id	sid	tid	token	lemma	pos
fcis-30390	1	1	frontiers	frontier	NOUN
fcis-30390	1	2	in	in	ADP
fcis-30390	1	3	computing	computing	NOUN
fcis-30390	1	4	and	and	CCONJ
fcis-30390	1	5	intelligent	intelligent	ADJ
fcis-30390	1	6	systems	system	NOUN
fcis-30390	1	7	issn	issn	VERB
fcis-30390	1	8	:	:	PUNCT
fcis-30390	1	9	2832	2832	NUM
fcis-30390	1	10	-	-	SYM
fcis-30390	1	11	6024	6024	NUM
fcis-30390	1	12	|	|	NOUN
fcis-30390	1	13	vol	vol	NOUN
fcis-30390	1	14	.	.	PROPN
fcis-30390	2	1	12	12	NUM
fcis-30390	2	2	,	,	PUNCT
fcis-30390	2	3	no	no	INTJ
fcis-30390	2	4	.	.	NOUN
fcis-30390	2	5	1	1	NUM
fcis-30390	2	6	,	,	PUNCT
fcis-30390	2	7	2025	2025	NUM
fcis-30390	2	8	35	35	NUM
fcis-30390	2	9	timemixer‐me	timemixer‐me	ADP
fcis-30390	2	10	:	:	PUNCT
fcis-30390	2	11	a	a	DET
fcis-30390	2	12	dual‐memory	dual‐memory	NOUN
fcis-30390	2	13	enhanced	enhance	VERB
fcis-30390	2	14	architecture	architecture	NOUN
fcis-30390	2	15	for	for	ADP
fcis-30390	2	16	time	time	NOUN
fcis-30390	2	17	series	series	PROPN
fcis-30390	2	18	forecasting	forecasting	PROPN
fcis-30390	2	19	dongze	dongze	PROPN
fcis-30390	2	20	wu	wu	PROPN
fcis-30390	2	21	,	,	PUNCT
fcis-30390	2	22	yonghai	yonghai	PROPN
fcis-30390	2	23	yang	yang	PROPN
fcis-30390	2	24	,	,	PUNCT
fcis-30390	2	25	dechuan	dechuan	PROPN
fcis-30390	2	26	qiu	qiu	PROPN
fcis-30390	3	1	*	*	PUNCT
fcis-30390	3	2	,	,	PUNCT
fcis-30390	3	3	guangming	guangming	PROPN
fcis-30390	3	4	li	li	PROPN
fcis-30390	3	5	university	university	PROPN
fcis-30390	3	6	of	of	ADP
fcis-30390	3	7	science	science	NOUN
fcis-30390	3	8	and	and	CCONJ
fcis-30390	3	9	technology	technology	PROPN
fcis-30390	3	10	west	west	PROPN
fcis-30390	3	11	campus	campus	PROPN
fcis-30390	3	12	,	,	PUNCT
fcis-30390	3	13	89	89	NUM
fcis-30390	3	14	beijing	beijing	PROPN
fcis-30390	3	15	road	road	NOUN
fcis-30390	3	16	,	,	PUNCT
fcis-30390	3	17	korla	korla	PROPN
fcis-30390	3	18	,	,	PUNCT
fcis-30390	3	19	bayingolin	bayingolin	PROPN
fcis-30390	3	20	mongol	mongol	PROPN
fcis-30390	3	21	autonomous	autonomous	PROPN
fcis-30390	3	22	prefecture	prefecture	NOUN
fcis-30390	3	23	841000	841000	NUM
fcis-30390	3	24	,	,	PUNCT
fcis-30390	3	25	xinjiang	xinjiang	PROPN
fcis-30390	3	26	,	,	PUNCT
fcis-30390	3	27	china	china	PROPN
fcis-30390	3	28	*	*	PUNCT
fcis-30390	3	29	corresponding	correspond	VERB
fcis-30390	3	30	author	author	NOUN
fcis-30390	3	31	:	:	PUNCT
fcis-30390	3	32	dechuan	dechuan	PROPN
fcis-30390	3	33	qiu	qiu	PROPN
fcis-30390	3	34	(	(	PUNCT
fcis-30390	3	35	email	email	NOUN
fcis-30390	3	36	:	:	PUNCT
fcis-30390	3	37	dimitriculter@foxmail.com	dimitriculter@foxmail.com	X
fcis-30390	3	38	)	)	PUNCT
fcis-30390	4	1	abstract	abstract	ADJ
fcis-30390	4	2	:	:	PUNCT
fcis-30390	4	3	time	time	NOUN
fcis-30390	4	4	series	series	PROPN
fcis-30390	4	5	prediction	prediction	PROPN
fcis-30390	4	6	,	,	PUNCT
fcis-30390	4	7	as	as	ADP
fcis-30390	4	8	a	a	DET
fcis-30390	4	9	research	research	NOUN
fcis-30390	4	10	field	field	NOUN
fcis-30390	4	11	with	with	ADP
fcis-30390	4	12	extensive	extensive	ADJ
fcis-30390	4	13	practical	practical	ADJ
fcis-30390	4	14	applications	application	NOUN
fcis-30390	4	15	,	,	PUNCT
fcis-30390	4	16	consistently	consistently	ADV
fcis-30390	4	17	faces	face	VERB
fcis-30390	4	18	a	a	DET
fcis-30390	4	19	key	key	ADJ
fcis-30390	4	20	challenge	challenge	NOUN
fcis-30390	4	21	:	:	PUNCT
fcis-30390	4	22	how	how	SCONJ
fcis-30390	4	23	to	to	PART
fcis-30390	4	24	improve	improve	VERB
fcis-30390	4	25	the	the	DET
fcis-30390	4	26	model	model	NOUN
fcis-30390	4	27	's	's	PART
fcis-30390	4	28	overall	overall	ADJ
fcis-30390	4	29	predictive	predictive	ADJ
fcis-30390	4	30	capability	capability	NOUN
fcis-30390	4	31	while	while	SCONJ
fcis-30390	4	32	maintaining	maintain	VERB
fcis-30390	4	33	short	short	ADJ
fcis-30390	4	34	-	-	PUNCT
fcis-30390	4	35	term	term	NOUN
fcis-30390	4	36	prediction	prediction	NOUN
fcis-30390	4	37	accuracy	accuracy	NOUN
fcis-30390	4	38	.	.	PUNCT
fcis-30390	5	1	to	to	PART
fcis-30390	5	2	address	address	VERB
fcis-30390	5	3	this	this	DET
fcis-30390	5	4	challenge	challenge	NOUN
fcis-30390	5	5	,	,	PUNCT
fcis-30390	5	6	this	this	DET
fcis-30390	5	7	research	research	NOUN
fcis-30390	5	8	innovatively	innovatively	ADV
fcis-30390	5	9	proposes	propose	VERB
fcis-30390	5	10	timemixer	timemixer	NOUN
fcis-30390	5	11	-	-	PUNCT
fcis-30390	5	12	me	i	PRON
fcis-30390	5	13	,	,	PUNCT
fcis-30390	5	14	a	a	DET
fcis-30390	5	15	dual	dual	ADJ
fcis-30390	5	16	-	-	PUNCT
fcis-30390	5	17	memory	memory	NOUN
fcis-30390	5	18	enhanced	enhance	VERB
fcis-30390	5	19	architecture	architecture	NOUN
fcis-30390	5	20	that	that	PRON
fcis-30390	5	21	ingeniously	ingeniously	ADV
fcis-30390	5	22	integrates	integrate	VERB
fcis-30390	5	23	cognitive	cognitive	ADJ
fcis-30390	5	24	science	science	NOUN
fcis-30390	5	25	's	's	PART
fcis-30390	5	26	dual	dual	ADJ
fcis-30390	5	27	memory	memory	NOUN
fcis-30390	5	28	system	system	NOUN
fcis-30390	5	29	theory	theory	NOUN
fcis-30390	5	30	with	with	ADP
fcis-30390	5	31	deep	deep	ADJ
fcis-30390	5	32	learning	learning	NOUN
fcis-30390	5	33	techniques	technique	NOUN
fcis-30390	5	34	,	,	PUNCT
fcis-30390	5	35	achieving	achieve	VERB
fcis-30390	5	36	adaptive	adaptive	ADJ
fcis-30390	5	37	multi	multi	ADJ
fcis-30390	5	38	-	-	NOUN
fcis-30390	5	39	scale	scale	ADJ
fcis-30390	5	40	modeling	modeling	NOUN
fcis-30390	5	41	of	of	ADP
fcis-30390	5	42	time	time	NOUN
fcis-30390	5	43	series	series	NOUN
fcis-30390	5	44	.	.	PUNCT
fcis-30390	6	1	through	through	ADP
fcis-30390	6	2	experimental	experimental	ADJ
fcis-30390	6	3	validation	validation	NOUN
fcis-30390	6	4	on	on	ADP
fcis-30390	6	5	multiple	multiple	ADJ
fcis-30390	6	6	weather	weather	NOUN
fcis-30390	6	7	datasets	dataset	NOUN
fcis-30390	6	8	from	from	ADP
fcis-30390	6	9	southern	southern	PROPN
fcis-30390	6	10	xinjiang	xinjiang	PROPN
fcis-30390	6	11	,	,	PUNCT
fcis-30390	6	12	timemixer	timemixer	NOUN
fcis-30390	6	13	-	-	PUNCT
fcis-30390	6	14	me	i	PRON
fcis-30390	6	15	demonstrates	demonstrate	VERB
fcis-30390	6	16	exceptional	exceptional	ADJ
fcis-30390	6	17	predictive	predictive	ADJ
fcis-30390	6	18	performance	performance	NOUN
fcis-30390	6	19	.	.	PUNCT
fcis-30390	7	1	compared	compare	VERB
fcis-30390	7	2	to	to	ADP
fcis-30390	7	3	the	the	DET
fcis-30390	7	4	baseline	baseline	PROPN
fcis-30390	7	5	timemixer	timemixer	PROPN
fcis-30390	7	6	model	model	PROPN
fcis-30390	7	7	,	,	PUNCT
fcis-30390	7	8	it	it	PRON
fcis-30390	7	9	achieved	achieve	VERB
fcis-30390	7	10	a	a	DET
fcis-30390	7	11	significant	significant	ADJ
fcis-30390	7	12	reduction	reduction	NOUN
fcis-30390	7	13	of	of	ADP
fcis-30390	7	14	2%-2.4	2%-2.4	NUM
fcis-30390	7	15	%	%	NOUN
fcis-30390	7	16	in	in	ADP
fcis-30390	7	17	mse	mse	NOUN
fcis-30390	7	18	metrics	metric	NOUN
fcis-30390	7	19	and	and	CCONJ
fcis-30390	7	20	a	a	DET
fcis-30390	7	21	remarkable	remarkable	ADJ
fcis-30390	7	22	improvement	improvement	NOUN
fcis-30390	7	23	of	of	ADP
fcis-30390	7	24	3.7%-4.2	3.7%-4.2	NUM
fcis-30390	7	25	%	%	NOUN
fcis-30390	7	26	in	in	ADP
fcis-30390	7	27	mae	mae	PROPN
fcis-30390	7	28	metrics	metric	NOUN
fcis-30390	7	29	.	.	PUNCT
fcis-30390	8	1	these	these	DET
fcis-30390	8	2	results	result	NOUN
fcis-30390	8	3	convincingly	convincingly	ADV
fcis-30390	8	4	demonstrate	demonstrate	VERB
fcis-30390	8	5	the	the	DET
fcis-30390	8	6	model	model	NOUN
fcis-30390	8	7	's	's	PART
fcis-30390	8	8	excellent	excellent	ADJ
fcis-30390	8	9	performance	performance	NOUN
fcis-30390	8	10	in	in	ADP
fcis-30390	8	11	short	short	ADJ
fcis-30390	8	12	-	-	PUNCT
fcis-30390	8	13	term	term	NOUN
fcis-30390	8	14	prediction	prediction	NOUN
fcis-30390	8	15	tasks	task	NOUN
fcis-30390	8	16	,	,	PUNCT
fcis-30390	8	17	providing	provide	VERB
fcis-30390	8	18	a	a	DET
fcis-30390	8	19	novel	novel	ADJ
fcis-30390	8	20	approach	approach	NOUN
fcis-30390	8	21	to	to	ADP
fcis-30390	8	22	addressing	address	VERB
fcis-30390	8	23	the	the	DET
fcis-30390	8	24	trade	trade	NOUN
fcis-30390	8	25	-	-	PUNCT
fcis-30390	8	26	off	off	NOUN
fcis-30390	8	27	between	between	ADP
fcis-30390	8	28	long	long	ADJ
fcis-30390	8	29	-	-	PUNCT
fcis-30390	8	30	term	term	NOUN
fcis-30390	8	31	and	and	CCONJ
fcis-30390	8	32	short	short	ADJ
fcis-30390	8	33	-	-	PUNCT
fcis-30390	8	34	term	term	NOUN
fcis-30390	8	35	dependencies	dependency	NOUN
fcis-30390	8	36	in	in	ADP
fcis-30390	8	37	time	time	NOUN
fcis-30390	8	38	series	series	PROPN
fcis-30390	8	39	prediction	prediction	NOUN
fcis-30390	8	40	.	.	PUNCT
fcis-30390	9	1	keywords	keyword	NOUN
fcis-30390	9	2	:	:	PUNCT
fcis-30390	9	3	dual	dual	ADJ
fcis-30390	9	4	memory	memory	NOUN
fcis-30390	9	5	-	-	PUNCT
fcis-30390	9	6	enhanced	enhance	VERB
fcis-30390	9	7	architecture	architecture	NOUN
fcis-30390	9	8	;	;	PUNCT
fcis-30390	9	9	cognitive	cognitive	ADJ
fcis-30390	9	10	science	science	NOUN
fcis-30390	9	11	;	;	PUNCT
fcis-30390	9	12	adaptive	adaptive	ADJ
fcis-30390	9	13	multi	multi	ADJ
fcis-30390	9	14	-	-	ADJ
fcis-30390	9	15	scale	scale	ADJ
fcis-30390	9	16	modeling	modeling	NOUN
fcis-30390	9	17	;	;	PUNCT
fcis-30390	9	18	short	short	ADJ
fcis-30390	9	19	-	-	PUNCT
fcis-30390	9	20	term	term	NOUN
fcis-30390	9	21	prediction	prediction	NOUN
fcis-30390	9	22	.	.	PUNCT
fcis-30390	10	1	1	1	X
fcis-30390	10	2	.	.	X
fcis-30390	10	3	introduction	introduction	NOUN
fcis-30390	10	4	in	in	ADP
fcis-30390	10	5	recent	recent	ADJ
fcis-30390	10	6	years	year	NOUN
fcis-30390	10	7	,	,	PUNCT
fcis-30390	10	8	the	the	DET
fcis-30390	10	9	southern	southern	ADJ
fcis-30390	10	10	region	region	NOUN
fcis-30390	10	11	of	of	ADP
fcis-30390	10	12	xinjiang	xinjiang	PROPN
fcis-30390	10	13	has	have	AUX
fcis-30390	10	14	experienced	experience	VERB
fcis-30390	10	15	a	a	DET
fcis-30390	10	16	continuous	continuous	ADJ
fcis-30390	10	17	increase	increase	NOUN
fcis-30390	10	18	in	in	ADP
fcis-30390	10	19	annual	annual	ADJ
fcis-30390	10	20	average	average	ADJ
fcis-30390	10	21	precipitation	precipitation	NOUN
fcis-30390	10	22	and	and	CCONJ
fcis-30390	10	23	frequent	frequent	ADJ
fcis-30390	10	24	extreme	extreme	ADJ
fcis-30390	10	25	weather	weather	NOUN
fcis-30390	10	26	events	event	NOUN
fcis-30390	10	27	,	,	PUNCT
fcis-30390	10	28	significantly	significantly	ADV
fcis-30390	10	29	impacting	impact	VERB
fcis-30390	10	30	local	local	ADJ
fcis-30390	10	31	agricultural	agricultural	ADJ
fcis-30390	10	32	production	production	NOUN
fcis-30390	10	33	and	and	CCONJ
fcis-30390	10	34	residents	resident	NOUN
fcis-30390	10	35	'	'	PART
fcis-30390	10	36	lives	life	NOUN
fcis-30390	10	37	.	.	PUNCT
fcis-30390	11	1	accurate	accurate	ADJ
fcis-30390	11	2	short	short	ADJ
fcis-30390	11	3	-	-	PUNCT
fcis-30390	11	4	term	term	NOUN
fcis-30390	11	5	precipitation	precipitation	NOUN
fcis-30390	11	6	forecasting	forecasting	NOUN
fcis-30390	11	7	is	be	AUX
fcis-30390	11	8	crucial	crucial	ADJ
fcis-30390	11	9	for	for	ADP
fcis-30390	11	10	disaster	disaster	NOUN
fcis-30390	11	11	prevention	prevention	NOUN
fcis-30390	11	12	.	.	PUNCT
fcis-30390	12	1	however	however	ADV
fcis-30390	12	2	,	,	PUNCT
fcis-30390	12	3	due	due	ADP
fcis-30390	12	4	to	to	ADP
fcis-30390	12	5	the	the	DET
fcis-30390	12	6	unique	unique	ADJ
fcis-30390	12	7	geographical	geographical	ADJ
fcis-30390	12	8	environment	environment	NOUN
fcis-30390	12	9	and	and	CCONJ
fcis-30390	12	10	climatic	climatic	ADJ
fcis-30390	12	11	characteristics	characteristic	NOUN
fcis-30390	12	12	of	of	ADP
fcis-30390	12	13	southern	southern	ADJ
fcis-30390	12	14	xinjiang	xinjiang	PROPN
fcis-30390	12	15	,	,	PUNCT
fcis-30390	12	16	existing	exist	VERB
fcis-30390	12	17	precipitation	precipitation	NOUN
fcis-30390	12	18	forecasting	forecasting	NOUN
fcis-30390	12	19	models	model	NOUN
fcis-30390	12	20	face	face	VERB
fcis-30390	12	21	multiple	multiple	ADJ
fcis-30390	12	22	challenges	challenge	NOUN
fcis-30390	12	23	in	in	ADP
fcis-30390	12	24	this	this	DET
fcis-30390	12	25	region	region	NOUN
fcis-30390	12	26	:	:	PUNCT
fcis-30390	12	27	located	locate	VERB
fcis-30390	12	28	between	between	ADP
fcis-30390	12	29	the	the	DET
fcis-30390	12	30	tianshan	tianshan	NOUN
fcis-30390	12	31	and	and	CCONJ
fcis-30390	12	32	kunlun	kunlun	PROPN
fcis-30390	12	33	mountains	mountains	PROPN
fcis-30390	12	34	,	,	PUNCT
fcis-30390	12	35	the	the	DET
fcis-30390	12	36	diverse	diverse	ADJ
fcis-30390	12	37	topographical	topographical	ADJ
fcis-30390	12	38	features	feature	NOUN
fcis-30390	12	39	including	include	VERB
fcis-30390	12	40	mountains	mountain	NOUN
fcis-30390	12	41	,	,	PUNCT
fcis-30390	12	42	basins	basin	NOUN
fcis-30390	12	43	,	,	PUNCT
fcis-30390	12	44	and	and	CCONJ
fcis-30390	12	45	gobi	gobi	PROPN
fcis-30390	12	46	desert	desert	PROPN
fcis-30390	12	47	lead	lead	NOUN
fcis-30390	12	48	to	to	ADP
fcis-30390	12	49	complex	complex	ADJ
fcis-30390	12	50	local	local	ADJ
fcis-30390	12	51	weather	weather	NOUN
fcis-30390	12	52	systems	system	NOUN
fcis-30390	12	53	.	.	PUNCT
fcis-30390	13	1	existing	exist	VERB
fcis-30390	13	2	models	model	NOUN
fcis-30390	13	3	perform	perform	VERB
fcis-30390	13	4	poorly	poorly	ADV
fcis-30390	13	5	in	in	ADP
fcis-30390	13	6	handling	handle	VERB
fcis-30390	13	7	mountain	mountain	NOUN
fcis-30390	13	8	-	-	PUNCT
fcis-30390	13	9	plain	plain	NOUN
fcis-30390	13	10	circulation	circulation	NOUN
fcis-30390	13	11	and	and	CCONJ
fcis-30390	13	12	local	local	ADJ
fcis-30390	13	13	convection	convection	NOUN
fcis-30390	13	14	systems	system	NOUN
fcis-30390	13	15	.	.	PUNCT
fcis-30390	14	1	furthermore	furthermore	ADV
fcis-30390	14	2	,	,	PUNCT
fcis-30390	14	3	meteorological	meteorological	ADJ
fcis-30390	14	4	elements	element	NOUN
fcis-30390	14	5	such	such	ADJ
fcis-30390	14	6	as	as	ADP
fcis-30390	14	7	temperature	temperature	NOUN
fcis-30390	14	8	,	,	PUNCT
fcis-30390	14	9	humidity	humidity	NOUN
fcis-30390	14	10	,	,	PUNCT
fcis-30390	14	11	pressure	pressure	NOUN
fcis-30390	14	12	,	,	PUNCT
fcis-30390	14	13	and	and	CCONJ
fcis-30390	14	14	wind	wind	NOUN
fcis-30390	14	15	fields	field	NOUN
fcis-30390	14	16	exhibit	exhibit	VERB
fcis-30390	14	17	unique	unique	ADJ
fcis-30390	14	18	spatiotemporal	spatiotemporal	ADJ
fcis-30390	14	19	variation	variation	NOUN
fcis-30390	14	20	characteristics	characteristic	NOUN
fcis-30390	14	21	and	and	CCONJ
fcis-30390	14	22	interaction	interaction	NOUN
fcis-30390	14	23	patterns	pattern	NOUN
fcis-30390	14	24	during	during	ADP
fcis-30390	14	25	precipitation	precipitation	NOUN
fcis-30390	14	26	formation	formation	NOUN
fcis-30390	14	27	,	,	PUNCT
fcis-30390	14	28	increasing	increase	VERB
fcis-30390	14	29	the	the	DET
fcis-30390	14	30	difficulty	difficulty	NOUN
fcis-30390	14	31	of	of	ADP
fcis-30390	14	32	capturing	capture	VERB
fcis-30390	14	33	coupling	couple	VERB
fcis-30390	14	34	relationships	relationship	NOUN
fcis-30390	14	35	.	.	PUNCT
fcis-30390	15	1	additionally	additionally	ADV
fcis-30390	15	2	,	,	PUNCT
fcis-30390	15	3	due	due	ADP
fcis-30390	15	4	to	to	ADP
fcis-30390	15	5	terrain	terrain	NOUN
fcis-30390	15	6	and	and	CCONJ
fcis-30390	15	7	observation	observation	NOUN
fcis-30390	15	8	station	station	NOUN
fcis-30390	15	9	distribution	distribution	NOUN
fcis-30390	15	10	limitations	limitation	NOUN
fcis-30390	15	11	,	,	PUNCT
fcis-30390	15	12	meteorological	meteorological	ADJ
fcis-30390	15	13	observation	observation	NOUN
fcis-30390	15	14	data	datum	NOUN
fcis-30390	15	15	in	in	ADP
fcis-30390	15	16	this	this	DET
fcis-30390	15	17	region	region	NOUN
fcis-30390	15	18	suffers	suffer	VERB
fcis-30390	15	19	from	from	ADP
fcis-30390	15	20	uneven	uneven	ADJ
fcis-30390	15	21	spatial	spatial	ADJ
fcis-30390	15	22	distribution	distribution	NOUN
fcis-30390	15	23	and	and	CCONJ
fcis-30390	15	24	discontinuous	discontinuous	ADJ
fcis-30390	15	25	time	time	NOUN
fcis-30390	15	26	series	series	NOUN
fcis-30390	15	27	.	.	PUNCT
fcis-30390	16	1	the	the	DET
fcis-30390	16	2	region	region	NOUN
fcis-30390	16	3	's	's	PART
fcis-30390	16	4	precipitation	precipitation	NOUN
fcis-30390	16	5	shows	show	VERB
fcis-30390	16	6	significant	significant	ADJ
fcis-30390	16	7	seasonal	seasonal	ADJ
fcis-30390	16	8	variation	variation	NOUN
fcis-30390	16	9	characteristics	characteristic	NOUN
fcis-30390	16	10	,	,	PUNCT
fcis-30390	16	11	with	with	ADP
fcis-30390	16	12	substantial	substantial	ADJ
fcis-30390	16	13	differences	difference	NOUN
fcis-30390	16	14	in	in	ADP
fcis-30390	16	15	precipitation	precipitation	NOUN
fcis-30390	16	16	mechanisms	mechanism	NOUN
fcis-30390	16	17	across	across	ADP
fcis-30390	16	18	seasons	season	NOUN
fcis-30390	16	19	,	,	PUNCT
fcis-30390	16	20	collectively	collectively	ADV
fcis-30390	16	21	making	make	VERB
fcis-30390	16	22	it	it	PRON
fcis-30390	16	23	difficult	difficult	ADJ
fcis-30390	16	24	for	for	SCONJ
fcis-30390	16	25	existing	exist	VERB
fcis-30390	16	26	forecasting	forecasting	NOUN
fcis-30390	16	27	models	model	NOUN
fcis-30390	16	28	to	to	PART
fcis-30390	16	29	maintain	maintain	VERB
fcis-30390	16	30	stable	stable	ADJ
fcis-30390	16	31	prediction	prediction	NOUN
fcis-30390	16	32	accuracy	accuracy	NOUN
fcis-30390	16	33	.	.	PUNCT
fcis-30390	17	1	to	to	PART
fcis-30390	17	2	address	address	VERB
fcis-30390	17	3	these	these	DET
fcis-30390	17	4	challenges	challenge	NOUN
fcis-30390	17	5	,	,	PUNCT
fcis-30390	17	6	this	this	DET
fcis-30390	17	7	paper	paper	NOUN
fcis-30390	17	8	proposes	propose	VERB
fcis-30390	17	9	a	a	DET
fcis-30390	17	10	novel	novel	ADJ
fcis-30390	17	11	dual	dual	ADJ
fcis-30390	17	12	-	-	PUNCT
fcis-30390	17	13	memory	memory	NOUN
fcis-30390	17	14	enhanced	enhance	VERB
fcis-30390	17	15	architecture	architecture	NOUN
fcis-30390	17	16	,	,	PUNCT
fcis-30390	17	17	timemixer	timemixer	NOUN
fcis-30390	17	18	-	-	PUNCT
fcis-30390	17	19	me	i	PRON
fcis-30390	17	20	,	,	PUNCT
fcis-30390	17	21	based	base	VERB
fcis-30390	17	22	on	on	ADP
fcis-30390	17	23	timemixer	timemixer	NOUN
fcis-30390	17	24	.	.	PUNCT
fcis-30390	18	1	the	the	DET
fcis-30390	18	2	model	model	NOUN
fcis-30390	18	3	achieves	achieve	VERB
fcis-30390	18	4	innovative	innovative	ADJ
fcis-30390	18	5	breakthroughs	breakthrough	NOUN
fcis-30390	18	6	in	in	ADP
fcis-30390	18	7	multiple	multiple	ADJ
fcis-30390	18	8	aspects	aspect	NOUN
fcis-30390	18	9	:	:	PUNCT
fcis-30390	18	10	designed	design	VERB
fcis-30390	18	11	for	for	ADP
fcis-30390	18	12	southern	southern	ADJ
fcis-30390	18	13	xinjiang	xinjiang	PROPN
fcis-30390	18	14	's	's	PART
fcis-30390	18	15	characteristics	characteristic	NOUN
fcis-30390	18	16	,	,	PUNCT
fcis-30390	18	17	it	it	PRON
fcis-30390	18	18	incorporates	incorporate	VERB
fcis-30390	18	19	a	a	DET
fcis-30390	18	20	dual	dual	ADJ
fcis-30390	18	21	-	-	PUNCT
fcis-30390	18	22	memory	memory	NOUN
fcis-30390	18	23	system	system	NOUN
fcis-30390	18	24	with	with	ADP
fcis-30390	18	25	short	short	ADJ
fcis-30390	18	26	-	-	PUNCT
fcis-30390	18	27	term	term	NOUN
fcis-30390	18	28	and	and	CCONJ
fcis-30390	18	29	long	long	ADJ
fcis-30390	18	30	-	-	PUNCT
fcis-30390	18	31	term	term	NOUN
fcis-30390	18	32	memory	memory	NOUN
fcis-30390	18	33	units	unit	NOUN
fcis-30390	18	34	.	.	PUNCT
fcis-30390	19	1	the	the	DET
fcis-30390	19	2	short	short	ADJ
fcis-30390	19	3	-	-	PUNCT
fcis-30390	19	4	term	term	NOUN
fcis-30390	19	5	memory	memory	NOUN
fcis-30390	19	6	unit	unit	NOUN
fcis-30390	19	7	employs	employ	VERB
fcis-30390	19	8	3×1	3×1	NUM
fcis-30390	19	9	depth	depth	NOUN
fcis-30390	19	10	-	-	PUNCT
fcis-30390	19	11	separable	separable	NOUN
fcis-30390	19	12	convolution	convolution	NOUN
fcis-30390	19	13	and	and	CCONJ
fcis-30390	19	14	4	4	NUM
fcis-30390	19	15	-	-	PUNCT
fcis-30390	19	16	head	head	NOUN
fcis-30390	19	17	attention	attention	NOUN
fcis-30390	19	18	mechanism	mechanism	NOUN
fcis-30390	19	19	to	to	PART
fcis-30390	19	20	capture	capture	VERB
fcis-30390	19	21	rapid	rapid	ADJ
fcis-30390	19	22	changes	change	NOUN
fcis-30390	19	23	in	in	ADP
fcis-30390	19	24	local	local	ADJ
fcis-30390	19	25	weather	weather	NOUN
fcis-30390	19	26	systems	system	NOUN
fcis-30390	19	27	,	,	PUNCT
fcis-30390	19	28	while	while	SCONJ
fcis-30390	19	29	the	the	DET
fcis-30390	19	30	long	long	ADJ
fcis-30390	19	31	-	-	PUNCT
fcis-30390	19	32	term	term	NOUN
fcis-30390	19	33	memory	memory	NOUN
fcis-30390	19	34	unit	unit	NOUN
fcis-30390	19	35	utilizes	utilize	VERB
fcis-30390	19	36	a	a	DET
fcis-30390	19	37	32	32	NUM
fcis-30390	19	38	-	-	PUNCT
fcis-30390	19	39	dimensional	dimensional	ADJ
fcis-30390	19	40	memory	memory	NOUN
fcis-30390	19	41	bank	bank	NOUN
fcis-30390	19	42	and	and	CCONJ
fcis-30390	19	43	adaptive	adaptive	ADJ
fcis-30390	19	44	pooling	pooling	NOUN
fcis-30390	19	45	mechanism	mechanism	NOUN
fcis-30390	19	46	to	to	PART
fcis-30390	19	47	store	store	VERB
fcis-30390	19	48	and	and	CCONJ
fcis-30390	19	49	leverage	leverage	VERB
fcis-30390	19	50	seasonal	seasonal	ADJ
fcis-30390	19	51	climate	climate	NOUN
fcis-30390	19	52	characteristics	characteristic	NOUN
fcis-30390	19	53	,	,	PUNCT
fcis-30390	19	54	enabling	enable	VERB
fcis-30390	19	55	effective	effective	ADJ
fcis-30390	19	56	modeling	modeling	NOUN
fcis-30390	19	57	of	of	ADP
fcis-30390	19	58	multi	multi	ADJ
fcis-30390	19	59	-	-	ADJ
fcis-30390	19	60	scale	scale	ADJ
fcis-30390	19	61	weather	weather	NOUN
fcis-30390	19	62	processes	process	NOUN
fcis-30390	19	63	.	.	PUNCT
fcis-30390	20	1	the	the	DET
fcis-30390	20	2	model	model	NOUN
fcis-30390	20	3	innovatively	innovatively	ADV
fcis-30390	20	4	introduces	introduce	VERB
fcis-30390	20	5	an	an	DET
fcis-30390	20	6	adaptive	adaptive	ADJ
fcis-30390	20	7	feature	feature	NOUN
fcis-30390	20	8	fusion	fusion	NOUN
fcis-30390	20	9	mechanism	mechanism	NOUN
fcis-30390	20	10	,	,	PUNCT
fcis-30390	20	11	dynamically	dynamically	ADV
fcis-30390	20	12	adjusting	adjust	VERB
fcis-30390	20	13	the	the	DET
fcis-30390	20	14	importance	importance	NOUN
fcis-30390	20	15	of	of	ADP
fcis-30390	20	16	different	different	ADJ
fcis-30390	20	17	meteorological	meteorological	ADJ
fcis-30390	20	18	elements	element	NOUN
fcis-30390	20	19	through	through	ADP
fcis-30390	20	20	learnable	learnable	ADJ
fcis-30390	20	21	weight	weight	NOUN
fcis-30390	20	22	matrices	matrix	NOUN
fcis-30390	20	23	,	,	PUNCT
fcis-30390	20	24	enhancing	enhance	VERB
fcis-30390	20	25	understanding	understanding	NOUN
fcis-30390	20	26	and	and	CCONJ
fcis-30390	20	27	prediction	prediction	NOUN
fcis-30390	20	28	capabilities	capability	NOUN
fcis-30390	20	29	of	of	ADP
fcis-30390	20	30	complex	complex	ADJ
fcis-30390	20	31	weather	weather	NOUN
fcis-30390	20	32	system	system	NOUN
fcis-30390	20	33	evolution	evolution	NOUN
fcis-30390	20	34	.	.	PUNCT
fcis-30390	21	1	furthermore	furthermore	ADV
fcis-30390	21	2	,	,	PUNCT
fcis-30390	21	3	through	through	ADP
fcis-30390	21	4	specialized	specialized	ADJ
fcis-30390	21	5	data	datum	NOUN
fcis-30390	21	6	preprocessing	preprocessing	NOUN
fcis-30390	21	7	and	and	CCONJ
fcis-30390	21	8	augmentation	augmentation	NOUN
fcis-30390	21	9	strategies	strategy	NOUN
fcis-30390	21	10	,	,	PUNCT
fcis-30390	21	11	and	and	CCONJ
fcis-30390	21	12	the	the	DET
fcis-30390	21	13	introduction	introduction	NOUN
fcis-30390	21	14	of	of	ADP
fcis-30390	21	15	a	a	DET
fcis-30390	21	16	periodic	periodic	ADJ
fcis-30390	21	17	detector	detector	NOUN
fcis-30390	21	18	based	base	VERB
fcis-30390	21	19	on	on	ADP
fcis-30390	21	20	7×1	7×1	NUM
fcis-30390	21	21	convolution	convolution	NOUN
fcis-30390	21	22	kernels	kernel	NOUN
fcis-30390	21	23	,	,	PUNCT
fcis-30390	21	24	the	the	DET
fcis-30390	21	25	model	model	NOUN
fcis-30390	21	26	significantly	significantly	ADV
fcis-30390	21	27	improves	improve	VERB
fcis-30390	21	28	its	its	PRON
fcis-30390	21	29	robustness	robustness	NOUN
fcis-30390	21	30	to	to	ADP
fcis-30390	21	31	incomplete	incomplete	ADJ
fcis-30390	21	32	and	and	CCONJ
fcis-30390	21	33	noisy	noisy	ADJ
fcis-30390	21	34	data	datum	NOUN
fcis-30390	21	35	.	.	PUNCT
fcis-30390	22	1	experiments	experiment	NOUN
fcis-30390	22	2	show	show	VERB
fcis-30390	22	3	that	that	SCONJ
fcis-30390	22	4	compared	compare	VERB
fcis-30390	22	5	to	to	ADP
fcis-30390	22	6	the	the	DET
fcis-30390	22	7	baseline	baseline	PROPN
fcis-30390	22	8	timemixer	timemixer	PROPN
fcis-30390	22	9	model	model	NOUN
fcis-30390	22	10	,	,	PUNCT
fcis-30390	22	11	timemixer	timemixer	NOUN
fcis-30390	22	12	-	-	PUNCT
fcis-30390	22	13	me	i	PRON
fcis-30390	22	14	achieves	achieve	VERB
fcis-30390	22	15	2%-2.4	2%-2.4	NUM
fcis-30390	22	16	%	%	NOUN
fcis-30390	22	17	lower	low	ADJ
fcis-30390	22	18	mse	mse	NOUN
fcis-30390	22	19	and	and	CCONJ
fcis-30390	22	20	3.7%-4.2	3.7%-4.2	NUM
fcis-30390	22	21	%	%	NOUN
