id	sid	tid	token	lemma	pos
esrj-117639	1	1	issn	issn	PROPN
esrj-117639	1	2	1794	1794	NUM
esrj-117639	1	3	-	-	SYM
esrj-117639	1	4	6190	6190	NUM
esrj-117639	1	5	e	e	NOUN
esrj-117639	1	6	-	-	NOUN
esrj-117639	1	7	issn	issn	PROPN
esrj-117639	1	8	2339	2339	NUM
esrj-117639	1	9	-	-	SYM
esrj-117639	1	10	3459	3459	NUM
esrj-117639	1	11	https://doi.org/10.15446/esrj.v28n4.115333	https://doi.org/10.15446/esrj.v28n4.115333	AUX
esrj-117639	1	12	earth	earth	PROPN
esrj-117639	1	13	sciences	sciences	PROPN
esrj-117639	1	14	research	research	PROPN
esrj-117639	1	15	journal	journal	PROPN
esrj-117639	1	16	earth	earth	PROPN
esrj-117639	1	17	sci	sci	PROPN
esrj-117639	1	18	.	.	PUNCT
esrj-117639	2	1	res	res	PROPN
esrj-117639	2	2	.	.	PUNCT
esrj-117639	3	1	j.	j.	PROPN
esrj-117639	3	2	vol	vol	PROPN
esrj-117639	3	3	.	.	PROPN
esrj-117639	4	1	28	28	NUM
esrj-117639	4	2	,	,	PUNCT
esrj-117639	4	3	no	no	INTJ
esrj-117639	4	4	.	.	NOUN
esrj-117639	4	5	4	4	NUM
esrj-117639	4	6	(	(	PUNCT
esrj-117639	4	7	december	december	PROPN
esrj-117639	4	8	,	,	PUNCT
esrj-117639	4	9	2024	2024	NUM
esrj-117639	4	10	):	):	PUNCT
esrj-117639	4	11	421	421	NUM
esrj-117639	4	12	426	426	NUM
esrj-117639	4	13	g	g	NOUN
esrj-117639	4	14	eo	eo	PROPN
esrj-117639	4	15	ph	ph	PROPN
esrj-117639	4	16	y	y	PROPN
esrj-117639	4	17	si	si	PROPN
esrj-117639	4	18	c	c	PROPN
esrj-117639	4	19	s	s	PART
esrj-117639	4	20	three	three	NUM
esrj-117639	4	21	-	-	PUNCT
esrj-117639	4	22	dimensional	dimensional	ADJ
esrj-117639	4	23	gravity	gravity	NOUN
esrj-117639	4	24	interface	interface	NOUN
esrj-117639	4	25	inversion	inversion	NOUN
esrj-117639	4	26	based	base	VERB
esrj-117639	4	27	on	on	ADP
esrj-117639	4	28	artificial	artificial	ADJ
esrj-117639	4	29	neural	neural	ADJ
esrj-117639	4	30	network	network	NOUN
esrj-117639	4	31	and	and	CCONJ
esrj-117639	4	32	discrete	discrete	ADJ
esrj-117639	4	33	cosine	cosine	NOUN
esrj-117639	4	34	transform	transform	NOUN
esrj-117639	4	35	algorithm	algorithm	NOUN
esrj-117639	4	36	huadong	huadong	NOUN
esrj-117639	4	37	song	song	NOUN
esrj-117639	4	38	,	,	PUNCT
esrj-117639	4	39	hui	hui	PROPN
esrj-117639	4	40	fu	fu	PROPN
esrj-117639	4	41	*	*	PROPN
esrj-117639	4	42	,	,	PUNCT
esrj-117639	4	43	jifeng	jifeng	PROPN
esrj-117639	4	44	qin	qin	PROPN
esrj-117639	4	45	huanghe	huanghe	PROPN
esrj-117639	4	46	science	science	PROPN
esrj-117639	4	47	and	and	CCONJ
esrj-117639	4	48	technology	technology	NOUN
esrj-117639	4	49	university	university	NOUN
esrj-117639	4	50	,	,	PUNCT
esrj-117639	4	51	zhengzhou	zhengzhou	PROPN
esrj-117639	4	52	450003	450003	NUM
esrj-117639	4	53	,	,	PUNCT
esrj-117639	4	54	china	china	PROPN
esrj-117639	4	55	*	*	PUNCT
esrj-117639	4	56	corresponding	correspond	VERB
esrj-117639	4	57	author	author	NOUN
esrj-117639	4	58	:	:	PUNCT
esrj-117639	4	59	17630076382@hhstu.cn	17630076382@hhstu.cn	NUM
esrj-117639	4	60	how	how	SCONJ
esrj-117639	4	61	to	to	PART
esrj-117639	4	62	cite	cite	VERB
esrj-117639	4	63	item	item	NOUN
esrj-117639	4	64	:	:	PUNCT
esrj-117639	4	65	song	song	PROPN
esrj-117639	4	66	,	,	PUNCT
esrj-117639	4	67	h.	h.	PROPN
esrj-117639	4	68	,	,	PUNCT
esrj-117639	4	69	fu	fu	PROPN
esrj-117639	4	70	,	,	PUNCT
esrj-117639	4	71	h.	h.	PROPN
esrj-117639	4	72	,	,	PUNCT
esrj-117639	4	73	&	&	CCONJ
esrj-117639	4	74	qin	qin	PROPN
esrj-117639	4	75	,	,	PUNCT
esrj-117639	4	76	j.	j.	PROPN
esrj-117639	4	77	(	(	PUNCT
esrj-117639	4	78	2024	2024	NUM
esrj-117639	4	79	)	)	PUNCT
esrj-117639	4	80	.	.	PUNCT
esrj-117639	5	1	three	three	NUM
esrj-117639	5	2	-	-	PUNCT
esrj-117639	5	3	dimensional	dimensional	ADJ
esrj-117639	5	4	gravity	gravity	NOUN
esrj-117639	5	5	interface	interface	NOUN
esrj-117639	5	6	inversion	inversion	NOUN
esrj-117639	5	7	based	base	VERB
esrj-117639	5	8	on	on	ADP
esrj-117639	5	9	artificial	artificial	ADJ
esrj-117639	5	10	neural	neural	ADJ
esrj-117639	5	11	network	network	NOUN
esrj-117639	5	12	and	and	CCONJ
esrj-117639	5	13	discrete	discrete	ADJ
esrj-117639	5	14	cosine	cosine	NOUN
esrj-117639	5	15	transform	transform	NOUN
esrj-117639	5	16	algorithm	algorithm	NOUN
esrj-117639	5	17	.	.	PUNCT
esrj-117639	6	1	earth	earth	PROPN
esrj-117639	6	2	sciences	sciences	PROPN
esrj-117639	6	3	research	research	PROPN
esrj-117639	6	4	journal	journal	PROPN
esrj-117639	6	5	,	,	PUNCT
esrj-117639	6	6	28(4	28(4	NUM
esrj-117639	6	7	)	)	PUNCT
esrj-117639	6	8	,	,	PUNCT
esrj-117639	6	9	421	421	NUM
esrj-117639	6	10	-	-	SYM
esrj-117639	6	11	426	426	NUM
esrj-117639	6	12	.	.	PUNCT
esrj-117639	7	1	https://doi.org/10.15446/esrj.v28n4.115333	https://doi.org/10.15446/esrj.v28n4.115333	AUX
esrj-117639	7	2	record	record	PROPN
esrj-117639	7	3	manuscript	manuscript	NOUN
esrj-117639	7	4	received	receive	VERB
esrj-117639	7	5	:	:	PUNCT
esrj-117639	7	6	19/11/2024	19/11/2024	NUM
esrj-117639	7	7	accepted	accept	VERB
esrj-117639	7	8	for	for	ADP
esrj-117639	7	9	publication	publication	NOUN
esrj-117639	7	10	:	:	PUNCT
esrj-117639	7	11	27/01/2025	27/01/2025	NUM
esrj-117639	7	12	abst	abst	PROPN
esrj-117639	7	13	r	r	PROPN
esrj-117639	7	14	ac	ac	PROPN
esrj-117639	7	15	t	t	PROPN
esrj-117639	7	16	this	this	DET
esrj-117639	7	17	study	study	NOUN
esrj-117639	7	18	applies	apply	VERB
esrj-117639	7	19	artificial	artificial	ADJ
esrj-117639	7	20	neural	neural	ADJ
esrj-117639	7	21	networks	network	NOUN
esrj-117639	7	22	to	to	ADP
esrj-117639	7	23	three	three	NUM
esrj-117639	7	24	-	-	PUNCT
esrj-117639	7	25	dimensional	dimensional	ADJ
esrj-117639	7	26	gravity	gravity	NOUN
esrj-117639	7	27	density	density	NOUN
esrj-117639	7	28	interface	interface	NOUN
esrj-117639	7	29	inversion	inversion	NOUN
esrj-117639	7	30	.	.	PUNCT
esrj-117639	8	1	parker	parker	PROPN
esrj-117639	8	2	’s	’s	PART
esrj-117639	8	3	formula	formula	NOUN
esrj-117639	8	4	based	base	VERB
esrj-117639	8	5	on	on	ADP
esrj-117639	8	6	fast	fast	ADJ
esrj-117639	8	7	fourier	fourier	NOUN
esrj-117639	8	8	transform	transform	NOUN
esrj-117639	8	9	method	method	NOUN
esrj-117639	8	10	plays	play	VERB
esrj-117639	8	11	an	an	DET
esrj-117639	8	12	important	important	ADJ
esrj-117639	8	13	role	role	NOUN
esrj-117639	8	14	in	in	ADP
esrj-117639	8	15	gravity	gravity	NOUN
esrj-117639	8	16	interface	interface	NOUN
esrj-117639	8	17	inversion	inversion	NOUN
esrj-117639	8	18	.	.	PUNCT
esrj-117639	9	1	in	in	ADP
esrj-117639	9	2	the	the	DET
esrj-117639	9	3	training	training	NOUN
esrj-117639	9	4	process	process	NOUN
esrj-117639	9	5	of	of	ADP
esrj-117639	9	6	the	the	DET
esrj-117639	9	7	artificial	artificial	ADJ
esrj-117639	9	8	neural	neural	ADJ
esrj-117639	9	9	network	network	NOUN
esrj-117639	9	10	,	,	PUNCT
esrj-117639	9	11	random	random	ADJ
esrj-117639	9	12	generated	generate	VERB
esrj-117639	9	13	underground	underground	ADJ
esrj-117639	9	14	interface	interface	NOUN
esrj-117639	9	15	geometries	geometry	NOUN
esrj-117639	9	16	are	be	AUX
esrj-117639	9	17	used	use	VERB
esrj-117639	9	18	as	as	ADP
esrj-117639	9	19	inputs	input	NOUN
esrj-117639	9	20	,	,	PUNCT
esrj-117639	9	21	while	while	SCONJ
esrj-117639	9	22	the	the	DET
esrj-117639	9	23	output	output	NOUN
esrj-117639	9	24	is	be	AUX
esrj-117639	9	25	the	the	DET
esrj-117639	9	26	gravity	gravity	NOUN
esrj-117639	9	27	anomaly	anomaly	NOUN
esrj-117639	9	28	data	datum	NOUN
esrj-117639	9	29	calculated	calculate	VERB
esrj-117639	9	30	by	by	ADP
esrj-117639	9	31	parker	parker	PROPN
esrj-117639	9	32	’s	’s	PART
esrj-117639	9	33	formula	formula	NOUN
esrj-117639	9	34	.	.	PUNCT
esrj-117639	10	1	a	a	DET
esrj-117639	10	2	large	large	ADJ
esrj-117639	10	3	-	-	PUNCT
esrj-117639	10	4	scale	scale	NOUN
esrj-117639	10	5	input	input	NOUN
esrj-117639	10	6	-	-	PUNCT
esrj-117639	10	7	output	output	NOUN
esrj-117639	10	8	training	training	NOUN
esrj-117639	10	9	set	set	NOUN
esrj-117639	10	10	is	be	AUX
esrj-117639	10	11	formed	form	VERB
esrj-117639	10	12	for	for	ADP
esrj-117639	10	13	the	the	DET
esrj-117639	10	14	training	training	NOUN
esrj-117639	10	15	process	process	NOUN
esrj-117639	10	16	of	of	ADP
esrj-117639	10	17	the	the	DET
esrj-117639	10	18	artificial	artificial	ADJ
esrj-117639	10	19	neural	neural	ADJ
esrj-117639	10	20	network	network	NOUN
esrj-117639	10	21	.	.	PUNCT
esrj-117639	11	1	in	in	ADP
esrj-117639	11	2	addition	addition	NOUN
esrj-117639	11	3	,	,	PUNCT
esrj-117639	11	4	discrete	discrete	ADJ
esrj-117639	11	5	cosine	cosine	NOUN
esrj-117639	11	6	transform	transform	NOUN
esrj-117639	11	7	(	(	PUNCT
esrj-117639	11	8	dft	dft	NOUN
esrj-117639	11	9	)	)	PUNCT
esrj-117639	11	10	is	be	AUX
esrj-117639	11	11	introduced	introduce	VERB
esrj-117639	11	12	to	to	PART
esrj-117639	11	13	compress	compress	VERB
esrj-117639	11	14	and	and	CCONJ
esrj-117639	11	15	store	store	NOUN
esrj-117639	11	16	matrices	matrix	NOUN
esrj-117639	11	17	,	,	PUNCT
esrj-117639	11	18	which	which	PRON
esrj-117639	11	19	reduces	reduce	VERB
esrj-117639	11	20	computational	computational	ADJ
esrj-117639	11	21	memory	memory	NOUN
esrj-117639	11	22	,	,	PUNCT
esrj-117639	11	23	decreases	decrease	VERB
esrj-117639	11	24	computation	computation	NOUN
esrj-117639	11	25	time	time	NOUN
esrj-117639	11	26	,	,	PUNCT
esrj-117639	11	27	and	and	CCONJ
esrj-117639	11	28	improves	improve	VERB
esrj-117639	11	29	computational	computational	ADJ
esrj-117639	11	30	efficiency	efficiency	NOUN
esrj-117639	11	31	in	in	ADP
esrj-117639	11	32	the	the	DET
esrj-117639	11	33	training	training	NOUN
esrj-117639	11	34	and	and	CCONJ
esrj-117639	11	35	testing	testing	NOUN
esrj-117639	11	36	processes	process	NOUN
esrj-117639	11	37	of	of	ADP
esrj-117639	11	38	the	the	DET
esrj-117639	11	39	artificial	artificial	ADJ
esrj-117639	11	40	neural	neural	ADJ
esrj-117639	11	41	network	network	NOUN
esrj-117639	11	42	.	.	PUNCT
esrj-117639	12	1	a	a	DET
esrj-117639	12	2	deep	deep	ADJ
esrj-117639	12	3	learning	learning	NOUN
esrj-117639	12	4	interface	interface	NOUN
esrj-117639	12	5	inversion	inversion	NOUN
esrj-117639	12	6	algorithm	algorithm	NOUN
esrj-117639	12	7	based	base	VERB
esrj-117639	12	8	on	on	ADP
esrj-117639	12	9	the	the	DET
esrj-117639	12	10	u	u	ADJ
esrj-117639	12	11	-	-	ADJ
esrj-117639	12	12	net	net	ADJ
esrj-117639	12	13	network	network	NOUN
esrj-117639	12	14	model	model	NOUN
esrj-117639	12	15	is	be	AUX
esrj-117639	12	16	designed	design	VERB
esrj-117639	12	17	.	.	PUNCT
esrj-117639	13	1	on	on	ADP
esrj-117639	13	2	the	the	DET
esrj-117639	13	3	basis	basis	NOUN
esrj-117639	13	4	of	of	ADP
esrj-117639	13	5	the	the	DET
esrj-117639	13	6	traditional	traditional	ADJ
esrj-117639	13	7	loss	loss	NOUN
esrj-117639	13	8	function	function	NOUN
esrj-117639	13	9	,	,	PUNCT
esrj-117639	13	10	a	a	DET
esrj-117639	13	11	smooth	smooth	ADJ
esrj-117639	13	12	loss	loss	NOUN
esrj-117639	13	13	term	term	NOUN
esrj-117639	13	14	and	and	CCONJ
esrj-117639	13	15	an	an	DET
esrj-117639	13	16	overfitting	overfitte	VERB
esrj-117639	13	17	suppression	suppression	NOUN
esrj-117639	13	18	term	term	NOUN
esrj-117639	13	19	are	be	AUX
esrj-117639	13	20	added	add	VERB
esrj-117639	13	21	to	to	PART
esrj-117639	13	22	improve	improve	VERB
esrj-117639	13	23	the	the	DET
esrj-117639	13	24	smoothness	smoothness	NOUN
esrj-117639	13	25	and	and	CCONJ
esrj-117639	13	26	convergence	convergence	NOUN
esrj-117639	13	27	efficiency	efficiency	NOUN
esrj-117639	13	28	of	of	ADP
esrj-117639	13	29	the	the	DET
esrj-117639	13	30	gravity	gravity	NOUN
esrj-117639	13	31	interface	interface	NOUN
esrj-117639	13	32	inversion	inversion	NOUN
esrj-117639	13	33	results	result	NOUN
esrj-117639	13	34	.	.	PUNCT
esrj-117639	14	1	finally	finally	ADV
esrj-117639	14	2	,	,	PUNCT
esrj-117639	14	3	the	the	DET
esrj-117639	14	4	inversion	inversion	NOUN
esrj-117639	14	5	prediction	prediction	NOUN
esrj-117639	14	6	is	be	AUX
esrj-117639	14	7	verified	verify	VERB
esrj-117639	14	8	through	through	ADP
esrj-117639	14	9	the	the	DET
esrj-117639	14	10	test	test	NOUN
esrj-117639	14	11	sample	sample	NOUN
esrj-117639	14	12	set	set	VERB
esrj-117639	14	13	to	to	PART
esrj-117639	14	14	validate	validate	VERB
esrj-117639	14	15	the	the	DET
esrj-117639	14	16	generalization	generalization	NOUN
esrj-117639	14	17	of	of	ADP
esrj-117639	14	18	the	the	DET
esrj-117639	14	19	established	establish	VERB
esrj-117639	14	20	deep	deep	ADJ
esrj-117639	14	21	learning	learning	NOUN
esrj-117639	14	22	network	network	NOUN
esrj-117639	14	23	model	model	NOUN
esrj-117639	14	24	.	.	PUNCT
esrj-117639	15	1	this	this	DET
esrj-117639	15	2	paper	paper	NOUN
esrj-117639	15	3	analyzes	analyze	VERB
esrj-117639	15	4	the	the	DET
esrj-117639	15	5	effectiveness	effectiveness	NOUN
esrj-117639	15	6	and	and	CCONJ
esrj-117639	15	7	practicality	practicality	NOUN
esrj-117639	15	8	of	of	ADP
esrj-117639	15	9	this	this	DET
esrj-117639	15	10	method	method	NOUN
esrj-117639	15	11	in	in	ADP
esrj-117639	15	12	density	density	NOUN
esrj-117639	15	13	interface	interface	NOUN
esrj-117639	15	14	inversion	inversion	NOUN
esrj-117639	15	15	through	through	ADP
esrj-117639	15	16	theoretical	theoretical	ADJ
esrj-117639	15	17	models	model	NOUN
esrj-117639	15	18	and	and	CCONJ
esrj-117639	15	19	actual	actual	ADJ
esrj-117639	15	20	data	datum	NOUN
esrj-117639	15	21	experiments	experiment	NOUN
esrj-117639	15	22	.	.	PUNCT
esrj-117639	16	1	the	the	DET
esrj-117639	16	2	deep	deep	ADJ
esrj-117639	16	3	learning	learning	NOUN
esrj-117639	16	4	interface	interface	NOUN
esrj-117639	16	5	inversion	inversion	NOUN
esrj-117639	16	6	method	method	NOUN
esrj-117639	16	7	based	base	VERB
esrj-117639	16	8	on	on	ADP
esrj-117639	16	9	the	the	DET
esrj-117639	16	10	improved	improve	VERB
esrj-117639	16	11	loss	loss	NOUN
esrj-117639	16	12	function	function	NOUN
esrj-117639	16	13	constraint	constraint	NOUN
esrj-117639	16	14	effectively	effectively	ADV
esrj-117639	16	15	improves	improve	VERB
esrj-117639	16	16	the	the	DET
esrj-117639	16	17	convergence	convergence	NOUN
esrj-117639	16	18	efficiency	efficiency	NOUN
esrj-117639	16	19	and	and	CCONJ
esrj-117639	16	20	computational	computational	ADJ
esrj-117639	16	21	stability	stability	NOUN
esrj-117639	16	22	of	of	ADP
esrj-117639	16	23	density	density	NOUN
esrj-117639	16	24	interface	interface	NOUN
esrj-117639	16	25	inversion	inversion	NOUN
esrj-117639	16	26	.	.	PUNCT
esrj-117639	17	1	applying	apply	VERB
esrj-117639	17	2	this	this	DET
esrj-117639	17	3	method	method	NOUN
esrj-117639	17	4	to	to	ADP
esrj-117639	17	5	synthetic	synthetic	ADJ
esrj-117639	17	6	data	datum	NOUN
esrj-117639	17	7	and	and	CCONJ
esrj-117639	17	8	actual	actual	ADJ
esrj-117639	17	9	measured	measured	ADJ
esrj-117639	17	10	data	datum	NOUN
esrj-117639	17	11	processing	processing	NOUN
esrj-117639	17	12	has	have	AUX
esrj-117639	17	13	achieved	achieve	VERB
esrj-117639	17	14	good	good	ADJ
esrj-117639	17	15	results	result	NOUN
esrj-117639	17	16	.	.	PUNCT
esrj-117639	18	1	keywords	keyword	NOUN
esrj-117639	18	2	:	:	PUNCT
esrj-117639	18	3	neural	neural	ADJ
esrj-117639	18	4	network	network	NOUN
esrj-117639	18	5	;	;	PUNCT
esrj-117639	18	6	discrete	discrete	ADJ
esrj-117639	18	7	cosine	cosine	NOUN
esrj-117639	18	8	transform	transform	NOUN
esrj-117639	18	9	;	;	PUNCT
esrj-117639	18	10	application	application	NOUN
esrj-117639	18	11	;	;	PUNCT
esrj-117639	18	12	3d	3d	NUM
esrj-117639	18	13	gravity	gravity	NOUN
esrj-117639	18	14	palabras	palabra	NOUN
esrj-117639	18	15	clave	clave	NOUN
esrj-117639	18	16	:	:	PUNCT
esrj-117639	18	17	red	red	ADJ
esrj-117639	18	18	neuronal	neuronal	NOUN
esrj-117639	18	19	;	;	PUNCT
esrj-117639	18	20	transformada	transformada	X
esrj-117639	18	21	discreta	discreta	PROPN
esrj-117639	18	22	de	de	X
esrj-117639	18	23	coseno	coseno	PROPN
esrj-117639	18	24	;	;	PUNCT
esrj-117639	18	25	aplicación	aplicación	PROPN
esrj-117639	18	26	;	;	PUNCT
esrj-117639	18	27	modelo	modelo	PROPN
esrj-117639	18	28	de	de	PROPN
esrj-117639	18	29	gravedad	gravedad	PROPN
esrj-117639	18	30	3d	3d	PROPN
esrj-117639	18	31	inversión	inversión	PROPN
esrj-117639	18	32	de	de	PROPN
esrj-117639	18	33	la	la	PROPN
esrj-117639	18	34	interfaz	interfaz	X
esrj-117639	18	35	gravitacional	gravitacional	ADJ
esrj-117639	18	36	tridimensional	tridimensional	ADJ
esrj-117639	18	37	con	con	NOUN
esrj-117639	18	38	base	base	NOUN
esrj-117639	18	39	en	en	PROPN
esrj-117639	18	40	los	los	PROPN
esrj-117639	18	41	algoritmos	algoritmos	PROPN
esrj-117639	18	42	de	de	PROPN
esrj-117639	18	43	transformación	transformación	PROPN
esrj-117639	18	44	de	de	PROPN
esrj-117639	18	45	redes	redes	PROPN
esrj-117639	18	46	neuronales	neuronales	PROPN
esrj-117639	18	47	artificiales	artificiales	PROPN
esrj-117639	18	48	y	y	PROPN
esrj-117639	18	49	coseno	coseno	PROPN
esrj-117639	18	50	discreto	discreto	PROPN
esrj-117639	18	51	este	este	PROPN
esrj-117639	18	52	estudio	estudio	PROPN
esrj-117639	18	53	aplica	aplica	PROPN
esrj-117639	18	54	redes	redes	PROPN
esrj-117639	18	55	neuronales	neuronales	PROPN
esrj-117639	18	56	artificiales	artificiale	VERB
esrj-117639	18	57	a	a	DET
esrj-117639	18	58	la	la	X
esrj-117639	18	59	interfaz	interfaz	X
esrj-117639	18	60	de	de	PROPN
esrj-117639	18	61	densidad	densidad	PROPN
esrj-117639	18	62	de	de	PROPN
esrj-117639	18	63	la	la	PROPN
esrj-117639	18	64	inversión	inversión	PROPN
esrj-117639	18	65	de	de	PROPN
esrj-117639	18	66	gravedad	gravedad	PROPN
esrj-117639	18	67	tridimensional	tridimensional	NOUN
esrj-117639	18	68	.	.	PUNCT
esrj-117639	19	1	la	la	PROPN
esrj-117639	19	2	formula	formula	PROPN
esrj-117639	19	3	de	de	PROPN
esrj-117639	19	4	parker	parker	PROPN
esrj-117639	19	5	basada	basada	PROPN
esrj-117639	19	6	en	en	PROPN
esrj-117639	19	7	la	la	PROPN
esrj-117639	19	8	transformada	transformada	PROPN
esrj-117639	19	9	rápida	rápida	PROPN
esrj-117639	19	10	de	de	PROPN
esrj-117639	19	11	fourier	fourier	PROPN
esrj-117639	19	12	juega	juega	PROPN
esrj-117639	19	13	un	un	PROPN
esrj-117639	19	14	papel	papel	PROPN
esrj-117639	19	15	importante	importante	PROPN
esrj-117639	19	16	en	en	X
esrj-117639	19	17	la	la	PROPN
esrj-117639	19	18	interfaz	interfaz	PROPN
esrj-117639	19	19	de	de	PROPN
esrj-117639	19	20	inversión	inversión	PROPN
esrj-117639	19	21	de	de	PROPN
esrj-117639	19	22	la	la	PROPN
esrj-117639	19	23	gravedad	gravedad	PROPN
esrj-117639	19	24	.	.	PUNCT
esrj-117639	20	1	en	en	PROPN
esrj-117639	20	2	el	el	PROPN
esrj-117639	20	3	proceso	proceso	PROPN
esrj-117639	20	4	de	de	PROPN
esrj-117639	20	5	entrenamiento	entrenamiento	PROPN
esrj-117639	20	6	de	de	PROPN
esrj-117639	20	7	las	las	PROPN
esrj-117639	20	8	redes	redes	PROPN
esrj-117639	20	9	neuronales	neuronale	VERB
esrj-117639	20	10	artificiales	artificiale	NOUN
esrj-117639	20	11	,	,	PUNCT
esrj-117639	20	12	las	las	PROPN
esrj-117639	20	13	geometrías	geometrías	PROPN
esrj-117639	20	14	de	de	X
esrj-117639	20	15	interfaz	interfaz	PROPN
esrj-117639	20	16	subterráneas	subterráneas	PROPN
esrj-117639	20	17	generadas	generadas	PROPN
esrj-117639	20	18	aleatoriamente	aleatoriamente	PROPN
esrj-117639	20	19	se	se	PROPN
esrj-117639	20	20	usaron	usaron	PROPN
esrj-117639	20	21	como	como	PROPN
esrj-117639	20	22	registros	registros	PROPN
esrj-117639	20	23	,	,	PUNCT
esrj-117639	20	24	mientras	mientras	X
esrj-117639	20	25	que	que	X
esrj-117639	20	26	los	los	PROPN
esrj-117639	20	27	resultados	resultados	PROPN
esrj-117639	20	28	son	son	NOUN
esrj-117639	20	29	la	la	PROPN
esrj-117639	20	30	información	información	PROPN
esrj-117639	20	31	de	de	X
esrj-117639	20	32	anomalías	anomalías	X
esrj-117639	20	33	de	de	X
esrj-117639	20	34	la	la	PROPN
esrj-117639	20	35	gravedad	gravedad	PROPN
esrj-117639	20	36	calculadas	calculada	VERB
esrj-117639	20	37	con	con	PROPN
esrj-117639	20	38	la	la	PROPN
esrj-117639	20	39	formula	formula	PROPN
esrj-117639	20	40	de	de	PROPN
esrj-117639	20	41	parker	parker	PROPN
esrj-117639	20	42	.	.	PUNCT
esrj-117639	21	1	un	un	PROPN
esrj-117639	21	2	amplio	amplio	PROPN
esrj-117639	21	3	conjunto	conjunto	PROPN
esrj-117639	21	4	de	de	X
esrj-117639	21	5	datos	datos	X
esrj-117639	21	6	de	de	PROPN
esrj-117639	21	7	registro	registro	PROPN
esrj-117639	21	8	y	y	PROPN
esrj-117639	21	9	resultado	resultado	PROPN
esrj-117639	21	10	se	se	PROPN
esrj-117639	21	11	definieron	definieron	PROPN
esrj-117639	21	12	para	para	PROPN
esrj-117639	21	13	el	el	PROPN
esrj-117639	21	14	proceso	proceso	PROPN
esrj-117639	21	15	de	de	PROPN
esrj-117639	21	16	entrenamiento	entrenamiento	PROPN
esrj-117639	21	17	de	de	PROPN
esrj-117639	21	18	la	la	X
esrj-117639	21	19	red	red	ADJ
esrj-117639	21	20	neuronal	neuronal	ADJ
esrj-117639	21	21	artificial	artificial	NOUN
esrj-117639	21	22	.	.	PUNCT
esrj-117639	22	1	adicionalmente	adicionalmente	NOUN
esrj-117639	22	2	,	,	PUNCT
esrj-117639	22	3	la	la	PROPN
esrj-117639	22	4	transformada	transformada	PROPN
esrj-117639	22	5	de	de	PROPN
esrj-117639	22	6	coseno	coseno	PROPN
esrj-117639	22	7	discreta	discreta	PROPN
esrj-117639	22	8	se	se	PROPN
esrj-117639	22	9	presenta	presenta	PROPN
esrj-117639	22	10	para	para	PROPN
esrj-117639	22	11	comprimir	comprimir	PROPN
esrj-117639	22	12	y	y	PROPN
esrj-117639	22	13	almacenar	almacenar	ADJ
esrj-117639	22	14	matrices	matrix	NOUN
esrj-117639	22	15	,	,	PUNCT
esrj-117639	22	16	y	y	PROPN
esrj-117639	22	17	así	así	PROPN
esrj-117639	22	18	se	se	PROPN
esrj-117639	22	19	reduce	reduce	VERB
esrj-117639	22	20	la	la	PROPN
esrj-117639	22	21	memoria	memoria	PROPN
esrj-117639	22	22	computacional	computacional	PROPN
esrj-117639	22	23	,	,	PUNCT
esrj-117639	22	24	se	se	X
esrj-117639	22	25	reduce	reduce	VERB
esrj-117639	22	26	el	el	PROPN
esrj-117639	22	27	tiempo	tiempo	PROPN
esrj-117639	22	28	de	de	X
esrj-117639	22	29	computación	computación	PROPN
esrj-117639	22	30	y	y	PROPN
esrj-117639	22	31	se	se	PROPN
esrj-117639	22	32	mejora	mejora	X
esrj-117639	22	33	la	la	X
esrj-117639	22	34	eficiencia	eficiencia	PROPN
esrj-117639	22	35	del	del	PROPN
esrj-117639	22	36	equipo	equipo	PROPN
esrj-117639	22	37	en	en	PROPN
esrj-117639	22	38	los	los	PROPN
esrj-117639	22	39	procesos	procesos	PROPN
esrj-117639	22	40	de	de	PROPN
esrj-117639	22	41	entrenamiento	entrenamiento	PROPN
esrj-117639	22	42	y	y	PROPN
esrj-117639	22	43	de	de	PROPN
esrj-117639	22	44	prueba	prueba	X
esrj-117639	22	45	de	de	X
esrj-117639	22	46	la	la	X
esrj-117639	22	47	red	red	ADJ
esrj-117639	22	48	neuronal	neuronal	ADJ
esrj-117639	22	49	artificial	artificial	NOUN
esrj-117639	22	50	.	.	PUNCT
esrj-117639	23	1	de	de	X
esrj-117639	23	2	esta	esta	PROPN
esrj-117639	23	3	forma	forma	PROPN
esrj-117639	23	4	se	se	PROPN
esrj-117639	23	5	diseñó	diseñó	PROPN
esrj-117639	23	6	un	un	PROPN
esrj-117639	23	7	algoritmo	algoritmo	PROPN
esrj-117639	23	8	de	de	PROPN
esrj-117639	23	9	aprendizaje	aprendizaje	PROPN
esrj-117639	23	10	profundo	profundo	PROPN
esrj-117639	23	11	para	para	PROPN
esrj-117639	23	12	la	la	PROPN
esrj-117639	23	13	interfaz	interfaz	PROPN
esrj-117639	23	14	de	de	PROPN
esrj-117639	23	15	inversión	inversión	PROPN
esrj-117639	23	16	basado	basado	PROPN
esrj-117639	23	17	en	en	PROPN
esrj-117639	23	18	el	el	PROPN
esrj-117639	23	19	modelo	modelo	PROPN
esrj-117639	23	20	de	de	PROPN
esrj-117639	23	21	red	red	PROPN
esrj-117639	23	22	tipo	tipo	PROPN
esrj-117639	23	23	u.	u.	PROPN
esrj-117639	23	24	con	con	PROPN
esrj-117639	23	25	base	base	PROPN
esrj-117639	23	26	en	en	X
esrj-117639	23	27	la	la	PROPN
esrj-117639	23	28	tradicional	tradicional	PROPN
esrj-117639	23	29	función	función	PROPN
esrj-117639	23	30	de	de	PROPN
esrj-117639	23	31	pérdida	pérdida	PROPN
esrj-117639	23	32	se	se	PROPN
esrj-117639	23	33	adicionaron	adicionaron	PROPN
esrj-117639	23	34	un	un	PROPN
esrj-117639	23	35	término	término	PROPN
esrj-117639	23	36	de	de	PROPN
esrj-117639	23	37	pérdida	pérdida	PROPN
esrj-117639	23	38	suave	suave	PROPN
esrj-117639	23	39	y	y	PROPN
esrj-117639	23	40	un	un	PROPN
esrj-117639	23	41	término	término	PROPN
esrj-117639	23	42	de	de	PROPN
esrj-117639	23	43	sobreajuste	sobreajuste	PROPN
esrj-117639	23	44	de	de	X
esrj-117639	23	45	supresión	supresión	PROPN
esrj-117639	23	46	para	para	PROPN
esrj-117639	23	47	mejorar	mejorar	NOUN
esrj-117639	23	48	la	la	X
esrj-117639	23	49	eficiencia	eficiencia	X
esrj-117639	23	50	de	de	X
esrj-117639	23	51	la	la	PROPN
esrj-117639	23	52	uniformidad	uniformidad	PROPN
esrj-117639	23	53	y	y	PROPN
esrj-117639	23	54	de	de	PROPN
esrj-117639	23	55	la	la	X
esrj-117639	23	56	convergencia	convergencia	PROPN
esrj-117639	23	57	de	de	X
esrj-117639	23	58	los	los	PROPN
esrj-117639	23	59	resultados	resultados	X
esrj-117639	23	60	de	de	X
esrj-117639	23	61	la	la	X
esrj-117639	23	62	interfaz	interfaz	X
esrj-117639	23	63	de	de	PROPN
esrj-117639	23	64	inversión	inversión	PROPN
esrj-117639	23	65	de	de	PROPN
esrj-117639	23	66	la	la	PROPN
esrj-117639	23	67	gravedad	gravedad	PROPN
esrj-117639	23	68	.	.	PUNCT
esrj-117639	24	1	finalmente	finalmente	PROPN
esrj-117639	24	2	la	la	PROPN
esrj-117639	24	3	predicción	predicción	PROPN
esrj-117639	24	4	de	de	PROPN
esrj-117639	24	5	la	la	PROPN
esrj-117639	24	6	inversión	inversión	PROPN
esrj-117639	24	7	se	se	PROPN
esrj-117639	24	8	verificó	verificó	VERB
esrj-117639	24	9	a	a	DET
esrj-117639	24	10	través	través	NOUN
esrj-117639	24	11	del	del	X
esrj-117639	24	12	ejemplo	ejemplo	PROPN
esrj-117639	24	13	de	de	PROPN
esrj-117639	24	14	prueba	prueba	PROPN
esrj-117639	24	15	para	para	PROPN
esrj-117639	24	16	validar	validar	PROPN
esrj-117639	24	17	la	la	PROPN
esrj-117639	24	18	generalización	generalización	PROPN
esrj-117639	24	19	del	del	PROPN
esrj-117639	24	20	modelo	modelo	PROPN
esrj-117639	24	21	de	de	PROPN
esrj-117639	24	22	red	red	PROPN
esrj-117639	24	23	de	de	PROPN
esrj-117639	24	24	aprendizaje	aprendizaje	PROPN
esrj-117639	24	25	profundo	profundo	PROPN
esrj-117639	24	26	que	que	PROPN
esrj-117639	24	27	se	se	PROPN
esrj-117639	24	28	estableció	estableció	PROPN
esrj-117639	24	29	.	.	PUNCT
esrj-117639	25	1	este	este	PROPN
esrj-117639	25	2	artículo	artículo	PROPN
esrj-117639	25	3	analiza	analiza	PROPN
esrj-117639	25	4	la	la	PROPN
esrj-117639	25	5	efectividad	efectividad	PROPN
esrj-117639	25	6	y	y	PROPN
esrj-117639	25	7	la	la	PROPN
esrj-117639	25	8	practicidad	practicidad	PROPN
esrj-117639	25	9	de	de	X
esrj-117639	25	10	este	este	X
esrj-117639	25	11	método	método	X
esrj-117639	25	12	de	de	PROPN
esrj-117639	25	13	inversión	inversión	PROPN
esrj-117639	25	14	de	de	PROPN
esrj-117639	25	15	la	la	X
esrj-117639	25	16	interfaz	interfaz	PROPN
esrj-117639	25	17	de	de	PROPN
esrj-117639	25	18	densidad	densidad	PROPN
esrj-117639	25	19	a	a	DET
esrj-117639	25	20	través	través	X
esrj-117639	25	21	de	de	X
esrj-117639	25	22	modelos	modelos	PROPN
esrj-117639	25	23	teóricos	teóricos	PROPN
esrj-117639	25	24	y	y	PROPN
esrj-117639	25	25	experimentos	experimentos	PROPN
esrj-117639	25	26	con	con	PROPN
esrj-117639	25	27	información	información	PROPN
esrj-117639	25	28	real	real	PROPN
esrj-117639	25	29	.	.	PUNCT
esrj-117639	26	1	este	este	X
esrj-117639	26	2	método	método	X
esrj-117639	26	3	de	de	PROPN
esrj-117639	26	4	inversión	inversión	PROPN
esrj-117639	26	5	de	de	X
esrj-117639	26	6	interfaz	interfaz	PROPN
esrj-117639	26	7	de	de	PROPN
esrj-117639	26	8	aprendizaje	aprendizaje	PROPN
esrj-117639	26	9	profundo	profundo	PROPN
esrj-117639	26	10	,	,	PUNCT
esrj-117639	26	11	basada	basada	PROPN
esrj-117639	26	12	en	en	X
esrj-117639	26	13	la	la	PROPN
esrj-117639	26	14	restricción	restricción	PROPN
esrj-117639	26	15	de	de	PROPN
esrj-117639	26	16	función	función	PROPN
esrj-117639	26	17	de	de	PROPN
esrj-117639	26	18	perdida	perdida	PROPN
esrj-117639	26	19	mejorada	mejorada	PROPN
esrj-117639	26	20	,	,	PUNCT
esrj-117639	26	21	efectivamente	efectivamente	PROPN
esrj-117639	26	22	mejora	mejora	PROPN
esrj-117639	26	23	la	la	X
esrj-117639	26	24	convergencia	convergencia	PROPN
esrj-117639	26	25	de	de	ADP
esrj-117639	26	26	eficiencia	eficiencia	PROPN
esrj-117639	26	27	y	y	PROPN
esrj-117639	26	28	la	la	PROPN
esrj-117639	26	29	estabilidad	estabilidad	PROPN
esrj-117639	26	30	computacional	computacional	PROPN
esrj-117639	26	31	de	de	PROPN
esrj-117639	26	32	la	la	PROPN
esrj-117639	26	33	inversión	inversión	PROPN
esrj-117639	26	34	de	de	PROPN
esrj-117639	26	35	la	la	X
esrj-117639	26	36	interfaz	interfaz	X
esrj-117639	26	37	de	de	PROPN
esrj-117639	26	38	densidad	densidad	PROPN
esrj-117639	26	39	.	.	PUNCT
esrj-117639	27	1	la	la	PROPN
esrj-117639	27	2	aplicación	aplicación	PROPN
esrj-117639	27	3	de	de	X
esrj-117639	27	4	este	este	X
esrj-117639	27	5	método	método	X
esrj-117639	27	6	a	a	DET
esrj-117639	27	7	procesos	procesos	PROPN
esrj-117639	27	8	de	de	PROPN
esrj-117639	27	9	información	información	PROPN
esrj-117639	27	10	sintética	sintética	PROPN
esrj-117639	27	11	y	y	PROPN
esrj-117639	27	12	de	de	PROPN
