id	sid	tid	token	lemma	pos
easat-5050	1	1	edelweiss	edelweiss	PROPN
easat-5050	1	2	applied	apply	VERB
easat-5050	1	3	science	science	NOUN
easat-5050	1	4	and	and	CCONJ
easat-5050	1	5	technology	technology	NOUN
easat-5050	1	6	issn	issn	PROPN
easat-5050	1	7	:	:	PUNCT
easat-5050	1	8	2576	2576	NUM
easat-5050	1	9	-	-	SYM
easat-5050	1	10	8484	8484	NUM
easat-5050	1	11	vol	vol	NOUN
easat-5050	1	12	.	.	PROPN
easat-5050	2	1	9	9	NUM
easat-5050	2	2	,	,	PUNCT
easat-5050	2	3	no	no	INTJ
easat-5050	2	4	.	.	NOUN
easat-5050	2	5	2	2	NUM
easat-5050	2	6	,	,	PUNCT
easat-5050	2	7	2132	2132	NUM
easat-5050	2	8	-	-	SYM
easat-5050	2	9	2146	2146	NUM
easat-5050	2	10	2025	2025	NUM
easat-5050	2	11	publisher	publisher	NOUN
easat-5050	2	12	:	:	PUNCT
easat-5050	2	13	learning	learn	VERB
easat-5050	2	14	gate	gate	NOUN
easat-5050	2	15	doi	doi	PROPN
easat-5050	2	16	:	:	PUNCT
easat-5050	2	17	10.55214/25768484.v9i2.5050	10.55214/25768484.v9i2.5050	NUM
easat-5050	2	18	©	©	PROPN
easat-5050	2	19	2025	2025	NUM
easat-5050	2	20	by	by	ADP
easat-5050	2	21	the	the	DET
easat-5050	2	22	authors	author	NOUN
easat-5050	2	23	;	;	PUNCT
easat-5050	2	24	licensee	licensee	PROPN
easat-5050	2	25	learning	learning	NOUN
easat-5050	2	26	gate	gate	NOUN
easat-5050	3	1	©	©	PROPN
easat-5050	3	2	2025	2025	NUM
easat-5050	3	3	by	by	ADP
easat-5050	3	4	the	the	DET
easat-5050	3	5	authors	author	NOUN
easat-5050	3	6	;	;	PUNCT
easat-5050	3	7	licensee	licensee	PROPN
easat-5050	3	8	learning	learning	NOUN
easat-5050	3	9	gate	gate	NOUN
easat-5050	3	10	history	history	NOUN
easat-5050	3	11	:	:	PUNCT
easat-5050	3	12	received	receive	VERB
easat-5050	3	13	:	:	PUNCT
easat-5050	3	14	20	20	NUM
easat-5050	3	15	december	december	PROPN
easat-5050	3	16	2024	2024	NUM
easat-5050	3	17	;	;	PUNCT
easat-5050	3	18	revised	revise	VERB
easat-5050	3	19	:	:	PUNCT
easat-5050	3	20	23	23	NUM
easat-5050	3	21	january	january	NOUN
easat-5050	3	22	2025	2025	NUM
easat-5050	3	23	;	;	PUNCT
easat-5050	3	24	accepted	accept	VERB
easat-5050	3	25	:	:	PUNCT
easat-5050	3	26	25	25	NUM
easat-5050	3	27	january	january	PROPN
easat-5050	3	28	2025	2025	NUM
easat-5050	3	29	;	;	PUNCT
easat-5050	3	30	published	publish	VERB
easat-5050	3	31	:	:	PUNCT
easat-5050	4	1	26	26	NUM
easat-5050	4	2	february	february	NOUN
easat-5050	4	3	2025	2025	NUM
easat-5050	4	4	*	*	PUNCT
easat-5050	4	5	correspondence	correspondence	NOUN
easat-5050	4	6	:	:	PUNCT
easat-5050	5	1	jtshirley@unimas.my	jtshirley@unimas.my	PROPN
easat-5050	5	2	bearing	bear	VERB
easat-5050	5	3	fault	fault	NOUN
easat-5050	5	4	diagnosis	diagnosis	NOUN
easat-5050	5	5	in	in	ADP
easat-5050	5	6	high	high	ADJ
easat-5050	5	7	noise	noise	NOUN
easat-5050	5	8	environment	environment	NOUN
easat-5050	5	9	using	use	VERB
easat-5050	5	10	multi	multi	ADJ
easat-5050	5	11	-	-	ADJ
easat-5050	5	12	scale	scale	ADJ
easat-5050	5	13	processing	processing	NOUN
easat-5050	5	14	,	,	PUNCT
easat-5050	5	15	channel	channel	NOUN
easat-5050	5	16	-	-	PUNCT
easat-5050	5	17	attention	attention	NOUN
easat-5050	5	18	and	and	CCONJ
easat-5050	5	19	feature	feature	NOUN
easat-5050	5	20	-	-	PUNCT
easat-5050	5	21	enhanced	enhance	VERB
easat-5050	5	22	convolutional	convolutional	ADJ
easat-5050	5	23	neural	neural	ADJ
easat-5050	5	24	network	network	NOUN
easat-5050	5	25	model	model	NOUN
easat-5050	5	26	xiuyu	xiuyu	PROPN
easat-5050	5	27	li1,2	li1,2	PROPN
easat-5050	5	28	,	,	PUNCT
easat-5050	5	29	shirley	shirley	PROPN
easat-5050	5	30	j.	j.	PROPN
easat-5050	5	31	tanjong1	tanjong1	PROPN
easat-5050	5	32	*	*	PROPN
easat-5050	5	33	1faculty	1faculty	NUM
easat-5050	5	34	of	of	ADP
easat-5050	5	35	engineering	engineering	NOUN
easat-5050	5	36	,	,	PUNCT
easat-5050	5	37	universiti	universiti	PROPN
easat-5050	5	38	malaysia	malaysia	PROPN
easat-5050	5	39	sarawak	sarawak	PROPN
easat-5050	5	40	,	,	PUNCT
easat-5050	5	41	94300	94300	NUM
easat-5050	5	42	,	,	PUNCT
easat-5050	5	43	kota	kota	PROPN
easat-5050	5	44	samarahan	samarahan	PROPN
easat-5050	5	45	,	,	PUNCT
easat-5050	5	46	sarawak	sarawak	NOUN
easat-5050	5	47	,	,	PUNCT
easat-5050	5	48	malaysia	malaysia	PROPN
easat-5050	5	49	;	;	PUNCT
easat-5050	5	50	jtshirley@unimas.my	jtshirley@unimas.my	PROPN
easat-5050	5	51	(	(	PUNCT
easat-5050	5	52	s.j.t	s.j.t	NOUN
easat-5050	5	53	.	.	PUNCT
easat-5050	5	54	)	)	PUNCT
easat-5050	5	55	.	.	PUNCT
easat-5050	6	1	2intelligent	2intelligent	NUM
easat-5050	6	2	manufacturing	manufacturing	NOUN
easat-5050	6	3	and	and	CCONJ
easat-5050	6	4	automobile	automobile	NOUN
easat-5050	6	5	academy	academy	PROPN
easat-5050	6	6	,	,	PUNCT
easat-5050	6	7	anhui	anhui	PROPN
easat-5050	6	8	business	business	PROPN
easat-5050	6	9	and	and	CCONJ
easat-5050	6	10	technology	technology	PROPN
easat-5050	6	11	college	college	PROPN
easat-5050	6	12	,	,	PUNCT
easat-5050	6	13	hefei	hefei	PROPN
easat-5050	6	14	,	,	PUNCT
easat-5050	6	15	231131	231131	NUM
easat-5050	6	16	,	,	PUNCT
easat-5050	6	17	china	china	PROPN
easat-5050	6	18	.	.	PUNCT
easat-5050	7	1	abstract	abstract	PROPN
easat-5050	7	2	:	:	PUNCT
easat-5050	7	3	this	this	DET
easat-5050	7	4	paper	paper	NOUN
easat-5050	7	5	presents	present	VERB
easat-5050	7	6	a	a	DET
easat-5050	7	7	model	model	NOUN
easat-5050	7	8	using	use	VERB
easat-5050	7	9	deep	deep	ADJ
easat-5050	7	10	learning	learning	NOUN
easat-5050	7	11	techniques	technique	NOUN
easat-5050	7	12	which	which	PRON
easat-5050	7	13	includes	include	VERB
easat-5050	7	14	multi	multi	ADJ
easat-5050	7	15	-	-	ADJ
easat-5050	7	16	scale	scale	ADJ
easat-5050	7	17	processing	processing	NOUN
easat-5050	7	18	,	,	PUNCT
easat-5050	7	19	channel	channel	NOUN
easat-5050	7	20	attention	attention	NOUN
easat-5050	7	21	,	,	PUNCT
easat-5050	7	22	feature	feature	NOUN
easat-5050	7	23	enhancement	enhancement	NOUN
easat-5050	7	24	,	,	PUNCT
easat-5050	7	25	and	and	CCONJ
easat-5050	7	26	anomaly	anomaly	NOUN
easat-5050	7	27	classification	classification	NOUN
easat-5050	7	28	layers	layer	NOUN
easat-5050	7	29	,	,	PUNCT
easat-5050	7	30	referred	refer	VERB
easat-5050	7	31	to	to	ADP
easat-5050	7	32	as	as	ADP
easat-5050	7	33	mcfcnn	mcfcnn	PROPN
easat-5050	7	34	,	,	PUNCT
easat-5050	7	35	for	for	ADP
easat-5050	7	36	bearing	bear	VERB
easat-5050	7	37	fault	fault	NOUN
easat-5050	7	38	diagnosis	diagnosis	NOUN
easat-5050	7	39	in	in	ADP
easat-5050	7	40	noisy	noisy	ADJ
easat-5050	7	41	industrial	industrial	ADJ
easat-5050	7	42	environments	environment	NOUN
easat-5050	7	43	.	.	PUNCT
easat-5050	8	1	the	the	DET
easat-5050	8	2	mcfcnn	mcfcnn	PROPN
easat-5050	8	3	network	network	NOUN
easat-5050	8	4	combines	combine	VERB
easat-5050	8	5	multi	multi	ADJ
easat-5050	8	6	-	-	ADJ
easat-5050	8	7	channel	channel	ADJ
easat-5050	8	8	parallel	parallel	ADJ
easat-5050	8	9	convolution	convolution	NOUN
easat-5050	8	10	,	,	PUNCT
easat-5050	8	11	effectively	effectively	ADV
easat-5050	8	12	capturing	capture	VERB
easat-5050	8	13	spatial	spatial	ADJ
easat-5050	8	14	information	information	NOUN
easat-5050	8	15	,	,	PUNCT
easat-5050	8	16	and	and	CCONJ
easat-5050	8	17	introduces	introduce	VERB
easat-5050	8	18	channel	channel	NOUN
easat-5050	8	19	attention	attention	NOUN
easat-5050	8	20	mechanisms	mechanism	NOUN
easat-5050	8	21	to	to	PART
easat-5050	8	22	adaptively	adaptively	ADV
easat-5050	8	23	recalibrate	recalibrate	VERB
easat-5050	8	24	channel	channel	NOUN
easat-5050	8	25	-	-	PUNCT
easat-5050	8	26	level	level	NOUN
easat-5050	8	27	feature	feature	NOUN
easat-5050	8	28	responses	response	NOUN
easat-5050	8	29	.	.	PUNCT
easat-5050	9	1	secondary	secondary	ADJ
easat-5050	9	2	neurons	neuron	NOUN
easat-5050	9	3	are	be	AUX
easat-5050	9	4	introduced	introduce	VERB
easat-5050	9	5	to	to	PART
easat-5050	9	6	enhance	enhance	VERB
easat-5050	9	7	the	the	DET
easat-5050	9	8	model	model	NOUN
easat-5050	9	9	’s	’s	PART
easat-5050	9	10	ability	ability	NOUN
easat-5050	9	11	to	to	PART
easat-5050	9	12	capture	capture	VERB
easat-5050	9	13	complex	complex	ADJ
easat-5050	9	14	nonlinear	nonlinear	ADJ
easat-5050	9	15	patterns	pattern	NOUN
easat-5050	9	16	related	relate	VERB
easat-5050	9	17	to	to	ADP
easat-5050	9	18	bearing	bear	VERB
easat-5050	9	19	faults	fault	NOUN
easat-5050	9	20	.	.	PUNCT
easat-5050	10	1	the	the	DET
easat-5050	10	2	model	model	NOUN
easat-5050	10	3	was	be	AUX
easat-5050	10	4	tested	test	VERB
easat-5050	10	5	and	and	CCONJ
easat-5050	10	6	compared	compare	VERB
easat-5050	10	7	to	to	ADP
easat-5050	10	8	other	other	ADJ
easat-5050	10	9	models	model	NOUN
easat-5050	10	10	using	use	VERB
easat-5050	10	11	a	a	DET
easat-5050	10	12	publicly	publicly	ADV
easat-5050	10	13	available	available	ADJ
easat-5050	10	14	data	datum	NOUN
easat-5050	10	15	set	set	VERB
easat-5050	10	16	.	.	PUNCT
easat-5050	11	1	in	in	ADP
easat-5050	11	2	a	a	DET
easat-5050	11	3	simulated	simulate	VERB
easat-5050	11	4	high	high	ADJ
easat-5050	11	5	-	-	PUNCT
easat-5050	11	6	noise	noise	NOUN
easat-5050	11	7	environment	environment	NOUN
easat-5050	11	8	,	,	PUNCT
easat-5050	11	9	the	the	DET
easat-5050	11	10	proposed	propose	VERB
easat-5050	11	11	model	model	NOUN
easat-5050	11	12	outperforms	outperform	VERB
easat-5050	11	13	existing	exist	VERB
easat-5050	11	14	models	model	NOUN
easat-5050	11	15	in	in	ADP
easat-5050	11	16	fault	fault	NOUN
easat-5050	11	17	diagnosis	diagnosis	NOUN
easat-5050	11	18	,	,	PUNCT
easat-5050	11	19	with	with	ADP
easat-5050	11	20	accuracy	accuracy	NOUN
easat-5050	11	21	greater	great	ADJ
easat-5050	11	22	than	than	ADP
easat-5050	11	23	80	80	NUM
easat-5050	11	24	%	%	NOUN
easat-5050	11	25	even	even	ADV
easat-5050	11	26	at	at	ADP
easat-5050	11	27	high	high	ADJ
easat-5050	11	28	signal	signal	NOUN
easat-5050	11	29	-	-	PUNCT
easat-5050	11	30	to	to	ADP
easat-5050	11	31	-	-	PUNCT
easat-5050	11	32	noise	noise	NOUN
easat-5050	11	33	(	(	PUNCT
easat-5050	11	34	snr	snr	NOUN
easat-5050	11	35	)	)	PUNCT
easat-5050	11	36	ratio	ratio	NOUN
easat-5050	11	37	.	.	PUNCT
easat-5050	12	1	at	at	ADP
easat-5050	12	2	snr	snr	PROPN
easat-5050	12	3	=	=	SYM
easat-5050	12	4	-6	-6	PROPN
easat-5050	12	5	,	,	PUNCT
easat-5050	12	6	the	the	DET
easat-5050	12	7	mcfcnn	mcfcnn	NOUN
easat-5050	12	8	records	record	VERB
easat-5050	12	9	higher	high	ADJ
easat-5050	12	10	accuracy	accuracy	NOUN
easat-5050	12	11	(	(	PUNCT
easat-5050	12	12	83	83	NUM
easat-5050	12	13	%	%	NOUN
easat-5050	12	14	)	)	PUNCT
easat-5050	12	15	,	,	PUNCT
easat-5050	12	16	precision	precision	NOUN
easat-5050	12	17	(	(	PUNCT
easat-5050	12	18	89	89	NUM
easat-5050	12	19	%	%	NOUN
easat-5050	12	20	)	)	PUNCT
easat-5050	12	21	,	,	PUNCT
easat-5050	12	22	and	and	CCONJ
easat-5050	12	23	recall	recall	NOUN
easat-5050	12	24	rates	rate	NOUN
easat-5050	12	25	(	(	PUNCT
easat-5050	12	26	84.5	84.5	NUM
easat-5050	12	27	%	%	NOUN
easat-5050	12	28	)	)	PUNCT
easat-5050	12	29	as	as	SCONJ
easat-5050	12	30	compared	compare	VERB
easat-5050	12	31	to	to	ADP
easat-5050	12	32	prior	prior	ADJ
easat-5050	12	33	models	model	NOUN
easat-5050	12	34	.	.	PUNCT
easat-5050	13	1	the	the	DET
easat-5050	13	2	proposed	propose	VERB
easat-5050	13	3	model	model	NOUN
easat-5050	13	4	can	can	AUX
easat-5050	13	5	be	be	AUX
easat-5050	13	6	integrated	integrate	VERB
easat-5050	13	7	into	into	ADP
easat-5050	13	8	the	the	DET
easat-5050	13	9	maintenance	maintenance	NOUN
easat-5050	13	10	management	management	NOUN
easat-5050	13	11	system	system	NOUN
easat-5050	13	12	to	to	PART
easat-5050	13	13	enhance	enhance	VERB
easat-5050	13	14	bearing	bear	VERB
easat-5050	13	15	health	health	NOUN
easat-5050	13	16	assessment	assessment	NOUN
easat-5050	13	17	and	and	CCONJ
easat-5050	13	18	prediction	prediction	NOUN
easat-5050	13	19	,	,	PUNCT
easat-5050	13	20	improving	improve	VERB
easat-5050	13	21	machine	machine	NOUN
easat-5050	13	22	prognostics	prognostic	NOUN
easat-5050	13	23	.	.	PUNCT
easat-5050	14	1	keywords	keyword	NOUN
easat-5050	14	2	:	:	PUNCT
easat-5050	14	3	bearing	bear	VERB
easat-5050	14	4	fault	fault	NOUN
easat-5050	14	5	diagnosis	diagnosis	NOUN
easat-5050	14	6	,	,	PUNCT
easat-5050	14	7	channel	channel	NOUN
easat-5050	14	8	attention	attention	NOUN
easat-5050	14	9	,	,	PUNCT
easat-5050	14	10	multi	multi	ADJ
easat-5050	14	11	-	-	ADJ
easat-5050	14	12	scale	scale	ADJ
easat-5050	14	13	processing	processing	NOUN
easat-5050	14	14	,	,	PUNCT
easat-5050	14	15	quadratic	quadratic	ADJ
easat-5050	14	16	convolutional	convolutional	ADJ
easat-5050	14	17	neural	neural	NOUN
easat-5050	14	18	.	.	PUNCT
easat-5050	15	1	1	1	X
easat-5050	15	2	.	.	X
easat-5050	15	3	introduction	introduction	NOUN
easat-5050	15	4	rotating	rotate	VERB
easat-5050	15	5	machines	machine	NOUN
easat-5050	15	6	are	be	AUX
easat-5050	15	7	core	core	ADJ
easat-5050	15	8	parts	part	NOUN
easat-5050	15	9	of	of	ADP
easat-5050	15	10	equipment	equipment	NOUN
easat-5050	15	11	in	in	ADP
easat-5050	15	12	power	power	NOUN
easat-5050	15	13	plants	plant	NOUN
easat-5050	15	14	and	and	CCONJ
easat-5050	15	15	industrial	industrial	ADJ
easat-5050	15	16	production	production	NOUN
easat-5050	15	17	.	.	PUNCT
easat-5050	16	1	bearings	bearing	NOUN
easat-5050	16	2	are	be	AUX
easat-5050	16	3	key	key	ADJ
easat-5050	16	4	components	component	NOUN
easat-5050	16	5	that	that	PRON
easat-5050	16	6	support	support	VERB
easat-5050	16	7	and	and	CCONJ
easat-5050	16	8	maintain	maintain	VERB
easat-5050	16	9	rotating	rotate	VERB
easat-5050	16	10	parts	part	NOUN
easat-5050	16	11	and	and	CCONJ
easat-5050	16	12	are	be	AUX
easat-5050	16	13	vital	vital	ADJ
easat-5050	16	14	in	in	ADP
easat-5050	16	15	ensuring	ensure	VERB
easat-5050	16	16	the	the	DET
easat-5050	16	17	smooth	smooth	ADJ
easat-5050	16	18	operation	operation	NOUN
easat-5050	16	19	and	and	CCONJ
easat-5050	16	20	long	long	ADJ
easat-5050	16	21	-	-	PUNCT
easat-5050	16	22	term	term	NOUN
easat-5050	16	23	reliability	reliability	NOUN
easat-5050	16	24	of	of	ADP
easat-5050	16	25	machinery	machinery	NOUN
easat-5050	16	26	.	.	PUNCT
easat-5050	17	1	however	however	ADV
easat-5050	17	2	,	,	PUNCT
easat-5050	17	3	owing	owe	VERB
easat-5050	17	4	to	to	ADP
easat-5050	17	5	the	the	DET
easat-5050	17	6	complex	complex	ADJ
easat-5050	17	7	working	working	NOUN
easat-5050	17	8	environment	environment	NOUN
easat-5050	17	9	,	,	PUNCT
easat-5050	17	10	high	high	ADJ
easat-5050	17	11	loads	load	NOUN
easat-5050	17	12	,	,	PUNCT
easat-5050	17	13	and	and	CCONJ
easat-5050	17	14	continuous	continuous	ADJ
easat-5050	17	15	operation	operation	NOUN
easat-5050	17	16	of	of	ADP
easat-5050	17	17	the	the	DET
easat-5050	17	18	rotating	rotate	VERB
easat-5050	17	19	machinery	machinery	NOUN
easat-5050	17	20	,	,	PUNCT
easat-5050	17	21	the	the	DET
easat-5050	17	22	failure	failure	NOUN
easat-5050	17	23	rate	rate	NOUN
easat-5050	17	24	is	be	AUX
easat-5050	17	25	high	high	ADJ
easat-5050	17	26	.	.	PUNCT
easat-5050	18	1	faults	fault	NOUN
easat-5050	18	2	in	in	ADP
easat-5050	18	3	components	component	NOUN
easat-5050	18	4	of	of	ADP
easat-5050	18	5	mechanical	mechanical	ADJ
easat-5050	18	6	equipment	equipment	NOUN
easat-5050	18	7	may	may	AUX
easat-5050	18	8	cause	cause	VERB
easat-5050	18	9	the	the	DET
easat-5050	18	10	entire	entire	ADJ
easat-5050	18	11	system	system	NOUN
easat-5050	18	12	inoperable	inoperable	ADJ
easat-5050	18	13	and	and	CCONJ
easat-5050	18	14	failures	failure	NOUN
easat-5050	18	15	of	of	ADP
easat-5050	18	16	critical	critical	ADJ
easat-5050	18	17	components	component	NOUN
easat-5050	18	18	can	can	AUX
easat-5050	18	19	lead	lead	VERB
easat-5050	18	20	to	to	ADP
easat-5050	18	21	significant	significant	ADJ
easat-5050	18	22	casualties	casualty	NOUN
easat-5050	18	23	and	and	CCONJ
easat-5050	18	24	economic	economic	ADJ
easat-5050	18	25	losses	loss	NOUN
easat-5050	18	26	(	(	PUNCT
easat-5050	18	27	rai	rai	PROPN
easat-5050	18	28	&	&	CCONJ
easat-5050	18	29	upadhyay	upadhyay	PROPN
easat-5050	18	30	,	,	PUNCT
easat-5050	18	31	2016	2016	NUM
easat-5050	18	32	)	)	PUNCT
easat-5050	18	33	.	.	PUNCT
easat-5050	19	1	bearing	bear	VERB
easat-5050	19	2	failures	failure	NOUN
easat-5050	19	3	account	account	VERB
easat-5050	19	4	for	for	ADP
easat-5050	19	5	a	a	DET
easat-5050	19	6	significant	significant	ADJ
easat-5050	19	7	proportion	proportion	NOUN
easat-5050	19	8	of	of	ADP
easat-5050	19	9	mechanical	mechanical	ADJ
easat-5050	19	10	equipment	equipment	NOUN
easat-5050	19	11	failures	failure	NOUN
easat-5050	19	12	(	(	PUNCT
easat-5050	19	13	qiu	qiu	PROPN
easat-5050	19	14	et	et	PROPN
easat-5050	19	15	al	al	PROPN
easat-5050	19	16	.	.	PROPN
easat-5050	19	17	,	,	PUNCT
easat-5050	19	18	2023	2023	NUM
easat-5050	19	19	)	)	PUNCT
easat-5050	19	20	.	.	PUNCT
easat-5050	20	1	bearing	bear	VERB
easat-5050	20	2	fault	fault	NOUN
easat-5050	20	3	signals	signal	NOUN
easat-5050	20	4	can	can	AUX
easat-5050	20	5	be	be	AUX
easat-5050	20	6	collected	collect	VERB
easat-5050	20	7	by	by	ADP
easat-5050	20	8	installing	instal	VERB
easat-5050	20	9	various	various	ADJ
easat-5050	20	10	types	type	NOUN
easat-5050	20	11	of	of	ADP
easat-5050	20	12	sensors	sensor	NOUN
easat-5050	20	13	to	to	PART
easat-5050	20	14	monitor	monitor	VERB
easat-5050	20	15	the	the	DET
easat-5050	20	16	working	work	VERB
easat-5050	20	17	conditions	condition	NOUN
easat-5050	20	18	of	of	ADP
easat-5050	20	19	bearings	bearing	NOUN
easat-5050	20	20	,	,	PUNCT
easat-5050	20	21	including	include	VERB
easat-5050	20	22	vibration	vibration	NOUN
easat-5050	20	23	(	(	PUNCT
easat-5050	20	24	h.-y	h.-y	NOUN
easat-5050	20	25	.	.	PUNCT
easat-5050	21	1	chen	chen	PROPN
easat-5050	21	2	&	&	CCONJ
easat-5050	21	3	lee	lee	PROPN
easat-5050	21	4	,	,	PUNCT
easat-5050	21	5	2020	2020	NUM
easat-5050	21	6	;	;	PUNCT
easat-5050	21	7	z.	z.	PROPN
easat-5050	21	8	chen	chen	PROPN
easat-5050	21	9	,	,	PUNCT
easat-5050	21	10	wang	wang	PROPN
easat-5050	21	11	,	,	PUNCT
easat-5050	21	12	zhou	zhou	PROPN
easat-5050	21	13	,	,	PUNCT
easat-5050	21	14	yang	yang	PROPN
easat-5050	21	15	,	,	PUNCT
easat-5050	21	16	&	&	CCONJ
easat-5050	21	17	liu	liu	PROPN
easat-5050	21	18	,	,	PUNCT
easat-5050	21	19	2024	2024	NUM
easat-5050	21	20	)	)	PUNCT
easat-5050	21	21	temperature	temperature	NOUN
easat-5050	21	22	(	(	PUNCT
easat-5050	21	23	eang	eang	PROPN
easat-5050	21	24	&	&	CCONJ
easat-5050	21	25	lee	lee	PROPN
easat-5050	21	26	,	,	PUNCT
easat-5050	21	27	2024	2024	NUM
easat-5050	21	28	;	;	PUNCT
easat-5050	21	29	gunerkar	gunerkar	NOUN
easat-5050	21	30	,	,	PUNCT
easat-5050	21	31	jalan	jalan	PROPN
easat-5050	21	32	,	,	PUNCT
easat-5050	21	33	&	&	CCONJ
easat-5050	21	34	belgamwar	belgamwar	PROPN
easat-5050	21	35	,	,	PUNCT
easat-5050	21	36	2019	2019	NUM
easat-5050	21	37	)	)	PUNCT
easat-5050	21	38	rotational	rotational	ADJ
easat-5050	21	39	speed	speed	NOUN
easat-5050	21	40	(	(	PUNCT
easat-5050	21	41	eang	eang	PROPN
easat-5050	21	42	&	&	CCONJ
easat-5050	21	43	lee	lee	PROPN
easat-5050	21	44	,	,	PUNCT
easat-5050	21	45	2024	2024	NUM
easat-5050	21	46	)	)	PUNCT
easat-5050	21	47	and	and	CCONJ
easat-5050	21	48	electrical	electrical	ADJ
easat-5050	21	49	current	current	ADJ
easat-5050	21	50	(	(	PUNCT
easat-5050	21	51	bouras	bouras	NUM
easat-5050	21	52	,	,	PUNCT
easat-5050	21	53	bennedjai	bennedjai	PROPN
easat-5050	21	54	,	,	PUNCT
easat-5050	21	55	&	&	CCONJ
easat-5050	21	56	bouras	boura	NOUN
easat-5050	21	57	,	,	PUNCT
easat-5050	21	58	2020	2020	NUM
easat-5050	21	59	)	)	PUNCT
easat-5050	21	60	among	among	ADP
easat-5050	21	61	others	other	NOUN
easat-5050	21	62	.	.	PUNCT
easat-5050	22	1	vibration	vibration	NOUN
easat-5050	22	2	-	-	PUNCT
easat-5050	22	3	based	base	VERB
easat-5050	22	4	signal	signal	NOUN
easat-5050	22	5	acquisition	acquisition	NOUN
easat-5050	22	6	has	have	AUX
easat-5050	22	7	been	be	AUX
easat-5050	22	8	extensively	extensively	ADV
easat-5050	22	9	researched	research	VERB
easat-5050	22	10	and	and	CCONJ
easat-5050	22	11	applied	apply	VERB
easat-5050	22	12	owing	owe	VERB
easat-5050	22	13	to	to	ADP
easat-5050	22	14	its	its	PRON
easat-5050	22	15	independence	independence	NOUN
easat-5050	22	16	from	from	ADP
easat-5050	22	17	the	the	DET
easat-5050	22	18	mechanical	mechanical	ADJ
easat-5050	22	19	structure	structure	NOUN
easat-5050	22	20	and	and	CCONJ
easat-5050	22	21	simplicity	simplicity	NOUN
easat-5050	22	22	of	of	ADP
easat-5050	22	23	testing	testing	NOUN
easat-5050	22	24	.	.	PUNCT
easat-5050	23	1	however	however	ADV
easat-5050	23	2	,	,	PUNCT
easat-5050	23	3	the	the	DET
easat-5050	23	4	complexity	complexity	NOUN
easat-5050	23	5	of	of	ADP
easat-5050	23	6	the	the	DET
easat-5050	23	7	environment	environment	NOUN
easat-5050	23	8	and	and	CCONJ
easat-5050	23	9	variability	variability	NOUN
easat-5050	23	10	of	of	ADP
easat-5050	23	11	the	the	DET
easat-5050	23	12	load	load	NOUN
easat-5050	23	13	result	result	NOUN
easat-5050	23	14	in	in	ADP
easat-5050	23	15	the	the	DET
easat-5050	23	16	interference	interference	NOUN
easat-5050	23	17	of	of	ADP
easat-5050	23	18	noise	noise	NOUN
easat-5050	23	19	in	in	ADP
easat-5050	23	20	the	the	DET
easat-5050	23	21	collected	collect	VERB
easat-5050	23	22	vibration	vibration	NOUN
easat-5050	23	23	signals	signal	NOUN
easat-5050	23	24	during	during	ADP
easat-5050	23	25	the	the	DET
easat-5050	23	26	fault	fault	NOUN
easat-5050	23	27	diagnosis	diagnosis	NOUN
easat-5050	23	28	process	process	NOUN
easat-5050	23	29	,	,	PUNCT
easat-5050	23	30	increasing	increase	VERB
easat-5050	23	31	the	the	DET
easat-5050	23	32	difficulty	difficulty	NOUN
easat-5050	23	33	of	of	ADP
easat-5050	23	34	fault	fault	NOUN
easat-5050	23	35	diagnosis	diagnosis	NOUN
easat-5050	23	36	.	.	PUNCT
easat-5050	24	1	figure	figure	NOUN
easat-5050	24	2	1	1	NUM
easat-5050	24	3	illustrates	illustrate	VERB
easat-5050	24	4	a	a	DET
easat-5050	24	5	vibration	vibration	NOUN
easat-5050	24	6	signal	signal	NOUN
easat-5050	24	7	without	without	ADP
easat-5050	24	8	noise	noise	NOUN
easat-5050	24	9	and	and	CCONJ
easat-5050	24	10	with	with	ADP
easat-5050	24	11	noise	noise	NOUN
easat-5050	24	12	.	.	PUNCT
easat-5050	25	1	2133	2133	NUM
easat-5050	25	2	edelweiss	edelweiss	PROPN
easat-5050	25	3	applied	apply	VERB
easat-5050	25	4	science	science	NOUN
easat-5050	25	5	and	and	CCONJ
easat-5050	25	6	technology	technology	NOUN
easat-5050	25	7	issn	issn	PROPN
easat-5050	25	8	:	:	PUNCT
easat-5050	25	9	2576	2576	NUM
easat-5050	25	10	-	-	SYM
easat-5050	25	11	8484	8484	NUM
easat-5050	25	12	vol	vol	NOUN
easat-5050	25	13	.	.	PROPN
easat-5050	26	1	9	9	NUM
easat-5050	26	2	,	,	PUNCT
easat-5050	26	3	no	no	INTJ
easat-5050	26	4	.	.	NOUN
easat-5050	26	5	2	2	NUM
easat-5050	26	6	:	:	SYM
easat-5050	26	7	2132	2132	NUM
easat-5050	26	8	-	-	SYM
easat-5050	26	9	2146	2146	NUM
easat-5050	26	10	,	,	PUNCT
easat-5050	26	11	2025	2025	NUM
easat-5050	26	12	doi	doi	NOUN
easat-5050	26	13	:	:	PUNCT
easat-5050	26	14	10.55214/25768484.v9i2.5050	10.55214/25768484.v9i2.5050	NUM
easat-5050	26	15	©	©	PROPN
easat-5050	26	16	2025	2025	NUM
easat-5050	26	17	by	by	ADP
easat-5050	26	18	the	the	DET
easat-5050	26	19	authors	author	NOUN
easat-5050	26	20	;	;	PUNCT
easat-5050	26	21	licensee	licensee	PROPN
easat-5050	26	22	learning	learn	VERB
easat-5050	26	23	gate	gate	NOUN
easat-5050	26	24	figure	figure	NOUN
easat-5050	26	25	1	1	NUM
easat-5050	26	26	.	.	PUNCT
easat-5050	27	1	sample	sample	NOUN
easat-5050	27	2	of	of	ADP
easat-5050	27	3	vibration	vibration	NOUN
easat-5050	27	4	signals	signal	NOUN
easat-5050	27	5	:	:	PUNCT
easat-5050	27	6	(	(	PUNCT
easat-5050	27	7	a	a	X
easat-5050	27	8	)	)	PUNCT
easat-5050	27	9	top	top	ADJ
easat-5050	27	10	graph	graph	NOUN
easat-5050	27	11	–	–	PUNCT
easat-5050	27	12	vibration	vibration	NOUN
easat-5050	27	13	data	datum	NOUN
easat-5050	27	14	without	without	ADP
easat-5050	27	15	noise	noise	NOUN
easat-5050	27	16	;	;	PUNCT
easat-5050	27	17	(	(	PUNCT
easat-5050	27	18	b	b	X
easat-5050	27	19	)	)	PUNCT
easat-5050	27	20	bottom	bottom	ADJ
easat-5050	27	21	graph	graph	NOUN
easat-5050	27	22	–	–	PUNCT
easat-5050	27	23	vibration	vibration	NOUN
easat-5050	27	24	data	datum	NOUN
easat-5050	27	25	with	with	ADP
easat-5050	27	26	noise	noise	NOUN
easat-5050	27	27	.	.	PUNCT
easat-5050	28	1	for	for	ADP
easat-5050	28	2	bearing	bear	VERB
easat-5050	28	3	fault	fault	NOUN
easat-5050	28	4	diagnosis	diagnosis	NOUN
easat-5050	28	5	in	in	ADP
easat-5050	28	6	a	a	DET
easat-5050	28	7	noisy	noisy	ADJ
easat-5050	28	8	environment	environment	NOUN
easat-5050	28	9	,	,	PUNCT
easat-5050	28	10	traditional	traditional	ADJ
easat-5050	28	11	methods	method	NOUN
easat-5050	28	12	struggle	struggle	VERB
easat-5050	28	13	to	to	PART
easat-5050	28	14	extract	extract	VERB
easat-5050	28	15	fault	fault	VERB
easat-5050	28	16	characteristic	characteristic	ADJ
easat-5050	28	17	frequencies	frequency	NOUN
easat-5050	28	18	directly	directly	ADV
easat-5050	28	19	(	(	PUNCT
easat-5050	28	20	cerrada	cerrada	PROPN
easat-5050	28	21	et	et	PROPN
easat-5050	28	22	al	al	PROPN
easat-5050	28	23	.	.	PROPN
easat-5050	28	24	,	,	PUNCT
easat-5050	28	25	2018	2018	NUM
easat-5050	28	26	)	)	PUNCT
easat-5050	28	27	.	.	PUNCT
easat-5050	29	1	frequently	frequently	ADV
easat-5050	29	2	,	,	PUNCT
easat-5050	29	3	fault	fault	VERB
easat-5050	29	4	feature	feature	NOUN
easat-5050	29	5	extraction	extraction	NOUN
easat-5050	29	6	is	be	AUX
easat-5050	29	7	achieved	achieve	VERB
easat-5050	29	8	through	through	ADP
easat-5050	29	9	time	time	NOUN
easat-5050	29	10	-	-	PUNCT
easat-5050	29	11	domain	domain	NOUN
easat-5050	29	12	(	(	PUNCT
easat-5050	29	13	bousseksou	bousseksou	ADJ
easat-5050	29	14	,	,	PUNCT
easat-5050	29	15	bessous	bessous	ADJ
easat-5050	29	16	,	,	PUNCT
easat-5050	29	17	&	&	CCONJ
easat-5050	29	18	mahmoud	mahmoud	PROPN
easat-5050	29	19	,	,	PUNCT
easat-5050	29	20	2024	2024	NUM
easat-5050	29	21	;	;	PUNCT
easat-5050	29	22	zhao	zhao	X
easat-5050	29	23	,	,	PUNCT
easat-5050	29	24	zhu	zhu	PROPN
easat-5050	29	25	,	,	PUNCT
easat-5050	29	26	&	&	CCONJ
easat-5050	29	27	he	he	PRON
easat-5050	29	28	,	,	PUNCT
easat-5050	29	29	2024	2024	NUM
easat-5050	29	30	)	)	PUNCT
easat-5050	29	31	frequencydomain	frequencydomain	ADJ
easat-5050	29	32	,	,	PUNCT
easat-5050	29	33	and	and	CCONJ
easat-5050	29	34	time	time	NOUN
easat-5050	29	35	–	–	PUNCT
easat-5050	29	36	frequency	frequency	NOUN
easat-5050	29	37	signal	signal	NOUN
easat-5050	29	38	processing	processing	NOUN
easat-5050	29	39	methods	method	NOUN
easat-5050	29	40	,	,	PUNCT
easat-5050	29	41	followed	follow	VERB
easat-5050	29	42	by	by	ADP
easat-5050	29	43	inputting	inputte	VERB
easat-5050	29	44	the	the	DET
easat-5050	29	45	extracted	extract	VERB
easat-5050	29	46	features	feature	NOUN
easat-5050	29	47	into	into	ADP
easat-5050	29	48	a	a	DET
easat-5050	29	49	diagnostic	diagnostic	ADJ
easat-5050	29	50	model	model	NOUN
easat-5050	29	51	for	for	ADP
easat-5050	29	52	fault	fault	NOUN
easat-5050	29	53	classification	classification	NOUN
easat-5050	29	54	.	.	PUNCT
easat-5050	30	1	common	common	ADJ
easat-5050	30	2	feature	feature	NOUN
easat-5050	30	3	-	-	PUNCT
easat-5050	30	4	extraction	extraction	NOUN
easat-5050	30	5	methods	method	NOUN
easat-5050	30	6	include	include	VERB
easat-5050	30	7	the	the	DET
easat-5050	30	8	shorttime	shorttime	NOUN
easat-5050	30	9	fourier	fourier	NOUN
easat-5050	30	10	transform	transform	NOUN
easat-5050	30	11	(	(	PUNCT
easat-5050	30	12	t.	t.	PROPN
easat-5050	30	13	wang	wang	PROPN
easat-5050	30	14	,	,	PUNCT
easat-5050	30	15	liang	liang	PROPN
easat-5050	30	16	,	,	PUNCT
easat-5050	30	17	li	li	PROPN
easat-5050	30	18	,	,	PUNCT
easat-5050	30	19	&	&	CCONJ
easat-5050	30	20	cheng	cheng	PROPN
easat-5050	30	21	,	,	PUNCT
easat-5050	30	22	2014	2014	NUM
easat-5050	30	23	)	)	PUNCT
easat-5050	30	24	fast	fast	ADJ
easat-5050	30	25	fourier	fourier	NOUN
easat-5050	30	26	transform	transform	NOUN
easat-5050	30	27	(	(	PUNCT
easat-5050	30	28	g.	g.	PROPN
easat-5050	30	29	chen	chen	PROPN
easat-5050	30	30	,	,	PUNCT
easat-5050	30	31	liu	liu	PROPN
easat-5050	30	32	,	,	PUNCT
easat-5050	30	33	&	&	CCONJ
easat-5050	30	34	chen	chen	PROPN
easat-5050	30	35	,	,	PUNCT
easat-5050	30	36	2020	2020	NUM
easat-5050	30	37	)	)	PUNCT
easat-5050	30	38	hilbert	hilbert	NOUN
easat-5050	30	39	transform	transform	NOUN
easat-5050	30	40	(	(	PUNCT
easat-5050	30	41	dai	dai	PROPN
easat-5050	30	42	et	et	PROPN
easat-5050	30	43	al	al	PROPN
easat-5050	30	44	.	.	PROPN
easat-5050	30	45	,	,	PUNCT
easat-5050	30	46	2020	2020	NUM
easat-5050	30	47	)	)	PUNCT
easat-5050	30	48	and	and	CCONJ
easat-5050	30	49	wavelet	wavelet	NOUN
easat-5050	30	50	transform	transform	NOUN
easat-5050	30	51	(	(	PUNCT
easat-5050	30	52	duong	duong	PROPN
easat-5050	30	53	&	&	CCONJ
easat-5050	30	54	kim	kim	PROPN
easat-5050	30	55	,	,	PUNCT
easat-5050	30	56	2018	2018	NUM
easat-5050	30	57	)	)	PUNCT
easat-5050	30	58	.	.	PUNCT
easat-5050	31	1	common	common	ADJ
easat-5050	31	2	machine	machine	NOUN
easat-5050	31	3	-	-	PUNCT
easat-5050	31	4	learning	learn	VERB
easat-5050	31	5	diagnostic	diagnostic	ADJ
easat-5050	31	6	models	model	NOUN
easat-5050	31	7	include	include	VERB
easat-5050	31	8	logistic	logistic	ADJ
easat-5050	31	9	regression	regression	NOUN
easat-5050	31	10	(	(	PUNCT
easat-5050	31	11	shang	shang	PROPN
easat-5050	31	12	et	et	PROPN
easat-5050	31	13	al	al	PROPN
easat-5050	31	14	.	.	PROPN
easat-5050	31	15	,	,	PUNCT
easat-5050	31	16	2017	2017	NUM
easat-5050	31	17	)	)	PUNCT
easat-5050	31	18	k	k	X
easat-5050	31	19	-	-	PUNCT
easat-5050	31	20	nearest	near	ADJ
easat-5050	31	21	neighbours	neighbour	NOUN
easat-5050	31	22	(	(	PUNCT
easat-5050	31	23	lu	lu	PROPN
easat-5050	31	24	et	et	PROPN
easat-5050	31	25	al	al	PROPN
easat-5050	31	26	.	.	PROPN
easat-5050	31	27	,	,	PUNCT
easat-5050	31	28	2021	2021	NUM
easat-5050	31	29	)	)	PUNCT
easat-5050	31	30	and	and	CCONJ
easat-5050	31	31	support	support	VERB
easat-5050	31	32	vector	vector	NOUN
easat-5050	31	33	machines	machine	NOUN
easat-5050	31	34	(	(	PUNCT
easat-5050	31	35	m.	m.	NOUN
easat-5050	31	36	wang	wang	PROPN
easat-5050	31	37	et	et	PROPN
easat-5050	31	38	al	al	PROPN
easat-5050	31	39	.	.	PROPN
easat-5050	31	40	,	,	PUNCT
easat-5050	31	41	2021	2021	NUM
easat-5050	31	42	)	)	PUNCT
easat-5050	31	43	.	.	PUNCT
easat-5050	32	1	however	however	ADV
easat-5050	32	2	,	,	PUNCT
easat-5050	32	3	traditional	traditional	ADJ
easat-5050	32	4	methods	method	NOUN
easat-5050	32	5	have	have	VERB
easat-5050	32	6	some	some	DET
easat-5050	32	7	limitations	limitation	NOUN
easat-5050	32	8	:	:	PUNCT
easat-5050	32	9	1	1	X
easat-5050	32	10	)	)	PUNCT
easat-5050	32	11	they	they	PRON
easat-5050	32	12	rely	rely	VERB
easat-5050	32	13	on	on	ADP
easat-5050	32	14	manual	manual	ADJ
easat-5050	32	15	feature	feature	NOUN
easat-5050	32	16	extraction	extraction	NOUN
easat-5050	32	17	,	,	PUNCT
easat-5050	32	18	which	which	PRON
easat-5050	32	19	results	result	VERB
easat-5050	32	20	in	in	ADP
easat-5050	32	21	complex	complex	ADJ
easat-5050	32	22	signal	signal	ADJ
easat-5050	32	23	processing	processing	NOUN
easat-5050	32	24	designs	design	NOUN
easat-5050	32	25	and	and	CCONJ
easat-5050	32	26	a	a	DET
easat-5050	32	27	heavy	heavy	ADJ
easat-5050	32	28	reliance	reliance	NOUN
easat-5050	32	29	on	on	ADP
easat-5050	32	30	expert	expert	NOUN
easat-5050	32	31	experience	experience	NOUN
easat-5050	32	32	.	.	PUNCT
easat-5050	33	1	when	when	SCONJ
easat-5050	33	2	the	the	DET
easat-5050	33	3	operating	operating	NOUN
easat-5050	33	4	conditions	condition	NOUN
easat-5050	33	5	are	be	AUX
easat-5050	33	6	complex	complex	ADJ
easat-5050	33	7	,	,	PUNCT
easat-5050	33	8	these	these	DET
easat-5050	33	9	methods	method	NOUN
easat-5050	33	10	fail	fail	VERB
easat-5050	33	11	to	to	PART
easat-5050	33	12	address	address	VERB
easat-5050	33	13	the	the	DET
easat-5050	33	14	problem	problem	NOUN
easat-5050	33	15	of	of	ADP
easat-5050	33	16	differences	difference	NOUN
easat-5050	33	17	in	in	ADP
easat-5050	33	18	data	datum	NOUN
easat-5050	33	19	distribution	distribution	NOUN
easat-5050	33	20	,	,	PUNCT
easat-5050	33	21	posing	pose	VERB
easat-5050	33	22	significant	significant	ADJ
easat-5050	33	23	challenges	challenge	NOUN
easat-5050	33	24	for	for	ADP
easat-5050	33	25	fault	fault	NOUN
easat-5050	33	26	diagnosis	diagnosis	NOUN
easat-5050	33	27	.	.	PUNCT
easat-5050	34	1	2	2	X
easat-5050	34	2	)	)	PUNCT
easat-5050	34	3	the	the	DET
easat-5050	34	4	diagnostic	diagnostic	ADJ
easat-5050	34	5	model	model	NOUN
easat-5050	34	6	requires	require	VERB
easat-5050	34	7	laborious	laborious	ADJ
easat-5050	34	8	design	design	NOUN
easat-5050	34	9	steps	step	NOUN
easat-5050	34	10	and	and	CCONJ
easat-5050	34	11	different	different	ADJ
easat-5050	34	12	model	model	NOUN
easat-5050	34	13	structure	structure	NOUN
easat-5050	34	14	parameters	parameter	NOUN
easat-5050	34	15	are	be	AUX
easat-5050	34	16	required	require	VERB
easat-5050	34	17	for	for	ADP
easat-5050	34	18	different	different	ADJ
easat-5050	34	19	fault	fault	NOUN
easat-5050	34	20	types	type	NOUN
easat-5050	34	21	.	.	PUNCT
easat-5050	35	1	when	when	SCONJ
easat-5050	35	2	signals	signal	NOUN
easat-5050	35	3	are	be	AUX
easat-5050	35	4	contaminated	contaminate	VERB
easat-5050	35	5	by	by	ADP
easat-5050	35	6	noise	noise	NOUN
easat-5050	35	7	,	,	PUNCT
easat-5050	35	8	the	the	DET
easat-5050	35	9	diagnostic	diagnostic	ADJ
easat-5050	35	10	performance	performance	NOUN
easat-5050	35	11	is	be	AUX
easat-5050	35	12	suboptimal	suboptimal	ADJ
easat-5050	35	13	.	.	PUNCT
easat-5050	36	1	therefore	therefore	ADV
easat-5050	36	2	,	,	PUNCT
easat-5050	36	3	current	current	ADJ
easat-5050	36	4	research	research	NOUN
easat-5050	36	5	on	on	ADP
easat-5050	36	6	bearing	bear	VERB
easat-5050	36	7	fault	fault	NOUN
easat-5050	36	8	diagnosis	diagnosis	NOUN
easat-5050	36	9	in	in	ADP
easat-5050	36	10	noisy	noisy	ADJ
easat-5050	36	11	environments	environment	NOUN
easat-5050	36	12	focuses	focus	VERB
easat-5050	36	13	on	on	ADP
easat-5050	36	14	deep	deep	ADJ
easat-5050	36	15	learning	learning	NOUN
easat-5050	36	16	methods	method	NOUN
easat-5050	36	17	.	.	PUNCT
easat-5050	37	1	deep	deep	ADJ
easat-5050	37	2	learning	learning	NOUN
easat-5050	37	3	has	have	VERB
easat-5050	37	4	deep	deep	ADJ
easat-5050	37	5	structures	structure	NOUN
easat-5050	37	6	and	and	CCONJ
easat-5050	37	7	effective	effective	ADJ
easat-5050	37	8	nonlinear	nonlinear	ADJ
easat-5050	37	9	feature	feature	NOUN
easat-5050	37	10	-	-	PUNCT
easat-5050	37	11	extraction	extraction	NOUN
easat-5050	37	12	capabilities	capability	NOUN
easat-5050	37	13	,	,	PUNCT
easat-5050	37	14	enabling	enable	VERB
easat-5050	37	15	it	it	PRON
easat-5050	37	16	to	to	PART
easat-5050	37	17	extract	extract	VERB
easat-5050	37	18	fault	fault	NOUN
easat-5050	37	19	features	feature	VERB
easat-5050	37	20	directly	directly	ADV
easat-5050	37	21	from	from	ADP
easat-5050	37	22	raw	raw	ADJ
easat-5050	37	23	data	datum	NOUN
easat-5050	37	24	and	and	CCONJ
easat-5050	37	25	achieve	achieve	VERB
easat-5050	37	26	end	end	NOUN
easat-5050	37	27	-	-	PUNCT
easat-5050	37	28	to	to	ADP
easat-5050	37	29	-	-	PUNCT
easat-5050	37	30	end	end	NOUN
easat-5050	37	31	fault	fault	NOUN
easat-5050	37	32	diagnosis	diagnosis	NOUN
easat-5050	37	33	,	,	PUNCT
easat-5050	37	34	avoiding	avoid	VERB
easat-5050	37	35	the	the	DET
easat-5050	37	36	tedious	tedious	ADJ
easat-5050	37	37	feature	feature	NOUN
easat-5050	37	38	-	-	PUNCT
easat-5050	37	39	extraction	extraction	NOUN
easat-5050	37	40	and	and	CCONJ
easat-5050	37	41	classification	classification	NOUN
easat-5050	37	42	steps	step	NOUN
easat-5050	37	43	of	of	ADP
easat-5050	37	44	traditional	traditional	ADJ
easat-5050	37	45	methods	method	NOUN
easat-5050	37	46	.	.	PUNCT
easat-5050	38	1	deep	deep	ADJ
easat-5050	38	2	learning	learning	NOUN
easat-5050	38	3	methods	method	NOUN
easat-5050	38	4	can	can	AUX
