id	sid	tid	token	lemma	pos
ejpam-5354	1	1	european	european	PROPN
ejpam-5354	1	2	journal	journal	PROPN
ejpam-5354	1	3	of	of	ADP
ejpam-5354	1	4	pure	pure	ADJ
ejpam-5354	1	5	and	and	CCONJ
ejpam-5354	1	6	applied	apply	VERB
ejpam-5354	1	7	mathematics	mathematic	NOUN
ejpam-5354	1	8	vol	vol	NOUN
ejpam-5354	1	9	.	.	PROPN
ejpam-5354	2	1	17	17	NUM
ejpam-5354	2	2	,	,	PUNCT
ejpam-5354	2	3	no	no	INTJ
ejpam-5354	2	4	.	.	NOUN
ejpam-5354	2	5	4	4	NUM
ejpam-5354	2	6	,	,	PUNCT
ejpam-5354	2	7	2024	2024	NUM
ejpam-5354	2	8	,	,	PUNCT
ejpam-5354	2	9	2692	2692	NUM
ejpam-5354	2	10	-	-	SYM
ejpam-5354	2	11	2705	2705	NUM
ejpam-5354	2	12	issn	issn	PROPN
ejpam-5354	2	13	1307	1307	NUM
ejpam-5354	2	14	-	-	SYM
ejpam-5354	2	15	5543	5543	NUM
ejpam-5354	2	16	–	–	PUNCT
ejpam-5354	2	17	ejpam.com	ejpam.com	X
ejpam-5354	2	18	published	publish	VERB
ejpam-5354	2	19	by	by	ADP
ejpam-5354	2	20	new	new	PROPN
ejpam-5354	2	21	york	york	PROPN
ejpam-5354	2	22	business	business	PROPN
ejpam-5354	2	23	global	global	PROPN
ejpam-5354	2	24	enhanced	enhance	VERB
ejpam-5354	2	25	conjugate	conjugate	ADJ
ejpam-5354	2	26	gradient	gradient	NOUN
ejpam-5354	2	27	method	method	NOUN
ejpam-5354	2	28	for	for	ADP
ejpam-5354	2	29	unconstrained	unconstrained	ADJ
ejpam-5354	2	30	optimization	optimization	NOUN
ejpam-5354	2	31	and	and	CCONJ
ejpam-5354	2	32	its	its	PRON
ejpam-5354	2	33	application	application	NOUN
ejpam-5354	2	34	in	in	ADP
ejpam-5354	2	35	neural	neural	ADJ
ejpam-5354	2	36	networks	network	NOUN
ejpam-5354	2	37	dlovan	dlovan	VERB
ejpam-5354	2	38	haji	haji	PROPN
ejpam-5354	2	39	omar1	omar1	PROPN
ejpam-5354	2	40	,	,	PUNCT
ejpam-5354	2	41	alaa	alaa	PROPN
ejpam-5354	2	42	luqman	luqman	PROPN
ejpam-5354	2	43	ibrahim1,∗	ibrahim1,∗	PROPN
ejpam-5354	2	44	,	,	PUNCT
ejpam-5354	2	45	masoud	masoud	PROPN
ejpam-5354	2	46	muhammed	muhammed	PROPN
ejpam-5354	2	47	hassan2	hassan2	PROPN
ejpam-5354	2	48	,	,	PUNCT
ejpam-5354	2	49	bayda	bayda	VERB
ejpam-5354	2	50	ghanim	ghanim	NOUN
ejpam-5354	2	51	fathi1	fathi1	PROPN
ejpam-5354	2	52	,	,	PUNCT
ejpam-5354	2	53	diman	diman	NOUN
ejpam-5354	2	54	abdulqader	abdulqader	NOUN
ejpam-5354	2	55	sulaiman1	sulaiman1	PROPN
ejpam-5354	2	56	1	1	NUM
ejpam-5354	2	57	department	department	NOUN
ejpam-5354	2	58	of	of	ADP
ejpam-5354	2	59	mathematics	mathematic	NOUN
ejpam-5354	2	60	,	,	PUNCT
ejpam-5354	2	61	college	college	NOUN
ejpam-5354	2	62	of	of	ADP
ejpam-5354	2	63	science	science	NOUN
ejpam-5354	2	64	,	,	PUNCT
ejpam-5354	2	65	university	university	NOUN
ejpam-5354	2	66	of	of	ADP
ejpam-5354	2	67	zakho	zakho	PROPN
ejpam-5354	2	68	,	,	PUNCT
ejpam-5354	2	69	kurdistan	kurdistan	ADJ
ejpam-5354	2	70	region	region	NOUN
ejpam-5354	2	71	,	,	PUNCT
ejpam-5354	2	72	iraq	iraq	PROPN
ejpam-5354	2	73	2	2	NUM
ejpam-5354	2	74	department	department	NOUN
ejpam-5354	2	75	of	of	ADP
ejpam-5354	2	76	computer	computer	NOUN
ejpam-5354	2	77	science	science	NOUN
ejpam-5354	2	78	,	,	PUNCT
ejpam-5354	2	79	college	college	NOUN
ejpam-5354	2	80	of	of	ADP
ejpam-5354	2	81	science	science	NOUN
ejpam-5354	2	82	,	,	PUNCT
ejpam-5354	2	83	university	university	NOUN
ejpam-5354	2	84	of	of	ADP
ejpam-5354	2	85	zakho	zakho	PROPN
ejpam-5354	2	86	,	,	PUNCT
ejpam-5354	2	87	kurdistan	kurdistan	ADJ
ejpam-5354	2	88	region	region	NOUN
ejpam-5354	2	89	,	,	PUNCT
ejpam-5354	2	90	iraq	iraq	PROPN
ejpam-5354	2	91	abstract	abstract	NOUN
ejpam-5354	2	92	.	.	PUNCT
ejpam-5354	3	1	in	in	ADP
ejpam-5354	3	2	this	this	DET
ejpam-5354	3	3	study	study	NOUN
ejpam-5354	3	4	,	,	PUNCT
ejpam-5354	3	5	we	we	PRON
ejpam-5354	3	6	present	present	VERB
ejpam-5354	3	7	a	a	DET
ejpam-5354	3	8	novel	novel	ADJ
ejpam-5354	3	9	conjugate	conjugate	ADJ
ejpam-5354	3	10	gradient	gradient	NOUN
ejpam-5354	3	11	method	method	NOUN
ejpam-5354	3	12	specifically	specifically	ADV
ejpam-5354	3	13	designed	design	VERB
ejpam-5354	3	14	for	for	ADP
ejpam-5354	3	15	addressing	address	VERB
ejpam-5354	3	16	with	with	ADP
ejpam-5354	3	17	unconstrained	unconstrained	ADJ
ejpam-5354	3	18	optimization	optimization	NOUN
ejpam-5354	3	19	problems	problem	NOUN
ejpam-5354	3	20	.	.	PUNCT
ejpam-5354	4	1	traditional	traditional	ADJ
ejpam-5354	4	2	conjugate	conjugate	ADJ
ejpam-5354	4	3	gradient	gradient	NOUN
ejpam-5354	4	4	methods	method	NOUN
ejpam-5354	4	5	have	have	AUX
ejpam-5354	4	6	shown	show	VERB
ejpam-5354	4	7	effectiveness	effectiveness	NOUN
ejpam-5354	4	8	in	in	ADP
ejpam-5354	4	9	solving	solve	VERB
ejpam-5354	4	10	optimization	optimization	NOUN
ejpam-5354	4	11	problems	problem	NOUN
ejpam-5354	4	12	,	,	PUNCT
ejpam-5354	4	13	but	but	CCONJ
ejpam-5354	4	14	they	they	PRON
ejpam-5354	4	15	may	may	AUX
ejpam-5354	4	16	encounter	encounter	VERB
ejpam-5354	4	17	challenges	challenge	NOUN
ejpam-5354	4	18	when	when	SCONJ
ejpam-5354	4	19	dealing	deal	VERB
ejpam-5354	4	20	with	with	ADP
ejpam-5354	4	21	unconstrained	unconstrained	ADJ
ejpam-5354	4	22	problems	problem	NOUN
ejpam-5354	4	23	.	.	PUNCT
ejpam-5354	5	1	our	our	PRON
ejpam-5354	5	2	method	method	NOUN
ejpam-5354	5	3	addresses	address	VERB
ejpam-5354	5	4	this	this	DET
ejpam-5354	5	5	issue	issue	NOUN
ejpam-5354	5	6	by	by	ADP
ejpam-5354	5	7	introducing	introduce	VERB
ejpam-5354	5	8	modifications	modification	NOUN
ejpam-5354	5	9	that	that	PRON
ejpam-5354	5	10	enhance	enhance	VERB
ejpam-5354	5	11	its	its	PRON
ejpam-5354	5	12	performance	performance	NOUN
ejpam-5354	5	13	in	in	ADP
ejpam-5354	5	14	the	the	DET
ejpam-5354	5	15	unconstrained	unconstrained	ADJ
ejpam-5354	5	16	setting	setting	NOUN
ejpam-5354	5	17	.	.	PUNCT
ejpam-5354	6	1	we	we	PRON
ejpam-5354	6	2	demonstrate	demonstrate	VERB
ejpam-5354	6	3	that	that	SCONJ
ejpam-5354	6	4	,	,	PUNCT
ejpam-5354	6	5	under	under	ADP
ejpam-5354	6	6	certain	certain	ADJ
ejpam-5354	6	7	conditions	condition	NOUN
ejpam-5354	6	8	,	,	PUNCT
ejpam-5354	6	9	our	our	PRON
ejpam-5354	6	10	method	method	NOUN
ejpam-5354	6	11	satisfies	satisfy	VERB
ejpam-5354	6	12	both	both	CCONJ
ejpam-5354	6	13	the	the	DET
ejpam-5354	6	14	sufficient	sufficient	ADJ
ejpam-5354	6	15	descent	descent	NOUN
ejpam-5354	6	16	criteria	criterion	NOUN
ejpam-5354	6	17	and	and	CCONJ
ejpam-5354	6	18	establishes	establish	VERB
ejpam-5354	6	19	global	global	ADJ
ejpam-5354	6	20	convergence	convergence	NOUN
ejpam-5354	6	21	,	,	PUNCT
ejpam-5354	6	22	ensuring	ensure	VERB
ejpam-5354	6	23	progress	progress	NOUN
ejpam-5354	6	24	towards	towards	ADP
ejpam-5354	6	25	the	the	DET
ejpam-5354	6	26	optimal	optimal	ADJ
ejpam-5354	6	27	solution	solution	NOUN
ejpam-5354	6	28	at	at	ADP
ejpam-5354	6	29	each	each	DET
ejpam-5354	6	30	iteration	iteration	NOUN
ejpam-5354	6	31	.	.	PUNCT
ejpam-5354	7	1	moreover	moreover	ADV
ejpam-5354	7	2	,	,	PUNCT
ejpam-5354	7	3	we	we	PRON
ejpam-5354	7	4	establish	establish	VERB
ejpam-5354	7	5	the	the	DET
ejpam-5354	7	6	global	global	ADJ
ejpam-5354	7	7	convergence	convergence	NOUN
ejpam-5354	7	8	of	of	ADP
ejpam-5354	7	9	our	our	PRON
ejpam-5354	7	10	method	method	NOUN
ejpam-5354	7	11	,	,	PUNCT
ejpam-5354	7	12	providing	provide	VERB
ejpam-5354	7	13	confidence	confidence	NOUN
ejpam-5354	7	14	in	in	ADP
ejpam-5354	7	15	its	its	PRON
ejpam-5354	7	16	ability	ability	NOUN
ejpam-5354	7	17	to	to	PART
ejpam-5354	7	18	find	find	VERB
ejpam-5354	7	19	the	the	DET
ejpam-5354	7	20	global	global	ADJ
ejpam-5354	7	21	optimum	optimum	NOUN
ejpam-5354	7	22	.	.	PUNCT
ejpam-5354	8	1	to	to	PART
ejpam-5354	8	2	showcase	showcase	VERB
ejpam-5354	8	3	the	the	DET
ejpam-5354	8	4	practical	practical	ADJ
ejpam-5354	8	5	applicability	applicability	NOUN
ejpam-5354	8	6	of	of	ADP
ejpam-5354	8	7	our	our	PRON
ejpam-5354	8	8	approach	approach	NOUN
ejpam-5354	8	9	,	,	PUNCT
ejpam-5354	8	10	we	we	PRON
ejpam-5354	8	11	apply	apply	VERB
ejpam-5354	8	12	this	this	DET
ejpam-5354	8	13	novel	novel	ADJ
ejpam-5354	8	14	method	method	NOUN
ejpam-5354	8	15	to	to	ADP
ejpam-5354	8	16	a	a	DET
ejpam-5354	8	17	dataset	dataset	NOUN
ejpam-5354	8	18	,	,	PUNCT
ejpam-5354	8	19	applying	apply	VERB
ejpam-5354	8	20	a	a	DET
ejpam-5354	8	21	feed	feed	NOUN
ejpam-5354	8	22	-	-	PUNCT
ejpam-5354	8	23	forward	forward	ADV
ejpam-5354	8	24	neural	neural	ADJ
ejpam-5354	8	25	network	network	NOUN
ejpam-5354	8	26	value	value	NOUN
ejpam-5354	8	27	estimation	estimation	NOUN
ejpam-5354	8	28	for	for	ADP
ejpam-5354	8	29	continuous	continuous	ADJ
ejpam-5354	8	30	trigonometric	trigonometric	ADJ
ejpam-5354	8	31	function	function	NOUN
ejpam-5354	8	32	value	value	NOUN
ejpam-5354	8	33	estimation	estimation	NOUN
ejpam-5354	8	34	.	.	PUNCT
ejpam-5354	9	1	to	to	PART
ejpam-5354	9	2	evaluate	evaluate	VERB
ejpam-5354	9	3	the	the	DET
ejpam-5354	9	4	efficiency	efficiency	NOUN
ejpam-5354	9	5	and	and	CCONJ
ejpam-5354	9	6	effectiveness	effectiveness	NOUN
ejpam-5354	9	7	of	of	ADP
ejpam-5354	9	8	our	our	PRON
ejpam-5354	9	9	modified	modify	VERB
ejpam-5354	9	10	approach	approach	NOUN
ejpam-5354	9	11	,	,	PUNCT
ejpam-5354	9	12	we	we	PRON
ejpam-5354	9	13	conducted	conduct	VERB
ejpam-5354	9	14	numerical	numerical	ADJ
ejpam-5354	9	15	experiments	experiment	NOUN
ejpam-5354	9	16	on	on	ADP
ejpam-5354	9	17	a	a	DET
ejpam-5354	9	18	set	set	NOUN
ejpam-5354	9	19	of	of	ADP
ejpam-5354	9	20	well	well	ADV
ejpam-5354	9	21	-	-	PUNCT
ejpam-5354	9	22	known	know	VERB
ejpam-5354	9	23	test	test	NOUN
ejpam-5354	9	24	functions	function	NOUN
ejpam-5354	9	25	.	.	PUNCT
ejpam-5354	10	1	these	these	DET
ejpam-5354	10	2	experiments	experiment	NOUN
ejpam-5354	10	3	reveal	reveal	VERB
ejpam-5354	10	4	that	that	SCONJ
ejpam-5354	10	5	our	our	PRON
ejpam-5354	10	6	algorithm	algorithm	NOUN
ejpam-5354	10	7	significantly	significantly	ADV
ejpam-5354	10	8	reduces	reduce	VERB
ejpam-5354	10	9	computational	computational	ADJ
ejpam-5354	10	10	time	time	NOUN
ejpam-5354	10	11	due	due	ADP
ejpam-5354	10	12	to	to	ADP
ejpam-5354	10	13	its	its	PRON
ejpam-5354	10	14	faster	fast	ADJ
ejpam-5354	10	15	convergence	convergence	NOUN
ejpam-5354	10	16	rates	rate	NOUN
ejpam-5354	10	17	and	and	CCONJ
ejpam-5354	10	18	increased	increase	VERB
ejpam-5354	10	19	speed	speed	NOUN
ejpam-5354	10	20	in	in	ADP
ejpam-5354	10	21	directional	directional	ADJ
ejpam-5354	10	22	minimization	minimization	NOUN
ejpam-5354	10	23	.	.	PUNCT
ejpam-5354	11	1	these	these	DET
ejpam-5354	11	2	compelling	compelling	ADJ
ejpam-5354	11	3	results	result	NOUN
ejpam-5354	11	4	highlight	highlight	VERB
ejpam-5354	11	5	the	the	DET
ejpam-5354	11	6	advantages	advantage	NOUN
ejpam-5354	11	7	of	of	ADP
ejpam-5354	11	8	our	our	PRON
ejpam-5354	11	9	approach	approach	NOUN
ejpam-5354	11	10	over	over	ADP
ejpam-5354	11	11	traditional	traditional	ADJ
ejpam-5354	11	12	conjugate	conjugate	ADJ
ejpam-5354	11	13	gradient	gradient	NOUN
ejpam-5354	11	14	methods	method	NOUN
ejpam-5354	11	15	in	in	ADP
ejpam-5354	11	16	the	the	DET
ejpam-5354	11	17	context	context	NOUN
ejpam-5354	11	18	of	of	ADP
ejpam-5354	11	19	unconstrained	unconstrained	ADJ
ejpam-5354	11	20	optimization	optimization	NOUN
ejpam-5354	11	21	problems	problem	NOUN
ejpam-5354	11	22	.	.	PUNCT
ejpam-5354	12	1	2020	2020	NUM
ejpam-5354	12	2	mathematics	mathematic	NOUN
ejpam-5354	12	3	subject	subject	NOUN
ejpam-5354	12	4	classifications	classification	NOUN
ejpam-5354	12	5	:	:	PUNCT
ejpam-5354	12	6	90c90	90c90	NUM
ejpam-5354	12	7	,	,	PUNCT
ejpam-5354	12	8	90c30	90c30	NUM
ejpam-5354	12	9	,	,	PUNCT
ejpam-5354	12	10	90c06	90c06	NUM
ejpam-5354	12	11	,	,	PUNCT
ejpam-5354	12	12	49m37	49m37	NUM
ejpam-5354	12	13	key	key	ADJ
ejpam-5354	12	14	words	word	NOUN
ejpam-5354	12	15	and	and	CCONJ
ejpam-5354	12	16	phrases	phrase	NOUN
ejpam-5354	12	17	:	:	PUNCT
ejpam-5354	12	18	optimization	optimization	NOUN
ejpam-5354	12	19	,	,	PUNCT
ejpam-5354	12	20	conjugate	conjugate	ADJ
ejpam-5354	12	21	gradient	gradient	NOUN
ejpam-5354	12	22	,	,	PUNCT
ejpam-5354	12	23	neural	neural	ADJ
ejpam-5354	12	24	networks	network	NOUN
ejpam-5354	12	25	,	,	PUNCT
ejpam-5354	12	26	convergence	convergence	NOUN
ejpam-5354	12	27	analysis	analysis	NOUN
ejpam-5354	12	28	∗corresponding	∗corresponde	VERB
ejpam-5354	12	29	author	author	NOUN
ejpam-5354	12	30	.	.	PUNCT
ejpam-5354	13	1	doi	doi	NOUN
ejpam-5354	13	2	:	:	PUNCT
ejpam-5354	13	3	https://doi.org/10.29020/nybg.ejpam.v17i4.5354	https://doi.org/10.29020/nybg.ejpam.v17i4.5354	PROPN
ejpam-5354	13	4	email	email	NOUN
ejpam-5354	13	5	addresses	address	NOUN
ejpam-5354	13	6	:	:	PUNCT
ejpam-5354	13	7	dilovan.omar@uoz.edu.krd	dilovan.omar@uoz.edu.krd	PROPN
ejpam-5354	13	8	(	(	PUNCT
ejpam-5354	13	9	d.	d.	PROPN
ejpam-5354	13	10	h.	h.	PROPN
ejpam-5354	13	11	omar	omar	PROPN
ejpam-5354	13	12	)	)	PUNCT
ejpam-5354	13	13	,	,	PUNCT
ejpam-5354	13	14	alaa.ibrahim@uoz.edu.krd	alaa.ibrahim@uoz.edu.krd	PROPN
ejpam-5354	13	15	(	(	PUNCT
ejpam-5354	13	16	alaa	alaa	PROPN
ejpam-5354	13	17	l.	l.	PROPN
ejpam-5354	13	18	ibrahim	ibrahim	PROPN
ejpam-5354	13	19	)	)	PUNCT
ejpam-5354	13	20	,	,	PUNCT
ejpam-5354	13	21	masoud.hassan@uoz.edu.krd	masoud.hassan@uoz.edu.krd	X
ejpam-5354	13	22	(	(	PUNCT
ejpam-5354	13	23	m.	m.	PROPN
ejpam-5354	13	24	m.	m.	PROPN
ejpam-5354	13	25	hassan	hassan	PROPN
ejpam-5354	13	26	)	)	PUNCT
ejpam-5354	13	27	,	,	PUNCT
ejpam-5354	13	28	bayda.fathi@uoz.edu.krd	bayda.fathi@uoz.edu.krd	PROPN
ejpam-5354	13	29	(	(	PUNCT
ejpam-5354	13	30	b.	b.	PROPN
ejpam-5354	13	31	g.	g.	PROPN
ejpam-5354	13	32	fathi	fathi	PROPN
ejpam-5354	13	33	)	)	PUNCT
ejpam-5354	13	34	,	,	PUNCT
ejpam-5354	14	1	dimanabdulqader90@gmail.com	dimanabdulqader90@gmail.com	PROPN
ejpam-5354	14	2	(	(	PUNCT
ejpam-5354	14	3	d.	d.	PROPN
ejpam-5354	14	4	a.	a.	PROPN
ejpam-5354	14	5	sulaiman	sulaiman	PROPN
ejpam-5354	14	6	)	)	PUNCT
ejpam-5354	14	7	https://www.ejpam.com	https://www.ejpam.com	X
ejpam-5354	14	8	2692	2692	NUM
ejpam-5354	15	1	copyright	copyright	NOUN
ejpam-5354	15	2	:	:	PUNCT
ejpam-5354	15	3	©	©	PROPN
ejpam-5354	15	4	2024	2024	NUM
ejpam-5354	15	5	the	the	DET
ejpam-5354	15	6	author(s	author(s	NOUN
ejpam-5354	15	7	)	)	PUNCT
ejpam-5354	15	8	.	.	PUNCT
ejpam-5354	16	1	(	(	PUNCT
ejpam-5354	16	2	cc	cc	NOUN
ejpam-5354	16	3	by	by	ADP
ejpam-5354	16	4	-	-	PUNCT
ejpam-5354	16	5	nc	nc	PROPN
ejpam-5354	16	6	4.0	4.0	NUM
ejpam-5354	16	7	)	)	PUNCT
ejpam-5354	17	1	alaa	alaa	PROPN
ejpam-5354	17	2	l.	l.	PROPN
ejpam-5354	17	3	ibrahim	ibrahim	PROPN
ejpam-5354	17	4	et	et	PROPN
ejpam-5354	17	5	al	al	PROPN
ejpam-5354	17	6	.	.	PUNCT
ejpam-5354	17	7	/	/	SYM
ejpam-5354	17	8	eur	eur	PROPN
ejpam-5354	17	9	.	.	PUNCT
ejpam-5354	18	1	j.	j.	PROPN
ejpam-5354	18	2	pure	pure	PROPN
ejpam-5354	18	3	appl	appl	PROPN
ejpam-5354	18	4	.	.	PROPN
ejpam-5354	18	5	math	math	PROPN
ejpam-5354	18	6	,	,	PUNCT
ejpam-5354	18	7	17	17	NUM
ejpam-5354	18	8	(	(	PUNCT
ejpam-5354	18	9	4	4	NUM
ejpam-5354	18	10	)	)	PUNCT
ejpam-5354	18	11	(	(	PUNCT
ejpam-5354	18	12	2024	2024	NUM
ejpam-5354	18	13	)	)	PUNCT
ejpam-5354	18	14	,	,	PUNCT
ejpam-5354	18	15	2692	2692	NUM
ejpam-5354	18	16	-	-	SYM
ejpam-5354	18	17	2705	2705	NUM
ejpam-5354	18	18	2693	2693	NUM
ejpam-5354	18	19	1	1	NUM
ejpam-5354	18	20	.	.	PUNCT
ejpam-5354	18	21	introduction	introduction	NOUN
ejpam-5354	18	22	the	the	DET
ejpam-5354	18	23	conjugate	conjugate	ADJ
ejpam-5354	18	24	gradient	gradient	NOUN
ejpam-5354	18	25	(	(	PUNCT
ejpam-5354	18	26	cg	cg	NOUN
ejpam-5354	18	27	)	)	PUNCT
ejpam-5354	18	28	method	method	NOUN
ejpam-5354	18	29	’s	’s	PART
ejpam-5354	18	30	ability	ability	NOUN
ejpam-5354	18	31	to	to	PART
ejpam-5354	18	32	effectively	effectively	ADV
ejpam-5354	18	33	solve	solve	VERB
ejpam-5354	18	34	a	a	DET
ejpam-5354	18	35	large	large	ADJ
ejpam-5354	18	36	number	number	NOUN
ejpam-5354	18	37	of	of	ADP
ejpam-5354	18	38	unconstrained	unconstrained	ADJ
ejpam-5354	18	39	optimization	optimization	NOUN
ejpam-5354	18	40	problems	problem	NOUN
ejpam-5354	18	41	in	in	ADP
ejpam-5354	18	42	a	a	DET
ejpam-5354	18	43	short	short	ADJ
ejpam-5354	18	44	amount	amount	NOUN
ejpam-5354	18	45	of	of	ADP
ejpam-5354	18	46	time	time	NOUN
ejpam-5354	18	47	and	and	CCONJ
ejpam-5354	18	48	with	with	ADP
ejpam-5354	18	49	fewer	few	ADJ
ejpam-5354	18	50	iterations	iteration	NOUN
ejpam-5354	18	51	is	be	AUX
ejpam-5354	18	52	one	one	NUM
ejpam-5354	18	53	of	of	ADP
ejpam-5354	18	54	its	its	PRON
ejpam-5354	18	55	main	main	ADJ
ejpam-5354	18	56	advantages	advantage	NOUN
ejpam-5354	18	57	.	.	PUNCT
ejpam-5354	19	1	because	because	SCONJ
ejpam-5354	19	2	cg	cg	NOUN
ejpam-5354	19	3	methods	method	NOUN
ejpam-5354	19	4	do	do	AUX
ejpam-5354	19	5	n’t	not	PART
ejpam-5354	19	6	rely	rely	VERB
ejpam-5354	19	7	on	on	ADP
ejpam-5354	19	8	the	the	DET
ejpam-5354	19	9	hessian	hessian	ADJ
ejpam-5354	19	10	matrix	matrix	NOUN
ejpam-5354	19	11	or	or	CCONJ
ejpam-5354	19	12	its	its	PRON
ejpam-5354	19	13	approximation	approximation	NOUN
ejpam-5354	19	14	,	,	PUNCT
ejpam-5354	19	15	they	they	PRON
ejpam-5354	19	16	have	have	VERB
ejpam-5354	19	17	lower	low	ADJ
ejpam-5354	19	18	computational	computational	ADJ
ejpam-5354	19	19	costs	cost	NOUN
ejpam-5354	19	20	and	and	CCONJ
ejpam-5354	19	21	storage	storage	NOUN
ejpam-5354	19	22	requirements	requirement	NOUN
ejpam-5354	19	23	.	.	PUNCT
ejpam-5354	20	1	furthermore	furthermore	ADV
ejpam-5354	20	2	,	,	PUNCT
ejpam-5354	20	3	the	the	DET
ejpam-5354	20	4	cg	cg	NOUN
ejpam-5354	20	5	method	method	NOUN
ejpam-5354	20	6	satisfies	satisfy	VERB
ejpam-5354	20	7	the	the	DET
ejpam-5354	20	8	descent	descent	NOUN
ejpam-5354	20	9	condition	condition	NOUN
ejpam-5354	20	10	and	and	CCONJ
ejpam-5354	20	11	exhibits	exhibit	VERB
ejpam-5354	20	12	fast	fast	ADJ
ejpam-5354	20	13	global	global	ADJ
ejpam-5354	20	14	convergence	convergence	NOUN
ejpam-5354	20	15	.	.	PUNCT
ejpam-5354	21	1	this	this	DET
ejpam-5354	21	2	method	method	NOUN
ejpam-5354	21	3	’s	’s	PART
ejpam-5354	21	4	unique	unique	ADJ
ejpam-5354	21	5	quality	quality	NOUN
ejpam-5354	21	6	is	be	AUX
ejpam-5354	21	7	its	its	PRON
ejpam-5354	21	8	simplicity	simplicity	NOUN
ejpam-5354	21	9	,	,	PUNCT
ejpam-5354	21	10	which	which	PRON
ejpam-5354	21	11	makes	make	VERB
ejpam-5354	21	12	the	the	DET
ejpam-5354	21	13	algebraic	algebraic	ADJ
ejpam-5354	21	14	processes	process	NOUN
ejpam-5354	21	15	and	and	CCONJ
ejpam-5354	21	16	writing	write	VERB
ejpam-5354	21	17	computer	computer	NOUN
ejpam-5354	21	18	code	code	NOUN
ejpam-5354	21	19	easy	easy	ADJ
ejpam-5354	21	20	to	to	PART
ejpam-5354	21	21	understand	understand	VERB
ejpam-5354	21	22	.	.	PUNCT
ejpam-5354	22	1	as	as	ADP
ejpam-5354	22	2	a	a	DET
ejpam-5354	22	3	result	result	NOUN
ejpam-5354	22	4	,	,	PUNCT
ejpam-5354	22	5	the	the	DET
ejpam-5354	22	6	approach	approach	NOUN
ejpam-5354	22	7	shows	show	VERB
ejpam-5354	22	8	promise	promise	NOUN
ejpam-5354	22	9	and	and	CCONJ
ejpam-5354	22	10	competence	competence	NOUN
ejpam-5354	22	11	in	in	ADP
ejpam-5354	22	12	handling	handle	VERB
ejpam-5354	22	13	large	large	ADJ
ejpam-5354	22	14	-	-	PUNCT
ejpam-5354	22	15	scale	scale	NOUN
ejpam-5354	22	16	unconstrained	unconstrained	ADJ
ejpam-5354	22	17	minimization	minimization	NOUN
ejpam-5354	22	18	problems	problem	NOUN
ejpam-5354	22	19	[	[	X
ejpam-5354	22	20	2	2	NUM
ejpam-5354	22	21	]	]	PUNCT
ejpam-5354	22	22	.	.	PUNCT
ejpam-5354	23	1	this	this	DET
ejpam-5354	23	2	study	study	NOUN
ejpam-5354	23	3	will	will	AUX
ejpam-5354	23	4	consider	consider	VERB
ejpam-5354	23	5	the	the	DET
ejpam-5354	23	6	following	follow	VERB
ejpam-5354	23	7	model	model	NOUN
ejpam-5354	23	8	minf	minf	PROPN
ejpam-5354	23	9	(	(	PUNCT
ejpam-5354	23	10	x	x	NOUN
ejpam-5354	23	11	)	)	PUNCT
ejpam-5354	23	12	∀	∀	X
ejpam-5354	24	1	x	x	SYM
ejpam-5354	24	2	∈	∈	PROPN
ejpam-5354	24	3	rn	rn	PROPN
ejpam-5354	24	4	,	,	PUNCT
ejpam-5354	24	5	(	(	PUNCT
ejpam-5354	24	6	1	1	X
ejpam-5354	24	7	)	)	PUNCT
ejpam-5354	24	8	where	where	SCONJ
ejpam-5354	24	9	f	f	PROPN
ejpam-5354	24	10	:	:	PUNCT
ejpam-5354	24	11	rn	rn	PROPN
ejpam-5354	24	12	→	→	SYM
ejpam-5354	24	13	r	r	NOUN
ejpam-5354	24	14	is	be	AUX
ejpam-5354	24	15	a	a	DET
ejpam-5354	24	16	continuously	continuously	ADV
ejpam-5354	24	17	differentiable	differentiable	ADJ
ejpam-5354	24	18	function	function	NOUN
ejpam-5354	24	19	,	,	PUNCT
ejpam-5354	24	20	gk	gk	PROPN
ejpam-5354	24	21	=	=	SYM
ejpam-5354	24	22	∇f	∇f	PROPN
ejpam-5354	24	23	(	(	PUNCT
ejpam-5354	24	24	xk	xk	INTJ
ejpam-5354	24	25	)	)	PUNCT
ejpam-5354	24	26	is	be	AUX
ejpam-5354	24	27	a	a	DET
ejpam-5354	24	28	gradient	gradient	NOUN
ejpam-5354	24	29	at	at	ADP
ejpam-5354	24	30	point	point	NOUN
ejpam-5354	24	31	xk	xk	PROPN
ejpam-5354	24	32	are	be	AUX
ejpam-5354	24	33	solved	solve	VERB
ejpam-5354	24	34	using	use	VERB
ejpam-5354	24	35	this	this	DET
ejpam-5354	24	36	expression	expression	NOUN
ejpam-5354	24	37	xk+1	xk+1	PUNCT
ejpam-5354	25	1	=	=	SYM
ejpam-5354	25	2	xk	xk	PROPN
ejpam-5354	26	1	+	+	CCONJ
ejpam-5354	26	2	αkdk	αkdk	NOUN
ejpam-5354	26	3	k	k	PROPN
ejpam-5354	26	4	=	=	SYM
ejpam-5354	26	5	0	0	NUM
ejpam-5354	26	6	,	,	PUNCT
ejpam-5354	26	7	1	1	NUM
ejpam-5354	26	8	,	,	PUNCT
ejpam-5354	26	9	2	2	NUM
ejpam-5354	26	10	,	,	PUNCT
ejpam-5354	26	11	.	.	PUNCT
ejpam-5354	26	12	.	.	PUNCT
ejpam-5354	27	1	.	.	PUNCT
ejpam-5354	28	1	(	(	PUNCT
ejpam-5354	28	2	2	2	X
ejpam-5354	28	3	)	)	PUNCT
ejpam-5354	28	4	where	where	SCONJ
ejpam-5354	28	5	xk	xk	PROPN
ejpam-5354	28	6	is	be	AUX
ejpam-5354	28	7	a	a	DET
ejpam-5354	28	8	solution	solution	NOUN
ejpam-5354	28	9	in	in	ADP
ejpam-5354	28	10	current	current	ADJ
ejpam-5354	28	11	iteration	iteration	NOUN
ejpam-5354	28	12	and	and	CCONJ
ejpam-5354	28	13	xk	xk	PROPN
ejpam-5354	28	14	is	be	AUX
ejpam-5354	28	15	a	a	DET
ejpam-5354	28	16	new	new	ADJ
ejpam-5354	28	17	iteration	iteration	NOUN
ejpam-5354	28	18	point	point	NOUN
ejpam-5354	28	19	,	,	PUNCT
ejpam-5354	28	20	αk	αk	SCONJ
ejpam-5354	28	21	>	>	X
ejpam-5354	28	22	0	0	NUM
ejpam-5354	28	23	is	be	AUX
ejpam-5354	28	24	a	a	DET
ejpam-5354	28	25	step	step	NOUN
ejpam-5354	28	26	size	size	NOUN
ejpam-5354	28	27	determined	determine	VERB
ejpam-5354	28	28	by	by	ADP
ejpam-5354	28	29	some	some	DET
ejpam-5354	28	30	line	line	NOUN
ejpam-5354	28	31	searches	search	NOUN
ejpam-5354	28	32	.	.	PUNCT
ejpam-5354	29	1	dk	dk	PROPN
ejpam-5354	29	2	is	be	AUX
ejpam-5354	29	3	a	a	DET
ejpam-5354	29	4	search	search	NOUN
ejpam-5354	29	5	direction	direction	NOUN
ejpam-5354	29	6	which	which	PRON
ejpam-5354	29	7	is	be	AUX
ejpam-5354	29	8	determined	determine	VERB
ejpam-5354	29	9	by	by	ADP
ejpam-5354	29	10	:	:	PUNCT
ejpam-5354	29	11	d0	d0	NOUN
ejpam-5354	29	12	=	=	SYM
ejpam-5354	29	13	−g0	−g0	PROPN
ejpam-5354	29	14	and	and	CCONJ
ejpam-5354	29	15	dk+1	dk+1	NOUN
ejpam-5354	29	16	=	=	SYM
ejpam-5354	29	17	−gk+1	−gk+1	PROPN
ejpam-5354	29	18	+	+	CCONJ
ejpam-5354	29	19	βkdk	βkdk	NOUN
ejpam-5354	29	20	for	for	ADP
ejpam-5354	29	21	k	k	PROPN
ejpam-5354	29	22	≥	≥	PROPN
ejpam-5354	29	23	1	1	NUM
ejpam-5354	29	24	,	,	PUNCT
ejpam-5354	29	25	(	(	PUNCT
ejpam-5354	29	26	3	3	X
ejpam-5354	29	27	)	)	PUNCT
ejpam-5354	29	28	where	where	SCONJ
ejpam-5354	29	29	βk	βk	NOUN
ejpam-5354	29	30	is	be	AUX
ejpam-5354	29	31	an	an	DET
ejpam-5354	29	32	important	important	ADJ
ejpam-5354	29	33	parameter	parameter	NOUN
ejpam-5354	29	34	.	.	PUNCT
ejpam-5354	30	1	the	the	DET
ejpam-5354	30	2	different	different	ADJ
ejpam-5354	30	3	choices	choice	NOUN
ejpam-5354	30	4	for	for	ADP
ejpam-5354	30	5	the	the	DET
ejpam-5354	30	6	parameter	parameter	NOUN
ejpam-5354	30	7	βk	βk	ADV
ejpam-5354	30	8	correspond	correspond	ADV
ejpam-5354	30	9	to	to	ADP
