id	sid	tid	token	lemma	pos
ejpam-1676	1	1	5_wu.dvi	5_wu.dvi	NUM
ejpam-1676	1	2	european	european	ADJ
ejpam-1676	1	3	journal	journal	PROPN
ejpam-1676	1	4	of	of	ADP
ejpam-1676	1	5	pure	pure	ADJ
ejpam-1676	1	6	and	and	CCONJ
ejpam-1676	1	7	applied	apply	VERB
ejpam-1676	1	8	mathematics	mathematic	NOUN
ejpam-1676	1	9	vol	vol	NOUN
ejpam-1676	1	10	.	.	PROPN
ejpam-1676	2	1	5	5	NUM
ejpam-1676	2	2	,	,	PUNCT
ejpam-1676	2	3	no	no	INTJ
ejpam-1676	2	4	.	.	NOUN
ejpam-1676	2	5	3	3	NUM
ejpam-1676	2	6	,	,	PUNCT
ejpam-1676	2	7	2012	2012	NUM
ejpam-1676	2	8	,	,	PUNCT
ejpam-1676	2	9	333	333	NUM
ejpam-1676	2	10	-	-	SYM
ejpam-1676	2	11	356	356	NUM
ejpam-1676	2	12	issn	issn	PROPN
ejpam-1676	2	13	1307	1307	NUM
ejpam-1676	2	14	-	-	SYM
ejpam-1676	2	15	5543	5543	NUM
ejpam-1676	2	16	–	–	PUNCT
ejpam-1676	2	17	www.ejpam.com	www.ejpam.com	X
ejpam-1676	2	18	data	data	NOUN
ejpam-1676	2	19	fusion	fusion	NOUN
ejpam-1676	2	20	using	use	VERB
ejpam-1676	2	21	weighted	weight	VERB
ejpam-1676	2	22	likelihood	likelihood	PROPN
ejpam-1676	2	23	pengfei	pengfei	PROPN
ejpam-1676	2	24	guo	guo	PROPN
ejpam-1676	2	25	,	,	PUNCT
ejpam-1676	2	26	xiaogang	xiaogang	PROPN
ejpam-1676	2	27	wang	wang	PROPN
ejpam-1676	2	28	,	,	PUNCT
ejpam-1676	2	29	yuehua	yuehua	PROPN
ejpam-1676	2	30	wu∗	wu∗	PROPN
ejpam-1676	2	31	department	department	PROPN
ejpam-1676	2	32	of	of	ADP
ejpam-1676	2	33	mathematics	mathematics	PROPN
ejpam-1676	2	34	and	and	CCONJ
ejpam-1676	2	35	statistics	statistic	NOUN
ejpam-1676	2	36	,	,	PUNCT
ejpam-1676	2	37	york	york	PROPN
ejpam-1676	2	38	university	university	PROPN
ejpam-1676	2	39	,	,	PUNCT
ejpam-1676	2	40	toronto	toronto	PROPN
ejpam-1676	2	41	,	,	PUNCT
ejpam-1676	2	42	canada	canada	PROPN
ejpam-1676	2	43	abstract	abstract	PROPN
ejpam-1676	2	44	.	.	PUNCT
ejpam-1676	3	1	this	this	DET
ejpam-1676	3	2	article	article	NOUN
ejpam-1676	3	3	proposes	propose	VERB
ejpam-1676	3	4	to	to	PART
ejpam-1676	3	5	perform	perform	VERB
ejpam-1676	3	6	data	datum	NOUN
ejpam-1676	3	7	fusion	fusion	NOUN
ejpam-1676	3	8	by	by	ADP
ejpam-1676	3	9	using	use	VERB
ejpam-1676	3	10	an	an	DET
ejpam-1676	3	11	adaptive	adaptive	ADJ
ejpam-1676	3	12	weighted	weight	VERB
ejpam-1676	3	13	likelihood	likelihood	NOUN
ejpam-1676	3	14	function	function	NOUN
ejpam-1676	3	15	when	when	SCONJ
ejpam-1676	3	16	data	datum	NOUN
ejpam-1676	3	17	sets	set	NOUN
ejpam-1676	3	18	are	be	AUX
ejpam-1676	3	19	available	available	ADJ
ejpam-1676	3	20	from	from	ADP
ejpam-1676	3	21	related	related	ADJ
ejpam-1676	3	22	populations	population	NOUN
ejpam-1676	3	23	.	.	PUNCT
ejpam-1676	4	1	the	the	DET
ejpam-1676	4	2	main	main	ADJ
ejpam-1676	4	3	objective	objective	NOUN
ejpam-1676	4	4	of	of	ADP
ejpam-1676	4	5	data	datum	NOUN
ejpam-1676	4	6	fusion	fusion	NOUN
ejpam-1676	4	7	is	be	AUX
ejpam-1676	4	8	to	to	PART
ejpam-1676	4	9	integrate	integrate	VERB
ejpam-1676	4	10	information	information	NOUN
ejpam-1676	4	11	from	from	ADP
ejpam-1676	4	12	different	different	ADJ
ejpam-1676	4	13	sources	source	NOUN
ejpam-1676	4	14	to	to	PART
ejpam-1676	4	15	improve	improve	VERB
ejpam-1676	4	16	the	the	DET
ejpam-1676	4	17	quality	quality	NOUN
ejpam-1676	4	18	of	of	ADP
ejpam-1676	4	19	inference	inference	NOUN
ejpam-1676	4	20	when	when	SCONJ
ejpam-1676	4	21	the	the	DET
ejpam-1676	4	22	sample	sample	NOUN
ejpam-1676	4	23	size	size	NOUN
ejpam-1676	4	24	from	from	ADP
ejpam-1676	4	25	the	the	DET
ejpam-1676	4	26	target	target	NOUN
ejpam-1676	4	27	population	population	NOUN
ejpam-1676	4	28	is	be	AUX
ejpam-1676	4	29	small	small	ADJ
ejpam-1676	4	30	or	or	CCONJ
ejpam-1676	4	31	moderate	moderate	ADJ
ejpam-1676	4	32	.	.	PUNCT
ejpam-1676	5	1	the	the	DET
ejpam-1676	5	2	weighted	weight	VERB
ejpam-1676	5	3	likelihood	likelihood	NOUN
ejpam-1676	5	4	function	function	NOUN
ejpam-1676	5	5	is	be	AUX
ejpam-1676	5	6	employed	employ	VERB
ejpam-1676	5	7	simply	simply	ADV
ejpam-1676	5	8	as	as	ADP
ejpam-1676	5	9	an	an	DET
ejpam-1676	5	10	instrument	instrument	NOUN
ejpam-1676	5	11	to	to	PART
ejpam-1676	5	12	facilitate	facilitate	VERB
ejpam-1676	5	13	the	the	DET
ejpam-1676	5	14	data	datum	NOUN
ejpam-1676	5	15	fusion	fusion	NOUN
ejpam-1676	5	16	process	process	NOUN
ejpam-1676	5	17	.	.	PUNCT
ejpam-1676	6	1	the	the	DET
ejpam-1676	6	2	weighted	weight	VERB
ejpam-1676	6	3	likelihood	likelihood	NOUN
ejpam-1676	6	4	method	method	NOUN
ejpam-1676	6	5	has	have	VERB
ejpam-1676	6	6	informationtheoretic	informationtheoretic	ADJ
ejpam-1676	6	7	justification	justification	NOUN
ejpam-1676	6	8	and	and	CCONJ
ejpam-1676	6	9	embraces	embrace	VERB
ejpam-1676	6	10	the	the	DET
ejpam-1676	6	11	widely	widely	ADV
ejpam-1676	6	12	used	use	VERB
ejpam-1676	6	13	classical	classical	ADJ
ejpam-1676	6	14	likelihood	likelihood	NOUN
ejpam-1676	6	15	method	method	NOUN
ejpam-1676	6	16	which	which	PRON
ejpam-1676	6	17	utilizes	utilize	VERB
ejpam-1676	6	18	only	only	ADV
ejpam-1676	6	19	on	on	ADP
ejpam-1676	6	20	the	the	DET
ejpam-1676	6	21	data	datum	NOUN
ejpam-1676	6	22	set	set	VERB
ejpam-1676	6	23	from	from	ADP
ejpam-1676	6	24	the	the	DET
ejpam-1676	6	25	target	target	NOUN
ejpam-1676	6	26	population	population	NOUN
ejpam-1676	6	27	.	.	PUNCT
ejpam-1676	7	1	the	the	DET
ejpam-1676	7	2	degree	degree	NOUN
ejpam-1676	7	3	of	of	ADP
ejpam-1676	7	4	information	information	NOUN
ejpam-1676	7	5	integration	integration	NOUN
ejpam-1676	7	6	in	in	ADP
ejpam-1676	7	7	the	the	DET
ejpam-1676	7	8	proposed	propose	VERB
ejpam-1676	7	9	data	data	NOUN
ejpam-1676	7	10	fusion	fusion	NOUN
ejpam-1676	7	11	process	process	NOUN
ejpam-1676	7	12	is	be	AUX
ejpam-1676	7	13	determined	determine	VERB
ejpam-1676	7	14	by	by	ADP
ejpam-1676	7	15	the	the	DET
ejpam-1676	7	16	likelihood	likelihood	NOUN
ejpam-1676	7	17	weights	weight	NOUN
ejpam-1676	7	18	which	which	PRON
ejpam-1676	7	19	should	should	AUX
ejpam-1676	7	20	be	be	AUX
ejpam-1676	7	21	chosen	choose	VERB
ejpam-1676	7	22	in	in	ADP
ejpam-1676	7	23	a	a	DET
ejpam-1676	7	24	reasonable	reasonable	ADJ
ejpam-1676	7	25	and	and	CCONJ
ejpam-1676	7	26	adaptive	adaptive	ADJ
ejpam-1676	7	27	way	way	NOUN
ejpam-1676	7	28	.	.	PUNCT
ejpam-1676	8	1	the	the	DET
ejpam-1676	8	2	major	major	ADJ
ejpam-1676	8	3	challenge	challenge	NOUN
ejpam-1676	8	4	in	in	ADP
ejpam-1676	8	5	the	the	DET
ejpam-1676	8	6	proposed	propose	VERB
ejpam-1676	8	7	data	data	NOUN
ejpam-1676	8	8	fusion	fusion	NOUN
ejpam-1676	8	9	process	process	NOUN
ejpam-1676	8	10	is	be	AUX
ejpam-1676	8	11	then	then	ADV
ejpam-1676	8	12	to	to	PART
ejpam-1676	8	13	choose	choose	VERB
ejpam-1676	8	14	likelihood	likelihood	NOUN
ejpam-1676	8	15	weights	weight	NOUN
ejpam-1676	8	16	adaptively	adaptively	ADV
ejpam-1676	8	17	and	and	CCONJ
ejpam-1676	8	18	effectively	effectively	ADV
ejpam-1676	8	19	when	when	SCONJ
ejpam-1676	8	20	the	the	DET
ejpam-1676	8	21	deterministic	deterministic	ADJ
ejpam-1676	8	22	relationships	relationship	NOUN
ejpam-1676	8	23	among	among	ADP
ejpam-1676	8	24	all	all	DET
ejpam-1676	8	25	related	related	ADJ
ejpam-1676	8	26	parameters	parameter	NOUN
ejpam-1676	8	27	are	be	AUX
ejpam-1676	8	28	unknown	unknown	ADJ
ejpam-1676	8	29	.	.	PUNCT
ejpam-1676	9	1	we	we	PRON
ejpam-1676	9	2	propose	propose	VERB
ejpam-1676	9	3	adaptive	adaptive	ADJ
ejpam-1676	9	4	likelihood	likelihood	NOUN
ejpam-1676	9	5	weights	weight	NOUN
ejpam-1676	9	6	based	base	VERB
ejpam-1676	9	7	on	on	ADP
ejpam-1676	9	8	the	the	DET
ejpam-1676	9	9	estimated	estimate	VERB
ejpam-1676	9	10	likelihood	likelihood	NOUN
ejpam-1676	9	11	ratio	ratio	NOUN
ejpam-1676	9	12	.	.	PUNCT
ejpam-1676	10	1	we	we	PRON
ejpam-1676	10	2	show	show	VERB
ejpam-1676	10	3	that	that	SCONJ
ejpam-1676	10	4	the	the	DET
ejpam-1676	10	5	data	data	NOUN
ejpam-1676	10	6	fusion	fusion	NOUN
ejpam-1676	10	7	involving	involve	VERB
ejpam-1676	10	8	all	all	DET
ejpam-1676	10	9	relevant	relevant	ADJ
ejpam-1676	10	10	data	data	NOUN
ejpam-1676	10	11	sets	set	NOUN
ejpam-1676	10	12	could	could	AUX
ejpam-1676	10	13	significantly	significantly	ADV
ejpam-1676	10	14	improve	improve	VERB
ejpam-1676	10	15	the	the	DET
ejpam-1676	10	16	mean	mean	ADJ
ejpam-1676	10	17	squared	square	VERB
ejpam-1676	10	18	error	error	NOUN
ejpam-1676	10	19	(	(	PUNCT
ejpam-1676	10	20	mse	mse	NOUN
ejpam-1676	10	21	)	)	PUNCT
ejpam-1676	10	22	of	of	ADP
ejpam-1676	10	23	the	the	DET
ejpam-1676	10	24	classical	classical	ADJ
ejpam-1676	10	25	maximum	maximum	ADJ
ejpam-1676	10	26	likelihood	likelihood	NOUN
ejpam-1676	10	27	estimator	estimator	NOUN
ejpam-1676	10	28	which	which	PRON
ejpam-1676	10	29	only	only	ADV
ejpam-1676	10	30	uses	use	VERB
ejpam-1676	10	31	data	datum	NOUN
ejpam-1676	10	32	set	set	VERB
ejpam-1676	10	33	from	from	ADP
ejpam-1676	10	34	the	the	DET
ejpam-1676	10	35	target	target	NOUN
ejpam-1676	10	36	population	population	NOUN
ejpam-1676	10	37	.	.	PUNCT
ejpam-1676	11	1	it	it	PRON
ejpam-1676	11	2	also	also	ADV
ejpam-1676	11	3	increases	increase	VERB
ejpam-1676	11	4	the	the	DET
ejpam-1676	11	5	power	power	NOUN
ejpam-1676	11	6	for	for	ADP
ejpam-1676	11	7	hypothesis	hypothesis	NOUN
ejpam-1676	11	8	testing	testing	NOUN
ejpam-1676	11	9	.	.	PUNCT
ejpam-1676	12	1	the	the	DET
ejpam-1676	12	2	proposed	propose	VERB
ejpam-1676	12	3	estimator	estimator	NOUN
ejpam-1676	12	4	is	be	AUX
ejpam-1676	12	5	shown	show	VERB
ejpam-1676	12	6	to	to	PART
ejpam-1676	12	7	be	be	AUX
ejpam-1676	12	8	consistent	consistent	ADJ
ejpam-1676	12	9	and	and	CCONJ
ejpam-1676	12	10	asymptotically	asymptotically	ADV
ejpam-1676	12	11	normally	normally	ADV
ejpam-1676	12	12	distributed	distribute	VERB
ejpam-1676	12	13	in	in	ADP
ejpam-1676	12	14	the	the	DET
ejpam-1676	12	15	framework	framework	NOUN
ejpam-1676	12	16	of	of	ADP
ejpam-1676	12	17	generalized	generalized	ADJ
ejpam-1676	12	18	linear	linear	ADJ
ejpam-1676	12	19	models	model	NOUN
ejpam-1676	12	20	.	.	PUNCT
ejpam-1676	13	1	the	the	DET
ejpam-1676	13	2	advantage	advantage	NOUN
ejpam-1676	13	3	of	of	ADP
ejpam-1676	13	4	the	the	DET
ejpam-1676	13	5	proposed	propose	VERB
ejpam-1676	13	6	weighted	weight	VERB
ejpam-1676	13	7	likelihood	likelihood	NOUN
ejpam-1676	13	8	estimator	estimator	NOUN
ejpam-1676	13	9	for	for	ADP
ejpam-1676	13	10	linear	linear	PROPN
ejpam-1676	13	11	models	model	NOUN
ejpam-1676	13	12	is	be	AUX
ejpam-1676	13	13	illustrated	illustrate	VERB
ejpam-1676	13	14	numerically	numerically	ADV
ejpam-1676	13	15	by	by	ADP
ejpam-1676	13	16	a	a	DET
ejpam-1676	13	17	simulation	simulation	NOUN
ejpam-1676	13	18	study	study	NOUN
ejpam-1676	13	19	.	.	PUNCT
ejpam-1676	14	1	a	a	DET
ejpam-1676	14	2	real	real	ADJ
ejpam-1676	14	3	data	data	NOUN
ejpam-1676	14	4	example	example	NOUN
ejpam-1676	14	5	is	be	AUX
ejpam-1676	14	6	also	also	ADV
ejpam-1676	14	7	provided	provide	VERB
ejpam-1676	14	8	.	.	PUNCT
ejpam-1676	15	1	2010	2010	NUM
ejpam-1676	15	2	mathematics	mathematic	NOUN
ejpam-1676	15	3	subject	subject	NOUN
ejpam-1676	15	4	classifications	classification	NOUN
ejpam-1676	15	5	:	:	PUNCT
ejpam-1676	15	6	62f12	62f12	NUM
ejpam-1676	15	7	,	,	PUNCT
ejpam-1676	15	8	62j05	62j05	NUM
ejpam-1676	15	9	,	,	PUNCT
ejpam-1676	15	10	62j12	62j12	NUM
ejpam-1676	15	11	key	key	ADJ
ejpam-1676	15	12	words	word	NOUN
ejpam-1676	15	13	and	and	CCONJ
ejpam-1676	15	14	phrases	phrase	NOUN
ejpam-1676	15	15	:	:	PUNCT
ejpam-1676	15	16	cross	cross	ADJ
ejpam-1676	15	17	-	-	ADJ
ejpam-1676	15	18	validation	validation	ADJ
ejpam-1676	15	19	,	,	PUNCT
ejpam-1676	15	20	nonparametric	nonparametric	NOUN
ejpam-1676	15	21	regression	regression	NOUN
ejpam-1676	15	22	,	,	PUNCT
ejpam-1676	15	23	relative	relative	ADJ
ejpam-1676	15	24	likelihood	likelihood	NOUN
ejpam-1676	15	25	ratio	ratio	NOUN
ejpam-1676	15	26	,	,	PUNCT
ejpam-1676	15	27	semiparametric	semiparametric	NOUN
ejpam-1676	15	28	estimation	estimation	NOUN
ejpam-1676	15	29	,	,	PUNCT
ejpam-1676	15	30	weighted	weight	VERB
ejpam-1676	15	31	likelihood	likelihood	NOUN
ejpam-1676	16	1	1	1	NUM
ejpam-1676	16	2	.	.	X
ejpam-1676	17	1	introduction	introduction	NOUN
ejpam-1676	17	2	in	in	ADP
ejpam-1676	17	3	clinical	clinical	ADJ
ejpam-1676	17	4	trials	trial	NOUN
ejpam-1676	17	5	for	for	ADP
ejpam-1676	17	6	medical	medical	ADJ
ejpam-1676	17	7	research	research	NOUN
ejpam-1676	17	8	,	,	PUNCT
ejpam-1676	17	9	a	a	DET
ejpam-1676	17	10	common	common	ADJ
ejpam-1676	17	11	problem	problem	NOUN
ejpam-1676	17	12	that	that	PRON
ejpam-1676	17	13	often	often	ADV
ejpam-1676	17	14	arises	arise	VERB
ejpam-1676	17	15	is	be	AUX
ejpam-1676	17	16	how	how	SCONJ
ejpam-1676	17	17	to	to	PART
ejpam-1676	17	18	efficiently	efficiently	ADV
ejpam-1676	17	19	estimate	estimate	VERB
ejpam-1676	17	20	the	the	DET
ejpam-1676	17	21	parameter	parameter	NOUN
ejpam-1676	17	22	of	of	ADP
ejpam-1676	17	23	interest	interest	NOUN
ejpam-1676	17	24	when	when	SCONJ
ejpam-1676	17	25	the	the	DET
ejpam-1676	17	26	sample	sample	NOUN
ejpam-1676	17	27	size	size	NOUN
ejpam-1676	17	28	from	from	ADP
ejpam-1676	17	29	the	the	DET
ejpam-1676	17	30	population	population	NOUN
ejpam-1676	17	31	of	of	ADP
ejpam-1676	17	32	inferential	inferential	ADJ
ejpam-1676	17	33	interest	interest	NOUN
ejpam-1676	17	34	is	be	AUX
ejpam-1676	17	35	small	small	ADJ
ejpam-1676	17	36	due	due	ADP
ejpam-1676	17	37	to	to	ADP
ejpam-1676	17	38	the	the	DET
ejpam-1676	17	39	cost	cost	NOUN
ejpam-1676	17	40	or	or	CCONJ
ejpam-1676	17	41	other	other	ADJ
ejpam-1676	17	42	limitations	limitation	NOUN
ejpam-1676	17	43	associated	associate	VERB
ejpam-1676	17	44	with	with	ADP
ejpam-1676	17	45	the	the	DET
ejpam-1676	17	46	experiment	experiment	NOUN
ejpam-1676	17	47	.	.	PUNCT
ejpam-1676	18	1	when	when	SCONJ
ejpam-1676	18	2	the	the	DET
ejpam-1676	18	3	sample	sample	NOUN
ejpam-1676	18	4	size	size	NOUN
ejpam-1676	18	5	from	from	ADP
ejpam-1676	18	6	the	the	DET
ejpam-1676	18	7	population	population	NOUN
ejpam-1676	18	8	of	of	ADP
ejpam-1676	18	9	inferential	inferential	ADJ
ejpam-1676	18	10	interest	interest	NOUN
ejpam-1676	18	11	is	be	AUX
ejpam-1676	18	12	small	small	ADJ
ejpam-1676	18	13	and	and	CCONJ
ejpam-1676	18	14	observations	observation	NOUN
ejpam-1676	18	15	from	from	ADP
ejpam-1676	18	16	related	related	ADJ
ejpam-1676	18	17	studies	study	NOUN
ejpam-1676	18	18	are	be	AUX
ejpam-1676	18	19	available	available	ADJ
ejpam-1676	18	20	,	,	PUNCT
ejpam-1676	18	21	one	one	NUM
ejpam-1676	18	22	important	important	ADJ
ejpam-1676	18	23	question	question	NOUN
ejpam-1676	18	24	is	be	AUX
ejpam-1676	18	25	whether	whether	SCONJ
ejpam-1676	18	26	there	there	PRON
ejpam-1676	18	27	is	be	VERB
ejpam-1676	18	28	any	any	DET
ejpam-1676	18	29	merit	merit	NOUN
ejpam-1676	18	30	to	to	ADP
ejpam-1676	18	31	∗corresponding	∗corresponde	VERB
ejpam-1676	18	32	author	author	NOUN
ejpam-1676	18	33	.	.	PUNCT
ejpam-1676	19	1	email	email	NOUN
ejpam-1676	19	2	addresses	address	NOUN
ejpam-1676	19	3	:	:	PUNCT
ejpam-1676	20	1	pguo�mathstat.yorku	pguo�mathstat.yorku	X
ejpam-1676	20	2	.	.	PUNCT
ejpam-1676	21	1	a	a	DET
ejpam-1676	21	2	(	(	PUNCT
ejpam-1676	21	3	p.	p.	NOUN
ejpam-1676	21	4	guo	guo	PROPN
ejpam-1676	21	5	)	)	PUNCT
ejpam-1676	21	6	,	,	PUNCT
ejpam-1676	21	7	stevenw�mathstat.yorku	stevenw�mathstat.yorku	PROPN
ejpam-1676	21	8	.	.	PROPN
ejpam-1676	22	1	a	a	PRON
ejpam-1676	22	2	(	(	PUNCT
ejpam-1676	22	3	s.	s.	PROPN
ejpam-1676	22	4	wang	wang	PROPN
ejpam-1676	22	5	)	)	PUNCT
ejpam-1676	22	6	,	,	PUNCT
ejpam-1676	22	7	wuyh�mathstat.yorku	wuyh�mathstat.yorku	PROPN
ejpam-1676	22	8	.	.	PUNCT
ejpam-1676	23	1	a	a	DET
ejpam-1676	23	2	(	(	PUNCT
ejpam-1676	23	3	y.	y.	PROPN
ejpam-1676	23	4	wu	wu	PROPN
ejpam-1676	23	5	)	)	PUNCT
ejpam-1676	23	6	http://www.ejpam.com	http://www.ejpam.com	X
ejpam-1676	24	1	333	333	NUM
ejpam-1676	24	2	c	c	X
ejpam-1676	24	3	©	©	PROPN
ejpam-1676	24	4	2012	2012	NUM
ejpam-1676	24	5	ejpam	ejpam	VERB
ejpam-1676	24	6	all	all	DET
ejpam-1676	24	7	rights	right	NOUN
ejpam-1676	24	8	reserved	reserve	VERB
ejpam-1676	24	9	.	.	PUNCT
ejpam-1676	25	1	p.	p.	PROPN
ejpam-1676	25	2	guo	guo	PROPN
ejpam-1676	25	3	,	,	PUNCT
ejpam-1676	25	4	x.	x.	PROPN
ejpam-1676	25	5	wang	wang	PROPN
ejpam-1676	25	6	,	,	PUNCT
ejpam-1676	25	7	y.	y.	PROPN
ejpam-1676	25	8	wu	wu	PROPN
ejpam-1676	25	9	/	/	SYM
ejpam-1676	25	10	eur	eur	PROPN
ejpam-1676	25	11	.	.	PUNCT
ejpam-1676	26	1	j.	j.	PROPN
ejpam-1676	26	2	pure	pure	PROPN
ejpam-1676	26	3	appl	appl	PROPN
ejpam-1676	26	4	.	.	PROPN
ejpam-1676	26	5	math	math	PROPN
ejpam-1676	26	6	,	,	PUNCT
ejpam-1676	26	7	5	5	NUM
ejpam-1676	26	8	(	(	PUNCT
ejpam-1676	26	9	2012	2012	NUM
ejpam-1676	26	10	)	)	PUNCT
ejpam-1676	26	11	,	,	PUNCT
ejpam-1676	26	12	333	333	NUM
ejpam-1676	26	13	-	-	SYM
ejpam-1676	26	14	356	356	NUM
ejpam-1676	26	15	334	334	NUM
ejpam-1676	26	16	combine	combine	VERB
ejpam-1676	26	17	information	information	NOUN
ejpam-1676	26	18	from	from	ADP
ejpam-1676	26	19	populations	population	NOUN
ejpam-1676	26	20	that	that	PRON
ejpam-1676	26	21	are	be	AUX
ejpam-1676	26	22	known	know	VERB
ejpam-1676	26	23	to	to	PART
ejpam-1676	26	24	be	be	AUX
ejpam-1676	26	25	different	different	ADJ
ejpam-1676	26	26	than	than	ADP
ejpam-1676	26	27	the	the	DET
ejpam-1676	26	28	population	population	NOUN
ejpam-1676	26	29	of	of	ADP
ejpam-1676	26	30	inferential	inferential	ADJ
ejpam-1676	26	31	interest	interest	NOUN
ejpam-1676	26	32	.	.	PUNCT
ejpam-1676	27	1	the	the	DET
ejpam-1676	27	2	classical	classical	ADJ
ejpam-1676	27	3	likelihood	likelihood	NOUN
ejpam-1676	27	4	method	method	NOUN
ejpam-1676	27	5	would	would	AUX
ejpam-1676	27	6	concentrate	concentrate	VERB
ejpam-1676	27	7	solely	solely	ADV
ejpam-1676	27	8	on	on	ADP
ejpam-1676	27	9	the	the	DET
ejpam-1676	27	10	sample	sample	NOUN
ejpam-1676	27	11	from	from	ADP
ejpam-1676	27	12	the	the	DET
ejpam-1676	27	13	the	the	DET
ejpam-1676	27	14	target	target	NOUN
ejpam-1676	27	15	population	population	NOUN
ejpam-1676	27	16	without	without	ADP
ejpam-1676	27	17	incorporating	incorporate	VERB
ejpam-1676	27	18	other	other	ADJ
ejpam-1676	27	19	relevant	relevant	ADJ
ejpam-1676	27	20	information	information	NOUN
ejpam-1676	27	21	.	.	PUNCT
ejpam-1676	28	1	the	the	DET
ejpam-1676	28	2	advantage	advantage	NOUN
ejpam-1676	28	3	of	of	ADP
ejpam-1676	28	4	doing	do	VERB
ejpam-1676	28	5	so	so	ADV
ejpam-1676	28	6	is	be	AUX
ejpam-1676	28	7	to	to	PART
ejpam-1676	28	8	avoid	avoid	VERB
ejpam-1676	28	9	possible	possible	ADJ
ejpam-1676	28	10	bias	bias	NOUN
ejpam-1676	28	11	or	or	CCONJ
ejpam-1676	28	12	contamination	contamination	NOUN
ejpam-1676	28	13	of	of	ADP
ejpam-1676	28	14	the	the	DET
ejpam-1676	28	15	data	datum	NOUN
ejpam-1676	28	16	sampled	sample	VERB
ejpam-1676	28	17	directly	directly	ADV
ejpam-1676	28	18	from	from	ADP
ejpam-1676	28	19	the	the	DET
ejpam-1676	28	20	target	target	NOUN
ejpam-1676	28	21	population	population	NOUN
ejpam-1676	28	22	.	.	PUNCT
ejpam-1676	29	1	the	the	DET
ejpam-1676	29	2	maximum	maximum	ADJ
ejpam-1676	29	3	likelihood	likelihood	NOUN
ejpam-1676	29	4	estimator	estimator	NOUN
ejpam-1676	29	5	(	(	PUNCT
ejpam-1676	29	6	mle	mle	PROPN
ejpam-1676	29	7	)	)	PUNCT
ejpam-1676	29	8	is	be	AUX
ejpam-1676	29	9	well	well	ADV
ejpam-1676	29	10	known	known	ADJ
ejpam-1676	29	11	to	to	PART
ejpam-1676	29	12	enjoy	enjoy	VERB
ejpam-1676	29	13	asymptotic	asymptotic	ADJ
ejpam-1676	29	14	properties	property	NOUN
ejpam-1676	29	15	such	such	ADJ
ejpam-1676	29	16	as	as	ADP
ejpam-1676	29	17	consistency	consistency	NOUN
ejpam-1676	29	18	and	and	CCONJ
ejpam-1676	29	19	asymptotic	asymptotic	ADJ
ejpam-1676	29	20	normality	normality	NOUN
ejpam-1676	29	21	.	.	PUNCT
ejpam-1676	30	1	it	it	PRON
ejpam-1676	30	2	is	be	AUX
ejpam-1676	30	3	one	one	NUM
ejpam-1676	30	4	of	of	ADP
ejpam-1676	30	5	the	the	DET
ejpam-1676	30	6	most	most	ADV
ejpam-1676	30	7	widely	widely	ADV
ejpam-1676	30	8	used	use	VERB
ejpam-1676	30	9	method	method	NOUN
ejpam-1676	30	10	in	in	ADP
ejpam-1676	30	11	statistical	statistical	ADJ
ejpam-1676	30	12	inference	inference	NOUN
ejpam-1676	30	13	.	.	PUNCT
ejpam-1676	31	1	however	however	ADV
ejpam-1676	31	2	,	,	PUNCT
ejpam-1676	31	3	the	the	DET
ejpam-1676	31	4	asymptotic	asymptotic	ADJ
ejpam-1676	31	5	properties	property	NOUN
ejpam-1676	31	6	are	be	AUX
ejpam-1676	31	7	of	of	ADP
ejpam-1676	31	8	little	little	ADJ
ejpam-1676	31	9	help	help	NOUN
ejpam-1676	31	10	when	when	SCONJ
ejpam-1676	31	11	the	the	DET
ejpam-1676	31	12	sample	sample	NOUN
ejpam-1676	31	13	size	size	NOUN
ejpam-1676	31	14	is	be	AUX
ejpam-1676	31	15	small	small	ADJ
ejpam-1676	31	16	or	or	CCONJ
ejpam-1676	31	17	very	very	ADV
ejpam-1676	31	18	moderate	moderate	ADJ
ejpam-1676	31	19	.	.	PUNCT
ejpam-1676	32	1	the	the	DET
ejpam-1676	32	2	mle	mle	PROPN
ejpam-1676	32	3	could	could	AUX
ejpam-1676	32	4	provide	provide	VERB
ejpam-1676	32	5	quite	quite	ADV
ejpam-1676	32	6	misleading	misleading	ADJ
ejpam-1676	32	7	results	result	NOUN
ejpam-1676	32	8	due	due	ADP
ejpam-1676	32	9	to	to	ADP
ejpam-1676	32	10	a	a	DET
ejpam-1676	32	11	insufficient	insufficient	ADJ
ejpam-1676	32	12	sample	sample	NOUN
ejpam-1676	32	13	size	size	NOUN
ejpam-1676	32	14	.	.	PUNCT
ejpam-1676	33	1	by	by	ADP
ejpam-1676	33	2	contrast	contrast	NOUN
ejpam-1676	33	3	,	,	PUNCT
ejpam-1676	33	4	the	the	DET
ejpam-1676	33	5	bayesian	bayesian	NOUN
ejpam-1676	33	6	method	method	NOUN
ejpam-1676	33	7	can	can	AUX
ejpam-1676	33	8	effectively	effectively	ADV
ejpam-1676	33	9	combine	combine	VERB
ejpam-1676	33	10	information	information	NOUN
ejpam-1676	33	11	when	when	SCONJ
ejpam-1676	33	12	the	the	DET
ejpam-1676	33	13	parameters	parameter	NOUN
ejpam-1676	33	14	from	from	ADP
ejpam-1676	33	15	these	these	DET
ejpam-1676	33	16	populations	population	NOUN
ejpam-1676	33	17	are	be	AUX
ejpam-1676	33	18	assumed	assume	VERB
ejpam-1676	33	19	to	to	PART
ejpam-1676	33	20	be	be	AUX
ejpam-1676	33	21	random	random	ADJ
ejpam-1676	33	22	variables	variable	NOUN
ejpam-1676	33	23	from	from	ADP
ejpam-1676	33	24	a	a	DET
ejpam-1676	33	25	hyper	hyper	NOUN
ejpam-1676	33	26	-	-	NOUN
ejpam-1676	33	27	distribution	distribution	NOUN
ejpam-1676	33	28	.	.	PUNCT
ejpam-1676	34	1	it	it	PRON
ejpam-1676	34	2	has	have	VERB
ejpam-1676	34	3	many	many	ADJ
ejpam-1676	34	4	advantages	advantage	NOUN
ejpam-1676	34	5	since	since	SCONJ
ejpam-1676	34	6	prior	prior	ADJ
ejpam-1676	34	7	information	information	NOUN
ejpam-1676	34	8	can	can	AUX
ejpam-1676	34	9	be	be	AUX
ejpam-1676	34	10	formally	formally	ADV
ejpam-1676	34	11	incorporated	incorporate	VERB
ejpam-1676	34	12	and	and	CCONJ
ejpam-1676	34	13	the	the	DET
ejpam-1676	34	14	inferences	inference	NOUN
ejpam-1676	34	15	are	be	AUX
ejpam-1676	34	16	conditional	conditional	ADJ
ejpam-1676	34	17	on	on	ADP
ejpam-1676	34	18	all	all	DET
ejpam-1676	34	19	the	the	DET
ejpam-1676	34	20	data	datum	NOUN
ejpam-1676	34	21	.	.	PUNCT
ejpam-1676	35	1	bayesian	bayesian	NOUN
ejpam-1676	35	2	inference	inference	NOUN
ejpam-1676	35	3	might	might	AUX
ejpam-1676	35	4	be	be	AUX
ejpam-1676	35	5	computationally	computationally	ADV
ejpam-1676	35	6	intensive	intensive	ADJ
ejpam-1676	35	7	for	for	ADP
ejpam-1676	35	8	nontrivial	nontrivial	ADJ
ejpam-1676	35	9	cases	case	NOUN
ejpam-1676	35	10	.	.	PUNCT
ejpam-1676	36	1	it	it	PRON
ejpam-1676	36	2	is	be	AUX
ejpam-1676	36	3	well	well	ADV
ejpam-1676	36	4	known	know	VERB
ejpam-1676	36	5	,	,	PUNCT
ejpam-1676	36	6	however	however	ADV
ejpam-1676	36	7	,	,	PUNCT
ejpam-1676	36	8	the	the	DET
ejpam-1676	36	9	powerful	powerful	ADJ
ejpam-1676	36	10	monte	monte	PROPN
ejpam-1676	36	11	carlo	carlo	PROPN
ejpam-1676	36	12	markov	markov	PROPN
ejpam-1676	36	13	chain	chain	NOUN
ejpam-1676	36	14	method	method	NOUN
ejpam-1676	36	15	could	could	AUX
ejpam-1676	36	16	be	be	AUX
ejpam-1676	36	17	computationally	computationally	ADV
ejpam-1676	36	18	intensive	intensive	ADJ
ejpam-1676	36	19	and	and	CCONJ
ejpam-1676	36	20	challenging	challenging	ADJ
ejpam-1676	36	21	for	for	ADP
ejpam-1676	36	22	the	the	DET
ejpam-1676	36	23	high	high	ADJ
ejpam-1676	36	24	dimensional	dimensional	ADJ
ejpam-1676	36	25	case	case	NOUN
ejpam-1676	36	26	when	when	SCONJ
ejpam-1676	36	27	a	a	DET
ejpam-1676	36	28	conjugate	conjugate	NOUN
ejpam-1676	36	29	prior	prior	ADV
ejpam-1676	36	30	can	can	AUX
ejpam-1676	36	31	not	not	PART
ejpam-1676	36	32	be	be	AUX
ejpam-1676	36	33	assumed	assume	VERB
ejpam-1676	36	34	.	.	PUNCT
ejpam-1676	37	1	for	for	ADP
ejpam-1676	37	2	exploratory	exploratory	ADJ
ejpam-1676	37	3	purposes	purpose	NOUN
ejpam-1676	37	4	,	,	PUNCT
ejpam-1676	37	5	we	we	PRON
ejpam-1676	37	6	propose	propose	VERB
ejpam-1676	37	7	to	to	PART
ejpam-1676	37	8	integrate	integrate	VERB
ejpam-1676	37	9	all	all	DET
ejpam-1676	37	10	available	available	ADJ
ejpam-1676	37	11	information	information	NOUN
ejpam-1676	37	12	through	through	ADP
ejpam-1676	37	13	a	a	DET
ejpam-1676	37	14	data	data	NOUN
ejpam-1676	37	15	fusion	fusion	NOUN
ejpam-1676	37	16	process	process	NOUN
ejpam-1676	37	17	based	base	VERB
ejpam-1676	37	18	on	on	ADP
ejpam-1676	37	19	an	an	DET
ejpam-1676	37	20	adaptive	adaptive	ADJ
ejpam-1676	37	21	weighted	weight	VERB
ejpam-1676	37	22	likelihood	likelihood	NOUN
ejpam-1676	37	23	function	function	NOUN
ejpam-1676	37	24	.	.	PUNCT
ejpam-1676	38	1	the	the	DET
ejpam-1676	38	2	weighted	weight	VERB
ejpam-1676	38	3	likelihood	likelihood	NOUN
ejpam-1676	38	4	function	function	NOUN
ejpam-1676	38	5	that	that	SCONJ
ejpam-1676	38	6	we	we	PRON
ejpam-1676	38	7	employ	employ	VERB
ejpam-1676	38	8	can	can	AUX
ejpam-1676	38	9	be	be	AUX
ejpam-1676	38	10	very	very	ADV
ejpam-1676	38	11	adaptive	adaptive	ADJ
ejpam-1676	38	12	and	and	CCONJ
ejpam-1676	38	13	easy	easy	ADJ
ejpam-1676	38	14	to	to	PART
ejpam-1676	38	15	implement	implement	VERB
ejpam-1676	38	16	.	.	PUNCT
ejpam-1676	39	1	since	since	SCONJ
ejpam-1676	39	2	there	there	PRON
ejpam-1676	39	3	are	be	VERB
ejpam-1676	39	4	many	many	ADJ
ejpam-1676	39	5	weighted	weight	VERB
ejpam-1676	39	6	likelihoods	likelihood	NOUN
ejpam-1676	39	7	proposed	propose	VERB
ejpam-1676	39	8	for	for	ADP
ejpam-1676	39	9	various	various	ADJ
ejpam-1676	39	10	purposes	purpose	NOUN
ejpam-1676	39	11	,	,	PUNCT
ejpam-1676	39	12	it	it	PRON
ejpam-1676	39	13	is	be	AUX
ejpam-1676	39	14	necessary	necessary	ADJ
ejpam-1676	39	15	to	to	PART
ejpam-1676	39	16	provide	provide	VERB
ejpam-1676	39	17	an	an	DET
ejpam-1676	39	18	brief	brief	ADJ
ejpam-1676	39	19	overview	overview	NOUN
ejpam-1676	39	20	and	and	CCONJ
ejpam-1676	39	21	avoid	avoid	VERB
ejpam-1676	39	22	any	any	DET
ejpam-1676	39	23	possible	possible	ADJ
ejpam-1676	39	24	confusion	confusion	NOUN
ejpam-1676	39	25	later	later	ADV
ejpam-1676	39	26	on	on	ADV
ejpam-1676	39	27	.	.	PUNCT
ejpam-1676	40	1	[	[	X
ejpam-1676	40	2	4	4	X
ejpam-1676	40	3	]	]	PUNCT
ejpam-1676	40	4	introduced	introduce	VERB
ejpam-1676	40	5	the	the	DET
ejpam-1676	40	6	idea	idea	NOUN
ejpam-1676	40	7	of	of	ADP
ejpam-1676	40	8	local	local	ADJ
ejpam-1676	40	9	likelihood	likelihood	NOUN
ejpam-1676	40	10	to	to	PART
ejpam-1676	40	11	derive	derive	VERB
ejpam-1676	40	12	local	local	ADJ
ejpam-1676	40	13	inference	inference	NOUN
ejpam-1676	40	14	.	.	PUNCT
ejpam-1676	41	1	[	[	X
ejpam-1676	41	2	2	2	X
ejpam-1676	41	3	]	]	PUNCT
ejpam-1676	41	4	defined	define	VERB
ejpam-1676	41	5	her	her	PRON
ejpam-1676	41	6	version	version	NOUN
ejpam-1676	41	7	of	of	ADP
ejpam-1676	41	8	local	local	ADJ
ejpam-1676	41	9	likelihood	likelihood	NOUN
ejpam-1676	41	10	in	in	ADP
ejpam-1676	41	11	the	the	DET
ejpam-1676	41	12	context	context	NOUN
ejpam-1676	41	13	of	of	ADP
ejpam-1676	41	14	non	non	ADJ
ejpam-1676	41	15	-	-	ADJ
ejpam-1676	41	16	parametric	parametric	ADJ
ejpam-1676	41	17	regression	regression	NOUN
ejpam-1676	41	18	.	.	PUNCT
ejpam-1676	42	1	[	[	X
ejpam-1676	42	2	17	17	NUM
ejpam-1676	42	3	]	]	PUNCT
ejpam-1676	42	4	presented	present	VERB
ejpam-1676	42	5	the	the	DET
ejpam-1676	42	6	general	general	ADJ
ejpam-1676	42	7	form	form	NOUN
ejpam-1676	42	8	of	of	ADP
ejpam-1676	42	9	the	the	DET
ejpam-1676	42	10	local	local	ADJ
ejpam-1676	42	11	likelihood	likelihood	NOUN
ejpam-1676	42	12	.	.	PUNCT
ejpam-1676	43	1	[	[	X
ejpam-1676	43	2	9	9	NUM
ejpam-1676	43	3	]	]	PUNCT
ejpam-1676	43	4	and	and	CCONJ
ejpam-1676	43	5	[	[	X
ejpam-1676	43	6	10	10	NUM
ejpam-1676	43	7	]	]	PUNCT
ejpam-1676	43	8	proposed	propose	VERB
ejpam-1676	43	9	their	their	PRON
ejpam-1676	43	10	relevant	relevant	ADJ
ejpam-1676	43	11	weighted	weight	VERB
ejpam-1676	43	12	likelihood	likelihood	NOUN
ejpam-1676	43	13	function	function	NOUN
ejpam-1676	43	14	to	to	PART
ejpam-1676	43	15	combine	combine	VERB
ejpam-1676	43	16	information	information	NOUN
ejpam-1676	43	17	in	in	ADP
ejpam-1676	43	18	which	which	PRON
ejpam-1676	43	19	the	the	DET
ejpam-1676	43	20	weights	weight	NOUN
ejpam-1676	43	21	are	be	AUX
ejpam-1676	43	22	very	very	ADV
ejpam-1676	43	23	general	general	ADJ
ejpam-1676	43	24	.	.	PUNCT
ejpam-1676	44	1	[	[	X
ejpam-1676	44	2	12	12	NUM
ejpam-1676	44	3	]	]	PUNCT
ejpam-1676	44	4	proposed	propose	VERB
ejpam-1676	44	5	their	their	PRON
ejpam-1676	44	6	adaptive	adaptive	ADJ
ejpam-1676	44	7	weights	weight	NOUN
ejpam-1676	44	8	when	when	SCONJ
ejpam-1676	44	9	the	the	DET
ejpam-1676	44	10	time	time	NOUN
ejpam-1676	44	11	trend	trend	NOUN
ejpam-1676	44	12	is	be	AUX
ejpam-1676	44	13	present	present	ADJ
ejpam-1676	44	14	.	.	PUNCT
ejpam-1676	45	1	[	[	X
ejpam-1676	45	2	11	11	NUM
ejpam-1676	45	3	]	]	PUNCT
ejpam-1676	45	4	reviewed	review	VERB
ejpam-1676	45	5	most	most	ADV
ejpam-1676	45	6	related	relate	VERB
ejpam-1676	45	7	weighted	weighted	ADJ
ejpam-1676	45	8	likelihood	likelihood	NOUN
ejpam-1676	45	9	methods	method	NOUN
ejpam-1676	45	10	proposed	propose	VERB
ejpam-1676	45	11	in	in	ADP
ejpam-1676	45	12	the	the	DET
ejpam-1676	45	13	literature	literature	NOUN
ejpam-1676	45	14	.	.	PUNCT
ejpam-1676	46	1	all	all	DET
ejpam-1676	46	2	the	the	DET
ejpam-1676	46	3	aforementioned	aforementioned	ADJ
ejpam-1676	46	4	weighted	weight	VERB
ejpam-1676	46	5	likelihood	likelihood	NOUN
ejpam-1676	46	6	methods	method	NOUN
ejpam-1676	46	7	focus	focus	VERB
ejpam-1676	46	8	on	on	ADP
ejpam-1676	46	9	a	a	DET
ejpam-1676	46	10	setting	setting	NOUN
ejpam-1676	46	11	in	in	ADP
ejpam-1676	46	12	which	which	PRON
ejpam-1676	46	13	the	the	DET
ejpam-1676	46	14	number	number	NOUN
ejpam-1676	46	15	of	of	ADP
ejpam-1676	46	16	populations	population	NOUN
ejpam-1676	46	17	goes	go	VERB
ejpam-1676	46	18	to	to	ADP
ejpam-1676	46	19	infinity	infinity	NOUN
ejpam-1676	46	20	.	.	PUNCT
ejpam-1676	47	1	in	in	ADP
ejpam-1676	47	2	this	this	DET
ejpam-1676	47	3	article	article	NOUN
ejpam-1676	47	4	,	,	PUNCT
ejpam-1676	47	5	however	however	ADV
ejpam-1676	47	6	,	,	PUNCT
ejpam-1676	47	7	we	we	PRON
ejpam-1676	47	8	focus	focus	VERB
ejpam-1676	47	9	on	on	ADP
ejpam-1676	47	10	situations	situation	NOUN
ejpam-1676	47	11	in	in	ADP
ejpam-1676	47	12	which	which	PRON
ejpam-1676	47	13	the	the	DET
ejpam-1676	47	14	number	number	NOUN
ejpam-1676	47	15	of	of	ADP
ejpam-1676	47	16	populations	population	NOUN
ejpam-1676	47	17	is	be	AUX
ejpam-1676	47	18	actually	actually	ADV
ejpam-1676	47	19	fixed	fix	VERB
ejpam-1676	47	20	.	.	PUNCT
ejpam-1676	48	1	for	for	ADP
ejpam-1676	48	2	example	example	NOUN
ejpam-1676	48	3	,	,	PUNCT
ejpam-1676	48	4	one	one	PRON
ejpam-1676	48	5	might	might	AUX
ejpam-1676	48	6	be	be	AUX
ejpam-1676	48	7	interested	interested	ADJ
ejpam-1676	48	8	in	in	ADP
ejpam-1676	48	9	the	the	DET
ejpam-1676	48	10	efficacy	efficacy	NOUN
ejpam-1676	48	11	of	of	ADP
ejpam-1676	48	12	one	one	NUM
ejpam-1676	48	13	particular	particular	ADJ
ejpam-1676	48	14	treatment	treatment	NOUN
ejpam-1676	48	15	when	when	SCONJ
ejpam-1676	48	16	the	the	DET
ejpam-1676	48	17	results	result	NOUN
ejpam-1676	48	18	of	of	ADP
ejpam-1676	48	19	the	the	DET
ejpam-1676	48	20	current	current	ADJ
ejpam-1676	48	21	and	and	CCONJ
ejpam-1676	48	22	several	several	ADJ
ejpam-1676	48	23	historical	historical	ADJ
ejpam-1676	48	24	clinical	clinical	ADJ
ejpam-1676	48	25	trials	trial	NOUN
ejpam-1676	48	26	are	be	AUX
ejpam-1676	48	27	available	available	ADJ
ejpam-1676	48	28	.	.	PUNCT
ejpam-1676	49	1	extending	extend	VERB
ejpam-1676	49	2	the	the	DET
ejpam-1676	49	3	relevant	relevant	ADJ
ejpam-1676	49	4	weighted	weight	VERB
ejpam-1676	49	5	likelihood	likelihood	NOUN
ejpam-1676	49	6	by	by	ADP
ejpam-1676	49	7	[	[	X
ejpam-1676	49	8	9	9	NUM
ejpam-1676	49	9	]	]	PUNCT
ejpam-1676	49	10	,	,	PUNCT
ejpam-1676	49	11	[	[	X
ejpam-1676	49	12	21	21	NUM
ejpam-1676	49	13	]	]	PUNCT
ejpam-1676	49	14	proposed	propose	VERB
ejpam-1676	49	15	their	their	PRON
ejpam-1676	49	16	version	version	NOUN
ejpam-1676	49	17	of	of	ADP
ejpam-1676	49	18	the	the	DET
ejpam-1676	49	19	wl	wl	X
ejpam-1676	49	20	when	when	SCONJ
ejpam-1676	49	21	the	the	DET
ejpam-1676	49	22	number	number	NOUN
ejpam-1676	49	23	of	of	ADP
ejpam-1676	49	24	populations	population	NOUN
ejpam-1676	49	25	is	be	AUX
ejpam-1676	49	26	fixed	fix	VERB
ejpam-1676	49	27	.	.	PUNCT
ejpam-1676	50	1	we	we	PRON
ejpam-1676	50	2	now	now	ADV
ejpam-1676	50	3	briefly	briefly	VERB
ejpam-1676	50	4	their	their	PRON
ejpam-1676	50	5	weighted	weight	VERB
ejpam-1676	50	6	likelihood	likelihood	NOUN
ejpam-1676	50	7	function	function	NOUN
ejpam-1676	50	8	as	as	SCONJ
ejpam-1676	50	9	follows	follow	VERB
ejpam-1676	50	10	.	.	PUNCT
ejpam-1676	51	1	suppose	suppose	VERB
ejpam-1676	51	2	that	that	SCONJ
ejpam-1676	51	3	we	we	PRON
ejpam-1676	51	4	observe	observe	VERB
ejpam-1676	51	5	independent	independent	ADJ
ejpam-1676	51	6	random	random	ADJ
ejpam-1676	51	7	response	response	NOUN
ejpam-1676	51	8	vectors	vector	NOUN
ejpam-1676	51	9	y	y	PROPN
ejpam-1676	51	10	1	1	NUM
ejpam-1676	51	11	,	,	PUNCT
ejpam-1676	51	12	.	.	PUNCT
ejpam-1676	51	13	.	.	PUNCT
ejpam-1676	52	1	.	.	PUNCT
ejpam-1676	53	1	,	,	PUNCT
ejpam-1676	54	1	y	y	PROPN
ejpam-1676	54	2	m	m	VERB
ejpam-1676	54	3	with	with	ADP
ejpam-1676	54	4	probability	probability	NOUN
ejpam-1676	54	5	density	density	NOUN
ejpam-1676	54	6	functions	function	NOUN
ejpam-1676	54	7	f	f	X
ejpam-1676	54	8	(	(	PUNCT
ejpam-1676	54	9	.;θ1	.;θ1	PROPN
ejpam-1676	54	10	)	)	PUNCT
ejpam-1676	54	11	,	,	PUNCT
ejpam-1676	54	12	.	.	PUNCT
ejpam-1676	54	13	.	.	PUNCT
ejpam-1676	54	14	.	.	PUNCT
ejpam-1676	55	1	,	,	PUNCT
ejpam-1676	55	2	f	f	PROPN
ejpam-1676	55	3	(	(	PUNCT
ejpam-1676	55	4	.;θm	.;θm	NOUN
ejpam-1676	55	5	)	)	PUNCT
ejpam-1676	55	6	,	,	PUNCT
ejpam-1676	55	7	where	where	SCONJ
ejpam-1676	55	8	y	y	PROPN
ejpam-1676	55	9	i	i	NOUN
ejpam-1676	55	10	=	=	PUNCT
ejpam-1676	55	11	(	(	PUNCT
ejpam-1676	55	12	yi1	yi1	NOUN
ejpam-1676	55	13	,	,	PUNCT
ejpam-1676	55	14	.	.	PUNCT
ejpam-1676	55	15	.	.	PUNCT
ejpam-1676	55	16	.	.	PUNCT
ejpam-1676	56	1	,	,	PUNCT
ejpam-1676	56	2	yini	yini	PROPN
ejpam-1676	56	3	)	)	PUNCT
ejpam-1676	56	4	t	t	PROPN
ejpam-1676	56	5	,	,	PUNCT
ejpam-1676	56	6	i	i	NOUN
ejpam-1676	56	7	=	=	SYM
ejpam-1676	56	8	1,2	1,2	NUM
ejpam-1676	56	9	,	,	PUNCT
ejpam-1676	56	10	·	·	PUNCT
ejpam-1676	56	11	·	·	PUNCT
ejpam-1676	56	12	·	·	PUNCT
ejpam-1676	56	13	,	,	PUNCT
ejpam-1676	56	14	m.	m.	NOUN
ejpam-1676	56	15	further	far	ADV
ejpam-1676	56	16	suppose	suppose	VERB
ejpam-1676	56	17	that	that	SCONJ
ejpam-1676	56	18	only	only	ADV
ejpam-1676	56	19	population	population	NOUN
ejpam-1676	56	20	1	1	NUM
ejpam-1676	56	21	,	,	PUNCT
ejpam-1676	56	22	in	in	ADP
ejpam-1676	56	23	particular	particular	ADJ
ejpam-1676	56	24	θ1	θ1	NOUN
ejpam-1676	56	25	,	,	PUNCT
ejpam-1676	56	26	an	an	DET
ejpam-1676	56	27	unknown	unknown	ADJ
ejpam-1676	56	28	vector	vector	NOUN
ejpam-1676	56	29	of	of	ADP
ejpam-1676	56	30	parameters	parameter	NOUN
ejpam-1676	56	31	,	,	PUNCT
ejpam-1676	56	32	is	be	AUX
ejpam-1676	56	33	of	of	ADP
ejpam-1676	56	34	inferential	inferential	ADJ
ejpam-1676	56	35	interest	interest	NOUN
ejpam-1676	56	36	.	.	PUNCT
ejpam-1676	57	1	given	give	VERB
ejpam-1676	57	2	data	data	PROPN
ejpam-1676	57	3	,	,	PUNCT
ejpam-1676	57	4	y	y	PROPN
ejpam-1676	57	5	1	1	NUM
ejpam-1676	57	6	=	=	SYM
ejpam-1676	57	7	y1	y1	PROPN
ejpam-1676	57	8	,	,	PUNCT
ejpam-1676	57	9	the	the	DET
ejpam-1676	57	10	classical	classical	ADJ
ejpam-1676	57	11	likelihood	likelihood	NOUN
ejpam-1676	57	12	would	would	AUX
ejpam-1676	57	13	be	be	AUX
ejpam-1676	57	14	l1(y1,θ1	l1(y1,θ1	PROPN
ejpam-1676	57	15	)	)	PUNCT
ejpam-1676	58	1	=	=	SYM
ejpam-1676	58	2	n1∏	n1∏	NOUN
ejpam-1676	58	3	j=1	j=1	PROPN
ejpam-1676	58	4	f	f	PROPN
ejpam-1676	58	5	(	(	PUNCT
ejpam-1676	58	6	y1	y1	INTJ
ejpam-1676	58	7	j;θ1	j;θ1	PROPN
ejpam-1676	58	8	)	)	PUNCT
ejpam-1676	58	9	.	.	PUNCT
ejpam-1676	59	1	when	when	SCONJ
ejpam-1676	59	2	the	the	DET
ejpam-1676	59	3	parameters	parameter	NOUN
ejpam-1676	59	4	θ2	θ2	PROPN
ejpam-1676	59	5	,	,	PUNCT
ejpam-1676	59	6	.	.	PUNCT
ejpam-1676	59	7	.	.	PUNCT
ejpam-1676	60	1	.	.	PUNCT
ejpam-1676	61	1	,	,	PUNCT
ejpam-1676	61	2	θm	θm	PROPN
ejpam-1676	61	3	are	be	AUX
ejpam-1676	61	4	thought	think	VERB
ejpam-1676	61	5	to	to	PART
ejpam-1676	61	6	interconnected	interconnect	VERB
ejpam-1676	61	7	with	with	ADP
ejpam-1676	61	8	θ1	θ1	NOUN
ejpam-1676	61	9	,	,	PUNCT
ejpam-1676	61	10	for	for	ADP
ejpam-1676	61	11	given	give	VERB
ejpam-1676	61	12	p.	p.	PROPN
ejpam-1676	61	13	guo	guo	PROPN
ejpam-1676	61	14	,	,	PUNCT
ejpam-1676	61	15	x.	x.	PROPN
ejpam-1676	61	16	wang	wang	PROPN
ejpam-1676	61	17	,	,	PUNCT
ejpam-1676	61	18	y.	y.	PROPN
ejpam-1676	61	19	wu	wu	PROPN
ejpam-1676	61	20	/	/	SYM
ejpam-1676	61	21	eur	eur	PROPN
ejpam-1676	61	22	.	.	PUNCT
ejpam-1676	62	1	j.	j.	PROPN
ejpam-1676	62	2	pure	pure	PROPN
ejpam-1676	62	3	appl	appl	PROPN
ejpam-1676	62	4	.	.	PROPN
ejpam-1676	62	5	math	math	PROPN
ejpam-1676	62	6	,	,	PUNCT
ejpam-1676	62	7	5	5	NUM
ejpam-1676	62	8	(	(	PUNCT
ejpam-1676	62	9	2012	2012	NUM
ejpam-1676	62	10	)	)	PUNCT
ejpam-1676	62	11	,	,	PUNCT
ejpam-1676	62	12	333	333	NUM
ejpam-1676	62	13	-	-	SYM
ejpam-1676	62	14	356	356	NUM
ejpam-1676	62	15	335	335	NUM
ejpam-1676	62	16	y	y	NOUN
ejpam-1676	62	17	=	=	SYM
ejpam-1676	62	18	(	(	PUNCT
ejpam-1676	62	19	y1	y1	PROPN
ejpam-1676	62	20	,	,	PUNCT
ejpam-1676	62	21	y2	y2	PROPN
ejpam-1676	62	22	,	,	PUNCT
ejpam-1676	62	23	·	·	PUNCT
ejpam-1676	62	24	·	·	PUNCT
ejpam-1676	62	25	·	·	PUNCT
ejpam-1676	62	26	,	,	PUNCT
ejpam-1676	62	27	ym	ym	PROPN
ejpam-1676	62	28	)	)	PUNCT
ejpam-1676	62	29	,	,	PUNCT
ejpam-1676	62	30	the	the	DET
ejpam-1676	62	31	weighted	weight	VERB
ejpam-1676	62	32	likelihood	likelihood	NOUN
ejpam-1676	62	33	(	(	PUNCT
ejpam-1676	62	34	wl	wl	NOUN
ejpam-1676	62	35	)	)	PUNCT
ejpam-1676	62	36	for	for	ADP
ejpam-1676	62	37	θ1	θ1	NOUN
ejpam-1676	62	38	is	be	AUX
ejpam-1676	62	39	defined	define	VERB
ejpam-1676	62	40	as	as	ADP
ejpam-1676	62	41	wl(y;θ1	wl(y;θ1	PROPN
ejpam-1676	62	42	)	)	PUNCT
ejpam-1676	63	1	=	=	SYM
ejpam-1676	63	2	m∏	m∏	PROPN
ejpam-1676	63	3	i=1	i=1	PUNCT
ejpam-1676	64	1	ni∏	ni∏	NOUN
ejpam-1676	65	1	j=1	j=1	ADJ
ejpam-1676	65	2	f	f	PROPN
ejpam-1676	65	3	(	(	PUNCT
ejpam-1676	65	4	yi	yi	PROPN
ejpam-1676	65	5	j;θ1	j;θ1	PROPN
ejpam-1676	65	6	)	)	PUNCT
ejpam-1676	65	7	λi	λi	INTJ
ejpam-1676	65	8	,	,	PUNCT
ejpam-1676	65	9	where	where	SCONJ
ejpam-1676	65	10	λ	λ	X
ejpam-1676	65	11	=	=	SYM
ejpam-1676	65	12	(	(	PUNCT
ejpam-1676	65	13	λ1	λ1	ADJ
ejpam-1676	65	14	,	,	PUNCT
ejpam-1676	65	15	.	.	PUNCT
ejpam-1676	65	16	.	.	PUNCT
ejpam-1676	65	17	.	.	PUNCT
ejpam-1676	66	1	,	,	PUNCT
ejpam-1676	66	2	λm	λm	X
ejpam-1676	66	3	)	)	PUNCT
ejpam-1676	66	4	,	,	PUNCT
ejpam-1676	66	5	the	the	DET
ejpam-1676	66	6	“	"	PUNCT
ejpam-1676	66	7	weights	weight	NOUN
ejpam-1676	66	8	vector	vector	NOUN
ejpam-1676	66	9	”	"	PUNCT
ejpam-1676	66	10	,	,	PUNCT
ejpam-1676	66	11	must	must	AUX
ejpam-1676	66	12	be	be	AUX
ejpam-1676	66	13	specified	specify	VERB
ejpam-1676	66	14	.	.	PUNCT
ejpam-1676	67	1	we	we	PRON
ejpam-1676	67	2	emphasize	emphasize	VERB
ejpam-1676	67	3	that	that	SCONJ
ejpam-1676	67	4	the	the	DET
ejpam-1676	67	5	parameters	parameter	NOUN
ejpam-1676	67	6	from	from	ADP
ejpam-1676	67	7	the	the	DET
ejpam-1676	67	8	related	related	ADJ
ejpam-1676	67	9	populations	population	NOUN
ejpam-1676	67	10	,	,	PUNCT
ejpam-1676	67	11	θ2	θ2	PROPN
ejpam-1676	67	12	,	,	PUNCT
ejpam-1676	67	13	.	.	PUNCT
ejpam-1676	67	14	.	.	PUNCT
ejpam-1676	68	1	.	.	PUNCT
ejpam-1676	69	1	,	,	PUNCT
ejpam-1676	69	2	θm	θm	PROPN
ejpam-1676	69	3	,	,	PUNCT
ejpam-1676	69	4	which	which	PRON
ejpam-1676	69	5	are	be	AUX
ejpam-1676	69	6	unknown	unknown	ADJ
ejpam-1676	69	7	,	,	PUNCT
ejpam-1676	69	8	do	do	AUX
ejpam-1676	69	9	not	not	PART
ejpam-1676	69	10	appear	appear	VERB
ejpam-1676	69	11	in	in	ADP
ejpam-1676	69	12	the	the	DET
ejpam-1676	69	13	wl	wl	PROPN
ejpam-1676	69	14	,	,	PUNCT
ejpam-1676	69	15	since	since	SCONJ
ejpam-1676	69	16	the	the	DET
ejpam-1676	69	17	inferential	inferential	ADJ
ejpam-1676	69	18	interest	interest	NOUN
ejpam-1676	69	19	focuses	focus	VERB
ejpam-1676	69	20	on	on	ADP
ejpam-1676	69	21	θ1	θ1	NOUN
ejpam-1676	69	22	of	of	ADP
ejpam-1676	69	23	the	the	DET
ejpam-1676	69	24	first	first	ADJ
ejpam-1676	69	25	population	population	NOUN
ejpam-1676	69	26	.	.	PUNCT
ejpam-1676	70	1	the	the	DET
ejpam-1676	70	2	wl	wl	PROPN
ejpam-1676	70	3	estimator	estimator	PROPN
ejpam-1676	70	4	(	(	PUNCT
ejpam-1676	70	5	wle	wle	PROPN
ejpam-1676	70	6	)	)	PUNCT
ejpam-1676	70	7	for	for	ADP
ejpam-1676	70	8	θ1	θ1	NOUN
ejpam-1676	70	9	,	,	PUNCT
ejpam-1676	70	10	say	say	VERB
ejpam-1676	70	11	θ̃1	θ̃1	NOUN
ejpam-1676	70	12	,	,	PUNCT
ejpam-1676	70	13	is	be	AUX
ejpam-1676	70	14	defined	define	VERB
ejpam-1676	70	15	as	as	ADP
ejpam-1676	70	16	the	the	DET
ejpam-1676	70	17	maximizer	maximizer	NOUN
ejpam-1676	70	18	of	of	ADP
ejpam-1676	70	19	the	the	DET
ejpam-1676	70	20	objective	objective	ADJ
ejpam-1676	70	21	function	function	NOUN
ejpam-1676	70	22	wl(y;θ1	wl(y;θ1	PROPN
ejpam-1676	70	23	):	):	PUNCT
ejpam-1676	70	24	θ̃1	θ̃1	NOUN
ejpam-1676	70	25	=	=	NOUN
ejpam-1676	70	26	arg	arg	NOUN
ejpam-1676	70	27	sup	sup	NOUN
ejpam-1676	70	28	θ1∈θ	θ1∈θ	NUM
ejpam-1676	70	29	wl(y;θ1	wl(y;θ1	NOUN
ejpam-1676	70	30	)	)	PUNCT
ejpam-1676	70	31	.	.	PUNCT
ejpam-1676	71	1	in	in	ADP
ejpam-1676	71	2	order	order	NOUN
ejpam-1676	71	3	for	for	SCONJ
ejpam-1676	71	4	the	the	DET
ejpam-1676	71	5	weighted	weight	VERB
ejpam-1676	71	6	likelihood	likelihood	NOUN
ejpam-1676	71	7	to	to	PART
ejpam-1676	71	8	be	be	AUX
ejpam-1676	71	9	effective	effective	ADJ
ejpam-1676	71	10	,	,	PUNCT
ejpam-1676	71	11	the	the	DET
ejpam-1676	71	12	weights	weight	NOUN
ejpam-1676	71	13	must	must	AUX
ejpam-1676	71	14	be	be	AUX
ejpam-1676	71	15	chosen	choose	VERB
ejpam-1676	71	16	adaptively	adaptively	ADV
ejpam-1676	71	17	so	so	SCONJ
ejpam-1676	71	18	that	that	SCONJ
ejpam-1676	71	19	all	all	DET
ejpam-1676	71	20	information	information	NOUN
ejpam-1676	71	21	obtained	obtain	VERB
ejpam-1676	71	22	from	from	ADP
ejpam-1676	71	23	related	related	ADJ
ejpam-1676	71	24	population	population	NOUN
ejpam-1676	71	25	must	must	AUX
ejpam-1676	71	26	be	be	AUX
ejpam-1676	71	27	evaluated	evaluate	VERB
ejpam-1676	71	28	to	to	PART
ejpam-1676	71	29	determine	determine	VERB
ejpam-1676	71	30	their	their	PRON
ejpam-1676	71	31	relevance	relevance	NOUN
ejpam-1676	71	32	.	.	PUNCT
ejpam-1676	72	1	[	[	X
ejpam-1676	72	2	7	7	X
ejpam-1676	72	3	]	]	PUNCT
ejpam-1676	72	4	proposed	propose	VERB
ejpam-1676	72	5	a	a	DET
ejpam-1676	72	6	cross	cross	ADJ
ejpam-1676	72	7	-	-	ADJ
ejpam-1676	72	8	validatory	validatory	ADJ
ejpam-1676	72	9	procedure	procedure	NOUN
ejpam-1676	72	10	for	for	ADP
ejpam-1676	72	11	weights	weight	NOUN
ejpam-1676	72	12	selection	selection	NOUN
ejpam-1676	72	13	and	and	CCONJ
ejpam-1676	72	14	showed	show	VERB
ejpam-1676	72	15	that	that	SCONJ
ejpam-1676	72	16	the	the	DET
ejpam-1676	72	17	resulting	result	VERB
ejpam-1676	72	18	wle	wle	NOUN
ejpam-1676	72	19	is	be	AUX
ejpam-1676	72	20	consistent	consistent	ADJ
ejpam-1676	72	21	and	and	CCONJ
ejpam-1676	72	22	asymptotically	asymptotically	ADV
ejpam-1676	72	23	normal	normal	ADJ
ejpam-1676	72	24	.	.	PUNCT
ejpam-1676	73	1	however	however	ADV
ejpam-1676	73	2	,	,	PUNCT
ejpam-1676	73	3	the	the	DET
ejpam-1676	73	4	cross	cross	ADJ
ejpam-1676	73	5	-	-	ADJ
ejpam-1676	73	6	validatory	validatory	ADJ
ejpam-1676	73	7	approach	approach	NOUN
ejpam-1676	73	8	is	be	AUX
ejpam-1676	73	9	challenged	challenge	VERB
ejpam-1676	73	10	with	with	ADP
ejpam-1676	73	11	the	the	DET
ejpam-1676	73	12	following	follow	VERB
ejpam-1676	73	13	problems	problem	NOUN
ejpam-1676	73	14	:	:	PUNCT
ejpam-1676	73	15	(	(	PUNCT
ejpam-1676	73	16	a	a	X
ejpam-1676	73	17	)	)	PUNCT
ejpam-1676	73	18	it	it	PRON
ejpam-1676	73	19	is	be	AUX
ejpam-1676	73	20	computationally	computationally	ADV
ejpam-1676	73	21	intensive	intensive	ADJ
ejpam-1676	73	22	if	if	SCONJ
ejpam-1676	73	23	the	the	DET
ejpam-1676	73	24	wle	wle	NOUN
ejpam-1676	73	25	does	do	AUX
ejpam-1676	73	26	not	not	PART
ejpam-1676	73	27	have	have	VERB
ejpam-1676	73	28	an	an	DET
ejpam-1676	73	29	analytical	analytical	ADJ
ejpam-1676	73	30	form	form	NOUN
ejpam-1676	73	31	,	,	PUNCT
ejpam-1676	73	32	and	and	CCONJ
ejpam-1676	73	33	(	(	PUNCT
ejpam-1676	73	34	b	b	X
ejpam-1676	73	35	)	)	PUNCT
ejpam-1676	73	36	numerical	numerical	ADJ
ejpam-1676	73	37	performances	performance	NOUN
ejpam-1676	73	38	could	could	AUX
ejpam-1676	73	39	be	be	AUX
ejpam-1676	73	40	unstable	unstable	ADJ
ejpam-1676	73	41	when	when	SCONJ
ejpam-1676	73	42	the	the	DET
ejpam-1676	73	43	sample	sample	NOUN
ejpam-1676	73	44	sizes	size	NOUN
ejpam-1676	73	45	are	be	AUX
ejpam-1676	73	46	unequal	unequal	ADJ
ejpam-1676	73	47	or	or	CCONJ
ejpam-1676	73	48	the	the	DET
ejpam-1676	73	49	sample	sample	NOUN
ejpam-1676	73	50	sizes	size	NOUN
ejpam-1676	73	51	are	be	AUX
ejpam-1676	73	52	very	very	ADV
ejpam-1676	73	53	small	small	ADJ
ejpam-1676	73	54	.	.	PUNCT
ejpam-1676	74	1	thus	thus	ADV
ejpam-1676	74	2	,	,	PUNCT
ejpam-1676	74	3	we	we	PRON
ejpam-1676	74	4	propose	propose	VERB
ejpam-1676	74	5	a	a	DET
ejpam-1676	74	6	simple	simple	ADJ
ejpam-1676	74	7	and	and	CCONJ
ejpam-1676	74	8	effective	effective	ADJ
ejpam-1676	74	9	method	method	NOUN
ejpam-1676	74	10	to	to	PART
ejpam-1676	74	11	choose	choose	VERB
ejpam-1676	74	12	the	the	DET
ejpam-1676	74	13	likelihood	likelihood	NOUN
ejpam-1676	74	14	weights	weight	VERB
ejpam-1676	74	15	adaptively	adaptively	ADV
ejpam-1676	74	16	.	.	PUNCT
ejpam-1676	75	1	this	this	DET
ejpam-1676	75	2	method	method	NOUN
ejpam-1676	75	3	is	be	AUX
ejpam-1676	75	4	better	well	ADJ
ejpam-1676	75	5	than	than	ADP
ejpam-1676	75	6	the	the	DET
ejpam-1676	75	7	cross	cross	ADJ
ejpam-1676	75	8	-	-	ADJ
ejpam-1676	75	9	validated	validated	ADJ