easat-5050	38	5	handle	handle	VERB
easat-5050	38	6	large	large	ADJ
easat-5050	38	7	-	-	PUNCT
easat-5050	38	8	scale	scale	NOUN
easat-5050	38	9	datasets	dataset	NOUN
easat-5050	38	10	and	and	CCONJ
easat-5050	38	11	complex	complex	ADJ
easat-5050	38	12	signal	signal	NOUN
easat-5050	38	13	conditions	condition	NOUN
easat-5050	38	14	and	and	CCONJ
easat-5050	38	15	exhibit	exhibit	VERB
easat-5050	38	16	robustness	robustness	NOUN
easat-5050	38	17	and	and	CCONJ
easat-5050	38	18	generalisation	generalisation	NOUN
easat-5050	38	19	performance	performance	NOUN
easat-5050	38	20	.	.	PUNCT
easat-5050	39	1	the	the	DET
easat-5050	39	2	vibration	vibration	NOUN
easat-5050	39	3	signals	signal	NOUN
easat-5050	39	4	collected	collect	VERB
easat-5050	39	5	for	for	ADP
easat-5050	39	6	bearing	bear	VERB
easat-5050	39	7	diagnostics	diagnostic	NOUN
easat-5050	39	8	are	be	AUX
easat-5050	39	9	generally	generally	ADV
easat-5050	39	10	one	one	NUM
easat-5050	39	11	-	-	PUNCT
easat-5050	39	12	dimensional	dimensional	ADJ
easat-5050	39	13	time	time	NOUN
easat-5050	39	14	series	series	NOUN
easat-5050	39	15	signals	signal	NOUN
easat-5050	39	16	.	.	PUNCT
easat-5050	40	1	currently	currently	ADV
easat-5050	40	2	,	,	PUNCT
easat-5050	40	3	two	two	NUM
easat-5050	40	4	methods	method	NOUN
easat-5050	40	5	are	be	AUX
easat-5050	40	6	used	use	VERB
easat-5050	40	7	for	for	ADP
easat-5050	40	8	bearing	bear	VERB
easat-5050	40	9	diagnostics	diagnostic	NOUN
easat-5050	40	10	based	base	VERB
easat-5050	40	11	on	on	ADP
easat-5050	40	12	the	the	DET
easat-5050	40	13	input	input	NOUN
easat-5050	40	14	dimensions	dimension	NOUN
easat-5050	40	15	of	of	ADP
easat-5050	40	16	the	the	DET
easat-5050	40	17	network	network	NOUN
easat-5050	40	18	.	.	PUNCT
easat-5050	41	1	one	one	NUM
easat-5050	41	2	approach	approach	NOUN
easat-5050	41	3	is	be	AUX
easat-5050	41	4	to	to	PART
easat-5050	41	5	directly	directly	ADV
easat-5050	41	6	process	process	VERB
easat-5050	41	7	one	one	NUM
easat-5050	41	8	-	-	PUNCT
easat-5050	41	9	dimensional	dimensional	ADJ
easat-5050	41	10	data	datum	NOUN
easat-5050	41	11	.	.	PUNCT
easat-5050	42	1	the	the	DET
easat-5050	42	2	1dcnn	1dcnn	NUM
easat-5050	42	3	model	model	NOUN
easat-5050	42	4	proposed	propose	VERB
easat-5050	42	5	in	in	ADP
easat-5050	42	6	eren	eren	PROPN
easat-5050	42	7	,	,	PUNCT
easat-5050	42	8	ince	ince	NOUN
easat-5050	42	9	,	,	PUNCT
easat-5050	42	10	and	and	CCONJ
easat-5050	42	11	kiranyaz	kiranyaz	NOUN
easat-5050	42	12	(	(	PUNCT
easat-5050	42	13	2019	2019	NUM
easat-5050	42	14	)	)	PUNCT
easat-5050	42	15	;	;	PUNCT
easat-5050	42	16	ozcan	ozcan	X
easat-5050	42	17	,	,	PUNCT
easat-5050	42	18	devecioglu	devecioglu	PROPN
easat-5050	42	19	,	,	PUNCT
easat-5050	42	20	ince	ince	NOUN
easat-5050	42	21	,	,	PUNCT
easat-5050	42	22	eren	eren	PROPN
easat-5050	42	23	,	,	PUNCT
easat-5050	42	24	and	and	CCONJ
easat-5050	42	25	askar	askar	PROPN
easat-5050	42	26	(	(	PUNCT
easat-5050	42	27	2022	2022	NUM
easat-5050	42	28	)	)	PUNCT
easat-5050	42	29	and	and	CCONJ
easat-5050	42	30	huan	huan	PROPN
easat-5050	42	31	wang	wang	PROPN
easat-5050	42	32	,	,	PUNCT
easat-5050	42	33	liu	liu	PROPN
easat-5050	42	34	,	,	PUNCT
easat-5050	42	35	peng	peng	PROPN
easat-5050	42	36	,	,	PUNCT
easat-5050	42	37	and	and	CCONJ
easat-5050	42	38	qin	qin	PROPN
easat-5050	42	39	(	(	PUNCT
easat-5050	42	40	2019	2019	NUM
easat-5050	42	41	)	)	PUNCT
easat-5050	42	42	has	have	AUX
easat-5050	42	43	been	be	AUX
easat-5050	42	44	proven	prove	VERB
easat-5050	42	45	to	to	PART
easat-5050	42	46	have	have	VERB
easat-5050	42	47	highly	highly	ADV
easat-5050	42	48	generalizable	generalizable	ADJ
easat-5050	42	49	performance	performance	NOUN
easat-5050	42	50	for	for	ADP
easat-5050	42	51	bearing	bear	VERB
easat-5050	42	52	fault	fault	NOUN
easat-5050	42	53	diagnosis	diagnosis	NOUN
easat-5050	42	54	.	.	PUNCT
easat-5050	43	1	another	another	DET
easat-5050	43	2	approach	approach	NOUN
easat-5050	43	3	involves	involve	VERB
easat-5050	43	4	the	the	DET
easat-5050	43	5	transformation	transformation	NOUN
easat-5050	43	6	of	of	ADP
easat-5050	43	7	one	one	NUM
easat-5050	43	8	-	-	PUNCT
easat-5050	43	9	dimensional	dimensional	ADJ
easat-5050	43	10	data	datum	NOUN
easat-5050	43	11	into	into	ADP
easat-5050	43	12	two	two	NUM
easat-5050	43	13	-	-	PUNCT
easat-5050	43	14	dimensional	dimensional	ADJ
easat-5050	43	15	data	datum	NOUN
easat-5050	43	16	for	for	ADP
easat-5050	43	17	further	further	ADJ
easat-5050	43	18	processing	processing	NOUN
easat-5050	43	19	.	.	PUNCT
easat-5050	44	1	some	some	DET
easat-5050	44	2	studies	study	NOUN
easat-5050	44	3	converted	convert	VERB
easat-5050	44	4	the	the	DET
easat-5050	44	5	time	time	NOUN
easat-5050	44	6	-	-	PUNCT
easat-5050	44	7	domain	domain	NOUN
easat-5050	44	8	vibration	vibration	NOUN
easat-5050	44	9	signals	signal	NOUN
easat-5050	44	10	of	of	ADP
easat-5050	44	11	bearings	bearing	NOUN
easat-5050	44	12	into	into	ADP
easat-5050	44	13	grayscale	grayscale	NOUN
easat-5050	44	14	images	image	NOUN
easat-5050	44	15	containing	contain	VERB
easat-5050	44	16	fault	fault	NOUN
easat-5050	44	17	features	feature	NOUN
easat-5050	44	18	,	,	PUNCT
easat-5050	44	19	which	which	PRON
easat-5050	44	20	were	be	AUX
easat-5050	44	21	then	then	ADV
easat-5050	44	22	input	input	VERB
easat-5050	44	23	into	into	ADP
easat-5050	44	24	a	a	DET
easat-5050	44	25	2dcnn	2dcnn	NUM
easat-5050	44	26	model	model	NOUN
easat-5050	44	27	(	(	PUNCT
easat-5050	44	28	wan	wan	PROPN
easat-5050	44	29	,	,	PUNCT
easat-5050	44	30	chen	chen	PROPN
easat-5050	44	31	,	,	PUNCT
easat-5050	44	32	li	li	PROPN
easat-5050	44	33	,	,	PUNCT
easat-5050	44	34	&	&	CCONJ
easat-5050	44	35	li	li	PROPN
easat-5050	44	36	,	,	PUNCT
easat-5050	44	37	2020	2020	NUM
easat-5050	44	38	;	;	PUNCT
easat-5050	44	39	y.	y.	PROPN
easat-5050	44	40	wang	wang	PROPN
easat-5050	44	41	,	,	PUNCT
easat-5050	44	42	ding	ding	PROPN
easat-5050	44	43	,	,	PUNCT
easat-5050	44	44	zeng	zeng	PROPN
easat-5050	44	45	,	,	PUNCT
easat-5050	44	46	wang	wang	PROPN
easat-5050	44	47	,	,	PUNCT
easat-5050	44	48	&	&	CCONJ
easat-5050	44	49	shao	shao	PROPN
easat-5050	44	50	,	,	PUNCT
easat-5050	44	51	2020	2020	NUM
easat-5050	44	52	;	;	PUNCT
easat-5050	44	53	xu	xu	PROPN
easat-5050	44	54	,	,	PUNCT
easat-5050	44	55	liu	liu	PROPN
easat-5050	44	56	,	,	PUNCT
easat-5050	44	57	jiang	jiang	PROPN
easat-5050	44	58	,	,	PUNCT
easat-5050	44	59	söffker	söffker	PROPN
easat-5050	44	60	,	,	PUNCT
easat-5050	44	61	&	&	CCONJ
easat-5050	44	62	shen	shen	PROPN
easat-5050	44	63	,	,	PUNCT
easat-5050	44	64	2019	2019	NUM
easat-5050	44	65	)	)	PUNCT
easat-5050	44	66	.	.	PUNCT
easat-5050	45	1	the	the	DET
easat-5050	45	2	trained	train	VERB
easat-5050	45	3	2134	2134	NUM
easat-5050	45	4	edelweiss	edelweiss	PROPN
easat-5050	45	5	applied	apply	VERB
easat-5050	45	6	science	science	NOUN
easat-5050	45	7	and	and	CCONJ
easat-5050	45	8	technology	technology	NOUN
easat-5050	45	9	issn	issn	PROPN
easat-5050	45	10	:	:	PUNCT
easat-5050	45	11	2576	2576	NUM
easat-5050	45	12	-	-	SYM
easat-5050	45	13	8484	8484	NUM
easat-5050	45	14	vol	vol	NOUN
easat-5050	45	15	.	.	PROPN
easat-5050	46	1	9	9	NUM
easat-5050	46	2	,	,	PUNCT
easat-5050	46	3	no	no	INTJ
easat-5050	46	4	.	.	NOUN
easat-5050	46	5	2	2	NUM
easat-5050	46	6	:	:	SYM
easat-5050	46	7	2132	2132	NUM
easat-5050	46	8	-	-	SYM
easat-5050	46	9	2146	2146	NUM
easat-5050	46	10	,	,	PUNCT
easat-5050	46	11	2025	2025	NUM
easat-5050	46	12	doi	doi	NOUN
easat-5050	46	13	:	:	PUNCT
easat-5050	46	14	10.55214/25768484.v9i2.5050	10.55214/25768484.v9i2.5050	NUM
easat-5050	46	15	©	©	PROPN
easat-5050	46	16	2025	2025	NUM
easat-5050	46	17	by	by	ADP
easat-5050	46	18	the	the	DET
easat-5050	46	19	authors	author	NOUN
easat-5050	46	20	;	;	PUNCT
easat-5050	46	21	licensee	licensee	PROPN
easat-5050	46	22	learning	learning	PROPN
easat-5050	46	23	gate	gate	NOUN
easat-5050	46	24	model	model	NOUN
easat-5050	46	25	can	can	AUX
easat-5050	46	26	effectively	effectively	ADV
easat-5050	46	27	recognise	recognise	VERB
easat-5050	46	28	the	the	DET
easat-5050	46	29	types	type	NOUN
easat-5050	46	30	of	of	ADP
easat-5050	46	31	faults	fault	NOUN
easat-5050	46	32	and	and	CCONJ
easat-5050	46	33	their	their	PRON
easat-5050	46	34	severity	severity	NOUN
easat-5050	46	35	.	.	PUNCT
easat-5050	47	1	in	in	ADP
easat-5050	47	2	comparison	comparison	NOUN
easat-5050	47	3	,	,	PUNCT
easat-5050	47	4	the	the	DET
easat-5050	47	5	1dcnn	1dcnn	NUM
easat-5050	47	6	network	network	NOUN
easat-5050	47	7	that	that	PRON
easat-5050	47	8	directly	directly	ADV
easat-5050	47	9	processes	process	VERB
easat-5050	47	10	one	one	NUM
easat-5050	47	11	-	-	PUNCT
easat-5050	47	12	dimensional	dimensional	ADJ
easat-5050	47	13	data	datum	NOUN
easat-5050	47	14	is	be	AUX
easat-5050	47	15	simpler	simple	ADJ
easat-5050	47	16	,	,	PUNCT
easat-5050	47	17	whereas	whereas	SCONJ
easat-5050	47	18	converting	convert	VERB
easat-5050	47	19	from	from	ADP
easat-5050	47	20	onedimensional	onedimensional	ADJ
easat-5050	47	21	to	to	ADP
easat-5050	47	22	two	two	NUM
easat-5050	47	23	-	-	PUNCT
easat-5050	47	24	dimensional	dimensional	ADJ
easat-5050	47	25	data	datum	NOUN
easat-5050	47	26	may	may	AUX
easat-5050	47	27	result	result	VERB
easat-5050	47	28	in	in	ADP
easat-5050	47	29	some	some	DET
easat-5050	47	30	information	information	NOUN
easat-5050	47	31	loss	loss	NOUN
easat-5050	47	32	.	.	PUNCT
easat-5050	48	1	therefore	therefore	ADV
easat-5050	48	2	,	,	PUNCT
easat-5050	48	3	a	a	DET
easat-5050	48	4	w1dcnn	w1dcnn	PROPN
easat-5050	48	5	model	model	NOUN
easat-5050	48	6	was	be	AUX
easat-5050	48	7	proposed	propose	VERB
easat-5050	48	8	in	in	ADP
easat-5050	48	9	zhang	zhang	PROPN
easat-5050	48	10	,	,	PUNCT
easat-5050	48	11	peng	peng	PROPN
easat-5050	48	12	,	,	PUNCT
easat-5050	48	13	li	li	PROPN
easat-5050	48	14	,	,	PUNCT
easat-5050	48	15	chen	chen	PROPN
easat-5050	48	16	,	,	PUNCT
easat-5050	48	17	and	and	CCONJ
easat-5050	48	18	zhang	zhang	PROPN
easat-5050	48	19	(	(	PUNCT
easat-5050	48	20	2017	2017	NUM
easat-5050	48	21	)	)	PUNCT
easat-5050	48	22	and	and	CCONJ
easat-5050	48	23	it	it	PRON
easat-5050	48	24	has	have	VERB
easat-5050	48	25	two	two	NUM
easat-5050	48	26	main	main	ADJ
easat-5050	48	27	characteristics	characteristic	NOUN
easat-5050	48	28	.	.	PUNCT
easat-5050	49	1	the	the	DET
easat-5050	49	2	first	first	ADJ
easat-5050	49	3	layer	layer	NOUN
easat-5050	49	4	uses	use	VERB
easat-5050	49	5	a	a	DET
easat-5050	49	6	wide	wide	ADJ
easat-5050	49	7	64	64	NUM
easat-5050	49	8	-	-	PUNCT
easat-5050	49	9	channel	channel	NOUN
easat-5050	49	10	convolutional	convolutional	ADJ
easat-5050	49	11	kernel	kernel	NOUN
easat-5050	49	12	,	,	PUNCT
easat-5050	49	13	whereas	whereas	SCONJ
easat-5050	49	14	the	the	DET
easat-5050	49	15	other	other	ADJ
easat-5050	49	16	layers	layer	NOUN
easat-5050	49	17	use	use	VERB
easat-5050	49	18	a	a	DET
easat-5050	49	19	network	network	NOUN
easat-5050	49	20	structure	structure	NOUN
easat-5050	49	21	with	with	ADP
easat-5050	49	22	small	small	ADJ
easat-5050	49	23	3×3	3×3	NUM
easat-5050	49	24	convolutional	convolutional	ADJ
easat-5050	49	25	kernels	kernel	NOUN
easat-5050	49	26	.	.	PUNCT
easat-5050	50	1	a	a	DET
easat-5050	50	2	high	high	ADJ
easat-5050	50	3	accuracy	accuracy	NOUN
easat-5050	50	4	was	be	AUX
easat-5050	50	5	achieved	achieve	VERB
easat-5050	50	6	with	with	ADP
easat-5050	50	7	this	this	DET
easat-5050	50	8	model	model	NOUN
easat-5050	50	9	.	.	PUNCT
easat-5050	51	1	based	base	VERB
easat-5050	51	2	on	on	ADP
easat-5050	51	3	the	the	DET
easat-5050	51	4	success	success	NOUN
easat-5050	51	5	of	of	ADP
easat-5050	51	6	the	the	DET
easat-5050	51	7	w1dcnn	w1dcnn	PROPN
easat-5050	51	8	in	in	ADP
easat-5050	51	9	diagnostics	diagnostic	NOUN
easat-5050	51	10	,	,	PUNCT
easat-5050	51	11	researchers	researcher	NOUN
easat-5050	51	12	have	have	AUX
easat-5050	51	13	proposed	propose	VERB
easat-5050	51	14	several	several	ADJ
easat-5050	51	15	improved	improved	ADJ
easat-5050	51	16	versions	version	NOUN
easat-5050	51	17	of	of	ADP
easat-5050	51	18	it	it	PRON
easat-5050	51	19	y.	y.	PROPN
easat-5050	51	20	wang	wang	PROPN
easat-5050	51	21	et	et	PROPN
easat-5050	51	22	al	al	PROPN
easat-5050	51	23	.	.	PROPN
easat-5050	52	1	(	(	PUNCT
easat-5050	52	2	2020	2020	NUM
easat-5050	52	3	)	)	PUNCT
easat-5050	52	4	and	and	CCONJ
easat-5050	52	5	shenfield	shenfield	NOUN
easat-5050	52	6	and	and	CCONJ
easat-5050	52	7	howarth	howarth	NOUN
easat-5050	52	8	(	(	PUNCT
easat-5050	52	9	2020	2020	NUM
easat-5050	52	10	)	)	PUNCT
easat-5050	52	11	.	.	PUNCT
easat-5050	53	1	recently	recently	ADV
easat-5050	53	2	,	,	PUNCT
easat-5050	53	3	owing	owe	VERB
easat-5050	53	4	to	to	ADP
easat-5050	53	5	the	the	DET
easat-5050	53	6	effective	effective	ADJ
easat-5050	53	7	feature	feature	NOUN
easat-5050	53	8	-	-	PUNCT
easat-5050	53	9	representation	representation	NOUN
easat-5050	53	10	capabilities	capability	NOUN
easat-5050	53	11	of	of	ADP
easat-5050	53	12	quadratic	quadratic	ADJ
easat-5050	53	13	neural	neural	ADJ
easat-5050	53	14	networks	network	NOUN
easat-5050	53	15	(	(	PUNCT
easat-5050	53	16	fan	fan	PROPN
easat-5050	53	17	,	,	PUNCT
easat-5050	53	18	xiong	xiong	PROPN
easat-5050	53	19	,	,	PUNCT
easat-5050	53	20	&	&	CCONJ
easat-5050	53	21	wang	wang	PROPN
easat-5050	53	22	,	,	PUNCT
easat-5050	53	23	2020	2020	NUM
easat-5050	53	24	;	;	PUNCT
easat-5050	53	25	j.x	j.x	PROPN
easat-5050	53	26	.	.	PROPN
easat-5050	53	27	liao	liao	PROPN
easat-5050	53	28	et	et	PROPN
easat-5050	53	29	al	al	PROPN
easat-5050	53	30	.	.	PROPN
easat-5050	53	31	,	,	PUNCT
easat-5050	53	32	2023	2023	NUM
easat-5050	53	33	)	)	PUNCT
easat-5050	53	34	,	,	PUNCT
easat-5050	53	35	the	the	DET
easat-5050	53	36	traditional	traditional	ADJ
easat-5050	53	37	linear	linear	ADJ
easat-5050	53	38	neural	neural	ADJ
easat-5050	53	39	network	network	NOUN
easat-5050	53	40	convolution	convolution	NOUN
easat-5050	53	41	in	in	ADP
easat-5050	53	42	the	the	DET
easat-5050	53	43	w1dcnn	w1dcnn	PROPN
easat-5050	53	44	network	network	NOUN
easat-5050	53	45	has	have	AUX
easat-5050	53	46	been	be	AUX
easat-5050	53	47	replaced	replace	VERB
easat-5050	53	48	with	with	ADP
easat-5050	53	49	quadratic	quadratic	ADJ
easat-5050	53	50	neural	neural	ADJ
easat-5050	53	51	elements	element	NOUN
easat-5050	53	52	,	,	PUNCT
easat-5050	53	53	resulting	result	VERB
easat-5050	53	54	in	in	ADP
easat-5050	53	55	a	a	DET
easat-5050	53	56	qcnn	qcnn	PROPN
easat-5050	53	57	model	model	NOUN
easat-5050	53	58	.	.	PUNCT
easat-5050	54	1	many	many	ADJ
easat-5050	54	2	researchers	researcher	NOUN
easat-5050	54	3	have	have	AUX
easat-5050	54	4	introduced	introduce	VERB
easat-5050	54	5	attention	attention	NOUN
easat-5050	54	6	mechanisms	mechanism	NOUN
easat-5050	54	7	into	into	ADP
easat-5050	54	8	bearing	bear	VERB
easat-5050	54	9	fault	fault	NOUN
easat-5050	54	10	diagnosis	diagnosis	NOUN
easat-5050	54	11	tasks	task	NOUN
easat-5050	54	12	,	,	PUNCT
easat-5050	54	13	enabling	enable	VERB
easat-5050	54	14	traditional	traditional	ADJ
easat-5050	54	15	cnn	cnn	PROPN
easat-5050	54	16	models	model	NOUN
easat-5050	54	17	to	to	PART
easat-5050	54	18	achieve	achieve	VERB
easat-5050	54	19	higher	high	ADJ
easat-5050	54	20	classification	classification	NOUN
easat-5050	54	21	accuracy	accuracy	NOUN
easat-5050	54	22	(	(	PUNCT
easat-5050	54	23	huan	huan	PROPN
easat-5050	54	24	wang	wang	PROPN
easat-5050	54	25	et	et	PROPN
easat-5050	54	26	al	al	PROPN
easat-5050	54	27	.	.	PROPN
easat-5050	54	28	,	,	PUNCT
easat-5050	54	29	2019	2019	NUM
easat-5050	54	30	;	;	PUNCT
easat-5050	54	31	hui	hui	PROPN
easat-5050	54	32	wang	wang	PROPN
easat-5050	54	33	,	,	PUNCT
easat-5050	54	34	xu	xu	PROPN
easat-5050	54	35	,	,	PUNCT
easat-5050	54	36	yan	yan	PROPN
easat-5050	54	37	,	,	PUNCT
easat-5050	54	38	sun	sun	PROPN
easat-5050	54	39	,	,	PUNCT
easat-5050	54	40	&	&	CCONJ
easat-5050	54	41	chen	chen	PROPN
easat-5050	54	42	,	,	PUNCT
easat-5050	54	43	2020	2020	NUM
easat-5050	54	44	)	)	PUNCT
easat-5050	54	45	.	.	PUNCT
easat-5050	55	1	although	although	SCONJ
easat-5050	55	2	deep	deep	ADJ
easat-5050	55	3	learning	learning	NOUN
easat-5050	55	4	has	have	AUX
easat-5050	55	5	achieved	achieve	VERB
easat-5050	55	6	significant	significant	ADJ
easat-5050	55	7	success	success	NOUN
easat-5050	55	8	in	in	ADP
easat-5050	55	9	bearing	bear	VERB
easat-5050	55	10	fault	fault	NOUN
easat-5050	55	11	diagnosis	diagnosis	NOUN
easat-5050	55	12	research	research	NOUN
easat-5050	55	13	,	,	PUNCT
easat-5050	55	14	current	current	ADJ
easat-5050	55	15	deep	deep	ADJ
easat-5050	55	16	learning	learning	NOUN
easat-5050	55	17	models	model	NOUN
easat-5050	55	18	have	have	VERB
easat-5050	55	19	some	some	DET
easat-5050	55	20	limitations	limitation	NOUN
easat-5050	55	21	.	.	PUNCT
easat-5050	56	1	many	many	ADJ
easat-5050	56	2	deep	deep	ADJ
easat-5050	56	3	learning	learning	NOUN
easat-5050	56	4	models	model	NOUN
easat-5050	56	5	perform	perform	VERB
easat-5050	56	6	well	well	ADV
easat-5050	56	7	on	on	ADP
easat-5050	56	8	clean	clean	ADJ
easat-5050	56	9	signals	signal	NOUN
easat-5050	56	10	with	with	ADP
easat-5050	56	11	consistent	consistent	ADJ
easat-5050	56	12	data	datum	NOUN
easat-5050	56	13	distributions	distribution	NOUN
easat-5050	56	14	.	.	PUNCT
easat-5050	57	1	however	however	ADV
easat-5050	57	2	,	,	PUNCT
easat-5050	57	3	when	when	SCONJ
easat-5050	57	4	the	the	DET
easat-5050	57	5	signals	signal	NOUN
easat-5050	57	6	contain	contain	VERB
easat-5050	57	7	a	a	DET
easat-5050	57	8	large	large	ADJ
easat-5050	57	9	amount	amount	NOUN
easat-5050	57	10	of	of	ADP
easat-5050	57	11	noise	noise	NOUN
easat-5050	57	12	,	,	PUNCT
easat-5050	57	13	the	the	DET
easat-5050	57	14	classification	classification	NOUN
easat-5050	57	15	performance	performance	NOUN
easat-5050	57	16	of	of	ADP
easat-5050	57	17	the	the	DET
easat-5050	57	18	models	model	NOUN
easat-5050	57	19	decreases	decrease	VERB
easat-5050	57	20	significantly	significantly	ADV
easat-5050	57	21	.	.	PUNCT
easat-5050	58	1	for	for	ADP
easat-5050	58	2	example	example	NOUN
easat-5050	58	3	,	,	PUNCT
easat-5050	58	4	when	when	SCONJ
easat-5050	58	5	a	a	DET
easat-5050	58	6	signal	signal	NOUN
easat-5050	58	7	-	-	PUNCT
easat-5050	58	8	to	to	ADP
easat-5050	58	9	-	-	PUNCT
easat-5050	58	10	noise	noise	NOUN
easat-5050	58	11	ratio	ratio	NOUN
easat-5050	58	12	(	(	PUNCT
easat-5050	58	13	snr	snr	PROPN
easat-5050	58	14	)	)	PUNCT
easat-5050	58	15	of	of	ADP
easat-5050	58	16	-6	-6	PROPN
easat-5050	58	17	is	be	AUX
easat-5050	58	18	added	add	VERB
easat-5050	58	19	to	to	ADP
easat-5050	58	20	the	the	DET
easat-5050	58	21	cwru	cwru	PROPN
easat-5050	58	22	dataset	dataset	PROPN
easat-5050	58	23	,	,	PUNCT
easat-5050	58	24	the	the	DET
easat-5050	58	25	classification	classification	NOUN
easat-5050	58	26	accuracies	accuracy	NOUN
easat-5050	58	27	of	of	ADP
easat-5050	58	28	aresnet	aresnet	PROPN
easat-5050	58	29	(	(	PUNCT
easat-5050	58	30	zhong	zhong	PROPN
easat-5050	58	31	,	,	PUNCT
easat-5050	58	32	wang	wang	PROPN
easat-5050	58	33	,	,	PUNCT
easat-5050	58	34	&	&	CCONJ
easat-5050	58	35	ban	ban	PROPN
easat-5050	58	36	,	,	PUNCT
easat-5050	58	37	2022	2022	NUM
easat-5050	58	38	)	)	PUNCT
easat-5050	58	39	w1dcnn	w1dcnn	PROPN
easat-5050	58	40	(	(	PUNCT
easat-5050	58	41	zhang	zhang	PROPN
easat-5050	58	42	et	et	PROPN
easat-5050	58	43	al	al	PROPN
easat-5050	58	44	.	.	PROPN
easat-5050	58	45	,	,	PUNCT
easat-5050	58	46	2017	2017	NUM
easat-5050	58	47	)	)	PUNCT
easat-5050	58	48	and	and	CCONJ
easat-5050	58	49	qcnn	qcnn	PROPN
easat-5050	58	50	(	(	PUNCT
easat-5050	58	51	j.-x	j.-x	PROPN
easat-5050	58	52	.	.	PUNCT
easat-5050	59	1	liao	liao	PROPN
easat-5050	59	2	et	et	PROPN
easat-5050	59	3	al	al	PROPN
easat-5050	59	4	.	.	PROPN
easat-5050	59	5	,	,	PUNCT
easat-5050	59	6	2023	2023	NUM
easat-5050	59	7	)	)	PUNCT
easat-5050	59	8	are	be	AUX
easat-5050	59	9	61.2	61.2	NUM
easat-5050	59	10	%	%	NOUN
easat-5050	59	11	,	,	PUNCT
easat-5050	59	12	66.5	66.5	NUM
easat-5050	59	13	%	%	NOUN
easat-5050	59	14	,	,	PUNCT
easat-5050	59	15	and	and	CCONJ
easat-5050	59	16	71.1	71.1	NUM
easat-5050	59	17	%	%	NOUN
easat-5050	59	18	,	,	PUNCT
easat-5050	59	19	respectively	respectively	ADV
easat-5050	59	20	.	.	PUNCT
easat-5050	60	1	these	these	DET
easat-5050	60	2	models	model	NOUN
easat-5050	60	3	require	require	VERB
easat-5050	60	4	further	further	ADJ
easat-5050	60	5	improvements	improvement	NOUN
easat-5050	60	6	in	in	ADP
easat-5050	60	7	their	their	PRON
easat-5050	60	8	classification	classification	NOUN
easat-5050	60	9	accuracy	accuracy	NOUN
easat-5050	60	10	in	in	ADP
easat-5050	60	11	noisy	noisy	ADJ
easat-5050	60	12	environments	environment	NOUN
easat-5050	60	13	.	.	PUNCT
easat-5050	61	1	noise	noise	NOUN
easat-5050	61	2	is	be	AUX
easat-5050	61	3	ubiquitous	ubiquitous	ADJ
easat-5050	61	4	in	in	ADP
easat-5050	61	5	industrial	industrial	ADJ
easat-5050	61	6	settings	setting	NOUN
easat-5050	61	7	,	,	PUNCT
easat-5050	61	8	making	make	VERB
easat-5050	61	9	it	it	PRON
easat-5050	61	10	crucial	crucial	ADJ
easat-5050	61	11	to	to	PART
easat-5050	61	12	efficiently	efficiently	ADV
easat-5050	61	13	diagnose	diagnose	VERB
easat-5050	61	14	signals	signal	NOUN
easat-5050	61	15	that	that	PRON
easat-5050	61	16	contain	contain	VERB
easat-5050	61	17	multiple	multiple	ADJ
easat-5050	61	18	sources	source	NOUN
easat-5050	61	19	of	of	ADP
easat-5050	61	20	noise	noise	NOUN
easat-5050	61	21	.	.	PUNCT
easat-5050	62	1	based	base	VERB
easat-5050	62	2	on	on	ADP
easat-5050	62	3	the	the	DET
easat-5050	62	4	aforementioned	aforementioned	ADJ
easat-5050	62	5	problems	problem	NOUN
easat-5050	62	6	,	,	PUNCT
easat-5050	62	7	this	this	DET
easat-5050	62	8	paper	paper	NOUN
easat-5050	62	9	proposes	propose	VERB
easat-5050	62	10	a	a	DET
easat-5050	62	11	new	new	ADJ
easat-5050	62	12	network	network	NOUN
easat-5050	62	13	architecture	architecture	NOUN
easat-5050	62	14	called	call	VERB
easat-5050	62	15	mcfcnn	mcfcnn	PROPN
easat-5050	62	16	.	.	PUNCT
easat-5050	63	1	the	the	DET
easat-5050	63	2	network	network	NOUN
easat-5050	63	3	model	model	NOUN
easat-5050	63	4	contains	contain	VERB
easat-5050	63	5	multi	multi	ADJ
easat-5050	63	6	-	-	ADJ
easat-5050	63	7	scale	scale	ADJ
easat-5050	63	8	processing	processing	NOUN
easat-5050	63	9	,	,	PUNCT
easat-5050	63	10	channel	channel	NOUN
easat-5050	63	11	attention	attention	NOUN
easat-5050	63	12	,	,	PUNCT
easat-5050	63	13	feature	feature	NOUN
easat-5050	63	14	enhancement	enhancement	NOUN
easat-5050	63	15	,	,	PUNCT
easat-5050	63	16	and	and	CCONJ
easat-5050	63	17	anomaly	anomaly	NOUN
easat-5050	63	18	classification	classification	NOUN
easat-5050	63	19	layers	layer	NOUN
easat-5050	63	20	,	,	PUNCT
easat-5050	63	21	which	which	PRON
easat-5050	63	22	are	be	AUX
easat-5050	63	23	used	use	VERB
easat-5050	63	24	for	for	ADP
easat-5050	63	25	the	the	DET
easat-5050	63	26	efficient	efficient	ADJ
easat-5050	63	27	diagnosis	diagnosis	NOUN
easat-5050	63	28	of	of	ADP
easat-5050	63	29	bearing	bear	VERB
easat-5050	63	30	faults	fault	NOUN
easat-5050	63	31	in	in	ADP
easat-5050	63	32	multi	multi	ADJ
easat-5050	63	33	-	-	ADJ
easat-5050	63	34	noise	noise	ADJ
easat-5050	63	35	environments	environment	NOUN
easat-5050	63	36	.	.	PUNCT
easat-5050	64	1	2	2	X
easat-5050	64	2	.	.	X
easat-5050	64	3	methods	method	NOUN
easat-5050	64	4	this	this	DET
easat-5050	64	5	paper	paper	NOUN
easat-5050	64	6	utilises	utilise	NOUN
easat-5050	64	7	time	time	NOUN
easat-5050	64	8	-	-	PUNCT
easat-5050	64	9	series	series	NOUN
easat-5050	64	10	vibration	vibration	NOUN
easat-5050	64	11	data	datum	NOUN
easat-5050	64	12	collected	collect	VERB
easat-5050	64	13	during	during	ADP
easat-5050	64	14	the	the	DET
easat-5050	64	15	operation	operation	NOUN
easat-5050	64	16	of	of	ADP
easat-5050	64	17	bearings	bearing	NOUN
easat-5050	64	18	for	for	ADP
easat-5050	64	19	fault	fault	NOUN
easat-5050	64	20	diagnosis	diagnosis	NOUN
easat-5050	64	21	.	.	PUNCT
easat-5050	65	1	the	the	DET
easat-5050	65	2	diagnostic	diagnostic	ADJ
easat-5050	65	3	problem	problem	NOUN
easat-5050	65	4	can	can	AUX
easat-5050	65	5	be	be	AUX
easat-5050	65	6	defined	define	VERB
easat-5050	65	7	as	as	SCONJ
easat-5050	65	8	follows	follow	VERB
easat-5050	65	9	:	:	PUNCT
easat-5050	65	10	the	the	DET
easat-5050	65	11	input	input	NOUN
easat-5050	65	12	is	be	AUX
easat-5050	65	13	𝑋𝑡	𝑋𝑡	PROPN
easat-5050	65	14	∈	∈	PROPN
easat-5050	65	15	𝑅𝑘	𝑅𝑘	NOUN
easat-5050	65	16	,	,	PUNCT
easat-5050	65	17	where	where	SCONJ
easat-5050	65	18	𝑡	𝑡	X
easat-5050	65	19	=	=	SYM
easat-5050	65	20	(	(	PUNCT
easat-5050	65	21	0	0	NUM
easat-5050	65	22	,	,	PUNCT
easat-5050	65	23	1	1	NUM
easat-5050	65	24	,	,	PUNCT
easat-5050	65	25	2	2	NUM
easat-5050	65	26	,	,	PUNCT
easat-5050	65	27	…	…	PUNCT
easat-5050	65	28	,	,	PUNCT
easat-5050	65	29	𝑇	𝑇	PROPN
easat-5050	65	30	)	)	PUNCT
easat-5050	65	31	.	.	PUNCT
easat-5050	66	1	here	here	ADV
easat-5050	66	2	,	,	PUNCT
easat-5050	66	3	𝑇	𝑇	PROPN
easat-5050	66	4	is	be	AUX
easat-5050	66	5	the	the	DET
easat-5050	66	6	time	time	NOUN
easat-5050	66	7	step	step	NOUN
easat-5050	66	8	,	,	PUNCT
easat-5050	66	9	and	and	CCONJ
easat-5050	66	10	𝑘	𝑘	PRON
easat-5050	66	11	is	be	AUX
easat-5050	66	12	the	the	DET
easat-5050	66	13	data	data	NOUN
easat-5050	66	14	type	type	NOUN
easat-5050	66	15	,	,	PUNCT
easat-5050	66	16	which	which	PRON
easat-5050	66	17	includes	include	VERB
easat-5050	66	18	both	both	DET
easat-5050	66	19	fault	fault	VERB
easat-5050	66	20	data	datum	NOUN
easat-5050	66	21	and	and	CCONJ
easat-5050	66	22	healthy	healthy	ADJ
easat-5050	66	23	data	datum	NOUN
easat-5050	66	24	of	of	ADP
easat-5050	66	25	the	the	DET
easat-5050	66	26	bearings	bearing	NOUN
easat-5050	66	27	.	.	PUNCT
easat-5050	67	1	the	the	DET
easat-5050	67	2	corresponding	corresponding	ADJ
easat-5050	67	3	output	output	NOUN
easat-5050	67	4	is	be	AUX
easat-5050	67	5	the	the	DET
easat-5050	67	6	fault	fault	NOUN
easat-5050	67	7	type	type	NOUN
easat-5050	67	8	𝑌𝑡	𝑌𝑡	PROPN
easat-5050	67	9	,	,	PUNCT
easat-5050	67	10	for	for	ADP
easat-5050	67	11	model	model	NOUN
easat-5050	67	12	fault	fault	NOUN
easat-5050	67	13	diagnosis	diagnosis	NOUN
easat-5050	67	14	.	.	PUNCT
easat-5050	68	1	the	the	DET
easat-5050	68	2	goal	goal	NOUN
easat-5050	68	3	is	be	AUX
easat-5050	68	4	to	to	PART
easat-5050	68	5	achieve	achieve	VERB
easat-5050	68	6	fault	fault	NOUN
easat-5050	68	7	diagnosis	diagnosis	NOUN
easat-5050	68	8	by	by	ADP
easat-5050	68	9	establishing	establish	VERB
easat-5050	68	10	a	a	DET
easat-5050	68	11	mapping	mapping	NOUN
easat-5050	68	12	relationship	relationship	NOUN
easat-5050	68	13	between	between	ADP
easat-5050	68	14	𝑌𝑡	𝑌𝑡	PROPN
easat-5050	68	15	and	and	CCONJ
easat-5050	68	16	the	the	DET
easat-5050	68	17	dataset	dataset	NOUN
easat-5050	68	18	,	,	PUNCT
easat-5050	68	19	as	as	SCONJ
easat-5050	68	20	follows	follow	VERB
easat-5050	68	21	:	:	PUNCT
easat-5050	68	22	𝑌𝑡	𝑌𝑡	PROPN
easat-5050	68	23	=	=	SYM
easat-5050	68	24	𝑓(𝑋	𝑓(𝑋	PROPN
easat-5050	68	25	)	)	PUNCT
easat-5050	68	26	(	(	PUNCT
easat-5050	68	27	1	1	X
easat-5050	68	28	)	)	PUNCT
easat-5050	68	29	where	where	SCONJ
easat-5050	68	30	𝑋	𝑋	NOUN
easat-5050	68	31	is	be	AUX
easat-5050	68	32	the	the	DET
easat-5050	68	33	bearing	bear	VERB
easat-5050	68	34	operation	operation	NOUN
easat-5050	68	35	data	datum	NOUN
easat-5050	68	36	,	,	PUNCT
easat-5050	68	37	𝑓	𝑓	PRON
easat-5050	68	38	is	be	AUX
easat-5050	68	39	the	the	DET
easat-5050	68	40	mapping	mapping	NOUN
easat-5050	68	41	function	function	NOUN
easat-5050	68	42	,	,	PUNCT
easat-5050	68	43	and	and	CCONJ
easat-5050	68	44	𝑌𝑡	𝑌𝑡	PROPN
easat-5050	68	45	is	be	AUX
easat-5050	68	46	the	the	DET
easat-5050	68	47	fault	fault	NOUN
easat-5050	68	48	classification	classification	NOUN
easat-5050	68	49	result	result	NOUN
easat-5050	68	50	of	of	ADP
easat-5050	68	51	𝑓	𝑓	DET
easat-5050	68	52	diagnosis	diagnosis	NOUN
easat-5050	68	53	.	.	PUNCT
easat-5050	69	1	the	the	DET
easat-5050	69	2	proposed	propose	VERB
easat-5050	69	3	bearing	bear	VERB
easat-5050	69	4	fault	fault	NOUN
easat-5050	69	5	diagnosis	diagnosis	NOUN
easat-5050	69	6	network	network	NOUN
easat-5050	69	7	is	be	AUX
easat-5050	69	8	shown	show	VERB
easat-5050	69	9	in	in	ADP
easat-5050	69	10	figure	figure	NOUN
easat-5050	69	11	2	2	NUM
easat-5050	69	12	.	.	PUNCT
easat-5050	70	1	the	the	DET
easat-5050	70	2	entire	entire	ADJ
easat-5050	70	3	network	network	NOUN
easat-5050	70	4	consists	consist	VERB
easat-5050	70	5	of	of	ADP
easat-5050	70	6	four	four	NUM
easat-5050	70	7	layers	layer	NOUN
easat-5050	70	8	:	:	PUNCT
easat-5050	70	9	multi	multi	ADJ
easat-5050	70	10	-	-	ADJ
easat-5050	70	11	scale	scale	ADJ
easat-5050	70	12	processing	processing	NOUN
easat-5050	70	13	,	,	PUNCT
easat-5050	70	14	channel	channel	NOUN
easat-5050	70	15	attention	attention	NOUN
easat-5050	70	16	,	,	PUNCT
easat-5050	70	17	feature	feature	NOUN
easat-5050	70	18	enhancement	enhancement	NOUN
easat-5050	70	19	,	,	PUNCT
easat-5050	70	20	and	and	CCONJ
easat-5050	70	21	anomaly	anomaly	NOUN
easat-5050	70	22	classification	classification	NOUN
easat-5050	70	23	layers	layer	NOUN
easat-5050	70	24	.	.	PUNCT
easat-5050	71	1	2135	2135	NUM
easat-5050	71	2	edelweiss	edelweiss	PROPN
easat-5050	71	3	applied	apply	VERB
easat-5050	71	4	science	science	NOUN
easat-5050	71	5	and	and	CCONJ
easat-5050	71	6	technology	technology	NOUN
easat-5050	71	7	issn	issn	PROPN
easat-5050	71	8	:	:	PUNCT
easat-5050	71	9	2576	2576	NUM
easat-5050	71	10	-	-	SYM
easat-5050	71	11	8484	8484	NUM
easat-5050	71	12	vol	vol	NOUN
easat-5050	71	13	.	.	PROPN
easat-5050	72	1	9	9	NUM
easat-5050	72	2	,	,	PUNCT
easat-5050	72	3	no	no	INTJ
easat-5050	72	4	.	.	NOUN
easat-5050	72	5	2	2	NUM
easat-5050	72	6	:	:	SYM
easat-5050	72	7	2132	2132	NUM
easat-5050	72	8	-	-	SYM
easat-5050	72	9	2146	2146	NUM
easat-5050	72	10	,	,	PUNCT
easat-5050	72	11	2025	2025	NUM
easat-5050	72	12	doi	doi	NOUN
easat-5050	72	13	:	:	PUNCT
easat-5050	72	14	10.55214/25768484.v9i2.5050	10.55214/25768484.v9i2.5050	NUM
easat-5050	72	15	©	©	PROPN
easat-5050	72	16	2025	2025	NUM
easat-5050	72	17	by	by	ADP
easat-5050	72	18	the	the	DET
easat-5050	72	19	authors	author	NOUN
easat-5050	72	20	;	;	PUNCT
easat-5050	72	21	licensee	licensee	PROPN
easat-5050	72	22	learning	learn	VERB
easat-5050	72	23	gate	gate	NOUN
easat-5050	72	24	figure	figure	NOUN
easat-5050	72	25	2	2	NUM
easat-5050	73	1	.	.	PUNCT
easat-5050	73	2	mcfcnn	mcfcnn	PROPN
easat-5050	73	3	network	network	NOUN
easat-5050	73	4	architecture	architecture	NOUN
easat-5050	73	5	.	.	PUNCT
easat-5050	74	1	2.1	2.1	NUM
easat-5050	74	2	.	.	PUNCT
easat-5050	75	1	multi	multi	ADJ
easat-5050	75	2	-	-	ADJ
easat-5050	75	3	scale	scale	ADJ
easat-5050	75	4	processing	processing	NOUN
easat-5050	75	5	layer	layer	NOUN
easat-5050	75	6	the	the	DET
easat-5050	75	7	multi	multi	ADJ
easat-5050	75	8	-	-	ADJ
easat-5050	75	9	scale	scale	ADJ
easat-5050	75	10	processing	processing	NOUN
easat-5050	75	11	layer	layer	NOUN
easat-5050	75	12	comprises	comprise	VERB
easat-5050	75	13	a	a	DET
easat-5050	75	14	parallel	parallel	ADJ
easat-5050	75	15	convolution	convolution	NOUN
easat-5050	75	16	operation	operation	NOUN
easat-5050	75	17	that	that	PRON
easat-5050	75	18	utilises	utilise	VERB
easat-5050	75	19	multiple	multiple	ADJ
easat-5050	75	20	parallel	parallel	ADJ
easat-5050	75	21	convolution	convolution	NOUN
easat-5050	75	22	branches	branch	NOUN
easat-5050	75	23	to	to	PART
easat-5050	75	24	extract	extract	VERB
easat-5050	75	25	different	different	ADJ
easat-5050	75	26	features	feature	NOUN
easat-5050	75	27	.	.	PUNCT
easat-5050	76	1	each	each	DET
easat-5050	76	2	branch	branch	NOUN
easat-5050	76	3	can	can	AUX
easat-5050	76	4	use	use	VERB
easat-5050	76	5	different	different	ADJ
easat-5050	76	6	convolution	convolution	NOUN
easat-5050	76	7	kernels	kernel	NOUN
easat-5050	76	8	to	to	PART
easat-5050	76	9	process	process	VERB
easat-5050	76	10	input	input	NOUN
easat-5050	76	11	data	datum	NOUN
easat-5050	76	12	and	and	CCONJ
easat-5050	76	13	generate	generate	VERB
easat-5050	76	14	a	a	DET
easat-5050	76	15	set	set	NOUN
easat-5050	76	16	of	of	ADP
easat-5050	76	17	feature	feature	NOUN
easat-5050	76	18	maps	map	NOUN
easat-5050	76	19	.	.	PUNCT
easat-5050	77	1	these	these	DET
easat-5050	77	2	feature	feature	NOUN
easat-5050	77	3	maps	map	NOUN
easat-5050	77	4	extract	extract	VERB
easat-5050	77	5	features	feature	NOUN
easat-5050	77	6	from	from	ADP
easat-5050	77	7	different	different	ADJ
easat-5050	77	8	aspects	aspect	NOUN
easat-5050	77	9	and	and	CCONJ
easat-5050	77	10	are	be	AUX
easat-5050	77	11	then	then	ADV
easat-5050	77	12	fused	fuse	VERB
easat-5050	77	13	to	to	PART
easat-5050	77	14	obtain	obtain	VERB
easat-5050	77	15	a	a	DET
easat-5050	77	16	more	more	ADV
easat-5050	77	17	comprehensive	comprehensive	ADJ
easat-5050	77	18	feature	feature	NOUN
easat-5050	77	19	representation	representation	NOUN
easat-5050	77	20	.	.	PUNCT
easat-5050	78	1	assuming	assume	VERB
easat-5050	78	2	the	the	DET
easat-5050	78	3	input	input	NOUN
easat-5050	78	4	data	datum	NOUN
easat-5050	78	5	are	be	AUX
easat-5050	78	6	𝑿	𝑿	PROPN
easat-5050	78	7	,	,	PUNCT
easat-5050	78	8	the	the	DET
easat-5050	78	9	number	number	NOUN
easat-5050	78	10	of	of	ADP
easat-5050	78	11	convolution	convolution	NOUN
easat-5050	78	12	kernels	kernel	NOUN
easat-5050	78	13	is	be	AUX
easat-5050	78	14	𝑵	𝑵	PROPN
easat-5050	78	15	,	,	PUNCT
easat-5050	78	16	the	the	DET
easat-5050	78	17	size	size	NOUN
easat-5050	78	18	of	of	ADP
easat-5050	78	19	the	the	DET
easat-5050	78	20	convolution	convolution	NOUN
easat-5050	78	21	kernel	kernel	NOUN
easat-5050	78	22	is	be	AUX
easat-5050	78	23	𝑲	𝑲	DET
easat-5050	78	24	×	×	PROPN
easat-5050	78	25	𝑲	𝑲	NOUN
easat-5050	78	26	,	,	PUNCT
easat-5050	78	27	the	the	DET
easat-5050	78	28	number	number	NOUN
easat-5050	78	29	of	of	ADP
easat-5050	78	30	channels	channel	NOUN
easat-5050	78	31	in	in	ADP
easat-5050	78	32	the	the	DET
easat-5050	78	33	input	input	NOUN
easat-5050	78	34	data	datum	NOUN
easat-5050	78	35	is	be	AUX
easat-5050	78	36	𝑪	𝑪	PROPN
easat-5050	78	37	,	,	PUNCT
easat-5050	78	38	and	and	CCONJ
easat-5050	78	39	the	the	DET
easat-5050	78	40	number	number	NOUN
easat-5050	78	41	of	of	ADP
easat-5050	78	42	channels	channel	NOUN
easat-5050	78	43	in	in	ADP
easat-5050	78	44	the	the	DET
easat-5050	78	45	output	output	NOUN
easat-5050	78	46	feature	feature	NOUN
easat-5050	78	47	maps	map	NOUN
easat-5050	78	48	is	be	AUX
easat-5050	78	49	𝑴.	𝑴.	PROPN
easat-5050	78	50	for	for	ADP
easat-5050	78	51	each	each	DET
easat-5050	78	52	convolution	convolution	NOUN
easat-5050	78	53	kernel	kernel	NOUN
easat-5050	78	54	𝒊	𝒊	X
easat-5050	78	55	(	(	PUNCT
easat-5050	78	56	𝟏	𝟏	PROPN
easat-5050	78	57	≤	≤	NUM
easat-5050	78	58	𝒊	𝒊	X
easat-5050	78	59	≤	≤	PROPN
easat-5050	78	60	𝑵	𝑵	PROPN
easat-5050	78	61	)	)	PUNCT
easat-5050	78	62	,	,	PUNCT
easat-5050	78	63	its	its	PRON
easat-5050	78	64	weight	weight	NOUN
easat-5050	78	65	matrix	matrix	NOUN
easat-5050	78	66	is	be	AUX
easat-5050	78	67	𝑾𝒊	𝑾𝒊	PROPN
easat-5050	78	68	with	with	ADP
easat-5050	78	69	a	a	DET
easat-5050	78	70	shape	shape	NOUN
easat-5050	78	71	of	of	ADP
easat-5050	78	72	𝑲	𝑲	DET
easat-5050	78	73	×	×	NOUN
easat-5050	78	74	𝑲	𝑲	PRON
easat-5050	78	75	×	×	NOUN
easat-5050	78	76	𝑪	𝑪	NOUN
easat-5050	78	77	,	,	PUNCT
easat-5050	78	78	and	and	CCONJ
easat-5050	78	79	the	the	DET
easat-5050	78	80	bias	bias	NOUN
easat-5050	78	81	term	term	NOUN
easat-5050	78	82	is	be	AUX
easat-5050	78	83	𝒃𝒊.	𝒃𝒊.	SCONJ
easat-5050	78	84	the	the	DET
easat-5050	78	85	output	output	NOUN
easat-5050	78	86	of	of	ADP
easat-5050	78	87	the	the	DET
easat-5050	78	88	multi	multi	ADJ
