id	sid	tid	token	lemma	pos
ajst-7347	1	1	academic	academic	ADJ
ajst-7347	1	2	journal	journal	NOUN
ajst-7347	1	3	of	of	ADP
ajst-7347	1	4	science	science	NOUN
ajst-7347	1	5	and	and	CCONJ
ajst-7347	1	6	technology	technology	NOUN
ajst-7347	1	7	issn	issn	NOUN
ajst-7347	1	8	:	:	PUNCT
ajst-7347	1	9	2771	2771	NUM
ajst-7347	1	10	-	-	SYM
ajst-7347	1	11	3032	3032	NUM
ajst-7347	1	12	|	|	NOUN
ajst-7347	1	13	vol	vol	NOUN
ajst-7347	1	14	.	.	PROPN
ajst-7347	2	1	5	5	NUM
ajst-7347	2	2	,	,	PUNCT
ajst-7347	2	3	no	no	INTJ
ajst-7347	2	4	.	.	NOUN
ajst-7347	2	5	3	3	NUM
ajst-7347	2	6	,	,	PUNCT
ajst-7347	2	7	2023	2023	NUM
ajst-7347	2	8	1	1	NUM
ajst-7347	2	9	prediction	prediction	NOUN
ajst-7347	2	10	of	of	ADP
ajst-7347	2	11	pm2.5	pm2.5	DET
ajst-7347	2	12	concentration	concentration	NOUN
ajst-7347	2	13	based	base	VERB
ajst-7347	2	14	on	on	ADP
ajst-7347	2	15	cnn‐	cnn‐	NOUN
ajst-7347	2	16	bigru	bigru	PROPN
ajst-7347	2	17	model	model	PROPN
ajst-7347	2	18	xinfang	xinfang	PROPN
ajst-7347	2	19	li	li	PROPN
ajst-7347	2	20	,	,	PUNCT
ajst-7347	2	21	hua	hua	PROPN
ajst-7347	2	22	huo	huo	PROPN
ajst-7347	2	23	henan	henan	PROPN
ajst-7347	2	24	university	university	PROPN
ajst-7347	2	25	of	of	ADP
ajst-7347	2	26	science	science	NOUN
ajst-7347	2	27	and	and	CCONJ
ajst-7347	2	28	technology	technology	NOUN
ajst-7347	2	29	,	,	PUNCT
ajst-7347	2	30	school	school	NOUN
ajst-7347	2	31	of	of	ADP
ajst-7347	2	32	information	information	NOUN
ajst-7347	2	33	engineering	engineering	NOUN
ajst-7347	2	34	,	,	PUNCT
ajst-7347	2	35	china	china	PROPN
ajst-7347	2	36	abstract	abstract	NOUN
ajst-7347	2	37	:	:	PUNCT
ajst-7347	2	38	the	the	DET
ajst-7347	2	39	issue	issue	NOUN
ajst-7347	2	40	of	of	ADP
ajst-7347	2	41	air	air	NOUN
ajst-7347	2	42	pollution	pollution	NOUN
ajst-7347	2	43	has	have	AUX
ajst-7347	2	44	always	always	ADV
ajst-7347	2	45	been	be	AUX
ajst-7347	2	46	a	a	DET
ajst-7347	2	47	concern	concern	NOUN
ajst-7347	2	48	.	.	PUNCT
ajst-7347	3	1	bad	bad	ADJ
ajst-7347	3	2	smog	smog	NOUN
ajst-7347	3	3	weather	weather	NOUN
ajst-7347	3	4	not	not	PART
ajst-7347	3	5	only	only	ADV
ajst-7347	3	6	brings	bring	VERB
ajst-7347	3	7	inconvenience	inconvenience	NOUN
ajst-7347	3	8	to	to	ADP
ajst-7347	3	9	people	people	NOUN
ajst-7347	3	10	's	's	PART
ajst-7347	3	11	travel	travel	NOUN
ajst-7347	3	12	,	,	PUNCT
ajst-7347	3	13	but	but	CCONJ
ajst-7347	3	14	also	also	ADV
ajst-7347	3	15	poses	pose	VERB
ajst-7347	3	16	a	a	DET
ajst-7347	3	17	threat	threat	NOUN
ajst-7347	3	18	to	to	ADP
ajst-7347	3	19	people	people	NOUN
ajst-7347	3	20	's	's	PART
ajst-7347	3	21	health	health	NOUN
ajst-7347	3	22	.	.	PUNCT
ajst-7347	4	1	pm2.5	pm2.5	DET
ajst-7347	4	2	concentration	concentration	NOUN
ajst-7347	4	3	is	be	AUX
ajst-7347	4	4	an	an	DET
ajst-7347	4	5	important	important	ADJ
ajst-7347	4	6	indicator	indicator	NOUN
ajst-7347	4	7	of	of	ADP
ajst-7347	4	8	air	air	NOUN
ajst-7347	4	9	conditions	condition	NOUN
ajst-7347	4	10	.	.	PUNCT
ajst-7347	5	1	therefore	therefore	ADV
ajst-7347	5	2	,	,	PUNCT
ajst-7347	5	3	it	it	PRON
ajst-7347	5	4	is	be	AUX
ajst-7347	5	5	of	of	ADP
ajst-7347	5	6	long	long	ADJ
ajst-7347	5	7	-	-	PUNCT
ajst-7347	5	8	term	term	NOUN
ajst-7347	5	9	significance	significance	NOUN
ajst-7347	5	10	to	to	PART
ajst-7347	5	11	analyze	analyze	VERB
ajst-7347	5	12	and	and	CCONJ
ajst-7347	5	13	predict	predict	VERB
ajst-7347	5	14	the	the	DET
ajst-7347	5	15	concentration	concentration	NOUN
ajst-7347	5	16	of	of	ADP
ajst-7347	5	17	pm2.5	pm2.5	PROPN
ajst-7347	5	18	.	.	PUNCT
ajst-7347	6	1	aiming	aim	VERB
ajst-7347	6	2	at	at	ADP
ajst-7347	6	3	the	the	DET
ajst-7347	6	4	problem	problem	NOUN
ajst-7347	6	5	that	that	SCONJ
ajst-7347	6	6	a	a	DET
ajst-7347	6	7	single	single	ADJ
ajst-7347	6	8	machine	machine	NOUN
ajst-7347	6	9	learning	learning	NOUN
ajst-7347	6	10	model	model	NOUN
ajst-7347	6	11	can	can	AUX
ajst-7347	6	12	not	not	PART
ajst-7347	6	13	consider	consider	VERB
ajst-7347	6	14	the	the	DET
ajst-7347	6	15	impact	impact	NOUN
ajst-7347	6	16	of	of	ADP
ajst-7347	6	17	multiple	multiple	ADJ
ajst-7347	6	18	factors	factor	NOUN
ajst-7347	6	19	on	on	ADP
ajst-7347	6	20	pm2.5	pm2.5	DET
ajst-7347	6	21	concentration	concentration	NOUN
ajst-7347	6	22	changes	change	NOUN
ajst-7347	6	23	,	,	PUNCT
ajst-7347	6	24	and	and	CCONJ
ajst-7347	6	25	the	the	DET
ajst-7347	6	26	data	data	NOUN
ajst-7347	6	27	characteristics	characteristic	NOUN
ajst-7347	6	28	are	be	AUX
ajst-7347	6	29	complex	complex	ADJ
ajst-7347	6	30	,	,	PUNCT
ajst-7347	6	31	which	which	PRON
ajst-7347	6	32	can	can	AUX
ajst-7347	6	33	not	not	PART
ajst-7347	6	34	better	well	ADV
ajst-7347	6	35	capture	capture	VERB
ajst-7347	6	36	all	all	DET
ajst-7347	6	37	the	the	DET
ajst-7347	6	38	characteristics	characteristic	NOUN
ajst-7347	6	39	of	of	ADP
ajst-7347	6	40	the	the	DET
ajst-7347	6	41	data	datum	NOUN
ajst-7347	6	42	,	,	PUNCT
ajst-7347	6	43	and	and	CCONJ
ajst-7347	6	44	can	can	AUX
ajst-7347	6	45	not	not	PART
ajst-7347	6	46	highlight	highlight	VERB
ajst-7347	6	47	the	the	DET
ajst-7347	6	48	regularity	regularity	NOUN
ajst-7347	6	49	of	of	ADP
ajst-7347	6	50	pm2.5	pm2.5	DET
ajst-7347	6	51	changes	change	NOUN
ajst-7347	6	52	over	over	ADP
ajst-7347	6	53	time	time	NOUN
ajst-7347	6	54	,	,	PUNCT
ajst-7347	6	55	the	the	DET
ajst-7347	6	56	construction	construction	NOUN
ajst-7347	6	57	of	of	ADP
ajst-7347	6	58	a	a	DET
ajst-7347	6	59	combined	combine	VERB
ajst-7347	6	60	model	model	NOUN
ajst-7347	6	61	further	far	ADV
ajst-7347	6	62	improves	improve	VERB
ajst-7347	6	63	the	the	DET
ajst-7347	6	64	prediction	prediction	NOUN
ajst-7347	6	65	accuracy	accuracy	NOUN
ajst-7347	6	66	.	.	PUNCT
ajst-7347	7	1	firstly	firstly	ADV
ajst-7347	7	2	,	,	PUNCT
ajst-7347	7	3	based	base	VERB
ajst-7347	7	4	on	on	ADP
ajst-7347	7	5	the	the	DET
ajst-7347	7	6	pm2.5	pm2.5	ADJ
ajst-7347	7	7	concentration	concentration	NOUN
ajst-7347	7	8	values	value	NOUN
ajst-7347	7	9	,	,	PUNCT
ajst-7347	7	10	air	air	NOUN
ajst-7347	7	11	quality	quality	NOUN
ajst-7347	7	12	data	datum	NOUN
ajst-7347	7	13	,	,	PUNCT
ajst-7347	7	14	and	and	CCONJ
ajst-7347	7	15	meteorological	meteorological	ADJ
ajst-7347	7	16	data	datum	NOUN
ajst-7347	7	17	at	at	ADP
ajst-7347	7	18	various	various	ADJ
ajst-7347	7	19	stations	station	NOUN
ajst-7347	7	20	in	in	ADP
ajst-7347	7	21	new	new	ADJ
ajst-7347	7	22	taipei	taipei	PROPN
ajst-7347	7	23	city	city	PROPN
ajst-7347	7	24	,	,	PUNCT
ajst-7347	7	25	taiwan	taiwan	PROPN
ajst-7347	7	26	province	province	PROPN
ajst-7347	7	27	,	,	PUNCT
ajst-7347	7	28	through	through	ADP
ajst-7347	7	29	analyzing	analyze	VERB
ajst-7347	7	30	the	the	DET
ajst-7347	7	31	spatiotemporal	spatiotemporal	ADJ
ajst-7347	7	32	distribution	distribution	NOUN
ajst-7347	7	33	characteristics	characteristic	NOUN
ajst-7347	7	34	of	of	ADP
ajst-7347	7	35	the	the	DET
ajst-7347	7	36	pm2.5	pm2.5	ADJ
ajst-7347	7	37	concentration	concentration	NOUN
ajst-7347	7	38	at	at	ADP
ajst-7347	7	39	the	the	DET
ajst-7347	7	40	target	target	NOUN
ajst-7347	7	41	station	station	NOUN
ajst-7347	7	42	,	,	PUNCT
ajst-7347	7	43	as	as	ADV
ajst-7347	7	44	well	well	ADV
ajst-7347	7	45	as	as	ADP
ajst-7347	7	46	the	the	DET
ajst-7347	7	47	correlation	correlation	NOUN
ajst-7347	7	48	with	with	ADP
ajst-7347	7	49	various	various	ADJ
ajst-7347	7	50	pollutant	pollutant	ADJ
ajst-7347	7	51	factors	factor	NOUN
ajst-7347	7	52	and	and	CCONJ
ajst-7347	7	53	meteorological	meteorological	ADJ
ajst-7347	7	54	factors	factor	NOUN
ajst-7347	7	55	,	,	PUNCT
ajst-7347	7	56	spearman	spearman	NOUN
ajst-7347	7	57	correlation	correlation	NOUN
ajst-7347	7	58	analysis	analysis	NOUN
ajst-7347	7	59	is	be	AUX
ajst-7347	7	60	used	use	VERB
ajst-7347	7	61	for	for	ADP
ajst-7347	7	62	feature	feature	NOUN
ajst-7347	7	63	selection	selection	NOUN
ajst-7347	7	64	.	.	PUNCT
ajst-7347	8	1	the	the	DET
ajst-7347	8	2	combined	combined	PROPN
ajst-7347	8	3	model	model	PROPN
ajst-7347	8	4	cnn	cnn	PROPN
ajst-7347	8	5	-	-	PUNCT
ajst-7347	8	6	bigru	bigru	PROPN
ajst-7347	8	7	constructed	construct	VERB
ajst-7347	8	8	in	in	ADP
ajst-7347	8	9	this	this	DET
ajst-7347	8	10	paper	paper	NOUN
ajst-7347	8	11	utilizes	utilize	VERB
ajst-7347	8	12	its	its	PRON
ajst-7347	8	13	unique	unique	ADJ
ajst-7347	8	14	convolution	convolution	NOUN
ajst-7347	8	15	operation	operation	NOUN
ajst-7347	8	16	to	to	PART
ajst-7347	8	17	extract	extract	VERB
ajst-7347	8	18	features	feature	NOUN
ajst-7347	8	19	from	from	ADP
ajst-7347	8	20	one	one	NUM
ajst-7347	8	21	-	-	PUNCT
ajst-7347	8	22	dimensional	dimensional	ADJ
ajst-7347	8	23	data	datum	NOUN
ajst-7347	8	24	,	,	PUNCT
ajst-7347	8	25	and	and	CCONJ
ajst-7347	8	26	combines	combine	VERB
ajst-7347	8	27	the	the	DET
ajst-7347	8	28	circular	circular	ADJ
ajst-7347	8	29	neural	neural	ADJ
ajst-7347	8	30	network	network	NOUN
ajst-7347	8	31	bigru	bigru	VERB
ajst-7347	8	32	with	with	ADP
ajst-7347	8	33	bidirectional	bidirectional	ADJ
ajst-7347	8	34	transmission	transmission	NOUN
ajst-7347	8	35	function	function	NOUN
ajst-7347	8	36	to	to	PART
ajst-7347	8	37	model	model	VERB
ajst-7347	8	38	and	and	CCONJ
ajst-7347	8	39	predict	predict	VERB
ajst-7347	8	40	pm2.5	pm2.5	DET
ajst-7347	8	41	concentration	concentration	NOUN
ajst-7347	8	42	based	base	VERB
ajst-7347	8	43	on	on	ADP
ajst-7347	8	44	the	the	DET
ajst-7347	8	45	functional	functional	ADJ
ajst-7347	8	46	advantages	advantage	NOUN
ajst-7347	8	47	of	of	ADP
ajst-7347	8	48	both	both	DET
ajst-7347	8	49	parties	party	NOUN
ajst-7347	8	50	.	.	PUNCT
ajst-7347	9	1	keywords	keyword	NOUN
ajst-7347	9	2	:	:	PUNCT
ajst-7347	9	3	pm2.5	pm2.5	NUM
ajst-7347	9	4	,	,	PUNCT
ajst-7347	9	5	cnn	cnn	PROPN
ajst-7347	9	6	,	,	PUNCT
ajst-7347	9	7	gru	gru	PROPN
ajst-7347	9	8	,	,	PUNCT
ajst-7347	9	9	combination	combination	NOUN
ajst-7347	9	10	model	model	NOUN
ajst-7347	9	11	.	.	PUNCT
ajst-7347	10	1	1	1	X
ajst-7347	10	2	.	.	X
ajst-7347	10	3	introduction	introduction	NOUN
ajst-7347	10	4	with	with	ADP
ajst-7347	10	5	the	the	DET
ajst-7347	10	6	acceleration	acceleration	NOUN
ajst-7347	10	7	of	of	ADP
ajst-7347	10	8	china	china	PROPN
ajst-7347	10	9	's	's	PART
ajst-7347	10	10	industrialization	industrialization	NOUN
ajst-7347	10	11	process	process	NOUN
ajst-7347	10	12	,	,	PUNCT
ajst-7347	10	13	the	the	DET
ajst-7347	10	14	accompanying	accompanying	ADJ
ajst-7347	10	15	air	air	NOUN
ajst-7347	10	16	pollution	pollution	NOUN
ajst-7347	10	17	problem	problem	NOUN
ajst-7347	10	18	is	be	AUX
ajst-7347	10	19	becoming	become	VERB
ajst-7347	10	20	increasingly	increasingly	ADV
ajst-7347	10	21	serious	serious	ADJ
ajst-7347	10	22	.	.	PUNCT
ajst-7347	11	1	poor	poor	ADJ
ajst-7347	11	2	air	air	NOUN
ajst-7347	11	3	quality	quality	NOUN
ajst-7347	11	4	can	can	AUX
ajst-7347	11	5	affect	affect	VERB
ajst-7347	11	6	our	our	PRON
ajst-7347	11	7	daily	daily	ADJ
ajst-7347	11	8	lives	life	NOUN
ajst-7347	11	9	and	and	CCONJ
ajst-7347	11	10	even	even	ADV
ajst-7347	11	11	our	our	PRON
ajst-7347	11	12	physical	physical	ADJ
ajst-7347	11	13	and	and	CCONJ
ajst-7347	11	14	mental	mental	ADJ
ajst-7347	11	15	health	health	NOUN
ajst-7347	11	16	,	,	PUNCT
ajst-7347	11	17	with	with	ADP
ajst-7347	11	18	pm2.5	pm2.5	PROPN
ajst-7347	11	19	causing	cause	VERB
ajst-7347	11	20	the	the	DET
ajst-7347	11	21	most	most	ADV
ajst-7347	11	22	serious	serious	ADJ
ajst-7347	11	23	harm	harm	NOUN
ajst-7347	11	24	to	to	ADP
ajst-7347	11	25	the	the	DET
ajst-7347	11	26	human	human	ADJ
ajst-7347	11	27	body[1,2	body[1,2	PROPN
ajst-7347	11	28	]	]	X
ajst-7347	11	29	.	.	PUNCT
ajst-7347	12	1	accurate	accurate	ADJ
ajst-7347	12	2	and	and	CCONJ
ajst-7347	12	3	timely	timely	ADJ
ajst-7347	12	4	prediction	prediction	NOUN
ajst-7347	12	5	of	of	ADP
ajst-7347	12	6	future	future	ADJ
ajst-7347	12	7	pm2.5	pm2.5	ADJ
ajst-7347	12	8	concentrations	concentration	NOUN
ajst-7347	12	9	can	can	AUX
ajst-7347	12	10	provide	provide	VERB
ajst-7347	12	11	assistance	assistance	NOUN
ajst-7347	12	12	for	for	ADP
ajst-7347	12	13	people	people	NOUN
ajst-7347	12	14	's	's	PART
ajst-7347	12	15	defense	defense	NOUN
ajst-7347	12	16	.	.	PUNCT
ajst-7347	13	1	in	in	ADP
ajst-7347	13	2	recent	recent	ADJ
ajst-7347	13	3	decades	decade	NOUN
ajst-7347	13	4	,	,	PUNCT
ajst-7347	13	5	many	many	ADJ
ajst-7347	13	6	scholars	scholar	NOUN
ajst-7347	13	7	have	have	AUX
ajst-7347	13	8	analyzed	analyze	VERB
ajst-7347	13	9	and	and	CCONJ
ajst-7347	13	10	studied	study	VERB
ajst-7347	13	11	the	the	DET
ajst-7347	13	12	formation	formation	NOUN
ajst-7347	13	13	,	,	PUNCT
ajst-7347	13	14	diffusion	diffusion	NOUN
ajst-7347	13	15	,	,	PUNCT
ajst-7347	13	16	and	and	CCONJ
ajst-7347	13	17	concentration	concentration	NOUN
ajst-7347	13	18	prediction	prediction	NOUN
ajst-7347	13	19	of	of	ADP
ajst-7347	13	20	pm2.5	pm2.5	PROPN
ajst-7347	13	21	.	.	PUNCT
ajst-7347	14	1	in	in	ADP
ajst-7347	14	2	previous	previous	ADJ
ajst-7347	14	3	years	year	NOUN
ajst-7347	14	4	,	,	PUNCT
ajst-7347	14	5	most	most	ADJ
ajst-7347	14	6	of	of	ADP
ajst-7347	14	7	the	the	DET
ajst-7347	14	8	research	research	NOUN
ajst-7347	14	9	on	on	ADP
ajst-7347	14	10	pm2.5	pm2.5	PROPN
ajst-7347	14	11	used	use	VERB
ajst-7347	14	12	traditional	traditional	ADJ
ajst-7347	14	13	prediction	prediction	NOUN
ajst-7347	14	14	methods	method	NOUN
ajst-7347	14	15	,	,	PUNCT
ajst-7347	14	16	mainly	mainly	ADV
ajst-7347	14	17	based	base	VERB
ajst-7347	14	18	on	on	ADP
ajst-7347	14	19	mechanism	mechanism	NOUN
ajst-7347	14	20	models	model	NOUN
ajst-7347	14	21	,	,	PUNCT
ajst-7347	14	22	which	which	PRON
ajst-7347	14	23	dynamically	dynamically	ADV
ajst-7347	14	24	modeled	model	VERB
ajst-7347	14	25	the	the	DET
ajst-7347	14	26	transmission	transmission	NOUN
ajst-7347	14	27	and	and	CCONJ
ajst-7347	14	28	formation	formation	NOUN
ajst-7347	14	29	of	of	ADP
ajst-7347	14	30	pm2.5	pm2.5	PROPN
ajst-7347	14	31	based	base	VERB
ajst-7347	14	32	on	on	ADP
ajst-7347	14	33	the	the	DET
ajst-7347	14	34	geographical	geographical	ADJ
ajst-7347	14	35	environment	environment	NOUN
ajst-7347	14	36	,	,	PUNCT
ajst-7347	14	37	meteorological	meteorological	ADJ
ajst-7347	14	38	conditions	condition	NOUN
ajst-7347	14	39	,	,	PUNCT
ajst-7347	14	40	and	and	CCONJ
ajst-7347	14	41	even	even	ADV
ajst-7347	14	42	industrial	industrial	ADJ
ajst-7347	14	43	level	level	NOUN
ajst-7347	14	44	of	of	ADP
ajst-7347	14	45	a	a	DET
ajst-7347	14	46	region	region	NOUN
ajst-7347	14	47	.	.	PUNCT
ajst-7347	15	1	and	and	CCONJ
ajst-7347	15	2	most	most	ADJ
ajst-7347	15	3	of	of	ADP
ajst-7347	15	4	them	they	PRON
ajst-7347	15	5	use	use	VERB
ajst-7347	15	6	camq	camq	NOUN
ajst-7347	15	7	models[3	models[3	NOUN
ajst-7347	15	8	-	-	PUNCT
ajst-7347	15	9	6].for	6].for	PROPN
ajst-7347	15	10	example	example	NOUN
ajst-7347	15	11	,	,	PUNCT
ajst-7347	15	12	lee	lee	PROPN
ajst-7347	15	13	et	et	PROPN
ajst-7347	15	14	al[7	al[7	PROPN
ajst-7347	15	15	]	]	PUNCT
ajst-7347	15	16	used	use	VERB
ajst-7347	15	17	the	the	DET
ajst-7347	15	18	mm5smoke	mm5smoke	NOUN
ajst-7347	15	19	-	-	PUNCT
ajst-7347	15	20	cmaq	cmaq	NOUN
ajst-7347	15	21	modeling	modeling	NOUN
ajst-7347	15	22	system	system	NOUN
ajst-7347	15	23	to	to	PART
ajst-7347	15	24	respectively	respectively	ADV
ajst-7347	15	25	generate	generate	VERB
ajst-7347	15	26	meteorological	meteorological	ADJ
ajst-7347	15	27	fields	field	NOUN
ajst-7347	15	28	,	,	PUNCT
ajst-7347	15	29	prepare	prepare	VERB
ajst-7347	15	30	emissions	emission	NOUN
ajst-7347	15	31	,	,	PUNCT
ajst-7347	15	32	and	and	CCONJ
ajst-7347	15	33	simulate	simulate	VERB
ajst-7347	15	34	air	air	NOUN
ajst-7347	15	35	quality	quality	NOUN
ajst-7347	15	36	;	;	PUNCT
ajst-7347	15	37	jiang	jiang	PROPN
ajst-7347	15	38	et	et	PROPN
ajst-7347	15	39	al[8	al[8	PROPN
ajst-7347	15	40	]	]	PUNCT
ajst-7347	15	41	used	use	VERB
ajst-7347	15	42	a	a	DET
ajst-7347	15	43	community	community	NOUN
ajst-7347	15	44	multiscale	multiscale	ADJ
ajst-7347	15	45	air	air	NOUN
ajst-7347	15	46	quality	quality	NOUN
ajst-7347	15	47	(	(	PUNCT
ajst-7347	15	48	cmaq	cmaq	NOUN
ajst-7347	15	49	)	)	PUNCT
ajst-7347	15	50	simulation	simulation	NOUN
ajst-7347	15	51	model	model	NOUN
ajst-7347	15	52	to	to	PART
ajst-7347	15	53	predict	predict	VERB
ajst-7347	15	54	the	the	DET
ajst-7347	15	55	daily	daily	ADJ
ajst-7347	15	56	average	average	ADJ
ajst-7347	15	57	concentration	concentration	NOUN
ajst-7347	15	58	of	of	ADP
ajst-7347	15	59	fine	fine	ADJ
ajst-7347	15	60	particulate	particulate	NOUN
ajst-7347	15	61	matter	matter	NOUN
ajst-7347	15	62	(	(	PUNCT
ajst-7347	15	63	pm2.5	pm2.5	NUM
ajst-7347	15	64	)	)	PUNCT
ajst-7347	15	65	.	.	PUNCT
ajst-7347	16	1	another	another	DET
ajst-7347	16	2	prediction	prediction	NOUN
ajst-7347	16	3	method	method	NOUN
ajst-7347	16	4	is	be	AUX
ajst-7347	16	5	to	to	PART
ajst-7347	16	6	establish	establish	VERB
ajst-7347	16	7	regression	regression	NOUN
ajst-7347	16	8	models	model	NOUN
ajst-7347	16	9	.	.	PUNCT
ajst-7347	17	1	for	for	ADP
ajst-7347	17	2	example	example	NOUN
ajst-7347	17	3	,	,	PUNCT
ajst-7347	17	4	wang	wang	PROPN
ajst-7347	17	5	weirud	weirud	PROPN
ajst-7347	17	6	et	et	PROPN
ajst-7347	17	7	al[9	al[9	PROPN
ajst-7347	17	8	]	]	PUNCT
ajst-7347	17	9	used	use	VERB
ajst-7347	17	10	the	the	DET
ajst-7347	17	11	box	box	NOUN
ajst-7347	17	12	-	-	PUNCT
ajst-7347	17	13	jenkins	jenkins	PROPN
ajst-7347	17	14	theory	theory	NOUN
ajst-7347	17	15	to	to	PART
ajst-7347	17	16	establish	establish	VERB
ajst-7347	17	17	an	an	DET
ajst-7347	17	18	arima	arima	NOUN
ajst-7347	17	19	model	model	NOUN
ajst-7347	17	20	to	to	PART
ajst-7347	17	21	predict	predict	VERB
ajst-7347	17	22	the	the	DET
ajst-7347	17	23	pm2.5	pm2.5	ADJ
ajst-7347	17	24	concentration	concentration	NOUN
ajst-7347	17	25	in	in	ADP
ajst-7347	17	26	hangzhou	hangzhou	PROPN
ajst-7347	17	27	,	,	PUNCT
ajst-7347	17	28	and	and	CCONJ
ajst-7347	17	29	found	find	VERB
ajst-7347	17	30	the	the	DET
ajst-7347	17	31	optimal	optimal	ADJ
ajst-7347	17	32	parameter	parameter	NOUN
ajst-7347	17	33	value	value	NOUN
ajst-7347	17	34	with	with	ADP
ajst-7347	17	35	good	good	ADJ
ajst-7347	17	36	prediction	prediction	NOUN
ajst-7347	17	37	accuracy	accuracy	NOUN
ajst-7347	17	38	;	;	PUNCT
ajst-7347	17	39	wang	wang	PROPN
ajst-7347	17	40	juan[10	juan[10	PROPN
ajst-7347	17	41	]	]	PUNCT
ajst-7347	17	42	used	use	VERB
ajst-7347	17	43	multiple	multiple	ADJ
ajst-7347	17	44	regression	regression	NOUN
ajst-7347	17	45	analysis	analysis	NOUN
ajst-7347	17	46	and	and	CCONJ
ajst-7347	17	47	gray	gray	ADJ
ajst-7347	17	48	correlation	correlation	NOUN
ajst-7347	17	49	2	2	NUM
ajst-7347	17	50	degree	degree	NOUN
ajst-7347	17	51	method	method	NOUN
ajst-7347	17	52	to	to	PART
ajst-7347	17	53	conduct	conduct	VERB
ajst-7347	17	54	air	air	NOUN
ajst-7347	17	55	pollution	pollution	NOUN
ajst-7347	17	56	quality	quality	NOUN
ajst-7347	17	57	research	research	NOUN
ajst-7347	17	58	in	in	ADP
ajst-7347	17	59	most	most	ADJ
ajst-7347	17	60	areas	area	NOUN
ajst-7347	17	61	of	of	ADP
ajst-7347	17	62	the	the	DET
ajst-7347	17	63	city	city	NOUN
ajst-7347	17	64	;	;	PUNCT
ajst-7347	17	65	wang	wang	PROPN
ajst-7347	17	66	jing	jing	PROPN
ajst-7347	17	67	et	et	PROPN
ajst-7347	17	68	al[11]proposed	al[11]propose	VERB
ajst-7347	17	69	a	a	DET
ajst-7347	17	70	hierarchical	hierarchical	ADJ
ajst-7347	17	71	autoregressive	autoregressive	ADJ
ajst-7347	17	72	model	model	NOUN
ajst-7347	17	73	based	base	VERB
ajst-7347	17	74	on	on	ADP
ajst-7347	17	75	bayesian	bayesian	NOUN
ajst-7347	17	76	to	to	PART
ajst-7347	17	77	handle	handle	VERB
ajst-7347	17	78	the	the	DET
ajst-7347	17	79	synchronous	synchronous	ADJ
ajst-7347	17	80	prediction	prediction	NOUN
ajst-7347	17	81	of	of	ADP
ajst-7347	17	82	pm2.5	pm2.5	NOUN
ajst-7347	17	83	;	;	PUNCT
ajst-7347	17	84	most	most	ADJ
ajst-7347	17	85	of	of	ADP
ajst-7347	17	86	these	these	DET
ajst-7347	17	87	traditional	traditional	ADJ
ajst-7347	17	88	pm2.5	pm2.5	ADJ
ajst-7347	17	89	concentration	concentration	NOUN
ajst-7347	17	90	modeling	modeling	NOUN
ajst-7347	17	91	methods	method	NOUN
ajst-7347	17	92	require	require	VERB
ajst-7347	17	93	complex	complex	ADJ
ajst-7347	17	94	calculations	calculation	NOUN
ajst-7347	17	95	and	and	CCONJ
ajst-7347	17	96	have	have	VERB
ajst-7347	17	97	a	a	DET
ajst-7347	17	98	problem	problem	NOUN
ajst-7347	17	99	of	of	ADP
ajst-7347	17	100	low	low	ADJ
ajst-7347	17	101	accuracy	accuracy	NOUN
ajst-7347	17	102	.	.	PUNCT