ejpam-1676	75	10	weights	weight	NOUN
ejpam-1676	75	11	proposed	propose	VERB
ejpam-1676	75	12	by	by	ADP
ejpam-1676	75	13	[	[	X
ejpam-1676	75	14	7	7	X
ejpam-1676	75	15	]	]	PUNCT
ejpam-1676	75	16	in	in	ADP
ejpam-1676	75	17	that	that	SCONJ
ejpam-1676	75	18	its	its	PRON
ejpam-1676	75	19	computation	computation	NOUN
ejpam-1676	75	20	is	be	AUX
ejpam-1676	75	21	robust	robust	ADJ
ejpam-1676	75	22	and	and	CCONJ
ejpam-1676	75	23	implementation	implementation	NOUN
ejpam-1676	75	24	is	be	AUX
ejpam-1676	75	25	straightforward	straightforward	ADJ
ejpam-1676	75	26	.	.	PUNCT
ejpam-1676	76	1	more	more	ADV
ejpam-1676	76	2	importantly	importantly	ADV
ejpam-1676	76	3	,	,	PUNCT
ejpam-1676	76	4	it	it	PRON
ejpam-1676	76	5	avoids	avoid	VERB
ejpam-1676	76	6	all	all	DET
ejpam-1676	76	7	the	the	DET
ejpam-1676	76	8	numerical	numerical	ADJ
ejpam-1676	76	9	problems	problem	NOUN
ejpam-1676	76	10	with	with	ADP
ejpam-1676	76	11	the	the	DET
ejpam-1676	76	12	cross	cross	ADJ
ejpam-1676	76	13	-	-	ADJ
ejpam-1676	76	14	validated	validated	ADJ
ejpam-1676	76	15	weights	weight	NOUN
ejpam-1676	76	16	when	when	SCONJ
ejpam-1676	76	17	the	the	DET
ejpam-1676	76	18	sample	sample	NOUN
ejpam-1676	76	19	sizes	size	NOUN
ejpam-1676	76	20	are	be	AUX
ejpam-1676	76	21	small	small	ADJ
ejpam-1676	76	22	or	or	CCONJ
ejpam-1676	76	23	unequal	unequal	ADJ
ejpam-1676	76	24	.	.	PUNCT
ejpam-1676	77	1	we	we	PRON
ejpam-1676	77	2	derive	derive	VERB
ejpam-1676	77	3	the	the	DET
ejpam-1676	77	4	consistency	consistency	NOUN
ejpam-1676	77	5	and	and	CCONJ
ejpam-1676	77	6	asymptotic	asymptotic	ADJ
ejpam-1676	77	7	distribution	distribution	NOUN
ejpam-1676	77	8	of	of	ADP
ejpam-1676	77	9	the	the	DET
ejpam-1676	77	10	resulting	result	VERB
ejpam-1676	77	11	wle	wle	NOUN
ejpam-1676	77	12	.	.	PUNCT
ejpam-1676	78	1	the	the	DET
ejpam-1676	78	2	article	article	NOUN
ejpam-1676	78	3	is	be	AUX
ejpam-1676	78	4	organized	organize	VERB
ejpam-1676	78	5	as	as	SCONJ
ejpam-1676	78	6	follows	follow	VERB
ejpam-1676	78	7	.	.	PUNCT
ejpam-1676	79	1	in	in	ADP
ejpam-1676	79	2	section	section	NOUN
ejpam-1676	79	3	2	2	NUM
ejpam-1676	79	4	,	,	PUNCT
ejpam-1676	79	5	we	we	PRON
ejpam-1676	79	6	develop	develop	VERB
ejpam-1676	79	7	a	a	DET
ejpam-1676	79	8	wl	wl	NOUN
ejpam-1676	79	9	procedure	procedure	NOUN
ejpam-1676	79	10	.	.	PUNCT
ejpam-1676	80	1	the	the	DET
ejpam-1676	80	2	asymptotic	asymptotic	ADJ
ejpam-1676	80	3	properties	property	NOUN
ejpam-1676	80	4	of	of	ADP
ejpam-1676	80	5	the	the	DET
ejpam-1676	80	6	wle	wle	NOUN
ejpam-1676	80	7	for	for	ADP
ejpam-1676	80	8	generalized	generalized	ADJ
ejpam-1676	80	9	linear	linear	NOUN
ejpam-1676	80	10	models	model	NOUN
ejpam-1676	80	11	using	use	VERB
ejpam-1676	80	12	the	the	DET
ejpam-1676	80	13	proposed	propose	VERB
ejpam-1676	80	14	adaptive	adaptive	ADJ
ejpam-1676	80	15	weights	weight	NOUN
ejpam-1676	80	16	are	be	AUX
ejpam-1676	80	17	presented	present	VERB
ejpam-1676	80	18	in	in	ADP
ejpam-1676	80	19	section	section	NOUN
ejpam-1676	80	20	3	3	NUM
ejpam-1676	80	21	.	.	PUNCT
ejpam-1676	80	22	section	section	NOUN
ejpam-1676	80	23	4	4	NUM
ejpam-1676	80	24	presents	present	VERB
ejpam-1676	80	25	the	the	DET
ejpam-1676	80	26	results	result	NOUN
ejpam-1676	80	27	from	from	ADP
ejpam-1676	80	28	simulation	simulation	NOUN
ejpam-1676	80	29	experiments	experiment	NOUN
ejpam-1676	80	30	that	that	PRON
ejpam-1676	80	31	illustrate	illustrate	VERB
ejpam-1676	80	32	the	the	DET
ejpam-1676	80	33	numerical	numerical	ADJ
ejpam-1676	80	34	performance	performance	NOUN
ejpam-1676	80	35	of	of	ADP
ejpam-1676	80	36	our	our	PRON
ejpam-1676	80	37	method	method	NOUN
ejpam-1676	80	38	.	.	PUNCT
ejpam-1676	81	1	section	section	NOUN
ejpam-1676	81	2	5	5	NUM
ejpam-1676	81	3	shows	show	VERB
ejpam-1676	81	4	how	how	SCONJ
ejpam-1676	81	5	to	to	PART
ejpam-1676	81	6	obtain	obtain	VERB
ejpam-1676	81	7	the	the	DET
ejpam-1676	81	8	empirical	empirical	ADJ
ejpam-1676	81	9	distribution	distribution	NOUN
ejpam-1676	81	10	of	of	ADP
ejpam-1676	81	11	the	the	DET
ejpam-1676	81	12	wle	wle	NOUN
ejpam-1676	81	13	by	by	ADP
ejpam-1676	81	14	bootstrap	bootstrap	NOUN
ejpam-1676	81	15	.	.	PUNCT
ejpam-1676	82	1	section	section	NOUN
ejpam-1676	82	2	6	6	NUM
ejpam-1676	82	3	analyzes	analyze	VERB
ejpam-1676	82	4	a	a	DET
ejpam-1676	82	5	real	real	ADJ
ejpam-1676	82	6	data	datum	NOUN
ejpam-1676	82	7	set	set	VERB
ejpam-1676	82	8	from	from	ADP
ejpam-1676	82	9	by	by	ADP
ejpam-1676	82	10	using	use	VERB
ejpam-1676	82	11	the	the	DET
ejpam-1676	82	12	proposed	propose	VERB
ejpam-1676	82	13	approach	approach	NOUN
ejpam-1676	82	14	.	.	PUNCT
ejpam-1676	83	1	section	section	NOUN
ejpam-1676	83	2	7	7	NUM
ejpam-1676	83	3	provides	provide	VERB
ejpam-1676	83	4	conclusions	conclusion	NOUN
ejpam-1676	83	5	and	and	CCONJ
ejpam-1676	83	6	discussions	discussion	NOUN
ejpam-1676	83	7	.	.	PUNCT
ejpam-1676	84	1	2	2	X
ejpam-1676	84	2	.	.	X
ejpam-1676	84	3	the	the	DET
ejpam-1676	84	4	weighted	weight	VERB
ejpam-1676	84	5	likelihood	likelihood	NOUN
ejpam-1676	84	6	and	and	CCONJ
ejpam-1676	84	7	adaptive	adaptive	ADJ
ejpam-1676	84	8	weights	weight	NOUN
ejpam-1676	84	9	2.1	2.1	NUM
ejpam-1676	84	10	.	.	PUNCT
ejpam-1676	85	1	the	the	DET
ejpam-1676	85	2	weighted	weight	VERB
ejpam-1676	85	3	likelihood	likelihood	NOUN
ejpam-1676	85	4	we	we	PRON
ejpam-1676	85	5	assume	assume	VERB
ejpam-1676	85	6	the	the	DET
ejpam-1676	85	7	existence	existence	NOUN
ejpam-1676	85	8	of	of	ADP
ejpam-1676	85	9	the	the	DET
ejpam-1676	85	10	m	m	PROPN
ejpam-1676	85	11	population	population	NOUN
ejpam-1676	85	12	density	density	NOUN
ejpam-1676	85	13	functions	function	NOUN
ejpam-1676	85	14	are	be	AUX
ejpam-1676	85	15	unknown	unknown	ADJ
ejpam-1676	85	16	and	and	CCONJ
ejpam-1676	85	17	play	play	VERB
ejpam-1676	85	18	purely	purely	ADV
ejpam-1676	85	19	conceptual	conceptual	ADJ
ejpam-1676	85	20	roles	role	NOUN
ejpam-1676	85	21	.	.	PUNCT
ejpam-1676	86	1	more	more	ADV
ejpam-1676	86	2	specifically	specifically	ADV
ejpam-1676	86	3	,	,	PUNCT
ejpam-1676	86	4	assume	assume	VERB
ejpam-1676	86	5	σ	σ	NOUN
ejpam-1676	86	6	-	-	ADJ
ejpam-1676	86	7	finite	finite	ADJ
ejpam-1676	86	8	probability	probability	NOUN
ejpam-1676	86	9	spaces	space	NOUN
ejpam-1676	86	10	(	(	PUNCT
ejpam-1676	86	11	ω	ω	PROPN
ejpam-1676	86	12	,	,	PUNCT
ejpam-1676	86	13	f	f	PROPN
ejpam-1676	86	14	,	,	PUNCT
ejpam-1676	86	15	µi	µi	PROPN
ejpam-1676	86	16	)	)	PUNCT
ejpam-1676	86	17	,	,	PUNCT
ejpam-1676	87	1	i	i	NOUN
ejpam-1676	87	2	=	=	NOUN
ejpam-1676	87	3	1,2	1,2	NUM
ejpam-1676	87	4	,	,	PUNCT
ejpam-1676	87	5	.	.	PUNCT
ejpam-1676	87	6	.	.	PUNCT
ejpam-1676	88	1	.	.	PUNCT
ejpam-1676	89	1	,	,	PUNCT
ejpam-1676	89	2	m	m	VERB
ejpam-1676	89	3	,	,	PUNCT
ejpam-1676	89	4	with	with	ADP
ejpam-1676	89	5	probability	probability	NOUN
ejpam-1676	89	6	measures	measure	NOUN
ejpam-1676	89	7	µi	µi	INTJ
ejpam-1676	89	8	’s	’	NOUN
ejpam-1676	89	9	that	that	PRON
ejpam-1676	89	10	are	be	AUX
ejpam-1676	89	11	absolutely	absolutely	ADV
ejpam-1676	89	12	continuous	continuous	ADJ
ejpam-1676	89	13	with	with	ADP
ejpam-1676	89	14	respect	respect	NOUN
ejpam-1676	89	15	to	to	ADP
ejpam-1676	89	16	one	one	NUM
ejpam-1676	89	17	another	another	DET
ejpam-1676	89	18	.	.	PUNCT
ejpam-1676	90	1	the	the	DET
ejpam-1676	90	2	existence	existence	NOUN
ejpam-1676	90	3	of	of	ADP
ejpam-1676	90	4	a	a	DET
ejpam-1676	90	5	σ	σ	NOUN
ejpam-1676	90	6	-	-	PUNCT
ejpam-1676	90	7	finite	finite	ADJ
ejpam-1676	90	8	measure	measure	NOUN
ejpam-1676	90	9	ν	ν	NOUN
ejpam-1676	90	10	that	that	PRON
ejpam-1676	90	11	dominates	dominate	VERB
ejpam-1676	90	12	the	the	DET
ejpam-1676	90	13	µi	µi	PROPN
ejpam-1676	90	14	’s	’	VERB
ejpam-1676	90	15	then	then	ADV
ejpam-1676	90	16	follows	follow	VERB
ejpam-1676	90	17	.	.	PUNCT
ejpam-1676	91	1	we	we	PRON
ejpam-1676	91	2	take	take	VERB
ejpam-1676	91	3	the	the	DET
ejpam-1676	91	4	fi	fi	NOUN
ejpam-1676	91	5	to	to	PART
ejpam-1676	91	6	be	be	AUX
ejpam-1676	91	7	the	the	DET
ejpam-1676	91	8	radon	radon	PROPN
ejpam-1676	91	9	-	-	PUNCT
ejpam-1676	91	10	nikodym	nikodym	ADJ
ejpam-1676	91	11	derivatives	derivative	NOUN
ejpam-1676	91	12	of	of	ADP
ejpam-1676	91	13	µi	µi	PROPN
ejpam-1676	91	14	with	with	ADP
ejpam-1676	91	15	respect	respect	NOUN
ejpam-1676	91	16	to	to	ADP
ejpam-1676	91	17	ν	ν	NOUN
ejpam-1676	91	18	for	for	ADP
ejpam-1676	91	19	i	i	X
ejpam-1676	91	20	=	=	SYM
ejpam-1676	91	21	1,2	1,2	NUM
ejpam-1676	91	22	,	,	PUNCT
ejpam-1676	91	23	.	.	PUNCT
ejpam-1676	91	24	.	.	PUNCT
ejpam-1676	92	1	.	.	PUNCT
ejpam-1676	93	1	,	,	PUNCT
ejpam-1676	93	2	m.	m.	NOUN
ejpam-1676	93	3	suppose	suppose	VERB
ejpam-1676	93	4	that	that	SCONJ
ejpam-1676	93	5	we	we	PRON
ejpam-1676	93	6	observe	observe	VERB
ejpam-1676	93	7	independent	independent	ADJ
ejpam-1676	93	8	random	random	ADJ
ejpam-1676	93	9	response	response	NOUN
ejpam-1676	93	10	vectors	vector	NOUN
ejpam-1676	93	11	y	y	PROPN
ejpam-1676	93	12	1	1	NUM
ejpam-1676	93	13	,	,	PUNCT
ejpam-1676	93	14	.	.	PUNCT
ejpam-1676	93	15	.	.	PUNCT
ejpam-1676	94	1	.	.	PUNCT
ejpam-1676	95	1	,	,	PUNCT
ejpam-1676	96	1	y	y	PROPN
ejpam-1676	96	2	m	m	VERB
ejpam-1676	96	3	with	with	ADP
ejpam-1676	96	4	probability	probability	NOUN
ejpam-1676	96	5	density	density	NOUN
ejpam-1676	96	6	functions	function	NOUN
ejpam-1676	96	7	f	f	X
ejpam-1676	96	8	(	(	PUNCT
ejpam-1676	96	9	.;θ1	.;θ1	PROPN
ejpam-1676	96	10	)	)	PUNCT
ejpam-1676	96	11	,	,	PUNCT
ejpam-1676	96	12	.	.	PUNCT
ejpam-1676	96	13	.	.	PUNCT
ejpam-1676	96	14	.	.	PUNCT
ejpam-1676	97	1	,	,	PUNCT
ejpam-1676	97	2	f	f	PROPN
ejpam-1676	97	3	(	(	PUNCT
ejpam-1676	97	4	.;θm	.;θm	NOUN
ejpam-1676	97	5	)	)	PUNCT
ejpam-1676	97	6	,	,	PUNCT
ejpam-1676	97	7	where	where	SCONJ
ejpam-1676	97	8	y	y	PROPN
ejpam-1676	97	9	i	i	NOUN
ejpam-1676	97	10	=	=	PUNCT
ejpam-1676	97	11	(	(	PUNCT
ejpam-1676	97	12	yi1	yi1	NOUN
ejpam-1676	97	13	,	,	PUNCT
ejpam-1676	97	14	.	.	PUNCT
ejpam-1676	97	15	.	.	PUNCT
ejpam-1676	97	16	.	.	PUNCT
ejpam-1676	98	1	,	,	PUNCT
ejpam-1676	98	2	yini	yini	PROPN
ejpam-1676	98	3	)	)	PUNCT
ejpam-1676	98	4	t	t	PROPN
ejpam-1676	98	5	,	,	PUNCT
ejpam-1676	98	6	i	i	NOUN
ejpam-1676	98	7	=	=	SYM
ejpam-1676	98	8	1,2	1,2	NUM
ejpam-1676	98	9	,	,	PUNCT
ejpam-1676	98	10	·	·	PUNCT
ejpam-1676	98	11	·	·	PUNCT
ejpam-1676	98	12	·	·	PUNCT
ejpam-1676	98	13	,	,	PUNCT
ejpam-1676	98	14	m.	m.	NOUN
ejpam-1676	98	15	further	further	ADJ
ejpam-1676	98	16	p.	p.	PROPN
ejpam-1676	98	17	guo	guo	PROPN
ejpam-1676	98	18	,	,	PUNCT
ejpam-1676	98	19	x.	x.	PROPN
ejpam-1676	98	20	wang	wang	PROPN
ejpam-1676	98	21	,	,	PUNCT
ejpam-1676	98	22	y.	y.	PROPN
ejpam-1676	98	23	wu	wu	PROPN
ejpam-1676	98	24	/	/	SYM
ejpam-1676	98	25	eur	eur	PROPN
ejpam-1676	98	26	.	.	PUNCT
ejpam-1676	99	1	j.	j.	PROPN
ejpam-1676	99	2	pure	pure	PROPN
ejpam-1676	99	3	appl	appl	PROPN
ejpam-1676	99	4	.	.	PROPN
ejpam-1676	99	5	math	math	PROPN
ejpam-1676	99	6	,	,	PUNCT
ejpam-1676	99	7	5	5	NUM
ejpam-1676	99	8	(	(	PUNCT
ejpam-1676	99	9	2012	2012	NUM
ejpam-1676	99	10	)	)	PUNCT
ejpam-1676	99	11	,	,	PUNCT
ejpam-1676	99	12	333	333	NUM
ejpam-1676	99	13	-	-	SYM
ejpam-1676	99	14	356	356	NUM
ejpam-1676	99	15	336	336	NUM
ejpam-1676	99	16	suppose	suppose	VERB
ejpam-1676	99	17	that	that	SCONJ
ejpam-1676	99	18	only	only	ADV
ejpam-1676	99	19	population	population	NOUN
ejpam-1676	99	20	1	1	NUM
ejpam-1676	99	21	,	,	PUNCT
ejpam-1676	99	22	in	in	ADP
ejpam-1676	99	23	particular	particular	ADJ
ejpam-1676	99	24	θ1	θ1	NOUN
ejpam-1676	99	25	,	,	PUNCT
ejpam-1676	99	26	an	an	DET
ejpam-1676	99	27	unknown	unknown	ADJ
ejpam-1676	99	28	vector	vector	NOUN
ejpam-1676	99	29	of	of	ADP
ejpam-1676	99	30	parameters	parameter	NOUN
ejpam-1676	99	31	,	,	PUNCT
ejpam-1676	99	32	is	be	AUX
ejpam-1676	99	33	of	of	ADP
ejpam-1676	99	34	inferential	inferential	ADJ
ejpam-1676	99	35	interest	interest	NOUN
ejpam-1676	99	36	.	.	PUNCT
ejpam-1676	100	1	given	give	VERB
ejpam-1676	100	2	data	datum	NOUN
ejpam-1676	100	3	,	,	PUNCT
ejpam-1676	100	4	y1	y1	NOUN
ejpam-1676	100	5	=	=	SYM
ejpam-1676	100	6	y1	y1	PROPN
ejpam-1676	100	7	,	,	PUNCT
ejpam-1676	100	8	the	the	DET
ejpam-1676	100	9	classical	classical	ADJ
ejpam-1676	100	10	likelihood	likelihood	NOUN
ejpam-1676	100	11	of	of	ADP
ejpam-1676	100	12	θ1	θ1	NOUN
ejpam-1676	100	13	is	be	AUX
ejpam-1676	100	14	defined	define	VERB
ejpam-1676	100	15	by	by	ADP
ejpam-1676	100	16	l1(y1;θ1	l1(y1;θ1	ADJ
ejpam-1676	100	17	)	)	PUNCT
ejpam-1676	100	18	=	=	SYM
ejpam-1676	100	19	n1∏	n1∏	NOUN
ejpam-1676	100	20	j=1	j=1	PROPN
ejpam-1676	100	21	f	f	PROPN
ejpam-1676	100	22	(	(	PUNCT
ejpam-1676	100	23	y1	y1	INTJ
ejpam-1676	100	24	j;θ1	j;θ1	PROPN
ejpam-1676	100	25	)	)	PUNCT
ejpam-1676	100	26	.	.	PUNCT
ejpam-1676	101	1	similarly	similarly	ADV
ejpam-1676	101	2	,	,	PUNCT
ejpam-1676	101	3	we	we	PRON
ejpam-1676	101	4	can	can	AUX
ejpam-1676	101	5	define	define	VERB
ejpam-1676	101	6	the	the	DET
ejpam-1676	101	7	likelihood	likelihood	NOUN
ejpam-1676	101	8	for	for	SCONJ
ejpam-1676	101	9	each	each	DET
ejpam-1676	101	10	θi	θi	NOUN
ejpam-1676	101	11	as	as	SCONJ
ejpam-1676	101	12	follows	follow	VERB
ejpam-1676	101	13	:	:	PUNCT
ejpam-1676	101	14	li(y	li(y	X
ejpam-1676	101	15	i	i	PRON
ejpam-1676	101	16	,	,	PUNCT
ejpam-1676	101	17	θ1	θ1	NOUN
ejpam-1676	101	18	)	)	PUNCT
ejpam-1676	101	19	=	=	PUNCT
ejpam-1676	102	1	ni∏	ni∏	NOUN
ejpam-1676	103	1	j=1	j=1	ADJ
ejpam-1676	103	2	f	f	PROPN
ejpam-1676	103	3	(	(	PUNCT
ejpam-1676	103	4	yi	yi	PROPN
ejpam-1676	103	5	j;θi	j;θi	PROPN
ejpam-1676	103	6	)	)	PUNCT
ejpam-1676	103	7	.	.	PUNCT
ejpam-1676	104	1	where	where	SCONJ
ejpam-1676	104	2	i	i	PRON
ejpam-1676	104	3	=	=	NOUN
ejpam-1676	104	4	2	2	NUM
ejpam-1676	104	5	,	,	PUNCT
ejpam-1676	104	6	·	·	PUNCT
ejpam-1676	104	7	·	·	PUNCT
ejpam-1676	104	8	·	·	PUNCT
ejpam-1676	104	9	,	,	PUNCT
ejpam-1676	104	10	m.	m.	NOUN
ejpam-1676	104	11	if	if	SCONJ
ejpam-1676	104	12	all	all	DET
ejpam-1676	104	13	parameters	parameter	NOUN
ejpam-1676	104	14	are	be	AUX
ejpam-1676	104	15	not	not	PART
ejpam-1676	104	16	related	relate	VERB
ejpam-1676	104	17	in	in	ADP
ejpam-1676	104	18	any	any	DET
ejpam-1676	104	19	way	way	NOUN
ejpam-1676	104	20	,	,	PUNCT
ejpam-1676	104	21	the	the	DET
ejpam-1676	104	22	correct	correct	ADJ
ejpam-1676	104	23	likelihood	likelihood	NOUN
ejpam-1676	104	24	for	for	ADP
ejpam-1676	104	25	θ1	θ1	NOUN
ejpam-1676	104	26	should	should	AUX
ejpam-1676	104	27	be	be	AUX
ejpam-1676	104	28	l1(y1;θ1	l1(y1;θ1	ADJ
ejpam-1676	104	29	)	)	PUNCT
ejpam-1676	104	30	.	.	PUNCT
ejpam-1676	105	1	however	however	ADV
ejpam-1676	105	2	,	,	PUNCT
ejpam-1676	105	3	l1(y1;θ1	l1(y1;θ1	ADJ
ejpam-1676	105	4	)	)	PUNCT
ejpam-1676	105	5	would	would	AUX
ejpam-1676	105	6	not	not	PART
ejpam-1676	105	7	be	be	AUX
ejpam-1676	105	8	the	the	DET
ejpam-1676	105	9	correct	correct	ADJ
ejpam-1676	105	10	likelihood	likelihood	NOUN
ejpam-1676	105	11	if	if	SCONJ
ejpam-1676	105	12	there	there	PRON
ejpam-1676	105	13	exist	exist	VERB
ejpam-1676	105	14	functional	functional	ADJ
ejpam-1676	105	15	relationship	relationship	NOUN
ejpam-1676	105	16	among	among	ADP
ejpam-1676	105	17	all	all	DET
ejpam-1676	105	18	parameters	parameter	NOUN
ejpam-1676	105	19	.	.	PUNCT
ejpam-1676	106	1	to	to	PART
ejpam-1676	106	2	be	be	AUX
ejpam-1676	106	3	more	more	ADV
ejpam-1676	106	4	specific	specific	ADJ
ejpam-1676	106	5	,	,	PUNCT
ejpam-1676	106	6	assume	assume	VERB
ejpam-1676	106	7	that	that	SCONJ
ejpam-1676	106	8	θ2	θ2	PROPN
ejpam-1676	106	9	=	=	PROPN
ejpam-1676	106	10	g1(θ1	g1(θ1	PROPN
ejpam-1676	106	11	)	)	PUNCT
ejpam-1676	106	12	,	,	PUNCT
ejpam-1676	106	13	.	.	PUNCT
ejpam-1676	106	14	.	.	PUNCT
ejpam-1676	107	1	.	.	PUNCT
ejpam-1676	108	1	,	,	PUNCT
ejpam-1676	108	2	θm	θm	PROPN
ejpam-1676	108	3	=	=	SYM
ejpam-1676	108	4	gm(θ1	gm(θ1	NOUN
ejpam-1676	108	5	)	)	PUNCT
ejpam-1676	108	6	,	,	PUNCT
ejpam-1676	108	7	where	where	SCONJ
ejpam-1676	108	8	gi	gi	NOUN
ejpam-1676	108	9	are	be	AUX
ejpam-1676	108	10	measurable	measurable	ADJ
ejpam-1676	108	11	functions	function	NOUN
ejpam-1676	108	12	of	of	ADP
ejpam-1676	108	13	θ1	θ1	NOUN
ejpam-1676	108	14	.	.	PUNCT
ejpam-1676	109	1	furthermore	furthermore	ADV
ejpam-1676	109	2	,	,	PUNCT
ejpam-1676	109	3	if	if	SCONJ
ejpam-1676	109	4	all	all	DET
ejpam-1676	109	5	observations	observation	NOUN
ejpam-1676	109	6	from	from	ADP
ejpam-1676	109	7	different	different	ADJ
ejpam-1676	109	8	populations	population	NOUN
ejpam-1676	109	9	are	be	AUX
ejpam-1676	109	10	independent	independent	ADJ
ejpam-1676	109	11	of	of	ADP
ejpam-1676	109	12	each	each	DET
ejpam-1676	109	13	other	other	ADJ
ejpam-1676	109	14	,	,	PUNCT
ejpam-1676	109	15	the	the	DET
ejpam-1676	109	16	likelihood	likelihood	NOUN
ejpam-1676	109	17	of	of	ADP
ejpam-1676	109	18	θ1	θ1	NOUN
ejpam-1676	109	19	involving	involve	VERB
ejpam-1676	109	20	all	all	DET
ejpam-1676	109	21	samples	sample	NOUN
ejpam-1676	109	22	should	should	AUX
ejpam-1676	109	23	be	be	AUX
ejpam-1676	109	24	l̃(y1	l̃(y1	PROPN
ejpam-1676	109	25	,	,	PUNCT
ejpam-1676	109	26	·	·	PUNCT
ejpam-1676	109	27	·	·	PUNCT
ejpam-1676	109	28	·	·	PUNCT
ejpam-1676	109	29	,	,	PUNCT
ejpam-1676	109	30	ym;θ1	ym;θ1	X
ejpam-1676	109	31	)	)	PUNCT
ejpam-1676	109	32	=	=	SYM
ejpam-1676	110	1	l1(y1,θi	l1(y1,θi	NUM
ejpam-1676	110	2	)	)	PUNCT
ejpam-1676	110	3	×	×	NOUN
ejpam-1676	110	4	m∏	m∏	PROPN
ejpam-1676	110	5	i=2	i=2	PROPN
ejpam-1676	110	6	li(y	li(y	PUNCT
ejpam-1676	110	7	i	i	PROPN
ejpam-1676	110	8	,	,	PUNCT
ejpam-1676	110	9	gi(θ1	gi(θ1	PROPN
ejpam-1676	110	10	)	)	PUNCT
ejpam-1676	110	11	)	)	PUNCT
ejpam-1676	110	12	.	.	PUNCT
ejpam-1676	111	1	we	we	PRON
ejpam-1676	111	2	observe	observe	VERB
ejpam-1676	111	3	that	that	SCONJ
ejpam-1676	111	4	all	all	DET
ejpam-1676	111	5	parameters	parameter	NOUN
ejpam-1676	111	6	,	,	PUNCT
ejpam-1676	111	7	θ2	θ2	PROPN
ejpam-1676	111	8	,	,	PUNCT
ejpam-1676	111	9	.	.	PUNCT
ejpam-1676	111	10	.	.	PUNCT
ejpam-1676	111	11	.	.	PUNCT
ejpam-1676	112	1	,	,	PUNCT
ejpam-1676	112	2	θm	θm	PROPN
ejpam-1676	112	3	disappeared	disappear	VERB
ejpam-1676	112	4	from	from	ADP
ejpam-1676	112	5	the	the	DET
ejpam-1676	112	6	above	above	ADJ
ejpam-1676	112	7	likelihood	likelihood	NOUN
ejpam-1676	112	8	since	since	SCONJ
ejpam-1676	112	9	they	they	PRON
ejpam-1676	112	10	are	be	AUX
ejpam-1676	112	11	replaced	replace	VERB
ejpam-1676	112	12	by	by	ADP
ejpam-1676	112	13	g2(θ1	g2(θ1	NOUN
ejpam-1676	112	14	)	)	PUNCT
ejpam-1676	112	15	,	,	PUNCT
ejpam-1676	112	16	.	.	PUNCT
ejpam-1676	112	17	.	.	PUNCT
ejpam-1676	113	1	.	.	PUNCT
ejpam-1676	114	1	,	,	PUNCT
ejpam-1676	114	2	gm(θm	gm(θm	PROPN
ejpam-1676	114	3	)	)	PUNCT
ejpam-1676	114	4	respectively	respectively	ADV
ejpam-1676	114	5	.	.	PUNCT
ejpam-1676	115	1	we	we	PRON
ejpam-1676	115	2	see	see	VERB
ejpam-1676	115	3	that	that	SCONJ
ejpam-1676	115	4	all	all	DET
ejpam-1676	115	5	data	datum	NOUN
ejpam-1676	115	6	sets	set	NOUN
ejpam-1676	115	7	are	be	AUX
ejpam-1676	115	8	integrated	integrate	VERB
ejpam-1676	115	9	into	into	ADP
ejpam-1676	115	10	one	one	NUM
ejpam-1676	115	11	likelihood	likelihood	NOUN
ejpam-1676	115	12	function	function	NOUN
ejpam-1676	115	13	that	that	PRON
ejpam-1676	115	14	we	we	PRON
ejpam-1676	115	15	call	call	VERB
ejpam-1676	115	16	fusion	fusion	NOUN
ejpam-1676	115	17	likelihood	likelihood	NOUN
ejpam-1676	115	18	.	.	PUNCT
ejpam-1676	116	1	when	when	SCONJ
ejpam-1676	116	2	the	the	DET
ejpam-1676	116	3	functional	functional	ADJ
ejpam-1676	116	4	relationship	relationship	NOUN
ejpam-1676	116	5	,	,	PUNCT
ejpam-1676	116	6	gi	gi	INTJ
ejpam-1676	116	7	,	,	PUNCT
ejpam-1676	116	8	are	be	AUX
ejpam-1676	116	9	indeed	indeed	ADV
ejpam-1676	116	10	known	know	VERB
ejpam-1676	116	11	,	,	PUNCT
ejpam-1676	116	12	we	we	PRON
ejpam-1676	116	13	can	can	AUX
ejpam-1676	116	14	derive	derive	VERB
ejpam-1676	116	15	the	the	DET
ejpam-1676	116	16	estimating	estimate	VERB
ejpam-1676	116	17	equation	equation	NOUN
ejpam-1676	116	18	by	by	ADP
ejpam-1676	116	19	consider	consider	VERB
ejpam-1676	116	20	the	the	DET
ejpam-1676	116	21	following	following	NOUN
ejpam-1676	116	22	:	:	PUNCT
ejpam-1676	116	23	log	log	VERB
ejpam-1676	116	24	l̃(y1	l̃(y1	PROPN
ejpam-1676	116	25	,	,	PUNCT
ejpam-1676	116	26	.	.	PUNCT
ejpam-1676	116	27	.	.	PUNCT
ejpam-1676	117	1	.	.	PUNCT
ejpam-1676	118	1	,	,	PUNCT
ejpam-1676	118	2	ym;θ1	ym;θ1	NOUN
ejpam-1676	118	3	)	)	PUNCT
ejpam-1676	118	4	∂	∂	NUM
ejpam-1676	118	5	θ1	θ1	NOUN
ejpam-1676	118	6	=	=	SYM
ejpam-1676	118	7	log	log	VERB
ejpam-1676	118	8	l1(y1;θ1	l1(y1;θ1	ADJ
ejpam-1676	118	9	)	)	PUNCT
ejpam-1676	118	10	∂	∂	NUM
ejpam-1676	118	11	θ1	θ1	NOUN
ejpam-1676	118	12	+	+	CCONJ
ejpam-1676	118	13	m∑	m∑	CCONJ
ejpam-1676	118	14	i=2	i=2	PROPN
ejpam-1676	118	15	log	log	VERB
ejpam-1676	118	16	li(y	li(y	X
ejpam-1676	118	17	i;θi	i;θi	PROPN
ejpam-1676	118	18	)	)	PUNCT
ejpam-1676	118	19	∂	∂	PROPN
ejpam-1676	118	20	θi	θi	PROPN
ejpam-1676	118	21	gi(θ1	gi(θ1	PROPN
ejpam-1676	118	22	)	)	PUNCT
ejpam-1676	118	23	∂	∂	NUM
ejpam-1676	118	24	θ1	θ1	NOUN
ejpam-1676	118	25	=	=	SYM
ejpam-1676	118	26	log	log	VERB
ejpam-1676	118	27	l1(y1;θ1	l1(y1;θ1	ADJ
ejpam-1676	118	28	)	)	PUNCT
ejpam-1676	118	29	∂	∂	NUM
ejpam-1676	118	30	θ1	θ1	NOUN
ejpam-1676	118	31	+	+	CCONJ
ejpam-1676	118	32	m∑	m∑	ADV
ejpam-1676	118	33	i=2	i=2	PROPN
ejpam-1676	118	34	wi(θ1	wi(θ1	PROPN
ejpam-1676	118	35	)	)	PUNCT
ejpam-1676	118	36	log	log	VERB
ejpam-1676	118	37	li(y	li(y	X
ejpam-1676	118	38	i;θi	i;θi	PROPN
ejpam-1676	118	39	)	)	PUNCT
ejpam-1676	118	40	∂	∂	NUM
ejpam-1676	118	41	θi	θi	NOUN
ejpam-1676	118	42	.	.	PUNCT
ejpam-1676	119	1	where	where	SCONJ
ejpam-1676	119	2	wi(θ1	wi(θ1	X
ejpam-1676	119	3	)	)	PUNCT
ejpam-1676	119	4	=	=	SYM
ejpam-1676	119	5	∂	∂	NUM
ejpam-1676	119	6	gi(θ1)/∂	gi(θ1)/∂	NOUN
ejpam-1676	119	7	θ1	θ1	NOUN
ejpam-1676	119	8	,	,	PUNCT
ejpam-1676	119	9	i	i	NOUN
ejpam-1676	119	10	=	=	NOUN
ejpam-1676	119	11	2	2	NUM
ejpam-1676	119	12	,	,	PUNCT
ejpam-1676	119	13	·	·	PUNCT
ejpam-1676	119	14	·	·	PUNCT
ejpam-1676	119	15	·	·	PUNCT
ejpam-1676	119	16	,	,	PUNCT
ejpam-1676	119	17	m.	m.	NOUN
ejpam-1676	119	18	by	by	ADP
ejpam-1676	119	19	setting	set	VERB
ejpam-1676	119	20	the	the	DET
ejpam-1676	119	21	above	above	ADJ
ejpam-1676	119	22	function	function	NOUN
ejpam-1676	119	23	to	to	PART
ejpam-1676	119	24	be	be	AUX
ejpam-1676	119	25	zero	zero	NUM
ejpam-1676	119	26	,	,	PUNCT
ejpam-1676	119	27	we	we	PRON
ejpam-1676	119	28	obtain	obtain	VERB
ejpam-1676	119	29	an	an	DET
ejpam-1676	119	30	estimating	estimate	VERB
ejpam-1676	119	31	equation	equation	NOUN
ejpam-1676	119	32	for	for	ADP
ejpam-1676	119	33	θ1	θ1	NOUN
ejpam-1676	119	34	which	which	PRON
ejpam-1676	119	35	is	be	AUX
ejpam-1676	119	36	based	base	VERB
ejpam-1676	119	37	on	on	ADP
ejpam-1676	119	38	all	all	DET
ejpam-1676	119	39	available	available	ADJ
ejpam-1676	119	40	data	datum	NOUN
ejpam-1676	119	41	.	.	PUNCT
ejpam-1676	120	1	in	in	ADP
ejpam-1676	120	2	practice	practice	NOUN
ejpam-1676	120	3	,	,	PUNCT
ejpam-1676	120	4	however	however	ADV
ejpam-1676	120	5	,	,	PUNCT
ejpam-1676	120	6	the	the	DET
ejpam-1676	120	7	functional	functional	ADJ
ejpam-1676	120	8	form	form	NOUN
ejpam-1676	120	9	of	of	ADP
ejpam-1676	120	10	gi	gi	PROPN
ejpam-1676	120	11	’s	’	VERB
ejpam-1676	120	12	are	be	AUX
ejpam-1676	120	13	not	not	PART
ejpam-1676	120	14	known	know	VERB
ejpam-1676	120	15	.	.	PUNCT
ejpam-1676	121	1	therefore	therefore	ADV
ejpam-1676	121	2	,	,	PUNCT
ejpam-1676	121	3	the	the	DET
ejpam-1676	121	4	analytical	analytical	ADJ
ejpam-1676	121	5	forms	form	NOUN
ejpam-1676	121	6	of	of	ADP
ejpam-1676	121	7	gi	gi	NOUN
ejpam-1676	121	8	’s	’	VERB
ejpam-1676	121	9	are	be	AUX
ejpam-1676	121	10	simply	simply	ADV
ejpam-1676	121	11	not	not	PART
ejpam-1676	121	12	available	available	ADJ
ejpam-1676	121	13	.	.	PUNCT
ejpam-1676	122	1	when	when	SCONJ
ejpam-1676	122	2	facing	face	VERB
ejpam-1676	122	3	such	such	DET
ejpam-1676	122	4	a	a	DET
ejpam-1676	122	5	difficulty	difficulty	NOUN
ejpam-1676	122	6	,	,	PUNCT
ejpam-1676	122	7	one	one	PRON
ejpam-1676	122	8	could	could	AUX
ejpam-1676	122	9	abandon	abandon	VERB
ejpam-1676	122	10	the	the	DET
ejpam-1676	122	11	data	datum	NOUN
ejpam-1676	122	12	fusion	fusion	NOUN
ejpam-1676	122	13	process	process	NOUN
ejpam-1676	122	14	and	and	CCONJ
ejpam-1676	122	15	proceed	proceed	VERB
ejpam-1676	122	16	with	with	ADP
ejpam-1676	122	17	l1	l1	PROPN
ejpam-1676	122	18	by	by	ADP
ejpam-1676	122	19	severing	sever	VERB
ejpam-1676	122	20	the	the	DET
ejpam-1676	122	21	connections	connection	NOUN
ejpam-1676	122	22	among	among	ADP
ejpam-1676	122	23	all	all	DET
ejpam-1676	122	24	parameters	parameter	NOUN
ejpam-1676	122	25	.	.	PUNCT
ejpam-1676	123	1	although	although	SCONJ
ejpam-1676	123	2	this	this	PRON
ejpam-1676	123	3	would	would	AUX
ejpam-1676	123	4	avoid	avoid	VERB
ejpam-1676	123	5	any	any	DET
ejpam-1676	123	6	possible	possible	ADJ
ejpam-1676	123	7	bias	bias	NOUN
ejpam-1676	123	8	,	,	PUNCT
ejpam-1676	123	9	this	this	DET
ejpam-1676	123	10	approach	approach	NOUN
ejpam-1676	123	11	would	would	AUX
ejpam-1676	123	12	clearly	clearly	ADV
ejpam-1676	123	13	lose	lose	VERB
ejpam-1676	123	14	some	some	DET
ejpam-1676	123	15	information	information	NOUN
ejpam-1676	123	16	.	.	PUNCT
ejpam-1676	124	1	the	the	DET
ejpam-1676	124	2	loss	loss	NOUN
ejpam-1676	124	3	of	of	ADP
ejpam-1676	124	4	information	information	NOUN
ejpam-1676	124	5	should	should	AUX
ejpam-1676	124	6	not	not	PART
ejpam-1676	124	7	be	be	AUX
ejpam-1676	124	8	of	of	ADP
ejpam-1676	124	9	great	great	ADJ
ejpam-1676	124	10	concern	concern	NOUN
ejpam-1676	124	11	if	if	SCONJ
ejpam-1676	124	12	the	the	DET
ejpam-1676	124	13	sample	sample	NOUN
ejpam-1676	124	14	size	size	NOUN
ejpam-1676	124	15	from	from	ADP
ejpam-1676	124	16	the	the	DET
ejpam-1676	124	17	target	target	NOUN
ejpam-1676	124	18	population	population	NOUN
ejpam-1676	124	19	is	be	AUX
ejpam-1676	124	20	large	large	ADJ
ejpam-1676	124	21	enough	enough	ADV
ejpam-1676	124	22	to	to	PART
ejpam-1676	124	23	ensure	ensure	VERB
ejpam-1676	124	24	a	a	DET
ejpam-1676	124	25	valid	valid	ADJ
ejpam-1676	124	26	and	and	CCONJ
ejpam-1676	124	27	effective	effective	ADJ
ejpam-1676	124	28	inference	inference	NOUN
ejpam-1676	124	29	.	.	PUNCT
ejpam-1676	125	1	the	the	DET
ejpam-1676	125	2	benefit	benefit	NOUN
ejpam-1676	125	3	of	of	ADP
ejpam-1676	125	4	integrating	integrate	VERB
ejpam-1676	125	5	information	information	NOUN
ejpam-1676	125	6	from	from	ADP
ejpam-1676	125	7	other	other	ADJ
ejpam-1676	125	8	sources	source	NOUN
ejpam-1676	125	9	is	be	AUX
ejpam-1676	125	10	clearly	clearly	ADV
ejpam-1676	125	11	negligible	negligible	ADJ
ejpam-1676	125	12	.	.	PUNCT
ejpam-1676	126	1	when	when	SCONJ
ejpam-1676	126	2	the	the	DET
ejpam-1676	126	3	sample	sample	NOUN
ejpam-1676	126	4	from	from	ADP
ejpam-1676	126	5	the	the	DET
ejpam-1676	126	6	target	target	NOUN
ejpam-1676	126	7	population	population	NOUN
ejpam-1676	126	8	is	be	AUX
ejpam-1676	126	9	very	very	ADV
ejpam-1676	126	10	small	small	ADJ
ejpam-1676	126	11	or	or	CCONJ
ejpam-1676	126	12	moderate	moderate	ADJ
ejpam-1676	126	13	,	,	PUNCT
ejpam-1676	126	14	the	the	DET
ejpam-1676	126	15	related	related	ADJ
ejpam-1676	126	16	data	data	NOUN
ejpam-1676	126	17	sets	set	NOUN
ejpam-1676	126	18	discarded	discard	VERB
ejpam-1676	126	19	as	as	ADP
ejpam-1676	126	20	the	the	DET
ejpam-1676	126	21	result	result	NOUN
ejpam-1676	126	22	of	of	ADP
ejpam-1676	126	23	avoiding	avoid	VERB
ejpam-1676	126	24	bias	bias	NOUN
ejpam-1676	126	25	could	could	AUX
ejpam-1676	126	26	contain	contain	VERB
ejpam-1676	126	27	very	very	ADV
ejpam-1676	126	28	valuable	valuable	ADJ
ejpam-1676	126	29	information	information	NOUN
ejpam-1676	126	30	.	.	PUNCT
ejpam-1676	127	1	p.	p.	NOUN
ejpam-1676	127	2	guo	guo	PROPN
ejpam-1676	127	3	,	,	PUNCT
ejpam-1676	127	4	x.	x.	PROPN
ejpam-1676	127	5	wang	wang	PROPN
ejpam-1676	127	6	,	,	PUNCT
ejpam-1676	127	7	y.	y.	PROPN
ejpam-1676	127	8	wu	wu	PROPN
ejpam-1676	127	9	/	/	SYM
ejpam-1676	127	10	eur	eur	PROPN
ejpam-1676	127	11	.	.	PUNCT
ejpam-1676	128	1	j.	j.	PROPN
ejpam-1676	128	2	pure	pure	PROPN
ejpam-1676	128	3	appl	appl	PROPN
ejpam-1676	128	4	.	.	PROPN
ejpam-1676	128	5	math	math	PROPN
ejpam-1676	128	6	,	,	PUNCT
ejpam-1676	128	7	5	5	NUM
ejpam-1676	128	8	(	(	PUNCT
ejpam-1676	128	9	2012	2012	NUM
ejpam-1676	128	10	)	)	PUNCT
ejpam-1676	128	11	,	,	PUNCT
ejpam-1676	128	12	333	333	NUM
ejpam-1676	128	13	-	-	SYM
ejpam-1676	128	14	356	356	NUM
ejpam-1676	128	15	337	337	NUM
ejpam-1676	128	16	the	the	DET
ejpam-1676	128	17	alternative	alternative	NOUN
ejpam-1676	128	18	is	be	AUX
ejpam-1676	128	19	try	try	VERB
ejpam-1676	128	20	to	to	PART
ejpam-1676	128	21	incorporate	incorporate	VERB
ejpam-1676	128	22	information	information	NOUN
ejpam-1676	128	23	from	from	ADP
ejpam-1676	128	24	other	other	ADJ
ejpam-1676	128	25	sources	source	NOUN
ejpam-1676	128	26	in	in	ADP
ejpam-1676	128	27	a	a	DET
ejpam-1676	128	28	meaningful	meaningful	ADJ
ejpam-1676	128	29	way	way	NOUN
ejpam-1676	128	30	.	.	PUNCT
ejpam-1676	129	1	this	this	PRON
ejpam-1676	129	2	must	must	AUX
ejpam-1676	129	3	be	be	AUX
ejpam-1676	129	4	done	do	VERB
ejpam-1676	129	5	in	in	ADP
ejpam-1676	129	6	a	a	DET
ejpam-1676	129	7	reasonable	reasonable	ADJ
ejpam-1676	129	8	and	and	CCONJ
ejpam-1676	129	9	careful	careful	ADJ
ejpam-1676	129	10	fashion	fashion	NOUN
ejpam-1676	129	11	without	without	ADP
ejpam-1676	129	12	the	the	DET
ejpam-1676	129	13	knowledge	knowledge	NOUN
ejpam-1676	129	14	of	of	ADP
ejpam-1676	129	15	the	the	DET
ejpam-1676	129	16	link	link	NOUN
ejpam-1676	129	17	functions	function	NOUN
ejpam-1676	129	18	gi	gi	ADP
ejpam-1676	129	19	,	,	PUNCT
ejpam-1676	129	20	i	i	NOUN
ejpam-1676	129	21	=	=	NOUN
ejpam-1676	129	22	2	2	NUM
ejpam-1676	129	23	,	,	PUNCT
ejpam-1676	129	24	·	·	PUNCT
ejpam-1676	129	25	·	·	PUNCT
ejpam-1676	129	26	·	·	PUNCT
ejpam-1676	129	27	,	,	PUNCT
ejpam-1676	129	28	m.	m.	NOUN
ejpam-1676	129	29	first	first	ADV
ejpam-1676	129	30	,	,	PUNCT
ejpam-1676	129	31	we	we	PRON
ejpam-1676	129	32	must	must	AUX
ejpam-1676	129	33	adopt	adopt	VERB
ejpam-1676	129	34	a	a	DET
ejpam-1676	129	35	meaningful	meaningful	ADJ
ejpam-1676	129	36	measure	measure	NOUN
ejpam-1676	129	37	to	to	PART
ejpam-1676	129	38	evaluate	evaluate	VERB
ejpam-1676	129	39	the	the	DET
ejpam-1676	129	40	discrepancy	discrepancy	NOUN
ejpam-1676	129	41	among	among	ADP
ejpam-1676	129	42	the	the	DET
ejpam-1676	129	43	probability	probability	NOUN
ejpam-1676	129	44	density	density	NOUN
ejpam-1676	129	45	distributions	distribution	NOUN
ejpam-1676	129	46	when	when	SCONJ
ejpam-1676	129	47	the	the	DET
ejpam-1676	129	48	parameters	parameter	NOUN
ejpam-1676	129	49	are	be	AUX
ejpam-1676	129	50	known	know	VERB
ejpam-1676	129	51	.	.	PUNCT
ejpam-1676	130	1	second	second	ADV
ejpam-1676	130	2	,	,	PUNCT
ejpam-1676	130	3	we	we	PRON
ejpam-1676	130	4	must	must	AUX
ejpam-1676	130	5	tackle	tackle	VERB
ejpam-1676	130	6	the	the	DET
ejpam-1676	130	7	problem	problem	NOUN
ejpam-1676	130	8	without	without	ADP
ejpam-1676	130	9	the	the	DET
ejpam-1676	130	10	knowledge	knowledge	NOUN
ejpam-1676	130	11	of	of	ADP
ejpam-1676	130	12	the	the	DET
ejpam-1676	130	13	true	true	ADJ
ejpam-1676	130	14	values	value	NOUN
ejpam-1676	130	15	of	of	ADP
ejpam-1676	130	16	the	the	DET
ejpam-1676	130	17	parameters	parameter	NOUN
ejpam-1676	130	18	.	.	PUNCT
ejpam-1676	131	1	to	to	PART
ejpam-1676	131	2	resolve	resolve	VERB
ejpam-1676	131	3	the	the	DET
ejpam-1676	131	4	first	first	ADJ
ejpam-1676	131	5	challenge	challenge	NOUN
ejpam-1676	131	6	,	,	PUNCT
ejpam-1676	131	7	we	we	PRON
ejpam-1676	131	8	propose	propose	VERB
ejpam-1676	131	9	to	to	PART
ejpam-1676	131	10	use	use	VERB
ejpam-1676	131	11	some	some	DET
ejpam-1676	131	12	general	general	ADJ
ejpam-1676	131	13	measure	measure	NOUN
ejpam-1676	131	14	for	for	ADP
ejpam-1676	131	15	information	information	NOUN
ejpam-1676	131	16	discrepancy	discrepancy	NOUN
ejpam-1676	131	17	to	to	PART
ejpam-1676	131	18	characterize	characterize	VERB
ejpam-1676	131	19	the	the	DET
ejpam-1676	131	20	connections	connection	NOUN
ejpam-1676	131	21	among	among	ADP
ejpam-1676	131	22	all	all	DET
ejpam-1676	131	23	parameters	parameter	NOUN
ejpam-1676	131	24	.	.	PUNCT
ejpam-1676	132	1	entropy	entropy	PROPN
ejpam-1676	132	2	or	or	CCONJ
ejpam-1676	132	3	relative	relative	ADJ
ejpam-1676	132	4	entropy	entropy	NOUN
ejpam-1676	132	5	has	have	AUX
ejpam-1676	132	6	been	be	AUX
ejpam-1676	132	7	widely	widely	ADV
ejpam-1676	132	8	used	use	VERB
ejpam-1676	132	9	in	in	ADP
ejpam-1676	132	10	information	information	NOUN
ejpam-1676	132	11	theory	theory	NOUN
ejpam-1676	132	12	.	.	PUNCT
ejpam-1676	133	1	for	for	ADP
ejpam-1676	133	2	density	density	NOUN
ejpam-1676	133	3	functions	function	NOUN
ejpam-1676	133	4	,	,	PUNCT
ejpam-1676	133	5	g1(x	g1(x	NOUN
ejpam-1676	133	6	)	)	PUNCT
ejpam-1676	133	7	and	and	CCONJ
ejpam-1676	133	8	g2(x	g2(x	PROPN
ejpam-1676	133	9	)	)	PUNCT
ejpam-1676	133	10	,	,	PUNCT
ejpam-1676	133	11	with	with	ADP
ejpam-1676	133	12	respect	respect	NOUN
ejpam-1676	133	13	to	to	ADP
ejpam-1676	133	14	a	a	DET
ejpam-1676	133	15	σ−finite	σ−finite	PROPN
ejpam-1676	133	16	measure	measure	NOUN
ejpam-1676	133	17	ν	ν	NOUN
ejpam-1676	133	18	,	,	PUNCT
ejpam-1676	133	19	the	the	DET
ejpam-1676	133	20	relative	relative	ADJ
ejpam-1676	133	21	entropy	entropy	NOUN
ejpam-1676	133	22	,	,	PUNCT
ejpam-1676	133	23	also	also	ADV
ejpam-1676	133	24	called	call	VERB
ejpam-1676	133	25	kullback	kullback	NOUN
ejpam-1676	133	26	-	-	PUNCT
ejpam-1676	133	27	leibler	leibler	NOUN
ejpam-1676	133	28	divergence	divergence	NOUN
ejpam-1676	133	29	,	,	PUNCT
ejpam-1676	133	30	is	be	AUX
ejpam-1676	133	31	defined	define	VERB
ejpam-1676	133	32	as	as	ADP
ejpam-1676	133	33	:	:	PUNCT
ejpam-1676	133	34	k	k	X
ejpam-1676	133	35	l(g1	l(g1	ADV
ejpam-1676	133	36	,	,	PUNCT
ejpam-1676	133	37	g2	g2	PROPN
ejpam-1676	133	38	)	)	PUNCT
ejpam-1676	133	39	=	=	SYM
ejpam-1676	133	40	e1	e1	PROPN
ejpam-1676	133	41	�	�	PROPN
ejpam-1676	133	42	log	log	NOUN
ejpam-1676	133	43	g1(x	g1(x	NOUN
ejpam-1676	133	44	)	)	PUNCT
ejpam-1676	133	45	g2(x	g2(x	NOUN
ejpam-1676	133	46	)	)	PUNCT
ejpam-1676	133	47	�	�	PROPN
ejpam-1676	133	48	=	=	SYM
ejpam-1676	133	49	∫	∫	PROPN
ejpam-1676	133	50	log	log	PROPN
ejpam-1676	133	51	g1(x	g1(x	NOUN
ejpam-1676	133	52	)	)	PUNCT
ejpam-1676	133	53	g2(x	g2(x	NOUN
ejpam-1676	133	54	)	)	PUNCT
ejpam-1676	134	1	g1(x)dν(x	g1(x)dν(x	NOUN
ejpam-1676	134	2	)	)	PUNCT
ejpam-1676	134	3	.	.	PUNCT
ejpam-1676	135	1	(	(	PUNCT
ejpam-1676	135	2	1	1	X
ejpam-1676	135	3	)	)	PUNCT
ejpam-1676	135	4	in	in	ADP
ejpam-1676	135	5	this	this	DET
ejpam-1676	135	6	expression	expression	NOUN
ejpam-1676	135	7	,	,	PUNCT
ejpam-1676	135	8	log(g1(x)/g2(x	log(g1(x)/g2(x	PROPN
ejpam-1676	135	9	)	)	PUNCT
ejpam-1676	135	10	)	)	PUNCT
ejpam-1676	135	11	is	be	AUX
ejpam-1676	135	12	defined	define	VERB
ejpam-1676	135	13	as	as	ADP
ejpam-1676	135	14	+	+	NOUN
ejpam-1676	135	15	∞	∞	PROPN
ejpam-1676	135	16	,	,	PUNCT
ejpam-1676	135	17	if	if	SCONJ
ejpam-1676	135	18	g1(x	g1(x	NOUN
ejpam-1676	135	19	)	)	PUNCT
ejpam-1676	135	20	>	>	X
ejpam-1676	135	21	0	0	PUNCT
ejpam-1676	135	22	and	and	CCONJ
ejpam-1676	135	23	g2(x	g2(x	PROPN
ejpam-1676	135	24	)	)	PUNCT
ejpam-1676	136	1	=	=	SYM
ejpam-1676	136	2	0	0	X
ejpam-1676	136	3	.	.	PUNCT
ejpam-1676	137	1	therefore	therefore	ADV
ejpam-1676	137	2	the	the	DET
ejpam-1676	137	3	expectation	expectation	NOUN
ejpam-1676	137	4	could	could	AUX
ejpam-1676	137	5	be	be	AUX
ejpam-1676	137	6	+	+	ADV
ejpam-1676	137	7	∞.	∞.	PROPN
ejpam-1676	137	8	although	although	SCONJ
ejpam-1676	137	9	log(g1(x)/g2(x	log(g1(x)/g2(x	PROPN
ejpam-1676	137	10	)	)	PUNCT
ejpam-1676	137	11	)	)	PUNCT
ejpam-1676	137	12	is	be	AUX
ejpam-1676	137	13	defined	define	VERB
ejpam-1676	137	14	as	as	ADP
ejpam-1676	137	15	−∞	−∞	ADP
ejpam-1676	137	16	when	when	SCONJ
ejpam-1676	137	17	g1(x	g1(x	NOUN
ejpam-1676	137	18	)	)	PUNCT
ejpam-1676	137	19	=	=	SYM
ejpam-1676	137	20	0	0	NUM
ejpam-1676	137	21	and	and	CCONJ
ejpam-1676	137	22	g2(x	g2(x	PROPN
ejpam-1676	137	23	)	)	PUNCT
ejpam-1676	137	24	>	>	X
ejpam-1676	137	25	0	0	NUM
ejpam-1676	137	26	,	,	PUNCT
ejpam-1676	137	27	the	the	DET
ejpam-1676	137	28	integrand	integrand	NOUN
ejpam-1676	137	29	,	,	PUNCT
ejpam-1676	137	30	log(g1(x)/g2(x))g1(x	log(g1(x)/g2(x))g1(x	ADJ
ejpam-1676	137	31	)	)	PUNCT
ejpam-1676	137	32	is	be	AUX
ejpam-1676	137	33	defined	define	VERB
ejpam-1676	137	34	as	as	ADP
ejpam-1676	137	35	zero	zero	NUM
ejpam-1676	137	36	in	in	ADP
ejpam-1676	137	37	this	this	DET
ejpam-1676	137	38	case	case	NOUN
ejpam-1676	137	39	.	.	PUNCT
ejpam-1676	138	1	it	it	PRON
ejpam-1676	138	2	is	be	AUX
ejpam-1676	138	3	widely	widely	ADV
ejpam-1676	138	4	used	use	VERB
ejpam-1676	138	5	in	in	ADP
ejpam-1676	138	6	information	information	NOUN
ejpam-1676	138	7	theory	theory	NOUN
ejpam-1676	138	8	.	.	PUNCT
ejpam-1676	139	1	detailed	detailed	ADJ
ejpam-1676	139	2	discussions	discussion	NOUN
ejpam-1676	139	3	and	and	CCONJ
ejpam-1676	139	4	application	application	NOUN
ejpam-1676	139	5	of	of	ADP
ejpam-1676	139	6	the	the	DET
ejpam-1676	139	7	entropy	entropy	NOUN
ejpam-1676	139	8	in	in	ADP
ejpam-1676	139	9	information	information	NOUN
ejpam-1676	139	10	theory	theory	NOUN
ejpam-1676	139	11	can	can	AUX
ejpam-1676	139	12	be	be	AUX
ejpam-1676	139	13	found	find	VERB
ejpam-1676	139	14	in	in	ADP
ejpam-1676	139	15	[	[	X
ejpam-1676	139	16	19	19	NUM
ejpam-1676	139	17	]	]	PUNCT
ejpam-1676	139	18	.	.	PUNCT
ejpam-1676	140	1	some	some	DET
ejpam-1676	140	2	theoretical	theoretical	ADJ
ejpam-1676	140	3	properties	property	NOUN
ejpam-1676	140	4	of	of	ADP
ejpam-1676	140	5	the	the	DET
ejpam-1676	140	6	entropy	entropy	NOUN
ejpam-1676	140	7	can	can	AUX
ejpam-1676	140	8	be	be	AUX
ejpam-1676	140	9	found	find	VERB
ejpam-1676	140	10	in	in	ADP
ejpam-1676	140	11	[	[	X
ejpam-1676	140	12	14	14	NUM
ejpam-1676	140	13	]	]	PUNCT
ejpam-1676	140	14	.	.	PUNCT
ejpam-1676	141	1	in	in	ADP
ejpam-1676	141	2	particular	particular	ADJ
ejpam-1676	141	3	,	,	PUNCT
ejpam-1676	141	4	the	the	DET
ejpam-1676	141	5	relative	relative	ADJ
ejpam-1676	141	6	entropy	entropy	NOUN
ejpam-1676	141	7	is	be	AUX
ejpam-1676	141	8	not	not	PART
ejpam-1676	141	9	symmetric	symmetric	ADJ
ejpam-1676	141	10	and	and	CCONJ
ejpam-1676	141	11	therefore	therefore	ADV
ejpam-1676	141	12	not	not	PART
ejpam-1676	141	13	a	a	DET
ejpam-1676	141	14	distance	distance	NOUN
ejpam-1676	141	15	.	.	PUNCT
ejpam-1676	142	1	the	the	DET
ejpam-1676	142	2	relative	relative	ADJ
ejpam-1676	142	3	entropy	entropy	NOUN
ejpam-1676	142	4	,	,	PUNCT
ejpam-1676	142	5	also	also	ADV
ejpam-1676	142	6	known	know	VERB
ejpam-1676	142	7	as	as	ADP
ejpam-1676	142	8	kullback	kullback	NOUN
ejpam-1676	142	9	-	-	PUNCT
ejpam-1676	142	10	leibler	leibler	NOUN
ejpam-1676	142	11	divergence	divergence	NOUN
ejpam-1676	142	12	,	,	PUNCT
ejpam-1676	142	13	is	be	AUX
ejpam-1676	142	14	also	also	ADV
ejpam-1676	142	15	called	call	VERB
ejpam-1676	142	16	the	the	DET
ejpam-1676	142	17	entropy	entropy	NOUN
ejpam-1676	142	18	loss	loss	NOUN
ejpam-1676	142	19	.	.	PUNCT
ejpam-1676	143	1	[	[	X
ejpam-1676	143	2	20	20	NUM
ejpam-1676	143	3	]	]	PUNCT
ejpam-1676	143	4	introduce	introduce	VERB
ejpam-1676	143	5	it	it	PRON
ejpam-1676	143	6	as	as	ADP
ejpam-1676	143	7	a	a	DET
ejpam-1676	143	8	performance	performance	NOUN
ejpam-1676	143	9	criterion	criterion	NOUN
ejpam-1676	143	10	in	in	ADP
ejpam-1676	143	11	estimating	estimate	VERB
ejpam-1676	143	12	the	the	DET
ejpam-1676	143	13	multinormal	multinormal	ADJ
ejpam-1676	143	14	variance	variance	NOUN
ejpam-1676	143	15	-	-	PUNCT
ejpam-1676	143	16	covariance	covariance	NOUN
ejpam-1676	143	17	matrix	matrix	NOUN
ejpam-1676	143	18	.	.	PUNCT
ejpam-1676	144	1	[	[	X
ejpam-1676	144	2	15	15	NUM
ejpam-1676	144	3	]	]	PUNCT
ejpam-1676	144	4	shows	show	VERB
ejpam-1676	144	5	the	the	DET
ejpam-1676	144	6	entropy	entropy	NOUN
ejpam-1676	144	7	is	be	AUX
ejpam-1676	144	8	a	a	DET
ejpam-1676	144	9	loss	loss	NOUN
ejpam-1676	144	10	function	function	NOUN
ejpam-1676	144	11	in	in	ADP
ejpam-1676	144	12	a	a	DET
ejpam-1676	144	13	bayesian	bayesian	NOUN
ejpam-1676	144	14	framework	framework	NOUN
ejpam-1676	144	15	.	.	PUNCT
ejpam-1676	145	1	[	[	X
ejpam-1676	145	2	18	18	NUM
ejpam-1676	145	3	]	]	PUNCT
ejpam-1676	145	4	provides	provide	VERB
ejpam-1676	145	5	detailed	detailed	ADJ
ejpam-1676	145	6	discussions	discussion	NOUN
ejpam-1676	145	7	on	on	ADP
ejpam-1676	145	8	the	the	DET
ejpam-1676	145	9	relative	relative	ADJ
ejpam-1676	145	10	entropy	entropy	NOUN
ejpam-1676	145	11	including	include	VERB
ejpam-1676	145	12	the	the	DET
ejpam-1676	145	13	connection	connection	NOUN
ejpam-1676	145	14	between	between	ADP
ejpam-1676	145	15	fisher	fisher	PROPN
ejpam-1676	145	16	information	information	NOUN
ejpam-1676	145	17	and	and	CCONJ
ejpam-1676	145	18	relative	relative	ADJ
ejpam-1676	145	19	entropy	entropy	NOUN
ejpam-1676	145	20	.	.	PUNCT
ejpam-1676	146	1	[	[	X
ejpam-1676	146	2	13	13	NUM
ejpam-1676	146	3	]	]	PUNCT
ejpam-1676	146	4	has	have	AUX
ejpam-1676	146	5	shown	show	VERB
ejpam-1676	146	6	that	that	SCONJ
ejpam-1676	146	7	the	the	DET
ejpam-1676	146	8	classical	classical	ADJ
ejpam-1676	146	9	maximum	maximum	ADJ
ejpam-1676	146	10	likelihood	likelihood	NOUN
ejpam-1676	146	11	principle	principle	NOUN
ejpam-1676	146	12	can	can	AUX
ejpam-1676	146	13	be	be	AUX
ejpam-1676	146	14	considered	consider	VERB
ejpam-1676	146	15	a	a	DET
ejpam-1676	146	16	method	method	NOUN
ejpam-1676	146	17	of	of	ADP
ejpam-1676	146	18	asymptotic	asymptotic	ADJ
ejpam-1676	146	19	realization	realization	NOUN
ejpam-1676	146	20	of	of	ADP
ejpam-1676	146	21	an	an	DET
ejpam-1676	146	22	optimum	optimum	ADJ
ejpam-1676	146	23	estimate	estimate	NOUN
ejpam-1676	146	24	with	with	ADP
ejpam-1676	146	25	respect	respect	NOUN
ejpam-1676	146	26	to	to	ADP
ejpam-1676	146	27	the	the	DET
ejpam-1676	146	28	relative	relative	ADJ
ejpam-1676	146	29	entropy	entropy	PROPN
ejpam-1676	146	30	.	.	PUNCT
ejpam-1676	147	1	since	since	SCONJ
ejpam-1676	147	2	the	the	DET
ejpam-1676	147	3	metric	metric	NOUN
ejpam-1676	147	4	of	of	ADP
ejpam-1676	147	5	the	the	DET
ejpam-1676	147	6	manifold	manifold	NOUN
ejpam-1676	147	7	that	that	PRON
ejpam-1676	147	8	connects	connect	VERB
ejpam-1676	147	9	all	all	DET
ejpam-1676	147	10	the	the	DET
ejpam-1676	147	11	parameters	parameter	NOUN
ejpam-1676	147	12	can	can	AUX
ejpam-1676	147	13	not	not	PART
ejpam-1676	147	14	be	be	AUX
ejpam-1676	147	15	specified	specify	VERB
ejpam-1676	147	16	,	,	PUNCT
ejpam-1676	147	17	we	we	PRON
ejpam-1676	147	18	now	now	ADV
ejpam-1676	147	19	define	define	VERB
ejpam-1676	147	20	the	the	DET
ejpam-1676	147	21	neighborhood	neighborhood	NOUN
ejpam-1676	147	22	enveloping	envelop	VERB
ejpam-1676	147	23	by	by	ADP
ejpam-1676	147	24	[	[	X
ejpam-1676	147	25	17	17	NUM
ejpam-1676	147	26	]	]	PUNCT
ejpam-1676	147	27	.	.	PUNCT
ejpam-1676	148	1	for	for	ADP
ejpam-1676	148	2	any	any	DET
ejpam-1676	148	3	fixed	fix	VERB
ejpam-1676	148	4	value	value	NOUN
ejpam-1676	148	5	of	of	ADP
ejpam-1676	148	6	θ1	θ1	NOUN
ejpam-1676	148	7	,	,	PUNCT
ejpam-1676	148	8	the	the	DET
ejpam-1676	148	9	parameter	parameter	NOUN
ejpam-1676	148	10	of	of	ADP
ejpam-1676	148	11	the	the	DET
ejpam-1676	148	12	population	population	NOUN
ejpam-1676	148	13	of	of	ADP
ejpam-1676	148	14	inferential	inferential	ADJ
ejpam-1676	148	15	interest	interest	NOUN
ejpam-1676	148	16	,	,	PUNCT
ejpam-1676	148	17	we	we	PRON
ejpam-1676	148	18	can	can	AUX
ejpam-1676	148	19	define	define	VERB
ejpam-1676	148	20	a	a	DET
ejpam-1676	148	21	neighborhood	neighborhood	NOUN
ejpam-1676	148	22	by	by	ADP
ejpam-1676	148	23	using	use	VERB
ejpam-1676	148	24	the	the	DET
ejpam-1676	148	25	relative	relative	ADJ
ejpam-1676	148	26	entropy	entropy	NOUN
ejpam-1676	148	27	as	as	ADP
ejpam-1676	148	28	n	n	X
ejpam-1676	148	29	f1	f1	NOUN
ejpam-1676	148	30	(	(	PUNCT
ejpam-1676	148	31	ε	ε	PROPN
ejpam-1676	148	32	)	)	PUNCT
ejpam-1676	148	33	=	=	SYM
ejpam-1676	148	34	∪θ∈θ	∪θ∈θ	NOUN
ejpam-1676	148	35	{	{	PUNCT
ejpam-1676	148	36	f	f	PROPN
ejpam-1676	148	37	(	(	PUNCT
ejpam-1676	148	38	θ	θ	PROPN
ejpam-1676	148	39	)	)	PUNCT
ejpam-1676	148	40	:	:	PUNCT
ejpam-1676	149	1	i	i	PRON
ejpam-1676	149	2	(	(	PUNCT
ejpam-1676	149	3	f	f	PROPN
ejpam-1676	149	4	(	(	PUNCT
ejpam-1676	149	5	θ	θ	NOUN
ejpam-1676	149	6	)	)	PUNCT
ejpam-1676	149	7	,	,	PUNCT
ejpam-1676	149	8	f1)≤	f1)≤	PROPN
ejpam-1676	149	9	ε	ε	PROPN
ejpam-1676	149	10	}	}	PUNCT
ejpam-1676	149	11	,	,	PUNCT
ejpam-1676	149	12	(	(	PUNCT
ejpam-1676	149	13	2	2	X
ejpam-1676	149	14	)	)	PUNCT
ejpam-1676	149	15	where	where	SCONJ
ejpam-1676	149	16	f1	f1	NOUN
ejpam-1676	149	17	=	=	SYM
ejpam-1676	149	18	f	f	PROPN
ejpam-1676	149	19	(	(	PUNCT
ejpam-1676	149	20	x	x	INTJ
ejpam-1676	149	21	,	,	PUNCT
ejpam-1676	149	22	θ1	θ1	NOUN
ejpam-1676	149	23	)	)	PUNCT
ejpam-1676	149	24	and	and	CCONJ
ejpam-1676	149	25	ε≥	ε≥	PROPN
ejpam-1676	149	26	0	0	PROPN
ejpam-1676	149	27	.	.	PUNCT
ejpam-1676	149	28	assume	assume	VERB
ejpam-1676	149	29	that	that	SCONJ
ejpam-1676	149	30	the	the	DET
ejpam-1676	149	31	density	density	NOUN
ejpam-1676	149	32	functions	function	NOUN
ejpam-1676	149	33	,	,	PUNCT
ejpam-1676	149	34	f1	f1	NOUN
ejpam-1676	149	35	,	,	PUNCT
ejpam-1676	149	36	.	.	PUNCT
ejpam-1676	149	37	.	.	PUNCT
ejpam-1676	149	38	.	.	PUNCT
ejpam-1676	150	1	,	,	PUNCT
ejpam-1676	150	2	fm	fm	PROPN
ejpam-1676	150	3	∈	∈	PROPN
ejpam-1676	150	4	v	v	NOUN
ejpam-1676	150	5	,	,	PUNCT
ejpam-1676	150	6	are	be	AUX
ejpam-1676	150	7	all	all	PRON
ejpam-1676	150	8	assumed	assume	VERB
ejpam-1676	150	9	to	to	PART
ejpam-1676	150	10	be	be	AUX
ejpam-1676	150	11	continuous	continuous	ADJ
ejpam-1676	150	12	where	where	SCONJ
ejpam-1676	150	13	v	v	NOUN
ejpam-1676	150	14	is	be	AUX
ejpam-1676	150	15	a	a	DET
ejpam-1676	150	16	reflexive	reflexive	ADJ
ejpam-1676	150	17	banach	banach	NOUN
ejpam-1676	150	18	space	space	NOUN
ejpam-1676	150	19	.	.	PUNCT
ejpam-1676	151	1	although	although	SCONJ
ejpam-1676	151	2	v	v	NOUN
ejpam-1676	151	3	can	can	AUX
ejpam-1676	151	4	be	be	AUX
ejpam-1676	151	5	quite	quite	ADV
ejpam-1676	151	6	arbitrary	arbitrary	ADJ
ejpam-1676	151	7	,	,	PUNCT
ejpam-1676	151	8	we	we	PRON
ejpam-1676	151	9	take	take	VERB
ejpam-1676	151	10	v	v	NOUN
ejpam-1676	151	11	=	=	SYM
ejpam-1676	151	12	lp	lp	NOUN
ejpam-1676	151	13	=	=	SYM
ejpam-1676	151	14	lp(ω	lp(ω	X
ejpam-1676	151	15	,	,	PUNCT
ejpam-1676	151	16	ν	ν	NOUN
