7	,	,	PUNCT
bjmsr-315	315	8	will	will	AUX
bjmsr-315	315	9	be	be	AUX
bjmsr-315	315	10	discussed	discuss	VERB
bjmsr-315	315	11	by	by	ADP
bjmsr-315	315	12	bidabad	bidabad	NOUN
bjmsr-315	315	13	(	(	PUNCT
bjmsr-315	315	14	1989a	1989a	NUM
bjmsr-315	315	15	,	,	PUNCT
bjmsr-315	315	16	b	b	NOUN
bjmsr-315	315	17	)	)	PUNCT
bjmsr-315	315	18	.	.	PUNCT
bjmsr-315	316	1	bidabad	bidabad	PROPN
bjmsr-315	316	2	(	(	PUNCT
bjmsr-315	316	3	1989a	1989a	NUM
bjmsr-315	316	4	,	,	PUNCT
bjmsr-315	316	5	b	b	X
bjmsr-315	316	6	)	)	PUNCT
bjmsr-315	316	7	introduces	introduce	VERB
bjmsr-315	316	8	four	four	NUM
bjmsr-315	316	9	descent	descent	NOUN
bjmsr-315	316	10	algorithms	algorithm	NOUN
bjmsr-315	316	11	which	which	PRON
bjmsr-315	316	12	two	two	NUM
bjmsr-315	316	13	of	of	ADP
bjmsr-315	316	14	them	they	PRON
bjmsr-315	316	15	are	be	AUX
bjmsr-315	316	16	for	for	ADP
bjmsr-315	316	17	simple	simple	ADJ
bjmsr-315	316	18	,	,	PUNCT
bjmsr-315	316	19	and	and	CCONJ
bjmsr-315	316	20	two	two	NUM
bjmsr-315	316	21	others	other	NOUN
bjmsr-315	316	22	are	be	AUX
bjmsr-315	316	23	for	for	ADP
bjmsr-315	316	24	multiple	multiple	ADJ
bjmsr-315	316	25	regression	regression	NOUN
bjmsr-315	316	26	models.his	models.his	PRON
bjmsr-315	316	27	first	first	ADJ
bjmsr-315	316	28	algorithm	algorithm	NOUN
bjmsr-315	316	29	is	be	AUX
bjmsr-315	316	30	crude	crude	ADJ
bjmsr-315	316	31	and	and	CCONJ
bjmsr-315	316	32	tries	try	VERB
bjmsr-315	316	33	to	to	PART
bjmsr-315	316	34	check	check	VERB
bjmsr-315	316	35	many	many	ADJ
bjmsr-315	316	36	points	point	NOUN
bjmsr-315	316	37	to	to	PART
bjmsr-315	316	38	find	find	VERB
bjmsr-315	316	39	the	the	DET
bjmsr-315	316	40	optimal	optimal	ADJ
bjmsr-315	316	41	solution	solution	NOUN
bjmsr-315	316	42	.	.	PUNCT
bjmsr-315	317	1	the	the	DET
bjmsr-315	317	2	second	second	ADJ
bjmsr-315	317	3	algorithm	algorithm	NOUN
bjmsr-315	317	4	is	be	AUX
bjmsr-315	317	5	more	more	ADV
bjmsr-315	317	6	efficient	efficient	ADJ
bjmsr-315	317	7	.	.	PUNCT
bjmsr-315	318	1	algorithm	algorithm	NOUN
bjmsr-315	318	2	three	three	NUM
bjmsr-315	318	3	is	be	AUX
bjmsr-315	318	4	a	a	DET
bjmsr-315	318	5	partial	partial	ADJ
bjmsr-315	318	6	descent	descent	NOUN
bjmsr-315	318	7	procedure	procedure	NOUN
bjmsr-315	318	8	for	for	ADP
bjmsr-315	318	9	the	the	DET
bjmsr-315	318	10	general	general	ADJ
bjmsr-315	318	11	linear	linear	PROPN
bjmsr-315	318	12	model	model	PROPN
bjmsr-315	318	13	.	.	PUNCT
bjmsr-315	319	1	algorithm	algorithm	PROPN
bjmsr-315	319	2	four	four	NUM
bjmsr-315	319	3	is	be	AUX
bjmsr-315	319	4	a	a	DET
bjmsr-315	319	5	full	full	ADJ
bjmsr-315	319	6	descent	descent	NOUN
bjmsr-315	319	7	method	method	NOUN
bjmsr-315	319	8	which	which	PRON
bjmsr-315	319	9	has	have	VERB
bjmsr-315	319	10	many	many	ADJ
bjmsr-315	319	11	proved	prove	VERB
bjmsr-315	319	12	properties	property	NOUN
bjmsr-315	319	13	.	.	PUNCT
bjmsr-315	320	1	he	he	PRON
bjmsr-315	320	2	also	also	ADV
bjmsr-315	320	3	proves	prove	VERB
bjmsr-315	320	4	the	the	DET
bjmsr-315	320	5	convergence	convergence	NOUN
bjmsr-315	320	6	of	of	ADP
bjmsr-315	320	7	all	all	DET
bjmsr-315	320	8	the	the	DET
bjmsr-315	320	9	above	above	ADJ
bjmsr-315	320	10	four	four	NUM
bjmsr-315	320	11	algorithms	algorithm	NOUN
bjmsr-315	320	12	.	.	PUNCT
bjmsr-315	321	1	simplex	simplex	NOUN
bjmsr-315	321	2	type	type	NOUN
bjmsr-315	321	3	algorithms	algorithm	VERB
bjmsr-315	321	4	the	the	DET
bjmsr-315	321	5	essence	essence	NOUN
bjmsr-315	321	6	of	of	ADP
bjmsr-315	321	7	linear	linear	PROPN
bjmsr-315	321	8	programming	programming	NOUN
bjmsr-315	321	9	in	in	ADP
bjmsr-315	321	10	solving	solve	VERB
bjmsr-315	321	11	l1	l1	PROPN
bjmsr-315	321	12	norm	norm	NOUN
bjmsr-315	321	13	problem	problem	NOUN
bjmsr-315	321	14	may	may	AUX
bjmsr-315	321	15	be	be	AUX
bjmsr-315	321	16	found	find	VERB
bjmsr-315	321	17	in	in	ADP
bjmsr-315	321	18	the	the	DET
bjmsr-315	321	19	work	work	NOUN
bjmsr-315	321	20	of	of	ADP
bjmsr-315	321	21	edgeworth	edgeworth	PROPN
bjmsr-315	321	22	(	(	PUNCT
bjmsr-315	321	23	1888	1888	NUM
bjmsr-315	321	24	)	)	PUNCT
bjmsr-315	321	25	.	.	PUNCT
bjmsr-315	322	1	harris	harris	PROPN
bjmsr-315	322	2	(	(	PUNCT
bjmsr-315	322	3	1950	1950	NUM
bjmsr-315	322	4	)	)	PUNCT
bjmsr-315	322	5	suggested	suggest	VERB
bjmsr-315	322	6	that	that	SCONJ
bjmsr-315	322	7	the	the	DET
bjmsr-315	322	8	l1	l1	PROPN
bjmsr-315	322	9	norm	norm	PROPN
bjmsr-315	322	10	estimation	estimation	NOUN
bjmsr-315	322	11	problem	problem	NOUN
bjmsr-315	322	12	is	be	AUX
bjmsr-315	322	13	connected	connect	VERB
bjmsr-315	322	14	with	with	ADP
bjmsr-315	322	15	linear	linear	ADJ
bjmsr-315	322	16	programming	programming	NOUN
bjmsr-315	322	17	.	.	PUNCT
bjmsr-315	323	1	charnes	charne	NOUN
bjmsr-315	323	2	and	and	CCONJ
bjmsr-315	323	3	cooper	cooper	PROPN
bjmsr-315	323	4	and	and	CCONJ
bjmsr-315	323	5	ferguson	ferguson	PROPN
bjmsr-315	323	6	(	(	PUNCT
bjmsr-315	323	7	1955	1955	NUM
bjmsr-315	323	8	)	)	PUNCT
bjmsr-315	323	9	formulated	formulate	VERB
bjmsr-315	323	10	the	the	DET
bjmsr-315	323	11	problem	problem	NOUN
bjmsr-315	323	12	as	as	ADP
bjmsr-315	323	13	a	a	DET
bjmsr-315	323	14	linear	linear	ADJ
bjmsr-315	323	15	programming	programming	NOUN
bjmsr-315	323	16	model	model	NOUN
bjmsr-315	323	17	.	.	PUNCT
bjmsr-315	324	1	this	this	DET
bjmsr-315	324	2	article	article	NOUN
bjmsr-315	324	3	is	be	AUX
bjmsr-315	324	4	the	the	DET
bjmsr-315	324	5	first	first	ADV
bjmsr-315	324	6	known	know	VERB
bjmsr-315	324	7	to	to	PART
bjmsr-315	324	8	use	use	VERB
bjmsr-315	324	9	linear	linear	ADJ
bjmsr-315	324	10	programming	programming	NOUN
bjmsr-315	324	11	for	for	ADP
bjmsr-315	324	12	this	this	DET
bjmsr-315	324	13	case	case	NOUN
bjmsr-315	324	14	.	.	PUNCT
bjmsr-315	325	1	adaptation	adaptation	NOUN
bjmsr-315	325	2	of	of	ADP
bjmsr-315	325	3	linear	linear	PROPN
bjmsr-315	325	4	programming	programming	NOUN
bjmsr-315	325	5	to	to	ADP
bjmsr-315	325	6	l1	l1	PROPN
bjmsr-315	325	7	norm	norm	PROPN
bjmsr-315	325	8	estimation	estimation	NOUN
bjmsr-315	325	9	problem	problem	NOUN
bjmsr-315	325	10	is	be	AUX
bjmsr-315	325	11	shown	show	VERB
bjmsr-315	325	12	below	below	ADP
bjmsr-315	325	13	:	:	PUNCT
bjmsr-315	325	14	min	min	NOUN
bjmsr-315	325	15	:	:	PUNCT
bjmsr-315	325	16	1n	1n	NUM
bjmsr-315	325	17	t(w+v	t(w+v	NOUN
bjmsr-315	325	18	)	)	PUNCT
bjmsr-315	325	19	ß	ß	PRON
bjmsr-315	325	20	s.to	s.to	PROPN
bjmsr-315	325	21	:	:	PUNCT
bjmsr-315	325	22	xß+in(w	xß+in(w	NOUN
bjmsr-315	325	23	-	-	PUNCT
bjmsr-315	325	24	v)=y	v)=y	NOUN
bjmsr-315	325	25	(	(	PUNCT
bjmsr-315	325	26	31	31	NUM
bjmsr-315	325	27	)	)	PUNCT
bjmsr-315	325	28	w	w	NOUN
bjmsr-315	325	29	,	,	PUNCT
bjmsr-315	325	30	v≥0	v≥0	PROPN
bjmsr-315	325	31	ß	ß	NOUN
bjmsr-315	325	32	unrestricted	unrestricted	ADJ
bjmsr-315	325	33	in	in	ADP
bjmsr-315	325	34	sign	sign	NOUN
bjmsr-315	325	35	where	where	SCONJ
bjmsr-315	325	36	1n	1n	PROPN
bjmsr-315	325	37	is	be	AUX
bjmsr-315	325	38	a	a	DET
bjmsr-315	325	39	vector	vector	NOUN
bjmsr-315	325	40	of	of	ADP
bjmsr-315	325	41	size	size	NOUN
bjmsr-315	325	42	nx1	nx1	NOUN
bjmsr-315	325	43	of	of	ADP
bjmsr-315	325	44	1	1	NUM
bjmsr-315	325	45	's	's	PART
bjmsr-315	325	46	and	and	CCONJ
bjmsr-315	325	47	in	in	ADP
bjmsr-315	325	48	is	be	AUX
bjmsr-315	325	49	a	a	DET
bjmsr-315	325	50	nth	nth	NOUN
bjmsr-315	325	51	order	order	NOUN
bjmsr-315	325	52	identity	identity	NOUN
bjmsr-315	325	53	matrix	matrix	NOUN
bjmsr-315	325	54	.	.	PUNCT
bjmsr-315	326	1	the	the	DET
bjmsr-315	326	2	vectors	vector	NOUN
bjmsr-315	326	3	v	v	ADP
bjmsr-315	326	4	and	and	CCONJ
bjmsr-315	326	5	w	w	NOUN
bjmsr-315	326	6	are	be	AUX
bjmsr-315	326	7	of	of	ADP
bjmsr-315	326	8	size	size	NOUN
bjmsr-315	326	9	nx1	nx1	NOUN
bjmsr-315	326	10	and	and	CCONJ
bjmsr-315	326	11	their	their	PRON
bjmsr-315	326	12	elements	element	NOUN
bjmsr-315	326	13	may	may	AUX
bjmsr-315	326	14	be	be	AUX
bjmsr-315	326	15	interpreted	interpret	VERB
bjmsr-315	326	16	as	as	ADP
bjmsr-315	326	17	vertical	vertical	ADJ
bjmsr-315	326	18	deviations	deviation	NOUN
bjmsr-315	326	19	above	above	ADP
bjmsr-315	326	20	and	and	CCONJ
bjmsr-315	326	21	below	below	ADP
bjmsr-315	326	22	the	the	DET
bjmsr-315	326	23	fitted	fit	VERB
bjmsr-315	326	24	regression	regression	NOUN
bjmsr-315	326	25	hyperplane	hyperplane	NOUN
bjmsr-315	326	26	respectively	respectively	ADV
bjmsr-315	326	27	.	.	PUNCT
bjmsr-315	327	1	this	this	DET
bjmsr-315	327	2	problem	problem	NOUN
bjmsr-315	327	3	has	have	VERB
bjmsr-315	327	4	n	n	NUM
bjmsr-315	327	5	equality	equality	NOUN
bjmsr-315	327	6	constraints	constraint	NOUN
bjmsr-315	327	7	in	in	ADP
bjmsr-315	327	8	m+2n	m+2n	PROPN
bjmsr-315	327	9	variables	variable	NOUN
bjmsr-315	327	10	.	.	PUNCT
bjmsr-315	328	1	when	when	SCONJ
bjmsr-315	328	2	n	n	PRON
bjmsr-315	328	3	is	be	AUX
bjmsr-315	328	4	large	large	ADJ
bjmsr-315	328	5	,	,	PUNCT
bjmsr-315	328	6	this	this	DET
bjmsr-315	328	7	formulation	formulation	NOUN
bjmsr-315	328	8	generally	generally	ADV
bjmsr-315	328	9	requires	require	VERB
bjmsr-315	328	10	a	a	DET
bjmsr-315	328	11	large	large	ADJ
bjmsr-315	328	12	amount	amount	NOUN
bjmsr-315	328	13	of	of	ADP
bjmsr-315	328	14	storage	storage	NOUN
bjmsr-315	328	15	and	and	CCONJ
bjmsr-315	328	16	computation	computation	NOUN
bjmsr-315	328	17	time	time	NOUN
bjmsr-315	328	18	.	.	PUNCT
bjmsr-315	329	1	wagner	wagner	PROPN
bjmsr-315	329	2	(	(	PUNCT
bjmsr-315	329	3	1959	1959	NUM
bjmsr-315	329	4	)	)	PUNCT
bjmsr-315	329	5	shows	show	VERB
bjmsr-315	329	6	that	that	SCONJ
bjmsr-315	329	7	the	the	DET
bjmsr-315	329	8	formulation	formulation	NOUN
bjmsr-315	329	9	of	of	ADP
bjmsr-315	329	10	the	the	DET
bjmsr-315	329	11	l1	l1	PROPN
bjmsr-315	329	12	norm	norm	NOUN
bjmsr-315	329	13	regression	regression	NOUN
bjmsr-315	329	14	may	may	AUX
bjmsr-315	329	15	be	be	AUX
bjmsr-315	329	16	reduced	reduce	VERB
bjmsr-315	329	17	to	to	ADP
bjmsr-315	329	18	m	m	PROPN
bjmsr-315	329	19	equality	equality	NOUN
bjmsr-315	329	20	constraints	constraint	NOUN
bjmsr-315	329	21	linear	linear	ADJ
bjmsr-315	329	22	programming	programming	NOUN
bjmsr-315	329	23	problem	problem	NOUN
bjmsr-315	329	24	.	.	PUNCT
bjmsr-315	330	1	thus	thus	ADV
bjmsr-315	330	2	,	,	PUNCT
bjmsr-315	330	3	this	this	DET
bjmsr-315	330	4	dual	dual	ADJ
bjmsr-315	330	5	formulation	formulation	NOUN
bjmsr-315	330	6	reduces	reduce	VERB
bjmsr-315	330	7	n	n	DET
bjmsr-315	330	8	equations	equation	NOUN
bjmsr-315	330	9	of	of	ADP
bjmsr-315	330	10	primal	primal	ADJ
bjmsr-315	330	11	form	form	NOUN
bjmsr-315	330	12	to	to	ADP
bjmsr-315	330	13	m	m	PROPN
bjmsr-315	330	14	equations	equation	NOUN
bjmsr-315	330	15	of	of	ADP
bjmsr-315	330	16	dual	dual	ADJ
bjmsr-315	330	17	form	form	NOUN
bjmsr-315	330	18	and	and	CCONJ
bjmsr-315	330	19	considerably	considerably	ADV
bjmsr-315	330	20	reduces	reduce	VERB
bjmsr-315	330	21	the	the	DET
bjmsr-315	330	22	storage	storage	NOUN
bjmsr-315	330	23	and	and	CCONJ
bjmsr-315	330	24	computation	computation	NOUN
bjmsr-315	330	25	time	time	NOUN
bjmsr-315	330	26	.	.	PUNCT
bjmsr-315	331	1	fisher	fisher	PROPN
bjmsr-315	331	2	(	(	PUNCT
bjmsr-315	331	3	1961	1961	NUM
bjmsr-315	331	4	)	)	PUNCT
bjmsr-315	331	5	reviews	review	VERB
bjmsr-315	331	6	the	the	DET
bjmsr-315	331	7	formulation	formulation	NOUN
bjmsr-315	331	8	of	of	ADP
bjmsr-315	331	9	the	the	DET
bjmsr-315	331	10	l1	l1	PROPN
bjmsr-315	331	11	norm	norm	NOUN
bjmsr-315	331	12	estimation	estimation	NOUN
bjmsr-315	331	13	in	in	ADP
bjmsr-315	331	14	relation	relation	NOUN
bjmsr-315	331	15	to	to	ADP
bjmsr-315	331	16	the	the	DET
bjmsr-315	331	17	primal	primal	ADJ
bjmsr-315	331	18	form	form	NOUN
bjmsr-315	331	19	of	of	ADP
bjmsr-315	331	20	linear	linear	PROPN
bjmsr-315	331	21	programming	programming	NOUN
bjmsr-315	331	22	.	.	PUNCT
bjmsr-315	332	1	barrodale	barrodale	NOUN
bjmsr-315	332	2	and	and	CCONJ
bjmsr-315	332	3	young	young	ADJ
bjmsr-315	332	4	(	(	PUNCT
bjmsr-315	332	5	1966	1966	NUM
bjmsr-315	332	6	)	)	PUNCT
bjmsr-315	332	7	developed	develop	VERB
bjmsr-315	332	8	a	a	DET
bjmsr-315	332	9	modified	modified	ADJ
bjmsr-315	332	10	simplex	simplex	NOUN
bjmsr-315	332	11	algorithm	algorithm	NOUN
bjmsr-315	332	12	for	for	ADP
bjmsr-315	332	13	determining	determine	VERB
bjmsr-315	332	14	the	the	DET
bjmsr-315	332	15	best	well	ADV
bjmsr-315	332	16	fitting	fitting	ADJ
bjmsr-315	332	17	function	function	NOUN
bjmsr-315	332	18	to	to	ADP
bjmsr-315	332	19	a	a	DET
bjmsr-315	332	20	set	set	NOUN
bjmsr-315	332	21	of	of	ADP
bjmsr-315	332	22	discrete	discrete	ADJ
bjmsr-315	332	23	data	datum	NOUN
bjmsr-315	332	24	under	under	ADP
bjmsr-315	332	25	the	the	DET
bjmsr-315	332	26	l1	l1	PROPN
bjmsr-315	332	27	norm	norm	NOUN
bjmsr-315	332	28	criterion	criterion	NOUN
bjmsr-315	332	29	.	.	PUNCT
bjmsr-315	333	1	the	the	DET
bjmsr-315	333	2	method	method	NOUN
bjmsr-315	333	3	is	be	AUX
bjmsr-315	333	4	given	give	VERB
bjmsr-315	333	5	as	as	ADP
bjmsr-315	333	6	algol	algol	NOUN
bjmsr-315	333	7	codes	code	NOUN
bjmsr-315	333	8	(	(	PUNCT
bjmsr-315	333	9	for	for	SCONJ
bjmsr-315	333	10	critics	critic	NOUN
bjmsr-315	333	11	see	see	VERB
bjmsr-315	333	12	,	,	PUNCT
bjmsr-315	333	13	mccormick	mccormick	PROPN
bjmsr-315	333	14	and	and	CCONJ
bjmsr-315	333	15	sposito	sposito	X
bjmsr-315	333	16	(	(	PUNCT
bjmsr-315	333	17	1975	1975	NUM
bjmsr-315	333	18	)	)	PUNCT
bjmsr-315	333	19	)	)	PUNCT
bjmsr-315	333	20	.	.	PUNCT
bjmsr-315	334	1	davies	davy	NOUN
bjmsr-315	334	2	(	(	PUNCT
bjmsr-315	334	3	1967	1967	NUM
bjmsr-315	334	4	)	)	PUNCT
bjmsr-315	334	5	demonstrates	demonstrate	VERB
bjmsr-315	334	6	the	the	DET
bjmsr-315	334	7	use	use	NOUN
bjmsr-315	334	8	of	of	ADP
bjmsr-315	334	9	the	the	DET
bjmsr-315	334	10	l1	l1	PROPN
bjmsr-315	334	11	norm	norm	NOUN
bjmsr-315	334	12	regression	regression	NOUN
bjmsr-315	334	13	estimates	estimate	NOUN
bjmsr-315	334	14	.	.	PUNCT
bjmsr-315	335	1	rabinowitz	rabinowitz	NOUN
bjmsr-315	335	2	(	(	PUNCT
bjmsr-315	335	3	1968	1968	NUM
bjmsr-315	335	4	)	)	PUNCT
bjmsr-315	335	5	also	also	ADV
bjmsr-315	335	6	discusses	discuss	VERB
bjmsr-315	335	7	the	the	DET
bjmsr-315	335	8	application	application	NOUN
bjmsr-315	335	9	of	of	ADP
bjmsr-315	335	10	linear	linear	PROPN
bjmsr-315	335	11	programming	programming	NOUN
bjmsr-315	335	12	in	in	ADP
bjmsr-315	335	13	this	this	DET
bjmsr-315	335	14	field	field	NOUN
bjmsr-315	335	15	.	.	PUNCT
bjmsr-315	336	1	crocker	crocker	PROPN
bjmsr-315	336	2	(	(	PUNCT
bjmsr-315	336	3	1969	1969	NUM
bjmsr-315	336	4	)	)	PUNCT
bjmsr-315	336	5	cautions	caution	VERB
bjmsr-315	336	6	against	against	ADP
bjmsr-315	336	7	using	use	VERB
bjmsr-315	336	8	the	the	DET
bjmsr-315	336	9	l1	l1	PROPN
bjmsr-315	336	10	norm	norm	NOUN
bjmsr-315	336	11	criterion	criterion	NOUN
bjmsr-315	336	12	merely	merely	ADV
bjmsr-315	336	13	to	to	PART
bjmsr-315	336	14	restrain	restrain	VERB
bjmsr-315	336	15	unwanted	unwanted	ADJ
bjmsr-315	336	16	negative	negative	ADJ
bjmsr-315	336	17	coefficient	coefficient	NOUN
bjmsr-315	336	18	estimates	estimate	NOUN
bjmsr-315	336	19	which	which	PRON
bjmsr-315	336	20	occur	occur	VERB
bjmsr-315	336	21	in	in	ADP
bjmsr-315	336	22	the	the	DET
bjmsr-315	336	23	least	least	ADJ
bjmsr-315	336	24	squares	square	NOUN
bjmsr-315	336	25	regression	regression	NOUN
bjmsr-315	336	26	.	.	PUNCT
bjmsr-315	337	1	multicolinearity	multicolinearity	NOUN
bjmsr-315	337	2	is	be	AUX
bjmsr-315	337	3	one	one	NUM
bjmsr-315	337	4	of	of	ADP
bjmsr-315	337	5	the	the	DET
bjmsr-315	337	6	cases	case	NOUN
bjmsr-315	337	7	which	which	PRON
bjmsr-315	337	8	causes	cause	VERB
bjmsr-315	337	9	this	this	DET
bjmsr-315	337	10	result	result	NOUN
bjmsr-315	337	11	.	.	PUNCT
bjmsr-315	338	1	robers	rober	NOUN
bjmsr-315	338	2	and	and	CCONJ
bjmsr-315	338	3	ben	ben	PROPN
bjmsr-315	338	4	-	-	PROPN
bjmsr-315	338	5	israel	israel	PROPN
bjmsr-315	338	6	(	(	PUNCT
bjmsr-315	338	7	1969	1969	NUM
bjmsr-315	338	8	)	)	PUNCT
bjmsr-315	338	9	by	by	ADP
bjmsr-315	338	10	using	use	VERB
bjmsr-315	338	11	interval	interval	NOUN
bjmsr-315	338	12	linear	linear	PROPN
bjmsr-315	338	13	programming	programming	NOUN
bjmsr-315	338	14	,	,	PUNCT
bjmsr-315	338	15	proposed	propose	VERB
bjmsr-315	338	16	an	an	DET
bjmsr-315	338	17	algorithm	algorithm	NOUN
bjmsr-315	338	18	to	to	PART
bjmsr-315	338	19	solve	solve	VERB
bjmsr-315	338	20	the	the	DET
bjmsr-315	338	21	l1	l1	PROPN
bjmsr-315	338	22	norm	norm	PROPN
bjmsr-315	338	23	estimation	estimation	NOUN
bjmsr-315	338	24	problem	problem	NOUN
bjmsr-315	338	25	.	.	PUNCT
bjmsr-315	339	1	rabinowitz	rabinowitz	NOUN
bjmsr-315	339	2	(	(	PUNCT
bjmsr-315	339	3	1970	1970	NUM
bjmsr-315	339	4	)	)	PUNCT
bjmsr-315	339	5	,	,	PUNCT
bjmsr-315	339	6	shanno	shanno	NOUN
bjmsr-315	339	7	and	and	CCONJ
bjmsr-315	339	8	weil	weil	PROPN
bjmsr-315	339	9	(	(	PUNCT
bjmsr-315	339	10	1970	1970	NUM
bjmsr-315	339	11	)	)	PUNCT
bjmsr-315	339	12	discuss	discuss	VERB
bjmsr-315	339	13	some	some	DET
bjmsr-315	339	14	connections	connection	NOUN
bjmsr-315	339	15	between	between	ADP
bjmsr-315	339	16	linear	linear	NOUN
bjmsr-315	339	17	programming	programming	NOUN
bjmsr-315	339	18	and	and	CCONJ
bjmsr-315	339	19	the	the	DET
bjmsr-315	339	20	approximation	approximation	NOUN
bjmsr-315	339	21	problem	problem	NOUN
bjmsr-315	339	22	.	.	PUNCT
bjmsr-315	340	1	barrodale	barrodale	NOUN
bjmsr-315	340	2	(	(	PUNCT
bjmsr-315	340	3	1970	1970	NUM
bjmsr-315	340	4	)	)	PUNCT
bjmsr-315	340	5	summarizes	summarize	VERB
bjmsr-315	340	6	the	the	DET
bjmsr-315	340	7	linear	linear	ADJ
bjmsr-315	340	8	and	and	CCONJ
bjmsr-315	340	9	nonlinear	nonlinear	PROPN
bjmsr-315	340	10	l1	l1	PROPN
bjmsr-315	340	11	norm	norm	NOUN
bjmsr-315	340	12	curve	curve	PROPN
bjmsr-315	340	13	fitting	fit	VERB
bjmsr-315	340	14	on	on	ADP
bjmsr-315	340	15	both	both	CCONJ
bjmsr-315	340	16	continuous	continuous	ADJ
bjmsr-315	340	17	copyright	copyright	NOUN
bjmsr-315	341	1	©	©	PROPN
bjmsr-315	341	2	cc	cc	NOUN
bjmsr-315	341	3	-	-	PUNCT
bjmsr-315	341	4	by	by	ADP
bjmsr-315	341	5	-	-	PUNCT
bjmsr-315	341	6	nc	nc	PROPN
bjmsr-315	341	7	2019	2019	NUM
bjmsr-315	341	8	,	,	PUNCT
bjmsr-315	341	9	bjmsr	bjmsr	PROPN
bjmsr-315	341	10	www.cribfb.com/journal/index.php/bjmsr	www.cribfb.com/journal/index.php/bjmsr	X
bjmsr-315	342	1	bangladesh	bangladesh	PROPN
bjmsr-315	342	2	journal	journal	PROPN
bjmsr-315	342	3	of	of	ADP
bjmsr-315	342	4	multidisciplinary	multidisciplinary	ADJ
bjmsr-315	342	5	scientific	scientific	ADJ
bjmsr-315	342	6	research	research	NOUN
bjmsr-315	342	7	vol	vol	NOUN
bjmsr-315	342	8	.	.	PROPN
bjmsr-315	342	9	1	1	NUM
bjmsr-315	342	10	,	,	PUNCT
bjmsr-315	342	11	no	no	INTJ
bjmsr-315	342	12	.	.	NOUN
bjmsr-315	342	13	1	1	NUM
bjmsr-315	342	14	;	;	PUNCT
bjmsr-315	342	15	2019	2019	NUM
bjmsr-315	342	16	59	59	NUM
bjmsr-315	342	17	and	and	CCONJ
bjmsr-315	342	18	discrete	discrete	ADJ
bjmsr-315	342	19	data	datum	NOUN
bjmsr-315	342	20	.	.	PUNCT
bjmsr-315	343	1	spyropoulos	spyropoulo	NOUN
bjmsr-315	343	2	and	and	CCONJ
bjmsr-315	343	3	kiountouzis	kiountouzis	NOUN
bjmsr-315	343	4	and	and	CCONJ
bjmsr-315	343	5	young	young	ADJ
bjmsr-315	343	6	(	(	PUNCT
bjmsr-315	343	7	1973	1973	NUM
bjmsr-315	343	8	)	)	PUNCT
bjmsr-315	343	9	suggest	suggest	VERB
bjmsr-315	343	10	two	two	NUM
bjmsr-315	343	11	algorithms	algorithm	NOUN
bjmsr-315	343	12	for	for	ADP
bjmsr-315	343	13	fitting	fitting	ADJ
bjmsr-315	343	14	general	general	ADJ
bjmsr-315	343	15	functions	function	NOUN
bjmsr-315	343	16	and	and	CCONJ
bjmsr-315	343	17	particularly	particularly	ADV
bjmsr-315	343	18	fast	fast	ADJ
bjmsr-315	343	19	algorithm	algorithm	NOUN
bjmsr-315	343	20	with	with	ADP
bjmsr-315	343	21	minimum	minimum	ADJ
bjmsr-315	343	22	storage	storage	NOUN
bjmsr-315	343	23	requirements	requirement	NOUN
bjmsr-315	343	24	for	for	ADP
bjmsr-315	343	25	fitting	fitting	ADJ
bjmsr-315	343	26	polynomials	polynomial	NOUN
bjmsr-315	343	27	based	base	VERB
bjmsr-315	343	28	on	on	ADP
bjmsr-315	343	29	the	the	DET
bjmsr-315	343	30	algebraic	algebraic	ADJ
bjmsr-315	343	31	properties	property	NOUN
bjmsr-315	343	32	of	of	ADP
bjmsr-315	343	33	linear	linear	ADJ
bjmsr-315	343	34	programming	programming	NOUN
bjmsr-315	343	35	formulation	formulation	NOUN
bjmsr-315	343	36	.	.	PUNCT
bjmsr-315	344	1	robers	rober	NOUN
bjmsr-315	344	2	and	and	CCONJ
bjmsr-315	344	3	robers	rober	NOUN
bjmsr-315	344	4	(	(	PUNCT
bjmsr-315	344	5	1973	1973	NUM
bjmsr-315	344	6	)	)	PUNCT
bjmsr-315	344	7	have	have	AUX
bjmsr-315	344	8	supplied	supply	VERB
bjmsr-315	344	9	a	a	DET
bjmsr-315	344	10	special	special	ADJ
bjmsr-315	344	11	version	version	NOUN
bjmsr-315	344	12	of	of	ADP
bjmsr-315	344	13	the	the	DET
bjmsr-315	344	14	general	general	ADJ
bjmsr-315	344	15	method	method	NOUN
bjmsr-315	344	16	of	of	ADP
bjmsr-315	344	17	robers	rober	NOUN
bjmsr-315	344	18	and	and	CCONJ
bjmsr-315	344	19	benisrael	benisrael	PROPN
bjmsr-315	344	20	(	(	PUNCT
bjmsr-315	344	21	1969	1969	NUM
bjmsr-315	344	22	)	)	PUNCT
bjmsr-315	344	23	,	,	PUNCT
bjmsr-315	344	24	which	which	PRON
bjmsr-315	344	25	is	be	AUX
bjmsr-315	344	26	designed	design	VERB
bjmsr-315	344	27	specifically	specifically	ADV
bjmsr-315	344	28	for	for	ADP
bjmsr-315	344	29	the	the	DET
bjmsr-315	344	30	l1	l1	PROPN
bjmsr-315	344	31	norm	norm	PROPN
bjmsr-315	344	32	problem	problem	NOUN
bjmsr-315	344	33	.	.	PUNCT
bjmsr-315	345	1	a	a	DET
bjmsr-315	345	2	related	relate	VERB
bjmsr-315	345	3	fortran	fortran	NOUN
bjmsr-315	345	4	code	code	NOUN
bjmsr-315	345	5	is	be	AUX
bjmsr-315	345	6	also	also	ADV
bjmsr-315	345	7	provided	provide	VERB
bjmsr-315	345	8	.	.	PUNCT
bjmsr-315	346	1	barrodale	barrodale	NOUN
bjmsr-315	346	2	and	and	CCONJ
bjmsr-315	346	3	roberts	roberts	PROPN
bjmsr-315	346	4	(	(	PUNCT
bjmsr-315	346	5	1973	1973	NUM
bjmsr-315	346	6	)	)	PUNCT
bjmsr-315	346	7	present	present	VERB
bjmsr-315	346	8	a	a	DET
bjmsr-315	346	9	modification	modification	NOUN
bjmsr-315	346	10	of	of	ADP
bjmsr-315	346	11	the	the	DET
bjmsr-315	346	12	simplex	simplex	NOUN
bjmsr-315	346	13	method	method	NOUN
bjmsr-315	346	14	,	,	PUNCT
bjmsr-315	346	15	which	which	PRON
bjmsr-315	346	16	needs	need	VERB
bjmsr-315	346	17	a	a	DET
bjmsr-315	346	18	smaller	small	ADJ
bjmsr-315	346	19	amount	amount	NOUN
bjmsr-315	346	20	of	of	ADP
bjmsr-315	346	21	storage	storage	NOUN
bjmsr-315	346	22	,	,	PUNCT
bjmsr-315	346	23	and	and	CCONJ
bjmsr-315	346	24	by	by	ADP
bjmsr-315	346	25	skipping	skip	VERB
bjmsr-315	346	26	over	over	ADP
bjmsr-315	346	27	simplex	simplex	NOUN
bjmsr-315	346	28	vertices	vertex	NOUN
bjmsr-315	346	29	is	be	AUX
bjmsr-315	346	30	more	more	ADV
bjmsr-315	346	31	efficient	efficient	ADJ
bjmsr-315	346	32	than	than	ADP
bjmsr-315	346	33	the	the	DET
bjmsr-315	346	34	usual	usual	ADJ
bjmsr-315	346	35	simplex	simplex	NOUN
bjmsr-315	346	36	procedure	procedure	NOUN
bjmsr-315	346	37	.	.	PUNCT
bjmsr-315	347	1	define	define	VERB
bjmsr-315	347	2	the	the	DET
bjmsr-315	347	3	vector	vector	NOUN
bjmsr-315	347	4	ß	ß	NOUN
bjmsr-315	347	5	as	as	ADP
bjmsr-315	347	6	a	a	DET
bjmsr-315	347	7	difference	difference	NOUN
bjmsr-315	347	8	of	of	ADP
bjmsr-315	347	9	two	two	NUM
bjmsr-315	347	10	nonnegative	nonnegative	ADJ
bjmsr-315	347	11	vectors	vector	NOUN
bjmsr-315	347	12	c	c	NOUN
bjmsr-315	347	13	and	and	CCONJ
bjmsr-315	347	14	d	d	NOUN
bjmsr-315	347	15	;	;	PUNCT
bjmsr-315	347	16	their	their	PRON
bjmsr-315	347	17	formulation	formulation	NOUN
bjmsr-315	347	18	can	can	AUX
bjmsr-315	347	19	be	be	AUX
bjmsr-315	347	20	stated	state	VERB
bjmsr-315	347	21	as	as	ADP
bjmsr-315	347	22	follows	follow	NOUN
bjmsr-315	347	23	,	,	PUNCT
bjmsr-315	347	24	min	min	NOUN
bjmsr-315	347	25	:	:	PUNCT
bjmsr-315	347	26	1n	1n	NUM
bjmsr-315	347	27	t(w+v	t(w+v	NOUN
bjmsr-315	347	28	)	)	PUNCT
bjmsr-315	348	1	c	c	NOUN
bjmsr-315	348	2	,	,	PUNCT
bjmsr-315	348	3	d	d	PROPN
bjmsr-315	348	4	s.to	s.to	NOUN
bjmsr-315	348	5	:	:	PUNCT
bjmsr-315	348	6	x(c	x(c	NOUN
bjmsr-315	348	7	-	-	PUNCT
bjmsr-315	348	8	d)+in(w	d)+in(w	NOUN
bjmsr-315	348	9	-	-	PUNCT
bjmsr-315	348	10	v)=y	v)=y	NOUN
bjmsr-315	348	11	(	(	PUNCT
bjmsr-315	348	12	32	32	NUM
bjmsr-315	348	13	)	)	PUNCT
bjmsr-315	348	14	w	w	PROPN
bjmsr-315	348	15	,	,	PUNCT
bjmsr-315	348	16	v	v	NOUN
bjmsr-315	348	17	,	,	PUNCT
bjmsr-315	348	18	c	c	NOUN
bjmsr-315	348	19	,	,	PUNCT
bjmsr-315	348	20	d≥0	d≥0	NOUN
bjmsr-315	348	21	because	because	SCONJ
bjmsr-315	348	22	of	of	ADP
bjmsr-315	348	23	the	the	DET
bjmsr-315	348	24	relationships	relationship	NOUN
bjmsr-315	348	25	among	among	ADP
bjmsr-315	348	26	variables	variable	NOUN
bjmsr-315	348	27	,	,	PUNCT
bjmsr-315	348	28	the	the	DET
bjmsr-315	348	29	computation	computation	NOUN
bjmsr-315	348	30	can	can	AUX
bjmsr-315	348	31	be	be	AUX
bjmsr-315	348	32	performed	perform	VERB
bjmsr-315	348	33	by	by	ADP
bjmsr-315	348	34	using	use	VERB
bjmsr-315	348	35	only	only	ADV
bjmsr-315	348	36	(	(	PUNCT
bjmsr-315	348	37	n+2)x(m+2	n+2)x(m+2	NOUN
bjmsr-315	348	38	)	)	PUNCT
bjmsr-315	348	39	amount	amount	NOUN
bjmsr-315	348	40	of	of	ADP
bjmsr-315	348	41	array	array	NOUN
bjmsr-315	348	42	storage	storage	NOUN
bjmsr-315	348	43	,	,	PUNCT
bjmsr-315	348	44	including	include	VERB
bjmsr-315	348	45	labels	label	NOUN
bjmsr-315	348	46	for	for	ADP
bjmsr-315	348	47	the	the	DET
bjmsr-315	348	48	basic	basic	ADJ
bjmsr-315	348	49	and	and	CCONJ
bjmsr-315	348	50	non	non	ADJ
bjmsr-315	348	51	-	-	ADJ
bjmsr-315	348	52	basic	basic	ADJ
bjmsr-315	348	53	vectors	vector	NOUN
bjmsr-315	348	54	.	.	PUNCT
bjmsr-315	349	1	an	an	DET
bjmsr-315	349	2	initial	initial	ADJ
bjmsr-315	349	3	basis	basis	NOUN
bjmsr-315	349	4	is	be	AUX
bjmsr-315	349	5	given	give	VERB
bjmsr-315	349	6	by	by	ADP
bjmsr-315	349	7	w	w	ADP
bjmsr-315	349	8	if	if	SCONJ
bjmsr-315	349	9	all	all	DET
bjmsr-315	349	10	yi	yi	NOUN
bjmsr-315	349	11	are	be	AUX
bjmsr-315	349	12	nonnegative	nonnegative	ADJ
bjmsr-315	349	13	.	.	PUNCT
bjmsr-315	350	1	if	if	SCONJ
bjmsr-315	350	2	a	a	DET
bjmsr-315	350	3	yi	yi	NOUN
bjmsr-315	350	4	is	be	AUX
bjmsr-315	350	5	negative	negative	ADJ
bjmsr-315	350	6	,	,	PUNCT
bjmsr-315	350	7	the	the	DET
bjmsr-315	350	8	sign	sign	NOUN
bjmsr-315	350	9	of	of	ADP
bjmsr-315	350	10	the	the	DET
bjmsr-315	350	11	corresponding	corresponding	NOUN
bjmsr-315	350	12	row	row	NOUN
bjmsr-315	350	13	is	be	AUX
bjmsr-315	350	14	changed	change	VERB
bjmsr-315	350	15	,	,	PUNCT
bjmsr-315	350	16	and	and	CCONJ
bjmsr-315	350	17	the	the	DET
bjmsr-315	350	18	unit	unit	NOUN
bjmsr-315	350	19	column	column	NOUN
bjmsr-315	350	20	from	from	ADP
bjmsr-315	350	21	the	the	DET
bjmsr-315	350	22	corresponding	corresponding	ADJ
bjmsr-315	350	23	element	element	NOUN
bjmsr-315	350	24	of	of	ADP
bjmsr-315	350	25	v	v	NOUN
bjmsr-315	350	26	is	be	AUX
bjmsr-315	350	27	taken	take	VERB
bjmsr-315	350	28	as	as	ADP
bjmsr-315	350	29	part	part	NOUN
bjmsr-315	350	30	of	of	ADP
bjmsr-315	350	31	the	the	DET
bjmsr-315	350	32	basis	basis	NOUN
bjmsr-315	350	33	.	.	PUNCT
bjmsr-315	351	1	the	the	DET
bjmsr-315	351	2	algorithm	algorithm	NOUN
bjmsr-315	351	3	is	be	AUX
bjmsr-315	351	4	implemented	implement	VERB
bjmsr-315	351	5	in	in	ADP
bjmsr-315	351	6	two	two	NUM
bjmsr-315	351	7	stages	stage	NOUN
bjmsr-315	351	8	.	.	PUNCT
bjmsr-315	352	1	the	the	DET
bjmsr-315	352	2	first	first	ADJ
bjmsr-315	352	3	stage	stage	NOUN
bjmsr-315	352	4	restricts	restrict	VERB
bjmsr-315	352	5	the	the	DET
bjmsr-315	352	6	choice	choice	NOUN
bjmsr-315	352	7	of	of	ADP
bjmsr-315	352	8	the	the	DET
bjmsr-315	352	9	pivotal	pivotal	ADJ
bjmsr-315	352	10	column	column	NOUN
bjmsr-315	352	11	during	during	ADP
bjmsr-315	352	12	the	the	DET
bjmsr-315	352	13	first	first	ADJ
bjmsr-315	352	14	m	m	NOUN
bjmsr-315	352	15	iterations	iteration	NOUN
bjmsr-315	352	16	to	to	ADP
bjmsr-315	352	17	the	the	DET
bjmsr-315	352	18	elements	element	NOUN
bjmsr-315	352	19	of	of	ADP
bjmsr-315	352	20	the	the	DET
bjmsr-315	352	21	vector	vector	NOUN
bjmsr-315	352	22	cj	cj	NOUN
bjmsr-315	352	23	and	and	CCONJ
bjmsr-315	352	24	dj	dj	NOUN
bjmsr-315	352	25	according	accord	VERB
bjmsr-315	352	26	to	to	ADP
bjmsr-315	352	27	the	the	DET
bjmsr-315	352	28	associated	associated	ADJ
bjmsr-315	352	29	maximum	maximum	ADJ
bjmsr-315	352	30	nonnegative	nonnegative	ADJ
bjmsr-315	352	31	marginal	marginal	ADJ
bjmsr-315	352	32	costs	cost	NOUN
bjmsr-315	352	33	.	.	PUNCT
bjmsr-315	353	1	the	the	DET
bjmsr-315	353	2	vector	vector	NOUN
bjmsr-315	353	3	that	that	PRON
bjmsr-315	353	4	leaves	leave	VERB
bjmsr-315	353	5	the	the	DET
bjmsr-315	353	6	basis	basis	NOUN
bjmsr-315	353	7	causes	cause	VERB
bjmsr-315	353	8	the	the	DET
bjmsr-315	353	9	maximum	maximum	ADJ
bjmsr-315	353	10	decrease	decrease	NOUN
bjmsr-315	353	11	in	in	ADP
bjmsr-315	353	12	the	the	DET
bjmsr-315	353	13	objective	objective	ADJ
bjmsr-315	353	14	function	function	NOUN
bjmsr-315	353	15	.	.	PUNCT
bjmsr-315	354	1	thus	thus	ADV
bjmsr-315	354	2	the	the	DET
bjmsr-315	354	3	pivot	pivot	NOUN
bjmsr-315	354	4	element	element	NOUN
bjmsr-315	354	5	is	be	AUX
bjmsr-315	354	6	not	not	PART
bjmsr-315	354	7	necessarily	necessarily	ADV
bjmsr-315	354	8	the	the	DET
bjmsr-315	354	9	same	same	ADJ
bjmsr-315	354	10	as	as	ADP
bjmsr-315	354	11	in	in	ADP
bjmsr-315	354	12	the	the	DET
bjmsr-315	354	13	usual	usual	ADJ
bjmsr-315	354	14	simplex	simplex	NOUN
bjmsr-315	354	15	.	.	PUNCT
bjmsr-315	355	1	the	the	DET
bjmsr-315	355	2	second	second	ADJ
bjmsr-315	355	3	stage	stage	NOUN
bjmsr-315	355	4	involves	involve	VERB
bjmsr-315	355	5	interchanging	interchange	VERB
bjmsr-315	355	6	nonbasic	nonbasic	PROPN
bjmsr-315	355	7	wi	wi	PROPN
bjmsr-315	355	8	or	or	CCONJ
bjmsr-315	355	9	vi	vi	PROPN
bjmsr-315	355	10	with	with	ADP
bjmsr-315	355	11	the	the	DET
bjmsr-315	355	12	basic	basic	ADJ
bjmsr-315	355	13	wi	wi	PROPN
bjmsr-315	355	14	or	or	CCONJ
bjmsr-315	355	15	vi	vi	PROPN
bjmsr-315	355	16	.	.	PUNCT
bjmsr-315	356	1	the	the	DET
bjmsr-315	356	2	basic	basic	ADJ
bjmsr-315	356	3	vectors	vector	NOUN
bjmsr-315	356	4	corresponding	correspond	VERB
bjmsr-315	356	5	to	to	ADP
bjmsr-315	356	6	cj	cj	PROPN
bjmsr-315	356	7	and	and	CCONJ
bjmsr-315	356	8	dj	dj	NOUN
bjmsr-315	356	9	are	be	AUX
bjmsr-315	356	10	not	not	PART
bjmsr-315	356	11	allowed	allow	VERB
bjmsr-315	356	12	to	to	PART
bjmsr-315	356	13	leave	leave	VERB
bjmsr-315	356	14	the	the	DET
bjmsr-315	356	15	basis	basis	NOUN
bjmsr-315	356	16	.	.	PUNCT
bjmsr-315	357	1	the	the	DET
bjmsr-315	357	2	algorithm	algorithm	NOUN
bjmsr-315	357	3	terminates	terminate	VERB
bjmsr-315	357	4	when	when	SCONJ
bjmsr-315	357	5	all	all	DET
bjmsr-315	357	6	marginal	marginal	ADJ
bjmsr-315	357	7	costs	cost	NOUN
bjmsr-315	357	8	are	be	AUX
bjmsr-315	357	9	nonpositive	nonpositive	ADJ
bjmsr-315	357	10	(	(	PUNCT
bjmsr-315	357	11	see	see	PROPN
bjmsr-315	357	12	,	,	PUNCT
bjmsr-315	357	13	kennedy	kennedy	PROPN
bjmsr-315	357	14	and	and	CCONJ
bjmsr-315	357	15	gentle	gentle	ADJ
bjmsr-315	357	16	(	(	PUNCT
bjmsr-315	357	17	1980	1980	NUM
bjmsr-315	357	18	)	)	PUNCT