ajst-7347	18	1	in	in	ADP
ajst-7347	18	2	recent	recent	ADJ
ajst-7347	18	3	years	year	NOUN
ajst-7347	18	4	,	,	PUNCT
ajst-7347	18	5	machine	machine	NOUN
ajst-7347	18	6	learning	learning	NOUN
ajst-7347	18	7	algorithms	algorithm	NOUN
ajst-7347	18	8	have	have	AUX
ajst-7347	18	9	become	become	VERB
ajst-7347	18	10	a	a	DET
ajst-7347	18	11	research	research	NOUN
ajst-7347	18	12	craze	craze	NOUN
ajst-7347	18	13	.	.	PUNCT
ajst-7347	19	1	ama	ama	PROPN
ajst-7347	19	2	et	et	NOUN
ajst-7347	19	3	al[12]established	al[12]established	NUM
ajst-7347	19	4	two	two	NUM
ajst-7347	19	5	different	different	ADJ
ajst-7347	19	6	models	model	NOUN
ajst-7347	19	7	,	,	PUNCT
ajst-7347	19	8	support	support	NOUN
ajst-7347	19	9	vector	vector	NOUN
ajst-7347	19	10	machine	machine	NOUN
ajst-7347	19	11	svm	svm	NOUN
ajst-7347	19	12	and	and	CCONJ
ajst-7347	19	13	artificial	artificial	ADJ
ajst-7347	19	14	neural	neural	ADJ
ajst-7347	19	15	network	network	NOUN
ajst-7347	19	16	ann	ann	PROPN
ajst-7347	19	17	.	.	PUNCT
ajst-7347	20	1	the	the	DET
ajst-7347	20	2	results	result	NOUN
ajst-7347	20	3	show	show	VERB
ajst-7347	20	4	that	that	SCONJ
ajst-7347	20	5	the	the	DET
ajst-7347	20	6	prediction	prediction	NOUN
ajst-7347	20	7	accuracy	accuracy	NOUN
ajst-7347	20	8	of	of	ADP
ajst-7347	20	9	neural	neural	ADJ
ajst-7347	20	10	network	network	NOUN
ajst-7347	20	11	for	for	ADP
ajst-7347	20	12	pm2.5	pm2.5	NUM
ajst-7347	20	13	is	be	AUX
ajst-7347	20	14	better	well	ADJ
ajst-7347	20	15	than	than	ADP
ajst-7347	20	16	that	that	PRON
ajst-7347	20	17	of	of	ADP
ajst-7347	20	18	support	support	NOUN
ajst-7347	20	19	vector	vector	NOUN
ajst-7347	20	20	machine	machine	NOUN
ajst-7347	20	21	;	;	PUNCT
ajst-7347	20	22	chen	chen	PROPN
ajst-7347	20	23	et	et	PROPN
ajst-7347	20	24	al[13	al[13	PROPN
ajst-7347	20	25	]	]	PUNCT
ajst-7347	20	26	established	establish	VERB
ajst-7347	20	27	a	a	DET
ajst-7347	20	28	random	random	ADJ
ajst-7347	20	29	forest	forest	NOUN
ajst-7347	20	30	model	model	NOUN
ajst-7347	20	31	and	and	CCONJ
ajst-7347	20	32	two	two	NUM
ajst-7347	20	33	traditional	traditional	ADJ
ajst-7347	20	34	regression	regression	NOUN
ajst-7347	20	35	models	model	NOUN
ajst-7347	20	36	to	to	PART
ajst-7347	20	37	estimate	estimate	VERB
ajst-7347	20	38	ground	ground	NOUN
ajst-7347	20	39	pm2.5	pm2.5	ADJ
ajst-7347	20	40	concentrations	concentration	NOUN
ajst-7347	20	41	.	.	PUNCT
ajst-7347	21	1	the	the	DET
ajst-7347	21	2	results	result	NOUN
ajst-7347	21	3	show	show	VERB
ajst-7347	21	4	that	that	SCONJ
ajst-7347	21	5	the	the	DET
ajst-7347	21	6	prediction	prediction	NOUN
ajst-7347	21	7	accuracy	accuracy	NOUN
ajst-7347	21	8	of	of	ADP
ajst-7347	21	9	the	the	DET
ajst-7347	21	10	random	random	ADJ
ajst-7347	21	11	forest	forest	NOUN
ajst-7347	21	12	model	model	NOUN
ajst-7347	21	13	is	be	AUX
ajst-7347	21	14	significantly	significantly	ADV
ajst-7347	21	15	higher	high	ADJ
ajst-7347	21	16	than	than	ADP
ajst-7347	21	17	that	that	PRON
ajst-7347	21	18	of	of	ADP
ajst-7347	21	19	the	the	DET
ajst-7347	21	20	two	two	NUM
ajst-7347	21	21	traditional	traditional	ADJ
ajst-7347	21	22	regression	regression	NOUN
ajst-7347	21	23	models	model	NOUN
ajst-7347	21	24	;	;	PUNCT
ajst-7347	21	25	joharestani	joharestani	PROPN
ajst-7347	21	26	et	et	PROPN
ajst-7347	21	27	al[14]studied	al[14]studie	VERB
ajst-7347	21	28	the	the	DET
ajst-7347	21	29	importance	importance	NOUN
ajst-7347	21	30	of	of	ADP
ajst-7347	21	31	the	the	DET
ajst-7347	21	32	characteristics	characteristic	NOUN
ajst-7347	21	33	of	of	ADP
ajst-7347	21	34	pm2.5	pm2.5	DET
ajst-7347	21	35	prediction	prediction	NOUN
ajst-7347	21	36	in	in	ADP
ajst-7347	21	37	urban	urban	ADJ
ajst-7347	21	38	tehran	tehran	PROPN
ajst-7347	21	39	,	,	PUNCT
ajst-7347	21	40	and	and	CCONJ
ajst-7347	21	41	constructed	construct	VERB
ajst-7347	21	42	random	random	ADJ
ajst-7347	21	43	forests	forest	NOUN
ajst-7347	21	44	,	,	PUNCT
ajst-7347	21	45	limit	limit	VERB
ajst-7347	21	46	gradient	gradient	ADJ
ajst-7347	21	47	lifting	lifting	NOUN
ajst-7347	21	48	,	,	PUNCT
ajst-7347	21	49	and	and	CCONJ
ajst-7347	21	50	deep	deep	ADJ
ajst-7347	21	51	learning	learning	NOUN
ajst-7347	21	52	machine	machine	NOUN
ajst-7347	21	53	learning	learning	NOUN
ajst-7347	21	54	methods	method	NOUN
ajst-7347	21	55	to	to	PART
ajst-7347	21	56	predict	predict	VERB
ajst-7347	21	57	it	it	PRON
ajst-7347	21	58	.	.	PUNCT
ajst-7347	22	1	liang	liang	PROPN
ajst-7347	22	2	xiguan[15	xiguan[15	PROPN
ajst-7347	22	3	]	]	PROPN
ajst-7347	22	4	,	,	PUNCT
ajst-7347	22	5	kong	kong	PROPN
ajst-7347	22	6	yu	yu	PROPN
ajst-7347	22	7	,	,	PUNCT
ajst-7347	22	8	et	et	PROPN
ajst-7347	22	9	al[16	al[16	NOUN
ajst-7347	22	10	]	]	PUNCT
ajst-7347	22	11	established	establish	VERB
ajst-7347	22	12	a	a	DET
ajst-7347	22	13	tree	tree	NOUN
ajst-7347	22	14	model	model	NOUN
ajst-7347	22	15	to	to	PART
ajst-7347	22	16	predict	predict	VERB
ajst-7347	22	17	pm2.5	pm2.5	NOUN
ajst-7347	22	18	,	,	PUNCT
ajst-7347	22	19	and	and	CCONJ
ajst-7347	22	20	obtained	obtain	VERB
ajst-7347	22	21	its	its	PRON
ajst-7347	22	22	strong	strong	ADJ
ajst-7347	22	23	stability	stability	NOUN
ajst-7347	22	24	compared	compare	VERB
ajst-7347	22	25	to	to	ADP
ajst-7347	22	26	other	other	ADJ
ajst-7347	22	27	model	model	NOUN
ajst-7347	22	28	tree	tree	NOUN
ajst-7347	22	29	models	model	NOUN
ajst-7347	22	30	.	.	PUNCT
ajst-7347	23	1	these	these	DET
ajst-7347	23	2	machine	machine	NOUN
ajst-7347	23	3	learning	learning	NOUN
ajst-7347	23	4	models	model	NOUN
ajst-7347	23	5	have	have	AUX
ajst-7347	23	6	achieved	achieve	VERB
ajst-7347	23	7	good	good	ADJ
ajst-7347	23	8	results	result	NOUN
ajst-7347	23	9	in	in	ADP
ajst-7347	23	10	the	the	DET
ajst-7347	23	11	prediction	prediction	NOUN
ajst-7347	23	12	of	of	ADP
ajst-7347	23	13	pm2.5	pm2.5	DET
ajst-7347	23	14	concentration	concentration	NOUN
ajst-7347	23	15	.	.	PUNCT
ajst-7347	24	1	however	however	ADV
ajst-7347	24	2	,	,	PUNCT
ajst-7347	24	3	for	for	ADP
ajst-7347	24	4	the	the	DET
ajst-7347	24	5	single	single	ADJ
ajst-7347	24	6	model	model	NOUN
ajst-7347	24	7	is	be	AUX
ajst-7347	24	8	not	not	PART
ajst-7347	24	9	accurate	accurate	ADJ
ajst-7347	24	10	in	in	ADP
ajst-7347	24	11	the	the	DET
ajst-7347	24	12	prediction	prediction	NOUN
ajst-7347	24	13	of	of	ADP
ajst-7347	24	14	pm2.5	pm2.5	DET
ajst-7347	24	15	concentration	concentration	NOUN
ajst-7347	24	16	,	,	PUNCT
ajst-7347	24	17	and	and	CCONJ
ajst-7347	24	18	can	can	AUX
ajst-7347	24	19	not	not	PART
ajst-7347	24	20	better	well	ADV
ajst-7347	24	21	capture	capture	VERB
ajst-7347	24	22	the	the	DET
ajst-7347	24	23	characteristics	characteristic	NOUN
ajst-7347	24	24	of	of	ADP
ajst-7347	24	25	multidimensional	multidimensional	ADJ
ajst-7347	24	26	data	datum	NOUN
ajst-7347	24	27	changes	change	NOUN
ajst-7347	24	28	,	,	PUNCT
ajst-7347	24	29	the	the	DET
ajst-7347	24	30	combined	combined	ADJ
ajst-7347	24	31	model	model	NOUN
ajst-7347	24	32	can	can	AUX
ajst-7347	24	33	further	far	ADV
ajst-7347	24	34	improve	improve	VERB
ajst-7347	24	35	the	the	DET
ajst-7347	24	36	prediction	prediction	NOUN
ajst-7347	24	37	accuracy	accuracy	NOUN
ajst-7347	24	38	.	.	PUNCT
ajst-7347	25	1	for	for	ADP
ajst-7347	25	2	example	example	NOUN
ajst-7347	25	3	,	,	PUNCT
ajst-7347	25	4	ding	de	VERB
ajst-7347	25	5	et	et	NOUN
ajst-7347	25	6	al[17	al[17	PROPN
ajst-7347	25	7	]	]	PUNCT
ajst-7347	25	8	combined	combine	VERB
ajst-7347	25	9	convolutional	convolutional	ADJ
ajst-7347	25	10	neural	neural	ADJ
ajst-7347	25	11	networks	network	NOUN
ajst-7347	25	12	(	(	PUNCT
ajst-7347	25	13	cnn	cnn	PROPN
ajst-7347	25	14	)	)	PUNCT
ajst-7347	25	15	with	with	ADP
ajst-7347	25	16	short	short	ADJ
ajst-7347	25	17	-	-	PUNCT
ajst-7347	25	18	term	term	NOUN
ajst-7347	25	19	and	and	CCONJ
ajst-7347	25	20	short	short	ADJ
ajst-7347	25	21	-	-	PUNCT
ajst-7347	25	22	term	term	NOUN
ajst-7347	25	23	memory	memory	NOUN
ajst-7347	25	24	(	(	PUNCT
ajst-7347	25	25	lstm	lstm	NOUN
ajst-7347	25	26	)	)	PUNCT
ajst-7347	25	27	to	to	PART
ajst-7347	25	28	achieve	achieve	VERB
ajst-7347	25	29	better	well	ADJ
ajst-7347	25	30	prediction	prediction	NOUN
ajst-7347	25	31	results	result	NOUN
ajst-7347	25	32	than	than	ADP
ajst-7347	25	33	other	other	ADJ
ajst-7347	25	34	machine	machine	NOUN
ajst-7347	25	35	learning	learning	NOUN
ajst-7347	25	36	methods	method	NOUN
ajst-7347	25	37	;	;	PUNCT
ajst-7347	25	38	qian	qian	PROPN
ajst-7347	25	39	et	et	NOUN
ajst-7347	25	40	al[18	al[18	PROPN
ajst-7347	25	41	]	]	PUNCT
ajst-7347	25	42	used	use	VERB
ajst-7347	25	43	a	a	DET
ajst-7347	25	44	generalized	generalize	VERB
ajst-7347	25	45	additive	additive	ADJ
ajst-7347	25	46	model	model	NOUN
ajst-7347	25	47	to	to	PART
ajst-7347	25	48	combine	combine	VERB
ajst-7347	25	49	neural	neural	ADJ
ajst-7347	25	50	networks	network	NOUN
ajst-7347	25	51	,	,	PUNCT
ajst-7347	25	52	random	random	ADJ
ajst-7347	25	53	forests	forest	NOUN
ajst-7347	25	54	,	,	PUNCT
ajst-7347	25	55	and	and	CCONJ
ajst-7347	25	56	gradient	gradient	NOUN
ajst-7347	25	57	enhanced	enhance	VERB
ajst-7347	25	58	pm2.5	pm2.5	PROPN
ajst-7347	25	59	estimates	estimate	NOUN
ajst-7347	25	60	;	;	PUNCT
ajst-7347	25	61	niu	niu	PROPN
ajst-7347	25	62	et	et	NOUN
ajst-7347	25	63	al[19	al[19	PROPN
ajst-7347	25	64	]	]	PUNCT
ajst-7347	25	65	used	use	VERB
ajst-7347	25	66	empirical	empirical	ADJ
ajst-7347	25	67	mode	mode	NOUN
ajst-7347	25	68	decomposition	decomposition	NOUN
ajst-7347	25	69	method	method	NOUN
ajst-7347	25	70	,	,	PUNCT
ajst-7347	25	71	gray	gray	ADJ
ajst-7347	25	72	wolf	wolf	PROPN
ajst-7347	25	73	optimization	optimization	NOUN
ajst-7347	25	74	algorithm	algorithm	NOUN
ajst-7347	25	75	,	,	PUNCT
ajst-7347	25	76	and	and	CCONJ
ajst-7347	25	77	svr	svr	PROPN
ajst-7347	25	78	model	model	NOUN
ajst-7347	25	79	to	to	PART
ajst-7347	25	80	achieve	achieve	VERB
ajst-7347	25	81	high	high	ADJ
ajst-7347	25	82	-	-	PUNCT
ajst-7347	25	83	precision	precision	NOUN
ajst-7347	25	84	prediction	prediction	NOUN
ajst-7347	25	85	of	of	ADP
ajst-7347	25	86	pm2.5	pm2.5	PROPN
ajst-7347	25	87	.	.	PUNCT
ajst-7347	26	1	these	these	DET
ajst-7347	26	2	experiments	experiment	NOUN
ajst-7347	26	3	all	all	PRON
ajst-7347	26	4	use	use	VERB
ajst-7347	26	5	a	a	DET
ajst-7347	26	6	single	single	ADJ
ajst-7347	26	7	model	model	NOUN
ajst-7347	26	8	for	for	ADP
ajst-7347	26	9	comparative	comparative	ADJ
ajst-7347	26	10	verification	verification	NOUN
ajst-7347	26	11	.	.	PUNCT
ajst-7347	27	1	therefore	therefore	ADV
ajst-7347	27	2	,	,	PUNCT
ajst-7347	27	3	it	it	PRON
ajst-7347	27	4	is	be	AUX
ajst-7347	27	5	proved	prove	VERB
ajst-7347	27	6	that	that	SCONJ
ajst-7347	27	7	the	the	DET
ajst-7347	27	8	combined	combined	ADJ
ajst-7347	27	9	model	model	NOUN
ajst-7347	27	10	can	can	AUX
ajst-7347	27	11	improve	improve	VERB
ajst-7347	27	12	the	the	DET
ajst-7347	27	13	prediction	prediction	NOUN
ajst-7347	27	14	accuracy	accuracy	NOUN
ajst-7347	27	15	to	to	ADP
ajst-7347	27	16	a	a	DET
ajst-7347	27	17	certain	certain	ADJ
ajst-7347	27	18	extent	extent	NOUN
ajst-7347	27	19	.	.	PUNCT
ajst-7347	28	1	in	in	ADP
ajst-7347	28	2	this	this	DET
ajst-7347	28	3	paper	paper	NOUN
ajst-7347	28	4	,	,	PUNCT
ajst-7347	28	5	a	a	DET
ajst-7347	28	6	cnn	cnn	PROPN
ajst-7347	28	7	-	-	PUNCT
ajst-7347	28	8	bigru	bigru	PROPN
ajst-7347	28	9	model	model	NOUN
ajst-7347	28	10	was	be	AUX
ajst-7347	28	11	established	establish	VERB
ajst-7347	28	12	to	to	PART
ajst-7347	28	13	analyze	analyze	VERB
ajst-7347	28	14	and	and	CCONJ
ajst-7347	28	15	predict	predict	VERB
ajst-7347	28	16	pm2.5	pm2.5	PROPN
ajst-7347	28	17	.	.	PUNCT
ajst-7347	29	1	its	its	PRON
ajst-7347	29	2	unique	unique	ADJ
ajst-7347	29	3	convolution	convolution	NOUN
ajst-7347	29	4	operation	operation	NOUN
ajst-7347	29	5	was	be	AUX
ajst-7347	29	6	used	use	VERB
ajst-7347	29	7	to	to	PART
ajst-7347	29	8	extract	extract	VERB
ajst-7347	29	9	features	feature	NOUN
ajst-7347	29	10	from	from	ADP
ajst-7347	29	11	one	one	NUM
ajst-7347	29	12	-	-	PUNCT
ajst-7347	29	13	dimensional	dimensional	ADJ
ajst-7347	29	14	data	datum	NOUN
ajst-7347	29	15	.	.	PUNCT
ajst-7347	30	1	at	at	ADP
ajst-7347	30	2	the	the	DET
ajst-7347	30	3	same	same	ADJ
ajst-7347	30	4	time	time	NOUN
ajst-7347	30	5	,	,	PUNCT
ajst-7347	30	6	the	the	DET
ajst-7347	30	7	circular	circular	ADJ
ajst-7347	30	8	neural	neural	ADJ
ajst-7347	30	9	network	network	NOUN
ajst-7347	30	10	bigru	bigru	VERB
ajst-7347	30	11	with	with	ADP
ajst-7347	30	12	bidirectional	bidirectional	ADJ
ajst-7347	30	13	transmission	transmission	NOUN
ajst-7347	30	14	function	function	NOUN
ajst-7347	30	15	was	be	AUX
ajst-7347	30	16	combined	combine	VERB
ajst-7347	30	17	to	to	PART
ajst-7347	30	18	model	model	VERB
ajst-7347	30	19	and	and	CCONJ
ajst-7347	30	20	predict	predict	VERB
ajst-7347	30	21	2	2	NUM
ajst-7347	30	22	pm2.5	pm2.5	PRON
ajst-7347	30	23	concentration	concentration	NOUN
ajst-7347	30	24	based	base	VERB
ajst-7347	30	25	on	on	ADP
ajst-7347	30	26	the	the	DET
ajst-7347	30	27	functional	functional	ADJ
ajst-7347	30	28	advantages	advantage	NOUN
ajst-7347	30	29	of	of	ADP
ajst-7347	30	30	both	both	DET
ajst-7347	30	31	sides	side	NOUN
ajst-7347	30	32	.	.	PUNCT
ajst-7347	31	1	2	2	X
ajst-7347	31	2	.	.	NUM
ajst-7347	31	3	proposed	propose	VERB
ajst-7347	31	4	algorithm	algorithm	NOUN
ajst-7347	31	5	2.1	2.1	NUM
ajst-7347	31	6	.	.	PUNCT
ajst-7347	32	1	cnn	cnn	PROPN
ajst-7347	32	2	convolutional	convolutional	ADJ
ajst-7347	32	3	neural	neural	ADJ
ajst-7347	32	4	networks	network	NOUN
ajst-7347	32	5	is	be	AUX
ajst-7347	32	6	one	one	NUM
ajst-7347	32	7	of	of	ADP
ajst-7347	32	8	the	the	DET
ajst-7347	32	9	most	most	ADV
ajst-7347	32	10	important	important	ADJ
ajst-7347	32	11	networks	network	NOUN
ajst-7347	32	12	in	in	ADP
ajst-7347	32	13	the	the	DET
ajst-7347	32	14	field	field	NOUN
ajst-7347	32	15	of	of	ADP
ajst-7347	32	16	deep	deep	ADJ
ajst-7347	32	17	learning	learning	NOUN
ajst-7347	32	18	.	.	PUNCT
ajst-7347	33	1	it	it	PRON
ajst-7347	33	2	is	be	AUX
ajst-7347	33	3	a	a	DET
ajst-7347	33	4	deep	deep	ADJ
ajst-7347	33	5	learning	learning	NOUN
ajst-7347	33	6	model	model	NOUN
ajst-7347	33	7	or	or	CCONJ
ajst-7347	33	8	multilayer	multilayer	ADJ
ajst-7347	33	9	perceptron	perceptron	PROPN
ajst-7347	33	10	similar	similar	ADJ
ajst-7347	33	11	to	to	ADP
ajst-7347	33	12	artificial	artificial	ADJ
ajst-7347	33	13	neural	neural	ADJ
ajst-7347	33	14	networks	network	NOUN
ajst-7347	33	15	,	,	PUNCT
ajst-7347	33	16	and	and	CCONJ
ajst-7347	33	17	is	be	AUX
ajst-7347	33	18	often	often	ADV
ajst-7347	33	19	used	use	VERB
ajst-7347	33	20	in	in	ADP
ajst-7347	33	21	analyzing	analyze	VERB
ajst-7347	33	22	visual	visual	ADJ
ajst-7347	33	23	images	image	NOUN
ajst-7347	33	24	(	(	PUNCT
ajst-7347	33	25	such	such	ADJ
ajst-7347	33	26	as	as	ADP
ajst-7347	33	27	face	face	NOUN
ajst-7347	33	28	recognition	recognition	NOUN
ajst-7347	33	29	,	,	PUNCT
ajst-7347	33	30	image	image	NOUN
ajst-7347	33	31	segmentation	segmentation	NOUN
ajst-7347	33	32	,	,	PUNCT
ajst-7347	33	33	and	and	CCONJ
ajst-7347	33	34	image	image	NOUN
ajst-7347	33	35	classification	classification	NOUN
ajst-7347	33	36	)	)	PUNCT
ajst-7347	33	37	and	and	CCONJ
ajst-7347	33	38	natural	natural	ADJ
ajst-7347	33	39	language	language	NOUN
ajst-7347	33	40	processing	processing	NOUN
ajst-7347	33	41	.	.	PUNCT
ajst-7347	34	1	cnn	cnn	PROPN
ajst-7347	34	2	is	be	AUX
ajst-7347	34	3	a	a	DET
ajst-7347	34	4	feedforward	feedforward	ADJ
ajst-7347	34	5	neural	neural	ADJ
ajst-7347	34	6	network	network	NOUN
ajst-7347	34	7	that	that	PRON
ajst-7347	34	8	can	can	AUX
ajst-7347	34	9	extract	extract	VERB
ajst-7347	34	10	features	feature	NOUN
ajst-7347	34	11	from	from	ADP
ajst-7347	34	12	data	datum	NOUN
ajst-7347	34	13	,	,	PUNCT
ajst-7347	34	14	with	with	ADP
ajst-7347	34	15	local	local	ADJ
ajst-7347	34	16	connectivity	connectivity	NOUN
ajst-7347	34	17	,	,	PUNCT
ajst-7347	34	18	weight	weight	NOUN
ajst-7347	34	19	sharing	sharing	NOUN
ajst-7347	34	20	,	,	PUNCT
ajst-7347	34	21	and	and	CCONJ
ajst-7347	34	22	other	other	ADJ
ajst-7347	34	23	characteristics	characteristic	NOUN
ajst-7347	34	24	.	.	PUNCT
ajst-7347	35	1	a	a	DET
ajst-7347	35	2	convolutional	convolutional	ADJ
ajst-7347	35	3	neural	neural	ADJ
ajst-7347	35	4	network	network	NOUN
ajst-7347	35	5	is	be	AUX
ajst-7347	35	6	mainly	mainly	ADV
ajst-7347	35	7	composed	compose	VERB
ajst-7347	35	8	of	of	ADP
ajst-7347	35	9	five	five	NUM
ajst-7347	35	10	layers	layer	NOUN
ajst-7347	35	11	:	:	PUNCT
ajst-7347	35	12	data	datum	NOUN
ajst-7347	35	13	input	input	NOUN
ajst-7347	35	14	layer	layer	NOUN
ajst-7347	35	15	,	,	PUNCT
ajst-7347	35	16	convolutional	convolutional	ADJ
ajst-7347	35	17	layer	layer	NOUN
ajst-7347	35	18	,	,	PUNCT
ajst-7347	35	19	pooling	pool	VERB
ajst-7347	35	20	layer	layer	NOUN
ajst-7347	35	21	,	,	PUNCT
ajst-7347	35	22	fully	fully	ADV
ajst-7347	35	23	connected	connected	ADJ
ajst-7347	35	24	layer	layer	NOUN
ajst-7347	35	25	,	,	PUNCT
ajst-7347	35	26	and	and	CCONJ
ajst-7347	35	27	output	output	NOUN
ajst-7347	35	28	layer[20	layer[20	NOUN
ajst-7347	35	29	]	]	PUNCT
ajst-7347	35	30	.	.	PUNCT
ajst-7347	36	1	2.2	2.2	NUM
ajst-7347	36	2	.	.	PUNCT
ajst-7347	37	1	bigru	bigru	PROPN
ajst-7347	37	2	bigru	bigru	NOUN
ajst-7347	37	3	(	(	PUNCT
ajst-7347	37	4	bidirectional	bidirectional	ADJ
ajst-7347	37	5	gate	gate	PROPN
ajst-7347	37	6	recurrent	recurrent	PROPN
ajst-7347	37	7	unit	unit	NOUN
ajst-7347	37	8	)	)	PUNCT
ajst-7347	37	9	is	be	AUX
ajst-7347	37	10	an	an	DET
ajst-7347	37	11	extended	extended	ADJ
ajst-7347	37	12	model	model	NOUN
ajst-7347	37	13	of	of	ADP
ajst-7347	37	14	gru	gru	PROPN
ajst-7347	37	15	,	,	PUNCT
ajst-7347	37	16	which	which	PRON
ajst-7347	37	17	consists	consist	VERB
ajst-7347	37	18	of	of	ADP
ajst-7347	37	19	two	two	NUM
ajst-7347	37	20	gru	gru	NOUN
ajst-7347	37	21	models	model	NOUN
ajst-7347	37	22	:	:	PUNCT
ajst-7347	37	23	a	a	DET
ajst-7347	37	24	forward	forward	ADV
ajst-7347	37	25	transitive	transitive	ADJ
ajst-7347	37	26	gru	gru	NOUN
ajst-7347	37	27	model	model	NOUN
ajst-7347	37	28	that	that	PRON
ajst-7347	37	29	receives	receive	VERB
ajst-7347	37	30	forward	forward	ADP
ajst-7347	37	31	inputs	input	NOUN
ajst-7347	37	32	;	;	PUNCT
ajst-7347	37	33	the	the	DET
ajst-7347	37	34	other	other	ADJ
ajst-7347	37	35	is	be	AUX
ajst-7347	37	36	the	the	DET
ajst-7347	37	37	reverse	reverse	PROPN
ajst-7347	37	38	transmission	transmission	NOUN
ajst-7347	37	39	gru	gru	NOUN
ajst-7347	37	40	model	model	NOUN
ajst-7347	37	41	,	,	PUNCT
ajst-7347	37	42	which	which	PRON
ajst-7347	37	43	learns	learn	VERB
ajst-7347	37	44	the	the	DET
ajst-7347	37	45	reverse	reverse	ADJ
ajst-7347	37	46	input	input	NOUN
ajst-7347	37	47	of	of	ADP
ajst-7347	37	48	the	the	DET
ajst-7347	37	49	model[21,22	model[21,22	NOUN
ajst-7347	37	50	]	]	PUNCT
ajst-7347	37	51	.	.	PUNCT
ajst-7347	38	1	as	as	SCONJ
ajst-7347	38	2	the	the	DET
ajst-7347	38	3	pm2.5	pm2.5	PROPN
ajst-7347	38	4	concentration	concentration	NOUN
ajst-7347	38	5	value	value	NOUN
ajst-7347	38	6	studied	study	VERB
ajst-7347	38	7	in	in	ADP
ajst-7347	38	8	this	this	DET
ajst-7347	38	9	article	article	NOUN
ajst-7347	38	10	belongs	belong	VERB
ajst-7347	38	11	to	to	ADP
ajst-7347	38	12	time	time	NOUN
ajst-7347	38	13	series	series	PROPN
ajst-7347	38	14	data	data	PROPN
ajst-7347	38	15	,	,	PUNCT
ajst-7347	38	16	which	which	PRON
ajst-7347	38	17	is	be	AUX
ajst-7347	38	18	greatly	greatly	ADV
ajst-7347	38	19	affected	affect	VERB
ajst-7347	38	20	by	by	ADP
ajst-7347	38	21	the	the	DET
ajst-7347	38	22	chronological	chronological	ADJ
ajst-7347	38	23	order	order	NOUN
ajst-7347	38	24	,	,	PUNCT
ajst-7347	38	25	the	the	DET
ajst-7347	38	26	bigru	bigru	NOUN
ajst-7347	38	27	model	model	NOUN
ajst-7347	38	28	is	be	AUX
ajst-7347	38	29	selected	select	VERB
ajst-7347	38	30	to	to	PART
ajst-7347	38	31	model	model	VERB
ajst-7347	38	32	it	it	PRON
ajst-7347	38	33	,	,	PUNCT
ajst-7347	38	34	which	which	PRON
ajst-7347	38	35	can	can	AUX
ajst-7347	38	36	not	not	PART
ajst-7347	38	37	only	only	ADV
ajst-7347	38	38	learn	learn	VERB
ajst-7347	38	39	the	the	DET
ajst-7347	38	40	impact	impact	NOUN
ajst-7347	38	41	of	of	ADP
ajst-7347	38	42	historical	historical	ADJ
ajst-7347	38	43	information	information	NOUN
ajst-7347	38	44	on	on	ADP
ajst-7347	38	45	subsequent	subsequent	ADJ
ajst-7347	38	46	information	information	NOUN
ajst-7347	38	47	,	,	PUNCT