ejpam-1676	151	17	)	)	PUNCT
ejpam-1676	151	18	.	.	PUNCT
ejpam-1676	152	1	it	it	PRON
ejpam-1676	152	2	is	be	AUX
ejpam-1676	152	3	known	know	VERB
ejpam-1676	152	4	that	that	SCONJ
ejpam-1676	152	5	the	the	DET
ejpam-1676	152	6	lp	lp	PROPN
ejpam-1676	152	7	spaces	space	NOUN
ejpam-1676	152	8	(	(	PUNCT
ejpam-1676	152	9	1	1	NUM
ejpam-1676	152	10	<	<	X
ejpam-1676	152	11	p	p	X
ejpam-1676	152	12	<	<	X
ejpam-1676	152	13	∞	∞	NUM
ejpam-1676	152	14	)	)	PUNCT
ejpam-1676	152	15	are	be	AUX
ejpam-1676	152	16	reflexive	reflexive	ADJ
ejpam-1676	152	17	but	but	CCONJ
ejpam-1676	152	18	that	that	SCONJ
ejpam-1676	152	19	l1	l1	PROPN
ejpam-1676	152	20	is	be	AUX
ejpam-1676	152	21	not	not	PART
ejpam-1676	152	22	[	[	PUNCT
ejpam-1676	152	23	see	see	VERB
ejpam-1676	152	24	16	16	NUM
ejpam-1676	152	25	,	,	PUNCT
ejpam-1676	152	26	for	for	ADP
ejpam-1676	152	27	example	example	NOUN
ejpam-1676	152	28	]	]	PUNCT
ejpam-1676	152	29	.	.	PUNCT
ejpam-1676	153	1	for	for	ADP
ejpam-1676	153	2	i	i	PRON
ejpam-1676	153	3	=	=	NOUN
ejpam-1676	153	4	2	2	NUM
ejpam-1676	153	5	,	,	PUNCT
ejpam-1676	153	6	.	.	PUNCT
ejpam-1676	153	7	.	.	PUNCT
ejpam-1676	154	1	.	.	PUNCT
ejpam-1676	155	1	,	,	PUNCT
ejpam-1676	155	2	m	m	X
ejpam-1676	155	3	,	,	PUNCT
ejpam-1676	155	4	we	we	PRON
ejpam-1676	155	5	define	define	VERB
ejpam-1676	155	6	ei	ei	X
ejpam-1676	155	7	=	=	PUNCT
ejpam-1676	155	8	{	{	PUNCT
ejpam-1676	155	9	g	g	PROPN
ejpam-1676	155	10	∈	∈	PROPN
ejpam-1676	155	11	lp	lp	NOUN
ejpam-1676	155	12	:	:	PUNCT
ejpam-1676	155	13	||g	||g	VERB
ejpam-1676	155	14	−	−	PROPN
ejpam-1676	155	15	fi	fi	NOUN
ejpam-1676	155	16	||p	||p	PROPN
ejpam-1676	155	17	<	<	X
ejpam-1676	155	18	ci	ci	PROPN
ejpam-1676	155	19	,	,	PUNCT
ejpam-1676	155	20	∫	∫	PROPN
ejpam-1676	155	21	fi(x	fi(x	NUM
ejpam-1676	155	22	)	)	PUNCT
ejpam-1676	155	23	log	log	VERB
ejpam-1676	155	24	g(x	g(x	NOUN
ejpam-1676	155	25	)	)	PUNCT
ejpam-1676	155	26	fi(x	fi(x	NUM
ejpam-1676	155	27	)	)	PUNCT
ejpam-1676	155	28	dν(x)≤	dν(x)≤	PROPN
ejpam-1676	155	29	ai	ai	PROPN
ejpam-1676	155	30	,	,	PUNCT
ejpam-1676	155	31	∫	∫	PROPN
ejpam-1676	155	32	g(x)dν(x	g(x)dν(x	X
ejpam-1676	155	33	)	)	PUNCT
ejpam-1676	155	34	=	=	SYM
ejpam-1676	155	35	1	1	NUM
ejpam-1676	155	36	,	,	PUNCT
ejpam-1676	155	37	g(x	g(x	NOUN
ejpam-1676	155	38	)	)	PUNCT
ejpam-1676	155	39	>	>	X
ejpam-1676	155	40	0	0	NUM
ejpam-1676	155	41	}	}	PUNCT
ejpam-1676	155	42	,	,	PUNCT
ejpam-1676	155	43	(	(	PUNCT
ejpam-1676	155	44	3	3	X
ejpam-1676	155	45	)	)	PUNCT
ejpam-1676	155	46	where	where	SCONJ
ejpam-1676	155	47	ai	ai	VERB
ejpam-1676	155	48	≥	≥	NOUN
ejpam-1676	155	49	0	0	NUM
ejpam-1676	155	50	and	and	CCONJ
ejpam-1676	155	51	ci	ci	NOUN
ejpam-1676	155	52	,	,	PUNCT
ejpam-1676	155	53	i	i	NOUN
ejpam-1676	155	54	=	=	NOUN
ejpam-1676	155	55	2,3	2,3	NUM
ejpam-1676	155	56	,	,	PUNCT
ejpam-1676	155	57	.	.	PUNCT
ejpam-1676	155	58	.	.	PUNCT
ejpam-1676	156	1	.	.	PUNCT
ejpam-1676	157	1	,	,	PUNCT
ejpam-1676	157	2	m	m	VERB
ejpam-1676	157	3	,	,	PUNCT
ejpam-1676	157	4	are	be	AUX
ejpam-1676	157	5	constants	constant	NOUN
ejpam-1676	157	6	.	.	PUNCT
ejpam-1676	158	1	we	we	PRON
ejpam-1676	158	2	then	then	ADV
ejpam-1676	158	3	define	define	VERB
ejpam-1676	158	4	the	the	DET
ejpam-1676	158	5	interception	interception	NOUN
ejpam-1676	158	6	of	of	ADP
ejpam-1676	158	7	these	these	DET
ejpam-1676	158	8	individual	individual	ADJ
ejpam-1676	158	9	enveloping	envelop	VERB
ejpam-1676	158	10	neighborhood	neighborhood	NOUN
ejpam-1676	158	11	.	.	PUNCT
ejpam-1676	159	1	e	e	X
ejpam-1676	159	2	=	=	PUNCT
ejpam-1676	159	3	∩m	∩m	PROPN
ejpam-1676	159	4	i=2ei	i=2ei	NOUN
ejpam-1676	159	5	.	.	PUNCT
ejpam-1676	160	1	(	(	PUNCT
ejpam-1676	160	2	4	4	X
ejpam-1676	160	3	)	)	PUNCT
ejpam-1676	160	4	p.	p.	NOUN
ejpam-1676	160	5	guo	guo	PROPN
ejpam-1676	160	6	,	,	PUNCT
ejpam-1676	160	7	x.	x.	PROPN
ejpam-1676	160	8	wang	wang	PROPN
ejpam-1676	160	9	,	,	PUNCT
ejpam-1676	160	10	y.	y.	PROPN
ejpam-1676	160	11	wu	wu	PROPN
ejpam-1676	160	12	/	/	SYM
ejpam-1676	160	13	eur	eur	PROPN
ejpam-1676	160	14	.	.	PUNCT
ejpam-1676	161	1	j.	j.	PROPN
ejpam-1676	161	2	pure	pure	PROPN
ejpam-1676	161	3	appl	appl	PROPN
ejpam-1676	161	4	.	.	PROPN
ejpam-1676	161	5	math	math	PROPN
ejpam-1676	161	6	,	,	PUNCT
ejpam-1676	161	7	5	5	NUM
ejpam-1676	161	8	(	(	PUNCT
ejpam-1676	161	9	2012	2012	NUM
ejpam-1676	161	10	)	)	PUNCT
ejpam-1676	161	11	,	,	PUNCT
ejpam-1676	161	12	333	333	NUM
ejpam-1676	161	13	-	-	SYM
ejpam-1676	161	14	356	356	NUM
ejpam-1676	161	15	338	338	NUM
ejpam-1676	161	16	we	we	PRON
ejpam-1676	161	17	remark	remark	VERB
ejpam-1676	161	18	that	that	SCONJ
ejpam-1676	161	19	the	the	DET
ejpam-1676	161	20	set	set	NOUN
ejpam-1676	161	21	e	e	NOUN
ejpam-1676	161	22	will	will	AUX
ejpam-1676	161	23	be	be	AUX
ejpam-1676	161	24	bounded	bound	VERB
ejpam-1676	161	25	with	with	ADP
ejpam-1676	161	26	respect	respect	NOUN
ejpam-1676	161	27	to	to	ADP
ejpam-1676	161	28	the	the	DET
ejpam-1676	161	29	lp	lp	ADJ
ejpam-1676	161	30	norm	norm	NOUN
ejpam-1676	161	31	and	and	CCONJ
ejpam-1676	161	32	non	non	ADJ
ejpam-1676	161	33	-	-	ADJ
ejpam-1676	161	34	empty	empty	ADJ
ejpam-1676	161	35	if	if	SCONJ
ejpam-1676	161	36	the	the	DET
ejpam-1676	161	37	constraints	constraint	NOUN
ejpam-1676	161	38	are	be	AUX
ejpam-1676	161	39	not	not	PART
ejpam-1676	161	40	too	too	ADV
ejpam-1676	161	41	restrictive	restrictive	ADJ
ejpam-1676	161	42	.	.	PUNCT
ejpam-1676	162	1	the	the	DET
ejpam-1676	162	2	latter	latter	ADJ
ejpam-1676	162	3	is	be	AUX
ejpam-1676	162	4	assumed	assume	VERB
ejpam-1676	162	5	throughout	throughout	ADP
ejpam-1676	162	6	.	.	PUNCT
ejpam-1676	163	1	thus	thus	ADV
ejpam-1676	163	2	,	,	PUNCT
ejpam-1676	163	3	for	for	ADP
ejpam-1676	163	4	a	a	DET
ejpam-1676	163	5	given	give	VERB
ejpam-1676	163	6	set	set	NOUN
ejpam-1676	163	7	of	of	ADP
ejpam-1676	163	8	density	density	NOUN
ejpam-1676	163	9	functions	function	NOUN
ejpam-1676	163	10	,	,	PUNCT
ejpam-1676	163	11	f1(x	f1(x	NOUN
ejpam-1676	163	12	)	)	PUNCT
ejpam-1676	163	13	being	be	AUX
ejpam-1676	163	14	primary	primary	ADJ
ejpam-1676	163	15	,	,	PUNCT
ejpam-1676	163	16	we	we	PRON
ejpam-1676	163	17	seek	seek	VERB
ejpam-1676	163	18	a	a	DET
ejpam-1676	163	19	probability	probability	NOUN
ejpam-1676	163	20	density	density	NOUN
ejpam-1676	163	21	function	function	NOUN
ejpam-1676	163	22	g	g	PROPN
ejpam-1676	163	23	∈	∈	PROPN
ejpam-1676	163	24	e	e	X
ejpam-1676	163	25	which	which	PRON
ejpam-1676	163	26	minimizes	minimize	VERB
ejpam-1676	163	27	i	i	PROPN
ejpam-1676	163	28	(	(	PUNCT
ejpam-1676	163	29	f1	f1	NOUN
ejpam-1676	163	30	,	,	PUNCT
ejpam-1676	163	31	g	g	NOUN
ejpam-1676	163	32	)	)	PUNCT
ejpam-1676	163	33	=	=	SYM
ejpam-1676	164	1	∫	∫	PROPN
ejpam-1676	164	2	f1(x	f1(x	NUM
ejpam-1676	164	3	)	)	PUNCT
ejpam-1676	164	4	log	log	NOUN
ejpam-1676	164	5	g(x	g(x	NOUN
ejpam-1676	164	6	)	)	PUNCT
ejpam-1676	164	7	f1(x	f1(x	NOUN
ejpam-1676	164	8	)	)	PUNCT
ejpam-1676	164	9	dν(x	dν(x	NOUN
ejpam-1676	164	10	)	)	PUNCT
ejpam-1676	164	11	over	over	ADP
ejpam-1676	164	12	all	all	DET
ejpam-1676	164	13	probability	probability	NOUN
ejpam-1676	164	14	densities	density	NOUN
ejpam-1676	164	15	,	,	PUNCT
ejpam-1676	164	16	g	g	NOUN
ejpam-1676	164	17	,	,	PUNCT
ejpam-1676	164	18	satisfying	satisfy	VERB
ejpam-1676	164	19	i	i	PRON
ejpam-1676	164	20	(	(	PUNCT
ejpam-1676	164	21	fi	fi	NOUN
ejpam-1676	164	22	,	,	PUNCT
ejpam-1676	164	23	g	g	NOUN
ejpam-1676	164	24	)	)	PUNCT
ejpam-1676	164	25	≤	≤	NOUN
ejpam-1676	164	26	ai	ai	VERB
ejpam-1676	164	27	,	,	PUNCT
ejpam-1676	164	28	i	i	PRON
ejpam-1676	164	29	=	=	NOUN
ejpam-1676	164	30	2	2	NUM
ejpam-1676	164	31	,	,	PUNCT
ejpam-1676	164	32	.	.	PUNCT
ejpam-1676	164	33	.	.	PUNCT
ejpam-1676	164	34	.	.	PUNCT
ejpam-1676	165	1	,	,	PUNCT
ejpam-1676	165	2	m	m	PRON
ejpam-1676	165	3	,	,	PUNCT
ejpam-1676	165	4	(	(	PUNCT
ejpam-1676	165	5	5	5	NUM
ejpam-1676	165	6	)	)	PUNCT
ejpam-1676	165	7	where	where	SCONJ
ejpam-1676	165	8	ai	ai	VERB
ejpam-1676	165	9	,	,	PUNCT
ejpam-1676	165	10	i	i	NOUN
ejpam-1676	165	11	=	=	NOUN
ejpam-1676	165	12	2,3	2,3	NUM
ejpam-1676	165	13	,	,	PUNCT
ejpam-1676	165	14	.	.	PUNCT
ejpam-1676	165	15	.	.	PUNCT
ejpam-1676	165	16	.	.	PUNCT
ejpam-1676	166	1	,	,	PUNCT
ejpam-1676	166	2	m	m	VERB
ejpam-1676	166	3	,	,	PUNCT
ejpam-1676	166	4	are	be	AUX
ejpam-1676	166	5	non	non	ADJ
ejpam-1676	166	6	-	-	ADJ
ejpam-1676	166	7	negative	negative	ADJ
ejpam-1676	166	8	constants	constant	NOUN
ejpam-1676	166	9	.	.	PUNCT
ejpam-1676	167	1	[	[	X
ejpam-1676	167	2	6	6	NUM
ejpam-1676	167	3	]	]	PUNCT
ejpam-1676	167	4	showed	show	VERB
ejpam-1676	167	5	that	that	SCONJ
ejpam-1676	167	6	the	the	DET
ejpam-1676	167	7	the	the	DET
ejpam-1676	167	8	optimal	optimal	ADJ
ejpam-1676	167	9	solution	solution	NOUN
ejpam-1676	167	10	takes	take	VERB
ejpam-1676	167	11	the	the	DET
ejpam-1676	167	12	form	form	NOUN
ejpam-1676	167	13	of	of	ADP
ejpam-1676	167	14	a	a	DET
ejpam-1676	167	15	mixing	mix	VERB
ejpam-1676	167	16	distribution	distribution	NOUN
ejpam-1676	167	17	:	:	PUNCT
ejpam-1676	167	18	g∗	g∗	PROPN
ejpam-1676	167	19	=	=	PUNCT
ejpam-1676	167	20	m∑	m∑	INTJ
ejpam-1676	167	21	i=1	i=1	PROPN
ejpam-1676	167	22	λi	λi	PROPN
ejpam-1676	167	23	fi(x	fi(x	NUM
ejpam-1676	167	24	)	)	PUNCT
ejpam-1676	167	25	,	,	PUNCT
ejpam-1676	167	26	(	(	PUNCT
ejpam-1676	167	27	6	6	NUM
ejpam-1676	167	28	)	)	PUNCT
ejpam-1676	167	29	where	where	SCONJ
ejpam-1676	167	30	g∗	g∗	PROPN
ejpam-1676	167	31	is	be	AUX
ejpam-1676	167	32	the	the	DET
ejpam-1676	167	33	optimal	optimal	ADJ
ejpam-1676	167	34	solution	solution	NOUN
ejpam-1676	167	35	for	for	ADP
ejpam-1676	167	36	optimization	optimization	NOUN
ejpam-1676	167	37	probelm	probelm	PROPN
ejpam-1676	167	38	described	describe	VERB
ejpam-1676	167	39	above	above	ADV
ejpam-1676	167	40	.	.	PUNCT
ejpam-1676	168	1	assuming	assume	VERB
ejpam-1676	168	2	that	that	DET
ejpam-1676	168	3	fi	fi	NOUN
ejpam-1676	168	4	are	be	AUX
ejpam-1676	168	5	members	member	NOUN
ejpam-1676	168	6	of	of	ADP
ejpam-1676	168	7	one	one	NUM
ejpam-1676	168	8	parametric	parametric	ADJ
ejpam-1676	168	9	family	family	NOUN
ejpam-1676	168	10	while	while	SCONJ
ejpam-1676	168	11	the	the	DET
ejpam-1676	168	12	difference	difference	NOUN
ejpam-1676	168	13	is	be	AUX
ejpam-1676	168	14	due	due	ADJ
ejpam-1676	168	15	to	to	ADP
ejpam-1676	168	16	different	different	ADJ
ejpam-1676	168	17	value	value	NOUN
ejpam-1676	168	18	for	for	ADP
ejpam-1676	168	19	the	the	DET
ejpam-1676	168	20	parameter	parameter	NOUN
ejpam-1676	168	21	,	,	PUNCT
ejpam-1676	168	22	i.e.	i.e.	X
ejpam-1676	168	23	fi(x	fi(x	X
ejpam-1676	168	24	)	)	PUNCT
ejpam-1676	168	25	=	=	SYM
ejpam-1676	168	26	f	f	X
ejpam-1676	168	27	(	(	PUNCT
ejpam-1676	168	28	x	x	NOUN
ejpam-1676	168	29	;	;	PUNCT
ejpam-1676	168	30	θi	θi	NUM
ejpam-1676	168	31	)	)	PUNCT
ejpam-1676	168	32	.	.	PUNCT
ejpam-1676	169	1	without	without	ADP
ejpam-1676	169	2	the	the	DET
ejpam-1676	169	3	knowledge	knowledge	NOUN
ejpam-1676	169	4	of	of	ADP
ejpam-1676	169	5	the	the	DET
ejpam-1676	169	6	exact	exact	ADJ
ejpam-1676	169	7	values	value	NOUN
ejpam-1676	169	8	for	for	ADP
ejpam-1676	169	9	θi	θi	NUM
ejpam-1676	169	10	,	,	PUNCT
ejpam-1676	169	11	[	[	X
ejpam-1676	169	12	6	6	NUM
ejpam-1676	169	13	]	]	PUNCT
ejpam-1676	169	14	show	show	VERB
ejpam-1676	169	15	that	that	SCONJ
ejpam-1676	169	16	the	the	DET
ejpam-1676	169	17	weighted	weight	VERB
ejpam-1676	169	18	likelihood	likelihood	NOUN
ejpam-1676	169	19	function	function	NOUN
ejpam-1676	169	20	:	:	PUNCT
ejpam-1676	169	21	wl(y;θ1	wl(y;θ1	ADJ
ejpam-1676	169	22	)	)	PUNCT
ejpam-1676	169	23	=	=	SYM
ejpam-1676	169	24	m∏	m∏	PROPN
ejpam-1676	169	25	i=1	i=1	PUNCT
ejpam-1676	170	1	ni∏	ni∏	NOUN
ejpam-1676	171	1	j=1	j=1	ADJ
ejpam-1676	171	2	f	f	PROPN
ejpam-1676	171	3	(	(	PUNCT
ejpam-1676	171	4	yi	yi	PROPN
ejpam-1676	171	5	j;θ1	j;θ1	PROPN
ejpam-1676	171	6	)	)	PUNCT
ejpam-1676	171	7	λi	λi	VERB
ejpam-1676	171	8	,	,	PUNCT
ejpam-1676	171	9	(	(	PUNCT
ejpam-1676	171	10	7	7	X
ejpam-1676	171	11	)	)	PUNCT
ejpam-1676	171	12	is	be	AUX
ejpam-1676	171	13	the	the	DET
ejpam-1676	171	14	correct	correct	ADJ
ejpam-1676	171	15	device	device	NOUN
ejpam-1676	171	16	to	to	PART
ejpam-1676	171	17	used	use	VERB
ejpam-1676	171	18	by	by	ADP
ejpam-1676	171	19	following	follow	VERB
ejpam-1676	171	20	the	the	DET
ejpam-1676	171	21	argument	argument	NOUN
ejpam-1676	171	22	of	of	ADP
ejpam-1676	171	23	[	[	X
ejpam-1676	171	24	13	13	NUM
ejpam-1676	171	25	]	]	PUNCT
ejpam-1676	171	26	.	.	PUNCT
ejpam-1676	172	1	the	the	DET
ejpam-1676	172	2	weights	weight	NOUN
ejpam-1676	172	3	are	be	AUX
ejpam-1676	172	4	functions	function	NOUN
ejpam-1676	172	5	of	of	ADP
ejpam-1676	172	6	the	the	DET
ejpam-1676	172	7	λi	λi	X
ejpam-1676	172	8	’s	’s	ADJ
ejpam-1676	172	9	which	which	PRON
ejpam-1676	172	10	measures	measure	VERB
ejpam-1676	172	11	the	the	DET
ejpam-1676	172	12	discrepancies	discrepancy	NOUN
ejpam-1676	172	13	among	among	ADP
ejpam-1676	172	14	all	all	DET
ejpam-1676	172	15	parameters	parameter	NOUN
ejpam-1676	172	16	.	.	PUNCT
ejpam-1676	173	1	we	we	PRON
ejpam-1676	173	2	remark	remark	VERB
ejpam-1676	173	3	that	that	SCONJ
ejpam-1676	173	4	this	this	DET
ejpam-1676	173	5	classical	classical	ADJ
ejpam-1676	173	6	likelihood	likelihood	NOUN
ejpam-1676	173	7	is	be	AUX
ejpam-1676	173	8	embraced	embrace	VERB
ejpam-1676	173	9	by	by	ADP
ejpam-1676	173	10	the	the	DET
ejpam-1676	173	11	proposed	propose	VERB
ejpam-1676	173	12	likelihood	likelihood	NOUN
ejpam-1676	173	13	by	by	ADP
ejpam-1676	173	14	setting	set	VERB
ejpam-1676	173	15	the	the	DET
ejpam-1676	173	16	weights	weight	NOUN
ejpam-1676	173	17	to	to	PART
ejpam-1676	173	18	be	be	AUX
ejpam-1676	173	19	zero	zero	NUM
ejpam-1676	173	20	for	for	ADP
ejpam-1676	173	21	relevant	relevant	ADJ
ejpam-1676	173	22	samples	sample	NOUN
ejpam-1676	173	23	.	.	PUNCT
ejpam-1676	174	1	in	in	ADP
ejpam-1676	174	2	addition	addition	NOUN
ejpam-1676	174	3	,	,	PUNCT
ejpam-1676	174	4	[	[	X
ejpam-1676	174	5	5	5	NUM
ejpam-1676	174	6	]	]	PUNCT
ejpam-1676	174	7	showed	show	VERB
ejpam-1676	174	8	that	that	SCONJ
ejpam-1676	174	9	the	the	DET
ejpam-1676	174	10	estimator	estimator	NOUN
ejpam-1676	174	11	derived	derive	VERB
ejpam-1676	174	12	from	from	ADP
ejpam-1676	174	13	the	the	DET
ejpam-1676	174	14	weighted	weight	VERB
ejpam-1676	174	15	likelihood	likelihood	NOUN
ejpam-1676	174	16	function	function	NOUN
ejpam-1676	174	17	can	can	AUX
ejpam-1676	174	18	be	be	AUX
ejpam-1676	174	19	viewed	view	VERB
ejpam-1676	174	20	as	as	ADP
ejpam-1676	174	21	an	an	DET
ejpam-1676	174	22	approximate	approximate	ADJ
ejpam-1676	174	23	bayes	bayes	NOUN
ejpam-1676	174	24	decision	decision	NOUN
ejpam-1676	174	25	by	by	ADP
ejpam-1676	174	26	following	follow	VERB
ejpam-1676	174	27	the	the	DET
ejpam-1676	174	28	same	same	ADJ
ejpam-1676	174	29	argument	argument	NOUN
ejpam-1676	174	30	by	by	ADP
ejpam-1676	174	31	[	[	X
ejpam-1676	174	32	3	3	NUM
ejpam-1676	174	33	]	]	PUNCT
ejpam-1676	174	34	.	.	PUNCT
ejpam-1676	175	1	the	the	DET
ejpam-1676	175	2	coefficients	coefficient	NOUN
ejpam-1676	175	3	ai	ai	VERB
ejpam-1676	175	4	’s	’s	ADV
ejpam-1676	175	5	,	,	PUNCT
ejpam-1676	175	6	however	however	ADV
ejpam-1676	175	7	,	,	PUNCT
ejpam-1676	175	8	are	be	AUX
ejpam-1676	175	9	not	not	PART
ejpam-1676	175	10	known	know	VERB
ejpam-1676	175	11	since	since	SCONJ
ejpam-1676	175	12	the	the	DET
ejpam-1676	175	13	exact	exact	ADJ
ejpam-1676	175	14	functional	functional	ADJ
ejpam-1676	175	15	form	form	NOUN
ejpam-1676	175	16	of	of	ADP
ejpam-1676	175	17	the	the	DET
ejpam-1676	175	18	connections	connection	NOUN
ejpam-1676	175	19	among	among	ADP
ejpam-1676	175	20	all	all	DET
ejpam-1676	175	21	parameters	parameter	NOUN
ejpam-1676	175	22	are	be	AUX
ejpam-1676	175	23	not	not	PART
ejpam-1676	175	24	assumed	assume	VERB
ejpam-1676	175	25	to	to	PART
ejpam-1676	175	26	be	be	AUX
ejpam-1676	175	27	known	know	VERB
ejpam-1676	175	28	.	.	PUNCT
ejpam-1676	176	1	therefore	therefore	ADV
ejpam-1676	176	2	,	,	PUNCT
ejpam-1676	176	3	the	the	DET
ejpam-1676	176	4	weights	weight	NOUN
ejpam-1676	176	5	must	must	AUX
ejpam-1676	176	6	be	be	AUX
ejpam-1676	176	7	chosen	choose	VERB
ejpam-1676	176	8	adaptively	adaptively	ADV
ejpam-1676	176	9	to	to	PART
ejpam-1676	176	10	avoid	avoid	VERB
ejpam-1676	176	11	introduction	introduction	NOUN
ejpam-1676	176	12	of	of	ADP
ejpam-1676	176	13	significant	significant	ADJ
ejpam-1676	176	14	bias	bias	NOUN
ejpam-1676	176	15	into	into	ADP
ejpam-1676	176	16	the	the	DET
ejpam-1676	176	17	inference	inference	NOUN
ejpam-1676	176	18	for	for	ADP
ejpam-1676	176	19	the	the	DET
ejpam-1676	176	20	target	target	NOUN
ejpam-1676	176	21	population	population	NOUN
ejpam-1676	176	22	.	.	PUNCT
ejpam-1676	177	1	2.2	2.2	NUM
ejpam-1676	177	2	.	.	PUNCT
ejpam-1676	178	1	adaptive	adaptive	ADJ
ejpam-1676	178	2	weights	weight	NOUN
ejpam-1676	178	3	based	base	VERB
ejpam-1676	178	4	on	on	ADP
ejpam-1676	178	5	likelihood	likelihood	NOUN
ejpam-1676	178	6	ratio	ratio	NOUN
ejpam-1676	178	7	since	since	SCONJ
ejpam-1676	178	8	the	the	DET
ejpam-1676	178	9	weighted	weight	VERB
ejpam-1676	178	10	likelihood	likelihood	NOUN
ejpam-1676	178	11	is	be	AUX
ejpam-1676	178	12	the	the	DET
ejpam-1676	178	13	chosen	choose	VERB
ejpam-1676	178	14	instrument	instrument	NOUN
ejpam-1676	178	15	for	for	ADP
ejpam-1676	178	16	the	the	DET
ejpam-1676	178	17	data	data	NOUN
ejpam-1676	178	18	fusion	fusion	NOUN
ejpam-1676	178	19	process	process	NOUN
ejpam-1676	178	20	,	,	PUNCT
ejpam-1676	178	21	the	the	DET
ejpam-1676	178	22	likelihood	likelihood	NOUN
ejpam-1676	178	23	weights	weight	NOUN
ejpam-1676	178	24	play	play	VERB
ejpam-1676	178	25	a	a	DET
ejpam-1676	178	26	central	central	ADJ
ejpam-1676	178	27	role	role	NOUN
ejpam-1676	178	28	in	in	ADP
ejpam-1676	178	29	integrating	integrate	VERB
ejpam-1676	178	30	all	all	DET
ejpam-1676	178	31	relevant	relevant	ADJ
ejpam-1676	178	32	information	information	NOUN
ejpam-1676	178	33	in	in	ADP
ejpam-1676	178	34	a	a	DET
ejpam-1676	178	35	meaningful	meaningful	ADJ
ejpam-1676	178	36	way	way	NOUN
ejpam-1676	178	37	.	.	PUNCT
ejpam-1676	179	1	as	as	SCONJ
ejpam-1676	179	2	we	we	PRON
ejpam-1676	179	3	have	have	AUX
ejpam-1676	179	4	seen	see	VERB
ejpam-1676	179	5	in	in	ADP
ejpam-1676	179	6	previous	previous	ADJ
ejpam-1676	179	7	section	section	NOUN
ejpam-1676	179	8	,	,	PUNCT
ejpam-1676	179	9	the	the	DET
ejpam-1676	179	10	likelihood	likelihood	NOUN
ejpam-1676	179	11	weights	weight	NOUN
ejpam-1676	179	12	should	should	AUX
ejpam-1676	179	13	be	be	AUX
ejpam-1676	179	14	chosen	choose	VERB
ejpam-1676	179	15	to	to	PART
ejpam-1676	179	16	reflect	reflect	VERB
ejpam-1676	179	17	the	the	DET
ejpam-1676	179	18	true	true	ADJ
ejpam-1676	179	19	relevance	relevance	NOUN
ejpam-1676	179	20	between	between	ADP
ejpam-1676	179	21	a	a	DET
ejpam-1676	179	22	relevant	relevant	ADJ
ejpam-1676	179	23	sample	sample	NOUN
ejpam-1676	179	24	and	and	CCONJ
ejpam-1676	179	25	the	the	DET
ejpam-1676	179	26	target	target	NOUN
ejpam-1676	179	27	population	population	NOUN
ejpam-1676	179	28	.	.	PUNCT
ejpam-1676	180	1	without	without	ADP
ejpam-1676	180	2	any	any	DET
ejpam-1676	180	3	prior	prior	ADJ
ejpam-1676	180	4	knowledge	knowledge	NOUN
ejpam-1676	180	5	,	,	PUNCT
ejpam-1676	180	6	the	the	DET
ejpam-1676	180	7	evidence	evidence	NOUN
ejpam-1676	180	8	expressed	express	VERB
ejpam-1676	180	9	in	in	ADP
ejpam-1676	180	10	the	the	DET
ejpam-1676	180	11	relevant	relevant	ADJ
ejpam-1676	180	12	sample	sample	NOUN
ejpam-1676	180	13	should	should	AUX
ejpam-1676	180	14	be	be	AUX
ejpam-1676	180	15	the	the	DET
ejpam-1676	180	16	only	only	ADJ
ejpam-1676	180	17	source	source	NOUN
ejpam-1676	180	18	for	for	ADP
ejpam-1676	180	19	evaluation	evaluation	NOUN
ejpam-1676	180	20	and	and	CCONJ
ejpam-1676	180	21	determination	determination	NOUN
ejpam-1676	180	22	of	of	ADP
ejpam-1676	180	23	a	a	DET
ejpam-1676	180	24	set	set	NOUN
ejpam-1676	180	25	of	of	ADP
ejpam-1676	180	26	appropriate	appropriate	ADJ
ejpam-1676	180	27	weights	weight	NOUN
ejpam-1676	180	28	.	.	PUNCT
ejpam-1676	181	1	the	the	DET
ejpam-1676	181	2	fundamental	fundamental	ADJ
ejpam-1676	181	3	question	question	NOUN
ejpam-1676	181	4	is	be	AUX
ejpam-1676	181	5	how	how	SCONJ
ejpam-1676	181	6	to	to	PART
ejpam-1676	181	7	measure	measure	VERB
ejpam-1676	181	8	the	the	DET
ejpam-1676	181	9	discrepancy	discrepancy	NOUN
ejpam-1676	181	10	between	between	ADP
ejpam-1676	181	11	to	to	ADP
ejpam-1676	181	12	two	two	NUM
ejpam-1676	181	13	populations	population	NOUN
ejpam-1676	181	14	that	that	PRON
ejpam-1676	181	15	could	could	AUX
ejpam-1676	181	16	be	be	AUX
ejpam-1676	181	17	related	relate	VERB
ejpam-1676	181	18	.	.	PUNCT
ejpam-1676	182	1	in	in	ADP
ejpam-1676	182	2	a	a	DET
ejpam-1676	182	3	likelihood	likelihood	NOUN
ejpam-1676	182	4	ratio	ratio	NOUN
ejpam-1676	182	5	context	context	NOUN
ejpam-1676	182	6	,	,	PUNCT
ejpam-1676	182	7	a	a	DET
ejpam-1676	182	8	likelihood	likelihood	NOUN
ejpam-1676	182	9	-	-	PUNCT
ejpam-1676	182	10	ratio	ratio	NOUN
ejpam-1676	182	11	test	test	NOUN
ejpam-1676	182	12	is	be	AUX
ejpam-1676	182	13	a	a	DET
ejpam-1676	182	14	statistical	statistical	ADJ
ejpam-1676	182	15	test	test	NOUN
ejpam-1676	182	16	for	for	ADP
ejpam-1676	182	17	making	make	VERB
ejpam-1676	182	18	a	a	DET
ejpam-1676	182	19	decision	decision	NOUN
ejpam-1676	182	20	between	between	ADP
ejpam-1676	182	21	two	two	NUM
ejpam-1676	182	22	hypotheses	hypothesis	NOUN
ejpam-1676	182	23	based	base	VERB
ejpam-1676	182	24	on	on	ADP
ejpam-1676	182	25	the	the	DET
ejpam-1676	182	26	value	value	NOUN
ejpam-1676	182	27	of	of	ADP
ejpam-1676	182	28	this	this	DET
ejpam-1676	182	29	ratio	ratio	NOUN
ejpam-1676	182	30	.	.	PUNCT
ejpam-1676	183	1	since	since	SCONJ
ejpam-1676	183	2	the	the	DET
ejpam-1676	183	3	parameters	parameter	NOUN
ejpam-1676	183	4	of	of	ADP
ejpam-1676	183	5	these	these	DET
ejpam-1676	183	6	two	two	NUM
ejpam-1676	183	7	populations	population	NOUN
ejpam-1676	183	8	are	be	AUX
ejpam-1676	183	9	not	not	PART
ejpam-1676	183	10	known	know	VERB
ejpam-1676	183	11	,	,	PUNCT
ejpam-1676	183	12	therefore	therefore	ADV
ejpam-1676	183	13	we	we	PRON
ejpam-1676	183	14	plug	plug	VERB
ejpam-1676	183	15	in	in	ADP
ejpam-1676	183	16	the	the	DET
ejpam-1676	183	17	mle	mle	PROPN
ejpam-1676	183	18	estimates	estimate	NOUN
ejpam-1676	183	19	.	.	PUNCT
ejpam-1676	184	1	when	when	SCONJ
ejpam-1676	184	2	the	the	DET
ejpam-1676	184	3	ratio	ratio	NOUN
ejpam-1676	184	4	measures	measure	VERB
ejpam-1676	184	5	the	the	DET
ejpam-1676	184	6	resemblance	resemblance	NOUN
ejpam-1676	184	7	between	between	ADP
ejpam-1676	184	8	the	the	DET
ejpam-1676	184	9	two	two	NUM
ejpam-1676	184	10	values	value	NOUN
ejpam-1676	184	11	.	.	PUNCT
ejpam-1676	185	1	similarly	similarly	ADV
ejpam-1676	185	2	p.	p.	PROPN
ejpam-1676	185	3	guo	guo	PROPN
ejpam-1676	185	4	,	,	PUNCT
ejpam-1676	185	5	x.	x.	PROPN
ejpam-1676	185	6	wang	wang	PROPN
ejpam-1676	185	7	,	,	PUNCT
ejpam-1676	185	8	y.	y.	PROPN
ejpam-1676	185	9	wu	wu	PROPN
ejpam-1676	185	10	/	/	SYM
ejpam-1676	185	11	eur	eur	PROPN
ejpam-1676	185	12	.	.	PUNCT
ejpam-1676	186	1	j.	j.	PROPN
ejpam-1676	186	2	pure	pure	PROPN
ejpam-1676	186	3	appl	appl	PROPN
ejpam-1676	186	4	.	.	PROPN
ejpam-1676	186	5	math	math	PROPN
ejpam-1676	186	6	,	,	PUNCT
ejpam-1676	186	7	5	5	NUM
ejpam-1676	186	8	(	(	PUNCT
ejpam-1676	186	9	2012	2012	NUM
ejpam-1676	186	10	)	)	PUNCT
ejpam-1676	186	11	,	,	PUNCT
ejpam-1676	186	12	333	333	NUM
ejpam-1676	186	13	-	-	SYM
ejpam-1676	186	14	356	356	NUM
ejpam-1676	186	15	339	339	NUM
ejpam-1676	186	16	to	to	ADP
ejpam-1676	186	17	the	the	DET
ejpam-1676	186	18	hypothesis	hypothesis	NOUN
ejpam-1676	186	19	test	test	NOUN
ejpam-1676	187	1	,	,	PUNCT
ejpam-1676	187	2	it	it	PRON
ejpam-1676	187	3	implies	imply	VERB
ejpam-1676	187	4	that	that	SCONJ
ejpam-1676	187	5	the	the	DET
ejpam-1676	187	6	two	two	NUM
ejpam-1676	187	7	parameter	parameter	NOUN
ejpam-1676	187	8	values	value	NOUN
ejpam-1676	187	9	are	be	AUX
ejpam-1676	187	10	too	too	ADV
ejpam-1676	187	11	different	different	ADJ
ejpam-1676	187	12	in	in	ADP
ejpam-1676	187	13	the	the	DET
ejpam-1676	187	14	context	context	NOUN
ejpam-1676	187	15	of	of	ADP
ejpam-1676	187	16	likelihood	likelihood	NOUN
ejpam-1676	187	17	ratio	ratio	NOUN
ejpam-1676	187	18	.	.	PUNCT
ejpam-1676	188	1	following	follow	VERB
ejpam-1676	188	2	the	the	DET
ejpam-1676	188	3	same	same	ADJ
ejpam-1676	188	4	line	line	NOUN
ejpam-1676	188	5	of	of	ADP
ejpam-1676	188	6	argument	argument	NOUN
ejpam-1676	188	7	,	,	PUNCT
ejpam-1676	188	8	we	we	PRON
ejpam-1676	188	9	propose	propose	VERB
ejpam-1676	188	10	to	to	PART
ejpam-1676	188	11	use	use	VERB
ejpam-1676	188	12	the	the	DET
ejpam-1676	188	13	likelihood	likelihood	NOUN
ejpam-1676	188	14	ratio	ratio	NOUN
ejpam-1676	188	15	as	as	ADP
ejpam-1676	188	16	a	a	DET
ejpam-1676	188	17	measure	measure	NOUN
ejpam-1676	188	18	of	of	ADP
ejpam-1676	188	19	discrepancy	discrepancy	NOUN
ejpam-1676	188	20	.	.	PUNCT
ejpam-1676	189	1	since	since	SCONJ
ejpam-1676	189	2	the	the	DET
ejpam-1676	189	3	true	true	ADJ
ejpam-1676	189	4	values	value	NOUN
ejpam-1676	189	5	of	of	ADP
ejpam-1676	189	6	the	the	DET
ejpam-1676	189	7	parameters	parameter	NOUN
ejpam-1676	189	8	are	be	AUX
ejpam-1676	189	9	not	not	PART
ejpam-1676	189	10	known	know	VERB
ejpam-1676	189	11	,	,	PUNCT
ejpam-1676	189	12	it	it	PRON
ejpam-1676	189	13	is	be	AUX
ejpam-1676	189	14	natural	natural	ADJ
ejpam-1676	189	15	to	to	PART
ejpam-1676	189	16	use	use	VERB
ejpam-1676	189	17	estimated	estimate	VERB
ejpam-1676	189	18	likelihood	likelihood	NOUN
ejpam-1676	189	19	ratio	ratio	NOUN
ejpam-1676	189	20	with	with	ADP
ejpam-1676	189	21	the	the	DET
ejpam-1676	189	22	parameter	parameter	NOUN
ejpam-1676	189	23	values	value	NOUN
ejpam-1676	189	24	replaced	replace	VERB
ejpam-1676	189	25	by	by	ADP
ejpam-1676	189	26	corresponding	correspond	VERB
ejpam-1676	189	27	mle	mle	PROPN
ejpam-1676	189	28	estimates	estimate	NOUN
ejpam-1676	189	29	.	.	PUNCT
ejpam-1676	190	1	given	give	VERB
ejpam-1676	190	2	a	a	DET
ejpam-1676	190	3	parametric	parametric	ADJ
ejpam-1676	190	4	density	density	NOUN
ejpam-1676	190	5	function	function	NOUN
ejpam-1676	190	6	f1	f1	NOUN
ejpam-1676	190	7	and	and	CCONJ
ejpam-1676	190	8	a	a	DET
ejpam-1676	190	9	sample	sample	NOUN
ejpam-1676	190	10	y	y	NOUN
ejpam-1676	190	11	1	1	NUM
ejpam-1676	190	12	=	=	SYM
ejpam-1676	190	13	(	(	PUNCT
ejpam-1676	190	14	y11	y11	NOUN
ejpam-1676	190	15	,	,	PUNCT
ejpam-1676	190	16	y12	y12	PROPN
ejpam-1676	190	17	,	,	PUNCT
ejpam-1676	190	18	.	.	PUNCT
ejpam-1676	190	19	.	.	PUNCT
ejpam-1676	190	20	.	.	PUNCT
ejpam-1676	191	1	,	,	PUNCT
ejpam-1676	191	2	y1n1	y1n1	PROPN
ejpam-1676	191	3	)	)	PUNCT
ejpam-1676	191	4	,	,	PUNCT
ejpam-1676	191	5	the	the	DET
ejpam-1676	191	6	classical	classical	ADJ
ejpam-1676	191	7	likelihood	likelihood	NOUN
ejpam-1676	191	8	ratio	ratio	NOUN
ejpam-1676	191	9	for	for	ADP
ejpam-1676	191	10	testing	test	VERB
ejpam-1676	191	11	a	a	DET
ejpam-1676	191	12	null	null	ADJ
ejpam-1676	191	13	hypothesis	hypothesis	NOUN
ejpam-1676	191	14	is	be	AUX
ejpam-1676	191	15	based	base	VERB
ejpam-1676	191	16	on	on	ADP
ejpam-1676	191	17	the	the	DET
ejpam-1676	191	18	following	follow	VERB
ejpam-1676	191	19	lr	lr	X
ejpam-1676	191	20	=	=	SYM
ejpam-1676	191	21	supωh	supωh	NOUN
ejpam-1676	191	22	n1∏	n1∏	NOUN
ejpam-1676	191	23	j=1	j=1	PROPN
ejpam-1676	191	24	f1(y1	f1(y1	PROPN
ejpam-1676	191	25	j;θ1	j;θ1	PROPN
ejpam-1676	191	26	)	)	PUNCT
ejpam-1676	191	27	supω	supω	NOUN
ejpam-1676	191	28	n1∏	n1∏	ADP
ejpam-1676	191	29	j=1	j=1	PROPN
ejpam-1676	191	30	f1(y1	f1(y1	PROPN
ejpam-1676	191	31	j;θ1	j;θ1	PROPN
ejpam-1676	191	32	)	)	PUNCT
ejpam-1676	191	33	,	,	PUNCT
ejpam-1676	191	34	where	where	SCONJ
ejpam-1676	191	35	ωh	ωh	ADP
ejpam-1676	191	36	is	be	AUX
ejpam-1676	191	37	a	a	DET
ejpam-1676	191	38	subspace	subspace	NOUN
ejpam-1676	191	39	in	in	ADP
ejpam-1676	191	40	the	the	DET
ejpam-1676	191	41	parameter	parameter	NOUN
ejpam-1676	191	42	space	space	NOUN
ejpam-1676	191	43	under	under	ADP
ejpam-1676	191	44	the	the	DET
ejpam-1676	191	45	null	null	ADJ
ejpam-1676	191	46	hypothesis	hypothesis	NOUN
ejpam-1676	191	47	h.	h.	PROPN
ejpam-1676	191	48	assume	assume	VERB
ejpam-1676	191	49	that	that	SCONJ
ejpam-1676	191	50	the	the	DET
ejpam-1676	191	51	likelihood	likelihood	NOUN
ejpam-1676	191	52	function	function	NOUN
ejpam-1676	191	53	attains	attain	VERB
ejpam-1676	191	54	an	an	DET
ejpam-1676	191	55	unique	unique	ADJ
ejpam-1676	191	56	maximum	maximum	NOUN
ejpam-1676	191	57	in	in	ADP
ejpam-1676	191	58	ωh	ωh	ADP
ejpam-1676	191	59	and	and	CCONJ
ejpam-1676	191	60	ω	ω	NUM
ejpam-1676	191	61	respectively	respectively	ADV
ejpam-1676	191	62	.	.	PUNCT
ejpam-1676	192	1	denote	denote	VERB
ejpam-1676	192	2	the	the	DET
ejpam-1676	192	3	maximizers	maximizer	NOUN
ejpam-1676	192	4	as	as	ADP
ejpam-1676	192	5	bθh	bθh	NOUN
ejpam-1676	192	6	1	1	NUM
ejpam-1676	192	7	and	and	CCONJ
ejpam-1676	192	8	bθ1	bθ1	NOUN
ejpam-1676	192	9	respectively	respectively	ADV
ejpam-1676	192	10	.	.	PUNCT
ejpam-1676	193	1	the	the	DET
ejpam-1676	193	2	likelihood	likelihood	NOUN
ejpam-1676	193	3	ratio	ratio	NOUN
ejpam-1676	193	4	(	(	PUNCT
ejpam-1676	193	5	lr	lr	NOUN
ejpam-1676	193	6	)	)	PUNCT
ejpam-1676	193	7	can	can	AUX
ejpam-1676	193	8	then	then	ADV
ejpam-1676	193	9	be	be	AUX
ejpam-1676	193	10	rewritten	rewrite	VERB
ejpam-1676	193	11	as	as	ADP
ejpam-1676	193	12	lr	lr	NOUN
ejpam-1676	193	13	=	=	PUNCT
ejpam-1676	193	14	n1∏	n1∏	NOUN
ejpam-1676	193	15	j=1	j=1	PROPN
ejpam-1676	193	16	f1(y1	f1(y1	PROPN
ejpam-1676	193	17	j	j	PROPN
ejpam-1676	193	18	;	;	PUNCT
ejpam-1676	193	19	bθh	bθh	NOUN
ejpam-1676	193	20	1	1	X
ejpam-1676	193	21	)	)	PUNCT
ejpam-1676	193	22	n1∏	n1∏	NOUN
ejpam-1676	193	23	j=1	j=1	PROPN
ejpam-1676	193	24	f1(y1	f1(y1	PROPN
ejpam-1676	193	25	j	j	PROPN
ejpam-1676	193	26	;	;	PUNCT
ejpam-1676	193	27	bθ1	bθ1	PROPN
ejpam-1676	193	28	)	)	PUNCT
ejpam-1676	193	29	.	.	PUNCT
ejpam-1676	194	1	therefore	therefore	ADV
ejpam-1676	194	2	,	,	PUNCT
ejpam-1676	194	3	the	the	DET
ejpam-1676	194	4	validity	validity	NOUN
ejpam-1676	194	5	of	of	ADP
ejpam-1676	194	6	the	the	DET
ejpam-1676	194	7	proposed	propose	VERB
ejpam-1676	194	8	hypothesis	hypothesis	NOUN
ejpam-1676	194	9	is	be	AUX
ejpam-1676	194	10	evaluated	evaluate	VERB
ejpam-1676	194	11	according	accord	VERB
ejpam-1676	194	12	to	to	ADP
ejpam-1676	194	13	the	the	DET
ejpam-1676	194	14	likelihood	likelihood	NOUN
ejpam-1676	194	15	ratio	ratio	NOUN
ejpam-1676	194	16	.	.	PUNCT
ejpam-1676	195	1	by	by	ADP
ejpam-1676	195	2	using	use	VERB
ejpam-1676	195	3	a	a	DET
ejpam-1676	195	4	similar	similar	ADJ
ejpam-1676	195	5	idea	idea	NOUN
ejpam-1676	195	6	,	,	PUNCT
ejpam-1676	195	7	we	we	PRON
ejpam-1676	195	8	propose	propose	VERB
ejpam-1676	195	9	to	to	PART
ejpam-1676	195	10	choose	choose	VERB
ejpam-1676	195	11	likelihood	likelihood	NOUN
ejpam-1676	195	12	weights	weight	NOUN
ejpam-1676	195	13	by	by	ADP
ejpam-1676	195	14	using	use	VERB
ejpam-1676	195	15	the	the	DET
ejpam-1676	195	16	estimated	estimate	VERB
ejpam-1676	195	17	pairwise	pairwise	NOUN
ejpam-1676	195	18	likelihood	likelihood	NOUN
ejpam-1676	195	19	ratio	ratio	NOUN
ejpam-1676	195	20	.	.	PUNCT
ejpam-1676	196	1	for	for	ADP
ejpam-1676	196	2	simplicity	simplicity	NOUN
ejpam-1676	196	3	,	,	PUNCT
ejpam-1676	196	4	let	let	VERB
ejpam-1676	196	5	m	m	NOUN
ejpam-1676	196	6	=	=	NOUN
ejpam-1676	196	7	2	2	NUM
ejpam-1676	196	8	,	,	PUNCT
ejpam-1676	196	9	i.e.	i.e.	X
ejpam-1676	196	10	,	,	PUNCT
ejpam-1676	196	11	there	there	PRON
ejpam-1676	196	12	is	be	VERB
ejpam-1676	196	13	only	only	ADV
ejpam-1676	196	14	one	one	NUM
ejpam-1676	196	15	related	relate	VERB
ejpam-1676	196	16	population	population	NOUN
ejpam-1676	196	17	together	together	ADV
ejpam-1676	196	18	with	with	ADP
ejpam-1676	196	19	the	the	DET
ejpam-1676	196	20	target	target	NOUN
ejpam-1676	196	21	population	population	NOUN
ejpam-1676	196	22	.	.	PUNCT
ejpam-1676	197	1	let	let	VERB
ejpam-1676	197	2	bθ2	bθ2	NOUN
ejpam-1676	197	3	denote	denote	VERB
ejpam-1676	197	4	the	the	DET
ejpam-1676	197	5	estimate	estimate	NOUN
ejpam-1676	197	6	derived	derive	VERB
ejpam-1676	197	7	from	from	ADP
ejpam-1676	197	8	the	the	DET
ejpam-1676	197	9	second	second	ADJ
ejpam-1676	197	10	sample	sample	NOUN
ejpam-1676	197	11	.	.	PUNCT
ejpam-1676	198	1	we	we	PRON
ejpam-1676	198	2	consider	consider	VERB
ejpam-1676	198	3	the	the	DET
ejpam-1676	198	4	estimated	estimate	VERB
ejpam-1676	198	5	pairwise	pairwise	NOUN
ejpam-1676	198	6	likelihood	likelihood	NOUN
ejpam-1676	198	7	ratio	ratio	NOUN
ejpam-1676	198	8	(	(	PUNCT
ejpam-1676	198	9	plr	plr	NOUN
ejpam-1676	198	10	)	)	PUNCT
ejpam-1676	198	11	as	as	SCONJ
ejpam-1676	198	12	follows	follow	VERB
ejpam-1676	198	13	:	:	PUNCT
ejpam-1676	199	1	γi	γi	ADP
ejpam-1676	199	2	=	=	PUNCT
ejpam-1676	199	3	p	p	X
ejpam-1676	199	4	lri	lri	PROPN
ejpam-1676	199	5	=	=	SYM
ejpam-1676	199	6	n1∏	n1∏	PROPN
ejpam-1676	199	7	j=1	j=1	PROPN
ejpam-1676	199	8	f1(y1	f1(y1	PROPN
ejpam-1676	199	9	j	j	PROPN
ejpam-1676	199	10	;	;	PUNCT
ejpam-1676	199	11	bθi	bθi	PROPN
ejpam-1676	199	12	)	)	PUNCT
ejpam-1676	199	13	n1∏	n1∏	PROPN
ejpam-1676	199	14	j=1	j=1	PROPN
ejpam-1676	199	15	f1(y1	f1(y1	PROPN
ejpam-1676	199	16	j	j	PROPN
ejpam-1676	199	17	;	;	PUNCT
ejpam-1676	199	18	bθ1	bθ1	PROPN
ejpam-1676	199	19	)	)	PUNCT
ejpam-1676	199	20	,	,	PUNCT
ejpam-1676	200	1	i	i	NOUN
ejpam-1676	200	2	=	=	NOUN
ejpam-1676	200	3	1,2	1,2	NUM
ejpam-1676	200	4	.	.	PUNCT
ejpam-1676	201	1	it	it	PRON
ejpam-1676	201	2	then	then	ADV
ejpam-1676	201	3	follows	follow	VERB
ejpam-1676	201	4	that	that	SCONJ
ejpam-1676	201	5	γ1	γ1	PROPN
ejpam-1676	201	6	equals	equal	VERB
ejpam-1676	201	7	to	to	ADP
ejpam-1676	201	8	1	1	NUM
ejpam-1676	201	9	.	.	PUNCT
ejpam-1676	202	1	furthermore	furthermore	ADV
ejpam-1676	202	2	,	,	PUNCT
ejpam-1676	202	3	we	we	PRON
ejpam-1676	202	4	have	have	VERB
ejpam-1676	202	5	γi	γi	NOUN
ejpam-1676	202	6	≤	≤	NUM
ejpam-1676	202	7	1	1	NUM
ejpam-1676	202	8	,	,	PUNCT
ejpam-1676	202	9	i	i	PRON
ejpam-1676	202	10	=	=	NOUN
ejpam-1676	202	11	2	2	NUM
ejpam-1676	202	12	,	,	PUNCT
ejpam-1676	202	13	.	.	PUNCT
ejpam-1676	202	14	.	.	PUNCT
ejpam-1676	202	15	.	.	PUNCT
ejpam-1676	203	1	,	,	PUNCT
ejpam-1676	203	2	m	m	VERB
ejpam-1676	203	3	by	by	ADP
ejpam-1676	203	4	the	the	DET
ejpam-1676	203	5	definition	definition	NOUN
ejpam-1676	203	6	of	of	ADP
ejpam-1676	203	7	the	the	DET
ejpam-1676	203	8	mle	mle	PROPN
ejpam-1676	203	9	.	.	PUNCT
ejpam-1676	204	1	the	the	DET
ejpam-1676	204	2	likelihood	likelihood	NOUN
ejpam-1676	204	3	weights	weight	VERB
ejpam-1676	204	4	λ1	λ1	ADJ
ejpam-1676	204	5	and	and	CCONJ
ejpam-1676	204	6	λ2	λ2	NOUN
ejpam-1676	204	7	will	will	AUX
ejpam-1676	204	8	then	then	ADV
ejpam-1676	204	9	be	be	AUX
ejpam-1676	204	10	chosen	choose	VERB
ejpam-1676	204	11	as	as	ADP
ejpam-1676	204	12	λ1	λ1	PROPN
ejpam-1676	204	13	=	=	SYM
ejpam-1676	204	14	1	1	NUM
ejpam-1676	204	15	1	1	NUM
ejpam-1676	204	16	+	+	NUM
ejpam-1676	204	17	γ2	γ2	ADJ
ejpam-1676	204	18	,	,	PUNCT
ejpam-1676	204	19	λ2	λ2	PROPN
ejpam-1676	204	20	=	=	SYM
ejpam-1676	204	21	γ2	γ2	PROPN
ejpam-1676	204	22	1	1	NUM
ejpam-1676	204	23	+	+	NOUN
ejpam-1676	204	24	γ2	γ2	ADJ
ejpam-1676	204	25	.	.	PUNCT
ejpam-1676	205	1	for	for	ADP
ejpam-1676	205	2	the	the	DET
ejpam-1676	205	3	simple	simple	ADJ
ejpam-1676	205	4	linear	linear	NOUN
ejpam-1676	205	5	models	model	NOUN
ejpam-1676	205	6	presented	present	VERB
ejpam-1676	205	7	before	before	ADV
ejpam-1676	205	8	,	,	PUNCT
ejpam-1676	205	9	a	a	DET
ejpam-1676	205	10	simplification	simplification	NOUN
ejpam-1676	205	11	yields	yield	VERB
ejpam-1676	205	12	that	that	PRON
ejpam-1676	205	13	the	the	DET
ejpam-1676	205	14	wle	wle	NOUN
ejpam-1676	205	15	of	of	ADP
ejpam-1676	205	16	β1	β1	PROPN
ejpam-1676	205	17	takes	take	VERB
ejpam-1676	205	18	the	the	DET
ejpam-1676	205	19	following	follow	VERB
ejpam-1676	205	20	form	form	NOUN
ejpam-1676	205	21	:	:	PUNCT
ejpam-1676	205	22	bβwle	bβwle	ADJ
ejpam-1676	205	23	1	1	NUM
ejpam-1676	205	24	=	=	SYM
ejpam-1676	205	25	w1	w1	NOUN
ejpam-1676	205	26	bβmle	bβmle	NOUN
ejpam-1676	205	27	1	1	NUM
ejpam-1676	206	1	+	+	NOUN
ejpam-1676	206	2	w2	w2	NOUN
ejpam-1676	206	3	bβmle	bβmle	NOUN
ejpam-1676	206	4	2	2	NUM
ejpam-1676	206	5	,	,	PUNCT
ejpam-1676	206	6	p.	p.	NOUN
ejpam-1676	206	7	guo	guo	PROPN
ejpam-1676	206	8	,	,	PUNCT
ejpam-1676	206	9	x.	x.	PROPN
ejpam-1676	206	10	wang	wang	PROPN
ejpam-1676	206	11	,	,	PUNCT
ejpam-1676	206	12	y.	y.	PROPN
ejpam-1676	206	13	wu	wu	PROPN
ejpam-1676	206	14	/	/	SYM
ejpam-1676	206	15	eur	eur	PROPN
ejpam-1676	206	16	.	.	PUNCT
ejpam-1676	207	1	j.	j.	PROPN
ejpam-1676	207	2	pure	pure	PROPN
ejpam-1676	207	3	appl	appl	PROPN
ejpam-1676	207	4	.	.	PROPN
ejpam-1676	207	5	math	math	PROPN
ejpam-1676	207	6	,	,	PUNCT
ejpam-1676	207	7	5	5	NUM
ejpam-1676	207	8	(	(	PUNCT
ejpam-1676	207	9	2012	2012	NUM
ejpam-1676	207	10	)	)	PUNCT
ejpam-1676	207	11	,	,	PUNCT
ejpam-1676	207	12	333	333	NUM
ejpam-1676	207	13	-	-	SYM
ejpam-1676	207	14	356	356	NUM
ejpam-1676	207	15	340	340	NUM
ejpam-1676	208	1	where	where	SCONJ
ejpam-1676	208	2	w1	w1	NOUN
ejpam-1676	208	3	=	=	SYM
ejpam-1676	208	4	n1∑	n1∑	PROPN
ejpam-1676	208	5	j=1	j=1	PROPN
ejpam-1676	209	1	x2	x2	PROPN
ejpam-1676	209	2	1	1	NUM
ejpam-1676	210	1	j	j	PROPN
ejpam-1676	210	2	n1∑	n1∑	PROPN
ejpam-1676	210	3	j=1	j=1	PROPN
ejpam-1676	211	1	x2	x2	PROPN
ejpam-1676	211	2	1	1	NUM
ejpam-1676	211	3	j	j	NOUN
ejpam-1676	211	4	+	+	CCONJ
ejpam-1676	211	5	γ2	γ2	ADJ
ejpam-1676	211	6	n2∑	n2∑	PROPN
ejpam-1676	211	7	k=1	k=1	PUNCT
ejpam-1676	212	1	x2	x2	PROPN
ejpam-1676	212	2	2k	2k	NOUN
ejpam-1676	212	3	and	and	CCONJ
ejpam-1676	212	4	w2	w2	NOUN
ejpam-1676	212	5	=	=	SYM
ejpam-1676	212	6	γ2	γ2	ADJ
ejpam-1676	212	7	n2∑	n2∑	PROPN
ejpam-1676	212	8	k=1	k=1	PUNCT
ejpam-1676	213	1	x2	x2	PROPN
ejpam-1676	213	2	2k	2k	NOUN
ejpam-1676	214	1	n1∑	n1∑	PROPN
ejpam-1676	214	2	j=1	j=1	PROPN
ejpam-1676	215	1	x2	x2	PROPN
ejpam-1676	215	2	1	1	NUM
ejpam-1676	215	3	j	j	NOUN
ejpam-1676	215	4	+	+	CCONJ
ejpam-1676	215	5	γ2	γ2	ADJ
ejpam-1676	215	6	n2∑	n2∑	PROPN
ejpam-1676	215	7	k=1	k=1	PUNCT
ejpam-1676	216	1	x2	x2	PROPN
ejpam-1676	216	2	2k	2k	PROPN
ejpam-1676	216	3	.	.	PUNCT
ejpam-1676	217	1	to	to	PART
ejpam-1676	217	2	further	far	ADV
ejpam-1676	217	3	illustrate	illustrate	VERB
ejpam-1676	217	4	the	the	DET
ejpam-1676	217	5	behavior	behavior	NOUN
ejpam-1676	217	6	of	of	ADP
ejpam-1676	217	7	the	the	DET
ejpam-1676	217	8	proposed	propose	VERB
ejpam-1676	217	9	likelihood	likelihood	NOUN
ejpam-1676	217	10	weights	weight	NOUN
ejpam-1676	217	11	,	,	PUNCT
ejpam-1676	217	12	we	we	PRON
ejpam-1676	217	13	revisit	revisit	VERB
ejpam-1676	217	14	the	the	DET
ejpam-1676	217	15	simple	simple	ADJ
ejpam-1676	217	16	linear	linear	NOUN
ejpam-1676	217	17	models	model	NOUN
ejpam-1676	217	18	given	give	VERB
ejpam-1676	217	19	by	by	ADP
ejpam-1676	217	20	equation	equation	NOUN
ejpam-1676	217	21	(	(	PUNCT
ejpam-1676	217	22	1	1	NUM
ejpam-1676	217	23	)	)	PUNCT
ejpam-1676	217	24	.	.	PUNCT
ejpam-1676	218	1	it	it	PRON
ejpam-1676	218	2	then	then	ADV
ejpam-1676	218	3	follows	follow	VERB
ejpam-1676	218	4	that	that	SCONJ
ejpam-1676	218	5	γ2	γ2	NOUN
ejpam-1676	218	6	=	=	PUNCT
ejpam-1676	218	7	exp	exp	NOUN
ejpam-1676	218	8			NOUN
ejpam-1676	218	9			VERB
ejpam-1676	218	10	−	−	PROPN
ejpam-1676	218	11	1	1	NUM
ejpam-1676	218	12	2σ2	2σ2	NUM
ejpam-1676	218	13	1	1	NUM
ejpam-1676	218	14	n1∑	n1∑	PROPN
ejpam-1676	218	15	j=1	j=1	PROPN
ejpam-1676	218	16	(	(	PUNCT
ejpam-1676	218	17	y1	y1	INTJ
ejpam-1676	218	18	j	j	PROPN
ejpam-1676	218	19	−	−	PROPN
ejpam-1676	218	20	bβmle	bβmle	VERB
ejpam-1676	218	21	2	2	NUM
ejpam-1676	218	22	x1	x1	PROPN
ejpam-1676	218	23	j	j	PROPN
ejpam-1676	218	24	)	)	PUNCT
ejpam-1676	218	25	2	2	NUM
ejpam-1676	219	1	+	+	SYM
ejpam-1676	219	2	1	1	NUM
ejpam-1676	219	3	2σ2	2σ2	NUM
ejpam-1676	219	4	1	1	NUM
ejpam-1676	219	5	n1∑	n1∑	PROPN
ejpam-1676	219	6	j=1	j=1	PROPN
ejpam-1676	219	7	(	(	PUNCT
ejpam-1676	219	8	y1	y1	INTJ
ejpam-1676	219	9	j	j	PROPN
ejpam-1676	219	10	−	−	PROPN
ejpam-1676	219	11	bβmle	bβmle	VERB
ejpam-1676	219	12	1	1	NUM
ejpam-1676	219	13	x1	x1	PROPN
ejpam-1676	219	14	j	j	PROPN
ejpam-1676	219	15	)	)	PUNCT
ejpam-1676	219	16	2	2	NUM
ejpam-1676	219	17			PROPN
ejpam-1676	219	18			PROPN
ejpam-1676	219	19			NOUN
ejpam-1676	219	20	=	=	SYM
ejpam-1676	219	21	exp	exp	NOUN
ejpam-1676	219	22			NOUN
ejpam-1676	219	23			ADJ
ejpam-1676	219	24	−(bβ	−(bβ	ADV
ejpam-1676	219	25	mle	mle	PROPN
ejpam-1676	219	26	1	1	NUM
ejpam-1676	219	27	−	−	NOUN
ejpam-1676	219	28	bβmle	bβmle	NOUN
ejpam-1676	219	29	2	2	NUM
ejpam-1676	219	30	)	)	PUNCT
ejpam-1676	219	31	2	2	NUM
ejpam-1676	220	1	n1∑	n1∑	PROPN
ejpam-1676	220	2	j=1	j=1	PROPN
ejpam-1676	221	1	x2	x2	PROPN
ejpam-1676	221	2	1	1	NUM
ejpam-1676	221	3	j/(2σ	j/(2σ	NOUN
ejpam-1676	221	4	2	2	NUM
ejpam-1676	221	5	1	1	NUM
ejpam-1676	221	6	)	)	PUNCT
ejpam-1676	221	7			PROPN
ejpam-1676	221	8			PROPN
ejpam-1676	221	9			NOUN
ejpam-1676	221	10	.	.	PUNCT
ejpam-1676	222	1	since	since	SCONJ
ejpam-1676	222	2	λ2	λ2	NOUN
ejpam-1676	222	3	=	=	SYM
ejpam-1676	222	4	1/(1	1/(1	NUM
ejpam-1676	222	5	+	+	SYM
ejpam-1676	222	6	γ2	γ2	ADJ
ejpam-1676	222	7	)	)	PUNCT
ejpam-1676	222	8	,	,	PUNCT
ejpam-1676	222	9	we	we	PRON
ejpam-1676	222	10	then	then	ADV
ejpam-1676	222	11	have	have	VERB
ejpam-1676	222	12	λ2	λ2	NOUN
ejpam-1676	222	13	=	=	SYM
ejpam-1676	223	1	1	1	NUM
ejpam-1676	223	2	1	1	NUM
ejpam-1676	223	3	+	+	NUM
ejpam-1676	223	4	exp	exp	NOUN
ejpam-1676	223	5	(	(	PUNCT
ejpam-1676	223	6	(	(	PUNCT
ejpam-1676	223	7	bβmle	bβmle	NOUN
ejpam-1676	223	8	1	1	NUM
ejpam-1676	223	9	−	−	NOUN
ejpam-1676	223	10	bβmle	bβmle	NOUN
ejpam-1676	223	11	2	2	NUM
ejpam-1676	223	12	)	)	SYM
ejpam-1676	223	13	2	2	NUM
ejpam-1676	223	14	n1∑	n1∑	PROPN
ejpam-1676	223	15	j=1	j=1	PROPN
ejpam-1676	224	1	x2	x2	PROPN
ejpam-1676	224	2	1	1	NUM
ejpam-1676	224	3	j	j	NOUN
ejpam-1676	224	4	/(2σ2	/(2σ2	NOUN
ejpam-1676	224	5	1	1	NUM
ejpam-1676	224	6	)	)	PUNCT
ejpam-1676	224	7	)	)	PUNCT
ejpam-1676	224	8	.	.	PUNCT
ejpam-1676	225	1	thus	thus	ADV
ejpam-1676	225	2	the	the	DET
ejpam-1676	225	3	proposed	propose	VERB
ejpam-1676	225	4	likelihood	likelihood	NOUN
ejpam-1676	225	5	weight	weight	NOUN
ejpam-1676	225	6	for	for	ADP
ejpam-1676	225	7	the	the	DET
ejpam-1676	225	8	second	second	ADJ
ejpam-1676	225	9	sample	sample	NOUN
ejpam-1676	225	10	is	be	AUX
ejpam-1676	225	11	a	a	DET
ejpam-1676	225	12	function	function	NOUN
ejpam-1676	225	13	of	of	ADP
ejpam-1676	225	14	the	the	DET
ejpam-1676	225	15	estimated	estimate	VERB
ejpam-1676	225	16	difference	difference	NOUN
ejpam-1676	225	17	between	between	ADP
ejpam-1676	225	18	the	the	DET
ejpam-1676	225	19	regression	regression	NOUN
ejpam-1676	225	20	coefficients	coefficient	NOUN
ejpam-1676	225	21	,	,	PUNCT
ejpam-1676	225	22	the	the	DET
ejpam-1676	225	23	variance	variance	NOUN
ejpam-1676	225	24	of	of	ADP
ejpam-1676	225	25	the	the	DET
ejpam-1676	225	26	error	error	NOUN
ejpam-1676	225	27	terms	term	NOUN
ejpam-1676	225	28	and	and	CCONJ
ejpam-1676	225	29	the	the	DET
ejpam-1676	225	30	“	"	PUNCT
ejpam-1676	225	31	variance	variance	NOUN
ejpam-1676	225	32	”	"	PUNCT
ejpam-1676	225	33	of	of	ADP
ejpam-1676	225	34	the	the	DET
ejpam-1676	225	35	covariate	covariate	NOUN
ejpam-1676	225	36	.	.	PUNCT
ejpam-1676	226	1	for	for	ADP
ejpam-1676	226	2	a	a	DET
ejpam-1676	226	3	fixed	fix	VERB
ejpam-1676	226	4	design	design	NOUN
ejpam-1676	226	5	matrix	matrix	NOUN
ejpam-1676	226	6	,	,	PUNCT
ejpam-1676	226	7	a	a	DET
ejpam-1676	226	8	large	large	ADJ
ejpam-1676	226	9	difference	difference	NOUN
ejpam-1676	226	10	between	between	ADP
ejpam-1676	226	11	bβmle	bβmle	NOUN
ejpam-1676	226	12	1	1	NUM
ejpam-1676	226	13	and	and	CCONJ
ejpam-1676	226	14	bβmle	bβmle	NOUN
ejpam-1676	226	15	2	2	NUM
ejpam-1676	226	16	will	will	AUX
ejpam-1676	226	17	reduce	reduce	VERB
ejpam-1676	226	18	the	the	DET
ejpam-1676	226	19	assigned	assign	VERB
ejpam-1676	226	20	importance	importance	NOUN
ejpam-1676	226	21	to	to	ADP
ejpam-1676	226	22	the	the	DET
ejpam-1676	226	23	second	second	ADJ
ejpam-1676	226	24	sample	sample	NOUN
ejpam-1676	226	25	.	.	PUNCT
ejpam-1676	227	1	on	on	ADP
ejpam-1676	227	2	the	the	DET
ejpam-1676	227	3	other	other	ADJ
ejpam-1676	227	4	hand	hand	NOUN
ejpam-1676	227	5	,	,	PUNCT
ejpam-1676	227	6	a	a	DET
ejpam-1676	227	7	large	large	ADJ
ejpam-1676	227	8	variance	variance	NOUN
ejpam-1676	227	9	in	in	ADP
ejpam-1676	227	10	the	the	DET
ejpam-1676	227	11	first	first	ADJ
ejpam-1676	227	12	model	model	NOUN
ejpam-1676	227	13	would	would	AUX
ejpam-1676	227	14	result	result	VERB
ejpam-1676	227	15	in	in	ADP
ejpam-1676	227	16	a	a	DET
ejpam-1676	227	17	bigger	big	ADJ
ejpam-1676	227	18	value	value	NOUN
ejpam-1676	227	19	for	for	ADP
ejpam-1676	227	20	the	the	DET
ejpam-1676	227	21	weight	weight	NOUN
ejpam-1676	227	22	assigned	assign	VERB
ejpam-1676	227	23	to	to	ADP
ejpam-1676	227	24	the	the	DET
ejpam-1676	227	25	second	second	ADJ
ejpam-1676	227	26	sample	sample	NOUN
ejpam-1676	227	27	.	.	PUNCT
ejpam-1676	228	1	2.3	2.3	NUM
ejpam-1676	228	2	.	.	PUNCT
ejpam-1676	228	3	estimation	estimation	NOUN
ejpam-1676	228	4	with	with	ADP
ejpam-1676	228	5	fixed	fix	VERB
ejpam-1676	228	6	likelihood	likelihood	NOUN
ejpam-1676	228	7	weights	weight	NOUN
ejpam-1676	228	8	in	in	ADP
ejpam-1676	228	9	this	this	DET
ejpam-1676	228	10	section	section	NOUN
ejpam-1676	228	11	we	we	PRON
ejpam-1676	228	12	develop	develop	VERB
ejpam-1676	228	13	the	the	DET
ejpam-1676	228	14	wl	wl	NOUN
ejpam-1676	228	15	procedure	procedure	NOUN
ejpam-1676	228	16	for	for	ADP
ejpam-1676	228	17	the	the	DET
ejpam-1676	228	18	linear	linear	ADJ
ejpam-1676	228	19	models	model	NOUN
ejpam-1676	228	20	as	as	ADP
ejpam-1676	228	21	an	an	DET
ejpam-1676	228	22	illustration	illustration	NOUN
ejpam-1676	228	23	.	.	PUNCT
ejpam-1676	229	1	for	for	ADP
ejpam-1676	229	2	simplicity	simplicity	NOUN
ejpam-1676	229	3	,	,	PUNCT
ejpam-1676	229	4	we	we	PRON
ejpam-1676	229	5	assume	assume	VERB
ejpam-1676	229	6	that	that	SCONJ
ejpam-1676	229	7	m	m	VERB
ejpam-1676	229	8	=	=	SYM
ejpam-1676	229	9	2	2	NUM
ejpam-1676	229	10	in	in	ADP
ejpam-1676	229	11	sections	section	NOUN
ejpam-1676	229	12	2.1	2.1	NUM
ejpam-1676	229	13	and	and	CCONJ
ejpam-1676	229	14	2.2	2.2	NUM
ejpam-1676	229	15	.	.	PUNCT
ejpam-1676	230	1	extension	extension	NOUN