bjmsr-315	357	19	)	)	PUNCT
bjmsr-315	357	20	.	.	PUNCT
bjmsr-315	358	1	fortran	fortran	NOUN
bjmsr-315	358	2	code	code	NOUN
bjmsr-315	358	3	for	for	ADP
bjmsr-315	358	4	this	this	DET
bjmsr-315	358	5	procedure	procedure	NOUN
bjmsr-315	358	6	is	be	AUX
bjmsr-315	358	7	given	give	VERB
bjmsr-315	358	8	by	by	ADP
bjmsr-315	358	9	barrodale	barrodale	NOUN
bjmsr-315	358	10	and	and	CCONJ
bjmsr-315	358	11	roberts	roberts	PROPN
bjmsr-315	358	12	(	(	PUNCT
bjmsr-315	358	13	1974	1974	NUM
bjmsr-315	358	14	)	)	PUNCT
bjmsr-315	358	15	.	.	PUNCT
bjmsr-315	359	1	peters	peters	PROPN
bjmsr-315	359	2	and	and	CCONJ
bjmsr-315	359	3	willms	willm	NOUN
bjmsr-315	359	4	(	(	PUNCT
bjmsr-315	359	5	1983	1983	NUM
bjmsr-315	359	6	)	)	PUNCT
bjmsr-315	359	7	give	give	VERB
bjmsr-315	359	8	algorithms	algorithm	NOUN
bjmsr-315	359	9	accompanying	accompany	VERB
bjmsr-315	359	10	with	with	ADP
bjmsr-315	359	11	computer	computer	NOUN
bjmsr-315	359	12	codes	code	NOUN
bjmsr-315	359	13	for	for	ADP
bjmsr-315	359	14	up	up	ADV
bjmsr-315	359	15	-	-	PUNCT
bjmsr-315	359	16	and	and	CCONJ
bjmsr-315	359	17	-	-	PUNCT
bjmsr-315	359	18	down	down	NOUN
bjmsr-315	359	19	dating	date	VERB
bjmsr-315	359	20	the	the	DET
bjmsr-315	359	21	solution	solution	NOUN
bjmsr-315	359	22	of	of	ADP
bjmsr-315	359	23	the	the	DET
bjmsr-315	359	24	problem	problem	NOUN
bjmsr-315	359	25	when	when	SCONJ
bjmsr-315	359	26	a	a	DET
bjmsr-315	359	27	column	column	NOUN
bjmsr-315	359	28	or	or	CCONJ
bjmsr-315	359	29	row	row	NOUN
bjmsr-315	359	30	inserted	insert	VERB
bjmsr-315	359	31	to	to	ADP
bjmsr-315	359	32	or	or	CCONJ
bjmsr-315	359	33	deleted	delete	VERB
bjmsr-315	359	34	from	from	ADP
bjmsr-315	359	35	x	x	PRON
bjmsr-315	359	36	,	,	PUNCT
bjmsr-315	359	37	or	or	CCONJ
bjmsr-315	359	38	y	y	PROPN
bjmsr-315	359	39	is	be	AUX
bjmsr-315	359	40	changed	change	VERB
bjmsr-315	359	41	.	.	PUNCT
bjmsr-315	360	1	these	these	DET
bjmsr-315	360	2	algorithms	algorithm	NOUN
bjmsr-315	360	3	are	be	AUX
bjmsr-315	360	4	all	all	PRON
bjmsr-315	360	5	based	base	VERB
bjmsr-315	360	6	on	on	ADP
bjmsr-315	360	7	barrodale	barrodale	NOUN
bjmsr-315	360	8	and	and	CCONJ
bjmsr-315	360	9	roberts	roberts	PROPN
bjmsr-315	360	10	(	(	PUNCT
bjmsr-315	360	11	1973,74	1973,74	NOUN
bjmsr-315	360	12	)	)	PUNCT
bjmsr-315	360	13	procedure	procedure	NOUN
bjmsr-315	360	14	.	.	PUNCT
bjmsr-315	361	1	abdelmalek	abdelmalek	PROPN
bjmsr-315	361	2	(	(	PUNCT
bjmsr-315	361	3	1974	1974	NUM
bjmsr-315	361	4	)	)	PUNCT
bjmsr-315	361	5	describes	describe	VERB
bjmsr-315	361	6	a	a	DET
bjmsr-315	361	7	dual	dual	ADJ
bjmsr-315	361	8	simplex	simplex	NOUN
bjmsr-315	361	9	algorithm	algorithm	NOUN
bjmsr-315	361	10	for	for	ADP
bjmsr-315	361	11	the	the	DET
bjmsr-315	361	12	l1	l1	PROPN
bjmsr-315	361	13	norm	norm	NOUN
bjmsr-315	361	14	problem	problem	NOUN
bjmsr-315	361	15	with	with	ADP
bjmsr-315	361	16	no	no	DET
bjmsr-315	361	17	use	use	NOUN
bjmsr-315	361	18	of	of	ADP
bjmsr-315	361	19	artificial	artificial	ADJ
bjmsr-315	361	20	variables	variable	NOUN
bjmsr-315	361	21	.	.	PUNCT
bjmsr-315	362	1	for	for	ADP
bjmsr-315	362	2	this	this	DET
bjmsr-315	362	3	algorithm	algorithm	NOUN
bjmsr-315	362	4	,	,	PUNCT
bjmsr-315	362	5	the	the	DET
bjmsr-315	362	6	haar	haar	NOUN
bjmsr-315	362	7	condition	condition	NOUN
bjmsr-315	362	8	(	(	PUNCT
bjmsr-315	362	9	see	see	INTJ
bjmsr-315	362	10	,	,	PUNCT
bjmsr-315	362	11	osborne	osborne	PROPN
bjmsr-315	362	12	(	(	PUNCT
bjmsr-315	362	13	1985	1985	NUM
bjmsr-315	362	14	)	)	PUNCT
bjmsr-315	362	15	,	,	PUNCT
bjmsr-315	362	16	moroney	moroney	PROPN
bjmsr-315	362	17	(	(	PUNCT
bjmsr-315	362	18	1961	1961	NUM
bjmsr-315	362	19	)	)	PUNCT
bjmsr-315	362	20	)	)	PUNCT
bjmsr-315	362	21	need	need	AUX
bjmsr-315	362	22	not	not	PART
bjmsr-315	362	23	be	be	AUX
bjmsr-315	362	24	satisfied	satisfied	ADJ
bjmsr-315	362	25	anymore	anymore	ADV
bjmsr-315	362	26	.	.	PUNCT
bjmsr-315	363	1	this	this	DET
bjmsr-315	363	2	algorithm	algorithm	NOUN
bjmsr-315	363	3	seemed	seem	VERB
bjmsr-315	363	4	to	to	PART
bjmsr-315	363	5	be	be	AUX
bjmsr-315	363	6	very	very	ADV
bjmsr-315	363	7	efficient	efficient	ADJ
bjmsr-315	363	8	at	at	ADP
bjmsr-315	363	9	the	the	DET
bjmsr-315	363	10	time	time	NOUN
bjmsr-315	363	11	of	of	ADP
bjmsr-315	363	12	publication	publication	NOUN
bjmsr-315	363	13	.	.	PUNCT
bjmsr-315	364	1	an	an	DET
bjmsr-315	364	2	improved	improved	ADJ
bjmsr-315	364	3	dual	dual	ADJ
bjmsr-315	364	4	simplex	simplex	NOUN
bjmsr-315	364	5	algorithm	algorithm	NOUN
bjmsr-315	364	6	for	for	ADP
bjmsr-315	364	7	l1	l1	PROPN
bjmsr-315	364	8	norm	norm	NOUN
bjmsr-315	364	9	approximation	approximation	NOUN
bjmsr-315	364	10	is	be	AUX
bjmsr-315	364	11	proposed	propose	VERB
bjmsr-315	364	12	by	by	ADP
bjmsr-315	364	13	abdelmalek	abdelmalek	PROPN
bjmsr-315	364	14	(	(	PUNCT
bjmsr-315	364	15	1975a	1975a	NUM
bjmsr-315	364	16	)	)	PUNCT
bjmsr-315	364	17	.	.	PUNCT
bjmsr-315	365	1	in	in	ADP
bjmsr-315	365	2	this	this	DET
bjmsr-315	365	3	algorithm	algorithm	NOUN
bjmsr-315	365	4	,	,	PUNCT
bjmsr-315	365	5	certain	certain	ADJ
bjmsr-315	365	6	intermediate	intermediate	ADJ
bjmsr-315	365	7	iterations	iteration	NOUN
bjmsr-315	365	8	are	be	AUX
bjmsr-315	365	9	skipped	skip	VERB
bjmsr-315	365	10	,	,	PUNCT
bjmsr-315	365	11	and	and	CCONJ
bjmsr-315	365	12	in	in	ADP
bjmsr-315	365	13	the	the	DET
bjmsr-315	365	14	case	case	NOUN
bjmsr-315	365	15	of	of	ADP
bjmsr-315	365	16	illconditioned	illconditione	VERB
bjmsr-315	365	17	problems	problem	NOUN
bjmsr-315	365	18	,	,	PUNCT
bjmsr-315	365	19	the	the	DET
bjmsr-315	365	20	basis	basis	NOUN
bjmsr-315	365	21	matrix	matrix	NOUN
bjmsr-315	365	22	can	can	AUX
bjmsr-315	365	23	lend	lend	VERB
bjmsr-315	365	24	itself	itself	PRON
bjmsr-315	365	25	to	to	AUX
bjmsr-315	365	26	triangular	triangular	NOUN
bjmsr-315	365	27	factorization	factorization	NOUN
bjmsr-315	365	28	and	and	CCONJ
bjmsr-315	365	29	thus	thus	ADV
bjmsr-315	365	30	ensure	ensure	VERB
bjmsr-315	365	31	a	a	DET
bjmsr-315	365	32	stable	stable	ADJ
bjmsr-315	365	33	solution	solution	NOUN
bjmsr-315	365	34	.	.	PUNCT
bjmsr-315	366	1	abdelmalek	abdelmalek	PROPN
bjmsr-315	366	2	(	(	PUNCT
bjmsr-315	366	3	1980a	1980a	NUM
bjmsr-315	366	4	)	)	PUNCT
bjmsr-315	366	5	improves	improve	VERB
bjmsr-315	366	6	his	his	PRON
bjmsr-315	366	7	previous	previous	ADJ
bjmsr-315	366	8	algorithm	algorithm	NOUN
bjmsr-315	366	9	by	by	ADP
bjmsr-315	366	10	using	use	VERB
bjmsr-315	366	11	triangular	triangular	NOUN
bjmsr-315	366	12	decomposition	decomposition	NOUN
bjmsr-315	366	13	.	.	PUNCT
bjmsr-315	367	1	a	a	DET
bjmsr-315	367	2	fortran	fortran	ADJ
bjmsr-315	367	3	translation	translation	NOUN
bjmsr-315	367	4	of	of	ADP
bjmsr-315	367	5	the	the	DET
bjmsr-315	367	6	algorithm	algorithm	NOUN
bjmsr-315	367	7	is	be	AUX
bjmsr-315	367	8	given	give	VERB
bjmsr-315	367	9	by	by	ADP
bjmsr-315	367	10	abdelmalek	abdelmalek	PROPN
bjmsr-315	367	11	(	(	PUNCT
bjmsr-315	367	12	1980b	1980b	NUM
bjmsr-315	367	13	)	)	PUNCT
bjmsr-315	367	14	.	.	PUNCT
bjmsr-315	368	1	sposito	sposito	PROPN
bjmsr-315	368	2	and	and	CCONJ
bjmsr-315	368	3	mccormick	mccormick	PROPN
bjmsr-315	368	4	and	and	CCONJ
bjmsr-315	368	5	kennedy	kennedy	PROPN
bjmsr-315	368	6	(	(	PUNCT
bjmsr-315	368	7	1975	1975	NUM
bjmsr-315	368	8	)	)	PUNCT
bjmsr-315	368	9	summarize	summarize	VERB
bjmsr-315	368	10	much	much	ADJ
bjmsr-315	368	11	of	of	ADP
bjmsr-315	368	12	the	the	DET
bjmsr-315	368	13	works	work	NOUN
bjmsr-315	368	14	on	on	ADP
bjmsr-315	368	15	l1	l1	PROPN
bjmsr-315	368	16	norm	norm	NOUN
bjmsr-315	368	17	estimation	estimation	NOUN
bjmsr-315	368	18	including	include	VERB
bjmsr-315	368	19	problem	problem	NOUN
bjmsr-315	368	20	statement	statement	NOUN
bjmsr-315	368	21	,	,	PUNCT
bjmsr-315	368	22	linear	linear	ADJ
bjmsr-315	368	23	programming	programming	NOUN
bjmsr-315	368	24	formulation	formulation	NOUN
bjmsr-315	368	25	,	,	PUNCT
bjmsr-315	368	26	efficient	efficient	ADJ
bjmsr-315	368	27	computational	computational	ADJ
bjmsr-315	368	28	algorithms	algorithm	NOUN
bjmsr-315	368	29	,	,	PUNCT
bjmsr-315	368	30	and	and	CCONJ
bjmsr-315	368	31	properties	property	NOUN
bjmsr-315	368	32	of	of	ADP
bjmsr-315	368	33	the	the	DET
bjmsr-315	368	34	estimators	estimator	NOUN
bjmsr-315	368	35	.	.	PUNCT
bjmsr-315	369	1	armstrong	armstrong	PROPN
bjmsr-315	369	2	and	and	CCONJ
bjmsr-315	369	3	kung	kung	PROPN
bjmsr-315	369	4	(	(	PUNCT
bjmsr-315	369	5	1978	1978	NUM
bjmsr-315	369	6	)	)	PUNCT
bjmsr-315	369	7	propose	propose	VERB
bjmsr-315	369	8	an	an	DET
bjmsr-315	369	9	algorithm	algorithm	NOUN
bjmsr-315	369	10	for	for	ADP
bjmsr-315	369	11	a	a	DET
bjmsr-315	369	12	simple	simple	ADJ
bjmsr-315	369	13	two	two	NUM
bjmsr-315	369	14	-	-	PUNCT
bjmsr-315	369	15	parameter	parameter	NOUN
bjmsr-315	369	16	l1	l1	PROPN
bjmsr-315	369	17	norm	norm	PROPN
bjmsr-315	369	18	regression	regression	PROPN
bjmsr-315	369	19	.	.	PUNCT
bjmsr-315	370	1	the	the	DET
bjmsr-315	370	2	method	method	NOUN
bjmsr-315	370	3	is	be	AUX
bjmsr-315	370	4	a	a	DET
bjmsr-315	370	5	specification	specification	NOUN
bjmsr-315	370	6	of	of	ADP
bjmsr-315	370	7	linear	linear	PROPN
bjmsr-315	370	8	programming	programming	NOUN
bjmsr-315	370	9	of	of	ADP
bjmsr-315	370	10	barrodale	barrodale	NOUN
bjmsr-315	370	11	and	and	CCONJ
bjmsr-315	370	12	roberts	roberts	PROPN
bjmsr-315	370	13	(	(	PUNCT
bjmsr-315	370	14	1973	1973	NUM
bjmsr-315	370	15	)	)	PUNCT
bjmsr-315	370	16	algorithm	algorithm	NOUN
bjmsr-315	370	17	.	.	PUNCT
bjmsr-315	371	1	a	a	DET
bjmsr-315	371	2	fortran	fortran	ADJ
bjmsr-315	371	3	code	code	NOUN
bjmsr-315	371	4	is	be	AUX
bjmsr-315	371	5	given	give	VERB
bjmsr-315	371	6	too	too	ADV
bjmsr-315	371	7	.	.	PUNCT
bjmsr-315	372	1	armstrong	armstrong	PROPN
bjmsr-315	372	2	and	and	CCONJ
bjmsr-315	372	3	frome	frome	PROPN
bjmsr-315	372	4	and	and	CCONJ
bjmsr-315	372	5	kung	kung	PROPN
bjmsr-315	372	6	(	(	PUNCT
bjmsr-315	372	7	1979	1979	NUM
bjmsr-315	372	8	)	)	PUNCT
bjmsr-315	372	9	use	use	NOUN
bjmsr-315	372	10	lu	lu	NOUN
bjmsr-315	372	11	(	(	PUNCT
bjmsr-315	372	12	lower	low	ADJ
bjmsr-315	372	13	-	-	PUNCT
bjmsr-315	372	14	upper	upper	ADJ
bjmsr-315	372	15	triangular	triangular	NOUN
bjmsr-315	372	16	)	)	PUNCT
bjmsr-315	372	17	decomposition	decomposition	NOUN
bjmsr-315	372	18	of	of	ADP
bjmsr-315	372	19	bartels	bartel	NOUN
bjmsr-315	372	20	and	and	CCONJ
bjmsr-315	372	21	golub	golub	PROPN
bjmsr-315	372	22	(	(	PUNCT
bjmsr-315	372	23	1969	1969	NUM
bjmsr-315	372	24	)	)	PUNCT
bjmsr-315	372	25	in	in	ADP
bjmsr-315	372	26	maintaining	maintain	VERB
bjmsr-315	372	27	the	the	DET
bjmsr-315	372	28	current	current	ADJ
bjmsr-315	372	29	basis	basis	NOUN
bjmsr-315	372	30	on	on	ADP
bjmsr-315	372	31	the	the	DET
bjmsr-315	372	32	revised	revise	VERB
bjmsr-315	372	33	simplex	simplex	NOUN
bjmsr-315	372	34	procedure	procedure	NOUN
bjmsr-315	372	35	.	.	PUNCT
bjmsr-315	373	1	a	a	DET
bjmsr-315	373	2	fortran	fortran	ADJ
bjmsr-315	373	3	translation	translation	NOUN
bjmsr-315	373	4	is	be	AUX
bjmsr-315	373	5	also	also	ADV
bjmsr-315	373	6	enclosed	enclose	VERB
bjmsr-315	373	7	.	.	PUNCT
bjmsr-315	374	1	armstrong	armstrong	PROPN
bjmsr-315	374	2	and	and	CCONJ
bjmsr-315	374	3	godfrey	godfrey	PROPN
bjmsr-315	374	4	(	(	PUNCT
bjmsr-315	374	5	1979	1979	NUM
bjmsr-315	374	6	)	)	PUNCT
bjmsr-315	374	7	show	show	VERB
bjmsr-315	374	8	that	that	SCONJ
bjmsr-315	374	9	the	the	DET
bjmsr-315	374	10	primal	primal	ADJ
bjmsr-315	374	11	method	method	NOUN
bjmsr-315	374	12	of	of	ADP
bjmsr-315	374	13	barrodale	barrodale	NOUN
bjmsr-315	374	14	and	and	CCONJ
bjmsr-315	374	15	roberts	roberts	PROPN
bjmsr-315	374	16	(	(	PUNCT
bjmsr-315	374	17	1973	1973	NUM
bjmsr-315	374	18	)	)	PUNCT
bjmsr-315	374	19	and	and	CCONJ
bjmsr-315	374	20	the	the	DET
bjmsr-315	374	21	dual	dual	ADJ
bjmsr-315	374	22	method	method	NOUN
bjmsr-315	374	23	of	of	ADP
bjmsr-315	374	24	abdelmalek	abdelmalek	PROPN
bjmsr-315	374	25	(	(	PUNCT
bjmsr-315	374	26	1975a	1975a	NUM
bjmsr-315	374	27	)	)	PUNCT
bjmsr-315	374	28	are	be	AUX
bjmsr-315	374	29	essentially	essentially	ADV
bjmsr-315	374	30	equivalent	equivalent	ADJ
bjmsr-315	374	31	.	.	PUNCT
bjmsr-315	375	1	with	with	ADP
bjmsr-315	375	2	a	a	DET
bjmsr-315	375	3	given	give	VERB
bjmsr-315	375	4	initial	initial	ADJ
bjmsr-315	375	5	basis	basis	NOUN
bjmsr-315	375	6	for	for	ADP
bjmsr-315	375	7	the	the	DET
bjmsr-315	375	8	two	two	NUM
bjmsr-315	375	9	methods	method	NOUN
bjmsr-315	375	10	,	,	PUNCT
bjmsr-315	375	11	they	they	PRON
bjmsr-315	375	12	show	show	VERB
bjmsr-315	375	13	that	that	SCONJ
bjmsr-315	375	14	both	both	DET
bjmsr-315	375	15	algorithms	algorithm	NOUN
bjmsr-315	375	16	will	will	AUX
bjmsr-315	375	17	generate	generate	VERB
bjmsr-315	375	18	corresponding	corresponding	ADJ
bjmsr-315	375	19	bases	basis	NOUN
bjmsr-315	375	20	at	at	ADP
bjmsr-315	375	21	each	each	DET
bjmsr-315	375	22	iteration	iteration	NOUN
bjmsr-315	375	23	.	.	PUNCT
bjmsr-315	376	1	the	the	DET
bjmsr-315	376	2	only	only	ADJ
bjmsr-315	376	3	difference	difference	NOUN
bjmsr-315	376	4	is	be	AUX
bjmsr-315	376	5	the	the	DET
bjmsr-315	376	6	choice	choice	NOUN
bjmsr-315	376	7	of	of	ADP
bjmsr-315	376	8	initial	initial	ADJ
bjmsr-315	376	9	basis	basis	NOUN
bjmsr-315	376	10	and	and	CCONJ
bjmsr-315	376	11	heuristic	heuristic	ADJ
bjmsr-315	376	12	rules	rule	NOUN
bjmsr-315	376	13	for	for	ADP
bjmsr-315	376	14	breaking	break	VERB
bjmsr-315	376	15	ties	tie	NOUN
bjmsr-315	376	16	.	.	PUNCT
bjmsr-315	377	1	armstrong	armstrong	PROPN
bjmsr-315	377	2	and	and	CCONJ
bjmsr-315	377	3	kung	kung	PROPN
bjmsr-315	377	4	(	(	PUNCT
bjmsr-315	377	5	1982b	1982b	NUM
bjmsr-315	377	6	)	)	PUNCT
bjmsr-315	377	7	present	present	VERB
bjmsr-315	377	8	a	a	DET
bjmsr-315	377	9	dual	dual	ADJ
bjmsr-315	377	10	linear	linear	NOUN
bjmsr-315	377	11	programming	programming	NOUN
bjmsr-315	377	12	formulation	formulation	NOUN
bjmsr-315	377	13	for	for	ADP
bjmsr-315	377	14	the	the	DET
bjmsr-315	377	15	problem	problem	NOUN
bjmsr-315	377	16	.	.	PUNCT
bjmsr-315	378	1	various	various	ADJ
bjmsr-315	378	2	basis	basis	NOUN
bjmsr-315	378	3	entry	entry	NOUN
bjmsr-315	378	4	and	and	CCONJ
bjmsr-315	378	5	initialization	initialization	NOUN
bjmsr-315	378	6	procedures	procedure	NOUN
bjmsr-315	378	7	are	be	AUX
bjmsr-315	378	8	considered	consider	VERB
bjmsr-315	378	9	.	.	PUNCT
bjmsr-315	379	1	it	it	PRON
bjmsr-315	379	2	has	have	AUX
bjmsr-315	379	3	been	be	AUX
bjmsr-315	379	4	shown	show	VERB
bjmsr-315	379	5	that	that	SCONJ
bjmsr-315	379	6	the	the	DET
bjmsr-315	379	7	dual	dual	ADJ
bjmsr-315	379	8	approach	approach	NOUN
bjmsr-315	379	9	is	be	AUX
bjmsr-315	379	10	superior	superior	ADJ
bjmsr-315	379	11	to	to	PART
bjmsr-315	379	12	primal	primal	ADJ
bjmsr-315	379	13	one	one	NUM
bjmsr-315	379	14	if	if	SCONJ
bjmsr-315	379	15	a	a	DET
bjmsr-315	379	16	good	good	ADJ
bjmsr-315	379	17	dual	dual	ADJ
bjmsr-315	379	18	feasible	feasible	ADJ
bjmsr-315	379	19	solution	solution	NOUN
bjmsr-315	379	20	is	be	AUX
bjmsr-315	379	21	readily	readily	ADV
bjmsr-315	379	22	available	available	ADJ
bjmsr-315	379	23	(	(	PUNCT
bjmsr-315	379	24	see	see	VERB
bjmsr-315	379	25	also	also	ADV
bjmsr-315	379	26	,	,	PUNCT
bjmsr-315	379	27	steiger	steiger	PROPN
bjmsr-315	379	28	(	(	PUNCT
bjmsr-315	379	29	1980	1980	NUM
bjmsr-315	379	30	)	)	PUNCT
bjmsr-315	379	31	)	)	PUNCT
bjmsr-315	379	32	.	.	PUNCT
bjmsr-315	380	1	banks	bank	NOUN
bjmsr-315	380	2	and	and	CCONJ
bjmsr-315	380	3	taylor	taylor	PROPN
bjmsr-315	380	4	(	(	PUNCT
bjmsr-315	380	5	1980	1980	NUM
bjmsr-315	380	6	)	)	PUNCT
bjmsr-315	380	7	suggest	suggest	VERB
bjmsr-315	380	8	a	a	DET
bjmsr-315	380	9	modification	modification	NOUN
bjmsr-315	380	10	of	of	ADP
bjmsr-315	380	11	barrodale	barrodale	NOUN
bjmsr-315	380	12	and	and	CCONJ
bjmsr-315	380	13	roberts	roberts	PROPN
bjmsr-315	380	14	(	(	PUNCT
bjmsr-315	380	15	1973	1973	NUM
bjmsr-315	380	16	)	)	PUNCT
bjmsr-315	380	17	algorithm	algorithm	NOUN
bjmsr-315	380	18	.	.	PUNCT
bjmsr-315	381	1	the	the	DET
bjmsr-315	381	2	objective	objective	ADJ
bjmsr-315	381	3	function	function	NOUN
bjmsr-315	381	4	is	be	AUX
bjmsr-315	381	5	altered	alter	VERB
bjmsr-315	381	6	to	to	PART
bjmsr-315	381	7	include	include	VERB
bjmsr-315	381	8	magnitudes	magnitude	NOUN
bjmsr-315	381	9	of	of	ADP
bjmsr-315	381	10	the	the	DET
bjmsr-315	381	11	elements	element	NOUN
bjmsr-315	381	12	of	of	ADP
bjmsr-315	381	13	both	both	DET
bjmsr-315	381	14	errors	error	NOUN
bjmsr-315	381	15	and	and	CCONJ
bjmsr-315	381	16	solution	solution	NOUN
bjmsr-315	381	17	vectors	vector	NOUN
bjmsr-315	381	18	.	.	PUNCT
bjmsr-315	382	1	for	for	ADP
bjmsr-315	382	2	a	a	DET
bjmsr-315	382	3	general	general	ADJ
bjmsr-315	382	4	discussion	discussion	NOUN
bjmsr-315	382	5	on	on	ADP
bjmsr-315	382	6	simplex	simplex	NOUN
bjmsr-315	382	7	for	for	ADP
bjmsr-315	382	8	piecewise	piecewise	NOUN
bjmsr-315	382	9	linear	linear	NOUN
bjmsr-315	382	10	programming	programming	NOUN
bjmsr-315	382	11	see	see	VERB
bjmsr-315	382	12	fourer	fourer	NOUN
bjmsr-315	382	13	(	(	PUNCT
bjmsr-315	382	14	1985a	1985a	NUM
bjmsr-315	382	15	,	,	PUNCT
bjmsr-315	382	16	b	b	NOUN
bjmsr-315	382	17	)	)	PUNCT
bjmsr-315	382	18	and	and	CCONJ
bjmsr-315	382	19	for	for	ADP
bjmsr-315	382	20	a	a	DET
bjmsr-315	382	21	survey	survey	NOUN
bjmsr-315	382	22	of	of	ADP
bjmsr-315	382	23	the	the	DET
bjmsr-315	382	24	corresponding	corresponding	ADJ
bjmsr-315	382	25	copyright	copyright	NOUN
bjmsr-315	382	26	©	©	PROPN
bjmsr-315	382	27	cc	cc	PROPN
bjmsr-315	382	28	-	-	PUNCT
bjmsr-315	382	29	by	by	ADP
bjmsr-315	382	30	-	-	PUNCT
bjmsr-315	382	31	nc	nc	PROPN
bjmsr-315	382	32	2019	2019	NUM
bjmsr-315	382	33	,	,	PUNCT
bjmsr-315	382	34	bjmsr	bjmsr	PROPN
bjmsr-315	382	35	www.cribfb.com/journal/index.php/bjmsr	www.cribfb.com/journal/index.php/bjmsr	X
bjmsr-315	382	36	bangladesh	bangladesh	PROPN
bjmsr-315	382	37	journal	journal	PROPN
bjmsr-315	382	38	of	of	ADP
bjmsr-315	382	39	multidisciplinary	multidisciplinary	ADJ
bjmsr-315	382	40	scientific	scientific	ADJ
bjmsr-315	382	41	research	research	NOUN
bjmsr-315	382	42	vol	vol	NOUN
bjmsr-315	382	43	.	.	PROPN
bjmsr-315	383	1	1	1	NUM
bjmsr-315	383	2	,	,	PUNCT
bjmsr-315	383	3	no	no	INTJ
bjmsr-315	383	4	.	.	NOUN
bjmsr-315	383	5	1	1	NUM
bjmsr-315	383	6	;	;	PUNCT
bjmsr-315	383	7	2019	2019	NUM
bjmsr-315	383	8	60	60	NUM
bjmsr-315	383	9	problem	problem	NOUN
bjmsr-315	383	10	on	on	ADP
bjmsr-315	383	11	the	the	DET
bjmsr-315	383	12	l1	l1	PROPN
bjmsr-315	383	13	norm	norm	NOUN
bjmsr-315	383	14	see	see	VERB
bjmsr-315	383	15	fourer	fourer	NOUN
bjmsr-315	383	16	(	(	PUNCT
bjmsr-315	383	17	1986	1986	NUM
bjmsr-315	383	18	)	)	PUNCT
bjmsr-315	383	19	.	.	PUNCT
bjmsr-315	384	1	narula	narula	NOUN
bjmsr-315	384	2	and	and	CCONJ
bjmsr-315	384	3	wellington	wellington	PROPN
bjmsr-315	384	4	(	(	PUNCT
bjmsr-315	384	5	1987	1987	NUM
bjmsr-315	384	6	)	)	PUNCT
bjmsr-315	384	7	propose	propose	VERB
bjmsr-315	384	8	an	an	DET
bjmsr-315	384	9	efficient	efficient	ADJ
bjmsr-315	384	10	linear	linear	NOUN
bjmsr-315	384	11	programming	programming	NOUN
bjmsr-315	384	12	algorithm	algorithm	NOUN
bjmsr-315	384	13	to	to	PART
bjmsr-315	384	14	solve	solve	VERB
bjmsr-315	384	15	both	both	DET
bjmsr-315	384	16	l1	l1	PROPN
bjmsr-315	384	17	and	and	CCONJ
bjmsr-315	384	18	l¥	l¥	PROPN
bjmsr-315	384	19	norms	norm	VERB
bjmsr-315	384	20	linear	linear	VERB
bjmsr-315	384	21	multiple	multiple	ADJ
bjmsr-315	384	22	regressions	regression	NOUN
bjmsr-315	384	23	.	.	PUNCT
bjmsr-315	385	1	the	the	DET
bjmsr-315	385	2	algorithm	algorithm	NOUN
bjmsr-315	385	3	exploits	exploit	VERB
bjmsr-315	385	4	the	the	DET
bjmsr-315	385	5	special	special	ADJ
bjmsr-315	385	6	structure	structure	NOUN
bjmsr-315	385	7	and	and	CCONJ
bjmsr-315	385	8	similarities	similarity	NOUN
bjmsr-315	385	9	between	between	ADP
bjmsr-315	385	10	the	the	DET
bjmsr-315	385	11	two	two	NUM
bjmsr-315	385	12	problems	problem	NOUN
bjmsr-315	385	13	.	.	PUNCT
bjmsr-315	386	1	brennan	brennan	PROPN
bjmsr-315	386	2	and	and	CCONJ
bjmsr-315	386	3	seiford	seiford	PROPN
bjmsr-315	386	4	(	(	PUNCT
bjmsr-315	386	5	1987	1987	NUM
bjmsr-315	386	6	)	)	PUNCT
bjmsr-315	386	7	develop	develop	VERB
bjmsr-315	386	8	a	a	DET
bjmsr-315	386	9	geometrical	geometrical	ADJ
bjmsr-315	386	10	interpretation	interpretation	NOUN
bjmsr-315	386	11	of	of	ADP
bjmsr-315	386	12	linear	linear	PROPN
bjmsr-315	386	13	programming	programming	NOUN
bjmsr-315	386	14	in	in	ADP
bjmsr-315	386	15	l1	l1	PROPN
bjmsr-315	386	16	norm	norm	PROPN
bjmsr-315	386	17	regression	regression	NOUN
bjmsr-315	386	18	.	.	PUNCT
bjmsr-315	387	1	they	they	PRON
bjmsr-315	387	2	give	give	VERB
bjmsr-315	387	3	a	a	DET
bjmsr-315	387	4	geometric	geometric	ADJ
bjmsr-315	387	5	insight	insight	NOUN
bjmsr-315	387	6	into	into	ADP
bjmsr-315	387	7	the	the	DET
bjmsr-315	387	8	solving	solving	NOUN
bjmsr-315	387	9	process	process	NOUN
bjmsr-315	387	10	in	in	ADP
bjmsr-315	387	11	the	the	DET
bjmsr-315	387	12	space	space	NOUN
bjmsr-315	387	13	of	of	ADP
bjmsr-315	387	14	observations	observation	NOUN
bjmsr-315	387	15	.	.	PUNCT
bjmsr-315	388	1	mcconnell	mcconnell	PROPN
bjmsr-315	388	2	(	(	PUNCT
bjmsr-315	388	3	1987	1987	NUM
bjmsr-315	388	4	)	)	PUNCT
bjmsr-315	388	5	shows	show	VERB
bjmsr-315	388	6	how	how	SCONJ
bjmsr-315	388	7	the	the	DET
bjmsr-315	388	8	method	method	NOUN
bjmsr-315	388	9	of	of	ADP
bjmsr-315	388	10	vanishing	vanish	VERB
bjmsr-315	388	11	jacobians	jacobian	NOUN
bjmsr-315	388	12	which	which	PRON
bjmsr-315	388	13	has	have	AUX
bjmsr-315	388	14	been	be	AUX
bjmsr-315	388	15	used	use	VERB
bjmsr-315	388	16	to	to	PART
bjmsr-315	388	17	optimize	optimize	VERB
bjmsr-315	388	18	quadratic	quadratic	ADJ
bjmsr-315	388	19	programming	programming	NOUN
bjmsr-315	388	20	problems	problem	NOUN
bjmsr-315	388	21	can	can	AUX
bjmsr-315	388	22	also	also	ADV
bjmsr-315	388	23	be	be	AUX
bjmsr-315	388	24	used	use	VERB
bjmsr-315	388	25	to	to	PART
bjmsr-315	388	26	solve	solve	VERB
bjmsr-315	388	27	the	the	DET
bjmsr-315	388	28	special	special	ADJ
bjmsr-315	388	29	linear	linear	NOUN
bjmsr-315	388	30	programming	programming	NOUN
bjmsr-315	388	31	problem	problem	NOUN
bjmsr-315	388	32	associated	associate	VERB
bjmsr-315	388	33	with	with	ADP
bjmsr-315	388	34	computing	compute	VERB
bjmsr-315	388	35	linear	linear	PROPN
bjmsr-315	388	36	discrete	discrete	ADJ
bjmsr-315	388	37	l1	l1	PROPN
bjmsr-315	388	38	norm	norm	NOUN
bjmsr-315	388	39	approximation	approximation	NOUN
bjmsr-315	388	40	.	.	PUNCT
bjmsr-315	389	1	for	for	ADP
bjmsr-315	389	2	the	the	DET
bjmsr-315	389	3	possibility	possibility	NOUN
bjmsr-315	389	4	of	of	ADP
bjmsr-315	389	5	applying	apply	VERB
bjmsr-315	389	6	other	other	ADJ
bjmsr-315	389	7	types	type	NOUN
bjmsr-315	389	8	of	of	ADP
bjmsr-315	389	9	linear	linear	ADJ
bjmsr-315	389	10	programming	programming	NOUN
bjmsr-315	389	11	solutions	solution	NOUN
bjmsr-315	389	12	such	such	ADJ
bjmsr-315	389	13	as	as	ADP
bjmsr-315	389	14	karmarkar	karmarkar	NOUN
bjmsr-315	389	15	solution	solution	NOUN
bjmsr-315	389	16	to	to	ADP
bjmsr-315	389	17	l1	l1	PROPN
bjmsr-315	389	18	norm	norm	PROPN
bjmsr-315	389	19	problem	problem	NOUN
bjmsr-315	389	20	see	see	VERB
bjmsr-315	389	21	meketon	meketon	NOUN
bjmsr-315	389	22	(	(	PUNCT
bjmsr-315	389	23	1986	1986	NUM
bjmsr-315	389	24	)	)	PUNCT
bjmsr-315	389	25	.	.	PUNCT
bjmsr-315	390	1	other	other	ADJ
bjmsr-315	390	2	algorithms	algorithm	NOUN
bjmsr-315	390	3	this	this	DET
bjmsr-315	390	4	category	category	NOUN
bjmsr-315	390	5	consists	consist	VERB
bjmsr-315	390	6	of	of	ADP
bjmsr-315	390	7	algorithms	algorithm	NOUN
bjmsr-315	390	8	which	which	PRON
bjmsr-315	390	9	were	be	AUX
bjmsr-315	390	10	not	not	PART
bjmsr-315	390	11	classified	classify	VERB
bjmsr-315	390	12	in	in	ADP
bjmsr-315	390	13	the	the	DET
bjmsr-315	390	14	two	two	NUM
bjmsr-315	390	15	last	last	ADJ
bjmsr-315	390	16	sections	section	NOUN
bjmsr-315	390	17	.	.	PUNCT
bjmsr-315	391	1	rice	rice	NOUN
bjmsr-315	391	2	(	(	PUNCT
bjmsr-315	391	3	1964c	1964c	NUM
bjmsr-315	391	4	)	)	PUNCT
bjmsr-315	391	5	applies	apply	VERB
bjmsr-315	391	6	the	the	DET
bjmsr-315	391	7	bisection	bisection	NOUN
bjmsr-315	391	8	method	method	NOUN
bjmsr-315	391	9	to	to	PART
bjmsr-315	391	10	l1	l1	PROPN
bjmsr-315	391	11	norm	norm	PROPN
bjmsr-315	391	12	regression	regression	NOUN
bjmsr-315	391	13	.	.	PUNCT
bjmsr-315	392	1	in	in	ADP
bjmsr-315	392	2	this	this	DET
bjmsr-315	392	3	method	method	NOUN
bjmsr-315	392	4	at	at	ADP
bjmsr-315	392	5	each	each	DET
bjmsr-315	392	6	step	step	NOUN
bjmsr-315	392	7	,	,	PUNCT
bjmsr-315	392	8	the	the	DET
bjmsr-315	392	9	domain	domain	NOUN
bjmsr-315	392	10	of	of	ADP
bjmsr-315	392	11	s	s	PROPN
bjmsr-315	392	12	is	be	AUX
bjmsr-315	392	13	broken	break	VERB
bjmsr-315	392	14	to	to	ADP
bjmsr-315	392	15	two	two	NUM
bjmsr-315	392	16	segments	segment	NOUN
bjmsr-315	392	17	,	,	PUNCT
bjmsr-315	392	18	and	and	CCONJ
bjmsr-315	392	19	the	the	DET
bjmsr-315	392	20	appropriate	appropriate	ADJ
bjmsr-315	392	21	segment	segment	NOUN
bjmsr-315	392	22	is	be	AUX
bjmsr-315	392	23	selected	select	VERB
bjmsr-315	392	24	for	for	ADP
bjmsr-315	392	25	the	the	DET
bjmsr-315	392	26	next	next	ADJ
bjmsr-315	392	27	iteration	iteration	NOUN
bjmsr-315	392	28	.	.	PUNCT
bjmsr-315	393	1	the	the	DET
bjmsr-315	393	2	solution	solution	NOUN
bjmsr-315	393	3	is	be	AUX
bjmsr-315	393	4	reached	reach	VERB
bjmsr-315	393	5	when	when	SCONJ
bjmsr-315	393	6	the	the	DET
bjmsr-315	393	7	last	last	ADJ
bjmsr-315	393	8	segment	segment	NOUN
bjmsr-315	393	9	is	be	AUX
bjmsr-315	393	10	less	less	ADJ
bjmsr-315	393	11	than	than	ADP
bjmsr-315	393	12	a	a	DET
bjmsr-315	393	13	predetermined	predetermine	VERB
bjmsr-315	393	14	small	small	ADJ
bjmsr-315	393	15	value	value	NOUN
bjmsr-315	393	16	.	.	PUNCT
bjmsr-315	394	1	abdelmalek	abdelmalek	PROPN
bjmsr-315	394	2	(	(	PUNCT
bjmsr-315	394	3	1971	1971	NUM
bjmsr-315	394	4	)	)	PUNCT
bjmsr-315	394	5	develops	develop	VERB
bjmsr-315	394	6	an	an	DET
bjmsr-315	394	7	algorithm	algorithm	NOUN
bjmsr-315	394	8	for	for	ADP
bjmsr-315	394	9	fitting	fitting	ADJ
bjmsr-315	394	10	functions	function	NOUN
bjmsr-315	394	11	to	to	PART
bjmsr-315	394	12	discrete	discrete	VERB
bjmsr-315	394	13	data	datum	NOUN
bjmsr-315	394	14	points	point	NOUN
bjmsr-315	394	15	and	and	CCONJ
bjmsr-315	394	16	solving	solve	VERB
bjmsr-315	394	17	the	the	DET
bjmsr-315	394	18	overdetermined	overdetermine	VERB
bjmsr-315	394	19	system	system	NOUN
bjmsr-315	394	20	of	of	ADP
bjmsr-315	394	21	linear	linear	PROPN
bjmsr-315	394	22	equations	equation	NOUN
bjmsr-315	394	23	.	.	PUNCT
bjmsr-315	395	1	the	the	DET
bjmsr-315	395	2	procedure	procedure	NOUN
bjmsr-315	395	3	is	be	AUX
bjmsr-315	395	4	based	base	VERB
bjmsr-315	395	5	on	on	ADP
bjmsr-315	395	6	determining	determine	VERB
bjmsr-315	395	7	l1	l1	PROPN
bjmsr-315	395	8	norm	norm	NOUN
bjmsr-315	395	9	solution	solution	NOUN
bjmsr-315	395	10	as	as	ADP
bjmsr-315	395	11	the	the	DET
bjmsr-315	395	12	limiting	limit	VERB
bjmsr-315	395	13	case	case	NOUN
bjmsr-315	395	14	of	of	ADP
bjmsr-315	395	15	lp	lp	PROPN
bjmsr-315	395	16	norm	norm	NOUN
bjmsr-315	395	17	approximation	approximation	NOUN
bjmsr-315	395	18	when	when	SCONJ
bjmsr-315	395	19	p	p	NOUN
bjmsr-315	395	20	tends	tend	VERB
bjmsr-315	395	21	to	to	ADP
bjmsr-315	395	22	one	one	NUM
bjmsr-315	395	23	from	from	ADP
bjmsr-315	395	24	right	right	ADV
bjmsr-315	395	25	in	in	ADP
bjmsr-315	395	26	the	the	DET
bjmsr-315	395	27	limit	limit	NOUN
bjmsr-315	395	28	.	.	PUNCT
bjmsr-315	396	1	this	this	DET
bjmsr-315	396	2	technique	technique	NOUN
bjmsr-315	396	3	thus	thus	ADV
bjmsr-315	396	4	obtains	obtain	VERB
bjmsr-315	396	5	a	a	DET
bjmsr-315	396	6	solution	solution	NOUN
bjmsr-315	396	7	to	to	ADP
bjmsr-315	396	8	a	a	DET
bjmsr-315	396	9	linear	linear	ADJ
bjmsr-315	396	10	problem	problem	NOUN
bjmsr-315	396	11	by	by	ADP
bjmsr-315	396	12	solving	solve	VERB
bjmsr-315	396	13	a	a	DET
bjmsr-315	396	14	sequence	sequence	NOUN
bjmsr-315	396	15	of	of	ADP
bjmsr-315	396	16	nonlinear	nonlinear	ADJ
bjmsr-315	396	17	problems	problem	NOUN
bjmsr-315	396	18	.	.	PUNCT
bjmsr-315	397	1	schlossmacher	schlossmacher	NOUN
bjmsr-315	397	2	(	(	PUNCT
bjmsr-315	397	3	1973	1973	NUM
bjmsr-315	397	4	)	)	PUNCT
bjmsr-315	397	5	computed	compute	VERB
bjmsr-315	397	6	the	the	DET
bjmsr-315	397	7	l1	l1	PROPN
bjmsr-315	397	8	norm	norm	NOUN
bjmsr-315	397	9	estimates	estimate	NOUN
bjmsr-315	397	10	of	of	ADP
bjmsr-315	397	11	regression	regression	NOUN
bjmsr-315	397	12	parameters	parameter	NOUN
bjmsr-315	397	13	by	by	ADP
bjmsr-315	397	14	an	an	DET
bjmsr-315	397	15	iterative	iterative	NOUN
bjmsr-315	397	16	weighted	weight	VERB
bjmsr-315	397	17	least	least	ADJ
bjmsr-315	397	18	squares	square	NOUN
bjmsr-315	397	19	procedure	procedure	NOUN
bjmsr-315	397	20	.	.	PUNCT
bjmsr-315	398	1	instead	instead	ADV
bjmsr-315	398	2	of	of	ADP
bjmsr-315	398	3	minimizing	minimize	VERB
bjmsr-315	398	4	the	the	DET
bjmsr-315	398	5	sum	sum	NOUN
bjmsr-315	398	6	of	of	ADP
bjmsr-315	398	7	absolute	absolute	ADJ
bjmsr-315	398	8	deviations	deviation	NOUN
bjmsr-315	398	9	,	,	PUNCT
bjmsr-315	398	10	he	he	PRON
bjmsr-315	398	11	minimized	minimize	VERB
bjmsr-315	398	12	he	he	PRON
bjmsr-315	398	13	sum	sum	NOUN
bjmsr-315	398	14	of	of	ADP
bjmsr-315	398	15	weighted	weight	VERB
bjmsr-315	398	16	squared	square	VERB
bjmsr-315	398	17	errors	error	NOUN
bjmsr-315	398	18	with	with	ADP
bjmsr-315	398	19	1/	1/	NUM
bjmsr-315	398	20	│	│	NUM
bjmsr-315	398	21	ui	ui	NOUN
bjmsr-315	398	22	│	│	PUNCT
bjmsr-315	398	23	as	as	ADP
bjmsr-315	398	24	weights	weight	NOUN
bjmsr-315	398	25	.	.	PUNCT
bjmsr-315	399	1	once	once	ADV
bjmsr-315	399	2	the	the	DET
bjmsr-315	399	3	least	least	ADJ
bjmsr-315	399	4	squares	square	NOUN
bjmsr-315	399	5	is	be	AUX
bjmsr-315	399	6	applied	apply	VERB
bjmsr-315	399	7	to	to	ADP
bjmsr-315	399	8	the	the	DET
bjmsr-315	399	9	problem	problem	NOUN
bjmsr-315	399	10	and	and	CCONJ
bjmsr-315	399	11	residuals	residual	NOUN
bjmsr-315	399	12	are	be	AUX
bjmsr-315	399	13	computed	compute	VERB
bjmsr-315	399	14	.	.	PUNCT
bjmsr-315	400	1	the	the	DET
bjmsr-315	400	2	absolute	absolute	ADJ
bjmsr-315	400	3	value	value	NOUN
bjmsr-315	400	4	of	of	ADP
bjmsr-315	400	5	the	the	DET
bjmsr-315	400	6	inverse	inverse	NOUN
bjmsr-315	400	7	of	of	ADP
bjmsr-315	400	8	the	the	DET
bjmsr-315	400	9	residuals	residual	NOUN
bjmsr-315	400	10	are	be	AUX
bjmsr-315	400	11	again	again	ADV
bjmsr-315	400	12	used	use	VERB
bjmsr-315	400	13	as	as	ADP
bjmsr-315	400	14	corresponding	correspond	VERB
bjmsr-315	400	15	weights	weight	NOUN
bjmsr-315	400	16	in	in	ADP
bjmsr-315	400	17	the	the	DET
bjmsr-315	400	18	next	next	ADJ
bjmsr-315	400	19	iteration	iteration	NOUN
bjmsr-315	400	20	for	for	ADP
bjmsr-315	400	21	minimizing	minimize	VERB
bjmsr-315	400	22	the	the	DET
bjmsr-315	400	23	sum	sum	NOUN
bjmsr-315	400	24	of	of	ADP