ajst-7347	38	48	but	but	CCONJ
ajst-7347	38	49	also	also	ADV
ajst-7347	38	50	ensure	ensure	VERB
ajst-7347	38	51	the	the	DET
ajst-7347	38	52	correlation	correlation	NOUN
ajst-7347	38	53	between	between	ADP
ajst-7347	38	54	historical	historical	ADJ
ajst-7347	38	55	information	information	NOUN
ajst-7347	38	56	and	and	CCONJ
ajst-7347	38	57	subsequent	subsequent	ADJ
ajst-7347	38	58	information	information	NOUN
ajst-7347	38	59	,	,	PUNCT
ajst-7347	38	60	thereby	thereby	ADV
ajst-7347	38	61	improving	improve	VERB
ajst-7347	38	62	the	the	DET
ajst-7347	38	63	prediction	prediction	NOUN
ajst-7347	38	64	accuracy	accuracy	NOUN
ajst-7347	38	65	of	of	ADP
ajst-7347	38	66	pm2.5	pm2.5	DET
ajst-7347	38	67	concentration	concentration	NOUN
ajst-7347	38	68	.	.	PUNCT
ajst-7347	39	1	the	the	DET
ajst-7347	39	2	structure	structure	NOUN
ajst-7347	39	3	diagram	diagram	NOUN
ajst-7347	39	4	of	of	ADP
ajst-7347	39	5	bigru	bigru	NOUN
ajst-7347	39	6	module	module	NOUN
ajst-7347	39	7	3	3	NUM
ajst-7347	39	8	is	be	AUX
ajst-7347	39	9	shown	show	VERB
ajst-7347	39	10	in	in	ADP
ajst-7347	39	11	the	the	DET
ajst-7347	39	12	figure	figure	NOUN
ajst-7347	39	13	1	1	NUM
ajst-7347	39	14	.	.	PUNCT
ajst-7347	39	15	figure	figure	NOUN
ajst-7347	39	16	1	1	NUM
ajst-7347	39	17	.	.	PUNCT
ajst-7347	40	1	bigru	bigru	PROPN
ajst-7347	40	2	model	model	NOUN
ajst-7347	40	3	structure	structure	NOUN
ajst-7347	40	4	3	3	X
ajst-7347	40	5	.	.	PUNCT
ajst-7347	40	6	experiment	experiment	NOUN
ajst-7347	40	7	3.1	3.1	NUM
ajst-7347	40	8	.	.	PUNCT
ajst-7347	41	1	feature	feature	NOUN
ajst-7347	41	2	selection	selection	NOUN
ajst-7347	41	3	pm2.5	pm2.5	ADJ
ajst-7347	41	4	concentration	concentration	NOUN
ajst-7347	41	5	is	be	AUX
ajst-7347	41	6	influenced	influence	VERB
ajst-7347	41	7	by	by	ADP
ajst-7347	41	8	various	various	ADJ
ajst-7347	41	9	factors	factor	NOUN
ajst-7347	41	10	and	and	CCONJ
ajst-7347	41	11	exhibits	exhibit	VERB
ajst-7347	41	12	a	a	DET
ajst-7347	41	13	state	state	NOUN
ajst-7347	41	14	of	of	ADP
ajst-7347	41	15	coexistence	coexistence	NOUN
ajst-7347	41	16	of	of	ADP
ajst-7347	41	17	multiple	multiple	ADJ
ajst-7347	41	18	change	change	NOUN
ajst-7347	41	19	modes[23,24	modes[23,24	NOUN
ajst-7347	41	20	]	]	PUNCT
ajst-7347	41	21	.	.	PUNCT
ajst-7347	42	1	therefore	therefore	ADV
ajst-7347	42	2	,	,	PUNCT
ajst-7347	42	3	in	in	ADP
ajst-7347	42	4	order	order	NOUN
ajst-7347	42	5	to	to	PART
ajst-7347	42	6	build	build	VERB
ajst-7347	42	7	an	an	DET
ajst-7347	42	8	effective	effective	ADJ
ajst-7347	42	9	model	model	NOUN
ajst-7347	42	10	and	and	CCONJ
ajst-7347	42	11	improve	improve	VERB
ajst-7347	42	12	prediction	prediction	NOUN
ajst-7347	42	13	accuracy	accuracy	NOUN
ajst-7347	42	14	,	,	PUNCT
ajst-7347	42	15	it	it	PRON
ajst-7347	42	16	is	be	AUX
ajst-7347	42	17	necessary	necessary	ADJ
ajst-7347	42	18	to	to	PART
ajst-7347	42	19	establish	establish	VERB
ajst-7347	42	20	a	a	DET
ajst-7347	42	21	reasonable	reasonable	ADJ
ajst-7347	42	22	input	input	NOUN
ajst-7347	42	23	characteristic	characteristic	ADJ
ajst-7347	42	24	matrix	matrix	NOUN
ajst-7347	42	25	.	.	PUNCT
ajst-7347	43	1	a	a	DET
ajst-7347	43	2	good	good	ADJ
ajst-7347	43	3	representation	representation	NOUN
ajst-7347	43	4	of	of	ADP
ajst-7347	43	5	the	the	DET
ajst-7347	43	6	relationship	relationship	NOUN
ajst-7347	43	7	between	between	ADP
ajst-7347	43	8	input	input	NOUN
ajst-7347	43	9	data	datum	NOUN
ajst-7347	43	10	and	and	CCONJ
ajst-7347	43	11	pm2.5	pm2.5	PROPN
ajst-7347	43	12	is	be	AUX
ajst-7347	43	13	a	a	DET
ajst-7347	43	14	key	key	ADJ
ajst-7347	43	15	part	part	NOUN
ajst-7347	43	16	of	of	ADP
ajst-7347	43	17	building	build	VERB
ajst-7347	43	18	a	a	DET
ajst-7347	43	19	prediction	prediction	NOUN
ajst-7347	43	20	model	model	NOUN
ajst-7347	43	21	.	.	PUNCT
ajst-7347	44	1	feature	feature	NOUN
ajst-7347	44	2	selection	selection	NOUN
ajst-7347	44	3	can	can	AUX
ajst-7347	44	4	effectively	effectively	ADV
ajst-7347	44	5	improve	improve	VERB
ajst-7347	44	6	the	the	DET
ajst-7347	44	7	accuracy	accuracy	NOUN
ajst-7347	44	8	of	of	ADP
ajst-7347	44	9	the	the	DET
ajst-7347	44	10	model	model	NOUN
ajst-7347	44	11	,	,	PUNCT
ajst-7347	44	12	reduce	reduce	VERB
ajst-7347	44	13	runtime	runtime	NOUN
ajst-7347	44	14	,	,	PUNCT
ajst-7347	44	15	and	and	CCONJ
ajst-7347	44	16	reduce	reduce	VERB
ajst-7347	44	17	the	the	DET
ajst-7347	44	18	risk	risk	NOUN
ajst-7347	44	19	of	of	ADP
ajst-7347	44	20	overfitting	overfitte	VERB
ajst-7347	44	21	by	by	ADP
ajst-7347	44	22	reducing	reduce	VERB
ajst-7347	44	23	redundant	redundant	ADJ
ajst-7347	44	24	and	and	CCONJ
ajst-7347	44	25	unrelated	unrelated	ADJ
ajst-7347	44	26	features	feature	NOUN
ajst-7347	44	27	.	.	PUNCT
ajst-7347	45	1	from	from	ADP
ajst-7347	45	2	the	the	DET
ajst-7347	45	3	correlation	correlation	NOUN
ajst-7347	45	4	analysis	analysis	NOUN
ajst-7347	45	5	in	in	ADP
ajst-7347	45	6	chapter	chapter	NOUN
ajst-7347	45	7	3	3	NUM
ajst-7347	45	8	,	,	PUNCT
ajst-7347	45	9	it	it	PRON
ajst-7347	45	10	can	can	AUX
ajst-7347	45	11	be	be	AUX
ajst-7347	45	12	seen	see	VERB
ajst-7347	45	13	that	that	SCONJ
ajst-7347	45	14	most	most	ADJ
ajst-7347	45	15	pollution	pollution	NOUN
ajst-7347	45	16	factors	factor	NOUN
ajst-7347	45	17	and	and	CCONJ
ajst-7347	45	18	meteorological	meteorological	ADJ
ajst-7347	45	19	factors	factor	NOUN
ajst-7347	45	20	are	be	AUX
ajst-7347	45	21	significantly	significantly	ADV
ajst-7347	45	22	related	relate	VERB
ajst-7347	45	23	to	to	ADP
ajst-7347	45	24	the	the	DET
ajst-7347	45	25	increase	increase	NOUN
ajst-7347	45	26	and	and	CCONJ
ajst-7347	45	27	decrease	decrease	NOUN
ajst-7347	45	28	of	of	ADP
ajst-7347	45	29	pm2.5	pm2.5	DET
ajst-7347	45	30	concentration	concentration	NOUN
ajst-7347	45	31	.	.	PUNCT
ajst-7347	46	1	in	in	ADP
ajst-7347	46	2	this	this	DET
ajst-7347	46	3	paper	paper	NOUN
ajst-7347	46	4	,	,	PUNCT
ajst-7347	46	5	spearman	spearman	NOUN
ajst-7347	46	6	correlation	correlation	NOUN
ajst-7347	46	7	coefficient	coefficient	NOUN
ajst-7347	46	8	is	be	AUX
ajst-7347	46	9	selected	select	VERB
ajst-7347	46	10	for	for	ADP
ajst-7347	46	11	feature	feature	NOUN
ajst-7347	46	12	selection[25	selection[25	NOUN
ajst-7347	46	13	]	]	PUNCT
ajst-7347	46	14	.	.	PUNCT
ajst-7347	47	1	take	take	VERB
ajst-7347	47	2	the	the	DET
ajst-7347	47	3	monitoring	monitoring	NOUN
ajst-7347	47	4	point	point	NOUN
ajst-7347	47	5	in	in	ADP
ajst-7347	47	6	xindian	xindian	ADJ
ajst-7347	47	7	district	district	NOUN
ajst-7347	47	8	as	as	ADP
ajst-7347	47	9	an	an	DET
ajst-7347	47	10	example	example	NOUN
ajst-7347	47	11	.	.	PUNCT
ajst-7347	48	1	figure	figure	NOUN
ajst-7347	48	2	2	2	NUM
ajst-7347	48	3	and	and	CCONJ
ajst-7347	48	4	figure	figure	VERB
ajst-7347	48	5	3	3	NUM
ajst-7347	48	6	show	show	VERB
ajst-7347	48	7	the	the	DET
ajst-7347	48	8	correlation	correlation	NOUN
ajst-7347	48	9	between	between	ADP
ajst-7347	48	10	pm2.5	pm2.5	ADJ
ajst-7347	48	11	and	and	CCONJ
ajst-7347	48	12	pollutants	pollutant	NOUN
ajst-7347	48	13	and	and	CCONJ
ajst-7347	48	14	the	the	DET
ajst-7347	48	15	correlation	correlation	NOUN
ajst-7347	48	16	between	between	ADP
ajst-7347	48	17	meteorological	meteorological	ADJ
ajst-7347	48	18	factors	factor	NOUN
ajst-7347	48	19	.	.	PUNCT
ajst-7347	49	1	you	you	PRON
ajst-7347	49	2	can	can	AUX
ajst-7347	49	3	clearly	clearly	ADV
ajst-7347	49	4	see	see	VERB
ajst-7347	49	5	the	the	DET
ajst-7347	49	6	spearman	spearman	NOUN
ajst-7347	49	7	correlation	correlation	NOUN
ajst-7347	49	8	between	between	ADP
ajst-7347	49	9	each	each	DET
ajst-7347	49	10	feature	feature	NOUN
ajst-7347	49	11	and	and	CCONJ
ajst-7347	49	12	pm2.5	pm2.5	PROPN
ajst-7347	49	13	.	.	PUNCT
ajst-7347	50	1	the	the	DET
ajst-7347	50	2	darker	dark	ADJ
ajst-7347	50	3	the	the	DET
ajst-7347	50	4	color	color	NOUN
ajst-7347	50	5	,	,	PUNCT
ajst-7347	50	6	the	the	PRON
ajst-7347	50	7	stronger	strong	ADJ
ajst-7347	50	8	the	the	DET
ajst-7347	50	9	correlation	correlation	NOUN
ajst-7347	50	10	.	.	PUNCT
ajst-7347	51	1	the	the	DET
ajst-7347	51	2	correlation	correlation	NOUN
ajst-7347	51	3	between	between	ADP
ajst-7347	51	4	pm10	pm10	PROPN
ajst-7347	51	5	and	and	CCONJ
ajst-7347	51	6	pm2.5	pm2.5	PROPN
ajst-7347	51	7	in	in	ADP
ajst-7347	51	8	pollutants	pollutant	NOUN
ajst-7347	51	9	reaches	reach	VERB
ajst-7347	51	10	0.83	0.83	NUM
ajst-7347	51	11	,	,	PUNCT
ajst-7347	51	12	with	with	ADP
ajst-7347	51	13	a	a	DET
ajst-7347	51	14	strong	strong	ADJ
ajst-7347	51	15	correlation	correlation	NOUN
ajst-7347	51	16	,	,	PUNCT
ajst-7347	51	17	followed	follow	VERB
ajst-7347	51	18	by	by	ADP
ajst-7347	51	19	co	co	NOUN
ajst-7347	51	20	,	,	PUNCT
ajst-7347	51	21	while	while	SCONJ
ajst-7347	51	22	the	the	DET
ajst-7347	51	23	correlation	correlation	NOUN
ajst-7347	51	24	between	between	ADP
ajst-7347	51	25	meteorological	meteorological	ADJ
ajst-7347	51	26	factors	factor	NOUN
ajst-7347	51	27	and	and	CCONJ
ajst-7347	51	28	pm2.5	pm2.5	PROPN
ajst-7347	51	29	is	be	AUX
ajst-7347	51	30	weak	weak	ADJ
ajst-7347	51	31	,	,	PUNCT
ajst-7347	51	32	and	and	CCONJ
ajst-7347	51	33	most	most	ADJ
ajst-7347	51	34	characteristic	characteristic	ADJ
ajst-7347	51	35	variables	variable	NOUN
ajst-7347	51	36	are	be	AUX
ajst-7347	51	37	inversely	inversely	ADV
ajst-7347	51	38	proportional	proportional	ADJ
ajst-7347	51	39	to	to	ADP
ajst-7347	51	40	them	they	PRON
ajst-7347	51	41	,	,	PUNCT
ajst-7347	51	42	such	such	ADJ
ajst-7347	51	43	as	as	ADP
ajst-7347	51	44	rh	rh	PROPN
ajst-7347	51	45	(	(	PUNCT
ajst-7347	51	46	relative	relative	ADJ
ajst-7347	51	47	humidity	humidity	NOUN
ajst-7347	51	48	)	)	PUNCT
ajst-7347	51	49	and	and	CCONJ
ajst-7347	51	50	amb_temp	amb_temp	NOUN
ajst-7347	51	51	(	(	PUNCT
ajst-7347	51	52	temperature	temperature	NOUN
ajst-7347	51	53	)	)	PUNCT
ajst-7347	51	54	,	,	PUNCT
ajst-7347	51	55	etc	etc	X
ajst-7347	51	56	.	.	X
ajst-7347	51	57	therefore	therefore	ADV
ajst-7347	51	58	,	,	PUNCT
ajst-7347	51	59	pm10	pm10	PROPN
ajst-7347	51	60	,	,	PUNCT
ajst-7347	51	61	co	co	NOUN
ajst-7347	51	62	,	,	PUNCT
ajst-7347	51	63	no2	no2	NOUN
ajst-7347	51	64	,	,	PUNCT
ajst-7347	51	65	and	and	CCONJ
ajst-7347	51	66	so2	so2	NOUN
ajst-7347	51	67	with	with	ADP
ajst-7347	51	68	strong	strong	ADJ
ajst-7347	51	69	correlation	correlation	NOUN
ajst-7347	51	70	are	be	AUX
ajst-7347	51	71	selected	select	VERB
ajst-7347	51	72	as	as	ADP
ajst-7347	51	73	pollutant	pollutant	ADJ
ajst-7347	51	74	factors	factor	NOUN
ajst-7347	51	75	,	,	PUNCT
ajst-7347	51	76	while	while	SCONJ
ajst-7347	51	77	rh	rh	PROPN
ajst-7347	51	78	and	and	CCONJ
ajst-7347	51	79	ws_hr	ws_hr	PROPN
ajst-7347	51	80	are	be	AUX
ajst-7347	51	81	selected	select	VERB
ajst-7347	51	82	as	as	ADP
ajst-7347	51	83	meteorological	meteorological	ADJ
ajst-7347	51	84	factors	factor	NOUN
ajst-7347	51	85	,	,	PUNCT
ajst-7347	51	86	which	which	PRON
ajst-7347	51	87	are	be	AUX
ajst-7347	51	88	used	use	VERB
ajst-7347	51	89	as	as	ADP
ajst-7347	51	90	a	a	DET
ajst-7347	51	91	correlation	correlation	NOUN
ajst-7347	51	92	factor	factor	NOUN
ajst-7347	51	93	auxiliary	auxiliary	ADJ
ajst-7347	51	94	variable	variable	NOUN
ajst-7347	51	95	.	.	PUNCT
ajst-7347	52	1	figure	figure	NOUN
ajst-7347	52	2	2	2	NUM
ajst-7347	52	3	.	.	PUNCT
ajst-7347	53	1	the	the	DET
ajst-7347	53	2	correlation	correlation	NOUN
ajst-7347	53	3	between	between	ADP
ajst-7347	53	4	pm2.5	pm2.5	PROPN
ajst-7347	53	5	and	and	CCONJ
ajst-7347	53	6	other	other	ADJ
ajst-7347	53	7	pollutants	pollutant	NOUN
ajst-7347	53	8	3	3	NUM
ajst-7347	53	9	figure	figure	NOUN
ajst-7347	53	10	3	3	NUM
ajst-7347	53	11	.	.	PUNCT
ajst-7347	54	1	the	the	DET
ajst-7347	54	2	correlation	correlation	NOUN
ajst-7347	54	3	between	between	ADP
ajst-7347	54	4	pm2.5	pm2.5	ADJ
ajst-7347	54	5	and	and	CCONJ
ajst-7347	54	6	meteorological	meteorological	ADJ
ajst-7347	54	7	factors	factor	NOUN
ajst-7347	54	8	3.2	3.2	NUM
ajst-7347	54	9	.	.	PUNCT
ajst-7347	55	1	data	datum	NOUN
ajst-7347	55	2	preprocessing	preprocesse	VERB
ajst-7347	55	3	missing	miss	VERB
ajst-7347	55	4	data	datum	NOUN
ajst-7347	55	5	will	will	AUX
ajst-7347	55	6	lead	lead	VERB
ajst-7347	55	7	to	to	ADP
ajst-7347	55	8	the	the	DET
ajst-7347	55	9	reduction	reduction	NOUN
ajst-7347	55	10	of	of	ADP
ajst-7347	55	11	sample	sample	NOUN
ajst-7347	55	12	information	information	NOUN
ajst-7347	55	13	,	,	PUNCT
ajst-7347	55	14	deviation	deviation	NOUN
ajst-7347	55	15	of	of	ADP
ajst-7347	55	16	data	datum	NOUN
ajst-7347	55	17	analysis	analysis	NOUN
ajst-7347	55	18	results	result	NOUN
ajst-7347	55	19	,	,	PUNCT
ajst-7347	55	20	and	and	CCONJ
ajst-7347	55	21	“	"	PUNCT
ajst-7347	55	22	pseudo	pseudo	NOUN
ajst-7347	55	23	regression	regression	NOUN
ajst-7347	55	24	”	"	PUNCT
ajst-7347	55	25	of	of	ADP
ajst-7347	55	26	data	datum	NOUN
ajst-7347	55	27	outliers	outlier	NOUN
ajst-7347	55	28	.	.	PUNCT
ajst-7347	56	1	to	to	PART
ajst-7347	56	2	solve	solve	VERB
ajst-7347	56	3	these	these	DET
ajst-7347	56	4	problems	problem	NOUN
ajst-7347	56	5	,	,	PUNCT
ajst-7347	56	6	this	this	DET
ajst-7347	56	7	paper	paper	NOUN
ajst-7347	56	8	adopts	adopt	VERB
ajst-7347	56	9	the	the	DET
ajst-7347	56	10	following	follow	VERB
ajst-7347	56	11	measures	measure	NOUN
ajst-7347	56	12	:	:	PUNCT
ajst-7347	56	13	data	datum	NOUN
ajst-7347	56	14	missing	missing	ADJ
ajst-7347	56	15	is	be	AUX
ajst-7347	56	16	filled	fill	VERB
ajst-7347	56	17	with	with	ADP
ajst-7347	56	18	the	the	DET
ajst-7347	56	19	nearest	near	ADJ
ajst-7347	56	20	neighbor	neighbor	NOUN
ajst-7347	56	21	(	(	PUNCT
ajst-7347	56	22	knn	knn	PROPN
ajst-7347	56	23	)	)	PUNCT
ajst-7347	56	24	algorithm	algorithm	NOUN
ajst-7347	56	25	.	.	PUNCT
ajst-7347	57	1	the	the	DET
ajst-7347	57	2	filling	filling	NOUN
ajst-7347	57	3	idea	idea	NOUN
ajst-7347	57	4	of	of	ADP
ajst-7347	57	5	knn	knn	PROPN
ajst-7347	57	6	considers	consider	VERB
ajst-7347	57	7	the	the	DET
ajst-7347	57	8	“	"	PUNCT
ajst-7347	57	9	distance	distance	NOUN
ajst-7347	57	10	”	"	PUNCT
ajst-7347	57	11	between	between	ADP
ajst-7347	57	12	two	two	NUM
ajst-7347	57	13	samples	sample	NOUN
ajst-7347	57	14	,	,	PUNCT
ajst-7347	57	15	and	and	CCONJ
ajst-7347	57	16	selects	select	VERB
ajst-7347	57	17	the	the	DET
ajst-7347	57	18	average	average	ADJ
ajst-7347	57	19	value	value	NOUN
ajst-7347	57	20	or	or	CCONJ
ajst-7347	57	21	distance	distance	NOUN
ajst-7347	57	22	weighting	weighting	NOUN
ajst-7347	57	23	of	of	ADP
ajst-7347	57	24	the	the	DET
ajst-7347	57	25	nearest	near	ADJ
ajst-7347	57	26	several	several	ADJ
ajst-7347	57	27	observations	observation	NOUN
ajst-7347	57	28	as	as	ADP
ajst-7347	57	29	the	the	DET
ajst-7347	57	30	filling	filling	NOUN
ajst-7347	57	31	value	value	NOUN
ajst-7347	57	32	of	of	ADP
ajst-7347	57	33	the	the	DET
ajst-7347	57	34	missing	miss	VERB
ajst-7347	57	35	samples	sample	NOUN
ajst-7347	57	36	.	.	PUNCT
ajst-7347	58	1	outlier	outlier	NOUN
ajst-7347	58	2	refers	refer	VERB
ajst-7347	58	3	to	to	ADP
ajst-7347	58	4	the	the	DET
ajst-7347	58	5	index	index	NOUN
ajst-7347	58	6	data	datum	NOUN
ajst-7347	58	7	containing	contain	VERB
ajst-7347	58	8	incorrect	incorrect	ADJ
ajst-7347	58	9	input	input	NOUN
ajst-7347	58	10	and	and	CCONJ
ajst-7347	58	11	illegal	illegal	ADJ
ajst-7347	58	12	data	datum	NOUN
ajst-7347	58	13	,	,	PUNCT
ajst-7347	58	14	that	that	ADV
ajst-7347	58	15	is	is	ADV
ajst-7347	58	16	,	,	PUNCT
ajst-7347	58	17	the	the	DET
ajst-7347	58	18	parameter	parameter	NOUN
ajst-7347	58	19	value	value	NOUN
ajst-7347	58	20	exceeds	exceed	VERB
ajst-7347	58	21	the	the	DET
ajst-7347	58	22	normal	normal	ADJ
ajst-7347	58	23	range	range	NOUN
ajst-7347	58	24	.	.	PUNCT
ajst-7347	59	1	the	the	DET
ajst-7347	59	2	box	box	PROPN
ajst-7347	59	3	diagram	diagram	NOUN
ajst-7347	59	4	method	method	NOUN
ajst-7347	59	5	is	be	AUX
ajst-7347	59	6	used	use	VERB
ajst-7347	59	7	to	to	PART
ajst-7347	59	8	treat	treat	VERB
ajst-7347	59	9	values	value	NOUN
ajst-7347	59	10	less	less	ADJ
ajst-7347	59	11	than	than	ADP
ajst-7347	59	12	ql-1.5iqr	ql-1.5iqr	ADJ
ajst-7347	59	13	or	or	CCONJ
ajst-7347	59	14	greater	great	ADJ
ajst-7347	59	15	than	than	ADP
ajst-7347	59	16	qu+1.5iqr	qu+1.5iqr	NUM
ajst-7347	59	17	as	as	ADP
ajst-7347	59	18	exceptions	exception	NOUN
ajst-7347	59	19	.	.	PUNCT
ajst-7347	60	1	qu	qu	PROPN
ajst-7347	60	2	and	and	CCONJ
ajst-7347	60	3	ql	ql	PRON
ajst-7347	60	4	are	be	AUX
ajst-7347	60	5	the	the	DET
ajst-7347	60	6	upper	upper	ADJ
ajst-7347	60	7	quartile	quartile	NOUN
ajst-7347	60	8	and	and	CCONJ
ajst-7347	60	9	the	the	DET
ajst-7347	60	10	lower	low	ADJ
ajst-7347	60	11	quartile	quartile	NOUN
ajst-7347	60	12	respectively	respectively	ADV
ajst-7347	60	13	,	,	PUNCT
ajst-7347	60	14	and	and	CCONJ
ajst-7347	60	15	iqr	iqr	PROPN
ajst-7347	60	16	is	be	AUX
ajst-7347	60	17	the	the	DET
ajst-7347	60	18	spacing	spacing	NOUN
ajst-7347	60	19	of	of	ADP
ajst-7347	60	20	the	the	DET
ajst-7347	60	21	quartiles	quartile	NOUN
ajst-7347	60	22	.	.	PUNCT
ajst-7347	61	1	the	the	DET
ajst-7347	61	2	detected	detect	VERB
ajst-7347	61	3	outliers	outlier	NOUN
ajst-7347	61	4	are	be	AUX
ajst-7347	61	5	regarded	regard	VERB
ajst-7347	61	6	as	as	ADP
ajst-7347	61	7	missing	miss	VERB
ajst-7347	61	8	values	value	NOUN
ajst-7347	61	9	and	and	CCONJ
ajst-7347	61	10	filled	fill	VERB
ajst-7347	61	11	with	with	ADP
ajst-7347	61	12	knn	knn	NOUN
ajst-7347	61	13	filling	filling	NOUN
ajst-7347	61	14	method	method	NOUN
ajst-7347	61	15	.	.	PUNCT
ajst-7347	62	1	the	the	DET
ajst-7347	62	2	data	data	NOUN
ajst-7347	62	3	range	range	NOUN
ajst-7347	62	4	of	of	ADP
ajst-7347	62	5	different	different	ADJ
ajst-7347	62	6	features	feature	NOUN
ajst-7347	62	7	is	be	AUX
ajst-7347	62	8	different	different	ADJ
ajst-7347	62	9	,	,	PUNCT
ajst-7347	62	10	and	and	CCONJ
ajst-7347	62	11	the	the	DET
ajst-7347	62	12	difference	difference	NOUN
ajst-7347	62	13	between	between	ADP
ajst-7347	62	14	values	value	NOUN
ajst-7347	62	15	may	may	AUX
ajst-7347	62	16	be	be	AUX
ajst-7347	62	17	large	large	ADJ
ajst-7347	62	18	,	,	PUNCT
ajst-7347	62	19	so	so	SCONJ
ajst-7347	62	20	the	the	DET
ajst-7347	62	21	data	datum	NOUN
ajst-7347	62	22	are	be	AUX
ajst-7347	62	23	standardized	standardize	VERB
ajst-7347	62	24	.	.	PUNCT
ajst-7347	63	1	standard	standard	ADJ
ajst-7347	63	2	deviation	deviation	NOUN
ajst-7347	63	3	is	be	AUX
ajst-7347	63	4	used	use	VERB
ajst-7347	63	5	to	to	PART
ajst-7347	63	6	standardize	standardize	VERB
ajst-7347	63	7	data	datum	NOUN
ajst-7347	63	8	.	.	PUNCT
ajst-7347	64	1	the	the	DET
ajst-7347	64	2	mean	mean	ADJ
ajst-7347	64	3	value	value	NOUN
ajst-7347	64	4	and	and	CCONJ
ajst-7347	64	5	standard	standard	ADJ
ajst-7347	64	6	deviation	deviation	NOUN
ajst-7347	64	7	of	of	ADP
ajst-7347	64	8	data	datum	NOUN
ajst-7347	64	9	processed	process	VERB
ajst-7347	64	10	by	by	ADP
ajst-7347	64	11	this	this	DET
ajst-7347	64	12	method	method	NOUN
ajst-7347	64	13	are	be	AUX
ajst-7347	64	14	0	0	NUM
ajst-7347	64	15	and	and	CCONJ
ajst-7347	64	16	1	1	NUM
ajst-7347	64	17	respectively	respectively	ADV
ajst-7347	64	18	.	.	PUNCT
ajst-7347	65	1	the	the	DET
ajst-7347	65	2	transformation	transformation	NOUN
ajst-7347	65	3	formula	formula	NOUN
ajst-7347	65	4	is	be	AUX
ajst-7347	65	5	shown	show	VERB
ajst-7347	65	6	in	in	ADP
ajst-7347	65	7	equation	equation	NOUN
ajst-7347	65	8	(	(	PUNCT
ajst-7347	65	9	20	20	NUM
ajst-7347	65	10	):	):	PUNCT
ajst-7347	65	11	*	*	PUNCT
ajst-7347	66	1	x	x	PUNCT
ajst-7347	66	2	x	x	PUNCT
ajst-7347	66	3	x	x	SYM
ajst-7347	66	4			NUM
ajst-7347	66	5			NOUN
ajst-7347	66	6			NUM
ajst-7347	66	7	(	(	PUNCT
ajst-7347	66	8	1	1	X
ajst-7347	66	9	)	)	PUNCT
ajst-7347	66	10	𝑋	𝑋	NOUN
ajst-7347	66	11	is	be	AUX
ajst-7347	66	12	the	the	DET
ajst-7347	66	13	mean	mean	ADJ
ajst-7347	66	14	value	value	NOUN
ajst-7347	66	15	of	of	ADP
ajst-7347	66	16	the	the	DET
ajst-7347	66	17	original	original	ADJ
ajst-7347	66	18	data	datum	NOUN
ajst-7347	66	19	,	,	PUNCT
ajst-7347	66	20	and	and	CCONJ
ajst-7347	66	21	δ	δ	PROPN
ajst-7347	66	22	is	be	AUX
ajst-7347	66	23	the	the	DET
ajst-7347	66	24	standard	standard	ADJ
ajst-7347	66	25	deviation	deviation	NOUN
ajst-7347	66	26	of	of	ADP
ajst-7347	66	27	the	the	DET
ajst-7347	66	28	original	original	ADJ
ajst-7347	66	29	data	datum	NOUN
ajst-7347	66	30	.	.	PUNCT
ajst-7347	67	1	3.3	3.3	NUM
ajst-7347	67	2	.	.	PUNCT
ajst-7347	68	1	model	model	NOUN
ajst-7347	68	2	construction	construction	NOUN