ejpam-1676	230	2	to	to	ADP
ejpam-1676	230	3	the	the	DET
ejpam-1676	230	4	case	case	NOUN
ejpam-1676	230	5	of	of	ADP
ejpam-1676	230	6	m	m	PROPN
ejpam-1676	230	7	>	>	X
ejpam-1676	230	8	2	2	NUM
ejpam-1676	230	9	is	be	AUX
ejpam-1676	230	10	straightforward	straightforward	ADJ
ejpam-1676	230	11	.	.	PUNCT
ejpam-1676	231	1	assume	assume	VERB
ejpam-1676	231	2	that	that	SCONJ
ejpam-1676	231	3	data	datum	NOUN
ejpam-1676	231	4	{	{	PUNCT
ejpam-1676	231	5	x1	x1	PROPN
ejpam-1676	231	6	j	j	PROPN
ejpam-1676	231	7	,	,	PUNCT
ejpam-1676	231	8	y1	y1	PROPN
ejpam-1676	231	9	j	j	PROPN
ejpam-1676	231	10	;	;	PUNCT
ejpam-1676	231	11	j	j	PROPN
ejpam-1676	231	12	=	=	SYM
ejpam-1676	231	13	1	1	NUM
ejpam-1676	231	14	,	,	PUNCT
ejpam-1676	231	15	.	.	PUNCT
ejpam-1676	231	16	.	.	PUNCT
ejpam-1676	231	17	.	.	PUNCT
ejpam-1676	232	1	,	,	PUNCT
ejpam-1676	232	2	n1	n1	NOUN
ejpam-1676	232	3	}	}	PUNCT
ejpam-1676	232	4	and	and	CCONJ
ejpam-1676	232	5	{	{	PUNCT
ejpam-1676	232	6	x2k	x2k	NOUN
ejpam-1676	232	7	,	,	PUNCT
ejpam-1676	232	8	y2k	y2k	NOUN
ejpam-1676	232	9	;	;	PUNCT
ejpam-1676	232	10	k	k	PROPN
ejpam-1676	232	11	=	=	SYM
ejpam-1676	232	12	1	1	NUM
ejpam-1676	232	13	,	,	PUNCT
ejpam-1676	232	14	.	.	PUNCT
ejpam-1676	232	15	.	.	PUNCT
ejpam-1676	232	16	.	.	PUNCT
ejpam-1676	233	1	,	,	PUNCT
ejpam-1676	233	2	n2	n2	PROPN
ejpam-1676	233	3	}	}	PUNCT
ejpam-1676	233	4	are	be	AUX
ejpam-1676	233	5	generated	generate	VERB
ejpam-1676	233	6	from	from	ADP
ejpam-1676	233	7	the	the	DET
ejpam-1676	233	8	following	follow	VERB
ejpam-1676	233	9	models	model	NOUN
ejpam-1676	233	10	:	:	PUNCT
ejpam-1676	233	11	y1	y1	INTJ
ejpam-1676	233	12	j	j	X
ejpam-1676	233	13	=	=	SYM
ejpam-1676	234	1	β1	β1	PROPN
ejpam-1676	234	2	x1	x1	PROPN
ejpam-1676	234	3	j	j	PROPN
ejpam-1676	235	1	+	+	CCONJ
ejpam-1676	235	2	ε	ε	PROPN
ejpam-1676	235	3	j	j	PROPN
ejpam-1676	235	4	,	,	PUNCT
ejpam-1676	235	5	ε	ε	PROPN
ejpam-1676	235	6	j	j	PROPN
ejpam-1676	235	7	∼	∼	NOUN
ejpam-1676	235	8	n(0,σ2	n(0,σ2	PROPN
ejpam-1676	235	9	1	1	NUM
ejpam-1676	235	10	)	)	PUNCT
ejpam-1676	235	11	,	,	PUNCT
ejpam-1676	235	12	j	j	PROPN
ejpam-1676	235	13	=	=	SYM
ejpam-1676	235	14	1	1	NUM
ejpam-1676	235	15	,	,	PUNCT
ejpam-1676	235	16	.	.	PUNCT
ejpam-1676	235	17	.	.	PUNCT
ejpam-1676	236	1	.	.	PUNCT
ejpam-1676	237	1	,	,	PUNCT
ejpam-1676	237	2	n1	n1	NOUN
ejpam-1676	237	3	,	,	PUNCT
ejpam-1676	237	4	y2k	y2k	PROPN
ejpam-1676	237	5	=	=	SYM
ejpam-1676	237	6	β2	β2	VERB
ejpam-1676	238	1	x2	x2	PROPN
ejpam-1676	238	2	j	j	PROPN
ejpam-1676	239	1	+	+	CCONJ
ejpam-1676	239	2	ek	ek	PROPN
ejpam-1676	239	3	;	;	PUNCT
ejpam-1676	239	4	ek	ek	X
ejpam-1676	239	5	∼	∼	NOUN
ejpam-1676	239	6	n(0,σ2	n(0,σ2	PROPN
ejpam-1676	239	7	2	2	NUM
ejpam-1676	239	8	)	)	PUNCT
ejpam-1676	239	9	,	,	PUNCT
ejpam-1676	239	10	k	k	X
ejpam-1676	239	11	=	=	SYM
ejpam-1676	239	12	1	1	NUM
ejpam-1676	239	13	,	,	PUNCT
ejpam-1676	239	14	.	.	PUNCT
ejpam-1676	239	15	.	.	PUNCT
ejpam-1676	240	1	.	.	PUNCT
ejpam-1676	241	1	,	,	PUNCT
ejpam-1676	241	2	n2	n2	NOUN
ejpam-1676	241	3	,	,	PUNCT
ejpam-1676	241	4	(	(	PUNCT
ejpam-1676	241	5	8)	8)	NUM
ejpam-1676	241	6	where	where	SCONJ
ejpam-1676	241	7	the	the	DET
ejpam-1676	241	8	{	{	PUNCT
ejpam-1676	241	9	x1	x1	PROPN
ejpam-1676	241	10	j	j	PROPN
ejpam-1676	241	11	,	,	PUNCT
ejpam-1676	241	12	j	j	PROPN
ejpam-1676	241	13	=	=	NOUN
ejpam-1676	241	14	1	1	NUM
ejpam-1676	241	15	,	,	PUNCT
ejpam-1676	241	16	.	.	PUNCT
ejpam-1676	241	17	.	.	PUNCT
ejpam-1676	242	1	.	.	PUNCT
ejpam-1676	243	1	,	,	PUNCT
ejpam-1676	243	2	n1	n1	NOUN
ejpam-1676	243	3	;	;	PUNCT
ejpam-1676	243	4	x2	x2	PROPN
ejpam-1676	243	5	j	j	PROPN
ejpam-1676	243	6	,	,	PUNCT
ejpam-1676	243	7	k	k	PROPN
ejpam-1676	243	8	=	=	SYM
ejpam-1676	243	9	1	1	NUM
ejpam-1676	243	10	,	,	PUNCT
ejpam-1676	243	11	.	.	PUNCT
ejpam-1676	243	12	.	.	PUNCT
ejpam-1676	244	1	.	.	PUNCT
ejpam-1676	245	1	,	,	PUNCT
ejpam-1676	245	2	n2	n2	PROPN
ejpam-1676	245	3	}	}	PUNCT
ejpam-1676	245	4	are	be	AUX
ejpam-1676	245	5	fixed	fix	VERB
ejpam-1676	245	6	.	.	PUNCT
ejpam-1676	246	1	assume	assume	VERB
ejpam-1676	246	2	that	that	SCONJ
ejpam-1676	246	3	{	{	PUNCT
ejpam-1676	246	4	ε	ε	PROPN
ejpam-1676	246	5	j	j	PROPN
ejpam-1676	246	6	}	}	PUNCT
ejpam-1676	246	7	’s	’s	PART
ejpam-1676	246	8	and	and	CCONJ
ejpam-1676	246	9	ek	ek	PROPN
ejpam-1676	246	10	’s	’	NOUN
ejpam-1676	246	11	are	be	AUX
ejpam-1676	246	12	i.i.d	i.i.d	ADP
ejpam-1676	246	13	.	.	PUNCT
ejpam-1676	246	14	respectively	respectively	ADV
ejpam-1676	246	15	.	.	PUNCT
ejpam-1676	247	1	the	the	DET
ejpam-1676	247	2	error	error	NOUN
ejpam-1676	247	3	terms	term	NOUN
ejpam-1676	247	4	from	from	ADP
ejpam-1676	247	5	these	these	DET
ejpam-1676	247	6	two	two	NUM
ejpam-1676	247	7	models	model	NOUN
ejpam-1676	247	8	are	be	AUX
ejpam-1676	247	9	assumed	assume	VERB
ejpam-1676	247	10	to	to	PART
ejpam-1676	247	11	be	be	AUX
ejpam-1676	247	12	independent	independent	ADJ
ejpam-1676	247	13	as	as	ADV
ejpam-1676	247	14	well	well	ADV
ejpam-1676	247	15	.	.	PUNCT
ejpam-1676	248	1	for	for	ADP
ejpam-1676	248	2	the	the	DET
ejpam-1676	248	3	purpose	purpose	NOUN
ejpam-1676	248	4	of	of	ADP
ejpam-1676	248	5	our	our	PRON
ejpam-1676	248	6	demonstration	demonstration	NOUN
ejpam-1676	248	7	we	we	PRON
ejpam-1676	248	8	assume	assume	VERB
ejpam-1676	248	9	σ1	σ1	PROPN
ejpam-1676	248	10	and	and	CCONJ
ejpam-1676	248	11	σ2	σ2	PROPN
ejpam-1676	248	12	are	be	AUX
ejpam-1676	248	13	known	know	VERB
ejpam-1676	248	14	although	although	SCONJ
ejpam-1676	248	15	that	that	PRON
ejpam-1676	248	16	would	would	AUX
ejpam-1676	248	17	rarely	rarely	ADV
ejpam-1676	248	18	be	be	AUX
ejpam-1676	248	19	the	the	DET
ejpam-1676	248	20	case	case	NOUN
ejpam-1676	248	21	in	in	ADP
ejpam-1676	248	22	practice	practice	NOUN
ejpam-1676	248	23	.	.	PUNCT
ejpam-1676	249	1	denote	denote	VERB
ejpam-1676	249	2	y1	y1	INTJ
ejpam-1676	250	1	=	=	PUNCT
ejpam-1676	250	2	(	(	PUNCT
ejpam-1676	250	3	y11	y11	NOUN
ejpam-1676	250	4	,	,	PUNCT
ejpam-1676	250	5	y12	y12	PROPN
ejpam-1676	250	6	,	,	PUNCT
ejpam-1676	250	7	.	.	PUNCT
ejpam-1676	250	8	.	.	PUNCT
ejpam-1676	251	1	.	.	PUNCT
ejpam-1676	252	1	,	,	PUNCT
ejpam-1676	252	2	y1n1	y1n1	PROPN
ejpam-1676	252	3	)	)	PUNCT
ejpam-1676	252	4	t	t	PROPN
ejpam-1676	252	5	,	,	PUNCT
ejpam-1676	252	6	and	and	CCONJ
ejpam-1676	252	7	y2	y2	PROPN
ejpam-1676	253	1	=	=	SYM
ejpam-1676	253	2	(	(	PUNCT
ejpam-1676	253	3	y21	y21	PROPN
ejpam-1676	253	4	,	,	PUNCT
ejpam-1676	253	5	y22	y22	PROPN
ejpam-1676	253	6	,	,	PUNCT
ejpam-1676	253	7	.	.	PUNCT
ejpam-1676	253	8	.	.	PUNCT
ejpam-1676	254	1	.	.	PUNCT
ejpam-1676	255	1	,	,	PUNCT
ejpam-1676	255	2	y2n2	y2n2	NOUN
ejpam-1676	255	3	)	)	PUNCT
ejpam-1676	255	4	t.	t.	NOUN
ejpam-1676	255	5	p.	p.	PROPN
ejpam-1676	255	6	guo	guo	PROPN
ejpam-1676	255	7	,	,	PUNCT
ejpam-1676	255	8	x.	x.	PROPN
ejpam-1676	255	9	wang	wang	PROPN
ejpam-1676	255	10	,	,	PUNCT
ejpam-1676	255	11	y.	y.	PROPN
ejpam-1676	255	12	wu	wu	PROPN
ejpam-1676	255	13	/	/	SYM
ejpam-1676	255	14	eur	eur	PROPN
ejpam-1676	255	15	.	.	PUNCT
ejpam-1676	256	1	j.	j.	PROPN
ejpam-1676	256	2	pure	pure	PROPN
ejpam-1676	256	3	appl	appl	PROPN
ejpam-1676	256	4	.	.	PROPN
ejpam-1676	256	5	math	math	PROPN
ejpam-1676	256	6	,	,	PUNCT
ejpam-1676	256	7	5	5	NUM
ejpam-1676	256	8	(	(	PUNCT
ejpam-1676	256	9	2012	2012	NUM
ejpam-1676	256	10	)	)	PUNCT
ejpam-1676	256	11	,	,	PUNCT
ejpam-1676	256	12	333	333	NUM
ejpam-1676	256	13	-	-	SYM
ejpam-1676	256	14	356	356	NUM
ejpam-1676	256	15	341	341	NUM
ejpam-1676	256	16	suppose	suppose	VERB
ejpam-1676	256	17	that	that	SCONJ
ejpam-1676	256	18	the	the	DET
ejpam-1676	256	19	parameter	parameter	NOUN
ejpam-1676	256	20	,	,	PUNCT
ejpam-1676	256	21	β1	β1	PROPN
ejpam-1676	256	22	,	,	PUNCT
ejpam-1676	256	23	is	be	AUX
ejpam-1676	256	24	of	of	ADP
ejpam-1676	256	25	primary	primary	ADJ
ejpam-1676	256	26	inferential	inferential	ADJ
ejpam-1676	256	27	interest	interest	NOUN
ejpam-1676	256	28	.	.	PUNCT
ejpam-1676	257	1	the	the	DET
ejpam-1676	257	2	parameter	parameter	NOUN
ejpam-1676	257	3	β2	β2	PROPN
ejpam-1676	257	4	is	be	AUX
ejpam-1676	257	5	thought	think	VERB
ejpam-1676	257	6	or	or	CCONJ
ejpam-1676	257	7	expected	expect	VERB
ejpam-1676	257	8	to	to	PART
ejpam-1676	257	9	be	be	AUX
ejpam-1676	257	10	not	not	PART
ejpam-1676	257	11	too	too	ADV
ejpam-1676	257	12	different	different	ADJ
ejpam-1676	257	13	from	from	ADP
ejpam-1676	257	14	β1	β1	PROPN
ejpam-1676	257	15	due	due	ADP
ejpam-1676	257	16	to	to	ADP
ejpam-1676	257	17	the	the	DET
ejpam-1676	257	18	“	"	PUNCT
ejpam-1676	257	19	similarity	similarity	NOUN
ejpam-1676	257	20	”	"	PUNCT
ejpam-1676	257	21	of	of	ADP
ejpam-1676	257	22	the	the	DET
ejpam-1676	257	23	two	two	NUM
ejpam-1676	257	24	experiments	experiment	NOUN
ejpam-1676	257	25	.	.	PUNCT
ejpam-1676	258	1	note	note	VERB
ejpam-1676	258	2	that	that	SCONJ
ejpam-1676	258	3	a	a	DET
ejpam-1676	258	4	bivariate	bivariate	ADJ
ejpam-1676	258	5	distribution	distribution	NOUN
ejpam-1676	258	6	is	be	AUX
ejpam-1676	258	7	not	not	PART
ejpam-1676	258	8	assumed	assume	VERB
ejpam-1676	258	9	in	in	ADP
ejpam-1676	258	10	the	the	DET
ejpam-1676	258	11	above	above	ADJ
ejpam-1676	258	12	model	model	NOUN
ejpam-1676	258	13	.	.	PUNCT
ejpam-1676	259	1	assuming	assume	VERB
ejpam-1676	259	2	marginal	marginal	ADJ
ejpam-1676	259	3	normality	normality	NOUN
ejpam-1676	259	4	,	,	PUNCT
ejpam-1676	259	5	the	the	DET
ejpam-1676	259	6	marginal	marginal	ADJ
ejpam-1676	259	7	likelihoods	likelihood	NOUN
ejpam-1676	259	8	for	for	ADP
ejpam-1676	259	9	β1	β1	PROPN
ejpam-1676	259	10	and	and	CCONJ
ejpam-1676	259	11	β2	β2	NOUN
ejpam-1676	259	12	are	be	AUX
ejpam-1676	259	13	l1(y1;β1)∝	l1(y1;β1)∝	ADJ
ejpam-1676	259	14	n1∏	n1∏	NOUN
ejpam-1676	259	15	j=1	j=1	PROPN
ejpam-1676	259	16	exp	exp	NOUN
ejpam-1676	259	17	¨	¨	NOUN
ejpam-1676	259	18	−	−	PROPN
ejpam-1676	260	1	(	(	PUNCT
ejpam-1676	260	2	y1	y1	INTJ
ejpam-1676	260	3	j	j	PROPN
ejpam-1676	260	4	−	−	PROPN
ejpam-1676	261	1	β1	β1	PROPN
ejpam-1676	261	2	x1	x1	PROPN
ejpam-1676	261	3	j	j	PROPN
ejpam-1676	261	4	)	)	PUNCT
ejpam-1676	261	5	2	2	NUM
ejpam-1676	261	6	2	2	NUM
ejpam-1676	261	7	σ2	σ2	NOUN
ejpam-1676	261	8	1	1	NUM
ejpam-1676	261	9	«	«	PUNCT
ejpam-1676	261	10	,	,	PUNCT
ejpam-1676	261	11	and	and	CCONJ
ejpam-1676	261	12	l2(y2;β2)∝	l2(y2;β2)∝	PROPN
ejpam-1676	261	13	n2∏	n2∏	VERB
ejpam-1676	261	14	k=1	k=1	X
ejpam-1676	261	15	exp	exp	NOUN
ejpam-1676	261	16	¨	¨	NOUN
ejpam-1676	261	17	−	−	PROPN
ejpam-1676	261	18	(	(	PUNCT
ejpam-1676	261	19	y2k	y2k	PROPN
ejpam-1676	261	20	−	−	PROPN
ejpam-1676	261	21	β2	β2	NOUN
ejpam-1676	261	22	x2k	x2k	NOUN
ejpam-1676	261	23	)	)	PUNCT
ejpam-1676	261	24	2	2	NUM
ejpam-1676	261	25	2	2	NUM
ejpam-1676	261	26	σ2	σ2	NOUN
ejpam-1676	261	27	2	2	NUM
ejpam-1676	261	28	«	«	PUNCT
ejpam-1676	261	29	.	.	PUNCT
ejpam-1676	262	1	a	a	DET
ejpam-1676	262	2	direct	direct	ADJ
ejpam-1676	262	3	calculation	calculation	NOUN
ejpam-1676	262	4	deduces	deduce	VERB
ejpam-1676	262	5	the	the	DET
ejpam-1676	262	6	mle	mle	NOUN
ejpam-1676	262	7	of	of	ADP
ejpam-1676	262	8	β1	β1	PROPN
ejpam-1676	262	9	and	and	CCONJ
ejpam-1676	262	10	β2	β2	PROPN
ejpam-1676	262	11	as	as	SCONJ
ejpam-1676	262	12	follows	follow	VERB
ejpam-1676	262	13	bβmle	bβmle	NOUN
ejpam-1676	262	14	1	1	NUM
ejpam-1676	262	15	=	=	SYM
ejpam-1676	262	16			NOUN
ejpam-1676	263	1			ADJ
ejpam-1676	263	2	n1∑	n1∑	PROPN
ejpam-1676	263	3	j=1	j=1	PROPN
ejpam-1676	264	1	x2	x2	PROPN
ejpam-1676	264	2	1	1	NUM
ejpam-1676	264	3	j	j	PROPN
ejpam-1676	264	4			PROPN
ejpam-1676	265	1			NOUN
ejpam-1676	265	2	−1	−1	VERB
ejpam-1676	266	1	n1∑	n1∑	PROPN
ejpam-1676	266	2	j=1	j=1	PROPN
ejpam-1676	267	1	x1	x1	PROPN
ejpam-1676	267	2	j	j	PROPN
ejpam-1676	267	3	y1	y1	PROPN
ejpam-1676	267	4	j	j	PROPN
ejpam-1676	267	5	,	,	PUNCT
ejpam-1676	267	6	and	and	CCONJ
ejpam-1676	267	7	bβmle	bβmle	NOUN
ejpam-1676	267	8	2	2	X
ejpam-1676	267	9	=	=	SYM
ejpam-1676	267	10	n2∑	n2∑	PROPN
ejpam-1676	267	11	k=1	k=1	PUNCT
ejpam-1676	268	1	x2	x2	PROPN
ejpam-1676	268	2	2k	2k	PROPN
ejpam-1676	268	3	!	!	PUNCT
ejpam-1676	269	1	−1	−1	NOUN
ejpam-1676	269	2	n2∑	n2∑	PROPN
ejpam-1676	269	3	k=1	k=1	PUNCT
ejpam-1676	270	1	x2k	x2k	PROPN
ejpam-1676	270	2	y2k	y2k	PROPN
ejpam-1676	270	3	.	.	PUNCT
ejpam-1676	271	1	if	if	SCONJ
ejpam-1676	271	2	one	one	PRON
ejpam-1676	271	3	knows	know	VERB
ejpam-1676	271	4	that	that	SCONJ
ejpam-1676	271	5	β1	β1	PROPN
ejpam-1676	271	6	and	and	CCONJ
ejpam-1676	271	7	β2	β2	NOUN
ejpam-1676	271	8	are	be	AUX
ejpam-1676	271	9	similar	similar	ADJ
ejpam-1676	271	10	to	to	ADP
ejpam-1676	271	11	each	each	DET
ejpam-1676	271	12	other	other	ADJ
ejpam-1676	271	13	according	accord	VERB
ejpam-1676	271	14	to	to	ADP
ejpam-1676	271	15	past	past	ADJ
ejpam-1676	271	16	studies	study	NOUN
ejpam-1676	271	17	or	or	CCONJ
ejpam-1676	271	18	expert	expert	NOUN
ejpam-1676	271	19	opinions	opinion	NOUN
ejpam-1676	271	20	,	,	PUNCT
ejpam-1676	271	21	then	then	ADV
ejpam-1676	271	22	it	it	PRON
ejpam-1676	271	23	is	be	AUX
ejpam-1676	271	24	reasonable	reasonable	ADJ
ejpam-1676	271	25	to	to	PART
ejpam-1676	271	26	expect	expect	VERB
ejpam-1676	271	27	that	that	SCONJ
ejpam-1676	271	28	this	this	DET
ejpam-1676	271	29	information	information	NOUN
ejpam-1676	271	30	might	might	AUX
ejpam-1676	271	31	be	be	AUX
ejpam-1676	271	32	used	use	VERB
ejpam-1676	271	33	to	to	PART
ejpam-1676	271	34	yield	yield	VERB
ejpam-1676	271	35	a	a	DET
ejpam-1676	271	36	better	well	ADJ
ejpam-1676	271	37	estimate	estimate	NOUN
ejpam-1676	271	38	of	of	ADP
ejpam-1676	271	39	the	the	DET
ejpam-1676	271	40	parameter	parameter	NOUN
ejpam-1676	271	41	β1	β1	PROPN
ejpam-1676	271	42	.	.	PUNCT
ejpam-1676	272	1	the	the	DET
ejpam-1676	272	2	wl	wl	PROPN
ejpam-1676	272	3	for	for	ADP
ejpam-1676	272	4	inference	inference	NOUN
ejpam-1676	272	5	about	about	ADP
ejpam-1676	272	6	β1	β1	PROPN
ejpam-1676	272	7	can	can	AUX
ejpam-1676	272	8	be	be	AUX
ejpam-1676	272	9	represented	represent	VERB
ejpam-1676	272	10	as	as	ADP
ejpam-1676	272	11	:	:	PUNCT
ejpam-1676	272	12	wl(β1	wl(β1	PROPN
ejpam-1676	272	13	;	;	PUNCT
ejpam-1676	272	14	y1	y1	NOUN
ejpam-1676	272	15	,	,	PUNCT
ejpam-1676	272	16	y2	y2	NOUN
ejpam-1676	272	17	)	)	PUNCT
ejpam-1676	272	18	=	=	PUNCT
ejpam-1676	273	1	l	l	NOUN
ejpam-1676	273	2	λ1	λ1	ADJ
ejpam-1676	273	3	1	1	NUM
ejpam-1676	273	4	(	(	PUNCT
ejpam-1676	273	5	y1;β1)l	y1;β1)l	NOUN
ejpam-1676	273	6	λ2	λ2	NOUN
ejpam-1676	273	7	2	2	NUM
ejpam-1676	273	8	(	(	PUNCT
ejpam-1676	273	9	y2;β1	y2;β1	NOUN
ejpam-1676	273	10	)	)	PUNCT
ejpam-1676	273	11	,	,	PUNCT
ejpam-1676	273	12	(	(	PUNCT
ejpam-1676	273	13	9	9	X
ejpam-1676	273	14	)	)	PUNCT
ejpam-1676	273	15	where	where	SCONJ
ejpam-1676	273	16	λ1	λ1	ADJ
ejpam-1676	273	17	and	and	CCONJ
ejpam-1676	273	18	λ2	λ2	NOUN
ejpam-1676	273	19	are	be	AUX
ejpam-1676	273	20	weights	weight	NOUN
ejpam-1676	273	21	selected	select	VERB
ejpam-1676	273	22	according	accord	VERB
ejpam-1676	273	23	to	to	ADP
ejpam-1676	273	24	the	the	DET
ejpam-1676	273	25	relevance	relevance	NOUN
ejpam-1676	273	26	of	of	ADP
ejpam-1676	273	27	the	the	DET
ejpam-1676	273	28	likelihood	likelihood	NOUN
ejpam-1676	273	29	to	to	PART
ejpam-1676	273	30	which	which	PRON
ejpam-1676	273	31	they	they	PRON
ejpam-1676	273	32	attached	attach	VERB
ejpam-1676	273	33	.	.	PUNCT
ejpam-1676	274	1	the	the	DET
ejpam-1676	274	2	non	non	ADJ
ejpam-1676	274	3	-	-	ADJ
ejpam-1676	274	4	negative	negative	ADJ
ejpam-1676	274	5	requirement	requirement	NOUN
ejpam-1676	274	6	for	for	ADP
ejpam-1676	274	7	the	the	DET
ejpam-1676	274	8	weights	weight	NOUN
ejpam-1676	274	9	is	be	AUX
ejpam-1676	274	10	not	not	PART
ejpam-1676	274	11	assumed	assume	VERB
ejpam-1676	274	12	in	in	ADP
ejpam-1676	274	13	the	the	DET
ejpam-1676	274	14	formulation	formulation	NOUN
ejpam-1676	274	15	although	although	SCONJ
ejpam-1676	274	16	the	the	DET
ejpam-1676	274	17	optimum	optimum	ADJ
ejpam-1676	274	18	weights	weight	NOUN
ejpam-1676	274	19	should	should	AUX
ejpam-1676	274	20	be	be	AUX
ejpam-1676	274	21	non	non	ADJ
ejpam-1676	274	22	-	-	ADJ
ejpam-1676	274	23	negative	negative	ADJ
ejpam-1676	274	24	according	accord	VERB
ejpam-1676	274	25	to	to	ADP
ejpam-1676	274	26	the	the	DET
ejpam-1676	274	27	assertion	assertion	NOUN
ejpam-1676	274	28	by	by	ADP
ejpam-1676	274	29	[	[	X
ejpam-1676	274	30	6	6	NUM
ejpam-1676	274	31	]	]	PUNCT
ejpam-1676	274	32	.	.	PUNCT
ejpam-1676	275	1	it	it	PRON
ejpam-1676	275	2	is	be	AUX
ejpam-1676	275	3	worth	worth	ADJ
ejpam-1676	275	4	noting	note	VERB
ejpam-1676	275	5	that	that	SCONJ
ejpam-1676	275	6	l2(y2;β1	l2(y2;β1	PROPN
ejpam-1676	275	7	)	)	PUNCT
ejpam-1676	275	8	instead	instead	ADV
ejpam-1676	275	9	of	of	ADP
ejpam-1676	275	10	l2(y2;β2	l2(y2;β2	NOUN
ejpam-1676	275	11	)	)	PUNCT
ejpam-1676	275	12	is	be	AUX
ejpam-1676	275	13	used	use	VERB
ejpam-1676	275	14	to	to	PART
ejpam-1676	275	15	define	define	VERB
ejpam-1676	275	16	wl(β1	wl(β1	NOUN
ejpam-1676	275	17	)	)	PUNCT
ejpam-1676	275	18	since	since	SCONJ
ejpam-1676	275	19	β1	β1	PROPN
ejpam-1676	275	20	is	be	AUX
ejpam-1676	275	21	of	of	ADP
ejpam-1676	275	22	our	our	PRON
ejpam-1676	275	23	primary	primary	ADJ
ejpam-1676	275	24	inferential	inferential	ADJ
ejpam-1676	275	25	interest	interest	NOUN
ejpam-1676	275	26	at	at	ADP
ejpam-1676	275	27	this	this	DET
ejpam-1676	275	28	stage	stage	NOUN
ejpam-1676	275	29	and	and	CCONJ
ejpam-1676	275	30	the	the	DET
ejpam-1676	275	31	marginal	marginal	ADJ
ejpam-1676	275	32	distributions	distribution	NOUN
ejpam-1676	275	33	of	of	ADP
ejpam-1676	275	34	the	the	DET
ejpam-1676	275	35	elements	element	NOUN
ejpam-1676	275	36	of	of	ADP
ejpam-1676	275	37	the	the	DET
ejpam-1676	275	38	y	y	PROPN
ejpam-1676	275	39	2	2	NUM
ejpam-1676	275	40	are	be	AUX
ejpam-1676	275	41	thought	think	VERB
ejpam-1676	275	42	to	to	PART
ejpam-1676	275	43	resemble	resemble	VERB
ejpam-1676	275	44	the	the	DET
ejpam-1676	275	45	marginal	marginal	ADJ
ejpam-1676	275	46	distributions	distribution	NOUN
ejpam-1676	275	47	of	of	ADP
ejpam-1676	275	48	those	those	PRON
ejpam-1676	275	49	of	of	ADP
ejpam-1676	275	50	the	the	DET
ejpam-1676	275	51	y1	y1	PROPN
ejpam-1676	275	52	.	.	PUNCT
ejpam-1676	276	1	note	note	VERB
ejpam-1676	276	2	that	that	SCONJ
ejpam-1676	276	3	the	the	DET
ejpam-1676	276	4	wl	wl	PROPN
ejpam-1676	276	5	for	for	ADP
ejpam-1676	276	6	θ1	θ1	NOUN
ejpam-1676	276	7	depends	depend	VERB
ejpam-1676	276	8	on	on	ADP
ejpam-1676	276	9	the	the	DET
ejpam-1676	276	10	distribution	distribution	NOUN
ejpam-1676	276	11	of	of	ADP
ejpam-1676	276	12	the	the	DET
ejpam-1676	276	13	y	y	PROPN
ejpam-1676	276	14	1	1	NUM
ejpam-1676	276	15	.	.	PUNCT
ejpam-1676	277	1	however	however	ADV
ejpam-1676	277	2	,	,	PUNCT
ejpam-1676	277	3	it	it	PRON
ejpam-1676	277	4	does	do	AUX
ejpam-1676	277	5	not	not	PART
ejpam-1676	277	6	depend	depend	VERB
ejpam-1676	277	7	on	on	ADP
ejpam-1676	277	8	the	the	DET
ejpam-1676	277	9	distribution	distribution	NOUN
ejpam-1676	277	10	of	of	ADP
ejpam-1676	277	11	y	y	PROPN
ejpam-1676	277	12	2	2	NUM
ejpam-1676	277	13	.	.	PUNCT
ejpam-1676	277	14	notice	notice	VERB
ejpam-1676	277	15	that	that	SCONJ
ejpam-1676	277	16	the	the	DET
ejpam-1676	277	17	joint	joint	ADJ
ejpam-1676	277	18	distribution	distribution	NOUN
ejpam-1676	277	19	of	of	ADP
ejpam-1676	277	20	the	the	DET
ejpam-1676	277	21	y	y	PROPN
ejpam-1676	277	22	1	1	NUM
ejpam-1676	277	23	and	and	CCONJ
ejpam-1676	277	24	y	y	PROPN
ejpam-1676	277	25	2	2	NUM
ejpam-1676	277	26	does	do	AUX
ejpam-1676	277	27	not	not	PART
ejpam-1676	277	28	appear	appear	VERB
ejpam-1676	277	29	in	in	ADP
ejpam-1676	277	30	the	the	DET
ejpam-1676	277	31	formulation	formulation	NOUN
ejpam-1676	277	32	of	of	ADP
ejpam-1676	277	33	the	the	DET
ejpam-1676	277	34	wl	wl	PROPN
ejpam-1676	277	35	for	for	ADP
ejpam-1676	277	36	θ1	θ1	NOUN
ejpam-1676	277	37	and	and	CCONJ
ejpam-1676	277	38	no	no	DET
ejpam-1676	277	39	assumptions	assumption	NOUN
ejpam-1676	277	40	are	be	AUX
ejpam-1676	277	41	made	make	VERB
ejpam-1676	277	42	about	about	ADP
ejpam-1676	277	43	it	it	PRON
ejpam-1676	277	44	.	.	PUNCT
ejpam-1676	278	1	the	the	DET
ejpam-1676	278	2	wle	wle	NOUN
ejpam-1676	278	3	of	of	ADP
ejpam-1676	278	4	β1	β1	PROPN
ejpam-1676	278	5	is	be	AUX
ejpam-1676	278	6	obtained	obtain	VERB
ejpam-1676	278	7	by	by	ADP
ejpam-1676	278	8	maximizing	maximize	VERB
ejpam-1676	278	9	the	the	DET
ejpam-1676	278	10	weighted	weight	VERB
ejpam-1676	278	11	likelihood	likelihood	NOUN
ejpam-1676	278	12	function	function	NOUN
ejpam-1676	278	13	for	for	ADP
ejpam-1676	278	14	given	give	VERB
ejpam-1676	278	15	weights	weight	NOUN
ejpam-1676	278	16	λ1	λ1	ADJ
ejpam-1676	278	17	and	and	CCONJ
ejpam-1676	278	18	λ2	λ2	NOUN
ejpam-1676	278	19	.	.	PUNCT
ejpam-1676	279	1	it	it	PRON
ejpam-1676	279	2	follows	follow	VERB
ejpam-1676	279	3	from	from	ADP
ejpam-1676	279	4	(	(	PUNCT
ejpam-1676	279	5	9	9	NUM
ejpam-1676	279	6	)	)	PUNCT
ejpam-1676	279	7	that	that	PRON
ejpam-1676	279	8	log{wl(β1	log{wl(β1	VERB
ejpam-1676	279	9	)	)	PUNCT
ejpam-1676	279	10	}	}	PUNCT
ejpam-1676	280	1	=	=	SYM
ejpam-1676	280	2	λ1	λ1	ADJ
ejpam-1676	280	3	log{l1(y1;β1)}+λ2	log{l1(y1;β1)}+λ2	ADJ
ejpam-1676	280	4	log{l2(y2;β1	log{l2(y2;β1	NOUN
ejpam-1676	280	5	)	)	PUNCT
ejpam-1676	280	6	}	}	PUNCT
ejpam-1676	280	7	.	.	PUNCT
ejpam-1676	281	1	note	note	VERB
ejpam-1676	281	2	that	that	SCONJ
ejpam-1676	281	3	∂	∂	NUM
ejpam-1676	281	4	log{wl(β1	log{wl(β1	NOUN
ejpam-1676	281	5	)	)	PUNCT
ejpam-1676	281	6	}	}	PUNCT
ejpam-1676	281	7	∂	∂	NUM
ejpam-1676	282	1	β1	β1	NOUN
ejpam-1676	282	2	=	=	SYM
ejpam-1676	282	3	λ1	λ1	PROPN
ejpam-1676	282	4	2σ2	2σ2	NUM
ejpam-1676	282	5	1	1	NUM
ejpam-1676	282	6	n1∑	n1∑	PROPN
ejpam-1676	282	7	j=1	j=1	PROPN
ejpam-1676	283	1	x1	x1	PROPN
ejpam-1676	283	2	j(y1	j(y1	PROPN
ejpam-1676	283	3	j	j	PROPN
ejpam-1676	284	1	−	−	PROPN
ejpam-1676	284	2	β1	β1	PROPN
ejpam-1676	284	3	x1	x1	PROPN
ejpam-1676	284	4	j)+	j)+	PROPN
ejpam-1676	284	5	λ2	λ2	NOUN
ejpam-1676	284	6	2σ2	2σ2	NUM
ejpam-1676	284	7	1	1	NUM
ejpam-1676	284	8	n2∑	n2∑	PROPN
ejpam-1676	284	9	k=1	k=1	X
ejpam-1676	284	10	x2k(y2k	x2k(y2k	PROPN
ejpam-1676	284	11	−β1	−β1	PROPN
ejpam-1676	284	12	x2k	x2k	NOUN
ejpam-1676	284	13	)	)	PUNCT
ejpam-1676	284	14	.	.	PUNCT
ejpam-1676	285	1	p.	p.	NOUN
ejpam-1676	285	2	guo	guo	PROPN
ejpam-1676	285	3	,	,	PUNCT
ejpam-1676	285	4	x.	x.	PROPN
ejpam-1676	285	5	wang	wang	PROPN
ejpam-1676	285	6	,	,	PUNCT
ejpam-1676	285	7	y.	y.	PROPN
ejpam-1676	285	8	wu	wu	PROPN
ejpam-1676	285	9	/	/	SYM
ejpam-1676	285	10	eur	eur	PROPN
ejpam-1676	285	11	.	.	PUNCT
ejpam-1676	286	1	j.	j.	PROPN
ejpam-1676	286	2	pure	pure	PROPN
ejpam-1676	286	3	appl	appl	PROPN
ejpam-1676	286	4	.	.	PROPN
ejpam-1676	286	5	math	math	PROPN
ejpam-1676	286	6	,	,	PUNCT
ejpam-1676	286	7	5	5	NUM
ejpam-1676	286	8	(	(	PUNCT
ejpam-1676	286	9	2012	2012	NUM
ejpam-1676	286	10	)	)	PUNCT
ejpam-1676	286	11	,	,	PUNCT
ejpam-1676	286	12	333	333	NUM
ejpam-1676	286	13	-	-	SYM
ejpam-1676	286	14	356	356	NUM
ejpam-1676	286	15	342	342	NUM
ejpam-1676	286	16	a	a	DET
ejpam-1676	286	17	simplification	simplification	NOUN
ejpam-1676	286	18	yields	yield	VERB
ejpam-1676	286	19	the	the	DET
ejpam-1676	286	20	wle	wle	NOUN
ejpam-1676	286	21	of	of	ADP
ejpam-1676	286	22	β1	β1	PROPN
ejpam-1676	286	23	as	as	SCONJ
ejpam-1676	286	24	follows	follow	VERB
ejpam-1676	286	25	bβwle	bβwle	ADJ
ejpam-1676	286	26	1	1	NUM
ejpam-1676	286	27	=	=	SYM
ejpam-1676	286	28	w1	w1	NOUN
ejpam-1676	286	29	bβmle	bβmle	NOUN
ejpam-1676	286	30	1	1	NUM
ejpam-1676	287	1	+	+	NOUN
ejpam-1676	287	2	w2	w2	NOUN
ejpam-1676	287	3	bβmle	bβmle	NOUN
ejpam-1676	287	4	2	2	NUM
ejpam-1676	287	5	.	.	PUNCT
ejpam-1676	288	1	(	(	PUNCT
ejpam-1676	288	2	10	10	NUM
ejpam-1676	288	3	)	)	PUNCT
ejpam-1676	288	4	where	where	SCONJ
ejpam-1676	289	1	w1	w1	NOUN
ejpam-1676	289	2	=	=	SYM
ejpam-1676	289	3	λ1	λ1	PROPN
ejpam-1676	289	4	n1∑	n1∑	PROPN
ejpam-1676	289	5	j=1	j=1	PROPN
ejpam-1676	290	1	x2	x2	PROPN
ejpam-1676	290	2	1	1	NUM
ejpam-1676	290	3	j	j	NOUN
ejpam-1676	290	4	λ1	λ1	PROPN
ejpam-1676	290	5	n1∑	n1∑	PROPN
ejpam-1676	290	6	j=1	j=1	PROPN
ejpam-1676	291	1	x2	x2	PROPN
ejpam-1676	291	2	1	1	NUM
ejpam-1676	291	3	j	j	PROPN
ejpam-1676	292	1	+	+	ADJ
ejpam-1676	292	2	λ2	λ2	NOUN
ejpam-1676	292	3	n2∑	n2∑	PROPN
ejpam-1676	292	4	k=1	k=1	PUNCT
ejpam-1676	293	1	x2	x2	PROPN
ejpam-1676	293	2	2k	2k	NOUN
ejpam-1676	293	3	and	and	CCONJ
ejpam-1676	293	4	w2	w2	NOUN
ejpam-1676	293	5	=	=	SYM
ejpam-1676	293	6	λ2	λ2	PROPN
ejpam-1676	293	7	n2∑	n2∑	PROPN
ejpam-1676	293	8	k=1	k=1	PUNCT
ejpam-1676	294	1	x2	x2	PROPN
ejpam-1676	294	2	2k	2k	NOUN
ejpam-1676	294	3	λ1	λ1	VERB
ejpam-1676	294	4	n1∑	n1∑	PROPN
ejpam-1676	294	5	j=1	j=1	PROPN
ejpam-1676	295	1	x2	x2	PROPN
ejpam-1676	295	2	1	1	NUM
ejpam-1676	295	3	j	j	PROPN
ejpam-1676	296	1	+	+	ADJ
ejpam-1676	296	2	λ2	λ2	NOUN
ejpam-1676	296	3	n2∑	n2∑	PROPN
ejpam-1676	296	4	k=1	k=1	PUNCT
ejpam-1676	297	1	x2	x2	PROPN
ejpam-1676	297	2	2k	2k	PROPN
ejpam-1676	297	3	.	.	PUNCT
ejpam-1676	298	1	it	it	PRON
ejpam-1676	298	2	can	can	AUX
ejpam-1676	298	3	be	be	AUX
ejpam-1676	298	4	seen	see	VERB
ejpam-1676	298	5	that	that	SCONJ
ejpam-1676	298	6	the	the	DET
ejpam-1676	298	7	wle	wle	NOUN
ejpam-1676	298	8	of	of	ADP
ejpam-1676	298	9	β1	β1	PROPN
ejpam-1676	298	10	is	be	AUX
ejpam-1676	298	11	a	a	DET
ejpam-1676	298	12	linear	linear	ADJ
ejpam-1676	298	13	combination	combination	NOUN
ejpam-1676	298	14	of	of	ADP
ejpam-1676	298	15	bβmle	bβmle	NOUN
ejpam-1676	298	16	1	1	NUM
ejpam-1676	298	17	and	and	CCONJ
ejpam-1676	298	18	bβmle	bβmle	NOUN
ejpam-1676	298	19	2	2	NUM
ejpam-1676	298	20	,	,	PUNCT
ejpam-1676	298	21	under	under	ADP
ejpam-1676	298	22	the	the	DET
ejpam-1676	298	23	oversimplified	oversimplified	ADJ
ejpam-1676	298	24	model	model	NOUN
ejpam-1676	298	25	.	.	PUNCT
ejpam-1676	299	1	the	the	DET
ejpam-1676	299	2	wle	wle	NOUN
ejpam-1676	299	3	of	of	ADP
ejpam-1676	299	4	β1	β1	PROPN
ejpam-1676	299	5	coincides	coincide	VERB
ejpam-1676	299	6	with	with	ADP
ejpam-1676	299	7	the	the	DET
ejpam-1676	299	8	mle	mle	NOUN
ejpam-1676	299	9	of	of	ADP
ejpam-1676	299	10	β1	β1	PROPN
ejpam-1676	299	11	obtained	obtain	VERB
ejpam-1676	299	12	from	from	ADP
ejpam-1676	299	13	the	the	DET
ejpam-1676	299	14	first	first	ADJ
ejpam-1676	299	15	if	if	SCONJ
ejpam-1676	299	16	the	the	DET
ejpam-1676	299	17	weight	weight	NOUN
ejpam-1676	299	18	for	for	ADP
ejpam-1676	299	19	the	the	DET
ejpam-1676	299	20	second	second	ADJ
ejpam-1676	299	21	marginal	marginal	ADJ
ejpam-1676	299	22	likelihood	likelihood	NOUN
ejpam-1676	299	23	function	function	NOUN
ejpam-1676	299	24	is	be	AUX
ejpam-1676	299	25	set	set	VERB
ejpam-1676	299	26	to	to	PART
ejpam-1676	299	27	be	be	AUX
ejpam-1676	299	28	zero	zero	NUM
ejpam-1676	299	29	.	.	PUNCT
ejpam-1676	300	1	therefore	therefore	ADV
ejpam-1676	300	2	,	,	PUNCT
ejpam-1676	300	3	the	the	DET
ejpam-1676	300	4	weights	weight	NOUN
ejpam-1676	300	5	w1	w1	NOUN
ejpam-1676	300	6	and	and	CCONJ
ejpam-1676	300	7	w2	w2	NOUN
ejpam-1676	300	8	reflect	reflect	VERB
ejpam-1676	300	9	the	the	DET
ejpam-1676	300	10	importance	importance	NOUN
ejpam-1676	300	11	of	of	ADP
ejpam-1676	300	12	bβmle	bβmle	NOUN
ejpam-1676	300	13	1	1	NUM
ejpam-1676	300	14	and	and	CCONJ
ejpam-1676	300	15	bβmle	bβmle	NOUN
ejpam-1676	300	16	2	2	NUM
ejpam-1676	300	17	.	.	PUNCT
ejpam-1676	301	1	let	let	VERB
ejpam-1676	301	2	β0	β0	PROPN
ejpam-1676	301	3	1	1	NUM
ejpam-1676	301	4	and	and	CCONJ
ejpam-1676	301	5	β0	β0	NOUN
ejpam-1676	301	6	2	2	NUM
ejpam-1676	301	7	be	be	AUX
ejpam-1676	301	8	the	the	DET
ejpam-1676	301	9	true	true	ADJ
ejpam-1676	301	10	values	value	NOUN
ejpam-1676	301	11	of	of	ADP
ejpam-1676	301	12	the	the	DET
ejpam-1676	301	13	parameters	parameter	NOUN
ejpam-1676	301	14	β1	β1	PROPN
ejpam-1676	301	15	and	and	CCONJ
ejpam-1676	301	16	β2	β2	NOUN
ejpam-1676	301	17	respectively	respectively	ADV
ejpam-1676	301	18	.	.	PUNCT
ejpam-1676	302	1	if	if	SCONJ
ejpam-1676	302	2	one	one	PRON
ejpam-1676	302	3	knows	know	VERB
ejpam-1676	302	4	that	that	SCONJ
ejpam-1676	302	5	|β0	|β0	PRON
ejpam-1676	302	6	1	1	NUM
ejpam-1676	302	7	−β0	−β0	NOUN
ejpam-1676	302	8	2	2	NUM
ejpam-1676	303	1	|	|	ADV
ejpam-1676	303	2	≤	≤	NUM
ejpam-1676	303	3	c	c	NOUN
ejpam-1676	303	4	,	,	PUNCT
ejpam-1676	303	5	where	where	SCONJ
ejpam-1676	303	6	c	c	PROPN
ejpam-1676	303	7	is	be	AUX
ejpam-1676	303	8	a	a	DET
ejpam-1676	303	9	known	know	VERB
ejpam-1676	303	10	constant	constant	ADJ
ejpam-1676	303	11	according	accord	VERB
ejpam-1676	303	12	to	to	ADP
ejpam-1676	303	13	past	past	ADJ
ejpam-1676	303	14	studies	study	NOUN
ejpam-1676	303	15	or	or	CCONJ
ejpam-1676	303	16	expert	expert	NOUN
ejpam-1676	303	17	opinions	opinion	NOUN
ejpam-1676	303	18	,	,	PUNCT
ejpam-1676	303	19	we	we	PRON
ejpam-1676	303	20	then	then	ADV
ejpam-1676	303	21	have	have	VERB
ejpam-1676	303	22	the	the	DET
ejpam-1676	303	23	following	follow	VERB
ejpam-1676	303	24	theorem	theorem	VERB
ejpam-1676	303	25	.	.	PUNCT
ejpam-1676	303	26	theorem	theorem	NOUN
ejpam-1676	303	27	1	1	NUM
ejpam-1676	303	28	.	.	PUNCT
ejpam-1676	304	1	under	under	ADP
ejpam-1676	304	2	the	the	DET
ejpam-1676	304	3	normality	normality	NOUN
ejpam-1676	304	4	assumption	assumption	NOUN
ejpam-1676	304	5	with	with	ADP
ejpam-1676	304	6	known	know	VERB
ejpam-1676	304	7	variances	variance	NOUN
ejpam-1676	304	8	,	,	PUNCT
ejpam-1676	304	9	the	the	DET
ejpam-1676	304	10	wle	wle	PROPN
ejpam-1676	304	11	β̃wle	β̃wle	NOUN
ejpam-1676	304	12	1	1	NUM
ejpam-1676	304	13	takes	take	VERB
ejpam-1676	304	14	the	the	DET
ejpam-1676	304	15	form	form	NOUN
ejpam-1676	304	16	β̃wle	β̃wle	VERB
ejpam-1676	304	17	1	1	NUM
ejpam-1676	304	18	(	(	PUNCT
ejpam-1676	304	19	w1	w1	NOUN
ejpam-1676	304	20	,	,	PUNCT
ejpam-1676	304	21	w2	w2	NOUN
ejpam-1676	304	22	)	)	PUNCT
ejpam-1676	304	23	=	=	SYM
ejpam-1676	305	1	w1	w1	NOUN
ejpam-1676	305	2	bβmle	bβmle	NOUN
ejpam-1676	305	3	1	1	NUM
ejpam-1676	306	1	+	+	NOUN
ejpam-1676	306	2	w2	w2	NOUN
ejpam-1676	306	3	bβmle	bβmle	NOUN
ejpam-1676	306	4	2	2	NUM
ejpam-1676	306	5	,	,	PUNCT
ejpam-1676	306	6	where	where	SCONJ
ejpam-1676	306	7	w1	w1	NOUN
ejpam-1676	306	8	+	+	CCONJ
ejpam-1676	306	9	w2	w2	NOUN
ejpam-1676	306	10	=	=	SYM
ejpam-1676	306	11	1,0	1,0	NUM
ejpam-1676	306	12	<	<	X
ejpam-1676	306	13	w1	w1	NOUN
ejpam-1676	306	14	≤	≤	NUM
ejpam-1676	306	15	1,0	1,0	NUM
ejpam-1676	306	16	≤	≤	NOUN
ejpam-1676	306	17	w2	w2	NOUN
ejpam-1676	306	18	<	<	X
ejpam-1676	306	19	1	1	NUM
ejpam-1676	306	20	.	.	PUNCT
ejpam-1676	307	1	furthermore	furthermore	ADV
ejpam-1676	307	2	,	,	PUNCT
ejpam-1676	307	3	|β0	|β0	VERB
ejpam-1676	307	4	1	1	NUM
ejpam-1676	307	5	−	−	NOUN
ejpam-1676	307	6	β0	β0	NOUN
ejpam-1676	307	7	2	2	NUM
ejpam-1676	307	8	|	|	ADV
ejpam-1676	307	9	≤	≤	NUM
ejpam-1676	307	10	c	c	X
ejpam-1676	307	11	,	,	PUNCT
ejpam-1676	307	12	c	c	NOUN
ejpam-1676	307	13	>	>	X
ejpam-1676	307	14	0	0	NUM
ejpam-1676	307	15	,	,	PUNCT
ejpam-1676	307	16	then	then	ADV
ejpam-1676	307	17	|e(β̃1−	|e(β̃1−	VERB
ejpam-1676	307	18	β0	β0	PROPN
ejpam-1676	307	19	1	1	NUM
ejpam-1676	307	20	)	)	PUNCT
ejpam-1676	307	21	|	|	ADV
ejpam-1676	307	22	≤	≤	NUM
ejpam-1676	307	23	w2	w2	NOUN
ejpam-1676	307	24	c.	c.	PROPN
ejpam-1676	307	25	in	in	ADP
ejpam-1676	307	26	addition	addition	NOUN
ejpam-1676	307	27	,	,	PUNCT
ejpam-1676	307	28	max	max	PROPN
ejpam-1676	307	29	w1,w2	w1,w2	PROPN
ejpam-1676	307	30	mse(β̃wle	mse(β̃wle	ADV
ejpam-1676	307	31	1	1	NUM
ejpam-1676	307	32	)	)	PUNCT
ejpam-1676	307	33	<	<	X
ejpam-1676	307	34	mse(bβmle	mse(bβmle	NOUN
ejpam-1676	307	35	1	1	NUM
ejpam-1676	307	36	)	)	PUNCT
ejpam-1676	307	37	if	if	SCONJ
ejpam-1676	307	38	and	and	CCONJ
ejpam-1676	307	39	only	only	ADV
ejpam-1676	307	40	if	if	SCONJ
ejpam-1676	307	41	m	m	VERB
ejpam-1676	307	42	m	m	VERB
ejpam-1676	307	43	+	+	ADJ
ejpam-1676	307	44	1	1	NUM
ejpam-1676	307	45	<	<	X
ejpam-1676	307	46	w1	w1	NOUN
ejpam-1676	307	47	<	<	X
ejpam-1676	307	48	1	1	NUM
ejpam-1676	307	49	,	,	PUNCT
ejpam-1676	307	50	where	where	SCONJ
ejpam-1676	307	51	m	m	NOUN
ejpam-1676	307	52	=	=	VERB
ejpam-1676	307	53	c2	c2	PROPN
ejpam-1676	307	54	+	+	PROPN
ejpam-1676	307	55	σ2	σ2	PROPN
ejpam-1676	307	56	2	2	NUM
ejpam-1676	307	57	�	�	NOUN
ejpam-1676	307	58	n2∑	n2∑	PROPN
ejpam-1676	307	59	k=1	k=1	PUNCT
ejpam-1676	308	1	x2	x2	PROPN
ejpam-1676	308	2	2k	2k	PROPN
ejpam-1676	308	3	�	�	PROPN
ejpam-1676	308	4	−1	−1	NOUN
ejpam-1676	308	5	−σ2	−σ2	ADJ
ejpam-1676	308	6	1	1	NUM
ejpam-1676	308	7	n1∑	n1∑	PROPN
ejpam-1676	308	8	j=1	j=1	PROPN
ejpam-1676	309	1	x2	x2	PROPN
ejpam-1676	309	2	1	1	NUM
ejpam-1676	309	3	j	j	NOUN
ejpam-1676	309	4	!	!	PUNCT
ejpam-1676	309	5	−1	−1	NOUN
ejpam-1676	310	1	2σ2	2σ2	NUM
ejpam-1676	310	2	1	1	NUM
ejpam-1676	310	3	n1∑	n1∑	PROPN
ejpam-1676	310	4	j=1	j=1	PROPN
ejpam-1676	310	5	x2	x2	PROPN
ejpam-1676	310	6	1	1	NUM
ejpam-1676	310	7	j	j	NOUN
ejpam-1676	310	8	!	!	PUNCT
ejpam-1676	310	9	−1	−1	NOUN
ejpam-1676	310	10	.	.	PUNCT
ejpam-1676	311	1	in	in	ADP
ejpam-1676	311	2	addition	addition	NOUN
ejpam-1676	311	3	,	,	PUNCT
ejpam-1676	311	4	var(β̃wle	var(β̃wle	ADV
ejpam-1676	311	5	1	1	NUM
ejpam-1676	311	6	)	)	PUNCT
ejpam-1676	311	7	<	<	X
ejpam-1676	311	8	var(bβmle	var(bβmle	ADP
ejpam-1676	311	9	1	1	NUM
ejpam-1676	311	10	)	)	PUNCT
ejpam-1676	311	11	if	if	SCONJ
ejpam-1676	311	12	maxw1,w2	maxw1,w2	VERB
ejpam-1676	311	13	mse(β̃wle	mse(β̃wle	ADV
ejpam-1676	311	14	1	1	NUM
ejpam-1676	311	15	)	)	PUNCT
ejpam-1676	311	16	<	<	X
ejpam-1676	311	17	mse(bβmle	mse(bβmle	NOUN
ejpam-1676	311	18	1	1	NUM
ejpam-1676	311	19	)	)	PUNCT
ejpam-1676	311	20	.	.	PUNCT
ejpam-1676	312	1	thus	thus	ADV
ejpam-1676	312	2	,	,	PUNCT
ejpam-1676	312	3	for	for	ADP
ejpam-1676	312	4	small	small	ADJ
ejpam-1676	312	5	value	value	NOUN
ejpam-1676	312	6	of	of	ADP
ejpam-1676	312	7	c	c	PROPN
ejpam-1676	312	8	,	,	PUNCT
ejpam-1676	312	9	the	the	DET
ejpam-1676	312	10	wle	wle	NOUN
ejpam-1676	312	11	could	could	AUX
ejpam-1676	312	12	have	have	VERB
ejpam-1676	312	13	smaller	small	ADJ
ejpam-1676	312	14	mse	mse	NOUN
ejpam-1676	312	15	and	and	CCONJ
ejpam-1676	312	16	variance	variance	NOUN
ejpam-1676	312	17	simultaneously	simultaneously	ADV
ejpam-1676	312	18	with	with	ADP
ejpam-1676	312	19	the	the	DET
ejpam-1676	312	20	cost	cost	NOUN
ejpam-1676	312	21	of	of	ADP
ejpam-1676	312	22	small	small	ADJ
ejpam-1676	312	23	bias	bias	NOUN
ejpam-1676	312	24	.	.	PUNCT
ejpam-1676	313	1	the	the	DET
ejpam-1676	313	2	magnitude	magnitude	NOUN
ejpam-1676	313	3	of	of	ADP
ejpam-1676	313	4	the	the	DET
ejpam-1676	313	5	bias	bias	NOUN
ejpam-1676	313	6	is	be	AUX
ejpam-1676	313	7	controlled	control	VERB
ejpam-1676	313	8	by	by	ADP
ejpam-1676	313	9	the	the	DET
ejpam-1676	313	10	weight	weight	NOUN
ejpam-1676	313	11	assigned	assign	VERB
ejpam-1676	313	12	to	to	ADP
ejpam-1676	313	13	the	the	DET
ejpam-1676	313	14	second	second	ADJ
ejpam-1676	313	15	population	population	NOUN
ejpam-1676	313	16	.	.	PUNCT
ejpam-1676	314	1	the	the	DET
ejpam-1676	314	2	lower	lower	ADV
ejpam-1676	314	3	bound	bind	VERB
ejpam-1676	314	4	for	for	ADP
ejpam-1676	314	5	the	the	DET
ejpam-1676	314	6	weight	weight	NOUN
ejpam-1676	314	7	w1	w1	NOUN
ejpam-1676	314	8	is	be	AUX
ejpam-1676	314	9	connected	connect	VERB
ejpam-1676	314	10	with	with	ADP
ejpam-1676	314	11	the	the	DET
ejpam-1676	314	12	worst	bad	ADJ
ejpam-1676	314	13	mse	mse	NOUN
ejpam-1676	314	14	of	of	ADP
ejpam-1676	314	15	the	the	DET
ejpam-1676	314	16	wle	wle	PROPN
ejpam-1676	314	17	instead	instead	ADV
ejpam-1676	314	18	of	of	ADP
ejpam-1676	314	19	the	the	DET
ejpam-1676	314	20	optimal	optimal	ADJ
ejpam-1676	314	21	one	one	NOUN
ejpam-1676	314	22	.	.	PUNCT
ejpam-1676	315	1	however	however	ADV
ejpam-1676	315	2	,	,	PUNCT
ejpam-1676	315	3	this	this	DET
ejpam-1676	315	4	theorem	theorem	NOUN
ejpam-1676	315	5	does	do	AUX
ejpam-1676	315	6	not	not	PART
ejpam-1676	315	7	provide	provide	VERB
ejpam-1676	315	8	any	any	DET
ejpam-1676	315	9	guidance	guidance	NOUN
ejpam-1676	315	10	on	on	ADP
ejpam-1676	315	11	how	how	SCONJ
ejpam-1676	315	12	to	to	PART
ejpam-1676	315	13	choose	choose	VERB
ejpam-1676	315	14	the	the	DET
ejpam-1676	315	15	optimum	optimum	ADJ
ejpam-1676	315	16	weights	weight	NOUN
ejpam-1676	315	17	since	since	SCONJ
ejpam-1676	315	18	the	the	DET
ejpam-1676	315	19	bound	bound	ADJ
ejpam-1676	315	20	c	c	NOUN
ejpam-1676	315	21	is	be	AUX
ejpam-1676	315	22	rarely	rarely	ADV
ejpam-1676	315	23	known	know	VERB
ejpam-1676	315	24	in	in	ADP
ejpam-1676	315	25	practice	practice	NOUN
ejpam-1676	315	26	.	.	PUNCT
ejpam-1676	316	1	even	even	ADV
ejpam-1676	316	2	though	though	SCONJ
ejpam-1676	316	3	the	the	DET
ejpam-1676	316	4	exact	exact	ADJ
ejpam-1676	316	5	value	value	NOUN
ejpam-1676	316	6	of	of	ADP
ejpam-1676	316	7	the	the	DET
ejpam-1676	316	8	bound	bind	VERB
ejpam-1676	316	9	is	be	AUX
ejpam-1676	316	10	known	know	VERB
ejpam-1676	316	11	or	or	CCONJ
ejpam-1676	316	12	can	can	AUX
ejpam-1676	316	13	be	be	AUX
ejpam-1676	316	14	estimated	estimate	VERB
ejpam-1676	316	15	fairly	fairly	ADV
ejpam-1676	316	16	accurately	accurately	ADV
ejpam-1676	316	17	,	,	PUNCT
ejpam-1676	316	18	this	this	DET
ejpam-1676	316	19	theorem	theorem	NOUN
ejpam-1676	316	20	still	still	ADV
ejpam-1676	316	21	could	could	AUX
ejpam-1676	316	22	not	not	PART
ejpam-1676	316	23	be	be	AUX
ejpam-1676	316	24	used	use	VERB
ejpam-1676	316	25	to	to	PART
ejpam-1676	316	26	choose	choose	VERB
ejpam-1676	316	27	the	the	DET
ejpam-1676	316	28	best	good	ADJ
ejpam-1676	316	29	set	set	NOUN
ejpam-1676	316	30	of	of	ADP
ejpam-1676	316	31	weights	weight	NOUN
ejpam-1676	316	32	.	.	PUNCT
ejpam-1676	317	1	for	for	ADP
ejpam-1676	317	2	example	example	NOUN
ejpam-1676	317	3	,	,	PUNCT
ejpam-1676	317	4	suppose	suppose	VERB
ejpam-1676	317	5	that	that	SCONJ
ejpam-1676	317	6	there	there	PRON
ejpam-1676	317	7	are	be	VERB
ejpam-1676	317	8	two	two	NUM
ejpam-1676	317	9	independent	independent	ADJ
ejpam-1676	317	10	experiments	experiment	NOUN
ejpam-1676	317	11	with	with	ADP
ejpam-1676	317	12	exactly	exactly	ADV
ejpam-1676	317	13	the	the	DET
ejpam-1676	317	14	same	same	ADJ
ejpam-1676	317	15	coefficients	coefficient	NOUN
ejpam-1676	317	16	and	and	CCONJ
ejpam-1676	317	17	design	design	NOUN
ejpam-1676	317	18	matrices	matrix	NOUN
ejpam-1676	317	19	.	.	PUNCT
ejpam-1676	318	1	therefore	therefore	ADV
ejpam-1676	318	2	,	,	PUNCT
ejpam-1676	318	3	we	we	PRON
ejpam-1676	318	4	should	should	AUX
ejpam-1676	318	5	simply	simply	ADV
ejpam-1676	318	6	merge	merge	VERB
ejpam-1676	318	7	these	these	DET
ejpam-1676	318	8	two	two	NUM
ejpam-1676	318	9	experiments	experiment	NOUN
ejpam-1676	318	10	.	.	PUNCT
ejpam-1676	319	1	however	however	ADV
ejpam-1676	319	2	,	,	PUNCT
ejpam-1676	319	3	the	the	DET
ejpam-1676	319	4	above	above	ADJ
ejpam-1676	319	5	theorem	theorem	NOUN
ejpam-1676	319	6	merely	merely	ADV
ejpam-1676	319	7	suggests	suggest	VERB
ejpam-1676	319	8	that	that	SCONJ
ejpam-1676	319	9	the	the	DET
ejpam-1676	319	10	lower	lower	ADV
ejpam-1676	319	11	bound	bind	VERB
ejpam-1676	319	12	is	be	AUX
ejpam-1676	319	13	zero	zero	NUM
ejpam-1676	319	14	for	for	ADP
ejpam-1676	319	15	w1	w1	NOUN
ejpam-1676	319	16	since	since	SCONJ
ejpam-1676	319	17	m	m	PROPN
ejpam-1676	319	18	is	be	AUX
ejpam-1676	319	19	zero	zero	NUM
ejpam-1676	319	20	in	in	ADP
ejpam-1676	319	21	this	this	DET
ejpam-1676	319	22	case	case	NOUN
ejpam-1676	319	23	.	.	PUNCT
ejpam-1676	320	1	p.	p.	NOUN
ejpam-1676	320	2	guo	guo	PROPN
ejpam-1676	320	3	,	,	PUNCT
ejpam-1676	320	4	x.	x.	PROPN
ejpam-1676	320	5	wang	wang	PROPN
ejpam-1676	320	6	,	,	PUNCT
ejpam-1676	320	7	y.	y.	PROPN
ejpam-1676	320	8	wu	wu	PROPN
ejpam-1676	320	9	/	/	SYM
ejpam-1676	320	10	eur	eur	PROPN
ejpam-1676	320	11	.	.	PUNCT
ejpam-1676	321	1	j.	j.	PROPN
ejpam-1676	321	2	pure	pure	PROPN
ejpam-1676	321	3	appl	appl	PROPN
ejpam-1676	321	4	.	.	PROPN
ejpam-1676	321	5	math	math	PROPN
ejpam-1676	321	6	,	,	PUNCT
ejpam-1676	321	7	5	5	NUM
ejpam-1676	321	8	(	(	PUNCT
ejpam-1676	321	9	2012	2012	NUM
ejpam-1676	321	10	)	)	PUNCT
ejpam-1676	321	11	,	,	PUNCT
ejpam-1676	321	12	333	333	NUM
ejpam-1676	321	13	-	-	SYM
ejpam-1676	321	14	356	356	NUM
ejpam-1676	321	15	343	343	NUM
ejpam-1676	321	16	3	3	NUM
ejpam-1676	321	17	.	.	PUNCT
ejpam-1676	322	1	asymptotic	asymptotic	ADJ
ejpam-1676	322	2	properties	property	NOUN
ejpam-1676	322	3	in	in	ADP
ejpam-1676	322	4	this	this	DET
ejpam-1676	322	5	section	section	NOUN
ejpam-1676	322	6	we	we	PRON
ejpam-1676	322	7	investigate	investigate	VERB
ejpam-1676	322	8	the	the	DET
ejpam-1676	322	9	asymptotic	asymptotic	ADJ
ejpam-1676	322	10	properties	property	NOUN
ejpam-1676	322	11	of	of	ADP
ejpam-1676	322	12	the	the	DET
ejpam-1676	322	13	wle	wle	NOUN
ejpam-1676	322	14	in	in	ADP
ejpam-1676	322	15	the	the	DET
ejpam-1676	322	16	framework	framework	NOUN
ejpam-1676	322	17	of	of	ADP
ejpam-1676	322	18	generalized	generalized	ADJ
ejpam-1676	322	19	linear	linear	NOUN
ejpam-1676	322	20	models	model	NOUN
ejpam-1676	322	21	(	(	PUNCT
ejpam-1676	322	22	glm	glm	PROPN
ejpam-1676	322	23	)	)	PUNCT
ejpam-1676	322	24	for	for	ADP
ejpam-1676	322	25	simple	simple	ADJ
ejpam-1676	322	26	presentation	presentation	NOUN
ejpam-1676	322	27	.	.	PUNCT
ejpam-1676	323	1	the	the	DET
ejpam-1676	323	2	glms	glm	NOUN
ejpam-1676	323	3	have	have	VERB
ejpam-1676	323	4	the	the	DET
ejpam-1676	323	5	following	follow	VERB
ejpam-1676	323	6	structure	structure	NOUN
ejpam-1676	323	7	.	.	PUNCT
ejpam-1676	324	1	the	the	DET
ejpam-1676	324	2	sample	sample	NOUN
ejpam-1676	324	3	y	y	PROPN
ejpam-1676	324	4	=	=	SYM
ejpam-1676	324	5	(	(	PUNCT
ejpam-1676	324	6	y1	y1	PROPN
ejpam-1676	324	7	,	,	PUNCT
ejpam-1676	324	8	.	.	PUNCT
ejpam-1676	324	9	.	.	PUNCT
ejpam-1676	325	1	.	.	PUNCT
ejpam-1676	326	1	,	,	PUNCT
ejpam-1676	326	2	yn	yn	X
ejpam-1676	326	3	)	)	PUNCT
ejpam-1676	326	4	∈	∈	PROPN
ejpam-1676	327	1	ℜn	ℜn	PROPN
ejpam-1676	327	2	has	have	VERB
ejpam-1676	327	3	independent	independent	ADJ
ejpam-1676	327	4	components	component	NOUN
ejpam-1676	327	5	and	and	CCONJ
ejpam-1676	327	6	yi	yi	PROPN
ejpam-1676	327	7	has	have	VERB
ejpam-1676	327	8	the	the	DET
ejpam-1676	327	9	p.d.f	p.d.f	NOUN
ejpam-1676	327	10	.	.	PUNCT
ejpam-1676	328	1	exp	exp	PROPN
ejpam-1676	328	2	�	�	PROPN
ejpam-1676	328	3	ηi	ηi	PROPN
ejpam-1676	328	4	yi	yi	PROPN
ejpam-1676	328	5	−	−	PROPN
ejpam-1676	328	6	ζ(ηi	ζ(ηi	PROPN
ejpam-1676	328	7	)	)	PUNCT
ejpam-1676	328	8	φi	φi	ADP
ejpam-1676	328	9	�	�	PROPN
ejpam-1676	328	10	h(yi	h(yi	NOUN
ejpam-1676	328	11	,	,	PUNCT
ejpam-1676	328	12	φi	φi	ADP
ejpam-1676	328	13	)	)	PUNCT
ejpam-1676	328	14	,	,	PUNCT
ejpam-1676	328	15	.	.	PUNCT
ejpam-1676	328	16	.	.	PUNCT
ejpam-1676	328	17	.	.	PUNCT
ejpam-1676	329	1	,	,	PUNCT
ejpam-1676	329	2	n	n	CCONJ
ejpam-1676	329	3	,	,	PUNCT
ejpam-1676	329	4	w.r.t	w.r.t	NOUN
ejpam-1676	329	5	.	.	PUNCT
ejpam-1676	330	1	a	a	DET
ejpam-1676	330	2	σ	σ	VERB
ejpam-1676	330	3	-	-	PUNCT
ejpam-1676	330	4	finite	finite	ADJ
ejpam-1676	330	5	measure	measure	NOUN
ejpam-1676	330	6	ν	ν	NOUN
ejpam-1676	330	7	,	,	PUNCT
ejpam-1676	330	8	where	where	SCONJ
ejpam-1676	330	9	ηi	ηi	NOUN
ejpam-1676	330	10	and	and	CCONJ
ejpam-1676	330	11	φi	φi	ADV
ejpam-1676	330	12	are	be	AUX
ejpam-1676	330	13	unknown	unknown	ADJ
ejpam-1676	330	14	,	,	PUNCT
ejpam-1676	330	15	φi	φi	ADP
ejpam-1676	330	16	>	>	X
ejpam-1676	330	17	0	0	NUM
ejpam-1676	330	18	,	,	PUNCT
ejpam-1676	330	19	ηi	ηi	NOUN
ejpam-1676	330	20	∈	∈	PROPN
ejpam-1676	330	21	ξ	ξ	X
ejpam-1676	330	22	=	=	SYM
ejpam-1676	330	23	{	{	PUNCT
ejpam-1676	330	24	η	η	PROPN
ejpam-1676	330	25	:	:	PUNCT
ejpam-1676	330	26	0	0	NUM
ejpam-1676	330	27	<	<	X
ejpam-1676	330	28	∫	∫	PROPN
ejpam-1676	330	29	h(y	h(y	ADV
ejpam-1676	330	30	,	,	PUNCT
ejpam-1676	330	31	φ)eηy	φ)eηy	PROPN
ejpam-1676	330	32	/	/	SYM
ejpam-1676	330	33	φdν(y)<∞	φdν(y)<∞	PROPN
ejpam-1676	330	34	}	}	PUNCT
ejpam-1676	330	35	⊂	⊂	PROPN
ejpam-1676	330	36	ℜ	ℜ	PROPN
ejpam-1676	330	37	for	for	ADP
ejpam-1676	330	38	all	all	DET
ejpam-1676	330	39	i	i	PROPN
ejpam-1676	330	40	,	,	PUNCT
ejpam-1676	330	41	ζ	ζ	NOUN
ejpam-1676	330	42	and	and	CCONJ
ejpam-1676	330	43	h	h	NOUN
ejpam-1676	330	44	are	be	AUX
ejpam-1676	330	45	known	know	VERB
ejpam-1676	330	46	functions	function	NOUN
ejpam-1676	330	47	,	,	PUNCT
ejpam-1676	330	48	and	and	CCONJ
ejpam-1676	330	49	ζ′′(η	ζ′′(η	NOUN
ejpam-1676	330	50	)	)	PUNCT
ejpam-1676	330	51	>	>	X
ejpam-1676	330	52	0	0	PUNCT
ejpam-1676	330	53	is	be	AUX
ejpam-1676	330	54	assumed	assume	VERB
ejpam-1676	330	55	for	for	ADP
ejpam-1676	330	56	all	all	DET
ejpam-1676	330	57	η	η	PROPN
ejpam-1676	330	58	∈	∈	PROPN
ejpam-1676	330	59	ξ	ξ	PROPN
ejpam-1676	330	60	◦	◦	NOUN
ejpam-1676	330	61	,	,	PUNCT
ejpam-1676	330	62	the	the	DET
ejpam-1676	330	63	inferior	inferior	NOUN
ejpam-1676	330	64	of	of	ADP
ejpam-1676	330	65	ξ	ξ	PROPN
ejpam-1676	330	66	.	.	PUNCT
ejpam-1676	330	67	note	note	VERB
ejpam-1676	330	68	that	that	SCONJ
ejpam-1676	330	69	the	the	DET
ejpam-1676	330	70	p.d.f	p.d.f	NOUN
ejpam-1676	330	71	.	.	PUNCT
ejpam-1676	330	72	belongs	belong	VERB
ejpam-1676	330	73	to	to	ADP
ejpam-1676	330	74	an	an	DET
ejpam-1676	330	75	exponential	exponential	ADJ
ejpam-1676	330	76	family	family	NOUN
ejpam-1676	330	77	if	if	SCONJ
ejpam-1676	330	78	φ	φ	PROPN
ejpam-1676	330	79	is	be	AUX
ejpam-1676	330	80	known	know	VERB
ejpam-1676	330	81	.	.	PUNCT
ejpam-1676	331	1	as	as	ADP
ejpam-1676	331	2	a	a	DET
ejpam-1676	331	3	consequence	consequence	NOUN
ejpam-1676	331	4	,	,	PUNCT
ejpam-1676	331	5	for	for	ADP
ejpam-1676	331	6	i	i	PROPN
ejpam-1676	331	7	=	=	NOUN
ejpam-1676	331	8	1	1	NUM
ejpam-1676	331	9	,	,	PUNCT
ejpam-1676	331	10	.	.	PUNCT
ejpam-1676	331	11	.	.	PUNCT
ejpam-1676	332	1	.	.	PUNCT
ejpam-1676	332	2	,	,	PUNCT
ejpam-1676	333	1	n	n	CCONJ
ejpam-1676	333	2	,	,	PUNCT
ejpam-1676	333	3	e(yi	e(yi	NOUN
ejpam-1676	333	4	)	)	PUNCT
ejpam-1676	334	1	=	=	SYM
ejpam-1676	334	2	ζ	ζ	NOUN
ejpam-1676	334	3	′(ηi	′(ηi	NOUN
ejpam-1676	334	4	)	)	PUNCT
ejpam-1676	334	5	△	△	X
ejpam-1676	334	6	=	=	SYM
ejpam-1676	334	7	µ(ηi	µ(ηi	NOUN
ejpam-1676	334	8	)	)	PUNCT
ejpam-1676	334	9	,	,	PUNCT
ejpam-1676	334	10	var(yi	var(yi	X
ejpam-1676	334	11	)	)	PUNCT
ejpam-1676	334	12	=	=	SYM
ejpam-1676	334	13	φζ	φζ	NOUN
ejpam-1676	334	14	′′(ηi	′′(ηi	NUM
ejpam-1676	334	15	)	)	PUNCT
ejpam-1676	334	16	.	.	PUNCT
ejpam-1676	335	1	it	it	PRON
ejpam-1676	335	2	is	be	AUX
ejpam-1676	335	3	assumed	assume	VERB
ejpam-1676	335	4	that	that	SCONJ
ejpam-1676	335	5	ηi	ηi	PROPN
ejpam-1676	335	6	is	be	AUX
ejpam-1676	335	7	related	relate	VERB
ejpam-1676	335	8	to	to	ADP
ejpam-1676	335	9	x	x	PROPN
ejpam-1676	335	10	i	i	PROPN
