id	sid	tid	token	lemma	pos
easat-11054	1	1	edelweiss	edelweiss	PROPN
easat-11054	1	2	applied	apply	VERB
easat-11054	1	3	science	science	NOUN
easat-11054	1	4	and	and	CCONJ
easat-11054	1	5	technology	technology	NOUN
easat-11054	1	6	issn	issn	PROPN
easat-11054	1	7	:	:	PUNCT
easat-11054	1	8	2576	2576	NUM
easat-11054	1	9	-	-	SYM
easat-11054	1	10	8484	8484	NUM
easat-11054	1	11	vol	vol	NOUN
easat-11054	1	12	.	.	PROPN
easat-11054	2	1	9	9	NUM
easat-11054	2	2	,	,	PUNCT
easat-11054	2	3	no	no	INTJ
easat-11054	2	4	.	.	NOUN
easat-11054	2	5	11	11	NUM
easat-11054	2	6	,	,	PUNCT
easat-11054	2	7	1053	1053	NUM
easat-11054	2	8	-	-	SYM
easat-11054	2	9	1075	1075	NUM
easat-11054	2	10	2025	2025	NUM
easat-11054	2	11	publisher	publisher	NOUN
easat-11054	2	12	:	:	PUNCT
easat-11054	2	13	learning	learn	VERB
easat-11054	2	14	gate	gate	NOUN
easat-11054	2	15	doi	doi	PROPN
easat-11054	2	16	:	:	PUNCT
easat-11054	2	17	10.55214/2576	10.55214/2576	NUM
easat-11054	2	18	-	-	SYM
easat-11054	2	19	8484.v9i11.11054	8484.v9i11.11054	NOUN
easat-11054	3	1	©	©	PROPN
easat-11054	3	2	2025	2025	NUM
easat-11054	3	3	by	by	ADP
easat-11054	3	4	the	the	DET
easat-11054	3	5	authors	author	NOUN
easat-11054	3	6	;	;	PUNCT
easat-11054	3	7	licensee	licensee	PROPN
easat-11054	3	8	learning	learning	NOUN
easat-11054	3	9	gate	gate	NOUN
easat-11054	3	10	©	©	PROPN
easat-11054	3	11	2025	2025	NUM
easat-11054	3	12	by	by	ADP
easat-11054	3	13	the	the	DET
easat-11054	3	14	authors	author	NOUN
easat-11054	3	15	;	;	PUNCT
easat-11054	3	16	licensee	licensee	PROPN
easat-11054	3	17	learning	learning	NOUN
easat-11054	3	18	gate	gate	NOUN
easat-11054	3	19	history	history	NOUN
easat-11054	3	20	:	:	PUNCT
easat-11054	3	21	received	receive	VERB
easat-11054	3	22	:	:	PUNCT
easat-11054	3	23	2	2	NUM
easat-11054	3	24	september	september	PROPN
easat-11054	3	25	2025	2025	NUM
easat-11054	3	26	;	;	PUNCT
easat-11054	3	27	revised	revise	VERB
easat-11054	3	28	:	:	PUNCT
easat-11054	3	29	10	10	NUM
easat-11054	3	30	october	october	NOUN
easat-11054	3	31	2025	2025	NUM
easat-11054	3	32	;	;	PUNCT
easat-11054	3	33	accepted	accept	VERB
easat-11054	3	34	:	:	PUNCT
easat-11054	3	35	14	14	NUM
easat-11054	3	36	october	october	NOUN
easat-11054	3	37	2025	2025	NUM
easat-11054	3	38	;	;	PUNCT
easat-11054	3	39	published	publish	VERB
easat-11054	3	40	:	:	PUNCT
easat-11054	4	1	17	17	NUM
easat-11054	4	2	november	november	PROPN
easat-11054	4	3	2025	2025	NUM
easat-11054	4	4	*	*	PUNCT
easat-11054	4	5	correspondence	correspondence	NOUN
easat-11054	4	6	:	:	PUNCT
easat-11054	4	7	abs-khako@hotmail.com	abs-khako@hotmail.com	X
easat-11054	4	8	hybrid	hybrid	ADJ
easat-11054	4	9	iiot	iiot	PROPN
easat-11054	4	10	intrusion	intrusion	NOUN
easat-11054	4	11	detection	detection	NOUN
easat-11054	4	12	enhancement	enhancement	NOUN
easat-11054	4	13	with	with	ADP
easat-11054	4	14	a3c	a3c	ADJ
easat-11054	4	15	abdulssalam	abdulssalam	PROPN
easat-11054	4	16	m	m	VERB
easat-11054	4	17	khako1	khako1	NOUN
easat-11054	4	18	*	*	PROPN