ejpam-5354	30	10	different	different	ADJ
ejpam-5354	30	11	cg	cg	NOUN
ejpam-5354	30	12	methods	method	NOUN
ejpam-5354	30	13	.	.	PUNCT
ejpam-5354	31	1	the	the	DET
ejpam-5354	31	2	fletcher	fletcher	PROPN
ejpam-5354	31	3	-	-	PUNCT
ejpam-5354	31	4	reeves	reeves	PROPN
ejpam-5354	31	5	(	(	PUNCT
ejpam-5354	31	6	fr	fr	NOUN
ejpam-5354	31	7	)	)	PUNCT
ejpam-5354	31	8	method	method	NOUN
ejpam-5354	31	9	[	[	X
ejpam-5354	31	10	12	12	NUM
ejpam-5354	31	11	]	]	PUNCT
ejpam-5354	31	12	,	,	PUNCT
ejpam-5354	31	13	the	the	DET
ejpam-5354	31	14	polak	polak	NOUN
ejpam-5354	31	15	-	-	PUNCT
ejpam-5354	31	16	ribiterepolyak	ribiterepolyak	NOUN
ejpam-5354	31	17	(	(	PUNCT
ejpam-5354	31	18	prp	prp	NOUN
ejpam-5354	31	19	)	)	PUNCT
ejpam-5354	31	20	method	method	NOUN
ejpam-5354	31	21	[	[	X
ejpam-5354	31	22	29	29	NUM
ejpam-5354	31	23	,	,	PUNCT
ejpam-5354	31	24	30	30	NUM
ejpam-5354	31	25	]	]	PUNCT
ejpam-5354	31	26	,	,	PUNCT
ejpam-5354	31	27	the	the	DET
ejpam-5354	31	28	dai	dai	PROPN
ejpam-5354	31	29	-	-	PUNCT
ejpam-5354	31	30	yuan	yuan	PROPN
ejpam-5354	31	31	(	(	PUNCT
ejpam-5354	31	32	dy	dy	NOUN
ejpam-5354	31	33	)	)	PUNCT
ejpam-5354	31	34	method	method	NOUN
ejpam-5354	31	35	[	[	X
ejpam-5354	31	36	8	8	NUM
ejpam-5354	31	37	]	]	PUNCT
ejpam-5354	31	38	,	,	PUNCT
ejpam-5354	31	39	the	the	DET
ejpam-5354	31	40	conjugate	conjugate	ADJ
ejpam-5354	31	41	descent	descent	NOUN
ejpam-5354	31	42	(	(	PUNCT
ejpam-5354	31	43	cd	cd	PROPN
ejpam-5354	31	44	)	)	PUNCT
ejpam-5354	31	45	method	method	NOUN
ejpam-5354	31	46	[	[	X
ejpam-5354	31	47	9	9	NUM
ejpam-5354	31	48	]	]	PUNCT
ejpam-5354	31	49	,	,	PUNCT
ejpam-5354	31	50	and	and	CCONJ
ejpam-5354	31	51	others	other	NOUN
ejpam-5354	32	1	[	[	X
ejpam-5354	32	2	20	20	NUM
ejpam-5354	32	3	]	]	PUNCT
ejpam-5354	32	4	,	,	PUNCT
ejpam-5354	32	5	are	be	AUX
ejpam-5354	32	6	some	some	PRON
ejpam-5354	32	7	of	of	ADP
ejpam-5354	32	8	the	the	DET
ejpam-5354	32	9	well	well	ADV
ejpam-5354	32	10	-	-	PUNCT
ejpam-5354	32	11	known	know	VERB
ejpam-5354	32	12	cg	cg	NOUN
ejpam-5354	32	13	methods	method	NOUN
ejpam-5354	32	14	used	use	VERB
ejpam-5354	32	15	in	in	ADP
ejpam-5354	32	16	the	the	DET
ejpam-5354	32	17	study	study	NOUN
ejpam-5354	32	18	of	of	ADP
ejpam-5354	32	19	optimization	optimization	NOUN
ejpam-5354	32	20	problems	problem	NOUN
ejpam-5354	32	21	to	to	PART
ejpam-5354	32	22	obtain	obtain	VERB
ejpam-5354	32	23	the	the	DET
ejpam-5354	32	24	solution	solution	NOUN
ejpam-5354	32	25	of	of	ADP
ejpam-5354	32	26	the	the	DET
ejpam-5354	32	27	models	model	NOUN
ejpam-5354	32	28	.	.	PUNCT
ejpam-5354	33	1	a	a	DET
ejpam-5354	33	2	number	number	NOUN
ejpam-5354	33	3	of	of	ADP
ejpam-5354	33	4	researchers	researcher	NOUN
ejpam-5354	33	5	[	[	X
ejpam-5354	33	6	4	4	NUM
ejpam-5354	33	7	]	]	PUNCT
ejpam-5354	33	8	,	,	PUNCT
ejpam-5354	33	9	have	have	AUX
ejpam-5354	33	10	discussed	discuss	VERB
ejpam-5354	33	11	the	the	DET
ejpam-5354	33	12	convergence	convergence	NOUN
ejpam-5354	33	13	of	of	ADP
ejpam-5354	33	14	the	the	DET
ejpam-5354	33	15	well	well	ADV
ejpam-5354	33	16	-	-	PUNCT
ejpam-5354	33	17	known	know	VERB
ejpam-5354	33	18	classical	classical	ADJ
ejpam-5354	33	19	cg	cg	NOUN
ejpam-5354	33	20	methods	method	NOUN
ejpam-5354	33	21	,	,	PUNCT
ejpam-5354	33	22	such	such	ADJ
ejpam-5354	33	23	as	as	ADP
ejpam-5354	33	24	hestenes	hestene	NOUN
ejpam-5354	33	25	–	–	PUNCT
ejpam-5354	33	26	stiefel	stiefel	NOUN
ejpam-5354	33	27	(	(	PUNCT
ejpam-5354	33	28	hs	hs	X
ejpam-5354	33	29	)	)	PUNCT
ejpam-5354	34	1	[	[	X
ejpam-5354	34	2	16	16	NUM
ejpam-5354	34	3	]	]	PUNCT
ejpam-5354	34	4	,	,	PUNCT
ejpam-5354	34	5	polak	polak	PROPN
ejpam-5354	34	6	–	–	PUNCT
ejpam-5354	34	7	ribiere	ribiere	X
ejpam-5354	34	8	–	–	PUNCT
ejpam-5354	34	9	polyak	polyak	NOUN
ejpam-5354	34	10	(	(	PUNCT
ejpam-5354	34	11	prp	prp	PROPN
ejpam-5354	34	12	)	)	PUNCT
ejpam-5354	35	1	[	[	X
ejpam-5354	35	2	13	13	NUM
ejpam-5354	35	3	,	,	PUNCT
ejpam-5354	35	4	30	30	NUM
ejpam-5354	35	5	]	]	PUNCT
ejpam-5354	35	6	,	,	PUNCT
ejpam-5354	35	7	fletcher	fletcher	PROPN
ejpam-5354	35	8	–	–	PUNCT
ejpam-5354	35	9	reeves	reeves	PROPN
ejpam-5354	35	10	(	(	PUNCT
ejpam-5354	35	11	fr	fr	PROPN
ejpam-5354	35	12	)	)	PUNCT
ejpam-5354	36	1	[	[	X
ejpam-5354	36	2	11	11	NUM
ejpam-5354	36	3	]	]	PUNCT
ejpam-5354	36	4	,	,	PUNCT
ejpam-5354	36	5	liu	liu	PROPN
ejpam-5354	36	6	–	–	PUNCT
ejpam-5354	36	7	storey	storey	NOUN
ejpam-5354	36	8	(	(	PUNCT
ejpam-5354	36	9	ls	ls	PROPN
ejpam-5354	36	10	)	)	PUNCT
ejpam-5354	37	1	[	[	X
ejpam-5354	37	2	26	26	NUM
ejpam-5354	37	3	]	]	PUNCT
ejpam-5354	37	4	,	,	PUNCT
ejpam-5354	37	5	conjugate	conjugate	ADJ
ejpam-5354	37	6	descent	descent	NOUN
ejpam-5354	37	7	(	(	PUNCT
ejpam-5354	37	8	cd	cd	PROPN
ejpam-5354	37	9	)	)	PUNCT
ejpam-5354	38	1	[	[	X
ejpam-5354	38	2	10	10	NUM
ejpam-5354	38	3	]	]	PUNCT
ejpam-5354	38	4	,	,	PUNCT
ejpam-5354	38	5	and	and	CCONJ
ejpam-5354	38	6	dai	dai	PROPN
ejpam-5354	38	7	–	–	PROPN
ejpam-5354	38	8	yuan	yuan	PROPN
ejpam-5354	38	9	(	(	PUNCT
ejpam-5354	38	10	dy	dy	X
ejpam-5354	38	11	)	)	PUNCT
ejpam-5354	39	1	[	[	X
ejpam-5354	39	2	7	7	NUM
ejpam-5354	39	3	]	]	PUNCT
ejpam-5354	39	4	.	.	PUNCT
ejpam-5354	40	1	although	although	SCONJ
ejpam-5354	40	2	the	the	DET
ejpam-5354	40	3	prp	prp	NOUN
ejpam-5354	40	4	formula	formula	NOUN
ejpam-5354	40	5	is	be	AUX
ejpam-5354	40	6	thought	think	VERB
ejpam-5354	40	7	to	to	PART
ejpam-5354	40	8	be	be	AUX
ejpam-5354	40	9	the	the	DET
ejpam-5354	40	10	most	most	ADV
ejpam-5354	40	11	efficient	efficient	ADJ
ejpam-5354	40	12	in	in	ADP
ejpam-5354	40	13	terms	term	NOUN
ejpam-5354	40	14	of	of	ADP
ejpam-5354	40	15	numerical	numerical	ADJ
ejpam-5354	40	16	computation	computation	NOUN
ejpam-5354	40	17	,	,	PUNCT
ejpam-5354	40	18	there	there	PRON
ejpam-5354	40	19	is	be	VERB
ejpam-5354	40	20	no	no	DET
ejpam-5354	40	21	guarantee	guarantee	NOUN
ejpam-5354	40	22	that	that	SCONJ
ejpam-5354	40	23	it	it	PRON
ejpam-5354	40	24	will	will	AUX
ejpam-5354	40	25	converge	converge	VERB
ejpam-5354	40	26	under	under	ADP
ejpam-5354	40	27	multiple	multiple	ADJ
ejpam-5354	40	28	line	line	NOUN
ejpam-5354	40	29	searches	search	NOUN
ejpam-5354	40	30	[	[	X
ejpam-5354	40	31	31	31	NUM
ejpam-5354	40	32	]	]	PUNCT
ejpam-5354	40	33	.	.	PUNCT
ejpam-5354	41	1	this	this	DET
ejpam-5354	41	2	disadvantage	disadvantage	NOUN
ejpam-5354	41	3	prompted	prompt	VERB
ejpam-5354	41	4	several	several	ADJ
ejpam-5354	41	5	changes	change	NOUN
ejpam-5354	41	6	to	to	ADP
ejpam-5354	41	7	the	the	DET
ejpam-5354	41	8	prp	prp	PROPN
ejpam-5354	41	9	parameter	parameter	NOUN
ejpam-5354	41	10	.	.	PUNCT
ejpam-5354	42	1	for	for	ADP
ejpam-5354	42	2	in	in	ADP
ejpam-5354	42	3	-	-	PUNCT
ejpam-5354	42	4	depth	depth	NOUN
ejpam-5354	42	5	discussions	discussion	NOUN
ejpam-5354	42	6	on	on	ADP
ejpam-5354	42	7	the	the	DET
ejpam-5354	42	8	prp	prp	NOUN
ejpam-5354	42	9	technique	technique	NOUN
ejpam-5354	42	10	,	,	PUNCT
ejpam-5354	42	11	see	see	VERB
ejpam-5354	42	12	yuan	yuan	PROPN
ejpam-5354	42	13	et	et	NOUN
ejpam-5354	42	14	al	al	PROPN
ejpam-5354	42	15	.	.	PUNCT
ejpam-5354	43	1	[	[	X
ejpam-5354	43	2	36	36	NUM
ejpam-5354	43	3	]	]	PUNCT
ejpam-5354	43	4	,	,	PUNCT
ejpam-5354	43	5	andrei	andrei	NOUN
ejpam-5354	43	6	[	[	X
ejpam-5354	43	7	4	4	NUM
ejpam-5354	43	8	]	]	PUNCT
ejpam-5354	43	9	,	,	PUNCT
ejpam-5354	43	10	and	and	CCONJ
ejpam-5354	43	11	zhang	zhang	PROPN
ejpam-5354	43	12	et	et	PROPN
ejpam-5354	43	13	al	al	PROPN
ejpam-5354	43	14	.	.	PUNCT
ejpam-5354	44	1	[	[	X
ejpam-5354	44	2	37	37	NUM
ejpam-5354	44	3	]	]	PUNCT
ejpam-5354	44	4	.	.	PUNCT
ejpam-5354	45	1	recently	recently	ADV
ejpam-5354	45	2	,	,	PUNCT
ejpam-5354	45	3	there	there	PRON
ejpam-5354	45	4	has	have	AUX
ejpam-5354	45	5	been	be	AUX
ejpam-5354	45	6	a	a	DET
ejpam-5354	45	7	significant	significant	ADJ
ejpam-5354	45	8	trend	trend	NOUN
ejpam-5354	45	9	among	among	ADP
ejpam-5354	45	10	researchers	researcher	NOUN
ejpam-5354	45	11	to	to	PART
ejpam-5354	45	12	develop	develop	VERB
ejpam-5354	45	13	the	the	DET
ejpam-5354	45	14	best	good	ADJ
ejpam-5354	45	15	and	and	CCONJ
ejpam-5354	45	16	fastest	fast	ADJ
ejpam-5354	45	17	optimization	optimization	NOUN
ejpam-5354	45	18	methods	method	NOUN
ejpam-5354	45	19	,	,	PUNCT
ejpam-5354	45	20	driven	drive	VERB
ejpam-5354	45	21	by	by	ADP
ejpam-5354	45	22	the	the	DET
ejpam-5354	45	23	increasing	increase	VERB
ejpam-5354	45	24	number	number	NOUN
ejpam-5354	45	25	of	of	ADP
ejpam-5354	45	26	real	real	ADJ
ejpam-5354	45	27	-	-	PUNCT
ejpam-5354	45	28	world	world	NOUN
ejpam-5354	45	29	applications	application	NOUN
ejpam-5354	45	30	that	that	PRON
ejpam-5354	45	31	rely	rely	VERB
ejpam-5354	45	32	on	on	ADP
ejpam-5354	45	33	optimization	optimization	NOUN
ejpam-5354	45	34	techniques	technique	NOUN
ejpam-5354	45	35	.	.	PUNCT
ejpam-5354	46	1	these	these	DET
ejpam-5354	46	2	applications	application	NOUN
ejpam-5354	46	3	span	span	VERB
ejpam-5354	46	4	various	various	ADJ
ejpam-5354	46	5	fields	field	NOUN
ejpam-5354	46	6	,	,	PUNCT
ejpam-5354	46	7	including	include	VERB
ejpam-5354	46	8	machine	machine	NOUN
ejpam-5354	46	9	learning	learning	NOUN
ejpam-5354	46	10	,	,	PUNCT
ejpam-5354	46	11	engineering	engineering	NOUN
ejpam-5354	46	12	design	design	NOUN
ejpam-5354	46	13	,	,	PUNCT
ejpam-5354	46	14	logistics	logistic	NOUN
ejpam-5354	46	15	,	,	PUNCT
ejpam-5354	46	16	finance	finance	NOUN
ejpam-5354	46	17	,	,	PUNCT
ejpam-5354	46	18	and	and	CCONJ
ejpam-5354	46	19	healthcare	healthcare	PROPN
ejpam-5354	46	20	,	,	PUNCT
ejpam-5354	46	21	where	where	SCONJ
ejpam-5354	46	22	efficient	efficient	ADJ
ejpam-5354	46	23	and	and	CCONJ
ejpam-5354	46	24	accurate	accurate	ADJ
ejpam-5354	46	25	optimization	optimization	NOUN
ejpam-5354	46	26	is	be	AUX
ejpam-5354	46	27	crucial	crucial	ADJ
ejpam-5354	46	28	for	for	ADP
ejpam-5354	46	29	improving	improve	VERB
ejpam-5354	46	30	performance	performance	NOUN
ejpam-5354	46	31	,	,	PUNCT
ejpam-5354	46	32	reducing	reduce	VERB
ejpam-5354	46	33	costs	cost	NOUN
ejpam-5354	46	34	,	,	PUNCT
ejpam-5354	46	35	and	and	CCONJ
ejpam-5354	46	36	solving	solve	VERB
ejpam-5354	46	37	complex	complex	ADJ
ejpam-5354	46	38	problems	problem	NOUN
ejpam-5354	46	39	.	.	PUNCT
ejpam-5354	47	1	the	the	DET
ejpam-5354	47	2	quest	quest	NOUN
ejpam-5354	47	3	for	for	ADP
ejpam-5354	47	4	more	more	ADV
ejpam-5354	47	5	effective	effective	ADJ
ejpam-5354	47	6	optimization	optimization	NOUN
ejpam-5354	47	7	methods	method	NOUN
ejpam-5354	47	8	is	be	AUX
ejpam-5354	47	9	fueled	fuel	VERB
ejpam-5354	47	10	by	by	ADP
ejpam-5354	47	11	the	the	DET
ejpam-5354	47	12	need	need	NOUN
ejpam-5354	47	13	to	to	PART
ejpam-5354	47	14	handle	handle	VERB
ejpam-5354	47	15	larger	large	ADJ
ejpam-5354	47	16	datasets	dataset	NOUN
ejpam-5354	47	17	,	,	PUNCT
ejpam-5354	47	18	higher	higher	ADV
ejpam-5354	47	19	-	-	PUNCT
ejpam-5354	47	20	dimensional	dimensional	ADJ
ejpam-5354	47	21	problems	problem	NOUN
ejpam-5354	47	22	,	,	PUNCT
ejpam-5354	47	23	and	and	CCONJ
ejpam-5354	47	24	more	more	ADV
ejpam-5354	47	25	intricate	intricate	ADJ
ejpam-5354	47	26	models	model	NOUN
ejpam-5354	47	27	,	,	PUNCT
ejpam-5354	47	28	making	make	VERB
ejpam-5354	47	29	optimization	optimization	NOUN
ejpam-5354	47	30	a	a	DET
ejpam-5354	47	31	key	key	ADJ
ejpam-5354	47	32	area	area	NOUN
ejpam-5354	47	33	of	of	ADP
ejpam-5354	47	34	research	research	NOUN
ejpam-5354	47	35	and	and	CCONJ
ejpam-5354	47	36	innovation	innovation	NOUN
ejpam-5354	47	37	see	see	VERB
ejpam-5354	47	38	[	[	X
ejpam-5354	47	39	1	1	NUM
ejpam-5354	47	40	,	,	PUNCT
ejpam-5354	47	41	6	6	NUM
ejpam-5354	47	42	,	,	PUNCT
ejpam-5354	47	43	15	15	NUM
ejpam-5354	47	44	,	,	PUNCT
ejpam-5354	47	45	19	19	NUM
ejpam-5354	47	46	,	,	PUNCT
ejpam-5354	47	47	22	22	NUM
ejpam-5354	47	48	,	,	PUNCT
ejpam-5354	47	49	24	24	NUM
ejpam-5354	47	50	,	,	PUNCT
ejpam-5354	47	51	34	34	NUM
ejpam-5354	47	52	]	]	PUNCT
ejpam-5354	47	53	.	.	PUNCT
ejpam-5354	48	1	alaa	alaa	PROPN
ejpam-5354	48	2	l.	l.	PROPN
ejpam-5354	48	3	ibrahim	ibrahim	PROPN
ejpam-5354	48	4	et	et	PROPN
ejpam-5354	48	5	al	al	PROPN
ejpam-5354	48	6	.	.	PUNCT
ejpam-5354	48	7	/	/	SYM
ejpam-5354	48	8	eur	eur	PROPN
ejpam-5354	48	9	.	.	PUNCT
ejpam-5354	49	1	j.	j.	PROPN
ejpam-5354	49	2	pure	pure	PROPN
ejpam-5354	49	3	appl	appl	PROPN
ejpam-5354	49	4	.	.	PROPN
ejpam-5354	49	5	math	math	PROPN
ejpam-5354	49	6	,	,	PUNCT
ejpam-5354	49	7	17	17	NUM
ejpam-5354	49	8	(	(	PUNCT
ejpam-5354	49	9	4	4	NUM
ejpam-5354	49	10	)	)	PUNCT
ejpam-5354	49	11	(	(	PUNCT
ejpam-5354	49	12	2024	2024	NUM
ejpam-5354	49	13	)	)	PUNCT
ejpam-5354	49	14	,	,	PUNCT
ejpam-5354	49	15	2692	2692	NUM
ejpam-5354	49	16	-	-	SYM
ejpam-5354	49	17	2705	2705	NUM
ejpam-5354	49	18	2694	2694	NUM
ejpam-5354	49	19	due	due	ADP
ejpam-5354	49	20	to	to	ADP
ejpam-5354	49	21	their	their	PRON
ejpam-5354	49	22	exceptional	exceptional	ADJ
ejpam-5354	49	23	capacity	capacity	NOUN
ejpam-5354	49	24	for	for	ADP
ejpam-5354	49	25	self	self	NOUN
ejpam-5354	49	26	-	-	PUNCT
ejpam-5354	49	27	learning	learn	VERB
ejpam-5354	49	28	and	and	CCONJ
ejpam-5354	49	29	self	self	NOUN
ejpam-5354	49	30	-	-	PUNCT
ejpam-5354	49	31	adaptation	adaptation	NOUN
ejpam-5354	49	32	,	,	PUNCT
ejpam-5354	49	33	artificial	artificial	ADJ
ejpam-5354	49	34	neural	neural	ADJ
ejpam-5354	49	35	networks	network	NOUN
ejpam-5354	49	36	(	(	PUNCT
ejpam-5354	49	37	anns	anns	PROPN
ejpam-5354	49	38	)	)	PUNCT
ejpam-5354	49	39	have	have	AUX
ejpam-5354	49	40	been	be	AUX
ejpam-5354	49	41	employed	employ	VERB
ejpam-5354	49	42	successfully	successfully	ADV
ejpam-5354	49	43	in	in	ADP
ejpam-5354	49	44	numerous	numerous	ADJ
ejpam-5354	49	45	machine	machine	NOUN
ejpam-5354	49	46	learning	learning	NOUN
ejpam-5354	49	47	applications	application	NOUN
ejpam-5354	49	48	for	for	ADP
ejpam-5354	49	49	decades	decade	NOUN
ejpam-5354	49	50	[	[	X
ejpam-5354	49	51	5	5	NUM
ejpam-5354	49	52	,	,	PUNCT
ejpam-5354	49	53	17	17	NUM
ejpam-5354	49	54	,	,	PUNCT
ejpam-5354	49	55	35	35	NUM
ejpam-5354	49	56	]	]	PUNCT
ejpam-5354	49	57	.	.	PUNCT
ejpam-5354	50	1	they	they	PRON
ejpam-5354	50	2	have	have	AUX
ejpam-5354	50	3	been	be	AUX
ejpam-5354	50	4	heavily	heavily	ADV
ejpam-5354	50	5	utilized	utilize	VERB
ejpam-5354	50	6	in	in	ADP
ejpam-5354	50	7	fields	field	NOUN
ejpam-5354	50	8	like	like	ADP
ejpam-5354	50	9	robotics	robotic	NOUN
ejpam-5354	50	10	,	,	PUNCT
ejpam-5354	50	11	security	security	NOUN
ejpam-5354	50	12	,	,	PUNCT
ejpam-5354	50	13	and	and	CCONJ
ejpam-5354	50	14	self	self	NOUN
ejpam-5354	50	15	-	-	PUNCT
ejpam-5354	50	16	driving	drive	VERB
ejpam-5354	50	17	automobiles	automobile	NOUN
ejpam-5354	50	18	.	.	PUNCT
ejpam-5354	51	1	because	because	SCONJ
ejpam-5354	51	2	of	of	ADP
ejpam-5354	51	3	their	their	PRON
ejpam-5354	51	4	ability	ability	NOUN
ejpam-5354	51	5	to	to	PART
ejpam-5354	51	6	solve	solve	VERB
ejpam-5354	51	7	problems	problem	NOUN
ejpam-5354	51	8	with	with	ADP
ejpam-5354	51	9	flexibility	flexibility	NOUN
ejpam-5354	51	10	and	and	CCONJ
ejpam-5354	51	11	support	support	NOUN
ejpam-5354	51	12	for	for	ADP
ejpam-5354	51	13	parallel	parallel	ADJ
ejpam-5354	51	14	processing	processing	NOUN
ejpam-5354	51	15	,	,	PUNCT
ejpam-5354	51	16	they	they	PRON
ejpam-5354	51	17	are	be	AUX
ejpam-5354	51	18	frequently	frequently	ADV
ejpam-5354	51	19	found	find	VERB
ejpam-5354	51	20	to	to	PART
ejpam-5354	51	21	be	be	AUX
ejpam-5354	51	22	more	more	ADV
ejpam-5354	51	23	active	active	ADJ
ejpam-5354	51	24	and	and	CCONJ
ejpam-5354	51	25	precise	precise	ADJ
ejpam-5354	51	26	than	than	ADP
ejpam-5354	51	27	other	other	ADJ
ejpam-5354	51	28	classification	classification	NOUN
ejpam-5354	51	29	techniques	technique	NOUN
ejpam-5354	51	30	[	[	X
ejpam-5354	51	31	25	25	NUM
ejpam-5354	51	32	,	,	PUNCT
ejpam-5354	51	33	33	33	NUM
ejpam-5354	51	34	]	]	PUNCT
ejpam-5354	51	35	.	.	PUNCT
ejpam-5354	52	1	while	while	SCONJ
ejpam-5354	52	2	there	there	PRON
ejpam-5354	52	3	are	be	VERB
ejpam-5354	52	4	a	a	DET
ejpam-5354	52	5	number	number	NOUN
ejpam-5354	52	6	of	of	ADP
ejpam-5354	52	7	proposed	propose	VERB
ejpam-5354	52	8	training	training	NOUN
ejpam-5354	52	9	techniques	technique	NOUN
ejpam-5354	52	10	,	,	PUNCT
ejpam-5354	52	11	feed	feed	VERB
ejpam-5354	52	12	forward	forward	ADV
ejpam-5354	52	13	neural	neural	ADJ
ejpam-5354	52	14	networks	network	NOUN
ejpam-5354	52	15	(	(	PUNCT
ejpam-5354	52	16	fnns	fnn	NOUN
ejpam-5354	52	17	)	)	PUNCT
ejpam-5354	52	18	are	be	AUX
ejpam-5354	52	19	one	one	NUM
ejpam-5354	52	20	of	of	ADP
ejpam-5354	52	21	the	the	DET
ejpam-5354	52	22	most	most	ADV
ejpam-5354	52	23	well	well	ADV
ejpam-5354	52	24	-	-	PUNCT
ejpam-5354	52	25	known	know	VERB
ejpam-5354	52	26	and	and	CCONJ
ejpam-5354	52	27	frequently	frequently	ADV
ejpam-5354	52	28	applied	apply	VERB
ejpam-5354	52	29	training	training	NOUN
ejpam-5354	52	30	styles	style	NOUN
ejpam-5354	52	31	across	across	ADP
ejpam-5354	52	32	a	a	DET
ejpam-5354	52	33	wide	wide	ADJ
ejpam-5354	52	34	range	range	NOUN
ejpam-5354	52	35	of	of	ADP
ejpam-5354	52	36	domains	domain	NOUN
ejpam-5354	52	37	and	and	CCONJ
ejpam-5354	52	38	applications	application	NOUN
ejpam-5354	52	39	.	.	PUNCT
ejpam-5354	53	1	in	in	ADP
ejpam-5354	53	2	recent	recent	ADJ
ejpam-5354	53	3	times	time	NOUN
ejpam-5354	53	4	,	,	PUNCT
ejpam-5354	53	5	there	there	PRON
ejpam-5354	53	6	has	have	AUX
ejpam-5354	53	7	been	be	AUX
ejpam-5354	53	8	a	a	DET
ejpam-5354	53	9	growing	grow	VERB
ejpam-5354	53	10	trend	trend	NOUN
ejpam-5354	53	11	towards	towards	ADP
ejpam-5354	53	12	utilizing	utilize	VERB
ejpam-5354	53	13	the	the	DET
ejpam-5354	53	14	conjugate	conjugate	ADJ
ejpam-5354	53	15	gradient	gradient	NOUN
ejpam-5354	53	16	method	method	NOUN
ejpam-5354	53	17	for	for	ADP
ejpam-5354	53	18	training	train	VERB
ejpam-5354	53	19	neural	neural	ADJ
ejpam-5354	53	20	networks	network	NOUN
ejpam-5354	53	21	,	,	PUNCT
ejpam-5354	53	22	with	with	ADP
ejpam-5354	53	23	various	various	ADJ
ejpam-5354	53	24	new	new	ADJ
ejpam-5354	53	25	approaches	approach	NOUN
ejpam-5354	53	26	and	and	CCONJ
ejpam-5354	53	27	techniques	technique	NOUN
ejpam-5354	53	28	being	be	AUX
ejpam-5354	53	29	introduced	introduce	VERB
ejpam-5354	53	30	to	to	PART
ejpam-5354	53	31	enhance	enhance	VERB
ejpam-5354	53	32	its	its	PRON
ejpam-5354	53	33	efficiency	efficiency	NOUN
ejpam-5354	53	34	and	and	CCONJ
ejpam-5354	53	35	applicability	applicability	NOUN
ejpam-5354	53	36	in	in	ADP
ejpam-5354	53	37	machine	machine	NOUN
ejpam-5354	53	38	learning	learning	NOUN
ejpam-5354	53	39	tasks	task	NOUN
ejpam-5354	53	40	[	[	X
ejpam-5354	53	41	18	18	NUM
ejpam-5354	53	42	,	,	PUNCT
ejpam-5354	53	43	21	21	NUM
ejpam-5354	53	44	,	,	PUNCT
ejpam-5354	53	45	23	23	NUM
ejpam-5354	53	46	]	]	PUNCT
ejpam-5354	53	47	.	.	PUNCT
ejpam-5354	54	1	motivated	motivate	VERB
ejpam-5354	54	2	by	by	ADP
ejpam-5354	54	3	the	the	DET
ejpam-5354	54	4	previously	previously	ADV
ejpam-5354	54	5	mentioned	mention	VERB
ejpam-5354	54	6	trend	trend	NOUN
ejpam-5354	54	7	,	,	PUNCT
ejpam-5354	54	8	this	this	DET
ejpam-5354	54	9	paper	paper	NOUN
ejpam-5354	54	10	presents	present	VERB
ejpam-5354	54	11	an	an	DET
ejpam-5354	54	12	improved	improved	ADJ
ejpam-5354	54	13	conjugate	conjugate	ADJ
ejpam-5354	54	14	gradient	gradient	NOUN
ejpam-5354	54	15	method	method	NOUN
ejpam-5354	54	16	and	and	CCONJ
ejpam-5354	54	17	uses	use	VERB
ejpam-5354	54	18	feed	feed	NOUN
ejpam-5354	54	19	-	-	PUNCT
ejpam-5354	54	20	forward	forward	NOUN
ejpam-5354	54	21	neural	neural	ADJ
ejpam-5354	54	22	networks	network	NOUN
ejpam-5354	54	23	to	to	PART
ejpam-5354	54	24	solve	solve	VERB
ejpam-5354	54	25	different	different	ADJ
ejpam-5354	54	26	datasets	dataset	NOUN
ejpam-5354	54	27	.	.	PUNCT
ejpam-5354	55	1	the	the	DET
ejpam-5354	55	2	remaining	remain	VERB
ejpam-5354	55	3	sections	section	NOUN
ejpam-5354	55	4	of	of	ADP
ejpam-5354	55	5	this	this	DET
ejpam-5354	55	6	paper	paper	NOUN
ejpam-5354	55	7	are	be	AUX
ejpam-5354	55	8	organized	organize	VERB
ejpam-5354	55	9	as	as	SCONJ
ejpam-5354	55	10	follows	follow	VERB
ejpam-5354	55	11	:	:	PUNCT
ejpam-5354	55	12	the	the	DET
ejpam-5354	55	13	new	new	ADJ
ejpam-5354	55	14	conjugate	conjugate	ADJ
ejpam-5354	55	15	parameter	parameter	NOUN
ejpam-5354	55	16	’s	’s	PART
ejpam-5354	55	17	derivation	derivation	NOUN
ejpam-5354	55	18	process	process	NOUN
ejpam-5354	55	19	and	and	CCONJ
ejpam-5354	55	20	a	a	DET
ejpam-5354	55	21	thorough	thorough	ADJ
ejpam-5354	55	22	explanation	explanation	NOUN
ejpam-5354	55	23	of	of	ADP
ejpam-5354	55	24	its	its	PRON
ejpam-5354	55	25	algorithm	algorithm	NOUN
ejpam-5354	55	26	are	be	AUX
ejpam-5354	55	27	presented	present	VERB
ejpam-5354	55	28	in	in	ADP
ejpam-5354	55	29	section	section	NOUN
ejpam-5354	55	30	2	2	NUM
ejpam-5354	55	31	.	.	PUNCT
ejpam-5354	55	32	section	section	NOUN
ejpam-5354	55	33	3	3	NUM
ejpam-5354	55	34	discusses	discuss	VERB
ejpam-5354	55	35	the	the	DET
ejpam-5354	55	36	global	global	ADJ
ejpam-5354	55	37	convergence	convergence	NOUN
ejpam-5354	55	38	properties	property	NOUN
ejpam-5354	55	39	of	of	ADP
ejpam-5354	55	40	the	the	DET
ejpam-5354	55	41	method	method	NOUN
ejpam-5354	55	42	,	,	PUNCT
ejpam-5354	55	43	along	along	ADP
ejpam-5354	55	44	with	with	ADP
ejpam-5354	55	45	the	the	DET
ejpam-5354	55	46	sufficient	sufficient	ADJ
ejpam-5354	55	47	descent	descent	NOUN
ejpam-5354	55	48	condition	condition	NOUN
ejpam-5354	55	49	.	.	PUNCT
ejpam-5354	56	1	the	the	DET
ejpam-5354	56	2	numerical	numerical	ADJ
ejpam-5354	56	3	results	result	NOUN
ejpam-5354	56	4	of	of	ADP
ejpam-5354	56	5	applying	apply	VERB
ejpam-5354	56	6	the	the	DET
ejpam-5354	56	7	new	new	ADJ
ejpam-5354	56	8	method	method	NOUN
ejpam-5354	56	9	to	to	ADP
ejpam-5354	56	10	a	a	DET
ejpam-5354	56	11	number	number	NOUN
ejpam-5354	56	12	of	of	ADP
ejpam-5354	56	13	benchmark	benchmark	NOUN
ejpam-5354	56	14	test	test	NOUN
ejpam-5354	56	15	problems	problem	NOUN
ejpam-5354	56	16	are	be	AUX
ejpam-5354	56	17	presented	present	VERB
ejpam-5354	56	18	in	in	ADP
ejpam-5354	56	19	section	section	NOUN
ejpam-5354	56	20	4	4	NUM
ejpam-5354	56	21	.	.	PUNCT
ejpam-5354	57	1	lastly	lastly	ADV
ejpam-5354	57	2	,	,	PUNCT
ejpam-5354	57	3	the	the	DET
ejpam-5354	57	4	approach	approach	NOUN
ejpam-5354	57	5	is	be	AUX
ejpam-5354	57	6	expanded	expand	VERB
ejpam-5354	57	7	to	to	PART
ejpam-5354	57	8	handle	handle	VERB
ejpam-5354	57	9	issues	issue	NOUN
ejpam-5354	57	10	with	with	ADP
ejpam-5354	57	11	feed	feed	NOUN
ejpam-5354	57	12	-	-	PUNCT
ejpam-5354	57	13	forward	forward	NOUN
ejpam-5354	57	14	neural	neural	ADJ
ejpam-5354	57	15	networks	network	NOUN
ejpam-5354	57	16	2	2	NUM
ejpam-5354	57	17	.	.	PUNCT
ejpam-5354	58	1	derivation	derivation	NOUN
ejpam-5354	58	2	of	of	ADP
ejpam-5354	58	3	new	new	ADJ
ejpam-5354	58	4	method	method	NOUN
ejpam-5354	58	5	and	and	CCONJ
ejpam-5354	58	6	algorithm	algorithm	NOUN
ejpam-5354	58	7	in	in	ADP
ejpam-5354	58	8	this	this	DET
ejpam-5354	58	9	section	section	NOUN
ejpam-5354	58	10	,	,	PUNCT
ejpam-5354	58	11	a	a	DET
ejpam-5354	58	12	new	new	ADJ
ejpam-5354	58	13	conjugate	conjugate	ADJ
ejpam-5354	58	14	gradient	gradient	NOUN
ejpam-5354	58	15	method	method	NOUN
ejpam-5354	58	16	for	for	ADP
ejpam-5354	58	17	unconstrained	unconstrained	ADJ
ejpam-5354	58	18	optimizations	optimization	NOUN
ejpam-5354	58	19	is	be	AUX
ejpam-5354	58	20	present	present	ADJ
ejpam-5354	58	21	based	base	VERB
ejpam-5354	58	22	on	on	ADP
ejpam-5354	58	23	suggested	suggest	VERB