ejpam-1676	335	11	,	,	PUNCT
ejpam-1676	335	12	the	the	DET
ejpam-1676	335	13	ith	ith	PROPN
ejpam-1676	335	14	value	value	NOUN
ejpam-1676	335	15	of	of	ADP
ejpam-1676	335	16	a	a	DET
ejpam-1676	335	17	p	p	ADJ
ejpam-1676	335	18	-	-	PUNCT
ejpam-1676	335	19	vector	vector	NOUN
ejpam-1676	335	20	covariates	covariate	NOUN
ejpam-1676	335	21	,	,	PUNCT
ejpam-1676	335	22	through	through	ADP
ejpam-1676	335	23	g(µ(ηi	g(µ(ηi	NOUN
ejpam-1676	335	24	)	)	PUNCT
ejpam-1676	335	25	)	)	PUNCT
ejpam-1676	336	1	=	=	PUNCT
ejpam-1676	336	2	β	β	X
ejpam-1676	336	3	t	t	X
ejpam-1676	336	4	x	x	X
ejpam-1676	336	5	i	i	PROPN
ejpam-1676	336	6	,	,	PUNCT
ejpam-1676	336	7	i	i	PRON
ejpam-1676	336	8	=	=	NOUN
ejpam-1676	336	9	1	1	NUM
ejpam-1676	336	10	,	,	PUNCT
ejpam-1676	336	11	.	.	PUNCT
ejpam-1676	336	12	.	.	PUNCT
ejpam-1676	336	13	.	.	PUNCT
ejpam-1676	337	1	,	,	PUNCT
ejpam-1676	337	2	n.	n.	NOUN
ejpam-1676	337	3	in	in	ADP
ejpam-1676	337	4	glm	glm	PROPN
ejpam-1676	337	5	,	,	PUNCT
ejpam-1676	337	6	β	β	X
ejpam-1676	337	7	is	be	AUX
ejpam-1676	337	8	the	the	DET
ejpam-1676	337	9	parameter	parameter	NOUN
ejpam-1676	337	10	of	of	ADP
ejpam-1676	337	11	interest	interest	NOUN
ejpam-1676	337	12	and	and	CCONJ
ejpam-1676	337	13	φi	φi	ADV
ejpam-1676	337	14	’s	’s	ADV
ejpam-1676	337	15	are	be	AUX
ejpam-1676	337	16	considered	consider	VERB
ejpam-1676	337	17	to	to	PART
ejpam-1676	337	18	be	be	AUX
ejpam-1676	337	19	nuisance	nuisance	ADJ
ejpam-1676	337	20	parameters	parameter	NOUN
ejpam-1676	337	21	.	.	PUNCT
ejpam-1676	338	1	the	the	DET
ejpam-1676	338	2	range	range	NOUN
ejpam-1676	338	3	of	of	ADP
ejpam-1676	338	4	β	β	PROPN
ejpam-1676	338	5	is	be	AUX
ejpam-1676	338	6	assumed	assume	VERB
ejpam-1676	338	7	to	to	PART
ejpam-1676	338	8	be	be	AUX
ejpam-1676	338	9	b	b	NOUN
ejpam-1676	338	10	=	=	SYM
ejpam-1676	338	11	{	{	PUNCT
ejpam-1676	338	12	β	β	X
ejpam-1676	338	13	:	:	PUNCT
ejpam-1676	338	14	(	(	PUNCT
ejpam-1676	338	15	g	g	NOUN
ejpam-1676	338	16	◦	◦	NOUN
ejpam-1676	338	17	µ)−1(β	µ)−1(β	PROPN
ejpam-1676	338	18	t	t	PROPN
ejpam-1676	338	19	x	x	SYM
ejpam-1676	338	20	)	)	PUNCT
ejpam-1676	338	21	∈	∈	PROPN
ejpam-1676	338	22	ξ	ξ	X
ejpam-1676	338	23	◦	◦	NOUN
ejpam-1676	338	24	∀x	∀x	X
ejpam-1676	338	25	∈	∈	NOUN
ejpam-1676	338	26	x	x	NOUN
ejpam-1676	338	27	}	}	PUNCT
ejpam-1676	338	28	,	,	PUNCT
ejpam-1676	338	29	where	where	SCONJ
ejpam-1676	338	30	x	x	PRON
ejpam-1676	338	31	is	be	AUX
ejpam-1676	338	32	the	the	DET
ejpam-1676	338	33	range	range	NOUN
ejpam-1676	338	34	of	of	ADP
ejpam-1676	338	35	x	x	PROPN
ejpam-1676	338	36	i	i	PRON
ejpam-1676	338	37	’	'	PUNCT
ejpam-1676	338	38	s.	s.	PROPN
ejpam-1676	338	39	it	it	PRON
ejpam-1676	338	40	is	be	AUX
ejpam-1676	338	41	often	often	ADV
ejpam-1676	338	42	assumed	assume	VERB
ejpam-1676	338	43	that	that	SCONJ
ejpam-1676	338	44	φi	φi	ADP
ejpam-1676	338	45	=	=	PROPN
ejpam-1676	338	46	φ	φ	PROPN
ejpam-1676	338	47	/	/	SYM
ejpam-1676	338	48	t	t	PROPN
ejpam-1676	339	1	i	i	PRON
ejpam-1676	339	2	,	,	PUNCT
ejpam-1676	339	3	i	i	PRON
ejpam-1676	339	4	=	=	NOUN
ejpam-1676	339	5	1	1	NUM
ejpam-1676	339	6	,	,	PUNCT
ejpam-1676	339	7	.	.	PUNCT
ejpam-1676	339	8	.	.	PUNCT
ejpam-1676	339	9	.	.	PUNCT
ejpam-1676	340	1	,	,	PUNCT
ejpam-1676	340	2	n	n	CCONJ
ejpam-1676	340	3	,	,	PUNCT
ejpam-1676	340	4	with	with	ADP
ejpam-1676	340	5	unknown	unknown	ADJ
ejpam-1676	340	6	φ	φ	PROPN
ejpam-1676	340	7	>	>	X
ejpam-1676	340	8	0	0	PUNCT
ejpam-1676	340	9	and	and	CCONJ
ejpam-1676	340	10	known	know	VERB
ejpam-1676	340	11	positive	positive	ADJ
ejpam-1676	340	12	t	t	NOUN
ejpam-1676	340	13	i	i	PRON
ejpam-1676	340	14	’	'	PUNCT
ejpam-1676	340	15	s.	s.	PROPN
ejpam-1676	340	16	let	let	VERB
ejpam-1676	340	17	θ	θ	PROPN
ejpam-1676	340	18	=	=	SYM
ejpam-1676	340	19	(	(	PUNCT
ejpam-1676	340	20	β	β	X
ejpam-1676	340	21	,	,	PUNCT
ejpam-1676	340	22	φ	φ	PROPN
ejpam-1676	340	23	)	)	PUNCT
ejpam-1676	340	24	and	and	CCONJ
ejpam-1676	340	25	ψ=	ψ=	NOUN
ejpam-1676	340	26	(	(	PUNCT
ejpam-1676	340	27	g	g	NOUN
ejpam-1676	340	28	◦	◦	NOUN
ejpam-1676	340	29	µ)−1	µ)−1	NOUN
ejpam-1676	340	30	,	,	PUNCT
ejpam-1676	340	31	then	then	ADV
ejpam-1676	340	32	the	the	DET
ejpam-1676	340	33	log	log	NOUN
ejpam-1676	340	34	-	-	PUNCT
ejpam-1676	340	35	likelihood	likelihood	NOUN
ejpam-1676	340	36	function	function	NOUN
ejpam-1676	340	37	is	be	AUX
ejpam-1676	340	38	log	log	VERB
ejpam-1676	340	39	l(θ	l(θ	NOUN
ejpam-1676	340	40	)	)	PUNCT
ejpam-1676	341	1	=	=	PUNCT
ejpam-1676	341	2	n∑	n∑	PROPN
ejpam-1676	341	3	i=1	i=1	PROPN
ejpam-1676	341	4	�	�	PROPN
ejpam-1676	341	5	log	log	VERB
ejpam-1676	341	6	h(yi	h(yi	PRON
ejpam-1676	341	7	,	,	PUNCT
ejpam-1676	341	8	φ	φ	PROPN
ejpam-1676	341	9	/	/	SYM
ejpam-1676	341	10	t	t	PROPN
ejpam-1676	341	11	i	i	PROPN
ejpam-1676	341	12	)	)	PUNCT
ejpam-1676	342	1	+	+	CCONJ
ejpam-1676	342	2	ψ(β	ψ(β	PROPN
ejpam-1676	342	3	t	t	NOUN
ejpam-1676	342	4	x	x	X
ejpam-1676	342	5	i)yi	i)yi	PROPN
ejpam-1676	342	6	−	−	PROPN
ejpam-1676	343	1	ζ(ψ(β	ζ(ψ(β	NOUN
ejpam-1676	343	2	t	t	NOUN
ejpam-1676	343	3	x	x	PROPN
ejpam-1676	343	4	i	i	NOUN
ejpam-1676	343	5	)	)	PUNCT
ejpam-1676	343	6	)	)	PUNCT
ejpam-1676	344	1	φ	φ	PROPN
ejpam-1676	344	2	/	/	SYM
ejpam-1676	344	3	t	t	PROPN
ejpam-1676	344	4	i	i	PRON
ejpam-1676	344	5	�	�	PROPN
ejpam-1676	344	6	and	and	CCONJ
ejpam-1676	344	7	the	the	DET
ejpam-1676	344	8	score	score	NOUN
ejpam-1676	344	9	function	function	NOUN
ejpam-1676	344	10	for	for	ADP
ejpam-1676	344	11	β	β	PROPN
ejpam-1676	344	12	is	be	AUX
ejpam-1676	344	13	sn(β	sn(β	PRON
ejpam-1676	344	14	)	)	PUNCT
ejpam-1676	345	1	=	=	SYM
ejpam-1676	346	1	∂	∂	NUM
ejpam-1676	346	2	log	log	NOUN
ejpam-1676	346	3	l(θ	l(θ	PROPN
ejpam-1676	346	4	)	)	PUNCT
ejpam-1676	346	5	∂	∂	PUNCT
ejpam-1676	346	6	β	β	X
ejpam-1676	346	7	=	=	SYM
ejpam-1676	346	8	1	1	NUM
ejpam-1676	346	9	φ	φ	NUM
ejpam-1676	346	10	n∑	n∑	NOUN
ejpam-1676	346	11	i=1	i=1	PROPN
ejpam-1676	347	1	{	{	PUNCT
ejpam-1676	347	2	[	[	X
ejpam-1676	347	3	yi	yi	X
ejpam-1676	347	4	−µ(ψ(β	−µ(ψ(β	PROPN
ejpam-1676	347	5	t	t	NOUN
ejpam-1676	347	6	x	x	SYM
ejpam-1676	347	7	i))]ψ	i))]ψ	NUM
ejpam-1676	347	8	′(β	′(β	NOUN
ejpam-1676	347	9	t	t	NOUN
ejpam-1676	347	10	x	x	PUNCT
ejpam-1676	347	11	i))t	i))t	PROPN
ejpam-1676	347	12	ix	ix	ADP
ejpam-1676	347	13	i	i	PROPN
ejpam-1676	347	14	}	}	PUNCT
ejpam-1676	347	15	.	.	PUNCT
ejpam-1676	348	1	let	let	VERB
ejpam-1676	348	2	mn(β	mn(β	X
ejpam-1676	348	3	)	)	PUNCT
ejpam-1676	349	1	=	=	SYM
ejpam-1676	349	2	n∑	n∑	NOUN
ejpam-1676	349	3	i=1	i=1	X
ejpam-1676	350	1	[	[	X
ejpam-1676	350	2	ψ′(β	ψ′(β	X
ejpam-1676	350	3	t	t	NOUN
ejpam-1676	350	4	x	x	PUNCT
ejpam-1676	350	5	i	i	NOUN
ejpam-1676	350	6	)	)	PUNCT
ejpam-1676	350	7	]	]	PUNCT
ejpam-1676	350	8	2ζ′′(ψ(β	2ζ′′(ψ(β	NUM
ejpam-1676	350	9	t	t	NOUN
ejpam-1676	350	10	x	x	PUNCT
ejpam-1676	350	11	i))t	i))t	PROPN
ejpam-1676	350	12	ix	ix	INTJ
ejpam-1676	351	1	ix	ix	ADP
ejpam-1676	351	2	t	t	PROPN
ejpam-1676	351	3	i	i	PRON
ejpam-1676	351	4	.	.	PUNCT
ejpam-1676	352	1	p.	p.	PROPN
ejpam-1676	352	2	guo	guo	PROPN
ejpam-1676	352	3	,	,	PUNCT
ejpam-1676	352	4	x.	x.	PROPN
ejpam-1676	352	5	wang	wang	PROPN
ejpam-1676	352	6	,	,	PUNCT
ejpam-1676	352	7	y.	y.	PROPN
ejpam-1676	352	8	wu	wu	PROPN
ejpam-1676	352	9	/	/	SYM
ejpam-1676	352	10	eur	eur	PROPN
ejpam-1676	352	11	.	.	PUNCT
ejpam-1676	353	1	j.	j.	PROPN
ejpam-1676	353	2	pure	pure	PROPN
ejpam-1676	353	3	appl	appl	PROPN
ejpam-1676	353	4	.	.	PROPN
ejpam-1676	353	5	math	math	PROPN
ejpam-1676	353	6	,	,	PUNCT
ejpam-1676	353	7	5	5	NUM
ejpam-1676	353	8	(	(	PUNCT
ejpam-1676	353	9	2012	2012	NUM
ejpam-1676	353	10	)	)	PUNCT
ejpam-1676	353	11	,	,	PUNCT
ejpam-1676	353	12	333	333	NUM
ejpam-1676	353	13	-	-	SYM
ejpam-1676	353	14	356	356	NUM
ejpam-1676	353	15	344	344	NUM
ejpam-1676	353	16	then	then	ADV
ejpam-1676	353	17	the	the	DET
ejpam-1676	353	18	fisher	fisher	PROPN
ejpam-1676	353	19	information	information	NOUN
ejpam-1676	353	20	evaluated	evaluate	VERB
ejpam-1676	353	21	at	at	ADP
ejpam-1676	353	22	β	β	PROPN
ejpam-1676	353	23	is	be	AUX
ejpam-1676	353	24	in(β	in(β	PUNCT
ejpam-1676	353	25	)	)	PUNCT
ejpam-1676	354	1	=	=	SYM
ejpam-1676	354	2	var	var	PROPN
ejpam-1676	354	3	�	�	PROPN
ejpam-1676	354	4	∂	∂	NUM
ejpam-1676	354	5	log	log	VERB
ejpam-1676	354	6	l(θ	l(θ	PROPN
ejpam-1676	354	7	)	)	PUNCT
ejpam-1676	354	8	∂	∂	NUM
ejpam-1676	354	9	β	β	X
ejpam-1676	354	10	�	�	PROPN
ejpam-1676	354	11	=	=	SYM
ejpam-1676	354	12	mn(β)/φ	mn(β)/φ	PROPN
ejpam-1676	354	13	.	.	PUNCT
ejpam-1676	355	1	we	we	PRON
ejpam-1676	355	2	now	now	ADV
ejpam-1676	355	3	present	present	VERB
ejpam-1676	355	4	the	the	DET
ejpam-1676	355	5	asymptotic	asymptotic	ADJ
ejpam-1676	355	6	properties	property	NOUN
ejpam-1676	355	7	for	for	ADP
ejpam-1676	355	8	the	the	DET
ejpam-1676	355	9	relevant	relevant	ADJ
ejpam-1676	355	10	weighted	weight	VERB
ejpam-1676	355	11	likelihood	likelihood	NOUN
ejpam-1676	355	12	estimator	estimator	NOUN
ejpam-1676	355	13	and	and	CCONJ
ejpam-1676	355	14	the	the	DET
ejpam-1676	355	15	hypothesis	hypothesis	NOUN
ejpam-1676	355	16	testing	testing	NOUN
ejpam-1676	355	17	based	base	VERB
ejpam-1676	355	18	on	on	ADP
ejpam-1676	355	19	this	this	DET
ejpam-1676	355	20	estimator	estimator	NOUN
ejpam-1676	355	21	.	.	PUNCT
ejpam-1676	356	1	we	we	PRON
ejpam-1676	356	2	use	use	VERB
ejpam-1676	356	3	double	double	ADJ
ejpam-1676	356	4	index	index	NOUN
ejpam-1676	356	5	as	as	ADP
ejpam-1676	356	6	the	the	DET
ejpam-1676	356	7	subscript	subscript	NOUN
ejpam-1676	356	8	of	of	ADP
ejpam-1676	356	9	observations	observation	NOUN
ejpam-1676	356	10	and	and	CCONJ
ejpam-1676	356	11	establish	establish	VERB
ejpam-1676	356	12	the	the	DET
ejpam-1676	356	13	following	following	ADJ
ejpam-1676	356	14	notations	notation	NOUN
ejpam-1676	356	15	.	.	PUNCT
ejpam-1676	357	1	•	•	NUM
ejpam-1676	357	2	β0	β0	PROPN
ejpam-1676	357	3	:	:	PUNCT
ejpam-1676	357	4	the	the	DET
ejpam-1676	357	5	true	true	ADJ
ejpam-1676	357	6	parameter	parameter	NOUN
ejpam-1676	357	7	value	value	NOUN
ejpam-1676	357	8	of	of	ADP
ejpam-1676	357	9	main	main	ADJ
ejpam-1676	357	10	group	group	NOUN
ejpam-1676	357	11	.	.	PUNCT
ejpam-1676	358	1	•	•	NUM
ejpam-1676	358	2	y1	y1	PROPN
ejpam-1676	358	3	j	j	PROPN
ejpam-1676	358	4	,	,	PUNCT
ejpam-1676	358	5	j	j	PROPN
ejpam-1676	358	6	=	=	NOUN
ejpam-1676	358	7	1	1	NUM
ejpam-1676	358	8	,	,	PUNCT
ejpam-1676	358	9	.	.	PUNCT
ejpam-1676	358	10	.	.	PUNCT
ejpam-1676	358	11	.	.	PUNCT
ejpam-1676	359	1	,	,	PUNCT
ejpam-1676	359	2	n1	n1	PROPN
ejpam-1676	359	3	are	be	AUX
ejpam-1676	359	4	the	the	DET
ejpam-1676	359	5	observations	observation	NOUN
ejpam-1676	359	6	from	from	ADP
ejpam-1676	359	7	main	main	ADJ
ejpam-1676	359	8	group	group	NOUN
ejpam-1676	359	9	.	.	PUNCT
ejpam-1676	360	1	•	•	NUM
ejpam-1676	360	2	y2	y2	INTJ
ejpam-1676	360	3	j	j	PROPN
ejpam-1676	360	4	,	,	PUNCT
ejpam-1676	360	5	j	j	PROPN
ejpam-1676	360	6	=	=	NOUN
ejpam-1676	360	7	1	1	NUM
ejpam-1676	360	8	,	,	PUNCT
ejpam-1676	360	9	.	.	PUNCT
ejpam-1676	360	10	.	.	PUNCT
ejpam-1676	360	11	.	.	PUNCT
ejpam-1676	361	1	,	,	PUNCT
ejpam-1676	361	2	n2	n2	PROPN
ejpam-1676	361	3	are	be	AUX
ejpam-1676	361	4	the	the	DET
ejpam-1676	361	5	observations	observation	NOUN
ejpam-1676	361	6	from	from	ADP
ejpam-1676	361	7	contamination	contamination	NOUN
ejpam-1676	361	8	group	group	NOUN
ejpam-1676	361	9	.	.	PUNCT
ejpam-1676	362	1	•	•	NUM
ejpam-1676	362	2	wl(β	wl(β	PUNCT
ejpam-1676	362	3	)	)	PUNCT
ejpam-1676	363	1	=	=	SYM
ejpam-1676	364	1	∑2	∑2	PROPN
ejpam-1676	364	2	i=1	i=1	X
ejpam-1676	364	3	∑ni	∑ni	VERB
ejpam-1676	365	1	j=1	j=1	PROPN
ejpam-1676	365	2	wi	wi	PROPN
ejpam-1676	365	3	j	j	PROPN
ejpam-1676	365	4	log	log	VERB
ejpam-1676	365	5	f	f	PROPN
ejpam-1676	365	6	(	(	PUNCT
ejpam-1676	365	7	yi	yi	NOUN
ejpam-1676	366	1	j	j	PROPN
ejpam-1676	366	2	|	|	ADV
ejpam-1676	366	3	x	x	X
ejpam-1676	366	4	i	i	PROPN
ejpam-1676	366	5	,	,	PUNCT
ejpam-1676	366	6	β	β	X
ejpam-1676	366	7	)	)	PUNCT
ejpam-1676	367	1	=	=	SYM
ejpam-1676	368	1	∑2	∑2	PROPN
ejpam-1676	368	2	i=1	i=1	X
ejpam-1676	368	3	∑ni	∑ni	VERB
ejpam-1676	369	1	j=1	j=1	PROPN
ejpam-1676	369	2	wi	wi	PROPN
ejpam-1676	369	3	j	j	PROPN
ejpam-1676	369	4	li	li	PROPN
ejpam-1676	369	5	j(β	j(β	PROPN
ejpam-1676	369	6	)	)	PUNCT
ejpam-1676	369	7	.	.	PUNCT
ejpam-1676	370	1	•	•	NUM
ejpam-1676	370	2	l(β	l(β	PROPN
ejpam-1676	370	3	)	)	PUNCT
ejpam-1676	371	1	=	=	PUNCT
ejpam-1676	372	1	∑2	∑2	PROPN
ejpam-1676	372	2	i=1	i=1	X
ejpam-1676	372	3	∑ni	∑ni	PUNCT
ejpam-1676	373	1	j=1	j=1	PROPN
ejpam-1676	373	2	log	log	NOUN
ejpam-1676	373	3	f	f	PROPN
ejpam-1676	374	1	(	(	PUNCT
ejpam-1676	374	2	yi	yi	NOUN
ejpam-1676	374	3	j	j	PROPN
ejpam-1676	375	1	|	|	ADV
ejpam-1676	375	2	x	x	X
ejpam-1676	375	3	i	i	PROPN
ejpam-1676	375	4	,	,	PUNCT
ejpam-1676	375	5	β	β	X
ejpam-1676	375	6	)	)	PUNCT
ejpam-1676	376	1	=	=	SYM
ejpam-1676	377	1	∑2	∑2	PROPN
ejpam-1676	377	2	i=1	i=1	X
ejpam-1676	377	3	∑ni	∑ni	PUNCT
ejpam-1676	377	4	j=1	j=1	PROPN
ejpam-1676	377	5	li	li	PROPN
ejpam-1676	377	6	j(β	j(β	PROPN
ejpam-1676	377	7	)	)	PUNCT
ejpam-1676	377	8	.	.	PUNCT
ejpam-1676	378	1	besides	besides	SCONJ
ejpam-1676	378	2	the	the	DET
ejpam-1676	378	3	regularity	regularity	NOUN
ejpam-1676	378	4	conditions	condition	NOUN
ejpam-1676	378	5	and	and	CCONJ
ejpam-1676	378	6	assumptions	assumption	NOUN
ejpam-1676	378	7	given	give	VERB
ejpam-1676	378	8	in	in	ADP
ejpam-1676	378	9	[	[	X
ejpam-1676	378	10	1	1	NUM
ejpam-1676	378	11	]	]	PUNCT
ejpam-1676	378	12	,	,	PUNCT
ejpam-1676	378	13	we	we	PRON
ejpam-1676	378	14	need	need	VERB
ejpam-1676	378	15	the	the	DET
ejpam-1676	378	16	following	follow	VERB
ejpam-1676	378	17	additional	additional	ADJ
ejpam-1676	378	18	assumptions	assumption	NOUN
ejpam-1676	378	19	for	for	ADP
ejpam-1676	378	20	the	the	DET
ejpam-1676	378	21	present	present	ADJ
ejpam-1676	378	22	problem	problem	NOUN
ejpam-1676	378	23	.	.	PUNCT
ejpam-1676	379	1	(	(	PUNCT
ejpam-1676	379	2	a1	a1	PROPN
ejpam-1676	379	3	)	)	PUNCT
ejpam-1676	379	4	the	the	DET
ejpam-1676	379	5	information	information	NOUN
ejpam-1676	379	6	matrix	matrix	NOUN
ejpam-1676	379	7	of	of	ADP
ejpam-1676	379	8	main	main	ADJ
ejpam-1676	379	9	group	group	NOUN
ejpam-1676	379	10	in1	in1	NOUN
ejpam-1676	379	11	(	(	PUNCT
ejpam-1676	379	12	β0	β0	NOUN
ejpam-1676	379	13	)	)	PUNCT
ejpam-1676	379	14	is	be	AUX
ejpam-1676	379	15	positive	positive	ADJ
ejpam-1676	379	16	definite	definite	ADJ
ejpam-1676	379	17	,	,	PUNCT
ejpam-1676	379	18	and	and	CCONJ
ejpam-1676	379	19	the	the	DET
ejpam-1676	379	20	eigenvalues	eigenvalue	NOUN
ejpam-1676	379	21	of	of	ADP
ejpam-1676	379	22	n−r	n−r	NOUN
ejpam-1676	379	23	1	1	NUM
ejpam-1676	379	24	in1	in1	NOUN
ejpam-1676	379	25	(	(	PUNCT
ejpam-1676	379	26	β0	β0	NOUN
ejpam-1676	379	27	)	)	PUNCT
ejpam-1676	379	28	are	be	AUX
ejpam-1676	379	29	bounded	bound	VERB
ejpam-1676	379	30	,	,	PUNCT
ejpam-1676	379	31	for	for	ADP
ejpam-1676	379	32	some	some	DET
ejpam-1676	379	33	r	r	NOUN
ejpam-1676	379	34	>	>	X
ejpam-1676	379	35	0	0	NUM
ejpam-1676	379	36	,	,	PUNCT
ejpam-1676	379	37	so	so	ADV
ejpam-1676	379	38	are	be	AUX
ejpam-1676	379	39	the	the	DET
ejpam-1676	379	40	eigenvalues	eigenvalue	NOUN
ejpam-1676	379	41	of	of	ADP
ejpam-1676	379	42	nr	nr	NOUN
ejpam-1676	379	43	1	1	NUM
ejpam-1676	379	44	i−1	i−1	PROPN
ejpam-1676	379	45	n1	n1	PROPN
ejpam-1676	379	46	(	(	PUNCT
ejpam-1676	379	47	β0	β0	NOUN
ejpam-1676	379	48	)	)	PUNCT
ejpam-1676	379	49	.	.	PUNCT
ejpam-1676	380	1	that	that	PRON
ejpam-1676	380	2	is	be	AUX
ejpam-1676	380	3	,	,	PUNCT
ejpam-1676	380	4	there	there	PRON
ejpam-1676	380	5	exist	exist	VERB
ejpam-1676	380	6	constants	constant	NOUN
ejpam-1676	380	7	c1	c1	PROPN
ejpam-1676	380	8	and	and	CCONJ
ejpam-1676	380	9	c2	c2	PROPN
ejpam-1676	380	10	,	,	PUNCT
ejpam-1676	380	11	such	such	ADJ
ejpam-1676	380	12	that	that	SCONJ
ejpam-1676	380	13	c1	c1	PROPN
ejpam-1676	380	14	≤	≤	PROPN
ejpam-1676	381	1	λminnr	λminnr	PROPN
ejpam-1676	381	2	1i−1	1i−1	NUM
ejpam-1676	381	3	n1	n1	PROPN
ejpam-1676	381	4	(	(	PUNCT
ejpam-1676	381	5	β0)≤	β0)≤	PUNCT
ejpam-1676	381	6	λmaxnr	λmaxnr	VERB
ejpam-1676	381	7	1i−1	1i−1	NUM
ejpam-1676	381	8	n1	n1	PROPN
ejpam-1676	381	9	(	(	PUNCT
ejpam-1676	381	10	β0)≤	β0)≤	NUM
ejpam-1676	381	11	c2	c2	PROPN
ejpam-1676	381	12	.	.	PUNCT
ejpam-1676	382	1	(	(	PUNCT
ejpam-1676	382	2	a2	a2	PROPN
ejpam-1676	382	3	)	)	PUNCT
ejpam-1676	382	4	there	there	PRON
ejpam-1676	382	5	exists	exist	VERB
ejpam-1676	382	6	a	a	DET
ejpam-1676	382	7	neighborhood	neighborhood	NOUN
ejpam-1676	382	8	of	of	ADP
ejpam-1676	382	9	β0	β0	ADJ
ejpam-1676	382	10	,	,	PUNCT
ejpam-1676	382	11	o	o	NOUN
ejpam-1676	382	12	,	,	PUNCT
ejpam-1676	382	13	such	such	ADJ
ejpam-1676	382	14	that	that	SCONJ
ejpam-1676	382	15	for	for	ADP
ejpam-1676	382	16	all	all	DET
ejpam-1676	382	17	β	β	NOUN
ejpam-1676	382	18	∈	∈	NOUN
ejpam-1676	382	19	o	o	NOUN
ejpam-1676	382	20	and	and	CCONJ
ejpam-1676	382	21	y2	y2	PROPN
ejpam-1676	382	22	j	j	PROPN
ejpam-1676	382	23	,	,	PUNCT
ejpam-1676	382	24	j	j	PROPN
ejpam-1676	382	25	=	=	NOUN
ejpam-1676	382	26	1	1	NUM
ejpam-1676	382	27	,	,	PUNCT
ejpam-1676	382	28	.	.	PUNCT
ejpam-1676	382	29	.	.	PUNCT
ejpam-1676	383	1	.	.	PUNCT
ejpam-1676	384	1	,	,	PUNCT
ejpam-1676	384	2	n2	n2	NOUN
ejpam-1676	384	3	,	,	PUNCT
ejpam-1676	384	4	e	e	X
ejpam-1676	384	5	(	(	PUNCT
ejpam-1676	384	6	sup	sup	PROPN
ejpam-1676	384	7	β∈o,1≤	β∈o,1≤	PROPN
ejpam-1676	384	8	j≤n2	j≤n2	PROPN
ejpam-1676	384	9	�	�	PROPN
ejpam-1676	384	10	�	�	PROPN
ejpam-1676	384	11	�	�	PROPN
ejpam-1676	384	12	�	�	PROPN
ejpam-1676	384	13	�	�	PROPN
ejpam-1676	384	14	log	log	NOUN
ejpam-1676	384	15	f	f	PROPN
ejpam-1676	385	1	(	(	PUNCT
ejpam-1676	385	2	y2	y2	PROPN
ejpam-1676	385	3	j	j	PROPN
ejpam-1676	385	4	,	,	PUNCT
ejpam-1676	385	5	β	β	X
ejpam-1676	385	6	)	)	PUNCT
ejpam-1676	385	7	f	f	PROPN
ejpam-1676	385	8	(	(	PUNCT
ejpam-1676	385	9	y2	y2	PROPN
ejpam-1676	385	10	j	j	PROPN
ejpam-1676	385	11	,	,	PUNCT
ejpam-1676	385	12	β0	β0	PROPN
ejpam-1676	385	13	)	)	PUNCT
ejpam-1676	385	14	�	�	PROPN
ejpam-1676	385	15	�	�	PROPN
ejpam-1676	385	16	�	�	PROPN
ejpam-1676	385	17	�	�	PROPN
ejpam-1676	385	18	�	�	PROPN
ejpam-1676	385	19	)	)	PUNCT
ejpam-1676	385	20	≤	≤	PUNCT
ejpam-1676	386	1	k	k	NOUN
ejpam-1676	386	2	,	,	PUNCT
ejpam-1676	386	3	where	where	SCONJ
ejpam-1676	386	4	k	k	PROPN
ejpam-1676	386	5	is	be	AUX
ejpam-1676	386	6	a	a	DET
ejpam-1676	386	7	constant	constant	ADJ
ejpam-1676	386	8	.	.	PUNCT
ejpam-1676	387	1	(	(	PUNCT
ejpam-1676	387	2	a3	a3	NOUN
ejpam-1676	387	3	)	)	PUNCT
ejpam-1676	387	4	the	the	DET
ejpam-1676	387	5	proportion	proportion	NOUN
ejpam-1676	387	6	of	of	ADP
ejpam-1676	387	7	contamination	contamination	NOUN
ejpam-1676	387	8	converges	converge	NOUN
ejpam-1676	387	9	to	to	ADP
ejpam-1676	387	10	0	0	NUM
ejpam-1676	387	11	in	in	ADP
ejpam-1676	387	12	the	the	DET
ejpam-1676	387	13	order	order	NOUN
ejpam-1676	387	14	of	of	ADP
ejpam-1676	387	15	n−1/2	n−1/2	PROPN
ejpam-1676	387	16	,	,	PUNCT
ejpam-1676	387	17	i.e.	i.e.	X
ejpam-1676	387	18	,	,	PUNCT
ejpam-1676	387	19	εn	εn	ADJ
ejpam-1676	387	20	=	=	PUNCT
ejpam-1676	387	21	o(n−1/2	o(n−1/2	PROPN
ejpam-1676	387	22	)	)	PUNCT
ejpam-1676	387	23	.	.	PUNCT
ejpam-1676	388	1	(	(	PUNCT
ejpam-1676	388	2	a4	a4	INTJ
ejpam-1676	388	3	)	)	PUNCT
ejpam-1676	388	4	the	the	DET
ejpam-1676	388	5	criteria	criterion	NOUN
ejpam-1676	388	6	to	to	PART
ejpam-1676	388	7	choose	choose	VERB
ejpam-1676	388	8	weights	weight	NOUN
ejpam-1676	388	9	guarantees	guarantee	NOUN
ejpam-1676	388	10	that	that	SCONJ
ejpam-1676	388	11	the	the	DET
ejpam-1676	388	12	weights	weight	NOUN
ejpam-1676	388	13	are	be	AUX
ejpam-1676	388	14	consistent	consistent	ADJ
ejpam-1676	388	15	.	.	PUNCT
ejpam-1676	389	1	assume	assume	VERB
ejpam-1676	389	2	the	the	DET
ejpam-1676	389	3	first	first	ADJ
ejpam-1676	389	4	n1	n1	PROPN
ejpam-1676	389	5	observations	observation	NOUN
ejpam-1676	389	6	are	be	AUX
ejpam-1676	389	7	from	from	ADP
ejpam-1676	389	8	main	main	ADJ
ejpam-1676	389	9	group	group	NOUN
ejpam-1676	389	10	and	and	CCONJ
ejpam-1676	389	11	the	the	DET
ejpam-1676	389	12	rest	rest	NOUN
ejpam-1676	389	13	are	be	AUX
ejpam-1676	389	14	from	from	ADP
ejpam-1676	389	15	contamination	contamination	NOUN
ejpam-1676	389	16	group	group	NOUN
ejpam-1676	389	17	.	.	PUNCT
ejpam-1676	390	1	then	then	ADV
ejpam-1676	390	2	as	as	ADP
ejpam-1676	390	3	n→∞	n→∞	NUM
ejpam-1676	390	4	and	and	CCONJ
ejpam-1676	390	5	εn→	εn→	NOUN
ejpam-1676	390	6	0	0	NUM
ejpam-1676	390	7	,	,	PUNCT
ejpam-1676	390	8	w	w	NOUN
ejpam-1676	390	9	=	=	SYM
ejpam-1676	390	10	(	(	PUNCT
ejpam-1676	390	11	w1	w1	NOUN
ejpam-1676	390	12	,	,	PUNCT
ejpam-1676	390	13	.	.	PUNCT
ejpam-1676	390	14	.	.	PUNCT
ejpam-1676	391	1	.	.	PUNCT
ejpam-1676	392	1	,	,	PUNCT
ejpam-1676	392	2	wn	wn	PROPN
ejpam-1676	392	3	)	)	PUNCT
ejpam-1676	392	4	t	t	PROPN
ejpam-1676	392	5	→	→	SYM
ejpam-1676	392	6	v	v	NOUN
ejpam-1676	392	7	=	=	PUNCT
ejpam-1676	392	8	(	(	PUNCT
ejpam-1676	392	9	v1	v1	NOUN
ejpam-1676	392	10	,	,	PUNCT
ejpam-1676	392	11	.	.	PUNCT
ejpam-1676	392	12	.	.	PUNCT
ejpam-1676	392	13	.	.	PUNCT
ejpam-1676	393	1	,	,	PUNCT
ejpam-1676	393	2	vn	vn	X
ejpam-1676	393	3	)	)	PUNCT
ejpam-1676	393	4	=	=	SYM
ejpam-1676	394	1	(	(	PUNCT
ejpam-1676	394	2	1	1	NUM
ejpam-1676	394	3	,	,	PUNCT
ejpam-1676	394	4	.	.	PUNCT
ejpam-1676	394	5	.	.	PUNCT
ejpam-1676	395	1	.	.	PUNCT
ejpam-1676	396	1	,	,	PUNCT
ejpam-1676	396	2	1︸	1︸	NUM
ejpam-1676	396	3	︷︷	︷︷	NOUN
ejpam-1676	396	4	︸	︸	X
ejpam-1676	396	5	n1	n1	PROPN
ejpam-1676	396	6	,	,	PUNCT
ejpam-1676	396	7	0	0	NUM
ejpam-1676	396	8	,	,	PUNCT
ejpam-1676	396	9	.	.	PUNCT
ejpam-1676	396	10	.	.	PUNCT
ejpam-1676	397	1	.	.	PUNCT
ejpam-1676	398	1	,	,	PUNCT
ejpam-1676	398	2	0︸	0︸	PUNCT
ejpam-1676	398	3	︷︷	︷︷	NOUN
ejpam-1676	398	4	︸	︸	X
ejpam-1676	398	5	n2	n2	PROPN
ejpam-1676	398	6	)	)	PUNCT
ejpam-1676	398	7	t	t	PROPN
ejpam-1676	398	8	.	.	PUNCT
ejpam-1676	399	1	furthermore	furthermore	ADV
ejpam-1676	399	2	,	,	PUNCT
ejpam-1676	399	3	maxi	maxi	ADJ
ejpam-1676	399	4	|wi	|wi	NOUN
ejpam-1676	399	5	−	−	PROPN
ejpam-1676	399	6	vi|	vi|	PROPN
ejpam-1676	399	7	=	=	SYM
ejpam-1676	399	8	o(n−1/2	o(n−1/2	PROPN
ejpam-1676	399	9	)	)	PUNCT
ejpam-1676	399	10	.	.	PUNCT
ejpam-1676	400	1	we	we	PRON
ejpam-1676	400	2	now	now	ADV
ejpam-1676	400	3	provide	provide	VERB
ejpam-1676	400	4	the	the	DET
ejpam-1676	400	5	asymptotic	asymptotic	ADJ
ejpam-1676	400	6	results	result	NOUN
ejpam-1676	400	7	in	in	ADP
ejpam-1676	400	8	theorems	theorem	NOUN
ejpam-1676	400	9	2	2	NUM
ejpam-1676	400	10	-	-	SYM
ejpam-1676	400	11	3	3	NUM
ejpam-1676	400	12	below	below	ADV
ejpam-1676	400	13	.	.	PUNCT
ejpam-1676	401	1	p.	p.	NOUN
ejpam-1676	401	2	guo	guo	PROPN
ejpam-1676	401	3	,	,	PUNCT
ejpam-1676	401	4	x.	x.	PROPN
ejpam-1676	401	5	wang	wang	PROPN
ejpam-1676	401	6	,	,	PUNCT
ejpam-1676	401	7	y.	y.	PROPN
ejpam-1676	401	8	wu	wu	PROPN
ejpam-1676	401	9	/	/	SYM
ejpam-1676	401	10	eur	eur	PROPN
ejpam-1676	401	11	.	.	PUNCT
ejpam-1676	402	1	j.	j.	PROPN
ejpam-1676	402	2	pure	pure	PROPN
ejpam-1676	402	3	appl	appl	PROPN
ejpam-1676	402	4	.	.	PROPN
ejpam-1676	402	5	math	math	PROPN
ejpam-1676	402	6	,	,	PUNCT
ejpam-1676	402	7	5	5	NUM
ejpam-1676	402	8	(	(	PUNCT
ejpam-1676	402	9	2012	2012	NUM
ejpam-1676	402	10	)	)	PUNCT
ejpam-1676	402	11	,	,	PUNCT
ejpam-1676	402	12	333	333	NUM
ejpam-1676	402	13	-	-	SYM
ejpam-1676	402	14	356	356	NUM
ejpam-1676	402	15	345	345	NUM
ejpam-1676	402	16	theorem	theorem	NOUN
ejpam-1676	402	17	2	2	NUM
ejpam-1676	402	18	.	.	PUNCT
ejpam-1676	403	1	under	under	ADP
ejpam-1676	403	2	the	the	DET
ejpam-1676	403	3	regularity	regularity	NOUN
ejpam-1676	403	4	conditions	condition	NOUN
ejpam-1676	403	5	specified	specify	VERB
ejpam-1676	403	6	in	in	ADP
ejpam-1676	403	7	[	[	X
ejpam-1676	403	8	1	1	NUM
ejpam-1676	403	9	]	]	PUNCT
ejpam-1676	403	10	and	and	CCONJ
ejpam-1676	403	11	assumptions	assumption	NOUN
ejpam-1676	403	12	(	(	PUNCT
ejpam-1676	403	13	a1)–(a4	a1)–(a4	INTJ
ejpam-1676	403	14	)	)	PUNCT
ejpam-1676	403	15	,	,	PUNCT
ejpam-1676	403	16	there	there	PRON
ejpam-1676	403	17	is	be	VERB
ejpam-1676	403	18	a	a	DET
ejpam-1676	403	19	sequence	sequence	NOUN
ejpam-1676	403	20	of	of	ADP
ejpam-1676	403	21	estimators	estimator	NOUN
ejpam-1676	403	22	{	{	PUNCT
ejpam-1676	403	23	β̂n	β̂n	NOUN
ejpam-1676	403	24	}	}	PUNCT
ejpam-1676	403	25	such	such	ADJ
ejpam-1676	403	26	that	that	SCONJ
ejpam-1676	403	27	p(sn(β̂n	p(sn(β̂n	NOUN
ejpam-1676	403	28	)	)	PUNCT
ejpam-1676	403	29	=	=	SYM
ejpam-1676	403	30	0)→	0)→	NOUN
ejpam-1676	403	31	1	1	NUM
ejpam-1676	403	32	and	and	CCONJ
ejpam-1676	403	33	β̂n→	β̂n→	NOUN
ejpam-1676	403	34	β0	β0	ADJ
ejpam-1676	403	35	in	in	ADP
ejpam-1676	403	36	probability	probability	NOUN
ejpam-1676	403	37	.	.	PUNCT
ejpam-1676	404	1	proof	proof	NOUN
ejpam-1676	404	2	.	.	PUNCT
ejpam-1676	405	1	in	in	ADP
ejpam-1676	405	2	order	order	NOUN
ejpam-1676	405	3	to	to	PART
ejpam-1676	405	4	present	present	VERB
ejpam-1676	405	5	the	the	DET
ejpam-1676	405	6	proof	proof	NOUN
ejpam-1676	405	7	clearly	clearly	ADV
ejpam-1676	405	8	,	,	PUNCT
ejpam-1676	405	9	we	we	PRON
ejpam-1676	405	10	use	use	VERB
ejpam-1676	405	11	double	double	ADJ
ejpam-1676	405	12	index	index	NOUN
ejpam-1676	405	13	as	as	ADP
ejpam-1676	405	14	the	the	DET
ejpam-1676	405	15	subscript	subscript	NOUN
ejpam-1676	405	16	of	of	ADP
ejpam-1676	405	17	observations	observation	NOUN
ejpam-1676	405	18	.	.	PUNCT
ejpam-1676	406	1	then	then	ADV
ejpam-1676	406	2	we	we	PRON
ejpam-1676	406	3	need	need	VERB
ejpam-1676	406	4	to	to	PART
ejpam-1676	406	5	specify	specify	VERB
ejpam-1676	406	6	the	the	DET
ejpam-1676	406	7	following	follow	VERB
ejpam-1676	406	8	notations	notation	NOUN
ejpam-1676	406	9	.	.	PUNCT
ejpam-1676	407	1	•	•	NUM
ejpam-1676	407	2	β0	β0	PROPN
ejpam-1676	407	3	:	:	PUNCT
ejpam-1676	407	4	the	the	DET
ejpam-1676	407	5	true	true	ADJ
ejpam-1676	407	6	parameter	parameter	NOUN
ejpam-1676	407	7	of	of	ADP
ejpam-1676	407	8	main	main	ADJ
ejpam-1676	407	9	group	group	NOUN
ejpam-1676	407	10	.	.	PUNCT
ejpam-1676	408	1	•	•	NUM
ejpam-1676	408	2	γ0	γ0	PROPN
ejpam-1676	408	3	:	:	PUNCT
ejpam-1676	408	4	the	the	DET
ejpam-1676	408	5	true	true	ADJ
ejpam-1676	408	6	parameter	parameter	NOUN
ejpam-1676	408	7	of	of	ADP
ejpam-1676	408	8	second	second	ADJ
ejpam-1676	408	9	group	group	NOUN
ejpam-1676	408	10	.	.	PUNCT
ejpam-1676	409	1	•	•	NUM
ejpam-1676	409	2	y1	y1	PROPN
ejpam-1676	409	3	j	j	PROPN
ejpam-1676	409	4	,	,	PUNCT
ejpam-1676	409	5	j	j	PROPN
ejpam-1676	409	6	=	=	NOUN
ejpam-1676	409	7	1	1	NUM
ejpam-1676	409	8	,	,	PUNCT
ejpam-1676	409	9	.	.	PUNCT
ejpam-1676	409	10	.	.	PUNCT
ejpam-1676	409	11	.	.	PUNCT
ejpam-1676	410	1	,	,	PUNCT
ejpam-1676	410	2	n1	n1	PROPN
ejpam-1676	410	3	are	be	AUX
ejpam-1676	410	4	the	the	DET
ejpam-1676	410	5	observations	observation	NOUN
ejpam-1676	410	6	from	from	ADP
ejpam-1676	410	7	main	main	ADJ
ejpam-1676	410	8	group	group	NOUN
ejpam-1676	410	9	.	.	PUNCT
ejpam-1676	411	1	•	•	NUM
ejpam-1676	411	2	y2	y2	INTJ
ejpam-1676	411	3	j	j	PROPN
ejpam-1676	411	4	,	,	PUNCT
ejpam-1676	411	5	j	j	PROPN
ejpam-1676	411	6	=	=	NOUN
ejpam-1676	411	7	1	1	NUM
ejpam-1676	411	8	,	,	PUNCT
ejpam-1676	411	9	.	.	PUNCT
ejpam-1676	411	10	.	.	PUNCT
ejpam-1676	411	11	.	.	PUNCT
ejpam-1676	412	1	,	,	PUNCT
ejpam-1676	412	2	n2	n2	PROPN
ejpam-1676	412	3	are	be	AUX
ejpam-1676	412	4	the	the	DET
ejpam-1676	412	5	observations	observation	NOUN
ejpam-1676	412	6	from	from	ADP
ejpam-1676	412	7	second	second	ADJ
ejpam-1676	412	8	group	group	NOUN
ejpam-1676	412	9	.	.	PUNCT
ejpam-1676	413	1	•	•	NUM
ejpam-1676	413	2	wl(β	wl(β	PUNCT
ejpam-1676	413	3	)	)	PUNCT
ejpam-1676	414	1	=	=	SYM
ejpam-1676	415	1	∑2	∑2	PROPN
ejpam-1676	415	2	i=1	i=1	X
ejpam-1676	415	3	∑ni	∑ni	VERB
ejpam-1676	416	1	j=1	j=1	PROPN
ejpam-1676	416	2	wi	wi	PROPN
ejpam-1676	416	3	j	j	PROPN
ejpam-1676	416	4	log	log	VERB
ejpam-1676	416	5	f	f	PROPN
ejpam-1676	417	1	(	(	PUNCT
ejpam-1676	417	2	yi	yi	PROPN
ejpam-1676	417	3	j	j	PROPN
ejpam-1676	417	4	|x	|x	PROPN
ejpam-1676	417	5	i	i	PRON
ejpam-1676	417	6	,	,	PUNCT
ejpam-1676	417	7	β	β	X
ejpam-1676	417	8	)	)	PUNCT
ejpam-1676	418	1	=	=	SYM
ejpam-1676	419	1	∑2	∑2	PROPN
ejpam-1676	419	2	i=1	i=1	X
ejpam-1676	419	3	∑ni	∑ni	VERB
ejpam-1676	420	1	j=1	j=1	PROPN
ejpam-1676	420	2	wi	wi	PROPN
ejpam-1676	420	3	j	j	PROPN
ejpam-1676	420	4	li	li	PROPN
ejpam-1676	420	5	j(β	j(β	PROPN
ejpam-1676	420	6	)	)	PUNCT
ejpam-1676	420	7	.	.	PUNCT
ejpam-1676	421	1	•	•	NUM
ejpam-1676	421	2	l(β	l(β	PROPN
ejpam-1676	421	3	)	)	PUNCT
ejpam-1676	422	1	=	=	PUNCT
ejpam-1676	423	1	∑2	∑2	PROPN
ejpam-1676	423	2	i=1	i=1	X
ejpam-1676	423	3	∑ni	∑ni	PUNCT
ejpam-1676	424	1	j=1	j=1	PROPN
ejpam-1676	424	2	log	log	NOUN
ejpam-1676	424	3	f	f	PROPN
ejpam-1676	425	1	(	(	PUNCT
ejpam-1676	425	2	yi	yi	PROPN
ejpam-1676	425	3	j	j	PROPN
ejpam-1676	425	4	|x	|x	PROPN
ejpam-1676	425	5	i	i	PRON
ejpam-1676	425	6	,	,	PUNCT
ejpam-1676	425	7	β	β	X
ejpam-1676	425	8	)	)	PUNCT
ejpam-1676	426	1	=	=	SYM
ejpam-1676	427	1	∑2	∑2	PROPN
ejpam-1676	427	2	i=1	i=1	X
ejpam-1676	427	3	∑ni	∑ni	PUNCT
ejpam-1676	427	4	j=1	j=1	PROPN
ejpam-1676	427	5	li	li	PROPN
ejpam-1676	427	6	j(β	j(β	PROPN
ejpam-1676	427	7	)	)	PUNCT
ejpam-1676	427	8	.	.	PUNCT
ejpam-1676	428	1	since	since	SCONJ
ejpam-1676	428	2	there	there	PRON
ejpam-1676	428	3	are	be	VERB
ejpam-1676	428	4	two	two	NUM
ejpam-1676	428	5	populations	population	NOUN
ejpam-1676	428	6	,	,	PUNCT
ejpam-1676	428	7	we	we	PRON
ejpam-1676	428	8	distinguish	distinguish	VERB
ejpam-1676	428	9	the	the	DET
ejpam-1676	428	10	score	score	NOUN
ejpam-1676	428	11	functions	function	NOUN
ejpam-1676	428	12	,	,	PUNCT
ejpam-1676	428	13	fisher	fisher	PROPN
ejpam-1676	428	14	information	information	NOUN
ejpam-1676	428	15	,	,	PUNCT
ejpam-1676	428	16	design	design	NOUN
ejpam-1676	428	17	matrices	matrix	NOUN
ejpam-1676	428	18	in	in	ADP
ejpam-1676	428	19	the	the	DET
ejpam-1676	428	20	following	following	ADJ
ejpam-1676	428	21	way	way	NOUN
ejpam-1676	428	22	.	.	PUNCT
ejpam-1676	429	1	for	for	ADP
ejpam-1676	429	2	main	main	ADJ
ejpam-1676	429	3	group	group	NOUN
ejpam-1676	429	4	,	,	PUNCT
ejpam-1676	429	5	we	we	PRON
ejpam-1676	429	6	denote	denote	VERB
ejpam-1676	429	7	them	they	PRON
ejpam-1676	429	8	as	as	ADP
ejpam-1676	429	9	sn1	sn1	PROPN
ejpam-1676	429	10	(	(	PUNCT
ejpam-1676	429	11	β	β	NOUN
ejpam-1676	429	12	)	)	PUNCT
ejpam-1676	429	13	,	,	PUNCT
ejpam-1676	429	14	in1	in1	NOUN
ejpam-1676	429	15	(	(	PUNCT
ejpam-1676	429	16	β	β	NOUN
ejpam-1676	429	17	)	)	PUNCT
ejpam-1676	429	18	and	and	CCONJ
ejpam-1676	429	19	x	x	INTJ
ejpam-1676	429	20	;	;	PUNCT
ejpam-1676	429	21	and	and	CCONJ
ejpam-1676	429	22	for	for	ADP
ejpam-1676	429	23	contamination	contamination	NOUN
ejpam-1676	429	24	group	group	NOUN
ejpam-1676	429	25	,	,	PUNCT
ejpam-1676	429	26	we	we	PRON
ejpam-1676	429	27	have	have	VERB
ejpam-1676	429	28	sn2	sn2	PROPN
ejpam-1676	429	29	(	(	PUNCT
ejpam-1676	429	30	β	β	NOUN
ejpam-1676	429	31	)	)	PUNCT
ejpam-1676	429	32	,	,	PUNCT
ejpam-1676	429	33	in2	in2	PROPN
ejpam-1676	429	34	(	(	PUNCT
ejpam-1676	429	35	β	β	NOUN
ejpam-1676	429	36	)	)	PUNCT
ejpam-1676	429	37	and	and	CCONJ
ejpam-1676	429	38	z	z	NOUN
ejpam-1676	429	39	.	.	PUNCT
ejpam-1676	430	1	for	for	ADP
ejpam-1676	430	2	simplicity	simplicity	NOUN
ejpam-1676	430	3	,	,	PUNCT
ejpam-1676	430	4	we	we	PRON
ejpam-1676	430	5	use	use	VERB
ejpam-1676	430	6	mn1	mn1	PROPN
ejpam-1676	430	7	to	to	PART
ejpam-1676	430	8	denote	denote	VERB
ejpam-1676	430	9	mn1	mn1	PROPN
ejpam-1676	430	10	(	(	PUNCT
ejpam-1676	430	11	β0	β0	NOUN
ejpam-1676	430	12	)	)	PUNCT
ejpam-1676	430	13	.	.	PUNCT
ejpam-1676	431	1	define	define	VERB
ejpam-1676	431	2	the	the	DET
ejpam-1676	431	3	neighborhood	neighborhood	NOUN
ejpam-1676	431	4	of	of	ADP
ejpam-1676	431	5	β0	β0	PROPN
ejpam-1676	431	6	,	,	PUNCT
ejpam-1676	431	7	nn1	nn1	PROPN
ejpam-1676	431	8	(	(	PUNCT
ejpam-1676	431	9	δ	δ	PROPN
ejpam-1676	431	10	)	)	PUNCT
ejpam-1676	431	11	,	,	PUNCT
ejpam-1676	431	12	δ	δ	PROPN
ejpam-1676	431	13	>	>	X
ejpam-1676	431	14	0	0	NUM
ejpam-1676	431	15	,	,	PUNCT
ejpam-1676	431	16	as	as	ADP
ejpam-1676	431	17	nn1	nn1	PROPN
ejpam-1676	431	18	(	(	PUNCT
ejpam-1676	431	19	δ	δ	PROPN
ejpam-1676	431	20	)	)	PUNCT
ejpam-1676	431	21	=	=	SYM
ejpam-1676	431	22	{	{	PUNCT
ejpam-1676	431	23	β	β	X
ejpam-1676	431	24	:	:	PUNCT
ejpam-1676	431	25	‖in1	‖in1	PROPN
ejpam-1676	431	26	(	(	PUNCT
ejpam-1676	431	27	β0	β0	PROPN
ejpam-1676	431	28	)	)	PUNCT
ejpam-1676	431	29	1/2(β	1/2(β	NUM
ejpam-1676	431	30	−β0)‖	−β0)‖	VERB
ejpam-1676	431	31	≤	≤	NUM
ejpam-1676	431	32	δ	δ	PROPN
ejpam-1676	431	33	)	)	PUNCT
ejpam-1676	431	34	}	}	PUNCT
ejpam-1676	431	35	,	,	PUNCT
ejpam-1676	431	36	where	where	SCONJ
ejpam-1676	431	37	in1	in1	PROPN
ejpam-1676	431	38	(	(	PUNCT
ejpam-1676	431	39	β0	β0	NOUN
ejpam-1676	431	40	)	)	PUNCT
ejpam-1676	431	41	is	be	AUX
ejpam-1676	431	42	the	the	DET
ejpam-1676	431	43	fisher	fisher	PROPN
ejpam-1676	431	44	information	information	NOUN
ejpam-1676	431	45	of	of	ADP
ejpam-1676	431	46	the	the	DET
ejpam-1676	431	47	main	main	ADJ
ejpam-1676	431	48	group	group	NOUN
ejpam-1676	431	49	data	datum	NOUN
ejpam-1676	431	50	evaluated	evaluate	VERB
ejpam-1676	431	51	at	at	ADP
ejpam-1676	431	52	true	true	ADJ
ejpam-1676	431	53	value	value	NOUN
ejpam-1676	431	54	β0	β0	NOUN
ejpam-1676	431	55	.	.	PUNCT
ejpam-1676	432	1	let	let	AUX
ejpam-1676	432	2	∂	∂	PROPN
ejpam-1676	432	3	nn1	nn1	PROPN
ejpam-1676	432	4	(	(	PUNCT
ejpam-1676	432	5	δ	δ	PROPN
ejpam-1676	432	6	)	)	PUNCT
ejpam-1676	432	7	be	be	VERB
ejpam-1676	432	8	the	the	DET
ejpam-1676	432	9	boundary	boundary	NOUN
ejpam-1676	432	10	of	of	ADP
ejpam-1676	432	11	nn1	nn1	PROPN
ejpam-1676	432	12	(	(	PUNCT
ejpam-1676	432	13	δ	δ	PROPN
ejpam-1676	432	14	)	)	PUNCT
ejpam-1676	432	15	.	.	PUNCT
ejpam-1676	433	1	to	to	PART
ejpam-1676	433	2	prove	prove	VERB
ejpam-1676	433	3	β	β	X
ejpam-1676	433	4	p→	p→	NOUN
ejpam-1676	433	5	β0	β0	PROPN
ejpam-1676	433	6	,	,	PUNCT
ejpam-1676	433	7	it	it	PRON
ejpam-1676	433	8	is	be	AUX
ejpam-1676	433	9	sufficient	sufficient	ADJ
ejpam-1676	433	10	to	to	PART
ejpam-1676	433	11	show	show	VERB
ejpam-1676	433	12	for	for	ADP
ejpam-1676	433	13	any	any	DET
ejpam-1676	433	14	η	η	PROPN
ejpam-1676	433	15	>	>	X
ejpam-1676	433	16	0	0	PROPN
ejpam-1676	433	17	,	,	PUNCT
ejpam-1676	433	18	there	there	PRON
ejpam-1676	433	19	exist	exist	VERB
ejpam-1676	433	20	δ	δ	PROPN
ejpam-1676	433	21	>	>	X
ejpam-1676	433	22	0	0	PUNCT
ejpam-1676	434	1	and	and	CCONJ
ejpam-1676	434	2	n	n	CCONJ
ejpam-1676	434	3	>	>	ADP
ejpam-1676	434	4	0	0	NUM
ejpam-1676	434	5	,	,	PUNCT
ejpam-1676	434	6	such	such	ADJ
ejpam-1676	434	7	that	that	PRON
ejpam-1676	434	8	for	for	ADP
ejpam-1676	434	9	n≥	n≥	PROPN
ejpam-1676	434	10	n	n	NOUN
ejpam-1676	434	11	and	and	CCONJ
ejpam-1676	434	12	all	all	DET
ejpam-1676	434	13	β	β	X
ejpam-1676	434	14	∈	∈	PROPN
ejpam-1676	434	15	∂	∂	NUM
ejpam-1676	434	16	nn1	nn1	PROPN
ejpam-1676	434	17	(	(	PUNCT
ejpam-1676	434	18	δ	δ	PROPN
ejpam-1676	434	19	)	)	PUNCT
ejpam-1676	434	20	,	,	PUNCT
ejpam-1676	434	21	we	we	PRON
ejpam-1676	434	22	have	have	VERB
ejpam-1676	434	23	p(wl(β)−wl(β0	p(wl(β)−wl(β0	PROPN
ejpam-1676	434	24	)	)	PUNCT
ejpam-1676	434	25	<	<	X
ejpam-1676	434	26	0	0	NUM
ejpam-1676	434	27	for	for	ADP
ejpam-1676	434	28	all	all	DET
ejpam-1676	434	29	β	β	X
ejpam-1676	434	30	∈	∈	PROPN
ejpam-1676	434	31	∂	∂	NUM
ejpam-1676	434	32	nn1	nn1	PROPN
ejpam-1676	434	33	(	(	PUNCT
ejpam-1676	434	34	δ	δ	PROPN
ejpam-1676	434	35	)	)	PUNCT
ejpam-1676	434	36	)	)	PUNCT
ejpam-1676	434	37	>	>	X
ejpam-1676	435	1	1−η	1−η	NUM
ejpam-1676	435	2	.	.	PUNCT
ejpam-1676	436	1	we	we	PRON
ejpam-1676	436	2	start	start	VERB
ejpam-1676	436	3	the	the	DET
ejpam-1676	436	4	proof	proof	NOUN
ejpam-1676	436	5	by	by	ADP
ejpam-1676	436	6	partitioning	partition	VERB
ejpam-1676	436	7	wl(β)−wl(β0	wl(β)−wl(β0	PROPN
ejpam-1676	436	8	)	)	PUNCT
ejpam-1676	436	9	.	.	PUNCT
ejpam-1676	437	1	wl(β)−wl(β0	wl(β)−wl(β0	PROPN
ejpam-1676	437	2	)	)	PUNCT
ejpam-1676	438	1	=	=	SYM
ejpam-1676	438	2	n1∑	n1∑	PROPN
ejpam-1676	438	3	j=1	j=1	PROPN
ejpam-1676	438	4	�	�	PROPN
ejpam-1676	438	5	l1	l1	PROPN
ejpam-1676	438	6	j(β)−	j(β)−	PROPN
ejpam-1676	438	7	l1	l1	PROPN
ejpam-1676	438	8	j(β0	j(β0	PROPN
ejpam-1676	438	9	)	)	PUNCT
ejpam-1676	438	10	�	�	PROPN
ejpam-1676	439	1	−	−	PROPN
ejpam-1676	440	1	n1∑	n1∑	PROPN
ejpam-1676	440	2	j=1	j=1	PROPN
ejpam-1676	440	3	(	(	PUNCT
ejpam-1676	440	4	1−w1	1−w1	NUM
ejpam-1676	440	5	j	j	PROPN
ejpam-1676	440	6	)	)	PUNCT
ejpam-1676	440	7	�	�	PROPN
ejpam-1676	440	8	l1	l1	PROPN
ejpam-1676	440	9	j(β)−	j(β)−	PROPN
ejpam-1676	440	10	l1	l1	PROPN
ejpam-1676	440	11	j(β0	j(β0	PROPN
ejpam-1676	440	12	)	)	PUNCT
ejpam-1676	440	13	�	�	PROPN
ejpam-1676	440	14	+	+	CCONJ
ejpam-1676	440	15	n2∑	n2∑	PROPN
ejpam-1676	440	16	j=1	j=1	PROPN
ejpam-1676	440	17	w2	w2	NOUN
ejpam-1676	440	18	j	j	PROPN
ejpam-1676	440	19	�	�	PROPN
ejpam-1676	440	20	l2	l2	PROPN
ejpam-1676	440	21	j(β)−	j(β)−	PROPN
ejpam-1676	440	22	l2	l2	NOUN
ejpam-1676	440	23	j(β0	j(β0	NOUN
ejpam-1676	440	24	)	)	PUNCT
ejpam-1676	440	25	�	�	PROPN
ejpam-1676	440	26	△	△	PROPN
ejpam-1676	440	27	=	=	SYM
ejpam-1676	440	28	a−	a−	PROPN
ejpam-1676	440	29	b	b	PROPN
ejpam-1676	440	30	+	+	CCONJ
ejpam-1676	440	31	c	c	NOUN
ejpam-1676	440	32	.	.	PUNCT
ejpam-1676	441	1	in	in	ADP
ejpam-1676	441	2	[	[	X
ejpam-1676	441	3	1	1	NUM
ejpam-1676	441	4	]	]	PUNCT
ejpam-1676	441	5	,	,	PUNCT
ejpam-1676	441	6	it	it	PRON
ejpam-1676	441	7	is	be	AUX
ejpam-1676	441	8	shown	show	VERB
ejpam-1676	441	9	that	that	SCONJ
ejpam-1676	441	10	there	there	PRON
ejpam-1676	441	11	exist	exist	VERB
ejpam-1676	441	12	δ	δ	PROPN
ejpam-1676	441	13	>	>	X
ejpam-1676	441	14	0	0	PUNCT
ejpam-1676	441	15	and	and	CCONJ
ejpam-1676	441	16	n	n	PRON
ejpam-1676	441	17	such	such	ADJ
ejpam-1676	441	18	that	that	PRON
ejpam-1676	441	19	for	for	ADP
ejpam-1676	441	20	n≥	n≥	PROPN
ejpam-1676	441	21	n	n	NOUN
ejpam-1676	441	22	and	and	CCONJ
ejpam-1676	441	23	all	all	DET
ejpam-1676	441	24	β	β	X
ejpam-1676	441	25	∈	∈	PROPN
ejpam-1676	441	26	∂	∂	NUM
ejpam-1676	441	27	nn1	nn1	PROPN
ejpam-1676	441	28	(	(	PUNCT
ejpam-1676	441	29	δ	δ	PROPN
ejpam-1676	441	30	)	)	PUNCT
ejpam-1676	441	31	,	,	PUNCT
ejpam-1676	441	32	p(a	p(a	PROPN
ejpam-1676	441	33	<	<	X
ejpam-1676	441	34	0	0	NUM
ejpam-1676	441	35	)	)	PUNCT
ejpam-1676	441	36	>	>	X
ejpam-1676	442	1	1−η	1−η	NUM
ejpam-1676	442	2	.	.	PUNCT
ejpam-1676	443	1	p.	p.	NOUN
ejpam-1676	443	2	guo	guo	PROPN
ejpam-1676	443	3	,	,	PUNCT
ejpam-1676	443	4	x.	x.	PROPN
ejpam-1676	443	5	wang	wang	PROPN
ejpam-1676	443	6	,	,	PUNCT
ejpam-1676	443	7	y.	y.	PROPN
ejpam-1676	443	8	wu	wu	PROPN
ejpam-1676	443	9	/	/	SYM
ejpam-1676	443	10	eur	eur	PROPN
ejpam-1676	443	11	.	.	PUNCT
ejpam-1676	444	1	j.	j.	PROPN
ejpam-1676	444	2	pure	pure	PROPN
ejpam-1676	444	3	appl	appl	PROPN
ejpam-1676	444	4	.	.	PROPN
ejpam-1676	444	5	math	math	PROPN
ejpam-1676	444	6	,	,	PUNCT
ejpam-1676	444	7	5	5	NUM
ejpam-1676	444	8	(	(	PUNCT
ejpam-1676	444	9	2012	2012	NUM
ejpam-1676	444	10	)	)	PUNCT
ejpam-1676	444	11	,	,	PUNCT
ejpam-1676	444	12	333	333	NUM
ejpam-1676	444	13	-	-	SYM
ejpam-1676	444	14	356	356	NUM
ejpam-1676	444	15	346	346	NUM
ejpam-1676	445	1	then	then	ADV
ejpam-1676	445	2	we	we	PRON
ejpam-1676	445	3	need	need	VERB
ejpam-1676	445	4	to	to	PART
ejpam-1676	445	5	show	show	VERB
ejpam-1676	445	6	that	that	PRON
ejpam-1676	445	7	b	b	NOUN
ejpam-1676	445	8	and	and	CCONJ
ejpam-1676	445	9	c	c	PROPN
ejpam-1676	445	10	are	be	AUX
ejpam-1676	445	11	dominated	dominate	VERB
ejpam-1676	445	12	by	by	ADP
ejpam-1676	445	13	a	a	PRON
ejpam-1676	445	14	,	,	PUNCT
ejpam-1676	445	15	which	which	PRON
ejpam-1676	445	16	will	will	AUX
ejpam-1676	445	17	follows	follow	VERB
ejpam-1676	445	18	from	from	ADP
ejpam-1676	445	19	b	b	PROPN
ejpam-1676	445	20	and	and	CCONJ
ejpam-1676	445	21	c	c	PROPN
ejpam-1676	445	22	are	be	AUX
ejpam-1676	445	23	op(1	op(1	NOUN
ejpam-1676	445	24	)	)	PUNCT
ejpam-1676	445	25	since	since	SCONJ
ejpam-1676	445	26	it	it	PRON
ejpam-1676	445	27	is	be	AUX
ejpam-1676	445	28	shown	show	VERB
ejpam-1676	445	29	that	that	SCONJ
ejpam-1676	445	30	a	a	PRON
ejpam-1676	445	31	is	be	AUX
ejpam-1676	445	32	op(1	op(1	NOUN
ejpam-1676	445	33	)	)	PUNCT
ejpam-1676	445	34	in	in	ADP
ejpam-1676	445	35	[	[	X
ejpam-1676	445	36	1	1	NUM
ejpam-1676	445	37	]	]	PUNCT
ejpam-1676	445	38	.	.	PUNCT
ejpam-1676	446	1	in	in	ADP
ejpam-1676	446	2	the	the	DET
ejpam-1676	446	3	following	following	NOUN
ejpam-1676	446	4	,	,	PUNCT
ejpam-1676	446	5	we	we	PRON
ejpam-1676	446	6	show	show	VERB
ejpam-1676	446	7	that	that	SCONJ
ejpam-1676	446	8	b	b	NOUN
ejpam-1676	446	9	is	be	AUX
ejpam-1676	446	10	dominated	dominate	VERB
ejpam-1676	446	11	by	by	ADP
ejpam-1676	446	12	a.	a.	NOUN
ejpam-1676	446	13	define	define	PROPN
ejpam-1676	446	14	s1−w	s1−w	PROPN
ejpam-1676	446	15	n1	n1	PROPN
ejpam-1676	446	16	(	(	PUNCT
ejpam-1676	446	17	β	β	NOUN
ejpam-1676	446	18	)	)	PUNCT
ejpam-1676	446	19	as	as	ADP
ejpam-1676	446	20	s1−w	s1−w	PROPN
ejpam-1676	446	21	n1	n1	PROPN
ejpam-1676	446	22	(	(	PUNCT
ejpam-1676	446	23	β	β	NOUN
ejpam-1676	446	24	)	)	PUNCT
ejpam-1676	446	25	=	=	SYM
ejpam-1676	446	26	∂	∂	NUM
ejpam-1676	446	27	∂	∂	NOUN
ejpam-1676	446	28	β	β	X
ejpam-1676	446	29	n1∑	n1∑	PROPN
ejpam-1676	446	30	j=1	j=1	PROPN
ejpam-1676	446	31	(	(	PUNCT
ejpam-1676	446	32	1−w1	1−w1	NUM
ejpam-1676	446	33	j)l1	j)l1	PROPN
ejpam-1676	446	34	j(β	j(β	PROPN
ejpam-1676	446	35	)	)	PUNCT
ejpam-1676	447	1	=	=	SYM
ejpam-1676	447	2	�	�	PROPN
ejpam-1676	447	3	∂	∂	NUM
ejpam-1676	447	4	log	log	NOUN
ejpam-1676	447	5	f	f	PROPN
ejpam-1676	447	6	(	(	PUNCT
ejpam-1676	447	7	y11,β	y11,β	NOUN
ejpam-1676	447	8	)	)	PUNCT
ejpam-1676	447	9	∂β	∂β	PROPN
ejpam-1676	447	10	,	,	PUNCT
ejpam-1676	447	11	.	.	PUNCT
ejpam-1676	447	12	.	.	PUNCT
ejpam-1676	448	1	.	.	PUNCT
ejpam-1676	449	1	,	,	PUNCT
ejpam-1676	449	2	∂	∂	NUM
ejpam-1676	450	1	log	log	NOUN
ejpam-1676	450	2	f	f	PROPN
ejpam-1676	450	3	(	(	PUNCT
ejpam-1676	450	4	y1n1	y1n1	PROPN
ejpam-1676	450	5	,	,	PUNCT
ejpam-1676	450	6	β	β	NOUN
ejpam-1676	450	7	)	)	PUNCT
ejpam-1676	450	8	∂	∂	PUNCT
ejpam-1676	450	9	β	β	X
ejpam-1676	450	10	�	�	PROPN
ejpam-1676	450	11			PROPN
ejpam-1676	450	12			NOUN
ejpam-1676	451	1	1−w11	1−w11	INTJ
ejpam-1676	451	2	...	...	PUNCT
ejpam-1676	452	1	1−w1n1	1−w1n1	INTJ
ejpam-1676	452	2			NOUN
ejpam-1676	452	3			PROPN
ejpam-1676	452	4	△	△	PROPN
ejpam-1676	453	1	=	=	SYM
ejpam-1676	453	2	�	�	PROPN
ejpam-1676	453	3	u1	u1	PROPN
ejpam-1676	453	4	,	,	PUNCT
ejpam-1676	453	5	.	.	PUNCT
ejpam-1676	453	6	.	.	PUNCT
ejpam-1676	453	7	.	.	PUNCT
ejpam-1676	454	1	,	,	PUNCT
ejpam-1676	454	2	un1	un1	PROPN
ejpam-1676	454	3	�	�	PROPN
ejpam-1676	454	4	w1	w1	NOUN
ejpam-1676	454	5	△	△	PROPN
ejpam-1676	454	6	=	=	SYM
ejpam-1676	454	7	s1w1	s1w1	PROPN
ejpam-1676	454	8	.	.	PUNCT
ejpam-1676	454	9	let	let	VERB
ejpam-1676	454	10	sn1	sn1	PROPN
ejpam-1676	454	11	(	(	PUNCT
ejpam-1676	454	12	β	β	NOUN
ejpam-1676	454	13	)	)	PUNCT
ejpam-1676	454	14	be	be	VERB
ejpam-1676	454	15	the	the	DET
ejpam-1676	454	16	score	score	NOUN
ejpam-1676	454	17	function	function	NOUN
ejpam-1676	454	18	for	for	ADP
ejpam-1676	454	19	the	the	DET
ejpam-1676	454	20	main	main	ADJ
ejpam-1676	454	21	group	group	NOUN
ejpam-1676	454	22	,	,	PUNCT
ejpam-1676	454	23	which	which	PRON
ejpam-1676	454	24	can	can	AUX
ejpam-1676	454	25	be	be	AUX
ejpam-1676	454	26	similarly	similarly	ADV
ejpam-1676	454	27	written	write	VERB
ejpam-1676	454	28	as	as	ADP
ejpam-1676	454	29	sn1	sn1	PROPN
ejpam-1676	454	30	(	(	PUNCT
ejpam-1676	454	31	β	β	X
ejpam-1676	454	32	)	)	PUNCT
ejpam-1676	455	1	=	=	SYM
ejpam-1676	455	2	s11n1	s11n1	NOUN
ejpam-1676	455	3	,	,	PUNCT
ejpam-1676	455	4	where	where	SCONJ
ejpam-1676	455	5	1n	1n	NOUN
ejpam-1676	455	6	denotes	denote	VERB
ejpam-1676	455	7	the	the	DET
ejpam-1676	455	8	length	length	NOUN
ejpam-1676	455	9	-	-	PUNCT
ejpam-1676	455	10	n	n	NOUN
ejpam-1676	455	11	vector	vector	NOUN
ejpam-1676	455	12	with	with	ADP
ejpam-1676	455	13	all	all	DET
ejpam-1676	455	14	element	element	NOUN
ejpam-1676	455	15	equal	equal	ADJ
ejpam-1676	455	16	to	to	ADP
ejpam-1676	455	17	1	1	NUM
ejpam-1676	455	18	.	.	PUNCT
ejpam-1676	456	1	then	then	ADV
ejpam-1676	456	2	by	by	ADP
ejpam-1676	456	3	taking	take	VERB
ejpam-1676	456	4	taylor	taylor	PROPN
ejpam-1676	456	5	expansion	expansion	NOUN
ejpam-1676	456	6	of	of	ADP
ejpam-1676	456	7	b	b	NOUN
ejpam-1676	456	8	,	,	PUNCT
ejpam-1676	456	9	we	we	PRON
ejpam-1676	456	10	have	have	VERB
ejpam-1676	456	11	b	b	NOUN
ejpam-1676	456	12	=	=	SYM
ejpam-1676	456	13	(	(	PUNCT
ejpam-1676	456	14	β	β	X
ejpam-1676	456	15	−β0	−β0	PROPN
ejpam-1676	456	16	)	)	PUNCT
ejpam-1676	456	17	t	t	PROPN
ejpam-1676	456	18	s1−w	s1−w	PROPN
ejpam-1676	456	19	n1	n1	PROPN
ejpam-1676	456	20	(	(	PUNCT
ejpam-1676	456	21	β0	β0	PROPN
ejpam-1676	456	22	)	)	PUNCT
ejpam-1676	456	23	+	+	CCONJ
ejpam-1676	456	24	1	1	NUM
ejpam-1676	456	25	2	2	NUM
ejpam-1676	456	26	(	(	PUNCT
ejpam-1676	456	27	β	β	X
ejpam-1676	456	28	−β0	−β0	PROPN
ejpam-1676	456	29	)	)	PUNCT
ejpam-1676	456	30	t∇s1−w	t∇s1−w	NOUN
ejpam-1676	456	31	n1	n1	NOUN
ejpam-1676	456	32	(	(	PUNCT
ejpam-1676	456	33	γ∗)(β	γ∗)(β	NOUN
ejpam-1676	456	34	−β0	−β0	PROPN
ejpam-1676	456	35	)	)	PUNCT
ejpam-1676	456	36	,	,	PUNCT
ejpam-1676	456	37	where	where	SCONJ
ejpam-1676	456	38	γ∗	γ∗	PROPN
ejpam-1676	456	39	∈	∈	PROPN
ejpam-1676	456	40	nn1	nn1	PROPN
ejpam-1676	456	41	(	(	PUNCT
ejpam-1676	456	42	δ	δ	PROPN
ejpam-1676	456	43	)	)	PUNCT
ejpam-1676	456	44	.	.	PUNCT
ejpam-1676	457	1	in	in	ADP
ejpam-1676	457	2	the	the	DET
ejpam-1676	457	3	following	following	NOUN
ejpam-1676	457	4	,	,	PUNCT
ejpam-1676	457	5	we	we	PRON
ejpam-1676	457	6	will	will	AUX
ejpam-1676	457	7	show	show	VERB
ejpam-1676	457	8	that	that	SCONJ
ejpam-1676	457	9	both	both	CCONJ
ejpam-1676	457	10	these	these	DET
ejpam-1676	457	11	two	two	NUM
ejpam-1676	457	12	terms	term	NOUN
ejpam-1676	457	13	are	be	AUX
ejpam-1676	457	14	op(1	op(1	NOUN
ejpam-1676	457	15	)	)	PUNCT
ejpam-1676	457	16	.	.	PUNCT
ejpam-1676	458	1	to	to	PART
ejpam-1676	458	2	show	show	VERB
ejpam-1676	458	3	(	(	PUNCT
ejpam-1676	458	4	β	β	X
ejpam-1676	458	5	−β0	−β0	PROPN
ejpam-1676	458	6	)	)	PUNCT
ejpam-1676	458	7	t	t	PROPN
ejpam-1676	458	8	s1−w	s1−w	PROPN
ejpam-1676	458	9	n1	n1	PROPN
ejpam-1676	458	10	(	(	PUNCT
ejpam-1676	458	11	β0	β0	NOUN
ejpam-1676	458	12	)	)	PUNCT
ejpam-1676	458	13	=	=	PUNCT
ejpam-1676	458	14	op(1	op(1	PROPN
ejpam-1676	458	15	)	)	PUNCT
ejpam-1676	458	16	,	,	PUNCT