easat-5050	78	89	-	-	ADJ
easat-5050	78	90	branch	branch	ADJ
easat-5050	78	91	convolution	convolution	NOUN
easat-5050	78	92	operation	operation	NOUN
easat-5050	78	93	for	for	ADP
easat-5050	78	94	the	the	DET
easat-5050	78	95	m	m	PROPN
easat-5050	78	96	-	-	PUNCT
easat-5050	78	97	th	th	VERB
easat-5050	78	98	feature	feature	NOUN
easat-5050	78	99	map	map	NOUN
easat-5050	78	100	of	of	ADP
easat-5050	78	101	the	the	DET
easat-5050	78	102	input	input	NOUN
easat-5050	78	103	data	datum	NOUN
easat-5050	78	104	𝑿	𝑿	PROPN
easat-5050	78	105	can	can	AUX
easat-5050	78	106	be	be	AUX
easat-5050	78	107	calculated	calculate	VERB
easat-5050	78	108	using	use	VERB
easat-5050	78	109	the	the	DET
easat-5050	78	110	following	follow	VERB
easat-5050	78	111	formula	formula	NOUN
easat-5050	78	112	:	:	PUNCT
easat-5050	78	113	𝑌𝑚	𝑌𝑚	PROPN
easat-5050	78	114	=	=	PUNCT
easat-5050	78	115	∑	∑	PUNCT
easat-5050	78	116	𝑛	𝑛	DET
easat-5050	78	117	𝑓(𝑐𝑜𝑛𝑣(𝑊𝑛	𝑓(𝑐𝑜𝑛𝑣(𝑊𝑛	PROPN
easat-5050	78	118	,	,	PUNCT
easat-5050	78	119	𝑋	𝑋	PROPN
easat-5050	78	120	)	)	PUNCT
easat-5050	78	121	+	+	NUM
easat-5050	78	122	𝑏𝑛	𝑏𝑛	NOUN
easat-5050	78	123	)	)	PUNCT
easat-5050	78	124	(	(	PUNCT
easat-5050	78	125	2	2	X
easat-5050	78	126	)	)	PUNCT
easat-5050	78	127	where	where	SCONJ
easat-5050	78	128	𝒄𝒐𝒏𝒗(𝑾𝒏	𝒄𝒐𝒏𝒗(𝑾𝒏	PROPN
easat-5050	78	129	,	,	PUNCT
easat-5050	78	130	𝑿	𝑿	PROPN
easat-5050	78	131	)	)	PUNCT
easat-5050	78	132	represents	represent	VERB
easat-5050	78	133	the	the	DET
easat-5050	78	134	convolution	convolution	NOUN
easat-5050	78	135	operation	operation	NOUN
easat-5050	78	136	between	between	ADP
easat-5050	78	137	the	the	DET
easat-5050	78	138	input	input	NOUN
easat-5050	78	139	data	datum	NOUN
easat-5050	78	140	𝑿	𝑿	PROPN
easat-5050	78	141	and	and	CCONJ
easat-5050	78	142	the	the	DET
easat-5050	78	143	convolutional	convolutional	ADJ
easat-5050	78	144	kernel	kernel	NOUN
easat-5050	78	145	𝑾𝒏	𝑾𝒏	NOUN
easat-5050	78	146	;	;	PUNCT
easat-5050	78	147	𝒃𝒏	𝒃𝒏	PROPN
easat-5050	78	148	represents	represent	VERB
easat-5050	78	149	the	the	DET
easat-5050	78	150	bias	bias	NOUN
easat-5050	78	151	term	term	NOUN
easat-5050	78	152	of	of	ADP
easat-5050	78	153	the	the	DET
easat-5050	78	154	𝒏-th	𝒏-th	PROPN
easat-5050	78	155	convolutional	convolutional	ADJ
easat-5050	78	156	kernel	kernel	NOUN
easat-5050	78	157	;	;	PUNCT
easat-5050	78	158	and	and	CCONJ
easat-5050	78	159	𝒇	𝒇	NOUN
easat-5050	78	160	represents	represent	VERB
easat-5050	78	161	the	the	DET
easat-5050	78	162	rectified	rectified	ADJ
easat-5050	78	163	linear	linear	ADJ
easat-5050	78	164	activation	activation	NOUN
easat-5050	78	165	function	function	NOUN
easat-5050	78	166	(	(	PUNCT
easat-5050	78	167	relu	relu	NOUN
easat-5050	78	168	)	)	PUNCT
easat-5050	78	169	function	function	NOUN
easat-5050	78	170	.	.	PUNCT
easat-5050	79	1	in	in	ADP
easat-5050	79	2	bearing	bear	VERB
easat-5050	79	3	fault	fault	NOUN
easat-5050	79	4	diagnosis	diagnosis	NOUN
easat-5050	79	5	,	,	PUNCT
easat-5050	79	6	different	different	ADJ
easat-5050	79	7	types	type	NOUN
easat-5050	79	8	of	of	ADP
easat-5050	79	9	faults	fault	NOUN
easat-5050	79	10	exhibit	exhibit	VERB
easat-5050	79	11	different	different	ADJ
easat-5050	79	12	fault	fault	NOUN
easat-5050	79	13	characteristic	characteristic	ADJ
easat-5050	79	14	frequencies	frequency	NOUN
easat-5050	79	15	,	,	PUNCT
easat-5050	79	16	which	which	PRON
easat-5050	79	17	result	result	VERB
easat-5050	79	18	in	in	ADP
easat-5050	79	19	different	different	ADJ
easat-5050	79	20	sensitivities	sensitivity	NOUN
easat-5050	79	21	of	of	ADP
easat-5050	79	22	different	different	ADJ
easat-5050	79	23	scales	scale	NOUN
easat-5050	79	24	of	of	ADP
easat-5050	79	25	features	feature	NOUN
easat-5050	79	26	to	to	ADP
easat-5050	79	27	different	different	ADJ
easat-5050	79	28	types	type	NOUN
easat-5050	79	29	of	of	ADP
easat-5050	79	30	faults	fault	NOUN
easat-5050	79	31	.	.	PUNCT
easat-5050	80	1	through	through	ADP
easat-5050	80	2	multi	multi	ADJ
easat-5050	80	3	-	-	ADJ
easat-5050	80	4	branch	branch	ADJ
easat-5050	80	5	convolutional	convolutional	ADJ
easat-5050	80	6	operations	operation	NOUN
easat-5050	80	7	,	,	PUNCT
easat-5050	80	8	the	the	DET
easat-5050	80	9	vibration	vibration	NOUN
easat-5050	80	10	features	feature	VERB
easat-5050	80	11	at	at	ADP
easat-5050	80	12	different	different	ADJ
easat-5050	80	13	scales	scale	NOUN
easat-5050	80	14	can	can	AUX
easat-5050	80	15	be	be	AUX
easat-5050	80	16	captured	capture	VERB
easat-5050	80	17	simultaneously	simultaneously	ADV
easat-5050	80	18	,	,	PUNCT
easat-5050	80	19	thereby	thereby	ADV
easat-5050	80	20	improving	improve	VERB
easat-5050	80	21	the	the	DET
easat-5050	80	22	detection	detection	NOUN
easat-5050	80	23	and	and	CCONJ
easat-5050	80	24	diagnostic	diagnostic	ADJ
easat-5050	80	25	capabilities	capability	NOUN
easat-5050	80	26	of	of	ADP
easat-5050	80	27	the	the	DET
easat-5050	80	28	model	model	NOUN
easat-5050	80	29	for	for	ADP
easat-5050	80	30	bearing	bear	VERB
easat-5050	80	31	faults	fault	NOUN
easat-5050	80	32	.	.	PUNCT
easat-5050	81	1	this	this	PRON
easat-5050	81	2	enhances	enhance	VERB
easat-5050	81	3	the	the	DET
easat-5050	81	4	adaptability	adaptability	NOUN
easat-5050	81	5	and	and	CCONJ
easat-5050	81	6	generalisation	generalisation	NOUN
easat-5050	81	7	ability	ability	NOUN
easat-5050	81	8	of	of	ADP
easat-5050	81	9	the	the	DET
easat-5050	81	10	model	model	NOUN
easat-5050	81	11	for	for	ADP
easat-5050	81	12	bearing	bear	VERB
easat-5050	81	13	faults	fault	NOUN
easat-5050	81	14	.	.	PUNCT
easat-5050	82	1	additionally	additionally	ADV
easat-5050	82	2	,	,	PUNCT
easat-5050	82	3	feature	feature	NOUN
easat-5050	82	4	fusion	fusion	NOUN
easat-5050	82	5	at	at	ADP
easat-5050	82	6	different	different	ADJ
easat-5050	82	7	scales	scale	NOUN
easat-5050	82	8	can	can	AUX
easat-5050	82	9	acquire	acquire	VERB
easat-5050	82	10	a	a	DET
easat-5050	82	11	more	more	ADV
easat-5050	82	12	comprehensive	comprehensive	ADJ
easat-5050	82	13	and	and	CCONJ
easat-5050	82	14	rich	rich	ADJ
easat-5050	82	15	feature	feature	NOUN
easat-5050	82	16	representation	representation	NOUN
easat-5050	82	17	.	.	PUNCT
easat-5050	83	1	this	this	PRON
easat-5050	83	2	aids	aid	VERB
easat-5050	83	3	the	the	DET
easat-5050	83	4	model	model	NOUN
easat-5050	83	5	in	in	ADP
easat-5050	83	6	better	well	ADJ
easat-5050	83	7	understanding	understand	VERB
easat-5050	83	8	the	the	DET
easat-5050	83	9	vibration	vibration	NOUN
easat-5050	83	10	signals	signal	NOUN
easat-5050	83	11	at	at	ADP
easat-5050	83	12	different	different	ADJ
easat-5050	83	13	scales	scale	NOUN
easat-5050	83	14	and	and	CCONJ
easat-5050	83	15	improves	improve	VERB
easat-5050	83	16	its	its	PRON
easat-5050	83	17	discriminative	discriminative	NOUN
easat-5050	83	18	power	power	NOUN
easat-5050	83	19	for	for	ADP
easat-5050	83	20	bearing	bear	VERB
easat-5050	83	21	faults	fault	NOUN
easat-5050	83	22	.	.	PUNCT
easat-5050	84	1	2.2	2.2	NUM
easat-5050	84	2	.	.	PUNCT
easat-5050	84	3	channel	channel	NOUN
easat-5050	84	4	attention	attention	NOUN
easat-5050	84	5	layer	layer	NOUN
easat-5050	84	6	the	the	DET
easat-5050	84	7	channel	channel	NOUN
easat-5050	84	8	attention	attention	NOUN
easat-5050	84	9	network	network	NOUN
easat-5050	84	10	by	by	ADP
easat-5050	84	11	bahdanau	bahdanau	PROPN
easat-5050	84	12	,	,	PUNCT
easat-5050	84	13	cho	cho	NOUN
easat-5050	84	14	,	,	PUNCT
easat-5050	84	15	and	and	CCONJ
easat-5050	84	16	bengio	bengio	PROPN
easat-5050	84	17	(	(	PUNCT
easat-5050	84	18	2014	2014	NUM
easat-5050	84	19	)	)	PUNCT
easat-5050	84	20	won	win	VERB
easat-5050	84	21	the	the	DET
easat-5050	84	22	championship	championship	NOUN
easat-5050	84	23	in	in	ADP
easat-5050	84	24	the	the	DET
easat-5050	84	25	imagenet	imagenet	NOUN
easat-5050	84	26	competition	competition	NOUN
easat-5050	84	27	for	for	ADP
easat-5050	84	28	classification	classification	NOUN
easat-5050	84	29	tasks	task	NOUN
easat-5050	84	30	in	in	ADP
easat-5050	84	31	2017	2017	NUM
easat-5050	84	32	.	.	PUNCT
easat-5050	85	1	it	it	PRON
easat-5050	85	2	can	can	AUX
easat-5050	85	3	be	be	AUX
easat-5050	85	4	used	use	VERB
easat-5050	85	5	to	to	PART
easat-5050	85	6	enhance	enhance	VERB
easat-5050	85	7	the	the	DET
easat-5050	85	8	attention	attention	NOUN
easat-5050	85	9	mechanism	mechanism	NOUN
easat-5050	85	10	of	of	ADP
easat-5050	85	11	convolutional	convolutional	ADJ
easat-5050	85	12	neural	neural	ADJ
easat-5050	85	13	networks	network	NOUN
easat-5050	85	14	.	.	PUNCT
easat-5050	86	1	it	it	PRON
easat-5050	86	2	improves	improve	VERB
easat-5050	86	3	network	network	NOUN
easat-5050	86	4	performance	performance	NOUN
easat-5050	86	5	by	by	ADP
easat-5050	86	6	adaptively	adaptively	ADV
easat-5050	86	7	adjusting	adjust	VERB
easat-5050	86	8	the	the	DET
easat-5050	86	9	importance	importance	NOUN
easat-5050	86	10	of	of	ADP
easat-5050	86	11	features	feature	NOUN
easat-5050	86	12	across	across	ADP
easat-5050	86	13	channels	channel	NOUN
easat-5050	86	14	.	.	PUNCT
easat-5050	87	1	the	the	DET
easat-5050	87	2	channel	channel	NOUN
easat-5050	87	3	attention	attention	NOUN
easat-5050	87	4	module	module	NOUN
easat-5050	87	5	consists	consist	VERB
easat-5050	87	6	of	of	ADP
easat-5050	87	7	two	two	NUM
easat-5050	87	8	key	key	ADJ
easat-5050	87	9	steps	step	NOUN
easat-5050	87	10	i.e.	i.e.	X
easat-5050	87	11	squeezing	squeezing	NOUN
easat-5050	87	12	and	and	CCONJ
easat-5050	87	13	excitation	excitation	NOUN
easat-5050	87	14	(	(	PUNCT
easat-5050	87	15	figure	figure	NOUN
easat-5050	87	16	3	3	NUM
easat-5050	87	17	)	)	PUNCT
easat-5050	87	18	.	.	PUNCT
easat-5050	88	1	2136	2136	NUM
easat-5050	88	2	edelweiss	edelweiss	PROPN
easat-5050	88	3	applied	apply	VERB
easat-5050	88	4	science	science	NOUN
easat-5050	88	5	and	and	CCONJ
easat-5050	88	6	technology	technology	NOUN
easat-5050	88	7	issn	issn	PROPN
easat-5050	88	8	:	:	PUNCT
easat-5050	88	9	2576	2576	NUM
easat-5050	88	10	-	-	SYM
easat-5050	88	11	8484	8484	NUM
easat-5050	88	12	vol	vol	NOUN
easat-5050	88	13	.	.	PROPN
easat-5050	89	1	9	9	NUM
easat-5050	89	2	,	,	PUNCT
easat-5050	89	3	no	no	INTJ
easat-5050	89	4	.	.	NOUN
easat-5050	89	5	2	2	NUM
easat-5050	89	6	:	:	SYM
easat-5050	89	7	2132	2132	NUM
easat-5050	89	8	-	-	SYM
easat-5050	89	9	2146	2146	NUM
easat-5050	89	10	,	,	PUNCT
easat-5050	89	11	2025	2025	NUM
easat-5050	89	12	doi	doi	NOUN
easat-5050	89	13	:	:	PUNCT
easat-5050	89	14	10.55214/25768484.v9i2.5050	10.55214/25768484.v9i2.5050	NUM
easat-5050	89	15	©	©	PROPN
easat-5050	89	16	2025	2025	NUM
easat-5050	89	17	by	by	ADP
easat-5050	89	18	the	the	DET
easat-5050	89	19	authors	author	NOUN
easat-5050	89	20	;	;	PUNCT
easat-5050	89	21	licensee	licensee	PROPN
easat-5050	89	22	learning	learn	VERB
easat-5050	89	23	gate	gate	PROPN
easat-5050	89	24	figure	figure	NOUN
easat-5050	89	25	3	3	NUM
easat-5050	89	26	.	.	NOUN
easat-5050	89	27	channel	channel	NOUN
easat-5050	89	28	attention	attention	NOUN
easat-5050	89	29	module	module	NOUN
easat-5050	89	30	source	source	NOUN
easat-5050	89	31	:	:	PUNCT
easat-5050	89	32	hu	hu	PROPN
easat-5050	89	33	,	,	PUNCT
easat-5050	89	34	shen	shen	NOUN
easat-5050	89	35	,	,	PUNCT
easat-5050	89	36	and	and	CCONJ
easat-5050	89	37	sun	sun	NOUN
easat-5050	89	38	(	(	PUNCT
easat-5050	89	39	2018	2018	NUM
easat-5050	89	40	)	)	PUNCT
easat-5050	89	41	.	.	PUNCT
easat-5050	90	1	squeezing	squeeze	VERB
easat-5050	90	2	involves	involve	VERB
easat-5050	90	3	global	global	ADJ
easat-5050	90	4	pooling	pooling	NOUN
easat-5050	90	5	on	on	ADP
easat-5050	90	6	the	the	DET
easat-5050	90	7	input	input	NOUN
easat-5050	90	8	feature	feature	NOUN
easat-5050	90	9	map	map	NOUN
easat-5050	90	10	along	along	ADP
easat-5050	90	11	the	the	DET
easat-5050	90	12	channel	channel	NOUN
easat-5050	90	13	dimensions	dimension	NOUN
easat-5050	90	14	to	to	PART
easat-5050	90	15	obtain	obtain	VERB
easat-5050	90	16	global	global	ADJ
easat-5050	90	17	information	information	NOUN
easat-5050	90	18	regarding	regard	VERB
easat-5050	90	19	the	the	DET
easat-5050	90	20	channels	channel	NOUN
easat-5050	90	21	.	.	PUNCT
easat-5050	91	1	assuming	assume	VERB
easat-5050	91	2	that	that	SCONJ
easat-5050	91	3	the	the	DET
easat-5050	91	4	size	size	NOUN
easat-5050	91	5	of	of	ADP
easat-5050	91	6	the	the	DET
easat-5050	91	7	input	input	NOUN
easat-5050	91	8	feature	feature	NOUN
easat-5050	91	9	map	map	NOUN
easat-5050	91	10	is	be	AUX
easat-5050	91	11	(	(	PUNCT
easat-5050	91	12	𝑯	𝑯	PROPN
easat-5050	91	13	,	,	PUNCT
easat-5050	91	14	𝑾	𝑾	ADJ
easat-5050	91	15	,	,	PUNCT
easat-5050	91	16	𝑪	𝑪	PROPN
easat-5050	91	17	)	)	PUNCT
easat-5050	91	18	,	,	PUNCT
easat-5050	91	19	the	the	DET
easat-5050	91	20	squeeze	squeeze	NOUN
easat-5050	91	21	operation	operation	NOUN
easat-5050	91	22	performs	perform	VERB
easat-5050	91	23	a	a	DET
easat-5050	91	24	global	global	ADJ
easat-5050	91	25	average	average	ADJ
easat-5050	91	26	pooling	pooling	NOUN
easat-5050	91	27	on	on	ADP
easat-5050	91	28	the	the	DET
easat-5050	91	29	input	input	NOUN
easat-5050	91	30	feature	feature	NOUN
easat-5050	91	31	map	map	NOUN
easat-5050	91	32	,	,	PUNCT
easat-5050	91	33	resulting	result	VERB
easat-5050	91	34	in	in	ADP
easat-5050	91	35	a	a	DET
easat-5050	91	36	𝑪dimensional	𝑪dimensional	ADJ
easat-5050	91	37	vector	vector	NOUN
easat-5050	91	38	that	that	PRON
easat-5050	91	39	represents	represent	VERB
easat-5050	91	40	the	the	DET
easat-5050	91	41	global	global	ADJ
easat-5050	91	42	average	average	ADJ
easat-5050	91	43	value	value	NOUN
easat-5050	91	44	for	for	ADP
easat-5050	91	45	each	each	DET
easat-5050	91	46	channel	channel	NOUN
easat-5050	91	47	.	.	PUNCT
easat-5050	92	1	this	this	DET
easat-5050	92	2	vector	vector	NOUN
easat-5050	92	3	represents	represent	VERB
easat-5050	92	4	the	the	DET
easat-5050	92	5	overall	overall	ADJ
easat-5050	92	6	importance	importance	NOUN
easat-5050	92	7	of	of	ADP
easat-5050	92	8	each	each	DET
easat-5050	92	9	channel	channel	NOUN
easat-5050	92	10	in	in	ADP
easat-5050	92	11	the	the	DET
easat-5050	92	12	input	input	NOUN
easat-5050	92	13	feature	feature	NOUN
easat-5050	92	14	map	map	NOUN
easat-5050	92	15	.	.	PUNCT
easat-5050	93	1	the	the	DET
easat-5050	93	2	resulting	result	VERB
easat-5050	93	3	output	output	NOUN
easat-5050	93	4	z	z	NOUN
easat-5050	93	5	can	can	AUX
easat-5050	93	6	be	be	AUX
easat-5050	93	7	expressed	express	VERB
easat-5050	93	8	as	as	ADP
easat-5050	93	9	𝑍	𝑍	NOUN
easat-5050	93	10	=	=	SYM
easat-5050	93	11	f𝑠𝑞(x	f𝑠𝑞(x	PROPN
easat-5050	93	12	)	)	PUNCT
easat-5050	93	13	=	=	SYM
easat-5050	93	14	1	1	NUM
easat-5050	93	15	𝐻×𝑊	𝐻×𝑊	NOUN
easat-5050	93	16	∑	∑	PUNCT
easat-5050	93	17	𝐻	𝐻	PROPN
easat-5050	93	18	𝑖=1	𝑖=1	PUNCT
easat-5050	93	19	∑	∑	PUNCT
easat-5050	93	20	𝑊	𝑊	PROPN
easat-5050	93	21	𝑗=1	𝑗=1	PROPN
easat-5050	93	22	𝑋(𝑖	𝑋(𝑖	PROPN
easat-5050	93	23	,	,	PUNCT
easat-5050	93	24	𝑗	𝑗	NOUN
easat-5050	93	25	)	)	PUNCT
easat-5050	93	26	(	(	PUNCT
easat-5050	93	27	3	3	X
easat-5050	93	28	)	)	PUNCT
easat-5050	93	29	excitation	excitation	NOUN
easat-5050	93	30	then	then	ADV
easat-5050	93	31	uses	use	VERB
easat-5050	93	32	a	a	DET
easat-5050	93	33	learned	learn	VERB
easat-5050	93	34	nonlinear	nonlinear	ADJ
easat-5050	93	35	function	function	NOUN
easat-5050	93	36	to	to	PART
easat-5050	93	37	transform	transform	VERB
easat-5050	93	38	the	the	DET
easat-5050	93	39	squeezed	squeeze	VERB
easat-5050	93	40	features	feature	NOUN
easat-5050	93	41	into	into	ADP
easat-5050	93	42	channel	channel	NOUN
easat-5050	93	43	attention	attention	NOUN
easat-5050	93	44	weights	weight	NOUN
easat-5050	93	45	,	,	PUNCT
easat-5050	93	46	which	which	PRON
easat-5050	93	47	are	be	AUX
easat-5050	93	48	used	use	VERB
easat-5050	93	49	to	to	PART
easat-5050	93	50	adjust	adjust	VERB
easat-5050	93	51	the	the	DET
easat-5050	93	52	importance	importance	NOUN
easat-5050	93	53	of	of	ADP
easat-5050	93	54	each	each	DET
easat-5050	93	55	channel	channel	NOUN
easat-5050	93	56	in	in	ADP
easat-5050	93	57	the	the	DET
easat-5050	93	58	input	input	NOUN
easat-5050	93	59	feature	feature	NOUN
easat-5050	93	60	map	map	NOUN
easat-5050	93	61	.	.	PUNCT
easat-5050	94	1	the	the	DET
easat-5050	94	2	excitation	excitation	NOUN
easat-5050	94	3	operation	operation	NOUN
easat-5050	94	4	generates	generate	VERB
easat-5050	94	5	the	the	DET
easat-5050	94	6	channel	channel	NOUN
easat-5050	94	7	attention	attention	NOUN
easat-5050	94	8	weights	weight	NOUN
easat-5050	94	9	by	by	ADP
easat-5050	94	10	learning	learn	VERB
easat-5050	94	11	a	a	DET
easat-5050	94	12	nonlinear	nonlinear	ADJ
easat-5050	94	13	mapping	mapping	NOUN
easat-5050	94	14	function	function	NOUN
easat-5050	94	15	.	.	PUNCT
easat-5050	95	1	this	this	DET
easat-5050	95	2	mapping	mapping	NOUN
easat-5050	95	3	function	function	NOUN
easat-5050	95	4	can	can	AUX
easat-5050	95	5	be	be	AUX
easat-5050	95	6	learned	learn	VERB
easat-5050	95	7	using	use	VERB
easat-5050	95	8	either	either	CCONJ
easat-5050	95	9	fully	fully	ADV
easat-5050	95	10	connected	connected	ADJ
easat-5050	95	11	or	or	CCONJ
easat-5050	95	12	one	one	NUM
easat-5050	95	13	-	-	PUNCT
easat-5050	95	14	dimensional	dimensional	ADJ
easat-5050	95	15	convolutional	convolutional	ADJ
easat-5050	95	16	layers	layer	NOUN
easat-5050	95	17	.	.	PUNCT
easat-5050	96	1	given	give	VERB
easat-5050	96	2	an	an	DET
easat-5050	96	3	input	input	NOUN
easat-5050	96	4	feature	feature	NOUN
easat-5050	96	5	vector	vector	NOUN
easat-5050	96	6	𝒁	𝒁	PROPN
easat-5050	96	7	,	,	PUNCT
easat-5050	96	8	the	the	DET
easat-5050	96	9	result	result	NOUN
easat-5050	96	10	of	of	ADP
easat-5050	96	11	the	the	DET
easat-5050	96	12	excitation	excitation	NOUN
easat-5050	96	13	operation	operation	NOUN
easat-5050	96	14	𝑺	𝑺	PROPN
easat-5050	96	15	can	can	AUX
easat-5050	96	16	be	be	AUX
easat-5050	96	17	expressed	express	VERB
easat-5050	96	18	as	as	ADP
easat-5050	96	19	𝑆	𝑆	PROPN
easat-5050	96	20	=	=	SYM
easat-5050	96	21	𝜎(𝑊2	𝜎(𝑊2	PROPN
easat-5050	96	22	∗	∗	NOUN
easat-5050	96	23	𝛿(𝑊1	𝛿(𝑊1	PROPN
easat-5050	96	24	∗	∗	NOUN
easat-5050	96	25	𝑍	𝑍	NOUN
easat-5050	96	26	)	)	PUNCT
easat-5050	96	27	)	)	PUNCT
easat-5050	96	28	(	(	PUNCT
easat-5050	96	29	4	4	X
easat-5050	96	30	)	)	PUNCT
easat-5050	96	31	𝑆	𝑆	PROPN
easat-5050	96	32	is	be	AUX
easat-5050	96	33	a	a	DET
easat-5050	96	34	𝐶-dimensional	𝐶-dimensional	ADJ
easat-5050	96	35	vector	vector	NOUN
easat-5050	96	36	representing	represent	VERB
easat-5050	96	37	the	the	DET
easat-5050	96	38	attention	attention	NOUN
easat-5050	96	39	weights	weight	NOUN
easat-5050	96	40	for	for	ADP
easat-5050	96	41	each	each	DET
easat-5050	96	42	channel	channel	NOUN
easat-5050	96	43	;	;	PUNCT
easat-5050	96	44	w1	w1	NOUN
easat-5050	96	45	and	and	CCONJ
easat-5050	96	46	w2	w2	NOUN
easat-5050	96	47	are	be	AUX
easat-5050	96	48	weight	weight	NOUN
easat-5050	96	49	matrices	matrix	NOUN
easat-5050	96	50	;	;	PUNCT
easat-5050	96	51	δ	δ	PROPN
easat-5050	96	52	represents	represent	VERB
easat-5050	96	53	the	the	DET
easat-5050	96	54	relu	relu	NOUN
easat-5050	96	55	activation	activation	NOUN
easat-5050	96	56	function	function	NOUN
easat-5050	96	57	;	;	PUNCT
easat-5050	96	58	and	and	CCONJ
easat-5050	96	59	σ	σ	PROPN
easat-5050	96	60	is	be	AUX
easat-5050	96	61	the	the	DET
easat-5050	96	62	sigmoid	sigmoid	NOUN
easat-5050	96	63	function	function	NOUN
easat-5050	96	64	.	.	PUNCT
easat-5050	97	1	scale	scale	NOUN
easat-5050	97	2	:	:	PUNCT
easat-5050	97	3	by	by	ADP
easat-5050	97	4	performing	perform	VERB
easat-5050	97	5	an	an	DET
easat-5050	97	6	element	element	ADJ
easat-5050	97	7	-	-	PUNCT
easat-5050	97	8	wise	wise	ADJ
easat-5050	97	9	multiplication	multiplication	NOUN
easat-5050	97	10	between	between	ADP
easat-5050	97	11	𝑆	𝑆	PROPN
easat-5050	97	12	and	and	CCONJ
easat-5050	97	13	𝑋	𝑋	PROPN
easat-5050	97	14	,	,	PUNCT
easat-5050	97	15	the	the	DET
easat-5050	97	16	channels	channel	NOUN
easat-5050	97	17	of	of	ADP
easat-5050	97	18	𝑋	𝑋	PROPN
easat-5050	97	19	can	can	AUX
easat-5050	97	20	be	be	AUX
easat-5050	97	21	reweighted	reweighte	VERB
easat-5050	97	22	.	.	PUNCT
easat-5050	98	1	therefore	therefore	ADV
easat-5050	98	2	,	,	PUNCT
easat-5050	98	3	the	the	DET
easat-5050	98	4	value	value	NOUN
easat-5050	98	5	of	of	ADP
easat-5050	98	6	the	the	DET
easat-5050	98	7	weighted	weight	VERB
easat-5050	98	8	feature	feature	NOUN
easat-5050	98	9	map	map	NOUN
easat-5050	98	10	𝑌	𝑌	PROPN
easat-5050	98	11	is	be	AUX
easat-5050	98	12	𝑌	𝑌	PROPN
easat-5050	98	13	=	=	SYM
easat-5050	98	14	𝐹𝑠𝑐𝑎𝑙𝑒(𝑆	𝐹𝑠𝑐𝑎𝑙𝑒(𝑆	PROPN
easat-5050	98	15	,	,	PUNCT
easat-5050	98	16	𝑋	𝑋	NOUN
easat-5050	98	17	)	)	PUNCT
easat-5050	98	18	=	=	SYM
easat-5050	98	19	𝑆	𝑆	PROPN
easat-5050	98	20	∗	∗	NOUN
easat-5050	98	21	𝑋	𝑋	PROPN
easat-5050	98	22	(	(	PUNCT
easat-5050	98	23	5	5	NUM
easat-5050	98	24	)	)	PUNCT
easat-5050	98	25	2137	2137	NUM
easat-5050	98	26	edelweiss	edelweiss	PROPN
easat-5050	98	27	applied	apply	VERB
easat-5050	98	28	science	science	NOUN
easat-5050	98	29	and	and	CCONJ
easat-5050	98	30	technology	technology	NOUN
easat-5050	98	31	issn	issn	PROPN
easat-5050	98	32	:	:	PUNCT
easat-5050	98	33	2576	2576	NUM
easat-5050	98	34	-	-	SYM
easat-5050	98	35	8484	8484	NUM
easat-5050	98	36	vol	vol	NOUN
easat-5050	98	37	.	.	PROPN
easat-5050	99	1	9	9	NUM
easat-5050	99	2	,	,	PUNCT
easat-5050	99	3	no	no	INTJ
easat-5050	99	4	.	.	NOUN
easat-5050	99	5	2	2	NUM
easat-5050	99	6	:	:	SYM
easat-5050	99	7	2132	2132	NUM
easat-5050	99	8	-	-	SYM
easat-5050	99	9	2146	2146	NUM
easat-5050	99	10	,	,	PUNCT
easat-5050	99	11	2025	2025	NUM
easat-5050	99	12	doi	doi	NOUN
easat-5050	99	13	:	:	PUNCT
easat-5050	99	14	10.55214/25768484.v9i2.5050	10.55214/25768484.v9i2.5050	NUM
easat-5050	99	15	©	©	PROPN
easat-5050	99	16	2025	2025	NUM
easat-5050	99	17	by	by	ADP
easat-5050	99	18	the	the	DET
easat-5050	99	19	authors	author	NOUN
easat-5050	99	20	;	;	PUNCT
easat-5050	99	21	licensee	licensee	PROPN
easat-5050	99	22	learning	learning	NOUN
easat-5050	99	23	gate	gate	VERB
easat-5050	99	24	the	the	DET
easat-5050	99	25	channel	channel	NOUN
easat-5050	99	26	attention	attention	NOUN
easat-5050	99	27	layer	layer	NOUN
easat-5050	99	28	can	can	AUX
easat-5050	99	29	adaptively	adaptively	ADV
easat-5050	99	30	learn	learn	VERB
easat-5050	99	31	the	the	DET
easat-5050	99	32	correlations	correlation	NOUN
easat-5050	99	33	between	between	ADP
easat-5050	99	34	feature	feature	NOUN
easat-5050	99	35	channels	channel	NOUN
easat-5050	99	36	and	and	CCONJ
easat-5050	99	37	enhance	enhance	VERB
easat-5050	99	38	the	the	DET
easat-5050	99	39	important	important	ADJ
easat-5050	99	40	features	feature	NOUN
easat-5050	99	41	in	in	ADP
easat-5050	99	42	bearing	bear	VERB
easat-5050	99	43	vibration	vibration	NOUN
easat-5050	99	44	signals	signal	NOUN
easat-5050	99	45	.	.	PUNCT
easat-5050	100	1	in	in	ADP
easat-5050	100	2	bearing	bear	VERB
easat-5050	100	3	fault	fault	NOUN
easat-5050	100	4	diagnosis	diagnosis	NOUN
easat-5050	100	5	,	,	PUNCT
easat-5050	100	6	samples	sample	NOUN
easat-5050	100	7	of	of	ADP
easat-5050	100	8	faulty	faulty	ADJ
easat-5050	100	9	cases	case	NOUN
easat-5050	100	10	are	be	AUX
easat-5050	100	11	often	often	ADV
easat-5050	100	12	limited	limit	VERB
easat-5050	100	13	,	,	PUNCT
easat-5050	100	14	which	which	PRON
easat-5050	100	15	means	mean	VERB
easat-5050	100	16	that	that	SCONJ
easat-5050	100	17	the	the	DET
easat-5050	100	18	model	model	NOUN
easat-5050	100	19	must	must	AUX
easat-5050	100	20	learn	learn	VERB
easat-5050	100	21	from	from	ADP
easat-5050	100	22	limited	limited	ADJ
easat-5050	100	23	data	datum	NOUN
easat-5050	100	24	and	and	CCONJ
easat-5050	100	25	generalise	generalise	VERB
easat-5050	100	26	to	to	ADP
easat-5050	100	27	unseen	unseen	ADJ
easat-5050	100	28	fault	fault	NOUN
easat-5050	100	29	scenarios	scenario	NOUN
easat-5050	100	30	.	.	PUNCT
easat-5050	101	1	the	the	DET
easat-5050	101	2	channel	channel	NOUN
easat-5050	101	3	attention	attention	NOUN
easat-5050	101	4	module	module	NOUN
easat-5050	101	5	can	can	AUX
easat-5050	101	6	enhance	enhance	VERB
easat-5050	101	7	the	the	DET
easat-5050	101	8	sensitivity	sensitivity	NOUN
easat-5050	101	9	and	and	CCONJ
easat-5050	101	10	discriminability	discriminability	NOUN
easat-5050	101	11	of	of	ADP
easat-5050	101	12	the	the	DET
easat-5050	101	13	model	model	NOUN
easat-5050	101	14	for	for	ADP
easat-5050	101	15	bearing	bear	VERB
easat-5050	101	16	fault	fault	NOUN
easat-5050	101	17	samples	sample	NOUN
easat-5050	101	18	.	.	PUNCT
easat-5050	102	1	by	by	ADP
easat-5050	102	2	adaptively	adaptively	ADV
easat-5050	102	3	adjusting	adjust	VERB
easat-5050	102	4	the	the	DET
easat-5050	102	5	weights	weight	NOUN
easat-5050	102	6	of	of	ADP
easat-5050	102	7	the	the	DET
easat-5050	102	8	feature	feature	NOUN
easat-5050	102	9	channels	channel	NOUN
easat-5050	102	10	,	,	PUNCT
easat-5050	102	11	the	the	DET
easat-5050	102	12	model	model	NOUN
easat-5050	102	13	can	can	AUX
easat-5050	102	14	focus	focus	VERB
easat-5050	102	15	more	more	ADV
easat-5050	102	16	on	on	ADP
easat-5050	102	17	features	feature	NOUN
easat-5050	102	18	relevant	relevant	ADJ
easat-5050	102	19	to	to	ADP
easat-5050	102	20	bearing	bear	VERB
easat-5050	102	21	faults	fault	NOUN
easat-5050	102	22	,	,	PUNCT
easat-5050	102	23	enabling	enable	VERB
easat-5050	102	24	a	a	DET
easat-5050	102	25	more	more	ADV
easat-5050	102	26	accurate	accurate	ADJ
easat-5050	102	27	identification	identification	NOUN
easat-5050	102	28	and	and	CCONJ
easat-5050	102	29	classification	classification	NOUN
easat-5050	102	30	of	of	ADP
easat-5050	102	31	faulty	faulty	ADJ
easat-5050	102	32	samples	sample	NOUN
easat-5050	102	33	.	.	PUNCT
easat-5050	103	1	2.3	2.3	NUM
easat-5050	103	2	.	.	PUNCT
easat-5050	104	1	feature	feature	NOUN
easat-5050	104	2	enhancement	enhancement	NOUN
easat-5050	104	3	layer	layer	NOUN
easat-5050	104	4	traditional	traditional	ADJ
easat-5050	104	5	neural	neural	ADJ
easat-5050	104	6	network	network	NOUN
easat-5050	104	7	models	model	NOUN
easat-5050	104	8	commonly	commonly	ADV
easat-5050	104	9	use	use	VERB
easat-5050	104	10	activation	activation	NOUN
easat-5050	104	11	functions	function	NOUN
easat-5050	104	12	such	such	ADJ
easat-5050	104	13	as	as	ADP
easat-5050	104	14	relu	relu	NOUN
easat-5050	104	15	and	and	CCONJ
easat-5050	104	16	sigmoid	sigmoid	NOUN
easat-5050	104	17	to	to	PART
easat-5050	104	18	capture	capture	VERB
easat-5050	104	19	non	non	ADJ
easat-5050	104	20	-	-	ADJ
easat-5050	104	21	linear	linear	ADJ
easat-5050	104	22	relationships	relationship	NOUN
easat-5050	104	23	.	.	PUNCT
easat-5050	105	1	however	however	ADV
easat-5050	105	2	,	,	PUNCT
easat-5050	105	3	these	these	DET
easat-5050	105	4	functions	function	NOUN
easat-5050	105	5	can	can	AUX
easat-5050	105	6	only	only	ADV
easat-5050	105	7	capture	capture	VERB
easat-5050	105	8	first	first	ADJ
easat-5050	105	9	-	-	PUNCT
easat-5050	105	10	order	order	NOUN
easat-5050	105	11	features	feature	NOUN
easat-5050	105	12	.	.	PUNCT
easat-5050	106	1	fan	fan	NOUN
easat-5050	106	2	et	et	PROPN
easat-5050	106	3	al	al	PROPN
easat-5050	106	4	.	.	PROPN
easat-5050	106	5	(	(	PUNCT
easat-5050	106	6	2020	2020	NUM
easat-5050	106	7	)	)	PUNCT
easat-5050	106	8	proposed	propose	VERB
easat-5050	106	9	a	a	DET
easat-5050	106	10	novel	novel	ADJ
easat-5050	106	11	neural	neural	ADJ
easat-5050	106	12	network	network	NOUN
easat-5050	106	13	that	that	PRON
easat-5050	106	14	used	use	VERB
easat-5050	106	15	quadratic	quadratic	ADJ
easat-5050	106	16	neurones	neurone	NOUN
easat-5050	106	17	to	to	PART
easat-5050	106	18	enhance	enhance	VERB
easat-5050	106	19	the	the	DET
easat-5050	106	20	expressive	expressive	ADJ
easat-5050	106	21	power	power	NOUN
easat-5050	106	22	of	of	ADP
easat-5050	106	23	a	a	DET
easat-5050	106	24	model	model	NOUN
easat-5050	106	25	and	and	CCONJ
easat-5050	106	26	capture	capture	VERB
easat-5050	106	27	higher	high	ADJ
easat-5050	106	28	-	-	PUNCT
easat-5050	106	29	order	order	NOUN
easat-5050	106	30	nonlinear	nonlinear	ADJ
easat-5050	106	31	features	feature	NOUN
easat-5050	106	32	.	.	PUNCT
easat-5050	107	1	this	this	PRON
easat-5050	107	2	promotes	promote	VERB
easat-5050	107	3	neuronal	neuronal	ADJ
easat-5050	107	4	diversity	diversity	NOUN
easat-5050	107	5	in	in	ADP
easat-5050	107	6	deep	deep	ADJ
easat-5050	107	7	learning	learning	NOUN
easat-5050	107	8	.	.	PUNCT
easat-5050	108	1	the	the	DET
easat-5050	108	2	distinctive	distinctive	ADJ
easat-5050	108	3	feature	feature	NOUN
easat-5050	108	4	of	of	ADP
easat-5050	108	5	this	this	DET
easat-5050	108	6	approach	approach	NOUN
easat-5050	108	7	is	be	AUX
easat-5050	108	8	the	the	DET
easat-5050	108	9	integration	integration	NOUN
easat-5050	108	10	of	of	ADP
easat-5050	108	11	the	the	DET
easat-5050	108	12	two	two	NUM
easat-5050	108	13	inner	inner	ADJ
easat-5050	108	14	products	product	NOUN
easat-5050	108	15	of	of	ADP
easat-5050	108	16	the	the	DET
easat-5050	108	17	input	input	NOUN
easat-5050	108	18	vector	vector	NOUN
easat-5050	108	19	and	and	CCONJ
easat-5050	108	20	the	the	DET
easat-5050	108	21	power	power	NOUN
easat-5050	108	22	term	term	NOUN
easat-5050	108	23	before	before	ADP
easat-5050	108	24	the	the	DET
easat-5050	108	25	activation	activation	NOUN
easat-5050	108	26	function	function	NOUN
easat-5050	108	27	.	.	PUNCT
easat-5050	109	1	if	if	SCONJ
easat-5050	109	2	the	the	DET
easat-5050	109	3	input	input	NOUN
easat-5050	109	4	vector	vector	NOUN
easat-5050	109	5	is	be	AUX
easat-5050	109	6	x	x	X
easat-5050	109	7	,	,	PUNCT
easat-5050	109	8	the	the	DET
easat-5050	109	9	output	output	NOUN
easat-5050	109	10	of	of	ADP
easat-5050	109	11	the	the	DET
easat-5050	109	12	quadratic	quadratic	ADJ
easat-5050	109	13	neuron	neuron	NOUN
easat-5050	109	14	,	,	PUNCT
easat-5050	109	15	denoted	denote	VERB
easat-5050	109	16	as	as	ADP
easat-5050	109	17	𝒇(𝒙	𝒇(𝒙	NOUN
easat-5050	109	18	)	)	PUNCT
easat-5050	109	19	,	,	PUNCT
easat-5050	109	20	is	be	AUX
easat-5050	109	21	𝜎(𝑓(𝑥	𝜎(𝑓(𝑥	NOUN
easat-5050	109	22	)	)	PUNCT
easat-5050	109	23	)	)	PUNCT
easat-5050	110	1	=	=	SYM
easat-5050	110	2	𝜎	𝜎	PROPN
easat-5050	110	3	(	(	PUNCT
easat-5050	110	4	(	(	PUNCT
easat-5050	110	5	∑	∑	PART
easat-5050	110	6	𝑛	𝑛	PRON
easat-5050	110	7	𝑖=1	𝑖=1	PROPN
easat-5050	110	8	𝑤𝑖	𝑤𝑖	PRON
easat-5050	110	9	𝑟𝑥𝑖	𝑟𝑥𝑖	NOUN
easat-5050	110	10	+	+	CCONJ
easat-5050	110	11	𝑏𝑟	𝑏𝑟	NOUN
easat-5050	110	12	)	)	PUNCT
easat-5050	110	13	(	(	PUNCT
easat-5050	110	14	∑	∑	PUNCT
easat-5050	110	15	𝑛	𝑛	PRON
easat-5050	110	16	𝑖=1	𝑖=1	PROPN
easat-5050	110	17	𝑤𝑖	𝑤𝑖	ADP
easat-5050	110	18	𝑔	𝑔	PROPN
easat-5050	110	19	𝑥𝑖	𝑥𝑖	PROPN
easat-5050	110	20	+	+	CCONJ
easat-5050	110	21	𝑏𝑔	𝑏𝑔	X
easat-5050	110	22	)	)	PUNCT
easat-5050	111	1	+	+	CCONJ
easat-5050	111	2	∑	∑	PROPN
easat-5050	111	3	𝑛	𝑛	PRON
easat-5050	111	4	𝑖=1	𝑖=1	PROPN
easat-5050	111	5	𝑤𝑖	𝑤𝑖	PRON
easat-5050	111	6	𝑏𝑥𝑖	𝑏𝑥𝑖	PROPN
easat-5050	111	7	2	2	NUM
easat-5050	111	8	+	+	CCONJ
easat-5050	111	9	𝑐	𝑐	X
easat-5050	111	10	)	)	PUNCT
easat-5050	111	11	=	=	PUNCT
easat-5050	111	12	𝝈((𝒙⊤𝒘𝒓	𝝈((𝒙⊤𝒘𝒓	X
easat-5050	112	1	+	+	CCONJ
easat-5050	112	2	𝒃𝒓)(𝒙⊤𝒘𝒈	𝒃𝒓)(𝒙⊤𝒘𝒈	PROPN
easat-5050	112	3	+	+	X
easat-5050	112	4	𝒃𝒈	𝒃𝒈	X
easat-5050	112	5	)	)	PUNCT
easat-5050	112	6	+	+	CCONJ
easat-5050	112	7	(	(	PUNCT
easat-5050	112	8	𝒙	𝒙	NUM
easat-5050	112	9	⊙	⊙	NOUN
easat-5050	112	10	𝒙)⊤𝒘𝒃	𝒙)⊤𝒘𝒃	PROPN
easat-5050	112	11	+	+	PROPN
easat-5050	112	12	𝒄	𝒄	X
easat-5050	112	13	)	)	PUNCT
easat-5050	112	14	(	(	PUNCT
easat-5050	112	15	6	6	X
easat-5050	112	16	)	)	PUNCT
easat-5050	112	17	σ	σ	NOUN
easat-5050	112	18	represents	represent	VERB
easat-5050	112	19	the	the	DET
easat-5050	112	20	activation	activation	NOUN
easat-5050	112	21	function	function	NOUN
easat-5050	112	22	;	;	PUNCT
easat-5050	112	23	⊙	⊙	PROPN
easat-5050	112	24	denotes	denote	VERB
easat-5050	112	25	the	the	DET
easat-5050	112	26	hadamard	hadamard	ADJ
easat-5050	112	27	product	product	NOUN
easat-5050	112	28	;	;	PUNCT
easat-5050	112	29	𝒘𝒓	𝒘𝒓	NOUN
easat-5050	112	30	,	,	PUNCT
easat-5050	112	31	𝒘𝒈	𝒘𝒈	NOUN
easat-5050	112	32	,	,	PUNCT
easat-5050	112	33	and	and	CCONJ
easat-5050	112	34	𝒘𝒃	𝒘𝒃	PROPN
easat-5050	112	35	are	be	AUX
easat-5050	112	36	weight	weight	NOUN
easat-5050	112	37	matrices	matrix	NOUN
easat-5050	112	38	;	;	PUNCT
easat-5050	112	39	𝒃𝒓	𝒃𝒓	ADV
easat-5050	112	40	,	,	PUNCT
easat-5050	112	41	𝒃𝒈	𝒃𝒈	ADP
easat-5050	112	42	,	,	PUNCT
easat-5050	112	43	and	and	CCONJ
easat-5050	112	44	𝒄	𝒄	PROPN
easat-5050	112	45	are	be	AUX
easat-5050	112	46	bias	bias	NOUN
easat-5050	112	47	vectors	vector	NOUN
easat-5050	112	48	.	.	PUNCT
easat-5050	113	1	quadratic	quadratic	ADJ
easat-5050	113	2	neural	neural	ADJ
easat-5050	113	3	networks	network	NOUN
easat-5050	113	4	have	have	VERB
easat-5050	113	5	higher	high	ADJ
easat-5050	113	6	feature	feature	NOUN
easat-5050	113	7	-	-	PUNCT
easat-5050	113	8	extraction	extraction	NOUN
easat-5050	113	9	capabilities	capability	NOUN
easat-5050	113	10	fan	fan	NOUN
easat-5050	113	11	et	et	PROPN
easat-5050	113	12	al	al	PROPN
easat-5050	113	13	.	.	PROPN
easat-5050	114	1	(	(	PUNCT
easat-5050	114	2	2020	2020	NUM
easat-5050	114	3	)	)	PUNCT
easat-5050	114	4	.	.	PUNCT
easat-5050	115	1	j.-x	j.-x	NOUN
easat-5050	115	2	.	.	PUNCT
easat-5050	116	1	liao	liao	PROPN
easat-5050	116	2	et	et	PROPN
easat-5050	116	3	al	al	PROPN
easat-5050	116	4	.	.	PROPN
easat-5050	117	1	(	(	PUNCT
easat-5050	117	2	2023	2023	NUM
easat-5050	117	3	)	)	PUNCT
easat-5050	117	4	explained	explain	VERB
easat-5050	117	5	that	that	SCONJ
easat-5050	117	6	the	the	DET
easat-5050	117	7	mathematical	mathematical	ADJ
easat-5050	117	8	formulation	formulation	NOUN
easat-5050	117	9	of	of	ADP
easat-5050	117	10	quadratic	quadratic	ADJ
easat-5050	117	11	neurones	neurone	NOUN
easat-5050	117	12	can	can	AUX
easat-5050	117	13	derive	derive	VERB
easat-5050	117	14	a	a	DET
easat-5050	117	15	module	module	NOUN
easat-5050	117	16	similar	similar	ADJ
easat-5050	117	17	to	to	ADP
easat-5050	117	18	an	an	DET
easat-5050	117	19	attention	attention	NOUN
easat-5050	117	20	mechanism	mechanism	NOUN
easat-5050	117	21	,	,	PUNCT
easat-5050	117	22	which	which	PRON
easat-5050	117	23	aids	aid	VERB
easat-5050	117	24	the	the	DET
easat-5050	117	25	network	network	NOUN
easat-5050	117	26	in	in	ADP
easat-5050	117	27	focusing	focus	VERB
easat-5050	117	28	more	more	ADV
easat-5050	117	29	on	on	ADP
easat-5050	117	30	local	local	ADJ
easat-5050	117	31	information	information	NOUN
easat-5050	117	32	.	.	PUNCT
easat-5050	118	1	bearing	bear	VERB
easat-5050	118	2	fault	fault	NOUN
easat-5050	118	3	signals	signal	NOUN
easat-5050	118	4	,	,	PUNCT
easat-5050	118	5	which	which	PRON
easat-5050	118	6	are	be	AUX
easat-5050	118	7	generally	generally	ADV
easat-5050	118	8	high	high	ADJ
easat-5050	118	9	frequency	frequency	NOUN
easat-5050	118	10	,	,	PUNCT
easat-5050	118	11	short	short	ADJ
easat-5050	118	12	range	range	NOUN
easat-5050	118	13	,	,	PUNCT
easat-5050	118	14	and	and	CCONJ
easat-5050	118	15	high	high	ADJ
easat-5050	118	16	amplitude	amplitude	NOUN
easat-5050	118	17	,	,	PUNCT
easat-5050	118	18	belong	belong	VERB
easat-5050	118	19	to	to	ADP
easat-5050	118	20	local	local	ADJ
easat-5050	118	21	feature	feature	NOUN
easat-5050	118	22	information	information	NOUN
easat-5050	118	23	.	.	PUNCT
easat-5050	119	1	therefore	therefore	ADV
easat-5050	119	2	,	,	PUNCT
easat-5050	119	3	in	in	ADP
easat-5050	119	4	bearing	bear	VERB
easat-5050	119	5	fault	fault	NOUN
easat-5050	119	6	diagnosis	diagnosis	NOUN
easat-5050	119	7	tasks	task	NOUN
easat-5050	119	8	,	,	PUNCT
easat-5050	119	9	particularly	particularly	ADV
easat-5050	119	10	in	in	ADP
easat-5050	119	11	complex	complex	ADJ
easat-5050	119	12	noisy	noisy	ADJ
easat-5050	119	13	environments	environment	NOUN
easat-5050	119	14	,	,	PUNCT
easat-5050	119	15	we	we	PRON
