id	sid	tid	token	lemma	pos
cuesj-105	1	1	14	14	NUM
cuesj-105	1	2	http://journals.cihanuniversity.edu.iq/index.php/cuesj	http://journals.cihanuniversity.edu.iq/index.php/cuesj	ADJ
cuesj-105	1	3	cuesj	cuesj	NOUN
cuesj-105	1	4	2019	2019	NUM
cuesj-105	1	5	,	,	PUNCT
cuesj-105	1	6	3	3	NUM
cuesj-105	1	7	(	(	PUNCT
cuesj-105	1	8	2	2	NUM
cuesj-105	1	9	):	):	PUNCT
cuesj-105	1	10	14	14	NUM
cuesj-105	1	11	-	-	SYM
cuesj-105	1	12	20	20	NUM
cuesj-105	1	13	research	research	NOUN
cuesj-105	1	14	article	article	NOUN
cuesj-105	1	15	a	a	DET
cuesj-105	1	16	face	face	NOUN
cuesj-105	1	17	recognition	recognition	NOUN
cuesj-105	1	18	system	system	NOUN
cuesj-105	1	19	based	base	VERB
cuesj-105	1	20	on	on	ADP
cuesj-105	1	21	principal	principal	ADJ
cuesj-105	1	22	component	component	NOUN
cuesj-105	1	23	analysis	analysis	NOUN
cuesj-105	1	24	-	-	PUNCT
cuesj-105	1	25	wavelet	wavelet	NOUN
cuesj-105	1	26	and	and	CCONJ
cuesj-105	1	27	support	support	NOUN
cuesj-105	1	28	vector	vector	NOUN
cuesj-105	1	29	machines	machine	NOUN
cuesj-105	1	30	laith	laith	PROPN
cuesj-105	1	31	r.	r.	PROPN
cuesj-105	1	32	fleah1	fleah1	PROPN
cuesj-105	1	33	*	*	PROPN
cuesj-105	1	34	,	,	PUNCT
cuesj-105	1	35	shaimaa	shaimaa	PROPN
cuesj-105	1	36	a.	a.	PROPN
cuesj-105	1	37	al	al	PROPN
cuesj-105	1	38	-	-	PUNCT
cuesj-105	1	39	aubi2	aubi2	PROPN
cuesj-105	1	40	1department	1department	NUM
cuesj-105	1	41	of	of	ADP
cuesj-105	1	42	computer	computer	NOUN
cuesj-105	1	43	science	science	NOUN
cuesj-105	1	44	,	,	PUNCT
cuesj-105	1	45	cihan	cihan	VERB
cuesj-105	1	46	university	university	NOUN
cuesj-105	1	47	-	-	PUNCT
cuesj-105	1	48	erbil	erbil	PROPN
cuesj-105	1	49	,	,	PUNCT
cuesj-105	1	50	iraq	iraq	PROPN
cuesj-105	1	51	,	,	PUNCT
cuesj-105	1	52	2department	2department	NUM
cuesj-105	1	53	of	of	ADP
cuesj-105	1	54	computer	computer	NOUN
cuesj-105	1	55	science	science	PROPN
cuesj-105	1	56	department	department	PROPN
cuesj-105	1	57	,	,	PUNCT
cuesj-105	1	58	college	college	NOUN
cuesj-105	1	59	of	of	ADP
cuesj-105	1	60	science	science	NOUN
cuesj-105	1	61	,	,	PUNCT
cuesj-105	1	62	salahaddin	salahaddin	VERB
cuesj-105	1	63	university	university	NOUN
cuesj-105	1	64	,	,	PUNCT
cuesj-105	1	65	erbil	erbil	PROPN
cuesj-105	1	66	,	,	PUNCT
cuesj-105	1	67	iraq	iraq	PROPN
cuesj-105	1	68	abstract	abstract	ADJ
cuesj-105	1	69	face	face	NOUN
cuesj-105	1	70	recognition	recognition	NOUN
cuesj-105	1	71	can	can	AUX
cuesj-105	1	72	represent	represent	VERB
cuesj-105	1	73	a	a	DET
cuesj-105	1	74	key	key	ADJ
cuesj-105	1	75	requirement	requirement	NOUN
cuesj-105	1	76	in	in	ADP
cuesj-105	1	77	various	various	ADJ
cuesj-105	1	78	types	type	NOUN
cuesj-105	1	79	of	of	ADP
cuesj-105	1	80	applications	application	NOUN
cuesj-105	1	81	such	such	ADJ
cuesj-105	1	82	as	as	ADP
cuesj-105	1	83	human	human	ADJ
cuesj-105	1	84	-	-	PUNCT
cuesj-105	1	85	computer	computer	NOUN
cuesj-105	1	86	interface	interface	NOUN
cuesj-105	1	87	,	,	PUNCT
cuesj-105	1	88	monitoring	monitoring	NOUN
cuesj-105	1	89	systems	system	NOUN
cuesj-105	1	90	,	,	PUNCT
cuesj-105	1	91	as	as	ADV
cuesj-105	1	92	well	well	ADV
cuesj-105	1	93	as	as	ADP
cuesj-105	1	94	personal	personal	ADJ
cuesj-105	1	95	identification	identification	NOUN
cuesj-105	1	96	.	.	PUNCT
cuesj-105	2	1	in	in	ADP
cuesj-105	2	2	this	this	DET
cuesj-105	2	3	paper	paper	NOUN
cuesj-105	2	4	,	,	PUNCT
cuesj-105	2	5	design	design	NOUN
cuesj-105	2	6	and	and	CCONJ
cuesj-105	2	7	implement	implement	NOUN
cuesj-105	2	8	of	of	ADP
cuesj-105	2	9	face	face	NOUN
cuesj-105	2	10	recognition	recognition	NOUN
cuesj-105	2	11	system	system	NOUN
cuesj-105	2	12	are	be	AUX
cuesj-105	2	13	introduced	introduce	VERB
cuesj-105	2	14	.	.	PUNCT
cuesj-105	3	1	in	in	ADP
cuesj-105	3	2	this	this	DET
cuesj-105	3	3	system	system	NOUN
cuesj-105	3	4	,	,	PUNCT
cuesj-105	3	5	a	a	DET
cuesj-105	3	6	combination	combination	NOUN
cuesj-105	3	7	of	of	ADP
cuesj-105	3	8	principal	principal	ADJ
cuesj-105	3	9	component	component	NOUN
cuesj-105	3	10	analysis	analysis	NOUN
cuesj-105	3	11	(	(	PUNCT
cuesj-105	3	12	pca	pca	NOUN
cuesj-105	3	13	)	)	PUNCT
cuesj-105	3	14	and	and	CCONJ
cuesj-105	3	15	wavelet	wavelet	NOUN
cuesj-105	3	16	feature	feature	NOUN
cuesj-105	3	17	extraction	extraction	NOUN
cuesj-105	3	18	algorithms	algorithm	NOUN
cuesj-105	3	19	with	with	ADP
cuesj-105	3	20	support	support	NOUN
cuesj-105	3	21	vector	vector	NOUN
cuesj-105	3	22	machine	machine	NOUN
cuesj-105	3	23	(	(	PUNCT
cuesj-105	3	24	svm	svm	PROPN
cuesj-105	3	25	)	)	PUNCT
cuesj-105	3	26	and	and	CCONJ
cuesj-105	3	27	k	k	NOUN
cuesj-105	3	28	-	-	PUNCT
cuesj-105	3	29	nearest	near	ADJ
cuesj-105	3	30	neighborhood	neighborhood	NOUN
cuesj-105	3	31	classifier	classifier	NOUN
cuesj-105	3	32	is	be	AUX
cuesj-105	3	33	used	use	VERB
cuesj-105	3	34	.	.	PUNCT
cuesj-105	4	1	pca	pca	PROPN
cuesj-105	4	2	and	and	CCONJ
cuesj-105	4	3	wavelet	wavelet	NOUN
cuesj-105	4	4	transform	transform	NOUN
cuesj-105	4	5	methods	method	NOUN
cuesj-105	4	6	are	be	AUX
cuesj-105	4	7	used	use	VERB
cuesj-105	4	8	to	to	PART
cuesj-105	4	9	extract	extract	VERB
cuesj-105	4	10	features	feature	NOUN
cuesj-105	4	11	from	from	ADP
cuesj-105	4	12	face	face	NOUN
cuesj-105	4	13	image	image	NOUN
cuesj-105	4	14	using	use	VERB
cuesj-105	4	15	and	and	CCONJ
cuesj-105	4	16	identify	identify	VERB
cuesj-105	4	17	the	the	DET
cuesj-105	4	18	image	image	NOUN
cuesj-105	4	19	of	of	ADP
cuesj-105	4	20	the	the	DET
cuesj-105	4	21	face	face	NOUN
cuesj-105	4	22	using	use	VERB
cuesj-105	4	23	svms	svms	NOUN
cuesj-105	4	24	classifier	classifier	NOUN
cuesj-105	4	25	as	as	ADV
cuesj-105	4	26	well	well	ADV
cuesj-105	4	27	as	as	ADP
cuesj-105	4	28	the	the	DET
cuesj-105	4	29	neural	neural	ADJ
cuesj-105	4	30	network	network	NOUN
cuesj-105	4	31	are	be	AUX
cuesj-105	4	32	used	use	VERB
cuesj-105	4	33	as	as	ADP
cuesj-105	4	34	a	a	DET
cuesj-105	4	35	classifier	classifier	NOUN
cuesj-105	4	36	to	to	PART
cuesj-105	4	37	compare	compare	VERB
cuesj-105	4	38	its	its	PRON
cuesj-105	4	39	results	result	NOUN
cuesj-105	4	40	with	with	ADP
cuesj-105	4	41	the	the	DET
cuesj-105	4	42	proposed	propose	VERB
cuesj-105	4	43	system	system	NOUN
cuesj-105	4	44	.	.	PUNCT
cuesj-105	5	1	for	for	ADP
cuesj-105	5	2	a	a	DET
cuesj-105	5	3	more	more	ADV
cuesj-105	5	4	comprehensive	comprehensive	ADJ
cuesj-105	5	5	comparison	comparison	NOUN
cuesj-105	5	6	,	,	PUNCT
cuesj-105	5	7	two	two	NUM
cuesj-105	5	8	face	face	NOUN
cuesj-105	5	9	image	image	NOUN
cuesj-105	5	10	databases	database	NOUN
cuesj-105	5	11	(	(	PUNCT
cuesj-105	5	12	yale	yale	NOUN
cuesj-105	5	13	and	and	CCONJ
cuesj-105	5	14	orl	orl	PROPN
cuesj-105	5	15	)	)	PUNCT
cuesj-105	5	16	are	be	AUX
cuesj-105	5	17	used	use	VERB
cuesj-105	5	18	to	to	PART
cuesj-105	5	19	test	test	VERB
cuesj-105	5	20	the	the	DET
cuesj-105	5	21	performance	performance	NOUN
cuesj-105	5	22	of	of	ADP
cuesj-105	5	23	the	the	DET
cuesj-105	5	24	system	system	NOUN
cuesj-105	5	25	.	.	PUNCT
cuesj-105	6	1	finally	finally	ADV
cuesj-105	6	2	,	,	PUNCT
cuesj-105	6	3	the	the	DET
cuesj-105	6	4	experimental	experimental	ADJ
cuesj-105	6	5	results	result	NOUN
cuesj-105	6	6	show	show	VERB
cuesj-105	6	7	the	the	DET
cuesj-105	6	8	efficiency	efficiency	NOUN
cuesj-105	6	9	and	and	CCONJ
cuesj-105	6	10	reliability	reliability	NOUN
cuesj-105	6	11	of	of	ADP
cuesj-105	6	12	face	face	NOUN
cuesj-105	6	13	recognition	recognition	NOUN
cuesj-105	6	14	system	system	NOUN
cuesj-105	6	15	,	,	PUNCT
cuesj-105	6	16	and	and	CCONJ
cuesj-105	6	17	the	the	DET
cuesj-105	6	18	results	result	NOUN
cuesj-105	6	19	demonstrate	demonstrate	VERB
cuesj-105	6	20	accuracy	accuracy	NOUN
cuesj-105	6	21	on	on	ADP
cuesj-105	6	22	two	two	NUM
cuesj-105	6	23	databases	database	NOUN
cuesj-105	6	24	indicated	indicate	VERB
cuesj-105	6	25	that	that	SCONJ
cuesj-105	6	26	the	the	DET
cuesj-105	6	27	results	result	NOUN
cuesj-105	6	28	enhancement	enhancement	VERB
cuesj-105	6	29	5	5	NUM
cuesj-105	6	30	%	%	NOUN
cuesj-105	6	31	using	use	VERB
cuesj-105	6	32	the	the	DET
cuesj-105	6	33	svm	svm	ADJ
cuesj-105	6	34	classifier	classifier	NOUN
cuesj-105	6	35	with	with	ADP
cuesj-105	6	36	polynomial	polynomial	ADJ
cuesj-105	6	37	kernel	kernel	NOUN
cuesj-105	6	38	function	function	NOUN
cuesj-105	6	39	compared	compare	VERB
cuesj-105	6	40	to	to	PART
cuesj-105	6	41	use	use	VERB
cuesj-105	6	42	feedforward	feedforward	ADJ
cuesj-105	6	43	neural	neural	ADJ
cuesj-105	6	44	network	network	NOUN
cuesj-105	6	45	classifier	classifier	NOUN
cuesj-105	6	46	.	.	PUNCT
cuesj-105	7	1	keywords	keyword	NOUN
cuesj-105	7	2	:	:	PUNCT
cuesj-105	7	3	face	face	NOUN
cuesj-105	7	4	recognition	recognition	NOUN
cuesj-105	7	5	,	,	PUNCT
cuesj-105	7	6	feedforward	feedforward	NOUN
cuesj-105	7	7	backpropagation	backpropagation	NOUN
cuesj-105	7	8	,	,	PUNCT
cuesj-105	7	9	neural	neural	ADJ
cuesj-105	7	10	network	network	NOUN
cuesj-105	7	11	and	and	CCONJ
cuesj-105	7	12	k	k	NOUN
cuesj-105	7	13	-	-	PUNCT
cuesj-105	7	14	nearest	near	ADJ
cuesj-105	7	15	neighborhood	neighborhood	NOUN
cuesj-105	7	16	,	,	PUNCT
cuesj-105	7	17	support	support	VERB
cuesj-105	7	18	vector	vector	NOUN
cuesj-105	7	19	machines	machine	NOUN
cuesj-105	7	20	introduction	introduction	NOUN
cuesj-105	7	21	facial	facial	ADJ
cuesj-105	7	22	recognition	recognition	NOUN
cuesj-105	7	23	is	be	AUX
cuesj-105	7	24	considered	consider	VERB
cuesj-105	7	25	as	as	ADP
cuesj-105	7	26	one	one	NUM
cuesj-105	7	27	of	of	ADP
cuesj-105	7	28	the	the	DET
cuesj-105	7	29	biometric	biometric	ADJ
cuesj-105	7	30	systems	system	NOUN
cuesj-105	7	31	and	and	CCONJ
cuesj-105	7	32	the	the	DET
cuesj-105	7	33	importance	importance	NOUN
cuesj-105	7	34	of	of	ADP
cuesj-105	7	35	this	this	DET
cuesj-105	7	36	field	field	NOUN
cuesj-105	7	37	as	as	ADP
cuesj-105	7	38	a	a	DET
cuesj-105	7	39	topic	topic	NOUN
cuesj-105	7	40	of	of	ADP
cuesj-105	7	41	research	research	NOUN
cuesj-105	7	42	has	have	AUX
cuesj-105	7	43	increased	increase	VERB
cuesj-105	7	44	in	in	ADP
cuesj-105	7	45	recent	recent	ADJ
cuesj-105	7	46	years	year	NOUN
cuesj-105	7	47	.	.	PUNCT
cuesj-105	8	1	face	face	NOUN
cuesj-105	8	2	recognition	recognition	NOUN
cuesj-105	8	3	system	system	NOUN
cuesj-105	8	4	uses	use	VERB
cuesj-105	8	5	biometric	biometric	ADJ
cuesj-105	8	6	information	information	NOUN
cuesj-105	8	7	which	which	PRON
cuesj-105	8	8	takes	take	VERB
cuesj-105	8	9	from	from	ADP
cuesj-105	8	10	human	human	NOUN
cuesj-105	8	11	,	,	PUNCT
cuesj-105	8	12	but	but	CCONJ
cuesj-105	8	13	the	the	DET
cuesj-105	8	14	use	use	NOUN
cuesj-105	8	15	of	of	ADP
cuesj-105	8	16	this	this	DET
cuesj-105	8	17	system	system	NOUN
cuesj-105	8	18	is	be	AUX
cuesj-105	8	19	more	more	ADV
cuesj-105	8	20	easily	easily	ADV
cuesj-105	8	21	than	than	ADP
cuesj-105	8	22	in	in	ADP
cuesj-105	8	23	another	another	DET
cuesj-105	8	24	biometric	biometric	ADJ
cuesj-105	8	25	system	system	NOUN
cuesj-105	8	26	such	such	ADJ
cuesj-105	8	27	as	as	ADP
cuesj-105	8	28	fingerprints	fingerprint	NOUN
cuesj-105	8	29	,	,	PUNCT
cuesj-105	8	30	iris	iris	NOUN
cuesj-105	8	31	,	,	PUNCT
cuesj-105	8	32	and	and	CCONJ
cuesj-105	8	33	signature	signature	NOUN
cuesj-105	8	34	,	,	PUNCT
cuesj-105	8	35	especially	especially	ADV
cuesj-105	8	36	with	with	ADP
cuesj-105	8	37	the	the	DET
cuesj-105	8	38	uncooperative	uncooperative	ADJ
cuesj-105	8	39	people	people	NOUN
cuesj-105	8	40	.	.	PUNCT
cuesj-105	9	1	face	face	NOUN
cuesj-105	9	2	recognition	recognition	NOUN
cuesj-105	9	3	systems	system	NOUN
cuesj-105	9	4	are	be	AUX
cuesj-105	9	5	usually	usually	ADV
cuesj-105	9	6	used	use	VERB
cuesj-105	9	7	in	in	ADP
cuesj-105	9	8	security	security	NOUN
cuesj-105	9	9	systems	system	NOUN
cuesj-105	9	10	to	to	PART
cuesj-105	9	11	monitor	monitor	VERB
cuesj-105	9	12	people	people	NOUN
cuesj-105	9	13	and	and	CCONJ
cuesj-105	9	14	to	to	PART
cuesj-105	9	15	discover	discover	VERB
cuesj-105	9	16	any	any	DET
cuesj-105	9	17	security	security	NOUN
cuesj-105	9	18	breach	breach	NOUN
cuesj-105	9	19	,	,	PUNCT
cuesj-105	9	20	for	for	ADP
cuesj-105	9	21	example	example	NOUN
cuesj-105	9	22	,	,	PUNCT
cuesj-105	9	23	face	face	NOUN
cuesj-105	9	24	recognition	recognition	NOUN
cuesj-105	9	25	system	system	NOUN
cuesj-105	9	26	uses	use	VERB
cuesj-105	9	27	in	in	ADP
cuesj-105	9	28	the	the	DET
cuesj-105	9	29	airport	airport	NOUN
cuesj-105	9	30	for	for	ADP
cuesj-105	9	31	monitoring	monitoring	NOUN
cuesj-105	9	32	and	and	CCONJ
cuesj-105	9	33	arrests	arrest	NOUN
cuesj-105	9	34	the	the	DET
cuesj-105	9	35	wanted	wanted	NOUN
cuesj-105	9	36	.	.	PUNCT
cuesj-105	10	1	therefore	therefore	ADV
cuesj-105	10	2	,	,	PUNCT
cuesj-105	10	3	face	face	NOUN
cuesj-105	10	4	recognition	recognition	NOUN
cuesj-105	10	5	system	system	NOUN
cuesj-105	10	6	is	be	AUX
cuesj-105	10	7	used	use	VERB
cuesj-105	10	8	in	in	ADP
cuesj-105	10	9	many	many	ADJ
cuesj-105	10	10	security	security	NOUN
cuesj-105	10	11	applications	application	NOUN
cuesj-105	10	12	such	such	ADJ
cuesj-105	10	13	as	as	ADP
cuesj-105	10	14	avoiding	avoid	VERB
cuesj-105	10	15	crime	crime	NOUN
cuesj-105	10	16	,	,	PUNCT
cuesj-105	10	17	monitoring	monitor	VERB
cuesj-105	10	18	through	through	ADP
cuesj-105	10	19	video	video	NOUN
cuesj-105	10	20	,	,	PUNCT
cuesj-105	10	21	and	and	CCONJ
cuesj-105	10	22	also	also	ADV
cuesj-105	10	23	for	for	ADP
cuesj-105	10	24	verification	verification	NOUN
cuesj-105	10	25	of	of	ADP
cuesj-105	10	26	the	the	DET
cuesj-105	10	27	identity	identity	NOUN
cuesj-105	10	28	of	of	ADP
cuesj-105	10	29	persons	person	NOUN
cuesj-105	10	30	.	.	PUNCT
cuesj-105	11	1	principal	principal	ADJ
cuesj-105	11	2	component	component	NOUN
cuesj-105	11	3	analysis	analysis	NOUN
cuesj-105	11	4	(	(	PUNCT
cuesj-105	11	5	pca	pca	NOUN
cuesj-105	11	6	)	)	PUNCT
cuesj-105	11	7	is	be	AUX
cuesj-105	11	8	one	one	NUM
cuesj-105	11	9	of	of	ADP
cuesj-105	11	10	the	the	DET
cuesj-105	11	11	linear	linear	ADJ
cuesj-105	11	12	transformation	transformation	NOUN
cuesj-105	11	13	techniques	technique	NOUN
cuesj-105	11	14	that	that	PRON
cuesj-105	11	15	are	be	AUX
cuesj-105	11	16	used	use	VERB
cuesj-105	11	17	to	to	PART
cuesj-105	11	18	obtain	obtain	VERB
cuesj-105	11	19	features	feature	NOUN
cuesj-105	11	20	from	from	ADP
cuesj-105	11	21	data	datum	NOUN
cuesj-105	11	22	or	or	CCONJ
cuesj-105	11	23	to	to	PART
cuesj-105	11	24	decompress	decompress	VERB
cuesj-105	11	25	the	the	DET
cuesj-105	11	26	information	information	NOUN
cuesj-105	11	27	.	.	PUNCT
cuesj-105	12	1	it	it	PRON
cuesj-105	12	2	is	be	AUX
cuesj-105	12	3	also	also	ADV
cuesj-105	12	4	known	know	VERB
cuesj-105	12	5	as	as	ADP
cuesj-105	12	6	one	one	NUM
cuesj-105	12	7	of	of	ADP
cuesj-105	12	8	the	the	DET
cuesj-105	12	9	famous	famous	ADJ
cuesj-105	12	10	techniques	technique	NOUN
cuesj-105	12	11	that	that	PRON
cuesj-105	12	12	are	be	AUX
cuesj-105	12	13	used	use	VERB
cuesj-105	12	14	to	to	PART
cuesj-105	12	15	take	take	VERB
cuesj-105	12	16	global	global	ADJ
cuesj-105	12	17	structures	structure	NOUN
cuesj-105	12	18	from	from	ADP
cuesj-105	12	19	the	the	DET
cuesj-105	12	20	high	high	ADV
cuesj-105	12	21	-	-	PUNCT
cuesj-105	12	22	dimensional	dimensional	ADJ
cuesj-105	12	23	information	information	NOUN
cuesj-105	12	24	set	set	VERB
cuesj-105	12	25	as	as	ADV
cuesj-105	12	26	well	well	ADV
cuesj-105	12	27	as	as	ADP
cuesj-105	12	28	it	it	PRON
cuesj-105	12	29	is	be	AUX
cuesj-105	12	30	applied	apply	VERB
cuesj-105	12	31	to	to	PART
cuesj-105	12	32	decompose	decompose	VERB
cuesj-105	12	33	dimensionality	dimensionality	NOUN
cuesj-105	12	34	of	of	ADP
cuesj-105	12	35	the	the	DET
cuesj-105	12	36	data	datum	NOUN
cuesj-105	12	37	and	and	CCONJ
cuesj-105	12	38	take	take	VERB
cuesj-105	12	39	unique	unique	ADJ
cuesj-105	12	40	features	feature	NOUN
cuesj-105	12	41	from	from	ADP
cuesj-105	12	42	the	the	DET
cuesj-105	12	43	faces	face	NOUN
cuesj-105	12	44	picture	picture	NOUN
cuesj-105	12	45	.	.	PUNCT
cuesj-105	13	1	this	this	DET
cuesj-105	13	2	technique	technique	NOUN
cuesj-105	13	3	can	can	AUX
cuesj-105	13	4	be	be	AUX
cuesj-105	13	5	used	use	VERB
cuesj-105	13	6	also	also	ADV
cuesj-105	13	7	to	to	PART
cuesj-105	13	8	distinguish	distinguish	VERB
cuesj-105	13	9	patterns	pattern	NOUN
cuesj-105	13	10	in	in	ADP
cuesj-105	13	11	data	datum	NOUN
cuesj-105	13	12	and	and	CCONJ
cuesj-105	13	13	express	express	VERB
cuesj-105	13	14	the	the	DET
cuesj-105	13	15	data	datum	NOUN
cuesj-105	13	16	so	so	SCONJ
cuesj-105	13	17	as	as	SCONJ
cuesj-105	13	18	to	to	PART
cuesj-105	13	19	highlight	highlight	VERB
cuesj-105	13	20	their	their	PRON
cuesj-105	13	21	differences	difference	NOUN
cuesj-105	13	22	and	and	CCONJ
cuesj-105	13	23	similarities	similarity	NOUN
cuesj-105	13	24	between	between	ADP
cuesj-105	13	25	them.[1	them.[1	NOUN
cuesj-105	13	26	]	]	PUNCT
cuesj-105	13	27	wavelet	wavelet	NOUN
cuesj-105	13	28	transforms	transform	VERB
cuesj-105	13	29	are	be	AUX
cuesj-105	13	30	a	a	DET
cuesj-105	13	31	very	very	ADV
cuesj-105	13	32	effective	effective	ADJ
cuesj-105	13	33	method	method	NOUN
cuesj-105	13	34	and	and	CCONJ
cuesj-105	13	35	specifically	specifically	ADV
cuesj-105	13	36	in	in	ADP
cuesj-105	13	37	image	image	NOUN
cuesj-105	13	38	compression	compression	NOUN
cuesj-105	13	39	and	and	CCONJ
cuesj-105	13	40	many	many	ADJ
cuesj-105	13	41	other	other	ADJ
cuesj-105	13	42	types	type	NOUN
cuesj-105	13	43	of	of	ADP
cuesj-105	13	44	data	datum	NOUN
cuesj-105	13	45	.	.	PUNCT
cuesj-105	14	1	as	as	SCONJ
cuesj-105	14	2	many	many	ADJ
cuesj-105	14	3	coefficients	coefficient	NOUN
cuesj-105	14	4	of	of	ADP
cuesj-105	14	5	wavelet	wavelet	NOUN
cuesj-105	14	6	transform	transform	NOUN
cuesj-105	14	7	method	method	NOUN
cuesj-105	14	8	become	become	VERB
cuesj-105	14	9	zero	zero	NUM
cuesj-105	14	10	or	or	CCONJ
cuesj-105	14	11	very	very	ADV
cuesj-105	14	12	small	small	ADJ
cuesj-105	14	13	.	.	PUNCT
cuesj-105	15	1	for	for	ADP
cuesj-105	15	2	this	this	DET
cuesj-105	15	3	reason	reason	NOUN
cuesj-105	15	4	,	,	PUNCT
cuesj-105	15	5	wavelet	wavelet	NOUN
cuesj-105	15	6	transforms	transform	VERB
cuesj-105	15	7	can	can	AUX
cuesj-105	15	8	consider	consider	VERB
cuesj-105	15	9	as	as	ADP
cuesj-105	15	10	a	a	DET
cuesj-105	15	11	very	very	ADV
cuesj-105	15	12	helpful	helpful	ADJ
cuesj-105	15	13	tool	tool	NOUN
cuesj-105	15	14	for	for	ADP
cuesj-105	15	15	image	image	NOUN
cuesj-105	15	16	compression	compression	NOUN
cuesj-105	15	17	.	.	PUNCT
cuesj-105	16	1	the	the	DET
cuesj-105	16	2	primary	primary	ADJ
cuesj-105	16	3	benefit	benefit	NOUN
cuesj-105	16	4	of	of	ADP
cuesj-105	16	5	wavelet	wavelet	NOUN
cuesj-105	16	6	transforms	transform	NOUN
cuesj-105	16	7	compared	compare	VERB
cuesj-105	16	8	than	than	ADP
cuesj-105	16	9	other	other	ADJ
cuesj-105	16	10	decomposition	decomposition	NOUN
cuesj-105	16	11	methods	method	NOUN
cuesj-105	16	12	is	be	AUX
cuesj-105	16	13	that	that	SCONJ
cuesj-105	16	14	the	the	DET
cuesj-105	16	15	basic	basic	ADJ
cuesj-105	16	16	functions	function	NOUN
cuesj-105	16	17	that	that	PRON
cuesj-105	16	18	related	relate	VERB
cuesj-105	16	19	with	with	ADP
cuesj-105	16	20	a	a	DET
cuesj-105	16	21	wavelet	wavelet	NOUN
cuesj-105	16	22	decomposition	decomposition	NOUN
cuesj-105	16	23	typically	typically	ADV
cuesj-105	16	24	have	have	VERB
cuesj-105	16	25	short	short	ADJ
cuesj-105	16	26	and	and	CCONJ
cuesj-105	16	27	long	long	ADJ
cuesj-105	16	28	support	support	NOUN
cuesj-105	16	29	.	.	PUNCT
cuesj-105	17	1	the	the	DET
cuesj-105	17	2	short	short	ADJ
cuesj-105	17	3	support	support	NOUN
cuesj-105	17	4	can	can	AUX
cuesj-105	17	5	effectively	effectively	ADV
cuesj-105	17	6	act	act	VERB
cuesj-105	17	7	on	on	ADP
cuesj-105	17	8	sharp	sharp	ADJ
cuesj-105	17	9	transitions	transition	NOUN
cuesj-105	17	10	as	as	ADP
cuesj-105	17	11	the	the	DET
cuesj-105	17	12	edges	edge	NOUN
cuesj-105	17	13	of	of	ADP
cuesj-105	17	14	the	the	DET
cuesj-105	17	15	image	image	NOUN
cuesj-105	17	16	;	;	PUNCT
cuesj-105	17	17	however	however	ADV
cuesj-105	17	18	,	,	PUNCT
cuesj-105	17	19	long	long	ADJ
cuesj-105	17	20	support	support	NOUN
cuesj-105	17	21	is	be	AUX
cuesj-105	17	22	operative	operative	ADJ
cuesj-105	17	23	in	in	ADP
cuesj-105	17	24	representing	represent	VERB
cuesj-105	17	25	slow	slow	ADJ
cuesj-105	17	26	variations	variation	NOUN
cuesj-105	17	27	of	of	ADP
cuesj-105	17	28	the	the	DET
cuesj-105	17	29	image.[2	image.[2	ADJ
cuesj-105	17	30	]	]	PUNCT
cuesj-105	17	31	support	support	NOUN
cuesj-105	17	32	vector	vector	NOUN
cuesj-105	17	33	machine	machine	NOUN
cuesj-105	17	34	(	(	PUNCT
cuesj-105	17	35	svm	svm	PROPN
cuesj-105	17	36	)	)	PUNCT
cuesj-105	17	37	is	be	AUX
cuesj-105	17	38	one	one	NUM
cuesj-105	17	39	of	of	ADP
cuesj-105	17	40	important	important	ADJ
cuesj-105	17	41	learning	learning	NOUN
cuesj-105	17	42	algorithms	algorithm	NOUN
cuesj-105	17	43	that	that	PRON
cuesj-105	17	44	are	be	AUX
cuesj-105	17	45	applied	apply	VERB
cuesj-105	17	46	in	in	ADP
cuesj-105	17	47	many	many	ADJ
cuesj-105	17	48	of	of	ADP
cuesj-105	17	49	applications	application	NOUN
cuesj-105	17	50	and	and	CCONJ
cuesj-105	17	51	it	it	PRON
cuesj-105	17	52	is	be	AUX
cuesj-105	17	53	currently	currently	ADV
cuesj-105	17	54	represented	represent	VERB
cuesj-105	17	55	one	one	NUM
cuesj-105	17	56	of	of	ADP
cuesj-105	17	57	the	the	DET
cuesj-105	17	58	very	very	ADV
cuesj-105	17	59	used	used	ADJ
cuesj-105	17	60	methods	method	NOUN
cuesj-105	17	61	in	in	ADP
cuesj-105	17	62	many	many	ADJ
cuesj-105	17	63	application	application	NOUN
cuesj-105	17	64	fields	field	NOUN
cuesj-105	17	65	.	.	PUNCT
cuesj-105	18	1	the	the	DET
cuesj-105	18	2	svm	svm	PROPN
cuesj-105	18	3	theories	theory	NOUN
cuesj-105	18	4	were	be	AUX
cuesj-105	18	5	introduced	introduce	VERB
cuesj-105	18	6	at	at	ADP
cuesj-105	18	7	first	first	ADV
cuesj-105	18	8	in	in	ADP