ejpam-1676	458	17	we	we	PRON
ejpam-1676	458	18	need	need	VERB
ejpam-1676	458	19	to	to	PART
ejpam-1676	458	20	show	show	VERB
ejpam-1676	458	21	for	for	ADP
ejpam-1676	458	22	any	any	DET
ejpam-1676	458	23	ǫ	ǫ	NOUN
ejpam-1676	458	24	>	>	X
ejpam-1676	458	25	0	0	PROPN
ejpam-1676	458	26	,	,	PUNCT
ejpam-1676	458	27	lim	lim	NOUN
ejpam-1676	458	28	n→∞	n→∞	PRON
ejpam-1676	458	29	p(‖(β	p(‖(β	NOUN
ejpam-1676	458	30	−β0	−β0	PROPN
ejpam-1676	458	31	)	)	PUNCT
ejpam-1676	458	32	t	t	PROPN
ejpam-1676	458	33	s1−w	s1−w	PROPN
ejpam-1676	458	34	n1	n1	PROPN
ejpam-1676	458	35	(	(	PUNCT
ejpam-1676	458	36	β0)‖	β0)‖	X
ejpam-1676	458	37	<	<	X
ejpam-1676	458	38	ǫ	ǫ	NOUN
ejpam-1676	458	39	)	)	PUNCT
ejpam-1676	458	40	=	=	SYM
ejpam-1676	458	41	1	1	NUM
ejpam-1676	458	42	by	by	ADP
ejpam-1676	458	43	markov	markov	NOUN
ejpam-1676	458	44	inequality	inequality	NOUN
ejpam-1676	458	45	,	,	PUNCT
ejpam-1676	458	46	we	we	PRON
ejpam-1676	458	47	have	have	VERB
ejpam-1676	458	48	p(‖(β	p(‖(β	NOUN
ejpam-1676	458	49	−β0	−β0	NOUN
ejpam-1676	458	50	)	)	PUNCT
ejpam-1676	458	51	t	t	PROPN
ejpam-1676	458	52	s1−w	s1−w	PROPN
ejpam-1676	458	53	n1	n1	PROPN
ejpam-1676	458	54	(	(	PUNCT
ejpam-1676	458	55	β0)‖	β0)‖	VERB
ejpam-1676	458	56	<	<	X
ejpam-1676	458	57	ǫ)≥	ǫ)≥	PROPN
ejpam-1676	458	58	1−	1−	NUM
ejpam-1676	458	59	(	(	PUNCT
ejpam-1676	458	60	1	1	NUM
ejpam-1676	458	61	ǫ	ǫ	NOUN
ejpam-1676	458	62	)	)	PUNCT
ejpam-1676	458	63	2e‖(β	2e‖(β	NUM
ejpam-1676	458	64	−β0	−β0	NOUN
ejpam-1676	458	65	)	)	PUNCT
ejpam-1676	458	66	t	t	PROPN
ejpam-1676	458	67	s1−w	s1−w	PROPN
ejpam-1676	458	68	n1	n1	PROPN
ejpam-1676	458	69	(	(	PUNCT
ejpam-1676	458	70	β0)‖2	β0)‖2	PROPN
ejpam-1676	458	71	and	and	CCONJ
ejpam-1676	458	72	so	so	ADV
ejpam-1676	458	73	it	it	PRON
ejpam-1676	458	74	is	be	AUX
ejpam-1676	458	75	sufficient	sufficient	ADJ
ejpam-1676	458	76	to	to	PART
ejpam-1676	458	77	show	show	VERB
ejpam-1676	458	78	e‖(β	e‖(β	NOUN
ejpam-1676	458	79	−β0	−β0	PROPN
ejpam-1676	458	80	)	)	PUNCT
ejpam-1676	458	81	t	t	PROPN
ejpam-1676	458	82	s1−w	s1−w	PROPN
ejpam-1676	458	83	n1	n1	PROPN
ejpam-1676	458	84	(	(	PUNCT
ejpam-1676	458	85	β0)‖2→	β0)‖2→	NOUN
ejpam-1676	458	86	0	0	NUM
ejpam-1676	458	87	.	.	PUNCT
ejpam-1676	459	1	let	let	VERB
ejpam-1676	459	2	λ=	λ=	ADJ
ejpam-1676	459	3	in1	in1	ADJ
ejpam-1676	459	4	(	(	PUNCT
ejpam-1676	459	5	β0	β0	PROPN
ejpam-1676	459	6	)	)	PUNCT
ejpam-1676	459	7	1/2(β	1/2(β	NUM
ejpam-1676	460	1	−β0)/δ	−β0)/δ	NOUN
ejpam-1676	460	2	,	,	PUNCT
ejpam-1676	460	3	then	then	ADV
ejpam-1676	460	4	‖λ‖=	‖λ‖=	PROPN
ejpam-1676	460	5	1	1	NUM
ejpam-1676	460	6	and	and	CCONJ
ejpam-1676	460	7	e‖(β	e‖(β	NOUN
ejpam-1676	460	8	−β0	−β0	PROPN
ejpam-1676	460	9	)	)	PUNCT
ejpam-1676	460	10	t	t	PROPN
ejpam-1676	460	11	s1−w	s1−w	PROPN
ejpam-1676	460	12	n1	n1	PROPN
ejpam-1676	460	13	(	(	PUNCT
ejpam-1676	460	14	β0)‖2	β0)‖2	PROPN
ejpam-1676	460	15	=	=	SYM
ejpam-1676	460	16	e‖(β	e‖(β	NOUN
ejpam-1676	460	17	−β0	−β0	PROPN
ejpam-1676	460	18	)	)	PUNCT
ejpam-1676	460	19	t	t	PROPN
ejpam-1676	460	20	s1w1‖2	s1w1‖2	PUNCT
ejpam-1676	460	21	=	=	SYM
ejpam-1676	460	22	e‖δλt	e‖δλt	X
ejpam-1676	460	23	i−1/2	i−1/2	PROPN
ejpam-1676	460	24	n1	n1	PROPN
ejpam-1676	460	25	(	(	PUNCT
ejpam-1676	460	26	β0)s1w1‖2	β0)s1w1‖2	PROPN
ejpam-1676	460	27	≤	≤	NUM
ejpam-1676	460	28	δ2‖w1‖2e‖i−1/2	δ2‖w1‖2e‖i−1/2	PROPN
ejpam-1676	460	29	n1	n1	PROPN
ejpam-1676	460	30	(	(	PUNCT
ejpam-1676	460	31	β0)s1‖2	β0)s1‖2	NOUN
ejpam-1676	460	32	=	=	PUNCT
ejpam-1676	460	33	δ2‖w1‖2e	δ2‖w1‖2e	PROPN
ejpam-1676	460	34	n	n	PROPN
ejpam-1676	460	35	t	t	X
ejpam-1676	460	36	r[st	r[st	NOUN
ejpam-1676	460	37	1	1	NUM
ejpam-1676	460	38	i−1	i−1	PROPN
ejpam-1676	460	39	n1	n1	PROPN
ejpam-1676	460	40	(	(	PUNCT
ejpam-1676	460	41	β0)s1	β0)s1	PROPN
ejpam-1676	460	42	]	]	X
ejpam-1676	460	43	o	o	X
ejpam-1676	461	1	=	=	PUNCT
ejpam-1676	461	2	δ2‖w1‖2e	δ2‖w1‖2e	NUM
ejpam-1676	461	3			VERB
ejpam-1676	461	4			ADP
ejpam-1676	461	5			ADJ
ejpam-1676	461	6	n1∑	n1∑	PROPN
ejpam-1676	461	7	j=1	j=1	PROPN
ejpam-1676	461	8	u	u	PROPN
ejpam-1676	461	9	t	t	X
ejpam-1676	461	10	j	j	PROPN
ejpam-1676	461	11	i−1	i−1	PROPN
ejpam-1676	461	12	n1	n1	PROPN
ejpam-1676	461	13	(	(	PUNCT
ejpam-1676	461	14	β0)u	β0)u	PROPN
ejpam-1676	461	15	j	j	PROPN
ejpam-1676	461	16			PROPN
ejpam-1676	461	17			PROPN
ejpam-1676	461	18			NOUN
ejpam-1676	461	19	.	.	PUNCT
ejpam-1676	462	1	p.	p.	NOUN
ejpam-1676	462	2	guo	guo	PROPN
ejpam-1676	462	3	,	,	PUNCT
ejpam-1676	462	4	x.	x.	PROPN
ejpam-1676	462	5	wang	wang	PROPN
ejpam-1676	462	6	,	,	PUNCT
ejpam-1676	462	7	y.	y.	PROPN
ejpam-1676	462	8	wu	wu	PROPN
ejpam-1676	462	9	/	/	SYM
ejpam-1676	462	10	eur	eur	PROPN
ejpam-1676	462	11	.	.	PUNCT
ejpam-1676	463	1	j.	j.	PROPN
ejpam-1676	463	2	pure	pure	PROPN
ejpam-1676	463	3	appl	appl	PROPN
ejpam-1676	463	4	.	.	PROPN
ejpam-1676	463	5	math	math	PROPN
ejpam-1676	463	6	,	,	PUNCT
ejpam-1676	463	7	5	5	NUM
ejpam-1676	463	8	(	(	PUNCT
ejpam-1676	463	9	2012	2012	NUM
ejpam-1676	463	10	)	)	PUNCT
ejpam-1676	463	11	,	,	PUNCT
ejpam-1676	463	12	333	333	NUM
ejpam-1676	463	13	-	-	SYM
ejpam-1676	463	14	356	356	NUM
ejpam-1676	463	15	347	347	NUM
ejpam-1676	463	16	because	because	SCONJ
ejpam-1676	463	17	the	the	DET
ejpam-1676	463	18	eigenvalues	eigenvalue	NOUN
ejpam-1676	463	19	of	of	ADP
ejpam-1676	463	20	in1	in1	NOUN
ejpam-1676	463	21	(	(	PUNCT
ejpam-1676	463	22	β0)/n	β0)/n	NOUN
ejpam-1676	463	23	r	r	NOUN
ejpam-1676	463	24	1	1	NUM
ejpam-1676	463	25	are	be	AUX
ejpam-1676	463	26	bounded	bound	VERB
ejpam-1676	463	27	by	by	ADP
ejpam-1676	463	28	assumption	assumption	NOUN
ejpam-1676	463	29	(	(	PUNCT
ejpam-1676	463	30	a4	a4	NOUN
ejpam-1676	463	31	)	)	PUNCT
ejpam-1676	463	32	,	,	PUNCT
ejpam-1676	463	33	so	so	ADV
ejpam-1676	463	34	are	be	AUX
ejpam-1676	463	35	eigenvalues	eigenvalue	NOUN
ejpam-1676	463	36	of	of	ADP
ejpam-1676	463	37	nr	nr	NOUN
ejpam-1676	463	38	1	1	NUM
ejpam-1676	463	39	i−1	i−1	PROPN
ejpam-1676	463	40	n1	n1	PROPN
ejpam-1676	463	41	(	(	PUNCT
ejpam-1676	463	42	β0	β0	NOUN
ejpam-1676	463	43	)	)	PUNCT
ejpam-1676	463	44	.	.	PUNCT
ejpam-1676	464	1	that	that	PRON
ejpam-1676	464	2	is	be	AUX
ejpam-1676	464	3	c1	c1	PROPN
ejpam-1676	464	4	≤	≤	NOUN
ejpam-1676	465	1	λmin(n	λmin(n	PROPN
ejpam-1676	465	2	r	r	NOUN
ejpam-1676	465	3	1i−1	1i−1	NUM
ejpam-1676	465	4	n1	n1	NOUN
ejpam-1676	465	5	(	(	PUNCT
ejpam-1676	465	6	β0))≤	β0))≤	PUNCT
ejpam-1676	465	7	λmax(n	λmax(n	NOUN
ejpam-1676	465	8	r	r	PROPN
ejpam-1676	465	9	1i−1	1i−1	NUM
ejpam-1676	465	10	n1	n1	PROPN
ejpam-1676	465	11	(	(	PUNCT
ejpam-1676	465	12	β0))≤	β0))≤	NUM
ejpam-1676	465	13	c2	c2	PROPN
ejpam-1676	465	14	,	,	PUNCT
ejpam-1676	465	15	where	where	SCONJ
ejpam-1676	465	16	c1	c1	PROPN
ejpam-1676	465	17	and	and	CCONJ
ejpam-1676	465	18	c2	c2	PROPN
ejpam-1676	465	19	are	be	AUX
ejpam-1676	465	20	constants	constant	NOUN
ejpam-1676	465	21	.	.	PUNCT
ejpam-1676	466	1	because	because	SCONJ
ejpam-1676	466	2	eut	eut	PROPN
ejpam-1676	466	3	j	j	PROPN
ejpam-1676	466	4	uk	uk	PROPN
ejpam-1676	466	5	=	=	SYM
ejpam-1676	466	6	0	0	NUM
ejpam-1676	466	7	,	,	PUNCT
ejpam-1676	466	8	let	let	VERB
ejpam-1676	466	9	ip	ip	NOUN
ejpam-1676	466	10	be	be	AUX
ejpam-1676	466	11	the	the	DET
ejpam-1676	466	12	p	p	PROPN
ejpam-1676	466	13	×	×	NOUN
ejpam-1676	466	14	p	p	NOUN
ejpam-1676	466	15	identity	identity	NOUN
ejpam-1676	466	16	matrix	matrix	NOUN
ejpam-1676	466	17	,	,	PUNCT
ejpam-1676	466	18	we	we	PRON
ejpam-1676	466	19	have	have	VERB
ejpam-1676	466	20	e‖(β	e‖(β	NOUN
ejpam-1676	466	21	−β0	−β0	PROPN
ejpam-1676	466	22	)	)	PUNCT
ejpam-1676	466	23	t	t	PROPN
ejpam-1676	466	24	s1−w	s1−w	PROPN
ejpam-1676	466	25	n1	n1	PROPN
ejpam-1676	466	26	(	(	PUNCT
ejpam-1676	466	27	β0)‖2	β0)‖2	PROPN
ejpam-1676	466	28	≤	≤	NUM
ejpam-1676	466	29	δ2‖w1‖2	δ2‖w1‖2	NOUN
ejpam-1676	466	30	1	1	NUM
ejpam-1676	466	31	n1	n1	NOUN
ejpam-1676	466	32	e	e	X
ejpam-1676	466	33			NOUN
ejpam-1676	466	34			ADP
ejpam-1676	466	35			ADJ
ejpam-1676	466	36	n1∑	n1∑	PROPN
ejpam-1676	466	37	j=1	j=1	PROPN
ejpam-1676	466	38	u	u	PROPN
ejpam-1676	466	39	t	t	PROPN
ejpam-1676	466	40	j	j	PROPN
ejpam-1676	466	41	(	(	PUNCT
ejpam-1676	466	42	n	n	CCONJ
ejpam-1676	466	43	r	r	NOUN
ejpam-1676	466	44	1i−1	1i−1	NUM
ejpam-1676	466	45	n1	n1	NOUN
ejpam-1676	466	46	(	(	PUNCT
ejpam-1676	466	47	β0))u	β0))u	X
ejpam-1676	466	48	j	j	PROPN
ejpam-1676	466	49			PROPN
ejpam-1676	466	50			PROPN
ejpam-1676	466	51			NOUN
ejpam-1676	466	52	≤	≤	ADV
ejpam-1676	466	53	δ2‖w1‖2	δ2‖w1‖2	SYM
ejpam-1676	466	54	1	1	NUM
ejpam-1676	466	55	n1	n1	NOUN
ejpam-1676	466	56	e	e	X
ejpam-1676	466	57			NOUN
ejpam-1676	466	58			ADP
ejpam-1676	466	59			ADJ
ejpam-1676	466	60	n1∑	n1∑	PROPN
ejpam-1676	466	61	j=1	j=1	PROPN
ejpam-1676	467	1	ut	ut	PROPN
ejpam-1676	467	2	j	j	PROPN
ejpam-1676	468	1	[	[	X
ejpam-1676	468	2	c2	c2	PROPN
ejpam-1676	468	3	ip]u	ip]u	VERB
ejpam-1676	468	4	j	j	PROPN
ejpam-1676	468	5			PROPN
ejpam-1676	468	6			PROPN
ejpam-1676	468	7			NOUN
ejpam-1676	468	8	≤	≤	SCONJ
ejpam-1676	468	9	δ2‖w1‖2	δ2‖w1‖2	AUX
ejpam-1676	468	10	c2	c2	VERB
ejpam-1676	468	11	nr	nr	PRON
ejpam-1676	468	12	1c1	1c1	NUM
ejpam-1676	468	13	e	e	NOUN
ejpam-1676	468	14	n	n	PRON
ejpam-1676	468	15	sn1	sn1	PROPN
ejpam-1676	468	16	(	(	PUNCT
ejpam-1676	468	17	β0	β0	PROPN
ejpam-1676	468	18	)	)	PUNCT
ejpam-1676	468	19	t	t	NOUN
ejpam-1676	469	1	[	[	X
ejpam-1676	469	2	nr	nr	PROPN
ejpam-1676	469	3	1i−1	1i−1	NUM
ejpam-1676	469	4	n1	n1	NOUN
ejpam-1676	469	5	(	(	PUNCT
ejpam-1676	469	6	β0)]sn1	β0)]sn1	NOUN
ejpam-1676	469	7	(	(	PUNCT
ejpam-1676	469	8	β0	β0	NOUN
ejpam-1676	469	9	)	)	PUNCT
ejpam-1676	469	10	o	o	NOUN
ejpam-1676	469	11	≤	≤	PROPN
ejpam-1676	470	1	δ2‖w1‖2	δ2‖w1‖2	PROPN
ejpam-1676	470	2	c2	c2	PROPN
ejpam-1676	470	3	nr	nr	PROPN
ejpam-1676	470	4	1	1	PROPN
ejpam-1676	470	5	c1	c1	PROPN
ejpam-1676	470	6	etr	etr	PROPN
ejpam-1676	470	7	n	n	PROPN
ejpam-1676	470	8	sn1	sn1	PROPN
ejpam-1676	470	9	(	(	PUNCT
ejpam-1676	470	10	β0	β0	NOUN
ejpam-1676	470	11	)	)	PUNCT
ejpam-1676	470	12	t[nr	t[nr	NOUN
ejpam-1676	470	13	1	1	NUM
ejpam-1676	470	14	i−1	i−1	PROPN
ejpam-1676	470	15	n1	n1	NOUN
ejpam-1676	470	16	(	(	PUNCT
ejpam-1676	470	17	β0)]sn1	β0)]sn1	NOUN
ejpam-1676	470	18	(	(	PUNCT
ejpam-1676	470	19	β0	β0	NOUN
ejpam-1676	470	20	)	)	PUNCT
ejpam-1676	470	21	o	o	NOUN
ejpam-1676	471	1	=	=	PUNCT
ejpam-1676	471	2	δ2‖w1‖2	δ2‖w1‖2	PROPN
ejpam-1676	471	3	c2p	c2p	NOUN
ejpam-1676	471	4	c1	c1	PROPN
ejpam-1676	471	5	.	.	PUNCT
ejpam-1676	472	1	because	because	SCONJ
ejpam-1676	472	2	maxi(1−wi	maxi(1−wi	PROPN
ejpam-1676	472	3	)	)	PUNCT
ejpam-1676	472	4	=	=	SYM
ejpam-1676	472	5	o(n−1/2	o(n−1/2	PROPN
ejpam-1676	472	6	)	)	PUNCT
ejpam-1676	472	7	,	,	PUNCT
ejpam-1676	472	8	‖w1‖2	‖w1‖2	NOUN
ejpam-1676	472	9	=	=	SYM
ejpam-1676	472	10	o(1	o(1	PROPN
ejpam-1676	472	11	)	)	PUNCT
ejpam-1676	472	12	.	.	PUNCT
ejpam-1676	473	1	then	then	ADV
ejpam-1676	473	2	δ2‖w1‖2	δ2‖w1‖2	PROPN
ejpam-1676	473	3	c2p	c2p	PROPN
ejpam-1676	473	4	c1	c1	PROPN
ejpam-1676	473	5	=	=	PUNCT
ejpam-1676	473	6	o(1	o(1	PROPN
ejpam-1676	473	7	)	)	PUNCT
ejpam-1676	473	8	.	.	PUNCT
ejpam-1676	474	1	so	so	ADV
ejpam-1676	474	2	e‖(β	e‖(β	PROPN
ejpam-1676	474	3	−β0	−β0	PROPN
ejpam-1676	474	4	)	)	PUNCT
ejpam-1676	474	5	t	t	PROPN
ejpam-1676	474	6	s1−w	s1−w	PROPN
ejpam-1676	474	7	n1	n1	PROPN
ejpam-1676	474	8	(	(	PUNCT
ejpam-1676	474	9	β0)‖2→	β0)‖2→	PROPN
ejpam-1676	474	10	0	0	NUM
ejpam-1676	474	11	.	.	PUNCT
ejpam-1676	475	1	this	this	PRON
ejpam-1676	475	2	proves	prove	VERB
ejpam-1676	475	3	that	that	SCONJ
ejpam-1676	475	4	the	the	DET
ejpam-1676	475	5	first	first	ADJ
ejpam-1676	475	6	term	term	NOUN
ejpam-1676	475	7	is	be	AUX
ejpam-1676	475	8	op(1	op(1	NOUN
ejpam-1676	475	9	)	)	PUNCT
ejpam-1676	475	10	.	.	PUNCT
ejpam-1676	476	1	in	in	ADP
ejpam-1676	476	2	order	order	NOUN
ejpam-1676	476	3	to	to	PART
ejpam-1676	476	4	show	show	VERB
ejpam-1676	476	5	the	the	DET
ejpam-1676	476	6	second	second	ADJ
ejpam-1676	476	7	term	term	NOUN
ejpam-1676	476	8	is	be	AUX
ejpam-1676	476	9	op(1	op(1	NOUN
ejpam-1676	476	10	)	)	PUNCT
ejpam-1676	476	11	,	,	PUNCT
ejpam-1676	476	12	we	we	PRON
ejpam-1676	476	13	need	need	VERB
ejpam-1676	476	14	to	to	PART
ejpam-1676	476	15	show	show	VERB
ejpam-1676	476	16	‖(β−β0	‖(β−β0	ADV
ejpam-1676	476	17	)	)	PUNCT
ejpam-1676	476	18	t∇s1−w	t∇s1−w	NOUN
ejpam-1676	476	19	n1	n1	NOUN
ejpam-1676	476	20	(	(	PUNCT
ejpam-1676	476	21	γ∗)(β−β0)‖	γ∗)(β−β0)‖	PROPN
ejpam-1676	476	22	is	be	AUX
ejpam-1676	476	23	op(1	op(1	NOUN
ejpam-1676	476	24	)	)	PUNCT
ejpam-1676	476	25	.	.	PUNCT
ejpam-1676	477	1	since	since	SCONJ
ejpam-1676	477	2	(	(	PUNCT
ejpam-1676	477	3	β	β	X
ejpam-1676	477	4	−β0	−β0	PROPN
ejpam-1676	477	5	)	)	PUNCT
ejpam-1676	477	6	t∇s1−w	t∇s1−w	NOUN
ejpam-1676	477	7	n1	n1	NOUN
ejpam-1676	477	8	(	(	PUNCT
ejpam-1676	477	9	γ)(β	γ)(β	NOUN
ejpam-1676	477	10	−β0	−β0	PROPN
ejpam-1676	477	11	)	)	PUNCT
ejpam-1676	477	12	=	=	SYM
ejpam-1676	477	13	δ	δ	PROPN
ejpam-1676	477	14	2λt	2λt	NOUN
ejpam-1676	477	15	i−1/2	i−1/2	PROPN
ejpam-1676	477	16	n1	n1	PROPN
ejpam-1676	477	17	(	(	PUNCT
ejpam-1676	477	18	β0)∇s1−w	β0)∇s1−w	PROPN
ejpam-1676	477	19	n1	n1	ADJ
ejpam-1676	477	20	(	(	PUNCT
ejpam-1676	477	21	γ)i−1/2	γ)i−1/2	PROPN
ejpam-1676	477	22	n1	n1	PROPN
ejpam-1676	477	23	(	(	PUNCT
ejpam-1676	477	24	β0)λ	β0)λ	ADJ
ejpam-1676	477	25	,	,	PUNCT
ejpam-1676	477	26	it	it	PRON
ejpam-1676	477	27	is	be	AUX
ejpam-1676	477	28	sufficient	sufficient	ADJ
ejpam-1676	477	29	to	to	PART
ejpam-1676	477	30	show	show	VERB
ejpam-1676	477	31	max	max	PROPN
ejpam-1676	477	32	γ∈nn1	γ∈nn1	PUNCT
ejpam-1676	477	33	(	(	PUNCT
ejpam-1676	477	34	δ	δ	PROPN
ejpam-1676	477	35	)	)	PUNCT
ejpam-1676	477	36	‖i−1/2	‖i−1/2	PROPN
ejpam-1676	477	37	n1	n1	PROPN
ejpam-1676	477	38	(	(	PUNCT
ejpam-1676	477	39	β0)∇s1−w	β0)∇s1−w	PROPN
ejpam-1676	477	40	n1	n1	ADJ
ejpam-1676	477	41	(	(	PUNCT
ejpam-1676	477	42	γ)i−1/2	γ)i−1/2	PROPN
ejpam-1676	477	43	n1	n1	NOUN
ejpam-1676	477	44	(	(	PUNCT
ejpam-1676	477	45	β0)‖	β0)‖	PUNCT
ejpam-1676	477	46	→	→	SYM
ejpam-1676	477	47	0	0	X
ejpam-1676	477	48	.	.	PUNCT
ejpam-1676	478	1	let	let	VERB
ejpam-1676	478	2	rw	rw	PROPN
ejpam-1676	478	3	n1	n1	PROPN
ejpam-1676	478	4	(	(	PUNCT
ejpam-1676	478	5	β	β	X
ejpam-1676	478	6	)	)	PUNCT
ejpam-1676	479	1	=	=	SYM
ejpam-1676	479	2	n1∑	n1∑	PROPN
ejpam-1676	479	3	j=1	j=1	PROPN
ejpam-1676	479	4	(	(	PUNCT
ejpam-1676	479	5	1−w1	1−w1	NUM
ejpam-1676	479	6	j)[y1	j)[y1	PROPN
ejpam-1676	479	7	j	j	PROPN
ejpam-1676	479	8	−µ(ψ(β	−µ(ψ(β	PROPN
ejpam-1676	479	9	t	t	PROPN
ejpam-1676	479	10	x	x	PART
ejpam-1676	479	11	j))]ψ	j))]ψ	PROPN
ejpam-1676	479	12	′′(β	′′(β	NOUN
ejpam-1676	479	13	t	t	PROPN
ejpam-1676	479	14	x	x	SYM
ejpam-1676	479	15	j)t	j)t	NOUN
ejpam-1676	479	16	jx	jx	PROPN
ejpam-1676	479	17	jx	jx	PROPN
ejpam-1676	479	18	t	t	PROPN
ejpam-1676	479	19	j	j	PROPN
ejpam-1676	479	20	,	,	PUNCT
ejpam-1676	479	21	m	m	PROPN
ejpam-1676	479	22	w	w	PROPN
ejpam-1676	479	23	n1	n1	PROPN
ejpam-1676	479	24	(	(	PUNCT
ejpam-1676	479	25	β	β	X
ejpam-1676	479	26	)	)	PUNCT
ejpam-1676	480	1	=	=	SYM
ejpam-1676	480	2	n1∑	n1∑	PROPN
ejpam-1676	480	3	j=1	j=1	PROPN
ejpam-1676	480	4	(	(	PUNCT
ejpam-1676	480	5	1−w1	1−w1	NUM
ejpam-1676	480	6	j)[ψ	j)[ψ	PROPN
ejpam-1676	480	7	′(β	′(β	VERB
ejpam-1676	480	8	t	t	PROPN
ejpam-1676	480	9	x	x	X
ejpam-1676	480	10	j	j	PROPN
ejpam-1676	480	11	)	)	PUNCT
ejpam-1676	480	12	]	]	PUNCT
ejpam-1676	480	13	2ζ′′(ψ(β	2ζ′′(ψ(β	NUM
ejpam-1676	480	14	t	t	X
ejpam-1676	480	15	x	x	PUNCT
ejpam-1676	480	16	j))t	j))t	PROPN
ejpam-1676	480	17	ix	ix	PROPN
ejpam-1676	480	18	jx	jx	PROPN
ejpam-1676	480	19	t	t	PROPN
ejpam-1676	480	20	j	j	PROPN
ejpam-1676	480	21	.	.	PUNCT
ejpam-1676	481	1	then	then	ADV
ejpam-1676	481	2	∇s1−w	∇s1−w	PROPN
ejpam-1676	481	3	n1	n1	NOUN
ejpam-1676	481	4	(	(	PUNCT
ejpam-1676	481	5	γ	γ	NOUN
ejpam-1676	481	6	)	)	PUNCT
ejpam-1676	481	7	=	=	PUNCT
ejpam-1676	482	1	[	[	X
ejpam-1676	482	2	rw	rw	NOUN
ejpam-1676	482	3	n1	n1	PROPN
ejpam-1676	482	4	(	(	PUNCT
ejpam-1676	482	5	γ)−m	γ)−m	PROPN
ejpam-1676	482	6	w	w	PROPN
ejpam-1676	482	7	n1	n1	PROPN
ejpam-1676	482	8	(	(	PUNCT
ejpam-1676	482	9	γ)]/φ	γ)]/φ	PROPN
ejpam-1676	482	10	.	.	PUNCT
ejpam-1676	483	1	so	so	ADV
ejpam-1676	483	2	it	it	PRON
ejpam-1676	483	3	suffices	suffice	VERB
ejpam-1676	483	4	to	to	PART
ejpam-1676	483	5	show	show	VERB
ejpam-1676	483	6	max	max	PROPN
ejpam-1676	483	7	γ∈nn1	γ∈nn1	PUNCT
ejpam-1676	483	8	(	(	PUNCT
ejpam-1676	483	9	δ	δ	NOUN
ejpam-1676	483	10	)	)	PUNCT
ejpam-1676	483	11	‖m−1/2	‖m−1/2	PROPN
ejpam-1676	484	1	n1	n1	PROPN
ejpam-1676	484	2	m	m	PROPN
ejpam-1676	484	3	w	w	PROPN
ejpam-1676	484	4	n1	n1	PROPN
ejpam-1676	484	5	(	(	PUNCT
ejpam-1676	484	6	γ)m−1/2	γ)m−1/2	PROPN
ejpam-1676	484	7	n1	n1	PROPN
ejpam-1676	484	8	‖	‖	PROPN
ejpam-1676	484	9	→	→	SYM
ejpam-1676	484	10	0	0	NUM
ejpam-1676	484	11	,	,	PUNCT
ejpam-1676	484	12	p.	p.	NOUN
ejpam-1676	484	13	guo	guo	PROPN
ejpam-1676	484	14	,	,	PUNCT
ejpam-1676	484	15	x.	x.	PROPN
ejpam-1676	484	16	wang	wang	PROPN
ejpam-1676	484	17	,	,	PUNCT
ejpam-1676	484	18	y.	y.	PROPN
ejpam-1676	484	19	wu	wu	PROPN
ejpam-1676	484	20	/	/	SYM
ejpam-1676	484	21	eur	eur	PROPN
ejpam-1676	484	22	.	.	PUNCT
ejpam-1676	485	1	j.	j.	PROPN
ejpam-1676	485	2	pure	pure	PROPN
ejpam-1676	485	3	appl	appl	PROPN
ejpam-1676	485	4	.	.	PROPN
ejpam-1676	485	5	math	math	PROPN
ejpam-1676	485	6	,	,	PUNCT
ejpam-1676	485	7	5	5	NUM
ejpam-1676	485	8	(	(	PUNCT
ejpam-1676	485	9	2012	2012	NUM
ejpam-1676	485	10	)	)	PUNCT
ejpam-1676	485	11	,	,	PUNCT
ejpam-1676	485	12	333	333	NUM
ejpam-1676	485	13	-	-	SYM
ejpam-1676	485	14	356	356	NUM
ejpam-1676	485	15	348	348	NUM
ejpam-1676	485	16	and	and	CCONJ
ejpam-1676	485	17	max	max	PROPN
ejpam-1676	485	18	γ∈nn1	γ∈nn1	PUNCT
ejpam-1676	485	19	(	(	PUNCT
ejpam-1676	485	20	δ	δ	NOUN
ejpam-1676	485	21	)	)	PUNCT
ejpam-1676	485	22	‖m−1/2	‖m−1/2	PROPN
ejpam-1676	485	23	n1	n1	PROPN
ejpam-1676	485	24	rw	rw	NOUN
ejpam-1676	485	25	n1	n1	PROPN
ejpam-1676	485	26	(	(	PUNCT
ejpam-1676	485	27	γ)m−1/2	γ)m−1/2	PROPN
ejpam-1676	485	28	n1	n1	PROPN
ejpam-1676	485	29	‖	‖	PROPN
ejpam-1676	485	30	→	→	SYM
ejpam-1676	485	31	0	0	NUM
ejpam-1676	485	32	.	.	PUNCT
ejpam-1676	486	1	because	because	SCONJ
ejpam-1676	486	2	‖m−1/2	‖m−1/2	PROPN
ejpam-1676	486	3	n1	n1	PROPN
ejpam-1676	486	4	m	m	PROPN
ejpam-1676	486	5	w	w	PROPN
ejpam-1676	486	6	n1	n1	PROPN
ejpam-1676	486	7	(	(	PUNCT
ejpam-1676	486	8	γ)m−1/2	γ)m−1/2	PROPN
ejpam-1676	486	9	n1	n1	PROPN
ejpam-1676	486	10	‖	‖	PROPN
ejpam-1676	486	11	≤max	≤max	PROPN
ejpam-1676	486	12	j	j	PROPN
ejpam-1676	486	13	(	(	PUNCT
ejpam-1676	486	14	1−w1	1−w1	NUM
ejpam-1676	486	15	j)‖m−1/2	j)‖m−1/2	PROPN
ejpam-1676	486	16	n1	n1	PROPN
ejpam-1676	486	17	mn1	mn1	PROPN
ejpam-1676	486	18	(	(	PUNCT
ejpam-1676	486	19	γ)m−1/2	γ)m−1/2	PROPN
ejpam-1676	486	20	n1	n1	ADJ
ejpam-1676	486	21	‖.	‖.	NOUN
ejpam-1676	486	22	using	use	VERB
ejpam-1676	486	23	similar	similar	ADJ
ejpam-1676	486	24	argument	argument	NOUN
ejpam-1676	486	25	as	as	ADP
ejpam-1676	486	26	in	in	ADP
ejpam-1676	486	27	[	[	X
ejpam-1676	486	28	1	1	NUM
ejpam-1676	486	29	]	]	PUNCT
ejpam-1676	486	30	,	,	PUNCT
ejpam-1676	486	31	‖m−1/2	‖m−1/2	ADJ
ejpam-1676	486	32	n1	n1	PROPN
ejpam-1676	486	33	mn1	mn1	PROPN
ejpam-1676	486	34	(	(	PUNCT
ejpam-1676	486	35	γ)m−1/2	γ)m−1/2	PROPN
ejpam-1676	486	36	n1	n1	PROPN
ejpam-1676	486	37	‖	‖	PROPN
ejpam-1676	486	38	is	be	AUX
ejpam-1676	486	39	bounded	bound	VERB
ejpam-1676	486	40	by	by	ADP
ejpam-1676	486	41	max	max	PROPN
ejpam-1676	486	42	γ∈nn1	γ∈nn1	PUNCT
ejpam-1676	486	43	(	(	PUNCT
ejpam-1676	486	44	δ	δ	NOUN
ejpam-1676	486	45	)	)	PUNCT
ejpam-1676	486	46	‖m−1/2	‖m−1/2	NOUN
ejpam-1676	486	47	n1	n1	PROPN
ejpam-1676	486	48	mn1	mn1	PROPN
ejpam-1676	486	49	(	(	PUNCT
ejpam-1676	486	50	γ)m−1/2	γ)m−1/2	PROPN
ejpam-1676	486	51	n1	n1	PROPN
ejpam-1676	486	52	‖	‖	PROPN
ejpam-1676	486	53	≤	≤	PROPN
ejpam-1676	487	1	pp	pp	ADP
ejpam-1676	487	2	max	max	PROPN
ejpam-1676	487	3	γ∈nn1	γ∈nn1	PUNCT
ejpam-1676	487	4	(	(	PUNCT
ejpam-1676	487	5	δ	δ	NOUN
ejpam-1676	487	6	)	)	PUNCT
ejpam-1676	487	7	,	,	PUNCT
ejpam-1676	487	8	j≤n1	j≤n1	PROPN
ejpam-1676	487	9	|ϕ(γt	|ϕ(γt	PROPN
ejpam-1676	487	10	x	x	SYM
ejpam-1676	487	11	j)/ϕ(β	j)/ϕ(β	NOUN
ejpam-1676	487	12	t	t	NOUN
ejpam-1676	487	13	x	x	SYM
ejpam-1676	487	14	j)|	j)|	PROPN
ejpam-1676	487	15	,	,	PUNCT
ejpam-1676	487	16	which	which	PRON
ejpam-1676	487	17	converges	converge	VERB
ejpam-1676	487	18	to	to	ADP
ejpam-1676	487	19	0	0	NUM
ejpam-1676	487	20	since	since	SCONJ
ejpam-1676	487	21	ϕ	ϕ	PROPN
ejpam-1676	487	22	is	be	AUX
ejpam-1676	487	23	continuous	continuous	ADJ
ejpam-1676	487	24	and	and	CCONJ
ejpam-1676	487	25	,	,	PUNCT
ejpam-1676	487	26	for	for	ADP
ejpam-1676	487	27	γ	γ	PROPN
ejpam-1676	487	28	∈	∈	PROPN
ejpam-1676	487	29	nn1	nn1	PROPN
ejpam-1676	487	30	(	(	PUNCT
ejpam-1676	487	31	δ	δ	PROPN
ejpam-1676	487	32	)	)	PUNCT
ejpam-1676	487	33	,	,	PUNCT
ejpam-1676	487	34	|γt	|γt	PROPN
ejpam-1676	487	35	x	x	SYM
ejpam-1676	487	36	j−β	j−β	PROPN
ejpam-1676	487	37	t	t	PROPN
ejpam-1676	487	38	x	x	SYM
ejpam-1676	487	39	j|2→	j|2→	PROPN
ejpam-1676	487	40	0	0	NUM
ejpam-1676	487	41	.	.	PUNCT
ejpam-1676	488	1	because	because	SCONJ
ejpam-1676	488	2	max	max	PROPN
ejpam-1676	488	3	j(1−w1	j(1−w1	PROPN
ejpam-1676	488	4	j)→	j)→	PROPN
ejpam-1676	488	5	0	0	NUM
ejpam-1676	488	6	,	,	PUNCT
ejpam-1676	488	7	we	we	PRON
ejpam-1676	488	8	have	have	VERB
ejpam-1676	488	9	max	max	PROPN
ejpam-1676	488	10	γ∈nn1	γ∈nn1	PUNCT
ejpam-1676	488	11	(	(	PUNCT
ejpam-1676	488	12	δ	δ	NOUN
ejpam-1676	488	13	)	)	PUNCT
ejpam-1676	488	14	‖m−1/2	‖m−1/2	PROPN
ejpam-1676	489	1	n1	n1	PROPN
ejpam-1676	489	2	m	m	PROPN
ejpam-1676	489	3	w	w	PROPN
ejpam-1676	489	4	n1	n1	PROPN
ejpam-1676	489	5	(	(	PUNCT
ejpam-1676	489	6	γ)m−1/2	γ)m−1/2	PROPN
ejpam-1676	489	7	n1	n1	PROPN
ejpam-1676	489	8	‖	‖	PROPN
ejpam-1676	489	9	→	→	SYM
ejpam-1676	489	10	0	0	X
ejpam-1676	489	11	.	.	PUNCT
ejpam-1676	490	1	let	let	VERB
ejpam-1676	491	1	e	e	NOUN
ejpam-1676	491	2	j	j	X
ejpam-1676	491	3	=	=	SYM
ejpam-1676	491	4	y1	y1	PROPN
ejpam-1676	491	5	j	j	X
ejpam-1676	491	6	−µ(ψ(βt	−µ(ψ(βt	PROPN
ejpam-1676	491	7	x	x	SYM
ejpam-1676	491	8	j	j	PROPN
ejpam-1676	491	9	)	)	PUNCT
ejpam-1676	491	10	)	)	PUNCT
ejpam-1676	491	11	,	,	PUNCT
ejpam-1676	491	12	uw	uw	PROPN
ejpam-1676	491	13	n1	n1	PROPN
ejpam-1676	491	14	(	(	PUNCT
ejpam-1676	491	15	γ	γ	NOUN
ejpam-1676	491	16	)	)	PUNCT
ejpam-1676	491	17	=	=	SYM
ejpam-1676	491	18	n1∑	n1∑	PROPN
ejpam-1676	491	19	j=1	j=1	PROPN
ejpam-1676	491	20	(	(	PUNCT
ejpam-1676	491	21	1−w1	1−w1	NUM
ejpam-1676	491	22	j)[µ(ψ(β	j)[µ(ψ(β	NOUN
ejpam-1676	491	23	t	t	NOUN
ejpam-1676	491	24	x	x	SYM
ejpam-1676	491	25	j))−µ(ψ(γt	j))−µ(ψ(γt	ADJ
ejpam-1676	491	26	x	x	PART
ejpam-1676	491	27	j))]ψ	j))]ψ	PROPN
ejpam-1676	491	28	′′(β	′′(β	NOUN
ejpam-1676	491	29	t	t	PROPN
ejpam-1676	491	30	x	x	SYM
ejpam-1676	491	31	j)t	j)t	NOUN
ejpam-1676	491	32	jx	jx	PROPN
ejpam-1676	491	33	jx	jx	PROPN
ejpam-1676	491	34	t	t	PROPN
ejpam-1676	491	35	j	j	PROPN
ejpam-1676	491	36	,	,	PUNCT
ejpam-1676	491	37	v	v	PROPN
ejpam-1676	491	38	w	w	PROPN
ejpam-1676	491	39	n1	n1	PROPN
ejpam-1676	491	40	(	(	PUNCT
ejpam-1676	491	41	γ	γ	NOUN
ejpam-1676	491	42	)	)	PUNCT
ejpam-1676	491	43	=	=	SYM
ejpam-1676	491	44	n1∑	n1∑	PROPN
ejpam-1676	491	45	j=1	j=1	PROPN
ejpam-1676	491	46	(	(	PUNCT
ejpam-1676	491	47	1−w1	1−w1	NUM
ejpam-1676	491	48	j)e	j)e	X
ejpam-1676	491	49	j[ψ	j[ψ	X
ejpam-1676	491	50	′′(γt	′′(γt	NUM
ejpam-1676	491	51	x	x	X
ejpam-1676	491	52	j)−ψ′′(β	j)−ψ′′(β	NOUN
ejpam-1676	491	53	t	t	NOUN
ejpam-1676	491	54	x	x	PUNCT
ejpam-1676	491	55	j)]t	j)]t	PROPN
ejpam-1676	491	56	ix	ix	INTJ
ejpam-1676	491	57	jx	jx	PROPN
ejpam-1676	491	58	t	t	PROPN
ejpam-1676	491	59	j	j	PROPN
ejpam-1676	491	60	,	,	PUNCT
ejpam-1676	491	61	w	w	PROPN
ejpam-1676	491	62	w	w	PROPN
ejpam-1676	491	63	n1	n1	PROPN
ejpam-1676	491	64	(	(	PUNCT
ejpam-1676	491	65	β	β	X
ejpam-1676	491	66	)	)	PUNCT
ejpam-1676	491	67	=	=	SYM
ejpam-1676	492	1	n1∑	n1∑	PROPN
ejpam-1676	492	2	j=1	j=1	PROPN
ejpam-1676	492	3	(	(	PUNCT
ejpam-1676	492	4	1−w1	1−w1	NUM
ejpam-1676	492	5	j)e	j)e	X
ejpam-1676	492	6	jψ	jψ	PROPN
ejpam-1676	492	7	′′(β	′′(β	PROPN
ejpam-1676	492	8	t	t	PROPN
ejpam-1676	492	9	x	x	PUNCT
ejpam-1676	492	10	j)]t	j)]t	ADV
ejpam-1676	492	11	ix	ix	INTJ
ejpam-1676	492	12	jx	jx	PROPN
ejpam-1676	492	13	t	t	PROPN
ejpam-1676	492	14	j	j	PROPN
ejpam-1676	492	15	.	.	PUNCT
ejpam-1676	493	1	then	then	ADV
ejpam-1676	493	2	rw	rw	PROPN
ejpam-1676	493	3	n1	n1	PROPN
ejpam-1676	493	4	(	(	PUNCT
ejpam-1676	493	5	γ	γ	NOUN
ejpam-1676	493	6	)	)	PUNCT
ejpam-1676	493	7	=	=	SYM
ejpam-1676	493	8	uw	uw	PROPN
ejpam-1676	493	9	n1	n1	PROPN
ejpam-1676	493	10	(	(	PUNCT
ejpam-1676	493	11	γ	γ	X
ejpam-1676	493	12	)	)	PUNCT
ejpam-1676	493	13	+	+	NOUN
ejpam-1676	493	14	v	v	NUM
ejpam-1676	493	15	w	w	NOUN
ejpam-1676	493	16	n1	n1	PROPN
ejpam-1676	493	17	(	(	PUNCT
ejpam-1676	493	18	γ	γ	X
ejpam-1676	493	19	)	)	PUNCT
ejpam-1676	494	1	+	+	PROPN
ejpam-1676	494	2	w	w	PROPN
ejpam-1676	494	3	w	w	PROPN
ejpam-1676	494	4	n1	n1	PROPN
ejpam-1676	494	5	(	(	PUNCT
ejpam-1676	494	6	β	β	NOUN
ejpam-1676	494	7	)	)	PUNCT
ejpam-1676	494	8	.	.	PUNCT
ejpam-1676	495	1	using	use	VERB
ejpam-1676	495	2	a	a	DET
ejpam-1676	495	3	similar	similar	ADJ
ejpam-1676	495	4	argument	argument	NOUN
ejpam-1676	495	5	as	as	ADP
ejpam-1676	495	6	above	above	ADV
ejpam-1676	495	7	,	,	PUNCT
ejpam-1676	495	8	we	we	PRON
ejpam-1676	495	9	can	can	AUX
ejpam-1676	495	10	show	show	VERB
ejpam-1676	495	11	that	that	SCONJ
ejpam-1676	495	12	max	max	PROPN
ejpam-1676	495	13	γ∈nn1	γ∈nn1	PUNCT
ejpam-1676	495	14	(	(	PUNCT
ejpam-1676	495	15	δ	δ	NOUN
ejpam-1676	495	16	)	)	PUNCT
ejpam-1676	495	17	‖m−1/2	‖m−1/2	PROPN
ejpam-1676	495	18	n1	n1	PROPN
ejpam-1676	495	19	uw	uw	PROPN
ejpam-1676	495	20	n1	n1	PROPN
ejpam-1676	495	21	(	(	PUNCT
ejpam-1676	495	22	γ)m−1/2	γ)m−1/2	PROPN
ejpam-1676	495	23	n1	n1	PROPN
ejpam-1676	495	24	‖	‖	PROPN
ejpam-1676	495	25	→	→	SYM
ejpam-1676	495	26	0	0	X
ejpam-1676	495	27	.	.	X
ejpam-1676	495	28	note	note	VERB
ejpam-1676	495	29	that	that	SCONJ
ejpam-1676	495	30	‖m−1/2	‖m−1/2	PROPN
ejpam-1676	495	31	n1	n1	PROPN
ejpam-1676	495	32	uw	uw	PROPN
ejpam-1676	495	33	n1	n1	PROPN
ejpam-1676	495	34	(	(	PUNCT
ejpam-1676	495	35	γ)m−1/2	γ)m−1/2	PROPN
ejpam-1676	495	36	n1	n1	PROPN
ejpam-1676	495	37	‖	‖	PROPN
ejpam-1676	495	38	is	be	AUX
ejpam-1676	495	39	bounded	bound	VERB
ejpam-1676	495	40	by	by	ADP
ejpam-1676	495	41	the	the	DET
ejpam-1676	495	42	product	product	NOUN
ejpam-1676	495	43	of	of	ADP
ejpam-1676	495	44	max	max	PROPN
ejpam-1676	495	45	j	j	PROPN
ejpam-1676	495	46	(	(	PUNCT
ejpam-1676	495	47	1−w1	1−w1	NUM
ejpam-1676	495	48	j)m	j)m	NOUN
ejpam-1676	495	49	−1/2	−1/2	VERB
ejpam-1676	495	50	n1	n1	PROPN
ejpam-1676	495	51	n1∑	n1∑	PROPN
ejpam-1676	495	52	j=1	j=1	PROPN
ejpam-1676	495	53	|e	|e	VERB
ejpam-1676	495	54	j	j	PROPN
ejpam-1676	495	55	|t	|t	PROPN
ejpam-1676	496	1	jx	jx	PROPN
ejpam-1676	496	2	jx	jx	PROPN
ejpam-1676	496	3	t	t	PROPN
ejpam-1676	496	4	j	j	PROPN
ejpam-1676	496	5	m−1/2	m−1/2	PROPN
ejpam-1676	496	6	n1	n1	PROPN
ejpam-1676	496	7	=	=	SYM
ejpam-1676	496	8	op(1	op(1	NOUN
ejpam-1676	496	9	)	)	PUNCT
ejpam-1676	496	10	and	and	CCONJ
ejpam-1676	496	11	max	max	PROPN
ejpam-1676	496	12	γ∈nn1	γ∈nn1	PUNCT
ejpam-1676	496	13	(	(	PUNCT
ejpam-1676	496	14	δ	δ	NOUN
ejpam-1676	496	15	)	)	PUNCT
ejpam-1676	496	16	,	,	PUNCT
ejpam-1676	496	17	j≤n1	j≤n1	ADV
ejpam-1676	496	18	|ψ′′(γt	|ψ′′(γt	NOUN
ejpam-1676	496	19	x	x	X
ejpam-1676	496	20	j)−ψ′′(β	j)−ψ′′(β	NOUN
ejpam-1676	496	21	t	t	PROPN
ejpam-1676	496	22	x	x	SYM
ejpam-1676	496	23	j)|	j)|	PROPN
ejpam-1676	496	24	,	,	PUNCT
ejpam-1676	496	25	which	which	PRON
ejpam-1676	496	26	can	can	AUX
ejpam-1676	496	27	be	be	AUX
ejpam-1676	496	28	shown	show	VERB
ejpam-1676	496	29	to	to	PART
ejpam-1676	496	30	be	be	AUX
ejpam-1676	496	31	o(1	o(1	NOUN
ejpam-1676	496	32	)	)	PUNCT
ejpam-1676	496	33	using	use	VERB
ejpam-1676	496	34	the	the	DET
ejpam-1676	496	35	same	same	ADJ
ejpam-1676	496	36	argument	argument	NOUN
ejpam-1676	496	37	.	.	PUNCT
ejpam-1676	497	1	hence	hence	ADV
ejpam-1676	497	2	,	,	PUNCT
ejpam-1676	497	3	max	max	PROPN
ejpam-1676	497	4	γ∈nn1	γ∈nn1	PUNCT
ejpam-1676	497	5	(	(	PUNCT
ejpam-1676	497	6	δ	δ	NOUN
ejpam-1676	497	7	)	)	PUNCT
ejpam-1676	497	8	‖m−1/2	‖m−1/2	PROPN
ejpam-1676	497	9	n1	n1	PROPN
ejpam-1676	497	10	v	v	PROPN
ejpam-1676	497	11	w	w	PROPN
ejpam-1676	497	12	n1	n1	PROPN
ejpam-1676	497	13	(	(	PUNCT
ejpam-1676	497	14	γ)m−1/2	γ)m−1/2	PROPN
ejpam-1676	497	15	n1	n1	PROPN
ejpam-1676	497	16	‖	‖	PROPN
ejpam-1676	497	17	→	→	SYM
ejpam-1676	497	18	0	0	NUM
ejpam-1676	497	19	.	.	PUNCT
ejpam-1676	498	1	p.	p.	PROPN
ejpam-1676	498	2	guo	guo	PROPN
ejpam-1676	498	3	,	,	PUNCT
ejpam-1676	498	4	x.	x.	PROPN
ejpam-1676	498	5	wang	wang	PROPN
ejpam-1676	498	6	,	,	PUNCT
ejpam-1676	498	7	y.	y.	PROPN
ejpam-1676	498	8	wu	wu	PROPN
ejpam-1676	498	9	/	/	SYM
ejpam-1676	498	10	eur	eur	PROPN
ejpam-1676	498	11	.	.	PUNCT
ejpam-1676	499	1	j.	j.	PROPN
ejpam-1676	499	2	pure	pure	PROPN
ejpam-1676	499	3	appl	appl	PROPN
ejpam-1676	499	4	.	.	PROPN
ejpam-1676	499	5	math	math	PROPN
ejpam-1676	499	6	,	,	PUNCT
ejpam-1676	499	7	5	5	NUM
ejpam-1676	499	8	(	(	PUNCT
ejpam-1676	499	9	2012	2012	NUM
ejpam-1676	499	10	)	)	PUNCT
ejpam-1676	499	11	,	,	PUNCT
ejpam-1676	499	12	333	333	NUM
ejpam-1676	499	13	-	-	SYM
ejpam-1676	499	14	356	356	NUM
ejpam-1676	499	15	349	349	NUM
ejpam-1676	499	16	finally	finally	ADV
ejpam-1676	499	17	we	we	PRON
ejpam-1676	499	18	need	need	VERB
ejpam-1676	499	19	to	to	PART
ejpam-1676	499	20	show	show	VERB
ejpam-1676	499	21	max	max	PROPN
ejpam-1676	499	22	γ∈nn1	γ∈nn1	PUNCT
ejpam-1676	499	23	(	(	PUNCT
ejpam-1676	499	24	δ	δ	NOUN
ejpam-1676	499	25	)	)	PUNCT
ejpam-1676	499	26	‖m−1/2	‖m−1/2	VERB
ejpam-1676	499	27	n1	n1	PROPN
ejpam-1676	499	28	w	w	PROPN
ejpam-1676	499	29	w	w	PROPN
ejpam-1676	499	30	n1	n1	PROPN
ejpam-1676	499	31	(	(	PUNCT
ejpam-1676	499	32	β)m−1/2	β)m−1/2	PROPN
ejpam-1676	499	33	n1	n1	PROPN
ejpam-1676	499	34	‖	‖	PROPN
ejpam-1676	499	35	→	→	SYM
ejpam-1676	499	36	0	0	NUM
ejpam-1676	499	37	.	.	PUNCT
ejpam-1676	500	1	since	since	SCONJ
ejpam-1676	500	2	e(e	e(e	PROPN
ejpam-1676	500	3	j	j	PROPN
ejpam-1676	500	4	)	)	PUNCT
ejpam-1676	500	5	=	=	SYM
ejpam-1676	500	6	0	0	NUM
ejpam-1676	500	7	and	and	CCONJ
ejpam-1676	500	8	e	e	X
ejpam-1676	500	9	j	j	PROPN
ejpam-1676	500	10	’s	’	VERB
ejpam-1676	500	11	are	be	AUX
ejpam-1676	500	12	independent	independent	ADJ
ejpam-1676	500	13	,	,	PUNCT
ejpam-1676	500	14	it	it	PRON
ejpam-1676	500	15	suffices	suffice	VERB
ejpam-1676	500	16	to	to	PART
ejpam-1676	500	17	show	show	VERB
ejpam-1676	500	18	that	that	SCONJ
ejpam-1676	500	19	n1∑	n1∑	PROPN
ejpam-1676	500	20	j=1	j=1	PROPN
ejpam-1676	500	21	e|(1−w1	e|(1−w1	PROPN
ejpam-1676	500	22	j)e	j)e	X
ejpam-1676	501	1	jψ	jψ	PROPN
ejpam-1676	501	2	′′(β	′′(β	PROPN
ejpam-1676	501	3	t	t	PROPN
ejpam-1676	501	4	x	x	SYM
ejpam-1676	501	5	j)t	j)t	NOUN
ejpam-1676	501	6	ix	ix	ADP
ejpam-1676	501	7	t	t	PROPN
ejpam-1676	501	8	j	j	PROPN
ejpam-1676	502	1	m−1	m−1	PROPN
ejpam-1676	502	2	n1	n1	PROPN
ejpam-1676	502	3	x	x	SYM
ejpam-1676	502	4	j|1+τ→	j|1+τ→	X
ejpam-1676	502	5	0	0	NUM
ejpam-1676	502	6	for	for	ADP
ejpam-1676	502	7	some	some	DET
ejpam-1676	502	8	τ	τ	PROPN
ejpam-1676	502	9	∈	∈	PROPN
ejpam-1676	502	10	(	(	PUNCT
ejpam-1676	502	11	0,1	0,1	NUM
ejpam-1676	502	12	)	)	PUNCT
ejpam-1676	502	13	.	.	PUNCT
ejpam-1676	503	1	it	it	PRON
ejpam-1676	503	2	is	be	AUX
ejpam-1676	503	3	trivial	trivial	ADJ
ejpam-1676	503	4	since	since	SCONJ
ejpam-1676	503	5	max	max	PROPN
ejpam-1676	503	6	j(1−w1	j(1−w1	PROPN
ejpam-1676	503	7	j)→	j)→	PROPN
ejpam-1676	503	8	0	0	PUNCT
ejpam-1676	503	9	and	and	CCONJ
ejpam-1676	503	10	it	it	PRON
ejpam-1676	503	11	is	be	AUX
ejpam-1676	503	12	shown	show	VERB
ejpam-1676	503	13	in	in	ADP
ejpam-1676	503	14	[	[	X
ejpam-1676	503	15	1	1	X
ejpam-1676	503	16	]	]	PUNCT
ejpam-1676	503	17	that	that	SCONJ
ejpam-1676	503	18	n1∑	n1∑	PROPN
ejpam-1676	503	19	j=1	j=1	PROPN
ejpam-1676	503	20	e|e	e|e	VERB
ejpam-1676	504	1	jψ	jψ	PROPN
ejpam-1676	504	2	′′(β	′′(β	PROPN
ejpam-1676	504	3	t	t	PROPN
ejpam-1676	504	4	x	x	SYM
ejpam-1676	504	5	j)t	j)t	NOUN
ejpam-1676	504	6	ix	ix	ADP
ejpam-1676	504	7	t	t	PROPN
ejpam-1676	504	8	j	j	PROPN
ejpam-1676	505	1	m−1	m−1	PROPN
ejpam-1676	505	2	n1	n1	PROPN
ejpam-1676	505	3	x	x	SYM
ejpam-1676	505	4	j|1+τ→	j|1+τ→	X
ejpam-1676	505	5	0	0	X
ejpam-1676	505	6	.	.	PUNCT
ejpam-1676	506	1	now	now	ADV
ejpam-1676	506	2	we	we	PRON
ejpam-1676	506	3	have	have	AUX
ejpam-1676	506	4	shown	show	VERB
ejpam-1676	506	5	that	that	SCONJ
ejpam-1676	506	6	the	the	DET
ejpam-1676	506	7	first	first	ADJ
ejpam-1676	506	8	term	term	NOUN
ejpam-1676	506	9	and	and	CCONJ
ejpam-1676	506	10	the	the	DET
ejpam-1676	506	11	remainder	remainder	NOUN
ejpam-1676	506	12	of	of	ADP
ejpam-1676	506	13	the	the	DET
ejpam-1676	506	14	taylor	taylor	PROPN
ejpam-1676	506	15	expansion	expansion	NOUN
ejpam-1676	506	16	of	of	ADP
ejpam-1676	506	17	b	b	NOUN
ejpam-1676	506	18	are	be	AUX
ejpam-1676	506	19	both	both	PRON
ejpam-1676	506	20	op(1	op(1	NOUN
ejpam-1676	506	21	)	)	PUNCT
ejpam-1676	506	22	.	.	PUNCT
ejpam-1676	507	1	so	so	ADV
ejpam-1676	507	2	b	b	X
ejpam-1676	507	3	=	=	PUNCT
ejpam-1676	507	4	op(1	op(1	PROPN
ejpam-1676	507	5	)	)	PUNCT
ejpam-1676	507	6	.	.	PUNCT
ejpam-1676	508	1	in	in	ADP
ejpam-1676	508	2	the	the	DET
ejpam-1676	508	3	following	following	NOUN
ejpam-1676	508	4	we	we	PRON
ejpam-1676	508	5	show	show	VERB
ejpam-1676	508	6	that	that	SCONJ
ejpam-1676	508	7	c	c	PROPN
ejpam-1676	508	8	is	be	AUX
ejpam-1676	508	9	dominated	dominate	VERB
ejpam-1676	508	10	by	by	ADP
ejpam-1676	508	11	a.	a.	NOUN
ejpam-1676	508	12	by	by	ADP
ejpam-1676	508	13	assumption	assumption	NOUN
ejpam-1676	508	14	(	(	PUNCT
ejpam-1676	508	15	a5	a5	PROPN
ejpam-1676	508	16	)	)	PUNCT
ejpam-1676	508	17	,	,	PUNCT
ejpam-1676	508	18	we	we	PRON
ejpam-1676	508	19	have	have	VERB
ejpam-1676	508	20	e|c	e|c	NOUN
ejpam-1676	509	1	|	|	ADV
ejpam-1676	509	2	=	=	SYM
ejpam-1676	509	3	e	e	PROPN
ejpam-1676	509	4	�	�	PROPN
ejpam-1676	509	5	�	�	PROPN
ejpam-1676	509	6	�	�	PROPN
ejpam-1676	509	7	�	�	PROPN
ejpam-1676	509	8	�	�	PROPN
ejpam-1676	509	9	�	�	PROPN
ejpam-1676	509	10	n2∑	n2∑	PROPN
ejpam-1676	509	11	j=1	j=1	PROPN
ejpam-1676	509	12	w2	w2	NOUN
ejpam-1676	509	13	j	j	PROPN
ejpam-1676	509	14	�	�	PROPN
ejpam-1676	509	15	l2	l2	PROPN
ejpam-1676	509	16	j(β)−	j(β)−	PROPN
ejpam-1676	509	17	l2	l2	NOUN
ejpam-1676	509	18	j(β0	j(β0	NOUN
ejpam-1676	509	19	)	)	PUNCT
ejpam-1676	509	20	�	�	PROPN
ejpam-1676	509	21	�	�	PROPN
ejpam-1676	509	22	�	�	PROPN
ejpam-1676	509	23	�	�	PROPN
ejpam-1676	509	24	�	�	PROPN
ejpam-1676	509	25	�	�	PROPN
ejpam-1676	509	26	�	�	PROPN
ejpam-1676	509	27	≤	≤	NUM
ejpam-1676	509	28	n2∑	n2∑	PROPN
ejpam-1676	509	29	j=1	j=1	PROPN
ejpam-1676	509	30	w2	w2	NOUN
ejpam-1676	509	31	j	j	PROPN
ejpam-1676	509	32	e	e	PROPN
ejpam-1676	509	33	�	�	PROPN
ejpam-1676	509	34	�	�	PROPN
ejpam-1676	509	35	�	�	PROPN
ejpam-1676	509	36	�	�	PROPN
ejpam-1676	509	37	�	�	PROPN
ejpam-1676	509	38	log	log	NOUN
ejpam-1676	509	39	f	f	PROPN
ejpam-1676	509	40	(	(	PUNCT
ejpam-1676	509	41	y2	y2	PROPN
ejpam-1676	509	42	j	j	PROPN
ejpam-1676	509	43	,	,	PUNCT
ejpam-1676	509	44	β	β	X
ejpam-1676	509	45	)	)	PUNCT
ejpam-1676	509	46	f	f	PROPN
ejpam-1676	510	1	(	(	PUNCT
ejpam-1676	510	2	y2	y2	PROPN
ejpam-1676	510	3	j	j	PROPN
ejpam-1676	510	4	,	,	PUNCT
ejpam-1676	510	5	β0	β0	PROPN
ejpam-1676	510	6	)	)	PUNCT
ejpam-1676	510	7	�	�	PROPN
ejpam-1676	510	8	�	�	PROPN
ejpam-1676	510	9	�	�	PROPN
ejpam-1676	510	10	�	�	PROPN
ejpam-1676	510	11	�	�	PROPN
ejpam-1676	510	12	≤	≤	NOUN
ejpam-1676	510	13	k	k	X
ejpam-1676	510	14	n2∑	n2∑	PROPN
ejpam-1676	510	15	j=1	j=1	PROPN
ejpam-1676	510	16	w2	w2	NOUN
ejpam-1676	510	17	j	j	PROPN
ejpam-1676	510	18	=	=	SYM
ejpam-1676	510	19	k	k	X
ejpam-1676	510	20	·	·	PUNCT
ejpam-1676	510	21	n2o(n−1/2	n2o(n−1/2	NUM
ejpam-1676	510	22	)	)	PUNCT
ejpam-1676	510	23	→	→	SYM
ejpam-1676	510	24	0	0	X
ejpam-1676	510	25	.	.	PUNCT
ejpam-1676	511	1	this	this	PRON
ejpam-1676	511	2	implies	imply	VERB
ejpam-1676	511	3	c	c	X
ejpam-1676	511	4	→p	→p	PROPN
ejpam-1676	511	5	0	0	NUM
ejpam-1676	511	6	.	.	PUNCT
ejpam-1676	511	7	theorem	theorem	NOUN
ejpam-1676	511	8	3	3	NUM
ejpam-1676	511	9	.	.	PUNCT
ejpam-1676	512	1	under	under	ADP
ejpam-1676	512	2	the	the	DET
ejpam-1676	512	3	assumptions	assumption	NOUN
ejpam-1676	512	4	of	of	ADP
ejpam-1676	512	5	theorem	theorem	NOUN
ejpam-1676	512	6	2	2	NUM
ejpam-1676	512	7	,	,	PUNCT
ejpam-1676	512	8	i1/2	i1/2	ADJ
ejpam-1676	512	9	n1	n1	NOUN
ejpam-1676	512	10	(	(	PUNCT
ejpam-1676	512	11	β0)(β̂	β0)(β̂	NOUN
ejpam-1676	512	12	−	−	NOUN
ejpam-1676	512	13	β0)→	β0)→	SYM
ejpam-1676	512	14	np(0	np(0	NOUN
ejpam-1676	512	15	,	,	PUNCT
ejpam-1676	512	16	ip	ip	NOUN
ejpam-1676	512	17	)	)	PUNCT
ejpam-1676	512	18	in	in	ADP
ejpam-1676	512	19	distribution	distribution	NOUN
ejpam-1676	512	20	.	.	PUNCT
ejpam-1676	513	1	proof	proof	NOUN
ejpam-1676	513	2	.	.	PUNCT
ejpam-1676	514	1	let	let	VERB
ejpam-1676	514	2	sw	sw	PROPN
ejpam-1676	514	3	n	n	PROPN
ejpam-1676	514	4	(	(	PUNCT
ejpam-1676	514	5	β	β	NOUN
ejpam-1676	514	6	)	)	PUNCT
ejpam-1676	514	7	=	=	SYM
ejpam-1676	515	1	∂	∂	NUM
ejpam-1676	516	1	∂β	∂β	PROPN
ejpam-1676	516	2	2∑	2∑	NUM
ejpam-1676	516	3	i=1	i=1	PUNCT
ejpam-1676	517	1	ni∑	ni∑	PROPN
ejpam-1676	517	2	j=1	j=1	PROPN
ejpam-1676	517	3	wi	wi	PROPN
ejpam-1676	517	4	j	j	PROPN
ejpam-1676	517	5	li	li	PROPN
ejpam-1676	517	6	j(β	j(β	PROPN
ejpam-1676	517	7	)	)	PUNCT
ejpam-1676	517	8	and	and	CCONJ
ejpam-1676	517	9	∇sw	∇sw	PROPN
ejpam-1676	517	10	n	n	X
ejpam-1676	517	11	(	(	PUNCT
ejpam-1676	517	12	β	β	NOUN
ejpam-1676	517	13	)	)	PUNCT
ejpam-1676	517	14	=	=	SYM
ejpam-1676	517	15	∂	∂	NUM
ejpam-1676	517	16	∂	∂	NOUN
ejpam-1676	517	17	β	β	X
ejpam-1676	517	18	sw	sw	PROPN
ejpam-1676	517	19	n	n	PROPN
ejpam-1676	517	20	(	(	PUNCT
ejpam-1676	517	21	β	β	NOUN
ejpam-1676	517	22	)	)	PUNCT
ejpam-1676	517	23	.	.	PUNCT
ejpam-1676	518	1	then	then	ADV
ejpam-1676	518	2	∇sw	∇sw	PROPN
ejpam-1676	518	3	n	n	X
ejpam-1676	518	4	(	(	PUNCT
ejpam-1676	518	5	β	β	NOUN
ejpam-1676	518	6	)	)	PUNCT
ejpam-1676	518	7	=	=	PROPN
ejpam-1676	518	8	∇sw	∇sw	PROPN
ejpam-1676	518	9	n1	n1	PROPN
ejpam-1676	518	10	(	(	PUNCT
ejpam-1676	518	11	β	β	X
ejpam-1676	518	12	)	)	PUNCT
ejpam-1676	518	13	+	+	ADJ
ejpam-1676	518	14	∇sw	∇sw	PROPN
ejpam-1676	518	15	n2	n2	NOUN
ejpam-1676	518	16	(	(	PUNCT
ejpam-1676	518	17	β	β	NOUN
ejpam-1676	518	18	)	)	PUNCT
ejpam-1676	518	19	,	,	PUNCT
ejpam-1676	518	20	p.	p.	PROPN
ejpam-1676	518	21	guo	guo	PROPN
ejpam-1676	518	22	,	,	PUNCT
ejpam-1676	518	23	x.	x.	PROPN
ejpam-1676	518	24	wang	wang	PROPN
ejpam-1676	518	25	,	,	PUNCT
ejpam-1676	518	26	y.	y.	PROPN
ejpam-1676	518	27	wu	wu	PROPN
ejpam-1676	518	28	/	/	SYM
ejpam-1676	518	29	eur	eur	PROPN
ejpam-1676	518	30	.	.	PUNCT
ejpam-1676	519	1	j.	j.	PROPN
ejpam-1676	519	2	pure	pure	PROPN
ejpam-1676	519	3	appl	appl	PROPN
ejpam-1676	519	4	.	.	PROPN
ejpam-1676	519	5	math	math	PROPN
ejpam-1676	519	6	,	,	PUNCT
ejpam-1676	519	7	5	5	NUM
ejpam-1676	519	8	(	(	PUNCT
ejpam-1676	519	9	2012	2012	NUM
ejpam-1676	519	10	)	)	PUNCT
ejpam-1676	519	11	,	,	PUNCT
ejpam-1676	519	12	333	333	NUM
ejpam-1676	519	13	-	-	SYM
ejpam-1676	519	14	356	356	NUM
ejpam-1676	519	15	350	350	NUM
ejpam-1676	519	16	where	where	SCONJ
ejpam-1676	519	17	∇sw	∇sw	ADJ
ejpam-1676	519	18	n1	n1	PROPN
ejpam-1676	519	19	(	(	PUNCT
ejpam-1676	519	20	β	β	NOUN
ejpam-1676	519	21	)	)	PUNCT
ejpam-1676	519	22	and	and	CCONJ
ejpam-1676	519	23	∇sw	∇sw	PROPN
ejpam-1676	519	24	n2	n2	NOUN
ejpam-1676	519	25	(	(	PUNCT
ejpam-1676	519	26	β	β	NOUN
ejpam-1676	519	27	)	)	PUNCT
ejpam-1676	519	28	are	be	AUX
ejpam-1676	519	29	similarly	similarly	ADV
ejpam-1676	519	30	defined	define	VERB
ejpam-1676	519	31	on	on	ADP
ejpam-1676	519	32	main	main	ADJ
ejpam-1676	519	33	group	group	NOUN
ejpam-1676	519	34	and	and	CCONJ
ejpam-1676	519	35	contamination	contamination	NOUN
ejpam-1676	519	36	group	group	NOUN
ejpam-1676	519	37	,	,	PUNCT
ejpam-1676	519	38	∇sw	∇sw	PROPN
ejpam-1676	519	39	n1	n1	PROPN
ejpam-1676	519	40	(	(	PUNCT
ejpam-1676	519	41	β	β	NOUN
ejpam-1676	519	42	)	)	PUNCT
ejpam-1676	519	43	=	=	SYM
ejpam-1676	519	44	∂	∂	NUM
ejpam-1676	520	1	∂β	∂β	PROPN
ejpam-1676	520	2	n1∑	n1∑	PROPN
ejpam-1676	520	3	j=1	j=1	PROPN
ejpam-1676	520	4	w1	w1	PROPN
ejpam-1676	520	5	j	j	PROPN
ejpam-1676	520	6	l1	l1	PROPN
ejpam-1676	520	7	j(β	j(β	PROPN
ejpam-1676	520	8	)	)	PUNCT
ejpam-1676	520	9	,	,	PUNCT
ejpam-1676	520	10	∇sw	∇sw	PROPN
ejpam-1676	520	11	n2	n2	NOUN
ejpam-1676	520	12	(	(	PUNCT
ejpam-1676	520	13	β	β	NOUN
ejpam-1676	520	14	)	)	PUNCT
ejpam-1676	520	15	=	=	SYM
ejpam-1676	520	16	∂	∂	NUM
ejpam-1676	520	17	∂β	∂β	PROPN
ejpam-1676	520	18	n2∑	n2∑	PROPN
ejpam-1676	520	19	j=1	j=1	PROPN
ejpam-1676	520	20	w2	w2	NOUN
ejpam-1676	520	21	j	j	PROPN
ejpam-1676	520	22	l2	l2	PROPN
ejpam-1676	520	23	j(β	j(β	PROPN
ejpam-1676	520	24	)	)	PUNCT
ejpam-1676	520	25	.	.	PUNCT
ejpam-1676	521	1	since	since	SCONJ
ejpam-1676	521	2	c	c	PROPN
ejpam-1676	521	3	=	=	SYM
ejpam-1676	521	4	op(1	op(1	NUM
ejpam-1676	521	5	)	)	PUNCT
ejpam-1676	521	6	,	,	PUNCT
ejpam-1676	521	7	we	we	PRON
ejpam-1676	521	8	know	know	VERB
ejpam-1676	521	9	that	that	SCONJ
ejpam-1676	521	10	∇sw	∇sw	PROPN
ejpam-1676	521	11	n2	n2	NOUN
ejpam-1676	521	12	(	(	PUNCT
ejpam-1676	521	13	β)→	β)→	ADV
ejpam-1676	521	14	0	0	NUM
ejpam-1676	521	15	.	.	PUNCT
ejpam-1676	522	1	by	by	ADP
ejpam-1676	522	2	the	the	DET
ejpam-1676	522	3	proof	proof	NOUN
ejpam-1676	522	4	of	of	ADP
ejpam-1676	522	5	b	b	PROPN
ejpam-1676	522	6	=	=	PUNCT
ejpam-1676	522	7	op(1	op(1	PROPN
ejpam-1676	522	8	)	)	PUNCT
ejpam-1676	522	9	,	,	PUNCT
ejpam-1676	522	10	we	we	PRON
ejpam-1676	522	11	have	have	VERB
ejpam-1676	522	12	max	max	PROPN
ejpam-1676	522	13	γ∈nn1	γ∈nn1	PUNCT
ejpam-1676	522	14	(	(	PUNCT
ejpam-1676	522	15	δ	δ	PROPN
ejpam-1676	522	16	)	)	PUNCT
ejpam-1676	522	17	‖i−1/2	‖i−1/2	PROPN
ejpam-1676	522	18	n1	n1	PROPN
ejpam-1676	522	19	(	(	PUNCT
ejpam-1676	522	20	β0)∇s1−w	β0)∇s1−w	PROPN
ejpam-1676	522	21	n1	n1	ADJ
ejpam-1676	522	22	(	(	PUNCT
ejpam-1676	522	23	γ)i−1/2	γ)i−1/2	PROPN
ejpam-1676	522	24	n1	n1	NOUN
ejpam-1676	522	25	(	(	PUNCT
ejpam-1676	522	26	β0)‖	β0)‖	NUM
ejpam-1676	522	27	→	→	SYM
ejpam-1676	522	28	0	0	NUM
ejpam-1676	522	29	,	,	PUNCT
ejpam-1676	522	30	so	so	ADV
ejpam-1676	522	31	,	,	PUNCT
ejpam-1676	522	32	as	as	ADP
ejpam-1676	522	33	∇sn1	∇sn1	PRON
ejpam-1676	522	34	(	(	PUNCT
ejpam-1676	522	35	γ	γ	NOUN
ejpam-1676	522	36	)	)	PUNCT
ejpam-1676	522	37	=	=	NOUN
ejpam-1676	522	38	∇s1−w	∇s1−w	PROPN
ejpam-1676	522	39	n1	n1	NOUN
ejpam-1676	522	40	(	(	PUNCT
ejpam-1676	522	41	γ	γ	X
ejpam-1676	522	42	)	)	PUNCT
ejpam-1676	522	43	+	+	ADJ
ejpam-1676	522	44	∇sw	∇sw	PROPN
ejpam-1676	522	45	n1	n1	PROPN
ejpam-1676	522	46	(	(	PUNCT
ejpam-1676	522	47	γ	γ	NOUN
ejpam-1676	522	48	)	)	PUNCT
ejpam-1676	522	49	and	and	CCONJ
ejpam-1676	522	50	nn1	nn1	PROPN
ejpam-1676	522	51	(	(	PUNCT
ejpam-1676	522	52	δ	δ	PROPN
ejpam-1676	522	53	)	)	PUNCT
ejpam-1676	522	54	converges	converge	VERB
ejpam-1676	522	55	to	to	ADP
ejpam-1676	522	56	β0	β0	PROPN
ejpam-1676	522	57	,	,	PUNCT
ejpam-1676	522	58	i−1/2	i−1/2	PROPN
ejpam-1676	522	59	n1	n1	PROPN
ejpam-1676	522	60	(	(	PUNCT
ejpam-1676	522	61	β0)∇sw	β0)∇sw	PROPN
ejpam-1676	522	62	n1	n1	PROPN
ejpam-1676	522	63	(	(	PUNCT
ejpam-1676	522	64	β0)i	β0)i	NOUN
ejpam-1676	522	65	−1/2	−1/2	VERB
ejpam-1676	522	66	n1	n1	PROPN
ejpam-1676	522	67	(	(	PUNCT
ejpam-1676	522	68	β0	β0	NOUN
ejpam-1676	522	69	)	)	PUNCT
ejpam-1676	523	1	=	=	SYM
ejpam-1676	523	2	i−1/2	i−1/2	PROPN
ejpam-1676	523	3	n1	n1	PROPN
ejpam-1676	523	4	(	(	PUNCT
ejpam-1676	523	5	β0)∇sn1	β0)∇sn1	X
ejpam-1676	523	6	(	(	PUNCT
ejpam-1676	523	7	β)i−1/2	β)i−1/2	PROPN
ejpam-1676	523	8	n1	n1	PROPN
ejpam-1676	523	9	(	(	PUNCT
ejpam-1676	523	10	β0	β0	NOUN
ejpam-1676	523	11	)	)	PUNCT
ejpam-1676	523	12	+	+	NUM
ejpam-1676	523	13	o(1	o(1	NOUN
ejpam-1676	523	14	)	)	PUNCT
ejpam-1676	523	15	→−ip	→−ip	NOUN
ejpam-1676	523	16	,	,	PUNCT
ejpam-1676	523	17	and	and	CCONJ
ejpam-1676	523	18	∇sw	∇sw	PROPN
ejpam-1676	523	19	n1	n1	NOUN
ejpam-1676	523	20	(	(	PUNCT
ejpam-1676	523	21	β0)→−in1	β0)→−in1	PUNCT
ejpam-1676	523	22	(	(	PUNCT
ejpam-1676	523	23	β0	β0	NOUN
ejpam-1676	523	24	)	)	PUNCT
ejpam-1676	523	25	.	.	PUNCT
ejpam-1676	524	1	then	then	ADV
ejpam-1676	524	2	∇sw	∇sw	PROPN
ejpam-1676	524	3	n	n	PROPN
ejpam-1676	524	4	(	(	PUNCT
ejpam-1676	524	5	β0	β0	NOUN
ejpam-1676	524	6	)	)	PUNCT
ejpam-1676	524	7	=	=	PROPN
ejpam-1676	524	8	∇sw	∇sw	PROPN
ejpam-1676	524	9	n1	n1	PROPN
ejpam-1676	524	10	(	(	PUNCT
ejpam-1676	524	11	β0	β0	PROPN
ejpam-1676	524	12	)	)	PUNCT
ejpam-1676	524	13	+	+	ADJ
ejpam-1676	524	14	∇sw	∇sw	PROPN
ejpam-1676	524	15	n2	n2	NOUN
ejpam-1676	524	16	(	(	PUNCT
ejpam-1676	524	17	β0)→−in1	β0)→−in1	NOUN
ejpam-1676	524	18	(	(	PUNCT
ejpam-1676	524	19	β0	β0	NOUN
ejpam-1676	524	20	)	)	PUNCT
ejpam-1676	524	21	.	.	PUNCT
ejpam-1676	525	1	taking	take	VERB
ejpam-1676	525	2	the	the	DET
ejpam-1676	525	3	taylor	taylor	PROPN
ejpam-1676	525	4	expansion	expansion	NOUN
ejpam-1676	525	5	of	of	ADP
ejpam-1676	525	6	sw	sw	PROPN
ejpam-1676	525	7	n	n	PROPN
ejpam-1676	525	8	(	(	PUNCT
ejpam-1676	525	9	β̂	β̂	ADP
ejpam-1676	525	10	)	)	PUNCT
ejpam-1676	525	11	at	at	ADP
ejpam-1676	525	12	β0	β0	PROPN
ejpam-1676	525	13	,	,	PUNCT
ejpam-1676	525	14	where	where	SCONJ
ejpam-1676	525	15	β̂	β̂	ADP
ejpam-1676	525	16	is	be	AUX
ejpam-1676	525	17	the	the	DET
ejpam-1676	525	18	wl	wl	NOUN
ejpam-1676	525	19	estimate	estimate	NOUN
ejpam-1676	525	20	,	,	PUNCT
ejpam-1676	525	21	it	it	PRON
ejpam-1676	525	22	follows	follow	VERB
ejpam-1676	525	23	that	that	SCONJ
ejpam-1676	525	24	sw	sw	PROPN
ejpam-1676	525	25	n	n	PROPN
ejpam-1676	525	26	(	(	PUNCT
ejpam-1676	525	27	β̂	β̂	NUM
ejpam-1676	525	28	)	)	PUNCT
ejpam-1676	525	29	=	=	SYM
ejpam-1676	525	30	sw	sw	PROPN
ejpam-1676	525	31	n	n	PROPN
ejpam-1676	525	32	(	(	PUNCT
ejpam-1676	525	33	β0	β0	PROPN
ejpam-1676	525	34	)	)	PUNCT
ejpam-1676	525	35	+	+	NOUN
ejpam-1676	525	36	∇sw	∇sw	ADJ
ejpam-1676	525	37	n	n	X
ejpam-1676	525	38	(	(	PUNCT
ejpam-1676	525	39	β0)(β̂	β0)(β̂	PROPN
ejpam-1676	525	40	−β0	−β0	PROPN
ejpam-1676	525	41	)	)	PUNCT
ejpam-1676	526	1	+	+	CCONJ
ejpam-1676	526	2	op(∇sw	op(∇sw	ADJ
ejpam-1676	526	3	n	n	CCONJ
ejpam-1676	526	4	(	(	PUNCT