esrj-117639	27	13	información	información	PROPN
esrj-117639	27	14	medida	medida	PROPN
esrj-117639	27	15	ha	ha	INTJ
esrj-117639	27	16	alcanzado	alcanzado	PROPN
esrj-117639	27	17	buenos	buenos	PROPN
esrj-117639	27	18	resultados	resultado	NOUN
esrj-117639	27	19	.	.	PUNCT
esrj-117639	28	1	resu	resu	PROPN
esrj-117639	28	2	m	m	VERB
esrj-117639	28	3	en	en	X
esrj-117639	28	4	https://doi.org/10.15446/esrj.v28n4.115333	https://doi.org/10.15446/esrj.v28n4.115333	X
esrj-117639	28	5	mailto:17630076382@hhstu.cn	mailto:17630076382@hhstu.cn	PROPN
esrj-117639	28	6	https://doi.org/10.15446/esrj.v28n4.115333	https://doi.org/10.15446/esrj.v28n4.115333	NUM
esrj-117639	28	7	422	422	NUM
esrj-117639	28	8	huadong	huadong	ADJ
esrj-117639	28	9	song	song	NOUN
esrj-117639	28	10	,	,	PUNCT
esrj-117639	28	11	hui	hui	PROPN
esrj-117639	28	12	fu	fu	PROPN
esrj-117639	28	13	,	,	PUNCT
esrj-117639	28	14	jifeng	jifeng	PROPN
esrj-117639	28	15	qin	qin	PROPN
esrj-117639	28	16	1	1	X
esrj-117639	28	17	.	.	PUNCT
esrj-117639	29	1	introduction	introduction	NOUN
esrj-117639	29	2	a	a	DET
esrj-117639	29	3	classic	classic	ADJ
esrj-117639	29	4	problem	problem	NOUN
esrj-117639	29	5	in	in	ADP
esrj-117639	29	6	geophysics	geophysic	NOUN
esrj-117639	29	7	is	be	AUX
esrj-117639	29	8	the	the	DET
esrj-117639	29	9	determination	determination	NOUN
esrj-117639	29	10	of	of	ADP
esrj-117639	29	11	the	the	DET
esrj-117639	29	12	geometric	geometric	ADJ
esrj-117639	29	13	shape	shape	NOUN
esrj-117639	29	14	of	of	ADP
esrj-117639	29	15	density	density	NOUN
esrj-117639	29	16	interfaces	interface	NOUN
esrj-117639	29	17	related	relate	VERB
esrj-117639	29	18	to	to	ADP
esrj-117639	29	19	gravity	gravity	NOUN
esrj-117639	29	20	anomalies	anomaly	NOUN
esrj-117639	29	21	,	,	PUNCT
esrj-117639	29	22	such	such	ADJ
esrj-117639	29	23	as	as	ADP
esrj-117639	29	24	the	the	DET
esrj-117639	29	25	basement	basement	NOUN
esrj-117639	29	26	shape	shape	NOUN
esrj-117639	29	27	of	of	ADP
esrj-117639	29	28	sedimentary	sedimentary	ADJ
esrj-117639	29	29	basins	basin	NOUN
esrj-117639	29	30	,	,	PUNCT
esrj-117639	29	31	the	the	DET
esrj-117639	29	32	moho	moho	NOUN
esrj-117639	29	33	interface	interface	NOUN
esrj-117639	29	34	,	,	PUNCT
esrj-117639	29	35	and	and	CCONJ
esrj-117639	29	36	seafloor	seafloor	NOUN
esrj-117639	29	37	topography	topography	NOUN
esrj-117639	29	38	.	.	PUNCT
esrj-117639	30	1	these	these	DET
esrj-117639	30	2	studies	study	NOUN
esrj-117639	30	3	provide	provide	VERB
esrj-117639	30	4	valuable	valuable	ADJ
esrj-117639	30	5	information	information	NOUN
esrj-117639	30	6	for	for	ADP
esrj-117639	30	7	understanding	understand	VERB
esrj-117639	30	8	deep	deep	ADJ
esrj-117639	30	9	-	-	PUNCT
esrj-117639	30	10	sea	sea	NOUN
esrj-117639	30	11	geological	geological	ADJ
esrj-117639	30	12	structures	structure	NOUN
esrj-117639	30	13	,	,	PUNCT
esrj-117639	30	14	regional	regional	ADJ
esrj-117639	30	15	tectonics	tectonic	NOUN
esrj-117639	30	16	,	,	PUNCT
esrj-117639	30	17	basin	basin	NOUN
esrj-117639	30	18	formation	formation	NOUN
esrj-117639	30	19	,	,	PUNCT
esrj-117639	30	20	seafloor	seafloor	NOUN
esrj-117639	30	21	topography	topography	NOUN
esrj-117639	30	22	formation	formation	NOUN
esrj-117639	30	23	processes	process	NOUN
esrj-117639	30	24	,	,	PUNCT
esrj-117639	30	25	and	and	CCONJ
esrj-117639	30	26	many	many	ADJ
esrj-117639	30	27	other	other	ADJ
esrj-117639	30	28	geological	geological	ADJ
esrj-117639	30	29	and	and	CCONJ
esrj-117639	30	30	geodynamic	geodynamic	ADJ
esrj-117639	30	31	phenomena	phenomenon	NOUN
esrj-117639	30	32	.	.	PUNCT
esrj-117639	31	1	however	however	ADV
esrj-117639	31	2	,	,	PUNCT
esrj-117639	31	3	estimating	estimate	VERB
esrj-117639	31	4	the	the	DET
esrj-117639	31	5	geometric	geometric	ADJ
esrj-117639	31	6	shape	shape	NOUN
esrj-117639	31	7	of	of	ADP
esrj-117639	31	8	a	a	DET
esrj-117639	31	9	specific	specific	ADJ
esrj-117639	31	10	geological	geological	ADJ
esrj-117639	31	11	structure	structure	NOUN
esrj-117639	31	12	’s	’s	PART
esrj-117639	31	13	interface	interface	NOUN
esrj-117639	31	14	using	use	VERB
esrj-117639	31	15	gravity	gravity	NOUN
esrj-117639	31	16	data	datum	NOUN
esrj-117639	31	17	is	be	AUX
esrj-117639	31	18	an	an	DET
esrj-117639	31	19	ill	ill	ADV
esrj-117639	31	20	-	-	PUNCT
esrj-117639	31	21	posed	pose	VERB
esrj-117639	31	22	and	and	CCONJ
esrj-117639	31	23	non	non	ADJ
esrj-117639	31	24	-	-	ADJ
esrj-117639	31	25	unique	unique	ADJ
esrj-117639	31	26	problem	problem	NOUN
esrj-117639	31	27	.	.	PUNCT
esrj-117639	32	1	it	it	PRON
esrj-117639	32	2	requires	require	VERB
esrj-117639	32	3	careful	careful	ADJ
esrj-117639	32	4	consideration	consideration	NOUN
esrj-117639	32	5	of	of	ADP
esrj-117639	32	6	the	the	DET
esrj-117639	32	7	selection	selection	NOUN
esrj-117639	32	8	of	of	ADP
esrj-117639	32	9	input	input	NOUN
esrj-117639	32	10	gravity	gravity	NOUN
esrj-117639	32	11	data	datum	NOUN
esrj-117639	32	12	,	,	PUNCT
esrj-117639	32	13	pre	pre	ADJ
esrj-117639	32	14	-	-	ADJ
esrj-117639	32	15	processing	processing	ADJ
esrj-117639	32	16	,	,	PUNCT
esrj-117639	32	17	inversion	inversion	NOUN
esrj-117639	32	18	parameters	parameter	NOUN
esrj-117639	32	19	(	(	PUNCT
esrj-117639	32	20	such	such	ADJ
esrj-117639	32	21	as	as	ADP
esrj-117639	32	22	average	average	ADJ
esrj-117639	32	23	depth	depth	NOUN
esrj-117639	32	24	and	and	CCONJ
esrj-117639	32	25	interface	interface	NOUN
esrj-117639	32	26	density	density	NOUN
esrj-117639	32	27	)	)	PUNCT
esrj-117639	32	28	,	,	PUNCT
esrj-117639	32	29	and	and	CCONJ
esrj-117639	32	30	other	other	ADJ
esrj-117639	32	31	geophysical	geophysical	ADJ
esrj-117639	32	32	constraints	constraint	NOUN
esrj-117639	32	33	to	to	PART
esrj-117639	32	34	stabilize	stabilize	VERB
esrj-117639	32	35	gravity	gravity	NOUN
esrj-117639	32	36	inversion	inversion	NOUN
esrj-117639	32	37	.	.	PUNCT
esrj-117639	33	1	the	the	DET
esrj-117639	33	2	bott	bott	PROPN
esrj-117639	33	3	method	method	NOUN
esrj-117639	33	4	(	(	PUNCT
esrj-117639	33	5	bott	bott	PROPN
esrj-117639	33	6	,	,	PUNCT
esrj-117639	33	7	1960	1960	NUM
esrj-117639	33	8	)	)	PUNCT
esrj-117639	33	9	was	be	AUX
esrj-117639	33	10	initially	initially	ADV
esrj-117639	33	11	used	use	VERB
esrj-117639	33	12	to	to	PART
esrj-117639	33	13	solve	solve	VERB
esrj-117639	33	14	this	this	DET
esrj-117639	33	15	problem	problem	NOUN
esrj-117639	33	16	.	.	PUNCT
esrj-117639	34	1	this	this	DET
esrj-117639	34	2	method	method	NOUN
esrj-117639	34	3	is	be	AUX
esrj-117639	34	4	based	base	VERB
esrj-117639	34	5	on	on	ADP
esrj-117639	34	6	iterative	iterative	ADJ
esrj-117639	34	7	solutions	solution	NOUN
esrj-117639	34	8	and	and	CCONJ
esrj-117639	34	9	assumes	assume	VERB
esrj-117639	34	10	constant	constant	ADJ
esrj-117639	34	11	density	density	NOUN
esrj-117639	34	12	differences	difference	NOUN
esrj-117639	34	13	.	.	PUNCT
esrj-117639	35	1	the	the	DET
esrj-117639	35	2	initial	initial	ADJ
esrj-117639	35	3	approximate	approximate	ADJ
esrj-117639	35	4	value	value	NOUN
esrj-117639	35	5	of	of	ADP
esrj-117639	35	6	sediment	sediment	NOUN
esrj-117639	35	7	thickness	thickness	NOUN
esrj-117639	35	8	corresponding	correspond	VERB
esrj-117639	35	9	to	to	ADP
esrj-117639	35	10	each	each	DET
esrj-117639	35	11	gravity	gravity	NOUN
esrj-117639	35	12	reference	reference	NOUN
esrj-117639	35	13	position	position	NOUN
esrj-117639	35	14	is	be	AUX
esrj-117639	35	15	obtained	obtain	VERB
esrj-117639	35	16	using	use	VERB
esrj-117639	35	17	the	the	DET
esrj-117639	35	18	flat	flat	ADJ
esrj-117639	35	19	plate	plate	NOUN
esrj-117639	35	20	formula	formula	NOUN
esrj-117639	35	21	.	.	PUNCT
esrj-117639	36	1	the	the	DET
esrj-117639	36	2	fitting	fitting	ADJ
esrj-117639	36	3	difference	difference	NOUN
esrj-117639	36	4	between	between	ADP
esrj-117639	36	5	the	the	DET
esrj-117639	36	6	observed	observe	VERB
esrj-117639	36	7	data	datum	NOUN
esrj-117639	36	8	and	and	CCONJ
esrj-117639	36	9	the	the	DET
esrj-117639	36	10	predicted	predict	VERB
esrj-117639	36	11	data	data	NOUN
esrj-117639	36	12	is	be	AUX
esrj-117639	36	13	then	then	ADV
esrj-117639	36	14	converted	convert	VERB
esrj-117639	36	15	into	into	ADP
esrj-117639	36	16	a	a	DET
esrj-117639	36	17	correction	correction	NOUN
esrj-117639	36	18	to	to	ADP
esrj-117639	36	19	the	the	DET
esrj-117639	36	20	initial	initial	ADJ
esrj-117639	36	21	model	model	NOUN
esrj-117639	36	22	(	(	PUNCT
esrj-117639	36	23	using	use	VERB
esrj-117639	36	24	the	the	DET
esrj-117639	36	25	flat	flat	ADJ
esrj-117639	36	26	plate	plate	NOUN
esrj-117639	36	27	formula	formula	NOUN
esrj-117639	36	28	again	again	ADV
esrj-117639	36	29	)	)	PUNCT
esrj-117639	36	30	,	,	PUNCT
esrj-117639	36	31	and	and	CCONJ
esrj-117639	36	32	the	the	DET
esrj-117639	36	33	iteration	iteration	NOUN
esrj-117639	36	34	process	process	NOUN
esrj-117639	36	35	stops	stop	VERB
esrj-117639	36	36	when	when	SCONJ
esrj-117639	36	37	the	the	DET
esrj-117639	36	38	fitting	fitting	ADJ
esrj-117639	36	39	difference	difference	NOUN
esrj-117639	36	40	drops	drop	VERB
esrj-117639	36	41	below	below	ADP
esrj-117639	36	42	a	a	DET
esrj-117639	36	43	certain	certain	ADJ
esrj-117639	36	44	threshold	threshold	NOUN
esrj-117639	36	45	.	.	PUNCT
esrj-117639	37	1	cordell	cordell	PROPN
esrj-117639	37	2	and	and	CCONJ
esrj-117639	37	3	henderson	henderson	PROPN
esrj-117639	37	4	(	(	PUNCT
esrj-117639	37	5	1968	1968	NUM
esrj-117639	37	6	)	)	PUNCT
esrj-117639	37	7	modified	modify	VERB
esrj-117639	37	8	the	the	DET
esrj-117639	37	9	iteration	iteration	NOUN
esrj-117639	37	10	step	step	NOUN
esrj-117639	37	11	length	length	NOUN
esrj-117639	37	12	in	in	ADP
esrj-117639	37	13	the	the	DET
esrj-117639	37	14	bott	bott	PROPN
esrj-117639	37	15	method	method	NOUN
esrj-117639	37	16	and	and	CCONJ
esrj-117639	37	17	introduced	introduce	VERB
esrj-117639	37	18	the	the	DET
esrj-117639	37	19	ratio	ratio	NOUN
esrj-117639	37	20	between	between	ADP
esrj-117639	37	21	the	the	DET
esrj-117639	37	22	previously	previously	ADV
esrj-117639	37	23	iterated	iterate	VERB
esrj-117639	37	24	model	model	NOUN
esrj-117639	37	25	-	-	PUNCT
esrj-117639	37	26	generated	generate	VERB
esrj-117639	37	27	observed	observed	ADJ
esrj-117639	37	28	data	datum	NOUN
esrj-117639	37	29	and	and	CCONJ
esrj-117639	37	30	calculated	calculate	VERB
esrj-117639	37	31	data	datum	NOUN
esrj-117639	37	32	to	to	PART
esrj-117639	37	33	update	update	VERB
esrj-117639	37	34	the	the	DET
esrj-117639	37	35	model	model	NOUN
esrj-117639	37	36	.	.	PUNCT
esrj-117639	38	1	silva	silva	PROPN
esrj-117639	38	2	et	et	PROPN
esrj-117639	38	3	al	al	PROPN
esrj-117639	38	4	.	.	PROPN
esrj-117639	39	1	(	(	PUNCT
esrj-117639	39	2	2014	2014	NUM
esrj-117639	39	3	)	)	PUNCT
esrj-117639	39	4	suggested	suggest	VERB
esrj-117639	39	5	accelerating	accelerate	VERB
esrj-117639	39	6	convergence	convergence	NOUN
esrj-117639	39	7	by	by	ADP
esrj-117639	39	8	dividing	divide	VERB
esrj-117639	39	9	the	the	DET
esrj-117639	39	10	residual	residual	ADJ
esrj-117639	39	11	(	(	PUNCT
esrj-117639	39	12	i.e.	i.e.	X
esrj-117639	39	13	,	,	PUNCT
esrj-117639	39	14	the	the	DET
esrj-117639	39	15	difference	difference	NOUN
esrj-117639	39	16	between	between	ADP
esrj-117639	39	17	the	the	DET
esrj-117639	39	18	observed	observe	VERB
esrj-117639	39	19	gravity	gravity	NOUN
esrj-117639	39	20	data	datum	NOUN
esrj-117639	39	21	and	and	CCONJ
esrj-117639	39	22	the	the	DET
esrj-117639	39	23	currently	currently	ADV
esrj-117639	39	24	calculated	calculate	VERB
esrj-117639	39	25	gravity	gravity	NOUN
esrj-117639	39	26	data	datum	NOUN
esrj-117639	39	27	)	)	PUNCT
esrj-117639	39	28	by	by	ADP
esrj-117639	39	29	an	an	DET
esrj-117639	39	30	arbitrarily	arbitrarily	ADV
esrj-117639	39	31	larger	large	ADJ
esrj-117639	39	32	positive	positive	ADJ
esrj-117639	39	33	value	value	NOUN
esrj-117639	39	34	(	(	PUNCT
esrj-117639	39	35	no	no	ADV
esrj-117639	39	36	longer	long	ADV
esrj-117639	39	37	using	use	VERB
esrj-117639	39	38	the	the	DET
esrj-117639	39	39	slab	slab	NOUN
esrj-117639	39	40	formula	formula	NOUN
esrj-117639	39	41	)	)	PUNCT
esrj-117639	39	42	,	,	PUNCT
esrj-117639	39	43	and	and	CCONJ
esrj-117639	39	44	accepting	accept	VERB
esrj-117639	39	45	the	the	DET
esrj-117639	39	46	obtained	obtain	VERB
esrj-117639	39	47	model	model	NOUN
esrj-117639	39	48	when	when	SCONJ
esrj-117639	39	49	the	the	DET
esrj-117639	39	50	l2	l2	NOUN
esrj-117639	39	51	norm	norm	NOUN
esrj-117639	39	52	of	of	ADP
esrj-117639	39	53	the	the	DET
esrj-117639	39	54	residual	residual	ADJ
esrj-117639	39	55	vector	vector	NOUN
esrj-117639	39	56	is	be	AUX
esrj-117639	39	57	less	less	ADJ
esrj-117639	39	58	than	than	ADP
esrj-117639	39	59	the	the	DET
esrj-117639	39	60	previous	previous	ADJ
esrj-117639	39	61	iteration	iteration	NOUN
esrj-117639	39	62	’s	’s	PART
esrj-117639	39	63	norm	norm	NOUN
esrj-117639	39	64	.	.	PUNCT
esrj-117639	40	1	many	many	ADJ
esrj-117639	40	2	scholars	scholar	NOUN
esrj-117639	40	3	have	have	AUX
esrj-117639	40	4	adopted	adopt	VERB
esrj-117639	40	5	similar	similar	ADJ
esrj-117639	40	6	methods	method	NOUN
esrj-117639	40	7	,	,	PUNCT
esrj-117639	40	8	which	which	PRON
esrj-117639	40	9	will	will	AUX
esrj-117639	40	10	not	not	PART
esrj-117639	40	11	be	be	AUX
esrj-117639	40	12	described	describe	VERB
esrj-117639	40	13	in	in	ADP
esrj-117639	40	14	detail	detail	NOUN
esrj-117639	40	15	here	here	ADV
esrj-117639	40	16	.	.	PUNCT
esrj-117639	41	1	currently	currently	ADV
esrj-117639	41	2	,	,	PUNCT
esrj-117639	41	3	gravity	gravity	NOUN
esrj-117639	41	4	inversions	inversion	NOUN
esrj-117639	41	5	are	be	AUX
esrj-117639	41	6	primarily	primarily	ADV
esrj-117639	41	7	based	base	VERB
esrj-117639	41	8	on	on	ADP
esrj-117639	41	9	structured	structured	ADJ
esrj-117639	41	10	grids	grid	NOUN
esrj-117639	41	11	,	,	PUNCT
esrj-117639	41	12	and	and	CCONJ
esrj-117639	41	13	research	research	NOUN
esrj-117639	41	14	on	on	ADP
esrj-117639	41	15	gravity	gravity	NOUN
esrj-117639	41	16	inversions	inversion	NOUN
esrj-117639	41	17	based	base	VERB
esrj-117639	41	18	on	on	ADP
esrj-117639	41	19	unstructured	unstructured	ADJ
esrj-117639	41	20	grids	grid	NOUN
esrj-117639	41	21	has	have	AUX
esrj-117639	41	22	only	only	ADV
esrj-117639	41	23	emerged	emerge	VERB
esrj-117639	41	24	as	as	ADP
esrj-117639	41	25	a	a	DET
esrj-117639	41	26	field	field	NOUN
esrj-117639	41	27	of	of	ADP
esrj-117639	41	28	study	study	NOUN
esrj-117639	41	29	in	in	ADP
esrj-117639	41	30	the	the	DET
esrj-117639	41	31	last	last	ADJ
esrj-117639	41	32	decade	decade	NOUN
esrj-117639	41	33	or	or	CCONJ
esrj-117639	41	34	so	so	ADV
esrj-117639	41	35	,	,	PUNCT
esrj-117639	41	36	with	with	ADP
esrj-117639	41	37	relatively	relatively	ADV
esrj-117639	41	38	limited	limited	ADJ
esrj-117639	41	39	literature	literature	NOUN
esrj-117639	41	40	available	available	ADJ
esrj-117639	41	41	for	for	ADP
esrj-117639	41	42	reference	reference	NOUN
esrj-117639	41	43	.	.	PUNCT
esrj-117639	42	1	among	among	ADP
esrj-117639	42	2	them	they	PRON
esrj-117639	42	3	,	,	PUNCT
esrj-117639	42	4	the	the	DET
esrj-117639	42	5	research	research	NOUN
esrj-117639	42	6	achievements	achievement	NOUN
esrj-117639	42	7	of	of	ADP
esrj-117639	42	8	memorial	memorial	ADJ
esrj-117639	42	9	university	university	NOUN
esrj-117639	42	10	of	of	ADP
esrj-117639	42	11	newfoundland	newfoundland	PROPN
esrj-117639	42	12	in	in	ADP
esrj-117639	42	13	canada	canada	PROPN
esrj-117639	42	14	stand	stand	VERB
esrj-117639	42	15	out	out	ADP
esrj-117639	42	16	in	in	ADP
esrj-117639	42	17	this	this	DET
esrj-117639	42	18	direction	direction	NOUN
esrj-117639	42	19	.	.	PUNCT
esrj-117639	43	1	lelièvre	lelièvre	PROPN
esrj-117639	43	2	et	et	PROPN
esrj-117639	43	3	al	al	PROPN
esrj-117639	43	4	.	.	PROPN
esrj-117639	44	1	(	(	PUNCT
esrj-117639	44	2	2012	2012	NUM
esrj-117639	44	3	)	)	PUNCT
esrj-117639	44	4	conducted	conduct	VERB
esrj-117639	44	5	a	a	DET
esrj-117639	44	6	systematic	systematic	ADJ
esrj-117639	44	7	study	study	NOUN
esrj-117639	44	8	on	on	ADP
esrj-117639	44	9	the	the	DET
esrj-117639	44	10	generation	generation	NOUN
esrj-117639	44	11	and	and	CCONJ
esrj-117639	44	12	storage	storage	NOUN
esrj-117639	44	13	of	of	ADP
esrj-117639	44	14	unstructured	unstructured	ADJ
esrj-117639	44	15	grids	grid	NOUN
esrj-117639	44	16	,	,	PUNCT
esrj-117639	44	17	as	as	ADV
esrj-117639	44	18	well	well	ADV
esrj-117639	44	19	as	as	ADP
esrj-117639	44	20	forward	forward	ADV
esrj-117639	44	21	modeling	modeling	NOUN
esrj-117639	44	22	and	and	CCONJ
esrj-117639	44	23	inversion	inversion	NOUN
esrj-117639	44	24	constraints	constraint	NOUN
esrj-117639	44	25	based	base	VERB
esrj-117639	44	26	on	on	ADP
esrj-117639	44	27	unstructured	unstructured	ADJ
esrj-117639	44	28	grids	grid	NOUN
esrj-117639	44	29	,	,	PUNCT
esrj-117639	44	30	and	and	CCONJ
esrj-117639	44	31	successfully	successfully	ADV
esrj-117639	44	32	applied	apply	VERB
esrj-117639	44	33	them	they	PRON
esrj-117639	44	34	to	to	ADP
esrj-117639	44	35	the	the	DET
esrj-117639	44	36	problem	problem	NOUN
esrj-117639	44	37	of	of	ADP
esrj-117639	44	38	gravity	gravity	NOUN
esrj-117639	44	39	inversion	inversion	NOUN
esrj-117639	44	40	.	.	PUNCT
esrj-117639	45	1	subsequently	subsequently	ADV
esrj-117639	45	2	,	,	PUNCT
esrj-117639	45	3	sun	sun	PROPN
esrj-117639	45	4	et	et	PROPN
esrj-117639	45	5	al	al	PROPN
esrj-117639	45	6	.	.	PROPN
esrj-117639	45	7	(	(	PUNCT
esrj-117639	45	8	2021	2021	NUM
esrj-117639	45	9	)	)	PUNCT
esrj-117639	45	10	also	also	ADV
esrj-117639	45	11	conducted	conduct	VERB
esrj-117639	45	12	research	research	NOUN
esrj-117639	45	13	on	on	ADP
esrj-117639	45	14	gravity	gravity	NOUN
esrj-117639	45	15	inversions	inversion	NOUN
esrj-117639	45	16	based	base	VERB
esrj-117639	45	17	on	on	ADP
esrj-117639	45	18	unstructured	unstructured	ADJ
esrj-117639	45	19	grids	grid	NOUN
esrj-117639	45	20	and	and	CCONJ
esrj-117639	45	21	discussed	discuss	VERB
esrj-117639	45	22	the	the	DET
esrj-117639	45	23	role	role	NOUN
esrj-117639	45	24	of	of	ADP
esrj-117639	45	25	depth	depth	NOUN
esrj-117639	45	26	weighting	weighting	NOUN
esrj-117639	45	27	functions	function	NOUN
esrj-117639	45	28	and	and	CCONJ
esrj-117639	45	29	fuzzy	fuzzy	ADJ
esrj-117639	45	30	c	c	NOUN
esrj-117639	45	31	-	-	PUNCT
esrj-117639	45	32	means	mean	VERB
esrj-117639	45	33	clustering	cluster	VERB
esrj-117639	45	34	algorithms	algorithm	NOUN
esrj-117639	45	35	in	in	ADP
esrj-117639	45	36	inversion	inversion	NOUN
esrj-117639	45	37	.	.	PUNCT
esrj-117639	46	1	ma	ma	PROPN
esrj-117639	46	2	et	et	PROPN
esrj-117639	46	3	al	al	PROPN
esrj-117639	46	4	.	.	PROPN
esrj-117639	47	1	(	(	PUNCT
esrj-117639	47	2	2022	2022	NUM
esrj-117639	47	3	)	)	PUNCT
esrj-117639	47	4	utilized	utilize	VERB
esrj-117639	47	5	the	the	DET
esrj-117639	47	6	tilt	tilt	NOUN
esrj-117639	47	7	angle	angle	NOUN
esrj-117639	47	8	method	method	NOUN
esrj-117639	47	9	to	to	PART
esrj-117639	47	10	conduct	conduct	VERB
esrj-117639	47	11	local	local	ADJ
esrj-117639	47	12	adaptive	adaptive	ADJ
esrj-117639	47	13	grid	grid	NOUN
esrj-117639	47	14	research	research	NOUN
esrj-117639	47	15	on	on	ADP
esrj-117639	47	16	gravity	gravity	NOUN
esrj-117639	47	17	inversion	inversion	NOUN
esrj-117639	47	18	based	base	VERB
esrj-117639	47	19	on	on	ADP
esrj-117639	47	20	unstructured	unstructured	ADJ
esrj-117639	47	21	grids	grid	NOUN
esrj-117639	47	22	,	,	PUNCT
esrj-117639	47	23	improving	improve	VERB
esrj-117639	47	24	inversion	inversion	NOUN
esrj-117639	47	25	efficiency	efficiency	NOUN
esrj-117639	47	26	by	by	ADP
esrj-117639	47	27	reducing	reduce	VERB
esrj-117639	47	28	the	the	DET
esrj-117639	47	29	number	number	NOUN
esrj-117639	47	30	of	of	ADP
esrj-117639	47	31	grids	grid	NOUN
esrj-117639	47	32	.	.	PUNCT
esrj-117639	48	1	the	the	DET
esrj-117639	48	2	aforementioned	aforementioned	ADJ
esrj-117639	48	3	methods	method	NOUN
esrj-117639	48	4	are	be	AUX
esrj-117639	48	5	all	all	DET
esrj-117639	48	6	calculations	calculation	NOUN
esrj-117639	48	7	performed	perform	VERB
esrj-117639	48	8	in	in	ADP
esrj-117639	48	9	the	the	DET
esrj-117639	48	10	spatial	spatial	ADJ
esrj-117639	48	11	domain	domain	NOUN
esrj-117639	48	12	.	.	PUNCT
esrj-117639	49	1	however	however	ADV
esrj-117639	49	2	,	,	PUNCT
esrj-117639	49	3	when	when	SCONJ
esrj-117639	49	4	the	the	DET
esrj-117639	49	5	model	model	NOUN
esrj-117639	49	6	is	be	AUX
esrj-117639	49	7	complex	complex	ADJ
esrj-117639	49	8	or	or	CCONJ
esrj-117639	49	9	there	there	PRON
esrj-117639	49	10	are	be	VERB
esrj-117639	49	11	a	a	DET
esrj-117639	49	12	large	large	ADJ
esrj-117639	49	13	number	number	NOUN
esrj-117639	49	14	of	of	ADP
esrj-117639	49	15	observation	observation	NOUN
esrj-117639	49	16	data	datum	NOUN
esrj-117639	49	17	,	,	PUNCT
esrj-117639	49	18	this	this	DET
esrj-117639	49	19	process	process	NOUN
esrj-117639	49	20	is	be	AUX
esrj-117639	49	21	computationally	computationally	ADV
esrj-117639	49	22	intensive	intensive	ADJ
esrj-117639	49	23	.	.	PUNCT
esrj-117639	50	1	since	since	SCONJ
esrj-117639	50	2	the	the	DET
esrj-117639	50	3	advent	advent	NOUN
esrj-117639	50	4	of	of	ADP
esrj-117639	50	5	the	the	DET
esrj-117639	50	6	fast	fast	ADJ
esrj-117639	50	7	fourier	fourier	NOUN
esrj-117639	50	8	transform	transform	NOUN
esrj-117639	50	9	,	,	PUNCT
esrj-117639	50	10	many	many	ADJ
esrj-117639	50	11	scholars	scholar	NOUN
esrj-117639	50	12	have	have	AUX
esrj-117639	50	13	tried	try	VERB
esrj-117639	50	14	to	to	PART
esrj-117639	50	15	apply	apply	VERB
esrj-117639	50	16	it	it	PRON
esrj-117639	50	17	to	to	ADP
esrj-117639	50	18	gravity	gravity	NOUN
esrj-117639	50	19	data	datum	NOUN
esrj-117639	50	20	inversion	inversion	NOUN
esrj-117639	50	21	.	.	PUNCT
esrj-117639	51	1	parker	parker	PROPN
esrj-117639	51	2	(	(	PUNCT
esrj-117639	51	3	1972	1972	NUM
esrj-117639	51	4	)	)	PUNCT
esrj-117639	51	5	provided	provide	VERB
esrj-117639	51	6	a	a	DET
esrj-117639	51	7	forward	forward	ADJ
esrj-117639	51	8	formula	formula	NOUN
esrj-117639	51	9	for	for	ADP
esrj-117639	51	10	applying	apply	VERB
esrj-117639	51	11	the	the	DET
esrj-117639	51	12	fast	fast	ADJ
esrj-117639	51	13	fourier	fourier	NOUN
esrj-117639	51	14	transform	transform	NOUN
esrj-117639	51	15	in	in	ADP
esrj-117639	51	16	the	the	DET
esrj-117639	51	17	gravity	gravity	NOUN
esrj-117639	51	18	method	method	NOUN
esrj-117639	51	19	,	,	PUNCT
esrj-117639	51	20	which	which	PRON
esrj-117639	51	21	is	be	AUX
esrj-117639	51	22	characterized	characterize	VERB
esrj-117639	51	23	by	by	ADP
esrj-117639	51	24	its	its	PRON
esrj-117639	51	25	speed	speed	NOUN
esrj-117639	51	26	and	and	CCONJ
esrj-117639	51	27	high	high	ADJ
esrj-117639	51	28	accuracy	accuracy	NOUN
esrj-117639	51	29	.	.	PUNCT
esrj-117639	52	1	with	with	ADP
esrj-117639	52	2	the	the	DET
esrj-117639	52	3	advancement	advancement	NOUN
esrj-117639	52	4	of	of	ADP
esrj-117639	52	5	computers	computer	NOUN
esrj-117639	52	6	,	,	PUNCT
esrj-117639	52	7	parker	parker	PROPN
esrj-117639	52	8	’s	’s	PART
esrj-117639	52	9	method	method	NOUN
esrj-117639	52	10	based	base	VERB
esrj-117639	52	11	on	on	ADP
esrj-117639	52	12	the	the	DET
esrj-117639	52	13	fast	fast	ADJ
esrj-117639	52	14	fourier	fourier	NOUN
esrj-117639	52	15	transform	transform	NOUN
esrj-117639	52	16	has	have	AUX
esrj-117639	52	17	become	become	VERB
esrj-117639	52	18	feasible	feasible	ADJ
esrj-117639	52	19	for	for	ADP
esrj-117639	52	20	fast	fast	ADJ
esrj-117639	52	21	computing	computing	NOUN
esrj-117639	52	22	of	of	ADP
esrj-117639	52	23	magnetic	magnetic	ADJ
esrj-117639	52	24	or	or	CCONJ
esrj-117639	52	25	gravity	gravity	NOUN
esrj-117639	52	26	fields	field	NOUN
esrj-117639	52	27	.	.	PUNCT
esrj-117639	53	1	oldenburg	oldenburg	PROPN
esrj-117639	53	2	(	(	PUNCT
esrj-117639	53	3	1974	1974	NUM
esrj-117639	53	4	)	)	PUNCT
esrj-117639	53	5	proved	prove	VERB
esrj-117639	53	6	that	that	SCONJ
esrj-117639	53	7	parker	parker	PROPN
esrj-117639	53	8	’s	’s	PART
esrj-117639	53	9	expression	expression	NOUN
esrj-117639	53	10	can	can	AUX
esrj-117639	53	11	be	be	AUX
esrj-117639	53	12	used	use	VERB
esrj-117639	53	13	to	to	PART
esrj-117639	53	14	determine	determine	VERB
esrj-117639	53	15	the	the	DET
esrj-117639	53	16	geometric	geometric	ADJ
esrj-117639	53	17	shape	shape	NOUN
esrj-117639	53	18	of	of	ADP
esrj-117639	53	19	the	the	DET
esrj-117639	53	20	interface	interface	NOUN
esrj-117639	53	21	by	by	ADP
esrj-117639	53	22	gravity	gravity	NOUN
esrj-117639	53	23	anomaly	anomaly	NOUN
esrj-117639	53	24	data	datum	NOUN
esrj-117639	53	25	.	.	PUNCT
esrj-117639	54	1	granser	granser	PROPN
esrj-117639	54	2	(	(	PUNCT
esrj-117639	54	3	1987a	1987a	NUM
esrj-117639	54	4	)	)	PUNCT
esrj-117639	54	5	rearranged	rearrange	VERB
esrj-117639	54	6	the	the	DET
esrj-117639	54	7	forward	forward	ADJ
esrj-117639	54	8	algorithm	algorithm	NOUN
esrj-117639	54	9	to	to	PART
esrj-117639	54	10	obtain	obtain	VERB
esrj-117639	54	11	the	the	DET
esrj-117639	54	12	depth	depth	NOUN
esrj-117639	54	13	of	of	ADP
esrj-117639	54	14	the	the	DET
esrj-117639	54	15	sedimentary	sedimentary	ADJ
esrj-117639	54	16	basin	basin	NOUN
esrj-117639	54	17	basement	basement	NOUN
esrj-117639	54	18	.	.	PUNCT
esrj-117639	55	1	this	this	DET
esrj-117639	55	2	method	method	NOUN
esrj-117639	55	3	is	be	AUX
esrj-117639	55	4	similar	similar	ADJ
esrj-117639	55	5	to	to	ADP
esrj-117639	55	6	the	the	DET
esrj-117639	55	7	inversion	inversion	NOUN
esrj-117639	55	8	program	program	NOUN
esrj-117639	55	9	developed	develop	VERB
esrj-117639	55	10	by	by	ADP
esrj-117639	55	11	oldenburg	oldenburg	PROPN
esrj-117639	55	12	.	.	PUNCT
esrj-117639	56	1	granser	granser	PROPN
esrj-117639	56	2	(	(	PUNCT
esrj-117639	56	3	1987b	1987b	NUM
esrj-117639	56	4	)	)	PUNCT
esrj-117639	56	5	also	also	ADV
esrj-117639	56	6	applied	apply	VERB
esrj-117639	56	7	another	another	DET
esrj-117639	56	8	nonlinear	nonlinear	ADJ
esrj-117639	56	9	frequency	frequency	NOUN
esrj-117639	56	10	-	-	PUNCT
esrj-117639	56	11	domain	domain	NOUN
esrj-117639	56	12	inversion	inversion	NOUN
esrj-117639	56	13	method	method	NOUN
esrj-117639	56	14	based	base	VERB
esrj-117639	56	15	on	on	ADP
esrj-117639	56	16	schmidt	schmidt	PROPN
esrj-117639	56	17	-	-	PUNCT
esrj-117639	56	18	lichtenstein	lichtenstein	PROPN
esrj-117639	56	19	to	to	ADP
esrj-117639	56	20	the	the	DET
esrj-117639	56	21	case	case	NOUN
esrj-117639	56	22	of	of	ADP
esrj-117639	56	23	a	a	DET
esrj-117639	56	24	3d	3d	NUM
esrj-117639	56	25	sedimentary	sedimentary	ADJ
esrj-117639	56	26	basin	basin	NOUN
esrj-117639	56	27	with	with	ADP
esrj-117639	56	28	uniform	uniform	ADJ
esrj-117639	56	29	density	density	NOUN
esrj-117639	56	30	differences	difference	NOUN
esrj-117639	56	31	.	.	PUNCT
esrj-117639	57	1	although	although	SCONJ
esrj-117639	57	2	algorithms	algorithm	NOUN
esrj-117639	57	3	based	base	VERB
esrj-117639	57	4	on	on	ADP
esrj-117639	57	5	the	the	DET
esrj-117639	57	6	fast	fast	ADJ
esrj-117639	57	7	fourier	fourier	NOUN
esrj-117639	57	8	transform	transform	NOUN
esrj-117639	57	9	reduce	reduce	VERB
esrj-117639	57	10	computation	computation	NOUN
esrj-117639	57	11	time	time	NOUN
esrj-117639	57	12	,	,	PUNCT
esrj-117639	57	13	they	they	PRON
esrj-117639	57	14	require	require	VERB
esrj-117639	57	15	an	an	DET
esrj-117639	57	16	average	average	ADJ
esrj-117639	57	17	depth	depth	NOUN
esrj-117639	57	18	of	of	ADP
esrj-117639	57	19	the	the	DET
esrj-117639	57	20	interface	interface	NOUN
esrj-117639	57	21	and	and	CCONJ
esrj-117639	57	22	a	a	DET
esrj-117639	57	23	low	low	ADJ
esrj-117639	57	24	-	-	PUNCT
esrj-117639	57	25	pass	pass	NOUN
esrj-117639	57	26	filter	filter	NOUN
esrj-117639	57	27	to	to	PART
esrj-117639	57	28	achieve	achieve	VERB
esrj-117639	57	29	convergence	convergence	NOUN
esrj-117639	57	30	in	in	ADP
esrj-117639	57	31	the	the	DET
esrj-117639	57	32	inversion	inversion	NOUN
esrj-117639	57	33	process	process	NOUN
esrj-117639	57	34	(	(	PUNCT
esrj-117639	57	35	oldenburg	oldenburg	PROPN
esrj-117639	57	36	,	,	PUNCT
esrj-117639	57	37	1974	1974	NUM
esrj-117639	57	38	;	;	PUNCT
esrj-117639	57	39	granser	granser	NOUN
esrj-117639	57	40	,	,	PUNCT
esrj-117639	57	41	1987a	1987a	NUM
esrj-117639	57	42	;	;	PUNCT
esrj-117639	57	43	granser	granser	NOUN
esrj-117639	57	44	,	,	PUNCT
esrj-117639	57	45	1987b	1987b	NUM
esrj-117639	57	46	)	)	PUNCT
esrj-117639	57	47	.	.	PUNCT
esrj-117639	58	1	based	base	VERB
esrj-117639	58	2	on	on	ADP
esrj-117639	58	3	the	the	DET
esrj-117639	58	4	parker	parker	PROPN
esrj-117639	58	5	-	-	PUNCT
esrj-117639	58	6	oldenburg	oldenburg	PROPN
esrj-117639	58	7	formula	formula	NOUN
esrj-117639	58	8	,	,	PUNCT
esrj-117639	58	9	some	some	DET
esrj-117639	58	10	authors	author	NOUN
esrj-117639	58	11	have	have	AUX
esrj-117639	58	12	developed	develop	VERB