cuesj-105	18	9	the	the	DET
cuesj-105	18	10	sixties	sixty	NOUN
cuesj-105	18	11	and	and	CCONJ
cuesj-105	18	12	seventies	seventy	NOUN
cuesj-105	18	13	from	from	ADP
cuesj-105	18	14	the	the	DET
cuesj-105	18	15	previous	previous	ADJ
cuesj-105	18	16	century	century	NOUN
cuesj-105	18	17	by	by	ADP
cuesj-105	18	18	the	the	DET
cuesj-105	18	19	scientists	scientist	NOUN
cuesj-105	18	20	vapnik	vapnik	X
cuesj-105	18	21	and	and	CCONJ
cuesj-105	18	22	chervonenkis	chervonenki	NOUN
cuesj-105	18	23	,	,	PUNCT
cuesj-105	18	24	but	but	CCONJ
cuesj-105	18	25	the	the	DET
cuesj-105	18	26	svms	svms	NOUN
cuesj-105	18	27	method	method	NOUN
cuesj-105	18	28	is	be	AUX
cuesj-105	18	29	still	still	ADV
cuesj-105	18	30	not	not	PART
cuesj-105	18	31	implementations	implementation	NOUN
cuesj-105	18	32	practical	practical	ADJ
cuesj-105	18	33	until	until	SCONJ
cuesj-105	18	34	the	the	DET
cuesj-105	18	35	early.[3	early.[3	NOUN
cuesj-105	18	36	]	]	PUNCT
cuesj-105	18	37	cihan	cihan	VERB
cuesj-105	18	38	university	university	NOUN
cuesj-105	18	39	-	-	PUNCT
cuesj-105	18	40	erbil	erbil	PROPN
cuesj-105	18	41	scientific	scientific	ADJ
cuesj-105	18	42	journal	journal	NOUN
cuesj-105	18	43	(	(	PUNCT
cuesj-105	18	44	cuesj	cuesj	PROPN
cuesj-105	18	45	)	)	PUNCT
cuesj-105	18	46	corresponding	correspond	VERB
cuesj-105	18	47	author	author	NOUN
cuesj-105	18	48	:	:	PUNCT
cuesj-105	18	49	laith	laith	PROPN
cuesj-105	18	50	r.	r.	PROPN
cuesj-105	18	51	fleah	fleah	PROPN
cuesj-105	18	52	,	,	PUNCT
cuesj-105	18	53	department	department	NOUN
cuesj-105	18	54	of	of	ADP
cuesj-105	18	55	computer	computer	NOUN
cuesj-105	18	56	science	science	NOUN
cuesj-105	18	57	,	,	PUNCT
cuesj-105	18	58	cihan	cihan	VERB
cuesj-105	18	59	university	university	NOUN
cuesj-105	18	60	-	-	PUNCT
cuesj-105	18	61	erbil	erbil	PROPN
cuesj-105	18	62	,	,	PUNCT
cuesj-105	18	63	iraq	iraq	PROPN
cuesj-105	18	64	.	.	PUNCT
cuesj-105	19	1	e	e	X
cuesj-105	19	2	-	-	NOUN
cuesj-105	19	3	mail	mail	NOUN
cuesj-105	19	4	:	:	PUNCT
cuesj-105	19	5	laith.flaih@cihanuniversity.edu.iq	laith.flaih@cihanuniversity.edu.iq	NOUN
cuesj-105	19	6	received	receive	VERB
cuesj-105	19	7	:	:	PUNCT
cuesj-105	19	8	apr	apr	NOUN
cuesj-105	19	9	18	18	NUM
cuesj-105	19	10	,	,	PUNCT
cuesj-105	19	11	2019	2019	NUM
cuesj-105	19	12	accepted	accept	VERB
cuesj-105	19	13	:	:	PUNCT
cuesj-105	19	14	apr	apr	NOUN
cuesj-105	19	15	25	25	NUM
cuesj-105	19	16	,	,	PUNCT
cuesj-105	19	17	2019	2019	NUM
cuesj-105	19	18	published	publish	VERB
cuesj-105	19	19	:	:	PUNCT
cuesj-105	19	20	aug	aug	PROPN
cuesj-105	19	21	20	20	NUM
cuesj-105	19	22	,	,	PUNCT
cuesj-105	19	23	2019	2019	NUM
cuesj-105	19	24	doi	doi	NOUN
cuesj-105	19	25	:	:	PUNCT
cuesj-105	19	26	10.24086	10.24086	NUM
cuesj-105	19	27	/	/	SYM
cuesj-105	19	28	cuesj.v3n2y2019.pp14	cuesj.v3n2y2019.pp14	ADP
cuesj-105	19	29	-	-	PUNCT
cuesj-105	19	30	20	20	NUM
cuesj-105	19	31	copyright	copyright	NOUN
cuesj-105	19	32	©	©	PROPN
cuesj-105	19	33	2019	2019	NUM
cuesj-105	19	34	laith	laith	PROPN
cuesj-105	19	35	r.	r.	PROPN
cuesj-105	19	36	fleah	fleah	PROPN
cuesj-105	19	37	,	,	PUNCT
cuesj-105	19	38	shaimaa	shaimaa	PROPN
cuesj-105	19	39	a.	a.	PROPN
cuesj-105	19	40	al	al	PROPN
cuesj-105	19	41	-	-	PUNCT
cuesj-105	19	42	aubi	aubi	PROPN
cuesj-105	19	43	.	.	PUNCT
cuesj-105	20	1	this	this	PRON
cuesj-105	20	2	is	be	AUX
cuesj-105	20	3	an	an	DET
cuesj-105	20	4	open	open	ADJ
cuesj-105	20	5	-	-	PUNCT
cuesj-105	20	6	access	access	NOUN
cuesj-105	20	7	article	article	NOUN
cuesj-105	20	8	distributed	distribute	VERB
cuesj-105	20	9	under	under	ADP
cuesj-105	20	10	the	the	DET
cuesj-105	20	11	creative	creative	ADJ
cuesj-105	20	12	commons	common	NOUN
cuesj-105	20	13	attribution	attribution	NOUN
cuesj-105	20	14	license	license	NOUN
cuesj-105	20	15	.	.	PUNCT
cuesj-105	21	1	fleah	fleah	NOUN
cuesj-105	21	2	and	and	CCONJ
cuesj-105	21	3	al	al	PROPN
cuesj-105	21	4	-	-	PUNCT
cuesj-105	21	5	aubi	aubi	PROPN
cuesj-105	21	6	:	:	PUNCT
cuesj-105	21	7	frs	frs	PROPN
cuesj-105	21	8	-	-	PUNCT
cuesj-105	21	9	pca	pca	NOUN
cuesj-105	21	10	and	and	CCONJ
cuesj-105	21	11	svm	svm	ADJ
cuesj-105	21	12	15	15	NUM
cuesj-105	21	13	http://journals.cihanuniversity.edu.iq/index.php/cuesj	http://journals.cihanuniversity.edu.iq/index.php/cuesj	ADJ
cuesj-105	21	14	cuesj	cuesj	NOUN
cuesj-105	21	15	2019	2019	NUM
cuesj-105	21	16	,	,	PUNCT
cuesj-105	21	17	3	3	NUM
cuesj-105	21	18	(	(	PUNCT
cuesj-105	21	19	2	2	NUM
cuesj-105	21	20	):	):	PUNCT
cuesj-105	21	21	14	14	NUM
cuesj-105	21	22	-	-	SYM
cuesj-105	21	23	20	20	NUM
cuesj-105	21	24	it	it	PRON
cuesj-105	21	25	has	have	AUX
cuesj-105	21	26	been	be	AUX
cuesj-105	21	27	applied	apply	VERB
cuesj-105	21	28	in	in	ADP
cuesj-105	21	29	many	many	ADJ
cuesj-105	21	30	of	of	ADP
cuesj-105	21	31	applications	application	NOUN
cuesj-105	21	32	in	in	ADP
cuesj-105	21	33	many	many	ADJ
cuesj-105	21	34	various	various	ADJ
cuesj-105	21	35	fields	field	NOUN
cuesj-105	21	36	such	such	ADJ
cuesj-105	21	37	as	as	ADP
cuesj-105	21	38	medical	medical	ADJ
cuesj-105	21	39	diagnosis	diagnosis	NOUN
cuesj-105	21	40	,	,	PUNCT
cuesj-105	21	41	text	text	NOUN
cuesj-105	21	42	or	or	CCONJ
cuesj-105	21	43	image	image	NOUN
cuesj-105	21	44	classification	classification	NOUN
cuesj-105	21	45	and	and	CCONJ
cuesj-105	21	46	also	also	ADV
cuesj-105	21	47	categorization	categorization	NOUN
cuesj-105	21	48	,	,	PUNCT
cuesj-105	21	49	spam	spam	NOUN
cuesj-105	21	50	categorization	categorization	NOUN
cuesj-105	21	51	,	,	PUNCT
cuesj-105	21	52	detection	detection	NOUN
cuesj-105	21	53	and	and	CCONJ
cuesj-105	21	54	recognition	recognition	NOUN
cuesj-105	21	55	of	of	ADP
cuesj-105	21	56	an	an	DET
cuesj-105	21	57	object	object	NOUN
cuesj-105	21	58	,	,	PUNCT
cuesj-105	21	59	face	face	NOUN
cuesj-105	21	60	detection	detection	NOUN
cuesj-105	21	61	,	,	PUNCT
cuesj-105	21	62	verification	verification	NOUN
cuesj-105	21	63	and	and	CCONJ
cuesj-105	21	64	recognition	recognition	NOUN
cuesj-105	21	65	,	,	PUNCT
cuesj-105	21	66	bioinformatics	bioinformatics	NOUN
cuesj-105	21	67	,	,	PUNCT
cuesj-105	21	68	signal	signal	NOUN
cuesj-105	21	69	processing	processing	NOUN
cuesj-105	21	70	,	,	PUNCT
cuesj-105	21	71	prediction	prediction	NOUN
cuesj-105	21	72	,	,	PUNCT
cuesj-105	21	73	information	information	NOUN
cuesj-105	21	74	,	,	PUNCT
cuesj-105	21	75	and	and	CCONJ
cuesj-105	21	76	image	image	NOUN
cuesj-105	21	77	retrieval.[4	retrieval.[4	NOUN
cuesj-105	21	78	]	]	X
cuesj-105	21	79	k	k	ADJ
cuesj-105	21	80	-	-	PUNCT
cuesj-105	21	81	nearest	near	ADJ
cuesj-105	21	82	neighbor	neighbor	NOUN
cuesj-105	21	83	[	[	X
cuesj-105	21	84	knn	knn	X
cuesj-105	21	85	]	]	X
cuesj-105	21	86	is	be	AUX
cuesj-105	21	87	one	one	NUM
cuesj-105	21	88	of	of	ADP
cuesj-105	21	89	the	the	DET
cuesj-105	21	90	most	most	ADV
cuesj-105	21	91	important	important	ADJ
cuesj-105	21	92	algorithms	algorithm	NOUN
cuesj-105	21	93	used	use	VERB
cuesj-105	21	94	in	in	ADP
cuesj-105	21	95	face	face	NOUN
cuesj-105	21	96	recognition	recognition	NOUN
cuesj-105	21	97	system	system	NOUN
cuesj-105	21	98	.	.	PUNCT
cuesj-105	22	1	this	this	DET
cuesj-105	22	2	algorithm	algorithm	NOUN
cuesj-105	22	3	is	be	AUX
cuesj-105	22	4	easy	easy	ADJ
cuesj-105	22	5	to	to	PART
cuesj-105	22	6	understand	understand	VERB
cuesj-105	22	7	,	,	PUNCT
cuesj-105	22	8	but	but	CCONJ
cuesj-105	22	9	in	in	ADP
cuesj-105	22	10	another	another	DET
cuesj-105	22	11	hand	hand	NOUN
cuesj-105	22	12	,	,	PUNCT
cuesj-105	22	13	it	it	PRON
cuesj-105	22	14	is	be	AUX
cuesj-105	22	15	useful	useful	ADJ
cuesj-105	22	16	in	in	ADP
cuesj-105	22	17	practice	practice	NOUN
cuesj-105	22	18	.	.	PUNCT
cuesj-105	23	1	furthermore	furthermore	ADV
cuesj-105	23	2	,	,	PUNCT
cuesj-105	23	3	it	it	PRON
cuesj-105	23	4	is	be	AUX
cuesj-105	23	5	used	use	VERB
cuesj-105	23	6	in	in	ADP
cuesj-105	23	7	several	several	ADJ
cuesj-105	23	8	of	of	ADP
cuesj-105	23	9	applications	application	NOUN
cuesj-105	23	10	such	such	ADJ
cuesj-105	23	11	as	as	ADP
cuesj-105	23	12	vision	vision	NOUN
cuesj-105	23	13	,	,	PUNCT
cuesj-105	23	14	proteins	protein	NOUN
cuesj-105	23	15	,	,	PUNCT
cuesj-105	23	16	computational	computational	ADJ
cuesj-105	23	17	geometry	geometry	NOUN
cuesj-105	23	18	,	,	PUNCT
cuesj-105	23	19	and	and	CCONJ
cuesj-105	23	20	graphs	graph	NOUN
cuesj-105	23	21	.	.	PUNCT
cuesj-105	24	1	it	it	PRON
cuesj-105	24	2	also	also	ADV
cuesj-105	24	3	considered	consider	VERB
cuesj-105	24	4	as	as	ADP
cuesj-105	24	5	one	one	NUM
cuesj-105	24	6	of	of	ADP
cuesj-105	24	7	the	the	DET
cuesj-105	24	8	best	good	ADJ
cuesj-105	24	9	10	10	NUM
cuesj-105	24	10	algorithms	algorithm	NOUN
cuesj-105	24	11	in	in	ADP
cuesj-105	24	12	data	datum	NOUN
cuesj-105	24	13	mining.[5,6	mining.[5,6	PROPN
cuesj-105	24	14	]	]	X
cuesj-105	24	15	knn	knn	PROPN
cuesj-105	24	16	is	be	AUX
cuesj-105	24	17	considered	consider	VERB
cuesj-105	24	18	as	as	ADP
cuesj-105	24	19	one	one	NUM
cuesj-105	24	20	of	of	ADP
cuesj-105	24	21	the	the	DET
cuesj-105	24	22	non	non	ADJ
cuesj-105	24	23	-	-	ADJ
cuesj-105	24	24	parametric	parametric	ADJ
cuesj-105	24	25	learning	learning	NOUN
cuesj-105	24	26	algorithms	algorithm	NOUN
cuesj-105	24	27	.	.	PUNCT
cuesj-105	25	1	this	this	PRON
cuesj-105	25	2	means	mean	VERB
cuesj-105	25	3	that	that	SCONJ
cuesj-105	25	4	it	it	PRON
cuesj-105	25	5	does	do	AUX
cuesj-105	25	6	not	not	PART
cuesj-105	25	7	produce	produce	VERB
cuesj-105	25	8	any	any	DET
cuesj-105	25	9	supposition	supposition	NOUN
cuesj-105	25	10	on	on	ADP
cuesj-105	25	11	the	the	DET
cuesj-105	25	12	implicit	implicit	ADJ
cuesj-105	25	13	data	datum	NOUN
cuesj-105	25	14	distribution	distribution	NOUN
cuesj-105	25	15	.	.	PUNCT
cuesj-105	26	1	this	this	PRON
cuesj-105	26	2	is	be	AUX
cuesj-105	26	3	very	very	ADV
cuesj-105	26	4	helpful	helpful	ADJ
cuesj-105	26	5	in	in	ADP
cuesj-105	26	6	the	the	DET
cuesj-105	26	7	real	real	ADJ
cuesj-105	26	8	-	-	PUNCT
cuesj-105	26	9	world	world	NOUN
cuesj-105	26	10	application	application	NOUN
cuesj-105	26	11	,	,	PUNCT
cuesj-105	26	12	many	many	ADJ
cuesj-105	26	13	of	of	ADP
cuesj-105	26	14	the	the	DET
cuesj-105	26	15	practical	practical	ADJ
cuesj-105	26	16	data	datum	NOUN
cuesj-105	26	17	do	do	AUX
cuesj-105	26	18	not	not	PART
cuesj-105	26	19	conform	conform	VERB
cuesj-105	26	20	the	the	DET
cuesj-105	26	21	classical	classical	ADJ
cuesj-105	26	22	theoretical	theoretical	ADJ
cuesj-105	26	23	supposition	supposition	NOUN
cuesj-105	26	24	made	make	VERB
cuesj-105	26	25	such	such	ADJ
cuesj-105	26	26	as	as	ADP
cuesj-105	26	27	linearly	linearly	ADV
cuesj-105	26	28	separable	separable	ADJ
cuesj-105	26	29	and	and	CCONJ
cuesj-105	26	30	gaussian	gaussian	ADJ
cuesj-105	26	31	mixtures.[5,6	mixtures.[5,6	NOUN
cuesj-105	26	32	]	]	X
cuesj-105	26	33	the	the	DET
cuesj-105	26	34	overall	overall	ADJ
cuesj-105	26	35	structure	structure	NOUN
cuesj-105	26	36	of	of	ADP
cuesj-105	26	37	the	the	DET
cuesj-105	26	38	paper	paper	NOUN
cuesj-105	26	39	is	be	AUX
cuesj-105	26	40	organized	organize	VERB
cuesj-105	26	41	as	as	SCONJ
cuesj-105	26	42	follows	follow	VERB
cuesj-105	26	43	:	:	PUNCT
cuesj-105	26	44	section	section	PROPN
cuesj-105	26	45	ii	ii	PROPN
cuesj-105	26	46	:	:	PUNCT
cuesj-105	26	47	introduced	introduce	VERB
cuesj-105	26	48	the	the	DET
cuesj-105	26	49	structure	structure	NOUN
cuesj-105	26	50	of	of	ADP
cuesj-105	26	51	the	the	DET
cuesj-105	26	52	system	system	NOUN
cuesj-105	26	53	,	,	PUNCT
cuesj-105	26	54	this	this	PRON
cuesj-105	26	55	is	be	AUX
cuesj-105	26	56	included	include	VERB
cuesj-105	26	57	pre	pre	ADJ
cuesj-105	26	58	-	-	ADJ
cuesj-105	26	59	processing	processing	ADJ
cuesj-105	26	60	methods	method	NOUN
cuesj-105	26	61	,	,	PUNCT
cuesj-105	26	62	feature	feature	NOUN
cuesj-105	26	63	extraction	extraction	NOUN
cuesj-105	26	64	methods	method	NOUN
cuesj-105	26	65	,	,	PUNCT
cuesj-105	26	66	and	and	CCONJ
cuesj-105	26	67	classification	classification	NOUN
cuesj-105	26	68	methods	method	NOUN
cuesj-105	26	69	.	.	PUNCT
cuesj-105	27	1	in	in	ADP
cuesj-105	27	2	section	section	PROPN
cuesj-105	27	3	iii	iii	PROPN
cuesj-105	27	4	,	,	PUNCT
cuesj-105	27	5	the	the	DET
cuesj-105	27	6	experimental	experimental	ADJ
cuesj-105	27	7	results	result	NOUN
cuesj-105	27	8	and	and	CCONJ
cuesj-105	27	9	its	its	PRON
cuesj-105	27	10	analysis	analysis	NOUN
cuesj-105	27	11	are	be	AUX
cuesj-105	27	12	introduced	introduce	VERB
cuesj-105	27	13	while	while	SCONJ
cuesj-105	27	14	section	section	NOUN
cuesj-105	27	15	iv	iv	NUM
cuesj-105	27	16	concludes	conclude	VERB
cuesj-105	27	17	the	the	DET
cuesj-105	27	18	paper	paper	NOUN
cuesj-105	27	19	.	.	PUNCT
cuesj-105	28	1	face	face	NOUN
cuesj-105	28	2	system	system	NOUN
cuesj-105	28	3	structure	structure	NOUN
cuesj-105	28	4	face	face	NOUN
cuesj-105	28	5	recognition	recognition	NOUN
cuesj-105	28	6	system	system	NOUN
cuesj-105	28	7	that	that	PRON
cuesj-105	28	8	introduced	introduce	VERB
cuesj-105	28	9	in	in	ADP
cuesj-105	28	10	this	this	DET
cuesj-105	28	11	paper	paper	NOUN
cuesj-105	28	12	consists	consist	VERB
cuesj-105	28	13	of	of	ADP
cuesj-105	28	14	the	the	DET
cuesj-105	28	15	following	follow	VERB
cuesj-105	28	16	steps	step	NOUN
cuesj-105	28	17	:	:	PUNCT
cuesj-105	28	18	pre	pre	ADJ
cuesj-105	28	19	-	-	ADJ
cuesj-105	28	20	processing	processing	ADJ
cuesj-105	28	21	,	,	PUNCT
cuesj-105	28	22	feature	feature	NOUN
cuesj-105	28	23	extraction	extraction	NOUN
cuesj-105	28	24	,	,	PUNCT
cuesj-105	28	25	classification	classification	NOUN
cuesj-105	28	26	,	,	PUNCT
cuesj-105	28	27	and	and	CCONJ
cuesj-105	28	28	in	in	ADP
cuesj-105	28	29	the	the	DET
cuesj-105	28	30	final	final	ADJ
cuesj-105	28	31	steps	step	NOUN
cuesj-105	28	32	is	be	AUX
cuesj-105	28	33	identification	identification	NOUN
cuesj-105	28	34	.	.	PUNCT
cuesj-105	29	1	the	the	DET
cuesj-105	29	2	stages	stage	NOUN
cuesj-105	29	3	of	of	ADP
cuesj-105	29	4	the	the	DET
cuesj-105	29	5	system	system	NOUN
cuesj-105	29	6	are	be	AUX
cuesj-105	29	7	shown	show	VERB
cuesj-105	29	8	in	in	ADP
cuesj-105	29	9	figure	figure	NOUN
cuesj-105	29	10	1	1	NUM
cuesj-105	29	11	.	.	PUNCT
cuesj-105	30	1	pre	pre	ADJ
cuesj-105	30	2	-	-	ADJ
cuesj-105	30	3	processing	processing	ADJ
cuesj-105	30	4	pre	pre	ADJ
cuesj-105	30	5	-	-	ADJ
cuesj-105	30	6	processing	processing	ADJ
cuesj-105	30	7	stage	stage	NOUN
cuesj-105	30	8	is	be	AUX
cuesj-105	30	9	the	the	DET
cuesj-105	30	10	first	first	ADJ
cuesj-105	30	11	process	process	NOUN
cuesj-105	30	12	that	that	PRON
cuesj-105	30	13	done	do	VERB
cuesj-105	30	14	in	in	ADP
cuesj-105	30	15	the	the	DET
cuesj-105	30	16	proposed	propose	VERB
cuesj-105	30	17	system	system	NOUN
cuesj-105	30	18	.	.	PUNCT
cuesj-105	31	1	it	it	PRON
cuesj-105	31	2	is	be	AUX
cuesj-105	31	3	applied	apply	VERB
cuesj-105	31	4	to	to	PART
cuesj-105	31	5	improve	improve	VERB
cuesj-105	31	6	the	the	DET
cuesj-105	31	7	image	image	NOUN
cuesj-105	31	8	and	and	CCONJ
cuesj-105	31	9	also	also	ADV
cuesj-105	31	10	to	to	PART
cuesj-105	31	11	improve	improve	VERB
cuesj-105	31	12	the	the	DET
cuesj-105	31	13	recognition	recognition	NOUN
cuesj-105	31	14	rate	rate	NOUN
cuesj-105	31	15	through	through	AUX
cuesj-105	31	16	remove	remove	VERB
cuesj-105	31	17	the	the	DET
cuesj-105	31	18	effect	effect	NOUN
cuesj-105	31	19	of	of	ADP
cuesj-105	31	20	the	the	DET
cuesj-105	31	21	light	light	NOUN
cuesj-105	31	22	on	on	ADP
cuesj-105	31	23	and	and	CCONJ
cuesj-105	31	24	also	also	ADV
cuesj-105	31	25	to	to	PART
cuesj-105	31	26	correct	correct	VERB
cuesj-105	31	27	the	the	DET
cuesj-105	31	28	gradient	gradient	NOUN
cuesj-105	31	29	of	of	ADP
cuesj-105	31	30	illumination	illumination	NOUN
cuesj-105	31	31	in	in	ADP
cuesj-105	31	32	the	the	DET
cuesj-105	31	33	image	image	NOUN
cuesj-105	31	34	.	.	PUNCT
cuesj-105	32	1	in	in	ADP
cuesj-105	32	2	the	the	DET
cuesj-105	32	3	following	following	NOUN
cuesj-105	32	4	,	,	PUNCT
cuesj-105	32	5	the	the	DET
cuesj-105	32	6	pre	pre	ADJ
cuesj-105	32	7	-	-	ADJ
cuesj-105	32	8	processing	processing	ADJ
cuesj-105	32	9	techniques	technique	NOUN
cuesj-105	32	10	that	that	PRON
cuesj-105	32	11	applied	apply	VERB
cuesj-105	32	12	at	at	ADP
cuesj-105	32	13	this	this	DET
cuesj-105	32	14	stage	stage	NOUN
cuesj-105	32	15	to	to	PART
cuesj-105	32	16	get	get	VERB
cuesj-105	32	17	the	the	DET
cuesj-105	32	18	good	good	ADJ
cuesj-105	32	19	results	result	NOUN
cuesj-105	32	20	:	:	PUNCT
cuesj-105	32	21	normalization	normalization	NOUN
cuesj-105	32	22	size	size	NOUN
cuesj-105	32	23	of	of	ADP
cuesj-105	32	24	the	the	DET
cuesj-105	32	25	image	image	NOUN
cuesj-105	32	26	:	:	PUNCT
cuesj-105	32	27	two	two	NUM
cuesj-105	32	28	types	type	NOUN
cuesj-105	32	29	of	of	ADP
cuesj-105	32	30	database	database	NOUN
cuesj-105	32	31	are	be	AUX
cuesj-105	32	32	used	use	VERB
cuesj-105	32	33	in	in	ADP
cuesj-105	32	34	this	this	DET
cuesj-105	32	35	thesis	thesis	NOUN
cuesj-105	32	36	and	and	CCONJ
cuesj-105	32	37	each	each	DET
cuesj-105	32	38	database	database	NOUN
cuesj-105	32	39	has	have	VERB
cuesj-105	32	40	different	different	ADJ
cuesj-105	32	41	size	size	NOUN
cuesj-105	32	42	of	of	ADP
cuesj-105	32	43	an	an	DET
cuesj-105	32	44	image	image	NOUN
cuesj-105	32	45	;	;	PUNCT
cuesj-105	32	46	therefore	therefore	ADV
cuesj-105	32	47	,	,	PUNCT
cuesj-105	32	48	this	this	DET
cuesj-105	32	49	technique	technique	NOUN
cuesj-105	32	50	is	be	AUX
cuesj-105	32	51	used	use	VERB
cuesj-105	32	52	to	to	PART
cuesj-105	32	53	make	make	VERB
cuesj-105	32	54	all	all	DET
cuesj-105	32	55	images	image	NOUN
cuesj-105	32	56	have	have	VERB
cuesj-105	32	57	the	the	DET
cuesj-105	32	58	same	same	ADJ
cuesj-105	32	59	size	size	NOUN
cuesj-105	32	60	.	.	PUNCT
cuesj-105	33	1	this	this	PRON
cuesj-105	33	2	is	be	AUX
cuesj-105	33	3	because	because	SCONJ
cuesj-105	33	4	the	the	DET
cuesj-105	33	5	next	next	ADJ
cuesj-105	33	6	stages	stage	NOUN
cuesj-105	33	7	have	have	VERB
cuesj-105	33	8	an	an	DET
cuesj-105	33	9	arithmetic	arithmetic	ADJ
cuesj-105	33	10	operation	operation	NOUN
cuesj-105	33	11	such	such	ADJ
cuesj-105	33	12	as	as	ADP
cuesj-105	33	13	multiplication	multiplication	NOUN
cuesj-105	33	14	and	and	CCONJ
cuesj-105	33	15	division	division	NOUN
cuesj-105	33	16	of	of	ADP
cuesj-105	33	17	array	array	NOUN
cuesj-105	33	18	.	.	PUNCT
cuesj-105	34	1	for	for	ADP
cuesj-105	34	2	all	all	DET
cuesj-105	34	3	these	these	DET
cuesj-105	34	4	reasons	reason	NOUN
cuesj-105	34	5	,	,	PUNCT
cuesj-105	34	6	the	the	DET
cuesj-105	34	7	size	size	NOUN
cuesj-105	34	8	normalization	normalization	NOUN
cuesj-105	34	9	is	be	AUX
cuesj-105	34	10	very	very	ADV
cuesj-105	34	11	significant	significant	ADJ
cuesj-105	34	12	.	.	PUNCT
cuesj-105	35	1	contrast	contrast	NOUN
cuesj-105	35	2	stretching	stretch	VERB
cuesj-105	35	3	:	:	PUNCT
cuesj-105	35	4	it	it	PRON
cuesj-105	35	5	is	be	AUX
cuesj-105	35	6	also	also	ADV
cuesj-105	35	7	known	know	VERB
cuesj-105	35	8	as	as	ADP
cuesj-105	35	9	histogram	histogram	NOUN
cuesj-105	35	10	stretching	stretching	NOUN
cuesj-105	35	11	.	.	PUNCT
cuesj-105	36	1	it	it	PRON
cuesj-105	36	2	is	be	AUX
cuesj-105	36	3	considered	consider	VERB
cuesj-105	36	4	as	as	ADP
cuesj-105	36	5	one	one	NUM
cuesj-105	36	6	of	of	ADP
cuesj-105	36	7	the	the	DET
cuesj-105	36	8	techniques	technique	NOUN
cuesj-105	36	9	that	that	PRON
cuesj-105	36	10	used	use	VERB
cuesj-105	36	11	to	to	PART
cuesj-105	36	12	enhance	enhance	VERB
cuesj-105	36	13	an	an	DET
cuesj-105	36	14	image	image	NOUN
cuesj-105	36	15	.	.	PUNCT
cuesj-105	37	1	the	the	DET
cuesj-105	37	2	goal	goal	NOUN
cuesj-105	37	3	of	of	ADP
cuesj-105	37	4	this	this	DET
cuesj-105	37	5	technique	technique	NOUN
cuesj-105	37	6	has	have	AUX
cuesj-105	37	7	improved	improve	VERB
cuesj-105	37	8	the	the	DET
cuesj-105	37	9	contrast	contrast	NOUN
cuesj-105	37	10	in	in	ADP
cuesj-105	37	11	an	an	DET
cuesj-105	37	12	image	image	NOUN
cuesj-105	37	13	through	through	ADP
cuesj-105	37	14	stretch	stretch	VERB
cuesj-105	37	15	the	the	DET
cuesj-105	37	16	range	range	NOUN
cuesj-105	37	17	of	of	ADP
cuesj-105	37	18	intensity	intensity	NOUN
cuesj-105	37	19	values	value	NOUN
cuesj-105	37	20	.	.	PUNCT
cuesj-105	38	1	in	in	ADP
cuesj-105	38	2	other	other	ADJ
cuesj-105	38	3	words	word	NOUN
cuesj-105	38	4	,	,	PUNCT
cuesj-105	38	5	this	this	PRON
cuesj-105	38	6	is	be	AUX
cuesj-105	38	7	applied	apply	VERB
cuesj-105	38	8	by	by	ADP
cuesj-105	38	9	raising	raise	VERB
cuesj-105	38	10	the	the	DET
cuesj-105	38	11	dynamic	dynamic	ADJ
cuesj-105	38	12	range	range	NOUN
cuesj-105	38	13	of	of	ADP
cuesj-105	38	14	the	the	DET
cuesj-105	38	15	gray	gray	ADJ
cuesj-105	38	16	levels	level	NOUN
cuesj-105	38	17	in	in	ADP
cuesj-105	38	18	the	the	DET
cuesj-105	38	19	image	image	NOUN
cuesj-105	38	20	.	.	PUNCT
cuesj-105	39	1	for	for	ADP
cuesj-105	39	2	example	example	NOUN
cuesj-105	39	3	,	,	PUNCT
cuesj-105	39	4	in	in	ADP
cuesj-105	39	5	an	an	DET
cuesj-105	39	6	8	8	NUM
cuesj-105	39	7	-	-	PUNCT
cuesj-105	39	8	bit	bit	NOUN
cuesj-105	39	9	system	system	NOUN
cuesj-105	39	10	,	,	PUNCT
cuesj-105	39	11	the	the	DET
cuesj-105	39	12	256	256	NUM
cuesj-105	39	13	gray	gray	ADJ
cuesj-105	39	14	levels	level	NOUN
cuesj-105	39	15	can	can	AUX
cuesj-105	39	16	be	be	AUX
cuesj-105	39	17	shown	show	VERB
cuesj-105	39	18	only	only	ADV
cuesj-105	39	19	but	but	CCONJ
cuesj-105	39	20	if	if	SCONJ
cuesj-105	39	21	the	the	DET
cuesj-105	39	22	number	number	NOUN
cuesj-105	39	23	of	of	ADP
cuesj-105	39	24	gray	gray	ADJ
cuesj-105	39	25	levels	level	NOUN
cuesj-105	39	26	in	in	ADP
cuesj-105	39	27	the	the	DET
cuesj-105	39	28	captured	capture	VERB
cuesj-105	39	29	image	image	NOUN
cuesj-105	39	30	prevalence	prevalence	NOUN
cuesj-105	39	31	through	through	ADP
cuesj-105	39	32	a	a	DET
cuesj-105	39	33	lesser	less	ADJ
cuesj-105	39	34	range	range	NOUN
cuesj-105	39	35	,	,	PUNCT