ejpam-5354	58	24	vector	vector	NOUN
ejpam-5354	58	25	y∗k	y∗k	PROPN
ejpam-5354	58	26	and	and	CCONJ
ejpam-5354	58	27	perry	perry	PROPN
ejpam-5354	58	28	conjugate	conjugate	PROPN
ejpam-5354	58	29	gradient	gradient	ADJ
ejpam-5354	58	30	parameter	parameter	NOUN
ejpam-5354	58	31	[	[	X
ejpam-5354	58	32	28	28	NUM
ejpam-5354	58	33	]	]	PUNCT
ejpam-5354	58	34	,	,	PUNCT
ejpam-5354	58	35	βk	βk	ADP
ejpam-5354	58	36	perry	perry	NOUN
ejpam-5354	58	37	=	=	PUNCT
ejpam-5354	58	38	gk+1	gk+1	PROPN
ejpam-5354	58	39	t	t	NOUN
ejpam-5354	58	40	(	(	PUNCT
ejpam-5354	58	41	yk	yk	PROPN
ejpam-5354	58	42	−	−	PROPN
ejpam-5354	58	43	νk	νk	NOUN
ejpam-5354	58	44	)	)	PUNCT
ejpam-5354	58	45	dk	dk	PROPN
ejpam-5354	58	46	t	t	PROPN
ejpam-5354	58	47	yk	yk	PROPN
ejpam-5354	58	48	.	.	PUNCT
ejpam-5354	59	1	(	(	PUNCT
ejpam-5354	59	2	4	4	X
ejpam-5354	59	3	)	)	PUNCT
ejpam-5354	59	4	consider	consider	VERB
ejpam-5354	59	5	the	the	DET
ejpam-5354	59	6	vector	vector	NOUN
ejpam-5354	59	7	y∗k	y∗k	PROPN
ejpam-5354	59	8	is	be	AUX
ejpam-5354	59	9	defined	define	VERB
ejpam-5354	59	10	by	by	ADP
ejpam-5354	59	11	y∗k	y∗k	PROPN
ejpam-5354	59	12	=	=	SYM
ejpam-5354	59	13	(	(	PUNCT
ejpam-5354	59	14	1	1	NUM
ejpam-5354	59	15	+	+	NUM
ejpam-5354	59	16	σ	σ	NOUN
ejpam-5354	60	1	+	+	X
ejpam-5354	60	2	λ)yk	λ)yk	PROPN
ejpam-5354	60	3	,	,	PUNCT
ejpam-5354	60	4	where	where	SCONJ
ejpam-5354	60	5	σ	σ	PROPN
ejpam-5354	60	6	and	and	CCONJ
ejpam-5354	60	7	λ	λ	NOUN
ejpam-5354	60	8	are	be	AUX
ejpam-5354	60	9	given	give	VERB
ejpam-5354	60	10	by	by	ADP
ejpam-5354	60	11	:	:	PUNCT
ejpam-5354	60	12	σ	σ	NOUN
ejpam-5354	60	13	=	=	SYM
ejpam-5354	60	14	1	1	NUM
ejpam-5354	60	15	1	1	NUM
ejpam-5354	60	16	+	+	CCONJ
ejpam-5354	60	17	e−∥yk∥	e−∥yk∥	PRON
ejpam-5354	60	18	,	,	PUNCT
ejpam-5354	60	19	λ	λ	X
ejpam-5354	60	20	=	=	VERB
ejpam-5354	60	21	µ	µ	PROPN
ejpam-5354	60	22	√	√	PROPN
ejpam-5354	60	23	m	m	VERB
ejpam-5354	60	24	∥gk∥	∥gk∥	NUM
ejpam-5354	60	25	(	(	PUNCT
ejpam-5354	60	26	1	1	NUM
ejpam-5354	60	27	+	+	NUM
ejpam-5354	60	28	∥gk+1∥	∥gk+1∥	NUM
ejpam-5354	60	29	)	)	PUNCT
ejpam-5354	61	1	[	[	X
ejpam-5354	61	2	3	3	X
ejpam-5354	61	3	]	]	PUNCT
ejpam-5354	61	4	with	with	ADP
ejpam-5354	61	5	m	m	PART
ejpam-5354	61	6	representing	represent	VERB
ejpam-5354	61	7	machine	machine	NOUN
ejpam-5354	61	8	accuracy	accuracy	NOUN
ejpam-5354	61	9	,	,	PUNCT
ejpam-5354	61	10	taking	take	VERB
ejpam-5354	61	11	(	(	PUNCT
ejpam-5354	61	12	m	m	PROPN
ejpam-5354	61	13	=	=	SYM
ejpam-5354	61	14	1∗	1∗	NUM
ejpam-5354	61	15	10−16	10−16	PROPN
ejpam-5354	61	16	)	)	PUNCT
ejpam-5354	61	17	,	,	PUNCT
ejpam-5354	61	18	and	and	CCONJ
ejpam-5354	61	19	µ	µ	X
ejpam-5354	61	20	=	=	SYM
ejpam-5354	61	21	0.1	0.1	NUM
ejpam-5354	61	22	.	.	PUNCT
ejpam-5354	61	23	substituting	substitute	VERB
ejpam-5354	61	24	these	these	PRON
ejpam-5354	61	25	into	into	ADP
ejpam-5354	61	26	the	the	DET
ejpam-5354	61	27	expression	expression	NOUN
ejpam-5354	61	28	for	for	ADP
ejpam-5354	61	29	y∗k	y∗k	NUM
ejpam-5354	61	30	gives	give	NOUN
ejpam-5354	61	31	:	:	PUNCT
ejpam-5354	61	32	y∗k	y∗k	PROPN
ejpam-5354	61	33	=	=	SYM
ejpam-5354	61	34	yk	yk	PROPN
ejpam-5354	61	35	+	+	CCONJ
ejpam-5354	61	36	1	1	NUM
ejpam-5354	61	37	1	1	NUM
ejpam-5354	61	38	+	+	CCONJ
ejpam-5354	61	39	e−∥yk∥	e−∥yk∥	PROPN
ejpam-5354	61	40	yk	yk	PROPN
ejpam-5354	61	41	+	+	PROPN
ejpam-5354	61	42	µ	µ	PROPN
ejpam-5354	61	43	√	√	PROPN
ejpam-5354	61	44	m	m	VERB
ejpam-5354	61	45	∥gk∥	∥gk∥	NUM
ejpam-5354	61	46	(	(	PUNCT
ejpam-5354	61	47	1	1	NUM
ejpam-5354	61	48	+	+	NUM
ejpam-5354	61	49	∥gk+1∥	∥gk+1∥	NUM
ejpam-5354	61	50	)	)	PUNCT
ejpam-5354	61	51	yk′	yk′	ADP
ejpam-5354	61	52	(	(	PUNCT
ejpam-5354	61	53	5	5	NUM
ejpam-5354	61	54	)	)	PUNCT
ejpam-5354	61	55	now	now	ADV
ejpam-5354	61	56	,	,	PUNCT
ejpam-5354	61	57	by	by	ADP
ejpam-5354	61	58	replacing	replace	VERB
ejpam-5354	61	59	yk	yk	PROPN
ejpam-5354	61	60	by	by	ADP
ejpam-5354	61	61	y∗k	y∗k	PROPN
ejpam-5354	61	62	in	in	ADP
ejpam-5354	61	63	equation	equation	NOUN
ejpam-5354	61	64	(	(	PUNCT
ejpam-5354	61	65	4	4	NUM
ejpam-5354	61	66	)	)	PUNCT
ejpam-5354	61	67	,	,	PUNCT
ejpam-5354	61	68	we	we	PRON
ejpam-5354	61	69	obtained	obtain	VERB
ejpam-5354	61	70	a	a	DET
ejpam-5354	61	71	new	new	ADJ
ejpam-5354	61	72	parameter	parameter	NOUN
ejpam-5354	61	73	βnew	βnew	NOUN
ejpam-5354	62	1	k	k	PROPN
ejpam-5354	62	2	=	=	PUNCT
ejpam-5354	62	3	gk+1	gk+1	PROPN
ejpam-5354	62	4	t	t	NOUN
ejpam-5354	62	5	(	(	PUNCT
ejpam-5354	62	6	y∗k	y∗k	NUM
ejpam-5354	62	7	−	−	PROPN
ejpam-5354	62	8	vk	vk	PROPN
ejpam-5354	62	9	)	)	PUNCT
ejpam-5354	62	10	dkt	dkt	PROPN
ejpam-5354	62	11	y	y	PROPN
ejpam-5354	62	12	∗	∗	NOUN
ejpam-5354	62	13	k	k	PROPN
ejpam-5354	63	1	=	=	PUNCT
ejpam-5354	63	2	gk+1	gk+1	NOUN
ejpam-5354	63	3	t	t	NOUN
ejpam-5354	63	4	(	(	PUNCT
ejpam-5354	63	5	(	(	PUNCT
ejpam-5354	63	6	1	1	NUM
ejpam-5354	63	7	+	+	NUM
ejpam-5354	63	8	σ	σ	NOUN
ejpam-5354	63	9	+	+	CCONJ
ejpam-5354	63	10	λ)yk	λ)yk	PROPN
ejpam-5354	63	11	−	−	PROPN
ejpam-5354	63	12	vk	vk	NOUN
ejpam-5354	63	13	)	)	PUNCT
ejpam-5354	63	14	(	(	PUNCT
ejpam-5354	63	15	1	1	NUM
ejpam-5354	63	16	+	+	NUM
ejpam-5354	63	17	σ	σ	NOUN
ejpam-5354	63	18	+	+	CCONJ
ejpam-5354	63	19	λ)dkt	λ)dkt	PROPN
ejpam-5354	63	20	yk	yk	PROPN
ejpam-5354	63	21	.	.	PUNCT
ejpam-5354	64	1	alaa	alaa	PROPN
ejpam-5354	64	2	l.	l.	PROPN
ejpam-5354	64	3	ibrahim	ibrahim	PROPN
ejpam-5354	64	4	et	et	PROPN
ejpam-5354	64	5	al	al	PROPN
ejpam-5354	64	6	.	.	PUNCT
ejpam-5354	64	7	/	/	SYM
ejpam-5354	64	8	eur	eur	PROPN
ejpam-5354	64	9	.	.	PUNCT
ejpam-5354	65	1	j.	j.	PROPN
ejpam-5354	65	2	pure	pure	PROPN
ejpam-5354	65	3	appl	appl	PROPN
ejpam-5354	65	4	.	.	PROPN
ejpam-5354	65	5	math	math	PROPN
ejpam-5354	65	6	,	,	PUNCT
ejpam-5354	65	7	17	17	NUM
ejpam-5354	65	8	(	(	PUNCT
ejpam-5354	65	9	4	4	NUM
ejpam-5354	65	10	)	)	PUNCT
ejpam-5354	65	11	(	(	PUNCT
ejpam-5354	65	12	2024	2024	NUM
ejpam-5354	65	13	)	)	PUNCT
ejpam-5354	65	14	,	,	PUNCT
ejpam-5354	65	15	2692	2692	NUM
ejpam-5354	65	16	-	-	SYM
ejpam-5354	65	17	2705	2705	NUM
ejpam-5354	65	18	2695	2695	NUM
ejpam-5354	65	19	or	or	CCONJ
ejpam-5354	65	20	βnew	βnew	ADJ
ejpam-5354	65	21	k	k	NOUN
ejpam-5354	65	22	=	=	PUNCT
ejpam-5354	65	23	gk+1	gk+1	PROPN
ejpam-5354	65	24	t	t	PROPN
ejpam-5354	65	25	yk	yk	PROPN
ejpam-5354	65	26	dtk	dtk	PROPN
ejpam-5354	65	27	yk	yk	PROPN
ejpam-5354	66	1	−	−	PROPN
ejpam-5354	66	2	gk+1	gk+1	NOUN
ejpam-5354	66	3	t	t	PROPN
ejpam-5354	66	4	vk	vk	X
ejpam-5354	66	5	(	(	PUNCT
ejpam-5354	66	6	1	1	NUM
ejpam-5354	66	7	+	+	NUM
ejpam-5354	66	8	σ	σ	NOUN
ejpam-5354	66	9	+	+	CCONJ
ejpam-5354	66	10	λ)dkt	λ)dkt	PROPN
ejpam-5354	66	11	yk	yk	PROPN
ejpam-5354	66	12	.	.	PUNCT
ejpam-5354	67	1	(	(	PUNCT
ejpam-5354	67	2	6	6	X
ejpam-5354	67	3	)	)	PUNCT
ejpam-5354	67	4	algorithm	algorithm	NOUN
ejpam-5354	67	5	2.1	2.1	NUM
ejpam-5354	67	6	:	:	PUNCT
ejpam-5354	67	7	new	new	ADJ
ejpam-5354	67	8	method	method	NOUN
ejpam-5354	67	9	require	require	NOUN
ejpam-5354	67	10	:	:	PUNCT
ejpam-5354	67	11	initial	initial	ADJ
ejpam-5354	67	12	point	point	NOUN
ejpam-5354	67	13	x0	x0	PROPN
ejpam-5354	67	14	∈	∈	PROPN
ejpam-5354	67	15	rn	rn	PROPN
ejpam-5354	67	16	.	.	PROPN
ejpam-5354	67	17	step	step	NOUN
ejpam-5354	67	18	1	1	NUM
ejpam-5354	67	19	:	:	PUNCT
ejpam-5354	67	20	set	set	VERB
ejpam-5354	67	21	k	k	PROPN
ejpam-5354	67	22	=	=	SYM
ejpam-5354	67	23	0	0	NUM
ejpam-5354	67	24	,	,	PUNCT
ejpam-5354	67	25	g0	g0	NOUN
ejpam-5354	67	26	=	=	SYM
ejpam-5354	67	27	∇f	∇f	PROPN
ejpam-5354	67	28	(	(	PUNCT
ejpam-5354	67	29	x0	x0	PROPN
ejpam-5354	67	30	)	)	PUNCT
ejpam-5354	67	31	,	,	PUNCT
ejpam-5354	67	32	d0	d0	NOUN
ejpam-5354	67	33	=	=	SYM
ejpam-5354	67	34	−g0	−g0	PROPN
ejpam-5354	67	35	.	.	PUNCT
ejpam-5354	68	1	if	if	SCONJ
ejpam-5354	68	2	g0	g0	PROPN
ejpam-5354	68	3	=	=	SYM
ejpam-5354	68	4	0	0	PROPN
ejpam-5354	68	5	,	,	PUNCT
ejpam-5354	68	6	then	then	ADV
ejpam-5354	68	7	stop	stop	VERB
ejpam-5354	68	8	.	.	PUNCT
ejpam-5354	69	1	step	step	NOUN
ejpam-5354	69	2	2	2	NUM
ejpam-5354	69	3	:	:	PUNCT
ejpam-5354	69	4	compute	compute	VERB
ejpam-5354	69	5	αk	αk	NOUN
ejpam-5354	69	6	to	to	PART
ejpam-5354	69	7	minimize	minimize	VERB
ejpam-5354	69	8	f(xk+1	f(xk+1	NOUN
ejpam-5354	69	9	)	)	PUNCT
ejpam-5354	69	10	(	(	PUNCT
ejpam-5354	69	11	i.e.	i.e.	X
ejpam-5354	69	12	,	,	PUNCT
ejpam-5354	69	13	fk+1	fk+1	ADJ
ejpam-5354	69	14	≤	≤	NUM
ejpam-5354	69	15	fk	fk	NOUN
ejpam-5354	69	16	)	)	PUNCT
ejpam-5354	69	17	using	use	VERB
ejpam-5354	69	18	cubic	cubic	ADJ
ejpam-5354	69	19	line	line	NOUN
ejpam-5354	69	20	search	search	NOUN
ejpam-5354	69	21	.	.	PUNCT
ejpam-5354	70	1	step	step	NOUN
ejpam-5354	70	2	3	3	NUM
ejpam-5354	70	3	:	:	PUNCT
ejpam-5354	70	4	update	update	NOUN
ejpam-5354	70	5	xk+1	xk+1	PROPN
ejpam-5354	70	6	=	=	SYM
ejpam-5354	70	7	xk	xk	PROPN
ejpam-5354	71	1	+	+	CCONJ
ejpam-5354	71	2	αkdk	αkdk	NOUN
ejpam-5354	71	3	.	.	PUNCT
ejpam-5354	72	1	step	step	NOUN
ejpam-5354	72	2	4	4	NUM
ejpam-5354	72	3	:	:	PUNCT
ejpam-5354	72	4	compute	compute	NOUN
ejpam-5354	72	5	gk+1	gk+1	NOUN
ejpam-5354	72	6	=	=	SYM
ejpam-5354	72	7	∇f	∇f	PROPN
ejpam-5354	72	8	(	(	PUNCT
ejpam-5354	72	9	xk+1	xk+1	ADJ
ejpam-5354	72	10	)	)	PUNCT
ejpam-5354	72	11	.	.	PUNCT
ejpam-5354	73	1	if	if	SCONJ
ejpam-5354	73	2	||gk+1||	||gk+1||	PROPN
ejpam-5354	73	3	≤	≤	PROPN
ejpam-5354	73	4	10−5	10−5	NUM
ejpam-5354	73	5	,	,	PUNCT
ejpam-5354	73	6	then	then	ADV
ejpam-5354	73	7	stop	stop	VERB
ejpam-5354	73	8	.	.	PUNCT
ejpam-5354	74	1	step	step	NOUN
ejpam-5354	74	2	5	5	NUM
ejpam-5354	74	3	:	:	PUNCT
ejpam-5354	74	4	compute	compute	VERB
ejpam-5354	74	5	βk	βk	ADP
ejpam-5354	74	6	from	from	ADP
ejpam-5354	74	7	equation	equation	NOUN
ejpam-5354	74	8	(	(	PUNCT
ejpam-5354	74	9	6	6	NUM
ejpam-5354	74	10	)	)	PUNCT
ejpam-5354	74	11	.	.	PUNCT
ejpam-5354	75	1	step	step	NOUN
ejpam-5354	75	2	6	6	NUM
ejpam-5354	75	3	:	:	PUNCT
ejpam-5354	75	4	set	set	VERB
ejpam-5354	75	5	dk+1	dk+1	NOUN
ejpam-5354	75	6	=	=	SYM
ejpam-5354	75	7	−gk+1	−gk+1	PROPN
ejpam-5354	75	8	+	+	CCONJ
ejpam-5354	75	9	βnew	βnew	ADJ
ejpam-5354	75	10	k	k	PROPN
ejpam-5354	75	11	dk	dk	PROPN
ejpam-5354	75	12	.	.	PROPN
ejpam-5354	75	13	step	step	NOUN
ejpam-5354	75	14	7	7	NUM
ejpam-5354	75	15	:	:	PUNCT
ejpam-5354	75	16	if	if	SCONJ
ejpam-5354	75	17	k	k	PROPN
ejpam-5354	75	18	=	=	PUNCT
ejpam-5354	75	19	n	n	PROPN
ejpam-5354	75	20	or	or	CCONJ
ejpam-5354	75	21	if	if	SCONJ
ejpam-5354	75	22	∣∣gtk	∣∣gtk	NUM
ejpam-5354	75	23	gk+1	gk+1	X
ejpam-5354	75	24	∣∣	∣∣	NUM
ejpam-5354	75	25	≥	≥	NOUN
ejpam-5354	75	26	0.2||gk+1||2	0.2||gk+1||2	NOUN
ejpam-5354	75	27	is	be	AUX
ejpam-5354	75	28	satisfied	satisfied	ADJ
ejpam-5354	75	29	,	,	PUNCT
ejpam-5354	75	30	go	go	VERB
ejpam-5354	75	31	to	to	PART
ejpam-5354	75	32	step	step	VERB
ejpam-5354	75	33	3	3	NUM
ejpam-5354	75	34	.	.	PUNCT
ejpam-5354	76	1	otherwise	otherwise	ADV
ejpam-5354	76	2	,	,	PUNCT
ejpam-5354	76	3	set	set	VERB
ejpam-5354	76	4	k	k	PROPN
ejpam-5354	77	1	=	=	PUNCT
ejpam-5354	77	2	k	k	PROPN
ejpam-5354	78	1	+	+	CCONJ
ejpam-5354	78	2	1	1	NUM
ejpam-5354	78	3	and	and	CCONJ
ejpam-5354	78	4	go	go	VERB
ejpam-5354	78	5	to	to	PART
ejpam-5354	78	6	step	step	VERB
ejpam-5354	78	7	4	4	NUM
ejpam-5354	78	8	.	.	PUNCT
ejpam-5354	79	1	theorem	theorem	NOUN
ejpam-5354	79	2	1	1	NUM
ejpam-5354	79	3	:	:	PUNCT
ejpam-5354	79	4	let	let	VERB
ejpam-5354	79	5	the	the	DET
ejpam-5354	79	6	sequences	sequence	NOUN
ejpam-5354	79	7	{	{	PUNCT
ejpam-5354	79	8	xk	xk	ADJ
ejpam-5354	79	9	}	}	PUNCT
ejpam-5354	79	10	and	and	CCONJ
ejpam-5354	79	11	{	{	PUNCT
ejpam-5354	79	12	dk	dk	NOUN
ejpam-5354	79	13	}	}	PUNCT
ejpam-5354	79	14	be	be	AUX
ejpam-5354	79	15	generated	generate	VERB
ejpam-5354	79	16	by	by	ADP
ejpam-5354	79	17	equations	equation	NOUN
ejpam-5354	79	18	(	(	PUNCT
ejpam-5354	79	19	2	2	NUM
ejpam-5354	79	20	)	)	PUNCT
ejpam-5354	79	21	,	,	PUNCT
ejpam-5354	79	22	(	(	PUNCT
ejpam-5354	79	23	3	3	NUM
ejpam-5354	79	24	)	)	PUNCT
ejpam-5354	79	25	,	,	PUNCT
ejpam-5354	79	26	and	and	CCONJ
ejpam-5354	79	27	(	(	PUNCT
ejpam-5354	79	28	6	6	NUM
ejpam-5354	79	29	)	)	PUNCT
ejpam-5354	79	30	.	.	PUNCT
ejpam-5354	80	1	then	then	ADV
ejpam-5354	80	2	the	the	DET
ejpam-5354	80	3	search	search	NOUN
ejpam-5354	80	4	direction	direction	NOUN
ejpam-5354	80	5	dk+1	dk+1	VERB
ejpam-5354	80	6	satisfies	satisfie	NOUN
ejpam-5354	80	7	the	the	DET
ejpam-5354	80	8	sufficient	sufficient	ADJ
ejpam-5354	80	9	descent	descent	NOUN
ejpam-5354	80	10	condition	condition	NOUN
ejpam-5354	80	11	:	:	PUNCT
ejpam-5354	80	12	gtk+1dk+1	gtk+1dk+1	NUM
ejpam-5354	80	13	≤	≤	NUM
ejpam-5354	80	14	−c∥gk+1∥2	−c∥gk+1∥2	NOUN
ejpam-5354	80	15	proof	proof	NOUN
ejpam-5354	80	16	:	:	PUNCT
ejpam-5354	80	17	from	from	ADP
ejpam-5354	80	18	equations	equation	NOUN
ejpam-5354	80	19	(	(	PUNCT
ejpam-5354	80	20	3	3	NUM
ejpam-5354	80	21	)	)	PUNCT
ejpam-5354	80	22	and	and	CCONJ
ejpam-5354	80	23	(	(	PUNCT
ejpam-5354	80	24	6	6	NUM
ejpam-5354	80	25	)	)	PUNCT
ejpam-5354	80	26	,	,	PUNCT
ejpam-5354	80	27	we	we	PRON
ejpam-5354	80	28	have	have	VERB
ejpam-5354	80	29	dk+1	dk+1	NOUN
ejpam-5354	80	30	=	=	SYM
ejpam-5354	80	31	−gk+1	−gk+1	PROPN
ejpam-5354	80	32	+	+	CCONJ
ejpam-5354	80	33	(	(	PUNCT
ejpam-5354	80	34	gk+1	gk+1	NOUN
ejpam-5354	80	35	t	t	NOUN
ejpam-5354	80	36	yk	yk	PROPN
ejpam-5354	81	1	dk	dk	PROPN
ejpam-5354	81	2	t	t	PROPN
ejpam-5354	81	3	yk	yk	PROPN
ejpam-5354	82	1	−	−	PROPN
ejpam-5354	82	2	gk+1	gk+1	NOUN
ejpam-5354	82	3	t	t	X
ejpam-5354	82	4	νk	νk	X
ejpam-5354	82	5	(	(	PUNCT
ejpam-5354	82	6	1	1	NUM
ejpam-5354	82	7	+	+	NUM
ejpam-5354	82	8	σ	σ	PROPN
ejpam-5354	82	9	+	+	CCONJ
ejpam-5354	82	10	λ	λ	NOUN
ejpam-5354	82	11	)	)	PUNCT
ejpam-5354	82	12	dk	dk	PROPN
ejpam-5354	82	13	t	t	PROPN
ejpam-5354	82	14	yk	yk	PROPN
ejpam-5354	82	15	)	)	PUNCT
ejpam-5354	83	1	dk	dk	PROPN
ejpam-5354	83	2	.	.	PUNCT
ejpam-5354	84	1	(	(	PUNCT
ejpam-5354	84	2	7	7	X
ejpam-5354	84	3	)	)	PUNCT
ejpam-5354	84	4	now	now	ADV
ejpam-5354	84	5	,	,	PUNCT
ejpam-5354	84	6	by	by	ADP
ejpam-5354	84	7	multiplying	multiply	VERB
ejpam-5354	84	8	the	the	DET
ejpam-5354	84	9	above	above	ADJ
ejpam-5354	84	10	equation	equation	NOUN
ejpam-5354	84	11	by	by	ADP
ejpam-5354	84	12	gtk+1	gtk+1	NOUN
ejpam-5354	84	13	from	from	ADP
ejpam-5354	84	14	both	both	DET
ejpam-5354	84	15	sides	side	NOUN
ejpam-5354	84	16	,	,	PUNCT
ejpam-5354	84	17	we	we	PRON
ejpam-5354	84	18	get	get	VERB
ejpam-5354	84	19	gtk+1dk+1	gtk+1dk+1	VERB
ejpam-5354	84	20	=	=	SYM
ejpam-5354	84	21	−∥gk+1∥2	−∥gk+1∥2	PROPN
ejpam-5354	84	22	+	+	CCONJ
ejpam-5354	84	23	βhs	βhs	NOUN
ejpam-5354	84	24	k	k	PROPN
ejpam-5354	84	25	gtk+1dk	gtk+1dk	PROPN
ejpam-5354	85	1	−	−	PROPN
ejpam-5354	85	2	αk	αk	NOUN
ejpam-5354	85	3	(	(	PUNCT
ejpam-5354	85	4	gtk+1dk	gtk+1dk	PROPN
ejpam-5354	85	5	)	)	PUNCT
ejpam-5354	85	6	2	2	NUM
ejpam-5354	85	7	(	(	PUNCT
ejpam-5354	85	8	1	1	NUM
ejpam-5354	85	9	+	+	NUM
ejpam-5354	85	10	σ	σ	NOUN
ejpam-5354	85	11	+	+	CCONJ
ejpam-5354	85	12	λ)dtk	λ)dtk	X
ejpam-5354	85	13	yk	yk	NOUN
ejpam-5354	85	14	.	.	PUNCT
ejpam-5354	86	1	(	(	PUNCT
ejpam-5354	86	2	8)	8)	NUM
ejpam-5354	86	3	if	if	SCONJ
ejpam-5354	86	4	dtk	dtk	PROPN
ejpam-5354	86	5	gk+1	gk+1	AUX
ejpam-5354	86	6	=	=	SYM
ejpam-5354	86	7	0	0	NUM
ejpam-5354	86	8	,	,	PUNCT
ejpam-5354	86	9	then	then	ADV
ejpam-5354	86	10	gtk+1dk+1	gtk+1dk+1	ADJ
ejpam-5354	86	11	=	=	SYM
ejpam-5354	86	12	−∥gk+1∥2	−∥gk+1∥2	ADJ
ejpam-5354	86	13	≤	≤	NOUN
ejpam-5354	86	14	0	0	NUM
ejpam-5354	86	15	.	.	PUNCT
ejpam-5354	87	1	then	then	ADV
ejpam-5354	87	2	the	the	DET
ejpam-5354	87	3	proof	proof	NOUN
ejpam-5354	87	4	is	be	AUX
ejpam-5354	87	5	complete	complete	ADJ
ejpam-5354	87	6	.	.	PUNCT
ejpam-5354	88	1	now	now	ADV
ejpam-5354	88	2	,	,	PUNCT
ejpam-5354	88	3	if	if	SCONJ
ejpam-5354	88	4	dtk	dtk	PROPN
ejpam-5354	88	5	gk+1	gk+1	VERB
ejpam-5354	88	6	̸=	̸=	PROPN
ejpam-5354	88	7	0	0	NUM
ejpam-5354	88	8	,	,	PUNCT
ejpam-5354	88	9	we	we	PRON
ejpam-5354	88	10	know	know	VERB
ejpam-5354	88	11	that	that	SCONJ
ejpam-5354	88	12	the	the	DET
ejpam-5354	88	13	first	first	ADJ
ejpam-5354	88	14	two	two	NUM
ejpam-5354	88	15	terms	term	NOUN
ejpam-5354	88	16	of	of	ADP
ejpam-5354	88	17	the	the	DET
ejpam-5354	88	18	above	above	ADJ
ejpam-5354	88	19	equation	equation	NOUN
ejpam-5354	88	20	correspond	correspond	VERB
ejpam-5354	88	21	to	to	ADP
ejpam-5354	88	22	the	the	DET
ejpam-5354	88	23	search	search	NOUN
ejpam-5354	88	24	direction	direction	NOUN
ejpam-5354	88	25	of	of	ADP
ejpam-5354	88	26	the	the	DET
ejpam-5354	88	27	hs	hs	PROPN
ejpam-5354	88	28	method	method	NOUN
ejpam-5354	88	29	,	,	PUNCT
ejpam-5354	88	30	so	so	SCONJ
ejpam-5354	88	31	it	it	PRON
ejpam-5354	88	32	is	be	AUX
ejpam-5354	88	33	less	less	ADJ
ejpam-5354	88	34	than	than	ADP
ejpam-5354	88	35	or	or	CCONJ
ejpam-5354	88	36	equal	equal	ADJ
ejpam-5354	88	37	to	to	ADP
ejpam-5354	88	38	zero	zero	NUM
ejpam-5354	88	39	.	.	PUNCT
ejpam-5354	89	1	and	and	CCONJ
ejpam-5354	89	2	it	it	PRON
ejpam-5354	89	3	is	be	AUX
ejpam-5354	89	4	obvious	obvious	ADJ
ejpam-5354	89	5	that	that	SCONJ
ejpam-5354	89	6	αk	αk	NOUN
ejpam-5354	89	7	,	,	PUNCT
ejpam-5354	89	8	1	1	NUM
ejpam-5354	89	9	+	+	SYM
ejpam-5354	89	10	σ	σ	PROPN
ejpam-5354	89	11	+	+	CCONJ
ejpam-5354	89	12	λ	λ	PROPN
ejpam-5354	89	13	,	,	PUNCT
ejpam-5354	89	14	dk	dk	PROPN
ejpam-5354	89	15	t	t	PROPN
ejpam-5354	89	16	yk	yk	PROPN
ejpam-5354	89	17	,	,	PUNCT
ejpam-5354	89	18	and	and	CCONJ
ejpam-5354	89	19	(	(	PUNCT
ejpam-5354	89	20	gtk+1dk	gtk+1dk	PROPN
ejpam-5354	89	21	)	)	PUNCT
ejpam-5354	89	22	2	2	NUM
ejpam-5354	89	23	are	be	AUX
ejpam-5354	89	24	positive	positive	ADJ
ejpam-5354	89	25	.	.	PUNCT
ejpam-5354	90	1	therefore	therefore	ADV
ejpam-5354	90	2	,	,	PUNCT
ejpam-5354	90	3	we	we	PRON
ejpam-5354	90	4	have	have	AUX
ejpam-5354	90	5	gtk+1dk+1	gtk+1dk+1	VERB
ejpam-5354	90	6	=	=	SYM
ejpam-5354	90	7	−∥gk+1∥2	−∥gk+1∥2	PROPN
ejpam-5354	90	8	+	+	CCONJ
ejpam-5354	90	9	βhs	βhs	NOUN
ejpam-5354	90	10	k	k	PROPN
ejpam-5354	90	11	gtk+1dk	gtk+1dk	PROPN
ejpam-5354	91	1	−	−	PROPN
ejpam-5354	91	2	αk	αk	NOUN
ejpam-5354	91	3	(	(	PUNCT
ejpam-5354	91	4	gtk+1dk	gtk+1dk	PROPN
ejpam-5354	91	5	)	)	PUNCT
ejpam-5354	91	6	2	2	NUM
ejpam-5354	91	7	(	(	PUNCT
ejpam-5354	91	8	1	1	NUM
ejpam-5354	91	9	+	+	NUM
ejpam-5354	91	10	σ	σ	NOUN
ejpam-5354	92	1	+	+	CCONJ
ejpam-5354	92	2	λ)dtk	λ)dtk	PUNCT
ejpam-5354	92	3	yk	yk	NOUN
ejpam-5354	92	4	≤	≤	NOUN
ejpam-5354	92	5	0	0	PUNCT
ejpam-5354	92	6	,	,	PUNCT
ejpam-5354	92	7	so	so	ADV
ejpam-5354	92	8	,	,	PUNCT
ejpam-5354	92	9	the	the	DET
ejpam-5354	92	10	above	above	ADJ
ejpam-5354	92	11	equation	equation	NOUN
ejpam-5354	92	12	can	can	AUX
ejpam-5354	92	13	be	be	AUX
ejpam-5354	92	14	written	write	VERB
ejpam-5354	92	15	as	as	ADP
ejpam-5354	92	16	gtk+1dk+1	gtk+1dk+1	VERB
ejpam-5354	92	17	≤	≤	NUM
ejpam-5354	92	18	−	−	NOUN
ejpam-5354	93	1	(	(	PUNCT
ejpam-5354	93	2	αk	αk	INTJ
ejpam-5354	93	3	(	(	PUNCT
ejpam-5354	93	4	gtk+1dk	gtk+1dk	PROPN
ejpam-5354	93	5	)	)	PUNCT
ejpam-5354	93	6	2	2	NUM
ejpam-5354	93	7	(	(	PUNCT
ejpam-5354	93	8	1	1	NUM
ejpam-5354	93	9	+	+	NUM
ejpam-5354	93	10	σ	σ	NOUN
ejpam-5354	93	11	+	+	X
ejpam-5354	93	12	λ)dtk	λ)dtk	X
ejpam-5354	93	13	yk∥gk+1∥2	yk∥gk+1∥2	NOUN
ejpam-5354	93	14	)	)	PUNCT
ejpam-5354	93	15	∥gk+1∥2	∥gk+1∥2	PROPN
ejpam-5354	93	16	.	.	PUNCT
ejpam-5354	94	1	(	(	PUNCT
ejpam-5354	94	2	9	9	NUM
ejpam-5354	94	3	)	)	PUNCT
ejpam-5354	94	4	then	then	ADV
ejpam-5354	94	5	,	,	PUNCT
ejpam-5354	94	6	gtk+1dk+1	gtk+1dk+1	NUM
ejpam-5354	94	7	≤	≤	NUM
ejpam-5354	94	8	−c∥gk+1∥2	−c∥gk+1∥2	NOUN
ejpam-5354	94	9	,	,	PUNCT
ejpam-5354	94	10	alaa	alaa	PROPN
ejpam-5354	94	11	l.	l.	PROPN
ejpam-5354	94	12	ibrahim	ibrahim	PROPN
ejpam-5354	94	13	et	et	PROPN
ejpam-5354	94	14	al	al	PROPN
ejpam-5354	94	15	.	.	PUNCT
ejpam-5354	94	16	/	/	SYM
ejpam-5354	94	17	eur	eur	PROPN
ejpam-5354	94	18	.	.	PUNCT
ejpam-5354	95	1	j.	j.	PROPN
ejpam-5354	95	2	pure	pure	PROPN
ejpam-5354	95	3	appl	appl	PROPN
ejpam-5354	95	4	.	.	PROPN
ejpam-5354	95	5	math	math	PROPN
ejpam-5354	95	6	,	,	PUNCT
ejpam-5354	95	7	17	17	NUM
ejpam-5354	95	8	(	(	PUNCT
ejpam-5354	95	9	4	4	NUM
ejpam-5354	95	10	)	)	PUNCT
ejpam-5354	95	11	(	(	PUNCT
ejpam-5354	95	12	2024	2024	NUM
ejpam-5354	95	13	)	)	PUNCT
ejpam-5354	95	14	,	,	PUNCT
ejpam-5354	95	15	2692	2692	NUM
ejpam-5354	95	16	-	-	SYM
ejpam-5354	95	17	2705	2705	NUM
ejpam-5354	95	18	2696	2696	NUM
ejpam-5354	96	1	where	where	SCONJ
ejpam-5354	96	2	c	c	NOUN
ejpam-5354	96	3	=	=	SYM
ejpam-5354	96	4	αk	αk	NOUN
ejpam-5354	96	5	(	(	PUNCT
ejpam-5354	96	6	gtk+1dk	gtk+1dk	PROPN
ejpam-5354	96	7	)	)	PUNCT
ejpam-5354	96	8	2	2	NUM
ejpam-5354	96	9	(	(	PUNCT
ejpam-5354	96	10	1	1	NUM
ejpam-5354	96	11	+	+	NUM
ejpam-5354	96	12	σ	σ	NOUN
ejpam-5354	96	13	+	+	X
ejpam-5354	96	14	λ)dtk	λ)dtk	X
ejpam-5354	96	15	yk∥gk+1∥2	yk∥gk+1∥2	PROPN
ejpam-5354	96	16	.	.	PUNCT
ejpam-5354	97	1	this	this	PRON
ejpam-5354	97	2	completes	complete	VERB
ejpam-5354	97	3	the	the	DET
ejpam-5354	97	4	proof	proof	NOUN
ejpam-5354	97	5	.	.	PUNCT
ejpam-5354	98	1	to	to	PART
ejpam-5354	98	2	prove	prove	VERB
ejpam-5354	98	3	the	the	DET
ejpam-5354	98	4	global	global	ADJ
ejpam-5354	98	5	convergence	convergence	NOUN
ejpam-5354	98	6	of	of	ADP
ejpam-5354	98	7	the	the	DET
ejpam-5354	98	8	new	new	ADJ
ejpam-5354	98	9	method	method	NOUN
ejpam-5354	98	10	,	,	PUNCT
ejpam-5354	98	11	we	we	PRON
ejpam-5354	98	12	need	need	VERB
ejpam-5354	98	13	to	to	PART
ejpam-5354	98	14	make	make	VERB
ejpam-5354	98	15	the	the	DET
ejpam-5354	98	16	following	follow	VERB
ejpam-5354	98	17	fundamental	fundamental	ADJ
ejpam-5354	98	18	assumptions	assumption	NOUN
ejpam-5354	98	19	about	about	ADP
ejpam-5354	98	20	the	the	DET
ejpam-5354	98	21	objective	objective	ADJ
ejpam-5354	98	22	function	function	NOUN
ejpam-5354	98	23	in	in	ADP
ejpam-5354	98	24	this	this	DET
ejpam-5354	98	25	section	section	NOUN
ejpam-5354	98	26	.	.	PUNCT
ejpam-5354	99	1	assumption	assumption	NOUN
ejpam-5354	100	1	[	[	X
ejpam-5354	100	2	38	38	NUM
ejpam-5354	100	3	]	]	PUNCT
ejpam-5354	100	4	:	:	PUNCT
ejpam-5354	100	5	(	(	PUNCT
ejpam-5354	100	6	i	i	NOUN
ejpam-5354	100	7	)	)	PUNCT
ejpam-5354	100	8	the	the	DET
ejpam-5354	100	9	level	level	NOUN
ejpam-5354	100	10	set	set	VERB
ejpam-5354	100	11	δ	δ	X
ejpam-5354	100	12	=	=	PRON
ejpam-5354	100	13	{	{	PUNCT
ejpam-5354	100	14	x	x	PROPN
ejpam-5354	100	15	|	|	ADV
ejpam-5354	100	16	f(x	f(x	PROPN
ejpam-5354	100	17	)	)	PUNCT
ejpam-5354	100	18	≤	≤	NUM
ejpam-5354	100	19	f(x0	f(x0	NOUN
ejpam-5354	100	20	)	)	PUNCT
ejpam-5354	100	21	}	}	PUNCT
ejpam-5354	100	22	is	be	AUX
ejpam-5354	100	23	bounded	bound	VERB
ejpam-5354	100	24	.	.	PUNCT
ejpam-5354	101	1	(	(	PUNCT
ejpam-5354	101	2	ii	ii	NOUN
ejpam-5354	101	3	)	)	PUNCT
ejpam-5354	101	4	in	in	ADP
ejpam-5354	101	5	some	some	DET
ejpam-5354	101	6	neighborhood	neighborhood	NOUN
ejpam-5354	101	7	n	n	NOUN
ejpam-5354	101	8	of	of	ADP
ejpam-5354	101	9	δ	δ	PROPN
ejpam-5354	101	10	,	,	PUNCT
ejpam-5354	101	11	f	f	PROPN
ejpam-5354	101	12	is	be	AUX
ejpam-5354	101	13	continuously	continuously	ADV
ejpam-5354	101	14	differentiable	differentiable	ADJ
ejpam-5354	101	15	,	,	PUNCT
ejpam-5354	101	16	and	and	CCONJ
ejpam-5354	101	17	its	its	PRON
ejpam-5354	101	18	gradient	gradient	NOUN
ejpam-5354	101	19	is	be	AUX
ejpam-5354	101	20	lipschitz	lipschitz	NOUN
ejpam-5354	101	21	continuous	continuous	ADJ
ejpam-5354	101	22	with	with	ADP
ejpam-5354	101	23	a	a	DET
ejpam-5354	101	24	lipschitz	lipschitz	NOUN
ejpam-5354	101	25	constant	constant	ADJ
ejpam-5354	101	26	δ	δ	PROPN