esrj-117639	58	13	computer	computer	NOUN
esrj-117639	58	14	programs	program	NOUN
esrj-117639	58	15	to	to	PART
esrj-117639	58	16	determine	determine	VERB
esrj-117639	58	17	the	the	DET
esrj-117639	58	18	geometric	geometric	ADJ
esrj-117639	58	19	shape	shape	NOUN
esrj-117639	58	20	of	of	ADP
esrj-117639	58	21	density	density	NOUN
esrj-117639	58	22	interfaces	interface	NOUN
esrj-117639	58	23	from	from	ADP
esrj-117639	58	24	gravity	gravity	NOUN
esrj-117639	58	25	anomaly	anomaly	NOUN
esrj-117639	58	26	data	datum	NOUN
esrj-117639	58	27	.	.	PUNCT
esrj-117639	59	1	nagendra	nagendra	PROPN
esrj-117639	59	2	et	et	PROPN
esrj-117639	59	3	al	al	PROPN
esrj-117639	59	4	.	.	PROPN
esrj-117639	59	5	(	(	PUNCT
esrj-117639	59	6	1996	1996	NUM
esrj-117639	59	7	)	)	PUNCT
esrj-117639	59	8	developed	develop	VERB
esrj-117639	59	9	fortran	fortran	ADJ
esrj-117639	59	10	language	language	NOUN
esrj-117639	59	11	program	program	NOUN
esrj-117639	59	12	codes	code	NOUN
esrj-117639	59	13	based	base	VERB
esrj-117639	59	14	on	on	ADP
esrj-117639	59	15	the	the	DET
esrj-117639	59	16	parker	parker	PROPN
esrj-117639	59	17	-	-	PUNCT
esrj-117639	59	18	oldenburg	oldenburg	PROPN
esrj-117639	59	19	method	method	NOUN
esrj-117639	59	20	and	and	CCONJ
esrj-117639	59	21	the	the	DET
esrj-117639	59	22	granser	granser	NOUN
esrj-117639	59	23	method	method	NOUN
esrj-117639	59	24	,	,	PUNCT
esrj-117639	59	25	respectively	respectively	ADV
esrj-117639	59	26	,	,	PUNCT
esrj-117639	59	27	for	for	ADP
esrj-117639	59	28	analyzing	analyze	VERB
esrj-117639	59	29	two	two	NUM
esrj-117639	59	30	-	-	PUNCT
esrj-117639	59	31	dimensional	dimensional	ADJ
esrj-117639	59	32	gravity	gravity	NOUN
esrj-117639	59	33	anomaly	anomaly	NOUN
esrj-117639	59	34	data	datum	NOUN
esrj-117639	59	35	.	.	PUNCT
esrj-117639	60	1	gómez	gómez	NOUN
esrj-117639	60	2	-	-	PUNCT
esrj-117639	60	3	ortiz	ortiz	PROPN
esrj-117639	60	4	and	and	CCONJ
esrj-117639	60	5	agarwal	agarwal	PROPN
esrj-117639	60	6	(	(	PUNCT
esrj-117639	60	7	2005	2005	NUM
esrj-117639	60	8	)	)	PUNCT
esrj-117639	60	9	proposed	propose	VERB
esrj-117639	60	10	a	a	DET
esrj-117639	60	11	matlab	matlab	PROPN
esrj-117639	60	12	language	language	NOUN
esrj-117639	60	13	program	program	NOUN
esrj-117639	60	14	code	code	NOUN
esrj-117639	60	15	for	for	ADP
esrj-117639	60	16	the	the	DET
esrj-117639	60	17	three	three	NUM
esrj-117639	60	18	-	-	PUNCT
esrj-117639	60	19	dimensional	dimensional	ADJ
esrj-117639	60	20	extension	extension	NOUN
esrj-117639	60	21	of	of	ADP
esrj-117639	60	22	the	the	DET
esrj-117639	60	23	parker	parker	PROPN
esrj-117639	60	24	-	-	PUNCT
esrj-117639	60	25	oldenburg	oldenburg	PROPN
esrj-117639	60	26	method	method	NOUN
esrj-117639	60	27	to	to	PART
esrj-117639	60	28	obtain	obtain	VERB
esrj-117639	60	29	the	the	DET
esrj-117639	60	30	geometric	geometric	ADJ
esrj-117639	60	31	shape	shape	NOUN
esrj-117639	60	32	of	of	ADP
esrj-117639	60	33	density	density	NOUN
esrj-117639	60	34	interfaces	interface	NOUN
esrj-117639	60	35	.	.	PUNCT
esrj-117639	61	1	the	the	DET
esrj-117639	61	2	linear	linear	ADJ
esrj-117639	61	3	inversion	inversion	NOUN
esrj-117639	61	4	method	method	NOUN
esrj-117639	61	5	is	be	AUX
esrj-117639	61	6	simple	simple	ADJ
esrj-117639	61	7	and	and	CCONJ
esrj-117639	61	8	widely	widely	ADV
esrj-117639	61	9	used	use	VERB
esrj-117639	61	10	due	due	ADP
esrj-117639	61	11	to	to	ADP
esrj-117639	61	12	its	its	PRON
esrj-117639	61	13	linearization	linearization	NOUN
esrj-117639	61	14	process	process	NOUN
esrj-117639	61	15	.	.	PUNCT
esrj-117639	62	1	however	however	ADV
esrj-117639	62	2	,	,	PUNCT
esrj-117639	62	3	it	it	PRON
esrj-117639	62	4	is	be	AUX
esrj-117639	62	5	highly	highly	ADV
esrj-117639	62	6	dependent	dependent	ADJ
esrj-117639	62	7	on	on	ADP
esrj-117639	62	8	the	the	DET
esrj-117639	62	9	selection	selection	NOUN
esrj-117639	62	10	of	of	ADP
esrj-117639	62	11	initial	initial	ADJ
esrj-117639	62	12	parameters	parameter	NOUN
esrj-117639	62	13	,	,	PUNCT
esrj-117639	62	14	leading	lead	VERB
esrj-117639	62	15	to	to	ADP
esrj-117639	62	16	a	a	DET
esrj-117639	62	17	decrease	decrease	NOUN
esrj-117639	62	18	in	in	ADP
esrj-117639	62	19	inversion	inversion	NOUN
esrj-117639	62	20	accuracy	accuracy	NOUN
esrj-117639	62	21	.	.	PUNCT
esrj-117639	63	1	therefore	therefore	ADV
esrj-117639	63	2	,	,	PUNCT
esrj-117639	63	3	in	in	ADP
esrj-117639	63	4	recent	recent	ADJ
esrj-117639	63	5	years	year	NOUN
esrj-117639	63	6	,	,	PUNCT
esrj-117639	63	7	non	non	ADJ
esrj-117639	63	8	-	-	ADJ
esrj-117639	63	9	linear	linear	ADJ
esrj-117639	63	10	inversion	inversion	NOUN
esrj-117639	63	11	methods	method	NOUN
esrj-117639	63	12	have	have	AUX
esrj-117639	63	13	attracted	attract	VERB
esrj-117639	63	14	the	the	DET
esrj-117639	63	15	attention	attention	NOUN
esrj-117639	63	16	of	of	ADP
esrj-117639	63	17	many	many	ADJ
esrj-117639	63	18	geophysicists	geophysicist	NOUN
esrj-117639	63	19	.	.	PUNCT
esrj-117639	64	1	artificial	artificial	ADJ
esrj-117639	64	2	neural	neural	ADJ
esrj-117639	64	3	networks	network	NOUN
esrj-117639	64	4	(	(	PUNCT
esrj-117639	64	5	anns	anns	PROPN
esrj-117639	64	6	)	)	PUNCT
esrj-117639	64	7	are	be	AUX
esrj-117639	64	8	one	one	NUM
esrj-117639	64	9	of	of	ADP
esrj-117639	64	10	the	the	DET
esrj-117639	64	11	most	most	ADV
esrj-117639	64	12	active	active	ADJ
esrj-117639	64	13	non	non	ADJ
esrj-117639	64	14	-	-	ADJ
esrj-117639	64	15	linear	linear	ADJ
esrj-117639	64	16	inversion	inversion	NOUN
esrj-117639	64	17	methods	method	NOUN
esrj-117639	64	18	,	,	PUNCT
esrj-117639	64	19	as	as	SCONJ
esrj-117639	64	20	they	they	PRON
esrj-117639	64	21	can	can	AUX
esrj-117639	64	22	approximate	approximate	VERB
esrj-117639	64	23	a	a	DET
esrj-117639	64	24	non	non	ADJ
esrj-117639	64	25	-	-	ADJ
esrj-117639	64	26	linear	linear	ADJ
esrj-117639	64	27	continuous	continuous	ADJ
esrj-117639	64	28	function	function	NOUN
esrj-117639	64	29	with	with	ADP
esrj-117639	64	30	any	any	DET
esrj-117639	64	31	desired	desire	VERB
esrj-117639	64	32	accuracy	accuracy	NOUN
esrj-117639	64	33	.	.	PUNCT
esrj-117639	65	1	anns	anns	PROPN
esrj-117639	65	2	use	use	VERB
esrj-117639	65	3	the	the	DET
esrj-117639	65	4	input	input	NOUN
esrj-117639	65	5	variables	variable	NOUN
esrj-117639	65	6	as	as	ADP
esrj-117639	65	7	independent	independent	ADJ
esrj-117639	65	8	variables	variable	NOUN
esrj-117639	65	9	,	,	PUNCT
esrj-117639	65	10	and	and	CCONJ
esrj-117639	65	11	the	the	DET
esrj-117639	65	12	output	output	NOUN
esrj-117639	65	13	variables	variable	VERB
esrj-117639	65	14	as	as	ADP
esrj-117639	65	15	dependent	dependent	ADJ
esrj-117639	65	16	variables	variable	NOUN
esrj-117639	65	17	,	,	PUNCT
esrj-117639	65	18	and	and	CCONJ
esrj-117639	65	19	can	can	AUX
esrj-117639	65	20	replace	replace	VERB
esrj-117639	65	21	this	this	DET
esrj-117639	65	22	function	function	NOUN
esrj-117639	65	23	with	with	ADP
esrj-117639	65	24	any	any	DET
esrj-117639	65	25	high	high	ADJ
esrj-117639	65	26	degree	degree	NOUN
esrj-117639	65	27	of	of	ADP
esrj-117639	65	28	accuracy	accuracy	NOUN
esrj-117639	65	29	(	(	PUNCT
esrj-117639	65	30	van	van	PROPN
esrj-117639	65	31	der	der	PROPN
esrj-117639	65	32	baan	baan	PROPN
esrj-117639	65	33	and	and	CCONJ
esrj-117639	65	34	jutten	jutten	PROPN
esrj-117639	65	35	,	,	PUNCT
esrj-117639	65	36	2000	2000	NUM
esrj-117639	65	37	)	)	PUNCT
esrj-117639	65	38	.	.	PUNCT
esrj-117639	66	1	since	since	SCONJ
esrj-117639	66	2	the	the	DET
esrj-117639	66	3	late	late	ADJ
esrj-117639	66	4	1980s	1980	NOUN
esrj-117639	66	5	,	,	PUNCT
esrj-117639	66	6	artificial	artificial	ADJ
esrj-117639	66	7	neural	neural	ADJ
esrj-117639	66	8	networks	network	NOUN
esrj-117639	66	9	(	(	PUNCT
esrj-117639	66	10	anns	anns	PROPN
esrj-117639	66	11	)	)	PUNCT
esrj-117639	66	12	have	have	AUX
esrj-117639	66	13	been	be	AUX
esrj-117639	66	14	applied	apply	VERB
esrj-117639	66	15	to	to	ADP
esrj-117639	66	16	geophysical	geophysical	ADJ
esrj-117639	66	17	problems	problem	NOUN
esrj-117639	66	18	(	(	PUNCT
esrj-117639	66	19	raiche	raiche	NOUN
esrj-117639	66	20	,	,	PUNCT
esrj-117639	66	21	1991	1991	NUM
esrj-117639	66	22	)	)	PUNCT
esrj-117639	66	23	and	and	CCONJ
esrj-117639	66	24	increasingly	increasingly	ADV
esrj-117639	66	25	used	use	VERB
esrj-117639	66	26	to	to	PART
esrj-117639	66	27	interpret	interpret	VERB
esrj-117639	66	28	geophysical	geophysical	ADJ
esrj-117639	66	29	data	datum	NOUN
esrj-117639	66	30	.	.	PUNCT
esrj-117639	67	1	anns	anns	PROPN
esrj-117639	67	2	,	,	PUNCT
esrj-117639	67	3	as	as	ADP
esrj-117639	67	4	a	a	DET
esrj-117639	67	5	computer	computer	NOUN
esrj-117639	67	6	processing	processing	NOUN
esrj-117639	67	7	method	method	NOUN
esrj-117639	67	8	,	,	PUNCT
esrj-117639	67	9	have	have	AUX
esrj-117639	67	10	shown	show	VERB
esrj-117639	67	11	great	great	ADJ
esrj-117639	67	12	potential	potential	NOUN
esrj-117639	67	13	for	for	ADP
esrj-117639	67	14	pattern	pattern	NOUN
esrj-117639	67	15	recognition	recognition	NOUN
esrj-117639	67	16	in	in	ADP
esrj-117639	67	17	geophysical	geophysical	ADJ
esrj-117639	67	18	exploration	exploration	NOUN
esrj-117639	67	19	problems	problem	NOUN
esrj-117639	67	20	,	,	PUNCT
esrj-117639	67	21	such	such	ADJ
esrj-117639	67	22	as	as	ADP
esrj-117639	67	23	log	log	NOUN
esrj-117639	67	24	interpretation	interpretation	NOUN
esrj-117639	67	25	(	(	PUNCT
esrj-117639	67	26	wiener	wiener	NOUN
esrj-117639	67	27	et	et	PROPN
esrj-117639	67	28	al	al	PROPN
esrj-117639	67	29	.	.	PROPN
esrj-117639	67	30	,	,	PUNCT
esrj-117639	67	31	1991	1991	NUM
esrj-117639	67	32	)	)	PUNCT
esrj-117639	67	33	,	,	PUNCT
esrj-117639	67	34	electromagnetic	electromagnetic	ADJ
esrj-117639	67	35	profiling	profiling	NOUN
esrj-117639	67	36	data	datum	NOUN
esrj-117639	67	37	processing	processing	NOUN
esrj-117639	67	38	(	(	PUNCT
esrj-117639	67	39	poulton	poulton	PROPN
esrj-117639	67	40	,	,	PUNCT
esrj-117639	67	41	1992	1992	NUM
esrj-117639	67	42	;	;	PUNCT
esrj-117639	67	43	winkler	winkler	NOUN
esrj-117639	67	44	,	,	PUNCT
esrj-117639	67	45	1994	1994	NUM
esrj-117639	67	46	)	)	PUNCT
esrj-117639	67	47	,	,	PUNCT
esrj-117639	67	48	ground	ground	NOUN
esrj-117639	67	49	penetrating	penetrate	VERB
esrj-117639	67	50	radar	radar	NOUN
esrj-117639	67	51	(	(	PUNCT
esrj-117639	67	52	poulton	poulton	PROPN
esrj-117639	67	53	&	&	CCONJ
esrj-117639	67	54	el	el	PROPN
esrj-117639	67	55	-	-	PUNCT
esrj-117639	67	56	fouly	fouly	ADV
esrj-117639	67	57	,	,	PUNCT
esrj-117639	67	58	1991	1991	NUM
esrj-117639	67	59	)	)	PUNCT
esrj-117639	67	60	,	,	PUNCT
esrj-117639	67	61	seismic	seismic	ADJ
esrj-117639	67	62	waveform	waveform	NOUN
esrj-117639	67	63	recognition	recognition	NOUN
esrj-117639	67	64	(	(	PUNCT
esrj-117639	67	65	dai	dai	PROPN
esrj-117639	67	66	&	&	CCONJ
esrj-117639	67	67	macbeth	macbeth	PROPN
esrj-117639	67	68	,	,	PUNCT
esrj-117639	67	69	1995	1995	NUM
esrj-117639	67	70	;	;	PUNCT
esrj-117639	67	71	ashida	ashida	PROPN
esrj-117639	67	72	,	,	PUNCT
esrj-117639	67	73	1996	1996	NUM
esrj-117639	67	74	)	)	PUNCT
esrj-117639	67	75	,	,	PUNCT
esrj-117639	67	76	and	and	CCONJ
esrj-117639	67	77	more	more	ADJ
esrj-117639	67	78	.	.	PUNCT
esrj-117639	68	1	in	in	ADP
esrj-117639	68	2	recent	recent	ADJ
esrj-117639	68	3	years	year	NOUN
esrj-117639	68	4	,	,	PUNCT
esrj-117639	68	5	the	the	DET
esrj-117639	68	6	petroleum	petroleum	NOUN
esrj-117639	68	7	industry	industry	NOUN
esrj-117639	68	8	has	have	AUX
esrj-117639	68	9	been	be	AUX
esrj-117639	68	10	using	use	VERB
esrj-117639	68	11	anns	anns	NOUN
esrj-117639	68	12	to	to	PART
esrj-117639	68	13	process	process	VERB
esrj-117639	68	14	measured	measure	VERB
esrj-117639	68	15	data	datum	NOUN
esrj-117639	68	16	.	.	PUNCT
esrj-117639	69	1	anns	anns	PROPN
esrj-117639	69	2	have	have	AUX
esrj-117639	69	3	also	also	ADV
esrj-117639	69	4	been	be	AUX
esrj-117639	69	5	applied	apply	VERB
esrj-117639	69	6	in	in	ADP
esrj-117639	69	7	the	the	DET
esrj-117639	69	8	interpretation	interpretation	NOUN
esrj-117639	69	9	of	of	ADP
esrj-117639	69	10	gravity	gravity	NOUN
esrj-117639	69	11	and	and	CCONJ
esrj-117639	69	12	magnetic	magnetic	ADJ
esrj-117639	69	13	data	datum	NOUN
esrj-117639	69	14	.	.	PUNCT
esrj-117639	70	1	over	over	ADP
esrj-117639	70	2	the	the	DET
esrj-117639	70	3	past	past	ADJ
esrj-117639	70	4	few	few	ADJ
esrj-117639	70	5	decades	decade	NOUN
esrj-117639	70	6	,	,	PUNCT
esrj-117639	70	7	with	with	ADP
esrj-117639	70	8	the	the	DET
esrj-117639	70	9	continuous	continuous	ADJ
esrj-117639	70	10	improvement	improvement	NOUN
esrj-117639	70	11	of	of	ADP
esrj-117639	70	12	computer	computer	NOUN
esrj-117639	70	13	performance	performance	NOUN
esrj-117639	70	14	,	,	PUNCT
esrj-117639	70	15	the	the	DET
esrj-117639	70	16	application	application	NOUN
esrj-117639	70	17	of	of	ADP
esrj-117639	70	18	machine	machine	NOUN
esrj-117639	70	19	learning	learning	NOUN
esrj-117639	70	20	in	in	ADP
esrj-117639	70	21	gravity	gravity	NOUN
esrj-117639	70	22	exploration	exploration	NOUN
esrj-117639	70	23	has	have	AUX
esrj-117639	70	24	gradually	gradually	ADV
esrj-117639	70	25	developed	develop	VERB
esrj-117639	70	26	into	into	ADP
esrj-117639	70	27	a	a	DET
esrj-117639	70	28	hot	hot	ADJ
esrj-117639	70	29	research	research	NOUN
esrj-117639	70	30	direction	direction	NOUN
esrj-117639	70	31	.	.	PUNCT
esrj-117639	71	1	especially	especially	ADV
esrj-117639	71	2	after	after	ADP
esrj-117639	71	3	adler	adler	PROPN
esrj-117639	71	4	&	&	CCONJ
esrj-117639	71	5	oktem	oktem	PROPN
esrj-117639	71	6	(	(	PUNCT
esrj-117639	71	7	2017	2017	NUM
esrj-117639	71	8	)	)	PUNCT
esrj-117639	71	9	and	and	CCONJ
esrj-117639	71	10	jin	jin	NOUN
esrj-117639	71	11	et	et	PROPN
esrj-117639	71	12	al	al	PROPN
esrj-117639	71	13	.	.	PROPN
esrj-117639	72	1	(	(	PUNCT
esrj-117639	72	2	2017	2017	NUM
esrj-117639	72	3	)	)	PUNCT
esrj-117639	72	4	demonstrated	demonstrate	VERB
esrj-117639	72	5	that	that	SCONJ
esrj-117639	72	6	machine	machine	NOUN
esrj-117639	72	7	learning	learning	NOUN
esrj-117639	72	8	can	can	AUX
esrj-117639	72	9	solve	solve	VERB
esrj-117639	72	10	ill	ill	ADV
esrj-117639	72	11	-	-	PUNCT
esrj-117639	72	12	posed	pose	VERB
esrj-117639	72	13	problems	problem	NOUN
esrj-117639	72	14	,	,	PUNCT
esrj-117639	72	15	various	various	ADJ
esrj-117639	72	16	geophysical	geophysical	ADJ
esrj-117639	72	17	inversion	inversion	NOUN
esrj-117639	72	18	methods	method	NOUN
esrj-117639	72	19	based	base	VERB
esrj-117639	72	20	on	on	ADP
esrj-117639	72	21	machine	machine	NOUN
esrj-117639	72	22	learning	learning	NOUN
esrj-117639	72	23	have	have	AUX
esrj-117639	72	24	emerged	emerge	VERB
esrj-117639	72	25	like	like	ADP
esrj-117639	72	26	mushrooms	mushroom	NOUN
esrj-117639	72	27	after	after	ADP
esrj-117639	72	28	a	a	DET
esrj-117639	72	29	spring	spring	NOUN
esrj-117639	72	30	rain	rain	NOUN
esrj-117639	72	31	.	.	PUNCT
esrj-117639	73	1	carpenter	carpenter	NOUN
esrj-117639	73	2	(	(	PUNCT
esrj-117639	73	3	2021	2021	NUM
esrj-117639	73	4	)	)	PUNCT
esrj-117639	73	5	was	be	AUX
esrj-117639	73	6	the	the	DET
esrj-117639	73	7	first	first	ADJ
esrj-117639	73	8	to	to	PART
esrj-117639	73	9	establish	establish	VERB
esrj-117639	73	10	a	a	DET
esrj-117639	73	11	mapping	mapping	NOUN
esrj-117639	73	12	relationship	relationship	NOUN
esrj-117639	73	13	between	between	ADP
esrj-117639	73	14	gravity	gravity	NOUN
esrj-117639	73	15	data	datum	NOUN
esrj-117639	73	16	and	and	CCONJ
esrj-117639	73	17	salt	salt	NOUN
esrj-117639	73	18	structure	structure	NOUN
esrj-117639	73	19	through	through	ADP
esrj-117639	73	20	machine	machine	NOUN
esrj-117639	73	21	learning	learning	NOUN
esrj-117639	73	22	,	,	PUNCT
esrj-117639	73	23	avoiding	avoid	VERB
esrj-117639	73	24	the	the	DET
esrj-117639	73	25	iterative	iterative	ADJ
esrj-117639	73	26	process	process	NOUN
esrj-117639	73	27	of	of	ADP
esrj-117639	73	28	conventional	conventional	ADJ
esrj-117639	73	29	gravity	gravity	NOUN
esrj-117639	73	30	inversion	inversion	NOUN
esrj-117639	73	31	;	;	PUNCT
esrj-117639	73	32	wang	wang	PROPN
esrj-117639	73	33	et	et	PROPN
esrj-117639	73	34	al	al	PROPN
esrj-117639	73	35	.	.	PROPN
esrj-117639	74	1	(	(	PUNCT
esrj-117639	74	2	2021	2021	NUM
esrj-117639	74	3	)	)	PUNCT
esrj-117639	74	4	used	use	VERB
esrj-117639	74	5	u	u	NOUN
esrj-117639	74	6	-	-	NOUN
esrj-117639	74	7	net	net	ADJ
esrj-117639	74	8	to	to	PART
esrj-117639	74	9	complete	complete	VERB
esrj-117639	74	10	the	the	DET
esrj-117639	74	11	three	three	NUM
esrj-117639	74	12	-	-	PUNCT
esrj-117639	74	13	dimensional	dimensional	ADJ
esrj-117639	74	14	inversion	inversion	NOUN
esrj-117639	74	15	of	of	ADP
esrj-117639	74	16	gravity	gravity	NOUN
esrj-117639	74	17	data	datum	NOUN
esrj-117639	74	18	in	in	ADP
esrj-117639	74	19	the	the	DET
esrj-117639	74	20	st	st	PROPN
esrj-117639	74	21	.	.	PROPN
esrj-117639	74	22	nicholas	nicholas	PROPN
esrj-117639	74	23	mining	mining	PROPN
esrj-117639	74	24	area	area	NOUN
esrj-117639	74	25	,	,	PUNCT
esrj-117639	74	26	and	and	CCONJ
esrj-117639	74	27	proved	prove	VERB
esrj-117639	74	28	through	through	ADP
esrj-117639	74	29	comparative	comparative	ADJ
esrj-117639	74	30	experiments	experiment	NOUN
esrj-117639	74	31	that	that	SCONJ
esrj-117639	74	32	the	the	DET
esrj-117639	74	33	machine	machine	NOUN
esrj-117639	74	34	learning	learn	VERB
esrj-117639	74	35	inversion	inversion	NOUN
esrj-117639	74	36	results	result	NOUN
esrj-117639	74	37	have	have	VERB
esrj-117639	74	38	higher	high	ADJ
esrj-117639	74	39	resolution	resolution	NOUN
esrj-117639	74	40	and	and	CCONJ
esrj-117639	74	41	clearer	clear	ADJ
esrj-117639	74	42	boundary	boundary	ADJ
esrj-117639	74	43	positions	position	NOUN
esrj-117639	74	44	than	than	ADP
esrj-117639	74	45	conventional	conventional	ADJ
esrj-117639	74	46	inversion	inversion	NOUN
esrj-117639	74	47	.	.	PUNCT
esrj-117639	75	1	li	li	PROPN
esrj-117639	75	2	et	et	PROPN
esrj-117639	75	3	al	al	PROPN
esrj-117639	75	4	.	.	PROPN
esrj-117639	75	5	(	(	PUNCT
esrj-117639	75	6	2022	2022	NUM
esrj-117639	75	7	)	)	PUNCT
esrj-117639	75	8	used	use	VERB
esrj-117639	75	9	a	a	DET
esrj-117639	75	10	self	self	NOUN
esrj-117639	75	11	-	-	PUNCT
esrj-117639	75	12	supervised	supervise	VERB
esrj-117639	75	13	deep	deep	ADJ
esrj-117639	75	14	learning	learning	NOUN
esrj-117639	75	15	method	method	NOUN
esrj-117639	75	16	to	to	PART
esrj-117639	75	17	establish	establish	VERB
esrj-117639	75	18	a	a	DET
esrj-117639	75	19	new	new	ADJ
esrj-117639	75	20	ssgi	ssgi	NOUN
esrj-117639	75	21	network	network	NOUN
esrj-117639	75	22	.	.	PUNCT
esrj-117639	76	1	this	this	DET
esrj-117639	76	2	method	method	NOUN
esrj-117639	76	3	incorporates	incorporate	VERB
esrj-117639	76	4	the	the	DET
esrj-117639	76	5	basic	basic	ADJ
esrj-117639	76	6	theory	theory	NOUN
esrj-117639	76	7	of	of	ADP
esrj-117639	76	8	the	the	DET
esrj-117639	76	9	gravity	gravity	NOUN
esrj-117639	76	10	field	field	NOUN
esrj-117639	76	11	into	into	ADP
esrj-117639	76	12	the	the	DET
esrj-117639	76	13	inversion	inversion	NOUN
esrj-117639	76	14	,	,	PUNCT
esrj-117639	76	15	optimizing	optimize	VERB
esrj-117639	76	16	by	by	ADP
esrj-117639	76	17	minimizing	minimize	VERB
esrj-117639	76	18	the	the	DET
esrj-117639	76	19	difference	difference	NOUN
esrj-117639	76	20	between	between	ADP
esrj-117639	76	21	the	the	DET
esrj-117639	76	22	observed	observe	VERB
esrj-117639	76	23	gravity	gravity	NOUN
esrj-117639	76	24	anomalies	anomaly	NOUN
esrj-117639	76	25	and	and	CCONJ
esrj-117639	76	26	the	the	DET
esrj-117639	76	27	reconstructed	reconstructed	ADJ
esrj-117639	76	28	gravity	gravity	NOUN
esrj-117639	76	29	anomalies	anomaly	NOUN
esrj-117639	76	30	,	,	PUNCT
esrj-117639	76	31	reducing	reduce	VERB
esrj-117639	76	32	the	the	DET
esrj-117639	76	33	uncertainty	uncertainty	NOUN
esrj-117639	76	34	of	of	ADP
esrj-117639	76	35	inversion	inversion	NOUN
esrj-117639	76	36	.	.	PUNCT
esrj-117639	77	1	zhang	zhang	PROPN
esrj-117639	77	2	et	et	PROPN
esrj-117639	77	3	al	al	PROPN
esrj-117639	77	4	.	.	PROPN
esrj-117639	77	5	(	(	PUNCT
esrj-117639	77	6	2022	2022	NUM
esrj-117639	77	7	)	)	PUNCT
esrj-117639	77	8	established	establish	VERB
esrj-117639	77	9	a	a	DET
esrj-117639	77	10	new	new	ADJ
esrj-117639	77	11	convolutional	convolutional	ADJ
esrj-117639	77	12	neural	neural	ADJ
esrj-117639	77	13	network	network	NOUN
esrj-117639	77	14	-	-	PUNCT
esrj-117639	77	15	decnet	decnet	NOUN
esrj-117639	77	16	for	for	ADP
esrj-117639	77	17	gravity	gravity	NOUN
esrj-117639	77	18	data	datum	NOUN
esrj-117639	77	19	inversion	inversion	NOUN
esrj-117639	77	20	,	,	PUNCT
esrj-117639	77	21	transforming	transform	VERB
esrj-117639	77	22	the	the	DET
esrj-117639	77	23	problem	problem	NOUN
esrj-117639	77	24	of	of	ADP
esrj-117639	77	25	mapping	mapping	NOUN
esrj-117639	77	26	from	from	ADP
esrj-117639	77	27	two	two	NUM
esrj-117639	77	28	dimensions	dimension	NOUN
esrj-117639	77	29	to	to	ADP
esrj-117639	77	30	three	three	NUM
esrj-117639	77	31	dimensions	dimension	NOUN
esrj-117639	77	32	into	into	ADP
esrj-117639	77	33	four	four	NUM
esrj-117639	77	34	problems	problem	NOUN
esrj-117639	77	35	of	of	ADP
esrj-117639	77	36	mapping	mapping	NOUN
esrj-117639	77	37	from	from	ADP
esrj-117639	77	38	two	two	NUM
esrj-117639	77	39	dimensions	dimension	NOUN
esrj-117639	77	40	to	to	ADP
esrj-117639	77	41	two	two	NUM
esrj-117639	77	42	dimensions	dimension	NOUN
esrj-117639	77	43	.	.	PUNCT
esrj-117639	78	1	this	this	DET
esrj-117639	78	2	method	method	NOUN
esrj-117639	78	3	has	have	VERB
esrj-117639	78	4	the	the	DET
esrj-117639	78	5	advantages	advantage	NOUN
esrj-117639	78	6	of	of	ADP
esrj-117639	78	7	high	high	ADJ
esrj-117639	78	8	accuracy	accuracy	NOUN
esrj-117639	78	9	,	,	PUNCT
esrj-117639	78	10	low	low	ADJ
esrj-117639	78	11	memory	memory	NOUN
esrj-117639	78	12	requirements	requirement	NOUN
esrj-117639	78	13	,	,	PUNCT
esrj-117639	78	14	and	and	CCONJ
esrj-117639	78	15	fast	fast	ADJ
esrj-117639	78	16	convergence	convergence	NOUN
esrj-117639	78	17	.	.	PUNCT
esrj-117639	79	1	this	this	DET
esrj-117639	79	2	article	article	NOUN
esrj-117639	79	3	presents	present	VERB
esrj-117639	79	4	a	a	DET
esrj-117639	79	5	study	study	NOUN
esrj-117639	79	6	on	on	ADP
esrj-117639	79	7	the	the	DET
esrj-117639	79	8	inversion	inversion	NOUN
esrj-117639	79	9	of	of	ADP
esrj-117639	79	10	3d	3d	NUM
esrj-117639	79	11	gravity	gravity	NOUN
esrj-117639	79	12	interfaces	interface	NOUN
esrj-117639	79	13	by	by	ADP
esrj-117639	79	14	combining	combine	VERB
esrj-117639	79	15	the	the	DET
esrj-117639	79	16	parker	parker	PROPN
esrj-117639	79	17	formula	formula	NOUN
esrj-117639	79	18	with	with	ADP
esrj-117639	79	19	artificial	artificial	ADJ
esrj-117639	79	20	neural	neural	ADJ
esrj-117639	79	21	networks	network	NOUN
esrj-117639	79	22	,	,	PUNCT
esrj-117639	79	23	and	and	CCONJ
esrj-117639	79	24	utilizing	utilize	VERB
esrj-117639	79	25	the	the	DET
esrj-117639	79	26	discrete	discrete	ADJ
esrj-117639	79	27	cosine	cosine	NOUN
esrj-117639	79	28	transform	transform	NOUN
esrj-117639	79	29	algorithm	algorithm	NOUN
esrj-117639	79	30	.	.	PUNCT
esrj-117639	80	1	the	the	DET
esrj-117639	80	2	article	article	NOUN
esrj-117639	80	3	begins	begin	VERB
esrj-117639	80	4	with	with	ADP
esrj-117639	80	5	a	a	DET
esrj-117639	80	6	review	review	NOUN
esrj-117639	80	7	of	of	ADP
esrj-117639	80	8	the	the	DET
esrj-117639	80	9	classic	classic	ADJ
esrj-117639	80	10	parker	parker	PROPN
esrj-117639	80	11	-	-	PUNCT
esrj-117639	80	12	oldenburg	oldenburg	PROPN
esrj-117639	80	13	formula	formula	NOUN
esrj-117639	80	14	,	,	PUNCT
esrj-117639	80	15	followed	follow	VERB
esrj-117639	80	16	by	by	ADP
esrj-117639	80	17	an	an	DET
esrj-117639	80	18	explanation	explanation	NOUN
esrj-117639	80	19	of	of	ADP
esrj-117639	80	20	the	the	DET
esrj-117639	80	21	artificial	artificial	ADJ
esrj-117639	80	22	neural	neural	ADJ
esrj-117639	80	23	network	network	NOUN
esrj-117639	80	24	algorithm	algorithm	NOUN
esrj-117639	80	25	process	process	NOUN
esrj-117639	80	26	and	and	CCONJ
esrj-117639	80	27	the	the	DET
esrj-117639	80	28	principles	principle	NOUN
esrj-117639	80	29	of	of	ADP
esrj-117639	80	30	the	the	DET
esrj-117639	80	31	discrete	discrete	ADJ
esrj-117639	80	32	cosine	cosine	NOUN
esrj-117639	80	33	transform	transform	NOUN
esrj-117639	80	34	algorithm	algorithm	NOUN
esrj-117639	80	35	.	.	PUNCT
esrj-117639	81	1	the	the	DET
esrj-117639	81	2	effectiveness	effectiveness	NOUN
esrj-117639	81	3	of	of	ADP
esrj-117639	81	4	the	the	DET
esrj-117639	81	5	proposed	propose	VERB
esrj-117639	81	6	algorithm	algorithm	NOUN
esrj-117639	81	7	is	be	AUX
esrj-117639	81	8	then	then	ADV
esrj-117639	81	9	demonstrated	demonstrate	VERB
esrj-117639	81	10	through	through	ADP
esrj-117639	81	11	synthetic	synthetic	ADJ
esrj-117639	81	12	and	and	CCONJ
esrj-117639	81	13	measured	measured	ADJ
esrj-117639	81	14	examples	example	NOUN
esrj-117639	81	15	.	.	PUNCT
esrj-117639	82	1	finally	finally	ADV
esrj-117639	82	2	,	,	PUNCT
esrj-117639	82	3	the	the	DET
esrj-117639	82	4	article	article	NOUN
esrj-117639	82	5	concludes	conclude	VERB
esrj-117639	82	6	with	with	ADP
esrj-117639	82	7	a	a	DET
esrj-117639	82	8	summary	summary	NOUN
esrj-117639	82	9	of	of	ADP
esrj-117639	82	10	the	the	DET
esrj-117639	82	11	algorithm	algorithm	NOUN
esrj-117639	82	12	and	and	CCONJ
esrj-117639	82	13	suggests	suggest	VERB
esrj-117639	82	14	possible	possible	ADJ
esrj-117639	82	15	improvements	improvement	NOUN
esrj-117639	82	16	.	.	PUNCT
esrj-117639	83	1	2	2	X
esrj-117639	83	2	.	.	X
esrj-117639	83	3	methods	method	NOUN
esrj-117639	83	4	2.1	2.1	NUM
esrj-117639	83	5	parker	parker	PROPN
esrj-117639	83	6	-	-	PUNCT
esrj-117639	83	7	oldenburg	oldenburg	PROPN
esrj-117639	83	8	formula	formula	NOUN
esrj-117639	83	9	the	the	DET
esrj-117639	83	10	parker	parker	PROPN
esrj-117639	83	11	formula	formula	NOUN
esrj-117639	83	12	,	,	PUNCT
esrj-117639	83	13	derived	derive	VERB
esrj-117639	83	14	by	by	ADP
esrj-117639	83	15	parker	parker	PROPN
esrj-117639	83	16	in	in	ADP
esrj-117639	83	17	1973	1973	NUM
esrj-117639	83	18	and	and	CCONJ
esrj-117639	83	19	1974	1974	NUM
esrj-117639	83	20	,	,	PUNCT
esrj-117639	83	21	provides	provide	VERB
esrj-117639	83	22	a	a	DET
esrj-117639	83	23	fourier	fourier	NOUN
esrj-117639	83	24	transform	transform	NOUN
esrj-117639	83	25	for	for	ADP
esrj-117639	83	26	calculating	calculate	VERB
esrj-117639	83	27	gravity	gravity	NOUN
esrj-117639	83	28	anomalies	anomaly	NOUN
esrj-117639	83	29	caused	cause	VERB
esrj-117639	83	30	by	by	ADP
esrj-117639	83	31	a	a	DET
esrj-117639	83	32	subsurface	subsurface	NOUN
esrj-117639	83	33	density	density	NOUN
esrj-117639	83	34	distribution	distribution	NOUN
esrj-117639	83	35	.	.	PUNCT
esrj-117639	84	1	the	the	DET
esrj-117639	84	2	equation	equation	NOUN
esrj-117639	84	3	1	1	NUM
esrj-117639	84	4	is	be	AUX
esrj-117639	84	5	expressed	express	VERB
esrj-117639	84	6	as	as	SCONJ
esrj-117639	84	7	follows	follow	VERB
esrj-117639	84	8	:	:	PUNCT
esrj-117639	84	9	[	[	PUNCT
esrj-117639	84	10	]	]	X
esrj-117639	84	11	=	=	SYM
esrj-117639	84	12	2	2	NUM
esrj-117639	84	13	(	(	PUNCT
esrj-117639	84	14	−|	−|	NOUN
esrj-117639	84	15	⃗	⃗	PROPN
esrj-117639	84	16	|	|	NOUN
esrj-117639	84	17	0	0	NUM
esrj-117639	84	18	)	)	PUNCT
esrj-117639	84	19	∑	∑	PUNCT
esrj-117639	84	20	∞	∞	PROPN
esrj-117639	84	21	|	|	ADV
esrj-117639	84	22	⃗	⃗	PROPN
esrj-117639	84	23	|	|	NOUN
esrj-117639	84	24	!	!	PUNCT
esrj-117639	85	1	[	[	X
esrj-117639	85	2	ℎ	ℎ	X
esrj-117639	85	3	(	(	PUNCT
esrj-117639	85	4	)	)	PUNCT
esrj-117639	85	5	]	]	PUNCT
esrj-117639	85	6	(	(	PUNCT
esrj-117639	85	7	1	1	X
esrj-117639	85	8	)	)	PUNCT
esrj-117639	85	9	here	here	ADV
esrj-117639	85	10	,	,	PUNCT
esrj-117639	85	11	f	f	PROPN
esrj-117639	85	12	is	be	AUX
esrj-117639	85	13	the	the	DET
esrj-117639	85	14	fourier	fourier	NOUN
esrj-117639	85	15	transform	transform	NOUN
esrj-117639	85	16	operator	operator	NOUN
esrj-117639	85	17	,	,	PUNCT
esrj-117639	85	18	g	g	PROPN
esrj-117639	85	19	is	be	AUX
esrj-117639	85	20	the	the	DET
esrj-117639	85	21	universal	universal	ADJ
esrj-117639	85	22	gravitational	gravitational	ADJ
esrj-117639	85	23	constant	constant	ADJ
esrj-117639	85	24	,	,	PUNCT
esrj-117639	85	25	ρ	ρ	PROPN
esrj-117639	85	26	is	be	AUX
esrj-117639	85	27	the	the	DET
esrj-117639	85	28	density	density	NOUN
esrj-117639	85	29	contrast	contrast	NOUN
esrj-117639	85	30	on	on	ADP
esrj-117639	85	31	the	the	DET
esrj-117639	85	32	interface	interface	NOUN
esrj-117639	85	33	,	,	PUNCT
esrj-117639	85	34	k→	k→	X
esrj-117639	85	35	is	be	AUX
esrj-117639	85	36	the	the	DET
esrj-117639	85	37	wave	wave	NOUN
esrj-117639	85	38	vector	vector	NOUN
esrj-117639	85	39	,	,	PUNCT
esrj-117639	85	40	h(r→	h(r→	NUM
esrj-117639	85	41	)	)	PUNCT
esrj-117639	85	42	is	be	AUX
esrj-117639	85	43	the	the	DET
esrj-117639	85	44	interface	interface	NOUN