fcis-30390	22	22	lower	low	ADJ
fcis-30390	22	23	mae	mae	PROPN
fcis-30390	22	24	on	on	ADP
fcis-30390	22	25	multiple	multiple	ADJ
fcis-30390	22	26	southern	southern	ADJ
fcis-30390	22	27	xinjiang	xinjiang	PROPN
fcis-30390	22	28	weather	weather	PROPN
fcis-30390	22	29	datasets	dataset	NOUN
fcis-30390	22	30	,	,	PUNCT
fcis-30390	22	31	demonstrating	demonstrate	VERB
fcis-30390	22	32	excellent	excellent	ADJ
fcis-30390	22	33	shortterm	shortterm	NOUN
fcis-30390	22	34	prediction	prediction	NOUN
fcis-30390	22	35	accuracy	accuracy	NOUN
fcis-30390	22	36	.	.	PUNCT
fcis-30390	23	1	these	these	DET
fcis-30390	23	2	results	result	NOUN
fcis-30390	23	3	confirm	confirm	VERB
fcis-30390	23	4	the	the	DET
fcis-30390	23	5	practical	practical	ADJ
fcis-30390	23	6	value	value	NOUN
fcis-30390	23	7	of	of	ADP
fcis-30390	23	8	the	the	DET
fcis-30390	23	9	proposed	propose	VERB
fcis-30390	23	10	model	model	NOUN
fcis-30390	23	11	in	in	ADP
fcis-30390	23	12	short	short	ADJ
fcis-30390	23	13	-	-	PUNCT
fcis-30390	23	14	term	term	NOUN
fcis-30390	23	15	precipitation	precipitation	NOUN
fcis-30390	23	16	forecasting	forecasting	NOUN
fcis-30390	23	17	for	for	ADP
fcis-30390	23	18	the	the	DET
fcis-30390	23	19	southern	southern	ADJ
fcis-30390	23	20	xinjiang	xinjiang	PROPN
fcis-30390	23	21	region	region	PROPN
fcis-30390	23	22	.	.	PUNCT
fcis-30390	24	1	the	the	DET
fcis-30390	24	2	remainder	remainder	NOUN
fcis-30390	24	3	of	of	ADP
fcis-30390	24	4	this	this	DET
fcis-30390	24	5	paper	paper	NOUN
fcis-30390	24	6	is	be	AUX
fcis-30390	24	7	organized	organize	VERB
fcis-30390	24	8	as	as	SCONJ
fcis-30390	24	9	follows	follow	VERB
fcis-30390	24	10	:	:	PUNCT
fcis-30390	24	11	section	section	NOUN
fcis-30390	24	12	2	2	NUM
fcis-30390	24	13	reviews	review	NOUN
fcis-30390	24	14	related	relate	VERB
fcis-30390	24	15	work	work	NOUN
fcis-30390	24	16	;	;	PUNCT
fcis-30390	24	17	section	section	NOUN
fcis-30390	24	18	3	3	NUM
fcis-30390	24	19	details	detail	NOUN
fcis-30390	24	20	the	the	DET
fcis-30390	24	21	timemixer	timemixer	NOUN
fcis-30390	24	22	-	-	PUNCT
fcis-30390	24	23	me	i	PRON
fcis-30390	24	24	model	model	NOUN
fcis-30390	24	25	architecture	architecture	NOUN
fcis-30390	24	26	;	;	PUNCT
fcis-30390	24	27	section	section	NOUN
fcis-30390	24	28	4	4	NUM
fcis-30390	24	29	presents	present	VERB
fcis-30390	24	30	experimental	experimental	ADJ
fcis-30390	24	31	results	result	NOUN
fcis-30390	24	32	and	and	CCONJ
fcis-30390	24	33	analysis	analysis	NOUN
fcis-30390	24	34	in	in	ADP
fcis-30390	24	35	southern	southern	ADJ
fcis-30390	24	36	xinjiang	xinjiang	PROPN
fcis-30390	24	37	;	;	PUNCT
fcis-30390	24	38	and	and	CCONJ
fcis-30390	24	39	finally	finally	ADV
fcis-30390	24	40	,	,	PUNCT
fcis-30390	24	41	section	section	NOUN
fcis-30390	24	42	5	5	NUM
fcis-30390	24	43	concludes	conclude	VERB
fcis-30390	24	44	the	the	DET
fcis-30390	24	45	paper	paper	NOUN
fcis-30390	24	46	and	and	CCONJ
fcis-30390	24	47	discusses	discuss	VERB
fcis-30390	24	48	future	future	ADJ
fcis-30390	24	49	improvement	improvement	NOUN
fcis-30390	24	50	directions	direction	NOUN
fcis-30390	24	51	.	.	PUNCT
fcis-30390	25	1	2	2	X
fcis-30390	25	2	.	.	NUM
fcis-30390	25	3	related	relate	VERB
fcis-30390	25	4	work	work	NOUN
fcis-30390	25	5	research	research	NOUN
fcis-30390	25	6	methods	method	NOUN
fcis-30390	25	7	in	in	ADP
fcis-30390	25	8	time	time	NOUN
fcis-30390	25	9	series	series	PROPN
fcis-30390	25	10	forecasting	forecasting	NOUN
fcis-30390	25	11	have	have	AUX
fcis-30390	25	12	undergone	undergo	VERB
fcis-30390	25	13	significant	significant	ADJ
fcis-30390	25	14	evolution	evolution	NOUN
fcis-30390	25	15	from	from	ADP
fcis-30390	25	16	traditional	traditional	ADJ
fcis-30390	25	17	statistical	statistical	ADJ
fcis-30390	25	18	approaches	approach	NOUN
fcis-30390	25	19	to	to	ADP
fcis-30390	25	20	deep	deep	ADJ
fcis-30390	25	21	learning	learning	NOUN
fcis-30390	25	22	methods	method	NOUN
fcis-30390	25	23	.	.	PUNCT
fcis-30390	26	1	early	early	ADJ
fcis-30390	26	2	research	research	NOUN
fcis-30390	26	3	was	be	AUX
fcis-30390	26	4	represented	represent	VERB
fcis-30390	26	5	by	by	ADP
fcis-30390	26	6	the	the	DET
fcis-30390	26	7	arima	arima	PROPN
fcis-30390	26	8	model	model	NOUN
fcis-30390	26	9	proposed	propose	VERB
fcis-30390	26	10	by	by	ADP
fcis-30390	26	11	box	box	PROPN
fcis-30390	26	12	et	et	PROPN
fcis-30390	26	13	al.[1	al.[1	PROPN
fcis-30390	26	14	]	]	PUNCT
fcis-30390	26	15	and	and	CCONJ
fcis-30390	26	16	systematic	systematic	ADJ
fcis-30390	26	17	forecasting	forecasting	NOUN
fcis-30390	26	18	principles	principle	NOUN
fcis-30390	26	19	by	by	ADP
fcis-30390	26	20	hyndman[2	hyndman[2	PROPN
fcis-30390	26	21	]	]	PUNCT
fcis-30390	26	22	,	,	PUNCT
fcis-30390	26	23	which	which	PRON
fcis-30390	26	24	established	establish	VERB
fcis-30390	26	25	a	a	DET
fcis-30390	26	26	solid	solid	ADJ
fcis-30390	26	27	theoretical	theoretical	ADJ
fcis-30390	26	28	foundation	foundation	NOUN
fcis-30390	26	29	for	for	ADP
fcis-30390	26	30	time	time	NOUN
fcis-30390	26	31	series	series	PROPN
fcis-30390	26	32	analysis	analysis	NOUN
fcis-30390	26	33	.	.	PUNCT
fcis-30390	27	1	the	the	DET
fcis-30390	27	2	advent	advent	NOUN
fcis-30390	27	3	of	of	ADP
fcis-30390	27	4	deep	deep	ADJ
fcis-30390	27	5	learning	learning	NOUN
fcis-30390	27	6	brought	bring	VERB
fcis-30390	27	7	revolutionary	revolutionary	ADJ
fcis-30390	27	8	breakthroughs	breakthrough	NOUN
fcis-30390	27	9	to	to	ADP
fcis-30390	27	10	the	the	DET
fcis-30390	27	11	field	field	NOUN
fcis-30390	27	12	,	,	PUNCT
fcis-30390	27	13	with	with	ADP
fcis-30390	27	14	graves	grave	NOUN
fcis-30390	27	15	et	et	NOUN
fcis-30390	27	16	al.[3	al.[3	NOUN
fcis-30390	27	17	]	]	PUNCT
fcis-30390	27	18	proposing	propose	VERB
fcis-30390	27	19	lstm	lstm	NOUN
fcis-30390	27	20	,	,	PUNCT
fcis-30390	27	21	which	which	PRON
fcis-30390	27	22	effectively	effectively	ADV
fcis-30390	27	23	addressed	address	VERB
fcis-30390	27	24	long	long	ADJ
fcis-30390	27	25	-	-	PUNCT
fcis-30390	27	26	term	term	NOUN
fcis-30390	27	27	dependency	dependency	NOUN
fcis-30390	27	28	problems	problem	NOUN
fcis-30390	27	29	through	through	ADP
fcis-30390	27	30	sophisticated	sophisticated	ADJ
fcis-30390	27	31	gating	gate	VERB
fcis-30390	27	32	mechanisms	mechanism	NOUN
fcis-30390	27	33	,	,	PUNCT
fcis-30390	27	34	while	while	SCONJ
fcis-30390	27	35	the	the	DET
fcis-30390	27	36	transformer	transformer	NOUN
fcis-30390	27	37	architecture	architecture	NOUN
fcis-30390	27	38	proposed	propose	VERB
fcis-30390	27	39	by	by	ADP
fcis-30390	27	40	vaswani	vaswani	PROPN
fcis-30390	27	41	et	et	NOUN
fcis-30390	27	42	al.[4	al.[4	PROPN
fcis-30390	27	43	]	]	PUNCT
fcis-30390	27	44	36	36	NUM
fcis-30390	27	45	initiated	initiate	VERB
fcis-30390	27	46	a	a	DET
fcis-30390	27	47	new	new	ADJ
fcis-30390	27	48	research	research	NOUN
fcis-30390	27	49	wave	wave	NOUN
fcis-30390	27	50	with	with	ADP
fcis-30390	27	51	its	its	PRON
fcis-30390	27	52	self	self	NOUN
fcis-30390	27	53	-	-	PUNCT
fcis-30390	27	54	attention	attention	NOUN
fcis-30390	27	55	mechanism.building	mechanism.builde	VERB
fcis-30390	27	56	upon	upon	SCONJ
fcis-30390	27	57	these	these	DET
fcis-30390	27	58	foundations	foundation	NOUN
fcis-30390	27	59	,	,	PUNCT
fcis-30390	27	60	fedformer	fedformer	NOUN
fcis-30390	27	61	proposed	propose	VERB
fcis-30390	27	62	by	by	ADP
fcis-30390	27	63	zhou	zhou	PROPN
fcis-30390	27	64	et	et	PROPN
fcis-30390	27	65	al.[5	al.[5	PROPN
fcis-30390	27	66	]	]	PUNCT
fcis-30390	27	67	enhanced	enhance	VERB
fcis-30390	27	68	periodic	periodic	ADJ
fcis-30390	27	69	pattern	pattern	NOUN
fcis-30390	27	70	modeling	model	VERB
fcis-30390	27	71	capabilities	capability	NOUN
fcis-30390	27	72	through	through	ADP
fcis-30390	27	73	frequency	frequency	NOUN
fcis-30390	27	74	domain	domain	NOUN
fcis-30390	27	75	decomposition	decomposition	NOUN
fcis-30390	27	76	,	,	PUNCT
fcis-30390	27	77	while	while	SCONJ
fcis-30390	27	78	wu	wu	PROPN
fcis-30390	27	79	et	et	PROPN
fcis-30390	27	80	al.[6	al.[6	PROPN
fcis-30390	27	81	]	]	PUNCT
fcis-30390	27	82	systematically	systematically	ADV
fcis-30390	27	83	validated	validate	VERB
fcis-30390	27	84	the	the	DET
fcis-30390	27	85	effectiveness	effectiveness	NOUN
fcis-30390	27	86	of	of	ADP
fcis-30390	27	87	transformer	transformer	NOUN
fcis-30390	27	88	architectures	architecture	NOUN
fcis-30390	27	89	in	in	ADP
fcis-30390	27	90	time	time	NOUN
fcis-30390	27	91	series	series	PROPN
fcis-30390	27	92	forecasting	forecasting	NOUN
fcis-30390	27	93	.	.	PUNCT
fcis-30390	28	1	addressing	address	VERB
fcis-30390	28	2	non	non	ADJ
fcis-30390	28	3	-	-	ADJ
fcis-30390	28	4	stationarity	stationarity	ADJ
fcis-30390	28	5	issues	issue	NOUN
fcis-30390	28	6	,	,	PUNCT
fcis-30390	28	7	liu	liu	PROPN
fcis-30390	28	8	et	et	PROPN
fcis-30390	28	9	al.[7	al.[7	PROPN
fcis-30390	28	10	]	]	PUNCT
fcis-30390	28	11	introduced	introduce	VERB
fcis-30390	28	12	the	the	DET
fcis-30390	28	13	non	non	ADJ
fcis-30390	28	14	-	-	ADJ
fcis-30390	28	15	stationary	stationary	ADJ
fcis-30390	28	16	transformer	transformer	NOUN
fcis-30390	28	17	with	with	ADP
fcis-30390	28	18	innovative	innovative	ADJ
fcis-30390	28	19	adaptive	adaptive	ADJ
fcis-30390	28	20	normalization	normalization	NOUN
fcis-30390	28	21	mechanisms.in	mechanisms.in	AUX
fcis-30390	28	22	terms	term	NOUN
fcis-30390	28	23	of	of	ADP
fcis-30390	28	24	multi	multi	ADJ
fcis-30390	28	25	-	-	ADJ
fcis-30390	28	26	scale	scale	ADJ
fcis-30390	28	27	modeling	modeling	NOUN
fcis-30390	28	28	,	,	PUNCT
fcis-30390	28	29	timesnet	timesnet	VERB
fcis-30390	28	30	by	by	ADP
fcis-30390	28	31	wu	wu	PROPN
fcis-30390	28	32	et	et	PROPN
fcis-30390	28	33	al.[8	al.[8	PROPN
fcis-30390	28	34	]	]	PUNCT
fcis-30390	28	35	achieved	achieve	VERB
fcis-30390	28	36	effective	effective	ADJ
fcis-30390	28	37	multiscale	multiscale	ADJ
fcis-30390	28	38	feature	feature	NOUN
fcis-30390	28	39	capture	capture	NOUN
fcis-30390	28	40	through	through	ADP
fcis-30390	28	41	temporal	temporal	ADJ
fcis-30390	28	42	2d	2d	NUM
fcis-30390	28	43	variation	variation	NOUN
fcis-30390	28	44	modeling	modeling	NOUN
fcis-30390	28	45	,	,	PUNCT
fcis-30390	28	46	while	while	SCONJ
fcis-30390	28	47	scaleformer	scaleformer	NOUN
fcis-30390	28	48	proposed	propose	VERB
fcis-30390	28	49	by	by	ADP
fcis-30390	28	50	shabani	shabani	PROPN
fcis-30390	28	51	et	et	PROPN
fcis-30390	28	52	al.[9	al.[9	PROPN
fcis-30390	28	53	]	]	PUNCT
fcis-30390	28	54	further	far	ADV
fcis-30390	28	55	improved	improve	VERB
fcis-30390	28	56	prediction	prediction	NOUN
fcis-30390	28	57	accuracy	accuracy	NOUN
fcis-30390	28	58	using	use	VERB
fcis-30390	28	59	iterative	iterative	NOUN
fcis-30390	28	60	refinement	refinement	NOUN
fcis-30390	28	61	strategies	strategy	NOUN
fcis-30390	28	62	.	.	PUNCT
fcis-30390	29	1	recent	recent	ADJ
fcis-30390	29	2	research	research	NOUN
fcis-30390	29	3	has	have	AUX
fcis-30390	29	4	focused	focus	VERB
fcis-30390	29	5	more	more	ADJ
fcis-30390	29	6	on	on	ADP
fcis-30390	29	7	quantifying	quantify	VERB
fcis-30390	29	8	prediction	prediction	NOUN
fcis-30390	29	9	uncertainty	uncertainty	NOUN
fcis-30390	29	10	,	,	PUNCT
fcis-30390	29	11	with	with	ADP
fcis-30390	29	12	lin	lin	PROPN
fcis-30390	29	13	et	et	PROPN
fcis-30390	29	14	al.[10	al.[10	PROPN
fcis-30390	29	15	]	]	PUNCT
fcis-30390	29	16	providing	provide	VERB
fcis-30390	29	17	a	a	DET
fcis-30390	29	18	comprehensive	comprehensive	ADJ
fcis-30390	29	19	review	review	NOUN
fcis-30390	29	20	of	of	ADP
fcis-30390	29	21	diffusion	diffusion	NOUN
fcis-30390	29	22	models	model	NOUN
fcis-30390	29	23	in	in	ADP
fcis-30390	29	24	time	time	NOUN
fcis-30390	29	25	series	series	PROPN
fcis-30390	29	26	applications	application	NOUN
fcis-30390	29	27	,	,	PUNCT
fcis-30390	29	28	while	while	SCONJ
fcis-30390	29	29	the	the	DET
fcis-30390	29	30	koopa	koopa	PROPN
fcis-30390	29	31	architecture	architecture	NOUN
fcis-30390	29	32	proposed	propose	VERB
fcis-30390	29	33	by	by	ADP
fcis-30390	29	34	liu	liu	PROPN
fcis-30390	29	35	et	et	PROPN
fcis-30390	29	36	al.[11	al.[11	PROPN
fcis-30390	29	37	]	]	PUNCT
fcis-30390	29	38	innovatively	innovatively	ADV
fcis-30390	29	39	incorporated	incorporate	VERB
fcis-30390	29	40	koopman	koopman	NOUN
fcis-30390	29	41	operator	operator	NOUN
fcis-30390	29	42	theory.regarding	theory.regarde	VERB
fcis-30390	29	43	memory	memory	NOUN
fcis-30390	29	44	enhancement	enhancement	NOUN
fcis-30390	29	45	mechanisms	mechanism	NOUN
fcis-30390	29	46	,	,	PUNCT
fcis-30390	29	47	lee	lee	PROPN
fcis-30390	29	48	et	et	PROPN
fcis-30390	29	49	al.[12	al.[12	PROPN
fcis-30390	29	50	]	]	PUNCT
fcis-30390	29	51	and	and	CCONJ
fcis-30390	29	52	liu	liu	PROPN
fcis-30390	29	53	et	et	PROPN
fcis-30390	29	54	al.[13	al.[13	NOUN
fcis-30390	29	55	]	]	PUNCT
fcis-30390	29	56	proposed	propose	VERB
fcis-30390	29	57	pattern	pattern	NOUN
fcis-30390	29	58	matching	matching	NOUN
fcis-30390	29	59	memory	memory	NOUN
fcis-30390	29	60	networks	network	NOUN
fcis-30390	29	61	for	for	ADP
fcis-30390	29	62	traffic	traffic	NOUN
fcis-30390	29	63	prediction	prediction	NOUN
fcis-30390	29	64	and	and	CCONJ
fcis-30390	29	65	graph	graph	NOUN
fcis-30390	29	66	learning	learn	VERB
fcis-30390	29	67	memory	memory	NOUN
fcis-30390	29	68	networks	network	NOUN
fcis-30390	29	69	for	for	ADP
fcis-30390	29	70	multivariate	multivariate	NOUN
fcis-30390	29	71	time	time	NOUN
fcis-30390	29	72	series	series	PROPN
fcis-30390	29	73	prediction	prediction	PROPN
fcis-30390	29	74	,	,	PUNCT
fcis-30390	29	75	respectively	respectively	ADV
fcis-30390	29	76	.	.	PUNCT
fcis-30390	30	1	additionally	additionally	ADV
fcis-30390	30	2	,	,	PUNCT
fcis-30390	30	3	crossformer	crossformer	NOUN
fcis-30390	30	4	proposed	propose	VERB
fcis-30390	30	5	by	by	ADP
fcis-30390	30	6	zhang	zhang	PROPN
fcis-30390	30	7	et	et	PROPN
fcis-30390	30	8	al.[14	al.[14	PROPN
fcis-30390	30	9	]	]	PUNCT
fcis-30390	30	10	provided	provide	VERB
fcis-30390	30	11	new	new	ADJ
fcis-30390	30	12	solutions	solution	NOUN
fcis-30390	30	13	for	for	ADP
fcis-30390	30	14	multivariate	multivariate	NOUN
fcis-30390	30	15	time	time	NOUN
fcis-30390	30	16	series	series	NOUN
fcis-30390	30	17	prediction	prediction	NOUN
fcis-30390	30	18	by	by	ADP
fcis-30390	30	19	leveraging	leverage	VERB
fcis-30390	30	20	cross	cross	ADJ
fcis-30390	30	21	-	-	ADJ
fcis-30390	30	22	dimensional	dimensional	ADJ
fcis-30390	30	23	dependencies	dependency	NOUN
fcis-30390	30	24	.	.	PUNCT
fcis-30390	31	1	3	3	X
fcis-30390	31	2	.	.	X
fcis-30390	31	3	timemixer	timemixer	NOUN
fcis-30390	31	4	-	-	PUNCT
fcis-30390	31	5	me	i	PRON
fcis-30390	31	6	input	input	PROPN
fcis-30390	31	7	series	series	PROPN
fcis-30390	31	8	dft	dft	PROPN
fcis-30390	31	9	series	series	PROPN
fcis-30390	31	10	decomposition	decomposition	NOUN
fcis-30390	31	11	frequency	frequency	NOUN
fcis-30390	31	12	domain	domain	NOUN
fcis-30390	31	13	decomposition	decomposition	NOUN
fcis-30390	31	14	multiscaleseasonmixing	multiscaleseasonmixe	VERB
fcis-30390	31	15	bottom	bottom	ADJ
fcis-30390	31	16	-	-	PUNCT
fcis-30390	31	17	up	up	ADP
fcis-30390	31	18	-	-	PUNCT
fcis-30390	31	19	mixing	mix	VERB
fcis-30390	31	20	multiscaletrendmixing	multiscaletrendmixe	VERB
fcis-30390	31	21	up	up	ADP
fcis-30390	31	22	-	-	PUNCT
fcis-30390	31	23	bottom	bottom	ADV
fcis-30390	31	24	-	-	PUNCT
fcis-30390	31	25	mixing	mix	VERB
fcis-30390	31	26	anyvariateattention	anyvariateattention	NOUN
fcis-30390	31	27	multi	multi	ADJ
fcis-30390	31	28	-	-	ADJ
fcis-30390	31	29	timescale	timescale	ADJ
fcis-30390	31	30	memory	memory	NOUN
fcis-30390	31	31	enhancement	enhancement	NOUN
fcis-30390	31	32	mechanism	mechanism	NOUN
fcis-30390	31	33	(	(	PUNCT
fcis-30390	31	34	mtmem	mtmem	NOUN
fcis-30390	31	35	)	)	PUNCT
fcis-30390	31	36	short_term	short_term	VERB
fcis-30390	31	37	_	_	NOUN
fcis-30390	31	38	memory	memory	NOUN
fcis-30390	31	39	central	central	ADJ
fcis-30390	31	40	memory	memory	NOUN
fcis-30390	31	41	long_term	long_term	PROPN
fcis-30390	31	42	_	_	NOUN
fcis-30390	31	43	memory	memory	NOUN
fcis-30390	31	44	result	result	NOUN
fcis-30390	31	45	figure	figure	NOUN
fcis-30390	31	46	1	1	NUM
fcis-30390	31	47	.	.	PUNCT
fcis-30390	32	1	the	the	DET
fcis-30390	32	2	overall	overall	ADJ
fcis-30390	32	3	architecture	architecture	NOUN
fcis-30390	32	4	of	of	ADP
fcis-30390	32	5	timemixer	timemixer	NOUN
fcis-30390	32	6	-	-	PUNCT
fcis-30390	32	7	me	i	PRON
fcis-30390	32	8	we	we	PRON
fcis-30390	32	9	extend	extend	VERB
fcis-30390	32	10	the	the	DET
fcis-30390	32	11	baseline	baseline	PROPN
fcis-30390	32	12	timemixer	timemixer	PROPN
fcis-30390	32	13	model	model	NOUN
fcis-30390	32	14	which	which	PRON
fcis-30390	32	15	includes	include	VERB
fcis-30390	32	16	dft	dft	PROPN
fcis-30390	32	17	series	series	NOUN
fcis-30390	32	18	decomposition	decomposition	NOUN
fcis-30390	32	19	and	and	CCONJ
fcis-30390	32	20	multiscalemixing	multiscalemixe	VERB
fcis-30390	32	21	architecture	architecture	NOUN
fcis-30390	32	22	modules	module	NOUN
fcis-30390	32	23	by	by	ADP
fcis-30390	32	24	introducing	introduce	VERB
fcis-30390	32	25	the	the	DET
fcis-30390	32	26	anyvariateattention	anyvariateattention	NOUN
fcis-30390	32	27	module	module	NOUN
fcis-30390	32	28	and	and	CCONJ
fcis-30390	32	29	proposing	propose	VERB
fcis-30390	32	30	the	the	DET
fcis-30390	32	31	mtmem	mtmem	NOUN
fcis-30390	32	32	module	module	NOUN
fcis-30390	32	33	.	.	PUNCT
fcis-30390	33	1	3.1	3.1	NUM
fcis-30390	33	2	.	.	PUNCT
fcis-30390	33	3	anyvariateattention	anyvariateattention	NOUN
fcis-30390	33	4	input	input	PROPN
fcis-30390	33	5	layernorm	layernorm	PROPN
fcis-30390	33	6	k‐>k(2d	k‐>k(2d	PROPN
fcis-30390	33	7	)	)	PUNCT
fcis-30390	33	8	v‐>v(d	v‐>v(d	NOUN
fcis-30390	33	9	)	)	PUNCT
fcis-30390	33	10	q‐>q(2d	q‐>q(2d	PROPN
fcis-30390	33	11	)	)	PUNCT
fcis-30390	33	12	v	v	ADP
fcis-30390	33	13	k	k	X
fcis-30390	33	14	q	q	PROPN
fcis-30390	33	15	scaled	scale	VERB
fcis-30390	33	16	dotproduct	dotproduct	NOUN
fcis-30390	33	17	attention	attention	NOUN
fcis-30390	33	18	+	+	CCONJ
fcis-30390	33	19	u1	u1	NOUN
fcis-30390	33	20	/	/	SYM
fcis-30390	33	21	u2	u2	PROPN
fcis-30390	33	22	softmax	softmax	PROPN
fcis-30390	33	23	dropout	dropout	PROPN
fcis-30390	33	24	output_proj	output_proj	PROPN
fcis-30390	33	25	ection	ection	NOUN
fcis-30390	33	26	+	+	ADP
fcis-30390	33	27	residual	residual	ADJ
fcis-30390	33	28	output	output	NOUN
fcis-30390	33	29	figure	figure	NOUN
fcis-30390	33	30	2	2	NUM
fcis-30390	33	31	.	.	PUNCT
fcis-30390	34	1	the	the	DET
fcis-30390	34	2	architectural	architectural	ADJ
fcis-30390	34	3	diagram	diagram	NOUN
fcis-30390	34	4	of	of	ADP
fcis-30390	34	5	the	the	DET
fcis-30390	34	6	anyvariateattention	anyvariateattention	NOUN
fcis-30390	34	7	module	module	NOUN
fcis-30390	34	8	the	the	DET
fcis-30390	34	9	input	input	NOUN
fcis-30390	34	10	features	feature	VERB
fcis-30390	34	11	first	first	ADV
fcis-30390	34	12	undergo	undergo	VERB
fcis-30390	34	13	layernorm	layernorm	NOUN
fcis-30390	34	14	preprocessing	preprocessing	NOUN
fcis-30390	34	15	,	,	PUNCT
fcis-30390	34	16	followed	follow	VERB
fcis-30390	34	17	by	by	ADP
fcis-30390	34	18	parallel	parallel	ADJ
fcis-30390	34	19	processing	processing	NOUN
fcis-30390	34	20	through	through	ADP
fcis-30390	34	21	three	three	NUM
fcis-30390	34	22	branches	branch	NOUN
fcis-30390	34	23	to	to	PART
fcis-30390	34	24	generate	generate	VERB
fcis-30390	34	25	query	query	NOUN
fcis-30390	34	26	(	(	PUNCT
fcis-30390	34	27	q	q	NOUN
fcis-30390	34	28	)	)	PUNCT
fcis-30390	34	29	,	,	PUNCT
fcis-30390	34	30	key	key	ADJ
fcis-30390	34	31	(	(	PUNCT
fcis-30390	34	32	k	k	NOUN
fcis-30390	34	33	)	)	PUNCT
fcis-30390	34	34	,	,	PUNCT
fcis-30390	34	35	and	and	CCONJ
fcis-30390	34	36	value	value	NOUN
fcis-30390	34	37	(	(	PUNCT
fcis-30390	34	38	v	v	NOUN
fcis-30390	34	39	)	)	PUNCT
fcis-30390	34	40	representations	representation	NOUN
fcis-30390	34	41	.	.	PUNCT
fcis-30390	35	1	these	these	DET
fcis-30390	35	2	representations	representation	NOUN
fcis-30390	35	3	are	be	AUX
fcis-30390	35	4	then	then	ADV
fcis-30390	35	5	transformed	transform	VERB
fcis-30390	35	6	into	into	ADP
fcis-30390	35	7	their	their	PRON
fcis-30390	35	8	respective	respective	ADJ
fcis-30390	35	9	2d	2d	NOUN
fcis-30390	35	10	formats	format	NOUN
fcis-30390	35	11	(	(	PUNCT
fcis-30390	35	12	q->q(2d	q->q(2d	PROPN
fcis-30390	35	13	)	)	PUNCT
fcis-30390	35	14	,	,	PUNCT
fcis-30390	35	15	k->k(2d	k->k(2d	PROPN
fcis-30390	35	16	)	)	PUNCT
fcis-30390	35	17	,	,	PUNCT
fcis-30390	35	18	v->v(d	v->v(d	NOUN
fcis-30390	35	19	)	)	PUNCT
fcis-30390	35	20	)	)	PUNCT
fcis-30390	35	21	.	.	PUNCT
fcis-30390	36	1	the	the	DET
fcis-30390	36	2	key	key	ADJ
fcis-30390	36	3	and	and	CCONJ
fcis-30390	36	4	query	query	NOUN
fcis-30390	36	5	branches	branch	NOUN
fcis-30390	36	6	feed	feed	VERB
fcis-30390	36	7	into	into	ADP
fcis-30390	36	8	a	a	DET
fcis-30390	36	9	scaled	scale	VERB
fcis-30390	36	10	dot	dot	NOUN
fcis-30390	36	11	-	-	PUNCT
fcis-30390	36	12	product	product	NOUN
fcis-30390	36	13	attention	attention	NOUN
fcis-30390	36	14	mechanism	mechanism	NOUN
fcis-30390	36	15	,	,	PUNCT
fcis-30390	36	16	which	which	PRON
fcis-30390	36	17	is	be	AUX
fcis-30390	36	18	enhanced	enhance	VERB
fcis-30390	36	19	by	by	ADP
fcis-30390	36	20	a	a	DET
fcis-30390	36	21	u1	u1	NOUN
fcis-30390	36	22	/	/	SYM
fcis-30390	36	23	u2	u2	NOUN
fcis-30390	36	24	scaling	scale	VERB
fcis-30390	36	25	factor	factor	NOUN
fcis-30390	36	26	before	before	ADP
fcis-30390	36	27	passing	pass	VERB
fcis-30390	36	28	through	through	ADP
fcis-30390	36	29	softmax	softmax	NOUN
fcis-30390	36	30	and	and	CCONJ
fcis-30390	36	31	dropout	dropout	NOUN
fcis-30390	36	32	layers	layer	NOUN
fcis-30390	36	33	.	.	PUNCT
fcis-30390	37	1	the	the	DET
fcis-30390	37	2	attention	attention	NOUN
fcis-30390	37	3	output	output	NOUN
fcis-30390	37	4	is	be	AUX
fcis-30390	37	5	combined	combine	VERB
fcis-30390	37	6	with	with	ADP
fcis-30390	37	7	the	the	DET
fcis-30390	37	8	processed	process	VERB
fcis-30390	37	9	value	value	NOUN
fcis-30390	37	10	branch	branch	NOUN
fcis-30390	37	11	in	in	ADP
fcis-30390	37	12	the	the	DET
fcis-30390	37	13	output_projection	output_projection	NOUN
fcis-30390	37	14	layer	layer	NOUN
fcis-30390	37	15	,	,	PUNCT
fcis-30390	37	16	and	and	CCONJ
fcis-30390	37	17	finally	finally	ADV
fcis-30390	37	18	merged	merge	VERB
fcis-30390	37	19	with	with	ADP
fcis-30390	37	20	a	a	DET
fcis-30390	37	21	residual	residual	ADJ
fcis-30390	37	22	connection	connection	NOUN
fcis-30390	37	23	to	to	PART
fcis-30390	37	24	produce	produce	VERB
fcis-30390	37	25	the	the	DET
fcis-30390	37	26	final	final	ADJ
fcis-30390	37	27	output	output	NOUN
fcis-30390	37	28	.	.	PUNCT
fcis-30390	38	1	the	the	DET
fcis-30390	38	2	complex	complex	ADJ
fcis-30390	38	3	interactions	interaction	NOUN
fcis-30390	38	4	between	between	ADP
fcis-30390	38	5	multivariate	multivariate	NOUN
fcis-30390	38	6	time	time	NOUN
fcis-30390	38	7	series	series	PROPN
fcis-30390	38	8	exhibit	exhibit	VERB
fcis-30390	38	9	unique	unique	ADJ
fcis-30390	38	10	coupling	coupling	NOUN
fcis-30390	38	11	characteristics	characteristic	NOUN
fcis-30390	38	12	across	across	ADP
fcis-30390	38	13	different	different	ADJ
fcis-30390	38	14	temporal	temporal	ADJ
fcis-30390	38	15	and	and	CCONJ
fcis-30390	38	16	feature	feature	NOUN
fcis-30390	38	17	dimensions	dimension	NOUN
fcis-30390	38	18	.	.	PUNCT
fcis-30390	39	1	given	give	VERB
fcis-30390	39	2	an	an	DET
fcis-30390	39	3	input	input	NOUN
fcis-30390	39	4	sequence	sequence	NOUN
fcis-30390	39	5	:	:	PUNCT
fcis-30390	39	6	∈	∈	PROPN
fcis-30390	39	7	1	1	NUM
fcis-30390	39	8	where	where	SCONJ
fcis-30390	39	9	b	b	NOUN
fcis-30390	39	10	represents	represent	VERB
fcis-30390	39	11	batch	batch	NOUN
fcis-30390	39	12	size	size	NOUN
fcis-30390	39	13	,	,	PUNCT
fcis-30390	39	14	t	t	PROPN
fcis-30390	39	15	denotes	denote	VERB
fcis-30390	39	16	sequence	sequence	NOUN
fcis-30390	39	17	length	length	NOUN
fcis-30390	39	18	,	,	PUNCT
fcis-30390	39	19	n	n	PRON
fcis-30390	39	20	indicates	indicate	VERB
fcis-30390	39	21	the	the	DET
fcis-30390	39	22	number	number	NOUN
fcis-30390	39	23	of	of	ADP
fcis-30390	39	24	variables	variable	NOUN
fcis-30390	39	25	,	,	PUNCT
fcis-30390	39	26	and	and	CCONJ
fcis-30390	39	27	d	d	NOUN
fcis-30390	39	28	is	be	AUX
fcis-30390	39	29	the	the	DET
fcis-30390	39	30	feature	feature	NOUN
fcis-30390	39	31	dimension	dimension	NOUN
fcis-30390	39	32	,	,	PUNCT
fcis-30390	39	33	anyvariateattention	anyvariateattention	NOUN
fcis-30390	39	34	enhances	enhance	VERB
fcis-30390	39	35	inter	inter	ADJ
fcis-30390	39	36	-	-	ADJ