ejpam-5354	101	27	>	>	X
ejpam-5354	101	28	0	0	NUM
ejpam-5354	101	29	,	,	PUNCT
ejpam-5354	101	30	i.e.	i.e.	X
ejpam-5354	101	31	,	,	PUNCT
ejpam-5354	101	32	∥g(x)−	∥g(x)−	X
ejpam-5354	101	33	g(y)∥	g(y)∥	VERB
ejpam-5354	101	34	≤	≤	PUNCT
ejpam-5354	101	35	δ∥x−	δ∥x−	ADV
ejpam-5354	101	36	y∥	y∥	NOUN
ejpam-5354	101	37	∀x	∀x	NUM
ejpam-5354	101	38	,	,	PUNCT
ejpam-5354	101	39	y	y	PROPN
ejpam-5354	101	40	∈	∈	PROPN
ejpam-5354	101	41	δ	δ	PROPN
ejpam-5354	101	42	.	.	PUNCT
ejpam-5354	102	1	(	(	PUNCT
ejpam-5354	102	2	iii	iii	NOUN
ejpam-5354	102	3	)	)	PUNCT
ejpam-5354	102	4	from	from	ADP
ejpam-5354	102	5	the	the	DET
ejpam-5354	102	6	above	above	ADJ
ejpam-5354	102	7	assumptions	assumption	NOUN
ejpam-5354	102	8	,	,	PUNCT
ejpam-5354	102	9	there	there	PRON
ejpam-5354	102	10	exists	exist	VERB
ejpam-5354	102	11	a	a	DET
ejpam-5354	102	12	positive	positive	ADJ
ejpam-5354	102	13	constant	constant	ADJ
ejpam-5354	102	14	b	b	NOUN
ejpam-5354	102	15	such	such	ADJ
ejpam-5354	102	16	that	that	PRON
ejpam-5354	102	17	:	:	PUNCT
ejpam-5354	103	1	∥g(x)∥	∥g(x)∥	PROPN
ejpam-5354	103	2	≤	≤	PROPN
ejpam-5354	103	3	b	b	PROPN
ejpam-5354	103	4	∀x	∀x	NUM
ejpam-5354	103	5	∈	∈	PROPN
ejpam-5354	103	6	δ	δ	PROPN
ejpam-5354	103	7	.	.	PUNCT
ejpam-5354	104	1	(	(	PUNCT
ejpam-5354	104	2	10	10	NUM
ejpam-5354	104	3	)	)	PUNCT
ejpam-5354	104	4	(	(	PUNCT
ejpam-5354	104	5	iv	iv	X
ejpam-5354	104	6	)	)	PUNCT
ejpam-5354	104	7	if	if	SCONJ
ejpam-5354	104	8	f	f	PROPN
ejpam-5354	104	9	is	be	AUX
ejpam-5354	104	10	a	a	DET
ejpam-5354	104	11	uniformly	uniformly	ADV
ejpam-5354	104	12	convex	convex	NOUN
ejpam-5354	104	13	function	function	NOUN
ejpam-5354	104	14	,	,	PUNCT
ejpam-5354	104	15	there	there	PRON
ejpam-5354	104	16	exists	exist	VERB
ejpam-5354	104	17	a	a	DET
ejpam-5354	104	18	constant	constant	ADJ
ejpam-5354	104	19	ϑ	ϑ	X
ejpam-5354	104	20	>	>	X
ejpam-5354	104	21	0	0	NUM
ejpam-5354	104	22	such	such	ADJ
ejpam-5354	104	23	that	that	PRON
ejpam-5354	104	24	:	:	PUNCT
ejpam-5354	104	25	(	(	PUNCT
ejpam-5354	104	26	g(x)−	g(x)−	PROPN
ejpam-5354	104	27	g(y))t	g(y))t	PROPN
ejpam-5354	104	28	(	(	PUNCT
ejpam-5354	104	29	x−	x−	PROPN
ejpam-5354	104	30	y	y	PROPN
ejpam-5354	104	31	)	)	PUNCT
ejpam-5354	104	32	≥	≥	NOUN
ejpam-5354	104	33	ϑ∥x−	ϑ∥x−	PROPN
ejpam-5354	104	34	y∥2	y∥2	NOUN
ejpam-5354	104	35	∀x	∀x	NUM
ejpam-5354	104	36	,	,	PUNCT
ejpam-5354	104	37	y	y	PROPN
ejpam-5354	104	38	∈	∈	PROPN
ejpam-5354	104	39	ω	ω	PROPN
ejpam-5354	104	40	,	,	PUNCT
ejpam-5354	104	41	(	(	PUNCT
ejpam-5354	104	42	11	11	NUM
ejpam-5354	104	43	)	)	PUNCT
ejpam-5354	104	44	can	can	AUX
ejpam-5354	104	45	be	be	AUX
ejpam-5354	104	46	rewrite	rewrite	VERB
ejpam-5354	104	47	(	(	PUNCT
ejpam-5354	104	48	11	11	NUM
ejpam-5354	104	49	)	)	PUNCT
ejpam-5354	104	50	,	,	PUNCT
ejpam-5354	104	51	in	in	ADP
ejpam-5354	104	52	the	the	DET
ejpam-5354	104	53	following	follow	VERB
ejpam-5354	104	54	manner	manner	NOUN
ejpam-5354	104	55	:	:	PUNCT
ejpam-5354	104	56	ytk	ytk	PROPN
ejpam-5354	104	57	vk	vk	PROPN
ejpam-5354	104	58	≥	≥	NOUN
ejpam-5354	104	59	ϑ||vk||2	ϑ||vk||2	NOUN
ejpam-5354	104	60	.	.	PUNCT
ejpam-5354	105	1	(	(	PUNCT
ejpam-5354	105	2	12	12	NUM
ejpam-5354	105	3	)	)	PUNCT
ejpam-5354	105	4	lemma	lemma	PROPN
ejpam-5354	105	5	1	1	NUM
ejpam-5354	105	6	:	:	PUNCT
ejpam-5354	106	1	[	[	X
ejpam-5354	106	2	37	37	NUM
ejpam-5354	106	3	]	]	PUNCT
ejpam-5354	106	4	let	let	VERB
ejpam-5354	106	5	the	the	DET
ejpam-5354	106	6	above	above	ADJ
ejpam-5354	106	7	assumptions	assumption	NOUN
ejpam-5354	106	8	hold	hold	VERB
ejpam-5354	106	9	.	.	PUNCT
ejpam-5354	107	1	consider	consider	VERB
ejpam-5354	107	2	the	the	DET
ejpam-5354	107	3	methods	method	NOUN
ejpam-5354	107	4	(	(	PUNCT
ejpam-5354	107	5	2	2	NUM
ejpam-5354	107	6	)	)	PUNCT
ejpam-5354	107	7	,	,	PUNCT
ejpam-5354	107	8	and	and	CCONJ
ejpam-5354	107	9	(	(	PUNCT
ejpam-5354	107	10	3	3	NUM
ejpam-5354	107	11	)	)	PUNCT
ejpam-5354	107	12	,	,	PUNCT
ejpam-5354	107	13	where	where	SCONJ
ejpam-5354	107	14	dk+1	dk+1	NOUN
ejpam-5354	107	15	is	be	AUX
ejpam-5354	107	16	a	a	DET
ejpam-5354	107	17	descent	descent	NOUN
ejpam-5354	107	18	direction	direction	NOUN
ejpam-5354	107	19	and	and	CCONJ
ejpam-5354	107	20	αk	αk	SCONJ
ejpam-5354	107	21	satisfies	satisfy	VERB
ejpam-5354	107	22	the	the	DET
ejpam-5354	107	23	standard	standard	ADJ
ejpam-5354	107	24	wolfe	wolfe	PROPN
ejpam-5354	107	25	line	line	PROPN
ejpam-5354	107	26	search	search	NOUN
ejpam-5354	107	27	.	.	PUNCT
ejpam-5354	108	1	if∑	if∑	NOUN
ejpam-5354	108	2	k≥1	k≥1	NOUN
ejpam-5354	108	3	1	1	NUM
ejpam-5354	108	4	∥dk∥2	∥dk∥2	NOUN
ejpam-5354	108	5	=	=	SYM
ejpam-5354	108	6	∞.	∞.	PROPN
ejpam-5354	108	7	then	then	ADV
ejpam-5354	108	8	,	,	PUNCT
ejpam-5354	108	9	lim	lim	PROPN
ejpam-5354	108	10	k→∞	k→∞	PROPN
ejpam-5354	108	11	inf	inf	PROPN
ejpam-5354	109	1	∥gk∥	∥gk∥	PROPN
ejpam-5354	109	2	=	=	SYM
ejpam-5354	109	3	0	0	X
ejpam-5354	109	4	.	.	PUNCT
ejpam-5354	110	1	theorem	theorem	NOUN
ejpam-5354	110	2	2	2	NUM
ejpam-5354	110	3	:	:	PUNCT
ejpam-5354	110	4	if	if	SCONJ
ejpam-5354	110	5	above	above	ADJ
ejpam-5354	110	6	assumptions	assumption	NOUN
ejpam-5354	110	7	are	be	AUX
ejpam-5354	110	8	true	true	ADJ
ejpam-5354	110	9	and	and	CCONJ
ejpam-5354	110	10	the	the	DET
ejpam-5354	110	11	corresponding	correspond	VERB
ejpam-5354	110	12	sequences	sequence	NOUN
ejpam-5354	110	13	{	{	PUNCT
ejpam-5354	110	14	xk	xk	NOUN
ejpam-5354	110	15	}	}	PUNCT
ejpam-5354	110	16	,	,	PUNCT
ejpam-5354	110	17	{	{	PUNCT
ejpam-5354	110	18	dk	dk	X
ejpam-5354	110	19	}	}	PUNCT
ejpam-5354	110	20	,	,	PUNCT
ejpam-5354	110	21	{	{	PUNCT
ejpam-5354	110	22	gk	gk	NOUN
ejpam-5354	110	23	}	}	PUNCT
ejpam-5354	110	24	,	,	PUNCT
ejpam-5354	110	25	{	{	PUNCT
ejpam-5354	110	26	αk	αk	AUX
ejpam-5354	110	27	}	}	PUNCT
ejpam-5354	110	28	are	be	AUX
ejpam-5354	110	29	generated	generate	VERB
ejpam-5354	110	30	by	by	ADP
ejpam-5354	110	31	the	the	DET
ejpam-5354	110	32	new	new	ADJ
ejpam-5354	110	33	method	method	NOUN
ejpam-5354	110	34	(	(	PUNCT
ejpam-5354	110	35	2	2	NUM
ejpam-5354	110	36	)	)	PUNCT
ejpam-5354	110	37	,	,	PUNCT
ejpam-5354	110	38	(	(	PUNCT
ejpam-5354	110	39	3	3	NUM
ejpam-5354	110	40	)	)	PUNCT
ejpam-5354	110	41	,	,	PUNCT
ejpam-5354	110	42	and	and	CCONJ
ejpam-5354	110	43	(	(	PUNCT
ejpam-5354	110	44	6	6	NUM
ejpam-5354	110	45	)	)	PUNCT
ejpam-5354	110	46	,	,	PUNCT
ejpam-5354	110	47	then	then	ADV
ejpam-5354	110	48	lim	lim	PROPN
ejpam-5354	110	49	k→∞	k→∞	PROPN
ejpam-5354	110	50	inf	inf	PROPN
ejpam-5354	110	51	∥gk+1∥	∥gk+1∥	NUM
ejpam-5354	110	52	=	=	NOUN
ejpam-5354	110	53	0	0	X
ejpam-5354	110	54	.	.	X
ejpam-5354	111	1	proof	proof	NOUN
ejpam-5354	111	2	:	:	PUNCT
ejpam-5354	111	3	from	from	ADP
ejpam-5354	111	4	equations	equation	NOUN
ejpam-5354	111	5	(	(	PUNCT
ejpam-5354	111	6	3	3	NUM
ejpam-5354	111	7	)	)	PUNCT
ejpam-5354	111	8	,	,	PUNCT
ejpam-5354	111	9	and	and	CCONJ
ejpam-5354	111	10	(	(	PUNCT
ejpam-5354	111	11	6	6	NUM
ejpam-5354	111	12	)	)	PUNCT
ejpam-5354	111	13	,	,	PUNCT
ejpam-5354	111	14	we	we	PRON
ejpam-5354	111	15	have	have	VERB
ejpam-5354	111	16	alaa	alaa	PROPN
ejpam-5354	111	17	l.	l.	PROPN
ejpam-5354	111	18	ibrahim	ibrahim	PROPN
ejpam-5354	111	19	et	et	PROPN
ejpam-5354	112	1	al	al	PROPN
ejpam-5354	112	2	.	.	PUNCT
ejpam-5354	112	3	/	/	SYM
ejpam-5354	112	4	eur	eur	PROPN
ejpam-5354	112	5	.	.	PUNCT
ejpam-5354	113	1	j.	j.	PROPN
ejpam-5354	113	2	pure	pure	PROPN
ejpam-5354	113	3	appl	appl	PROPN
ejpam-5354	113	4	.	.	PROPN
ejpam-5354	113	5	math	math	PROPN
ejpam-5354	113	6	,	,	PUNCT
ejpam-5354	113	7	17	17	NUM
ejpam-5354	113	8	(	(	PUNCT
ejpam-5354	113	9	4	4	NUM
ejpam-5354	113	10	)	)	PUNCT
ejpam-5354	113	11	(	(	PUNCT
ejpam-5354	113	12	2024	2024	NUM
ejpam-5354	113	13	)	)	PUNCT
ejpam-5354	113	14	,	,	PUNCT
ejpam-5354	113	15	2692	2692	NUM
ejpam-5354	113	16	-	-	SYM
ejpam-5354	113	17	2705	2705	NUM
ejpam-5354	113	18	2697	2697	NUM
ejpam-5354	113	19	∥dk+1∥	∥dk+1∥	NUM
ejpam-5354	113	20	≤	≤	NUM
ejpam-5354	113	21	∥gk+1∥+	∥gk+1∥+	X
ejpam-5354	114	1	(	(	PUNCT
ejpam-5354	114	2	∣∣∣∣∣gtk+1yk	∣∣∣∣∣gtk+1yk	PROPN
ejpam-5354	114	3	dtk	dtk	PROPN
ejpam-5354	114	4	yk	yk	PROPN
ejpam-5354	114	5	−	−	PROPN
ejpam-5354	114	6	gtk+1vk	gtk+1vk	PROPN
ejpam-5354	114	7	(	(	PUNCT
ejpam-5354	114	8	1	1	NUM
ejpam-5354	114	9	+	+	NUM
ejpam-5354	114	10	σ	σ	NOUN
ejpam-5354	114	11	+	+	CCONJ
ejpam-5354	114	12	λ)dtk	λ)dtk	X
ejpam-5354	114	13	yk	yk	PROPN
ejpam-5354	114	14	∣∣∣∣∣	∣∣∣∣∣	ADJ
ejpam-5354	114	15	)	)	PUNCT
ejpam-5354	114	16	∥dk∥	∥dk∥	NOUN
ejpam-5354	114	17	,	,	PUNCT
ejpam-5354	114	18	≤∥gk+1∥+	≤∥gk+1∥+	PROPN
ejpam-5354	114	19	∣∣∣∣∣gtk+1yk	∣∣∣∣∣gtk+1yk	PROPN
ejpam-5354	114	20	dtk	dtk	PROPN
ejpam-5354	114	21	yk	yk	PROPN
ejpam-5354	114	22	∣∣∣∣∣	∣∣∣∣∣	PROPN
ejpam-5354	114	23	∥dk∥+	∥dk∥+	PROPN
ejpam-5354	115	1	∣∣∣∣∣	∣∣∣∣∣	PROPN
ejpam-5354	116	1	gtk+1vk	gtk+1vk	X
ejpam-5354	116	2	(	(	PUNCT
ejpam-5354	116	3	1	1	NUM
ejpam-5354	116	4	+	+	NUM
ejpam-5354	116	5	σ	σ	NOUN
ejpam-5354	116	6	+	+	CCONJ
ejpam-5354	116	7	λ)dkt	λ)dkt	PROPN
ejpam-5354	116	8	yk	yk	PROPN
ejpam-5354	116	9	∣∣∣∣∣	∣∣∣∣∣	PROPN
ejpam-5354	116	10	∥dk∥	∥dk∥	PROPN
ejpam-5354	116	11	.	.	PUNCT
ejpam-5354	117	1	(	(	PUNCT
ejpam-5354	117	2	13	13	NUM
ejpam-5354	117	3	)	)	PUNCT
ejpam-5354	117	4	given	give	VERB
ejpam-5354	117	5	that	that	SCONJ
ejpam-5354	117	6	gtk+1vk	gtk+1vk	PROPN
ejpam-5354	117	7	≤	≤	PROPN
ejpam-5354	117	8	αkd	αkd	NOUN
ejpam-5354	117	9	t	t	PROPN
ejpam-5354	117	10	k	k	PROPN
ejpam-5354	117	11	yk	yk	PROPN
ejpam-5354	117	12	and	and	CCONJ
ejpam-5354	117	13	considering	consider	VERB
ejpam-5354	117	14	the	the	DET
ejpam-5354	117	15	lipschitz	lipschitz	NOUN
ejpam-5354	117	16	condition	condition	NOUN
ejpam-5354	117	17	∥yk∥	∥yk∥	PRON
ejpam-5354	117	18	≤	≤	ADJ
ejpam-5354	117	19	l∥vk∥	l∥vk∥	NOUN
ejpam-5354	117	20	with	with	ADP
ejpam-5354	117	21	l	l	PROPN
ejpam-5354	117	22	>	>	X
ejpam-5354	117	23	0	0	NUM
ejpam-5354	117	24	,	,	PUNCT
ejpam-5354	117	25	along	along	ADP
ejpam-5354	117	26	with	with	ADP
ejpam-5354	117	27	the	the	DET
ejpam-5354	117	28	use	use	NOUN
ejpam-5354	117	29	of	of	ADP
ejpam-5354	117	30	equation	equation	NOUN
ejpam-5354	117	31	(	(	PUNCT
ejpam-5354	117	32	12	12	NUM
ejpam-5354	117	33	)	)	PUNCT
ejpam-5354	117	34	,	,	PUNCT
ejpam-5354	117	35	it	it	PRON
ejpam-5354	117	36	follows	follow	VERB
ejpam-5354	117	37	that	that	SCONJ
ejpam-5354	117	38	∥dk+1∥	∥dk+1∥	NOUN
ejpam-5354	117	39	≤	≤	X
ejpam-5354	117	40	b+	b+	ADJ
ejpam-5354	117	41	αkbl	αkbl	NOUN
ejpam-5354	117	42	∥vk∥	∥vk∥	PROPN
ejpam-5354	117	43	ϑ	ϑ	PROPN
ejpam-5354	117	44	∥vk∥2	∥vk∥2	PROPN
ejpam-5354	117	45	∥dk∥+	∥dk∥+	PROPN
ejpam-5354	118	1	αk	αk	CCONJ
ejpam-5354	118	2	(	(	PUNCT
ejpam-5354	118	3	1	1	NUM
ejpam-5354	118	4	+	+	NUM
ejpam-5354	118	5	σ	σ	PROPN
ejpam-5354	118	6	+	+	CCONJ
ejpam-5354	118	7	λ	λ	PROPN
ejpam-5354	118	8	)	)	PUNCT
ejpam-5354	118	9	∥dk∥	∥dk∥	NOUN
ejpam-5354	118	10	.	.	PUNCT
ejpam-5354	119	1	(	(	PUNCT
ejpam-5354	119	2	14	14	NUM
ejpam-5354	119	3	)	)	PUNCT
ejpam-5354	119	4	since	since	SCONJ
ejpam-5354	119	5	,	,	PUNCT
ejpam-5354	119	6	∥vk∥	∥vk∥	PROPN
ejpam-5354	119	7	=	=	SYM
ejpam-5354	119	8	∥x−	∥x−	PROPN
ejpam-5354	119	9	xk∥	xk∥	PROPN
ejpam-5354	119	10	,	,	PUNCT
ejpam-5354	119	11	d	d	X
ejpam-5354	119	12	=	=	X
ejpam-5354	119	13	max	max	PROPN
ejpam-5354	119	14	{	{	PUNCT
ejpam-5354	119	15	∥x−	∥x−	PROPN
ejpam-5354	119	16	xk∥	xk∥	PROPN
ejpam-5354	119	17	}	}	PUNCT
ejpam-5354	119	18	,	,	PUNCT
ejpam-5354	119	19	∀x	∀x	X
ejpam-5354	119	20	,	,	PUNCT
ejpam-5354	119	21	xk	xk	PROPN
ejpam-5354	119	22	∈	∈	PROPN
ejpam-5354	119	23	r	r	X
ejpam-5354	119	24	}	}	PUNCT
ejpam-5354	119	25	.	.	PUNCT
ejpam-5354	120	1	hence	hence	ADV
ejpam-5354	120	2	equations	equation	NOUN
ejpam-5354	120	3	(	(	PUNCT
ejpam-5354	120	4	14	14	NUM
ejpam-5354	120	5	)	)	PUNCT
ejpam-5354	120	6	,	,	PUNCT
ejpam-5354	120	7	becomes	become	VERB
ejpam-5354	120	8	∥dk+1∥	∥dk+1∥	NUM
ejpam-5354	120	9	≤	≤	X
ejpam-5354	120	10	b+	b+	ADJ
ejpam-5354	120	11	(	(	PUNCT
ejpam-5354	120	12	bl	bl	INTJ
ejpam-5354	120	13	ϑ	ϑ	PROPN
ejpam-5354	121	1	+	+	X
ejpam-5354	121	2	d	d	X
ejpam-5354	121	3	(	(	PUNCT
ejpam-5354	121	4	1	1	NUM
ejpam-5354	121	5	+	+	NUM
ejpam-5354	121	6	σ	σ	PROPN
ejpam-5354	121	7	+	+	CCONJ
ejpam-5354	121	8	λ	λ	NOUN
ejpam-5354	121	9	)	)	PUNCT
ejpam-5354	121	10	)	)	PUNCT
ejpam-5354	122	1	=	=	PUNCT
ejpam-5354	122	2	β	β	X
ejpam-5354	122	3	⇒	⇒	VERB
ejpam-5354	122	4	∑	∑	PUNCT
ejpam-5354	122	5	k≥1	k≥1	PROPN
ejpam-5354	122	6	1	1	NUM
ejpam-5354	122	7	∥dk+1∥2	∥dk+1∥2	ADJ
ejpam-5354	122	8	≥	≥	NOUN
ejpam-5354	122	9	∑	∑	ADV
ejpam-5354	122	10	k≥1	k≥1	PROPN
ejpam-5354	122	11	1	1	NUM
ejpam-5354	122	12	β2	β2	NOUN
ejpam-5354	122	13	=	=	SYM
ejpam-5354	122	14	∞	∞	PROPN
ejpam-5354	122	15	which	which	PRON
ejpam-5354	122	16	completes	complete	VERB
ejpam-5354	122	17	the	the	DET
ejpam-5354	122	18	proof	proof	NOUN
ejpam-5354	122	19	.	.	PUNCT
ejpam-5354	123	1	3	3	X
ejpam-5354	123	2	.	.	X
ejpam-5354	123	3	numerical	numerical	PROPN
ejpam-5354	123	4	results	result	NOUN
ejpam-5354	123	5	the	the	DET
ejpam-5354	123	6	performance	performance	NOUN
ejpam-5354	123	7	of	of	ADP
ejpam-5354	123	8	the	the	DET
ejpam-5354	123	9	new	new	ADJ
ejpam-5354	123	10	method	method	NOUN
ejpam-5354	123	11	was	be	AUX
ejpam-5354	123	12	evaluated	evaluate	VERB
ejpam-5354	123	13	using	use	VERB
ejpam-5354	123	14	standard	standard	ADJ
ejpam-5354	123	15	test	test	NOUN
ejpam-5354	123	16	functions	function	NOUN
ejpam-5354	123	17	as	as	SCONJ
ejpam-5354	123	18	described	describe	VERB
ejpam-5354	123	19	by	by	ADP
ejpam-5354	123	20	andrei	andrei	NOUN
ejpam-5354	124	1	[	[	X
ejpam-5354	124	2	2	2	NUM
ejpam-5354	124	3	]	]	PUNCT
ejpam-5354	124	4	,	,	PUNCT
ejpam-5354	124	5	with	with	ADP
ejpam-5354	124	6	initial	initial	ADJ
ejpam-5354	124	7	points	point	NOUN
ejpam-5354	124	8	ranging	range	VERB
ejpam-5354	124	9	from	from	ADP
ejpam-5354	124	10	5	5	NUM
ejpam-5354	124	11	to	to	ADP
ejpam-5354	124	12	5000	5000	NUM
ejpam-5354	124	13	.	.	PUNCT
ejpam-5354	125	1	the	the	DET
ejpam-5354	125	2	assessment	assessment	NOUN
ejpam-5354	125	3	was	be	AUX
ejpam-5354	125	4	conducted	conduct	VERB
ejpam-5354	125	5	by	by	ADP
ejpam-5354	125	6	comparing	compare	VERB
ejpam-5354	125	7	the	the	DET
ejpam-5354	125	8	new	new	ADJ
ejpam-5354	125	9	method	method	NOUN
ejpam-5354	125	10	against	against	ADP
ejpam-5354	125	11	the	the	DET
ejpam-5354	125	12	classical	classical	ADJ
ejpam-5354	125	13	hestenes	hestene	NOUN
ejpam-5354	125	14	-	-	PUNCT
ejpam-5354	125	15	stiefel	stiefel	NOUN
ejpam-5354	125	16	(	(	PUNCT
ejpam-5354	125	17	hs	hs	NOUN
ejpam-5354	125	18	)	)	PUNCT
ejpam-5354	125	19	conjugate	conjugate	ADJ
ejpam-5354	125	20	gradient	gradient	ADJ
ejpam-5354	125	21	method	method	NOUN
ejpam-5354	125	22	.	.	PUNCT
ejpam-5354	126	1	key	key	ADJ
ejpam-5354	126	2	metrics	metric	NOUN
ejpam-5354	126	3	for	for	ADP
ejpam-5354	126	4	comparison	comparison	NOUN
ejpam-5354	126	5	include	include	VERB
ejpam-5354	126	6	the	the	DET
ejpam-5354	126	7	number	number	NOUN
ejpam-5354	126	8	of	of	ADP
ejpam-5354	126	9	iterations	iteration	NOUN
ejpam-5354	126	10	(	(	PUNCT
ejpam-5354	126	11	noi	noi	PROPN
ejpam-5354	126	12	)	)	PUNCT
ejpam-5354	126	13	required	require	VERB
ejpam-5354	126	14	for	for	ADP
ejpam-5354	126	15	convergence	convergence	NOUN
ejpam-5354	126	16	and	and	CCONJ
ejpam-5354	126	17	the	the	DET
ejpam-5354	126	18	number	number	NOUN
ejpam-5354	126	19	of	of	ADP
ejpam-5354	126	20	function	function	NOUN
ejpam-5354	126	21	evaluations	evaluation	NOUN
ejpam-5354	126	22	(	(	PUNCT
ejpam-5354	126	23	nof	nof	PROPN
ejpam-5354	126	24	)	)	PUNCT
ejpam-5354	126	25	during	during	ADP
ejpam-5354	126	26	the	the	DET
ejpam-5354	126	27	optimization	optimization	NOUN
ejpam-5354	126	28	process	process	NOUN
ejpam-5354	126	29	.	.	PUNCT
ejpam-5354	127	1	all	all	DET
ejpam-5354	127	2	programs	program	NOUN
ejpam-5354	127	3	were	be	AUX
ejpam-5354	127	4	implemented	implement	VERB
ejpam-5354	127	5	in	in	ADP
ejpam-5354	127	6	fortran	fortran	NOUN
ejpam-5354	127	7	95	95	NUM
ejpam-5354	127	8	,	,	PUNCT
ejpam-5354	127	9	and	and	CCONJ
ejpam-5354	127	10	a	a	DET
ejpam-5354	127	11	cubic	cubic	ADJ
ejpam-5354	127	12	interpolation	interpolation	NOUN
ejpam-5354	127	13	line	line	NOUN
ejpam-5354	127	14	search	search	NOUN
ejpam-5354	127	15	method	method	NOUN
ejpam-5354	127	16	,	,	PUNCT
ejpam-5354	127	17	which	which	PRON
ejpam-5354	127	18	uses	use	VERB
ejpam-5354	127	19	both	both	PRON
ejpam-5354	127	20	function	function	NOUN
ejpam-5354	127	21	and	and	CCONJ
ejpam-5354	127	22	gradient	gradient	ADJ
ejpam-5354	127	23	information	information	NOUN
ejpam-5354	127	24	,	,	PUNCT
ejpam-5354	127	25	was	be	AUX
ejpam-5354	127	26	employed	employ	VERB
ejpam-5354	127	27	.	.	PUNCT
ejpam-5354	128	1	the	the	DET
ejpam-5354	128	2	results	result	NOUN
ejpam-5354	128	3	are	be	AUX
ejpam-5354	128	4	summarized	summarize	VERB
ejpam-5354	128	5	in	in	ADP
ejpam-5354	128	6	tables	table	NOUN
ejpam-5354	128	7	1	1	NUM
ejpam-5354	128	8	and	and	CCONJ
ejpam-5354	128	9	2	2	NUM
ejpam-5354	128	10	.	.	NOUN
ejpam-5354	128	11	table	table	NOUN
ejpam-5354	128	12	1	1	NUM
ejpam-5354	128	13	:	:	PUNCT
ejpam-5354	128	14	the	the	DET
ejpam-5354	128	15	results	result	NOUN
ejpam-5354	128	16	of	of	ADP
ejpam-5354	128	17	the	the	DET
ejpam-5354	128	18	new	new	ADJ
ejpam-5354	128	19	method	method	NOUN
ejpam-5354	128	20	with	with	ADP
ejpam-5354	128	21	standard	standard	ADJ
ejpam-5354	128	22	method	method	NOUN
ejpam-5354	128	23	.	.	PUNCT
ejpam-5354	129	1	test	test	NOUN
ejpam-5354	129	2	function	function	PROPN
ejpam-5354	129	3	n	n	PROPN
ejpam-5354	129	4	hs	hs	PROPN
ejpam-5354	129	5	new	new	PROPN
ejpam-5354	129	6	noi	noi	PROPN
ejpam-5354	129	7	nof	nof	PROPN
ejpam-5354	129	8	noi	noi	PROPN
ejpam-5354	129	9	nof	nof	PROPN
ejpam-5354	129	10	mile	mile	PROPN
ejpam-5354	129	11	4	4	NUM
ejpam-5354	129	12	28	28	NUM
ejpam-5354	129	13	85	85	NUM
ejpam-5354	129	14	27	27	NUM
ejpam-5354	129	15	83	83	NUM
ejpam-5354	129	16	10	10	NUM
ejpam-5354	129	17	31	31	NUM
ejpam-5354	129	18	102	102	NUM
ejpam-5354	129	19	27	27	NUM
ejpam-5354	129	20	83	83	NUM
ejpam-5354	129	21	100	100	NUM
ejpam-5354	129	22	33	33	NUM
ejpam-5354	129	23	114	114	NUM
ejpam-5354	129	24	30	30	NUM
ejpam-5354	129	25	98	98	NUM
ejpam-5354	129	26	500	500	NUM
ejpam-5354	129	27	40	40	NUM
ejpam-5354	129	28	147	147	NUM
ejpam-5354	129	29	32	32	NUM
ejpam-5354	129	30	105	105	NUM
ejpam-5354	129	31	1000	1000	NUM
ejpam-5354	129	32	46	46	NUM
ejpam-5354	129	33	176	176	NUM
ejpam-5354	129	34	33	33	NUM
ejpam-5354	129	35	129	129	NUM
ejpam-5354	129	36	5000	5000	NUM
ejpam-5354	129	37	54	54	NUM
ejpam-5354	129	38	211	211	NUM
ejpam-5354	129	39	33	33	NUM
ejpam-5354	129	40	129	129	NUM
ejpam-5354	130	1	alaa	alaa	PROPN
ejpam-5354	130	2	l.	l.	PROPN
ejpam-5354	130	3	ibrahim	ibrahim	PROPN
ejpam-5354	130	4	et	et	PROPN
ejpam-5354	130	5	al	al	PROPN
ejpam-5354	130	6	.	.	PUNCT
ejpam-5354	130	7	/	/	SYM
ejpam-5354	130	8	eur	eur	PROPN
ejpam-5354	130	9	.	.	PUNCT
ejpam-5354	131	1	j.	j.	PROPN
ejpam-5354	131	2	pure	pure	PROPN
ejpam-5354	131	3	appl	appl	PROPN
ejpam-5354	131	4	.	.	PROPN
ejpam-5354	131	5	math	math	PROPN
ejpam-5354	131	6	,	,	PUNCT
ejpam-5354	131	7	17	17	NUM
ejpam-5354	131	8	(	(	PUNCT
ejpam-5354	131	9	4	4	NUM
ejpam-5354	131	10	)	)	PUNCT
ejpam-5354	131	11	(	(	PUNCT
ejpam-5354	131	12	2024	2024	NUM
ejpam-5354	131	13	)	)	PUNCT
ejpam-5354	131	14	,	,	PUNCT
ejpam-5354	131	15	2692	2692	NUM
ejpam-5354	131	16	-	-	SYM
ejpam-5354	131	17	2705	2705	NUM
ejpam-5354	131	18	2698	2698	NUM
ejpam-5354	131	19	powell	powell	NOUN
ejpam-5354	131	20	4	4	NUM
ejpam-5354	131	21	38	38	NUM
ejpam-5354	131	22	108	108	NUM
ejpam-5354	131	23	31	31	NUM
ejpam-5354	131	24	87	87	NUM
ejpam-5354	131	25	10	10	NUM
ejpam-5354	131	26	38	38	NUM
ejpam-5354	131	27	108	108	NUM
ejpam-5354	131	28	31	31	NUM
ejpam-5354	131	29	87	87	NUM
ejpam-5354	131	30	100	100	NUM
ejpam-5354	131	31	40	40	NUM
ejpam-5354	131	32	122	122	NUM
ejpam-5354	131	33	33	33	NUM
ejpam-5354	131	34	102	102	NUM
ejpam-5354	131	35	500	500	NUM
ejpam-5354	131	36	41	41	NUM
ejpam-5354	131	37	124	124	NUM
ejpam-5354	131	38	33	33	NUM
ejpam-5354	131	39	102	102	NUM
ejpam-5354	131	40	1000	1000	NUM
ejpam-5354	131	41	41	41	NUM
ejpam-5354	132	1	124	124	NUM
ejpam-5354	132	2	33	33	NUM
ejpam-5354	132	3	102	102	NUM
ejpam-5354	132	4	5000	5000	NUM
ejpam-5354	132	5	41	41	NUM
ejpam-5354	132	6	124	124	NUM
ejpam-5354	132	7	33	33	NUM
ejpam-5354	132	8	102	102	NUM
ejpam-5354	132	9	wolfe	wolfe	PROPN
ejpam-5354	132	10	5	5	NUM
ejpam-5354	132	11	11	11	NUM
ejpam-5354	132	12	24	24	NUM
ejpam-5354	132	13	11	11	NUM
ejpam-5354	132	14	24	24	NUM
ejpam-5354	132	15	10	10	NUM
ejpam-5354	132	16	32	32	NUM
ejpam-5354	132	17	65	65	NUM
ejpam-5354	132	18	32	32	NUM
ejpam-5354	132	19	65	65	NUM
ejpam-5354	132	20	100	100	NUM
ejpam-5354	132	21	49	49	NUM
ejpam-5354	132	22	99	99	NUM
ejpam-5354	132	23	49	49	NUM
ejpam-5354	132	24	99	99	NUM
ejpam-5354	132	25	500	500	NUM
ejpam-5354	132	26	52	52	NUM
ejpam-5354	132	27	105	105	NUM
ejpam-5354	132	28	48	48	NUM
ejpam-5354	132	29	99	99	NUM
ejpam-5354	132	30	1000	1000	NUM
ejpam-5354	132	31	70	70	NUM
ejpam-5354	132	32	141	141	NUM
ejpam-5354	132	33	43	43	NUM
ejpam-5354	132	34	91	91	NUM
ejpam-5354	132	35	5000	5000	NUM
ejpam-5354	132	36	165	165	NUM
ejpam-5354	132	37	348	348	NUM
ejpam-5354	132	38	102	102	NUM
ejpam-5354	132	39	227	227	NUM
ejpam-5354	132	40	g	g	NOUN
ejpam-5354	132	41	-	-	PUNCT
ejpam-5354	132	42	cantrel	cantrel	NOUN
ejpam-5354	132	43	4	4	NUM
ejpam-5354	132	44	22	22	NUM
ejpam-5354	132	45	159	159	NUM
ejpam-5354	132	46	14	14	NUM
ejpam-5354	132	47	47	47	NUM
ejpam-5354	132	48	10	10	NUM
ejpam-5354	132	49	22	22	NUM
ejpam-5354	132	50	159	159	NUM
ejpam-5354	132	51	15	15	NUM
ejpam-5354	132	52	58	58	NUM
ejpam-5354	132	53	100	100	NUM
ejpam-5354	132	54	22	22	NUM
ejpam-5354	132	55	159	159	NUM
ejpam-5354	132	56	17	17	NUM
ejpam-5354	132	57	74	74	NUM
ejpam-5354	132	58	500	500	NUM
ejpam-5354	132	59	23	23	NUM
ejpam-5354	132	60	171	171	NUM
ejpam-5354	132	61	17	17	NUM
ejpam-5354	132	62	74	74	NUM
ejpam-5354	132	63	1000	1000	NUM
ejpam-5354	132	64	23	23	NUM
ejpam-5354	132	65	171	171	NUM
ejpam-5354	132	66	21	21	NUM
ejpam-5354	132	67	110	110	NUM
ejpam-5354	132	68	5000	5000	NUM
ejpam-5354	132	69	28	28	NUM
ejpam-5354	132	70	248	248	NUM
ejpam-5354	132	71	23	23	NUM
ejpam-5354	132	72	126	126	NUM
ejpam-5354	132	73	osp	osp	PROPN
ejpam-5354	132	74	.	.	PUNCT
ejpam-5354	133	1	4	4	NUM
ejpam-5354	133	2	8	8	NUM
ejpam-5354	133	3	45	45	NUM
ejpam-5354	133	4	8	8	NUM
ejpam-5354	133	5	43	43	NUM
ejpam-5354	133	6	10	10	NUM
ejpam-5354	133	7	13	13	NUM
ejpam-5354	133	8	58	58	NUM
ejpam-5354	133	9	13	13	NUM
ejpam-5354	133	10	53	53	NUM
ejpam-5354	133	11	100	100	NUM
ejpam-5354	133	12	49	49	NUM
ejpam-5354	133	13	185	185	NUM
ejpam-5354	133	14	45	45	NUM
ejpam-5354	133	15	160	160	NUM
ejpam-5354	133	16	500	500	NUM
ejpam-5354	133	17	112	112	NUM
ejpam-5354	133	18	353	353	NUM
ejpam-5354	133	19	100	100	NUM
ejpam-5354	133	20	316	316	NUM
ejpam-5354	133	21	1000	1000	NUM
ejpam-5354	133	22	156	156	NUM
ejpam-5354	133	23	473	473	NUM
ejpam-5354	133	24	150	150	NUM
ejpam-5354	133	25	471	471	NUM
ejpam-5354	133	26	5000	5000	NUM
ejpam-5354	133	27	256	256	NUM
ejpam-5354	133	28	774	774	NUM
ejpam-5354	133	29	250	250	NUM
ejpam-5354	133	30	772	772	NUM
ejpam-5354	133	31	wood	wood	NOUN
ejpam-5354	133	32	5	5	NUM
ejpam-5354	133	33	30	30	NUM
ejpam-5354	133	34	68	68	NUM