bjmsr-315	400	25	weighted	weight	VERB
bjmsr-315	400	26	squared	square	VERB
bjmsr-315	400	27	errors	error	NOUN
bjmsr-315	400	28	(	(	PUNCT
bjmsr-315	400	29	see	see	VERB
bjmsr-315	400	30	also	also	ADV
bjmsr-315	400	31	,	,	PUNCT
bjmsr-315	400	32	holland	holland	PROPN
bjmsr-315	400	33	and	and	CCONJ
bjmsr-315	400	34	welsh	welsh	PROPN
bjmsr-315	400	35	(	(	PUNCT
bjmsr-315	400	36	1977	1977	NUM
bjmsr-315	400	37	)	)	PUNCT
bjmsr-315	400	38	)	)	PUNCT
bjmsr-315	400	39	.	.	PUNCT
bjmsr-315	401	1	fair	fair	ADJ
bjmsr-315	401	2	(	(	PUNCT
bjmsr-315	401	3	1974	1974	NUM
bjmsr-315	401	4	)	)	PUNCT
bjmsr-315	401	5	observed	observe	VERB
bjmsr-315	401	6	that	that	SCONJ
bjmsr-315	401	7	the	the	DET
bjmsr-315	401	8	estimated	estimate	VERB
bjmsr-315	401	9	values	value	NOUN
bjmsr-315	401	10	of	of	ADP
bjmsr-315	401	11	ß	ß	PRON
bjmsr-315	401	12	did	do	AUX
bjmsr-315	401	13	not	not	PART
bjmsr-315	401	14	change	change	VERB
bjmsr-315	401	15	after	after	ADP
bjmsr-315	401	16	the	the	DET
bjmsr-315	401	17	second	second	ADJ
bjmsr-315	401	18	or	or	CCONJ
bjmsr-315	401	19	third	third	ADJ
bjmsr-315	401	20	iterations	iteration	NOUN
bjmsr-315	401	21	.	.	PUNCT
bjmsr-315	402	1	in	in	ADP
bjmsr-315	402	2	cases	case	NOUN
bjmsr-315	402	3	where	where	SCONJ
bjmsr-315	402	4	any	any	DET
bjmsr-315	402	5	residual	residual	ADJ
bjmsr-315	402	6	is	be	AUX
bjmsr-315	402	7	zero	zero	NUM
bjmsr-315	402	8	,	,	PUNCT
bjmsr-315	402	9	the	the	DET
bjmsr-315	402	10	continuation	continuation	NOUN
bjmsr-315	402	11	of	of	ADP
bjmsr-315	402	12	the	the	DET
bjmsr-315	402	13	procedure	procedure	NOUN
bjmsr-315	402	14	is	be	AUX
bjmsr-315	402	15	impossible	impossible	ADJ
bjmsr-315	402	16	,	,	PUNCT
bjmsr-315	402	17	because	because	SCONJ
bjmsr-315	402	18	the	the	DET
bjmsr-315	402	19	corresponding	corresponding	ADJ
bjmsr-315	402	20	weight	weight	NOUN
bjmsr-315	402	21	to	to	ADP
bjmsr-315	402	22	this	this	DET
bjmsr-315	402	23	residual	residual	ADJ
bjmsr-315	402	24	is	be	AUX
bjmsr-315	402	25	infinite	infinite	ADJ
bjmsr-315	402	26	.	.	PUNCT
bjmsr-315	403	1	this	this	DET
bjmsr-315	403	2	problem	problem	NOUN
bjmsr-315	403	3	is	be	AUX
bjmsr-315	403	4	also	also	ADV
bjmsr-315	403	5	discussed	discuss	VERB
bjmsr-315	403	6	by	by	ADP
bjmsr-315	403	7	sposito	sposito	NOUN
bjmsr-315	403	8	and	and	CCONJ
bjmsr-315	403	9	kennedy	kennedy	PROPN
bjmsr-315	403	10	and	and	CCONJ
bjmsr-315	403	11	gentle	gentle	ADJ
bjmsr-315	403	12	(	(	PUNCT
bjmsr-315	403	13	1977	1977	NUM
bjmsr-315	403	14	)	)	PUNCT
bjmsr-315	403	15	,	,	PUNCT
bjmsr-315	403	16	soliman	soliman	NOUN
bjmsr-315	403	17	and	and	CCONJ
bjmsr-315	403	18	christensen	christensen	PROPN
bjmsr-315	403	19	and	and	CCONJ
bjmsr-315	403	20	rouhi	rouhi	PROPN
bjmsr-315	403	21	(	(	PUNCT
bjmsr-315	403	22	1988	1988	NUM
bjmsr-315	403	23	)	)	PUNCT
bjmsr-315	403	24	.	.	PUNCT
bjmsr-315	404	1	absolute	absolute	ADJ
bjmsr-315	404	2	convergence	convergence	NOUN
bjmsr-315	404	3	of	of	ADP
bjmsr-315	404	4	this	this	DET
bjmsr-315	404	5	algorithm	algorithm	NOUN
bjmsr-315	404	6	has	have	AUX
bjmsr-315	404	7	not	not	PART
bjmsr-315	404	8	been	be	AUX
bjmsr-315	404	9	proved	prove	VERB
bjmsr-315	404	10	,	,	PUNCT
bjmsr-315	404	11	but	but	CCONJ
bjmsr-315	404	12	the	the	DET
bjmsr-315	404	13	non	non	ADJ
bjmsr-315	404	14	-	-	ADJ
bjmsr-315	404	15	convergent	convergent	ADJ
bjmsr-315	404	16	experiment	experiment	NOUN
bjmsr-315	404	17	has	have	AUX
bjmsr-315	404	18	not	not	PART
bjmsr-315	404	19	been	be	AUX
bjmsr-315	404	20	reported	report	VERB
bjmsr-315	404	21	.	.	PUNCT
bjmsr-315	405	1	soliman	soliman	NOUN
bjmsr-315	405	2	and	and	CCONJ
bjmsr-315	405	3	christensen	christensen	PROPN
bjmsr-315	405	4	and	and	CCONJ
bjmsr-315	405	5	rouhi	rouhi	PROPN
bjmsr-315	405	6	(	(	PUNCT
bjmsr-315	405	7	1988	1988	NUM
bjmsr-315	405	8	)	)	PUNCT
bjmsr-315	405	9	used	use	VERB
bjmsr-315	405	10	left	left	ADJ
bjmsr-315	405	11	pseudoinverse	pseudoinverse	NOUN
bjmsr-315	405	12	(	(	PUNCT
bjmsr-315	405	13	see	see	VERB
bjmsr-315	405	14	,	,	PUNCT
bjmsr-315	405	15	dhrymes	dhryme	NOUN
bjmsr-315	405	16	(	(	PUNCT
bjmsr-315	405	17	1978	1978	NUM
bjmsr-315	405	18	)	)	PUNCT
bjmsr-315	405	19	for	for	ADP
bjmsr-315	405	20	a	a	DET
bjmsr-315	405	21	description	description	NOUN
bjmsr-315	405	22	of	of	ADP
bjmsr-315	405	23	this	this	DET
bjmsr-315	405	24	inverse	inverse	NOUN
bjmsr-315	405	25	)	)	PUNCT
bjmsr-315	405	26	to	to	PART
bjmsr-315	405	27	solve	solve	VERB
bjmsr-315	405	28	the	the	DET
bjmsr-315	405	29	general	general	ADJ
bjmsr-315	405	30	linear	linear	PROPN
bjmsr-315	405	31	l1	l1	PROPN
bjmsr-315	405	32	norm	norm	PROPN
bjmsr-315	405	33	regression	regression	PROPN
bjmsr-315	405	34	.	.	PUNCT
bjmsr-315	406	1	according	accord	VERB
bjmsr-315	406	2	to	to	ADP
bjmsr-315	406	3	this	this	DET
bjmsr-315	406	4	procedure	procedure	NOUN
bjmsr-315	406	5	,	,	PUNCT
bjmsr-315	406	6	one	one	PRON
bjmsr-315	406	7	should	should	AUX
bjmsr-315	406	8	calculate	calculate	VERB
bjmsr-315	406	9	the	the	DET
bjmsr-315	406	10	least	least	ADJ
bjmsr-315	406	11	squares	square	NOUN
bjmsr-315	406	12	solution	solution	NOUN
bjmsr-315	406	13	using	use	VERB
bjmsr-315	406	14	the	the	DET
bjmsr-315	406	15	left	left	ADJ
bjmsr-315	406	16	pseudoinverse	pseudoinverse	NOUN
bjmsr-315	406	17	or	or	CCONJ
bjmsr-315	406	18	least	least	ADJ
bjmsr-315	406	19	squares	square	NOUN
bjmsr-315	406	20	approximation	approximation	NOUN
bjmsr-315	406	21	as	as	ADP
bjmsr-315	406	22	ß^=(xtx)-1xty	ß^=(xtx)-1xty	PROPN
bjmsr-315	406	23	.	.	PROPN
bjmsr-315	406	24	calculate	calculate	VERB
bjmsr-315	406	25	the	the	DET
bjmsr-315	406	26	residuals	residual	NOUN
bjmsr-315	406	27	as	as	ADP
bjmsr-315	406	28	u=	u=	NOUN
bjmsr-315	406	29	│	│	NUM
bjmsr-315	406	30	yxß^	yxß^	NOUN
bjmsr-315	406	31	│	│	NOUN
bjmsr-315	406	32	;	;	PUNCT
bjmsr-315	406	33	where	where	SCONJ
bjmsr-315	406	34	u	u	NOUN
bjmsr-315	406	35	is	be	AUX
bjmsr-315	406	36	nx1	nx1	ADP
bjmsr-315	406	37	column	column	NOUN
bjmsr-315	406	38	vector	vector	PROPN
bjmsr-315	406	39	.	.	PUNCT
bjmsr-315	407	1	select	select	VERB
bjmsr-315	407	2	the	the	DET
bjmsr-315	407	3	m	m	PROPN
bjmsr-315	407	4	observations	observation	NOUN
bjmsr-315	407	5	with	with	ADP
bjmsr-315	407	6	the	the	DET
bjmsr-315	407	7	smallest	small	ADJ
bjmsr-315	407	8	absolute	absolute	ADJ
bjmsr-315	407	9	values	value	NOUN
bjmsr-315	407	10	of	of	ADP
bjmsr-315	407	11	the	the	DET
bjmsr-315	407	12	residuals	residual	NOUN
bjmsr-315	407	13	and	and	CCONJ
bjmsr-315	407	14	partition	partition	VERB
bjmsr-315	407	15	the	the	DET
bjmsr-315	407	16	matrices	matrix	NOUN
bjmsr-315	407	17	as	as	ADP
bjmsr-315	407	18	the	the	DET
bjmsr-315	407	19	selected	select	VERB
bjmsr-315	407	20	observations	observation	NOUN
bjmsr-315	407	21	locate	locate	VERB
bjmsr-315	407	22	on	on	ADP
bjmsr-315	407	23	the	the	DET
bjmsr-315	407	24	top	top	NOUN
bjmsr-315	407	25	,	,	PUNCT
bjmsr-315	407	26	,	,	PUNCT
bjmsr-315	407	27	,	,	PUNCT
bjmsr-315	407	28	.	.	PUNCT
bjmsr-315	408	1			PROPN
bjmsr-315	408	2			PROPN
bjmsr-315	408	3			INTJ
bjmsr-315	408	4			PROPN
bjmsr-315	408	5			ADJ
bjmsr-315	408	6			NOUN
bjmsr-315	408	7			PROPN
bjmsr-315	408	8			NOUN
bjmsr-315	408	9			NOUN
bjmsr-315	408	10			NOUN
bjmsr-315	409	1			NUM
bjmsr-315	410	1			NUM
bjmsr-315	411	1			ADJ
bjmsr-315	411	2			NOUN
bjmsr-315	411	3			PROPN
bjmsr-315	411	4			NOUN
bjmsr-315	411	5			NOUN
bjmsr-315	411	6			NOUN
bjmsr-315	411	7			NOUN
bjmsr-315	411	8			NOUN
bjmsr-315	411	9			NOUN
bjmsr-315	411	10			NOUN
bjmsr-315	411	11			NOUN
bjmsr-315	411	12			PROPN
bjmsr-315	411	13			VERB
bjmsr-315	411	14			PROPN
bjmsr-315	411	15			PROPN
bjmsr-315	411	16	u	u	NOUN
bjmsr-315	411	17	x	x	X
bjmsr-315	411	18	y	y	PROPN
bjmsr-315	411	19	u	u	NOUN
bjmsr-315	411	20	x	x	X
bjmsr-315	411	21	y	y	VERB
bjmsr-315	411	22	u	u	X
bjmsr-315	411	23	yx	yx	PROPN
bjmsr-315	411	24	(	(	PUNCT
bjmsr-315	411	25	33	33	NUM
bjmsr-315	411	26	)	)	PUNCT
bjmsr-315	411	27	solve	solve	VERB
bjmsr-315	411	28	y^=x^ß^	y^=x^ß^	NOUN
bjmsr-315	411	29	for	for	ADP
bjmsr-315	411	30	the	the	DET
bjmsr-315	411	31	top	top	ADJ
bjmsr-315	411	32	partitions	partition	NOUN
bjmsr-315	411	33	as	as	ADP
bjmsr-315	411	34	ß^=x^-1y	ß^=x^-1y	NOUN
bjmsr-315	411	35	.	.	PUNCT
bjmsr-315	412	1	although	although	SCONJ
bjmsr-315	412	2	this	this	DET
bjmsr-315	412	3	procedure	procedure	NOUN
bjmsr-315	412	4	is	be	AUX
bjmsr-315	412	5	operationally	operationally	ADV
bjmsr-315	412	6	simple	simple	ADJ
bjmsr-315	412	7	,	,	PUNCT
bjmsr-315	412	8	its	its	PRON
bjmsr-315	412	9	solution	solution	NOUN
bjmsr-315	412	10	is	be	AUX
bjmsr-315	412	11	not	not	PART
bjmsr-315	412	12	the	the	DET
bjmsr-315	412	13	same	same	ADJ
bjmsr-315	412	14	as	as	ADP
bjmsr-315	412	15	other	other	ADJ
bjmsr-315	412	16	exact	exact	ADJ
bjmsr-315	412	17	methods	method	NOUN
bjmsr-315	412	18	,	,	PUNCT
bjmsr-315	412	19	and	and	CCONJ
bjmsr-315	412	20	no	no	DET
bjmsr-315	412	21	proof	proof	NOUN
bjmsr-315	412	22	is	be	AUX
bjmsr-315	412	23	presented	present	VERB
bjmsr-315	412	24	to	to	PART
bjmsr-315	412	25	show	show	VERB
bjmsr-315	412	26	that	that	SCONJ
bjmsr-315	412	27	the	the	DET
bjmsr-315	412	28	solution	solution	NOUN
bjmsr-315	412	29	is	be	AUX
bjmsr-315	412	30	in	in	ADP
bjmsr-315	412	31	the	the	DET
bjmsr-315	412	32	neighborhood	neighborhood	NOUN
bjmsr-315	412	33	of	of	ADP
bjmsr-315	412	34	the	the	DET
bjmsr-315	412	35	exact	exact	ADJ
bjmsr-315	412	36	solution	solution	NOUN
bjmsr-315	412	37	of	of	ADP
bjmsr-315	412	38	the	the	DET
bjmsr-315	412	39	l1	l1	PROPN
bjmsr-315	412	40	norm	norm	PROPN
bjmsr-315	412	41	minimization	minimization	PROPN
bjmsr-315	412	42	problem	problem	NOUN
bjmsr-315	412	43	.	.	PUNCT
bjmsr-315	413	1	application	application	NOUN
bjmsr-315	413	2	of	of	ADP
bjmsr-315	413	3	median	median	ADJ
bjmsr-315	413	4	polish	polish	NOUN
bjmsr-315	413	5	(	(	PUNCT
bjmsr-315	413	6	see	see	VERB
bjmsr-315	413	7	,	,	PUNCT
bjmsr-315	413	8	tukey	tukey	NOUN
bjmsr-315	413	9	(	(	PUNCT
bjmsr-315	413	10	1977	1977	NUM
bjmsr-315	413	11	)	)	PUNCT
bjmsr-315	413	12	)	)	PUNCT
bjmsr-315	413	13	and	and	CCONJ
bjmsr-315	413	14	e	e	NOUN
bjmsr-315	413	15	-	-	ADJ
bjmsr-315	413	16	median	median	ADJ
bjmsr-315	413	17	polish	polish	NOUN
bjmsr-315	413	18	to	to	ADP
bjmsr-315	413	19	l1	l1	PROPN
bjmsr-315	413	20	norm	norm	NOUN
bjmsr-315	413	21	estimation	estimation	NOUN
bjmsr-315	413	22	are	be	AUX
bjmsr-315	413	23	discussed	discuss	VERB
bjmsr-315	413	24	and	and	CCONJ
bjmsr-315	413	25	developed	develop	VERB
bjmsr-315	413	26	by	by	ADP
bjmsr-315	413	27	bloomfield	bloomfield	PROPN
bjmsr-315	413	28	and	and	CCONJ
bjmsr-315	413	29	steiger	steiger	PROPN
bjmsr-315	413	30	(	(	PUNCT
bjmsr-315	413	31	1983	1983	NUM
bjmsr-315	413	32	)	)	PUNCT
bjmsr-315	413	33	,	,	PUNCT
bjmsr-315	413	34	kemperman	kemperman	NOUN
bjmsr-315	413	35	(	(	PUNCT
bjmsr-315	413	36	1984	1984	NUM
bjmsr-315	413	37	)	)	PUNCT
bjmsr-315	413	38	,	,	PUNCT
bjmsr-315	413	39	sposito	sposito	X
bjmsr-315	413	40	(	(	PUNCT
bjmsr-315	413	41	1987a	1987a	NUM
bjmsr-315	413	42	)	)	PUNCT
bjmsr-315	413	43	,	,	PUNCT
bjmsr-315	413	44	bradu	bradu	NOUN
bjmsr-315	413	45	(	(	PUNCT
bjmsr-315	413	46	1987a	1987a	NUM
bjmsr-315	413	47	,	,	PUNCT
bjmsr-315	413	48	b	b	NOUN
bjmsr-315	413	49	)	)	PUNCT
bjmsr-315	413	50	.	.	PUNCT
bjmsr-315	414	1	copyright	copyright	NOUN
bjmsr-315	414	2	©	©	PROPN
bjmsr-315	414	3	cc	cc	PROPN
bjmsr-315	414	4	-	-	PUNCT
bjmsr-315	414	5	by	by	ADP
bjmsr-315	414	6	-	-	PUNCT
bjmsr-315	414	7	nc	nc	PROPN
bjmsr-315	414	8	2019	2019	NUM
bjmsr-315	414	9	,	,	PUNCT
bjmsr-315	414	10	bjmsr	bjmsr	PROPN
bjmsr-315	414	11	www.cribfb.com/journal/index.php/bjmsr	www.cribfb.com/journal/index.php/bjmsr	X
bjmsr-315	415	1	bangladesh	bangladesh	PROPN
bjmsr-315	415	2	journal	journal	PROPN
bjmsr-315	415	3	of	of	ADP
bjmsr-315	415	4	multidisciplinary	multidisciplinary	ADJ
bjmsr-315	415	5	scientific	scientific	ADJ
bjmsr-315	415	6	research	research	NOUN
bjmsr-315	415	7	vol	vol	NOUN
bjmsr-315	415	8	.	.	PROPN
bjmsr-315	415	9	1	1	NUM
bjmsr-315	415	10	,	,	PUNCT
bjmsr-315	415	11	no	no	INTJ
bjmsr-315	415	12	.	.	NOUN
bjmsr-315	415	13	1	1	NUM
bjmsr-315	415	14	;	;	PUNCT
bjmsr-315	415	15	2019	2019	NUM
bjmsr-315	415	16	61	61	NUM
bjmsr-315	415	17	application	application	NOUN
bjmsr-315	415	18	of	of	ADP
bjmsr-315	415	19	karmarkar	karmarkar	NOUN
bjmsr-315	415	20	's	's	PART
bjmsr-315	415	21	algorithm	algorithm	NOUN
bjmsr-315	415	22	for	for	ADP
bjmsr-315	415	23	linear	linear	NOUN
bjmsr-315	415	24	programming	programming	NOUN
bjmsr-315	415	25	and	and	CCONJ
bjmsr-315	415	26	its	its	PRON
bjmsr-315	415	27	relation	relation	NOUN
bjmsr-315	415	28	to	to	ADP
bjmsr-315	415	29	l1	l1	PROPN
bjmsr-315	415	30	norm	norm	NOUN
bjmsr-315	415	31	is	be	AUX
bjmsr-315	415	32	given	give	VERB
bjmsr-315	415	33	by	by	ADP
bjmsr-315	415	34	sherali	sherali	ADJ
bjmsr-315	415	35	and	and	CCONJ
bjmsr-315	415	36	skarpness	skarpness	NOUN
bjmsr-315	415	37	and	and	CCONJ
bjmsr-315	415	38	kim	kim	PROPN
bjmsr-315	415	39	(	(	PUNCT
bjmsr-315	415	40	1987	1987	NUM
bjmsr-315	415	41	)	)	PUNCT
bjmsr-315	415	42	.	.	PUNCT
bjmsr-315	416	1	for	for	ADP
bjmsr-315	416	2	using	use	VERB
bjmsr-315	416	3	homotopy	homotopy	NOUN
bjmsr-315	416	4	method	method	NOUN
bjmsr-315	416	5	in	in	ADP
bjmsr-315	416	6	l1	l1	PROPN
bjmsr-315	416	7	norm	norm	NOUN
bjmsr-315	416	8	,	,	PUNCT
bjmsr-315	416	9	see	see	VERB
bjmsr-315	416	10	garcia	garcia	PROPN
bjmsr-315	416	11	and	and	CCONJ
bjmsr-315	416	12	gould	gould	PROPN
bjmsr-315	416	13	(	(	PUNCT
bjmsr-315	416	14	1983	1983	NUM
bjmsr-315	416	15	)	)	PUNCT
bjmsr-315	416	16	,	,	PUNCT
bjmsr-315	416	17	schellhorn	schellhorn	ADJ
bjmsr-315	416	18	(	(	PUNCT
bjmsr-315	416	19	1987	1987	NUM
bjmsr-315	416	20	)	)	PUNCT
bjmsr-315	416	21	.	.	PUNCT
bjmsr-315	417	1	an	an	DET
bjmsr-315	417	2	algorithm	algorithm	NOUN
bjmsr-315	417	3	for	for	ADP
bjmsr-315	417	4	linear	linear	PROPN
bjmsr-315	417	5	l1	l1	PROPN
bjmsr-315	417	6	norm	norm	NOUN
bjmsr-315	417	7	approximation	approximation	NOUN
bjmsr-315	417	8	for	for	ADP
bjmsr-315	417	9	the	the	DET
bjmsr-315	417	10	continuous	continuous	ADJ
bjmsr-315	417	11	function	function	NOUN
bjmsr-315	417	12	is	be	AUX
bjmsr-315	417	13	given	give	VERB
bjmsr-315	417	14	by	by	ADP
bjmsr-315	417	15	watson	watson	PROPN
bjmsr-315	417	16	(	(	PUNCT
bjmsr-315	417	17	1981	1981	NUM
bjmsr-315	417	18	)	)	PUNCT
bjmsr-315	417	19	,	,	PUNCT
bjmsr-315	417	20	(	(	PUNCT
bjmsr-315	417	21	see	see	VERB
bjmsr-315	417	22	also	also	ADV
bjmsr-315	417	23	,	,	PUNCT
bjmsr-315	417	24	baboolal	baboolal	NOUN
bjmsr-315	417	25	and	and	CCONJ
bjmsr-315	417	26	watson	watson	PROPN
bjmsr-315	417	27	(	(	PUNCT
bjmsr-315	417	28	1981	1981	NUM
bjmsr-315	417	29	)	)	PUNCT
bjmsr-315	417	30	)	)	PUNCT
bjmsr-315	417	31	.	.	PUNCT
bjmsr-315	418	1	references	reference	NOUN
bjmsr-315	418	2	n.n	n.n	PROPN
bjmsr-315	418	3	.	.	PROPN
bjmsr-315	418	4	abdelmalek	abdelmalek	PROPN
bjmsr-315	418	5	(	(	PUNCT
bjmsr-315	418	6	1971	1971	NUM
bjmsr-315	418	7	)	)	PUNCT
bjmsr-315	418	8	linear	linear	ADJ
bjmsr-315	418	9	approximation	approximation	NOUN
bjmsr-315	418	10	for	for	ADP
bjmsr-315	418	11	a	a	DET
bjmsr-315	418	12	discrete	discrete	ADJ
bjmsr-315	418	13	point	point	NOUN
bjmsr-315	418	14	set	set	VERB
bjmsr-315	418	15	and	and	CCONJ
bjmsr-315	418	16	l1	l1	PROPN
bjmsr-315	418	17	solutions	solution	NOUN
bjmsr-315	418	18	of	of	ADP
bjmsr-315	418	19	overdetermined	overdetermine	VERB
bjmsr-315	418	20	linear	linear	PROPN
bjmsr-315	418	21	equations	equation	NOUN
bjmsr-315	418	22	.	.	PUNCT
bjmsr-315	419	1	j.	j.	PROPN
bjmsr-315	419	2	acm	acm	PROPN
bjmsr-315	419	3	,	,	PUNCT
bjmsr-315	419	4	18	18	NUM
bjmsr-315	419	5	,	,	PUNCT
bjmsr-315	419	6	41	41	NUM
bjmsr-315	419	7	-	-	SYM
bjmsr-315	419	8	47	47	NUM
bjmsr-315	419	9	.	.	PUNCT
bjmsr-315	420	1	n.n	n.n	PROPN
bjmsr-315	420	2	.	.	PROPN
bjmsr-315	420	3	abdelmalek	abdelmalek	PROPN
bjmsr-315	420	4	(	(	PUNCT
bjmsr-315	420	5	1974	1974	NUM
bjmsr-315	420	6	)	)	PUNCT
bjmsr-315	420	7	on	on	ADP
bjmsr-315	420	8	the	the	DET
bjmsr-315	420	9	discrete	discrete	ADJ
bjmsr-315	420	10	linear	linear	PROPN
bjmsr-315	420	11	l1	l1	PROPN
bjmsr-315	420	12	approximation	approximation	NOUN
bjmsr-315	420	13	and	and	CCONJ
bjmsr-315	420	14	l1	l1	PROPN
bjmsr-315	420	15	solutions	solution	NOUN
bjmsr-315	420	16	of	of	ADP
bjmsr-315	420	17	overdetermined	overdetermine	VERB
bjmsr-315	420	18	linear	linear	PROPN
bjmsr-315	420	19	equations	equation	NOUN
bjmsr-315	420	20	.	.	PUNCT
bjmsr-315	421	1	j.	j.	PROPN
bjmsr-315	421	2	of	of	ADP
bjmsr-315	421	3	approx	approx	PROPN
bjmsr-315	421	4	.	.	PUNCT
bjmsr-315	422	1	theory	theory	NOUN
bjmsr-315	422	2	,	,	PUNCT
bjmsr-315	422	3	11	11	NUM
bjmsr-315	422	4	,	,	PUNCT
bjmsr-315	422	5	38	38	NUM
bjmsr-315	422	6	-	-	SYM
bjmsr-315	422	7	53	53	NUM
bjmsr-315	422	8	.	.	PUNCT
bjmsr-315	423	1	n.n	n.n	PROPN
bjmsr-315	423	2	.	.	PROPN
bjmsr-315	423	3	abdelmalek	abdelmalek	PROPN
bjmsr-315	423	4	(	(	PUNCT
bjmsr-315	423	5	1975a	1975a	NUM
bjmsr-315	423	6	)	)	PUNCT
bjmsr-315	423	7	an	an	DET
bjmsr-315	423	8	efficient	efficient	ADJ
bjmsr-315	423	9	method	method	NOUN
bjmsr-315	423	10	for	for	ADP
bjmsr-315	423	11	the	the	DET
bjmsr-315	423	12	discrete	discrete	ADJ
bjmsr-315	423	13	l1	l1	PROPN
bjmsr-315	423	14	approximation	approximation	NOUN
bjmsr-315	423	15	problem	problem	NOUN
bjmsr-315	423	16	.	.	PUNCT
bjmsr-315	424	1	math	math	NOUN
bjmsr-315	424	2	.	.	PUNCT
bjmsr-315	425	1	comput	comput	NOUN
bjmsr-315	425	2	.	.	PUNCT
bjmsr-315	425	3	,	,	PUNCT
bjmsr-315	425	4	29	29	NUM
bjmsr-315	425	5	,	,	PUNCT
bjmsr-315	425	6	844	844	NUM
bjmsr-315	425	7	-	-	SYM
bjmsr-315	425	8	850	850	NUM
bjmsr-315	425	9	.	.	PUNCT
bjmsr-315	426	1	n.n	n.n	PROPN
bjmsr-315	426	2	.	.	PROPN
bjmsr-315	426	3	abdelmalek	abdelmalek	PROPN
bjmsr-315	426	4	(	(	PUNCT
bjmsr-315	426	5	1980a	1980a	NUM
bjmsr-315	426	6	)	)	PUNCT
bjmsr-315	426	7	l1	l1	PROPN
bjmsr-315	426	8	solution	solution	NOUN
bjmsr-315	426	9	of	of	ADP
bjmsr-315	426	10	overdetermined	overdetermine	VERB
bjmsr-315	426	11	systems	system	NOUN
bjmsr-315	426	12	of	of	ADP
bjmsr-315	426	13	linear	linear	PROPN
bjmsr-315	426	14	equations	equation	NOUN
bjmsr-315	426	15	.	.	PUNCT
bjmsr-315	427	1	acm	acm	PROPN
bjmsr-315	427	2	trans	trans	PROPN
bjmsr-315	427	3	.	.	PROPN
bjmsr-315	428	1	math	math	PROPN
bjmsr-315	428	2	.	.	PUNCT
bjmsr-315	429	1	soft	soft	ADJ
bjmsr-315	429	2	.	.	PUNCT
bjmsr-315	430	1	,	,	PUNCT
bjmsr-315	430	2	6	6	NUM
bjmsr-315	430	3	,	,	PUNCT
bjmsr-315	430	4	220	220	NUM
bjmsr-315	430	5	-	-	SYM
bjmsr-315	430	6	227	227	NUM
bjmsr-315	430	7	.	.	PUNCT
bjmsr-315	431	1	n.n	n.n	PROPN
bjmsr-315	431	2	.	.	PROPN
bjmsr-315	431	3	abdelmalek	abdelmalek	PROPN
bjmsr-315	431	4	(	(	PUNCT
bjmsr-315	431	5	1980b	1980b	NUM
bjmsr-315	431	6	)	)	PUNCT
bjmsr-315	431	7	a	a	DET
bjmsr-315	431	8	fortran	fortran	NOUN
bjmsr-315	431	9	subroutine	subroutine	NOUN
bjmsr-315	431	10	for	for	ADP
bjmsr-315	431	11	the	the	DET
bjmsr-315	431	12	l1	l1	PROPN
bjmsr-315	431	13	solution	solution	NOUN
bjmsr-315	431	14	of	of	ADP
bjmsr-315	431	15	overdetermined	overdetermine	VERB
bjmsr-315	431	16	systems	system	NOUN
bjmsr-315	431	17	of	of	ADP
bjmsr-315	431	18	linear	linear	PROPN
bjmsr-315	431	19	equations	equation	NOUN
bjmsr-315	431	20	.	.	PUNCT
bjmsr-315	432	1	acm	acm	PROPN
bjmsr-315	432	2	trans	trans	PROPN
bjmsr-315	432	3	.	.	PROPN
bjmsr-315	433	1	math	math	PROPN
bjmsr-315	433	2	.	.	PUNCT
bjmsr-315	434	1	soft	soft	ADJ
bjmsr-315	434	2	.	.	PUNCT
bjmsr-315	435	1	,	,	PUNCT
bjmsr-315	435	2	6	6	NUM
bjmsr-315	435	3	,	,	PUNCT
bjmsr-315	435	4	228	228	NUM
bjmsr-315	435	5	-	-	SYM
bjmsr-315	435	6	30	30	NUM
bjmsr-315	435	7	.	.	PUNCT
bjmsr-315	436	1	d.h	d.h	PROPN
bjmsr-315	436	2	.	.	PROPN
bjmsr-315	436	3	anderson	anderson	PROPN
bjmsr-315	436	4	,	,	PUNCT
bjmsr-315	436	5	m.r	m.r	PROPN
bjmsr-315	436	6	.	.	PROPN
bjmsr-315	436	7	osborne	osborne	PROPN
bjmsr-315	436	8	(	(	PUNCT
bjmsr-315	436	9	1976	1976	NUM
bjmsr-315	436	10	)	)	PUNCT
bjmsr-315	436	11	discrete	discrete	ADJ
bjmsr-315	436	12	linear	linear	ADJ
bjmsr-315	436	13	approximation	approximation	NOUN
bjmsr-315	436	14	problems	problem	NOUN
bjmsr-315	436	15	in	in	ADP
bjmsr-315	436	16	polyhedral	polyhedral	ADJ
bjmsr-315	436	17	norms	norm	NOUN
bjmsr-315	436	18	.	.	PUNCT
bjmsr-315	437	1	numer	numer	PROPN
bjmsr-315	437	2	.	.	PUNCT
bjmsr-315	437	3	math	math	NOUN
bjmsr-315	437	4	.	.	PUNCT
bjmsr-315	438	1	26	26	NUM
bjmsr-315	438	2	,	,	PUNCT
bjmsr-315	438	3	179189	179189	NUM
bjmsr-315	438	4	.	.	PUNCT
bjmsr-315	439	1	r.d	r.d	PROPN
bjmsr-315	439	2	.	.	PROPN
bjmsr-315	440	1	armstrong	armstrong	PROPN
bjmsr-315	440	2	,	,	PUNCT
bjmsr-315	440	3	e.l	e.l	PROPN
bjmsr-315	440	4	.	.	PROPN
bjmsr-315	440	5	frome	frome	PROPN
bjmsr-315	440	6	,	,	PUNCT
bjmsr-315	440	7	d.s	d.s	PROPN
bjmsr-315	440	8	.	.	PROPN
bjmsr-315	440	9	kung	kung	PROPN
bjmsr-315	440	10	(	(	PUNCT
bjmsr-315	440	11	1979	1979	NUM
bjmsr-315	440	12	)	)	PUNCT
bjmsr-315	440	13	a	a	DET
bjmsr-315	440	14	revised	revise	VERB
bjmsr-315	440	15	simplex	simplex	NOUN
bjmsr-315	440	16	algorithm	algorithm	NOUN
bjmsr-315	440	17	for	for	ADP
bjmsr-315	440	18	the	the	DET
bjmsr-315	440	19	absolute	absolute	ADJ
bjmsr-315	440	20	deviation	deviation	NOUN
bjmsr-315	440	21	curve	curve	NOUN
bjmsr-315	440	22	fitting	fitting	ADJ
bjmsr-315	440	23	problem	problem	NOUN
bjmsr-315	440	24	.	.	PUNCT
bjmsr-315	441	1	commun	commun	PROPN
bjmsr-315	441	2	.	.	PUNCT
bjmsr-315	442	1	stat	stat	PROPN
bjmsr-315	442	2	.	.	PUNCT
bjmsr-315	443	1	b8	b8	PROPN
bjmsr-315	443	2	,	,	PUNCT
bjmsr-315	443	3	175	175	NUM
bjmsr-315	443	4	-	-	SYM
bjmsr-315	443	5	190	190	NUM
bjmsr-315	443	6	.	.	PUNCT
bjmsr-315	444	1	r.d	r.d	PROPN
bjmsr-315	444	2	.	.	PROPN
bjmsr-315	444	3	armstrong	armstrong	PROPN
bjmsr-315	444	4	,	,	PUNCT
bjmsr-315	444	5	j.	j.	PROPN
bjmsr-315	444	6	godfrey	godfrey	PROPN
bjmsr-315	444	7	(	(	PUNCT
bjmsr-315	444	8	1979	1979	NUM
bjmsr-315	444	9	)	)	PUNCT
bjmsr-315	444	10	two	two	NUM
bjmsr-315	444	11	linear	linear	ADJ
bjmsr-315	444	12	programming	programming	NOUN
bjmsr-315	444	13	algorithms	algorithm	NOUN
bjmsr-315	444	14	for	for	ADP
bjmsr-315	444	15	the	the	DET
bjmsr-315	444	16	discrete	discrete	ADJ
bjmsr-315	444	17	l1	l1	PROPN
bjmsr-315	444	18	problem	problem	NOUN
bjmsr-315	444	19	.	.	PUNCT
bjmsr-315	445	1	math	math	NOUN
bjmsr-315	445	2	.	.	PUNCT
bjmsr-315	446	1	comput	comput	NOUN
bjmsr-315	446	2	.	.	PUNCT
bjmsr-315	446	3	,	,	PUNCT
bjmsr-315	446	4	33	33	NUM
bjmsr-315	446	5	,	,	PUNCT
bjmsr-315	446	6	289300	289300	NUM
bjmsr-315	446	7	.	.	PUNCT
bjmsr-315	447	1	r.d	r.d	PROPN
bjmsr-315	447	2	.	.	PROPN
bjmsr-315	448	1	armstrong	armstrong	PROPN
bjmsr-315	448	2	,	,	PUNCT
bjmsr-315	448	3	d.s	d.s	PROPN
bjmsr-315	448	4	.	.	PROPN
bjmsr-315	448	5	kung	kung	PROPN
bjmsr-315	448	6	(	(	PUNCT
bjmsr-315	448	7	1978	1978	NUM
bjmsr-315	448	8	)	)	PUNCT
bjmsr-315	448	9	as132	as132	PROPN
bjmsr-315	448	10	:	:	PUNCT
bjmsr-315	448	11	least	least	ADJ
bjmsr-315	448	12	absolute	absolute	ADJ
bjmsr-315	448	13	value	value	NOUN
bjmsr-315	448	14	estimates	estimate	NOUN
bjmsr-315	448	15	for	for	ADP
bjmsr-315	448	16	a	a	DET
bjmsr-315	448	17	simple	simple	ADJ
bjmsr-315	448	18	linear	linear	NOUN
bjmsr-315	448	19	regression	regression	NOUN
bjmsr-315	448	20	problem	problem	NOUN
bjmsr-315	448	21	.	.	PUNCT
bjmsr-315	449	1	appl	appl	PROPN
bjmsr-315	449	2	.	.	PUNCT
bjmsr-315	450	1	stat	stat	PROPN
bjmsr-315	450	2	.	.	PUNCT
bjmsr-315	450	3	,	,	PUNCT
bjmsr-315	450	4	27	27	NUM
bjmsr-315	450	5	,	,	PUNCT
bjmsr-315	450	6	363	363	NUM
bjmsr-315	450	7	-	-	SYM
bjmsr-315	450	8	366	366	NUM
bjmsr-315	450	9	.	.	PUNCT
bjmsr-315	451	1	r.d	r.d	PROPN
bjmsr-315	451	2	.	.	PROPN
bjmsr-315	452	1	armstrong	armstrong	PROPN
bjmsr-315	452	2	,	,	PUNCT
bjmsr-315	452	3	d.s	d.s	PROPN
bjmsr-315	452	4	.	.	PROPN
bjmsr-315	452	5	kung	kung	PROPN
bjmsr-315	452	6	(	(	PUNCT
bjmsr-315	452	7	1982b	1982b	NUM
bjmsr-315	452	8	)	)	PUNCT
bjmsr-315	452	9	a	a	DET
bjmsr-315	452	10	dual	dual	ADJ
bjmsr-315	452	11	algorithm	algorithm	NOUN
bjmsr-315	452	12	to	to	PART
bjmsr-315	452	13	solve	solve	VERB
bjmsr-315	452	14	linear	linear	ADV
bjmsr-315	452	15	least	least	ADV
bjmsr-315	452	16	absolute	absolute	ADJ
bjmsr-315	452	17	value	value	NOUN
bjmsr-315	452	18	problems	problem	NOUN
bjmsr-315	452	19	.	.	PUNCT
bjmsr-315	453	1	j.	j.	PROPN
bjmsr-315	453	2	oper	oper	PROPN
bjmsr-315	453	3	.	.	PUNCT
bjmsr-315	454	1	res	res	PROPN
bjmsr-315	454	2	.	.	PUNCT
bjmsr-315	454	3	soc	soc	PROPN
bjmsr-315	454	4	.	.	PUNCT
bjmsr-315	455	1	,	,	PUNCT
bjmsr-315	455	2	33	33	NUM
bjmsr-315	455	3	,	,	PUNCT
bjmsr-315	455	4	931	931	NUM
bjmsr-315	455	5	-	-	SYM
bjmsr-315	455	6	936	936	NUM
bjmsr-315	455	7	.	.	PUNCT
bjmsr-315	456	1	t.s	t.s	PROPN
bjmsr-315	456	2	.	.	PROPN
bjmsr-315	456	3	arthanari	arthanari	PROPN
bjmsr-315	456	4	,	,	PUNCT
bjmsr-315	456	5	y.	y.	PROPN
bjmsr-315	456	6	dodge	dodge	PROPN
bjmsr-315	456	7	(	(	PUNCT
bjmsr-315	456	8	1981	1981	NUM
bjmsr-315	456	9	)	)	PUNCT
bjmsr-315	456	10	mathematical	mathematical	ADJ
bjmsr-315	456	11	programming	programming	NOUN
bjmsr-315	456	12	in	in	ADP
bjmsr-315	456	13	statistics	statistic	NOUN
bjmsr-315	456	14	.	.	PUNCT
bjmsr-315	457	1	john	john	PROPN
bjmsr-315	457	2	wiley	wiley	PROPN
bjmsr-315	457	3	,	,	PUNCT
bjmsr-315	457	4	interscience	interscience	NOUN
bjmsr-315	457	5	division	division	NOUN
bjmsr-315	457	6	,	,	PUNCT
bjmsr-315	457	7	new	new	PROPN
bjmsr-315	457	8	york	york	PROPN
bjmsr-315	457	9	.	.	PUNCT
bjmsr-315	458	1	s.	s.	PROPN
bjmsr-315	458	2	baboolal	baboolal	PROPN
bjmsr-315	458	3	,	,	PUNCT
bjmsr-315	458	4	g.a	g.a	PROPN
bjmsr-315	458	5	.	.	PROPN
bjmsr-315	458	6	watson	watson	PROPN
bjmsr-315	458	7	(	(	PUNCT
bjmsr-315	458	8	1981	1981	NUM
bjmsr-315	458	9	)	)	PUNCT
bjmsr-315	458	10	computational	computational	ADJ
bjmsr-315	458	11	experience	experience	NOUN
bjmsr-315	458	12	with	with	ADP
bjmsr-315	458	13	an	an	DET
bjmsr-315	458	14	algorithm	algorithm	NOUN
bjmsr-315	458	15	for	for	ADP
bjmsr-315	458	16	discrete	discrete	ADJ
bjmsr-315	458	17	l1	l1	PROPN
bjmsr-315	458	18	approximation	approximation	NOUN
bjmsr-315	458	19	.	.	PUNCT
bjmsr-315	459	1	computing	computing	NOUN
bjmsr-315	459	2	,	,	PUNCT
bjmsr-315	459	3	27	27	NUM
bjmsr-315	459	4	,	,	PUNCT
bjmsr-315	459	5	245	245	NUM
bjmsr-315	459	6	-	-	SYM
bjmsr-315	459	7	252	252	NUM
bjmsr-315	459	8	.	.	PUNCT
bjmsr-315	460	1	s.c	s.c	PROPN
bjmsr-315	460	2	.	.	PROPN
bjmsr-315	460	3	banks	bank	NOUN
bjmsr-315	460	4	,	,	PUNCT
bjmsr-315	460	5	h.l	h.l	PROPN
bjmsr-315	460	6	.	.	PROPN
bjmsr-315	460	7	taylor	taylor	PROPN
bjmsr-315	460	8	(	(	PUNCT
bjmsr-315	460	9	1980	1980	NUM
bjmsr-315	460	10	)	)	PUNCT
bjmsr-315	460	11	a	a	DET
bjmsr-315	460	12	modification	modification	NOUN
bjmsr-315	460	13	to	to	ADP
bjmsr-315	460	14	the	the	DET
bjmsr-315	460	15	discrete	discrete	ADJ
bjmsr-315	460	16	l1	l1	PROPN
bjmsr-315	460	17	linear	linear	PROPN
bjmsr-315	460	18	approximation	approximation	NOUN
bjmsr-315	460	19	algorithm	algorithm	NOUN
bjmsr-315	460	20	of	of	ADP
bjmsr-315	460	21	barrodale	barrodale	NOUN
bjmsr-315	460	22	and	and	CCONJ
bjmsr-315	460	23	roberts	roberts	PROPN
bjmsr-315	460	24	.	.	PUNCT
bjmsr-315	461	1	siam	siam	PROPN
bjmsr-315	461	2	j.	j.	PROPN
bjmsr-315	461	3	on	on	ADP
bjmsr-315	461	4	scientific	scientific	ADJ
bjmsr-315	461	5	and	and	CCONJ
bjmsr-315	461	6	stat	stat	NOUN
bjmsr-315	461	7	.	.	PUNCT
bjmsr-315	462	1	comput	comput	NOUN
bjmsr-315	462	2	.	.	PUNCT
bjmsr-315	463	1	1	1	NUM
bjmsr-315	463	2	,	,	PUNCT
bjmsr-315	463	3	187	187	NUM
bjmsr-315	463	4	-	-	SYM
bjmsr-315	463	5	190	190	NUM
bjmsr-315	463	6	.	.	PUNCT
bjmsr-315	463	7	barrodale	barrodale	NOUN
bjmsr-315	463	8	(	(	PUNCT
bjmsr-315	463	9	1970	1970	NUM
bjmsr-315	463	10	)	)	PUNCT
bjmsr-315	463	11	on	on	ADP
bjmsr-315	463	12	computing	compute	VERB
bjmsr-315	463	13	best	good	ADJ
bjmsr-315	463	14	l1	l1	PROPN
bjmsr-315	463	15	approximations	approximation	NOUN
bjmsr-315	463	16	.	.	PUNCT
bjmsr-315	464	1	in	in	ADP
bjmsr-315	464	2	a.	a.	PROPN
bjmsr-315	464	3	talbot	talbot	PROPN
bjmsr-315	464	4	,	,	PUNCT
bjmsr-315	464	5	approximation	approximation	NOUN
bjmsr-315	464	6	theory	theory	NOUN
bjmsr-315	464	7	academic	academic	ADJ
bjmsr-315	464	8	press	press	NOUN
bjmsr-315	464	9	,	,	PUNCT
bjmsr-315	464	10	new	new	PROPN
bjmsr-315	464	11	york	york	PROPN
bjmsr-315	464	12	,	,	PUNCT
bjmsr-315	464	13	205	205	NUM
bjmsr-315	464	14	-	-	SYM
bjmsr-315	464	15	215	215	NUM
bjmsr-315	464	16	.	.	PUNCT
bjmsr-315	465	1	barrodale	barrodale	NOUN
bjmsr-315	465	2	,	,	PUNCT
bjmsr-315	465	3	f.d.k	f.d.k	NOUN
bjmsr-315	465	4	.	.	PUNCT
bjmsr-315	466	1	roberts	roberts	PROPN
bjmsr-315	466	2	(	(	PUNCT
bjmsr-315	466	3	1973	1973	NUM
bjmsr-315	466	4	)	)	PUNCT
bjmsr-315	466	5	an	an	DET
bjmsr-315	466	6	improved	improved	ADJ
bjmsr-315	466	7	algorithm	algorithm	NOUN
bjmsr-315	466	8	for	for	ADP
bjmsr-315	466	9	discrete	discrete	ADJ
bjmsr-315	466	10	l1	l1	PROPN
bjmsr-315	466	11	linear	linear	PROPN
bjmsr-315	466	12	approximation	approximation	NOUN
bjmsr-315	466	13	.	.	PUNCT
bjmsr-315	467	1	siam	siam	PROPN
bjmsr-315	467	2	j.	j.	PROPN
bjmsr-315	467	3	numer	numer	PROPN
bjmsr-315	467	4	.	.	PUNCT
bjmsr-315	468	1	anal	anal	PROPN
bjmsr-315	468	2	.	.	PROPN
bjmsr-315	468	3	,	,	PUNCT
bjmsr-315	468	4	10	10	NUM
bjmsr-315	468	5	,	,	PUNCT
bjmsr-315	468	6	839	839	NUM
bjmsr-315	468	7	-	-	SYM
bjmsr-315	468	8	848	848	NUM
bjmsr-315	468	9	.	.	PUNCT
bjmsr-315	469	1	barrodale	barrodale	NOUN
bjmsr-315	469	2	,	,	PUNCT
bjmsr-315	469	3	f.d.k	f.d.k	NOUN
bjmsr-315	469	4	.	.	PUNCT
bjmsr-315	470	1	roberts	roberts	PROPN
bjmsr-315	470	2	(	(	PUNCT
bjmsr-315	470	3	1974	1974	NUM
bjmsr-315	470	4	)	)	PUNCT
bjmsr-315	470	5	algorithm	algorithm	NOUN
bjmsr-315	470	6	478	478	NUM
bjmsr-315	470	7	:	:	PUNCT
bjmsr-315	470	8	solution	solution	NOUN
bjmsr-315	470	9	of	of	ADP
bjmsr-315	470	10	an	an	DET
bjmsr-315	470	11	overdetermined	overdetermine	VERB
bjmsr-315	470	12	system	system	NOUN
bjmsr-315	470	13	of	of	ADP
bjmsr-315	470	14	equations	equation	NOUN
bjmsr-315	470	15	in	in	ADP
bjmsr-315	470	16	the	the	DET
bjmsr-315	470	17	l1	l1	PROPN
bjmsr-315	470	18	norm	norm	NOUN
bjmsr-315	470	19	.	.	PUNCT
bjmsr-315	471	1	commun	commun	PROPN
bjmsr-315	471	2	.	.	PUNCT
bjmsr-315	472	1	acm	acm	PROPN
bjmsr-315	472	2	,	,	PUNCT
bjmsr-315	472	3	17	17	NUM
bjmsr-315	472	4	,	,	PUNCT
bjmsr-315	472	5	319	319	NUM
bjmsr-315	472	6	-	-	SYM
bjmsr-315	472	7	320	320	NUM
bjmsr-315	472	8	.	.	PUNCT
bjmsr-315	473	1	barrodale	barrodale	NOUN
bjmsr-315	473	2	,	,	PUNCT
bjmsr-315	473	3	a.	a.	NOUN
bjmsr-315	473	4	young	young	PROPN
bjmsr-315	473	5	(	(	PUNCT
bjmsr-315	473	6	1966	1966	NUM
bjmsr-315	473	7	)	)	PUNCT
bjmsr-315	473	8	algorithms	algorithm	NOUN
bjmsr-315	473	9	for	for	ADP
bjmsr-315	473	10	best	good	ADJ
bjmsr-315	473	11	l1	l1	PROPN