easat-5050	119	16	choose	choose	VERB
easat-5050	119	17	to	to	PART
easat-5050	119	18	replace	replace	VERB
easat-5050	119	19	traditional	traditional	ADJ
easat-5050	119	20	linear	linear	ADJ
easat-5050	119	21	convolution	convolution	NOUN
easat-5050	119	22	operations	operation	NOUN
easat-5050	119	23	with	with	ADP
easat-5050	119	24	quadratic	quadratic	ADJ
easat-5050	119	25	neural	neural	ADJ
easat-5050	119	26	network	network	NOUN
easat-5050	119	27	convolutions	convolution	NOUN
easat-5050	119	28	.	.	PUNCT
easat-5050	120	1	2.4	2.4	NUM
easat-5050	120	2	.	.	PUNCT
easat-5050	120	3	anomaly	anomaly	NOUN
easat-5050	120	4	classification	classification	NOUN
easat-5050	120	5	layer	layer	NOUN
easat-5050	120	6	after	after	ADP
easat-5050	120	7	a	a	DET
easat-5050	120	8	series	series	NOUN
easat-5050	120	9	of	of	ADP
easat-5050	120	10	feature	feature	NOUN
easat-5050	120	11	extractions	extraction	NOUN
easat-5050	120	12	,	,	PUNCT
easat-5050	120	13	the	the	DET
easat-5050	120	14	vibration	vibration	NOUN
easat-5050	120	15	signal	signal	NOUN
easat-5050	120	16	of	of	ADP
easat-5050	120	17	the	the	DET
easat-5050	120	18	bearing	bearing	NOUN
easat-5050	120	19	is	be	AUX
easat-5050	120	20	mapped	map	VERB
easat-5050	120	21	to	to	ADP
easat-5050	120	22	the	the	DET
easat-5050	120	23	sample	sample	NOUN
easat-5050	120	24	label	label	NOUN
easat-5050	120	25	space	space	NOUN
easat-5050	120	26	using	use	VERB
easat-5050	120	27	an	an	DET
easat-5050	120	28	anomaly	anomaly	NOUN
easat-5050	120	29	classification	classification	NOUN
easat-5050	120	30	layer	layer	NOUN
easat-5050	120	31	to	to	PART
easat-5050	120	32	achieve	achieve	VERB
easat-5050	120	33	fault	fault	NOUN
easat-5050	120	34	classification	classification	NOUN
easat-5050	120	35	.	.	PUNCT
easat-5050	121	1	the	the	DET
easat-5050	121	2	mcfcnn	mcfcnn	PROPN
easat-5050	121	3	model	model	NOUN
easat-5050	121	4	employs	employ	VERB
easat-5050	121	5	two	two	NUM
easat-5050	121	6	linear	linear	ADJ
easat-5050	121	7	layers	layer	NOUN
easat-5050	121	8	to	to	PART
easat-5050	121	9	reduce	reduce	VERB
easat-5050	121	10	the	the	DET
easat-5050	121	11	dimensionality	dimensionality	NOUN
easat-5050	121	12	of	of	ADP
easat-5050	121	13	the	the	DET
easat-5050	121	14	features	feature	NOUN
easat-5050	121	15	to	to	ADP
easat-5050	121	16	10	10	NUM
easat-5050	121	17	and	and	CCONJ
easat-5050	121	18	then	then	ADV
easat-5050	121	19	performs	perform	VERB
easat-5050	121	20	fault	fault	NOUN
easat-5050	121	21	feature	feature	NOUN
easat-5050	121	22	classification	classification	NOUN
easat-5050	121	23	using	use	VERB
easat-5050	121	24	the	the	DET
easat-5050	121	25	𝑺𝒐𝒇𝒕𝑴𝒂𝒙	𝑺𝒐𝒇𝒕𝑴𝒂𝒙	PROPN
easat-5050	121	26	function	function	NOUN
easat-5050	121	27	.	.	PUNCT
easat-5050	122	1	specifically	specifically	ADV
easat-5050	122	2	,	,	PUNCT
easat-5050	122	3	assuming	assume	VERB
easat-5050	122	4	that	that	SCONJ
easat-5050	122	5	the	the	DET
easat-5050	122	6	input	input	NOUN
easat-5050	122	7	vector	vector	NOUN
easat-5050	122	8	is	be	AUX
easat-5050	122	9	𝒙	𝒙	NUM
easat-5050	122	10	,	,	PUNCT
easat-5050	122	11	the	the	DET
easat-5050	122	12	linear	linear	ADJ
easat-5050	122	13	layer	layer	NOUN
easat-5050	122	14	𝒉(𝒙	𝒉(𝒙	PROPN
easat-5050	122	15	)	)	PUNCT
easat-5050	122	16	can	can	AUX
easat-5050	122	17	be	be	AUX
easat-5050	122	18	expressed	express	VERB
easat-5050	122	19	as	as	ADP
easat-5050	122	20	ℎ(𝑥	ℎ(𝑥	NUM
easat-5050	122	21	)	)	PUNCT
easat-5050	122	22	=	=	SYM
easat-5050	122	23	𝑓(𝑤𝑥	𝑓(𝑤𝑥	PROPN
easat-5050	122	24	+	+	CCONJ
easat-5050	122	25	𝑏	𝑏	NOUN
easat-5050	122	26	)	)	PUNCT
easat-5050	122	27	(	(	PUNCT
easat-5050	122	28	7	7	X
easat-5050	122	29	)	)	PUNCT
easat-5050	122	30	the	the	DET
easat-5050	122	31	output	output	NOUN
easat-5050	122	32	of	of	ADP
easat-5050	122	33	the	the	DET
easat-5050	122	34	anomaly	anomaly	NOUN
easat-5050	122	35	classification	classification	NOUN
easat-5050	122	36	layer	layer	NOUN
easat-5050	122	37	𝑌(𝑥	𝑌(𝑥	NOUN
easat-5050	122	38	)	)	PUNCT
easat-5050	122	39	of	of	ADP
easat-5050	122	40	the	the	DET
easat-5050	122	41	mcfcnn	mcfcnn	NOUN
easat-5050	122	42	can	can	AUX
easat-5050	122	43	be	be	AUX
easat-5050	122	44	expressed	express	VERB
easat-5050	122	45	as	as	ADP
easat-5050	122	46	𝑌(𝑥	𝑌(𝑥	X
easat-5050	122	47	)	)	PUNCT
easat-5050	122	48	=	=	SYM
easat-5050	122	49	𝑆𝑜𝑓𝑡𝑀𝑎𝑥(𝑓(ℎ(𝑥)𝑤1	𝑆𝑜𝑓𝑡𝑀𝑎𝑥(𝑓(ℎ(𝑥)𝑤1	PUNCT
easat-5050	122	50	+	+	PUNCT
easat-5050	122	51	𝑏1	𝑏1	ADJ
easat-5050	122	52	)	)	PUNCT
easat-5050	122	53	)	)	PUNCT
easat-5050	122	54	(	(	PUNCT
easat-5050	122	55	8)	8)	NUM
easat-5050	122	56	2138	2138	NUM
easat-5050	122	57	edelweiss	edelweiss	PROPN
easat-5050	122	58	applied	apply	VERB
easat-5050	122	59	science	science	NOUN
easat-5050	122	60	and	and	CCONJ
easat-5050	122	61	technology	technology	NOUN
easat-5050	122	62	issn	issn	PROPN
easat-5050	122	63	:	:	PUNCT
easat-5050	122	64	2576	2576	NUM
easat-5050	122	65	-	-	SYM
easat-5050	122	66	8484	8484	NUM
easat-5050	122	67	vol	vol	NOUN
easat-5050	122	68	.	.	PROPN
easat-5050	123	1	9	9	NUM
easat-5050	123	2	,	,	PUNCT
easat-5050	123	3	no	no	INTJ
easat-5050	123	4	.	.	NOUN
easat-5050	123	5	2	2	NUM
easat-5050	123	6	:	:	SYM
easat-5050	123	7	2132	2132	NUM
easat-5050	123	8	-	-	SYM
easat-5050	123	9	2146	2146	NUM
easat-5050	123	10	,	,	PUNCT
easat-5050	123	11	2025	2025	NUM
easat-5050	123	12	doi	doi	NOUN
easat-5050	123	13	:	:	PUNCT
easat-5050	123	14	10.55214/25768484.v9i2.5050	10.55214/25768484.v9i2.5050	NUM
easat-5050	123	15	©	©	PROPN
easat-5050	123	16	2025	2025	NUM
easat-5050	123	17	by	by	ADP
easat-5050	123	18	the	the	DET
easat-5050	123	19	authors	author	NOUN
easat-5050	123	20	;	;	PUNCT
easat-5050	123	21	licensee	licensee	PROPN
easat-5050	123	22	learning	learning	NOUN
easat-5050	123	23	gate	gate	NOUN
easat-5050	123	24	where	where	SCONJ
easat-5050	123	25	𝒘	𝒘	NOUN
easat-5050	123	26	and	and	CCONJ
easat-5050	123	27	𝒘𝟏	𝒘𝟏	ADJ
easat-5050	123	28	are	be	AUX
easat-5050	123	29	weight	weight	NOUN
easat-5050	123	30	matrices	matrix	NOUN
easat-5050	123	31	,	,	PUNCT
easat-5050	123	32	𝒃	𝒃	NOUN
easat-5050	123	33	and	and	CCONJ
easat-5050	123	34	𝒃𝟏	𝒃𝟏	VERB
easat-5050	123	35	in	in	ADP
easat-5050	123	36	equations	equation	NOUN
easat-5050	123	37	(	(	PUNCT
easat-5050	123	38	7	7	NUM
easat-5050	123	39	)	)	PUNCT
easat-5050	123	40	and	and	CCONJ
easat-5050	123	41	(	(	PUNCT
easat-5050	123	42	8)	8)	NUM
easat-5050	123	43	are	be	AUX
easat-5050	123	44	bias	bias	NOUN
easat-5050	123	45	vectors	vector	NOUN
easat-5050	123	46	;	;	PUNCT
easat-5050	123	47	and	and	CCONJ
easat-5050	123	48	h(x	h(x	PROPN
easat-5050	123	49	)	)	PUNCT
easat-5050	123	50	represents	represent	VERB
easat-5050	123	51	the	the	DET
easat-5050	123	52	output	output	NOUN
easat-5050	123	53	of	of	ADP
easat-5050	123	54	the	the	DET
easat-5050	123	55	first	first	ADJ
easat-5050	123	56	layer	layer	NOUN
easat-5050	123	57	.	.	PUNCT
easat-5050	124	1	3	3	X
easat-5050	124	2	.	.	X
easat-5050	124	3	experiments	experiment	NOUN
easat-5050	124	4	a	a	DET
easat-5050	124	5	publicly	publicly	ADV
easat-5050	124	6	available	available	ADJ
easat-5050	124	7	bearing	bearing	NOUN
easat-5050	124	8	dataset	dataset	NOUN
easat-5050	124	9	is	be	AUX
easat-5050	124	10	used	use	VERB
easat-5050	124	11	to	to	PART
easat-5050	124	12	validate	validate	VERB
easat-5050	124	13	the	the	DET
easat-5050	124	14	performance	performance	NOUN
easat-5050	124	15	of	of	ADP
easat-5050	124	16	the	the	DET
easat-5050	124	17	mcfcnn	mcfcnn	PROPN
easat-5050	124	18	model	model	NOUN
easat-5050	124	19	.	.	PUNCT
easat-5050	125	1	the	the	DET
easat-5050	125	2	proposed	propose	VERB
easat-5050	125	3	method	method	NOUN
easat-5050	125	4	is	be	AUX
easat-5050	125	5	compared	compare	VERB
easat-5050	125	6	with	with	ADP
easat-5050	125	7	other	other	ADJ
easat-5050	125	8	methods	method	NOUN
easat-5050	125	9	under	under	ADP
easat-5050	125	10	different	different	ADJ
easat-5050	125	11	noise	noise	NOUN
easat-5050	125	12	conditions	condition	NOUN
easat-5050	125	13	.	.	PUNCT
easat-5050	126	1	3.1	3.1	NUM
easat-5050	126	2	.	.	PUNCT
easat-5050	126	3	dataset	dataset	ADJ
easat-5050	126	4	description	description	NOUN
easat-5050	126	5	the	the	DET
easat-5050	126	6	case	case	NOUN
easat-5050	126	7	western	western	PROPN
easat-5050	126	8	reserve	reserve	PROPN
easat-5050	126	9	university	university	PROPN
easat-5050	126	10	bearing	bear	VERB
easat-5050	126	11	dataset	dataset	NOUN
easat-5050	126	12	(	(	PUNCT
easat-5050	126	13	cwru	cwru	PROPN
easat-5050	126	14	)	)	PUNCT
easat-5050	126	15	(	(	PUNCT
easat-5050	126	16	achen	achen	NOUN
easat-5050	126	17	/	/	SYM
easat-5050	126	18	case	case	NOUN
easat-5050	126	19	,	,	PUNCT
easat-5050	126	20	2024	2024	NUM
easat-5050	126	21	)	)	PUNCT
easat-5050	126	22	is	be	AUX
easat-5050	126	23	a	a	DET
easat-5050	126	24	publicly	publicly	ADV
easat-5050	126	25	available	available	ADJ
easat-5050	126	26	dataset	dataset	NOUN
easat-5050	126	27	for	for	ADP
easat-5050	126	28	bearing	bear	VERB
easat-5050	126	29	fault	fault	NOUN
easat-5050	126	30	diagnosis	diagnosis	NOUN
easat-5050	126	31	,	,	PUNCT
easat-5050	126	32	created	create	VERB
easat-5050	126	33	in	in	ADP
easat-5050	126	34	2003	2003	NUM
easat-5050	126	35	by	by	ADP
easat-5050	126	36	dr	dr	PROPN
easat-5050	126	37	.	.	PROPN
easat-5050	126	38	james	james	PROPN
easat-5050	126	39	t.	t.	PROPN
easat-5050	126	40	allison	allison	PROPN
easat-5050	126	41	and	and	CCONJ
easat-5050	126	42	colleagues	colleague	NOUN
easat-5050	126	43	from	from	ADP
easat-5050	126	44	the	the	DET
easat-5050	126	45	department	department	NOUN
easat-5050	126	46	of	of	ADP
easat-5050	126	47	mechanical	mechanical	ADJ
easat-5050	126	48	engineering	engineering	NOUN
easat-5050	126	49	at	at	ADP
easat-5050	126	50	case	case	NOUN
easat-5050	126	51	western	western	PROPN
easat-5050	126	52	reserve	reserve	PROPN
easat-5050	126	53	university	university	PROPN
easat-5050	126	54	in	in	ADP
easat-5050	126	55	the	the	DET
easat-5050	126	56	united	united	PROPN
easat-5050	126	57	states	states	PROPN
easat-5050	126	58	.	.	PUNCT
easat-5050	127	1	this	this	DET
easat-5050	127	2	dataset	dataset	NOUN
easat-5050	127	3	consists	consist	VERB
easat-5050	127	4	of	of	ADP
easat-5050	127	5	accelerometer	accelerometer	NOUN
easat-5050	127	6	signals	signal	NOUN
easat-5050	127	7	from	from	ADP
easat-5050	127	8	four	four	NUM
easat-5050	127	9	bearings	bearing	NOUN
easat-5050	127	10	,	,	PUNCT
easat-5050	127	11	each	each	PRON
easat-5050	127	12	containing	contain	VERB
easat-5050	127	13	data	datum	NOUN
easat-5050	127	14	from	from	ADP
easat-5050	127	15	normal	normal	ADJ
easat-5050	127	16	operation	operation	NOUN
easat-5050	127	17	as	as	ADV
easat-5050	127	18	well	well	ADV
easat-5050	127	19	as	as	ADP
easat-5050	127	20	various	various	ADJ
easat-5050	127	21	types	type	NOUN
easat-5050	127	22	of	of	ADP
easat-5050	127	23	fault	fault	NOUN
easat-5050	127	24	conditions	condition	NOUN
easat-5050	127	25	.	.	PUNCT
easat-5050	128	1	the	the	DET
easat-5050	128	2	fault	fault	NOUN
easat-5050	128	3	types	type	NOUN
easat-5050	128	4	include	include	VERB
easat-5050	128	5	inner	inner	ADJ
easat-5050	128	6	race	race	NOUN
easat-5050	128	7	faults	fault	NOUN
easat-5050	128	8	,	,	PUNCT
easat-5050	128	9	outer	outer	ADJ
easat-5050	128	10	race	race	NOUN
easat-5050	128	11	faults	fault	NOUN
easat-5050	128	12	,	,	PUNCT
easat-5050	128	13	and	and	CCONJ
easat-5050	128	14	rolling	rolling	ADJ
easat-5050	128	15	element	element	NOUN
easat-5050	128	16	faults	fault	NOUN
easat-5050	128	17	.	.	PUNCT
easat-5050	129	1	this	this	DET
easat-5050	129	2	dataset	dataset	NOUN
easat-5050	129	3	is	be	AUX
easat-5050	129	4	utilised	utilise	VERB
easat-5050	129	5	for	for	ADP
easat-5050	129	6	research	research	NOUN
easat-5050	129	7	and	and	CCONJ
easat-5050	129	8	experimentation	experimentation	NOUN
easat-5050	129	9	in	in	ADP
easat-5050	129	10	mechanical	mechanical	ADJ
easat-5050	129	11	fault	fault	NOUN
easat-5050	129	12	diagnosis	diagnosis	NOUN
easat-5050	129	13	and	and	CCONJ
easat-5050	129	14	predictive	predictive	ADJ
easat-5050	129	15	analytics	analytic	NOUN
easat-5050	129	16	by	by	ADP
easat-5050	129	17	many	many	ADJ
easat-5050	129	18	other	other	ADJ
easat-5050	129	19	researchers	researcher	NOUN
easat-5050	129	20	such	such	ADJ
easat-5050	129	21	as	as	ADP
easat-5050	129	22	in	in	ADP
easat-5050	129	23	jalayer	jalayer	PROPN
easat-5050	129	24	,	,	PUNCT
easat-5050	129	25	jalayer	jalayer	PROPN
easat-5050	129	26	,	,	PUNCT
easat-5050	129	27	mor	mor	ADV
easat-5050	129	28	,	,	PUNCT
easat-5050	129	29	orsenigo	orsenigo	ADJ
easat-5050	129	30	,	,	PUNCT
easat-5050	129	31	and	and	CCONJ
easat-5050	129	32	vercellis	vercellis	X
easat-5050	129	33	(	(	PUNCT
easat-5050	129	34	2024	2024	NUM
easat-5050	129	35	)	)	PUNCT
easat-5050	129	36	;	;	PUNCT
easat-5050	129	37	(	(	PUNCT
easat-5050	129	38	y.	y.	PROPN
easat-5050	129	39	liao	liao	PROPN
easat-5050	129	40	et	et	PROPN
easat-5050	129	41	al	al	PROPN
easat-5050	129	42	.	.	PROPN
easat-5050	129	43	,	,	PUNCT
easat-5050	129	44	2023	2023	NUM
easat-5050	129	45	)	)	PUNCT
easat-5050	129	46	;	;	PUNCT
easat-5050	129	47	rosa	rosa	PROPN
easat-5050	129	48	,	,	PUNCT
easat-5050	129	49	braga	braga	NOUN
easat-5050	129	50	,	,	PUNCT
easat-5050	129	51	and	and	CCONJ
easat-5050	129	52	silva	silva	NOUN
easat-5050	129	53	(	(	PUNCT
easat-5050	129	54	2024	2024	NUM
easat-5050	129	55	)	)	PUNCT
easat-5050	129	56	and	and	CCONJ
easat-5050	129	57	yoo	yoo	PROPN
easat-5050	129	58	,	,	PUNCT
easat-5050	129	59	jo	jo	PROPN
easat-5050	129	60	,	,	PUNCT
easat-5050	129	61	and	and	CCONJ
easat-5050	129	62	ban	ban	NOUN
easat-5050	129	63	(	(	PUNCT
easat-5050	129	64	2023	2023	NUM
easat-5050	129	65	)	)	PUNCT
easat-5050	129	66	.	.	PUNCT
easat-5050	130	1	it	it	PRON
easat-5050	130	2	contains	contain	VERB
easat-5050	130	3	extensive	extensive	ADJ
easat-5050	130	4	time	time	NOUN
easat-5050	130	5	-	-	PUNCT
easat-5050	130	6	series	series	NOUN
easat-5050	130	7	signals	signal	NOUN
easat-5050	130	8	suitable	suitable	ADJ
easat-5050	130	9	for	for	ADP
easat-5050	130	10	studying	study	VERB
easat-5050	130	11	the	the	DET
easat-5050	130	12	performance	performance	NOUN
easat-5050	130	13	of	of	ADP
easat-5050	130	14	vibration	vibration	NOUN
easat-5050	130	15	analysis	analysis	NOUN
easat-5050	130	16	and	and	CCONJ
easat-5050	130	17	signal	signal	NOUN
easat-5050	130	18	processing	processing	NOUN
easat-5050	130	19	algorithms	algorithm	NOUN
easat-5050	130	20	.	.	PUNCT
easat-5050	131	1	the	the	DET
easat-5050	131	2	test	test	NOUN
easat-5050	131	3	rig	rig	NOUN
easat-5050	131	4	consisted	consist	VERB
easat-5050	131	5	of	of	ADP
easat-5050	131	6	a	a	DET
easat-5050	131	7	2	2	NUM
easat-5050	131	8	-	-	PUNCT
easat-5050	131	9	horsepower	horsepower	NOUN
easat-5050	131	10	motor	motor	NOUN
easat-5050	131	11	,	,	PUNCT
easat-5050	131	12	a	a	DET
easat-5050	131	13	torque	torque	NOUN
easat-5050	131	14	sensor	sensor	NOUN
easat-5050	131	15	/	/	SYM
easat-5050	131	16	encoder	encoder	NOUN
easat-5050	131	17	,	,	PUNCT
easat-5050	131	18	and	and	CCONJ
easat-5050	131	19	a	a	DET
easat-5050	131	20	dynamometer	dynamometer	NOUN
easat-5050	131	21	.	.	PUNCT
easat-5050	132	1	the	the	DET
easat-5050	132	2	tested	test	VERB
easat-5050	132	3	bearings	bearing	NOUN
easat-5050	132	4	supported	support	VERB
easat-5050	132	5	the	the	DET
easat-5050	132	6	motor	motor	NOUN
easat-5050	132	7	shaft	shaft	NOUN
easat-5050	132	8	.	.	PUNCT
easat-5050	133	1	faults	fault	NOUN
easat-5050	133	2	of	of	ADP
easat-5050	133	3	different	different	ADJ
easat-5050	133	4	sizes	size	NOUN
easat-5050	133	5	were	be	AUX
easat-5050	133	6	induced	induce	VERB
easat-5050	133	7	on	on	ADP
easat-5050	133	8	the	the	DET
easat-5050	133	9	inner	inner	ADJ
easat-5050	133	10	race	race	NOUN
easat-5050	133	11	,	,	PUNCT
easat-5050	133	12	rolling	roll	VERB
easat-5050	133	13	elements	element	NOUN
easat-5050	133	14	,	,	PUNCT
easat-5050	133	15	and	and	CCONJ
easat-5050	133	16	outer	outer	ADJ
easat-5050	133	17	race	race	NOUN
easat-5050	133	18	using	use	VERB
easat-5050	133	19	electrical	electrical	ADJ
easat-5050	133	20	discharge	discharge	NOUN
easat-5050	133	21	machining	machining	NOUN
easat-5050	133	22	.	.	PUNCT
easat-5050	134	1	the	the	DET
easat-5050	134	2	vibration	vibration	NOUN
easat-5050	134	3	data	datum	NOUN
easat-5050	134	4	were	be	AUX
easat-5050	134	5	collected	collect	VERB
easat-5050	134	6	at	at	ADP
easat-5050	134	7	two	two	NUM
easat-5050	134	8	sampling	sample	VERB
easat-5050	134	9	rates	rate	NOUN
easat-5050	134	10	,	,	PUNCT
easat-5050	134	11	12	12	NUM
easat-5050	134	12	and	and	CCONJ
easat-5050	134	13	48	48	NUM
easat-5050	134	14	khz	khz	NOUN
easat-5050	134	15	,	,	PUNCT
easat-5050	134	16	from	from	ADP
easat-5050	134	17	both	both	CCONJ
easat-5050	134	18	the	the	DET
easat-5050	134	19	fan	fan	NOUN
easat-5050	134	20	and	and	CCONJ
easat-5050	134	21	drive	drive	ADJ
easat-5050	134	22	ends	end	NOUN
easat-5050	134	23	of	of	ADP
easat-5050	134	24	the	the	DET
easat-5050	134	25	motor	motor	NOUN
easat-5050	134	26	.	.	PUNCT
easat-5050	135	1	additionally	additionally	ADV
easat-5050	135	2	,	,	PUNCT
easat-5050	135	3	the	the	DET
easat-5050	135	4	data	datum	NOUN
easat-5050	135	5	included	include	VERB
easat-5050	135	6	four	four	NUM
easat-5050	135	7	load	load	NOUN
easat-5050	135	8	conditions	condition	NOUN
easat-5050	135	9	(	(	PUNCT
easat-5050	135	10	0	0	NUM
easat-5050	135	11	,	,	PUNCT
easat-5050	135	12	1	1	NUM
easat-5050	135	13	,	,	PUNCT
easat-5050	135	14	2	2	NUM
easat-5050	135	15	,	,	PUNCT
easat-5050	135	16	and	and	CCONJ
easat-5050	135	17	3	3	NUM
easat-5050	135	18	hp	hp	NOUN
easat-5050	135	19	)	)	PUNCT
easat-5050	135	20	,	,	PUNCT
easat-5050	135	21	with	with	ADP
easat-5050	135	22	motor	motor	NOUN
easat-5050	135	23	speeds	speed	NOUN
easat-5050	135	24	varying	vary	VERB
easat-5050	135	25	between	between	ADP
easat-5050	135	26	1730	1730	NUM
easat-5050	135	27	and	and	CCONJ
easat-5050	135	28	1797	1797	NUM
easat-5050	135	29	rpm	rpm	NOUN
easat-5050	135	30	,	,	PUNCT
easat-5050	135	31	depending	depend	VERB
easat-5050	135	32	on	on	ADP
easat-5050	135	33	the	the	DET
easat-5050	135	34	load	load	NOUN
easat-5050	135	35	.	.	PUNCT
easat-5050	136	1	in	in	ADP
easat-5050	136	2	this	this	DET
easat-5050	136	3	study	study	NOUN
easat-5050	136	4	,	,	PUNCT
easat-5050	136	5	we	we	PRON
easat-5050	136	6	utilized	utilize	VERB
easat-5050	136	7	vibration	vibration	NOUN
easat-5050	136	8	signals	signal	NOUN
easat-5050	136	9	collected	collect	VERB
easat-5050	136	10	at	at	ADP
easat-5050	136	11	the	the	DET
easat-5050	136	12	fan	fan	NOUN
easat-5050	136	13	end	end	NOUN
easat-5050	136	14	at	at	ADP
easat-5050	136	15	a	a	DET
easat-5050	136	16	sampling	sample	VERB
easat-5050	136	17	rate	rate	NOUN
easat-5050	136	18	of	of	ADP
easat-5050	136	19	12	12	NUM
easat-5050	136	20	khz	khz	NOUN
easat-5050	136	21	.	.	PUNCT
easat-5050	137	1	the	the	DET
easat-5050	137	2	dataset	dataset	NOUN
easat-5050	137	3	consists	consist	VERB
easat-5050	137	4	of	of	ADP
easat-5050	137	5	acceleration	acceleration	NOUN
easat-5050	137	6	signals	signal	NOUN
easat-5050	137	7	from	from	ADP
easat-5050	137	8	four	four	NUM
easat-5050	137	9	bearings	bearing	NOUN
easat-5050	137	10	,	,	PUNCT
easat-5050	137	11	each	each	PRON
easat-5050	137	12	containing	contain	VERB
easat-5050	137	13	data	datum	NOUN
easat-5050	137	14	for	for	ADP
easat-5050	137	15	the	the	DET
easat-5050	137	16	normal	normal	ADJ
easat-5050	137	17	state	state	NOUN
easat-5050	137	18	and	and	CCONJ
easat-5050	137	19	different	different	ADJ
easat-5050	137	20	types	type	NOUN
easat-5050	137	21	of	of	ADP
easat-5050	137	22	fault	fault	NOUN
easat-5050	137	23	states	state	NOUN
easat-5050	137	24	.	.	PUNCT
easat-5050	138	1	the	the	DET
easat-5050	138	2	fault	fault	NOUN
easat-5050	138	3	types	type	NOUN
easat-5050	138	4	include	include	VERB
easat-5050	138	5	inner	inner	ADJ
easat-5050	138	6	race	race	NOUN
easat-5050	138	7	,	,	PUNCT
easat-5050	138	8	outer	outer	ADJ
easat-5050	138	9	race	race	NOUN
easat-5050	138	10	,	,	PUNCT
easat-5050	138	11	and	and	CCONJ
easat-5050	138	12	rolling	rolling	ADJ
easat-5050	138	13	element	element	NOUN
easat-5050	138	14	faults	fault	NOUN
easat-5050	138	15	.	.	PUNCT
easat-5050	139	1	each	each	DET
easat-5050	139	2	fault	fault	NOUN
easat-5050	139	3	type	type	NOUN
easat-5050	139	4	has	have	VERB
easat-5050	139	5	three	three	NUM
easat-5050	139	6	diameters	diameter	NOUN
easat-5050	139	7	.	.	PUNCT
easat-5050	140	1	therefore	therefore	ADV
easat-5050	140	2	,	,	PUNCT
easat-5050	140	3	the	the	DET
easat-5050	140	4	dataset	dataset	NOUN
easat-5050	140	5	includes	include	VERB
easat-5050	140	6	nine	nine	NUM
easat-5050	140	7	types	type	NOUN
easat-5050	140	8	of	of	ADP
easat-5050	140	9	faulty	faulty	ADJ
easat-5050	140	10	bearings	bearing	NOUN
easat-5050	140	11	and	and	CCONJ
easat-5050	140	12	one	one	NUM
easat-5050	140	13	healthy	healthy	ADJ
easat-5050	140	14	bearing	bearing	NOUN
easat-5050	140	15	.	.	PUNCT
easat-5050	141	1	the	the	DET
easat-5050	141	2	fault	fault	NOUN
easat-5050	141	3	types	type	NOUN
easat-5050	141	4	and	and	CCONJ
easat-5050	141	5	labels	label	NOUN
easat-5050	141	6	are	be	AUX
easat-5050	141	7	summarized	summarize	VERB
easat-5050	141	8	in	in	ADP
easat-5050	141	9	table	table	NOUN
easat-5050	141	10	1	1	NUM
easat-5050	141	11	.	.	PUNCT
easat-5050	141	12	table	table	NOUN
easat-5050	141	13	1	1	NUM
easat-5050	141	14	.	.	NOUN
easat-5050	141	15	ten	ten	NUM
easat-5050	141	16	classes	class	NOUN
easat-5050	141	17	of	of	ADP
easat-5050	141	18	bearing	bear	VERB
easat-5050	141	19	condition	condition	NOUN
easat-5050	141	20	in	in	ADP
easat-5050	141	21	cwru	cwru	PROPN
easat-5050	141	22	data	datum	NOUN
easat-5050	141	23	set	set	VERB
easat-5050	141	24	.	.	PUNCT
easat-5050	142	1	label	label	NOUN
easat-5050	142	2	fault	fault	NOUN
easat-5050	142	3	mode	mode	NOUN
easat-5050	142	4	label	label	NOUN
easat-5050	142	5	fault	fault	NOUN
easat-5050	142	6	mode	mode	NOUN
easat-5050	142	7	0	0	NUM
easat-5050	142	8	ball	ball	NOUN
easat-5050	142	9	(	(	PUNCT
easat-5050	142	10	slight	slight	ADJ
easat-5050	142	11	)	)	PUNCT
easat-5050	142	12	5	5	NUM
easat-5050	142	13	inner	inner	ADJ
easat-5050	142	14	race	race	NOUN
easat-5050	142	15	(	(	PUNCT
easat-5050	142	16	serious	serious	ADJ
easat-5050	142	17	)	)	PUNCT
easat-5050	142	18	1	1	NUM
easat-5050	142	19	ball	ball	NOUN
easat-5050	142	20	(	(	PUNCT
easat-5050	142	21	medium	medium	NOUN
easat-5050	142	22	)	)	PUNCT
easat-5050	142	23	6	6	NUM
easat-5050	142	24	outer	outer	ADJ
easat-5050	142	25	race	race	NOUN
easat-5050	142	26	(	(	PUNCT
easat-5050	142	27	slight	slight	ADV
easat-5050	142	28	)	)	PUNCT
easat-5050	142	29	2	2	NUM
easat-5050	142	30	ball	ball	NOUN
easat-5050	142	31	(	(	PUNCT
easat-5050	142	32	serious	serious	ADJ
easat-5050	142	33	)	)	PUNCT
easat-5050	142	34	7	7	NUM
easat-5050	142	35	outer	outer	ADJ
easat-5050	142	36	race	race	NOUN
easat-5050	142	37	(	(	PUNCT
easat-5050	142	38	medium	medium	NOUN
easat-5050	142	39	)	)	PUNCT
easat-5050	142	40	3	3	NUM
easat-5050	142	41	inner	inner	ADJ
easat-5050	142	42	race	race	NOUN
easat-5050	142	43	(	(	PUNCT
easat-5050	142	44	slight	slight	ADV
easat-5050	142	45	)	)	PUNCT
easat-5050	142	46	8	8	NUM
easat-5050	142	47	outer	outer	ADJ
easat-5050	142	48	race	race	NOUN
easat-5050	142	49	(	(	PUNCT
easat-5050	142	50	serious	serious	ADJ
easat-5050	142	51	)	)	PUNCT
easat-5050	142	52	4	4	NUM
easat-5050	142	53	inner	inner	ADJ
easat-5050	142	54	race	race	NOUN
easat-5050	142	55	(	(	PUNCT
easat-5050	142	56	medium	medium	NOUN
easat-5050	142	57	)	)	PUNCT
easat-5050	142	58	9	9	NUM
easat-5050	142	59	healthy	healthy	ADJ
easat-5050	142	60	3.2	3.2	NUM
easat-5050	142	61	.	.	PUNCT
easat-5050	143	1	experimental	experimental	ADJ
easat-5050	143	2	setup	setup	NOUN
easat-5050	143	3	cwru	cwru	PROPN
easat-5050	143	4	dataset	dataset	NOUN
easat-5050	143	5	contains	contain	VERB
easat-5050	143	6	very	very	ADV
easat-5050	143	7	long	long	ADJ
easat-5050	143	8	vibration	vibration	NOUN
easat-5050	143	9	signals	signal	NOUN
easat-5050	143	10	,	,	PUNCT
easat-5050	143	11	thus	thus	ADV
easat-5050	143	12	each	each	DET
easat-5050	143	13	sample	sample	NOUN
easat-5050	143	14	was	be	AUX
easat-5050	143	15	defined	define	VERB
easat-5050	143	16	as	as	ADP
easat-5050	143	17	a	a	DET
easat-5050	143	18	sequence	sequence	NOUN
easat-5050	143	19	with	with	ADP
easat-5050	143	20	a	a	DET
easat-5050	143	21	length	length	NOUN
easat-5050	143	22	of	of	ADP
easat-5050	143	23	2048	2048	NUM
easat-5050	143	24	.	.	PUNCT
easat-5050	144	1	data	datum	NOUN
easat-5050	144	2	augmentation	augmentation	NOUN
easat-5050	144	3	techniques	technique	NOUN
easat-5050	144	4	were	be	AUX
easat-5050	144	5	applied	apply	VERB
easat-5050	144	6	to	to	PART
easat-5050	144	7	randomly	randomly	VERB
easat-5050	144	8	crop	crop	NOUN
easat-5050	144	9	1000	1000	NUM
easat-5050	144	10	samples	sample	NOUN
easat-5050	144	11	to	to	PART
easat-5050	144	12	form	form	VERB
easat-5050	144	13	a	a	DET
easat-5050	144	14	dataset	dataset	NOUN
easat-5050	144	15	for	for	ADP
easat-5050	144	16	each	each	DET
easat-5050	144	17	sample	sample	NOUN
easat-5050	144	18	.	.	PUNCT
easat-5050	145	1	the	the	DET
easat-5050	145	2	total	total	ADJ
easat-5050	145	3	number	number	NOUN
easat-5050	145	4	of	of	ADP
easat-5050	145	5	samples	sample	NOUN
easat-5050	145	6	was	be	AUX
easat-5050	145	7	10,000	10,000	NUM
easat-5050	145	8	.	.	PUNCT
easat-5050	146	1	all	all	DET
easat-5050	146	2	samples	sample	NOUN
easat-5050	146	3	were	be	AUX
easat-5050	146	4	normalised	normalise	VERB
easat-5050	146	5	.	.	PUNCT
easat-5050	147	1	finally	finally	ADV
easat-5050	147	2	,	,	PUNCT
easat-5050	147	3	the	the	DET
easat-5050	147	4	dataset	dataset	NOUN
easat-5050	147	5	was	be	AUX
easat-5050	147	6	split	split	VERB
easat-5050	147	7	into	into	ADP
easat-5050	147	8	training	training	NOUN
easat-5050	147	9	,	,	PUNCT
easat-5050	147	10	validation	validation	NOUN
easat-5050	147	11	,	,	PUNCT
easat-5050	147	12	and	and	CCONJ
easat-5050	147	13	testing	testing	NOUN
easat-5050	147	14	sets	set	NOUN
easat-5050	147	15	at	at	ADP
easat-5050	147	16	a	a	DET
easat-5050	147	17	ratio	ratio	NOUN
easat-5050	147	18	of	of	ADP
easat-5050	147	19	0.5:0.25:0.25	0.5:0.25:0.25	PROPN
easat-5050	147	20	.	.	PUNCT
easat-5050	148	1	the	the	DET
easat-5050	148	2	loss	loss	NOUN
easat-5050	148	3	function	function	NOUN
easat-5050	148	4	used	use	VERB
easat-5050	148	5	for	for	ADP
easat-5050	148	6	training	training	NOUN
easat-5050	148	7	was	be	AUX
easat-5050	148	8	cross	cross	ADJ
easat-5050	148	9	-	-	NOUN
easat-5050	148	10	entropy	entropy	ADJ
easat-5050	148	11	,	,	PUNCT
easat-5050	148	12	and	and	CCONJ
easat-5050	148	13	the	the	DET
easat-5050	148	14	network	network	NOUN
easat-5050	148	15	was	be	AUX
easat-5050	148	16	optimised	optimise	VERB
easat-5050	148	17	using	use	VERB
easat-5050	148	18	a	a	DET
easat-5050	148	19	stochastic	stochastic	ADJ
easat-5050	148	20	gradient	gradient	ADJ
easat-5050	148	21	descent	descent	NOUN
easat-5050	148	22	(	(	PUNCT
easat-5050	148	23	sgd	sgd	NOUN
easat-5050	148	24	)	)	PUNCT
easat-5050	148	25	optimiser	optimiser	NOUN
easat-5050	148	26	.	.	PUNCT
easat-5050	149	1	the	the	DET
easat-5050	149	2	gradient	gradient	NOUN
easat-5050	149	3	clipping	clip	VERB
easat-5050	149	4	technique	technique	NOUN
easat-5050	149	5	was	be	AUX
easat-5050	149	6	applied	apply	VERB
easat-5050	149	7	during	during	ADP
easat-5050	149	8	the	the	DET
easat-5050	149	9	iteration	iteration	NOUN
easat-5050	149	10	to	to	PART
easat-5050	149	11	prevent	prevent	VERB
easat-5050	149	12	a	a	DET
easat-5050	149	13	gradient	gradient	ADJ
easat-5050	149	14	explosion	explosion	NOUN
easat-5050	149	15	.	.	PUNCT
easat-5050	150	1	the	the	DET
easat-5050	150	2	batch	batch	NOUN
easat-5050	150	3	size	size	NOUN
easat-5050	150	4	for	for	SCONJ
easat-5050	150	5	all	all	DET
easat-5050	150	6	the	the	DET
easat-5050	150	7	networks	network	NOUN
easat-5050	150	8	was	be	AUX
easat-5050	150	9	set	set	VERB
easat-5050	150	10	to	to	ADP
easat-5050	150	11	64	64	NUM
easat-5050	150	12	,	,	PUNCT
easat-5050	150	13	the	the	DET
easat-5050	150	14	learning	learning	NOUN
easat-5050	150	15	rate	rate	NOUN
easat-5050	150	16	was	be	AUX
easat-5050	150	17	0.1	0.1	NUM
easat-5050	150	18	,	,	PUNCT
easat-5050	150	19	and	and	CCONJ
easat-5050	150	20	the	the	DET
easat-5050	150	21	momentum	momentum	NOUN
easat-5050	150	22	was	be	AUX
easat-5050	150	23	0.9	0.9	NUM
easat-5050	150	24	.	.	PUNCT
easat-5050	151	1	through	through	ADP
easat-5050	151	2	the	the	DET
easat-5050	151	3	iterative	iterative	NOUN
easat-5050	151	4	training	training	NOUN
easat-5050	151	5	of	of	ADP
easat-5050	151	6	the	the	DET
easat-5050	151	7	model	model	NOUN
easat-5050	151	8	,	,	PUNCT
easat-5050	151	9	the	the	DET
easat-5050	151	10	specific	specific	ADJ
easat-5050	151	11	architecture	architecture	NOUN
easat-5050	151	12	of	of	ADP
easat-5050	151	13	the	the	DET
easat-5050	151	14	mcfcnn	mcfcnn	PROPN
easat-5050	151	15	network	network	NOUN
easat-5050	151	16	was	be	AUX
easat-5050	151	17	defined	define	VERB
easat-5050	151	18	,	,	PUNCT
easat-5050	151	19	as	as	SCONJ
easat-5050	151	20	shown	show	VERB
easat-5050	151	21	in	in	ADP
easat-5050	151	22	table	table	NOUN
easat-5050	151	23	2	2	NUM
easat-5050	151	24	.	.	SYM
easat-5050	151	25	2139	2139	NUM
easat-5050	151	26	edelweiss	edelweiss	PROPN
easat-5050	151	27	applied	apply	VERB
easat-5050	151	28	science	science	NOUN
easat-5050	151	29	and	and	CCONJ
easat-5050	151	30	technology	technology	NOUN
easat-5050	151	31	issn	issn	PROPN
easat-5050	151	32	:	:	PUNCT
easat-5050	151	33	2576	2576	NUM
easat-5050	151	34	-	-	SYM
easat-5050	151	35	8484	8484	NUM
easat-5050	151	36	vol	vol	NOUN
easat-5050	151	37	.	.	PROPN
easat-5050	152	1	9	9	NUM
easat-5050	152	2	,	,	PUNCT
easat-5050	152	3	no	no	INTJ
easat-5050	152	4	.	.	NOUN
easat-5050	152	5	2	2	NUM
easat-5050	152	6	:	:	SYM
easat-5050	152	7	2132	2132	NUM
easat-5050	152	8	-	-	SYM
easat-5050	152	9	2146	2146	NUM
easat-5050	152	10	,	,	PUNCT
easat-5050	152	11	2025	2025	NUM
easat-5050	152	12	doi	doi	NOUN
easat-5050	152	13	:	:	PUNCT
easat-5050	152	14	10.55214/25768484.v9i2.5050	10.55214/25768484.v9i2.5050	NUM
easat-5050	152	15	©	©	PROPN
easat-5050	152	16	2025	2025	NUM
easat-5050	152	17	by	by	ADP
easat-5050	152	18	the	the	DET
easat-5050	152	19	authors	author	NOUN
easat-5050	152	20	;	;	PUNCT
easat-5050	152	21	licensee	licensee	PROPN
easat-5050	152	22	learning	learn	VERB
easat-5050	152	23	gate	gate	NOUN
easat-5050	152	24	table	table	NOUN
easat-5050	152	25	2	2	NUM
easat-5050	152	26	.	.	PUNCT
easat-5050	153	1	the	the	DET
easat-5050	153	2	architecture	architecture	NOUN
easat-5050	153	3	configuration	configuration	NOUN
easat-5050	153	4	of	of	ADP
easat-5050	153	5	mcfcnn	mcfcnn	PROPN
easat-5050	153	6	.	.	PUNCT
easat-5050	154	1	layer	layer	NOUN
easat-5050	154	2	type	type	NOUN
easat-5050	154	3	kernel	kernel	PROPN
easat-5050	154	4	channel	channel	PROPN
easat-5050	154	5	stride	stride	NOUN
easat-5050	154	6	padding	padding	NOUN
easat-5050	154	7	output	output	NOUN
easat-5050	154	8	0	0	NUM
easat-5050	154	9	input	input	NOUN
easat-5050	154	10	1x2048	1x2048	PROPN
easat-5050	154	11	1	1	NUM
easat-5050	154	12	conv+bn+relu(4	conv+bn+relu(4	NOUN
easat-5050	154	13	)	)	PUNCT
easat-5050	154	14	3x1	3x1	NUM
easat-5050	154	15	4	4	NUM
easat-5050	154	16	1	1	NUM
easat-5050	154	17	1	1	NUM
easat-5050	154	18	4x2048	4x2048	NOUN
easat-5050	154	19	2	2	NUM
easat-5050	155	1	conv+bn+relu	conv+bn+relu	NOUN
easat-5050	155	2	3x1	3x1	NUM
easat-5050	155	3	32	32	NUM
easat-5050	155	4	1	1	NUM
easat-5050	155	5	1	1	NUM
easat-5050	155	6	32x2048	32x2048	NUM
easat-5050	155	7	3	3	NUM
easat-5050	155	8	squeeze	squeeze	NOUN
easat-5050	155	9	32x1	32x1	NUM
easat-5050	155	10	4	4	NUM
easat-5050	155	11	excitation	excitation	NOUN
easat-5050	155	12	32x2048	32x2048	NUM
easat-5050	155	13	5	5	NUM
easat-5050	155	14	qconv+bn+relu	qconv+bn+relu	PROPN
easat-5050	155	15	64x1	64x1	NUM
easat-5050	155	16	16	16	NUM
easat-5050	155	17	8	8	NUM
easat-5050	155	18	28	28	NUM
easat-5050	155	19	16x256	16x256	NUM
easat-5050	155	20	6	6	NUM
easat-5050	155	21	maxpool	maxpool	NOUN
easat-5050	155	22	2x1	2x1	NUM
easat-5050	155	23	2	2	NUM
easat-5050	155	24	16x128	16x128	NOUN
easat-5050	155	25	7	7	NUM
easat-5050	155	26	qconv+bn+relu	qconv+bn+relu	PROPN
easat-5050	155	27	3x1	3x1	NUM
easat-5050	155	28	32	32	NUM
easat-5050	155	29	1	1	NUM
easat-5050	155	30	1	1	NUM
easat-5050	155	31	32x128	32x128	NOUN
easat-5050	155	32	8	8	NUM
easat-5050	155	33	maxpool	maxpool	NOUN
easat-5050	155	34	2x1	2x1	NUM
easat-5050	155	35	2	2	NUM
easat-5050	155	36	32x64	32x64	NUM
easat-5050	155	37	9	9	NUM
easat-5050	155	38	qconv+bn+relu	qconv+bn+relu	PROPN
easat-5050	155	39	3x1	3x1	NUM
easat-5050	155	40	64	64	NUM
easat-5050	155	41	1	1	NUM
easat-5050	155	42	1	1	NUM
easat-5050	155	43	64x64	64x64	NUM
easat-5050	155	44	10	10	NUM
easat-5050	155	45	maxpool	maxpool	NOUN
easat-5050	155	46	2x1	2x1	NUM
easat-5050	155	47	2	2	NUM
easat-5050	155	48	64x32	64x32	NUM
easat-5050	155	49	11	11	NUM
easat-5050	155	50	qconv+bn+relu	qconv+bn+relu	PROPN
easat-5050	155	51	3x1	3x1	NUM
easat-5050	155	52	64	64	NUM
easat-5050	155	53	1	1	NUM
easat-5050	155	54	1	1	NUM
easat-5050	155	55	64x32	64x32	NUM
easat-5050	155	56	12	12	NUM
easat-5050	155	57	maxpool	maxpool	NOUN
easat-5050	155	58	2x1	2x1	NUM
easat-5050	155	59	2	2	NUM
easat-5050	155	60	64x16	64x16	NUM
easat-5050	155	61	13	13	NUM
easat-5050	155	62	qconv+bn+relu	qconv+bn+relu	PROPN
easat-5050	155	63	3x1	3x1	NUM
easat-5050	155	64	64	64	NUM
easat-5050	155	65	1	1	NUM
easat-5050	155	66	1	1	NUM
easat-5050	155	67	64x16	64x16	NUM
easat-5050	155	68	14	14	NUM
easat-5050	155	69	maxpool	maxpool	NOUN
easat-5050	155	70	2x1	2x1	NUM
easat-5050	155	71	2	2	NUM
easat-5050	155	72	64x8	64x8	NUM
easat-5050	155	73	15	15	NUM
easat-5050	155	74	qconv+bn+relu	qconv+bn+relu	PROPN
easat-5050	155	75	3x1	3x1	NUM
easat-5050	155	76	64	64	NUM
easat-5050	155	77	1	1	NUM
easat-5050	155	78	64x6	64x6	NUM
easat-5050	155	79	16	16	NUM
easat-5050	155	80	maxpool	maxpool	NOUN
easat-5050	155	81	2x1	2x1	NUM
easat-5050	155	82	2	2	NUM
easat-5050	155	83	64x3	64x3	NUM
easat-5050	155	84	17	17	NUM
easat-5050	155	85	linear+relu	linear+relu	PROPN
easat-5050	155	86	192x100	192x100	PROPN
easat-5050	155	87	18	18	NUM
easat-5050	155	88	linear	linear	NOUN
easat-5050	155	89	100x10	100x10	NUM
easat-5050	155	90	19	19	NUM
easat-5050	155	91	output	output	NOUN
easat-5050	155	92	10	10	NUM
easat-5050	155	93	classes	class	NOUN
easat-5050	155	94	the	the	DET
easat-5050	155	95	feature	feature	NOUN
easat-5050	155	96	enhancement	enhancement	NOUN
easat-5050	155	97	layer	layer	NOUN
easat-5050	155	98	consisted	consist	VERB
easat-5050	155	99	of	of	ADP
easat-5050	155	100	six	six	NUM
easat-5050	155	101	layers	layer	NOUN
easat-5050	155	102	:	:	PUNCT
easat-5050	155	103	the	the	DET
easat-5050	155	104	first	first	ADJ
easat-5050	155	105	layer	layer	NOUN
easat-5050	155	106	used	use	VERB
easat-5050	155	107	a	a	DET
easat-5050	155	108	large	large	ADJ
easat-5050	155	109	convolutional	convolutional	ADJ
easat-5050	155	110	kernel	kernel	NOUN
easat-5050	155	111	with	with	ADP
easat-5050	155	112	a	a	DET
easat-5050	155	113	width	width	NOUN
easat-5050	155	114	of	of	ADP
easat-5050	155	115	64	64	NUM
easat-5050	155	116	,	,	PUNCT
easat-5050	155	117	and	and	CCONJ
easat-5050	155	118	the	the	DET
easat-5050	155	119	remaining	remain	VERB
easat-5050	155	120	five	five	NUM
easat-5050	155	121	layers	layer	NOUN
easat-5050	155	122	used	use	VERB
easat-5050	155	123	small	small	ADJ
easat-5050	155	124	convolutional	convolutional	ADJ
easat-5050	155	125	kernels	kernel	NOUN