ejpam-5354	133	35	28	28	NUM
ejpam-5354	133	36	62	62	NUM
ejpam-5354	133	37	10	10	NUM
ejpam-5354	133	38	30	30	NUM
ejpam-5354	133	39	68	68	NUM
ejpam-5354	133	40	28	28	NUM
ejpam-5354	133	41	62	62	NUM
ejpam-5354	133	42	100	100	NUM
ejpam-5354	133	43	30	30	NUM
ejpam-5354	133	44	68	68	NUM
ejpam-5354	133	45	28	28	NUM
ejpam-5354	133	46	62	62	NUM
ejpam-5354	133	47	500	500	NUM
ejpam-5354	133	48	30	30	NUM
ejpam-5354	133	49	68	68	NUM
ejpam-5354	133	50	28	28	NUM
ejpam-5354	133	51	62	62	NUM
ejpam-5354	133	52	1000	1000	NUM
ejpam-5354	133	53	30	30	NUM
ejpam-5354	133	54	68	68	NUM
ejpam-5354	133	55	28	28	NUM
ejpam-5354	133	56	62	62	NUM
ejpam-5354	133	57	5000	5000	NUM
ejpam-5354	133	58	30	30	NUM
ejpam-5354	133	59	68	68	NUM
ejpam-5354	133	60	28	28	NUM
ejpam-5354	133	61	62	62	NUM
ejpam-5354	133	62	rosen	rosen	NOUN
ejpam-5354	133	63	4	4	NUM
ejpam-5354	133	64	30	30	NUM
ejpam-5354	133	65	83	83	NUM
ejpam-5354	133	66	29	29	NUM
ejpam-5354	133	67	80	80	NUM
ejpam-5354	133	68	10	10	NUM
ejpam-5354	133	69	30	30	NUM
ejpam-5354	133	70	83	83	NUM
ejpam-5354	133	71	29	29	NUM
ejpam-5354	133	72	80	80	NUM
ejpam-5354	133	73	100	100	NUM
ejpam-5354	133	74	30	30	NUM
ejpam-5354	133	75	83	83	NUM
ejpam-5354	133	76	29	29	NUM
ejpam-5354	133	77	80	80	NUM
ejpam-5354	133	78	500	500	NUM
ejpam-5354	133	79	30	30	NUM
ejpam-5354	133	80	83	83	NUM
ejpam-5354	133	81	29	29	NUM
ejpam-5354	133	82	80	80	NUM
ejpam-5354	133	83	1000	1000	NUM
ejpam-5354	133	84	30	30	NUM
ejpam-5354	133	85	83	83	NUM
ejpam-5354	133	86	29	29	NUM
ejpam-5354	133	87	80	80	NUM
ejpam-5354	133	88	5000	5000	NUM
ejpam-5354	133	89	30	30	NUM
ejpam-5354	133	90	83	83	NUM
ejpam-5354	133	91	29	29	NUM
ejpam-5354	133	92	80	80	NUM
ejpam-5354	133	93	edger	edger	NOUN
ejpam-5354	133	94	4	4	NUM
ejpam-5354	133	95	5	5	NUM
ejpam-5354	133	96	14	14	NUM
ejpam-5354	133	97	5	5	NUM
ejpam-5354	133	98	14	14	NUM
ejpam-5354	133	99	10	10	NUM
ejpam-5354	133	100	5	5	NUM
ejpam-5354	133	101	14	14	NUM
ejpam-5354	133	102	5	5	NUM
ejpam-5354	133	103	14	14	NUM
ejpam-5354	133	104	100	100	NUM
ejpam-5354	133	105	5	5	NUM
ejpam-5354	133	106	14	14	NUM
ejpam-5354	133	107	5	5	NUM
ejpam-5354	133	108	14	14	NUM
ejpam-5354	133	109	500	500	NUM
ejpam-5354	133	110	6	6	NUM
ejpam-5354	133	111	16	16	NUM
ejpam-5354	133	112	5	5	NUM
ejpam-5354	133	113	14	14	NUM
ejpam-5354	133	114	1000	1000	NUM
ejpam-5354	133	115	6	6	NUM
ejpam-5354	133	116	16	16	NUM
ejpam-5354	133	117	6	6	NUM
ejpam-5354	133	118	16	16	NUM
ejpam-5354	133	119	5000	5000	NUM
ejpam-5354	133	120	6	6	NUM
ejpam-5354	133	121	16	16	NUM
ejpam-5354	133	122	6	6	NUM
ejpam-5354	133	123	16	16	NUM
ejpam-5354	133	124	alaa	alaa	PROPN
ejpam-5354	133	125	l.	l.	PROPN
ejpam-5354	133	126	ibrahim	ibrahim	PROPN
ejpam-5354	133	127	et	et	PROPN
ejpam-5354	133	128	al	al	PROPN
ejpam-5354	133	129	.	.	PUNCT
ejpam-5354	133	130	/	/	SYM
ejpam-5354	133	131	eur	eur	PROPN
ejpam-5354	133	132	.	.	PUNCT
ejpam-5354	134	1	j.	j.	PROPN
ejpam-5354	134	2	pure	pure	PROPN
ejpam-5354	134	3	appl	appl	PROPN
ejpam-5354	134	4	.	.	PROPN
ejpam-5354	134	5	math	math	PROPN
ejpam-5354	134	6	,	,	PUNCT
ejpam-5354	134	7	17	17	NUM
ejpam-5354	134	8	(	(	PUNCT
ejpam-5354	134	9	4	4	NUM
ejpam-5354	134	10	)	)	PUNCT
ejpam-5354	134	11	(	(	PUNCT
ejpam-5354	134	12	2024	2024	NUM
ejpam-5354	134	13	)	)	PUNCT
ejpam-5354	134	14	,	,	PUNCT
ejpam-5354	134	15	2692	2692	NUM
ejpam-5354	134	16	-	-	SYM
ejpam-5354	134	17	2705	2705	NUM
ejpam-5354	134	18	2699	2699	NUM
ejpam-5354	134	19	shallow	shallow	ADJ
ejpam-5354	134	20	4	4	NUM
ejpam-5354	134	21	8	8	NUM
ejpam-5354	134	22	21	21	NUM
ejpam-5354	134	23	8	8	NUM
ejpam-5354	134	24	21	21	NUM
ejpam-5354	134	25	10	10	NUM
ejpam-5354	134	26	8	8	NUM
ejpam-5354	134	27	21	21	NUM
ejpam-5354	134	28	8	8	NUM
ejpam-5354	134	29	21	21	NUM
ejpam-5354	134	30	100	100	NUM
ejpam-5354	134	31	8	8	NUM
ejpam-5354	134	32	21	21	NUM
ejpam-5354	134	33	8	8	NUM
ejpam-5354	134	34	21	21	NUM
ejpam-5354	134	35	500	500	NUM
ejpam-5354	134	36	8	8	NUM
ejpam-5354	134	37	21	21	NUM
ejpam-5354	134	38	8	8	NUM
ejpam-5354	134	39	21	21	NUM
ejpam-5354	134	40	1000	1000	NUM
ejpam-5354	134	41	9	9	NUM
ejpam-5354	134	42	24	24	NUM
ejpam-5354	134	43	9	9	NUM
ejpam-5354	134	44	24	24	NUM
ejpam-5354	134	45	5000	5000	NUM
ejpam-5354	134	46	9	9	NUM
ejpam-5354	134	47	24	24	NUM
ejpam-5354	134	48	9	9	NUM
ejpam-5354	134	49	24	24	NUM
ejpam-5354	134	50	total	total	ADJ
ejpam-5354	134	51	2027	2027	NUM
ejpam-5354	134	52	6410	6410	NUM
ejpam-5354	134	53	1758	1758	NUM
ejpam-5354	134	54	5190	5190	NUM
ejpam-5354	134	55	table	table	NOUN
ejpam-5354	134	56	2	2	NUM
ejpam-5354	134	57	:	:	PUNCT
ejpam-5354	134	58	the	the	DET
ejpam-5354	134	59	improvement	improvement	NOUN
ejpam-5354	134	60	percentage	percentage	NOUN
ejpam-5354	134	61	of	of	ADP
ejpam-5354	134	62	the	the	DET
ejpam-5354	134	63	new	new	ADJ
ejpam-5354	134	64	method	method	NOUN
ejpam-5354	134	65	compared	compare	VERB
ejpam-5354	134	66	to	to	ADP
ejpam-5354	134	67	the	the	DET
ejpam-5354	134	68	standard	standard	ADJ
ejpam-5354	134	69	method	method	NOUN
ejpam-5354	134	70	.	.	PUNCT
ejpam-5354	135	1	tools	tool	NOUN
ejpam-5354	135	2	hs	hs	PROPN
ejpam-5354	135	3	(	(	PUNCT
ejpam-5354	135	4	standard	standard	NOUN
ejpam-5354	135	5	)	)	PUNCT
ejpam-5354	135	6	new	new	ADJ
ejpam-5354	135	7	(	(	PUNCT
ejpam-5354	135	8	proposed	propose	VERB
ejpam-5354	135	9	)	)	PUNCT
ejpam-5354	135	10	noi	noi	PROPN
ejpam-5354	135	11	100	100	NUM
ejpam-5354	135	12	%	%	NOUN
ejpam-5354	135	13	86.7292	86.7292	NUM
ejpam-5354	135	14	%	%	NOUN
ejpam-5354	135	15	nof	nof	NOUN
ejpam-5354	135	16	100	100	NUM
ejpam-5354	135	17	%	%	NOUN
ejpam-5354	135	18	80.967	80.967	NUM
ejpam-5354	135	19	%	%	NOUN
ejpam-5354	135	20	table	table	NOUN
ejpam-5354	135	21	2	2	NUM
ejpam-5354	135	22	presents	present	VERB
ejpam-5354	135	23	the	the	DET
ejpam-5354	135	24	improvement	improvement	NOUN
ejpam-5354	135	25	percentages	percentage	NOUN
ejpam-5354	135	26	of	of	ADP
ejpam-5354	135	27	the	the	DET
ejpam-5354	135	28	new	new	ADJ
ejpam-5354	135	29	method	method	NOUN
ejpam-5354	135	30	compared	compare	VERB
ejpam-5354	135	31	to	to	ADP
ejpam-5354	135	32	the	the	DET
ejpam-5354	135	33	standard	standard	ADJ
ejpam-5354	135	34	hs	hs	PROPN
ejpam-5354	135	35	method	method	NOUN
ejpam-5354	135	36	across	across	ADP
ejpam-5354	135	37	two	two	NUM
ejpam-5354	135	38	key	key	ADJ
ejpam-5354	135	39	metrics	metric	NOUN
ejpam-5354	135	40	:	:	PUNCT
ejpam-5354	135	41	the	the	DET
ejpam-5354	135	42	number	number	NOUN
ejpam-5354	135	43	of	of	ADP
ejpam-5354	135	44	iterations	iteration	NOUN
ejpam-5354	135	45	(	(	PUNCT
ejpam-5354	135	46	noi	noi	PROPN
ejpam-5354	135	47	)	)	PUNCT
ejpam-5354	135	48	and	and	CCONJ
ejpam-5354	135	49	the	the	DET
ejpam-5354	135	50	number	number	NOUN
ejpam-5354	135	51	of	of	ADP
ejpam-5354	135	52	function	function	NOUN
ejpam-5354	135	53	evaluations	evaluation	NOUN
ejpam-5354	135	54	(	(	PUNCT
ejpam-5354	135	55	nof	nof	PROPN
ejpam-5354	135	56	)	)	PUNCT
ejpam-5354	135	57	.	.	PUNCT
ejpam-5354	136	1	the	the	DET
ejpam-5354	136	2	new	new	ADJ
ejpam-5354	136	3	method	method	NOUN
ejpam-5354	136	4	achieves	achieve	VERB
ejpam-5354	136	5	a	a	DET
ejpam-5354	136	6	reduction	reduction	NOUN
ejpam-5354	136	7	of	of	ADP
ejpam-5354	136	8	approximately	approximately	ADV
ejpam-5354	136	9	13.27	13.27	NUM
ejpam-5354	136	10	%	%	NOUN
ejpam-5354	136	11	in	in	ADP
ejpam-5354	136	12	the	the	DET
ejpam-5354	136	13	number	number	NOUN
ejpam-5354	136	14	of	of	ADP
ejpam-5354	136	15	iterations	iteration	NOUN
ejpam-5354	136	16	and	and	CCONJ
ejpam-5354	136	17	19.03	19.03	NUM
ejpam-5354	136	18	%	%	NOUN
ejpam-5354	136	19	in	in	ADP
ejpam-5354	136	20	the	the	DET
ejpam-5354	136	21	number	number	NOUN
ejpam-5354	136	22	of	of	ADP
ejpam-5354	136	23	function	function	NOUN
ejpam-5354	136	24	evaluations	evaluation	NOUN
ejpam-5354	136	25	,	,	PUNCT
ejpam-5354	136	26	demonstrating	demonstrate	VERB
ejpam-5354	136	27	its	its	PRON
ejpam-5354	136	28	enhanced	enhance	VERB
ejpam-5354	136	29	efficiency	efficiency	NOUN
ejpam-5354	136	30	over	over	ADP
ejpam-5354	136	31	the	the	DET
ejpam-5354	136	32	standard	standard	ADJ
ejpam-5354	136	33	hs	hs	PROPN
ejpam-5354	136	34	method	method	NOUN
ejpam-5354	136	35	.	.	PUNCT
ejpam-5354	137	1	4	4	X
ejpam-5354	137	2	.	.	X
ejpam-5354	137	3	application	application	NOUN
ejpam-5354	137	4	in	in	ADP
ejpam-5354	137	5	neural	neural	ADJ
ejpam-5354	137	6	networks	network	NOUN
ejpam-5354	137	7	artificial	artificial	ADJ
ejpam-5354	137	8	intelligence	intelligence	NOUN
ejpam-5354	137	9	(	(	PUNCT
ejpam-5354	137	10	ai	ai	NOUN
ejpam-5354	137	11	)	)	PUNCT
ejpam-5354	137	12	is	be	AUX
ejpam-5354	137	13	designed	design	VERB
ejpam-5354	137	14	to	to	PART
ejpam-5354	137	15	emulate	emulate	VERB
ejpam-5354	137	16	human	human	ADJ
ejpam-5354	137	17	brain	brain	NOUN
ejpam-5354	137	18	functionality	functionality	NOUN
ejpam-5354	137	19	[	[	X
ejpam-5354	137	20	24	24	NUM
ejpam-5354	137	21	]	]	PUNCT
ejpam-5354	137	22	.	.	PUNCT
ejpam-5354	138	1	among	among	ADP
ejpam-5354	138	2	the	the	DET
ejpam-5354	138	3	various	various	ADJ
ejpam-5354	138	4	branches	branch	NOUN
ejpam-5354	138	5	of	of	ADP
ejpam-5354	138	6	ai	ai	PROPN
ejpam-5354	138	7	,	,	PUNCT
ejpam-5354	138	8	artificial	artificial	ADJ
ejpam-5354	138	9	neural	neural	ADJ
ejpam-5354	138	10	networks	network	NOUN
ejpam-5354	138	11	(	(	PUNCT
ejpam-5354	138	12	anns	anns	PROPN
ejpam-5354	138	13	)	)	PUNCT
ejpam-5354	138	14	have	have	AUX
ejpam-5354	138	15	gained	gain	VERB
ejpam-5354	138	16	significant	significant	ADJ
ejpam-5354	138	17	popularity	popularity	NOUN
ejpam-5354	138	18	[	[	X
ejpam-5354	138	19	14	14	NUM
ejpam-5354	138	20	]	]	PUNCT
ejpam-5354	138	21	.	.	PUNCT
ejpam-5354	139	1	anns	anns	PROPN
ejpam-5354	139	2	are	be	AUX
ejpam-5354	139	3	employed	employ	VERB
ejpam-5354	139	4	in	in	ADP
ejpam-5354	139	5	tasks	task	NOUN
ejpam-5354	139	6	such	such	ADJ
ejpam-5354	139	7	as	as	ADP
ejpam-5354	139	8	classification	classification	NOUN
ejpam-5354	139	9	,	,	PUNCT
ejpam-5354	139	10	optimization	optimization	NOUN
ejpam-5354	139	11	,	,	PUNCT
ejpam-5354	139	12	and	and	CCONJ
ejpam-5354	139	13	prediction	prediction	NOUN
ejpam-5354	139	14	of	of	ADP
ejpam-5354	139	15	given	give	VERB
ejpam-5354	139	16	datasets	dataset	NOUN
ejpam-5354	139	17	to	to	PART
ejpam-5354	139	18	produce	produce	VERB
ejpam-5354	139	19	appropriate	appropriate	ADJ
ejpam-5354	139	20	results	result	NOUN
ejpam-5354	139	21	or	or	CCONJ
ejpam-5354	139	22	outputs	output	NOUN
ejpam-5354	139	23	.	.	PUNCT
ejpam-5354	140	1	the	the	DET
ejpam-5354	140	2	training	training	NOUN
ejpam-5354	140	3	and	and	CCONJ
ejpam-5354	140	4	testing	testing	NOUN
ejpam-5354	140	5	phases	phase	NOUN
ejpam-5354	140	6	are	be	AUX
ejpam-5354	140	7	pivotal	pivotal	ADJ
ejpam-5354	140	8	in	in	ADP
ejpam-5354	140	9	anns	ann	NOUN
ejpam-5354	140	10	,	,	PUNCT
ejpam-5354	140	11	allowing	allow	VERB
ejpam-5354	140	12	the	the	DET
ejpam-5354	140	13	network	network	NOUN
ejpam-5354	140	14	to	to	PART
ejpam-5354	140	15	discern	discern	VERB
ejpam-5354	140	16	patterns	pattern	NOUN
ejpam-5354	140	17	within	within	ADP
ejpam-5354	140	18	the	the	DET
ejpam-5354	140	19	data	datum	NOUN
ejpam-5354	140	20	,	,	PUNCT
ejpam-5354	140	21	even	even	ADV
ejpam-5354	140	22	when	when	SCONJ
ejpam-5354	140	23	the	the	DET
ejpam-5354	140	24	dataset	dataset	NOUN
ejpam-5354	140	25	is	be	AUX
ejpam-5354	140	26	incomplete	incomplete	ADJ
ejpam-5354	140	27	[	[	X
ejpam-5354	140	28	19	19	NUM
ejpam-5354	140	29	,	,	PUNCT
ejpam-5354	140	30	27	27	NUM
ejpam-5354	140	31	,	,	PUNCT
ejpam-5354	140	32	32	32	NUM
ejpam-5354	140	33	]	]	PUNCT
ejpam-5354	140	34	.	.	PUNCT
ejpam-5354	141	1	in	in	ADP
ejpam-5354	141	2	this	this	DET
ejpam-5354	141	3	section	section	NOUN
ejpam-5354	141	4	,	,	PUNCT
ejpam-5354	141	5	the	the	DET
ejpam-5354	141	6	performance	performance	NOUN
ejpam-5354	141	7	of	of	ADP
ejpam-5354	141	8	the	the	DET
ejpam-5354	141	9	hs	hs	PROPN
ejpam-5354	141	10	method	method	NOUN
ejpam-5354	141	11	is	be	AUX
ejpam-5354	141	12	evaluated	evaluate	VERB
ejpam-5354	141	13	in	in	ADP
ejpam-5354	141	14	comparison	comparison	NOUN
ejpam-5354	141	15	with	with	ADP
ejpam-5354	141	16	the	the	DET
ejpam-5354	141	17	new	new	ADJ
ejpam-5354	141	18	method	method	NOUN
ejpam-5354	141	19	for	for	ADP
ejpam-5354	141	20	training	train	VERB
ejpam-5354	141	21	neural	neural	ADJ
ejpam-5354	141	22	networks	network	NOUN
ejpam-5354	141	23	.	.	PUNCT
ejpam-5354	142	1	the	the	DET
ejpam-5354	142	2	algorithms	algorithm	NOUN
ejpam-5354	142	3	were	be	AUX
ejpam-5354	142	4	implemented	implement	VERB
ejpam-5354	142	5	using	use	VERB
ejpam-5354	142	6	matlab	matlab	PROPN
ejpam-5354	142	7	(	(	PUNCT
ejpam-5354	142	8	2013a	2013a	NUM
ejpam-5354	142	9	)	)	PUNCT
ejpam-5354	142	10	and	and	CCONJ
ejpam-5354	142	11	the	the	DET
ejpam-5354	142	12	matlab	matlab	PROPN
ejpam-5354	142	13	neural	neural	ADJ
ejpam-5354	142	14	network	network	PROPN
ejpam-5354	142	15	toolbox	toolbox	PROPN
ejpam-5354	142	16	version	version	PROPN
ejpam-5354	142	17	8.1	8.1	NUM
ejpam-5354	142	18	for	for	ADP
ejpam-5354	142	19	conjugate	conjugate	ADJ
ejpam-5354	142	20	gradient	gradient	ADJ
ejpam-5354	142	21	optimization	optimization	NOUN
ejpam-5354	142	22	.	.	PUNCT
ejpam-5354	143	1	the	the	DET
ejpam-5354	143	2	network	network	NOUN
ejpam-5354	143	3	is	be	AUX
ejpam-5354	143	4	trained	train	VERB
ejpam-5354	143	5	until	until	SCONJ
ejpam-5354	143	6	the	the	DET
ejpam-5354	143	7	mean	mean	ADJ
ejpam-5354	143	8	squared	square	VERB
ejpam-5354	143	9	errors	error	NOUN
ejpam-5354	143	10	fall	fall	VERB
ejpam-5354	143	11	below	below	ADP
ejpam-5354	143	12	the	the	DET
ejpam-5354	143	13	specified	specify	VERB
ejpam-5354	143	14	error	error	NOUN
ejpam-5354	143	15	target	target	NOUN
ejpam-5354	143	16	,	,	PUNCT
ejpam-5354	143	17	indicating	indicate	VERB
ejpam-5354	143	18	a	a	DET
ejpam-5354	143	19	decreasing	decrease	VERB
ejpam-5354	143	20	trend	trend	NOUN
ejpam-5354	143	21	in	in	ADP
ejpam-5354	143	22	the	the	DET
ejpam-5354	143	23	error	error	NOUN
ejpam-5354	143	24	function	function	NOUN
ejpam-5354	143	25	.	.	PUNCT
ejpam-5354	144	1	for	for	ADP
ejpam-5354	144	2	consistency	consistency	NOUN
ejpam-5354	144	3	,	,	PUNCT
ejpam-5354	144	4	the	the	DET
ejpam-5354	144	5	same	same	ADJ
ejpam-5354	144	6	initial	initial	ADJ
ejpam-5354	144	7	weights	weight	NOUN
ejpam-5354	144	8	were	be	AUX
ejpam-5354	144	9	used	use	VERB
ejpam-5354	144	10	across	across	ADP
ejpam-5354	144	11	all	all	DET
ejpam-5354	144	12	algorithms	algorithm	NOUN
ejpam-5354	144	13	,	,	PUNCT
ejpam-5354	144	14	which	which	PRON
ejpam-5354	144	15	were	be	AUX
ejpam-5354	144	16	randomly	randomly	ADV
ejpam-5354	144	17	initialized	initialize	VERB
ejpam-5354	144	18	within	within	ADP
ejpam-5354	144	19	the	the	DET
ejpam-5354	144	20	range	range	NOUN
ejpam-5354	144	21	(	(	PUNCT
ejpam-5354	144	22	0	0	NUM
ejpam-5354	144	23	,	,	PUNCT
ejpam-5354	144	24	1	1	NUM
ejpam-5354	144	25	)	)	PUNCT
ejpam-5354	144	26	.	.	PUNCT
ejpam-5354	145	1	the	the	DET
ejpam-5354	145	2	problems	problem	NOUN
ejpam-5354	145	3	addressed	address	VERB
ejpam-5354	145	4	in	in	ADP
ejpam-5354	145	5	this	this	DET
ejpam-5354	145	6	evaluation	evaluation	NOUN
ejpam-5354	145	7	are	be	AUX
ejpam-5354	145	8	:	:	PUNCT
ejpam-5354	145	9	(	(	PUNCT
ejpam-5354	145	10	i	i	NOUN
ejpam-5354	145	11	)	)	PUNCT
ejpam-5354	145	12	input	input	NOUN
ejpam-5354	145	13	:	:	PUNCT
ejpam-5354	145	14	p	p	X
ejpam-5354	146	1	=	=	X
ejpam-5354	147	1	[	[	X
ejpam-5354	147	2	−1,−1	−1,−1	NOUN
ejpam-5354	147	3	,	,	PUNCT
ejpam-5354	147	4	2	2	NUM
ejpam-5354	147	5	,	,	PUNCT
ejpam-5354	147	6	2	2	NUM
ejpam-5354	147	7	;	;	PUNCT
ejpam-5354	147	8	0	0	NUM
ejpam-5354	147	9	,	,	PUNCT
ejpam-5354	147	10	5	5	NUM
ejpam-5354	147	11	,	,	PUNCT
ejpam-5354	147	12	0	0	NUM
ejpam-5354	147	13	,	,	PUNCT
ejpam-5354	147	14	5	5	NUM
ejpam-5354	147	15	]	]	PUNCT
ejpam-5354	147	16	and	and	CCONJ
ejpam-5354	147	17	target	target	NOUN
ejpam-5354	147	18	:	:	PUNCT
ejpam-5354	147	19	t	t	NOUN
ejpam-5354	147	20	=	=	PUNCT
ejpam-5354	148	1	[	[	X
ejpam-5354	148	2	−1,−1	−1,−1	NOUN
ejpam-5354	148	3	,	,	PUNCT
ejpam-5354	148	4	1	1	NUM
ejpam-5354	148	5	,	,	PUNCT
ejpam-5354	148	6	1	1	NUM
ejpam-5354	148	7	]	]	PUNCT
ejpam-5354	148	8	.	.	PUNCT
ejpam-5354	149	1	the	the	DET
ejpam-5354	149	2	target	target	NOUN
ejpam-5354	149	3	error	error	NOUN
ejpam-5354	149	4	was	be	AUX
ejpam-5354	149	5	set	set	VERB
ejpam-5354	149	6	to	to	ADP
ejpam-5354	149	7	1×	1×	PROPN
ejpam-5354	149	8	10−10	10−10	NUM
ejpam-5354	149	9	and	and	CCONJ
ejpam-5354	149	10	the	the	DET
ejpam-5354	149	11	maximum	maximum	ADJ
ejpam-5354	149	12	number	number	NOUN
ejpam-5354	149	13	of	of	ADP
ejpam-5354	149	14	epochs	epoch	NOUN
ejpam-5354	149	15	was	be	AUX
ejpam-5354	149	16	set	set	VERB
ejpam-5354	149	17	to	to	ADP
ejpam-5354	149	18	1000	1000	NUM
ejpam-5354	149	19	.	.	PUNCT
ejpam-5354	150	1	(	(	PUNCT
ejpam-5354	150	2	ii	ii	NOUN
ejpam-5354	150	3	)	)	PUNCT
ejpam-5354	150	4	input	input	NOUN
ejpam-5354	150	5	:	:	PUNCT
ejpam-5354	150	6	continuous	continuous	ADJ
ejpam-5354	150	7	trigonometric	trigonometric	ADJ
ejpam-5354	150	8	function	function	NOUN
ejpam-5354	150	9	f(x	f(x	PROPN
ejpam-5354	150	10	)	)	PUNCT
ejpam-5354	151	1	=	=	SYM
ejpam-5354	151	2	sin(x	sin(x	PROPN
ejpam-5354	151	3	)	)	PUNCT
ejpam-5354	152	1	+	+	NUM
ejpam-5354	153	1	cos(2x	cos(2x	VERB
ejpam-5354	153	2	)	)	PUNCT
ejpam-5354	153	3	,	,	PUNCT
ejpam-5354	153	4	where	where	SCONJ
ejpam-5354	153	5	x	x	X
ejpam-5354	153	6	∈	∈	PROPN
ejpam-5354	154	1	[	[	X
ejpam-5354	154	2	0	0	NUM
ejpam-5354	154	3	,	,	PUNCT
ejpam-5354	154	4	π	π	NOUN
ejpam-5354	154	5	]	]	X
ejpam-5354	154	6	.	.	PUNCT
ejpam-5354	155	1	the	the	DET
ejpam-5354	155	2	target	target	NOUN
ejpam-5354	155	3	error	error	NOUN
ejpam-5354	155	4	was	be	AUX
ejpam-5354	155	5	set	set	VERB
ejpam-5354	155	6	to	to	ADP
ejpam-5354	155	7	1	1	NUM
ejpam-5354	155	8	×	×	NOUN
ejpam-5354	155	9	10−10	10−10	NUM
ejpam-5354	155	10	and	and	CCONJ
ejpam-5354	155	11	the	the	DET
ejpam-5354	155	12	maximum	maximum	ADJ
ejpam-5354	155	13	number	number	NOUN
ejpam-5354	155	14	of	of	ADP
ejpam-5354	155	15	epochs	epoch	NOUN
ejpam-5354	155	16	was	be	AUX
ejpam-5354	155	17	set	set	VERB
ejpam-5354	155	18	to	to	ADP
ejpam-5354	155	19	1000	1000	NUM
ejpam-5354	155	20	.	.	PUNCT
ejpam-5354	156	1	alaa	alaa	PROPN
ejpam-5354	156	2	l.	l.	PROPN
ejpam-5354	156	3	ibrahim	ibrahim	PROPN
ejpam-5354	156	4	et	et	PROPN
ejpam-5354	156	5	al	al	PROPN
ejpam-5354	156	6	.	.	PUNCT
ejpam-5354	156	7	/	/	SYM
ejpam-5354	156	8	eur	eur	PROPN
ejpam-5354	156	9	.	.	PUNCT
ejpam-5354	157	1	j.	j.	PROPN
ejpam-5354	157	2	pure	pure	PROPN
ejpam-5354	157	3	appl	appl	PROPN
ejpam-5354	157	4	.	.	PROPN
ejpam-5354	157	5	math	math	PROPN
ejpam-5354	157	6	,	,	PUNCT
ejpam-5354	157	7	17	17	NUM
ejpam-5354	157	8	(	(	PUNCT
ejpam-5354	157	9	4	4	NUM
ejpam-5354	157	10	)	)	PUNCT
ejpam-5354	157	11	(	(	PUNCT
ejpam-5354	157	12	2024	2024	NUM
ejpam-5354	157	13	)	)	PUNCT
ejpam-5354	157	14	,	,	PUNCT
ejpam-5354	157	15	2692	2692	NUM
ejpam-5354	157	16	-	-	SYM
ejpam-5354	157	17	2705	2705	NUM
ejpam-5354	157	18	2700	2700	NUM
ejpam-5354	157	19	the	the	DET
ejpam-5354	157	20	results	result	NOUN
ejpam-5354	157	21	of	of	ADP
ejpam-5354	157	22	the	the	DET
ejpam-5354	157	23	training	training	NOUN
ejpam-5354	157	24	methods	method	NOUN
ejpam-5354	157	25	are	be	AUX
ejpam-5354	157	26	presented	present	VERB
ejpam-5354	157	27	in	in	ADP
ejpam-5354	157	28	tables	table	NOUN
ejpam-5354	157	29	3	3	NUM
ejpam-5354	157	30	and	and	CCONJ
ejpam-5354	157	31	figures	figure	NOUN
ejpam-5354	157	32	1	1	NUM
ejpam-5354	157	33	,	,	PUNCT
ejpam-5354	157	34	2	2	NUM
ejpam-5354	157	35	,	,	PUNCT
ejpam-5354	157	36	3	3	NUM
ejpam-5354	157	37	,	,	PUNCT
ejpam-5354	157	38	and	and	CCONJ
ejpam-5354	157	39	4	4	NUM
ejpam-5354	157	40	below	below	ADV
ejpam-5354	157	41	.	.	PUNCT
ejpam-5354	158	1	table	table	NOUN
ejpam-5354	158	2	3	3	NUM
ejpam-5354	158	3	:	:	PUNCT
ejpam-5354	158	4	comparing	compare	VERB
ejpam-5354	158	5	the	the	DET
ejpam-5354	158	6	performance	performance	NOUN
ejpam-5354	158	7	of	of	ADP
ejpam-5354	158	8	new	new	ADJ
ejpam-5354	158	9	method	method	NOUN
ejpam-5354	158	10	with	with	ADP
ejpam-5354	158	11	standard	standard	ADJ
ejpam-5354	158	12	method	method	NOUN
ejpam-5354	158	13	for	for	ADP
ejpam-5354	158	14	training	train	VERB
ejpam-5354	158	15	neural	neural	ADJ
ejpam-5354	158	16	network	network	NOUN
ejpam-5354	158	17	methods	method	NOUN
ejpam-5354	159	1	no	no	INTJ
ejpam-5354	159	2	.	.	PUNCT
ejpam-5354	160	1	running	run	VERB
ejpam-5354	160	2	epochs	epoch	NOUN
ejpam-5354	160	3	cputime(s	cputime(s	NOUN
ejpam-5354	160	4	)	)	PUNCT
ejpam-5354	160	5	gradient	gradient	ADJ
ejpam-5354	160	6	problem1	problem1	NOUN
ejpam-5354	160	7	standard	standard	NOUN
ejpam-5354	160	8	1	1	NUM
ejpam-5354	160	9	1000	1000	NUM
ejpam-5354	160	10	0:00:04	0:00:04	NUM
ejpam-5354	160	11	0.000376	0.000376	NUM
ejpam-5354	160	12	2	2	NUM
ejpam-5354	160	13	34	34	NUM
ejpam-5354	160	14	0:00:00	0:00:00	NUM
ejpam-5354	160	15	0.000239	0.000239	NUM
ejpam-5354	160	16	3	3	NUM
ejpam-5354	160	17	41	41	NUM
ejpam-5354	161	1	0:00:00	0:00:00	NUM
ejpam-5354	161	2	0.000173	0.000173	NUM
ejpam-5354	161	3	4	4	NUM
ejpam-5354	161	4	21	21	NUM
ejpam-5354	161	5	0:00:00	0:00:00	NUM
ejpam-5354	161	6	0.000209	0.000209	NUM
ejpam-5354	161	7	5	5	NUM
ejpam-5354	161	8	38	38	NUM
ejpam-5354	161	9	0:00:00	0:00:00	NUM
ejpam-5354	161	10	0.000157	0.000157	NUM
ejpam-5354	161	11	new	new	ADJ
ejpam-5354	161	12	cg	cg	NOUN
ejpam-5354	161	13	1	1	NUM
ejpam-5354	161	14	6	6	NUM
ejpam-5354	161	15	0:00:00	0:00:00	NUM
ejpam-5354	161	16	0.153	0.153	NUM
ejpam-5354	161	17	2	2	NUM
ejpam-5354	161	18	4	4	NUM
ejpam-5354	161	19	0:00:00	0:00:00	NUM
ejpam-5354	161	20	0.121	0.121	NUM
ejpam-5354	161	21	3	3	NUM
ejpam-5354	161	22	6	6	NUM
ejpam-5354	161	23	0:00:00	0:00:00	NUM
ejpam-5354	161	24	0.0649	0.0649	NUM
ejpam-5354	161	25	4	4	NUM
ejpam-5354	161	26	6	6	NUM
ejpam-5354	161	27	0:00:00	0:00:00	NUM
ejpam-5354	161	28	0.0499	0.0499	NUM
ejpam-5354	161	29	5	5	NUM
ejpam-5354	161	30	4	4	NUM
ejpam-5354	161	31	0:00:00	0:00:00	NUM
ejpam-5354	161	32	0.382	0.382	NUM
ejpam-5354	161	33	problem	problem	NOUN
ejpam-5354	161	34	2	2	NUM
ejpam-5354	161	35	standard	standard	NOUN
ejpam-5354	161	36	1	1	NUM
ejpam-5354	161	37	296	296	NUM
ejpam-5354	161	38	0:00:01	0:00:01	NUM
ejpam-5354	161	39	0.000244	0.000244	NUM
ejpam-5354	161	40	2	2	NUM
ejpam-5354	161	41	1000	1000	NUM
ejpam-5354	161	42	0:00:02	0:00:02	NUM
ejpam-5354	162	1	0.00334	0.00334	NUM
ejpam-5354	162	2	3	3	NUM
ejpam-5354	162	3	69	69	NUM
ejpam-5354	162	4	0:00:00	0:00:00	NUM
ejpam-5354	162	5	0.000186	0.000186	NUM
ejpam-5354	162	6	4	4	NUM
ejpam-5354	162	7	1000	1000	NUM
ejpam-5354	162	8	0:00:02	0:00:02	NUM
ejpam-5354	162	9	0.00219	0.00219	NUM
ejpam-5354	162	10	5	5	NUM
ejpam-5354	162	11	1000	1000	NUM
ejpam-5354	162	12	0:00:02	0:00:02	NUM
ejpam-5354	163	1	0.00374	0.00374	NUM
ejpam-5354	163	2	new	new	ADJ
ejpam-5354	163	3	cg	cg	NOUN
ejpam-5354	163	4	1	1	NUM
ejpam-5354	163	5	2	2	NUM
ejpam-5354	163	6	0:00:00	0:00:00	NUM
ejpam-5354	163	7	0.89	0.89	NUM
ejpam-5354	163	8	2	2	NUM
ejpam-5354	163	9	2	2	NUM
ejpam-5354	163	10	0:00:00	0:00:00	NUM
ejpam-5354	163	11	0.975	0.975	NUM
ejpam-5354	163	12	3	3	NUM
ejpam-5354	163	13	2	2	NUM
ejpam-5354	163	14	0:00:00	0:00:00	NUM
ejpam-5354	163	15	0.525	0.525	NUM