bjmsr-315	473	12	and	and	CCONJ
bjmsr-315	473	13	l∞	l∞	NOUN
bjmsr-315	473	14	linear	linear	ADJ
bjmsr-315	473	15	approximations	approximation	NOUN
bjmsr-315	473	16	on	on	ADP
bjmsr-315	473	17	a	a	DET
bjmsr-315	473	18	discrete	discrete	ADJ
bjmsr-315	473	19	set	set	NOUN
bjmsr-315	473	20	.	.	PUNCT
bjmsr-315	474	1	numer	numer	PROPN
bjmsr-315	474	2	.	.	PUNCT
bjmsr-315	474	3	math	math	PROPN
bjmsr-315	474	4	.	.	PUNCT
bjmsr-315	474	5	,	,	PUNCT
bjmsr-315	474	6	8	8	NUM
bjmsr-315	474	7	,	,	PUNCT
bjmsr-315	474	8	295	295	NUM
bjmsr-315	474	9	-	-	SYM
bjmsr-315	474	10	306	306	NUM
bjmsr-315	474	11	.	.	PUNCT
bjmsr-315	475	1	copyright	copyright	NOUN
bjmsr-315	475	2	©	©	PROPN
bjmsr-315	475	3	cc	cc	PROPN
bjmsr-315	475	4	-	-	PUNCT
bjmsr-315	475	5	by	by	ADP
bjmsr-315	475	6	-	-	PUNCT
bjmsr-315	475	7	nc	nc	PROPN
bjmsr-315	475	8	2019	2019	NUM
bjmsr-315	475	9	,	,	PUNCT
bjmsr-315	475	10	bjmsr	bjmsr	PROPN
bjmsr-315	475	11	www.cribfb.com/journal/index.php/bjmsr	www.cribfb.com/journal/index.php/bjmsr	X
bjmsr-315	476	1	bangladesh	bangladesh	PROPN
bjmsr-315	476	2	journal	journal	PROPN
bjmsr-315	476	3	of	of	ADP
bjmsr-315	476	4	multidisciplinary	multidisciplinary	ADJ
bjmsr-315	476	5	scientific	scientific	ADJ
bjmsr-315	476	6	research	research	NOUN
bjmsr-315	476	7	vol	vol	NOUN
bjmsr-315	476	8	.	.	PROPN
bjmsr-315	476	9	1	1	NUM
bjmsr-315	476	10	,	,	PUNCT
bjmsr-315	476	11	no	no	INTJ
bjmsr-315	476	12	.	.	NOUN
bjmsr-315	476	13	1	1	NUM
bjmsr-315	476	14	;	;	PUNCT
bjmsr-315	476	15	2019	2019	NUM
bjmsr-315	476	16	62	62	NUM
bjmsr-315	476	17	r.h	r.h	PROPN
bjmsr-315	476	18	.	.	PROPN
bjmsr-315	476	19	bartels	bartels	PROPN
bjmsr-315	476	20	,	,	PUNCT
bjmsr-315	476	21	a.r	a.r	PROPN
bjmsr-315	476	22	.	.	PROPN
bjmsr-315	476	23	conn	conn	PROPN
bjmsr-315	476	24	(	(	PUNCT
bjmsr-315	476	25	1977	1977	NUM
bjmsr-315	476	26	)	)	PUNCT
bjmsr-315	476	27	lav	lav	NOUN
bjmsr-315	476	28	regression	regression	NOUN
bjmsr-315	476	29	:	:	PUNCT
bjmsr-315	476	30	a	a	DET
bjmsr-315	476	31	special	special	ADJ
bjmsr-315	476	32	case	case	NOUN
bjmsr-315	476	33	of	of	ADP
bjmsr-315	476	34	piecewise	piecewise	PROPN
bjmsr-315	476	35	linear	linear	PROPN
bjmsr-315	476	36	minimization	minimization	NOUN
bjmsr-315	476	37	.	.	PUNCT
bjmsr-315	477	1	commun	commun	PROPN
bjmsr-315	477	2	.	.	PUNCT
bjmsr-315	478	1	stat	stat	PROPN
bjmsr-315	478	2	.	.	PUNCT
bjmsr-315	478	3	,	,	PUNCT
bjmsr-315	478	4	b6	b6	NOUN
bjmsr-315	478	5	,	,	PUNCT
bjmsr-315	478	6	329340	329340	NUM
bjmsr-315	478	7	.	.	PUNCT
bjmsr-315	479	1	r.h	r.h	PROPN
bjmsr-315	479	2	.	.	PROPN
bjmsr-315	479	3	bartels	bartels	PROPN
bjmsr-315	479	4	,	,	PUNCT
bjmsr-315	479	5	a.r	a.r	PROPN
bjmsr-315	479	6	.	.	PROPN
bjmsr-315	479	7	conn	conn	PROPN
bjmsr-315	479	8	,	,	PUNCT
bjmsr-315	479	9	j.	j.	PROPN
bjmsr-315	479	10	sinclair	sinclair	PROPN
bjmsr-315	479	11	(	(	PUNCT
bjmsr-315	479	12	1976	1976	NUM
bjmsr-315	479	13	)	)	PUNCT
bjmsr-315	479	14	the	the	DET
bjmsr-315	479	15	l1	l1	PROPN
bjmsr-315	479	16	solution	solution	NOUN
bjmsr-315	479	17	to	to	ADP
bjmsr-315	479	18	an	an	DET
bjmsr-315	479	19	overdetermined	overdetermine	VERB
bjmsr-315	479	20	linear	linear	NOUN
bjmsr-315	479	21	system	system	NOUN
bjmsr-315	479	22	.	.	PUNCT
bjmsr-315	480	1	proc	proc	NOUN
bjmsr-315	480	2	.	.	PUNCT
bjmsr-315	481	1	9th	9th	PROPN
bjmsr-315	481	2	ann	ann	PROPN
bjmsr-315	481	3	.	.	PUNCT
bjmsr-315	481	4	symp	symp	PROPN
bjmsr-315	481	5	.	.	PUNCT
bjmsr-315	482	1	interface	interface	PROPN
bjmsr-315	482	2	statist	statist	NOUN
bjmsr-315	482	3	.	.	PUNCT
bjmsr-315	483	1	in	in	ADP
bjmsr-315	483	2	c.d.c	c.d.c	PROPN
bjmsr-315	483	3	.	.	PUNCT
bjmsr-315	484	1	hoaglin	hoaglin	PROPN
bjmsr-315	484	2	(	(	PUNCT
bjmsr-315	484	3	ed	ed	NOUN
bjmsr-315	484	4	.	.	PUNCT
bjmsr-315	484	5	)	)	PUNCT
bjmsr-315	484	6	boston	boston	PROPN
bjmsr-315	484	7	,	,	PUNCT
bjmsr-315	484	8	prindle	prindle	PROPN
bjmsr-315	484	9	,	,	PUNCT
bjmsr-315	484	10	weber	weber	PROPN
bjmsr-315	484	11	and	and	CCONJ
bjmsr-315	484	12	schmidt	schmidt	PROPN
bjmsr-315	484	13	inc	inc	PROPN
bjmsr-315	484	14	.	.	PROPN
bjmsr-315	484	15	,	,	PUNCT
bjmsr-315	484	16	120	120	NUM
bjmsr-315	484	17	-	-	SYM
bjmsr-315	484	18	7	7	NUM
bjmsr-315	484	19	.	.	PUNCT
bjmsr-315	485	1	r.h	r.h	PROPN
bjmsr-315	485	2	.	.	PROPN
bjmsr-315	485	3	bartels	bartels	PROPN
bjmsr-315	485	4	,	,	PUNCT
bjmsr-315	485	5	a.r	a.r	PROPN
bjmsr-315	485	6	.	.	PROPN
bjmsr-315	485	7	conn	conn	PROPN
bjmsr-315	485	8	,	,	PUNCT
bjmsr-315	485	9	j.	j.	PROPN
bjmsr-315	485	10	sinclair	sinclair	PROPN
bjmsr-315	485	11	(	(	PUNCT
bjmsr-315	485	12	1978	1978	NUM
bjmsr-315	485	13	)	)	PUNCT
bjmsr-315	485	14	minimization	minimization	NOUN
bjmsr-315	485	15	technique	technique	NOUN
bjmsr-315	485	16	for	for	ADP
bjmsr-315	485	17	piecewise	piecewise	NOUN
bjmsr-315	485	18	differentiable	differentiable	ADJ
bjmsr-315	485	19	functions	function	NOUN
bjmsr-315	485	20	:	:	PUNCT
bjmsr-315	485	21	the	the	DET
bjmsr-315	485	22	l1	l1	PROPN
bjmsr-315	485	23	solution	solution	NOUN
bjmsr-315	485	24	to	to	ADP
bjmsr-315	485	25	an	an	DET
bjmsr-315	485	26	overdetermined	overdetermine	VERB
bjmsr-315	485	27	linear	linear	NOUN
bjmsr-315	485	28	system	system	NOUN
bjmsr-315	485	29	.	.	PUNCT
bjmsr-315	486	1	siam	siam	PROPN
bjmsr-315	486	2	j.	j.	PROPN
bjmsr-315	486	3	numer	numer	PROPN
bjmsr-315	486	4	.	.	PUNCT
bjmsr-315	487	1	anal	anal	PROPN
bjmsr-315	487	2	.	.	PUNCT
bjmsr-315	488	1	15	15	NUM
bjmsr-315	488	2	,	,	PUNCT
bjmsr-315	488	3	224	224	NUM
bjmsr-315	488	4	-	-	SYM
bjmsr-315	488	5	241	241	NUM
bjmsr-315	488	6	.	.	PUNCT
bjmsr-315	489	1	r.h	r.h	PROPN
bjmsr-315	489	2	.	.	PROPN
bjmsr-315	489	3	bartels	bartels	PROPN
bjmsr-315	489	4	,	,	PUNCT
bjmsr-315	489	5	g.h	g.h	PROPN
bjmsr-315	489	6	.	.	PROPN
bjmsr-315	489	7	golub	golub	PROPN
bjmsr-315	489	8	(	(	PUNCT
bjmsr-315	489	9	1969	1969	NUM
bjmsr-315	489	10	)	)	PUNCT
bjmsr-315	489	11	the	the	DET
bjmsr-315	489	12	simplex	simplex	NOUN
bjmsr-315	489	13	method	method	NOUN
bjmsr-315	489	14	of	of	ADP
bjmsr-315	489	15	linear	linear	ADJ
bjmsr-315	489	16	programming	programming	NOUN
bjmsr-315	489	17	using	use	VERB
bjmsr-315	489	18	lu	lu	NOUN
bjmsr-315	489	19	decomposition	decomposition	NOUN
bjmsr-315	489	20	.	.	PUNCT
bjmsr-315	490	1	commun	commun	PROPN
bjmsr-315	490	2	.	.	PUNCT
bjmsr-315	491	1	acm	acm	PROPN
bjmsr-315	491	2	,	,	PUNCT
bjmsr-315	491	3	12	12	NUM
bjmsr-315	491	4	,	,	PUNCT
bjmsr-315	491	5	266	266	NUM
bjmsr-315	491	6	-	-	SYM
bjmsr-315	491	7	268	268	NUM
bjmsr-315	491	8	.	.	PUNCT
bjmsr-315	492	1	j.	j.	PROPN
bjmsr-315	492	2	bejar	bejar	PROPN
bjmsr-315	492	3	(	(	PUNCT
bjmsr-315	492	4	1956	1956	NUM
bjmsr-315	492	5	)	)	PUNCT
bjmsr-315	492	6	regression	regression	NOUN
bjmsr-315	492	7	en	en	PROPN
bjmsr-315	492	8	mediana	mediana	PROPN
bjmsr-315	492	9	y	y	PROPN
bjmsr-315	492	10	la	la	PROPN
bjmsr-315	492	11	programación	programación	PROPN
bjmsr-315	492	12	lineal	lineal	NOUN
bjmsr-315	492	13	,	,	PUNCT
bjmsr-315	492	14	trabajos	trabajos	PROPN
bjmsr-315	492	15	de	de	PROPN
bjmsr-315	492	16	estadistica	estadistica	PROPN
bjmsr-315	492	17	7	7	NUM
bjmsr-315	492	18	,	,	PUNCT
bjmsr-315	492	19	141	141	NUM
bjmsr-315	492	20	-	-	SYM
bjmsr-315	492	21	58	58	NUM
bjmsr-315	492	22	.	.	PUNCT
bjmsr-315	493	1	j.	j.	PROPN
bjmsr-315	493	2	bejar	bejar	PROPN
bjmsr-315	493	3	(	(	PUNCT
bjmsr-315	493	4	1957	1957	NUM
bjmsr-315	493	5	)	)	PUNCT
bjmsr-315	493	6	calculo	calculo	PROPN
bjmsr-315	493	7	practico	practico	NOUN
bjmsr-315	493	8	de	de	X
bjmsr-315	493	9	la	la	PROPN
bjmsr-315	493	10	regression	regression	PROPN
bjmsr-315	493	11	en	en	PROPN
bjmsr-315	493	12	mediana	mediana	PROPN
bjmsr-315	493	13	trabajos	trabajos	PROPN
bjmsr-315	493	14	de	de	PROPN
bjmsr-315	493	15	estadistica	estadistica	PROPN
bjmsr-315	493	16	,	,	PUNCT
bjmsr-315	493	17	8	8	NUM
bjmsr-315	493	18	,	,	PUNCT
bjmsr-315	493	19	157	157	NUM
bjmsr-315	493	20	-	-	SYM
bjmsr-315	493	21	173	173	NUM
bjmsr-315	493	22	.	.	PUNCT
bjmsr-315	494	1	bijan	bijan	PROPN
bjmsr-315	494	2	bidabad	bidabad	NOUN
bjmsr-315	494	3	(	(	PUNCT
bjmsr-315	494	4	1987a	1987a	NUM
bjmsr-315	494	5	)	)	PUNCT
bjmsr-315	494	6	least	least	ADJ
bjmsr-315	494	7	absolute	absolute	ADJ
bjmsr-315	494	8	error	error	NOUN
bjmsr-315	494	9	estimation	estimation	NOUN
bjmsr-315	494	10	.	.	PUNCT
bjmsr-315	495	1	the	the	DET
bjmsr-315	495	2	first	first	ADJ
bjmsr-315	495	3	international	international	ADJ
bjmsr-315	495	4	conference	conference	NOUN
bjmsr-315	495	5	on	on	ADP
bjmsr-315	495	6	statistical	statistical	ADJ
bjmsr-315	495	7	data	datum	NOUN
bjmsr-315	495	8	analysis	analysis	NOUN
bjmsr-315	495	9	based	base	VERB
bjmsr-315	495	10	on	on	ADP
bjmsr-315	495	11	the	the	DET
bjmsr-315	495	12	l1‎‎	l1‎‎	ADJ
bjmsr-315	495	13	norm	norm	NOUN
bjmsr-315	495	14	and	and	CCONJ
bjmsr-315	495	15	related	related	ADJ
bjmsr-315	495	16	methods	method	NOUN
bjmsr-315	495	17	,	,	PUNCT
bjmsr-315	495	18	neuchatel	neuchatel	NOUN
bjmsr-315	495	19	,	,	PUNCT
bjmsr-315	495	20	switzerland	switzerland	PROPN
bjmsr-315	495	21	.	.	PUNCT
bjmsr-315	496	1	http://www.bidabad.com/doc/lae-i.pdf	http://www.bidabad.com/doc/lae-i.pdf	PROPN
bjmsr-315	496	2	bijan	bijan	PROPN
bjmsr-315	496	3	bidabad	bidabad	NOUN
bjmsr-315	496	4	(	(	PUNCT
bjmsr-315	496	5	1987b	1987b	NUM
bjmsr-315	496	6	)	)	PUNCT
bjmsr-315	496	7	least	least	ADJ
bjmsr-315	496	8	absolute	absolute	ADJ
bjmsr-315	496	9	error	error	NOUN
bjmsr-315	496	10	estimation	estimation	NOUN
bjmsr-315	496	11	,	,	PUNCT
bjmsr-315	496	12	part	part	PROPN
bjmsr-315	496	13	ii	ii	PROPN
bjmsr-315	496	14	.	.	PROPN
bjmsr-315	496	15	submitted	submit	VERB
bjmsr-315	496	16	to	to	ADP
bjmsr-315	496	17	the	the	DET
bjmsr-315	496	18	first	first	ADJ
bjmsr-315	496	19	international	international	ADJ
bjmsr-315	496	20	conference	conference	NOUN
bjmsr-315	496	21	on	on	ADP
bjmsr-315	496	22	statistical	statistical	ADJ
bjmsr-315	496	23	data	datum	NOUN
bjmsr-315	496	24	analysis	analysis	NOUN
bjmsr-315	496	25	based	base	VERB
bjmsr-315	496	26	on	on	ADP
bjmsr-315	496	27	the	the	DET
bjmsr-315	496	28	l1‎‎	l1‎‎	ADJ
bjmsr-315	496	29	norm	norm	NOUN
bjmsr-315	496	30	and	and	CCONJ
bjmsr-315	496	31	related	related	ADJ
bjmsr-315	496	32	methods	method	NOUN
bjmsr-315	496	33	,	,	PUNCT
bjmsr-315	496	34	neuchatel	neuchatel	NOUN
bjmsr-315	496	35	,	,	PUNCT
bjmsr-315	496	36	switzerland	switzerland	PROPN
bjmsr-315	496	37	.	.	PUNCT
bjmsr-315	497	1	http://www.bidabad.com/doc/lae-ii.pdf	http://www.bidabad.com/doc/lae-ii.pdf	PROPN
bjmsr-315	497	2	bijan	bijan	PROPN
bjmsr-315	497	3	bidabad	bidabad	NOUN
bjmsr-315	497	4	(	(	PUNCT
bjmsr-315	497	5	1988a	1988a	NUM
bjmsr-315	497	6	)	)	PUNCT
bjmsr-315	497	7	a	a	DET
bjmsr-315	497	8	proposed	propose	VERB
bjmsr-315	497	9	algorithm	algorithm	NOUN
bjmsr-315	497	10	for	for	ADP
bjmsr-315	497	11	least	least	ADJ
bjmsr-315	497	12	absolute	absolute	ADJ
bjmsr-315	497	13	error	error	NOUN
bjmsr-315	497	14	estimation	estimation	NOUN
bjmsr-315	497	15	.	.	PUNCT
bjmsr-315	498	1	proc	proc	PROPN
bjmsr-315	498	2	.	.	PUNCT
bjmsr-315	499	1	of	of	ADP
bjmsr-315	499	2	the	the	DET
bjmsr-315	499	3	third	third	ADJ
bjmsr-315	499	4	seminar	seminar	NOUN
bjmsr-315	499	5	of	of	ADP
bjmsr-315	499	6	mathematical	mathematical	ADJ
bjmsr-315	499	7	analysis	analysis	NOUN
bjmsr-315	499	8	.	.	PUNCT
bjmsr-315	500	1	shiraz	shiraz	PROPN
bjmsr-315	500	2	univ	univ	PROPN
bjmsr-315	500	3	.	.	PROPN
bjmsr-315	500	4	,	,	PUNCT
bjmsr-315	500	5	24	24	NUM
bjmsr-315	500	6	-	-	SYM
bjmsr-315	500	7	34	34	NUM
bjmsr-315	500	8	,	,	PUNCT
bjmsr-315	500	9	shiraz	shiraz	PROPN
bjmsr-315	500	10	,	,	PUNCT
bjmsr-315	500	11	iran	iran	PROPN
bjmsr-315	500	12	.	.	PUNCT
bjmsr-315	501	1	bijan	bijan	PROPN
bjmsr-315	501	2	bidabad	bidabad	NOUN
bjmsr-315	501	3	(	(	PUNCT
bjmsr-315	501	4	1988b	1988b	NUM
bjmsr-315	501	5	)	)	PUNCT
bjmsr-315	501	6	a	a	DET
bjmsr-315	501	7	proposed	propose	VERB
bjmsr-315	501	8	algorithm	algorithm	NOUN
bjmsr-315	501	9	for	for	ADP
bjmsr-315	501	10	least	least	ADJ
bjmsr-315	501	11	absolute	absolute	ADJ
bjmsr-315	501	12	error	error	NOUN
bjmsr-315	501	13	estimation	estimation	NOUN
bjmsr-315	501	14	,	,	PUNCT
bjmsr-315	501	15	part	part	PROPN
bjmsr-315	501	16	ii	ii	PROPN
bjmsr-315	501	17	.	.	PUNCT
bjmsr-315	501	18	proc	proc	PROPN
bjmsr-315	501	19	.	.	PUNCT
bjmsr-315	502	1	of	of	ADP
bjmsr-315	502	2	the	the	DET
bjmsr-315	502	3	third	third	ADJ
bjmsr-315	502	4	seminar	seminar	NOUN
bjmsr-315	502	5	of	of	ADP
bjmsr-315	502	6	mathematical	mathematical	ADJ
bjmsr-315	502	7	analysis	analysis	NOUN
bjmsr-315	502	8	,	,	PUNCT
bjmsr-315	502	9	shiraz	shiraz	PROPN
bjmsr-315	502	10	univ	univ	PROPN
bjmsr-315	502	11	.	.	PROPN
bjmsr-315	502	12	,	,	PUNCT
bjmsr-315	502	13	35	35	NUM
bjmsr-315	502	14	-	-	SYM
bjmsr-315	502	15	50	50	NUM
bjmsr-315	502	16	,	,	PUNCT
bjmsr-315	502	17	shiraz	shiraz	NOUN
bjmsr-315	502	18	,	,	PUNCT
bjmsr-315	502	19	iran	iran	PROPN
bjmsr-315	502	20	.	.	PUNCT
bjmsr-315	503	1	bijan	bijan	PROPN
bjmsr-315	503	2	bidabad	bidabad	NOUN
bjmsr-315	503	3	(	(	PUNCT
bjmsr-315	503	4	1989a	1989a	NUM
bjmsr-315	503	5	)	)	PUNCT
bjmsr-315	503	6	discrete	discrete	ADJ
bjmsr-315	503	7	and	and	CCONJ
bjmsr-315	503	8	continuous	continuous	ADJ
bjmsr-315	503	9	l1‎‎	l1‎‎	ADJ
bjmsr-315	503	10	norm	norm	NOUN
bjmsr-315	503	11	regressions	regression	NOUN
bjmsr-315	503	12	,	,	PUNCT
bjmsr-315	503	13	proposition	proposition	NOUN
bjmsr-315	503	14	of	of	ADP
bjmsr-315	503	15	discrete	discrete	ADJ
bjmsr-315	503	16	approximation	approximation	NOUN
bjmsr-315	503	17	algorithms	algorithm	NOUN
bjmsr-315	503	18	and	and	CCONJ
bjmsr-315	503	19	continuous	continuous	ADJ
bjmsr-315	503	20	smoothing	smoothing	NOUN
bjmsr-315	503	21	of	of	ADP
bjmsr-315	503	22	concentration	concentration	NOUN
bjmsr-315	503	23	surface	surface	NOUN
bjmsr-315	503	24	,	,	PUNCT
bjmsr-315	503	25	ph.d	ph.d	PROPN
bjmsr-315	503	26	.	.	PUNCT
bjmsr-315	504	1	thesis	thesis	PROPN
bjmsr-315	504	2	,	,	PUNCT
bjmsr-315	504	3	islamic	islamic	PROPN
bjmsr-315	504	4	azad	azad	PROPN
bjmsr-315	504	5	univ	univ	PROPN
bjmsr-315	504	6	.	.	PROPN
bjmsr-315	504	7	,	,	PUNCT
bjmsr-315	504	8	tehran	tehran	PROPN
bjmsr-315	504	9	,	,	PUNCT
bjmsr-315	504	10	iran	iran	PROPN
bjmsr-315	504	11	.	.	PUNCT
bjmsr-315	505	1	http://www.bidabad.com/doc/l1-norm-thesis-en.pdf	http://www.bidabad.com/doc/l1-norm-thesis-en.pdf	PROPN
bjmsr-315	505	2	bijan	bijan	PROPN
bjmsr-315	505	3	bidabad	bidabad	NOUN
bjmsr-315	505	4	(	(	PUNCT
bjmsr-315	505	5	1989b	1989b	NUM
bjmsr-315	505	6	)	)	PUNCT
bjmsr-315	505	7	discrete	discrete	ADJ
bjmsr-315	505	8	and	and	CCONJ
bjmsr-315	505	9	continuous	continuous	ADJ
bjmsr-315	505	10	l1‎‎	l1‎‎	ADJ
bjmsr-315	505	11	norm	norm	NOUN
bjmsr-315	505	12	regressions	regression	NOUN
bjmsr-315	505	13	,	,	PUNCT
bjmsr-315	505	14	proposition	proposition	NOUN
bjmsr-315	505	15	of	of	ADP
bjmsr-315	505	16	discrete	discrete	ADJ
bjmsr-315	505	17	approximation	approximation	NOUN
bjmsr-315	505	18	algorithms	algorithm	NOUN
bjmsr-315	505	19	and	and	CCONJ
bjmsr-315	505	20	continuous	continuous	ADJ
bjmsr-315	505	21	smoothing	smoothing	NOUN
bjmsr-315	505	22	of	of	ADP
bjmsr-315	505	23	concentration	concentration	NOUN
bjmsr-315	505	24	surface	surface	NOUN
bjmsr-315	505	25	,	,	PUNCT
bjmsr-315	505	26	ph.d	ph.d	PROPN
bjmsr-315	505	27	.	.	PUNCT
bjmsr-315	506	1	thesis	thesis	PROPN
bjmsr-315	506	2	,	,	PUNCT
bjmsr-315	506	3	islamic	islamic	PROPN
bjmsr-315	506	4	azad	azad	PROPN
bjmsr-315	506	5	univ	univ	PROPN
bjmsr-315	506	6	.	.	PROPN
bjmsr-315	506	7	,	,	PUNCT
bjmsr-315	506	8	tehran	tehran	PROPN
bjmsr-315	506	9	,	,	PUNCT
bjmsr-315	506	10	iran	iran	PROPN
bjmsr-315	506	11	.	.	PUNCT
bjmsr-315	507	1	farsi	farsi	PROPN
bjmsr-315	507	2	translation	translation	NOUN
bjmsr-315	507	3	.	.	PUNCT
bjmsr-315	508	1	http://www.bidabad.com/doc/l1-norm-thesis-fa.pdf	http://www.bidabad.com/doc/l1-norm-thesis-fa.pdf	PROPN
bjmsr-315	508	2	bijan	bijan	PROPN
bjmsr-315	508	3	bidabad	bidabad	NOUN
bjmsr-315	508	4	(	(	PUNCT
bjmsr-315	508	5	2005	2005	NUM
bjmsr-315	508	6	)	)	PUNCT
bjmsr-315	508	7	.	.	PUNCT
bjmsr-315	509	1	l1	l1	PROPN
bjmsr-315	509	2	norm	norm	PROPN
bjmsr-315	509	3	based	base	VERB
bjmsr-315	509	4	computational	computational	ADJ
bjmsr-315	509	5	algorithms	algorithm	NOUN
bjmsr-315	509	6	.	.	PUNCT
bjmsr-315	510	1	http://www.bidabad.com/doc/l1-article6.pdf	http://www.bidabad.com/doc/l1-article6.pdf	PROPN
bjmsr-315	510	2	bijan	bijan	PROPN
bjmsr-315	510	3	bidabad	bidabad	NOUN
bjmsr-315	510	4	(	(	PUNCT
bjmsr-315	510	5	2005	2005	NUM
bjmsr-315	510	6	)	)	PUNCT
bjmsr-315	510	7	.	.	PUNCT
bjmsr-315	511	1	l1	l1	PROPN
bjmsr-315	511	2	norm	norm	NOUN
bjmsr-315	511	3	solution	solution	NOUN
bjmsr-315	511	4	of	of	ADP
bjmsr-315	511	5	overdetermined	overdetermined	ADJ
bjmsr-315	511	6	system	system	NOUN
bjmsr-315	511	7	of	of	ADP
bjmsr-315	511	8	linear	linear	PROPN
bjmsr-315	511	9	equations	equation	NOUN
bjmsr-315	511	10	.	.	PUNCT
bjmsr-315	512	1	http://www.bidabad.com/doc/l1article5.pdf	http://www.bidabad.com/doc/l1article5.pdf	PROPN
bjmsr-315	512	2	bijan	bijan	PROPN
bjmsr-315	512	3	bidabad	bidabad	NOUN
bjmsr-315	512	4	(	(	PUNCT
bjmsr-315	512	5	2005	2005	NUM
bjmsr-315	512	6	)	)	PUNCT
bjmsr-315	512	7	.	.	PUNCT
bjmsr-315	513	1	l1	l1	PROPN
bjmsr-315	513	2	norm	norm	PROPN
bjmsr-315	513	3	based	base	VERB
bjmsr-315	513	4	data	datum	NOUN
bjmsr-315	513	5	analysis	analysis	NOUN
bjmsr-315	513	6	and	and	CCONJ
bjmsr-315	513	7	related	related	ADJ
bjmsr-315	513	8	methods	method	NOUN
bjmsr-315	513	9	.	.	PUNCT
bjmsr-315	514	1	http://www.bidabad.com/doc/l1-articl1.pdf	http://www.bidabad.com/doc/l1-articl1.pdf	PROPN
bjmsr-315	514	2	bijan	bijan	PROPN
bjmsr-315	514	3	bidabad	bidabad	NOUN
bjmsr-315	514	4	(	(	PUNCT
bjmsr-315	514	5	2005	2005	NUM
bjmsr-315	514	6	)	)	PUNCT
bjmsr-315	514	7	.	.	PUNCT
bjmsr-315	515	1	new	new	ADJ
bjmsr-315	515	2	algorithms	algorithm	NOUN
bjmsr-315	515	3	for	for	ADP
bjmsr-315	515	4	the	the	DET
bjmsr-315	515	5	l1	l1	PROPN
bjmsr-315	515	6	norm	norm	NOUN
bjmsr-315	515	7	regression	regression	NOUN
bjmsr-315	515	8	.	.	PUNCT
bjmsr-315	516	1	http://www.bidabad.com/doc/l1-article2.pdf	http://www.bidabad.com/doc/l1-article2.pdf	ADJ
bjmsr-315	516	2	bijan	bijan	NOUN
bjmsr-315	516	3	bidabad	bidabad	NOUN
bjmsr-315	516	4	(	(	PUNCT
bjmsr-315	516	5	2005	2005	NUM
bjmsr-315	516	6	)	)	PUNCT
bjmsr-315	516	7	.	.	PUNCT
bjmsr-315	517	1	comparative	comparative	ADJ
bjmsr-315	517	2	study	study	NOUN
bjmsr-315	517	3	of	of	ADP
bjmsr-315	517	4	the	the	DET
bjmsr-315	517	5	l1	l1	PROPN
bjmsr-315	517	6	norm	norm	PROPN
bjmsr-315	517	7	regression	regression	NOUN
bjmsr-315	517	8	algorithms	algorithm	NOUN
bjmsr-315	517	9	.	.	PUNCT
bjmsr-315	518	1	http://www.bidabad.com/doc/l1-articl3.pdf	http://www.bidabad.com/doc/l1-articl3.pdf	PROPN
bjmsr-315	518	2	bijan	bijan	PROPN
bjmsr-315	518	3	bidabad	bidabad	NOUN
bjmsr-315	518	4	(	(	PUNCT
bjmsr-315	518	5	2005	2005	NUM
bjmsr-315	518	6	)	)	PUNCT
bjmsr-315	518	7	.	.	PUNCT
bjmsr-315	519	1	continuous	continuous	ADJ
bjmsr-315	519	2	l1	l1	PROPN
bjmsr-315	519	3	norm	norm	NOUN
bjmsr-315	519	4	estimation	estimation	NOUN
bjmsr-315	519	5	of	of	ADP
bjmsr-315	519	6	lorenz	lorenz	PROPN
bjmsr-315	519	7	curve	curve	PROPN
bjmsr-315	519	8	.	.	PUNCT
bjmsr-315	520	1	http://www.bidabad.com/doc/l1-articl4.pdf	http://www.bidabad.com/doc/l1-articl4.pdf	PROPN
bjmsr-315	520	2	bijan	bijan	PROPN
bjmsr-315	520	3	bidabad	bidabad	NOUN
bjmsr-315	520	4	(	(	PUNCT
bjmsr-315	520	5	1993	1993	NUM
bjmsr-315	520	6	)	)	PUNCT
bjmsr-315	520	7	.	.	PUNCT
bjmsr-315	521	1	estimating	estimate	VERB
bjmsr-315	521	2	lorenz	lorenz	PROPN
bjmsr-315	521	3	curve	curve	NOUN
bjmsr-315	521	4	for	for	ADP
bjmsr-315	521	5	iran	iran	PROPN
bjmsr-315	521	6	by	by	ADP
bjmsr-315	521	7	using	use	VERB
bjmsr-315	521	8	continuous	continuous	ADJ
bjmsr-315	521	9	l1	l1	PROPN
bjmsr-315	521	10	norm	norm	NOUN
bjmsr-315	521	11	estimation	estimation	PROPN
bjmsr-315	521	12	,	,	PUNCT
bjmsr-315	521	13	economics	economic	NOUN
bjmsr-315	521	14	and	and	CCONJ
bjmsr-315	521	15	management	management	NOUN
bjmsr-315	521	16	journal	journal	NOUN
bjmsr-315	521	17	,	,	PUNCT
bjmsr-315	521	18	islamic	islamic	PROPN
bjmsr-315	521	19	azad	azad	PROPN
bjmsr-315	521	20	university	university	PROPN
bjmsr-315	521	21	,	,	PUNCT
bjmsr-315	521	22	no	no	INTJ
bjmsr-315	521	23	.	.	NOUN
bjmsr-315	521	24	19	19	NUM
bjmsr-315	521	25	,	,	PUNCT
bjmsr-315	521	26	winter	winter	NOUN
bjmsr-315	521	27	1993	1993	NUM
bjmsr-315	521	28	,	,	PUNCT
bjmsr-315	521	29	pp	pp	ADV
bjmsr-315	521	30	.	.	PUNCT
bjmsr-315	522	1	83	83	NUM
bjmsr-315	522	2	-	-	SYM
bjmsr-315	522	3	101	101	NUM
bjmsr-315	522	4	.	.	PUNCT
bjmsr-315	523	1	http://www.bidabad.com/doc/iraninc-l1.pdf	http://www.bidabad.com/doc/iraninc-l1.pdf	PROPN
bjmsr-315	523	2	bijan	bijan	PROPN
bjmsr-315	523	3	bidabad	bidabad	NOUN
bjmsr-315	523	4	(	(	PUNCT
bjmsr-315	523	5	2005	2005	NUM
bjmsr-315	523	6	)	)	PUNCT
bjmsr-315	523	7	.	.	PUNCT
bjmsr-315	524	1	continuous	continuous	ADJ
bjmsr-315	524	2	l1	l1	PROPN
bjmsr-315	524	3	norm	norm	NOUN
bjmsr-315	524	4	estimation	estimation	NOUN
bjmsr-315	524	5	of	of	ADP
bjmsr-315	524	6	lorenz	lorenz	PROPN
bjmsr-315	524	7	curve	curve	VERB
bjmsr-315	524	8	when	when	SCONJ
bjmsr-315	524	9	probability	probability	NOUN
bjmsr-315	524	10	density	density	NOUN
bjmsr-315	524	11	function	function	NOUN
bjmsr-315	524	12	is	be	AUX
bjmsr-315	524	13	known	know	VERB
bjmsr-315	524	14	.	.	PUNCT
bjmsr-315	525	1	bijan	bijan	PROPN
bjmsr-315	525	2	bidabad	bidabad	NOUN
bjmsr-315	525	3	(	(	PUNCT
bjmsr-315	525	4	2005	2005	NUM
bjmsr-315	525	5	)	)	PUNCT
bjmsr-315	525	6	.	.	PUNCT
bjmsr-315	526	1	usa	usa	PROPN
bjmsr-315	526	2	income	income	PROPN
bjmsr-315	526	3	distribution	distribution	PROPN
bjmsr-315	526	4	counter	counter	NOUN
bjmsr-315	526	5	-	-	NOUN
bjmsr-315	526	6	business	business	NOUN
bjmsr-315	526	7	-	-	PUNCT
bjmsr-315	526	8	cyclical	cyclical	ADJ
bjmsr-315	526	9	trend	trend	NOUN
bjmsr-315	526	10	(	(	PUNCT
bjmsr-315	526	11	estimating	estimate	VERB
bjmsr-315	526	12	lorenz	lorenz	PROPN
bjmsr-315	526	13	curve	curve	NOUN
bjmsr-315	526	14	using	use	VERB
bjmsr-315	526	15	continuous	continuous	ADJ
bjmsr-315	526	16	l1	l1	PROPN
bjmsr-315	526	17	norm	norm	NOUN
bjmsr-315	526	18	estimation	estimation	PROPN
bjmsr-315	526	19	)	)	PUNCT
bjmsr-315	526	20	.	.	PUNCT
bjmsr-315	527	1	first	first	ADJ
bjmsr-315	527	2	meeting	meeting	NOUN
bjmsr-315	527	3	of	of	ADP
bjmsr-315	527	4	the	the	DET
bjmsr-315	527	5	society	society	NOUN
bjmsr-315	527	6	for	for	ADP
bjmsr-315	527	7	the	the	DET
bjmsr-315	527	8	study	study	NOUN
bjmsr-315	527	9	of	of	ADP
bjmsr-315	527	10	economic	economic	ADJ
bjmsr-315	527	11	inequality	inequality	NOUN
bjmsr-315	527	12	(	(	PUNCT
bjmsr-315	527	13	ecineq	ecineq	PROPN
bjmsr-315	527	14	)	)	PUNCT
bjmsr-315	527	15	,	,	PUNCT
bjmsr-315	527	16	palma	palma	PROPN
bjmsr-315	527	17	de	de	PROPN
bjmsr-315	527	18	mallorca	mallorca	PROPN
bjmsr-315	527	19	,	,	PUNCT
bjmsr-315	527	20	spain	spain	PROPN
bjmsr-315	527	21	,	,	PUNCT
bjmsr-315	527	22	july	july	PROPN
bjmsr-315	527	23	20	20	NUM
bjmsr-315	527	24	-	-	SYM
bjmsr-315	527	25	22	22	NUM
bjmsr-315	527	26	,	,	PUNCT
bjmsr-315	527	27	2005	2005	NUM
bjmsr-315	527	28	.	.	PUNCT
bjmsr-315	528	1	http://www.bidabad.com/doc/lae-i.pdf	http://www.bidabad.com/doc/lae-i.pdf	PROPN
bjmsr-315	528	2	http://www.bidabad.com/doc/lae-ii.pdf	http://www.bidabad.com/doc/lae-ii.pdf	PROPN
bjmsr-315	528	3	http://www.bidabad.com/doc/l1-norm-thesis-en.pdf	http://www.bidabad.com/doc/l1-norm-thesis-en.pdf	PROPN
bjmsr-315	528	4	http://www.bidabad.com/doc/l1-norm-thesis-fa.pdf	http://www.bidabad.com/doc/l1-norm-thesis-fa.pdf	PROPN
bjmsr-315	528	5	http://www.bidabad.com/doc/l1-article6.pdf	http://www.bidabad.com/doc/l1-article6.pdf	PROPN
bjmsr-315	528	6	http://www.bidabad.com/doc/l1-article5.pdf	http://www.bidabad.com/doc/l1-article5.pdf	PROPN
bjmsr-315	528	7	http://www.bidabad.com/doc/l1-article5.pdf	http://www.bidabad.com/doc/l1-article5.pdf	PROPN
bjmsr-315	528	8	http://www.bidabad.com/doc/l1-article1.pdf	http://www.bidabad.com/doc/l1-article1.pdf	PROPN
bjmsr-315	528	9	http://www.bidabad.com/doc/l1-article2.pdf	http://www.bidabad.com/doc/l1-article2.pdf	PROPN
bjmsr-315	528	10	http://www.bidabad.com/doc/l1-article3.pdf	http://www.bidabad.com/doc/l1-article3.pdf	PROPN
bjmsr-315	528	11	http://www.bidabad.com/doc/l1-article3.pdf	http://www.bidabad.com/doc/l1-article3.pdf	PROPN
bjmsr-315	528	12	http://www.bidabad.com/doc/l1-article4.pdf	http://www.bidabad.com/doc/l1-article4.pdf	PROPN
bjmsr-315	528	13	http://www.bidabad.com/doc/iraninc-l1.pdf	http://www.bidabad.com/doc/iraninc-l1.pdf	ADP
bjmsr-315	528	14	copyright	copyright	NOUN
bjmsr-315	528	15	©	©	PROPN
bjmsr-315	528	16	cc	cc	PROPN
bjmsr-315	528	17	-	-	PUNCT
bjmsr-315	528	18	by	by	ADP
bjmsr-315	528	19	-	-	PUNCT
bjmsr-315	528	20	nc	nc	PROPN
bjmsr-315	528	21	2019	2019	NUM
bjmsr-315	528	22	,	,	PUNCT
bjmsr-315	528	23	bjmsr	bjmsr	PROPN
bjmsr-315	528	24	www.cribfb.com/journal/index.php/bjmsr	www.cribfb.com/journal/index.php/bjmsr	X
bjmsr-315	528	25	bangladesh	bangladesh	PROPN
bjmsr-315	528	26	journal	journal	PROPN
bjmsr-315	528	27	of	of	ADP
bjmsr-315	528	28	multidisciplinary	multidisciplinary	ADJ
bjmsr-315	528	29	scientific	scientific	ADJ
bjmsr-315	528	30	research	research	NOUN
bjmsr-315	528	31	vol	vol	NOUN
bjmsr-315	528	32	.	.	PROPN
bjmsr-315	529	1	1	1	NUM
bjmsr-315	529	2	,	,	PUNCT
bjmsr-315	529	3	no	no	INTJ
bjmsr-315	529	4	.	.	NOUN
bjmsr-315	529	5	1	1	NUM
bjmsr-315	529	6	;	;	PUNCT
bjmsr-315	529	7	2019	2019	NUM
bjmsr-315	529	8	63	63	NUM
bjmsr-315	529	9	http://www.uib.es/congres/ecopub/ecineq/general.html	http://www.uib.es/congres/ecopub/ecineq/general.html	PROPN
bjmsr-315	529	10	http://www.uib.es/congres/ecopub/ecineq/papers/039bidabab.pdf	http://www.uib.es/congres/ecopub/ecineq/papers/039bidabab.pdf	PROPN
bjmsr-315	529	11	http://www.bidabad.com/doc/estimating-lorenz-us.pdf	http://www.bidabad.com/doc/estimating-lorenz-us.pdf	PROPN
bjmsr-315	529	12	bijan	bijan	PROPN
bjmsr-315	529	13	bidabad	bidabad	PROPN
bjmsr-315	529	14	,	,	PUNCT
bjmsr-315	529	15	hamid	hamid	PROPN
bjmsr-315	529	16	shahrestani	shahrestani	PROPN
bjmsr-315	529	17	.	.	PUNCT
bjmsr-315	530	1	(	(	PUNCT
bjmsr-315	530	2	2008	2008	NUM
bjmsr-315	530	3	)	)	PUNCT
bjmsr-315	530	4	an	an	DET
bjmsr-315	530	5	implied	imply	VERB
bjmsr-315	530	6	inequality	inequality	NOUN
bjmsr-315	530	7	index	index	NOUN
bjmsr-315	530	8	using	use	VERB
bjmsr-315	530	9	l1	l1	PROPN
bjmsr-315	530	10	norm	norm	NOUN
bjmsr-315	530	11	estimation	estimation	NOUN
bjmsr-315	530	12	of	of	ADP
bjmsr-315	530	13	lorenz	lorenz	PROPN
bjmsr-315	530	14	curve	curve	PROPN
bjmsr-315	530	15	.	.	PUNCT
bjmsr-315	531	1	global	global	ADJ
bjmsr-315	531	2	conference	conference	NOUN
bjmsr-315	531	3	on	on	ADP
bjmsr-315	531	4	business	business	NOUN
bjmsr-315	531	5	and	and	CCONJ
bjmsr-315	531	6	finance	finance	NOUN
bjmsr-315	531	7	proceedings	proceeding	NOUN
bjmsr-315	531	8	.	.	PUNCT
bjmsr-315	532	1	mercedes	mercede	NOUN
bjmsr-315	532	2	jalbert	jalbert	PROPN
bjmsr-315	532	3	,	,	PUNCT
bjmsr-315	532	4	managing	managing	NOUN
bjmsr-315	532	5	editor	editor	NOUN
bjmsr-315	532	6	,	,	PUNCT
bjmsr-315	532	7	issn	issn	PROPN
bjmsr-315	532	8	1931	1931	NUM
bjmsr-315	532	9	-	-	SYM
bjmsr-315	532	10	0285	0285	NUM
bjmsr-315	532	11	cd	cd	PROPN
bjmsr-315	532	12	,	,	PUNCT
bjmsr-315	532	13	issn	issn	PROPN
bjmsr-315	532	14	1941	1941	NUM
bjmsr-315	532	15	-	-	SYM
bjmsr-315	532	16	9589	9589	NUM
bjmsr-315	532	17	online	online	NOUN
bjmsr-315	532	18	,	,	PUNCT
bjmsr-315	532	19	volume	volume	NOUN
bjmsr-315	532	20	3	3	NUM
bjmsr-315	532	21	,	,	PUNCT
bjmsr-315	532	22	number	number	NOUN
bjmsr-315	532	23	2	2	NUM
bjmsr-315	532	24	,	,	PUNCT
bjmsr-315	532	25	2008	2008	NUM
bjmsr-315	532	26	,	,	PUNCT
bjmsr-315	532	27	the	the	DET
bjmsr-315	532	28	institute	institute	NOUN
bjmsr-315	532	29	for	for	ADP
bjmsr-315	532	30	business	business	NOUN
bjmsr-315	532	31	and	and	CCONJ
bjmsr-315	532	32	finance	finance	NOUN
bjmsr-315	532	33	research	research	NOUN
bjmsr-315	532	34	,	,	PUNCT
bjmsr-315	532	35	ramada	ramada	PROPN
bjmsr-315	532	36	plaza	plaza	PROPN
bjmsr-315	532	37	herradura	herradura	NOUN
bjmsr-315	532	38	,	,	PUNCT
bjmsr-315	532	39	san	san	PROPN
bjmsr-315	532	40	jose	jose	PROPN
bjmsr-315	532	41	,	,	PUNCT
bjmsr-315	532	42	costa	costa	PROPN
bjmsr-315	532	43	rica	rica	PROPN
bjmsr-315	532	44	,	,	PUNCT
bjmsr-315	532	45	may	may	AUX
bjmsr-315	532	46	28	28	NUM
bjmsr-315	532	47	-	-	SYM
bjmsr-315	532	48	31	31	NUM
bjmsr-315	532	49	,	,	PUNCT
bjmsr-315	532	50	2008	2008	NUM
bjmsr-315	532	51	,	,	PUNCT
bjmsr-315	532	52	pp	pp	ADJ
bjmsr-315	532	53	.	.	PUNCT
bjmsr-315	533	1	148	148	NUM
bjmsr-315	533	2	-	-	SYM
bjmsr-315	533	3	163	163	NUM
bjmsr-315	533	4	.	.	PUNCT
bjmsr-315	534	1	global	global	ADJ
bjmsr-315	534	2	journal	journal	PROPN
bjmsr-315	534	3	of	of	ADP
bjmsr-315	534	4	business	business	NOUN
bjmsr-315	534	5	research	research	NOUN
bjmsr-315	534	6	,	,	PUNCT
bjmsr-315	534	7	vol	vol	NOUN
bjmsr-315	534	8	.	.	PROPN
bjmsr-315	534	9	4	4	NUM
bjmsr-315	534	10	,	,	PUNCT
bjmsr-315	534	11	no	no	INTJ
bjmsr-315	534	12	.	.	NOUN
bjmsr-315	534	13	1	1	NUM
bjmsr-315	534	14	,	,	PUNCT
bjmsr-315	534	15	2010	2010	NUM
bjmsr-315	534	16	,	,	PUNCT
bjmsr-315	534	17	pp.29	pp.29	NOUN
bjmsr-315	534	18	-	-	PUNCT
bjmsr-315	534	19	45	45	NUM
bjmsr-315	534	20	.	.	PUNCT
bjmsr-315	535	1	http://www.bidabad.com/doc/l1-implied-inequality-index-4.pdf	http://www.bidabad.com/doc/l1-implied-inequality-index-4.pdf	PROPN
bjmsr-315	535	2	http://www.theibfr.com/archive/issn-1941-9589-v3-n2-2008.pdf	http://www.theibfr.com/archive/issn-1941-9589-v3-n2-2008.pdf	PROPN
bjmsr-315	535	3	http://www.bidabad.com/doc/ssrn-id1631861.pdf	http://www.bidabad.com/doc/ssrn-id1631861.pdf	PROPN
bjmsr-315	535	4	p.	p.	PROPN
bjmsr-315	535	5	bloomfield	bloomfield	PROPN
bjmsr-315	535	6	,	,	PUNCT
bjmsr-315	535	7	w.	w.	PROPN
bjmsr-315	535	8	steiger	steiger	PROPN
bjmsr-315	535	9	(	(	PUNCT
bjmsr-315	535	10	1980	1980	NUM
bjmsr-315	535	11	)	)	PUNCT
bjmsr-315	535	12	least	least	ADJ
bjmsr-315	535	13	absolute	absolute	ADJ
bjmsr-315	535	14	deviations	deviation	NOUN
bjmsr-315	535	15	curve	curve	VERB
bjmsr-315	535	16	fitting	fitting	ADJ
bjmsr-315	535	17	.	.	PUNCT
bjmsr-315	536	1	siam	siam	PROPN
bjmsr-315	536	2	j.	j.	PROPN
bjmsr-315	536	3	sci	sci	PROPN
bjmsr-315	536	4	.	.	PROPN
bjmsr-315	536	5	statist	statist	PROPN
bjmsr-315	536	6	.	.	PUNCT
bjmsr-315	537	1	comput	comput	NOUN
bjmsr-315	537	2	.	.	PUNCT
bjmsr-315	538	1	1	1	NUM
bjmsr-315	538	2	,	,	PUNCT
bjmsr-315	538	3	290	290	NUM
bjmsr-315	538	4	-	-	SYM
bjmsr-315	538	5	301	301	NUM
bjmsr-315	538	6	.	.	PUNCT
bjmsr-315	539	1	p.	p.	PROPN
bjmsr-315	539	2	bloomfield	bloomfield	PROPN