easat-5050	155	126	with	with	ADP
easat-5050	155	127	a	a	DET
easat-5050	155	128	width	width	NOUN
easat-5050	155	129	of	of	ADP
easat-5050	155	130	3	3	NUM
easat-5050	155	131	.	.	PUNCT
easat-5050	156	1	this	this	DET
easat-5050	156	2	architecture	architecture	NOUN
easat-5050	156	3	was	be	AUX
easat-5050	156	4	obtained	obtain	VERB
easat-5050	156	5	from	from	ADP
easat-5050	156	6	the	the	DET
easat-5050	156	7	w1dcnn	w1dcnn	PROPN
easat-5050	156	8	(	(	PUNCT
easat-5050	156	9	zhang	zhang	PROPN
easat-5050	156	10	et	et	PROPN
easat-5050	156	11	al	al	PROPN
easat-5050	156	12	.	.	PROPN
easat-5050	156	13	,	,	PUNCT
easat-5050	156	14	2017	2017	NUM
easat-5050	156	15	)	)	PUNCT
easat-5050	156	16	.	.	PUNCT
easat-5050	157	1	all	all	DET
easat-5050	157	2	experiments	experiment	NOUN
easat-5050	157	3	were	be	AUX
easat-5050	157	4	conducted	conduct	VERB
easat-5050	157	5	on	on	ADP
easat-5050	157	6	a	a	DET
easat-5050	157	7	windows	window	NOUN
easat-5050	157	8	10	10	NUM
easat-5050	157	9	operating	operating	NOUN
easat-5050	157	10	system	system	NOUN
easat-5050	157	11	equipped	equip	VERB
easat-5050	157	12	with	with	ADP
easat-5050	157	13	an	an	DET
easat-5050	157	14	intel	intel	PROPN
easat-5050	157	15	i512490f	i512490f	PROPN
easat-5050	157	16	cpu	cpu	NOUN
easat-5050	157	17	and	and	CCONJ
easat-5050	157	18	an	an	DET
easat-5050	157	19	nvidia	nvidia	PROPN
easat-5050	157	20	rtx	rtx	PROPN
easat-5050	157	21	3060	3060	NUM
easat-5050	157	22	12	12	NUM
easat-5050	157	23	gb	gb	PROPN
easat-5050	157	24	gpu	gpu	PROPN
easat-5050	157	25	.	.	PUNCT
easat-5050	158	1	our	our	PRON
easat-5050	158	2	code	code	NOUN
easat-5050	158	3	was	be	AUX
easat-5050	158	4	written	write	VERB
easat-5050	158	5	using	use	VERB
easat-5050	158	6	python	python	PROPN
easat-5050	158	7	3.8	3.8	NUM
easat-5050	158	8	and	and	CCONJ
easat-5050	158	9	pytorch	pytorch	NOUN
easat-5050	158	10	.	.	PUNCT
easat-5050	159	1	in	in	ADP
easat-5050	159	2	this	this	DET
easat-5050	159	3	study	study	NOUN
easat-5050	159	4	,	,	PUNCT
easat-5050	159	5	three	three	NUM
easat-5050	159	6	performance	performance	NOUN
easat-5050	159	7	metrics	metric	NOUN
easat-5050	159	8	(	(	PUNCT
easat-5050	159	9	accuracy	accuracy	NOUN
easat-5050	159	10	,	,	PUNCT
easat-5050	159	11	precision	precision	NOUN
easat-5050	159	12	,	,	PUNCT
easat-5050	159	13	and	and	CCONJ
easat-5050	159	14	recall	recall	NOUN
easat-5050	159	15	)	)	PUNCT
easat-5050	159	16	were	be	AUX
easat-5050	159	17	employed	employ	VERB
easat-5050	159	18	to	to	PART
easat-5050	159	19	evaluate	evaluate	VERB
easat-5050	159	20	the	the	DET
easat-5050	159	21	diagnostic	diagnostic	ADJ
easat-5050	159	22	performance	performance	NOUN
easat-5050	159	23	for	for	ADP
easat-5050	159	24	faults	fault	NOUN
easat-5050	159	25	.	.	PUNCT
easat-5050	160	1	their	their	PRON
easat-5050	160	2	definitions	definition	NOUN
easat-5050	160	3	are	be	AUX
easat-5050	160	4	as	as	SCONJ
easat-5050	160	5	follows	follow	VERB
easat-5050	160	6	:	:	PUNCT
easat-5050	160	7	𝐴𝑐𝑐𝑢𝑟𝑎𝑐𝑦	𝐴𝑐𝑐𝑢𝑟𝑎𝑐𝑦	PROPN
easat-5050	160	8	=	=	SYM
easat-5050	160	9	𝑇𝑃+𝑇𝑁	𝑇𝑃+𝑇𝑁	PROPN
easat-5050	160	10	𝑇𝑃+𝐹𝑁+𝐹𝑃+𝑇𝑁	𝑇𝑃+𝐹𝑁+𝐹𝑃+𝑇𝑁	NUM
easat-5050	160	11	×	×	NOUN
easat-5050	160	12	100	100	NUM
easat-5050	160	13	%	%	NOUN
easat-5050	160	14	(	(	PUNCT
easat-5050	160	15	9	9	NUM
easat-5050	160	16	)	)	PUNCT
easat-5050	160	17	𝑃𝑟𝑒𝑐𝑖𝑠𝑖𝑜𝑛	𝑃𝑟𝑒𝑐𝑖𝑠𝑖𝑜𝑛	PROPN
easat-5050	160	18	=	=	SYM
easat-5050	160	19	𝑇𝑃	𝑇𝑃	PROPN
easat-5050	161	1	𝑇𝑃+𝐹𝑃	𝑇𝑃+𝐹𝑃	NOUN
easat-5050	161	2	×	×	NOUN
easat-5050	161	3	100	100	NUM
easat-5050	161	4	%	%	NOUN
easat-5050	161	5	(	(	PUNCT
easat-5050	161	6	10	10	NUM
easat-5050	161	7	)	)	PUNCT
easat-5050	161	8	𝑅𝑒𝑐𝑎𝑙𝑙	𝑅𝑒𝑐𝑎𝑙𝑙	PROPN
easat-5050	161	9	=	=	SYM
easat-5050	161	10	𝑇𝑃	𝑇𝑃	PROPN
easat-5050	161	11	𝑇𝑃+𝐹𝑁	𝑇𝑃+𝐹𝑁	NOUN
easat-5050	161	12	×	×	NOUN
easat-5050	161	13	100	100	NUM
easat-5050	161	14	%	%	NOUN
easat-5050	161	15	(	(	PUNCT
easat-5050	161	16	11	11	NUM
easat-5050	161	17	)	)	PUNCT
easat-5050	161	18	tp	tp	NOUN
easat-5050	161	19	,	,	PUNCT
easat-5050	161	20	fp	fp	PROPN
easat-5050	161	21	,	,	PUNCT
easat-5050	161	22	tn	tn	PROPN
easat-5050	161	23	,	,	PUNCT
easat-5050	161	24	and	and	CCONJ
easat-5050	162	1	fn	fn	NOUN
easat-5050	162	2	mean	mean	ADJ
easat-5050	162	3	true	true	ADJ
easat-5050	162	4	positive	positive	ADJ
easat-5050	162	5	,	,	PUNCT
easat-5050	162	6	false	false	ADJ
easat-5050	162	7	positive	positive	ADJ
easat-5050	162	8	,	,	PUNCT
easat-5050	162	9	true	true	ADJ
easat-5050	162	10	negative	negative	ADJ
easat-5050	162	11	,	,	PUNCT
easat-5050	162	12	and	and	CCONJ
easat-5050	162	13	false	false	ADJ
easat-5050	162	14	negative	negative	ADJ
easat-5050	162	15	,	,	PUNCT
easat-5050	162	16	respectively	respectively	ADV
easat-5050	162	17	.	.	PUNCT
easat-5050	163	1	gaussian	gaussian	ADJ
easat-5050	163	2	white	white	PROPN
easat-5050	163	3	noise	noise	NOUN
easat-5050	163	4	was	be	AUX
easat-5050	163	5	added	add	VERB
easat-5050	163	6	to	to	ADP
easat-5050	163	7	the	the	DET
easat-5050	163	8	dataset	dataset	NOUN
easat-5050	163	9	and	and	CCONJ
easat-5050	163	10	the	the	DET
easat-5050	163	11	performance	performance	NOUN
easat-5050	163	12	metrics	metric	NOUN
easat-5050	163	13	of	of	ADP
easat-5050	163	14	the	the	DET
easat-5050	163	15	model	model	NOUN
easat-5050	163	16	under	under	ADP
easat-5050	163	17	noisy	noisy	ADJ
easat-5050	163	18	conditions	condition	NOUN
easat-5050	163	19	were	be	AUX
easat-5050	163	20	evaluated	evaluate	VERB
easat-5050	163	21	.	.	PUNCT
easat-5050	164	1	the	the	DET
easat-5050	164	2	snr	snr	PROPN
easat-5050	164	3	is	be	AUX
easat-5050	164	4	defined	define	VERB
easat-5050	164	5	as	as	SCONJ
easat-5050	164	6	follows	follow	VERB
easat-5050	164	7	:	:	PUNCT
easat-5050	164	8	snrdb	snrdb	NOUN
easat-5050	164	9	=	=	PROPN
easat-5050	165	1	10	10	NUM
easat-5050	165	2	log10	log10	PROPN
easat-5050	165	3	(	(	PUNCT
easat-5050	165	4	psignal	psignal	ADJ
easat-5050	165	5	pnoise	pnoise	NOUN
easat-5050	165	6	)	)	PUNCT
easat-5050	165	7	(	(	PUNCT
easat-5050	165	8	12	12	NUM
easat-5050	165	9	)	)	PUNCT
easat-5050	165	10	2140	2140	NUM
easat-5050	165	11	edelweiss	edelweiss	PROPN
easat-5050	165	12	applied	apply	VERB
easat-5050	165	13	science	science	NOUN
easat-5050	165	14	and	and	CCONJ
easat-5050	165	15	technology	technology	NOUN
easat-5050	165	16	issn	issn	PROPN
easat-5050	165	17	:	:	PUNCT
easat-5050	165	18	2576	2576	NUM
easat-5050	165	19	-	-	SYM
easat-5050	165	20	8484	8484	NUM
easat-5050	165	21	vol	vol	NOUN
easat-5050	165	22	.	.	PROPN
easat-5050	166	1	9	9	NUM
easat-5050	166	2	,	,	PUNCT
easat-5050	166	3	no	no	INTJ
easat-5050	166	4	.	.	NOUN
easat-5050	166	5	2	2	NUM
easat-5050	166	6	:	:	SYM
easat-5050	166	7	2132	2132	NUM
easat-5050	166	8	-	-	SYM
easat-5050	166	9	2146	2146	NUM
easat-5050	166	10	,	,	PUNCT
easat-5050	166	11	2025	2025	NUM
easat-5050	166	12	doi	doi	NOUN
easat-5050	166	13	:	:	PUNCT
easat-5050	166	14	10.55214/25768484.v9i2.5050	10.55214/25768484.v9i2.5050	NUM
easat-5050	166	15	©	©	PROPN
easat-5050	166	16	2025	2025	NUM
easat-5050	166	17	by	by	ADP
easat-5050	166	18	the	the	DET
easat-5050	166	19	authors	author	NOUN
easat-5050	166	20	;	;	PUNCT
easat-5050	166	21	licensee	licensee	PROPN
easat-5050	166	22	learning	learning	NOUN
easat-5050	166	23	gate	gate	NOUN
easat-5050	166	24	𝑷𝒔𝒊𝒈𝒏𝒂𝒍	𝑷𝒔𝒊𝒈𝒏𝒂𝒍	PROPN
easat-5050	166	25	and	and	CCONJ
easat-5050	166	26	𝑷𝒏𝒐𝒊𝒔𝒆	𝑷𝒏𝒐𝒊𝒔𝒆	PROPN
easat-5050	166	27	are	be	AUX
easat-5050	166	28	the	the	DET
easat-5050	166	29	powers	power	NOUN
easat-5050	166	30	of	of	ADP
easat-5050	166	31	the	the	DET
easat-5050	166	32	signal	signal	NOUN
easat-5050	166	33	and	and	CCONJ
easat-5050	166	34	noise	noise	NOUN
easat-5050	166	35	,	,	PUNCT
easat-5050	166	36	respectively	respectively	ADV
easat-5050	166	37	.	.	PUNCT
easat-5050	167	1	3.3	3.3	NUM
easat-5050	167	2	.	.	PUNCT
easat-5050	168	1	comparison	comparison	NOUN
easat-5050	168	2	with	with	ADP
easat-5050	168	3	other	other	ADJ
easat-5050	168	4	models	model	NOUN
easat-5050	168	5	accuracy	accuracy	NOUN
easat-5050	168	6	:	:	PUNCT
easat-5050	168	7	performance	performance	NOUN
easat-5050	168	8	of	of	ADP
easat-5050	168	9	the	the	DET
easat-5050	168	10	mcfcnn	mcfcnn	NOUN
easat-5050	168	11	model	model	NOUN
easat-5050	168	12	in	in	ADP
easat-5050	168	13	different	different	ADJ
easat-5050	168	14	snr	snr	NOUN
easat-5050	168	15	environments	environment	NOUN
easat-5050	168	16	is	be	AUX
easat-5050	168	17	compared	compare	VERB
easat-5050	168	18	with	with	ADP
easat-5050	168	19	the	the	DET
easat-5050	168	20	w1dcnn	w1dcnn	PROPN
easat-5050	168	21	and	and	CCONJ
easat-5050	168	22	qcnn	qcnn	PROPN
easat-5050	168	23	models	model	NOUN
easat-5050	168	24	using	use	VERB
easat-5050	168	25	the	the	DET
easat-5050	168	26	same	same	ADJ
easat-5050	168	27	data	datum	NOUN
easat-5050	168	28	preprocessing	preprocesse	VERB
easat-5050	168	29	techniques	technique	NOUN
easat-5050	168	30	.	.	PUNCT
easat-5050	169	1	results	result	NOUN
easat-5050	169	2	presented	present	VERB
easat-5050	169	3	in	in	ADP
easat-5050	169	4	table	table	NOUN
easat-5050	169	5	3	3	NUM
easat-5050	169	6	compares	compare	VERB
easat-5050	169	7	the	the	DET
easat-5050	169	8	recognition	recognition	NOUN
easat-5050	169	9	accuracy	accuracy	NOUN
easat-5050	169	10	under	under	ADP
easat-5050	169	11	seven	seven	NUM
easat-5050	169	12	different	different	ADJ
easat-5050	169	13	snr	snr	NOUN
easat-5050	169	14	conditions	condition	NOUN
easat-5050	169	15	,	,	PUNCT
easat-5050	169	16	and	and	CCONJ
easat-5050	169	17	all	all	DET
easat-5050	169	18	the	the	DET
easat-5050	169	19	results	result	NOUN
easat-5050	169	20	were	be	AUX
easat-5050	169	21	averages	average	NOUN
easat-5050	169	22	of	of	ADP
easat-5050	169	23	10	10	NUM
easat-5050	169	24	runs	run	NOUN
easat-5050	169	25	.	.	PUNCT
easat-5050	170	1	the	the	DET
easat-5050	170	2	best	good	ADJ
easat-5050	170	3	results	result	NOUN
easat-5050	170	4	among	among	ADP
easat-5050	170	5	the	the	DET
easat-5050	170	6	three	three	NUM
easat-5050	170	7	models	model	NOUN
easat-5050	170	8	are	be	AUX
easat-5050	170	9	highlighted	highlight	VERB
easat-5050	170	10	in	in	ADP
easat-5050	170	11	bold	bold	ADJ
easat-5050	170	12	,	,	PUNCT
easat-5050	170	13	indicating	indicate	VERB
easat-5050	170	14	that	that	SCONJ
easat-5050	170	15	the	the	DET
easat-5050	170	16	mcfcnn	mcfcnn	PROPN
easat-5050	170	17	model	model	NOUN
easat-5050	170	18	outperformed	outperform	VERB
easat-5050	170	19	the	the	DET
easat-5050	170	20	other	other	ADJ
easat-5050	170	21	two	two	NUM
easat-5050	170	22	models	model	NOUN
easat-5050	170	23	in	in	ADP
easat-5050	170	24	almost	almost	ADV
easat-5050	170	25	every	every	PRON
easat-5050	170	26	snr	snr	PROPN
easat-5050	170	27	environment	environment	NOUN
easat-5050	170	28	.	.	PUNCT
easat-5050	171	1	table	table	NOUN
easat-5050	171	2	3	3	NUM
easat-5050	171	3	.	.	PUNCT
easat-5050	172	1	accuracy	accuracy	NOUN
easat-5050	172	2	of	of	ADP
easat-5050	172	3	different	different	ADJ
easat-5050	172	4	models	model	NOUN
easat-5050	172	5	at	at	ADP
easat-5050	172	6	different	different	ADJ
easat-5050	172	7	snrs	snrs	NOUN
easat-5050	172	8	.	.	PUNCT
easat-5050	173	1	snr	snr	PROPN
easat-5050	174	1	-6	-6	INTJ
easat-5050	174	2	-4	-4	INTJ
easat-5050	175	1	-2	-2	NOUN
easat-5050	175	2	0	0	NUM
easat-5050	175	3	2	2	NUM
easat-5050	175	4	4	4	NUM
easat-5050	175	5	cwru	cwru	NOUN
easat-5050	175	6	(	(	PUNCT
easat-5050	175	7	0hp	0hp	NOUN
easat-5050	175	8	)	)	PUNCT
easat-5050	175	9	w1dcnn	w1dcnn	PROPN
easat-5050	175	10	66.52	66.52	NUM
easat-5050	175	11	%	%	NOUN
easat-5050	175	12	77.36	77.36	NUM
easat-5050	175	13	%	%	NOUN
easat-5050	175	14	87.12	87.12	NUM
easat-5050	175	15	%	%	NOUN
easat-5050	175	16	94.20	94.20	NUM
easat-5050	175	17	%	%	NOUN
easat-5050	175	18	98.36	98.36	NUM
easat-5050	175	19	%	%	NOUN
easat-5050	175	20	99.52	99.52	NUM
easat-5050	175	21	%	%	NOUN
easat-5050	175	22	qcnn	qcnn	NOUN
easat-5050	175	23	71.12	71.12	NUM
easat-5050	175	24	%	%	NOUN
easat-5050	175	25	79.92	79.92	NUM
easat-5050	175	26	%	%	NOUN
easat-5050	175	27	89.28	89.28	NUM
easat-5050	175	28	%	%	NOUN
easat-5050	175	29	95.76	95.76	NUM
easat-5050	175	30	%	%	NOUN
easat-5050	175	31	98.52	98.52	NUM
easat-5050	175	32	%	%	NOUN
easat-5050	175	33	99.64	99.64	NUM
easat-5050	175	34	%	%	NOUN
easat-5050	175	35	mcfcnn	mcfcnn	NOUN
easat-5050	175	36	81.40	81.40	NUM
easat-5050	175	37	%	%	NOUN
easat-5050	175	38	91.04	91.04	NUM
easat-5050	175	39	%	%	NOUN
easat-5050	175	40	96.84	96.84	NUM
easat-5050	175	41	%	%	NOUN
easat-5050	175	42	99.72	99.72	NUM
easat-5050	175	43	%	%	NOUN
easat-5050	175	44	99.92	99.92	NUM
easat-5050	175	45	%	%	NOUN
easat-5050	175	46	99.96	99.96	NUM
easat-5050	175	47	%	%	NOUN
easat-5050	175	48	cwru	cwru	NOUN
easat-5050	175	49	(	(	PUNCT
easat-5050	175	50	1hp	1hp	NOUN
easat-5050	175	51	)	)	PUNCT
easat-5050	175	52	w1dcnn	w1dcnn	PROPN
easat-5050	175	53	69.68	69.68	NUM
easat-5050	175	54	%	%	NOUN
easat-5050	175	55	72.68	72.68	NUM
easat-5050	175	56	%	%	NOUN
easat-5050	175	57	88.60	88.60	NUM
easat-5050	175	58	%	%	NOUN
easat-5050	175	59	92.28	92.28	NUM
easat-5050	175	60	%	%	NOUN
easat-5050	175	61	98.12	98.12	NUM
easat-5050	175	62	%	%	NOUN
easat-5050	175	63	97.92	97.92	NUM
easat-5050	175	64	%	%	NOUN
easat-5050	175	65	qcnn	qcnn	NOUN
easat-5050	175	66	70.80	70.80	NUM
easat-5050	175	67	%	%	NOUN
easat-5050	175	68	81.44	81.44	NUM
easat-5050	175	69	%	%	NOUN
easat-5050	175	70	89.82	89.82	NUM
easat-5050	175	71	%	%	NOUN
easat-5050	175	72	96.24	96.24	NUM
easat-5050	175	73	%	%	NOUN
easat-5050	175	74	98.72	98.72	NUM
easat-5050	175	75	%	%	NOUN
easat-5050	175	76	99.40	99.40	NUM
easat-5050	175	77	%	%	NOUN
easat-5050	175	78	mcfcnn	mcfcnn	NOUN
easat-5050	175	79	81.88	81.88	NUM
easat-5050	175	80	%	%	NOUN
easat-5050	175	81	92.32	92.32	NUM
easat-5050	175	82	%	%	NOUN
easat-5050	175	83	97.44	97.44	NUM
easat-5050	175	84	%	%	NOUN
easat-5050	175	85	96.64	96.64	NUM
easat-5050	175	86	%	%	NOUN
easat-5050	175	87	98.84	98.84	NUM
easat-5050	175	88	%	%	NOUN
easat-5050	175	89	99.16	99.16	NUM
easat-5050	175	90	%	%	NOUN
easat-5050	175	91	cwru	cwru	NOUN
easat-5050	175	92	(	(	PUNCT
easat-5050	175	93	2hp	2hp	NOUN
easat-5050	175	94	)	)	PUNCT
easat-5050	175	95	w1dcnn	w1dcnn	PROPN
easat-5050	175	96	71.92	71.92	NUM
easat-5050	175	97	%	%	NOUN
easat-5050	175	98	83.72	83.72	NUM
easat-5050	175	99	%	%	NOUN
easat-5050	175	100	90.76	90.76	NUM
easat-5050	175	101	%	%	NOUN
easat-5050	175	102	97.80	97.80	NUM
easat-5050	175	103	%	%	NOUN
easat-5050	175	104	99.08	99.08	NUM
easat-5050	175	105	%	%	NOUN
easat-5050	175	106	99.84	99.84	NUM
easat-5050	175	107	%	%	NOUN
easat-5050	175	108	qcnn	qcnn	NOUN
easat-5050	175	109	75.88	75.88	NUM
easat-5050	175	110	%	%	NOUN
easat-5050	175	111	85.16	85.16	NUM
easat-5050	175	112	%	%	NOUN
easat-5050	175	113	93.52	93.52	NUM
easat-5050	175	114	%	%	NOUN
easat-5050	175	115	97.96	97.96	NUM
easat-5050	175	116	%	%	NOUN
easat-5050	175	117	99.64	99.64	NUM
easat-5050	175	118	%	%	NOUN
easat-5050	175	119	100	100	NUM
easat-5050	175	120	%	%	NOUN
easat-5050	175	121	mcfcnn	mcfcnn	NOUN
easat-5050	175	122	87.6	87.6	NUM
easat-5050	175	123	%	%	NOUN
easat-5050	175	124	97.28	97.28	NUM
easat-5050	175	125	%	%	NOUN
easat-5050	175	126	96.64	96.64	NUM
easat-5050	175	127	%	%	NOUN
easat-5050	175	128	99.72	99.72	NUM
easat-5050	175	129	%	%	NOUN
easat-5050	175	130	99.88	99.88	NUM
easat-5050	175	131	%	%	NOUN
easat-5050	175	132	100	100	NUM
easat-5050	175	133	%	%	NOUN
easat-5050	175	134	cwru	cwru	NOUN
easat-5050	175	135	(	(	PUNCT
easat-5050	175	136	3hp	3hp	NOUN
easat-5050	175	137	)	)	PUNCT
easat-5050	175	138	w1dcnn	w1dcnn	PROPN
easat-5050	175	139	72.40	72.40	NUM
easat-5050	175	140	%	%	NOUN
easat-5050	175	141	82.64	82.64	NUM
easat-5050	175	142	%	%	NOUN
easat-5050	175	143	90.68	90.68	NUM
easat-5050	175	144	%	%	NOUN
easat-5050	175	145	96.60	96.60	NUM
easat-5050	175	146	%	%	NOUN
easat-5050	175	147	99.68	99.68	NUM
easat-5050	175	148	%	%	NOUN
easat-5050	175	149	99.88	99.88	NUM
easat-5050	175	150	%	%	NOUN
easat-5050	175	151	qcnn	qcnn	PROPN
easat-5050	175	152	73.96	73.96	NUM
easat-5050	175	153	%	%	NOUN
easat-5050	175	154	84.44	84.44	NUM
easat-5050	175	155	%	%	NOUN
easat-5050	175	156	92.68	92.68	NUM
easat-5050	175	157	%	%	NOUN
easat-5050	175	158	97.56	97.56	NUM
easat-5050	175	159	%	%	NOUN
easat-5050	175	160	99.52	99.52	NUM
easat-5050	175	161	%	%	NOUN
easat-5050	175	162	100	100	NUM
easat-5050	175	163	%	%	NOUN
easat-5050	175	164	mcfcnn	mcfcnn	NOUN
easat-5050	175	165	87.66	87.66	NUM
easat-5050	175	166	%	%	NOUN
easat-5050	175	167	92.84	92.84	NUM
easat-5050	175	168	%	%	NOUN
easat-5050	175	169	98.12	98.12	NUM
easat-5050	175	170	%	%	NOUN
easat-5050	175	171	99.72	99.72	NUM
easat-5050	175	172	%	%	NOUN
easat-5050	175	173	100	100	NUM
easat-5050	175	174	%	%	NOUN
easat-5050	175	175	100	100	NUM
easat-5050	175	176	%	%	NOUN
easat-5050	175	177	based	base	VERB
easat-5050	175	178	on	on	ADP
easat-5050	175	179	the	the	DET
easat-5050	175	180	diagnostic	diagnostic	ADJ
easat-5050	175	181	results	result	NOUN
easat-5050	175	182	in	in	ADP
easat-5050	175	183	table	table	NOUN
easat-5050	175	184	3	3	NUM
easat-5050	175	185	,	,	PUNCT
easat-5050	175	186	we	we	PRON
easat-5050	175	187	calculate	calculate	VERB
easat-5050	175	188	the	the	DET
easat-5050	175	189	average	average	ADJ
easat-5050	175	190	accuracy	accuracy	NOUN
easat-5050	175	191	of	of	ADP
easat-5050	175	192	each	each	DET
easat-5050	175	193	model	model	NOUN
easat-5050	175	194	for	for	ADP
easat-5050	175	195	the	the	DET
easat-5050	175	196	four	four	NUM
easat-5050	175	197	operating	operating	NOUN
easat-5050	175	198	conditions	condition	NOUN
easat-5050	175	199	and	and	CCONJ
easat-5050	175	200	plotted	plot	VERB
easat-5050	175	201	them	they	PRON
easat-5050	175	202	as	as	ADP
easat-5050	175	203	line	line	NOUN
easat-5050	175	204	graphs	graph	NOUN
easat-5050	175	205	,	,	PUNCT
easat-5050	175	206	as	as	SCONJ
easat-5050	175	207	shown	show	VERB
easat-5050	175	208	in	in	ADP
easat-5050	175	209	figure	figure	NOUN
easat-5050	175	210	4	4	NUM
easat-5050	175	211	.	.	PUNCT
easat-5050	175	212	figure	figure	VERB
easat-5050	175	213	4	4	NUM
easat-5050	175	214	.	.	PUNCT
easat-5050	175	215	average	average	ADJ
easat-5050	175	216	accuracy	accuracy	NOUN
easat-5050	175	217	of	of	ADP
easat-5050	175	218	the	the	DET
easat-5050	175	219	model	model	NOUN
easat-5050	175	220	under	under	ADP
easat-5050	175	221	different	different	ADJ
easat-5050	175	222	snrs	snrs	NOUN
easat-5050	175	223	.	.	PUNCT
easat-5050	176	1	2141	2141	NUM
easat-5050	176	2	edelweiss	edelweiss	PROPN
easat-5050	176	3	applied	apply	VERB
easat-5050	176	4	science	science	NOUN
easat-5050	176	5	and	and	CCONJ
easat-5050	176	6	technology	technology	NOUN
easat-5050	176	7	issn	issn	PROPN
easat-5050	176	8	:	:	PUNCT
easat-5050	176	9	2576	2576	NUM
easat-5050	176	10	-	-	SYM
easat-5050	176	11	8484	8484	NUM
easat-5050	176	12	vol	vol	NOUN
easat-5050	176	13	.	.	PROPN
easat-5050	177	1	9	9	NUM
easat-5050	177	2	,	,	PUNCT
easat-5050	177	3	no	no	INTJ
easat-5050	177	4	.	.	NOUN
easat-5050	177	5	2	2	NUM
easat-5050	177	6	:	:	SYM
easat-5050	177	7	2132	2132	NUM
easat-5050	177	8	-	-	SYM
easat-5050	177	9	2146	2146	NUM
easat-5050	177	10	,	,	PUNCT
easat-5050	177	11	2025	2025	NUM
easat-5050	177	12	doi	doi	NOUN
easat-5050	177	13	:	:	PUNCT
easat-5050	177	14	10.55214/25768484.v9i2.5050	10.55214/25768484.v9i2.5050	NUM
easat-5050	177	15	©	©	PROPN
easat-5050	177	16	2025	2025	NUM
easat-5050	177	17	by	by	ADP
easat-5050	177	18	the	the	DET
easat-5050	177	19	authors	author	NOUN
easat-5050	177	20	;	;	PUNCT
easat-5050	177	21	licensee	licensee	PROPN
easat-5050	177	22	learning	learn	VERB
easat-5050	177	23	gate	gate	NOUN
easat-5050	177	24	figure	figure	NOUN
easat-5050	177	25	4(a)-(d	4(a)-(d	NUM
easat-5050	177	26	)	)	PUNCT
easat-5050	177	27	describes	describe	VERB
easat-5050	177	28	that	that	SCONJ
easat-5050	177	29	as	as	SCONJ
easat-5050	177	30	the	the	DET
easat-5050	177	31	snr	snr	NOUN
easat-5050	177	32	increased	increase	VERB
easat-5050	177	33	from	from	ADP
easat-5050	177	34	-6	-6	PROPN
easat-5050	177	35	to	to	ADP
easat-5050	177	36	6	6	NUM
easat-5050	177	37	,	,	PUNCT
easat-5050	177	38	the	the	DET
easat-5050	177	39	diagnostic	diagnostic	ADJ
easat-5050	177	40	accuracy	accuracy	NOUN
easat-5050	177	41	of	of	ADP
easat-5050	177	42	all	all	DET
easat-5050	177	43	three	three	NUM
easat-5050	177	44	models	model	NOUN
easat-5050	177	45	improved	improve	VERB
easat-5050	177	46	.	.	PUNCT
easat-5050	178	1	this	this	PRON
easat-5050	178	2	implied	imply	VERB
easat-5050	178	3	that	that	SCONJ
easat-5050	178	4	when	when	SCONJ
easat-5050	178	5	the	the	DET
easat-5050	178	6	noise	noise	NOUN
easat-5050	178	7	in	in	ADP
easat-5050	178	8	the	the	DET
easat-5050	178	9	signal	signal	NOUN
easat-5050	178	10	is	be	AUX
easat-5050	178	11	relatively	relatively	ADV
easat-5050	178	12	low	low	ADJ
easat-5050	178	13	,	,	PUNCT
easat-5050	178	14	and	and	CCONJ
easat-5050	178	15	the	the	DET
easat-5050	178	16	difference	difference	NOUN
easat-5050	178	17	between	between	ADP
easat-5050	178	18	the	the	DET
easat-5050	178	19	signal	signal	NOUN
easat-5050	178	20	and	and	CCONJ
easat-5050	178	21	background	background	NOUN
easat-5050	178	22	noise	noise	NOUN
easat-5050	178	23	is	be	AUX
easat-5050	178	24	more	more	ADV
easat-5050	178	25	pronounced	pronounce	VERB
easat-5050	178	26	,	,	PUNCT
easat-5050	178	27	the	the	DET
easat-5050	178	28	models	model	NOUN
easat-5050	178	29	are	be	AUX
easat-5050	178	30	more	more	ADV
easat-5050	178	31	likely	likely	ADJ
easat-5050	178	32	to	to	PART
easat-5050	178	33	accurately	accurately	ADV
easat-5050	178	34	identify	identify	VERB
easat-5050	178	35	and	and	CCONJ
easat-5050	178	36	classify	classify	VERB
easat-5050	178	37	the	the	DET
easat-5050	178	38	faults	fault	NOUN
easat-5050	178	39	.	.	PUNCT
easat-5050	179	1	figure	figure	VERB
easat-5050	179	2	4(c	4(c	NUM
easat-5050	179	3	)	)	PUNCT
easat-5050	179	4	and	and	CCONJ
easat-5050	179	5	(	(	PUNCT
easat-5050	179	6	d	d	X
easat-5050	179	7	)	)	PUNCT
easat-5050	179	8	demonstrates	demonstrate	VERB
easat-5050	179	9	that	that	SCONJ
easat-5050	179	10	the	the	DET
easat-5050	179	11	mcfcnn	mcfcnn	NOUN
easat-5050	179	12	model	model	NOUN
easat-5050	179	13	achieved	achieve	VERB
easat-5050	179	14	an	an	DET
easat-5050	179	15	average	average	ADJ
easat-5050	179	16	accuracy	accuracy	NOUN
easat-5050	179	17	10	10	NUM
easat-5050	179	18	-	-	SYM
easat-5050	179	19	20	20	NUM
easat-5050	179	20	%	%	NOUN
easat-5050	179	21	higher	high	ADJ
easat-5050	179	22	than	than	ADP
easat-5050	179	23	the	the	DET
easat-5050	179	24	other	other	ADJ
easat-5050	179	25	two	two	NUM
easat-5050	179	26	models	model	NOUN
easat-5050	179	27	when	when	SCONJ
easat-5050	179	28	the	the	DET
easat-5050	179	29	snr	snr	NOUN
easat-5050	179	30	was	be	AUX
easat-5050	180	1	-6	-6	ADV
easat-5050	180	2	.	.	PUNCT
easat-5050	181	1	additionally	additionally	ADV
easat-5050	181	2	,	,	PUNCT
easat-5050	181	3	when	when	SCONJ
easat-5050	181	4	considering	consider	VERB
easat-5050	181	5	cases	case	NOUN
easat-5050	181	6	in	in	ADP
easat-5050	181	7	which	which	PRON
easat-5050	181	8	the	the	DET
easat-5050	181	9	diagnostic	diagnostic	ADJ
easat-5050	181	10	results	result	NOUN
easat-5050	181	11	were	be	AUX
easat-5050	181	12	100	100	NUM
easat-5050	181	13	%	%	NOUN
easat-5050	181	14	accurate	accurate	ADJ
easat-5050	181	15	across	across	ADP
easat-5050	181	16	all	all	DET
easat-5050	181	17	seven	seven	NUM
easat-5050	181	18	snr	snr	NOUN
easat-5050	181	19	levels	level	NOUN
easat-5050	181	20	,	,	PUNCT
easat-5050	181	21	the	the	DET
easat-5050	181	22	w1dcnn	w1dcnn	PROPN
easat-5050	181	23	achieved	achieve	VERB
easat-5050	181	24	this	this	PRON
easat-5050	181	25	only	only	ADV
easat-5050	181	26	once	once	ADV
easat-5050	181	27	,	,	PUNCT
easat-5050	181	28	qcnn	qcnn	PROPN
easat-5050	181	29	achieved	achieve	VERB
easat-5050	181	30	it	it	PRON
easat-5050	181	31	four	four	NUM
easat-5050	181	32	times	time	NOUN
easat-5050	181	33	,	,	PUNCT
easat-5050	181	34	and	and	CCONJ
easat-5050	181	35	mcfcnn	mcfcnn	NOUN
easat-5050	181	36	achieved	achieve	VERB
easat-5050	181	37	it	it	PRON
easat-5050	181	38	six	six	NUM
easat-5050	181	39	times	time	NOUN
easat-5050	181	40	.	.	PUNCT
easat-5050	182	1	therefore	therefore	ADV
easat-5050	182	2	,	,	PUNCT
easat-5050	182	3	it	it	PRON
easat-5050	182	4	can	can	AUX
easat-5050	182	5	be	be	AUX
easat-5050	182	6	concluded	conclude	VERB
easat-5050	182	7	that	that	SCONJ
easat-5050	182	8	the	the	DET
easat-5050	182	9	mcfcnn	mcfcnn	NOUN
easat-5050	182	10	consistently	consistently	ADV
easat-5050	182	11	provides	provide	VERB
easat-5050	182	12	high	high	ADJ
easat-5050	182	13	performance	performance	NOUN
easat-5050	182	14	across	across	ADP
easat-5050	182	15	various	various	ADJ
easat-5050	182	16	noise	noise	NOUN
easat-5050	182	17	environments	environment	NOUN
easat-5050	182	18	.	.	PUNCT
easat-5050	183	1	figure	figure	NOUN
easat-5050	183	2	5	5	NUM
easat-5050	183	3	shows	show	VERB
easat-5050	183	4	the	the	DET
easat-5050	183	5	training	training	NOUN
easat-5050	183	6	loss	loss	NOUN
easat-5050	183	7	graphs	graph	NOUN
easat-5050	183	8	for	for	ADP
easat-5050	183	9	the	the	DET
easat-5050	183	10	three	three	NUM
easat-5050	183	11	models	model	NOUN
easat-5050	183	12	under	under	ADP
easat-5050	183	13	the	the	DET
easat-5050	183	14	1	1	NUM
easat-5050	183	15	hp	hp	NOUN
easat-5050	183	16	condition	condition	NOUN
easat-5050	183	17	.	.	PUNCT
easat-5050	184	1	the	the	DET
easat-5050	184	2	graph	graph	NOUN
easat-5050	184	3	shows	show	VERB
easat-5050	184	4	that	that	SCONJ
easat-5050	184	5	the	the	DET
easat-5050	184	6	loss	loss	NOUN
easat-5050	184	7	exhibited	exhibit	VERB
easat-5050	184	8	an	an	DET
easat-5050	184	9	overall	overall	ADJ
easat-5050	184	10	decreasing	decrease	VERB
easat-5050	184	11	trend	trend	NOUN
easat-5050	184	12	.	.	PUNCT
easat-5050	185	1	as	as	SCONJ
easat-5050	185	2	the	the	DET
easat-5050	185	3	training	training	NOUN
easat-5050	185	4	progressed	progress	VERB
easat-5050	185	5	,	,	PUNCT
easat-5050	185	6	the	the	DET
easat-5050	185	7	loss	loss	NOUN
easat-5050	185	8	gradually	gradually	ADV
easat-5050	185	9	decreased	decrease	VERB
easat-5050	185	10	,	,	PUNCT
easat-5050	185	11	indicating	indicate	VERB
easat-5050	185	12	that	that	SCONJ
easat-5050	185	13	the	the	DET
easat-5050	185	14	models	model	NOUN
easat-5050	185	15	were	be	AUX
easat-5050	185	16	optimised	optimise	VERB
easat-5050	185	17	during	during	ADP
easat-5050	185	18	the	the	DET
easat-5050	185	19	learning	learning	NOUN
easat-5050	185	20	process	process	NOUN
easat-5050	185	21	.	.	PUNCT
easat-5050	186	1	specifically	specifically	ADV
easat-5050	186	2	,	,	PUNCT
easat-5050	186	3	w1dcnn	w1dcnn	PROPN
easat-5050	186	4	had	have	VERB
easat-5050	186	5	the	the	DET
easat-5050	186	6	highest	high	ADJ
easat-5050	186	7	loss	loss	NOUN
easat-5050	186	8	value	value	NOUN
easat-5050	186	9	,	,	PUNCT
easat-5050	186	10	followed	follow	VERB
easat-5050	186	11	by	by	ADP
easat-5050	186	12	qcnn	qcnn	PROPN
easat-5050	186	13	,	,	PUNCT
easat-5050	186	14	whereas	whereas	SCONJ
easat-5050	186	15	mcfcnn	mcfcnn	PRON
easat-5050	186	16	had	have	VERB
easat-5050	186	17	the	the	DET
easat-5050	186	18	lowest	low	ADJ
easat-5050	186	19	loss	loss	NOUN
easat-5050	186	20	.	.	PUNCT
easat-5050	187	1	this	this	PRON
easat-5050	187	2	suggested	suggest	VERB
easat-5050	187	3	that	that	SCONJ
easat-5050	187	4	the	the	DET
easat-5050	187	5	mcfcnn	mcfcnn	NOUN
easat-5050	187	6	model	model	NOUN
easat-5050	187	7	performed	perform	VERB
easat-5050	187	8	relatively	relatively	ADV
easat-5050	187	9	well	well	ADV
easat-5050	187	10	in	in	ADP
easat-5050	187	11	terms	term	NOUN
easat-5050	187	12	of	of	ADP
easat-5050	187	13	loss	loss	NOUN
easat-5050	187	14	.	.	PUNCT
easat-5050	188	1	the	the	DET
easat-5050	188	2	stability	stability	NOUN
easat-5050	188	3	of	of	ADP
easat-5050	188	4	the	the	DET
easat-5050	188	5	model	model	NOUN
easat-5050	188	6	can	can	AUX
easat-5050	188	7	be	be	AUX
easat-5050	188	8	preliminarily	preliminarily	ADV
easat-5050	188	9	assessed	assess	VERB
easat-5050	188	10	by	by	ADP
easat-5050	188	11	observing	observe	VERB
easat-5050	188	12	fluctuations	fluctuation	NOUN
easat-5050	188	13	in	in	ADP
easat-5050	188	14	the	the	DET
easat-5050	188	15	loss	loss	NOUN
easat-5050	188	16	.	.	PUNCT
easat-5050	189	1	mcfcnn	mcfcnn	PROPN
easat-5050	189	2	had	have	VERB
easat-5050	189	3	relatively	relatively	ADV
easat-5050	189	4	small	small	ADJ
easat-5050	189	5	fluctuations	fluctuation	NOUN
easat-5050	189	6	in	in	ADP
easat-5050	189	7	loss	loss	NOUN
easat-5050	189	8	,	,	PUNCT
easat-5050	189	9	indicating	indicate	VERB
easat-5050	189	10	that	that	SCONJ
easat-5050	189	11	it	it	PRON
easat-5050	189	12	is	be	AUX
easat-5050	189	13	a	a	DET
easat-5050	189	14	more	more	ADV
easat-5050	189	15	stable	stable	ADJ
easat-5050	189	16	model	model	NOUN
easat-5050	189	17	with	with	ADP
easat-5050	189	18	better	well	ADJ
easat-5050	189	19	adaptability	adaptability	NOUN
easat-5050	189	20	to	to	ADP
easat-5050	189	21	input	input	NOUN
easat-5050	189	22	data	datum	NOUN
easat-5050	189	23	.	.	PUNCT
easat-5050	190	1	figure	figure	NOUN
easat-5050	190	2	5	5	NUM
easat-5050	190	3	.	.	PUNCT
easat-5050	191	1	losses	loss	NOUN
easat-5050	191	2	of	of	ADP
easat-5050	191	3	different	different	ADJ
easat-5050	191	4	models	model	NOUN
easat-5050	191	5	(	(	PUNCT
easat-5050	191	6	1	1	NUM
easat-5050	191	7	hp	hp	NOUN
easat-5050	191	8	)	)	PUNCT
easat-5050	191	9	.	.	PUNCT
easat-5050	192	1	t	t	PROPN
easat-5050	192	2	-	-	PUNCT
easat-5050	192	3	sne	sne	NOUN
easat-5050	192	4	visualisation	visualisation	NOUN
easat-5050	192	5	:	:	PUNCT
easat-5050	192	6	when	when	SCONJ
easat-5050	192	7	a	a	DET
easat-5050	192	8	signal	signal	NOUN
easat-5050	192	9	is	be	AUX
easat-5050	192	10	heavily	heavily	ADV
easat-5050	192	11	disturbed	disturb	VERB
easat-5050	192	12	by	by	ADP
easat-5050	192	13	noise	noise	NOUN
easat-5050	192	14	,	,	PUNCT
easat-5050	192	15	the	the	DET
easat-5050	192	16	fault	fault	NOUN
easat-5050	192	17	recognition	recognition	NOUN
easat-5050	192	18	of	of	ADP
easat-5050	192	19	bearings	bearing	NOUN
easat-5050	192	20	is	be	AUX
easat-5050	192	21	also	also	ADV
easat-5050	192	22	compromised	compromise	VERB
easat-5050	192	23	.	.	PUNCT
easat-5050	193	1	for	for	ADP
easat-5050	193	2	example	example	NOUN
easat-5050	193	3	,	,	PUNCT
easat-5050	193	4	when	when	SCONJ
easat-5050	193	5	𝑆𝑁𝑅	𝑆𝑁𝑅	PROPN
easat-5050	193	6	=	=	SYM
easat-5050	193	7	−6	−6	NOUN
easat-5050	193	8	and	and	CCONJ
easat-5050	193	9	the	the	DET
easat-5050	193	10	load	load	NOUN
easat-5050	193	11	was	be	AUX
easat-5050	193	12	0	0	NUM
easat-5050	193	13	hp	hp	PROPN
easat-5050	193	14	,	,	PUNCT
easat-5050	193	15	the	the	DET
easat-5050	193	16	classification	classification	NOUN
easat-5050	193	17	accuracy	accuracy	NOUN
easat-5050	193	18	of	of	ADP
easat-5050	193	19	the	the	DET
easat-5050	193	20	w1dcnn	w1dcnn	PROPN
easat-5050	193	21	model	model	NOUN
easat-5050	193	22	was	be	AUX
easat-5050	193	23	66.52	66.52	NUM
easat-5050	193	24	%	%	NOUN
easat-5050	193	25	and	and	CCONJ
easat-5050	193	26	that	that	PRON
easat-5050	193	27	of	of	ADP
easat-5050	193	28	qcnn	qcnn	PROPN
easat-5050	193	29	was	be	AUX
easat-5050	193	30	71.12	71.12	NUM
easat-5050	193	31	%	%	NOUN
easat-5050	193	32	;	;	PUNCT
easat-5050	193	33	however	however	ADV
easat-5050	193	34	,	,	PUNCT
easat-5050	193	35	mcfcnn	mcfcnn	PRON
easat-5050	193	36	still	still	ADV
easat-5050	193	37	maintained	maintain	VERB
easat-5050	193	38	a	a	DET
easat-5050	193	39	high	high	ADJ
easat-5050	193	40	recognition	recognition	NOUN
easat-5050	193	41	rate	rate	NOUN
easat-5050	193	42	of	of	ADP
easat-5050	193	43	81.40	81.40	NUM
easat-5050	193	44	%	%	NOUN
easat-5050	193	45	.	.	PUNCT
easat-5050	194	1	because	because	SCONJ
easat-5050	194	2	qcnn	qcnn	PROPN
easat-5050	194	3	and	and	CCONJ
easat-5050	194	4	mcfcnn	mcfcnn	PRON
easat-5050	194	5	have	have	VERB
easat-5050	194	6	similar	similar	ADJ
easat-5050	194	7	network	network	NOUN
easat-5050	194	8	architectures	architecture	NOUN
easat-5050	194	9	and	and	CCONJ
easat-5050	194	10	quadratic	quadratic	ADJ
easat-5050	194	11	neuron	neuron	NOUN
easat-5050	194	12	networks	network	NOUN
easat-5050	194	13	,	,	PUNCT
easat-5050	194	14	the	the	DET
easat-5050	194	15	parallel	parallel	ADJ
easat-5050	194	16	convolution	convolution	NOUN
easat-5050	194	17	and	and	CCONJ
easat-5050	194	18	se	se	X
easat-5050	194	19	modules	module	NOUN
easat-5050	194	20	in	in	ADP
easat-5050	194	21	our	our	PRON
easat-5050	194	22	proposed	propose	VERB
easat-5050	194	23	network	network	NOUN
easat-5050	194	24	mcfcnn	mcfcnn	PROPN
easat-5050	194	25	,	,	PUNCT
easat-5050	194	26	play	play	VERB
easat-5050	194	27	a	a	DET
easat-5050	194	28	positive	positive	ADJ
easat-5050	194	29	role	role	NOUN
easat-5050	194	30	in	in	ADP
easat-5050	194	31	recognising	recognise	VERB
easat-5050	194	32	fault	fault	NOUN
easat-5050	194	33	features	feature	NOUN
easat-5050	194	34	.	.	PUNCT
easat-5050	195	1	to	to	PART
easat-5050	195	2	better	well	ADV
easat-5050	195	3	compare	compare	VERB
easat-5050	195	4	the	the	DET
easat-5050	195	5	patterns	pattern	NOUN
easat-5050	195	6	learned	learn	VERB
easat-5050	195	7	by	by	ADP
easat-5050	195	8	the	the	DET
easat-5050	195	9	three	three	NUM
easat-5050	195	10	models	model	NOUN
easat-5050	195	11	in	in	ADP
easat-5050	195	12	the	the	DET
easat-5050	195	13	training	training	NOUN
easat-5050	195	14	data	datum	NOUN
easat-5050	195	15	,	,	PUNCT
easat-5050	195	16	we	we	PRON
easat-5050	195	17	output	output	VERB
easat-5050	195	18	the	the	DET
easat-5050	195	19	results	result	NOUN
easat-5050	195	20	of	of	ADP
easat-5050	195	21	the	the	DET
easat-5050	195	22	last	last	ADJ
easat-5050	195	23	convolutional	convolutional	ADJ
easat-5050	195	24	layer	layer	NOUN
easat-5050	195	25	and	and	CCONJ
easat-5050	195	26	then	then	ADV
easat-5050	195	27	reduced	reduce	VERB
easat-5050	195	28	the	the	DET
easat-5050	195	29	high	high	ADV
easat-5050	195	30	-	-	PUNCT
easat-5050	195	31	dimensional	dimensional	ADJ
easat-5050	195	32	data	datum	NOUN
easat-5050	195	33	to	to	ADP
easat-5050	195	34	2d	2d	NOUN
easat-5050	195	35	.	.	PUNCT
easat-5050	196	1	the	the	DET
easat-5050	196	2	data	datum	NOUN
easat-5050	196	3	as	as	SCONJ
easat-5050	196	4	divided	divide	VERB
easat-5050	196	5	into	into	ADP
easat-5050	196	6	ten	ten	NUM
easat-5050	196	7	fault	fault	NOUN
easat-5050	196	8	categories	category	NOUN
easat-5050	196	9	and	and	CCONJ
easat-5050	196	10	used	use	VERB
easat-5050	196	11	t	t	PROPN
easat-5050	196	12	-	-	PUNCT
easat-5050	196	13	sne	sne	PROPN
easat-5050	196	14	(	(	PUNCT
easat-5050	196	15	pouyet	pouyet	PROPN
easat-5050	196	16	,	,	PUNCT
easat-5050	196	17	rohani	rohani	PROPN
easat-5050	196	18	,	,	PUNCT
easat-5050	196	19	katsaggelos	katsaggelo	NOUN
easat-5050	196	20	,	,	PUNCT
easat-5050	196	21	cossairt	cossairt	NOUN
easat-5050	196	22	,	,	PUNCT
easat-5050	196	23	&	&	CCONJ
easat-5050	196	24	walton	walton	PROPN
easat-5050	196	25	,	,	PUNCT
easat-5050	196	26	2018	2018	NUM
easat-5050	196	27	)	)	PUNCT
easat-5050	196	28	for	for	ADP
easat-5050	196	29	visualisation	visualisation	NOUN
easat-5050	196	30	,	,	PUNCT
easat-5050	196	31	as	as	SCONJ
easat-5050	196	32	shown	show	VERB
easat-5050	196	33	in	in	ADP
easat-5050	196	34	figure	figure	NOUN
easat-5050	196	35	6	6	NUM
easat-5050	196	36	.	.	X
easat-5050	196	37	2142	2142	NUM
easat-5050	196	38	edelweiss	edelweiss	PROPN
easat-5050	196	39	applied	apply	VERB
easat-5050	196	40	science	science	NOUN
easat-5050	196	41	and	and	CCONJ
easat-5050	196	42	technology	technology	NOUN
easat-5050	196	43	issn	issn	PROPN
easat-5050	196	44	:	:	PUNCT
easat-5050	196	45	2576	2576	NUM