ejpam-5354	163	16	4	4	NUM
ejpam-5354	163	17	2	2	NUM
ejpam-5354	163	18	0:00:00	0:00:00	NUM
ejpam-5354	163	19	0.697	0.697	NUM
ejpam-5354	163	20	5	5	NUM
ejpam-5354	163	21	3	3	NUM
ejpam-5354	163	22	0:00:02	0:00:02	NUM
ejpam-5354	163	23	0.332	0.332	NUM
ejpam-5354	163	24	table	table	NOUN
ejpam-5354	163	25	3	3	NUM
ejpam-5354	163	26	compares	compare	VERB
ejpam-5354	163	27	the	the	DET
ejpam-5354	163	28	performance	performance	NOUN
ejpam-5354	163	29	of	of	ADP
ejpam-5354	163	30	the	the	DET
ejpam-5354	163	31	new	new	ADJ
ejpam-5354	163	32	(	(	PUNCT
ejpam-5354	163	33	cg	cg	NOUN
ejpam-5354	163	34	)	)	PUNCT
ejpam-5354	163	35	method	method	NOUN
ejpam-5354	163	36	with	with	ADP
ejpam-5354	163	37	the	the	DET
ejpam-5354	163	38	standard	standard	ADJ
ejpam-5354	163	39	method	method	NOUN
ejpam-5354	163	40	in	in	ADP
ejpam-5354	163	41	training	train	VERB
ejpam-5354	163	42	neural	neural	ADJ
ejpam-5354	163	43	networks	network	NOUN
ejpam-5354	163	44	.	.	PUNCT
ejpam-5354	164	1	the	the	DET
ejpam-5354	164	2	new	new	ADJ
ejpam-5354	164	3	cg	cg	NOUN
ejpam-5354	164	4	method	method	NOUN
ejpam-5354	164	5	significantly	significantly	ADV
ejpam-5354	164	6	reduces	reduce	VERB
ejpam-5354	164	7	the	the	DET
ejpam-5354	164	8	number	number	NOUN
ejpam-5354	164	9	of	of	ADP
ejpam-5354	164	10	epochs	epoch	NOUN
ejpam-5354	164	11	needed	need	VERB
ejpam-5354	164	12	to	to	PART
ejpam-5354	164	13	converge	converge	VERB
ejpam-5354	164	14	,	,	PUNCT
ejpam-5354	164	15	often	often	ADV
ejpam-5354	164	16	achieving	achieve	VERB
ejpam-5354	164	17	results	result	NOUN
ejpam-5354	164	18	in	in	ADP
ejpam-5354	164	19	just	just	ADV
ejpam-5354	164	20	2	2	NUM
ejpam-5354	164	21	to	to	PART
ejpam-5354	164	22	6	6	NUM
ejpam-5354	164	23	epochs	epoch	NOUN
ejpam-5354	164	24	compared	compare	VERB
ejpam-5354	164	25	to	to	ADP
ejpam-5354	164	26	the	the	DET
ejpam-5354	164	27	standard	standard	ADJ
ejpam-5354	164	28	method	method	NOUN
ejpam-5354	164	29	’s	’s	PART
ejpam-5354	164	30	1000	1000	NUM
ejpam-5354	164	31	epochs	epoch	NOUN
ejpam-5354	164	32	in	in	ADP
ejpam-5354	164	33	several	several	ADJ
ejpam-5354	164	34	cases	case	NOUN
ejpam-5354	164	35	.	.	PUNCT
ejpam-5354	165	1	both	both	DET
ejpam-5354	165	2	methods	method	NOUN
ejpam-5354	165	3	show	show	VERB
ejpam-5354	165	4	minimal	minimal	ADJ
ejpam-5354	165	5	cpu	cpu	NOUN
ejpam-5354	165	6	time	time	NOUN
ejpam-5354	165	7	per	per	ADP
ejpam-5354	165	8	epoch	epoch	NOUN
ejpam-5354	165	9	,	,	PUNCT
ejpam-5354	165	10	but	but	CCONJ
ejpam-5354	165	11	the	the	DET
ejpam-5354	165	12	new	new	ADJ
ejpam-5354	165	13	cg	cg	NOUN
ejpam-5354	165	14	method	method	NOUN
ejpam-5354	165	15	generally	generally	ADV
ejpam-5354	165	16	achieves	achieve	VERB
ejpam-5354	165	17	higher	high	ADJ
ejpam-5354	165	18	gradient	gradient	ADJ
ejpam-5354	165	19	values	value	NOUN
ejpam-5354	165	20	in	in	ADP
ejpam-5354	165	21	fewer	few	ADJ
ejpam-5354	165	22	epochs	epoch	NOUN
ejpam-5354	165	23	,	,	PUNCT
ejpam-5354	165	24	indicating	indicate	VERB
ejpam-5354	165	25	more	more	ADV
ejpam-5354	165	26	efficient	efficient	ADJ
ejpam-5354	165	27	optimization	optimization	NOUN
ejpam-5354	165	28	.	.	PUNCT
ejpam-5354	166	1	overall	overall	ADV
ejpam-5354	166	2	,	,	PUNCT
ejpam-5354	166	3	the	the	DET
ejpam-5354	166	4	new	new	ADJ
ejpam-5354	166	5	cg	cg	NOUN
ejpam-5354	166	6	method	method	NOUN
ejpam-5354	166	7	outperforms	outperform	VERB
ejpam-5354	166	8	the	the	DET
ejpam-5354	166	9	standard	standard	ADJ
ejpam-5354	166	10	method	method	NOUN
ejpam-5354	166	11	in	in	ADP
ejpam-5354	166	12	terms	term	NOUN
ejpam-5354	166	13	of	of	ADP
ejpam-5354	166	14	faster	fast	ADJ
ejpam-5354	166	15	convergence	convergence	NOUN
ejpam-5354	166	16	and	and	CCONJ
ejpam-5354	166	17	computational	computational	ADJ
ejpam-5354	166	18	efficiency	efficiency	NOUN
ejpam-5354	166	19	.	.	PUNCT
ejpam-5354	167	1	alaa	alaa	PROPN
ejpam-5354	167	2	l.	l.	PROPN
ejpam-5354	167	3	ibrahim	ibrahim	PROPN
ejpam-5354	167	4	et	et	PROPN
ejpam-5354	167	5	al	al	PROPN
ejpam-5354	167	6	.	.	PUNCT
ejpam-5354	167	7	/	/	SYM
ejpam-5354	167	8	eur	eur	PROPN
ejpam-5354	167	9	.	.	PUNCT
ejpam-5354	168	1	j.	j.	PROPN
ejpam-5354	168	2	pure	pure	PROPN
ejpam-5354	168	3	appl	appl	PROPN
ejpam-5354	168	4	.	.	PROPN
ejpam-5354	168	5	math	math	PROPN
ejpam-5354	168	6	,	,	PUNCT
ejpam-5354	168	7	17	17	NUM
ejpam-5354	168	8	(	(	PUNCT
ejpam-5354	168	9	4	4	NUM
ejpam-5354	168	10	)	)	PUNCT
ejpam-5354	168	11	(	(	PUNCT
ejpam-5354	168	12	2024	2024	NUM
ejpam-5354	168	13	)	)	PUNCT
ejpam-5354	168	14	,	,	PUNCT
ejpam-5354	168	15	2692	2692	NUM
ejpam-5354	168	16	-	-	SYM
ejpam-5354	168	17	2705	2705	NUM
ejpam-5354	168	18	2701	2701	NUM
ejpam-5354	168	19	0	0	NUM
ejpam-5354	169	1	200	200	NUM
ejpam-5354	169	2	400	400	NUM
ejpam-5354	169	3	600	600	NUM
ejpam-5354	169	4	800	800	NUM
ejpam-5354	169	5	1000	1000	NUM
ejpam-5354	169	6	10	10	NUM
ejpam-5354	169	7	−15	−15	NOUN
ejpam-5354	169	8	10	10	NUM
ejpam-5354	169	9	−10	−10	NUM
ejpam-5354	169	10	10	10	NUM
ejpam-5354	169	11	−5	−5	ADV
ejpam-5354	169	12	10	10	NUM
ejpam-5354	169	13	0	0	NUM
ejpam-5354	169	14	best	good	ADJ
ejpam-5354	169	15	training	training	NOUN
ejpam-5354	169	16	performance	performance	NOUN
ejpam-5354	169	17	is	be	AUX
ejpam-5354	169	18	8.735e−06	8.735e−06	NUM
ejpam-5354	169	19	at	at	ADP
ejpam-5354	169	20	epoch	epoch	NOUN
ejpam-5354	169	21	1000	1000	NUM
ejpam-5354	169	22	m	m	NOUN
ejpam-5354	169	23	s	s	PART
ejpam-5354	169	24	e	e	NOUN
ejpam-5354	169	25	1000	1000	NUM
ejpam-5354	169	26	epochs	epoch	NOUN
ejpam-5354	169	27	train	train	VERB
ejpam-5354	169	28	best	good	ADJ
ejpam-5354	169	29	goal	goal	NOUN
ejpam-5354	169	30	0	0	NUM
ejpam-5354	169	31	1	1	NUM
ejpam-5354	169	32	2	2	NUM
ejpam-5354	169	33	3	3	NUM
ejpam-5354	169	34	4	4	NUM
ejpam-5354	169	35	5	5	NUM
ejpam-5354	169	36	6	6	NUM
ejpam-5354	169	37	10	10	NUM
ejpam-5354	169	38	−15	−15	NOUN
ejpam-5354	169	39	10	10	NUM
ejpam-5354	169	40	−10	−10	NUM
ejpam-5354	169	41	10	10	NUM
ejpam-5354	169	42	−5	−5	ADV
ejpam-5354	169	43	10	10	NUM
ejpam-5354	169	44	0	0	NUM
ejpam-5354	169	45	best	good	ADJ
ejpam-5354	169	46	training	training	NOUN
ejpam-5354	169	47	performance	performance	NOUN
ejpam-5354	169	48	is	be	AUX
ejpam-5354	169	49	0.012145	0.012145	NUM
ejpam-5354	169	50	at	at	ADP
ejpam-5354	169	51	epoch	epoch	NOUN
ejpam-5354	169	52	5	5	NUM
ejpam-5354	170	1	m	m	NOUN
ejpam-5354	170	2	s	s	NOUN
ejpam-5354	170	3	e	e	NOUN
ejpam-5354	170	4	6	6	NUM
ejpam-5354	170	5	epochs	epoch	NOUN
ejpam-5354	170	6	train	train	VERB
ejpam-5354	170	7	best	good	ADJ
ejpam-5354	170	8	goal	goal	NOUN
ejpam-5354	170	9	figure	figure	NOUN
ejpam-5354	170	10	1	1	NUM
ejpam-5354	170	11	:	:	PUNCT
ejpam-5354	170	12	validation	validation	NOUN
ejpam-5354	170	13	performance	performance	NOUN
ejpam-5354	170	14	hs	hs	PROPN
ejpam-5354	170	15	method	method	NOUN
ejpam-5354	170	16	(	(	PUNCT
ejpam-5354	170	17	left	left	ADJ
ejpam-5354	170	18	)	)	PUNCT
ejpam-5354	170	19	and	and	CCONJ
ejpam-5354	170	20	new	new	ADJ
ejpam-5354	170	21	method	method	NOUN
ejpam-5354	170	22	(	(	PUNCT
ejpam-5354	170	23	right	right	ADJ
ejpam-5354	170	24	)	)	PUNCT
ejpam-5354	170	25	,	,	PUNCT
ejpam-5354	170	26	of	of	ADP
ejpam-5354	170	27	problem	problem	NOUN
ejpam-5354	170	28	1	1	NUM
ejpam-5354	170	29	10	10	NUM
ejpam-5354	170	30	−5	−5	ADV
ejpam-5354	170	31	10	10	NUM
ejpam-5354	170	32	0	0	NUM
ejpam-5354	170	33	10	10	NUM
ejpam-5354	170	34	5	5	NUM
ejpam-5354	170	35	g	g	NOUN
ejpam-5354	170	36	ra	ra	PROPN
ejpam-5354	171	1	d	d	NOUN
ejpam-5354	171	2	ie	ie	X
ejpam-5354	171	3	n	n	PROPN
ejpam-5354	171	4	t	t	NOUN
ejpam-5354	171	5	gradient	gradient	NOUN
ejpam-5354	171	6	=	=	SYM
ejpam-5354	171	7	0.00037554	0.00037554	NUM
ejpam-5354	171	8	,	,	PUNCT
ejpam-5354	171	9	at	at	ADP
ejpam-5354	171	10	epoch	epoch	NOUN
ejpam-5354	171	11	1000	1000	NUM
ejpam-5354	171	12	−1	−1	NOUN
ejpam-5354	171	13	0	0	NUM
ejpam-5354	171	14	1	1	NUM
ejpam-5354	171	15	v	v	ADP
ejpam-5354	171	16	a	a	DET
ejpam-5354	171	17	l	l	NOUN
ejpam-5354	171	18	fa	fa	X
ejpam-5354	171	19	il	il	PROPN
ejpam-5354	171	20	validation	validation	NOUN
ejpam-5354	171	21	checks	check	NOUN
ejpam-5354	171	22	=	=	SYM
ejpam-5354	171	23	0	0	NUM
ejpam-5354	171	24	,	,	PUNCT
ejpam-5354	171	25	at	at	ADP
ejpam-5354	171	26	epoch	epoch	NOUN
ejpam-5354	171	27	1000	1000	NUM
ejpam-5354	171	28	0	0	NUM
ejpam-5354	171	29	200	200	NUM
ejpam-5354	171	30	400	400	NUM
ejpam-5354	171	31	600	600	NUM
ejpam-5354	171	32	800	800	NUM
ejpam-5354	171	33	1000	1000	NUM
ejpam-5354	171	34	10	10	NUM
ejpam-5354	171	35	−5	−5	ADV
ejpam-5354	171	36	10	10	NUM
ejpam-5354	171	37	0	0	NUM
ejpam-5354	171	38	10	10	NUM
ejpam-5354	171	39	5	5	NUM
ejpam-5354	171	40	a	a	DET
ejpam-5354	171	41	1000	1000	NUM
ejpam-5354	171	42	epochs	epoch	NOUN
ejpam-5354	171	43	step	step	NOUN
ejpam-5354	171	44	size	size	NOUN
ejpam-5354	171	45	=	=	NOUN
ejpam-5354	171	46	0.0001	0.0001	NUM
ejpam-5354	171	47	,	,	PUNCT
ejpam-5354	171	48	at	at	ADP
ejpam-5354	171	49	epoch	epoch	NOUN
ejpam-5354	171	50	1000	1000	NUM
ejpam-5354	171	51	10	10	NUM
ejpam-5354	171	52	0	0	NUM
ejpam-5354	172	1	g	g	NOUN
ejpam-5354	172	2	ra	ra	PROPN
ejpam-5354	173	1	d	d	NOUN
ejpam-5354	173	2	ie	ie	X
ejpam-5354	173	3	n	n	PROPN
ejpam-5354	173	4	t	t	NOUN
ejpam-5354	173	5	gradient	gradient	NOUN
ejpam-5354	173	6	=	=	SYM
ejpam-5354	173	7	0.15284	0.15284	NUM
ejpam-5354	173	8	,	,	PUNCT
ejpam-5354	173	9	at	at	ADP
ejpam-5354	173	10	epoch	epoch	NOUN
ejpam-5354	173	11	6	6	NUM
ejpam-5354	173	12	−1	−1	NOUN
ejpam-5354	173	13	0	0	NUM
ejpam-5354	173	14	1	1	NUM
ejpam-5354	173	15	v	v	ADP
ejpam-5354	173	16	a	a	DET
ejpam-5354	173	17	l	l	NOUN
ejpam-5354	173	18	fa	fa	X
ejpam-5354	173	19	il	il	PROPN
ejpam-5354	173	20	validation	validation	NOUN
ejpam-5354	173	21	checks	check	NOUN
ejpam-5354	173	22	=	=	SYM
ejpam-5354	173	23	0	0	NUM
ejpam-5354	173	24	,	,	PUNCT
ejpam-5354	173	25	at	at	ADP
ejpam-5354	173	26	epoch	epoch	NOUN
ejpam-5354	173	27	6	6	NUM
ejpam-5354	173	28	0	0	NUM
ejpam-5354	173	29	1	1	NUM
ejpam-5354	173	30	2	2	NUM
ejpam-5354	173	31	3	3	NUM
ejpam-5354	173	32	4	4	NUM
ejpam-5354	173	33	5	5	NUM
ejpam-5354	173	34	6	6	NUM
ejpam-5354	173	35	10	10	NUM
ejpam-5354	173	36	−2	−2	NOUN
ejpam-5354	173	37	10	10	NUM
ejpam-5354	173	38	0	0	NUM
ejpam-5354	173	39	10	10	NUM
ejpam-5354	173	40	2	2	NUM
ejpam-5354	173	41	a	a	DET
ejpam-5354	173	42	6	6	NUM
ejpam-5354	173	43	epochs	epoch	NOUN
ejpam-5354	173	44	step	step	NOUN
ejpam-5354	173	45	size	size	NOUN
ejpam-5354	173	46	=	=	SYM
ejpam-5354	173	47	0	0	NUM
ejpam-5354	173	48	,	,	PUNCT
ejpam-5354	173	49	at	at	ADP
ejpam-5354	173	50	epoch	epoch	NOUN
ejpam-5354	173	51	6	6	NUM
ejpam-5354	173	52	figure	figure	NOUN
ejpam-5354	173	53	2	2	NUM
ejpam-5354	173	54	:	:	PUNCT
ejpam-5354	173	55	training	training	NOUN
ejpam-5354	173	56	performance	performance	NOUN
ejpam-5354	173	57	hs	hs	PROPN
ejpam-5354	173	58	method	method	NOUN
ejpam-5354	173	59	(	(	PUNCT
ejpam-5354	173	60	left	left	ADJ
ejpam-5354	173	61	)	)	PUNCT
ejpam-5354	173	62	and	and	CCONJ
ejpam-5354	173	63	new	new	ADJ
ejpam-5354	173	64	method	method	NOUN
ejpam-5354	173	65	(	(	PUNCT
ejpam-5354	173	66	right	right	ADJ
ejpam-5354	173	67	)	)	PUNCT
ejpam-5354	173	68	,	,	PUNCT
ejpam-5354	173	69	of	of	ADP
ejpam-5354	173	70	problem	problem	NOUN
ejpam-5354	173	71	1	1	X
ejpam-5354	173	72	.	.	PUNCT
ejpam-5354	174	1	alaa	alaa	PROPN
ejpam-5354	174	2	l.	l.	PROPN
ejpam-5354	174	3	ibrahim	ibrahim	PROPN
ejpam-5354	174	4	et	et	PROPN
ejpam-5354	174	5	al	al	PROPN
ejpam-5354	174	6	.	.	PUNCT
ejpam-5354	174	7	/	/	SYM
ejpam-5354	174	8	eur	eur	PROPN
ejpam-5354	174	9	.	.	PUNCT
ejpam-5354	175	1	j.	j.	PROPN
ejpam-5354	175	2	pure	pure	PROPN
ejpam-5354	175	3	appl	appl	PROPN
ejpam-5354	175	4	.	.	PROPN
ejpam-5354	175	5	math	math	PROPN
ejpam-5354	175	6	,	,	PUNCT
ejpam-5354	175	7	17	17	NUM
ejpam-5354	175	8	(	(	PUNCT
ejpam-5354	175	9	4	4	NUM
ejpam-5354	175	10	)	)	PUNCT
ejpam-5354	175	11	(	(	PUNCT
ejpam-5354	175	12	2024	2024	NUM
ejpam-5354	175	13	)	)	PUNCT
ejpam-5354	175	14	,	,	PUNCT
ejpam-5354	175	15	2692	2692	NUM
ejpam-5354	175	16	-	-	SYM
ejpam-5354	175	17	2705	2705	NUM
ejpam-5354	175	18	2702	2702	NUM
ejpam-5354	175	19	0	0	NUM
ejpam-5354	176	1	50	50	NUM
ejpam-5354	176	2	100	100	NUM
ejpam-5354	176	3	150	150	NUM
ejpam-5354	176	4	200	200	NUM
ejpam-5354	176	5	250	250	NUM
ejpam-5354	176	6	10	10	NUM
ejpam-5354	176	7	−15	−15	NOUN
ejpam-5354	176	8	10	10	NUM
ejpam-5354	176	9	−10	−10	NUM
ejpam-5354	176	10	10	10	NUM
ejpam-5354	176	11	−5	−5	ADV
ejpam-5354	176	12	10	10	NUM
ejpam-5354	176	13	0	0	NUM
ejpam-5354	176	14	best	good	ADJ
ejpam-5354	176	15	training	training	NOUN
ejpam-5354	176	16	performance	performance	NOUN
ejpam-5354	176	17	is	be	AUX
ejpam-5354	176	18	6.2779e−09	6.2779e−09	NUM
ejpam-5354	176	19	at	at	ADP
ejpam-5354	176	20	epoch	epoch	NOUN
ejpam-5354	176	21	295	295	NUM
ejpam-5354	176	22	m	m	NOUN
ejpam-5354	176	23	s	s	NOUN
ejpam-5354	176	24	e	e	NOUN
ejpam-5354	176	25	296	296	NUM
ejpam-5354	176	26	epochs	epoch	NOUN
ejpam-5354	176	27	train	train	VERB
ejpam-5354	176	28	best	good	ADJ
ejpam-5354	176	29	goal	goal	NOUN
ejpam-5354	176	30	0	0	NUM
ejpam-5354	176	31	0.5	0.5	NUM
ejpam-5354	176	32	1	1	NUM
ejpam-5354	176	33	1.5	1.5	NUM
ejpam-5354	176	34	2	2	NUM
ejpam-5354	176	35	10	10	NUM
ejpam-5354	176	36	−15	−15	NOUN
ejpam-5354	176	37	10	10	NUM
ejpam-5354	176	38	−10	−10	NUM
ejpam-5354	176	39	10	10	NUM
ejpam-5354	176	40	−5	−5	ADV
ejpam-5354	176	41	10	10	NUM
ejpam-5354	176	42	0	0	NUM
ejpam-5354	176	43	best	good	ADJ
ejpam-5354	176	44	training	training	NOUN
ejpam-5354	176	45	performance	performance	NOUN
ejpam-5354	176	46	is	be	AUX
ejpam-5354	176	47	0.33619	0.33619	NUM
ejpam-5354	176	48	at	at	ADP
ejpam-5354	176	49	epoch	epoch	NOUN
ejpam-5354	176	50	1	1	NUM
ejpam-5354	176	51	m	m	NOUN
ejpam-5354	176	52	s	s	NOUN
ejpam-5354	176	53	e	e	NOUN
ejpam-5354	176	54	2	2	NUM
ejpam-5354	176	55	epochs	epoch	NOUN
ejpam-5354	176	56	train	train	VERB
ejpam-5354	176	57	best	good	ADJ
ejpam-5354	176	58	goal	goal	NOUN
ejpam-5354	176	59	figure	figure	NOUN
ejpam-5354	176	60	3	3	NUM
ejpam-5354	176	61	:	:	PUNCT
ejpam-5354	176	62	validation	validation	NOUN
ejpam-5354	176	63	performance	performance	NOUN
ejpam-5354	176	64	hs	hs	PROPN
ejpam-5354	176	65	method	method	NOUN
ejpam-5354	176	66	(	(	PUNCT
ejpam-5354	176	67	left	left	ADJ
ejpam-5354	176	68	)	)	PUNCT
ejpam-5354	176	69	and	and	CCONJ
ejpam-5354	176	70	new	new	ADJ
ejpam-5354	176	71	method	method	NOUN
ejpam-5354	176	72	(	(	PUNCT
ejpam-5354	176	73	right	right	ADJ
ejpam-5354	176	74	)	)	PUNCT
ejpam-5354	176	75	,	,	PUNCT
ejpam-5354	176	76	of	of	ADP
ejpam-5354	176	77	problem	problem	NOUN
ejpam-5354	176	78	2	2	NUM
ejpam-5354	176	79	.	.	NOUN
ejpam-5354	176	80	10	10	NUM
ejpam-5354	176	81	−5	−5	NOUN
ejpam-5354	176	82	10	10	NUM
ejpam-5354	176	83	0	0	NUM
ejpam-5354	176	84	10	10	NUM
ejpam-5354	176	85	5	5	NUM
ejpam-5354	176	86	g	g	NOUN
ejpam-5354	176	87	ra	ra	PROPN
ejpam-5354	177	1	d	d	NOUN
ejpam-5354	177	2	ie	ie	X
ejpam-5354	177	3	n	n	PROPN
ejpam-5354	177	4	t	t	NOUN
ejpam-5354	177	5	gradient	gradient	NOUN
ejpam-5354	177	6	=	=	SYM
ejpam-5354	177	7	0.00024378	0.00024378	NUM
ejpam-5354	177	8	,	,	PUNCT
ejpam-5354	177	9	at	at	ADP
ejpam-5354	177	10	epoch	epoch	NOUN
ejpam-5354	177	11	296	296	NUM
ejpam-5354	177	12	−1	−1	NOUN
ejpam-5354	177	13	0	0	NUM
ejpam-5354	177	14	1	1	NUM
ejpam-5354	177	15	v	v	ADP
ejpam-5354	177	16	a	a	DET
ejpam-5354	177	17	l	l	NOUN
ejpam-5354	177	18	fa	fa	X
ejpam-5354	177	19	il	il	PROPN
ejpam-5354	177	20	validation	validation	NOUN
ejpam-5354	177	21	checks	check	NOUN
ejpam-5354	177	22	=	=	SYM
ejpam-5354	177	23	0	0	NUM
ejpam-5354	177	24	,	,	PUNCT
ejpam-5354	177	25	at	at	ADP
ejpam-5354	177	26	epoch	epoch	NOUN
ejpam-5354	177	27	296	296	NUM
ejpam-5354	177	28	0	0	NUM
ejpam-5354	177	29	50	50	NUM
ejpam-5354	177	30	100	100	NUM
ejpam-5354	177	31	150	150	NUM
ejpam-5354	177	32	200	200	NUM
ejpam-5354	177	33	250	250	NUM
ejpam-5354	177	34	10	10	NUM
ejpam-5354	177	35	−4	−4	NUM
ejpam-5354	177	36	10	10	NUM
ejpam-5354	177	37	−2	−2	NOUN
ejpam-5354	177	38	10	10	NUM
ejpam-5354	177	39	0	0	NUM
ejpam-5354	177	40	a	a	DET
ejpam-5354	177	41	296	296	NUM
ejpam-5354	177	42	epochs	epoch	NOUN
ejpam-5354	177	43	step	step	NOUN
ejpam-5354	177	44	size	size	NOUN
ejpam-5354	177	45	=	=	SYM
ejpam-5354	177	46	0	0	NUM
ejpam-5354	177	47	,	,	PUNCT
ejpam-5354	177	48	at	at	ADP
ejpam-5354	177	49	epoch	epoch	NOUN
ejpam-5354	177	50	296	296	NUM
ejpam-5354	177	51	10	10	NUM
ejpam-5354	177	52	0	0	NUM
ejpam-5354	178	1	g	g	NOUN
ejpam-5354	178	2	ra	ra	PROPN
ejpam-5354	179	1	d	d	NOUN
ejpam-5354	179	2	ie	ie	X
ejpam-5354	179	3	n	n	PROPN
ejpam-5354	179	4	t	t	NOUN
ejpam-5354	179	5	gradient	gradient	NOUN
ejpam-5354	179	6	=	=	SYM
ejpam-5354	179	7	0.89041	0.89041	NUM
ejpam-5354	179	8	,	,	PUNCT
ejpam-5354	179	9	at	at	ADP
ejpam-5354	179	10	epoch	epoch	NOUN
ejpam-5354	179	11	2	2	NUM
ejpam-5354	179	12	−1	−1	NOUN
ejpam-5354	179	13	0	0	NUM
ejpam-5354	179	14	1	1	NUM
ejpam-5354	179	15	v	v	ADP
ejpam-5354	179	16	a	a	DET
ejpam-5354	179	17	l	l	NOUN
ejpam-5354	179	18	fa	fa	X
ejpam-5354	179	19	il	il	PROPN
ejpam-5354	179	20	validation	validation	NOUN
ejpam-5354	179	21	checks	check	NOUN
ejpam-5354	179	22	=	=	SYM
ejpam-5354	179	23	0	0	NUM
ejpam-5354	179	24	,	,	PUNCT
ejpam-5354	179	25	at	at	ADP
ejpam-5354	179	26	epoch	epoch	NOUN
ejpam-5354	179	27	2	2	NUM
ejpam-5354	179	28	0	0	NUM
ejpam-5354	179	29	0.5	0.5	NUM
ejpam-5354	179	30	1	1	NUM
ejpam-5354	179	31	1.5	1.5	NUM
ejpam-5354	179	32	2	2	NUM
ejpam-5354	179	33	10	10	NUM
ejpam-5354	179	34	−0.03	−0.03	NOUN
ejpam-5354	179	35	10	10	NUM
ejpam-5354	179	36	0	0	NUM
ejpam-5354	179	37	a	a	DET
ejpam-5354	179	38	2	2	NUM
ejpam-5354	179	39	epochs	epoch	NOUN
ejpam-5354	179	40	step	step	NOUN
ejpam-5354	179	41	size	size	NOUN
ejpam-5354	179	42	=	=	SYM
ejpam-5354	179	43	0	0	NUM
ejpam-5354	179	44	,	,	PUNCT
ejpam-5354	179	45	at	at	ADP
ejpam-5354	179	46	epoch	epoch	NOUN
ejpam-5354	179	47	2	2	NUM
ejpam-5354	179	48	figure	figure	NOUN
ejpam-5354	179	49	4	4	NUM
ejpam-5354	179	50	:	:	PUNCT
ejpam-5354	179	51	training	training	NOUN
ejpam-5354	179	52	performance	performance	NOUN
ejpam-5354	179	53	hs	hs	PROPN
ejpam-5354	179	54	method	method	NOUN
ejpam-5354	179	55	(	(	PUNCT
ejpam-5354	179	56	left	left	ADJ
ejpam-5354	179	57	)	)	PUNCT
ejpam-5354	179	58	and	and	CCONJ
ejpam-5354	179	59	new	new	ADJ
ejpam-5354	179	60	method	method	NOUN
ejpam-5354	179	61	(	(	PUNCT
ejpam-5354	179	62	right	right	ADJ
ejpam-5354	179	63	)	)	PUNCT
ejpam-5354	179	64	,	,	PUNCT
ejpam-5354	179	65	of	of	ADP
ejpam-5354	179	66	problem	problem	NOUN
ejpam-5354	179	67	2	2	NUM
ejpam-5354	179	68	.	.	NOUN
ejpam-5354	179	69	5	5	NUM
ejpam-5354	179	70	.	.	X
ejpam-5354	179	71	conclusion	conclusion	NOUN
ejpam-5354	179	72	in	in	ADP
ejpam-5354	179	73	this	this	DET
ejpam-5354	179	74	paper	paper	NOUN
ejpam-5354	179	75	,	,	PUNCT
ejpam-5354	179	76	we	we	PRON
ejpam-5354	179	77	have	have	AUX
ejpam-5354	179	78	presented	present	VERB
ejpam-5354	179	79	a	a	DET
ejpam-5354	179	80	modified	modify	VERB
ejpam-5354	179	81	conjugate	conjugate	ADJ
ejpam-5354	179	82	gradient	gradient	NOUN
ejpam-5354	179	83	method	method	NOUN
ejpam-5354	179	84	specifically	specifically	ADV
ejpam-5354	179	85	designed	design	VERB
ejpam-5354	179	86	for	for	ADP
ejpam-5354	179	87	addressing	address	VERB
ejpam-5354	179	88	unconstrained	unconstrained	ADJ
ejpam-5354	179	89	optimization	optimization	NOUN
ejpam-5354	179	90	problems	problem	NOUN
ejpam-5354	179	91	.	.	PUNCT
ejpam-5354	180	1	our	our	PRON
ejpam-5354	180	2	new	new	ADJ
ejpam-5354	180	3	method	method	NOUN
ejpam-5354	180	4	builds	build	VERB
ejpam-5354	180	5	upon	upon	SCONJ
ejpam-5354	180	6	the	the	DET
ejpam-5354	180	7	classical	classical	ADJ
ejpam-5354	180	8	conjugate	conjugate	ADJ
ejpam-5354	180	9	gradient	gradient	NOUN
ejpam-5354	180	10	method	method	NOUN
ejpam-5354	180	11	while	while	SCONJ
ejpam-5354	180	12	incorporating	incorporate	VERB
ejpam-5354	180	13	enhancements	enhancement	NOUN
ejpam-5354	180	14	that	that	PRON
ejpam-5354	180	15	preserve	preserve	VERB
ejpam-5354	180	16	the	the	DET
ejpam-5354	180	17	sufficient	sufficient	ADJ
ejpam-5354	180	18	descent	descent	NOUN
ejpam-5354	180	19	properties	property	NOUN
ejpam-5354	180	20	.	.	PUNCT
ejpam-5354	181	1	moreover	moreover	ADV
ejpam-5354	181	2	,	,	PUNCT
ejpam-5354	181	3	we	we	PRON
ejpam-5354	181	4	have	have	AUX
ejpam-5354	181	5	explored	explore	VERB
ejpam-5354	181	6	the	the	DET
ejpam-5354	181	7	global	global	ADJ
ejpam-5354	181	8	convergence	convergence	NOUN
ejpam-5354	181	9	of	of	ADP
ejpam-5354	181	10	our	our	PRON
ejpam-5354	181	11	method	method	NOUN
ejpam-5354	181	12	under	under	ADP
ejpam-5354	181	13	appropriate	appropriate	ADJ
ejpam-5354	181	14	conditions	condition	NOUN
ejpam-5354	181	15	,	,	PUNCT
ejpam-5354	181	16	providing	provide	VERB
ejpam-5354	181	17	reassurance	reassurance	NOUN
ejpam-5354	181	18	of	of	ADP
ejpam-5354	181	19	its	its	PRON
ejpam-5354	181	20	ability	ability	NOUN
ejpam-5354	181	21	to	to	PART
ejpam-5354	181	22	converge	converge	VERB
ejpam-5354	181	23	to	to	ADP
ejpam-5354	181	24	the	the	DET
ejpam-5354	181	25	optimal	optimal	ADJ
ejpam-5354	181	26	solution	solution	NOUN
ejpam-5354	181	27	.	.	PUNCT
ejpam-5354	182	1	through	through	ADP
ejpam-5354	182	2	extensive	extensive	ADJ
ejpam-5354	182	3	numerical	numerical	ADJ
ejpam-5354	182	4	computations	computation	NOUN
ejpam-5354	182	5	and	and	CCONJ
ejpam-5354	182	6	experiments	experiment	NOUN
ejpam-5354	182	7	,	,	PUNCT
ejpam-5354	182	8	we	we	PRON
ejpam-5354	182	9	references	reference	VERB
ejpam-5354	182	10	2703	2703	NUM
ejpam-5354	182	11	have	have	AUX
ejpam-5354	182	12	demonstrated	demonstrate	VERB
ejpam-5354	182	13	the	the	DET
ejpam-5354	182	14	promising	promising	ADJ
ejpam-5354	182	15	performance	performance	NOUN
ejpam-5354	182	16	of	of	ADP
ejpam-5354	182	17	our	our	PRON
ejpam-5354	182	18	modified	modify	VERB
ejpam-5354	182	19	conjugate	conjugate	ADJ
ejpam-5354	182	20	gradient	gradient	NOUN
ejpam-5354	182	21	method	method	NOUN
ejpam-5354	182	22	.	.	PUNCT
ejpam-5354	183	1	the	the	DET
ejpam-5354	183	2	results	result	NOUN
ejpam-5354	183	3	showcase	showcase	VERB
ejpam-5354	183	4	its	its	PRON
ejpam-5354	183	5	effectiveness	effectiveness	NOUN
ejpam-5354	183	6	in	in	ADP
ejpam-5354	183	7	solving	solve	VERB
ejpam-5354	183	8	unconstrained	unconstrained	ADJ
ejpam-5354	183	9	optimization	optimization	NOUN
ejpam-5354	183	10	problems	problem	NOUN
ejpam-5354	183	11	,	,	PUNCT
ejpam-5354	183	12	surpassing	surpass	VERB
ejpam-5354	183	13	the	the	DET
ejpam-5354	183	14	classical	classical	ADJ
ejpam-5354	183	15	hs	hs	INTJ
ejpam-5354	183	16	training	training	NOUN
ejpam-5354	183	17	function	function	NOUN
ejpam-5354	183	18	in	in	ADP
ejpam-5354	183	19	terms	term	NOUN
ejpam-5354	183	20	of	of	ADP
ejpam-5354	183	21	reducing	reduce	VERB
ejpam-5354	183	22	the	the	DET
ejpam-5354	183	23	number	number	NOUN
ejpam-5354	183	24	of	of	ADP
ejpam-5354	183	25	iterations	iteration	NOUN
ejpam-5354	183	26	and	and	CCONJ
ejpam-5354	183	27	cpu	cpu	NOUN
ejpam-5354	183	28	time	time	NOUN
ejpam-5354	183	29	.	.	PUNCT
ejpam-5354	184	1	furthermore	furthermore	ADV
ejpam-5354	184	2	,	,	PUNCT
ejpam-5354	184	3	we	we	PRON
ejpam-5354	184	4	have	have	AUX
ejpam-5354	184	5	extended	extend	VERB
ejpam-5354	184	6	the	the	DET
ejpam-5354	184	7	application	application	NOUN