esrj-117639	85	45	depth	depth	NOUN
esrj-117639	85	46	,	,	PUNCT
esrj-117639	85	47	and	and	CCONJ
esrj-117639	85	48	z0	z0	PROPN
esrj-117639	85	49	is	be	AUX
esrj-117639	85	50	the	the	DET
esrj-117639	85	51	average	average	ADJ
esrj-117639	85	52	depth	depth	NOUN
esrj-117639	85	53	of	of	ADP
esrj-117639	85	54	the	the	DET
esrj-117639	85	55	horizontal	horizontal	ADJ
esrj-117639	85	56	interface	interface	NOUN
esrj-117639	85	57	.	.	PUNCT
esrj-117639	86	1	oldenburg	oldenburg	PROPN
esrj-117639	86	2	rearranged	rearrange	VERB
esrj-117639	86	3	the	the	DET
esrj-117639	86	4	formula	formula	NOUN
esrj-117639	86	5	in	in	ADP
esrj-117639	86	6	1974	1974	NUM
esrj-117639	86	7	,	,	PUNCT
esrj-117639	86	8	and	and	CCONJ
esrj-117639	86	9	through	through	ADP
esrj-117639	86	10	an	an	DET
esrj-117639	86	11	iterative	iterative	NOUN
esrj-117639	86	12	process	process	NOUN
esrj-117639	86	13	,	,	PUNCT
esrj-117639	86	14	the	the	DET
esrj-117639	86	15	depth	depth	NOUN
esrj-117639	86	16	of	of	ADP
esrj-117639	86	17	the	the	DET
esrj-117639	86	18	undulating	undulating	NOUN
esrj-117639	86	19	interface	interface	NOUN
esrj-117639	86	20	can	can	AUX
esrj-117639	86	21	be	be	AUX
esrj-117639	86	22	obtained	obtain	VERB
esrj-117639	86	23	as	as	ADP
esrj-117639	86	24	follows	follow	VERB
esrj-117639	86	25	:	:	PUNCT
esrj-117639	86	26	423three	423three	ADJ
esrj-117639	86	27	-	-	PUNCT
esrj-117639	86	28	dimensional	dimensional	ADJ
esrj-117639	86	29	gravity	gravity	NOUN
esrj-117639	86	30	interface	interface	NOUN
esrj-117639	86	31	inversion	inversion	NOUN
esrj-117639	86	32	based	base	VERB
esrj-117639	86	33	on	on	ADP
esrj-117639	86	34	artificial	artificial	ADJ
esrj-117639	86	35	neural	neural	ADJ
esrj-117639	86	36	network	network	NOUN
esrj-117639	86	37	and	and	CCONJ
esrj-117639	86	38	discrete	discrete	ADJ
esrj-117639	86	39	cosine	cosine	NOUN
esrj-117639	86	40	transform	transform	NOUN
esrj-117639	86	41	algorithm	algorithm	NOUN
esrj-117639	86	42	[	[	X
esrj-117639	86	43	ℎ	ℎ	X
esrj-117639	86	44	(	(	PUNCT
esrj-117639	86	45	)	)	PUNCT
esrj-117639	86	46	]	]	PUNCT
esrj-117639	87	1	=	=	PUNCT
esrj-117639	87	2	−	−	PROPN
esrj-117639	87	3	[	[	PUNCT
esrj-117639	87	4	(	(	PUNCT
esrj-117639	87	5	)	)	PUNCT
esrj-117639	87	6	]	]	X
esrj-117639	87	7	e	e	X
esrj-117639	87	8	(	(	PUNCT
esrj-117639	87	9	)	)	PUNCT
esrj-117639	87	10	2	2	NUM
esrj-117639	87	11	−	−	NOUN
esrj-117639	87	12	∑	∑	PUNCT
esrj-117639	87	13	∞	∞	PROPN
esrj-117639	87	14	!	!	PUNCT
esrj-117639	88	1	[	[	X
esrj-117639	88	2	ℎ	ℎ	X
esrj-117639	88	3	(	(	PUNCT
esrj-117639	88	4	)	)	PUNCT
esrj-117639	88	5	]	]	PUNCT
esrj-117639	88	6	(	(	PUNCT
esrj-117639	88	7	2	2	X
esrj-117639	88	8	)	)	PUNCT
esrj-117639	88	9	in	in	ADP
esrj-117639	88	10	this	this	DET
esrj-117639	88	11	inversion	inversion	NOUN
esrj-117639	88	12	process	process	NOUN
esrj-117639	88	13	,	,	PUNCT
esrj-117639	88	14	we	we	PRON
esrj-117639	88	15	need	need	VERB
esrj-117639	88	16	to	to	PART
esrj-117639	88	17	assume	assume	VERB
esrj-117639	88	18	the	the	DET
esrj-117639	88	19	average	average	ADJ
esrj-117639	88	20	depth	depth	NOUN
esrj-117639	88	21	z0	z0	NOUN
esrj-117639	88	22	of	of	ADP
esrj-117639	88	23	the	the	DET
esrj-117639	88	24	interface	interface	NOUN
esrj-117639	88	25	and	and	CCONJ
esrj-117639	88	26	the	the	DET
esrj-117639	88	27	density	density	NOUN
esrj-117639	88	28	contrast	contrast	NOUN
esrj-117639	88	29	ρ	ρ	X
esrj-117639	88	30	.	.	PUNCT
esrj-117639	89	1	first	first	ADV
esrj-117639	89	2	,	,	PUNCT
esrj-117639	89	3	we	we	PRON
esrj-117639	89	4	substitute	substitute	VERB
esrj-117639	89	5	h(x	h(x	PROPN
esrj-117639	89	6	)	)	PUNCT
esrj-117639	90	1	=	=	SYM
esrj-117639	90	2	0	0	PUNCT
esrj-117639	90	3	into	into	ADP
esrj-117639	90	4	the	the	DET
esrj-117639	90	5	second	second	ADJ
esrj-117639	90	6	term	term	NOUN
esrj-117639	90	7	on	on	ADP
esrj-117639	90	8	the	the	DET
esrj-117639	90	9	right	right	ADJ
esrj-117639	90	10	-	-	PUNCT
esrj-117639	90	11	hand	hand	NOUN
esrj-117639	90	12	side	side	NOUN
esrj-117639	90	13	of	of	ADP
esrj-117639	90	14	equation	equation	NOUN
esrj-117639	90	15	(	(	PUNCT
esrj-117639	90	16	2	2	NUM
esrj-117639	90	17	)	)	PUNCT
esrj-117639	90	18	and	and	CCONJ
esrj-117639	90	19	calculate	calculate	VERB
esrj-117639	90	20	the	the	DET
esrj-117639	90	21	first	first	ADJ
esrj-117639	90	22	term	term	NOUN
esrj-117639	90	23	on	on	ADP
esrj-117639	90	24	the	the	DET
esrj-117639	90	25	right	right	ADJ
esrj-117639	90	26	-	-	PUNCT
esrj-117639	90	27	hand	hand	NOUN
esrj-117639	90	28	side	side	NOUN
esrj-117639	90	29	.	.	PUNCT
esrj-117639	91	1	then	then	ADV
esrj-117639	91	2	,	,	PUNCT
esrj-117639	91	3	we	we	PRON
esrj-117639	91	4	perform	perform	VERB
esrj-117639	91	5	the	the	DET
esrj-117639	91	6	inverse	inverse	NOUN
esrj-117639	91	7	fourier	fourier	NOUN
esrj-117639	91	8	transform	transform	NOUN
esrj-117639	91	9	to	to	PART
esrj-117639	91	10	obtain	obtain	VERB
esrj-117639	91	11	the	the	DET
esrj-117639	91	12	first	first	ADJ
esrj-117639	91	13	approximate	approximate	ADJ
esrj-117639	91	14	value	value	NOUN
esrj-117639	91	15	of	of	ADP
esrj-117639	91	16	h(x	h(x	PROPN
esrj-117639	91	17	)	)	PUNCT
esrj-117639	91	18	.	.	PUNCT
esrj-117639	92	1	we	we	PRON
esrj-117639	92	2	then	then	ADV
esrj-117639	92	3	substitute	substitute	VERB
esrj-117639	92	4	this	this	DET
esrj-117639	92	5	value	value	NOUN
esrj-117639	92	6	into	into	ADP
esrj-117639	92	7	the	the	DET
esrj-117639	92	8	right	right	ADJ
esrj-117639	92	9	-	-	PUNCT
esrj-117639	92	10	hand	hand	NOUN
esrj-117639	92	11	side	side	NOUN
esrj-117639	92	12	of	of	ADP
esrj-117639	92	13	equation	equation	NOUN
esrj-117639	92	14	(	(	PUNCT
esrj-117639	92	15	2	2	NUM
esrj-117639	92	16	)	)	PUNCT
esrj-117639	92	17	and	and	CCONJ
esrj-117639	92	18	iteratively	iteratively	ADV
esrj-117639	92	19	solve	solve	VERB
esrj-117639	92	20	until	until	SCONJ
esrj-117639	92	21	a	a	DET
esrj-117639	92	22	suitable	suitable	ADJ
esrj-117639	92	23	solution	solution	NOUN
esrj-117639	92	24	is	be	AUX
esrj-117639	92	25	obtained	obtain	VERB
esrj-117639	92	26	.	.	PUNCT
esrj-117639	93	1	2.2	2.2	NUM
esrj-117639	93	2	artificial	artificial	ADJ
esrj-117639	93	3	neural	neural	ADJ
esrj-117639	93	4	networks	network	NOUN
esrj-117639	93	5	in	in	ADP
esrj-117639	93	6	geophysics	geophysic	NOUN
esrj-117639	93	7	,	,	PUNCT
esrj-117639	93	8	given	give	VERB
esrj-117639	93	9	an	an	DET
esrj-117639	93	10	underground	underground	ADJ
esrj-117639	93	11	model	model	NOUN
esrj-117639	93	12	and	and	CCONJ
esrj-117639	93	13	a	a	DET
esrj-117639	93	14	forward	forward	ADJ
esrj-117639	93	15	function	function	NOUN
esrj-117639	93	16	,	,	PUNCT
esrj-117639	93	17	the	the	DET
esrj-117639	93	18	corresponding	corresponding	ADJ
esrj-117639	93	19	geophysical	geophysical	ADJ
esrj-117639	93	20	data	datum	NOUN
esrj-117639	93	21	can	can	AUX
esrj-117639	93	22	be	be	AUX
esrj-117639	93	23	represented	represent	VERB
esrj-117639	93	24	by	by	ADP
esrj-117639	93	25	the	the	DET
esrj-117639	93	26	equation	equation	NOUN
esrj-117639	93	27	3	3	NUM
esrj-117639	93	28	:	:	PUNCT
esrj-117639	93	29	d	d	NOUN
esrj-117639	93	30	=	=	PUNCT
esrj-117639	93	31	g(m	g(m	PROPN
esrj-117639	93	32	)	)	PUNCT
esrj-117639	93	33	(	(	PUNCT
esrj-117639	93	34	3	3	X
esrj-117639	93	35	)	)	PUNCT
esrj-117639	93	36	the	the	DET
esrj-117639	93	37	inversion	inversion	NOUN
esrj-117639	93	38	process	process	NOUN
esrj-117639	93	39	is	be	AUX
esrj-117639	93	40	to	to	PART
esrj-117639	93	41	find	find	VERB
esrj-117639	93	42	an	an	DET
esrj-117639	93	43	optimal	optimal	ADJ
esrj-117639	93	44	model	model	NOUN
esrj-117639	93	45	whose	whose	DET
esrj-117639	93	46	forward	forward	ADJ
esrj-117639	93	47	response	response	NOUN
esrj-117639	93	48	data	datum	NOUN
esrj-117639	93	49	can	can	AUX
esrj-117639	93	50	best	well	ADV
esrj-117639	93	51	fit	fit	VERB
esrj-117639	93	52	the	the	DET
esrj-117639	93	53	observed	observe	VERB
esrj-117639	93	54	data	data	NOUN
esrj-117639	93	55	dobs	dob	NOUN
esrj-117639	93	56	.	.	PUNCT
esrj-117639	94	1	artificial	artificial	ADJ
esrj-117639	94	2	neural	neural	ADJ
esrj-117639	94	3	networks	network	NOUN
esrj-117639	94	4	consist	consist	VERB
esrj-117639	94	5	of	of	ADP
esrj-117639	94	6	network	network	NOUN
esrj-117639	94	7	units	unit	NOUN
esrj-117639	94	8	called	call	VERB
esrj-117639	94	9	neurons	neuron	NOUN
esrj-117639	94	10	,	,	PUNCT
esrj-117639	94	11	which	which	PRON
esrj-117639	94	12	are	be	AUX
esrj-117639	94	13	arranged	arrange	VERB
esrj-117639	94	14	in	in	ADP
esrj-117639	94	15	layers	layer	NOUN
esrj-117639	94	16	.	.	PUNCT
esrj-117639	95	1	generally	generally	ADV
esrj-117639	95	2	,	,	PUNCT
esrj-117639	95	3	there	there	PRON
esrj-117639	95	4	are	be	VERB
esrj-117639	95	5	input	input	ADJ
esrj-117639	95	6	layers	layer	NOUN
esrj-117639	95	7	,	,	PUNCT
esrj-117639	95	8	hidden	hidden	ADJ
esrj-117639	95	9	layers	layer	NOUN
esrj-117639	95	10	,	,	PUNCT
esrj-117639	95	11	and	and	CCONJ
esrj-117639	95	12	output	output	NOUN
esrj-117639	95	13	layers	layer	NOUN
esrj-117639	95	14	.	.	PUNCT
esrj-117639	96	1	when	when	SCONJ
esrj-117639	96	2	the	the	DET
esrj-117639	96	3	neural	neural	ADJ
esrj-117639	96	4	network	network	NOUN
esrj-117639	96	5	is	be	AUX
esrj-117639	96	6	fully	fully	ADV
esrj-117639	96	7	connected	connect	VERB
esrj-117639	96	8	,	,	PUNCT
esrj-117639	96	9	each	each	DET
esrj-117639	96	10	neuron	neuron	NOUN
esrj-117639	96	11	in	in	ADP
esrj-117639	96	12	the	the	DET
esrj-117639	96	13	input	input	NOUN
esrj-117639	96	14	layer	layer	NOUN
esrj-117639	96	15	is	be	AUX
esrj-117639	96	16	connected	connect	VERB
esrj-117639	96	17	to	to	ADP
esrj-117639	96	18	any	any	DET
esrj-117639	96	19	neuron	neuron	NOUN
esrj-117639	96	20	in	in	ADP
esrj-117639	96	21	the	the	DET
esrj-117639	96	22	hidden	hide	VERB
esrj-117639	96	23	layer	layer	NOUN
esrj-117639	96	24	,	,	PUNCT
esrj-117639	96	25	and	and	CCONJ
esrj-117639	96	26	each	each	DET
esrj-117639	96	27	neuron	neuron	NOUN
esrj-117639	96	28	in	in	ADP
esrj-117639	96	29	the	the	DET
esrj-117639	96	30	hidden	hide	VERB
esrj-117639	96	31	layer	layer	NOUN
esrj-117639	96	32	is	be	AUX
esrj-117639	96	33	also	also	ADV
esrj-117639	96	34	connected	connect	VERB
esrj-117639	96	35	to	to	ADP
esrj-117639	96	36	any	any	DET
esrj-117639	96	37	neuron	neuron	NOUN
esrj-117639	96	38	in	in	ADP
esrj-117639	96	39	the	the	DET
esrj-117639	96	40	output	output	NOUN
esrj-117639	96	41	layer	layer	NOUN
esrj-117639	96	42	.	.	PUNCT
esrj-117639	97	1	the	the	DET
esrj-117639	97	2	connection	connection	NOUN
esrj-117639	97	3	method	method	NOUN
esrj-117639	97	4	is	be	AUX
esrj-117639	97	5	expressed	express	VERB
esrj-117639	97	6	in	in	ADP
esrj-117639	97	7	a	a	DET
esrj-117639	97	8	mathematical	mathematical	ADJ
esrj-117639	97	9	expression	expression	NOUN
esrj-117639	97	10	of	of	ADP
esrj-117639	97	11	weighted	weight	VERB
esrj-117639	97	12	sum	sum	NOUN
esrj-117639	97	13	,	,	PUNCT
esrj-117639	97	14	and	and	CCONJ
esrj-117639	97	15	is	be	AUX
esrj-117639	97	16	subjected	subject	VERB
esrj-117639	97	17	to	to	ADP
esrj-117639	97	18	the	the	DET
esrj-117639	97	19	action	action	NOUN
esrj-117639	97	20	of	of	ADP
esrj-117639	97	21	an	an	DET
esrj-117639	97	22	activation	activation	NOUN
esrj-117639	97	23	function	function	NOUN
esrj-117639	97	24	in	in	ADP
esrj-117639	97	25	the	the	DET
esrj-117639	97	26	hidden	hide	VERB
esrj-117639	97	27	layer	layer	NOUN
esrj-117639	97	28	.	.	PUNCT
esrj-117639	98	1	taking	take	VERB
esrj-117639	98	2	the	the	DET
esrj-117639	98	3	input	input	NOUN
esrj-117639	98	4	vector	vector	NOUN
esrj-117639	98	5	xn	xn	PROPN
esrj-117639	98	6	as	as	ADP
esrj-117639	98	7	an	an	DET
esrj-117639	98	8	example	example	NOUN
esrj-117639	98	9	,	,	PUNCT
esrj-117639	98	10	the	the	DET
esrj-117639	98	11	mathematical	mathematical	ADJ
esrj-117639	98	12	expression	expression	NOUN
esrj-117639	98	13	an	an	DET
esrj-117639	98	14	m	m	NOUN
esrj-117639	98	15	corresponding	correspond	VERB
esrj-117639	98	16	to	to	ADP
esrj-117639	98	17	the	the	DET
esrj-117639	98	18	mth	mth	NOUN
esrj-117639	98	19	neuron	neuron	NOUN
esrj-117639	98	20	in	in	ADP
esrj-117639	98	21	the	the	DET
esrj-117639	98	22	hidden	hide	VERB
esrj-117639	98	23	layer	layer	NOUN
esrj-117639	98	24	is	be	AUX
esrj-117639	98	25	:	:	PUNCT
esrj-117639	98	26	=	=	SYM
esrj-117639	98	27	(	(	PUNCT
esrj-117639	98	28	h	h	NOUN
esrj-117639	98	29	=	=	NOUN
esrj-117639	98	30	1	1	NUM
esrj-117639	98	31	+	+	CCONJ
esrj-117639	98	32	)	)	PUNCT
esrj-117639	98	33	(	(	PUNCT
esrj-117639	98	34	4	4	X
esrj-117639	98	35	)	)	PUNCT
esrj-117639	98	36	where	where	SCONJ
esrj-117639	98	37	f	f	PROPN
esrj-117639	98	38	represents	represent	VERB
esrj-117639	98	39	the	the	DET
esrj-117639	98	40	activation	activation	NOUN
esrj-117639	98	41	function	function	NOUN
esrj-117639	98	42	,	,	PUNCT
esrj-117639	98	43	m	m	VERB
esrj-117639	98	44	=	=	NOUN
esrj-117639	98	45	1,2	1,2	NUM
esrj-117639	98	46	,	,	PUNCT
esrj-117639	98	47	...	...	PUNCT
esrj-117639	98	48	,	,	PUNCT
esrj-117639	98	49	l	l	NOUN
esrj-117639	98	50	,	,	PUNCT
esrj-117639	98	51	the	the	DET
esrj-117639	98	52	superscript	superscript	PROPN
esrj-117639	98	53	h	h	NOUN
esrj-117639	98	54	represents	represent	VERB
esrj-117639	98	55	the	the	DET
esrj-117639	98	56	hidden	hide	VERB
esrj-117639	98	57	layer	layer	NOUN
esrj-117639	98	58	,	,	PUNCT
esrj-117639	98	59	ωh	ωh	NUM
esrj-117639	98	60	mi	mi	PROPN
esrj-117639	98	61	represents	represent	VERB
esrj-117639	98	62	the	the	DET
esrj-117639	98	63	weighting	weighting	NOUN
esrj-117639	98	64	coefficient	coefficient	NOUN
esrj-117639	98	65	of	of	ADP
esrj-117639	98	66	the	the	DET
esrj-117639	98	67	ith	ith	PROPN
esrj-117639	98	68	neuron	neuron	NOUN
esrj-117639	98	69	in	in	ADP
esrj-117639	98	70	the	the	DET
esrj-117639	98	71	input	input	NOUN
esrj-117639	98	72	layer	layer	NOUN
esrj-117639	98	73	for	for	ADP
esrj-117639	98	74	the	the	DET
esrj-117639	98	75	m	m	PROPN
esrj-117639	98	76	neuron	neuron	NOUN
esrj-117639	98	77	in	in	ADP
esrj-117639	98	78	the	the	DET
esrj-117639	98	79	hidden	hide	VERB
esrj-117639	98	80	layer	layer	NOUN
esrj-117639	98	81	,	,	PUNCT
esrj-117639	98	82	xn	xn	PROPN
esrj-117639	99	1	i	i	PRON
esrj-117639	99	2	represents	represent	VERB
esrj-117639	99	3	the	the	DET
esrj-117639	99	4	ith	ith	PROPN
esrj-117639	99	5	component	component	NOUN
esrj-117639	99	6	of	of	ADP
esrj-117639	99	7	the	the	DET
esrj-117639	99	8	input	input	NOUN
esrj-117639	99	9	vector	vector	NOUN
esrj-117639	99	10	xn	xn	PROPN
esrj-117639	99	11	,	,	PUNCT
esrj-117639	99	12	and	and	CCONJ
esrj-117639	99	13	bm	bm	PROPN
esrj-117639	99	14	represents	represent	VERB
esrj-117639	99	15	the	the	DET
esrj-117639	99	16	bias	bias	NOUN
esrj-117639	99	17	term	term	NOUN
esrj-117639	99	18	of	of	ADP
esrj-117639	99	19	the	the	DET
esrj-117639	99	20	mth	mth	NOUN
esrj-117639	99	21	neuron	neuron	NOUN
esrj-117639	99	22	in	in	ADP
esrj-117639	99	23	the	the	DET
esrj-117639	99	24	hidden	hide	VERB
esrj-117639	99	25	layer	layer	NOUN
esrj-117639	99	26	.	.	PUNCT
esrj-117639	100	1	the	the	DET
esrj-117639	100	2	activation	activation	NOUN
esrj-117639	100	3	function	function	NOUN
esrj-117639	100	4	f	f	PROPN
esrj-117639	100	5	is	be	AUX
esrj-117639	100	6	a	a	DET
esrj-117639	100	7	bounded	bounded	ADJ
esrj-117639	100	8	nonlinear	nonlinear	ADJ
esrj-117639	100	9	function	function	NOUN
esrj-117639	100	10	,	,	PUNCT
esrj-117639	100	11	and	and	CCONJ
esrj-117639	100	12	the	the	DET
esrj-117639	100	13	sigmoid	sigmoid	NOUN
esrj-117639	100	14	function	function	NOUN
esrj-117639	100	15	is	be	AUX
esrj-117639	100	16	commonly	commonly	ADV
esrj-117639	100	17	used	use	VERB
esrj-117639	100	18	:	:	PUNCT
esrj-117639	100	19	(	(	PUNCT
esrj-117639	100	20	)	)	PUNCT
esrj-117639	100	21	=	=	SYM
esrj-117639	100	22	1	1	NUM
esrj-117639	100	23	1+exp	1+exp	NOUN
esrj-117639	100	24	(	(	PUNCT
esrj-117639	100	25	)	)	PUNCT
esrj-117639	100	26	(	(	PUNCT
esrj-117639	100	27	5	5	X
esrj-117639	100	28	)	)	PUNCT
esrj-117639	100	29	where	where	SCONJ
esrj-117639	100	30	controls	control	VERB
esrj-117639	100	31	the	the	DET
esrj-117639	100	32	rate	rate	NOUN
esrj-117639	100	33	at	at	ADP
esrj-117639	100	34	which	which	PRON
esrj-117639	100	35	the	the	DET
esrj-117639	100	36	function	function	NOUN
esrj-117639	100	37	value	value	NOUN
esrj-117639	100	38	tends	tend	VERB
esrj-117639	100	39	to	to	ADP
esrj-117639	100	40	0.5	0.5	NUM
esrj-117639	100	41	.	.	PUNCT
esrj-117639	101	1	the	the	DET
esrj-117639	101	2	activation	activation	NOUN
esrj-117639	101	3	function	function	NOUN
esrj-117639	101	4	endows	endow	VERB
esrj-117639	101	5	the	the	DET
esrj-117639	101	6	neural	neural	ADJ
esrj-117639	101	7	network	network	NOUN
esrj-117639	101	8	with	with	ADP
esrj-117639	101	9	highly	highly	ADV
esrj-117639	101	10	nonlinear	nonlinear	ADJ
esrj-117639	101	11	expressive	expressive	ADJ
esrj-117639	101	12	power	power	NOUN
esrj-117639	101	13	,	,	PUNCT
esrj-117639	101	14	making	make	VERB
esrj-117639	101	15	it	it	PRON
esrj-117639	101	16	suitable	suitable	ADJ
esrj-117639	101	17	for	for	ADP
esrj-117639	101	18	solving	solve	VERB
esrj-117639	101	19	nonlinear	nonlinear	ADJ
esrj-117639	101	20	problems	problem	NOUN
esrj-117639	101	21	.	.	PUNCT
esrj-117639	102	1	similar	similar	ADJ
esrj-117639	102	2	to	to	ADP
esrj-117639	102	3	equaion	equaion	NOUN
esrj-117639	102	4	(	(	PUNCT
esrj-117639	102	5	4	4	NUM
esrj-117639	102	6	)	)	PUNCT
esrj-117639	102	7	,	,	PUNCT
esrj-117639	102	8	the	the	DET
esrj-117639	102	9	kth	kth	PROPN
esrj-117639	102	10	neuron	neuron	NOUN
esrj-117639	102	11	on	on	ADP
esrj-117639	102	12	k	k	PROPN
esrj-117639	102	13	in	in	ADP
esrj-117639	102	14	the	the	DET
esrj-117639	102	15	output	output	NOUN
esrj-117639	102	16	layer	layer	NOUN
esrj-117639	102	17	can	can	AUX
esrj-117639	102	18	be	be	AUX
esrj-117639	102	19	represented	represent	VERB
esrj-117639	102	20	as	as	ADP
esrj-117639	102	21	:	:	PUNCT
esrj-117639	102	22	=	=	PUNCT
esrj-117639	102	23	∘	∘	X
esrj-117639	102	24	(	(	PUNCT
esrj-117639	102	25	∘	∘	X
esrj-117639	102	26	+	+	CCONJ
esrj-117639	102	27	)	)	PUNCT
esrj-117639	102	28	(	(	PUNCT
esrj-117639	102	29	6	6	NUM
esrj-117639	102	30	)	)	PUNCT
esrj-117639	102	31	where	where	SCONJ
esrj-117639	102	32	k	k	NOUN
esrj-117639	102	33	=	=	SYM
esrj-117639	102	34	1,2	1,2	NUM
esrj-117639	102	35	,	,	PUNCT
esrj-117639	102	36	...	...	PUNCT
esrj-117639	102	37	,	,	PUNCT
esrj-117639	102	38	m	m	PROPN
esrj-117639	102	39	,	,	PUNCT
esrj-117639	102	40	the	the	DET
esrj-117639	102	41	superscript	superscript	PROPN
esrj-117639	102	42	o	o	PROPN
esrj-117639	102	43	represents	represent	VERB
esrj-117639	102	44	the	the	DET
esrj-117639	102	45	output	output	NOUN
esrj-117639	102	46	layer	layer	NOUN
esrj-117639	102	47	,	,	PUNCT
esrj-117639	102	48	ωo	ωo	ADP
esrj-117639	102	49	km	km	NOUN
esrj-117639	102	50	represents	represent	VERB
esrj-117639	102	51	the	the	DET
esrj-117639	102	52	weighting	weighting	NOUN
esrj-117639	102	53	coefficient	coefficient	NOUN
esrj-117639	102	54	of	of	ADP
esrj-117639	102	55	the	the	DET
esrj-117639	102	56	mth	mth	NOUN
esrj-117639	102	57	neuron	neuron	NOUN
esrj-117639	102	58	in	in	ADP
esrj-117639	102	59	the	the	DET
esrj-117639	102	60	hidden	hide	VERB
esrj-117639	102	61	layer	layer	NOUN
esrj-117639	102	62	for	for	ADP
esrj-117639	102	63	the	the	DET
esrj-117639	102	64	kth	kth	PROPN
esrj-117639	102	65	neuron	neuron	PROPN
esrj-117639	102	66	in	in	ADP
esrj-117639	102	67	the	the	DET
esrj-117639	102	68	output	output	NOUN
esrj-117639	102	69	layer	layer	NOUN
esrj-117639	102	70	,	,	PUNCT
esrj-117639	102	71	and	and	CCONJ
esrj-117639	102	72	ck	ck	PROPN
esrj-117639	102	73	represents	represent	VERB
esrj-117639	102	74	the	the	DET
esrj-117639	102	75	bias	bias	NOUN
esrj-117639	102	76	term	term	NOUN
esrj-117639	102	77	of	of	ADP
esrj-117639	102	78	the	the	DET
esrj-117639	102	79	kth	kth	PROPN
esrj-117639	102	80	neuron	neuron	PROPN
esrj-117639	102	81	in	in	ADP
esrj-117639	102	82	the	the	DET
esrj-117639	102	83	output	output	NOUN
esrj-117639	102	84	layer	layer	NOUN
esrj-117639	102	85	.	.	PUNCT
esrj-117639	103	1	the	the	DET
esrj-117639	103	2	vector	vector	NOUN
esrj-117639	103	3	is	be	AUX
esrj-117639	103	4	the	the	DET
esrj-117639	103	5	expected	expect	VERB
esrj-117639	103	6	output	output	NOUN
esrj-117639	103	7	of	of	ADP
esrj-117639	103	8	the	the	DET
esrj-117639	103	9	neural	neural	ADJ
esrj-117639	103	10	network	network	NOUN
esrj-117639	103	11	,	,	PUNCT
esrj-117639	103	12	and	and	CCONJ
esrj-117639	103	13	there	there	PRON
esrj-117639	103	14	is	be	VERB
esrj-117639	103	15	an	an	DET
esrj-117639	103	16	error	error	NOUN
esrj-117639	103	17	between	between	ADP
esrj-117639	103	18	it	it	PRON
esrj-117639	103	19	and	and	CCONJ
esrj-117639	103	20	the	the	DET
esrj-117639	103	21	actual	actual	ADJ
esrj-117639	103	22	output	output	NOUN
esrj-117639	103	23	of	of	ADP
esrj-117639	103	24	the	the	DET
esrj-117639	103	25	neural	neural	ADJ
esrj-117639	103	26	network	network	NOUN
esrj-117639	103	27	:	:	PUNCT
esrj-117639	103	28	=	=	SYM
esrj-117639	103	29	1	1	NUM
esrj-117639	103	30	2	2	NUM
esrj-117639	103	31	(	(	PUNCT
esrj-117639	103	32	)	)	PUNCT
esrj-117639	103	33	2σ	2σ	PROPN
esrj-117639	104	1	σ	σ	X
esrj-117639	104	2	(	(	PUNCT
esrj-117639	104	3	7	7	X
esrj-117639	104	4	)	)	PUNCT
esrj-117639	104	5	the	the	DET
esrj-117639	104	6	training	training	NOUN
esrj-117639	104	7	of	of	ADP
esrj-117639	104	8	the	the	DET
esrj-117639	104	9	neural	neural	ADJ
esrj-117639	104	10	network	network	NOUN
esrj-117639	104	11	is	be	AUX
esrj-117639	104	12	an	an	DET
esrj-117639	104	13	optimization	optimization	NOUN
esrj-117639	104	14	process	process	NOUN
esrj-117639	104	15	with	with	ADP
esrj-117639	104	16	e	e	PROPN
esrj-117639	104	17	as	as	ADP
esrj-117639	104	18	the	the	DET
esrj-117639	104	19	objective	objective	ADJ
esrj-117639	104	20	function	function	NOUN
esrj-117639	104	21	.	.	PUNCT
esrj-117639	105	1	common	common	ADJ
esrj-117639	105	2	optimization	optimization	NOUN
esrj-117639	105	3	algorithms	algorithm	NOUN
esrj-117639	105	4	include	include	VERB
esrj-117639	105	5	levenbergmarquardt	levenbergmarquardt	ADJ
esrj-117639	105	6	algorithm	algorithm	NOUN
esrj-117639	105	7	,	,	PUNCT
esrj-117639	105	8	conjugate	conjugate	ADJ
esrj-117639	105	9	gradient	gradient	NOUN
esrj-117639	105	10	method	method	NOUN
esrj-117639	105	11	,	,	PUNCT
esrj-117639	105	12	simulated	simulate	VERB
esrj-117639	105	13	annealing	annealing	NOUN
esrj-117639	105	14	method	method	NOUN
esrj-117639	105	15	,	,	PUNCT
esrj-117639	105	16	and	and	CCONJ
esrj-117639	105	17	genetic	genetic	ADJ
esrj-117639	105	18	algorithm	algorithm	NOUN
esrj-117639	105	19	,	,	PUNCT
esrj-117639	105	20	etc	etc	X
esrj-117639	105	21	.	.	X
esrj-117639	105	22	2.3	2.3	NUM
esrj-117639	105	23	u	u	NOUN
esrj-117639	105	24	-	-	ADJ
esrj-117639	105	25	net	net	ADJ
esrj-117639	105	26	neural	neural	ADJ
esrj-117639	105	27	network	network	NOUN
esrj-117639	105	28	the	the	DET
esrj-117639	105	29	u	u	PROPN
esrj-117639	105	30	-	-	ADJ
esrj-117639	105	31	net	net	ADJ
esrj-117639	105	32	neural	neural	ADJ
esrj-117639	105	33	network	network	NOUN
esrj-117639	105	34	model	model	NOUN
esrj-117639	105	35	used	use	VERB
esrj-117639	105	36	in	in	ADP
esrj-117639	105	37	this	this	DET
esrj-117639	105	38	paper	paper	NOUN
esrj-117639	105	39	is	be	AUX
esrj-117639	105	40	shown	show	VERB
esrj-117639	105	41	in	in	ADP
esrj-117639	105	42	figure	figure	NOUN
esrj-117639	105	43	1	1	NUM
esrj-117639	105	44	.	.	PUNCT
esrj-117639	106	1	the	the	DET
esrj-117639	106	2	u	u	ADJ
esrj-117639	106	3	-	-	ADJ
esrj-117639	106	4	net	net	ADJ
esrj-117639	106	5	network	network	NOUN
esrj-117639	106	6	model	model	NOUN
esrj-117639	106	7	extracts	extract	NOUN
esrj-117639	106	8	detailed	detail	VERB
esrj-117639	106	9	features	feature	NOUN
esrj-117639	106	10	through	through	ADP
esrj-117639	106	11	a	a	DET
esrj-117639	106	12	down	down	ADV
esrj-117639	106	13	-	-	PUNCT
esrj-117639	106	14	sampling	sample	VERB
esrj-117639	106	15	process	process	NOUN
esrj-117639	106	16	and	and	CCONJ
esrj-117639	106	17	restores	restore	VERB
esrj-117639	106	18	dimensions	dimension	NOUN
esrj-117639	106	19	through	through	ADP
esrj-117639	106	20	an	an	DET
esrj-117639	106	21	up	up	ADV
esrj-117639	106	22	-	-	PUNCT
esrj-117639	106	23	sampling	sample	VERB
esrj-117639	106	24	process	process	NOUN
esrj-117639	106	25	.	.	PUNCT
esrj-117639	107	1	it	it	PRON
esrj-117639	107	2	concatenates	concatenate	VERB
esrj-117639	107	3	shallow	shallow	ADJ
esrj-117639	107	4	background	background	NOUN
esrj-117639	107	5	information	information	NOUN
esrj-117639	107	6	with	with	ADP
esrj-117639	107	7	deep	deep	ADJ
esrj-117639	107	8	local	local	ADJ
esrj-117639	107	9	data	datum	NOUN
esrj-117639	107	10	,	,	PUNCT
esrj-117639	107	11	ensuring	ensure	VERB
esrj-117639	107	12	that	that	SCONJ
esrj-117639	107	13	positional	positional	ADJ
esrj-117639	107	14	and	and	CCONJ
esrj-117639	107	15	other	other	ADJ
esrj-117639	107	16	shallow	shallow	ADJ
esrj-117639	107	17	layer	layer	NOUN
esrj-117639	107	18	information	information	NOUN
esrj-117639	107	19	is	be	AUX
esrj-117639	107	20	well	well	ADV
esrj-117639	107	21	retained	retain	VERB
esrj-117639	107	22	until	until	ADP
esrj-117639	107	23	the	the	DET
esrj-117639	107	24	data	data	NOUN
esrj-117639	107	25	output	output	NOUN
esrj-117639	107	26	,	,	PUNCT
esrj-117639	107	27	making	make	VERB
esrj-117639	107	28	the	the	DET
esrj-117639	107	29	network	network	NOUN
esrj-117639	107	30	more	more	ADV
esrj-117639	107	31	sensitive	sensitive	ADJ
esrj-117639	107	32	to	to	ADP
esrj-117639	107	33	positional	positional	ADJ
esrj-117639	107	34	information	information	NOUN
esrj-117639	107	35	(	(	PUNCT
esrj-117639	107	36	ronneberger	ronneberger	NOUN
esrj-117639	107	37	et	et	PROPN
esrj-117639	107	38	al	al	PROPN
esrj-117639	107	39	.	.	PROPN
esrj-117639	107	40	,	,	PUNCT
esrj-117639	107	41	2015	2015	NUM
esrj-117639	107	42	)	)	PUNCT
esrj-117639	107	43	.	.	PUNCT
esrj-117639	108	1	additionally	additionally	ADV
esrj-117639	108	2	,	,	PUNCT
esrj-117639	108	3	as	as	ADP
esrj-117639	108	4	a	a	DET
esrj-117639	108	5	fully	fully	ADV
esrj-117639	108	6	convolutional	convolutional	ADJ
esrj-117639	108	7	neural	neural	ADJ
esrj-117639	108	8	network	network	NOUN
esrj-117639	108	9	,	,	PUNCT
esrj-117639	108	10	u	u	NOUN
esrj-117639	108	11	-	-	ADJ
esrj-117639	108	12	net	net	ADJ
esrj-117639	108	13	abandons	abandon	NOUN
esrj-117639	108	14	complex	complex	ADJ
esrj-117639	108	15	fully	fully	ADV
esrj-117639	108	16	connected	connect	VERB
esrj-117639	108	17	layers	layer	NOUN
esrj-117639	108	18	and	and	CCONJ
esrj-117639	108	19	directly	directly	ADV
esrj-117639	108	20	outputs	output	VERB
esrj-117639	108	21	through	through	ADP
esrj-117639	108	22	a	a	DET
esrj-117639	108	23	1×1	1×1	ADJ
esrj-117639	108	24	convolutional	convolutional	ADJ
esrj-117639	108	25	layer	layer	NOUN
esrj-117639	108	26	,	,	PUNCT
esrj-117639	108	27	making	make	VERB
esrj-117639	108	28	the	the	DET
esrj-117639	108	29	network	network	NOUN
esrj-117639	108	30	parameters	parameter	NOUN
esrj-117639	108	31	simpler	simple	ADJ
esrj-117639	108	32	and	and	CCONJ
esrj-117639	108	33	easier	easy	ADJ
esrj-117639	108	34	to	to	PART
esrj-117639	108	35	train	train	VERB
esrj-117639	108	36	.	.	PUNCT
esrj-117639	109	1	in	in	ADP
esrj-117639	109	2	the	the	DET
esrj-117639	109	3	gravity	gravity	NOUN
esrj-117639	109	4	data	datum	NOUN
esrj-117639	109	5	density	density	NOUN
esrj-117639	109	6	interface	interface	NOUN
esrj-117639	109	7	inversion	inversion	NOUN
esrj-117639	109	8	,	,	PUNCT
esrj-117639	109	9	there	there	PRON
esrj-117639	109	10	is	be	VERB
esrj-117639	109	11	a	a	DET
esrj-117639	109	12	clear	clear	ADJ
esrj-117639	109	13	positional	positional	ADJ
esrj-117639	109	14	feature	feature	NOUN
esrj-117639	109	15	between	between	ADP
esrj-117639	109	16	the	the	DET
esrj-117639	109	17	input	input	NOUN
esrj-117639	109	18	and	and	CCONJ
esrj-117639	109	19	output	output	NOUN
esrj-117639	109	20	data	datum	NOUN
esrj-117639	109	21	.	.	PUNCT
esrj-117639	110	1	using	use	VERB
esrj-117639	110	2	the	the	DET
esrj-117639	110	3	u	u	ADJ
esrj-117639	110	4	-	-	ADJ
esrj-117639	110	5	net	net	ADJ
esrj-117639	110	6	neural	neural	ADJ
esrj-117639	110	7	network	network	NOUN
esrj-117639	110	8	can	can	AUX
esrj-117639	110	9	more	more	ADV
esrj-117639	110	10	effectively	effectively	ADV
esrj-117639	110	11	complete	complete	VERB
esrj-117639	110	12	the	the	DET
esrj-117639	110	13	inversion	inversion	NOUN
esrj-117639	110	14	of	of	ADP
esrj-117639	110	15	the	the	DET
esrj-117639	110	16	interface	interface	NOUN
esrj-117639	110	17	undulation	undulation	NOUN
esrj-117639	110	18	depth	depth	NOUN
esrj-117639	110	19	.	.	PUNCT
esrj-117639	111	1	figure	figure	NOUN
esrj-117639	111	2	1	1	NUM
