id	sid	tid	token	lemma	pos
ejpam-4568	1	1	european	european	PROPN
ejpam-4568	1	2	journal	journal	PROPN
ejpam-4568	1	3	of	of	ADP
ejpam-4568	1	4	pure	pure	ADJ
ejpam-4568	1	5	and	and	CCONJ
ejpam-4568	1	6	applied	apply	VERB
ejpam-4568	1	7	mathematics	mathematic	NOUN
ejpam-4568	1	8	vol	vol	NOUN
ejpam-4568	1	9	.	.	PROPN
ejpam-4568	2	1	15	15	NUM
ejpam-4568	2	2	,	,	PUNCT
ejpam-4568	2	3	no	no	INTJ
ejpam-4568	2	4	.	.	NOUN
ejpam-4568	2	5	4	4	NUM
ejpam-4568	2	6	,	,	PUNCT
ejpam-4568	2	7	2022	2022	NUM
ejpam-4568	2	8	,	,	PUNCT
ejpam-4568	2	9	2011	2011	NUM
ejpam-4568	2	10	-	-	SYM
ejpam-4568	2	11	2021	2021	NUM
ejpam-4568	2	12	issn	issn	PROPN
ejpam-4568	2	13	1307	1307	NUM
ejpam-4568	2	14	-	-	SYM
ejpam-4568	2	15	5543	5543	NUM
ejpam-4568	2	16	–	–	PUNCT
ejpam-4568	2	17	ejpam.com	ejpam.com	X
ejpam-4568	2	18	published	publish	VERB
ejpam-4568	2	19	by	by	ADP
ejpam-4568	2	20	new	new	PROPN
ejpam-4568	2	21	york	york	PROPN
ejpam-4568	2	22	business	business	PROPN
ejpam-4568	2	23	global	global	ADJ
ejpam-4568	2	24	efficient	efficient	ADJ
ejpam-4568	2	25	new	new	ADJ
ejpam-4568	2	26	conjugate	conjugate	ADJ
ejpam-4568	2	27	gradient	gradient	NOUN
ejpam-4568	2	28	methods	method	NOUN
ejpam-4568	2	29	for	for	ADP
ejpam-4568	2	30	removing	remove	VERB
ejpam-4568	2	31	impulse	impulse	ADJ
ejpam-4568	2	32	noise	noise	NOUN
ejpam-4568	2	33	images	image	NOUN
ejpam-4568	2	34	basim	basim	PROPN
ejpam-4568	2	35	a.	a.	PROPN
ejpam-4568	2	36	hassan1,∗	hassan1,∗	PROPN
ejpam-4568	2	37	,	,	PUNCT
ejpam-4568	2	38	hameed	hameed	NOUN
ejpam-4568	2	39	m.	m.	NOUN
ejpam-4568	2	40	sadiq1	sadiq1	PROPN
ejpam-4568	2	41	1	1	NUM
ejpam-4568	2	42	department	department	NOUN
ejpam-4568	2	43	of	of	ADP
ejpam-4568	2	44	mathematics	mathematic	NOUN
ejpam-4568	2	45	,	,	PUNCT
ejpam-4568	2	46	college	college	NOUN
ejpam-4568	2	47	of	of	ADP
ejpam-4568	2	48	computers	computer	NOUN
ejpam-4568	2	49	sciences	sciences	PROPN
ejpam-4568	2	50	and	and	CCONJ
ejpam-4568	2	51	mathematics	mathematic	NOUN
ejpam-4568	2	52	,	,	PUNCT
ejpam-4568	2	53	university	university	PROPN
ejpam-4568	2	54	of	of	ADP
ejpam-4568	2	55	mosul	mosul	PROPN
ejpam-4568	2	56	,	,	PUNCT
ejpam-4568	2	57	iraq	iraq	PROPN
ejpam-4568	2	58	abstract	abstract	NOUN
ejpam-4568	2	59	.	.	PUNCT
ejpam-4568	3	1	in	in	ADP
ejpam-4568	3	2	most	most	ADJ
ejpam-4568	3	3	applications	application	NOUN
ejpam-4568	3	4	,	,	PUNCT
ejpam-4568	3	5	denoising	denoise	VERB
ejpam-4568	3	6	image	image	NOUN
ejpam-4568	3	7	is	be	AUX
ejpam-4568	3	8	fundamental	fundamental	ADJ
ejpam-4568	3	9	to	to	ADP
ejpam-4568	3	10	subsequent	subsequent	ADJ
ejpam-4568	3	11	image	image	NOUN
ejpam-4568	3	12	processing	processing	NOUN
ejpam-4568	3	13	operations	operation	NOUN
ejpam-4568	3	14	.	.	PUNCT
ejpam-4568	4	1	in	in	ADP
ejpam-4568	4	2	this	this	DET
ejpam-4568	4	3	research	research	NOUN
ejpam-4568	4	4	,	,	PUNCT
ejpam-4568	4	5	we	we	PRON
ejpam-4568	4	6	derivation	derivation	VERB
ejpam-4568	4	7	a	a	DET
ejpam-4568	4	8	new	new	ADJ
ejpam-4568	4	9	formula	formula	NOUN
ejpam-4568	4	10	of	of	ADP
ejpam-4568	4	11	conjugate	conjugate	ADJ
ejpam-4568	4	12	gradient	gradient	NOUN
ejpam-4568	4	13	methods	method	NOUN
ejpam-4568	4	14	based	base	VERB
ejpam-4568	4	15	on	on	ADP
ejpam-4568	4	16	the	the	DET
ejpam-4568	4	17	quadratic	quadratic	ADJ
ejpam-4568	4	18	model	model	NOUN
ejpam-4568	4	19	.	.	PUNCT
ejpam-4568	5	1	the	the	DET
ejpam-4568	5	2	fact	fact	NOUN
ejpam-4568	5	3	that	that	SCONJ
ejpam-4568	5	4	the	the	DET
ejpam-4568	5	5	search	search	NOUN
ejpam-4568	5	6	direction	direction	NOUN
ejpam-4568	5	7	created	create	VERB
ejpam-4568	5	8	at	at	ADP
ejpam-4568	5	9	each	each	DET
ejpam-4568	5	10	iteration	iteration	NOUN
ejpam-4568	5	11	of	of	ADP
ejpam-4568	5	12	the	the	DET
ejpam-4568	5	13	proposed	propose	VERB
ejpam-4568	5	14	approach	approach	NOUN
ejpam-4568	5	15	is	be	AUX
ejpam-4568	5	16	descending	descend	VERB
ejpam-4568	5	17	and	and	CCONJ
ejpam-4568	5	18	independent	independent	ADJ
ejpam-4568	5	19	of	of	ADP
ejpam-4568	5	20	the	the	DET
ejpam-4568	5	21	line	line	NOUN
ejpam-4568	5	22	search	search	NOUN
ejpam-4568	5	23	makes	make	VERB
ejpam-4568	5	24	it	it	PRON
ejpam-4568	5	25	interesting	interesting	ADJ
ejpam-4568	5	26	.	.	PUNCT
ejpam-4568	6	1	the	the	DET
ejpam-4568	6	2	use	use	NOUN
ejpam-4568	6	3	of	of	ADP
ejpam-4568	6	4	wolfe	wolfe	PROPN
ejpam-4568	6	5	conditions	condition	NOUN
ejpam-4568	6	6	also	also	ADV
ejpam-4568	6	7	determines	determine	VERB
ejpam-4568	6	8	the	the	DET
ejpam-4568	6	9	global	global	ADJ
ejpam-4568	6	10	convergence	convergence	NOUN
ejpam-4568	6	11	of	of	ADP
ejpam-4568	6	12	the	the	DET
ejpam-4568	6	13	suggested	suggest	VERB
ejpam-4568	6	14	approach	approach	NOUN
ejpam-4568	6	15	.	.	PUNCT
ejpam-4568	7	1	to	to	PART
ejpam-4568	7	2	prove	prove	VERB
ejpam-4568	7	3	the	the	DET
ejpam-4568	7	4	viability	viability	NOUN
ejpam-4568	7	5	of	of	ADP
ejpam-4568	7	6	the	the	DET
ejpam-4568	7	7	suggested	suggest	VERB
ejpam-4568	7	8	approach	approach	NOUN
ejpam-4568	7	9	,	,	PUNCT
ejpam-4568	7	10	comparison	comparison	NOUN
ejpam-4568	7	11	tests	test	VERB
ejpam-4568	7	12	on	on	ADP
ejpam-4568	7	13	impulse	impulse	ADJ
ejpam-4568	7	14	noise	noise	NOUN
ejpam-4568	7	15	reduction	reduction	NOUN
ejpam-4568	7	16	are	be	AUX
ejpam-4568	7	17	given	give	VERB
ejpam-4568	7	18	.	.	PUNCT
ejpam-4568	8	1	2020	2020	NUM
ejpam-4568	8	2	mathematics	mathematic	NOUN
ejpam-4568	8	3	subject	subject	NOUN
ejpam-4568	8	4	classifications	classification	NOUN
ejpam-4568	8	5	:	:	PUNCT
ejpam-4568	8	6	90c30	90c30	NUM
ejpam-4568	8	7	,	,	PUNCT
ejpam-4568	8	8	65k05	65k05	NUM
ejpam-4568	8	9	,	,	PUNCT
ejpam-4568	8	10	49m37	49m37	NUM
ejpam-4568	8	11	key	key	ADJ
ejpam-4568	8	12	words	word	NOUN
ejpam-4568	8	13	and	and	CCONJ
ejpam-4568	8	14	phrases	phrase	NOUN
ejpam-4568	8	15	:	:	PUNCT
ejpam-4568	8	16	optimization	optimization	NOUN
ejpam-4568	8	17	,	,	PUNCT
ejpam-4568	8	18	impulse	impulse	ADJ
ejpam-4568	8	19	noise	noise	NOUN
ejpam-4568	8	20	,	,	PUNCT
ejpam-4568	8	21	conjugate	conjugate	ADJ
ejpam-4568	8	22	gradient	gradient	NOUN
ejpam-4568	8	23	method	method	NOUN
ejpam-4568	8	24	,	,	PUNCT
ejpam-4568	8	25	global	global	ADJ
ejpam-4568	8	26	convergence	convergence	NOUN
ejpam-4568	8	27	1	1	NUM
ejpam-4568	8	28	.	.	PUNCT
ejpam-4568	9	1	introduction	introduction	NOUN
ejpam-4568	9	2	image	image	NOUN
ejpam-4568	9	3	denoising	denoising	NOUN
ejpam-4568	9	4	is	be	AUX
ejpam-4568	9	5	a	a	DET
ejpam-4568	9	6	fundamental	fundamental	ADJ
ejpam-4568	9	7	problem	problem	NOUN
ejpam-4568	9	8	in	in	ADP
ejpam-4568	9	9	image	image	NOUN
ejpam-4568	9	10	processing	processing	NOUN
ejpam-4568	9	11	operations	operation	NOUN
ejpam-4568	9	12	.	.	PUNCT
ejpam-4568	10	1	one	one	NUM
ejpam-4568	10	2	of	of	ADP
ejpam-4568	10	3	its	its	PRON
ejpam-4568	10	4	more	more	ADV
ejpam-4568	10	5	investigated	investigate	VERB
ejpam-4568	10	6	domains	domain	NOUN
ejpam-4568	10	7	is	be	AUX
ejpam-4568	10	8	image	image	NOUN
ejpam-4568	10	9	denoising	denoising	NOUN
ejpam-4568	10	10	which	which	PRON
ejpam-4568	10	11	plays	play	VERB
ejpam-4568	10	12	an	an	DET
ejpam-4568	10	13	adequate	adequate	ADJ
ejpam-4568	10	14	contribution	contribution	NOUN
ejpam-4568	10	15	in	in	ADP
ejpam-4568	10	16	many	many	ADJ
ejpam-4568	10	17	applications	application	NOUN
ejpam-4568	10	18	.	.	PUNCT
ejpam-4568	11	1	there	there	PRON
ejpam-4568	11	2	are	be	VERB
ejpam-4568	11	3	two	two	NUM
ejpam-4568	11	4	main	main	ADJ
ejpam-4568	11	5	models	model	NOUN
ejpam-4568	11	6	to	to	PART
ejpam-4568	11	7	represent	represent	VERB
ejpam-4568	11	8	impulse	impulse	ADJ
ejpam-4568	11	9	noise	noise	NOUN
ejpam-4568	11	10	[	[	X
ejpam-4568	11	11	17	17	NUM
ejpam-4568	11	12	]	]	PUNCT
ejpam-4568	11	13	.	.	PUNCT
ejpam-4568	12	1	one	one	NUM
ejpam-4568	12	2	type	type	NOUN
ejpam-4568	12	3	of	of	ADP
ejpam-4568	12	4	noise	noise	NOUN
ejpam-4568	12	5	is	be	AUX
ejpam-4568	12	6	known	know	VERB
ejpam-4568	12	7	as	as	ADP
ejpam-4568	12	8	salt	salt	NOUN
ejpam-4568	12	9	-	-	PUNCT
ejpam-4568	12	10	and	and	CCONJ
ejpam-4568	12	11	-	-	PUNCT
ejpam-4568	12	12	pepper	pepper	NOUN
ejpam-4568	12	13	noise	noise	NOUN
ejpam-4568	12	14	,	,	PUNCT
ejpam-4568	12	15	in	in	ADP
ejpam-4568	12	16	which	which	PRON
ejpam-4568	12	17	the	the	DET
ejpam-4568	12	18	noisy	noisy	ADJ
ejpam-4568	12	19	pixels	pixel	NOUN
ejpam-4568	12	20	can	can	AUX
ejpam-4568	12	21	only	only	ADV
ejpam-4568	12	22	accept	accept	VERB
ejpam-4568	12	23	the	the	DET
ejpam-4568	12	24	maximum	maximum	ADJ
ejpam-4568	12	25	and	and	CCONJ
ejpam-4568	12	26	minimum	minimum	ADJ
ejpam-4568	12	27	pixel	pixel	NOUN
ejpam-4568	12	28	values	value	NOUN
ejpam-4568	12	29	possible	possible	ADJ
ejpam-4568	12	30	within	within	ADP
ejpam-4568	12	31	the	the	DET
ejpam-4568	12	32	dynamic	dynamic	ADJ
ejpam-4568	12	33	range	range	NOUN
ejpam-4568	12	34	of	of	ADP
ejpam-4568	12	35	the	the	DET
ejpam-4568	12	36	source	source	NOUN
ejpam-4568	12	37	image	image	NOUN
ejpam-4568	12	38	.	.	PUNCT
ejpam-4568	13	1	the	the	DET
ejpam-4568	13	2	random	random	ADV
ejpam-4568	13	3	-	-	PUNCT
ejpam-4568	13	4	valued	value	VERB
ejpam-4568	13	5	noise	noise	NOUN
ejpam-4568	13	6	,	,	PUNCT
ejpam-4568	13	7	which	which	PRON
ejpam-4568	13	8	may	may	AUX
ejpam-4568	13	9	have	have	VERB
ejpam-4568	13	10	any	any	DET
ejpam-4568	13	11	random	random	ADJ
ejpam-4568	13	12	value	value	NOUN
ejpam-4568	13	13	between	between	ADP
ejpam-4568	13	14	the	the	DET
ejpam-4568	13	15	maximum	maximum	ADJ
ejpam-4568	13	16	and	and	CCONJ
ejpam-4568	13	17	minimum	minimum	ADJ
ejpam-4568	13	18	pixel	pixel	NOUN
ejpam-4568	13	19	values	value	NOUN
ejpam-4568	13	20	of	of	ADP
ejpam-4568	13	21	dynamic	dynamic	ADJ
ejpam-4568	13	22	range	range	NOUN
ejpam-4568	13	23	,	,	PUNCT
ejpam-4568	13	24	is	be	AUX
ejpam-4568	13	25	another	another	DET
ejpam-4568	13	26	type	type	NOUN
ejpam-4568	13	27	of	of	ADP
ejpam-4568	13	28	impulsive	impulsive	ADJ
ejpam-4568	13	29	noise	noise	NOUN
ejpam-4568	13	30	[	[	X
ejpam-4568	13	31	18	18	NUM
ejpam-4568	13	32	]	]	PUNCT
ejpam-4568	13	33	.	.	PUNCT
ejpam-4568	14	1	one	one	NUM
ejpam-4568	14	2	of	of	ADP
ejpam-4568	14	3	the	the	DET
ejpam-4568	14	4	most	most	ADV
ejpam-4568	14	5	significant	significant	ADJ
ejpam-4568	14	6	issues	issue	NOUN
ejpam-4568	14	7	in	in	ADP
ejpam-4568	14	8	picture	picture	NOUN
ejpam-4568	14	9	processing	processing	NOUN
ejpam-4568	14	10	is	be	AUX
ejpam-4568	14	11	the	the	DET
ejpam-4568	14	12	removal	removal	NOUN
ejpam-4568	14	13	of	of	ADP
ejpam-4568	14	14	above	above	ADP
ejpam-4568	14	15	both	both	DET
ejpam-4568	14	16	noise	noise	NOUN
ejpam-4568	14	17	.	.	PUNCT
ejpam-4568	15	1	the	the	DET
ejpam-4568	15	2	average	average	ADJ
ejpam-4568	15	3	filter	filter	NOUN
ejpam-4568	15	4	and	and	CCONJ
ejpam-4568	15	5	its	its	PRON
ejpam-4568	15	6	variations	variation	NOUN
ejpam-4568	15	7	[	[	X
ejpam-4568	15	8	7	7	X
ejpam-4568	15	9	]	]	PUNCT
ejpam-4568	15	10	may	may	AUX
ejpam-4568	15	11	find	find	VERB
ejpam-4568	15	12	the	the	DET
ejpam-4568	15	13	noisy	noisy	ADJ
ejpam-4568	15	14	pixels	pixel	NOUN
ejpam-4568	15	15	but	but	CCONJ
ejpam-4568	15	16	return	return	VERB
ejpam-4568	15	17	them	they	PRON
ejpam-4568	15	18	badly	badly	ADV
ejpam-4568	15	19	when	when	SCONJ
ejpam-4568	15	20	the	the	DET
ejpam-4568	15	21	noise	noise	NOUN
ejpam-4568	15	22	ratio	ratio	NOUN
ejpam-4568	15	23	is	be	AUX
ejpam-4568	15	24	large	large	ADJ
ejpam-4568	15	25	.	.	PUNCT
ejpam-4568	16	1	these	these	DET
ejpam-4568	16	2	two	two	NUM
ejpam-4568	16	3	approaches	approach	NOUN
ejpam-4568	16	4	are	be	AUX
ejpam-4568	16	5	quite	quite	ADV
ejpam-4568	16	6	common	common	ADJ
ejpam-4568	16	7	for	for	ADP
ejpam-4568	16	8	this	this	DET
ejpam-4568	16	9	purpose	purpose	NOUN
ejpam-4568	16	10	.	.	PUNCT
ejpam-4568	17	1	unaltered	unaltered	ADJ
ejpam-4568	17	2	gray	gray	ADJ
ejpam-4568	17	3	levels	level	NOUN
ejpam-4568	17	4	exist	exist	VERB
ejpam-4568	17	5	in	in	ADP
ejpam-4568	17	6	unharmed	unharmed	ADJ
ejpam-4568	17	7	pixels	pixel	NOUN
ejpam-4568	17	8	.	.	PUNCT
ejpam-4568	18	1	while	while	SCONJ
ejpam-4568	18	2	the	the	DET
ejpam-4568	18	3	variational	variational	ADJ
ejpam-4568	18	4	technique	technique	NOUN
ejpam-4568	18	5	is	be	AUX
ejpam-4568	18	6	capable	capable	ADJ
ejpam-4568	18	7	of	of	ADP
ejpam-4568	18	8	keeping	keep	VERB
ejpam-4568	18	9	the	the	DET
ejpam-4568	18	10	features	feature	NOUN
ejpam-4568	18	11	and	and	CCONJ
ejpam-4568	18	12	edges	edge	NOUN
ejpam-4568	18	13	well	well	ADV
ejpam-4568	18	14	,	,	PUNCT
ejpam-4568	18	15	every	every	DET
ejpam-4568	18	16	pixel	pixel	PROPN
ejpam-4568	18	17	’s	’s	PART
ejpam-4568	18	18	gray	gray	ADJ
ejpam-4568	18	19	level	level	NOUN
ejpam-4568	18	20	,	,	PUNCT
ejpam-4568	18	21	including	include	VERB
ejpam-4568	18	22	those	those	PRON
ejpam-4568	18	23	that	that	PRON
ejpam-4568	18	24	are	be	AUX
ejpam-4568	18	25	n’t	not	PART
ejpam-4568	18	26	damaged	damage	VERB
ejpam-4568	18	27	,	,	PUNCT
ejpam-4568	18	28	is	be	AUX
ejpam-4568	18	29	altered	alter	VERB
ejpam-4568	18	30	,	,	PUNCT
ejpam-4568	18	31	see	see	VERB
ejpam-4568	18	32	[	[	X
ejpam-4568	18	33	19	19	NUM
ejpam-4568	18	34	]	]	PUNCT
ejpam-4568	18	35	.	.	PUNCT
ejpam-4568	19	1	the	the	DET
ejpam-4568	19	2	recovered	recovered	ADJ
ejpam-4568	19	3	picture	picture	NOUN
ejpam-4568	19	4	may	may	AUX
ejpam-4568	19	5	also	also	ADV
ejpam-4568	19	6	lose	lose	VERB
ejpam-4568	19	7	its	its	PRON
ejpam-4568	19	8	details	detail	NOUN
ejpam-4568	19	9	and	and	CCONJ
ejpam-4568	19	10	become	become	VERB
ejpam-4568	19	11	distorted	distort	VERB
ejpam-4568	19	12	.	.	PUNCT
ejpam-4568	20	1	∗corresponding	∗corresponde	VERB
ejpam-4568	20	2	author	author	NOUN
ejpam-4568	20	3	.	.	PUNCT
ejpam-4568	21	1	doi	doi	NOUN
ejpam-4568	21	2	:	:	PUNCT
ejpam-4568	21	3	https://doi.org/10.29020/nybg.ejpam.v15i4.4568	https://doi.org/10.29020/nybg.ejpam.v15i4.4568	PROPN
ejpam-4568	21	4	email	email	NOUN
ejpam-4568	21	5	addresses	address	NOUN
ejpam-4568	21	6	:	:	PUNCT
ejpam-4568	21	7	basimah@uomosul.edu.iq	basimah@uomosul.edu.iq	PROPN
ejpam-4568	21	8	(	(	PUNCT
ejpam-4568	21	9	b.	b.	PROPN
ejpam-4568	21	10	a.	a.	PROPN
ejpam-4568	21	11	hassan	hassan	PROPN
ejpam-4568	21	12	)	)	PUNCT
ejpam-4568	21	13	,	,	PUNCT
ejpam-4568	21	14	hameedsadiq1975@gmail.com	hameedsadiq1975@gmail.com	X
ejpam-4568	21	15	(	(	PUNCT
ejpam-4568	21	16	h.	h.	PROPN
ejpam-4568	21	17	m.	m.	PROPN
ejpam-4568	21	18	sadiq	sadiq	PROPN
ejpam-4568	21	19	)	)	PUNCT
ejpam-4568	21	20	https://www.ejpam.com	https://www.ejpam.com	NOUN
ejpam-4568	21	21	2011	2011	NUM
ejpam-4568	22	1	©	©	PROPN
ejpam-4568	22	2	2022	2022	NUM
ejpam-4568	22	3	ejpam	ejpam	VERB
ejpam-4568	22	4	all	all	DET
ejpam-4568	22	5	rights	right	NOUN
ejpam-4568	22	6	reserved	reserve	VERB
ejpam-4568	22	7	.	.	PUNCT
ejpam-4568	23	1	b.	b.	PROPN
ejpam-4568	23	2	a.	a.	PROPN
ejpam-4568	23	3	hassan	hassan	PROPN
ejpam-4568	23	4	,	,	PUNCT
ejpam-4568	23	5	h.	h.	PROPN
ejpam-4568	23	6	m.	m.	PROPN
ejpam-4568	23	7	sadiq	sadiq	PROPN
ejpam-4568	23	8	/	/	SYM
ejpam-4568	23	9	eur	eur	PROPN
ejpam-4568	23	10	.	.	PUNCT
ejpam-4568	24	1	j.	j.	PROPN
ejpam-4568	24	2	pure	pure	PROPN
ejpam-4568	24	3	appl	appl	PROPN
ejpam-4568	24	4	.	.	PROPN
ejpam-4568	24	5	math	math	PROPN
ejpam-4568	24	6	,	,	PUNCT
ejpam-4568	24	7	15	15	NUM
ejpam-4568	24	8	(	(	PUNCT
ejpam-4568	24	9	4	4	NUM
ejpam-4568	24	10	)	)	PUNCT
ejpam-4568	24	11	(	(	PUNCT
ejpam-4568	24	12	2022	2022	NUM
ejpam-4568	24	13	)	)	PUNCT
ejpam-4568	24	14	,	,	PUNCT
ejpam-4568	24	15	2011	2011	NUM
ejpam-4568	24	16	-	-	SYM
ejpam-4568	24	17	2021	2021	NUM
ejpam-4568	24	18	2012	2012	NUM
ejpam-4568	24	19	the	the	DET
ejpam-4568	24	20	two	two	NUM
ejpam-4568	24	21	-	-	PUNCT
ejpam-4568	24	22	phase	phase	NOUN
ejpam-4568	24	23	method	method	NOUN
ejpam-4568	24	24	in	in	ADP
ejpam-4568	24	25	[	[	X
ejpam-4568	24	26	6	6	NUM
ejpam-4568	24	27	]	]	PUNCT
ejpam-4568	24	28	unify	unify	VERB
ejpam-4568	24	29	the	the	DET
ejpam-4568	24	30	advantages	advantage	NOUN
ejpam-4568	24	31	of	of	ADP
ejpam-4568	24	32	both	both	DET
ejpam-4568	24	33	methods	method	NOUN
ejpam-4568	24	34	and	and	CCONJ
ejpam-4568	24	35	we	we	PRON
ejpam-4568	24	36	will	will	AUX
ejpam-4568	24	37	give	give	VERB
ejpam-4568	24	38	a	a	DET