ajst-7347	68	3	the	the	DET
ajst-7347	68	4	hybrid	hybrid	ADJ
ajst-7347	68	5	cnn	cnn	PROPN
ajst-7347	68	6	-	-	PUNCT
ajst-7347	68	7	bigru	bigru	PROPN
ajst-7347	68	8	model	model	NOUN
ajst-7347	68	9	constructed	construct	VERB
ajst-7347	68	10	in	in	ADP
ajst-7347	68	11	this	this	DET
ajst-7347	68	12	paper	paper	NOUN
ajst-7347	68	13	first	first	ADV
ajst-7347	68	14	uses	use	VERB
ajst-7347	68	15	one	one	NUM
ajst-7347	68	16	-	-	PUNCT
ajst-7347	68	17	dimensional	dimensional	ADJ
ajst-7347	68	18	cnn	cnn	NOUN
ajst-7347	68	19	to	to	PART
ajst-7347	68	20	mine	mine	PRON
ajst-7347	68	21	the	the	DET
ajst-7347	68	22	nonlinear	nonlinear	ADJ
ajst-7347	68	23	characteristics	characteristic	NOUN
ajst-7347	68	24	of	of	ADP
ajst-7347	68	25	data	datum	NOUN
ajst-7347	68	26	,	,	PUNCT
ajst-7347	68	27	improve	improve	VERB
ajst-7347	68	28	the	the	DET
ajst-7347	68	29	running	running	NOUN
ajst-7347	68	30	speed	speed	NOUN
ajst-7347	68	31	,	,	PUNCT
ajst-7347	68	32	and	and	CCONJ
ajst-7347	68	33	also	also	ADV
ajst-7347	68	34	eliminate	eliminate	VERB
ajst-7347	68	35	some	some	DET
ajst-7347	68	36	unstable	unstable	ADJ
ajst-7347	68	37	factors	factor	NOUN
ajst-7347	68	38	.	.	PUNCT
ajst-7347	69	1	then	then	ADV
ajst-7347	69	2	,	,	PUNCT
ajst-7347	69	3	its	its	PRON
ajst-7347	69	4	output	output	NOUN
ajst-7347	69	5	is	be	AUX
ajst-7347	69	6	used	use	VERB
ajst-7347	69	7	as	as	ADP
ajst-7347	69	8	the	the	DET
ajst-7347	69	9	input	input	NOUN
ajst-7347	69	10	of	of	ADP
ajst-7347	69	11	the	the	DET
ajst-7347	69	12	next	next	ADJ
ajst-7347	69	13	time	time	NOUN
ajst-7347	69	14	series	series	PROPN
ajst-7347	69	15	model	model	PROPN
ajst-7347	69	16	,	,	PUNCT
ajst-7347	69	17	bigru	bigru	NOUN
ajst-7347	69	18	,	,	PUNCT
ajst-7347	69	19	to	to	PART
ajst-7347	69	20	extract	extract	VERB
ajst-7347	69	21	hidden	hide	VERB
ajst-7347	69	22	temporal	temporal	ADJ
ajst-7347	69	23	rules	rule	NOUN
ajst-7347	69	24	,	,	PUNCT
ajst-7347	69	25	which	which	PRON
ajst-7347	69	26	can	can	AUX
ajst-7347	69	27	capture	capture	VERB
ajst-7347	69	28	long	long	ADJ
ajst-7347	69	29	-	-	PUNCT
ajst-7347	69	30	term	term	NOUN
ajst-7347	69	31	and	and	CCONJ
ajst-7347	69	32	short	short	ADJ
ajst-7347	69	33	-	-	PUNCT
ajst-7347	69	34	term	term	NOUN
ajst-7347	69	35	dependencies	dependency	NOUN
ajst-7347	69	36	in	in	ADP
ajst-7347	69	37	time	time	NOUN
ajst-7347	69	38	series	series	NOUN
ajst-7347	69	39	,	,	PUNCT
ajst-7347	69	40	and	and	CCONJ
ajst-7347	69	41	feature	feature	NOUN
ajst-7347	69	42	vectors	vector	NOUN
ajst-7347	69	43	can	can	AUX
ajst-7347	69	44	be	be	AUX
ajst-7347	69	45	extracted	extract	VERB
ajst-7347	69	46	through	through	ADP
ajst-7347	69	47	forward	forward	ADV
ajst-7347	69	48	and	and	CCONJ
ajst-7347	69	49	backward	backward	ADJ
ajst-7347	69	50	transmission	transmission	NOUN
ajst-7347	69	51	.	.	PUNCT
ajst-7347	70	1	this	this	DET
ajst-7347	70	2	model	model	NOUN
ajst-7347	70	3	combines	combine	VERB
ajst-7347	70	4	the	the	DET
ajst-7347	70	5	advantages	advantage	NOUN
ajst-7347	70	6	of	of	ADP
ajst-7347	70	7	cnn	cnn	PROPN
ajst-7347	70	8	in	in	ADP
ajst-7347	70	9	learning	learn	VERB
ajst-7347	70	10	local	local	ADJ
ajst-7347	70	11	and	and	CCONJ
ajst-7347	70	12	related	related	ADJ
ajst-7347	70	13	features	feature	NOUN
ajst-7347	70	14	with	with	ADP
ajst-7347	70	15	the	the	DET
ajst-7347	70	16	advantages	advantage	NOUN
ajst-7347	70	17	of	of	ADP
ajst-7347	70	18	learning	learn	VERB
ajst-7347	70	19	temporal	temporal	ADJ
ajst-7347	70	20	rule	rule	NOUN
ajst-7347	70	21	representations	representation	NOUN
ajst-7347	70	22	from	from	ADP
ajst-7347	70	23	bigru	bigru	NOUN
ajst-7347	70	24	,	,	PUNCT
ajst-7347	70	25	further	far	ADV
ajst-7347	70	26	improving	improve	VERB
ajst-7347	70	27	the	the	DET
ajst-7347	70	28	prediction	prediction	NOUN
ajst-7347	70	29	accuracy	accuracy	NOUN
ajst-7347	70	30	of	of	ADP
ajst-7347	70	31	pm2.5	pm2.5	DET
ajst-7347	70	32	concentrations	concentration	NOUN
ajst-7347	70	33	.	.	PUNCT
ajst-7347	71	1	in	in	ADP
ajst-7347	71	2	this	this	DET
ajst-7347	71	3	study	study	NOUN
ajst-7347	71	4	,	,	PUNCT
ajst-7347	71	5	there	there	PRON
ajst-7347	71	6	are	be	VERB
ajst-7347	71	7	eight	eight	NUM
ajst-7347	71	8	data	datum	NOUN
ajst-7347	71	9	features	feature	NOUN
ajst-7347	71	10	used	use	VERB
ajst-7347	71	11	.	.	PUNCT
ajst-7347	72	1	the	the	DET
ajst-7347	72	2	input	input	NOUN
ajst-7347	72	3	is	be	AUX
ajst-7347	72	4	the	the	DET
ajst-7347	72	5	pollutant	pollutant	ADJ
ajst-7347	72	6	concentration	concentration	NOUN
ajst-7347	72	7	data	datum	NOUN
ajst-7347	72	8	for	for	ADP
ajst-7347	72	9	the	the	DET
ajst-7347	72	10	past	past	ADJ
ajst-7347	72	11	hour	hour	NOUN
ajst-7347	72	12	,	,	PUNCT
ajst-7347	72	13	and	and	CCONJ
ajst-7347	72	14	the	the	DET
ajst-7347	72	15	output	output	NOUN
ajst-7347	72	16	is	be	AUX
ajst-7347	72	17	the	the	DET
ajst-7347	72	18	concentration	concentration	NOUN
ajst-7347	72	19	value	value	NOUN
ajst-7347	72	20	of	of	ADP
ajst-7347	72	21	pm2.5	pm2.5	PROPN
ajst-7347	72	22	for	for	ADP
ajst-7347	72	23	the	the	DET
ajst-7347	72	24	next	next	ADJ
ajst-7347	72	25	hour	hour	NOUN
ajst-7347	72	26	.	.	PUNCT
ajst-7347	73	1	that	that	PRON
ajst-7347	73	2	is	is	ADV
ajst-7347	73	3	,	,	PUNCT
ajst-7347	73	4	the	the	DET
ajst-7347	73	5	input	input	NOUN
ajst-7347	73	6	is	be	AUX
ajst-7347	73	7	a	a	DET
ajst-7347	73	8	matrix	matrix	NOUN
ajst-7347	73	9	of	of	ADP
ajst-7347	73	10	(	(	PUNCT
ajst-7347	73	11	1	1	NUM
ajst-7347	73	12	×	×	NOUN
ajst-7347	73	13	8)	8)	NUM
ajst-7347	73	14	size	size	NOUN
ajst-7347	73	15	,	,	PUNCT
ajst-7347	73	16	and	and	CCONJ
ajst-7347	73	17	the	the	DET
ajst-7347	73	18	output	output	NOUN
ajst-7347	73	19	matrix	matrix	NOUN
ajst-7347	73	20	is	be	AUX
ajst-7347	73	21	(	(	PUNCT
ajst-7347	73	22	1	1	NUM
ajst-7347	73	23	×	×	NOUN
ajst-7347	73	24	1	1	NUM
ajst-7347	73	25	)	)	PUNCT
ajst-7347	73	26	.	.	PUNCT
ajst-7347	74	1	the	the	DET
ajst-7347	74	2	construction	construction	NOUN
ajst-7347	74	3	of	of	ADP
ajst-7347	74	4	cnn	cnn	PROPN
ajst-7347	74	5	-	-	PUNCT
ajst-7347	74	6	bigru	bigru	PROPN
ajst-7347	74	7	model	model	NOUN
ajst-7347	74	8	is	be	AUX
ajst-7347	74	9	mainly	mainly	ADV
ajst-7347	74	10	divided	divide	VERB
ajst-7347	74	11	into	into	ADP
ajst-7347	74	12	two	two	NUM
ajst-7347	74	13	parts	part	NOUN
ajst-7347	74	14	,	,	PUNCT
ajst-7347	74	15	the	the	DET
ajst-7347	74	16	structural	structural	ADJ
ajst-7347	74	17	design	design	NOUN
ajst-7347	74	18	of	of	ADP
ajst-7347	74	19	one	one	NUM
ajst-7347	74	20	-	-	PUNCT
ajst-7347	74	21	dimensional	dimensional	ADJ
ajst-7347	74	22	cnn	cnn	NOUN
ajst-7347	74	23	and	and	CCONJ
ajst-7347	74	24	the	the	DET
ajst-7347	74	25	structural	structural	ADJ
ajst-7347	74	26	design	design	NOUN
ajst-7347	74	27	of	of	ADP
ajst-7347	74	28	bigru	bigru	NOUN
ajst-7347	74	29	model	model	NOUN
ajst-7347	74	30	.	.	PUNCT
ajst-7347	75	1	the	the	DET
ajst-7347	75	2	input	input	NOUN
ajst-7347	75	3	data	datum	NOUN
ajst-7347	75	4	input	input	NOUN
ajst-7347	75	5	sample	sample	NOUN
ajst-7347	75	6	constructed	construct	VERB
ajst-7347	75	7	in	in	ADP
ajst-7347	75	8	this	this	DET
ajst-7347	75	9	paper	paper	NOUN
ajst-7347	75	10	is	be	AUX
ajst-7347	75	11	divided	divide	VERB
ajst-7347	75	12	into	into	ADP
ajst-7347	75	13	data	datum	NOUN
ajst-7347	75	14	segments	segment	NOUN
ajst-7347	75	15	of	of	ADP
ajst-7347	75	16	a	a	DET
ajst-7347	75	17	certain	certain	ADJ
ajst-7347	75	18	length	length	NOUN
ajst-7347	75	19	.	.	PUNCT
ajst-7347	76	1	using	use	VERB
ajst-7347	76	2	one	one	NUM
ajst-7347	76	3	-	-	PUNCT
ajst-7347	76	4	dimensional	dimensional	ADJ
ajst-7347	76	5	cnn	cnn	NOUN
ajst-7347	76	6	can	can	AUX
ajst-7347	76	7	effectively	effectively	ADV
ajst-7347	76	8	extract	extract	VERB
ajst-7347	76	9	local	local	ADJ
ajst-7347	76	10	features	feature	NOUN
ajst-7347	76	11	.	.	PUNCT
ajst-7347	77	1	since	since	SCONJ
ajst-7347	77	2	this	this	DET
ajst-7347	77	3	article	article	NOUN
ajst-7347	77	4	uses	use	VERB
ajst-7347	77	5	historical	historical	ADJ
ajst-7347	77	6	1	1	NUM
ajst-7347	77	7	-	-	PUNCT
ajst-7347	77	8	hour	hour	NOUN
ajst-7347	77	9	data	datum	NOUN
ajst-7347	77	10	of	of	ADP
ajst-7347	77	11	various	various	ADJ
ajst-7347	77	12	pollutants	pollutant	NOUN
ajst-7347	77	13	and	and	CCONJ
ajst-7347	77	14	meteorological	meteorological	ADJ
ajst-7347	77	15	factors	factor	NOUN
ajst-7347	77	16	to	to	PART
ajst-7347	77	17	predict	predict	VERB
ajst-7347	77	18	the	the	DET
ajst-7347	77	19	pm2.5	pm2.5	ADJ
ajst-7347	77	20	concentration	concentration	NOUN
ajst-7347	77	21	in	in	ADP
ajst-7347	77	22	the	the	DET
ajst-7347	77	23	next	next	ADJ
ajst-7347	77	24	hour	hour	NOUN
ajst-7347	77	25	,	,	PUNCT
ajst-7347	77	26	the	the	DET
ajst-7347	77	27	design	design	NOUN
ajst-7347	77	28	of	of	ADP
ajst-7347	77	29	the	the	DET
ajst-7347	77	30	network	network	NOUN
ajst-7347	77	31	layer	layer	NOUN
ajst-7347	77	32	should	should	AUX
ajst-7347	77	33	not	not	PART
ajst-7347	77	34	be	be	AUX
ajst-7347	77	35	complex	complex	ADJ
ajst-7347	77	36	.	.	PUNCT
ajst-7347	78	1	based	base	VERB
ajst-7347	78	2	on	on	ADP
ajst-7347	78	3	previous	previous	ADJ
ajst-7347	78	4	experience	experience	NOUN
ajst-7347	78	5	,	,	PUNCT
ajst-7347	78	6	a	a	DET
ajst-7347	78	7	layer	layer	NOUN
ajst-7347	78	8	of	of	ADP
ajst-7347	78	9	convolution	convolution	NOUN
ajst-7347	78	10	layer	layer	NOUN
ajst-7347	78	11	and	and	CCONJ
ajst-7347	78	12	pooling	pool	VERB
ajst-7347	78	13	layer	layer	NOUN
ajst-7347	78	14	has	have	AUX
ajst-7347	78	15	been	be	AUX
ajst-7347	78	16	designed	design	VERB
ajst-7347	78	17	to	to	PART
ajst-7347	78	18	avoid	avoid	VERB
ajst-7347	78	19	increasing	increase	VERB
ajst-7347	78	20	the	the	DET
ajst-7347	78	21	calculation	calculation	NOUN
ajst-7347	78	22	amount	amount	NOUN
ajst-7347	78	23	.	.	PUNCT
ajst-7347	79	1	the	the	DET
ajst-7347	79	2	structural	structural	ADJ
ajst-7347	79	3	design	design	NOUN
ajst-7347	79	4	of	of	ADP
ajst-7347	79	5	the	the	DET
ajst-7347	79	6	cnn	cnn	PROPN
ajst-7347	79	7	-	-	PUNCT
ajst-7347	79	8	bigru	bigru	PROPN
ajst-7347	79	9	model	model	NOUN
ajst-7347	79	10	is	be	AUX
ajst-7347	79	11	also	also	ADV
ajst-7347	79	12	based	base	VERB
ajst-7347	79	13	on	on	ADP
ajst-7347	79	14	past	past	ADJ
ajst-7347	79	15	experience	experience	NOUN
ajst-7347	79	16	.	.	PUNCT
ajst-7347	80	1	it	it	PRON
ajst-7347	80	2	mainly	mainly	ADV
ajst-7347	80	3	consists	consist	VERB
ajst-7347	80	4	of	of	ADP
ajst-7347	80	5	the	the	DET
ajst-7347	80	6	following	follow	VERB
ajst-7347	80	7	parts	part	NOUN
ajst-7347	80	8	:	:	PUNCT
ajst-7347	80	9	input	input	NOUN
ajst-7347	80	10	layer	layer	NOUN
ajst-7347	80	11	,	,	PUNCT
ajst-7347	80	12	convolution	convolution	NOUN
ajst-7347	80	13	layer	layer	NOUN
ajst-7347	80	14	,	,	PUNCT
ajst-7347	80	15	pooling	pool	VERB
ajst-7347	80	16	layer	layer	NOUN
ajst-7347	80	17	,	,	PUNCT
ajst-7347	80	18	bigru	bigru	NOUN
ajst-7347	80	19	layer	layer	NOUN
ajst-7347	80	20	,	,	PUNCT
ajst-7347	80	21	full	full	ADJ
ajst-7347	80	22	connection	connection	NOUN
ajst-7347	80	23	layer	layer	NOUN
ajst-7347	80	24	,	,	PUNCT
ajst-7347	80	25	and	and	CCONJ
ajst-7347	80	26	output	output	NOUN
ajst-7347	80	27	layer	layer	NOUN
ajst-7347	80	28	.	.	PUNCT
ajst-7347	81	1	the	the	DET
ajst-7347	81	2	input	input	NOUN
ajst-7347	81	3	layer	layer	NOUN
ajst-7347	81	4	is	be	AUX
ajst-7347	81	5	mainly	mainly	ADV
ajst-7347	81	6	used	use	VERB
ajst-7347	81	7	for	for	ADP
ajst-7347	81	8	data	data	NOUN
ajst-7347	81	9	preparation	preparation	NOUN
ajst-7347	81	10	.	.	PUNCT
ajst-7347	82	1	the	the	DET
ajst-7347	82	2	convolution	convolution	NOUN
ajst-7347	82	3	layer	layer	NOUN
ajst-7347	82	4	extracts	extract	NOUN
ajst-7347	82	5	features	feature	NOUN
ajst-7347	82	6	,	,	PUNCT
ajst-7347	82	7	and	and	CCONJ
ajst-7347	82	8	the	the	DET
ajst-7347	82	9	pooling	pooling	NOUN
ajst-7347	82	10	layer	layer	NOUN
ajst-7347	82	11	is	be	AUX
ajst-7347	82	12	used	use	VERB
ajst-7347	82	13	for	for	ADP
ajst-7347	82	14	data	data	NOUN
ajst-7347	82	15	sampling	sampling	NOUN
ajst-7347	82	16	.	.	PUNCT
ajst-7347	83	1	generally	generally	ADV
ajst-7347	83	2	,	,	PUNCT
ajst-7347	83	3	the	the	DET
ajst-7347	83	4	convolution	convolution	NOUN
ajst-7347	83	5	layer	layer	NOUN
ajst-7347	83	6	and	and	CCONJ
ajst-7347	83	7	pooling	pool	VERB
ajst-7347	83	8	layer	layer	NOUN
ajst-7347	83	9	of	of	ADP
ajst-7347	83	10	a	a	DET
ajst-7347	83	11	convolutional	convolutional	ADJ
ajst-7347	83	12	neural	neural	ADJ
ajst-7347	83	13	network	network	NOUN
ajst-7347	83	14	are	be	AUX
ajst-7347	83	15	nested	nest	VERB
ajst-7347	83	16	and	and	CCONJ
ajst-7347	83	17	used	use	VERB
ajst-7347	83	18	to	to	PART
ajst-7347	83	19	extract	extract	VERB
ajst-7347	83	20	and	and	CCONJ
ajst-7347	83	21	compress	compress	VERB
ajst-7347	83	22	temporal	temporal	ADJ
ajst-7347	83	23	dimension	dimension	NOUN
ajst-7347	83	24	features	feature	NOUN
ajst-7347	83	25	.	.	PUNCT
ajst-7347	84	1	the	the	DET
ajst-7347	84	2	convolution	convolution	NOUN
ajst-7347	84	3	layer	layer	NOUN
ajst-7347	84	4	transmits	transmit	VERB
ajst-7347	84	5	the	the	DET
ajst-7347	84	6	extracted	extract	VERB
ajst-7347	84	7	data	datum	NOUN
ajst-7347	84	8	information	information	NOUN
ajst-7347	84	9	to	to	ADP
ajst-7347	84	10	the	the	DET
ajst-7347	84	11	bi	bi	ADJ
ajst-7347	84	12	-	-	ADJ
ajst-7347	84	13	directional	directional	ADJ
ajst-7347	84	14	circular	circular	ADJ
ajst-7347	84	15	neural	neural	ADJ
ajst-7347	84	16	network	network	NOUN
ajst-7347	84	17	bigru	bigru	VERB
ajst-7347	84	18	layer	layer	NOUN
ajst-7347	84	19	,	,	PUNCT
ajst-7347	84	20	which	which	PRON
ajst-7347	84	21	performs	perform	VERB
ajst-7347	84	22	forward	forward	ADV
ajst-7347	84	23	and	and	CCONJ
ajst-7347	84	24	reverse	reverse	VERB
ajst-7347	84	25	transmission	transmission	NOUN
ajst-7347	84	26	of	of	ADP
ajst-7347	84	27	data	datum	NOUN
ajst-7347	84	28	,	,	PUNCT
ajst-7347	84	29	followed	follow	VERB
ajst-7347	84	30	by	by	ADP
ajst-7347	84	31	a	a	DET
ajst-7347	84	32	fully	fully	ADV
ajst-7347	84	33	connected	connect	VERB
ajst-7347	84	34	layer	layer	NOUN
ajst-7347	84	35	.	.	PUNCT
ajst-7347	85	1	the	the	DET
ajst-7347	85	2	model	model	NOUN
ajst-7347	85	3	structure	structure	NOUN
ajst-7347	85	4	diagram	diagram	NOUN
ajst-7347	85	5	is	be	AUX
ajst-7347	85	6	shown	show	VERB
ajst-7347	85	7	in	in	ADP
ajst-7347	85	8	figure	figure	NOUN
ajst-7347	85	9	4	4	NUM
ajst-7347	85	10	.	.	NOUN
ajst-7347	85	11	4	4	NUM
ajst-7347	85	12	figure	figure	NOUN
ajst-7347	85	13	4	4	NUM
ajst-7347	85	14	.	.	PUNCT
ajst-7347	86	1	structure	structure	NOUN
ajst-7347	86	2	of	of	ADP
ajst-7347	86	3	the	the	DET
ajst-7347	86	4	cnn	cnn	PROPN
ajst-7347	86	5	-	-	PUNCT
ajst-7347	86	6	bigru	bigru	PROPN
ajst-7347	86	7	model	model	NOUN
ajst-7347	86	8	due	due	ADP
ajst-7347	86	9	to	to	ADP
ajst-7347	86	10	the	the	DET
ajst-7347	86	11	small	small	ADJ
ajst-7347	86	12	dimension	dimension	NOUN
ajst-7347	86	13	of	of	ADP
ajst-7347	86	14	the	the	DET
ajst-7347	86	15	input	input	NOUN
ajst-7347	86	16	matrix	matrix	NOUN
ajst-7347	86	17	in	in	ADP
ajst-7347	86	18	this	this	DET
ajst-7347	86	19	article	article	NOUN
ajst-7347	86	20	,	,	PUNCT
ajst-7347	86	21	the	the	DET
ajst-7347	86	22	number	number	NOUN
ajst-7347	86	23	of	of	ADP
ajst-7347	86	24	layers	layer	NOUN
ajst-7347	86	25	for	for	ADP
ajst-7347	86	26	constructing	construct	VERB
ajst-7347	86	27	the	the	DET
ajst-7347	86	28	network	network	NOUN
ajst-7347	86	29	should	should	AUX
ajst-7347	86	30	not	not	PART
ajst-7347	86	31	be	be	AUX
ajst-7347	86	32	too	too	ADV
ajst-7347	86	33	large	large	ADJ
ajst-7347	86	34	.	.	PUNCT
ajst-7347	87	1	this	this	DET
ajst-7347	87	2	article	article	NOUN
ajst-7347	87	3	sets	set	VERB
ajst-7347	87	4	up	up	ADP
ajst-7347	87	5	a	a	DET
ajst-7347	87	6	one	one	NUM
ajst-7347	87	7	-	-	PUNCT
ajst-7347	87	8	dimensional	dimensional	ADJ
ajst-7347	87	9	convolutional	convolutional	ADJ
ajst-7347	87	10	layer	layer	NOUN
ajst-7347	87	11	,	,	PUNCT
ajst-7347	87	12	with	with	ADP
ajst-7347	87	13	a	a	DET
ajst-7347	87	14	convolutional	convolutional	ADJ
ajst-7347	87	15	core	core	NOUN
ajst-7347	87	16	size	size	NOUN
ajst-7347	87	17	of	of	ADP
ajst-7347	87	18	2	2	NUM
ajst-7347	87	19	,	,	PUNCT
ajst-7347	87	20	a	a	DET
ajst-7347	87	21	convolutional	convolutional	ADJ
ajst-7347	87	22	core	core	NOUN
ajst-7347	87	23	number	number	NOUN
ajst-7347	87	24	of	of	ADP
ajst-7347	87	25	64	64	NUM
ajst-7347	87	26	,	,	PUNCT
ajst-7347	87	27	and	and	CCONJ
ajst-7347	87	28	a	a	DET
ajst-7347	87	29	step	step	NOUN
ajst-7347	87	30	size	size	NOUN
ajst-7347	87	31	of	of	ADP
ajst-7347	87	32	1	1	NUM
ajst-7347	87	33	.	.	PUNCT
ajst-7347	88	1	the	the	DET
ajst-7347	88	2	maximum	maximum	ADJ
ajst-7347	88	3	pooling	pooling	NOUN
ajst-7347	88	4	is	be	AUX
ajst-7347	88	5	selected	select	VERB
ajst-7347	88	6	for	for	ADP
ajst-7347	88	7	the	the	DET
ajst-7347	88	8	pooling	pool	VERB
ajst-7347	88	9	layer	layer	NOUN
ajst-7347	88	10	.	.	PUNCT
ajst-7347	89	1	the	the	DET
ajst-7347	89	2	bigru	bigru	NOUN
ajst-7347	89	3	layer	layer	NOUN
ajst-7347	89	4	is	be	AUX
ajst-7347	89	5	also	also	ADV
ajst-7347	89	6	set	set	VERB
ajst-7347	89	7	to	to	ADP
ajst-7347	89	8	one	one	NUM
ajst-7347	89	9	layer	layer	NOUN
ajst-7347	89	10	with	with	ADP
ajst-7347	89	11	a	a	DET
ajst-7347	89	12	number	number	NOUN
ajst-7347	89	13	of	of	ADP
ajst-7347	89	14	neurons	neuron	NOUN
ajst-7347	89	15	of	of	ADP
ajst-7347	89	16	200	200	NUM
ajst-7347	89	17	,	,	PUNCT
ajst-7347	89	18	which	which	PRON
ajst-7347	89	19	can	can	AUX
ajst-7347	89	20	save	save	VERB
ajst-7347	89	21	time	time	NOUN
ajst-7347	89	22	while	while	SCONJ
ajst-7347	89	23	maintaining	maintain	VERB
ajst-7347	89	24	good	good	ADJ
ajst-7347	89	25	accuracy	accuracy	NOUN
ajst-7347	89	26	.	.	PUNCT
ajst-7347	90	1	3.4	3.4	NUM
ajst-7347	90	2	.	.	PUNCT
ajst-7347	90	3	evaluation	evaluation	NOUN
ajst-7347	90	4	criteria	criterion	NOUN
ajst-7347	90	5	in	in	ADP
ajst-7347	90	6	this	this	DET
ajst-7347	90	7	paper	paper	NOUN
ajst-7347	90	8	,	,	PUNCT
ajst-7347	90	9	root	root	NOUN
ajst-7347	90	10	mean	mean	VERB
ajst-7347	90	11	square	square	ADJ
ajst-7347	90	12	error	error	NOUN
ajst-7347	90	13	(	(	PUNCT
ajst-7347	90	14	rmse	rmse	NOUN
ajst-7347	90	15	)	)	PUNCT
ajst-7347	90	16	,	,	PUNCT
ajst-7347	90	17	mean	mean	VERB
ajst-7347	90	18	absolute	absolute	ADJ
ajst-7347	90	19	error	error	NOUN
ajst-7347	90	20	(	(	PUNCT
ajst-7347	90	21	mae	mae	PROPN
ajst-7347	90	22	)	)	PUNCT
ajst-7347	90	23	and	and	CCONJ
ajst-7347	90	24	determination	determination	NOUN
ajst-7347	90	25	coefficient	coefficient	PROPN
ajst-7347	90	26	𝑅	𝑅	PROPN
ajst-7347	90	27	are	be	AUX
ajst-7347	90	28	selected	select	VERB
ajst-7347	90	29	to	to	PART
ajst-7347	90	30	evaluate	evaluate	VERB
ajst-7347	90	31	the	the	DET
ajst-7347	90	32	prediction	prediction	NOUN
ajst-7347	90	33	performance	performance	NOUN
ajst-7347	90	34	of	of	ADP
ajst-7347	90	35	the	the	DET
ajst-7347	90	36	model	model	NOUN
ajst-7347	90	37	.	.	PUNCT
ajst-7347	91	1	its	its	PRON
ajst-7347	91	2	formula	formula	NOUN
ajst-7347	91	3	is	be	AUX
ajst-7347	91	4	defined	define	VERB
ajst-7347	91	5	as	as	SCONJ
ajst-7347	91	6	follows	follow	VERB
ajst-7347	91	7	:	:	PUNCT
ajst-7347	91	8	�	�	PROPN
ajst-7347	91	9	2	2	NUM
ajst-7347	91	10	1	1	NUM
ajst-7347	91	11	1	1	NUM
ajst-7347	91	12	(	(	PUNCT
ajst-7347	91	13	)	)	PUNCT
ajst-7347	92	1	n	n	CCONJ
ajst-7347	92	2	i	i	PRON
ajst-7347	92	3	ii	ii	PROPN
ajst-7347	92	4	rmse	rmse	NOUN
ajst-7347	93	1	y	y	PROPN
ajst-7347	93	2	y	y	PROPN
ajst-7347	93	3	n	n	CCONJ
ajst-7347	93	4			NUM
ajst-7347	93	5			NUM
ajst-7347	94	1			X
ajst-7347	94	2	(	(	PUNCT
ajst-7347	94	3	2	2	NUM
ajst-7347	94	4	)	)	SYM
ajst-7347	94	5	1	1	NUM
ajst-7347	94	6	n	n	NOUN
ajst-7347	95	1	i	i	PRON
ajst-7347	95	2	ii	ii	VERB
ajst-7347	95	3	y	y	VERB
ajst-7347	95	4	x	x	SYM
ajst-7347	95	5	mae	mae	PROPN
ajst-7347	95	6	n	n	PROPN
ajst-7347	95	7			PROPN
ajst-7347	95	8			PROPN
ajst-7347	95	9			NOUN
ajst-7347	95	10			X
ajst-7347	95	11	(	(	PUNCT
ajst-7347	95	12	3	3	X
ajst-7347	95	13	)	)	PUNCT
ajst-7347	95	14	�	�	NOUN
ajst-7347	95	15	2	2	NUM
ajst-7347	95	16	2	2	NUM
ajst-7347	95	17	2	2	NUM
ajst-7347	95	18	(	(	PUNCT
ajst-7347	95	19	)	)	PUNCT
ajst-7347	95	20	1	1	NUM
ajst-7347	95	21	(	(	PUNCT
ajst-7347	95	22	)	)	PUNCT
ajst-7347	95	23	l	l	PROPN
ajst-7347	95	24	ii	ii	PROPN
ajst-7347	95	25	ii	ii	X
ajst-7347	95	26	y	y	PROPN
ajst-7347	95	27	y	y	PROPN
ajst-7347	95	28	r	r	PROPN