cuesj-105	39	36	the	the	DET
cuesj-105	39	37	images	image	NOUN
cuesj-105	39	38	can	can	AUX
cuesj-105	39	39	be	be	AUX
cuesj-105	39	40	improved	improve	VERB
cuesj-105	39	41	by	by	ADP
cuesj-105	39	42	extended	extend	VERB
cuesj-105	39	43	this	this	DET
cuesj-105	39	44	number	number	NOUN
cuesj-105	39	45	to	to	ADP
cuesj-105	39	46	a	a	DET
cuesj-105	39	47	wider	wide	ADJ
cuesj-105	39	48	range	range	NOUN
cuesj-105	39	49	.	.	PUNCT
cuesj-105	40	1	features	feature	NOUN
cuesj-105	40	2	extraction	extraction	VERB
cuesj-105	40	3	the	the	DET
cuesj-105	40	4	second	second	ADJ
cuesj-105	40	5	stage	stage	NOUN
cuesj-105	40	6	of	of	ADP
cuesj-105	40	7	this	this	DET
cuesj-105	40	8	system	system	NOUN
cuesj-105	40	9	is	be	AUX
cuesj-105	40	10	feature	feature	NOUN
cuesj-105	40	11	extraction	extraction	NOUN
cuesj-105	40	12	.	.	PUNCT
cuesj-105	41	1	it	it	PRON
cuesj-105	41	2	is	be	AUX
cuesj-105	41	3	used	use	VERB
cuesj-105	41	4	to	to	PART
cuesj-105	41	5	obtain	obtain	VERB
cuesj-105	41	6	the	the	DET
cuesj-105	41	7	significant	significant	ADJ
cuesj-105	41	8	features	feature	NOUN
cuesj-105	41	9	from	from	ADP
cuesj-105	41	10	the	the	DET
cuesj-105	41	11	picture	picture	NOUN
cuesj-105	41	12	of	of	ADP
cuesj-105	41	13	the	the	DET
cuesj-105	41	14	face	face	NOUN
cuesj-105	41	15	and	and	CCONJ
cuesj-105	41	16	then	then	ADV
cuesj-105	41	17	applied	apply	VERB
cuesj-105	41	18	these	these	DET
cuesj-105	41	19	features	feature	NOUN
cuesj-105	41	20	to	to	ADP
cuesj-105	41	21	the	the	DET
cuesj-105	41	22	next	next	ADJ
cuesj-105	41	23	stage	stage	NOUN
cuesj-105	41	24	(	(	PUNCT
cuesj-105	41	25	identification	identification	NOUN
cuesj-105	41	26	and	and	CCONJ
cuesj-105	41	27	classification	classification	NOUN
cuesj-105	41	28	stages	stage	NOUN
cuesj-105	41	29	)	)	PUNCT
cuesj-105	41	30	.	.	PUNCT
cuesj-105	42	1	without	without	ADP
cuesj-105	42	2	this	this	DET
cuesj-105	42	3	stage	stage	NOUN
cuesj-105	42	4	,	,	PUNCT
cuesj-105	42	5	the	the	DET
cuesj-105	42	6	function	function	NOUN
cuesj-105	42	7	of	of	ADP
cuesj-105	42	8	recognition	recognition	NOUN
cuesj-105	42	9	and	and	CCONJ
cuesj-105	42	10	classification	classification	NOUN
cuesj-105	42	11	becomes	become	VERB
cuesj-105	42	12	very	very	ADV
cuesj-105	42	13	hard	hard	ADJ
cuesj-105	42	14	and	and	CCONJ
cuesj-105	42	15	very	very	ADV
cuesj-105	42	16	difficult	difficult	ADJ
cuesj-105	42	17	because	because	SCONJ
cuesj-105	42	18	the	the	DET
cuesj-105	42	19	picture	picture	NOUN
cuesj-105	42	20	consists	consist	VERB
cuesj-105	42	21	of	of	ADP
cuesj-105	42	22	many	many	ADJ
cuesj-105	42	23	features	feature	NOUN
cuesj-105	42	24	and	and	CCONJ
cuesj-105	42	25	most	most	ADJ
cuesj-105	42	26	of	of	ADP
cuesj-105	42	27	these	these	DET
cuesj-105	42	28	features	feature	NOUN
cuesj-105	42	29	can	can	AUX
cuesj-105	42	30	not	not	PART
cuesj-105	42	31	be	be	AUX
cuesj-105	42	32	applied	apply	VERB
cuesj-105	42	33	to	to	PART
cuesj-105	42	34	implement	implement	VERB
cuesj-105	42	35	classification	classification	NOUN
cuesj-105	42	36	.	.	PUNCT
cuesj-105	43	1	in	in	ADP
cuesj-105	43	2	this	this	DET
cuesj-105	43	3	paper	paper	NOUN
cuesj-105	43	4	,	,	PUNCT
cuesj-105	43	5	each	each	PRON
cuesj-105	43	6	of	of	ADP
cuesj-105	43	7	pca	pca	PROPN
cuesj-105	43	8	and	and	CCONJ
cuesj-105	43	9	wavelet	wavelet	NOUN
cuesj-105	43	10	transform	transform	NOUN
cuesj-105	43	11	techniques	technique	NOUN
cuesj-105	43	12	is	be	AUX
cuesj-105	43	13	used	use	VERB
cuesj-105	43	14	to	to	PART
cuesj-105	43	15	obtain	obtain	VERB
cuesj-105	43	16	the	the	DET
cuesj-105	43	17	significant	significant	ADJ
cuesj-105	43	18	features	feature	NOUN
cuesj-105	43	19	that	that	PRON
cuesj-105	43	20	are	be	AUX
cuesj-105	43	21	important	important	ADJ
cuesj-105	43	22	to	to	PART
cuesj-105	43	23	identify	identify	VERB
cuesj-105	43	24	and	and	CCONJ
cuesj-105	43	25	classify	classify	VERB
cuesj-105	43	26	the	the	DET
cuesj-105	43	27	face	face	NOUN
cuesj-105	43	28	pictures	picture	NOUN
cuesj-105	43	29	.	.	PUNCT
cuesj-105	44	1	after	after	SCONJ
cuesj-105	44	2	preprocessing	preprocesse	VERB
cuesj-105	44	3	stage	stage	NOUN
cuesj-105	44	4	is	be	AUX
cuesj-105	44	5	implemented	implement	VERB
cuesj-105	44	6	,	,	PUNCT
cuesj-105	44	7	the	the	DET
cuesj-105	44	8	face	face	NOUN
cuesj-105	44	9	picture	picture	NOUN
cuesj-105	44	10	i	i	PRON
cuesj-105	44	11	(	(	PUNCT
cuesj-105	44	12	x	x	PROPN
cuesj-105	44	13	,	,	PUNCT
cuesj-105	44	14	y	y	PROPN
cuesj-105	44	15	)	)	PUNCT
cuesj-105	44	16	that	that	PRON
cuesj-105	44	17	is	be	AUX
cuesj-105	44	18	a	a	DET
cuesj-105	44	19	two	two	NUM
cuesj-105	44	20	-	-	PUNCT
cuesj-105	44	21	dimensional	dimensional	ADJ
cuesj-105	44	22	matrix	matrix	NOUN
cuesj-105	44	23	(	(	PUNCT
cuesj-105	44	24	n	n	CCONJ
cuesj-105	44	25	×	×	NOUN
cuesj-105	44	26	n	n	CCONJ
cuesj-105	44	27	)	)	PUNCT
cuesj-105	44	28	is	be	AUX
cuesj-105	44	29	then	then	ADV
cuesj-105	44	30	transformed	transform	VERB
cuesj-105	44	31	to	to	ADP
cuesj-105	44	32	a	a	DET
cuesj-105	44	33	vector	vector	NOUN
cuesj-105	44	34	and	and	CCONJ
cuesj-105	44	35	that	that	DET
cuesj-105	44	36	length	length	NOUN
cuesj-105	44	37	is	be	AUX
cuesj-105	44	38	had	have	VERB
cuesj-105	44	39	dimension	dimension	NOUN
cuesj-105	44	40	n²	n²	PROPN
cuesj-105	44	41	.	.	PUNCT
cuesj-105	45	1	thus	thus	ADV
cuesj-105	45	2	,	,	PUNCT
cuesj-105	45	3	the	the	DET
cuesj-105	45	4	picture	picture	NOUN
cuesj-105	45	5	with	with	ADP
cuesj-105	45	6	dimension	dimension	NOUN
cuesj-105	45	7	64	64	NUM
cuesj-105	45	8	×	×	NOUN
cuesj-105	45	9	64	64	NUM
cuesj-105	45	10	will	will	AUX
cuesj-105	45	11	transform	transform	VERB
cuesj-105	45	12	to	to	ADP
cuesj-105	45	13	a	a	DET
cuesj-105	45	14	vector	vector	NOUN
cuesj-105	45	15	of	of	ADP
cuesj-105	45	16	dimension	dimension	NOUN
cuesj-105	45	17	4096	4096	NUM
cuesj-105	45	18	,	,	PUNCT
cuesj-105	45	19	but	but	CCONJ
cuesj-105	45	20	this	this	DET
cuesj-105	45	21	vector	vector	NOUN
cuesj-105	45	22	can	can	AUX
cuesj-105	45	23	not	not	PART
cuesj-105	45	24	be	be	AUX
cuesj-105	45	25	applied	apply	VERB
cuesj-105	45	26	directly	directly	ADV
cuesj-105	45	27	to	to	PART
cuesj-105	45	28	recognition	recognition	VERB
cuesj-105	45	29	directly	directly	ADV
cuesj-105	45	30	because	because	SCONJ
cuesj-105	45	31	the	the	DET
cuesj-105	45	32	face	face	NOUN
cuesj-105	45	33	of	of	ADP
cuesj-105	45	34	human	human	NOUN
cuesj-105	45	35	is	be	AUX
cuesj-105	45	36	very	very	ADV
cuesj-105	45	37	comparable	comparable	ADJ
cuesj-105	45	38	from	from	ADP
cuesj-105	45	39	each	each	DET
cuesj-105	45	40	other	other	ADJ
cuesj-105	45	41	and	and	CCONJ
cuesj-105	45	42	another	another	DET
cuesj-105	45	43	reason	reason	NOUN
cuesj-105	45	44	the	the	DET
cuesj-105	45	45	4096	4096	NUM
cuesj-105	45	46	-	-	PUNCT
cuesj-105	45	47	dimension	dimension	NOUN
cuesj-105	45	48	space	space	NOUN
cuesj-105	45	49	is	be	AUX
cuesj-105	45	50	big	big	ADJ
cuesj-105	45	51	and	and	CCONJ
cuesj-105	45	52	it	it	PRON
cuesj-105	45	53	is	be	AUX
cuesj-105	45	54	needed	need	VERB
cuesj-105	45	55	much	much	ADJ
cuesj-105	45	56	time	time	NOUN
cuesj-105	45	57	to	to	ADP
cuesj-105	45	58	process	process	NOUN
cuesj-105	45	59	.	.	PUNCT
cuesj-105	46	1	therefore	therefore	ADV
cuesj-105	46	2	,	,	PUNCT
cuesj-105	46	3	wavelet	wavelet	NOUN
cuesj-105	46	4	and	and	CCONJ
cuesj-105	46	5	pca	pca	NOUN
cuesj-105	46	6	techniques	technique	NOUN
cuesj-105	46	7	are	be	AUX
cuesj-105	46	8	applied	apply	VERB
cuesj-105	46	9	to	to	PART
cuesj-105	46	10	obtain	obtain	VERB
cuesj-105	46	11	a	a	DET
cuesj-105	46	12	better	well	ADJ
cuesj-105	46	13	feature	feature	NOUN
cuesj-105	46	14	from	from	ADP
cuesj-105	46	15	face	face	NOUN
cuesj-105	46	16	picture	picture	NOUN
cuesj-105	46	17	and	and	CCONJ
cuesj-105	46	18	also	also	ADV
cuesj-105	46	19	to	to	PART
cuesj-105	46	20	enhance	enhance	VERB
cuesj-105	46	21	the	the	DET
cuesj-105	46	22	capability	capability	NOUN
cuesj-105	46	23	of	of	ADP
cuesj-105	46	24	the	the	DET
cuesj-105	46	25	recognition	recognition	NOUN
cuesj-105	46	26	rate	rate	NOUN
cuesj-105	46	27	.	.	PUNCT
cuesj-105	47	1	pca	pca	NOUN
cuesj-105	47	2	algorithm	algorithm	NOUN
cuesj-105	47	3	at	at	ADP
cuesj-105	47	4	first	first	ADV
cuesj-105	47	5	,	,	PUNCT
cuesj-105	47	6	the	the	DET
cuesj-105	47	7	training	training	NOUN
cuesj-105	47	8	set	set	NOUN
cuesj-105	47	9	is	be	AUX
cuesj-105	47	10	generated	generate	VERB
cuesj-105	47	11	and	and	CCONJ
cuesj-105	47	12	this	this	DET
cuesj-105	47	13	training	training	NOUN
cuesj-105	47	14	set	set	NOUN
cuesj-105	47	15	consists	consist	VERB
cuesj-105	47	16	of	of	ADP
cuesj-105	47	17	a	a	DET
cuesj-105	47	18	number	number	NOUN
cuesj-105	47	19	of	of	ADP
cuesj-105	47	20	images	image	NOUN
cuesj-105	47	21	(	(	PUNCT
cuesj-105	47	22	m	m	NOUN
cuesj-105	47	23	)	)	PUNCT
cuesj-105	47	24	and	and	CCONJ
cuesj-105	47	25	each	each	DET
cuesj-105	47	26	image	image	NOUN
cuesj-105	47	27	in	in	ADP
cuesj-105	47	28	its	its	PRON
cuesj-105	47	29	to	to	PART
cuesj-105	47	30	be	be	AUX
cuesj-105	47	31	transformed	transform	VERB
cuesj-105	47	32	to	to	ADP
cuesj-105	47	33	the	the	DET
cuesj-105	47	34	vector	vector	NOUN
cuesj-105	47	35	(	(	PUNCT
cuesj-105	47	36	n2	n2	ADJ
cuesj-105	47	37	)	)	PUNCT
cuesj-105	47	38	.	.	PUNCT
cuesj-105	48	1	the	the	DET
cuesj-105	48	2	step	step	NOUN
cuesj-105	48	3	after	after	SCONJ
cuesj-105	48	4	generated	generate	VERB
cuesj-105	48	5	the	the	DET
cuesj-105	48	6	training	training	NOUN
cuesj-105	48	7	set	set	NOUN
cuesj-105	48	8	is	be	AUX
cuesj-105	48	9	calculated	calculate	VERB
cuesj-105	48	10	the	the	DET
cuesj-105	48	11	mean	mean	NOUN
cuesj-105	48	12	of	of	ADP
cuesj-105	48	13	images	image	NOUN
cuesj-105	48	14	as	as	SCONJ
cuesj-105	48	15	this	this	PRON
cuesj-105	48	16	is	be	AUX
cuesj-105	48	17	described	describe	VERB
cuesj-105	48	18	in	in	ADP
cuesj-105	48	19	equation	equation	NOUN
cuesj-105	48	20	1	1	NUM
cuesj-105	48	21	:	:	PUNCT
cuesj-105	48	22	figure	figure	NOUN
cuesj-105	48	23	1	1	NUM
cuesj-105	48	24	:	:	PUNCT
cuesj-105	48	25	the	the	DET
cuesj-105	48	26	structure	structure	NOUN
cuesj-105	48	27	of	of	ADP
cuesj-105	48	28	a	a	DET
cuesj-105	48	29	proposed	propose	VERB
cuesj-105	48	30	system	system	NOUN
cuesj-105	48	31	fleah	fleah	PROPN
cuesj-105	48	32	and	and	CCONJ
cuesj-105	48	33	al	al	PROPN
cuesj-105	48	34	-	-	PUNCT
cuesj-105	48	35	aubi	aubi	PROPN
cuesj-105	48	36	:	:	PUNCT
cuesj-105	48	37	frs	frs	PROPN
cuesj-105	48	38	-	-	PUNCT
cuesj-105	48	39	pca	pca	NOUN
cuesj-105	48	40	and	and	CCONJ
cuesj-105	48	41	svm	svm	PROPN
cuesj-105	48	42	16	16	NUM
cuesj-105	48	43	http://journals.cihanuniversity.edu.iq/index.php/cuesj	http://journals.cihanuniversity.edu.iq/index.php/cuesj	ADJ
cuesj-105	48	44	cuesj	cuesj	NOUN
cuesj-105	48	45	2019	2019	NUM
cuesj-105	48	46	,	,	PUNCT
cuesj-105	48	47	3	3	NUM
cuesj-105	48	48	(	(	PUNCT
cuesj-105	48	49	2	2	NUM
cuesj-105	48	50	):	):	PUNCT
cuesj-105	48	51	14	14	NUM
cuesj-105	48	52	-	-	SYM
cuesj-105	48	53	20	20	NUM
cuesj-105	48	54	ψ	ψ	NOUN
cuesj-105	48	55	γ=	γ=	NOUN
cuesj-105	48	56	=	=	SYM
cuesj-105	48	57	∑1	∑1	NOUN
cuesj-105	48	58	1	1	NUM
cuesj-105	48	59	m	m	NOUN
cuesj-105	48	60	n	n	ADP
cuesj-105	48	61	m	m	VERB
cuesj-105	48	62	i	i	NOUN
cuesj-105	48	63	(	(	PUNCT
cuesj-105	48	64	1	1	X
cuesj-105	48	65	)	)	PUNCT
cuesj-105	48	66	γi	γi	NOUN
cuesj-105	48	67	in	in	ADP
cuesj-105	48	68	the	the	DET
cuesj-105	48	69	question	question	NOUN
cuesj-105	48	70	(	(	PUNCT
cuesj-105	48	71	1	1	X
cuesj-105	48	72	)	)	PUNCT
cuesj-105	48	73	refers	refer	VERB
cuesj-105	48	74	to	to	ADP
cuesj-105	48	75	the	the	DET
cuesj-105	48	76	image	image	NOUN
cuesj-105	48	77	in	in	ADP
cuesj-105	48	78	the	the	DET
cuesj-105	48	79	training	training	NOUN
cuesj-105	48	80	group	group	NOUN
cuesj-105	48	81	.	.	PUNCT
cuesj-105	49	1	then	then	ADV
cuesj-105	49	2	,	,	PUNCT
cuesj-105	49	3	the	the	DET
cuesj-105	49	4	normalized	normalize	VERB
cuesj-105	49	5	is	be	AUX
cuesj-105	49	6	implemented	implement	VERB
cuesj-105	49	7	in	in	ADP
cuesj-105	49	8	training	training	NOUN
cuesj-105	49	9	set	set	VERB
cuesj-105	49	10	to	to	PART
cuesj-105	49	11	achieve	achieve	VERB
cuesj-105	49	12	the	the	DET
cuesj-105	49	13	zero	zero	NUM
cuesj-105	49	14	mean	mean	NOUN
cuesj-105	49	15	that	that	PRON
cuesj-105	49	16	describes	describe	VERB
cuesj-105	49	17	the	the	DET
cuesj-105	49	18	degree	degree	NOUN
cuesj-105	49	19	of	of	ADP
cuesj-105	49	20	different	different	ADJ
cuesj-105	49	21	between	between	ADP
cuesj-105	49	22	the	the	DET
cuesj-105	49	23	mean	mean	NOUN
cuesj-105	49	24	of	of	ADP
cuesj-105	49	25	image	image	NOUN
cuesj-105	49	26	and	and	CCONJ
cuesj-105	49	27	image	image	NOUN
cuesj-105	49	28	,	,	PUNCT
cuesj-105	49	29	and	and	CCONJ
cuesj-105	49	30	this	this	PRON
cuesj-105	49	31	is	be	AUX
cuesj-105	49	32	described	describe	VERB
cuesj-105	49	33	in	in	ADP
cuesj-105	49	34	equation	equation	NOUN
cuesj-105	49	35	:	:	PUNCT
cuesj-105	49	36	ϕi	ϕi	ADP
cuesj-105	49	37	=	=	SYM
cuesj-105	49	38	γi	γi	NOUN
cuesj-105	49	39	–	–	PUNCT
cuesj-105	49	40	ψi	ψi	ADJ
cuesj-105	49	41	1	1	NUM
cuesj-105	49	42	,	,	PUNCT
cuesj-105	49	43	2	2	NUM
cuesj-105	49	44	…	…	SYM
cuesj-105	49	45	m	m	VERB
cuesj-105	49	46	(	(	PUNCT
cuesj-105	49	47	2	2	NUM
cuesj-105	49	48	)	)	PUNCT
cuesj-105	49	49	the	the	DET
cuesj-105	49	50	covariance	covariance	NOUN
cuesj-105	49	51	matrix	matrix	NOUN
cuesj-105	49	52	c	c	NOUN
cuesj-105	49	53	of	of	ADP
cuesj-105	49	54	the	the	DET
cuesj-105	49	55	training	training	NOUN
cuesj-105	49	56	set	set	NOUN
cuesj-105	49	57	can	can	AUX
cuesj-105	49	58	be	be	AUX
cuesj-105	49	59	obtained	obtain	VERB
cuesj-105	49	60	using	use	VERB
cuesj-105	49	61	equation	equation	NOUN
cuesj-105	49	62	2	2	NUM
cuesj-105	49	63	.	.	PUNCT
cuesj-105	50	1	ψ	ψ	NOUN
cuesj-105	50	2	=	=	X
cuesj-105	50	3	=	=	PUNCT
cuesj-105	50	4	∑1	∑1	NOUN
cuesj-105	50	5	1	1	NUM
cuesj-105	50	6	m	m	NOUN
cuesj-105	50	7	n	n	NOUN
cuesj-105	50	8	m	m	VERB
cuesj-105	51	1	i	i	PRON
cuesj-105	52	1	i	i	PRON
cuesj-105	52	2	t	t	VERB
cuesj-105	52	3			NOUN
cuesj-105	52	4	(	(	PUNCT
cuesj-105	52	5	3	3	NUM
cuesj-105	52	6	)	)	PUNCT
cuesj-105	52	7	i̇f	i̇f	NOUN
cuesj-105	52	8	a	a	DET
cuesj-105	52	9	vector	vector	NOUN
cuesj-105	52	10	u	u	NOUN
cuesj-105	52	11	satisfies	satisfy	VERB
cuesj-105	52	12	the	the	DET
cuesj-105	52	13	condition	condition	NOUN
cuesj-105	52	14	in	in	ADP
cuesj-105	52	15	question	question	NOUN
cuesj-105	52	16	(	(	PUNCT
cuesj-105	52	17	4	4	NUM
cuesj-105	52	18	)	)	PUNCT
cuesj-105	52	19	,	,	PUNCT
cuesj-105	52	20	the	the	DET
cuesj-105	52	21	eigenvector	eigenvector	NOUN
cuesj-105	52	22	of	of	ADP
cuesj-105	52	23	the	the	DET
cuesj-105	52	24	covariance	covariance	NOUN
cuesj-105	52	25	matrix	matrix	NOUN
cuesj-105	52	26	c	c	NOUN
cuesj-105	52	27	is	be	AUX
cuesj-105	52	28	considered	consider	VERB
cuesj-105	52	29	a	a	DET
cuesj-105	52	30	nonzero	nonzero	NOUN
cuesj-105	52	31	vector	vector	NOUN
cuesj-105	52	32	.	.	PUNCT
cuesj-105	53	1	c	c	PROPN
cuesj-105	53	2	uk	uk	PROPN
cuesj-105	53	3	=	=	SYM
cuesj-105	53	4	vk	vk	PROPN
cuesj-105	53	5	uk	uk	PROPN
cuesj-105	53	6	(	(	PUNCT
cuesj-105	53	7	4	4	NUM
cuesj-105	53	8	)	)	PUNCT
cuesj-105	53	9	where	where	SCONJ
cuesj-105	53	10	,	,	PUNCT
cuesj-105	53	11	vk	vk	NOUN
cuesj-105	53	12	refers	refer	VERB
cuesj-105	53	13	to	to	ADP
cuesj-105	53	14	the	the	DET
cuesj-105	53	15	corresponding	corresponding	ADJ
cuesj-105	53	16	eigenvalues	eigenvalue	NOUN
cuesj-105	53	17	.	.	PUNCT
cuesj-105	54	1	the	the	DET
cuesj-105	54	2	size	size	NOUN
cuesj-105	54	3	of	of	ADP
cuesj-105	54	4	matrix	matrix	NOUN
cuesj-105	54	5	c	c	NOUN
cuesj-105	54	6	with	with	ADP
cuesj-105	54	7	the	the	DET
cuesj-105	54	8	dimension	dimension	NOUN
cuesj-105	54	9	of	of	ADP
cuesj-105	54	10	n	n	NUM
cuesj-105	54	11	×	×	NOUN
cuesj-105	54	12	n	n	PRON
cuesj-105	54	13	is	be	AUX
cuesj-105	54	14	a	a	DET
cuesj-105	54	15	large	large	ADJ
cuesj-105	54	16	size	size	NOUN
cuesj-105	54	17	and	and	CCONJ
cuesj-105	54	18	needs	need	VERB
cuesj-105	54	19	much	much	ADJ
cuesj-105	54	20	computational	computational	ADJ
cuesj-105	54	21	time	time	NOUN
cuesj-105	54	22	and	and	CCONJ
cuesj-105	54	23	size	size	NOUN
cuesj-105	54	24	memory	memory	NOUN
cuesj-105	54	25	to	to	PART
cuesj-105	54	26	obtain	obtain	VERB
cuesj-105	54	27	results	result	NOUN
cuesj-105	54	28	from	from	ADP
cuesj-105	54	29	it	it	PRON
cuesj-105	54	30	,	,	PUNCT
cuesj-105	54	31	to	to	PART
cuesj-105	54	32	solve	solve	VERB
cuesj-105	54	33	this	this	DET
cuesj-105	54	34	problem	problem	NOUN
cuesj-105	54	35	,	,	PUNCT
cuesj-105	54	36	the	the	DET
cuesj-105	54	37	covariance	covariance	NOUN
cuesj-105	54	38	matrix	matrix	NOUN
cuesj-105	54	39	l	l	NOUN
cuesj-105	54	40	that	that	PRON
cuesj-105	54	41	has	have	VERB
cuesj-105	54	42	small	small	ADJ
cuesj-105	54	43	size	size	NOUN
cuesj-105	54	44	compare	compare	VERB
cuesj-105	54	45	than	than	ADP
cuesj-105	54	46	the	the	DET
cuesj-105	54	47	original	original	ADJ
cuesj-105	54	48	covariance	covariance	NOUN
cuesj-105	54	49	matrix	matrix	NOUN
cuesj-105	54	50	c	c	NOUN
cuesj-105	54	51	is	be	AUX
cuesj-105	54	52	calculated	calculate	VERB
cuesj-105	54	53	.	.	PUNCT
cuesj-105	55	1	equation	equation	NOUN
cuesj-105	55	2	5	5	NUM
cuesj-105	55	3	describes	describe	VERB
cuesj-105	55	4	how	how	SCONJ
cuesj-105	55	5	the	the	DET
cuesj-105	55	6	matrix	matrix	NOUN
cuesj-105	55	7	c	c	NOUN
cuesj-105	55	8	can	can	AUX
cuesj-105	55	9	be	be	AUX
cuesj-105	55	10	decomposed	decompose	VERB
cuesj-105	55	11	:	:	PUNCT
cuesj-105	55	12	c	c	X
cuesj-105	55	13	=	=	SYM
cuesj-105	55	14	aat	aat	X
cuesj-105	55	15	(	(	PUNCT
cuesj-105	55	16	5	5	NUM
cuesj-105	55	17	)	)	PUNCT
cuesj-105	55	18	where	where	SCONJ
cuesj-105	55	19	,	,	PUNCT
cuesj-105	55	20	a	a	PRON
cuesj-105	55	21	represents	represent	VERB
cuesj-105	55	22	zero	zero	NUM
cuesj-105	55	23	means	mean	VERB
cuesj-105	55	24	matrix	matrix	NOUN
cuesj-105	55	25	[	[	X
cuesj-105	55	26	ϕ1	ϕ1	NOUN
cuesj-105	55	27	,	,	PUNCT
cuesj-105	55	28	ϕ2	ϕ2	ADV
cuesj-105	55	29	…	…	PUNCT
cuesj-105	55	30	,	,	PUNCT
cuesj-105	55	31	ϕi	ϕi	ADP
cuesj-105	55	32	]	]	PUNCT
cuesj-105	55	33	and	and	CCONJ
cuesj-105	55	34	the	the	DET
cuesj-105	55	35	eigenvectors	eigenvector	NOUN
cuesj-105	55	36	vi	vi	PROPN
cuesj-105	55	37	can	can	AUX
cuesj-105	55	38	be	be	AUX
cuesj-105	55	39	consideration	consideration	NOUN
cuesj-105	55	40	according	accord	VERB
cuesj-105	55	41	to	to	ADP
cuesj-105	55	42	the	the	DET
cuesj-105	55	43	following	following	NOUN
cuesj-105	55	44	:	:	PUNCT
cuesj-105	56	1	aat	aat	PROPN
cuesj-105	56	2	vi	vi	NOUN
cuesj-105	57	1	=	=	PUNCT
cuesj-105	57	2	µi	µi	PROPN
cuesj-105	57	3	vi	vi	PROPN
cuesj-105	57	4	(	(	PUNCT
cuesj-105	57	5	6	6	NUM
cuesj-105	57	6	)	)	PUNCT
cuesj-105	57	7	now	now	ADV
cuesj-105	57	8	,	,	PUNCT
cuesj-105	57	9	by	by	ADP
cuesj-105	57	10	multiply	multiply	VERB
cuesj-105	57	11	each	each	DET
cuesj-105	57	12	side	side	NOUN
cuesj-105	57	13	of	of	ADP
cuesj-105	57	14	equation	equation	NOUN
cuesj-105	57	15	6	6	NUM
cuesj-105	57	16	by	by	ADP
cuesj-105	57	17	a	a	PRON
cuesj-105	57	18	,	,	PUNCT
cuesj-105	57	19	we	we	PRON
cuesj-105	57	20	can	can	AUX
cuesj-105	57	21	obtain	obtain	VERB
cuesj-105	57	22	the	the	DET
cuesj-105	57	23	following	following	NOUN
cuesj-105	57	24	:	:	PUNCT
cuesj-105	57	25	aat	aat	PROPN
cuesj-105	57	26	avi	avi	NOUN
cuesj-105	57	27	=	=	PUNCT
cuesj-105	57	28	µi	µi	ADP
cuesj-105	57	29	avi	avi	NOUN
cuesj-105	57	30	(	(	PUNCT
cuesj-105	57	31	7	7	NUM
cuesj-105	57	32	)	)	PUNCT
cuesj-105	57	33	from	from	ADP
cuesj-105	57	34	equation	equation	NOUN
cuesj-105	57	35	(	(	PUNCT
cuesj-105	57	36	7	7	NUM
cuesj-105	57	37	)	)	PUNCT
cuesj-105	57	38	,	,	PUNCT
cuesj-105	57	39	we	we	PRON
cuesj-105	57	40	can	can	AUX
cuesj-105	57	41	see	see	VERB
cuesj-105	57	42	that	that	PRON
cuesj-105	57	43	represents	represent	VERB
cuesj-105	57	44	the	the	DET
cuesj-105	57	45	eigenvectors	eigenvector	NOUN
cuesj-105	57	46	of	of	ADP
cuesj-105	57	47	the	the	DET
cuesj-105	57	48	covariance	covariance	NOUN
cuesj-105	57	49	matrix	matrix	NOUN
cuesj-105	57	50	c	c	NOUN
cuesj-105	57	51	and	and	CCONJ
cuesj-105	57	52	the	the	DET
cuesj-105	57	53	matrix	matrix	NOUN
cuesj-105	57	54	l	l	NOUN
cuesj-105	57	55	with	with	ADP
cuesj-105	57	56	dimension	dimension	NOUN
cuesj-105	57	57	(	(	PUNCT
cuesj-105	57	58	m	m	PROPN
cuesj-105	57	59	×	×	PROPN
cuesj-105	57	60	m	m	VERB
cuesj-105	57	61	)	)	PUNCT
cuesj-105	57	62	can	can	AUX
cuesj-105	57	63	be	be	AUX
cuesj-105	57	64	rewritten	rewrite	VERB
cuesj-105	57	65	as	as	SCONJ
cuesj-105	57	66	shown	show	VERB
cuesj-105	57	67	in	in	ADP
cuesj-105	57	68	equation	equation	NOUN
cuesj-105	57	69	(	(	PUNCT
cuesj-105	57	70	8)	8)	NUM
cuesj-105	57	71	:	:	PUNCT
cuesj-105	57	72	l	l	NOUN
cuesj-105	57	73	=	=	PUNCT
cuesj-105	57	74	aat	aat	X
cuesj-105	57	75	(	(	PUNCT
cuesj-105	57	76	8)	8)	NUM
cuesj-105	57	77	now	now	ADV
cuesj-105	57	78	,	,	PUNCT
cuesj-105	57	79	we	we	PRON
cuesj-105	57	80	obtained	obtain	VERB
cuesj-105	57	81	the	the	DET
cuesj-105	57	82	eigenvectors	eigenvector	NOUN
cuesj-105	57	83	vi	vi	ADV
cuesj-105	57	84	of	of	ADP
cuesj-105	57	85	smaller	small	ADJ
cuesj-105	57	86	covariance	covariance	NOUN
cuesj-105	57	87	matrix	matrix	NOUN
cuesj-105	57	88	l	l	NOUN
cuesj-105	57	89	that	that	PRON
cuesj-105	57	90	has	have	AUX
cuesj-105	57	91	dimension	dimension	NOUN
cuesj-105	57	92	m	m	PROPN
cuesj-105	57	93	and	and	CCONJ
cuesj-105	57	94	this	this	DET
cuesj-105	57	95	way	way	NOUN
cuesj-105	57	96	,	,	PUNCT
cuesj-105	57	97	the	the	DET
cuesj-105	57	98	linear	linear	ADJ
cuesj-105	57	99	combinations	combination	NOUN
cuesj-105	57	100	of	of	ADP
cuesj-105	57	101	the	the	DET
cuesj-105	57	102	training	training	NOUN
cuesj-105	57	103	data	datum	NOUN
cuesj-105	57	104	of	of	ADP
cuesj-105	57	105	the	the	DET
cuesj-105	57	106	images	image	NOUN
cuesj-105	57	107	to	to	ADP