fcis-30390	39	37	variable	variable	ADJ
fcis-30390	39	38	feature	feature	NOUN
fcis-30390	39	39	representation	representation	NOUN
fcis-30390	39	40	through	through	ADP
fcis-30390	39	41	the	the	DET
fcis-30390	39	42	following	follow	VERB
fcis-30390	39	43	integrated	integrate	VERB
fcis-30390	39	44	mechanism	mechanism	NOUN
fcis-30390	39	45	:	:	PUNCT
fcis-30390	39	46	the	the	DET
fcis-30390	39	47	module	module	NOUN
fcis-30390	39	48	first	first	ADJ
fcis-30390	39	49	projects	project	NOUN
fcis-30390	39	50	input	input	NOUN
fcis-30390	39	51	features	feature	NOUN
fcis-30390	39	52	into	into	ADP
fcis-30390	39	53	a	a	DET
fcis-30390	39	54	37	37	NUM
fcis-30390	39	55	higher	higher	ADV
fcis-30390	39	56	-	-	PUNCT
fcis-30390	39	57	dimensional	dimensional	ADJ
fcis-30390	39	58	representation	representation	NOUN
fcis-30390	39	59	space	space	NOUN
fcis-30390	39	60	to	to	PART
fcis-30390	39	61	better	well	ADV
fcis-30390	39	62	capture	capture	VERB
fcis-30390	39	63	multi	multi	ADJ
fcis-30390	39	64	-	-	ADJ
fcis-30390	39	65	scale	scale	ADJ
fcis-30390	39	66	feature	feature	NOUN
fcis-30390	39	67	relationships	relationship	NOUN
fcis-30390	39	68	,	,	PUNCT
fcis-30390	39	69	where	where	SCONJ
fcis-30390	39	70	:	:	PUNCT
fcis-30390	39	71	∈	∈	PROPN
fcis-30390	39	72	2	2	NUM
fcis-30390	39	73	∈	∈	NOUN
fcis-30390	39	74	3	3	NUM
fcis-30390	39	75	∈	∈	NOUN
fcis-30390	39	76	4	4	NUM
fcis-30390	39	77	this	this	DET
fcis-30390	39	78	double	double	ADJ
fcis-30390	39	79	-	-	PUNCT
fcis-30390	39	80	dimension	dimension	NOUN
fcis-30390	39	81	projection	projection	NOUN
fcis-30390	39	82	design	design	NOUN
fcis-30390	39	83	significantly	significantly	ADV
fcis-30390	39	84	enhances	enhance	VERB
fcis-30390	39	85	the	the	DET
fcis-30390	39	86	model	model	NOUN
fcis-30390	39	87	's	's	PART
fcis-30390	39	88	expressiveness	expressiveness	NOUN
fcis-30390	39	89	.	.	PUNCT
fcis-30390	40	1	to	to	PART
fcis-30390	40	2	strengthen	strengthen	VERB
fcis-30390	40	3	positional	positional	ADJ
fcis-30390	40	4	awareness	awareness	NOUN
fcis-30390	40	5	,	,	PUNCT
fcis-30390	40	6	a	a	DET
fcis-30390	40	7	learnable	learnable	ADJ
fcis-30390	40	8	rotary	rotary	NOUN
fcis-30390	40	9	position	position	NOUN
fcis-30390	40	10	encoding	encoding	NOUN
fcis-30390	40	11	matrix	matrix	NOUN
fcis-30390	40	12	r	r	NOUN
fcis-30390	40	13	is	be	AUX
fcis-30390	40	14	introduced	introduce	VERB
fcis-30390	40	15	,	,	PUNCT
fcis-30390	40	16	where	where	SCONJ
fcis-30390	40	17	:	:	PUNCT
fcis-30390	40	18	,	,	PUNCT
fcis-30390	40	19	∈	∈	PROPN
fcis-30390	40	20	5	5	NUM
fcis-30390	40	21	which	which	PRON
fcis-30390	40	22	adaptively	adaptively	ADV
fcis-30390	40	23	learns	learn	VERB
fcis-30390	40	24	positional	positional	ADJ
fcis-30390	40	25	dependencies	dependency	NOUN
fcis-30390	40	26	through	through	ADP
fcis-30390	40	27	gradient	gradient	ADJ
fcis-30390	40	28	descent	descent	NOUN
fcis-30390	40	29	.	.	PUNCT
fcis-30390	41	1	the	the	DET
fcis-30390	41	2	attention	attention	NOUN
fcis-30390	41	3	computation	computation	NOUN
fcis-30390	41	4	considers	consider	VERB
fcis-30390	41	5	intervariable	intervariable	ADJ
fcis-30390	41	6	relationships	relationship	NOUN
fcis-30390	41	7	through	through	ADP
fcis-30390	41	8	:	:	PUNCT
fcis-30390	41	9	√2	√2	NOUN
fcis-30390	41	10	∈	∈	NOUN
fcis-30390	41	11	6	6	NUM
fcis-30390	41	12	incorporating	incorporate	VERB
fcis-30390	41	13	variable	variable	ADJ
fcis-30390	41	14	-	-	PUNCT
fcis-30390	41	15	specific	specific	ADJ
fcis-30390	41	16	bias	bias	NOUN
fcis-30390	41	17	terms	term	NOUN
fcis-30390	41	18	:	:	PUNCT
fcis-30390	41	19	1	1	NUM
fcis-30390	41	20	7	7	NUM
fcis-30390	41	21	to	to	PART
fcis-30390	41	22	distinguish	distinguish	VERB
fcis-30390	41	23	relationships	relationship	NOUN
fcis-30390	41	24	between	between	ADP
fcis-30390	41	25	same	same	ADJ
fcis-30390	41	26	and	and	CCONJ
fcis-30390	41	27	different	different	ADJ
fcis-30390	41	28	variables	variable	NOUN
fcis-30390	41	29	,	,	PUNCT
fcis-30390	41	30	resulting	result	VERB
fcis-30390	41	31	in	in	ADP
fcis-30390	41	32	:	:	PUNCT
fcis-30390	41	33	8	8	NUM
fcis-30390	41	34	where	where	SCONJ
fcis-30390	41	35	and	and	CCONJ
fcis-30390	41	36	are	be	AUX
fcis-30390	41	37	learnable	learnable	ADJ
fcis-30390	41	38	scalar	scalar	ADJ
fcis-30390	41	39	parameters	parameter	NOUN
fcis-30390	41	40	and	and	CCONJ
fcis-30390	41	41	is	be	AUX
fcis-30390	41	42	the	the	DET
fcis-30390	41	43	n	n	ADV
fcis-30390	41	44	-	-	PUNCT
fcis-30390	41	45	dimensional	dimensional	ADJ
fcis-30390	41	46	identity	identity	NOUN
fcis-30390	41	47	matrix	matrix	NOUN
fcis-30390	41	48	.	.	PUNCT
fcis-30390	42	1	the	the	DET
fcis-30390	42	2	final	final	ADJ
fcis-30390	42	3	attention	attention	NOUN
fcis-30390	42	4	weights	weight	NOUN
fcis-30390	42	5	are	be	AUX
fcis-30390	42	6	computed	compute	VERB
fcis-30390	42	7	through	through	ADP
fcis-30390	42	8	:	:	PUNCT
fcis-30390	42	9	softmax	softmax	NOUN
fcis-30390	42	10	∈	∈	NOUN
fcis-30390	42	11	9	9	NUM
fcis-30390	42	12	followed	follow	VERB
fcis-30390	42	13	by	by	ADP
fcis-30390	42	14	feature	feature	NOUN
fcis-30390	42	15	aggregation	aggregation	NOUN
fcis-30390	42	16	:	:	PUNCT
fcis-30390	42	17	∈	∈	PROPN
fcis-30390	42	18	10	10	NUM
fcis-30390	42	19	layer	layer	NOUN
fcis-30390	42	20	normalization	normalization	NOUN
fcis-30390	42	21	and	and	CCONJ
fcis-30390	42	22	residual	residual	ADJ
fcis-30390	42	23	connections	connection	NOUN
fcis-30390	42	24	ensure	ensure	VERB
fcis-30390	42	25	feature	feature	NOUN
fcis-30390	42	26	distribution	distribution	NOUN
fcis-30390	42	27	stability	stability	NOUN
fcis-30390	42	28	through	through	ADP
fcis-30390	42	29	layernorm	layernorm	NOUN
fcis-30390	42	30	11	11	NUM
fcis-30390	42	31	and	and	CCONJ
fcis-30390	42	32	12	12	NUM
fcis-30390	42	33	where	where	SCONJ
fcis-30390	42	34	is	be	AUX
fcis-30390	42	35	the	the	DET
fcis-30390	42	36	output	output	NOUN
fcis-30390	42	37	projection	projection	NOUN
fcis-30390	42	38	matrix	matrix	NOUN
fcis-30390	42	39	.	.	PUNCT
fcis-30390	43	1	the	the	DET
fcis-30390	43	2	complete	complete	ADJ
fcis-30390	43	3	forward	forward	ADJ
fcis-30390	43	4	propagation	propagation	NOUN
fcis-30390	43	5	process	process	NOUN
fcis-30390	43	6	can	can	AUX
fcis-30390	43	7	be	be	AUX
fcis-30390	43	8	expressed	express	VERB
fcis-30390	43	9	as	as	ADP
fcis-30390	43	10	:	:	PUNCT
fcis-30390	43	11	anyvariateattention	anyvariateattention	NOUN
fcis-30390	43	12	layernorm	layernorm	NOUN
fcis-30390	43	13	⋅	⋅	PROPN
fcis-30390	43	14	13	13	NUM
fcis-30390	43	15	where	where	SCONJ
fcis-30390	43	16	:	:	PUNCT
fcis-30390	43	17	softmax	softmax	PROPN
fcis-30390	43	18	√2	√2	NOUN
fcis-30390	43	19	1	1	NUM
fcis-30390	43	20	14	14	NUM
fcis-30390	43	21	this	this	DET
fcis-30390	43	22	design	design	NOUN
fcis-30390	43	23	achieves	achieve	VERB
fcis-30390	43	24	an	an	DET
fcis-30390	43	25	efficient	efficient	ADJ
fcis-30390	43	26	balance	balance	NOUN
fcis-30390	43	27	between	between	ADP
fcis-30390	43	28	computational	computational	ADJ
fcis-30390	43	29	complexity	complexity	NOUN
fcis-30390	43	30	and	and	CCONJ
fcis-30390	43	31	model	model	NOUN
fcis-30390	43	32	expressiveness	expressiveness	NOUN
fcis-30390	43	33	.	.	PUNCT
fcis-30390	44	1	through	through	ADP
fcis-30390	44	2	adaptive	adaptive	ADJ
fcis-30390	44	3	variable	variable	ADJ
fcis-30390	44	4	attention	attention	NOUN
fcis-30390	44	5	mechanisms	mechanism	NOUN
fcis-30390	44	6	,	,	PUNCT
fcis-30390	44	7	the	the	DET
fcis-30390	44	8	model	model	NOUN
fcis-30390	44	9	effectively	effectively	ADV
fcis-30390	44	10	captures	capture	VERB
fcis-30390	44	11	dynamic	dynamic	ADJ
fcis-30390	44	12	correlations	correlation	NOUN
fcis-30390	44	13	in	in	ADP
fcis-30390	44	14	multivariate	multivariate	NOUN
fcis-30390	44	15	time	time	NOUN
fcis-30390	44	16	series	series	NOUN
fcis-30390	44	17	while	while	SCONJ
fcis-30390	44	18	differentiating	differentiate	VERB
fcis-30390	44	19	between	between	ADP
fcis-30390	44	20	same	same	ADJ
fcis-30390	44	21	-	-	PUNCT
fcis-30390	44	22	variable	variable	ADJ
fcis-30390	44	23	and	and	CCONJ
fcis-30390	44	24	cross	cross	ADJ
fcis-30390	44	25	-	-	ADJ
fcis-30390	44	26	variable	variable	ADJ
fcis-30390	44	27	relationships	relationship	NOUN
fcis-30390	44	28	through	through	ADP
fcis-30390	44	29	bias	bias	NOUN
fcis-30390	44	30	terms	term	NOUN
fcis-30390	44	31	and	and	CCONJ
fcis-30390	44	32	.	.	PUNCT
fcis-30390	45	1	the	the	DET
fcis-30390	45	2	temporal	temporal	ADJ
fcis-30390	45	3	structure	structure	NOUN
fcis-30390	45	4	is	be	AUX
fcis-30390	45	5	preserved	preserve	VERB
fcis-30390	45	6	through	through	ADP
fcis-30390	45	7	rotary	rotary	NOUN
fcis-30390	45	8	position	position	NOUN
fcis-30390	45	9	encoding	encoding	NOUN
fcis-30390	45	10	and	and	CCONJ
fcis-30390	45	11	residual	residual	ADJ
fcis-30390	45	12	connections	connection	NOUN
fcis-30390	45	13	,	,	PUNCT
fcis-30390	45	14	while	while	SCONJ
fcis-30390	45	15	learnable	learnable	ADJ
fcis-30390	45	16	projection	projection	NOUN
fcis-30390	45	17	matrices	matrix	NOUN
fcis-30390	45	18	and	and	CCONJ
fcis-30390	45	19	layer	layer	NOUN
fcis-30390	45	20	normalization	normalization	NOUN
fcis-30390	45	21	enable	enable	VERB
fcis-30390	45	22	adaptive	adaptive	ADJ
fcis-30390	45	23	feature	feature	NOUN
fcis-30390	45	24	representation	representation	NOUN
fcis-30390	45	25	adjustment	adjustment	NOUN
fcis-30390	45	26	,	,	PUNCT
fcis-30390	45	27	significantly	significantly	ADV
fcis-30390	45	28	enhancing	enhance	VERB
fcis-30390	45	29	the	the	DET
fcis-30390	45	30	model	model	NOUN
fcis-30390	45	31	's	's	PART
fcis-30390	45	32	capability	capability	NOUN
fcis-30390	45	33	in	in	ADP
fcis-30390	45	34	complex	complex	ADJ
fcis-30390	45	35	multivariate	multivariate	NOUN
fcis-30390	45	36	time	time	NOUN
fcis-30390	45	37	series	series	PROPN
fcis-30390	45	38	modeling	modeling	NOUN
fcis-30390	45	39	while	while	SCONJ
fcis-30390	45	40	maintaining	maintain	VERB
fcis-30390	45	41	computational	computational	ADJ
fcis-30390	45	42	efficiency	efficiency	NOUN
fcis-30390	45	43	.	.	PUNCT
fcis-30390	46	1	3.2	3.2	NUM
fcis-30390	46	2	.	.	PUNCT
fcis-30390	47	1	multi	multi	ADJ
fcis-30390	47	2	-	-	ADJ
fcis-30390	47	3	time	time	NOUN
fcis-30390	47	4	-	-	PUNCT
fcis-30390	47	5	scale	scale	NOUN
fcis-30390	47	6	memory	memory	NOUN
fcis-30390	47	7	enhancement	enhancement	NOUN
fcis-30390	47	8	module	module	NOUN
fcis-30390	47	9	input	input	NOUN
fcis-30390	47	10	w	w	ADP
fcis-30390	47	11	₁	₁	PROPN
fcis-30390	47	12	w	w	PROPN
fcis-30390	47	13	₂	₂	NUM
fcis-30390	47	14	w	w	NOUN
fcis-30390	47	15	₃	₃	PROPN
fcis-30390	47	16	w	w	PROPN
fcis-30390	47	17	₄	₄	NUM
fcis-30390	47	18	w	w	PROPN
fcis-30390	47	19	₅	₅	PROPN
fcis-30390	47	20	w	w	PROPN
fcis-30390	47	21	₆	₆	NUM
fcis-30390	47	22	w	w	NOUN
fcis-30390	47	23	₇	₇	NUM
fcis-30390	47	24	7	7	NUM
fcis-30390	47	25	1	1	NUM
fcis-30390	47	26	conv	conv	NOUN
fcis-30390	47	27	extracted	extract	VERB
fcis-30390	47	28	periodic	periodic	ADJ
fcis-30390	47	29	features	feature	NOUN
fcis-30390	47	30	g	g	PROPN
fcis-30390	47	31	e	e	PROPN
fcis-30390	47	32	l	l	NOUN
fcis-30390	47	33	u	u	PROPN
fcis-30390	47	34	enhanced	enhance	VERB
fcis-30390	47	35	periodicity	periodicity	NOUN
fcis-30390	47	36	features	feature	VERB
fcis-30390	47	37	3×1	3×1	NUM
fcis-30390	47	38	local	local	ADJ
fcis-30390	47	39	convolution	convolution	NOUN
fcis-30390	47	40	w	w	ADP
fcis-30390	47	41	₁	₁	PROPN
fcis-30390	47	42	w	w	PROPN
fcis-30390	47	43	₂	₂	NUM
fcis-30390	47	44	w	w	NOUN
fcis-30390	47	45	₃	₃	NUM
fcis-30390	47	46	s	s	NOUN
fcis-30390	47	47	to	to	ADP
fcis-30390	47	48	ra	ra	PROPN
fcis-30390	47	49	ge	ge	PROPN
fcis-30390	47	50	u	u	PROPN
fcis-30390	47	51	pd	pd	PROPN
fcis-30390	47	52	at	at	ADP
fcis-30390	47	53	e	e	PROPN
fcis-30390	47	54	r	r	VERB
fcis-30390	47	55	et	et	PROPN
fcis-30390	47	56	ri	ri	PROPN
fcis-30390	47	57	ev	ev	PROPN
fcis-30390	47	58	al	al	PROPN
fcis-30390	47	59	short_term	short_term	PROPN
fcis-30390	47	60	memory	memory	NOUN
fcis-30390	47	61	unit	unit	NOUN
fcis-30390	47	62	q	q	PROPN
fcis-30390	48	1	k	k	PROPN
fcis-30390	48	2	v	v	ADP
fcis-30390	48	3	four_head	four_head	ADJ
fcis-30390	48	4	attention	attention	NOUN
fcis-30390	48	5	enhanced	enhance	VERB
fcis-30390	48	6	short	short	ADJ
fcis-30390	48	7	-	-	PUNCT
fcis-30390	48	8	term	term	NOUN
fcis-30390	48	9	dependency	dependency	NOUN
fcis-30390	48	10	features	feature	VERB
fcis-30390	48	11	gap	gap	NOUN
fcis-30390	48	12	s	s	VERB
fcis-30390	48	13	to	to	ADP
fcis-30390	48	14	ra	ra	PROPN
fcis-30390	48	15	ge	ge	PROPN
fcis-30390	48	16	u	u	PROPN
fcis-30390	48	17	pd	pd	PROPN
fcis-30390	48	18	at	at	ADP
fcis-30390	48	19	e	e	PROPN
fcis-30390	48	20	r	r	VERB
fcis-30390	48	21	et	et	PROPN
fcis-30390	48	22	ri	ri	NOUN
fcis-30390	48	23	ev	ev	ADP
fcis-30390	48	24	a	a	DET
fcis-30390	48	25	long_term	long_term	ADJ
fcis-30390	48	26	memory	memory	NOUN
fcis-30390	48	27	unit	unit	NOUN
fcis-30390	48	28	enhanced	enhance	VERB
fcis-30390	48	29	long	long	ADJ
fcis-30390	48	30	-	-	PUNCT
fcis-30390	48	31	term	term	NOUN
fcis-30390	48	32	dependency	dependency	NOUN
fcis-30390	48	33	features	feature	VERB
fcis-30390	48	34	periodicity	periodicity	NOUN
fcis-30390	48	35	detector	detector	NOUN
fcis-30390	48	36	short_term	short_term	NOUN
fcis-30390	48	37	m	m	VERB
fcis-30390	48	38	em	em	PROPN
fcis-30390	48	39	ory	ory	NOUN
fcis-30390	48	40	unit	unit	NOUN
fcis-30390	48	41	l	l	PROPN
fcis-30390	48	42	ong_term	ong_term	NOUN
fcis-30390	48	43	m	m	VERB
fcis-30390	48	44	em	em	PROPN
fcis-30390	48	45	ory	ory	NOUN
fcis-30390	48	46	unit	unit	NOUN
fcis-30390	48	47	feature	feature	NOUN
fcis-30390	48	48	integration	integration	NOUN
fcis-30390	48	49	layernorm	layernorm	NOUN
fcis-30390	48	50	dropout	dropout	PROPN
fcis-30390	48	51	result	result	NOUN
fcis-30390	48	52	figure	figure	NOUN
fcis-30390	48	53	3	3	NUM
fcis-30390	48	54	.	.	PUNCT
fcis-30390	49	1	the	the	DET
fcis-30390	49	2	architecture	architecture	NOUN
fcis-30390	49	3	of	of	ADP
fcis-30390	49	4	mtmem	mtmem	NOUN
fcis-30390	49	5	this	this	DET
fcis-30390	49	6	module	module	NOUN
fcis-30390	49	7	consists	consist	VERB
fcis-30390	49	8	of	of	ADP
fcis-30390	49	9	three	three	NUM
fcis-30390	49	10	parallel	parallel	ADJ
fcis-30390	49	11	processing	processing	NOUN
fcis-30390	49	12	tracks	track	NOUN
fcis-30390	49	13	for	for	ADP
fcis-30390	49	14	comprehensive	comprehensive	ADJ
fcis-30390	49	15	feature	feature	NOUN
fcis-30390	49	16	extraction	extraction	NOUN
fcis-30390	49	17	and	and	CCONJ
fcis-30390	49	18	enhancement	enhancement	NOUN
fcis-30390	49	19	.	.	PUNCT
fcis-30390	50	1	the	the	DET
fcis-30390	50	2	upper	upper	ADJ
fcis-30390	50	3	track	track	NOUN
fcis-30390	50	4	employs	employ	VERB
fcis-30390	50	5	gap	gap	NOUN
fcis-30390	50	6	(	(	PUNCT
fcis-30390	50	7	global	global	ADJ
fcis-30390	50	8	average	average	ADJ
fcis-30390	50	9	pooling	pooling	NOUN
fcis-30390	50	10	)	)	PUNCT
fcis-30390	50	11	followed	follow	VERB
fcis-30390	50	12	by	by	ADP
fcis-30390	50	13	a	a	DET
fcis-30390	50	14	long	long	ADJ
fcis-30390	50	15	-	-	PUNCT
fcis-30390	50	16	term	term	NOUN
fcis-30390	50	17	memory	memory	NOUN
fcis-30390	50	18	unit	unit	NOUN
fcis-30390	50	19	to	to	PART
fcis-30390	50	20	capture	capture	VERB
fcis-30390	50	21	global	global	ADJ
fcis-30390	50	22	dependencies	dependency	NOUN
fcis-30390	50	23	.	.	PUNCT
fcis-30390	51	1	the	the	DET
fcis-30390	51	2	middle	middle	ADJ
fcis-30390	51	3	track	track	NOUN
fcis-30390	51	4	utilizes	utilize	VERB
fcis-30390	51	5	a	a	DET
fcis-30390	51	6	3×1	3×1	NUM
fcis-30390	51	7	local	local	ADJ
fcis-30390	51	8	convolution	convolution	NOUN
fcis-30390	51	9	and	and	CCONJ
fcis-30390	51	10	shortterm	shortterm	PROPN
fcis-30390	51	11	memory	memory	NOUN
fcis-30390	51	12	unit	unit	NOUN
fcis-30390	51	13	,	,	PUNCT
fcis-30390	51	14	coupled	couple	VERB
fcis-30390	51	15	with	with	ADP
fcis-30390	51	16	four	four	NUM
fcis-30390	51	17	-	-	PUNCT
fcis-30390	51	18	head	head	NOUN
fcis-30390	51	19	attention	attention	NOUN
fcis-30390	51	20	mechanism	mechanism	NOUN
fcis-30390	51	21	for	for	ADP
fcis-30390	51	22	local	local	ADJ
fcis-30390	51	23	feature	feature	NOUN
fcis-30390	51	24	processing	processing	NOUN
fcis-30390	51	25	.	.	PUNCT
fcis-30390	52	1	the	the	DET
fcis-30390	52	2	bottom	bottom	ADJ
fcis-30390	52	3	track	track	NOUN
fcis-30390	52	4	implements	implement	VERB
fcis-30390	52	5	a	a	DET
fcis-30390	52	6	7×1	7×1	NUM
fcis-30390	52	7	convolutional	convolutional	ADJ
fcis-30390	52	8	layer	layer	NOUN
fcis-30390	52	9	and	and	CCONJ
fcis-30390	52	10	gelu	gelu	ADJ
fcis-30390	52	11	activation	activation	NOUN
fcis-30390	52	12	to	to	PART
fcis-30390	52	13	extract	extract	VERB
fcis-30390	52	14	periodic	periodic	ADJ
fcis-30390	52	15	features	feature	NOUN
fcis-30390	52	16	.	.	PUNCT
fcis-30390	53	1	each	each	DET
fcis-30390	53	2	track	track	NOUN
fcis-30390	53	3	enhances	enhance	VERB
fcis-30390	53	4	specific	specific	ADJ
fcis-30390	53	5	feature	feature	NOUN
fcis-30390	53	6	types	type	NOUN
fcis-30390	53	7	(	(	PUNCT
fcis-30390	53	8	long	long	ADJ
fcis-30390	53	9	-	-	PUNCT
fcis-30390	53	10	term	term	NOUN
fcis-30390	53	11	dependency	dependency	NOUN
fcis-30390	53	12	,	,	PUNCT
fcis-30390	53	13	short	short	ADJ
fcis-30390	53	14	-	-	PUNCT
fcis-30390	53	15	term	term	NOUN
fcis-30390	53	16	dependency	dependency	NOUN
fcis-30390	53	17	,	,	PUNCT
fcis-30390	53	18	and	and	CCONJ
fcis-30390	53	19	periodicity	periodicity	NOUN
fcis-30390	53	20	)	)	PUNCT
fcis-30390	53	21	before	before	ADP
fcis-30390	53	22	being	be	AUX
fcis-30390	53	23	integrated	integrate	VERB
fcis-30390	53	24	through	through	ADP
fcis-30390	53	25	a	a	DET
fcis-30390	53	26	feature	feature	NOUN
fcis-30390	53	27	integration	integration	NOUN
fcis-30390	53	28	module	module	NOUN
fcis-30390	53	29	.	.	PUNCT
fcis-30390	54	1	the	the	DET
fcis-30390	54	2	final	final	ADJ
fcis-30390	54	3	output	output	NOUN
fcis-30390	54	4	undergoes	undergoe	NOUN
fcis-30390	54	5	layernorm	layernorm	NOUN
fcis-30390	54	6	and	and	CCONJ
fcis-30390	54	7	dropout	dropout	NOUN
fcis-30390	54	8	operations	operation	NOUN
fcis-30390	54	9	to	to	PART
fcis-30390	54	10	produce	produce	VERB
fcis-30390	54	11	the	the	DET
fcis-30390	54	12	result	result	NOUN
fcis-30390	54	13	.	.	PUNCT
fcis-30390	55	1	time	time	NOUN
fcis-30390	55	2	series	series	PROPN
fcis-30390	55	3	contain	contain	VERB
fcis-30390	55	4	rich	rich	ADJ
fcis-30390	55	5	multi	multi	ADJ
fcis-30390	55	6	-	-	ADJ
fcis-30390	55	7	scale	scale	ADJ
fcis-30390	55	8	temporal	temporal	ADJ
fcis-30390	55	9	features	feature	NOUN
fcis-30390	55	10	that	that	PRON
fcis-30390	55	11	exhibit	exhibit	VERB
fcis-30390	55	12	unique	unique	ADJ
fcis-30390	55	13	statistical	statistical	ADJ
fcis-30390	55	14	properties	property	NOUN
fcis-30390	55	15	at	at	ADP
fcis-30390	55	16	different	different	ADJ
fcis-30390	55	17	time	time	NOUN
fcis-30390	55	18	scales	scale	NOUN
fcis-30390	55	19	.	.	PUNCT
fcis-30390	56	1	given	give	VERB
fcis-30390	56	2	the	the	DET
fcis-30390	56	3	input	input	NOUN
fcis-30390	56	4	sequence	sequence	NOUN
fcis-30390	56	5	:	:	PUNCT
fcis-30390	56	6	∈	∈	PROPN
fcis-30390	56	7	15	15	NUM
fcis-30390	57	1	the	the	DET
fcis-30390	57	2	multi	multi	ADJ
fcis-30390	57	3	-	-	ADJ
fcis-30390	57	4	time	time	ADJ
fcis-30390	57	5	-	-	PUNCT
fcis-30390	57	6	scale	scale	NOUN
fcis-30390	57	7	memory	memory	NOUN
fcis-30390	57	8	enhancement	enhancement	NOUN
fcis-30390	57	9	module	module	NOUN
fcis-30390	57	10	(	(	PUNCT
fcis-30390	57	11	mtmem	mtmem	NOUN
fcis-30390	57	12	)	)	PUNCT
fcis-30390	57	13	enhances	enhance	VERB
fcis-30390	57	14	time	time	NOUN
fcis-30390	57	15	series	series	PROPN
fcis-30390	57	16	representation	representation	NOUN
fcis-30390	57	17	through	through	ADP
fcis-30390	57	18	a	a	DET
fcis-30390	57	19	multi	multi	ADJ
fcis-30390	57	20	-	-	ADJ
fcis-30390	57	21	level	level	ADJ
fcis-30390	57	22	feature	feature	NOUN
fcis-30390	57	23	extraction	extraction	NOUN
fcis-30390	57	24	and	and	CCONJ
fcis-30390	57	25	fusion	fusion	NOUN
fcis-30390	57	26	mechanism	mechanism	NOUN
fcis-30390	57	27	,	,	PUNCT
fcis-30390	57	28	expressed	express	VERB
fcis-30390	57	29	as	as	ADP
fcis-30390	57	30	:	:	PUNCT
fcis-30390	57	31	,	,	PUNCT
fcis-30390	57	32	ℓ	ℓ	INTJ
fcis-30390	57	33	,	,	PUNCT
fcis-30390	57	34	16	16	NUM
fcis-30390	57	35	38	38	NUM
fcis-30390	57	36	the	the	DET
fcis-30390	57	37	core	core	NOUN
fcis-30390	57	38	architecture	architecture	NOUN
fcis-30390	57	39	of	of	ADP
fcis-30390	57	40	mtmem	mtmem	NOUN
fcis-30390	57	41	consists	consist	NOUN
fcis-30390	57	42	of	of	ADP
fcis-30390	57	43	three	three	NUM
fcis-30390	57	44	complementary	complementary	ADJ
fcis-30390	57	45	feature	feature	NOUN
fcis-30390	57	46	extraction	extraction	NOUN
fcis-30390	57	47	modules	module	NOUN
fcis-30390	57	48	,	,	PUNCT
fcis-30390	57	49	each	each	PRON
fcis-30390	57	50	focusing	focus	VERB
fcis-30390	57	51	on	on	ADP
fcis-30390	57	52	capturing	capture	VERB
fcis-30390	57	53	temporal	temporal	ADJ
fcis-30390	57	54	patterns	pattern	NOUN
fcis-30390	57	55	at	at	ADP
fcis-30390	57	56	specific	specific	ADJ
fcis-30390	57	57	scales	scale	NOUN
fcis-30390	57	58	.	.	PUNCT
fcis-30390	58	1	first	first	ADJ
fcis-30390	58	2	,	,	PUNCT
fcis-30390	58	3	short	short	ADJ
fcis-30390	58	4	-	-	PUNCT
fcis-30390	58	5	term	term	NOUN
fcis-30390	58	6	dependency	dependency	NOUN
fcis-30390	58	7	features	feature	NOUN
fcis-30390	58	8	are	be	AUX
fcis-30390	58	9	captured	capture	VERB
fcis-30390	58	10	through	through	ADP
fcis-30390	58	11	a	a	DET
fcis-30390	58	12	combination	combination	NOUN
fcis-30390	58	13	of	of	ADP
fcis-30390	58	14	local	local	ADJ
fcis-30390	58	15	convolution	convolution	NOUN
fcis-30390	58	16	and	and	CCONJ
fcis-30390	58	17	attention	attention	NOUN
fcis-30390	58	18	mechanisms	mechanism	NOUN
fcis-30390	58	19	.	.	PUNCT
fcis-30390	59	1	specifically	specifically	ADV
fcis-30390	59	2	,	,	PUNCT
fcis-30390	59	3	depthwise	depthwise	VERB
fcis-30390	59	4	separable	separable	ADJ
fcis-30390	59	5	convolution	convolution	NOUN
fcis-30390	59	6	extracts	extract	NOUN
fcis-30390	59	7	features	feature	NOUN
fcis-30390	59	8	within	within	ADP
fcis-30390	59	9	a	a	DET
fcis-30390	59	10	local	local	ADJ
fcis-30390	59	11	time	time	NOUN
fcis-30390	59	12	window	window	NOUN
fcis-30390	59	13	as	as	ADP
fcis-30390	59	14	:	:	PUNCT
fcis-30390	59	15	∗	∗	NOUN
fcis-30390	59	16	,	,	PUNCT
fcis-30390	59	17	∈	∈	PROPN
fcis-30390	59	18	17	17	NUM
fcis-30390	59	19	this	this	DET
fcis-30390	59	20	convolution	convolution	NOUN
fcis-30390	59	21	structure	structure	NOUN
fcis-30390	59	22	significantly	significantly	ADV
fcis-30390	59	23	reduces	reduce	VERB
fcis-30390	59	24	the	the	DET
fcis-30390	59	25	number	number	NOUN
fcis-30390	59	26	of	of	ADP
fcis-30390	59	27	parameters	parameter	NOUN
fcis-30390	59	28	while	while	SCONJ
fcis-30390	59	29	maintaining	maintain	VERB
fcis-30390	59	30	feature	feature	NOUN
fcis-30390	59	31	extraction	extraction	NOUN
fcis-30390	59	32	capability	capability	NOUN
fcis-30390	59	33	.	.	PUNCT
fcis-30390	60	1	subsequently	subsequently	ADV
fcis-30390	60	2	,	,	PUNCT
fcis-30390	60	3	a	a	DET
fcis-30390	60	4	multi	multi	ADJ
fcis-30390	60	5	-	-	ADJ
fcis-30390	60	6	head	head	ADJ
fcis-30390	60	7	self	self	NOUN
fcis-30390	60	8	-	-	PUNCT
fcis-30390	60	9	attention	attention	NOUN
fcis-30390	60	10	mechanism	mechanism	NOUN
fcis-30390	60	11	further	far	ADV
fcis-30390	60	12	enhances	enhance	VERB
fcis-30390	60	13	the	the	DET
fcis-30390	60	14	expression	expression	NOUN
fcis-30390	60	15	of	of	ADP
fcis-30390	60	16	local	local	ADJ
fcis-30390	60	17	features	feature	NOUN
fcis-30390	60	18	:	:	PUNCT
fcis-30390	60	19	softmax	softmax	NOUN
fcis-30390	60	20	18	18	NUM
fcis-30390	60	21	where	where	SCONJ
fcis-30390	60	22	the	the	DET
fcis-30390	60	23	attention	attention	NOUN
fcis-30390	60	24	weights	weight	NOUN
fcis-30390	60	25	consider	consider	VERB
fcis-30390	60	26	the	the	DET
fcis-30390	60	27	local	local	ADJ
fcis-30390	60	28	correlations	correlation	NOUN
fcis-30390	60	29	of	of	ADP
fcis-30390	60	30	features	feature	NOUN
fcis-30390	60	31	,	,	PUNCT
fcis-30390	60	32	and	and	CCONJ
fcis-30390	60	33	h	h	NOUN
fcis-30390	60	34	attention	attention	NOUN
fcis-30390	60	35	heads	head	NOUN
fcis-30390	60	36	collaboratively	collaboratively	ADV
fcis-30390	60	37	construct	construct	VERB
fcis-30390	60	38	multi	multi	ADJ
fcis-30390	60	39	-	-	ADJ
fcis-30390	60	40	faceted	faceted	ADJ
fcis-30390	60	41	feature	feature	NOUN
fcis-30390	60	42	representations	representation	NOUN
fcis-30390	60	43	.	.	PUNCT
fcis-30390	61	1	for	for	ADP
fcis-30390	61	2	long	long	ADJ