easat-11054	4	19	,	,	PUNCT
easat-11054	4	20	sharaf	sharaf	VERB
easat-11054	4	21	a	a	DET
easat-11054	4	22	alhomdy2	alhomdy2	NOUN
easat-11054	5	1	1,2department	1,2department	NUM
easat-11054	5	2	of	of	ADP
easat-11054	5	3	information	information	NOUN
easat-11054	5	4	technology	technology	NOUN
easat-11054	5	5	,	,	PUNCT
easat-11054	5	6	faculty	faculty	NOUN
easat-11054	5	7	of	of	ADP
easat-11054	5	8	computer	computer	NOUN
easat-11054	5	9	and	and	CCONJ
easat-11054	5	10	information	information	NOUN
easat-11054	5	11	technology	technology	NOUN
easat-11054	5	12	,	,	PUNCT
easat-11054	5	13	sana’a	sana’a	NOUN
easat-11054	5	14	university	university	NOUN
easat-11054	5	15	,	,	PUNCT
easat-11054	5	16	sana’a	sana’a	NOUN
easat-11054	5	17	,	,	PUNCT
easat-11054	5	18	sanaa	sanaa	PROPN
easat-11054	5	19	,	,	PUNCT
easat-11054	5	20	yemen	yemen	PROPN
easat-11054	5	21	;	;	PUNCT
easat-11054	5	22	abs-khako@hotmail.com	abs-khako@hotmail.com	X
easat-11054	5	23	(	(	PUNCT
easat-11054	5	24	a.m.k	a.m.k	PROPN
easat-11054	5	25	.	.	PUNCT
easat-11054	5	26	)	)	PUNCT
easat-11054	6	1	sha.alhomdy@su.edu.ye	sha.alhomdy@su.edu.ye	PUNCT
easat-11054	6	2	(	(	PUNCT
easat-11054	6	3	s.a.a	s.a.a	PROPN
easat-11054	6	4	.	.	PUNCT
easat-11054	6	5	)	)	PUNCT
easat-11054	6	6	.	.	PUNCT
easat-11054	7	1	abstract	abstract	ADJ
easat-11054	7	2	:	:	PUNCT
easat-11054	7	3	traditional	traditional	ADJ
easat-11054	7	4	intrusion	intrusion	NOUN
easat-11054	7	5	detection	detection	NOUN
easat-11054	7	6	systems	system	NOUN
easat-11054	7	7	often	often	ADV
easat-11054	7	8	face	face	VERB
easat-11054	7	9	difficulties	difficulty	NOUN
easat-11054	7	10	in	in	ADP
easat-11054	7	11	detecting	detect	VERB
easat-11054	7	12	the	the	DET
easat-11054	7	13	diverse	diverse	ADJ
easat-11054	7	14	classes	class	NOUN
easat-11054	7	15	of	of	ADP
easat-11054	7	16	intrusions	intrusion	NOUN
easat-11054	7	17	in	in	ADP
easat-11054	7	18	the	the	DET
easat-11054	7	19	industrial	industrial	ADJ
easat-11054	7	20	internet	internet	NOUN
easat-11054	7	21	of	of	ADP
easat-11054	7	22	things	thing	NOUN
easat-11054	7	23	(	(	PUNCT
easat-11054	7	24	iiot	iiot	NOUN
easat-11054	7	25	)	)	PUNCT
easat-11054	7	26	environment	environment	NOUN
easat-11054	7	27	due	due	ADP
easat-11054	7	28	to	to	ADP
easat-11054	7	29	its	its	PRON
easat-11054	7	30	heterogeneity	heterogeneity	NOUN
easat-11054	7	31	and	and	CCONJ
easat-11054	7	32	complexity	complexity	NOUN
easat-11054	7	33	.	.	PUNCT
easat-11054	8	1	such	such	ADJ
easat-11054	8	2	environments	environment	NOUN
easat-11054	8	3	encounter	encounter	VERB
easat-11054	8	4	a	a	DET
easat-11054	8	5	mixture	mixture	NOUN
easat-11054	8	6	of	of	ADP
easat-11054	8	7	deterministic	deterministic	ADJ
easat-11054	8	8	and	and	CCONJ
easat-11054	8	9	stochastic	stochastic	ADJ
easat-11054	8	10	cyberattack	cyberattack	NOUN
easat-11054	8	11	patterns	pattern	NOUN
easat-11054	8	12	,	,	PUNCT
easat-11054	8	13	corresponding	correspond	VERB
easat-11054	8	14	to	to	PART