cuesj-105	57	108	the	the	DET
cuesj-105	57	109	form	form	NOUN
cuesj-105	57	110	of	of	ADP
cuesj-105	57	111	eigenfaces	eigenface	NOUN
cuesj-105	57	112	ui	ui	NOUN
cuesj-105	57	113	can	can	AUX
cuesj-105	57	114	be	be	AUX
cuesj-105	57	115	obtained	obtain	VERB
cuesj-105	57	116	from	from	ADP
cuesj-105	57	117	equation	equation	NOUN
cuesj-105	57	118	below	below	ADV
cuesj-105	57	119	:	:	PUNCT
cuesj-105	57	120	u	u	PROPN
cuesj-105	57	121	vi	vi	PROPN
cuesj-105	58	1	k	k	X
cuesj-105	58	2	m	m	VERB
cuesj-105	58	3	ik	ik	X
cuesj-105	58	4	k=	k=	X
cuesj-105	58	5	=	=	PUNCT
cuesj-105	58	6	∑	∑	PROPN
cuesj-105	58	7	1	1	NUM
cuesj-105	58	8			NOUN
cuesj-105	58	9	(	(	PUNCT
cuesj-105	58	10	8)	8)	NUM
cuesj-105	58	11	this	this	DET
cuesj-105	58	12	eigenvector	eigenvector	NOUN
cuesj-105	58	13	that	that	PRON
cuesj-105	58	14	has	have	VERB
cuesj-105	58	15	symbol	symbol	NOUN
cuesj-105	58	16	ui	ui	PROPN
cuesj-105	58	17	is	be	AUX
cuesj-105	58	18	also	also	ADV
cuesj-105	58	19	called	call	VERB
cuesj-105	58	20	as	as	ADP
cuesj-105	58	21	eigenfaces	eigenface	NOUN
cuesj-105	58	22	because	because	SCONJ
cuesj-105	58	23	when	when	SCONJ
cuesj-105	58	24	transfer	transfer	VERB
cuesj-105	58	25	these	these	DET
cuesj-105	58	26	eigenvectors	eigenvector	NOUN
cuesj-105	58	27	form	form	VERB
cuesj-105	58	28	a	a	DET
cuesj-105	58	29	vector	vector	NOUN
cuesj-105	58	30	of	of	ADP
cuesj-105	58	31	length	length	NOUN
cuesj-105	58	32	(	(	PUNCT
cuesj-105	58	33	n	n	CCONJ
cuesj-105	58	34	)	)	PUNCT
cuesj-105	58	35	to	to	ADP
cuesj-105	58	36	a	a	DET
cuesj-105	58	37	two	two	NUM
cuesj-105	58	38	-	-	PUNCT
cuesj-105	58	39	dimensional	dimensional	ADJ
cuesj-105	58	40	matrix	matrix	NOUN
cuesj-105	58	41	with	with	ADP
cuesj-105	58	42	dimension	dimension	NOUN
cuesj-105	58	43	(	(	PUNCT
cuesj-105	58	44	n	n	CCONJ
cuesj-105	58	45	×	×	NOUN
cuesj-105	58	46	n	n	CCONJ
cuesj-105	58	47	)	)	PUNCT
cuesj-105	58	48	and	and	CCONJ
cuesj-105	58	49	show	show	VERB
cuesj-105	58	50	it	it	PRON
cuesj-105	58	51	.	.	PUNCT
cuesj-105	59	1	it	it	PRON
cuesj-105	59	2	is	be	AUX
cuesj-105	59	3	shown	show	VERB
cuesj-105	59	4	as	as	ADP
cuesj-105	59	5	the	the	DET
cuesj-105	59	6	human	human	ADJ
cuesj-105	59	7	face	face	NOUN
cuesj-105	59	8	.	.	PUNCT
cuesj-105	60	1	wavelet	wavelet	NOUN
cuesj-105	60	2	transform	transform	NOUN
cuesj-105	60	3	it	it	PRON
cuesj-105	60	4	is	be	AUX
cuesj-105	60	5	the	the	DET
cuesj-105	60	6	first	first	ADJ
cuesj-105	60	7	step	step	NOUN
cuesj-105	60	8	of	of	ADP
cuesj-105	60	9	this	this	DET
cuesj-105	60	10	stage	stage	NOUN
cuesj-105	60	11	and	and	CCONJ
cuesj-105	60	12	it	it	PRON
cuesj-105	60	13	is	be	AUX
cuesj-105	60	14	used	use	VERB
cuesj-105	60	15	wavelet	wavelet	NOUN
cuesj-105	60	16	transform	transform	NOUN
cuesj-105	60	17	to	to	PART
cuesj-105	60	18	remove	remove	VERB
cuesj-105	60	19	the	the	DET
cuesj-105	60	20	noise	noise	NOUN
cuesj-105	60	21	from	from	ADP
cuesj-105	60	22	the	the	DET
cuesj-105	60	23	image	image	NOUN
cuesj-105	60	24	with	with	ADP
cuesj-105	60	25	saving	save	VERB
cuesj-105	60	26	the	the	DET
cuesj-105	60	27	quality	quality	NOUN
cuesj-105	60	28	of	of	ADP
cuesj-105	60	29	an	an	DET
cuesj-105	60	30	image	image	NOUN
cuesj-105	60	31	.	.	PUNCT
cuesj-105	61	1	the	the	DET
cuesj-105	61	2	second	second	ADJ
cuesj-105	61	3	benefit	benefit	NOUN
cuesj-105	61	4	is	be	AUX
cuesj-105	61	5	to	to	PART
cuesj-105	61	6	obtain	obtain	VERB
cuesj-105	61	7	some	some	PRON
cuesj-105	61	8	of	of	ADP
cuesj-105	61	9	a	a	DET
cuesj-105	61	10	feature	feature	NOUN
cuesj-105	61	11	from	from	ADP
cuesj-105	61	12	the	the	DET
cuesj-105	61	13	image	image	NOUN
cuesj-105	61	14	and	and	CCONJ
cuesj-105	61	15	also	also	ADV
cuesj-105	61	16	to	to	PART
cuesj-105	61	17	decompose	decompose	VERB
cuesj-105	61	18	the	the	DET
cuesj-105	61	19	size	size	NOUN
cuesj-105	61	20	of	of	ADP
cuesj-105	61	21	the	the	DET
cuesj-105	61	22	image	image	NOUN
cuesj-105	61	23	and	and	CCONJ
cuesj-105	61	24	by	by	ADP
cuesj-105	61	25	this	this	DET
cuesj-105	61	26	way	way	NOUN
cuesj-105	61	27	,	,	PUNCT
cuesj-105	61	28	thus	thus	ADV
cuesj-105	61	29	reduced	reduce	VERB
cuesj-105	61	30	the	the	DET
cuesj-105	61	31	time	time	NOUN
cuesj-105	61	32	of	of	ADP
cuesj-105	61	33	the	the	DET
cuesj-105	61	34	process	process	NOUN
cuesj-105	61	35	.	.	PUNCT
cuesj-105	62	1	the	the	DET
cuesj-105	62	2	output	output	NOUN
cuesj-105	62	3	of	of	ADP
cuesj-105	62	4	the	the	DET
cuesj-105	62	5	wavelet	wavelet	NOUN
cuesj-105	62	6	transform	transform	NOUN
cuesj-105	62	7	on	on	ADP
cuesj-105	62	8	an	an	DET
cuesj-105	62	9	image	image	NOUN
cuesj-105	62	10	,	,	PUNCT
cuesj-105	62	11	it	it	PRON
cuesj-105	62	12	consists	consist	VERB
cuesj-105	62	13	of	of	ADP
cuesj-105	62	14	four	four	NUM
cuesj-105	62	15	sub	sub	NOUN
cuesj-105	62	16	-	-	NOUN
cuesj-105	62	17	bands	band	NOUN
cuesj-105	62	18	.	.	PUNCT
cuesj-105	63	1	these	these	DET
cuesj-105	63	2	sub	sub	NOUN
cuesj-105	63	3	-	-	ADJ
cuesj-105	63	4	bands	band	NOUN
cuesj-105	63	5	are	be	AUX
cuesj-105	63	6	the	the	DET
cuesj-105	63	7	following	follow	VERB
cuesj-105	63	8	:	:	PUNCT
cuesj-105	63	9	low	low	ADJ
cuesj-105	63	10	-	-	PUNCT
cuesj-105	63	11	low	low	ADJ
cuesj-105	63	12	(	(	PUNCT
cuesj-105	63	13	ll	ll	NOUN
cuesj-105	63	14	)	)	PUNCT
cuesj-105	63	15	,	,	PUNCT
cuesj-105	63	16	high	high	ADJ
cuesj-105	63	17	-	-	PUNCT
cuesj-105	63	18	low	low	ADJ
cuesj-105	63	19	(	(	PUNCT
cuesj-105	63	20	hl	hl	NOUN
cuesj-105	63	21	)	)	PUNCT
cuesj-105	63	22	,	,	PUNCT
cuesj-105	63	23	low	low	ADJ
cuesj-105	63	24	-	-	PUNCT
cuesj-105	63	25	high	high	ADJ
cuesj-105	63	26	(	(	PUNCT
cuesj-105	63	27	lh	lh	PROPN
cuesj-105	63	28	)	)	PUNCT
cuesj-105	63	29	,	,	PUNCT
cuesj-105	63	30	and	and	CCONJ
cuesj-105	63	31	high	high	ADJ
cuesj-105	63	32	-	-	PUNCT
cuesj-105	63	33	high	high	ADJ
cuesj-105	63	34	.	.	PUNCT
cuesj-105	64	1	the	the	DET
cuesj-105	64	2	ll	ll	PROPN
cuesj-105	64	3	sub	sub	NOUN
cuesj-105	64	4	-	-	NOUN
cuesj-105	64	5	band	band	NOUN
cuesj-105	64	6	gives	give	VERB
cuesj-105	64	7	the	the	DET
cuesj-105	64	8	most	most	ADV
cuesj-105	64	9	important	important	ADJ
cuesj-105	64	10	features	feature	NOUN
cuesj-105	64	11	of	of	ADP
cuesj-105	64	12	the	the	DET
cuesj-105	64	13	image	image	NOUN
cuesj-105	64	14	and	and	CCONJ
cuesj-105	64	15	hl	hl	NOUN
cuesj-105	64	16	and	and	CCONJ
cuesj-105	64	17	lh	lh	PROPN
cuesj-105	64	18	sub	sub	NOUN
cuesj-105	64	19	-	-	NOUN
cuesj-105	64	20	bands	band	NOUN
cuesj-105	64	21	give	give	VERB
cuesj-105	64	22	the	the	DET
cuesj-105	64	23	horizontal	horizontal	ADJ
cuesj-105	64	24	edges	edge	NOUN
cuesj-105	64	25	and	and	CCONJ
cuesj-105	64	26	vertical	vertical	ADJ
cuesj-105	64	27	edges	edge	NOUN
cuesj-105	64	28	,	,	PUNCT
cuesj-105	64	29	and	and	CCONJ
cuesj-105	64	30	finally	finally	ADV
cuesj-105	64	31	,	,	PUNCT
cuesj-105	64	32	the	the	PRON
cuesj-105	64	33	high	high	ADJ
cuesj-105	64	34	sub	sub	NOUN
cuesj-105	64	35	-	-	NOUN
cuesj-105	64	36	band	band	NOUN
cuesj-105	64	37	gives	give	VERB
cuesj-105	64	38	the	the	DET
cuesj-105	64	39	diagonal	diagonal	ADJ
cuesj-105	64	40	edges	edge	NOUN
cuesj-105	64	41	.	.	PUNCT
cuesj-105	65	1	the	the	DET
cuesj-105	65	2	results	result	NOUN
cuesj-105	65	3	that	that	PRON
cuesj-105	65	4	obtained	obtain	VERB
cuesj-105	65	5	after	after	ADP
cuesj-105	65	6	applied	apply	VERB
cuesj-105	65	7	wavelet	wavelet	NOUN
cuesj-105	65	8	transform	transform	NOUN
cuesj-105	65	9	are	be	AUX
cuesj-105	65	10	shown	show	VERB
cuesj-105	65	11	in	in	ADP
cuesj-105	65	12	figure	figure	NOUN
cuesj-105	65	13	2	2	NUM
cuesj-105	65	14	.	.	PUNCT
cuesj-105	65	15	when	when	SCONJ
cuesj-105	65	16	this	this	DET
cuesj-105	65	17	technique	technique	NOUN
cuesj-105	65	18	is	be	AUX
cuesj-105	65	19	applied	apply	VERB
cuesj-105	65	20	for	for	ADP
cuesj-105	65	21	1	1	NUM
cuesj-105	65	22	time	time	NOUN
cuesj-105	65	23	to	to	ADP
cuesj-105	65	24	the	the	DET
cuesj-105	65	25	image	image	NOUN
cuesj-105	65	26	with	with	ADP
cuesj-105	65	27	dimensions	dimension	NOUN
cuesj-105	65	28	(	(	PUNCT
cuesj-105	65	29	112	112	NUM
cuesj-105	65	30	×	×	NOUN
cuesj-105	65	31	92	92	NUM
cuesj-105	65	32	)	)	PUNCT
cuesj-105	65	33	,	,	PUNCT
cuesj-105	65	34	the	the	DET
cuesj-105	65	35	dimensions	dimension	NOUN
cuesj-105	65	36	of	of	ADP
cuesj-105	65	37	the	the	DET
cuesj-105	65	38	four	four	NUM
cuesj-105	65	39	sub	sub	NOUN
cuesj-105	65	40	-	-	ADJ
cuesj-105	65	41	bands	band	NOUN
cuesj-105	65	42	are	be	AUX
cuesj-105	65	43	56	56	NUM
cuesj-105	65	44	×	×	NOUN
cuesj-105	65	45	46	46	NUM
cuesj-105	65	46	.	.	PUNCT
cuesj-105	66	1	we	we	PRON
cuesj-105	66	2	can	can	AUX
cuesj-105	66	3	conclude	conclude	VERB
cuesj-105	66	4	from	from	ADP
cuesj-105	66	5	the	the	DET
cuesj-105	66	6	previous	previous	ADJ
cuesj-105	66	7	paragraph	paragraph	NOUN
cuesj-105	66	8	after	after	SCONJ
cuesj-105	66	9	used	use	VERB
cuesj-105	66	10	one	one	NUM
cuesj-105	66	11	stage	stage	NOUN
cuesj-105	66	12	of	of	ADP
cuesj-105	66	13	the	the	DET
cuesj-105	66	14	wavelet	wavelet	NOUN
cuesj-105	66	15	transform	transform	NOUN
cuesj-105	66	16	,	,	PUNCT
cuesj-105	66	17	the	the	DET
cuesj-105	66	18	size	size	NOUN
cuesj-105	66	19	of	of	ADP
cuesj-105	66	20	the	the	DET
cuesj-105	66	21	image	image	NOUN
cuesj-105	66	22	is	be	AUX
cuesj-105	66	23	decreased	decrease	VERB
cuesj-105	66	24	to	to	ADP
cuesj-105	66	25	the	the	DET
cuesj-105	66	26	half	half	NOUN
cuesj-105	66	27	.	.	PUNCT
cuesj-105	67	1	in	in	ADP
cuesj-105	67	2	this	this	DET
cuesj-105	67	3	work	work	NOUN
cuesj-105	67	4	,	,	PUNCT
cuesj-105	67	5	two	two	NUM
cuesj-105	67	6	levels	level	NOUN
cuesj-105	67	7	of	of	ADP
cuesj-105	67	8	the	the	DET
cuesj-105	67	9	wavelet	wavelet	NOUN
cuesj-105	67	10	transform	transform	NOUN
cuesj-105	67	11	are	be	AUX
cuesj-105	67	12	used	use	VERB
cuesj-105	67	13	.	.	PUNCT
cuesj-105	68	1	classification	classification	NOUN
cuesj-105	68	2	stage	stage	NOUN
cuesj-105	68	3	a	a	DET
cuesj-105	68	4	classification	classification	NOUN
cuesj-105	68	5	is	be	AUX
cuesj-105	68	6	the	the	DET
cuesj-105	68	7	third	third	ADJ
cuesj-105	68	8	stage	stage	NOUN
cuesj-105	68	9	of	of	ADP
cuesj-105	68	10	this	this	DET
cuesj-105	68	11	recognition	recognition	NOUN
cuesj-105	68	12	system	system	NOUN
cuesj-105	68	13	.	.	PUNCT
cuesj-105	69	1	it	it	PRON
cuesj-105	69	2	is	be	AUX
cuesj-105	69	3	used	use	VERB
cuesj-105	69	4	after	after	SCONJ
cuesj-105	69	5	each	each	PRON
cuesj-105	69	6	of	of	ADP
cuesj-105	69	7	the	the	DET
cuesj-105	69	8	previous	previous	ADJ
cuesj-105	69	9	stages	stage	NOUN
cuesj-105	69	10	is	be	AUX
cuesj-105	69	11	done	do	VERB
cuesj-105	69	12	.	.	PUNCT
cuesj-105	70	1	it	it	PRON
cuesj-105	70	2	is	be	AUX
cuesj-105	70	3	a	a	DET
cuesj-105	70	4	very	very	ADV
cuesj-105	70	5	significant	significant	ADJ
cuesj-105	70	6	stage	stage	NOUN
cuesj-105	70	7	.	.	PUNCT
cuesj-105	71	1	the	the	DET
cuesj-105	71	2	goals	goal	NOUN
cuesj-105	71	3	of	of	ADP
cuesj-105	71	4	this	this	DET
cuesj-105	71	5	stage	stage	NOUN
cuesj-105	71	6	are	be	AUX
cuesj-105	71	7	classified	classify	VERB
cuesj-105	71	8	the	the	DET
cuesj-105	71	9	interimages	interimage	NOUN
cuesj-105	71	10	based	base	VERB
cuesj-105	71	11	on	on	ADP
cuesj-105	71	12	the	the	DET
cuesj-105	71	13	information	information	NOUN
cuesj-105	71	14	that	that	PRON
cuesj-105	71	15	obtained	obtain	VERB
cuesj-105	71	16	through	through	ADP
cuesj-105	71	17	the	the	DET
cuesj-105	71	18	training	training	NOUN
cuesj-105	71	19	stage	stage	NOUN
cuesj-105	71	20	.	.	PUNCT
cuesj-105	72	1	there	there	PRON
cuesj-105	72	2	are	be	VERB
cuesj-105	72	3	a	a	DET
cuesj-105	72	4	number	number	NOUN
cuesj-105	72	5	of	of	ADP
cuesj-105	72	6	classification	classification	NOUN
cuesj-105	72	7	methods	method	NOUN
cuesj-105	72	8	that	that	PRON
cuesj-105	72	9	can	can	AUX
cuesj-105	72	10	be	be	AUX
cuesj-105	72	11	used	use	VERB
cuesj-105	72	12	to	to	PART
cuesj-105	72	13	do	do	VERB
cuesj-105	72	14	this	this	DET
cuesj-105	72	15	job	job	NOUN
cuesj-105	72	16	.	.	PUNCT
cuesj-105	73	1	in	in	ADP
cuesj-105	73	2	this	this	DET
cuesj-105	73	3	recognition	recognition	NOUN
cuesj-105	73	4	system	system	NOUN
cuesj-105	73	5	,	,	PUNCT
cuesj-105	73	6	svms	svms	NOUN
cuesj-105	73	7	that	that	PRON
cuesj-105	73	8	are	be	AUX
cuesj-105	73	9	considered	consider	VERB
cuesj-105	73	10	as	as	ADP
cuesj-105	73	11	one	one	NUM
cuesj-105	73	12	of	of	ADP
cuesj-105	73	13	the	the	DET
cuesj-105	73	14	famous	famous	ADJ
cuesj-105	73	15	classification	classification	NOUN
cuesj-105	73	16	methods	method	NOUN
cuesj-105	73	17	are	be	AUX
cuesj-105	73	18	used	use	VERB
cuesj-105	73	19	to	to	PART
cuesj-105	73	20	do	do	VERB
cuesj-105	73	21	this	this	DET
cuesj-105	73	22	function	function	NOUN
cuesj-105	73	23	.	.	PUNCT
cuesj-105	74	1	svms	svms	NOUN
cuesj-105	74	2	are	be	AUX
cuesj-105	74	3	one	one	NUM
cuesj-105	74	4	of	of	ADP
cuesj-105	74	5	the	the	DET
cuesj-105	74	6	most	most	ADV
cuesj-105	74	7	important	important	ADJ
cuesj-105	74	8	learning	learning	NOUN
cuesj-105	74	9	algorithms	algorithm	NOUN
cuesj-105	74	10	which	which	PRON
cuesj-105	74	11	are	be	AUX
cuesj-105	74	12	applied	apply	VERB
cuesj-105	74	13	in	in	ADP
cuesj-105	74	14	many	many	ADJ
cuesj-105	74	15	of	of	ADP
cuesj-105	74	16	applications	application	NOUN
cuesj-105	74	17	.	.	PUNCT
cuesj-105	75	1	in	in	ADP
cuesj-105	75	2	this	this	DET
cuesj-105	75	3	work	work	NOUN
cuesj-105	75	4	,	,	PUNCT
cuesj-105	75	5	we	we	PRON
cuesj-105	75	6	also	also	ADV
cuesj-105	75	7	aim	aim	VERB
cuesj-105	75	8	to	to	PART
cuesj-105	75	9	make	make	VERB
cuesj-105	75	10	a	a	DET
cuesj-105	75	11	comparison	comparison	NOUN
cuesj-105	75	12	between	between	ADP
cuesj-105	75	13	different	different	ADJ
cuesj-105	75	14	kinds	kind	NOUN
cuesj-105	75	15	of	of	ADP
cuesj-105	75	16	the	the	DET
cuesj-105	75	17	svms	svms	NOUN
cuesj-105	75	18	kernel	kernel	PROPN
cuesj-105	75	19	to	to	PART
cuesj-105	75	20	analyze	analyze	VERB
cuesj-105	75	21	its	its	PRON
cuesj-105	75	22	results	result	NOUN
cuesj-105	75	23	and	and	CCONJ
cuesj-105	75	24	determine	determine	VERB
cuesj-105	75	25	any	any	DET
cuesj-105	75	26	kernels	kernel	NOUN
cuesj-105	75	27	of	of	ADP
cuesj-105	75	28	them	they	PRON
cuesj-105	75	29	can	can	AUX
cuesj-105	75	30	give	give	VERB
cuesj-105	75	31	a	a	DET
cuesj-105	75	32	good	good	ADJ
cuesj-105	75	33	result	result	NOUN
cuesj-105	75	34	with	with	ADP
cuesj-105	75	35	high	high	ADJ
cuesj-105	75	36	accuracy	accuracy	NOUN
cuesj-105	75	37	rate	rate	NOUN
cuesj-105	75	38	and	and	CCONJ
cuesj-105	75	39	also	also	ADV
cuesj-105	75	40	we	we	PRON
cuesj-105	75	41	make	make	VERB
cuesj-105	75	42	a	a	DET
cuesj-105	75	43	comparison	comparison	NOUN
cuesj-105	75	44	with	with	ADP
cuesj-105	75	45	a	a	DET
cuesj-105	75	46	feedforward	feedforward	ADJ
cuesj-105	75	47	backpropagation	backpropagation	NOUN
cuesj-105	75	48	neural	neural	ADJ
cuesj-105	75	49	network	network	NOUN
cuesj-105	75	50	(	(	PUNCT
cuesj-105	75	51	ffbpnn	ffbpnn	PROPN
cuesj-105	75	52	)	)	PUNCT
cuesj-105	75	53	that	that	PRON
cuesj-105	75	54	used	use	VERB
cuesj-105	75	55	as	as	ADP
cuesj-105	75	56	a	a	DET
cuesj-105	75	57	classifier	classifier	NOUN
cuesj-105	75	58	to	to	PART
cuesj-105	75	59	ensure	ensure	VERB
cuesj-105	75	60	from	from	ADP
cuesj-105	75	61	the	the	DET
cuesj-105	75	62	power	power	NOUN
cuesj-105	75	63	of	of	ADP
cuesj-105	75	64	our	our	PRON
cuesj-105	75	65	methods	method	NOUN
cuesj-105	75	66	.	.	PUNCT
cuesj-105	76	1	svm	svm	PROPN
cuesj-105	76	2	classifier	classifier	PROPN
cuesj-105	76	3	svm	svm	PROPN
cuesj-105	76	4	is	be	AUX
cuesj-105	76	5	considering	consider	VERB
cuesj-105	76	6	as	as	ADP
cuesj-105	76	7	learning	learn	VERB
cuesj-105	76	8	algorithm	algorithm	NOUN
cuesj-105	76	9	method	method	NOUN
cuesj-105	76	10	depended	depend	VERB
cuesj-105	76	11	on	on	ADP
cuesj-105	76	12	statistical	statistical	ADJ
cuesj-105	76	13	learning	learning	NOUN
cuesj-105	76	14	algorithms	algorithm	NOUN
cuesj-105	76	15	.	.	PUNCT
cuesj-105	77	1	it	it	PRON
cuesj-105	77	2	is	be	AUX
cuesj-105	77	3	used	use	VERB
cuesj-105	77	4	to	to	PART
cuesj-105	77	5	figure	figure	VERB
cuesj-105	77	6	2	2	NUM
cuesj-105	77	7	:	:	PUNCT
cuesj-105	77	8	the	the	DET
cuesj-105	77	9	result	result	NOUN
cuesj-105	77	10	after	after	SCONJ
cuesj-105	77	11	used	use	VERB
cuesj-105	77	12	one	one	NUM
cuesj-105	77	13	level	level	NOUN
cuesj-105	77	14	of	of	ADP
cuesj-105	77	15	the	the	DET
cuesj-105	77	16	wavelet	wavelet	NOUN
cuesj-105	77	17	transform	transform	NOUN
cuesj-105	77	18	fleah	fleah	PROPN
cuesj-105	77	19	and	and	CCONJ
cuesj-105	77	20	al	al	PROPN
cuesj-105	77	21	-	-	PUNCT
cuesj-105	77	22	aubi	aubi	PROPN
cuesj-105	77	23	:	:	PUNCT
cuesj-105	77	24	frs	frs	PROPN
cuesj-105	77	25	-	-	PUNCT
cuesj-105	77	26	pca	pca	NOUN
cuesj-105	77	27	and	and	CCONJ
cuesj-105	77	28	svm	svm	PROPN
cuesj-105	77	29	17	17	NUM
cuesj-105	77	30	http://journals.cihanuniversity.edu.iq/index.php/cuesj	http://journals.cihanuniversity.edu.iq/index.php/cuesj	ADJ
cuesj-105	77	31	cuesj	cuesj	NOUN
cuesj-105	77	32	2019	2019	NUM
cuesj-105	77	33	,	,	PUNCT
cuesj-105	77	34	3	3	NUM
cuesj-105	77	35	(	(	PUNCT
cuesj-105	77	36	2	2	NUM
cuesj-105	77	37	):	):	PUNCT
cuesj-105	77	38	14	14	NUM
cuesj-105	77	39	-	-	SYM
cuesj-105	77	40	20	20	NUM
cuesj-105	77	41	classify	classify	VERB
cuesj-105	77	42	the	the	DET
cuesj-105	77	43	feature	feature	NOUN
cuesj-105	77	44	of	of	ADP
cuesj-105	77	45	the	the	DET
cuesj-105	77	46	data	datum	NOUN
cuesj-105	77	47	by	by	ADP
cuesj-105	77	48	finding	find	VERB
cuesj-105	77	49	a	a	DET
cuesj-105	77	50	maximum	maximum	ADJ
cuesj-105	77	51	marginal	marginal	ADJ
cuesj-105	77	52	hyperplan	hyperplan	NOUN
cuesj-105	77	53	that	that	SCONJ
cuesj-105	77	54	separating	separate	VERB
cuesj-105	77	55	between	between	ADP
cuesj-105	77	56	two	two	NUM
cuesj-105	77	57	classes	class	NOUN
cuesj-105	77	58	.	.	PUNCT
cuesj-105	78	1	it	it	PRON
cuesj-105	78	2	also	also	ADV
cuesj-105	78	3	can	can	AUX
cuesj-105	78	4	be	be	AUX
cuesj-105	78	5	applied	apply	VERB
cuesj-105	78	6	for	for	ADP
cuesj-105	78	7	linear	linear	ADJ
cuesj-105	78	8	and	and	CCONJ
cuesj-105	78	9	non	non	ADJ
cuesj-105	78	10	-	-	ADJ
cuesj-105	78	11	linear	linear	ADJ
cuesj-105	78	12	classification	classification	NOUN
cuesj-105	78	13	using	use	VERB
cuesj-105	78	14	different	different	ADJ
cuesj-105	78	15	types	type	NOUN
cuesj-105	78	16	of	of	ADP
cuesj-105	78	17	kernel	kernel	PROPN
cuesj-105	78	18	functions	function	NOUN
cuesj-105	78	19	.	.	PUNCT
cuesj-105	79	1	basic	basic	ADJ
cuesj-105	79	2	svm	svm	PROPN
cuesj-105	79	3	formulation	formulation	NOUN
cuesj-105	79	4	svms	svms	NOUN
cuesj-105	79	5	are	be	AUX
cuesj-105	79	6	used	use	VERB
cuesj-105	79	7	to	to	PART
cuesj-105	79	8	make	make	VERB
cuesj-105	79	9	classification	classification	NOUN
cuesj-105	79	10	among	among	ADP
cuesj-105	79	11	two	two	NUM
cuesj-105	79	12	classes	class	NOUN
cuesj-105	79	13	to	to	PART
cuesj-105	79	14	give	give	VERB
cuesj-105	79	15	a	a	DET
cuesj-105	79	16	decision	decision	NOUN
cuesj-105	79	17	surface	surface	NOUN
cuesj-105	79	18	.	.	PUNCT
cuesj-105	80	1	this	this	DET
cuesj-105	80	2	decision	decision	NOUN
cuesj-105	80	3	surface	surface	NOUN
cuesj-105	80	4	is	be	AUX
cuesj-105	80	5	represented	represent	VERB
cuesj-105	80	6	a	a	DET
cuesj-105	80	7	maximum	maximum	ADJ
cuesj-105	80	8	distance	distance	NOUN
cuesj-105	80	9	that	that	PRON
cuesj-105	80	10	can	can	AUX
cuesj-105	80	11	depend	depend	VERB
cuesj-105	80	12	on	on	ADP
cuesj-105	80	13	it	it	PRON
cuesj-105	80	14	to	to	PART
cuesj-105	80	15	separate	separate	VERB
cuesj-105	80	16	closest	close	ADJ
cuesj-105	80	17	points	point	NOUN
cuesj-105	80	18	in	in	ADP
cuesj-105	80	19	the	the	DET
cuesj-105	80	20	classification	classification	NOUN
cuesj-105	80	21	’s	’s	PART
cuesj-105	80	22	class	class	NOUN
cuesj-105	80	23	.	.	PUNCT
cuesj-105	81	1	the	the	DET
cuesj-105	81	2	closest	close	ADJ
cuesj-105	81	3	points	point	NOUN
cuesj-105	81	4	are	be	AUX
cuesj-105	81	5	known	know	VERB
cuesj-105	81	6	as	as	ADP
cuesj-105	81	7	support	support	NOUN
cuesj-105	81	8	vectors	vector	NOUN
cuesj-105	81	9	.	.	PUNCT
cuesj-105	82	1	figure	figure	NOUN
cuesj-105	82	2	3	3	NUM
cuesj-105	82	3	describes	describe	VERB
cuesj-105	82	4	the	the	DET
cuesj-105	82	5	hyperplane	hyperplane	NOUN
cuesj-105	82	6	of	of	ADP
cuesj-105	82	7	support	support	NOUN
cuesj-105	82	8	vector	vector	NOUN
cuesj-105	82	9	separating	separate	VERB
cuesj-105	82	10	.	.	PUNCT
cuesj-105	83	1	in	in	ADP
cuesj-105	83	2	this	this	DET
cuesj-105	83	3	part	part	NOUN
cuesj-105	83	4	of	of	ADP
cuesj-105	83	5	the	the	DET
cuesj-105	83	6	chapter	chapter	NOUN
cuesj-105	83	7	,	,	PUNCT
cuesj-105	83	8	we	we	PRON
cuesj-105	83	9	will	will	AUX
cuesj-105	83	10	consider	consider	VERB
cuesj-105	83	11	svms	svms	NOUN
cuesj-105	83	12	as	as	ADP
cuesj-105	83	13	maximum	maximum	ADJ
cuesj-105	83	14	margin	margin	NOUN
cuesj-105	83	15	classifiers	classifier	NOUN
cuesj-105	83	16	.	.	PUNCT
cuesj-105	84	1	to	to	PART
cuesj-105	84	2	facilitate	facilitate	VERB
cuesj-105	84	3	the	the	DET
cuesj-105	84	4	condition	condition	NOUN
cuesj-105	84	5	that	that	PRON
cuesj-105	84	6	supposed	suppose	VERB
cuesj-105	84	7	,	,	PUNCT
cuesj-105	84	8	we	we	PRON
cuesj-105	84	9	will	will	AUX
cuesj-105	84	10	consider	consider	VERB
cuesj-105	84	11	that	that	DET
cuesj-105	84	12	input	input	NOUN
cuesj-105	84	13	data	datum	NOUN
cuesj-105	84	14	can	can	AUX
cuesj-105	84	15	separate	separate	VERB
cuesj-105	84	16	linearly	linearly	ADV
cuesj-105	84	17	.	.	PUNCT
cuesj-105	85	1	under	under	ADP
cuesj-105	85	2	this	this	DET
cuesj-105	85	3	supposition	supposition	NOUN
cuesj-105	85	4	,	,	PUNCT
cuesj-105	85	5	we	we	PRON
cuesj-105	85	6	can	can	AUX
cuesj-105	85	7	separate	separate	VERB
cuesj-105	85	8	any	any	DET
cuesj-105	85	9	two	two	NUM
cuesj-105	85	10	classes	class	NOUN
cuesj-105	85	11	by	by	ADP