esrj-117639	111	3	.	.	X
esrj-117639	112	1	u	u	ADJ
esrj-117639	112	2	-	-	ADJ
esrj-117639	112	3	net	net	ADJ
esrj-117639	112	4	neural	neural	ADJ
esrj-117639	112	5	network	network	NOUN
esrj-117639	112	6	structure	structure	NOUN
esrj-117639	112	7	in	in	ADP
esrj-117639	112	8	this	this	DET
esrj-117639	112	9	paper	paper	NOUN
esrj-117639	112	10	,	,	PUNCT
esrj-117639	112	11	adam	adam	PROPN
esrj-117639	112	12	,	,	PUNCT
esrj-117639	112	13	which	which	PRON
esrj-117639	112	14	has	have	VERB
esrj-117639	112	15	strong	strong	ADJ
esrj-117639	112	16	robustness	robustness	NOUN
esrj-117639	112	17	,	,	PUNCT
esrj-117639	112	18	is	be	AUX
esrj-117639	112	19	adopted	adopt	VERB
esrj-117639	112	20	as	as	ADP
esrj-117639	112	21	the	the	DET
esrj-117639	112	22	optimization	optimization	NOUN
esrj-117639	112	23	algorithm	algorithm	NOUN
esrj-117639	112	24	,	,	PUNCT
esrj-117639	112	25	and	and	CCONJ
esrj-117639	112	26	the	the	DET
esrj-117639	112	27	minimum	minimum	ADJ
esrj-117639	112	28	square	square	NOUN
esrj-117639	112	29	error	error	NOUN
esrj-117639	112	30	φmse	φmse	NOUN
esrj-117639	112	31	is	be	AUX
esrj-117639	112	32	chosen	choose	VERB
esrj-117639	112	33	as	as	ADP
esrj-117639	112	34	the	the	DET
esrj-117639	112	35	basic	basic	ADJ
esrj-117639	112	36	loss	loss	NOUN
esrj-117639	112	37	function	function	NOUN
esrj-117639	112	38	.	.	PUNCT
esrj-117639	113	1	moreover	moreover	ADV
esrj-117639	113	2	,	,	PUNCT
esrj-117639	113	3	since	since	SCONJ
esrj-117639	113	4	it	it	PRON
esrj-117639	113	5	is	be	AUX
esrj-117639	113	6	assumed	assume	VERB
esrj-117639	113	7	that	that	SCONJ
esrj-117639	113	8	the	the	DET
esrj-117639	113	9	underground	underground	ADJ
esrj-117639	113	10	density	density	NOUN
esrj-117639	113	11	interface	interface	NOUN
esrj-117639	113	12	is	be	AUX
esrj-117639	113	13	smooth	smooth	ADJ
esrj-117639	113	14	,	,	PUNCT
esrj-117639	113	15	a	a	DET
esrj-117639	113	16	smoothness	smoothness	ADJ
esrj-117639	113	17	loss	loss	NOUN
esrj-117639	113	18	term	term	NOUN
esrj-117639	113	19	φsmooth	φsmooth	NOUN
esrj-117639	113	20	is	be	AUX
esrj-117639	113	21	added	add	VERB
esrj-117639	113	22	to	to	ADP
esrj-117639	113	23	the	the	DET
esrj-117639	113	24	basic	basic	ADJ
esrj-117639	113	25	loss	loss	NOUN
esrj-117639	113	26	function	function	NOUN
esrj-117639	113	27	to	to	PART
esrj-117639	113	28	penalize	penalize	VERB
esrj-117639	113	29	rough	rough	ADJ
esrj-117639	113	30	results	result	NOUN
esrj-117639	113	31	.	.	PUNCT
esrj-117639	114	1	at	at	ADP
esrj-117639	114	2	the	the	DET
esrj-117639	114	3	same	same	ADJ
esrj-117639	114	4	time	time	NOUN
esrj-117639	114	5	,	,	PUNCT
esrj-117639	114	6	to	to	PART
esrj-117639	114	7	address	address	VERB
esrj-117639	114	8	the	the	DET
esrj-117639	114	9	overfitting	overfitte	VERB
esrj-117639	114	10	issue	issue	NOUN
esrj-117639	114	11	during	during	ADP
esrj-117639	114	12	training	training	NOUN
esrj-117639	114	13	,	,	PUNCT
esrj-117639	114	14	an	an	DET
esrj-117639	114	15	l2	l2	NOUN
esrj-117639	114	16	norm	norm	NOUN
esrj-117639	114	17	is	be	AUX
esrj-117639	114	18	introduced	introduce	VERB
esrj-117639	114	19	to	to	PART
esrj-117639	114	20	constrain	constrain	VERB
esrj-117639	114	21	the	the	DET
esrj-117639	114	22	network	network	NOUN
esrj-117639	114	23	parameters	parameter	NOUN
esrj-117639	114	24	.	.	PUNCT
esrj-117639	115	1	the	the	DET
esrj-117639	115	2	final	final	ADJ
esrj-117639	115	3	expression	expression	NOUN
esrj-117639	115	4	of	of	ADP
esrj-117639	115	5	the	the	DET
esrj-117639	115	6	loss	loss	NOUN
esrj-117639	115	7	function	function	NOUN
esrj-117639	115	8	proposed	propose	VERB
esrj-117639	115	9	in	in	ADP
esrj-117639	115	10	this	this	DET
esrj-117639	115	11	paper	paper	NOUN
esrj-117639	115	12	is	be	AUX
esrj-117639	115	13	:	:	PUNCT
esrj-117639	115	14	φ	φ	PROPN
esrj-117639	115	15	(	(	PUNCT
esrj-117639	115	16	,	,	PUNCT
esrj-117639	115	17	ℎpre	ℎpre	NOUN
esrj-117639	115	18	)	)	PUNCT
esrj-117639	116	1	=	=	PUNCT
esrj-117639	116	2	φmsee	φmsee	ADJ
esrj-117639	116	3	(	(	PUNCT
esrj-117639	116	4	ℎpre	ℎpre	PROPN
esrj-117639	116	5	)	)	PUNCT
esrj-117639	117	1	+	+	CCONJ
esrj-117639	117	2	φsmooth	φsmooth	NOUN
esrj-117639	117	3	(	(	PUNCT
esrj-117639	117	4	ℎpre	ℎpre	PROPN
esrj-117639	117	5	)	)	PUNCT
esrj-117639	118	1	+	+	CCONJ
esrj-117639	118	2	φ	φ	NUM
esrj-117639	118	3	2	2	NUM
esrj-117639	118	4	(	(	PUNCT
esrj-117639	118	5	)	)	PUNCT
esrj-117639	118	6	(	(	PUNCT
esrj-117639	118	7	8)	8)	NUM
esrj-117639	118	8	where	where	SCONJ
esrj-117639	118	9	φmse	φmse	ADJ
esrj-117639	118	10	=	=	SYM
esrj-117639	118	11	∥	∥	PROPN
esrj-117639	118	12	ℎpre	ℎpre	NOUN
esrj-117639	118	13	−	−	NOUN
esrj-117639	118	14	ℎ	ℎ	PROPN
esrj-117639	118	15	∥2	∥2	NOUN
esrj-117639	118	16	2	2	NUM
esrj-117639	118	17	φsmooth	φsmooth	ADJ
esrj-117639	118	18	=	=	NOUN
esrj-117639	118	19	∥	∥	NOUN
esrj-117639	118	20	ℎpre	ℎpre	NOUN
esrj-117639	118	21	∥2	∥2	NOUN
esrj-117639	118	22	2	2	NUM
esrj-117639	118	23	φ	φ	NOUN
esrj-117639	118	24	2	2	NUM
esrj-117639	118	25	=	=	SYM
esrj-117639	118	26	∥	∥	X
esrj-117639	118	27	∥2	∥2	NOUN
esrj-117639	118	28	2	2	NUM
esrj-117639	118	29	{	{	PUNCT
esrj-117639	118	30	(	(	PUNCT
esrj-117639	118	31	9	9	NUM
esrj-117639	118	32	)	)	PUNCT
esrj-117639	118	33	in	in	ADP
esrj-117639	118	34	the	the	DET
esrj-117639	118	35	context	context	NOUN
esrj-117639	118	36	provided	provide	VERB
esrj-117639	118	37	,	,	PUNCT
esrj-117639	118	38	θ	θ	PROPN
esrj-117639	118	39	represents	represent	VERB
esrj-117639	118	40	the	the	DET
esrj-117639	118	41	weights	weight	NOUN
esrj-117639	118	42	and	and	CCONJ
esrj-117639	118	43	biases	bias	NOUN
esrj-117639	118	44	of	of	ADP
esrj-117639	118	45	the	the	DET
esrj-117639	118	46	neural	neural	ADJ
esrj-117639	118	47	network	network	NOUN
esrj-117639	118	48	,	,	PUNCT
esrj-117639	118	49	while	while	SCONJ
esrj-117639	118	50	hpre	hpre	PROPN
esrj-117639	118	51	is	be	AUX
esrj-117639	118	52	the	the	DET
esrj-117639	118	53	interface	interface	NOUN
esrj-117639	118	54	depth	depth	NOUN
esrj-117639	118	55	data	datum	NOUN
esrj-117639	118	56	obtained	obtain	VERB
esrj-117639	118	57	from	from	ADP
esrj-117639	118	58	the	the	DET
esrj-117639	118	59	neural	neural	ADJ
esrj-117639	118	60	network	network	NOUN
esrj-117639	118	61	inversion	inversion	NOUN
esrj-117639	118	62	.	.	PUNCT
esrj-117639	119	1	the	the	DET
esrj-117639	119	2	h	h	NOUN
esrj-117639	119	3	is	be	AUX
esrj-117639	119	4	the	the	DET
esrj-117639	119	5	desired	desire	VERB
esrj-117639	119	6	output	output	NOUN
esrj-117639	119	7	interface	interface	NOUN
esrj-117639	119	8	data	datum	NOUN
esrj-117639	119	9	,	,	PUNCT
esrj-117639	119	10	which	which	PRON
esrj-117639	119	11	is	be	AUX
esrj-117639	119	12	generated	generate	VERB
esrj-117639	119	13	by	by	ADP
esrj-117639	119	14	the	the	DET
esrj-117639	119	15	generator	generator	NOUN
esrj-117639	119	16	and	and	CCONJ
esrj-117639	119	17	obtained	obtain	VERB
esrj-117639	119	18	after	after	ADP
esrj-117639	119	19	preprocessing	preprocesse	VERB
esrj-117639	119	20	.	.	PUNCT
esrj-117639	120	1	2.4	2.4	NUM
esrj-117639	120	2	discrete	discrete	ADJ
esrj-117639	120	3	cosine	cosine	NOUN
esrj-117639	120	4	transform	transform	NOUN
esrj-117639	120	5	discrete	discrete	ADJ
esrj-117639	120	6	cosine	cosine	NOUN
esrj-117639	120	7	transform	transform	NOUN
esrj-117639	120	8	(	(	PUNCT
esrj-117639	120	9	dct	dct	PROPN
esrj-117639	120	10	)	)	PUNCT
esrj-117639	120	11	is	be	AUX
esrj-117639	120	12	widely	widely	ADV
esrj-117639	120	13	used	use	VERB
esrj-117639	120	14	in	in	ADP
esrj-117639	120	15	image	image	NOUN
esrj-117639	120	16	compression	compression	NOUN
esrj-117639	120	17	.	.	PUNCT
esrj-117639	121	1	dct	dct	PROPN
esrj-117639	121	2	represents	represent	VERB
esrj-117639	121	3	an	an	DET
esrj-117639	121	4	image	image	NOUN
esrj-117639	121	5	as	as	ADP
esrj-117639	121	6	a	a	DET
esrj-117639	121	7	sum	sum	NOUN
esrj-117639	121	8	of	of	ADP
esrj-117639	121	9	sine	sine	ADJ
esrj-117639	121	10	waves	wave	NOUN
esrj-117639	121	11	with	with	ADP
esrj-117639	121	12	different	different	ADJ
esrj-117639	121	13	amplitudes	amplitude	NOUN
esrj-117639	121	14	and	and	CCONJ
esrj-117639	121	15	frequencies	frequency	NOUN
esrj-117639	121	16	.	.	PUNCT
esrj-117639	122	1	by	by	ADP
esrj-117639	122	2	performing	perform	VERB
esrj-117639	122	3	a	a	DET
esrj-117639	122	4	discrete	discrete	ADJ
esrj-117639	122	5	cosine	cosine	NOUN
esrj-117639	122	6	forward	forward	ADV
esrj-117639	122	7	transform	transform	VERB
esrj-117639	122	8	on	on	ADP
esrj-117639	122	9	the	the	DET
esrj-117639	122	10	original	original	ADJ
esrj-117639	122	11	object	object	NOUN
esrj-117639	122	12	,	,	PUNCT
esrj-117639	122	13	the	the	DET
esrj-117639	122	14	transformed	transform	VERB
esrj-117639	122	15	coefficients	coefficient	NOUN
esrj-117639	122	16	are	be	AUX
esrj-117639	122	17	mainly	mainly	ADV
esrj-117639	122	18	concentrated	concentrate	VERB
esrj-117639	122	19	in	in	ADP
esrj-117639	122	20	the	the	DET
esrj-117639	122	21	upper	upper	ADJ
esrj-117639	122	22	left	left	NOUN
esrj-117639	122	23	corner	corner	NOUN
esrj-117639	122	24	,	,	PUNCT
esrj-117639	122	25	and	and	CCONJ
esrj-117639	122	26	most	most	ADJ
esrj-117639	122	27	of	of	ADP
esrj-117639	122	28	the	the	DET
esrj-117639	122	29	remaining	remain	VERB
esrj-117639	122	30	coefficients	coefficient	NOUN
esrj-117639	122	31	are	be	AUX
esrj-117639	122	32	close	close	ADJ
esrj-117639	122	33	to	to	ADP
esrj-117639	122	34	zero	zero	NUM
esrj-117639	122	35	.	.	PUNCT
esrj-117639	123	1	the	the	DET
esrj-117639	123	2	threshold	threshold	NOUN
esrj-117639	123	3	operation	operation	NOUN
esrj-117639	123	4	is	be	AUX
esrj-117639	123	5	performed	perform	VERB
esrj-117639	123	6	on	on	ADP
esrj-117639	123	7	the	the	DET
esrj-117639	123	8	transformed	transform	VERB
esrj-117639	123	9	dct	dct	PROPN
esrj-117639	123	10	coefficients	coefficient	NOUN
esrj-117639	123	11	,	,	PUNCT
esrj-117639	123	12	and	and	CCONJ
esrj-117639	123	13	coefficients	coefficient	VERB
esrj-117639	123	14	424	424	NUM
esrj-117639	123	15	huadong	huadong	ADJ
esrj-117639	123	16	song	song	NOUN
esrj-117639	123	17	,	,	PUNCT
esrj-117639	123	18	hui	hui	PROPN
esrj-117639	123	19	fu	fu	PROPN
esrj-117639	123	20	,	,	PUNCT
esrj-117639	123	21	jifeng	jifeng	PROPN
esrj-117639	123	22	qin	qin	PROPN
esrj-117639	123	23	smaller	small	ADJ
esrj-117639	123	24	than	than	ADP
esrj-117639	123	25	a	a	DET
esrj-117639	123	26	certain	certain	ADJ
esrj-117639	123	27	value	value	NOUN
esrj-117639	123	28	are	be	AUX
esrj-117639	123	29	set	set	VERB
esrj-117639	123	30	to	to	ADP
esrj-117639	123	31	zero	zero	NUM
esrj-117639	123	32	,	,	PUNCT
esrj-117639	123	33	which	which	PRON
esrj-117639	123	34	achieves	achieve	VERB
esrj-117639	123	35	the	the	DET
esrj-117639	123	36	quantization	quantization	NOUN
esrj-117639	123	37	process	process	NOUN
esrj-117639	123	38	in	in	ADP
esrj-117639	123	39	image	image	NOUN
esrj-117639	123	40	compression	compression	NOUN
esrj-117639	123	41	.	.	PUNCT
esrj-117639	124	1	then	then	ADV
esrj-117639	124	2	,	,	PUNCT
esrj-117639	124	3	the	the	DET
esrj-117639	124	4	discrete	discrete	ADJ
esrj-117639	124	5	cosine	cosine	NOUN
esrj-117639	124	6	inverse	inverse	NOUN
esrj-117639	124	7	transform	transform	NOUN
esrj-117639	124	8	is	be	AUX
esrj-117639	124	9	performed	perform	VERB
esrj-117639	124	10	to	to	PART
esrj-117639	124	11	obtain	obtain	VERB
esrj-117639	124	12	the	the	DET
esrj-117639	124	13	compressed	compressed	ADJ
esrj-117639	124	14	image	image	NOUN
esrj-117639	124	15	.	.	PUNCT
esrj-117639	125	1	the	the	DET
esrj-117639	125	2	2d	2d	NUM
esrj-117639	125	3	dft	dft	NOUN
esrj-117639	125	4	formula	formula	NOUN
esrj-117639	125	5	for	for	ADP
esrj-117639	125	6	an	an	DET
esrj-117639	125	7	input	input	NOUN
esrj-117639	125	8	image	image	NOUN
esrj-117639	125	9	data	datum	NOUN
esrj-117639	125	10	matrix	matrix	NOUN
esrj-117639	125	11	a	a	PRON
esrj-117639	125	12	with	with	ADP
esrj-117639	125	13	dimensions	dimension	NOUN
esrj-117639	125	14	m	m	VERB
esrj-117639	125	15	*	*	PUNCT
esrj-117639	125	16	n	n	PRON
esrj-117639	125	17	is	be	AUX
esrj-117639	125	18	:	:	PUNCT
esrj-117639	125	19	=	=	SYM
esrj-117639	125	20	∑	∑	PUNCT
esrj-117639	125	21	∑	∑	PROPN
esrj-117639	125	22	cos	cos	PROPN
esrj-117639	125	23	(	(	PUNCT
esrj-117639	125	24	)	)	PUNCT
esrj-117639	125	25	cos	cos	PROPN
esrj-117639	125	26	(	(	PUNCT
esrj-117639	125	27	)	)	PUNCT
esrj-117639	125	28	,	,	PUNCT
esrj-117639	125	29	0	0	NUM
esrj-117639	125	30	1	1	NUM
esrj-117639	125	31	0	0	NUM
esrj-117639	125	32	1	1	NUM
esrj-117639	125	33	(	(	PUNCT
esrj-117639	125	34	10	10	NUM
esrj-117639	125	35	)	)	PUNCT
esrj-117639	125	36	the	the	DET
esrj-117639	125	37	value	value	NOUN
esrj-117639	125	38	bpq	bpq	NOUN
esrj-117639	125	39	is	be	AUX
esrj-117639	125	40	the	the	DET
esrj-117639	125	41	dct	dct	PROPN
esrj-117639	125	42	coefficient	coefficient	NOUN
esrj-117639	125	43	of	of	ADP
esrj-117639	125	44	the	the	DET
esrj-117639	125	45	image	image	NOUN
esrj-117639	125	46	data	datum	NOUN
esrj-117639	125	47	matrix	matrix	NOUN
esrj-117639	125	48	a.	a.	PROPN
esrj-117639	125	49	dct	dct	PROPN
esrj-117639	125	50	is	be	AUX
esrj-117639	125	51	a	a	DET
esrj-117639	125	52	reversible	reversible	ADJ
esrj-117639	125	53	transform	transform	NOUN
esrj-117639	125	54	,	,	PUNCT
esrj-117639	125	55	and	and	CCONJ
esrj-117639	125	56	its	its	PRON
esrj-117639	125	57	inverse	inverse	NOUN
esrj-117639	125	58	transform	transform	NOUN
esrj-117639	125	59	(	(	PUNCT
esrj-117639	125	60	idft	idft	NOUN
esrj-117639	125	61	)	)	PUNCT
esrj-117639	125	62	equation	equation	NOUN
esrj-117639	125	63	is	be	AUX
esrj-117639	125	64	:	:	PUNCT
esrj-117639	125	65	=	=	SYM
esrj-117639	125	66	∑	∑	PUNCT
esrj-117639	125	67	∑	∑	PUNCT
esrj-117639	125	68	=	=	PROPN
esrj-117639	125	69	0	0	X
esrj-117639	125	70	cos	cos	PROPN
esrj-117639	125	71	(	(	PUNCT
esrj-117639	125	72	)	)	PUNCT
esrj-117639	125	73	cos	cos	PROPN
esrj-117639	125	74	(	(	PUNCT
esrj-117639	125	75	)	)	PUNCT
esrj-117639	125	76	,	,	PUNCT
esrj-117639	125	77	0	0	NUM
esrj-117639	125	78	1	1	NUM
esrj-117639	125	79	0	0	NUM
esrj-117639	125	80	1	1	NUM
esrj-117639	125	81	(	(	PUNCT
esrj-117639	125	82	11	11	NUM
esrj-117639	125	83	)	)	PUNCT
esrj-117639	125	84	where	where	SCONJ
esrj-117639	125	85	=	=	NOUN
esrj-117639	125	86	1/√	1/√	NUM
esrj-117639	125	87	,	,	PUNCT
esrj-117639	125	88	=	=	SYM
esrj-117639	125	89	0	0	NUM
esrj-117639	125	90	2/	2/	NUM
esrj-117639	125	91	,	,	PUNCT
esrj-117639	125	92	1	1	NUM
esrj-117639	125	93	1	1	NUM
esrj-117639	125	94	{	{	PUNCT
esrj-117639	125	95	(	(	PUNCT
esrj-117639	125	96	12	12	NUM
esrj-117639	125	97	)	)	PUNCT
esrj-117639	125	98	=	=	NOUN
esrj-117639	125	99	1/√	1/√	NUM
esrj-117639	125	100	,	,	PUNCT
esrj-117639	125	101	=	=	SYM
esrj-117639	125	102	0	0	NUM
esrj-117639	125	103	2/	2/	NUM
esrj-117639	125	104	,	,	PUNCT
esrj-117639	125	105	1	1	NUM
esrj-117639	125	106	1	1	NUM
esrj-117639	125	107	{	{	PUNCT
esrj-117639	125	108	(	(	PUNCT
esrj-117639	125	109	13	13	NUM
esrj-117639	125	110	)	)	PUNCT
esrj-117639	125	111	assuming	assume	VERB
esrj-117639	125	112	the	the	DET
esrj-117639	125	113	randomly	randomly	ADV
esrj-117639	125	114	generated	generate	VERB
esrj-117639	125	115	2d	2d	NUM
esrj-117639	125	116	subsurface	subsurface	NOUN
esrj-117639	125	117	medium	medium	NOUN
esrj-117639	125	118	as	as	ADP
esrj-117639	125	119	input	input	NOUN
esrj-117639	125	120	data	datum	NOUN
esrj-117639	125	121	,	,	PUNCT
esrj-117639	125	122	its	its	PRON
esrj-117639	125	123	large	large	ADJ
esrj-117639	125	124	dimension	dimension	NOUN
esrj-117639	125	125	requires	require	VERB
esrj-117639	125	126	a	a	DET
esrj-117639	125	127	large	large	ADJ
esrj-117639	125	128	number	number	NOUN
esrj-117639	125	129	of	of	ADP
esrj-117639	125	130	neural	neural	ADJ
esrj-117639	125	131	network	network	NOUN
esrj-117639	125	132	units	unit	NOUN
esrj-117639	125	133	for	for	ADP
esrj-117639	125	134	training	training	NOUN
esrj-117639	125	135	,	,	PUNCT
esrj-117639	125	136	which	which	PRON
esrj-117639	125	137	results	result	VERB
esrj-117639	125	138	in	in	ADP
esrj-117639	125	139	slow	slow	ADJ
esrj-117639	125	140	training	training	NOUN
esrj-117639	125	141	speed	speed	NOUN
esrj-117639	125	142	.	.	PUNCT
esrj-117639	126	1	to	to	PART
esrj-117639	126	2	solve	solve	VERB
esrj-117639	126	3	this	this	DET
esrj-117639	126	4	problem	problem	NOUN
esrj-117639	126	5	,	,	PUNCT
esrj-117639	126	6	we	we	PRON
esrj-117639	126	7	compress	compress	VERB
esrj-117639	126	8	its	its	PRON
esrj-117639	126	9	data	data	NOUN
esrj-117639	126	10	matrix	matrix	NOUN
esrj-117639	126	11	based	base	VERB
esrj-117639	126	12	on	on	ADP
esrj-117639	126	13	discrete	discrete	ADJ
esrj-117639	126	14	cosine	cosine	NOUN
esrj-117639	126	15	transform	transform	NOUN
esrj-117639	126	16	to	to	PART
esrj-117639	126	17	reduce	reduce	VERB
esrj-117639	126	18	the	the	DET
esrj-117639	126	19	dimension	dimension	NOUN
esrj-117639	126	20	of	of	ADP
esrj-117639	126	21	the	the	DET
esrj-117639	126	22	data	datum	NOUN
esrj-117639	126	23	.	.	PUNCT
esrj-117639	127	1	for	for	ADP
esrj-117639	127	2	a	a	DET
esrj-117639	127	3	2d	2d	NUM
esrj-117639	127	4	input	input	NOUN
esrj-117639	127	5	data	datum	NOUN
esrj-117639	127	6	matrix	matrix	NOUN
esrj-117639	127	7	,	,	PUNCT
esrj-117639	127	8	the	the	DET
esrj-117639	127	9	energy	energy	NOUN
esrj-117639	127	10	of	of	ADP
esrj-117639	127	11	the	the	DET
esrj-117639	127	12	dct	dct	PROPN
esrj-117639	127	13	coefficients	coefficient	NOUN
esrj-117639	127	14	is	be	AUX
esrj-117639	127	15	mainly	mainly	ADV
esrj-117639	127	16	concentrated	concentrate	VERB
esrj-117639	127	17	in	in	ADP
esrj-117639	127	18	the	the	DET
esrj-117639	127	19	low	low	ADJ
esrj-117639	127	20	-	-	PUNCT
esrj-117639	127	21	frequency	frequency	NOUN
esrj-117639	127	22	part	part	NOUN
esrj-117639	127	23	,	,	PUNCT
esrj-117639	127	24	and	and	CCONJ
esrj-117639	127	25	the	the	DET
esrj-117639	127	26	high	high	ADJ
esrj-117639	127	27	-	-	PUNCT
esrj-117639	127	28	frequency	frequency	NOUN
esrj-117639	127	29	part	part	NOUN
esrj-117639	127	30	is	be	AUX
esrj-117639	127	31	close	close	ADJ
esrj-117639	127	32	to	to	ADP
esrj-117639	127	33	zero	zero	NUM
esrj-117639	127	34	.	.	PUNCT
esrj-117639	128	1	in	in	ADP
esrj-117639	128	2	this	this	DET
esrj-117639	128	3	study	study	NOUN
esrj-117639	128	4	,	,	PUNCT
esrj-117639	128	5	we	we	PRON
esrj-117639	128	6	first	first	ADV
esrj-117639	128	7	use	use	VERB
esrj-117639	128	8	only	only	ADV
esrj-117639	128	9	the	the	DET
esrj-117639	128	10	low	low	ADJ
esrj-117639	128	11	-	-	PUNCT
esrj-117639	128	12	frequency	frequency	NOUN
esrj-117639	128	13	coefficient	coefficient	NOUN
esrj-117639	128	14	parts	part	NOUN
esrj-117639	128	15	of	of	ADP
esrj-117639	128	16	the	the	DET
esrj-117639	128	17	input	input	NOUN
esrj-117639	128	18	and	and	CCONJ
esrj-117639	128	19	output	output	NOUN
esrj-117639	128	20	for	for	ADP
esrj-117639	128	21	training	training	NOUN
esrj-117639	128	22	to	to	PART
esrj-117639	128	23	obtain	obtain	VERB
esrj-117639	128	24	an	an	DET
esrj-117639	128	25	optimal	optimal	ADJ
esrj-117639	128	26	neural	neural	ADJ
esrj-117639	128	27	network	network	NOUN
esrj-117639	128	28	.	.	PUNCT
esrj-117639	129	1	then	then	ADV
esrj-117639	129	2	,	,	PUNCT
esrj-117639	129	3	we	we	PRON
esrj-117639	129	4	use	use	VERB
esrj-117639	129	5	this	this	DET
esrj-117639	129	6	neural	neural	ADJ
esrj-117639	129	7	network	network	NOUN
esrj-117639	129	8	to	to	PART
esrj-117639	129	9	predict	predict	VERB
esrj-117639	129	10	new	new	ADJ
esrj-117639	129	11	inputs	input	NOUN
esrj-117639	129	12	and	and	CCONJ
esrj-117639	129	13	obtain	obtain	VERB
esrj-117639	129	14	the	the	DET
esrj-117639	129	15	low	low	ADJ
esrj-117639	129	16	-	-	PUNCT
esrj-117639	129	17	frequency	frequency	NOUN
esrj-117639	129	18	coefficients	coefficient	NOUN
esrj-117639	129	19	of	of	ADP
esrj-117639	129	20	the	the	DET
esrj-117639	129	21	new	new	ADJ
esrj-117639	129	22	output	output	NOUN
esrj-117639	129	23	.	.	PUNCT
esrj-117639	130	1	finally	finally	ADV
esrj-117639	130	2	,	,	PUNCT
esrj-117639	130	3	we	we	PRON
esrj-117639	130	4	use	use	VERB
esrj-117639	130	5	the	the	DET
esrj-117639	130	6	inverse	inverse	NOUN
esrj-117639	130	7	discrete	discrete	ADJ
esrj-117639	130	8	cosine	cosine	NOUN
esrj-117639	130	9	transform	transform	NOUN
esrj-117639	130	10	to	to	PART
esrj-117639	130	11	reconstruct	reconstruct	VERB
esrj-117639	130	12	the	the	DET
esrj-117639	130	13	full	full	ADJ
esrj-117639	130	14	-	-	PUNCT
esrj-117639	130	15	frequency	frequency	NOUN
esrj-117639	130	16	data	datum	NOUN
esrj-117639	130	17	.	.	PUNCT
esrj-117639	131	1	3	3	X
esrj-117639	131	2	.	.	X
esrj-117639	131	3	application	application	NOUN
esrj-117639	131	4	3.1	3.1	NUM
esrj-117639	131	5	synthetic	synthetic	ADJ
esrj-117639	131	6	data	datum	NOUN
esrj-117639	131	7	during	during	ADP
esrj-117639	131	8	the	the	DET
esrj-117639	131	9	neural	neural	ADJ
esrj-117639	131	10	network	network	NOUN
esrj-117639	131	11	training	training	NOUN
esrj-117639	131	12	process	process	NOUN
esrj-117639	131	13	,	,	PUNCT
esrj-117639	131	14	we	we	PRON
esrj-117639	131	15	used	used	AUX
esrj-117639	131	16	randomly	randomly	ADV
esrj-117639	131	17	generated	generate	VERB
esrj-117639	131	18	underground	underground	ADJ
esrj-117639	131	19	interfaces	interface	NOUN
esrj-117639	131	20	to	to	PART
esrj-117639	131	21	create	create	VERB
esrj-117639	131	22	a	a	DET
esrj-117639	131	23	training	training	NOUN
esrj-117639	131	24	set	set	NOUN
esrj-117639	131	25	of	of	ADP
esrj-117639	131	26	100,000	100,000	NUM
esrj-117639	131	27	“	"	PUNCT
esrj-117639	131	28	interface	interface	NOUN
esrj-117639	131	29	modelgravity	modelgravity	NOUN
esrj-117639	131	30	anomaly	anomaly	NOUN
esrj-117639	131	31	data	datum	NOUN
esrj-117639	131	32	”	"	PUNCT
esrj-117639	131	33	pairs	pair	NOUN
esrj-117639	131	34	,	,	PUNCT
esrj-117639	131	35	and	and	CCONJ
esrj-117639	131	36	then	then	ADV
esrj-117639	131	37	trained	train	VERB
esrj-117639	131	38	an	an	DET
esrj-117639	131	39	artificial	artificial	ADJ
esrj-117639	131	40	neural	neural	ADJ
esrj-117639	131	41	network	network	NOUN
esrj-117639	131	42	.	.	PUNCT
esrj-117639	132	1	we	we	PRON
esrj-117639	132	2	used	use	VERB
esrj-117639	132	3	randomly	randomly	ADV
esrj-117639	132	4	synthesized	synthesize	VERB
esrj-117639	132	5	underground	underground	ADJ
esrj-117639	132	6	interfaces	interface	NOUN
esrj-117639	132	7	to	to	PART
esrj-117639	132	8	verify	verify	VERB
esrj-117639	132	9	the	the	DET
esrj-117639	132	10	correctness	correctness	NOUN
esrj-117639	132	11	of	of	ADP
esrj-117639	132	12	the	the	DET
esrj-117639	132	13	algorithm	algorithm	NOUN
esrj-117639	132	14	proposed	propose	VERB
esrj-117639	132	15	in	in	ADP
esrj-117639	132	16	this	this	DET
esrj-117639	132	17	paper	paper	NOUN
esrj-117639	132	18	.	.	PUNCT
esrj-117639	133	1	the	the	DET
esrj-117639	133	2	average	average	ADJ
esrj-117639	133	3	depth	depth	NOUN
esrj-117639	133	4	of	of	ADP
esrj-117639	133	5	the	the	DET
esrj-117639	133	6	model	model	NOUN
esrj-117639	133	7	is	be	AUX
esrj-117639	133	8	25	25	NUM
esrj-117639	133	9	km	km	NOUN
esrj-117639	133	10	,	,	PUNCT
esrj-117639	133	11	with	with	ADP
esrj-117639	133	12	a	a	DET
esrj-117639	133	13	range	range	NOUN
esrj-117639	133	14	of	of	ADP
esrj-117639	133	15	250	250	NUM
esrj-117639	133	16	km	km	NOUN
esrj-117639	133	17	×	×	NOUN
esrj-117639	133	18	250	250	NUM
esrj-117639	133	19	km	km	NOUN
esrj-117639	133	20	.	.	PUNCT
esrj-117639	134	1	in	in	ADP
esrj-117639	134	2	this	this	DET
esrj-117639	134	3	model	model	NOUN
esrj-117639	134	4	,	,	PUNCT
esrj-117639	134	5	the	the	DET
esrj-117639	134	6	northeast	northeast	NOUN
esrj-117639	134	7	is	be	AUX
esrj-117639	134	8	a	a	DET
esrj-117639	134	9	large	large	ADJ
esrj-117639	134	10	dome	dome	NOUN
esrj-117639	134	11	with	with	ADP
esrj-117639	134	12	a	a	DET
esrj-117639	134	13	minimum	minimum	ADJ
esrj-117639	134	14	interface	interface	NOUN
esrj-117639	134	15	depth	depth	NOUN
esrj-117639	134	16	of	of	ADP
esrj-117639	134	17	about	about	ADV
esrj-117639	134	18	16	16	NUM
esrj-117639	134	19	km	km	NOUN
esrj-117639	134	20	,	,	PUNCT
esrj-117639	134	21	and	and	CCONJ
esrj-117639	134	22	the	the	DET
esrj-117639	134	23	south	south	NOUN
esrj-117639	134	24	consists	consist	VERB
esrj-117639	134	25	of	of	ADP
esrj-117639	134	26	several	several	ADJ
esrj-117639	134	27	discontinuous	discontinuous	ADJ
esrj-117639	134	28	basins	basin	NOUN
esrj-117639	134	29	with	with	ADP
esrj-117639	134	30	a	a	DET
esrj-117639	134	31	maximum	maximum	ADJ
esrj-117639	134	32	depth	depth	NOUN
esrj-117639	134	33	of	of	ADP
esrj-117639	134	34	about	about	ADV
esrj-117639	134	35	24	24	NUM
esrj-117639	134	36	km	km	NOUN
esrj-117639	134	37	.	.	PUNCT
esrj-117639	135	1	the	the	DET
esrj-117639	135	2	density	density	NOUN
esrj-117639	135	3	contrast	contrast	NOUN
esrj-117639	135	4	of	of	ADP
esrj-117639	135	5	the	the	DET
esrj-117639	135	6	interface	interface	NOUN
esrj-117639	135	7	is	be	AUX
esrj-117639	135	8	0.4g	0.4g	ADJ
esrj-117639	135	9	/	/	SYM
esrj-117639	135	10	cm3	cm3	NOUN
esrj-117639	135	11	,	,	PUNCT
esrj-117639	135	12	and	and	CCONJ
esrj-117639	135	13	the	the	DET
esrj-117639	135	14	average	average	ADJ
esrj-117639	135	15	depth	depth	NOUN
esrj-117639	135	16	of	of	ADP
esrj-117639	135	17	the	the	DET
esrj-117639	135	18	interface	interface	NOUN
esrj-117639	135	19	is	be	AUX
esrj-117639	135	20	20	20	NUM
esrj-117639	135	21	km	km	NOUN
esrj-117639	135	22	.	.	PUNCT
esrj-117639	136	1	the	the	DET
esrj-117639	136	2	observation	observation	NOUN
esrj-117639	136	3	points	point	NOUN
esrj-117639	136	4	on	on	ADP
esrj-117639	136	5	the	the	DET
esrj-117639	136	6	surface	surface	NOUN
esrj-117639	136	7	are	be	AUX
esrj-117639	136	8	spaced	space	VERB
esrj-117639	136	9	"	"	PUNCT
esrj-117639	136	10	1	1	NUM
esrj-117639	136	11	km	km	NOUN
esrj-117639	136	12	×1	×1	NOUN
esrj-117639	136	13	km	km	NOUN
esrj-117639	136	14	"	"	PUNCT
esrj-117639	136	15	apart	apart	ADV
esrj-117639	136	16	,	,	PUNCT
esrj-117639	136	17	and	and	CCONJ
esrj-117639	136	18	the	the	DET
esrj-117639	136	19	underground	underground	ADJ
esrj-117639	136	20	medium	medium	NOUN
esrj-117639	136	21	is	be	AUX
esrj-117639	136	22	divided	divide	VERB
esrj-117639	136	23	into	into	ADP
esrj-117639	136	24	a	a	DET
esrj-117639	136	25	grid	grid	NOUN
esrj-117639	136	26	of	of	ADP
esrj-117639	136	27	256	256	NUM
esrj-117639	136	28	×	×	NOUN
esrj-117639	136	29	256	256	NUM
esrj-117639	136	30	with	with	ADP
esrj-117639	136	31	a	a	DET
esrj-117639	136	32	grid	grid	NOUN
esrj-117639	136	33	spacing	spacing	NOUN
esrj-117639	136	34	of	of	ADP
esrj-117639	136	35	"	"	PUNCT
esrj-117639	136	36	1	1	NUM
esrj-117639	136	37	km	km	NOUN
esrj-117639	136	38	"	"	PUNCT
esrj-117639	136	39	.	.	PUNCT
esrj-117639	137	1	a	a	DET
esrj-117639	137	2	5	5	NUM
esrj-117639	137	3	%	%	NOUN
esrj-117639	137	4	gaussian	gaussian	ADJ
esrj-117639	137	5	random	random	ADJ
esrj-117639	137	6	error	error	NOUN
esrj-117639	137	7	was	be	AUX
esrj-117639	137	8	added	add	VERB
esrj-117639	137	9	during	during	ADP
esrj-117639	137	10	the	the	DET
esrj-117639	137	11	artificial	artificial	ADJ
esrj-117639	137	12	neural	neural	ADJ
esrj-117639	137	13	network	network	NOUN
esrj-117639	137	14	training	training	NOUN
esrj-117639	137	15	process	process	NOUN
esrj-117639	137	16	.	.	PUNCT
esrj-117639	138	1	during	during	ADP
esrj-117639	138	2	the	the	DET
esrj-117639	138	3	training	training	NOUN
esrj-117639	138	4	process	process	NOUN
esrj-117639	138	5	,	,	PUNCT
esrj-117639	138	6	the	the	DET
esrj-117639	138	7	discrete	discrete	ADJ
esrj-117639	138	8	cosine	cosine	NOUN
esrj-117639	138	9	transform	transform	NOUN
esrj-117639	138	10	algorithm	algorithm	NOUN
esrj-117639	138	11	was	be	AUX
esrj-117639	138	12	used	use	VERB
esrj-117639	138	13	to	to	PART
esrj-117639	138	14	compress	compress	VERB
esrj-117639	138	15	the	the	DET
esrj-117639	138	16	data	datum	NOUN
esrj-117639	138	17	matrix	matrix	NOUN
esrj-117639	138	18	.	.	PUNCT
esrj-117639	139	1	the	the	DET
esrj-117639	139	2	size	size	NOUN
esrj-117639	139	3	of	of	ADP
esrj-117639	139	4	the	the	DET
esrj-117639	139	5	observation	observation	NOUN
esrj-117639	139	6	data	datum	NOUN
esrj-117639	139	7	matrix	matrix	NOUN
esrj-117639	139	8	is	be	AUX
esrj-117639	139	9	known	know	VERB
esrj-117639	139	10	to	to	PART
esrj-117639	139	11	be	be	AUX
esrj-117639	139	12	250	250	NUM
esrj-117639	139	13	×	×	NOUN
esrj-117639	139	14	250	250	NUM
esrj-117639	139	15	,	,	PUNCT
esrj-117639	139	16	and	and	CCONJ
esrj-117639	139	17	the	the	DET
esrj-117639	139	18	size	size	NOUN
esrj-117639	139	19	of	of	ADP
esrj-117639	139	20	the	the	DET
esrj-117639	139	21	underground	underground	ADJ
esrj-117639	139	22	medium	medium	ADJ
esrj-117639	139	23	matrix	matrix	NOUN
esrj-117639	139	24	is	be	AUX