ejpam-4568	24	39	brief	brief	ADJ
ejpam-4568	24	40	description	description	NOUN
ejpam-4568	24	41	here	here	ADV
ejpam-4568	24	42	.	.	PUNCT
ejpam-4568	25	1	in	in	ADP
ejpam-4568	25	2	the	the	DET
ejpam-4568	25	3	following	following	NOUN
ejpam-4568	25	4	,	,	PUNCT
ejpam-4568	25	5	we	we	PRON
ejpam-4568	25	6	let	let	VERB
ejpam-4568	25	7	xi	xi	PROPN
ejpam-4568	25	8	,	,	PUNCT
ejpam-4568	25	9	j	j	PROPN
ejpam-4568	25	10	for	for	ADP
ejpam-4568	25	11	(	(	PUNCT
ejpam-4568	25	12	i	i	PROPN
ejpam-4568	25	13	,	,	PUNCT
ejpam-4568	25	14	j	j	PROPN
ejpam-4568	25	15	)	)	PUNCT
ejpam-4568	25	16	∈	∈	PROPN
ejpam-4568	26	1	a	a	DET
ejpam-4568	26	2	=	=	SYM
ejpam-4568	26	3	1	1	NUM
ejpam-4568	26	4	,	,	PUNCT
ejpam-4568	26	5	2	2	NUM
ejpam-4568	26	6	,	,	PUNCT
ejpam-4568	26	7	3	3	NUM
ejpam-4568	26	8	,	,	PUNCT
ejpam-4568	26	9	...	...	PUNCT
ejpam-4568	26	10	,	,	PUNCT
ejpam-4568	26	11	m	m	VERB
ejpam-4568	26	12	×	×	NOUN
ejpam-4568	26	13	1	1	NUM
ejpam-4568	26	14	,	,	PUNCT
ejpam-4568	26	15	2	2	NUM
ejpam-4568	26	16	,	,	PUNCT
ejpam-4568	26	17	3	3	NUM
ejpam-4568	26	18	,	,	PUNCT
ejpam-4568	26	19	...	...	PUNCT
ejpam-4568	26	20	,	,	PUNCT
ejpam-4568	26	21	n	n	AUX
ejpam-4568	26	22	,	,	PUNCT
ejpam-4568	26	23	be	be	AUX
ejpam-4568	26	24	the	the	DET
ejpam-4568	26	25	gray	gray	ADJ
ejpam-4568	26	26	level	level	NOUN
ejpam-4568	26	27	of	of	ADP
ejpam-4568	26	28	a	a	DET
ejpam-4568	26	29	true	true	ADJ
ejpam-4568	26	30	m	m	NOUN
ejpam-4568	26	31	by	by	ADP
ejpam-4568	26	32	n.	n.	NOUN
ejpam-4568	26	33	let	let	VERB
ejpam-4568	26	34	the	the	DET
ejpam-4568	26	35	set	set	NOUN
ejpam-4568	26	36	of	of	ADP
ejpam-4568	26	37	indices	index	NOUN
ejpam-4568	26	38	of	of	ADP
ejpam-4568	26	39	the	the	DET
ejpam-4568	26	40	noisy	noisy	ADJ
ejpam-4568	26	41	pixels	pixel	NOUN
ejpam-4568	26	42	uncover	uncover	NOUN
ejpam-4568	26	43	in	in	ADP
ejpam-4568	26	44	the	the	DET
ejpam-4568	26	45	first	first	ADJ
ejpam-4568	26	46	phase	phase	NOUN
ejpam-4568	26	47	denote	denote	VERB
ejpam-4568	26	48	by	by	ADP
ejpam-4568	26	49	n	n	CCONJ
ejpam-4568	26	50	,	,	PUNCT
ejpam-4568	26	51	where	where	SCONJ
ejpam-4568	26	52	n	n	X
ejpam-4568	26	53	⊂	⊂	PROPN
ejpam-4568	26	54	a.	a.	NOUN
ejpam-4568	26	55	let	let	VERB
ejpam-4568	26	56	ui	ui	PROPN
ejpam-4568	26	57	,	,	PUNCT
ejpam-4568	26	58	j	j	PROPN
ejpam-4568	27	1	=	=	PUNCT
ejpam-4568	28	1	[	[	X
ejpam-4568	28	2	ui	ui	PROPN
ejpam-4568	28	3	,	,	PUNCT
ejpam-4568	28	4	j	j	PROPN
ejpam-4568	28	5	]	]	X
ejpam-4568	28	6	(	(	PUNCT
ejpam-4568	28	7	i	i	PRON
ejpam-4568	28	8	,	,	PUNCT
ejpam-4568	28	9	j)∈n	j)∈n	PROPN
ejpam-4568	28	10	be	be	AUX
ejpam-4568	28	11	a	a	DET
ejpam-4568	28	12	column	column	NOUN
ejpam-4568	28	13	vector	vector	NOUN
ejpam-4568	28	14	of	of	ADP
ejpam-4568	28	15	length	length	NOUN
ejpam-4568	28	16	c	c	PROPN
ejpam-4568	28	17	ordered	order	VERB
ejpam-4568	28	18	lexicographically	lexicographically	ADV
ejpam-4568	28	19	(	(	PUNCT
ejpam-4568	28	20	c	c	NOUN
ejpam-4568	28	21	is	be	AUX
ejpam-4568	28	22	the	the	DET
ejpam-4568	28	23	number	number	NOUN
ejpam-4568	28	24	of	of	ADP
ejpam-4568	28	25	elements	element	NOUN
ejpam-4568	28	26	of	of	ADP
ejpam-4568	28	27	n	n	CCONJ
ejpam-4568	28	28	)	)	PUNCT
ejpam-4568	28	29	,	,	PUNCT
ejpam-4568	28	30	and	and	CCONJ
ejpam-4568	28	31	yi	yi	PROPN
ejpam-4568	28	32	,	,	PUNCT
ejpam-4568	28	33	j	j	PROPN
ejpam-4568	28	34	denote	denote	VERB
ejpam-4568	28	35	the	the	DET
ejpam-4568	28	36	observed	observed	ADJ
ejpam-4568	28	37	pixel	pixel	NOUN
ejpam-4568	28	38	value	value	NOUN
ejpam-4568	28	39	at	at	ADP
ejpam-4568	28	40	position	position	NOUN
ejpam-4568	28	41	(	(	PUNCT
ejpam-4568	28	42	i	i	PROPN
ejpam-4568	28	43	,	,	PUNCT
ejpam-4568	28	44	j	j	PROPN
ejpam-4568	28	45	)	)	PUNCT
ejpam-4568	28	46	.	.	PUNCT
ejpam-4568	29	1	then	then	ADV
ejpam-4568	29	2	,	,	PUNCT
ejpam-4568	29	3	the	the	DET
ejpam-4568	29	4	second	second	ADJ
ejpam-4568	29	5	phase	phase	NOUN
ejpam-4568	29	6	it	it	PRON
ejpam-4568	29	7	to	to	PART
ejpam-4568	29	8	recover	recover	VERB
ejpam-4568	29	9	the	the	DET
ejpam-4568	29	10	noisy	noisy	ADJ
ejpam-4568	29	11	pixels	pixel	NOUN
ejpam-4568	29	12	by	by	ADP
ejpam-4568	29	13	minimizing	minimize	VERB
ejpam-4568	29	14	the	the	DET
ejpam-4568	29	15	following	follow	VERB
ejpam-4568	29	16	edge	edge	NOUN
ejpam-4568	29	17	-	-	PUNCT
ejpam-4568	29	18	preserving	preserve	VERB
ejpam-4568	29	19	regularization	regularization	NOUN
ejpam-4568	29	20	function	function	NOUN
ejpam-4568	29	21	:	:	PUNCT
ejpam-4568	29	22	fα(u	fα(u	X
ejpam-4568	29	23	)	)	PUNCT
ejpam-4568	29	24	=	=	PUNCT
ejpam-4568	29	25	∑	∑	PUNCT
ejpam-4568	29	26	(	(	PUNCT
ejpam-4568	29	27	i	i	PROPN
ejpam-4568	29	28	,	,	PUNCT
ejpam-4568	29	29	j)∈n	j)∈n	PROPN
ejpam-4568	29	30	[	[	PUNCT
ejpam-4568	29	31	|ui	|ui	X
ejpam-4568	29	32	,	,	PUNCT
ejpam-4568	29	33	j	j	PROPN
ejpam-4568	29	34	−	−	PROPN
ejpam-4568	29	35	yi	yi	PROPN
ejpam-4568	29	36	,	,	PUNCT
ejpam-4568	29	37	j	j	PROPN
ejpam-4568	29	38	|+	|+	NOUN
ejpam-4568	29	39	β	β	X
ejpam-4568	29	40	2	2	NUM
ejpam-4568	29	41	(	(	PUNCT
ejpam-4568	29	42	2×	2×	NUM
ejpam-4568	29	43	s1	s1	PROPN
ejpam-4568	29	44	i	i	PROPN
ejpam-4568	29	45	,	,	PUNCT
ejpam-4568	29	46	j	j	PROPN
ejpam-4568	29	47	+	+	CCONJ
ejpam-4568	29	48	s2	s2	VERB
ejpam-4568	29	49	i	i	PROPN
ejpam-4568	29	50	,	,	PUNCT
ejpam-4568	29	51	j	j	PROPN
ejpam-4568	29	52	)	)	PUNCT
ejpam-4568	29	53	]	]	PUNCT
ejpam-4568	30	1	(	(	PUNCT
ejpam-4568	30	2	1	1	X
ejpam-4568	30	3	)	)	PUNCT
ejpam-4568	30	4	where	where	SCONJ
ejpam-4568	30	5	s1	s1	PROPN
ejpam-4568	30	6	i	i	PROPN
ejpam-4568	30	7	,	,	PUNCT
ejpam-4568	30	8	j	j	PROPN
ejpam-4568	30	9	=	=	SYM
ejpam-4568	30	10	2	2	NUM
ejpam-4568	30	11	∑	∑	PUNCT
ejpam-4568	30	12	(	(	PUNCT
ejpam-4568	30	13	m	m	PROPN
ejpam-4568	30	14	,	,	PUNCT
ejpam-4568	30	15	n)∈pi	n)∈pi	PROPN
ejpam-4568	30	16	,	,	PUNCT
ejpam-4568	30	17	j∩nc	j∩nc	PROPN
ejpam-4568	30	18	φα(ui	φα(ui	PROPN
ejpam-4568	30	19	,	,	PUNCT
ejpam-4568	30	20	j	j	PROPN
ejpam-4568	30	21	−	−	PROPN
ejpam-4568	30	22	ym	ym	PROPN
ejpam-4568	30	23	,	,	PUNCT
ejpam-4568	30	24	n	n	CCONJ
ejpam-4568	30	25	)	)	PUNCT
ejpam-4568	30	26	,	,	PUNCT
ejpam-4568	30	27	s	s	VERB
ejpam-4568	30	28	2	2	NUM
ejpam-4568	30	29	i	i	PROPN
ejpam-4568	30	30	,	,	PUNCT
ejpam-4568	30	31	j	j	PROPN
ejpam-4568	30	32	=	=	SYM
ejpam-4568	30	33	∑	∑	PUNCT
ejpam-4568	30	34	(	(	PUNCT
ejpam-4568	30	35	m	m	PROPN
ejpam-4568	30	36	,	,	PUNCT
ejpam-4568	30	37	n)∈pi	n)∈pi	PROPN
ejpam-4568	30	38	,	,	PUNCT
ejpam-4568	30	39	j∩n	j∩n	PROPN
ejpam-4568	30	40	φα(ui	φα(ui	PROPN
ejpam-4568	30	41	,	,	PUNCT
ejpam-4568	30	42	j	j	PROPN
ejpam-4568	30	43	−	−	PROPN
ejpam-4568	30	44	ym	ym	PROPN
ejpam-4568	30	45	,	,	PUNCT
ejpam-4568	30	46	n	n	CCONJ
ejpam-4568	30	47	)	)	PUNCT
ejpam-4568	30	48	and	and	CCONJ
ejpam-4568	30	49	φα	φα	PROPN
ejpam-4568	30	50	is	be	AUX
ejpam-4568	30	51	an	an	DET
ejpam-4568	30	52	edge	edge	NOUN
ejpam-4568	30	53	-	-	PUNCT
ejpam-4568	30	54	preserving	preserve	VERB
ejpam-4568	30	55	potential	potential	ADJ
ejpam-4568	30	56	function	function	NOUN
ejpam-4568	30	57	having	have	VERB
ejpam-4568	30	58	the	the	DET
ejpam-4568	30	59	parameter	parameter	NOUN
ejpam-4568	30	60	α	α	NOUN
ejpam-4568	30	61	.	.	PUNCT
ejpam-4568	31	1	examples	example	NOUN
ejpam-4568	31	2	of	of	ADP
ejpam-4568	31	3	such	such	ADJ
ejpam-4568	31	4	φα(x	φα(x	NUM
ejpam-4568	31	5	)	)	PUNCT
ejpam-4568	31	6	are	be	AUX
ejpam-4568	31	7	:	:	PUNCT
ejpam-4568	31	8	φα(x	φα(x	X
ejpam-4568	31	9	)	)	PUNCT
ejpam-4568	32	1	=	=	SYM
ejpam-4568	33	1	√	√	NUM
ejpam-4568	33	2	α+	α+	X
ejpam-4568	33	3	x2	x2	PROPN
ejpam-4568	33	4	,	,	PUNCT
ejpam-4568	33	5	α	α	X
ejpam-4568	33	6	>	>	X
ejpam-4568	33	7	0	0	PROPN
ejpam-4568	33	8	.	.	PUNCT
ejpam-4568	34	1	however	however	ADV
ejpam-4568	34	2	,	,	PUNCT
ejpam-4568	34	3	because	because	SCONJ
ejpam-4568	34	4	of	of	ADP
ejpam-4568	34	5	the	the	DET
ejpam-4568	34	6	|ui	|ui	NOUN
ejpam-4568	34	7	,	,	PUNCT
ejpam-4568	34	8	j	j	PROPN
ejpam-4568	34	9	−	−	PROPN
ejpam-4568	34	10	yi	yi	PROPN
ejpam-4568	34	11	,	,	PUNCT
ejpam-4568	34	12	j	j	PROPN
ejpam-4568	34	13	|	|	ADV
ejpam-4568	34	14	term	term	NOUN
ejpam-4568	34	15	,	,	PUNCT
ejpam-4568	34	16	the	the	DET
ejpam-4568	34	17	functional	functional	ADJ
ejpam-4568	34	18	of	of	ADP
ejpam-4568	34	19	problem	problem	NOUN
ejpam-4568	34	20	(	(	PUNCT
ejpam-4568	34	21	1	1	X
ejpam-4568	34	22	)	)	PUNCT
ejpam-4568	34	23	is	be	AUX
ejpam-4568	34	24	nonsmooth	nonsmooth	ADJ
ejpam-4568	34	25	.	.	PUNCT
ejpam-4568	35	1	it	it	PRON
ejpam-4568	35	2	is	be	AUX
ejpam-4568	35	3	commonly	commonly	ADV
ejpam-4568	35	4	accepted	accept	VERB
ejpam-4568	35	5	that	that	SCONJ
ejpam-4568	35	6	this	this	DET
ejpam-4568	35	7	nonsmooth	nonsmooth	ADJ
ejpam-4568	35	8	term	term	NOUN
ejpam-4568	35	9	can	can	AUX
ejpam-4568	35	10	separate	separate	VERB
ejpam-4568	35	11	from	from	ADP
ejpam-4568	35	12	(	(	PUNCT
ejpam-4568	35	13	1	1	NUM
ejpam-4568	35	14	)	)	PUNCT
ejpam-4568	35	15	because	because	SCONJ
ejpam-4568	35	16	,	,	PUNCT
ejpam-4568	35	17	on	on	ADP
ejpam-4568	35	18	the	the	DET
ejpam-4568	35	19	one	one	NUM
ejpam-4568	35	20	hand	hand	NOUN
ejpam-4568	35	21	,	,	PUNCT
ejpam-4568	35	22	it	it	PRON
ejpam-4568	35	23	keeps	keep	VERB
ejpam-4568	35	24	the	the	DET
ejpam-4568	35	25	minimizer	minimizer	NOUN
ejpam-4568	35	26	u	u	NOUN
ejpam-4568	35	27	close	close	ADJ
ejpam-4568	35	28	to	to	ADP
ejpam-4568	35	29	the	the	DET
ejpam-4568	35	30	original	original	ADJ
ejpam-4568	35	31	picture	picture	NOUN
ejpam-4568	35	32	y	y	NOUN
ejpam-4568	35	33	,	,	PUNCT
ejpam-4568	35	34	ensuring	ensure	VERB
ejpam-4568	35	35	that	that	SCONJ
ejpam-4568	35	36	the	the	DET
ejpam-4568	35	37	original	original	ADJ
ejpam-4568	35	38	image	image	NOUN
ejpam-4568	35	39	’s	’s	PART
ejpam-4568	35	40	unaltered	unaltered	ADJ
ejpam-4568	35	41	pixels	pixel	NOUN
ejpam-4568	35	42	are	be	AUX
ejpam-4568	35	43	preserved	preserve	VERB
ejpam-4568	35	44	.	.	PUNCT
ejpam-4568	36	1	however	however	ADV
ejpam-4568	36	2	,	,	PUNCT
ejpam-4568	36	3	the	the	DET
ejpam-4568	36	4	two	two	NUM
ejpam-4568	36	5	-	-	PUNCT
ejpam-4568	36	6	phase	phase	NOUN
ejpam-4568	36	7	approach	approach	NOUN
ejpam-4568	36	8	simply	simply	ADV
ejpam-4568	36	9	cleans	clean	VERB
ejpam-4568	36	10	the	the	DET
ejpam-4568	36	11	noisy	noisy	ADJ
ejpam-4568	36	12	pixels	pixel	NOUN
ejpam-4568	36	13	,	,	PUNCT
ejpam-4568	36	14	leaving	leave	VERB
ejpam-4568	36	15	the	the	DET
ejpam-4568	36	16	unharmed	unharmed	ADJ
ejpam-4568	36	17	pixels	pixel	NOUN
ejpam-4568	36	18	unaltered	unaltered	ADJ
ejpam-4568	36	19	,	,	PUNCT
ejpam-4568	36	20	making	make	VERB
ejpam-4568	36	21	issue	issue	NOUN
ejpam-4568	36	22	(	(	PUNCT
ejpam-4568	36	23	1	1	NUM
ejpam-4568	36	24	)	)	PUNCT
ejpam-4568	36	25	functional	functional	ADJ
ejpam-4568	36	26	.	.	PUNCT
ejpam-4568	37	1	consequently	consequently	ADV
ejpam-4568	37	2	,	,	PUNCT
ejpam-4568	37	3	the	the	DET
ejpam-4568	37	4	nonsmooth	nonsmooth	NOUN
ejpam-4568	37	5	term	term	NOUN
ejpam-4568	37	6	is	be	AUX
ejpam-4568	37	7	not	not	PART
ejpam-4568	37	8	necessary	necessary	ADJ
ejpam-4568	37	9	.	.	PUNCT
ejpam-4568	38	1	the	the	DET
ejpam-4568	38	2	word	word	NOUN
ejpam-4568	38	3	”	"	PUNCT
ejpam-4568	38	4	data	data	NOUN
ejpam-4568	38	5	-	-	PUNCT
ejpam-4568	38	6	fitting	fit	VERB
ejpam-4568	38	7	”	"	PUNCT
ejpam-4568	38	8	in	in	ADP
ejpam-4568	38	9	[	[	X
ejpam-4568	38	10	4	4	NUM
ejpam-4568	38	11	]	]	PUNCT
ejpam-4568	38	12	should	should	AUX
ejpam-4568	38	13	be	be	AUX
ejpam-4568	38	14	removed	remove	VERB
ejpam-4568	38	15	,	,	PUNCT
ejpam-4568	38	16	according	accord	VERB
ejpam-4568	38	17	to	to	ADP
ejpam-4568	38	18	cai	cai	PROPN
ejpam-4568	38	19	et	et	PROPN
ejpam-4568	38	20	al	al	PROPN
ejpam-4568	38	21	.	.	PROPN
ejpam-4568	39	1	with	with	ADP
ejpam-4568	39	2	this	this	DET
ejpam-4568	39	3	process	process	NOUN
ejpam-4568	39	4	,	,	PUNCT
ejpam-4568	39	5	fα(u	fα(u	NOUN
ejpam-4568	39	6	)	)	PUNCT
ejpam-4568	39	7	may	may	AUX
ejpam-4568	39	8	be	be	AUX
ejpam-4568	39	9	converted	convert	VERB
ejpam-4568	39	10	into	into	ADP
ejpam-4568	39	11	a	a	DET
ejpam-4568	39	12	smooth	smooth	ADJ
ejpam-4568	39	13	function	function	NOUN
ejpam-4568	39	14	that	that	PRON
ejpam-4568	39	15	can	can	AUX
ejpam-4568	39	16	be	be	AUX
ejpam-4568	39	17	effectively	effectively	ADV
ejpam-4568	39	18	reduced	reduce	VERB
ejpam-4568	39	19	.	.	PUNCT
ejpam-4568	40	1	as	as	ADP
ejpam-4568	40	2	a	a	DET
ejpam-4568	40	3	result	result	NOUN
ejpam-4568	40	4	,	,	PUNCT
ejpam-4568	40	5	the	the	DET
ejpam-4568	40	6	objective	objective	ADJ
ejpam-4568	40	7	function	function	NOUN
ejpam-4568	40	8	that	that	PRON
ejpam-4568	40	9	we	we	PRON
ejpam-4568	40	10	will	will	AUX
ejpam-4568	40	11	reduce	reduce	VERB
ejpam-4568	40	12	in	in	ADP
ejpam-4568	40	13	this	this	DET
ejpam-4568	40	14	essay	essay	NOUN
ejpam-4568	40	15	has	have	VERB
ejpam-4568	40	16	the	the	DET
ejpam-4568	40	17	following	follow	VERB
ejpam-4568	40	18	shape	shape	NOUN
ejpam-4568	40	19	:	:	PUNCT
ejpam-4568	40	20	fα(u	fα(u	X
ejpam-4568	40	21	)	)	PUNCT
ejpam-4568	40	22	=	=	PUNCT
ejpam-4568	40	23	∑	∑	PUNCT
ejpam-4568	40	24	(	(	PUNCT
ejpam-4568	40	25	i	i	PROPN
ejpam-4568	40	26	,	,	PUNCT
ejpam-4568	40	27	j)∈n	j)∈n	PROPN
ejpam-4568	40	28	[	[	PUNCT
ejpam-4568	40	29	(	(	PUNCT
ejpam-4568	40	30	2×	2×	NUM
ejpam-4568	40	31	s1	s1	PROPN
ejpam-4568	40	32	i	i	PROPN
ejpam-4568	40	33	,	,	PUNCT
ejpam-4568	40	34	j	j	PROPN
ejpam-4568	40	35	+	+	CCONJ
ejpam-4568	40	36	s2	s2	VERB
ejpam-4568	40	37	i	i	PROPN
ejpam-4568	40	38	,	,	PUNCT
ejpam-4568	40	39	j	j	PROPN
ejpam-4568	40	40	)	)	PUNCT
ejpam-4568	40	41	]	]	PUNCT
ejpam-4568	40	42	(	(	PUNCT
ejpam-4568	40	43	2	2	NUM
ejpam-4568	40	44	)	)	PUNCT
ejpam-4568	40	45	nowadays	nowadays	ADV
ejpam-4568	40	46	,	,	PUNCT
ejpam-4568	40	47	conjugate	conjugate	ADJ
ejpam-4568	40	48	gradient	gradient	NOUN
ejpam-4568	40	49	(	(	PUNCT
ejpam-4568	40	50	cg	cg	NOUN
ejpam-4568	40	51	)	)	PUNCT
ejpam-4568	40	52	methods	method	NOUN
ejpam-4568	40	53	are	be	AUX
ejpam-4568	40	54	regarded	regard	VERB
ejpam-4568	40	55	as	as	ADP
ejpam-4568	40	56	popular	popular	ADJ
ejpam-4568	40	57	and	and	CCONJ
ejpam-4568	40	58	efficient	efficient	ADJ
ejpam-4568	40	59	algorithms	algorithm	NOUN
ejpam-4568	40	60	to	to	PART
ejpam-4568	40	61	deal	deal	VERB
ejpam-4568	40	62	problems	problem	NOUN
ejpam-4568	40	63	noise	noise	NOUN
ejpam-4568	40	64	of	of	ADP
ejpam-4568	40	65	image	image	NOUN
ejpam-4568	40	66	.	.	PUNCT
ejpam-4568	41	1	in	in	ADP
ejpam-4568	41	2	general	general	ADJ
ejpam-4568	41	3	,	,	PUNCT
ejpam-4568	41	4	the	the	DET
ejpam-4568	41	5	method	method	NOUN
ejpam-4568	41	6	has	have	VERB
ejpam-4568	41	7	the	the	DET
ejpam-4568	41	8	following	follow	VERB
ejpam-4568	41	9	form	form	NOUN
ejpam-4568	41	10	:	:	PUNCT
ejpam-4568	41	11	minf(u);u	minf(u);u	PROPN
ejpam-4568	41	12	∈	∈	PROPN
ejpam-4568	41	13	rn	rn	PROPN
ejpam-4568	41	14	(	(	PUNCT
ejpam-4568	41	15	3	3	NUM
ejpam-4568	41	16	)	)	PUNCT
ejpam-4568	41	17	where	where	SCONJ
ejpam-4568	41	18	f	f	PROPN
ejpam-4568	41	19	is	be	AUX
ejpam-4568	41	20	a	a	DET
ejpam-4568	41	21	continuously	continuously	ADV
ejpam-4568	41	22	differentiable	differentiable	ADJ
ejpam-4568	41	23	function	function	NOUN
ejpam-4568	41	24	.	.	PUNCT
ejpam-4568	42	1	this	this	DET
ejpam-4568	42	2	problem	problem	NOUN
ejpam-4568	42	3	may	may	AUX
ejpam-4568	42	4	be	be	AUX
ejpam-4568	42	5	effectively	effectively	ADV
ejpam-4568	42	6	solved	solve	VERB
ejpam-4568	42	7	using	use	VERB
ejpam-4568	42	8	an	an	DET
ejpam-4568	42	9	iterative	iterative	NOUN
ejpam-4568	42	10	technique	technique	NOUN
ejpam-4568	42	11	at	at	ADP
ejpam-4568	42	12	the	the	DET
ejpam-4568	42	13	(	(	PUNCT
ejpam-4568	42	14	k	k	NOUN
ejpam-4568	42	15	+	+	NOUN
ejpam-4568	42	16	1	1	X
ejpam-4568	42	17	)	)	PUNCT
ejpam-4568	42	18	iteration	iteration	NOUN
ejpam-4568	42	19	by	by	ADP
ejpam-4568	42	20	using	use	VERB
ejpam-4568	42	21	the	the	DET
ejpam-4568	42	22	iteration	iteration	NOUN
ejpam-4568	42	23	form	form	NOUN
ejpam-4568	42	24	shown	show	VERB
ejpam-4568	42	25	below	below	ADP
ejpam-4568	42	26	:	:	PUNCT
ejpam-4568	42	27	uk+1	uk+1	PROPN
ejpam-4568	42	28	=	=	PUNCT
ejpam-4568	42	29	uk	uk	PROPN
ejpam-4568	42	30	+	+	CCONJ
ejpam-4568	42	31	αkdk	αkdk	NOUN
ejpam-4568	42	32	(	(	PUNCT