ejpam-1676	526	5	β0)(β̂	β0)(β̂	PROPN
ejpam-1676	526	6	−β0	−β0	PROPN
ejpam-1676	526	7	)	)	PUNCT
ejpam-1676	526	8	)	)	PUNCT
ejpam-1676	527	1	and	and	CCONJ
ejpam-1676	527	2	so	so	ADV
ejpam-1676	527	3	sw	sw	PROPN
ejpam-1676	527	4	n	n	PROPN
ejpam-1676	527	5	(	(	PUNCT
ejpam-1676	527	6	β̂	β̂	NUM
ejpam-1676	527	7	)	)	PUNCT
ejpam-1676	527	8	=	=	SYM
ejpam-1676	527	9	sw	sw	PROPN
ejpam-1676	527	10	n	n	CCONJ
ejpam-1676	527	11	(	(	PUNCT
ejpam-1676	527	12	β0)−	β0)−	NOUN
ejpam-1676	527	13	in1	in1	NOUN
ejpam-1676	527	14	(	(	PUNCT
ejpam-1676	527	15	β0)(β̂	β0)(β̂	NOUN
ejpam-1676	527	16	−β0	−β0	PROPN
ejpam-1676	527	17	)	)	PUNCT
ejpam-1676	528	1	+	+	CCONJ
ejpam-1676	528	2	op(in1	op(in1	NUM
ejpam-1676	528	3	(	(	PUNCT
ejpam-1676	528	4	β0)(β̂	β0)(β̂	PROPN
ejpam-1676	528	5	−β0	−β0	PROPN
ejpam-1676	528	6	)	)	PUNCT
ejpam-1676	528	7	)	)	PUNCT
ejpam-1676	529	1	(	(	PUNCT
ejpam-1676	529	2	11	11	X
ejpam-1676	529	3	)	)	PUNCT
ejpam-1676	529	4	setting	set	VERB
ejpam-1676	529	5	sw	sw	PROPN
ejpam-1676	529	6	n	n	PROPN
ejpam-1676	529	7	(	(	PUNCT
ejpam-1676	529	8	β̂	β̂	NUM
ejpam-1676	529	9	)	)	PUNCT
ejpam-1676	529	10	=	=	SYM
ejpam-1676	529	11	0	0	NUM
ejpam-1676	529	12	,	,	PUNCT
ejpam-1676	529	13	and	and	CCONJ
ejpam-1676	529	14	multiplying	multiply	VERB
ejpam-1676	529	15	the	the	DET
ejpam-1676	529	16	both	both	DET
ejpam-1676	529	17	sides	side	NOUN
ejpam-1676	529	18	of	of	ADP
ejpam-1676	529	19	(	(	PUNCT
ejpam-1676	529	20	11	11	NUM
ejpam-1676	529	21	)	)	PUNCT
ejpam-1676	529	22	by	by	ADP
ejpam-1676	529	23	i−1/2	i−1/2	PROPN
ejpam-1676	529	24	n1	n1	PROPN
ejpam-1676	529	25	(	(	PUNCT
ejpam-1676	529	26	β0	β0	NOUN
ejpam-1676	529	27	)	)	PUNCT
ejpam-1676	529	28	,	,	PUNCT
ejpam-1676	529	29	we	we	PRON
ejpam-1676	529	30	obtain	obtain	VERB
ejpam-1676	529	31	i1/2	i1/2	ADJ
ejpam-1676	529	32	n1	n1	NOUN
ejpam-1676	529	33	(	(	PUNCT
ejpam-1676	529	34	β0)(β̂	β0)(β̂	NOUN
ejpam-1676	529	35	−β0	−β0	PROPN
ejpam-1676	529	36	)	)	PUNCT
ejpam-1676	529	37	=	=	PUNCT
ejpam-1676	530	1	i−1/2	i−1/2	PROPN
ejpam-1676	530	2	n1	n1	PROPN
ejpam-1676	530	3	(	(	PUNCT
ejpam-1676	530	4	β0)s	β0)s	ADJ
ejpam-1676	530	5	w	w	PROPN
ejpam-1676	530	6	n	n	PROPN
ejpam-1676	530	7	(	(	PUNCT
ejpam-1676	530	8	β0	β0	PROPN
ejpam-1676	530	9	)	)	PUNCT
ejpam-1676	530	10	+	+	CCONJ
ejpam-1676	530	11	op(1	op(1	NOUN
ejpam-1676	530	12	)	)	PUNCT
ejpam-1676	530	13	.	.	PUNCT
ejpam-1676	531	1	because	because	SCONJ
ejpam-1676	531	2	c	c	PROPN
ejpam-1676	531	3	=	=	SYM
ejpam-1676	531	4	op(1	op(1	NUM
ejpam-1676	531	5	)	)	PUNCT
ejpam-1676	531	6	,	,	PUNCT
ejpam-1676	531	7	we	we	PRON
ejpam-1676	531	8	have	have	VERB
ejpam-1676	531	9	var(sw	var(sw	NOUN
ejpam-1676	531	10	n2	n2	NOUN
ejpam-1676	531	11	(	(	PUNCT
ejpam-1676	531	12	β))→	β))→	PROPN
ejpam-1676	531	13	0	0	NUM
ejpam-1676	531	14	,	,	PUNCT
ejpam-1676	531	15	then	then	ADV
ejpam-1676	531	16	var(sw	var(sw	PUNCT
ejpam-1676	531	17	n	n	PROPN
ejpam-1676	531	18	(	(	PUNCT
ejpam-1676	531	19	β0	β0	NOUN
ejpam-1676	531	20	)	)	PUNCT
ejpam-1676	531	21	)	)	PUNCT
ejpam-1676	532	1	=	=	SYM
ejpam-1676	532	2	var(sw	var(sw	ADP
ejpam-1676	532	3	n1	n1	PROPN
ejpam-1676	532	4	(	(	PUNCT
ejpam-1676	532	5	β0	β0	NOUN
ejpam-1676	532	6	)	)	PUNCT
ejpam-1676	532	7	)	)	PUNCT
ejpam-1676	533	1	+	+	CCONJ
ejpam-1676	533	2	var(sw	var(sw	X
ejpam-1676	533	3	n2	n2	PROPN
ejpam-1676	533	4	(	(	PUNCT
ejpam-1676	533	5	β0	β0	NOUN
ejpam-1676	533	6	)	)	PUNCT
ejpam-1676	533	7	)	)	PUNCT
ejpam-1676	534	1	=	=	SYM
ejpam-1676	534	2	1	1	NUM
ejpam-1676	534	3	φ	φ	PROPN
ejpam-1676	534	4			PROPN
ejpam-1676	534	5			VERB
ejpam-1676	534	6	n1∑	n1∑	PROPN
ejpam-1676	534	7	j=1	j=1	PROPN
ejpam-1676	534	8	w2	w2	NOUN
ejpam-1676	534	9	1	1	NUM
ejpam-1676	534	10	jϕ((β	jϕ((β	PROPN
ejpam-1676	534	11	t	t	PROPN
ejpam-1676	534	12	0	0	NUM
ejpam-1676	534	13	x	x	SYM
ejpam-1676	534	14	j))t1	j))t1	PROPN
ejpam-1676	534	15	jx	jx	PROPN
ejpam-1676	534	16	jx	jx	PROPN
ejpam-1676	534	17	t	t	PROPN
ejpam-1676	534	18	j	j	PROPN
ejpam-1676	534	19			PROPN
ejpam-1676	534	20	+	+	VERB
ejpam-1676	534	21	op(1	op(1	NOUN
ejpam-1676	534	22	)	)	PUNCT
ejpam-1676	534	23	p.	p.	NOUN
ejpam-1676	534	24	guo	guo	PROPN
ejpam-1676	534	25	,	,	PUNCT
ejpam-1676	534	26	x.	x.	PROPN
ejpam-1676	534	27	wang	wang	PROPN
ejpam-1676	534	28	,	,	PUNCT
ejpam-1676	534	29	y.	y.	PROPN
ejpam-1676	534	30	wu	wu	PROPN
ejpam-1676	534	31	/	/	SYM
ejpam-1676	534	32	eur	eur	PROPN
ejpam-1676	534	33	.	.	PUNCT
ejpam-1676	535	1	j.	j.	PROPN
ejpam-1676	535	2	pure	pure	PROPN
ejpam-1676	535	3	appl	appl	PROPN
ejpam-1676	535	4	.	.	PROPN
ejpam-1676	535	5	math	math	PROPN
ejpam-1676	535	6	,	,	PUNCT
ejpam-1676	535	7	5	5	NUM
ejpam-1676	535	8	(	(	PUNCT
ejpam-1676	535	9	2012	2012	NUM
ejpam-1676	535	10	)	)	PUNCT
ejpam-1676	535	11	,	,	PUNCT
ejpam-1676	535	12	333	333	NUM
ejpam-1676	535	13	-	-	SYM
ejpam-1676	535	14	356	356	NUM
ejpam-1676	535	15	351	351	NUM
ejpam-1676	535	16	=	=	NOUN
ejpam-1676	535	17	in1	in1	NOUN
ejpam-1676	535	18	(	(	PUNCT
ejpam-1676	535	19	β0)−	β0)−	NOUN
ejpam-1676	535	20	1	1	NUM
ejpam-1676	535	21	φ	φ	PROPN
ejpam-1676	535	22			PROPN
ejpam-1676	535	23			VERB
ejpam-1676	535	24	n1∑	n1∑	PROPN
ejpam-1676	535	25	j=1	j=1	PROPN
ejpam-1676	535	26	(	(	PUNCT
ejpam-1676	535	27	1−w2	1−w2	NUM
ejpam-1676	535	28	1	1	NUM
ejpam-1676	535	29	j)ϕ((β	j)ϕ((β	X
ejpam-1676	535	30	t	t	PROPN
ejpam-1676	535	31	0	0	NUM
ejpam-1676	535	32	x	x	SYM
ejpam-1676	535	33	j))t1	j))t1	PROPN
ejpam-1676	536	1	jx	jx	PROPN
ejpam-1676	536	2	jx	jx	PROPN
ejpam-1676	536	3	t	t	PROPN
ejpam-1676	536	4	j	j	PROPN
ejpam-1676	536	5			PROPN
ejpam-1676	536	6	+	+	VERB
ejpam-1676	536	7	op(1	op(1	PRON
ejpam-1676	536	8	)	)	PUNCT
ejpam-1676	536	9	=	=	SYM
ejpam-1676	537	1	in1	in1	NOUN
ejpam-1676	537	2	(	(	PUNCT
ejpam-1676	537	3	β0)−	β0)−	NOUN
ejpam-1676	537	4	op(var(sw	op(var(sw	PROPN
ejpam-1676	537	5	n	n	PROPN
ejpam-1676	537	6	(	(	PUNCT
ejpam-1676	537	7	β0	β0	NOUN
ejpam-1676	537	8	)	)	PUNCT
ejpam-1676	537	9	)	)	PUNCT
ejpam-1676	537	10	)	)	PUNCT
ejpam-1676	537	11	,	,	PUNCT
ejpam-1676	537	12	and	and	CCONJ
ejpam-1676	537	13	hence	hence	ADV
ejpam-1676	537	14	in1	in1	ADJ
ejpam-1676	537	15	(	(	PUNCT
ejpam-1676	537	16	β0	β0	NOUN
ejpam-1676	537	17	)	)	PUNCT
ejpam-1676	537	18	=	=	PUNCT
ejpam-1676	537	19	var(sw	var(sw	PROPN
ejpam-1676	537	20	n	n	PROPN
ejpam-1676	537	21	(	(	PUNCT
ejpam-1676	537	22	β0	β0	NOUN
ejpam-1676	537	23	)	)	PUNCT
ejpam-1676	537	24	)	)	PUNCT
ejpam-1676	538	1	+	+	CCONJ
ejpam-1676	538	2	op(var(sw	op(var(sw	PROPN
ejpam-1676	538	3	n	n	PROPN
ejpam-1676	538	4	(	(	PUNCT
ejpam-1676	538	5	β0	β0	NOUN
ejpam-1676	538	6	)	)	PUNCT
ejpam-1676	538	7	)	)	PUNCT
ejpam-1676	538	8	)	)	PUNCT
ejpam-1676	538	9	.	.	PUNCT
ejpam-1676	539	1	by	by	ADP
ejpam-1676	539	2	the	the	DET
ejpam-1676	539	3	clt	clt	PROPN
ejpam-1676	539	4	[	[	X
ejpam-1676	539	5	corrollary	corrollary	ADJ
ejpam-1676	539	6	1.3	1.3	NUM
ejpam-1676	539	7	,	,	PUNCT
ejpam-1676	539	8	1	1	NUM
ejpam-1676	539	9	]	]	PUNCT
ejpam-1676	539	10	and	and	CCONJ
ejpam-1676	539	11	slutsky	slutsky	PROPN
ejpam-1676	539	12	’s	’s	PART
ejpam-1676	539	13	theorem	theorem	PROPN
ejpam-1676	539	14	,	,	PUNCT
ejpam-1676	539	15	we	we	PRON
ejpam-1676	539	16	have	have	VERB
ejpam-1676	539	17	i−1/2	i−1/2	PROPN
ejpam-1676	539	18	n1	n1	PROPN
ejpam-1676	539	19	(	(	PUNCT
ejpam-1676	539	20	β0)s	β0)s	ADJ
ejpam-1676	539	21	w	w	PROPN
ejpam-1676	539	22	n	n	PROPN
ejpam-1676	539	23	(	(	PUNCT
ejpam-1676	539	24	β0	β0	NOUN
ejpam-1676	539	25	)	)	PUNCT
ejpam-1676	539	26	d→	d→	VERB
ejpam-1676	539	27	np(0	np(0	NOUN
ejpam-1676	539	28	,	,	PUNCT
ejpam-1676	539	29	ip	ip	NOUN
ejpam-1676	539	30	)	)	PUNCT
ejpam-1676	539	31	,	,	PUNCT
ejpam-1676	539	32	and	and	CCONJ
ejpam-1676	539	33	hence	hence	ADV
ejpam-1676	539	34	i1/2	i1/2	PROPN
ejpam-1676	539	35	n1	n1	PROPN
ejpam-1676	539	36	(	(	PUNCT
ejpam-1676	539	37	β0)(β̂	β0)(β̂	PROPN
ejpam-1676	539	38	−β0	−β0	PROPN
ejpam-1676	539	39	)	)	PUNCT
ejpam-1676	539	40	d→	d→	VERB
ejpam-1676	539	41	np(0	np(0	NOUN
ejpam-1676	539	42	,	,	PUNCT
ejpam-1676	539	43	ip	ip	NOUN
ejpam-1676	539	44	)	)	PUNCT
ejpam-1676	539	45	.	.	PUNCT
ejpam-1676	540	1	4	4	X
ejpam-1676	540	2	.	.	X
ejpam-1676	540	3	empirical	empirical	ADJ
ejpam-1676	540	4	distribution	distribution	NOUN
ejpam-1676	540	5	by	by	ADP
ejpam-1676	540	6	bootstrap	bootstrap	NOUN
ejpam-1676	540	7	we	we	PRON
ejpam-1676	540	8	have	have	AUX
ejpam-1676	540	9	given	give	VERB
ejpam-1676	540	10	the	the	DET
ejpam-1676	540	11	asymptotic	asymptotic	ADJ
ejpam-1676	540	12	distribution	distribution	NOUN
ejpam-1676	540	13	of	of	ADP
ejpam-1676	540	14	wle	wle	PROPN
ejpam-1676	540	15	,	,	PUNCT
ejpam-1676	540	16	however	however	ADV
ejpam-1676	540	17	,	,	PUNCT
ejpam-1676	540	18	the	the	DET
ejpam-1676	540	19	asymptotic	asymptotic	ADJ
ejpam-1676	540	20	distribution	distribution	NOUN
ejpam-1676	540	21	is	be	AUX
ejpam-1676	540	22	hard	hard	ADJ
ejpam-1676	540	23	to	to	PART
ejpam-1676	540	24	use	use	VERB
ejpam-1676	540	25	in	in	ADP
ejpam-1676	540	26	real	real	ADJ
ejpam-1676	540	27	data	datum	NOUN
ejpam-1676	540	28	analysis	analysis	NOUN
ejpam-1676	540	29	.	.	PUNCT
ejpam-1676	541	1	at	at	ADP
ejpam-1676	541	2	the	the	DET
ejpam-1676	541	3	same	same	ADJ
ejpam-1676	541	4	time	time	NOUN
ejpam-1676	541	5	,	,	PUNCT
ejpam-1676	541	6	the	the	DET
ejpam-1676	541	7	sample	sample	NOUN
ejpam-1676	541	8	size	size	NOUN
ejpam-1676	541	9	of	of	ADP
ejpam-1676	541	10	data	datum	NOUN
ejpam-1676	541	11	sometime	sometime	ADV
ejpam-1676	541	12	is	be	AUX
ejpam-1676	541	13	not	not	PART
ejpam-1676	541	14	big	big	ADJ
ejpam-1676	541	15	enough	enough	ADV
ejpam-1676	541	16	,	,	PUNCT
ejpam-1676	541	17	and	and	CCONJ
ejpam-1676	541	18	therefore	therefore	ADV
ejpam-1676	541	19	it	it	PRON
ejpam-1676	541	20	may	may	AUX
ejpam-1676	541	21	cause	cause	VERB
ejpam-1676	541	22	bias	bias	NOUN
ejpam-1676	541	23	to	to	PART
ejpam-1676	541	24	use	use	VERB
ejpam-1676	541	25	the	the	DET
ejpam-1676	541	26	asymptotic	asymptotic	ADJ
ejpam-1676	541	27	distribution	distribution	NOUN
ejpam-1676	541	28	.	.	PUNCT
ejpam-1676	542	1	in	in	ADP
ejpam-1676	542	2	order	order	NOUN
ejpam-1676	542	3	to	to	PART
ejpam-1676	542	4	assess	assess	VERB
ejpam-1676	542	5	the	the	DET
ejpam-1676	542	6	significance	significance	NOUN
ejpam-1676	542	7	of	of	ADP
ejpam-1676	542	8	a	a	DET
ejpam-1676	542	9	parameter	parameter	NOUN
ejpam-1676	542	10	based	base	VERB
ejpam-1676	542	11	on	on	ADP
ejpam-1676	542	12	wle	wle	PROPN
ejpam-1676	542	13	,	,	PUNCT
ejpam-1676	542	14	we	we	PRON
ejpam-1676	542	15	proposed	propose	VERB
ejpam-1676	542	16	the	the	DET
ejpam-1676	542	17	bootstrap	bootstrap	NOUN
ejpam-1676	542	18	method	method	NOUN
ejpam-1676	542	19	to	to	PART
ejpam-1676	542	20	generate	generate	VERB
ejpam-1676	542	21	the	the	DET
ejpam-1676	542	22	empirical	empirical	ADJ
ejpam-1676	542	23	distribution	distribution	NOUN
ejpam-1676	542	24	of	of	ADP
ejpam-1676	542	25	wle	wle	PROPN
ejpam-1676	542	26	.	.	PUNCT
ejpam-1676	543	1	without	without	ADP
ejpam-1676	543	2	loss	loss	NOUN
ejpam-1676	543	3	of	of	ADP
ejpam-1676	543	4	generality	generality	NOUN
ejpam-1676	543	5	,	,	PUNCT
ejpam-1676	543	6	we	we	PRON
ejpam-1676	543	7	assume	assume	VERB
ejpam-1676	543	8	that	that	SCONJ
ejpam-1676	543	9	m	m	PROPN
ejpam-1676	543	10	=	=	SYM
ejpam-1676	543	11	2	2	NUM
ejpam-1676	543	12	and	and	CCONJ
ejpam-1676	543	13	number	number	NOUN
ejpam-1676	543	14	of	of	ADP
ejpam-1676	543	15	independent	independent	ADJ
ejpam-1676	543	16	variables	variable	NOUN
ejpam-1676	543	17	to	to	PART
ejpam-1676	543	18	be	be	AUX
ejpam-1676	543	19	1	1	NUM
ejpam-1676	543	20	.	.	PUNCT
ejpam-1676	543	21	suppose	suppose	VERB
ejpam-1676	543	22	the	the	DET
ejpam-1676	543	23	data	datum	NOUN
ejpam-1676	543	24	we	we	PRON
ejpam-1676	543	25	have	have	VERB
ejpam-1676	543	26	is	be	AUX
ejpam-1676	543	27	(	(	PUNCT
ejpam-1676	543	28	x	x	PROPN
ejpam-1676	543	29	i	i	PROPN
ejpam-1676	543	30	,	,	PUNCT
ejpam-1676	543	31	yi	yi	PROPN
ejpam-1676	543	32	)	)	PUNCT
ejpam-1676	543	33	for	for	ADP
ejpam-1676	543	34	i	i	PROPN
ejpam-1676	543	35	=	=	NOUN
ejpam-1676	543	36	1	1	NUM
ejpam-1676	543	37	,	,	PUNCT
ejpam-1676	543	38	.	.	PUNCT
ejpam-1676	543	39	.	.	PUNCT
ejpam-1676	544	1	.	.	PUNCT
ejpam-1676	545	1	,	,	PUNCT
ejpam-1676	545	2	n.	n.	PROPN
ejpam-1676	545	3	below	below	ADV
ejpam-1676	545	4	are	be	AUX
ejpam-1676	545	5	the	the	DET
ejpam-1676	545	6	main	main	ADJ
ejpam-1676	545	7	steps	step	NOUN
ejpam-1676	545	8	to	to	PART
ejpam-1676	545	9	generate	generate	VERB
ejpam-1676	545	10	the	the	DET
ejpam-1676	545	11	empirical	empirical	ADJ
ejpam-1676	545	12	distribution	distribution	NOUN
ejpam-1676	545	13	.	.	PUNCT
ejpam-1676	546	1	step	step	NOUN
ejpam-1676	546	2	1	1	NUM
ejpam-1676	546	3	:	:	PUNCT
ejpam-1676	546	4	in	in	ADP
ejpam-1676	546	5	the	the	DET
ejpam-1676	546	6	main	main	ADJ
ejpam-1676	546	7	group	group	NOUN
ejpam-1676	546	8	,	,	PUNCT
ejpam-1676	546	9	let	let	VERB
ejpam-1676	546	10	z	z	PROPN
ejpam-1676	546	11	j	j	PROPN
ejpam-1676	546	12	(	(	PUNCT
ejpam-1676	546	13	j	j	PROPN
ejpam-1676	546	14	=	=	SYM
ejpam-1676	546	15	1	1	NUM
ejpam-1676	546	16	,	,	PUNCT
ejpam-1676	546	17	.	.	PUNCT
ejpam-1676	546	18	.	.	PUNCT
ejpam-1676	547	1	.	.	PUNCT
ejpam-1676	548	1	,	,	PUNCT
ejpam-1676	548	2	k	k	X
ejpam-1676	548	3	)	)	PUNCT
ejpam-1676	548	4	to	to	PART
ejpam-1676	548	5	be	be	AUX
ejpam-1676	548	6	the	the	DET
ejpam-1676	548	7	unique	unique	ADJ
ejpam-1676	548	8	values	value	NOUN
ejpam-1676	548	9	of	of	ADP
ejpam-1676	548	10	x	x	PUNCT
ejpam-1676	548	11	i	i	NOUN
ejpam-1676	548	12	(	(	PUNCT
ejpam-1676	548	13	i	i	NOUN
ejpam-1676	548	14	=	=	NOUN
ejpam-1676	548	15	1	1	NUM
ejpam-1676	548	16	,	,	PUNCT
ejpam-1676	548	17	.	.	PUNCT
ejpam-1676	548	18	.	.	PUNCT
ejpam-1676	549	1	.	.	PUNCT
ejpam-1676	550	1	,	,	PUNCT
ejpam-1676	550	2	n1	n1	NOUN
ejpam-1676	550	3	)	)	PUNCT
ejpam-1676	550	4	,	,	PUNCT
ejpam-1676	550	5	where	where	SCONJ
ejpam-1676	550	6	k	k	PROPN
ejpam-1676	550	7	≤	≤	PROPN
ejpam-1676	550	8	n1	n1	NOUN
ejpam-1676	550	9	.	.	PUNCT
ejpam-1676	551	1	calculate	calculate	VERB
ejpam-1676	551	2	the	the	DET
ejpam-1676	551	3	frequency	frequency	NOUN
ejpam-1676	551	4	of	of	ADP
ejpam-1676	551	5	z	z	PROPN
ejpam-1676	551	6	j	j	PROPN
ejpam-1676	551	7	and	and	CCONJ
ejpam-1676	551	8	let	let	VERB
ejpam-1676	551	9	p0	p0	NOUN
ejpam-1676	551	10	=	=	SYM
ejpam-1676	551	11	∑	∑	PUNCT
ejpam-1676	551	12	yi	yi	PROPN
ejpam-1676	551	13	/	/	SYM
ejpam-1676	551	14	n1	n1	NOUN
ejpam-1676	551	15	.	.	PUNCT
ejpam-1676	552	1	step	step	NOUN
ejpam-1676	552	2	2	2	NUM
ejpam-1676	552	3	:	:	PUNCT
ejpam-1676	552	4	draw	draw	VERB
ejpam-1676	552	5	n1	n1	ADJ
ejpam-1676	552	6	sample	sample	NOUN
ejpam-1676	552	7	from	from	ADP
ejpam-1676	552	8	z	z	PROPN
ejpam-1676	552	9	j	j	PROPN
ejpam-1676	552	10	,	,	PUNCT
ejpam-1676	552	11	with	with	ADP
ejpam-1676	552	12	probabilities	probability	NOUN
ejpam-1676	552	13	corresponds	correspond	VERB
ejpam-1676	552	14	to	to	ADP
ejpam-1676	552	15	frequencies	frequency	NOUN
ejpam-1676	552	16	of	of	ADP
ejpam-1676	552	17	zi	zi	NOUN
ejpam-1676	552	18	.	.	PUNCT
ejpam-1676	553	1	denotes	denote	VERB
ejpam-1676	553	2	the	the	DET
ejpam-1676	553	3	new	new	ADJ
ejpam-1676	553	4	sample	sample	NOUN
ejpam-1676	553	5	of	of	ADP
ejpam-1676	553	6	the	the	DET
ejpam-1676	553	7	main	main	ADJ
ejpam-1676	553	8	group	group	NOUN
ejpam-1676	553	9	as	as	ADP
ejpam-1676	553	10	x	x	PROPN
ejpam-1676	553	11	′i	′i	NOUN
ejpam-1676	553	12	(	(	PUNCT
ejpam-1676	553	13	i	i	NOUN
ejpam-1676	553	14	=	=	NOUN
ejpam-1676	553	15	1	1	NUM
ejpam-1676	553	16	,	,	PUNCT
ejpam-1676	553	17	.	.	PUNCT
ejpam-1676	553	18	.	.	PUNCT
ejpam-1676	554	1	.	.	PUNCT
ejpam-1676	555	1	,	,	PUNCT
ejpam-1676	555	2	n1	n1	NOUN
ejpam-1676	555	3	)	)	PUNCT
ejpam-1676	555	4	.	.	PUNCT
ejpam-1676	556	1	step	step	NOUN
ejpam-1676	556	2	3	3	NUM
ejpam-1676	556	3	:	:	PUNCT
ejpam-1676	556	4	regenerate	regenerate	VERB
ejpam-1676	556	5	y	y	PROPN
ejpam-1676	556	6	′i	′i	NOUN
ejpam-1676	556	7	(	(	PUNCT
ejpam-1676	556	8	i	i	NOUN
ejpam-1676	556	9	=	=	NOUN
ejpam-1676	556	10	1	1	NUM
ejpam-1676	556	11	,	,	PUNCT
ejpam-1676	556	12	.	.	PUNCT
ejpam-1676	556	13	.	.	PUNCT
ejpam-1676	557	1	.	.	PUNCT
ejpam-1676	558	1	,	,	PUNCT
ejpam-1676	558	2	n1	n1	NOUN
ejpam-1676	558	3	)	)	PUNCT
ejpam-1676	558	4	from	from	ADP
ejpam-1676	558	5	binomial	binomial	ADJ
ejpam-1676	558	6	distribution	distribution	NOUN
ejpam-1676	558	7	bin(n1	bin(n1	NOUN
ejpam-1676	558	8	,	,	PUNCT
ejpam-1676	558	9	p0	p0	NOUN
ejpam-1676	558	10	)	)	PUNCT
ejpam-1676	558	11	.	.	PUNCT
ejpam-1676	559	1	step	step	NOUN
ejpam-1676	559	2	4	4	NUM
ejpam-1676	559	3	:	:	PUNCT
ejpam-1676	559	4	combine	combine	VERB
ejpam-1676	559	5	(	(	PUNCT
ejpam-1676	559	6	x	x	X
ejpam-1676	559	7	′	′	NUM
ejpam-1676	560	1	i	i	PRON
ejpam-1676	560	2	,	,	PUNCT
ejpam-1676	560	3	y	y	PROPN
ejpam-1676	560	4	′	′	NUM
ejpam-1676	560	5	i	i	INTJ
ejpam-1676	560	6	)	)	PUNCT
ejpam-1676	561	1	(	(	PUNCT
ejpam-1676	561	2	i	i	NOUN
ejpam-1676	561	3	=	=	NOUN
ejpam-1676	561	4	1	1	NUM
ejpam-1676	561	5	,	,	PUNCT
ejpam-1676	561	6	.	.	PUNCT
ejpam-1676	561	7	.	.	PUNCT
ejpam-1676	561	8	.	.	PUNCT
ejpam-1676	562	1	,	,	PUNCT
ejpam-1676	562	2	n1	n1	NOUN
ejpam-1676	562	3	)	)	PUNCT
ejpam-1676	562	4	and	and	CCONJ
ejpam-1676	562	5	(	(	PUNCT
ejpam-1676	562	6	x	x	X
ejpam-1676	562	7	i	i	PROPN
ejpam-1676	562	8	,	,	PUNCT
ejpam-1676	562	9	yi	yi	PROPN
ejpam-1676	562	10	)	)	PUNCT
ejpam-1676	563	1	i	i	NOUN
ejpam-1676	563	2	=	=	SYM
ejpam-1676	563	3	n1	n1	PROPN
ejpam-1676	563	4	+	+	CCONJ
ejpam-1676	563	5	1	1	NUM
ejpam-1676	563	6	,	,	PUNCT
ejpam-1676	563	7	.	.	PUNCT
ejpam-1676	563	8	.	.	PUNCT
ejpam-1676	564	1	.	.	PUNCT
ejpam-1676	565	1	,	,	PUNCT
ejpam-1676	565	2	n	n	CCONJ
ejpam-1676	565	3	,	,	PUNCT
ejpam-1676	565	4	calculate	calculate	VERB
ejpam-1676	565	5	the	the	DET
ejpam-1676	565	6	wle	wle	NOUN
ejpam-1676	565	7	of	of	ADP
ejpam-1676	565	8	this	this	DET
ejpam-1676	565	9	new	new	ADJ
ejpam-1676	565	10	sample	sample	NOUN
ejpam-1676	565	11	.	.	PUNCT
ejpam-1676	566	1	repeat	repeat	NOUN
ejpam-1676	566	2	step	step	NOUN
ejpam-1676	566	3	2	2	NUM
ejpam-1676	566	4	to	to	ADP
ejpam-1676	566	5	4	4	NUM
ejpam-1676	566	6	for	for	ADP
ejpam-1676	566	7	n	n	PRON
ejpam-1676	566	8	times	time	NOUN
ejpam-1676	566	9	,	,	PUNCT
ejpam-1676	566	10	to	to	PART
ejpam-1676	566	11	get	get	VERB
ejpam-1676	566	12	the	the	DET
ejpam-1676	566	13	empirical	empirical	ADJ
ejpam-1676	566	14	distribution	distribution	NOUN
ejpam-1676	566	15	of	of	ADP
ejpam-1676	566	16	wle	wle	NOUN
ejpam-1676	566	17	with	with	ADP
ejpam-1676	566	18	n	n	DET
ejpam-1676	566	19	values	value	NOUN
ejpam-1676	566	20	.	.	PUNCT
ejpam-1676	567	1	5	5	X
ejpam-1676	567	2	.	.	X
ejpam-1676	567	3	simulation	simulation	NOUN
ejpam-1676	567	4	study	study	NOUN
ejpam-1676	567	5	in	in	ADP
ejpam-1676	567	6	this	this	DET
ejpam-1676	567	7	section	section	NOUN
ejpam-1676	567	8	we	we	PRON
ejpam-1676	567	9	present	present	VERB
ejpam-1676	567	10	the	the	DET
ejpam-1676	567	11	simulation	simulation	NOUN
ejpam-1676	567	12	results	result	VERB
ejpam-1676	567	13	to	to	PART
ejpam-1676	567	14	illustrate	illustrate	VERB
ejpam-1676	567	15	the	the	DET
ejpam-1676	567	16	numerical	numerical	ADJ
ejpam-1676	567	17	performance	performance	NOUN
ejpam-1676	567	18	of	of	ADP
ejpam-1676	567	19	the	the	DET
ejpam-1676	567	20	proposed	propose	VERB
ejpam-1676	567	21	methods	method	NOUN
ejpam-1676	567	22	.	.	PUNCT
ejpam-1676	568	1	we	we	PRON
ejpam-1676	568	2	mainly	mainly	ADV
ejpam-1676	568	3	compare	compare	VERB
ejpam-1676	568	4	the	the	DET
ejpam-1676	568	5	estimated	estimate	VERB
ejpam-1676	568	6	values	value	NOUN
ejpam-1676	568	7	and	and	CCONJ
ejpam-1676	568	8	the	the	DET
ejpam-1676	568	9	standard	standard	ADJ
ejpam-1676	568	10	deviations	deviation	NOUN
ejpam-1676	568	11	based	base	VERB
ejpam-1676	568	12	on	on	ADP
ejpam-1676	568	13	the	the	DET
ejpam-1676	568	14	mle	mle	PROPN
ejpam-1676	568	15	and	and	CCONJ
ejpam-1676	568	16	wle	wle	NOUN
ejpam-1676	568	17	estimators	estimator	NOUN
ejpam-1676	568	18	and	and	CCONJ
ejpam-1676	568	19	present	present	VERB
ejpam-1676	568	20	the	the	DET
ejpam-1676	568	21	ratios	ratio	NOUN
ejpam-1676	568	22	of	of	ADP
ejpam-1676	568	23	the	the	DET
ejpam-1676	568	24	mse	mse	NOUN
ejpam-1676	568	25	of	of	ADP
ejpam-1676	568	26	the	the	DET
ejpam-1676	568	27	wle	wle	NOUN
ejpam-1676	568	28	to	to	ADP
ejpam-1676	568	29	the	the	DET
ejpam-1676	568	30	mse	mse	NOUN
ejpam-1676	568	31	of	of	ADP
ejpam-1676	568	32	the	the	DET
ejpam-1676	568	33	mle	mle	NOUN
ejpam-1676	568	34	under	under	ADP
ejpam-1676	568	35	different	different	ADJ
ejpam-1676	568	36	simulation	simulation	NOUN
ejpam-1676	568	37	scenarios	scenario	NOUN
ejpam-1676	568	38	.	.	PUNCT
ejpam-1676	569	1	p.	p.	NOUN
ejpam-1676	569	2	guo	guo	PROPN
ejpam-1676	569	3	,	,	PUNCT
ejpam-1676	569	4	x.	x.	PROPN
ejpam-1676	569	5	wang	wang	PROPN
ejpam-1676	569	6	,	,	PUNCT
ejpam-1676	569	7	y.	y.	PROPN
ejpam-1676	569	8	wu	wu	PROPN
ejpam-1676	569	9	/	/	SYM
ejpam-1676	569	10	eur	eur	PROPN
ejpam-1676	569	11	.	.	PUNCT
ejpam-1676	570	1	j.	j.	PROPN
ejpam-1676	570	2	pure	pure	PROPN
ejpam-1676	570	3	appl	appl	PROPN
ejpam-1676	570	4	.	.	PROPN
ejpam-1676	570	5	math	math	PROPN
ejpam-1676	570	6	,	,	PUNCT
ejpam-1676	570	7	5	5	NUM
ejpam-1676	570	8	(	(	PUNCT
ejpam-1676	570	9	2012	2012	NUM
ejpam-1676	570	10	)	)	PUNCT
ejpam-1676	570	11	,	,	PUNCT
ejpam-1676	570	12	333	333	NUM
ejpam-1676	570	13	-	-	SYM
ejpam-1676	570	14	356	356	NUM
ejpam-1676	570	15	352	352	NUM
ejpam-1676	570	16	example	example	NOUN
ejpam-1676	570	17	1	1	NUM
ejpam-1676	570	18	.	.	PUNCT
ejpam-1676	571	1	the	the	DET
ejpam-1676	571	2	data	datum	NOUN
ejpam-1676	571	3	are	be	AUX
ejpam-1676	571	4	generated	generate	VERB
ejpam-1676	571	5	from	from	ADP
ejpam-1676	571	6	models	model	NOUN
ejpam-1676	571	7	(	(	PUNCT
ejpam-1676	571	8	8)	8)	NUM
ejpam-1676	571	9	with	with	ADP
ejpam-1676	571	10	β1	β1	PROPN
ejpam-1676	571	11	=	=	PUNCT
ejpam-1676	571	12	1	1	NUM
ejpam-1676	571	13	and	and	CCONJ
ejpam-1676	571	14	β2	β2	NOUN
ejpam-1676	571	15	=	=	PROPN
ejpam-1676	571	16	0.92	0.92	NUM
ejpam-1676	571	17	.	.	PUNCT
ejpam-1676	572	1	let	let	VERB
ejpam-1676	572	2	x1	x1	PROPN
ejpam-1676	572	3	and	and	CCONJ
ejpam-1676	572	4	x2	x2	PROPN
ejpam-1676	572	5	are	be	AUX
ejpam-1676	572	6	independent	independent	ADJ
ejpam-1676	572	7	and	and	CCONJ
ejpam-1676	572	8	identically	identically	ADV
ejpam-1676	572	9	distributed	distribute	VERB
ejpam-1676	572	10	with	with	ADP
ejpam-1676	572	11	n(0,1	n(0,1	NOUN
ejpam-1676	572	12	)	)	PUNCT
ejpam-1676	572	13	.	.	PUNCT
ejpam-1676	573	1	also	also	ADV
ejpam-1676	573	2	,	,	PUNCT
ejpam-1676	573	3	we	we	PRON
ejpam-1676	573	4	set	set	VERB
ejpam-1676	573	5	that	that	DET
ejpam-1676	573	6	σ1	σ1	PROPN
ejpam-1676	573	7	=	=	PROPN
ejpam-1676	573	8	σ2	σ2	PROPN
ejpam-1676	573	9	=	=	NOUN
ejpam-1676	573	10	0.3	0.3	NUM
ejpam-1676	573	11	.	.	PUNCT
ejpam-1676	574	1	the	the	DET
ejpam-1676	574	2	simulation	simulation	NOUN
ejpam-1676	574	3	study	study	NOUN
ejpam-1676	574	4	is	be	AUX
ejpam-1676	574	5	conducted	conduct	VERB
ejpam-1676	574	6	as	as	SCONJ
ejpam-1676	574	7	follows	follow	VERB
ejpam-1676	574	8	.	.	PUNCT
ejpam-1676	575	1	we	we	PRON
ejpam-1676	575	2	consider	consider	VERB
ejpam-1676	575	3	three	three	NUM
ejpam-1676	575	4	scenarios	scenario	NOUN
ejpam-1676	575	5	:	:	PUNCT
ejpam-1676	575	6	(	(	PUNCT
ejpam-1676	575	7	i	i	NOUN
ejpam-1676	575	8	)	)	PUNCT
ejpam-1676	575	9	n1	n1	PROPN
ejpam-1676	575	10	<	<	X
ejpam-1676	575	11	n2	n2	PROPN
ejpam-1676	575	12	;	;	PUNCT
ejpam-1676	575	13	(	(	PUNCT
ejpam-1676	575	14	ii	ii	NOUN
ejpam-1676	575	15	)	)	PUNCT
ejpam-1676	575	16	n1	n1	PROPN
ejpam-1676	575	17	=	=	SYM
ejpam-1676	575	18	n2	n2	NOUN
ejpam-1676	575	19	;	;	PUNCT
ejpam-1676	575	20	and	and	CCONJ
ejpam-1676	575	21	(	(	PUNCT
ejpam-1676	575	22	iii	iii	X
ejpam-1676	575	23	)	)	PUNCT
ejpam-1676	575	24	n1	n1	PROPN
ejpam-1676	575	25	>	>	X
ejpam-1676	575	26	n2	n2	PROPN
ejpam-1676	575	27	.	.	PUNCT
ejpam-1676	576	1	the	the	DET
ejpam-1676	576	2	range	range	NOUN
ejpam-1676	576	3	of	of	ADP
ejpam-1676	576	4	n1	n1	NOUN
ejpam-1676	576	5	and	and	CCONJ
ejpam-1676	576	6	n2	n2	NOUN
ejpam-1676	576	7	is	be	AUX
ejpam-1676	576	8	between	between	ADP
ejpam-1676	576	9	8	8	NUM
ejpam-1676	576	10	and	and	CCONJ
ejpam-1676	576	11	90	90	NUM
ejpam-1676	576	12	.	.	PUNCT
ejpam-1676	577	1	for	for	ADP
ejpam-1676	577	2	each	each	DET
ejpam-1676	577	3	fixed	fix	VERB
ejpam-1676	577	4	sample	sample	NOUN
ejpam-1676	577	5	size	size	NOUN
ejpam-1676	577	6	,	,	PUNCT
ejpam-1676	577	7	1000	1000	NUM
ejpam-1676	577	8	independent	independent	ADJ
ejpam-1676	577	9	data	datum	NOUN
ejpam-1676	577	10	sets	set	NOUN
ejpam-1676	577	11	are	be	AUX
ejpam-1676	577	12	generated	generate	VERB
ejpam-1676	577	13	.	.	PUNCT
ejpam-1676	578	1	table	table	NOUN
ejpam-1676	578	2	1	1	NUM
ejpam-1676	578	3	summarizes	summarize	NOUN
ejpam-1676	578	4	the	the	DET
ejpam-1676	578	5	results	result	NOUN
ejpam-1676	578	6	.	.	PUNCT
ejpam-1676	579	1	it	it	PRON
ejpam-1676	579	2	can	can	AUX
ejpam-1676	579	3	be	be	AUX
ejpam-1676	579	4	seen	see	VERB
ejpam-1676	579	5	that	that	SCONJ
ejpam-1676	579	6	both	both	CCONJ
ejpam-1676	579	7	the	the	DET
ejpam-1676	579	8	mle	mle	PROPN
ejpam-1676	579	9	and	and	CCONJ
ejpam-1676	579	10	wle	wle	PROPN
ejpam-1676	579	11	are	be	AUX
ejpam-1676	579	12	close	close	ADJ
ejpam-1676	579	13	to	to	ADP
ejpam-1676	579	14	the	the	DET
ejpam-1676	579	15	true	true	ADJ
ejpam-1676	579	16	values	value	NOUN
ejpam-1676	579	17	.	.	PUNCT
ejpam-1676	580	1	the	the	DET
ejpam-1676	580	2	standard	standard	ADJ
ejpam-1676	580	3	deviations	deviation	NOUN
ejpam-1676	580	4	of	of	ADP
ejpam-1676	580	5	the	the	DET
ejpam-1676	580	6	wle	wle	PROPN
ejpam-1676	580	7	’s	’s	PART
ejpam-1676	580	8	are	be	AUX
ejpam-1676	580	9	consistently	consistently	ADV
ejpam-1676	580	10	smaller	small	ADJ
ejpam-1676	580	11	than	than	ADP
ejpam-1676	580	12	those	those	PRON
ejpam-1676	580	13	of	of	ADP
ejpam-1676	580	14	the	the	DET
ejpam-1676	580	15	mle	mle	NOUN
ejpam-1676	580	16	.	.	PUNCT
ejpam-1676	581	1	furthermore	furthermore	ADV
ejpam-1676	581	2	,	,	PUNCT
ejpam-1676	581	3	the	the	DET
ejpam-1676	581	4	ratios	ratio	NOUN
ejpam-1676	581	5	of	of	ADP
ejpam-1676	581	6	the	the	DET
ejpam-1676	581	7	mse	mse	NOUN
ejpam-1676	581	8	of	of	ADP
ejpam-1676	581	9	the	the	DET
ejpam-1676	581	10	wle	wle	NOUN
ejpam-1676	581	11	to	to	ADP
ejpam-1676	581	12	the	the	DET
ejpam-1676	581	13	mse	mse	NOUN
ejpam-1676	581	14	of	of	ADP
ejpam-1676	581	15	the	the	DET
ejpam-1676	581	16	mle	mle	NOUN
ejpam-1676	581	17	are	be	AUX
ejpam-1676	581	18	always	always	ADV
ejpam-1676	581	19	smaller	small	ADJ
ejpam-1676	581	20	than	than	ADP
ejpam-1676	581	21	1	1	NUM
ejpam-1676	581	22	.	.	PUNCT
ejpam-1676	582	1	this	this	PRON
ejpam-1676	582	2	implies	imply	VERB
ejpam-1676	582	3	that	that	SCONJ
ejpam-1676	582	4	the	the	DET
ejpam-1676	582	5	wle	wle	NOUN
ejpam-1676	582	6	could	could	AUX
ejpam-1676	582	7	reduce	reduce	VERB
ejpam-1676	582	8	the	the	DET
ejpam-1676	582	9	mse	mse	NOUN
ejpam-1676	582	10	in	in	ADP
ejpam-1676	582	11	contrast	contrast	NOUN
ejpam-1676	582	12	with	with	ADP
ejpam-1676	582	13	the	the	DET
ejpam-1676	582	14	mle	mle	NOUN
ejpam-1676	582	15	.	.	PROPN
ejpam-1676	582	16	to	to	PART
ejpam-1676	582	17	assess	assess	VERB
ejpam-1676	582	18	the	the	DET
ejpam-1676	582	19	sensitivity	sensitivity	NOUN
ejpam-1676	582	20	of	of	ADP
ejpam-1676	582	21	the	the	DET
ejpam-1676	582	22	ratio	ratio	NOUN
ejpam-1676	582	23	to	to	ADP
ejpam-1676	582	24	actual	actual	ADJ
ejpam-1676	582	25	value	value	NOUN
ejpam-1676	582	26	of	of	ADP
ejpam-1676	582	27	the	the	DET
ejpam-1676	582	28	related	relate	VERB
ejpam-1676	582	29	parameter	parameter	NOUN
ejpam-1676	582	30	β2	β2	NOUN
ejpam-1676	582	31	,	,	PUNCT
ejpam-1676	582	32	we	we	PRON
ejpam-1676	582	33	run	run	VERB
ejpam-1676	582	34	simulations	simulation	NOUN
ejpam-1676	582	35	with	with	ADP
ejpam-1676	582	36	different	different	ADJ
ejpam-1676	582	37	β2	β2	NOUN
ejpam-1676	582	38	values	value	NOUN
ejpam-1676	582	39	ranging	range	VERB
ejpam-1676	582	40	from	from	ADP
ejpam-1676	582	41	0.7	0.7	NUM
ejpam-1676	582	42	to	to	PART
ejpam-1676	582	43	0.98	0.98	NUM
ejpam-1676	582	44	.	.	PUNCT
ejpam-1676	583	1	the	the	DET
ejpam-1676	583	2	trend	trend	NOUN
ejpam-1676	583	3	of	of	ADP
ejpam-1676	583	4	the	the	DET
ejpam-1676	583	5	ratios	ratio	NOUN
ejpam-1676	583	6	against	against	ADP
ejpam-1676	583	7	the	the	DET
ejpam-1676	583	8	values	value	NOUN
ejpam-1676	583	9	of	of	ADP
ejpam-1676	583	10	β2	β2	PROPN
ejpam-1676	583	11	is	be	AUX
ejpam-1676	583	12	provided	provide	VERB
ejpam-1676	583	13	in	in	ADP
ejpam-1676	583	14	figure	figure	NOUN
ejpam-1676	583	15	1	1	NUM
ejpam-1676	583	16	.	.	PUNCT
ejpam-1676	584	1	in	in	ADP
ejpam-1676	584	2	general	general	ADJ
ejpam-1676	584	3	,	,	PUNCT
ejpam-1676	584	4	we	we	PRON
ejpam-1676	584	5	find	find	VERB
ejpam-1676	584	6	that	that	SCONJ
ejpam-1676	584	7	the	the	DET
ejpam-1676	584	8	smaller	small	ADJ
ejpam-1676	584	9	n1	n1	NOUN
ejpam-1676	584	10	,	,	PUNCT
ejpam-1676	584	11	the	the	PRON
ejpam-1676	584	12	smaller	small	ADJ
ejpam-1676	584	13	the	the	DET
ejpam-1676	584	14	ratio	ratio	NOUN
ejpam-1676	584	15	under	under	ADP
ejpam-1676	584	16	the	the	DET
ejpam-1676	584	17	current	current	ADJ
ejpam-1676	584	18	setup	setup	NOUN
ejpam-1676	584	19	.	.	PUNCT
ejpam-1676	585	1	for	for	ADP
ejpam-1676	585	2	each	each	DET
ejpam-1676	585	3	fixed	fix	VERB
ejpam-1676	585	4	n1	n1	NOUN
ejpam-1676	585	5	,	,	PUNCT
ejpam-1676	585	6	the	the	DET
ejpam-1676	585	7	ratios	ratio	NOUN
ejpam-1676	585	8	are	be	AUX
ejpam-1676	585	9	inversely	inversely	ADV
ejpam-1676	585	10	proportional	proportional	ADJ
ejpam-1676	585	11	to	to	ADP
ejpam-1676	585	12	the	the	DET
ejpam-1676	585	13	value	value	NOUN
ejpam-1676	585	14	of	of	ADP
ejpam-1676	585	15	β2	β2	NOUN
ejpam-1676	585	16	.	.	PUNCT
ejpam-1676	586	1	they	they	PRON
ejpam-1676	586	2	would	would	AUX
ejpam-1676	586	3	also	also	ADV
ejpam-1676	586	4	increase	increase	VERB
ejpam-1676	586	5	with	with	ADP
ejpam-1676	586	6	larger	large	ADJ
ejpam-1676	586	7	value	value	NOUN
ejpam-1676	586	8	for	for	ADP
ejpam-1676	586	9	n2	n2	NOUN
ejpam-1676	586	10	.	.	PUNCT
ejpam-1676	587	1	we	we	PRON
ejpam-1676	587	2	see	see	VERB
ejpam-1676	587	3	that	that	SCONJ
ejpam-1676	587	4	the	the	DET
ejpam-1676	587	5	wle	wle	NOUN
ejpam-1676	587	6	outperforms	outperform	VERB
ejpam-1676	587	7	the	the	DET
ejpam-1676	587	8	mle	mle	NOUN
ejpam-1676	587	9	remarkably	remarkably	ADV
ejpam-1676	587	10	in	in	ADP
ejpam-1676	587	11	small	small	ADJ
ejpam-1676	587	12	sample	sample	NOUN
ejpam-1676	587	13	size	size	NOUN
ejpam-1676	587	14	n1	n1	PROPN
ejpam-1676	587	15	.	.	PUNCT
ejpam-1676	588	1	this	this	DET
ejpam-1676	588	2	feature	feature	NOUN
ejpam-1676	588	3	could	could	AUX
ejpam-1676	588	4	be	be	AUX
ejpam-1676	588	5	useful	useful	ADJ
ejpam-1676	588	6	in	in	ADP
ejpam-1676	588	7	practice	practice	NOUN
ejpam-1676	588	8	if	if	SCONJ
ejpam-1676	588	9	the	the	DET
ejpam-1676	588	10	first	first	ADJ
ejpam-1676	588	11	sample	sample	NOUN
ejpam-1676	588	12	size	size	NOUN
ejpam-1676	588	13	is	be	AUX
ejpam-1676	588	14	small	small	ADJ
ejpam-1676	588	15	and	and	CCONJ
ejpam-1676	588	16	we	we	PRON
ejpam-1676	588	17	have	have	VERB
ejpam-1676	588	18	abundant	abundant	ADJ
ejpam-1676	588	19	related	related	ADJ
ejpam-1676	588	20	information	information	NOUN
ejpam-1676	588	21	.	.	PUNCT
ejpam-1676	589	1	0.70	0.70	NUM
ejpam-1676	589	2	0.80	0.80	NUM
ejpam-1676	589	3	0.90	0.90	NUM
ejpam-1676	589	4	0	0	NUM
ejpam-1676	589	5	.	.	PROPN
ejpam-1676	590	1	6	6	NUM
ejpam-1676	590	2	0	0	NUM
ejpam-1676	590	3	.	.	NOUN
ejpam-1676	590	4	8	8	NUM
ejpam-1676	590	5	1	1	NUM
ejpam-1676	590	6	.	.	NOUN
ejpam-1676	590	7	0	0	NUM
ejpam-1676	591	1	1	1	NUM
ejpam-1676	591	2	.	.	SYM
ejpam-1676	591	3	2	2	NUM
ejpam-1676	591	4	beta2	beta2	NOUN
ejpam-1676	591	5	ra	ra	PROPN
ejpam-1676	591	6	tio	tio	PROPN
ejpam-1676	591	7	n1=10;n2=8	n1=10;n2=8	PROPN
ejpam-1676	591	8	0.70	0.70	NUM
ejpam-1676	591	9	0.80	0.80	NUM
ejpam-1676	591	10	0.90	0.90	NUM
ejpam-1676	591	11	0	0	NUM
ejpam-1676	591	12	.	.	PROPN
ejpam-1676	592	1	6	6	NUM
ejpam-1676	592	2	0	0	NUM
ejpam-1676	592	3	.	.	NOUN
ejpam-1676	592	4	8	8	NUM
ejpam-1676	592	5	1	1	NUM
ejpam-1676	592	6	.	.	NOUN
ejpam-1676	592	7	0	0	NUM
ejpam-1676	593	1	1	1	NUM
ejpam-1676	593	2	.	.	SYM
ejpam-1676	593	3	2	2	NUM
ejpam-1676	593	4	beta2	beta2	NOUN
ejpam-1676	593	5	ra	ra	PROPN
ejpam-1676	593	6	tio	tio	PROPN
ejpam-1676	593	7	n1=10;n2=10	n1=10;n2=10	PROPN
ejpam-1676	593	8	0.70	0.70	NUM
ejpam-1676	593	9	0.80	0.80	NUM
ejpam-1676	593	10	0.90	0.90	NUM
ejpam-1676	593	11	0	0	NUM
ejpam-1676	593	12	.	.	PROPN
ejpam-1676	594	1	6	6	NUM
ejpam-1676	594	2	0	0	NUM
ejpam-1676	594	3	.	.	NOUN
ejpam-1676	594	4	8	8	NUM
ejpam-1676	594	5	1	1	NUM
ejpam-1676	594	6	.	.	NOUN
ejpam-1676	594	7	0	0	NUM
ejpam-1676	595	1	1	1	NUM
ejpam-1676	595	2	.	.	SYM
ejpam-1676	595	3	2	2	NUM
ejpam-1676	595	4	beta2	beta2	NOUN
ejpam-1676	595	5	ra	ra	NOUN
ejpam-1676	595	6	tio	tio	NOUN
ejpam-1676	595	7	n1=10;n2=15	n1=10;n2=15	PROPN
ejpam-1676	595	8	0.70	0.70	NUM
ejpam-1676	595	9	0.80	0.80	NUM
ejpam-1676	595	10	0.90	0.90	NUM
ejpam-1676	595	11	0	0	NUM
ejpam-1676	595	12	.	.	PROPN
ejpam-1676	596	1	6	6	NUM
ejpam-1676	596	2	0	0	NUM
ejpam-1676	596	3	.	.	NOUN
ejpam-1676	596	4	8	8	NUM
ejpam-1676	596	5	1	1	NUM
ejpam-1676	596	6	.	.	NOUN
ejpam-1676	596	7	0	0	NUM
ejpam-1676	597	1	1	1	NUM
ejpam-1676	597	2	.	.	SYM
ejpam-1676	597	3	2	2	NUM
ejpam-1676	597	4	beta2	beta2	NOUN
ejpam-1676	597	5	ra	ra	PROPN
ejpam-1676	597	6	tio	tio	PROPN
ejpam-1676	597	7	n1=20;n2=16	n1=20;n2=16	PROPN
ejpam-1676	597	8	0.70	0.70	NUM
ejpam-1676	597	9	0.80	0.80	NUM
ejpam-1676	597	10	0.90	0.90	NUM
ejpam-1676	597	11	0	0	NUM
ejpam-1676	597	12	.	.	PROPN
ejpam-1676	598	1	6	6	NUM
ejpam-1676	598	2	0	0	NUM
ejpam-1676	598	3	.	.	NOUN
ejpam-1676	598	4	8	8	NUM
ejpam-1676	598	5	1	1	NUM
ejpam-1676	598	6	.	.	NOUN
ejpam-1676	598	7	0	0	NUM
ejpam-1676	599	1	1	1	NUM
ejpam-1676	599	2	.	.	SYM
ejpam-1676	599	3	2	2	NUM
ejpam-1676	599	4	beta2	beta2	NOUN
ejpam-1676	599	5	ra	ra	PROPN
ejpam-1676	599	6	tio	tio	PROPN
ejpam-1676	599	7	n1=20;n2=20	n1=20;n2=20	PROPN
ejpam-1676	599	8	0.70	0.70	NUM
ejpam-1676	599	9	0.80	0.80	NUM
ejpam-1676	599	10	0.90	0.90	NUM
ejpam-1676	599	11	0	0	NUM
ejpam-1676	599	12	.	.	PROPN
ejpam-1676	600	1	6	6	NUM
ejpam-1676	600	2	0	0	NUM
ejpam-1676	600	3	.	.	NOUN
ejpam-1676	600	4	8	8	NUM
ejpam-1676	600	5	1	1	NUM
ejpam-1676	600	6	.	.	NOUN
ejpam-1676	600	7	0	0	NUM
ejpam-1676	601	1	1	1	NUM
ejpam-1676	601	2	.	.	SYM
ejpam-1676	601	3	2	2	NUM
ejpam-1676	601	4	beta2	beta2	NOUN
ejpam-1676	601	5	ra	ra	NOUN
ejpam-1676	601	6	tio	tio	NOUN
ejpam-1676	601	7	n1=20;n2=30	n1=20;n2=30	ADP
ejpam-1676	601	8	0.70	0.70	NUM
ejpam-1676	601	9	0.80	0.80	NUM
ejpam-1676	601	10	0.90	0.90	NUM
ejpam-1676	601	11	0	0	NUM
ejpam-1676	601	12	.	.	PROPN
ejpam-1676	602	1	6	6	NUM
ejpam-1676	602	2	0	0	NUM
ejpam-1676	602	3	.	.	NOUN
ejpam-1676	602	4	8	8	NUM
ejpam-1676	602	5	1	1	NUM
ejpam-1676	602	6	.	.	NOUN
ejpam-1676	602	7	0	0	NUM
ejpam-1676	603	1	1	1	NUM
ejpam-1676	603	2	.	.	SYM
ejpam-1676	603	3	2	2	NUM
ejpam-1676	603	4	beta2	beta2	NOUN
ejpam-1676	603	5	ra	ra	PROPN
ejpam-1676	603	6	tio	tio	NOUN
ejpam-1676	603	7	n1=30;n2=24	n1=30;n2=24	NOUN
ejpam-1676	603	8	0.70	0.70	NUM
ejpam-1676	603	9	0.80	0.80	NUM
ejpam-1676	603	10	0.90	0.90	NUM
ejpam-1676	603	11	0	0	NUM
ejpam-1676	603	12	.	.	PROPN
ejpam-1676	604	1	6	6	NUM
ejpam-1676	604	2	0	0	NUM
ejpam-1676	604	3	.	.	NOUN
ejpam-1676	604	4	8	8	NUM
ejpam-1676	604	5	1	1	NUM
ejpam-1676	604	6	.	.	NOUN
ejpam-1676	604	7	0	0	NUM
ejpam-1676	605	1	1	1	NUM
ejpam-1676	605	2	.	.	SYM
ejpam-1676	605	3	2	2	NUM
ejpam-1676	605	4	beta2	beta2	NOUN
ejpam-1676	605	5	ra	ra	NOUN
ejpam-1676	605	6	tio	tio	NOUN
ejpam-1676	605	7	n1=30;n2=30	n1=30;n2=30	PROPN
ejpam-1676	606	1	0.70	0.70	NUM
ejpam-1676	606	2	0.80	0.80	NUM
ejpam-1676	606	3	0.90	0.90	NUM
ejpam-1676	606	4	0	0	NUM
ejpam-1676	606	5	.	.	PROPN
ejpam-1676	607	1	6	6	NUM
ejpam-1676	607	2	0	0	NUM
ejpam-1676	607	3	.	.	NOUN
ejpam-1676	607	4	8	8	NUM
ejpam-1676	607	5	1	1	NUM
ejpam-1676	607	6	.	.	NOUN
ejpam-1676	607	7	0	0	NUM
ejpam-1676	608	1	1	1	NUM
ejpam-1676	608	2	.	.	SYM
ejpam-1676	608	3	2	2	NUM
ejpam-1676	608	4	beta2	beta2	NOUN
ejpam-1676	608	5	ra	ra	PROPN
ejpam-1676	608	6	tio	tio	PROPN
ejpam-1676	608	7	n1=30;n2=45	n1=30;n2=45	PROPN
ejpam-1676	608	8	figure	figure	NOUN
ejpam-1676	608	9	1	1	NUM
ejpam-1676	608	10	:	:	PUNCT
ejpam-1676	608	11	the	the	DET
ejpam-1676	608	12	trend	trend	NOUN
ejpam-1676	608	13	of	of	ADP
ejpam-1676	608	14	the	the	DET
ejpam-1676	608	15	ratio	ratio	NOUN
ejpam-1676	608	16	values	value	NOUN
ejpam-1676	608	17	against	against	ADP
ejpam-1676	608	18	β2	β2	NOUN
ejpam-1676	608	19	for	for	ADP
ejpam-1676	608	20	9	9	NUM
ejpam-1676	608	21	ombinations	ombination	NOUN
ejpam-1676	608	22	of	of	ADP
ejpam-1676	608	23	(	(	PUNCT
ejpam-1676	608	24	n1	n1	NOUN
ejpam-1676	608	25	,	,	PUNCT
ejpam-1676	608	26	n2	n2	ADJ
ejpam-1676	608	27	)	)	PUNCT
ejpam-1676	608	28	.	.	PUNCT
ejpam-1676	609	1	the	the	DET
ejpam-1676	609	2	horizontal	horizontal	ADJ
ejpam-1676	609	3	line	line	NOUN
ejpam-1676	609	4	indi	indi	PROPN
ejpam-1676	609	5	ates	ates	PROPN
ejpam-1676	609	6	ratio=	ratio=	PROPN
ejpam-1676	609	7	1	1	NUM
ejpam-1676	609	8	.	.	PUNCT
ejpam-1676	610	1	p.	p.	NOUN
ejpam-1676	610	2	g	g	PROPN
ejpam-1676	611	1	u	u	NOUN
ejpam-1676	611	2	o	o	PROPN
ejpam-1676	611	3	,	,	PUNCT
ejpam-1676	611	4	x	x	PROPN
ejpam-1676	611	5	.	.	PUNCT
ejpam-1676	612	1	w	w	PROPN
ejpam-1676	612	2	a	a	DET
ejpam-1676	612	3	n	n	ADV
ejpam-1676	612	4	g	g	NOUN
ejpam-1676	612	5	,	,	PUNCT
ejpam-1676	612	6	y.	y.	PROPN
ejpam-1676	612	7	w	w	PROPN
ejpam-1676	612	8	u	u	PROPN
ejpam-1676	612	9	/	/	SYM
ejpam-1676	612	10	e	e	X
ejpam-1676	612	11	u	u	PROPN
ejpam-1676	612	12	r.	r.	PROPN
ejpam-1676	612	13	j.	j.	PROPN
ejpam-1676	613	1	p	p	PROPN
ejpam-1676	613	2	u	u	PROPN
ejpam-1676	613	3	re	re	ADP
ejpam-1676	613	4	a	a	DET
ejpam-1676	613	5	p	p	X
ejpam-1676	613	6	p	p	X
ejpam-1676	613	7	l.	l.	PROPN
ejpam-1676	613	8	m	m	PROPN
ejpam-1676	613	9	a	a	DET
ejpam-1676	613	10	th	th	X
ejpam-1676	613	11	,	,	PUNCT
ejpam-1676	613	12	5	5	NUM
ejpam-1676	613	13	(	(	PUNCT
ejpam-1676	613	14	2	2	NUM
ejpam-1676	613	15	0	0	NUM
ejpam-1676	613	16	1	1	NUM
ejpam-1676	613	17	2	2	NUM
ejpam-1676	613	18	)	)	PUNCT
ejpam-1676	613	19	,	,	PUNCT
ejpam-1676	613	20	3	3	NUM
ejpam-1676	613	21	3	3	NUM
ejpam-1676	613	22	3	3	NUM
ejpam-1676	613	23	-3	-3	SYM
ejpam-1676	613	24	5	5	NUM
ejpam-1676	613	25	6	6	NUM
ejpam-1676	613	26	3	3	NUM
ejpam-1676	613	27	5	5	NUM
ejpam-1676	613	28	3	3	NUM
ejpam-1676	613	29	table	table	NOUN
ejpam-1676	613	30	1	1	NUM
ejpam-1676	613	31	:	:	PUNCT
ejpam-1676	613	32	estimated	estimate	VERB
ejpam-1676	613	33	values	value	NOUN
ejpam-1676	613	34	based	base	VERB
ejpam-1676	613	35	on	on	ADP
ejpam-1676	613	36	the	the	DET
ejpam-1676	613	37	mle	mle	PROPN
ejpam-1676	613	38	and	and	CCONJ
ejpam-1676	613	39	wle	wle	PROPN
ejpam-1676	613	40	,	,	PUNCT
ejpam-1676	613	41	the	the	DET
ejpam-1676	613	42	orresponding	orresponde	VERB
ejpam-1676	613	43	mse	mse	NOUN
ejpam-1676	613	44	and	and	CCONJ
ejpam-1676	613	45	the	the	DET
ejpam-1676	613	46	ratio	ratio	NOUN
ejpam-1676	613	47	of	of	ADP
ejpam-1676	613	48	the	the	DET
ejpam-1676	613	49	mse(wle	mse(wle	NOUN
ejpam-1676	613	50	)	)	PUNCT
ejpam-1676	613	51	to	to	ADP
ejpam-1676	613	52	the	the	DET
ejpam-1676	613	53	mse(mle	mse(mle	NOUN
ejpam-1676	613	54	)	)	PUNCT
ejpam-1676	613	55	for	for	ADP
ejpam-1676	613	56	example	example	NOUN
ejpam-1676	613	57	1	1	X
ejpam-1676	613	58	.	.	X
ejpam-1676	613	59	mle	mle	PROPN
ejpam-1676	613	60	wle	wle	PROPN
ejpam-1676	613	61	n1	n1	PROPN
ejpam-1676	613	62	n2	n2	PROPN
ejpam-1676	613	63	β̂(std	β̂(std	PROPN
ejpam-1676	613	64	)	)	PUNCT
ejpam-1676	613	65	mse(std	mse(std	NOUN
ejpam-1676	613	66	)	)	PUNCT
ejpam-1676	613	67	β̂(std	β̂(std	PROPN
ejpam-1676	613	68	)	)	PUNCT
ejpam-1676	613	69	mse(std	mse(std	NOUN
ejpam-1676	613	70	)	)	PUNCT
ejpam-1676	613	71	λ̂(std	λ̂(std	PROPN
ejpam-1676	613	72	)	)	PUNCT
ejpam-1676	613	73	ratio	ratio	NOUN
ejpam-1676	613	74	10	10	NUM
ejpam-1676	613	75	8	8	NUM
ejpam-1676	613	76	1.009(0.239	1.009(0.239	NUM
ejpam-1676	613	77	)	)	PUNCT
ejpam-1676	613	78	0.057(0.093	0.057(0.093	ADJ
ejpam-1676	613	79	)	)	PUNCT
ejpam-1676	613	80	0.997(0.21	0.997(0.21	PROPN
ejpam-1676	613	81	)	)	PUNCT
ejpam-1676	613	82	0.044(0.069	0.044(0.069	NOUN
ejpam-1676	613	83	)	)	PUNCT
ejpam-1676	613	84	0.744(0.187	0.744(0.187	NOUN
ejpam-1676	613	85	)	)	PUNCT
ejpam-1676	613	86	0.768	0.768	NUM
ejpam-1676	613	87	20	20	NUM
ejpam-1676	613	88	16	16	NUM
ejpam-1676	613	89	1.008(0.166	1.008(0.166	NUM
ejpam-1676	613	90	)	)	PUNCT
ejpam-1676	613	91	0.028(0.041	0.028(0.041	NOUN
ejpam-1676	613	92	)	)	PUNCT
ejpam-1676	613	93	0.992(0.149	0.992(0.149	NOUN
ejpam-1676	613	94	)	)	PUNCT
ejpam-1676	613	95	0.022(0.033	0.022(0.033	NOUN
ejpam-1676	613	96	)	)	PUNCT
ejpam-1676	613	97	0.757(0.186	0.757(0.186	PROPN
ejpam-1676	613	98	)	)	PUNCT
ejpam-1676	613	99	0.807	0.807	NUM
ejpam-1676	613	100	30	30	NUM
ejpam-1676	613	101	24	24	NUM
ejpam-1676	613	102	1.006(0.128	1.006(0.128	NUM
ejpam-1676	613	103	)	)	PUNCT
ejpam-1676	613	104	0.016(0.025	0.016(0.025	NOUN
ejpam-1676	613	105	)	)	PUNCT
ejpam-1676	613	106	0.993(0.113	0.993(0.113	ADJ
ejpam-1676	613	107	)	)	PUNCT
ejpam-1676	613	108	0.013(0.02	0.013(0.02	NUM
ejpam-1676	613	109	)	)	PUNCT
ejpam-1676	613	110	0.754(0.188	0.754(0.188	NOUN
ejpam-1676	613	111	)	)	PUNCT
ejpam-1676	613	112	0.778	0.778	NUM
ejpam-1676	613	113	40	40	NUM
ejpam-1676	613	114	32	32	NUM
ejpam-1676	613	115	1.001(0.114	1.001(0.114	NUM
ejpam-1676	613	116	)	)	PUNCT
ejpam-1676	613	117	0.013(0.019	0.013(0.019	NOUN
ejpam-1676	613	118	)	)	PUNCT
ejpam-1676	613	119	0.991(0.102	0.991(0.102	NOUN
ejpam-1676	613	120	)	)	PUNCT
ejpam-1676	613	121	0.01(0.016	0.01(0.016	NUM
ejpam-1676	613	122	)	)	PUNCT
ejpam-1676	613	123	0.767(0.187	0.767(0.187	NUM
ejpam-1676	613	124	)	)	PUNCT
ejpam-1676	613	125	0.813	0.813	NUM
ejpam-1676	613	126	50	50	NUM
ejpam-1676	613	127	40	40	NUM
ejpam-1676	613	128	1.005(0.098	1.005(0.098	NUM
ejpam-1676	613	129	)	)	PUNCT
ejpam-1676	613	130	0.01(0.014	0.01(0.014	NOUN
ejpam-1676	613	131	)	)	PUNCT
ejpam-1676	613	132	0.994(0.09	0.994(0.09	PROPN
ejpam-1676	613	133	)	)	PUNCT
ejpam-1676	613	134	0.008(0.012	0.008(0.012	NOUN
ejpam-1676	613	135	)	)	PUNCT
ejpam-1676	613	136	0.781(0.194	0.781(0.194	NOUN
ejpam-1676	613	137	)	)	PUNCT
ejpam-1676	613	138	0.851	0.851	NUM
ejpam-1676	613	139	60	60	NUM
ejpam-1676	613	140	48	48	NUM
ejpam-1676	613	141	0.999(0.094	0.999(0.094	NOUN
ejpam-1676	613	142	)	)	PUNCT
ejpam-1676	613	143	0.009(0.013	0.009(0.013	NOUN
ejpam-1676	613	144	)	)	PUNCT
ejpam-1676	613	145	0.99(0.088	0.99(0.088	NUM
ejpam-1676	613	146	)	)	PUNCT
ejpam-1676	613	147	0.008(0.012	0.008(0.012	NOUN
ejpam-1676	613	148	)	)	PUNCT
ejpam-1676	613	149	0.774(0.192	0.774(0.192	NOUN
ejpam-1676	613	150	)	)	PUNCT
ejpam-1676	613	151	0.875	0.875	NUM
ejpam-1676	613	152	10	10	NUM
ejpam-1676	613	153	10	10	NUM
ejpam-1676	613	154	1.004(0.241	1.004(0.241	NUM
ejpam-1676	613	155	)	)	PUNCT
ejpam-1676	613	156	0.058(0.085	0.058(0.085	NOUN
ejpam-1676	613	157	)	)	PUNCT
ejpam-1676	613	158	0.985(0.206	0.985(0.206	NOUN
ejpam-1676	613	159	)	)	PUNCT
ejpam-1676	613	160	0.042(0.063	0.042(0.063	NUM
ejpam-1676	613	161	)	)	PUNCT
ejpam-1676	613	162	0.742(0.186	0.742(0.186	PROPN
ejpam-1676	613	163	)	)	PUNCT
ejpam-1676	613	164	0.734	0.734	NUM
ejpam-1676	613	165	20	20	NUM
ejpam-1676	613	166	20	20	NUM
ejpam-1676	613	167	1.011(0.164	1.011(0.164	NOUN
ejpam-1676	613	168	)	)	PUNCT
ejpam-1676	613	169	0.027(0.04	0.027(0.04	NOUN
ejpam-1676	613	170	)	)	PUNCT
ejpam-1676	613	171	0.996(0.143	0.996(0.143	NOUN
ejpam-1676	613	172	)	)	PUNCT
ejpam-1676	613	173	0.02(0.032	0.02(0.032	NOUN
ejpam-1676	613	174	)	)	PUNCT
ejpam-1676	613	175	0.743(0.188	0.743(0.188	PROPN
ejpam-1676	613	176	)	)	PUNCT
ejpam-1676	613	177	0.751	0.751	NUM
ejpam-1676	613	178	30	30	NUM
ejpam-1676	613	179	30	30	NUM
ejpam-1676	613	180	1.003(0.134	1.003(0.134	NUM
ejpam-1676	613	181	)	)	PUNCT
ejpam-1676	613	182	0.018(0.026	0.018(0.026	PROPN
ejpam-1676	613	183	)	)	PUNCT
ejpam-1676	613	184	0.989(0.117	0.989(0.117	ADJ
ejpam-1676	613	185	)	)	PUNCT
ejpam-1676	613	186	0.014(0.02	0.014(0.02	NUM
ejpam-1676	613	187	)	)	PUNCT
ejpam-1676	613	188	0.758(0.188	0.758(0.188	NOUN
ejpam-1676	613	189	)	)	PUNCT
ejpam-1676	613	190	0.765	0.765	NUM
ejpam-1676	613	191	40	40	NUM
ejpam-1676	613	192	40	40	NUM
ejpam-1676	613	193	1.001(0.112	1.001(0.112	NUM
ejpam-1676	613	194	)	)	PUNCT
ejpam-1676	613	195	0.013(0.018	0.013(0.018	NOUN
ejpam-1676	613	196	)	)	PUNCT
ejpam-1676	613	197	0.988(0.1	0.988(0.1	NOUN
ejpam-1676	613	198	)	)	PUNCT
ejpam-1676	613	199	0.01(0.015	0.01(0.015	NUM
ejpam-1676	613	200	)	)	PUNCT
ejpam-1676	613	201	0.761(0.188	0.761(0.188	NOUN
ejpam-1676	613	202	)	)	PUNCT
ejpam-1676	613	203	0.808	0.808	NUM
ejpam-1676	613	204	50	50	NUM
ejpam-1676	613	205	50	50	NUM
ejpam-1676	613	206	0.994(0.099	0.994(0.099	NUM
ejpam-1676	613	207	)	)	PUNCT
ejpam-1676	613	208	0.01(0.015	0.01(0.015	ADV
ejpam-1676	613	209	)	)	PUNCT
ejpam-1676	613	210	0.981(0.088	0.981(0.088	NUM
ejpam-1676	613	211	)	)	PUNCT
ejpam-1676	613	212	0.008(0.012	0.008(0.012	NOUN
ejpam-1676	613	213	)	)	PUNCT
ejpam-1676	613	214	0.764(0.189	0.764(0.189	NUM
ejpam-1676	613	215	)	)	PUNCT
ejpam-1676	613	216	0.812	0.812	NUM
ejpam-1676	613	217	60	60	NUM
ejpam-1676	613	218	60	60	NUM
ejpam-1676	613	219	0.999(0.088	0.999(0.088	NOUN
ejpam-1676	613	220	)	)	PUNCT
ejpam-1676	613	221	0.008(0.011	0.008(0.011	ADV
ejpam-1676	613	222	)	)	PUNCT
ejpam-1676	613	223	0.986(0.081	0.986(0.081	NOUN
ejpam-1676	613	224	)	)	PUNCT
ejpam-1676	613	225	0.007(0.009	0.007(0.009	NOUN
ejpam-1676	613	226	)	)	PUNCT
ejpam-1676	613	227	0.77(0.187	0.77(0.187	NOUN
ejpam-1676	613	228	)	)	PUNCT
ejpam-1676	613	229	0.867	0.867	NUM
ejpam-1676	613	230	10	10	NUM
ejpam-1676	613	231	15	15	NUM
ejpam-1676	613	232	0.986(0.249	0.986(0.249	NUM
ejpam-1676	613	233	)	)	PUNCT
ejpam-1676	613	234	0.062(0.108	0.062(0.108	NUM
ejpam-1676	613	235	)	)	PUNCT
ejpam-1676	613	236	0.966(0.196	0.966(0.196	ADJ
ejpam-1676	613	237	)	)	PUNCT
ejpam-1676	613	238	0.04(0.077	0.04(0.077	NOUN
ejpam-1676	613	239	)	)	PUNCT
ejpam-1676	613	240	0.717(0.181	0.717(0.181	NOUN