bjmsr-315	539	3	,	,	PUNCT
bjmsr-315	539	4	w.	w.	PROPN
bjmsr-315	539	5	steiger	steiger	PROPN
bjmsr-315	539	6	(	(	PUNCT
bjmsr-315	539	7	1983	1983	NUM
bjmsr-315	539	8	)	)	PUNCT
bjmsr-315	539	9	least	least	ADJ
bjmsr-315	539	10	absolute	absolute	ADJ
bjmsr-315	539	11	deviations	deviation	NOUN
bjmsr-315	539	12	:	:	PUNCT
bjmsr-315	540	1	theory	theory	NOUN
bjmsr-315	540	2	,	,	PUNCT
bjmsr-315	540	3	applications	application	NOUN
bjmsr-315	540	4	and	and	CCONJ
bjmsr-315	540	5	algorithms	algorithm	NOUN
bjmsr-315	540	6	.	.	PUNCT
bjmsr-315	541	1	birkhauser	birkhauser	PROPN
bjmsr-315	541	2	,	,	PUNCT
bjmsr-315	541	3	boston	boston	PROPN
bjmsr-315	541	4	.	.	PUNCT
bjmsr-315	542	1	r.j	r.j	PROPN
bjmsr-315	542	2	.	.	PROPN
bjmsr-315	542	3	boscovich	boscovich	PROPN
bjmsr-315	542	4	(	(	PUNCT
bjmsr-315	542	5	1757	1757	NUM
bjmsr-315	542	6	)	)	PUNCT
bjmsr-315	542	7	de	de	X
bjmsr-315	542	8	litteraria	litteraria	X
bjmsr-315	542	9	expeditione	expeditione	NOUN
bjmsr-315	542	10	per	per	ADP
bjmsr-315	542	11	pontificiam	pontificiam	PROPN
bjmsr-315	542	12	ditionem	ditionem	PROPN
bjmsr-315	542	13	,	,	PUNCT
bjmsr-315	542	14	et	et	PROPN
bjmsr-315	542	15	synopsis	synopsis	PROPN
bjmsr-315	542	16	amplioris	amplioris	PROPN
bjmsr-315	542	17	operis	operis	PROPN
bjmsr-315	542	18	...	...	PUNCT
bjmsr-315	542	19	,	,	PUNCT
bjmsr-315	542	20	'	'	PUNCT
bjmsr-315	542	21	bononiensi	bononiensi	VERB
bjmsr-315	542	22	'	'	PUNCT
bjmsr-315	542	23	scientiarum	scientiarum	PROPN
bjmsr-315	542	24	et	et	PROPN
bjmsr-315	542	25	artum	artum	PROPN
bjmsr-315	542	26	instituto	instituto	PROPN
bjmsr-315	542	27	atque	atque	PROPN
bjmsr-315	542	28	academia	academia	PROPN
bjmsr-315	542	29	commetarii	commetarii	PROPN
bjmsr-315	542	30	,	,	PUNCT
bjmsr-315	542	31	vol.4	vol.4	PROPN
bjmsr-315	542	32	,	,	PUNCT
bjmsr-315	542	33	353	353	NUM
bjmsr-315	542	34	-	-	SYM
bjmsr-315	542	35	396	396	NUM
bjmsr-315	542	36	.	.	PUNCT
bjmsr-315	543	1	reprinted	reprint	VERB
bjmsr-315	543	2	with	with	ADP
bjmsr-315	543	3	a	a	DET
bjmsr-315	543	4	serbo	serbo	NOUN
bjmsr-315	543	5	-	-	ADJ
bjmsr-315	543	6	croatian	croatian	ADJ
bjmsr-315	543	7	translation	translation	NOUN
bjmsr-315	543	8	by	by	ADP
bjmsr-315	543	9	n.	n.	PROPN
bjmsr-315	543	10	cubranic	cubranic	PROPN
bjmsr-315	543	11	,	,	PUNCT
bjmsr-315	543	12	institute	institute	NOUN
bjmsr-315	543	13	of	of	ADP
bjmsr-315	543	14	higher	high	ADJ
bjmsr-315	543	15	geodesy	geodesy	PROPN
bjmsr-315	543	16	,	,	PUNCT
bjmsr-315	543	17	university	university	NOUN
bjmsr-315	543	18	of	of	ADP
bjmsr-315	543	19	zagreb	zagreb	PROPN
bjmsr-315	543	20	1961	1961	NUM
bjmsr-315	543	21	.	.	PUNCT
bjmsr-315	544	1	r.j	r.j	PROPN
bjmsr-315	544	2	.	.	PROPN
bjmsr-315	544	3	boscovich	boscovich	PROPN
bjmsr-315	544	4	(	(	PUNCT
bjmsr-315	544	5	1760	1760	NUM
bjmsr-315	544	6	)	)	PUNCT
bjmsr-315	544	7	de	de	PROPN
bjmsr-315	544	8	recentissimis	recentissimis	NOUN
bjmsr-315	544	9	graduum	graduum	PROPN
bjmsr-315	544	10	dimensionibus	dimensionibus	PROPN
bjmsr-315	544	11	et	et	PROPN
bjmsr-315	544	12	figura	figura	PROPN
bjmsr-315	544	13	,	,	PUNCT
bjmsr-315	544	14	ac	ac	PROPN
bjmsr-315	544	15	magnitudine	magnitudine	PROPN
bjmsr-315	544	16	terrae	terrae	PROPN
bjmsr-315	544	17	inde	inde	PROPN
bjmsr-315	544	18	derivanda	derivanda	PROPN
bjmsr-315	544	19	.	.	PUNCT
bjmsr-315	545	1	philosophiae	philosophiae	NOUN
bjmsr-315	545	2	recentioris	recentioris	PROPN
bjmsr-315	545	3	,	,	PUNCT
bjmsr-315	545	4	a	a	DET
bjmsr-315	545	5	benedicto	benedicto	NOUN
bjmsr-315	545	6	stay	stay	VERB
bjmsr-315	545	7	in	in	ADP
bjmsr-315	545	8	romano	romano	NOUN
bjmsr-315	545	9	archigynasis	archigynasis	NOUN
bjmsr-315	545	10	publico	publico	PROPN
bjmsr-315	545	11	eloquentare	eloquentare	NOUN
bjmsr-315	545	12	professore	professore	PROPN
bjmsr-315	545	13	,	,	PUNCT
bjmsr-315	545	14	vesibus	vesibus	NOUN
bjmsr-315	545	15	traditae	traditae	NOUN
bjmsr-315	545	16	,	,	PUNCT
bjmsr-315	545	17	libri	libri	NOUN
bjmsr-315	545	18	x	x	PROPN
bjmsr-315	545	19	,	,	PUNCT
bjmsr-315	545	20	cum	cum	PROPN
bjmsr-315	545	21	adnotianibus	adnotianibus	PROPN
bjmsr-315	545	22	et	et	PROPN
bjmsr-315	545	23	supplementas	supplementas	PROPN
bjmsr-315	545	24	p.	p.	PROPN
bjmsr-315	545	25	rugerii	rugerii	PROPN
bjmsr-315	545	26	joseph	joseph	PROPN
bjmsr-315	545	27	boscovich	boscovich	PROPN
bjmsr-315	545	28	,	,	PUNCT
bjmsr-315	545	29	s.j	s.j	PROPN
bjmsr-315	545	30	.	.	PROPN
bjmsr-315	545	31	,	,	PUNCT
bjmsr-315	545	32	2	2	NUM
bjmsr-315	545	33	,	,	PUNCT
bjmsr-315	545	34	406	406	NUM
bjmsr-315	545	35	-	-	SYM
bjmsr-315	545	36	426	426	NUM
bjmsr-315	545	37	.	.	PUNCT
bjmsr-315	546	1	a.l	a.l	PROPN
bjmsr-315	546	2	.	.	PROPN
bjmsr-315	546	3	bowley	bowley	PROPN
bjmsr-315	546	4	(	(	PUNCT
bjmsr-315	546	5	1902	1902	NUM
bjmsr-315	546	6	)	)	PUNCT
bjmsr-315	546	7	methods	method	NOUN
bjmsr-315	546	8	of	of	ADP
bjmsr-315	546	9	representing	represent	VERB
bjmsr-315	546	10	the	the	DET
bjmsr-315	546	11	statistics	statistic	NOUN
bjmsr-315	546	12	of	of	ADP
bjmsr-315	546	13	wages	wage	NOUN
bjmsr-315	546	14	and	and	CCONJ
bjmsr-315	546	15	other	other	ADJ
bjmsr-315	546	16	groups	group	NOUN
bjmsr-315	546	17	not	not	PART
bjmsr-315	546	18	fulfilling	fulfil	VERB
bjmsr-315	546	19	the	the	DET
bjmsr-315	546	20	normal	normal	ADJ
bjmsr-315	546	21	law	law	NOUN
bjmsr-315	546	22	of	of	ADP
bjmsr-315	546	23	error	error	NOUN
bjmsr-315	546	24	,	,	PUNCT
bjmsr-315	546	25	ii	ii	PROPN
bjmsr-315	546	26	:	:	PUNCT
bjmsr-315	546	27	applications	application	NOUN
bjmsr-315	546	28	to	to	PART
bjmsr-315	546	29	wage	wage	VERB
bjmsr-315	546	30	statistics	statistic	NOUN
bjmsr-315	546	31	and	and	CCONJ
bjmsr-315	546	32	other	other	ADJ
bjmsr-315	546	33	groups	group	NOUN
bjmsr-315	546	34	.	.	PUNCT
bjmsr-315	547	1	j.	j.	PROPN
bjmsr-315	547	2	of	of	ADP
bjmsr-315	547	3	the	the	DET
bjmsr-315	547	4	roy	roy	PROPN
bjmsr-315	547	5	.	.	PROPN
bjmsr-315	547	6	stat	stat	PROPN
bjmsr-315	547	7	.	.	PUNCT
bjmsr-315	548	1	soc	soc	PROPN
bjmsr-315	548	2	.	.	PROPN
bjmsr-315	548	3	,	,	PUNCT
bjmsr-315	548	4	65	65	NUM
bjmsr-315	548	5	,	,	PUNCT
bjmsr-315	548	6	331	331	NUM
bjmsr-315	548	7	-	-	SYM
bjmsr-315	548	8	54	54	NUM
bjmsr-315	548	9	.	.	PUNCT
bjmsr-315	549	1	a.l	a.l	PROPN
bjmsr-315	549	2	.	.	PROPN
bjmsr-315	549	3	bowley	bowley	PROPN
bjmsr-315	549	4	(	(	PUNCT
bjmsr-315	549	5	1928	1928	NUM
bjmsr-315	549	6	)	)	PUNCT
bjmsr-315	549	7	f.y	f.y	PROPN
bjmsr-315	549	8	.	.	PROPN
bjmsr-315	549	9	edgeworth	edgeworth	PROPN
bjmsr-315	549	10	's	's	PART
bjmsr-315	549	11	contributions	contribution	NOUN
bjmsr-315	549	12	to	to	ADP
bjmsr-315	549	13	mathematical	mathematical	ADJ
bjmsr-315	549	14	statistics	statistic	NOUN
bjmsr-315	549	15	.	.	PUNCT
bjmsr-315	550	1	london	london	PROPN
bjmsr-315	550	2	,	,	PUNCT
bjmsr-315	550	3	roy	roy	PROPN
bjmsr-315	550	4	.	.	PROPN
bjmsr-315	550	5	stat	stat	PROPN
bjmsr-315	550	6	.	.	PUNCT
bjmsr-315	551	1	soc	soc	PROPN
bjmsr-315	551	2	..	..	PUNCT
bjmsr-315	551	3	d.	d.	PROPN
bjmsr-315	551	4	bradu	bradu	PROPN
bjmsr-315	551	5	(	(	PUNCT
bjmsr-315	551	6	1987a	1987a	NUM
bjmsr-315	551	7	)	)	PUNCT
bjmsr-315	551	8	l1	l1	PROPN
bjmsr-315	551	9	fit	fit	PROPN
bjmsr-315	551	10	,	,	PUNCT
bjmsr-315	551	11	median	median	ADJ
bjmsr-315	551	12	polish	polish	NOUN
bjmsr-315	551	13	and	and	CCONJ
bjmsr-315	551	14	conjugate	conjugate	ADJ
bjmsr-315	551	15	gradients	gradient	NOUN
bjmsr-315	551	16	.	.	PUNCT
bjmsr-315	552	1	csir	csir	PROPN
bjmsr-315	552	2	tech	tech	PROPN
bjmsr-315	552	3	.	.	PUNCT
bjmsr-315	553	1	rep	rep	PROPN
bjmsr-315	553	2	.	.	PROPN
bjmsr-315	553	3	twisk	twisk	PROPN
bjmsr-315	553	4	509	509	NUM
bjmsr-315	553	5	,	,	PUNCT
bjmsr-315	553	6	national	national	ADJ
bjmsr-315	553	7	res	re	NOUN
bjmsr-315	553	8	.	.	PUNCT
bjmsr-315	553	9	inst	inst	PROPN
bjmsr-315	553	10	.	.	PROPN
bjmsr-315	554	1	for	for	ADP
bjmsr-315	554	2	math	math	PROPN
bjmsr-315	554	3	sci	sci	PROPN
bjmsr-315	554	4	.	.	PUNCT
bjmsr-315	554	5	csir	csir	PROPN
bjmsr-315	554	6	,	,	PUNCT
bjmsr-315	554	7	pretoria	pretoria	PROPN
bjmsr-315	554	8	.	.	PUNCT
bjmsr-315	554	9	d.	d.	PROPN
bjmsr-315	554	10	bradu	bradu	PROPN
bjmsr-315	554	11	(	(	PUNCT
bjmsr-315	554	12	1987b	1987b	PROPN
bjmsr-315	554	13	)	)	PUNCT
bjmsr-315	554	14	an	an	DET
bjmsr-315	554	15	ε	ε	PROPN
bjmsr-315	554	16	-	-	PUNCT
bjmsr-315	554	17	median	median	ADJ
bjmsr-315	554	18	polish	polish	NOUN
bjmsr-315	554	19	algorithm	algorithm	NOUN
bjmsr-315	554	20	.	.	PUNCT
bjmsr-315	555	1	csda	csda	NOUN
bjmsr-315	555	2	,	,	PUNCT
bjmsr-315	555	3	5	5	NUM
bjmsr-315	555	4	,	,	PUNCT
bjmsr-315	555	5	327	327	NUM
bjmsr-315	555	6	-	-	SYM
bjmsr-315	555	7	336	336	NUM
bjmsr-315	555	8	.	.	PUNCT
bjmsr-315	556	1	j.j	j.j	PROPN
bjmsr-315	556	2	.	.	PROPN
bjmsr-315	556	3	brennan	brennan	PROPN
bjmsr-315	556	4	,	,	PUNCT
bjmsr-315	556	5	l.m	l.m	PROPN
bjmsr-315	556	6	.	.	PROPN
bjmsr-315	556	7	seiford	seiford	PROPN
bjmsr-315	556	8	(	(	PUNCT
bjmsr-315	556	9	1987	1987	NUM
bjmsr-315	556	10	)	)	PUNCT
bjmsr-315	556	11	linear	linear	NOUN
bjmsr-315	556	12	programming	programming	NOUN
bjmsr-315	556	13	and	and	CCONJ
bjmsr-315	556	14	l1	l1	PROPN
bjmsr-315	556	15	approximation	approximation	NOUN
bjmsr-315	556	16	using	use	VERB
bjmsr-315	556	17	the	the	DET
bjmsr-315	556	18	method	method	NOUN
bjmsr-315	556	19	of	of	ADP
bjmsr-315	556	20	vanishing	vanish	VERB
bjmsr-315	556	21	jacobians	jacobian	NOUN
bjmsr-315	556	22	.	.	PUNCT
bjmsr-315	557	1	csda	csda	NOUN
bjmsr-315	557	2	,	,	PUNCT
bjmsr-315	557	3	5	5	NUM
bjmsr-315	557	4	,	,	PUNCT
bjmsr-315	557	5	263	263	NUM
bjmsr-315	557	6	-	-	SYM
bjmsr-315	557	7	276	276	NUM
bjmsr-315	557	8	.	.	PUNCT
bjmsr-315	558	1	b.m	b.m	PROPN
bjmsr-315	558	2	.	.	PROPN
bjmsr-315	558	3	brown	brown	PROPN
bjmsr-315	558	4	(	(	PUNCT
bjmsr-315	558	5	1980	1980	NUM
bjmsr-315	558	6	)	)	PUNCT
bjmsr-315	558	7	median	median	ADJ
bjmsr-315	558	8	estimates	estimate	NOUN
bjmsr-315	558	9	in	in	ADP
bjmsr-315	558	10	a	a	DET
bjmsr-315	558	11	simple	simple	ADJ
bjmsr-315	558	12	linear	linear	ADJ
bjmsr-315	558	13	regression	regression	NOUN
bjmsr-315	558	14	.	.	PUNCT
bjmsr-315	559	1	australian	australian	ADJ
bjmsr-315	559	2	j.	j.	PROPN
bjmsr-315	559	3	of	of	ADP
bjmsr-315	559	4	stat	stat	PROPN
bjmsr-315	559	5	.	.	PUNCT
bjmsr-315	559	6	,	,	PUNCT
bjmsr-315	559	7	22	22	NUM
bjmsr-315	559	8	,	,	PUNCT
bjmsr-315	559	9	154	154	NUM
bjmsr-315	559	10	-	-	SYM
bjmsr-315	559	11	165	165	NUM
bjmsr-315	559	12	.	.	PUNCT
bjmsr-315	560	1	c.	c.	NOUN
bjmsr-315	560	2	bruen	bruen	PROPN
bjmsr-315	560	3	(	(	PUNCT
bjmsr-315	560	4	1938	1938	NUM
bjmsr-315	560	5	)	)	PUNCT
bjmsr-315	560	6	methods	method	NOUN
bjmsr-315	560	7	for	for	ADP
bjmsr-315	560	8	the	the	DET
bjmsr-315	560	9	combination	combination	NOUN
bjmsr-315	560	10	of	of	ADP
bjmsr-315	560	11	observations	observation	NOUN
bjmsr-315	560	12	modal	modal	ADJ
bjmsr-315	560	13	points	point	NOUN
bjmsr-315	560	14	or	or	CCONJ
bjmsr-315	560	15	most	most	ADV
bjmsr-315	560	16	lesser	less	ADJ
bjmsr-315	560	17	-	-	PUNCT
bjmsr-315	560	18	deviations	deviation	NOUN
bjmsr-315	560	19	,	,	PUNCT
bjmsr-315	560	20	mean	mean	ADJ
bjmsr-315	560	21	loci	locus	NOUN
bjmsr-315	560	22	or	or	CCONJ
bjmsr-315	560	23	least	least	ADJ
bjmsr-315	560	24	squares	square	NOUN
bjmsr-315	560	25	,	,	PUNCT
bjmsr-315	560	26	and	and	CCONJ
bjmsr-315	560	27	mid	mid	ADJ
bjmsr-315	560	28	point	point	NOUN
bjmsr-315	560	29	of	of	ADP
bjmsr-315	560	30	least	least	ADJ
bjmsr-315	560	31	range	range	NOUN
bjmsr-315	560	32	or	or	CCONJ
bjmsr-315	560	33	least	least	ADJ
bjmsr-315	560	34	greatest	great	ADJ
bjmsr-315	560	35	-	-	PUNCT
bjmsr-315	560	36	deviation	deviation	NOUN
bjmsr-315	560	37	.	.	PUNCT
bjmsr-315	561	1	metron	metron	PROPN
bjmsr-315	561	2	13	13	NUM
bjmsr-315	561	3	,	,	PUNCT
bjmsr-315	561	4	61	61	NUM
bjmsr-315	561	5	-	-	SYM
bjmsr-315	561	6	140	140	NUM
bjmsr-315	561	7	.	.	PUNCT
bjmsr-315	562	1	j.	j.	PROPN
bjmsr-315	562	2	chamber	chamber	PROPN
bjmsr-315	562	3	(	(	PUNCT
bjmsr-315	562	4	1971	1971	NUM
bjmsr-315	562	5	)	)	PUNCT
bjmsr-315	562	6	algorithm	algorithm	NOUN
bjmsr-315	562	7	410	410	NUM
bjmsr-315	562	8	:	:	PUNCT
bjmsr-315	562	9	partial	partial	ADJ
bjmsr-315	562	10	sorting	sorting	NOUN
bjmsr-315	562	11	.	.	PUNCT
bjmsr-315	563	1	comm	comm	NOUN
bjmsr-315	563	2	.	.	PUNCT
bjmsr-315	564	1	acm	acm	PROPN
bjmsr-315	564	2	,	,	PUNCT
bjmsr-315	564	3	14	14	NUM
bjmsr-315	564	4	,	,	PUNCT
bjmsr-315	564	5	357	357	NUM
bjmsr-315	564	6	-	-	SYM
bjmsr-315	564	7	358	358	NUM
bjmsr-315	564	8	.	.	PUNCT
bjmsr-315	565	1	j.m	j.m	PROPN
bjmsr-315	565	2	.	.	PROPN
bjmsr-315	565	3	chambers	chambers	PROPN
bjmsr-315	565	4	(	(	PUNCT
bjmsr-315	565	5	1977	1977	NUM
bjmsr-315	565	6	)	)	PUNCT
bjmsr-315	565	7	computational	computational	ADJ
bjmsr-315	565	8	methods	method	NOUN
bjmsr-315	565	9	for	for	ADP
bjmsr-315	565	10	data	datum	NOUN
bjmsr-315	565	11	analysis	analysis	NOUN
bjmsr-315	565	12	wiley	wiley	NOUN
bjmsr-315	565	13	,	,	PUNCT
bjmsr-315	565	14	new	new	PROPN
bjmsr-315	565	15	york	york	PROPN
bjmsr-315	565	16	.	.	PUNCT
bjmsr-315	566	1	charnes	charnes	PROPN
bjmsr-315	566	2	,	,	PUNCT
bjmsr-315	566	3	w.w	w.w	PROPN
bjmsr-315	566	4	.	.	PROPN
bjmsr-315	566	5	cooper	cooper	PROPN
bjmsr-315	566	6	,	,	PUNCT
bjmsr-315	566	7	r.o	r.o	PROPN
bjmsr-315	566	8	.	.	PROPN
bjmsr-315	566	9	ferguson	ferguson	PROPN
bjmsr-315	566	10	(	(	PUNCT
bjmsr-315	566	11	1955	1955	NUM
bjmsr-315	566	12	)	)	PUNCT
bjmsr-315	566	13	optimal	optimal	ADJ
bjmsr-315	566	14	estimation	estimation	NOUN
bjmsr-315	566	15	of	of	ADP
bjmsr-315	566	16	executive	executive	ADJ
bjmsr-315	566	17	compensation	compensation	NOUN
bjmsr-315	566	18	by	by	ADP
bjmsr-315	566	19	linear	linear	PROPN
bjmsr-315	566	20	programming	programming	NOUN
bjmsr-315	566	21	.	.	PUNCT
bjmsr-315	567	1	manag	manag	PROPN
bjmsr-315	567	2	.	.	PUNCT
bjmsr-315	568	1	sci	sci	PROPN
bjmsr-315	568	2	.	.	PROPN
bjmsr-315	568	3	1	1	NUM
bjmsr-315	568	4	,	,	PUNCT
bjmsr-315	568	5	138	138	NUM
bjmsr-315	568	6	-	-	SYM
bjmsr-315	568	7	151	151	NUM
bjmsr-315	568	8	.	.	PUNCT
bjmsr-315	569	1	e.w	e.w	PROPN
bjmsr-315	569	2	.	.	PROPN
bjmsr-315	569	3	cheney	cheney	PROPN
bjmsr-315	569	4	(	(	PUNCT
bjmsr-315	569	5	1966	1966	NUM
bjmsr-315	569	6	)	)	PUNCT
bjmsr-315	569	7	introduction	introduction	NOUN
bjmsr-315	569	8	to	to	ADP
bjmsr-315	569	9	approximation	approximation	NOUN
bjmsr-315	569	10	theory	theory	NOUN
bjmsr-315	569	11	,	,	PUNCT
bjmsr-315	569	12	mcgraw	mcgraw	PROPN
bjmsr-315	569	13	-	-	PUNCT
bjmsr-315	569	14	hill	hill	PROPN
bjmsr-315	569	15	,	,	PUNCT
bjmsr-315	569	16	new	new	PROPN
bjmsr-315	569	17	york	york	PROPN
bjmsr-315	569	18	.	.	PUNCT
bjmsr-315	570	1	http://www.uib.es/congres/ecopub/ecineq/general.htm	http://www.uib.es/congres/ecopub/ecineq/general.htm	PROPN
bjmsr-315	570	2	http://www.uib.es/congres/ecopub/ecineq/papers/039bidabab.pdf	http://www.uib.es/congres/ecopub/ecineq/papers/039bidabab.pdf	PROPN
bjmsr-315	570	3	http://www.bidabad.com/doc/estimating-lorenz-us.pdf	http://www.bidabad.com/doc/estimating-lorenz-us.pdf	PROPN
bjmsr-315	570	4	http://www.bidabad.com/doc/l1-implied-inequality-index-4.pdf	http://www.bidabad.com/doc/l1-implied-inequality-index-4.pdf	PROPN
bjmsr-315	570	5	http://www.theibfr.com/archive/issn-1941-9589-v3-n2-2008.pdf	http://www.theibfr.com/archive/issn-1941-9589-v3-n2-2008.pdf	PROPN
bjmsr-315	570	6	http://www.bidabad.com/doc/ssrn-id1631861.pdf	http://www.bidabad.com/doc/ssrn-id1631861.pdf	PROPN
bjmsr-315	570	7	copyright	copyright	NOUN
bjmsr-315	570	8	©	©	PROPN
bjmsr-315	570	9	cc	cc	PROPN
bjmsr-315	570	10	-	-	PUNCT
bjmsr-315	570	11	by	by	ADP
bjmsr-315	570	12	-	-	PUNCT
bjmsr-315	570	13	nc	nc	PROPN
bjmsr-315	570	14	2019	2019	NUM
bjmsr-315	570	15	,	,	PUNCT
bjmsr-315	570	16	bjmsr	bjmsr	PROPN
bjmsr-315	570	17	www.cribfb.com/journal/index.php/bjmsr	www.cribfb.com/journal/index.php/bjmsr	X
bjmsr-315	570	18	bangladesh	bangladesh	PROPN
bjmsr-315	570	19	journal	journal	PROPN
bjmsr-315	570	20	of	of	ADP
bjmsr-315	570	21	multidisciplinary	multidisciplinary	ADJ
bjmsr-315	570	22	scientific	scientific	ADJ
bjmsr-315	570	23	research	research	NOUN
bjmsr-315	570	24	vol	vol	NOUN
bjmsr-315	570	25	.	.	PROPN
bjmsr-315	571	1	1	1	NUM
bjmsr-315	571	2	,	,	PUNCT
bjmsr-315	571	3	no	no	INTJ
bjmsr-315	571	4	.	.	NOUN
bjmsr-315	571	5	1	1	NUM
bjmsr-315	571	6	;	;	PUNCT
bjmsr-315	571	7	2019	2019	NUM
bjmsr-315	571	8	64	64	NUM
bjmsr-315	571	9	f.h	f.h	PROPN
bjmsr-315	571	10	.	.	PROPN
bjmsr-315	571	11	clarke	clarke	PROPN
bjmsr-315	571	12	(	(	PUNCT
bjmsr-315	571	13	1983	1983	NUM
bjmsr-315	571	14	)	)	PUNCT
bjmsr-315	571	15	optimization	optimization	NOUN
bjmsr-315	571	16	and	and	CCONJ
bjmsr-315	571	17	nonsmooth	nonsmooth	ADJ
bjmsr-315	571	18	analysis	analysis	NOUN
bjmsr-315	571	19	.	.	PUNCT
bjmsr-315	572	1	wiley	wiley	PROPN
bjmsr-315	572	2	,	,	PUNCT
bjmsr-315	572	3	new	new	PROPN
bjmsr-315	572	4	york	york	PROPN
bjmsr-315	572	5	a.r	a.r	PROPN
bjmsr-315	572	6	.	.	PROPN
bjmsr-315	572	7	conn	conn	PROPN
bjmsr-315	572	8	(	(	PUNCT
bjmsr-315	572	9	1976	1976	NUM
bjmsr-315	572	10	)	)	PUNCT
bjmsr-315	572	11	linear	linear	NOUN
bjmsr-315	572	12	programming	programming	NOUN
bjmsr-315	572	13	via	via	ADP
bjmsr-315	572	14	a	a	DET
bjmsr-315	572	15	nondifferentiable	nondifferentiable	ADJ
bjmsr-315	572	16	penalty	penalty	NOUN
bjmsr-315	572	17	function	function	NOUN
bjmsr-315	572	18	.	.	PUNCT
bjmsr-315	573	1	siam	siam	PROPN
bjmsr-315	573	2	j.	j.	PROPN
bjmsr-315	573	3	numer	numer	PROPN
bjmsr-315	573	4	.	.	PUNCT
bjmsr-315	574	1	anal	anal	PROPN
bjmsr-315	574	2	.	.	PROPN
bjmsr-315	574	3	,	,	PUNCT
bjmsr-315	574	4	13	13	NUM
bjmsr-315	574	5	,	,	PUNCT
bjmsr-315	574	6	145	145	NUM
bjmsr-315	574	7	-	-	SYM
bjmsr-315	574	8	154	154	NUM
bjmsr-315	574	9	.	.	PUNCT
bjmsr-315	575	1	d.c	d.c	PROPN
bjmsr-315	575	2	.	.	PUNCT
bjmsr-315	575	3	crocker	crocker	PROPN
bjmsr-315	575	4	(	(	PUNCT
bjmsr-315	575	5	1969	1969	NUM
bjmsr-315	575	6	)	)	PUNCT
bjmsr-315	575	7	linear	linear	NOUN
bjmsr-315	575	8	programming	programming	NOUN
bjmsr-315	575	9	technique	technique	NOUN
bjmsr-315	575	10	in	in	ADP
bjmsr-315	575	11	regression	regression	NOUN
bjmsr-315	575	12	analysis	analysis	NOUN
bjmsr-315	575	13	,	,	PUNCT
bjmsr-315	575	14	the	the	DET
bjmsr-315	575	15	hidden	hidden	ADJ
bjmsr-315	575	16	danger	danger	NOUN
bjmsr-315	575	17	.	.	PUNCT
bjmsr-315	576	1	a.i.e.e	a.i.e.e	PROPN
bjmsr-315	576	2	.	.	PROPN
bjmsr-315	576	3	trans	trans	PROPN
bjmsr-315	576	4	.	.	PROPN
bjmsr-315	576	5	,	,	PUNCT
bjmsr-315	576	6	1	1	NUM
bjmsr-315	576	7	,	,	PUNCT
bjmsr-315	576	8	112	112	NUM
bjmsr-315	576	9	-	-	SYM
bjmsr-315	576	10	126	126	NUM
bjmsr-315	576	11	.	.	PUNCT
bjmsr-315	577	1	m.	m.	NOUN
bjmsr-315	577	2	davies	davy	NOUN
bjmsr-315	577	3	(	(	PUNCT
bjmsr-315	577	4	1976	1976	NUM
bjmsr-315	577	5	)	)	PUNCT
bjmsr-315	577	6	linear	linear	ADJ
bjmsr-315	577	7	approximation	approximation	NOUN
bjmsr-315	577	8	using	use	VERB
bjmsr-315	577	9	the	the	DET
bjmsr-315	577	10	criterion	criterion	NOUN
bjmsr-315	577	11	of	of	ADP
bjmsr-315	577	12	least	least	ADJ
bjmsr-315	577	13	total	total	ADJ
bjmsr-315	577	14	deviations	deviation	NOUN
bjmsr-315	577	15	.	.	PUNCT
bjmsr-315	578	1	j.	j.	PROPN
bjmsr-315	578	2	roy	roy	PROPN
bjmsr-315	578	3	.	.	PROPN
bjmsr-315	578	4	stat	stat	PROPN
bjmsr-315	578	5	.	.	PUNCT
bjmsr-315	579	1	soc	soc	PROPN
bjmsr-315	579	2	.	.	PUNCT
bjmsr-315	580	1	b29	b29	NOUN
bjmsr-315	580	2	,	,	PUNCT
bjmsr-315	580	3	101	101	NUM
bjmsr-315	580	4	-	-	SYM
bjmsr-315	580	5	109	109	NUM
bjmsr-315	580	6	.	.	PUNCT
bjmsr-315	581	1	t.e	t.e	PROPN
bjmsr-315	581	2	.	.	PROPN
bjmsr-315	581	3	dielman	dielman	PROPN
bjmsr-315	581	4	(	(	PUNCT
bjmsr-315	581	5	1984	1984	NUM
bjmsr-315	581	6	)	)	PUNCT
bjmsr-315	581	7	least	least	ADJ
bjmsr-315	581	8	absolute	absolute	ADJ
bjmsr-315	581	9	value	value	NOUN
bjmsr-315	581	10	estimation	estimation	NOUN
bjmsr-315	581	11	in	in	ADP
bjmsr-315	581	12	regression	regression	NOUN
bjmsr-315	581	13	models	model	NOUN
bjmsr-315	581	14	:	:	PUNCT
bjmsr-315	581	15	an	an	DET
bjmsr-315	581	16	annotated	annotate	VERB
bjmsr-315	581	17	bibliography	bibliography	NOUN
bjmsr-315	581	18	.	.	PUNCT
bjmsr-315	582	1	comm	comm	NOUN
bjmsr-315	582	2	.	.	PUNCT
bjmsr-315	583	1	stat	stat	PROPN
bjmsr-315	583	2	.	.	PUNCT
bjmsr-315	584	1	13	13	NUM
bjmsr-315	584	2	,	,	PUNCT
bjmsr-315	584	3	51341	51341	NUM
bjmsr-315	584	4	.	.	PUNCT
bjmsr-315	585	1	t.	t.	PROPN
bjmsr-315	585	2	dielman	dielman	PROPN
bjmsr-315	585	3	,	,	PUNCT
bjmsr-315	585	4	r.	r.	PROPN
bjmsr-315	585	5	pfaffenberger	pfaffenberger	ADV
bjmsr-315	585	6	(	(	PUNCT
bjmsr-315	585	7	1982	1982	NUM
bjmsr-315	585	8	)	)	PUNCT
bjmsr-315	585	9	lav	lav	NOUN
bjmsr-315	585	10	(	(	PUNCT
bjmsr-315	585	11	least	least	ADJ
bjmsr-315	585	12	absolute	absolute	ADJ
bjmsr-315	585	13	value	value	NOUN
bjmsr-315	585	14	)	)	PUNCT
bjmsr-315	585	15	estimation	estimation	NOUN
bjmsr-315	585	16	in	in	ADP
bjmsr-315	585	17	linear	linear	PROPN
bjmsr-315	585	18	regression	regression	NOUN
bjmsr-315	585	19	:	:	PUNCT
bjmsr-315	585	20	a	a	DET
bjmsr-315	585	21	review	review	NOUN
bjmsr-315	585	22	,	,	PUNCT
bjmsr-315	585	23	tims	tims	PROPN
bjmsr-315	585	24	studies	study	NOUN
bjmsr-315	585	25	in	in	ADP
bjmsr-315	585	26	the	the	DET
bjmsr-315	585	27	manag	manag	NOUN
bjmsr-315	585	28	.	.	PUNCT
bjmsr-315	586	1	sci	sci	PROPN
bjmsr-315	586	2	.	.	PROPN
bjmsr-315	586	3	,19	,19	PROPN
bjmsr-315	586	4	,	,	PUNCT
bjmsr-315	586	5	31	31	NUM
bjmsr-315	586	6	-	-	SYM
bjmsr-315	586	7	52	52	NUM
bjmsr-315	586	8	.	.	PUNCT
bjmsr-315	587	1	t.	t.	PROPN
bjmsr-315	587	2	dielman	dielman	PROPN
bjmsr-315	587	3	,	,	PUNCT
bjmsr-315	587	4	r.	r.	PROPN
bjmsr-315	587	5	pfaffenberger	pfaffenberger	ADV
bjmsr-315	587	6	(	(	PUNCT
bjmsr-315	587	7	1984	1984	NUM
bjmsr-315	587	8	)	)	PUNCT
bjmsr-315	587	9	computational	computational	ADJ
bjmsr-315	587	10	algorithms	algorithm	NOUN
bjmsr-315	587	11	for	for	ADP
bjmsr-315	587	12	calculating	calculate	VERB
bjmsr-315	587	13	least	least	ADJ
bjmsr-315	587	14	absolute	absolute	ADJ
bjmsr-315	587	15	value	value	NOUN
bjmsr-315	587	16	and	and	CCONJ
bjmsr-315	587	17	chebyshev	chebyshev	NOUN
bjmsr-315	587	18	estimates	estimate	NOUN
bjmsr-315	587	19	for	for	ADP
bjmsr-315	587	20	multiple	multiple	ADJ
bjmsr-315	587	21	regression	regression	NOUN
bjmsr-315	587	22	.	.	PUNCT
bjmsr-315	588	1	amer	amer	PROPN
bjmsr-315	588	2	.	.	PUNCT
bjmsr-315	589	1	j.	j.	PROPN
bjmsr-315	589	2	math	math	PROPN
bjmsr-315	589	3	.	.	PUNCT
bjmsr-315	590	1	manag	manag	PROPN
bjmsr-315	590	2	.	.	PUNCT
bjmsr-315	591	1	sci	sci	PROPN
bjmsr-315	591	2	.	.	PROPN
bjmsr-315	591	3	,	,	PUNCT
bjmsr-315	591	4	4	4	NUM
bjmsr-315	591	5	,	,	PUNCT
bjmsr-315	591	6	169	169	NUM
bjmsr-315	591	7	-	-	SYM
bjmsr-315	591	8	197	197	NUM
bjmsr-315	591	9	.	.	PUNCT
bjmsr-315	592	1	p.j	p.j	PROPN
bjmsr-315	592	2	.	.	PROPN
bjmsr-315	592	3	dhrymes	dhryme	NOUN
bjmsr-315	592	4	(	(	PUNCT
bjmsr-315	592	5	1978	1978	NUM
bjmsr-315	592	6	)	)	PUNCT
bjmsr-315	592	7	mathematics	mathematic	NOUN
bjmsr-315	592	8	for	for	ADP
bjmsr-315	592	9	econometrics	econometric	NOUN
bjmsr-315	592	10	.	.	PUNCT
bjmsr-315	593	1	springer	springer	NOUN
bjmsr-315	593	2	-	-	PUNCT
bjmsr-315	593	3	verlag	verlag	PROPN
bjmsr-315	593	4	,	,	PUNCT
bjmsr-315	593	5	new	new	PROPN
bjmsr-315	593	6	york	york	PROPN
bjmsr-315	593	7	.	.	PUNCT
bjmsr-315	594	1	y.	y.	PROPN
bjmsr-315	594	2	dodge	dodge	PROPN
bjmsr-315	594	3	(	(	PUNCT
bjmsr-315	594	4	1987	1987	NUM
bjmsr-315	594	5	)	)	PUNCT
bjmsr-315	594	6	an	an	DET
bjmsr-315	594	7	introduction	introduction	NOUN
bjmsr-315	594	8	to	to	ADP
bjmsr-315	594	9	statistical	statistical	ADJ
bjmsr-315	594	10	data	datum	NOUN
bjmsr-315	594	11	analysis	analysis	NOUN
bjmsr-315	594	12	l1	l1	PROPN
bjmsr-315	594	13	-	-	PUNCT
bjmsr-315	594	14	norm	norm	NOUN
bjmsr-315	594	15	based	base	VERB
bjmsr-315	594	16	.	.	PUNCT
bjmsr-315	595	1	in	in	ADP
bjmsr-315	595	2	y.	y.	PROPN
bjmsr-315	595	3	dodge	dodge	PROPN
bjmsr-315	595	4	(	(	PUNCT
bjmsr-315	595	5	ed	ed	NOUN
bjmsr-315	595	6	.	.	PUNCT
bjmsr-315	595	7	)	)	PUNCT
bjmsr-315	595	8	statistical	statistical	ADJ
bjmsr-315	595	9	data	datum	NOUN
bjmsr-315	595	10	analysis	analysis	NOUN
bjmsr-315	595	11	based	base	VERB
bjmsr-315	595	12	on	on	ADP
bjmsr-315	595	13	the	the	DET
bjmsr-315	595	14	l1	l1	PROPN
bjmsr-315	595	15	norm	norm	NOUN
bjmsr-315	595	16	and	and	CCONJ
bjmsr-315	595	17	related	related	ADJ
bjmsr-315	595	18	methods	method	NOUN
bjmsr-315	595	19	.	.	PUNCT
bjmsr-315	596	1	north	north	NOUN
bjmsr-315	596	2	-	-	PUNCT
bjmsr-315	596	3	holland	holland	PROPN
bjmsr-315	596	4	.	.	PUNCT
bjmsr-315	597	1	reprinted	reprint	VERB
bjmsr-315	597	2	in	in	ADP
bjmsr-315	597	3	csda	csda	NOUN
bjmsr-315	597	4	,	,	PUNCT
bjmsr-315	597	5	5	5	NUM
bjmsr-315	597	6	,	,	PUNCT
bjmsr-315	597	7	239	239	NUM
bjmsr-315	597	8	-	-	SYM
bjmsr-315	597	9	254	254	NUM
bjmsr-315	597	10	.	.	PUNCT
bjmsr-315	598	1	a.f	a.f	PROPN
bjmsr-315	598	2	.	.	PROPN
bjmsr-315	598	3	dufton	dufton	PROPN
bjmsr-315	598	4	(	(	PUNCT
bjmsr-315	598	5	1928	1928	NUM
bjmsr-315	598	6	)	)	PUNCT
bjmsr-315	598	7	correlation	correlation	NOUN
bjmsr-315	598	8	.	.	PUNCT
bjmsr-315	599	1	nature	nature	NOUN
bjmsr-315	599	2	,	,	PUNCT
bjmsr-315	599	3	121	121	NUM
bjmsr-315	599	4	,	,	PUNCT
bjmsr-315	599	5	866	866	NUM
bjmsr-315	599	6	.	.	PUNCT
bjmsr-315	600	1	f.y	f.y	PROPN
bjmsr-315	600	2	.	.	PROPN
bjmsr-315	600	3	edgeworth	edgeworth	PROPN
bjmsr-315	600	4	(	(	PUNCT
bjmsr-315	600	5	1883	1883	NUM
bjmsr-315	600	6	)	)	PUNCT
bjmsr-315	600	7	the	the	DET
bjmsr-315	600	8	method	method	NOUN
bjmsr-315	600	9	of	of	ADP
bjmsr-315	600	10	least	least	ADJ
bjmsr-315	600	11	squares	square	NOUN
bjmsr-315	600	12	.	.	PUNCT
bjmsr-315	601	1	philosophical	philosophical	ADJ
bjmsr-315	601	2	magazine	magazine	NOUN
bjmsr-315	601	3	,	,	PUNCT
bjmsr-315	601	4	16	16	NUM
bjmsr-315	601	5	,	,	PUNCT
bjmsr-315	601	6	360	360	NUM
bjmsr-315	601	7	-	-	SYM
bjmsr-315	601	8	375	375	NUM
bjmsr-315	601	9	.	.	PUNCT
bjmsr-315	602	1	f.y	f.y	PROPN
bjmsr-315	602	2	.	.	PROPN
bjmsr-315	602	3	edgeworth	edgeworth	PROPN
bjmsr-315	602	4	(	(	PUNCT
bjmsr-315	602	5	1887a	1887a	NUM
bjmsr-315	602	6	)	)	PUNCT
bjmsr-315	602	7	on	on	ADP
bjmsr-315	602	8	observations	observation	NOUN
bjmsr-315	602	9	relating	relate	VERB
bjmsr-315	602	10	to	to	ADP
bjmsr-315	602	11	several	several	ADJ
bjmsr-315	602	12	quantities	quantity	NOUN
bjmsr-315	602	13	.	.	PUNCT
bjmsr-315	603	1	hermathena	hermathena	NOUN
bjmsr-315	603	2	,	,	PUNCT
bjmsr-315	603	3	6	6	NUM
bjmsr-315	603	4	,	,	PUNCT
bjmsr-315	603	5	279	279	NUM
bjmsr-315	603	6	-	-	SYM
bjmsr-315	603	7	285	285	NUM
bjmsr-315	603	8	.	.	PUNCT
bjmsr-315	604	1	f.y	f.y	PROPN
bjmsr-315	604	2	.	.	PROPN
bjmsr-315	604	3	edgeworth	edgeworth	PROPN
bjmsr-315	604	4	(	(	PUNCT
bjmsr-315	604	5	1887b	1887b	NUM
bjmsr-315	604	6	)	)	PUNCT
bjmsr-315	604	7	a	a	DET
bjmsr-315	604	8	new	new	ADJ
bjmsr-315	604	9	method	method	NOUN
bjmsr-315	604	10	of	of	ADP
bjmsr-315	604	11	reducing	reduce	VERB
bjmsr-315	604	12	observations	observation	NOUN
bjmsr-315	604	13	relating	relate	VERB
bjmsr-315	604	14	to	to	ADP
bjmsr-315	604	15	several	several	ADJ
bjmsr-315	604	16	quantities	quantity	NOUN
bjmsr-315	604	17	.	.	PUNCT
bjmsr-315	605	1	philosophical	philosophical	ADJ
bjmsr-315	605	2	magazine	magazine	NOUN
bjmsr-315	605	3	,	,	PUNCT
bjmsr-315	605	4	24	24	NUM
bjmsr-315	605	5	,	,	PUNCT
bjmsr-315	605	6	222	222	NUM
bjmsr-315	605	7	-	-	SYM
bjmsr-315	605	8	223	223	NUM
bjmsr-315	605	9	.	.	PUNCT
bjmsr-315	606	1	f.y	f.y	PROPN
bjmsr-315	606	2	.	.	PROPN
bjmsr-315	606	3	edgeworth	edgeworth	PROPN
bjmsr-315	606	4	(	(	PUNCT
bjmsr-315	606	5	1888	1888	NUM
bjmsr-315	606	6	)	)	PUNCT
bjmsr-315	606	7	on	on	ADP
bjmsr-315	606	8	a	a	DET
bjmsr-315	606	9	new	new	ADJ
bjmsr-315	606	10	method	method	NOUN
bjmsr-315	606	11	of	of	ADP
bjmsr-315	606	12	reducing	reduce	VERB
bjmsr-315	606	13	observation	observation	NOUN
bjmsr-315	606	14	relating	relate	VERB
bjmsr-315	606	15	to	to	ADP
bjmsr-315	606	16	several	several	ADJ
bjmsr-315	606	17	quantities	quantity	NOUN
bjmsr-315	606	18	.	.	PUNCT
bjmsr-315	607	1	philosophical	philosophical	ADJ
bjmsr-315	607	2	magazine	magazine	NOUN
bjmsr-315	607	3	,	,	PUNCT
bjmsr-315	607	4	25	25	NUM
bjmsr-315	607	5	,	,	PUNCT
bjmsr-315	607	6	184	184	NUM
bjmsr-315	607	7	-	-	SYM
bjmsr-315	607	8	191	191	NUM
bjmsr-315	607	9	.	.	PUNCT
bjmsr-315	608	1	f.y	f.y	PROPN
bjmsr-315	608	2	.	.	PROPN
bjmsr-315	608	3	edgeworth	edgeworth	PROPN
bjmsr-315	608	4	(	(	PUNCT
bjmsr-315	608	5	1902	1902	NUM
bjmsr-315	608	6	)	)	PUNCT
bjmsr-315	608	7	method	method	NOUN
bjmsr-315	608	8	of	of	ADP
bjmsr-315	608	9	representing	represent	VERB
bjmsr-315	608	10	statistics	statistic	NOUN
bjmsr-315	608	11	of	of	ADP
bjmsr-315	608	12	wage	wage	NOUN
bjmsr-315	608	13	and	and	CCONJ
bjmsr-315	608	14	other	other	ADJ
bjmsr-315	608	15	groups	group	NOUN
bjmsr-315	608	16	not	not	PART
bjmsr-315	608	17	fulfilling	fulfil	VERB
bjmsr-315	608	18	the	the	DET
bjmsr-315	608	19	normal	normal	ADJ
bjmsr-315	608	20	law	law	NOUN
bjmsr-315	608	21	of	of	ADP
bjmsr-315	608	22	error	error	NOUN
bjmsr-315	608	23	,	,	PUNCT
bjmsr-315	608	24	i	i	PRON
bjmsr-315	608	25	:	:	PUNCT
bjmsr-315	608	26	mathematical	mathematical	ADJ
bjmsr-315	608	27	considerations	consideration	NOUN
bjmsr-315	608	28	.	.	PUNCT
bjmsr-315	609	1	j.	j.	PROPN
bjmsr-315	609	2	roy	roy	PROPN
bjmsr-315	609	3	.	.	PROPN
bjmsr-315	609	4	stat	stat	PROPN
bjmsr-315	609	5	.	.	PUNCT
bjmsr-315	610	1	soc	soc	PROPN
bjmsr-315	610	2	.	.	PROPN
bjmsr-315	610	3	,	,	PUNCT
bjmsr-315	610	4	65	65	NUM
bjmsr-315	610	5	,	,	PUNCT
bjmsr-315	610	6	325	325	NUM
bjmsr-315	610	7	-	-	SYM
bjmsr-315	610	8	331	331	NUM
bjmsr-315	610	9	.	.	PUNCT
bjmsr-315	611	1	f.y	f.y	PROPN
bjmsr-315	611	2	.	.	PROPN
bjmsr-315	611	3	edgeworth	edgeworth	PROPN
bjmsr-315	611	4	(	(	PUNCT
bjmsr-315	611	5	1923	1923	NUM
bjmsr-315	611	6	)	)	PUNCT
bjmsr-315	611	7	on	on	ADP
bjmsr-315	611	8	the	the	DET
bjmsr-315	611	9	use	use	NOUN
bjmsr-315	611	10	of	of	ADP
bjmsr-315	611	11	medians	median	NOUN
bjmsr-315	611	12	for	for	ADP
bjmsr-315	611	13	reducing	reduce	VERB
bjmsr-315	611	14	observations	observation	NOUN
bjmsr-315	611	15	relating	relate	VERB
bjmsr-315	611	16	to	to	ADP
bjmsr-315	611	17	several	several	ADJ
bjmsr-315	611	18	quantities	quantity	NOUN
bjmsr-315	611	19	.	.	PUNCT
bjmsr-315	612	1	philosophical	philosophical	ADJ
bjmsr-315	612	2	magazine	magazine	NOUN
bjmsr-315	612	3	,	,	PUNCT
bjmsr-315	612	4	6th	6th	ADJ
bjmsr-315	612	5	series	series	NOUN
bjmsr-315	612	6	,	,	PUNCT
bjmsr-315	612	7	46	46	NUM
bjmsr-315	612	8	,	,	PUNCT
bjmsr-315	612	9	1074	1074	NUM
bjmsr-315	612	10	-	-	SYM
bjmsr-315	612	11	1088	1088	NUM
bjmsr-315	612	12	.	.	PUNCT
bjmsr-315	613	1	eisenhart	eisenhart	NOUN
bjmsr-315	613	2	(	(	PUNCT
bjmsr-315	613	3	1961	1961	NUM
bjmsr-315	613	4	)	)	PUNCT