easat-11054	8	15	discrete	discrete	VERB
easat-11054	8	16	and	and	CCONJ
easat-11054	8	17	continuous	continuous	ADJ
easat-11054	8	18	features	feature	NOUN
easat-11054	8	19	.	.	PUNCT
easat-11054	9	1	this	this	DET
easat-11054	9	2	challenge	challenge	NOUN
easat-11054	9	3	results	result	NOUN
easat-11054	9	4	in	in	ADP
easat-11054	9	5	inefficient	inefficient	ADJ
easat-11054	9	6	feature	feature	NOUN
easat-11054	9	7	extraction	extraction	NOUN
easat-11054	9	8	in	in	ADP
easat-11054	9	9	current	current	ADJ
easat-11054	9	10	asynchronous	asynchronous	ADJ
easat-11054	9	11	advantage	advantage	NOUN
easat-11054	9	12	actor	actor	NOUN
easat-11054	9	13	-	-	PUNCT
easat-11054	9	14	critic	critic	NOUN
easat-11054	9	15	(	(	PUNCT
easat-11054	9	16	a3c	a3c	NUM
easat-11054	9	17	)	)	PUNCT
easat-11054	9	18	algorithms	algorithm	NOUN
easat-11054	9	19	.	.	PUNCT
easat-11054	10	1	to	to	PART
easat-11054	10	2	address	address	VERB
easat-11054	10	3	these	these	DET
easat-11054	10	4	issues	issue	NOUN
easat-11054	10	5	,	,	PUNCT
easat-11054	10	6	this	this	DET
easat-11054	10	7	article	article	NOUN
easat-11054	10	8	proposes	propose	VERB
easat-11054	10	9	a	a	DET
easat-11054	10	10	hybrid	hybrid	NOUN
easat-11054	10	11	a3c	a3c	PUNCT
easat-11054	10	12	with	with	ADP
easat-11054	10	13	decision	decision	NOUN
easat-11054	10	14	tree	tree	NOUN
easat-11054	10	15	(	(	PUNCT
easat-11054	10	16	hadt	hadt	NOUN
easat-11054	10	17	)	)	PUNCT
easat-11054	10	18	algorithm	algorithm	NOUN
easat-11054	10	19	for	for	ADP
easat-11054	10	20	intrusion	intrusion	NOUN
easat-11054	10	21	detection	detection	NOUN
easat-11054	10	22	.	.	PUNCT
easat-11054	11	1	in	in	ADP
easat-11054	11	2	this	this	DET
easat-11054	11	3	proposed	propose	VERB
easat-11054	11	4	algorithm	algorithm	NOUN
easat-11054	11	5	,	,	PUNCT
easat-11054	11	6	the	the	DET
easat-11054	11	7	simple	simple	ADJ
easat-11054	11	8	a3c	a3c	NUM
easat-11054	11	9	is	be	AUX
easat-11054	11	10	used	use	VERB
easat-11054	11	11	to	to	PART
easat-11054	11	12	detect	detect	VERB
easat-11054	11	13	stochastic	stochastic	ADJ
easat-11054	11	14	intrusions	intrusion	NOUN
easat-11054	11	15	related	relate	VERB
easat-11054	11	16	to	to	ADP
easat-11054	11	17	continuous	continuous	ADJ
easat-11054	11	18	features	feature	NOUN
easat-11054	11	19	,	,	PUNCT
easat-11054	11	20	while	while	SCONJ
easat-11054	11	21	the	the	DET
easat-11054	11	22	decision	decision	NOUN
easat-11054	11	23	tree	tree	NOUN
easat-11054	11	24	(	(	PUNCT
easat-11054	11	25	dt	dt	NOUN
easat-11054	11	26	)	)	PUNCT
easat-11054	11	27	algorithm	algorithm	NOUN
easat-11054	11	28	is	be	AUX
easat-11054	11	29	employed	employ	VERB
easat-11054	11	30	for	for	ADP
easat-11054	11	31	detecting	detect	VERB
easat-11054	11	32	fixed	fix	VERB
easat-11054	11	33	-	-	PUNCT
easat-11054	11	34	pattern	pattern	NOUN
easat-11054	11	35	(	(	PUNCT
easat-11054	11	36	deterministic	deterministic	ADJ
easat-11054	11	37	)	)	PUNCT
easat-11054	11	38	intrusions	intrusion	NOUN
easat-11054	11	39	related	relate	VERB