ajst-7347	95	29	y	y	PROPN
ajst-7347	95	30	y	y	PROPN
ajst-7347	95	31			PROPN
ajst-7347	95	32			PROPN
ajst-7347	95	33			PROPN
ajst-7347	95	34			PROPN
ajst-7347	95	35			X
ajst-7347	95	36			X
ajst-7347	95	37	(	(	PUNCT
ajst-7347	95	38	4	4	X
ajst-7347	95	39	)	)	PUNCT
ajst-7347	95	40	the	the	PRON
ajst-7347	95	41	smaller	small	ADJ
ajst-7347	95	42	the	the	DET
ajst-7347	95	43	rmse	rmse	NOUN
ajst-7347	95	44	and	and	CCONJ
ajst-7347	95	45	mae	mae	PROPN
ajst-7347	95	46	values	value	NOUN
ajst-7347	95	47	are	be	AUX
ajst-7347	95	48	,	,	PUNCT
ajst-7347	95	49	the	the	PRON
ajst-7347	95	50	better	well	ADJ
ajst-7347	95	51	the	the	DET
ajst-7347	95	52	effect	effect	NOUN
ajst-7347	95	53	of	of	ADP
ajst-7347	95	54	the	the	DET
ajst-7347	95	55	model	model	NOUN
ajst-7347	95	56	will	will	AUX
ajst-7347	95	57	be	be	AUX
ajst-7347	95	58	.	.	PUNCT
ajst-7347	96	1	the	the	PRON
ajst-7347	96	2	larger	large	ADJ
ajst-7347	96	3	the	the	DET
ajst-7347	96	4	𝑅	𝑅	PROPN
ajst-7347	96	5	value	value	NOUN
ajst-7347	96	6	is	be	AUX
ajst-7347	96	7	,	,	PUNCT
ajst-7347	96	8	the	the	PRON
ajst-7347	96	9	better	well	ADJ
ajst-7347	96	10	the	the	DET
ajst-7347	96	11	effect	effect	NOUN
ajst-7347	96	12	of	of	ADP
ajst-7347	96	13	the	the	DET
ajst-7347	96	14	model	model	NOUN
ajst-7347	96	15	will	will	AUX
ajst-7347	96	16	be	be	AUX
ajst-7347	96	17	.	.	PUNCT
ajst-7347	97	1	3.5	3.5	NUM
ajst-7347	97	2	.	.	PUNCT
ajst-7347	98	1	result	result	VERB
ajst-7347	98	2	analysis	analysis	NOUN
ajst-7347	98	3	from	from	ADP
ajst-7347	98	4	the	the	DET
ajst-7347	98	5	scatter	scatter	NOUN
ajst-7347	98	6	diagram	diagram	NOUN
ajst-7347	98	7	figure	figure	NOUN
ajst-7347	98	8	5	5	NUM
ajst-7347	98	9	(	(	PUNCT
ajst-7347	98	10	a	a	NOUN
ajst-7347	98	11	)	)	PUNCT
ajst-7347	98	12	(	(	PUNCT
ajst-7347	98	13	b	b	X
ajst-7347	98	14	)	)	PUNCT
ajst-7347	98	15	(	(	PUNCT
ajst-7347	98	16	c	c	NOUN
ajst-7347	98	17	)	)	PUNCT
ajst-7347	98	18	,	,	PUNCT
ajst-7347	98	19	it	it	PRON
ajst-7347	98	20	can	can	AUX
ajst-7347	98	21	be	be	AUX
ajst-7347	98	22	observed	observe	VERB
ajst-7347	98	23	that	that	SCONJ
ajst-7347	98	24	the	the	DET
ajst-7347	98	25	𝑅	𝑅	PROPN
ajst-7347	98	26	values	value	NOUN
ajst-7347	98	27	of	of	ADP
ajst-7347	98	28	the	the	DET
ajst-7347	98	29	three	three	NUM
ajst-7347	98	30	monitoring	monitoring	NOUN
ajst-7347	98	31	points	point	NOUN
ajst-7347	98	32	are	be	AUX
ajst-7347	98	33	0.9111	0.9111	NUM
ajst-7347	98	34	,	,	PUNCT
ajst-7347	98	35	0.8890	0.8890	NUM
ajst-7347	98	36	,	,	PUNCT
ajst-7347	98	37	and	and	CCONJ
ajst-7347	98	38	0.8768	0.8768	NUM
ajst-7347	98	39	,	,	PUNCT
ajst-7347	98	40	respectively	respectively	ADV
ajst-7347	98	41	,	,	PUNCT
ajst-7347	98	42	which	which	PRON
ajst-7347	98	43	are	be	AUX
ajst-7347	98	44	greater	great	ADJ
ajst-7347	98	45	than	than	ADP
ajst-7347	98	46	0.85	0.85	NUM
ajst-7347	98	47	,	,	PUNCT
ajst-7347	98	48	which	which	PRON
ajst-7347	98	49	can	can	AUX
ajst-7347	98	50	prove	prove	VERB
ajst-7347	98	51	that	that	SCONJ
ajst-7347	98	52	the	the	DET
ajst-7347	98	53	fitting	fitting	ADJ
ajst-7347	98	54	effect	effect	NOUN
ajst-7347	98	55	of	of	ADP
ajst-7347	98	56	the	the	DET
ajst-7347	98	57	combined	combined	ADJ
ajst-7347	98	58	model	model	NOUN
ajst-7347	98	59	at	at	ADP
ajst-7347	98	60	the	the	DET
ajst-7347	98	61	three	three	NUM
ajst-7347	98	62	selected	select	VERB
ajst-7347	98	63	stations	station	NOUN
ajst-7347	98	64	is	be	AUX
ajst-7347	98	65	very	very	ADV
ajst-7347	98	66	good	good	ADJ
ajst-7347	98	67	.	.	PUNCT
ajst-7347	99	1	the	the	DET
ajst-7347	99	2	univariate	univariate	ADJ
ajst-7347	99	3	linear	linear	ADJ
ajst-7347	99	4	fitting	fitting	ADJ
ajst-7347	99	5	function	function	NOUN
ajst-7347	99	6	for	for	ADP
ajst-7347	99	7	the	the	DET
ajst-7347	99	8	monitoring	monitoring	NOUN
ajst-7347	99	9	points	point	NOUN
ajst-7347	99	10	in	in	ADP
ajst-7347	99	11	xindian	xindian	ADJ
ajst-7347	99	12	district	district	NOUN
ajst-7347	99	13	is	be	AUX
ajst-7347	99	14	:	:	PUNCT
ajst-7347	99	15	y=0.9318x+1.3457	y=0.9318x+1.3457	PROPN
ajst-7347	99	16	,	,	PUNCT
ajst-7347	99	17	with	with	ADP
ajst-7347	99	18	a	a	DET
ajst-7347	99	19	slope	slope	NOUN
ajst-7347	99	20	of	of	ADP
ajst-7347	99	21	0.9318	0.9318	NUM
ajst-7347	99	22	,	,	PUNCT
ajst-7347	99	23	which	which	PRON
ajst-7347	99	24	is	be	AUX
ajst-7347	99	25	closer	close	ADJ
ajst-7347	99	26	to	to	ADP
ajst-7347	99	27	1	1	NUM
ajst-7347	99	28	.	.	PUNCT
ajst-7347	99	29	overall	overall	ADV
ajst-7347	99	30	,	,	PUNCT
ajst-7347	99	31	the	the	DET
ajst-7347	99	32	actual	actual	ADJ
ajst-7347	99	33	value	value	NOUN
ajst-7347	99	34	and	and	CCONJ
ajst-7347	99	35	predicted	predict	VERB
ajst-7347	99	36	value	value	NOUN
ajst-7347	99	37	are	be	AUX
ajst-7347	99	38	relatively	relatively	ADV
ajst-7347	99	39	close	close	ADJ
ajst-7347	99	40	.	.	PUNCT
ajst-7347	100	1	the	the	DET
ajst-7347	100	2	slopes	slope	NOUN
ajst-7347	100	3	of	of	ADP
ajst-7347	100	4	the	the	DET
ajst-7347	100	5	fitting	fitting	ADJ
ajst-7347	100	6	equations	equation	NOUN
ajst-7347	100	7	for	for	ADP
ajst-7347	100	8	taoyuan	taoyuan	PROPN
ajst-7347	100	9	district	district	PROPN
ajst-7347	100	10	and	and	CCONJ
ajst-7347	100	11	songshan	songshan	NOUN
ajst-7347	100	12	district	district	NOUN
ajst-7347	100	13	are	be	AUX
ajst-7347	100	14	0.87	0.87	NUM
ajst-7347	100	15	and	and	CCONJ
ajst-7347	100	16	0.88	0.88	NUM
ajst-7347	100	17	,	,	PUNCT
ajst-7347	100	18	and	and	CCONJ
ajst-7347	100	19	their	their	PRON
ajst-7347	100	20	data	datum	NOUN
ajst-7347	100	21	fitting	fitting	ADJ
ajst-7347	100	22	is	be	AUX
ajst-7347	100	23	also	also	ADV
ajst-7347	100	24	relatively	relatively	ADV
ajst-7347	100	25	concentrated	concentrated	ADJ
ajst-7347	100	26	.	.	PUNCT
ajst-7347	101	1	figure	figure	NOUN
ajst-7347	101	2	6	6	NUM
ajst-7347	101	3	(	(	PUNCT
ajst-7347	101	4	a	a	NOUN
ajst-7347	101	5	)	)	PUNCT
ajst-7347	101	6	(	(	PUNCT
ajst-7347	101	7	b	b	X
ajst-7347	101	8	)	)	PUNCT
ajst-7347	101	9	(	(	PUNCT
ajst-7347	101	10	c	c	X
ajst-7347	101	11	)	)	PUNCT
ajst-7347	101	12	show	show	VERB
ajst-7347	101	13	a	a	DET
ajst-7347	101	14	broken	broken	ADJ
ajst-7347	101	15	line	line	NOUN
ajst-7347	101	16	diagram	diagram	NOUN
ajst-7347	101	17	of	of	ADP
ajst-7347	101	18	the	the	DET
ajst-7347	101	19	actual	actual	ADJ
ajst-7347	101	20	and	and	CCONJ
ajst-7347	101	21	predicted	predict	VERB
ajst-7347	101	22	values	value	NOUN
ajst-7347	101	23	of	of	ADP
ajst-7347	101	24	the	the	DET
ajst-7347	101	25	three	three	NUM
ajst-7347	101	26	stations	station	NOUN
ajst-7347	101	27	,	,	PUNCT
ajst-7347	101	28	and	and	CCONJ
ajst-7347	101	29	selects	select	VERB
ajst-7347	101	30	the	the	DET
ajst-7347	101	31	pm2.5	pm2.5	ADJ
ajst-7347	101	32	concentration	concentration	NOUN
ajst-7347	101	33	values	value	NOUN
ajst-7347	101	34	for	for	ADP
ajst-7347	101	35	the	the	DET
ajst-7347	101	36	first	first	ADJ
ajst-7347	101	37	200	200	NUM
ajst-7347	101	38	hours	hour	NOUN
ajst-7347	101	39	.	.	PUNCT
ajst-7347	102	1	it	it	PRON
ajst-7347	102	2	can	can	AUX
ajst-7347	102	3	be	be	AUX
ajst-7347	102	4	seen	see	VERB
ajst-7347	102	5	that	that	SCONJ
ajst-7347	102	6	when	when	SCONJ
ajst-7347	102	7	the	the	DET
ajst-7347	102	8	pm2.5	pm2.5	ADJ
ajst-7347	102	9	concentration	concentration	NOUN
ajst-7347	102	10	value	value	NOUN
ajst-7347	102	11	is	be	AUX
ajst-7347	102	12	less	less	ADJ
ajst-7347	102	13	than	than	ADP
ajst-7347	102	14	5	5	NUM
ajst-7347	102	15	μg/𝑚	μg/𝑚	NOUN
ajst-7347	102	16	,	,	PUNCT
ajst-7347	102	17	the	the	DET
ajst-7347	102	18	fitting	fitting	ADJ
ajst-7347	102	19	degree	degree	NOUN
ajst-7347	102	20	of	of	ADP
ajst-7347	102	21	the	the	DET
ajst-7347	102	22	predicted	predict	VERB
ajst-7347	102	23	value	value	NOUN
ajst-7347	102	24	of	of	ADP
ajst-7347	102	25	the	the	DET
ajst-7347	102	26	model	model	NOUN
ajst-7347	102	27	is	be	AUX
ajst-7347	102	28	low	low	ADJ
ajst-7347	102	29	,	,	PUNCT
ajst-7347	102	30	and	and	CCONJ
ajst-7347	102	31	when	when	SCONJ
ajst-7347	102	32	the	the	DET
ajst-7347	102	33	pm2.5	pm2.5	ADJ
ajst-7347	102	34	concentration	concentration	NOUN
ajst-7347	102	35	value	value	NOUN
ajst-7347	102	36	is	be	AUX
ajst-7347	102	37	between	between	ADP
ajst-7347	102	38	10	10	NUM
ajst-7347	102	39	-	-	SYM
ajst-7347	102	40	25	25	NUM
ajst-7347	102	41	μg/𝑚	μg/𝑚	NOUN
ajst-7347	102	42	,	,	PUNCT
ajst-7347	102	43	the	the	DET
ajst-7347	102	44	fitting	fitting	ADJ
ajst-7347	102	45	effect	effect	NOUN
ajst-7347	102	46	is	be	AUX
ajst-7347	102	47	good	good	ADJ
ajst-7347	102	48	.	.	PUNCT
ajst-7347	103	1	therefore	therefore	ADV
ajst-7347	103	2	,	,	PUNCT
ajst-7347	103	3	it	it	PRON
ajst-7347	103	4	can	can	AUX
ajst-7347	103	5	be	be	AUX
ajst-7347	103	6	concluded	conclude	VERB
ajst-7347	103	7	that	that	SCONJ
ajst-7347	103	8	the	the	DET
ajst-7347	103	9	cnn	cnn	PROPN
ajst-7347	103	10	-	-	PUNCT
ajst-7347	103	11	bigru	bigru	PROPN
ajst-7347	103	12	combined	combined	ADJ
ajst-7347	103	13	model	model	NOUN
ajst-7347	103	14	has	have	VERB
ajst-7347	103	15	a	a	DET
ajst-7347	103	16	relatively	relatively	ADV
ajst-7347	103	17	good	good	ADJ
ajst-7347	103	18	effect	effect	NOUN
ajst-7347	103	19	in	in	ADP
ajst-7347	103	20	predicting	predict	VERB
ajst-7347	103	21	pm2.5	pm2.5	DET
ajst-7347	103	22	concentration	concentration	NOUN
ajst-7347	103	23	.	.	PUNCT
ajst-7347	104	1	figure	figure	NOUN
ajst-7347	104	2	5	5	NUM
ajst-7347	104	3	.	.	PUNCT
ajst-7347	104	4	forecast	forecast	NOUN
ajst-7347	104	5	results	result	NOUN
ajst-7347	104	6	for	for	ADP
ajst-7347	104	7	each	each	DET
ajst-7347	104	8	site	site	NOUN
ajst-7347	104	9	5	5	NUM
ajst-7347	104	10	in	in	ADP
ajst-7347	104	11	figure	figure	NOUN
ajst-7347	104	12	6(a)(b	6(a)(b	NUM
ajst-7347	104	13	)	)	PUNCT
ajst-7347	104	14	and	and	CCONJ
ajst-7347	104	15	(	(	PUNCT
ajst-7347	104	16	c	c	X
ajst-7347	104	17	)	)	PUNCT
ajst-7347	104	18	are	be	AUX
ajst-7347	104	19	the	the	DET
ajst-7347	104	20	loss	loss	NOUN
ajst-7347	104	21	iteration	iteration	NOUN
ajst-7347	104	22	diagrams	diagram	NOUN
ajst-7347	104	23	for	for	ADP
ajst-7347	104	24	the	the	DET
ajst-7347	104	25	lstm	lstm	PROPN
ajst-7347	104	26	model	model	NOUN
ajst-7347	104	27	,	,	PUNCT
ajst-7347	104	28	cnn	cnn	PROPN
ajst-7347	104	29	-	-	PUNCT
ajst-7347	104	30	gru	gru	PROPN
ajst-7347	104	31	model	model	NOUN
ajst-7347	104	32	,	,	PUNCT
ajst-7347	104	33	and	and	CCONJ
ajst-7347	104	34	cnn	cnn	PROPN
ajst-7347	104	35	-	-	PUNCT
ajst-7347	104	36	bigru	bigru	PROPN
ajst-7347	104	37	model	model	NOUN
ajst-7347	104	38	,	,	PUNCT
ajst-7347	104	39	respectively	respectively	ADV
ajst-7347	104	40	.	.	PUNCT
ajst-7347	105	1	the	the	DET
ajst-7347	105	2	evaluation	evaluation	NOUN
ajst-7347	105	3	index	index	NOUN
ajst-7347	105	4	selects	select	VERB
ajst-7347	105	5	the	the	DET
ajst-7347	105	6	mae	mae	PROPN
ajst-7347	105	7	graph	graph	NOUN
ajst-7347	105	8	,	,	PUNCT
ajst-7347	105	9	with	with	ADP
ajst-7347	105	10	the	the	DET
ajst-7347	105	11	ordinate	ordinate	NOUN
ajst-7347	105	12	loss	loss	NOUN
ajst-7347	105	13	representing	represent	VERB
ajst-7347	105	14	the	the	DET
ajst-7347	105	15	loss	loss	NOUN
ajst-7347	105	16	value	value	NOUN
ajst-7347	105	17	,	,	PUNCT
ajst-7347	105	18	and	and	CCONJ
ajst-7347	105	19	the	the	DET
ajst-7347	105	20	abscissa	abscissa	ADJ
ajst-7347	105	21	epoch	epoch	NOUN
ajst-7347	105	22	representing	represent	VERB
ajst-7347	105	23	the	the	DET
ajst-7347	105	24	number	number	NOUN
ajst-7347	105	25	of	of	ADP
ajst-7347	105	26	iterations	iteration	NOUN
ajst-7347	105	27	.	.	PUNCT
ajst-7347	106	1	the	the	DET
ajst-7347	106	2	epoch	epoch	NOUN
ajst-7347	106	3	for	for	ADP
ajst-7347	106	4	all	all	DET
ajst-7347	106	5	models	model	NOUN
ajst-7347	106	6	is	be	AUX
ajst-7347	106	7	set	set	VERB
ajst-7347	106	8	to	to	ADP
ajst-7347	106	9	60	60	NUM
ajst-7347	106	10	.	.	PUNCT
ajst-7347	107	1	as	as	SCONJ
ajst-7347	107	2	can	can	AUX
ajst-7347	107	3	be	be	AUX
ajst-7347	107	4	observed	observe	VERB
ajst-7347	107	5	in	in	ADP
ajst-7347	107	6	the	the	DET
ajst-7347	107	7	figure	figure	NOUN
ajst-7347	107	8	,	,	PUNCT
ajst-7347	107	9	the	the	DET
ajst-7347	107	10	loss	loss	NOUN
ajst-7347	107	11	rates	rate	NOUN
ajst-7347	107	12	of	of	ADP
ajst-7347	107	13	the	the	DET
ajst-7347	107	14	three	three	NUM
ajst-7347	107	15	models	model	NOUN
ajst-7347	107	16	all	all	PRON
ajst-7347	107	17	decreased	decrease	VERB
ajst-7347	107	18	rapidly	rapidly	ADV
ajst-7347	107	19	during	during	ADP
ajst-7347	107	20	the	the	DET
ajst-7347	107	21	first	first	ADJ
ajst-7347	107	22	few	few	ADJ
ajst-7347	107	23	epochs	epoch	NOUN
ajst-7347	107	24	,	,	PUNCT
ajst-7347	107	25	and	and	CCONJ
ajst-7347	107	26	then	then	ADV
ajst-7347	107	27	decreased	decrease	VERB
ajst-7347	107	28	slightly	slightly	ADV
ajst-7347	107	29	.	.	PUNCT
ajst-7347	108	1	when	when	SCONJ
ajst-7347	108	2	the	the	DET
ajst-7347	108	3	number	number	NOUN
ajst-7347	108	4	of	of	ADP
ajst-7347	108	5	iterations	iteration	NOUN
ajst-7347	108	6	is	be	AUX
ajst-7347	108	7	5	5	NUM
ajst-7347	108	8	,	,	PUNCT
ajst-7347	108	9	it	it	PRON
ajst-7347	108	10	can	can	AUX
ajst-7347	108	11	be	be	AUX
ajst-7347	108	12	observed	observe	VERB
ajst-7347	108	13	that	that	SCONJ
ajst-7347	108	14	the	the	DET
ajst-7347	108	15	loss	loss	NOUN
ajst-7347	108	16	value	value	NOUN
ajst-7347	108	17	of	of	ADP
ajst-7347	108	18	the	the	DET
ajst-7347	108	19	lstm	lstm	PROPN
ajst-7347	108	20	model	model	NOUN
ajst-7347	108	21	is	be	AUX
ajst-7347	108	22	about	about	ADV
ajst-7347	108	23	2.9	2.9	NUM
ajst-7347	108	24	,	,	PUNCT
ajst-7347	108	25	the	the	DET
ajst-7347	108	26	loss	loss	NOUN
ajst-7347	108	27	value	value	NOUN
ajst-7347	108	28	of	of	ADP
ajst-7347	108	29	the	the	DET
ajst-7347	108	30	cnn	cnn	PROPN
ajst-7347	108	31	-	-	PUNCT
ajst-7347	108	32	gru	gru	PROPN
ajst-7347	108	33	model	model	NOUN
ajst-7347	108	34	is	be	AUX
ajst-7347	108	35	about	about	ADV
ajst-7347	108	36	3	3	NUM
ajst-7347	108	37	,	,	PUNCT
ajst-7347	108	38	and	and	CCONJ
ajst-7347	108	39	the	the	DET
ajst-7347	108	40	training	training	NOUN
ajst-7347	108	41	set	set	VERB
ajst-7347	108	42	loss	loss	NOUN
ajst-7347	108	43	value	value	NOUN
ajst-7347	108	44	and	and	CCONJ
ajst-7347	108	45	test	test	NOUN
ajst-7347	108	46	set	set	VERB
ajst-7347	108	47	loss	loss	NOUN
ajst-7347	108	48	value	value	NOUN
ajst-7347	108	49	of	of	ADP
ajst-7347	108	50	the	the	DET
ajst-7347	108	51	cnn	cnn	PROPN
ajst-7347	108	52	-	-	PUNCT
ajst-7347	108	53	bigru	bigru	PROPN
ajst-7347	108	54	model	model	NOUN
ajst-7347	108	55	are	be	AUX
ajst-7347	108	56	both	both	ADV
ajst-7347	108	57	less	less	ADJ
ajst-7347	108	58	than	than	ADP
ajst-7347	108	59	3	3	NUM
ajst-7347	108	60	.	.	PUNCT
ajst-7347	109	1	it	it	PRON
ajst-7347	109	2	can	can	AUX
ajst-7347	109	3	be	be	AUX
ajst-7347	109	4	concluded	conclude	VERB
ajst-7347	109	5	that	that	SCONJ
ajst-7347	109	6	the	the	DET
ajst-7347	109	7	model	model	NOUN
ajst-7347	109	8	used	use	VERB
ajst-7347	109	9	in	in	ADP
ajst-7347	109	10	this	this	DET
ajst-7347	109	11	article	article	NOUN
ajst-7347	109	12	,	,	PUNCT
ajst-7347	109	13	cnn	cnn	PROPN
ajst-7347	109	14	-	-	PUNCT
ajst-7347	109	15	bigru	bigru	PROPN
ajst-7347	109	16	,	,	PUNCT
ajst-7347	109	17	can	can	AUX
ajst-7347	109	18	achieve	achieve	VERB
ajst-7347	109	19	the	the	DET
ajst-7347	109	20	minimum	minimum	ADJ
ajst-7347	109	21	loss	loss	NOUN
ajst-7347	109	22	value	value	NOUN
ajst-7347	109	23	faster	fast	ADV
ajst-7347	109	24	.	.	PUNCT
ajst-7347	110	1	continue	continue	VERB
ajst-7347	110	2	to	to	PART
ajst-7347	110	3	observe	observe	VERB
ajst-7347	110	4	that	that	SCONJ
ajst-7347	110	5	the	the	DET
ajst-7347	110	6	lstm	lstm	PROPN
ajst-7347	110	7	's	's	PART
ajst-7347	110	8	loss	loss	NOUN
ajst-7347	110	9	value	value	NOUN
ajst-7347	110	10	in	in	ADP
ajst-7347	110	11	figure	figure	NOUN
ajst-7347	110	12	6	6	NUM
ajst-7347	110	13	(	(	PUNCT
ajst-7347	110	14	a	a	PRON
ajst-7347	110	15	)	)	PUNCT
ajst-7347	110	16	reaches	reach	VERB
ajst-7347	110	17	the	the	DET
ajst-7347	110	18	minimum	minimum	ADJ
ajst-7347	110	19	value	value	NOUN
ajst-7347	110	20	when	when	SCONJ
ajst-7347	110	21	epoch	epoch	NOUN
ajst-7347	110	22	is	be	AUX
ajst-7347	110	23	approximately	approximately	ADV
ajst-7347	110	24	45	45	NUM
ajst-7347	110	25	times	time	NOUN
ajst-7347	110	26	,	,	PUNCT
ajst-7347	110	27	while	while	SCONJ
ajst-7347	110	28	the	the	DET
ajst-7347	110	29	cnn	cnn	PROPN
ajst-7347	110	30	-	-	PUNCT
ajst-7347	110	31	gru	gru	PROPN
ajst-7347	110	32	model	model	NOUN
ajst-7347	110	33	in	in	ADP
ajst-7347	110	34	figure	figure	NOUN
ajst-7347	110	35	6	6	NUM
ajst-7347	110	36	(	(	PUNCT
ajst-7347	110	37	b	b	NOUN
ajst-7347	110	38	)	)	PUNCT
ajst-7347	110	39	reaches	reach	VERB
ajst-7347	110	40	the	the	DET
ajst-7347	110	41	minimum	minimum	ADJ
ajst-7347	110	42	value	value	NOUN
ajst-7347	110	43	when	when	SCONJ
ajst-7347	110	44	epoch	epoch	PROPN
ajst-7347	110	45	is	be	AUX
ajst-7347	110	46	50	50	NUM
ajst-7347	110	47	times	time	NOUN
ajst-7347	110	48	,	,	PUNCT
ajst-7347	110	49	and	and	CCONJ
ajst-7347	110	50	the	the	DET
ajst-7347	110	51	number	number	NOUN
ajst-7347	110	52	of	of	ADP
ajst-7347	110	53	iterations	iteration	NOUN
ajst-7347	110	54	for	for	ADP
ajst-7347	110	55	the	the	DET
ajst-7347	110	56	cnn	cnn	PROPN
ajst-7347	110	57	-	-	PUNCT
ajst-7347	110	58	bigru	bigru	PROPN
ajst-7347	110	59	model	model	NOUN
ajst-7347	110	60	reaches	reach	VERB
ajst-7347	110	61	the	the	DET
ajst-7347	110	62	minimum	minimum	ADJ
ajst-7347	110	63	value	value	NOUN
ajst-7347	110	64	when	when	SCONJ
ajst-7347	110	65	the	the	DET
ajst-7347	110	66	loss	loss	NOUN
ajst-7347	110	67	reaches	reach	VERB
ajst-7347	110	68	the	the	DET
ajst-7347	110	69	minimum	minimum	ADJ
ajst-7347	110	70	value	value	NOUN
ajst-7347	110	71	is	be	AUX
ajst-7347	110	72	approximately	approximately	ADV
ajst-7347	110	73	30	30	NUM
ajst-7347	110	74	.	.	PUNCT
ajst-7347	111	1	figure	figure	NOUN
ajst-7347	111	2	6	6	NUM
ajst-7347	111	3	.	.	PUNCT
ajst-7347	112	1	loss	loss	NOUN
ajst-7347	112	2	iteration	iteration	NOUN
ajst-7347	112	3	diagram	diagram	NOUN
ajst-7347	112	4	of	of	ADP
ajst-7347	112	5	each	each	DET
ajst-7347	112	6	model	model	NOUN
ajst-7347	112	7	from	from	ADP
ajst-7347	112	8	the	the	DET
ajst-7347	112	9	histograms	histogram	NOUN
ajst-7347	112	10	7	7	NUM
ajst-7347	112	11	and	and	CCONJ
ajst-7347	112	12	8	8	NUM
ajst-7347	112	13	,	,	PUNCT
ajst-7347	112	14	it	it	PRON
ajst-7347	112	15	can	can	AUX
ajst-7347	112	16	be	be	AUX
ajst-7347	112	17	seen	see	VERB
ajst-7347	112	18	that	that	SCONJ
ajst-7347	112	19	the	the	DET
ajst-7347	112	20	cnnbigru	cnnbigru	PROPN
ajst-7347	112	21	model	model	NOUN
ajst-7347	112	22	has	have	VERB
ajst-7347	112	23	the	the	DET
ajst-7347	112	24	lowest	low	ADJ
ajst-7347	112	25	mae	mae	PROPN
ajst-7347	112	26	and	and	CCONJ
ajst-7347	112	27	rmse	rmse	ADJ
ajst-7347	112	28	values	value	NOUN
ajst-7347	112	29	at	at	ADP
ajst-7347	112	30	the	the	DET
ajst-7347	112	31	three	three	NUM
ajst-7347	112	32	target	target	NOUN
ajst-7347	112	33	monitoring	monitoring	NOUN
ajst-7347	112	34	points	point	NOUN
ajst-7347	112	35	,	,	PUNCT
ajst-7347	112	36	and	and	CCONJ
ajst-7347	112	37	the	the	DET
ajst-7347	112	38	mae	mae	PROPN
ajst-7347	112	39	values	value	NOUN
ajst-7347	112	40	of	of	ADP
ajst-7347	112	41	the	the	DET
ajst-7347	112	42	four	four	NUM
ajst-7347	112	43	models	model	NOUN
ajst-7347	112	44	are	be	AUX
ajst-7347	112	45	lstm	lstm	PROPN
ajst-7347	112	46	>	>	X
ajst-7347	112	47	gru	gru	PROPN
ajst-7347	112	48	>	>	PROPN
ajst-7347	112	49	cnn	cnn	PROPN
ajst-7347	112	50	-	-	PUNCT
ajst-7347	112	51	gru	gru	PROPN
ajst-7347	112	52	>	>	X
ajst-7347	112	53	cnnbigru	cnnbigru	PROPN
ajst-7347	112	54	from	from	ADP
ajst-7347	112	55	high	high	ADJ
ajst-7347	112	56	to	to	PART
ajst-7347	112	57	low	low	VERB
ajst-7347	112	58	among	among	ADP
ajst-7347	112	59	them	they	PRON
ajst-7347	112	60	,	,	PUNCT
ajst-7347	112	61	the	the	DET
ajst-7347	112	62	mae	mae	PROPN
ajst-7347	112	63	predicted	predict	VERB
ajst-7347	112	64	by	by	ADP
ajst-7347	112	65	each	each	DET
ajst-7347	112	66	model	model	NOUN
ajst-7347	112	67	at	at	ADP
ajst-7347	112	68	the	the	DET
ajst-7347	112	69	monitoring	monitoring	NOUN
ajst-7347	112	70	points	point	NOUN
ajst-7347	112	71	in	in	ADP
ajst-7347	112	72	xindian	xindian	ADJ
ajst-7347	112	73	district	district	NOUN
ajst-7347	112	74	were	be	AUX
ajst-7347	112	75	2.5932	2.5932	NUM
ajst-7347	112	76	,	,	PUNCT
ajst-7347	112	77	2.6157	2.6157	NUM