bjmsr-315	613	5	boscovich	boscovich	NOUN
bjmsr-315	613	6	and	and	CCONJ
bjmsr-315	613	7	the	the	DET
bjmsr-315	613	8	combination	combination	NOUN
bjmsr-315	613	9	of	of	ADP
bjmsr-315	613	10	observations	observation	NOUN
bjmsr-315	613	11	.	.	PUNCT
bjmsr-315	614	1	ch	ch	NOUN
bjmsr-315	614	2	.	.	PROPN
bjmsr-315	614	3	9	9	NUM
bjmsr-315	614	4	of	of	ADP
bjmsr-315	614	5	whyte	whyte	PROPN
bjmsr-315	614	6	(	(	PUNCT
bjmsr-315	614	7	1961	1961	NUM
bjmsr-315	614	8	,	,	PUNCT
bjmsr-315	614	9	200	200	NUM
bjmsr-315	614	10	-	-	SYM
bjmsr-315	614	11	212	212	NUM
bjmsr-315	614	12	)	)	PUNCT
bjmsr-315	614	13	reprinted	reprint	VERB
bjmsr-315	614	14	in	in	ADP
bjmsr-315	614	15	kendall	kendall	PROPN
bjmsr-315	614	16	and	and	CCONJ
bjmsr-315	614	17	plackett	plackett	PROPN
bjmsr-315	614	18	(	(	PUNCT
bjmsr-315	614	19	1977	1977	NUM
bjmsr-315	614	20	)	)	PUNCT
bjmsr-315	614	21	studies	study	NOUN
bjmsr-315	614	22	in	in	ADP
bjmsr-315	614	23	the	the	DET
bjmsr-315	614	24	history	history	NOUN
bjmsr-315	614	25	of	of	ADP
bjmsr-315	614	26	statistics	statistic	NOUN
bjmsr-315	614	27	and	and	CCONJ
bjmsr-315	614	28	probability	probability	NOUN
bjmsr-315	614	29	,	,	PUNCT
bjmsr-315	614	30	vol.ii	vol.ii	PROPN
bjmsr-315	614	31	,	,	PUNCT
bjmsr-315	614	32	charles	charles	PROPN
bjmsr-315	614	33	griffin	griffin	PROPN
bjmsr-315	614	34	and	and	CCONJ
bjmsr-315	614	35	co.	co.	PROPN
bjmsr-315	614	36	ltd	ltd	PROPN
bjmsr-315	614	37	.	.	PROPN
bjmsr-315	614	38	,	,	PUNCT
bjmsr-315	614	39	high	high	PROPN
bjmsr-315	614	40	wycombe	wycombe	PROPN
bjmsr-315	614	41	88	88	NUM
bjmsr-315	614	42	-	-	SYM
bjmsr-315	614	43	100	100	NUM
bjmsr-315	614	44	.	.	PUNCT
bjmsr-315	615	1	j.e	j.e	PROPN
bjmsr-315	615	2	.	.	PROPN
bjmsr-315	615	3	estienne	estienne	PROPN
bjmsr-315	615	4	(	(	PUNCT
bjmsr-315	615	5	1926	1926	NUM
bjmsr-315	615	6	-	-	SYM
bjmsr-315	615	7	28	28	NUM
bjmsr-315	615	8	)	)	PUNCT
bjmsr-315	615	9	introduction	introduction	NOUN
bjmsr-315	615	10	a	a	DET
bjmsr-315	615	11	une	une	PROPN
bjmsr-315	615	12	theorie	theorie	PROPN
bjmsr-315	615	13	rationnelle	rationnelle	PROPN
bjmsr-315	615	14	des	des	PROPN
bjmsr-315	615	15	erreurs	erreurs	PROPN
bjmsr-315	615	16	d'observation	d'observation	PROPN
bjmsr-315	615	17	.	.	PUNCT
bjmsr-315	616	1	revue	revue	PROPN
bjmsr-315	616	2	d'artillerie	d'artillerie	PROPN
bjmsr-315	616	3	97(1926	97(1926	NUM
bjmsr-315	616	4	)	)	PUNCT
bjmsr-315	616	5	,	,	PUNCT
bjmsr-315	616	6	421	421	NUM
bjmsr-315	616	7	-	-	SYM
bjmsr-315	616	8	441	441	NUM
bjmsr-315	616	9	;	;	PUNCT
bjmsr-315	616	10	98(1928	98(1928	NUM
bjmsr-315	616	11	)	)	PUNCT
bjmsr-315	616	12	,	,	PUNCT
bjmsr-315	616	13	542	542	NUM
bjmsr-315	616	14	-	-	SYM
bjmsr-315	616	15	562	562	NUM
bjmsr-315	616	16	;	;	PUNCT
bjmsr-315	616	17	100(1927	100(1927	NUM
bjmsr-315	616	18	)	)	PUNCT
bjmsr-315	616	19	,	,	PUNCT
bjmsr-315	616	20	471	471	NUM
bjmsr-315	616	21	-	-	SYM
bjmsr-315	616	22	487	487	NUM
bjmsr-315	616	23	.	.	PUNCT
bjmsr-315	617	1	r.c	r.c	PROPN
bjmsr-315	617	2	.	.	PROPN
bjmsr-315	617	3	fair	fair	PROPN
bjmsr-315	617	4	(	(	PUNCT
bjmsr-315	617	5	1974	1974	NUM
bjmsr-315	617	6	)	)	PUNCT
bjmsr-315	617	7	on	on	ADP
bjmsr-315	617	8	the	the	DET
bjmsr-315	617	9	robust	robust	ADJ
bjmsr-315	617	10	estimation	estimation	NOUN
bjmsr-315	617	11	of	of	ADP
bjmsr-315	617	12	econometric	econometric	ADJ
bjmsr-315	617	13	models	model	NOUN
bjmsr-315	617	14	.	.	PUNCT
bjmsr-315	618	1	ann	ann	PROPN
bjmsr-315	618	2	.	.	PROPN
bjmsr-315	618	3	econ	econ	PROPN
bjmsr-315	618	4	.	.	PUNCT
bjmsr-315	618	5	soc	soc	PROPN
bjmsr-315	618	6	.	.	PUNCT
bjmsr-315	619	1	measurement	measurement	PROPN
bjmsr-315	619	2	,	,	PUNCT
bjmsr-315	619	3	3	3	NUM
bjmsr-315	619	4	,	,	PUNCT
bjmsr-315	619	5	667	667	NUM
bjmsr-315	619	6	-	-	SYM
bjmsr-315	619	7	77	77	NUM
bjmsr-315	619	8	.	.	PUNCT
bjmsr-315	620	1	r.w	r.w	PROPN
bjmsr-315	620	2	.	.	PROPN
bjmsr-315	620	3	farebrother	farebrother	PROPN
bjmsr-315	620	4	(	(	PUNCT
bjmsr-315	620	5	1987b	1987b	NUM
bjmsr-315	620	6	)	)	PUNCT
bjmsr-315	620	7	the	the	DET
bjmsr-315	620	8	historical	historical	ADJ
bjmsr-315	620	9	development	development	NOUN
bjmsr-315	620	10	of	of	ADP
bjmsr-315	620	11	the	the	DET
bjmsr-315	620	12	l1	l1	PROPN
bjmsr-315	620	13	and	and	CCONJ
bjmsr-315	620	14	l∞	l∞	NOUN
bjmsr-315	620	15	estimation	estimation	NOUN
bjmsr-315	620	16	procedures	procedure	NOUN
bjmsr-315	620	17	.	.	PUNCT
bjmsr-315	621	1	in	in	ADP
bjmsr-315	621	2	y.	y.	PROPN
bjmsr-315	621	3	dodge	dodge	PROPN
bjmsr-315	621	4	(	(	PUNCT
bjmsr-315	621	5	ed	ed	NOUN
bjmsr-315	621	6	.	.	PUNCT
bjmsr-315	621	7	)	)	PUNCT
bjmsr-315	621	8	statistical	statistical	ADJ
bjmsr-315	621	9	data	datum	NOUN
bjmsr-315	621	10	analysis	analysis	NOUN
bjmsr-315	621	11	based	base	VERB
bjmsr-315	621	12	on	on	ADP
bjmsr-315	621	13	the	the	DET
bjmsr-315	621	14	l1	l1	PROPN
bjmsr-315	621	15	norm	norm	NOUN
bjmsr-315	621	16	and	and	CCONJ
bjmsr-315	621	17	related	related	ADJ
bjmsr-315	621	18	methods	method	NOUN
bjmsr-315	621	19	.	.	PUNCT
bjmsr-315	622	1	north	north	NOUN
bjmsr-315	622	2	-	-	PUNCT
bjmsr-315	622	3	holland	holland	PROPN
bjmsr-315	622	4	.	.	PUNCT
bjmsr-315	623	1	37	37	NUM
bjmsr-315	623	2	-	-	SYM
bjmsr-315	623	3	64	64	NUM
bjmsr-315	623	4	.	.	PUNCT
bjmsr-315	624	1	r.w	r.w	PROPN
bjmsr-315	624	2	.	.	PROPN
bjmsr-315	624	3	farebrother	farebrother	PROPN
bjmsr-315	624	4	(	(	PUNCT
bjmsr-315	624	5	1987c	1987c	NUM
bjmsr-315	624	6	)	)	PUNCT
bjmsr-315	624	7	a	a	DET
bjmsr-315	624	8	simple	simple	ADJ
bjmsr-315	624	9	recursive	recursive	ADJ
bjmsr-315	624	10	procedure	procedure	NOUN
bjmsr-315	624	11	for	for	ADP
bjmsr-315	624	12	the	the	DET
bjmsr-315	624	13	l1	l1	PROPN
bjmsr-315	624	14	norm	norm	NOUN
bjmsr-315	624	15	fitting	fitting	ADJ
bjmsr-315	624	16	of	of	ADP
bjmsr-315	624	17	a	a	DET
bjmsr-315	624	18	straight	straight	ADJ
bjmsr-315	624	19	line	line	NOUN
bjmsr-315	624	20	.	.	PUNCT
bjmsr-315	625	1	work	work	NOUN
bjmsr-315	625	2	.	.	PUNCT
bjmsr-315	626	1	pap	pap	NOUN
bjmsr-315	626	2	.	.	PROPN
bjmsr-315	626	3	,	,	PUNCT
bjmsr-315	626	4	dept	dept	PROPN
bjmsr-315	626	5	.	.	PROPN
bjmsr-315	626	6	of	of	ADP
bjmsr-315	626	7	econometrics	econometric	NOUN
bjmsr-315	626	8	and	and	CCONJ
bjmsr-315	626	9	social	social	ADJ
bjmsr-315	626	10	stat	stat	NOUN
bjmsr-315	626	11	.	.	PUNCT
bjmsr-315	627	1	university	university	PROPN
bjmsr-315	627	2	of	of	ADP
bjmsr-315	627	3	manchester	manchester	PROPN
bjmsr-315	627	4	,	,	PUNCT
bjmsr-315	627	5	manchester	manchester	PROPN
bjmsr-315	627	6	,	,	PUNCT
bjmsr-315	627	7	m13	m13	ADJ
bjmsr-315	627	8	9pl	9pl	NOUN
bjmsr-315	627	9	,	,	PUNCT
bjmsr-315	627	10	uk	uk	PROPN
bjmsr-315	627	11	.	.	PROPN
bjmsr-315	627	12	w.d	w.d	PROPN
bjmsr-315	627	13	.	.	PROPN
bjmsr-315	627	14	fisher	fisher	PROPN
bjmsr-315	627	15	(	(	PUNCT
bjmsr-315	627	16	1961	1961	NUM
bjmsr-315	627	17	)	)	PUNCT
bjmsr-315	627	18	a	a	DET
bjmsr-315	627	19	note	note	NOUN
bjmsr-315	627	20	on	on	ADP
bjmsr-315	627	21	curve	curve	NOUN
bjmsr-315	627	22	fitting	fit	VERB
bjmsr-315	627	23	with	with	ADP
bjmsr-315	627	24	minimum	minimum	ADJ
bjmsr-315	627	25	deviations	deviation	NOUN
bjmsr-315	627	26	by	by	ADP
bjmsr-315	627	27	linear	linear	PROPN
bjmsr-315	627	28	programming	programming	NOUN
bjmsr-315	627	29	.	.	PUNCT
bjmsr-315	628	1	jasa	jasa	PROPN
bjmsr-315	628	2	,	,	PUNCT
bjmsr-315	628	3	11	11	NUM
bjmsr-315	628	4	,	,	PUNCT
bjmsr-315	628	5	359	359	NUM
bjmsr-315	628	6	-	-	SYM
bjmsr-315	628	7	362	362	NUM
bjmsr-315	628	8	.	.	PUNCT
bjmsr-315	629	1	copyright	copyright	NOUN
bjmsr-315	629	2	©	©	PROPN
bjmsr-315	629	3	cc	cc	PROPN
bjmsr-315	629	4	-	-	PUNCT
bjmsr-315	629	5	by	by	ADP
bjmsr-315	629	6	-	-	PUNCT
bjmsr-315	629	7	nc	nc	PROPN
bjmsr-315	629	8	2019	2019	NUM
bjmsr-315	629	9	,	,	PUNCT
bjmsr-315	629	10	bjmsr	bjmsr	PROPN
bjmsr-315	629	11	www.cribfb.com/journal/index.php/bjmsr	www.cribfb.com/journal/index.php/bjmsr	X
bjmsr-315	630	1	bangladesh	bangladesh	PROPN
bjmsr-315	630	2	journal	journal	PROPN
bjmsr-315	630	3	of	of	ADP
bjmsr-315	630	4	multidisciplinary	multidisciplinary	ADJ
bjmsr-315	630	5	scientific	scientific	ADJ
bjmsr-315	630	6	research	research	NOUN
bjmsr-315	630	7	vol	vol	NOUN
bjmsr-315	630	8	.	.	PROPN
bjmsr-315	630	9	1	1	NUM
bjmsr-315	630	10	,	,	PUNCT
bjmsr-315	630	11	no	no	INTJ
bjmsr-315	630	12	.	.	NOUN
bjmsr-315	630	13	1	1	NUM
bjmsr-315	630	14	;	;	PUNCT
bjmsr-315	630	15	2019	2019	NUM
bjmsr-315	630	16	65	65	NUM
bjmsr-315	630	17	r.	r.	NOUN
bjmsr-315	630	18	fourer	fourer	NOUN
bjmsr-315	630	19	(	(	PUNCT
bjmsr-315	630	20	1985a	1985a	NUM
bjmsr-315	630	21	)	)	PUNCT
bjmsr-315	630	22	a	a	DET
bjmsr-315	630	23	simplex	simplex	NOUN
bjmsr-315	630	24	algorithm	algorithm	NOUN
bjmsr-315	630	25	for	for	ADP
bjmsr-315	630	26	piecewise	piecewise	NOUN
bjmsr-315	630	27	-	-	PUNCT
bjmsr-315	630	28	linear	linear	NOUN
bjmsr-315	630	29	programming	programming	NOUN
bjmsr-315	630	30	i	i	NOUN
bjmsr-315	630	31	:	:	PUNCT
bjmsr-315	630	32	derivation	derivation	NOUN
bjmsr-315	630	33	and	and	CCONJ
bjmsr-315	630	34	proof	proof	NOUN
bjmsr-315	630	35	.	.	PUNCT
bjmsr-315	631	1	math	math	NOUN
bjmsr-315	631	2	.	.	PUNCT
bjmsr-315	632	1	prog	prog	PROPN
bjmsr-315	632	2	.	.	PROPN
bjmsr-315	633	1	,	,	PUNCT
bjmsr-315	633	2	33	33	NUM
bjmsr-315	633	3	,	,	PUNCT
bjmsr-315	633	4	204	204	NUM
bjmsr-315	633	5	-	-	SYM
bjmsr-315	633	6	233	233	NUM
bjmsr-315	633	7	.	.	PUNCT
bjmsr-315	634	1	r.	r.	PROPN
bjmsr-315	634	2	fourer	fourer	PROPN
bjmsr-315	634	3	(	(	PUNCT
bjmsr-315	634	4	1985b	1985b	NUM
bjmsr-315	634	5	)	)	PUNCT
bjmsr-315	634	6	a	a	DET
bjmsr-315	634	7	simplex	simplex	NOUN
bjmsr-315	634	8	algorithm	algorithm	NOUN
bjmsr-315	634	9	for	for	ADP
bjmsr-315	634	10	piecewise	piecewise	NOUN
bjmsr-315	634	11	-	-	PUNCT
bjmsr-315	634	12	linear	linear	NOUN
bjmsr-315	634	13	programming	programming	PROPN
bjmsr-315	634	14	ii	ii	PROPN
bjmsr-315	634	15	:	:	PUNCT
bjmsr-315	634	16	finiteness	finiteness	NOUN
bjmsr-315	634	17	,	,	PUNCT
bjmsr-315	634	18	feasibility	feasibility	NOUN
bjmsr-315	634	19	and	and	CCONJ
bjmsr-315	634	20	degeneracy	degeneracy	PROPN
bjmsr-315	634	21	.	.	PUNCT
bjmsr-315	635	1	tech	tech	PROPN
bjmsr-315	635	2	.	.	PUNCT
bjmsr-315	636	1	rep	rep	PROPN
bjmsr-315	636	2	.	.	PROPN
bjmsr-315	636	3	,	,	PUNCT
bjmsr-315	636	4	85	85	NUM
bjmsr-315	636	5	-	-	SYM
bjmsr-315	636	6	03	03	NUM
bjmsr-315	636	7	(	(	PUNCT
bjmsr-315	636	8	revised	revise	VERB
bjmsr-315	636	9	)	)	PUNCT
bjmsr-315	636	10	,	,	PUNCT
bjmsr-315	636	11	dept	dept	NOUN
bjmsr-315	636	12	.	.	PROPN
bjmsr-315	636	13	of	of	ADP
bjmsr-315	636	14	ind	ind	PROPN
bjmsr-315	636	15	.	.	PUNCT
bjmsr-315	637	1	engin	engin	PROPN
bjmsr-315	637	2	.	.	PUNCT
bjmsr-315	638	1	and	and	CCONJ
bjmsr-315	638	2	manag	manag	PROPN
bjmsr-315	638	3	.	.	PUNCT
bjmsr-315	639	1	sci	sci	PROPN
bjmsr-315	639	2	.	.	PROPN
bjmsr-315	639	3	,the	,the	PROPN
bjmsr-315	639	4	tech	tech	PROPN
bjmsr-315	639	5	.	.	PUNCT
bjmsr-315	639	6	inst	inst	PROPN
bjmsr-315	639	7	.	.	PROPN
bjmsr-315	639	8	,	,	PUNCT
bjmsr-315	639	9	northwestern	northwestern	ADJ
bjmsr-315	639	10	university	university	PROPN
bjmsr-315	639	11	,	,	PUNCT
bjmsr-315	639	12	evanston	evanston	PROPN
bjmsr-315	639	13	,	,	PUNCT
bjmsr-315	639	14	illinois	illinois	PROPN
bjmsr-315	639	15	.	.	PUNCT
bjmsr-315	639	16	r.	r.	PROPN
bjmsr-315	639	17	fourer	fourer	PROPN
bjmsr-315	639	18	(	(	PUNCT
bjmsr-315	639	19	1986	1986	NUM
bjmsr-315	639	20	)	)	PUNCT
bjmsr-315	639	21	a	a	DET
bjmsr-315	639	22	simplex	simplex	NOUN
bjmsr-315	639	23	algorithm	algorithm	NOUN
bjmsr-315	639	24	for	for	ADP
bjmsr-315	639	25	piecewise	piecewise	NOUN
bjmsr-315	639	26	-	-	PUNCT
bjmsr-315	639	27	linear	linear	NOUN
bjmsr-315	639	28	programming	programming	NOUN
bjmsr-315	639	29	iii	iii	PROPN
bjmsr-315	639	30	:	:	PUNCT
bjmsr-315	639	31	computational	computational	ADJ
bjmsr-315	639	32	analysis	analysis	NOUN
bjmsr-315	639	33	and	and	CCONJ
bjmsr-315	639	34	applications	application	NOUN
bjmsr-315	639	35	.	.	PUNCT
bjmsr-315	640	1	tech	tech	NOUN
bjmsr-315	640	2	.	.	PUNCT
bjmsr-315	641	1	rep	rep	PROPN
bjmsr-315	641	2	.	.	PROPN
bjmsr-315	641	3	,	,	PUNCT
bjmsr-315	641	4	86	86	NUM
bjmsr-315	641	5	-	-	SYM
bjmsr-315	641	6	03	03	NUM
bjmsr-315	641	7	,	,	PUNCT
bjmsr-315	641	8	dept	dept	NOUN
bjmsr-315	641	9	.	.	PROPN
bjmsr-315	641	10	of	of	ADP
bjmsr-315	641	11	ind	ind	PROPN
bjmsr-315	641	12	.	.	PUNCT
bjmsr-315	642	1	engin	engin	PROPN
bjmsr-315	642	2	.	.	PUNCT
bjmsr-315	643	1	and	and	CCONJ
bjmsr-315	643	2	manag	manag	PROPN
bjmsr-315	643	3	.	.	PUNCT
bjmsr-315	644	1	sci	sci	PROPN
bjmsr-315	644	2	.	.	PROPN
bjmsr-315	644	3	,	,	PUNCT
bjmsr-315	645	1	the	the	DET
bjmsr-315	645	2	tech	tech	NOUN
bjmsr-315	645	3	.	.	PUNCT
bjmsr-315	645	4	inst	inst	PROPN
bjmsr-315	645	5	.	.	PROPN
bjmsr-315	645	6	,	,	PUNCT
bjmsr-315	645	7	northwestern	northwestern	ADJ
bjmsr-315	645	8	university	university	PROPN
bjmsr-315	645	9	,	,	PUNCT
bjmsr-315	645	10	evanston	evanston	PROPN
bjmsr-315	645	11	,	,	PUNCT
bjmsr-315	645	12	illinois	illinois	PROPN
bjmsr-315	645	13	.	.	PUNCT
bjmsr-315	646	1	j.b.i	j.b.i	PROPN
bjmsr-315	646	2	.	.	PUNCT
bjmsr-315	647	1	fourier	fourier	PROPN
bjmsr-315	647	2	(	(	PUNCT
bjmsr-315	647	3	1824	1824	NUM
bjmsr-315	647	4	)	)	PUNCT
bjmsr-315	647	5	solution	solution	NOUN
bjmsr-315	647	6	d'une	d'une	PROPN
bjmsr-315	647	7	question	question	NOUN
bjmsr-315	647	8	particuliere	particuliere	NOUN
bjmsr-315	647	9	au	au	PROPN
bjmsr-315	647	10	calcul	calcul	PROPN
bjmsr-315	647	11	des	des	PROPN
bjmsr-315	647	12	inegalites	inegalite	NOUN
bjmsr-315	647	13	,	,	PUNCT
bjmsr-315	647	14	second	second	ADJ
bjmsr-315	647	15	extrait	extrait	NOUN
bjmsr-315	647	16	.	.	PUNCT
bjmsr-315	648	1	histoire	histoire	PROPN
bjmsr-315	648	2	de	de	PROPN
bjmsr-315	648	3	l'academie	l'academie	PROPN
bjmsr-315	648	4	des	des	PROPN
bjmsr-315	648	5	sciences	sciences	PROPN
bjmsr-315	648	6	pour	pour	VERB
bjmsr-315	648	7	1824	1824	NUM
bjmsr-315	648	8	,	,	PUNCT
bjmsr-315	648	9	47	47	NUM
bjmsr-315	648	10	-	-	SYM
bjmsr-315	648	11	55	55	NUM
bjmsr-315	648	12	.	.	PUNCT
bjmsr-315	649	1	reprinted	reprint	VERB
bjmsr-315	649	2	in	in	ADP
bjmsr-315	649	3	oeuvres	oeuvre	NOUN
bjmsr-315	649	4	de	de	X
bjmsr-315	649	5	fourier	fourier	NOUN
bjmsr-315	649	6	,	,	PUNCT
bjmsr-315	649	7	2	2	NUM
bjmsr-315	649	8	.	.	X
bjmsr-315	650	1	paris	paris	PROPN
bjmsr-315	650	2	,	,	PUNCT
bjmsr-315	650	3	1980	1980	NUM
bjmsr-315	650	4	,	,	PUNCT
bjmsr-315	650	5	gauthier	gauthier	NOUN
bjmsr-315	650	6	-	-	PUNCT
bjmsr-315	650	7	villars	villar	NOUN
bjmsr-315	650	8	,	,	PUNCT
bjmsr-315	650	9	325	325	NUM
bjmsr-315	650	10	-	-	SYM
bjmsr-315	650	11	328	328	NUM
bjmsr-315	650	12	.	.	PUNCT
bjmsr-315	651	1	g.	g.	PROPN
bjmsr-315	651	2	galilei	galilei	PROPN
bjmsr-315	651	3	(	(	PUNCT
bjmsr-315	651	4	1632	1632	NUM
bjmsr-315	651	5	)	)	PUNCT
bjmsr-315	651	6	dialogo	dialogo	PROPN
bjmsr-315	651	7	dei	dei	PROPN
bjmsr-315	651	8	massimi	massimi	PROPN
bjmsr-315	651	9	sistemi	sistemi	PROPN
bjmsr-315	651	10	.	.	PUNCT
bjmsr-315	652	1	c.b	c.b	PROPN
bjmsr-315	652	2	.	.	PROPN
bjmsr-315	652	3	garcia	garcia	PROPN
bjmsr-315	652	4	,	,	PUNCT
bjmsr-315	652	5	f.g	f.g	PROPN
bjmsr-315	652	6	.	.	PROPN
bjmsr-315	652	7	gould	gould	PROPN
bjmsr-315	652	8	(	(	PUNCT
bjmsr-315	652	9	1983	1983	NUM
bjmsr-315	652	10	)	)	PUNCT
bjmsr-315	652	11	an	an	DET
bjmsr-315	652	12	application	application	NOUN
bjmsr-315	652	13	of	of	ADP
bjmsr-315	652	14	homotopy	homotopy	NOUN
bjmsr-315	652	15	to	to	ADP
bjmsr-315	652	16	solving	solve	VERB
bjmsr-315	652	17	linear	linear	NOUN
bjmsr-315	652	18	programs	program	NOUN
bjmsr-315	652	19	.	.	PUNCT
bjmsr-315	653	1	math	math	NOUN
bjmsr-315	653	2	.	.	PUNCT
bjmsr-315	654	1	prog	prog	NOUN
bjmsr-315	654	2	.	.	PROPN
bjmsr-315	655	1	27	27	NUM
bjmsr-315	655	2	,	,	PUNCT
bjmsr-315	655	3	263	263	NUM
bjmsr-315	655	4	-	-	SYM
bjmsr-315	655	5	282	282	NUM
bjmsr-315	655	6	.	.	PUNCT
bjmsr-315	656	1	c.f	c.f	PROPN
bjmsr-315	656	2	.	.	PROPN
bjmsr-315	656	3	gauss	gauss	PROPN
bjmsr-315	656	4	(	(	PUNCT
bjmsr-315	656	5	1809	1809	NUM
bjmsr-315	656	6	)	)	PUNCT
bjmsr-315	656	7	theoria	theoria	PROPN
bjmsr-315	656	8	motus	motus	PROPN
bjmsr-315	656	9	corporum	corporum	PROPN
bjmsr-315	656	10	coelestium	coelestium	NOUN
bjmsr-315	656	11	.	.	PUNCT
bjmsr-315	657	1	in	in	ADP
bjmsr-315	657	2	f.	f.	PROPN
bjmsr-315	657	3	perthes	perthes	PROPN
bjmsr-315	657	4	,	,	PUNCT
bjmsr-315	657	5	i.h	i.h	PROPN
bjmsr-315	657	6	.	.	PROPN
bjmsr-315	657	7	besser	besser	NOUN
bjmsr-315	657	8	,	,	PUNCT
bjmsr-315	657	9	sectionbus	sectionbus	NOUN
bjmsr-315	657	10	conicis	conicis	PROPN
bjmsr-315	657	11	solem	solem	PROPN
bjmsr-315	657	12	ambientium	ambientium	PROPN
bjmsr-315	657	13	,	,	PUNCT
bjmsr-315	657	14	hamburg	hamburg	PROPN
bjmsr-315	657	15	.	.	PUNCT
bjmsr-315	658	1	reprinted	reprint	VERB
bjmsr-315	658	2	in	in	ADP
bjmsr-315	658	3	his	his	PRON
bjmsr-315	658	4	werke	werke	NOUN
bjmsr-315	658	5	,	,	PUNCT
bjmsr-315	658	6	vol	vol	NOUN
bjmsr-315	658	7	.	.	PROPN
bjmsr-315	658	8	7	7	NUM
bjmsr-315	658	9	,	,	PUNCT
bjmsr-315	658	10	f.	f.	PROPN
bjmsr-315	658	11	pethes	pethes	PROPN
bjmsr-315	658	12	,	,	PUNCT
bjmsr-315	658	13	gotha	gotha	PROPN
bjmsr-315	658	14	1871	1871	NUM
bjmsr-315	658	15	.	.	PUNCT
bjmsr-315	659	1	english	english	ADJ
bjmsr-315	659	2	translation	translation	NOUN
bjmsr-315	659	3	by	by	ADP
bjmsr-315	659	4	c.h	c.h	PROPN
bjmsr-315	659	5	.	.	PROPN
bjmsr-315	659	6	davis	davis	PROPN
bjmsr-315	659	7	,	,	PUNCT
bjmsr-315	659	8	little	little	ADJ
bjmsr-315	659	9	,	,	PUNCT
bjmsr-315	659	10	brown	brown	ADJ
bjmsr-315	659	11	and	and	CCONJ
bjmsr-315	659	12	co.	co.	PROPN
bjmsr-315	659	13	,	,	PUNCT
bjmsr-315	659	14	boston	boston	PROPN
bjmsr-315	659	15	,	,	PUNCT
bjmsr-315	659	16	1857	1857	NUM
bjmsr-315	659	17	.	.	PUNCT
bjmsr-315	660	1	reprinted	reprint	VERB
bjmsr-315	660	2	by	by	ADP
bjmsr-315	660	3	dover	dover	PROPN
bjmsr-315	660	4	pub	pub	NOUN
bjmsr-315	660	5	.	.	PUNCT
bjmsr-315	661	1	new	new	PROPN
bjmsr-315	661	2	york	york	PROPN
bjmsr-315	661	3	,	,	PUNCT
bjmsr-315	661	4	1963	1963	NUM
bjmsr-315	661	5	.	.	PUNCT
bjmsr-315	662	1	t.e	t.e	PROPN
bjmsr-315	662	2	.	.	PROPN
bjmsr-315	662	3	harris	harris	PROPN
bjmsr-315	662	4	(	(	PUNCT
bjmsr-315	662	5	1950	1950	NUM
bjmsr-315	662	6	)	)	PUNCT
bjmsr-315	662	7	regression	regression	NOUN
bjmsr-315	662	8	using	use	VERB
bjmsr-315	662	9	minimum	minimum	ADJ
bjmsr-315	662	10	absolute	absolute	ADJ
bjmsr-315	662	11	deviations	deviation	NOUN
bjmsr-315	662	12	.	.	PUNCT
bjmsr-315	663	1	am	be	AUX
bjmsr-315	663	2	.	.	PUNCT
bjmsr-315	663	3	statist	statist	PROPN
bjmsr-315	663	4	.	.	PUNCT
bjmsr-315	663	5	,	,	PUNCT
bjmsr-315	663	6	4	4	NUM
bjmsr-315	663	7	,	,	PUNCT
bjmsr-315	663	8	14	14	NUM
bjmsr-315	663	9	-	-	SYM
bjmsr-315	663	10	15	15	NUM
bjmsr-315	663	11	.	.	PUNCT
bjmsr-315	664	1	h.l	h.l	PROPN
bjmsr-315	664	2	.	.	PROPN
bjmsr-315	664	3	harter	harter	PROPN
bjmsr-315	664	4	(	(	PUNCT
bjmsr-315	664	5	1974a	1974a	NUM
bjmsr-315	664	6	)	)	PUNCT
bjmsr-315	664	7	the	the	DET
bjmsr-315	664	8	method	method	NOUN
bjmsr-315	664	9	of	of	ADP
bjmsr-315	664	10	least	least	ADJ
bjmsr-315	664	11	squares	square	NOUN
bjmsr-315	664	12	and	and	CCONJ
bjmsr-315	664	13	some	some	DET
bjmsr-315	664	14	alternative	alternative	NOUN
bjmsr-315	664	15	,	,	PUNCT
bjmsr-315	664	16	i.	i.	PROPN
bjmsr-315	664	17	int	int	PROPN
bjmsr-315	664	18	.	.	PUNCT
bjmsr-315	665	1	stat	stat	PROPN
bjmsr-315	665	2	.	.	PUNCT
bjmsr-315	666	1	rev	rev	PROPN
bjmsr-315	666	2	.	.	PROPN
bjmsr-315	666	3	,	,	PUNCT
bjmsr-315	666	4	42	42	NUM
bjmsr-315	666	5	,	,	PUNCT
bjmsr-315	666	6	147	147	NUM
bjmsr-315	666	7	-	-	SYM
bjmsr-315	666	8	174	174	NUM
bjmsr-315	666	9	.	.	PUNCT
bjmsr-315	667	1	h.l	h.l	PROPN
bjmsr-315	667	2	.	.	PROPN
bjmsr-315	667	3	harter	harter	PROPN
bjmsr-315	667	4	(	(	PUNCT
bjmsr-315	667	5	1974b	1974b	NUM
bjmsr-315	667	6	)	)	PUNCT
bjmsr-315	667	7	the	the	DET
bjmsr-315	667	8	method	method	NOUN
bjmsr-315	667	9	of	of	ADP
bjmsr-315	667	10	least	least	ADJ
bjmsr-315	667	11	squares	square	NOUN
bjmsr-315	667	12	and	and	CCONJ
bjmsr-315	667	13	some	some	DET
bjmsr-315	667	14	alternative	alternative	NOUN
bjmsr-315	667	15	,	,	PUNCT
bjmsr-315	667	16	ii	ii	PROPN
bjmsr-315	667	17	.	.	PUNCT
bjmsr-315	667	18	int	int	PROPN
bjmsr-315	667	19	.	.	PUNCT
bjmsr-315	668	1	stat	stat	PROPN
bjmsr-315	668	2	.	.	PUNCT
bjmsr-315	669	1	rev	rev	PROPN
bjmsr-315	669	2	.	.	PROPN
bjmsr-315	669	3	,	,	PUNCT
bjmsr-315	669	4	42	42	NUM
bjmsr-315	669	5	,	,	PUNCT
bjmsr-315	669	6	235	235	NUM
bjmsr-315	669	7	-	-	SYM
bjmsr-315	669	8	264	264	NUM
bjmsr-315	669	9	.	.	PUNCT
bjmsr-315	670	1	h.l	h.l	PROPN
bjmsr-315	670	2	.	.	PROPN
bjmsr-315	670	3	harter	harter	PROPN
bjmsr-315	670	4	(	(	PUNCT
bjmsr-315	670	5	1975a	1975a	NUM
bjmsr-315	670	6	)	)	PUNCT
bjmsr-315	670	7	the	the	DET
bjmsr-315	670	8	method	method	NOUN
bjmsr-315	670	9	of	of	ADP
bjmsr-315	670	10	least	least	ADJ
bjmsr-315	670	11	squares	square	NOUN
bjmsr-315	670	12	and	and	CCONJ
bjmsr-315	670	13	some	some	DET
bjmsr-315	670	14	alternative	alternative	ADJ
bjmsr-315	670	15	,	,	PUNCT
bjmsr-315	670	16	iii	iii	PROPN
bjmsr-315	670	17	.	.	PUNCT
bjmsr-315	670	18	int	int	NOUN
bjmsr-315	670	19	.	.	PUNCT
bjmsr-315	671	1	stat	stat	PROPN
bjmsr-315	671	2	.	.	PUNCT
bjmsr-315	672	1	rev	rev	PROPN
bjmsr-315	672	2	.	.	PROPN
bjmsr-315	672	3	,	,	PUNCT
bjmsr-315	672	4	43	43	NUM
bjmsr-315	672	5	,	,	PUNCT
bjmsr-315	672	6	1	1	NUM
bjmsr-315	672	7	-	-	SYM
bjmsr-315	672	8	44	44	NUM
bjmsr-315	672	9	.	.	PUNCT
bjmsr-315	673	1	h.l	h.l	PROPN
bjmsr-315	673	2	.	.	PROPN
bjmsr-315	673	3	harter	harter	PROPN
bjmsr-315	673	4	(	(	PUNCT
bjmsr-315	673	5	1975b	1975b	NUM
bjmsr-315	673	6	)	)	PUNCT
bjmsr-315	673	7	the	the	DET
bjmsr-315	673	8	method	method	NOUN
bjmsr-315	673	9	of	of	ADP
bjmsr-315	673	10	least	least	ADJ
bjmsr-315	673	11	squares	square	NOUN
bjmsr-315	673	12	and	and	CCONJ
bjmsr-315	673	13	some	some	DET
bjmsr-315	673	14	alternative	alternative	NOUN
bjmsr-315	673	15	,	,	PUNCT
bjmsr-315	673	16	iv	iv	NUM
bjmsr-315	673	17	int	int	NOUN
bjmsr-315	673	18	.	.	PUNCT
bjmsr-315	674	1	stat	stat	PROPN
bjmsr-315	674	2	.	.	PUNCT
bjmsr-315	675	1	rev	rev	PROPN
bjmsr-315	675	2	.	.	PROPN
bjmsr-315	675	3	,	,	PUNCT
bjmsr-315	675	4	43	43	NUM
bjmsr-315	675	5	,	,	PUNCT
bjmsr-315	675	6	125	125	NUM
bjmsr-315	675	7	-	-	SYM
bjmsr-315	675	8	190	190	NUM
bjmsr-315	675	9	,	,	PUNCT
bjmsr-315	675	10	273	273	NUM
bjmsr-315	675	11	-	-	SYM
bjmsr-315	675	12	278	278	NUM
bjmsr-315	675	13	.	.	PUNCT
bjmsr-315	676	1	h.l	h.l	PROPN
bjmsr-315	676	2	.	.	PROPN
bjmsr-315	676	3	harter	harter	PROPN
bjmsr-315	676	4	(	(	PUNCT
bjmsr-315	676	5	1975c	1975c	NUM
bjmsr-315	676	6	)	)	PUNCT
bjmsr-315	676	7	the	the	DET
bjmsr-315	676	8	method	method	NOUN
bjmsr-315	676	9	of	of	ADP
bjmsr-315	676	10	least	least	ADJ
bjmsr-315	676	11	squares	square	NOUN
bjmsr-315	676	12	and	and	CCONJ
bjmsr-315	676	13	some	some	DET
bjmsr-315	676	14	alternative	alternative	NOUN
bjmsr-315	676	15	,	,	PUNCT
bjmsr-315	676	16	v.	v.	ADP
bjmsr-315	676	17	int	int	NOUN
bjmsr-315	676	18	.	.	PUNCT
bjmsr-315	677	1	stat	stat	PROPN
bjmsr-315	677	2	.	.	PUNCT
bjmsr-315	678	1	rev	rev	PROPN
bjmsr-315	678	2	.	.	PROPN
bjmsr-315	678	3	,	,	PUNCT
bjmsr-315	678	4	43	43	NUM
bjmsr-315	678	5	,	,	PUNCT
bjmsr-315	678	6	269	269	NUM
bjmsr-315	678	7	-	-	SYM
bjmsr-315	678	8	272	272	NUM
bjmsr-315	678	9	.	.	PUNCT
bjmsr-315	679	1	h.l	h.l	PROPN
bjmsr-315	679	2	.	.	PROPN
bjmsr-315	679	3	harter	harter	PROPN
bjmsr-315	679	4	(	(	PUNCT
bjmsr-315	679	5	1976	1976	NUM
bjmsr-315	679	6	)	)	PUNCT
bjmsr-315	679	7	the	the	DET
bjmsr-315	679	8	method	method	NOUN
bjmsr-315	679	9	of	of	ADP
bjmsr-315	679	10	least	least	ADJ
bjmsr-315	679	11	squares	square	NOUN
bjmsr-315	679	12	and	and	CCONJ
bjmsr-315	679	13	some	some	DET
bjmsr-315	679	14	alternative	alternative	NOUN
bjmsr-315	679	15	,	,	PUNCT
bjmsr-315	679	16	vi	vi	PROPN
bjmsr-315	679	17	.	.	PROPN
bjmsr-315	679	18	int	int	NOUN
bjmsr-315	679	19	.	.	PUNCT
bjmsr-315	680	1	stat	stat	PROPN
bjmsr-315	680	2	.	.	PUNCT
bjmsr-315	681	1	rev	rev	PROPN
bjmsr-315	681	2	.	.	PROPN
bjmsr-315	681	3	,	,	PUNCT
bjmsr-315	681	4	44	44	NUM
bjmsr-315	681	5	,	,	PUNCT
bjmsr-315	681	6	113	113	NUM
bjmsr-315	681	7	-	-	SYM
bjmsr-315	681	8	159	159	NUM
bjmsr-315	681	9	.	.	PUNCT
bjmsr-315	682	1	p.w	p.w	PROPN
bjmsr-315	682	2	.	.	PROPN
bjmsr-315	682	3	holland	holland	PROPN
bjmsr-315	682	4	,	,	PUNCT
bjmsr-315	682	5	r.e	r.e	PROPN
bjmsr-315	682	6	.	.	PROPN
bjmsr-315	682	7	welsch	welsch	PROPN
bjmsr-315	682	8	(	(	PUNCT
bjmsr-315	682	9	1977	1977	NUM
bjmsr-315	682	10	)	)	PUNCT
bjmsr-315	682	11	robust	robust	ADJ
bjmsr-315	682	12	regression	regression	NOUN
bjmsr-315	682	13	using	use	VERB
bjmsr-315	682	14	iteratively	iteratively	ADV
bjmsr-315	682	15	reweighted	reweighte	VERB
bjmsr-315	682	16	least	least	ADJ
bjmsr-315	682	17	-	-	PUNCT
bjmsr-315	682	18	squares	square	NOUN
bjmsr-315	682	19	.	.	PUNCT
bjmsr-315	683	1	comm	comm	NOUN
bjmsr-315	683	2	.	.	PUNCT
bjmsr-315	684	1	stat	stat	PROPN
bjmsr-315	684	2	.	.	PUNCT
bjmsr-315	684	3	,	,	PUNCT
bjmsr-315	684	4	a6	a6	NOUN
bjmsr-315	684	5	,	,	PUNCT
bjmsr-315	684	6	813	813	NUM
bjmsr-315	684	7	-	-	SYM
bjmsr-315	684	8	827	827	NUM
bjmsr-315	684	9	.	.	PUNCT
bjmsr-315	685	1	l.	l.	PROPN
bjmsr-315	685	2	horvath	horvath	PROPN
bjmsr-315	685	3	(	(	PUNCT
bjmsr-315	685	4	1987	1987	NUM
bjmsr-315	685	5	)	)	PUNCT
bjmsr-315	685	6	asymptotic	asymptotic	ADJ
bjmsr-315	685	7	normality	normality	NOUN
bjmsr-315	685	8	of	of	ADP
bjmsr-315	685	9	lp	lp	NOUN
bjmsr-315	685	10	-	-	PUNCT
bjmsr-315	685	11	norms	norm	NOUN
bjmsr-315	685	12	of	of	ADP
bjmsr-315	685	13	density	density	NOUN
bjmsr-315	685	14	estimators	estimator	NOUN
bjmsr-315	685	15	.	.	PUNCT
bjmsr-315	686	1	tech	tech	NOUN
bjmsr-315	686	2	.	.	PUNCT
bjmsr-315	686	3	rep	rep	PROPN
bjmsr-315	686	4	.	.	PROPN
bjmsr-315	686	5	series	series	PROPN
bjmsr-315	686	6	of	of	ADP
bjmsr-315	686	7	lab	lab	NOUN
bjmsr-315	686	8	res	re	NOUN
bjmsr-315	686	9	.	.	PUNCT
bjmsr-315	686	10	stat	stat	PROPN
bjmsr-315	686	11	.	.	PUNCT
bjmsr-315	687	1	prob	prob	PROPN
bjmsr-315	687	2	.	.	PROPN
bjmsr-315	687	3	,	,	PUNCT
bjmsr-315	687	4	no.3	no.3	PROPN
bjmsr-315	687	5	,	,	PUNCT
bjmsr-315	687	6	carleton	carleton	PROPN
bjmsr-315	687	7	university	university	PROPN
bjmsr-315	687	8	,	,	PUNCT
bjmsr-315	687	9	ottawa	ottawa	PROPN
bjmsr-315	687	10	,	,	PUNCT
bjmsr-315	687	11	canada	canada	PROPN
bjmsr-315	687	12	.	.	PUNCT
bjmsr-315	688	1	h.	h.	PROPN
bjmsr-315	688	2	imai	imai	PROPN
bjmsr-315	688	3	,	,	PUNCT
bjmsr-315	688	4	k.	k.	PROPN
bjmsr-315	688	5	kato	kato	PROPN
bjmsr-315	688	6	,	,	PUNCT
bjmsr-315	688	7	p.	p.	NOUN
bjmsr-315	688	8	yamamoto	yamamoto	PROPN
bjmsr-315	688	9	(	(	PUNCT
bjmsr-315	688	10	1987	1987	NUM
bjmsr-315	688	11	)	)	PUNCT
bjmsr-315	688	12	a	a	DET
bjmsr-315	688	13	linear	linear	ADJ
bjmsr-315	688	14	-	-	PUNCT
bjmsr-315	688	15	time	time	NOUN
bjmsr-315	688	16	algorithm	algorithm	NOUN
bjmsr-315	688	17	for	for	ADP
bjmsr-315	688	18	linear	linear	PROPN
bjmsr-315	688	19	l1	l1	PROPN
bjmsr-315	688	20	approximation	approximation	NOUN
bjmsr-315	688	21	of	of	ADP
bjmsr-315	688	22	points	point	NOUN
bjmsr-315	688	23	.	.	PUNCT
bjmsr-315	689	1	tech	tech	NOUN
bjmsr-315	689	2	.	.	PUNCT
bjmsr-315	690	1	rep	rep	PROPN
bjmsr-315	690	2	.	.	PROPN
bjmsr-315	690	3	csce-87c30	csce-87c30	PROPN
bjmsr-315	690	4	.	.	PUNCT
bjmsr-315	691	1	dept	dept	PROPN
bjmsr-315	691	2	.	.	PROPN
bjmsr-315	692	1	of	of	ADP
bjmsr-315	692	2	comp	comp	PROPN
bjmsr-315	692	3	.	.	PUNCT
bjmsr-315	693	1	sci	sci	PROPN
bjmsr-315	693	2	.	.	PROPN
bjmsr-315	693	3	and	and	CCONJ
bjmsr-315	693	4	commun	commun	PROPN
bjmsr-315	693	5	.	.	PUNCT
bjmsr-315	694	1	engin	engin	PROPN
bjmsr-315	694	2	.	.	PROPN
bjmsr-315	694	3	,	,	PUNCT
bjmsr-315	694	4	kyushu	kyushu	PROPN
bjmsr-315	694	5	university	university	PROPN
bjmsr-315	694	6	36	36	NUM
bjmsr-315	694	7	,	,	PUNCT
bjmsr-315	694	8	fukuoka	fukuoka	PROPN
bjmsr-315	694	9	812	812	NUM
bjmsr-315	694	10	,	,	PUNCT
bjmsr-315	694	11	japan	japan	PROPN
bjmsr-315	694	12	.	.	PUNCT
bjmsr-315	695	1	l.a	l.a	PROPN
bjmsr-315	695	2	.	.	PROPN
bjmsr-315	695	3	josvanger	josvanger	PROPN
bjmsr-315	695	4	,	,	PUNCT
bjmsr-315	695	5	v.a	v.a	PROPN
bjmsr-315	695	6	.	.	PROPN
bjmsr-315	695	7	sposito	sposito	PROPN
bjmsr-315	695	8	(	(	PUNCT
bjmsr-315	695	9	1983	1983	NUM
bjmsr-315	695	10	)	)	PUNCT
bjmsr-315	695	11	l1	l1	PROPN
bjmsr-315	695	12	-	-	PUNCT
bjmsr-315	695	13	norm	norm	NOUN
bjmsr-315	695	14	estimates	estimate	NOUN
bjmsr-315	695	15	for	for	ADP
bjmsr-315	695	16	the	the	DET
bjmsr-315	695	17	simple	simple	ADJ
bjmsr-315	695	18	regression	regression	NOUN
bjmsr-315	695	19	problem	problem	NOUN
bjmsr-315	695	20	.	.	PUNCT
bjmsr-315	696	1	comm	comm	NOUN
bjmsr-315	696	2	.	.	PUNCT
bjmsr-315	697	1	stat	stat	PROPN
bjmsr-315	697	2	.	.	PUNCT
bjmsr-315	698	1	b12	b12	NOUN
bjmsr-315	698	2	,	,	PUNCT
bjmsr-315	698	3	215	215	NUM
bjmsr-315	698	4	-	-	SYM
bjmsr-315	698	5	21	21	NUM
bjmsr-315	698	6	.	.	PUNCT
bjmsr-315	699	1	o.j	o.j	PROPN
bjmsr-315	699	2	.	.	PROPN
bjmsr-315	699	3	karst	karst	PROPN
bjmsr-315	699	4	(	(	PUNCT
bjmsr-315	699	5	1958	1958	NUM
bjmsr-315	699	6	)	)	PUNCT
bjmsr-315	699	7	linear	linear	NOUN
bjmsr-315	699	8	curve	curve	NOUN
bjmsr-315	699	9	fitting	fitting	ADJ
bjmsr-315	699	10	using	use	VERB
bjmsr-315	699	11	least	least	ADJ
bjmsr-315	699	12	deviations	deviation	NOUN
bjmsr-315	699	13	.	.	PUNCT
bjmsr-315	700	1	jasa	jasa	PROPN
bjmsr-315	700	2	,	,	PUNCT
bjmsr-315	700	3	53	53	NUM
bjmsr-315	700	4	,	,	PUNCT
bjmsr-315	700	5	118	118	NUM
bjmsr-315	700	6	-	-	SYM
bjmsr-315	700	7	132	132	NUM
bjmsr-315	700	8	.	.	PUNCT
bjmsr-315	701	1	y.	y.	PROPN
bjmsr-315	701	2	kawara	kawara	PROPN
bjmsr-315	701	3	(	(	PUNCT
bjmsr-315	701	4	1979	1979	NUM
bjmsr-315	701	5	)	)	PUNCT
bjmsr-315	701	6	straight	straight	ADJ
bjmsr-315	701	7	line	line	NOUN
bjmsr-315	701	8	fitting	fit	VERB
bjmsr-315	701	9	by	by	ADP
bjmsr-315	701	10	minimizing	minimize	VERB
bjmsr-315	701	11	the	the	DET
bjmsr-315	701	12	sum	sum	NOUN
bjmsr-315	701	13	of	of	ADP
bjmsr-315	701	14	absolute	absolute	ADJ
bjmsr-315	701	15	deviations	deviation	NOUN
bjmsr-315	701	16	.	.	PUNCT
bjmsr-315	702	1	j.	j.	PROPN
bjmsr-315	702	2	of	of	ADP
bjmsr-315	702	3	the	the	DET
bjmsr-315	702	4	japan	japan	PROPN
bjmsr-315	702	5	stat	stat	PROPN
bjmsr-315	702	6	.	.	PUNCT
bjmsr-315	703	1	soc	soc	PROPN
bjmsr-315	703	2	.	.	PUNCT
bjmsr-315	703	3	,	,	PUNCT
bjmsr-315	703	4	9	9	NUM
bjmsr-315	703	5	,	,	PUNCT
bjmsr-315	703	6	47	47	NUM
bjmsr-315	703	7	-	-	SYM
bjmsr-315	703	8	64	64	NUM
bjmsr-315	703	9	.	.	PUNCT
bjmsr-315	704	1	j.h.b	j.h.b	PROPN
bjmsr-315	704	2	.	.	PUNCT
bjmsr-315	705	1	kemperman	kemperman	PROPN
bjmsr-315	705	2	(	(	PUNCT
bjmsr-315	705	3	1984	1984	NUM
bjmsr-315	705	4	)	)	PUNCT
bjmsr-315	705	5	least	least	ADJ
bjmsr-315	705	6	absolute	absolute	ADJ