ejpam-4568	42	33	4	4	NUM
ejpam-4568	42	34	)	)	PUNCT
ejpam-4568	42	35	where	where	SCONJ
ejpam-4568	42	36	uk	uk	PROPN
ejpam-4568	42	37	is	be	AUX
ejpam-4568	42	38	the	the	DET
ejpam-4568	42	39	current	current	ADJ
ejpam-4568	42	40	iterate	iterate	NOUN
ejpam-4568	42	41	point	point	NOUN
ejpam-4568	42	42	,	,	PUNCT
ejpam-4568	42	43	dk	dk	PROPN
ejpam-4568	42	44	is	be	AUX
ejpam-4568	42	45	a	a	DET
ejpam-4568	42	46	direction	direction	NOUN
ejpam-4568	42	47	of	of	ADP
ejpam-4568	42	48	f	f	PROPN
ejpam-4568	42	49	at	at	ADP
ejpam-4568	42	50	uk	uk	PROPN
ejpam-4568	42	51	,	,	PUNCT
ejpam-4568	42	52	and	and	CCONJ
ejpam-4568	42	53	αk	αk	SCONJ
ejpam-4568	42	54	>	>	X
ejpam-4568	42	55	0	0	PUNCT
ejpam-4568	42	56	is	be	AUX
ejpam-4568	42	57	step	step	NOUN
ejpam-4568	42	58	size	size	NOUN
ejpam-4568	42	59	obtained	obtain	VERB
ejpam-4568	42	60	by	by	ADP
ejpam-4568	42	61	a	a	DET
ejpam-4568	42	62	one	one	NUM
ejpam-4568	42	63	-	-	PUNCT
ejpam-4568	42	64	dimensional	dimensional	ADJ
ejpam-4568	42	65	line	line	NOUN
ejpam-4568	42	66	search	search	NOUN
ejpam-4568	42	67	.	.	PUNCT
ejpam-4568	43	1	step	step	NOUN
ejpam-4568	43	2	size	size	NOUN
ejpam-4568	43	3	αk	αk	NOUN
ejpam-4568	43	4	is	be	AUX
ejpam-4568	43	5	obtained	obtain	VERB
ejpam-4568	43	6	using	use	VERB
ejpam-4568	43	7	several	several	ADJ
ejpam-4568	43	8	forms	form	NOUN
ejpam-4568	43	9	of	of	ADP
ejpam-4568	43	10	line	line	NOUN
ejpam-4568	43	11	search	search	NOUN
ejpam-4568	43	12	,	,	PUNCT
ejpam-4568	43	13	i.e.	i.e.	X
ejpam-4568	43	14	,	,	PUNCT
ejpam-4568	43	15	exact	exact	ADJ
ejpam-4568	43	16	line	line	NOUN
ejpam-4568	43	17	search	search	NOUN
ejpam-4568	43	18	with	with	ADP
ejpam-4568	43	19	quadratic	quadratic	ADJ
ejpam-4568	43	20	model	model	NOUN
ejpam-4568	43	21	[	[	X
ejpam-4568	43	22	20	20	NUM
ejpam-4568	43	23	]	]	PUNCT
ejpam-4568	43	24	as	as	SCONJ
ejpam-4568	43	25	follows	follow	VERB
ejpam-4568	43	26	:	:	PUNCT
ejpam-4568	43	27	αk	αk	ADP
ejpam-4568	43	28	=	=	PUNCT
ejpam-4568	43	29	−	−	PROPN
ejpam-4568	43	30	gtk	gtk	NOUN
ejpam-4568	43	31	dk	dk	PROPN
ejpam-4568	43	32	dtkqdk	dtkqdk	NOUN
ejpam-4568	43	33	(	(	PUNCT
ejpam-4568	43	34	5	5	NUM
ejpam-4568	43	35	)	)	PUNCT
ejpam-4568	43	36	the	the	DET
ejpam-4568	43	37	strong	strong	ADJ
ejpam-4568	43	38	wolfe	wolfe	PROPN
ejpam-4568	43	39	inexact	inexact	ADJ
ejpam-4568	43	40	line	line	NOUN
ejpam-4568	43	41	search	search	NOUN
ejpam-4568	43	42	is	be	AUX
ejpam-4568	43	43	frequently	frequently	ADV
ejpam-4568	43	44	taken	take	VERB
ejpam-4568	43	45	into	into	ADP
ejpam-4568	43	46	consideration	consideration	NOUN
ejpam-4568	43	47	in	in	ADP
ejpam-4568	43	48	the	the	DET
ejpam-4568	43	49	convergence	convergence	NOUN
ejpam-4568	43	50	analysis	analysis	NOUN
ejpam-4568	43	51	implementation	implementation	NOUN
ejpam-4568	43	52	of	of	ADP
ejpam-4568	43	53	nonlinear	nonlinear	ADJ
ejpam-4568	43	54	conjugate	conjugate	ADJ
ejpam-4568	43	55	gradient	gradient	NOUN
ejpam-4568	43	56	techniques	technique	NOUN
ejpam-4568	43	57	since	since	SCONJ
ejpam-4568	43	58	exact	exact	ADJ
ejpam-4568	43	59	line	line	NOUN
ejpam-4568	43	60	b.	b.	PROPN
ejpam-4568	43	61	a.	a.	PROPN
ejpam-4568	43	62	hassan	hassan	PROPN
ejpam-4568	43	63	,	,	PUNCT
ejpam-4568	43	64	h.	h.	PROPN
ejpam-4568	43	65	m.	m.	PROPN
ejpam-4568	43	66	sadiq	sadiq	PROPN
ejpam-4568	43	67	/	/	SYM
ejpam-4568	43	68	eur	eur	PROPN
ejpam-4568	43	69	.	.	PUNCT
ejpam-4568	44	1	j.	j.	PROPN
ejpam-4568	44	2	pure	pure	PROPN
ejpam-4568	44	3	appl	appl	PROPN
ejpam-4568	44	4	.	.	PROPN
ejpam-4568	44	5	math	math	PROPN
ejpam-4568	44	6	,	,	PUNCT
ejpam-4568	44	7	15	15	NUM
ejpam-4568	44	8	(	(	PUNCT
ejpam-4568	44	9	4	4	NUM
ejpam-4568	44	10	)	)	PUNCT
ejpam-4568	44	11	(	(	PUNCT
ejpam-4568	44	12	2022	2022	NUM
ejpam-4568	44	13	)	)	PUNCT
ejpam-4568	44	14	,	,	PUNCT
ejpam-4568	44	15	2011	2011	NUM
ejpam-4568	44	16	-	-	SYM
ejpam-4568	44	17	2021	2021	NUM
ejpam-4568	44	18	2013	2013	NUM
ejpam-4568	44	19	search	search	NOUN
ejpam-4568	44	20	for	for	ADP
ejpam-4568	44	21	searching	search	VERB
ejpam-4568	44	22	αk	αk	NOUN
ejpam-4568	44	23	is	be	AUX
ejpam-4568	44	24	typically	typically	ADV
ejpam-4568	44	25	costly	costly	ADJ
ejpam-4568	44	26	and	and	CCONJ
ejpam-4568	44	27	impracticable	impracticable	ADJ
ejpam-4568	44	28	.	.	PUNCT
ejpam-4568	45	1	it	it	PRON
ejpam-4568	45	2	seeks	seek	VERB
ejpam-4568	45	3	to	to	PART
ejpam-4568	45	4	identify	identify	VERB
ejpam-4568	45	5	a	a	DET
ejpam-4568	45	6	step	step	NOUN
ejpam-4568	45	7	size	size	NOUN
ejpam-4568	45	8	αk	αk	NOUN
ejpam-4568	45	9	that	that	PRON
ejpam-4568	45	10	satisfies	satisfy	VERB
ejpam-4568	45	11	the	the	DET
ejpam-4568	45	12	two	two	NUM
ejpam-4568	45	13	strong	strong	ADJ
ejpam-4568	45	14	wolfe	wolfe	PROPN
ejpam-4568	45	15	requirements	requirement	NOUN
ejpam-4568	45	16	listed	list	VERB
ejpam-4568	45	17	below	below	ADP
ejpam-4568	45	18	[	[	X
ejpam-4568	45	19	5	5	NUM
ejpam-4568	45	20	]	]	PUNCT
ejpam-4568	45	21	,	,	PUNCT
ejpam-4568	45	22	namely	namely	ADV
ejpam-4568	45	23	:	:	PUNCT
ejpam-4568	45	24	f(uk	f(uk	PROPN
ejpam-4568	45	25	+	+	NUM
ejpam-4568	45	26	αkdk	αkdk	NOUN
ejpam-4568	45	27	)	)	PUNCT
ejpam-4568	45	28	≤	≤	PROPN
ejpam-4568	45	29	f(uk	f(uk	PROPN
ejpam-4568	45	30	)	)	PUNCT
ejpam-4568	45	31	+	+	CCONJ
ejpam-4568	46	1	δαkg	δαkg	NOUN
ejpam-4568	46	2	t	t	PROPN
ejpam-4568	46	3	k	k	PROPN
ejpam-4568	46	4	dk	dk	PROPN
ejpam-4568	46	5	(	(	PUNCT
ejpam-4568	46	6	6)∣∣g(uk	6)∣∣g(uk	PROPN
ejpam-4568	46	7	+	+	CCONJ
ejpam-4568	46	8	αkdk	αkdk	NOUN
ejpam-4568	46	9	)	)	PUNCT
ejpam-4568	46	10	tdk	tdk	NOUN
ejpam-4568	46	11	∣∣	∣∣	X
ejpam-4568	46	12	≤	≤	PROPN
ejpam-4568	46	13	σ	σ	NOUN
ejpam-4568	46	14	∣∣gtk	∣∣gtk	NUM
ejpam-4568	46	15	dk∣∣	dk∣∣	NOUN
ejpam-4568	46	16	(	(	PUNCT
ejpam-4568	46	17	7	7	NUM
ejpam-4568	46	18	)	)	PUNCT
ejpam-4568	46	19	where	where	SCONJ
ejpam-4568	46	20	0	0	NUM
ejpam-4568	46	21	<	<	X
ejpam-4568	46	22	δ	δ	X
ejpam-4568	46	23	<	<	X
ejpam-4568	46	24	σ	σ	X
ejpam-4568	46	25	<	<	X
ejpam-4568	46	26	1	1	NUM
ejpam-4568	46	27	are	be	AUX
ejpam-4568	46	28	arbitrary	arbitrary	ADJ
ejpam-4568	46	29	constants	constant	NOUN
ejpam-4568	46	30	and	and	CCONJ
ejpam-4568	46	31	gk	gk	NOUN
ejpam-4568	46	32	=	=	SYM
ejpam-4568	46	33	∇f(xk	∇f(xk	NOUN
ejpam-4568	46	34	)	)	PUNCT
ejpam-4568	46	35	.	.	PUNCT
ejpam-4568	47	1	the	the	DET
ejpam-4568	47	2	search	search	NOUN
ejpam-4568	47	3	direction	direction	NOUN
ejpam-4568	47	4	dk	dk	PROPN
ejpam-4568	47	5	is	be	AUX
ejpam-4568	47	6	computed	compute	VERB
ejpam-4568	47	7	by	by	ADP
ejpam-4568	47	8	:	:	PUNCT
ejpam-4568	47	9	dk+1	dk+1	X
ejpam-4568	47	10	=	=	SYM
ejpam-4568	47	11	−gk+1	−gk+1	PROPN
ejpam-4568	47	12	+	+	CCONJ
ejpam-4568	47	13	βksk	βksk	NOUN
ejpam-4568	47	14	(	(	PUNCT
ejpam-4568	47	15	8)	8)	NUM
ejpam-4568	47	16	βk	βk	NOUN
ejpam-4568	47	17	is	be	AUX
ejpam-4568	47	18	the	the	DET
ejpam-4568	47	19	conjugate	conjugate	ADJ
ejpam-4568	47	20	gradient	gradient	NOUN
ejpam-4568	47	21	parameter	parameter	NOUN
ejpam-4568	47	22	that	that	PRON
ejpam-4568	47	23	evaluates	evaluate	VERB
ejpam-4568	47	24	the	the	DET
ejpam-4568	47	25	performance	performance	NOUN
ejpam-4568	47	26	and	and	CCONJ
ejpam-4568	47	27	global	global	ADJ
ejpam-4568	47	28	convergence	convergence	NOUN
ejpam-4568	47	29	characteristics	characteristic	NOUN
ejpam-4568	47	30	of	of	ADP
ejpam-4568	47	31	several	several	ADJ
ejpam-4568	47	32	conjugate	conjugate	ADJ
ejpam-4568	47	33	gradient	gradient	ADJ
ejpam-4568	47	34	technique	technique	NOUN
ejpam-4568	47	35	types	type	NOUN
ejpam-4568	47	36	.	.	PUNCT
ejpam-4568	48	1	the	the	DET
ejpam-4568	48	2	nonlinear	nonlinear	ADJ
ejpam-4568	48	3	conjugate	conjugate	ADJ
ejpam-4568	48	4	gradient	gradient	ADJ
ejpam-4568	48	5	parameters	parameter	NOUN
ejpam-4568	48	6	include	include	VERB
ejpam-4568	48	7	some	some	DET
ejpam-4568	48	8	well	well	ADV
ejpam-4568	48	9	-	-	PUNCT
ejpam-4568	48	10	known	know	VERB
ejpam-4568	48	11	ones	one	NOUN
ejpam-4568	48	12	like	like	ADP
ejpam-4568	48	13	the	the	DET
ejpam-4568	48	14	fletcher	fletcher	NOUN
ejpam-4568	48	15	and	and	CCONJ
ejpam-4568	48	16	reeves	reeves	PROPN
ejpam-4568	48	17	(	(	PUNCT
ejpam-4568	48	18	fr)[10	fr)[10	NOUN
ejpam-4568	48	19	]	]	PUNCT
ejpam-4568	48	20	,	,	PUNCT
ejpam-4568	48	21	conjugate	conjugate	ADJ
ejpam-4568	48	22	descent	descent	NOUN
ejpam-4568	48	23	(	(	PUNCT
ejpam-4568	48	24	cd)[9	cd)[9	PROPN
ejpam-4568	48	25	]	]	X
ejpam-4568	48	26	,	,	PUNCT
ejpam-4568	48	27	dai	dai	PROPN
ejpam-4568	48	28	and	and	CCONJ
ejpam-4568	48	29	yuan	yuan	PROPN
ejpam-4568	48	30	(	(	PUNCT
ejpam-4568	48	31	dy	dy	X
ejpam-4568	48	32	)	)	PUNCT
ejpam-4568	49	1	[	[	X
ejpam-4568	49	2	8	8	NUM
ejpam-4568	49	3	]	]	PUNCT
ejpam-4568	49	4	.	.	PUNCT
ejpam-4568	50	1	these	these	DET
ejpam-4568	50	2	parameters	parameter	NOUN
ejpam-4568	50	3	are	be	AUX
ejpam-4568	50	4	given	give	VERB
ejpam-4568	50	5	by	by	ADP
ejpam-4568	50	6	the	the	DET
ejpam-4568	50	7	following	follow	VERB
ejpam-4568	50	8	formulae	formulae	NOUN
ejpam-4568	50	9	:	:	PUNCT
ejpam-4568	50	10	βfr	βfr	PROPN
ejpam-4568	50	11	k	k	NOUN
ejpam-4568	50	12	=	=	SYM
ejpam-4568	50	13	gtk+1gk+1	gtk+1gk+1	NOUN
ejpam-4568	50	14	gtk	gtk	NOUN
ejpam-4568	50	15	gk	gk	PROPN
ejpam-4568	50	16	,	,	PUNCT
ejpam-4568	50	17	βcd	βcd	VERB
ejpam-4568	50	18	k	k	NOUN
ejpam-4568	51	1	=	=	SYM
ejpam-4568	51	2	−	−	PROPN
ejpam-4568	51	3	gtk+1gk+1	gtk+1gk+1	NOUN
ejpam-4568	51	4	dtk	dtk	PROPN
ejpam-4568	51	5	gk	gk	PROPN
ejpam-4568	51	6	βdy	βdy	PROPN
ejpam-4568	51	7	k	k	PROPN
ejpam-4568	52	1	=	=	PROPN
ejpam-4568	52	2	gtk+1gk+1	gtk+1gk+1	PROPN
ejpam-4568	52	3	dtk	dtk	PROPN
ejpam-4568	52	4	yk	yk	PROPN
ejpam-4568	52	5	,	,	PUNCT
ejpam-4568	52	6	(	(	PUNCT
ejpam-4568	52	7	9	9	X
ejpam-4568	52	8	)	)	PUNCT
ejpam-4568	52	9	more	more	ADJ
ejpam-4568	52	10	information	information	NOUN
ejpam-4568	52	11	about	about	ADP
ejpam-4568	52	12	other	other	ADJ
ejpam-4568	52	13	conjugate	conjugate	ADJ
ejpam-4568	52	14	gradient	gradient	NOUN
ejpam-4568	52	15	methods	method	NOUN
ejpam-4568	52	16	[	[	X
ejpam-4568	52	17	16	16	NUM
ejpam-4568	52	18	]	]	PUNCT
ejpam-4568	52	19	.	.	PUNCT
ejpam-4568	53	1	al	al	PROPN
ejpam-4568	53	2	-	-	PUNCT
ejpam-4568	53	3	baali	baali	PROPN
ejpam-4568	53	4	[	[	X
ejpam-4568	53	5	1	1	NUM
ejpam-4568	53	6	]	]	PUNCT
ejpam-4568	53	7	extended	extend	VERB
ejpam-4568	53	8	this	this	DET
ejpam-4568	53	9	result	result	NOUN
ejpam-4568	53	10	to	to	ADP
ejpam-4568	53	11	an	an	DET
ejpam-4568	53	12	inexact	inexact	ADJ
ejpam-4568	53	13	line	line	NOUN
ejpam-4568	53	14	search	search	NOUN
ejpam-4568	53	15	and	and	CCONJ
ejpam-4568	53	16	showed	show	VERB
ejpam-4568	53	17	that	that	SCONJ
ejpam-4568	53	18	the	the	DET
ejpam-4568	53	19	method	method	NOUN
ejpam-4568	53	20	generates	generate	VERB
ejpam-4568	53	21	sufficient	sufficient	ADJ
ejpam-4568	53	22	descent	descent	NOUN
ejpam-4568	53	23	direction	direction	NOUN
ejpam-4568	53	24	under	under	ADP
ejpam-4568	53	25	the	the	DET
ejpam-4568	53	26	strong	strong	ADJ
ejpam-4568	53	27	wolfe	wolfe	PROPN
ejpam-4568	53	28	conditions	condition	NOUN
ejpam-4568	53	29	using	use	VERB
ejpam-4568	53	30	the	the	DET
ejpam-4568	53	31	constraint	constraint	NOUN
ejpam-4568	53	32	σ	σ	NOUN
ejpam-4568	53	33	<	<	X
ejpam-4568	53	34	1/2	1/2	NUM
ejpam-4568	53	35	.	.	PUNCT
ejpam-4568	54	1	the	the	DET
ejpam-4568	54	2	global	global	ADJ
ejpam-4568	54	3	convergence	convergence	NOUN
ejpam-4568	54	4	properties	property	NOUN
ejpam-4568	54	5	of	of	ADP
ejpam-4568	54	6	fr	fr	NOUN
ejpam-4568	54	7	,	,	PUNCT
ejpam-4568	54	8	dy	dy	NOUN
ejpam-4568	54	9	,	,	PUNCT
ejpam-4568	54	10	and	and	CCONJ
ejpam-4568	54	11	cd	cd	PROPN
ejpam-4568	54	12	methods	method	NOUN
ejpam-4568	54	13	with	with	ADP
ejpam-4568	54	14	exalt	exalt	VERB
ejpam-4568	54	15	line	line	NOUN
ejpam-4568	54	16	are	be	AUX
ejpam-4568	54	17	strong	strong	ADJ
ejpam-4568	54	18	,	,	PUNCT
ejpam-4568	54	19	but	but	CCONJ
ejpam-4568	54	20	they	they	PRON
ejpam-4568	54	21	are	be	AUX
ejpam-4568	54	22	prone	prone	ADJ
ejpam-4568	54	23	to	to	ADP
ejpam-4568	54	24	taking	take	VERB
ejpam-4568	54	25	many	many	ADJ
ejpam-4568	54	26	short	short	ADJ
ejpam-4568	54	27	steps	step	NOUN
ejpam-4568	54	28	without	without	ADP
ejpam-4568	54	29	making	make	VERB
ejpam-4568	54	30	sufficient	sufficient	ADJ
ejpam-4568	54	31	advancement	advancement	NOUN
ejpam-4568	54	32	to	to	ADP
ejpam-4568	54	33	the	the	DET
ejpam-4568	54	34	minimum	minimum	NOUN
ejpam-4568	54	35	by	by	ADP
ejpam-4568	54	36	hager	hager	NOUN
ejpam-4568	54	37	and	and	CCONJ
ejpam-4568	54	38	zhang	zhang	PROPN
ejpam-4568	55	1	[	[	X
ejpam-4568	55	2	11	11	NUM
ejpam-4568	55	3	]	]	PUNCT
ejpam-4568	55	4	.	.	PUNCT
ejpam-4568	56	1	to	to	PART
ejpam-4568	56	2	provide	provide	VERB
ejpam-4568	56	3	a	a	DET
ejpam-4568	56	4	broader	broad	ADJ
ejpam-4568	56	5	context	context	NOUN
ejpam-4568	56	6	for	for	ADP
ejpam-4568	56	7	developing	develop	VERB
ejpam-4568	56	8	conjugate	conjugate	ADJ
ejpam-4568	56	9	gradient	gradient	ADJ
ejpam-4568	56	10	methods	method	NOUN
ejpam-4568	56	11	,	,	PUNCT
ejpam-4568	56	12	perry	perry	PROPN
ejpam-4568	56	13	extended	extend	VERB
ejpam-4568	56	14	the	the	DET
ejpam-4568	56	15	classical	classical	ADJ
ejpam-4568	56	16	conjugate	conjugate	ADJ
ejpam-4568	56	17	condition	condition	NOUN
ejpam-4568	56	18	to	to	PART
ejpam-4568	56	19	:	:	PUNCT
ejpam-4568	56	20	dtk+1yk	dtk+1yk	VERB
ejpam-4568	56	21	=	=	SYM
ejpam-4568	56	22	−(hk+1gk+1	−(hk+1gk+1	PROPN
ejpam-4568	56	23	)	)	PUNCT
ejpam-4568	56	24	t	t	PROPN
ejpam-4568	56	25	yk	yk	NOUN
ejpam-4568	56	26	=	=	SYM
ejpam-4568	56	27	−gk+1(hk+1yk	−gk+1(hk+1yk	PROPN
ejpam-4568	56	28	)	)	PUNCT
ejpam-4568	56	29	=	=	SYM
ejpam-4568	56	30	−gtk+1sk	−gtk+1sk	PROPN
ejpam-4568	56	31	=	=	SYM
ejpam-4568	56	32	0	0	NUM
ejpam-4568	56	33	(	(	PUNCT
ejpam-4568	56	34	10	10	NUM
ejpam-4568	56	35	)	)	PUNCT
ejpam-4568	56	36	many	many	ADJ
ejpam-4568	56	37	efforts	effort	NOUN
ejpam-4568	56	38	have	have	AUX
ejpam-4568	56	39	been	be	AUX
ejpam-4568	56	40	made	make	VERB
ejpam-4568	56	41	in	in	ADP
ejpam-4568	56	42	few	few	ADJ
ejpam-4568	56	43	recent	recent	ADJ
ejpam-4568	56	44	years	year	NOUN
ejpam-4568	56	45	to	to	PART
ejpam-4568	56	46	design	design	VERB
ejpam-4568	56	47	new	new	ADJ
ejpam-4568	56	48	formulas	formula	NOUN
ejpam-4568	56	49	for	for	ADP
ejpam-4568	56	50	conjugate	conjugate	ADJ
ejpam-4568	56	51	gradient	gradient	NOUN
ejpam-4568	56	52	method	method	NOUN
ejpam-4568	56	53	which	which	PRON
ejpam-4568	56	54	are	be	AUX
ejpam-4568	56	55	not	not	PART
ejpam-4568	56	56	only	only	ADV
ejpam-4568	56	57	satisfied	satisfied	ADJ
ejpam-4568	56	58	global	global	ADJ
ejpam-4568	56	59	convergence	convergence	NOUN
ejpam-4568	56	60	but	but	CCONJ
ejpam-4568	56	61	also	also	ADV
ejpam-4568	56	62	improve	improve	VERB
ejpam-4568	56	63	numerical	numerical	ADJ
ejpam-4568	56	64	performance	performance	NOUN
ejpam-4568	56	65	for	for	ADP
ejpam-4568	56	66	method	method	NOUN
ejpam-4568	56	67	.	.	PUNCT
ejpam-4568	57	1	remainder	remainder	VERB
ejpam-4568	57	2	the	the	DET
ejpam-4568	57	3	conjugate	conjugate	ADJ
ejpam-4568	57	4	gradient	gradient	NOUN
ejpam-4568	57	5	methods	method	NOUN
ejpam-4568	57	6	have	have	VERB
ejpam-4568	57	7	many	many	ADJ
ejpam-4568	57	8	application	application	NOUN
ejpam-4568	57	9	in	in	ADP
ejpam-4568	57	10	real	real	ADJ
ejpam-4568	57	11	life	life	NOUN
ejpam-4568	57	12	in	in	ADP
ejpam-4568	57	13	our	our	PRON
ejpam-4568	57	14	work	work	NOUN
ejpam-4568	57	15	,	,	PUNCT
ejpam-4568	57	16	we	we	PRON
ejpam-4568	57	17	found	find	VERB
ejpam-4568	57	18	a	a	DET
ejpam-4568	57	19	new	new	ADJ
ejpam-4568	57	20	formula	formula	NOUN
ejpam-4568	57	21	for	for	ADP
ejpam-4568	57	22	conjugate	conjugate	ADJ
ejpam-4568	57	23	gradient	gradient	NOUN
ejpam-4568	57	24	method	method	NOUN
ejpam-4568	57	25	with	with	ADP
ejpam-4568	57	26	wolfe	wolfe	PROPN
ejpam-4568	57	27	–	–	PUNCT
ejpam-4568	57	28	powell	powell	PROPN
ejpam-4568	57	29	generate	generate	VERB
ejpam-4568	57	30	a	a	DET
ejpam-4568	57	31	descent	descent	NOUN
ejpam-4568	57	32	direction	direction	NOUN
ejpam-4568	57	33	at	at	ADP
ejpam-4568	57	34	each	each	DET
ejpam-4568	57	35	iteration	iteration	NOUN
ejpam-4568	57	36	in	in	ADP
ejpam-4568	57	37	section	section	NOUN