cuesj-105	85	12	obtained	obtain	VERB
cuesj-105	85	13	the	the	DET
cuesj-105	85	14	linear	linear	ADJ
cuesj-105	85	15	hyperplane	hyperplane	NOUN
cuesj-105	85	16	between	between	ADP
cuesj-105	85	17	them	they	PRON
cuesj-105	85	18	.	.	PUNCT
cuesj-105	86	1	figure	figure	VERB
cuesj-105	86	2	3	3	NUM
cuesj-105	86	3	describes	describe	VERB
cuesj-105	86	4	the	the	DET
cuesj-105	86	5	good	good	ADJ
cuesj-105	86	6	and	and	CCONJ
cuesj-105	86	7	bad	bad	ADJ
cuesj-105	86	8	separation	separation	NOUN
cuesj-105	86	9	between	between	ADP
cuesj-105	86	10	two	two	NUM
cuesj-105	86	11	classes	class	NOUN
cuesj-105	86	12	.	.	PUNCT
cuesj-105	87	1	let	let	VERB
cuesj-105	87	2	us	we	PRON
cuesj-105	87	3	summed	sum	VERB
cuesj-105	87	4	that	that	SCONJ
cuesj-105	87	5	,	,	PUNCT
cuesj-105	87	6	a	a	DET
cuesj-105	87	7	binary	binary	ADJ
cuesj-105	87	8	classification	classification	NOUN
cuesj-105	87	9	with	with	ADP
cuesj-105	87	10	xi	xi	PROPN
cuesj-105	87	11	as	as	ADP
cuesj-105	87	12	the	the	DET
cuesj-105	87	13	input	input	NOUN
cuesj-105	87	14	data	datum	NOUN
cuesj-105	87	15	,	,	PUNCT
cuesj-105	87	16	where	where	SCONJ
cuesj-105	87	17	i	i	PRON
cuesj-105	87	18	=	=	NOUN
cuesj-105	87	19	1	1	NUM
cuesj-105	87	20	,	,	PUNCT
cuesj-105	87	21	2	2	NUM
cuesj-105	87	22	,	,	PUNCT
cuesj-105	87	23	3	3	NUM
cuesj-105	87	24	..	..	PUNCT
cuesj-105	87	25	,	,	PUNCT
cuesj-105	87	26	,	,	PUNCT
cuesj-105	87	27	l.	l.	NOUN
cuesj-105	87	28	with	with	ADP
cuesj-105	87	29	identical	identical	ADJ
cuesj-105	87	30	labels	label	NOUN
cuesj-105	87	31	for	for	ADP
cuesj-105	87	32	the	the	DET
cuesj-105	87	33	two	two	NUM
cuesj-105	87	34	class	class	NOUN
cuesj-105	87	35	,	,	PUNCT
cuesj-105	87	36	now	now	ADV
cuesj-105	87	37	let	let	AUX
cuesj-105	87	38	assumed	assume	VERB
cuesj-105	87	39	that	that	SCONJ
cuesj-105	87	40	the	the	DET
cuesj-105	87	41	function	function	NOUN
cuesj-105	87	42	of	of	ADP
cuesj-105	87	43	decision	decision	NOUN
cuesj-105	87	44	knows	know	VERB
cuesj-105	87	45	as	as	SCONJ
cuesj-105	87	46	shown	show	VERB
cuesj-105	87	47	in	in	ADP
cuesj-105	87	48	the	the	DET
cuesj-105	87	49	question	question	NOUN
cuesj-105	87	50	below	below	ADV
cuesj-105	87	51	:	:	PUNCT
cuesj-105	88	1	f	f	PROPN
cuesj-105	88	2	(	(	PUNCT
cuesj-105	88	3	x	x	X
cuesj-105	88	4	)	)	PUNCT
cuesj-105	88	5	=	=	NOUN
cuesj-105	88	6	sign	sign	NOUN
cuesj-105	88	7	(	(	PUNCT
cuesj-105	88	8	w	w	NOUN
cuesj-105	88	9	x	x	X
cuesj-105	88	10	+	+	NUM
cuesj-105	88	11	b	b	NOUN
cuesj-105	88	12	)	)	PUNCT
cuesj-105	88	13	(	(	PUNCT
cuesj-105	88	14	9	9	NUM
cuesj-105	88	15	)	)	PUNCT
cuesj-105	88	16	where	where	SCONJ
cuesj-105	88	17	,	,	PUNCT
cuesj-105	88	18	symbol	symbol	NOUN
cuesj-105	88	19	represented	represent	VERB
cuesj-105	88	20	the	the	DET
cuesj-105	88	21	inner	inner	ADJ
cuesj-105	88	22	product	product	NOUN
cuesj-105	88	23	and	and	CCONJ
cuesj-105	88	24	symbol	symbol	NOUN
cuesj-105	88	25	b	b	PROPN
cuesj-105	88	26	represent	represent	VERB
cuesj-105	88	27	the	the	DET
cuesj-105	88	28	bias	bias	NOUN
cuesj-105	88	29	of	of	ADP
cuesj-105	88	30	the	the	DET
cuesj-105	88	31	function	function	NOUN
cuesj-105	88	32	and	and	CCONJ
cuesj-105	88	33	point	point	NOUN
cuesj-105	88	34	x	x	PUNCT
cuesj-105	88	35	take	take	VERB
cuesj-105	88	36	position	position	NOUN
cuesj-105	88	37	directly	directly	ADV
cuesj-105	88	38	on	on	ADP
cuesj-105	88	39	the	the	DET
cuesj-105	88	40	hyperplane	hyperplane	NOUN
cuesj-105	88	41	and	and	CCONJ
cuesj-105	88	42	it	it	PRON
cuesj-105	88	43	achieves	achieve	VERB
cuesj-105	88	44	the	the	DET
cuesj-105	88	45	condition	condition	NOUN
cuesj-105	88	46	.	.	PUNCT
cuesj-105	89	1	w	w	NOUN
cuesj-105	89	2	x	x	PUNCT
cuesj-105	90	1	+	+	NUM
cuesj-105	90	2	b	b	X
cuesj-105	90	3	=	=	SYM
cuesj-105	90	4	0	0	NUM
cuesj-105	90	5	(	(	PUNCT
cuesj-105	90	6	10	10	NUM
cuesj-105	90	7	)	)	PUNCT
cuesj-105	90	8	from	from	ADP
cuesj-105	90	9	calculating	calculate	VERB
cuesj-105	90	10	the	the	DET
cuesj-105	90	11	left	left	ADJ
cuesj-105	90	12	side	side	NOUN
cuesj-105	90	13	of	of	ADP
cuesj-105	90	14	equation	equation	NOUN
cuesj-105	90	15	(	(	PUNCT
cuesj-105	90	16	10	10	NUM
cuesj-105	90	17	)	)	PUNCT
cuesj-105	90	18	,	,	PUNCT
cuesj-105	90	19	we	we	PRON
cuesj-105	90	20	can	can	AUX
cuesj-105	90	21	obtain	obtain	VERB
cuesj-105	90	22	the	the	DET
cuesj-105	90	23	label	label	NOUN
cuesj-105	90	24	yi	yi	NOUN
cuesj-105	90	25	of	of	ADP
cuesj-105	90	26	a	a	DET
cuesj-105	90	27	data	datum	NOUN
cuesj-105	90	28	,	,	PUNCT
cuesj-105	90	29	where	where	SCONJ
cuesj-105	90	30	sign	sign	NOUN
cuesj-105	90	31	refers	refer	VERB
cuesj-105	90	32	to	to	ADP
cuesj-105	90	33	the	the	DET
cuesj-105	90	34	label	label	NOUN
cuesj-105	90	35	of	of	ADP
cuesj-105	90	36	class	class	NOUN
cuesj-105	90	37	.	.	PUNCT
cuesj-105	91	1	the	the	DET
cuesj-105	91	2	point	point	NOUN
cuesj-105	91	3	’s	’	VERB
cuesj-105	91	4	xi	xi	X
cuesj-105	91	5	on	on	ADP
cuesj-105	91	6	both	both	DET
cuesj-105	91	7	sides	side	NOUN
cuesj-105	91	8	of	of	ADP
cuesj-105	91	9	the	the	DET
cuesj-105	91	10	hyperplane	hyperplane	NOUN
cuesj-105	91	11	must	must	AUX
cuesj-105	91	12	achieve	achieve	VERB
cuesj-105	91	13	the	the	DET
cuesj-105	91	14	following	following	ADJ
cuesj-105	91	15	conditions	condition	NOUN
cuesj-105	91	16	:	:	PUNCT
cuesj-105	91	17	xi	xi	X
cuesj-105	91	18	w	w	PROPN
cuesj-105	92	1	+	+	NUM
cuesj-105	92	2	b	b	X
cuesj-105	92	3	<	<	X
cuesj-105	92	4	10	10	NUM
cuesj-105	92	5	(	(	PUNCT
cuesj-105	92	6	11	11	NUM
cuesj-105	92	7	)	)	PUNCT
cuesj-105	92	8	xi	xi	ADP
cuesj-105	93	1	w	w	PROPN
cuesj-105	93	2	+	+	NUM
cuesj-105	93	3	b	b	X
cuesj-105	93	4	>	>	X
cuesj-105	93	5	10	10	NUM
cuesj-105	93	6	(	(	PUNCT
cuesj-105	93	7	12	12	NUM
cuesj-105	93	8	)	)	PUNCT
cuesj-105	93	9	equations	equation	NOUN
cuesj-105	93	10	(	(	PUNCT
cuesj-105	93	11	11	11	NUM
cuesj-105	93	12	)	)	PUNCT
cuesj-105	93	13	and	and	CCONJ
cuesj-105	93	14	(	(	PUNCT
cuesj-105	93	15	12	12	NUM
cuesj-105	93	16	)	)	PUNCT
cuesj-105	93	17	can	can	AUX
cuesj-105	93	18	be	be	AUX
cuesj-105	93	19	rewritten	rewrite	VERB
cuesj-105	93	20	as	as	SCONJ
cuesj-105	93	21	follows	follow	VERB
cuesj-105	93	22	:	:	PUNCT
cuesj-105	93	23	yi	yi	PROPN
cuesj-105	93	24	(	(	PUNCT
cuesj-105	93	25	xi	xi	PROPN
cuesj-105	93	26	+	+	NUM
cuesj-105	93	27	b	b	NOUN
cuesj-105	93	28	≥	≥	NUM
cuesj-105	93	29	10	10	NUM
cuesj-105	93	30	)	)	PUNCT
cuesj-105	93	31	(	(	PUNCT
cuesj-105	93	32	13	13	NUM
cuesj-105	93	33	)	)	PUNCT
cuesj-105	93	34	using	use	VERB
cuesj-105	93	35	the	the	DET
cuesj-105	93	36	formulation	formulation	NOUN
cuesj-105	93	37	of	of	ADP
cuesj-105	93	38	lagrangian	lagrangian	ADJ
cuesj-105	93	39	,	,	PUNCT
cuesj-105	93	40	the	the	DET
cuesj-105	93	41	svms	svms	NOUN
cuesj-105	93	42	prediction	prediction	NOUN
cuesj-105	93	43	can	can	AUX
cuesj-105	93	44	obtain	obtain	VERB
cuesj-105	93	45	using	use	VERB
cuesj-105	93	46	the	the	DET
cuesj-105	93	47	following	follow	VERB
cuesj-105	93	48	equation	equation	NOUN
cuesj-105	93	49	:	:	PUNCT
cuesj-105	94	1	f	f	PROPN
cuesj-105	94	2	x	x	PUNCT
cuesj-105	94	3	y	y	NOUN
cuesj-105	94	4	x	x	X
cuesj-105	94	5	xi	xi	INTJ
cuesj-105	95	1	i	i	PRON
cuesj-105	96	1	si	si	PROPN
cuesj-105	96	2	m	m	VERB
cuesj-105	96	3	(	(	PUNCT
cuesj-105	96	4	)	)	PUNCT
cuesj-105	96	5	,	,	PUNCT
cuesj-105	96	6	�	�	X
cuesj-105	96	7	=	=	SYM
cuesj-105	96	8	<	<	X
cuesj-105	96	9	=	=	PROPN
cuesj-105	96	10	∑	∑	NOUN
cuesj-105	96	11			X
cuesj-105	96	12	0	0	NUM
cuesj-105	96	13	(	(	PUNCT
cuesj-105	96	14	14	14	NUM
cuesj-105	96	15	)	)	PUNCT
cuesj-105	96	16	where	where	SCONJ
cuesj-105	96	17	,	,	PUNCT
cuesj-105	96	18	m	m	VERB
cuesj-105	96	19	is	be	AUX
cuesj-105	96	20	represented	represent	VERB
cuesj-105	96	21	the	the	DET
cuesj-105	96	22	number	number	NOUN
cuesj-105	96	23	of	of	ADP
cuesj-105	96	24	support	support	NOUN
cuesj-105	96	25	vectors	vector	NOUN
cuesj-105	96	26	,	,	PUNCT
cuesj-105	96	27	xi	xi	X
cuesj-105	96	28	is	be	AUX
cuesj-105	96	29	a	a	DET
cuesj-105	96	30	support	support	NOUN
cuesj-105	96	31	vector	vector	NOUN
cuesj-105	96	32	,	,	PUNCT
cuesj-105	96	33	and	and	CCONJ
cuesj-105	96	34	αi	αi	PROPN
cuesj-105	96	35	is	be	AUX
cuesj-105	96	36	represented	represent	VERB
cuesj-105	96	37	the	the	DET
cuesj-105	96	38	corresponding	corresponding	ADJ
cuesj-105	96	39	lagrange	lagrange	NOUN
cuesj-105	96	40	multiplier	multiplier	ADV
cuesj-105	96	41	,	,	PUNCT
cuesj-105	96	42	and	and	CCONJ
cuesj-105	96	43	finally	finally	ADV
cuesj-105	96	44	,	,	PUNCT
cuesj-105	96	45	a	a	DET
cuesj-105	96	46	sign	sign	NOUN
cuesj-105	96	47	of	of	ADP
cuesj-105	96	48	f	f	PROPN
cuesj-105	96	49	(	(	PUNCT
cuesj-105	96	50	x	x	X
cuesj-105	96	51	)	)	PUNCT
cuesj-105	96	52	is	be	AUX
cuesj-105	96	53	classified	classify	VERB
cuesj-105	96	54	every	every	DET
cuesj-105	96	55	test	test	NOUN
cuesj-105	96	56	vector	vector	NOUN
cuesj-105	96	57	.	.	PUNCT
cuesj-105	97	1	we	we	PRON
cuesj-105	97	2	can	can	AUX
cuesj-105	97	3	say	say	VERB
cuesj-105	97	4	that	that	SCONJ
cuesj-105	97	5	the	the	DET
cuesj-105	97	6	output	output	NOUN
cuesj-105	97	7	of	of	ADP
cuesj-105	97	8	classification	classification	NOUN
cuesj-105	97	9	is	be	AUX
cuesj-105	97	10	correct	correct	ADJ
cuesj-105	97	11	if	if	SCONJ
cuesj-105	97	12	equation	equation	NOUN
cuesj-105	97	13	(	(	PUNCT
cuesj-105	97	14	14	14	NUM
cuesj-105	97	15	)	)	PUNCT
cuesj-105	97	16	holds	hold	VERB
cuesj-105	97	17	for	for	ADP
cuesj-105	97	18	all	all	DET
cuesj-105	97	19	input	input	NOUN
cuesj-105	97	20	points	point	NOUN
cuesj-105	97	21	.	.	PUNCT
cuesj-105	98	1	the	the	DET
cuesj-105	98	2	number	number	NOUN
cuesj-105	98	3	of	of	ADP
cuesj-105	98	4	possible	possible	ADJ
cuesj-105	98	5	combinations	combination	NOUN
cuesj-105	98	6	of	of	ADP
cuesj-105	98	7	both	both	DET
cuesj-105	98	8	weights	weight	NOUN
cuesj-105	98	9	and	and	CCONJ
cuesj-105	98	10	bias	bias	NOUN
cuesj-105	98	11	are	be	AUX
cuesj-105	98	12	big	big	ADJ
cuesj-105	98	13	and	and	CCONJ
cuesj-105	98	14	sometimes	sometimes	ADV
cuesj-105	98	15	not	not	PART
cuesj-105	98	16	optimal	optimal	ADJ
cuesj-105	98	17	.	.	PUNCT
cuesj-105	99	1	in	in	ADP
cuesj-105	99	2	general	general	ADJ
cuesj-105	99	3	,	,	PUNCT
cuesj-105	99	4	in	in	ADP
cuesj-105	99	5	the	the	DET
cuesj-105	99	6	binary	binary	ADJ
cuesj-105	99	7	classification	classification	NOUN
cuesj-105	99	8	,	,	PUNCT
cuesj-105	99	9	there	there	PRON
cuesj-105	99	10	is	be	VERB
cuesj-105	99	11	one	one	NUM
cuesj-105	99	12	ideal	ideal	ADJ
cuesj-105	99	13	separating	separate	VERB
cuesj-105	99	14	hyperplane	hyperplane	NOUN
cuesj-105	99	15	is	be	AUX
cuesj-105	99	16	offer	offer	NOUN
cuesj-105	99	17	.	.	PUNCT
cuesj-105	100	1	on	on	ADP
cuesj-105	100	2	the	the	DET
cuesj-105	100	3	ideal	ideal	ADJ
cuesj-105	100	4	hyperplane	hyperplane	NOUN
cuesj-105	100	5	,	,	PUNCT
cuesj-105	100	6	the	the	DET
cuesj-105	100	7	margin	margin	NOUN
cuesj-105	100	8	among	among	ADP
cuesj-105	100	9	two	two	NUM
cuesj-105	100	10	data	datum	NOUN
cuesj-105	100	11	is	be	AUX
cuesj-105	100	12	maximized	maximize	VERB
cuesj-105	100	13	.	.	PUNCT
cuesj-105	101	1	as	as	SCONJ
cuesj-105	101	2	shown	show	VERB
cuesj-105	101	3	in	in	ADP
cuesj-105	101	4	figure	figure	NOUN
cuesj-105	101	5	4	4	NUM
cuesj-105	101	6	,	,	PUNCT
cuesj-105	101	7	the	the	DET
cuesj-105	101	8	margin	margin	NOUN
cuesj-105	101	9	among	among	ADP
cuesj-105	101	10	two	two	NUM
cuesj-105	101	11	sets	set	NOUN
cuesj-105	101	12	of	of	ADP
cuesj-105	101	13	data	datum	NOUN
cuesj-105	101	14	is	be	AUX
cuesj-105	101	15	represented	represent	VERB
cuesj-105	101	16	by	by	ADP
cuesj-105	101	17	d+	d+	NOUN
cuesj-105	101	18	and	and	CCONJ
cuesj-105	101	19	d−	d−	PROPN
cuesj-105	101	20	which	which	PRON
cuesj-105	101	21	represented	represent	VERB
cuesj-105	101	22	the	the	DET
cuesj-105	101	23	distances	distance	NOUN
cuesj-105	101	24	between	between	ADP
cuesj-105	101	25	them	they	PRON
cuesj-105	101	26	.	.	PUNCT
cuesj-105	102	1	figure	figure	NOUN
cuesj-105	102	2	5	5	NUM
cuesj-105	102	3	describes	describe	VERB
cuesj-105	102	4	the	the	DET
cuesj-105	102	5	two	two	NUM
cuesj-105	102	6	separating	separate	VERB
cuesj-105	102	7	lines	line	NOUN
cuesj-105	102	8	,	,	PUNCT
cuesj-105	102	9	one	one	NUM
cuesj-105	102	10	of	of	ADP
cuesj-105	102	11	them	they	PRON
cuesj-105	102	12	is	be	AUX
cuesj-105	102	13	ideal	ideal	ADJ
cuesj-105	102	14	and	and	CCONJ
cuesj-105	102	15	another	another	PRON
cuesj-105	102	16	is	be	AUX
cuesj-105	102	17	random.[7	random.[7	NOUN
cuesj-105	102	18	]	]	PUNCT
cuesj-105	102	19	identification	identification	NOUN
cuesj-105	102	20	the	the	DET
cuesj-105	102	21	final	final	ADJ
cuesj-105	102	22	stage	stage	NOUN
cuesj-105	102	23	of	of	ADP
cuesj-105	102	24	the	the	DET
cuesj-105	102	25	system	system	NOUN
cuesj-105	102	26	is	be	AUX
cuesj-105	102	27	identification	identification	NOUN
cuesj-105	102	28	.	.	PUNCT
cuesj-105	103	1	the	the	DET
cuesj-105	103	2	aim	aim	NOUN
cuesj-105	103	3	of	of	ADP
cuesj-105	103	4	this	this	DET
cuesj-105	103	5	stage	stage	NOUN
cuesj-105	103	6	is	be	AUX
cuesj-105	103	7	to	to	PART
cuesj-105	103	8	find	find	VERB
cuesj-105	103	9	the	the	DET
cuesj-105	103	10	nearest	near	ADJ
cuesj-105	103	11	image	image	NOUN
cuesj-105	103	12	for	for	ADP
cuesj-105	103	13	the	the	DET
cuesj-105	103	14	new	new	ADJ
cuesj-105	103	15	image	image	NOUN
cuesj-105	103	16	figure	figure	NOUN
cuesj-105	103	17	4	4	NUM
cuesj-105	103	18	:	:	PUNCT
cuesj-105	103	19	the	the	DET
cuesj-105	103	20	support	support	NOUN
cuesj-105	103	21	vector	vector	NOUN
cuesj-105	103	22	and	and	CCONJ
cuesj-105	103	23	the	the	DET
cuesj-105	103	24	separating	separate	VERB
cuesj-105	103	25	hyperplane	hyperplane	NOUN
cuesj-105	103	26	figure	figure	NOUN
cuesj-105	103	27	5	5	NUM
cuesj-105	103	28	:	:	PUNCT
cuesj-105	103	29	samples	sample	NOUN
cuesj-105	103	30	of	of	ADP
cuesj-105	103	31	yale	yale	PROPN
cuesj-105	103	32	database	database	PROPN
cuesj-105	103	33	figure	figure	NOUN
cuesj-105	103	34	3	3	NUM
cuesj-105	103	35	:	:	PUNCT
cuesj-105	103	36	the	the	DET
cuesj-105	103	37	support	support	NOUN
cuesj-105	103	38	vector	vector	NOUN
cuesj-105	103	39	and	and	CCONJ
cuesj-105	103	40	the	the	DET
cuesj-105	103	41	separating	separate	VERB
cuesj-105	103	42	hyperplane	hyperplane	NOUN
cuesj-105	103	43	fleah	fleah	PROPN
cuesj-105	103	44	and	and	CCONJ
cuesj-105	103	45	al	al	PROPN
cuesj-105	103	46	-	-	PUNCT
cuesj-105	103	47	aubi	aubi	PROPN
cuesj-105	103	48	:	:	PUNCT
cuesj-105	103	49	frs	frs	PROPN
cuesj-105	103	50	-	-	PUNCT
cuesj-105	103	51	pca	pca	PROPN
cuesj-105	103	52	and	and	CCONJ
cuesj-105	103	53	svm	svm	PROPN
cuesj-105	103	54	18	18	NUM
cuesj-105	103	55	http://journals.cihanuniversity.edu.iq/index.php/cuesj	http://journals.cihanuniversity.edu.iq/index.php/cuesj	ADJ
cuesj-105	103	56	cuesj	cuesj	NOUN
cuesj-105	103	57	2019	2019	NUM
cuesj-105	103	58	,	,	PUNCT
cuesj-105	103	59	3	3	NUM
cuesj-105	103	60	(	(	PUNCT
cuesj-105	103	61	2	2	NUM
cuesj-105	103	62	):	):	PUNCT
cuesj-105	103	63	14	14	NUM
cuesj-105	103	64	-	-	SYM
cuesj-105	103	65	20	20	NUM
cuesj-105	103	66	that	that	PRON
cuesj-105	103	67	classified	classify	VERB
cuesj-105	103	68	before	before	ADP
cuesj-105	103	69	that	that	PRON
cuesj-105	103	70	using	use	VERB
cuesj-105	103	71	svm	svm	PROPN
cuesj-105	103	72	or	or	CCONJ
cuesj-105	103	73	ann	ann	PROPN
cuesj-105	103	74	.	.	PUNCT
cuesj-105	104	1	knn	knn	PROPN
cuesj-105	104	2	is	be	AUX
cuesj-105	104	3	used	use	VERB
cuesj-105	104	4	to	to	ADP
cuesj-105	104	5	implementing	implement	VERB
cuesj-105	104	6	that	that	PRON
cuesj-105	104	7	knn	knn	PROPN
cuesj-105	104	8	is	be	AUX
cuesj-105	104	9	not	not	PART
cuesj-105	104	10	like	like	ADP
cuesj-105	104	11	svm	svm	PROPN
cuesj-105	104	12	or	or	CCONJ
cuesj-105	104	13	ann	ann	PROPN
cuesj-105	104	14	.	.	PUNCT
cuesj-105	105	1	it	it	PRON
cuesj-105	105	2	does	do	AUX
cuesj-105	105	3	not	not	PART
cuesj-105	105	4	need	need	VERB
cuesj-105	105	5	to	to	ADP
cuesj-105	105	6	the	the	DET
cuesj-105	105	7	training	training	NOUN
cuesj-105	105	8	phase	phase	NOUN
cuesj-105	105	9	,	,	PUNCT
cuesj-105	105	10	and	and	CCONJ
cuesj-105	105	11	therefore	therefore	ADV
cuesj-105	105	12	,	,	PUNCT
cuesj-105	105	13	all	all	DET
cuesj-105	105	14	the	the	DET
cuesj-105	105	15	data	datum	NOUN
cuesj-105	105	16	are	be	AUX
cuesj-105	105	17	needed	need	VERB
cuesj-105	105	18	for	for	ADP
cuesj-105	105	19	the	the	DET
cuesj-105	105	20	testing	testing	NOUN
cuesj-105	105	21	phase	phase	NOUN
cuesj-105	105	22	.	.	PUNCT
cuesj-105	106	1	the	the	DET
cuesj-105	106	2	data	datum	NOUN
cuesj-105	106	3	that	that	PRON
cuesj-105	106	4	need	need	VERB
cuesj-105	106	5	for	for	ADP
cuesj-105	106	6	testing	testing	NOUN
cuesj-105	106	7	consist	consist	NOUN
cuesj-105	106	8	of	of	ADP
cuesj-105	106	9	a	a	DET
cuesj-105	106	10	set	set	NOUN
cuesj-105	106	11	of	of	ADP
cuesj-105	106	12	vectors	vector	NOUN
cuesj-105	106	13	and	and	CCONJ
cuesj-105	106	14	class	class	NOUN
cuesj-105	106	15	label	label	NOUN
cuesj-105	106	16	associated	associate	VERB
cuesj-105	106	17	with	with	ADP
cuesj-105	106	18	each	each	DET
cuesj-105	106	19	vector	vector	NOUN
cuesj-105	106	20	only	only	ADV
cuesj-105	106	21	.	.	PUNCT
cuesj-105	107	1	knn	knn	VERB
cuesj-105	107	2	for	for	ADP
cuesj-105	107	3	classification	classification	NOUN
cuesj-105	107	4	in	in	ADP
cuesj-105	107	5	the	the	DET
cuesj-105	107	6	section	section	NOUN
cuesj-105	107	7	,	,	PUNCT
cuesj-105	107	8	we	we	PRON
cuesj-105	107	9	will	will	AUX
cuesj-105	107	10	see	see	VERB
cuesj-105	107	11	how	how	SCONJ
cuesj-105	107	12	can	can	AUX
cuesj-105	107	13	be	be	AUX
cuesj-105	107	14	used	use	VERB
cuesj-105	107	15	knn	knn	PROPN
cuesj-105	107	16	as	as	ADP
cuesj-105	107	17	a	a	DET
cuesj-105	107	18	classifier	classifier	NOUN
cuesj-105	107	19	.	.	PUNCT
cuesj-105	108	1	to	to	PART
cuesj-105	108	2	do	do	VERB
cuesj-105	108	3	that	that	PRON
cuesj-105	108	4	,	,	PUNCT
cuesj-105	108	5	two	two	NUM
cuesj-105	108	6	data	data	NOUN
cuesj-105	108	7	sets	set	NOUN
cuesj-105	108	8	will	will	AUX
cuesj-105	108	9	create	create	VERB
cuesj-105	108	10	.	.	PUNCT
cuesj-105	109	1	one	one	NUM
cuesj-105	109	2	of	of	ADP
cuesj-105	109	3	them	they	PRON
cuesj-105	109	4	is	be	AUX
cuesj-105	109	5	training	train	VERB
cuesj-105	109	6	set	set	VERB
cuesj-105	109	7	and	and	CCONJ
cuesj-105	109	8	another	another	DET
cuesj-105	109	9	set	set	NOUN
cuesj-105	109	10	is	be	AUX
cuesj-105	109	11	testing	test	VERB
cuesj-105	109	12	set	set	VERB
cuesj-105	109	13	.	.	PUNCT
cuesj-105	110	1	the	the	DET
cuesj-105	110	2	goal	goal	NOUN
cuesj-105	110	3	of	of	ADP
cuesj-105	110	4	this	this	DET
cuesj-105	110	5	algorithm	algorithm	NOUN
cuesj-105	110	6	is	be	AUX
cuesj-105	110	7	to	to	PART
cuesj-105	110	8	obtain	obtain	VERB
cuesj-105	110	9	the	the	DET
cuesj-105	110	10	class	class	NOUN
cuesj-105	110	11	label	label	NOUN
cuesj-105	110	12	of	of	ADP
cuesj-105	110	13	the	the	DET
cuesj-105	110	14	new	new	ADJ
cuesj-105	110	15	point	point	NOUN
cuesj-105	110	16	.	.	PUNCT
cuesj-105	111	1	the	the	DET
cuesj-105	111	2	algorithm	algorithm	NOUN
cuesj-105	111	3	has	have	VERB
cuesj-105	111	4	various	various	ADJ
cuesj-105	111	5	conducts	conduct	NOUN
cuesj-105	111	6	depended	depend	VERB
cuesj-105	111	7	on	on	ADP
cuesj-105	111	8	the	the	DET
cuesj-105	111	9	value	value	NOUN
cuesj-105	111	10	of	of	ADP
cuesj-105	111	11	k	k	PROPN
cuesj-105	111	12	parameter	parameter	NOUN
cuesj-105	111	13	.	.	PUNCT
cuesj-105	112	1	the	the	DET
cuesj-105	112	2	first	first	ADJ
cuesj-105	112	3	hypothesis	hypothesis	NOUN
cuesj-105	112	4	:	:	PUNCT
cuesj-105	112	5	the	the	DET
cuesj-105	112	6	value	value	NOUN
cuesj-105	112	7	of	of	ADP
cuesj-105	112	8	k	k	PROPN
cuesj-105	112	9	parameter	parameter	PROPN
cuesj-105	112	10	is	be	AUX
cuesj-105	112	11	equal	equal	ADJ
cuesj-105	112	12	to	to	ADP
cuesj-105	112	13	one	one	NUM
cuesj-105	112	14	(	(	PUNCT
cuesj-105	112	15	k=1	k=1	NOUN
cuesj-105	112	16	)	)	PUNCT
cuesj-105	112	17	to	to	PART
cuesj-105	112	18	explain	explain	VERB
cuesj-105	112	19	the	the	DET
cuesj-105	112	20	first	first	ADJ
cuesj-105	112	21	hypothesis	hypothesis	NOUN
cuesj-105	112	22	,	,	PUNCT
cuesj-105	112	23	let	let	AUX
cuesj-105	112	24	consider	consider	VERB
cuesj-105	112	25	that	that	PRON
cuesj-105	112	26	x	x	PRON
cuesj-105	112	27	is	be	AUX
cuesj-105	112	28	the	the	DET
cuesj-105	112	29	point	point	NOUN
cuesj-105	112	30	to	to	PART
cuesj-105	112	31	be	be	AUX
cuesj-105	112	32	labeled	label	VERB
cuesj-105	112	33	and	and	CCONJ
cuesj-105	112	34	to	to	PART
cuesj-105	112	35	obtain	obtain	VERB
cuesj-105	112	36	the	the	DET
cuesj-105	112	37	point	point	NOUN
cuesj-105	112	38	that	that	SCONJ
cuesj-105	112	39	closest	close	ADJ
cuesj-105	112	40	to	to	ADP
cuesj-105	112	41	it	it	PRON
cuesj-105	112	42	.	.	PUNCT
cuesj-105	113	1	let	let	AUX
cuesj-105	113	2	consider	consider	VERB
cuesj-105	113	3	the	the	DET
cuesj-105	113	4	closest	close	ADJ
cuesj-105	113	5	point	point	NOUN
cuesj-105	113	6	y.	y.	NOUN
cuesj-105	113	7	in	in	ADP
cuesj-105	113	8	this	this	DET
cuesj-105	113	9	way	way	NOUN
cuesj-105	113	10	,	,	PUNCT
cuesj-105	113	11	the	the	DET
cuesj-105	113	12	nn	nn	PROPN
cuesj-105	113	13	rule	rule	NOUN
cuesj-105	113	14	asks	ask	VERB
cuesj-105	113	15	to	to	PART
cuesj-105	113	16	assign	assign	VERB
cuesj-105	113	17	the	the	DET
cuesj-105	113	18	label	label	NOUN
cuesj-105	113	19	of	of	ADP
cuesj-105	113	20	y	y	PRON
cuesj-105	113	21	to	to	PART
cuesj-105	113	22	x.	x.	VERB
cuesj-105	113	23	this	this	PRON
cuesj-105	113	24	appears	appear	VERB
cuesj-105	113	25	very	very	ADV
cuesj-105	113	26	simplistic	simplistic	ADJ
cuesj-105	113	27	,	,	PUNCT
cuesj-105	113	28	but	but	CCONJ
cuesj-105	113	29	this	this	DET
cuesj-105	113	30	way	way	NOUN
cuesj-105	113	31	is	be	AUX
cuesj-105	113	32	correct	correct	ADJ
cuesj-105	113	33	only	only	ADV
cuesj-105	113	34	when	when	SCONJ
cuesj-105	113	35	the	the	DET
cuesj-105	113	36	number	number	NOUN