easat-5050	196	46	-	-	SYM
easat-5050	196	47	8484	8484	NUM
easat-5050	196	48	vol	vol	NOUN
easat-5050	196	49	.	.	PROPN
easat-5050	197	1	9	9	NUM
easat-5050	197	2	,	,	PUNCT
easat-5050	197	3	no	no	INTJ
easat-5050	197	4	.	.	NOUN
easat-5050	197	5	2	2	NUM
easat-5050	197	6	:	:	SYM
easat-5050	197	7	2132	2132	NUM
easat-5050	197	8	-	-	SYM
easat-5050	197	9	2146	2146	NUM
easat-5050	197	10	,	,	PUNCT
easat-5050	197	11	2025	2025	NUM
easat-5050	197	12	doi	doi	NOUN
easat-5050	197	13	:	:	PUNCT
easat-5050	197	14	10.55214/25768484.v9i2.5050	10.55214/25768484.v9i2.5050	NUM
easat-5050	197	15	©	©	PROPN
easat-5050	197	16	2025	2025	NUM
easat-5050	197	17	by	by	ADP
easat-5050	197	18	the	the	DET
easat-5050	197	19	authors	author	NOUN
easat-5050	197	20	;	;	PUNCT
easat-5050	197	21	licensee	licensee	PROPN
easat-5050	197	22	learning	learn	VERB
easat-5050	197	23	gate	gate	PROPN
easat-5050	197	24	figure	figure	NOUN
easat-5050	197	25	6	6	NUM
easat-5050	197	26	.	.	PUNCT
easat-5050	198	1	different	different	ADJ
easat-5050	198	2	models	model	NOUN
easat-5050	198	3	'	'	PART
easat-5050	198	4	t	t	PROPN
easat-5050	198	5	-	-	PUNCT
easat-5050	198	6	sne	sne	NOUN
easat-5050	198	7	visualization	visualization	NOUN
easat-5050	198	8	results	result	NOUN
easat-5050	198	9	on	on	ADP
easat-5050	198	10	the	the	DET
easat-5050	198	11	cwru	cwru	NOUN
easat-5050	198	12	dataset	dataset	NOUN
easat-5050	198	13	(	(	PUNCT
easat-5050	198	14	1hp	1hp	NOUN
easat-5050	198	15	,	,	PUNCT
easat-5050	198	16	𝑆𝑁𝑅	𝑆𝑁𝑅	PROPN
easat-5050	198	17	=	=	SYM
easat-5050	198	18	−4	−4	PROPN
easat-5050	198	19	)	)	PUNCT
easat-5050	198	20	.	.	PUNCT
easat-5050	199	1	the	the	DET
easat-5050	199	2	visualisation	visualisation	NOUN
easat-5050	199	3	results	result	NOUN
easat-5050	199	4	showed	show	VERB
easat-5050	199	5	that	that	SCONJ
easat-5050	199	6	all	all	DET
easat-5050	199	7	three	three	NUM
easat-5050	199	8	models	model	NOUN
easat-5050	199	9	learned	learn	VERB
easat-5050	199	10	patterns	pattern	NOUN
easat-5050	199	11	of	of	ADP
easat-5050	199	12	different	different	ADJ
easat-5050	199	13	categories	category	NOUN
easat-5050	199	14	,	,	PUNCT
easat-5050	199	15	and	and	CCONJ
easat-5050	199	16	the	the	DET
easat-5050	199	17	same	same	ADJ
easat-5050	199	18	fault	fault	NOUN
easat-5050	199	19	categories	category	NOUN
easat-5050	199	20	were	be	AUX
easat-5050	199	21	clustered	cluster	VERB
easat-5050	199	22	.	.	PUNCT
easat-5050	200	1	however	however	ADV
easat-5050	200	2	,	,	PUNCT
easat-5050	200	3	some	some	DET
easat-5050	200	4	overlaps	overlap	NOUN
easat-5050	200	5	occurred	occur	VERB
easat-5050	200	6	in	in	ADP
easat-5050	200	7	the	the	DET
easat-5050	200	8	classification	classification	NOUN
easat-5050	200	9	of	of	ADP
easat-5050	200	10	certain	certain	ADJ
easat-5050	200	11	categories	category	NOUN
easat-5050	200	12	.	.	PUNCT
easat-5050	201	1	for	for	ADP
easat-5050	201	2	example	example	NOUN
easat-5050	201	3	,	,	PUNCT
easat-5050	201	4	in	in	ADP
easat-5050	201	5	the	the	DET
easat-5050	201	6	w1dcnn	w1dcnn	PROPN
easat-5050	201	7	results	result	NOUN
easat-5050	201	8	,	,	PUNCT
easat-5050	201	9	the	the	DET
easat-5050	201	10	orange	orange	ADJ
easat-5050	201	11	,	,	PUNCT
easat-5050	201	12	blue	blue	ADJ
easat-5050	201	13	,	,	PUNCT
easat-5050	201	14	and	and	CCONJ
easat-5050	201	15	green	green	ADJ
easat-5050	201	16	categories	category	NOUN
easat-5050	201	17	were	be	AUX
easat-5050	201	18	not	not	PART
easat-5050	201	19	well	well	ADV
easat-5050	201	20	distinguished	distinguished	ADJ
easat-5050	201	21	at	at	ADP
easat-5050	201	22	the	the	DET
easat-5050	201	23	boundaries	boundary	NOUN
easat-5050	201	24	,	,	PUNCT
easat-5050	201	25	and	and	CCONJ
easat-5050	201	26	the	the	DET
easat-5050	201	27	green	green	ADJ
easat-5050	201	28	category	category	NOUN
easat-5050	201	29	was	be	AUX
easat-5050	201	30	scattered	scatter	VERB
easat-5050	201	31	in	in	ADP
easat-5050	201	32	clusters	cluster	NOUN
easat-5050	201	33	,	,	PUNCT
easat-5050	201	34	indicating	indicate	VERB
easat-5050	201	35	that	that	SCONJ
easat-5050	201	36	this	this	DET
easat-5050	201	37	model	model	NOUN
easat-5050	201	38	did	do	AUX
easat-5050	201	39	not	not	PART
easat-5050	201	40	learn	learn	VERB
easat-5050	201	41	well	well	ADV
easat-5050	201	42	for	for	ADP
easat-5050	201	43	the	the	DET
easat-5050	201	44	green	green	ADJ
easat-5050	201	45	category	category	NOUN
easat-5050	201	46	.	.	PUNCT
easat-5050	202	1	in	in	ADP
easat-5050	202	2	the	the	DET
easat-5050	202	3	qcnn	qcnn	PROPN
easat-5050	202	4	results	result	VERB
easat-5050	202	5	,	,	PUNCT
easat-5050	202	6	some	some	DET
easat-5050	202	7	improvement	improvement	NOUN
easat-5050	202	8	was	be	AUX
easat-5050	202	9	observed	observe	VERB
easat-5050	202	10	,	,	PUNCT
easat-5050	202	11	as	as	SCONJ
easat-5050	202	12	the	the	DET
easat-5050	202	13	orange	orange	PROPN
easat-5050	202	14	category	category	NOUN
easat-5050	202	15	was	be	AUX
easat-5050	202	16	completely	completely	ADV
easat-5050	202	17	separated	separate	VERB
easat-5050	202	18	,	,	PUNCT
easat-5050	202	19	but	but	CCONJ
easat-5050	202	20	the	the	DET
easat-5050	202	21	blue	blue	ADJ
easat-5050	202	22	and	and	CCONJ
easat-5050	202	23	green	green	ADJ
easat-5050	202	24	categories	category	NOUN
easat-5050	202	25	were	be	AUX
easat-5050	202	26	not	not	PART
easat-5050	202	27	completely	completely	ADV
easat-5050	202	28	distinct	distinct	ADJ
easat-5050	202	29	,	,	PUNCT
easat-5050	202	30	and	and	CCONJ
easat-5050	202	31	the	the	DET
easat-5050	202	32	purple	purple	ADJ
easat-5050	202	33	and	and	CCONJ
easat-5050	202	34	dark	dark	ADJ
easat-5050	202	35	yellow	yellow	ADJ
easat-5050	202	36	categories	category	NOUN
easat-5050	202	37	were	be	AUX
easat-5050	202	38	close	close	ADJ
easat-5050	202	39	to	to	ADP
easat-5050	202	40	each	each	DET
easat-5050	202	41	other	other	ADJ
easat-5050	202	42	at	at	ADP
easat-5050	202	43	the	the	DET
easat-5050	202	44	boundaries	boundary	NOUN
easat-5050	202	45	.	.	PUNCT
easat-5050	203	1	in	in	ADP
easat-5050	203	2	the	the	DET
easat-5050	203	3	mcfcnn	mcfcnn	NOUN
easat-5050	203	4	results	result	VERB
easat-5050	203	5	,	,	PUNCT
easat-5050	203	6	although	although	SCONJ
easat-5050	203	7	some	some	DET
easat-5050	203	8	points	point	NOUN
easat-5050	203	9	in	in	ADP
easat-5050	203	10	the	the	DET
easat-5050	203	11	blue	blue	ADJ
easat-5050	203	12	and	and	CCONJ
easat-5050	203	13	green	green	ADJ
easat-5050	203	14	categories	category	NOUN
easat-5050	203	15	were	be	AUX
easat-5050	203	16	close	close	ADJ
easat-5050	203	17	2143	2143	NUM
easat-5050	203	18	edelweiss	edelweiss	PROPN
easat-5050	203	19	applied	apply	VERB
easat-5050	203	20	science	science	NOUN
easat-5050	203	21	and	and	CCONJ
easat-5050	203	22	technology	technology	NOUN
easat-5050	203	23	issn	issn	PROPN
easat-5050	203	24	:	:	PUNCT
easat-5050	203	25	2576	2576	NUM
easat-5050	203	26	-	-	SYM
easat-5050	203	27	8484	8484	NUM
easat-5050	203	28	vol	vol	NOUN
easat-5050	203	29	.	.	PROPN
easat-5050	204	1	9	9	NUM
easat-5050	204	2	,	,	PUNCT
easat-5050	204	3	no	no	INTJ
easat-5050	204	4	.	.	NOUN
easat-5050	204	5	2	2	NUM
easat-5050	204	6	:	:	SYM
easat-5050	204	7	2132	2132	NUM
easat-5050	204	8	-	-	SYM
easat-5050	204	9	2146	2146	NUM
easat-5050	204	10	,	,	PUNCT
easat-5050	204	11	2025	2025	NUM
easat-5050	204	12	doi	doi	NOUN
easat-5050	204	13	:	:	PUNCT
easat-5050	204	14	10.55214/25768484.v9i2.5050	10.55214/25768484.v9i2.5050	NUM
easat-5050	204	15	©	©	PROPN
easat-5050	204	16	2025	2025	NUM
easat-5050	204	17	by	by	ADP
easat-5050	204	18	the	the	DET
easat-5050	204	19	authors	author	NOUN
easat-5050	204	20	;	;	PUNCT
easat-5050	204	21	licensee	licensee	PROPN
easat-5050	204	22	learning	learn	VERB
easat-5050	204	23	gate	gate	NOUN
easat-5050	204	24	to	to	ADP
easat-5050	204	25	the	the	DET
easat-5050	204	26	orange	orange	PROPN
easat-5050	204	27	category	category	NOUN
easat-5050	204	28	,	,	PUNCT
easat-5050	204	29	their	their	PRON
easat-5050	204	30	numbers	number	NOUN
easat-5050	204	31	were	be	AUX
easat-5050	204	32	relatively	relatively	ADV
easat-5050	204	33	small	small	ADJ
easat-5050	204	34	,	,	PUNCT
easat-5050	204	35	and	and	CCONJ
easat-5050	204	36	overall	overall	ADV
easat-5050	204	37	,	,	PUNCT
easat-5050	204	38	the	the	DET
easat-5050	204	39	boundaries	boundary	NOUN
easat-5050	204	40	of	of	ADP
easat-5050	204	41	the	the	DET
easat-5050	204	42	three	three	NUM
easat-5050	204	43	categories	category	NOUN
easat-5050	204	44	were	be	AUX
easat-5050	204	45	distinguished	distinguish	VERB
easat-5050	204	46	.	.	PUNCT
easat-5050	205	1	confusion	confusion	NOUN
easat-5050	205	2	matrix	matrix	NOUN
easat-5050	205	3	comparison	comparison	NOUN
easat-5050	205	4	:	:	PUNCT
easat-5050	205	5	the	the	DET
easat-5050	205	6	classification	classification	NOUN
easat-5050	205	7	performance	performance	NOUN
easat-5050	205	8	of	of	ADP
easat-5050	205	9	the	the	DET
easat-5050	205	10	fault	fault	NOUN
easat-5050	205	11	modes	mode	NOUN
easat-5050	205	12	under	under	ADP
easat-5050	205	13	strong	strong	ADJ
easat-5050	205	14	noise	noise	NOUN
easat-5050	205	15	was	be	AUX
easat-5050	205	16	analysed	analyse	VERB
easat-5050	205	17	in	in	ADP
easat-5050	205	18	detail	detail	NOUN
easat-5050	205	19	by	by	ADP
easat-5050	205	20	calculating	calculate	VERB
easat-5050	205	21	the	the	DET
easat-5050	205	22	confusion	confusion	NOUN
easat-5050	205	23	matrices	matrix	NOUN
easat-5050	205	24	of	of	ADP
easat-5050	205	25	the	the	DET
easat-5050	205	26	mcfcnn	mcfcnn	PROPN
easat-5050	205	27	and	and	CCONJ
easat-5050	205	28	qcnn	qcnn	PROPN
easat-5050	205	29	models	model	NOUN
easat-5050	205	30	.	.	PUNCT
easat-5050	206	1	the	the	DET
easat-5050	206	2	values	value	NOUN
easat-5050	206	3	on	on	ADP
easat-5050	206	4	the	the	DET
easat-5050	206	5	diagonal	diagonal	NOUN
easat-5050	206	6	of	of	ADP
easat-5050	206	7	the	the	DET
easat-5050	206	8	confusion	confusion	NOUN
easat-5050	206	9	matrix	matrix	NOUN
easat-5050	206	10	represent	represent	VERB
easat-5050	206	11	the	the	DET
easat-5050	206	12	numbers	number	NOUN
easat-5050	206	13	of	of	ADP
easat-5050	206	14	correctly	correctly	ADV
easat-5050	206	15	identified	identify	VERB
easat-5050	206	16	instances	instance	NOUN
easat-5050	206	17	.	.	PUNCT
easat-5050	207	1	the	the	DET
easat-5050	207	2	numbers	number	NOUN
easat-5050	207	3	outside	outside	ADP
easat-5050	207	4	the	the	DET
easat-5050	207	5	diagonal	diagonal	NOUN
easat-5050	207	6	represented	represent	VERB
easat-5050	207	7	the	the	DET
easat-5050	207	8	number	number	NOUN
easat-5050	207	9	of	of	ADP
easat-5050	207	10	misclassified	misclassified	ADJ
easat-5050	207	11	instances	instance	NOUN
easat-5050	207	12	.	.	PUNCT
easat-5050	208	1	under	under	ADP
easat-5050	208	2	the	the	DET
easat-5050	208	3	interference	interference	NOUN
easat-5050	208	4	of	of	ADP
easat-5050	208	5	strong	strong	ADJ
easat-5050	208	6	noise	noise	NOUN
easat-5050	208	7	,	,	PUNCT
easat-5050	208	8	the	the	DET
easat-5050	208	9	qcnn	qcnn	PROPN
easat-5050	208	10	model	model	NOUN
easat-5050	208	11	accurately	accurately	ADV
easat-5050	208	12	identified	identify	VERB
easat-5050	208	13	four	four	NUM
easat-5050	208	14	categories	category	NOUN
easat-5050	208	15	.	.	PUNCT
easat-5050	209	1	the	the	DET
easat-5050	209	2	mcfcnn	mcfcnn	PROPN
easat-5050	209	3	model	model	NOUN
easat-5050	209	4	identified	identify	VERB
easat-5050	209	5	five	five	NUM
easat-5050	209	6	categories	category	NOUN
easat-5050	209	7	correctly	correctly	ADV
easat-5050	209	8	.	.	PUNCT
easat-5050	210	1	both	both	DET
easat-5050	210	2	models	model	NOUN
easat-5050	210	3	accurately	accurately	ADV
easat-5050	210	4	identified	identify	VERB
easat-5050	210	5	healthy	healthy	ADJ
easat-5050	210	6	bearings	bearing	NOUN
easat-5050	210	7	.	.	PUNCT
easat-5050	211	1	both	both	DET
easat-5050	211	2	models	model	NOUN
easat-5050	211	3	misclassified	misclassifie	VERB
easat-5050	211	4	52	52	NUM
easat-5050	211	5	instances	instance	NOUN
easat-5050	211	6	as	as	ADP
easat-5050	211	7	outer	outer	ADJ
easat-5050	211	8	race	race	NOUN
easat-5050	211	9	faults	fault	NOUN
easat-5050	211	10	,	,	PUNCT
easat-5050	211	11	indicating	indicate	VERB
easat-5050	211	12	errors	error	NOUN
easat-5050	211	13	in	in	ADP
easat-5050	211	14	identifying	identify	VERB
easat-5050	211	15	the	the	DET
easat-5050	211	16	size	size	NOUN
easat-5050	211	17	of	of	ADP
easat-5050	211	18	the	the	DET
easat-5050	211	19	outer	outer	ADJ
easat-5050	211	20	race	race	NOUN
easat-5050	211	21	faults	fault	NOUN
easat-5050	211	22	.	.	PUNCT
easat-5050	212	1	for	for	ADP
easat-5050	212	2	inner	inner	ADJ
easat-5050	212	3	race	race	NOUN
easat-5050	212	4	faults	fault	NOUN
easat-5050	212	5	,	,	PUNCT
easat-5050	212	6	the	the	DET
easat-5050	212	7	mcfcnn	mcfcnn	NOUN
easat-5050	212	8	model	model	NOUN
easat-5050	212	9	misclassified	misclassifie	VERB
easat-5050	212	10	20	20	NUM
easat-5050	212	11	instances	instance	NOUN
easat-5050	212	12	,	,	PUNCT
easat-5050	212	13	whereas	whereas	SCONJ
easat-5050	212	14	the	the	DET
easat-5050	212	15	qcnn	qcnn	PROPN
easat-5050	212	16	model	model	NOUN
easat-5050	212	17	misclassified	misclassifie	VERB
easat-5050	212	18	77	77	NUM
easat-5050	212	19	instances	instance	NOUN
easat-5050	212	20	;	;	PUNCT
easat-5050	212	21	the	the	DET
easat-5050	212	22	qcnn	qcnn	PROPN
easat-5050	212	23	model	model	NOUN
easat-5050	212	24	mistakenly	mistakenly	ADV
easat-5050	212	25	classified	classify	VERB
easat-5050	212	26	inner	inner	ADJ
easat-5050	212	27	race	race	NOUN
easat-5050	212	28	faults	fault	NOUN
easat-5050	212	29	as	as	ADP
easat-5050	212	30	rolling	roll	VERB
easat-5050	212	31	element	element	NOUN
easat-5050	212	32	faults	fault	NOUN
easat-5050	212	33	.	.	PUNCT
easat-5050	213	1	the	the	DET
easat-5050	213	2	performances	performance	NOUN
easat-5050	213	3	of	of	ADP
easat-5050	213	4	both	both	DET
easat-5050	213	5	models	model	NOUN
easat-5050	213	6	in	in	ADP
easat-5050	213	7	classifying	classify	VERB
easat-5050	213	8	rolling	rolling	ADJ
easat-5050	213	9	element	element	NOUN
easat-5050	213	10	bearing	bear	VERB
easat-5050	213	11	faults	fault	NOUN
easat-5050	213	12	were	be	AUX
easat-5050	213	13	relatively	relatively	ADV
easat-5050	213	14	low	low	ADJ
easat-5050	213	15	,	,	PUNCT
easat-5050	213	16	primarily	primarily	ADV
easat-5050	213	17	because	because	SCONJ
easat-5050	213	18	of	of	ADP
easat-5050	213	19	their	their	PRON
easat-5050	213	20	inability	inability	NOUN
easat-5050	213	21	to	to	PART
easat-5050	213	22	accurately	accurately	ADV
easat-5050	213	23	identify	identify	VERB
easat-5050	213	24	the	the	DET
easat-5050	213	25	fault	fault	NOUN
easat-5050	213	26	sizes	size	NOUN
easat-5050	213	27	of	of	ADP
easat-5050	213	28	the	the	DET
easat-5050	213	29	rolling	rolling	ADJ
easat-5050	213	30	elements	element	NOUN
easat-5050	213	31	.	.	PUNCT
easat-5050	214	1	the	the	DET
easat-5050	214	2	mcfcnn	mcfcnn	PROPN
easat-5050	214	3	model	model	NOUN
easat-5050	214	4	(	(	PUNCT
easat-5050	214	5	475/750	475/750	PROPN
easat-5050	214	6	=	=	SYM
easat-5050	214	7	63.33	63.33	NUM
easat-5050	214	8	%	%	NOUN
easat-5050	214	9	)	)	PUNCT
easat-5050	214	10	accurately	accurately	ADV
easat-5050	214	11	identified	identify	VERB
easat-5050	214	12	more	more	ADJ
easat-5050	214	13	rolling	rolling	ADJ
easat-5050	214	14	element	element	NOUN
easat-5050	214	15	faults	fault	NOUN
easat-5050	214	16	than	than	ADP
easat-5050	214	17	the	the	DET
easat-5050	214	18	qcnn	qcnn	PROPN
easat-5050	214	19	model	model	NOUN
easat-5050	214	20	(	(	PUNCT
easat-5050	214	21	333/750	333/750	NUM
easat-5050	214	22	=	=	SYM
easat-5050	214	23	44.4	44.4	NUM
easat-5050	214	24	%	%	NOUN
easat-5050	214	25	)	)	PUNCT
easat-5050	214	26	.	.	PUNCT
easat-5050	215	1	finally	finally	ADV
easat-5050	215	2	,	,	PUNCT
easat-5050	215	3	we	we	PRON
easat-5050	215	4	were	be	AUX
easat-5050	215	5	more	more	ADV
easat-5050	215	6	interested	interested	ADJ
easat-5050	215	7	in	in	ADP
easat-5050	215	8	the	the	DET
easat-5050	215	9	number	number	NOUN
easat-5050	215	10	of	of	ADP
easat-5050	215	11	misclassified	misclassifie	VERB
easat-5050	215	12	samples	sample	NOUN
easat-5050	215	13	for	for	ADP
easat-5050	215	14	fault	fault	NOUN
easat-5050	215	15	-	-	PUNCT
easat-5050	215	16	type	type	NOUN
easat-5050	215	17	identification	identification	NOUN
easat-5050	215	18	.	.	PUNCT
easat-5050	216	1	despite	despite	SCONJ
easat-5050	216	2	the	the	DET
easat-5050	216	3	strong	strong	ADJ
easat-5050	216	4	noise	noise	NOUN
easat-5050	216	5	interference	interference	NOUN
easat-5050	216	6	,	,	PUNCT
easat-5050	216	7	the	the	DET
easat-5050	216	8	mcfcnn	mcfcnn	NOUN
easat-5050	216	9	model	model	NOUN
easat-5050	216	10	achieved	achieve	VERB
easat-5050	216	11	100	100	NUM
easat-5050	216	12	%	%	NOUN
easat-5050	216	13	accuracy	accuracy	NOUN
easat-5050	216	14	in	in	ADP
easat-5050	216	15	recognising	recognise	VERB
easat-5050	216	16	fault	fault	NOUN
easat-5050	216	17	types	type	NOUN
easat-5050	216	18	,	,	PUNCT
easat-5050	216	19	whereas	whereas	SCONJ
easat-5050	216	20	the	the	DET
easat-5050	216	21	qcnn	qcnn	NOUN
easat-5050	216	22	misidentified	misidentifie	VERB
easat-5050	216	23	77	77	NUM
easat-5050	216	24	samples	sample	NOUN
easat-5050	216	25	of	of	ADP
easat-5050	216	26	inner	inner	ADJ
easat-5050	216	27	race	race	NOUN
easat-5050	216	28	faults	fault	NOUN
easat-5050	216	29	as	as	ADP
easat-5050	216	30	rolling	roll	VERB
easat-5050	216	31	element	element	NOUN
easat-5050	216	32	faults	fault	NOUN
easat-5050	216	33	.	.	PUNCT
easat-5050	217	1	all	all	DET
easat-5050	217	2	results	result	NOUN
easat-5050	217	3	indicated	indicate	VERB
easat-5050	217	4	that	that	SCONJ
easat-5050	217	5	the	the	DET
easat-5050	217	6	mcfcnn	mcfcnn	NOUN
easat-5050	217	7	model	model	NOUN
easat-5050	217	8	performed	perform	VERB
easat-5050	217	9	better	well	ADV
easat-5050	217	10	.	.	PUNCT
easat-5050	218	1	3.4	3.4	NUM
easat-5050	218	2	.	.	PUNCT
easat-5050	218	3	analysis	analysis	NOUN
easat-5050	218	4	of	of	ADP
easat-5050	218	5	mcfcnn	mcfcnn	PRON
easat-5050	218	6	to	to	PART
easat-5050	218	7	determine	determine	VERB
easat-5050	218	8	which	which	DET
easat-5050	218	9	part	part	NOUN
easat-5050	218	10	of	of	ADP
easat-5050	218	11	the	the	DET
easat-5050	218	12	model	model	NOUN
easat-5050	218	13	has	have	VERB
easat-5050	218	14	a	a	DET
easat-5050	218	15	better	well	ADJ
easat-5050	218	16	ability	ability	NOUN
easat-5050	218	17	to	to	PART
easat-5050	218	18	extract	extract	VERB
easat-5050	218	19	fault	fault	NOUN
easat-5050	218	20	signals	signal	NOUN
easat-5050	218	21	from	from	ADP
easat-5050	218	22	the	the	DET
easat-5050	218	23	bearing	bearing	NOUN
easat-5050	218	24	,	,	PUNCT
easat-5050	218	25	we	we	PRON
easat-5050	218	26	decomposed	decompose	VERB
easat-5050	218	27	the	the	DET
easat-5050	218	28	mcfcnn	mcfcnn	NOUN
easat-5050	218	29	model	model	NOUN
easat-5050	218	30	.	.	PUNCT
easat-5050	219	1	we	we	PRON
easat-5050	219	2	separately	separately	ADV
easat-5050	219	3	added	add	VERB
easat-5050	219	4	a	a	DET
easat-5050	219	5	multi	multi	ADJ
easat-5050	219	6	-	-	ADJ
easat-5050	219	7	scale	scale	ADJ
easat-5050	219	8	processing	processing	NOUN
easat-5050	219	9	layer	layer	NOUN
easat-5050	219	10	and	and	CCONJ
easat-5050	219	11	channel	channel	NOUN
easat-5050	219	12	attention	attention	NOUN
easat-5050	219	13	layer	layer	NOUN
easat-5050	219	14	to	to	ADP
easat-5050	219	15	the	the	DET
easat-5050	219	16	network	network	NOUN
easat-5050	219	17	for	for	ADP
easat-5050	219	18	comparison	comparison	NOUN
easat-5050	219	19	.	.	PUNCT
easat-5050	220	1	for	for	ADP
easat-5050	220	2	ease	ease	NOUN
easat-5050	220	3	of	of	ADP
easat-5050	220	4	identification	identification	NOUN
easat-5050	220	5	,	,	PUNCT
easat-5050	220	6	the	the	DET
easat-5050	220	7	network	network	NOUN
easat-5050	220	8	with	with	ADP
easat-5050	220	9	only	only	ADV
easat-5050	220	10	the	the	DET
easat-5050	220	11	multi	multi	ADJ
easat-5050	220	12	-	-	ADJ
easat-5050	220	13	scale	scale	ADJ
easat-5050	220	14	processing	processing	NOUN
easat-5050	220	15	layer	layer	NOUN
easat-5050	220	16	is	be	AUX
easat-5050	220	17	referred	refer	VERB
easat-5050	220	18	to	to	ADP
easat-5050	220	19	as	as	ADP
easat-5050	220	20	mfcnn	mfcnn	NOUN
easat-5050	220	21	,	,	PUNCT
easat-5050	220	22	and	and	CCONJ
easat-5050	220	23	the	the	DET
easat-5050	220	24	network	network	NOUN
easat-5050	220	25	with	with	ADP
easat-5050	220	26	only	only	ADV
easat-5050	220	27	the	the	DET
easat-5050	220	28	channel	channel	NOUN
easat-5050	220	29	attention	attention	NOUN
easat-5050	220	30	layer	layer	NOUN
easat-5050	220	31	is	be	AUX
easat-5050	220	32	referred	refer	VERB
easat-5050	220	33	to	to	ADP
easat-5050	220	34	as	as	ADP
easat-5050	220	35	cfcnn	cfcnn	PROPN
easat-5050	220	36	.	.	PUNCT
easat-5050	221	1	according	accord	VERB
easat-5050	221	2	to	to	ADP
easat-5050	221	3	table	table	NOUN
easat-5050	221	4	4	4	NUM
easat-5050	221	5	,	,	PUNCT
easat-5050	221	6	adding	add	VERB
easat-5050	221	7	these	these	DET
easat-5050	221	8	two	two	NUM
easat-5050	221	9	layers	layer	NOUN
easat-5050	221	10	separately	separately	ADV
easat-5050	221	11	improved	improve	VERB
easat-5050	221	12	the	the	DET
easat-5050	221	13	classification	classification	NOUN
easat-5050	221	14	performance	performance	NOUN
easat-5050	221	15	of	of	ADP
easat-5050	221	16	the	the	DET
easat-5050	221	17	model	model	NOUN
easat-5050	221	18	.	.	PUNCT
easat-5050	222	1	specifically	specifically	ADV
easat-5050	222	2	,	,	PUNCT
easat-5050	222	3	adding	add	VERB
easat-5050	222	4	only	only	ADV
easat-5050	222	5	the	the	DET
easat-5050	222	6	channel	channel	NOUN
easat-5050	222	7	attention	attention	NOUN
easat-5050	222	8	layer	layer	NOUN
easat-5050	222	9	outperformed	outperform	VERB
easat-5050	222	10	adding	add	VERB
easat-5050	222	11	only	only	ADV
easat-5050	222	12	the	the	DET
easat-5050	222	13	multiscale	multiscale	ADJ
easat-5050	222	14	processing	processing	NOUN
easat-5050	222	15	layer	layer	NOUN
easat-5050	222	16	by	by	ADP
easat-5050	222	17	3.46	3.46	NUM
easat-5050	222	18	%	%	NOUN
easat-5050	222	19	.	.	PUNCT
easat-5050	223	1	this	this	PRON
easat-5050	223	2	indicated	indicate	VERB
easat-5050	223	3	that	that	SCONJ
easat-5050	223	4	in	in	ADP
easat-5050	223	5	a	a	DET
easat-5050	223	6	noisy	noisy	ADJ
easat-5050	223	7	environment	environment	NOUN
easat-5050	223	8	,	,	PUNCT
easat-5050	223	9	the	the	DET
easat-5050	223	10	channel	channel	NOUN
easat-5050	223	11	attention	attention	NOUN
easat-5050	223	12	layer	layer	NOUN
easat-5050	223	13	has	have	VERB
easat-5050	223	14	a	a	DET
easat-5050	223	15	better	well	ADJ
easat-5050	223	16	capability	capability	NOUN
easat-5050	223	17	of	of	ADP
easat-5050	223	18	extracting	extract	VERB
easat-5050	223	19	fault	fault	NOUN
easat-5050	223	20	features	feature	NOUN
easat-5050	223	21	and	and	CCONJ
easat-5050	223	22	assigning	assign	VERB
easat-5050	223	23	more	more	ADJ
easat-5050	223	24	weights	weight	NOUN
easat-5050	223	25	to	to	ADP
easat-5050	223	26	them	they	PRON
easat-5050	223	27	.	.	PUNCT
easat-5050	224	1	however	however	ADV
easat-5050	224	2	,	,	PUNCT
easat-5050	224	3	in	in	ADP
easat-5050	224	4	a	a	DET
easat-5050	224	5	strongly	strongly	ADV
easat-5050	224	6	noisy	noisy	ADJ
easat-5050	224	7	environment	environment	NOUN
easat-5050	224	8	,	,	PUNCT
easat-5050	224	9	the	the	DET
easat-5050	224	10	mcfcnn	mcfcnn	PROPN
easat-5050	224	11	model	model	NOUN
easat-5050	224	12	exhibited	exhibit	VERB
easat-5050	224	13	higher	high	ADJ
easat-5050	224	14	accuracy	accuracy	NOUN
easat-5050	224	15	,	,	PUNCT
easat-5050	224	16	precision	precision	NOUN
easat-5050	224	17	,	,	PUNCT
easat-5050	224	18	and	and	CCONJ
easat-5050	224	19	recall	recall	NOUN
easat-5050	224	20	rates	rate	NOUN
easat-5050	224	21	.	.	PUNCT
easat-5050	225	1	table	table	NOUN
easat-5050	225	2	4	4	NUM
easat-5050	225	3	.	.	PUNCT
easat-5050	225	4	comparison	comparison	NOUN
easat-5050	225	5	of	of	ADP
easat-5050	225	6	model	model	NOUN
easat-5050	225	7	results	result	NOUN
easat-5050	225	8	(	(	PUNCT
easat-5050	225	9	snr	snr	PROPN
easat-5050	225	10	=	=	SYM
easat-5050	225	11	-6	-6	PROPN
easat-5050	225	12	)	)	PUNCT
easat-5050	225	13	.	.	PUNCT
easat-5050	226	1	methods	method	NOUN
easat-5050	226	2	accuracy	accuracy	PROPN
easat-5050	226	3	precision	precision	NOUN
easat-5050	226	4	recall	recall	NOUN
easat-5050	226	5	mfcnn	mfcnn	VERB
easat-5050	226	6	0.7716	0.7716	NUM
easat-5050	226	7	0.7933	0.7933	NUM
easat-5050	226	8	0.7908	0.7908	NUM
easat-5050	226	9	cfcnn	cfcnn	NOUN
easat-5050	226	10	0.8065	0.8065	NUM
easat-5050	227	1	0.8215	0.8215	NUM
easat-5050	227	2	0.8192	0.8192	NUM
easat-5050	227	3	mcfcnn	mcfcnn	NOUN
easat-5050	227	4	0.8312	0.8312	NUM
easat-5050	227	5	0.8899	0.8899	NUM
easat-5050	227	6	0.8452	0.8452	NUM
easat-5050	227	7	to	to	PART
easat-5050	227	8	investigate	investigate	VERB
easat-5050	227	9	the	the	DET
easat-5050	227	10	patterns	pattern	NOUN
easat-5050	227	11	learned	learn	VERB
easat-5050	227	12	by	by	ADP
easat-5050	227	13	the	the	DET
easat-5050	227	14	multi	multi	ADJ
easat-5050	227	15	-	-	ADJ
easat-5050	227	16	scale	scale	ADJ
easat-5050	227	17	processing	processing	NOUN
easat-5050	227	18	and	and	CCONJ
easat-5050	227	19	channel	channel	NOUN
easat-5050	227	20	attention	attention	NOUN
easat-5050	227	21	layers	layer	NOUN
easat-5050	227	22	in	in	ADP
easat-5050	227	23	network	network	NOUN
easat-5050	227	24	training	training	NOUN
easat-5050	227	25	,	,	PUNCT
easat-5050	227	26	we	we	PRON
easat-5050	227	27	defined	define	VERB
easat-5050	227	28	them	they	PRON
easat-5050	227	29	as	as	ADP
easat-5050	227	30	the	the	DET
easat-5050	227	31	first	first	ADJ
easat-5050	227	32	layer	layer	NOUN
easat-5050	227	33	of	of	ADP
easat-5050	227	34	the	the	DET
easat-5050	227	35	mcfcnn	mcfcnn	PROPN
easat-5050	227	36	network	network	NOUN
easat-5050	227	37	.	.	PUNCT
easat-5050	228	1	visualisation	visualisation	NOUN
easat-5050	228	2	techniques	technique	NOUN
easat-5050	228	3	were	be	AUX
easat-5050	228	4	applied	apply	VERB
easat-5050	228	5	to	to	PART
easat-5050	228	6	observe	observe	VERB
easat-5050	228	7	the	the	DET
easat-5050	228	8	information	information	NOUN
easat-5050	228	9	focused	focus	VERB
easat-5050	228	10	on	on	ADP
easat-5050	228	11	by	by	ADP
easat-5050	228	12	the	the	DET
easat-5050	228	13	first	first	ADJ
easat-5050	228	14	layer	layer	NOUN
easat-5050	228	15	of	of	ADP
easat-5050	228	16	the	the	DET
easat-5050	228	17	network	network	NOUN
easat-5050	228	18	when	when	SCONJ
easat-5050	228	19	fed	fed	PROPN
easat-5050	228	20	noisy	noisy	ADJ
easat-5050	228	21	,	,	PUNCT
easat-5050	228	22	healthy	healthy	ADJ
easat-5050	228	23	,	,	PUNCT
easat-5050	228	24	and	and	CCONJ
easat-5050	228	25	faulty	faulty	ADJ
easat-5050	228	26	bearing	bearing	NOUN
easat-5050	228	27	signals	signal	NOUN
easat-5050	228	28	.	.	PUNCT
easat-5050	229	1	figure	figure	NOUN
easat-5050	229	2	6(a	6(a	NUM
easat-5050	229	3	)	)	PUNCT
easat-5050	229	4	shows	show	VERB
easat-5050	229	5	the	the	DET
easat-5050	229	6	healthy	healthy	ADJ
easat-5050	229	7	bearing	bearing	NOUN
easat-5050	229	8	signal	signal	NOUN
easat-5050	229	9	and	and	CCONJ
easat-5050	229	10	figure	figure	VERB
easat-5050	229	11	6(b	6(b	NUM
easat-5050	229	12	)	)	PUNCT
easat-5050	229	13	the	the	DET
easat-5050	229	14	faulty	faulty	ADJ
easat-5050	229	15	bearing	bearing	NOUN
easat-5050	229	16	signal	signal	NOUN
easat-5050	229	17	.	.	PUNCT
easat-5050	230	1	2144	2144	NUM
easat-5050	230	2	edelweiss	edelweiss	PROPN
easat-5050	230	3	applied	apply	VERB
easat-5050	230	4	science	science	NOUN
easat-5050	230	5	and	and	CCONJ
easat-5050	230	6	technology	technology	NOUN
easat-5050	230	7	issn	issn	PROPN
easat-5050	230	8	:	:	PUNCT
easat-5050	230	9	2576	2576	NUM
easat-5050	230	10	-	-	SYM
easat-5050	230	11	8484	8484	NUM
easat-5050	230	12	vol	vol	NOUN
easat-5050	230	13	.	.	PROPN
easat-5050	231	1	9	9	NUM
easat-5050	231	2	,	,	PUNCT
easat-5050	231	3	no	no	INTJ
easat-5050	231	4	.	.	NOUN
easat-5050	231	5	2	2	NUM
easat-5050	231	6	:	:	SYM
easat-5050	231	7	2132	2132	NUM
easat-5050	231	8	-	-	SYM
easat-5050	231	9	2146	2146	NUM
easat-5050	231	10	,	,	PUNCT
easat-5050	231	11	2025	2025	NUM
easat-5050	231	12	doi	doi	NOUN
easat-5050	231	13	:	:	PUNCT
easat-5050	231	14	10.55214/25768484.v9i2.5050	10.55214/25768484.v9i2.5050	NUM
easat-5050	231	15	©	©	PROPN
easat-5050	231	16	2025	2025	NUM
easat-5050	231	17	by	by	ADP
easat-5050	231	18	the	the	DET
easat-5050	231	19	authors	author	NOUN
easat-5050	231	20	;	;	PUNCT
easat-5050	231	21	licensee	licensee	PROPN
easat-5050	231	22	learning	learn	VERB
easat-5050	231	23	gate	gate	NOUN
easat-5050	231	24	figure	figure	NOUN
easat-5050	231	25	7	7	NUM
easat-5050	231	26	.	.	PUNCT
easat-5050	231	27	output	output	NOUN
easat-5050	231	28	feature	feature	NOUN
easat-5050	231	29	map	map	NOUN
easat-5050	231	30	of	of	ADP
easat-5050	231	31	the	the	DET
easat-5050	231	32	first	first	ADJ
easat-5050	231	33	layer	layer	NOUN
easat-5050	231	34	of	of	ADP
easat-5050	231	35	the	the	DET
easat-5050	231	36	mcfcnn	mcfcnn	NOUN
easat-5050	231	37	.	.	PUNCT
easat-5050	232	1	figure	figure	NOUN
easat-5050	232	2	7(a	7(a	NUM
easat-5050	232	3	)	)	PUNCT
easat-5050	232	4	shows	show	VERB
easat-5050	232	5	the	the	DET
easat-5050	232	6	original	original	ADJ
easat-5050	232	7	signal	signal	NOUN
easat-5050	232	8	,	,	PUNCT
easat-5050	232	9	signal	signal	VERB
easat-5050	232	10	with	with	ADP
easat-5050	232	11	added	add	VERB
easat-5050	232	12	noise	noise	NOUN
easat-5050	232	13	,	,	PUNCT
easat-5050	232	14	and	and	CCONJ
easat-5050	232	15	output	output	NOUN
easat-5050	232	16	signal	signal	NOUN
easat-5050	232	17	after	after	ADP
easat-5050	232	18	passing	pass	VERB
easat-5050	232	19	through	through	ADP
easat-5050	232	20	the	the	DET
easat-5050	232	21	first	first	ADJ
easat-5050	232	22	layer	layer	NOUN
easat-5050	232	23	of	of	ADP
easat-5050	232	24	the	the	DET
easat-5050	232	25	mcfcnn	mcfcnn	NOUN
easat-5050	232	26	.	.	PUNCT
easat-5050	233	1	figure	figure	VERB
easat-5050	233	2	7(b	7(b	NOUN
easat-5050	233	3	)	)	PUNCT
easat-5050	233	4	shows	show	VERB
easat-5050	233	5	that	that	SCONJ
easat-5050	233	6	despite	despite	SCONJ
easat-5050	233	7	the	the	DET
easat-5050	233	8	severe	severe	ADJ
easat-5050	233	9	noise	noise	NOUN
easat-5050	233	10	interference	interference	NOUN
easat-5050	233	11	in	in	ADP
easat-5050	233	12	the	the	DET
easat-5050	233	13	original	original	ADJ
easat-5050	233	14	signal	signal	NOUN
easat-5050	233	15	,	,	PUNCT
easat-5050	233	16	the	the	DET
easat-5050	233	17	network	network	NOUN
easat-5050	233	18	could	could	AUX
easat-5050	233	19	still	still	ADV
easat-5050	233	20	capture	capture	VERB
easat-5050	233	21	the	the	DET
easat-5050	233	22	location	location	NOUN
easat-5050	233	23	of	of	ADP
easat-5050	233	24	the	the	DET
easat-5050	233	25	fault	fault	NOUN
easat-5050	233	26	signal	signal	NOUN
easat-5050	233	27	.	.	PUNCT
easat-5050	234	1	this	this	PRON
easat-5050	234	2	also	also	ADV
easat-5050	234	3	explains	explain	VERB
easat-5050	234	4	why	why	SCONJ
easat-5050	234	5	the	the	DET
easat-5050	234	6	mcfcnn	mcfcnn	NOUN
easat-5050	234	7	maintained	maintain	VERB
easat-5050	234	8	good	good	ADJ
easat-5050	234	9	classification	classification	NOUN
easat-5050	234	10	performance	performance	NOUN
easat-5050	234	11	even	even	ADV
easat-5050	234	12	under	under	ADP
easat-5050	234	13	noisy	noisy	ADJ
easat-5050	234	14	conditions	condition	NOUN
easat-5050	234	15	.	.	PUNCT
easat-5050	235	1	4	4	X
easat-5050	235	2	.	.	X
easat-5050	235	3	conclusions	conclusion	NOUN
easat-5050	235	4	in	in	ADP
easat-5050	235	5	this	this	DET
easat-5050	235	6	article	article	NOUN
easat-5050	235	7	,	,	PUNCT
easat-5050	235	8	we	we	PRON
easat-5050	235	9	proposed	propose	VERB
easat-5050	235	10	a	a	DET
easat-5050	235	11	convolutional	convolutional	ADJ
easat-5050	235	12	network	network	NOUN
easat-5050	235	13	called	call	VERB
easat-5050	235	14	mcfcnn	mcfcnn	PRON
easat-5050	235	15	for	for	ADP
easat-5050	235	16	end	end	NOUN
easat-5050	235	17	-	-	PUNCT
easat-5050	235	18	to	to	ADP
easat-5050	235	19	-	-	PUNCT
easat-5050	235	20	end	end	NOUN
easat-5050	235	21	bearing	bear	VERB
easat-5050	235	22	fault	fault	NOUN
easat-5050	235	23	diagnosis	diagnosis	NOUN
easat-5050	235	24	.	.	PUNCT
easat-5050	236	1	the	the	DET
easat-5050	236	2	proposed	propose	VERB
easat-5050	236	3	model	model	NOUN
easat-5050	236	4	incorporates	incorporate	VERB
easat-5050	236	5	quadratic	quadratic	ADJ
easat-5050	236	6	neuron	neuron	NOUN
easat-5050	236	7	convolutional	convolutional	ADJ
easat-5050	236	8	network	network	NOUN
easat-5050	236	9	operations	operation	NOUN
easat-5050	236	10	and	and	CCONJ
easat-5050	236	11	introduces	introduce	VERB
easat-5050	236	12	multi	multi	ADJ
easat-5050	236	13	-	-	ADJ
easat-5050	236	14	scale	scale	ADJ
easat-5050	236	15	processing	processing	NOUN
easat-5050	236	16	and	and	CCONJ
easat-5050	236	17	channel	channel	NOUN
easat-5050	236	18	attention	attention	NOUN
easat-5050	236	19	layers	layer	NOUN
easat-5050	236	20	to	to	PART
easat-5050	236	21	enhance	enhance	VERB
easat-5050	236	22	the	the	DET
easat-5050	236	23	diagnostic	diagnostic	ADJ
easat-5050	236	24	capability	capability	NOUN
easat-5050	236	25	of	of	ADP
easat-5050	236	26	the	the	DET
easat-5050	236	27	model	model	NOUN
easat-5050	236	28	in	in	ADP
easat-5050	236	29	the	the	DET
easat-5050	236	30	presence	presence	NOUN
easat-5050	236	31	of	of	ADP
easat-5050	236	32	strong	strong	ADJ
easat-5050	236	33	noise	noise	NOUN
easat-5050	236	34	.	.	PUNCT
easat-5050	237	1	the	the	DET
easat-5050	237	2	network	network	NOUN
easat-5050	237	3	achieved	achieve	VERB
easat-5050	237	4	good	good	ADJ
easat-5050	237	5	classification	classification	NOUN
easat-5050	237	6	performance	performance	NOUN
easat-5050	237	7	on	on	ADP
easat-5050	237	8	the	the	DET
easat-5050	237	9	cwru	cwru	NOUN
easat-5050	237	10	-	-	PUNCT
easat-5050	237	11	bearing	bear	VERB
easat-5050	237	12	dataset	dataset	NOUN
easat-5050	237	13	,	,	PUNCT
easat-5050	237	14	maintaining	maintain	VERB
easat-5050	237	15	an	an	DET
easat-5050	237	16	accuracy	accuracy	NOUN
easat-5050	237	17	of	of	ADP
easat-5050	237	18	over	over	ADP
easat-5050	237	19	80	80	NUM
easat-5050	237	20	%	%	NOUN
easat-5050	237	21	,	,	PUNCT
easat-5050	237	22	even	even	ADV
easat-5050	237	23	in	in	ADP
easat-5050	237	24	the	the	DET
easat-5050	237	25	presence	presence	NOUN
easat-5050	237	26	of	of	ADP
easat-5050	237	27	strong	strong	ADJ
easat-5050	237	28	noise	noise	NOUN
easat-5050	237	29	interference	interference	NOUN
easat-5050	237	30	.	.	PUNCT
easat-5050	238	1	furthermore	furthermore	ADV
easat-5050	238	2	,	,	PUNCT
easat-5050	238	3	classification	classification	NOUN
easat-5050	238	4	tests	test	NOUN
easat-5050	238	5	were	be	AUX
easat-5050	238	6	conducted	conduct	VERB
easat-5050	238	7	on	on	ADP
easat-5050	238	8	the	the	DET
easat-5050	238	9	multi	multi	ADJ
easat-5050	238	10	-	-	ADJ
easat-5050	238	11	scale	scale	ADJ
easat-5050	238	12	processing	processing	NOUN
easat-5050	238	13	and	and	CCONJ
easat-5050	238	14	channel	channel	NOUN
easat-5050	238	15	attention	attention	NOUN
easat-5050	238	16	layers	layer	NOUN
easat-5050	238	17	to	to	PART
easat-5050	238	18	validate	validate	VERB
easat-5050	238	19	their	their	PRON
easat-5050	238	20	importance	importance	NOUN
easat-5050	238	21	in	in	ADP
easat-5050	238	22	extracting	extract	VERB
easat-5050	238	23	bearing	bear	VERB
easat-5050	238	24	fault	fault	NOUN
easat-5050	238	25	features	feature	NOUN
easat-5050	238	26	.	.	PUNCT
easat-5050	239	1	visualisation	visualisation	NOUN
easat-5050	239	2	techniques	technique	NOUN
easat-5050	239	3	were	be	AUX
easat-5050	239	4	employed	employ	VERB
easat-5050	239	5	to	to	PART