fcis-30390	61	3	-	-	PUNCT
fcis-30390	61	4	term	term	NOUN
fcis-30390	61	5	dependency	dependency	NOUN
fcis-30390	61	6	feature	feature	NOUN
fcis-30390	61	7	modeling	modeling	NOUN
fcis-30390	61	8	,	,	PUNCT
fcis-30390	61	9	global	global	ADJ
fcis-30390	61	10	contextual	contextual	ADJ
fcis-30390	61	11	features	feature	NOUN
fcis-30390	61	12	are	be	AUX
fcis-30390	61	13	first	first	ADV
fcis-30390	61	14	obtained	obtain	VERB
fcis-30390	61	15	through	through	ADP
fcis-30390	61	16	pooling	pool	VERB
fcis-30390	61	17	operations	operation	NOUN
fcis-30390	61	18	along	along	ADP
fcis-30390	61	19	the	the	DET
fcis-30390	61	20	time	time	NOUN
fcis-30390	61	21	dimension	dimension	NOUN
fcis-30390	61	22	:	:	PUNCT
fcis-30390	61	23	pool	pool	NOUN
fcis-30390	61	24	∈	∈	PROPN
fcis-30390	61	25	19	19	NUM
fcis-30390	61	26	the	the	DET
fcis-30390	61	27	memory	memory	NOUN
fcis-30390	61	28	retrieval	retrieval	NOUN
fcis-30390	61	29	mechanism	mechanism	NOUN
fcis-30390	61	30	is	be	AUX
fcis-30390	61	31	implemented	implement	VERB
fcis-30390	61	32	through	through	ADP
fcis-30390	61	33	a	a	DET
fcis-30390	61	34	trainable	trainable	ADJ
fcis-30390	61	35	memory	memory	NOUN
fcis-30390	61	36	bank	bank	NOUN
fcis-30390	61	37	:	:	PUNCT
fcis-30390	61	38	∈	∈	PROPN
fcis-30390	61	39	20	20	NUM
fcis-30390	61	40	where	where	SCONJ
fcis-30390	61	41	21	21	NUM
fcis-30390	61	42	22	22	NUM
fcis-30390	61	43	the	the	DET
fcis-30390	61	44	long	long	ADJ
fcis-30390	61	45	-	-	PUNCT
fcis-30390	61	46	term	term	NOUN
fcis-30390	61	47	features	feature	NOUN
fcis-30390	61	48	are	be	AUX
fcis-30390	61	49	then	then	ADV
fcis-30390	61	50	computed	compute	VERB
fcis-30390	61	51	as	as	ADP
fcis-30390	61	52	:	:	PUNCT
fcis-30390	61	53	ℓ	ℓ	PROPN
fcis-30390	61	54	softmax	softmax	NOUN
fcis-30390	62	1	√	√	ADV
fcis-30390	62	2	23	23	NUM
fcis-30390	62	3	enabling	enable	VERB
fcis-30390	62	4	the	the	DET
fcis-30390	62	5	model	model	NOUN
fcis-30390	62	6	to	to	PART
fcis-30390	62	7	store	store	VERB
fcis-30390	62	8	and	and	CCONJ
fcis-30390	62	9	retrieve	retrieve	VERB
fcis-30390	62	10	long	long	ADJ
fcis-30390	62	11	-	-	PUNCT
fcis-30390	62	12	term	term	NOUN
fcis-30390	62	13	temporal	temporal	ADJ
fcis-30390	62	14	patterns	pattern	NOUN
fcis-30390	62	15	,	,	PUNCT
fcis-30390	62	16	thereby	thereby	ADV
fcis-30390	62	17	enhancing	enhance	VERB
fcis-30390	62	18	its	its	PRON
fcis-30390	62	19	modeling	modeling	NOUN
fcis-30390	62	20	capability	capability	NOUN
fcis-30390	62	21	for	for	ADP
fcis-30390	62	22	longterm	longterm	ADJ
fcis-30390	62	23	dependencies	dependency	NOUN
fcis-30390	62	24	.	.	PUNCT
fcis-30390	63	1	periodic	periodic	ADJ
fcis-30390	63	2	features	feature	NOUN
fcis-30390	63	3	are	be	AUX
fcis-30390	63	4	extracted	extract	VERB
fcis-30390	63	5	through	through	ADP
fcis-30390	63	6	a	a	DET
fcis-30390	63	7	dedicated	dedicated	ADJ
fcis-30390	63	8	convolutional	convolutional	ADJ
fcis-30390	63	9	network	network	NOUN
fcis-30390	63	10	:	:	PUNCT
fcis-30390	63	11	pool	pool	NOUN
fcis-30390	63	12	gelu	gelu	ADJ
fcis-30390	63	13	conv	conv	NOUN
fcis-30390	63	14	;	;	PUNCT
fcis-30390	63	15	24	24	NUM
fcis-30390	63	16	the	the	DET
fcis-30390	63	17	choice	choice	NOUN
fcis-30390	63	18	of	of	ADP
fcis-30390	63	19	convolution	convolution	NOUN
fcis-30390	63	20	kernel	kernel	NOUN
fcis-30390	63	21	size	size	NOUN
fcis-30390	63	22	is	be	AUX
fcis-30390	63	23	based	base	VERB
fcis-30390	63	24	on	on	ADP
fcis-30390	63	25	the	the	DET
fcis-30390	63	26	theoretical	theoretical	ADJ
fcis-30390	63	27	foundations	foundation	NOUN
fcis-30390	63	28	of	of	ADP
fcis-30390	63	29	time	time	NOUN
fcis-30390	63	30	series	series	NOUN
fcis-30390	63	31	analysis	analysis	NOUN
fcis-30390	63	32	and	and	CCONJ
fcis-30390	63	33	practical	practical	ADJ
fcis-30390	63	34	application	application	NOUN
fcis-30390	63	35	needs	need	NOUN
fcis-30390	63	36	.	.	PUNCT
fcis-30390	64	1	according	accord	VERB
fcis-30390	64	2	to	to	ADP
fcis-30390	64	3	the	the	DET
fcis-30390	64	4	nyquist	nyquist	NOUN
fcis-30390	64	5	sampling	sample	VERB
fcis-30390	64	6	theorem	theorem	NOUN
fcis-30390	64	7	,	,	PUNCT
fcis-30390	64	8	to	to	PART
fcis-30390	64	9	capture	capture	VERB
fcis-30390	64	10	a	a	DET
fcis-30390	64	11	signal	signal	NOUN
fcis-30390	64	12	with	with	ADP
fcis-30390	64	13	frequency	frequency	NOUN
fcis-30390	64	14	,	,	PUNCT
fcis-30390	64	15	the	the	DET
fcis-30390	64	16	sampling	sample	VERB
fcis-30390	64	17	frequency	frequency	NOUN
fcis-30390	64	18	must	must	AUX
fcis-30390	64	19	reach	reach	VERB
fcis-30390	64	20	2	2	NUM
fcis-30390	64	21	which	which	PRON
fcis-30390	64	22	translates	translate	VERB
fcis-30390	64	23	to	to	ADP
fcis-30390	64	24	the	the	DET
fcis-30390	64	25	requirement	requirement	NOUN
fcis-30390	64	26	that	that	SCONJ
fcis-30390	64	27	the	the	DET
fcis-30390	64	28	convolution	convolution	NOUN
fcis-30390	64	29	kernel	kernel	PROPN
fcis-30390	64	30	size	size	NOUN
fcis-30390	64	31	must	must	AUX
fcis-30390	64	32	satisfy	satisfy	VERB
fcis-30390	64	33	2	2	NUM
fcis-30390	64	34	where	where	SCONJ
fcis-30390	64	35	is	be	AUX
fcis-30390	64	36	the	the	DET
fcis-30390	64	37	minimum	minimum	ADJ
fcis-30390	64	38	period	period	NOUN
fcis-30390	64	39	length	length	NOUN
fcis-30390	64	40	.	.	PUNCT
fcis-30390	65	1	additionally	additionally	ADV
fcis-30390	65	2	,	,	PUNCT
fcis-30390	65	3	considering	consider	VERB
fcis-30390	65	4	the	the	DET
fcis-30390	65	5	effective	effective	ADJ
fcis-30390	65	6	receptive	receptive	ADJ
fcis-30390	65	7	field	field	NOUN
fcis-30390	65	8	:	:	PUNCT
fcis-30390	65	9	1	1	NUM
fcis-30390	65	10	1	1	NUM
fcis-30390	65	11	25	25	NUM
fcis-30390	65	12	and	and	CCONJ
fcis-30390	65	13	the	the	DET
fcis-30390	65	14	periodic	periodic	ADJ
fcis-30390	65	15	resolution	resolution	NOUN
fcis-30390	65	16	capability	capability	NOUN
fcis-30390	65	17	2	2	NUM
fcis-30390	65	18	/	/	SYM
fcis-30390	65	19	26	26	NUM
fcis-30390	65	20	where	where	SCONJ
fcis-30390	65	21	is	be	AUX
fcis-30390	65	22	the	the	DET
fcis-30390	65	23	input	input	NOUN
fcis-30390	65	24	sequence	sequence	NOUN
fcis-30390	65	25	length	length	NOUN
fcis-30390	65	26	.	.	PUNCT
fcis-30390	66	1	the	the	DET
fcis-30390	66	2	choice	choice	NOUN
fcis-30390	66	3	of	of	ADP
fcis-30390	66	4	2	2	NUM
fcis-30390	66	5	=	=	SYM
fcis-30390	66	6	7	7	NUM
fcis-30390	66	7	under	under	ADP
fcis-30390	66	8	the	the	DET
fcis-30390	66	9	default	default	NOUN
fcis-30390	66	10	sequence	sequence	NOUN
fcis-30390	66	11	length	length	NOUN
fcis-30390	66	12	=	=	SYM
fcis-30390	66	13	96	96	NUM
fcis-30390	66	14	ensures	ensure	VERB
fcis-30390	66	15	the	the	DET
fcis-30390	66	16	symmetry	symmetry	NOUN
fcis-30390	66	17	of	of	ADP
fcis-30390	66	18	the	the	DET
fcis-30390	66	19	convolution	convolution	NOUN
fcis-30390	66	20	operation	operation	NOUN
fcis-30390	66	21	while	while	SCONJ
fcis-30390	66	22	achieving	achieve	VERB
fcis-30390	66	23	0.146	0.146	NUM
fcis-30390	66	24	this	this	PRON
fcis-30390	66	25	allows	allow	VERB
fcis-30390	66	26	for	for	ADP
fcis-30390	66	27	the	the	DET
fcis-30390	66	28	capture	capture	NOUN
fcis-30390	66	29	of	of	ADP
fcis-30390	66	30	the	the	DET
fcis-30390	66	31	minimum	minimum	ADJ
fcis-30390	66	32	periodic	periodic	ADJ
fcis-30390	66	33	pattern	pattern	NOUN
fcis-30390	66	34	of	of	ADP
fcis-30390	66	35	approximately	approximately	ADV
fcis-30390	66	36	7	7	NUM
fcis-30390	66	37	time	time	NOUN
fcis-30390	66	38	steps	step	NOUN
fcis-30390	66	39	,	,	PUNCT
fcis-30390	66	40	while	while	SCONJ
fcis-30390	66	41	reaching	reach	VERB
fcis-30390	66	42	an	an	DET
fcis-30390	66	43	optimal	optimal	ADJ
fcis-30390	66	44	balance	balance	NOUN
fcis-30390	66	45	between	between	ADP
fcis-30390	66	46	computational	computational	ADJ
fcis-30390	66	47	complexity	complexity	NOUN
fcis-30390	66	48	and	and	CCONJ
fcis-30390	66	49	feature	feature	NOUN
fcis-30390	66	50	extraction	extraction	NOUN
fcis-30390	66	51	capability	capability	NOUN
fcis-30390	66	52	.	.	PUNCT
fcis-30390	67	1	this	this	DET
fcis-30390	67	2	choice	choice	NOUN
fcis-30390	67	3	is	be	AUX
fcis-30390	67	4	well	well	ADV
fcis-30390	67	5	-	-	PUNCT
fcis-30390	67	6	supported	support	VERB
fcis-30390	67	7	by	by	ADP
fcis-30390	67	8	theoretical	theoretical	ADJ
fcis-30390	67	9	analysis	analysis	NOUN
fcis-30390	67	10	and	and	CCONJ
fcis-30390	67	11	experimental	experimental	ADJ
fcis-30390	67	12	validation	validation	NOUN
fcis-30390	67	13	.	.	PUNCT
fcis-30390	68	1	multi	multi	ADJ
fcis-30390	68	2	-	-	ADJ
fcis-30390	68	3	scale	scale	ADJ
fcis-30390	68	4	features	feature	NOUN
fcis-30390	68	5	are	be	AUX
fcis-30390	68	6	fused	fuse	VERB
fcis-30390	68	7	through	through	ADP
fcis-30390	68	8	a	a	DET
fcis-30390	68	9	dynamic	dynamic	ADJ
fcis-30390	68	10	weight	weight	NOUN
fcis-30390	68	11	mechanism	mechanism	NOUN
fcis-30390	68	12	,	,	PUNCT
fcis-30390	68	13	where	where	SCONJ
fcis-30390	68	14	27	27	NUM
fcis-30390	68	15	and	and	CCONJ
fcis-30390	68	16	28	28	NUM
fcis-30390	68	17	this	this	DET
fcis-30390	68	18	fusion	fusion	NOUN
fcis-30390	68	19	approach	approach	NOUN
fcis-30390	68	20	ensures	ensure	VERB
fcis-30390	68	21	the	the	DET
fcis-30390	68	22	full	full	ADJ
fcis-30390	68	23	utilization	utilization	NOUN
fcis-30390	68	24	of	of	ADP
fcis-30390	68	25	information	information	NOUN
fcis-30390	68	26	and	and	CCONJ
fcis-30390	68	27	stable	stable	ADJ
fcis-30390	68	28	gradient	gradient	NOUN
fcis-30390	68	29	propagation	propagation	NOUN
fcis-30390	68	30	.	.	PUNCT
fcis-30390	69	1	the	the	DET
fcis-30390	69	2	model	model	NOUN
fcis-30390	69	3	's	's	PART
fcis-30390	69	4	computational	computational	ADJ
fcis-30390	69	5	complexity	complexity	NOUN
fcis-30390	69	6	is	be	AUX
fcis-30390	69	7	,	,	PUNCT
fcis-30390	69	8	and	and	CCONJ
fcis-30390	69	9	the	the	DET
fcis-30390	69	10	spatial	spatial	ADJ
fcis-30390	69	11	complexity	complexity	NOUN
fcis-30390	69	12	is	be	AUX
fcis-30390	69	13	.	.	PUNCT
fcis-30390	70	1	through	through	ADP
fcis-30390	70	2	hierarchical	hierarchical	ADJ
fcis-30390	70	3	feature	feature	NOUN
fcis-30390	70	4	extraction	extraction	NOUN
fcis-30390	70	5	and	and	CCONJ
fcis-30390	70	6	adaptive	adaptive	ADJ
fcis-30390	70	7	fusion	fusion	NOUN
fcis-30390	70	8	,	,	PUNCT
fcis-30390	70	9	the	the	DET
fcis-30390	70	10	model	model	NOUN
fcis-30390	70	11	accurately	accurately	ADV
fcis-30390	70	12	captures	capture	VERB
fcis-30390	70	13	the	the	DET
fcis-30390	70	14	dynamic	dynamic	ADJ
fcis-30390	70	15	features	feature	NOUN
fcis-30390	70	16	of	of	ADP
fcis-30390	70	17	time	time	NOUN
fcis-30390	70	18	series	series	NOUN
fcis-30390	70	19	across	across	ADP
fcis-30390	70	20	different	different	ADJ
fcis-30390	70	21	time	time	NOUN
fcis-30390	70	22	scales	scale	NOUN
fcis-30390	70	23	,	,	PUNCT
fcis-30390	70	24	providing	provide	VERB
fcis-30390	70	25	a	a	DET
fcis-30390	70	26	new	new	ADJ
fcis-30390	70	27	paradigm	paradigm	NOUN
fcis-30390	70	28	for	for	ADP
fcis-30390	70	29	time	time	NOUN
fcis-30390	70	30	series	series	PROPN
fcis-30390	70	31	analysis	analysis	NOUN
fcis-30390	70	32	.	.	PUNCT
fcis-30390	71	1	experimental	experimental	ADJ
fcis-30390	71	2	results	result	NOUN
fcis-30390	71	3	demonstrate	demonstrate	VERB
fcis-30390	71	4	that	that	SCONJ
fcis-30390	71	5	this	this	DET
fcis-30390	71	6	multi	multi	ADJ
fcis-30390	71	7	-	-	ADJ
fcis-30390	71	8	scale	scale	ADJ
fcis-30390	71	9	feature	feature	NOUN
fcis-30390	71	10	extraction	extraction	NOUN
fcis-30390	71	11	and	and	CCONJ
fcis-30390	71	12	fusion	fusion	NOUN
fcis-30390	71	13	mechanism	mechanism	NOUN
fcis-30390	71	14	achieves	achieve	VERB
fcis-30390	71	15	significant	significant	ADJ
fcis-30390	71	16	performance	performance	NOUN
fcis-30390	71	17	improvements	improvement	NOUN
fcis-30390	71	18	across	across	ADP
fcis-30390	71	19	various	various	ADJ
fcis-30390	71	20	time	time	NOUN
fcis-30390	71	21	series	series	PROPN
fcis-30390	71	22	forecasting	forecasting	NOUN
fcis-30390	71	23	tasks	task	NOUN
fcis-30390	71	24	.	.	PUNCT
fcis-30390	72	1	4	4	X
fcis-30390	72	2	.	.	X
fcis-30390	72	3	experiment	experiment	NOUN
fcis-30390	72	4	4.1	4.1	NUM
fcis-30390	72	5	.	.	PUNCT
fcis-30390	73	1	datasets	dataset	VERB
fcis-30390	73	2	the	the	DET
fcis-30390	73	3	southern	southern	ADJ
fcis-30390	73	4	region	region	NOUN
fcis-30390	73	5	of	of	ADP
fcis-30390	73	6	xinjiang	xinjiang	PROPN
fcis-30390	73	7	,	,	PUNCT
fcis-30390	73	8	characterized	characterize	VERB
fcis-30390	73	9	as	as	ADP
fcis-30390	73	10	a	a	DET
fcis-30390	73	11	typical	typical	ADJ
fcis-30390	73	12	temperate	temperate	ADJ
fcis-30390	73	13	continental	continental	ADJ
fcis-30390	73	14	arid	arid	NOUN
fcis-30390	73	15	climate	climate	NOUN
fcis-30390	73	16	zone	zone	NOUN
fcis-30390	73	17	,	,	PUNCT
fcis-30390	73	18	exhibits	exhibit	VERB
fcis-30390	73	19	distinctive	distinctive	ADJ
fcis-30390	73	20	features	feature	NOUN
fcis-30390	73	21	including	include	VERB
fcis-30390	73	22	sparse	sparse	ADJ
fcis-30390	73	23	precipitation	precipitation	NOUN
fcis-30390	73	24	,	,	PUNCT
fcis-30390	73	25	intense	intense	ADJ
fcis-30390	73	26	evaporation	evaporation	NOUN
fcis-30390	73	27	,	,	PUNCT
fcis-30390	73	28	and	and	CCONJ
fcis-30390	73	29	significant	significant	ADJ
fcis-30390	73	30	daily	daily	ADJ
fcis-30390	73	31	temperature	temperature	NOUN
fcis-30390	73	32	variations	variation	NOUN
fcis-30390	73	33	.	.	PUNCT
fcis-30390	74	1	the	the	DET
fcis-30390	74	2	ratio	ratio	NOUN
fcis-30390	74	3	of	of	ADP
fcis-30390	74	4	annual	annual	ADJ
fcis-30390	74	5	average	average	ADJ
fcis-30390	74	6	precipitation	precipitation	NOUN
fcis-30390	74	7	to	to	ADP
fcis-30390	74	8	potential	potential	ADJ
fcis-30390	74	9	evaporation	evaporation	NOUN
fcis-30390	74	10	reaches	reach	VERB
fcis-30390	74	11	0.6	0.6	NUM
fcis-30390	74	12	,	,	PUNCT
fcis-30390	74	13	indicating	indicate	VERB
fcis-30390	74	14	that	that	SCONJ
fcis-30390	74	15	precipitation	precipitation	NOUN
fcis-30390	74	16	amounts	amount	VERB
fcis-30390	74	17	to	to	ADP
fcis-30390	74	18	approximately	approximately	ADV
fcis-30390	74	19	60	60	NUM
fcis-30390	74	20	%	%	NOUN
fcis-30390	74	21	of	of	ADP
fcis-30390	74	22	potential	potential	ADJ
fcis-30390	74	23	evaporation	evaporation	NOUN
fcis-30390	74	24	.	.	PUNCT
fcis-30390	75	1	these	these	DET
fcis-30390	75	2	unique	unique	ADJ
fcis-30390	75	3	climatic	climatic	ADJ
fcis-30390	75	4	conditions	condition	NOUN
fcis-30390	75	5	have	have	AUX
fcis-30390	75	6	resulted	result	VERB
fcis-30390	75	7	in	in	ADP
fcis-30390	75	8	an	an	DET
fcis-30390	75	9	extremely	extremely	ADV
fcis-30390	75	10	fragile	fragile	ADJ
fcis-30390	75	11	ecosystem	ecosystem	NOUN
fcis-30390	75	12	,	,	PUNCT
fcis-30390	75	13	rendering	render	VERB
fcis-30390	75	14	the	the	DET
fcis-30390	75	15	region	region	NOUN
fcis-30390	75	16	highly	highly	ADV
fcis-30390	75	17	sensitive	sensitive	ADJ
fcis-30390	75	18	to	to	ADP
fcis-30390	75	19	climate	climate	NOUN
fcis-30390	75	20	changes	change	NOUN
fcis-30390	75	21	and	and	CCONJ
fcis-30390	75	22	posing	pose	VERB
fcis-30390	75	23	significant	significant	ADJ
fcis-30390	75	24	challenges	challenge	NOUN
fcis-30390	75	25	in	in	ADP
fcis-30390	75	26	ecological	ecological	ADJ
fcis-30390	75	27	conservation	conservation	NOUN
fcis-30390	75	28	.	.	PUNCT
fcis-30390	76	1	in	in	ADP
fcis-30390	76	2	this	this	DET
fcis-30390	76	3	study	study	NOUN
fcis-30390	76	4	,	,	PUNCT
fcis-30390	76	5	we	we	PRON
fcis-30390	76	6	selected	select	VERB
fcis-30390	76	7	meteorological	meteorological	ADJ
fcis-30390	76	8	data	datum	NOUN
fcis-30390	76	9	from	from	ADP
fcis-30390	76	10	five	five	NUM
fcis-30390	76	11	representative	representative	ADJ
fcis-30390	76	12	weather	weather	NOUN
fcis-30390	76	13	stations	station	NOUN
fcis-30390	76	14	:	:	PUNCT
fcis-30390	76	15	korla	korla	PROPN
fcis-30390	76	16	,	,	PUNCT
fcis-30390	76	17	kuche	kuche	PROPN
fcis-30390	76	18	,	,	PUNCT
fcis-30390	76	19	aral	aral	PROPN
fcis-30390	76	20	,	,	PUNCT
fcis-30390	76	21	kashgar	kashgar	NOUN
fcis-30390	76	22	,	,	PUNCT
fcis-30390	76	23	and	and	CCONJ
fcis-30390	76	24	hotan	hotan	NOUN
fcis-30390	76	25	.	.	PUNCT
fcis-30390	77	1	the	the	DET
fcis-30390	77	2	dataset	dataset	NOUN
fcis-30390	77	3	was	be	AUX
fcis-30390	77	4	sourced	source	VERB
fcis-30390	77	5	from	from	ADP
fcis-30390	77	6	the	the	DET
fcis-30390	77	7	meteorological	meteorological	ADJ
fcis-30390	77	8	website	website	NOUN
fcis-30390	77	9	https://rp5.ru/	https://rp5.ru/	NOUN
fcis-30390	77	10	,	,	PUNCT
fcis-30390	77	11	spanning	span	VERB
fcis-30390	77	12	from	from	ADP
fcis-30390	77	13	january	january	PROPN
fcis-30390	77	14	2005	2005	NUM
fcis-30390	77	15	to	to	ADP
fcis-30390	77	16	january	january	PROPN
fcis-30390	77	17	2025	2025	NUM
fcis-30390	77	18	.	.	PUNCT
fcis-30390	78	1	data	datum	NOUN
fcis-30390	78	2	sampling	sample	VERB
fcis-30390	78	3	was	be	AUX
fcis-30390	78	4	conducted	conduct	VERB
fcis-30390	78	5	at	at	ADP
fcis-30390	78	6	three	three	NUM
fcis-30390	78	7	-	-	PUNCT
fcis-30390	78	8	hour	hour	NOUN
fcis-30390	78	9	intervals	interval	NOUN
fcis-30390	78	10	across	across	ADP
fcis-30390	78	11	eight	eight	NUM
fcis-30390	78	12	daily	daily	ADJ
fcis-30390	78	13	timestamps	timestamp	NOUN
fcis-30390	78	14	(	(	PUNCT
fcis-30390	78	15	02:00	02:00	NUM
fcis-30390	78	16	,	,	PUNCT
fcis-30390	78	17	05:00	05:00	NUM
fcis-30390	78	18	,	,	PUNCT
fcis-30390	78	19	08:00	08:00	NUM
fcis-30390	78	20	,	,	PUNCT
fcis-30390	78	21	11:00	11:00	NUM
fcis-30390	78	22	,	,	PUNCT
fcis-30390	78	23	14:00	14:00	NUM
fcis-30390	78	24	,	,	PUNCT
fcis-30390	78	25	17:00	17:00	NUM
fcis-30390	78	26	,	,	PUNCT
fcis-30390	78	27	20:00	20:00	NUM
fcis-30390	78	28	,	,	PUNCT
fcis-30390	78	29	and	and	CCONJ
fcis-30390	78	30	23:00	23:00	NUM
fcis-30390	78	31	)	)	PUNCT
fcis-30390	78	32	,	,	PUNCT
fcis-30390	78	33	encompassing	encompass	VERB
fcis-30390	78	34	one	one	NUM
fcis-30390	78	35	temporal	temporal	ADJ
fcis-30390	78	36	feature	feature	NOUN
fcis-30390	78	37	and	and	CCONJ
fcis-30390	78	38	28	28	NUM
fcis-30390	78	39	meteorological	meteorological	ADJ
fcis-30390	78	40	parameters	parameter	NOUN
fcis-30390	78	41	in	in	ADP
fcis-30390	78	42	its	its	PRON
fcis-30390	78	43	dimensional	dimensional	ADJ
fcis-30390	78	44	space	space	NOUN
fcis-30390	78	45	.	.	PUNCT
fcis-30390	79	1	due	due	ADP
fcis-30390	79	2	to	to	ADP
fcis-30390	79	3	multiple	multiple	ADJ
fcis-30390	79	4	factors	factor	NOUN
fcis-30390	79	5	including	include	VERB
fcis-30390	79	6	sensor	sensor	NOUN
fcis-30390	79	7	malfunctions	malfunction	NOUN
fcis-30390	79	8	,	,	PUNCT
fcis-30390	79	9	network	network	NOUN
fcis-30390	79	10	communication	communication	NOUN
fcis-30390	79	11	issues	issue	NOUN
fcis-30390	79	12	,	,	PUNCT
fcis-30390	79	13	and	and	CCONJ
fcis-30390	79	14	data	datum	NOUN
fcis-30390	79	15	storage	storage	NOUN
fcis-30390	79	16	/	/	SYM
fcis-30390	79	17	transmission	transmission	NOUN
fcis-30390	79	18	problems	problem	NOUN
fcis-30390	79	19	,	,	PUNCT
fcis-30390	79	20	missing	miss	VERB
fcis-30390	79	21	values	value	NOUN
fcis-30390	79	22	were	be	AUX
fcis-30390	79	23	inevitable	inevitable	ADJ
fcis-30390	79	24	in	in	ADP
fcis-30390	79	25	the	the	DET
fcis-30390	79	26	raw	raw	ADJ
fcis-30390	79	27	dataset	dataset	NOUN
fcis-30390	79	28	,	,	PUNCT
fcis-30390	79	29	particularly	particularly	ADV
fcis-30390	79	30	prominent	prominent	ADJ
fcis-30390	79	31	in	in	ADP
fcis-30390	79	32	earlier	early	ADJ
fcis-30390	79	33	years	year	NOUN
fcis-30390	79	34	,	,	PUNCT
fcis-30390	79	35	though	though	SCONJ
fcis-30390	79	36	showing	show	VERB
fcis-30390	79	37	significant	significant	ADJ
fcis-30390	79	38	improvement	improvement	NOUN
fcis-30390	79	39	in	in	ADP
fcis-30390	79	40	recent	recent	ADJ
fcis-30390	79	41	data	datum	NOUN
fcis-30390	79	42	collection	collection	NOUN
fcis-30390	79	43	.	.	PUNCT
fcis-30390	80	1	feature	feature	NOUN
fcis-30390	80	2	elimination	elimination	NOUN
fcis-30390	80	3	was	be	AUX
fcis-30390	80	4	performed	perform	VERB
fcis-30390	80	5	for	for	ADP
fcis-30390	80	6	columns	column	NOUN
fcis-30390	80	7	with	with	ADP
fcis-30390	80	8	extensive	extensive	ADJ
fcis-30390	80	9	missing	missing	ADJ
fcis-30390	80	10	data	datum	NOUN
fcis-30390	80	11	or	or	CCONJ
fcis-30390	80	12	no	no	DET
fcis-30390	80	13	data	datum	NOUN
fcis-30390	80	14	,	,	PUNCT
fcis-30390	80	15	specifically	specifically	ADV
fcis-30390	80	16	for	for	ADP
fcis-30390	80	17	parameters	parameter	NOUN
fcis-30390	80	18	including	include	VERB
fcis-30390	80	19	ff10	ff10	PROPN
fcis-30390	80	20	,	,	PUNCT
fcis-30390	80	21	n	n	CCONJ
fcis-30390	80	22	,	,	PUNCT
fcis-30390	80	23	cl	cl	NOUN
fcis-30390	80	24	,	,	PUNCT
fcis-30390	80	25	nh	nh	PROPN
fcis-30390	80	26	,	,	PUNCT
fcis-30390	80	27	cm	cm	PROPN
fcis-30390	80	28	,	,	PUNCT
fcis-30390	80	29	ch	ch	NOUN
fcis-30390	80	30	,	,	PUNCT
fcis-30390	80	31	e	e	PROPN
fcis-30390	80	32	,	,	PUNCT
fcis-30390	80	33	tg	tg	PROPN
fcis-30390	80	34	,	,	PUNCT
fcis-30390	80	35	sss	sss	PROPN
fcis-30390	80	36	,	,	PUNCT
fcis-30390	80	37	tr	tr	VERB
fcis-30390	80	38	,	,	PUNCT
fcis-30390	80	39	e	e	NOUN
fcis-30390	80	40	,	,	PUNCT
fcis-30390	80	41	and	and	CCONJ
fcis-30390	80	42	rrr	rrr	NOUN
fcis-30390	80	43	.	.	PUNCT
fcis-30390	81	1	for	for	ADP
fcis-30390	81	2	the	the	DET
fcis-30390	81	3	remaining	remain	VERB
fcis-30390	81	4	missing	miss	VERB
fcis-30390	81	5	values	value	NOUN
fcis-30390	81	6	,	,	PUNCT
fcis-30390	81	7	considering	consider	VERB
fcis-30390	81	8	the	the	DET
fcis-30390	81	9	smooth	smooth	ADJ
fcis-30390	81	10	nature	nature	NOUN
fcis-30390	81	11	of	of	ADP
fcis-30390	81	12	weather	weather	NOUN
fcis-30390	81	13	variations	variation	NOUN
fcis-30390	81	14	within	within	ADP
fcis-30390	81	15	three	three	NUM
fcis-30390	81	16	-	-	PUNCT
fcis-30390	81	17	hour	hour	NOUN
fcis-30390	81	18	sampling	sampling	NOUN
fcis-30390	81	19	intervals	interval	NOUN
fcis-30390	81	20	,	,	PUNCT
fcis-30390	81	21	we	we	PRON
fcis-30390	81	22	employed	employ	VERB
fcis-30390	81	23	a	a	DET
fcis-30390	81	24	combination	combination	NOUN
fcis-30390	81	25	of	of	ADP
fcis-30390	81	26	data	datum	NOUN
fcis-30390	81	27	imputation	imputation	NOUN
fcis-30390	81	28	techniques	technique	NOUN
fcis-30390	81	29	including	include	VERB
fcis-30390	81	30	forward	forward	ADJ
fcis-30390	81	31	fill	fill	NOUN
fcis-30390	81	32	imputation	imputation	NOUN
fcis-30390	81	33	,	,	PUNCT
fcis-30390	81	34	random	random	ADJ
fcis-30390	81	35	imputation	imputation	NOUN
fcis-30390	81	36	,	,	PUNCT
fcis-30390	81	37	and	and	CCONJ
fcis-30390	81	38	linear	linear	ADJ
fcis-30390	81	39	interpolation	interpolation	NOUN
fcis-30390	81	40	to	to	PART
fcis-30390	81	41	ensure	ensure	VERB
fcis-30390	81	42	scientific	scientific	ADJ
fcis-30390	81	43	integrity	integrity	NOUN
fcis-30390	81	44	in	in	ADP
fcis-30390	81	45	our	our	PRON
fcis-30390	81	46	data	data	NOUN
fcis-30390	81	47	reconstruction	reconstruction	NOUN
fcis-30390	81	48	process	process	NOUN
fcis-30390	81	49	while	while	SCONJ
fcis-30390	81	50	maintaining	maintain	VERB
fcis-30390	81	51	the	the	DET
fcis-30390	81	52	temporal	temporal	ADJ
fcis-30390	81	53	consistency	consistency	NOUN
fcis-30390	81	54	of	of	ADP
fcis-30390	81	55	meteorological	meteorological	ADJ
fcis-30390	81	56	measurements	measurement	NOUN
fcis-30390	81	57	.	.	PUNCT
fcis-30390	82	1	table	table	NOUN
fcis-30390	82	2	1	1	NUM
fcis-30390	82	3	presents	present	VERB
fcis-30390	82	4	the	the	DET
fcis-30390	82	5	definitions	definition	NOUN
fcis-30390	82	6	and	and	CCONJ
fcis-30390	82	7	units	unit	NOUN
fcis-30390	82	8	of	of	ADP
fcis-30390	82	9	the	the	DET
fcis-30390	82	10	date	date	NOUN
fcis-30390	82	11	and	and	CCONJ
fcis-30390	82	12	meteorological	meteorological	ADJ
fcis-30390	82	13	attributes	attribute	NOUN
fcis-30390	82	14	in	in	ADP
fcis-30390	82	15	the	the	DET
fcis-30390	82	16	dataset	dataset	NOUN
fcis-30390	82	17	and	and	CCONJ
fcis-30390	82	18	table	table	NOUN
fcis-30390	82	19	2	2	NUM
fcis-30390	82	20	details	detail	NOUN
fcis-30390	82	21	their	their	PRON
fcis-30390	82	22	latitude	latitude	NOUN
fcis-30390	82	23	,	,	PUNCT
fcis-30390	82	24	longitude	longitude	NOUN
fcis-30390	82	25	,	,	PUNCT
fcis-30390	82	26	and	and	CCONJ
fcis-30390	82	27	altitude	altitude	NOUN
fcis-30390	82	28	information	information	NOUN
fcis-30390	82	29	.	.	PUNCT
fcis-30390	83	1	4.2	4.2	NUM
fcis-30390	83	2	.	.	PUNCT
fcis-30390	84	1	experimental	experimental	ADJ
fcis-30390	84	2	settings	setting	NOUN
fcis-30390	84	3	our	our	PRON
fcis-30390	84	4	model	model	NOUN
fcis-30390	84	5	implementation	implementation	NOUN
fcis-30390	84	6	is	be	AUX
fcis-30390	84	7	based	base	VERB