bjmsr-315	705	7	value	value	NOUN
bjmsr-315	705	8	and	and	CCONJ
bjmsr-315	705	9	median	median	ADJ
bjmsr-315	705	10	polish	polish	NOUN
bjmsr-315	705	11	.	.	PUNCT
bjmsr-315	706	1	in	in	ADP
bjmsr-315	706	2	y.l	y.l	PROPN
bjmsr-315	706	3	.	.	PUNCT
bjmsr-315	706	4	tong	tong	PROPN
bjmsr-315	706	5	(	(	PUNCT
bjmsr-315	706	6	ed	ed	NOUN
bjmsr-315	706	7	.	.	PUNCT
bjmsr-315	706	8	)	)	PUNCT
bjmsr-315	706	9	,	,	PUNCT
bjmsr-315	706	10	inequalities	inequality	NOUN
bjmsr-315	706	11	in	in	ADP
bjmsr-315	706	12	statistics	statistic	NOUN
bjmsr-315	706	13	and	and	CCONJ
bjmsr-315	706	14	probability	probability	NOUN
bjmsr-315	706	15	(	(	PUNCT
bjmsr-315	706	16	ims	im	NOUN
bjmsr-315	706	17	lecture	lecture	NOUN
bjmsr-315	706	18	notes	note	NOUN
bjmsr-315	706	19	monograph	monograph	PROPN
bjmsr-315	706	20	series	series	PROPN
bjmsr-315	706	21	,	,	PUNCT
bjmsr-315	706	22	vol.5	vol.5	PROPN
bjmsr-315	706	23	)	)	PUNCT
bjmsr-315	706	24	,	,	PUNCT
bjmsr-315	706	25	inst	inst	PROPN
bjmsr-315	706	26	.	.	PROPN
bjmsr-315	706	27	of	of	ADP
bjmsr-315	706	28	math	math	NOUN
bjmsr-315	706	29	.	.	PUNCT
bjmsr-315	707	1	stat	stat	PROPN
bjmsr-315	707	2	.	.	PUNCT
bjmsr-315	707	3	,	,	PUNCT
bjmsr-315	707	4	hayward	hayward	PROPN
bjmsr-315	707	5	,	,	PUNCT
bjmsr-315	707	6	ca	can	AUX
bjmsr-315	707	7	,	,	PUNCT
bjmsr-315	707	8	84	84	NUM
bjmsr-315	707	9	-	-	SYM
bjmsr-315	707	10	113	113	NUM
bjmsr-315	707	11	.	.	PUNCT
bjmsr-315	708	1	w.j	w.j	PROPN
bjmsr-315	708	2	.	.	PROPN
bjmsr-315	708	3	kennedy	kennedy	PROPN
bjmsr-315	708	4	,	,	PUNCT
bjmsr-315	708	5	j.e	j.e	PROPN
bjmsr-315	708	6	.	.	PROPN
bjmsr-315	708	7	gentle	gentle	ADJ
bjmsr-315	708	8	(	(	PUNCT
bjmsr-315	708	9	1980	1980	NUM
bjmsr-315	708	10	)	)	PUNCT
bjmsr-315	708	11	statistical	statistical	ADJ
bjmsr-315	708	12	computing	computing	NOUN
bjmsr-315	708	13	.	.	PUNCT
bjmsr-315	709	1	new	new	PROPN
bjmsr-315	709	2	york	york	PROPN
bjmsr-315	709	3	,	,	PUNCT
bjmsr-315	709	4	marcel	marcel	PROPN
bjmsr-315	709	5	dekker	dekker	PROPN
bjmsr-315	709	6	.	.	PUNCT
bjmsr-315	710	1	p.s	p.s	PROPN
bjmsr-315	710	2	.	.	PROPN
bjmsr-315	710	3	laplace	laplace	PROPN
bjmsr-315	710	4	(	(	PUNCT
bjmsr-315	710	5	1793	1793	NUM
bjmsr-315	710	6	)	)	PUNCT
bjmsr-315	710	7	sur	sur	PROPN
bjmsr-315	710	8	quelques	quelques	PROPN
bjmsr-315	710	9	points	point	NOUN
bjmsr-315	710	10	du	du	PROPN
bjmsr-315	710	11	system	system	NOUN
bjmsr-315	710	12	du	du	PROPN
bjmsr-315	710	13	monde	monde	PROPN
bjmsr-315	710	14	.	.	PUNCT
bjmsr-315	711	1	memoires	memoires	PROPN
bjmsr-315	711	2	de	de	PROPN
bjmsr-315	711	3	l'academie	l'academie	PROPN
bjmsr-315	711	4	royale	royale	PROPN
bjmsr-315	711	5	des	des	PROPN
bjmsr-315	711	6	science	science	PROPN
bjmsr-315	711	7	de	de	PROPN
bjmsr-315	711	8	paris	paris	PROPN
bjmsr-315	711	9	.	.	PUNCT
bjmsr-315	712	1	annee	annee	PROPN
bjmsr-315	712	2	1789	1789	NUM
bjmsr-315	712	3	,	,	PUNCT
bjmsr-315	712	4	1	1	NUM
bjmsr-315	712	5	-	-	SYM
bjmsr-315	712	6	87	87	NUM
bjmsr-315	712	7	.	.	PUNCT
bjmsr-315	713	1	reprinted	reprint	VERB
bjmsr-315	713	2	in	in	ADP
bjmsr-315	713	3	oeuvres	oeuvre	NOUN
bjmsr-315	713	4	completes	complete	VERB
bjmsr-315	713	5	de	de	PROPN
bjmsr-315	713	6	laplace	laplace	PROPN
bjmsr-315	713	7	ii	ii	PROPN
bjmsr-315	713	8	.	.	PUNCT
bjmsr-315	714	1	paris	paris	PROPN
bjmsr-315	714	2	,	,	PUNCT
bjmsr-315	714	3	gauthier	gauthier	NOUN
bjmsr-315	714	4	-	-	PUNCT
bjmsr-315	714	5	villars	villar	NOUN
bjmsr-315	714	6	,	,	PUNCT
bjmsr-315	714	7	1985	1985	NUM
bjmsr-315	714	8	,	,	PUNCT
bjmsr-315	714	9	477	477	NUM
bjmsr-315	714	10	-	-	SYM
bjmsr-315	714	11	558	558	NUM
bjmsr-315	714	12	.	.	PUNCT
bjmsr-315	715	1	copyright	copyright	NOUN
bjmsr-315	715	2	©	©	PROPN
bjmsr-315	715	3	cc	cc	PROPN
bjmsr-315	715	4	-	-	PUNCT
bjmsr-315	715	5	by	by	ADP
bjmsr-315	715	6	-	-	PUNCT
bjmsr-315	715	7	nc	nc	PROPN
bjmsr-315	715	8	2019	2019	NUM
bjmsr-315	715	9	,	,	PUNCT
bjmsr-315	715	10	bjmsr	bjmsr	PROPN
bjmsr-315	715	11	www.cribfb.com/journal/index.php/bjmsr	www.cribfb.com/journal/index.php/bjmsr	X
bjmsr-315	715	12	bangladesh	bangladesh	PROPN
bjmsr-315	715	13	journal	journal	PROPN
bjmsr-315	715	14	of	of	ADP
bjmsr-315	715	15	multidisciplinary	multidisciplinary	ADJ
bjmsr-315	715	16	scientific	scientific	ADJ
bjmsr-315	715	17	research	research	NOUN
bjmsr-315	715	18	vol	vol	NOUN
bjmsr-315	715	19	.	.	PROPN
bjmsr-315	716	1	1	1	NUM
bjmsr-315	716	2	,	,	PUNCT
bjmsr-315	716	3	no	no	INTJ
bjmsr-315	716	4	.	.	NOUN
bjmsr-315	716	5	1	1	NUM
bjmsr-315	716	6	;	;	PUNCT
bjmsr-315	717	1	2019	2019	NUM
bjmsr-315	717	2	66	66	NUM
bjmsr-315	717	3	p.s	p.s	PROPN
bjmsr-315	717	4	.	.	PROPN
bjmsr-315	717	5	laplace	laplace	PROPN
bjmsr-315	717	6	(	(	PUNCT
bjmsr-315	717	7	1799	1799	NUM
bjmsr-315	717	8	)	)	PUNCT
bjmsr-315	717	9	traite	traite	PROPN
bjmsr-315	717	10	des	des	PROPN
bjmsr-315	717	11	mecanique	mecanique	PROPN
bjmsr-315	717	12	celeste	celeste	PROPN
bjmsr-315	717	13	,	,	PUNCT
bjmsr-315	717	14	2	2	NUM
bjmsr-315	717	15	.	.	X
bjmsr-315	717	16	paris	paris	PROPN
bjmsr-315	717	17	;	;	PUNCT
bjmsr-315	717	18	j.b.m	j.b.m	PROPN
bjmsr-315	717	19	.	.	PUNCT
bjmsr-315	717	20	depart	depart	PROPN
bjmsr-315	717	21	.	.	PUNCT
bjmsr-315	718	1	reprinted	reprint	VERB
bjmsr-315	718	2	as	as	ADP
bjmsr-315	718	3	oeuvres	oeuvre	NOUN
bjmsr-315	718	4	completes	complete	VERB
bjmsr-315	718	5	de	de	ADP
bjmsr-315	718	6	laplace	laplace	NOUN
bjmsr-315	718	7	,	,	PUNCT
bjmsr-315	718	8	2	2	NUM
bjmsr-315	718	9	.	.	X
bjmsr-315	718	10	paris	paris	PROPN
bjmsr-315	718	11	;	;	PUNCT
bjmsr-315	718	12	gauthier	gauthier	NOUN
bjmsr-315	718	13	-	-	PUNCT
bjmsr-315	718	14	villars	villar	NOUN
bjmsr-315	718	15	1878	1878	NUM
bjmsr-315	718	16	,	,	PUNCT
bjmsr-315	718	17	116	116	NUM
bjmsr-315	718	18	-	-	SYM
bjmsr-315	718	19	165	165	NUM
bjmsr-315	718	20	.	.	PUNCT
bjmsr-315	719	1	p.s	p.s	PROPN
bjmsr-315	719	2	.	.	PROPN
bjmsr-315	719	3	laplace	laplace	PROPN
bjmsr-315	719	4	(	(	PUNCT
bjmsr-315	719	5	1812	1812	NUM
bjmsr-315	719	6	)	)	PUNCT
bjmsr-315	719	7	theorie	theorie	PROPN
bjmsr-315	719	8	analytique	analytique	PROPN
bjmsr-315	719	9	des	des	X
bjmsr-315	719	10	probabilites	probabilite	NOUN
bjmsr-315	719	11	,	,	PUNCT
bjmsr-315	719	12	mme	mme	X
bjmsr-315	719	13	courcier	courcier	PROPN
bjmsr-315	719	14	paris	paris	PROPN
bjmsr-315	719	15	1820	1820	NUM
bjmsr-315	719	16	reprinted	reprint	VERB
bjmsr-315	719	17	in	in	ADP
bjmsr-315	719	18	his	his	PRON
bjmsr-315	719	19	oeuvres	oeuvre	NOUN
bjmsr-315	719	20	,	,	PUNCT
bjmsr-315	719	21	vol.7	vol.7	PROPN
bjmsr-315	719	22	,	,	PUNCT
bjmsr-315	719	23	imprimerie	imprimerie	PROPN
bjmsr-315	719	24	royale	royale	PROPN
bjmsr-315	719	25	,	,	PUNCT
bjmsr-315	719	26	paris	paris	PROPN
bjmsr-315	719	27	,	,	PUNCT
bjmsr-315	719	28	1847	1847	NUM
bjmsr-315	719	29	,	,	PUNCT
bjmsr-315	719	30	and	and	CCONJ
bjmsr-315	719	31	gauthier	gauthier	NOUN
bjmsr-315	719	32	-	-	PUNCT
bjmsr-315	719	33	villars	villar	NOUN
bjmsr-315	719	34	et	et	NOUN
bjmsr-315	719	35	fils	fil	NOUN
bjmsr-315	719	36	,	,	PUNCT
bjmsr-315	719	37	paris	paris	PROPN
bjmsr-315	719	38	1886	1886	NUM
bjmsr-315	719	39	.	.	PUNCT
bjmsr-315	720	1	p.s	p.s	PROPN
bjmsr-315	720	2	.	.	PROPN
bjmsr-315	720	3	laplace	laplace	PROPN
bjmsr-315	720	4	(	(	PUNCT
bjmsr-315	720	5	1818	1818	NUM
bjmsr-315	720	6	)	)	PUNCT
bjmsr-315	720	7	duexieme	duexieme	NOUN
bjmsr-315	720	8	supplement	supplement	NOUN
bjmsr-315	720	9	to	to	ADP
bjmsr-315	720	10	laplace	laplace	NOUN
bjmsr-315	720	11	(	(	PUNCT
bjmsr-315	720	12	1812	1812	NUM
bjmsr-315	720	13	)	)	PUNCT
bjmsr-315	720	14	.	.	PUNCT
bjmsr-315	721	1	c.l	c.l	PROPN
bjmsr-315	721	2	.	.	PROPN
bjmsr-315	721	3	mathieu	mathieu	PROPN
bjmsr-315	721	4	(	(	PUNCT
bjmsr-315	721	5	1816	1816	NUM
bjmsr-315	721	6	)	)	PUNCT
bjmsr-315	721	7	sur	sur	PROPN
bjmsr-315	721	8	les	les	PROPN
bjmsr-315	721	9	experiences	experience	NOUN
bjmsr-315	721	10	du	du	NOUN
bjmsr-315	721	11	pendule	pendule	NOUN
bjmsr-315	721	12	,	,	PUNCT
bjmsr-315	721	13	faites	faites	PROPN
bjmsr-315	721	14	par	par	PROPN
bjmsr-315	721	15	les	les	PROPN
bjmsr-315	721	16	navigateurs	navigateurs	PROPN
bjmsr-315	721	17	espagnol	espagnol	PROPN
bjmsr-315	721	18	,	,	PUNCT
bjmsr-315	721	19	en	en	ADP
bjmsr-315	721	20	differens	differen	NOUN
bjmsr-315	721	21	points	point	VERB
bjmsr-315	721	22	du	du	PROPN
bjmsr-315	721	23	globe	globe	PROPN
bjmsr-315	721	24	.	.	PUNCT
bjmsr-315	722	1	connaissance	connaissance	PROPN
bjmsr-315	722	2	des	des	PROPN
bjmsr-315	722	3	tems	tems	PROPN
bjmsr-315	722	4	,	,	PUNCT
bjmsr-315	722	5	314	314	NUM
bjmsr-315	722	6	-	-	SYM
bjmsr-315	722	7	332	332	NUM
bjmsr-315	722	8	.	.	PUNCT
bjmsr-315	723	1	c.r	c.r	PROPN
bjmsr-315	723	2	.	.	PROPN
bjmsr-315	723	3	mcconnell	mcconnell	PROPN
bjmsr-315	723	4	(	(	PUNCT
bjmsr-315	723	5	1987	1987	NUM
bjmsr-315	723	6	)	)	PUNCT
bjmsr-315	723	7	on	on	ADP
bjmsr-315	723	8	computing	compute	VERB
bjmsr-315	723	9	a	a	DET
bjmsr-315	723	10	best	good	ADJ
bjmsr-315	723	11	discrete	discrete	ADJ
bjmsr-315	723	12	l1	l1	PROPN
bjmsr-315	723	13	approximation	approximation	NOUN
bjmsr-315	723	14	using	use	VERB
bjmsr-315	723	15	the	the	DET
bjmsr-315	723	16	method	method	NOUN
bjmsr-315	723	17	of	of	ADP
bjmsr-315	723	18	vanishing	vanish	VERB
bjmsr-315	723	19	jacobians	jacobian	NOUN
bjmsr-315	723	20	.	.	PUNCT
bjmsr-315	724	1	csda	csda	NOUN
bjmsr-315	724	2	,	,	PUNCT
bjmsr-315	724	3	5	5	NUM
bjmsr-315	724	4	,	,	PUNCT
bjmsr-315	724	5	277	277	NUM
bjmsr-315	724	6	-	-	SYM
bjmsr-315	724	7	288	288	NUM
bjmsr-315	724	8	.	.	PUNCT
bjmsr-315	725	1	g.f	g.f	PROPN
bjmsr-315	725	2	.	.	PROPN
bjmsr-315	725	3	mccormick	mccormick	PROPN
bjmsr-315	725	4	,	,	PUNCT
bjmsr-315	725	5	v.a	v.a	PROPN
bjmsr-315	725	6	.	.	PROPN
bjmsr-315	725	7	sposito	sposito	PROPN
bjmsr-315	725	8	(	(	PUNCT
bjmsr-315	725	9	1975	1975	NUM
bjmsr-315	725	10	)	)	PUNCT
bjmsr-315	725	11	a	a	DET
bjmsr-315	725	12	note	note	NOUN
bjmsr-315	725	13	on	on	ADP
bjmsr-315	725	14	l1	l1	PROPN
bjmsr-315	725	15	estimation	estimation	NOUN
bjmsr-315	725	16	based	base	VERB
bjmsr-315	725	17	on	on	ADP
bjmsr-315	725	18	the	the	DET
bjmsr-315	725	19	median	median	ADJ
bjmsr-315	725	20	positive	positive	ADJ
bjmsr-315	725	21	quotient	quotient	NOUN
bjmsr-315	725	22	.	.	PUNCT
bjmsr-315	726	1	appl	appl	PROPN
bjmsr-315	726	2	.	.	PUNCT
bjmsr-315	727	1	stat	stat	PROPN
bjmsr-315	727	2	.	.	PUNCT
bjmsr-315	727	3	,	,	PUNCT
bjmsr-315	727	4	24	24	NUM
bjmsr-315	727	5	,	,	PUNCT
bjmsr-315	727	6	347350	347350	NUM
bjmsr-315	727	7	.	.	PUNCT
bjmsr-315	728	1	m.s	m.s	PROPN
bjmsr-315	728	2	.	.	PROPN
bjmsr-315	728	3	meketon	meketon	PROPN
bjmsr-315	728	4	(	(	PUNCT
bjmsr-315	728	5	1986	1986	NUM
bjmsr-315	728	6	)	)	PUNCT
bjmsr-315	728	7	least	least	ADJ
bjmsr-315	728	8	absolute	absolute	ADJ
bjmsr-315	728	9	value	value	NOUN
bjmsr-315	728	10	regression	regression	NOUN
bjmsr-315	728	11	.	.	PUNCT
bjmsr-315	729	1	work	work	NOUN
bjmsr-315	729	2	.	.	PUNCT
bjmsr-315	730	1	pap	pap	NOUN
bjmsr-315	730	2	.	.	PROPN
bjmsr-315	730	3	,	,	PUNCT
bjmsr-315	730	4	at&t	at&t	PROPN
bjmsr-315	730	5	bell	bell	PROPN
bjmsr-315	730	6	laboratories	laboratory	NOUN
bjmsr-315	730	7	,	,	PUNCT
bjmsr-315	730	8	holmdel	holmdel	PROPN
bjmsr-315	730	9	,	,	PUNCT
bjmsr-315	730	10	n.j	n.j	PROPN
bjmsr-315	730	11	.	.	PROPN
bjmsr-315	730	12	r.m	r.m	PROPN
bjmsr-315	730	13	.	.	PROPN
bjmsr-315	730	14	moroney	moroney	PROPN
bjmsr-315	730	15	(	(	PUNCT
bjmsr-315	730	16	1961	1961	NUM
bjmsr-315	730	17	)	)	PUNCT
bjmsr-315	730	18	the	the	DET
bjmsr-315	730	19	haar	haar	PROPN
bjmsr-315	730	20	problem	problem	NOUN
bjmsr-315	730	21	in	in	ADP
bjmsr-315	730	22	l1	l1	PROPN
bjmsr-315	730	23	.	.	PUNCT
bjmsr-315	731	1	proc	proc	PROPN
bjmsr-315	731	2	.	.	PUNCT
bjmsr-315	732	1	amer	amer	PROPN
bjmsr-315	732	2	.	.	PUNCT
bjmsr-315	732	3	math	math	PROPN
bjmsr-315	732	4	.	.	PUNCT
bjmsr-315	733	1	soc	soc	PROPN
bjmsr-315	733	2	.	.	PUNCT
bjmsr-315	733	3	,	,	PUNCT
bjmsr-315	733	4	12	12	NUM
bjmsr-315	733	5	,	,	PUNCT
bjmsr-315	733	6	793	793	NUM
bjmsr-315	733	7	-	-	SYM
bjmsr-315	733	8	795	795	NUM
bjmsr-315	733	9	.	.	PUNCT
bjmsr-315	734	1	s.c	s.c	PROPN
bjmsr-315	734	2	.	.	PROPN
bjmsr-315	734	3	narula	narula	PROPN
bjmsr-315	734	4	,	,	PUNCT
bjmsr-315	734	5	j.f	j.f	PROPN
bjmsr-315	734	6	.	.	PROPN
bjmsr-315	734	7	wellington	wellington	PROPN
bjmsr-315	734	8	(	(	PUNCT
bjmsr-315	734	9	1985	1985	NUM
bjmsr-315	734	10	)	)	PUNCT
bjmsr-315	734	11	interior	interior	ADJ
bjmsr-315	734	12	analysis	analysis	NOUN
bjmsr-315	734	13	for	for	ADP
bjmsr-315	734	14	the	the	DET
bjmsr-315	734	15	minimum	minimum	ADJ
bjmsr-315	734	16	sum	sum	NOUN
bjmsr-315	734	17	of	of	ADP
bjmsr-315	734	18	absolute	absolute	ADJ
bjmsr-315	734	19	errors	error	NOUN
bjmsr-315	734	20	regression	regression	NOUN
bjmsr-315	734	21	.	.	PUNCT
bjmsr-315	735	1	technometrics	technometric	NOUN
bjmsr-315	735	2	,	,	PUNCT
bjmsr-315	735	3	27	27	NUM
bjmsr-315	735	4	,	,	PUNCT
bjmsr-315	735	5	181	181	NUM
bjmsr-315	735	6	-	-	SYM
bjmsr-315	735	7	188	188	NUM
bjmsr-315	735	8	.	.	PUNCT
bjmsr-315	736	1	s.c	s.c	PROPN
bjmsr-315	736	2	.	.	PROPN
bjmsr-315	736	3	narula	narula	PROPN
bjmsr-315	736	4	,	,	PUNCT
bjmsr-315	736	5	j.f	j.f	PROPN
bjmsr-315	736	6	.	.	PROPN
bjmsr-315	736	7	wellington	wellington	PROPN
bjmsr-315	736	8	(	(	PUNCT
bjmsr-315	736	9	1987	1987	NUM
bjmsr-315	736	10	)	)	PUNCT
bjmsr-315	736	11	an	an	DET
bjmsr-315	736	12	efficient	efficient	ADJ
bjmsr-315	736	13	algorithm	algorithm	NOUN
bjmsr-315	736	14	for	for	ADP
bjmsr-315	736	15	the	the	DET
bjmsr-315	736	16	msae	msae	NOUN
bjmsr-315	736	17	and	and	CCONJ
bjmsr-315	736	18	mmae	mmae	PROPN
bjmsr-315	736	19	regression	regression	NOUN
bjmsr-315	736	20	problems	problem	NOUN
bjmsr-315	736	21	.	.	PUNCT
bjmsr-315	737	1	work	work	NOUN
bjmsr-315	737	2	.	.	PUNCT
bjmsr-315	738	1	pap	pap	NOUN
bjmsr-315	738	2	.	.	PROPN
bjmsr-315	738	3	,	,	PUNCT
bjmsr-315	738	4	virginia	virginia	PROPN
bjmsr-315	738	5	commonwealth	commonwealth	PROPN
bjmsr-315	738	6	university	university	PROPN
bjmsr-315	738	7	,	,	PUNCT
bjmsr-315	738	8	richmond	richmond	PROPN
bjmsr-315	738	9	,	,	PUNCT
bjmsr-315	738	10	va	va	PROPN
bjmsr-315	738	11	23284	23284	NUM
bjmsr-315	738	12	.	.	PUNCT
bjmsr-315	739	1	m.r	m.r	PROPN
bjmsr-315	739	2	.	.	PROPN
bjmsr-315	739	3	osborne	osborne	PROPN
bjmsr-315	739	4	(	(	PUNCT
bjmsr-315	739	5	1987	1987	NUM
bjmsr-315	739	6	)	)	PUNCT
bjmsr-315	739	7	the	the	DET
bjmsr-315	739	8	reduced	reduce	VERB
bjmsr-315	739	9	gradient	gradient	ADJ
bjmsr-315	739	10	algorithm	algorithm	NOUN
bjmsr-315	739	11	.	.	PUNCT
bjmsr-315	740	1	in	in	ADP
bjmsr-315	740	2	y.	y.	PROPN
bjmsr-315	740	3	dodge	dodge	PROPN
bjmsr-315	740	4	(	(	PUNCT
bjmsr-315	740	5	ed	ed	NOUN
bjmsr-315	740	6	.	.	PUNCT
bjmsr-315	740	7	)	)	PUNCT
bjmsr-315	740	8	statistical	statistical	ADJ
bjmsr-315	740	9	data	datum	NOUN
bjmsr-315	740	10	analysis	analysis	NOUN
bjmsr-315	740	11	based	base	VERB
bjmsr-315	740	12	on	on	ADP
bjmsr-315	740	13	the	the	DET
bjmsr-315	740	14	l1	l1	PROPN
bjmsr-315	740	15	norm	norm	NOUN
bjmsr-315	740	16	and	and	CCONJ
bjmsr-315	740	17	related	related	ADJ
bjmsr-315	740	18	methods	method	NOUN
bjmsr-315	740	19	.	.	PUNCT
bjmsr-315	741	1	north	north	NOUN
bjmsr-315	741	2	-	-	PUNCT
bjmsr-315	741	3	holland	holland	PROPN
bjmsr-315	741	4	.	.	PUNCT
bjmsr-315	742	1	95	95	NUM
bjmsr-315	742	2	-	-	SYM
bjmsr-315	742	3	108	108	NUM
bjmsr-315	742	4	.	.	PUNCT
bjmsr-315	743	1	m.r	m.r	PROPN
bjmsr-315	743	2	.	.	PROPN
bjmsr-315	743	3	osborne	osborne	PROPN
bjmsr-315	743	4	,	,	PUNCT
bjmsr-315	743	5	s.a	s.a	PROPN
bjmsr-315	743	6	.	.	PROPN
bjmsr-315	743	7	pruess	pruess	PROPN
bjmsr-315	743	8	,	,	PUNCT
bjmsr-315	743	9	r.s	r.s	PROPN
bjmsr-315	743	10	.	.	PROPN
bjmsr-315	743	11	womersley	womersley	PROPN
bjmsr-315	743	12	(	(	PUNCT
bjmsr-315	743	13	1986	1986	NUM
bjmsr-315	743	14	)	)	PUNCT
bjmsr-315	743	15	concise	concise	ADJ
bjmsr-315	743	16	representation	representation	NOUN
bjmsr-315	743	17	of	of	ADP
bjmsr-315	743	18	generalized	generalized	ADJ
bjmsr-315	743	19	gradients	gradient	NOUN
bjmsr-315	743	20	.	.	PUNCT
bjmsr-315	744	1	j.	j.	PROPN
bjmsr-315	744	2	of	of	ADP
bjmsr-315	744	3	austra	austra	PROPN
bjmsr-315	744	4	.	.	PROPN
bjmsr-315	744	5	math	math	PROPN
bjmsr-315	744	6	.	.	PUNCT
bjmsr-315	745	1	soc	soc	PROPN
bjmsr-315	745	2	.	.	PUNCT
bjmsr-315	745	3	,	,	PUNCT
bjmsr-315	745	4	ser	ser	PROPN
bjmsr-315	745	5	.	.	PROPN
bjmsr-315	746	1	b	b	NUM
bjmsr-315	746	2	,	,	PUNCT
bjmsr-315	746	3	28	28	NUM
bjmsr-315	746	4	,	,	PUNCT
bjmsr-315	746	5	57	57	NUM
bjmsr-315	746	6	-	-	SYM
bjmsr-315	746	7	74	74	NUM
bjmsr-315	746	8	.	.	PUNCT
bjmsr-315	747	1	m.r	m.r	PROPN
bjmsr-315	747	2	.	.	PROPN
bjmsr-315	747	3	osborne	osborne	PROPN
bjmsr-315	747	4	,	,	PUNCT
bjmsr-315	747	5	g.a	g.a	PROPN
bjmsr-315	747	6	.	.	PROPN
bjmsr-315	747	7	watson	watson	PROPN
bjmsr-315	747	8	(	(	PUNCT
bjmsr-315	747	9	1985	1985	NUM
bjmsr-315	747	10	)	)	PUNCT
bjmsr-315	747	11	an	an	DET
bjmsr-315	747	12	analysis	analysis	NOUN
bjmsr-315	747	13	of	of	ADP
bjmsr-315	747	14	the	the	DET
bjmsr-315	747	15	total	total	ADJ
bjmsr-315	747	16	approximation	approximation	NOUN
bjmsr-315	747	17	problem	problem	NOUN
bjmsr-315	747	18	in	in	ADP
bjmsr-315	747	19	separable	separable	ADJ
bjmsr-315	747	20	norms	norm	NOUN
bjmsr-315	747	21	,	,	PUNCT
bjmsr-315	747	22	and	and	CCONJ
bjmsr-315	747	23	an	an	DET
bjmsr-315	747	24	algorithm	algorithm	NOUN
bjmsr-315	747	25	for	for	ADP
bjmsr-315	747	26	the	the	DET
bjmsr-315	747	27	total	total	ADJ
bjmsr-315	747	28	l1	l1	PROPN
bjmsr-315	747	29	problem	problem	NOUN
bjmsr-315	747	30	.	.	PUNCT
bjmsr-315	748	1	siam	siam	PROPN
bjmsr-315	748	2	j.	j.	PROPN
bjmsr-315	748	3	sci	sci	PROPN
bjmsr-315	748	4	.	.	PROPN
bjmsr-315	748	5	stat	stat	PROPN
bjmsr-315	748	6	.	.	PUNCT
bjmsr-315	749	1	comp	comp	PROPN
bjmsr-315	749	2	.	.	PUNCT
bjmsr-315	749	3	,	,	PUNCT
bjmsr-315	749	4	6	6	NUM
bjmsr-315	749	5	,	,	PUNCT
bjmsr-315	749	6	410	410	NUM
bjmsr-315	749	7	-	-	SYM
bjmsr-315	749	8	424	424	NUM
bjmsr-315	749	9	.	.	PUNCT
bjmsr-315	750	1	m.j	m.j	PROPN
bjmsr-315	750	2	.	.	PROPN
bjmsr-315	750	3	panik	panik	PROPN
bjmsr-315	750	4	(	(	PUNCT
bjmsr-315	750	5	1976	1976	NUM
bjmsr-315	750	6	)	)	PUNCT
bjmsr-315	750	7	classical	classical	ADJ
bjmsr-315	750	8	optimization	optimization	NOUN
bjmsr-315	750	9	:	:	PUNCT
bjmsr-315	750	10	foundation	foundation	NOUN
bjmsr-315	750	11	and	and	CCONJ
bjmsr-315	750	12	extensions	extension	NOUN
bjmsr-315	750	13	.	.	PUNCT
bjmsr-315	751	1	north	north	NOUN
bjmsr-315	751	2	-	-	PUNCT
bjmsr-315	751	3	holland	holland	PROPN
bjmsr-315	751	4	,	,	PUNCT
bjmsr-315	751	5	amsterdam	amsterdam	PROPN
bjmsr-315	751	6	.	.	PUNCT
bjmsr-315	752	1	u.	u.	PROPN
bjmsr-315	752	2	peters	peters	PROPN
bjmsr-315	752	3	,	,	PUNCT
bjmsr-315	752	4	c.	c.	PROPN
bjmsr-315	752	5	willms	willm	NOUN
bjmsr-315	752	6	(	(	PUNCT
bjmsr-315	752	7	1983	1983	NUM
bjmsr-315	752	8	)	)	PUNCT
bjmsr-315	752	9	upand	upand	VERB
bjmsr-315	752	10	down	down	ADV
bjmsr-315	752	11	-	-	PUNCT
bjmsr-315	752	12	dating	date	VERB
bjmsr-315	752	13	procedures	procedure	NOUN
bjmsr-315	752	14	for	for	ADP
bjmsr-315	752	15	linear	linear	PROPN
bjmsr-315	752	16	l1	l1	PROPN
bjmsr-315	752	17	regression	regression	NOUN
bjmsr-315	752	18	.	.	PUNCT
bjmsr-315	753	1	or	or	CCONJ
bjmsr-315	753	2	spektrum	spektrum	PROPN
bjmsr-315	753	3	5	5	NUM
bjmsr-315	753	4	,	,	PUNCT
bjmsr-315	753	5	229	229	NUM
bjmsr-315	753	6	-	-	SYM
bjmsr-315	753	7	239	239	NUM
bjmsr-315	753	8	.	.	PUNCT
bjmsr-315	754	1	p.	p.	NOUN
bjmsr-315	754	2	pilibossian	pilibossian	NOUN
bjmsr-315	754	3	(	(	PUNCT
bjmsr-315	754	4	1987	1987	NUM
bjmsr-315	754	5	)	)	PUNCT
bjmsr-315	754	6	a	a	DET
bjmsr-315	754	7	direct	direct	ADJ
bjmsr-315	754	8	solving	solving	NOUN
bjmsr-315	754	9	algorithm	algorithm	NOUN
bjmsr-315	754	10	for	for	ADP
bjmsr-315	754	11	a	a	DET
bjmsr-315	754	12	linear	linear	ADJ
bjmsr-315	754	13	regression	regression	NOUN
bjmsr-315	754	14	according	accord	VERB
bjmsr-315	754	15	to	to	ADP
bjmsr-315	754	16	l1	l1	PROPN
bjmsr-315	754	17	-	-	PUNCT
bjmsr-315	754	18	norm	norm	NOUN
bjmsr-315	754	19	criteria	criterion	NOUN
bjmsr-315	754	20	.	.	PUNCT
bjmsr-315	755	1	work	work	NOUN
bjmsr-315	755	2	.	.	PUNCT
bjmsr-315	756	1	pap	pap	NOUN
bjmsr-315	756	2	.	.	PROPN
bjmsr-315	756	3	,	,	PUNCT
bjmsr-315	756	4	l.s.t.a	l.s.t.a	PROPN
bjmsr-315	756	5	.	.	PROPN
bjmsr-315	756	6	universite	universite	PROPN
bjmsr-315	756	7	,	,	PUNCT
bjmsr-315	756	8	paris	paris	PROPN
bjmsr-315	756	9	vi	vi	PROPN
bjmsr-315	757	1	p.	p.	NOUN
bjmsr-315	757	2	rabinowitz	rabinowitz	PROPN
bjmsr-315	757	3	(	(	PUNCT
bjmsr-315	757	4	1968	1968	NUM
bjmsr-315	757	5	)	)	PUNCT
bjmsr-315	757	6	application	application	NOUN
bjmsr-315	757	7	of	of	ADP
bjmsr-315	757	8	linear	linear	PROPN
bjmsr-315	757	9	programming	programming	NOUN
bjmsr-315	757	10	to	to	ADP
bjmsr-315	757	11	numerical	numerical	ADJ
bjmsr-315	757	12	analysis	analysis	NOUN
bjmsr-315	757	13	.	.	PUNCT
bjmsr-315	758	1	siam	siam	PROPN
bjmsr-315	758	2	rev	rev	PROPN
bjmsr-315	758	3	.	.	PROPN
bjmsr-315	758	4	,	,	PUNCT
bjmsr-315	758	5	10	10	NUM
bjmsr-315	758	6	,	,	PUNCT
bjmsr-315	758	7	121	121	NUM
bjmsr-315	758	8	-	-	SYM
bjmsr-315	758	9	159	159	NUM
bjmsr-315	758	10	.	.	PUNCT
bjmsr-315	759	1	p.	p.	NOUN
bjmsr-315	759	2	rabinowitz	rabinowitz	PROPN
bjmsr-315	759	3	(	(	PUNCT
bjmsr-315	759	4	1970	1970	NUM
bjmsr-315	759	5	)	)	PUNCT
bjmsr-315	759	6	mathematical	mathematical	ADJ
bjmsr-315	759	7	programming	programming	NOUN
bjmsr-315	759	8	and	and	CCONJ
bjmsr-315	759	9	approximation	approximation	NOUN
bjmsr-315	759	10	.	.	PUNCT
bjmsr-315	760	1	in	in	ADP
bjmsr-315	760	2	a.	a.	PROPN
bjmsr-315	760	3	talbot	talbot	PROPN
bjmsr-315	760	4	(	(	PUNCT
bjmsr-315	760	5	ed	ed	NOUN
bjmsr-315	760	6	.	.	PUNCT
bjmsr-315	760	7	)	)	PUNCT
bjmsr-315	761	1	approximation	approximation	NOUN
bjmsr-315	761	2	theory	theory	NOUN
bjmsr-315	761	3	.	.	PUNCT
bjmsr-315	762	1	academic	academic	ADJ
bjmsr-315	762	2	press	press	NOUN
bjmsr-315	762	3	,	,	PUNCT
bjmsr-315	762	4	217	217	NUM
bjmsr-315	762	5	-	-	SYM
bjmsr-315	762	6	231	231	NUM
bjmsr-315	762	7	.	.	PUNCT
bjmsr-315	763	1	m.r	m.r	PROPN
bjmsr-315	763	2	.	.	PROPN
bjmsr-315	763	3	rao	rao	PROPN
bjmsr-315	763	4	,	,	PUNCT
bjmsr-315	763	5	v.	v.	PROPN
bjmsr-315	763	6	srinivasan	srinivasan	NOUN
bjmsr-315	763	7	(	(	PUNCT
bjmsr-315	763	8	1972	1972	NUM
bjmsr-315	763	9	)	)	PUNCT
bjmsr-315	763	10	a	a	DET
bjmsr-315	763	11	note	note	NOUN
bjmsr-315	763	12	on	on	ADP
bjmsr-315	763	13	sharpe	sharpe	PROPN
bjmsr-315	763	14	's	's	PART
bjmsr-315	763	15	algorithm	algorithm	NOUN
bjmsr-315	763	16	for	for	ADP
bjmsr-315	763	17	minimum	minimum	ADJ
bjmsr-315	763	18	sum	sum	NOUN
bjmsr-315	763	19	of	of	ADP
bjmsr-315	763	20	absolute	absolute	ADJ
bjmsr-315	763	21	deviations	deviation	NOUN
bjmsr-315	763	22	in	in	ADP
bjmsr-315	763	23	a	a	DET
bjmsr-315	763	24	simple	simple	ADJ
bjmsr-315	763	25	regression	regression	NOUN
bjmsr-315	763	26	problem	problem	NOUN
bjmsr-315	763	27	.	.	PUNCT
bjmsr-315	764	1	manag	manag	PROPN
bjmsr-315	764	2	.	.	PUNCT
bjmsr-315	765	1	sci	sci	PROPN
bjmsr-315	765	2	.	.	PROPN
bjmsr-315	765	3	,	,	PUNCT
bjmsr-315	765	4	19	19	NUM
bjmsr-315	765	5	,	,	PUNCT
bjmsr-315	765	6	222	222	NUM
bjmsr-315	765	7	-	-	SYM
bjmsr-315	765	8	225	225	NUM
bjmsr-315	765	9	.	.	PUNCT
bjmsr-315	766	1	e.c	e.c	PROPN
bjmsr-315	766	2	.	.	PROPN
bjmsr-315	766	3	rhodes	rhodes	PROPN
bjmsr-315	766	4	(	(	PUNCT
bjmsr-315	766	5	1930	1930	NUM
bjmsr-315	766	6	)	)	PUNCT
bjmsr-315	766	7	reducing	reduce	VERB
bjmsr-315	766	8	observations	observation	NOUN
bjmsr-315	766	9	by	by	ADP
bjmsr-315	766	10	the	the	DET
bjmsr-315	766	11	method	method	NOUN
bjmsr-315	766	12	of	of	ADP
bjmsr-315	766	13	minimum	minimum	NOUN
bjmsr-315	766	14	deviations	deviation	NOUN
bjmsr-315	766	15	.	.	PUNCT
bjmsr-315	767	1	philo	philo	PROPN
bjmsr-315	767	2	.	.	PUNCT
bjmsr-315	768	1	mag	mag	INTJ
bjmsr-315	768	2	.	.	PROPN
bjmsr-315	768	3	,	,	PUNCT
bjmsr-315	768	4	7th	7th	ADJ
bjmsr-315	768	5	series	series	NOUN
bjmsr-315	768	6	,	,	PUNCT
bjmsr-315	768	7	9	9	NUM
bjmsr-315	768	8	,	,	PUNCT
bjmsr-315	768	9	974	974	NUM
bjmsr-315	768	10	-	-	SYM
bjmsr-315	768	11	92	92	NUM
bjmsr-315	768	12	.	.	PUNCT
bjmsr-315	769	1	j.r	j.r	PROPN
bjmsr-315	769	2	.	.	PROPN
bjmsr-315	769	3	rice	rice	PROPN
bjmsr-315	769	4	(	(	PUNCT
bjmsr-315	769	5	1964c	1964c	NUM
bjmsr-315	769	6	)	)	PUNCT
bjmsr-315	769	7	the	the	DET
bjmsr-315	769	8	approximation	approximation	NOUN
bjmsr-315	769	9	of	of	ADP
bjmsr-315	769	10	functions	function	NOUN
bjmsr-315	769	11	,	,	PUNCT
bjmsr-315	769	12	vol	vol	NOUN
bjmsr-315	769	13	.	.	PUNCT
bjmsr-315	770	1	i	i	PRON
bjmsr-315	770	2	,	,	PUNCT
bjmsr-315	770	3	linear	linear	PROPN
bjmsr-315	770	4	theory	theory	NOUN
bjmsr-315	770	5	.	.	PUNCT
bjmsr-315	771	1	reading	read	VERB
bjmsr-315	771	2	mass	mass	PROPN
bjmsr-315	771	3	:	:	PUNCT
bjmsr-315	771	4	,	,	PUNCT
bjmsr-315	771	5	addison	addison	PROPN
bjmsr-315	771	6	-	-	PUNCT
bjmsr-315	771	7	wesley	wesley	PROPN
bjmsr-315	771	8	.	.	PUNCT
bjmsr-315	772	1	copyright	copyright	NOUN
bjmsr-315	772	2	©	©	PROPN
bjmsr-315	772	3	cc	cc	PROPN
bjmsr-315	772	4	-	-	PUNCT
bjmsr-315	772	5	by	by	ADP
bjmsr-315	772	6	-	-	PUNCT
bjmsr-315	772	7	nc	nc	PROPN
bjmsr-315	772	8	2019	2019	NUM
bjmsr-315	772	9	,	,	PUNCT
bjmsr-315	772	10	bjmsr	bjmsr	PROPN
bjmsr-315	772	11	www.cribfb.com/journal/index.php/bjmsr	www.cribfb.com/journal/index.php/bjmsr	X
bjmsr-315	772	12	bangladesh	bangladesh	PROPN
bjmsr-315	772	13	journal	journal	PROPN
bjmsr-315	772	14	of	of	ADP
bjmsr-315	772	15	multidisciplinary	multidisciplinary	ADJ
bjmsr-315	772	16	scientific	scientific	ADJ
bjmsr-315	772	17	research	research	NOUN
bjmsr-315	772	18	vol	vol	NOUN
bjmsr-315	772	19	.	.	PROPN
bjmsr-315	773	1	1	1	NUM
bjmsr-315	773	2	,	,	PUNCT
bjmsr-315	773	3	no	no	INTJ
bjmsr-315	773	4	.	.	NOUN
bjmsr-315	773	5	1	1	NUM
bjmsr-315	773	6	;	;	PUNCT
bjmsr-315	773	7	2019	2019	NUM
bjmsr-315	773	8	67	67	NUM
bjmsr-315	773	9	p.d	p.d	PROPN
bjmsr-315	773	10	.	.	PROPN
bjmsr-315	773	11	robers	rober	NOUN
bjmsr-315	773	12	,	,	PUNCT
bjmsr-315	773	13	a.	a.	PROPN
bjmsr-315	773	14	ben	ben	PROPN
bjmsr-315	773	15	-	-	PROPN
bjmsr-315	773	16	israel	israel	PROPN
bjmsr-315	773	17	(	(	PUNCT
bjmsr-315	773	18	1969	1969	NUM
bjmsr-315	773	19	)	)	PUNCT
bjmsr-315	773	20	an	an	DET
bjmsr-315	773	21	interval	interval	NOUN
bjmsr-315	773	22	programming	programming	NOUN
bjmsr-315	773	23	algorithm	algorithm	NOUN
bjmsr-315	773	24	for	for	ADP
bjmsr-315	773	25	discrete	discrete	ADJ
bjmsr-315	773	26	linear	linear	PROPN
bjmsr-315	773	27	l1	l1	PROPN
bjmsr-315	773	28	approximation	approximation	NOUN
bjmsr-315	773	29	problem	problem	NOUN
bjmsr-315	773	30	.	.	PUNCT
bjmsr-315	774	1	j.	j.	PROPN
bjmsr-315	774	2	approx	approx	PROPN
bjmsr-315	774	3	.	.	PUNCT
bjmsr-315	775	1	theory	theory	NOUN
bjmsr-315	775	2	,	,	PUNCT
bjmsr-315	775	3	2	2	NUM
bjmsr-315	775	4	,	,	PUNCT
bjmsr-315	775	5	323	323	NUM
bjmsr-315	775	6	-	-	SYM
bjmsr-315	775	7	336	336	NUM
bjmsr-315	775	8	.	.	PUNCT
bjmsr-315	776	1	p.d	p.d	PROPN
bjmsr-315	776	2	.	.	PROPN
bjmsr-315	776	3	robers	rober	NOUN
bjmsr-315	776	4	,	,	PUNCT
bjmsr-315	776	5	s.s	s.s	PROPN
bjmsr-315	776	6	.	.	PROPN
bjmsr-315	776	7	robers	rober	NOUN
bjmsr-315	776	8	(	(	PUNCT
bjmsr-315	776	9	1973	1973	NUM
bjmsr-315	776	10	)	)	PUNCT
bjmsr-315	776	11	algorithm	algorithm	NOUN
bjmsr-315	776	12	458	458	NUM
bjmsr-315	776	13	:	:	PUNCT
bjmsr-315	776	14	discrete	discrete	ADJ
bjmsr-315	776	15	linear	linear	PROPN
bjmsr-315	776	16	l1	l1	PROPN
bjmsr-315	776	17	approximation	approximation	NOUN
bjmsr-315	776	18	by	by	ADP
bjmsr-315	776	19	interval	interval	NOUN
bjmsr-315	776	20	linear	linear	PROPN
bjmsr-315	776	21	programming	programming	NOUN
bjmsr-315	776	22	.	.	PUNCT
bjmsr-315	777	1	comm	comm	NOUN
bjmsr-315	777	2	.	.	PUNCT
bjmsr-315	777	3	acm	acm	PROPN
bjmsr-315	777	4	,	,	PUNCT
bjmsr-315	777	5	16	16	NUM
bjmsr-315	777	6	,	,	PUNCT
bjmsr-315	777	7	629	629	NUM
bjmsr-315	777	8	-	-	SYM
bjmsr-315	777	9	633	633	NUM
bjmsr-315	777	10	.	.	PUNCT
bjmsr-315	778	1	e.	e.	PROPN
bjmsr-315	778	2	ronchetti	ronchetti	PROPN
bjmsr-315	778	3	(	(	PUNCT
bjmsr-315	778	4	1987	1987	NUM
bjmsr-315	778	5	)	)	PUNCT
bjmsr-315	778	6	bounded	bound	VERB
bjmsr-315	778	7	influence	influence	NOUN
bjmsr-315	778	8	in	in	ADP
bjmsr-315	778	9	regression	regression	NOUN
bjmsr-315	778	10	:	:	PUNCT
bjmsr-315	778	11	a	a	DET
bjmsr-315	778	12	review	review	NOUN
bjmsr-315	778	13	.	.	PUNCT
bjmsr-315	779	1	in	in	ADP
bjmsr-315	779	2	y.	y.	PROPN
bjmsr-315	779	3	dodge	dodge	PROPN
bjmsr-315	779	4	(	(	PUNCT
bjmsr-315	779	5	ed	ed	NOUN
bjmsr-315	779	6	.	.	PUNCT
bjmsr-315	779	7	)	)	PUNCT
bjmsr-315	779	8	statistical	statistical	ADJ
bjmsr-315	779	9	data	datum	NOUN
bjmsr-315	779	10	analysis	analysis	NOUN
bjmsr-315	779	11	based	base	VERB
bjmsr-315	779	12	on	on	ADP
bjmsr-315	779	13	the	the	DET
bjmsr-315	779	14	l	l	NOUN
bjmsr-315	779	15	1	1	NUM
bjmsr-315	779	16	norm	norm	NOUN
bjmsr-315	779	17	and	and	CCONJ
bjmsr-315	779	18	related	related	ADJ
bjmsr-315	779	19	methods	method	NOUN
bjmsr-315	779	20	.	.	PUNCT
bjmsr-315	780	1	north	north	NOUN
bjmsr-315	780	2	-	-	PUNCT
bjmsr-315	780	3	holland	holland	PROPN
bjmsr-315	780	4	,	,	PUNCT
bjmsr-315	780	5	65	65	NUM
bjmsr-315	780	6	-	-	SYM
bjmsr-315	780	7	80	80	NUM
bjmsr-315	780	8	.	.	PUNCT
bjmsr-315	781	1	a.n	a.n	PROPN
bjmsr-315	781	2	.	.	PROPN
bjmsr-315	781	3	sadovski	sadovski	PROPN
bjmsr-315	781	4	(	(	PUNCT
bjmsr-315	781	5	1974	1974	NUM
bjmsr-315	781	6	)	)	PUNCT
bjmsr-315	782	1	as74	as74	PROPN
bjmsr-315	782	2	:	:	PUNCT
bjmsr-315	782	3	l1	l1	PROPN
bjmsr-315	782	4	-	-	PUNCT
bjmsr-315	782	5	norm	norm	NOUN
bjmsr-315	782	6	fit	fit	NOUN
bjmsr-315	782	7	of	of	ADP
bjmsr-315	782	8	a	a	DET
bjmsr-315	782	9	straight	straight	ADJ
bjmsr-315	782	10	line	line	NOUN
bjmsr-315	782	11	.	.	PUNCT
bjmsr-315	783	1	appl	appl	PROPN
bjmsr-315	783	2	.	.	PUNCT
bjmsr-315	784	1	stat	stat	PROPN
bjmsr-315	784	2	.	.	PUNCT
bjmsr-315	785	1	23	23	NUM
bjmsr-315	785	2	,	,	PUNCT
bjmsr-315	785	3	244	244	NUM
bjmsr-315	785	4	-	-	SYM
bjmsr-315	785	5	248	248	NUM
bjmsr-315	785	6	.	.	PUNCT
bjmsr-315	786	1	j.p	j.p	PROPN
bjmsr-315	786	2	.	.	PROPN
bjmsr-315	786	3	schellhorn	schellhorn	ADJ
bjmsr-315	786	4	(	(	PUNCT
bjmsr-315	786	5	1987	1987	NUM
bjmsr-315	786	6	)	)	PUNCT
bjmsr-315	786	7	fitting	fit	VERB
bjmsr-315	786	8	data	datum	NOUN
bjmsr-315	786	9	through	through	ADP
bjmsr-315	786	10	homotopy	homotopy	NOUN
bjmsr-315	786	11	methods	method	NOUN
bjmsr-315	786	12	in	in	ADP
bjmsr-315	786	13	y.	y.	PROPN
bjmsr-315	786	14	dodge	dodge	PROPN
bjmsr-315	786	15	(	(	PUNCT
bjmsr-315	786	16	ed	ed	NOUN
bjmsr-315	786	17	.	.	PUNCT
bjmsr-315	786	18	)	)	PUNCT
bjmsr-315	787	1	statistical	statistical	ADJ
bjmsr-315	787	2	data	datum	NOUN
bjmsr-315	787	3	analysis	analysis	NOUN
bjmsr-315	787	4	based	base	VERB