ejpam-4568	57	38	2	2	NUM
ejpam-4568	57	39	and	and	CCONJ
ejpam-4568	57	40	the	the	DET
ejpam-4568	57	41	new	new	ADJ
ejpam-4568	57	42	formula	formula	NOUN
ejpam-4568	57	43	for	for	ADP
ejpam-4568	57	44	conjugate	conjugate	ADJ
ejpam-4568	57	45	gradient	gradient	NOUN
ejpam-4568	57	46	method	method	NOUN
ejpam-4568	57	47	which	which	PRON
ejpam-4568	57	48	is	be	AUX
ejpam-4568	57	49	satisfied	satisfied	ADJ
ejpam-4568	57	50	the	the	DET
ejpam-4568	57	51	global	global	ADJ
ejpam-4568	57	52	convergence	convergence	NOUN
ejpam-4568	57	53	in	in	ADP
ejpam-4568	57	54	section	section	NOUN
ejpam-4568	57	55	3	3	NUM
ejpam-4568	57	56	.	.	PUNCT
ejpam-4568	58	1	in	in	ADP
ejpam-4568	58	2	section	section	NOUN
ejpam-4568	58	3	4	4	NUM
ejpam-4568	58	4	,	,	PUNCT
ejpam-4568	58	5	we	we	PRON
ejpam-4568	58	6	present	present	VERB
ejpam-4568	58	7	the	the	DET
ejpam-4568	58	8	numerical	numerical	ADJ
ejpam-4568	58	9	behavior	behavior	NOUN
ejpam-4568	58	10	of	of	ADP
ejpam-4568	58	11	the	the	DET
ejpam-4568	58	12	method	method	NOUN
ejpam-4568	58	13	.	.	PUNCT
ejpam-4568	59	1	the	the	DET
ejpam-4568	59	2	last	last	ADJ
ejpam-4568	59	3	section	section	NOUN
ejpam-4568	59	4	proposes	propose	VERB
ejpam-4568	59	5	one	one	NUM
ejpam-4568	59	6	conclusion	conclusion	NOUN
ejpam-4568	59	7	.	.	PUNCT
ejpam-4568	60	1	2	2	X
ejpam-4568	60	2	.	.	X
ejpam-4568	60	3	propose	propose	VERB
ejpam-4568	60	4	new	new	ADJ
ejpam-4568	60	5	conjugate	conjugate	ADJ
ejpam-4568	60	6	gradient	gradient	NOUN
ejpam-4568	60	7	method	method	NOUN
ejpam-4568	60	8	:	:	PUNCT
ejpam-4568	60	9	we	we	PRON
ejpam-4568	60	10	will	will	AUX
ejpam-4568	60	11	discuss	discuss	VERB
ejpam-4568	60	12	the	the	DET
ejpam-4568	60	13	new	new	ADJ
ejpam-4568	60	14	parameter	parameter	NOUN
ejpam-4568	60	15	choice	choice	NOUN
ejpam-4568	60	16	.	.	PUNCT
ejpam-4568	61	1	now	now	ADV
ejpam-4568	61	2	using	use	VERB
ejpam-4568	61	3	the	the	DET
ejpam-4568	61	4	quadratic	quadratic	ADJ
ejpam-4568	61	5	formula	formula	NOUN
ejpam-4568	61	6	for	for	ADP
ejpam-4568	61	7	the	the	DET
ejpam-4568	61	8	objective	objective	ADJ
ejpam-4568	61	9	function	function	NOUN
ejpam-4568	61	10	f(x	f(x	PROPN
ejpam-4568	61	11	)	)	PUNCT
ejpam-4568	61	12	we	we	PRON
ejpam-4568	61	13	have	have	AUX
ejpam-4568	61	14	:	:	PUNCT
ejpam-4568	61	15	fk+1	fk+1	NUM
ejpam-4568	61	16	=	=	SYM
ejpam-4568	61	17	fk	fk	NOUN
ejpam-4568	62	1	+	+	NOUN
ejpam-4568	62	2	stk	stk	PROPN
ejpam-4568	62	3	gk	gk	PROPN
ejpam-4568	62	4	+	+	NOUN
ejpam-4568	62	5	1	1	NUM
ejpam-4568	62	6	2	2	NUM
ejpam-4568	62	7	stkq(uk)sk	stkq(uk)sk	PROPN
ejpam-4568	62	8	(	(	PUNCT
ejpam-4568	62	9	11	11	NUM
ejpam-4568	62	10	)	)	PUNCT
ejpam-4568	62	11	b.	b.	PROPN
ejpam-4568	62	12	a.	a.	PROPN
ejpam-4568	62	13	hassan	hassan	PROPN
ejpam-4568	62	14	,	,	PUNCT
ejpam-4568	62	15	h.	h.	PROPN
ejpam-4568	62	16	m.	m.	PROPN
ejpam-4568	62	17	sadiq	sadiq	PROPN
ejpam-4568	62	18	/	/	SYM
ejpam-4568	62	19	eur	eur	PROPN
ejpam-4568	62	20	.	.	PUNCT
ejpam-4568	63	1	j.	j.	PROPN
ejpam-4568	63	2	pure	pure	PROPN
ejpam-4568	63	3	appl	appl	PROPN
ejpam-4568	63	4	.	.	PROPN
ejpam-4568	63	5	math	math	PROPN
ejpam-4568	63	6	,	,	PUNCT
ejpam-4568	63	7	15	15	NUM
ejpam-4568	63	8	(	(	PUNCT
ejpam-4568	63	9	4	4	NUM
ejpam-4568	63	10	)	)	PUNCT
ejpam-4568	63	11	(	(	PUNCT
ejpam-4568	63	12	2022	2022	NUM
ejpam-4568	63	13	)	)	PUNCT
ejpam-4568	63	14	,	,	PUNCT
ejpam-4568	63	15	2011	2011	NUM
ejpam-4568	63	16	-	-	SYM
ejpam-4568	63	17	2021	2021	NUM
ejpam-4568	63	18	2014	2014	NUM
ejpam-4568	63	19	where	where	SCONJ
ejpam-4568	63	20	q(uk	q(uk	VERB
ejpam-4568	63	21	)	)	PUNCT
ejpam-4568	63	22	is	be	AUX
ejpam-4568	63	23	the	the	DET
ejpam-4568	63	24	hessian	hessian	ADJ
ejpam-4568	63	25	matrix	matrix	NOUN
ejpam-4568	63	26	of	of	ADP
ejpam-4568	63	27	second	second	ADJ
ejpam-4568	63	28	-	-	PUNCT
ejpam-4568	63	29	order	order	NOUN
ejpam-4568	63	30	derivatives	derivative	NOUN
ejpam-4568	63	31	.	.	PUNCT
ejpam-4568	64	1	we	we	PRON
ejpam-4568	64	2	derivative	derivative	VERB
ejpam-4568	64	3	both	both	DET
ejpam-4568	64	4	sides	side	NOUN
ejpam-4568	64	5	of	of	ADP
ejpam-4568	64	6	(	(	PUNCT
ejpam-4568	64	7	11	11	NUM
ejpam-4568	64	8	)	)	PUNCT
ejpam-4568	64	9	for	for	ADP
ejpam-4568	64	10	sk	sk	NOUN
ejpam-4568	64	11	,	,	PUNCT
ejpam-4568	64	12	is	be	AUX
ejpam-4568	64	13	presented	present	VERB
ejpam-4568	64	14	as	as	SCONJ
ejpam-4568	64	15	follows	follow	VERB
ejpam-4568	64	16	:	:	PUNCT
ejpam-4568	65	1	∇fk+1	∇fk+1	PROPN
ejpam-4568	65	2	=	=	PROPN
ejpam-4568	65	3	gk	gk	PROPN
ejpam-4568	65	4	+	+	NOUN
ejpam-4568	65	5	q(uk)sk	q(uk)sk	NOUN
ejpam-4568	65	6	=	=	SYM
ejpam-4568	65	7	0	0	NUM
ejpam-4568	65	8	(	(	PUNCT
ejpam-4568	65	9	12	12	NUM
ejpam-4568	65	10	)	)	PUNCT
ejpam-4568	65	11	using	use	VERB
ejpam-4568	65	12	(	(	PUNCT
ejpam-4568	65	13	5	5	NUM
ejpam-4568	65	14	)	)	PUNCT
ejpam-4568	65	15	and	and	CCONJ
ejpam-4568	65	16	(	(	PUNCT
ejpam-4568	65	17	12	12	NUM
ejpam-4568	65	18	)	)	PUNCT
ejpam-4568	65	19	in	in	ADP
ejpam-4568	65	20	(	(	PUNCT
ejpam-4568	65	21	11	11	NUM
ejpam-4568	65	22	)	)	PUNCT
ejpam-4568	65	23	,	,	PUNCT
ejpam-4568	65	24	we	we	PRON
ejpam-4568	65	25	have	have	VERB
ejpam-4568	65	26	:	:	PUNCT
ejpam-4568	66	1	stkq(uk)sk	stkq(uk)sk	AUX
ejpam-4568	66	2	=	=	SYM
ejpam-4568	66	3	fk	fk	INTJ
ejpam-4568	66	4	−	−	NOUN
ejpam-4568	66	5	fk+1	fk+1	NOUN
ejpam-4568	66	6	+	+	CCONJ
ejpam-4568	66	7	1	1	NUM
ejpam-4568	66	8	2	2	NUM
ejpam-4568	66	9	stk	stk	PROPN
ejpam-4568	66	10	yk	yk	PROPN
ejpam-4568	66	11	(	(	PUNCT
ejpam-4568	66	12	13	13	NUM
ejpam-4568	66	13	)	)	PUNCT
ejpam-4568	66	14	after	after	ADP
ejpam-4568	66	15	some	some	DET
ejpam-4568	66	16	algebra	algebra	NOUN
ejpam-4568	66	17	from	from	ADP
ejpam-4568	66	18	(	(	PUNCT
ejpam-4568	66	19	11	11	NUM
ejpam-4568	66	20	)	)	PUNCT
ejpam-4568	66	21	and	and	CCONJ
ejpam-4568	66	22	(	(	PUNCT
ejpam-4568	66	23	13	13	NUM
ejpam-4568	66	24	)	)	PUNCT
ejpam-4568	66	25	,	,	PUNCT
ejpam-4568	66	26	as	as	ADP
ejpam-4568	66	27	a	a	DET
ejpam-4568	66	28	result	result	NOUN
ejpam-4568	66	29	:	:	PUNCT
ejpam-4568	66	30	βk	βk	ADP
ejpam-4568	66	31	=	=	NOUN
ejpam-4568	66	32	1	1	NUM
ejpam-4568	66	33	2	2	NUM
ejpam-4568	66	34	g	g	NOUN
ejpam-4568	66	35	t	t	PROPN
ejpam-4568	66	36	k+1yk	k+1yk	PROPN
ejpam-4568	66	37	+	+	CCONJ
ejpam-4568	66	38	(	(	PUNCT
ejpam-4568	66	39	fk−fk+1	fk−fk+1	NOUN
ejpam-4568	66	40	)	)	PUNCT
ejpam-4568	66	41	stk	stk	PROPN
ejpam-4568	66	42	yk	yk	PROPN
ejpam-4568	66	43	gtk+1yk	gtk+1yk	PROPN
ejpam-4568	66	44	dtk	dtk	PROPN
ejpam-4568	66	45	yk	yk	PROPN
ejpam-4568	66	46	.	.	PUNCT
ejpam-4568	67	1	(	(	PUNCT
ejpam-4568	67	2	14	14	NUM
ejpam-4568	67	3	)	)	PUNCT
ejpam-4568	67	4	if	if	SCONJ
ejpam-4568	67	5	exact	exact	ADJ
ejpam-4568	67	6	line	line	NOUN
ejpam-4568	67	7	search	search	NOUN
ejpam-4568	67	8	is	be	AUX
ejpam-4568	67	9	utilized	utilize	VERB
ejpam-4568	67	10	,	,	PUNCT
ejpam-4568	67	11	then	then	ADV
ejpam-4568	67	12	βk	βk	NOUN
ejpam-4568	67	13	is	be	AUX
ejpam-4568	67	14	such	such	ADJ
ejpam-4568	67	15	that	that	SCONJ
ejpam-4568	67	16	:	:	PUNCT
ejpam-4568	67	17	βbv	βbv	VERB
ejpam-4568	67	18	1	1	NUM
ejpam-4568	67	19	k	k	NOUN
ejpam-4568	67	20	=	=	SYM
ejpam-4568	67	21	1	1	NUM
ejpam-4568	67	22	2∥gk+1∥2	2∥gk+1∥2	NUM
ejpam-4568	67	23	+	+	CCONJ
ejpam-4568	67	24	(	(	PUNCT
ejpam-4568	67	25	fk−fk+1	fk−fk+1	NOUN
ejpam-4568	67	26	)	)	PUNCT
ejpam-4568	67	27	stk	stk	PROPN
ejpam-4568	67	28	yk	yk	PROPN
ejpam-4568	67	29	∥gk+1∥2	∥gk+1∥2	PROPN
ejpam-4568	67	30	dtk	dtk	PROPN
ejpam-4568	67	31	yk	yk	PROPN
ejpam-4568	67	32	.	.	PUNCT
ejpam-4568	68	1	(	(	PUNCT
ejpam-4568	68	2	15	15	NUM
ejpam-4568	68	3	)	)	PUNCT
ejpam-4568	68	4	in	in	ADP
ejpam-4568	68	5	particular	particular	ADJ
ejpam-4568	68	6	,	,	PUNCT
ejpam-4568	68	7	conclude	conclude	VERB
ejpam-4568	68	8	:	:	PUNCT
ejpam-4568	68	9	βbv	βbv	VERB
ejpam-4568	68	10	2	2	NUM
ejpam-4568	68	11	k	k	NOUN
ejpam-4568	68	12	=	=	SYM
ejpam-4568	68	13	1	1	NUM
ejpam-4568	68	14	2∥gk+1∥2	2∥gk+1∥2	NUM
ejpam-4568	68	15	−	−	NOUN
ejpam-4568	68	16	(	(	PUNCT
ejpam-4568	68	17	fk−fk+1	fk−fk+1	NOUN
ejpam-4568	68	18	)	)	PUNCT
ejpam-4568	68	19	stk	stk	PROPN
ejpam-4568	68	20	gk	gk	PROPN
ejpam-4568	68	21	∥gk+1∥2	∥gk+1∥2	PROPN
ejpam-4568	69	1	−dtk	−dtk	ADV
ejpam-4568	69	2	gk	gk	PROPN
ejpam-4568	69	3	(	(	PUNCT
ejpam-4568	69	4	16	16	NUM
ejpam-4568	69	5	)	)	PUNCT
ejpam-4568	69	6	and	and	CCONJ
ejpam-4568	69	7	βbv	βbv	VERB
ejpam-4568	69	8	3	3	NUM
ejpam-4568	69	9	k	k	NOUN
ejpam-4568	69	10	=	=	SYM
ejpam-4568	69	11	1	1	NUM
ejpam-4568	69	12	2∥gk+1∥2	2∥gk+1∥2	NUM
ejpam-4568	69	13	+	+	CCONJ
ejpam-4568	69	14	(	(	PUNCT
ejpam-4568	69	15	fk−fk+1	fk−fk+1	NOUN
ejpam-4568	69	16	)	)	PUNCT
ejpam-4568	69	17	ak∥gk∥2	ak∥gk∥2	NOUN
ejpam-4568	69	18	∥gk+1∥2	∥gk+1∥2	PROPN
ejpam-4568	69	19	∥gk∥2	∥gk∥2	INTJ
ejpam-4568	69	20	.	.	PUNCT
ejpam-4568	70	1	(	(	PUNCT
ejpam-4568	70	2	17	17	NUM
ejpam-4568	70	3	)	)	PUNCT
ejpam-4568	70	4	thus	thus	ADV
ejpam-4568	70	5	,	,	PUNCT
ejpam-4568	70	6	bv1	bv1	PROPN
ejpam-4568	70	7	,	,	PUNCT
ejpam-4568	70	8	bv2	bv2	NOUN
ejpam-4568	70	9	and	and	CCONJ
ejpam-4568	70	10	bv3	bv3	ADV
ejpam-4568	70	11	are	be	AUX
ejpam-4568	70	12	the	the	DET
ejpam-4568	70	13	new	new	ADJ
ejpam-4568	70	14	parameters	parameter	NOUN
ejpam-4568	70	15	of	of	ADP
ejpam-4568	70	16	conjugate	conjugate	ADJ
ejpam-4568	70	17	gradient	gradient	NOUN
ejpam-4568	70	18	.	.	PUNCT
ejpam-4568	71	1	based	base	VERB
ejpam-4568	71	2	on	on	ADP
ejpam-4568	71	3	the	the	DET
ejpam-4568	71	4	above	above	ADJ
ejpam-4568	71	5	discussions	discussion	NOUN
ejpam-4568	71	6	,	,	PUNCT
ejpam-4568	71	7	the	the	DET
ejpam-4568	71	8	presented	present	VERB
ejpam-4568	71	9	algorithm	algorithm	NOUN
ejpam-4568	71	10	is	be	AUX
ejpam-4568	71	11	stated	state	VERB
ejpam-4568	71	12	as	as	SCONJ
ejpam-4568	71	13	follows	follow	VERB
ejpam-4568	71	14	:	:	PUNCT
ejpam-4568	71	15	step	step	NOUN
ejpam-4568	71	16	1	1	NUM
ejpam-4568	71	17	:	:	PUNCT
ejpam-4568	71	18	given	give	VERB
ejpam-4568	71	19	a	a	DET
ejpam-4568	71	20	starting	starting	NOUN
ejpam-4568	71	21	point	point	NOUN
ejpam-4568	71	22	u1	u1	NOUN
ejpam-4568	71	23	.	.	PUNCT
ejpam-4568	72	1	set	set	VERB
ejpam-4568	72	2	k	k	PROPN
ejpam-4568	73	1	=	=	PUNCT
ejpam-4568	73	2	0	0	NUM
ejpam-4568	73	3	and	and	CCONJ
ejpam-4568	73	4	d	d	NOUN
ejpam-4568	73	5	◦	◦	NOUN
ejpam-4568	73	6	=	=	SYM
ejpam-4568	73	7	−g	−g	NOUN
ejpam-4568	73	8	◦	◦	NOUN
ejpam-4568	73	9	.	.	PUNCT
ejpam-4568	74	1	step	step	NOUN
ejpam-4568	74	2	2	2	NUM
ejpam-4568	74	3	:	:	PUNCT
ejpam-4568	74	4	compute	compute	VERB
ejpam-4568	74	5	βk	βk	ADV
ejpam-4568	74	6	by	by	ADP
ejpam-4568	74	7	(	(	PUNCT
ejpam-4568	74	8	15	15	NUM
ejpam-4568	74	9	)	)	PUNCT
ejpam-4568	74	10	,	,	PUNCT
ejpam-4568	74	11	(	(	PUNCT
ejpam-4568	74	12	16	16	NUM
ejpam-4568	74	13	)	)	PUNCT
ejpam-4568	74	14	and	and	CCONJ
ejpam-4568	74	15	(	(	PUNCT
ejpam-4568	74	16	17	17	NUM
ejpam-4568	74	17	)	)	PUNCT
ejpam-4568	74	18	.	.	PUNCT
ejpam-4568	75	1	step	step	NOUN
ejpam-4568	75	2	3	3	NUM
ejpam-4568	75	3	:	:	PUNCT
ejpam-4568	75	4	compute	compute	NOUN
ejpam-4568	75	5	dk	dk	PROPN
ejpam-4568	75	6	by	by	ADP
ejpam-4568	75	7	(	(	PUNCT
ejpam-4568	75	8	8)	8)	NUM
ejpam-4568	75	9	and	and	CCONJ
ejpam-4568	75	10	(	(	PUNCT
ejpam-4568	75	11	15	15	NUM
ejpam-4568	75	12	)	)	PUNCT
ejpam-4568	75	13	.	.	PUNCT
ejpam-4568	76	1	if	if	SCONJ
ejpam-4568	76	2	∥gk∥	∥gk∥	NUM
ejpam-4568	76	3	=	=	SYM
ejpam-4568	76	4	0	0	NUM
ejpam-4568	76	5	,	,	PUNCT
ejpam-4568	76	6	then	then	ADV
ejpam-4568	76	7	stop	stop	VERB
ejpam-4568	76	8	.	.	PUNCT
ejpam-4568	77	1	step	step	VERB
ejpam-4568	77	2	4	4	NUM
ejpam-4568	77	3	:	:	PUNCT
ejpam-4568	77	4	evaluate	evaluate	VERB
ejpam-4568	77	5	αk	αk	NOUN
ejpam-4568	77	6	satisfy	satisfy	VERB
ejpam-4568	77	7	the	the	DET
ejpam-4568	77	8	conditions	condition	NOUN
ejpam-4568	77	9	(	(	PUNCT
ejpam-4568	77	10	6	6	NUM
ejpam-4568	77	11	)	)	PUNCT
ejpam-4568	77	12	and	and	CCONJ
ejpam-4568	77	13	(	(	PUNCT
ejpam-4568	77	14	7	7	NUM
ejpam-4568	77	15	)	)	PUNCT
ejpam-4568	77	16	.	.	PUNCT
ejpam-4568	78	1	step	step	NOUN
ejpam-4568	78	2	5	5	NUM
ejpam-4568	78	3	:	:	PUNCT
ejpam-4568	78	4	update	update	VERB
ejpam-4568	78	5	the	the	DET
ejpam-4568	78	6	new	new	ADJ
ejpam-4568	78	7	point	point	NOUN
ejpam-4568	78	8	by	by	ADP
ejpam-4568	78	9	the	the	DET
ejpam-4568	78	10	recurrence	recurrence	NOUN
ejpam-4568	78	11	expression	expression	NOUN
ejpam-4568	78	12	(	(	PUNCT
ejpam-4568	78	13	4	4	NUM
ejpam-4568	78	14	)	)	PUNCT
ejpam-4568	78	15	.	.	PUNCT
ejpam-4568	79	1	step	step	VERB
ejpam-4568	79	2	6	6	NUM
ejpam-4568	79	3	:	:	PUNCT
ejpam-4568	79	4	if	if	SCONJ
ejpam-4568	79	5	f(uk+1	f(uk+1	NOUN
ejpam-4568	79	6	)	)	PUNCT
ejpam-4568	79	7	<	<	X
ejpam-4568	79	8	f(uk	f(uk	PROPN
ejpam-4568	79	9	)	)	PUNCT
ejpam-4568	79	10	and	and	CCONJ
ejpam-4568	79	11	∥gk∥	∥gk∥	NOUN
ejpam-4568	79	12	<	<	X
ejpam-4568	79	13	ε	ε	PROPN
ejpam-4568	79	14	then	then	ADV
ejpam-4568	79	15	stop	stop	VERB
ejpam-4568	79	16	.	.	PUNCT
ejpam-4568	80	1	otherwise	otherwise	ADV
ejpam-4568	80	2	go	go	VERB
ejpam-4568	80	3	to	to	PART
ejpam-4568	80	4	step	step	VERB
ejpam-4568	80	5	2	2	NUM
ejpam-4568	80	6	with	with	ADP
ejpam-4568	80	7	k	k	PROPN
ejpam-4568	80	8	=	=	PUNCT
ejpam-4568	80	9	k+1	k+1	PROPN
ejpam-4568	80	10	.	.	PUNCT
ejpam-4568	81	1	an	an	DET
ejpam-4568	81	2	important	important	ADJ
ejpam-4568	81	3	feature	feature	NOUN
ejpam-4568	81	4	for	for	ADP
ejpam-4568	81	5	any	any	DET
ejpam-4568	81	6	minimization	minimization	NOUN
ejpam-4568	81	7	algorithm	algorithm	NOUN
ejpam-4568	81	8	is	be	AUX
ejpam-4568	81	9	the	the	DET
ejpam-4568	81	10	descent	descent	NOUN
ejpam-4568	81	11	or	or	CCONJ
ejpam-4568	81	12	the	the	DET
ejpam-4568	81	13	sufficient	sufficient	ADJ
ejpam-4568	81	14	descent	descent	NOUN
ejpam-4568	81	15	property	property	NOUN
ejpam-4568	81	16	.	.	PUNCT
ejpam-4568	82	1	the	the	DET
ejpam-4568	82	2	following	follow	VERB
ejpam-4568	82	3	theorem	theorem	NOUN
ejpam-4568	82	4	indicates	indicate	VERB
ejpam-4568	82	5	that	that	SCONJ
ejpam-4568	82	6	search	search	NOUN
ejpam-4568	82	7	direction	direction	NOUN
ejpam-4568	82	8	dk	dk	PROPN
ejpam-4568	82	9	satisfies	satisfy	VERB
ejpam-4568	82	10	descent	descent	NOUN
ejpam-4568	82	11	property	property	NOUN
ejpam-4568	82	12	of	of	ADP
ejpam-4568	82	13	our	our	PRON
ejpam-4568	82	14	algorithms	algorithm	NOUN
ejpam-4568	82	15	.	.	PUNCT
ejpam-4568	83	1	theorem	theorem	NOUN
ejpam-4568	83	2	1	1	NUM
ejpam-4568	83	3	.	.	PUNCT
ejpam-4568	84	1	in	in	ADP
ejpam-4568	84	2	the	the	DET
ejpam-4568	84	3	algorithm	algorithm	NOUN
ejpam-4568	84	4	(	(	PUNCT
ejpam-4568	84	5	4	4	NUM
ejpam-4568	84	6	)	)	PUNCT
ejpam-4568	84	7	,	,	PUNCT
ejpam-4568	84	8	(	(	PUNCT
ejpam-4568	84	9	8)	8)	NUM
ejpam-4568	84	10	,	,	PUNCT
ejpam-4568	84	11	(	(	PUNCT
ejpam-4568	84	12	15	15	NUM
ejpam-4568	84	13	)	)	PUNCT
ejpam-4568	84	14	,	,	PUNCT
ejpam-4568	84	15	assume	assume	VERB
ejpam-4568	84	16	that	that	SCONJ
ejpam-4568	84	17	αk	αk	NOUN
ejpam-4568	84	18	determined	determine	VERB
ejpam-4568	84	19	by	by	ADP
ejpam-4568	84	20	the	the	DET
ejpam-4568	84	21	wolfe	wolfe	PROPN
ejpam-4568	84	22	line	line	PROPN
ejpam-4568	84	23	search	search	NOUN
ejpam-4568	84	24	(	(	PUNCT
ejpam-4568	84	25	6)-(7	6)-(7	NOUN
ejpam-4568	84	26	)	)	PUNCT
ejpam-4568	84	27	then	then	ADV
ejpam-4568	84	28	the	the	DET
ejpam-4568	84	29	direction	direction	NOUN
ejpam-4568	84	30	dk+1	dk+1	NOUN
ejpam-4568	84	31	given	give	VERB
ejpam-4568	84	32	by	by	ADP
ejpam-4568	84	33	(	(	PUNCT
ejpam-4568	84	34	8)	8)	NUM
ejpam-4568	84	35	is	be	AUX
ejpam-4568	84	36	a	a	DET
ejpam-4568	84	37	descent	descent	NOUN
ejpam-4568	84	38	direction	direction	NOUN
ejpam-4568	84	39	.	.	PUNCT
ejpam-4568	85	1	b.	b.	PROPN
ejpam-4568	85	2	a.	a.	PROPN
ejpam-4568	85	3	hassan	hassan	PROPN
ejpam-4568	85	4	,	,	PUNCT
ejpam-4568	85	5	h.	h.	PROPN
ejpam-4568	85	6	m.	m.	PROPN
ejpam-4568	85	7	sadiq	sadiq	PROPN
ejpam-4568	85	8	/	/	SYM
ejpam-4568	85	9	eur	eur	PROPN
ejpam-4568	85	10	.	.	PUNCT
ejpam-4568	86	1	j.	j.	PROPN
ejpam-4568	86	2	pure	pure	PROPN
ejpam-4568	86	3	appl	appl	PROPN
ejpam-4568	86	4	.	.	PROPN
ejpam-4568	86	5	math	math	PROPN
ejpam-4568	86	6	,	,	PUNCT
ejpam-4568	86	7	15	15	NUM
ejpam-4568	86	8	(	(	PUNCT
ejpam-4568	86	9	4	4	NUM