fcis-30390	84	8	on	on	ADP
fcis-30390	84	9	the	the	DET
fcis-30390	84	10	pytorch	pytorch	NOUN
fcis-30390	84	11	framework	framework	NOUN
fcis-30390	84	12	,	,	PUNCT
fcis-30390	84	13	with	with	ADP
fcis-30390	84	14	all	all	DET
fcis-30390	84	15	experiments	experiment	NOUN
fcis-30390	84	16	conducted	conduct	VERB
fcis-30390	84	17	on	on	ADP
fcis-30390	84	18	a	a	DET
fcis-30390	84	19	single	single	ADJ
fcis-30390	84	20	nvidia	nvidia	PROPN
fcis-30390	84	21	gpu	gpu	NOUN
fcis-30390	84	22	for	for	ADP
fcis-30390	84	23	training	training	NOUN
fcis-30390	84	24	and	and	CCONJ
fcis-30390	84	25	evaluation	evaluation	NOUN
fcis-30390	84	26	.	.	PUNCT
fcis-30390	85	1	to	to	PART
fcis-30390	85	2	ensure	ensure	VERB
fcis-30390	85	3	experimental	experimental	ADJ
fcis-30390	85	4	reproducibility	reproducibility	NOUN
fcis-30390	85	5	,	,	PUNCT
fcis-30390	85	6	we	we	PRON
fcis-30390	85	7	set	set	VERB
fcis-30390	85	8	the	the	DET
fcis-30390	85	9	random	random	ADJ
fcis-30390	85	10	seed	seed	NOUN
fcis-30390	85	11	to	to	ADP
fcis-30390	85	12	2021	2021	NUM
fcis-30390	85	13	.	.	PUNCT
fcis-30390	86	1	the	the	DET
fcis-30390	86	2	specific	specific	ADJ
fcis-30390	86	3	model	model	NOUN
fcis-30390	86	4	hyperparameters	hyperparameter	NOUN
fcis-30390	86	5	are	be	AUX
fcis-30390	86	6	as	as	SCONJ
fcis-30390	86	7	follows	follow	VERB
fcis-30390	86	8	table	table	NOUN
fcis-30390	86	9	3	3	NUM
fcis-30390	86	10	:	:	SYM
fcis-30390	86	11	39	39	NUM
fcis-30390	86	12	table	table	NOUN
fcis-30390	86	13	1	1	NUM
fcis-30390	86	14	.	.	PUNCT
fcis-30390	86	15	original	original	ADJ
fcis-30390	86	16	data	datum	NOUN
fcis-30390	86	17	feature	feature	NOUN
fcis-30390	86	18	information	information	NOUN
fcis-30390	86	19	.	.	PUNCT
fcis-30390	87	1	contains	contain	VERB
fcis-30390	87	2	the	the	DET
fcis-30390	87	3	meanings	meaning	NOUN
fcis-30390	87	4	of	of	ADP
fcis-30390	87	5	29	29	NUM
fcis-30390	87	6	meteorological	meteorological	ADJ
fcis-30390	87	7	features	feature	NOUN
fcis-30390	87	8	and	and	CCONJ
fcis-30390	87	9	their	their	PRON
fcis-30390	87	10	units	unit	NOUN
fcis-30390	87	11	,	,	PUNCT
fcis-30390	87	12	including	include	VERB
fcis-30390	87	13	one	one	NUM
fcis-30390	87	14	date	date	NOUN
fcis-30390	87	15	feature	feature	NOUN
fcis-30390	87	16	and	and	CCONJ
fcis-30390	87	17	28	28	NUM
fcis-30390	87	18	meteorological	meteorological	ADJ
fcis-30390	87	19	features	feature	NOUN
fcis-30390	87	20	.	.	PUNCT
fcis-30390	88	1	feature	feature	NOUN
fcis-30390	88	2	connotation	connotation	PROPN
fcis-30390	88	3	unit	unit	NOUN
fcis-30390	88	4	date	date	NOUN
fcis-30390	88	5	the	the	DET
fcis-30390	88	6	time	time	NOUN
fcis-30390	88	7	at	at	ADP
fcis-30390	88	8	this	this	DET
fcis-30390	88	9	weather	weather	NOUN
fcis-30390	88	10	point	point	NOUN
fcis-30390	88	11	,	,	PUNCT
fcis-30390	88	12	taking	take	VERB
fcis-30390	88	13	into	into	ADP
fcis-30390	88	14	account	account	NOUN
fcis-30390	88	15	daylight	daylight	NOUN
fcis-30390	88	16	saving	save	VERB
fcis-30390	88	17	time	time	NOUN
fcis-30390	88	18	/	/	SYM
fcis-30390	88	19	winter	winter	NOUN
fcis-30390	88	20	time	time	NOUN
fcis-30390	88	21	/	/	SYM
fcis-30390	88	22	t	t	X
fcis-30390	88	23	atmospheric	atmospheric	ADJ
fcis-30390	88	24	temperature	temperature	NOUN
fcis-30390	88	25	at	at	ADP
fcis-30390	88	26	2	2	NUM
fcis-30390	88	27	m	m	NOUN
fcis-30390	88	28	above	above	ADP
fcis-30390	88	29	ground	ground	NOUN
fcis-30390	88	30	level	level	NOUN
fcis-30390	88	31	°	°	ADP
fcis-30390	88	32	c	c	PROPN
fcis-30390	88	33	po	po	X
fcis-30390	88	34	atmospheric	atmospheric	ADJ
fcis-30390	88	35	pressure	pressure	NOUN
fcis-30390	88	36	at	at	ADP
fcis-30390	88	37	the	the	DET
fcis-30390	88	38	level	level	NOUN
fcis-30390	88	39	of	of	ADP
fcis-30390	88	40	the	the	DET
fcis-30390	88	41	weather	weather	NOUN
fcis-30390	88	42	station	station	NOUN
fcis-30390	88	43	mmhg	mmhg	PROPN
fcis-30390	88	44	p	p	PROPN
fcis-30390	88	45	atmospheric	atmospheric	ADJ
fcis-30390	88	46	pressure	pressure	NOUN
fcis-30390	88	47	at	at	ADP
fcis-30390	88	48	mean	mean	PROPN
fcis-30390	88	49	sea	sea	NOUN
fcis-30390	88	50	level	level	NOUN
fcis-30390	88	51	mmhg	mmhg	PROPN
fcis-30390	88	52	pa	pa	PROPN
fcis-30390	88	53	change	change	NOUN
fcis-30390	88	54	in	in	ADP
fcis-30390	88	55	atmospheric	atmospheric	ADJ
fcis-30390	88	56	pressure	pressure	NOUN
fcis-30390	88	57	during	during	ADP
fcis-30390	88	58	the	the	DET
fcis-30390	88	59	3	3	NUM
fcis-30390	88	60	h	h	NOUN
fcis-30390	88	61	preceding	precede	VERB
fcis-30390	88	62	the	the	DET
fcis-30390	88	63	observation	observation	NOUN
fcis-30390	88	64	mmhg	mmhg	PROPN
fcis-30390	88	65	u	u	PROPN
fcis-30390	88	66	relative	relative	ADJ
fcis-30390	88	67	humidity	humidity	NOUN
fcis-30390	88	68	at	at	ADP
fcis-30390	88	69	2	2	NUM
fcis-30390	88	70	m	m	NOUN
fcis-30390	88	71	above	above	ADP
fcis-30390	88	72	ground	ground	NOUN
fcis-30390	88	73	level	level	NOUN
fcis-30390	88	74	%	%	NOUN
fcis-30390	88	75	dd	dd	VERB
fcis-30390	88	76	the	the	DET
fcis-30390	88	77	wind	wind	NOUN
fcis-30390	88	78	direction	direction	NOUN
fcis-30390	88	79	at	at	ADP
fcis-30390	88	80	10	10	NUM
fcis-30390	88	81	-	-	SYM
fcis-30390	88	82	12	12	NUM
fcis-30390	88	83	m	m	NOUN
fcis-30390	88	84	above	above	ADP
fcis-30390	88	85	ground	ground	NOUN
fcis-30390	88	86	level	level	NOUN
fcis-30390	88	87	during	during	ADP
fcis-30390	88	88	the	the	DET
fcis-30390	88	89	10	10	NUM
fcis-30390	88	90	min	min	NOUN
fcis-30390	88	91	prior	prior	ADV
fcis-30390	88	92	to	to	ADP
fcis-30390	88	93	observation	observation	NOUN
fcis-30390	88	94	/	/	SYM
fcis-30390	88	95	ff	ff	NOUN
fcis-30390	88	96	average	average	ADJ
fcis-30390	88	97	wind	wind	NOUN
fcis-30390	88	98	speed	speed	NOUN
fcis-30390	88	99	at	at	ADP
fcis-30390	88	100	ground	ground	NOUN
fcis-30390	88	101	level	level	NOUN
fcis-30390	88	102	of	of	ADP
fcis-30390	88	103	10	10	NUM
fcis-30390	88	104	-	-	SYM
fcis-30390	88	105	12	12	NUM
fcis-30390	88	106	m	m	NOUN
fcis-30390	88	107	during	during	ADP
fcis-30390	88	108	the	the	DET
fcis-30390	88	109	10	10	NUM
fcis-30390	88	110	min	min	NOUN
fcis-30390	88	111	prior	prior	ADV
fcis-30390	88	112	to	to	ADP
fcis-30390	88	113	observation	observation	NOUN
fcis-30390	88	114	m	m	PROPN
fcis-30390	88	115	/	/	SYM
fcis-30390	88	116	s	s	PART
fcis-30390	88	117	ff10	ff10	PROPN
fcis-30390	88	118	maximum	maximum	ADJ
fcis-30390	88	119	gusts	gust	NOUN
fcis-30390	88	120	at	at	ADP
fcis-30390	88	121	10	10	NUM
fcis-30390	88	122	-	-	SYM
fcis-30390	88	123	12	12	NUM
fcis-30390	88	124	meters	meter	NOUN
fcis-30390	88	125	above	above	ADP
fcis-30390	88	126	ground	ground	NOUN
fcis-30390	88	127	level	level	NOUN
fcis-30390	88	128	during	during	ADP
fcis-30390	88	129	the	the	DET
fcis-30390	88	130	10	10	NUM
fcis-30390	88	131	min	min	NOUN
fcis-30390	88	132	prior	prior	ADV
fcis-30390	88	133	to	to	ADP
fcis-30390	88	134	observation	observation	NOUN
fcis-30390	88	135	m	m	PROPN
fcis-30390	88	136	/	/	SYM
fcis-30390	88	137	s	s	PART
fcis-30390	88	138	ff3	ff3	ADJ
fcis-30390	88	139	maximum	maximum	ADJ
fcis-30390	88	140	gusts	gust	NOUN
fcis-30390	88	141	at	at	ADP
fcis-30390	88	142	ground	ground	NOUN
fcis-30390	88	143	level	level	NOUN
fcis-30390	88	144	of	of	ADP
fcis-30390	88	145	10	10	NUM
fcis-30390	88	146	-	-	SYM
fcis-30390	88	147	12	12	NUM
fcis-30390	88	148	m	m	NOUN
fcis-30390	88	149	between	between	ADP
fcis-30390	88	150	observations	observation	NOUN
fcis-30390	88	151	m	m	PROPN
fcis-30390	88	152	/	/	SYM
fcis-30390	88	153	s	s	PART
fcis-30390	88	154	n	n	NUM
fcis-30390	88	155	total	total	ADJ
fcis-30390	88	156	cloud	cloud	NOUN
fcis-30390	88	157	cover	cover	NOUN
fcis-30390	88	158	/	/	SYM
fcis-30390	88	159	ww	ww	PROPN
fcis-30390	88	160	weather	weather	NOUN
fcis-30390	88	161	conditions	condition	NOUN
fcis-30390	88	162	for	for	ADP
fcis-30390	88	163	the	the	DET
fcis-30390	88	164	day	day	NOUN
fcis-30390	88	165	as	as	SCONJ
fcis-30390	88	166	announced	announce	VERB
fcis-30390	88	167	by	by	ADP
fcis-30390	88	168	the	the	DET
fcis-30390	88	169	weather	weather	NOUN
fcis-30390	88	170	station	station	NOUN
fcis-30390	88	171	/	/	SYM
fcis-30390	88	172	w1	w1	PROPN
fcis-30390	88	173	past	past	PROPN
fcis-30390	88	174	weather	weather	NOUN
fcis-30390	88	175	during	during	ADP
fcis-30390	88	176	the	the	DET
fcis-30390	88	177	observation	observation	NOUN
fcis-30390	88	178	period	period	NOUN
fcis-30390	88	179	1	1	NUM
fcis-30390	88	180	/	/	SYM
fcis-30390	88	181	w2	w2	NOUN
fcis-30390	88	182	past	past	ADJ
fcis-30390	88	183	weather	weather	NOUN
fcis-30390	88	184	during	during	ADP
fcis-30390	88	185	the	the	DET
fcis-30390	88	186	observation	observation	NOUN
fcis-30390	88	187	period	period	NOUN
fcis-30390	88	188	2	2	NUM
fcis-30390	88	189	/	/	SYM
fcis-30390	88	190	tn	tn	NOUN
fcis-30390	88	191	lowest	low	ADJ
fcis-30390	88	192	temperature	temperature	NOUN
fcis-30390	88	193	in	in	ADP
fcis-30390	88	194	the	the	DET
fcis-30390	88	195	past	past	ADJ
fcis-30390	88	196	period	period	NOUN
fcis-30390	88	197	(	(	PUNCT
fcis-30390	88	198	not	not	PART
fcis-30390	88	199	more	more	ADJ
fcis-30390	88	200	than	than	ADP
fcis-30390	88	201	12	12	NUM
fcis-30390	88	202	h	h	NOUN
fcis-30390	88	203	)	)	PUNCT
fcis-30390	89	1	°	°	PROPN
fcis-30390	89	2	c	c	PROPN
fcis-30390	89	3	tx	tx	ADP
fcis-30390	89	4	maximum	maximum	ADJ
fcis-30390	89	5	temperature	temperature	NOUN
fcis-30390	89	6	in	in	ADP
fcis-30390	89	7	the	the	DET
fcis-30390	89	8	past	past	ADJ
fcis-30390	89	9	period	period	NOUN
fcis-30390	89	10	(	(	PUNCT
fcis-30390	89	11	not	not	PART
fcis-30390	89	12	more	more	ADJ
fcis-30390	89	13	than	than	ADP
fcis-30390	89	14	12	12	NUM
fcis-30390	89	15	h	h	NOUN
fcis-30390	89	16	)	)	PUNCT
fcis-30390	89	17	°	°	PROPN
fcis-30390	89	18	c	c	NOUN
fcis-30390	89	19	cl	cl	NOUN
fcis-30390	89	20	stratocumulus	stratocumulus	NOUN
fcis-30390	89	21	,	,	PUNCT
fcis-30390	89	22	stratus	stratus	NOUN
fcis-30390	89	23	,	,	PUNCT
fcis-30390	89	24	cumulus	cumulus	NOUN
fcis-30390	89	25	and	and	CCONJ
fcis-30390	89	26	cumulonimbus	cumulonimbus	NOUN
fcis-30390	89	27	/	/	SYM
fcis-30390	89	28	nh	nh	PROPN
fcis-30390	89	29	number	number	NOUN
fcis-30390	89	30	of	of	ADP
fcis-30390	89	31	all	all	DET
fcis-30390	89	32	clouds	cloud	NOUN
fcis-30390	89	33	c1	c1	PROPN
fcis-30390	89	34	observed	observe	VERB
fcis-30390	89	35	,	,	PUNCT
fcis-30390	89	36	number	number	NOUN
fcis-30390	89	37	of	of	ADP
fcis-30390	89	38	all	all	DET
fcis-30390	89	39	clouds	cloud	NOUN
fcis-30390	89	40	cm	cm	NOUN
fcis-30390	89	41	observed	observe	VERB
fcis-30390	89	42	in	in	ADP
fcis-30390	89	43	the	the	DET
fcis-30390	89	44	absence	absence	NOUN
fcis-30390	89	45	of	of	ADP
fcis-30390	89	46	cloud	cloud	PROPN
fcis-30390	89	47	c1	c1	PROPN
fcis-30390	89	48	/	/	SYM
fcis-30390	89	49	h	h	PROPN
fcis-30390	89	50	cloud	cloud	PROPN
fcis-30390	89	51	base	base	NOUN
fcis-30390	89	52	height	height	NOUN
fcis-30390	89	53	m	m	NOUN
fcis-30390	89	54	cm	cm	NOUN
fcis-30390	89	55	altocumulus	altocumulus	ADJ
fcis-30390	89	56	,	,	PUNCT
fcis-30390	89	57	cirrus	cirrus	NOUN
fcis-30390	89	58	and	and	CCONJ
fcis-30390	89	59	nimbostratus	nimbostratus	NOUN
fcis-30390	89	60	/	/	SYM
fcis-30390	89	61	ch	ch	NOUN
fcis-30390	89	62	cirrus	cirrus	NOUN
fcis-30390	89	63	,	,	PUNCT
fcis-30390	89	64	cirrocumulus	cirrocumulus	NOUN
fcis-30390	89	65	and	and	CCONJ
fcis-30390	89	66	cirrostratus	cirrostratus	NOUN
fcis-30390	89	67	/	/	SYM
fcis-30390	90	1	vv	vv	CCONJ
fcis-30390	90	2	horizontal	horizontal	ADJ
fcis-30390	90	3	visibility	visibility	NOUN
fcis-30390	90	4	km	km	NOUN
fcis-30390	90	5	td	td	NOUN
fcis-30390	90	6	dew	dew	NOUN
fcis-30390	90	7	point	point	VERB
fcis-30390	90	8	temperature	temperature	NOUN
fcis-30390	90	9	at	at	ADP
fcis-30390	90	10	2	2	NUM
fcis-30390	90	11	m	m	NOUN
fcis-30390	90	12	above	above	ADP
fcis-30390	90	13	the	the	DET
fcis-30390	90	14	ground	ground	NOUN
fcis-30390	90	15	°	°	ADP
fcis-30390	90	16	c	c	NOUN
fcis-30390	90	17	rrr	rrr	NOUN
fcis-30390	90	18	precipitation	precipitation	NOUN
fcis-30390	90	19	within	within	ADP
fcis-30390	90	20	3	3	NUM
fcis-30390	90	21	h	h	NOUN
fcis-30390	90	22	mm	mm	NOUN
fcis-30390	90	23	tr	tr	NOUN
fcis-30390	90	24	time	time	NOUN
fcis-30390	90	25	to	to	PART
fcis-30390	90	26	reach	reach	VERB
fcis-30390	90	27	prescribed	prescribed	ADJ
fcis-30390	90	28	precipitation	precipitation	NOUN
fcis-30390	90	29	/	/	SYM
fcis-30390	91	1	e	e	NOUN
fcis-30390	91	2	condition	condition	NOUN
fcis-30390	91	3	of	of	ADP
fcis-30390	91	4	soil	soil	NOUN
fcis-30390	91	5	surface	surface	NOUN
fcis-30390	91	6	without	without	ADP
fcis-30390	91	7	snow	snow	NOUN
fcis-30390	91	8	or	or	CCONJ
fcis-30390	91	9	ice	ice	NOUN
fcis-30390	91	10	cover	cover	NOUN
fcis-30390	91	11	/	/	SYM
fcis-30390	91	12	tg	tg	X
fcis-30390	91	13	minimum	minimum	ADJ
fcis-30390	91	14	temperature	temperature	NOUN
fcis-30390	91	15	of	of	ADP
fcis-30390	91	16	the	the	DET
fcis-30390	91	17	soil	soil	NOUN
fcis-30390	91	18	surface	surface	NOUN
fcis-30390	91	19	at	at	ADP
fcis-30390	91	20	night	night	NOUN
fcis-30390	91	21	°	°	ADP
fcis-30390	91	22	c	c	NOUN
fcis-30390	91	23	e	e	NOUN
fcis-30390	91	24	’	'	PUNCT
fcis-30390	91	25	condition	condition	NOUN
fcis-30390	91	26	of	of	ADP
fcis-30390	91	27	soil	soil	NOUN
fcis-30390	91	28	surfaces	surface	NOUN
fcis-30390	91	29	with	with	ADP
fcis-30390	91	30	snow	snow	NOUN
fcis-30390	91	31	or	or	CCONJ
fcis-30390	91	32	ice	ice	NOUN
fcis-30390	91	33	cover	cover	NOUN
fcis-30390	91	34	/	/	SYM
fcis-30390	91	35	sss	sss	NOUN
fcis-30390	91	36	depth	depth	NOUN
fcis-30390	91	37	of	of	ADP
fcis-30390	91	38	snow	snow	NOUN
fcis-30390	91	39	cm	cm	NOUN
fcis-30390	91	40	table	table	NOUN
fcis-30390	91	41	2	2	NUM
fcis-30390	91	42	.	.	X
fcis-30390	91	43	information	information	NOUN
fcis-30390	91	44	on	on	ADP
fcis-30390	91	45	weather	weather	NOUN
fcis-30390	91	46	station	station	NOUN
fcis-30390	91	47	data	datum	NOUN
fcis-30390	91	48	station	station	NOUN
fcis-30390	91	49	no	no	PROPN
fcis-30390	91	50	.	.	PUNCT
fcis-30390	91	51	station	station	NOUN
fcis-30390	91	52	name	name	NOUN
fcis-30390	91	53	latitude	latitude	NOUN
fcis-30390	91	54	(	(	PUNCT
fcis-30390	91	55	°	°	PROPN
fcis-30390	91	56	)	)	PUNCT
fcis-30390	91	57	longitude	longitude	NOUN
fcis-30390	91	58	(	(	PUNCT
fcis-30390	91	59	°	°	PRON
fcis-30390	91	60	)	)	PUNCT
fcis-30390	91	61	altitude	altitude	NOUN
fcis-30390	91	62	(	(	PUNCT
fcis-30390	91	63	m	m	NOUN
fcis-30390	91	64	)	)	PUNCT
fcis-30390	91	65	51656	51656	NUM
fcis-30390	91	66	korla	korla	NOUN
fcis-30390	91	67	41.46	41.46	NUM
fcis-30390	91	68	86.09	86.09	NUM
fcis-30390	91	69	944	944	NUM
fcis-30390	91	70	51644	51644	NUM
fcis-30390	91	71	kuche	kuche	NOUN
fcis-30390	91	72	41.43	41.43	NUM
fcis-30390	91	73	82.57	82.57	NUM
fcis-30390	91	74	1100	1100	NUM
fcis-30390	91	75	51730	51730	NUM
fcis-30390	91	76	aral	aral	PROPN
fcis-30390	91	77	40.33	40.33	NUM
fcis-30390	91	78	81.17	81.17	NUM
fcis-30390	91	79	1012	1012	NUM
fcis-30390	91	80	57109	57109	NUM
fcis-30390	91	81	kashgar	kashgar	NOUN
fcis-30390	91	82	39.28	39.28	NUM
fcis-30390	91	83	75.59	75.59	NUM
fcis-30390	91	84	1269	1269	NUM
fcis-30390	91	85	51828	51828	NUM
fcis-30390	91	86	hotan	hotan	NOUN
fcis-30390	91	87	37.07	37.07	NUM
fcis-30390	91	88	79.55	79.55	NUM
fcis-30390	91	89	1377	1377	NUM
fcis-30390	91	90	table	table	NOUN
fcis-30390	91	91	3	3	NUM
fcis-30390	91	92	.	.	PUNCT
fcis-30390	91	93	model	model	NOUN
fcis-30390	91	94	architecture	architecture	NOUN
fcis-30390	91	95	.	.	PUNCT
fcis-30390	92	1	parameter	parameter	NOUN
fcis-30390	92	2	value	value	NOUN
fcis-30390	92	3	seq_len	seq_len	X
fcis-30390	92	4	96	96	NUM
fcis-30390	92	5	pred_len	pred_len	NUM
fcis-30390	92	6	6	6	NUM
fcis-30390	92	7	,	,	PUNCT
fcis-30390	92	8	12	12	NUM
fcis-30390	92	9	,	,	PUNCT
fcis-30390	92	10	18	18	NUM
fcis-30390	92	11	,	,	PUNCT
fcis-30390	92	12	24	24	NUM
fcis-30390	92	13	,	,	PUNCT
fcis-30390	92	14	30	30	NUM
fcis-30390	92	15	,	,	PUNCT
fcis-30390	92	16	36	36	NUM
fcis-30390	92	17	,	,	PUNCT
fcis-30390	92	18	42	42	NUM
fcis-30390	92	19	,	,	PUNCT
fcis-30390	92	20	48	48	NUM
fcis-30390	92	21	d_model	d_model	NOUN
fcis-30390	92	22	16	16	NUM
fcis-30390	92	23	n_heads	n_head	NOUN
fcis-30390	92	24	4	4	NUM
fcis-30390	92	25	e_layers	e_layer	NOUN
fcis-30390	92	26	3	3	NUM
fcis-30390	92	27	d_layers	d_layer	NOUN
fcis-30390	92	28	1	1	NUM
fcis-30390	92	29	d_ff	d_ff	NUM
fcis-30390	92	30	32	32	NUM
fcis-30390	92	31	moving	move	VERB
fcis-30390	92	32	average	average	ADJ
fcis-30390	92	33	window	window	NOUN
fcis-30390	92	34	size	size	NOUN
fcis-30390	92	35	25	25	NUM
fcis-30390	92	36	to	to	PART
fcis-30390	92	37	comprehensively	comprehensively	ADV
fcis-30390	92	38	evaluate	evaluate	VERB
fcis-30390	92	39	the	the	DET
fcis-30390	92	40	model	model	NOUN
fcis-30390	92	41	's	's	PART
fcis-30390	92	42	predictive	predictive	ADJ
fcis-30390	92	43	performance	performance	NOUN
fcis-30390	92	44	across	across	ADP
fcis-30390	92	45	various	various	ADJ
fcis-30390	92	46	temporal	temporal	ADJ
fcis-30390	92	47	scales	scale	NOUN
fcis-30390	92	48	,	,	PUNCT
fcis-30390	92	49	we	we	PRON
fcis-30390	92	50	employed	employ	VERB
fcis-30390	92	51	two	two	NUM
fcis-30390	92	52	primary	primary	ADJ
fcis-30390	92	53	evaluation	evaluation	NOUN
fcis-30390	92	54	metrics	metric	NOUN
fcis-30390	92	55	:	:	PUNCT
fcis-30390	92	56	mean	mean	VERB
fcis-30390	92	57	square	square	ADJ
fcis-30390	92	58	error	error	NOUN
fcis-30390	92	59	(	(	PUNCT
fcis-30390	92	60	mse	mse	NOUN
fcis-30390	92	61	)	)	PUNCT
fcis-30390	92	62	and	and	CCONJ
fcis-30390	92	63	mean	mean	VERB
fcis-30390	92	64	absolute	absolute	ADJ
fcis-30390	92	65	error	error	NOUN
fcis-30390	92	66	(	(	PUNCT
fcis-30390	92	67	mae	mae	PROPN
fcis-30390	92	68	)	)	PUNCT
fcis-30390	92	69	.	.	PUNCT
fcis-30390	93	1	through	through	ADP
fcis-30390	93	2	systematic	systematic	ADJ
fcis-30390	93	3	testing	testing	NOUN
fcis-30390	93	4	with	with	ADP
fcis-30390	93	5	prediction	prediction	NOUN
fcis-30390	93	6	horizons	horizon	NOUN
fcis-30390	93	7	ranging	range	VERB
fcis-30390	93	8	from	from	ADP
fcis-30390	93	9	6	6	NUM
fcis-30390	93	10	to	to	ADP
fcis-30390	93	11	48	48	NUM
fcis-30390	93	12	hours	hour	NOUN
fcis-30390	93	13	,	,	PUNCT
fcis-30390	93	14	we	we	PRON
fcis-30390	93	15	conducted	conduct	VERB
fcis-30390	93	16	a	a	DET
fcis-30390	93	17	thorough	thorough	ADJ
fcis-30390	93	18	assessment	assessment	NOUN
fcis-30390	93	19	of	of	ADP
fcis-30390	93	20	the	the	DET
fcis-30390	93	21	model	model	NOUN
fcis-30390	93	22	's	's	PART
fcis-30390	93	23	performance	performance	NOUN
fcis-30390	93	24	in	in	ADP
fcis-30390	93	25	short	short	ADJ
fcis-30390	93	26	-	-	PUNCT
fcis-30390	93	27	term	term	NOUN
fcis-30390	93	28	forecasting	forecasting	NOUN
fcis-30390	93	29	tasks	task	NOUN
fcis-30390	93	30	.	.	PUNCT
fcis-30390	94	1	this	this	DET
fcis-30390	94	2	progressive	progressive	ADJ
fcis-30390	94	3	evaluation	evaluation	NOUN
fcis-30390	94	4	approach	approach	NOUN
fcis-30390	94	5	not	not	PART
fcis-30390	94	6	only	only	ADV
fcis-30390	94	7	demonstrates	demonstrate	VERB
fcis-30390	94	8	the	the	DET
fcis-30390	94	9	model	model	NOUN
fcis-30390	94	10	's	's	PART
fcis-30390	94	11	predictive	predictive	ADJ
fcis-30390	94	12	capabilities	capability	NOUN
fcis-30390	94	13	in	in	ADP
fcis-30390	94	14	short	short	ADJ
fcis-30390	94	15	-	-	PUNCT
fcis-30390	94	16	term	term	NOUN
fcis-30390	94	17	scenarios	scenario	NOUN
fcis-30390	94	18	but	but	CCONJ
fcis-30390	94	19	also	also	ADV
fcis-30390	94	20	provides	provide	VERB
fcis-30390	94	21	substantial	substantial	ADJ
fcis-30390	94	22	evidence	evidence	NOUN
fcis-30390	94	23	for	for	ADP
fcis-30390	94	24	its	its	PRON
fcis-30390	94	25	reliability	reliability	NOUN
fcis-30390	94	26	in	in	ADP
fcis-30390	94	27	real	real	ADJ
fcis-30390	94	28	-	-	PUNCT
fcis-30390	94	29	world	world	NOUN
fcis-30390	94	30	applications	application	NOUN
fcis-30390	94	31	.	.	PUNCT
fcis-30390	95	1	time	time	NOUN
fcis-30390	95	2	series	series	PROPN
fcis-30390	95	3	forecasting	forecasting	NOUN
fcis-30390	95	4	commonly	commonly	ADV
fcis-30390	95	5	employs	employ	VERB
fcis-30390	95	6	three	three	NUM
fcis-30390	95	7	methodological	methodological	ADJ
fcis-30390	95	8	approaches	approach	NOUN
fcis-30390	95	9	:	:	PUNCT
fcis-30390	95	10	s	s	X
fcis-30390	95	11	(	(	PUNCT
fcis-30390	95	12	univariate	univariate	ADJ
fcis-30390	95	13	forecasting	forecasting	NOUN
fcis-30390	95	14	)	)	PUNCT
fcis-30390	95	15	,	,	PUNCT
fcis-30390	95	16	m	m	PROPN
fcis-30390	95	17	(	(	PUNCT
fcis-30390	95	18	multivariate	multivariate	NOUN
fcis-30390	95	19	forecasting	forecasting	NOUN
fcis-30390	95	20	)	)	PUNCT
fcis-30390	95	21	,	,	PUNCT
fcis-30390	95	22	and	and	CCONJ
fcis-30390	95	23	ms	ms	NOUN
fcis-30390	95	24	(	(	PUNCT
fcis-30390	95	25	multivariate	multivariate	NOUN
fcis-30390	95	26	-	-	PUNCT
fcis-30390	95	27	tounivariate	tounivariate	ADJ
fcis-30390	95	28	forecasting	forecasting	NOUN
fcis-30390	95	29	)	)	PUNCT
fcis-30390	95	30	,	,	PUNCT
fcis-30390	95	31	detailed	detail	VERB
fcis-30390	95	32	as	as	SCONJ
fcis-30390	95	33	follows	follow	VERB
fcis-30390	95	34	:	:	PUNCT
fcis-30390	95	35	s	s	X
fcis-30390	95	36	(	(	PUNCT
fcis-30390	95	37	univariate	univariate	ADJ
fcis-30390	95	38	forecasting	forecasting	NOUN
fcis-30390	95	39	):	):	PUNCT
fcis-30390	95	40	this	this	DET
fcis-30390	95	41	approach	approach	NOUN
fcis-30390	95	42	involves	involve	VERB
fcis-30390	95	43	predicting	predict	VERB
fcis-30390	95	44	a	a	DET
fcis-30390	95	45	single	single	ADJ
fcis-30390	95	46	variable	variable	NOUN
fcis-30390	95	47	using	use	VERB
fcis-30390	95	48	only	only	ADV
fcis-30390	95	49	its	its	PRON
fcis-30390	95	50	historical	historical	ADJ
fcis-30390	95	51	data	datum	NOUN
fcis-30390	95	52	.	.	PUNCT
fcis-30390	96	1	the	the	DET
fcis-30390	96	2	model	model	NOUN
fcis-30390	96	3	focuses	focus	VERB
fcis-30390	96	4	on	on	ADP
fcis-30390	96	5	the	the	DET
fcis-30390	96	6	temporal	temporal	ADJ
fcis-30390	96	7	patterns	pattern	NOUN
fcis-30390	96	8	within	within	ADP
fcis-30390	96	9	a	a	DET
fcis-30390	96	10	single	single	ADJ
fcis-30390	96	11	feature	feature	NOUN
fcis-30390	96	12	's	's	PART
fcis-30390	96	13	historical	historical	ADJ
fcis-30390	96	14	observations	observation	NOUN
fcis-30390	96	15	to	to	PART
fcis-30390	96	16	forecast	forecast	VERB
fcis-30390	96	17	future	future	ADJ
fcis-30390	96	18	values	value	NOUN
fcis-30390	96	19	,	,	PUNCT
fcis-30390	96	20	based	base	VERB
fcis-30390	96	21	on	on	ADP
fcis-30390	96	22	distributional	distributional	ADJ
fcis-30390	96	23	assumptions	assumption	NOUN
fcis-30390	96	24	.	.	PUNCT
fcis-30390	97	1	m	m	VERB
fcis-30390	97	2	(	(	PUNCT
fcis-30390	97	3	multivariate	multivariate	NOUN
fcis-30390	97	4	forecasting	forecasting	NOUN
fcis-30390	97	5	):	):	PUNCT
fcis-30390	97	6	this	this	DET
fcis-30390	97	7	method	method	NOUN
fcis-30390	97	8	encompasses	encompass	VERB
fcis-30390	97	9	predicting	predict	VERB
fcis-30390	97	10	multiple	multiple	ADJ
fcis-30390	97	11	variables	variable	NOUN
fcis-30390	97	12	simultaneously	simultaneously	ADV
fcis-30390	97	13	using	use	VERB
fcis-30390	97	14	multiple	multiple	ADJ
fcis-30390	97	15	related	relate	VERB
fcis-30390	97	16	features	feature	NOUN
fcis-30390	97	17	.	.	PUNCT
fcis-30390	98	1	the	the	DET
fcis-30390	98	2	model	model	NOUN
fcis-30390	98	3	leverages	leverage	VERB
fcis-30390	98	4	historical	historical	ADJ
fcis-30390	98	5	data	datum	NOUN
fcis-30390	98	6	from	from	ADP
fcis-30390	98	7	multiple	multiple	ADJ
fcis-30390	98	8	correlated	correlate	VERB
fcis-30390	98	9	variables	variable	NOUN
fcis-30390	98	10	to	to	PART
fcis-30390	98	11	forecast	forecast	VERB
fcis-30390	98	12	future	future	ADJ
fcis-30390	98	13	values	value	NOUN
fcis-30390	98	14	for	for	ADP
fcis-30390	98	15	all	all	DET
fcis-30390	98	16	target	target	NOUN
fcis-30390	98	17	variables	variable	NOUN
fcis-30390	98	18	concurrently	concurrently	ADV
fcis-30390	98	19	.	.	PUNCT
fcis-30390	99	1	ms(multivariate	ms(multivariate	PROPN
fcis-30390	99	2	-	-	PUNCT
fcis-30390	99	3	to	to	ADP
fcis-30390	99	4	-	-	PUNCT
fcis-30390	99	5	univariateforecasting):this	univariateforecasting):this	ADJ
fcis-30390	99	6	approach	approach	NOUN
fcis-30390	99	7	utilizes	utilize	VERB
fcis-30390	99	8	multiple	multiple	ADJ
fcis-30390	99	9	features	feature	NOUN
fcis-30390	99	10	to	to	PART
fcis-30390	99	11	predict	predict	VERB
fcis-30390	99	12	a	a	DET
fcis-30390	99	13	single	single	ADJ
fcis-30390	99	14	target	target	NOUN
fcis-30390	99	15	variable	variable	NOUN
fcis-30390	99	16	.	.	PUNCT
fcis-30390	100	1	given	give	VERB