cuesj-105	113	37	of	of	ADP
cuesj-105	113	38	points	point	NOUN
cuesj-105	113	39	is	be	AUX
cuesj-105	113	40	not	not	PART
cuesj-105	113	41	very	very	ADV
cuesj-105	113	42	large	large	ADJ
cuesj-105	113	43	.	.	PUNCT
cuesj-105	114	1	however	however	ADV
cuesj-105	114	2	,	,	PUNCT
cuesj-105	114	3	in	in	ADP
cuesj-105	114	4	other	other	ADJ
cuesj-105	114	5	situations	situation	NOUN
cuesj-105	114	6	,	,	PUNCT
cuesj-105	114	7	when	when	SCONJ
cuesj-105	114	8	the	the	DET
cuesj-105	114	9	number	number	NOUN
cuesj-105	114	10	of	of	ADP
cuesj-105	114	11	data	datum	NOUN
cuesj-105	114	12	points	point	NOUN
cuesj-105	114	13	is	be	AUX
cuesj-105	114	14	very	very	ADV
cuesj-105	114	15	big	big	ADJ
cuesj-105	114	16	.	.	PUNCT
cuesj-105	115	1	the	the	DET
cuesj-105	115	2	label	label	NOUN
cuesj-105	115	3	of	of	ADP
cuesj-105	115	4	x	x	PUNCT
cuesj-105	115	5	and	and	CCONJ
cuesj-105	115	6	y	y	PROPN
cuesj-105	115	7	may	may	AUX
cuesj-105	115	8	be	be	AUX
cuesj-105	115	9	same	same	ADJ
cuesj-105	115	10	.	.	PUNCT
cuesj-105	116	1	to	to	PART
cuesj-105	116	2	simplest	simplest	VERB
cuesj-105	116	3	the	the	DET
cuesj-105	116	4	explain	explain	NOUN
cuesj-105	116	5	–	–	PUNCT
cuesj-105	116	6	let	let	AUX
cuesj-105	116	7	assume	assume	VERB
cuesj-105	116	8	that	that	SCONJ
cuesj-105	116	9	the	the	DET
cuesj-105	116	10	potentially	potentially	ADV
cuesj-105	116	11	biased	biased	ADJ
cuesj-105	116	12	coin	coin	NOUN
cuesj-105	116	13	.	.	PUNCT
cuesj-105	117	1	then	then	ADV
cuesj-105	117	2	,	,	PUNCT
cuesj-105	117	3	maybe	maybe	ADV
cuesj-105	117	4	the	the	DET
cuesj-105	117	5	head	head	NOUN
cuesj-105	117	6	will	will	AUX
cuesj-105	117	7	appear	appear	VERB
cuesj-105	117	8	in	in	ADP
cuesj-105	117	9	next	next	ADJ
cuesj-105	117	10	call	call	NOUN
cuesj-105	117	11	.	.	PUNCT
cuesj-105	118	1	the	the	DET
cuesj-105	118	2	same	same	ADJ
cuesj-105	118	3	argument	argument	NOUN
cuesj-105	118	4	can	can	AUX
cuesj-105	118	5	be	be	AUX
cuesj-105	118	6	applied	apply	VERB
cuesj-105	118	7	in	in	ADP
cuesj-105	118	8	this	this	DET
cuesj-105	118	9	situation	situation	NOUN
cuesj-105	118	10	.	.	PUNCT
cuesj-105	119	1	now	now	ADV
cuesj-105	119	2	,	,	PUNCT
cuesj-105	119	3	lets	let	VERB
cuesj-105	119	4	us	we	PRON
cuesj-105	119	5	apply	apply	VERB
cuesj-105	119	6	another	another	DET
cuesj-105	119	7	case	case	NOUN
cuesj-105	119	8	here	here	ADV
cuesj-105	119	9	–	–	PUNCT
cuesj-105	119	10	suppose	suppose	VERB
cuesj-105	119	11	that	that	SCONJ
cuesj-105	119	12	every	every	DET
cuesj-105	119	13	point	point	NOUN
cuesj-105	119	14	is	be	AUX
cuesj-105	119	15	represented	represent	VERB
cuesj-105	119	16	in	in	ADP
cuesj-105	119	17	a	a	DET
cuesj-105	119	18	d	d	ADJ
cuesj-105	119	19	-	-	ADJ
cuesj-105	119	20	dimensional	dimensional	ADJ
cuesj-105	119	21	plane	plane	NOUN
cuesj-105	119	22	and	and	CCONJ
cuesj-105	119	23	the	the	DET
cuesj-105	119	24	number	number	NOUN
cuesj-105	119	25	of	of	ADP
cuesj-105	119	26	points	point	NOUN
cuesj-105	119	27	is	be	AUX
cuesj-105	119	28	very	very	ADV
cuesj-105	119	29	huge	huge	ADJ
cuesj-105	119	30	.	.	PUNCT
cuesj-105	120	1	this	this	PRON
cuesj-105	120	2	led	lead	VERB
cuesj-105	120	3	to	to	PART
cuesj-105	120	4	conclude	conclude	VERB
cuesj-105	120	5	that	that	SCONJ
cuesj-105	120	6	the	the	DET
cuesj-105	120	7	density	density	NOUN
cuesj-105	120	8	of	of	ADP
cuesj-105	120	9	the	the	DET
cuesj-105	120	10	plane	plane	NOUN
cuesj-105	120	11	on	on	ADP
cuesj-105	120	12	each	each	DET
cuesj-105	120	13	point	point	NOUN
cuesj-105	120	14	of	of	ADP
cuesj-105	120	15	data	datum	NOUN
cuesj-105	120	16	is	be	AUX
cuesj-105	120	17	high	high	ADJ
cuesj-105	120	18	.	.	PUNCT
cuesj-105	121	1	in	in	ADP
cuesj-105	121	2	other	other	ADJ
cuesj-105	121	3	description	description	NOUN
cuesj-105	121	4	,	,	PUNCT
cuesj-105	121	5	in	in	ADP
cuesj-105	121	6	each	each	DET
cuesj-105	121	7	subspace	subspace	NOUN
cuesj-105	121	8	,	,	PUNCT
cuesj-105	121	9	a	a	DET
cuesj-105	121	10	large	large	ADJ
cuesj-105	121	11	number	number	NOUN
cuesj-105	121	12	of	of	ADP
cuesj-105	121	13	points	point	NOUN
cuesj-105	121	14	is	be	AUX
cuesj-105	121	15	offered	offer	VERB
cuesj-105	121	16	and	and	CCONJ
cuesj-105	121	17	consider	consider	VERB
cuesj-105	121	18	x	x	NOUN
cuesj-105	121	19	point	point	NOUN
cuesj-105	121	20	is	be	AUX
cuesj-105	121	21	in	in	ADP
cuesj-105	121	22	the	the	DET
cuesj-105	121	23	subspace	subspace	NOUN
cuesj-105	121	24	and	and	CCONJ
cuesj-105	121	25	is	be	AUX
cuesj-105	121	26	many	many	ADJ
cuesj-105	121	27	of	of	ADP
cuesj-105	121	28	neighbors	neighbor	NOUN
cuesj-105	121	29	,	,	PUNCT
cuesj-105	121	30	but	but	CCONJ
cuesj-105	121	31	in	in	ADP
cuesj-105	121	32	this	this	DET
cuesj-105	121	33	time	time	NOUN
cuesj-105	121	34	,	,	PUNCT
cuesj-105	121	35	consider	consider	VERB
cuesj-105	121	36	y	y	PRON
cuesj-105	121	37	be	be	AUX
cuesj-105	121	38	the	the	DET
cuesj-105	121	39	nn	nn	NOUN
cuesj-105	121	40	to	to	ADP
cuesj-105	121	41	the	the	DET
cuesj-105	121	42	point	point	NOUN
cuesj-105	121	43	x.	x.	NOUN
cuesj-105	122	1	we	we	PRON
cuesj-105	122	2	can	can	AUX
cuesj-105	122	3	say	say	VERB
cuesj-105	122	4	that	that	SCONJ
cuesj-105	122	5	,	,	PUNCT
cuesj-105	122	6	if	if	SCONJ
cuesj-105	122	7	x	x	PRON
cuesj-105	122	8	point	point	NOUN
cuesj-105	122	9	and	and	CCONJ
cuesj-105	122	10	y	y	PROPN
cuesj-105	122	11	point	point	NOUN
cuesj-105	122	12	are	be	AUX
cuesj-105	122	13	very	very	ADV
cuesj-105	122	14	close	close	ADJ
cuesj-105	122	15	together	together	ADV
cuesj-105	122	16	,	,	PUNCT
cuesj-105	122	17	then	then	ADV
cuesj-105	122	18	the	the	DET
cuesj-105	122	19	following	following	NOUN
cuesj-105	122	20	can	can	AUX
cuesj-105	122	21	propose	propose	VERB
cuesj-105	122	22	,	,	PUNCT
cuesj-105	122	23	the	the	DET
cuesj-105	122	24	probability	probability	NOUN
cuesj-105	122	25	of	of	ADP
cuesj-105	122	26	point	point	NOUN
cuesj-105	122	27	x	x	PUNCT
cuesj-105	122	28	and	and	CCONJ
cuesj-105	122	29	point	point	NOUN
cuesj-105	122	30	x	x	PUNCT
cuesj-105	122	31	related	relate	VERB
cuesj-105	122	32	to	to	ADP
cuesj-105	122	33	the	the	DET
cuesj-105	122	34	one	one	NUM
cuesj-105	122	35	class	class	NOUN
cuesj-105	122	36	is	be	AUX
cuesj-105	122	37	very	very	ADV
cuesj-105	122	38	big	big	ADJ
cuesj-105	122	39	,	,	PUNCT
cuesj-105	122	40	then	then	ADV
cuesj-105	122	41	using	use	VERB
cuesj-105	122	42	decision	decision	NOUN
cuesj-105	122	43	theory	theory	NOUN
cuesj-105	122	44	,	,	PUNCT
cuesj-105	122	45	the	the	DET
cuesj-105	122	46	following	follow	VERB
cuesj-105	122	47	is	be	AUX
cuesj-105	122	48	obtained	obtain	VERB
cuesj-105	122	49	,	,	PUNCT
cuesj-105	122	50	both	both	DET
cuesj-105	122	51	points	point	NOUN
cuesj-105	122	52	related	relate	VERB
cuesj-105	122	53	to	to	ADP
cuesj-105	122	54	the	the	DET
cuesj-105	122	55	one	one	NUM
cuesj-105	122	56	class	class	NOUN
cuesj-105	122	57	.	.	PUNCT
cuesj-105	123	1	bellhumer	bellhumer	PROPN
cuesj-105	123	2	et	et	PROPN
cuesj-105	123	3	al.[8	al.[8	PROPN
cuesj-105	123	4	]	]	PUNCT
cuesj-105	123	5	provided	provide	VERB
cuesj-105	123	6	a	a	DET
cuesj-105	123	7	very	very	ADV
cuesj-105	123	8	good	good	ADJ
cuesj-105	123	9	argument	argument	NOUN
cuesj-105	123	10	about	about	ADP
cuesj-105	123	11	the	the	DET
cuesj-105	123	12	rule	rule	NOUN
cuesj-105	123	13	of	of	ADP
cuesj-105	123	14	nn	nn	PROPN
cuesj-105	123	15	.	.	PROPN
cuesj-105	123	16	one	one	NUM
cuesj-105	123	17	of	of	ADP
cuesj-105	123	18	the	the	DET
cuesj-105	123	19	important	important	ADJ
cuesj-105	123	20	results	result	NOUN
cuesj-105	123	21	is	be	AUX
cuesj-105	123	22	to	to	PART
cuesj-105	123	23	obtain	obtain	VERB
cuesj-105	123	24	a	a	DET
cuesj-105	123	25	very	very	ADV
cuesj-105	123	26	small	small	ADJ
cuesj-105	123	27	error	error	NOUN
cuesj-105	123	28	bound	bind	VERB
cuesj-105	123	29	.	.	PUNCT
cuesj-105	124	1	this	this	DET
cuesj-105	124	2	bound	bind	VERB
cuesj-105	124	3	is	be	AUX
cuesj-105	124	4	described	describe	VERB
cuesj-105	124	5	in	in	ADP
cuesj-105	124	6	question	question	NOUN
cuesj-105	124	7	:	:	PUNCT
cuesj-105	124	8	*	*	PUNCT
cuesj-105	125	1	*	*	PUNCT
cuesj-105	125	2	(	(	PUNCT
cuesj-105	125	3	2	2	NUM
cuesj-105	125	4	*	*	PUNCT
cuesj-105	125	5	)	)	PUNCT
cuesj-105	125	6	1	1	NUM
cuesj-105	125	7	c	c	NOUN
cuesj-105	125	8	p	p	NOUN
cuesj-105	125	9	p	p	X
cuesj-105	125	10	p	p	X
cuesj-105	125	11	p	p	X
cuesj-105	125	12	c	c	NOUN
cuesj-105	125	13	≤	≤	NUM
cuesj-105	125	14	≤	≤	NUM
cuesj-105	125	15			PROPN
cuesj-105	125	16			NOUN
cuesj-105	125	17	(	(	PUNCT
cuesj-105	125	18	15	15	NUM
cuesj-105	125	19	)	)	PUNCT
cuesj-105	125	20	where	where	SCONJ
cuesj-105	125	21	the	the	DET
cuesj-105	125	22	bayes	bayes	NOUN
cuesj-105	125	23	error	error	NOUN
cuesj-105	125	24	rate	rate	NOUN
cuesj-105	125	25	,	,	PUNCT
cuesj-105	125	26	the	the	DET
cuesj-105	125	27	symbol	symbol	NOUN
cuesj-105	125	28	c	c	NOUN
cuesj-105	125	29	is	be	AUX
cuesj-105	125	30	represented	represent	VERB
cuesj-105	125	31	the	the	DET
cuesj-105	125	32	number	number	NOUN
cuesj-105	125	33	of	of	ADP
cuesj-105	125	34	classes	class	NOUN
cuesj-105	125	35	and	and	CCONJ
cuesj-105	125	36	the	the	DET
cuesj-105	125	37	symbol	symbol	NOUN
cuesj-105	125	38	p	p	NOUN
cuesj-105	125	39	is	be	AUX
cuesj-105	125	40	represented	represent	VERB
cuesj-105	125	41	the	the	DET
cuesj-105	125	42	error	error	NOUN
cuesj-105	125	43	rate	rate	NOUN
cuesj-105	125	44	of	of	ADP
cuesj-105	125	45	nn	nn	PROPN
cuesj-105	125	46	.	.	PUNCT
cuesj-105	125	47	the	the	DET
cuesj-105	125	48	result	result	NOUN
cuesj-105	125	49	is	be	AUX
cuesj-105	125	50	given	give	VERB
cuesj-105	125	51	the	the	DET
cuesj-105	125	52	following	follow	VERB
cuesj-105	125	53	conclusion	conclusion	NOUN
cuesj-105	125	54	if	if	SCONJ
cuesj-105	125	55	the	the	DET
cuesj-105	125	56	number	number	NOUN
cuesj-105	125	57	of	of	ADP
cuesj-105	125	58	points	point	NOUN
cuesj-105	125	59	very	very	ADV
cuesj-105	125	60	huge	huge	ADJ
cuesj-105	125	61	,	,	PUNCT
cuesj-105	125	62	then	then	ADV
cuesj-105	125	63	the	the	DET
cuesj-105	125	64	error	error	NOUN
cuesj-105	125	65	rate	rate	NOUN
cuesj-105	125	66	of	of	ADP
cuesj-105	125	67	nn	nn	X
cuesj-105	125	68	becomes	become	VERB
cuesj-105	125	69	less	less	ADJ
cuesj-105	125	70	than	than	ADP
cuesj-105	125	71	twice	twice	ADJ
cuesj-105	125	72	time	time	NOUN
cuesj-105	125	73	from	from	ADP
cuesj-105	125	74	the	the	DET
cuesj-105	125	75	bayes	bayes	NOUN
cuesj-105	125	76	error	error	NOUN
cuesj-105	125	77	rate	rate	NOUN
cuesj-105	125	78	and	and	CCONJ
cuesj-105	125	79	this	this	PRON
cuesj-105	125	80	is	be	AUX
cuesj-105	125	81	very	very	ADV
cuesj-105	125	82	good	good	ADJ
cuesj-105	125	83	results	result	NOUN
cuesj-105	125	84	obtained	obtain	VERB
cuesj-105	125	85	from	from	ADP
cuesj-105	125	86	algorithm	algorithm	NOUN
cuesj-105	125	87	as	as	ADP
cuesj-105	125	88	knn	knn	PROPN
cuesj-105	125	89	.	.	PUNCT
cuesj-105	126	1	the	the	DET
cuesj-105	126	2	second	second	ADJ
cuesj-105	126	3	hypothesis	hypothesis	NOUN
cuesj-105	126	4	:	:	PUNCT
cuesj-105	126	5	the	the	DET
cuesj-105	126	6	value	value	NOUN
cuesj-105	126	7	of	of	ADP
cuesj-105	126	8	k	k	PROPN
cuesj-105	126	9	parameter	parameter	PROPN
cuesj-105	126	10	is	be	AUX
cuesj-105	126	11	equal	equal	ADJ
cuesj-105	126	12	to	to	ADP
cuesj-105	126	13	two	two	NUM
cuesj-105	126	14	(	(	PUNCT
cuesj-105	126	15	k=2	k=2	PROPN
cuesj-105	126	16	)	)	PUNCT
cuesj-105	126	17	in	in	ADP
cuesj-105	126	18	general	general	ADJ
cuesj-105	126	19	,	,	PUNCT
cuesj-105	126	20	we	we	PRON
cuesj-105	126	21	aim	aim	VERB
cuesj-105	126	22	to	to	PART
cuesj-105	126	23	obtain	obtain	VERB
cuesj-105	126	24	the	the	DET
cuesj-105	126	25	knn	knn	NOUN
cuesj-105	126	26	and	and	CCONJ
cuesj-105	126	27	then	then	ADV
cuesj-105	126	28	make	make	VERB
cuesj-105	126	29	a	a	DET
cuesj-105	126	30	majority	majority	NOUN
cuesj-105	126	31	voting	voting	NOUN
cuesj-105	126	32	.	.	PUNCT
cuesj-105	127	1	as	as	SCONJ
cuesj-105	127	2	we	we	PRON
cuesj-105	127	3	mentioned	mention	VERB
cuesj-105	127	4	previously	previously	ADV
cuesj-105	127	5	,	,	PUNCT
cuesj-105	127	6	the	the	DET
cuesj-105	127	7	value	value	NOUN
cuesj-105	127	8	of	of	ADP
cuesj-105	127	9	k	k	PROPN
cuesj-105	127	10	is	be	AUX
cuesj-105	127	11	odd	odd	ADJ
cuesj-105	127	12	when	when	SCONJ
cuesj-105	127	13	there	there	PRON
cuesj-105	127	14	are	be	VERB
cuesj-105	127	15	two	two	NUM
cuesj-105	127	16	numbers	number	NOUN
cuesj-105	127	17	of	of	ADP
cuesj-105	127	18	classes	class	NOUN
cuesj-105	127	19	.	.	PUNCT
cuesj-105	128	1	let	let	VERB
cuesj-105	128	2	’s	’s	PROPN
cuesj-105	128	3	assumed	assume	VERB
cuesj-105	128	4	that	that	SCONJ
cuesj-105	128	5	k	k	PROPN
cuesj-105	128	6	is	be	AUX
cuesj-105	128	7	equal	equal	ADJ
cuesj-105	128	8	5	5	NUM
cuesj-105	128	9	and	and	CCONJ
cuesj-105	128	10	there	there	PRON
cuesj-105	128	11	are	be	VERB
cuesj-105	128	12	three	three	NUM
cuesj-105	128	13	instances	instance	NOUN
cuesj-105	128	14	of	of	ADP
cuesj-105	128	15	c1	c1	NOUN
cuesj-105	128	16	and	and	CCONJ
cuesj-105	128	17	two	two	NUM
cuesj-105	128	18	instances	instance	NOUN
cuesj-105	128	19	of	of	ADP
cuesj-105	128	20	c2	c2	PROPN
cuesj-105	128	21	.	.	PUNCT
cuesj-105	129	1	in	in	ADP
cuesj-105	129	2	this	this	DET
cuesj-105	129	3	situation	situation	NOUN
cuesj-105	129	4	,	,	PUNCT
cuesj-105	129	5	knn	knn	PROPN
cuesj-105	129	6	method	method	PROPN
cuesj-105	129	7	will	will	AUX
cuesj-105	129	8	label	label	VERB
cuesj-105	129	9	the	the	DET
cuesj-105	129	10	new	new	ADJ
cuesj-105	129	11	point	point	NOUN
cuesj-105	129	12	as	as	ADP
cuesj-105	129	13	c1	c1	PROPN
cuesj-105	129	14	.	.	PUNCT
cuesj-105	130	1	this	this	PRON
cuesj-105	130	2	is	be	AUX
cuesj-105	130	3	similar	similar	ADJ
cuesj-105	130	4	when	when	SCONJ
cuesj-105	130	5	there	there	PRON
cuesj-105	130	6	is	be	VERB
cuesj-105	130	7	multiclass	multiclass	ADJ
cuesj-105	130	8	want	want	NOUN
cuesj-105	130	9	to	to	PART
cuesj-105	130	10	label	label	VERB
cuesj-105	130	11	.	.	PUNCT
cuesj-105	131	1	one	one	NUM
cuesj-105	131	2	of	of	ADP
cuesj-105	131	3	the	the	DET
cuesj-105	131	4	knn	knn	PROPN
cuesj-105	131	5	techniques	technique	NOUN
cuesj-105	131	6	is	be	AUX
cuesj-105	131	7	not	not	PART
cuesj-105	131	8	given	give	VERB
cuesj-105	131	9	1	1	NUM
cuesj-105	131	10	to	to	ADP
cuesj-105	131	11	every	every	DET
cuesj-105	131	12	neighbor	neighbor	NOUN
cuesj-105	131	13	and	and	CCONJ
cuesj-105	131	14	another	another	DET
cuesj-105	131	15	common	common	ADJ
cuesj-105	131	16	technique	technique	NOUN
cuesj-105	131	17	is	be	AUX
cuesj-105	131	18	give	give	VERB
cuesj-105	131	19	weight	weight	NOUN
cuesj-105	131	20	to	to	ADP
cuesj-105	131	21	each	each	DET
cuesj-105	131	22	point	point	NOUN
cuesj-105	131	23	according	accord	VERB
cuesj-105	131	24	to	to	ADP
cuesj-105	131	25	the	the	DET
cuesj-105	131	26	distance	distance	NOUN
cuesj-105	131	27	that	that	PRON
cuesj-105	131	28	calculated	calculate	VERB
cuesj-105	131	29	.	.	PUNCT
cuesj-105	132	1	in	in	ADP
cuesj-105	132	2	another	another	DET
cuesj-105	132	3	way	way	NOUN
cuesj-105	132	4	,	,	PUNCT
cuesj-105	132	5	the	the	DET
cuesj-105	132	6	values	value	NOUN
cuesj-105	132	7	of	of	ADP
cuesj-105	132	8	each	each	DET
cuesj-105	132	9	weight	weight	NOUN
cuesj-105	132	10	of	of	ADP
cuesj-105	132	11	point	point	NOUN
cuesj-105	132	12	are	be	AUX
cuesj-105	132	13	inversely	inversely	ADV
cuesj-105	132	14	proportional	proportional	ADJ
cuesj-105	132	15	to	to	ADP
cuesj-105	132	16	the	the	DET
cuesj-105	132	17	distance	distance	NOUN
cuesj-105	132	18	of	of	ADP
cuesj-105	132	19	the	the	DET
cuesj-105	132	20	point	point	NOUN
cuesj-105	132	21	that	that	PRON
cuesj-105	132	22	will	will	AUX
cuesj-105	132	23	classify	classify	VERB
cuesj-105	132	24	.	.	PUNCT
cuesj-105	133	1	from	from	ADP
cuesj-105	133	2	the	the	DET
cuesj-105	133	3	sentence	sentence	NOUN
cuesj-105	133	4	above	above	ADV
cuesj-105	133	5	,	,	PUNCT
cuesj-105	133	6	the	the	DET
cuesj-105	133	7	neighboring	neighboring	NOUN
cuesj-105	133	8	points	point	NOUN
cuesj-105	133	9	have	have	VERB
cuesj-105	133	10	higher	high	ADJ
cuesj-105	133	11	values	value	NOUN
cuesj-105	133	12	compared	compare	VERB
cuesj-105	133	13	than	than	ADP
cuesj-105	133	14	the	the	DET
cuesj-105	133	15	ultimate	ultimate	ADJ
cuesj-105	133	16	points	point	NOUN
cuesj-105	133	17	.	.	PUNCT
cuesj-105	134	1	experimental	experimental	ADJ
cuesj-105	134	2	results	result	VERB
cuesj-105	134	3	a	a	DET
cuesj-105	134	4	number	number	NOUN
cuesj-105	134	5	of	of	ADP
cuesj-105	134	6	experiments	experiment	NOUN
cuesj-105	134	7	on	on	ADP
cuesj-105	134	8	the	the	DET
cuesj-105	134	9	two	two	NUM
cuesj-105	134	10	different	different	ADJ
cuesj-105	134	11	databases	database	NOUN
cuesj-105	134	12	are	be	AUX
cuesj-105	134	13	used	use	VERB
cuesj-105	134	14	.	.	PUNCT
cuesj-105	135	1	the	the	DET
cuesj-105	135	2	first	first	ADJ
cuesj-105	135	3	experiment	experiment	NOUN
cuesj-105	135	4	is	be	AUX
cuesj-105	135	5	applied	apply	VERB
cuesj-105	135	6	on	on	ADP
cuesj-105	135	7	the	the	DET
cuesj-105	135	8	yale	yale	PROPN
cuesj-105	135	9	database[8	database[8	PROPN
cuesj-105	135	10	]	]	PUNCT
cuesj-105	135	11	and	and	CCONJ
cuesj-105	135	12	the	the	DET
cuesj-105	135	13	second	second	ADJ
cuesj-105	135	14	experiment	experiment	NOUN
cuesj-105	135	15	is	be	AUX
cuesj-105	135	16	applied	apply	VERB
cuesj-105	135	17	on	on	ADP
cuesj-105	135	18	the	the	DET
cuesj-105	135	19	orl	orl	PROPN
cuesj-105	135	20	database.[9	database.[9	X
cuesj-105	135	21	]	]	X
cuesj-105	135	22	both	both	CCONJ
cuesj-105	135	23	these	these	DET
cuesj-105	135	24	database	database	NOUN
cuesj-105	135	25	are	be	AUX
cuesj-105	135	26	very	very	ADV
cuesj-105	135	27	famous	famous	ADJ
cuesj-105	135	28	databases	database	NOUN
cuesj-105	135	29	and	and	CCONJ
cuesj-105	135	30	are	be	AUX
cuesj-105	135	31	used	use	VERB
cuesj-105	135	32	by	by	ADP
cuesj-105	135	33	many	many	ADJ
cuesj-105	135	34	researchers	researcher	NOUN
cuesj-105	135	35	.	.	PUNCT
cuesj-105	136	1	in	in	ADP
cuesj-105	136	2	every	every	DET
cuesj-105	136	3	database	database	NOUN
cuesj-105	136	4	,	,	PUNCT
cuesj-105	136	5	two	two	NUM
cuesj-105	136	6	experiments	experiment	NOUN
cuesj-105	136	7	are	be	AUX
cuesj-105	136	8	done	do	VERB
cuesj-105	136	9	.	.	PUNCT
cuesj-105	137	1	the	the	DET
cuesj-105	137	2	first	first	ADJ
cuesj-105	137	3	experiment	experiment	NOUN
cuesj-105	137	4	is	be	AUX
cuesj-105	137	5	compared	compare	VERB
cuesj-105	137	6	between	between	ADP
cuesj-105	137	7	different	different	ADJ
cuesj-105	137	8	svms	svms	NOUN
cuesj-105	137	9	kernels	kernel	NOUN
cuesj-105	137	10	to	to	PART
cuesj-105	137	11	show	show	VERB
cuesj-105	137	12	any	any	DET
cuesj-105	137	13	kernel	kernel	NOUN
cuesj-105	137	14	give	give	VERB
cuesj-105	137	15	best	good	ADJ
cuesj-105	137	16	results	result	NOUN
cuesj-105	137	17	and	and	CCONJ
cuesj-105	137	18	the	the	DET
cuesj-105	137	19	second	second	ADJ
cuesj-105	137	20	experiment	experiment	NOUN
cuesj-105	137	21	is	be	AUX
cuesj-105	137	22	between	between	ADP
cuesj-105	137	23	svm	svm	ADJ
cuesj-105	137	24	classifier	classifier	NOUN
cuesj-105	137	25	and	and	CCONJ
cuesj-105	137	26	neural	neural	ADJ
cuesj-105	137	27	network	network	NOUN
cuesj-105	137	28	classifier	classifier	NOUN
cuesj-105	137	29	to	to	PART
cuesj-105	137	30	calculate	calculate	VERB
cuesj-105	137	31	the	the	DET
cuesj-105	137	32	performance	performance	NOUN
cuesj-105	137	33	of	of	ADP
cuesj-105	137	34	the	the	DET
cuesj-105	137	35	classifier	classifier	NOUN
cuesj-105	137	36	that	that	PRON
cuesj-105	137	37	used	use	VERB
cuesj-105	137	38	the	the	DET
cuesj-105	137	39	system	system	NOUN
cuesj-105	137	40	.	.	PUNCT
cuesj-105	138	1	yale	yale	PROPN
cuesj-105	138	2	experiment	experiment	PROPN
cuesj-105	138	3	yale	yale	PROPN
cuesj-105	138	4	database	database	PROPN
cuesj-105	138	5	is	be	AUX
cuesj-105	138	6	used	use	VERB
cuesj-105	138	7	in	in	ADP
cuesj-105	138	8	this	this	DET
cuesj-105	138	9	experiment	experiment	NOUN
cuesj-105	138	10	.	.	PUNCT
cuesj-105	139	1	this	this	DET
cuesj-105	139	2	database	database	NOUN
cuesj-105	139	3	is	be	AUX
cuesj-105	139	4	created	create	VERB
cuesj-105	139	5	by	by	ADP
cuesj-105	139	6	yale	yale	PROPN
cuesj-105	139	7	university	university	PROPN
cuesj-105	139	8	and	and	CCONJ
cuesj-105	139	9	it	it	PRON
cuesj-105	139	10	is	be	AUX
cuesj-105	139	11	provided	provide	VERB
cuesj-105	139	12	free	free	ADJ
cuesj-105	139	13	on	on	ADP
cuesj-105	139	14	the	the	DET
cuesj-105	139	15	internet	internet	NOUN
cuesj-105	139	16	and	and	CCONJ
cuesj-105	139	17	anyone	anyone	PRON
cuesj-105	139	18	can	can	AUX
cuesj-105	139	19	download	download	VERB
cuesj-105	139	20	it	it	PRON
cuesj-105	139	21	.	.	PUNCT
cuesj-105	140	1	this	this	DET
cuesj-105	140	2	database	database	NOUN
cuesj-105	140	3	contains	contain	VERB
cuesj-105	140	4	from	from	ADP
cuesj-105	140	5	15	15	NUM
cuesj-105	140	6	subjects	subject	NOUN
cuesj-105	140	7	and	and	CCONJ
cuesj-105	140	8	each	each	DET
cuesj-105	140	9	subject	subject	NOUN
cuesj-105	140	10	has	have	VERB
cuesj-105	140	11	11	11	NUM
cuesj-105	140	12	images	image	NOUN
cuesj-105	140	13	take	take	VERB
cuesj-105	140	14	under	under	ADP
cuesj-105	140	15	various	various	ADJ
cuesj-105	140	16	conditions	condition	NOUN
cuesj-105	140	17	.	.	PUNCT
cuesj-105	141	1	from	from	ADP
cuesj-105	141	2	that	that	DET
cuesj-105	141	3	mention	mention	NOUN
cuesj-105	141	4	above	above	ADV
cuesj-105	141	5	,	,	PUNCT
cuesj-105	141	6	this	this	PRON
cuesj-105	141	7	means	mean	VERB
cuesj-105	141	8	that	that	SCONJ
cuesj-105	141	9	the	the	DET
cuesj-105	141	10	database	database	NOUN
cuesj-105	141	11	consists	consist	VERB
cuesj-105	141	12	of	of	ADP
cuesj-105	141	13	165	165	NUM
cuesj-105	141	14	.	.	PUNCT
cuesj-105	142	1	this	this	DET
cuesj-105	142	2	database	database	NOUN
cuesj-105	142	3	is	be	AUX
cuesj-105	142	4	applied	apply	VERB