easat-11054	11	40	to	to	PART
easat-11054	11	41	discrete	discrete	VERB
easat-11054	11	42	features	feature	NOUN
easat-11054	11	43	.	.	PUNCT
easat-11054	12	1	the	the	DET
easat-11054	12	2	outputs	output	NOUN
easat-11054	12	3	of	of	ADP
easat-11054	12	4	these	these	DET
easat-11054	12	5	algorithms	algorithm	NOUN
easat-11054	12	6	are	be	AUX
easat-11054	12	7	combined	combine	VERB
easat-11054	12	8	in	in	ADP
easat-11054	12	9	a	a	DET
easat-11054	12	10	feedforward	feedforward	ADJ
easat-11054	12	11	neural	neural	ADJ
easat-11054	12	12	network	network	NOUN
easat-11054	12	13	(	(	PUNCT
easat-11054	12	14	ffnn	ffnn	NOUN
easat-11054	12	15	)	)	PUNCT
easat-11054	12	16	layer	layer	NOUN
easat-11054	12	17	to	to	PART
easat-11054	12	18	accurately	accurately	ADV
easat-11054	12	19	classify	classify	VERB
easat-11054	12	20	the	the	DET
easat-11054	12	21	intrusion	intrusion	NOUN
easat-11054	12	22	type	type	NOUN
easat-11054	12	23	.	.	PUNCT
easat-11054	13	1	additionally	additionally	ADV
easat-11054	13	2	,	,	PUNCT
easat-11054	13	3	the	the	DET
easat-11054	13	4	x	x	ADJ
easat-11054	13	5	-	-	ADJ
easat-11054	13	6	iiotid	iiotid	ADJ
easat-11054	13	7	dataset	dataset	NOUN
easat-11054	13	8	,	,	PUNCT
easat-11054	13	9	obtained	obtain	VERB
easat-11054	13	10	from	from	ADP
easat-11054	13	11	a	a	DET
easat-11054	13	12	real	real	ADJ
easat-11054	13	13	iiot	iiot	NOUN
easat-11054	13	14	environment	environment	NOUN
easat-11054	13	15	,	,	PUNCT
easat-11054	13	16	was	be	AUX
easat-11054	13	17	used	use	VERB
easat-11054	13	18	to	to	PART
easat-11054	13	19	compare	compare	VERB
easat-11054	13	20	the	the	DET
easat-11054	13	21	performance	performance	NOUN
easat-11054	13	22	of	of	ADP
easat-11054	13	23	hadt	hadt	NOUN
easat-11054	13	24	with	with	ADP
easat-11054	13	25	existing	exist	VERB
easat-11054	13	26	a3c	a3c	NUM
easat-11054	13	27	algorithms	algorithm	NOUN
easat-11054	13	28	in	in	ADP
easat-11054	13	29	related	related	ADJ
easat-11054	13	30	studies	study	NOUN
easat-11054	13	31	.	.	PUNCT
easat-11054	14	1	experimental	experimental	ADJ
easat-11054	14	2	results	result	NOUN
easat-11054	14	3	demonstrate	demonstrate	VERB
easat-11054	14	4	that	that	SCONJ
easat-11054	14	5	the	the	DET
easat-11054	14	6	hadt	hadt	NOUN
easat-11054	14	7	algorithm	algorithm	NOUN
easat-11054	14	8	outperforms	outperform	NOUN
easat-11054	14	9	baseline	baseline	VERB
easat-11054	14	10	a3c	a3c	ADJ
easat-11054	14	11	algorithms	algorithm	NOUN
easat-11054	14	12	in	in	ADP
easat-11054	14	13	terms	term	NOUN
easat-11054	14	14	of	of	ADP
easat-11054	14	15	accuracy	accuracy	NOUN
easat-11054	14	16	,	,	PUNCT
easat-11054	14	17	precision	precision	NOUN
easat-11054	14	18	,	,	PUNCT
easat-11054	14	19	recall	recall	NOUN
easat-11054	14	20	,	,	PUNCT
easat-11054	14	21	and	and	CCONJ
easat-11054	14	22	f1	f1	PROPN
easat-11054	14	23	score	score	NOUN
easat-11054	14	24	,	,	PUNCT
easat-11054	14	25	maintaining	maintain	VERB
easat-11054	14	26	a	a	DET
easat-11054	14	27	detection	detection	NOUN
easat-11054	14	28	rate	rate	NOUN