fcis-30390	100	2	that	that	DET
fcis-30390	100	3	precipitation	precipitation	NOUN
fcis-30390	100	4	is	be	AUX
fcis-30390	100	5	influenced	influence	VERB
fcis-30390	100	6	by	by	ADP
fcis-30390	100	7	multiple	multiple	ADJ
fcis-30390	100	8	factors	factor	NOUN
fcis-30390	100	9	exhibiting	exhibit	VERB
fcis-30390	100	10	complex	complex	ADJ
fcis-30390	100	11	non	non	ADJ
fcis-30390	100	12	-	-	ADJ
fcis-30390	100	13	linear	linear	ADJ
fcis-30390	100	14	relationships	relationship	NOUN
fcis-30390	100	15	,	,	PUNCT
fcis-30390	100	16	employing	employ	VERB
fcis-30390	100	17	a	a	DET
fcis-30390	100	18	univariate	univariate	ADJ
fcis-30390	100	19	approach	approach	NOUN
fcis-30390	100	20	(	(	PUNCT
fcis-30390	100	21	using	use	VERB
fcis-30390	100	22	only	only	ADJ
fcis-30390	100	23	historical	historical	ADJ
fcis-30390	100	24	precipitation	precipitation	NOUN
fcis-30390	100	25	data	datum	NOUN
fcis-30390	100	26	)	)	PUNCT
fcis-30390	100	27	would	would	AUX
fcis-30390	100	28	compromise	compromise	VERB
fcis-30390	100	29	prediction	prediction	NOUN
fcis-30390	100	30	accuracy	accuracy	NOUN
fcis-30390	100	31	.	.	PUNCT
fcis-30390	101	1	therefore	therefore	ADV
fcis-30390	101	2	,	,	PUNCT
fcis-30390	101	3	we	we	PRON
fcis-30390	101	4	opted	opt	VERB
fcis-30390	101	5	for	for	ADP
fcis-30390	101	6	the	the	DET
fcis-30390	101	7	multivariate	multivariate	NOUN
fcis-30390	101	8	-	-	PUNCT
fcis-30390	101	9	to	to	ADP
fcis-30390	101	10	-	-	PUNCT
fcis-30390	101	11	univariate	univariate	ADJ
fcis-30390	101	12	forecasting	forecasting	NOUN
fcis-30390	101	13	method	method	NOUN
fcis-30390	101	14	,	,	PUNCT
fcis-30390	101	15	which	which	PRON
fcis-30390	101	16	incorporates	incorporate	VERB
fcis-30390	101	17	historical	historical	ADJ
fcis-30390	101	18	weather	weather	NOUN
fcis-30390	101	19	data	datum	NOUN
fcis-30390	101	20	from	from	ADP
fcis-30390	101	21	multiple	multiple	ADJ
fcis-30390	101	22	influential	influential	ADJ
fcis-30390	101	23	factors	factor	NOUN
fcis-30390	101	24	(	(	PUNCT
fcis-30390	101	25	such	such	ADJ
fcis-30390	101	26	as	as	ADP
fcis-30390	101	27	temperature	temperature	NOUN
fcis-30390	101	28	,	,	PUNCT
fcis-30390	101	29	humidity	humidity	NOUN
fcis-30390	101	30	,	,	PUNCT
fcis-30390	101	31	and	and	CCONJ
fcis-30390	101	32	wind	wind	NOUN
fcis-30390	101	33	speed	speed	NOUN
fcis-30390	101	34	)	)	PUNCT
fcis-30390	101	35	to	to	PART
fcis-30390	101	36	predict	predict	VERB
fcis-30390	101	37	future	future	ADJ
fcis-30390	101	38	precipitation	precipitation	NOUN
fcis-30390	101	39	levels	level	NOUN
fcis-30390	101	40	.	.	PUNCT
fcis-30390	102	1	by	by	ADP
fcis-30390	102	2	integrating	integrate	VERB
fcis-30390	102	3	information	information	NOUN
fcis-30390	102	4	from	from	ADP
fcis-30390	102	5	multiple	multiple	ADJ
fcis-30390	102	6	features	feature	NOUN
fcis-30390	102	7	,	,	PUNCT
fcis-30390	102	8	we	we	PRON
fcis-30390	102	9	can	can	AUX
fcis-30390	102	10	enhance	enhance	VERB
fcis-30390	102	11	the	the	DET
fcis-30390	102	12	accuracy	accuracy	NOUN
fcis-30390	102	13	of	of	ADP
fcis-30390	102	14	predictions	prediction	NOUN
fcis-30390	102	15	for	for	ADP
fcis-30390	102	16	the	the	DET
fcis-30390	102	17	single	single	ADJ
fcis-30390	102	18	target	target	NOUN
fcis-30390	102	19	feature	feature	NOUN
fcis-30390	102	20	(	(	PUNCT
fcis-30390	102	21	precipitation	precipitation	NOUN
fcis-30390	102	22	)	)	PUNCT
fcis-30390	102	23	.	.	PUNCT
fcis-30390	103	1	4.3	4.3	NUM
fcis-30390	103	2	.	.	PUNCT
fcis-30390	103	3	ablation	ablation	NOUN
fcis-30390	103	4	experiment	experiment	NOUN
fcis-30390	103	5	to	to	PART
fcis-30390	103	6	conduct	conduct	VERB
fcis-30390	103	7	ablation	ablation	NOUN
fcis-30390	103	8	studies	study	NOUN
fcis-30390	103	9	,	,	PUNCT
fcis-30390	103	10	we	we	PRON
fcis-30390	103	11	designed	design	VERB
fcis-30390	103	12	the	the	DET
fcis-30390	103	13	following	following	ADJ
fcis-30390	103	14	model	model	NOUN
fcis-30390	103	15	variants	variant	NOUN
fcis-30390	103	16	:	:	PUNCT
fcis-30390	104	1	•base	•base	PROPN
fcis-30390	104	2	:	:	PUNCT
fcis-30390	104	3	the	the	DET
fcis-30390	104	4	baseline	baseline	PROPN
fcis-30390	104	5	model	model	NOUN
fcis-30390	104	6	40	40	NUM
fcis-30390	104	7	•base_any	•base_any	NOUN
fcis-30390	104	8	:	:	PUNCT
fcis-30390	104	9	the	the	DET
fcis-30390	104	10	baseline	baseline	PROPN
fcis-30390	104	11	model	model	NOUN
fcis-30390	104	12	enhanced	enhance	VERB
fcis-30390	104	13	with	with	ADP
fcis-30390	104	14	the	the	DET
fcis-30390	104	15	anyvariateattention	anyvariateattention	NOUN
fcis-30390	104	16	module	module	NOUN
fcis-30390	104	17	•base_any_mtmem	•base_any_mtmem	PROPN
fcis-30390	104	18	:	:	PUNCT
fcis-30390	104	19	the	the	DET
fcis-30390	104	20	base_any	base_any	PROPN
fcis-30390	104	21	model	model	NOUN
fcis-30390	104	22	augmented	augment	VERB
fcis-30390	104	23	with	with	ADP
fcis-30390	104	24	the	the	DET
fcis-30390	104	25	mtmem	mtmem	NOUN
fcis-30390	104	26	table	table	NOUN
fcis-30390	104	27	4.the	4.the	DET
fcis-30390	104	28	results	result	NOUN
fcis-30390	104	29	of	of	ADP
fcis-30390	104	30	the	the	DET
fcis-30390	104	31	ablation	ablation	NOUN
fcis-30390	104	32	experiments	experiment	NOUN
fcis-30390	104	33	models	model	NOUN
fcis-30390	104	34	base	base	VERB
fcis-30390	104	35	base_any	base_any	PROPN
fcis-30390	104	36	base_any_mtmem	base_any_mtmem	ADJ
fcis-30390	104	37	metrics	metric	NOUN
fcis-30390	104	38	mse	mse	PROPN
fcis-30390	104	39	mae	mae	PROPN
fcis-30390	104	40	mse	mse	PROPN
fcis-30390	104	41	mae	mae	PROPN
fcis-30390	104	42	mse	mse	PROPN
fcis-30390	104	43	mae	mae	PROPN
fcis-30390	104	44	aral	aral	PROPN
fcis-30390	104	45	6h	6h	NUM
fcis-30390	104	46	0.178	0.178	NUM
fcis-30390	104	47	0.188	0.188	NUM
fcis-30390	104	48	0.175	0.175	NUM
fcis-30390	104	49	0.186	0.186	NUM
fcis-30390	104	50	0.173	0.173	NUM
fcis-30390	104	51	0.184	0.184	NUM
fcis-30390	104	52	12h	12h	NUM
fcis-30390	104	53	0.201	0.201	NUM
fcis-30390	104	54	0.209	0.209	NUM
fcis-30390	104	55	0.198	0.198	NUM
fcis-30390	104	56	0.207	0.207	NUM
fcis-30390	104	57	0.196	0.196	NUM
fcis-30390	104	58	0.205	0.205	NUM
fcis-30390	104	59	18h	18h	NOUN
fcis-30390	104	60	0.217	0.217	NUM
fcis-30390	104	61	0.222	0.222	NUM
fcis-30390	104	62	0.214	0.214	NUM
fcis-30390	104	63	0.22	0.22	NUM
fcis-30390	104	64	0.212	0.212	NUM
fcis-30390	104	65	0.218	0.218	NUM
fcis-30390	104	66	24h	24h	NOUN
fcis-30390	104	67	0.228	0.228	NUM
fcis-30390	104	68	0.23	0.23	NUM
fcis-30390	104	69	0.225	0.225	NUM
fcis-30390	104	70	0.228	0.228	NUM
fcis-30390	104	71	0.223	0.223	NUM
fcis-30390	104	72	0.226	0.226	NUM
fcis-30390	104	73	30h	30h	NOUN
fcis-30390	104	74	0.24	0.24	NUM
fcis-30390	104	75	0.24	0.24	NUM
fcis-30390	104	76	0.237	0.237	NUM
fcis-30390	104	77	0.238	0.238	NUM
fcis-30390	104	78	0.235	0.235	NUM
fcis-30390	104	79	0.236	0.236	NUM
fcis-30390	104	80	36h	36h	NOUN
fcis-30390	104	81	0.245	0.245	NUM
fcis-30390	104	82	0.244	0.244	NUM
fcis-30390	104	83	0.242	0.242	NUM
fcis-30390	104	84	0.242	0.242	NUM
fcis-30390	104	85	0.24	0.24	NUM
fcis-30390	104	86	0.24	0.24	NUM
fcis-30390	104	87	42h	42h	NOUN
fcis-30390	104	88	0.253	0.253	NUM
fcis-30390	104	89	0.249	0.249	NUM
fcis-30390	104	90	0.25	0.25	NUM
fcis-30390	104	91	0.247	0.247	NUM
fcis-30390	104	92	0.248	0.248	NUM
fcis-30390	104	93	0.245	0.245	NUM
fcis-30390	104	94	48h	48h	NOUN
fcis-30390	104	95	0.263	0.263	NUM
fcis-30390	104	96	0.257	0.257	NUM
fcis-30390	104	97	0.26	0.26	NUM
fcis-30390	104	98	0.255	0.255	NUM
fcis-30390	104	99	0.258	0.258	NUM
fcis-30390	104	100	0.253	0.253	NUM
fcis-30390	104	101	kashgar	kashgar	ADJ
fcis-30390	104	102	6h	6h	NOUN
fcis-30390	104	103	0.127	0.127	NUM
fcis-30390	104	104	0.156	0.156	NUM
fcis-30390	104	105	0.124	0.124	NUM
fcis-30390	104	106	0.153	0.153	NUM
fcis-30390	104	107	0.123	0.123	NUM
fcis-30390	104	108	0.152	0.152	NUM
fcis-30390	104	109	12h	12h	NUM
fcis-30390	104	110	0.147	0.147	NUM
fcis-30390	104	111	0.176	0.176	NUM
fcis-30390	104	112	0.144	0.144	NUM
fcis-30390	104	113	0.173	0.173	NUM
fcis-30390	104	114	0.141	0.141	NUM
fcis-30390	104	115	0.171	0.171	NUM
fcis-30390	104	116	18h	18h	NOUN
fcis-30390	104	117	0.162	0.162	NUM
fcis-30390	104	118	0.19	0.19	NUM
fcis-30390	104	119	0.159	0.159	NUM
fcis-30390	104	120	0.187	0.187	NUM
fcis-30390	104	121	0.156	0.156	NUM
fcis-30390	104	122	0.184	0.184	NUM
fcis-30390	104	123	24h	24h	VERB
fcis-30390	104	124	0.176	0.176	NUM
fcis-30390	104	125	0.201	0.201	NUM
fcis-30390	104	126	0.173	0.173	NUM
fcis-30390	104	127	0.198	0.198	NUM
fcis-30390	104	128	0.17	0.17	NUM
fcis-30390	104	129	0.195	0.195	NUM
fcis-30390	104	130	30h	30h	NOUN
fcis-30390	104	131	0.188	0.188	NUM
fcis-30390	104	132	0.21	0.21	NUM
fcis-30390	104	133	0.185	0.185	NUM
fcis-30390	104	134	0.207	0.207	NUM
fcis-30390	104	135	0.184	0.184	NUM
fcis-30390	104	136	0.205	0.205	NUM
fcis-30390	104	137	36h	36h	NOUN
fcis-30390	104	138	0.199	0.199	NUM
fcis-30390	104	139	0.218	0.218	NUM
fcis-30390	104	140	0.196	0.196	NUM
fcis-30390	104	141	0.215	0.215	NUM
fcis-30390	104	142	0.194	0.194	NUM
fcis-30390	104	143	0.214	0.214	NUM
fcis-30390	104	144	42h	42h	NOUN
fcis-30390	104	145	0.21	0.21	NUM
fcis-30390	104	146	0.226	0.226	NUM
fcis-30390	104	147	0.207	0.207	NUM
fcis-30390	104	148	0.223	0.223	NUM
fcis-30390	104	149	0.205	0.205	NUM
fcis-30390	104	150	0.222	0.222	NUM
fcis-30390	104	151	48h	48h	NOUN
fcis-30390	104	152	0.219	0.219	NUM
fcis-30390	104	153	0.232	0.232	NUM
fcis-30390	104	154	0.216	0.216	NUM
fcis-30390	104	155	0.229	0.229	NUM
fcis-30390	104	156	0.213	0.213	NUM
fcis-30390	104	157	0.225	0.225	NUM
fcis-30390	104	158	korla	korla	PROPN
fcis-30390	104	159	6h	6h	NUM
fcis-30390	104	160	0.171	0.171	NUM
fcis-30390	104	161	0.185	0.185	NUM
fcis-30390	104	162	0.168	0.168	NUM
fcis-30390	104	163	0.181	0.181	NUM
fcis-30390	104	164	0.167	0.167	NUM
fcis-30390	104	165	0.18	0.18	NUM
fcis-30390	104	166	12h	12h	NUM
fcis-30390	104	167	0.19	0.19	NUM
fcis-30390	104	168	0.206	0.206	NUM
fcis-30390	104	169	0.188	0.188	NUM
fcis-30390	104	170	0.205	0.205	NUM
fcis-30390	104	171	0.186	0.186	NUM
fcis-30390	104	172	0.201	0.201	NUM
fcis-30390	104	173	18h	18h	NOUN
fcis-30390	104	174	0.203	0.203	NUM
fcis-30390	104	175	0.218	0.218	NUM
fcis-30390	104	176	0.201	0.201	NUM
fcis-30390	104	177	0.217	0.217	NUM
fcis-30390	104	178	0.2	0.2	NUM
fcis-30390	104	179	0.213	0.213	NUM
fcis-30390	104	180	24h	24h	NOUN
fcis-30390	104	181	0.212	0.212	NUM
fcis-30390	104	182	0.226	0.226	NUM
fcis-30390	104	183	0.209	0.209	NUM
fcis-30390	104	184	0.224	0.224	NUM
fcis-30390	104	185	0.207	0.207	NUM
fcis-30390	104	186	0.221	0.221	NUM
fcis-30390	104	187	30h	30h	NOUN
fcis-30390	104	188	0.219	0.219	NUM
fcis-30390	104	189	0.233	0.233	NUM
fcis-30390	104	190	0.217	0.217	NUM
fcis-30390	104	191	0.231	0.231	NUM
fcis-30390	104	192	0.215	0.215	NUM
fcis-30390	104	193	0.228	0.228	NUM
fcis-30390	104	194	36h	36h	NOUN
fcis-30390	104	195	0.226	0.226	NUM
fcis-30390	104	196	0.24	0.24	NUM
fcis-30390	104	197	0.223	0.223	NUM
fcis-30390	104	198	0.239	0.239	NUM
fcis-30390	104	199	0.221	0.221	NUM
fcis-30390	104	200	0.235	0.235	NUM
fcis-30390	104	201	42h	42h	NOUN
fcis-30390	104	202	0.233	0.233	NUM
fcis-30390	104	203	0.245	0.245	NUM
fcis-30390	104	204	0.23	0.23	NUM
fcis-30390	104	205	0.243	0.243	NUM
fcis-30390	104	206	0.226	0.226	NUM
fcis-30390	104	207	0.24	0.24	NUM
fcis-30390	104	208	48h	48h	NOUN
fcis-30390	104	209	0.24	0.24	NUM
fcis-30390	104	210	0.252	0.252	NUM
fcis-30390	104	211	0.238	0.238	NUM
fcis-30390	104	212	0.249	0.249	NUM
fcis-30390	104	213	0.235	0.235	NUM
fcis-30390	104	214	0.247	0.247	NUM
fcis-30390	104	215	hotan	hotan	NOUN
fcis-30390	104	216	6h	6h	NOUN
fcis-30390	104	217	0.165	0.165	NUM
fcis-30390	104	218	0.19	0.19	NUM
fcis-30390	104	219	0.169	0.169	NUM
fcis-30390	104	220	0.195	0.195	NUM
fcis-30390	104	221	0.166	0.166	NUM
fcis-30390	104	222	0.193	0.193	NUM
fcis-30390	104	223	12h	12h	NOUN
fcis-30390	104	224	0.187	0.187	NUM
fcis-30390	104	225	0.211	0.211	NUM
fcis-30390	104	226	0.189	0.189	NUM
fcis-30390	104	227	0.213	0.213	NUM
fcis-30390	104	228	0.19	0.19	NUM
fcis-30390	104	229	0.215	0.215	NUM
fcis-30390	104	230	18h	18h	NOUN
fcis-30390	104	231	0.203	0.203	NUM
fcis-30390	104	232	0.225	0.225	NUM
fcis-30390	104	233	0.207	0.207	NUM
fcis-30390	104	234	0.229	0.229	NUM
fcis-30390	104	235	0.206	0.206	NUM
fcis-30390	104	236	0.227	0.227	NUM
fcis-30390	104	237	24h	24h	NOUN
fcis-30390	104	238	0.216	0.216	NUM
fcis-30390	104	239	0.235	0.235	NUM
fcis-30390	104	240	0.219	0.219	NUM
fcis-30390	104	241	0.239	0.239	NUM
fcis-30390	104	242	0.219	0.219	NUM
fcis-30390	104	243	0.238	0.238	NUM
fcis-30390	104	244	30h	30h	NOUN
fcis-30390	104	245	0.228	0.228	NUM
fcis-30390	104	246	0.245	0.245	NUM
fcis-30390	104	247	0.231	0.231	NUM
fcis-30390	104	248	0.248	0.248	NUM
fcis-30390	104	249	0.23	0.23	NUM
fcis-30390	104	250	0.246	0.246	NUM
fcis-30390	104	251	36h	36h	NOUN
fcis-30390	104	252	0.236	0.236	NUM
fcis-30390	104	253	0.251	0.251	NUM
fcis-30390	104	254	0.241	0.241	NUM
fcis-30390	104	255	0.255	0.255	NUM
fcis-30390	104	256	0.239	0.239	NUM
fcis-30390	104	257	0.254	0.254	NUM
fcis-30390	104	258	42h	42h	NOUN
fcis-30390	104	259	0.247	0.247	NUM
fcis-30390	104	260	0.259	0.259	NUM
fcis-30390	104	261	0.248	0.248	NUM
fcis-30390	104	262	0.261	0.261	NUM
fcis-30390	104	263	0.248	0.248	NUM
fcis-30390	104	264	0.26	0.26	NUM
fcis-30390	104	265	48h	48h	NOUN
fcis-30390	104	266	0.254	0.254	NUM
fcis-30390	104	267	0.264	0.264	NUM
fcis-30390	104	268	0.254	0.254	NUM
fcis-30390	104	269	0.266	0.266	NUM
fcis-30390	104	270	0.255	0.255	NUM
fcis-30390	104	271	0.267	0.267	NUM
fcis-30390	104	272	kuche	kuche	X
fcis-30390	104	273	6h	6h	NUM
fcis-30390	104	274	0.15	0.15	NUM
fcis-30390	104	275	0.17	0.17	NUM
fcis-30390	104	276	0.153	0.153	NUM
fcis-30390	104	277	0.175	0.175	NUM
fcis-30390	104	278	0.154	0.154	NUM
fcis-30390	104	279	0.176	0.176	NUM
fcis-30390	104	280	12h	12h	NUM
fcis-30390	104	281	0.166	0.166	NUM
fcis-30390	104	282	0.188	0.188	NUM
fcis-30390	104	283	0.169	0.169	NUM
fcis-30390	104	284	0.192	0.192	NUM
fcis-30390	104	285	0.17	0.17	NUM
fcis-30390	104	286	0.192	0.192	NUM
fcis-30390	104	287	18h	18h	NOUN
fcis-30390	104	288	0.178	0.178	NUM
fcis-30390	104	289	0.2	0.2	NUM
fcis-30390	104	290	0.182	0.182	NUM
fcis-30390	104	291	0.205	0.205	NUM
fcis-30390	104	292	0.181	0.181	NUM
fcis-30390	104	293	0.203	0.203	NUM
fcis-30390	104	294	24h	24h	NOUN
fcis-30390	104	295	0.188	0.188	NUM
fcis-30390	104	296	0.209	0.209	NUM
fcis-30390	104	297	0.19	0.19	NUM
fcis-30390	104	298	0.212	0.212	NUM
fcis-30390	104	299	0.191	0.191	NUM
fcis-30390	104	300	0.213	0.213	NUM
fcis-30390	104	301	30h	30h	NOUN
fcis-30390	104	302	0.196	0.196	NUM
fcis-30390	104	303	0.217	0.217	NUM
fcis-30390	104	304	0.2	0.2	NUM
fcis-30390	104	305	0.222	0.222	NUM
fcis-30390	104	306	0.2	0.2	NUM
fcis-30390	104	307	0.222	0.222	NUM
fcis-30390	104	308	36h	36h	NOUN
fcis-30390	104	309	0.201	0.201	NUM
fcis-30390	104	310	0.222	0.222	NUM
fcis-30390	104	311	0.207	0.207	NUM
fcis-30390	104	312	0.228	0.228	NUM
fcis-30390	104	313	0.204	0.204	NUM
fcis-30390	104	314	0.224	0.224	NUM
fcis-30390	104	315	42h	42h	NOUN
fcis-30390	104	316	0.207	0.207	NUM
fcis-30390	104	317	0.227	0.227	NUM
fcis-30390	104	318	0.211	0.211	NUM
fcis-30390	104	319	0.232	0.232	NUM
fcis-30390	104	320	0.211	0.211	NUM
fcis-30390	104	321	0.232	0.232	NUM
fcis-30390	104	322	48h	48h	NOUN
fcis-30390	104	323	0.212	0.212	NUM
fcis-30390	104	324	0.232	0.232	NUM
fcis-30390	104	325	0.216	0.216	NUM
fcis-30390	104	326	0.236	0.236	NUM
fcis-30390	104	327	0.215	0.215	NUM
fcis-30390	104	328	0.234	0.234	NUM
fcis-30390	104	329	table	table	NOUN
fcis-30390	104	330	5	5	NUM
fcis-30390	104	331	.	.	PUNCT
fcis-30390	104	332	final	final	ADJ
fcis-30390	104	333	mse	mse	PROPN
fcis-30390	104	334	and	and	CCONJ
fcis-30390	104	335	mae	mae	PROPN
fcis-30390	104	336	values	value	NOUN
fcis-30390	104	337	for	for	ADP
fcis-30390	104	338	each	each	DET
fcis-30390	104	339	region	region	NOUN
fcis-30390	104	340	model	model	NOUN
fcis-30390	104	341	base	base	PROPN
fcis-30390	104	342	base_any	base_any	PROPN
fcis-30390	104	343	base_any_mtmem	base_any_mtmem	ADJ
fcis-30390	104	344	metrics	metric	NOUN
fcis-30390	104	345	mse	mse	PROPN
fcis-30390	104	346	mae	mae	PROPN
fcis-30390	104	347	mse	mse	PROPN
fcis-30390	104	348	mae	mae	PROPN
fcis-30390	104	349	mse	mse	PROPN
fcis-30390	104	350	mae	mae	PROPN
fcis-30390	104	351	aral	aral	PROPN
fcis-30390	104	352	1.825	1.825	NUM
fcis-30390	104	353	1.839	1.839	NUM
fcis-30390	104	354	1.801(2.4	1.801(2.4	NUM
fcis-30390	104	355	%	%	NOUN
fcis-30390	104	356	)	)	PUNCT
fcis-30390	104	357	1.823(1.6	1.823(1.6	NUM
fcis-30390	104	358	%	%	NOUN
fcis-30390	104	359	)	)	PUNCT
fcis-30390	104	360	1.785(4	1.785(4	NUM
fcis-30390	104	361	%	%	NOUN
fcis-30390	104	362	)	)	PUNCT
fcis-30390	104	363	1.807(3.2	1.807(3.2	NUM
fcis-30390	104	364	%	%	NOUN
fcis-30390	104	365	)	)	PUNCT
fcis-30390	104	366	kashgar	kashgar	NOUN
fcis-30390	104	367	1.428	1.428	NUM
fcis-30390	104	368	1.609	1.609	NUM
fcis-30390	104	369	1.404(2.4	1.404(2.4	NUM
fcis-30390	104	370	%	%	NOUN
fcis-30390	104	371	)	)	PUNCT
fcis-30390	104	372	1.585(2.4	1.585(2.4	NUM
fcis-30390	104	373	%	%	NOUN
fcis-30390	104	374	)	)	PUNCT
fcis-30390	104	375	1.386(4.2	1.386(4.2	NUM
fcis-30390	104	376	%	%	NOUN
fcis-30390	104	377	)	)	PUNCT
fcis-30390	104	378	1.568(4.1	1.568(4.1	NUM
fcis-30390	104	379	%	%	NOUN
fcis-30390	104	380	)	)	PUNCT
fcis-30390	104	381	korla	korla	NOUN
fcis-30390	104	382	1.694	1.694	NUM
fcis-30390	104	383	1.805	1.805	NUM
fcis-30390	104	384	1.674(2	1.674(2	NUM
fcis-30390	104	385	%	%	NOUN
fcis-30390	104	386	)	)	PUNCT
fcis-30390	104	387	1.789(1.6	1.789(1.6	NUM
fcis-30390	104	388	%	%	NOUN
fcis-30390	104	389	)	)	PUNCT
fcis-30390	104	390	1.657(3.7	1.657(3.7	NUM
fcis-30390	104	391	%	%	NOUN
fcis-30390	104	392	)	)	PUNCT
fcis-30390	104	393	1.765(4	1.765(4	NUM
fcis-30390	104	394	%	%	NOUN
fcis-30390	104	395	)	)	PUNCT
fcis-30390	104	396	hotan	hotan	NOUN
fcis-30390	104	397	1.736	1.736	NUM
fcis-30390	104	398	1.88	1.88	NUM
fcis-30390	104	399	1.758(2.2	1.758(2.2	NUM
fcis-30390	104	400	%	%	NOUN
fcis-30390	104	401	)	)	PUNCT
fcis-30390	104	402	1.906(2.8	1.906(2.8	NUM
fcis-30390	104	403	%	%	NOUN
fcis-30390	104	404	)	)	PUNCT
fcis-30390	104	405	1.753(1.7	1.753(1.7	NUM
fcis-30390	104	406	%	%	NOUN
fcis-30390	104	407	)	)	PUNCT
fcis-30390	104	408	1.9(2	1.9(2	NUM
fcis-30390	104	409	%	%	NOUN
fcis-30390	104	410	)	)	PUNCT
fcis-30390	104	411	kuche	kuche	NOUN
fcis-30390	104	412	1.498	1.498	NUM
fcis-30390	104	413	1.665	1.665	NUM
fcis-30390	104	414	1.528(3	1.528(3	NUM
fcis-30390	104	415	%	%	NOUN
fcis-30390	104	416	)	)	PUNCT
fcis-30390	104	417	1.702(3.7	1.702(3.7	NUM
fcis-30390	104	418	%	%	NOUN
fcis-30390	104	419	)	)	PUNCT
fcis-30390	104	420	1.526(2.8	1.526(2.8	NUM
fcis-30390	104	421	%	%	NOUN
fcis-30390	104	422	)	)	PUNCT
fcis-30390	104	423	1.696(3.1	1.696(3.1	NUM
fcis-30390	104	424	%	%	NOUN
fcis-30390	104	425	)	)	PUNCT
fcis-30390	104	426	5	5	NUM
fcis-30390	104	427	.	.	X
fcis-30390	104	428	results	result	NOUN
fcis-30390	104	429	and	and	CCONJ
fcis-30390	104	430	analysis	analysis	NOUN
fcis-30390	104	431	5.1	5.1	NUM
fcis-30390	104	432	.	.	PUNCT
fcis-30390	105	1	overall	overall	ADJ
fcis-30390	105	2	performance	performance	NOUN
fcis-30390	105	3	in	in	ADP
fcis-30390	105	4	our	our	PRON
fcis-30390	105	5	comprehensive	comprehensive	ADJ
fcis-30390	105	6	evaluation	evaluation	NOUN
fcis-30390	105	7	of	of	ADP
fcis-30390	105	8	weather	weather	NOUN
fcis-30390	105	9	data	datum	NOUN
fcis-30390	105	10	across	across	ADP
fcis-30390	105	11	five	five	NUM
fcis-30390	105	12	representative	representative	ADJ
fcis-30390	105	13	regions	region	NOUN
fcis-30390	105	14	in	in	ADP
fcis-30390	105	15	xinjiang	xinjiang	PROPN
fcis-30390	105	16	,	,	PUNCT
fcis-30390	105	17	we	we	PRON
fcis-30390	105	18	assessed	assess	VERB
fcis-30390	105	19	various	various	ADJ
fcis-30390	105	20	model	model	NOUN
fcis-30390	105	21	variants	variant	NOUN
fcis-30390	105	22	over	over	ADP
fcis-30390	105	23	prediction	prediction	NOUN
fcis-30390	105	24	horizons	horizon	NOUN
fcis-30390	105	25	ranging	range	VERB
fcis-30390	105	26	from	from	ADP
fcis-30390	105	27	6	6	NUM
fcis-30390	105	28	to	to	ADP
fcis-30390	105	29	48	48	NUM
fcis-30390	105	30	hours	hour	NOUN
fcis-30390	105	31	,	,	PUNCT
fcis-30390	105	32	yielding	yield	VERB
fcis-30390	105	33	outstanding	outstanding	ADJ
fcis-30390	105	34	results	result	NOUN
fcis-30390	105	35	presented	present	VERB
fcis-30390	105	36	in	in	ADP
fcis-30390	105	37	table	table	NOUN
fcis-30390	105	38	5	5	NUM
fcis-30390	105	39	.	.	PUNCT
fcis-30390	106	1	the	the	DET
fcis-30390	106	2	total	total	ADJ
fcis-30390	106	3	mse	mse	PROPN
fcis-30390	106	4	and	and	CCONJ
fcis-30390	106	5	mae	mae	PROPN
fcis-30390	106	6	values	value	NOUN
fcis-30390	106	7	for	for	ADP
fcis-30390	106	8	the	the	DET
fcis-30390	106	9	three	three	NUM
fcis-30390	106	10	models	model	NOUN
fcis-30390	106	11	were	be	AUX
fcis-30390	106	12	calculated	calculate	VERB
fcis-30390	106	13	and	and	CCONJ
fcis-30390	106	14	displayed	display	VERB
fcis-30390	106	15	in	in	ADP
fcis-30390	106	16	table	table	NOUN
fcis-30390	106	17	6	6	NUM
fcis-30390	106	18	.	.	PUNCT
fcis-30390	107	1	the	the	DET
fcis-30390	107	2	overall	overall	ADJ
fcis-30390	107	3	mse	mse	NOUN
fcis-30390	107	4	evaluation	evaluation	NOUN
fcis-30390	107	5	indicates	indicate	VERB
fcis-30390	107	6	that	that	SCONJ
fcis-30390	107	7	the	the	DET
fcis-30390	107	8	anyvariateattention+mtmem	anyvariateattention+mtmem	PROPN
fcis-30390	107	9	model	model	NOUN
fcis-30390	107	10	demonstrates	demonstrate	VERB
fcis-30390	107	11	optimal	optimal	ADJ
fcis-30390	107	12	predictive	predictive	ADJ
fcis-30390	107	13	performance	performance	NOUN
fcis-30390	107	14	,	,	PUNCT
fcis-30390	107	15	significantly	significantly	ADV
fcis-30390	107	16	outperforming	outperform	VERB
fcis-30390	107	17	both	both	CCONJ
fcis-30390	107	18	the	the	DET
fcis-30390	107	19	standalone	standalone	ADJ
fcis-30390	107	20	anyvariateattention	anyvariateattention	NOUN
fcis-30390	107	21	model	model	NOUN
fcis-30390	107	22	and	and	CCONJ
fcis-30390	107	23	the	the	DET
fcis-30390	107	24	baseline	baseline	PROPN
fcis-30390	107	25	model	model	NOUN
fcis-30390	107	26	.	.	PUNCT
fcis-30390	108	1	specifically	specifically	ADV
fcis-30390	108	2	,	,	PUNCT
fcis-30390	108	3	the	the	DET
fcis-30390	108	4	performance	performance	NOUN
fcis-30390	108	5	improvement	improvement	NOUN
fcis-30390	108	6	is	be	AUX
fcis-30390	108	7	most	most	ADV
fcis-30390	108	8	pronounced	pronounce	VERB
fcis-30390	108	9	in	in	ADP
fcis-30390	108	10	the	the	DET
fcis-30390	108	11	kashgar	kashgar	NOUN
fcis-30390	108	12	region	region	NOUN
fcis-30390	108	13	,	,	PUNCT
fcis-30390	108	14	where	where	SCONJ
fcis-30390	108	15	the	the	DET
fcis-30390	108	16	anyvariateattention+mtmem	anyvariateattention+mtmem	PROPN
fcis-30390	108	17	model	model	NOUN
fcis-30390	108	18	achieves	achieve	VERB
fcis-30390	108	19	a	a	DET
fcis-30390	108	20	4.2	4.2	NUM
fcis-30390	108	21	%	%	NOUN
fcis-30390	108	22	enhancement	enhancement	NOUN
fcis-30390	108	23	compared	compare	VERB
fcis-30390	108	24	to	to	ADP
fcis-30390	108	25	the	the	DET
fcis-30390	108	26	baseline	baseline	PROPN
fcis-30390	108	27	model	model	NOUN
fcis-30390	108	28	.	.	PUNCT
fcis-30390	109	1	this	this	PRON
fcis-30390	109	2	is	be	AUX
fcis-30390	109	3	followed	follow	VERB
fcis-30390	109	4	by	by	ADP
fcis-30390	109	5	a	a	DET
fcis-30390	109	6	4.0	4.0	NUM
fcis-30390	109	7	%	%	NOUN
fcis-30390	109	8	improvement	improvement	NOUN
fcis-30390	109	9	in	in	ADP
fcis-30390	109	10	the	the	DET
fcis-30390	109	11	aral	aral	ADJ
fcis-30390	109	12	region	region	NOUN
fcis-30390	109	13	and	and	CCONJ
fcis-30390	109	14	a	a	DET
fcis-30390	109	15	3.7	3.7	NUM
fcis-30390	109	16	%	%	NOUN
fcis-30390	109	17	increase	increase	NOUN
fcis-30390	109	18	in	in	ADP
fcis-30390	109	19	korla	korla	NOUN
fcis-30390	109	20	.	.	PUNCT
fcis-30390	110	1	in	in	ADP
fcis-30390	110	2	contrast	contrast	NOUN
fcis-30390	110	3	,	,	PUNCT
fcis-30390	110	4	both	both	CCONJ
fcis-30390	110	5	hotan	hotan	NOUN
fcis-30390	110	6	and	and	CCONJ
fcis-30390	110	7	kuche	kuche	PROPN
fcis-30390	110	8	regions	region	NOUN
fcis-30390	110	9	41	41	NUM
fcis-30390	110	10	exhibited	exhibit	VERB
fcis-30390	110	11	a	a	DET
fcis-30390	110	12	downward	downward	ADJ
fcis-30390	110	13	trend	trend	NOUN
fcis-30390	110	14	,	,	PUNCT
fcis-30390	110	15	with	with	ADP
fcis-30390	110	16	decreases	decrease	NOUN