cuesj-105	142	5	because	because	SCONJ
cuesj-105	142	6	it	it	PRON
cuesj-105	142	7	consists	consist	VERB
cuesj-105	142	8	of	of	ADP
cuesj-105	142	9	a	a	DET
cuesj-105	142	10	number	number	NOUN
cuesj-105	142	11	of	of	ADP
cuesj-105	142	12	images	image	NOUN
cuesj-105	142	13	captured	capture	VERB
cuesj-105	142	14	at	at	ADP
cuesj-105	142	15	various	various	ADJ
cuesj-105	142	16	conditions	condition	NOUN
cuesj-105	142	17	;	;	PUNCT
cuesj-105	142	18	this	this	PRON
cuesj-105	142	19	is	be	AUX
cuesj-105	142	20	included	include	VERB
cuesj-105	142	21	lighting	lighting	NOUN
cuesj-105	142	22	condition	condition	NOUN
cuesj-105	142	23	or	or	CCONJ
cuesj-105	142	24	other	other	ADJ
cuesj-105	142	25	things	thing	NOUN
cuesj-105	142	26	;	;	PUNCT
cuesj-105	142	27	therefore	therefore	ADV
cuesj-105	142	28	,	,	PUNCT
cuesj-105	142	29	it	it	PRON
cuesj-105	142	30	can	can	AUX
cuesj-105	142	31	be	be	AUX
cuesj-105	142	32	applied	apply	VERB
cuesj-105	142	33	to	to	PART
cuesj-105	142	34	test	test	VERB
cuesj-105	142	35	the	the	DET
cuesj-105	142	36	recognition	recognition	NOUN
cuesj-105	142	37	rate	rate	NOUN
cuesj-105	142	38	of	of	ADP
cuesj-105	142	39	the	the	DET
cuesj-105	142	40	proposed	propose	VERB
cuesj-105	142	41	system	system	NOUN
cuesj-105	142	42	and	and	CCONJ
cuesj-105	142	43	it	it	PRON
cuesj-105	142	44	is	be	AUX
cuesj-105	142	45	also	also	ADV
cuesj-105	142	46	very	very	ADV
cuesj-105	142	47	widely	widely	ADV
cuesj-105	142	48	used	use	VERB
cuesj-105	142	49	by	by	ADP
cuesj-105	142	50	many	many	ADJ
cuesj-105	142	51	of	of	ADP
cuesj-105	142	52	researchers	researcher	NOUN
cuesj-105	142	53	.	.	PUNCT
cuesj-105	143	1	in	in	ADP
cuesj-105	143	2	our	our	PRON
cuesj-105	143	3	experiment	experiment	NOUN
cuesj-105	143	4	,	,	PUNCT
cuesj-105	143	5	the	the	DET
cuesj-105	143	6	database	database	NOUN
cuesj-105	143	7	is	be	AUX
cuesj-105	143	8	divided	divide	VERB
cuesj-105	143	9	into	into	ADP
cuesj-105	143	10	two	two	NUM
cuesj-105	143	11	groups	group	NOUN
cuesj-105	143	12	;	;	PUNCT
cuesj-105	143	13	the	the	DET
cuesj-105	143	14	first	first	ADJ
cuesj-105	143	15	group	group	NOUN
cuesj-105	143	16	is	be	AUX
cuesj-105	143	17	for	for	ADP
cuesj-105	143	18	train	train	NOUN
cuesj-105	143	19	and	and	CCONJ
cuesj-105	143	20	the	the	DET
cuesj-105	143	21	second	second	ADJ
cuesj-105	143	22	group	group	NOUN
cuesj-105	143	23	is	be	AUX
cuesj-105	143	24	for	for	ADP
cuesj-105	143	25	the	the	DET
cuesj-105	143	26	test	test	NOUN
cuesj-105	143	27	.	.	PUNCT
cuesj-105	144	1	the	the	DET
cuesj-105	144	2	training	training	NOUN
cuesj-105	144	3	group	group	NOUN
cuesj-105	144	4	is	be	AUX
cuesj-105	144	5	applied	apply	VERB
cuesj-105	144	6	to	to	PART
cuesj-105	144	7	learn	learn	VERB
cuesj-105	144	8	the	the	DET
cuesj-105	144	9	system	system	NOUN
cuesj-105	144	10	and	and	CCONJ
cuesj-105	144	11	the	the	DET
cuesj-105	144	12	training	training	NOUN
cuesj-105	144	13	group	group	NOUN
cuesj-105	144	14	consists	consist	VERB
cuesj-105	144	15	of	of	ADP
cuesj-105	144	16	five	five	NUM
cuesj-105	144	17	images	image	NOUN
cuesj-105	144	18	of	of	ADP
cuesj-105	144	19	the	the	DET
cuesj-105	144	20	first	first	ADJ
cuesj-105	144	21	12	12	NUM
cuesj-105	144	22	classes	class	NOUN
cuesj-105	144	23	.	.	PUNCT
cuesj-105	145	1	the	the	DET
cuesj-105	145	2	testing	testing	NOUN
cuesj-105	145	3	group	group	NOUN
cuesj-105	145	4	is	be	AUX
cuesj-105	145	5	applied	apply	VERB
cuesj-105	145	6	to	to	PART
cuesj-105	145	7	calculate	calculate	VERB
cuesj-105	145	8	recognition	recognition	NOUN
cuesj-105	145	9	rate	rate	NOUN
cuesj-105	145	10	of	of	ADP
cuesj-105	145	11	the	the	DET
cuesj-105	145	12	proposed	propose	VERB
cuesj-105	145	13	system	system	NOUN
cuesj-105	145	14	and	and	CCONJ
cuesj-105	145	15	it	it	PRON
cuesj-105	145	16	consists	consist	VERB
cuesj-105	145	17	of	of	ADP
cuesj-105	145	18	the	the	DET
cuesj-105	145	19	other	other	ADJ
cuesj-105	145	20	five	five	NUM
cuesj-105	145	21	images	image	NOUN
cuesj-105	145	22	of	of	ADP
cuesj-105	145	23	the	the	DET
cuesj-105	145	24	same	same	ADJ
cuesj-105	145	25	12	12	NUM
cuesj-105	145	26	classes	class	NOUN
cuesj-105	145	27	as	as	ADV
cuesj-105	145	28	well	well	ADV
cuesj-105	145	29	as	as	ADP
cuesj-105	145	30	all	all	DET
cuesj-105	145	31	images	image	NOUN
cuesj-105	145	32	of	of	ADP
cuesj-105	145	33	the	the	DET
cuesj-105	145	34	last	last	ADJ
cuesj-105	145	35	three	three	NUM
cuesj-105	145	36	classes	class	NOUN
cuesj-105	145	37	.	.	PUNCT
cuesj-105	146	1	this	this	PRON
cuesj-105	146	2	is	be	AUX
cuesj-105	146	3	meant	mean	VERB
cuesj-105	146	4	that	that	SCONJ
cuesj-105	146	5	40	40	NUM
cuesj-105	146	6	%	%	NOUN
cuesj-105	146	7	of	of	ADP
cuesj-105	146	8	the	the	DET
cuesj-105	146	9	database	database	NOUN
cuesj-105	146	10	is	be	AUX
cuesj-105	146	11	applied	apply	VERB
cuesj-105	146	12	for	for	ADP
cuesj-105	146	13	training	training	NOUN
cuesj-105	146	14	and	and	CCONJ
cuesj-105	146	15	60	60	NUM
cuesj-105	146	16	%	%	NOUN
cuesj-105	146	17	of	of	ADP
cuesj-105	146	18	the	the	DET
cuesj-105	146	19	image	image	NOUN
cuesj-105	146	20	is	be	AUX
cuesj-105	146	21	applied	apply	VERB
cuesj-105	146	22	for	for	ADP
cuesj-105	146	23	the	the	DET
cuesj-105	146	24	test	test	NOUN
cuesj-105	146	25	.	.	PUNCT
cuesj-105	147	1	in	in	ADP
cuesj-105	147	2	this	this	DET
cuesj-105	147	3	experiment	experiment	NOUN
cuesj-105	147	4	,	,	PUNCT
cuesj-105	147	5	the	the	DET
cuesj-105	147	6	combinations	combination	NOUN
cuesj-105	147	7	of	of	ADP
cuesj-105	147	8	the	the	DET
cuesj-105	147	9	pca	pca	NOUN
cuesj-105	147	10	-	-	PUNCT
cuesj-105	147	11	wavelet	wavelet	NOUN
cuesj-105	147	12	transform	transform	NOUN
cuesj-105	147	13	are	be	AUX
cuesj-105	147	14	applied	apply	VERB
cuesj-105	147	15	for	for	ADP
cuesj-105	147	16	pre	pre	ADJ
cuesj-105	147	17	-	-	ADJ
cuesj-105	147	18	processing	processing	ADJ
cuesj-105	147	19	and	and	CCONJ
cuesj-105	147	20	feature	feature	NOUN
cuesj-105	147	21	extraction	extraction	NOUN
cuesj-105	147	22	,	,	PUNCT
cuesj-105	147	23	and	and	CCONJ
cuesj-105	147	24	comparison	comparison	NOUN
cuesj-105	147	25	between	between	ADP
cuesj-105	147	26	various	various	ADJ
cuesj-105	147	27	types	type	NOUN
cuesj-105	147	28	of	of	ADP
cuesj-105	147	29	the	the	DET
cuesj-105	147	30	svms	svms	NOUN
cuesj-105	147	31	kernel	kernel	PROPN
cuesj-105	147	32	is	be	AUX
cuesj-105	147	33	done	do	VERB
cuesj-105	147	34	.	.	PUNCT
cuesj-105	148	1	figure	figure	NOUN
cuesj-105	148	2	6	6	NUM
cuesj-105	148	3	describes	describe	VERB
cuesj-105	148	4	the	the	DET
cuesj-105	148	5	results	result	NOUN
cuesj-105	148	6	of	of	ADP
cuesj-105	148	7	the	the	DET
cuesj-105	148	8	svms	svms	NOUN
cuesj-105	148	9	classifier	classifier	NOUN
cuesj-105	148	10	using	use	VERB
cuesj-105	148	11	various	various	ADJ
cuesj-105	148	12	kernel	kernel	NOUN
cuesj-105	148	13	functions	function	NOUN
cuesj-105	148	14	in	in	ADP
cuesj-105	148	15	yale	yale	PROPN
cuesj-105	148	16	database	database	NOUN
cuesj-105	148	17	.	.	PUNCT
cuesj-105	149	1	in	in	ADP
cuesj-105	149	2	the	the	DET
cuesj-105	149	3	second	second	ADJ
cuesj-105	149	4	experiment	experiment	NOUN
cuesj-105	149	5	,	,	PUNCT
cuesj-105	149	6	ffbpnn	ffbpnn	PROPN
cuesj-105	149	7	is	be	AUX
cuesj-105	149	8	applied	apply	VERB
cuesj-105	149	9	as	as	ADP
cuesj-105	149	10	classifier	classifier	NOUN
cuesj-105	149	11	instead	instead	ADV
cuesj-105	149	12	of	of	ADP
cuesj-105	149	13	the	the	DET
cuesj-105	149	14	svms	svms	NOUN
cuesj-105	149	15	.	.	PUNCT
cuesj-105	150	1	this	this	PRON
cuesj-105	150	2	is	be	AUX
cuesj-105	150	3	done	do	VERB
cuesj-105	150	4	to	to	PART
cuesj-105	150	5	make	make	VERB
cuesj-105	150	6	a	a	DET
cuesj-105	150	7	comparison	comparison	NOUN
cuesj-105	150	8	between	between	ADP
cuesj-105	150	9	the	the	DET
cuesj-105	150	10	results	result	NOUN
cuesj-105	150	11	of	of	ADP
cuesj-105	150	12	svms	svms	NOUN
cuesj-105	150	13	and	and	CCONJ
cuesj-105	150	14	anns	anns	NOUN
cuesj-105	150	15	classifiers	classifier	NOUN
cuesj-105	150	16	as	as	ADV
cuesj-105	150	17	well	well	ADV
cuesj-105	150	18	as	as	ADP
cuesj-105	150	19	to	to	PART
cuesj-105	150	20	calculate	calculate	VERB
cuesj-105	150	21	the	the	DET
cuesj-105	150	22	performance	performance	NOUN
cuesj-105	150	23	of	of	ADP
cuesj-105	150	24	the	the	DET
cuesj-105	150	25	system	system	NOUN
cuesj-105	150	26	.	.	PUNCT
cuesj-105	151	1	figure	figure	NOUN
cuesj-105	151	2	7	7	NUM
cuesj-105	151	3	describes	describe	VERB
cuesj-105	151	4	the	the	DET
cuesj-105	151	5	recognition	recognition	NOUN
cuesj-105	151	6	rate	rate	NOUN
cuesj-105	151	7	results	result	NOUN
cuesj-105	151	8	of	of	ADP
cuesj-105	151	9	the	the	DET
cuesj-105	151	10	svms	svms	NOUN
cuesj-105	151	11	classifier	classifier	NOUN
cuesj-105	151	12	using	use	VERB
cuesj-105	151	13	polynomial	polynomial	ADJ
cuesj-105	151	14	kernel	kernel	NOUN
cuesj-105	151	15	function	function	NOUN
cuesj-105	151	16	and	and	CCONJ
cuesj-105	151	17	artificial	artificial	ADJ
cuesj-105	151	18	neural	neural	ADJ
cuesj-105	151	19	network	network	NOUN
cuesj-105	151	20	.	.	PUNCT
cuesj-105	152	1	orl	orl	PROPN
cuesj-105	152	2	experiment	experiment	NOUN
cuesj-105	152	3	orl	orl	PROPN
cuesj-105	152	4	database	database	PROPN
cuesj-105	152	5	is	be	AUX
cuesj-105	152	6	used	use	VERB
cuesj-105	152	7	in	in	ADP
cuesj-105	152	8	the	the	DET
cuesj-105	152	9	second	second	ADJ
cuesj-105	152	10	experiment	experiment	NOUN
cuesj-105	152	11	.	.	PUNCT
cuesj-105	153	1	this	this	DET
cuesj-105	153	2	database	database	NOUN
cuesj-105	153	3	is	be	AUX
cuesj-105	153	4	created	create	VERB
cuesj-105	153	5	by	by	ADP
cuesj-105	153	6	cambridge	cambridge	PROPN
cuesj-105	153	7	university	university	PROPN
cuesj-105	153	8	computer	computer	NOUN
cuesj-105	153	9	and	and	CCONJ
cuesj-105	153	10	anyone	anyone	PRON
cuesj-105	153	11	can	can	AUX
cuesj-105	153	12	download	download	VERB
cuesj-105	153	13	it	it	PRON
cuesj-105	153	14	from	from	ADP
cuesj-105	153	15	the	the	DET
cuesj-105	153	16	internet	internet	NOUN
cuesj-105	153	17	without	without	ADP
cuesj-105	153	18	fees	fee	NOUN
cuesj-105	153	19	.	.	PUNCT
cuesj-105	154	1	this	this	DET
cuesj-105	154	2	database	database	NOUN
cuesj-105	154	3	fleah	fleah	PROPN
cuesj-105	154	4	and	and	CCONJ
cuesj-105	154	5	al	al	PROPN
cuesj-105	154	6	-	-	PUNCT
cuesj-105	154	7	aubi	aubi	PROPN
cuesj-105	154	8	:	:	PUNCT
cuesj-105	154	9	frs	frs	PROPN
cuesj-105	154	10	-	-	PUNCT
cuesj-105	154	11	pca	pca	PROPN
cuesj-105	154	12	and	and	CCONJ
cuesj-105	154	13	svm	svm	ADJ
cuesj-105	154	14	19	19	NUM
cuesj-105	154	15	http://journals.cihanuniversity.edu.iq/index.php/cuesj	http://journals.cihanuniversity.edu.iq/index.php/cuesj	ADJ
cuesj-105	154	16	cuesj	cuesj	NOUN
cuesj-105	154	17	2019	2019	NUM
cuesj-105	154	18	,	,	PUNCT
cuesj-105	154	19	3	3	NUM
cuesj-105	154	20	(	(	PUNCT
cuesj-105	154	21	2	2	NUM
cuesj-105	154	22	):	):	PUNCT
cuesj-105	154	23	14	14	NUM
cuesj-105	154	24	-	-	SYM
cuesj-105	154	25	20	20	NUM
cuesj-105	154	26	consists	consist	VERB
cuesj-105	154	27	of	of	ADP
cuesj-105	154	28	40	40	NUM
cuesj-105	154	29	classes	class	NOUN
cuesj-105	154	30	with	with	ADP
cuesj-105	154	31	10	10	NUM
cuesj-105	154	32	images	image	NOUN
cuesj-105	154	33	for	for	ADP
cuesj-105	154	34	each	each	DET
cuesj-105	154	35	class	class	NOUN
cuesj-105	154	36	.	.	PUNCT
cuesj-105	155	1	this	this	PRON
cuesj-105	155	2	means	mean	VERB
cuesj-105	155	3	that	that	SCONJ
cuesj-105	155	4	database	database	NOUN
cuesj-105	155	5	consists	consist	VERB
cuesj-105	155	6	of	of	ADP
cuesj-105	155	7	400	400	NUM
cuesj-105	155	8	images	image	NOUN
cuesj-105	155	9	.	.	PUNCT
cuesj-105	156	1	in	in	ADP
cuesj-105	156	2	our	our	PRON
cuesj-105	156	3	experiment	experiment	NOUN
cuesj-105	156	4	,	,	PUNCT
cuesj-105	156	5	the	the	DET
cuesj-105	156	6	database	database	NOUN
cuesj-105	156	7	is	be	AUX
cuesj-105	156	8	divided	divide	VERB
cuesj-105	156	9	into	into	ADP
cuesj-105	156	10	two	two	NUM
cuesj-105	156	11	groups	group	NOUN
cuesj-105	156	12	;	;	PUNCT
cuesj-105	156	13	the	the	DET
cuesj-105	156	14	first	first	ADJ
cuesj-105	156	15	group	group	NOUN
cuesj-105	156	16	is	be	AUX
cuesj-105	156	17	for	for	ADP
cuesj-105	156	18	train	train	NOUN
cuesj-105	156	19	and	and	CCONJ
cuesj-105	156	20	the	the	DET
cuesj-105	156	21	second	second	ADJ
cuesj-105	156	22	group	group	NOUN
cuesj-105	156	23	is	be	AUX
cuesj-105	156	24	for	for	ADP
cuesj-105	156	25	the	the	DET
cuesj-105	156	26	test	test	NOUN
cuesj-105	156	27	.	.	PUNCT
cuesj-105	157	1	the	the	DET
cuesj-105	157	2	training	training	NOUN
cuesj-105	157	3	group	group	NOUN
cuesj-105	157	4	is	be	AUX
cuesj-105	157	5	applied	apply	VERB
cuesj-105	157	6	to	to	PART
cuesj-105	157	7	learn	learn	VERB
cuesj-105	157	8	the	the	DET
cuesj-105	157	9	system	system	NOUN
cuesj-105	157	10	and	and	CCONJ
cuesj-105	157	11	the	the	DET
cuesj-105	157	12	training	training	NOUN
cuesj-105	157	13	group	group	NOUN
cuesj-105	157	14	consists	consist	VERB
cuesj-105	157	15	of	of	ADP
cuesj-105	157	16	five	five	NUM
cuesj-105	157	17	images	image	NOUN
cuesj-105	157	18	of	of	ADP
cuesj-105	157	19	the	the	DET
cuesj-105	157	20	first	first	ADJ
cuesj-105	157	21	32	32	NUM
cuesj-105	157	22	classes	class	NOUN
cuesj-105	157	23	.	.	PUNCT
cuesj-105	158	1	the	the	DET
cuesj-105	158	2	testing	testing	NOUN
cuesj-105	158	3	group	group	NOUN
cuesj-105	158	4	is	be	AUX
cuesj-105	158	5	applied	apply	VERB
cuesj-105	158	6	to	to	PART
cuesj-105	158	7	calculate	calculate	VERB
cuesj-105	158	8	recognition	recognition	NOUN
cuesj-105	158	9	rate	rate	NOUN
cuesj-105	158	10	of	of	ADP
cuesj-105	158	11	the	the	DET
cuesj-105	158	12	proposed	propose	VERB
cuesj-105	158	13	system	system	NOUN
cuesj-105	158	14	and	and	CCONJ
cuesj-105	158	15	it	it	PRON
cuesj-105	158	16	consists	consist	VERB
cuesj-105	158	17	of	of	ADP
cuesj-105	158	18	the	the	DET
cuesj-105	158	19	other	other	ADJ
cuesj-105	158	20	six	six	NUM
cuesj-105	158	21	images	image	NOUN
cuesj-105	158	22	of	of	ADP
cuesj-105	158	23	the	the	DET
cuesj-105	158	24	same	same	ADJ
cuesj-105	158	25	32	32	NUM
cuesj-105	158	26	classes	class	NOUN
cuesj-105	158	27	as	as	ADV
cuesj-105	158	28	well	well	ADV
cuesj-105	158	29	as	as	ADP
cuesj-105	158	30	all	all	DET
cuesj-105	158	31	images	image	NOUN
cuesj-105	158	32	of	of	ADP
cuesj-105	158	33	the	the	DET
cuesj-105	158	34	last	last	ADJ
cuesj-105	158	35	eight	eight	NUM
cuesj-105	158	36	classes	class	NOUN
cuesj-105	158	37	.	.	PUNCT
cuesj-105	159	1	these	these	DET
cuesj-105	159	2	three	three	NUM
cuesj-105	159	3	classes	class	NOUN
cuesj-105	159	4	are	be	AUX
cuesj-105	159	5	used	use	VERB
cuesj-105	159	6	for	for	ADP
cuesj-105	159	7	test	test	NOUN
cuesj-105	159	8	only	only	ADV
cuesj-105	159	9	.	.	PUNCT
cuesj-105	160	1	this	this	PRON
cuesj-105	160	2	is	be	AUX
cuesj-105	160	3	meant	mean	VERB
cuesj-105	160	4	that	that	SCONJ
cuesj-105	160	5	40	40	NUM
cuesj-105	160	6	%	%	NOUN
cuesj-105	160	7	of	of	ADP
cuesj-105	160	8	the	the	DET
cuesj-105	160	9	database	database	NOUN
cuesj-105	160	10	is	be	AUX
cuesj-105	160	11	applied	apply	VERB
cuesj-105	160	12	for	for	ADP
cuesj-105	160	13	training	training	NOUN
cuesj-105	160	14	and	and	CCONJ
cuesj-105	160	15	60	60	NUM
cuesj-105	160	16	%	%	NOUN
cuesj-105	160	17	of	of	ADP
cuesj-105	160	18	the	the	DET
cuesj-105	160	19	image	image	NOUN
cuesj-105	160	20	is	be	AUX
cuesj-105	160	21	applied	apply	VERB
cuesj-105	160	22	for	for	ADP
cuesj-105	160	23	the	the	DET
cuesj-105	160	24	test	test	NOUN
cuesj-105	160	25	.	.	PUNCT
cuesj-105	161	1	figure	figure	NOUN
cuesj-105	161	2	8	8	NUM
cuesj-105	161	3	describes	describe	VERB
cuesj-105	161	4	the	the	DET
cuesj-105	161	5	results	result	NOUN
cuesj-105	161	6	of	of	ADP
cuesj-105	161	7	the	the	DET
cuesj-105	161	8	svms	svms	NOUN
cuesj-105	161	9	classifier	classifier	NOUN
cuesj-105	161	10	using	use	VERB
cuesj-105	161	11	different	different	ADJ
cuesj-105	161	12	kernel	kernel	NOUN
cuesj-105	161	13	function	function	NOUN
cuesj-105	161	14	in	in	ADP
cuesj-105	161	15	orl	orl	PROPN
cuesj-105	161	16	database	database	NOUN
cuesj-105	161	17	.	.	PUNCT
cuesj-105	162	1	in	in	ADP
cuesj-105	162	2	the	the	DET
cuesj-105	162	3	second	second	ADJ
cuesj-105	162	4	experiment	experiment	NOUN
cuesj-105	162	5	,	,	PUNCT
cuesj-105	162	6	ffbpnn	ffbpnn	PROPN
cuesj-105	162	7	is	be	AUX
cuesj-105	162	8	applied	apply	VERB
cuesj-105	162	9	as	as	ADP
cuesj-105	162	10	classifier	classifier	NOUN
cuesj-105	162	11	instead	instead	ADV
cuesj-105	162	12	of	of	ADP
cuesj-105	162	13	the	the	DET
cuesj-105	162	14	svms	svms	NOUN
cuesj-105	162	15	.	.	PUNCT
cuesj-105	163	1	this	this	PRON
cuesj-105	163	2	is	be	AUX
cuesj-105	163	3	done	do	VERB
cuesj-105	163	4	to	to	PART
cuesj-105	163	5	make	make	VERB
cuesj-105	163	6	a	a	DET
cuesj-105	163	7	comparison	comparison	NOUN
cuesj-105	163	8	between	between	ADP
cuesj-105	163	9	the	the	DET
cuesj-105	163	10	results	result	NOUN
cuesj-105	163	11	of	of	ADP
cuesj-105	163	12	svms	svms	NOUN
cuesj-105	163	13	and	and	CCONJ
cuesj-105	163	14	anns	anns	NOUN
cuesj-105	163	15	classifiers	classifier	NOUN
cuesj-105	163	16	as	as	ADV
cuesj-105	163	17	well	well	ADV
cuesj-105	163	18	as	as	ADP
cuesj-105	163	19	to	to	PART
cuesj-105	163	20	calculate	calculate	VERB
cuesj-105	163	21	the	the	DET
cuesj-105	163	22	performance	performance	NOUN
cuesj-105	163	23	of	of	ADP
cuesj-105	163	24	the	the	DET
cuesj-105	163	25	system	system	NOUN
cuesj-105	163	26	.	.	PUNCT
cuesj-105	164	1	figure	figure	NOUN
cuesj-105	164	2	9	9	NUM
cuesj-105	164	3	describes	describe	VERB
cuesj-105	164	4	the	the	DET
cuesj-105	164	5	recognition	recognition	NOUN
cuesj-105	164	6	rate	rate	NOUN
cuesj-105	164	7	results	result	NOUN
cuesj-105	164	8	of	of	ADP
cuesj-105	164	9	the	the	DET
cuesj-105	164	10	svms	svms	NOUN
cuesj-105	164	11	classifier	classifier	NOUN
cuesj-105	164	12	using	use	VERB
cuesj-105	164	13	polynomial	polynomial	ADJ
cuesj-105	164	14	kernel	kernel	NOUN
cuesj-105	164	15	function	function	NOUN
cuesj-105	164	16	and	and	CCONJ
cuesj-105	164	17	feedforward	feedforward	NOUN
cuesj-105	164	18	backpropagation	backpropagation	NOUN
cuesj-105	164	19	neural	neural	ADJ
cuesj-105	164	20	network	network	NOUN
cuesj-105	164	21	classifier	classifier	NOUN
cuesj-105	164	22	.	.	PUNCT
cuesj-105	165	1	conclusions	conclusion	NOUN
cuesj-105	165	2	face	face	VERB
cuesj-105	165	3	recognition	recognition	NOUN
cuesj-105	165	4	systems	system	NOUN
cuesj-105	165	5	are	be	AUX
cuesj-105	165	6	one	one	NUM
cuesj-105	165	7	of	of	ADP
cuesj-105	165	8	the	the	DET
cuesj-105	165	9	computer	computer	NOUN
cuesj-105	165	10	vision	vision	NOUN
cuesj-105	165	11	applications	application	NOUN
cuesj-105	165	12	and	and	CCONJ
cuesj-105	165	13	it	it	PRON
cuesj-105	165	14	is	be	AUX
cuesj-105	165	15	significant	significant	ADJ
cuesj-105	165	16	in	in	ADP
cuesj-105	165	17	many	many	ADJ
cuesj-105	165	18	of	of	ADP
cuesj-105	165	19	.	.	PUNCT
cuesj-105	166	1	in	in	ADP
cuesj-105	166	2	this	this	DET
cuesj-105	166	3	work	work	NOUN
cuesj-105	166	4	,	,	PUNCT
cuesj-105	166	5	face	face	NOUN
cuesj-105	166	6	recognition	recognition	NOUN
cuesj-105	166	7	system	system	NOUN
cuesj-105	166	8	is	be	AUX
cuesj-105	166	9	designing	design	VERB
cuesj-105	166	10	and	and	CCONJ
cuesj-105	166	11	implementation	implementation	NOUN
cuesj-105	166	12	.	.	PUNCT
cuesj-105	167	1	in	in	ADP
cuesj-105	167	2	this	this	DET
cuesj-105	167	3	system	system	NOUN
cuesj-105	167	4	,	,	PUNCT
cuesj-105	167	5	combines	combine	VERB
cuesj-105	167	6	between	between	ADP
cuesj-105	167	7	each	each	PRON
cuesj-105	167	8	of	of	ADP
cuesj-105	167	9	wavelet	wavelet	NOUN
cuesj-105	167	10	transform	transform	NOUN
cuesj-105	167	11	and	and	CCONJ
cuesj-105	167	12	pca	pca	NOUN
cuesj-105	167	13	methods	method	NOUN
cuesj-105	167	14	are	be	AUX
cuesj-105	167	15	applied	apply	VERB
cuesj-105	167	16	for	for	ADP
cuesj-105	167	17	obtained	obtain	VERB
cuesj-105	167	18	feature	feature	NOUN
cuesj-105	167	19	from	from	ADP
cuesj-105	167	20	the	the	DET
cuesj-105	167	21	image	image	NOUN
cuesj-105	167	22	after	after	SCONJ
cuesj-105	167	23	that	that	DET
cuesj-105	167	24	svm	svm	NOUN
cuesj-105	167	25	is	be	AUX
cuesj-105	167	26	applied	apply	VERB
cuesj-105	167	27	to	to	PART
cuesj-105	167	28	classify	classify	VERB
cuesj-105	167	29	these	these	DET
cuesj-105	167	30	features	feature	NOUN
cuesj-105	167	31	.	.	PUNCT
cuesj-105	168	1	the	the	DET
cuesj-105	168	2	proposed	propose	VERB
cuesj-105	168	3	system	system	NOUN
cuesj-105	168	4	is	be	AUX
cuesj-105	168	5	tested	test	VERB
cuesj-105	168	6	using	use	VERB
cuesj-105	168	7	two	two	NUM
cuesj-105	168	8	types	type	NOUN
cuesj-105	168	9	of	of	ADP
cuesj-105	168	10	databases	database	NOUN
cuesj-105	168	11	to	to	PART
cuesj-105	168	12	guarantee	guarantee	VERB
cuesj-105	168	13	from	from	ADP
cuesj-105	168	14	the	the	DET
cuesj-105	168	15	power	power	NOUN
cuesj-105	168	16	of	of	ADP
cuesj-105	168	17	the	the	DET
cuesj-105	168	18	system	system	NOUN
cuesj-105	168	19	.	.	PUNCT
cuesj-105	169	1	comparison	comparison	NOUN
cuesj-105	169	2	between	between	ADP
cuesj-105	169	3	the	the	DET
cuesj-105	169	4	results	result	NOUN
cuesj-105	169	5	91.4285	91.4285	NUM
cuesj-105	169	6	93.333395.2	93.333395.2	NUM
cuesj-105	170	1	95.2	95.2	NUM
cuesj-105	170	2	73.333	73.333	NUM
cuesj-105	170	3	0	0	NUM
cuesj-105	170	4	10	10	NUM
cuesj-105	170	5	20	20	NUM
cuesj-105	170	6	30	30	NUM
cuesj-105	170	7	40	40	NUM
cuesj-105	170	8	50	50	NUM
cuesj-105	170	9	60	60	NUM
cuesj-105	170	10	70	70	NUM
cuesj-105	170	11	80	80	NUM
cuesj-105	170	12	90	90	NUM
cuesj-105	170	13	100	100	NUM
cuesj-105	170	14	accuracy	accuracy	NOUN
cuesj-105	170	15	in	in	ADP
cuesj-105	170	16	yale	yale	PROPN
cuesj-105	170	17	database	database	PROPN
cuesj-105	170	18	linear	linear	PROPN
cuesj-105	170	19	quard	quard	PROPN
cuesj-105	170	20	polynomial	polynomial	ADJ