ajst-7347	112	78	,	,	PUNCT
ajst-7347	112	79	2.4596	2.4596	NUM
ajst-7347	112	80	,	,	PUNCT
ajst-7347	112	81	and	and	CCONJ
ajst-7347	112	82	8.24341	8.24341	NUM
ajst-7347	112	83	.	.	PUNCT
ajst-7347	113	1	the	the	DET
ajst-7347	113	2	mae	mae	PROPN
ajst-7347	113	3	value	value	NOUN
ajst-7347	113	4	of	of	ADP
ajst-7347	113	5	the	the	DET
ajst-7347	113	6	cnnbigru	cnnbigru	PROPN
ajst-7347	113	7	model	model	NOUN
ajst-7347	113	8	in	in	ADP
ajst-7347	113	9	xindian	xindian	ADJ
ajst-7347	113	10	district	district	NOUN
ajst-7347	113	11	increased	increase	VERB
ajst-7347	113	12	6.5	6.5	NUM
ajst-7347	113	13	%	%	NOUN
ajst-7347	113	14	compared	compare	VERB
ajst-7347	113	15	to	to	ADP
ajst-7347	113	16	the	the	DET
ajst-7347	113	17	mae	mae	PROPN
ajst-7347	113	18	value	value	NOUN
ajst-7347	113	19	of	of	ADP
ajst-7347	113	20	lstm	lstm	PROPN
ajst-7347	113	21	,	,	PUNCT
ajst-7347	113	22	increased	increase	VERB
ajst-7347	113	23	by	by	ADP
ajst-7347	113	24	7.5	7.5	NUM
ajst-7347	113	25	%	%	NOUN
ajst-7347	113	26	compared	compare	VERB
ajst-7347	113	27	to	to	ADP
ajst-7347	113	28	gru	gru	VERB
ajst-7347	113	29	,	,	PUNCT
ajst-7347	113	30	and	and	CCONJ
ajst-7347	113	31	increased	increase	VERB
ajst-7347	113	32	by	by	ADP
ajst-7347	113	33	1	1	NUM
ajst-7347	113	34	%	%	NOUN
ajst-7347	113	35	compared	compare	VERB
ajst-7347	113	36	to	to	ADP
ajst-7347	113	37	cnn	cnn	PROPN
ajst-7347	113	38	-	-	PUNCT
ajst-7347	113	39	gru	gru	PROPN
ajst-7347	113	40	.	.	PUNCT
ajst-7347	114	1	the	the	DET
ajst-7347	114	2	monitoring	monitoring	NOUN
ajst-7347	114	3	points	point	NOUN
ajst-7347	114	4	in	in	ADP
ajst-7347	114	5	taoyuan	taoyuan	PROPN
ajst-7347	114	6	district	district	PROPN
ajst-7347	114	7	increased	increase	VERB
ajst-7347	114	8	by	by	ADP
ajst-7347	114	9	1.6	1.6	NUM
ajst-7347	114	10	%	%	NOUN
ajst-7347	114	11	,	,	PUNCT
ajst-7347	114	12	1.3	1.3	NUM
ajst-7347	114	13	%	%	NOUN
ajst-7347	114	14	,	,	PUNCT
ajst-7347	114	15	and	and	CCONJ
ajst-7347	114	16	1.6	1.6	NUM
ajst-7347	114	17	%	%	NOUN
ajst-7347	114	18	respectively	respectively	ADV
ajst-7347	114	19	.	.	PUNCT
ajst-7347	115	1	the	the	DET
ajst-7347	115	2	monitoring	monitoring	NOUN
ajst-7347	115	3	points	point	NOUN
ajst-7347	115	4	in	in	ADP
ajst-7347	115	5	songshan	songshan	NOUN
ajst-7347	115	6	district	district	NOUN
ajst-7347	115	7	increased	increase	VERB
ajst-7347	115	8	by	by	ADP
ajst-7347	115	9	2.5	2.5	NUM
ajst-7347	115	10	%	%	NOUN
ajst-7347	115	11	,	,	PUNCT
ajst-7347	115	12	2.4	2.4	NUM
ajst-7347	115	13	%	%	NOUN
ajst-7347	115	14	,	,	PUNCT
ajst-7347	115	15	0.2	0.2	NUM
ajst-7347	115	16	%	%	NOUN
ajst-7347	115	17	model	model	PROPN
ajst-7347	115	18	cnn	cnn	PROPN
ajst-7347	115	19	-	-	PUNCT
ajst-7347	115	20	bigru	bigru	PROPN
ajst-7347	115	21	has	have	VERB
ajst-7347	115	22	significant	significant	ADJ
ajst-7347	115	23	advantages	advantage	NOUN
ajst-7347	115	24	over	over	ADP
ajst-7347	115	25	single	single	ADJ
ajst-7347	115	26	model	model	NOUN
ajst-7347	115	27	lstm	lstm	PROPN
ajst-7347	115	28	and	and	CCONJ
ajst-7347	115	29	gru	gru	PROPN
ajst-7347	115	30	.	.	PROPN
ajst-7347	116	1	among	among	ADP
ajst-7347	116	2	the	the	DET
ajst-7347	116	3	different	different	ADJ
ajst-7347	116	4	evaluation	evaluation	NOUN
ajst-7347	116	5	index	index	NOUN
ajst-7347	116	6	values	value	NOUN
ajst-7347	116	7	for	for	ADP
ajst-7347	116	8	the	the	DET
ajst-7347	116	9	three	three	NUM
ajst-7347	116	10	target	target	NOUN
ajst-7347	116	11	monitoring	monitoring	NOUN
ajst-7347	116	12	points	point	NOUN
ajst-7347	116	13	,	,	PUNCT
ajst-7347	116	14	the	the	DET
ajst-7347	116	15	r2	r2	PROPN
ajst-7347	116	16	predicted	predict	VERB
ajst-7347	116	17	by	by	ADP
ajst-7347	116	18	the	the	DET
ajst-7347	116	19	combined	combine	VERB
ajst-7347	116	20	model	model	NOUN
ajst-7347	116	21	for	for	ADP
ajst-7347	116	22	pm2.5	pm2.5	DET
ajst-7347	116	23	concentration	concentration	NOUN
ajst-7347	116	24	at	at	ADP
ajst-7347	116	25	the	the	DET
ajst-7347	116	26	three	three	NUM
ajst-7347	116	27	stations	station	NOUN
ajst-7347	116	28	reached	reach	VERB
ajst-7347	116	29	0.9111	0.9111	NUM
ajst-7347	116	30	,	,	PUNCT
ajst-7347	116	31	0.8890	0.8890	NUM
ajst-7347	116	32	,	,	PUNCT
ajst-7347	116	33	and	and	CCONJ
ajst-7347	116	34	0.8768	0.8768	NUM
ajst-7347	116	35	,	,	PUNCT
ajst-7347	116	36	respectively	respectively	ADV
ajst-7347	116	37	.	.	PUNCT
ajst-7347	117	1	compared	compare	VERB
ajst-7347	117	2	with	with	ADP
ajst-7347	117	3	lstm	lstm	PROPN
ajst-7347	117	4	's	's	PART
ajst-7347	117	5	0.8901	0.8901	NUM
ajst-7347	117	6	,	,	PUNCT
ajst-7347	117	7	0.8847	0.8847	NUM
ajst-7347	117	8	,	,	PUNCT
ajst-7347	117	9	and	and	CCONJ
ajst-7347	117	10	0.8593	0.8593	NUM
ajst-7347	117	11	,	,	PUNCT
ajst-7347	117	12	the	the	DET
ajst-7347	117	13	prediction	prediction	NOUN
ajst-7347	117	14	accuracy	accuracy	NOUN
ajst-7347	117	15	of	of	ADP
ajst-7347	117	16	the	the	DET
ajst-7347	117	17	combined	combined	ADJ
ajst-7347	117	18	model	model	NOUN
ajst-7347	117	19	for	for	ADP
ajst-7347	117	20	pm2.5	pm2.5	DET
ajst-7347	117	21	concentration	concentration	NOUN
ajst-7347	117	22	increased	increase	VERB
ajst-7347	117	23	by	by	ADP
ajst-7347	117	24	2.3	2.3	NUM
ajst-7347	117	25	%	%	NOUN
ajst-7347	117	26	,	,	PUNCT
ajst-7347	117	27	0.49	0.49	NUM
ajst-7347	117	28	%	%	NOUN
ajst-7347	117	29	,	,	PUNCT
ajst-7347	117	30	and	and	CCONJ
ajst-7347	117	31	2	2	NUM
ajst-7347	117	32	%	%	NOUN
ajst-7347	117	33	,	,	PUNCT
ajst-7347	117	34	respectively	respectively	ADV
ajst-7347	117	35	.	.	PUNCT
ajst-7347	118	1	compared	compare	VERB
ajst-7347	118	2	with	with	ADP
ajst-7347	118	3	gru	gru	PROPN
ajst-7347	118	4	's	's	PART
ajst-7347	118	5	0.8949	0.8949	NUM
ajst-7347	118	6	,	,	PUNCT
ajst-7347	118	7	0.8848	0.8848	NUM
ajst-7347	118	8	,	,	PUNCT
ajst-7347	118	9	and	and	CCONJ
ajst-7347	118	10	0.8613	0.8613	NUM
ajst-7347	118	11	,	,	PUNCT
ajst-7347	118	12	the	the	DET
ajst-7347	118	13	prediction	prediction	NOUN
ajst-7347	118	14	accuracy	accuracy	NOUN
ajst-7347	118	15	of	of	ADP
ajst-7347	118	16	the	the	DET
ajst-7347	118	17	combined	combined	ADJ
ajst-7347	118	18	model	model	NOUN
ajst-7347	118	19	for	for	ADP
ajst-7347	118	20	pm2.5	pm2.5	DET
ajst-7347	118	21	concentration	concentration	NOUN
ajst-7347	118	22	increased	increase	VERB
ajst-7347	118	23	by	by	ADP
ajst-7347	118	24	1.8	1.8	NUM
ajst-7347	118	25	%	%	NOUN
ajst-7347	118	26	,	,	PUNCT
ajst-7347	118	27	respectively	respectively	ADV
ajst-7347	118	28	0.5	0.5	NUM
ajst-7347	118	29	%	%	NOUN
ajst-7347	118	30	and	and	CCONJ
ajst-7347	118	31	1.8	1.8	NUM
ajst-7347	118	32	%	%	NOUN
ajst-7347	118	33	.	.	PUNCT
ajst-7347	119	1	6	6	NUM
ajst-7347	119	2	figure	figure	NOUN
ajst-7347	119	3	7	7	NUM
ajst-7347	119	4	.	.	PUNCT
ajst-7347	120	1	mae	mae	PROPN
ajst-7347	120	2	diagram	diagram	PROPN
ajst-7347	120	3	of	of	ADP
ajst-7347	120	4	each	each	DET
ajst-7347	120	5	model	model	NOUN
ajst-7347	120	6	figure	figure	NOUN
ajst-7347	120	7	8	8	NUM
ajst-7347	120	8	.	.	PUNCT
ajst-7347	121	1	mae	mae	PROPN
ajst-7347	121	2	diagram	diagram	PROPN
ajst-7347	121	3	of	of	ADP
ajst-7347	121	4	each	each	DET
ajst-7347	121	5	model	model	NOUN
ajst-7347	121	6	the	the	DET
ajst-7347	121	7	𝑅	𝑅	PROPN
ajst-7347	121	8	of	of	ADP
ajst-7347	121	9	the	the	DET
ajst-7347	121	10	cnn	cnn	PROPN
ajst-7347	121	11	-	-	PUNCT
ajst-7347	121	12	gru	gru	PROPN
ajst-7347	121	13	model	model	NOUN
ajst-7347	121	14	is	be	AUX
ajst-7347	121	15	0.9100	0.9100	NUM
ajst-7347	121	16	,	,	PUNCT
ajst-7347	121	17	0.8876	0.8876	NUM
ajst-7347	121	18	,	,	PUNCT
ajst-7347	121	19	and	and	CCONJ
ajst-7347	121	20	0.8692	0.8692	NUM
ajst-7347	121	21	,	,	PUNCT
ajst-7347	121	22	respectively	respectively	ADV
ajst-7347	121	23	,	,	PUNCT
ajst-7347	121	24	and	and	CCONJ
ajst-7347	121	25	the	the	DET
ajst-7347	121	26	prediction	prediction	NOUN
ajst-7347	121	27	accuracy	accuracy	NOUN
ajst-7347	121	28	of	of	ADP
ajst-7347	121	29	the	the	DET
ajst-7347	121	30	proposed	propose	VERB
ajst-7347	121	31	combined	combine	VERB
ajst-7347	121	32	model	model	NOUN
ajst-7347	121	33	has	have	AUX
ajst-7347	121	34	been	be	AUX
ajst-7347	121	35	improved	improve	VERB
ajst-7347	121	36	.	.	PUNCT
ajst-7347	122	1	its	its	PRON
ajst-7347	122	2	rmse	rmse	NOUN
ajst-7347	122	3	can	can	AUX
ajst-7347	122	4	also	also	ADV
ajst-7347	122	5	be	be	AUX
ajst-7347	122	6	calculated	calculate	VERB
ajst-7347	122	7	as	as	ADP
ajst-7347	122	8	a	a	DET
ajst-7347	122	9	percentage	percentage	NOUN
ajst-7347	122	10	of	of	ADP
ajst-7347	122	11	its	its	PRON
ajst-7347	122	12	improvement.it	improvement.it	PRON
ajst-7347	122	13	can	can	AUX
ajst-7347	122	14	be	be	AUX
ajst-7347	122	15	concluded	conclude	VERB
ajst-7347	122	16	that	that	SCONJ
ajst-7347	122	17	this	this	DET
ajst-7347	122	18	model	model	NOUN
ajst-7347	122	19	has	have	VERB
ajst-7347	122	20	a	a	DET
ajst-7347	122	21	significant	significant	ADJ
ajst-7347	122	22	improvement	improvement	NOUN
ajst-7347	122	23	compared	compare	VERB
ajst-7347	122	24	to	to	ADP
ajst-7347	122	25	lstm	lstm	PROPN
ajst-7347	122	26	,	,	PUNCT
ajst-7347	122	27	especially	especially	ADV
ajst-7347	122	28	for	for	ADP
ajst-7347	122	29	monitoring	monitor	VERB
ajst-7347	122	30	points	point	NOUN
ajst-7347	122	31	in	in	ADP
ajst-7347	122	32	the	the	DET
ajst-7347	122	33	new	new	ADJ
ajst-7347	122	34	store	store	NOUN
ajst-7347	122	35	area	area	NOUN
ajst-7347	122	36	.	.	PUNCT
ajst-7347	123	1	compared	compare	VERB
ajst-7347	123	2	to	to	ADP
ajst-7347	123	3	the	the	DET
ajst-7347	123	4	combined	combine	VERB
ajst-7347	123	5	module	module	NOUN
ajst-7347	123	6	cnngru	cnngru	NOUN
ajst-7347	123	7	,	,	PUNCT
ajst-7347	123	8	the	the	DET
ajst-7347	123	9	improvement	improvement	NOUN
ajst-7347	123	10	is	be	AUX
ajst-7347	123	11	not	not	PART
ajst-7347	123	12	significant	significant	ADJ
ajst-7347	123	13	,	,	PUNCT
ajst-7347	123	14	but	but	CCONJ
ajst-7347	123	15	there	there	PRON
ajst-7347	123	16	is	be	VERB
ajst-7347	123	17	also	also	ADV
ajst-7347	123	18	progress	progress	NOUN
ajst-7347	123	19	,	,	PUNCT
ajst-7347	123	20	because	because	SCONJ
ajst-7347	123	21	cnn	cnn	PROPN
ajst-7347	123	22	-	-	PUNCT
ajst-7347	123	23	gru	gru	PROPN
ajst-7347	123	24	also	also	ADV
ajst-7347	123	25	combines	combine	VERB
ajst-7347	123	26	the	the	DET
ajst-7347	123	27	ability	ability	NOUN
ajst-7347	123	28	of	of	ADP
ajst-7347	123	29	cnn	cnn	PROPN
ajst-7347	123	30	to	to	ADP
ajst-7347	123	31	mine	mine	PRON
ajst-7347	123	32	local	local	ADJ
ajst-7347	123	33	features	feature	NOUN
ajst-7347	123	34	of	of	ADP
ajst-7347	123	35	data	datum	NOUN
ajst-7347	123	36	and	and	CCONJ
ajst-7347	123	37	the	the	DET
ajst-7347	123	38	ability	ability	NOUN
ajst-7347	123	39	of	of	ADP
ajst-7347	123	40	gru	gru	NOUN
ajst-7347	123	41	to	to	PART
ajst-7347	123	42	process	process	VERB
ajst-7347	123	43	long	long	ADJ
ajst-7347	123	44	-	-	PUNCT
ajst-7347	123	45	term	term	NOUN
ajst-7347	123	46	and	and	CCONJ
ajst-7347	123	47	long	long	ADJ
ajst-7347	123	48	-	-	PUNCT
ajst-7347	123	49	term	term	NOUN
ajst-7347	123	50	memory	memory	NOUN
ajst-7347	123	51	.	.	PUNCT
ajst-7347	124	1	therefore	therefore	ADV
ajst-7347	124	2	,	,	PUNCT
ajst-7347	124	3	it	it	PRON
ajst-7347	124	4	can	can	AUX
ajst-7347	124	5	be	be	AUX
ajst-7347	124	6	concluded	conclude	VERB
ajst-7347	124	7	that	that	SCONJ
ajst-7347	124	8	the	the	DET
ajst-7347	124	9	combination	combination	NOUN
ajst-7347	124	10	model	model	NOUN
ajst-7347	124	11	is	be	AUX
ajst-7347	124	12	better	well	ADJ
ajst-7347	124	13	than	than	ADP
ajst-7347	124	14	the	the	DET
ajst-7347	124	15	single	single	ADJ
ajst-7347	124	16	model	model	NOUN
ajst-7347	124	17	,	,	PUNCT
ajst-7347	124	18	and	and	CCONJ
ajst-7347	124	19	the	the	DET
ajst-7347	124	20	proposed	propose	VERB
ajst-7347	124	21	combination	combination	NOUN
ajst-7347	124	22	model	model	NOUN
ajst-7347	124	23	has	have	VERB
ajst-7347	124	24	good	good	ADJ
ajst-7347	124	25	prediction	prediction	NOUN
ajst-7347	124	26	performance	performance	NOUN
ajst-7347	124	27	in	in	ADP
ajst-7347	124	28	most	most	ADJ
ajst-7347	124	29	cases	case	NOUN
ajst-7347	124	30	.	.	PUNCT
ajst-7347	125	1	figure	figure	NOUN
ajst-7347	125	2	9	9	NUM
ajst-7347	125	3	.	.	PUNCT
ajst-7347	126	1	mae	mae	PROPN
ajst-7347	126	2	diagram	diagram	PROPN
ajst-7347	126	3	and	and	CCONJ
ajst-7347	126	4	rmse	rmse	ADJ
ajst-7347	126	5	diagram	diagram	NOUN
ajst-7347	126	6	of	of	ADP
ajst-7347	126	7	each	each	DET
ajst-7347	126	8	model	model	NOUN
ajst-7347	126	9	7	7	NUM
ajst-7347	126	10	data	datum	NOUN
ajst-7347	126	11	from	from	ADP
ajst-7347	126	12	10	10	NUM
ajst-7347	126	13	monitoring	monitoring	NOUN
ajst-7347	126	14	stations	station	NOUN
ajst-7347	126	15	in	in	ADP
ajst-7347	126	16	xinbei	xinbei	PROPN
ajst-7347	126	17	city	city	NOUN
ajst-7347	126	18	were	be	AUX
ajst-7347	126	19	selected	select	VERB
ajst-7347	126	20	,	,	PUNCT
ajst-7347	126	21	and	and	CCONJ
ajst-7347	126	22	the	the	DET
ajst-7347	126	23	average	average	ADJ
ajst-7347	126	24	values	value	NOUN
ajst-7347	126	25	of	of	ADP
ajst-7347	126	26	each	each	DET
ajst-7347	126	27	characteristic	characteristic	ADJ
ajst-7347	126	28	were	be	AUX
ajst-7347	126	29	taken	take	VERB
ajst-7347	126	30	to	to	PART
ajst-7347	126	31	predict	predict	VERB
ajst-7347	126	32	the	the	DET
ajst-7347	126	33	pm2.5	pm2.5	ADJ
ajst-7347	126	34	concentration	concentration	NOUN
ajst-7347	126	35	value	value	NOUN
ajst-7347	126	36	in	in	ADP
ajst-7347	126	37	xinbei	xinbei	PROPN
ajst-7347	126	38	city	city	NOUN
ajst-7347	126	39	.	.	PUNCT
ajst-7347	127	1	first	first	ADV
ajst-7347	127	2	,	,	PUNCT
ajst-7347	127	3	construct	construct	VERB
ajst-7347	127	4	input	input	NOUN
ajst-7347	127	5	features	feature	NOUN
ajst-7347	127	6	,	,	PUNCT
ajst-7347	127	7	select	select	VERB
ajst-7347	127	8	relevant	relevant	ADJ
ajst-7347	127	9	feature	feature	NOUN
ajst-7347	127	10	variables	variable	NOUN
ajst-7347	127	11	,	,	PUNCT
ajst-7347	127	12	and	and	CCONJ
ajst-7347	127	13	then	then	ADV
ajst-7347	127	14	adjust	adjust	VERB
ajst-7347	127	15	parameters	parameter	NOUN
ajst-7347	127	16	.	.	PUNCT
ajst-7347	128	1	input	input	VERB
ajst-7347	128	2	the	the	DET
ajst-7347	128	3	constructed	construct	VERB
ajst-7347	128	4	cnnbigru	cnnbigru	ADJ
ajst-7347	128	5	combination	combination	NOUN
ajst-7347	128	6	model	model	NOUN
ajst-7347	128	7	to	to	PART
ajst-7347	128	8	predict	predict	VERB
ajst-7347	128	9	the	the	DET
ajst-7347	128	10	pm2.5	pm2.5	ADJ
ajst-7347	128	11	concentration	concentration	NOUN
ajst-7347	128	12	in	in	ADP
ajst-7347	128	13	xinbei	xinbei	NOUN
ajst-7347	128	14	city	city	NOUN
ajst-7347	128	15	,	,	PUNCT
ajst-7347	128	16	and	and	CCONJ
ajst-7347	128	17	also	also	ADV
ajst-7347	128	18	select	select	VERB
ajst-7347	128	19	lstm	lstm	PROPN
ajst-7347	128	20	,	,	PUNCT
ajst-7347	128	21	gru	gru	PROPN
ajst-7347	128	22	,	,	PUNCT
ajst-7347	128	23	and	and	CCONJ
ajst-7347	128	24	cnn	cnn	PROPN
ajst-7347	128	25	-	-	PUNCT
ajst-7347	128	26	gru	gru	PROPN
ajst-7347	128	27	models	model	NOUN
ajst-7347	128	28	as	as	ADP
ajst-7347	128	29	comparison	comparison	NOUN
ajst-7347	128	30	models	model	NOUN
ajst-7347	128	31	.	.	PUNCT
ajst-7347	129	1	figure	figure	NOUN
ajst-7347	129	2	9	9	NUM
ajst-7347	129	3	shows	show	VERB
ajst-7347	129	4	the	the	DET
ajst-7347	129	5	histogram	histogram	NOUN
ajst-7347	129	6	of	of	ADP
ajst-7347	129	7	the	the	DET
ajst-7347	129	8	evaluation	evaluation	NOUN
ajst-7347	129	9	indicators	indicator	NOUN
ajst-7347	129	10	for	for	ADP
ajst-7347	129	11	predicting	predict	VERB
ajst-7347	129	12	pm2.5	pm2.5	DET
ajst-7347	129	13	concentration	concentration	NOUN
ajst-7347	129	14	in	in	ADP
ajst-7347	129	15	xinbei	xinbei	NOUN
ajst-7347	129	16	city	city	NOUN
ajst-7347	129	17	by	by	ADP
ajst-7347	129	18	all	all	DET
ajst-7347	129	19	models	model	NOUN
ajst-7347	129	20	.	.	PUNCT
ajst-7347	130	1	the	the	DET
ajst-7347	130	2	evaluation	evaluation	NOUN
ajst-7347	130	3	indicators	indicator	NOUN
ajst-7347	130	4	are	be	AUX
ajst-7347	130	5	mae	mae	PROPN
ajst-7347	130	6	and	and	CCONJ
ajst-7347	130	7	rmse	rmse	NOUN
ajst-7347	130	8	.	.	PUNCT
ajst-7347	131	1	it	it	PRON
ajst-7347	131	2	can	can	AUX
ajst-7347	131	3	be	be	AUX
ajst-7347	131	4	observed	observe	VERB
ajst-7347	131	5	that	that	SCONJ
ajst-7347	131	6	compared	compare	VERB
ajst-7347	131	7	to	to	ADP
ajst-7347	131	8	using	use	VERB
ajst-7347	131	9	data	datum	NOUN
ajst-7347	131	10	from	from	ADP
ajst-7347	131	11	a	a	DET
ajst-7347	131	12	single	single	ADJ
ajst-7347	131	13	station	station	NOUN
ajst-7347	131	14	to	to	PART
ajst-7347	131	15	predict	predict	VERB
ajst-7347	131	16	,	,	PUNCT
ajst-7347	131	17	the	the	DET
ajst-7347	131	18	value	value	NOUN
ajst-7347	131	19	of	of	ADP
ajst-7347	131	20	rmse	rmse	NOUN
ajst-7347	131	21	is	be	AUX
ajst-7347	131	22	higher	high	ADJ
ajst-7347	131	23	,	,	PUNCT
ajst-7347	131	24	which	which	PRON
ajst-7347	131	25	is	be	AUX
ajst-7347	131	26	due	due	ADJ
ajst-7347	131	27	to	to	ADP
ajst-7347	131	28	differences	difference	NOUN
ajst-7347	131	29	in	in	ADP
ajst-7347	131	30	pollutants	pollutant	NOUN
ajst-7347	131	31	and	and	CCONJ
ajst-7347	131	32	meteorological	meteorological	ADJ
ajst-7347	131	33	conditions	condition	NOUN
ajst-7347	131	34	across	across	ADP
ajst-7347	131	35	different	different	ADJ
ajst-7347	131	36	regions	region	NOUN
ajst-7347	131	37	,	,	PUNCT
ajst-7347	131	38	showing	show	VERB
ajst-7347	131	39	strong	strong	ADJ
ajst-7347	131	40	regional	regional	ADJ
ajst-7347	131	41	characteristics	characteristic	NOUN
ajst-7347	131	42	.	.	PUNCT
ajst-7347	132	1	however	however	ADV
ajst-7347	132	2	,	,	PUNCT
ajst-7347	132	3	in	in	ADP
ajst-7347	132	4	the	the	DET
ajst-7347	132	5	mae	mae	PROPN
ajst-7347	132	6	and	and	CCONJ
ajst-7347	132	7	rmse	rmse	ADJ
ajst-7347	132	8	diagrams	diagram	NOUN
ajst-7347	132	9	of	of	ADP
ajst-7347	132	10	each	each	DET
ajst-7347	132	11	model	model	NOUN
ajst-7347	132	12	,	,	PUNCT
ajst-7347	132	13	the	the	DET
ajst-7347	132	14	value	value	NOUN
ajst-7347	132	15	of	of	ADP
ajst-7347	132	16	cnn	cnn	PROPN
ajst-7347	132	17	-	-	PUNCT
ajst-7347	132	18	bigru	bigru	PROPN
ajst-7347	132	19	is	be	AUX
ajst-7347	132	20	the	the	DET
ajst-7347	132	21	lowest	low	ADJ
ajst-7347	132	22	,	,	PUNCT
ajst-7347	132	23	followed	follow	VERB
ajst-7347	132	24	by	by	ADP
ajst-7347	132	25	the	the	DET
ajst-7347	132	26	cnn	cnn	PROPN
ajst-7347	132	27	-	-	PUNCT
ajst-7347	132	28	gru	gru	PROPN
ajst-7347	132	29	model	model	NOUN
ajst-7347	132	30	,	,	PUNCT
ajst-7347	132	31	and	and	CCONJ
ajst-7347	132	32	the	the	DET
ajst-7347	132	33	error	error	NOUN
ajst-7347	132	34	value	value	NOUN
ajst-7347	132	35	of	of	ADP
ajst-7347	132	36	the	the	DET
ajst-7347	132	37	two	two	NUM
ajst-7347	132	38	single	single	ADJ
ajst-7347	132	39	models	model	NOUN
ajst-7347	132	40	is	be	AUX
ajst-7347	132	41	still	still	ADV
ajst-7347	132	42	the	the	DET
ajst-7347	132	43	highest	high	ADJ
ajst-7347	132	44	.	.	PUNCT
ajst-7347	133	1	therefore	therefore	ADV
ajst-7347	133	2	,	,	PUNCT
ajst-7347	133	3	it	it	PRON
ajst-7347	133	4	can	can	AUX
ajst-7347	133	5	be	be	AUX
ajst-7347	133	6	concluded	conclude	VERB
ajst-7347	133	7	that	that	SCONJ
ajst-7347	133	8	the	the	DET
ajst-7347	133	9	effect	effect	NOUN
ajst-7347	133	10	of	of	ADP
ajst-7347	133	11	this	this	DET
ajst-7347	133	12	combined	combine	VERB
ajst-7347	133	13	model	model	NOUN
ajst-7347	133	14	in	in	ADP
ajst-7347	133	15	testing	test	VERB
ajst-7347	133	16	the	the	DET
ajst-7347	133	17	pm2.5	pm2.5	ADJ
ajst-7347	133	18	concentration	concentration	NOUN
ajst-7347	133	19	in	in	ADP
ajst-7347	133	20	the	the	DET
ajst-7347	133	21	entire	entire	ADJ
ajst-7347	133	22	new	new	ADJ
ajst-7347	133	23	taipei	taipei	PROPN
ajst-7347	133	24	city	city	NOUN
ajst-7347	133	25	is	be	AUX
ajst-7347	133	26	still	still	ADV
ajst-7347	133	27	significant	significant	ADJ
ajst-7347	133	28	.	.	PUNCT
ajst-7347	134	1	in	in	ADP
ajst-7347	134	2	summary	summary	NOUN
ajst-7347	134	3	,	,	PUNCT
ajst-7347	134	4	the	the	DET
ajst-7347	134	5	cnn	cnn	PROPN
ajst-7347	134	6	-	-	PUNCT
ajst-7347	134	7	bigru	bigru	NOUN
ajst-7347	134	8	combination	combination	NOUN
ajst-7347	134	9	model	model	NOUN
ajst-7347	134	10	used	use	VERB
ajst-7347	134	11	in	in	ADP
ajst-7347	134	12	this	this	DET
ajst-7347	134	13	article	article	NOUN
ajst-7347	134	14	can	can	AUX
ajst-7347	134	15	fully	fully	ADV
ajst-7347	134	16	utilize	utilize	VERB
ajst-7347	134	17	the	the	DET
ajst-7347	134	18	advantages	advantage	NOUN