fcis-30390	110	17	ranging	range	VERB
fcis-30390	110	18	from	from	ADP
fcis-30390	110	19	1.7	1.7	NUM
fcis-30390	110	20	%	%	NOUN
fcis-30390	110	21	to	to	ADP
fcis-30390	110	22	3.7	3.7	NUM
fcis-30390	110	23	%	%	NOUN
fcis-30390	110	24	.	.	PUNCT
fcis-30390	111	1	in	in	ADP
fcis-30390	111	2	short	short	ADJ
fcis-30390	111	3	-	-	PUNCT
fcis-30390	111	4	term	term	NOUN
fcis-30390	111	5	forecasting	forecasting	NOUN
fcis-30390	111	6	tasks	task	NOUN
fcis-30390	111	7	,	,	PUNCT
fcis-30390	111	8	the	the	DET
fcis-30390	111	9	model	model	NOUN
fcis-30390	111	10	exhibits	exhibit	VERB
fcis-30390	111	11	exceptional	exceptional	ADJ
fcis-30390	111	12	performance	performance	NOUN
fcis-30390	111	13	advantages	advantage	NOUN
fcis-30390	111	14	,	,	PUNCT
fcis-30390	111	15	particularly	particularly	ADV
fcis-30390	111	16	within	within	ADP
fcis-30390	111	17	the	the	DET
fcis-30390	111	18	6	6	NUM
fcis-30390	111	19	-	-	PUNCT
fcis-30390	111	20	hour	hour	NOUN
fcis-30390	111	21	prediction	prediction	NOUN
fcis-30390	111	22	window	window	NOUN
fcis-30390	111	23	,	,	PUNCT
fcis-30390	111	24	where	where	SCONJ
fcis-30390	111	25	the	the	DET
fcis-30390	111	26	anyvariateattention	anyvariateattention	NOUN
fcis-30390	111	27	+	+	NOUN
fcis-30390	111	28	mtmem	mtmem	NOUN
fcis-30390	111	29	model	model	NOUN
fcis-30390	111	30	achieves	achieve	VERB
fcis-30390	111	31	the	the	DET
fcis-30390	111	32	best	good	ADJ
fcis-30390	111	33	performance	performance	NOUN
fcis-30390	111	34	across	across	ADP
fcis-30390	111	35	all	all	DET
fcis-30390	111	36	test	test	NOUN
fcis-30390	111	37	regions	region	NOUN
fcis-30390	111	38	.	.	PUNCT
fcis-30390	112	1	this	this	DET
fcis-30390	112	2	advantage	advantage	NOUN
fcis-30390	112	3	is	be	AUX
fcis-30390	112	4	consistently	consistently	ADV
fcis-30390	112	5	maintained	maintain	VERB
fcis-30390	112	6	in	in	ADP
fcis-30390	112	7	the	the	DET
fcis-30390	112	8	12	12	NUM
fcis-30390	112	9	to	to	PART
fcis-30390	112	10	24	24	NUM
fcis-30390	112	11	-	-	PUNCT
fcis-30390	112	12	hour	hour	NOUN
fcis-30390	112	13	prediction	prediction	NOUN
fcis-30390	112	14	range	range	NOUN
fcis-30390	112	15	.	.	PUNCT
fcis-30390	113	1	notably	notably	ADV
fcis-30390	113	2	,	,	PUNCT
fcis-30390	113	3	even	even	ADV
fcis-30390	113	4	in	in	ADP
fcis-30390	113	5	medium	medium	ADJ
fcis-30390	113	6	to	to	ADP
fcis-30390	113	7	long	long	ADJ
fcis-30390	113	8	-	-	PUNCT
fcis-30390	113	9	term	term	NOUN
fcis-30390	113	10	predictions	prediction	NOUN
fcis-30390	113	11	(	(	PUNCT
fcis-30390	113	12	30	30	NUM
fcis-30390	113	13	to	to	PART
fcis-30390	113	14	48	48	NUM
fcis-30390	113	15	hours	hour	NOUN
fcis-30390	113	16	)	)	PUNCT
fcis-30390	113	17	,	,	PUNCT
fcis-30390	113	18	although	although	SCONJ
fcis-30390	113	19	the	the	DET
fcis-30390	113	20	performance	performance	NOUN
fcis-30390	113	21	improvement	improvement	NOUN
fcis-30390	113	22	margin	margin	NOUN
fcis-30390	113	23	slightly	slightly	ADV
fcis-30390	113	24	narrows	narrow	VERB
fcis-30390	113	25	,	,	PUNCT
fcis-30390	113	26	the	the	DET
fcis-30390	113	27	model	model	NOUN
fcis-30390	113	28	retains	retain	VERB
fcis-30390	113	29	significant	significant	ADJ
fcis-30390	113	30	advantages	advantage	NOUN
fcis-30390	113	31	and	and	CCONJ
fcis-30390	113	32	stable	stable	ADJ
fcis-30390	113	33	performance	performance	NOUN
fcis-30390	113	34	.	.	PUNCT
fcis-30390	114	1	from	from	ADP
fcis-30390	114	2	a	a	DET
fcis-30390	114	3	regional	regional	ADJ
fcis-30390	114	4	analysis	analysis	NOUN
fcis-30390	114	5	perspective	perspective	NOUN
fcis-30390	114	6	,	,	PUNCT
fcis-30390	114	7	kashgar	kashgar	PROPN
fcis-30390	114	8	demonstrates	demonstrate	VERB
fcis-30390	114	9	optimal	optimal	ADJ
fcis-30390	114	10	overall	overall	ADJ
fcis-30390	114	11	performance	performance	NOUN
fcis-30390	114	12	with	with	ADP
fcis-30390	114	13	the	the	DET
fcis-30390	114	14	lowest	low	ADJ
fcis-30390	114	15	total	total	ADJ
fcis-30390	114	16	mse	mse	NOUN
fcis-30390	114	17	of	of	ADP
fcis-30390	114	18	1.386	1.386	NUM
fcis-30390	114	19	,	,	PUNCT
fcis-30390	114	20	where	where	SCONJ
fcis-30390	114	21	the	the	DET
fcis-30390	114	22	4.2	4.2	NUM
fcis-30390	114	23	%	%	NOUN
fcis-30390	114	24	improvement	improvement	NOUN
fcis-30390	114	25	brought	bring	VERB
fcis-30390	114	26	by	by	ADP
fcis-30390	114	27	mtmem	mtmem	NOUN
fcis-30390	114	28	is	be	AUX
fcis-30390	114	29	particularly	particularly	ADV
fcis-30390	114	30	significant	significant	ADJ
fcis-30390	114	31	.	.	PUNCT
fcis-30390	115	1	the	the	DET
fcis-30390	115	2	aral	aral	PROPN
fcis-30390	115	3	region	region	NOUN
fcis-30390	115	4	also	also	ADV
fcis-30390	115	5	shows	show	VERB
fcis-30390	115	6	stable	stable	ADJ
fcis-30390	115	7	performance	performance	NOUN
fcis-30390	115	8	improvements	improvement	NOUN
fcis-30390	115	9	and	and	CCONJ
fcis-30390	115	10	balanced	balanced	ADJ
fcis-30390	115	11	behavior	behavior	NOUN
fcis-30390	115	12	across	across	ADP
fcis-30390	115	13	different	different	ADJ
fcis-30390	115	14	time	time	NOUN
fcis-30390	115	15	scales	scale	NOUN
fcis-30390	115	16	.	.	PUNCT
fcis-30390	116	1	however	however	ADV
fcis-30390	116	2	,	,	PUNCT
fcis-30390	116	3	in	in	ADP
fcis-30390	116	4	contrast	contrast	NOUN
fcis-30390	116	5	,	,	PUNCT
fcis-30390	116	6	both	both	CCONJ
fcis-30390	116	7	hotan	hotan	NOUN
fcis-30390	116	8	and	and	CCONJ
fcis-30390	116	9	kuche	kuche	NOUN
fcis-30390	116	10	regions	region	NOUN
fcis-30390	116	11	experienced	experience	VERB
fcis-30390	116	12	declines	decline	NOUN
fcis-30390	116	13	in	in	ADP
fcis-30390	116	14	mse	mse	PROPN
fcis-30390	116	15	and	and	CCONJ
fcis-30390	116	16	mae	mae	PROPN
fcis-30390	116	17	.	.	PROPN
fcis-30390	117	1	further	further	ADJ
fcis-30390	117	2	analysis	analysis	NOUN
fcis-30390	117	3	revealed	reveal	VERB
fcis-30390	117	4	that	that	SCONJ
fcis-30390	117	5	the	the	DET
fcis-30390	117	6	performance	performance	NOUN
fcis-30390	117	7	degradation	degradation	NOUN
fcis-30390	117	8	is	be	AUX
fcis-30390	117	9	attributed	attribute	VERB
fcis-30390	117	10	to	to	ADP
fcis-30390	117	11	data	datum	NOUN
fcis-30390	117	12	sparsity	sparsity	NOUN
fcis-30390	117	13	due	due	ADP
fcis-30390	117	14	to	to	ADP
fcis-30390	117	15	untimely	untimely	ADJ
fcis-30390	117	16	meteorological	meteorological	ADJ
fcis-30390	117	17	station	station	NOUN
fcis-30390	117	18	records	record	NOUN
fcis-30390	117	19	and	and	CCONJ
fcis-30390	117	20	the	the	DET
fcis-30390	117	21	impact	impact	NOUN
fcis-30390	117	22	of	of	ADP
fcis-30390	117	23	climate	climate	NOUN
fcis-30390	117	24	change	change	NOUN
fcis-30390	117	25	,	,	PUNCT
fcis-30390	117	26	which	which	PRON
fcis-30390	117	27	hindered	hinder	VERB
fcis-30390	117	28	the	the	DET
fcis-30390	117	29	historical	historical	ADJ
fcis-30390	117	30	data	datum	NOUN
fcis-30390	117	31	's	's	PART
fcis-30390	117	32	ability	ability	NOUN
fcis-30390	117	33	to	to	PART
fcis-30390	117	34	effectively	effectively	ADV
fcis-30390	117	35	reflect	reflect	VERB
fcis-30390	117	36	future	future	ADJ
fcis-30390	117	37	trends	trend	NOUN
fcis-30390	117	38	.	.	PUNCT
fcis-30390	118	1	additionally	additionally	ADV
fcis-30390	118	2	,	,	PUNCT
fcis-30390	118	3	inappropriate	inappropriate	ADJ
fcis-30390	118	4	model	model	NOUN
fcis-30390	118	5	selection	selection	NOUN
fcis-30390	118	6	,	,	PUNCT
fcis-30390	118	7	data	datum	NOUN
fcis-30390	118	8	processing	processing	NOUN
fcis-30390	118	9	methods	method	NOUN
fcis-30390	118	10	,	,	PUNCT
fcis-30390	118	11	parameter	parameter	NOUN
fcis-30390	118	12	tuning	tuning	NOUN
fcis-30390	118	13	,	,	PUNCT
fcis-30390	118	14	time	time	NOUN
fcis-30390	118	15	lag	lag	NOUN
fcis-30390	118	16	effects	effect	NOUN
fcis-30390	118	17	,	,	PUNCT
fcis-30390	118	18	data	datum	NOUN
fcis-30390	118	19	noise	noise	NOUN
fcis-30390	118	20	,	,	PUNCT
fcis-30390	118	21	and	and	CCONJ
fcis-30390	118	22	limitations	limitation	NOUN
fcis-30390	118	23	of	of	ADP
fcis-30390	118	24	model	model	NOUN
fcis-30390	118	25	assumptions	assumption	NOUN
fcis-30390	118	26	further	far	ADV
fcis-30390	118	27	affected	affect	VERB
fcis-30390	118	28	the	the	DET
fcis-30390	118	29	model	model	NOUN
fcis-30390	118	30	's	's	PART
fcis-30390	118	31	accuracy	accuracy	NOUN
fcis-30390	118	32	.	.	PUNCT
fcis-30390	119	1	these	these	DET
fcis-30390	119	2	factors	factor	NOUN
fcis-30390	119	3	collectively	collectively	ADV
fcis-30390	119	4	hindered	hinder	VERB
fcis-30390	119	5	the	the	DET
fcis-30390	119	6	model	model	NOUN
fcis-30390	119	7	's	's	PART
fcis-30390	119	8	ability	ability	NOUN
fcis-30390	119	9	to	to	PART
fcis-30390	119	10	accurately	accurately	ADV
fcis-30390	119	11	capture	capture	VERB
fcis-30390	119	12	the	the	DET
fcis-30390	119	13	complexities	complexity	NOUN
fcis-30390	119	14	of	of	ADP
fcis-30390	119	15	climate	climate	NOUN
fcis-30390	119	16	change	change	NOUN
fcis-30390	119	17	in	in	ADP
fcis-30390	119	18	these	these	DET
fcis-30390	119	19	two	two	NUM
fcis-30390	119	20	regions	region	NOUN
fcis-30390	119	21	,	,	PUNCT
fcis-30390	119	22	thereby	thereby	ADV
fcis-30390	119	23	impacting	impact	VERB
fcis-30390	119	24	the	the	DET
fcis-30390	119	25	forecasting	forecasting	NOUN
fcis-30390	119	26	results	result	NOUN
fcis-30390	119	27	.	.	PUNCT
fcis-30390	120	1	5.2	5.2	X
fcis-30390	120	2	.	.	PUNCT
fcis-30390	121	1	key	key	ADJ
fcis-30390	121	2	findings	finding	NOUN
fcis-30390	121	3	across	across	ADP
fcis-30390	121	4	all	all	DET
fcis-30390	121	5	regions	region	NOUN
fcis-30390	121	6	and	and	CCONJ
fcis-30390	121	7	prediction	prediction	NOUN
fcis-30390	121	8	horizons	horizon	NOUN
fcis-30390	121	9	,	,	PUNCT
fcis-30390	121	10	the	the	DET
fcis-30390	121	11	multitemporal	multitemporal	ADJ
fcis-30390	121	12	memory	memory	NOUN
fcis-30390	121	13	enhancement	enhancement	NOUN
fcis-30390	121	14	module	module	NOUN
fcis-30390	121	15	(	(	PUNCT
fcis-30390	121	16	mtmem	mtmem	NOUN
fcis-30390	121	17	)	)	PUNCT
fcis-30390	121	18	consistently	consistently	ADV
fcis-30390	121	19	delivered	deliver	VERB
fcis-30390	121	20	performance	performance	NOUN
fcis-30390	121	21	improvements	improvement	NOUN
fcis-30390	121	22	,	,	PUNCT
fcis-30390	121	23	ranging	range	VERB
fcis-30390	121	24	from	from	ADP
fcis-30390	121	25	1.7	1.7	NUM
fcis-30390	121	26	%	%	NOUN
fcis-30390	121	27	to	to	ADP
fcis-30390	121	28	4.2	4.2	NUM
fcis-30390	121	29	%	%	NOUN
fcis-30390	121	30	.	.	PUNCT
fcis-30390	122	1	the	the	DET
fcis-30390	122	2	most	most	ADV
fcis-30390	122	3	significant	significant	ADJ
fcis-30390	122	4	enhancements	enhancement	NOUN
fcis-30390	122	5	were	be	AUX
fcis-30390	122	6	observed	observe	VERB
fcis-30390	122	7	in	in	ADP
fcis-30390	122	8	short	short	ADJ
fcis-30390	122	9	-	-	PUNCT
fcis-30390	122	10	term	term	NOUN
fcis-30390	122	11	forecasts	forecast	NOUN
fcis-30390	122	12	(	(	PUNCT
fcis-30390	122	13	6h-24h	6h-24h	NOUN
fcis-30390	122	14	)	)	PUNCT
fcis-30390	122	15	,	,	PUNCT
fcis-30390	122	16	with	with	ADP
fcis-30390	122	17	a	a	DET
fcis-30390	122	18	slower	slow	ADJ
fcis-30390	122	19	degradation	degradation	NOUN
fcis-30390	122	20	in	in	ADP
fcis-30390	122	21	performance	performance	NOUN
fcis-30390	122	22	as	as	SCONJ
fcis-30390	122	23	the	the	DET
fcis-30390	122	24	prediction	prediction	NOUN
fcis-30390	122	25	horizon	horizon	NOUN
fcis-30390	122	26	increased	increase	VERB
fcis-30390	122	27	.	.	PUNCT
fcis-30390	123	1	the	the	DET
fcis-30390	123	2	model	model	NOUN
fcis-30390	123	3	demonstrated	demonstrate	VERB
fcis-30390	123	4	good	good	ADJ
fcis-30390	123	5	adaptability	adaptability	NOUN
fcis-30390	123	6	across	across	ADP
fcis-30390	123	7	different	different	ADJ
fcis-30390	123	8	geographical	geographical	ADJ
fcis-30390	123	9	locations	location	NOUN
fcis-30390	123	10	,	,	PUNCT
fcis-30390	123	11	with	with	ADP
fcis-30390	123	12	the	the	DET
fcis-30390	123	13	degree	degree	NOUN
fcis-30390	123	14	of	of	ADP
fcis-30390	123	15	performance	performance	NOUN
fcis-30390	123	16	improvement	improvement	NOUN
fcis-30390	123	17	varying	vary	VERB
fcis-30390	123	18	by	by	ADP
fcis-30390	123	19	region	region	NOUN
fcis-30390	123	20	,	,	PUNCT
fcis-30390	123	21	indicating	indicate	VERB
fcis-30390	123	22	its	its	PRON
fcis-30390	123	23	ability	ability	NOUN
fcis-30390	123	24	to	to	PART
fcis-30390	123	25	capture	capture	VERB
fcis-30390	123	26	region	region	NOUN
fcis-30390	123	27	-	-	PUNCT
fcis-30390	123	28	specific	specific	ADJ
fcis-30390	123	29	weather	weather	NOUN
fcis-30390	123	30	patterns	pattern	NOUN
fcis-30390	123	31	.	.	PUNCT
fcis-30390	124	1	these	these	DET
fcis-30390	124	2	results	result	NOUN
fcis-30390	124	3	suggest	suggest	VERB
fcis-30390	124	4	that	that	SCONJ
fcis-30390	124	5	the	the	DET
fcis-30390	124	6	proposed	propose	VERB
fcis-30390	124	7	mtmem	mtmem	NOUN
fcis-30390	124	8	module	module	NOUN
fcis-30390	124	9	effectively	effectively	ADV
fcis-30390	124	10	enhances	enhance	VERB
fcis-30390	124	11	weather	weather	NOUN
fcis-30390	124	12	forecasting	forecasting	NOUN
fcis-30390	124	13	performance	performance	NOUN
fcis-30390	124	14	,	,	PUNCT
fcis-30390	124	15	particularly	particularly	ADV
fcis-30390	124	16	excelling	excel	VERB
fcis-30390	124	17	in	in	ADP
fcis-30390	124	18	short	short	ADJ
fcis-30390	124	19	-	-	PUNCT
fcis-30390	124	20	term	term	NOUN
fcis-30390	124	21	prediction	prediction	NOUN
fcis-30390	124	22	tasks	task	NOUN
fcis-30390	124	23	.	.	PUNCT
fcis-30390	125	1	the	the	DET
fcis-30390	125	2	model	model	NOUN
fcis-30390	125	3	exhibits	exhibit	VERB
fcis-30390	125	4	stable	stable	ADJ
fcis-30390	125	5	performance	performance	NOUN
fcis-30390	125	6	improvements	improvement	NOUN
fcis-30390	125	7	across	across	ADP
fcis-30390	125	8	various	various	ADJ
fcis-30390	125	9	geographical	geographical	ADJ
fcis-30390	125	10	locations	location	NOUN
fcis-30390	125	11	and	and	CCONJ
fcis-30390	125	12	time	time	NOUN
fcis-30390	125	13	scales	scale	NOUN
fcis-30390	125	14	,	,	PUNCT
fcis-30390	125	15	validating	validate	VERB
fcis-30390	125	16	its	its	PRON
fcis-30390	125	17	reliability	reliability	NOUN
fcis-30390	125	18	and	and	CCONJ
fcis-30390	125	19	effectiveness	effectiveness	NOUN
fcis-30390	125	20	in	in	ADP
fcis-30390	125	21	practical	practical	ADJ
fcis-30390	125	22	applications	application	NOUN
fcis-30390	125	23	.	.	PUNCT
fcis-30390	126	1	figures	figure	NOUN
fcis-30390	126	2	4	4	NUM
fcis-30390	126	3	illustrates	illustrate	VERB
fcis-30390	126	4	the	the	DET
fcis-30390	126	5	visualization	visualization	NOUN
fcis-30390	126	6	of	of	ADP
fcis-30390	126	7	experimental	experimental	ADJ
fcis-30390	126	8	results	result	NOUN
fcis-30390	126	9	for	for	ADP
fcis-30390	126	10	the	the	DET
fcis-30390	126	11	aforementioned	aforementioned	ADJ
fcis-30390	126	12	five	five	NUM
fcis-30390	126	13	regions	region	NOUN
fcis-30390	126	14	,	,	PUNCT
fcis-30390	126	15	where	where	SCONJ
fcis-30390	126	16	the	the	DET
fcis-30390	126	17	blue	blue	ADJ
fcis-30390	126	18	lines	line	NOUN
fcis-30390	126	19	represent	represent	VERB
fcis-30390	126	20	the	the	DET
fcis-30390	126	21	ground	ground	NOUN
fcis-30390	126	22	truth	truth	NOUN
fcis-30390	126	23	values	value	NOUN
fcis-30390	126	24	and	and	CCONJ
fcis-30390	126	25	the	the	DET
fcis-30390	126	26	orange	orange	PROPN
fcis-30390	126	27	lines	line	NOUN
fcis-30390	126	28	indicate	indicate	VERB
fcis-30390	126	29	the	the	DET
fcis-30390	126	30	predicted	predict	VERB
fcis-30390	126	31	values	value	NOUN
fcis-30390	126	32	.	.	PUNCT
fcis-30390	127	1	figure	figure	VERB
fcis-30390	127	2	4	4	NUM
fcis-30390	127	3	.	.	PUNCT
fcis-30390	127	4	visualization	visualization	NOUN
fcis-30390	127	5	of	of	ADP
fcis-30390	127	6	experiments	experiment	NOUN
fcis-30390	127	7	6	6	NUM
fcis-30390	127	8	.	.	PUNCT
fcis-30390	128	1	conclusion	conclusion	NOUN
fcis-30390	128	2	this	this	DET
fcis-30390	128	3	study	study	NOUN
fcis-30390	128	4	presents	present	VERB
fcis-30390	128	5	the	the	DET
fcis-30390	128	6	timemixer	timemixer	NOUN
fcis-30390	128	7	-	-	PUNCT
fcis-30390	128	8	me	i	PRON
fcis-30390	128	9	model	model	NOUN
fcis-30390	128	10	,	,	PUNCT
fcis-30390	128	11	a	a	DET
fcis-30390	128	12	dualmemory	dualmemory	ADJ
fcis-30390	128	13	enhancement	enhancement	NOUN
fcis-30390	128	14	framework	framework	NOUN
fcis-30390	128	15	based	base	VERB
fcis-30390	128	16	on	on	ADP
fcis-30390	128	17	the	the	DET
fcis-30390	128	18	timemixer	timemixer	NOUN
fcis-30390	128	19	architecture	architecture	NOUN
fcis-30390	128	20	,	,	PUNCT
fcis-30390	128	21	which	which	PRON
fcis-30390	128	22	has	have	AUX
fcis-30390	128	23	achieved	achieve	VERB
fcis-30390	128	24	significant	significant	ADJ
fcis-30390	128	25	breakthroughs	breakthrough	NOUN
fcis-30390	128	26	in	in	ADP
fcis-30390	128	27	short	short	ADJ
fcis-30390	128	28	-	-	PUNCT
fcis-30390	128	29	term	term	NOUN
fcis-30390	128	30	weather	weather	NOUN
fcis-30390	128	31	forecasting	forecasting	NOUN
fcis-30390	128	32	tasks	task	NOUN
fcis-30390	128	33	in	in	ADP
fcis-30390	128	34	the	the	DET
fcis-30390	128	35	southern	southern	ADJ
fcis-30390	128	36	xinjiang	xinjiang	PROPN
fcis-30390	128	37	region	region	PROPN
fcis-30390	128	38	.	.	PUNCT
fcis-30390	129	1	in	in	ADP
fcis-30390	129	2	prediction	prediction	NOUN
fcis-30390	129	3	tasks	task	NOUN
fcis-30390	129	4	using	use	VERB
fcis-30390	129	5	datasets	dataset	NOUN
fcis-30390	129	6	from	from	ADP
fcis-30390	129	7	aral	aral	ADJ
fcis-30390	129	8	,	,	PUNCT
fcis-30390	129	9	kashgar	kashgar	NOUN
fcis-30390	129	10	,	,	PUNCT
fcis-30390	129	11	and	and	CCONJ
fcis-30390	129	12	korla	korla	PROPN
fcis-30390	129	13	,	,	PUNCT
fcis-30390	129	14	the	the	DET
fcis-30390	129	15	model	model	NOUN
fcis-30390	129	16	demonstrated	demonstrate	VERB
fcis-30390	129	17	substantial	substantial	ADJ
fcis-30390	129	18	performance	performance	NOUN
fcis-30390	129	19	improvements	improvement	NOUN
fcis-30390	129	20	compared	compare	VERB
fcis-30390	129	21	to	to	ADP
fcis-30390	129	22	the	the	DET
fcis-30390	129	23	baseline	baseline	NOUN
fcis-30390	129	24	timemixer	timemixer	NOUN
fcis-30390	129	25	,	,	PUNCT
fcis-30390	129	26	with	with	ADP
fcis-30390	129	27	mse	mse	NOUN
fcis-30390	129	28	metrics	metric	NOUN
fcis-30390	129	29	reduced	reduce	VERB
fcis-30390	129	30	by	by	ADP
fcis-30390	129	31	4	4	NUM
fcis-30390	129	32	%	%	NOUN
fcis-30390	129	33	,	,	PUNCT
fcis-30390	129	34	4.2	4.2	NUM
fcis-30390	129	35	%	%	NOUN
fcis-30390	129	36	,	,	PUNCT
fcis-30390	129	37	and	and	CCONJ
fcis-30390	129	38	3.7	3.7	NUM
fcis-30390	129	39	%	%	NOUN
fcis-30390	129	40	,	,	PUNCT
fcis-30390	129	41	respectively	respectively	ADV
fcis-30390	129	42	,	,	PUNCT
fcis-30390	129	43	and	and	CCONJ
fcis-30390	129	44	mae	mae	PROPN
fcis-30390	129	45	metrics	metric	NOUN
fcis-30390	129	46	showing	show	VERB
fcis-30390	129	47	comprehensive	comprehensive	ADJ
fcis-30390	129	48	enhancements	enhancement	NOUN
fcis-30390	129	49	ranging	range	VERB
fcis-30390	129	50	from	from	ADP
fcis-30390	129	51	3.7	3.7	NUM
fcis-30390	129	52	%	%	NOUN
fcis-30390	129	53	to	to	ADP
fcis-30390	129	54	4.2	4.2	NUM
fcis-30390	129	55	%	%	NOUN
fcis-30390	129	56	.	.	PUNCT
fcis-30390	130	1	these	these	DET
fcis-30390	130	2	achievements	achievement	NOUN
fcis-30390	130	3	are	be	AUX
fcis-30390	130	4	attributed	attribute	VERB
fcis-30390	130	5	to	to	ADP
fcis-30390	130	6	theoretical	theoretical	ADJ
fcis-30390	130	7	innovations	innovation	NOUN
fcis-30390	130	8	within	within	ADP
fcis-30390	130	9	the	the	DET
fcis-30390	130	10	model	model	NOUN
fcis-30390	130	11	,	,	PUNCT
fcis-30390	130	12	particularly	particularly	ADV
fcis-30390	130	13	the	the	DET
fcis-30390	130	14	clever	clever	ADJ
fcis-30390	130	15	integration	integration	NOUN
fcis-30390	130	16	of	of	ADP
fcis-30390	130	17	the	the	DET
fcis-30390	130	18	dual	dual	ADJ
fcis-30390	130	19	-	-	PUNCT
fcis-30390	130	20	memory	memory	NOUN
fcis-30390	130	21	mechanism	mechanism	NOUN
fcis-30390	130	22	from	from	ADP
fcis-30390	130	23	cognitive	cognitive	ADJ
fcis-30390	130	24	science	science	NOUN
fcis-30390	130	25	into	into	ADP
fcis-30390	130	26	the	the	DET
fcis-30390	130	27	time	time	NOUN
fcis-30390	130	28	series	series	PROPN
fcis-30390	130	29	forecasting	forecasting	NOUN
fcis-30390	130	30	domain	domain	NOUN
fcis-30390	130	31	.	.	PUNCT
fcis-30390	131	1	by	by	ADP
fcis-30390	131	2	effectively	effectively	ADV
fcis-30390	131	3	coordinating	coordinate	VERB
fcis-30390	131	4	short	short	ADJ
fcis-30390	131	5	-	-	PUNCT
fcis-30390	131	6	term	term	NOUN
fcis-30390	131	7	working	working	NOUN
fcis-30390	131	8	memory	memory	NOUN
fcis-30390	131	9	and	and	CCONJ
fcis-30390	131	10	long	long	ADJ
fcis-30390	131	11	-	-	PUNCT
fcis-30390	131	12	term	term	NOUN
fcis-30390	131	13	reference	reference	NOUN
fcis-30390	131	14	memory	memory	NOUN
fcis-30390	131	15	,	,	PUNCT
fcis-30390	131	16	the	the	DET
fcis-30390	131	17	model	model	NOUN
fcis-30390	131	18	accurately	accurately	ADV
fcis-30390	131	19	captures	capture	VERB
fcis-30390	131	20	the	the	DET
fcis-30390	131	21	dynamic	dynamic	ADJ
fcis-30390	131	22	characteristics	characteristic	NOUN
fcis-30390	131	23	of	of	ADP
fcis-30390	131	24	weather	weather	NOUN
fcis-30390	131	25	systems	system	NOUN
fcis-30390	131	26	.	.	PUNCT
fcis-30390	132	1	additionally	additionally	ADV
fcis-30390	132	2	,	,	PUNCT
fcis-30390	132	3	the	the	DET
fcis-30390	132	4	designed	design	VERB
fcis-30390	132	5	adaptive	adaptive	ADJ
fcis-30390	132	6	feature	feature	NOUN
fcis-30390	132	7	fusion	fusion	NOUN
fcis-30390	132	8	mechanism	mechanism	NOUN
fcis-30390	132	9	allows	allow	VERB
fcis-30390	132	10	for	for	ADP
fcis-30390	132	11	flexible	flexible	ADJ
fcis-30390	132	12	adjustment	adjustment	NOUN
fcis-30390	132	13	of	of	ADP
fcis-30390	132	14	feature	feature	NOUN
fcis-30390	132	15	weights	weight	NOUN
fcis-30390	132	16	based	base	VERB
fcis-30390	132	17	on	on	ADP
fcis-30390	132	18	changes	change	NOUN
fcis-30390	132	19	in	in	ADP
fcis-30390	132	20	time	time	NOUN
fcis-30390	132	21	scales	scale	NOUN
fcis-30390	132	22	.	.	PUNCT
fcis-30390	133	1	however	however	ADV
fcis-30390	133	2	,	,	PUNCT
fcis-30390	133	3	we	we	PRON
fcis-30390	133	4	also	also	ADV
fcis-30390	133	5	recognize	recognize	VERB
fcis-30390	133	6	that	that	SCONJ
fcis-30390	133	7	there	there	PRON
fcis-30390	133	8	is	be	VERB
fcis-30390	133	9	still	still	ADV
fcis-30390	133	10	room	room	NOUN
fcis-30390	133	11	for	for	ADP
fcis-30390	133	12	improvement	improvement	NOUN
fcis-30390	133	13	in	in	ADP
fcis-30390	133	14	the	the	DET
fcis-30390	133	15	model	model	NOUN
fcis-30390	133	16	's	's	PART
fcis-30390	133	17	42	42	NUM
fcis-30390	133	18	predictive	predictive	ADJ
fcis-30390	133	19	performance	performance	NOUN
fcis-30390	133	20	under	under	ADP
fcis-30390	133	21	extreme	extreme	ADJ
fcis-30390	133	22	weather	weather	NOUN
fcis-30390	133	23	conditions	condition	NOUN
fcis-30390	133	24	.	.	PUNCT
fcis-30390	134	1	future	future	ADJ
fcis-30390	134	2	research	research	NOUN
fcis-30390	134	3	directions	direction	NOUN
fcis-30390	134	4	may	may	AUX
fcis-30390	134	5	focus	focus	VERB
fcis-30390	134	6	on	on	ADP
fcis-30390	134	7	incorporating	incorporate	VERB
fcis-30390	134	8	external	external	ADJ
fcis-30390	134	9	factors	factor	NOUN
fcis-30390	134	10	such	such	ADJ
fcis-30390	134	11	as	as	ADP
fcis-30390	134	12	topography	topography	NOUN
fcis-30390	134	13	and	and	CCONJ
fcis-30390	134	14	vegetation	vegetation	NOUN
fcis-30390	134	15	to	to	PART
fcis-30390	134	16	further	far	ADV
fcis-30390	134	17	enhance	enhance	VERB
fcis-30390	134	18	prediction	prediction	NOUN
fcis-30390	134	19	accuracy	accuracy	NOUN
fcis-30390	134	20	,	,	PUNCT
fcis-30390	134	21	as	as	ADV
fcis-30390	134	22	well	well	ADV
fcis-30390	134	23	as	as	ADP
fcis-30390	134	24	exploring	explore	VERB
fcis-30390	134	25	the	the	DET
fcis-30390	134	26	extension	extension	NOUN
fcis-30390	134	27	of	of	ADP
fcis-30390	134	28	the	the	DET
fcis-30390	134	29	model	model	NOUN
fcis-30390	134	30	's	's	PART
fcis-30390	134	31	capabilities	capability	NOUN
fcis-30390	134	32	to	to	ADP
fcis-30390	134	33	longer	long	ADJ
fcis-30390	134	34	time	time	NOUN
fcis-30390	134	35	-	-	PUNCT
fcis-30390	134	36	scale	scale	NOUN
fcis-30390	134	37	weather	weather	NOUN
fcis-30390	134	38	forecasting	forecasting	NOUN
fcis-30390	134	39	tasks	task	NOUN
fcis-30390	134	40	,	,	PUNCT
fcis-30390	134	41	aiming	aim	VERB
fcis-30390	134	42	for	for	ADP
fcis-30390	134	43	a	a	DET
fcis-30390	134	44	more	more	ADV
fcis-30390	134	45	comprehensive	comprehensive	ADJ
fcis-30390	134	46	weather	weather	NOUN
fcis-30390	134	47	prediction	prediction	NOUN
fcis-30390	134	48	capability	capability	NOUN