ejpam-4568	86	10	)	)	PUNCT
ejpam-4568	86	11	(	(	PUNCT
ejpam-4568	86	12	2022	2022	NUM
ejpam-4568	86	13	)	)	PUNCT
ejpam-4568	86	14	,	,	PUNCT
ejpam-4568	86	15	2011	2011	NUM
ejpam-4568	86	16	-	-	SYM
ejpam-4568	86	17	2021	2021	NUM
ejpam-4568	86	18	2015	2015	NUM
ejpam-4568	86	19	proof	proof	NOUN
ejpam-4568	86	20	.	.	PUNCT
ejpam-4568	87	1	if	if	SCONJ
ejpam-4568	87	2	k	k	PROPN
ejpam-4568	87	3	=	=	SYM
ejpam-4568	87	4	0	0	NUM
ejpam-4568	87	5	,	,	PUNCT
ejpam-4568	87	6	then	then	ADV
ejpam-4568	87	7	d1	d1	PROPN
ejpam-4568	87	8	=	=	PUNCT
ejpam-4568	87	9	−g1	−g1	VERB
ejpam-4568	87	10	,	,	PUNCT
ejpam-4568	87	11	so	so	ADV
ejpam-4568	87	12	dt1	dt1	PROPN
ejpam-4568	87	13	g1	g1	PROPN
ejpam-4568	87	14	=	=	PUNCT
ejpam-4568	87	15	−∥g1∥2	−∥g1∥2	ADJ
ejpam-4568	87	16	<	<	X
ejpam-4568	87	17	0	0	X
ejpam-4568	87	18	.	.	PUNCT
ejpam-4568	87	19	suppose	suppose	VERB
ejpam-4568	87	20	that	that	SCONJ
ejpam-4568	87	21	dtk	dtk	PROPN
ejpam-4568	87	22	gk	gk	PROPN
ejpam-4568	87	23	<	<	X
ejpam-4568	87	24	0	0	PUNCT
ejpam-4568	87	25	for	for	ADP
ejpam-4568	87	26	all	all	DET
ejpam-4568	87	27	k.	k.	PROPN
ejpam-4568	87	28	multiply	multiply	PROPN
ejpam-4568	87	29	(	(	PUNCT
ejpam-4568	87	30	8)	8)	NUM
ejpam-4568	87	31	by	by	ADP
ejpam-4568	87	32	gtk+1	gtk+1	NOUN
ejpam-4568	87	33	,	,	PUNCT
ejpam-4568	87	34	will	will	AUX
ejpam-4568	87	35	give	give	VERB
ejpam-4568	87	36	:	:	PUNCT
ejpam-4568	87	37	dtk+1gk+1	dtk+1gk+1	PROPN
ejpam-4568	87	38	=	=	SYM
ejpam-4568	87	39	−∥gk+1∥2	−∥gk+1∥2	PROPN
ejpam-4568	87	40	+	+	CCONJ
ejpam-4568	87	41	βbv	βbv	NOUN
ejpam-4568	87	42	1	1	NUM
ejpam-4568	87	43	k	k	PROPN
ejpam-4568	87	44	dtk	dtk	PROPN
ejpam-4568	87	45	gk+1	gk+1	NOUN
ejpam-4568	87	46	(	(	PUNCT
ejpam-4568	87	47	18	18	NUM
ejpam-4568	87	48	)	)	PUNCT
ejpam-4568	87	49	since	since	SCONJ
ejpam-4568	87	50	∥gk+1∥2	∥gk+1∥2	NOUN
ejpam-4568	87	51	=	=	SYM
ejpam-4568	87	52	βkd	βkd	PROPN
ejpam-4568	88	1	t	t	PROPN
ejpam-4568	88	2	k	k	PROPN
ejpam-4568	88	3	yk	yk	PROPN
ejpam-4568	88	4	(	(	PUNCT
ejpam-4568	88	5	19	19	NUM
ejpam-4568	88	6	)	)	PUNCT
ejpam-4568	88	7	now	now	ADV
ejpam-4568	88	8	,	,	PUNCT
ejpam-4568	88	9	put	put	VERB
ejpam-4568	88	10	(	(	PUNCT
ejpam-4568	88	11	19	19	NUM
ejpam-4568	88	12	)	)	PUNCT
ejpam-4568	88	13	in	in	ADP
ejpam-4568	88	14	(	(	PUNCT
ejpam-4568	88	15	18	18	NUM
ejpam-4568	88	16	)	)	PUNCT
ejpam-4568	88	17	we	we	PRON
ejpam-4568	88	18	obtain	obtain	VERB
ejpam-4568	88	19	:	:	PUNCT
ejpam-4568	88	20	dtk+1gk+1	dtk+1gk+1	PROPN
ejpam-4568	89	1	=	=	SYM
ejpam-4568	89	2	−βkd	−βkd	NOUN
ejpam-4568	89	3	t	t	PROPN
ejpam-4568	89	4	k	k	PROPN
ejpam-4568	89	5	yk	yk	PROPN
ejpam-4568	89	6	+	+	CCONJ
ejpam-4568	89	7	βkd	βkd	PROPN
ejpam-4568	89	8	t	t	PROPN
ejpam-4568	89	9	k	k	NOUN
ejpam-4568	89	10	gk+1	gk+1	NOUN
ejpam-4568	89	11	=	=	PUNCT
ejpam-4568	89	12	βk[−dtk	βk[−dtk	SYM
ejpam-4568	89	13	yk	yk	PROPN
ejpam-4568	89	14	+	+	NUM
ejpam-4568	89	15	dtk	dtk	PROPN
ejpam-4568	89	16	gk+1	gk+1	AUX
ejpam-4568	89	17	]	]	X
ejpam-4568	89	18	=	=	PUNCT
ejpam-4568	90	1	βkd	βkd	PROPN
ejpam-4568	90	2	t	t	PROPN
ejpam-4568	90	3	k	k	PROPN
ejpam-4568	90	4	gk	gk	PROPN
ejpam-4568	90	5	<	<	X
ejpam-4568	90	6	0	0	NUM
ejpam-4568	91	1	(	(	PUNCT
ejpam-4568	91	2	20	20	NUM
ejpam-4568	91	3	)	)	PUNCT
ejpam-4568	91	4	the	the	DET
ejpam-4568	91	5	proof	proof	NOUN
ejpam-4568	91	6	is	be	AUX
ejpam-4568	91	7	complete	complete	ADJ
ejpam-4568	91	8	.	.	PUNCT
ejpam-4568	92	1	the	the	DET
ejpam-4568	92	2	proof	proof	ADJ
ejpam-4568	92	3	descent	descent	NOUN
ejpam-4568	92	4	property	property	NOUN
ejpam-4568	92	5	of	of	ADP
ejpam-4568	92	6	bv2	bv2	NOUN
ejpam-4568	92	7	and	and	CCONJ
ejpam-4568	92	8	bv3	bv3	ADV
ejpam-4568	92	9	is	be	AUX
ejpam-4568	92	10	similar	similar	ADJ
ejpam-4568	92	11	to	to	ADP
ejpam-4568	92	12	proof	proof	NOUN
ejpam-4568	92	13	bv1	bv1	PROPN
ejpam-4568	92	14	.	.	PROPN
ejpam-4568	93	1	3	3	X
ejpam-4568	93	2	.	.	X
ejpam-4568	93	3	convergence	convergence	NOUN
ejpam-4568	93	4	analysis	analysis	NOUN
ejpam-4568	93	5	:	:	PUNCT
ejpam-4568	93	6	in	in	ADP
ejpam-4568	93	7	order	order	NOUN
ejpam-4568	93	8	to	to	PART
ejpam-4568	93	9	establish	establish	VERB
ejpam-4568	93	10	the	the	DET
ejpam-4568	93	11	global	global	ADJ
ejpam-4568	93	12	convergence	convergence	NOUN
ejpam-4568	93	13	property	property	NOUN
ejpam-4568	93	14	of	of	ADP
ejpam-4568	93	15	the	the	DET
ejpam-4568	93	16	method	method	NOUN
ejpam-4568	93	17	,	,	PUNCT
ejpam-4568	93	18	we	we	PRON
ejpam-4568	93	19	make	make	VERB
ejpam-4568	93	20	the	the	DET
ejpam-4568	93	21	following	follow	VERB
ejpam-4568	93	22	standard	standard	ADJ
ejpam-4568	93	23	assumptions	assumption	NOUN
ejpam-4568	93	24	for	for	ADP
ejpam-4568	93	25	the	the	DET
ejpam-4568	93	26	objective	objective	ADJ
ejpam-4568	93	27	function	function	NOUN
ejpam-4568	93	28	:	:	PUNCT
ejpam-4568	93	29	•	•	NOUN
ejpam-4568	93	30	for	for	ADP
ejpam-4568	93	31	any	any	DET
ejpam-4568	93	32	initial	initial	ADJ
ejpam-4568	93	33	point	point	NOUN
ejpam-4568	93	34	x1	x1	PROPN
ejpam-4568	93	35	∈	∈	PROPN
ejpam-4568	93	36	rn	rn	PROPN
ejpam-4568	93	37	,	,	PUNCT
ejpam-4568	93	38	the	the	DET
ejpam-4568	93	39	level	level	NOUN
ejpam-4568	93	40	set	set	VERB
ejpam-4568	93	41	o	o	X
ejpam-4568	93	42	=	=	PUNCT
ejpam-4568	93	43	{	{	PUNCT
ejpam-4568	93	44	x	x	PROPN
ejpam-4568	93	45	∈	∈	PROPN
ejpam-4568	93	46	rn\f(x	rn\f(x	PROPN
ejpam-4568	93	47	)	)	PUNCT
ejpam-4568	93	48	<	<	X
ejpam-4568	93	49	f(x1	f(x1	NOUN
ejpam-4568	93	50	)	)	PUNCT
ejpam-4568	93	51	}	}	PUNCT
ejpam-4568	93	52	is	be	AUX
ejpam-4568	93	53	bounded	bound	VERB
ejpam-4568	93	54	.	.	PUNCT
ejpam-4568	94	1	•	•	NUM
ejpam-4568	94	2	f(x	f(x	PROPN
ejpam-4568	94	3	)	)	PUNCT
ejpam-4568	94	4	is	be	AUX
ejpam-4568	94	5	continuously	continuously	ADV
ejpam-4568	94	6	differentiable	differentiable	ADJ
ejpam-4568	94	7	in	in	ADP
ejpam-4568	94	8	a	a	DET
ejpam-4568	94	9	neighborhood	neighborhood	NOUN
ejpam-4568	94	10	u	u	NOUN
ejpam-4568	94	11	of	of	ADP
ejpam-4568	94	12	ω	ω	PROPN
ejpam-4568	94	13	,	,	PUNCT
ejpam-4568	94	14	and	and	CCONJ
ejpam-4568	94	15	its	its	PRON
ejpam-4568	94	16	gradient	gradient	NOUN
ejpam-4568	94	17	g(x	g(x	NOUN
ejpam-4568	94	18	)	)	PUNCT
ejpam-4568	94	19	is	be	AUX
ejpam-4568	94	20	lipschitz	lipschitz	ADJ
ejpam-4568	94	21	continuous	continuous	ADJ
ejpam-4568	94	22	,	,	PUNCT
ejpam-4568	94	23	namely	namely	ADV
ejpam-4568	94	24	,	,	PUNCT
ejpam-4568	94	25	there	there	PRON
ejpam-4568	94	26	exists	exist	VERB
ejpam-4568	94	27	a	a	DET
ejpam-4568	94	28	constant	constant	ADJ
ejpam-4568	94	29	l	l	NOUN
ejpam-4568	94	30	>	>	X
ejpam-4568	94	31	0	0	NUM
ejpam-4568	95	1	such	such	ADJ
ejpam-4568	95	2	that	that	SCONJ
ejpam-4568	95	3	:	:	PUNCT
ejpam-4568	95	4	∥g(x)−	∥g(x)−	ADJ
ejpam-4568	95	5	g(y)∥	g(y)∥	NOUN
ejpam-4568	95	6	=	=	SYM
ejpam-4568	95	7	l	l	NOUN
ejpam-4568	95	8	∥x−	∥x−	NOUN
ejpam-4568	95	9	y∥	y∥	NOUN
ejpam-4568	95	10	,	,	PUNCT
ejpam-4568	95	11	∀x	∀x	NUM
ejpam-4568	95	12	,	,	PUNCT
ejpam-4568	95	13	y	y	PROPN
ejpam-4568	95	14	∈	∈	PROPN
ejpam-4568	95	15	u.	u.	PROPN
ejpam-4568	95	16	(	(	PUNCT
ejpam-4568	95	17	21	21	NUM
ejpam-4568	95	18	)	)	PUNCT
ejpam-4568	95	19	to	to	PART
ejpam-4568	95	20	proceed	proceed	VERB
ejpam-4568	95	21	,	,	PUNCT
ejpam-4568	95	22	the	the	DET
ejpam-4568	95	23	well	well	ADV
ejpam-4568	95	24	-	-	PUNCT
ejpam-4568	95	25	known	know	VERB
ejpam-4568	95	26	zoutendijk	zoutendijk	NOUN
ejpam-4568	95	27	condition	condition	NOUN
ejpam-4568	95	28	[	[	X
ejpam-4568	95	29	22	22	NUM
ejpam-4568	95	30	]	]	PUNCT
ejpam-4568	95	31	is	be	AUX
ejpam-4568	95	32	reviewed	review	VERB
ejpam-4568	95	33	in	in	ADP
ejpam-4568	95	34	the	the	DET
ejpam-4568	95	35	following	following	NOUN
ejpam-4568	95	36	.	.	PUNCT
ejpam-4568	96	1	lemma	lemma	PROPN
ejpam-4568	96	2	1	1	NUM
ejpam-4568	96	3	.	.	PUNCT
ejpam-4568	96	4	suppose	suppose	VERB
ejpam-4568	96	5	that	that	SCONJ
ejpam-4568	96	6	assumptions	assumption	NOUN
ejpam-4568	96	7	holds	hold	VERB
ejpam-4568	96	8	true	true	ADJ
ejpam-4568	96	9	.	.	PUNCT
ejpam-4568	97	1	for	for	ADP
ejpam-4568	97	2	any	any	DET
ejpam-4568	97	3	cg	cg	NOUN
ejpam-4568	97	4	iterative	iterative	NOUN
ejpam-4568	97	5	algorithm	algorithm	NOUN
ejpam-4568	97	6	defined	define	VERB
ejpam-4568	97	7	by	by	ADP
ejpam-4568	97	8	(	(	PUNCT
ejpam-4568	97	9	4	4	NUM
ejpam-4568	97	10	)	)	PUNCT
ejpam-4568	97	11	,	,	PUNCT
ejpam-4568	97	12	where	where	SCONJ
ejpam-4568	97	13	dk	dk	PROPN
ejpam-4568	97	14	is	be	AUX
ejpam-4568	97	15	defined	define	VERB
ejpam-4568	97	16	by	by	ADP
ejpam-4568	97	17	(	(	PUNCT
ejpam-4568	97	18	8)	8)	NUM
ejpam-4568	97	19	,	,	PUNCT
ejpam-4568	97	20	and	and	CCONJ
ejpam-4568	97	21	the	the	DET
ejpam-4568	97	22	step	step	NOUN
ejpam-4568	97	23	-	-	PUNCT
ejpam-4568	97	24	size	size	NOUN
ejpam-4568	97	25	αk	αk	NOUN
ejpam-4568	97	26	is	be	AUX
ejpam-4568	97	27	obtained	obtain	VERB
ejpam-4568	97	28	by	by	ADP
ejpam-4568	97	29	the	the	DET
ejpam-4568	97	30	wolfe	wolfe	PROPN
ejpam-4568	97	31	line	line	PROPN
ejpam-4568	97	32	search	search	NOUN
ejpam-4568	97	33	.	.	PUNCT
ejpam-4568	98	1	then	then	ADV
ejpam-4568	98	2	:	:	PUNCT
ejpam-4568	98	3	∞∑	∞∑	NUM
ejpam-4568	98	4	k=1	k=1	X
ejpam-4568	98	5	(	(	PUNCT
ejpam-4568	98	6	gtk	gtk	NOUN
ejpam-4568	98	7	dk	dk	PROPN
ejpam-4568	98	8	)	)	PUNCT
ejpam-4568	98	9	2	2	NUM
ejpam-4568	98	10	∥dk∥2	∥dk∥2	NOUN
ejpam-4568	98	11	<	<	X
ejpam-4568	98	12	∞.	∞.	PROPN
ejpam-4568	98	13	(	(	PUNCT
ejpam-4568	98	14	22	22	NUM
ejpam-4568	98	15	)	)	PUNCT
ejpam-4568	98	16	in	in	ADP
ejpam-4568	98	17	the	the	DET
ejpam-4568	98	18	following	following	NOUN
ejpam-4568	98	19	theorem	theorem	NOUN
ejpam-4568	98	20	,	,	PUNCT
ejpam-4568	98	21	the	the	DET
ejpam-4568	98	22	convergence	convergence	NOUN
ejpam-4568	98	23	property	property	NOUN
ejpam-4568	98	24	of	of	ADP
ejpam-4568	98	25	new	new	ADJ
ejpam-4568	98	26	algorithm	algorithm	NOUN
ejpam-4568	98	27	is	be	AUX
ejpam-4568	98	28	proved	prove	VERB
ejpam-4568	98	29	.	.	PUNCT
ejpam-4568	99	1	theorem	theorem	NOUN
ejpam-4568	99	2	2	2	NUM
ejpam-4568	99	3	.	.	PUNCT
ejpam-4568	99	4	suppose	suppose	VERB
ejpam-4568	99	5	the	the	DET
ejpam-4568	99	6	all	all	DET
ejpam-4568	99	7	assumptions	assumption	NOUN
ejpam-4568	99	8	holds	hold	VERB
ejpam-4568	99	9	true	true	ADJ
ejpam-4568	99	10	.	.	PUNCT
ejpam-4568	100	1	consider	consider	VERB
ejpam-4568	100	2	the	the	DET
ejpam-4568	100	3	sequence{gk	sequence{gk	VERB
ejpam-4568	100	4	}	}	PUNCT
ejpam-4568	100	5	and	and	CCONJ
ejpam-4568	100	6	{	{	PUNCT
ejpam-4568	100	7	dk	dk	NOUN
ejpam-4568	100	8	}	}	PUNCT
ejpam-4568	100	9	generated	generate	VERB
ejpam-4568	100	10	by	by	ADP
ejpam-4568	100	11	the	the	DET
ejpam-4568	100	12	proposed	propose	VERB
ejpam-4568	100	13	method	method	NOUN
ejpam-4568	100	14	,	,	PUNCT
ejpam-4568	100	15	where	where	SCONJ
ejpam-4568	100	16	βk	βk	NOUN
ejpam-4568	100	17	is	be	AUX
ejpam-4568	100	18	given	give	VERB
ejpam-4568	100	19	by	by	ADP
ejpam-4568	100	20	(	(	PUNCT
ejpam-4568	100	21	15	15	NUM
ejpam-4568	100	22	)	)	PUNCT
ejpam-4568	100	23	,	,	PUNCT
ejpam-4568	100	24	and	and	CCONJ
ejpam-4568	100	25	αk	αk	SCONJ
ejpam-4568	100	26	satisfies	satisfy	VERB
ejpam-4568	100	27	the	the	DET
ejpam-4568	100	28	wolfe	wolfe	PROPN
ejpam-4568	100	29	line	line	PROPN
ejpam-4568	100	30	search	search	NOUN
ejpam-4568	100	31	,	,	PUNCT
ejpam-4568	100	32	then	then	ADV
ejpam-4568	100	33	,	,	PUNCT
ejpam-4568	100	34	lim	lim	PROPN
ejpam-4568	100	35	k−→∞	k−→∞	PROPN
ejpam-4568	100	36	inf	inf	PROPN
ejpam-4568	100	37	∥gk∥	∥gk∥	NOUN
ejpam-4568	100	38	=	=	SYM
ejpam-4568	100	39	0	0	NUM
ejpam-4568	100	40	(	(	PUNCT
ejpam-4568	100	41	23	23	NUM
ejpam-4568	100	42	)	)	PUNCT
ejpam-4568	100	43	proof	proof	NOUN
ejpam-4568	100	44	.	.	PUNCT
ejpam-4568	101	1	by	by	ADP
ejpam-4568	101	2	contradiction	contradiction	NOUN
ejpam-4568	101	3	,	,	PUNCT
ejpam-4568	101	4	suppose	suppose	VERB
ejpam-4568	101	5	that	that	SCONJ
ejpam-4568	101	6	(	(	PUNCT
ejpam-4568	101	7	23	23	NUM
ejpam-4568	101	8	)	)	PUNCT
ejpam-4568	101	9	is	be	AUX
ejpam-4568	101	10	not	not	PART
ejpam-4568	101	11	correct	correct	ADJ
ejpam-4568	101	12	.	.	PUNCT
ejpam-4568	102	1	therefore	therefore	ADV
ejpam-4568	102	2	,	,	PUNCT
ejpam-4568	102	3	there	there	PRON
ejpam-4568	102	4	exists	exist	VERB
ejpam-4568	102	5	a	a	DET
ejpam-4568	102	6	constant	constant	ADJ
ejpam-4568	102	7	ε	ε	PROPN
ejpam-4568	102	8	>	>	X
ejpam-4568	102	9	0	0	NUM
ejpam-4568	102	10	such	such	ADJ
ejpam-4568	102	11	that	that	SCONJ
ejpam-4568	102	12	:	:	PUNCT
ejpam-4568	102	13	∥gk+1∥	∥gk+1∥	NUM
ejpam-4568	102	14	>	>	X
ejpam-4568	102	15	ε	ε	PROPN
ejpam-4568	102	16	(	(	PUNCT
ejpam-4568	102	17	24	24	NUM
ejpam-4568	102	18	)	)	PUNCT
ejpam-4568	102	19	upon	upon	SCONJ
ejpam-4568	102	20	squaring	square	VERB
ejpam-4568	102	21	both	both	DET
ejpam-4568	102	22	sides	side	NOUN
ejpam-4568	102	23	of	of	ADP
ejpam-4568	102	24	(	(	PUNCT
ejpam-4568	102	25	8)	8)	NUM
ejpam-4568	102	26	,	,	PUNCT
ejpam-4568	102	27	we	we	PRON
ejpam-4568	102	28	obtain	obtain	VERB
ejpam-4568	102	29	:	:	PUNCT
ejpam-4568	102	30	b.	b.	PROPN
ejpam-4568	102	31	a.	a.	PROPN
ejpam-4568	102	32	hassan	hassan	PROPN
ejpam-4568	102	33	,	,	PUNCT
ejpam-4568	102	34	h.	h.	PROPN
ejpam-4568	102	35	m.	m.	PROPN
ejpam-4568	102	36	sadiq	sadiq	PROPN
ejpam-4568	102	37	/	/	SYM
ejpam-4568	102	38	eur	eur	PROPN
ejpam-4568	102	39	.	.	PUNCT
ejpam-4568	103	1	j.	j.	PROPN
ejpam-4568	103	2	pure	pure	PROPN
ejpam-4568	103	3	appl	appl	PROPN
ejpam-4568	103	4	.	.	PROPN
ejpam-4568	103	5	math	math	PROPN
ejpam-4568	103	6	,	,	PUNCT
ejpam-4568	103	7	15	15	NUM
ejpam-4568	103	8	(	(	PUNCT
ejpam-4568	103	9	4	4	NUM
ejpam-4568	103	10	)	)	PUNCT
ejpam-4568	103	11	(	(	PUNCT
ejpam-4568	103	12	2022	2022	NUM
ejpam-4568	103	13	)	)	PUNCT
ejpam-4568	103	14	,	,	PUNCT
ejpam-4568	103	15	2011	2011	NUM
ejpam-4568	103	16	-	-	SYM
ejpam-4568	103	17	2021	2021	NUM
ejpam-4568	103	18	2016	2016	NUM
ejpam-4568	103	19	∥dk+1∥2	∥dk+1∥2	NOUN
ejpam-4568	103	20	+	+	CCONJ
ejpam-4568	103	21	∥gk+1∥2	∥gk+1∥2	ADJ
ejpam-4568	103	22	+	+	CCONJ
ejpam-4568	103	23	2dtk+1gk+1	2dtk+1gk+1	NUM
ejpam-4568	103	24	=	=	SYM
ejpam-4568	103	25	(	(	PUNCT
ejpam-4568	103	26	βk	βk	NOUN
ejpam-4568	103	27	)	)	PUNCT
ejpam-4568	103	28	2	2	NUM
ejpam-4568	103	29	∥dk∥2	∥dk∥2	PROPN
ejpam-4568	103	30	(	(	PUNCT
ejpam-4568	103	31	25	25	NUM
ejpam-4568	103	32	)	)	PUNCT
ejpam-4568	103	33	next	next	ADV
ejpam-4568	103	34	,	,	PUNCT
ejpam-4568	103	35	dividing	divide	VERB
ejpam-4568	103	36	both	both	DET
ejpam-4568	103	37	sides	side	NOUN
ejpam-4568	103	38	of	of	ADP
ejpam-4568	103	39	the	the	DET
ejpam-4568	103	40	above	above	ADJ
ejpam-4568	103	41	inequality	inequality	NOUN
ejpam-4568	103	42	by	by	ADP
ejpam-4568	103	43	(	(	PUNCT
ejpam-4568	103	44	gtk+1dk+1	gtk+1dk+1	NOUN
ejpam-4568	103	45	)	)	PUNCT
ejpam-4568	103	46	2	2	NUM
ejpam-4568	103	47	,	,	PUNCT
ejpam-4568	103	48	we	we	PRON
ejpam-4568	103	49	have	have	AUX
ejpam-4568	103	50	:	:	PUNCT
ejpam-4568	103	51	∥dk+1∥2	∥dk+1∥2	ADJ
ejpam-4568	103	52	(	(	PUNCT
ejpam-4568	103	53	dtk+1gk+1)2	dtk+1gk+1)2	VERB
ejpam-4568	103	54	=	=	SYM
ejpam-4568	103	55	(	(	PUNCT
ejpam-4568	103	56	βk	βk	NOUN
ejpam-4568	103	57	)	)	PUNCT
ejpam-4568	103	58	2∥dk∥2	2∥dk∥2	NUM
ejpam-4568	103	59	(	(	PUNCT
ejpam-4568	103	60	dtk+1gk+1)2	dtk+1gk+1)2	VERB
ejpam-4568	103	61	−	−	NUM
ejpam-4568	103	62	2	2	NUM
ejpam-4568	103	63	(	(	PUNCT
ejpam-4568	103	64	dtk+1gk+1)2	dtk+1gk+1)2	VERB
ejpam-4568	103	65	−	−	PROPN
ejpam-4568	103	66	∥gk+1∥2	∥gk+1∥2	NOUN
ejpam-4568	103	67	(	(	PUNCT
ejpam-4568	103	68	dtk+1gk+1)2	dtk+1gk+1)2	VERB
ejpam-4568	103	69	=	=	SYM
ejpam-4568	103	70	(	(	PUNCT
ejpam-4568	103	71	βk	βk	NOUN
ejpam-4568	103	72	)	)	PUNCT
ejpam-4568	103	73	2∥dk∥2	2∥dk∥2	NUM
ejpam-4568	103	74	(	(	PUNCT
ejpam-4568	103	75	dtk+1gk+1)2	dtk+1gk+1)2	NOUN
ejpam-4568	103	76	−	−	PROPN
ejpam-4568	103	77	(	(	PUNCT
ejpam-4568	103	78	1	1	NUM
ejpam-4568	103	79	∥gk+1∥2	∥gk+1∥2	NOUN
ejpam-4568	103	80	+	+	CCONJ
ejpam-4568	103	81	∥gk+1∥	∥gk+1∥	NUM
ejpam-4568	103	82	(	(	PUNCT
ejpam-4568	103	83	dtk+1gk+1	dtk+1gk+1	NOUN
ejpam-4568	103	84	)	)	PUNCT
ejpam-4568	103	85	)	)	PUNCT
ejpam-4568	103	86	2	2	NUM
ejpam-4568	104	1	+	+	SYM
ejpam-4568	104	2	1	1	NUM
ejpam-4568	104	3	∥gk+1∥2	∥gk+1∥2	NOUN