ejpam-5354	184	8	of	of	ADP
ejpam-5354	184	9	our	our	PRON
ejpam-5354	184	10	method	method	NOUN
ejpam-5354	184	11	to	to	PART
ejpam-5354	184	12	feed	feed	VERB
ejpam-5354	184	13	-	-	PUNCT
ejpam-5354	184	14	forward	forward	ADV
ejpam-5354	184	15	neural	neural	ADJ
ejpam-5354	184	16	networks	network	NOUN
ejpam-5354	184	17	,	,	PUNCT
ejpam-5354	184	18	demonstrating	demonstrate	VERB
ejpam-5354	184	19	its	its	PRON
ejpam-5354	184	20	compatibility	compatibility	NOUN
ejpam-5354	184	21	with	with	ADP
ejpam-5354	184	22	complex	complex	ADJ
ejpam-5354	184	23	tasks	task	NOUN
ejpam-5354	184	24	such	such	ADJ
ejpam-5354	184	25	as	as	ADP
ejpam-5354	184	26	value	value	NOUN
ejpam-5354	184	27	estimation	estimation	NOUN
ejpam-5354	184	28	for	for	ADP
ejpam-5354	184	29	continuous	continuous	ADJ
ejpam-5354	184	30	trigonometric	trigonometric	ADJ
ejpam-5354	184	31	functions	function	NOUN
ejpam-5354	184	32	.	.	PUNCT
ejpam-5354	185	1	the	the	DET
ejpam-5354	185	2	outcomes	outcome	NOUN
ejpam-5354	185	3	of	of	ADP
ejpam-5354	185	4	these	these	DET
ejpam-5354	185	5	experiments	experiment	NOUN
ejpam-5354	185	6	have	have	AUX
ejpam-5354	185	7	highlighted	highlight	VERB
ejpam-5354	185	8	the	the	DET
ejpam-5354	185	9	advantages	advantage	NOUN
ejpam-5354	185	10	of	of	ADP
ejpam-5354	185	11	our	our	PRON
ejpam-5354	185	12	approach	approach	NOUN
ejpam-5354	185	13	,	,	PUNCT
ejpam-5354	185	14	showcasing	showcase	VERB
ejpam-5354	185	15	its	its	PRON
ejpam-5354	185	16	ability	ability	NOUN
ejpam-5354	185	17	to	to	PART
ejpam-5354	185	18	enhance	enhance	VERB
ejpam-5354	185	19	the	the	DET
ejpam-5354	185	20	efficiency	efficiency	NOUN
ejpam-5354	185	21	and	and	CCONJ
ejpam-5354	185	22	effectiveness	effectiveness	NOUN
ejpam-5354	185	23	of	of	ADP
ejpam-5354	185	24	neural	neural	ADJ
ejpam-5354	185	25	network	network	NOUN
ejpam-5354	185	26	training	training	NOUN
ejpam-5354	185	27	.	.	PUNCT
ejpam-5354	186	1	overall	overall	ADV
ejpam-5354	186	2	,	,	PUNCT
ejpam-5354	186	3	the	the	DET
ejpam-5354	186	4	introduction	introduction	NOUN
ejpam-5354	186	5	of	of	ADP
ejpam-5354	186	6	our	our	PRON
ejpam-5354	186	7	modified	modify	VERB
ejpam-5354	186	8	conjugate	conjugate	ADJ
ejpam-5354	186	9	gradient	gradient	NOUN
ejpam-5354	186	10	method	method	NOUN
ejpam-5354	186	11	presents	present	VERB
ejpam-5354	186	12	a	a	DET
ejpam-5354	186	13	significant	significant	ADJ
ejpam-5354	186	14	advancement	advancement	NOUN
ejpam-5354	186	15	in	in	ADP
ejpam-5354	186	16	the	the	DET
ejpam-5354	186	17	field	field	NOUN
ejpam-5354	186	18	of	of	ADP
ejpam-5354	186	19	unconstrained	unconstrained	ADJ
ejpam-5354	186	20	optimization	optimization	NOUN
ejpam-5354	186	21	.	.	PUNCT
ejpam-5354	187	1	by	by	ADP
ejpam-5354	187	2	combining	combine	VERB
ejpam-5354	187	3	the	the	DET
ejpam-5354	187	4	desirable	desirable	ADJ
ejpam-5354	187	5	properties	property	NOUN
ejpam-5354	187	6	of	of	ADP
ejpam-5354	187	7	the	the	DET
ejpam-5354	187	8	classical	classical	ADJ
ejpam-5354	187	9	conjugate	conjugate	ADJ
ejpam-5354	187	10	gradient	gradient	NOUN
ejpam-5354	187	11	method	method	NOUN
ejpam-5354	187	12	with	with	ADP
ejpam-5354	187	13	enhancements	enhancement	NOUN
ejpam-5354	187	14	tailored	tailor	VERB
ejpam-5354	187	15	for	for	ADP
ejpam-5354	187	16	unconstrained	unconstrained	ADJ
ejpam-5354	187	17	problems	problem	NOUN
ejpam-5354	187	18	,	,	PUNCT
ejpam-5354	187	19	our	our	PRON
ejpam-5354	187	20	approach	approach	NOUN
ejpam-5354	187	21	offers	offer	VERB
ejpam-5354	187	22	improved	improved	ADJ
ejpam-5354	187	23	convergence	convergence	NOUN
ejpam-5354	187	24	and	and	CCONJ
ejpam-5354	187	25	computational	computational	ADJ
ejpam-5354	187	26	efficiency	efficiency	NOUN
ejpam-5354	187	27	.	.	PUNCT
ejpam-5354	188	1	these	these	DET
ejpam-5354	188	2	findings	finding	NOUN
ejpam-5354	188	3	have	have	VERB
ejpam-5354	188	4	implications	implication	NOUN
ejpam-5354	188	5	for	for	ADP
ejpam-5354	188	6	a	a	DET
ejpam-5354	188	7	wide	wide	ADJ
ejpam-5354	188	8	range	range	NOUN
ejpam-5354	188	9	of	of	ADP
ejpam-5354	188	10	domains	domain	NOUN
ejpam-5354	188	11	where	where	SCONJ
ejpam-5354	188	12	optimization	optimization	NOUN
ejpam-5354	188	13	plays	play	VERB
ejpam-5354	188	14	a	a	DET
ejpam-5354	188	15	crucial	crucial	ADJ
ejpam-5354	188	16	role	role	NOUN
ejpam-5354	188	17	,	,	PUNCT
ejpam-5354	188	18	paving	pave	VERB
ejpam-5354	188	19	the	the	DET
ejpam-5354	188	20	way	way	NOUN
ejpam-5354	188	21	for	for	ADP
ejpam-5354	188	22	more	more	ADV
ejpam-5354	188	23	efficient	efficient	ADJ
ejpam-5354	188	24	and	and	CCONJ
ejpam-5354	188	25	effective	effective	ADJ
ejpam-5354	188	26	problem	problem	NOUN
ejpam-5354	188	27	-	-	PUNCT
ejpam-5354	188	28	solving	solve	VERB
ejpam-5354	188	29	methodologies	methodology	NOUN
ejpam-5354	188	30	.	.	PUNCT
ejpam-5354	189	1	references	reference	NOUN
ejpam-5354	189	2	[	[	X
ejpam-5354	189	3	1	1	NUM
ejpam-5354	189	4	]	]	PUNCT
ejpam-5354	189	5	a.	a.	NOUN
ejpam-5354	189	6	alhawarat	alhawarat	PROPN
ejpam-5354	189	7	,	,	PUNCT
ejpam-5354	189	8	z.	z.	PROPN
ejpam-5354	189	9	salleh	salleh	PROPN
ejpam-5354	189	10	,	,	PUNCT
ejpam-5354	189	11	h.	h.	PROPN
ejpam-5354	189	12	alolaiyan	alolaiyan	PROPN
ejpam-5354	189	13	,	,	PUNCT
ejpam-5354	189	14	h.	h.	PROPN
ejpam-5354	189	15	el	el	PROPN
ejpam-5354	189	16	hor	hor	PROPN
ejpam-5354	189	17	,	,	PUNCT
ejpam-5354	189	18	and	and	CCONJ
ejpam-5354	189	19	s.	s.	PROPN
ejpam-5354	189	20	ismail	ismail	PROPN
ejpam-5354	189	21	.	.	PUNCT
ejpam-5354	190	1	a	a	DET
ejpam-5354	190	2	three	three	NUM
ejpam-5354	190	3	-	-	PUNCT
ejpam-5354	190	4	term	term	NOUN
ejpam-5354	190	5	conjugate	conjugate	ADJ
ejpam-5354	190	6	gradient	gradient	ADJ
ejpam-5354	190	7	descent	descent	NOUN
ejpam-5354	190	8	method	method	NOUN
ejpam-5354	190	9	with	with	ADP
ejpam-5354	190	10	some	some	DET
ejpam-5354	190	11	applications	application	NOUN
ejpam-5354	190	12	.	.	PUNCT
ejpam-5354	191	1	j.	j.	PROPN
ejpam-5354	191	2	inequalities	inequalities	PROPN
ejpam-5354	191	3	appl	appl	PROPN
ejpam-5354	191	4	.	.	PROPN
ejpam-5354	191	5	,	,	PUNCT
ejpam-5354	191	6	2024	2024	NUM
ejpam-5354	191	7	.	.	PUNCT
ejpam-5354	192	1	[	[	X
ejpam-5354	192	2	2	2	NUM
ejpam-5354	192	3	]	]	X
ejpam-5354	192	4	n.	n.	NOUN
ejpam-5354	192	5	andrei	andrei	PROPN
ejpam-5354	192	6	.	.	PUNCT
ejpam-5354	193	1	an	an	DET
ejpam-5354	193	2	unconstrained	unconstrained	ADJ
ejpam-5354	193	3	optimization	optimization	NOUN
ejpam-5354	193	4	test	test	NOUN
ejpam-5354	193	5	functions	function	NOUN
ejpam-5354	193	6	collection	collection	NOUN
ejpam-5354	193	7	.	.	PUNCT
ejpam-5354	194	1	adv	adv	PROPN
ejpam-5354	194	2	.	.	PUNCT
ejpam-5354	194	3	model	model	PROPN
ejpam-5354	194	4	.	.	PUNCT
ejpam-5354	195	1	optim	optim	PROPN
ejpam-5354	195	2	.	.	PROPN
ejpam-5354	195	3	,	,	PUNCT
ejpam-5354	195	4	10:147–161	10:147–161	PROPN
ejpam-5354	195	5	,	,	PUNCT
ejpam-5354	195	6	2008	2008	NUM
ejpam-5354	195	7	.	.	PUNCT
ejpam-5354	196	1	[	[	X
ejpam-5354	196	2	3	3	X
ejpam-5354	196	3	]	]	X
ejpam-5354	196	4	n.	n.	PROPN
ejpam-5354	196	5	andrei	andrei	PROPN
ejpam-5354	196	6	.	.	PUNCT
ejpam-5354	197	1	accelerated	accelerate	VERB
ejpam-5354	197	2	conjugate	conjugate	ADJ
ejpam-5354	197	3	gradient	gradient	ADJ
ejpam-5354	197	4	algorithm	algorithm	NOUN
ejpam-5354	197	5	with	with	ADP
ejpam-5354	197	6	finite	finite	ADJ
ejpam-5354	197	7	difference	difference	NOUN
ejpam-5354	197	8	hessian	hessian	NOUN
ejpam-5354	197	9	/	/	SYM
ejpam-5354	197	10	vector	vector	NOUN
ejpam-5354	197	11	product	product	NOUN
ejpam-5354	197	12	approximation	approximation	NOUN
ejpam-5354	197	13	for	for	ADP
ejpam-5354	197	14	unconstrained	unconstrained	ADJ
ejpam-5354	197	15	optimization	optimization	NOUN
ejpam-5354	197	16	.	.	PUNCT
ejpam-5354	198	1	j.	j.	PROPN
ejpam-5354	198	2	comput	comput	PROPN
ejpam-5354	198	3	.	.	PUNCT
ejpam-5354	199	1	appl	appl	PROPN
ejpam-5354	199	2	.	.	PROPN
ejpam-5354	199	3	math	math	PROPN
ejpam-5354	199	4	.	.	PUNCT
ejpam-5354	199	5	,	,	PUNCT
ejpam-5354	199	6	230:570–582	230:570–582	NUM
ejpam-5354	199	7	,	,	PUNCT
ejpam-5354	199	8	2009	2009	NUM
ejpam-5354	199	9	.	.	PUNCT
ejpam-5354	200	1	[	[	X
ejpam-5354	200	2	4	4	X
ejpam-5354	200	3	]	]	X
ejpam-5354	200	4	n.	n.	PROPN
ejpam-5354	200	5	andrei	andrei	PROPN
ejpam-5354	200	6	.	.	PUNCT
ejpam-5354	201	1	nonlinear	nonlinear	ADJ
ejpam-5354	201	2	conjugate	conjugate	ADJ
ejpam-5354	201	3	gradient	gradient	ADJ
ejpam-5354	201	4	methods	method	NOUN
ejpam-5354	201	5	for	for	ADP
ejpam-5354	201	6	unconstrained	unconstrained	ADJ
ejpam-5354	201	7	optimization	optimization	NOUN
ejpam-5354	201	8	.	.	PUNCT
ejpam-5354	202	1	springer	springer	NOUN
ejpam-5354	202	2	optimization	optimization	NOUN
ejpam-5354	202	3	and	and	CCONJ
ejpam-5354	202	4	its	its	PRON
ejpam-5354	202	5	application	application	NOUN
ejpam-5354	202	6	,	,	PUNCT
ejpam-5354	202	7	2020	2020	NUM
ejpam-5354	202	8	.	.	PUNCT
ejpam-5354	203	1	[	[	X
ejpam-5354	203	2	5	5	NUM
ejpam-5354	203	3	]	]	PUNCT
ejpam-5354	203	4	c.	c.	PROPN
ejpam-5354	203	5	m.	m.	NOUN
ejpam-5354	203	6	bishop	bishop	PROPN
ejpam-5354	203	7	.	.	PUNCT
ejpam-5354	204	1	neural	neural	ADJ
ejpam-5354	204	2	networks	network	NOUN
ejpam-5354	204	3	for	for	ADP
ejpam-5354	204	4	pattern	pattern	NOUN
ejpam-5354	204	5	recognition	recognition	NOUN
ejpam-5354	204	6	.	.	PUNCT
ejpam-5354	205	1	1995	1995	NUM
ejpam-5354	205	2	.	.	PUNCT
ejpam-5354	206	1	[	[	X
ejpam-5354	206	2	6	6	NUM
ejpam-5354	206	3	]	]	PUNCT
ejpam-5354	206	4	z.	z.	PROPN
ejpam-5354	206	5	chen	chen	PROPN
ejpam-5354	206	6	,	,	PUNCT
ejpam-5354	206	7	h.	h.	PROPN
ejpam-5354	206	8	shao	shao	PROPN
ejpam-5354	206	9	,	,	PUNCT
ejpam-5354	206	10	p.	p.	PROPN
ejpam-5354	206	11	liu	liu	PROPN
ejpam-5354	206	12	,	,	PUNCT
ejpam-5354	206	13	g.	g.	PROPN
ejpam-5354	206	14	li	li	PROPN
ejpam-5354	206	15	,	,	PUNCT
ejpam-5354	206	16	and	and	CCONJ
ejpam-5354	206	17	x.	x.	PROPN
ejpam-5354	206	18	rong	rong	PROPN
ejpam-5354	206	19	.	.	PUNCT
ejpam-5354	207	1	an	an	DET
ejpam-5354	207	2	efficient	efficient	ADJ
ejpam-5354	207	3	hybrid	hybrid	ADJ
ejpam-5354	207	4	conjugate	conjugate	ADJ
ejpam-5354	207	5	gradient	gradient	NOUN
ejpam-5354	207	6	method	method	NOUN
ejpam-5354	207	7	with	with	ADP
ejpam-5354	207	8	an	an	DET
ejpam-5354	207	9	adaptive	adaptive	ADJ
ejpam-5354	207	10	strategy	strategy	NOUN
ejpam-5354	207	11	and	and	CCONJ
ejpam-5354	207	12	applications	application	NOUN
ejpam-5354	207	13	in	in	ADP
ejpam-5354	207	14	image	image	NOUN
ejpam-5354	207	15	restoration	restoration	NOUN
ejpam-5354	207	16	problems	problem	NOUN
ejpam-5354	207	17	.	.	PUNCT
ejpam-5354	208	1	appl	appl	PROPN
ejpam-5354	208	2	.	.	PUNCT
ejpam-5354	209	1	numer	numer	PROPN
ejpam-5354	209	2	.	.	PUNCT
ejpam-5354	209	3	math	math	PROPN
ejpam-5354	209	4	.	.	PUNCT
ejpam-5354	209	5	,	,	PUNCT
ejpam-5354	209	6	204:362–379	204:362–379	NUM
ejpam-5354	209	7	,	,	PUNCT
ejpam-5354	209	8	2024	2024	NUM
ejpam-5354	209	9	.	.	PUNCT
ejpam-5354	210	1	[	[	X
ejpam-5354	210	2	7	7	X
ejpam-5354	210	3	]	]	X
ejpam-5354	210	4	y.	y.	PROPN
ejpam-5354	210	5	dai	dai	PROPN
ejpam-5354	210	6	,	,	PUNCT
ejpam-5354	210	7	j.	j.	PROPN
ejpam-5354	210	8	han	han	PROPN
ejpam-5354	210	9	,	,	PUNCT
ejpam-5354	210	10	g.	g.	PROPN
ejpam-5354	210	11	liu	liu	PROPN
ejpam-5354	210	12	,	,	PUNCT
ejpam-5354	210	13	d.	d.	PROPN
ejpam-5354	210	14	sun	sun	PROPN
ejpam-5354	210	15	,	,	PUNCT
ejpam-5354	210	16	h.	h.	PROPN
ejpam-5354	210	17	yin	yin	PROPN
ejpam-5354	210	18	,	,	PUNCT
ejpam-5354	210	19	and	and	CCONJ
ejpam-5354	210	20	y.-x	y.-x	PROPN
ejpam-5354	210	21	.	.	PUNCT
ejpam-5354	211	1	yuan	yuan	PROPN
ejpam-5354	211	2	.	.	PUNCT
ejpam-5354	212	1	convergence	convergence	NOUN
ejpam-5354	212	2	properties	property	NOUN
ejpam-5354	212	3	of	of	ADP
ejpam-5354	212	4	nonlinear	nonlinear	ADJ
ejpam-5354	212	5	conjugate	conjugate	ADJ
ejpam-5354	212	6	gradient	gradient	ADJ
ejpam-5354	212	7	methods	method	NOUN
ejpam-5354	212	8	.	.	PUNCT
ejpam-5354	213	1	siam	siam	PROPN
ejpam-5354	213	2	j.	j.	PROPN
ejpam-5354	213	3	optim	optim	PROPN
ejpam-5354	213	4	.	.	PROPN
ejpam-5354	213	5	,	,	PUNCT
ejpam-5354	213	6	10:345–358	10:345–358	PROPN
ejpam-5354	213	7	,	,	PUNCT
ejpam-5354	213	8	2000	2000	NUM
ejpam-5354	213	9	.	.	PUNCT
ejpam-5354	214	1	[	[	X
ejpam-5354	214	2	8	8	X
ejpam-5354	214	3	]	]	X
ejpam-5354	214	4	y.	y.	PROPN
ejpam-5354	214	5	h.	h.	PROPN
ejpam-5354	214	6	dai	dai	PROPN
ejpam-5354	214	7	and	and	CCONJ
ejpam-5354	214	8	y.	y.	PROPN
ejpam-5354	214	9	yuan	yuan	PROPN
ejpam-5354	214	10	.	.	PUNCT
ejpam-5354	215	1	a	a	DET
ejpam-5354	215	2	nonlinear	nonlinear	ADJ
ejpam-5354	215	3	conjugate	conjugate	ADJ
ejpam-5354	215	4	gradient	gradient	NOUN
ejpam-5354	215	5	method	method	NOUN
ejpam-5354	215	6	with	with	ADP
ejpam-5354	215	7	a	a	DET
ejpam-5354	215	8	strong	strong	ADJ
ejpam-5354	215	9	global	global	ADJ
ejpam-5354	215	10	convergence	convergence	NOUN
ejpam-5354	215	11	property	property	NOUN
ejpam-5354	215	12	.	.	PUNCT
ejpam-5354	216	1	siam	siam	PROPN
ejpam-5354	216	2	j.	j.	PROPN
ejpam-5354	216	3	optim	optim	PROPN
ejpam-5354	216	4	.	.	PROPN
ejpam-5354	216	5	,	,	PUNCT
ejpam-5354	216	6	10:177–182	10:177–182	PROPN
ejpam-5354	216	7	,	,	PUNCT
ejpam-5354	216	8	1999	1999	NUM
ejpam-5354	216	9	.	.	PUNCT
ejpam-5354	217	1	[	[	X
ejpam-5354	217	2	9	9	NUM
ejpam-5354	217	3	]	]	PUNCT
ejpam-5354	217	4	r.	r.	PROPN
ejpam-5354	217	5	fletcher	fletcher	PROPN
ejpam-5354	217	6	.	.	PUNCT
ejpam-5354	218	1	practical	practical	ADJ
ejpam-5354	218	2	methods	method	NOUN
ejpam-5354	218	3	of	of	ADP
ejpam-5354	218	4	optimization	optimization	NOUN
ejpam-5354	218	5	.	.	PUNCT
ejpam-5354	219	1	john	john	PROPN
ejpam-5354	219	2	wiley	wiley	PROPN
ejpam-5354	219	3	&	&	CCONJ
ejpam-5354	219	4	sons	son	NOUN
ejpam-5354	219	5	,	,	PUNCT
ejpam-5354	219	6	new	new	PROPN
ejpam-5354	219	7	york	york	PROPN
ejpam-5354	219	8	,	,	PUNCT
ejpam-5354	219	9	2013	2013	NUM
ejpam-5354	219	10	.	.	PUNCT
ejpam-5354	220	1	[	[	X
ejpam-5354	220	2	10	10	NUM
ejpam-5354	220	3	]	]	X
ejpam-5354	220	4	r.	r.	PROPN
ejpam-5354	220	5	fletcher	fletcher	PROPN
ejpam-5354	220	6	.	.	PUNCT
ejpam-5354	221	1	practical	practical	ADJ
ejpam-5354	221	2	methods	method	NOUN
ejpam-5354	221	3	of	of	ADP
ejpam-5354	221	4	optimization	optimization	NOUN
ejpam-5354	221	5	.	.	PUNCT
ejpam-5354	222	1	john	john	PROPN
ejpam-5354	222	2	wiley	wiley	PROPN
ejpam-5354	222	3	&	&	CCONJ
ejpam-5354	222	4	sons	son	NOUN
ejpam-5354	222	5	,	,	PUNCT
ejpam-5354	222	6	2020	2020	NUM
ejpam-5354	222	7	.	.	PUNCT
ejpam-5354	223	1	references	reference	NOUN
ejpam-5354	223	2	2704	2704	NUM
ejpam-5354	224	1	[	[	X
ejpam-5354	224	2	11	11	NUM
ejpam-5354	224	3	]	]	PUNCT
ejpam-5354	224	4	r.	r.	PROPN
ejpam-5354	224	5	fletcher	fletcher	PROPN
ejpam-5354	224	6	and	and	CCONJ
ejpam-5354	224	7	m.	m.	PROPN
ejpam-5354	224	8	j.	j.	PROPN
ejpam-5354	224	9	d.	d.	PROPN
ejpam-5354	224	10	powell	powell	PROPN
ejpam-5354	224	11	.	.	PUNCT
ejpam-5354	225	1	a	a	DET
ejpam-5354	225	2	rapidly	rapidly	ADV
ejpam-5354	225	3	convergent	convergent	ADJ
ejpam-5354	225	4	descent	descent	NOUN
ejpam-5354	225	5	method	method	NOUN
ejpam-5354	225	6	for	for	ADP
ejpam-5354	225	7	minimization	minimization	NOUN
ejpam-5354	225	8	.	.	PUNCT
ejpam-5354	226	1	comput	comput	NOUN
ejpam-5354	226	2	.	.	PUNCT
ejpam-5354	227	1	j.	j.	PROPN
ejpam-5354	227	2	,	,	PUNCT
ejpam-5354	227	3	6:163–168	6:163–168	PROPN
ejpam-5354	227	4	,	,	PUNCT
ejpam-5354	227	5	1963	1963	NUM
ejpam-5354	227	6	.	.	PUNCT
ejpam-5354	228	1	[	[	X
ejpam-5354	228	2	12	12	NUM
ejpam-5354	228	3	]	]	PUNCT
ejpam-5354	228	4	r.	r.	PROPN
ejpam-5354	228	5	fletcher	fletcher	PROPN
ejpam-5354	228	6	and	and	CCONJ
ejpam-5354	228	7	c.	c.	PROPN
ejpam-5354	228	8	m.	m.	PROPN
ejpam-5354	228	9	reeves	reeves	PROPN
ejpam-5354	228	10	.	.	PUNCT
ejpam-5354	229	1	function	function	NOUN
ejpam-5354	229	2	minimization	minimization	NOUN
ejpam-5354	229	3	by	by	ADP
ejpam-5354	229	4	conjugate	conjugate	ADJ
ejpam-5354	229	5	gradients	gradient	NOUN
ejpam-5354	229	6	.	.	PUNCT
ejpam-5354	230	1	comput	comput	NOUN
ejpam-5354	230	2	.	.	PUNCT
ejpam-5354	231	1	j.	j.	PROPN
ejpam-5354	231	2	,	,	PUNCT
ejpam-5354	231	3	7:149–154	7:149–154	PROPN
ejpam-5354	231	4	,	,	PUNCT
ejpam-5354	231	5	1964	1964	NUM
ejpam-5354	231	6	.	.	PUNCT
ejpam-5354	232	1	[	[	X
ejpam-5354	232	2	13	13	NUM
ejpam-5354	232	3	]	]	PUNCT
ejpam-5354	232	4	t.	t.	PROPN
ejpam-5354	232	5	ginsburg	ginsburg	PROPN
ejpam-5354	232	6	.	.	PUNCT
ejpam-5354	233	1	the	the	DET
ejpam-5354	233	2	conjugate	conjugate	ADJ
ejpam-5354	233	3	gradient	gradient	NOUN
ejpam-5354	233	4	method	method	NOUN
ejpam-5354	233	5	in	in	ADP
ejpam-5354	233	6	extremal	extremal	ADJ
ejpam-5354	233	7	problems	problem	NOUN
ejpam-5354	233	8	.	.	PUNCT
ejpam-5354	234	1	ussr	ussr	ADJ
ejpam-5354	234	2	comput	comput	PROPN
ejpam-5354	234	3	.	.	PUNCT
ejpam-5354	235	1	math	math	NOUN
ejpam-5354	235	2	.	.	PUNCT
ejpam-5354	236	1	math	math	NOUN
ejpam-5354	236	2	.	.	PUNCT
ejpam-5354	237	1	phys	phy	NOUN
ejpam-5354	237	2	.	.	PUNCT
ejpam-5354	237	3	,	,	PUNCT
ejpam-5354	237	4	9:94–112	9:94–112	NUM
ejpam-5354	237	5	,	,	PUNCT
ejpam-5354	237	6	1969	1969	NUM
ejpam-5354	237	7	.	.	PUNCT
ejpam-5354	238	1	[	[	X
ejpam-5354	238	2	14	14	NUM
ejpam-5354	238	3	]	]	X
ejpam-5354	238	4	w.	w.	PROPN
ejpam-5354	238	5	hager	hager	PROPN
ejpam-5354	238	6	and	and	CCONJ
ejpam-5354	238	7	h.	h.	PROPN
ejpam-5354	238	8	zhang	zhang	PROPN
ejpam-5354	238	9	.	.	PUNCT
ejpam-5354	239	1	a	a	DET
ejpam-5354	239	2	survey	survey	NOUN
ejpam-5354	239	3	of	of	ADP
ejpam-5354	239	4	nonlinear	nonlinear	ADJ
ejpam-5354	239	5	conjugate	conjugate	ADJ
ejpam-5354	239	6	gradient	gradient	ADJ
ejpam-5354	239	7	methods	method	NOUN
ejpam-5354	239	8	.	.	PUNCT
ejpam-5354	240	1	pacific	pacific	PROPN
ejpam-5354	240	2	j.	j.	PROPN
ejpam-5354	240	3	optim	optim	PROPN
ejpam-5354	240	4	.	.	PROPN
ejpam-5354	240	5	,	,	PUNCT
ejpam-5354	240	6	2:35–58	2:35–58	NUM
ejpam-5354	240	7	,	,	PUNCT
ejpam-5354	240	8	2006	2006	NUM
ejpam-5354	240	9	.	.	PUNCT
ejpam-5354	241	1	[	[	X
ejpam-5354	241	2	15	15	NUM
ejpam-5354	241	3	]	]	X
ejpam-5354	241	4	a.	a.	NOUN
ejpam-5354	241	5	hassan	hassan	PROPN
ejpam-5354	241	6	,	,	PUNCT
ejpam-5354	241	7	b.	b.	PROPN
ejpam-5354	241	8	jabbar	jabbar	PROPN
ejpam-5354	241	9	,	,	PUNCT
ejpam-5354	241	10	and	and	CCONJ
ejpam-5354	241	11	a.	a.	NOUN
ejpam-5354	241	12	bayati	bayati	PROPN
ejpam-5354	241	13	.	.	PUNCT
ejpam-5354	242	1	a	a	DET
ejpam-5354	242	2	new	new	ADJ
ejpam-5354	242	3	class	class	NOUN
ejpam-5354	242	4	of	of	ADP
ejpam-5354	242	5	nonlinear	nonlinear	ADJ
ejpam-5354	242	6	conjugate	conjugate	ADJ
ejpam-5354	242	7	gradient	gradient	NOUN
ejpam-5354	242	8	method	method	NOUN
ejpam-5354	242	9	for	for	ADP
ejpam-5354	242	10	solving	solve	VERB
ejpam-5354	242	11	unconstrained	unconstrained	ADJ
ejpam-5354	242	12	minimization	minimization	NOUN
ejpam-5354	242	13	problems	problem	NOUN
ejpam-5354	242	14	.	.	PUNCT
ejpam-5354	243	1	in	in	ADP
ejpam-5354	243	2	presented	present	VERB
ejpam-5354	243	3	at	at	ADP
ejpam-5354	243	4	the	the	DET
ejpam-5354	243	5	(	(	PUNCT
ejpam-5354	243	6	2019	2019	NUM
ejpam-5354	243	7	)	)	PUNCT
ejpam-5354	243	8	.	.	PUNCT
ejpam-5354	244	1	[	[	X
ejpam-5354	244	2	16	16	NUM
ejpam-5354	244	3	]	]	PUNCT
ejpam-5354	244	4	m.	m.	PROPN
ejpam-5354	244	5	r.	r.	PROPN
ejpam-5354	244	6	hestenes	hestenes	PROPN
ejpam-5354	244	7	and	and	CCONJ
ejpam-5354	244	8	e.	e.	PROPN
ejpam-5354	244	9	stiefel	stiefel	PROPN
ejpam-5354	244	10	.	.	PUNCT
ejpam-5354	245	1	methods	method	NOUN
ejpam-5354	245	2	of	of	ADP
ejpam-5354	245	3	conjugate	conjugate	ADJ
ejpam-5354	245	4	gradients	gradient	NOUN
ejpam-5354	245	5	for	for	ADP
ejpam-5354	245	6	solving	solve	VERB
ejpam-5354	245	7	linear	linear	NOUN
ejpam-5354	245	8	systems	system	NOUN
ejpam-5354	245	9	.	.	PUNCT
ejpam-5354	246	1	j.	j.	PROPN
ejpam-5354	246	2	res	res	PROPN
ejpam-5354	246	3	.	.	PROPN
ejpam-5354	246	4	natl	natl	PROPN
ejpam-5354	246	5	.	.	PUNCT
ejpam-5354	246	6	bur	bur	PROPN
ejpam-5354	246	7	.	.	PUNCT
ejpam-5354	246	8	stand	stand	VERB
ejpam-5354	246	9	.	.	PUNCT
ejpam-5354	247	1	(	(	PUNCT
ejpam-5354	247	2	1934	1934	NUM
ejpam-5354	247	3	)	)	PUNCT
ejpam-5354	247	4	,	,	PUNCT
ejpam-5354	247	5	49:409	49:409	NUM
ejpam-5354	247	6	,	,	PUNCT
ejpam-5354	247	7	1952	1952	NUM
ejpam-5354	247	8	.	.	PUNCT
ejpam-5354	248	1	[	[	X
ejpam-5354	248	2	17	17	NUM
ejpam-5354	248	3	]	]	PUNCT
ejpam-5354	248	4	a.	a.	NOUN
ejpam-5354	248	5	hmich	hmich	NOUN
ejpam-5354	248	6	,	,	PUNCT
ejpam-5354	248	7	a.	a.	NOUN
ejpam-5354	248	8	badri	badri	PROPN
ejpam-5354	248	9	,	,	PUNCT
ejpam-5354	248	10	and	and	CCONJ
ejpam-5354	248	11	a.	a.	NOUN
ejpam-5354	248	12	sahel	sahel	PROPN
ejpam-5354	248	13	.	.	PUNCT
ejpam-5354	249	1	automatic	automatic	ADJ
ejpam-5354	249	2	speaker	speaker	NOUN
ejpam-5354	249	3	identification	identification	NOUN
ejpam-5354	249	4	by	by	ADP
ejpam-5354	249	5	using	use	VERB
ejpam-5354	249	6	the	the	DET
ejpam-5354	249	7	neural	neural	ADJ
ejpam-5354	249	8	network	network	NOUN
ejpam-5354	249	9	.	.	PUNCT
ejpam-5354	250	1	in	in	ADP
ejpam-5354	250	2	2011	2011	NUM
ejpam-5354	250	3	international	international	ADJ
ejpam-5354	250	4	conference	conference	NOUN
ejpam-5354	250	5	on	on	ADP
ejpam-5354	250	6	multimedia	multimedia	NOUN
ejpam-5354	250	7	computing	computing	NOUN
ejpam-5354	250	8	and	and	CCONJ
ejpam-5354	250	9	systems	system	NOUN
ejpam-5354	250	10	,	,	PUNCT
ejpam-5354	250	11	pages	page	NOUN
ejpam-5354	250	12	1–5	1–5	NUM
ejpam-5354	250	13	,	,	PUNCT
ejpam-5354	250	14	2011	2011	NUM
ejpam-5354	250	15	.	.	PUNCT
ejpam-5354	251	1	[	[	X
ejpam-5354	251	2	18	18	NUM
ejpam-5354	251	3	]	]	PUNCT
ejpam-5354	251	4	a.	a.	NOUN
ejpam-5354	251	5	l.	l.	PROPN
ejpam-5354	251	6	ibrahim	ibrahim	PROPN
ejpam-5354	251	7	and	and	CCONJ
ejpam-5354	251	8	m.	m.	PROPN
ejpam-5354	251	9	g.	g.	PROPN
ejpam-5354	251	10	mohammed	mohammed	PROPN
ejpam-5354	251	11	.	.	PUNCT
ejpam-5354	252	1	a	a	DET
ejpam-5354	252	2	new	new	ADJ
ejpam-5354	252	3	three	three	NUM
ejpam-5354	252	4	-	-	PUNCT
ejpam-5354	252	5	term	term	NOUN
ejpam-5354	252	6	conjugate	conjugate	ADJ
ejpam-5354	252	7	gradient	gradient	NOUN
ejpam-5354	252	8	method	method	NOUN
ejpam-5354	252	9	for	for	ADP
ejpam-5354	252	10	training	train	VERB
ejpam-5354	252	11	neural	neural	ADJ
ejpam-5354	252	12	networks	network	NOUN
ejpam-5354	252	13	with	with	ADP
ejpam-5354	252	14	global	global	ADJ
ejpam-5354	252	15	convergence	convergence	NOUN
ejpam-5354	252	16	.	.	PUNCT
ejpam-5354	253	1	indones	indone	NOUN
ejpam-5354	253	2	.	.	PUNCT
ejpam-5354	254	1	j.	j.	PROPN
ejpam-5354	254	2	electr	electr	PROPN
ejpam-5354	254	3	.	.	PUNCT
ejpam-5354	255	1	eng	eng	PROPN
ejpam-5354	255	2	.	.	PUNCT
ejpam-5354	256	1	comput	comput	PROPN
ejpam-5354	256	2	.	.	PUNCT
ejpam-5354	257	1	sci	sci	PROPN
ejpam-5354	257	2	.	.	PROPN
ejpam-5354	257	3	,	,	PUNCT
ejpam-5354	257	4	28:551–558	28:551–558	PROPN
ejpam-5354	257	5	,	,	PUNCT
ejpam-5354	257	6	2022	2022	NUM
ejpam-5354	257	7	.	.	PUNCT
ejpam-5354	258	1	[	[	X
ejpam-5354	258	2	19	19	NUM
ejpam-5354	258	3	]	]	PUNCT
ejpam-5354	258	4	a.	a.	NOUN
ejpam-5354	258	5	l.	l.	PROPN
ejpam-5354	258	6	ibrahim	ibrahim	PROPN
ejpam-5354	258	7	and	and	CCONJ
ejpam-5354	258	8	m.	m.	PROPN
ejpam-5354	258	9	g.	g.	PROPN
ejpam-5354	258	10	mohammed	mohammed	PROPN
ejpam-5354	258	11	.	.	PUNCT
ejpam-5354	259	1	a	a	DET
ejpam-5354	259	2	new	new	ADJ