bjmsr-315	787	5	on	on	ADP
bjmsr-315	787	6	the	the	DET
bjmsr-315	787	7	l	l	NOUN
bjmsr-315	787	8	1	1	NUM
bjmsr-315	787	9	norm	norm	NOUN
bjmsr-315	787	10	and	and	CCONJ
bjmsr-315	787	11	related	related	ADJ
bjmsr-315	787	12	methods	method	NOUN
bjmsr-315	787	13	.	.	PUNCT
bjmsr-315	788	1	north	north	NOUN
bjmsr-315	788	2	-	-	PUNCT
bjmsr-315	788	3	holland	holland	PROPN
bjmsr-315	788	4	.	.	PUNCT
bjmsr-315	789	1	131	131	NUM
bjmsr-315	789	2	-	-	SYM
bjmsr-315	789	3	138	138	NUM
bjmsr-315	789	4	.	.	PUNCT
bjmsr-315	789	5	e.j	e.j	PROPN
bjmsr-315	789	6	.	.	PROPN
bjmsr-315	789	7	schlossmacher	schlossmacher	PROPN
bjmsr-315	789	8	(	(	PUNCT
bjmsr-315	789	9	1973	1973	NUM
bjmsr-315	789	10	)	)	PUNCT
bjmsr-315	789	11	an	an	DET
bjmsr-315	789	12	iterative	iterative	NOUN
bjmsr-315	789	13	technique	technique	NOUN
bjmsr-315	789	14	for	for	ADP
bjmsr-315	789	15	absolute	absolute	ADJ
bjmsr-315	789	16	deviations	deviation	NOUN
bjmsr-315	789	17	curve	curve	VERB
bjmsr-315	789	18	fitting	fitting	ADJ
bjmsr-315	789	19	.	.	PUNCT
bjmsr-315	790	1	jasa	jasa	PROPN
bjmsr-315	790	2	68	68	NUM
bjmsr-315	790	3	,	,	PUNCT
bjmsr-315	790	4	857	857	NUM
bjmsr-315	790	5	-	-	SYM
bjmsr-315	790	6	865	865	NUM
bjmsr-315	790	7	.	.	PUNCT
bjmsr-315	791	1	e.	e.	PROPN
bjmsr-315	791	2	seneta	seneta	PROPN
bjmsr-315	791	3	(	(	PUNCT
bjmsr-315	791	4	1983	1983	NUM
bjmsr-315	791	5	)	)	PUNCT
bjmsr-315	791	6	the	the	DET
bjmsr-315	791	7	weighted	weight	VERB
bjmsr-315	791	8	median	median	NOUN
bjmsr-315	791	9	and	and	CCONJ
bjmsr-315	791	10	multiple	multiple	ADJ
bjmsr-315	791	11	regression	regression	NOUN
bjmsr-315	791	12	.	.	PUNCT
bjmsr-315	792	1	austral	austral	ADJ
bjmsr-315	792	2	.	.	PUNCT
bjmsr-315	793	1	j.	j.	PROPN
bjmsr-315	793	2	stat	stat	PROPN
bjmsr-315	793	3	.	.	PUNCT
bjmsr-315	793	4	,	,	PUNCT
bjmsr-315	793	5	25(2	25(2	NUM
bjmsr-315	793	6	)	)	PUNCT
bjmsr-315	793	7	,	,	PUNCT
bjmsr-315	793	8	370	370	NUM
bjmsr-315	793	9	-	-	SYM
bjmsr-315	793	10	377	377	NUM
bjmsr-315	793	11	.	.	PUNCT
bjmsr-315	794	1	e.	e.	PROPN
bjmsr-315	794	2	seneta	seneta	PROPN
bjmsr-315	794	3	,	,	PUNCT
bjmsr-315	794	4	w.l	w.l	PROPN
bjmsr-315	794	5	.	.	PROPN
bjmsr-315	794	6	steiger	steiger	PROPN
bjmsr-315	794	7	(	(	PUNCT
bjmsr-315	794	8	1984	1984	NUM
bjmsr-315	794	9	)	)	PUNCT
bjmsr-315	794	10	a	a	DET
bjmsr-315	794	11	new	new	ADJ
bjmsr-315	794	12	lad	lad	NOUN
bjmsr-315	794	13	curve	curve	NOUN
bjmsr-315	794	14	-	-	PUNCT
bjmsr-315	794	15	fitting	fit	VERB
bjmsr-315	794	16	algorithm	algorithm	NOUN
bjmsr-315	794	17	:	:	PUNCT
bjmsr-315	794	18	slightly	slightly	ADV
bjmsr-315	794	19	overdetermined	overdetermine	VERB
bjmsr-315	794	20	equation	equation	NOUN
bjmsr-315	794	21	system	system	NOUN
bjmsr-315	794	22	in	in	ADP
bjmsr-315	794	23	l1	l1	PROPN
bjmsr-315	794	24	.	.	PUNCT
bjmsr-315	795	1	discrete	discrete	ADJ
bjmsr-315	795	2	applied	apply	VERB
bjmsr-315	795	3	math	math	NOUN
bjmsr-315	795	4	.	.	PUNCT
bjmsr-315	795	5	,	,	PUNCT
bjmsr-315	795	6	7	7	NUM
bjmsr-315	795	7	,	,	PUNCT
bjmsr-315	795	8	79	79	NUM
bjmsr-315	795	9	-	-	SYM
bjmsr-315	795	10	91	91	NUM
bjmsr-315	795	11	.	.	PUNCT
bjmsr-315	796	1	shanno	shanno	PROPN
bjmsr-315	796	2	,	,	PUNCT
bjmsr-315	796	3	r.l	r.l	PROPN
bjmsr-315	796	4	.	.	PROPN
bjmsr-315	796	5	weil	weil	PROPN
bjmsr-315	796	6	(	(	PUNCT
bjmsr-315	796	7	1970	1970	NUM
bjmsr-315	796	8	)	)	PUNCT
bjmsr-315	796	9	linear	linear	NOUN
bjmsr-315	796	10	programming	programming	NOUN
bjmsr-315	796	11	with	with	ADP
bjmsr-315	796	12	absolute	absolute	ADJ
bjmsr-315	796	13	value	value	NOUN
bjmsr-315	796	14	functionals	functional	NOUN
bjmsr-315	796	15	.	.	PUNCT
bjmsr-315	797	1	oper	oper	NOUN
bjmsr-315	797	2	.	.	PUNCT
bjmsr-315	797	3	res	res	PROPN
bjmsr-315	797	4	.	.	PROPN
bjmsr-315	797	5	,	,	PUNCT
bjmsr-315	797	6	19	19	NUM
bjmsr-315	797	7	,	,	PUNCT
bjmsr-315	797	8	120	120	NUM
bjmsr-315	797	9	-	-	SYM
bjmsr-315	797	10	124	124	NUM
bjmsr-315	797	11	.	.	PUNCT
bjmsr-315	798	1	w.f	w.f	PROPN
bjmsr-315	798	2	.	.	PROPN
bjmsr-315	798	3	sharpe	sharpe	PROPN
bjmsr-315	798	4	(	(	PUNCT
bjmsr-315	798	5	1971	1971	NUM
bjmsr-315	798	6	)	)	PUNCT
bjmsr-315	798	7	mean	mean	ADJ
bjmsr-315	798	8	-	-	PUNCT
bjmsr-315	798	9	absolute	absolute	ADJ
bjmsr-315	798	10	deviation	deviation	NOUN
bjmsr-315	798	11	characteristic	characteristic	ADJ
bjmsr-315	798	12	lines	line	NOUN
bjmsr-315	798	13	for	for	ADP
bjmsr-315	798	14	securities	security	NOUN
bjmsr-315	798	15	and	and	CCONJ
bjmsr-315	798	16	portfolios	portfolio	NOUN
bjmsr-315	798	17	.	.	PUNCT
bjmsr-315	799	1	manag	manag	PROPN
bjmsr-315	799	2	.	.	PUNCT
bjmsr-315	800	1	sci	sci	PROPN
bjmsr-315	800	2	.	.	PROPN
bjmsr-315	800	3	,	,	PUNCT
bjmsr-315	800	4	18	18	NUM
bjmsr-315	800	5	,	,	PUNCT
bjmsr-315	800	6	b1	b1	NOUN
bjmsr-315	800	7	-	-	PUNCT
bjmsr-315	800	8	b13	b13	NOUN
bjmsr-315	800	9	.	.	PUNCT
bjmsr-315	801	1	h.d	h.d	PROPN
bjmsr-315	801	2	.	.	PROPN
bjmsr-315	801	3	sherali	sherali	PROPN
bjmsr-315	801	4	,	,	PUNCT
bjmsr-315	801	5	b.o	b.o	PROPN
bjmsr-315	801	6	.	.	PROPN
bjmsr-315	801	7	skarpness	skarpness	PROPN
bjmsr-315	801	8	,	,	PUNCT
bjmsr-315	801	9	b.	b.	PROPN
bjmsr-315	801	10	kim	kim	PROPN
bjmsr-315	801	11	(	(	PUNCT
bjmsr-315	801	12	1987	1987	NUM
bjmsr-315	801	13	)	)	PUNCT
bjmsr-315	801	14	an	an	DET
bjmsr-315	801	15	assumption	assumption	NOUN
bjmsr-315	801	16	-	-	PUNCT
bjmsr-315	801	17	free	free	ADJ
bjmsr-315	801	18	convergence	convergence	NOUN
bjmsr-315	801	19	analysis	analysis	NOUN
bjmsr-315	801	20	for	for	ADP
bjmsr-315	801	21	a	a	DET
bjmsr-315	801	22	perturbation	perturbation	NOUN
bjmsr-315	801	23	of	of	ADP
bjmsr-315	801	24	the	the	DET
bjmsr-315	801	25	scaling	scale	VERB
bjmsr-315	801	26	algorithm	algorithm	NOUN
bjmsr-315	801	27	for	for	ADP
bjmsr-315	801	28	linear	linear	NOUN
bjmsr-315	801	29	programs	program	NOUN
bjmsr-315	801	30	,	,	PUNCT
bjmsr-315	801	31	with	with	ADP
bjmsr-315	801	32	application	application	NOUN
bjmsr-315	801	33	to	to	ADP
bjmsr-315	801	34	the	the	DET
bjmsr-315	801	35	l1	l1	PROPN
bjmsr-315	801	36	estimation	estimation	NOUN
bjmsr-315	801	37	problem	problem	NOUN
bjmsr-315	801	38	.	.	PUNCT
bjmsr-315	802	1	dept	dept	PROPN
bjmsr-315	802	2	.	.	PROPN
bjmsr-315	802	3	of	of	ADP
bjmsr-315	802	4	ind	ind	PROPN
bjmsr-315	802	5	.	.	PUNCT
bjmsr-315	803	1	engin	engin	PROPN
bjmsr-315	803	2	.	.	PUNCT
bjmsr-315	804	1	and	and	CCONJ
bjmsr-315	804	2	or	or	CCONJ
bjmsr-315	804	3	,	,	PUNCT
bjmsr-315	804	4	virginia	virginia	PROPN
bjmsr-315	804	5	polytechnic	polytechnic	PROPN
bjmsr-315	804	6	inst	inst	PROPN
bjmsr-315	804	7	.	.	PROPN
bjmsr-315	805	1	and	and	CCONJ
bjmsr-315	805	2	state	state	PROPN
bjmsr-315	805	3	university	university	PROPN
bjmsr-315	805	4	,	,	PUNCT
bjmsr-315	805	5	blacksburg	blacksburg	PROPN
bjmsr-315	805	6	,	,	PUNCT
bjmsr-315	805	7	virginia	virginia	PROPN
bjmsr-315	805	8	.	.	PUNCT
bjmsr-315	806	1	o.b	o.b	PROPN
bjmsr-315	806	2	.	.	PROPN
bjmsr-315	806	3	sheynin	sheynin	PROPN
bjmsr-315	806	4	(	(	PUNCT
bjmsr-315	806	5	1973	1973	NUM
bjmsr-315	806	6	)	)	PUNCT
bjmsr-315	806	7	r.j	r.j	PROPN
bjmsr-315	806	8	.	.	PROPN
bjmsr-315	806	9	boscovich	boscovich	PROPN
bjmsr-315	806	10	's	's	PART
bjmsr-315	806	11	work	work	NOUN
bjmsr-315	806	12	on	on	ADP
bjmsr-315	806	13	probability	probability	NOUN
bjmsr-315	806	14	.	.	PUNCT
bjmsr-315	807	1	archive	archive	NOUN
bjmsr-315	807	2	for	for	ADP
bjmsr-315	807	3	history	history	NOUN
bjmsr-315	807	4	of	of	ADP
bjmsr-315	807	5	exact	exact	ADJ
bjmsr-315	807	6	sciences	science	NOUN
bjmsr-315	807	7	,	,	PUNCT
bjmsr-315	807	8	vol	vol	NOUN
bjmsr-315	807	9	.	.	NOUN
bjmsr-315	807	10	9	9	NUM
bjmsr-315	807	11	,	,	PUNCT
bjmsr-315	807	12	306	306	NUM
bjmsr-315	807	13	-	-	SYM
bjmsr-315	807	14	324	324	NUM
bjmsr-315	807	15	,	,	PUNCT
bjmsr-315	807	16	and	and	CCONJ
bjmsr-315	807	17	vol	vol	NOUN
bjmsr-315	807	18	.	.	PROPN
bjmsr-315	807	19	28	28	NUM
bjmsr-315	807	20	,	,	PUNCT
bjmsr-315	807	21	173	173	NUM
bjmsr-315	807	22	.	.	PUNCT
bjmsr-315	808	1	r.r	r.r	PROPN
bjmsr-315	808	2	.	.	PROPN
bjmsr-315	808	3	singleton	singleton	PROPN
bjmsr-315	808	4	(	(	PUNCT
bjmsr-315	808	5	1940	1940	NUM
bjmsr-315	808	6	)	)	PUNCT
bjmsr-315	808	7	a	a	DET
bjmsr-315	808	8	method	method	NOUN
bjmsr-315	808	9	for	for	ADP
bjmsr-315	808	10	minimizing	minimize	VERB
bjmsr-315	808	11	the	the	DET
bjmsr-315	808	12	sum	sum	NOUN
bjmsr-315	808	13	of	of	ADP
bjmsr-315	808	14	absolute	absolute	ADJ
bjmsr-315	808	15	values	value	NOUN
bjmsr-315	808	16	of	of	ADP
bjmsr-315	808	17	deviations	deviation	NOUN
bjmsr-315	808	18	.	.	PUNCT
bjmsr-315	809	1	annals	annal	NOUN
bjmsr-315	809	2	of	of	ADP
bjmsr-315	809	3	math	math	NOUN
bjmsr-315	809	4	.	.	PUNCT
bjmsr-315	810	1	stat	stat	PROPN
bjmsr-315	810	2	.	.	PUNCT
bjmsr-315	810	3	,	,	PUNCT
bjmsr-315	810	4	11	11	NUM
bjmsr-315	810	5	,	,	PUNCT
bjmsr-315	810	6	301	301	NUM
bjmsr-315	810	7	-	-	SYM
bjmsr-315	810	8	310	310	NUM
bjmsr-315	810	9	.	.	PUNCT
bjmsr-315	811	1	s.a	s.a	PROPN
bjmsr-315	811	2	.	.	PROPN
bjmsr-315	811	3	soliman	soliman	PROPN
bjmsr-315	811	4	,	,	PUNCT
bjmsr-315	811	5	g.s	g.s	PROPN
bjmsr-315	811	6	.	.	PROPN
bjmsr-315	811	7	christensen	christensen	PROPN
bjmsr-315	811	8	,	,	PUNCT
bjmsr-315	811	9	a.	a.	NOUN
bjmsr-315	811	10	rouhi	rouhi	NOUN
bjmsr-315	811	11	(	(	PUNCT
bjmsr-315	811	12	1988	1988	NUM
bjmsr-315	811	13	)	)	PUNCT
bjmsr-315	811	14	a	a	DET
bjmsr-315	811	15	new	new	ADJ
bjmsr-315	811	16	technique	technique	NOUN
bjmsr-315	811	17	for	for	ADP
bjmsr-315	811	18	curve	curve	NOUN
bjmsr-315	811	19	fitting	fitting	NOUN
bjmsr-315	811	20	based	base	VERB
bjmsr-315	811	21	on	on	ADP
bjmsr-315	811	22	minimum	minimum	ADJ
bjmsr-315	811	23	absolute	absolute	ADJ
bjmsr-315	811	24	deviations	deviation	NOUN
bjmsr-315	811	25	.	.	PUNCT
bjmsr-315	812	1	csda	csda	NOUN
bjmsr-315	812	2	,	,	PUNCT
bjmsr-315	812	3	6(4	6(4	PROPN
bjmsr-315	812	4	)	)	PUNCT
bjmsr-315	812	5	,	,	PUNCT
bjmsr-315	812	6	341	341	NUM
bjmsr-315	812	7	-	-	SYM
bjmsr-315	812	8	352	352	NUM
bjmsr-315	812	9	.	.	PUNCT
bjmsr-315	813	1	v.a	v.a	PROPN
bjmsr-315	813	2	.	.	PROPN
bjmsr-315	813	3	sposito	sposito	PROPN
bjmsr-315	813	4	(	(	PUNCT
bjmsr-315	813	5	1976	1976	NUM
bjmsr-315	813	6	)	)	PUNCT
bjmsr-315	813	7	a	a	DET
bjmsr-315	813	8	remark	remark	NOUN
bjmsr-315	813	9	on	on	ADP
bjmsr-315	813	10	algorithm	algorithm	NOUN
bjmsr-315	813	11	as74	as74	PROPN
bjmsr-315	813	12	,	,	PUNCT
bjmsr-315	813	13	l1	l1	PROPN
bjmsr-315	813	14	norm	norm	PROPN
bjmsr-315	813	15	fit	fit	NOUN
bjmsr-315	813	16	of	of	ADP
bjmsr-315	813	17	a	a	DET
bjmsr-315	813	18	straight	straight	ADJ
bjmsr-315	813	19	line	line	NOUN
bjmsr-315	813	20	.	.	PUNCT
bjmsr-315	814	1	appl	appl	PROPN
bjmsr-315	814	2	.	.	PUNCT
bjmsr-315	815	1	stat	stat	PROPN
bjmsr-315	815	2	.	.	PUNCT
bjmsr-315	815	3	,	,	PUNCT
bjmsr-315	815	4	25	25	NUM
bjmsr-315	815	5	,	,	PUNCT
bjmsr-315	815	6	96	96	NUM
bjmsr-315	815	7	-	-	SYM
bjmsr-315	815	8	97	97	NUM
bjmsr-315	815	9	.	.	PUNCT
bjmsr-315	816	1	v.a	v.a	PROPN
bjmsr-315	816	2	.	.	PROPN
bjmsr-315	816	3	sposito	sposito	PROPN
bjmsr-315	816	4	(	(	PUNCT
bjmsr-315	816	5	1987a	1987a	NUM
bjmsr-315	816	6	)	)	PUNCT
bjmsr-315	816	7	on	on	ADP
bjmsr-315	816	8	median	median	ADJ
bjmsr-315	816	9	polish	polish	NOUN
bjmsr-315	816	10	and	and	CCONJ
bjmsr-315	816	11	l1	l1	PROPN
bjmsr-315	816	12	estimators.csda	estimators.csda	PROPN
bjmsr-315	816	13	,	,	PUNCT
bjmsr-315	816	14	5	5	NUM
bjmsr-315	816	15	,	,	PUNCT
bjmsr-315	816	16	155	155	NUM
bjmsr-315	816	17	-	-	SYM
bjmsr-315	816	18	162	162	NUM
bjmsr-315	816	19	.	.	PUNCT
bjmsr-315	817	1	v.a	v.a	PROPN
bjmsr-315	817	2	.	.	PROPN
bjmsr-315	817	3	sposito	sposito	PROPN
bjmsr-315	817	4	,	,	PUNCT
bjmsr-315	817	5	w.j	w.j	PROPN
bjmsr-315	817	6	.	.	PROPN
bjmsr-315	817	7	kennedy	kennedy	PROPN
bjmsr-315	817	8	,	,	PUNCT
bjmsr-315	817	9	j.e	j.e	PROPN
bjmsr-315	817	10	.	.	PROPN
bjmsr-315	817	11	gentle	gentle	ADJ
bjmsr-315	817	12	(	(	PUNCT
bjmsr-315	817	13	1977	1977	NUM
bjmsr-315	817	14	)	)	PUNCT
bjmsr-315	817	15	as110	as110	PROPN
bjmsr-315	817	16	:	:	PUNCT
bjmsr-315	817	17	lp	lp	PROPN
bjmsr-315	817	18	norm	norm	NOUN
bjmsr-315	817	19	fit	fit	NOUN
bjmsr-315	817	20	of	of	ADP
bjmsr-315	817	21	a	a	DET
bjmsr-315	817	22	straight	straight	ADJ
bjmsr-315	817	23	line	line	NOUN
bjmsr-315	817	24	.	.	PUNCT
bjmsr-315	818	1	appl	appl	PROPN
bjmsr-315	818	2	.	.	PUNCT
bjmsr-315	819	1	stat	stat	PROPN
bjmsr-315	819	2	.	.	PUNCT
bjmsr-315	819	3	,	,	PUNCT
bjmsr-315	819	4	26	26	NUM
bjmsr-315	819	5	,	,	PUNCT
bjmsr-315	819	6	114	114	NUM
bjmsr-315	819	7	-	-	SYM
bjmsr-315	819	8	118	118	NUM
bjmsr-315	819	9	.	.	PUNCT
bjmsr-315	820	1	v.a	v.a	PROPN
bjmsr-315	820	2	.	.	PROPN
bjmsr-315	820	3	sposito	sposito	PROPN
bjmsr-315	820	4	,	,	PUNCT
bjmsr-315	820	5	g.f	g.f	PROPN
bjmsr-315	820	6	.	.	PROPN
bjmsr-315	820	7	mccormick	mccormick	PROPN
bjmsr-315	820	8	,	,	PUNCT
bjmsr-315	820	9	w.j	w.j	PROPN
bjmsr-315	820	10	.	.	PROPN
bjmsr-315	820	11	kennedy	kennedy	PROPN
bjmsr-315	820	12	(	(	PUNCT
bjmsr-315	820	13	1975	1975	NUM
bjmsr-315	820	14	)	)	PUNCT
bjmsr-315	820	15	l1	l1	PROPN
bjmsr-315	820	16	estimation	estimation	NOUN
bjmsr-315	820	17	strategies	strategy	NOUN
bjmsr-315	820	18	based	base	VERB
bjmsr-315	820	19	on	on	ADP
bjmsr-315	820	20	the	the	DET
bjmsr-315	820	21	simplex	simplex	NOUN
bjmsr-315	820	22	algorithm	algorithm	NOUN
bjmsr-315	820	23	.	.	PUNCT
bjmsr-315	821	1	in	in	ADP
bjmsr-315	821	2	proc	proc	PROPN
bjmsr-315	821	3	.	.	PUNCT
bjmsr-315	822	1	of	of	ADP
bjmsr-315	822	2	the	the	DET
bjmsr-315	822	3	eighth	eighth	ADJ
bjmsr-315	822	4	symposium	symposium	NOUN
bjmsr-315	822	5	on	on	ADP
bjmsr-315	822	6	the	the	DET
bjmsr-315	822	7	interface	interface	NOUN
bjmsr-315	822	8	,	,	PUNCT
bjmsr-315	822	9	j.w	j.w	PROPN
bjmsr-315	822	10	.	.	PROPN
bjmsr-315	822	11	france	france	PROPN
bjmsr-315	822	12	(	(	PUNCT
bjmsr-315	822	13	ed	ed	NOUN
bjmsr-315	822	14	.	.	PUNCT
bjmsr-315	822	15	)	)	PUNCT
bjmsr-315	823	1	health	health	NOUN
bjmsr-315	823	2	science	science	NOUN
bjmsr-315	823	3	computing	computing	PROPN
bjmsr-315	823	4	facility	facility	NOUN
bjmsr-315	823	5	.	.	PUNCT
bjmsr-315	824	1	ucla	ucla	PROPN
bjmsr-315	824	2	,	,	PUNCT
bjmsr-315	824	3	los	los	PROPN
bjmsr-315	824	4	angeles	angeles	PROPN
bjmsr-315	824	5	.	.	PUNCT
bjmsr-315	825	1	v.a	v.a	PROPN
bjmsr-315	825	2	.	.	PROPN
bjmsr-315	825	3	sposito	sposito	PROPN
bjmsr-315	825	4	,	,	PUNCT
bjmsr-315	825	5	w.c	w.c	PROPN
bjmsr-315	825	6	.	.	PROPN
bjmsr-315	825	7	smith	smith	PROPN
bjmsr-315	825	8	(	(	PUNCT
bjmsr-315	825	9	1976	1976	NUM
bjmsr-315	825	10	)	)	PUNCT
bjmsr-315	825	11	on	on	ADP
bjmsr-315	825	12	a	a	DET
bjmsr-315	825	13	sufficient	sufficient	ADJ
bjmsr-315	825	14	and	and	CCONJ
bjmsr-315	825	15	necessary	necessary	ADJ
bjmsr-315	825	16	condition	condition	NOUN
bjmsr-315	825	17	for	for	ADP
bjmsr-315	825	18	l1	l1	PROPN
bjmsr-315	825	19	estimation	estimation	PROPN
bjmsr-315	825	20	.	.	PUNCT
bjmsr-315	826	1	appl	appl	PROPN
bjmsr-315	826	2	.	.	PUNCT
bjmsr-315	826	3	stat	stat	PROPN
bjmsr-315	826	4	.	.	PUNCT
bjmsr-315	826	5	,	,	PUNCT
bjmsr-315	826	6	25	25	NUM
bjmsr-315	826	7	,	,	PUNCT
bjmsr-315	826	8	154	154	NUM
bjmsr-315	826	9	-	-	SYM
bjmsr-315	826	10	157	157	NUM
bjmsr-315	826	11	.	.	PUNCT
bjmsr-315	827	1	k.	k.	PROPN
bjmsr-315	827	2	spyropoulos	spyropoulos	PROPN
bjmsr-315	827	3	,	,	PUNCT
bjmsr-315	827	4	e.	e.	PROPN
bjmsr-315	827	5	kiountouzis	kiountouzis	PROPN
bjmsr-315	827	6	,	,	PUNCT
bjmsr-315	827	7	a.	a.	PROPN
bjmsr-315	827	8	young	young	PROPN
bjmsr-315	827	9	(	(	PUNCT
bjmsr-315	827	10	1973	1973	NUM
bjmsr-315	827	11	)	)	PUNCT
bjmsr-315	827	12	discrete	discrete	ADJ
bjmsr-315	827	13	approximation	approximation	NOUN
bjmsr-315	827	14	in	in	ADP
bjmsr-315	827	15	the	the	DET
bjmsr-315	827	16	l1	l1	PROPN
bjmsr-315	827	17	norm	norm	NOUN
bjmsr-315	827	18	.	.	PUNCT
bjmsr-315	828	1	comp	comp	PROPN
bjmsr-315	828	2	.	.	PUNCT
bjmsr-315	829	1	j.	j.	PROPN
bjmsr-315	829	2	,	,	PUNCT
bjmsr-315	829	3	16	16	NUM
bjmsr-315	829	4	,	,	PUNCT
bjmsr-315	829	5	180	180	NUM
bjmsr-315	829	6	-	-	SYM
bjmsr-315	829	7	186	186	NUM
bjmsr-315	829	8	.	.	PUNCT
bjmsr-315	830	1	w.l	w.l	PROPN
bjmsr-315	830	2	.	.	PROPN
bjmsr-315	830	3	steiger	steiger	PROPN
bjmsr-315	830	4	(	(	PUNCT
bjmsr-315	830	5	1980	1980	NUM
bjmsr-315	830	6	)	)	PUNCT
bjmsr-315	830	7	linear	linear	NOUN
bjmsr-315	830	8	programming	programming	NOUN
bjmsr-315	830	9	via	via	ADP
bjmsr-315	830	10	l1	l1	PROPN
bjmsr-315	830	11	curve	curve	PROPN
bjmsr-315	830	12	fitting	fitting	ADJ
bjmsr-315	830	13	beats	beat	NOUN
bjmsr-315	830	14	simplex	simplex	NOUN
bjmsr-315	830	15	.	.	PUNCT
bjmsr-315	831	1	abstracts	abstract	NOUN
bjmsr-315	831	2	,	,	PUNCT
bjmsr-315	831	3	ams	am	NOUN
bjmsr-315	831	4	,	,	PUNCT
bjmsr-315	831	5	80t	80t	NOUN
bjmsr-315	831	6	-	-	NOUN
bjmsr-315	831	7	c26	c26	NOUN
bjmsr-315	831	8	,	,	PUNCT
bjmsr-315	831	9	385	385	NUM
bjmsr-315	831	10	-	-	SYM
bjmsr-315	831	11	386	386	NUM
bjmsr-315	831	12	.	.	PUNCT
bjmsr-315	832	1	s.m	s.m	PROPN
bjmsr-315	832	2	.	.	PROPN
bjmsr-315	832	3	stigler	stigler	NOUN
bjmsr-315	832	4	(	(	PUNCT
bjmsr-315	832	5	1981	1981	NUM
bjmsr-315	832	6	)	)	PUNCT
bjmsr-315	832	7	gauss	gauss	NOUN
bjmsr-315	832	8	and	and	CCONJ
bjmsr-315	832	9	invention	invention	NOUN
bjmsr-315	832	10	of	of	ADP
bjmsr-315	832	11	least	least	ADJ
bjmsr-315	832	12	squares	square	NOUN
bjmsr-315	832	13	.	.	PUNCT
bjmsr-315	833	1	annals	annal	NOUN
bjmsr-315	833	2	of	of	ADP
bjmsr-315	833	3	stat	stat	PROPN
bjmsr-315	833	4	.	.	PUNCT
bjmsr-315	833	5	,	,	PUNCT
bjmsr-315	833	6	9	9	NUM
bjmsr-315	833	7	,	,	PUNCT
bjmsr-315	833	8	465	465	NUM
bjmsr-315	833	9	-	-	SYM
bjmsr-315	833	10	474	474	NUM
bjmsr-315	833	11	.	.	PUNCT
bjmsr-315	834	1	s.m	s.m	PROPN
bjmsr-315	834	2	.	.	PROPN
bjmsr-315	834	3	stigler	stigler	NOUN
bjmsr-315	834	4	(	(	PUNCT
bjmsr-315	834	5	1984	1984	NUM
bjmsr-315	834	6	)	)	PUNCT
bjmsr-315	834	7	studies	study	NOUN
bjmsr-315	834	8	in	in	ADP
bjmsr-315	834	9	the	the	DET
bjmsr-315	834	10	history	history	NOUN
bjmsr-315	834	11	of	of	ADP
bjmsr-315	834	12	probability	probability	NOUN
bjmsr-315	834	13	and	and	CCONJ
bjmsr-315	834	14	statistics	statistic	NOUN
bjmsr-315	834	15	xl	xl	PROPN
bjmsr-315	834	16	,	,	PUNCT
bjmsr-315	834	17	boscovich	boscovich	PROPN
bjmsr-315	834	18	,	,	PUNCT
bjmsr-315	834	19	simpson	simpson	PROPN
bjmsr-315	834	20	and	and	CCONJ
bjmsr-315	834	21	a	a	DET
bjmsr-315	834	22	1760	1760	NUM
bjmsr-315	834	23	manuscript	manuscript	NOUN
bjmsr-315	834	24	note	note	NOUN
bjmsr-315	834	25	on	on	ADP
bjmsr-315	834	26	fitting	fit	VERB
bjmsr-315	834	27	a	a	DET
bjmsr-315	834	28	linear	linear	NOUN
bjmsr-315	834	29	relation	relation	NOUN
bjmsr-315	834	30	.	.	PUNCT
bjmsr-315	835	1	biometrica	biometrica	PROPN
bjmsr-315	835	2	,	,	PUNCT
bjmsr-315	835	3	71	71	NUM
bjmsr-315	835	4	,	,	PUNCT
bjmsr-315	835	5	3	3	NUM
bjmsr-315	835	6	,	,	PUNCT
bjmsr-315	835	7	615	615	NUM
bjmsr-315	835	8	-	-	SYM
bjmsr-315	835	9	620	620	NUM
bjmsr-315	835	10	.	.	PUNCT
bjmsr-315	836	1	copyright	copyright	NOUN
bjmsr-315	836	2	©	©	PROPN
bjmsr-315	836	3	cc	cc	PROPN
bjmsr-315	836	4	-	-	PUNCT
bjmsr-315	836	5	by	by	ADP
bjmsr-315	836	6	-	-	PUNCT
bjmsr-315	836	7	nc	nc	PROPN
bjmsr-315	836	8	2019	2019	NUM
bjmsr-315	836	9	,	,	PUNCT
bjmsr-315	836	10	bjmsr	bjmsr	PROPN
bjmsr-315	836	11	www.cribfb.com/journal/index.php/bjmsr	www.cribfb.com/journal/index.php/bjmsr	X
bjmsr-315	837	1	bangladesh	bangladesh	PROPN
bjmsr-315	837	2	journal	journal	PROPN
bjmsr-315	837	3	of	of	ADP
bjmsr-315	837	4	multidisciplinary	multidisciplinary	ADJ
bjmsr-315	837	5	scientific	scientific	ADJ
bjmsr-315	837	6	research	research	NOUN
bjmsr-315	837	7	vol	vol	NOUN
bjmsr-315	837	8	.	.	PROPN
bjmsr-315	837	9	1	1	NUM
bjmsr-315	837	10	,	,	PUNCT
bjmsr-315	837	11	no	no	INTJ
bjmsr-315	837	12	.	.	NOUN
bjmsr-315	837	13	1	1	NUM
bjmsr-315	837	14	;	;	PUNCT
bjmsr-315	837	15	2019	2019	NUM
bjmsr-315	837	16	68	68	NUM
bjmsr-315	837	17	j.	j.	PROPN
bjmsr-315	837	18	svanberg	svanberg	PROPN
bjmsr-315	837	19	(	(	PUNCT
bjmsr-315	837	20	1805	1805	NUM
bjmsr-315	837	21	)	)	PUNCT
bjmsr-315	837	22	exposition	exposition	NOUN
bjmsr-315	837	23	des	des	PROPN
bjmsr-315	837	24	operations	operation	NOUN
bjmsr-315	837	25	faites	faite	NOUN
bjmsr-315	837	26	en	en	ADP
bjmsr-315	837	27	lappnie	lappnie	PROPN
bjmsr-315	837	28	pour	pour	VERB
bjmsr-315	837	29	la	la	PROPN
bjmsr-315	837	30	determination	determination	NOUN
bjmsr-315	837	31	d'un	d'un	PROPN
bjmsr-315	837	32	arc	arc	PROPN
bjmsr-315	837	33	du	du	PROPN
bjmsr-315	837	34	meridien	meridien	PROPN
bjmsr-315	837	35	en	en	PROPN
bjmsr-315	837	36	1801	1801	NUM
bjmsr-315	837	37	,	,	PUNCT
bjmsr-315	837	38	1802	1802	NUM
bjmsr-315	837	39	et	et	NOUN
bjmsr-315	837	40	1803	1803	NUM
bjmsr-315	837	41	,	,	PUNCT
bjmsr-315	837	42	...	...	PUNCT
bjmsr-315	838	1	stockholm	stockholm	PROPN
bjmsr-315	838	2	.	.	PUNCT
bjmsr-315	839	1	j.w	j.w	PROPN
bjmsr-315	839	2	.	.	PROPN
bjmsr-315	839	3	tukey	tukey	PROPN
bjmsr-315	839	4	(	(	PUNCT
bjmsr-315	839	5	1977	1977	NUM
bjmsr-315	839	6	)	)	PUNCT
bjmsr-315	839	7	exploratory	exploratory	ADJ
bjmsr-315	839	8	data	datum	NOUN
bjmsr-315	839	9	analysis	analysis	NOUN
bjmsr-315	839	10	.	.	PUNCT
bjmsr-315	840	1	reading	reading	NOUN
bjmsr-315	840	2	,	,	PUNCT
bjmsr-315	840	3	mass	mass	PROPN
bjmsr-315	840	4	.	.	PROPN
bjmsr-315	840	5	addison	addison	PROPN
bjmsr-315	840	6	-	-	PUNCT
bjmsr-315	840	7	wesley	wesley	PROPN
bjmsr-315	840	8	.	.	PUNCT
bjmsr-315	841	1	h.h	h.h	PROPN
bjmsr-315	841	2	.	.	PROPN
bjmsr-315	841	3	turner	turner	PROPN
bjmsr-315	841	4	(	(	PUNCT
bjmsr-315	841	5	1887	1887	NUM
bjmsr-315	841	6	)	)	PUNCT
bjmsr-315	841	7	on	on	ADP
bjmsr-315	841	8	mr	mr	PROPN
bjmsr-315	841	9	.	.	PROPN
bjmsr-315	841	10	edgeworth	edgeworth	PROPN
bjmsr-315	841	11	's	's	PART
bjmsr-315	841	12	method	method	NOUN
bjmsr-315	841	13	of	of	ADP
bjmsr-315	841	14	reducing	reduce	VERB
bjmsr-315	841	15	observations	observation	NOUN
bjmsr-315	841	16	relating	relate	VERB
bjmsr-315	841	17	to	to	ADP
bjmsr-315	841	18	several	several	ADJ
bjmsr-315	841	19	quantities	quantity	NOUN
bjmsr-315	841	20	.	.	PUNCT
bjmsr-315	842	1	phil	phil	PROPN
bjmsr-315	842	2	.	.	PUNCT
bjmsr-315	843	1	mag	mag	INTJ
bjmsr-315	843	2	.	.	PUNCT
bjmsr-315	844	1	(	(	PUNCT
bjmsr-315	844	2	5	5	NUM
bjmsr-315	844	3	th	th	NUM
bjmsr-315	844	4	series	series	NOUN
bjmsr-315	844	5	)	)	PUNCT
bjmsr-315	844	6	,	,	PUNCT
bjmsr-315	844	7	24	24	NUM
bjmsr-315	844	8	,	,	PUNCT
bjmsr-315	844	9	466	466	NUM
bjmsr-315	844	10	-	-	SYM
bjmsr-315	844	11	470	470	NUM
bjmsr-315	844	12	.	.	PUNCT
bjmsr-315	845	1	k.h	k.h	PROPN
bjmsr-315	845	2	.	.	PROPN
bjmsr-315	845	3	usow	usow	PROPN
bjmsr-315	845	4	(	(	PUNCT
bjmsr-315	845	5	1967a	1967a	NUM
bjmsr-315	845	6	)	)	PUNCT
bjmsr-315	845	7	on	on	ADP
bjmsr-315	845	8	l1	l1	PROPN
bjmsr-315	845	9	approximation	approximation	NOUN
bjmsr-315	845	10	:	:	PUNCT
bjmsr-315	845	11	computation	computation	NOUN
bjmsr-315	845	12	for	for	ADP
bjmsr-315	845	13	continuous	continuous	ADJ
bjmsr-315	845	14	functions	function	NOUN
bjmsr-315	845	15	and	and	CCONJ
bjmsr-315	845	16	continuous	continuous	ADJ
bjmsr-315	845	17	dependence	dependence	NOUN
bjmsr-315	845	18	.	.	PUNCT
bjmsr-315	846	1	siam	siam	PROPN
bjmsr-315	846	2	j.	j.	PROPN
bjmsr-315	846	3	of	of	ADP
bjmsr-315	846	4	numer	numer	PROPN
bjmsr-315	846	5	.	.	PUNCT
bjmsr-315	847	1	anal	anal	PROPN
bjmsr-315	847	2	.	.	PROPN
bjmsr-315	847	3	,	,	PUNCT
bjmsr-315	847	4	4	4	NUM
bjmsr-315	847	5	,	,	PUNCT
bjmsr-315	847	6	70	70	NUM
bjmsr-315	847	7	-	-	SYM
bjmsr-315	847	8	88	88	NUM
bjmsr-315	847	9	.	.	PUNCT
bjmsr-315	848	1	k.h	k.h	PROPN
bjmsr-315	848	2	.	.	PROPN
bjmsr-315	848	3	usow	usow	PROPN
bjmsr-315	848	4	(	(	PUNCT
bjmsr-315	848	5	1967b	1967b	NUM
bjmsr-315	848	6	)	)	PUNCT
bjmsr-315	848	7	on	on	ADP
bjmsr-315	848	8	l1	l1	PROPN
bjmsr-315	848	9	approximation	approximation	NOUN
bjmsr-315	848	10	:	:	PUNCT
bjmsr-315	848	11	computation	computation	NOUN
bjmsr-315	848	12	for	for	ADP
bjmsr-315	848	13	discrete	discrete	ADJ
bjmsr-315	848	14	functions	function	NOUN
bjmsr-315	848	15	and	and	CCONJ
bjmsr-315	848	16	discretization	discretization	NOUN
bjmsr-315	848	17	effect	effect	NOUN
bjmsr-315	848	18	.	.	PUNCT
bjmsr-315	849	1	siam	siam	PROPN
bjmsr-315	849	2	j.	j.	PROPN
bjmsr-315	849	3	numer	numer	PROPN
bjmsr-315	849	4	.	.	PUNCT
bjmsr-315	850	1	anal	anal	PROPN
bjmsr-315	850	2	.	.	PROPN
bjmsr-315	850	3	,	,	PUNCT
bjmsr-315	850	4	4	4	NUM
bjmsr-315	850	5	,	,	PUNCT
bjmsr-315	850	6	233	233	NUM
bjmsr-315	850	7	-	-	SYM
bjmsr-315	850	8	244	244	NUM
bjmsr-315	850	9	.	.	PUNCT
bjmsr-315	851	1	j.f	j.f	PROPN
bjmsr-315	851	2	.	.	PROPN
bjmsr-315	851	3	van	van	PROPN
bjmsr-315	851	4	beeck	beeck	NOUN
bjmsr-315	851	5	-	-	PUNCT
bjmsr-315	851	6	calkoen	calkoen	NOUN
bjmsr-315	851	7	(	(	PUNCT
bjmsr-315	851	8	1816	1816	NUM
bjmsr-315	851	9	)	)	PUNCT
bjmsr-315	851	10	ver	ver	PROPN
bjmsr-315	851	11	de	de	X
bjmsr-315	851	12	theoric	theoric	PROPN
bjmsr-315	851	13	der	der	PROPN
bjmsr-315	851	14	gemiddelde	gemiddelde	PROPN
bjmsr-315	851	15	waardij	waardij	PROPN
bjmsr-315	851	16	.	.	PUNCT
bjmsr-315	852	1	verhandlingen	verhandlingen	PROPN
bjmsr-315	852	2	der	der	PROPN
bjmsr-315	852	3	k.	k.	PROPN
bjmsr-315	852	4	nederlandandsch	nederlandandsch	PROPN
bjmsr-315	852	5	instituut	instituut	PROPN
bjmsr-315	852	6	can	can	AUX
bjmsr-315	852	7	wetenschappen	wetenschappen	VERB
bjmsr-315	852	8	,	,	PUNCT
bjmsr-315	852	9	2	2	NUM
bjmsr-315	852	10	,	,	PUNCT
bjmsr-315	852	11	1	1	NUM
bjmsr-315	852	12	-	-	SYM
bjmsr-315	852	13	19	19	NUM
bjmsr-315	852	14	.	.	PUNCT
bjmsr-315	852	15	b.a	b.a	PROPN
bjmsr-315	852	16	.	.	PROPN
bjmsr-315	852	17	von	von	PROPN
bjmsr-315	852	18	lindenau	lindenau	PROPN
bjmsr-315	852	19	(	(	PUNCT
bjmsr-315	852	20	1806	1806	NUM
bjmsr-315	852	21	)	)	PUNCT
bjmsr-315	852	22	uber	uber	ADJ
bjmsr-315	852	23	den	den	NOUN
bjmsr-315	852	24	gebrauch	gebrauch	ADJ
bjmsr-315	852	25	der	der	NOUN
bjmsr-315	852	26	gradmessungen	gradmessungen	PROPN
bjmsr-315	852	27	zur	zur	PROPN
bjmsr-315	852	28	bestimmung	bestimmung	VERB
bjmsr-315	852	29	der	der	ADJ
bjmsr-315	852	30	gestalt	gestalt	NOUN
bjmsr-315	852	31	der	der	PROPN
bjmsr-315	852	32	erde	erde	PROPN
bjmsr-315	852	33	.	.	PUNCT
bjmsr-315	853	1	monatliche	monatliche	PROPN
bjmsr-315	853	2	correspondenz	correspondenz	PROPN
bjmsr-315	853	3	zur	zur	PROPN
bjmsr-315	853	4	befar	befar	PROPN
bjmsr-315	853	5	derung	derung	VERB
bjmsr-315	853	6	der	der	NOUN
bjmsr-315	853	7	erd	erd	PROPN
bjmsr-315	853	8	-	-	PUNCT
bjmsr-315	853	9	und	und	NOUN
bjmsr-315	853	10	himmels	himmel	NOUN
bjmsr-315	853	11	-	-	PUNCT
bjmsr-315	853	12	kunde	kunde	NOUN
bjmsr-315	853	13	,	,	PUNCT
bjmsr-315	853	14	14	14	NUM
bjmsr-315	853	15	,	,	PUNCT
bjmsr-315	853	16	113	113	NUM
bjmsr-315	853	17	-	-	SYM
bjmsr-315	853	18	158	158	NUM
bjmsr-315	853	19	.	.	PUNCT
bjmsr-315	854	1	h.m	h.m	PROPN
bjmsr-315	854	2	.	.	PROPN
bjmsr-315	854	3	wagner	wagner	PROPN
bjmsr-315	854	4	(	(	PUNCT
bjmsr-315	854	5	1959	1959	NUM
bjmsr-315	854	6	)	)	PUNCT
bjmsr-315	854	7	linear	linear	NOUN
bjmsr-315	854	8	programming	programming	NOUN
bjmsr-315	854	9	technique	technique	NOUN
bjmsr-315	854	10	for	for	ADP
bjmsr-315	854	11	regression	regression	NOUN
bjmsr-315	854	12	analysis	analysis	NOUN
bjmsr-315	854	13	.	.	PUNCT
bjmsr-315	855	1	jasa	jasa	PROPN
bjmsr-315	855	2	,	,	PUNCT
bjmsr-315	855	3	54	54	NUM
bjmsr-315	855	4	,	,	PUNCT
bjmsr-315	855	5	202	202	NUM
bjmsr-315	855	6	-	-	SYM
bjmsr-315	855	7	212	212	NUM
bjmsr-315	855	8	.	.	PUNCT
bjmsr-315	856	1	g.a	g.a	PROPN
bjmsr-315	856	2	.	.	PROPN
bjmsr-315	856	3	watson	watson	PROPN
bjmsr-315	856	4	(	(	PUNCT
bjmsr-315	856	5	1981	1981	NUM
bjmsr-315	856	6	)	)	PUNCT
bjmsr-315	856	7	an	an	DET
bjmsr-315	856	8	algorithm	algorithm	NOUN
bjmsr-315	856	9	for	for	ADP
bjmsr-315	856	10	linear	linear	PROPN
bjmsr-315	856	11	l1	l1	PROPN
bjmsr-315	856	12	approximation	approximation	NOUN
bjmsr-315	856	13	of	of	ADP
bjmsr-315	856	14	continuous	continuous	ADJ
bjmsr-315	856	15	functions	function	NOUN
bjmsr-315	856	16	.	.	PUNCT
bjmsr-315	857	1	i	i	PRON
bjmsr-315	857	2	m	m	VERB
bjmsr-315	858	1	a	a	PROPN
bjmsr-315	858	2	j.	j.	PROPN
bjmsr-315	858	3	num	num	PROPN
bjmsr-315	858	4	.	.	PROPN
bjmsr-315	858	5	anal	anal	PROPN
bjmsr-315	858	6	.	.	PROPN
bjmsr-315	858	7	,	,	PUNCT
bjmsr-315	858	8	1	1	NUM
bjmsr-315	858	9	,	,	PUNCT
bjmsr-315	858	10	157	157	NUM
bjmsr-315	858	11	-	-	SYM
bjmsr-315	858	12	167	167	NUM
bjmsr-315	858	13	.	.	PUNCT
bjmsr-315	859	1	g.o	g.o	NOUN
bjmsr-315	859	2	wesolowsky	wesolowsky	PROPN
bjmsr-315	859	3	(	(	PUNCT
bjmsr-315	859	4	1981	1981	NUM
bjmsr-315	859	5	)	)	PUNCT
bjmsr-315	859	6	a	a	DET
bjmsr-315	859	7	new	new	ADJ
bjmsr-315	859	8	descent	descent	NOUN
bjmsr-315	859	9	algorithm	algorithm	NOUN
bjmsr-315	859	10	for	for	ADP
bjmsr-315	859	11	least	least	ADJ
bjmsr-315	859	12	absolute	absolute	ADJ
bjmsr-315	859	13	value	value	NOUN
bjmsr-315	859	14	regression	regression	NOUN
bjmsr-315	859	15	problem	problem	NOUN
bjmsr-315	859	16	.	.	PUNCT
bjmsr-315	860	1	comm	comm	NOUN
bjmsr-315	860	2	.	.	PUNCT
bjmsr-315	861	1	stat	stat	PROPN
bjmsr-315	861	2	.	.	PUNCT
bjmsr-315	861	3	,	,	PUNCT
bjmsr-315	861	4	b10	b10	PROPN
bjmsr-315	861	5	,	,	PUNCT
bjmsr-315	861	6	479	479	NUM
bjmsr-315	861	7	-	-	SYM
bjmsr-315	861	8	491	491	NUM
bjmsr-315	861	9	.	.	PUNCT
bjmsr-315	862	1	copyrights	copyright	VERB
bjmsr-315	862	2	copyright	copyright	NOUN
bjmsr-315	862	3	for	for	ADP
bjmsr-315	862	4	this	this	DET
bjmsr-315	862	5	article	article	NOUN
bjmsr-315	862	6	is	be	AUX
bjmsr-315	862	7	retained	retain	VERB
bjmsr-315	862	8	by	by	ADP
bjmsr-315	862	9	the	the	DET
bjmsr-315	862	10	author(s	author(s	PROPN
bjmsr-315	862	11	)	)	PUNCT
bjmsr-315	862	12	,	,	PUNCT
bjmsr-315	862	13	with	with	ADP
bjmsr-315	862	14	first	first	ADJ
bjmsr-315	862	15	publication	publication	NOUN
bjmsr-315	862	16	rights	right	NOUN
bjmsr-315	862	17	granted	grant	VERB
bjmsr-315	862	18	to	to	ADP
bjmsr-315	862	19	the	the	DET
bjmsr-315	862	20	journal	journal	NOUN
bjmsr-315	862	21	.	.	PUNCT
bjmsr-315	863	1	this	this	PRON
bjmsr-315	863	2	is	be	AUX
bjmsr-315	863	3	an	an	DET
bjmsr-315	863	4	open	open	ADJ
bjmsr-315	863	5	-	-	PUNCT
bjmsr-315	863	6	access	access	NOUN
bjmsr-315	863	7	article	article	NOUN
bjmsr-315	863	8	distributed	distribute	VERB
bjmsr-315	863	9	under	under	ADP
bjmsr-315	863	10	the	the	DET
bjmsr-315	863	11	terms	term	NOUN
bjmsr-315	863	12	and	and	CCONJ
bjmsr-315	863	13	conditions	condition	NOUN
bjmsr-315	863	14	of	of	ADP
bjmsr-315	863	15	the	the	DET
bjmsr-315	863	16	creative	creative	ADJ
bjmsr-315	863	17	commons	common	NOUN
bjmsr-315	863	18	attribution	attribution	NOUN
bjmsr-315	863	19	license	license	NOUN
bjmsr-315	863	20	(	(	PUNCT
bjmsr-315	863	21	http://creativecommons.org/licenses/by/4.0/	http://creativecommons.org/licenses/by/4.0/	PROPN
bjmsr-315	863	22	)	)	PUNCT
bjmsr-315	863	23	.	.	PUNCT