easat-11054	14	29	above	above	ADP
easat-11054	14	30	99.8	99.8	NUM
easat-11054	14	31	%	%	NOUN
easat-11054	14	32	,	,	PUNCT
easat-11054	14	33	especially	especially	ADV
easat-11054	14	34	for	for	ADP
easat-11054	14	35	intrusions	intrusion	NOUN
easat-11054	14	36	that	that	PRON
easat-11054	14	37	are	be	AUX
easat-11054	14	38	underrepresented	underrepresente	VERB
easat-11054	14	39	in	in	ADP
easat-11054	14	40	the	the	DET
easat-11054	14	41	dataset	dataset	NOUN
easat-11054	14	42	,	,	PUNCT
easat-11054	14	43	while	while	SCONJ
easat-11054	14	44	also	also	ADV
easat-11054	14	45	consuming	consume	VERB
easat-11054	14	46	less	less	ADJ
easat-11054	14	47	power	power	NOUN
easat-11054	14	48	.	.	PUNCT
easat-11054	15	1	keywords	keyword	NOUN
easat-11054	15	2	:	:	PUNCT
easat-11054	15	3	a3c	a3c	VERB
easat-11054	15	4	,	,	PUNCT
easat-11054	15	5	dt	dt	INTJ
easat-11054	15	6	,	,	PUNCT
easat-11054	15	7	ffnn	ffnn	NOUN
easat-11054	15	8	,	,	PUNCT
easat-11054	15	9	hadt	hadt	NOUN
easat-11054	15	10	,	,	PUNCT
easat-11054	15	11	iiot	iiot	NOUN
easat-11054	15	12	,	,	PUNCT
easat-11054	15	13	x	x	NOUN
easat-11054	15	14	-	-	NOUN
easat-11054	15	15	iiotid	iiotid	NOUN
easat-11054	15	16	.	.	PUNCT
easat-11054	16	1	1	1	X
easat-11054	16	2	.	.	X
easat-11054	16	3	introduction	introduction	NOUN
easat-11054	16	4	in	in	ADP
easat-11054	16	5	the	the	DET
easat-11054	16	6	iiot	iiot	NOUN
easat-11054	16	7	environment	environment	NOUN
easat-11054	16	8	,	,	PUNCT
easat-11054	16	9	making	make	VERB
easat-11054	16	10	informed	informed	ADJ
easat-11054	16	11	decisions	decision	NOUN
easat-11054	16	12	is	be	AUX
easat-11054	16	13	crucial	crucial	ADJ
easat-11054	16	14	when	when	SCONJ
easat-11054	16	15	exchanging	exchange	VERB
easat-11054	16	16	data	datum	NOUN
easat-11054	16	17	efficiently	efficiently	ADV
easat-11054	16	18	across	across	ADP
easat-11054	16	19	multiple	multiple	ADJ
easat-11054	16	20	domains	domain	NOUN
easat-11054	16	21	.	.	PUNCT
easat-11054	17	1	however	however	ADV
easat-11054	17	2	,	,	PUNCT
easat-11054	17	3	the	the	DET
easat-11054	17	4	rapid	rapid	ADJ
easat-11054	17	5	growth	growth	NOUN
easat-11054	17	6	of	of	ADP
easat-11054	17	7	this	this	DET
easat-11054	17	8	environment	environment	NOUN
easat-11054	17	9	has	have	AUX
easat-11054	17	10	created	create	VERB
easat-11054	17	11	significant	significant	ADJ
easat-11054	17	12	challenges	challenge	NOUN
easat-11054	17	13	in	in	ADP
easat-11054	17	14	cybersecurity	cybersecurity	NOUN
easat-11054	17	15	,	,	PUNCT
easat-11054	17	16	requiring	require	VERB
easat-11054	17	17	sophisticated	sophisticated	ADJ
easat-11054	17	18	intrusion	intrusion	NOUN
easat-11054	17	19	detection	detection	NOUN
easat-11054	17	20	systems	system	NOUN
easat-11054	17	21	(	(	PUNCT
easat-11054	17	22	ids	id	NOUN
easat-11054	17	23	)	)	PUNCT
easat-11054	18	1	[	[	X
easat-11054	18	2	1	1	NUM
easat-11054	18	3	]	]	PUNCT
easat-11054	18	4	.	.	PUNCT
easat-11054	19	1	although	although	SCONJ
easat-11054	19	2	basic	basic	ADJ