easat-5050	239	6	interpret	interpret	VERB
easat-5050	239	7	the	the	DET
easat-5050	239	8	output	output	NOUN
easat-5050	239	9	feature	feature	NOUN
easat-5050	239	10	maps	map	NOUN
easat-5050	239	11	.	.	PUNCT
easat-5050	240	1	in	in	ADP
easat-5050	240	2	the	the	DET
easat-5050	240	3	future	future	NOUN
easat-5050	240	4	,	,	PUNCT
easat-5050	240	5	further	further	ADJ
easat-5050	240	6	improvements	improvement	NOUN
easat-5050	240	7	to	to	ADP
easat-5050	240	8	the	the	DET
easat-5050	240	9	network	network	NOUN
easat-5050	240	10	will	will	AUX
easat-5050	240	11	be	be	AUX
easat-5050	240	12	required	require	VERB
easat-5050	240	13	to	to	PART
easat-5050	240	14	enhance	enhance	VERB
easat-5050	240	15	its	its	PRON
easat-5050	240	16	practical	practical	ADJ
easat-5050	240	17	applications	application	NOUN
easat-5050	240	18	.	.	PUNCT
easat-5050	241	1	transparency	transparency	NOUN
easat-5050	241	2	:	:	PUNCT
easat-5050	241	3	the	the	DET
easat-5050	241	4	authors	author	NOUN
easat-5050	241	5	confirm	confirm	VERB
easat-5050	241	6	that	that	SCONJ
easat-5050	241	7	the	the	DET
easat-5050	241	8	manuscript	manuscript	NOUN
easat-5050	241	9	is	be	AUX
easat-5050	241	10	an	an	DET
easat-5050	241	11	honest	honest	ADJ
easat-5050	241	12	,	,	PUNCT
easat-5050	241	13	accurate	accurate	ADJ
easat-5050	241	14	,	,	PUNCT
easat-5050	241	15	and	and	CCONJ
easat-5050	241	16	transparent	transparent	ADJ
easat-5050	241	17	account	account	NOUN
easat-5050	241	18	of	of	ADP
easat-5050	241	19	the	the	DET
easat-5050	241	20	study	study	NOUN
easat-5050	241	21	;	;	PUNCT
easat-5050	241	22	that	that	SCONJ
easat-5050	241	23	no	no	DET
easat-5050	241	24	vital	vital	ADJ
easat-5050	241	25	features	feature	NOUN
easat-5050	241	26	of	of	ADP
easat-5050	241	27	the	the	DET
easat-5050	241	28	study	study	NOUN
easat-5050	241	29	have	have	AUX
easat-5050	241	30	been	be	AUX
easat-5050	241	31	omitted	omit	VERB
easat-5050	241	32	;	;	PUNCT
easat-5050	241	33	and	and	CCONJ
easat-5050	241	34	that	that	SCONJ
easat-5050	241	35	any	any	DET
easat-5050	241	36	discrepancies	discrepancy	NOUN
easat-5050	241	37	from	from	ADP
easat-5050	241	38	the	the	DET
easat-5050	241	39	study	study	NOUN
easat-5050	241	40	as	as	SCONJ
easat-5050	241	41	planned	plan	VERB
easat-5050	241	42	have	have	AUX
easat-5050	241	43	been	be	AUX
easat-5050	241	44	explained	explain	VERB
easat-5050	241	45	.	.	PUNCT
easat-5050	242	1	this	this	DET
easat-5050	242	2	study	study	NOUN
easat-5050	242	3	followed	follow	VERB
easat-5050	242	4	all	all	DET
easat-5050	242	5	ethical	ethical	ADJ
easat-5050	242	6	practices	practice	NOUN
easat-5050	242	7	during	during	ADP
easat-5050	242	8	writing	writing	NOUN
easat-5050	242	9	.	.	PUNCT
easat-5050	243	1	acknowledgements	acknowledgement	NOUN
easat-5050	243	2	:	:	PUNCT
easat-5050	243	3	this	this	DET
easat-5050	243	4	work	work	NOUN
easat-5050	243	5	was	be	AUX
easat-5050	243	6	supported	support	VERB
easat-5050	243	7	by	by	ADP
easat-5050	243	8	the	the	DET
easat-5050	243	9	natural	natural	ADJ
easat-5050	243	10	science	science	NOUN
easat-5050	243	11	key	key	NOUN
easat-5050	243	12	program	program	NOUN
easat-5050	243	13	of	of	ADP
easat-5050	243	14	anhui	anhui	PROPN
easat-5050	243	15	province	province	PROPN
easat-5050	243	16	(	(	PUNCT
easat-5050	243	17	2022ah052792	2022ah052792	NUM
easat-5050	243	18	)	)	PUNCT
easat-5050	243	19	,	,	PUNCT
easat-5050	243	20	the	the	DET
easat-5050	243	21	excellent	excellent	ADJ
easat-5050	243	22	young	young	ADJ
easat-5050	243	23	talents	talent	NOUN
easat-5050	243	24	support	support	VERB
easat-5050	243	25	program	program	NOUN
easat-5050	243	26	of	of	ADP
easat-5050	243	27	anhui	anhui	PROPN
easat-5050	243	28	province	province	PROPN
easat-5050	243	29	university	university	PROPN
easat-5050	243	30	(	(	PUNCT
easat-5050	243	31	gxyq2022274	gxyq2022274	PROPN
easat-5050	243	32	)	)	PUNCT
easat-5050	243	33	,	,	PUNCT
easat-5050	243	34	and	and	CCONJ
easat-5050	243	35	universiti	universiti	PROPN
easat-5050	243	36	malaysia	malaysia	PROPN
easat-5050	243	37	sarawak	sarawak	PROPN
easat-5050	243	38	,	,	PUNCT
easat-5050	243	39	malaysia	malaysia	PROPN
easat-5050	243	40	.	.	PUNCT
easat-5050	244	1	2145	2145	NUM
easat-5050	244	2	edelweiss	edelweiss	PROPN
easat-5050	244	3	applied	apply	VERB
easat-5050	244	4	science	science	NOUN
easat-5050	244	5	and	and	CCONJ
easat-5050	244	6	technology	technology	NOUN
easat-5050	244	7	issn	issn	PROPN
easat-5050	244	8	:	:	PUNCT
easat-5050	244	9	2576	2576	NUM
easat-5050	244	10	-	-	SYM
easat-5050	244	11	8484	8484	NUM
easat-5050	244	12	vol	vol	NOUN
easat-5050	244	13	.	.	PROPN
easat-5050	245	1	9	9	NUM
easat-5050	245	2	,	,	PUNCT
easat-5050	245	3	no	no	INTJ
easat-5050	245	4	.	.	NOUN
easat-5050	245	5	2	2	NUM
easat-5050	245	6	:	:	SYM
easat-5050	245	7	2132	2132	NUM
easat-5050	245	8	-	-	SYM
easat-5050	245	9	2146	2146	NUM
easat-5050	245	10	,	,	PUNCT
easat-5050	245	11	2025	2025	NUM
easat-5050	245	12	doi	doi	NOUN
easat-5050	245	13	:	:	PUNCT
easat-5050	245	14	10.55214/25768484.v9i2.5050	10.55214/25768484.v9i2.5050	NUM
easat-5050	245	15	©	©	PROPN
easat-5050	245	16	2025	2025	NUM
easat-5050	245	17	by	by	ADP
easat-5050	245	18	the	the	DET
easat-5050	245	19	authors	author	NOUN
easat-5050	245	20	;	;	PUNCT
easat-5050	246	1	licensee	licensee	PROPN
easat-5050	246	2	learning	learn	VERB
easat-5050	246	3	gate	gate	NOUN
easat-5050	246	4	copyright	copyright	NOUN
easat-5050	246	5	:	:	PUNCT
easat-5050	246	6	©	©	PROPN
easat-5050	246	7	2025	2025	NUM
easat-5050	246	8	by	by	ADP
easat-5050	246	9	the	the	DET
easat-5050	246	10	authors	author	NOUN
easat-5050	246	11	.	.	PUNCT
easat-5050	247	1	this	this	DET
easat-5050	247	2	open	open	ADJ
easat-5050	247	3	-	-	PUNCT
easat-5050	247	4	access	access	NOUN
easat-5050	247	5	article	article	NOUN
easat-5050	247	6	is	be	AUX
easat-5050	247	7	distributed	distribute	VERB
easat-5050	247	8	under	under	ADP
easat-5050	247	9	the	the	DET
easat-5050	247	10	terms	term	NOUN
easat-5050	247	11	and	and	CCONJ
easat-5050	247	12	conditions	condition	NOUN
easat-5050	247	13	of	of	ADP
easat-5050	247	14	the	the	DET
easat-5050	247	15	creative	creative	ADJ
easat-5050	247	16	commons	common	NOUN
easat-5050	247	17	attribution	attribution	NOUN
easat-5050	247	18	(	(	PUNCT
easat-5050	247	19	cc	cc	NOUN
easat-5050	247	20	by	by	ADP
easat-5050	247	21	)	)	PUNCT
easat-5050	247	22	license	license	NOUN
easat-5050	247	23	(	(	PUNCT
easat-5050	247	24	https://creativecommons.org/licenses/by/4.0/	https://creativecommons.org/licenses/by/4.0/	PROPN
easat-5050	247	25	)	)	PUNCT
easat-5050	247	26	.	.	PUNCT
easat-5050	248	1	references	reference	NOUN
easat-5050	248	2	achen	achen	NOUN
easat-5050	248	3	/	/	SYM
easat-5050	248	4	case	case	NOUN
easat-5050	248	5	.	.	PUNCT
easat-5050	249	1	(	(	PUNCT
easat-5050	249	2	2024	2024	NUM
easat-5050	249	3	)	)	PUNCT
easat-5050	249	4	.	.	PUNCT
easat-5050	250	1	achen	achen	PROPN
easat-5050	250	2	/	/	SYM
easat-5050	250	3	casewesternreserveuniversitydata	casewesternreserveuniversitydata	PROPN
easat-5050	250	4	’	'	PUNCT
easat-5050	250	5	,	,	PUNCT
easat-5050	250	6	gitee	gitee	NOUN
easat-5050	250	7	.	.	PUNCT
easat-5050	251	1	retrieved	retrieve	VERB
easat-5050	251	2	from	from	ADP
easat-5050	251	3	https://gitee.com/zthcool/casewesternreserveuniversitydata	https://gitee.com/zthcool/casewesternreserveuniversitydata	PROPN
easat-5050	251	4	bahdanau	bahdanau	PROPN
easat-5050	251	5	,	,	PUNCT
easat-5050	251	6	d.	d.	PROPN
easat-5050	251	7	,	,	PUNCT
easat-5050	251	8	cho	cho	PROPN
easat-5050	251	9	,	,	PUNCT
easat-5050	251	10	k.	k.	PROPN
easat-5050	251	11	,	,	PUNCT
easat-5050	251	12	&	&	CCONJ
easat-5050	251	13	bengio	bengio	PROPN
easat-5050	251	14	,	,	PUNCT
easat-5050	251	15	y.	y.	PROPN
easat-5050	251	16	(	(	PUNCT
easat-5050	251	17	2014	2014	NUM
easat-5050	251	18	)	)	PUNCT
easat-5050	251	19	.	.	PUNCT
easat-5050	252	1	neural	neural	ADJ
easat-5050	252	2	machine	machine	NOUN
easat-5050	252	3	translation	translation	NOUN
easat-5050	252	4	by	by	ADP
easat-5050	252	5	jointly	jointly	ADV
easat-5050	252	6	learning	learn	VERB
easat-5050	252	7	to	to	PART
easat-5050	252	8	align	align	VERB
easat-5050	252	9	and	and	CCONJ
easat-5050	252	10	translate	translate	VERB
easat-5050	252	11	.	.	PUNCT
easat-5050	253	1	arxiv	arxiv	PROPN
easat-5050	253	2	preprintarxiv:1409.0473	preprintarxiv:1409.0473	PROPN
easat-5050	253	3	.	.	PUNCT
easat-5050	254	1	bouras	boura	NOUN
easat-5050	254	2	,	,	PUNCT
easat-5050	254	3	a.	a.	NOUN
easat-5050	254	4	,	,	PUNCT
easat-5050	254	5	bennedjai	bennedjai	PROPN
easat-5050	254	6	,	,	PUNCT
easat-5050	254	7	s.	s.	PROPN
easat-5050	254	8	,	,	PUNCT
easat-5050	254	9	&	&	CCONJ
easat-5050	254	10	bouras	boura	NOUN
easat-5050	254	11	,	,	PUNCT
easat-5050	254	12	s.	s.	PROPN
easat-5050	254	13	(	(	PUNCT
easat-5050	254	14	2020	2020	NUM
easat-5050	254	15	)	)	PUNCT
easat-5050	254	16	.	.	PUNCT
easat-5050	255	1	experimental	experimental	ADJ
easat-5050	255	2	detection	detection	NOUN
easat-5050	255	3	of	of	ADP
easat-5050	255	4	defects	defect	NOUN
easat-5050	255	5	in	in	ADP
easat-5050	255	6	variable	variable	ADJ
easat-5050	255	7	speed	speed	NOUN
easat-5050	255	8	fan	fan	NOUN
easat-5050	255	9	bearing	bearing	NOUN
easat-5050	255	10	using	use	VERB
easat-5050	255	11	stator	stator	NOUN
easat-5050	255	12	current	current	ADJ
easat-5050	255	13	monitoring	monitoring	NOUN
easat-5050	255	14	.	.	PUNCT
easat-5050	256	1	sn	sn	PROPN
easat-5050	256	2	applied	apply	VERB
easat-5050	256	3	sciences	science	NOUN
easat-5050	256	4	,	,	PUNCT
easat-5050	256	5	2	2	NUM
easat-5050	256	6	,	,	PUNCT
easat-5050	256	7	1	1	NUM
easat-5050	256	8	-	-	SYM
easat-5050	256	9	8	8	NUM
easat-5050	256	10	.	.	PUNCT
easat-5050	257	1	https://doi.org/10.1007/s42452-020-2687-2	https://doi.org/10.1007/s42452-020-2687-2	NUM
easat-5050	257	2	bousseksou	bousseksou	PROPN
easat-5050	257	3	,	,	PUNCT
easat-5050	257	4	r.	r.	PROPN
easat-5050	257	5	,	,	PUNCT
easat-5050	257	6	bessous	bessous	ADJ
easat-5050	257	7	,	,	PUNCT
easat-5050	257	8	n.	n.	NOUN
easat-5050	257	9	,	,	PUNCT
easat-5050	257	10	&	&	CCONJ
easat-5050	257	11	mahmoud	mahmoud	PROPN
easat-5050	257	12	,	,	PUNCT
easat-5050	257	13	m.	m.	NOUN
easat-5050	257	14	m.	m.	NOUN
easat-5050	257	15	(	(	PUNCT
easat-5050	257	16	2024	2024	NUM
easat-5050	257	17	)	)	PUNCT
easat-5050	257	18	.	.	PUNCT
easat-5050	258	1	diagnosis	diagnosis	NOUN
easat-5050	258	2	of	of	ADP
easat-5050	258	3	bearing	bear	VERB
easat-5050	258	4	faults	fault	NOUN
easat-5050	258	5	using	use	VERB
easat-5050	258	6	an	an	DET
easat-5050	258	7	advanced	advanced	ADJ
easat-5050	258	8	feature	feature	NOUN
easat-5050	258	9	extraction	extraction	NOUN
easat-5050	258	10	based	base	VERB
easat-5050	258	11	on	on	ADP
easat-5050	258	12	wpt	wpt	PROPN
easat-5050	258	13	and	and	CCONJ
easat-5050	258	14	eev	eev	PROPN
easat-5050	258	15	methods	method	NOUN
easat-5050	258	16	.	.	PUNCT
easat-5050	259	1	studies	study	NOUN
easat-5050	259	2	in	in	ADP
easat-5050	259	3	engineering	engineering	NOUN
easat-5050	259	4	and	and	CCONJ
easat-5050	259	5	exact	exact	ADJ
easat-5050	259	6	sciences	science	NOUN
easat-5050	259	7	,	,	PUNCT
easat-5050	259	8	5(2	5(2	NUM
easat-5050	259	9	)	)	PUNCT
easat-5050	259	10	,	,	PUNCT
easat-5050	259	11	e12305	e12305	PROPN
easat-5050	259	12	-	-	PUNCT
easat-5050	259	13	e12305	e12305	PROPN
easat-5050	259	14	.	.	PUNCT
easat-5050	260	1	https://doi.org/10.54021/seesv5n2-809	https://doi.org/10.54021/seesv5n2-809	PROPN
easat-5050	260	2	cerrada	cerrada	PROPN
easat-5050	260	3	,	,	PUNCT
easat-5050	260	4	m.	m.	NOUN
easat-5050	260	5	,	,	PUNCT
easat-5050	260	6	sánchez	sánchez	PROPN
easat-5050	260	7	,	,	PUNCT
easat-5050	260	8	r.-v	r.-v	PROPN
easat-5050	260	9	.	.	PROPN
easat-5050	260	10	,	,	PUNCT
easat-5050	260	11	li	li	PROPN
easat-5050	260	12	,	,	PUNCT
easat-5050	260	13	c.	c.	PROPN
easat-5050	260	14	,	,	PUNCT
easat-5050	260	15	pacheco	pacheco	PROPN
easat-5050	260	16	,	,	PUNCT
easat-5050	260	17	f.	f.	PROPN
easat-5050	260	18	,	,	PUNCT
easat-5050	260	19	cabrera	cabrera	PROPN
easat-5050	260	20	,	,	PUNCT
easat-5050	260	21	d.	d.	PROPN
easat-5050	260	22	,	,	PUNCT
easat-5050	260	23	de	de	PROPN
easat-5050	260	24	oliveira	oliveira	PROPN
easat-5050	260	25	,	,	PUNCT
easat-5050	260	26	j.	j.	PROPN
easat-5050	260	27	v.	v.	PROPN
easat-5050	260	28	,	,	PUNCT
easat-5050	260	29	&	&	CCONJ
easat-5050	260	30	vásquez	vásquez	PROPN
easat-5050	260	31	,	,	PUNCT
easat-5050	260	32	r.	r.	PROPN
easat-5050	260	33	e.	e.	PROPN
easat-5050	260	34	(	(	PUNCT
easat-5050	260	35	2018	2018	NUM
easat-5050	260	36	)	)	PUNCT
easat-5050	260	37	.	.	PUNCT
easat-5050	261	1	a	a	DET
easat-5050	261	2	review	review	NOUN
easat-5050	261	3	on	on	ADP
easat-5050	261	4	datadriven	datadriven	ADJ
easat-5050	261	5	fault	fault	NOUN
easat-5050	261	6	severity	severity	NOUN
easat-5050	261	7	assessment	assessment	NOUN
easat-5050	261	8	in	in	ADP
easat-5050	261	9	rolling	rolling	ADJ
easat-5050	261	10	bearings	bearing	NOUN
easat-5050	261	11	.	.	PUNCT
easat-5050	262	1	mechanical	mechanical	ADJ
easat-5050	262	2	systems	system	NOUN
easat-5050	262	3	and	and	CCONJ
easat-5050	262	4	signal	signal	PROPN
easat-5050	262	5	processing	processing	NOUN
easat-5050	262	6	,	,	PUNCT
easat-5050	262	7	99	99	NUM
easat-5050	262	8	,	,	PUNCT
easat-5050	262	9	169	169	NUM
easat-5050	262	10	-	-	SYM
easat-5050	262	11	196	196	NUM
easat-5050	262	12	.	.	PUNCT
easat-5050	263	1	https://doi.org/10.1016/j.ymssp.2017.06.012	https://doi.org/10.1016/j.ymssp.2017.06.012	VERB
easat-5050	263	2	chen	chen	PROPN
easat-5050	263	3	,	,	PUNCT
easat-5050	263	4	g.	g.	PROPN
easat-5050	263	5	,	,	PUNCT
easat-5050	263	6	liu	liu	PROPN
easat-5050	263	7	,	,	PUNCT
easat-5050	263	8	m.	m.	NOUN
easat-5050	263	9	,	,	PUNCT
easat-5050	263	10	&	&	CCONJ
easat-5050	263	11	chen	chen	PROPN
easat-5050	263	12	,	,	PUNCT
easat-5050	263	13	j.	j.	PROPN
easat-5050	263	14	(	(	PUNCT
easat-5050	263	15	2020	2020	NUM
easat-5050	263	16	)	)	PUNCT
easat-5050	263	17	.	.	PUNCT
easat-5050	264	1	frequency	frequency	NOUN
easat-5050	264	2	-	-	PUNCT
easat-5050	264	3	temporal	temporal	ADJ
easat-5050	264	4	-	-	PUNCT
easat-5050	264	5	logic	logic	NOUN
easat-5050	264	6	-	-	PUNCT
easat-5050	264	7	based	base	VERB
easat-5050	264	8	bearing	bearing	NOUN
easat-5050	264	9	fault	fault	NOUN
easat-5050	264	10	diagnosis	diagnosis	NOUN
easat-5050	264	11	and	and	CCONJ
easat-5050	264	12	fault	fault	VERB
easat-5050	264	13	interpretation	interpretation	NOUN
easat-5050	264	14	using	use	VERB
easat-5050	264	15	bayesian	bayesian	NOUN
easat-5050	264	16	optimization	optimization	NOUN
easat-5050	264	17	with	with	ADP
easat-5050	264	18	bayesian	bayesian	NOUN
easat-5050	264	19	neural	neural	ADJ
easat-5050	264	20	networks	network	NOUN
easat-5050	264	21	.	.	PUNCT
easat-5050	265	1	mechanical	mechanical	ADJ
easat-5050	265	2	systems	system	NOUN
easat-5050	265	3	and	and	CCONJ
easat-5050	265	4	signal	signal	NOUN
easat-5050	265	5	processing	processing	NOUN
easat-5050	265	6	,	,	PUNCT
easat-5050	265	7	145	145	NUM
easat-5050	265	8	,	,	PUNCT
easat-5050	265	9	106951	106951	NUM
easat-5050	265	10	.	.	PUNCT
easat-5050	266	1	https://doi.org/10.1016/j.ymssp.2020.106951	https://doi.org/10.1016/j.ymssp.2020.106951	PROPN
easat-5050	266	2	chen	chen	PROPN
easat-5050	266	3	,	,	PUNCT
easat-5050	266	4	h.-y	h.-y	NOUN
easat-5050	266	5	.	.	PUNCT
easat-5050	266	6	,	,	PUNCT
easat-5050	266	7	&	&	CCONJ
easat-5050	266	8	lee	lee	PROPN
easat-5050	266	9	,	,	PUNCT
easat-5050	266	10	c.-h	c.-h	PROPN
easat-5050	266	11	.	.	PUNCT
easat-5050	267	1	(	(	PUNCT
easat-5050	267	2	2020	2020	NUM
easat-5050	267	3	)	)	PUNCT
easat-5050	267	4	.	.	PUNCT
easat-5050	268	1	vibration	vibration	NOUN
easat-5050	268	2	signals	signal	VERB
easat-5050	268	3	analysis	analysis	NOUN
easat-5050	268	4	by	by	ADP
easat-5050	268	5	explainable	explainable	ADJ
easat-5050	268	6	artificial	artificial	ADJ
easat-5050	268	7	intelligence	intelligence	NOUN
easat-5050	268	8	(	(	PUNCT
easat-5050	268	9	xai	xai	PROPN
easat-5050	268	10	)	)	PUNCT
easat-5050	268	11	approach	approach	NOUN
easat-5050	268	12	:	:	PUNCT
easat-5050	268	13	application	application	NOUN
easat-5050	268	14	on	on	ADP
easat-5050	268	15	bearing	bear	VERB
easat-5050	268	16	faults	fault	NOUN
easat-5050	268	17	diagnosis	diagnosis	NOUN
easat-5050	268	18	.	.	PUNCT
easat-5050	269	1	ieee	ieee	NOUN
easat-5050	269	2	access	access	NOUN
easat-5050	269	3	,	,	PUNCT
easat-5050	269	4	8	8	NUM
easat-5050	269	5	,	,	PUNCT
easat-5050	269	6	134246	134246	NUM
easat-5050	269	7	-	-	SYM
easat-5050	269	8	134256	134256	NUM
easat-5050	269	9	.	.	PUNCT
easat-5050	270	1	https://doi.org/10.1109/access.2020.3006491	https://doi.org/10.1109/access.2020.3006491	PROPN
easat-5050	270	2	chen	chen	PROPN
easat-5050	270	3	,	,	PUNCT
easat-5050	270	4	z.	z.	PROPN
easat-5050	270	5	,	,	PUNCT
easat-5050	270	6	wang	wang	PROPN
easat-5050	270	7	,	,	PUNCT
easat-5050	270	8	h.	h.	PROPN
easat-5050	270	9	,	,	PUNCT
easat-5050	270	10	zhou	zhou	PROPN
easat-5050	270	11	,	,	PUNCT
easat-5050	270	12	y.	y.	PROPN
easat-5050	270	13	,	,	PUNCT
easat-5050	270	14	yang	yang	PROPN
easat-5050	270	15	,	,	PUNCT
easat-5050	270	16	y.	y.	PROPN
easat-5050	270	17	,	,	PUNCT
easat-5050	270	18	&	&	CCONJ
easat-5050	270	19	liu	liu	PROPN
easat-5050	270	20	,	,	PUNCT
easat-5050	270	21	y.	y.	PROPN
easat-5050	270	22	(	(	PUNCT
easat-5050	270	23	2024	2024	NUM
easat-5050	270	24	)	)	PUNCT
easat-5050	270	25	.	.	PUNCT
easat-5050	271	1	a	a	DET
easat-5050	271	2	multi	multi	ADJ
easat-5050	271	3	-	-	ADJ
easat-5050	271	4	scale	scale	ADJ
easat-5050	271	5	feature	feature	NOUN
easat-5050	271	6	extraction	extraction	NOUN
easat-5050	271	7	and	and	CCONJ
easat-5050	271	8	fusion	fusion	NOUN
easat-5050	271	9	method	method	NOUN
easat-5050	271	10	for	for	ADP
easat-5050	271	11	diagnosing	diagnose	VERB
easat-5050	271	12	bearing	bear	VERB
easat-5050	271	13	faults	fault	NOUN
easat-5050	271	14	.	.	PUNCT
easat-5050	272	1	journal	journal	PROPN
easat-5050	272	2	of	of	ADP
easat-5050	272	3	dynamics	dynamic	NOUN
easat-5050	272	4	,	,	PUNCT
easat-5050	272	5	monitoring	monitoring	NOUN
easat-5050	272	6	and	and	CCONJ
easat-5050	272	7	diagnostics	diagnostic	NOUN
easat-5050	272	8	,	,	PUNCT
easat-5050	272	9	3(4	3(4	NUM
easat-5050	272	10	)	)	PUNCT
easat-5050	272	11	,	,	PUNCT
easat-5050	272	12	268	268	NUM
easat-5050	272	13	-	-	SYM
easat-5050	272	14	278	278	NUM
easat-5050	272	15	.	.	PUNCT
easat-5050	273	1	https://doi.org/10.37965/jdmd.2024.560	https://doi.org/10.37965/jdmd.2024.560	PROPN
easat-5050	273	2	dai	dai	PROPN
easat-5050	273	3	,	,	PUNCT
easat-5050	273	4	w.	w.	PROPN
easat-5050	273	5	,	,	PUNCT
easat-5050	273	6	mo	mo	PROPN
easat-5050	273	7	,	,	PUNCT
easat-5050	273	8	z.	z.	PROPN
easat-5050	273	9	,	,	PUNCT
easat-5050	273	10	luo	luo	PROPN
easat-5050	273	11	,	,	PUNCT
easat-5050	273	12	c.	c.	PROPN
easat-5050	273	13	,	,	PUNCT
easat-5050	273	14	jiang	jiang	PROPN
easat-5050	273	15	,	,	PUNCT
easat-5050	273	16	j.	j.	PROPN
easat-5050	273	17	,	,	PUNCT
easat-5050	273	18	zhang	zhang	PROPN
easat-5050	273	19	,	,	PUNCT
easat-5050	273	20	h.	h.	PROPN
easat-5050	273	21	,	,	PUNCT
easat-5050	273	22	&	&	CCONJ
easat-5050	273	23	miao	miao	PROPN
easat-5050	273	24	,	,	PUNCT
easat-5050	273	25	q.	q.	PROPN
easat-5050	273	26	(	(	PUNCT
easat-5050	273	27	2020	2020	NUM
easat-5050	273	28	)	)	PUNCT
easat-5050	273	29	.	.	PUNCT
easat-5050	274	1	fault	fault	VERB
easat-5050	274	2	diagnosis	diagnosis	NOUN
easat-5050	274	3	of	of	ADP
easat-5050	274	4	rotating	rotate	VERB
easat-5050	274	5	machinery	machinery	NOUN
easat-5050	274	6	based	base	VERB
easat-5050	274	7	on	on	ADP
easat-5050	274	8	deep	deep	ADJ
easat-5050	274	9	reinforcement	reinforcement	NOUN
easat-5050	274	10	learning	learning	NOUN
easat-5050	274	11	and	and	CCONJ
easat-5050	274	12	reciprocal	reciprocal	ADJ
easat-5050	274	13	of	of	ADP
easat-5050	274	14	smoothness	smoothness	ADJ
easat-5050	274	15	index	index	NOUN
easat-5050	274	16	.	.	PUNCT
easat-5050	275	1	ieee	ieee	PROPN
easat-5050	275	2	sensors	sensor	NOUN
easat-5050	275	3	journal	journal	PROPN
easat-5050	275	4	,	,	PUNCT
easat-5050	275	5	20(15	20(15	NUM
easat-5050	275	6	)	)	PUNCT
easat-5050	275	7	,	,	PUNCT
easat-5050	275	8	8307	8307	NUM
easat-5050	275	9	-	-	SYM
easat-5050	275	10	8315	8315	NUM
easat-5050	275	11	.	.	PUNCT
easat-5050	276	1	https://doi.org/10.1109/jsen.2020.2970747	https://doi.org/10.1109/jsen.2020.2970747	PROPN
easat-5050	276	2	duong	duong	PROPN
easat-5050	276	3	,	,	PUNCT
easat-5050	276	4	b.	b.	PROPN
easat-5050	276	5	p.	p.	PROPN
easat-5050	276	6	,	,	PUNCT
easat-5050	276	7	&	&	CCONJ
easat-5050	276	8	kim	kim	PROPN
easat-5050	276	9	,	,	PUNCT
easat-5050	276	10	j.-m	j.-m	PROPN
easat-5050	276	11	.	.	PUNCT
easat-5050	277	1	(	(	PUNCT
easat-5050	277	2	2018	2018	NUM
easat-5050	277	3	)	)	PUNCT
easat-5050	277	4	.	.	PUNCT
easat-5050	278	1	non	non	ADJ
easat-5050	278	2	-	-	ADJ
easat-5050	278	3	mutually	mutually	ADV
easat-5050	278	4	exclusive	exclusive	ADJ
easat-5050	278	5	deep	deep	ADJ
easat-5050	278	6	neural	neural	ADJ
easat-5050	278	7	network	network	NOUN
easat-5050	278	8	classifier	classifier	NOUN
easat-5050	278	9	for	for	ADP
easat-5050	278	10	combined	combined	ADJ
easat-5050	278	11	modes	mode	NOUN
easat-5050	278	12	of	of	ADP
easat-5050	278	13	bearing	bear	VERB
easat-5050	278	14	fault	fault	NOUN
easat-5050	278	15	diagnosis	diagnosis	NOUN
easat-5050	278	16	.	.	PUNCT
easat-5050	279	1	sensors	sensor	NOUN
easat-5050	279	2	,	,	PUNCT
easat-5050	279	3	18(4	18(4	NUM
easat-5050	279	4	)	)	PUNCT
easat-5050	279	5	,	,	PUNCT
easat-5050	279	6	1129	1129	NUM
easat-5050	279	7	.	.	PUNCT
easat-5050	280	1	https://doi.org/10.3390/s18041129	https://doi.org/10.3390/s18041129	PROPN
easat-5050	280	2	eang	eang	PROPN
easat-5050	280	3	,	,	PUNCT
easat-5050	280	4	c.	c.	PROPN
easat-5050	280	5	,	,	PUNCT
easat-5050	280	6	&	&	CCONJ
easat-5050	280	7	lee	lee	PROPN
easat-5050	280	8	,	,	PUNCT
easat-5050	280	9	s.	s.	PROPN
easat-5050	280	10	(	(	PUNCT
easat-5050	280	11	2024	2024	NUM
easat-5050	280	12	)	)	PUNCT
easat-5050	280	13	.	.	PUNCT
easat-5050	281	1	predictive	predictive	ADJ
easat-5050	281	2	maintenance	maintenance	NOUN
easat-5050	281	3	and	and	CCONJ
easat-5050	281	4	fault	fault	VERB
easat-5050	281	5	detection	detection	NOUN
easat-5050	281	6	for	for	ADP
easat-5050	281	7	motor	motor	NOUN
easat-5050	281	8	drive	drive	NOUN
easat-5050	281	9	control	control	NOUN
easat-5050	281	10	systems	system	NOUN
easat-5050	281	11	in	in	ADP
easat-5050	281	12	industrial	industrial	ADJ
easat-5050	281	13	robots	robot	NOUN
easat-5050	281	14	using	use	VERB
easat-5050	281	15	cnn	cnn	PROPN
easat-5050	281	16	-	-	PUNCT
easat-5050	281	17	rnn	rnn	PROPN
easat-5050	281	18	-	-	PUNCT
easat-5050	281	19	based	base	VERB
easat-5050	281	20	observers	observer	NOUN
easat-5050	281	21	.	.	PUNCT
easat-5050	282	1	sensors	sensor	NOUN
easat-5050	282	2	,	,	PUNCT
easat-5050	282	3	25(1	25(1	NUM
easat-5050	282	4	)	)	PUNCT
easat-5050	282	5	,	,	PUNCT
easat-5050	282	6	25	25	NUM
easat-5050	282	7	.	.	PUNCT
easat-5050	283	1	https://doi.org/10.3390/s25010025	https://doi.org/10.3390/s25010025	NUM
easat-5050	283	2	eren	eren	PROPN
easat-5050	283	3	,	,	PUNCT
easat-5050	283	4	l.	l.	PROPN
easat-5050	283	5	,	,	PUNCT
easat-5050	283	6	ince	ince	NOUN
easat-5050	283	7	,	,	PUNCT
easat-5050	283	8	t.	t.	PROPN
easat-5050	283	9	,	,	PUNCT
easat-5050	283	10	&	&	CCONJ
easat-5050	283	11	kiranyaz	kiranyaz	PROPN
easat-5050	283	12	,	,	PUNCT
easat-5050	283	13	s.	s.	PROPN
easat-5050	283	14	(	(	PUNCT
easat-5050	283	15	2019	2019	NUM
easat-5050	283	16	)	)	PUNCT
easat-5050	283	17	.	.	PUNCT
easat-5050	284	1	a	a	DET
easat-5050	284	2	generic	generic	ADJ
easat-5050	284	3	intelligent	intelligent	ADJ
easat-5050	284	4	bearing	bearing	NOUN
easat-5050	284	5	fault	fault	NOUN
easat-5050	284	6	diagnosis	diagnosis	NOUN
easat-5050	284	7	system	system	NOUN
easat-5050	284	8	using	use	VERB
easat-5050	284	9	compact	compact	ADJ
easat-5050	284	10	adaptive	adaptive	ADJ
easat-5050	284	11	1d	1d	NUM
easat-5050	284	12	cnn	cnn	PROPN
easat-5050	284	13	classifier	classifier	NOUN
easat-5050	284	14	.	.	PUNCT
easat-5050	285	1	journal	journal	PROPN
easat-5050	285	2	of	of	ADP
easat-5050	285	3	signal	signal	NOUN
easat-5050	285	4	processing	processing	NOUN
easat-5050	285	5	systems	system	NOUN
easat-5050	285	6	,	,	PUNCT
easat-5050	285	7	91(2	91(2	NUM
easat-5050	285	8	)	)	PUNCT
easat-5050	285	9	,	,	PUNCT
easat-5050	285	10	179	179	NUM
easat-5050	285	11	-	-	SYM
easat-5050	285	12	189	189	NUM
easat-5050	285	13	.	.	PUNCT
easat-5050	286	1	https://doi.org/10.1007/s11265-018-1378-3	https://doi.org/10.1007/s11265-018-1378-3	NUM
easat-5050	286	2	fan	fan	PROPN
easat-5050	286	3	,	,	PUNCT
easat-5050	286	4	f.	f.	PROPN
easat-5050	286	5	,	,	PUNCT
easat-5050	286	6	xiong	xiong	PROPN
easat-5050	286	7	,	,	PUNCT
easat-5050	286	8	j.	j.	PROPN
easat-5050	286	9	,	,	PUNCT
easat-5050	286	10	&	&	CCONJ
easat-5050	286	11	wang	wang	PROPN
easat-5050	286	12	,	,	PUNCT
easat-5050	286	13	g.	g.	PROPN
easat-5050	286	14	(	(	PUNCT
easat-5050	286	15	2020	2020	NUM
easat-5050	286	16	)	)	PUNCT
easat-5050	286	17	.	.	PUNCT
easat-5050	287	1	universal	universal	ADJ
easat-5050	287	2	approximation	approximation	NOUN
easat-5050	287	3	with	with	ADP
easat-5050	287	4	quadratic	quadratic	ADJ
easat-5050	287	5	deep	deep	ADJ
easat-5050	287	6	networks	network	NOUN
easat-5050	287	7	.	.	PUNCT
easat-5050	288	1	neural	neural	ADJ
easat-5050	288	2	networks	network	NOUN
easat-5050	288	3	,	,	PUNCT
easat-5050	288	4	124	124	NUM
easat-5050	288	5	,	,	PUNCT
easat-5050	288	6	383	383	NUM
easat-5050	288	7	-	-	SYM
easat-5050	288	8	392	392	NUM
easat-5050	288	9	.	.	PUNCT
easat-5050	288	10	https://doi.org/10.1016/j.neunet.2020.01.007	https://doi.org/10.1016/j.neunet.2020.01.007	NOUN
easat-5050	288	11	gunerkar	gunerkar	NOUN
easat-5050	288	12	,	,	PUNCT
easat-5050	288	13	r.	r.	PROPN
easat-5050	288	14	s.	s.	PROPN
easat-5050	288	15	,	,	PUNCT
easat-5050	288	16	jalan	jalan	PROPN
easat-5050	288	17	,	,	PUNCT
easat-5050	288	18	a.	a.	PROPN
easat-5050	288	19	k.	k.	PROPN
easat-5050	288	20	,	,	PUNCT
easat-5050	288	21	&	&	CCONJ
easat-5050	288	22	belgamwar	belgamwar	PROPN
easat-5050	288	23	,	,	PUNCT
easat-5050	288	24	s.	s.	PROPN
easat-5050	288	25	u.	u.	PROPN
easat-5050	288	26	(	(	PUNCT
easat-5050	288	27	2019	2019	NUM
easat-5050	288	28	)	)	PUNCT
easat-5050	288	29	.	.	PUNCT
easat-5050	289	1	fault	fault	VERB
easat-5050	289	2	diagnosis	diagnosis	NOUN
easat-5050	289	3	of	of	ADP
easat-5050	289	4	rolling	rolling	ADJ
easat-5050	289	5	element	element	NOUN
easat-5050	289	6	bearing	bearing	NOUN
easat-5050	289	7	based	base	VERB
easat-5050	289	8	on	on	ADP
easat-5050	289	9	artificial	artificial	ADJ
easat-5050	289	10	neural	neural	ADJ
easat-5050	289	11	network	network	NOUN
easat-5050	289	12	.	.	PUNCT
easat-5050	290	1	journal	journal	PROPN
easat-5050	290	2	of	of	ADP
easat-5050	290	3	mechanical	mechanical	ADJ
easat-5050	290	4	science	science	NOUN
easat-5050	290	5	and	and	CCONJ
easat-5050	290	6	technology	technology	NOUN
easat-5050	290	7	,	,	PUNCT
easat-5050	290	8	33	33	NUM
easat-5050	290	9	,	,	PUNCT
easat-5050	290	10	505	505	NUM
easat-5050	290	11	-	-	SYM
easat-5050	290	12	511	511	NUM
easat-5050	290	13	.	.	PUNCT
easat-5050	291	1	https://doi.org/10.1007/s12206-019-0103-x	https://doi.org/10.1007/s12206-019-0103-x	PROPN
easat-5050	291	2	hu	hu	PROPN
easat-5050	291	3	,	,	PUNCT
easat-5050	291	4	j.	j.	PROPN
easat-5050	291	5	,	,	PUNCT
easat-5050	291	6	shen	shen	PROPN
easat-5050	291	7	,	,	PUNCT
easat-5050	291	8	l.	l.	PROPN
easat-5050	291	9	,	,	PUNCT
easat-5050	291	10	&	&	CCONJ
easat-5050	291	11	sun	sun	PROPN
easat-5050	291	12	,	,	PUNCT
easat-5050	291	13	g.	g.	PROPN
easat-5050	291	14	(	(	PUNCT
easat-5050	291	15	2018	2018	NUM
easat-5050	291	16	)	)	PUNCT
easat-5050	291	17	.	.	PUNCT
easat-5050	292	1	squeeze	squeeze	NOUN
easat-5050	292	2	-	-	PUNCT
easat-5050	292	3	and	and	CCONJ
easat-5050	292	4	-	-	PUNCT
easat-5050	292	5	excitation	excitation	NOUN
easat-5050	292	6	networks	network	NOUN
easat-5050	292	7	.	.	PUNCT
easat-5050	293	1	paper	paper	NOUN
easat-5050	293	2	presented	present	VERB
easat-5050	293	3	at	at	ADP
easat-5050	293	4	the	the	DET
easat-5050	293	5	proceedings	proceeding	NOUN
easat-5050	293	6	of	of	ADP
easat-5050	293	7	the	the	DET
easat-5050	293	8	ieee	ieee	NOUN
easat-5050	293	9	conference	conference	NOUN
easat-5050	293	10	on	on	ADP
easat-5050	293	11	computer	computer	NOUN
easat-5050	293	12	vision	vision	NOUN
easat-5050	293	13	and	and	CCONJ
easat-5050	293	14	pattern	pattern	NOUN
easat-5050	293	15	recognition	recognition	NOUN
easat-5050	293	16	.	.	PUNCT
easat-5050	294	1	jalayer	jalayer	PROPN
easat-5050	294	2	,	,	PUNCT
easat-5050	294	3	r.	r.	PROPN
easat-5050	294	4	,	,	PUNCT
easat-5050	294	5	jalayer	jalayer	PROPN
easat-5050	294	6	,	,	PUNCT
easat-5050	294	7	m.	m.	NOUN
easat-5050	294	8	,	,	PUNCT
easat-5050	294	9	mor	mor	PROPN
easat-5050	294	10	,	,	PUNCT
easat-5050	294	11	a.	a.	NOUN
easat-5050	294	12	,	,	PUNCT
easat-5050	294	13	orsenigo	orsenigo	PROPN
easat-5050	294	14	,	,	PUNCT
easat-5050	294	15	c.	c.	PROPN
easat-5050	294	16	,	,	PUNCT
easat-5050	294	17	&	&	CCONJ
easat-5050	294	18	vercellis	vercellis	PROPN
easat-5050	294	19	,	,	PUNCT
easat-5050	294	20	c.	c.	PROPN
easat-5050	294	21	(	(	PUNCT
easat-5050	294	22	2024	2024	NUM
easat-5050	294	23	)	)	PUNCT
easat-5050	294	24	.	.	PUNCT
easat-5050	295	1	evaluating	evaluate	VERB
easat-5050	295	2	deep	deep	ADJ
easat-5050	295	3	learning	learning	NOUN
easat-5050	295	4	models	model	NOUN
easat-5050	295	5	for	for	ADP
easat-5050	295	6	fault	fault	NOUN
easat-5050	295	7	diagnosis	diagnosis	NOUN
easat-5050	295	8	of	of	ADP
easat-5050	295	9	a	a	DET
easat-5050	295	10	rotating	rotate	VERB
easat-5050	295	11	machinery	machinery	NOUN
easat-5050	295	12	with	with	ADP
easat-5050	295	13	epistemic	epistemic	ADJ
easat-5050	295	14	and	and	CCONJ
easat-5050	295	15	aleatoric	aleatoric	ADJ
easat-5050	295	16	uncertainty	uncertainty	NOUN
easat-5050	295	17	.	.	PUNCT
easat-5050	296	1	arxiv	arxiv	PROPN
easat-5050	296	2	preprint	preprint	NOUN
easat-5050	296	3	arxiv:2412.18980	arxiv:2412.18980	NOUN
easat-5050	296	4	.	.	PUNCT
easat-5050	297	1	https://doi.org/10.48550/arxiv.2412.18980	https://doi.org/10.48550/arxiv.2412.18980	PROPN
easat-5050	297	2	liao	liao	PROPN
easat-5050	297	3	,	,	PUNCT
easat-5050	297	4	j.-x	j.-x	NOUN
easat-5050	297	5	.	.	PUNCT
easat-5050	297	6	,	,	PUNCT
easat-5050	297	7	dong	dong	PROPN
easat-5050	297	8	,	,	PUNCT
easat-5050	297	9	h.-c	h.-c	PROPN
easat-5050	297	10	.	.	PUNCT
easat-5050	297	11	,	,	PUNCT
easat-5050	297	12	sun	sun	PROPN
easat-5050	297	13	,	,	PUNCT
easat-5050	297	14	z.-q	z.-q	PROPN
easat-5050	297	15	.	.	PROPN
easat-5050	297	16	,	,	PUNCT
easat-5050	297	17	sun	sun	PROPN
easat-5050	297	18	,	,	PUNCT
easat-5050	297	19	j.	j.	PROPN
easat-5050	297	20	,	,	PUNCT
easat-5050	297	21	zhang	zhang	PROPN
easat-5050	297	22	,	,	PUNCT
easat-5050	297	23	s.	s.	PROPN
easat-5050	297	24	,	,	PUNCT
easat-5050	297	25	&	&	CCONJ
easat-5050	297	26	fan	fan	PROPN
easat-5050	297	27	,	,	PUNCT
easat-5050	297	28	f.-l	f.-l	PROPN
easat-5050	297	29	.	.	PUNCT
easat-5050	297	30	(	(	PUNCT
easat-5050	297	31	2023	2023	NUM
easat-5050	297	32	)	)	PUNCT
easat-5050	297	33	.	.	PUNCT
easat-5050	298	1	attention	attention	NOUN
easat-5050	298	2	-	-	PUNCT
easat-5050	298	3	embedded	embed	VERB
easat-5050	298	4	quadratic	quadratic	ADJ
easat-5050	298	5	network	network	NOUN
easat-5050	298	6	(	(	PUNCT
easat-5050	298	7	qttention	qttention	NOUN
easat-5050	298	8	)	)	PUNCT
easat-5050	298	9	for	for	ADP
easat-5050	298	10	effective	effective	ADJ
easat-5050	298	11	and	and	CCONJ
easat-5050	298	12	interpretable	interpretable	ADJ
easat-5050	298	13	bearing	bear	VERB
easat-5050	298	14	fault	fault	NOUN
easat-5050	298	15	diagnosis	diagnosis	NOUN
easat-5050	298	16	.	.	PUNCT
easat-5050	299	1	ieee	ieee	NOUN
easat-5050	299	2	transactions	transaction	NOUN
easat-5050	299	3	on	on	ADP
easat-5050	299	4	instrumentation	instrumentation	NOUN
easat-5050	299	5	and	and	CCONJ
easat-5050	299	6	measurement	measurement	NOUN
easat-5050	299	7	,	,	PUNCT
easat-5050	299	8	72	72	NUM
easat-5050	299	9	,	,	PUNCT
easat-5050	299	10	1	1	NUM
easat-5050	299	11	-	-	SYM
easat-5050	299	12	13	13	NUM
easat-5050	299	13	.	.	PUNCT
easat-5050	300	1	https://doi.org/10.1109/tim.2023.3259031	https://doi.org/10.1109/tim.2023.3259031	PROPN
easat-5050	300	2	liao	liao	PROPN
easat-5050	300	3	,	,	PUNCT
easat-5050	300	4	y.	y.	PROPN
easat-5050	300	5	,	,	PUNCT
easat-5050	300	6	li	li	PROPN
easat-5050	300	7	,	,	PUNCT
easat-5050	300	8	h.	h.	PROPN
easat-5050	300	9	,	,	PUNCT
easat-5050	300	10	zhang	zhang	PROPN
easat-5050	300	11	,	,	PUNCT
easat-5050	300	12	l.	l.	PROPN
easat-5050	300	13	,	,	PUNCT
easat-5050	300	14	yan	yan	PROPN
easat-5050	300	15	,	,	PUNCT
easat-5050	300	16	q.	q.	PROPN
easat-5050	300	17	,	,	PUNCT
easat-5050	300	18	yang	yang	PROPN
easat-5050	300	19	,	,	PUNCT
easat-5050	300	20	y.	y.	PROPN
easat-5050	300	21	,	,	PUNCT
easat-5050	300	22	&	&	CCONJ
easat-5050	300	23	li	li	PROPN
easat-5050	300	24	,	,	PUNCT
easat-5050	300	25	b.	b.	PROPN
easat-5050	300	26	(	(	PUNCT
easat-5050	300	27	2023	2023	NUM
easat-5050	300	28	)	)	PUNCT
easat-5050	300	29	.	.	PUNCT
easat-5050	301	1	anomaly	anomaly	PROPN
easat-5050	301	2	transformer	transformer	NOUN
easat-5050	301	3	for	for	ADP
easat-5050	301	4	vibration	vibration	NOUN
easat-5050	301	5	data	datum	NOUN
easat-5050	301	6	analysis	analysis	NOUN
easat-5050	301	7	:	:	PUNCT
easat-5050	301	8	case	case	NOUN
easat-5050	301	9	study	study	NOUN
easat-5050	301	10	on	on	ADP
easat-5050	301	11	cwru	cwru	PROPN
easat-5050	301	12	dataset	dataset	PROPN
easat-5050	301	13	.	.	PUNCT
easat-5050	302	1	paper	paper	NOUN
easat-5050	302	2	presented	present	VERB
easat-5050	302	3	at	at	ADP
easat-5050	302	4	the	the	DET
easat-5050	302	5	2023	2023	NUM
easat-5050	302	6	international	international	ADJ
easat-5050	302	7	conference	conference	NOUN
easat-5050	302	8	on	on	ADP
easat-5050	302	9	image	image	NOUN
easat-5050	302	10	processing	processing	NOUN
easat-5050	302	11	,	,	PUNCT
easat-5050	302	12	computer	computer	NOUN
easat-5050	302	13	vision	vision	NOUN
easat-5050	302	14	and	and	CCONJ
easat-5050	302	15	machine	machine	NOUN
easat-5050	302	16	learning	learning	NOUN
easat-5050	302	17	(	(	PUNCT
easat-5050	302	18	icicml	icicml	NOUN
easat-5050	302	19	)	)	PUNCT
easat-5050	302	20	,	,	PUNCT
easat-5050	302	21	chengdu	chengdu	PROPN
easat-5050	302	22	,	,	PUNCT
easat-5050	302	23	china	china	PROPN
easat-5050	302	24	:	:	PUNCT
easat-5050	302	25	ieee	ieee	PROPN
easat-5050	302	26	.	.	PUNCT
easat-5050	303	1	https://doi.org/10.1109/icicml60161.2023.10424762	https://doi.org/10.1109/icicml60161.2023.10424762	PROPN
easat-5050	303	2	.	.	PUNCT
easat-5050	304	1	lu	lu	PROPN
easat-5050	304	2	,	,	PUNCT
easat-5050	304	3	q.	q.	PROPN
easat-5050	304	4	,	,	PUNCT
easat-5050	304	5	shen	shen	PROPN
easat-5050	304	6	,	,	PUNCT
easat-5050	304	7	x.	x.	PROPN
easat-5050	304	8	,	,	PUNCT
easat-5050	304	9	wang	wang	PROPN
easat-5050	304	10	,	,	PUNCT
easat-5050	304	11	x.	x.	PROPN
easat-5050	304	12	,	,	PUNCT
easat-5050	304	13	li	li	PROPN
easat-5050	304	14	,	,	PUNCT
easat-5050	304	15	m.	m.	NOUN
easat-5050	304	16	,	,	PUNCT
easat-5050	304	17	li	li	PROPN
easat-5050	304	18	,	,	PUNCT
easat-5050	304	19	j.	j.	PROPN
easat-5050	304	20	,	,	PUNCT
easat-5050	304	21	&	&	CCONJ
easat-5050	304	22	zhang	zhang	PROPN
easat-5050	304	23	,	,	PUNCT
easat-5050	304	24	m.	m.	NOUN
easat-5050	304	25	(	(	PUNCT
easat-5050	304	26	2021	2021	NUM
easat-5050	304	27	)	)	PUNCT
easat-5050	304	28	.	.	PUNCT