fcis-30390	134	49	.	.	PUNCT
fcis-30390	135	1	acknowledgments	acknowledgment	NOUN
fcis-30390	135	2	thanks	thank	NOUN
fcis-30390	135	3	to	to	ADP
fcis-30390	135	4	xinjiang	xinjiang	PROPN
fcis-30390	135	5	university	university	PROPN
fcis-30390	135	6	of	of	ADP
fcis-30390	135	7	science	science	NOUN
fcis-30390	135	8	and	and	CCONJ
fcis-30390	135	9	technology	technology	NOUN
fcis-30390	135	10	for	for	ADP
fcis-30390	135	11	its	its	PRON
fcis-30390	135	12	support	support	NOUN
fcis-30390	135	13	of	of	ADP
fcis-30390	135	14	the	the	DET
fcis-30390	135	15	undergraduate	undergraduate	NOUN
fcis-30390	135	16	innovation	innovation	NOUN
fcis-30390	135	17	and	and	CCONJ
fcis-30390	135	18	entrepreneurship	entrepreneurship	NOUN
fcis-30390	135	19	program	program	NOUN
fcis-30390	135	20	fund	fund	NOUN
fcis-30390	135	21	.	.	PUNCT
fcis-30390	136	1	project	project	NOUN
fcis-30390	136	2	number	number	NOUN
fcis-30390	136	3	:	:	PUNCT
fcis-30390	136	4	x202513	x202513	PROPN
fcis-30390	136	5	561108	561108	NUM
fcis-30390	136	6	.	.	PUNCT
fcis-30390	137	1	references	reference	NOUN
fcis-30390	137	2	[	[	X
fcis-30390	137	3	1	1	NUM
fcis-30390	137	4	]	]	X
fcis-30390	137	5	box	box	NOUN
fcis-30390	137	6	,	,	PUNCT
fcis-30390	137	7	g.	g.	PROPN
fcis-30390	137	8	e.	e.	PROPN
fcis-30390	137	9	p.	p.	PROPN
fcis-30390	137	10	,	,	PUNCT
fcis-30390	137	11	gwilym	gwilym	PROPN
fcis-30390	137	12	m.	m.	PROPN
fcis-30390	137	13	jenkins	jenkins	PROPN
fcis-30390	137	14	,	,	PUNCT
fcis-30390	137	15	gregory	gregory	PROPN
fcis-30390	137	16	c.	c.	PROPN
fcis-30390	137	17	reinsel	reinsel	PROPN
fcis-30390	137	18	and	and	CCONJ
fcis-30390	137	19	greta	greta	PROPN
fcis-30390	137	20	m.	m.	PROPN
fcis-30390	137	21	ljung	ljung	PROPN
fcis-30390	137	22	.	.	PUNCT
fcis-30390	138	1	“	"	PUNCT
fcis-30390	138	2	time	time	NOUN
fcis-30390	138	3	series	series	PROPN
fcis-30390	138	4	analysis	analysis	NOUN
fcis-30390	138	5	:	:	PUNCT
fcis-30390	138	6	forecasting	forecasting	NOUN
fcis-30390	138	7	and	and	CCONJ
fcis-30390	138	8	control	control	NOUN
fcis-30390	138	9	.	.	PUNCT
fcis-30390	138	10	”	"	PUNCT
fcis-30390	139	1	the	the	DET
fcis-30390	139	2	statistician	statistician	ADJ
fcis-30390	139	3	27	27	NUM
fcis-30390	139	4	(	(	PUNCT
fcis-30390	139	5	1978	1978	NUM
fcis-30390	139	6	):	):	PUNCT
fcis-30390	139	7	265	265	NUM
fcis-30390	139	8	-	-	SYM
fcis-30390	139	9	265	265	NUM
fcis-30390	139	10	.	.	PUNCT
fcis-30390	140	1	[	[	X
fcis-30390	140	2	2	2	NUM
fcis-30390	140	3	]	]	PUNCT
fcis-30390	140	4	hyndman	hyndman	NOUN
fcis-30390	140	5	,	,	PUNCT
fcis-30390	140	6	rob	rob	PROPN
fcis-30390	140	7	j	j	PROPN
fcis-30390	140	8	and	and	CCONJ
fcis-30390	140	9	george	george	PROPN
fcis-30390	140	10	athanasopoulos	athanasopoulo	NOUN
fcis-30390	140	11	.	.	PUNCT
fcis-30390	141	1	“	"	PUNCT
fcis-30390	141	2	forecasting	forecasting	NOUN
fcis-30390	141	3	:	:	PUNCT
fcis-30390	141	4	principles	principle	NOUN
fcis-30390	141	5	and	and	CCONJ
fcis-30390	141	6	practice	practice	NOUN
fcis-30390	141	7	.	.	PUNCT
fcis-30390	141	8	”	"	PUNCT
fcis-30390	142	1	(	(	PUNCT
fcis-30390	142	2	2013	2013	NUM
fcis-30390	142	3	)	)	PUNCT
fcis-30390	142	4	.	.	PUNCT
fcis-30390	143	1	[	[	X
fcis-30390	143	2	3	3	NUM
fcis-30390	143	3	]	]	X
fcis-30390	143	4	hochreiter	hochreiter	PROPN
fcis-30390	143	5	,	,	PUNCT
fcis-30390	143	6	sepp	sepp	PROPN
fcis-30390	143	7	and	and	CCONJ
fcis-30390	143	8	jürgen	jürgen	PROPN
fcis-30390	143	9	schmidhuber	schmidhuber	PROPN
fcis-30390	143	10	.	.	PUNCT
fcis-30390	144	1	“	"	PUNCT
fcis-30390	144	2	long	long	ADJ
fcis-30390	144	3	short	short	ADJ
fcis-30390	144	4	-	-	PUNCT
fcis-30390	144	5	term	term	NOUN
fcis-30390	144	6	memory	memory	NOUN
fcis-30390	144	7	.	.	PUNCT
fcis-30390	144	8	”	"	PUNCT
fcis-30390	145	1	neural	neural	ADJ
fcis-30390	145	2	computation	computation	NOUN
fcis-30390	145	3	9	9	NUM
fcis-30390	145	4	(	(	PUNCT
fcis-30390	145	5	1997	1997	NUM
fcis-30390	145	6	):	):	PUNCT
fcis-30390	145	7	1735	1735	NUM
fcis-30390	145	8	-	-	SYM
fcis-30390	145	9	1780	1780	NUM
fcis-30390	145	10	.	.	PUNCT
fcis-30390	146	1	[	[	X
fcis-30390	146	2	4	4	NUM
fcis-30390	146	3	]	]	X
fcis-30390	146	4	vaswani	vaswani	NOUN
fcis-30390	146	5	,	,	PUNCT
fcis-30390	146	6	ashish	ashish	PROPN
fcis-30390	146	7	,	,	PUNCT
fcis-30390	146	8	noam	noam	PROPN
fcis-30390	146	9	m.	m.	PROPN
fcis-30390	146	10	shazeer	shazeer	PROPN
fcis-30390	146	11	,	,	PUNCT
fcis-30390	146	12	niki	niki	PROPN
fcis-30390	146	13	parmar	parmar	PROPN
fcis-30390	146	14	,	,	PUNCT
fcis-30390	146	15	jakob	jakob	PROPN
fcis-30390	146	16	uszkoreit	uszkoreit	PROPN
fcis-30390	146	17	,	,	PUNCT
fcis-30390	146	18	llion	llion	NOUN
fcis-30390	146	19	jones	jones	PROPN
fcis-30390	146	20	,	,	PUNCT
fcis-30390	146	21	aidan	aidan	PROPN
fcis-30390	146	22	n.	n.	PROPN
fcis-30390	146	23	gomez	gomez	PROPN
fcis-30390	146	24	,	,	PUNCT
fcis-30390	146	25	lukasz	lukasz	PROPN
fcis-30390	146	26	kaiser	kaiser	PROPN
fcis-30390	146	27	and	and	CCONJ
fcis-30390	146	28	illia	illia	PROPN
fcis-30390	146	29	polosukhin	polosukhin	PROPN
fcis-30390	146	30	.	.	PUNCT
fcis-30390	147	1	“	"	PUNCT
fcis-30390	147	2	attention	attention	NOUN
fcis-30390	147	3	is	be	AUX
fcis-30390	147	4	all	all	PRON
fcis-30390	147	5	you	you	PRON
fcis-30390	147	6	need	need	VERB
fcis-30390	147	7	.	.	PUNCT
fcis-30390	147	8	”	"	PUNCT
fcis-30390	148	1	neural	neural	ADJ
fcis-30390	148	2	information	information	NOUN
fcis-30390	148	3	processing	processing	NOUN
fcis-30390	148	4	systems	system	NOUN
fcis-30390	148	5	(	(	PUNCT
fcis-30390	148	6	2017	2017	NUM
fcis-30390	148	7	)	)	PUNCT
fcis-30390	148	8	.	.	PUNCT
fcis-30390	149	1	[	[	X
fcis-30390	149	2	5	5	NUM
fcis-30390	149	3	]	]	X
fcis-30390	149	4	zhou	zhou	PROPN
fcis-30390	149	5	,	,	PUNCT
fcis-30390	149	6	tian	tian	PROPN
fcis-30390	149	7	,	,	PUNCT
fcis-30390	149	8	ziqing	ziqing	PROPN
fcis-30390	149	9	ma	ma	PROPN
fcis-30390	149	10	,	,	PUNCT
fcis-30390	149	11	qingsong	qingsong	PROPN
fcis-30390	149	12	wen	wen	PROPN
fcis-30390	149	13	,	,	PUNCT
fcis-30390	149	14	xue	xue	PROPN
fcis-30390	149	15	wang	wang	PROPN
fcis-30390	149	16	,	,	PUNCT
fcis-30390	149	17	liang	liang	PROPN
fcis-30390	149	18	sun	sun	PROPN
fcis-30390	149	19	and	and	CCONJ
fcis-30390	149	20	rong	rong	PROPN
fcis-30390	149	21	jin	jin	PROPN
fcis-30390	149	22	.	.	PUNCT
fcis-30390	150	1	“	"	PUNCT
fcis-30390	150	2	fedformer	fedformer	X
fcis-30390	150	3	:	:	PUNCT
fcis-30390	150	4	frequency	frequency	NOUN
fcis-30390	150	5	enhanced	enhance	VERB
fcis-30390	150	6	decomposed	decomposed	ADJ
fcis-30390	150	7	transformer	transformer	NOUN
fcis-30390	150	8	for	for	ADP
fcis-30390	150	9	long	long	ADJ
fcis-30390	150	10	-	-	PUNCT
fcis-30390	150	11	term	term	NOUN
fcis-30390	150	12	series	series	NOUN
fcis-30390	150	13	forecasting	forecasting	NOUN
fcis-30390	150	14	.	.	PUNCT
fcis-30390	150	15	”	"	PUNCT
fcis-30390	151	1	international	international	ADJ
fcis-30390	151	2	conference	conference	NOUN
fcis-30390	151	3	on	on	ADP
fcis-30390	151	4	machine	machine	NOUN
fcis-30390	151	5	learning	learning	NOUN
fcis-30390	151	6	(	(	PUNCT
fcis-30390	151	7	2022	2022	NUM
fcis-30390	151	8	)	)	PUNCT
fcis-30390	151	9	.	.	PUNCT
fcis-30390	152	1	[	[	X
fcis-30390	152	2	6	6	NUM
fcis-30390	152	3	]	]	X
fcis-30390	152	4	zeng	zeng	X
fcis-30390	152	5	,	,	PUNCT
fcis-30390	152	6	ailing	ailing	NOUN
fcis-30390	152	7	,	,	PUNCT
fcis-30390	152	8	mu	mu	PROPN
fcis-30390	152	9	-	-	PUNCT
fcis-30390	152	10	hwa	hwa	PROPN
fcis-30390	152	11	chen	chen	PROPN
fcis-30390	152	12	,	,	PUNCT
fcis-30390	152	13	l.	l.	PROPN
fcis-30390	152	14	zhang	zhang	PROPN
fcis-30390	152	15	and	and	CCONJ
fcis-30390	152	16	qiang	qiang	PROPN
fcis-30390	152	17	xu	xu	PROPN
fcis-30390	152	18	.	.	PUNCT
fcis-30390	153	1	“	"	PUNCT
fcis-30390	153	2	are	be	AUX
fcis-30390	153	3	transformers	transformer	NOUN
fcis-30390	153	4	effective	effective	ADJ
fcis-30390	153	5	for	for	ADP
fcis-30390	153	6	time	time	NOUN
fcis-30390	153	7	series	series	PROPN
fcis-30390	153	8	forecasting	forecasting	NOUN
fcis-30390	153	9	?	?	PUNCT
fcis-30390	153	10	”	"	PUNCT
fcis-30390	154	1	aaai	aaai	PROPN
fcis-30390	154	2	conference	conference	NOUN
fcis-30390	154	3	on	on	ADP
fcis-30390	154	4	artificial	artificial	ADJ
fcis-30390	154	5	intelligence	intelligence	NOUN
fcis-30390	154	6	(	(	PUNCT
fcis-30390	154	7	2022	2022	NUM
fcis-30390	154	8	)	)	PUNCT
fcis-30390	154	9	.	.	PUNCT
fcis-30390	155	1	[	[	X
fcis-30390	155	2	7	7	X
fcis-30390	155	3	]	]	X
fcis-30390	155	4	liu	liu	PROPN
fcis-30390	155	5	,	,	PUNCT
fcis-30390	155	6	yong	yong	PROPN
fcis-30390	155	7	,	,	PUNCT
fcis-30390	155	8	haixu	haixu	PROPN
fcis-30390	155	9	wu	wu	PROPN
fcis-30390	155	10	,	,	PUNCT
fcis-30390	155	11	jianmin	jianmin	PROPN
fcis-30390	155	12	wang	wang	PROPN
fcis-30390	155	13	and	and	CCONJ
fcis-30390	155	14	mingsheng	mingsheng	PROPN
fcis-30390	155	15	long	long	ADV
fcis-30390	155	16	.	.	PUNCT
fcis-30390	156	1	“	"	PUNCT
fcis-30390	156	2	non	non	ADJ
fcis-30390	156	3	-	-	ADJ
fcis-30390	156	4	stationary	stationary	ADJ
fcis-30390	156	5	transformers	transformer	NOUN
fcis-30390	156	6	:	:	PUNCT
fcis-30390	156	7	exploring	explore	VERB
fcis-30390	156	8	the	the	DET
fcis-30390	156	9	stationarity	stationarity	NOUN
fcis-30390	156	10	in	in	ADP
fcis-30390	156	11	time	time	NOUN
fcis-30390	156	12	series	series	PROPN
fcis-30390	156	13	forecasting	forecasting	PROPN
fcis-30390	156	14	.	.	PUNCT
fcis-30390	156	15	”	"	PUNCT
fcis-30390	156	16	neural	neural	ADJ
fcis-30390	156	17	information	information	NOUN
fcis-30390	156	18	processing	processing	NOUN
fcis-30390	156	19	systems	system	NOUN
fcis-30390	156	20	(	(	PUNCT
fcis-30390	156	21	2022	2022	NUM
fcis-30390	156	22	)	)	PUNCT
fcis-30390	156	23	.	.	PUNCT
fcis-30390	157	1	[	[	X
fcis-30390	157	2	8	8	NUM
fcis-30390	157	3	]	]	SYM
fcis-30390	157	4	wu	wu	PROPN
fcis-30390	157	5	,	,	PUNCT
fcis-30390	157	6	haixu	haixu	ADJ
fcis-30390	157	7	,	,	PUNCT
fcis-30390	157	8	teng	teng	PROPN
fcis-30390	157	9	hu	hu	PROPN
fcis-30390	157	10	,	,	PUNCT
fcis-30390	157	11	yong	yong	PROPN
fcis-30390	157	12	liu	liu	PROPN
fcis-30390	157	13	,	,	PUNCT
fcis-30390	157	14	hang	hang	PROPN
fcis-30390	157	15	zhou	zhou	PROPN
fcis-30390	157	16	,	,	PUNCT
fcis-30390	157	17	jianmin	jianmin	PROPN
fcis-30390	157	18	wang	wang	PROPN
fcis-30390	157	19	and	and	CCONJ
fcis-30390	157	20	mingsheng	mingsheng	PROPN
fcis-30390	157	21	long	long	ADV
fcis-30390	157	22	.	.	PUNCT
fcis-30390	158	1	“	"	PUNCT
fcis-30390	158	2	timesnet	timesnet	NOUN
fcis-30390	158	3	:	:	PUNCT
fcis-30390	158	4	temporal	temporal	ADJ
fcis-30390	158	5	2d	2d	NOUN
fcis-30390	158	6	-	-	PUNCT
fcis-30390	158	7	variation	variation	NOUN
fcis-30390	158	8	modeling	modeling	NOUN
fcis-30390	158	9	for	for	ADP
fcis-30390	158	10	general	general	ADJ
fcis-30390	158	11	time	time	NOUN
fcis-30390	158	12	series	series	PROPN
fcis-30390	158	13	analysis	analysis	NOUN
fcis-30390	158	14	.	.	PUNCT
fcis-30390	158	15	”	"	PUNCT
fcis-30390	159	1	arxiv	arxiv	PROPN
fcis-30390	159	2	abs/2210.02186	abs/2210.02186	PROPN
fcis-30390	159	3	(	(	PUNCT
fcis-30390	159	4	2022	2022	NUM
fcis-30390	159	5	):	):	PUNCT
fcis-30390	159	6	n.	n.	PROPN
fcis-30390	159	7	pag	pag	PROPN
fcis-30390	159	8	.	.	PUNCT
fcis-30390	160	1	[	[	X
fcis-30390	160	2	9	9	NUM
fcis-30390	160	3	]	]	SYM
fcis-30390	160	4	shabani	shabani	PROPN
fcis-30390	160	5	,	,	PUNCT
fcis-30390	160	6	amin	amin	PROPN
fcis-30390	160	7	,	,	PUNCT
fcis-30390	160	8	amir	amir	PROPN
fcis-30390	160	9	hossein	hossein	PROPN
fcis-30390	160	10	sayyad	sayyad	PROPN
fcis-30390	160	11	abdi	abdi	PROPN
fcis-30390	160	12	,	,	PUNCT
fcis-30390	160	13	li	li	PROPN
fcis-30390	160	14	meng	meng	PROPN
fcis-30390	160	15	and	and	CCONJ
fcis-30390	160	16	tristan	tristan	PROPN
fcis-30390	160	17	sylvain	sylvain	PROPN
fcis-30390	160	18	.	.	PUNCT
fcis-30390	161	1	“	"	PUNCT
fcis-30390	161	2	scaleformer	scaleformer	ADJ
fcis-30390	161	3	:	:	PUNCT
fcis-30390	161	4	iterative	iterative	ADJ
fcis-30390	161	5	multi	multi	ADJ
fcis-30390	161	6	-	-	ADJ
fcis-30390	161	7	scale	scale	ADJ
fcis-30390	161	8	refining	refining	NOUN
fcis-30390	161	9	transformers	transformer	NOUN
fcis-30390	161	10	for	for	ADP
fcis-30390	161	11	time	time	NOUN
fcis-30390	161	12	series	series	PROPN
fcis-30390	161	13	forecasting	forecasting	PROPN
fcis-30390	161	14	.	.	PUNCT
fcis-30390	161	15	”	"	PUNCT
fcis-30390	162	1	arxiv	arxiv	PROPN
fcis-30390	162	2	abs/2206.04038	abs/2206.04038	NOUN
fcis-30390	162	3	(	(	PUNCT
fcis-30390	162	4	2022	2022	NUM
fcis-30390	162	5	):	):	PUNCT
fcis-30390	162	6	n.	n.	PROPN
fcis-30390	162	7	pag	pag	PROPN
fcis-30390	162	8	.	.	PUNCT
fcis-30390	163	1	[	[	X
fcis-30390	163	2	10	10	NUM
fcis-30390	163	3	]	]	X
fcis-30390	163	4	lin	lin	PROPN
fcis-30390	163	5	,	,	PUNCT
fcis-30390	163	6	lequan	lequan	PROPN
fcis-30390	163	7	,	,	PUNCT
fcis-30390	163	8	zhengkun	zhengkun	PROPN
fcis-30390	163	9	li	li	PROPN
fcis-30390	163	10	,	,	PUNCT
fcis-30390	163	11	ruikun	ruikun	PROPN
fcis-30390	163	12	li	li	PROPN
fcis-30390	163	13	,	,	PUNCT
fcis-30390	163	14	xuliang	xuliang	PROPN
fcis-30390	163	15	li	li	PROPN
fcis-30390	163	16	and	and	CCONJ
fcis-30390	163	17	junbin	junbin	PROPN
fcis-30390	163	18	gao	gao	PROPN
fcis-30390	163	19	.	.	PUNCT
fcis-30390	164	1	“	"	PUNCT
fcis-30390	164	2	diffusion	diffusion	NOUN
fcis-30390	164	3	models	model	NOUN
fcis-30390	164	4	for	for	ADP
fcis-30390	164	5	time	time	NOUN
fcis-30390	164	6	-	-	PUNCT
fcis-30390	164	7	series	series	NOUN
fcis-30390	164	8	applications	application	NOUN
fcis-30390	164	9	:	:	PUNCT
fcis-30390	164	10	a	a	DET
fcis-30390	164	11	survey	survey	NOUN
fcis-30390	164	12	.	.	PUNCT
fcis-30390	164	13	”	"	PUNCT
fcis-30390	165	1	frontiers	frontier	NOUN
fcis-30390	165	2	of	of	ADP
fcis-30390	165	3	information	information	NOUN
fcis-30390	165	4	technology	technology	PROPN
fcis-30390	165	5	&	&	CCONJ
fcis-30390	165	6	electronic	electronic	ADJ
fcis-30390	165	7	engineering	engineering	NOUN
fcis-30390	165	8	25	25	NUM
fcis-30390	165	9	(	(	PUNCT
fcis-30390	165	10	2023	2023	NUM
fcis-30390	165	11	):	):	PUNCT
fcis-30390	165	12	19	19	NUM
fcis-30390	165	13	-	-	SYM
fcis-30390	165	14	41	41	NUM
fcis-30390	165	15	.	.	PUNCT
fcis-30390	166	1	[	[	X
fcis-30390	166	2	11	11	NUM
fcis-30390	166	3	]	]	X
fcis-30390	166	4	liu	liu	PROPN
fcis-30390	166	5	,	,	PUNCT
fcis-30390	166	6	yong	yong	PROPN
fcis-30390	166	7	,	,	PUNCT
fcis-30390	166	8	chenyu	chenyu	ADJ
fcis-30390	166	9	li	li	PROPN
fcis-30390	166	10	,	,	PUNCT
fcis-30390	166	11	jianmin	jianmin	PROPN
fcis-30390	166	12	wang	wang	PROPN
fcis-30390	166	13	and	and	CCONJ
fcis-30390	166	14	mingsheng	mingsheng	PROPN
fcis-30390	166	15	long	long	ADV
fcis-30390	166	16	.	.	PUNCT
fcis-30390	167	1	“	"	PUNCT
fcis-30390	167	2	koopa	koopa	X
fcis-30390	167	3	:	:	PUNCT
fcis-30390	167	4	learning	learn	VERB
fcis-30390	167	5	non	non	ADJ
fcis-30390	167	6	-	-	ADJ
fcis-30390	167	7	stationary	stationary	ADJ
fcis-30390	167	8	time	time	NOUN
fcis-30390	167	9	series	series	PROPN
fcis-30390	167	10	dynamics	dynamic	NOUN
fcis-30390	167	11	with	with	ADP
fcis-30390	167	12	koopman	koopman	PROPN
fcis-30390	167	13	predictors	predictor	NOUN
fcis-30390	167	14	.	.	PUNCT
fcis-30390	167	15	”	"	PUNCT
fcis-30390	168	1	arxiv	arxiv	PROPN
fcis-30390	168	2	abs/2305.18803	abs/2305.18803	NOUN
fcis-30390	168	3	(	(	PUNCT
fcis-30390	168	4	2023	2023	NUM
fcis-30390	168	5	):	):	PUNCT
fcis-30390	168	6	n.	n.	PROPN
fcis-30390	168	7	pag	pag	PROPN
fcis-30390	168	8	.	.	PUNCT
fcis-30390	169	1	[	[	X
fcis-30390	169	2	12	12	NUM
fcis-30390	169	3	]	]	X
fcis-30390	169	4	lee	lee	PROPN
fcis-30390	169	5	,	,	PUNCT
fcis-30390	169	6	hyunwoo	hyunwoo	PROPN
fcis-30390	169	7	,	,	PUNCT
fcis-30390	169	8	seungmin	seungmin	ADJ
fcis-30390	169	9	jin	jin	PROPN
fcis-30390	169	10	,	,	PUNCT
fcis-30390	169	11	hyeshin	hyeshin	PROPN
fcis-30390	169	12	chu	chu	PROPN
fcis-30390	169	13	,	,	PUNCT
fcis-30390	169	14	hong	hong	PROPN
fcis-30390	169	15	sik	sik	PROPN
fcis-30390	169	16	lim	lim	PROPN
fcis-30390	169	17	and	and	CCONJ
fcis-30390	169	18	sungahn	sungahn	PROPN
fcis-30390	169	19	ko	ko	PROPN
fcis-30390	169	20	.	.	PUNCT
fcis-30390	170	1	“	"	PUNCT
fcis-30390	170	2	learning	learn	VERB
fcis-30390	170	3	to	to	PART
fcis-30390	170	4	remember	remember	VERB
fcis-30390	170	5	patterns	pattern	NOUN
fcis-30390	170	6	:	:	PUNCT
fcis-30390	170	7	pattern	pattern	NOUN
fcis-30390	170	8	matching	matching	NOUN
fcis-30390	170	9	memory	memory	NOUN
fcis-30390	170	10	networks	network	NOUN
fcis-30390	170	11	for	for	ADP
fcis-30390	170	12	traffic	traffic	NOUN
fcis-30390	170	13	forecasting	forecasting	NOUN
fcis-30390	170	14	.	.	PUNCT
fcis-30390	170	15	”	"	PUNCT
fcis-30390	171	1	arxiv	arxiv	PROPN
fcis-30390	171	2	abs/2110.10380	abs/2110.10380	ADP
fcis-30390	171	3	(	(	PUNCT
fcis-30390	171	4	2021	2021	NUM
fcis-30390	171	5	):	):	PUNCT
fcis-30390	171	6	n.	n.	PROPN
fcis-30390	171	7	pag	pag	PROPN
fcis-30390	171	8	.	.	PUNCT
fcis-30390	172	1	[	[	X
fcis-30390	172	2	13	13	NUM
fcis-30390	172	3	]	]	X
fcis-30390	172	4	liu	liu	PROPN
fcis-30390	172	5	,	,	PUNCT
fcis-30390	172	6	xiangyue	xiangyue	PROPN
fcis-30390	172	7	,	,	PUNCT
fcis-30390	172	8	xinqi	xinqi	PROPN
fcis-30390	172	9	lyu	lyu	NOUN
fcis-30390	172	10	,	,	PUNCT
fcis-30390	172	11	xiangchi	xiangchi	PROPN
fcis-30390	172	12	zhang	zhang	PROPN
fcis-30390	172	13	,	,	PUNCT
fcis-30390	172	14	jianliang	jianliang	PROPN
fcis-30390	172	15	gao	gao	PROPN
fcis-30390	172	16	and	and	CCONJ
fcis-30390	172	17	jiamin	jiamin	PROPN
fcis-30390	172	18	chen	chen	PROPN
fcis-30390	172	19	.	.	PUNCT
fcis-30390	173	1	“	"	PUNCT
fcis-30390	173	2	memory	memory	NOUN
fcis-30390	173	3	augmented	augment	VERB
fcis-30390	173	4	graph	graph	NOUN
fcis-30390	173	5	learning	learn	VERB
fcis-30390	173	6	networks	network	NOUN
fcis-30390	173	7	for	for	ADP
fcis-30390	173	8	multivariate	multivariate	NOUN
fcis-30390	173	9	time	time	NOUN
fcis-30390	173	10	series	series	PROPN
fcis-30390	173	11	forecasting	forecasting	PROPN
fcis-30390	173	12	.	.	PUNCT
fcis-30390	173	13	”	"	PUNCT
fcis-30390	173	14	proceedings	proceeding	NOUN
fcis-30390	173	15	of	of	ADP
fcis-30390	173	16	the	the	DET
fcis-30390	173	17	31st	31st	PROPN
fcis-30390	173	18	acm	acm	PROPN
fcis-30390	173	19	international	international	ADJ
fcis-30390	173	20	conference	conference	NOUN
fcis-30390	173	21	on	on	ADP
fcis-30390	173	22	information	information	NOUN
fcis-30390	173	23	&	&	CCONJ
fcis-30390	173	24	knowledge	knowledge	PROPN
fcis-30390	173	25	management	management	PROPN
fcis-30390	173	26	(	(	PUNCT
fcis-30390	173	27	2022	2022	NUM
fcis-30390	173	28	):	):	PUNCT
fcis-30390	173	29	n.	n.	PROPN
fcis-30390	173	30	pag	pag	PROPN
fcis-30390	173	31	.	.	PUNCT
fcis-30390	174	1	[	[	X
fcis-30390	174	2	14	14	NUM
fcis-30390	174	3	]	]	X
fcis-30390	174	4	zhang	zhang	PROPN
fcis-30390	174	5	,	,	PUNCT
fcis-30390	174	6	yunhao	yunhao	PROPN
fcis-30390	174	7	and	and	CCONJ
fcis-30390	174	8	junchi	junchi	PROPN
fcis-30390	174	9	yan	yan	PROPN
fcis-30390	174	10	.	.	PUNCT
fcis-30390	175	1	“	"	PUNCT
fcis-30390	175	2	crossformer	crossformer	NOUN
fcis-30390	175	3	:	:	PUNCT
fcis-30390	175	4	transformer	transformer	NOUN
fcis-30390	175	5	utilizing	utilize	VERB
fcis-30390	175	6	cross	cross	ADJ
fcis-30390	175	7	-	-	ADJ
fcis-30390	175	8	dimension	dimension	ADJ
fcis-30390	175	9	dependency	dependency	NOUN
fcis-30390	175	10	for	for	ADP
fcis-30390	175	11	multivariate	multivariate	NOUN
fcis-30390	175	12	time	time	NOUN
fcis-30390	175	13	series	series	PROPN
fcis-30390	175	14	forecasting	forecasting	PROPN
fcis-30390	175	15	.	.	PUNCT
fcis-30390	175	16	”	"	PUNCT
fcis-30390	176	1	international	international	ADJ
fcis-30390	176	2	conference	conference	NOUN
fcis-30390	176	3	on	on	ADP
fcis-30390	176	4	learning	learn	VERB
fcis-30390	176	5	representations	representation	NOUN
fcis-30390	176	6	(	(	PUNCT
fcis-30390	176	7	2023	2023	NUM
fcis-30390	176	8	)	)	PUNCT
fcis-30390	176	9	.	.	PUNCT
fcis-30390	177	1	[	[	X
fcis-30390	177	2	15	15	NUM
fcis-30390	177	3	]	]	X
fcis-30390	177	4	wang	wang	PROPN
fcis-30390	177	5	,	,	PUNCT
fcis-30390	177	6	shiyu	shiyu	PROPN
fcis-30390	177	7	,	,	PUNCT
fcis-30390	177	8	haixu	haixu	ADJ
fcis-30390	177	9	wu	wu	PROPN
fcis-30390	177	10	,	,	PUNCT
fcis-30390	177	11	xiao	xiao	PROPN
fcis-30390	177	12	long	long	PROPN
fcis-30390	177	13	shi	shi	PROPN
fcis-30390	177	14	,	,	PUNCT
fcis-30390	177	15	tengge	tengge	PROPN
fcis-30390	177	16	hu	hu	PROPN
fcis-30390	177	17	,	,	PUNCT
fcis-30390	177	18	huakun	huakun	PROPN
fcis-30390	177	19	luo	luo	PROPN
fcis-30390	177	20	,	,	PUNCT
fcis-30390	177	21	lintao	lintao	PROPN
fcis-30390	177	22	ma	ma	PROPN
fcis-30390	177	23	,	,	PUNCT
fcis-30390	177	24	james	james	PROPN
fcis-30390	177	25	y.	y.	PROPN
fcis-30390	177	26	zhang	zhang	PROPN
fcis-30390	177	27	and	and	CCONJ
fcis-30390	177	28	jun	jun	PROPN
fcis-30390	177	29	zhou	zhou	PROPN
fcis-30390	177	30	.	.	PUNCT
fcis-30390	178	1	“	"	PUNCT
fcis-30390	178	2	timemixer	timemixer	NOUN
fcis-30390	178	3	:	:	PUNCT
fcis-30390	178	4	decomposable	decomposable	ADJ
fcis-30390	178	5	multiscale	multiscale	NOUN
fcis-30390	178	6	mixing	mix	VERB
fcis-30390	178	7	for	for	ADP
fcis-30390	178	8	time	time	NOUN
fcis-30390	178	9	series	series	PROPN
fcis-30390	178	10	forecasting	forecasting	PROPN
fcis-30390	178	11	.	.	PUNCT
fcis-30390	178	12	”	"	PUNCT
fcis-30390	179	1	arxiv	arxiv	PROPN
fcis-30390	179	2	abs/2405.14616	abs/2405.14616	PROPN
fcis-30390	179	3	(	(	PUNCT
fcis-30390	179	4	2024	2024	NUM
fcis-30390	179	5	):	):	PUNCT
fcis-30390	179	6	n.	n.	PROPN
fcis-30390	179	7	pag	pag	PROPN
fcis-30390	179	8	.	.	PUNCT
fcis-30390	180	1	[	[	X
fcis-30390	180	2	16	16	NUM
fcis-30390	180	3	]	]	PUNCT
fcis-30390	180	4	xu	xu	PROPN
fcis-30390	180	5	,	,	PUNCT
fcis-30390	180	6	luwen	luwen	NOUN
fcis-30390	180	7	,	,	PUNCT
fcis-30390	180	8	jiwei	jiwei	PROPN
fcis-30390	180	9	qin	qin	PROPN
fcis-30390	180	10	,	,	PUNCT
fcis-30390	180	11	dezhi	dezhi	PROPN
fcis-30390	180	12	sun	sun	PROPN
fcis-30390	180	13	,	,	PUNCT
fcis-30390	180	14	yuanyuan	yuanyuan	PROPN
fcis-30390	180	15	liao	liao	PROPN
fcis-30390	180	16	and	and	CCONJ
fcis-30390	180	17	jiong	jiong	PROPN
fcis-30390	180	18	zheng	zheng	PROPN
fcis-30390	180	19	.	.	PUNCT
fcis-30390	181	1	“	"	PUNCT
fcis-30390	181	2	pfformer	pfformer	NOUN
fcis-30390	181	3	:	:	PUNCT
fcis-30390	181	4	a	a	DET
fcis-30390	181	5	time	time	NOUN
fcis-30390	181	6	-	-	PUNCT
fcis-30390	181	7	series	series	NOUN
fcis-30390	181	8	forecasting	forecasting	NOUN
fcis-30390	181	9	model	model	NOUN
fcis-30390	181	10	for	for	ADP
fcis-30390	181	11	short	short	ADJ
fcis-30390	181	12	-	-	PUNCT
fcis-30390	181	13	term	term	NOUN
fcis-30390	181	14	precipitation	precipitation	NOUN
fcis-30390	181	15	forecasting	forecasting	NOUN
fcis-30390	181	16	.	.	PUNCT
fcis-30390	181	17	”	"	PUNCT
fcis-30390	182	1	ieee	ieee	NOUN
fcis-30390	182	2	access	access	NOUN
fcis-30390	182	3	12	12	NUM
fcis-30390	182	4	(	(	PUNCT
fcis-30390	182	5	2024	2024	NUM
fcis-30390	182	6	):	):	PUNCT
fcis-30390	182	7	130948	130948	NUM
fcis-30390	182	8	-	-	SYM
fcis-30390	182	9	130961	130961	NUM
fcis-30390	182	10	.	.	PUNCT
fcis-30390	183	1	[	[	X
fcis-30390	183	2	17	17	NUM
fcis-30390	183	3	]	]	X
fcis-30390	183	4	woo	woo	PROPN
fcis-30390	183	5	,	,	PUNCT
fcis-30390	183	6	gerald	gerald	PROPN
fcis-30390	183	7	,	,	PUNCT
fcis-30390	183	8	chenghao	chenghao	PROPN
fcis-30390	183	9	liu	liu	PROPN
fcis-30390	183	10	,	,	PUNCT
fcis-30390	183	11	akshat	akshat	PROPN
fcis-30390	183	12	kumar	kumar	PROPN
fcis-30390	183	13	,	,	PUNCT
fcis-30390	183	14	caiming	caime	VERB
fcis-30390	183	15	xiong	xiong	PROPN
fcis-30390	183	16	,	,	PUNCT
fcis-30390	183	17	silvio	silvio	PROPN
fcis-30390	183	18	savarese	savarese	PROPN
fcis-30390	183	19	and	and	CCONJ
fcis-30390	183	20	doyen	doyen	PROPN
fcis-30390	183	21	sahoo	sahoo	PROPN
fcis-30390	183	22	.	.	PUNCT
fcis-30390	184	1	“	"	PUNCT
fcis-30390	184	2	unified	unified	ADJ
fcis-30390	184	3	training	training	NOUN
fcis-30390	184	4	of	of	ADP
fcis-30390	184	5	universal	universal	ADJ
fcis-30390	184	6	time	time	NOUN
fcis-30390	184	7	series	series	PROPN
fcis-30390	184	8	forecasting	forecasting	NOUN
fcis-30390	184	9	transformers	transformer	NOUN
fcis-30390	184	10	.	.	PUNCT
fcis-30390	184	11	”	"	PUNCT
fcis-30390	185	1	arxiv	arxiv	PROPN
fcis-30390	185	2	abs/2402.02592	abs/2402.02592	NOUN
fcis-30390	185	3	(	(	PUNCT
fcis-30390	185	4	2024	2024	NUM
fcis-30390	185	5	):	):	PUNCT
fcis-30390	185	6	n.	n.	PROPN
fcis-30390	185	7	pag	pag	PROPN
fcis-30390	185	8	.	.	PUNCT