esrj-117639	139	25	256	256	NUM
esrj-117639	139	26	×	×	NOUN
esrj-117639	139	27	256	256	NUM
esrj-117639	139	28	.	.	PUNCT
esrj-117639	140	1	figure	figure	NOUN
esrj-117639	140	2	2(a	2(a	NUM
esrj-117639	140	3	)	)	PUNCT
esrj-117639	140	4	and	and	CCONJ
esrj-117639	140	5	2(b	2(b	NUM
esrj-117639	140	6	)	)	PUNCT
esrj-117639	141	1	show	show	VERB
esrj-117639	141	2	the	the	DET
esrj-117639	141	3	discrete	discrete	ADJ
esrj-117639	141	4	cosine	cosine	NOUN
esrj-117639	141	5	transform	transform	NOUN
esrj-117639	141	6	coefficients	coefficient	NOUN
esrj-117639	141	7	of	of	ADP
esrj-117639	141	8	the	the	DET
esrj-117639	141	9	true	true	ADJ
esrj-117639	141	10	depth	depth	NOUN
esrj-117639	141	11	model	model	NOUN
esrj-117639	141	12	and	and	CCONJ
esrj-117639	141	13	the	the	DET
esrj-117639	141	14	observation	observation	NOUN
esrj-117639	141	15	data	datum	NOUN
esrj-117639	141	16	,	,	PUNCT
esrj-117639	141	17	respectively	respectively	ADV
esrj-117639	141	18	.	.	PUNCT
esrj-117639	142	1	it	it	PRON
esrj-117639	142	2	can	can	AUX
esrj-117639	142	3	be	be	AUX
esrj-117639	142	4	seen	see	VERB
esrj-117639	142	5	from	from	ADP
esrj-117639	142	6	the	the	DET
esrj-117639	142	7	figures	figure	NOUN
esrj-117639	142	8	that	that	SCONJ
esrj-117639	142	9	the	the	DET
esrj-117639	142	10	energy	energy	NOUN
esrj-117639	142	11	of	of	ADP
esrj-117639	142	12	the	the	DET
esrj-117639	142	13	coefficients	coefficient	NOUN
esrj-117639	142	14	after	after	ADP
esrj-117639	142	15	the	the	DET
esrj-117639	142	16	discrete	discrete	ADJ
esrj-117639	142	17	cosine	cosine	NOUN
esrj-117639	142	18	transform	transform	NOUN
esrj-117639	142	19	is	be	AUX
esrj-117639	142	20	mainly	mainly	ADV
esrj-117639	142	21	concentrated	concentrate	VERB
esrj-117639	142	22	in	in	ADP
esrj-117639	142	23	the	the	DET
esrj-117639	142	24	upper	upper	ADJ
esrj-117639	142	25	left	left	ADJ
esrj-117639	142	26	corner	corner	NOUN
esrj-117639	142	27	(	(	PUNCT
esrj-117639	142	28	shown	show	VERB
esrj-117639	142	29	in	in	ADP
esrj-117639	142	30	the	the	DET
esrj-117639	142	31	red	red	PROPN
esrj-117639	142	32	box	box	PROPN
esrj-117639	142	33	,	,	PUNCT
esrj-117639	142	34	at	at	ADP
esrj-117639	142	35	approximately	approximately	ADV
esrj-117639	142	36	position	position	NOUN
esrj-117639	142	37	10	10	NUM
esrj-117639	142	38	×	×	NOUN
esrj-117639	142	39	10	10	NUM
esrj-117639	142	40	)	)	PUNCT
esrj-117639	142	41	,	,	PUNCT
esrj-117639	142	42	and	and	CCONJ
esrj-117639	142	43	most	most	ADJ
esrj-117639	142	44	of	of	ADP
esrj-117639	142	45	the	the	DET
esrj-117639	142	46	other	other	ADJ
esrj-117639	142	47	coefficients	coefficient	NOUN
esrj-117639	142	48	are	be	AUX
esrj-117639	142	49	close	close	ADJ
esrj-117639	142	50	to	to	ADP
esrj-117639	142	51	zero	zero	NUM
esrj-117639	142	52	.	.	PUNCT
esrj-117639	143	1	only	only	ADV
esrj-117639	143	2	the	the	DET
esrj-117639	143	3	coefficients	coefficient	NOUN
esrj-117639	143	4	within	within	ADP
esrj-117639	143	5	the	the	DET
esrj-117639	143	6	red	red	PROPN
esrj-117639	143	7	box	box	PROPN
esrj-117639	143	8	were	be	AUX
esrj-117639	143	9	used	use	VERB
esrj-117639	143	10	during	during	ADP
esrj-117639	143	11	the	the	DET
esrj-117639	143	12	artificial	artificial	ADJ
esrj-117639	143	13	neural	neural	ADJ
esrj-117639	143	14	network	network	NOUN
esrj-117639	143	15	training	training	NOUN
esrj-117639	143	16	,	,	PUNCT
esrj-117639	143	17	compressing	compress	VERB
esrj-117639	143	18	the	the	DET
esrj-117639	143	19	256	256	NUM
esrj-117639	143	20	×	×	NOUN
esrj-117639	143	21	256	256	NUM
esrj-117639	143	22	data	datum	NOUN
esrj-117639	143	23	matrix	matrix	NOUN
esrj-117639	143	24	into	into	ADP
esrj-117639	143	25	a	a	DET
esrj-117639	143	26	10	10	NUM
esrj-117639	143	27	×	×	NOUN
esrj-117639	143	28	10	10	NUM
esrj-117639	143	29	data	datum	NOUN
esrj-117639	143	30	matrix	matrix	NOUN
esrj-117639	143	31	,	,	PUNCT
esrj-117639	143	32	greatly	greatly	ADV
esrj-117639	143	33	reducing	reduce	VERB
esrj-117639	143	34	the	the	DET
esrj-117639	143	35	computational	computational	ADJ
esrj-117639	143	36	memory	memory	NOUN
esrj-117639	143	37	,	,	PUNCT
esrj-117639	143	38	shortening	shorten	VERB
esrj-117639	143	39	the	the	DET
esrj-117639	143	40	calculation	calculation	NOUN
esrj-117639	143	41	time	time	NOUN
esrj-117639	143	42	,	,	PUNCT
esrj-117639	143	43	and	and	CCONJ
esrj-117639	143	44	improving	improve	VERB
esrj-117639	143	45	the	the	DET
esrj-117639	143	46	efficiency	efficiency	NOUN
esrj-117639	143	47	of	of	ADP
esrj-117639	143	48	artificial	artificial	ADJ
esrj-117639	143	49	neural	neural	ADJ
esrj-117639	143	50	network	network	NOUN
esrj-117639	143	51	training	training	NOUN
esrj-117639	143	52	.	.	PUNCT
esrj-117639	144	1	by	by	ADP
esrj-117639	144	2	comparison	comparison	NOUN
esrj-117639	144	3	,	,	PUNCT
esrj-117639	144	4	we	we	PRON
esrj-117639	144	5	know	know	VERB
esrj-117639	144	6	that	that	SCONJ
esrj-117639	144	7	training	training	NOUN
esrj-117639	144	8	without	without	ADP
esrj-117639	144	9	discrete	discrete	ADJ
esrj-117639	144	10	cosine	cosine	NOUN
esrj-117639	144	11	transform	transform	NOUN
esrj-117639	144	12	compression	compression	NOUN
esrj-117639	144	13	takes	take	VERB
esrj-117639	144	14	about	about	ADV
esrj-117639	144	15	68,000	68,000	NUM
esrj-117639	144	16	seconds	second	NOUN
esrj-117639	144	17	,	,	PUNCT
esrj-117639	144	18	while	while	SCONJ
esrj-117639	144	19	after	after	ADP
esrj-117639	144	20	compression	compression	NOUN
esrj-117639	144	21	,	,	PUNCT
esrj-117639	144	22	the	the	DET
esrj-117639	144	23	neural	neural	ADJ
esrj-117639	144	24	network	network	NOUN
esrj-117639	144	25	training	training	NOUN
esrj-117639	144	26	only	only	ADV
esrj-117639	144	27	takes	take	VERB
esrj-117639	144	28	about	about	ADV
esrj-117639	144	29	2,400	2,400	NUM
esrj-117639	144	30	seconds	second	NOUN
esrj-117639	144	31	.	.	PUNCT
esrj-117639	145	1	from	from	ADP
esrj-117639	145	2	figure	figure	NOUN
esrj-117639	145	3	3(a	3(a	NUM
esrj-117639	145	4	)	)	PUNCT
esrj-117639	145	5	and	and	CCONJ
esrj-117639	145	6	3(c	3(c	NUM
esrj-117639	145	7	)	)	PUNCT
esrj-117639	145	8	,	,	PUNCT
esrj-117639	145	9	it	it	PRON
esrj-117639	145	10	can	can	AUX
esrj-117639	145	11	be	be	AUX
esrj-117639	145	12	seen	see	VERB
esrj-117639	145	13	that	that	SCONJ
esrj-117639	145	14	the	the	DET
esrj-117639	145	15	inversion	inversion	NOUN
esrj-117639	145	16	results	result	NOUN
esrj-117639	145	17	of	of	ADP
esrj-117639	145	18	the	the	DET
esrj-117639	145	19	synthetic	synthetic	ADJ
esrj-117639	145	20	model	model	NOUN
esrj-117639	145	21	match	match	NOUN
esrj-117639	145	22	well	well	ADV
esrj-117639	145	23	with	with	ADP
esrj-117639	145	24	the	the	DET
esrj-117639	145	25	true	true	ADJ
esrj-117639	145	26	model	model	NOUN
esrj-117639	145	27	,	,	PUNCT
esrj-117639	145	28	and	and	CCONJ
esrj-117639	145	29	both	both	CCONJ
esrj-117639	145	30	the	the	DET
esrj-117639	145	31	minimum	minimum	ADJ
esrj-117639	145	32	interface	interface	NOUN
esrj-117639	145	33	depth	depth	NOUN
esrj-117639	145	34	of	of	ADP
esrj-117639	145	35	16	16	NUM
esrj-117639	145	36	km	km	NOUN
esrj-117639	145	37	and	and	CCONJ
esrj-117639	145	38	the	the	DET
esrj-117639	145	39	maximum	maximum	ADJ
esrj-117639	145	40	interface	interface	NOUN
esrj-117639	145	41	depth	depth	NOUN
esrj-117639	145	42	of	of	ADP
esrj-117639	145	43	24	24	NUM
esrj-117639	145	44	km	km	NOUN
esrj-117639	145	45	can	can	AUX
esrj-117639	145	46	be	be	AUX
esrj-117639	145	47	well	well	ADV
esrj-117639	145	48	recovered	recover	VERB
esrj-117639	145	49	.	.	PUNCT
esrj-117639	146	1	in	in	ADP
esrj-117639	146	2	addition	addition	NOUN
esrj-117639	146	3	,	,	PUNCT
esrj-117639	146	4	as	as	SCONJ
esrj-117639	146	5	shown	show	VERB
esrj-117639	146	6	in	in	ADP
esrj-117639	146	7	figure	figure	NOUN
esrj-117639	146	8	3(b	3(b	NUM
esrj-117639	146	9	)	)	PUNCT
esrj-117639	146	10	and	and	CCONJ
esrj-117639	146	11	3(d	3(d	NUM
esrj-117639	146	12	)	)	PUNCT
esrj-117639	146	13	,	,	PUNCT
esrj-117639	146	14	the	the	DET
esrj-117639	146	15	gravity	gravity	NOUN
esrj-117639	146	16	anomaly	anomaly	NOUN
esrj-117639	146	17	data	datum	NOUN
esrj-117639	146	18	calculated	calculate	VERB
esrj-117639	146	19	from	from	ADP
esrj-117639	146	20	the	the	DET
esrj-117639	146	21	inverted	inverted	ADJ
esrj-117639	146	22	depth	depth	NOUN
esrj-117639	146	23	interfaces	interface	NOUN
esrj-117639	146	24	are	be	AUX
esrj-117639	146	25	also	also	ADV
esrj-117639	146	26	consistent	consistent	ADJ
esrj-117639	146	27	with	with	ADP
esrj-117639	146	28	the	the	DET
esrj-117639	146	29	observed	observe	VERB
esrj-117639	146	30	gravity	gravity	NOUN
esrj-117639	146	31	anomaly	anomaly	NOUN
esrj-117639	146	32	data	datum	NOUN
esrj-117639	146	33	.	.	PUNCT
esrj-117639	147	1	figure	figure	NOUN
esrj-117639	147	2	3(e	3(e	NUM
esrj-117639	147	3	)	)	PUNCT
esrj-117639	147	4	and	and	CCONJ
esrj-117639	147	5	3(f	3(f	NUM
esrj-117639	147	6	)	)	PUNCT
esrj-117639	147	7	show	show	VERB
esrj-117639	147	8	the	the	DET
esrj-117639	147	9	differences	difference	NOUN
esrj-117639	147	10	in	in	ADP
esrj-117639	147	11	depth	depth	NOUN
esrj-117639	147	12	interface	interface	NOUN
esrj-117639	147	13	models	model	NOUN
esrj-117639	147	14	and	and	CCONJ
esrj-117639	147	15	gravity	gravity	NOUN
esrj-117639	147	16	anomaly	anomaly	NOUN
esrj-117639	147	17	data	datum	NOUN
esrj-117639	147	18	,	,	PUNCT
esrj-117639	147	19	respectively	respectively	ADV
esrj-117639	147	20	.	.	PUNCT
esrj-117639	148	1	it	it	PRON
esrj-117639	148	2	can	can	AUX
esrj-117639	148	3	be	be	AUX
esrj-117639	148	4	seen	see	VERB
esrj-117639	148	5	that	that	SCONJ
esrj-117639	148	6	the	the	DET
esrj-117639	148	7	differences	difference	NOUN
esrj-117639	148	8	in	in	ADP
esrj-117639	148	9	the	the	DET
esrj-117639	148	10	depth	depth	NOUN
esrj-117639	148	11	interface	interface	NOUN
esrj-117639	148	12	models	model	NOUN
esrj-117639	148	13	are	be	AUX
esrj-117639	148	14	very	very	ADV
esrj-117639	148	15	small	small	ADJ
esrj-117639	148	16	,	,	PUNCT
esrj-117639	148	17	with	with	ADP
esrj-117639	148	18	a	a	DET
esrj-117639	148	19	range	range	NOUN
esrj-117639	148	20	of	of	ADP
esrj-117639	148	21	approximately	approximately	ADV
esrj-117639	148	22	-1.0	-1.0	PROPN
esrj-117639	148	23	~	~	PUNCT
esrj-117639	148	24	1.0	1.0	NUM
esrj-117639	148	25	km	km	NOUN
esrj-117639	148	26	,	,	PUNCT
esrj-117639	148	27	while	while	SCONJ
esrj-117639	148	28	the	the	DET
esrj-117639	148	29	differences	difference	NOUN
esrj-117639	148	30	in	in	ADP
esrj-117639	148	31	the	the	DET
esrj-117639	148	32	gravity	gravity	NOUN
esrj-117639	148	33	anomaly	anomaly	NOUN
esrj-117639	148	34	data	datum	NOUN
esrj-117639	148	35	are	be	AUX
esrj-117639	148	36	also	also	ADV
esrj-117639	148	37	very	very	ADV
esrj-117639	148	38	small	small	ADJ
esrj-117639	148	39	,	,	PUNCT
esrj-117639	148	40	with	with	ADP
esrj-117639	148	41	a	a	DET
esrj-117639	148	42	range	range	NOUN
esrj-117639	148	43	of	of	ADP
esrj-117639	148	44	approximately	approximately	ADV
esrj-117639	148	45	-2.5	-2.5	ADJ
esrj-117639	148	46	~	~	SYM
esrj-117639	148	47	1.5	1.5	NUM
esrj-117639	148	48	mgal	mgal	NOUN
esrj-117639	148	49	.	.	PUNCT
esrj-117639	149	1	figure	figure	NOUN
esrj-117639	149	2	2	2	NUM
esrj-117639	149	3	.	.	NOUN
esrj-117639	149	4	discrete	discrete	ADJ
esrj-117639	149	5	cosine	cosine	NOUN
esrj-117639	149	6	transform	transform	NOUN
esrj-117639	149	7	coefficient	coefficient	NOUN
esrj-117639	149	8	of	of	ADP
esrj-117639	149	9	the	the	DET
esrj-117639	149	10	random	random	ADJ
esrj-117639	149	11	medium	medium	NOUN
esrj-117639	149	12	model	model	NOUN
esrj-117639	149	13	(	(	PUNCT
esrj-117639	149	14	a)-depth	a)-depth	NOUN
esrj-117639	149	15	model	model	PROPN
esrj-117639	149	16	dct	dct	PROPN
esrj-117639	149	17	coefficient	coefficient	PROPN
esrj-117639	149	18	(	(	PUNCT
esrj-117639	149	19	b)-gravity	b)-gravity	NOUN
esrj-117639	149	20	data	datum	NOUN
esrj-117639	149	21	dct	dct	PROPN
esrj-117639	149	22	coefficient	coefficient	PROPN
esrj-117639	149	23	figure	figure	NOUN
esrj-117639	149	24	3	3	NUM
esrj-117639	149	25	.	.	PUNCT
esrj-117639	150	1	the	the	DET
esrj-117639	150	2	random	random	ADJ
esrj-117639	150	3	medium	medium	NOUN
esrj-117639	150	4	model	model	NOUN
esrj-117639	150	5	(	(	PUNCT
esrj-117639	150	6	a)-true	a)-true	PUNCT
esrj-117639	150	7	depth	depth	NOUN
esrj-117639	150	8	model	model	NOUN
esrj-117639	150	9	(	(	PUNCT
esrj-117639	150	10	b)-observed	b)-observe	VERB
esrj-117639	150	11	gravity	gravity	NOUN
esrj-117639	150	12	(	(	PUNCT
esrj-117639	150	13	c)-invert	c)-invert	NOUN
esrj-117639	150	14	depth	depth	NOUN
esrj-117639	150	15	model	model	NOUN
esrj-117639	150	16	(	(	PUNCT
esrj-117639	150	17	d)-calculated	d)-calculate	VERB
esrj-117639	150	18	gravity	gravity	NOUN
esrj-117639	150	19	(	(	PUNCT
esrj-117639	150	20	e)-difference	e)-difference	NOUN
esrj-117639	150	21	of	of	ADP
esrj-117639	150	22	depth	depth	NOUN
esrj-117639	150	23	model	model	NOUN
esrj-117639	150	24	(	(	PUNCT
esrj-117639	150	25	f)-difference	f)-difference	NOUN
esrj-117639	150	26	of	of	ADP
esrj-117639	150	27	gravity	gravity	NOUN
esrj-117639	150	28	data	datum	NOUN
esrj-117639	150	29	425three	425three	ADJ
esrj-117639	150	30	-	-	ADJ
esrj-117639	150	31	dimensional	dimensional	ADJ
esrj-117639	150	32	gravity	gravity	NOUN
esrj-117639	150	33	interface	interface	NOUN
esrj-117639	150	34	inversion	inversion	NOUN
esrj-117639	150	35	based	base	VERB
esrj-117639	150	36	on	on	ADP
esrj-117639	150	37	artificial	artificial	ADJ
esrj-117639	150	38	neural	neural	ADJ
esrj-117639	150	39	network	network	NOUN
esrj-117639	150	40	and	and	CCONJ
esrj-117639	150	41	discrete	discrete	ADJ
esrj-117639	150	42	cosine	cosine	NOUN
esrj-117639	150	43	transform	transform	NOUN
esrj-117639	150	44	algorithm	algorithm	NOUN
esrj-117639	150	45	3.2	3.2	NUM
esrj-117639	150	46	field	field	NOUN
esrj-117639	150	47	data	data	VERB
esrj-117639	150	48	the	the	DET
esrj-117639	150	49	real	real	ADJ
esrj-117639	150	50	-	-	PUNCT
esrj-117639	150	51	world	world	NOUN
esrj-117639	150	52	example	example	NOUN
esrj-117639	150	53	comes	come	VERB
esrj-117639	150	54	from	from	ADP
esrj-117639	150	55	the	the	DET
esrj-117639	150	56	open	open	ADJ
esrj-117639	150	57	-	-	PUNCT
esrj-117639	150	58	source	source	NOUN
esrj-117639	150	59	data	datum	NOUN
esrj-117639	150	60	in	in	ADP
esrj-117639	150	61	the	the	DET
esrj-117639	150	62	article	article	NOUN
esrj-117639	150	63	by	by	ADP
esrj-117639	150	64	gomez	gomez	PROPN
esrj-117639	150	65	-	-	PUNCT
esrj-117639	150	66	ortiz	ortiz	PROPN
esrj-117639	150	67	and	and	CCONJ
esrj-117639	150	68	agarwal	agarwal	PROPN
esrj-117639	150	69	(	(	PUNCT
esrj-117639	150	70	2005	2005	NUM
esrj-117639	150	71	)	)	PUNCT
esrj-117639	150	72	,	,	PUNCT
esrj-117639	150	73	including	include	VERB
esrj-117639	150	74	a	a	DET
esrj-117639	150	75	subset	subset	NOUN
esrj-117639	150	76	of	of	ADP
esrj-117639	150	77	gravity	gravity	NOUN
esrj-117639	150	78	anomaly	anomaly	NOUN
esrj-117639	150	79	data	datum	NOUN
esrj-117639	150	80	and	and	CCONJ
esrj-117639	150	81	the	the	DET
esrj-117639	150	82	inverted	inverted	ADJ
esrj-117639	150	83	depth	depth	NOUN
esrj-117639	150	84	model	model	NOUN
esrj-117639	150	85	from	from	ADP
esrj-117639	150	86	the	the	DET
esrj-117639	150	87	brittany	brittany	NOUN
esrj-117639	150	88	region	region	NOUN
esrj-117639	150	89	of	of	ADP
esrj-117639	150	90	france	france	PROPN
esrj-117639	150	91	.	.	PUNCT
esrj-117639	151	1	the	the	DET
esrj-117639	151	2	area	area	NOUN
esrj-117639	151	3	of	of	ADP
esrj-117639	151	4	this	this	DET
esrj-117639	151	5	region	region	NOUN
esrj-117639	151	6	is	be	AUX
esrj-117639	151	7	300	300	NUM
esrj-117639	151	8	×	×	NOUN
esrj-117639	151	9	300	300	NUM
esrj-117639	151	10	km	km	NOUN
esrj-117639	151	11	,	,	PUNCT
esrj-117639	151	12	with	with	ADP
esrj-117639	151	13	two	two	NUM
esrj-117639	151	14	discontinuous	discontinuous	ADJ
esrj-117639	151	15	basins	basin	NOUN
esrj-117639	151	16	from	from	ADP
esrj-117639	151	17	northwest	northwest	NOUN
esrj-117639	151	18	to	to	ADP
esrj-117639	151	19	southeast	southeast	ADJ
esrj-117639	151	20	and	and	CCONJ
esrj-117639	151	21	several	several	ADJ
esrj-117639	151	22	dome	dome	NOUN
esrj-117639	151	23	structures	structure	NOUN
esrj-117639	151	24	around	around	ADP
esrj-117639	151	25	them	they	PRON
esrj-117639	151	26	.	.	PUNCT
esrj-117639	152	1	the	the	DET
esrj-117639	152	2	minimum	minimum	ADJ
esrj-117639	152	3	interface	interface	NOUN
esrj-117639	152	4	depth	depth	NOUN
esrj-117639	152	5	is	be	AUX
esrj-117639	152	6	about	about	ADV
esrj-117639	152	7	27.5	27.5	NUM
esrj-117639	152	8	km	km	NOUN
esrj-117639	152	9	and	and	CCONJ
esrj-117639	152	10	the	the	DET
esrj-117639	152	11	maximum	maximum	ADJ
esrj-117639	152	12	interface	interface	NOUN
esrj-117639	152	13	depth	depth	NOUN
esrj-117639	152	14	is	be	AUX
esrj-117639	152	15	about	about	ADV
esrj-117639	152	16	32	32	NUM
esrj-117639	152	17	km	km	NOUN
esrj-117639	152	18	.	.	PUNCT
esrj-117639	153	1	the	the	DET
esrj-117639	153	2	density	density	NOUN
esrj-117639	153	3	difference	difference	NOUN
esrj-117639	153	4	between	between	ADP
esrj-117639	153	5	the	the	DET
esrj-117639	153	6	crust	crust	NOUN
esrj-117639	153	7	and	and	CCONJ
esrj-117639	153	8	mantle	mantle	NOUN
esrj-117639	153	9	is	be	AUX
esrj-117639	153	10	0.4g	0.4g	ADJ
esrj-117639	153	11	/	/	SYM
esrj-117639	153	12	cm3	cm3	NOUN
esrj-117639	153	13	,	,	PUNCT
esrj-117639	153	14	and	and	CCONJ
esrj-117639	153	15	the	the	DET
esrj-117639	153	16	average	average	ADJ
esrj-117639	153	17	depth	depth	NOUN
esrj-117639	153	18	of	of	ADP
esrj-117639	153	19	the	the	DET
esrj-117639	153	20	crust	crust	NOUN
esrj-117639	153	21	-	-	PUNCT
esrj-117639	153	22	mantle	mantle	NOUN
esrj-117639	153	23	boundary	boundary	NOUN
esrj-117639	153	24	is	be	AUX
esrj-117639	153	25	30	30	NUM
esrj-117639	153	26	km	km	NOUN
esrj-117639	153	27	.	.	PUNCT
esrj-117639	154	1	the	the	DET
esrj-117639	154	2	observation	observation	NOUN
esrj-117639	154	3	point	point	NOUN
esrj-117639	154	4	interval	interval	NOUN
esrj-117639	154	5	is	be	AUX
esrj-117639	154	6	6	6	NUM
esrj-117639	154	7	km	km	NOUN
esrj-117639	154	8	,	,	PUNCT
esrj-117639	154	9	and	and	CCONJ
esrj-117639	154	10	the	the	DET
esrj-117639	154	11	underground	underground	ADJ
esrj-117639	154	12	medium	medium	ADJ
esrj-117639	154	13	grid	grid	NOUN
esrj-117639	154	14	is	be	AUX
esrj-117639	154	15	divided	divide	VERB
esrj-117639	154	16	into	into	ADP
esrj-117639	154	17	112	112	NUM
esrj-117639	154	18	×	×	NOUN
esrj-117639	154	19	112	112	NUM
esrj-117639	154	20	with	with	ADP
esrj-117639	154	21	a	a	DET
esrj-117639	154	22	grid	grid	NOUN
esrj-117639	154	23	spacing	spacing	NOUN
esrj-117639	154	24	of	of	ADP
esrj-117639	154	25	6	6	NUM
esrj-117639	154	26	km	km	NOUN
esrj-117639	154	27	.	.	PUNCT
esrj-117639	155	1	the	the	DET
esrj-117639	155	2	training	training	NOUN
esrj-117639	155	3	process	process	NOUN
esrj-117639	155	4	of	of	ADP
esrj-117639	155	5	the	the	DET
esrj-117639	155	6	artificial	artificial	ADJ
esrj-117639	155	7	neural	neural	ADJ
esrj-117639	155	8	network	network	NOUN
esrj-117639	155	9	designed	design	VERB
esrj-117639	155	10	for	for	ADP
esrj-117639	155	11	this	this	DET
esrj-117639	155	12	real	real	ADJ
esrj-117639	155	13	-	-	PUNCT
esrj-117639	155	14	world	world	NOUN
esrj-117639	155	15	data	datum	NOUN
esrj-117639	155	16	also	also	ADV
esrj-117639	155	17	used	use	VERB
esrj-117639	155	18	the	the	DET
esrj-117639	155	19	discrete	discrete	ADJ
esrj-117639	155	20	cosine	cosine	NOUN
esrj-117639	155	21	transform	transform	NOUN
esrj-117639	155	22	to	to	PART
esrj-117639	155	23	compress	compress	VERB
esrj-117639	155	24	the	the	DET
esrj-117639	155	25	data	datum	NOUN
esrj-117639	155	26	matrix	matrix	NOUN
esrj-117639	155	27	.	.	PUNCT
esrj-117639	156	1	in	in	ADP
esrj-117639	156	2	this	this	DET
esrj-117639	156	3	real	real	ADJ
esrj-117639	156	4	-	-	PUNCT
esrj-117639	156	5	world	world	NOUN
esrj-117639	156	6	example	example	NOUN
esrj-117639	156	7	,	,	PUNCT
esrj-117639	156	8	the	the	DET
esrj-117639	156	9	size	size	NOUN
esrj-117639	156	10	of	of	ADP
esrj-117639	156	11	the	the	DET
esrj-117639	156	12	observation	observation	NOUN
esrj-117639	156	13	data	datum	NOUN
esrj-117639	156	14	matrix	matrix	NOUN
esrj-117639	156	15	is	be	AUX
esrj-117639	156	16	112	112	NUM
esrj-117639	156	17	×	×	NOUN
esrj-117639	156	18	112	112	NUM
esrj-117639	156	19	,	,	PUNCT
esrj-117639	156	20	and	and	CCONJ
esrj-117639	156	21	the	the	DET
esrj-117639	156	22	size	size	NOUN
esrj-117639	156	23	of	of	ADP
esrj-117639	156	24	the	the	DET
esrj-117639	156	25	underground	underground	ADJ
esrj-117639	156	26	medium	medium	ADJ
esrj-117639	156	27	grid	grid	NOUN
esrj-117639	156	28	is	be	AUX
esrj-117639	156	29	also	also	ADV
esrj-117639	156	30	112	112	NUM
esrj-117639	156	31	×	×	NOUN
esrj-117639	156	32	112	112	NUM
esrj-117639	156	33	.	.	PUNCT
esrj-117639	157	1	as	as	SCONJ
esrj-117639	157	2	shown	show	VERB
esrj-117639	157	3	in	in	ADP
esrj-117639	157	4	figure	figure	NOUN
esrj-117639	157	5	4(a	4(a	NUM
esrj-117639	157	6	)	)	PUNCT
esrj-117639	157	7	and	and	CCONJ
esrj-117639	157	8	4(b	4(b	NUM
esrj-117639	157	9	)	)	PUNCT
esrj-117639	157	10	,	,	PUNCT
esrj-117639	157	11	the	the	DET
esrj-117639	157	12	energy	energy	NOUN
esrj-117639	157	13	of	of	ADP
esrj-117639	157	14	the	the	DET
esrj-117639	157	15	coefficients	coefficient	NOUN
esrj-117639	157	16	after	after	SCONJ
esrj-117639	157	17	the	the	DET
esrj-117639	157	18	cosine	cosine	NOUN
esrj-117639	157	19	transform	transform	NOUN
esrj-117639	157	20	is	be	AUX
esrj-117639	157	21	mainly	mainly	ADV
esrj-117639	157	22	concentrated	concentrate	VERB
esrj-117639	157	23	in	in	ADP
esrj-117639	157	24	the	the	DET
esrj-117639	157	25	upper	upper	ADJ
esrj-117639	157	26	left	left	ADJ
esrj-117639	157	27	corner	corner	NOUN
esrj-117639	157	28	(	(	PUNCT
esrj-117639	157	29	shown	show	VERB
esrj-117639	157	30	in	in	ADP
esrj-117639	157	31	the	the	DET
esrj-117639	157	32	red	red	PROPN
esrj-117639	157	33	box	box	PROPN
esrj-117639	157	34	,	,	PUNCT
esrj-117639	157	35	at	at	ADP
esrj-117639	157	36	approximately	approximately	ADV
esrj-117639	157	37	position	position	NOUN
esrj-117639	157	38	15	15	NUM
esrj-117639	157	39	×	×	NOUN
esrj-117639	157	40	15	15	NUM
esrj-117639	157	41	)	)	PUNCT
esrj-117639	157	42	,	,	PUNCT
esrj-117639	157	43	and	and	CCONJ
esrj-117639	157	44	most	most	ADJ
esrj-117639	157	45	of	of	ADP
esrj-117639	157	46	the	the	DET
esrj-117639	157	47	other	other	ADJ
esrj-117639	157	48	coefficients	coefficient	NOUN
esrj-117639	157	49	are	be	AUX
esrj-117639	157	50	close	close	ADJ
esrj-117639	157	51	to	to	ADP
esrj-117639	157	52	zero	zero	NUM
esrj-117639	157	53	.	.	PUNCT
esrj-117639	158	1	after	after	ADP
esrj-117639	158	2	applying	apply	VERB
esrj-117639	158	3	the	the	DET
esrj-117639	158	4	discrete	discrete	ADJ
esrj-117639	158	5	cosine	cosine	NOUN
esrj-117639	158	6	transform	transform	NOUN
esrj-117639	158	7	algorithm	algorithm	NOUN
esrj-117639	158	8	,	,	PUNCT
esrj-117639	158	9	we	we	PRON
esrj-117639	158	10	compressed	compress	VERB
esrj-117639	158	11	the	the	DET
esrj-117639	158	12	112	112	NUM
esrj-117639	158	13	×	×	NOUN
esrj-117639	158	14	112	112	NUM
esrj-117639	158	15	data	datum	NOUN
esrj-117639	158	16	matrix	matrix	NOUN
esrj-117639	158	17	into	into	ADP
esrj-117639	158	18	a	a	DET
esrj-117639	158	19	15	15	NUM
esrj-117639	158	20	×	×	NOUN
esrj-117639	158	21	15	15	NUM
esrj-117639	158	22	data	datum	NOUN
esrj-117639	158	23	matrix	matrix	NOUN
esrj-117639	158	24	,	,	PUNCT
esrj-117639	158	25	greatly	greatly	ADV
esrj-117639	158	26	improving	improve	VERB
esrj-117639	158	27	the	the	DET
esrj-117639	158	28	efficiency	efficiency	NOUN
esrj-117639	158	29	of	of	ADP
esrj-117639	158	30	artificial	artificial	ADJ
esrj-117639	158	31	neural	neural	ADJ
esrj-117639	158	32	network	network	NOUN
esrj-117639	158	33	training	training	NOUN
esrj-117639	158	34	.	.	PUNCT
esrj-117639	159	1	after	after	ADP
esrj-117639	159	2	comparison	comparison	NOUN
esrj-117639	159	3	,	,	PUNCT
esrj-117639	159	4	we	we	PRON
esrj-117639	159	5	find	find	VERB
esrj-117639	159	6	that	that	SCONJ
esrj-117639	159	7	training	training	NOUN
esrj-117639	159	8	without	without	ADP
esrj-117639	159	9	discrete	discrete	ADJ
esrj-117639	159	10	cosine	cosine	NOUN
esrj-117639	159	11	transform	transform	NOUN
esrj-117639	159	12	(	(	PUNCT
esrj-117639	159	13	dct	dct	PROPN
esrj-117639	159	14	)	)	PUNCT
esrj-117639	159	15	compression	compression	NOUN
esrj-117639	159	16	takes	take	VERB
esrj-117639	159	17	approximately	approximately	ADV
esrj-117639	159	18	24,000	24,000	NUM
esrj-117639	159	19	seconds	second	NOUN
esrj-117639	159	20	,	,	PUNCT
esrj-117639	159	21	whereas	whereas	SCONJ
esrj-117639	159	22	after	after	ADP
esrj-117639	159	23	compression	compression	NOUN
esrj-117639	159	24	,	,	PUNCT
esrj-117639	159	25	the	the	DET
esrj-117639	159	26	neural	neural	ADJ
esrj-117639	159	27	network	network	NOUN
esrj-117639	159	28	training	training	NOUN
esrj-117639	159	29	only	only	ADV
esrj-117639	159	30	requires	require	VERB
esrj-117639	159	31	about	about	ADP
esrj-117639	159	32	3,600	3,600	NUM
esrj-117639	159	33	seconds	second	NOUN
esrj-117639	159	34	.	.	PUNCT
esrj-117639	160	1	figure	figure	VERB
esrj-117639	160	2	4	4	NUM
esrj-117639	160	3	.	.	NOUN
esrj-117639	160	4	discrete	discrete	ADJ
esrj-117639	160	5	cosine	cosine	NOUN
esrj-117639	160	6	transform	transform	NOUN
esrj-117639	160	7	coefficient	coefficient	NOUN
esrj-117639	160	8	of	of	ADP
esrj-117639	160	9	the	the	DET
esrj-117639	160	10	gomez	gomez	PROPN
esrj-117639	160	11	-	-	PUNCT
esrj-117639	160	12	ortiz	ortiz	PROPN
esrj-117639	160	13	and	and	CCONJ
esrj-117639	160	14	agarwal	agarwal	PROPN
esrj-117639	160	15	’s	’s	PART
esrj-117639	160	16	invert	invert	NOUN
esrj-117639	160	17	depth	depth	NOUN
esrj-117639	160	18	model	model	NOUN
esrj-117639	160	19	(	(	PUNCT
esrj-117639	160	20	a	a	NOUN
esrj-117639	160	21	)	)	PUNCT
esrj-117639	160	22	and	and	CCONJ
esrj-117639	160	23	brittany	brittany	PROPN
esrj-117639	160	24	’s	’s	PART
esrj-117639	160	25	observed	observe	VERB
esrj-117639	160	26	gravity	gravity	NOUN
esrj-117639	160	27	data	datum	NOUN
esrj-117639	160	28	(	(	PUNCT
esrj-117639	160	29	b	b	NOUN
esrj-117639	160	30	)	)	PUNCT
esrj-117639	160	31	figure	figure	NOUN
esrj-117639	160	32	5	5	NUM
esrj-117639	160	33	shows	show	VERB
esrj-117639	160	34	the	the	DET
esrj-117639	160	35	comparison	comparison	NOUN
esrj-117639	160	36	between	between	ADP
esrj-117639	160	37	the	the	DET
esrj-117639	160	38	results	result	NOUN
esrj-117639	160	39	obtained	obtain	VERB
esrj-117639	160	40	by	by	ADP
esrj-117639	160	41	applying	apply	VERB
esrj-117639	160	42	our	our	PRON
esrj-117639	160	43	algorithm	algorithm	NOUN
esrj-117639	160	44	and	and	CCONJ
esrj-117639	160	45	the	the	DET
esrj-117639	160	46	results	result	NOUN
esrj-117639	160	47	in	in	ADP
esrj-117639	160	48	the	the	DET
esrj-117639	160	49	article	article	NOUN
esrj-117639	160	50	by	by	ADP
esrj-117639	160	51	gomez	gomez	PROPN
esrj-117639	160	52	-	-	PUNCT
esrj-117639	160	53	ortiz	ortiz	PROPN
esrj-117639	160	54	and	and	CCONJ
esrj-117639	160	55	agarwal	agarwal	PROPN
esrj-117639	160	56	(	(	PUNCT
esrj-117639	160	57	2005	2005	NUM
esrj-117639	160	58	)	)	PUNCT
esrj-117639	160	59	.	.	PUNCT
esrj-117639	161	1	as	as	SCONJ
esrj-117639	161	2	can	can	AUX
esrj-117639	161	3	be	be	AUX
esrj-117639	161	4	seen	see	VERB
esrj-117639	161	5	from	from	ADP
esrj-117639	161	6	the	the	DET
esrj-117639	161	7	comparison	comparison	NOUN
esrj-117639	161	8	between	between	ADP
esrj-117639	161	9	figure	figure	NOUN
esrj-117639	161	10	5(a	5(a	NUM
esrj-117639	161	11	)	)	PUNCT
esrj-117639	161	12	and	and	CCONJ
esrj-117639	161	13	5(c	5(c	NUM
esrj-117639	161	14	)	)	PUNCT
esrj-117639	161	15	,	,	PUNCT
esrj-117639	161	16	the	the	DET
esrj-117639	161	17	morphology	morphology	NOUN
esrj-117639	161	18	and	and	CCONJ
esrj-117639	161	19	numerical	numerical	ADJ
esrj-117639	161	20	values	value	NOUN
esrj-117639	161	21	of	of	ADP
esrj-117639	161	22	the	the	DET
esrj-117639	161	23	inversion	inversion	NOUN
esrj-117639	161	24	results	result	NOUN
esrj-117639	161	25	obtained	obtain	VERB
esrj-117639	161	26	by	by	ADP
esrj-117639	161	27	our	our	PRON
esrj-117639	161	28	algorithm	algorithm	NOUN
esrj-117639	161	29	are	be	AUX
esrj-117639	161	30	basically	basically	ADV
esrj-117639	161	31	consistent	consistent	ADJ
esrj-117639	161	32	with	with	ADP
esrj-117639	161	33	those	those	PRON
esrj-117639	161	34	in	in	ADP
esrj-117639	161	35	the	the	DET
esrj-117639	161	36	article	article	NOUN
esrj-117639	161	37	by	by	ADP
esrj-117639	161	38	gomez	gomez	PROPN
esrj-117639	161	39	-	-	PUNCT
esrj-117639	161	40	ortiz	ortiz	PROPN
esrj-117639	161	41	and	and	CCONJ