ajst-7347	134	19	of	of	ADP
ajst-7347	134	20	convolutional	convolutional	ADJ
ajst-7347	134	21	neural	neural	ADJ
ajst-7347	134	22	networks	network	NOUN
ajst-7347	134	23	and	and	CCONJ
ajst-7347	134	24	two	two	NUM
ajst-7347	134	25	-	-	PUNCT
ajst-7347	134	26	way	way	NOUN
ajst-7347	134	27	gated	gated	ADJ
ajst-7347	134	28	cyclic	cyclic	ADJ
ajst-7347	134	29	unit	unit	NOUN
ajst-7347	134	30	networks	network	NOUN
ajst-7347	134	31	.	.	PUNCT
ajst-7347	135	1	cnn	cnn	PROPN
ajst-7347	135	2	with	with	ADP
ajst-7347	135	3	strong	strong	ADJ
ajst-7347	135	4	feature	feature	NOUN
ajst-7347	135	5	extraction	extraction	NOUN
ajst-7347	135	6	capabilities	capability	NOUN
ajst-7347	135	7	will	will	AUX
ajst-7347	135	8	input	input	VERB
ajst-7347	135	9	output	output	NOUN
ajst-7347	135	10	features	feature	NOUN
ajst-7347	135	11	into	into	ADP
ajst-7347	135	12	the	the	DET
ajst-7347	135	13	bigru	bigru	NOUN
ajst-7347	135	14	network	network	NOUN
ajst-7347	135	15	,	,	PUNCT
ajst-7347	135	16	thereby	thereby	ADV
ajst-7347	135	17	avoiding	avoid	VERB
ajst-7347	135	18	the	the	DET
ajst-7347	135	19	singularity	singularity	NOUN
ajst-7347	135	20	of	of	ADP
ajst-7347	135	21	single	single	ADJ
ajst-7347	135	22	model	model	NOUN
ajst-7347	135	23	prediction	prediction	NOUN
ajst-7347	135	24	,	,	PUNCT
ajst-7347	135	25	and	and	CCONJ
ajst-7347	135	26	improving	improve	VERB
ajst-7347	135	27	the	the	DET
ajst-7347	135	28	prediction	prediction	NOUN
ajst-7347	135	29	ability	ability	NOUN
ajst-7347	135	30	.	.	PUNCT
ajst-7347	136	1	4	4	X
ajst-7347	136	2	.	.	X
ajst-7347	136	3	conclusion	conclusion	NOUN
ajst-7347	136	4	in	in	ADP
ajst-7347	136	5	this	this	DET
ajst-7347	136	6	paper	paper	NOUN
ajst-7347	136	7	,	,	PUNCT
ajst-7347	136	8	we	we	PRON
ajst-7347	136	9	conducted	conduct	VERB
ajst-7347	136	10	a	a	DET
ajst-7347	136	11	predictive	predictive	ADJ
ajst-7347	136	12	analysis	analysis	NOUN
ajst-7347	136	13	of	of	ADP
ajst-7347	136	14	pm2.5	pm2.5	DET
ajst-7347	136	15	concentration	concentration	NOUN
ajst-7347	136	16	data	datum	NOUN
ajst-7347	136	17	from	from	ADP
ajst-7347	136	18	three	three	NUM
ajst-7347	136	19	monitoring	monitoring	NOUN
ajst-7347	136	20	stations	station	NOUN
ajst-7347	136	21	in	in	ADP
ajst-7347	136	22	xinbei	xinbei	PROPN
ajst-7347	136	23	city	city	PROPN
ajst-7347	136	24	,	,	PUNCT
ajst-7347	136	25	taiwan	taiwan	PROPN
ajst-7347	136	26	province	province	PROPN
ajst-7347	136	27	.	.	PUNCT
ajst-7347	137	1	first	first	ADV
ajst-7347	137	2	,	,	PUNCT
ajst-7347	137	3	we	we	PRON
ajst-7347	137	4	selected	select	VERB
ajst-7347	137	5	characteristic	characteristic	ADJ
ajst-7347	137	6	variables	variable	NOUN
ajst-7347	137	7	with	with	ADP
ajst-7347	137	8	strong	strong	ADJ
ajst-7347	137	9	correlation	correlation	NOUN
ajst-7347	137	10	with	with	ADP
ajst-7347	137	11	them	they	PRON
ajst-7347	137	12	as	as	ADP
ajst-7347	137	13	predictive	predictive	ADJ
ajst-7347	137	14	auxiliary	auxiliary	ADJ
ajst-7347	137	15	factors	factor	NOUN
ajst-7347	137	16	,	,	PUNCT
ajst-7347	137	17	and	and	CCONJ
ajst-7347	137	18	then	then	ADV
ajst-7347	137	19	conducted	conduct	VERB
ajst-7347	137	20	modeling	modeling	NOUN
ajst-7347	137	21	and	and	CCONJ
ajst-7347	137	22	prediction	prediction	NOUN
ajst-7347	137	23	of	of	ADP
ajst-7347	137	24	pm2.5	pm2.5	DET
ajst-7347	137	25	concentration	concentration	NOUN
ajst-7347	137	26	.	.	PUNCT
ajst-7347	138	1	we	we	PRON
ajst-7347	138	2	analyzed	analyze	VERB
ajst-7347	138	3	three	three	NUM
ajst-7347	138	4	evaluation	evaluation	NOUN
ajst-7347	138	5	indicators	indicator	NOUN
ajst-7347	138	6	,	,	PUNCT
ajst-7347	138	7	and	and	CCONJ
ajst-7347	138	8	designed	design	VERB
ajst-7347	138	9	three	three	NUM
ajst-7347	138	10	models	model	NOUN
ajst-7347	138	11	for	for	ADP
ajst-7347	138	12	comparison	comparison	NOUN
ajst-7347	138	13	:	:	PUNCT
ajst-7347	138	14	lstm	lstm	ADJ
ajst-7347	138	15	model	model	PROPN
ajst-7347	138	16	,	,	PUNCT
ajst-7347	138	17	gru	gru	NOUN
ajst-7347	138	18	model	model	NOUN
ajst-7347	138	19	,	,	PUNCT
ajst-7347	138	20	and	and	CCONJ
ajst-7347	138	21	cnn	cnn	PROPN
ajst-7347	138	22	-	-	PUNCT
ajst-7347	138	23	gru	gru	PROPN
ajst-7347	138	24	model	model	NOUN
ajst-7347	138	25	.	.	PUNCT
ajst-7347	139	1	it	it	PRON
ajst-7347	139	2	was	be	AUX
ajst-7347	139	3	concluded	conclude	VERB
ajst-7347	139	4	that	that	SCONJ
ajst-7347	139	5	the	the	DET
ajst-7347	139	6	prediction	prediction	NOUN
ajst-7347	139	7	performance	performance	NOUN
ajst-7347	139	8	of	of	ADP
ajst-7347	139	9	the	the	DET
ajst-7347	139	10	cnn	cnn	PROPN
ajst-7347	139	11	-	-	PUNCT
ajst-7347	139	12	bigru	bigru	PROPN
ajst-7347	139	13	model	model	NOUN
ajst-7347	139	14	at	at	ADP
ajst-7347	139	15	the	the	DET
ajst-7347	139	16	three	three	NUM
ajst-7347	139	17	target	target	NOUN
ajst-7347	139	18	stations	station	NOUN
ajst-7347	139	19	was	be	AUX
ajst-7347	139	20	good	good	ADJ
ajst-7347	139	21	,	,	PUNCT
ajst-7347	139	22	finally	finally	ADV
ajst-7347	139	23	,	,	PUNCT
ajst-7347	139	24	a	a	DET
ajst-7347	139	25	prediction	prediction	NOUN
ajst-7347	139	26	analysis	analysis	NOUN
ajst-7347	139	27	was	be	AUX
ajst-7347	139	28	conducted	conduct	VERB
ajst-7347	139	29	on	on	ADP
ajst-7347	139	30	the	the	DET
ajst-7347	139	31	average	average	ADJ
ajst-7347	139	32	pm2.5	pm2.5	PART
ajst-7347	139	33	concentration	concentration	NOUN
ajst-7347	139	34	in	in	ADP
ajst-7347	139	35	xinbei	xinbei	PROPN
ajst-7347	139	36	city	city	NOUN
ajst-7347	139	37	,	,	PUNCT
ajst-7347	139	38	further	far	ADV
ajst-7347	139	39	verifying	verify	VERB
ajst-7347	139	40	that	that	SCONJ
ajst-7347	139	41	the	the	DET
ajst-7347	139	42	prediction	prediction	NOUN
ajst-7347	139	43	effect	effect	NOUN
ajst-7347	139	44	of	of	ADP
ajst-7347	139	45	the	the	DET
ajst-7347	139	46	combined	combined	ADJ
ajst-7347	139	47	model	model	NOUN
ajst-7347	139	48	is	be	AUX
ajst-7347	139	49	better	well	ADJ
ajst-7347	139	50	than	than	ADP
ajst-7347	139	51	that	that	PRON
ajst-7347	139	52	of	of	ADP
ajst-7347	139	53	a	a	DET
ajst-7347	139	54	single	single	ADJ
ajst-7347	139	55	model	model	NOUN
ajst-7347	139	56	,	,	PUNCT
ajst-7347	139	57	and	and	CCONJ
ajst-7347	139	58	that	that	SCONJ
ajst-7347	139	59	the	the	DET
ajst-7347	139	60	cnn	cnn	PROPN
ajst-7347	139	61	-	-	PUNCT
ajst-7347	139	62	bigru	bigru	PROPN
ajst-7347	139	63	model	model	NOUN
ajst-7347	139	64	has	have	VERB
ajst-7347	139	65	more	more	ADJ
ajst-7347	139	66	advantages	advantage	NOUN
ajst-7347	139	67	in	in	ADP
ajst-7347	139	68	terms	term	NOUN
ajst-7347	139	69	of	of	ADP
ajst-7347	139	70	measurement	measurement	NOUN
ajst-7347	139	71	performance	performance	NOUN
ajst-7347	139	72	.	.	PUNCT
ajst-7347	140	1	references	reference	NOUN
ajst-7347	140	2	[	[	X
ajst-7347	140	3	1	1	NUM
ajst-7347	140	4	]	]	X
ajst-7347	140	5	chen	chen	PROPN
ajst-7347	140	6	z	z	PROPN
ajst-7347	140	7	,	,	PUNCT
ajst-7347	140	8	chen	chen	PROPN
ajst-7347	140	9	d	d	PROPN
ajst-7347	140	10	,	,	PUNCT
ajst-7347	140	11	zhao	zhao	PROPN
ajst-7347	140	12	c	c	PROPN
ajst-7347	140	13	,	,	PUNCT
ajst-7347	141	1	et	et	PROPN
ajst-7347	141	2	al	al	PROPN
ajst-7347	141	3	.	.	PUNCT
ajst-7347	141	4	influence	influence	NOUN
ajst-7347	141	5	of	of	ADP
ajst-7347	141	6	meteorological	meteorological	ADJ
ajst-7347	141	7	conditions	condition	NOUN
ajst-7347	141	8	on	on	ADP
ajst-7347	141	9	pm2	pm2	NOUN
ajst-7347	141	10	.	.	PUNCT
ajst-7347	142	1	5	5	NUM
ajst-7347	142	2	concentrations	concentration	NOUN
ajst-7347	142	3	across	across	ADP
ajst-7347	142	4	china	china	PROPN
ajst-7347	142	5	:	:	PUNCT
ajst-7347	142	6	a	a	DET
ajst-7347	142	7	review	review	NOUN
ajst-7347	142	8	of	of	ADP
ajst-7347	142	9	methodology	methodology	NOUN
ajst-7347	142	10	and	and	CCONJ
ajst-7347	142	11	mechanism[j	mechanism[j	PROPN
ajst-7347	142	12	]	]	PUNCT
ajst-7347	142	13	.	.	PUNCT
ajst-7347	143	1	environment	environment	PROPN
ajst-7347	143	2	international	international	PROPN
ajst-7347	143	3	,	,	PUNCT
ajst-7347	143	4	2020	2020	NUM
ajst-7347	143	5	,	,	PUNCT
ajst-7347	143	6	139	139	NUM
ajst-7347	143	7	:	:	SYM
ajst-7347	143	8	105558	105558	NUM
ajst-7347	143	9	.	.	PUNCT
ajst-7347	144	1	[	[	X
ajst-7347	144	2	2	2	NUM
ajst-7347	144	3	]	]	PUNCT
ajst-7347	144	4	rajagopalan	rajagopalan	PROPN
ajst-7347	144	5	s	s	PROPN
ajst-7347	144	6	,	,	PUNCT
ajst-7347	144	7	al	al	ADJ
ajst-7347	144	8	-	-	PUNCT
ajst-7347	144	9	kindi	kindi	NOUN
ajst-7347	144	10	s	s	PART
ajst-7347	144	11	g	g	NOUN
ajst-7347	144	12	,	,	PUNCT
ajst-7347	144	13	brook	brook	PROPN
ajst-7347	144	14	r	r	PROPN
ajst-7347	144	15	d.	d.	PROPN
ajst-7347	144	16	air	air	PROPN
ajst-7347	144	17	pollution	pollution	PROPN
ajst-7347	144	18	and	and	CCONJ
ajst-7347	144	19	cardiovascular	cardiovascular	ADJ
ajst-7347	144	20	disease	disease	NOUN
ajst-7347	144	21	:	:	PUNCT
ajst-7347	144	22	jacc	jacc	VERB
ajst-7347	144	23	state	state	NOUN
ajst-7347	144	24	-	-	PUNCT
ajst-7347	144	25	of	of	ADP
ajst-7347	144	26	-	-	PUNCT
ajst-7347	144	27	the	the	DET
ajst-7347	144	28	-	-	PUNCT
ajst-7347	144	29	art	art	NOUN
ajst-7347	144	30	review[j	review[j	PROPN
ajst-7347	144	31	]	]	PUNCT
ajst-7347	144	32	.	.	PUNCT
ajst-7347	145	1	journal	journal	PROPN
ajst-7347	145	2	of	of	ADP
ajst-7347	145	3	the	the	DET
ajst-7347	145	4	american	american	PROPN
ajst-7347	145	5	college	college	PROPN
ajst-7347	145	6	of	of	ADP
ajst-7347	145	7	cardiology	cardiology	NOUN
ajst-7347	145	8	,	,	PUNCT
ajst-7347	145	9	2018	2018	NUM
ajst-7347	145	10	,	,	PUNCT
ajst-7347	145	11	72(17	72(17	NOUN
ajst-7347	145	12	):	):	PUNCT
ajst-7347	145	13	2054	2054	NUM
ajst-7347	145	14	-	-	SYM
ajst-7347	145	15	2070	2070	NUM
ajst-7347	145	16	.	.	PUNCT
ajst-7347	146	1	[	[	X
ajst-7347	146	2	3	3	X
ajst-7347	146	3	]	]	X
ajst-7347	146	4	hogrefe	hogrefe	ADJ
ajst-7347	146	5	c	c	NOUN
ajst-7347	146	6	,	,	PUNCT
ajst-7347	146	7	lynn	lynn	PROPN
ajst-7347	146	8	b	b	PROPN
ajst-7347	146	9	,	,	PUNCT
ajst-7347	146	10	goldberg	goldberg	PROPN
ajst-7347	146	11	r	r	PROPN
ajst-7347	146	12	,	,	PUNCT
ajst-7347	146	13	et	et	PROPN
ajst-7347	146	14	al	al	PROPN
ajst-7347	146	15	.	.	PUNCT
ajst-7347	147	1	a	a	DET
ajst-7347	147	2	combined	combined	ADJ
ajst-7347	147	3	model	model	NOUN
ajst-7347	147	4	–	–	PUNCT
ajst-7347	147	5	observation	observation	NOUN
ajst-7347	147	6	approach	approach	NOUN
ajst-7347	147	7	to	to	PART
ajst-7347	147	8	estimate	estimate	VERB
ajst-7347	147	9	historic	historic	ADJ
ajst-7347	147	10	gridded	gridde	VERB
ajst-7347	147	11	fields	field	NOUN
ajst-7347	147	12	of	of	ADP
ajst-7347	147	13	pm2	pm2	NOUN
ajst-7347	147	14	.	.	PUNCT
ajst-7347	148	1	5	5	NUM
ajst-7347	148	2	mass	mass	NOUN
ajst-7347	148	3	and	and	CCONJ
ajst-7347	148	4	species	species	NOUN
ajst-7347	148	5	concentrations[j	concentrations[j	NOUN
ajst-7347	148	6	]	]	PUNCT
ajst-7347	148	7	.	.	PUNCT
ajst-7347	149	1	atmospheric	atmospheric	PROPN
ajst-7347	149	2	environment	environment	PROPN
ajst-7347	149	3	,	,	PUNCT
ajst-7347	149	4	2009	2009	NUM
ajst-7347	149	5	,	,	PUNCT
ajst-7347	149	6	43(16	43(16	NUM
ajst-7347	149	7	):	):	PUNCT
ajst-7347	149	8	2561	2561	NUM
ajst-7347	149	9	-	-	SYM
ajst-7347	149	10	2570	2570	NUM
ajst-7347	149	11	.	.	PUNCT
ajst-7347	150	1	[	[	X
ajst-7347	150	2	4	4	X
ajst-7347	150	3	]	]	X
ajst-7347	150	4	zhang	zhang	PROPN
ajst-7347	150	5	y	y	PROPN
ajst-7347	150	6	,	,	PUNCT
ajst-7347	150	7	pun	pun	PROPN
ajst-7347	150	8	b	b	NOUN
ajst-7347	150	9	,	,	PUNCT
ajst-7347	150	10	wu	wu	PROPN
ajst-7347	150	11	s	s	PROPN
ajst-7347	150	12	y	y	PROPN
ajst-7347	150	13	,	,	PUNCT
ajst-7347	150	14	et	et	PROPN
ajst-7347	150	15	al	al	PROPN
ajst-7347	150	16	.	.	PUNCT
ajst-7347	150	17	application	application	NOUN
ajst-7347	150	18	and	and	CCONJ
ajst-7347	150	19	evaluation	evaluation	NOUN
ajst-7347	150	20	of	of	ADP
ajst-7347	150	21	two	two	NUM
ajst-7347	150	22	air	air	NOUN
ajst-7347	150	23	quality	quality	NOUN
ajst-7347	150	24	models	model	NOUN
ajst-7347	150	25	for	for	ADP
ajst-7347	150	26	particulate	particulate	NOUN
ajst-7347	150	27	matter	matter	NOUN
ajst-7347	150	28	for	for	ADP
ajst-7347	150	29	a	a	DET
ajst-7347	150	30	southeastern	southeastern	ADJ
ajst-7347	150	31	us	us	PROPN
ajst-7347	150	32	episode[j	episode[j	PROPN
ajst-7347	150	33	]	]	PUNCT
ajst-7347	150	34	.	.	PUNCT
ajst-7347	151	1	journal	journal	PROPN
ajst-7347	151	2	of	of	ADP
ajst-7347	151	3	the	the	DET
ajst-7347	151	4	air	air	PROPN
ajst-7347	151	5	&	&	CCONJ
ajst-7347	151	6	waste	waste	PROPN
ajst-7347	151	7	management	management	PROPN
ajst-7347	151	8	association	association	PROPN
ajst-7347	151	9	,	,	PUNCT
ajst-7347	151	10	2004	2004	NUM
ajst-7347	151	11	,	,	PUNCT
ajst-7347	151	12	54(12	54(12	NUM
ajst-7347	151	13	):	):	PUNCT
ajst-7347	151	14	1478	1478	NUM
ajst-7347	151	15	-	-	SYM
ajst-7347	151	16	1493	1493	NUM
ajst-7347	151	17	.	.	PUNCT
ajst-7347	152	1	[	[	X
ajst-7347	152	2	5	5	X
ajst-7347	152	3	]	]	X
ajst-7347	152	4	chuang	chuang	PROPN
ajst-7347	152	5	m	m	PROPN
ajst-7347	152	6	t	t	PROPN
ajst-7347	152	7	,	,	PUNCT
ajst-7347	152	8	fu	fu	PROPN
ajst-7347	152	9	j	j	PROPN
ajst-7347	152	10	s	s	PROPN
ajst-7347	152	11	,	,	PUNCT
ajst-7347	152	12	jang	jang	PROPN
ajst-7347	152	13	c	c	PROPN
ajst-7347	152	14	j	j	PROPN
ajst-7347	152	15	,	,	PUNCT
ajst-7347	152	16	et	et	PROPN
ajst-7347	152	17	al	al	PROPN
ajst-7347	152	18	.	.	PROPN
ajst-7347	152	19	simulation	simulation	NOUN
ajst-7347	152	20	of	of	ADP
ajst-7347	152	21	long	long	ADJ
ajst-7347	152	22	-	-	PUNCT
ajst-7347	152	23	range	range	NOUN
ajst-7347	152	24	transport	transport	NOUN
ajst-7347	152	25	aerosols	aerosol	NOUN
ajst-7347	152	26	from	from	ADP
ajst-7347	152	27	the	the	DET
ajst-7347	152	28	asian	asian	ADJ
ajst-7347	152	29	continent	continent	NOUN
ajst-7347	152	30	to	to	ADP
ajst-7347	152	31	taiwan	taiwan	PROPN
ajst-7347	152	32	by	by	ADP
ajst-7347	152	33	a	a	DET
ajst-7347	152	34	southward	southward	ADJ
ajst-7347	152	35	asian	asian	ADJ
ajst-7347	152	36	high	high	ADJ
ajst-7347	152	37	-	-	PUNCT
ajst-7347	152	38	pressure	pressure	NOUN
ajst-7347	152	39	system[j	system[j	NOUN
ajst-7347	152	40	]	]	PUNCT
ajst-7347	152	41	.	.	PUNCT
ajst-7347	153	1	science	science	NOUN
ajst-7347	153	2	of	of	ADP
ajst-7347	153	3	the	the	DET
ajst-7347	153	4	total	total	ADJ
ajst-7347	153	5	environment	environment	NOUN
ajst-7347	153	6	,	,	PUNCT
ajst-7347	153	7	2008	2008	NUM
ajst-7347	153	8	,	,	PUNCT
ajst-7347	153	9	406(1	406(1	PROPN
ajst-7347	153	10	-	-	SYM
ajst-7347	153	11	2	2	NUM
ajst-7347	153	12	):	):	PUNCT
ajst-7347	153	13	168	168	NUM
ajst-7347	153	14	-	-	SYM
ajst-7347	153	15	179	179	NUM
ajst-7347	153	16	.	.	PUNCT
ajst-7347	154	1	[	[	X
ajst-7347	154	2	6	6	NUM
ajst-7347	154	3	]	]	PUNCT
ajst-7347	154	4	san	san	PROPN
ajst-7347	154	5	josé	josé	PROPN
ajst-7347	154	6	r	r	PROPN
ajst-7347	154	7	,	,	PUNCT
ajst-7347	154	8	pérez	pérez	NOUN
ajst-7347	154	9	j	j	PROPN
ajst-7347	154	10	l	l	PROPN
ajst-7347	154	11	,	,	PUNCT
ajst-7347	154	12	morant	morant	PROPN
ajst-7347	154	13	j	j	PROPN
ajst-7347	154	14	l	l	PROPN
ajst-7347	154	15	,	,	PUNCT
ajst-7347	154	16	et	et	PROPN
ajst-7347	154	17	al	al	PROPN
ajst-7347	154	18	.	.	PROPN
ajst-7347	154	19	improved	improve	VERB
ajst-7347	154	20	modelling	model	VERB
ajst-7347	154	21	experiment	experiment	NOUN
ajst-7347	154	22	for	for	ADP
ajst-7347	154	23	elevated	elevated	ADJ
ajst-7347	154	24	pm10	pm10	PROPN
ajst-7347	154	25	and	and	CCONJ
ajst-7347	154	26	pm2	pm2	PROPN
ajst-7347	154	27	.	.	PUNCT
ajst-7347	155	1	5	5	NUM
ajst-7347	155	2	concentrations	concentration	NOUN
ajst-7347	155	3	in	in	ADP
ajst-7347	155	4	europe	europe	PROPN
ajst-7347	155	5	with	with	ADP
ajst-7347	155	6	mm5	mm5	PROPN
ajst-7347	155	7	-	-	PUNCT
ajst-7347	155	8	cmaq	cmaq	PROPN
ajst-7347	155	9	and	and	CCONJ
ajst-7347	155	10	wrf	wrf	PROPN
ajst-7347	155	11	/	/	SYM
ajst-7347	155	12	chem[j	chem[j	PROPN
ajst-7347	155	13	]	]	PUNCT
ajst-7347	155	14	.	.	PUNCT
ajst-7347	155	15	air	air	NOUN
ajst-7347	155	16	pollution	pollution	NOUN
ajst-7347	155	17	,	,	PUNCT
ajst-7347	155	18	2009	2009	NUM
ajst-7347	155	19	:	:	PUNCT
ajst-7347	155	20	377	377	NUM
ajst-7347	155	21	-	-	SYM
ajst-7347	155	22	86	86	NUM
ajst-7347	155	23	.	.	PUNCT
ajst-7347	156	1	[	[	X
ajst-7347	156	2	7	7	X
ajst-7347	156	3	]	]	X
ajst-7347	156	4	lee	lee	PROPN
ajst-7347	156	5	d	d	PROPN
ajst-7347	156	6	g.	g.	PROPN
ajst-7347	156	7	comparison	comparison	NOUN
ajst-7347	156	8	between	between	ADP
ajst-7347	156	9	atmospheric	atmospheric	ADJ
ajst-7347	156	10	chemistry	chemistry	NOUN
ajst-7347	156	11	model	model	NOUN
ajst-7347	156	12	and	and	CCONJ
ajst-7347	156	13	observations	observation	NOUN
ajst-7347	156	14	utilizing	utilize	VERB
ajst-7347	156	15	the	the	DET
ajst-7347	156	16	raqms	raqms	NOUN
ajst-7347	156	17	-	-	PUNCT
ajst-7347	156	18	cmaq	cmaq	NOUN
ajst-7347	156	19	linkage	linkage	NOUN
ajst-7347	156	20	,	,	PUNCT
ajst-7347	156	21	part	part	NOUN
ajst-7347	156	22	ii	ii	NOUN
ajst-7347	156	23	:	:	PUNCT
ajst-7347	156	24	impact	impact	NOUN
ajst-7347	156	25	on	on	ADP
ajst-7347	156	26	pm2	pm2	NOUN
ajst-7347	156	27	.	.	PUNCT
ajst-7347	157	1	5	5	NUM
ajst-7347	157	2	mass	mass	NOUN
ajst-7347	157	3	concentrations	concentration	NOUN
ajst-7347	157	4	simulated[j	simulated[j	VERB
ajst-7347	157	5	]	]	PUNCT
ajst-7347	157	6	.	.	PUNCT
ajst-7347	158	1	asian	asian	ADJ
ajst-7347	158	2	journal	journal	PROPN
ajst-7347	158	3	of	of	ADP
ajst-7347	158	4	atmospheric	atmospheric	PROPN
ajst-7347	158	5	environment	environment	NOUN
ajst-7347	158	6	,	,	PUNCT
ajst-7347	158	7	2014	2014	NUM
ajst-7347	158	8	,	,	PUNCT
ajst-7347	158	9	8(2	8(2	NUM
ajst-7347	158	10	):	):	PUNCT
ajst-7347	158	11	108	108	NUM
ajst-7347	158	12	-	-	SYM
ajst-7347	158	13	114	114	NUM
ajst-7347	158	14	.	.	PUNCT
ajst-7347	159	1	[	[	X
ajst-7347	159	2	8	8	NUM
ajst-7347	159	3	]	]	X
ajst-7347	159	4	jiang	jiang	PROPN
ajst-7347	159	5	x	x	PROPN
ajst-7347	159	6	,	,	PUNCT
ajst-7347	159	7	yoo	yoo	PROPN
ajst-7347	159	8	e.	e.	PROPN
ajst-7347	159	9	the	the	DET
ajst-7347	159	10	importance	importance	NOUN
ajst-7347	159	11	of	of	ADP
ajst-7347	159	12	spatial	spatial	ADJ
ajst-7347	159	13	resolutions	resolution	NOUN
ajst-7347	159	14	of	of	ADP
ajst-7347	159	15	community	community	NOUN
ajst-7347	159	16	multiscale	multiscale	ADJ
ajst-7347	159	17	air	air	NOUN
ajst-7347	159	18	quality	quality	NOUN
ajst-7347	159	19	(	(	PUNCT
ajst-7347	159	20	cmaq	cmaq	NOUN
ajst-7347	159	21	)	)	PUNCT
ajst-7347	159	22	models	model	NOUN
ajst-7347	159	23	on	on	ADP
ajst-7347	159	24	health	health	NOUN
ajst-7347	159	25	impact	impact	NOUN
ajst-7347	159	26	assessment[j	assessment[j	ADV
ajst-7347	159	27	]	]	PUNCT
ajst-7347	159	28	.	.	PUNCT
ajst-7347	160	1	science	science	NOUN
ajst-7347	160	2	of	of	ADP
ajst-7347	160	3	the	the	DET
ajst-7347	160	4	total	total	ADJ
ajst-7347	160	5	environment	environment	NOUN
ajst-7347	160	6	,	,	PUNCT
ajst-7347	160	7	2018	2018	NUM
ajst-7347	160	8	,	,	PUNCT
ajst-7347	160	9	627	627	NUM
ajst-7347	160	10	:	:	PUNCT
ajst-7347	160	11	1528	1528	NUM
ajst-7347	160	12	-	-	SYM
ajst-7347	160	13	1543	1543	NUM
ajst-7347	160	14	.	.	PUNCT
ajst-7347	161	1	[	[	X
ajst-7347	161	2	9	9	NUM
ajst-7347	161	3	]	]	X
ajst-7347	161	4	wang	wang	PROPN
ajst-7347	161	5	w	w	PROPN
ajst-7347	161	6	,	,	PUNCT
ajst-7347	161	7	wu	wu	PROPN
ajst-7347	161	8	t	t	PROPN
ajst-7347	161	9	,	,	PUNCT
ajst-7347	161	10	zhang	zhang	PROPN
ajst-7347	161	11	z.	z.	PROPN
ajst-7347	161	12	pm2	pm2	PROPN
ajst-7347	161	13	.	.	PUNCT
ajst-7347	162	1	5	5	NUM
ajst-7347	162	2	prediction	prediction	NOUN
ajst-7347	162	3	of	of	ADP
ajst-7347	162	4	hangzhou	hangzhou	PROPN
ajst-7347	162	5	based	base	VERB
ajst-7347	162	6	on	on	ADP
ajst-7347	162	7	arima	arima	PROPN
ajst-7347	162	8	model	model	NOUN
ajst-7347	163	1	[	[	X
ajst-7347	163	2	j	j	X
ajst-7347	163	3	]	]	X
ajst-7347	163	4	.	.	PUNCT
ajst-7347	164	1	journal	journal	PROPN
ajst-7347	164	2	of	of	ADP
ajst-7347	164	3	natural	natural	ADJ
ajst-7347	164	4	sciences	science	NOUN
ajst-7347	164	5	,	,	PUNCT
ajst-7347	164	6	harbin	harbin	PROPN
ajst-7347	164	7	normal	normal	ADJ
ajst-7347	164	8	university	university	NOUN
ajst-7347	164	9	,	,	PUNCT
ajst-7347	164	10	2018	2018	NUM
ajst-7347	164	11	,	,	PUNCT
ajst-7347	164	12	34(3	34(3	NUM
ajst-7347	164	13	):	):	PUNCT
ajst-7347	164	14	49	49	NUM
ajst-7347	164	15	-	-	SYM
ajst-7347	164	16	55	55	NUM
ajst-7347	164	17	.	.	PUNCT
ajst-7347	165	1	[	[	X
ajst-7347	165	2	10	10	NUM
ajst-7347	165	3	]	]	X