ejpam-5354	259	3	conjugate	conjugate	ADJ
ejpam-5354	259	4	gradient	gradient	NOUN
ejpam-5354	259	5	for	for	ADP
ejpam-5354	259	6	unconstrained	unconstrained	ADJ
ejpam-5354	259	7	optimization	optimization	NOUN
ejpam-5354	259	8	problems	problem	NOUN
ejpam-5354	259	9	and	and	CCONJ
ejpam-5354	259	10	its	its	PRON
ejpam-5354	259	11	applications	application	NOUN
ejpam-5354	259	12	in	in	ADP
ejpam-5354	259	13	neural	neural	ADJ
ejpam-5354	259	14	networks	network	NOUN
ejpam-5354	259	15	.	.	PUNCT
ejpam-5354	260	1	indones	indone	NOUN
ejpam-5354	260	2	.	.	PUNCT
ejpam-5354	261	1	j.	j.	PROPN
ejpam-5354	261	2	electr	electr	PROPN
ejpam-5354	261	3	.	.	PUNCT
ejpam-5354	262	1	eng	eng	PROPN
ejpam-5354	262	2	.	.	PUNCT
ejpam-5354	263	1	comput	comput	PROPN
ejpam-5354	263	2	.	.	PUNCT
ejpam-5354	264	1	sci	sci	PROPN
ejpam-5354	264	2	.	.	PROPN
ejpam-5354	264	3	,	,	PUNCT
ejpam-5354	264	4	33:93–100	33:93–100	NUM
ejpam-5354	264	5	,	,	PUNCT
ejpam-5354	264	6	2024	2024	NUM
ejpam-5354	264	7	.	.	PUNCT
ejpam-5354	265	1	[	[	X
ejpam-5354	265	2	20	20	NUM
ejpam-5354	265	3	]	]	PUNCT
ejpam-5354	265	4	a.	a.	PROPN
ejpam-5354	265	5	l.	l.	PROPN
ejpam-5354	265	6	ibrahim	ibrahim	PROPN
ejpam-5354	265	7	,	,	PUNCT
ejpam-5354	265	8	m.	m.	NOUN
ejpam-5354	265	9	a.	a.	PROPN
ejpam-5354	265	10	sadiq	sadiq	PROPN
ejpam-5354	265	11	,	,	PUNCT
ejpam-5354	265	12	and	and	CCONJ
ejpam-5354	265	13	s.	s.	PROPN
ejpam-5354	265	14	g.	g.	PROPN
ejpam-5354	265	15	shareef	shareef	PROPN
ejpam-5354	265	16	.	.	PUNCT
ejpam-5354	266	1	a	a	DET
ejpam-5354	266	2	new	new	ADJ
ejpam-5354	266	3	conjugate	conjugate	ADJ
ejpam-5354	266	4	gradient	gradient	ADJ
ejpam-5354	266	5	coefficient	coefficient	NOUN
ejpam-5354	266	6	for	for	ADP
ejpam-5354	266	7	unconstrained	unconstrained	ADJ
ejpam-5354	266	8	optimization	optimization	NOUN
ejpam-5354	266	9	based	base	VERB
ejpam-5354	266	10	on	on	ADP
ejpam-5354	266	11	dai	dai	PROPN
ejpam-5354	266	12	-	-	PUNCT
ejpam-5354	266	13	liao	liao	PROPN
ejpam-5354	266	14	.	.	PUNCT
ejpam-5354	267	1	sci	sci	PROPN
ejpam-5354	267	2	.	.	PUNCT
ejpam-5354	267	3	j.	j.	PROPN
ejpam-5354	267	4	univ	univ	PROPN
ejpam-5354	267	5	.	.	PUNCT
ejpam-5354	268	1	zakho	zakho	PROPN
ejpam-5354	268	2	.	.	PROPN
ejpam-5354	268	3	,	,	PUNCT
ejpam-5354	268	4	7:34–36	7:34–36	NUM
ejpam-5354	268	5	,	,	PUNCT
ejpam-5354	268	6	2019	2019	NUM
ejpam-5354	268	7	.	.	PUNCT
ejpam-5354	269	1	[	[	X
ejpam-5354	269	2	21	21	NUM
ejpam-5354	269	3	]	]	PUNCT
ejpam-5354	269	4	a.	a.	NOUN
ejpam-5354	269	5	l.	l.	PROPN
ejpam-5354	269	6	ibrahim	ibrahim	PROPN
ejpam-5354	269	7	and	and	CCONJ
ejpam-5354	269	8	s.	s.	PROPN
ejpam-5354	269	9	g.	g.	PROPN
ejpam-5354	269	10	shareef	shareef	PROPN
ejpam-5354	269	11	.	.	PROPN
ejpam-5354	269	12	modified	modify	VERB
ejpam-5354	269	13	conjugate	conjugate	ADJ
ejpam-5354	269	14	gradient	gradient	NOUN
ejpam-5354	269	15	method	method	NOUN
ejpam-5354	269	16	for	for	ADP
ejpam-5354	269	17	training	train	VERB
ejpam-5354	269	18	neural	neural	ADJ
ejpam-5354	269	19	networks	network	NOUN
ejpam-5354	269	20	based	base	VERB
ejpam-5354	269	21	on	on	ADP
ejpam-5354	269	22	logistic	logistic	ADJ
ejpam-5354	269	23	mapping	mapping	NOUN
ejpam-5354	269	24	.	.	PUNCT
ejpam-5354	270	1	2:45–51	2:45–51	NUM
ejpam-5354	270	2	,	,	PUNCT
ejpam-5354	270	3	2019	2019	NUM
ejpam-5354	270	4	.	.	PUNCT
ejpam-5354	271	1	[	[	X
ejpam-5354	271	2	22	22	NUM
ejpam-5354	271	3	]	]	PUNCT
ejpam-5354	271	4	x.	x.	PROPN
ejpam-5354	271	5	jiang	jiang	PROPN
ejpam-5354	271	6	,	,	PUNCT
ejpam-5354	271	7	h.	h.	PROPN
ejpam-5354	271	8	yang	yang	PROPN
ejpam-5354	271	9	,	,	PUNCT
ejpam-5354	271	10	j.	j.	PROPN
ejpam-5354	271	11	yin	yin	PROPN
ejpam-5354	271	12	,	,	PUNCT
ejpam-5354	271	13	and	and	CCONJ
ejpam-5354	271	14	w.	w.	PROPN
ejpam-5354	271	15	liao	liao	PROPN
ejpam-5354	271	16	.	.	PUNCT
ejpam-5354	272	1	a	a	DET
ejpam-5354	272	2	three	three	NUM
ejpam-5354	272	3	-	-	PUNCT
ejpam-5354	272	4	term	term	NOUN
ejpam-5354	272	5	conjugate	conjugate	ADJ
ejpam-5354	272	6	gradient	gradient	ADJ
ejpam-5354	272	7	algorithm	algorithm	NOUN
ejpam-5354	272	8	with	with	ADP
ejpam-5354	272	9	restart	restart	NOUN
ejpam-5354	272	10	procedure	procedure	NOUN
ejpam-5354	272	11	to	to	PART
ejpam-5354	272	12	solve	solve	VERB
ejpam-5354	272	13	image	image	NOUN
ejpam-5354	272	14	restoration	restoration	NOUN
ejpam-5354	272	15	problems	problem	NOUN
ejpam-5354	272	16	.	.	PUNCT
ejpam-5354	273	1	j.	j.	PROPN
ejpam-5354	273	2	comput	comput	PROPN
ejpam-5354	273	3	.	.	PUNCT
ejpam-5354	274	1	appl	appl	PROPN
ejpam-5354	274	2	.	.	PROPN
ejpam-5354	274	3	math	math	PROPN
ejpam-5354	274	4	.	.	PUNCT
ejpam-5354	274	5	,	,	PUNCT
ejpam-5354	274	6	424:115020	424:115020	NUM
ejpam-5354	274	7	,	,	PUNCT
ejpam-5354	274	8	2022	2022	NUM
ejpam-5354	274	9	.	.	PUNCT
ejpam-5354	275	1	[	[	X
ejpam-5354	275	2	23	23	NUM
ejpam-5354	275	3	]	]	PUNCT
ejpam-5354	275	4	k.	k.	PROPN
ejpam-5354	275	5	kamilu	kamilu	PROPN
ejpam-5354	275	6	,	,	PUNCT
ejpam-5354	275	7	m.	m.	NOUN
ejpam-5354	275	8	i.	i.	PROPN
ejpam-5354	275	9	sulaiman	sulaiman	PROPN
ejpam-5354	275	10	,	,	PUNCT
ejpam-5354	275	11	a.	a.	PROPN
ejpam-5354	275	12	muhammad	muhammad	PROPN
ejpam-5354	275	13	,	,	PUNCT
ejpam-5354	275	14	a.	a.	PROPN
ejpam-5354	275	15	w.	w.	PROPN
ejpam-5354	275	16	mohamad	mohamad	PROPN
ejpam-5354	275	17	,	,	PUNCT
ejpam-5354	275	18	and	and	CCONJ
ejpam-5354	275	19	m.	m.	NOUN
ejpam-5354	275	20	mamat	mamat	PROPN
ejpam-5354	275	21	.	.	PUNCT
ejpam-5354	276	1	performance	performance	NOUN
ejpam-5354	276	2	evaluation	evaluation	NOUN
ejpam-5354	276	3	of	of	ADP
ejpam-5354	276	4	a	a	DET
ejpam-5354	276	5	novel	novel	ADJ
ejpam-5354	276	6	conjugate	conjugate	ADJ
ejpam-5354	276	7	gradient	gradient	NOUN
ejpam-5354	276	8	method	method	NOUN
ejpam-5354	276	9	for	for	ADP
ejpam-5354	276	10	training	training	NOUN
ejpam-5354	276	11	feed	feed	NOUN
ejpam-5354	276	12	forward	forward	ADV
ejpam-5354	276	13	neural	neural	ADJ
ejpam-5354	276	14	network	network	NOUN
ejpam-5354	276	15	.	.	PUNCT
ejpam-5354	277	1	math	math	NOUN
ejpam-5354	277	2	.	.	PUNCT
ejpam-5354	278	1	model	model	PROPN
ejpam-5354	278	2	.	.	PUNCT
ejpam-5354	279	1	comput	comput	NOUN
ejpam-5354	279	2	.	.	PUNCT
ejpam-5354	279	3	,	,	PUNCT
ejpam-5354	279	4	2023	2023	NUM
ejpam-5354	279	5	.	.	PUNCT
ejpam-5354	280	1	[	[	X
ejpam-5354	280	2	24	24	NUM
ejpam-5354	280	3	]	]	X
ejpam-5354	280	4	y.	y.	NOUN
ejpam-5354	280	5	a.	a.	NOUN
ejpam-5354	280	6	laylani	laylani	PROPN
ejpam-5354	280	7	,	,	PUNCT
ejpam-5354	280	8	k.	k.	PROPN
ejpam-5354	280	9	k.	k.	PROPN
ejpam-5354	280	10	abbo	abbo	PROPN
ejpam-5354	280	11	,	,	PUNCT
ejpam-5354	280	12	and	and	CCONJ
ejpam-5354	280	13	h.	h.	PROPN
ejpam-5354	280	14	m.	m.	PROPN
ejpam-5354	280	15	khudhur	khudhur	PROPN
ejpam-5354	280	16	.	.	PUNCT
ejpam-5354	281	1	training	training	NOUN
ejpam-5354	281	2	feed	feed	NOUN
ejpam-5354	281	3	forward	forward	ADV
ejpam-5354	281	4	neural	neural	ADJ
ejpam-5354	281	5	network	network	NOUN
ejpam-5354	281	6	with	with	ADP
ejpam-5354	281	7	modified	modify	VERB
ejpam-5354	281	8	fletcher	fletcher	PROPN
ejpam-5354	281	9	-	-	PUNCT
ejpam-5354	281	10	reeves	reeves	PROPN
ejpam-5354	281	11	method	method	NOUN
ejpam-5354	281	12	.	.	PUNCT
ejpam-5354	282	1	j.	j.	PROPN
ejpam-5354	282	2	multidiscip	multidiscip	PROPN
ejpam-5354	282	3	.	.	PUNCT
ejpam-5354	283	1	model	model	PROPN
ejpam-5354	283	2	.	.	PUNCT
ejpam-5354	284	1	optim	optim	PROPN
ejpam-5354	284	2	.	.	PROPN
ejpam-5354	284	3	,	,	PUNCT
ejpam-5354	284	4	1:14–22	1:14–22	NOUN
ejpam-5354	284	5	,	,	PUNCT
ejpam-5354	284	6	2018	2018	NUM
ejpam-5354	284	7	.	.	PUNCT
ejpam-5354	285	1	references	reference	NOUN
ejpam-5354	285	2	2705	2705	NUM
ejpam-5354	285	3	[	[	X
ejpam-5354	285	4	25	25	NUM
ejpam-5354	285	5	]	]	X
ejpam-5354	285	6	b.	b.	PROPN
ejpam-5354	285	7	lerner	lerner	PROPN
ejpam-5354	285	8	,	,	PUNCT
ejpam-5354	285	9	h.	h.	PROPN
ejpam-5354	285	10	guterman	guterman	PROPN
ejpam-5354	285	11	,	,	PUNCT
ejpam-5354	285	12	m.	m.	NOUN
ejpam-5354	285	13	aladjem	aladjem	PROPN
ejpam-5354	285	14	,	,	PUNCT
ejpam-5354	285	15	and	and	CCONJ
ejpam-5354	285	16	i.	i.	PROPN
ejpam-5354	285	17	dinstein	dinstein	PROPN
ejpam-5354	285	18	.	.	PUNCT
ejpam-5354	286	1	a	a	DET
ejpam-5354	286	2	comparative	comparative	ADJ
ejpam-5354	286	3	study	study	NOUN
ejpam-5354	286	4	of	of	ADP
ejpam-5354	286	5	neural	neural	ADJ
ejpam-5354	286	6	network	network	NOUN
ejpam-5354	286	7	based	base	VERB
ejpam-5354	286	8	feature	feature	NOUN
ejpam-5354	286	9	extraction	extraction	NOUN
ejpam-5354	286	10	paradigms	paradigm	NOUN
ejpam-5354	286	11	.	.	PUNCT
ejpam-5354	287	1	pattern	pattern	NOUN
ejpam-5354	287	2	recognit	recognit	VERB
ejpam-5354	287	3	.	.	PUNCT
ejpam-5354	288	1	lett	lett	PROPN
ejpam-5354	288	2	.	.	PROPN
ejpam-5354	288	3	,	,	PUNCT
ejpam-5354	288	4	20:7–14	20:7–14	PROPN
ejpam-5354	288	5	,	,	PUNCT
ejpam-5354	288	6	1999	1999	NUM
ejpam-5354	288	7	.	.	PUNCT
ejpam-5354	289	1	[	[	X
ejpam-5354	289	2	26	26	NUM
ejpam-5354	289	3	]	]	X
ejpam-5354	289	4	y.	y.	PROPN
ejpam-5354	289	5	liu	liu	PROPN
ejpam-5354	289	6	and	and	CCONJ
ejpam-5354	289	7	c.	c.	PROPN
ejpam-5354	289	8	storey	storey	PROPN
ejpam-5354	289	9	.	.	PUNCT
ejpam-5354	290	1	efficient	efficient	ADJ
ejpam-5354	290	2	generalized	generalize	VERB
ejpam-5354	290	3	conjugate	conjugate	ADJ
ejpam-5354	290	4	gradient	gradient	ADJ
ejpam-5354	290	5	algorithms	algorithm	NOUN
ejpam-5354	290	6	,	,	PUNCT
ejpam-5354	290	7	part	part	NOUN
ejpam-5354	290	8	1	1	NUM
ejpam-5354	290	9	:	:	PUNCT
ejpam-5354	290	10	theory	theory	NOUN
ejpam-5354	290	11	.	.	PUNCT
ejpam-5354	291	1	j.	j.	PROPN
ejpam-5354	291	2	optim	optim	PROPN
ejpam-5354	291	3	.	.	PUNCT
ejpam-5354	292	1	theory	theory	NOUN
ejpam-5354	292	2	appl	appl	PROPN
ejpam-5354	292	3	.	.	PROPN
ejpam-5354	292	4	,	,	PUNCT
ejpam-5354	293	1	69:129–137	69:129–137	NUM
ejpam-5354	293	2	,	,	PUNCT
ejpam-5354	293	3	1991	1991	NUM
ejpam-5354	293	4	.	.	PUNCT
ejpam-5354	294	1	[	[	X
ejpam-5354	294	2	27	27	NUM
ejpam-5354	294	3	]	]	X
ejpam-5354	294	4	i.	i.	PROPN
ejpam-5354	294	5	e.	e.	PROPN
ejpam-5354	294	6	livieris	livieris	PROPN
ejpam-5354	294	7	and	and	CCONJ
ejpam-5354	294	8	p.	p.	NOUN
ejpam-5354	294	9	pintelas	pintela	NOUN
ejpam-5354	294	10	.	.	PUNCT
ejpam-5354	295	1	performance	performance	NOUN
ejpam-5354	295	2	evaluation	evaluation	NOUN
ejpam-5354	295	3	of	of	ADP
ejpam-5354	295	4	descent	descent	NOUN
ejpam-5354	295	5	cg	cg	NOUN
ejpam-5354	295	6	methods	method	NOUN
ejpam-5354	295	7	for	for	ADP
ejpam-5354	295	8	neural	neural	ADJ
ejpam-5354	295	9	network	network	NOUN
ejpam-5354	295	10	training	training	NOUN
ejpam-5354	295	11	.	.	PUNCT
ejpam-5354	296	1	hell	hell	INTJ
ejpam-5354	296	2	.	.	PUNCT
ejpam-5354	297	1	eur	eur	PROPN
ejpam-5354	297	2	.	.	PUNCT
ejpam-5354	298	1	res	re	NOUN
ejpam-5354	298	2	.	.	PUNCT
ejpam-5354	299	1	comput	comput	PROPN
ejpam-5354	299	2	.	.	PUNCT
ejpam-5354	300	1	math	math	NOUN
ejpam-5354	300	2	.	.	PUNCT
ejpam-5354	301	1	its	its	PRON
ejpam-5354	301	2	appl	appl	NOUN
ejpam-5354	301	3	.	.	PROPN
ejpam-5354	301	4	,	,	PUNCT
ejpam-5354	301	5	2009	2009	NUM
ejpam-5354	301	6	.	.	PUNCT
ejpam-5354	302	1	[	[	X
ejpam-5354	302	2	28	28	NUM
ejpam-5354	302	3	]	]	X
ejpam-5354	302	4	a.	a.	PROPN
ejpam-5354	302	5	perry	perry	PROPN
ejpam-5354	302	6	.	.	PUNCT
ejpam-5354	303	1	technical	technical	ADJ
ejpam-5354	303	2	note	note	NOUN
ejpam-5354	303	3	—	—	PUNCT
ejpam-5354	303	4	a	a	DET
ejpam-5354	303	5	modified	modify	VERB
ejpam-5354	303	6	conjugate	conjugate	ADJ
ejpam-5354	303	7	gradient	gradient	ADJ
ejpam-5354	303	8	algorithm	algorithm	NOUN
ejpam-5354	303	9	.	.	PUNCT
ejpam-5354	304	1	oper	oper	PROPN
ejpam-5354	304	2	.	.	PUNCT
ejpam-5354	304	3	res	res	PROPN
ejpam-5354	304	4	.	.	PROPN
ejpam-5354	304	5	,	,	PUNCT
ejpam-5354	304	6	26:1073–1078	26:1073–1078	NUM
ejpam-5354	304	7	,	,	PUNCT
ejpam-5354	304	8	1978	1978	NUM
ejpam-5354	304	9	.	.	PUNCT
ejpam-5354	305	1	[	[	X
ejpam-5354	305	2	29	29	NUM
ejpam-5354	305	3	]	]	X
ejpam-5354	305	4	b.	b.	PROPN
ejpam-5354	305	5	t.	t.	PROPN
ejpam-5354	305	6	polyak	polyak	PROPN
ejpam-5354	305	7	.	.	PUNCT
ejpam-5354	306	1	the	the	DET
ejpam-5354	306	2	conjugate	conjugate	ADJ
ejpam-5354	306	3	gradient	gradient	NOUN
ejpam-5354	306	4	method	method	NOUN
ejpam-5354	306	5	in	in	ADP
ejpam-5354	306	6	extremal	extremal	ADJ
ejpam-5354	306	7	problems	problem	NOUN
ejpam-5354	306	8	.	.	PUNCT
ejpam-5354	307	1	zh	zh	X
ejpam-5354	307	2	.	.	PUNCT
ejpam-5354	307	3	vychisl	vychisl	PROPN
ejpam-5354	307	4	.	.	PUNCT
ejpam-5354	308	1	mat	mat	NOUN
ejpam-5354	308	2	.	.	PUNCT
ejpam-5354	308	3	mat	mat	PROPN
ejpam-5354	308	4	.	.	PUNCT
ejpam-5354	308	5	fiz	fiz	PROPN
ejpam-5354	308	6	.	.	PROPN
ejpam-5354	308	7	,	,	PUNCT
ejpam-5354	308	8	9:807–821	9:807–821	PROPN
ejpam-5354	308	9	,	,	PUNCT
ejpam-5354	308	10	1969	1969	NUM
ejpam-5354	308	11	.	.	PUNCT
ejpam-5354	309	1	[	[	X
ejpam-5354	309	2	30	30	NUM
ejpam-5354	309	3	]	]	PUNCT
ejpam-5354	309	4	e.	e.	PROPN
ejpam-5354	309	5	p.	p.	PROPN
ejpam-5354	309	6	ribiere	ribiere	ADV
ejpam-5354	310	1	and	and	CCONJ
ejpam-5354	310	2	g.	g.	PROPN
ejpam-5354	310	3	note	note	VERB
ejpam-5354	310	4	sur	sur	PROPN
ejpam-5354	310	5	la	la	X
ejpam-5354	310	6	convergence	convergence	X
ejpam-5354	310	7	de	de	X
ejpam-5354	310	8	methods	method	NOUN
ejpam-5354	310	9	de	de	X
ejpam-5354	310	10	directions	direction	NOUN
ejpam-5354	310	11	conjuguees	conjuguee	NOUN
ejpam-5354	310	12	.	.	PUNCT
ejpam-5354	311	1	rev	rev	PROPN
ejpam-5354	311	2	.	.	PUNCT
ejpam-5354	312	1	fr	fr	PROPN
ejpam-5354	312	2	.	.	PUNCT
ejpam-5354	313	1	d’informatique	d’informatique	NOUN
ejpam-5354	313	2	rech	rech	NOUN
ejpam-5354	313	3	.	.	PUNCT
ejpam-5354	314	1	oper	oper	PROPN
ejpam-5354	314	2	.	.	PROPN
ejpam-5354	314	3	,	,	PUNCT
ejpam-5354	314	4	16:35–43	16:35–43	NUM
ejpam-5354	314	5	,	,	PUNCT
ejpam-5354	314	6	1969	1969	NUM
ejpam-5354	314	7	.	.	PUNCT
ejpam-5354	315	1	[	[	X
ejpam-5354	315	2	31	31	NUM
ejpam-5354	315	3	]	]	PUNCT
ejpam-5354	315	4	m.	m.	NOUN
ejpam-5354	315	5	rivaie	rivaie	PROPN
ejpam-5354	315	6	,	,	PUNCT
ejpam-5354	315	7	m.	m.	NOUN
ejpam-5354	315	8	mamat	mamat	PROPN
ejpam-5354	315	9	,	,	PUNCT
ejpam-5354	315	10	l.	l.	PROPN
ejpam-5354	315	11	w.	w.	PROPN
ejpam-5354	315	12	june	june	PROPN
ejpam-5354	315	13	,	,	PUNCT
ejpam-5354	315	14	and	and	CCONJ
ejpam-5354	315	15	i.	i.	PROPN
ejpam-5354	315	16	mohd	mohd	PROPN
ejpam-5354	315	17	.	.	PUNCT
ejpam-5354	316	1	a	a	DET
ejpam-5354	316	2	new	new	ADJ
ejpam-5354	316	3	conjugate	conjugate	ADJ
ejpam-5354	316	4	gradient	gradient	ADJ
ejpam-5354	316	5	coefficient	coefficient	NOUN
ejpam-5354	316	6	for	for	ADP
ejpam-5354	316	7	large	large	ADJ
ejpam-5354	316	8	scale	scale	NOUN
ejpam-5354	316	9	nonlinear	nonlinear	ADJ
ejpam-5354	316	10	unconstrained	unconstrained	ADJ
ejpam-5354	316	11	optimization	optimization	NOUN
ejpam-5354	316	12	.	.	PUNCT
ejpam-5354	317	1	int	int	NOUN
ejpam-5354	317	2	.	.	PUNCT
ejpam-5354	318	1	j.	j.	PROPN
ejpam-5354	318	2	math	math	PROPN
ejpam-5354	318	3	.	.	PUNCT
ejpam-5354	319	1	anal	anal	PROPN
ejpam-5354	319	2	.	.	PUNCT
ejpam-5354	319	3	,	,	PUNCT
ejpam-5354	319	4	6:1131–1146	6:1131–1146	NUM
ejpam-5354	319	5	,	,	PUNCT
ejpam-5354	319	6	2012	2012	NUM
ejpam-5354	319	7	.	.	PUNCT
ejpam-5354	320	1	[	[	X
ejpam-5354	320	2	32	32	NUM
ejpam-5354	320	3	]	]	PUNCT
ejpam-5354	320	4	d.	d.	PROPN
ejpam-5354	320	5	e.	e.	PROPN
ejpam-5354	320	6	rumelhart	rumelhart	PROPN
ejpam-5354	320	7	,	,	PUNCT
ejpam-5354	320	8	g.	g.	PROPN
ejpam-5354	320	9	e.	e.	PROPN
ejpam-5354	320	10	hinton	hinton	PROPN
ejpam-5354	320	11	,	,	PUNCT
ejpam-5354	320	12	and	and	CCONJ
ejpam-5354	320	13	r.	r.	PROPN
ejpam-5354	320	14	j.	j.	PROPN
ejpam-5354	320	15	williams	williams	PROPN
ejpam-5354	320	16	.	.	PUNCT
ejpam-5354	321	1	learning	learn	VERB
ejpam-5354	321	2	representations	representation	NOUN
ejpam-5354	321	3	by	by	ADP
ejpam-5354	321	4	back	back	ADV
ejpam-5354	321	5	-	-	PUNCT
ejpam-5354	321	6	propagating	propagate	VERB
ejpam-5354	321	7	errors	error	NOUN
ejpam-5354	321	8	.	.	PUNCT
ejpam-5354	322	1	nature	nature	NOUN
ejpam-5354	322	2	,	,	PUNCT
ejpam-5354	322	3	323:533–536	323:533–536	NUM
ejpam-5354	322	4	,	,	PUNCT
ejpam-5354	322	5	1986	1986	NUM
ejpam-5354	322	6	.	.	PUNCT
ejpam-5354	323	1	[	[	X
ejpam-5354	323	2	33	33	NUM
ejpam-5354	323	3	]	]	PUNCT
ejpam-5354	323	4	c.	c.	PROPN
ejpam-5354	323	5	j.	j.	PROPN
ejpam-5354	323	6	spoerer	spoerer	PROPN
ejpam-5354	323	7	,	,	PUNCT
ejpam-5354	323	8	t.	t.	PROPN
ejpam-5354	323	9	c.	c.	PROPN
ejpam-5354	323	10	kietzmann	kietzmann	PROPN
ejpam-5354	323	11	,	,	PUNCT
ejpam-5354	323	12	j.	j.	PROPN
ejpam-5354	323	13	mehrer	mehrer	PROPN
ejpam-5354	323	14	,	,	PUNCT
ejpam-5354	323	15	i.	i.	PROPN
ejpam-5354	323	16	charest	charest	PROPN
ejpam-5354	323	17	,	,	PUNCT
ejpam-5354	323	18	and	and	CCONJ
ejpam-5354	323	19	n.	n.	PROPN
ejpam-5354	323	20	kriegeskorte	kriegeskorte	PROPN
ejpam-5354	323	21	.	.	PUNCT
ejpam-5354	324	1	recurrent	recurrent	ADJ
ejpam-5354	324	2	neural	neural	ADJ
ejpam-5354	324	3	networks	network	NOUN
ejpam-5354	324	4	can	can	AUX
ejpam-5354	324	5	explain	explain	VERB
ejpam-5354	324	6	flexible	flexible	ADJ
ejpam-5354	324	7	trading	trading	NOUN
ejpam-5354	324	8	of	of	ADP
ejpam-5354	324	9	speed	speed	NOUN
ejpam-5354	324	10	and	and	CCONJ
ejpam-5354	324	11	accuracy	accuracy	NOUN
ejpam-5354	324	12	in	in	ADP
ejpam-5354	324	13	biological	biological	ADJ
ejpam-5354	324	14	vision	vision	NOUN
ejpam-5354	324	15	.	.	PUNCT
ejpam-5354	325	1	plos	plos	PROPN
ejpam-5354	325	2	comput	comput	PROPN
ejpam-5354	325	3	.	.	PUNCT
ejpam-5354	326	1	biol	biol	PROPN
ejpam-5354	326	2	.	.	PUNCT
ejpam-5354	326	3	,	,	PUNCT
ejpam-5354	326	4	16:1–27	16:1–27	NUM
ejpam-5354	326	5	,	,	PUNCT
ejpam-5354	326	6	2020	2020	NUM
ejpam-5354	326	7	.	.	PUNCT
ejpam-5354	327	1	[	[	X
ejpam-5354	327	2	34	34	NUM
ejpam-5354	327	3	]	]	PUNCT
ejpam-5354	327	4	x.	x.	NOUN
ejpam-5354	327	5	wang	wang	PROPN
ejpam-5354	327	6	,	,	PUNCT
ejpam-5354	327	7	g.	g.	PROPN
ejpam-5354	327	8	yuan	yuan	PROPN
ejpam-5354	327	9	,	,	PUNCT
ejpam-5354	327	10	and	and	CCONJ
ejpam-5354	327	11	l.	l.	PROPN
ejpam-5354	327	12	pang	pang	PROPN
ejpam-5354	327	13	.	.	PUNCT
ejpam-5354	328	1	a	a	DET
ejpam-5354	328	2	class	class	NOUN
ejpam-5354	328	3	of	of	ADP
ejpam-5354	328	4	new	new	ADJ
ejpam-5354	328	5	three	three	NUM
ejpam-5354	328	6	-	-	PUNCT
ejpam-5354	328	7	term	term	NOUN
ejpam-5354	328	8	descent	descent	NOUN
ejpam-5354	328	9	conjugate	conjugate	VERB
ejpam-5354	328	10	gradient	gradient	ADJ
ejpam-5354	328	11	algorithms	algorithm	NOUN
ejpam-5354	328	12	for	for	ADP
ejpam-5354	328	13	large	large	ADJ
ejpam-5354	328	14	-	-	PUNCT
ejpam-5354	328	15	scale	scale	NOUN
ejpam-5354	328	16	unconstrained	unconstrained	ADJ
ejpam-5354	328	17	optimization	optimization	NOUN
ejpam-5354	328	18	and	and	CCONJ
ejpam-5354	328	19	applications	application	NOUN
ejpam-5354	328	20	to	to	PART
ejpam-5354	328	21	image	image	NOUN
ejpam-5354	328	22	restoration	restoration	NOUN
ejpam-5354	328	23	problems	problem	NOUN
ejpam-5354	328	24	.	.	PUNCT
ejpam-5354	329	1	numer	numer	PROPN
ejpam-5354	329	2	.	.	PUNCT
ejpam-5354	330	1	algorithms	algorithms	PROPN
ejpam-5354	330	2	,	,	PUNCT
ejpam-5354	330	3	93	93	NUM
ejpam-5354	330	4	,	,	PUNCT
ejpam-5354	330	5	2022	2022	NUM
ejpam-5354	330	6	.	.	PUNCT
ejpam-5354	331	1	[	[	X
ejpam-5354	331	2	35	35	NUM
ejpam-5354	331	3	]	]	X
ejpam-5354	331	4	c.	c.	PROPN
ejpam-5354	331	5	h.	h.	PROPN
ejpam-5354	331	6	wu	wu	PROPN
ejpam-5354	331	7	.	.	PUNCT
ejpam-5354	332	1	gene	gene	NOUN
ejpam-5354	332	2	classification	classification	NOUN
ejpam-5354	332	3	artificial	artificial	ADJ
ejpam-5354	332	4	neural	neural	ADJ
ejpam-5354	332	5	system	system	NOUN
ejpam-5354	332	6	,	,	PUNCT
ejpam-5354	332	7	pages	page	NOUN
ejpam-5354	332	8	71–88	71–88	NUM
ejpam-5354	332	9	.	.	PUNCT
ejpam-5354	333	1	academic	academic	ADJ
ejpam-5354	333	2	press	press	NOUN
ejpam-5354	333	3	,	,	PUNCT
ejpam-5354	333	4	1996	1996	NUM
ejpam-5354	333	5	.	.	PUNCT
ejpam-5354	334	1	[	[	X
ejpam-5354	334	2	36	36	NUM
ejpam-5354	334	3	]	]	X
ejpam-5354	334	4	g.	g.	PROPN
ejpam-5354	334	5	yuan	yuan	PROPN
ejpam-5354	334	6	,	,	PUNCT
ejpam-5354	334	7	z.	z.	PROPN
ejpam-5354	334	8	wei	wei	PROPN
ejpam-5354	334	9	,	,	PUNCT
ejpam-5354	334	10	and	and	CCONJ
ejpam-5354	334	11	x.	x.	NOUN
ejpam-5354	334	12	lu	lu	PROPN
ejpam-5354	334	13	.	.	PUNCT
ejpam-5354	335	1	global	global	ADJ
ejpam-5354	335	2	convergence	convergence	NOUN
ejpam-5354	335	3	of	of	ADP
ejpam-5354	335	4	bfgs	bfgs	ADJ
ejpam-5354	335	5	and	and	CCONJ
ejpam-5354	335	6	prp	prp	NOUN
ejpam-5354	335	7	methods	method	NOUN
ejpam-5354	335	8	under	under	ADP
ejpam-5354	335	9	a	a	DET
ejpam-5354	335	10	modified	modify	VERB
ejpam-5354	335	11	weak	weak	ADJ
ejpam-5354	335	12	wolfe	wolfe	PROPN
ejpam-5354	335	13	–	–	PUNCT
ejpam-5354	335	14	powell	powell	PROPN
ejpam-5354	335	15	line	line	NOUN
ejpam-5354	335	16	search	search	NOUN
ejpam-5354	335	17	.	.	PUNCT
ejpam-5354	336	1	appl	appl	PROPN
ejpam-5354	336	2	.	.	PROPN
ejpam-5354	336	3	math	math	PROPN
ejpam-5354	336	4	.	.	PUNCT
ejpam-5354	337	1	model	model	PROPN
ejpam-5354	337	2	.	.	PROPN
ejpam-5354	338	1	,	,	PUNCT
ejpam-5354	338	2	47:811–825	47:811–825	PROPN
ejpam-5354	338	3	,	,	PUNCT
ejpam-5354	338	4	2017	2017	NUM
ejpam-5354	338	5	.	.	PUNCT
ejpam-5354	339	1	[	[	X
ejpam-5354	339	2	37	37	NUM
ejpam-5354	339	3	]	]	PUNCT
ejpam-5354	339	4	l.	l.	PROPN
ejpam-5354	339	5	zhang	zhang	PROPN
ejpam-5354	339	6	,	,	PUNCT
ejpam-5354	339	7	w.	w.	PROPN
ejpam-5354	339	8	zhou	zhou	PROPN
ejpam-5354	339	9	,	,	PUNCT
ejpam-5354	339	10	and	and	CCONJ
ejpam-5354	339	11	d.-h	d.-h	NOUN
ejpam-5354	339	12	.	.	PUNCT
ejpam-5354	340	1	li	li	PROPN
ejpam-5354	340	2	.	.	PUNCT
ejpam-5354	341	1	a	a	DET
ejpam-5354	341	2	descent	descent	NOUN
ejpam-5354	341	3	modified	modify	VERB
ejpam-5354	341	4	polak	polak	PROPN
ejpam-5354	341	5	–	–	PUNCT
ejpam-5354	341	6	ribiere	ribiere	X
ejpam-5354	341	7	–	–	PUNCT
ejpam-5354	341	8	polyak	polyak	NOUN
ejpam-5354	341	9	conjugate	conjugate	ADJ
ejpam-5354	341	10	gradient	gradient	NOUN
ejpam-5354	341	11	method	method	NOUN
ejpam-5354	341	12	and	and	CCONJ
ejpam-5354	341	13	its	its	PRON
ejpam-5354	341	14	global	global	ADJ
ejpam-5354	341	15	convergence	convergence	NOUN
ejpam-5354	341	16	.	.	PUNCT
ejpam-5354	342	1	i	i	PRON
ejpam-5354	342	2	m	m	VERB
ejpam-5354	342	3	a	a	PROPN
ejpam-5354	342	4	j.	j.	PROPN
ejpam-5354	342	5	numer	numer	PROPN
ejpam-5354	342	6	.	.	PUNCT
ejpam-5354	343	1	anal	anal	PROPN
ejpam-5354	343	2	.	.	PUNCT
ejpam-5354	343	3	,	,	PUNCT
ejpam-5354	344	1	26:629–640	26:629–640	NUM
ejpam-5354	344	2	,	,	PUNCT
ejpam-5354	344	3	2006	2006	NUM
ejpam-5354	344	4	.	.	PUNCT
ejpam-5354	345	1	[	[	X
ejpam-5354	345	2	38	38	NUM
ejpam-5354	345	3	]	]	PUNCT
ejpam-5354	345	4	g.	g.	PROPN
ejpam-5354	345	5	zoutendijk	zoutendijk	PROPN
ejpam-5354	345	6	.	.	PUNCT
ejpam-5354	346	1	nonlinear	nonlinear	ADJ
ejpam-5354	346	2	programming	programming	ADJ
ejpam-5354	346	3	computational	computational	ADJ
ejpam-5354	346	4	methods	method	NOUN
ejpam-5354	346	5	,	,	PUNCT
ejpam-5354	346	6	pages	page	NOUN
ejpam-5354	346	7	37–86	37–86	NUM
ejpam-5354	346	8	.	.	PUNCT
ejpam-5354	347	1	northholland	northholland	PROPN
ejpam-5354	347	2	,	,	PUNCT
ejpam-5354	347	3	amsterdam	amsterdam	PROPN
ejpam-5354	347	4	,	,	PUNCT
ejpam-5354	347	5	1970	1970	NUM
ejpam-5354	347	6	.	.	PUNCT