esrj-117639	161	42	agarwal	agarwal	PROPN
esrj-117639	161	43	(	(	PUNCT
esrj-117639	161	44	2005	2005	NUM
esrj-117639	161	45	)	)	PUNCT
esrj-117639	161	46	.	.	PUNCT
esrj-117639	162	1	from	from	ADP
esrj-117639	162	2	figure	figure	NOUN
esrj-117639	162	3	5(b	5(b	NUM
esrj-117639	162	4	)	)	PUNCT
esrj-117639	162	5	and	and	CCONJ
esrj-117639	162	6	5(d	5(d	NUM
esrj-117639	162	7	)	)	PUNCT
esrj-117639	162	8	,	,	PUNCT
esrj-117639	162	9	it	it	PRON
esrj-117639	162	10	can	can	AUX
esrj-117639	162	11	also	also	ADV
esrj-117639	162	12	be	be	AUX
esrj-117639	162	13	seen	see	VERB
esrj-117639	162	14	that	that	SCONJ
esrj-117639	162	15	the	the	DET
esrj-117639	162	16	gravity	gravity	NOUN
esrj-117639	162	17	anomaly	anomaly	NOUN
esrj-117639	162	18	data	datum	NOUN
esrj-117639	162	19	calculated	calculate	VERB
esrj-117639	162	20	from	from	ADP
esrj-117639	162	21	the	the	DET
esrj-117639	162	22	inversion	inversion	NOUN
esrj-117639	162	23	results	result	NOUN
esrj-117639	162	24	obtained	obtain	VERB
esrj-117639	162	25	by	by	ADP
esrj-117639	162	26	our	our	PRON
esrj-117639	162	27	algorithm	algorithm	NOUN
esrj-117639	162	28	also	also	ADV
esrj-117639	162	29	match	match	VERB
esrj-117639	162	30	well	well	ADV
esrj-117639	162	31	with	with	ADP
esrj-117639	162	32	the	the	DET
esrj-117639	162	33	observed	observe	VERB
esrj-117639	162	34	gravity	gravity	NOUN
esrj-117639	162	35	anomaly	anomaly	NOUN
esrj-117639	162	36	data	datum	NOUN
esrj-117639	162	37	.	.	PUNCT
esrj-117639	163	1	figure	figure	NOUN
esrj-117639	163	2	5(e	5(e	NUM
esrj-117639	163	3	)	)	PUNCT
esrj-117639	163	4	and	and	CCONJ
esrj-117639	163	5	5(f	5(f	NUM
esrj-117639	163	6	)	)	PUNCT
esrj-117639	163	7	further	far	ADV
esrj-117639	163	8	demonstrate	demonstrate	VERB
esrj-117639	163	9	the	the	DET
esrj-117639	163	10	practicality	practicality	NOUN
esrj-117639	163	11	of	of	ADP
esrj-117639	163	12	our	our	PRON
esrj-117639	163	13	algorithm	algorithm	NOUN
esrj-117639	163	14	.	.	PUNCT
esrj-117639	164	1	the	the	DET
esrj-117639	164	2	difference	difference	NOUN
esrj-117639	164	3	range	range	NOUN
esrj-117639	164	4	of	of	ADP
esrj-117639	164	5	the	the	DET
esrj-117639	164	6	depth	depth	NOUN
esrj-117639	164	7	model	model	NOUN
esrj-117639	164	8	is	be	AUX
esrj-117639	164	9	approximately	approximately	ADV
esrj-117639	164	10	between	between	ADP
esrj-117639	164	11	-1.5	-1.5	PUNCT
esrj-117639	164	12	~	~	PUNCT
esrj-117639	164	13	2.0	2.0	NUM
esrj-117639	164	14	km	km	NOUN
esrj-117639	164	15	,	,	PUNCT
esrj-117639	164	16	while	while	SCONJ
esrj-117639	164	17	the	the	DET
esrj-117639	164	18	difference	difference	NOUN
esrj-117639	164	19	range	range	NOUN
esrj-117639	164	20	of	of	ADP
esrj-117639	164	21	the	the	DET
esrj-117639	164	22	gravity	gravity	NOUN
esrj-117639	164	23	anomaly	anomaly	NOUN
esrj-117639	164	24	data	datum	NOUN
esrj-117639	164	25	is	be	AUX
esrj-117639	164	26	approximately	approximately	ADV
esrj-117639	164	27	between	between	ADP
esrj-117639	164	28	-3.0	-3.0	ADJ
esrj-117639	164	29	~	~	PUNCT
esrj-117639	164	30	2.0	2.0	NUM
esrj-117639	164	31	mgal	mgal	NOUN
esrj-117639	164	32	.	.	PUNCT
esrj-117639	165	1	4	4	X
esrj-117639	165	2	.	.	X
esrj-117639	165	3	conclusion	conclusion	NOUN
esrj-117639	165	4	in	in	ADP
esrj-117639	165	5	this	this	DET
esrj-117639	165	6	paper	paper	NOUN
esrj-117639	165	7	,	,	PUNCT
esrj-117639	165	8	we	we	PRON
esrj-117639	165	9	introduced	introduce	VERB
esrj-117639	165	10	the	the	DET
esrj-117639	165	11	artificial	artificial	ADJ
esrj-117639	165	12	neural	neural	ADJ
esrj-117639	165	13	network	network	NOUN
esrj-117639	165	14	algorithm	algorithm	NOUN
esrj-117639	165	15	and	and	CCONJ
esrj-117639	165	16	the	the	DET
esrj-117639	165	17	discrete	discrete	ADJ
esrj-117639	165	18	cosine	cosine	NOUN
esrj-117639	165	19	transform	transform	NOUN
esrj-117639	165	20	algorithm	algorithm	NOUN
esrj-117639	165	21	into	into	ADP
esrj-117639	165	22	the	the	DET
esrj-117639	165	23	research	research	NOUN
esrj-117639	165	24	of	of	ADP
esrj-117639	165	25	three	three	NUM
esrj-117639	165	26	-	-	PUNCT
esrj-117639	165	27	dimensional	dimensional	ADJ
esrj-117639	165	28	gravity	gravity	NOUN
esrj-117639	165	29	interface	interface	NOUN
esrj-117639	165	30	inversion	inversion	NOUN
esrj-117639	165	31	.	.	PUNCT
esrj-117639	166	1	by	by	ADP
esrj-117639	166	2	using	use	VERB
esrj-117639	166	3	the	the	DET
esrj-117639	166	4	method	method	NOUN
esrj-117639	166	5	of	of	ADP
esrj-117639	166	6	random	random	ADJ
esrj-117639	166	7	modeling	modeling	NOUN
esrj-117639	166	8	,	,	PUNCT
esrj-117639	166	9	the	the	DET
esrj-117639	166	10	underground	underground	ADJ
esrj-117639	166	11	interface	interface	NOUN
esrj-117639	166	12	geometry	geometry	NOUN
esrj-117639	166	13	shape	shape	NOUN
esrj-117639	166	14	is	be	AUX
esrj-117639	166	15	randomly	randomly	ADV
esrj-117639	166	16	generated	generate	VERB
esrj-117639	166	17	as	as	ADP
esrj-117639	166	18	the	the	DET
esrj-117639	166	19	output	output	NOUN
esrj-117639	166	20	,	,	PUNCT
esrj-117639	166	21	and	and	CCONJ
esrj-117639	166	22	the	the	DET
esrj-117639	166	23	gravity	gravity	NOUN
esrj-117639	166	24	anomaly	anomaly	NOUN
esrj-117639	166	25	data	datum	NOUN
esrj-117639	166	26	calculated	calculate	VERB
esrj-117639	166	27	based	base	VERB
esrj-117639	166	28	on	on	ADP
esrj-117639	166	29	parker	parker	PROPN
esrj-117639	166	30	formula	formula	NOUN
esrj-117639	166	31	using	use	VERB
esrj-117639	166	32	fast	fast	ADJ
esrj-117639	166	33	fourier	fourier	NOUN
esrj-117639	166	34	transform	transform	NOUN
esrj-117639	166	35	is	be	AUX
esrj-117639	166	36	used	use	VERB
esrj-117639	166	37	as	as	ADP
esrj-117639	166	38	the	the	DET
esrj-117639	166	39	input	input	NOUN
esrj-117639	166	40	,	,	PUNCT
esrj-117639	166	41	forming	form	VERB
esrj-117639	166	42	a	a	DET
esrj-117639	166	43	large	large	ADJ
esrj-117639	166	44	-	-	PUNCT
esrj-117639	166	45	scale	scale	NOUN
esrj-117639	166	46	input	input	NOUN
esrj-117639	166	47	-	-	PUNCT
esrj-117639	166	48	output	output	NOUN
esrj-117639	166	49	training	training	NOUN
esrj-117639	166	50	set	set	NOUN
esrj-117639	166	51	for	for	ADP
esrj-117639	166	52	the	the	DET
esrj-117639	166	53	training	training	NOUN
esrj-117639	166	54	process	process	NOUN
esrj-117639	166	55	of	of	ADP
esrj-117639	166	56	the	the	DET
esrj-117639	166	57	artificial	artificial	ADJ
esrj-117639	166	58	neural	neural	ADJ
esrj-117639	166	59	network	network	NOUN
esrj-117639	166	60	.	.	PUNCT
esrj-117639	167	1	in	in	ADP
esrj-117639	167	2	order	order	NOUN
esrj-117639	167	3	to	to	PART
esrj-117639	167	4	solve	solve	VERB
esrj-117639	167	5	the	the	DET
esrj-117639	167	6	problem	problem	NOUN
esrj-117639	167	7	of	of	ADP
esrj-117639	167	8	long	long	ADJ
esrj-117639	167	9	computation	computation	NOUN
esrj-117639	167	10	time	time	NOUN
esrj-117639	167	11	and	and	CCONJ
esrj-117639	167	12	high	high	ADJ
esrj-117639	167	13	memory	memory	NOUN
esrj-117639	167	14	requirements	requirement	NOUN
esrj-117639	167	15	faced	face	VERB
esrj-117639	167	16	by	by	ADP
esrj-117639	167	17	artificial	artificial	ADJ
esrj-117639	167	18	neural	neural	ADJ
esrj-117639	167	19	network	network	NOUN
esrj-117639	167	20	training	training	NOUN
esrj-117639	167	21	on	on	ADP
esrj-117639	167	22	large	large	ADJ
esrj-117639	167	23	data	datum	NOUN
esrj-117639	167	24	sets	set	NOUN
esrj-117639	167	25	,	,	PUNCT
esrj-117639	167	26	a	a	DET
esrj-117639	167	27	data	data	NOUN
esrj-117639	167	28	matrix	matrix	NOUN
esrj-117639	167	29	compression	compression	NOUN
esrj-117639	167	30	algorithm	algorithm	NOUN
esrj-117639	167	31	based	base	VERB
esrj-117639	167	32	on	on	ADP
esrj-117639	167	33	cosine	cosine	NOUN
esrj-117639	167	34	transform	transform	NOUN
esrj-117639	167	35	was	be	AUX
esrj-117639	167	36	introduced	introduce	VERB
esrj-117639	167	37	,	,	PUNCT
esrj-117639	167	38	greatly	greatly	ADV
esrj-117639	167	39	reducing	reduce	VERB
esrj-117639	167	40	the	the	DET
esrj-117639	167	41	computation	computation	NOUN
esrj-117639	167	42	time	time	NOUN
esrj-117639	167	43	,	,	PUNCT
esrj-117639	167	44	reducing	reduce	VERB
esrj-117639	167	45	the	the	DET
esrj-117639	167	46	memory	memory	NOUN
esrj-117639	167	47	requirement	requirement	NOUN
esrj-117639	167	48	,	,	PUNCT
esrj-117639	167	49	and	and	CCONJ
esrj-117639	167	50	improving	improve	VERB
esrj-117639	167	51	the	the	DET
esrj-117639	167	52	computation	computation	NOUN
esrj-117639	167	53	efficiency	efficiency	NOUN
esrj-117639	167	54	.	.	PUNCT
esrj-117639	168	1	the	the	DET
esrj-117639	168	2	trained	train	VERB
esrj-117639	168	3	neural	neural	ADJ
esrj-117639	168	4	network	network	NOUN
esrj-117639	168	5	was	be	AUX
esrj-117639	168	6	used	use	VERB
esrj-117639	168	7	for	for	ADP
esrj-117639	168	8	the	the	DET
esrj-117639	168	9	verification	verification	NOUN
esrj-117639	168	10	of	of	ADP
esrj-117639	168	11	synthetic	synthetic	ADJ
esrj-117639	168	12	and	and	CCONJ
esrj-117639	168	13	real	real	ADJ
esrj-117639	168	14	-	-	PUNCT
esrj-117639	168	15	world	world	NOUN
esrj-117639	168	16	data	datum	NOUN
esrj-117639	168	17	,	,	PUNCT
esrj-117639	168	18	and	and	CCONJ
esrj-117639	168	19	satisfactory	satisfactory	ADJ
esrj-117639	168	20	results	result	NOUN
esrj-117639	168	21	were	be	AUX
esrj-117639	168	22	obtained	obtain	VERB
esrj-117639	168	23	,	,	PUNCT
esrj-117639	168	24	indicating	indicate	VERB
esrj-117639	168	25	that	that	SCONJ
esrj-117639	168	26	the	the	DET
esrj-117639	168	27	trained	train	VERB
esrj-117639	168	28	artificial	artificial	ADJ
esrj-117639	168	29	neural	neural	ADJ
esrj-117639	168	30	network	network	NOUN
esrj-117639	168	31	has	have	VERB
esrj-117639	168	32	good	good	ADJ
esrj-117639	168	33	generalization	generalization	NOUN
esrj-117639	168	34	ability	ability	NOUN
esrj-117639	168	35	and	and	CCONJ
esrj-117639	168	36	can	can	AUX
esrj-117639	168	37	be	be	AUX
esrj-117639	168	38	a	a	DET
esrj-117639	168	39	good	good	ADJ
esrj-117639	168	40	method	method	NOUN
esrj-117639	168	41	for	for	ADP
esrj-117639	168	42	three	three	NUM
esrj-117639	168	43	-	-	PUNCT
esrj-117639	168	44	dimensional	dimensional	ADJ
esrj-117639	168	45	gravity	gravity	NOUN
esrj-117639	168	46	interface	interface	NOUN
esrj-117639	168	47	inversion	inversion	NOUN
esrj-117639	168	48	.	.	PUNCT
esrj-117639	169	1	the	the	DET
esrj-117639	169	2	experimental	experimental	ADJ
esrj-117639	169	3	results	result	NOUN
esrj-117639	169	4	from	from	ADP
esrj-117639	169	5	theoretical	theoretical	ADJ
esrj-117639	169	6	models	model	NOUN
esrj-117639	169	7	and	and	CCONJ
esrj-117639	169	8	actual	actual	ADJ
esrj-117639	169	9	data	datum	NOUN
esrj-117639	169	10	demonstrate	demonstrate	VERB
esrj-117639	169	11	that	that	SCONJ
esrj-117639	169	12	the	the	DET
esrj-117639	169	13	density	density	NOUN
esrj-117639	169	14	interface	interface	NOUN
esrj-117639	169	15	inversion	inversion	NOUN
esrj-117639	169	16	method	method	NOUN
esrj-117639	169	17	based	base	VERB
esrj-117639	169	18	on	on	ADP
esrj-117639	169	19	the	the	DET
esrj-117639	169	20	u	u	ADJ
esrj-117639	169	21	-	-	ADJ
esrj-117639	169	22	net	net	ADJ
esrj-117639	169	23	network	network	NOUN
esrj-117639	169	24	,	,	PUNCT
esrj-117639	169	25	which	which	PRON
esrj-117639	169	26	we	we	PRON
esrj-117639	169	27	have	have	AUX
esrj-117639	169	28	established	establish	VERB
esrj-117639	169	29	in	in	ADP
esrj-117639	169	30	this	this	DET
esrj-117639	169	31	paper	paper	NOUN
esrj-117639	169	32	,	,	PUNCT
esrj-117639	169	33	has	have	VERB
esrj-117639	169	34	superior	superior	ADJ
esrj-117639	169	35	inversion	inversion	NOUN
esrj-117639	169	36	capabilities	capability	NOUN
esrj-117639	169	37	for	for	ADP
esrj-117639	169	38	deep	deep	ADJ
esrj-117639	169	39	and	and	CCONJ
esrj-117639	169	40	gentle	gentle	ADJ
esrj-117639	169	41	density	density	NOUN
esrj-117639	169	42	interface	interface	NOUN
esrj-117639	169	43	undulations	undulation	NOUN
esrj-117639	169	44	.	.	PUNCT
esrj-117639	170	1	for	for	ADP
esrj-117639	170	2	density	density	NOUN
esrj-117639	170	3	interface	interface	NOUN
esrj-117639	170	4	inversion	inversion	NOUN
esrj-117639	170	5	problems	problem	NOUN
esrj-117639	170	6	characterized	characterize	VERB
esrj-117639	170	7	by	by	ADP
esrj-117639	170	8	steep	steep	ADJ
esrj-117639	170	9	interface	interface	NOUN
esrj-117639	170	10	undulations	undulation	NOUN
esrj-117639	170	11	,	,	PUNCT
esrj-117639	170	12	further	further	ADJ
esrj-117639	170	13	methodological	methodological	ADJ
esrj-117639	170	14	design	design	NOUN
esrj-117639	170	15	is	be	AUX
esrj-117639	170	16	required	require	VERB
esrj-117639	170	17	in	in	ADP
esrj-117639	170	18	aspects	aspect	NOUN
esrj-117639	170	19	such	such	ADJ
esrj-117639	170	20	as	as	ADP
esrj-117639	170	21	network	network	NOUN
esrj-117639	170	22	structure	structure	NOUN
esrj-117639	170	23	design	design	NOUN
esrj-117639	170	24	and	and	CCONJ
esrj-117639	170	25	the	the	DET
esrj-117639	170	26	establishment	establishment	NOUN
esrj-117639	170	27	of	of	ADP
esrj-117639	170	28	the	the	DET
esrj-117639	170	29	sample	sample	NOUN
esrj-117639	170	30	set	set	NOUN
esrj-117639	170	31	.	.	PUNCT
esrj-117639	171	1	this	this	PRON
esrj-117639	171	2	will	will	AUX
esrj-117639	171	3	pave	pave	VERB
esrj-117639	171	4	the	the	DET
esrj-117639	171	5	way	way	NOUN
esrj-117639	171	6	for	for	ADP
esrj-117639	171	7	an	an	DET
esrj-117639	171	8	intelligent	intelligent	ADJ
esrj-117639	171	9	and	and	CCONJ
esrj-117639	171	10	universal	universal	ADJ
esrj-117639	171	11	method	method	NOUN
esrj-117639	171	12	for	for	ADP
esrj-117639	171	13	the	the	DET
esrj-117639	171	14	computation	computation	NOUN
esrj-117639	171	15	of	of	ADP
esrj-117639	171	16	density	density	NOUN
esrj-117639	171	17	interfaces	interface	NOUN
esrj-117639	171	18	.	.	PUNCT
esrj-117639	172	1	figure	figure	NOUN
esrj-117639	172	2	5	5	NUM
esrj-117639	172	3	.	.	PUNCT
esrj-117639	172	4	3d	3d	NUM
esrj-117639	172	5	moho	moho	PROPN
esrj-117639	172	6	relief	relief	NOUN
esrj-117639	172	7	in	in	ADP
esrj-117639	172	8	brittany	brittany	PROPN
esrj-117639	172	9	(	(	PUNCT
esrj-117639	172	10	france	france	PROPN
esrj-117639	172	11	)	)	PUNCT
esrj-117639	172	12	(	(	PUNCT
esrj-117639	172	13	a)-gomez	a)-gomez	NOUN
esrj-117639	172	14	-	-	PUNCT
esrj-117639	172	15	ortiz	ortiz	PROPN
esrj-117639	172	16	and	and	CCONJ
esrj-117639	172	17	agarwal	agarwal	PROPN
esrj-117639	172	18	’s	’s	PART
esrj-117639	172	19	invert	invert	NOUN
esrj-117639	172	20	depth	depth	NOUN
esrj-117639	172	21	model	model	NOUN
esrj-117639	172	22	(	(	PUNCT
esrj-117639	172	23	b)-brittany	b)-brittany	NOUN
esrj-117639	172	24	’s	’s	PART
esrj-117639	172	25	observed	observe	VERB
esrj-117639	172	26	gravity	gravity	NOUN
esrj-117639	172	27	data	datum	NOUN
esrj-117639	172	28	.	.	PUNCT
esrj-117639	173	1	(	(	PUNCT
esrj-117639	173	2	c)our	c)our	PROPN
esrj-117639	173	3	program	program	NOUN
esrj-117639	173	4	’s	’s	PART
esrj-117639	173	5	invert	invert	NOUN
esrj-117639	173	6	depth	depth	NOUN
esrj-117639	173	7	model	model	NOUN
esrj-117639	173	8	(	(	PUNCT
esrj-117639	173	9	d)-our	d)-our	DET
esrj-117639	173	10	program	program	NOUN
esrj-117639	173	11	’s	’s	PART
esrj-117639	173	12	calculated	calculate	VERB
esrj-117639	173	13	gravity	gravity	NOUN
esrj-117639	173	14	data	datum	NOUN
esrj-117639	173	15	(	(	PUNCT
esrj-117639	173	16	e)-difference	e)-difference	NOUN
esrj-117639	173	17	of	of	ADP
esrj-117639	173	18	depth	depth	NOUN
esrj-117639	173	19	model	model	NOUN
esrj-117639	173	20	(	(	PUNCT
esrj-117639	173	21	f)-difference	f)-difference	NOUN
esrj-117639	173	22	of	of	ADP
esrj-117639	173	23	gravity	gravity	NOUN
esrj-117639	173	24	data	datum	NOUN
esrj-117639	173	25	references	reference	NOUN
esrj-117639	173	26	adler	adler	PROPN
esrj-117639	173	27	,	,	PUNCT
esrj-117639	173	28	j.	j.	PROPN
esrj-117639	173	29	,	,	PUNCT
esrj-117639	173	30	&	&	CCONJ
esrj-117639	173	31	öktem	öktem	PROPN
esrj-117639	173	32	,	,	PUNCT
esrj-117639	173	33	o.	o.	PROPN
esrj-117639	173	34	(	(	PUNCT
esrj-117639	173	35	2017	2017	NUM
esrj-117639	173	36	)	)	PUNCT
esrj-117639	173	37	.	.	PUNCT
esrj-117639	174	1	solving	solve	VERB
esrj-117639	174	2	ill	ill	ADV
esrj-117639	174	3	-	-	PUNCT
esrj-117639	174	4	posed	pose	VERB
esrj-117639	174	5	inverse	inverse	NOUN
esrj-117639	174	6	problems	problem	NOUN
esrj-117639	174	7	using	use	VERB
esrj-117639	174	8	iterative	iterative	NOUN
esrj-117639	174	9	deep	deep	ADJ
esrj-117639	174	10	neural	neural	ADJ
esrj-117639	174	11	networks	network	NOUN
esrj-117639	174	12	.	.	PUNCT
esrj-117639	175	1	inverse	inverse	NOUN
esrj-117639	175	2	problems	problem	NOUN
esrj-117639	175	3	.	.	PUNCT
esrj-117639	176	1	https://doi.org/10.1088/13616420/aa9581	https://doi.org/10.1088/13616420/aa9581	PROPN
esrj-117639	176	2	ashida	ashida	PROPN
esrj-117639	176	3	,	,	PUNCT
esrj-117639	176	4	y.	y.	PROPN
esrj-117639	176	5	(	(	PUNCT
esrj-117639	176	6	1996	1996	NUM
esrj-117639	176	7	)	)	PUNCT
esrj-117639	176	8	.	.	PUNCT
esrj-117639	177	1	data	datum	NOUN
esrj-117639	177	2	processing	processing	NOUN
esrj-117639	177	3	of	of	ADP
esrj-117639	177	4	reflection	reflection	NOUN
esrj-117639	177	5	seismic	seismic	ADJ
esrj-117639	177	6	data	datum	NOUN
esrj-117639	177	7	by	by	ADP
esrj-117639	177	8	use	use	NOUN
esrj-117639	177	9	of	of	ADP
esrj-117639	177	10	neural	neural	ADJ
esrj-117639	177	11	network	network	NOUN
esrj-117639	177	12	.	.	PUNCT
esrj-117639	178	1	journal	journal	PROPN
esrj-117639	178	2	of	of	ADP
esrj-117639	178	3	applied	apply	VERB
esrj-117639	178	4	geophysics	geophysic	NOUN
esrj-117639	178	5	,	,	PUNCT
esrj-117639	178	6	35(2	35(2	NUM
esrj-117639	178	7	-	-	SYM
esrj-117639	178	8	3	3	NUM
esrj-117639	178	9	)	)	PUNCT
esrj-117639	178	10	,	,	PUNCT
esrj-117639	178	11	89	89	NUM
esrj-117639	178	12	-	-	SYM
esrj-117639	178	13	98	98	NUM
esrj-117639	178	14	.	.	PUNCT
esrj-117639	179	1	https://doi	https://doi	PROPN
esrj-117639	179	2	.	.	PUNCT
esrj-117639	179	3	org/10.1016/0926	org/10.1016/0926	PROPN
esrj-117639	179	4	-	-	PUNCT
esrj-117639	179	5	9851(96)00010	9851(96)00010	NUM
esrj-117639	179	6	-	-	SYM
esrj-117639	179	7	9	9	NUM
esrj-117639	179	8	bott	bott	NOUN
esrj-117639	179	9	,	,	PUNCT
esrj-117639	179	10	m.h.p	m.h.p	PROPN
esrj-117639	179	11	.	.	PROPN
esrj-117639	179	12	(	(	PUNCT
esrj-117639	179	13	1960	1960	NUM
esrj-117639	179	14	)	)	PUNCT
esrj-117639	179	15	.	.	PUNCT
esrj-117639	180	1	the	the	DET
esrj-117639	180	2	use	use	NOUN
esrj-117639	180	3	of	of	ADP
esrj-117639	180	4	rapid	rapid	ADJ
esrj-117639	180	5	digital	digital	ADJ
esrj-117639	180	6	computing	computing	NOUN
esrj-117639	180	7	methods	method	NOUN
esrj-117639	180	8	for	for	ADP
esrj-117639	180	9	direct	direct	ADJ
esrj-117639	180	10	gravity	gravity	NOUN
esrj-117639	180	11	interpretation	interpretation	NOUN
esrj-117639	180	12	of	of	ADP
esrj-117639	180	13	sedimentary	sedimentary	ADJ
esrj-117639	180	14	basins	basin	NOUN
esrj-117639	180	15	.	.	PUNCT
esrj-117639	181	1	geophysical	geophysical	ADJ
esrj-117639	181	2	journal	journal	NOUN
esrj-117639	181	3	of	of	ADP
esrj-117639	181	4	the	the	DET
esrj-117639	181	5	royal	royal	ADJ
esrj-117639	181	6	astronomical	astronomical	ADJ
esrj-117639	181	7	society	society	NOUN
esrj-117639	181	8	,	,	PUNCT
esrj-117639	181	9	3	3	NUM
esrj-117639	181	10	,	,	PUNCT
esrj-117639	181	11	63	63	NUM
esrj-117639	181	12	-	-	SYM
esrj-117639	181	13	67	67	NUM
esrj-117639	181	14	.	.	PUNCT
esrj-117639	182	1	https://doi.org/10.1111/j.1365246x.1960.tb00065.x	https://doi.org/10.1111/j.1365246x.1960.tb00065.x	NOUN
esrj-117639	182	2	https://doi.org/10.1088/1361-6420/aa9581	https://doi.org/10.1088/1361-6420/aa9581	NOUN
esrj-117639	182	3	https://doi.org/10.1088/1361-6420/aa9581	https://doi.org/10.1088/1361-6420/aa9581	X
esrj-117639	182	4	https://doi.org/10.1016/0926-9851(96)00010-9	https://doi.org/10.1016/0926-9851(96)00010-9	PUNCT
esrj-117639	182	5	https://doi.org/10.1016/0926-9851(96)00010-9	https://doi.org/10.1016/0926-9851(96)00010-9	PUNCT
esrj-117639	182	6	https://doi.org/10.1111/j.1365-246x.1960.tb00065.x	https://doi.org/10.1111/j.1365-246x.1960.tb00065.x	NOUN
esrj-117639	182	7	https://doi.org/10.1111/j.1365-246x.1960.tb00065.x	https://doi.org/10.1111/j.1365-246x.1960.tb00065.x	NOUN
esrj-117639	182	8	426	426	NUM
esrj-117639	182	9	huadong	huadong	ADJ
esrj-117639	182	10	song	song	NOUN
esrj-117639	182	11	,	,	PUNCT
esrj-117639	182	12	hui	hui	PROPN
esrj-117639	182	13	fu	fu	PROPN
esrj-117639	182	14	,	,	PUNCT
esrj-117639	182	15	jifeng	jifeng	PROPN
esrj-117639	182	16	qin	qin	PROPN
esrj-117639	182	17	carpenter	carpenter	PROPN
esrj-117639	182	18	,	,	PUNCT
esrj-117639	182	19	c.	c.	NOUN
esrj-117639	182	20	(	(	PUNCT
esrj-117639	182	21	2021	2021	NUM
esrj-117639	182	22	)	)	PUNCT
esrj-117639	182	23	.	.	PUNCT
esrj-117639	183	1	machine	machine	NOUN
esrj-117639	183	2	-	-	PUNCT
esrj-117639	183	3	learning	learn	VERB
esrj-117639	183	4	method	method	NOUN
esrj-117639	183	5	determines	determine	VERB
esrj-117639	183	6	salt	salt	NOUN
esrj-117639	183	7	structures	structure	NOUN
esrj-117639	183	8	from	from	ADP
esrj-117639	183	9	gravity	gravity	NOUN
esrj-117639	183	10	data	datum	NOUN
esrj-117639	183	11	.	.	PUNCT
esrj-117639	184	1	journal	journal	PROPN
esrj-117639	184	2	of	of	ADP
esrj-117639	184	3	petroleum	petroleum	NOUN
esrj-117639	184	4	technology	technology	NOUN
esrj-117639	184	5	,	,	PUNCT
esrj-117639	184	6	73(02	73(02	NOUN
esrj-117639	184	7	)	)	PUNCT
esrj-117639	184	8	,	,	PUNCT
esrj-117639	184	9	70–71	70–71	NOUN
esrj-117639	184	10	.	.	PUNCT
esrj-117639	185	1	https://doi	https://doi	NOUN
esrj-117639	185	2	.	.	PUNCT
esrj-117639	186	1	org/10.2118/0221	org/10.2118/0221	NUM
esrj-117639	186	2	-	-	PUNCT
esrj-117639	186	3	0070	0070	NUM
esrj-117639	186	4	-	-	PUNCT
esrj-117639	186	5	jpt	jpt	PROPN
esrj-117639	186	6	cordell	cordell	PROPN
esrj-117639	186	7	,	,	PUNCT
esrj-117639	186	8	l.	l.	PROPN
esrj-117639	186	9	,	,	PUNCT
esrj-117639	186	10	&	&	CCONJ
esrj-117639	186	11	henderson	henderson	PROPN
esrj-117639	186	12	,	,	PUNCT
esrj-117639	186	13	r.	r.	PROPN
esrj-117639	186	14	g.	g.	PROPN
esrj-117639	186	15	(	(	PUNCT
esrj-117639	186	16	1968	1968	NUM
esrj-117639	186	17	)	)	PUNCT
esrj-117639	186	18	.	.	PUNCT
esrj-117639	187	1	iterative	iterative	VERB
esrj-117639	187	2	three	three	NUM
esrj-117639	187	3	-	-	PUNCT
esrj-117639	187	4	dimensional	dimensional	ADJ
esrj-117639	187	5	solution	solution	NOUN
esrj-117639	187	6	of	of	ADP
esrj-117639	187	7	gravity	gravity	NOUN
esrj-117639	187	8	anomaly	anomaly	NOUN
esrj-117639	187	9	data	datum	NOUN
esrj-117639	187	10	using	use	VERB
esrj-117639	187	11	a	a	DET
esrj-117639	187	12	digital	digital	ADJ
esrj-117639	187	13	computer	computer	NOUN
esrj-117639	187	14	.	.	PUNCT
esrj-117639	188	1	geophysics	geophysic	NOUN
esrj-117639	188	2	,	,	PUNCT
esrj-117639	188	3	33	33	NUM
esrj-117639	188	4	,	,	PUNCT
esrj-117639	188	5	596	596	NUM
esrj-117639	188	6	-	-	SYM
esrj-117639	188	7	601	601	NUM
esrj-117639	188	8	.	.	PUNCT
esrj-117639	188	9	https://doi.org/10.1190/1.1439955	https://doi.org/10.1190/1.1439955	PROPN
esrj-117639	188	10	dai	dai	PROPN
esrj-117639	188	11	,	,	PUNCT
esrj-117639	188	12	h.	h.	PROPN
esrj-117639	188	13	,	,	PUNCT
esrj-117639	188	14	&	&	CCONJ
esrj-117639	188	15	macbeth	macbeth	PROPN
esrj-117639	188	16	,	,	PUNCT
esrj-117639	188	17	c.	c.	PROPN
esrj-117639	188	18	(	(	PUNCT
esrj-117639	188	19	1995	1995	NUM
esrj-117639	188	20	)	)	PUNCT
esrj-117639	188	21	.	.	PUNCT
esrj-117639	189	1	automatic	automatic	ADJ
esrj-117639	189	2	picking	picking	NOUN
esrj-117639	189	3	of	of	ADP
esrj-117639	189	4	seismic	seismic	ADJ
esrj-117639	189	5	arrivals	arrival	NOUN
esrj-117639	189	6	in	in	ADP
esrj-117639	189	7	local	local	ADJ
esrj-117639	189	8	earthquake	earthquake	NOUN
esrj-117639	189	9	data	datum	NOUN
esrj-117639	189	10	using	use	VERB
esrj-117639	189	11	an	an	DET
esrj-117639	189	12	artificial	artificial	ADJ
esrj-117639	189	13	neural	neural	ADJ
esrj-117639	189	14	network	network	NOUN
esrj-117639	189	15	.	.	PUNCT
esrj-117639	190	1	geophysical	geophysical	ADJ
esrj-117639	190	2	journal	journal	PROPN
esrj-117639	190	3	international	international	PROPN
esrj-117639	190	4	,	,	PUNCT
esrj-117639	190	5	120(3	120(3	NUM
esrj-117639	190	6	)	)	PUNCT
esrj-117639	190	7	,	,	PUNCT
esrj-117639	190	8	758	758	NUM
esrj-117639	190	9	-	-	SYM
esrj-117639	190	10	774	774	NUM
esrj-117639	190	11	.	.	PUNCT
esrj-117639	191	1	https://doi.org/10.1111/j.1365-246x.1995	https://doi.org/10.1111/j.1365-246x.1995	PROPN
esrj-117639	191	2	.	.	PUNCT
esrj-117639	192	1	tb01851.x	tb01851.x	NOUN
esrj-117639	192	2	gómez	gómez	NOUN
esrj-117639	192	3	-	-	PUNCT
esrj-117639	192	4	ortiz	ortiz	PROPN
esrj-117639	192	5	,	,	PUNCT
esrj-117639	192	6	d.	d.	PROPN
esrj-117639	192	7	,	,	PUNCT
esrj-117639	192	8	&	&	CCONJ
esrj-117639	192	9	agarwal	agarwal	PROPN
esrj-117639	192	10	,	,	PUNCT
esrj-117639	192	11	b.	b.	PROPN
esrj-117639	192	12	n.	n.	PROPN
esrj-117639	192	13	(	(	PUNCT
esrj-117639	192	14	2005	2005	NUM
esrj-117639	192	15	)	)	PUNCT
esrj-117639	192	16	.	.	PUNCT
esrj-117639	193	1	3dinver.m	3dinver.m	NUM
esrj-117639	193	2	:	:	PUNCT
esrj-117639	193	3	a	a	DET
esrj-117639	193	4	matlab	matlab	PROPN
esrj-117639	193	5	program	program	NOUN
esrj-117639	193	6	to	to	PART
esrj-117639	193	7	invert	invert	VERB
esrj-117639	193	8	the	the	DET
esrj-117639	193	9	gravity	gravity	NOUN
esrj-117639	193	10	anomaly	anomaly	NOUN
esrj-117639	193	11	over	over	ADP
esrj-117639	193	12	a	a	DET
esrj-117639	193	13	3d	3d	NUM
esrj-117639	193	14	horizontal	horizontal	ADJ
esrj-117639	193	15	density	density	NOUN
esrj-117639	193	16	interface	interface	NOUN
esrj-117639	193	17	by	by	ADP
esrj-117639	193	18	parker	parker	PROPN
esrj-117639	193	19	–	–	PUNCT
esrj-117639	193	20	oldenburg	oldenburg	PROPN
esrj-117639	193	21	’s	’s	PART
esrj-117639	193	22	algorithm	algorithm	NOUN
esrj-117639	193	23	.	.	PUNCT
esrj-117639	194	1	computers	computer	NOUN
esrj-117639	194	2	&	&	CCONJ
esrj-117639	194	3	geosciences	geoscience	NOUN
esrj-117639	194	4	,	,	PUNCT
esrj-117639	194	5	31(4	31(4	NUM
esrj-117639	194	6	)	)	PUNCT
esrj-117639	194	7	,	,	PUNCT
esrj-117639	194	8	513520	513520	NUM
esrj-117639	194	9	.	.	PUNCT
esrj-117639	195	1	https://doi.org/10.1016/j.cageo.2004.11.004	https://doi.org/10.1016/j.cageo.2004.11.004	PROPN
esrj-117639	195	2	granser	granser	PROPN
esrj-117639	195	3	,	,	PUNCT
esrj-117639	195	4	h.	h.	PROPN
esrj-117639	195	5	(	(	PUNCT
esrj-117639	195	6	1987a	1987a	NUM
esrj-117639	195	7	)	)	PUNCT
esrj-117639	195	8	.	.	PUNCT
esrj-117639	196	1	three	three	NUM
esrj-117639	196	2	-	-	PUNCT
esrj-117639	196	3	dimensional	dimensional	ADJ
esrj-117639	196	4	interpretation	interpretation	NOUN
esrj-117639	196	5	of	of	ADP
esrj-117639	196	6	gravity	gravity	NOUN
esrj-117639	196	7	data	datum	NOUN
esrj-117639	196	8	from	from	ADP
esrj-117639	196	9	sedimentary	sedimentary	ADJ
esrj-117639	196	10	basins	basin	NOUN
esrj-117639	196	11	using	use	VERB
esrj-117639	196	12	an	an	DET
esrj-117639	196	13	exponential	exponential	ADJ
esrj-117639	196	14	density	density	NOUN
esrj-117639	196	15	-	-	PUNCT
esrj-117639	196	16	depth	depth	NOUN
esrj-117639	196	17	function	function	NOUN
esrj-117639	196	18	.	.	PUNCT
esrj-117639	197	1	geophysical	geophysical	ADJ
esrj-117639	197	2	prospecting	prospecting	NOUN
esrj-117639	197	3	,	,	PUNCT
esrj-117639	197	4	35	35	NUM
esrj-117639	197	5	,	,	PUNCT
esrj-117639	197	6	1030	1030	NUM
esrj-117639	197	7	-	-	SYM
esrj-117639	197	8	1041	1041	NUM
esrj-117639	197	9	.	.	PUNCT
esrj-117639	198	1	https://doi.org/10.1111/j.1365-2478.1987	https://doi.org/10.1111/j.1365-2478.1987	PROPN
esrj-117639	198	2	.	.	PUNCT
esrj-117639	199	1	tb00858.x	tb00858.x	PROPN
esrj-117639	199	2	granser	granser	PROPN
esrj-117639	199	3	,	,	PUNCT
esrj-117639	199	4	h.	h.	PROPN
esrj-117639	199	5	(	(	PUNCT
esrj-117639	199	6	1987b	1987b	NUM
esrj-117639	199	7	)	)	PUNCT
esrj-117639	199	8	.	.	PUNCT
esrj-117639	200	1	nonlinear	nonlinear	ADJ
esrj-117639	200	2	inversion	inversion	NOUN
esrj-117639	200	3	of	of	ADP
esrj-117639	200	4	gravity	gravity	NOUN
esrj-117639	200	5	data	datum	NOUN
esrj-117639	200	6	using	use	VERB
esrj-117639	200	7	the	the	DET
esrj-117639	200	8	schmidt	schmidt	ADJ
esrj-117639	200	9	-	-	PUNCT
esrj-117639	200	10	lichtenstein	lichtenstein	PROPN
esrj-117639	200	11	approach	approach	NOUN
esrj-117639	200	12	.	.	PUNCT
esrj-117639	201	1	geophysics	geophysic	NOUN
esrj-117639	201	2	,	,	PUNCT
esrj-117639	201	3	52	52	NUM
esrj-117639	201	4	,	,	PUNCT
esrj-117639	201	5	88	88	NUM
esrj-117639	201	6	-	-	SYM
esrj-117639	201	7	93	93	NUM
esrj-117639	201	8	.	.	PUNCT
esrj-117639	202	1	https://doi.org/10.1190/1.1442243	https://doi.org/10.1190/1.1442243	PROPN
esrj-117639	202	2	jin	jin	PROPN
esrj-117639	202	3	,	,	PUNCT
esrj-117639	202	4	k.	k.	PROPN