ajst-7347	165	4	wang	wang	PROPN
ajst-7347	165	5	j.	j.	PROPN
ajst-7347	165	6	prediction	prediction	PROPN
ajst-7347	165	7	of	of	ADP
ajst-7347	165	8	pm2	pm2	NOUN
ajst-7347	165	9	.	.	PUNCT
ajst-7347	166	1	5	5	NUM
ajst-7347	166	2	based	base	VERB
ajst-7347	166	3	on	on	ADP
ajst-7347	166	4	multiple	multiple	ADJ
ajst-7347	166	5	regression	regression	NOUN
ajst-7347	166	6	analysis	analysis	NOUN
ajst-7347	166	7	[	[	X
ajst-7347	166	8	j	j	X
ajst-7347	166	9	]	]	X
ajst-7347	166	10	.	.	PUNCT
ajst-7347	167	1	microcomputer	microcomputer	NOUN
ajst-7347	167	2	application	application	NOUN
ajst-7347	167	3	,	,	PUNCT
ajst-7347	167	4	2020	2020	NUM
ajst-7347	167	5	,	,	PUNCT
ajst-7347	167	6	3	3	X
ajst-7347	167	7	.	.	PUNCT
ajst-7347	168	1	[	[	X
ajst-7347	168	2	11	11	NUM
ajst-7347	168	3	]	]	PUNCT
ajst-7347	168	4	wang	wang	PROPN
ajst-7347	168	5	j	j	PROPN
ajst-7347	168	6	,	,	PUNCT
ajst-7347	168	7	cao	cao	PROPN
ajst-7347	168	8	c.	c.	PROPN
ajst-7347	168	9	pm_(2.5	pm_(2.5	PROPN
ajst-7347	168	10	)	)	PUNCT
ajst-7347	168	11	prediction	prediction	NOUN
ajst-7347	168	12	of	of	ADP
ajst-7347	168	13	beijing	beijing	PROPN
ajst-7347	168	14	based	base	VERB
ajst-7347	168	15	on	on	ADP
ajst-7347	168	16	bayesian	bayesian	NOUN
ajst-7347	168	17	hierarchical	hierarchical	ADJ
ajst-7347	168	18	autoregressive	autoregressive	ADJ
ajst-7347	168	19	spatio	spatio	NOUN
ajst-7347	168	20	-	-	PUNCT
ajst-7347	168	21	temporal	temporal	ADJ
ajst-7347	168	22	model	model	NOUN
ajst-7347	168	23	[	[	X
ajst-7347	168	24	j	j	X
ajst-7347	168	25	]	]	X
ajst-7347	168	26	.	.	PUNCT
ajst-7347	169	1	journal	journal	PROPN
ajst-7347	169	2	of	of	ADP
ajst-7347	169	3	nanjing	nanjing	PROPN
ajst-7347	169	4	university	university	PROPN
ajst-7347	169	5	of	of	ADP
ajst-7347	169	6	information	information	NOUN
ajst-7347	169	7	science	science	PROPN
ajst-7347	169	8	&	&	CCONJ
ajst-7347	169	9	technology	technology	PROPN
ajst-7347	169	10	(	(	PUNCT
ajst-7347	169	11	natural	natural	ADJ
ajst-7347	169	12	science),2023,15(01):34	science),2023,15(01):34	NOUN
ajst-7347	169	13	-	-	PUNCT
ajst-7347	169	14	41	41	NUM
ajst-7347	169	15	.	.	PUNCT
ajst-7347	170	1	[	[	X
ajst-7347	170	2	12	12	NUM
ajst-7347	170	3	]	]	X
ajst-7347	170	4	masood	masood	PROPN
ajst-7347	170	5	a	a	PROPN
ajst-7347	170	6	,	,	PUNCT
ajst-7347	170	7	ahmad	ahmad	PROPN
ajst-7347	170	8	k.	k.	PROPN
ajst-7347	170	9	a	a	DET
ajst-7347	170	10	model	model	NOUN
ajst-7347	170	11	for	for	ADP
ajst-7347	170	12	particulate	particulate	NOUN
ajst-7347	170	13	matter	matter	NOUN
ajst-7347	170	14	(	(	PUNCT
ajst-7347	170	15	pm2	pm2	NOUN
ajst-7347	170	16	.	.	NOUN
ajst-7347	170	17	5	5	NUM
ajst-7347	170	18	)	)	PUNCT
ajst-7347	170	19	prediction	prediction	NOUN
ajst-7347	170	20	for	for	ADP
ajst-7347	170	21	delhi	delhi	PROPN
ajst-7347	170	22	based	base	VERB
ajst-7347	170	23	on	on	ADP
ajst-7347	170	24	machine	machine	NOUN
ajst-7347	170	25	learning	learn	VERB
ajst-7347	170	26	approaches[j	approaches[j	PROPN
ajst-7347	170	27	]	]	PUNCT
ajst-7347	170	28	.	.	PUNCT
ajst-7347	171	1	procedia	procedia	PROPN
ajst-7347	171	2	computer	computer	NOUN
ajst-7347	171	3	science	science	NOUN
ajst-7347	171	4	,	,	PUNCT
ajst-7347	171	5	2020	2020	NUM
ajst-7347	171	6	,	,	PUNCT
ajst-7347	171	7	167	167	NUM
ajst-7347	171	8	:	:	SYM
ajst-7347	171	9	2101	2101	NUM
ajst-7347	171	10	-	-	SYM
ajst-7347	171	11	2110	2110	NUM
ajst-7347	171	12	.	.	PUNCT
ajst-7347	172	1	[	[	X
ajst-7347	172	2	13	13	NUM
ajst-7347	172	3	]	]	X
ajst-7347	172	4	chen	chen	PROPN
ajst-7347	172	5	g	g	PROPN
ajst-7347	172	6	,	,	PUNCT
ajst-7347	172	7	li	li	PROPN
ajst-7347	172	8	s	s	PROPN
ajst-7347	172	9	,	,	PUNCT
ajst-7347	172	10	knibbs	knibbs	X
ajst-7347	172	11	l	l	NOUN
ajst-7347	172	12	d	d	PROPN
ajst-7347	172	13	,	,	PUNCT
ajst-7347	172	14	et	et	PROPN
ajst-7347	172	15	al	al	PROPN
ajst-7347	172	16	.	.	PUNCT
ajst-7347	173	1	a	a	DET
ajst-7347	173	2	machine	machine	NOUN
ajst-7347	173	3	learning	learn	VERB
ajst-7347	173	4	method	method	NOUN
ajst-7347	173	5	to	to	PART
ajst-7347	173	6	estimate	estimate	VERB
ajst-7347	173	7	pm2	pm2	NOUN
ajst-7347	173	8	.	.	PUNCT
ajst-7347	174	1	5	5	NUM
ajst-7347	174	2	concentrations	concentration	NOUN
ajst-7347	174	3	across	across	ADP
ajst-7347	174	4	china	china	PROPN
ajst-7347	174	5	with	with	ADP
ajst-7347	174	6	remote	remote	ADJ
ajst-7347	174	7	sensing	sensing	NOUN
ajst-7347	174	8	,	,	PUNCT
ajst-7347	174	9	meteorological	meteorological	ADJ
ajst-7347	174	10	and	and	CCONJ
ajst-7347	174	11	land	land	NOUN
ajst-7347	174	12	use	use	NOUN
ajst-7347	174	13	information[j	information[j	PROPN
ajst-7347	174	14	]	]	PUNCT
ajst-7347	174	15	.	.	PUNCT
ajst-7347	175	1	science	science	NOUN
ajst-7347	175	2	of	of	ADP
ajst-7347	175	3	the	the	DET
ajst-7347	175	4	total	total	ADJ
ajst-7347	175	5	environment	environment	NOUN
ajst-7347	175	6	,	,	PUNCT
ajst-7347	175	7	2018	2018	NUM
ajst-7347	175	8	,	,	PUNCT
ajst-7347	175	9	636	636	NUM
ajst-7347	175	10	:	:	PUNCT
ajst-7347	175	11	52	52	NUM
ajst-7347	175	12	-	-	SYM
ajst-7347	175	13	60	60	NUM
ajst-7347	175	14	.	.	PUNCT
ajst-7347	176	1	[	[	X
ajst-7347	176	2	14	14	NUM
ajst-7347	176	3	]	]	X
ajst-7347	176	4	zamani	zamani	PROPN
ajst-7347	176	5	joharestani	joharestani	PROPN
ajst-7347	176	6	m	m	PROPN
ajst-7347	176	7	,	,	PUNCT
ajst-7347	176	8	cao	cao	PROPN
ajst-7347	176	9	c	c	PROPN
ajst-7347	176	10	,	,	PUNCT
ajst-7347	176	11	ni	ni	PROPN
ajst-7347	176	12	x	x	PROPN
ajst-7347	176	13	,	,	PUNCT
ajst-7347	176	14	et	et	PROPN
ajst-7347	176	15	al	al	PROPN
ajst-7347	176	16	.	.	PUNCT
ajst-7347	177	1	pm2	pm2	PROPN
ajst-7347	177	2	.	.	PUNCT
ajst-7347	178	1	5	5	NUM
ajst-7347	178	2	prediction	prediction	NOUN
ajst-7347	178	3	based	base	VERB
ajst-7347	178	4	on	on	ADP
ajst-7347	178	5	random	random	ADJ
ajst-7347	178	6	forest	forest	NOUN
ajst-7347	178	7	,	,	PUNCT
ajst-7347	178	8	xgboost	xgboost	ADV
ajst-7347	178	9	,	,	PUNCT
ajst-7347	178	10	and	and	CCONJ
ajst-7347	178	11	deep	deep	ADJ
ajst-7347	178	12	learning	learning	NOUN
ajst-7347	178	13	using	use	VERB
ajst-7347	178	14	multisource	multisource	NOUN
ajst-7347	178	15	remote	remote	ADJ
ajst-7347	178	16	sensing	sense	VERB
ajst-7347	178	17	data[j	data[j	NOUN
ajst-7347	178	18	]	]	PUNCT
ajst-7347	178	19	.	.	PUNCT
ajst-7347	179	1	atmosphere	atmosphere	NOUN
ajst-7347	179	2	,	,	PUNCT
ajst-7347	179	3	2019	2019	NUM
ajst-7347	179	4	,	,	PUNCT
ajst-7347	179	5	10(7	10(7	NUM
ajst-7347	179	6	):	):	PUNCT
ajst-7347	179	7	373	373	NUM
ajst-7347	179	8	.	.	PUNCT
ajst-7347	180	1	[	[	X
ajst-7347	180	2	15	15	NUM
ajst-7347	180	3	]	]	X
ajst-7347	180	4	liang	liang	PROPN
ajst-7347	180	5	y	y	PROPN
ajst-7347	180	6	,	,	PUNCT
ajst-7347	180	7	li	li	PROPN
ajst-7347	180	8	z	z	PROPN
ajst-7347	180	9	,	,	PUNCT
ajst-7347	180	10	jin	jin	PROPN
ajst-7347	180	11	y	y	PROPN
ajst-7347	180	12	,	,	PUNCT
ajst-7347	180	13	et	et	PROPN
ajst-7347	180	14	al	al	PROPN
ajst-7347	180	15	.	.	PUNCT
ajst-7347	181	1	comparison	comparison	NOUN
ajst-7347	181	2	of	of	ADP
ajst-7347	181	3	pm_(2.5	pm_(2.5	NUM
ajst-7347	181	4	)	)	PUNCT
ajst-7347	181	5	prediction	prediction	NOUN
ajst-7347	181	6	effect	effect	NOUN
ajst-7347	181	7	in	in	ADP
ajst-7347	181	8	beijing	beijing	PROPN
ajst-7347	181	9	based	base	VERB
ajst-7347	181	10	on	on	ADP
ajst-7347	181	11	tree	tree	NOUN
ajst-7347	181	12	model	model	NOUN
ajst-7347	182	1	[	[	X
ajst-7347	182	2	j	j	X
ajst-7347	182	3	]	]	X
ajst-7347	182	4	.	.	PUNCT
ajst-7347	183	1	environmental	environmental	PROPN
ajst-7347	183	2	engineering:1	engineering:1	PROPN
ajst-7347	183	3	-	-	PROPN
ajst-7347	183	4	13[2023	13[2023	NUM
ajst-7347	183	5	-	-	PUNCT
ajst-7347	183	6	03	03	NUM
ajst-7347	183	7	-	-	SYM
ajst-7347	183	8	13	13	NUM
ajst-7347	183	9	]	]	PUNCT
ajst-7347	183	10	.	.	PUNCT
ajst-7347	184	1	[	[	X
ajst-7347	184	2	16	16	NUM
ajst-7347	184	3	]	]	X
ajst-7347	184	4	kong	kong	PROPN
ajst-7347	184	5	y	y	PROPN
ajst-7347	184	6	,	,	PUNCT
ajst-7347	184	7	wang	wang	PROPN
ajst-7347	184	8	h	h	PROPN
ajst-7347	184	9	,	,	PUNCT
ajst-7347	184	10	zhang	zhang	PROPN
ajst-7347	184	11	h	h	PROPN
ajst-7347	184	12	,	,	PUNCT
ajst-7347	184	13	et	et	PROPN
ajst-7347	184	14	al	al	PROPN
ajst-7347	184	15	.	.	PUNCT
ajst-7347	185	1	pm2.5	pm2.5	PART
ajst-7347	185	2	concentration	concentration	NOUN
ajst-7347	185	3	prediction	prediction	NOUN
ajst-7347	185	4	based	base	VERB
ajst-7347	185	5	on	on	ADP
ajst-7347	185	6	integrated	integrated	ADJ
ajst-7347	185	7	learning	learning	NOUN
ajst-7347	185	8	algorithm	algorithm	NOUN
ajst-7347	186	1	[	[	X
ajst-7347	186	2	j	j	X
ajst-7347	186	3	]	]	X
ajst-7347	186	4	.	.	PUNCT
ajst-7347	187	1	environmental	environmental	ADJ
ajst-7347	187	2	protection	protection	PROPN
ajst-7347	187	3	science,2021,47(4):17	science,2021,47(4):17	NOUN
ajst-7347	187	4	-	-	NOUN
ajst-7347	187	5	23.i	23.i	NUM
ajst-7347	187	6	[	[	SYM
ajst-7347	187	7	17	17	NUM
ajst-7347	187	8	]	]	PUNCT
ajst-7347	188	1	ding	de	VERB
ajst-7347	188	2	c	c	NOUN
ajst-7347	188	3	,	,	PUNCT
ajst-7347	188	4	wang	wang	PROPN
ajst-7347	188	5	g	g	PROPN
ajst-7347	188	6	,	,	PUNCT
ajst-7347	188	7	zhang	zhang	PROPN
ajst-7347	188	8	x	x	PROPN
ajst-7347	188	9	,	,	PUNCT
ajst-7347	188	10	et	et	PROPN
ajst-7347	188	11	al	al	PROPN
ajst-7347	188	12	.	.	PUNCT
ajst-7347	189	1	a	a	DET
ajst-7347	189	2	hybrid	hybrid	ADJ
ajst-7347	189	3	cnn	cnn	PROPN
ajst-7347	189	4	-	-	PUNCT
ajst-7347	189	5	lstm	lstm	PROPN
ajst-7347	189	6	model	model	NOUN
ajst-7347	189	7	for	for	ADP
ajst-7347	189	8	predicting	predict	VERB
ajst-7347	189	9	pm2	pm2	NOUN
ajst-7347	189	10	.	.	PROPN
ajst-7347	189	11	5	5	NUM
ajst-7347	189	12	in	in	ADP
ajst-7347	189	13	beijing	beijing	PROPN
ajst-7347	189	14	based	base	VERB
ajst-7347	189	15	on	on	ADP
ajst-7347	189	16	spatiotemporal	spatiotemporal	ADJ
ajst-7347	189	17	correlation[j	correlation[j	PROPN
ajst-7347	189	18	]	]	PUNCT
ajst-7347	189	19	.	.	PUNCT
ajst-7347	190	1	environmental	environmental	ADJ
ajst-7347	190	2	and	and	CCONJ
ajst-7347	190	3	ecological	ecological	ADJ
ajst-7347	190	4	statistics	statistic	NOUN
ajst-7347	190	5	,	,	PUNCT
ajst-7347	190	6	2021	2021	NUM
ajst-7347	190	7	,	,	PUNCT
ajst-7347	190	8	28(3	28(3	NUM
ajst-7347	190	9	):	):	PUNCT
ajst-7347	190	10	503	503	NUM
ajst-7347	190	11	-	-	SYM
ajst-7347	190	12	522	522	NUM
ajst-7347	190	13	.	.	PUNCT
ajst-7347	191	1	[	[	X
ajst-7347	191	2	18	18	NUM
ajst-7347	191	3	]	]	X
ajst-7347	191	4	di	di	X
ajst-7347	191	5	q	q	PROPN
ajst-7347	191	6	,	,	PUNCT
ajst-7347	191	7	amini	amini	PROPN
ajst-7347	191	8	h	h	PROPN
ajst-7347	191	9	,	,	PUNCT
ajst-7347	191	10	shi	shi	PROPN
ajst-7347	191	11	l	l	PROPN
ajst-7347	191	12	,	,	PUNCT
ajst-7347	191	13	et	et	PROPN
ajst-7347	191	14	al	al	PROPN
ajst-7347	191	15	.	.	PUNCT
ajst-7347	192	1	an	an	DET
ajst-7347	192	2	ensemble	ensemble	ADJ
ajst-7347	192	3	-	-	PUNCT
ajst-7347	192	4	based	base	VERB
ajst-7347	192	5	model	model	NOUN
ajst-7347	192	6	of	of	ADP
ajst-7347	192	7	pm2	pm2	NOUN
ajst-7347	192	8	.	.	PUNCT
ajst-7347	193	1	5	5	NUM
ajst-7347	193	2	concentration	concentration	NOUN
ajst-7347	193	3	across	across	ADP
ajst-7347	193	4	the	the	DET
ajst-7347	193	5	contiguous	contiguous	ADJ
ajst-7347	193	6	united	united	ADJ
ajst-7347	193	7	states	states	PROPN
ajst-7347	193	8	with	with	ADP
ajst-7347	193	9	high	high	ADJ
ajst-7347	193	10	spatiotemporal	spatiotemporal	ADJ
ajst-7347	193	11	resolution[j	resolution[j	PROPN
ajst-7347	193	12	]	]	PUNCT
ajst-7347	193	13	.	.	PUNCT
ajst-7347	194	1	environment	environment	PROPN
ajst-7347	194	2	international	international	PROPN
ajst-7347	194	3	,	,	PUNCT
ajst-7347	194	4	2019	2019	NUM
ajst-7347	194	5	,	,	PUNCT
ajst-7347	194	6	130	130	NUM
ajst-7347	194	7	:	:	SYM
ajst-7347	194	8	104909	104909	NUM
ajst-7347	194	9	.	.	PUNCT
ajst-7347	195	1	[	[	X
ajst-7347	195	2	19	19	NUM
ajst-7347	195	3	]	]	X
ajst-7347	195	4	niu	niu	PROPN
ajst-7347	195	5	m	m	PROPN
ajst-7347	195	6	,	,	PUNCT
ajst-7347	195	7	wang	wang	PROPN
ajst-7347	195	8	y	y	PROPN
ajst-7347	195	9	,	,	PUNCT
ajst-7347	195	10	sun	sun	PROPN
ajst-7347	195	11	s	s	PROPN
ajst-7347	195	12	,	,	PUNCT
ajst-7347	195	13	et	et	PROPN
ajst-7347	195	14	al	al	PROPN
ajst-7347	195	15	.	.	PUNCT
ajst-7347	196	1	a	a	DET
ajst-7347	196	2	novel	novel	ADJ
ajst-7347	196	3	hybrid	hybrid	ADJ
ajst-7347	196	4	decompositionand	decompositionand	NOUN
ajst-7347	196	5	-	-	PUNCT
ajst-7347	196	6	ensemble	ensemble	ADJ
ajst-7347	196	7	model	model	NOUN
ajst-7347	196	8	based	base	VERB
ajst-7347	196	9	on	on	ADP
ajst-7347	196	10	ceemd	ceemd	PROPN
ajst-7347	196	11	and	and	CCONJ
ajst-7347	196	12	gwo	gwo	VERB
ajst-7347	196	13	for	for	ADP
ajst-7347	196	14	shortterm	shortterm	PROPN
ajst-7347	196	15	pm2	pm2	PROPN
ajst-7347	196	16	.	.	PUNCT
ajst-7347	197	1	5	5	NUM
ajst-7347	197	2	concentration	concentration	NOUN
ajst-7347	197	3	forecasting[j	forecasting[j	NOUN
ajst-7347	197	4	]	]	PUNCT
ajst-7347	197	5	.	.	PUNCT
ajst-7347	198	1	atmospheric	atmospheric	ADJ
ajst-7347	198	2	environment	environment	NOUN
ajst-7347	198	3	,	,	PUNCT
ajst-7347	198	4	2016	2016	NUM
ajst-7347	198	5	,	,	PUNCT
ajst-7347	198	6	134	134	NUM
ajst-7347	198	7	:	:	SYM
ajst-7347	198	8	168	168	NUM
ajst-7347	198	9	-	-	SYM
ajst-7347	198	10	180	180	NUM
ajst-7347	198	11	.	.	PUNCT
ajst-7347	199	1	[	[	X
ajst-7347	199	2	20	20	NUM
ajst-7347	199	3	]	]	X
ajst-7347	199	4	zhou	zhou	PROPN
ajst-7347	199	5	f	f	PROPN
ajst-7347	199	6	,	,	PUNCT
ajst-7347	199	7	jin	jin	PROPN
ajst-7347	199	8	l	l	PROPN
ajst-7347	199	9	,	,	PUNCT
ajst-7347	199	10	dong	dong	PROPN
ajst-7347	199	11	j.	j.	PROPN
ajst-7347	199	12	review	review	PROPN
ajst-7347	199	13	of	of	ADP
ajst-7347	199	14	convolutional	convolutional	ADJ
ajst-7347	199	15	neural	neural	ADJ
ajst-7347	199	16	network	network	NOUN
ajst-7347	199	17	[	[	X
ajst-7347	199	18	j	j	X
ajst-7347	199	19	]	]	X
ajst-7347	199	20	.	.	PUNCT
ajst-7347	200	1	journal	journal	PROPN
ajst-7347	200	2	of	of	ADP
ajst-7347	200	3	computer	computer	NOUN
ajst-7347	200	4	science,2017,40(6):12291251	science,2017,40(6):12291251	PROPN
ajst-7347	200	5	.	.	PUNCT
ajst-7347	201	1	[	[	X
ajst-7347	201	2	21	21	NUM
ajst-7347	201	3	]	]	X
ajst-7347	201	4	liang	liang	PROPN
ajst-7347	201	5	r	r	PROPN
ajst-7347	201	6	,	,	PUNCT
ajst-7347	201	7	chang	chang	PROPN
ajst-7347	201	8	x	x	PROPN
ajst-7347	201	9	,	,	PUNCT
ajst-7347	201	10	jia	jia	PROPN
ajst-7347	201	11	p	p	PROPN
ajst-7347	201	12	,	,	PUNCT
ajst-7347	201	13	et	et	PROPN
ajst-7347	201	14	al	al	PROPN
ajst-7347	201	15	.	.	PROPN
ajst-7347	201	16	mine	mine	PROPN
ajst-7347	201	17	gas	gas	NOUN
ajst-7347	201	18	concentration	concentration	NOUN
ajst-7347	201	19	forecasting	forecasting	NOUN
ajst-7347	201	20	model	model	NOUN
ajst-7347	201	21	based	base	VERB
ajst-7347	201	22	on	on	ADP
ajst-7347	201	23	an	an	DET
ajst-7347	201	24	optimized	optimize	VERB
ajst-7347	201	25	bigru	bigru	NOUN
ajst-7347	201	26	network[j	network[j	PROPN
ajst-7347	201	27	]	]	PUNCT
ajst-7347	201	28	.	.	PUNCT
ajst-7347	202	1	acs	acs	PROPN
ajst-7347	202	2	omega	omega	NOUN
ajst-7347	202	3	,	,	PUNCT
ajst-7347	202	4	2020	2020	NUM
ajst-7347	202	5	,	,	PUNCT
ajst-7347	202	6	5(44	5(44	NUM
ajst-7347	202	7	):	):	PUNCT
ajst-7347	202	8	28579	28579	NUM
ajst-7347	202	9	-	-	SYM
ajst-7347	202	10	28586	28586	NUM
ajst-7347	202	11	.	.	PUNCT
ajst-7347	203	1	8	8	NUM
ajst-7347	204	1	[	[	SYM
ajst-7347	204	2	22	22	NUM
ajst-7347	204	3	]	]	X
ajst-7347	204	4	zhang	zhang	PROPN
ajst-7347	204	5	b	b	PROPN
ajst-7347	204	6	,	,	PUNCT
ajst-7347	204	7	jia	jia	PROPN
ajst-7347	204	8	m	m	PROPN
ajst-7347	204	9	,	,	PUNCT
ajst-7347	204	10	xu	xu	PROPN
ajst-7347	204	11	j	j	PROPN
ajst-7347	204	12	,	,	PUNCT
ajst-7347	204	13	et	et	PROPN
ajst-7347	204	14	al	al	PROPN
ajst-7347	204	15	.	.	PROPN
ajst-7347	204	16	network	network	PROPN
ajst-7347	204	17	security	security	NOUN
ajst-7347	204	18	situation	situation	NOUN
ajst-7347	204	19	prediction	prediction	NOUN
ajst-7347	204	20	model	model	NOUN
ajst-7347	204	21	based	base	VERB
ajst-7347	204	22	on	on	ADP
ajst-7347	204	23	emd	emd	PROPN
ajst-7347	204	24	and	and	CCONJ
ajst-7347	204	25	elpso	elpso	NOUN
ajst-7347	204	26	optimized	optimize	VERB
ajst-7347	204	27	bigru	bigru	NOUN
ajst-7347	204	28	neural	neural	ADJ
ajst-7347	204	29	network[j	network[j	PROPN
ajst-7347	204	30	]	]	PUNCT
ajst-7347	204	31	.	.	PUNCT
ajst-7347	205	1	computational	computational	ADJ
ajst-7347	205	2	intelligence	intelligence	NOUN
ajst-7347	205	3	and	and	CCONJ
ajst-7347	205	4	neuroscience	neuroscience	NOUN
ajst-7347	205	5	,	,	PUNCT
ajst-7347	205	6	2022	2022	NUM
ajst-7347	205	7	,	,	PUNCT
ajst-7347	205	8	2022	2022	NUM
ajst-7347	205	9	.	.	PUNCT
ajst-7347	206	1	[	[	X
ajst-7347	206	2	23	23	NUM
ajst-7347	206	3	]	]	X
ajst-7347	206	4	zhang	zhang	PROPN
ajst-7347	206	5	h	h	PROPN
ajst-7347	206	6	,	,	PUNCT
ajst-7347	206	7	wang	wang	PROPN
ajst-7347	206	8	y	y	PROPN
ajst-7347	206	9	,	,	PUNCT
ajst-7347	206	10	hu	hu	PROPN
ajst-7347	206	11	j	j	PROPN
ajst-7347	206	12	,	,	PUNCT
ajst-7347	206	13	et	et	PROPN
ajst-7347	206	14	al	al	PROPN
ajst-7347	206	15	.	.	PUNCT
ajst-7347	206	16	relationships	relationship	NOUN
ajst-7347	206	17	between	between	ADP
ajst-7347	206	18	meteorological	meteorological	ADJ
ajst-7347	206	19	parameters	parameter	NOUN
ajst-7347	206	20	and	and	CCONJ
ajst-7347	206	21	criteria	criterion	NOUN
ajst-7347	206	22	air	air	NOUN
ajst-7347	206	23	pollutants	pollutant	NOUN
ajst-7347	206	24	in	in	ADP
ajst-7347	206	25	three	three	NUM
ajst-7347	206	26	megacities	megacitie	NOUN
ajst-7347	206	27	in	in	ADP
ajst-7347	206	28	china[j	china[j	NOUN
ajst-7347	206	29	]	]	PUNCT
ajst-7347	206	30	.	.	PUNCT
ajst-7347	207	1	environmental	environmental	ADJ
ajst-7347	207	2	research	research	NOUN
ajst-7347	207	3	,	,	PUNCT
ajst-7347	207	4	2015	2015	NUM
ajst-7347	207	5	,	,	PUNCT
ajst-7347	207	6	140	140	NUM
ajst-7347	207	7	:	:	PUNCT
ajst-7347	207	8	242	242	NUM
ajst-7347	207	9	-	-	SYM
ajst-7347	207	10	254	254	NUM
ajst-7347	207	11	.	.	PUNCT
ajst-7347	208	1	[	[	X
ajst-7347	208	2	24	24	NUM
ajst-7347	208	3	]	]	X
ajst-7347	208	4	rojas	rojas	PROPN
ajst-7347	208	5	n	n	AUX
ajst-7347	208	6	,	,	PUNCT
ajst-7347	208	7	galvis	galvis	PROPN
ajst-7347	208	8	b.	b.	PROPN
ajst-7347	208	9	relationship	relationship	NOUN
ajst-7347	208	10	between	between	ADP
ajst-7347	208	11	pm2	pm2	NOUN
ajst-7347	208	12	.	.	PROPN
ajst-7347	208	13	5	5	NUM
ajst-7347	208	14	and	and	CCONJ
ajst-7347	208	15	pm10	pm10	PROPN
ajst-7347	208	16	in	in	ADP
ajst-7347	208	17	bogotá[j	bogotá[j	PROPN
ajst-7347	208	18	]	]	PUNCT
ajst-7347	208	19	.	.	PUNCT
ajst-7347	209	1	revista	revista	PROPN
ajst-7347	209	2	de	de	PROPN
ajst-7347	209	3	ingeniería	ingeniería	PROPN
ajst-7347	209	4	,	,	PUNCT
ajst-7347	209	5	2005	2005	NUM
ajst-7347	209	6	(	(	PUNCT
ajst-7347	209	7	22	22	NUM
ajst-7347	209	8	):	):	PUNCT
ajst-7347	209	9	54	54	NUM
ajst-7347	209	10	-	-	SYM
ajst-7347	209	11	60	60	NUM
ajst-7347	209	12	.	.	PUNCT
ajst-7347	210	1	[	[	X
ajst-7347	210	2	25	25	NUM
ajst-7347	210	3	]	]	X
ajst-7347	210	4	liu	liu	PROPN
ajst-7347	210	5	d	d	PROPN
ajst-7347	210	6	,	,	PUNCT
ajst-7347	210	7	cho	cho	PROPN
ajst-7347	210	8	s	s	PROPN
ajst-7347	210	9	y	y	PROPN
ajst-7347	210	10	,	,	PUNCT
ajst-7347	210	11	sun	sun	PROPN
ajst-7347	210	12	d	d	PROPN
ajst-7347	210	13	,	,	PUNCT
ajst-7347	210	14	et	et	PROPN
ajst-7347	210	15	al	al	PROPN
ajst-7347	210	16	.	.	PUNCT
ajst-7347	211	1	a	a	DET
ajst-7347	211	2	spearman	spearman	NOUN
ajst-7347	211	3	correlation	correlation	NOUN
ajst-7347	211	4	coefficient	coefficient	NOUN
ajst-7347	211	5	ranking	rank	VERB
ajst-7347	211	6	for	for	ADP
ajst-7347	211	7	matching	match	VERB
ajst-7347	211	8	-	-	PUNCT
ajst-7347	211	9	score	score	NOUN
ajst-7347	211	10	fusion	fusion	NOUN
ajst-7347	211	11	on	on	ADP
ajst-7347	211	12	speaker	speaker	NOUN
ajst-7347	211	13	recognition[c	recognition[c	PROPN
ajst-7347	211	14	]	]	PUNCT
ajst-7347	211	15	.	.	PUNCT
ajst-7347	212	1	in	in	ADP
ajst-7347	212	2	tencon	tencon	NOUN
ajst-7347	212	3	2010	2010	NUM
ajst-7347	212	4	-	-	SYM
ajst-7347	212	5	2010	2010	NUM
ajst-7347	212	6	ieee	ieee	NOUN
ajst-7347	212	7	region	region	NOUN
ajst-7347	212	8	10	10	NUM
ajst-7347	212	9	conference	conference	NOUN
ajst-7347	212	10	.	.	PUNCT
ajst-7347	213	1	ieee	ieee	PROPN
ajst-7347	213	2	,	,	PUNCT
ajst-7347	213	3	2010	2010	NUM
ajst-7347	213	4	:	:	PUNCT
ajst-7347	213	5	736	736	NUM
ajst-7347	213	6	-	-	SYM
ajst-7347	213	7	741	741	NUM
ajst-7347	213	8	.	.	PUNCT