ejpam-4568	104	4	(	(	PUNCT
ejpam-4568	104	5	26	26	NUM
ejpam-4568	104	6	)	)	PUNCT
ejpam-4568	104	7	which	which	PRON
ejpam-4568	104	8	reduces	reduce	VERB
ejpam-4568	104	9	to	to	PART
ejpam-4568	104	10	:	:	PUNCT
ejpam-4568	104	11	∥dk+1∥2	∥dk+1∥2	X
ejpam-4568	104	12	(	(	PUNCT
ejpam-4568	104	13	dtk+1gk+1)2	dtk+1gk+1)2	VERB
ejpam-4568	104	14	=	=	SYM
ejpam-4568	104	15	(	(	PUNCT
ejpam-4568	104	16	βk	βk	NOUN
ejpam-4568	104	17	)	)	PUNCT
ejpam-4568	104	18	2∥dk∥2	2∥dk∥2	NUM
ejpam-4568	104	19	(	(	PUNCT
ejpam-4568	104	20	dtk+1gk+1)2	dtk+1gk+1)2	ADJ
ejpam-4568	104	21	+	+	NUM
ejpam-4568	104	22	1	1	NUM
ejpam-4568	104	23	∥gk+1∥2	∥gk+1∥2	NOUN
ejpam-4568	104	24	(	(	PUNCT
ejpam-4568	104	25	27	27	NUM
ejpam-4568	104	26	)	)	PUNCT
ejpam-4568	104	27	however	however	ADV
ejpam-4568	104	28	,	,	PUNCT
ejpam-4568	104	29	from	from	ADP
ejpam-4568	104	30	(	(	PUNCT
ejpam-4568	104	31	20	20	NUM
ejpam-4568	104	32	)	)	PUNCT
ejpam-4568	104	33	we	we	PRON
ejpam-4568	104	34	have	have	AUX
ejpam-4568	104	35	βk	βk	VERB
ejpam-4568	104	36	=	=	PUNCT
ejpam-4568	104	37	dtk+1gk+1	dtk+1gk+1	PROPN
ejpam-4568	104	38	dtk	dtk	PROPN
ejpam-4568	104	39	gk	gk	PROPN
ejpam-4568	104	40	,	,	PUNCT
ejpam-4568	104	41	substituting	substitute	VERB
ejpam-4568	104	42	in	in	ADP
ejpam-4568	104	43	(	(	PUNCT
ejpam-4568	104	44	27	27	NUM
ejpam-4568	104	45	)	)	PUNCT
ejpam-4568	104	46	,	,	PUNCT
ejpam-4568	104	47	we	we	PRON
ejpam-4568	104	48	obtain	obtain	VERB
ejpam-4568	104	49	:	:	PUNCT
ejpam-4568	104	50	∥dk+1∥2	∥dk+1∥2	X
ejpam-4568	104	51	(	(	PUNCT
ejpam-4568	104	52	dtk+1gk+1)2	dtk+1gk+1)2	VERB
ejpam-4568	104	53	=	=	SYM
ejpam-4568	104	54	(	(	PUNCT
ejpam-4568	104	55	dtk+1gk+1	dtk+1gk+1	NOUN
ejpam-4568	104	56	)	)	PUNCT
ejpam-4568	104	57	2	2	NUM
ejpam-4568	104	58	(	(	PUNCT
ejpam-4568	104	59	dtk	dtk	PROPN
ejpam-4568	104	60	gk	gk	PROPN
ejpam-4568	104	61	)	)	PUNCT
ejpam-4568	104	62	2	2	NUM
ejpam-4568	104	63	∥dk∥2	∥dk∥2	PROPN
ejpam-4568	105	1	(	(	PUNCT
ejpam-4568	105	2	dtk+1gk+1)2	dtk+1gk+1)2	VERB
ejpam-4568	105	3	+	+	NUM
ejpam-4568	105	4	1	1	NUM
ejpam-4568	105	5	∥gk+1∥2	∥gk+1∥2	NOUN
ejpam-4568	105	6	=	=	SYM
ejpam-4568	105	7	∥dk∥2	∥dk∥2	PROPN
ejpam-4568	106	1	(	(	PUNCT
ejpam-4568	106	2	dtk	dtk	PROPN
ejpam-4568	106	3	gk	gk	PROPN
ejpam-4568	106	4	)	)	PUNCT
ejpam-4568	106	5	2	2	NUM
ejpam-4568	107	1	+	+	SYM
ejpam-4568	107	2	1	1	NUM
ejpam-4568	107	3	∥gk+1∥2	∥gk+1∥2	NOUN
ejpam-4568	107	4	(	(	PUNCT
ejpam-4568	107	5	28	28	NUM
ejpam-4568	107	6	)	)	PUNCT
ejpam-4568	107	7	notice	notice	VERB
ejpam-4568	107	8	that	that	SCONJ
ejpam-4568	107	9	∥d1∥2	∥d1∥2	ADV
ejpam-4568	107	10	=	=	SYM
ejpam-4568	107	11	−gt1	−gt1	X
ejpam-4568	107	12	d1	d1	NOUN
ejpam-4568	107	13	=	=	SYM
ejpam-4568	107	14	∥g1∥2	∥g1∥2	PROPN
ejpam-4568	107	15	,	,	PUNCT
ejpam-4568	107	16	which	which	PRON
ejpam-4568	107	17	implies	imply	VERB
ejpam-4568	107	18	:	:	PUNCT
ejpam-4568	107	19	∥dk+1∥2	∥dk+1∥2	ADJ
ejpam-4568	107	20	(	(	PUNCT
ejpam-4568	107	21	dtk+1gk+1)2	dtk+1gk+1)2	VERB
ejpam-4568	107	22	=	=	SYM
ejpam-4568	107	23	k+1∑	k+1∑	PROPN
ejpam-4568	107	24	i=1	i=1	PROPN
ejpam-4568	107	25	1	1	NUM
ejpam-4568	107	26	∥gi∥2	∥gi∥2	NOUN
ejpam-4568	107	27	.	.	PUNCT
ejpam-4568	108	1	(	(	PUNCT
ejpam-4568	108	2	29	29	NUM
ejpam-4568	108	3	)	)	PUNCT
ejpam-4568	108	4	then	then	ADV
ejpam-4568	108	5	we	we	PRON
ejpam-4568	108	6	get	get	VERB
ejpam-4568	108	7	from	from	ADP
ejpam-4568	108	8	(	(	PUNCT
ejpam-4568	108	9	29	29	NUM
ejpam-4568	108	10	)	)	PUNCT
ejpam-4568	108	11	and	and	CCONJ
ejpam-4568	108	12	(	(	PUNCT
ejpam-4568	108	13	24	24	NUM
ejpam-4568	108	14	)	)	PUNCT
ejpam-4568	109	1	that	that	PRON
ejpam-4568	109	2	:	:	PUNCT
ejpam-4568	109	3	(	(	PUNCT
ejpam-4568	109	4	dtk	dtk	PROPN
ejpam-4568	109	5	gk	gk	PROPN
ejpam-4568	109	6	)	)	PUNCT
ejpam-4568	109	7	2	2	NUM
ejpam-4568	109	8	∥dk∥2	∥dk∥2	NOUN
ejpam-4568	109	9	=	=	PUNCT
ejpam-4568	109	10	∈2	∈2	PROPN
ejpam-4568	109	11	k	k	X
ejpam-4568	109	12	.	.	PUNCT
ejpam-4568	110	1	(	(	PUNCT
ejpam-4568	110	2	30	30	NUM
ejpam-4568	110	3	)	)	PUNCT
ejpam-4568	110	4	therefore	therefore	ADV
ejpam-4568	110	5	:	:	PUNCT
ejpam-4568	111	1	∞∑	∞∑	NUM
ejpam-4568	111	2	k=1	k=1	X
ejpam-4568	111	3	(	(	PUNCT
ejpam-4568	111	4	gtk	gtk	NOUN
ejpam-4568	111	5	dk	dk	PROPN
ejpam-4568	111	6	)	)	PUNCT
ejpam-4568	111	7	2	2	NUM
ejpam-4568	111	8	∥dk∥2	∥dk∥2	NOUN
ejpam-4568	111	9	=	=	PRON
ejpam-4568	111	10	∞.	∞.	PROPN
ejpam-4568	111	11	(	(	PUNCT
ejpam-4568	111	12	31	31	NUM
ejpam-4568	111	13	)	)	PUNCT
ejpam-4568	111	14	this	this	DET
ejpam-4568	111	15	result	result	NOUN
ejpam-4568	111	16	contradicts	contradict	VERB
ejpam-4568	111	17	(	(	PUNCT
ejpam-4568	111	18	22	22	NUM
ejpam-4568	111	19	)	)	PUNCT
ejpam-4568	111	20	,	,	PUNCT
ejpam-4568	111	21	the	the	DET
ejpam-4568	111	22	proof	proof	NOUN
ejpam-4568	111	23	is	be	AUX
ejpam-4568	111	24	completed	complete	VERB
ejpam-4568	111	25	.	.	PUNCT
ejpam-4568	112	1	global	global	ADJ
ejpam-4568	112	2	convergence	convergence	NOUN
ejpam-4568	112	3	property	property	NOUN
ejpam-4568	112	4	of	of	ADP
ejpam-4568	112	5	bv2	bv2	PROPN
ejpam-4568	112	6	,	,	PUNCT
ejpam-4568	112	7	bv3	bv3	ADJ
ejpam-4568	112	8	algorithm	algorithm	NOUN
ejpam-4568	112	9	are	be	AUX
ejpam-4568	112	10	similar	similar	ADJ
ejpam-4568	112	11	those	those	PRON
ejpam-4568	112	12	of	of	ADP
ejpam-4568	112	13	bv1	bv1	PROPN
ejpam-4568	112	14	algorithm	algorithm	NOUN
ejpam-4568	112	15	.	.	PUNCT
ejpam-4568	113	1	4	4	X
ejpam-4568	113	2	.	.	NOUN
ejpam-4568	113	3	numerical	numerical	ADJ
ejpam-4568	113	4	experiments	experiment	NOUN
ejpam-4568	113	5	to	to	PART
ejpam-4568	113	6	illustrate	illustrate	VERB
ejpam-4568	113	7	the	the	DET
ejpam-4568	113	8	effectiveness	effectiveness	NOUN
ejpam-4568	113	9	of	of	ADP
ejpam-4568	113	10	the	the	DET
ejpam-4568	113	11	suggested	suggest	VERB
ejpam-4568	113	12	approach	approach	NOUN
ejpam-4568	113	13	for	for	ADP
ejpam-4568	113	14	salt	salt	NOUN
ejpam-4568	113	15	-	-	PUNCT
ejpam-4568	113	16	and	and	CCONJ
ejpam-4568	113	17	-	-	PUNCT
ejpam-4568	113	18	pepper	pepper	NOUN
ejpam-4568	113	19	impulse	impulse	ADJ
ejpam-4568	113	20	noise	noise	NOUN
ejpam-4568	113	21	,	,	PUNCT
ejpam-4568	113	22	we	we	PRON
ejpam-4568	113	23	present	present	VERB
ejpam-4568	113	24	some	some	DET
ejpam-4568	113	25	numerical	numerical	ADJ
ejpam-4568	113	26	results	result	NOUN
ejpam-4568	113	27	in	in	ADP
ejpam-4568	113	28	this	this	DET
ejpam-4568	113	29	section	section	NOUN
ejpam-4568	113	30	.	.	PUNCT
ejpam-4568	114	1	table	table	NOUN
ejpam-4568	114	2	1	1	NUM
ejpam-4568	114	3	lists	list	VERB
ejpam-4568	114	4	the	the	DET
ejpam-4568	114	5	experimental	experimental	ADJ
ejpam-4568	114	6	outcomes	outcome	NOUN
ejpam-4568	114	7	.	.	PUNCT
ejpam-4568	115	1	we	we	PRON
ejpam-4568	115	2	report	report	VERB
ejpam-4568	115	3	the	the	DET
ejpam-4568	115	4	number	number	NOUN
ejpam-4568	115	5	of	of	ADP
ejpam-4568	115	6	iterations	iteration	NOUN
ejpam-4568	115	7	(	(	PUNCT
ejpam-4568	115	8	ni	ni	NOUN
ejpam-4568	115	9	)	)	PUNCT
ejpam-4568	115	10	,	,	PUNCT
ejpam-4568	115	11	the	the	DET
ejpam-4568	115	12	number	number	NOUN
ejpam-4568	115	13	of	of	ADP
ejpam-4568	115	14	function	function	NOUN
ejpam-4568	115	15	evaluations	evaluation	NOUN
ejpam-4568	115	16	b.	b.	PROPN
ejpam-4568	115	17	a.	a.	PROPN
ejpam-4568	115	18	hassan	hassan	PROPN
ejpam-4568	115	19	,	,	PUNCT
ejpam-4568	115	20	h.	h.	PROPN
ejpam-4568	115	21	m.	m.	PROPN
ejpam-4568	115	22	sadiq	sadiq	PROPN
ejpam-4568	115	23	/	/	SYM
ejpam-4568	115	24	eur	eur	PROPN
ejpam-4568	115	25	.	.	PUNCT
ejpam-4568	116	1	j.	j.	PROPN
ejpam-4568	116	2	pure	pure	PROPN
ejpam-4568	116	3	appl	appl	PROPN
ejpam-4568	116	4	.	.	PROPN
ejpam-4568	116	5	math	math	PROPN
ejpam-4568	116	6	,	,	PUNCT
ejpam-4568	116	7	15	15	NUM
ejpam-4568	116	8	(	(	PUNCT
ejpam-4568	116	9	4	4	NUM
ejpam-4568	116	10	)	)	PUNCT
ejpam-4568	116	11	(	(	PUNCT
ejpam-4568	116	12	2022	2022	NUM
ejpam-4568	116	13	)	)	PUNCT
ejpam-4568	116	14	,	,	PUNCT
ejpam-4568	116	15	2011	2011	NUM
ejpam-4568	116	16	-	-	SYM
ejpam-4568	116	17	2021	2021	NUM
ejpam-4568	116	18	2017	2017	NUM
ejpam-4568	116	19	(	(	PUNCT
ejpam-4568	116	20	nof	nof	PROPN
ejpam-4568	116	21	)	)	PUNCT
ejpam-4568	116	22	and	and	CCONJ
ejpam-4568	116	23	the	the	DET
ejpam-4568	116	24	evaluation	evaluation	NOUN
ejpam-4568	116	25	indexes	index	NOUN
ejpam-4568	116	26	used	use	VERB
ejpam-4568	116	27	in	in	ADP
ejpam-4568	116	28	the	the	DET
ejpam-4568	116	29	experiments	experiment	NOUN
ejpam-4568	116	30	were	be	AUX
ejpam-4568	116	31	the	the	DET
ejpam-4568	116	32	psnr	psnr	NOUN
ejpam-4568	116	33	(	(	PUNCT
ejpam-4568	116	34	peak	peak	NOUN
ejpam-4568	116	35	signal	signal	NOUN
ejpam-4568	116	36	to	to	PART
ejpam-4568	116	37	noise	noise	VERB
ejpam-4568	116	38	ratio	ratio	NOUN
ejpam-4568	116	39	)	)	PUNCT
ejpam-4568	116	40	see	see	VERB
ejpam-4568	116	41	[	[	X
ejpam-4568	116	42	3	3	NUM
ejpam-4568	116	43	]	]	PUNCT
ejpam-4568	116	44	,	,	PUNCT
ejpam-4568	116	45	which	which	PRON
ejpam-4568	116	46	is	be	AUX
ejpam-4568	116	47	defined	define	VERB
ejpam-4568	116	48	as	as	ADP
ejpam-4568	116	49	:	:	PUNCT
ejpam-4568	116	50	psnr	psnr	NOUN
ejpam-4568	116	51	=	=	SYM
ejpam-4568	116	52	10log10	10log10	NUM
ejpam-4568	116	53	2552	2552	NUM
ejpam-4568	116	54	1	1	NUM
ejpam-4568	116	55	mn	mn	PROPN
ejpam-4568	116	56	∑	∑	PROPN
ejpam-4568	116	57	i	i	PROPN
ejpam-4568	116	58	,	,	PUNCT
ejpam-4568	116	59	j	j	PROPN
ejpam-4568	116	60	(	(	PUNCT
ejpam-4568	116	61	u	u	NOUN
ejpam-4568	116	62	r	r	VERB
ejpam-4568	116	63	i	i	PROPN
ejpam-4568	116	64	,	,	PUNCT
ejpam-4568	116	65	j	j	PROPN
ejpam-4568	116	66	−	−	PROPN
ejpam-4568	116	67	u∗i	u∗i	NOUN
ejpam-4568	116	68	,	,	PUNCT
ejpam-4568	116	69	j	j	NOUN
ejpam-4568	116	70	)	)	PUNCT
ejpam-4568	116	71	2	2	NUM
ejpam-4568	116	72	(	(	PUNCT
ejpam-4568	116	73	32	32	NUM
ejpam-4568	116	74	)	)	PUNCT
ejpam-4568	116	75	where	where	SCONJ
ejpam-4568	116	76	uri	uri	PROPN
ejpam-4568	116	77	,	,	PUNCT
ejpam-4568	116	78	j	j	PROPN
ejpam-4568	116	79	and	and	CCONJ
ejpam-4568	116	80	u∗i	u∗i	NOUN
ejpam-4568	116	81	,	,	PUNCT
ejpam-4568	116	82	j	j	PROPN
ejpam-4568	116	83	denote	denote	VERB
ejpam-4568	116	84	the	the	DET
ejpam-4568	116	85	pixel	pixel	PROPN
ejpam-4568	116	86	values	value	NOUN
ejpam-4568	116	87	of	of	ADP
ejpam-4568	116	88	the	the	DET
ejpam-4568	116	89	restored	restore	VERB
ejpam-4568	116	90	image	image	NOUN
ejpam-4568	116	91	and	and	CCONJ
ejpam-4568	116	92	the	the	DET
ejpam-4568	116	93	original	original	ADJ
ejpam-4568	116	94	image	image	NOUN
ejpam-4568	116	95	.	.	PUNCT
ejpam-4568	117	1	we	we	PRON
ejpam-4568	117	2	stop	stop	VERB
ejpam-4568	117	3	the	the	DET
ejpam-4568	117	4	iteration	iteration	NOUN
ejpam-4568	117	5	if	if	SCONJ
ejpam-4568	117	6	the	the	DET
ejpam-4568	117	7	inequality	inequality	NOUN
ejpam-4568	117	8	:	:	PUNCT
ejpam-4568	117	9	|f(uk)−	|f(uk)−	ADJ
ejpam-4568	117	10	f(uk−1)|	f(uk−1)|	PROPN
ejpam-4568	117	11	|f(uk)|	|f(uk)|	PUNCT
ejpam-4568	117	12	=	=	SYM
ejpam-4568	117	13	10−4	10−4	NUM
ejpam-4568	117	14	and	and	CCONJ
ejpam-4568	117	15	∥f(uk)∥	∥f(uk)∥	PROPN
ejpam-4568	117	16	=	=	SYM
ejpam-4568	117	17	10−4(1	10−4(1	NUM
ejpam-4568	117	18	+	+	NUM
ejpam-4568	117	19	|f(uk)|	|f(uk)|	NOUN
ejpam-4568	117	20	)	)	PUNCT
ejpam-4568	117	21	(	(	PUNCT
ejpam-4568	117	22	33	33	NUM
ejpam-4568	117	23	)	)	PUNCT
ejpam-4568	117	24	are	be	AUX
ejpam-4568	117	25	satisfied	satisfied	ADJ
ejpam-4568	117	26	.	.	PUNCT
ejpam-4568	118	1	in	in	ADP
ejpam-4568	118	2	the	the	DET
ejpam-4568	118	3	experiment	experiment	NOUN
ejpam-4568	118	4	,	,	PUNCT
ejpam-4568	118	5	a	a	DET
ejpam-4568	118	6	picture	picture	NOUN
ejpam-4568	118	7	that	that	PRON
ejpam-4568	118	8	has	have	AUX
ejpam-4568	118	9	been	be	AUX
ejpam-4568	118	10	lost	lose	VERB
ejpam-4568	118	11	or	or	CCONJ
ejpam-4568	118	12	become	become	VERB
ejpam-4568	118	13	hazy	hazy	ADJ
ejpam-4568	118	14	is	be	AUX
ejpam-4568	118	15	recovered	recover	VERB
ejpam-4568	118	16	or	or	CCONJ
ejpam-4568	118	17	recreated	recreate	VERB
ejpam-4568	118	18	.	.	PUNCT
ejpam-4568	119	1	four	four	NUM
ejpam-4568	119	2	different	different	ADJ
ejpam-4568	119	3	images	image	NOUN
ejpam-4568	119	4	are	be	AUX
ejpam-4568	119	5	employed	employ	VERB
ejpam-4568	119	6	for	for	ADP
ejpam-4568	119	7	the	the	DET
ejpam-4568	119	8	experiment	experiment	NOUN
ejpam-4568	119	9	,	,	PUNCT
ejpam-4568	119	10	which	which	PRON
ejpam-4568	119	11	includes	include	VERB
ejpam-4568	119	12	,	,	PUNCT
ejpam-4568	119	13	placelena	placelena	NOUN
ejpam-4568	119	14	,	,	PUNCT
ejpam-4568	119	15	house	house	NOUN
ejpam-4568	119	16	,	,	PUNCT
ejpam-4568	119	17	cameraman	cameraman	NOUN
ejpam-4568	119	18	and	and	CCONJ
ejpam-4568	119	19	elaine	elaine	NOUN
ejpam-4568	119	20	by	by	ADP
ejpam-4568	119	21	employing	employ	VERB
ejpam-4568	119	22	test	test	NOUN
ejpam-4568	119	23	images	image	NOUN
ejpam-4568	119	24	in	in	ADP
ejpam-4568	119	25	[	[	X
ejpam-4568	119	26	2],[21	2],[21	NOUN
ejpam-4568	119	27	]	]	PUNCT
ejpam-4568	119	28	results	result	NOUN
ejpam-4568	119	29	of	of	ADP
ejpam-4568	119	30	experiment	experiment	NOUN
ejpam-4568	119	31	to	to	ADP
ejpam-4568	119	32	images	image	NOUN
ejpam-4568	119	33	shown	show	VERB
ejpam-4568	119	34	in	in	ADP
ejpam-4568	119	35	table	table	NOUN
ejpam-4568	119	36	(	(	PUNCT
ejpam-4568	119	37	1	1	NUM
ejpam-4568	119	38	)	)	PUNCT
ejpam-4568	119	39	.	.	PUNCT
ejpam-4568	120	1	images	image	NOUN
ejpam-4568	120	2	show	show	VERB
ejpam-4568	120	3	that	that	SCONJ
ejpam-4568	120	4	new	new	ADJ
ejpam-4568	120	5	algorithm	algorithm	NOUN
ejpam-4568	120	6	and	and	CCONJ
ejpam-4568	120	7	fr	fr	PROPN
ejpam-4568	120	8	algorithm	algorithm	NOUN
ejpam-4568	120	9	have	have	VERB
ejpam-4568	120	10	good	good	ADJ
ejpam-4568	120	11	performance	performance	NOUN
ejpam-4568	120	12	to	to	PART
ejpam-4568	120	13	solve	solve	VERB
ejpam-4568	120	14	the	the	DET
ejpam-4568	120	15	image	image	NOUN
ejpam-4568	120	16	restoration	restoration	NOUN
ejpam-4568	120	17	and	and	CCONJ
ejpam-4568	120	18	it	it	PRON
ejpam-4568	120	19	can	can	AUX
ejpam-4568	120	20	successfully	successfully	ADV
ejpam-4568	120	21	do	do	VERB
ejpam-4568	120	22	this	this	DET
ejpam-4568	120	23	problem	problem	NOUN
ejpam-4568	120	24	,	,	PUNCT
ejpam-4568	120	25	for	for	SCONJ
ejpam-4568	120	26	more	more	ADJ
ejpam-4568	120	27	details	detail	NOUN
ejpam-4568	120	28	in	in	ADP
ejpam-4568	120	29	this	this	DET
ejpam-4568	120	30	field	field	NOUN
ejpam-4568	120	31	see	see	VERB
ejpam-4568	120	32	[	[	X
ejpam-4568	120	33	12–15	12–15	NUM
ejpam-4568	120	34	]	]	PUNCT
ejpam-4568	120	35	.	.	PUNCT
ejpam-4568	121	1	the	the	DET
ejpam-4568	121	2	numerical	numerical	ADJ
ejpam-4568	121	3	results	result	NOUN
ejpam-4568	121	4	show	show	VERB
ejpam-4568	121	5	that	that	SCONJ
ejpam-4568	121	6	for	for	ADP
ejpam-4568	121	7	some	some	DET
ejpam-4568	121	8	situations	situation	NOUN
ejpam-4568	121	9	,	,	PUNCT
ejpam-4568	121	10	the	the	DET
ejpam-4568	121	11	suggested	suggest	VERB
ejpam-4568	121	12	solution	solution	NOUN
ejpam-4568	121	13	outperforms	outperform	VERB
ejpam-4568	121	14	the	the	DET
ejpam-4568	121	15	fr	fr	ADJ
ejpam-4568	121	16	method	method	NOUN
ejpam-4568	121	17	.	.	PUNCT
ejpam-4568	122	1	table	table	NOUN
ejpam-4568	122	2	1	1	NUM
ejpam-4568	122	3	:	:	PUNCT
ejpam-4568	122	4	numerical	numerical	ADJ
ejpam-4568	122	5	results	result	NOUN
ejpam-4568	122	6	of	of	ADP
ejpam-4568	122	7	fr	fr	PROPN
ejpam-4568	122	8	,	,	PUNCT
ejpam-4568	122	9	bv1	bv1	PROPN
ejpam-4568	122	10	,	,	PUNCT
ejpam-4568	122	11	bv2	bv2	NOUN
ejpam-4568	122	12	and	and	CCONJ
ejpam-4568	122	13	bv3	bv3	ADJ
ejpam-4568	122	14	algorithms	algorithm	NOUN
ejpam-4568	122	15	.	.	PUNCT
ejpam-4568	123	1	in	in	ADP
ejpam-4568	123	2	terms	term	NOUN
ejpam-4568	123	3	of	of	ADP
ejpam-4568	123	4	the	the	DET
ejpam-4568	123	5	number	number	NOUN
ejpam-4568	123	6	of	of	ADP
ejpam-4568	123	7	iterations	iteration	NOUN
ejpam-4568	123	8	and	and	CCONJ
ejpam-4568	123	9	function	function	NOUN
ejpam-4568	123	10	evaluations	evaluation	NOUN
ejpam-4568	123	11	,	,	PUNCT
ejpam-4568	123	12	as	as	ADV
ejpam-4568	123	13	well	well	ADV
ejpam-4568	123	14	as	as	ADP
ejpam-4568	123	15	the	the	DET
ejpam-4568	123	16	peak	peak	NOUN
ejpam-4568	123	17	signal	signal	NOUN