esrj-117639	202	5	h.	h.	PROPN
esrj-117639	202	6	,	,	PUNCT
esrj-117639	202	7	mccann	mccann	PROPN
esrj-117639	202	8	,	,	PUNCT
esrj-117639	202	9	m.	m.	NOUN
esrj-117639	202	10	t.	t.	PROPN
esrj-117639	202	11	,	,	PUNCT
esrj-117639	202	12	froustey	froustey	PROPN
esrj-117639	202	13	,	,	PUNCT
esrj-117639	202	14	e.	e.	PROPN
esrj-117639	202	15	,	,	PUNCT
esrj-117639	202	16	&	&	CCONJ
esrj-117639	202	17	unser	unser	PROPN
esrj-117639	202	18	,	,	PUNCT
esrj-117639	202	19	m.	m.	NOUN
esrj-117639	202	20	(	(	PUNCT
esrj-117639	202	21	2017	2017	NUM
esrj-117639	202	22	)	)	PUNCT
esrj-117639	202	23	.	.	PUNCT
esrj-117639	203	1	deep	deep	ADJ
esrj-117639	203	2	convolutional	convolutional	ADJ
esrj-117639	203	3	neural	neural	ADJ
esrj-117639	203	4	network	network	NOUN
esrj-117639	203	5	for	for	ADP
esrj-117639	203	6	inverse	inverse	NOUN
esrj-117639	203	7	problems	problem	NOUN
esrj-117639	203	8	in	in	ADP
esrj-117639	203	9	imaging	imaging	NOUN
esrj-117639	203	10	.	.	PUNCT
esrj-117639	204	1	ieee	ieee	NOUN
esrj-117639	204	2	transactions	transaction	NOUN
esrj-117639	204	3	on	on	ADP
esrj-117639	204	4	image	image	NOUN
esrj-117639	204	5	processing	processing	NOUN
esrj-117639	204	6	,	,	PUNCT
esrj-117639	204	7	26(9	26(9	NUM
esrj-117639	204	8	)	)	PUNCT
esrj-117639	204	9	,	,	PUNCT
esrj-117639	204	10	4509–4522	4509–4522	NUM
esrj-117639	204	11	.	.	PUNCT
esrj-117639	205	1	doi	doi	NOUN
esrj-117639	205	2	:	:	PUNCT
esrj-117639	205	3	10.1109	10.1109	NUM
esrj-117639	205	4	/	/	SYM
esrj-117639	205	5	tip.2017.2713099	tip.2017.2713099	PROPN
esrj-117639	205	6	lelièvre	lelièvre	PROPN
esrj-117639	205	7	peter	peter	PROPN
esrj-117639	205	8	,	,	PUNCT
esrj-117639	205	9	g.	g.	PROPN
esrj-117639	205	10	,	,	PUNCT
esrj-117639	205	11	farquharson	farquharson	PROPN
esrj-117639	205	12	colin	colin	PROPN
esrj-117639	205	13	,	,	PUNCT
esrj-117639	205	14	g.	g.	PROPN
esrj-117639	205	15	,	,	PUNCT
esrj-117639	205	16	&	&	CCONJ
esrj-117639	205	17	hurich	hurich	PROPN
esrj-117639	205	18	charles	charles	PROPN
esrj-117639	205	19	,	,	PUNCT
esrj-117639	205	20	a.	a.	NOUN
esrj-117639	205	21	(	(	PUNCT
esrj-117639	205	22	2012	2012	NUM
esrj-117639	205	23	)	)	PUNCT
esrj-117639	205	24	.	.	PUNCT
esrj-117639	206	1	joint	joint	ADJ
esrj-117639	206	2	inversion	inversion	NOUN
esrj-117639	206	3	of	of	ADP
esrj-117639	206	4	seismic	seismic	ADJ
esrj-117639	206	5	traveltimes	traveltime	NOUN
esrj-117639	206	6	and	and	CCONJ
esrj-117639	206	7	gravity	gravity	NOUN
esrj-117639	206	8	data	datum	NOUN
esrj-117639	206	9	on	on	ADP
esrj-117639	206	10	unstructured	unstructured	ADJ
esrj-117639	206	11	grids	grid	NOUN
esrj-117639	206	12	with	with	ADP
esrj-117639	206	13	application	application	NOUN
esrj-117639	206	14	to	to	ADP
esrj-117639	206	15	mineral	mineral	NOUN
esrj-117639	206	16	exploration	exploration	NOUN
esrj-117639	206	17	.	.	PUNCT
esrj-117639	207	1	geophysics	geophysic	NOUN
esrj-117639	207	2	,	,	PUNCT
esrj-117639	207	3	77(1	77(1	NUM
esrj-117639	207	4	)	)	PUNCT
esrj-117639	207	5	,	,	PUNCT
esrj-117639	207	6	k1	k1	NOUN
esrj-117639	207	7	-	-	PUNCT
esrj-117639	207	8	k15	k15	NOUN
esrj-117639	207	9	.	.	PUNCT
esrj-117639	208	1	https://doi.org/10.1190/geo2011-0154.1	https://doi.org/10.1190/geo2011-0154.1	PROPN
esrj-117639	209	1	li	li	PROPN
esrj-117639	209	2	,	,	PUNCT
esrj-117639	209	3	y.	y.	PROPN
esrj-117639	209	4	,	,	PUNCT
esrj-117639	209	5	jia	jia	PROPN
esrj-117639	209	6	,	,	PUNCT
esrj-117639	209	7	z.	z.	PROPN
esrj-117639	209	8	,	,	PUNCT
esrj-117639	209	9	&	&	CCONJ
esrj-117639	209	10	lu	lu	PROPN
esrj-117639	209	11	,	,	PUNCT
esrj-117639	209	12	w.	w.	PROPN
esrj-117639	209	13	(	(	PUNCT
esrj-117639	209	14	2022	2022	NUM
esrj-117639	209	15	)	)	PUNCT
esrj-117639	209	16	.	.	PUNCT
esrj-117639	210	1	self	self	NOUN
esrj-117639	210	2	-	-	PUNCT
esrj-117639	210	3	supervised	supervise	VERB
esrj-117639	210	4	deep	deep	ADJ
esrj-117639	210	5	learning	learning	NOUN
esrj-117639	210	6	for	for	ADP
esrj-117639	210	7	3d	3d	NUM
esrj-117639	210	8	gravity	gravity	NOUN
esrj-117639	210	9	inversion	inversion	NOUN
esrj-117639	210	10	.	.	PUNCT
esrj-117639	211	1	ieee	ieee	NOUN
esrj-117639	211	2	transactions	transaction	NOUN
esrj-117639	211	3	on	on	ADP
esrj-117639	211	4	geoscience	geoscience	NOUN
esrj-117639	211	5	and	and	CCONJ
esrj-117639	211	6	remote	remote	ADJ
esrj-117639	211	7	sensing	sensing	NOUN
esrj-117639	211	8	,	,	PUNCT
esrj-117639	211	9	60	60	NUM
esrj-117639	211	10	,	,	PUNCT
esrj-117639	211	11	1	1	NUM
esrj-117639	211	12	-	-	SYM
esrj-117639	211	13	1	1	NUM
esrj-117639	211	14	.	.	PUNCT
esrj-117639	211	15	doi	doi	NOUN
esrj-117639	211	16	:	:	PUNCT
esrj-117639	211	17	10.1109	10.1109	NUM
esrj-117639	211	18	/	/	SYM
esrj-117639	211	19	tgrs.2022.3225449	tgrs.2022.3225449	PROPN
esrj-117639	211	20	ma	ma	PROPN
esrj-117639	211	21	,	,	PUNCT
esrj-117639	211	22	g.	g.	PROPN
esrj-117639	211	23	,	,	PUNCT
esrj-117639	211	24	niu	niu	PROPN
esrj-117639	211	25	,	,	PUNCT
esrj-117639	211	26	r.	r.	PROPN
esrj-117639	211	27	,	,	PUNCT
esrj-117639	211	28	gao	gao	PROPN
esrj-117639	211	29	,	,	PUNCT
esrj-117639	211	30	t.	t.	PROPN
esrj-117639	211	31	,	,	PUNCT
esrj-117639	211	32	li	li	PROPN
esrj-117639	211	33	,	,	PUNCT
esrj-117639	211	34	l.	l.	PROPN
esrj-117639	211	35	,	,	PUNCT
esrj-117639	211	36	wang	wang	PROPN
esrj-117639	211	37	,	,	PUNCT
esrj-117639	211	38	t.	t.	PROPN
esrj-117639	211	39	,	,	PUNCT
esrj-117639	211	40	&	&	CCONJ
esrj-117639	211	41	meng	meng	PROPN
esrj-117639	211	42	,	,	PUNCT
esrj-117639	211	43	q.	q.	PROPN
esrj-117639	211	44	(	(	PUNCT
esrj-117639	211	45	2022	2022	NUM
esrj-117639	211	46	)	)	PUNCT
esrj-117639	211	47	.	.	PUNCT
esrj-117639	212	1	high	high	ADJ
esrj-117639	212	2	-	-	PUNCT
esrj-117639	212	3	efficiency	efficiency	NOUN
esrj-117639	212	4	gravity	gravity	NOUN
esrj-117639	212	5	data	data	NOUN
esrj-117639	212	6	inversion	inversion	NOUN
esrj-117639	212	7	method	method	NOUN
esrj-117639	212	8	based	base	VERB
esrj-117639	212	9	on	on	ADP
esrj-117639	212	10	locally	locally	ADV
esrj-117639	212	11	adaptive	adaptive	ADJ
esrj-117639	212	12	unstructured	unstructured	ADJ
esrj-117639	212	13	meshing	meshing	NOUN
esrj-117639	212	14	.	.	PUNCT
esrj-117639	213	1	ieee	ieee	NOUN
esrj-117639	213	2	transactions	transaction	NOUN
esrj-117639	213	3	on	on	ADP
esrj-117639	213	4	geoscience	geoscience	NOUN
esrj-117639	213	5	and	and	CCONJ
esrj-117639	213	6	remote	remote	ADJ
esrj-117639	213	7	sensing	sensing	NOUN
esrj-117639	213	8	,	,	PUNCT
esrj-117639	213	9	60	60	NUM
esrj-117639	213	10	.	.	PUNCT
esrj-117639	214	1	doi	doi	NOUN
esrj-117639	214	2	:	:	PUNCT
esrj-117639	214	3	10.1109	10.1109	NUM
esrj-117639	214	4	/	/	SYM
esrj-117639	214	5	tgrs.2022.3142042	tgrs.2022.3142042	PROPN
esrj-117639	214	6	nagendra	nagendra	PROPN
esrj-117639	214	7	,	,	PUNCT
esrj-117639	214	8	r.	r.	PROPN
esrj-117639	214	9	,	,	PUNCT
esrj-117639	214	10	prasad	prasad	PROPN
esrj-117639	214	11	,	,	PUNCT
esrj-117639	214	12	p.	p.	NOUN
esrj-117639	214	13	,	,	PUNCT
esrj-117639	214	14	&	&	CCONJ
esrj-117639	214	15	bhimasankaram	bhimasankaram	PROPN
esrj-117639	214	16	,	,	PUNCT
esrj-117639	214	17	v.	v.	PROPN
esrj-117639	214	18	(	(	PUNCT
esrj-117639	214	19	1996	1996	NUM
esrj-117639	214	20	)	)	PUNCT
esrj-117639	214	21	.	.	PUNCT
esrj-117639	215	1	forward	forward	ADV
esrj-117639	215	2	and	and	CCONJ
esrj-117639	215	3	inverse	inverse	NOUN
esrj-117639	215	4	computer	computer	NOUN
esrj-117639	215	5	modeling	modeling	NOUN
esrj-117639	215	6	of	of	ADP
esrj-117639	215	7	a	a	DET
esrj-117639	215	8	gravity	gravity	NOUN
esrj-117639	215	9	field	field	NOUN
esrj-117639	215	10	resulting	result	VERB
esrj-117639	215	11	from	from	ADP
esrj-117639	215	12	a	a	DET
esrj-117639	215	13	density	density	NOUN
esrj-117639	215	14	interface	interface	NOUN
esrj-117639	215	15	using	use	VERB
esrj-117639	215	16	parker	parker	PROPN
esrj-117639	215	17	-	-	PUNCT
esrj-117639	215	18	oldenberg	oldenberg	PROPN
esrj-117639	215	19	method	method	NOUN
esrj-117639	215	20	.	.	PUNCT
esrj-117639	216	1	computers	computer	NOUN
esrj-117639	216	2	&	&	CCONJ
esrj-117639	216	3	geosciences	geoscience	NOUN
esrj-117639	216	4	,	,	PUNCT
esrj-117639	216	5	22(3	22(3	NUM
esrj-117639	216	6	)	)	PUNCT
esrj-117639	216	7	,	,	PUNCT
esrj-117639	216	8	227	227	NUM
esrj-117639	216	9	-	-	SYM
esrj-117639	216	10	237	237	NUM
esrj-117639	216	11	.	.	PUNCT
esrj-117639	217	1	https://doi.org/10.1016/0098-3004(95)00075-5	https://doi.org/10.1016/0098-3004(95)00075-5	PROPN
esrj-117639	217	2	oldenburg	oldenburg	PROPN
esrj-117639	217	3	,	,	PUNCT
esrj-117639	217	4	d.	d.	PROPN
esrj-117639	217	5	w.	w.	PROPN
esrj-117639	217	6	(	(	PUNCT
esrj-117639	217	7	1974	1974	NUM
esrj-117639	217	8	)	)	PUNCT
esrj-117639	217	9	.	.	PUNCT
esrj-117639	218	1	the	the	DET
esrj-117639	218	2	inversion	inversion	NOUN
esrj-117639	218	3	and	and	CCONJ
esrj-117639	218	4	interpretation	interpretation	NOUN
esrj-117639	218	5	of	of	ADP
esrj-117639	218	6	gravity	gravity	NOUN
esrj-117639	218	7	anomalies	anomaly	NOUN
esrj-117639	218	8	.	.	PUNCT
esrj-117639	219	1	geophysics	geophysic	NOUN
esrj-117639	219	2	,	,	PUNCT
esrj-117639	219	3	39(4	39(4	NOUN
esrj-117639	219	4	)	)	PUNCT
esrj-117639	219	5	,	,	PUNCT
esrj-117639	219	6	526	526	NUM
esrj-117639	219	7	-	-	SYM
esrj-117639	219	8	536	536	NUM
esrj-117639	219	9	.	.	PUNCT
esrj-117639	220	1	https://doi.org/10.1190/1.1440444	https://doi.org/10.1190/1.1440444	PROPN
esrj-117639	220	2	parker	parker	PROPN
esrj-117639	220	3	,	,	PUNCT
esrj-117639	220	4	r.	r.	PROPN
esrj-117639	220	5	l.	l.	PROPN
esrj-117639	220	6	(	(	PUNCT
esrj-117639	220	7	1972	1972	NUM
esrj-117639	220	8	)	)	PUNCT
esrj-117639	220	9	.	.	PUNCT
esrj-117639	221	1	the	the	DET
esrj-117639	221	2	rapid	rapid	ADJ
esrj-117639	221	3	calculation	calculation	NOUN
esrj-117639	221	4	of	of	ADP
esrj-117639	221	5	potential	potential	ADJ
esrj-117639	221	6	anomalies	anomaly	NOUN
esrj-117639	221	7	.	.	PUNCT
esrj-117639	222	1	geophysical	geophysical	ADJ
esrj-117639	222	2	journal	journal	NOUN
esrj-117639	222	3	of	of	ADP
esrj-117639	222	4	the	the	DET
esrj-117639	222	5	royal	royal	ADJ
esrj-117639	222	6	astronomical	astronomical	ADJ
esrj-117639	222	7	society	society	NOUN
esrj-117639	222	8	,	,	PUNCT
esrj-117639	222	9	31	31	NUM
esrj-117639	222	10	,	,	PUNCT
esrj-117639	222	11	447	447	NUM
esrj-117639	222	12	-	-	SYM
esrj-117639	222	13	455	455	NUM
esrj-117639	222	14	.	.	PUNCT
esrj-117639	223	1	https://doi.org/10.1111/j.1365-246x.1973.tb06513.x	https://doi.org/10.1111/j.1365-246x.1973.tb06513.x	X
esrj-117639	224	1	poulton	poulton	PROPN
esrj-117639	224	2	,	,	PUNCT
esrj-117639	224	3	m.	m.	NOUN
esrj-117639	224	4	,	,	PUNCT
esrj-117639	224	5	&	&	CCONJ
esrj-117639	224	6	el	el	PROPN
esrj-117639	224	7	-	-	PUNCT
esrj-117639	224	8	fouly	fouly	ADV
esrj-117639	224	9	,	,	PUNCT
esrj-117639	224	10	a.	a.	NOUN
esrj-117639	224	11	(	(	PUNCT
esrj-117639	224	12	1991	1991	NUM
esrj-117639	224	13	)	)	PUNCT
esrj-117639	224	14	.	.	PUNCT
esrj-117639	225	1	preprocessing	preprocesse	VERB
esrj-117639	225	2	gpr	gpr	PROPN
esrj-117639	225	3	signatures	signature	NOUN
esrj-117639	225	4	for	for	ADP
esrj-117639	225	5	cascading	cascade	VERB
esrj-117639	225	6	neural	neural	ADJ
esrj-117639	225	7	network	network	NOUN
esrj-117639	225	8	classification	classification	NOUN
esrj-117639	225	9	.	.	PUNCT
esrj-117639	226	1	61st	61st	PROPN
esrj-117639	226	2	seg	seg	PROPN
esrj-117639	226	3	meeting	meeting	PROPN
esrj-117639	226	4	,	,	PUNCT
esrj-117639	226	5	houston	houston	PROPN
esrj-117639	226	6	,	,	PUNCT
esrj-117639	226	7	usa	usa	PROPN
esrj-117639	226	8	,	,	PUNCT
esrj-117639	226	9	expanded	expand	VERB
esrj-117639	226	10	abstracts	abstract	NOUN
esrj-117639	226	11	,	,	PUNCT
esrj-117639	226	12	507–509	507–509	NUM
esrj-117639	226	13	.	.	PUNCT
esrj-117639	227	1	https://doi.org/10.1190/1.1888789	https://doi.org/10.1190/1.1888789	PROPN
esrj-117639	227	2	poulton	poulton	PROPN
esrj-117639	227	3	,	,	PUNCT
esrj-117639	227	4	m.	m.	NOUN
esrj-117639	227	5	m.	m.	NOUN
esrj-117639	227	6	,	,	PUNCT
esrj-117639	227	7	sternberg	sternberg	PROPN
esrj-117639	227	8	,	,	PUNCT
esrj-117639	227	9	b.	b.	PROPN
esrj-117639	227	10	k.	k.	PROPN
esrj-117639	227	11	,	,	PUNCT
esrj-117639	227	12	&	&	CCONJ
esrj-117639	227	13	glass	glass	NOUN
esrj-117639	227	14	,	,	PUNCT
esrj-117639	227	15	c.	c.	PROPN
esrj-117639	227	16	e.	e.	PROPN
esrj-117639	227	17	(	(	PUNCT
esrj-117639	227	18	1992	1992	NUM
esrj-117639	227	19	)	)	PUNCT
esrj-117639	227	20	.	.	PUNCT
esrj-117639	228	1	neural	neural	ADJ
esrj-117639	228	2	network	network	NOUN
esrj-117639	228	3	pattern	pattern	NOUN
esrj-117639	228	4	recognition	recognition	NOUN
esrj-117639	228	5	of	of	ADP
esrj-117639	228	6	subsurface	subsurface	NOUN
esrj-117639	228	7	em	em	PRON
esrj-117639	228	8	images	image	NOUN
esrj-117639	228	9	.	.	PUNCT
esrj-117639	229	1	journal	journal	NOUN
esrj-117639	229	2	of	of	ADP
esrj-117639	229	3	applied	apply	VERB
esrj-117639	229	4	geophysics	geophysic	NOUN
esrj-117639	229	5	,	,	PUNCT
esrj-117639	229	6	29(1	29(1	NUM
esrj-117639	229	7	)	)	PUNCT
esrj-117639	229	8	,	,	PUNCT
esrj-117639	229	9	21	21	NUM
esrj-117639	229	10	-	-	SYM
esrj-117639	229	11	36	36	NUM
esrj-117639	229	12	.	.	PUNCT
esrj-117639	230	1	https://doi.org/10.1016/0926-9851(92)90010-i	https://doi.org/10.1016/0926-9851(92)90010-i	PROPN
esrj-117639	230	2	raiche	raiche	NOUN
esrj-117639	230	3	,	,	PUNCT
esrj-117639	230	4	a.	a.	NOUN
esrj-117639	230	5	(	(	PUNCT
esrj-117639	230	6	1991	1991	NUM
esrj-117639	230	7	)	)	PUNCT
esrj-117639	230	8	.	.	PUNCT
esrj-117639	231	1	a	a	DET
esrj-117639	231	2	pattern	pattern	NOUN
esrj-117639	231	3	recognition	recognition	NOUN
esrj-117639	231	4	approach	approach	NOUN
esrj-117639	231	5	to	to	ADP
esrj-117639	231	6	geophysical	geophysical	ADJ
esrj-117639	231	7	inversion	inversion	NOUN
esrj-117639	231	8	using	use	VERB
esrj-117639	231	9	neural	neural	ADJ
esrj-117639	231	10	networks	network	NOUN
esrj-117639	231	11	.	.	PUNCT
esrj-117639	232	1	geophysical	geophysical	ADJ
esrj-117639	232	2	journal	journal	PROPN
esrj-117639	232	3	international	international	PROPN
esrj-117639	232	4	,	,	PUNCT
esrj-117639	232	5	105	105	NUM
esrj-117639	232	6	,	,	PUNCT
esrj-117639	232	7	629–648	629–648	NUM
esrj-117639	232	8	.	.	PUNCT
esrj-117639	233	1	https://doi.org/10.1111/j.1365-246x.1991.tb00801.x	https://doi.org/10.1111/j.1365-246x.1991.tb00801.x	PROPN
esrj-117639	233	2	ronneberger	ronneberger	NOUN
esrj-117639	233	3	,	,	PUNCT
esrj-117639	233	4	o.	o.	PROPN
esrj-117639	233	5	,	,	PUNCT
esrj-117639	233	6	fischer	fischer	PROPN
esrj-117639	233	7	,	,	PUNCT
esrj-117639	233	8	p.	p.	PROPN
esrj-117639	233	9	,	,	PUNCT
esrj-117639	233	10	&	&	CCONJ
esrj-117639	233	11	brox	brox	PROPN
esrj-117639	233	12	,	,	PUNCT
esrj-117639	233	13	t.	t.	PROPN
esrj-117639	233	14	(	(	PUNCT
esrj-117639	233	15	2015	2015	NUM
esrj-117639	233	16	)	)	PUNCT
esrj-117639	233	17	.	.	PUNCT
esrj-117639	234	1	u	u	NOUN
esrj-117639	234	2	-	-	NOUN
esrj-117639	234	3	net	net	ADJ
esrj-117639	234	4	:	:	PUNCT
esrj-117639	234	5	convolutional	convolutional	ADJ
esrj-117639	234	6	networks	network	NOUN
esrj-117639	234	7	for	for	ADP
esrj-117639	234	8	biomedical	biomedical	ADJ
esrj-117639	234	9	image	image	NOUN
esrj-117639	234	10	segmentation	segmentation	NOUN
esrj-117639	234	11	.	.	PUNCT
esrj-117639	235	1	18th	18th	ADJ
esrj-117639	235	2	international	international	ADJ
esrj-117639	235	3	conference	conference	NOUN
esrj-117639	235	4	on	on	ADP
esrj-117639	235	5	medical	medical	ADJ
esrj-117639	235	6	image	image	NOUN
esrj-117639	235	7	computing	computing	NOUN
esrj-117639	235	8	and	and	CCONJ
esrj-117639	235	9	computer	computer	NOUN
esrj-117639	235	10	-	-	PUNCT
esrj-117639	235	11	assisted	assist	VERB
esrj-117639	235	12	intervention	intervention	NOUN
esrj-117639	235	13	.	.	PUNCT
esrj-117639	236	1	munich	munich	PROPN
esrj-117639	236	2	,	,	PUNCT
esrj-117639	236	3	germany	germany	PROPN
esrj-117639	236	4	:	:	PUNCT
esrj-117639	236	5	springer	springer	NOUN
esrj-117639	236	6	,	,	PUNCT
esrj-117639	236	7	234	234	NUM
esrj-117639	236	8	-	-	SYM
esrj-117639	236	9	241	241	NUM
esrj-117639	236	10	.	.	PUNCT
esrj-117639	237	1	https://doi.org/10.1007/978-3-31924574-4_28	https://doi.org/10.1007/978-3-31924574-4_28	PRON
esrj-117639	237	2	silva	silva	NOUN
esrj-117639	237	3	,	,	PUNCT
esrj-117639	237	4	j.	j.	PROPN
esrj-117639	237	5	b.	b.	PROPN
esrj-117639	237	6	c.	c.	PROPN
esrj-117639	237	7	,	,	PUNCT
esrj-117639	237	8	santos	santos	PROPN
esrj-117639	237	9	,	,	PUNCT
esrj-117639	237	10	d.	d.	PROPN
esrj-117639	237	11	f.	f.	PROPN
esrj-117639	237	12	,	,	PUNCT
esrj-117639	237	13	&	&	CCONJ
esrj-117639	237	14	gomes	gomes	PROPN
esrj-117639	237	15	,	,	PUNCT
esrj-117639	237	16	k.	k.	PROPN
esrj-117639	237	17	p.	p.	PROPN
esrj-117639	237	18	(	(	PUNCT
esrj-117639	237	19	2014	2014	NUM
esrj-117639	237	20	)	)	PUNCT
esrj-117639	237	21	.	.	PUNCT
esrj-117639	238	1	fast	fast	ADJ
esrj-117639	238	2	gravity	gravity	NOUN
esrj-117639	238	3	inversion	inversion	NOUN
esrj-117639	238	4	of	of	ADP
esrj-117639	238	5	basement	basement	NOUN
esrj-117639	238	6	relief	relief	NOUN
esrj-117639	238	7	.	.	PUNCT
esrj-117639	239	1	geophysics	geophysic	NOUN
esrj-117639	239	2	,	,	PUNCT
esrj-117639	239	3	79(5	79(5	NOUN
esrj-117639	239	4	)	)	PUNCT
esrj-117639	239	5	,	,	PUNCT
esrj-117639	239	6	79	79	NUM
esrj-117639	239	7	-	-	SYM
esrj-117639	239	8	91	91	NUM
esrj-117639	239	9	.	.	PUNCT
esrj-117639	240	1	https://doi.org/10.1190/geo20140024.1	https://doi.org/10.1190/geo20140024.1	PROPN
esrj-117639	240	2	sun	sun	PROPN
esrj-117639	240	3	s.	s.	PROPN
esrj-117639	240	4	,	,	PUNCT
esrj-117639	240	5	yin	yin	PROPN
esrj-117639	240	6	c.	c.	PROPN
esrj-117639	240	7	,	,	PUNCT
esrj-117639	240	8	&	&	CCONJ
esrj-117639	240	9	gao	gao	PROPN
esrj-117639	240	10	x.	x.	PROPN
esrj-117639	240	11	(	(	PUNCT
esrj-117639	240	12	2021	2021	NUM
esrj-117639	240	13	)	)	PUNCT
esrj-117639	240	14	.	.	PUNCT
esrj-117639	241	1	3d	3d	NUM
esrj-117639	241	2	gravity	gravity	NOUN
esrj-117639	241	3	inversion	inversion	NOUN
esrj-117639	241	4	on	on	ADP
esrj-117639	241	5	unstructured	unstructured	ADJ
esrj-117639	241	6	grids	grid	NOUN
esrj-117639	241	7	.	.	PUNCT
esrj-117639	242	1	applied	apply	VERB
esrj-117639	242	2	sciences	science	NOUN
esrj-117639	242	3	,	,	PUNCT
esrj-117639	242	4	11(2	11(2	NUM
esrj-117639	242	5	)	)	PUNCT
esrj-117639	242	6	,	,	PUNCT
esrj-117639	242	7	722	722	NUM
esrj-117639	242	8	.	.	PUNCT
esrj-117639	243	1	https://doi.org/10.3390/app11020722	https://doi.org/10.3390/app11020722	NUM
esrj-117639	243	2	van	van	PROPN
esrj-117639	243	3	der	der	PROPN
esrj-117639	243	4	baan	baan	PROPN
esrj-117639	243	5	,	,	PUNCT
esrj-117639	243	6	m.	m.	NOUN
esrj-117639	243	7	,	,	PUNCT
esrj-117639	243	8	&	&	CCONJ
esrj-117639	243	9	jutten	jutten	PROPN
esrj-117639	243	10	,	,	PUNCT
esrj-117639	243	11	c.	c.	PROPN
esrj-117639	243	12	(	(	PUNCT
esrj-117639	243	13	2000	2000	NUM
esrj-117639	243	14	)	)	PUNCT
esrj-117639	243	15	.	.	PUNCT
esrj-117639	244	1	neural	neural	ADJ
esrj-117639	244	2	networks	network	NOUN
esrj-117639	244	3	in	in	ADP
esrj-117639	244	4	geophysical	geophysical	ADJ
esrj-117639	244	5	applications	application	NOUN
esrj-117639	244	6	.	.	PUNCT
esrj-117639	245	1	geophysics	geophysic	NOUN
esrj-117639	245	2	,	,	PUNCT
esrj-117639	245	3	65(4	65(4	PROPN
esrj-117639	245	4	)	)	PUNCT
esrj-117639	245	5	,	,	PUNCT
esrj-117639	245	6	1032	1032	NUM
esrj-117639	245	7	-	-	SYM
esrj-117639	245	8	1047	1047	NUM
esrj-117639	245	9	.	.	PUNCT
esrj-117639	246	1	https://doi.org/10.1190/1.1444797	https://doi.org/10.1190/1.1444797	PROPN
esrj-117639	246	2	wang	wang	PROPN
esrj-117639	246	3	,	,	PUNCT
esrj-117639	246	4	y.	y.	PROPN
esrj-117639	246	5	f.	f.	PROPN
esrj-117639	246	6	,	,	PUNCT
esrj-117639	246	7	zhang	zhang	PROPN
esrj-117639	246	8	,	,	PUNCT
esrj-117639	246	9	y.	y.	PROPN
esrj-117639	246	10	j.	j.	PROPN
esrj-117639	246	11	,	,	PUNCT
esrj-117639	246	12	fu	fu	PROPN
esrj-117639	246	13	,	,	PUNCT
esrj-117639	246	14	l.	l.	PROPN
esrj-117639	246	15	h.	h.	PROPN
esrj-117639	246	16	,	,	PUNCT
esrj-117639	246	17	&	&	CCONJ
esrj-117639	246	18	li	li	PROPN
esrj-117639	246	19	,	,	PUNCT
esrj-117639	246	20	h.	h.	PROPN
esrj-117639	246	21	w.	w.	PROPN
esrj-117639	246	22	(	(	PUNCT
esrj-117639	246	23	2021	2021	NUM
esrj-117639	246	24	)	)	PUNCT
esrj-117639	246	25	.	.	PUNCT
esrj-117639	247	1	three	three	NUM
esrj-117639	247	2	-	-	PUNCT
esrj-117639	247	3	dimensional	dimensional	ADJ
esrj-117639	247	4	gravity	gravity	NOUN
esrj-117639	247	5	inversion	inversion	NOUN
esrj-117639	247	6	based	base	VERB
esrj-117639	247	7	on	on	ADP
esrj-117639	247	8	3d	3d	NUM
esrj-117639	247	9	u	u	PROPN
esrj-117639	247	10	-	-	PROPN
esrj-117639	247	11	net++	net++	PROPN
esrj-117639	247	12	.	.	PUNCT
esrj-117639	247	13	applied	apply	VERB
esrj-117639	247	14	geophysics	geophysic	NOUN
esrj-117639	247	15	,	,	PUNCT
esrj-117639	247	16	18(4	18(4	NUM
esrj-117639	247	17	)	)	PUNCT
esrj-117639	247	18	,	,	PUNCT
esrj-117639	247	19	451–460	451–460	NUM
esrj-117639	247	20	.	.	PUNCT
esrj-117639	248	1	https://doi.org/10.1007/s11770-021-0909-z	https://doi.org/10.1007/s11770-021-0909-z	PROPN
esrj-117639	248	2	wiener	wiener	NOUN
esrj-117639	248	3	,	,	PUNCT
esrj-117639	248	4	j.	j.	PROPN
esrj-117639	248	5	,	,	PUNCT
esrj-117639	248	6	rogers	rogers	PROPN
esrj-117639	248	7	,	,	PUNCT
esrj-117639	248	8	j.	j.	PROPN
esrj-117639	248	9	a.	a.	PROPN
esrj-117639	248	10	,	,	PUNCT
esrj-117639	248	11	rogers	rogers	PROPN
esrj-117639	248	12	,	,	PUNCT
esrj-117639	248	13	j.	j.	PROPN
esrj-117639	248	14	r.	r.	PROPN
esrj-117639	248	15	,	,	PUNCT
esrj-117639	248	16	&	&	CCONJ
esrj-117639	248	17	moll	moll	PROPN
esrj-117639	248	18	,	,	PUNCT
esrj-117639	248	19	r.	r.	PROPN
esrj-117639	248	20	(	(	PUNCT
esrj-117639	248	21	1991	1991	NUM
esrj-117639	248	22	)	)	PUNCT
esrj-117639	248	23	.	.	PUNCT
esrj-117639	249	1	predicting	predict	VERB
esrj-117639	249	2	carbonate	carbonate	NOUN
esrj-117639	249	3	permeabilities	permeability	NOUN
esrj-117639	249	4	from	from	ADP
esrj-117639	249	5	wireline	wireline	NOUN
esrj-117639	249	6	logs	log	NOUN
esrj-117639	249	7	using	use	VERB
esrj-117639	249	8	a	a	DET
esrj-117639	249	9	back	back	ADJ
esrj-117639	249	10	-	-	PUNCT
esrj-117639	249	11	propagation	propagation	NOUN
esrj-117639	249	12	neural	neural	ADJ
esrj-117639	249	13	network	network	NOUN
esrj-117639	249	14	.	.	PUNCT
esrj-117639	250	1	61st	61st	PROPN
esrj-117639	250	2	seg	seg	PROPN
esrj-117639	250	3	meeting	meeting	PROPN
esrj-117639	250	4	,	,	PUNCT
esrj-117639	250	5	houston	houston	PROPN
esrj-117639	250	6	,	,	PUNCT
esrj-117639	250	7	usa	usa	PROPN
esrj-117639	250	8	,	,	PUNCT
esrj-117639	250	9	expanded	expand	VERB
esrj-117639	250	10	abstracts	abstract	NOUN
esrj-117639	250	11	,	,	PUNCT
esrj-117639	250	12	285	285	NUM
esrj-117639	250	13	-	-	SYM
esrj-117639	250	14	288	288	NUM
esrj-117639	250	15	.	.	PUNCT
esrj-117639	251	1	https://doi.org/10.1190/1.1888943	https://doi.org/10.1190/1.1888943	PROPN
esrj-117639	251	2	winkler	winkler	PROPN
esrj-117639	251	3	,	,	PUNCT
esrj-117639	251	4	e.	e.	PROPN
esrj-117639	251	5	(	(	PUNCT
esrj-117639	251	6	1994	1994	NUM
esrj-117639	251	7	)	)	PUNCT
esrj-117639	251	8	.	.	PUNCT
esrj-117639	252	1	inversion	inversion	NOUN
esrj-117639	252	2	of	of	ADP
esrj-117639	252	3	electromagnetic	electromagnetic	ADJ
esrj-117639	252	4	data	datum	NOUN
esrj-117639	252	5	using	use	VERB
esrj-117639	252	6	neural	neural	ADJ
esrj-117639	252	7	networks	network	NOUN
esrj-117639	252	8	.	.	PUNCT
esrj-117639	253	1	56th	56th	ADJ
esrj-117639	253	2	eaeg	eaeg	NOUN
esrj-117639	253	3	meeting	meeting	NOUN
esrj-117639	253	4	,	,	PUNCT
esrj-117639	253	5	vienna	vienna	PROPN
esrj-117639	253	6	,	,	PUNCT
esrj-117639	253	7	austria	austria	PROPN
esrj-117639	253	8	,	,	PUNCT
esrj-117639	253	9	extended	extended	ADJ
esrj-117639	253	10	abstracts	abstract	NOUN
esrj-117639	253	11	,	,	PUNCT
esrj-117639	253	12	p124	p124	PROPN
esrj-117639	253	13	.	.	PUNCT
esrj-117639	254	1	zhang	zhang	PROPN
esrj-117639	254	2	,	,	PUNCT
esrj-117639	254	3	s.	s.	PROPN
esrj-117639	254	4	,	,	PUNCT
esrj-117639	254	5	yin	yin	PROPN
esrj-117639	254	6	,	,	PUNCT
esrj-117639	254	7	c.	c.	PROPN
esrj-117639	254	8	,	,	PUNCT
esrj-117639	254	9	cao	cao	PROPN
esrj-117639	254	10	,	,	PUNCT
esrj-117639	254	11	x.	x.	NOUN
esrj-117639	254	12	,	,	PUNCT
esrj-117639	254	13	sun	sun	PROPN
esrj-117639	254	14	,	,	PUNCT
esrj-117639	254	15	s.	s.	PROPN
esrj-117639	254	16	,	,	PUNCT
esrj-117639	254	17	liu	liu	PROPN
esrj-117639	254	18	,	,	PUNCT
esrj-117639	254	19	y.	y.	PROPN
esrj-117639	254	20	,	,	PUNCT
esrj-117639	254	21	&	&	CCONJ
esrj-117639	254	22	ren	ren	PROPN
esrj-117639	254	23	,	,	PUNCT
esrj-117639	254	24	x.	x.	NOUN
esrj-117639	254	25	(	(	PUNCT
esrj-117639	254	26	2022	2022	NUM
esrj-117639	254	27	)	)	PUNCT
esrj-117639	254	28	.	.	PUNCT
esrj-117639	255	1	decnet	decnet	NOUN
esrj-117639	255	2	:	:	PUNCT
esrj-117639	255	3	decomposition	decomposition	NOUN
esrj-117639	255	4	network	network	NOUN
esrj-117639	255	5	for	for	ADP
esrj-117639	255	6	3d	3d	NUM
esrj-117639	255	7	gravity	gravity	NOUN
esrj-117639	255	8	inversion	inversion	NOUN
esrj-117639	255	9	.	.	PUNCT
esrj-117639	256	1	geophysics	geophysic	NOUN
esrj-117639	256	2	,	,	PUNCT
esrj-117639	256	3	87(5	87(5	NUM
esrj-117639	256	4	):	):	PUNCT
esrj-117639	256	5	1so	1so	NOUN
esrj-117639	256	6	-	-	PUNCT
esrj-117639	256	7	v558	v558	PROPN
esrj-117639	256	8	.	.	PUNCT
esrj-117639	257	1	https://doi.org/10.1190/geo2021-0744.1	https://doi.org/10.1190/geo2021-0744.1	NOUN
esrj-117639	257	2	https://doi.org/10.2118/0221-0070-jpt	https://doi.org/10.2118/0221-0070-jpt	NOUN
esrj-117639	257	3	https://doi.org/10.2118/0221-0070-jpt	https://doi.org/10.2118/0221-0070-jpt	NOUN
esrj-117639	257	4	https://doi.org/10.1190/1.1439955	https://doi.org/10.1190/1.1439955	NOUN
esrj-117639	257	5	https://doi.org/10.1111/j.1365-246x.1995.tb01851.x	https://doi.org/10.1111/j.1365-246x.1995.tb01851.x	NOUN
esrj-117639	257	6	https://doi.org/10.1111/j.1365-246x.1995.tb01851.x	https://doi.org/10.1111/j.1365-246x.1995.tb01851.x	PROPN
esrj-117639	257	7	https://doi.org/10.1016/j.cageo.2004.11.004	https://doi.org/10.1016/j.cageo.2004.11.004	PROPN
esrj-117639	258	1	https://doi.org/10.1111/j.1365-2478.1987.tb00858.x	https://doi.org/10.1111/j.1365-2478.1987.tb00858.x	NOUN
esrj-117639	259	1	https://doi.org/10.1111/j.1365-2478.1987.tb00858.x	https://doi.org/10.1111/j.1365-2478.1987.tb00858.x	INTJ
esrj-117639	259	2	https://doi.org/10.1190/1.1442243	https://doi.org/10.1190/1.1442243	NOUN
esrj-117639	259	3	https://doi.org/10.1190/geo2011-0154.1	https://doi.org/10.1190/geo2011-0154.1	VERB
esrj-117639	259	4	https://doi.org/10.1190/geo2011-0154.1	https://doi.org/10.1190/geo2011-0154.1	PROPN
esrj-117639	259	5	https://doi.org/10.1016/0098-3004(95)00075-5	https://doi.org/10.1016/0098-3004(95)00075-5	PROPN
esrj-117639	259	6	https://doi.org/10.1190/1.1440444	https://doi.org/10.1190/1.1440444	ADP
esrj-117639	259	7	https://doi.org/10.1111/j.1365-246x.1973.tb06513.x	https://doi.org/10.1111/j.1365-246x.1973.tb06513.x	NOUN
esrj-117639	259	8	https://doi.org/10.1111/j.1365-246x.1973.tb06513.x	https://doi.org/10.1111/j.1365-246x.1973.tb06513.x	VERB
esrj-117639	259	9	https://doi.org/10.1190/1.1888789	https://doi.org/10.1190/1.1888789	NOUN
esrj-117639	260	1	https://doi.org/10.1016/0926-9851(92)90010-i	https://doi.org/10.1016/0926-9851(92)90010-i	INTJ
esrj-117639	260	2	https://doi.org/10.1111/j.1365-246x.1991.tb00801.x	https://doi.org/10.1111/j.1365-246x.1991.tb00801.x	INTJ
esrj-117639	260	3	https://doi.org/10.1111/j.1365-246x.1991.tb00801.x	https://doi.org/10.1111/j.1365-246x.1991.tb00801.x	VERB
esrj-117639	261	1	https://doi.org/10.1007/978-3-319-24574-4_28	https://doi.org/10.1007/978-3-319-24574-4_28	PROPN
esrj-117639	261	2	https://doi.org/10.1007/978-3-319-24574-4_28	https://doi.org/10.1007/978-3-319-24574-4_28	PROPN
esrj-117639	262	1	https://doi.org/10.1190/geo2014-0024.1	https://doi.org/10.1190/geo2014-0024.1	ADV
esrj-117639	263	1	https://doi.org/10.1190/geo2014-0024.1	https://doi.org/10.1190/geo2014-0024.1	ADJ
esrj-117639	263	2	https://doi.org/10.3390/app11020722	https://doi.org/10.3390/app11020722	NUM
esrj-117639	263	3	https://doi.org/10.1190/1.1444797	https://doi.org/10.1190/1.1444797	NOUN
esrj-117639	263	4	https://doi.org/10.1007/s11770-021-0909-z	https://doi.org/10.1007/s11770-021-0909-z	ADJ
esrj-117639	263	5	https://doi.org/10.1190/1.1888943	https://doi.org/10.1190/1.1888943	PROPN
esrj-117639	263	6	https://doi.org/10.1190/geo2021-0744.1	https://doi.org/10.1190/geo2021-0744.1	PROPN
esrj-117639	264	1	_	_	NOUN
esrj-117639	264	2	hlk132811074	hlk132811074	ADJ