cuesj-105	170	21	raduis	raduis	NOUN
cuesj-105	170	22	bias	bias	PROPN
cuesj-105	170	23	func	func	PROPN
cuesj-105	170	24	�	�	PROPN
cuesj-105	170	25	on	on	ADP
cuesj-105	170	26	mul	mul	PROPN
cuesj-105	170	27	�	�	PROPN
cuesj-105	170	28	peceptrom	peceptrom	ADP
cuesj-105	170	29	func	func	PROPN
cuesj-105	170	30	�	�	PROPN
cuesj-105	170	31	on	on	ADP
cuesj-105	170	32	figure	figure	NOUN
cuesj-105	170	33	6	6	NUM
cuesj-105	170	34	:	:	PUNCT
cuesj-105	170	35	the	the	DET
cuesj-105	170	36	results	result	NOUN
cuesj-105	170	37	of	of	ADP
cuesj-105	170	38	the	the	DET
cuesj-105	170	39	support	support	NOUN
cuesj-105	170	40	vector	vector	NOUN
cuesj-105	170	41	machines	machine	NOUN
cuesj-105	170	42	classifier	classifier	NOUN
cuesj-105	170	43	using	use	VERB
cuesj-105	170	44	different	different	ADJ
cuesj-105	170	45	kernel	kernel	NOUN
cuesj-105	170	46	functions	function	NOUN
cuesj-105	170	47	in	in	ADP
cuesj-105	170	48	yale	yale	PROPN
cuesj-105	170	49	database	database	NOUN
cuesj-105	170	50	95.2	95.2	NUM
cuesj-105	170	51	87.619	87.619	NUM
cuesj-105	170	52	0	0	NUM
cuesj-105	170	53	10	10	NUM
cuesj-105	170	54	20	20	NUM
cuesj-105	170	55	30	30	NUM
cuesj-105	170	56	40	40	NUM
cuesj-105	170	57	50	50	NUM
cuesj-105	170	58	60	60	NUM
cuesj-105	170	59	70	70	NUM
cuesj-105	170	60	80	80	NUM
cuesj-105	170	61	90	90	NUM
cuesj-105	170	62	100	100	NUM
cuesj-105	170	63	svms	svms	NOUN
cuesj-105	170	64	using	use	VERB
cuesj-105	170	65	ploynomial	ploynomial	ADJ
cuesj-105	170	66	kernel	kernel	NOUN
cuesj-105	170	67	func	func	PROPN
cuesj-105	170	68	�	�	PROPN
cuesj-105	170	69	on	on	ADP
cuesj-105	170	70	feed	feed	NOUN
cuesj-105	170	71	forward	forward	ADV
cuesj-105	170	72	backproga	backproga	PROPN
cuesj-105	170	73	�	�	PROPN
cuesj-105	170	74	on	on	ADP
cuesj-105	170	75	neural	neural	ADJ
cuesj-105	170	76	network	network	NOUN
cuesj-105	170	77	figure	figure	NOUN
cuesj-105	170	78	7	7	NUM
cuesj-105	170	79	:	:	PUNCT
cuesj-105	170	80	the	the	DET
cuesj-105	170	81	accuracy	accuracy	NOUN
cuesj-105	170	82	results	result	NOUN
cuesj-105	170	83	of	of	ADP
cuesj-105	170	84	the	the	DET
cuesj-105	170	85	support	support	NOUN
cuesj-105	170	86	vector	vector	NOUN
cuesj-105	170	87	machines	machine	NOUN
cuesj-105	170	88	classifier	classifier	NOUN
cuesj-105	170	89	using	use	VERB
cuesj-105	170	90	polynomial	polynomial	ADJ
cuesj-105	170	91	kernel	kernel	NOUN
cuesj-105	170	92	function	function	NOUN
cuesj-105	170	93	and	and	CCONJ
cuesj-105	170	94	artificial	artificial	ADJ
cuesj-105	170	95	neural	neural	ADJ
cuesj-105	170	96	network	network	NOUN
cuesj-105	170	97	88.3391.25	88.3391.25	NOUN
cuesj-105	170	98	96.25	96.25	NUM
cuesj-105	170	99	91.667	91.667	NUM
cuesj-105	170	100	65.833	65.833	NUM
cuesj-105	170	101	0	0	NUM
cuesj-105	170	102	10	10	NUM
cuesj-105	170	103	20	20	NUM
cuesj-105	170	104	30	30	NUM
cuesj-105	170	105	40	40	NUM
cuesj-105	170	106	50	50	NUM
cuesj-105	170	107	60	60	NUM
cuesj-105	170	108	70	70	NUM
cuesj-105	170	109	80	80	NUM
cuesj-105	170	110	90	90	NUM
cuesj-105	170	111	100	100	NUM
cuesj-105	170	112	110	110	NUM
cuesj-105	170	113	linear	linear	ADJ
cuesj-105	170	114	quard	quard	NOUN
cuesj-105	170	115	polynomial	polynomial	ADJ
cuesj-105	170	116	raduis	raduis	NOUN
cuesj-105	170	117	bias	bias	NOUN
cuesj-105	170	118	func	func	PROPN
cuesj-105	170	119	�	�	PROPN
cuesj-105	170	120	on	on	ADP
cuesj-105	170	121	figure	figure	NOUN
cuesj-105	170	122	8	8	NUM
cuesj-105	170	123	:	:	PUNCT
cuesj-105	170	124	the	the	DET
cuesj-105	170	125	results	result	NOUN
cuesj-105	170	126	of	of	ADP
cuesj-105	170	127	the	the	DET
cuesj-105	170	128	support	support	NOUN
cuesj-105	170	129	vector	vector	NOUN
cuesj-105	170	130	machines	machine	NOUN
cuesj-105	170	131	classifier	classifier	NOUN
cuesj-105	170	132	using	use	VERB
cuesj-105	170	133	different	different	ADJ
cuesj-105	170	134	kernel	kernel	NOUN
cuesj-105	170	135	function	function	NOUN
cuesj-105	170	136	in	in	ADP
cuesj-105	170	137	orl	orl	PROPN
cuesj-105	170	138	database	database	NOUN
cuesj-105	170	139	96.25	96.25	NUM
cuesj-105	170	140	91.667	91.667	NUM
cuesj-105	170	141	0	0	NUM
cuesj-105	170	142	10	10	NUM
cuesj-105	170	143	20	20	NUM
cuesj-105	170	144	30	30	NUM
cuesj-105	170	145	40	40	NUM
cuesj-105	170	146	50	50	NUM
cuesj-105	170	147	60	60	NUM
cuesj-105	170	148	70	70	NUM
cuesj-105	170	149	80	80	NUM
cuesj-105	170	150	90	90	NUM
cuesj-105	170	151	100	100	NUM
cuesj-105	170	152	svms	svms	NOUN
cuesj-105	170	153	using	use	VERB
cuesj-105	170	154	ploynomial	ploynomial	ADJ
cuesj-105	170	155	kernel	kernel	NOUN
cuesj-105	170	156	func	func	PROPN
cuesj-105	170	157	�	�	PROPN
cuesj-105	170	158	on	on	ADP
cuesj-105	170	159	feed	feed	NOUN
cuesj-105	170	160	forward	forward	ADV
cuesj-105	170	161	backproga	backproga	PROPN
cuesj-105	170	162	�	�	PROPN
cuesj-105	170	163	on	on	ADP
cuesj-105	170	164	neural	neural	ADJ
cuesj-105	170	165	network	network	NOUN
cuesj-105	170	166	figure	figure	NOUN
cuesj-105	170	167	9	9	NUM
cuesj-105	170	168	:	:	PUNCT
cuesj-105	170	169	the	the	DET
cuesj-105	170	170	accuracy	accuracy	NOUN
cuesj-105	170	171	results	result	NOUN
cuesj-105	170	172	of	of	ADP
cuesj-105	170	173	the	the	DET
cuesj-105	170	174	support	support	NOUN
cuesj-105	170	175	vector	vector	NOUN
cuesj-105	170	176	machines	machine	NOUN
cuesj-105	170	177	classifier	classifier	NOUN
cuesj-105	170	178	using	use	VERB
cuesj-105	170	179	polynomial	polynomial	ADJ
cuesj-105	170	180	kernel	kernel	NOUN
cuesj-105	170	181	function	function	NOUN
cuesj-105	170	182	and	and	CCONJ
cuesj-105	170	183	artificial	artificial	ADJ
cuesj-105	170	184	neural	neural	ADJ
cuesj-105	170	185	network	network	NOUN
cuesj-105	170	186	fleah	fleah	PROPN
cuesj-105	170	187	and	and	CCONJ
cuesj-105	170	188	al	al	PROPN
cuesj-105	170	189	-	-	PUNCT
cuesj-105	170	190	aubi	aubi	PROPN
cuesj-105	170	191	:	:	PUNCT
cuesj-105	171	1	frs	frs	PROPN
cuesj-105	171	2	-	-	PUNCT
cuesj-105	171	3	pca	pca	NOUN
cuesj-105	171	4	and	and	CCONJ
cuesj-105	171	5	svm	svm	ADJ
cuesj-105	171	6	20	20	NUM
cuesj-105	171	7	http://journals.cihanuniversity.edu.iq/index.php/cuesj	http://journals.cihanuniversity.edu.iq/index.php/cuesj	ADJ
cuesj-105	171	8	cuesj	cuesj	NOUN
cuesj-105	171	9	2019	2019	NUM
cuesj-105	171	10	,	,	PUNCT
cuesj-105	171	11	3	3	NUM
cuesj-105	171	12	(	(	PUNCT
cuesj-105	171	13	2	2	NUM
cuesj-105	171	14	):	):	PUNCT
cuesj-105	171	15	14	14	NUM
cuesj-105	171	16	-	-	SYM
cuesj-105	171	17	20	20	NUM
cuesj-105	171	18	of	of	ADP
cuesj-105	171	19	proposed	propose	VERB
cuesj-105	171	20	system	system	NOUN
cuesj-105	171	21	and	and	CCONJ
cuesj-105	171	22	ffbnn	ffbnn	PROPN
cuesj-105	171	23	of	of	ADP
cuesj-105	171	24	a	a	DET
cuesj-105	171	25	classifier	classifier	NOUN
cuesj-105	171	26	is	be	AUX
cuesj-105	171	27	implemented	implement	VERB
cuesj-105	171	28	.	.	PUNCT
cuesj-105	172	1	according	accord	VERB
cuesj-105	172	2	to	to	ADP
cuesj-105	172	3	the	the	DET
cuesj-105	172	4	obtained	obtain	VERB
cuesj-105	172	5	results	result	NOUN
cuesj-105	172	6	,	,	PUNCT
cuesj-105	172	7	the	the	DET
cuesj-105	172	8	following	follow	VERB
cuesj-105	172	9	conclusions	conclusion	NOUN
cuesj-105	172	10	can	can	AUX
cuesj-105	172	11	be	be	AUX
cuesj-105	172	12	stated	state	VERB
cuesj-105	172	13	:	:	PUNCT
cuesj-105	172	14	•	•	ADV
cuesj-105	172	15	applying	apply	VERB
cuesj-105	172	16	wavelet	wavelet	NOUN
cuesj-105	172	17	-	-	PUNCT
cuesj-105	172	18	pca	pca	NOUN
cuesj-105	172	19	feature	feature	NOUN
cuesj-105	172	20	extraction	extraction	NOUN
cuesj-105	172	21	algorithms	algorithm	NOUN
cuesj-105	172	22	enhanced	enhance	VERB
cuesj-105	172	23	the	the	DET
cuesj-105	172	24	recognition	recognition	NOUN
cuesj-105	172	25	rate	rate	NOUN
cuesj-105	172	26	of	of	ADP
cuesj-105	172	27	the	the	DET
cuesj-105	172	28	system	system	NOUN
cuesj-105	172	29	compared	compare	VERB
cuesj-105	172	30	to	to	ADP
cuesj-105	172	31	the	the	DET
cuesj-105	172	32	using	use	VERB
cuesj-105	172	33	pca	pca	NOUN
cuesj-105	172	34	only	only	ADV
cuesj-105	172	35	•	•	ADP
cuesj-105	172	36	using	use	VERB
cuesj-105	172	37	svm	svm	PROPN
cuesj-105	172	38	classifier	classifier	NOUN
cuesj-105	172	39	based	base	VERB
cuesj-105	172	40	on	on	ADP
cuesj-105	172	41	polynomial	polynomial	ADJ
cuesj-105	172	42	kernel	kernel	NOUN
cuesj-105	172	43	function	function	NOUN
cuesj-105	172	44	increases	increase	VERB
cuesj-105	172	45	the	the	DET
cuesj-105	172	46	accuracy	accuracy	NOUN
cuesj-105	172	47	of	of	ADP
cuesj-105	172	48	recognition	recognition	NOUN
cuesj-105	172	49	rate	rate	NOUN
cuesj-105	172	50	of	of	ADP
cuesj-105	172	51	the	the	DET
cuesj-105	172	52	system	system	NOUN
cuesj-105	172	53	by	by	ADP
cuesj-105	172	54	5	5	NUM
cuesj-105	172	55	%	%	NOUN
cuesj-105	172	56	compared	compare	VERB
cuesj-105	172	57	to	to	ADP
cuesj-105	172	58	feedforward	feedforward	VERB
cuesj-105	172	59	neural	neural	ADJ
cuesj-105	172	60	network	network	NOUN
cuesj-105	172	61	classifier	classifier	NOUN
cuesj-105	172	62	,	,	PUNCT
cuesj-105	172	63	especially	especially	ADV
cuesj-105	172	64	in	in	ADP
cuesj-105	172	65	cases	case	NOUN
cuesj-105	172	66	where	where	SCONJ
cuesj-105	172	67	the	the	DET
cuesj-105	172	68	features	feature	NOUN
cuesj-105	172	69	of	of	ADP
cuesj-105	172	70	a	a	DET
cuesj-105	172	71	person	person	NOUN
cuesj-105	172	72	’s	’s	PART
cuesj-105	172	73	face	face	NOUN
cuesj-105	172	74	are	be	AUX
cuesj-105	172	75	modified	modify	VERB
cuesj-105	172	76	with	with	ADP
cuesj-105	172	77	different	different	ADJ
cuesj-105	172	78	hairstyles	hairstyle	NOUN
cuesj-105	172	79	or	or	CCONJ
cuesj-105	172	80	wearing	wear	VERB
cuesj-105	172	81	glasses	glass	NOUN
cuesj-105	172	82	.	.	PUNCT
cuesj-105	173	1	the	the	DET
cuesj-105	173	2	comparison	comparison	NOUN
cuesj-105	173	3	between	between	ADP
cuesj-105	173	4	svm	svm	PROPN
cuesj-105	173	5	and	and	CCONJ
cuesj-105	173	6	ffbnn	ffbnn	PROPN
cuesj-105	173	7	classifiers	classifier	NOUN
cuesj-105	173	8	proved	prove	VERB
cuesj-105	173	9	that	that	SCONJ
cuesj-105	173	10	svm	svm	PROPN
cuesj-105	173	11	classifier	classifier	NOUN
cuesj-105	173	12	is	be	AUX
cuesj-105	173	13	more	more	ADV
cuesj-105	173	14	accurate	accurate	ADJ
cuesj-105	173	15	than	than	ADP
cuesj-105	173	16	ffbnn	ffbnn	PROPN
cuesj-105	173	17	classifier	classifier	PROPN
cuesj-105	173	18	in	in	ADP
cuesj-105	173	19	recognition	recognition	NOUN
cuesj-105	173	20	.	.	PUNCT
cuesj-105	174	1	references	reference	NOUN
cuesj-105	174	2	1	1	NUM
cuesj-105	174	3	.	.	PUNCT
cuesj-105	174	4	s.	s.	PROPN
cuesj-105	174	5	gupta	gupta	PROPN
cuesj-105	174	6	,	,	PUNCT
cuesj-105	174	7	o.	o.	PROPN
cuesj-105	174	8	p.	p.	PROPN
cuesj-105	174	9	sahu	sahu	PROPN
cuesj-105	174	10	,	,	PUNCT
cuesj-105	174	11	r.	r.	PROPN
cuesj-105	174	12	gupta	gupta	PROPN
cuesj-105	174	13	and	and	CCONJ
cuesj-105	174	14	a.	a.	NOUN
cuesj-105	174	15	goel	goel	PROPN
cuesj-105	174	16	.	.	PUNCT
cuesj-105	175	1	“	"	PUNCT
cuesj-105	175	2	a	a	DET
cuesj-105	175	3	bespoke	bespoke	NOUN
cuesj-105	175	4	approach	approach	NOUN
cuesj-105	175	5	for	for	ADP
cuesj-105	175	6	face	face	NOUN
cuesj-105	175	7	-	-	PUNCT
cuesj-105	175	8	recognition	recognition	NOUN
cuesj-105	175	9	using	use	VERB
cuesj-105	175	10	pca	pca	PROPN
cuesj-105	175	11	”	"	PUNCT
cuesj-105	175	12	.	.	PUNCT
cuesj-105	176	1	international	international	ADJ
cuesj-105	176	2	journal	journal	PROPN
cuesj-105	176	3	of	of	ADP
cuesj-105	176	4	computer	computer	NOUN
cuesj-105	176	5	science	science	NOUN
cuesj-105	176	6	and	and	CCONJ
cuesj-105	176	7	engineering	engineering	NOUN
cuesj-105	176	8	,	,	PUNCT
cuesj-105	176	9	vol	vol	NOUN
cuesj-105	176	10	.	.	PROPN
cuesj-105	176	11	2	2	NUM
cuesj-105	176	12	,	,	PUNCT
cuesj-105	176	13	no	no	INTJ
cuesj-105	176	14	.	.	NOUN
cuesj-105	176	15	02	02	NUM
cuesj-105	176	16	,	,	PUNCT
cuesj-105	176	17	pp.155	pp.155	PROPN
cuesj-105	176	18	-	-	X
cuesj-105	176	19	158	158	NUM
cuesj-105	176	20	,	,	PUNCT
cuesj-105	176	21	2010	2010	NUM
cuesj-105	176	22	.	.	PUNCT
cuesj-105	177	1	2	2	X
cuesj-105	177	2	.	.	X
cuesj-105	177	3	m.	m.	PROPN
cuesj-105	177	4	david	david	PROPN
cuesj-105	177	5	.	.	PUNCT
cuesj-105	178	1	“	"	PUNCT
cuesj-105	178	2	reversible	reversible	ADJ
cuesj-105	178	3	integer	integer	NOUN
cuesj-105	178	4	to	to	ADP
cuesj-105	178	5	integer	integer	NOUN
cuesj-105	178	6	wavelet	wavelet	NOUN
cuesj-105	178	7	transforms	transform	VERB
cuesj-105	178	8	for	for	ADP
cuesj-105	178	9	image	image	NOUN
cuesj-105	178	10	coding	coding	NOUN
cuesj-105	178	11	”	"	PUNCT
cuesj-105	178	12	.	.	PUNCT
cuesj-105	179	1	thesis	thesis	NOUN
cuesj-105	179	2	,	,	PUNCT
cuesj-105	179	3	university	university	PROPN
cuesj-105	179	4	of	of	ADP
cuesj-105	179	5	british	british	PROPN
cuesj-105	179	6	columbia	columbia	PROPN
cuesj-105	179	7	,	,	PUNCT
cuesj-105	179	8	2002	2002	NUM
cuesj-105	179	9	.	.	PUNCT
cuesj-105	180	1	3	3	X
cuesj-105	180	2	.	.	PUNCT
cuesj-105	180	3	j.	j.	PROPN
cuesj-105	180	4	s.	s.	PROPN
cuesj-105	180	5	taneja	taneja	PROPN
cuesj-105	180	6	.	.	PUNCT
cuesj-105	181	1	“	"	PUNCT
cuesj-105	181	2	analysis	analysis	NOUN
cuesj-105	181	3	of	of	ADP
cuesj-105	181	4	e.	e.	PROPN
cuesj-105	181	5	coli	coli	NOUN
cuesj-105	181	6	promoters	promoter	NOUN
cuesj-105	181	7	using	use	VERB
cuesj-105	181	8	support	support	NOUN
cuesj-105	181	9	vector	vector	NOUN
cuesj-105	181	10	machine	machine	NOUN
cuesj-105	181	11	”	"	PUNCT
cuesj-105	181	12	.	.	PUNCT
cuesj-105	182	1	thesis	thesis	NOUN
cuesj-105	182	2	,	,	PUNCT
cuesj-105	182	3	thapar	thapar	PROPN
cuesj-105	182	4	institute	institute	PROPN
cuesj-105	182	5	of	of	ADP
cuesj-105	182	6	engineering	engineering	NOUN
cuesj-105	182	7	and	and	CCONJ
cuesj-105	182	8	technology	technology	NOUN
cuesj-105	182	9	,	,	PUNCT
cuesj-105	182	10	2006	2006	NUM
cuesj-105	182	11	.	.	PUNCT
cuesj-105	183	1	4	4	X
cuesj-105	183	2	.	.	X
cuesj-105	183	3	m.	m.	PROPN
cuesj-105	183	4	e.	e.	PROPN
cuesj-105	183	5	mavroforakis	mavroforakis	PROPN
cuesj-105	183	6	.	.	PUNCT
cuesj-105	184	1	“	"	PUNCT
cuesj-105	184	2	geometric	geometric	ADJ
cuesj-105	184	3	approach	approach	NOUN
cuesj-105	184	4	to	to	ADP
cuesj-105	184	5	statistical	statistical	ADJ
cuesj-105	184	6	learning	learning	NOUN
cuesj-105	184	7	theory	theory	NOUN
cuesj-105	184	8	through	through	ADP
cuesj-105	184	9	support	support	NOUN
cuesj-105	184	10	vector	vector	NOUN
cuesj-105	184	11	machines	machine	NOUN
cuesj-105	184	12	(	(	PUNCT
cuesj-105	184	13	svm	svm	PROPN
cuesj-105	184	14	)	)	PUNCT
cuesj-105	184	15	with	with	ADP
cuesj-105	184	16	application	application	NOUN
cuesj-105	184	17	to	to	ADP
cuesj-105	184	18	medical	medical	ADJ
cuesj-105	184	19	diagnosis	diagnosis	NOUN
cuesj-105	184	20	”	"	PUNCT
cuesj-105	184	21	.	.	PUNCT
cuesj-105	185	1	thesis	thesis	NOUN
cuesj-105	185	2	,	,	PUNCT
cuesj-105	185	3	national	national	ADJ
cuesj-105	185	4	and	and	CCONJ
cuesj-105	185	5	kapodistrian	kapodistrian	ADJ
cuesj-105	185	6	university	university	PROPN
cuesj-105	185	7	of	of	ADP
cuesj-105	185	8	athens	athens	PROPN
cuesj-105	185	9	,	,	PUNCT
cuesj-105	185	10	2008	2008	NUM
cuesj-105	185	11	.	.	PUNCT
cuesj-105	186	1	5	5	NUM
cuesj-105	186	2	.	.	X
cuesj-105	186	3	x.	x.	PROPN
cuesj-105	186	4	wu	wu	PROPN
cuesj-105	186	5	,	,	PUNCT
cuesj-105	186	6	v.	v.	PROPN
cuesj-105	186	7	kumar	kumar	PROPN
cuesj-105	186	8	,	,	PUNCT
cuesj-105	186	9	j.	j.	PROPN
cuesj-105	186	10	r.	r.	PROPN
cuesj-105	186	11	quinlan	quinlan	PROPN
cuesj-105	186	12	and	and	CCONJ
cuesj-105	186	13	j.	j.	PROPN
cuesj-105	186	14	ghosh	ghosh	PROPN
cuesj-105	186	15	.	.	PUNCT
cuesj-105	187	1	“	"	PUNCT
cuesj-105	187	2	top	top	ADJ
cuesj-105	187	3	10	10	NUM
cuesj-105	187	4	algorithms	algorithm	NOUN
cuesj-105	187	5	in	in	ADP
cuesj-105	187	6	data	data	NOUN
cuesj-105	187	7	mining	mining	NOUN
cuesj-105	187	8	”	"	PUNCT
cuesj-105	187	9	.	.	PUNCT
cuesj-105	188	1	knowledge	knowledge	NOUN
cuesj-105	188	2	and	and	CCONJ
cuesj-105	188	3	information	information	NOUN
cuesj-105	188	4	systems	system	NOUN
cuesj-105	188	5	,	,	PUNCT
cuesj-105	188	6	vol	vol	NOUN
cuesj-105	188	7	.	.	PROPN
cuesj-105	189	1	14	14	NUM
cuesj-105	189	2	,	,	PUNCT
cuesj-105	189	3	no	no	INTJ
cuesj-105	189	4	.	.	NOUN
cuesj-105	189	5	1	1	NUM
cuesj-105	189	6	,	,	PUNCT
cuesj-105	189	7	pp	pp	ADJ
cuesj-105	189	8	.	.	PUNCT
cuesj-105	190	1	1	1	NUM
cuesj-105	190	2	-	-	SYM
cuesj-105	190	3	37	37	NUM
cuesj-105	190	4	,	,	PUNCT
cuesj-105	190	5	2008	2008	NUM
cuesj-105	190	6	.	.	PUNCT
cuesj-105	191	1	6	6	NUM
cuesj-105	191	2	.	.	X
cuesj-105	191	3	m.	m.	NOUN
cuesj-105	191	4	zuhaer	zuhaer	PROPN
cuesj-105	191	5	and	and	CCONJ
cuesj-105	191	6	n.	n.	PROPN
cuesj-105	191	7	al	al	PROPN
cuesj-105	191	8	-	-	PUNCT
cuesj-105	191	9	dabagh	dabagh	PROPN
cuesj-105	191	10	.	.	PUNCT
cuesj-105	192	1	“	"	PUNCT
cuesj-105	192	2	face	face	VERB
cuesj-105	192	3	recognition	recognition	NOUN
cuesj-105	192	4	using	use	VERB
cuesj-105	192	5	lbp	lbp	PROPN
cuesj-105	192	6	,	,	PUNCT
cuesj-105	192	7	fld	fld	PROPN
cuesj-105	192	8	and	and	CCONJ
cuesj-105	192	9	svm	svm	VERB
cuesj-105	192	10	with	with	ADP
cuesj-105	192	11	single	single	ADJ
cuesj-105	192	12	training	training	NOUN
cuesj-105	192	13	sample	sample	NOUN
cuesj-105	192	14	per	per	ADP
cuesj-105	192	15	person	person	NOUN
cuesj-105	192	16	”	"	PUNCT
cuesj-105	192	17	.	.	PUNCT
cuesj-105	193	1	international	international	ADJ
cuesj-105	193	2	journal	journal	PROPN
cuesj-105	193	3	of	of	ADP
cuesj-105	193	4	scientific	scientific	ADJ
cuesj-105	193	5	and	and	CCONJ
cuesj-105	193	6	engineering	engineering	NOUN
cuesj-105	193	7	research	research	NOUN
cuesj-105	193	8	,	,	PUNCT
cuesj-105	193	9	vol	vol	NOUN
cuesj-105	193	10	.	.	PROPN
cuesj-105	193	11	5	5	NUM
cuesj-105	193	12	,	,	PUNCT
cuesj-105	193	13	no	no	INTJ
cuesj-105	193	14	.	.	NOUN
cuesj-105	193	15	5	5	NUM
cuesj-105	193	16	,	,	PUNCT
cuesj-105	193	17	pp	pp	ADJ
cuesj-105	193	18	.	.	PUNCT
cuesj-105	193	19	180	180	NUM
cuesj-105	193	20	,	,	PUNCT
cuesj-105	193	21	may	may	AUX
cuesj-105	193	22	2014	2014	NUM
cuesj-105	193	23	.	.	PUNCT
cuesj-105	194	1	7	7	X
cuesj-105	194	2	.	.	X
cuesj-105	194	3	r.	r.	PROPN
cuesj-105	194	4	o.	o.	PROPN
cuesj-105	194	5	duda	duda	PROPN
cuesj-105	194	6	,	,	PUNCT
cuesj-105	194	7	p.	p.	PROPN
cuesj-105	194	8	e.	e.	PROPN
cuesj-105	194	9	hart	hart	PROPN
cuesj-105	194	10	and	and	CCONJ
cuesj-105	194	11	d.	d.	PROPN
cuesj-105	194	12	g.	g.	PROPN
cuesj-105	194	13	stork	stork	PROPN
cuesj-105	194	14	.	.	PUNCT
cuesj-105	195	1	“	"	PUNCT
cuesj-105	195	2	pattern	pattern	NOUN
cuesj-105	195	3	classification	classification	NOUN
cuesj-105	195	4	”	"	PUNCT
cuesj-105	195	5	.	.	PUNCT
cuesj-105	196	1	2nd	2nd	ADJ
cuesj-105	196	2	ed	ed	NOUN
cuesj-105	196	3	.	.	PUNCT
cuesj-105	197	1	new	new	PROPN
cuesj-105	197	2	york	york	PROPN
cuesj-105	197	3	:	:	PUNCT
cuesj-105	198	1	john	john	PROPN
cuesj-105	198	2	wiley	wiley	PROPN
cuesj-105	198	3	and	and	CCONJ
cuesj-105	198	4	sons	son	NOUN
cuesj-105	198	5	,	,	PUNCT
cuesj-105	198	6	inc	inc	PROPN
cuesj-105	198	7	.	.	PROPN
cuesj-105	198	8	,	,	PUNCT
cuesj-105	198	9	2000	2000	NUM
cuesj-105	198	10	.	.	PUNCT
cuesj-105	199	1	8	8	X
cuesj-105	199	2	.	.	PUNCT
cuesj-105	200	1	p.	p.	NOUN
cuesj-105	200	2	n.	n.	PROPN
cuesj-105	200	3	bellhumer	bellhumer	PROPN
cuesj-105	200	4	,	,	PUNCT
cuesj-105	200	5	j.	j.	PROPN
cuesj-105	200	6	hespanha	hespanha	PROPN
cuesj-105	200	7	and	and	CCONJ
cuesj-105	200	8	d.	d.	PROPN
cuesj-105	200	9	kriegman	kriegman	PROPN
cuesj-105	200	10	.	.	PUNCT
cuesj-105	201	1	eigenfaces	eigenface	NOUN
cuesj-105	201	2	vs.	vs.	ADP
cuesj-105	201	3	fisherfaces	fisherface	NOUN
cuesj-105	201	4	:	:	PUNCT
cuesj-105	201	5	recognition	recognition	NOUN
cuesj-105	201	6	using	use	VERB
cuesj-105	201	7	class	class	NOUN
cuesj-105	201	8	specific	specific	ADJ
cuesj-105	201	9	linear	linear	PROPN
cuesj-105	201	10	projection	projection	NOUN
cuesj-105	201	11	.	.	PUNCT
cuesj-105	202	1	ieee	ieee	NOUN
cuesj-105	202	2	transactions	transaction	NOUN
cuesj-105	202	3	on	on	ADP
cuesj-105	202	4	pattern	pattern	NOUN
cuesj-105	202	5	analysis	analysis	NOUN
cuesj-105	202	6	and	and	CCONJ
cuesj-105	202	7	machine	machine	NOUN
cuesj-105	202	8	intelligence	intelligence	NOUN
cuesj-105	202	9	,	,	PUNCT
cuesj-105	202	10	vol	vol	NOUN
cuesj-105	202	11	.	.	PROPN
cuesj-105	202	12	19	19	NUM
cuesj-105	202	13	,	,	PUNCT
cuesj-105	202	14	no	no	INTJ
cuesj-105	202	15	.	.	NOUN
cuesj-105	202	16	7	7	NUM
cuesj-105	202	17	,	,	PUNCT
cuesj-105	202	18	pp	pp	ADJ
cuesj-105	202	19	.	.	PUNCT
cuesj-105	203	1	711	711	NUM
cuesj-105	203	2	-	-	SYM
cuesj-105	203	3	720	720	NUM
cuesj-105	203	4	,	,	PUNCT
cuesj-105	203	5	july	july	PROPN
cuesj-105	203	6	1997	1997	NUM
cuesj-105	203	7	.	.	PUNCT
cuesj-105	204	1	9	9	X
cuesj-105	204	2	.	.	PUNCT
cuesj-105	204	3	f.	f.	PROPN
cuesj-105	204	4	samaria	samaria	PROPN
cuesj-105	204	5	and	and	CCONJ
cuesj-105	204	6	a.	a.	NOUN
cuesj-105	204	7	harter	harter	PROPN
cuesj-105	204	8	.	.	PUNCT
cuesj-105	205	1	d.	d.	PROPN
cuesj-105	205	2	parameterisation	parameterisation	NOUN
cuesj-105	205	3	of	of	ADP
cuesj-105	205	4	a	a	DET
cuesj-105	205	5	stochastic	stochastic	ADJ
cuesj-105	205	6	model	model	NOUN
cuesj-105	205	7	for	for	ADP
cuesj-105	205	8	human	human	ADJ
cuesj-105	205	9	face	face	NOUN
cuesj-105	205	10	identificatio	identificatio	NOUN
cuesj-105	205	11	.	.	PUNCT
cuesj-105	206	1	conference	conference	NOUN
cuesj-105	206	2	:	:	PUNCT
cuesj-105	206	3	applications	application	NOUN
cuesj-105	206	4	of	of	ADP
cuesj-105	206	5	computer	computer	NOUN
cuesj-105	206	6	vision	vision	NOUN
cuesj-105	206	7	.	.	PUNCT
cuesj-105	207	1	proceedings	proceeding	NOUN
cuesj-105	207	2	of	of	ADP
cuesj-105	207	3	the	the	DET
cuesj-105	207	4	second	second	ADJ
cuesj-105	207	5	ieee	ieee	NOUN
cuesj-105	207	6	workshop	workshop	NOUN
cuesj-105	207	7	,	,	PUNCT
cuesj-105	207	8	1994	1994	NUM
cuesj-105	207	9	.	.	PUNCT