ejpam-4568	123	18	to	to	PART
ejpam-4568	123	19	noise	noise	NOUN
ejpam-4568	123	20	ratio	ratio	NOUN
ejpam-4568	123	21	,	,	PUNCT
ejpam-4568	123	22	the	the	DET
ejpam-4568	123	23	recommended	recommend	VERB
ejpam-4568	123	24	algorithms	algorithm	NOUN
ejpam-4568	123	25	surpass	surpass	VERB
ejpam-4568	123	26	the	the	DET
ejpam-4568	123	27	fr	fr	ADJ
ejpam-4568	123	28	technique	technique	NOUN
ejpam-4568	123	29	,	,	PUNCT
ejpam-4568	123	30	as	as	SCONJ
ejpam-4568	123	31	shown	show	VERB
ejpam-4568	123	32	in	in	ADP
ejpam-4568	123	33	table	table	NOUN
ejpam-4568	123	34	(	(	PUNCT
ejpam-4568	123	35	1	1	NUM
ejpam-4568	123	36	)	)	PUNCT
ejpam-4568	123	37	.	.	PUNCT
ejpam-4568	124	1	5	5	X
ejpam-4568	124	2	.	.	X
ejpam-4568	124	3	conclusions	conclusion	NOUN
ejpam-4568	124	4	we	we	PRON
ejpam-4568	124	5	presented	present	VERB
ejpam-4568	124	6	a	a	DET
ejpam-4568	124	7	powerful	powerful	ADJ
ejpam-4568	124	8	conjugate	conjugate	ADJ
ejpam-4568	124	9	gradient	gradient	NOUN
ejpam-4568	124	10	technique	technique	NOUN
ejpam-4568	124	11	.	.	PUNCT
ejpam-4568	125	1	in	in	ADP
ejpam-4568	125	2	addition	addition	NOUN
ejpam-4568	125	3	to	to	ADP
ejpam-4568	125	4	meeting	meet	VERB
ejpam-4568	125	5	the	the	DET
ejpam-4568	125	6	adequate	adequate	ADJ
ejpam-4568	125	7	descent	descent	NOUN
ejpam-4568	125	8	criterion	criterion	NOUN
ejpam-4568	125	9	,	,	PUNCT
ejpam-4568	125	10	the	the	DET
ejpam-4568	125	11	proposed	propose	VERB
ejpam-4568	125	12	approach	approach	NOUN
ejpam-4568	125	13	is	be	AUX
ejpam-4568	125	14	globally	globally	ADV
ejpam-4568	125	15	convergent	convergent	ADJ
ejpam-4568	125	16	.	.	PUNCT
ejpam-4568	126	1	according	accord	VERB
ejpam-4568	126	2	to	to	ADP
ejpam-4568	126	3	numerical	numerical	ADJ
ejpam-4568	126	4	findings	finding	NOUN
ejpam-4568	126	5	,	,	PUNCT
ejpam-4568	126	6	the	the	DET
ejpam-4568	126	7	approach	approach	NOUN
ejpam-4568	126	8	operates	operate	VERB
ejpam-4568	126	9	well	well	ADV
ejpam-4568	126	10	in	in	ADP
ejpam-4568	126	11	practice	practice	NOUN
ejpam-4568	126	12	and	and	CCONJ
ejpam-4568	126	13	is	be	AUX
ejpam-4568	126	14	superior	superior	ADJ
ejpam-4568	126	15	than	than	ADP
ejpam-4568	126	16	the	the	DET
ejpam-4568	126	17	widely	widely	ADV
ejpam-4568	126	18	used	use	VERB
ejpam-4568	126	19	fr	fr	PROPN
ejpam-4568	126	20	method	method	NOUN
ejpam-4568	126	21	.	.	PUNCT
ejpam-4568	127	1	we	we	PRON
ejpam-4568	127	2	also	also	ADV
ejpam-4568	127	3	looked	look	VERB
ejpam-4568	127	4	at	at	ADP
ejpam-4568	127	5	our	our	PRON
ejpam-4568	127	6	approach	approach	NOUN
ejpam-4568	127	7	’s	’s	PART
ejpam-4568	127	8	aptitude	aptitude	NOUN
ejpam-4568	127	9	for	for	ADP
ejpam-4568	127	10	resolving	resolve	VERB
ejpam-4568	127	11	several	several	ADJ
ejpam-4568	127	12	practical	practical	ADJ
ejpam-4568	127	13	problems	problem	NOUN
ejpam-4568	127	14	.	.	PUNCT
ejpam-4568	128	1	in	in	ADP
ejpam-4568	128	2	this	this	DET
ejpam-4568	128	3	manner	manner	NOUN
ejpam-4568	128	4	,	,	PUNCT
ejpam-4568	128	5	a	a	DET
ejpam-4568	128	6	typical	typical	ADJ
ejpam-4568	128	7	issue	issue	NOUN
ejpam-4568	128	8	from	from	ADP
ejpam-4568	128	9	applications	application	NOUN
ejpam-4568	128	10	for	for	ADP
ejpam-4568	128	11	image	image	NOUN
ejpam-4568	128	12	processing	processing	NOUN
ejpam-4568	128	13	was	be	AUX
ejpam-4568	128	14	taken	take	VERB
ejpam-4568	128	15	into	into	ADP
ejpam-4568	128	16	account	account	NOUN
ejpam-4568	128	17	.	.	PUNCT
ejpam-4568	129	1	we	we	PRON
ejpam-4568	129	2	demonstrated	demonstrate	VERB
ejpam-4568	129	3	the	the	DET
ejpam-4568	129	4	acceptability	acceptability	NOUN
ejpam-4568	129	5	of	of	ADP
ejpam-4568	129	6	the	the	DET
ejpam-4568	129	7	image	image	NOUN
ejpam-4568	129	8	that	that	PRON
ejpam-4568	129	9	our	our	PRON
ejpam-4568	129	10	approach	approach	NOUN
ejpam-4568	129	11	restored	restore	VERB
ejpam-4568	129	12	.	.	PUNCT
ejpam-4568	130	1	b.	b.	PROPN
ejpam-4568	130	2	a.	a.	PROPN
ejpam-4568	130	3	hassan	hassan	PROPN
ejpam-4568	130	4	,	,	PUNCT
ejpam-4568	130	5	h.	h.	PROPN
ejpam-4568	130	6	m.	m.	PROPN
ejpam-4568	130	7	sadiq	sadiq	PROPN
ejpam-4568	130	8	/	/	SYM
ejpam-4568	130	9	eur	eur	PROPN
ejpam-4568	130	10	.	.	PUNCT
ejpam-4568	131	1	j.	j.	PROPN
ejpam-4568	131	2	pure	pure	PROPN
ejpam-4568	131	3	appl	appl	PROPN
ejpam-4568	131	4	.	.	PROPN
ejpam-4568	131	5	math	math	PROPN
ejpam-4568	131	6	,	,	PUNCT
ejpam-4568	131	7	15	15	NUM
ejpam-4568	131	8	(	(	PUNCT
ejpam-4568	131	9	4	4	NUM
ejpam-4568	131	10	)	)	PUNCT
ejpam-4568	131	11	(	(	PUNCT
ejpam-4568	131	12	2022	2022	NUM
ejpam-4568	131	13	)	)	PUNCT
ejpam-4568	131	14	,	,	PUNCT
ejpam-4568	131	15	2011	2011	NUM
ejpam-4568	131	16	-	-	SYM
ejpam-4568	131	17	2021	2021	NUM
ejpam-4568	131	18	2018	2018	NUM
ejpam-4568	131	19	figure	figure	NOUN
ejpam-4568	131	20	1	1	NUM
ejpam-4568	131	21	:	:	PUNCT
ejpam-4568	131	22	demonstrates	demonstrate	VERB
ejpam-4568	131	23	the	the	DET
ejpam-4568	131	24	results	result	NOUN
ejpam-4568	131	25	of	of	ADP
ejpam-4568	131	26	algorithms	algorithm	NOUN
ejpam-4568	131	27	fr	fr	PROPN
ejpam-4568	131	28	,	,	PUNCT
ejpam-4568	131	29	bv1	bv1	PROPN
ejpam-4568	131	30	,	,	PUNCT
ejpam-4568	131	31	bv2	bv2	NOUN
ejpam-4568	131	32	and	and	CCONJ
ejpam-4568	131	33	bv3	bv3	ADV
ejpam-4568	131	34	of	of	ADP
ejpam-4568	131	35	256×	256×	NUM
ejpam-4568	131	36	256	256	NUM
ejpam-4568	131	37	lena	lena	PROPN
ejpam-4568	131	38	image	image	NOUN
ejpam-4568	131	39	.	.	PUNCT
ejpam-4568	132	1	figure	figure	NOUN
ejpam-4568	132	2	2	2	NUM
ejpam-4568	132	3	:	:	PUNCT
ejpam-4568	132	4	demonstrates	demonstrate	VERB
ejpam-4568	132	5	the	the	DET
ejpam-4568	132	6	results	result	NOUN
ejpam-4568	132	7	of	of	ADP
ejpam-4568	132	8	algorithms	algorithm	NOUN
ejpam-4568	132	9	fr	fr	PROPN
ejpam-4568	132	10	,	,	PUNCT
ejpam-4568	132	11	bv1	bv1	PROPN
ejpam-4568	132	12	,	,	PUNCT
ejpam-4568	132	13	bv2	bv2	NOUN
ejpam-4568	132	14	and	and	CCONJ
ejpam-4568	132	15	bv3	bv3	ADV
ejpam-4568	132	16	of	of	ADP
ejpam-4568	132	17	256×	256×	NUM
ejpam-4568	132	18	256	256	NUM
ejpam-4568	132	19	house	house	NOUN
ejpam-4568	132	20	image	image	NOUN
ejpam-4568	132	21	.	.	PUNCT
ejpam-4568	133	1	b.	b.	PROPN
ejpam-4568	133	2	a.	a.	PROPN
ejpam-4568	133	3	hassan	hassan	PROPN
ejpam-4568	133	4	,	,	PUNCT
ejpam-4568	133	5	h.	h.	PROPN
ejpam-4568	133	6	m.	m.	PROPN
ejpam-4568	133	7	sadiq	sadiq	PROPN
ejpam-4568	133	8	/	/	SYM
ejpam-4568	133	9	eur	eur	PROPN
ejpam-4568	133	10	.	.	PUNCT
ejpam-4568	134	1	j.	j.	PROPN
ejpam-4568	134	2	pure	pure	PROPN
ejpam-4568	134	3	appl	appl	PROPN
ejpam-4568	134	4	.	.	PROPN
ejpam-4568	134	5	math	math	PROPN
ejpam-4568	134	6	,	,	PUNCT
ejpam-4568	134	7	15	15	NUM
ejpam-4568	134	8	(	(	PUNCT
ejpam-4568	134	9	4	4	NUM
ejpam-4568	134	10	)	)	PUNCT
ejpam-4568	134	11	(	(	PUNCT
ejpam-4568	134	12	2022	2022	NUM
ejpam-4568	134	13	)	)	PUNCT
ejpam-4568	134	14	,	,	PUNCT
ejpam-4568	134	15	2011	2011	NUM
ejpam-4568	134	16	-	-	SYM
ejpam-4568	134	17	2021	2021	NUM
ejpam-4568	134	18	2019	2019	NUM
ejpam-4568	134	19	figure	figure	NOUN
ejpam-4568	134	20	3	3	NUM
ejpam-4568	134	21	:	:	PUNCT
ejpam-4568	134	22	demonstrates	demonstrate	VERB
ejpam-4568	134	23	the	the	DET
ejpam-4568	134	24	results	result	NOUN
ejpam-4568	134	25	of	of	ADP
ejpam-4568	134	26	algorithms	algorithm	NOUN
ejpam-4568	134	27	fr	fr	PROPN
ejpam-4568	134	28	,	,	PUNCT
ejpam-4568	134	29	bv1	bv1	PROPN
ejpam-4568	134	30	,	,	PUNCT
ejpam-4568	134	31	bv2	bv2	NOUN
ejpam-4568	134	32	and	and	CCONJ
ejpam-4568	134	33	bv3	bv3	ADV
ejpam-4568	134	34	of	of	ADP
ejpam-4568	134	35	256×	256×	NUM
ejpam-4568	134	36	256	256	NUM
ejpam-4568	134	37	elaine	elaine	PROPN
ejpam-4568	134	38	image	image	NOUN
ejpam-4568	134	39	.	.	PUNCT
ejpam-4568	135	1	figure	figure	VERB
ejpam-4568	135	2	4	4	NUM
ejpam-4568	135	3	:	:	PUNCT
ejpam-4568	135	4	demonstrates	demonstrate	VERB
ejpam-4568	135	5	the	the	DET
ejpam-4568	135	6	results	result	NOUN
ejpam-4568	135	7	of	of	ADP
ejpam-4568	135	8	algorithms	algorithm	NOUN
ejpam-4568	135	9	fr	fr	PROPN
ejpam-4568	135	10	,	,	PUNCT
ejpam-4568	135	11	bv1	bv1	PROPN
ejpam-4568	135	12	,	,	PUNCT
ejpam-4568	135	13	bv2	bv2	NOUN
ejpam-4568	135	14	and	and	CCONJ
ejpam-4568	135	15	bv3	bv3	ADV
ejpam-4568	135	16	of	of	ADP
ejpam-4568	135	17	256×	256×	NUM
ejpam-4568	135	18	256	256	NUM
ejpam-4568	135	19	cameraman	cameraman	ADJ
ejpam-4568	135	20	image	image	NOUN
ejpam-4568	135	21	.	.	PUNCT
ejpam-4568	136	1	references	reference	NOUN
ejpam-4568	136	2	2020	2020	NUM
ejpam-4568	136	3	references	reference	NOUN
ejpam-4568	136	4	[	[	X
ejpam-4568	136	5	1	1	NUM
ejpam-4568	136	6	]	]	PUNCT
ejpam-4568	136	7	mehiddin	mehiddin	PROPN
ejpam-4568	136	8	al	al	PROPN
ejpam-4568	136	9	-	-	PUNCT
ejpam-4568	136	10	baali	baali	PROPN
ejpam-4568	136	11	.	.	PUNCT
ejpam-4568	137	1	descent	descent	NOUN
ejpam-4568	137	2	property	property	NOUN
ejpam-4568	137	3	and	and	CCONJ
ejpam-4568	137	4	global	global	ADJ
ejpam-4568	137	5	convergence	convergence	NOUN
ejpam-4568	137	6	of	of	ADP
ejpam-4568	137	7	the	the	DET
ejpam-4568	137	8	fletcher	fletcher	NOUN
ejpam-4568	137	9	—	—	PUNCT
ejpam-4568	137	10	reeves	reeve	NOUN
ejpam-4568	137	11	method	method	VERB
ejpam-4568	137	12	with	with	ADP
ejpam-4568	137	13	inexact	inexact	ADJ
ejpam-4568	137	14	line	line	NOUN
ejpam-4568	137	15	search	search	NOUN
ejpam-4568	137	16	.	.	PUNCT
ejpam-4568	138	1	i	i	PRON
ejpam-4568	138	2	m	m	VERB
ejpam-4568	138	3	a	a	DET
ejpam-4568	138	4	journal	journal	NOUN
ejpam-4568	138	5	of	of	ADP
ejpam-4568	138	6	numerical	numerical	ADJ
ejpam-4568	138	7	analysis	analysis	NOUN
ejpam-4568	138	8	,	,	PUNCT
ejpam-4568	138	9	5(1):121–124	5(1):121–124	NUM
ejpam-4568	138	10	,	,	PUNCT
ejpam-4568	138	11	1985	1985	NUM
ejpam-4568	138	12	.	.	PUNCT
ejpam-4568	139	1	[	[	X
ejpam-4568	139	2	2	2	NUM
ejpam-4568	139	3	]	]	X
ejpam-4568	139	4	neculai	neculai	ADJ
ejpam-4568	139	5	andrei	andrei	NOUN
ejpam-4568	139	6	.	.	PUNCT
ejpam-4568	140	1	an	an	DET
ejpam-4568	140	2	unconstrained	unconstrained	ADJ
ejpam-4568	140	3	optimization	optimization	NOUN
ejpam-4568	140	4	test	test	NOUN
ejpam-4568	140	5	functions	function	NOUN
ejpam-4568	140	6	collection	collection	NOUN
ejpam-4568	140	7	.	.	PUNCT
ejpam-4568	141	1	adv	adv	PROPN
ejpam-4568	141	2	.	.	PUNCT
ejpam-4568	141	3	model	model	PROPN
ejpam-4568	141	4	.	.	PUNCT
ejpam-4568	142	1	optim	optim	ADJ
ejpam-4568	142	2	,	,	PUNCT
ejpam-4568	142	3	10(1):147–161	10(1):147–161	PROPN
ejpam-4568	142	4	,	,	PUNCT
ejpam-4568	142	5	2008	2008	NUM
ejpam-4568	142	6	.	.	PUNCT
ejpam-4568	143	1	[	[	X
ejpam-4568	143	2	3	3	X
ejpam-4568	143	3	]	]	PUNCT
ejpam-4568	143	4	alan	alan	PROPN
ejpam-4568	143	5	c	c	PROPN
ejpam-4568	143	6	bovik	bovik	NOUN
ejpam-4568	143	7	.	.	PUNCT
ejpam-4568	144	1	handbook	handbook	NOUN
ejpam-4568	144	2	of	of	ADP
ejpam-4568	144	3	image	image	NOUN
ejpam-4568	144	4	and	and	CCONJ
ejpam-4568	144	5	video	video	NOUN
ejpam-4568	144	6	processing	processing	NOUN
ejpam-4568	144	7	.	.	PUNCT
ejpam-4568	145	1	academic	academic	ADJ
ejpam-4568	145	2	press	press	NOUN
ejpam-4568	145	3	,	,	PUNCT
ejpam-4568	145	4	2010	2010	NUM
ejpam-4568	145	5	.	.	PUNCT
ejpam-4568	146	1	[	[	X
ejpam-4568	146	2	4	4	NUM
ejpam-4568	146	3	]	]	X
ejpam-4568	146	4	jian	jian	PROPN
ejpam-4568	146	5	-	-	PUNCT
ejpam-4568	146	6	feng	feng	PROPN
ejpam-4568	146	7	cai	cai	PROPN
ejpam-4568	146	8	,	,	PUNCT
ejpam-4568	146	9	raymond	raymond	PROPN
ejpam-4568	146	10	chan	chan	PROPN
ejpam-4568	146	11	,	,	PUNCT
ejpam-4568	146	12	and	and	CCONJ
ejpam-4568	146	13	benedetta	benedetta	PROPN
ejpam-4568	146	14	morini	morini	PROPN
ejpam-4568	146	15	.	.	PROPN
ejpam-4568	146	16	minimization	minimization	NOUN
ejpam-4568	146	17	of	of	ADP
ejpam-4568	146	18	an	an	DET
ejpam-4568	146	19	edgepreserving	edgepreserving	ADJ
ejpam-4568	146	20	regularization	regularization	NOUN
ejpam-4568	146	21	functional	functional	ADJ
ejpam-4568	146	22	by	by	ADP
ejpam-4568	146	23	conjugate	conjugate	ADJ
ejpam-4568	146	24	gradient	gradient	ADJ
ejpam-4568	146	25	type	type	NOUN
ejpam-4568	146	26	methods	method	NOUN
ejpam-4568	146	27	.	.	PUNCT
ejpam-4568	147	1	in	in	ADP
ejpam-4568	147	2	image	image	NOUN
ejpam-4568	147	3	processing	processing	NOUN
ejpam-4568	147	4	based	base	VERB
ejpam-4568	147	5	on	on	ADP
ejpam-4568	147	6	partial	partial	ADJ
ejpam-4568	147	7	differential	differential	NOUN
ejpam-4568	147	8	equations	equation	NOUN
ejpam-4568	147	9	,	,	PUNCT
ejpam-4568	147	10	pages	page	NOUN
ejpam-4568	147	11	109–122	109–122	NUM
ejpam-4568	147	12	.	.	PUNCT
ejpam-4568	147	13	springer	springer	NOUN
ejpam-4568	147	14	,	,	PUNCT
ejpam-4568	147	15	2007	2007	NUM
ejpam-4568	147	16	.	.	PUNCT
ejpam-4568	148	1	[	[	X
ejpam-4568	148	2	5	5	NUM
ejpam-4568	148	3	]	]	X
ejpam-4568	148	4	jian	jian	PROPN
ejpam-4568	148	5	-	-	PUNCT
ejpam-4568	148	6	feng	feng	PROPN
ejpam-4568	148	7	cai	cai	PROPN
ejpam-4568	148	8	,	,	PUNCT
ejpam-4568	148	9	raymond	raymond	PROPN
ejpam-4568	148	10	h	h	PROPN
ejpam-4568	148	11	chan	chan	PROPN
ejpam-4568	148	12	,	,	PUNCT
ejpam-4568	148	13	and	and	CCONJ
ejpam-4568	148	14	mila	mila	PROPN
ejpam-4568	148	15	nikolova	nikolova	INTJ
ejpam-4568	148	16	.	.	PUNCT
ejpam-4568	149	1	fast	fast	ADJ
ejpam-4568	149	2	two	two	NUM
ejpam-4568	149	3	-	-	PUNCT
ejpam-4568	149	4	phase	phase	NOUN
ejpam-4568	149	5	image	image	NOUN
ejpam-4568	149	6	deblurring	deblurre	VERB
ejpam-4568	149	7	under	under	ADP
ejpam-4568	149	8	impulse	impulse	ADJ
ejpam-4568	149	9	noise	noise	NOUN
ejpam-4568	149	10	.	.	PUNCT
ejpam-4568	149	11	journal	journal	NOUN
ejpam-4568	149	12	of	of	ADP
ejpam-4568	149	13	mathematical	mathematical	ADJ
ejpam-4568	149	14	imaging	imaging	NOUN
ejpam-4568	149	15	and	and	CCONJ
ejpam-4568	149	16	vision	vision	NOUN
ejpam-4568	149	17	,	,	PUNCT
ejpam-4568	149	18	36(1):46–53	36(1):46–53	NUM
ejpam-4568	149	19	,	,	PUNCT
ejpam-4568	149	20	2010	2010	NUM
ejpam-4568	149	21	.	.	PUNCT
ejpam-4568	150	1	[	[	X
ejpam-4568	150	2	6	6	NUM
ejpam-4568	150	3	]	]	X
ejpam-4568	150	4	raymond	raymond	PROPN
ejpam-4568	150	5	h	h	PROPN
ejpam-4568	150	6	chan	chan	PROPN
ejpam-4568	150	7	,	,	PUNCT
ejpam-4568	150	8	chung	chung	PROPN
ejpam-4568	150	9	-	-	PUNCT
ejpam-4568	150	10	wa	wa	X
ejpam-4568	150	11	ho	ho	PROPN
ejpam-4568	150	12	,	,	PUNCT
ejpam-4568	150	13	and	and	CCONJ
ejpam-4568	150	14	mila	mila	PROPN
ejpam-4568	150	15	nikolova	nikolova	VERB
ejpam-4568	150	16	.	.	PUNCT
ejpam-4568	151	1	salt	salt	NOUN
ejpam-4568	151	2	-	-	PUNCT
ejpam-4568	151	3	and	and	CCONJ
ejpam-4568	151	4	-	-	PUNCT
ejpam-4568	151	5	pepper	pepper	NOUN
ejpam-4568	151	6	noise	noise	NOUN
ejpam-4568	151	7	removal	removal	NOUN
ejpam-4568	151	8	by	by	ADP
ejpam-4568	151	9	median	median	ADJ
ejpam-4568	151	10	-	-	PUNCT
ejpam-4568	151	11	type	type	NOUN
ejpam-4568	151	12	noise	noise	NOUN
ejpam-4568	151	13	detectors	detector	NOUN
ejpam-4568	151	14	and	and	CCONJ
ejpam-4568	151	15	detail	detail	NOUN
ejpam-4568	151	16	-	-	PUNCT
ejpam-4568	151	17	preserving	preserve	VERB
ejpam-4568	151	18	regularization	regularization	NOUN
ejpam-4568	151	19	.	.	PUNCT
ejpam-4568	152	1	ieee	ieee	NOUN
ejpam-4568	152	2	transactions	transaction	NOUN
ejpam-4568	152	3	on	on	ADP
ejpam-4568	152	4	image	image	NOUN
ejpam-4568	152	5	processing	processing	NOUN
ejpam-4568	152	6	,	,	PUNCT
ejpam-4568	152	7	14(10):1479–1485	14(10):1479–1485	NUM
ejpam-4568	152	8	,	,	PUNCT
ejpam-4568	152	9	2005	2005	NUM
ejpam-4568	152	10	.	.	PUNCT
ejpam-4568	153	1	[	[	X
ejpam-4568	153	2	7	7	X
ejpam-4568	153	3	]	]	X
ejpam-4568	153	4	tao	tao	PROPN
ejpam-4568	153	5	chen	chen	PROPN
ejpam-4568	153	6	and	and	CCONJ
ejpam-4568	153	7	hong	hong	PROPN
ejpam-4568	153	8	ren	ren	PROPN
ejpam-4568	153	9	wu	wu	PROPN
ejpam-4568	153	10	.	.	PUNCT
ejpam-4568	154	1	adaptive	adaptive	ADJ
ejpam-4568	154	2	impulse	impulse	ADJ
ejpam-4568	154	3	detection	detection	NOUN
ejpam-4568	154	4	using	use	VERB
ejpam-4568	154	5	center	center	NOUN
ejpam-4568	154	6	-	-	PUNCT
ejpam-4568	154	7	weighted	weight	VERB
ejpam-4568	154	8	median	median	ADJ
ejpam-4568	154	9	filters	filter	NOUN
ejpam-4568	154	10	.	.	PUNCT
ejpam-4568	155	1	ieee	ieee	NOUN
ejpam-4568	155	2	signal	signal	NOUN
ejpam-4568	155	3	processing	processing	NOUN
ejpam-4568	155	4	letters	letter	NOUN