ejpam-1676	613	241	)	)	PUNCT
ejpam-1676	613	242	0.638	0.638	NUM
ejpam-1676	613	243	20	20	NUM
ejpam-1676	613	244	30	30	NUM
ejpam-1676	613	245	1.006(0.161	1.006(0.161	NUM
ejpam-1676	613	246	)	)	PUNCT
ejpam-1676	613	247	0.026(0.039	0.026(0.039	NOUN
ejpam-1676	613	248	)	)	PUNCT
ejpam-1676	613	249	0.98(0.131	0.98(0.131	NUM
ejpam-1676	613	250	)	)	PUNCT
ejpam-1676	613	251	0.017(0.026	0.017(0.026	NUM
ejpam-1676	613	252	)	)	PUNCT
ejpam-1676	613	253	0.72(0.186	0.72(0.186	NOUN
ejpam-1676	613	254	)	)	PUNCT
ejpam-1676	613	255	0.675	0.675	NUM
ejpam-1676	613	256	30	30	NUM
ejpam-1676	613	257	45	45	NUM
ejpam-1676	613	258	0.995(0.13	0.995(0.13	NUM
ejpam-1676	613	259	)	)	PUNCT
ejpam-1676	613	260	0.017(0.025	0.017(0.025	NUM
ejpam-1676	613	261	)	)	PUNCT
ejpam-1676	613	262	0.977(0.106	0.977(0.106	NOUN
ejpam-1676	613	263	)	)	PUNCT
ejpam-1676	613	264	0.012(0.018	0.012(0.018	NOUN
ejpam-1676	613	265	)	)	PUNCT
ejpam-1676	613	266	0.74(0.184	0.74(0.184	NUM
ejpam-1676	613	267	)	)	PUNCT
ejpam-1676	613	268	0.693	0.693	NUM
ejpam-1676	613	269	40	40	NUM
ejpam-1676	613	270	60	60	NUM
ejpam-1676	613	271	1.002(0.109	1.002(0.109	NUM
ejpam-1676	613	272	)	)	PUNCT
ejpam-1676	613	273	0.012(0.017	0.012(0.017	NOUN
ejpam-1676	613	274	)	)	PUNCT
ejpam-1676	613	275	0.984(0.095	0.984(0.095	NOUN
ejpam-1676	613	276	)	)	PUNCT
ejpam-1676	613	277	0.009(0.014	0.009(0.014	NOUN
ejpam-1676	613	278	)	)	PUNCT
ejpam-1676	613	279	0.753(0.188	0.753(0.188	NOUN
ejpam-1676	613	280	)	)	PUNCT
ejpam-1676	613	281	0.78	0.78	NUM
ejpam-1676	613	282	50	50	NUM
ejpam-1676	613	283	75	75	NUM
ejpam-1676	613	284	1.002(0.101	1.002(0.101	NUM
ejpam-1676	613	285	)	)	PUNCT
ejpam-1676	613	286	0.01(0.015	0.01(0.015	NUM
ejpam-1676	613	287	)	)	PUNCT
ejpam-1676	613	288	0.982(0.084	0.982(0.084	NOUN
ejpam-1676	613	289	)	)	PUNCT
ejpam-1676	613	290	0.007(0.011	0.007(0.011	X
ejpam-1676	613	291	)	)	PUNCT
ejpam-1676	613	292	0.754(0.184	0.754(0.184	NOUN
ejpam-1676	613	293	)	)	PUNCT
ejpam-1676	613	294	0.729	0.729	NUM
ejpam-1676	613	295	60	60	NUM
ejpam-1676	613	296	90	90	NUM
ejpam-1676	613	297	1.002(0.096	1.002(0.096	NUM
ejpam-1676	613	298	)	)	PUNCT
ejpam-1676	613	299	0.009(0.014	0.009(0.014	NOUN
ejpam-1676	613	300	)	)	PUNCT
ejpam-1676	613	301	0.986(0.086	0.986(0.086	NOUN
ejpam-1676	613	302	)	)	PUNCT
ejpam-1676	613	303	0.007(0.011	0.007(0.011	NOUN
ejpam-1676	613	304	)	)	PUNCT
ejpam-1676	613	305	0.761(0.185	0.761(0.185	NOUN
ejpam-1676	613	306	)	)	PUNCT
ejpam-1676	613	307	0.815	0.815	NUM
ejpam-1676	613	308	p.	p.	NOUN
ejpam-1676	613	309	guo	guo	PROPN
ejpam-1676	613	310	,	,	PUNCT
ejpam-1676	613	311	x.	x.	PROPN
ejpam-1676	613	312	wang	wang	PROPN
ejpam-1676	613	313	,	,	PUNCT
ejpam-1676	613	314	y.	y.	PROPN
ejpam-1676	613	315	wu	wu	PROPN
ejpam-1676	613	316	/	/	SYM
ejpam-1676	613	317	eur	eur	PROPN
ejpam-1676	613	318	.	.	PUNCT
ejpam-1676	614	1	j.	j.	PROPN
ejpam-1676	614	2	pure	pure	PROPN
ejpam-1676	614	3	appl	appl	PROPN
ejpam-1676	614	4	.	.	PROPN
ejpam-1676	614	5	math	math	PROPN
ejpam-1676	614	6	,	,	PUNCT
ejpam-1676	614	7	5	5	NUM
ejpam-1676	614	8	(	(	PUNCT
ejpam-1676	614	9	2012	2012	NUM
ejpam-1676	614	10	)	)	PUNCT
ejpam-1676	614	11	,	,	PUNCT
ejpam-1676	614	12	333	333	NUM
ejpam-1676	614	13	-	-	SYM
ejpam-1676	614	14	356	356	NUM
ejpam-1676	614	15	354	354	NUM
ejpam-1676	614	16	6	6	NUM
ejpam-1676	614	17	.	.	PUNCT
ejpam-1676	615	1	a	a	DET
ejpam-1676	615	2	real	real	ADJ
ejpam-1676	615	3	data	datum	NOUN
ejpam-1676	615	4	example	example	NOUN
ejpam-1676	615	5	the	the	DET
ejpam-1676	615	6	low	low	ADJ
ejpam-1676	615	7	birth	birth	NOUN
ejpam-1676	615	8	weight	weight	NOUN
ejpam-1676	615	9	data	datum	NOUN
ejpam-1676	615	10	comes	come	VERB
ejpam-1676	615	11	from	from	ADP
ejpam-1676	615	12	a	a	DET
ejpam-1676	615	13	study	study	NOUN
ejpam-1676	615	14	of	of	ADP
ejpam-1676	615	15	risk	risk	NOUN
ejpam-1676	615	16	factors	factor	NOUN
ejpam-1676	615	17	associated	associate	VERB
ejpam-1676	615	18	with	with	ADP
ejpam-1676	615	19	low	low	ADJ
ejpam-1676	615	20	infant	infant	NOUN
ejpam-1676	615	21	birth	birth	NOUN
ejpam-1676	615	22	weight.the	weight.the	DET
ejpam-1676	615	23	data	datum	NOUN
ejpam-1676	615	24	were	be	AUX
ejpam-1676	615	25	collected	collect	VERB
ejpam-1676	615	26	at	at	ADP
ejpam-1676	615	27	baystate	baystate	PROPN
ejpam-1676	615	28	medical	medical	ADJ
ejpam-1676	615	29	center	center	PROPN
ejpam-1676	615	30	,	,	PUNCT
ejpam-1676	615	31	springfield	springfield	PROPN
ejpam-1676	615	32	,	,	PUNCT
ejpam-1676	615	33	massachusetts	massachusetts	PROPN
ejpam-1676	615	34	,	,	PUNCT
ejpam-1676	615	35	during	during	ADP
ejpam-1676	615	36	1986	1986	NUM
ejpam-1676	615	37	and	and	CCONJ
ejpam-1676	615	38	presented	present	VERB
ejpam-1676	615	39	in	in	ADP
ejpam-1676	615	40	[	[	X
ejpam-1676	615	41	8	8	NUM
ejpam-1676	615	42	]	]	PUNCT
ejpam-1676	615	43	.	.	PUNCT
ejpam-1676	616	1	the	the	DET
ejpam-1676	616	2	goal	goal	NOUN
ejpam-1676	616	3	of	of	ADP
ejpam-1676	616	4	this	this	DET
ejpam-1676	616	5	study	study	NOUN
ejpam-1676	616	6	was	be	AUX
ejpam-1676	616	7	to	to	PART
ejpam-1676	616	8	identify	identify	VERB
ejpam-1676	616	9	risk	risk	NOUN
ejpam-1676	616	10	factors	factor	NOUN
ejpam-1676	616	11	associated	associate	VERB
ejpam-1676	616	12	with	with	ADP
ejpam-1676	616	13	giving	give	VERB
ejpam-1676	616	14	birth	birth	NOUN
ejpam-1676	616	15	to	to	ADP
ejpam-1676	616	16	a	a	DET
ejpam-1676	616	17	low	low	ADJ
ejpam-1676	616	18	birth	birth	NOUN
ejpam-1676	616	19	weight	weight	NOUN
ejpam-1676	616	20	baby	baby	NOUN
ejpam-1676	616	21	who	who	PRON
ejpam-1676	616	22	weighs	weigh	VERB
ejpam-1676	616	23	less	less	ADJ
ejpam-1676	616	24	than	than	ADP
ejpam-1676	616	25	2500	2500	NUM
ejpam-1676	616	26	grams	gram	NOUN
ejpam-1676	616	27	.	.	PUNCT
ejpam-1676	617	1	in	in	ADP
ejpam-1676	617	2	this	this	DET
ejpam-1676	617	3	study	study	NOUN
ejpam-1676	617	4	data	datum	NOUN
ejpam-1676	617	5	were	be	AUX
ejpam-1676	617	6	collected	collect	VERB
ejpam-1676	617	7	on	on	ADP
ejpam-1676	617	8	189	189	NUM
ejpam-1676	617	9	women	woman	NOUN
ejpam-1676	617	10	,	,	PUNCT
ejpam-1676	617	11	59	59	NUM
ejpam-1676	617	12	of	of	ADP
ejpam-1676	617	13	which	which	PRON
ejpam-1676	617	14	had	have	VERB
ejpam-1676	617	15	low	low	ADJ
ejpam-1676	617	16	birth	birth	NOUN
ejpam-1676	617	17	weight	weight	NOUN
ejpam-1676	617	18	babies	baby	NOUN
ejpam-1676	617	19	and	and	CCONJ
ejpam-1676	617	20	130	130	NUM
ejpam-1676	617	21	of	of	ADP
ejpam-1676	617	22	which	which	PRON
ejpam-1676	617	23	had	have	VERB
ejpam-1676	617	24	normal	normal	ADJ
ejpam-1676	617	25	birth	birth	NOUN
ejpam-1676	617	26	weight	weight	NOUN
ejpam-1676	617	27	babies	baby	NOUN
ejpam-1676	617	28	.	.	PUNCT
ejpam-1676	618	1	variables	variable	NOUN
ejpam-1676	618	2	which	which	PRON
ejpam-1676	618	3	were	be	AUX
ejpam-1676	618	4	thought	think	VERB
ejpam-1676	618	5	to	to	PART
ejpam-1676	618	6	be	be	AUX
ejpam-1676	618	7	of	of	ADP
ejpam-1676	618	8	importance	importance	NOUN
ejpam-1676	618	9	were	be	AUX
ejpam-1676	618	10	weight	weight	NOUN
ejpam-1676	618	11	of	of	ADP
ejpam-1676	618	12	the	the	DET
ejpam-1676	618	13	subject	subject	NOUN
ejpam-1676	618	14	at	at	ADP
ejpam-1676	618	15	her	her	PRON
ejpam-1676	618	16	last	last	ADJ
ejpam-1676	618	17	menstrual	menstrual	ADJ
ejpam-1676	618	18	period	period	NOUN
ejpam-1676	618	19	and	and	CCONJ
ejpam-1676	618	20	race	race	NOUN
ejpam-1676	618	21	.	.	PUNCT
ejpam-1676	619	1	these	these	DET
ejpam-1676	619	2	two	two	NUM
ejpam-1676	619	3	variables	variable	NOUN
ejpam-1676	619	4	are	be	AUX
ejpam-1676	619	5	denoted	denote	VERB
ejpam-1676	619	6	as	as	ADP
ejpam-1676	619	7	"	"	PUNCT
ejpam-1676	619	8	lwt	lwt	NOUN
ejpam-1676	619	9	"	"	PUNCT
ejpam-1676	619	10	and	and	CCONJ
ejpam-1676	619	11	"	"	PUNCT
ejpam-1676	619	12	race	race	NOUN
ejpam-1676	619	13	"	"	PUNCT
ejpam-1676	619	14	.	.	PUNCT
ejpam-1676	620	1	the	the	DET
ejpam-1676	620	2	response	response	NOUN
ejpam-1676	620	3	"	"	PUNCT
ejpam-1676	620	4	low	low	ADJ
ejpam-1676	620	5	"	"	PUNCT
ejpam-1676	620	6	,	,	PUNCT
ejpam-1676	620	7	is	be	AUX
ejpam-1676	620	8	a	a	DET
ejpam-1676	620	9	binary	binary	ADJ
ejpam-1676	620	10	variable	variable	NOUN
ejpam-1676	620	11	which	which	PRON
ejpam-1676	620	12	indicates	indicate	VERB
ejpam-1676	620	13	a	a	DET
ejpam-1676	620	14	low	low	ADJ
ejpam-1676	620	15	birth	birth	NOUN
ejpam-1676	620	16	weight	weight	NOUN
ejpam-1676	620	17	baby	baby	NOUN
ejpam-1676	620	18	if	if	SCONJ
ejpam-1676	620	19	equals	equal	VERB
ejpam-1676	620	20	to	to	ADP
ejpam-1676	620	21	1	1	NUM
ejpam-1676	620	22	,	,	PUNCT
ejpam-1676	620	23	and	and	CCONJ
ejpam-1676	620	24	normal	normal	ADJ
ejpam-1676	620	25	birth	birth	NOUN
ejpam-1676	620	26	weight	weight	NOUN
ejpam-1676	620	27	if	if	SCONJ
ejpam-1676	620	28	equals	equal	VERB
ejpam-1676	620	29	to	to	ADP
ejpam-1676	620	30	0	0	NUM
ejpam-1676	620	31	.	.	PUNCT
ejpam-1676	621	1	the	the	DET
ejpam-1676	621	2	variable	variable	ADJ
ejpam-1676	621	3	"	"	PUNCT
ejpam-1676	621	4	race	race	NOUN
ejpam-1676	621	5	"	"	PUNCT
ejpam-1676	621	6	has	have	AUX
ejpam-1676	621	7	been	be	AUX
ejpam-1676	621	8	recoded	recode	VERB
ejpam-1676	621	9	using	use	VERB
ejpam-1676	621	10	the	the	DET
ejpam-1676	621	11	two	two	NUM
ejpam-1676	621	12	dummy	dummy	ADJ
ejpam-1676	621	13	variables	variable	NOUN
ejpam-1676	621	14	.	.	PUNCT
ejpam-1676	622	1	first	first	ADV
ejpam-1676	622	2	of	of	ADP
ejpam-1676	622	3	all	all	PRON
ejpam-1676	622	4	,	,	PUNCT
ejpam-1676	622	5	we	we	PRON
ejpam-1676	622	6	investigate	investigate	VERB
ejpam-1676	622	7	on	on	ADP
ejpam-1676	622	8	the	the	DET
ejpam-1676	622	9	model	model	NOUN
ejpam-1676	622	10	g(µ	g(µ	PROPN
ejpam-1676	622	11	)	)	PUNCT
ejpam-1676	622	12	=	=	SYM
ejpam-1676	623	1	β0	β0	PROPN
ejpam-1676	623	2	+	+	CCONJ
ejpam-1676	623	3	β1	β1	PROPN
ejpam-1676	623	4	lw	lw	PROPN
ejpam-1676	623	5	t	t	PROPN
ejpam-1676	623	6	,	,	PUNCT
ejpam-1676	623	7	where	where	SCONJ
ejpam-1676	623	8	µ	µ	NOUN
ejpam-1676	623	9	is	be	AUX
ejpam-1676	623	10	the	the	DET
ejpam-1676	623	11	expected	expect	VERB
ejpam-1676	623	12	value	value	NOUN
ejpam-1676	623	13	of	of	ADP
ejpam-1676	623	14	the	the	DET
ejpam-1676	623	15	binary	binary	ADJ
ejpam-1676	623	16	response	response	NOUN
ejpam-1676	623	17	and	and	CCONJ
ejpam-1676	623	18	g	g	NOUN
ejpam-1676	623	19	is	be	AUX
ejpam-1676	623	20	the	the	DET
ejpam-1676	623	21	link	link	NOUN
ejpam-1676	623	22	function	function	NOUN
ejpam-1676	623	23	for	for	ADP
ejpam-1676	623	24	binomial	binomial	ADJ
ejpam-1676	623	25	family	family	NOUN
ejpam-1676	623	26	.	.	PUNCT
ejpam-1676	624	1	we	we	PRON
ejpam-1676	624	2	find	find	VERB
ejpam-1676	624	3	that	that	SCONJ
ejpam-1676	624	4	the	the	DET
ejpam-1676	624	5	variable	variable	ADJ
ejpam-1676	624	6	lwt	lwt	NOUN
ejpam-1676	624	7	is	be	AUX
ejpam-1676	624	8	significant	significant	ADJ
ejpam-1676	624	9	.	.	PUNCT
ejpam-1676	625	1	the	the	DET
ejpam-1676	625	2	estimate	estimate	NOUN
ejpam-1676	625	3	of	of	ADP
ejpam-1676	625	4	the	the	DET
ejpam-1676	625	5	coefficients	coefficient	NOUN
ejpam-1676	625	6	and	and	CCONJ
ejpam-1676	625	7	the	the	DET
ejpam-1676	625	8	corresponding	correspond	VERB
ejpam-1676	625	9	p	p	NOUN
ejpam-1676	625	10	-	-	PUNCT
ejpam-1676	625	11	values	value	NOUN
ejpam-1676	625	12	are	be	AUX
ejpam-1676	625	13	given	give	VERB
ejpam-1676	625	14	in	in	ADP
ejpam-1676	625	15	table	table	NOUN
ejpam-1676	625	16	2	2	NUM
ejpam-1676	625	17	.	.	PUNCT
ejpam-1676	625	18	one	one	PRON
ejpam-1676	625	19	can	can	AUX
ejpam-1676	625	20	also	also	ADV
ejpam-1676	625	21	verify	verify	VERB
ejpam-1676	625	22	that	that	SCONJ
ejpam-1676	625	23	the	the	DET
ejpam-1676	625	24	interaction	interaction	NOUN
ejpam-1676	625	25	between	between	ADP
ejpam-1676	625	26	lwt	lwt	NOUN
ejpam-1676	625	27	and	and	CCONJ
ejpam-1676	625	28	race	race	NOUN
ejpam-1676	625	29	is	be	AUX
ejpam-1676	625	30	not	not	PART
ejpam-1676	625	31	significant	significant	ADJ
ejpam-1676	625	32	.	.	PUNCT
ejpam-1676	626	1	table	table	NOUN
ejpam-1676	626	2	2	2	NUM
ejpam-1676	626	3	:	:	PUNCT
ejpam-1676	626	4	estimated	estimate	VERB
ejpam-1676	626	5	coe	coe	PROPN
ejpam-1676	626	6	�	�	PROPN
ejpam-1676	626	7	ients	ient	NOUN
ejpam-1676	626	8	for	for	ADP
ejpam-1676	626	9	general	general	ADJ
ejpam-1676	626	10	logisti	logisti	PROPN
ejpam-1676	626	11	regression	regression	NOUN
ejpam-1676	626	12	model	model	NOUN
ejpam-1676	626	13	using	use	VERB
ejpam-1676	626	14	the	the	DET
ejpam-1676	626	15	variables	variable	NOUN
ejpam-1676	626	16	lwt	lwt	PROPN
ejpam-1676	626	17	.	.	PUNCT
ejpam-1676	626	18	estimate	estimate	VERB
ejpam-1676	626	19	p	p	NUM
ejpam-1676	626	20	-	-	PUNCT
ejpam-1676	626	21	value	value	NOUN
ejpam-1676	626	22	intercept	intercept	NOUN
ejpam-1676	626	23	0.9983	0.9983	NUM
ejpam-1676	626	24	0.2036	0.2036	NUM
ejpam-1676	626	25	lwt	lwt	NOUN
ejpam-1676	626	26	-0.0141	-0.0141	PUNCT
ejpam-1676	626	27	0.0227	0.0227	NUM
ejpam-1676	627	1	then	then	ADV
ejpam-1676	627	2	we	we	PRON
ejpam-1676	627	3	check	check	VERB
ejpam-1676	627	4	the	the	DET
ejpam-1676	627	5	significance	significance	NOUN
ejpam-1676	627	6	of	of	ADP
ejpam-1676	627	7	lwt	lwt	NOUN
ejpam-1676	627	8	within	within	ADP
ejpam-1676	627	9	each	each	DET
ejpam-1676	627	10	race	race	NOUN
ejpam-1676	627	11	individually	individually	ADV
ejpam-1676	627	12	.	.	PUNCT
ejpam-1676	628	1	the	the	DET
ejpam-1676	628	2	p	p	NOUN
ejpam-1676	628	3	-	-	PUNCT
ejpam-1676	628	4	values	value	NOUN
ejpam-1676	628	5	are	be	AUX
ejpam-1676	628	6	all	all	ADV
ejpam-1676	628	7	greater	great	ADJ
ejpam-1676	628	8	than	than	ADP
ejpam-1676	628	9	the	the	DET
ejpam-1676	628	10	p	p	NOUN
ejpam-1676	628	11	-	-	PUNCT
ejpam-1676	628	12	value	value	NOUN
ejpam-1676	628	13	of	of	ADP
ejpam-1676	628	14	lwt	lwt	NOUN
ejpam-1676	628	15	for	for	ADP
ejpam-1676	628	16	each	each	DET
ejpam-1676	628	17	group	group	NOUN
ejpam-1676	628	18	is	be	AUX
ejpam-1676	628	19	given	give	VERB
ejpam-1676	628	20	in	in	ADP
ejpam-1676	628	21	table	table	NOUN
ejpam-1676	628	22	3	3	NUM
ejpam-1676	628	23	.	.	PUNCT
ejpam-1676	629	1	despite	despite	SCONJ
ejpam-1676	629	2	the	the	DET
ejpam-1676	629	3	fact	fact	NOUN
ejpam-1676	629	4	that	that	SCONJ
ejpam-1676	629	5	the	the	DET
ejpam-1676	629	6	variable	variable	ADJ
ejpam-1676	629	7	lwt	lwt	NOUN
ejpam-1676	629	8	is	be	AUX
ejpam-1676	629	9	significant	significant	ADJ
ejpam-1676	629	10	for	for	ADP
ejpam-1676	629	11	the	the	DET
ejpam-1676	629	12	overall	overall	ADJ
ejpam-1676	629	13	model	model	NOUN
ejpam-1676	629	14	,	,	PUNCT
ejpam-1676	629	15	it	it	PRON
ejpam-1676	629	16	is	be	AUX
ejpam-1676	629	17	very	very	ADV
ejpam-1676	629	18	surprising	surprising	ADJ
ejpam-1676	629	19	to	to	PART
ejpam-1676	629	20	see	see	VERB
ejpam-1676	629	21	that	that	SCONJ
ejpam-1676	629	22	the	the	DET
ejpam-1676	629	23	lwt	lwt	NOUN
ejpam-1676	629	24	is	be	AUX
ejpam-1676	629	25	not	not	PART
ejpam-1676	629	26	significant	significant	ADJ
ejpam-1676	629	27	within	within	ADP
ejpam-1676	629	28	any	any	DET
ejpam-1676	629	29	group	group	NOUN
ejpam-1676	629	30	.	.	PUNCT
ejpam-1676	630	1	by	by	ADP
ejpam-1676	630	2	applying	apply	VERB
ejpam-1676	630	3	the	the	DET
ejpam-1676	630	4	wle	wle	NOUN
ejpam-1676	630	5	method	method	NOUN
ejpam-1676	630	6	based	base	VERB
ejpam-1676	630	7	on	on	ADP
ejpam-1676	630	8	distribution	distribution	NOUN
ejpam-1676	630	9	from	from	ADP
ejpam-1676	630	10	bootstraping	bootstrap	VERB
ejpam-1676	630	11	,	,	PUNCT
ejpam-1676	630	12	we	we	PRON
ejpam-1676	630	13	find	find	VERB
ejpam-1676	630	14	the	the	DET
ejpam-1676	630	15	lwt	lwt	NOUN
ejpam-1676	630	16	is	be	AUX
ejpam-1676	630	17	significant	significant	ADJ
ejpam-1676	630	18	in	in	ADP
ejpam-1676	630	19	"	"	PUNCT
ejpam-1676	630	20	other	other	ADJ
ejpam-1676	630	21	"	"	PUNCT
ejpam-1676	630	22	group	group	NOUN
ejpam-1676	630	23	.	.	PUNCT
ejpam-1676	631	1	table	table	NOUN
ejpam-1676	631	2	3	3	NUM
ejpam-1676	631	3	:	:	PUNCT
ejpam-1676	632	1	estimated	estimate	VERB
ejpam-1676	632	2	coe	coe	PROPN
ejpam-1676	632	3	�	�	PROPN
ejpam-1676	632	4	ients	ient	NOUN
ejpam-1676	632	5	for	for	ADP
ejpam-1676	632	6	general	general	ADJ
ejpam-1676	632	7	logisti	logisti	PROPN
ejpam-1676	632	8	regression	regression	NOUN
ejpam-1676	632	9	model	model	NOUN
ejpam-1676	632	10	using	use	VERB
ejpam-1676	632	11	the	the	DET
ejpam-1676	632	12	variables	variable	NOUN
ejpam-1676	632	13	lwt	lwt	NOUN
ejpam-1676	632	14	in	in	ADP
ejpam-1676	632	15	ea	ea	NUM
ejpam-1676	632	16	h	h	PROPN
ejpam-1676	632	17	race	race	NOUN
ejpam-1676	632	18	group	group	NOUN
ejpam-1676	632	19	.	.	PUNCT
ejpam-1676	633	1	estimate	estimate	VERB
ejpam-1676	633	2	p	p	NOUN
ejpam-1676	633	3	-	-	PUNCT
ejpam-1676	633	4	value	value	NOUN
ejpam-1676	633	5	“	"	PUNCT
ejpam-1676	633	6	white	white	ADJ
ejpam-1676	633	7	”	"	PUNCT
ejpam-1676	633	8	group	group	NOUN
ejpam-1676	633	9	intercept	intercept	VERB
ejpam-1676	633	10	0.7925	0.7925	NUM
ejpam-1676	633	11	0.528	0.528	NUM
ejpam-1676	633	12	lwt	lwt	NOUN
ejpam-1676	633	13	-0.0151	-0.0151	NOUN
ejpam-1676	633	14	0.123	0.123	NUM
ejpam-1676	633	15	“	"	PUNCT
ejpam-1676	633	16	black	black	ADJ
ejpam-1676	633	17	”	"	PUNCT
ejpam-1676	633	18	group	group	NOUN
ejpam-1676	633	19	intercept	intercept	VERB
ejpam-1676	633	20	0.6941	0.6941	NUM
ejpam-1676	633	21	0.663	0.663	NUM
ejpam-1676	633	22	lwt	lwt	PROPN
ejpam-1676	633	23	-0.0069	-0.0069	NOUN
ejpam-1676	633	24	0.517	0.517	NUM
ejpam-1676	633	25	“	"	PUNCT
ejpam-1676	633	26	other	other	ADJ
ejpam-1676	633	27	”	"	PUNCT
ejpam-1676	633	28	group	group	NOUN
ejpam-1676	633	29	intercept	intercept	VERB
ejpam-1676	633	30	2.7135	2.7135	NUM
ejpam-1676	633	31	0.1066	0.1066	NUM
ejpam-1676	633	32	lwt	lwt	NOUN
ejpam-1676	633	33	-0.0275	-0.0275	NOUN
ejpam-1676	633	34	0.0559	0.0559	NUM
ejpam-1676	633	35	references	reference	NOUN
ejpam-1676	633	36	355	355	NUM
ejpam-1676	633	37	table	table	NOUN
ejpam-1676	633	38	4	4	NUM
ejpam-1676	633	39	:	:	PUNCT
ejpam-1676	633	40	estimated	estimate	VERB
ejpam-1676	633	41	coe	coe	PROPN
ejpam-1676	633	42	�	�	PROPN
ejpam-1676	633	43	ients	ient	NOUN
ejpam-1676	633	44	for	for	ADP
ejpam-1676	633	45	weighted	weight	VERB
ejpam-1676	633	46	logisti	logisti	PROPN
ejpam-1676	633	47	regression	regression	NOUN
ejpam-1676	633	48	model	model	NOUN
ejpam-1676	633	49	using	use	VERB
ejpam-1676	633	50	the	the	DET
ejpam-1676	633	51	variables	variable	NOUN
ejpam-1676	633	52	lwt	lwt	NOUN
ejpam-1676	633	53	with	with	ADP
ejpam-1676	633	54	ea	ea	PROPN
ejpam-1676	633	55	h	h	PROPN
ejpam-1676	633	56	race	race	NOUN
ejpam-1676	633	57	group	group	NOUN
ejpam-1676	633	58	as	as	ADP
ejpam-1676	633	59	the	the	DET
ejpam-1676	633	60	main	main	ADJ
ejpam-1676	633	61	group	group	NOUN
ejpam-1676	633	62	.	.	PUNCT
ejpam-1676	634	1	p	p	X
ejpam-1676	634	2	-	-	PUNCT
ejpam-1676	634	3	values	value	NOUN
ejpam-1676	634	4	are	be	AUX
ejpam-1676	634	5	based	base	VERB
ejpam-1676	634	6	on	on	ADP
ejpam-1676	634	7	bootstrap	bootstrap	NOUN
ejpam-1676	634	8	distribution	distribution	NOUN
ejpam-1676	634	9	.	.	PUNCT
ejpam-1676	635	1	estimate	estimate	VERB
ejpam-1676	635	2	p	p	NOUN
ejpam-1676	635	3	-	-	PUNCT
ejpam-1676	635	4	value	value	NOUN
ejpam-1676	635	5	“	"	PUNCT
ejpam-1676	635	6	white	white	ADJ
ejpam-1676	635	7	”	"	PUNCT
ejpam-1676	635	8	group	group	NOUN
ejpam-1676	635	9	intercept	intercept	VERB
ejpam-1676	635	10	0.7925	0.7925	NUM
ejpam-1676	635	11	0.1296	0.1296	NUM
ejpam-1676	635	12	lwt	lwt	PROPN
ejpam-1676	635	13	-0.0148	-0.0148	NOUN
ejpam-1676	635	14	0.1332	0.1332	NUM
ejpam-1676	635	15	“	"	PUNCT
ejpam-1676	635	16	black	black	ADJ
ejpam-1676	635	17	”	"	PUNCT
ejpam-1676	635	18	group	group	NOUN
ejpam-1676	635	19	intercept	intercept	VERB
ejpam-1676	635	20	0.6941	0.6941	NUM
ejpam-1676	635	21	0.5396	0.5396	NUM
ejpam-1676	635	22	lwt	lwt	NOUN
ejpam-1676	635	23	-0.0075	-0.0075	VERB
ejpam-1676	635	24	0.5788	0.5788	NUM
ejpam-1676	635	25	“	"	PUNCT
ejpam-1676	635	26	other	other	ADJ
ejpam-1676	635	27	”	"	PUNCT
ejpam-1676	635	28	group	group	NOUN
ejpam-1676	635	29	intercept	intercept	VERB
ejpam-1676	635	30	2.7135	2.7135	NUM
ejpam-1676	635	31	0.0408	0.0408	NUM
ejpam-1676	635	32	lwt	lwt	NOUN
ejpam-1676	635	33	-0.0276	-0.0276	NOUN
ejpam-1676	635	34	0.0432	0.0432	NUM
ejpam-1676	635	35	7	7	NUM
ejpam-1676	635	36	.	.	PUNCT
ejpam-1676	636	1	discussion	discussion	NOUN
ejpam-1676	636	2	to	to	PART
ejpam-1676	636	3	obtain	obtain	VERB
ejpam-1676	636	4	a	a	DET
ejpam-1676	636	5	more	more	ADV
ejpam-1676	636	6	efficient	efficient	ADJ
ejpam-1676	636	7	estimator	estimator	NOUN
ejpam-1676	636	8	with	with	ADP
ejpam-1676	636	9	a	a	DET
ejpam-1676	636	10	smaller	small	ADJ
ejpam-1676	636	11	mse	mse	NOUN
ejpam-1676	636	12	when	when	SCONJ
ejpam-1676	636	13	related	related	ADJ
ejpam-1676	636	14	information	information	NOUN
ejpam-1676	636	15	is	be	AUX
ejpam-1676	636	16	available	available	ADJ
ejpam-1676	637	1	,	,	PUNCT
ejpam-1676	637	2	we	we	PRON
ejpam-1676	637	3	proposed	propose	VERB
ejpam-1676	637	4	the	the	DET
ejpam-1676	637	5	weighted	weight	VERB
ejpam-1676	637	6	likelihood	likelihood	NOUN
ejpam-1676	637	7	inference	inference	NOUN
ejpam-1676	637	8	for	for	ADP
ejpam-1676	637	9	the	the	DET
ejpam-1676	637	10	parameters	parameter	NOUN
ejpam-1676	637	11	in	in	ADP
ejpam-1676	637	12	linear	linear	PROPN
ejpam-1676	637	13	and	and	CCONJ
ejpam-1676	637	14	partially	partially	ADV
ejpam-1676	637	15	linear	linear	ADJ
ejpam-1676	637	16	models	model	NOUN
ejpam-1676	637	17	.	.	PUNCT
ejpam-1676	638	1	we	we	PRON
ejpam-1676	638	2	developed	develop	VERB
ejpam-1676	638	3	a	a	DET
ejpam-1676	638	4	data	data	NOUN
ejpam-1676	638	5	-	-	PUNCT
ejpam-1676	638	6	driven	drive	VERB
ejpam-1676	638	7	approach	approach	NOUN
ejpam-1676	638	8	to	to	PART
ejpam-1676	638	9	estimate	estimate	VERB
ejpam-1676	638	10	the	the	DET
ejpam-1676	638	11	weights	weight	NOUN
ejpam-1676	638	12	to	to	PART
ejpam-1676	638	13	address	address	VERB
ejpam-1676	638	14	the	the	DET
ejpam-1676	638	15	vexing	vexing	NOUN
ejpam-1676	638	16	issued	issue	VERB
ejpam-1676	638	17	in	in	ADP
ejpam-1676	638	18	weighted	weight	VERB
ejpam-1676	638	19	likelihood	likelihood	NOUN
ejpam-1676	638	20	inference	inference	NOUN
ejpam-1676	638	21	.	.	PUNCT
ejpam-1676	639	1	the	the	DET
ejpam-1676	639	2	proposed	propose	VERB
ejpam-1676	639	3	estimators	estimator	NOUN
ejpam-1676	639	4	have	have	VERB
ejpam-1676	639	5	the	the	DET
ejpam-1676	639	6	same	same	ADJ
ejpam-1676	639	7	asymptotically	asymptotically	ADV
ejpam-1676	639	8	normal	normal	ADJ
ejpam-1676	639	9	distribution	distribution	NOUN
ejpam-1676	639	10	as	as	ADP
ejpam-1676	639	11	the	the	DET
ejpam-1676	639	12	maximum	maximum	ADJ
ejpam-1676	639	13	likelihood	likelihood	NOUN
ejpam-1676	639	14	estimators	estimator	NOUN
ejpam-1676	639	15	.	.	PUNCT
ejpam-1676	640	1	the	the	DET
ejpam-1676	640	2	advantage	advantage	NOUN
ejpam-1676	640	3	of	of	ADP
ejpam-1676	640	4	the	the	DET
ejpam-1676	640	5	wle	wle	NOUN
ejpam-1676	640	6	over	over	ADP
ejpam-1676	640	7	the	the	DET
ejpam-1676	640	8	classical	classical	ADJ
ejpam-1676	640	9	mle	mle	NOUN
ejpam-1676	640	10	was	be	AUX
ejpam-1676	640	11	illustrated	illustrate	VERB
ejpam-1676	640	12	by	by	ADP
ejpam-1676	640	13	simulation	simulation	NOUN
ejpam-1676	640	14	studies	study	NOUN
ejpam-1676	640	15	.	.	PUNCT
ejpam-1676	641	1	results	result	NOUN
ejpam-1676	641	2	from	from	ADP
ejpam-1676	641	3	these	these	DET
ejpam-1676	641	4	simulation	simulation	NOUN
ejpam-1676	641	5	studies	study	NOUN
ejpam-1676	641	6	suggest	suggest	VERB
ejpam-1676	641	7	that	that	SCONJ
ejpam-1676	641	8	the	the	DET
ejpam-1676	641	9	proposed	propose	VERB
ejpam-1676	641	10	estimators	estimator	NOUN
ejpam-1676	641	11	have	have	VERB
ejpam-1676	641	12	prospective	prospective	ADJ
ejpam-1676	641	13	promises	promise	NOUN
ejpam-1676	641	14	.	.	PUNCT
ejpam-1676	642	1	the	the	DET
ejpam-1676	642	2	simplicity	simplicity	NOUN
ejpam-1676	642	3	and	and	CCONJ
ejpam-1676	642	4	effectiveness	effectiveness	NOUN
ejpam-1676	642	5	of	of	ADP
ejpam-1676	642	6	the	the	DET
ejpam-1676	642	7	proposed	propose	VERB
ejpam-1676	642	8	adaptive	adaptive	ADJ
ejpam-1676	642	9	weights	weight	NOUN
ejpam-1676	642	10	are	be	AUX
ejpam-1676	642	11	also	also	ADV
ejpam-1676	642	12	appealing	appeal	VERB
ejpam-1676	642	13	.	.	PUNCT
ejpam-1676	643	1	we	we	PRON
ejpam-1676	643	2	remark	remark	VERB
ejpam-1676	643	3	that	that	SCONJ
ejpam-1676	643	4	the	the	DET
ejpam-1676	643	5	real	real	ADJ
ejpam-1676	643	6	data	datum	NOUN
ejpam-1676	643	7	set	set	VERB
ejpam-1676	643	8	used	use	VERB
ejpam-1676	643	9	in	in	ADP
ejpam-1676	643	10	this	this	DET
ejpam-1676	643	11	paper	paper	NOUN
ejpam-1676	643	12	actually	actually	ADV
ejpam-1676	643	13	came	come	VERB
ejpam-1676	643	14	from	from	ADP
ejpam-1676	643	15	a	a	DET
ejpam-1676	643	16	longitudinal	longitudinal	ADJ
ejpam-1676	643	17	study	study	NOUN
ejpam-1676	643	18	.	.	PUNCT
ejpam-1676	644	1	thus	thus	ADV
ejpam-1676	644	2	one	one	PRON
ejpam-1676	644	3	must	must	AUX
ejpam-1676	644	4	recognize	recognize	VERB
ejpam-1676	644	5	the	the	DET
ejpam-1676	644	6	within	within	NOUN
ejpam-1676	644	7	-	-	PUNCT
ejpam-1676	644	8	subject	subject	NOUN
ejpam-1676	644	9	and	and	CCONJ
ejpam-1676	644	10	between	between	ADP
ejpam-1676	644	11	-	-	PUNCT
ejpam-1676	644	12	subject	subject	NOUN
ejpam-1676	644	13	variations	variation	NOUN
ejpam-1676	644	14	.	.	PUNCT
ejpam-1676	645	1	further	further	ADJ
ejpam-1676	645	2	investigation	investigation	NOUN
ejpam-1676	645	3	of	of	ADP
ejpam-1676	645	4	the	the	DET
ejpam-1676	645	5	inference	inference	NOUN
ejpam-1676	645	6	based	base	VERB
ejpam-1676	645	7	on	on	ADP
ejpam-1676	645	8	the	the	DET
ejpam-1676	645	9	weighted	weight	VERB
ejpam-1676	645	10	likelihood	likelihood	NOUN
ejpam-1676	645	11	for	for	ADP
ejpam-1676	645	12	mixed	mixed	ADJ
ejpam-1676	645	13	-	-	PUNCT
ejpam-1676	645	14	effect	effect	NOUN
ejpam-1676	645	15	models	model	NOUN
ejpam-1676	645	16	is	be	AUX
ejpam-1676	645	17	required	require	VERB
ejpam-1676	645	18	.	.	PUNCT
ejpam-1676	646	1	acknowledgements	acknowledgement	NOUN
ejpam-1676	646	2	the	the	DET
ejpam-1676	646	3	research	research	NOUN
ejpam-1676	646	4	was	be	AUX
ejpam-1676	646	5	partially	partially	ADV
ejpam-1676	646	6	supported	support	VERB
ejpam-1676	646	7	by	by	ADP
ejpam-1676	646	8	the	the	DET
ejpam-1676	646	9	natural	natural	ADJ
ejpam-1676	646	10	sciences	sciences	PROPN
ejpam-1676	646	11	and	and	CCONJ
ejpam-1676	646	12	engineering	engineering	NOUN
ejpam-1676	646	13	research	research	NOUN
ejpam-1676	646	14	council	council	PROPN
ejpam-1676	646	15	of	of	ADP
ejpam-1676	646	16	canada	canada	PROPN
ejpam-1676	646	17	.	.	PUNCT
ejpam-1676	647	1	references	reference	NOUN
ejpam-1676	647	2	[	[	X
ejpam-1676	647	3	1	1	NUM
ejpam-1676	647	4	]	]	X
ejpam-1676	647	5	j	j	PROPN
ejpam-1676	647	6	shao	shao	PROPN
ejpam-1676	647	7	.	.	PUNCT
ejpam-1676	648	1	mathematical	mathematical	ADJ
ejpam-1676	648	2	statistics	statistic	NOUN
ejpam-1676	648	3	.	.	PUNCT
ejpam-1676	649	1	springer	springer	NOUN
ejpam-1676	649	2	,	,	PUNCT
ejpam-1676	649	3	new	new	PROPN
ejpam-1676	649	4	york	york	PROPN
ejpam-1676	649	5	,	,	PUNCT
ejpam-1676	649	6	2003	2003	NUM
ejpam-1676	649	7	.	.	PUNCT
ejpam-1676	650	1	[	[	X
ejpam-1676	650	2	2	2	NUM
ejpam-1676	650	3	]	]	X
ejpam-1676	650	4	j	j	PROPN
ejpam-1676	650	5	staniswalis	staniswalis	PROPN
ejpam-1676	650	6	.	.	PUNCT
ejpam-1676	651	1	the	the	DET
ejpam-1676	651	2	kernel	kernel	PROPN
ejpam-1676	651	3	estimate	estimate	NOUN
ejpam-1676	651	4	of	of	ADP
ejpam-1676	651	5	a	a	DET
ejpam-1676	651	6	regression	regression	NOUN
ejpam-1676	651	7	function	function	NOUN
ejpam-1676	651	8	in	in	ADP
ejpam-1676	651	9	a	a	DET
ejpam-1676	651	10	likelihood	likelihood	NOUN
ejpam-1676	651	11	-	-	PUNCT
ejpam-1676	651	12	based	base	VERB
ejpam-1676	651	13	models	model	NOUN
ejpam-1676	651	14	.	.	PUNCT
ejpam-1676	652	1	j.	j.	PROPN
ejpam-1676	652	2	am	am	PROPN
ejpam-1676	652	3	.	.	PUNCT
ejpam-1676	653	1	statist	statist	PROPN
ejpam-1676	653	2	.	.	PUNCT
ejpam-1676	654	1	assoc	assoc	PROPN
ejpam-1676	654	2	.	.	PROPN
ejpam-1676	654	3	,	,	PUNCT
ejpam-1676	654	4	84:276–83	84:276–83	NOUN
ejpam-1676	654	5	,	,	PUNCT
ejpam-1676	654	6	1989	1989	NUM
ejpam-1676	654	7	.	.	PUNCT
ejpam-1676	655	1	[	[	X
ejpam-1676	655	2	3	3	X
ejpam-1676	655	3	]	]	X
ejpam-1676	655	4	m	m	VERB
ejpam-1676	655	5	stone	stone	NOUN
ejpam-1676	655	6	.	.	PUNCT
ejpam-1676	656	1	strong	strong	ADJ
ejpam-1676	656	2	inconsistencies	inconsistency	NOUN
ejpam-1676	656	3	from	from	ADP
ejpam-1676	656	4	uniform	uniform	ADJ
ejpam-1676	656	5	priors	prior	NOUN
ejpam-1676	656	6	.	.	PUNCT
ejpam-1676	657	1	j.	j.	PROPN
ejpam-1676	657	2	am	am	PROPN
ejpam-1676	657	3	.	.	PUNCT
ejpam-1676	658	1	statist	statist	PROPN
ejpam-1676	658	2	.	.	PUNCT
ejpam-1676	659	1	assoc	assoc	PROPN
ejpam-1676	659	2	.	.	PROPN
ejpam-1676	659	3	,	,	PUNCT
ejpam-1676	659	4	71:114–116	71:114–116	PROPN
ejpam-1676	659	5	,	,	PUNCT
ejpam-1676	659	6	1976	1976	NUM
ejpam-1676	659	7	.	.	PUNCT
ejpam-1676	660	1	[	[	X
ejpam-1676	660	2	4	4	NUM
ejpam-1676	660	3	]	]	SYM
ejpam-1676	660	4	r	r	NOUN
ejpam-1676	660	5	tibshirani	tibshirani	NOUN
ejpam-1676	660	6	and	and	CCONJ
ejpam-1676	660	7	t	t	PROPN
ejpam-1676	660	8	hastie	hastie	PROPN
ejpam-1676	660	9	.	.	PUNCT
ejpam-1676	661	1	local	local	ADJ
ejpam-1676	661	2	likelihood	likelihood	NOUN
ejpam-1676	661	3	estimation	estimation	NOUN
ejpam-1676	661	4	.	.	PUNCT
ejpam-1676	662	1	j.	j.	PROPN
ejpam-1676	662	2	am	am	PROPN
ejpam-1676	662	3	.	.	PUNCT
ejpam-1676	663	1	statist	statist	PROPN
ejpam-1676	663	2	.	.	PUNCT
ejpam-1676	664	1	assoc	assoc	PROPN
ejpam-1676	664	2	.	.	PROPN
ejpam-1676	664	3	,	,	PUNCT
ejpam-1676	664	4	82:559–67	82:559–67	NUM
ejpam-1676	664	5	,	,	PUNCT
ejpam-1676	664	6	1987	1987	NUM
ejpam-1676	664	7	.	.	PUNCT
ejpam-1676	665	1	references	reference	NOUN
ejpam-1676	665	2	356	356	NUM
ejpam-1676	665	3	[	[	X
ejpam-1676	665	4	5	5	NUM
ejpam-1676	665	5	]	]	PUNCT
ejpam-1676	665	6	x	x	PART
ejpam-1676	665	7	wang	wang	PROPN
ejpam-1676	665	8	.	.	PUNCT
ejpam-1676	666	1	approximating	approximate	VERB
ejpam-1676	666	2	bayesian	bayesian	NOUN
ejpam-1676	666	3	inference	inference	NOUN
ejpam-1676	666	4	by	by	ADP
ejpam-1676	666	5	weighted	weighted	ADJ
ejpam-1676	666	6	likelihood	likelihood	NOUN
ejpam-1676	666	7	.	.	PUNCT
ejpam-1676	667	1	can	can	AUX
ejpam-1676	667	2	.	.	PUNCT
ejpam-1676	668	1	j.	j.	PROPN
ejpam-1676	668	2	statist	statist	PROPN
ejpam-1676	668	3	.	.	PROPN
ejpam-1676	668	4	,	,	PUNCT
ejpam-1676	668	5	34:279–298	34:279–298	NUM
ejpam-1676	668	6	,	,	PUNCT
ejpam-1676	668	7	2006	2006	NUM
ejpam-1676	668	8	.	.	PUNCT
ejpam-1676	669	1	[	[	X
ejpam-1676	669	2	6	6	NUM
ejpam-1676	669	3	]	]	PUNCT
ejpam-1676	669	4	x	x	PROPN
ejpam-1676	669	5	wang	wang	PROPN
ejpam-1676	669	6	and	and	CCONJ
ejpam-1676	669	7	j	j	PROPN
ejpam-1676	669	8	zidek	zidek	PROPN
ejpam-1676	669	9	.	.	PUNCT
ejpam-1676	670	1	derivation	derivation	NOUN
ejpam-1676	670	2	of	of	ADP
ejpam-1676	670	3	mixture	mixture	NOUN
ejpam-1676	670	4	distribution	distribution	NOUN
ejpam-1676	670	5	and	and	CCONJ
ejpam-1676	670	6	weighted	weighted	ADJ
ejpam-1676	670	7	likelihood	likelihood	NOUN
ejpam-1676	670	8	as	as	ADP
ejpam-1676	670	9	minimizers	minimizer	NOUN
ejpam-1676	670	10	of	of	ADP
ejpam-1676	670	11	kl	kl	NOUN
ejpam-1676	670	12	-	-	NOUN
ejpam-1676	670	13	divergence	divergence	NOUN
ejpam-1676	670	14	subject	subject	NOUN
ejpam-1676	670	15	to	to	ADP
ejpam-1676	670	16	constraints	constraint	NOUN
ejpam-1676	670	17	.	.	PUNCT
ejpam-1676	671	1	ann	ann	PROPN
ejpam-1676	671	2	.	.	PROPN
ejpam-1676	671	3	inst	inst	PROPN
ejpam-1676	671	4	.	.	PUNCT
ejpam-1676	672	1	statist	statist	PROPN
ejpam-1676	672	2	.	.	PUNCT
ejpam-1676	673	1	math	math	NOUN
ejpam-1676	673	2	.	.	PUNCT
ejpam-1676	673	3	,	,	PUNCT
ejpam-1676	673	4	57:687	57:687	NUM
ejpam-1676	673	5	–	–	PUNCT
ejpam-1676	673	6	701	701	NUM
ejpam-1676	673	7	,	,	PUNCT
ejpam-1676	673	8	2005	2005	NUM
ejpam-1676	673	9	.	.	PUNCT
ejpam-1676	674	1	[	[	X
ejpam-1676	674	2	7	7	NUM
ejpam-1676	674	3	]	]	PUNCT
ejpam-1676	674	4	x	x	PART
ejpam-1676	674	5	wang	wang	PROPN
ejpam-1676	674	6	and	and	CCONJ
ejpam-1676	674	7	j	j	PROPN
ejpam-1676	674	8	zidek	zidek	PROPN
ejpam-1676	674	9	.	.	PUNCT
ejpam-1676	675	1	selecting	select	VERB
ejpam-1676	675	2	likelihood	likelihood	NOUN
ejpam-1676	675	3	weights	weight	NOUN
ejpam-1676	675	4	by	by	ADP
ejpam-1676	675	5	cross	cross	NOUN
ejpam-1676	675	6	-	-	NOUN
ejpam-1676	675	7	validation	validation	NOUN
ejpam-1676	675	8	.	.	PUNCT
ejpam-1676	676	1	ann	ann	PROPN
ejpam-1676	676	2	.	.	PUNCT
ejpam-1676	676	3	statist	statist	PROPN
ejpam-1676	676	4	.	.	PROPN
ejpam-1676	676	5	,	,	PUNCT
ejpam-1676	676	6	33:463–500	33:463–500	PROPN
ejpam-1676	676	7	,	,	PUNCT
ejpam-1676	676	8	2005	2005	NUM
ejpam-1676	676	9	.	.	PUNCT
ejpam-1676	677	1	[	[	X
ejpam-1676	677	2	8	8	NUM
ejpam-1676	677	3	]	]	X
ejpam-1676	677	4	d.	d.	PROPN
ejpam-1676	677	5	hosmer	hosmer	PROPN
ejpam-1676	677	6	and	and	CCONJ
ejpam-1676	677	7	s.	s.	PROPN
ejpam-1676	677	8	lemeshow	lemeshow	PROPN
ejpam-1676	677	9	.	.	PUNCT
ejpam-1676	678	1	applied	apply	VERB
ejpam-1676	678	2	logistic	logistic	ADJ
ejpam-1676	678	3	regression	regression	NOUN
ejpam-1676	678	4	.	.	PUNCT
ejpam-1676	679	1	wiley	wiley	PROPN
ejpam-1676	679	2	,	,	PUNCT
ejpam-1676	679	3	new	new	PROPN
ejpam-1676	679	4	york	york	PROPN
ejpam-1676	679	5	,	,	PUNCT
ejpam-1676	679	6	1989	1989	NUM
ejpam-1676	679	7	.	.	PUNCT
ejpam-1676	680	1	[	[	X
ejpam-1676	680	2	9	9	NUM
ejpam-1676	680	3	]	]	X
ejpam-1676	680	4	f.	f.	PROPN
ejpam-1676	680	5	hu	hu	PROPN
ejpam-1676	680	6	.	.	PUNCT
ejpam-1676	681	1	the	the	DET
ejpam-1676	681	2	asymptotic	asymptotic	ADJ
ejpam-1676	681	3	properties	property	NOUN
ejpam-1676	681	4	of	of	ADP
ejpam-1676	681	5	the	the	DET
ejpam-1676	681	6	maximum	maximum	ADJ
ejpam-1676	681	7	-	-	PUNCT
ejpam-1676	681	8	relevance	relevance	NOUN
ejpam-1676	681	9	weighted	weight	VERB
ejpam-1676	681	10	likelihood	likelihood	NOUN
ejpam-1676	681	11	estimators	estimator	NOUN
ejpam-1676	681	12	.	.	PUNCT
ejpam-1676	682	1	can	can	AUX
ejpam-1676	682	2	.	.	PUNCT
ejpam-1676	683	1	j.	j.	PROPN
ejpam-1676	683	2	statist	statist	PROPN
ejpam-1676	683	3	.	.	PROPN
ejpam-1676	683	4	,	,	PUNCT
ejpam-1676	683	5	25:45–59	25:45–59	NUM
ejpam-1676	683	6	,	,	PUNCT
ejpam-1676	683	7	1997	1997	NUM
ejpam-1676	683	8	.	.	PUNCT
ejpam-1676	684	1	[	[	X
ejpam-1676	684	2	10	10	NUM
ejpam-1676	684	3	]	]	X
ejpam-1676	684	4	f.	f.	PROPN
ejpam-1676	684	5	hu	hu	PROPN
ejpam-1676	684	6	and	and	CCONJ
ejpam-1676	684	7	j.	j.	PROPN
ejpam-1676	684	8	zidek	zidek	PROPN
ejpam-1676	684	9	.	.	PUNCT
ejpam-1676	685	1	the	the	DET
ejpam-1676	685	2	relevance	relevance	NOUN
ejpam-1676	685	3	weighted	weight	VERB
ejpam-1676	685	4	likelihood	likelihood	NOUN
ejpam-1676	685	5	with	with	ADP
ejpam-1676	685	6	applications	application	NOUN
ejpam-1676	685	7	.	.	PUNCT
ejpam-1676	686	1	in	in	ADP
ejpam-1676	686	2	ahmed	ahmed	PROPN
ejpam-1676	686	3	,	,	PUNCT
ejpam-1676	686	4	s.e	s.e	PROPN
ejpam-1676	686	5	.	.	PROPN
ejpam-1676	686	6	&	&	CCONJ
ejpam-1676	686	7	reid	reid	PROPN
ejpam-1676	686	8	,	,	PUNCT
ejpam-1676	686	9	n.	n.	NOUN
ejpam-1676	686	10	,	,	PUNCT
ejpam-1676	686	11	editor	editor	NOUN
ejpam-1676	686	12	,	,	PUNCT
ejpam-1676	686	13	empirical	empirical	ADJ
ejpam-1676	686	14	bayes	bayes	NOUN
ejpam-1676	686	15	and	and	CCONJ
ejpam-1676	686	16	likelihood	likelihood	NOUN
ejpam-1676	686	17	inference	inference	NOUN
ejpam-1676	686	18	.	.	PUNCT
ejpam-1676	686	19	,	,	PUNCT
ejpam-1676	686	20	pages	page	NOUN
ejpam-1676	686	21	211–235	211–235	NUM
ejpam-1676	686	22	,	,	PUNCT
ejpam-1676	686	23	new	new	PROPN
ejpam-1676	686	24	york	york	PROPN
ejpam-1676	686	25	.	.	PROPN
ejpam-1676	686	26	,	,	PUNCT
ejpam-1676	686	27	2001	2001	NUM
ejpam-1676	686	28	.	.	PUNCT
ejpam-1676	687	1	springer	springer	NOUN
ejpam-1676	687	2	verlag	verlag	PROPN
ejpam-1676	687	3	.	.	PUNCT
ejpam-1676	688	1	[	[	X
ejpam-1676	688	2	11	11	NUM
ejpam-1676	688	3	]	]	X
ejpam-1676	688	4	f	f	PROPN
ejpam-1676	688	5	hu	hu	PROPN
ejpam-1676	688	6	and	and	CCONJ
ejpam-1676	688	7	j	j	PROPN
ejpam-1676	688	8	zidek	zidek	PROPN
ejpam-1676	688	9	.	.	PUNCT
ejpam-1676	689	1	the	the	DET
ejpam-1676	689	2	weighted	weighted	ADJ
ejpam-1676	689	3	likelihood	likelihood	NOUN
ejpam-1676	689	4	.	.	PUNCT
ejpam-1676	690	1	can	can	AUX
ejpam-1676	690	2	.	.	PUNCT
ejpam-1676	691	1	j.	j.	PROPN
ejpam-1676	691	2	statist	statist	PROPN
ejpam-1676	691	3	.	.	PROPN
ejpam-1676	691	4	,	,	PUNCT
ejpam-1676	691	5	25:347–71	25:347–71	NOUN
ejpam-1676	691	6	,	,	PUNCT
ejpam-1676	691	7	2002	2002	NUM
ejpam-1676	691	8	.	.	PUNCT
ejpam-1676	692	1	[	[	X
ejpam-1676	692	2	12	12	NUM
ejpam-1676	692	3	]	]	X
ejpam-1676	692	4	f.	f.	PROPN
ejpam-1676	692	5	hu	hu	PROPN
ejpam-1676	692	6	and	and	CCONJ
ejpam-1676	692	7	w.	w.	PROPN
ejpam-1676	692	8	rosenberger	rosenberger	PROPN
ejpam-1676	692	9	.	.	PUNCT
ejpam-1676	693	1	analysis	analysis	NOUN
ejpam-1676	693	2	of	of	ADP
ejpam-1676	693	3	time	time	NOUN
ejpam-1676	693	4	trends	trend	NOUN
ejpam-1676	693	5	in	in	ADP
ejpam-1676	693	6	adaptive	adaptive	ADJ
ejpam-1676	693	7	designs	design	NOUN
ejpam-1676	693	8	with	with	ADP
ejpam-1676	693	9	applications	application	NOUN
ejpam-1676	693	10	to	to	ADP
ejpam-1676	693	11	a	a	DET
ejpam-1676	693	12	neurophysiology	neurophysiology	NOUN
ejpam-1676	693	13	experiments	experiment	NOUN
ejpam-1676	693	14	.	.	PUNCT
ejpam-1676	694	1	statist	statist	NOUN
ejpam-1676	694	2	.	.	PUNCT
ejpam-1676	695	1	med	med	PROPN
ejpam-1676	695	2	.	.	PROPN
ejpam-1676	695	3	,	,	PUNCT
ejpam-1676	695	4	19:2067–75	19:2067–75	NUM
ejpam-1676	695	5	,	,	PUNCT
ejpam-1676	695	6	2000	2000	NUM
ejpam-1676	695	7	.	.	PUNCT
ejpam-1676	696	1	[	[	X
ejpam-1676	696	2	13	13	NUM
ejpam-1676	696	3	]	]	X
ejpam-1676	696	4	h.	h.	NOUN
ejpam-1676	696	5	akaike	akaike	PROPN
ejpam-1676	696	6	.	.	PUNCT
ejpam-1676	697	1	information	information	NOUN
ejpam-1676	697	2	theory	theory	NOUN
ejpam-1676	697	3	and	and	CCONJ
ejpam-1676	697	4	an	an	DET
ejpam-1676	697	5	extension	extension	NOUN
ejpam-1676	697	6	of	of	ADP
ejpam-1676	697	7	the	the	DET
ejpam-1676	697	8	maximum	maximum	ADJ
ejpam-1676	697	9	likelihood	likelihood	NOUN
ejpam-1676	697	10	principle	principle	NOUN
ejpam-1676	697	11	.	.	PUNCT
ejpam-1676	698	1	in	in	ADP
ejpam-1676	698	2	petrov	petrov	PROPN
ejpam-1676	698	3	,	,	PUNCT
ejpam-1676	698	4	b.n	b.n	PROPN
ejpam-1676	698	5	.	.	PROPN
ejpam-1676	698	6	&	&	CCONJ
ejpam-1676	698	7	csaki	csaki	PROPN
ejpam-1676	698	8	,	,	PUNCT
ejpam-1676	698	9	f.	f.	PROPN
ejpam-1676	698	10	,	,	PUNCT
ejpam-1676	698	11	editor	editor	NOUN
ejpam-1676	698	12	,	,	PUNCT
ejpam-1676	698	13	proc	proc	NOUN
ejpam-1676	698	14	.	.	PUNCT
ejpam-1676	699	1	2nd	2nd	ADJ
ejpam-1676	699	2	international	international	ADJ
ejpam-1676	699	3	symposium	symposium	NOUN
ejpam-1676	699	4	on	on	ADP
ejpam-1676	699	5	information	information	NOUN
ejpam-1676	699	6	theory	theory	NOUN
ejpam-1676	699	7	.	.	PUNCT
ejpam-1676	700	1	,	,	PUNCT
ejpam-1676	700	2	pages	page	NOUN
ejpam-1676	700	3	267–281	267–281	NUM
ejpam-1676	700	4	,	,	PUNCT
ejpam-1676	700	5	budapest	budapest	NOUN
ejpam-1676	700	6	.	.	PUNCT
ejpam-1676	700	7	,	,	PUNCT
ejpam-1676	700	8	1973	1973	NUM
ejpam-1676	700	9	.	.	PUNCT
ejpam-1676	701	1	akademiai	akademiai	PROPN
ejpam-1676	701	2	kiado	kiado	PROPN
ejpam-1676	701	3	.	.	PUNCT
ejpam-1676	702	1	[	[	X
ejpam-1676	702	2	14	14	NUM
ejpam-1676	702	3	]	]	X
ejpam-1676	702	4	i.	i.	PROPN
ejpam-1676	702	5	csiszár	csiszár	PROPN
ejpam-1676	702	6	.	.	PUNCT
ejpam-1676	703	1	i	i	PRON
ejpam-1676	703	2	-	-	PUNCT
ejpam-1676	703	3	divergence	divergence	NOUN
ejpam-1676	703	4	geometry	geometry	NOUN
ejpam-1676	703	5	of	of	ADP
ejpam-1676	703	6	probability	probability	NOUN
ejpam-1676	703	7	distributions	distribution	NOUN
ejpam-1676	703	8	and	and	CCONJ
ejpam-1676	703	9	minimization	minimization	NOUN
ejpam-1676	703	10	problems	problem	NOUN
ejpam-1676	703	11	.	.	PUNCT
ejpam-1676	703	12	.	.	PUNCT
ejpam-1676	704	1	ann	ann	PROPN
ejpam-1676	704	2	.	.	PROPN
ejpam-1676	704	3	statist	statist	PROPN
ejpam-1676	704	4	.	.	PUNCT
ejpam-1676	704	5	,	,	PUNCT
ejpam-1676	704	6	3:146–158	3:146–158	NUM
ejpam-1676	704	7	,	,	PUNCT
ejpam-1676	704	8	1975	1975	NUM
ejpam-1676	704	9	.	.	PUNCT
ejpam-1676	705	1	[	[	X
ejpam-1676	705	2	15	15	NUM
ejpam-1676	705	3	]	]	X
ejpam-1676	705	4	j.	j.	PROPN
ejpam-1676	705	5	bernardo	bernardo	PROPN
ejpam-1676	705	6	.	.	PUNCT
ejpam-1676	706	1	a	a	DET
ejpam-1676	706	2	maximum	maximum	ADJ
ejpam-1676	706	3	likelihood	likelihood	NOUN
ejpam-1676	706	4	methodology	methodology	NOUN
ejpam-1676	706	5	for	for	ADP
ejpam-1676	706	6	clusterwise	clusterwise	ADJ
ejpam-1676	706	7	linear	linear	PROPN
ejpam-1676	706	8	regression	regression	NOUN
ejpam-1676	706	9	.	.	PUNCT
ejpam-1676	707	1	ann	ann	PROPN
ejpam-1676	707	2	.	.	PUNCT
ejpam-1676	707	3	statist	statist	PROPN
ejpam-1676	707	4	.	.	PUNCT
ejpam-1676	707	5	,	,	PUNCT
ejpam-1676	707	6	7:686–690	7:686–690	PROPN
ejpam-1676	707	7	,	,	PUNCT
ejpam-1676	707	8	1979	1979	NUM
ejpam-1676	707	9	.	.	PUNCT
ejpam-1676	708	1	[	[	X
ejpam-1676	708	2	16	16	NUM
ejpam-1676	708	3	]	]	X
ejpam-1676	708	4	h	h	NOUN
ejpam-1676	708	5	royden	royden	VERB
ejpam-1676	708	6	.	.	PUNCT
ejpam-1676	709	1	real	real	ADJ
ejpam-1676	709	2	analysis	analysis	NOUN
ejpam-1676	709	3	.	.	PUNCT
ejpam-1676	710	1	prentice	prentice	PROPN
ejpam-1676	710	2	hall	hall	PROPN
ejpam-1676	710	3	,	,	PUNCT
ejpam-1676	710	4	new	new	PROPN
ejpam-1676	710	5	york	york	PROPN
ejpam-1676	710	6	,	,	PUNCT
ejpam-1676	710	7	3	3	NUM
ejpam-1676	710	8	edition	edition	NOUN
ejpam-1676	710	9	,	,	PUNCT
ejpam-1676	710	10	1988	1988	NUM
ejpam-1676	710	11	.	.	PUNCT
ejpam-1676	711	1	[	[	X
ejpam-1676	711	2	17	17	NUM
ejpam-1676	711	3	]	]	X
ejpam-1676	711	4	s.	s.	PROPN
ejpam-1676	711	5	eguchi	eguchi	PROPN
ejpam-1676	711	6	and	and	CCONJ
ejpam-1676	711	7	j.	j.	PROPN
ejpam-1676	711	8	copas	copas	PROPN
ejpam-1676	711	9	.	.	PUNCT
ejpam-1676	712	1	a	a	DET
ejpam-1676	712	2	class	class	NOUN
ejpam-1676	712	3	of	of	ADP
ejpam-1676	712	4	local	local	ADJ
ejpam-1676	712	5	likelihood	likelihood	NOUN
ejpam-1676	712	6	methods	method	NOUN
ejpam-1676	712	7	and	and	CCONJ
ejpam-1676	712	8	near	near	ADJ
ejpam-1676	712	9	-	-	PUNCT
ejpam-1676	712	10	parametric	parametric	ADJ
ejpam-1676	712	11	asymptotics	asymptotic	NOUN
ejpam-1676	712	12	.	.	PUNCT
ejpam-1676	713	1	j.	j.	PROPN
ejpam-1676	713	2	r.	r.	PROPN
ejpam-1676	713	3	statist	statist	PROPN
ejpam-1676	713	4	.	.	PUNCT
ejpam-1676	714	1	soc	soc	PROPN
ejpam-1676	714	2	.	.	PUNCT
ejpam-1676	715	1	b	b	X
ejpam-1676	715	2	,	,	PUNCT
ejpam-1676	715	3	60:709–24	60:709–24	PROPN
ejpam-1676	715	4	,	,	PUNCT
ejpam-1676	715	5	1998	1998	NUM
ejpam-1676	715	6	.	.	PUNCT
ejpam-1676	716	1	[	[	X
ejpam-1676	716	2	18	18	NUM
ejpam-1676	716	3	]	]	PUNCT
ejpam-1676	716	4	s	s	VERB
ejpam-1676	716	5	kullback	kullback	NOUN
ejpam-1676	716	6	.	.	PUNCT
ejpam-1676	717	1	information	information	NOUN
ejpam-1676	717	2	theory	theory	NOUN
ejpam-1676	717	3	and	and	CCONJ
ejpam-1676	717	4	statistics	statistic	NOUN
ejpam-1676	717	5	.	.	PUNCT
ejpam-1676	718	1	wiley	wiley	PROPN
ejpam-1676	718	2	,	,	PUNCT
ejpam-1676	718	3	new	new	PROPN
ejpam-1676	718	4	york	york	PROPN
ejpam-1676	718	5	,	,	PUNCT
ejpam-1676	718	6	1959	1959	NUM
ejpam-1676	718	7	.	.	PUNCT
ejpam-1676	719	1	[	[	X
ejpam-1676	719	2	19	19	NUM
ejpam-1676	719	3	]	]	X
ejpam-1676	719	4	t.	t.	NOUN
ejpam-1676	719	5	cover	cover	NOUN
ejpam-1676	719	6	and	and	CCONJ
ejpam-1676	719	7	j.	j.	PROPN
ejpam-1676	719	8	thomas	thomas	PROPN
ejpam-1676	719	9	.	.	PUNCT
ejpam-1676	720	1	elements	element	NOUN
ejpam-1676	720	2	of	of	ADP
ejpam-1676	720	3	information	information	NOUN
ejpam-1676	720	4	theory	theory	NOUN
ejpam-1676	720	5	.	.	PUNCT
ejpam-1676	721	1	wiley	wiley	PROPN
ejpam-1676	721	2	,	,	PUNCT
ejpam-1676	721	3	new	new	PROPN
ejpam-1676	721	4	york	york	PROPN
ejpam-1676	721	5	,	,	PUNCT
ejpam-1676	721	6	1991	1991	NUM
ejpam-1676	721	7	.	.	PUNCT
ejpam-1676	722	1	[	[	X
ejpam-1676	722	2	20	20	NUM
ejpam-1676	722	3	]	]	X
ejpam-1676	722	4	w	w	PROPN
ejpam-1676	722	5	james	james	PROPN
ejpam-1676	722	6	and	and	CCONJ
ejpam-1676	722	7	c	c	PROPN
ejpam-1676	722	8	stein	stein	PROPN
ejpam-1676	722	9	.	.	PUNCT
ejpam-1676	723	1	estimation	estimation	NOUN
ejpam-1676	723	2	with	with	ADP
ejpam-1676	723	3	quadratic	quadratic	ADJ
ejpam-1676	723	4	loss	loss	NOUN
ejpam-1676	723	5	.	.	PUNCT
ejpam-1676	724	1	in	in	ADP
ejpam-1676	724	2	proc	proc	PROPN
ejpam-1676	724	3	4th	4th	PROPN
ejpam-1676	724	4	berkeley	berkeley	PROPN
ejpam-1676	724	5	symp	symp	PROPN
ejpam-1676	724	6	.	.	PUNCT
ejpam-1676	725	1	math	math	PROPN
ejpam-1676	725	2	statist	statist	PROPN
ejpam-1676	725	3	prob	prob	PROPN
ejpam-1676	725	4	.	.	PROPN
ejpam-1676	725	5	,	,	PUNCT
ejpam-1676	725	6	volume	volume	NOUN
ejpam-1676	725	7	1	1	NUM
ejpam-1676	725	8	,	,	PUNCT
ejpam-1676	725	9	pages	page	VERB
ejpam-1676	725	10	361–379	361–379	NUM
ejpam-1676	725	11	.	.	PUNCT
ejpam-1676	726	1	berkely	berkely	PROPN
ejpam-1676	726	2	:	:	PUNCT
ejpam-1676	726	3	university	university	NOUN
ejpam-1676	726	4	of	of	ADP
ejpam-1676	726	5	california	california	PROPN
ejpam-1676	726	6	press	press	PROPN
ejpam-1676	726	7	.	.	PUNCT
ejpam-1676	726	8	,	,	PUNCT
ejpam-1676	726	9	1961	1961	NUM
ejpam-1676	726	10	.	.	PUNCT
ejpam-1676	727	1	[	[	X
ejpam-1676	727	2	21	21	NUM
ejpam-1676	727	3	]	]	X
ejpam-1676	727	4	x	x	X
ejpam-1676	727	5	wang	wang	PROPN
ejpam-1676	727	6	,	,	PUNCT
ejpam-1676	727	7	c	c	PROPN
ejpam-1676	727	8	van	van	PROPN
ejpam-1676	727	9	emden	emden	PROPN
ejpam-1676	727	10	and	and	CCONJ
ejpam-1676	727	11	j	j	PROPN
ejpam-1676	727	12	zidek	zidek	PROPN
ejpam-1676	727	13	.	.	PUNCT
ejpam-1676	728	1	asymptotic	asymptotic	ADJ
ejpam-1676	728	2	properties	property	NOUN
ejpam-1676	728	3	of	of	ADP
ejpam-1676	728	4	maximum	maximum	ADJ
ejpam-1676	728	5	weighted	weight	VERB
ejpam-1676	728	6	likelihood	likelihood	NOUN
ejpam-1676	728	7	estimator	estimator	NOUN
ejpam-1676	728	8	.	.	PUNCT
ejpam-1676	729	1	j.	j.	PROPN
ejpam-1676	729	2	statist	statist	PROPN
ejpam-1676	729	3	.	.	PUNCT
ejpam-1676	730	1	plan	plan	NOUN
ejpam-1676	730	2	.	.	PUNCT
ejpam-1676	731	1	infer	infer	VERB
ejpam-1676	731	2	.	.	PUNCT
ejpam-1676	731	3	,	,	PUNCT
ejpam-1676	731	4	119:37–54	119:37–54	NUM
ejpam-1676	731	5	,	,	PUNCT
ejpam-1676	731	6	2004	2004	NUM
ejpam-1676	731	7	.	.	PUNCT