easat-5050	305	1	fault	fault	VERB
easat-5050	305	2	diagnosis	diagnosis	NOUN
easat-5050	305	3	of	of	ADP
easat-5050	305	4	rolling	rolling	ADJ
easat-5050	305	5	bearing	bearing	NOUN
easat-5050	305	6	based	base	VERB
easat-5050	305	7	on	on	ADP
easat-5050	305	8	improved	improved	ADJ
easat-5050	305	9	vmd	vmd	PROPN
easat-5050	305	10	and	and	CCONJ
easat-5050	305	11	knn	knn	PROPN
easat-5050	305	12	.	.	PROPN
easat-5050	306	1	mathematical	mathematical	ADJ
easat-5050	306	2	problems	problem	NOUN
easat-5050	306	3	in	in	ADP
easat-5050	306	4	engineering	engineering	NOUN
easat-5050	306	5	,	,	PUNCT
easat-5050	306	6	2021(1	2021(1	NUM
easat-5050	306	7	)	)	PUNCT
easat-5050	306	8	,	,	PUNCT
easat-5050	306	9	2530315	2530315	NUM
easat-5050	306	10	.	.	PUNCT
easat-5050	307	1	https://doi.org/10.1155/2021/2530315	https://doi.org/10.1155/2021/2530315	ADJ
easat-5050	307	2	ozcan	ozcan	PROPN
easat-5050	307	3	,	,	PUNCT
easat-5050	307	4	i.	i.	PROPN
easat-5050	307	5	h.	h.	PROPN
easat-5050	307	6	,	,	PUNCT
easat-5050	307	7	devecioglu	devecioglu	PROPN
easat-5050	307	8	,	,	PUNCT
easat-5050	307	9	o.	o.	PROPN
easat-5050	307	10	c.	c.	PROPN
easat-5050	307	11	,	,	PUNCT
easat-5050	307	12	ince	ince	NOUN
easat-5050	307	13	,	,	PUNCT
easat-5050	307	14	t.	t.	PROPN
easat-5050	307	15	,	,	PUNCT
easat-5050	307	16	eren	eren	PROPN
easat-5050	307	17	,	,	PUNCT
easat-5050	307	18	l.	l.	PROPN
easat-5050	307	19	,	,	PUNCT
easat-5050	307	20	&	&	CCONJ
easat-5050	307	21	askar	askar	NOUN
easat-5050	307	22	,	,	PUNCT
easat-5050	307	23	m.	m.	NOUN
easat-5050	307	24	(	(	PUNCT
easat-5050	307	25	2022	2022	NUM
easat-5050	307	26	)	)	PUNCT
easat-5050	307	27	.	.	PUNCT
easat-5050	308	1	enhanced	enhance	VERB
easat-5050	308	2	bearing	bear	VERB
easat-5050	308	3	fault	fault	NOUN
easat-5050	308	4	detection	detection	NOUN
easat-5050	308	5	using	use	VERB
easat-5050	308	6	multichannel	multichannel	NOUN
easat-5050	308	7	,	,	PUNCT
easat-5050	308	8	multilevel	multilevel	PROPN
easat-5050	308	9	1d	1d	NUM
easat-5050	308	10	cnn	cnn	PROPN
easat-5050	308	11	classifier	classifier	NOUN
easat-5050	308	12	.	.	PUNCT
easat-5050	309	1	electrical	electrical	ADJ
easat-5050	309	2	engineering	engineering	NOUN
easat-5050	309	3	,	,	PUNCT
easat-5050	309	4	104(2	104(2	NUM
easat-5050	309	5	)	)	PUNCT
easat-5050	309	6	,	,	PUNCT
easat-5050	309	7	435	435	NUM
easat-5050	309	8	-	-	SYM
easat-5050	309	9	447	447	NUM
easat-5050	309	10	.	.	PUNCT
easat-5050	310	1	https://doi.org/10.1007/s00202-021-01309-2	https://doi.org/10.1007/s00202-021-01309-2	NOUN
easat-5050	310	2	pouyet	pouyet	ADV
easat-5050	310	3	,	,	PUNCT
easat-5050	310	4	e.	e.	PROPN
easat-5050	310	5	,	,	PUNCT
easat-5050	310	6	rohani	rohani	PROPN
easat-5050	310	7	,	,	PUNCT
easat-5050	310	8	n.	n.	PROPN
easat-5050	310	9	,	,	PUNCT
easat-5050	310	10	katsaggelos	katsaggelo	NOUN
easat-5050	310	11	,	,	PUNCT
easat-5050	310	12	a.	a.	PROPN
easat-5050	310	13	k.	k.	PROPN
easat-5050	310	14	,	,	PUNCT
easat-5050	310	15	cossairt	cossairt	PROPN
easat-5050	310	16	,	,	PUNCT
easat-5050	310	17	o.	o.	PROPN
easat-5050	310	18	,	,	PUNCT
easat-5050	310	19	&	&	CCONJ
easat-5050	310	20	walton	walton	PROPN
easat-5050	310	21	,	,	PUNCT
easat-5050	310	22	m.	m.	NOUN
easat-5050	310	23	(	(	PUNCT
easat-5050	310	24	2018	2018	NUM
easat-5050	310	25	)	)	PUNCT
easat-5050	310	26	.	.	PUNCT
easat-5050	311	1	innovative	innovative	ADJ
easat-5050	311	2	data	datum	NOUN
easat-5050	311	3	reduction	reduction	NOUN
easat-5050	311	4	and	and	CCONJ
easat-5050	311	5	visualization	visualization	NOUN
easat-5050	311	6	strategy	strategy	NOUN
easat-5050	311	7	for	for	ADP
easat-5050	311	8	hyperspectral	hyperspectral	ADJ
easat-5050	311	9	imaging	imaging	NOUN
easat-5050	311	10	datasets	dataset	NOUN
easat-5050	311	11	using	use	VERB
easat-5050	311	12	t	t	PROPN
easat-5050	311	13	-	-	PUNCT
easat-5050	311	14	sne	sne	NOUN
easat-5050	311	15	approach	approach	NOUN
easat-5050	311	16	.	.	PUNCT
easat-5050	312	1	pure	pure	ADJ
easat-5050	312	2	and	and	CCONJ
easat-5050	312	3	applied	applied	ADJ
easat-5050	312	4	chemistry	chemistry	NOUN
easat-5050	312	5	,	,	PUNCT
easat-5050	312	6	90(3	90(3	NOUN
easat-5050	312	7	)	)	PUNCT
easat-5050	312	8	,	,	PUNCT
easat-5050	312	9	493	493	NUM
easat-5050	312	10	-	-	SYM
easat-5050	312	11	506	506	NUM
easat-5050	312	12	.	.	PUNCT
easat-5050	313	1	https://doi.org/10.1515/pac-2017-0907	https://doi.org/10.1515/pac-2017-0907	PROPN
easat-5050	313	2	qiu	qiu	PROPN
easat-5050	313	3	,	,	PUNCT
easat-5050	313	4	s.	s.	PROPN
easat-5050	313	5	,	,	PUNCT
easat-5050	313	6	cui	cui	PROPN
easat-5050	313	7	,	,	PUNCT
easat-5050	313	8	x.	x.	NOUN
easat-5050	313	9	,	,	PUNCT
easat-5050	313	10	ping	ping	PROPN
easat-5050	313	11	,	,	PUNCT
easat-5050	313	12	z.	z.	PROPN
easat-5050	313	13	,	,	PUNCT
easat-5050	313	14	shan	shan	PROPN
easat-5050	313	15	,	,	PUNCT
easat-5050	313	16	n.	n.	PROPN
easat-5050	313	17	,	,	PUNCT
easat-5050	313	18	li	li	PROPN
easat-5050	313	19	,	,	PUNCT
easat-5050	313	20	z.	z.	PROPN
easat-5050	313	21	,	,	PUNCT
easat-5050	313	22	bao	bao	PROPN
easat-5050	313	23	,	,	PUNCT
easat-5050	313	24	x.	x.	NOUN
easat-5050	313	25	,	,	PUNCT
easat-5050	313	26	&	&	CCONJ
easat-5050	313	27	xu	xu	PROPN
easat-5050	313	28	,	,	PUNCT
easat-5050	313	29	x.	x.	NOUN
easat-5050	313	30	(	(	PUNCT
easat-5050	313	31	2023	2023	NUM
easat-5050	313	32	)	)	PUNCT
easat-5050	313	33	.	.	PUNCT
easat-5050	314	1	deep	deep	ADJ
easat-5050	314	2	learning	learning	NOUN
easat-5050	314	3	techniques	technique	NOUN
easat-5050	314	4	in	in	ADP
easat-5050	314	5	intelligent	intelligent	ADJ
easat-5050	314	6	fault	fault	NOUN
easat-5050	314	7	diagnosis	diagnosis	NOUN
easat-5050	314	8	and	and	CCONJ
easat-5050	314	9	prognosis	prognosis	NOUN
easat-5050	314	10	for	for	ADP
easat-5050	314	11	industrial	industrial	ADJ
easat-5050	314	12	systems	system	NOUN
easat-5050	314	13	:	:	PUNCT
easat-5050	314	14	a	a	DET
easat-5050	314	15	review	review	NOUN
easat-5050	314	16	.	.	PUNCT
easat-5050	315	1	sensors	sensor	NOUN
easat-5050	315	2	,	,	PUNCT
easat-5050	315	3	23(3	23(3	NUM
easat-5050	315	4	)	)	PUNCT
easat-5050	315	5	,	,	PUNCT
easat-5050	315	6	1305	1305	NUM
easat-5050	315	7	.	.	PUNCT
easat-5050	316	1	https://doi.org/10.3390/s23031305	https://doi.org/10.3390/s23031305	NUM
easat-5050	316	2	https://creativecommons.org/licenses/by/4.0/	https://creativecommons.org/licenses/by/4.0/	PROPN
easat-5050	316	3	https://gitee.com/zthcool/casewesternreserveuniversitydata	https://gitee.com/zthcool/casewesternreserveuniversitydata	NOUN
easat-5050	316	4	https://doi.org/10.1007/s42452-020-2687-2	https://doi.org/10.1007/s42452-020-2687-2	NUM
easat-5050	316	5	https://doi.org/10.54021/seesv5n2-809	https://doi.org/10.54021/seesv5n2-809	X
easat-5050	316	6	https://doi.org/10.1016/j.ymssp.2017.06.012	https://doi.org/10.1016/j.ymssp.2017.06.012	VERB
easat-5050	316	7	https://doi.org/10.1016/j.ymssp.2020.106951	https://doi.org/10.1016/j.ymssp.2020.106951	NOUN
easat-5050	316	8	https://doi.org/10.1109/access.2020.3006491	https://doi.org/10.1109/access.2020.3006491	VERB
easat-5050	317	1	https://doi.org/10.37965/jdmd.2024.560	https://doi.org/10.37965/jdmd.2024.560	ADJ
easat-5050	317	2	https://doi.org/10.1109/jsen.2020.2970747	https://doi.org/10.1109/jsen.2020.2970747	PROPN
easat-5050	318	1	https://doi.org/10.3390/s18041129	https://doi.org/10.3390/s18041129	PROPN
easat-5050	319	1	https://doi.org/10.3390/s25010025	https://doi.org/10.3390/s25010025	NUM
easat-5050	319	2	https://doi.org/10.1007/s11265-018-1378-3	https://doi.org/10.1007/s11265-018-1378-3	NUM
easat-5050	319	3	https://doi.org/10.1016/j.neunet.2020.01.007	https://doi.org/10.1016/j.neunet.2020.01.007	NOUN
easat-5050	319	4	https://doi.org/10.1007/s12206-019-0103-x	https://doi.org/10.1007/s12206-019-0103-x	VERB
easat-5050	319	5	https://doi.org/10.48550/arxiv.2412.18980	https://doi.org/10.48550/arxiv.2412.18980	PROPN
easat-5050	319	6	https://doi.org/10.1109/tim.2023.3259031	https://doi.org/10.1109/tim.2023.3259031	NOUN
easat-5050	319	7	https://doi.org/10.1109/icicml60161.2023.10424762	https://doi.org/10.1109/icicml60161.2023.10424762	PROPN
easat-5050	319	8	https://doi.org/10.1155/2021/2530315	https://doi.org/10.1155/2021/2530315	PROPN
easat-5050	319	9	https://doi.org/10.1007/s00202-021-01309-2	https://doi.org/10.1007/s00202-021-01309-2	NOUN
easat-5050	319	10	https://doi.org/10.1515/pac-2017-0907	https://doi.org/10.1515/pac-2017-0907	VERB
easat-5050	319	11	https://doi.org/10.3390/s23031305	https://doi.org/10.3390/s23031305	NUM
easat-5050	319	12	2146	2146	NUM
easat-5050	319	13	edelweiss	edelweiss	PROPN
easat-5050	319	14	applied	apply	VERB
easat-5050	319	15	science	science	NOUN
easat-5050	319	16	and	and	CCONJ
easat-5050	319	17	technology	technology	NOUN
easat-5050	319	18	issn	issn	PROPN
easat-5050	319	19	:	:	PUNCT
easat-5050	319	20	2576	2576	NUM
easat-5050	319	21	-	-	SYM
easat-5050	319	22	8484	8484	NUM
easat-5050	319	23	vol	vol	NOUN
easat-5050	319	24	.	.	PROPN
easat-5050	320	1	9	9	NUM
easat-5050	320	2	,	,	PUNCT
easat-5050	320	3	no	no	INTJ
easat-5050	320	4	.	.	NOUN
easat-5050	320	5	2	2	NUM
easat-5050	320	6	:	:	SYM
easat-5050	320	7	2132	2132	NUM
easat-5050	320	8	-	-	SYM
easat-5050	320	9	2146	2146	NUM
easat-5050	320	10	,	,	PUNCT
easat-5050	320	11	2025	2025	NUM
easat-5050	320	12	doi	doi	NOUN
easat-5050	320	13	:	:	PUNCT
easat-5050	320	14	10.55214/25768484.v9i2.5050	10.55214/25768484.v9i2.5050	NUM
easat-5050	320	15	©	©	PROPN
easat-5050	320	16	2025	2025	NUM
easat-5050	320	17	by	by	ADP
easat-5050	320	18	the	the	DET
easat-5050	320	19	authors	author	NOUN
easat-5050	320	20	;	;	PUNCT
easat-5050	320	21	licensee	licensee	PROPN
easat-5050	320	22	learning	learning	PROPN
easat-5050	320	23	gate	gate	PROPN
easat-5050	320	24	rai	rai	PROPN
easat-5050	320	25	,	,	PUNCT
easat-5050	320	26	a.	a.	PROPN
easat-5050	320	27	,	,	PUNCT
easat-5050	320	28	&	&	CCONJ
easat-5050	320	29	upadhyay	upadhyay	PROPN
easat-5050	320	30	,	,	PUNCT
easat-5050	320	31	s.	s.	PROPN
easat-5050	320	32	h.	h.	PROPN
easat-5050	320	33	(	(	PUNCT
easat-5050	320	34	2016	2016	NUM
easat-5050	320	35	)	)	PUNCT
easat-5050	320	36	.	.	PUNCT
easat-5050	321	1	a	a	DET
easat-5050	321	2	review	review	NOUN
easat-5050	321	3	on	on	ADP
easat-5050	321	4	signal	signal	ADJ
easat-5050	321	5	processing	processing	NOUN
easat-5050	321	6	techniques	technique	NOUN
easat-5050	321	7	utilized	utilize	VERB
easat-5050	321	8	in	in	ADP
easat-5050	321	9	the	the	DET
easat-5050	321	10	fault	fault	NOUN
easat-5050	321	11	diagnosis	diagnosis	NOUN
easat-5050	321	12	of	of	ADP
easat-5050	321	13	rolling	rolling	ADJ
easat-5050	321	14	element	element	NOUN
easat-5050	321	15	bearings	bearing	NOUN
easat-5050	321	16	.	.	PUNCT
easat-5050	322	1	tribology	tribology	PROPN
easat-5050	322	2	international	international	PROPN
easat-5050	322	3	,	,	PUNCT
easat-5050	322	4	96	96	NUM
easat-5050	322	5	,	,	PUNCT
easat-5050	322	6	289	289	NUM
easat-5050	322	7	-	-	SYM
easat-5050	322	8	306	306	NUM
easat-5050	322	9	.	.	PUNCT
easat-5050	323	1	https://doi.org/10.1016/j.triboint.2015.12.037	https://doi.org/10.1016/j.triboint.2015.12.037	PROPN
easat-5050	323	2	rosa	rosa	PROPN
easat-5050	323	3	,	,	PUNCT
easat-5050	323	4	r.	r.	PROPN
easat-5050	323	5	k.	k.	PROPN
easat-5050	323	6	,	,	PUNCT
easat-5050	323	7	braga	braga	PROPN
easat-5050	323	8	,	,	PUNCT
easat-5050	323	9	d.	d.	PROPN
easat-5050	323	10	,	,	PUNCT
easat-5050	323	11	&	&	CCONJ
easat-5050	323	12	silva	silva	PROPN
easat-5050	323	13	,	,	PUNCT
easat-5050	323	14	d.	d.	PROPN
easat-5050	323	15	(	(	PUNCT
easat-5050	323	16	2024	2024	NUM
easat-5050	323	17	)	)	PUNCT
easat-5050	323	18	.	.	PUNCT
easat-5050	324	1	benchmarking	benchmarke	VERB
easat-5050	324	2	deep	deep	ADJ
easat-5050	324	3	learning	learning	NOUN
easat-5050	324	4	models	model	NOUN
easat-5050	324	5	for	for	ADP
easat-5050	324	6	bearing	bear	VERB
easat-5050	324	7	fault	fault	NOUN
easat-5050	324	8	diagnosis	diagnosis	NOUN
easat-5050	324	9	using	use	VERB
easat-5050	324	10	the	the	DET
easat-5050	324	11	cwru	cwru	NOUN
easat-5050	324	12	dataset	dataset	NOUN
easat-5050	324	13	:	:	PUNCT
easat-5050	324	14	a	a	DET
easat-5050	324	15	multi	multi	ADJ
easat-5050	324	16	-	-	ADJ
easat-5050	324	17	label	label	ADJ
easat-5050	324	18	approach	approach	NOUN
easat-5050	324	19	.	.	PUNCT
easat-5050	325	1	arxiv	arxiv	PROPN
easat-5050	325	2	preprint	preprint	VERB
easat-5050	325	3	arxiv:2407.14625	arxiv:2407.14625	NOUN
easat-5050	325	4	.	.	PUNCT
easat-5050	326	1	https://doi.org/10.48550/arxiv.2407.14625	https://doi.org/10.48550/arxiv.2407.14625	PROPN
easat-5050	326	2	shang	shang	PROPN
easat-5050	326	3	,	,	PUNCT
easat-5050	326	4	j.	j.	PROPN
easat-5050	326	5	,	,	PUNCT
easat-5050	326	6	chen	chen	PROPN
easat-5050	326	7	,	,	PUNCT
easat-5050	326	8	m.	m.	NOUN
easat-5050	326	9	,	,	PUNCT
easat-5050	326	10	ji	ji	PROPN
easat-5050	326	11	,	,	PUNCT
easat-5050	326	12	h.	h.	PROPN
easat-5050	326	13	,	,	PUNCT
easat-5050	326	14	zhou	zhou	PROPN
easat-5050	326	15	,	,	PUNCT
easat-5050	326	16	d.	d.	PROPN
easat-5050	326	17	,	,	PUNCT
easat-5050	326	18	zhang	zhang	PROPN
easat-5050	326	19	,	,	PUNCT
easat-5050	326	20	h.	h.	PROPN
easat-5050	326	21	,	,	PUNCT
easat-5050	326	22	&	&	CCONJ
easat-5050	326	23	li	li	PROPN
easat-5050	326	24	,	,	PUNCT
easat-5050	326	25	m.	m.	NOUN
easat-5050	326	26	(	(	PUNCT
easat-5050	326	27	2017	2017	NUM
easat-5050	326	28	)	)	PUNCT
easat-5050	326	29	.	.	PUNCT
easat-5050	327	1	dominant	dominant	ADJ
easat-5050	327	2	trend	trend	NOUN
easat-5050	327	3	based	base	VERB
easat-5050	327	4	logistic	logistic	ADJ
easat-5050	327	5	regression	regression	NOUN
easat-5050	327	6	for	for	ADP
easat-5050	327	7	fault	fault	NOUN
easat-5050	327	8	diagnosis	diagnosis	NOUN
easat-5050	327	9	in	in	ADP
easat-5050	327	10	nonstationary	nonstationary	ADJ
easat-5050	327	11	processes	process	NOUN
easat-5050	327	12	.	.	PUNCT
easat-5050	328	1	control	control	NOUN
easat-5050	328	2	engineering	engineering	NOUN
easat-5050	328	3	practice	practice	NOUN
easat-5050	328	4	,	,	PUNCT
easat-5050	328	5	66	66	NUM
easat-5050	328	6	,	,	PUNCT
easat-5050	328	7	156	156	NUM
easat-5050	328	8	-	-	SYM
easat-5050	328	9	168	168	NUM
easat-5050	328	10	.	.	PUNCT
easat-5050	329	1	https://doi.org/10.1016/j.conengprac.2017.06.011	https://doi.org/10.1016/j.conengprac.2017.06.011	PROPN
easat-5050	329	2	shenfield	shenfield	NOUN
easat-5050	329	3	,	,	PUNCT
easat-5050	329	4	a.	a.	NOUN
easat-5050	329	5	,	,	PUNCT
easat-5050	329	6	&	&	CCONJ
easat-5050	329	7	howarth	howarth	PROPN
easat-5050	329	8	,	,	PUNCT
easat-5050	329	9	m.	m.	NOUN
easat-5050	329	10	(	(	PUNCT
easat-5050	329	11	2020	2020	NUM
easat-5050	329	12	)	)	PUNCT
easat-5050	329	13	.	.	PUNCT
easat-5050	330	1	a	a	DET
easat-5050	330	2	novel	novel	ADJ
easat-5050	330	3	deep	deep	ADJ
easat-5050	330	4	learning	learning	NOUN
easat-5050	330	5	model	model	NOUN
easat-5050	330	6	for	for	ADP
easat-5050	330	7	the	the	DET
easat-5050	330	8	detection	detection	NOUN
easat-5050	330	9	and	and	CCONJ
easat-5050	330	10	identification	identification	NOUN
easat-5050	330	11	of	of	ADP
easat-5050	330	12	rolling	roll	VERB
easat-5050	330	13	elementbearing	elementbearing	NOUN
easat-5050	330	14	faults	fault	NOUN
easat-5050	330	15	.	.	PUNCT
easat-5050	331	1	sensors	sensor	NOUN
easat-5050	331	2	,	,	PUNCT
easat-5050	331	3	20(18	20(18	NUM
easat-5050	331	4	)	)	PUNCT
easat-5050	331	5	,	,	PUNCT
easat-5050	331	6	5112	5112	NUM
easat-5050	331	7	.	.	PUNCT
easat-5050	332	1	https://doi.org/10.3390/s20185112	https://doi.org/10.3390/s20185112	NUM
easat-5050	332	2	wan	wan	PROPN
easat-5050	332	3	,	,	PUNCT
easat-5050	332	4	l.	l.	PROPN
easat-5050	332	5	,	,	PUNCT
easat-5050	332	6	chen	chen	PROPN
easat-5050	332	7	,	,	PUNCT
easat-5050	332	8	y.	y.	PROPN
easat-5050	332	9	,	,	PUNCT
easat-5050	332	10	li	li	PROPN
easat-5050	332	11	,	,	PUNCT
easat-5050	332	12	h.	h.	PROPN
easat-5050	332	13	,	,	PUNCT
easat-5050	332	14	&	&	CCONJ
easat-5050	332	15	li	li	PROPN
easat-5050	332	16	,	,	PUNCT
easat-5050	332	17	c.	c.	PROPN
easat-5050	332	18	(	(	PUNCT
easat-5050	332	19	2020	2020	NUM
easat-5050	332	20	)	)	PUNCT
easat-5050	332	21	.	.	PUNCT
easat-5050	333	1	rolling	roll	VERB
easat-5050	333	2	-	-	PUNCT
easat-5050	333	3	element	element	NOUN
easat-5050	333	4	bearing	bear	VERB
easat-5050	333	5	fault	fault	NOUN
easat-5050	333	6	diagnosis	diagnosis	NOUN
easat-5050	333	7	using	use	VERB
easat-5050	333	8	improved	improve	VERB
easat-5050	333	9	lenet-5	lenet-5	NUM
easat-5050	333	10	network	network	NOUN
easat-5050	333	11	.	.	PUNCT
easat-5050	334	1	sensors	sensor	NOUN
easat-5050	334	2	,	,	PUNCT
easat-5050	334	3	20(6	20(6	NOUN
easat-5050	334	4	)	)	PUNCT
easat-5050	334	5	,	,	PUNCT
easat-5050	334	6	1693	1693	NUM
easat-5050	334	7	.	.	PUNCT
easat-5050	335	1	https://doi.org/10.3390/s20061693	https://doi.org/10.3390/s20061693	NUM
easat-5050	335	2	wang	wang	PROPN
easat-5050	335	3	,	,	PUNCT
easat-5050	335	4	h.	h.	PROPN
easat-5050	335	5	,	,	PUNCT
easat-5050	335	6	liu	liu	PROPN
easat-5050	335	7	,	,	PUNCT
easat-5050	335	8	z.	z.	PROPN
easat-5050	335	9	,	,	PUNCT
easat-5050	335	10	peng	peng	PROPN
easat-5050	335	11	,	,	PUNCT
easat-5050	335	12	d.	d.	PROPN
easat-5050	335	13	,	,	PUNCT
easat-5050	335	14	&	&	CCONJ
easat-5050	335	15	qin	qin	PROPN
easat-5050	335	16	,	,	PUNCT
easat-5050	335	17	y.	y.	PROPN
easat-5050	335	18	(	(	PUNCT
easat-5050	335	19	2019	2019	NUM
easat-5050	335	20	)	)	PUNCT
easat-5050	335	21	.	.	PUNCT
easat-5050	336	1	understanding	understanding	NOUN
easat-5050	336	2	and	and	CCONJ
easat-5050	336	3	learning	learn	VERB
easat-5050	336	4	discriminant	discriminant	NOUN
easat-5050	336	5	features	feature	NOUN
easat-5050	336	6	based	base	VERB
easat-5050	336	7	on	on	ADP
easat-5050	336	8	multiattention	multiattention	PROPN
easat-5050	336	9	1dcnn	1dcnn	NUM
easat-5050	336	10	for	for	ADP
easat-5050	336	11	wheelset	wheelset	NOUN
easat-5050	336	12	bearing	bear	VERB
easat-5050	336	13	fault	fault	NOUN
easat-5050	336	14	diagnosis	diagnosis	NOUN
easat-5050	336	15	.	.	PUNCT
easat-5050	337	1	ieee	ieee	NOUN
easat-5050	337	2	transactions	transaction	NOUN
easat-5050	337	3	on	on	ADP
easat-5050	337	4	industrial	industrial	ADJ
easat-5050	337	5	informatics	informatic	NOUN
easat-5050	337	6	,	,	PUNCT
easat-5050	337	7	16(9	16(9	NUM
easat-5050	337	8	)	)	PUNCT
easat-5050	337	9	,	,	PUNCT
easat-5050	337	10	5735	5735	NUM
easat-5050	337	11	-	-	SYM
easat-5050	337	12	5745	5745	NUM
easat-5050	337	13	.	.	PUNCT
easat-5050	338	1	https://doi.org/10.1109/tii.2019.2955540	https://doi.org/10.1109/tii.2019.2955540	PROPN
easat-5050	338	2	wang	wang	PROPN
easat-5050	338	3	,	,	PUNCT
easat-5050	338	4	h.	h.	PROPN
easat-5050	338	5	,	,	PUNCT
easat-5050	338	6	xu	xu	PROPN
easat-5050	338	7	,	,	PUNCT
easat-5050	338	8	j.	j.	PROPN
easat-5050	338	9	,	,	PUNCT
easat-5050	338	10	yan	yan	PROPN
easat-5050	338	11	,	,	PUNCT
easat-5050	338	12	r.	r.	PROPN
easat-5050	338	13	,	,	PUNCT
easat-5050	338	14	sun	sun	PROPN
easat-5050	338	15	,	,	PUNCT
easat-5050	338	16	c.	c.	PROPN
easat-5050	338	17	,	,	PUNCT
easat-5050	338	18	&	&	CCONJ
easat-5050	338	19	chen	chen	PROPN
easat-5050	338	20	,	,	PUNCT
easat-5050	338	21	x.	x.	PROPN
easat-5050	338	22	(	(	PUNCT
easat-5050	338	23	2020	2020	NUM
easat-5050	338	24	)	)	PUNCT
easat-5050	338	25	.	.	PUNCT
easat-5050	339	1	intelligent	intelligent	ADJ
easat-5050	339	2	bearing	bearing	NOUN
easat-5050	339	3	fault	fault	NOUN
easat-5050	339	4	diagnosis	diagnosis	NOUN
easat-5050	339	5	using	use	VERB
easat-5050	339	6	multi	multi	ADJ
easat-5050	339	7	-	-	ADJ
easat-5050	339	8	head	head	ADJ
easat-5050	339	9	attention	attention	NOUN
easat-5050	339	10	-	-	PUNCT
easat-5050	339	11	based	base	VERB
easat-5050	339	12	cnn	cnn	PROPN
easat-5050	339	13	.	.	PUNCT
easat-5050	340	1	procedia	procedia	PROPN
easat-5050	340	2	manufacturing	manufacturing	PROPN
easat-5050	340	3	,	,	PUNCT
easat-5050	340	4	49	49	NUM
easat-5050	340	5	,	,	PUNCT
easat-5050	340	6	112	112	NUM
easat-5050	340	7	-	-	SYM
easat-5050	340	8	118	118	NUM
easat-5050	340	9	.	.	PUNCT
easat-5050	341	1	https://doi.org/10.1016/j.promfg.2020.07.005	https://doi.org/10.1016/j.promfg.2020.07.005	PROPN
easat-5050	341	2	wang	wang	PROPN
easat-5050	341	3	,	,	PUNCT
easat-5050	341	4	m.	m.	NOUN
easat-5050	341	5	,	,	PUNCT
easat-5050	341	6	chen	chen	PROPN
easat-5050	341	7	,	,	PUNCT
easat-5050	341	8	y.	y.	PROPN
easat-5050	341	9	,	,	PUNCT
easat-5050	341	10	zhang	zhang	PROPN
easat-5050	341	11	,	,	PUNCT
easat-5050	341	12	x.	x.	PROPN
easat-5050	341	13	,	,	PUNCT
easat-5050	341	14	chau	chau	NOUN
easat-5050	341	15	,	,	PUNCT
easat-5050	341	16	t.	t.	PROPN
easat-5050	341	17	k.	k.	PROPN
easat-5050	341	18	,	,	PUNCT
easat-5050	341	19	ching	ching	PROPN
easat-5050	341	20	iu	iu	ADP
easat-5050	341	21	,	,	PUNCT
easat-5050	341	22	h.	h.	PROPN
easat-5050	341	23	h.	h.	PROPN
easat-5050	341	24	,	,	PUNCT
easat-5050	341	25	fernando	fernando	PROPN
easat-5050	341	26	,	,	PUNCT
easat-5050	341	27	t.	t.	PROPN
easat-5050	341	28	,	,	PUNCT
easat-5050	341	29	.	.	PUNCT
easat-5050	341	30	.	.	PUNCT
easat-5050	341	31	.	.	PUNCT
easat-5050	342	1	ma	ma	PROPN
easat-5050	342	2	,	,	PUNCT
easat-5050	342	3	m.	m.	NOUN
easat-5050	342	4	(	(	PUNCT
easat-5050	342	5	2021	2021	NUM
easat-5050	342	6	)	)	PUNCT
easat-5050	342	7	.	.	PUNCT
easat-5050	343	1	roller	roller	NOUN
easat-5050	343	2	bearing	bear	VERB
easat-5050	343	3	fault	fault	NOUN
easat-5050	343	4	diagnosis	diagnosis	NOUN
easat-5050	343	5	based	base	VERB
easat-5050	343	6	on	on	ADP
easat-5050	343	7	integrated	integrate	VERB
easat-5050	343	8	fault	fault	NOUN
easat-5050	343	9	feature	feature	NOUN
easat-5050	343	10	and	and	CCONJ
easat-5050	343	11	svm	svm	ADJ
easat-5050	343	12	.	.	PROPN
easat-5050	343	13	journal	journal	PROPN
easat-5050	343	14	of	of	ADP
easat-5050	343	15	vibration	vibration	NOUN
easat-5050	343	16	engineering	engineering	NOUN
easat-5050	343	17	&	&	CCONJ
easat-5050	343	18	technologies	technology	NOUN
easat-5050	343	19	,	,	PUNCT
easat-5050	343	20	10(3	10(3	NUM
easat-5050	343	21	)	)	PUNCT
easat-5050	343	22	,	,	PUNCT
easat-5050	343	23	853–862	853–862	NUM
easat-5050	343	24	.	.	PUNCT
easat-5050	344	1	https://doi.org/10.1007/s42417-021-00414-7	https://doi.org/10.1007/s42417-021-00414-7	NUM
easat-5050	344	2	wang	wang	PROPN
easat-5050	344	3	,	,	PUNCT
easat-5050	344	4	t.	t.	PROPN
easat-5050	344	5	,	,	PUNCT
easat-5050	344	6	liang	liang	PROPN
easat-5050	344	7	,	,	PUNCT
easat-5050	344	8	m.	m.	NOUN
easat-5050	344	9	,	,	PUNCT
easat-5050	344	10	li	li	PROPN
easat-5050	344	11	,	,	PUNCT
easat-5050	344	12	j.	j.	PROPN
easat-5050	344	13	,	,	PUNCT
easat-5050	344	14	&	&	CCONJ
easat-5050	344	15	cheng	cheng	PROPN
easat-5050	344	16	,	,	PUNCT
easat-5050	344	17	w.	w.	PROPN
easat-5050	344	18	(	(	PUNCT
easat-5050	344	19	2014	2014	NUM
easat-5050	344	20	)	)	PUNCT
easat-5050	344	21	.	.	PUNCT
easat-5050	345	1	rolling	roll	VERB
easat-5050	345	2	element	element	NOUN
easat-5050	345	3	bearing	bear	VERB
easat-5050	345	4	fault	fault	NOUN
easat-5050	345	5	diagnosis	diagnosis	NOUN
easat-5050	345	6	via	via	ADP
easat-5050	345	7	fault	fault	NOUN
easat-5050	345	8	characteristic	characteristic	ADJ
easat-5050	345	9	order	order	NOUN
easat-5050	345	10	(	(	PUNCT
easat-5050	345	11	fco	fco	PROPN
easat-5050	345	12	)	)	PUNCT
easat-5050	345	13	analysis	analysis	NOUN
easat-5050	345	14	.	.	PUNCT
easat-5050	346	1	mechanical	mechanical	ADJ
easat-5050	346	2	systems	system	NOUN
easat-5050	346	3	and	and	CCONJ
easat-5050	346	4	signal	signal	PROPN
easat-5050	346	5	processing	processing	NOUN
easat-5050	346	6	,	,	PUNCT
easat-5050	346	7	45(1	45(1	NOUN
easat-5050	346	8	)	)	PUNCT
easat-5050	346	9	,	,	PUNCT
easat-5050	346	10	139	139	NUM
easat-5050	346	11	-	-	SYM
easat-5050	346	12	153	153	NUM
easat-5050	346	13	.	.	PUNCT
easat-5050	347	1	https://doi.org/10.1016/j.ymssp.2013.11.011	https://doi.org/10.1016/j.ymssp.2013.11.011	NOUN
easat-5050	347	2	wang	wang	PROPN
easat-5050	347	3	,	,	PUNCT
easat-5050	347	4	y.	y.	PROPN
easat-5050	347	5	,	,	PUNCT
easat-5050	347	6	ding	ding	NOUN
easat-5050	347	7	,	,	PUNCT
easat-5050	347	8	x.	x.	PROPN
easat-5050	347	9	,	,	PUNCT
easat-5050	347	10	zeng	zeng	PROPN
easat-5050	347	11	,	,	PUNCT
easat-5050	347	12	q.	q.	PROPN
easat-5050	347	13	,	,	PUNCT
easat-5050	347	14	wang	wang	PROPN
easat-5050	347	15	,	,	PUNCT
easat-5050	347	16	l.	l.	PROPN
easat-5050	347	17	,	,	PUNCT
easat-5050	347	18	&	&	CCONJ
easat-5050	347	19	shao	shao	PROPN
easat-5050	347	20	,	,	PUNCT
easat-5050	347	21	y.	y.	PROPN
easat-5050	347	22	(	(	PUNCT
easat-5050	347	23	2020	2020	NUM
easat-5050	347	24	)	)	PUNCT
easat-5050	347	25	.	.	PUNCT
easat-5050	348	1	intelligent	intelligent	ADJ
easat-5050	348	2	rolling	rolling	ADJ
easat-5050	348	3	bearing	bear	VERB
easat-5050	348	4	fault	fault	NOUN
easat-5050	348	5	diagnosis	diagnosis	NOUN
easat-5050	348	6	via	via	ADP
easat-5050	348	7	vision	vision	NOUN
easat-5050	348	8	convnet	convnet	PROPN
easat-5050	348	9	.	.	PUNCT
easat-5050	349	1	ieee	ieee	PROPN
easat-5050	349	2	sensors	sensor	NOUN
easat-5050	349	3	journal	journal	PROPN
easat-5050	349	4	,	,	PUNCT
easat-5050	349	5	21(5	21(5	NUM
easat-5050	349	6	)	)	PUNCT
easat-5050	349	7	,	,	PUNCT
easat-5050	349	8	6600	6600	NUM
easat-5050	349	9	-	-	SYM
easat-5050	349	10	6609	6609	NUM
easat-5050	349	11	.	.	PUNCT
easat-5050	350	1	https://doi.org/10.1109/jsen.2020.3042182	https://doi.org/10.1109/jsen.2020.3042182	PROPN
easat-5050	350	2	xu	xu	PROPN
easat-5050	350	3	,	,	PUNCT
easat-5050	350	4	g.	g.	PROPN
easat-5050	350	5	,	,	PUNCT
easat-5050	350	6	liu	liu	PROPN
easat-5050	350	7	,	,	PUNCT
easat-5050	350	8	m.	m.	NOUN
easat-5050	350	9	,	,	PUNCT
easat-5050	350	10	jiang	jiang	PROPN
easat-5050	350	11	,	,	PUNCT
easat-5050	350	12	z.	z.	PROPN
easat-5050	350	13	,	,	PUNCT
easat-5050	350	14	söffker	söffker	PROPN
easat-5050	350	15	,	,	PUNCT
easat-5050	350	16	d.	d.	PROPN
easat-5050	350	17	,	,	PUNCT
easat-5050	350	18	&	&	CCONJ
easat-5050	350	19	shen	shen	PROPN
easat-5050	350	20	,	,	PUNCT
easat-5050	350	21	w.	w.	PROPN
easat-5050	350	22	(	(	PUNCT
easat-5050	350	23	2019	2019	NUM
easat-5050	350	24	)	)	PUNCT
easat-5050	350	25	.	.	PUNCT
easat-5050	351	1	bearing	bear	VERB
easat-5050	351	2	fault	fault	NOUN
easat-5050	351	3	diagnosis	diagnosis	NOUN
easat-5050	351	4	method	method	NOUN
easat-5050	351	5	based	base	VERB
easat-5050	351	6	on	on	ADP
easat-5050	351	7	deep	deep	ADJ
easat-5050	351	8	convolutional	convolutional	ADJ
easat-5050	351	9	neural	neural	ADJ
easat-5050	351	10	network	network	NOUN
easat-5050	351	11	and	and	CCONJ
easat-5050	351	12	random	random	ADJ
easat-5050	351	13	forest	forest	NOUN
easat-5050	351	14	ensemble	ensemble	ADJ
easat-5050	351	15	learning	learning	NOUN
easat-5050	351	16	.	.	PUNCT
easat-5050	352	1	sensors	sensor	NOUN
easat-5050	352	2	,	,	PUNCT
easat-5050	352	3	19(5	19(5	NUM
easat-5050	352	4	)	)	PUNCT
easat-5050	352	5	,	,	PUNCT
easat-5050	352	6	1088	1088	NUM
easat-5050	352	7	.	.	PUNCT
easat-5050	353	1	https://doi.org/10.3390/s19051088	https://doi.org/10.3390/s19051088	NUM
easat-5050	353	2	yoo	yoo	PROPN
easat-5050	353	3	,	,	PUNCT
easat-5050	353	4	y.	y.	PROPN
easat-5050	353	5	,	,	PUNCT
easat-5050	353	6	jo	jo	PROPN
easat-5050	353	7	,	,	PUNCT
easat-5050	353	8	h.	h.	PROPN
easat-5050	353	9	,	,	PUNCT
easat-5050	353	10	&	&	CCONJ
easat-5050	353	11	ban	ban	PROPN
easat-5050	353	12	,	,	PUNCT
easat-5050	353	13	s.-w	s.-w	PROPN
easat-5050	353	14	.	.	PUNCT
easat-5050	353	15	(	(	PUNCT
easat-5050	353	16	2023	2023	NUM
easat-5050	353	17	)	)	PUNCT
easat-5050	353	18	.	.	PUNCT
easat-5050	354	1	lite	lite	NOUN
easat-5050	354	2	and	and	CCONJ
easat-5050	354	3	efficient	efficient	ADJ
easat-5050	354	4	deep	deep	ADJ
easat-5050	354	5	learning	learning	NOUN
easat-5050	354	6	model	model	NOUN
easat-5050	354	7	for	for	ADP
easat-5050	354	8	bearing	bear	VERB
easat-5050	354	9	fault	fault	NOUN
easat-5050	354	10	diagnosis	diagnosis	NOUN
easat-5050	354	11	using	use	VERB
easat-5050	354	12	the	the	DET
easat-5050	354	13	cwru	cwru	NOUN
easat-5050	354	14	dataset	dataset	NOUN
easat-5050	354	15	.	.	PUNCT
easat-5050	355	1	sensors	sensor	NOUN
easat-5050	355	2	,	,	PUNCT
easat-5050	355	3	23(6	23(6	NUM
easat-5050	355	4	)	)	PUNCT
easat-5050	355	5	,	,	PUNCT
easat-5050	355	6	3157	3157	NUM
easat-5050	355	7	.	.	PUNCT
easat-5050	356	1	https://doi.org/10.3390/s23063157	https://doi.org/10.3390/s23063157	PROPN
easat-5050	356	2	zhang	zhang	PROPN
easat-5050	356	3	,	,	PUNCT
easat-5050	356	4	w.	w.	PROPN
easat-5050	356	5	,	,	PUNCT
easat-5050	356	6	peng	peng	PROPN
easat-5050	356	7	,	,	PUNCT
easat-5050	356	8	g.	g.	PROPN
easat-5050	356	9	,	,	PUNCT
easat-5050	356	10	li	li	PROPN
easat-5050	356	11	,	,	PUNCT
easat-5050	356	12	c.	c.	PROPN
easat-5050	356	13	,	,	PUNCT
easat-5050	356	14	chen	chen	PROPN
easat-5050	356	15	,	,	PUNCT
easat-5050	356	16	y.	y.	PROPN
easat-5050	356	17	,	,	PUNCT
easat-5050	356	18	&	&	CCONJ
easat-5050	356	19	zhang	zhang	PROPN
easat-5050	356	20	,	,	PUNCT
easat-5050	356	21	z.	z.	PROPN
easat-5050	356	22	(	(	PUNCT
easat-5050	356	23	2017	2017	NUM
easat-5050	356	24	)	)	PUNCT
easat-5050	356	25	.	.	PUNCT
easat-5050	357	1	a	a	DET
easat-5050	357	2	new	new	ADJ
easat-5050	357	3	deep	deep	ADJ
easat-5050	357	4	learning	learning	NOUN
easat-5050	357	5	model	model	NOUN
easat-5050	357	6	for	for	ADP
easat-5050	357	7	fault	fault	NOUN
easat-5050	357	8	diagnosis	diagnosis	NOUN
easat-5050	357	9	with	with	ADP
easat-5050	357	10	good	good	ADJ
easat-5050	357	11	anti	anti	ADJ
easat-5050	357	12	-	-	ADJ
easat-5050	357	13	noise	noise	ADJ
easat-5050	357	14	and	and	CCONJ
easat-5050	357	15	domain	domain	NOUN
easat-5050	357	16	adaptation	adaptation	NOUN
easat-5050	357	17	ability	ability	NOUN
easat-5050	357	18	on	on	ADP
easat-5050	357	19	raw	raw	ADJ
easat-5050	357	20	vibration	vibration	NOUN
easat-5050	357	21	signals	signal	NOUN
easat-5050	357	22	.	.	PUNCT
easat-5050	358	1	sensors	sensor	NOUN
easat-5050	358	2	,	,	PUNCT
easat-5050	358	3	17(2	17(2	NUM
easat-5050	358	4	)	)	PUNCT
easat-5050	358	5	,	,	PUNCT
easat-5050	358	6	425	425	NUM
easat-5050	358	7	.	.	PUNCT
easat-5050	359	1	https://doi.org/10.3390/s17020425	https://doi.org/10.3390/s17020425	NUM
easat-5050	359	2	zhao	zhao	PROPN
easat-5050	359	3	,	,	PUNCT
easat-5050	359	4	s.	s.	PROPN
easat-5050	359	5	,	,	PUNCT
easat-5050	359	6	zhu	zhu	PROPN
easat-5050	359	7	,	,	PUNCT
easat-5050	359	8	z.	z.	PROPN
easat-5050	359	9	,	,	PUNCT
easat-5050	359	10	&	&	CCONJ
easat-5050	359	11	he	he	PRON
easat-5050	359	12	,	,	PUNCT
easat-5050	359	13	j.	j.	PROPN
easat-5050	359	14	(	(	PUNCT
easat-5050	359	15	2024	2024	NUM
easat-5050	359	16	)	)	PUNCT
easat-5050	359	17	.	.	PUNCT
easat-5050	360	1	research	research	NOUN
easat-5050	360	2	and	and	CCONJ
easat-5050	360	3	optimization	optimization	NOUN
easat-5050	360	4	of	of	ADP
easat-5050	360	5	feature	feature	NOUN
easat-5050	360	6	extraction	extraction	NOUN
easat-5050	360	7	algorithms	algorithm	NOUN
easat-5050	360	8	based	base	VERB
easat-5050	360	9	on	on	ADP
easat-5050	360	10	bearing	bear	VERB
easat-5050	360	11	vibration	vibration	NOUN
easat-5050	360	12	signals	signal	NOUN
easat-5050	360	13	.	.	PUNCT
easat-5050	361	1	paper	paper	NOUN
easat-5050	361	2	presented	present	VERB
easat-5050	361	3	at	at	ADP
easat-5050	361	4	the	the	DET
easat-5050	361	5	international	international	ADJ
easat-5050	361	6	conference	conference	NOUN
easat-5050	361	7	on	on	ADP
easat-5050	361	8	cloud	cloud	NOUN
easat-5050	361	9	computing	computing	NOUN
easat-5050	361	10	and	and	CCONJ
easat-5050	361	11	communication	communication	NOUN
easat-5050	361	12	engineering	engineering	NOUN
easat-5050	361	13	(	(	PUNCT
easat-5050	361	14	ccce	ccce	NOUN
easat-5050	361	15	2024	2024	NUM
easat-5050	361	16	)	)	PUNCT
easat-5050	361	17	,	,	PUNCT
easat-5050	361	18	y.	y.	PROPN
easat-5050	361	19	yue	yue	PROPN
easat-5050	361	20	and	and	CCONJ
easat-5050	361	21	h.	h.	PROPN
easat-5050	361	22	xing	xing	PROPN
easat-5050	361	23	,	,	PUNCT
easat-5050	361	24	eds	eds	PROPN
easat-5050	361	25	.	.	PROPN
easat-5050	361	26	,	,	PUNCT
easat-5050	361	27	nanjing	nanjing	PROPN
easat-5050	361	28	,	,	PUNCT
easat-5050	361	29	china	china	PROPN
easat-5050	361	30	:	:	PUNCT
easat-5050	361	31	spie	spie	ADJ
easat-5050	361	32	.	.	PUNCT
easat-5050	361	33	https://doi.org/10.1117/12.3049948	https://doi.org/10.1117/12.3049948	PROPN
easat-5050	361	34	.	.	PUNCT
easat-5050	362	1	zhong	zhong	PROPN
easat-5050	362	2	,	,	PUNCT
easat-5050	362	3	x.	x.	PROPN
easat-5050	362	4	,	,	PUNCT
easat-5050	362	5	wang	wang	PROPN
easat-5050	362	6	,	,	PUNCT
easat-5050	362	7	f.	f.	PROPN
easat-5050	362	8	,	,	PUNCT
easat-5050	362	9	&	&	CCONJ
easat-5050	362	10	ban	ban	PROPN
easat-5050	362	11	,	,	PUNCT
easat-5050	362	12	h.	h.	PROPN
easat-5050	362	13	(	(	PUNCT
easat-5050	362	14	2022	2022	NUM
easat-5050	362	15	)	)	PUNCT
easat-5050	362	16	.	.	PUNCT
easat-5050	363	1	development	development	NOUN
easat-5050	363	2	of	of	ADP
easat-5050	363	3	a	a	DET
easat-5050	363	4	plug	plug	VERB
easat-5050	363	5	-	-	PUNCT
easat-5050	363	6	and	and	CCONJ
easat-5050	363	7	-	-	PUNCT
easat-5050	363	8	play	play	VERB
easat-5050	363	9	anti	anti	ADJ
easat-5050	363	10	-	-	ADJ
easat-5050	363	11	noise	noise	ADJ
easat-5050	363	12	module	module	NOUN
easat-5050	363	13	for	for	ADP
easat-5050	363	14	fault	fault	NOUN
easat-5050	363	15	diagnosis	diagnosis	NOUN
easat-5050	363	16	of	of	ADP
easat-5050	363	17	rotating	rotate	VERB
easat-5050	363	18	machines	machine	NOUN
easat-5050	363	19	in	in	ADP
easat-5050	363	20	nuclear	nuclear	ADJ
easat-5050	363	21	power	power	NOUN
easat-5050	363	22	plants	plant	NOUN
easat-5050	363	23	.	.	PUNCT
easat-5050	364	1	progress	progress	NOUN
easat-5050	364	2	in	in	ADP
easat-5050	364	3	nuclear	nuclear	ADJ
easat-5050	364	4	energy	energy	NOUN
easat-5050	364	5	,	,	PUNCT
easat-5050	364	6	151	151	NUM
easat-5050	364	7	,	,	PUNCT
easat-5050	364	8	104344	104344	NUM
easat-5050	364	9	.	.	PUNCT
easat-5050	365	1	https://doi.org/10.1016/j.pnucene.2022.104344	https://doi.org/10.1016/j.pnucene.2022.104344	PROPN
easat-5050	365	2	https://doi.org/10.1016/j.triboint.2015.12.037	https://doi.org/10.1016/j.triboint.2015.12.037	PROPN
easat-5050	365	3	https://doi.org/10.48550/arxiv.2407.14625	https://doi.org/10.48550/arxiv.2407.14625	PROPN
easat-5050	365	4	https://doi.org/10.1016/j.conengprac.2017.06.011	https://doi.org/10.1016/j.conengprac.2017.06.011	VERB
easat-5050	365	5	https://doi.org/10.3390/s20185112	https://doi.org/10.3390/s20185112	PROPN
easat-5050	365	6	https://doi.org/10.3390/s20061693	https://doi.org/10.3390/s20061693	NUM
easat-5050	365	7	https://doi.org/10.1109/tii.2019.2955540	https://doi.org/10.1109/tii.2019.2955540	INTJ
easat-5050	365	8	https://doi.org/10.1016/j.promfg.2020.07.005	https://doi.org/10.1016/j.promfg.2020.07.005	PROPN
easat-5050	365	9	https://doi.org/10.1007/s42417-021-00414-7	https://doi.org/10.1007/s42417-021-00414-7	NUM
easat-5050	365	10	https://doi.org/10.1016/j.ymssp.2013.11.011	https://doi.org/10.1016/j.ymssp.2013.11.011	NOUN
easat-5050	365	11	https://doi.org/10.1109/jsen.2020.3042182	https://doi.org/10.1109/jsen.2020.3042182	VERB
easat-5050	365	12	https://doi.org/10.3390/s19051088	https://doi.org/10.3390/s19051088	NUM
easat-5050	365	13	https://doi.org/10.3390/s23063157	https://doi.org/10.3390/s23063157	NOUN
easat-5050	365	14	https://doi.org/10.3390/s17020425	https://doi.org/10.3390/s17020425	NUM
easat-5050	365	15	https://doi.org/10.1117/12.3049948	https://doi.org/10.1117/12.3049948	NOUN
easat-5050	365	16	https://doi.org/10.1016/j.pnucene.2022.104344	https://doi.org/10.1016/j.pnucene.2022.104344	PROPN