ejpam-4568	155	5	,	,	PUNCT
ejpam-4568	155	6	8(1):1–3	8(1):1–3	NUM
ejpam-4568	155	7	,	,	PUNCT
ejpam-4568	155	8	2001	2001	NUM
ejpam-4568	155	9	.	.	PUNCT
ejpam-4568	156	1	[	[	X
ejpam-4568	156	2	8	8	NUM
ejpam-4568	156	3	]	]	X
ejpam-4568	156	4	yu	yu	PROPN
ejpam-4568	156	5	-	-	PROPN
ejpam-4568	156	6	hong	hong	PROPN
ejpam-4568	156	7	dai	dai	PROPN
ejpam-4568	156	8	and	and	CCONJ
ejpam-4568	156	9	yaxiang	yaxiang	PROPN
ejpam-4568	156	10	yuan	yuan	PROPN
ejpam-4568	156	11	.	.	PUNCT
ejpam-4568	157	1	a	a	DET
ejpam-4568	157	2	nonlinear	nonlinear	ADJ
ejpam-4568	157	3	conjugate	conjugate	ADJ
ejpam-4568	157	4	gradient	gradient	NOUN
ejpam-4568	157	5	method	method	NOUN
ejpam-4568	157	6	with	with	ADP
ejpam-4568	157	7	a	a	DET
ejpam-4568	157	8	strong	strong	ADJ
ejpam-4568	157	9	global	global	ADJ
ejpam-4568	157	10	convergence	convergence	NOUN
ejpam-4568	157	11	property	property	NOUN
ejpam-4568	157	12	.	.	PUNCT
ejpam-4568	158	1	siam	siam	PROPN
ejpam-4568	158	2	journal	journal	PROPN
ejpam-4568	158	3	on	on	ADP
ejpam-4568	158	4	optimization	optimization	NOUN
ejpam-4568	158	5	,	,	PUNCT
ejpam-4568	158	6	10(1):177–182	10(1):177–182	NUM
ejpam-4568	158	7	,	,	PUNCT
ejpam-4568	158	8	1999	1999	NUM
ejpam-4568	158	9	.	.	PUNCT
ejpam-4568	159	1	[	[	X
ejpam-4568	159	2	9	9	NUM
ejpam-4568	159	3	]	]	X
ejpam-4568	159	4	john	john	PROPN
ejpam-4568	159	5	e	e	PROPN
ejpam-4568	159	6	dennis	dennis	PROPN
ejpam-4568	159	7	jr	jr	PROPN
ejpam-4568	159	8	.	.	PROPN
ejpam-4568	159	9	practical	practical	ADJ
ejpam-4568	159	10	methods	method	NOUN
ejpam-4568	159	11	of	of	ADP
ejpam-4568	159	12	optimization	optimization	NOUN
ejpam-4568	159	13	,	,	PUNCT
ejpam-4568	159	14	vol	vol	NOUN
ejpam-4568	159	15	.	.	PROPN
ejpam-4568	159	16	1	1	NUM
ejpam-4568	159	17	:	:	PUNCT
ejpam-4568	159	18	unconstrained	unconstrained	ADJ
ejpam-4568	159	19	optimization	optimization	NOUN
ejpam-4568	159	20	(	(	PUNCT
ejpam-4568	159	21	r.	r.	PROPN
ejpam-4568	159	22	fletcher	fletcher	PROPN
ejpam-4568	159	23	)	)	PUNCT
ejpam-4568	159	24	,	,	PUNCT
ejpam-4568	159	25	1982	1982	NUM
ejpam-4568	159	26	.	.	PUNCT
ejpam-4568	160	1	[	[	X
ejpam-4568	160	2	10	10	NUM
ejpam-4568	160	3	]	]	X
ejpam-4568	160	4	reeves	reeves	PROPN
ejpam-4568	160	5	fletcher	fletcher	PROPN
ejpam-4568	160	6	and	and	CCONJ
ejpam-4568	160	7	colin	colin	PROPN
ejpam-4568	160	8	m	m	PROPN
ejpam-4568	160	9	reeves	reeves	PROPN
ejpam-4568	160	10	.	.	PUNCT
ejpam-4568	161	1	function	function	NOUN
ejpam-4568	161	2	minimization	minimization	NOUN
ejpam-4568	161	3	by	by	ADP
ejpam-4568	161	4	conjugate	conjugate	ADJ
ejpam-4568	161	5	gradients	gradient	NOUN
ejpam-4568	161	6	.	.	PUNCT
ejpam-4568	162	1	the	the	DET
ejpam-4568	162	2	computer	computer	NOUN
ejpam-4568	162	3	journal	journal	NOUN
ejpam-4568	162	4	,	,	PUNCT
ejpam-4568	162	5	7(2):149–154	7(2):149–154	NUM
ejpam-4568	162	6	,	,	PUNCT
ejpam-4568	162	7	1964	1964	NUM
ejpam-4568	162	8	.	.	PUNCT
ejpam-4568	163	1	[	[	X
ejpam-4568	163	2	11	11	NUM
ejpam-4568	163	3	]	]	X
ejpam-4568	163	4	william	william	PROPN
ejpam-4568	163	5	w	w	PROPN
ejpam-4568	163	6	hager	hager	PROPN
ejpam-4568	163	7	and	and	CCONJ
ejpam-4568	163	8	hongchao	hongchao	PROPN
ejpam-4568	163	9	zhang	zhang	PROPN
ejpam-4568	163	10	.	.	PUNCT
ejpam-4568	164	1	a	a	DET
ejpam-4568	164	2	survey	survey	NOUN
ejpam-4568	164	3	of	of	ADP
ejpam-4568	164	4	nonlinear	nonlinear	ADJ
ejpam-4568	164	5	conjugate	conjugate	ADJ
ejpam-4568	164	6	gradient	gradient	ADJ
ejpam-4568	164	7	methods	method	NOUN
ejpam-4568	164	8	.	.	PUNCT
ejpam-4568	165	1	pacific	pacific	PROPN
ejpam-4568	165	2	journal	journal	PROPN
ejpam-4568	165	3	of	of	ADP
ejpam-4568	165	4	optimization	optimization	NOUN
ejpam-4568	165	5	,	,	PUNCT
ejpam-4568	165	6	2(1):35–58	2(1):35–58	NUM
ejpam-4568	165	7	,	,	PUNCT
ejpam-4568	165	8	2006	2006	NUM
ejpam-4568	165	9	.	.	PUNCT
ejpam-4568	166	1	[	[	X
ejpam-4568	166	2	12	12	NUM
ejpam-4568	166	3	]	]	X
ejpam-4568	166	4	basim	basim	NOUN
ejpam-4568	166	5	a	a	DET
ejpam-4568	166	6	hassan	hassan	PROPN
ejpam-4568	166	7	,	,	PUNCT
ejpam-4568	166	8	fadhil	fadhil	NOUN
ejpam-4568	166	9	alfarag	alfarag	VERB
ejpam-4568	166	10	,	,	PUNCT
ejpam-4568	166	11	and	and	CCONJ
ejpam-4568	166	12	snezana	snezana	PROPN
ejpam-4568	166	13	djordjevic	djordjevic	PROPN
ejpam-4568	166	14	.	.	PUNCT
ejpam-4568	167	1	new	new	ADJ
ejpam-4568	167	2	step	step	NOUN
ejpam-4568	167	3	sizes	size	NOUN
ejpam-4568	167	4	of	of	ADP
ejpam-4568	167	5	the	the	DET
ejpam-4568	167	6	gradient	gradient	ADJ
ejpam-4568	167	7	methods	method	NOUN
ejpam-4568	167	8	for	for	ADP
ejpam-4568	167	9	unconstrained	unconstrained	ADJ
ejpam-4568	167	10	optimization	optimization	NOUN
ejpam-4568	167	11	problem	problem	NOUN
ejpam-4568	167	12	.	.	PUNCT
ejpam-4568	168	1	italian	italian	ADJ
ejpam-4568	168	2	journal	journal	NOUN
ejpam-4568	168	3	of	of	ADP
ejpam-4568	168	4	pure	pure	ADJ
ejpam-4568	168	5	and	and	CCONJ
ejpam-4568	168	6	applied	applied	ADJ
ejpam-4568	168	7	mathematics	mathematic	NOUN
ejpam-4568	168	8	,	,	PUNCT
ejpam-4568	168	9	page	page	NOUN
ejpam-4568	168	10	583	583	NUM
ejpam-4568	168	11	.	.	PUNCT
ejpam-4568	169	1	[	[	X
ejpam-4568	169	2	13	13	NUM
ejpam-4568	169	3	]	]	SYM
ejpam-4568	169	4	basim	basim	NOUN
ejpam-4568	169	5	a	a	DET
ejpam-4568	169	6	hassan	hassan	PROPN
ejpam-4568	169	7	and	and	CCONJ
ejpam-4568	169	8	abdulrahman	abdulrahman	PROPN
ejpam-4568	169	9	r	r	PROPN
ejpam-4568	169	10	ayoob	ayoob	PROPN
ejpam-4568	169	11	.	.	PUNCT
ejpam-4568	170	1	an	an	DET
ejpam-4568	170	2	adaptive	adaptive	ADJ
ejpam-4568	170	3	quasi	quasi	ADJ
ejpam-4568	170	4	-	-	ADJ
ejpam-4568	170	5	newton	newton	PROPN
ejpam-4568	170	6	equation	equation	NOUN
ejpam-4568	170	7	for	for	ADP
ejpam-4568	170	8	unconstrained	unconstrained	ADJ
ejpam-4568	170	9	optimization	optimization	NOUN
ejpam-4568	170	10	.	.	PUNCT
ejpam-4568	171	1	in	in	ADP
ejpam-4568	171	2	2021	2021	NUM
ejpam-4568	171	3	2nd	2nd	ADJ
ejpam-4568	171	4	information	information	NOUN
ejpam-4568	171	5	technology	technology	NOUN
ejpam-4568	171	6	to	to	PART
ejpam-4568	171	7	enhance	enhance	VERB
ejpam-4568	171	8	e	e	NOUN
ejpam-4568	171	9	-	-	ADJ
ejpam-4568	171	10	learning	learning	ADJ
ejpam-4568	171	11	and	and	CCONJ
ejpam-4568	171	12	other	other	ADJ
ejpam-4568	171	13	application	application	NOUN
ejpam-4568	171	14	(	(	PUNCT
ejpam-4568	171	15	it	it	PRON
ejpam-4568	171	16	-	-	PUNCT
ejpam-4568	171	17	ela	ela	PROPN
ejpam-4568	171	18	)	)	PUNCT
ejpam-4568	171	19	,	,	PUNCT
ejpam-4568	171	20	pages	page	NOUN
ejpam-4568	171	21	1–5	1–5	PROPN
ejpam-4568	171	22	.	.	PUNCT
ejpam-4568	171	23	ieee	ieee	PROPN
ejpam-4568	171	24	,	,	PUNCT
ejpam-4568	171	25	2021	2021	NUM
ejpam-4568	171	26	.	.	PUNCT
ejpam-4568	172	1	references	reference	NOUN
ejpam-4568	172	2	2021	2021	NUM
ejpam-4568	172	3	[	[	X
ejpam-4568	172	4	14	14	NUM
ejpam-4568	172	5	]	]	X
ejpam-4568	172	6	basim	basim	NOUN
ejpam-4568	172	7	a	a	DET
ejpam-4568	172	8	hassan	hassan	PROPN
ejpam-4568	172	9	and	and	CCONJ
ejpam-4568	172	10	abdulrahman	abdulrahman	PROPN
ejpam-4568	172	11	r	r	PROPN
ejpam-4568	172	12	ayoob	ayoob	NOUN
ejpam-4568	172	13	.	.	PUNCT
ejpam-4568	173	1	on	on	ADP
ejpam-4568	173	2	the	the	DET
ejpam-4568	173	3	new	new	ADJ
ejpam-4568	173	4	quasi	quasi	PROPN
ejpam-4568	173	5	-	-	ADJ
ejpam-4568	173	6	newton	newton	PROPN
ejpam-4568	173	7	equation	equation	NOUN
ejpam-4568	173	8	for	for	ADP
ejpam-4568	173	9	unconstrained	unconstrained	ADJ
ejpam-4568	173	10	optimization	optimization	NOUN
ejpam-4568	173	11	.	.	PUNCT
ejpam-4568	174	1	in	in	ADP
ejpam-4568	174	2	2022	2022	NUM
ejpam-4568	174	3	8th	8th	ADJ
ejpam-4568	174	4	international	international	ADJ
ejpam-4568	174	5	engineering	engineering	NOUN
ejpam-4568	174	6	conference	conference	NOUN
ejpam-4568	174	7	on	on	ADP
ejpam-4568	174	8	sustainable	sustainable	ADJ
ejpam-4568	174	9	technology	technology	NOUN
ejpam-4568	174	10	and	and	CCONJ
ejpam-4568	174	11	development	development	NOUN
ejpam-4568	174	12	(	(	PUNCT
ejpam-4568	174	13	iec	iec	PROPN
ejpam-4568	174	14	)	)	PUNCT
ejpam-4568	174	15	,	,	PUNCT
ejpam-4568	174	16	pages	page	VERB
ejpam-4568	174	17	168–172	168–172	NUM
ejpam-4568	174	18	.	.	PUNCT
ejpam-4568	175	1	ieee	ieee	NOUN
ejpam-4568	175	2	,	,	PUNCT
ejpam-4568	175	3	2022	2022	NUM
ejpam-4568	175	4	.	.	PUNCT
ejpam-4568	176	1	[	[	X
ejpam-4568	176	2	15	15	NUM
ejpam-4568	176	3	]	]	X
ejpam-4568	176	4	basim	basim	NOUN
ejpam-4568	176	5	a	a	DET
ejpam-4568	176	6	hassan	hassan	PROPN
ejpam-4568	176	7	and	and	CCONJ
ejpam-4568	176	8	ranen	ranen	PROPN
ejpam-4568	176	9	m.	m.	PROPN
ejpam-4568	176	10	sulaiman	sulaiman	PROPN
ejpam-4568	176	11	.	.	PUNCT
ejpam-4568	177	1	using	use	VERB
ejpam-4568	177	2	a	a	DET
ejpam-4568	177	3	new	new	ADJ
ejpam-4568	177	4	type	type	NOUN
ejpam-4568	177	5	of	of	ADP
ejpam-4568	177	6	formula	formula	NOUN
ejpam-4568	177	7	conjugate	conjugate	VERB
ejpam-4568	177	8	on	on	ADP
ejpam-4568	177	9	the	the	DET
ejpam-4568	177	10	gradient	gradient	ADJ
ejpam-4568	177	11	methods	method	NOUN
ejpam-4568	177	12	.	.	PUNCT
ejpam-4568	178	1	indonesian	indonesian	ADJ
ejpam-4568	178	2	journal	journal	PROPN
ejpam-4568	178	3	of	of	ADP
ejpam-4568	178	4	electrical	electrical	ADJ
ejpam-4568	178	5	engineering	engineering	NOUN
ejpam-4568	178	6	and	and	CCONJ
ejpam-4568	178	7	computer	computer	NOUN
ejpam-4568	178	8	science	science	NOUN
ejpam-4568	178	9	,	,	PUNCT
ejpam-4568	178	10	27(1):86–91	27(1):86–91	NUM
ejpam-4568	178	11	,	,	PUNCT
ejpam-4568	178	12	2022	2022	NUM
ejpam-4568	178	13	.	.	PUNCT
ejpam-4568	179	1	[	[	X
ejpam-4568	179	2	16	16	NUM
ejpam-4568	179	3	]	]	X
ejpam-4568	179	4	p	p	X
ejpam-4568	179	5	kaelo	kaelo	PROPN
ejpam-4568	179	6	,	,	PUNCT
ejpam-4568	179	7	p	p	NOUN
ejpam-4568	179	8	mtagulwa	mtagulwa	NOUN
ejpam-4568	179	9	,	,	PUNCT
ejpam-4568	179	10	and	and	CCONJ
ejpam-4568	179	11	mv	mv	PROPN
ejpam-4568	179	12	thuto	thuto	PROPN
ejpam-4568	179	13	.	.	PUNCT
ejpam-4568	180	1	a	a	DET
ejpam-4568	180	2	globally	globally	ADV
ejpam-4568	180	3	convergent	convergent	ADJ
ejpam-4568	180	4	hybrid	hybrid	ADJ
ejpam-4568	180	5	conjugate	conjugate	ADJ
ejpam-4568	180	6	gradient	gradient	NOUN
ejpam-4568	180	7	method	method	NOUN
ejpam-4568	180	8	with	with	ADP
ejpam-4568	180	9	strong	strong	ADJ
ejpam-4568	180	10	wolfe	wolfe	PROPN
ejpam-4568	180	11	conditions	condition	NOUN
ejpam-4568	180	12	for	for	ADP
ejpam-4568	180	13	unconstrained	unconstrained	ADJ
ejpam-4568	180	14	optimization	optimization	NOUN
ejpam-4568	180	15	.	.	PUNCT
ejpam-4568	181	1	mathematical	mathematical	ADJ
ejpam-4568	181	2	sciences	sciences	PROPN
ejpam-4568	181	3	,	,	PUNCT
ejpam-4568	181	4	14(1):1–9	14(1):1–9	NUM
ejpam-4568	181	5	,	,	PUNCT
ejpam-4568	181	6	2020	2020	NUM
ejpam-4568	181	7	.	.	PUNCT
ejpam-4568	182	1	[	[	X
ejpam-4568	182	2	17	17	NUM
ejpam-4568	182	3	]	]	X
ejpam-4568	182	4	morteza	morteza	NOUN
ejpam-4568	182	5	kimiaei	kimiaei	ADJ
ejpam-4568	182	6	and	and	CCONJ
ejpam-4568	182	7	majid	majid	PROPN
ejpam-4568	182	8	rostami	rostami	PROPN
ejpam-4568	182	9	.	.	PUNCT
ejpam-4568	183	1	impulse	impulse	ADJ
ejpam-4568	183	2	noise	noise	NOUN
ejpam-4568	183	3	removal	removal	NOUN
ejpam-4568	183	4	based	base	VERB
ejpam-4568	183	5	on	on	ADP
ejpam-4568	183	6	new	new	ADJ
ejpam-4568	183	7	hybrid	hybrid	ADJ
ejpam-4568	183	8	conjugate	conjugate	ADJ
ejpam-4568	183	9	gradient	gradient	ADJ
ejpam-4568	183	10	approach	approach	NOUN
ejpam-4568	183	11	.	.	PUNCT
ejpam-4568	184	1	kybernetika	kybernetika	PROPN
ejpam-4568	184	2	,	,	PUNCT
ejpam-4568	184	3	52(5):791–823	52(5):791–823	NUM
ejpam-4568	184	4	,	,	PUNCT
ejpam-4568	184	5	2016	2016	NUM
ejpam-4568	184	6	.	.	PUNCT
ejpam-4568	185	1	[	[	X
ejpam-4568	185	2	18	18	NUM
ejpam-4568	185	3	]	]	X
ejpam-4568	185	4	peter	peter	PROPN
ejpam-4568	185	5	mtagulwa	mtagulwa	PROPN
ejpam-4568	185	6	and	and	CCONJ
ejpam-4568	185	7	p	p	PRON
ejpam-4568	185	8	kaelo	kaelo	PROPN
ejpam-4568	185	9	.	.	PUNCT
ejpam-4568	186	1	a	a	DET
ejpam-4568	186	2	convergent	convergent	NOUN
ejpam-4568	186	3	modified	modify	VERB
ejpam-4568	186	4	hs	hs	ADJ
ejpam-4568	186	5	-	-	PUNCT
ejpam-4568	186	6	dy	dy	VERB
ejpam-4568	186	7	hybrid	hybrid	ADJ
ejpam-4568	186	8	conjugate	conjugate	ADJ
ejpam-4568	186	9	gradient	gradient	NOUN
ejpam-4568	186	10	method	method	NOUN
ejpam-4568	186	11	for	for	ADP
ejpam-4568	186	12	unconstrained	unconstrained	ADJ
ejpam-4568	186	13	optimization	optimization	NOUN
ejpam-4568	186	14	problems	problem	NOUN
ejpam-4568	186	15	.	.	PUNCT
ejpam-4568	187	1	journal	journal	NOUN
ejpam-4568	187	2	of	of	ADP
ejpam-4568	187	3	information	information	NOUN
ejpam-4568	187	4	and	and	CCONJ
ejpam-4568	187	5	optimization	optimization	NOUN
ejpam-4568	187	6	sciences	science	NOUN
ejpam-4568	187	7	,	,	PUNCT
ejpam-4568	187	8	40(1):97–113	40(1):97–113	NUM
ejpam-4568	187	9	,	,	PUNCT
ejpam-4568	187	10	2019	2019	NUM
ejpam-4568	187	11	.	.	PUNCT
ejpam-4568	188	1	[	[	X
ejpam-4568	188	2	19	19	NUM
ejpam-4568	188	3	]	]	X
ejpam-4568	188	4	mila	mila	NOUN
ejpam-4568	188	5	nikolova	nikolova	VERB
ejpam-4568	188	6	.	.	PUNCT
ejpam-4568	189	1	a	a	DET
ejpam-4568	189	2	variational	variational	ADJ
ejpam-4568	189	3	approach	approach	NOUN
ejpam-4568	189	4	to	to	PART
ejpam-4568	189	5	remove	remove	VERB
ejpam-4568	189	6	outliers	outlier	NOUN
ejpam-4568	189	7	and	and	CCONJ
ejpam-4568	189	8	impulse	impulse	ADJ
ejpam-4568	189	9	noise	noise	NOUN
ejpam-4568	189	10	.	.	PUNCT
ejpam-4568	190	1	journal	journal	NOUN
ejpam-4568	190	2	of	of	ADP
ejpam-4568	190	3	mathematical	mathematical	ADJ
ejpam-4568	190	4	imaging	imaging	NOUN
ejpam-4568	190	5	and	and	CCONJ
ejpam-4568	190	6	vision	vision	NOUN
ejpam-4568	190	7	,	,	PUNCT
ejpam-4568	190	8	20(1):99–120	20(1):99–120	NUM
ejpam-4568	190	9	,	,	PUNCT
ejpam-4568	190	10	2004	2004	NUM
ejpam-4568	190	11	.	.	PUNCT
ejpam-4568	191	1	[	[	X
ejpam-4568	191	2	20	20	NUM
ejpam-4568	191	3	]	]	PUNCT
ejpam-4568	191	4	jorge	jorge	NOUN
ejpam-4568	191	5	nocedal	nocedal	PROPN
ejpam-4568	191	6	and	and	CCONJ
ejpam-4568	191	7	stephen	stephen	PROPN
ejpam-4568	191	8	j	j	PROPN
ejpam-4568	191	9	wright	wright	PROPN
ejpam-4568	191	10	.	.	PUNCT
ejpam-4568	191	11	numerical	numerical	PROPN
ejpam-4568	191	12	optimization	optimization	PROPN
ejpam-4568	191	13	,	,	PUNCT
ejpam-4568	191	14	2006	2006	NUM
ejpam-4568	191	15	.	.	PUNCT
ejpam-4568	192	1	[	[	X
ejpam-4568	192	2	21	21	NUM
ejpam-4568	192	3	]	]	X
ejpam-4568	192	4	gaohang	gaohang	PROPN
ejpam-4568	192	5	yu	yu	PROPN
ejpam-4568	192	6	,	,	PUNCT
ejpam-4568	192	7	liqun	liqun	PROPN
ejpam-4568	192	8	qi	qi	PROPN
ejpam-4568	192	9	,	,	PUNCT
ejpam-4568	192	10	yimin	yimin	PROPN
ejpam-4568	192	11	sun	sun	PROPN
ejpam-4568	192	12	,	,	PUNCT
ejpam-4568	192	13	and	and	CCONJ
ejpam-4568	192	14	yi	yi	PROPN
ejpam-4568	192	15	zhou	zhou	PROPN
ejpam-4568	192	16	.	.	PUNCT
ejpam-4568	193	1	impulse	impulse	ADJ
ejpam-4568	193	2	noise	noise	NOUN
ejpam-4568	193	3	removal	removal	NOUN
ejpam-4568	193	4	by	by	ADP
ejpam-4568	193	5	a	a	DET
ejpam-4568	193	6	nonmonotone	nonmonotone	NOUN
ejpam-4568	193	7	adaptive	adaptive	PROPN
ejpam-4568	193	8	gradient	gradient	NOUN
ejpam-4568	193	9	method	method	NOUN
ejpam-4568	193	10	.	.	PUNCT
ejpam-4568	194	1	signal	signal	NOUN
ejpam-4568	194	2	processing	processing	NOUN
ejpam-4568	194	3	,	,	PUNCT
ejpam-4568	194	4	90(10):2891–2897	90(10):2891–2897	NUM
ejpam-4568	194	5	,	,	PUNCT
ejpam-4568	194	6	2010	2010	NUM
ejpam-4568	194	7	.	.	PUNCT
ejpam-4568	195	1	[	[	X
ejpam-4568	195	2	22	22	NUM
ejpam-4568	195	3	]	]	X
ejpam-4568	195	4	g	g	PROPN
ejpam-4568	195	5	zoutendijk	zoutendijk	PROPN
ejpam-4568	195	6	.	.	PUNCT
ejpam-4568	196	1	nonlinear	nonlinear	ADJ
ejpam-4568	196	2	programming	programming	NOUN
ejpam-4568	196	3	,	,	PUNCT
ejpam-4568	196	4	computational	computational	ADJ
ejpam-4568	196	5	methods	method	NOUN
ejpam-4568	196	6	.	.	PUNCT
ejpam-4568	197	1	integer	integer	NOUN
ejpam-4568	197	2	and	and	CCONJ
ejpam-4568	197	3	nonlinear	nonlinear	ADJ
ejpam-4568	197	4	programming	programming	NOUN
ejpam-4568	197	5	,	,	PUNCT
ejpam-4568	197	6	pages	page	NOUN
ejpam-4568	197	7	37–86	37–86	NUM
ejpam-4568	197	8	,	,	PUNCT
ejpam-4568	197	9	1970	1970	NUM
ejpam-4568	197	10	.	.	PUNCT