easat-11054	19	3	protection	protection	NOUN
easat-11054	19	4	is	be	AUX
easat-11054	19	5	provided	provide	VERB
easat-11054	19	6	for	for	ADP
easat-11054	19	7	the	the	DET
easat-11054	19	8	iiot	iiot	NOUN
easat-11054	19	9	network	network	NOUN
easat-11054	19	10	,	,	PUNCT
easat-11054	19	11	current	current	ADJ
easat-11054	19	12	security	security	NOUN
easat-11054	19	13	technologies	technology	NOUN
easat-11054	19	14	,	,	PUNCT
easat-11054	19	15	such	such	ADJ
easat-11054	19	16	as	as	ADP
easat-11054	19	17	firewalls	firewall	NOUN
easat-11054	19	18	and	and	CCONJ
easat-11054	19	19	anti	anti	ADJ
easat-11054	19	20	-	-	ADJ
easat-11054	19	21	theft	theft	ADJ
easat-11054	19	22	mechanisms	mechanism	NOUN
easat-11054	19	23	,	,	PUNCT
easat-11054	19	24	are	be	AUX
easat-11054	19	25	still	still	ADV
easat-11054	19	26	insufficient	insufficient	ADJ
easat-11054	19	27	to	to	PART
easat-11054	19	28	address	address	VERB
easat-11054	19	29	these	these	DET
easat-11054	19	30	cyber	cyber	ADJ
easat-11054	19	31	challenges	challenge	NOUN
easat-11054	19	32	[	[	X
easat-11054	19	33	2	2	NUM
easat-11054	19	34	]	]	PUNCT
easat-11054	19	35	.	.	PUNCT
easat-11054	20	1	these	these	DET
easat-11054	20	2	technologies	technology	NOUN
easat-11054	20	3	often	often	ADV
easat-11054	20	4	fail	fail	VERB
easat-11054	20	5	to	to	PART
easat-11054	20	6	detect	detect	VERB
easat-11054	20	7	skilled	skilled	ADJ
easat-11054	20	8	hackers	hacker	NOUN
easat-11054	20	9	who	who	PRON
easat-11054	20	10	use	use	VERB
easat-11054	20	11	complex	complex	ADJ
easat-11054	20	12	and	and	CCONJ
easat-11054	20	13	sophisticated	sophisticated	ADJ
easat-11054	20	14	intrusions	intrusion	NOUN
easat-11054	20	15	[	[	X
easat-11054	20	16	3	3	NUM
easat-11054	20	17	]	]	PUNCT
easat-11054	20	18	.	.	PUNCT
easat-11054	21	1	these	these	DET
easat-11054	21	2	intrusions	intrusion	NOUN
easat-11054	21	3	could	could	AUX
easat-11054	21	4	lead	lead	VERB
easat-11054	21	5	to	to	ADP
easat-11054	21	6	huge	huge	ADJ
easat-11054	21	7	financial	financial	ADJ
easat-11054	21	8	and	and	CCONJ
easat-11054	21	9	economic	economic	ADJ
easat-11054	21	10	losses	loss	NOUN
easat-11054	21	11	.	.	PUNCT
easat-11054	22	1	current	current	ADJ
easat-11054	22	2	idss	idss	NOUN
easat-11054	22	3	used	use	VERB
easat-11054	22	4	for	for	ADP
easat-11054	22	5	intensive	intensive	ADJ
easat-11054	22	6	and	and	CCONJ
easat-11054	22	7	anomaly	anomaly	NOUN
easat-11054	22	8	detection	detection	NOUN
easat-11054	22	9	have	have	AUX
easat-11054	22	10	faced	face	VERB
easat-11054	22	11	several	several	ADJ
easat-11054	22	12	performance	performance	NOUN
easat-11054	22	13	measurement	measurement	NOUN
easat-11054	22	14	problems	problem	NOUN
easat-11054	22	15	,	,	PUNCT
easat-11054	22	16	including	include	VERB
easat-11054	22	17	limited	limited	ADJ
easat-11054	22	18	detection	detection	NOUN
easat-11054	22	19	accuracy	accuracy	NOUN
easat-11054	22	20	,	,	PUNCT
easat-11054	22	21	precision	precision	NOUN
easat-11054	22	22	,	,	PUNCT
easat-11054	22	23	recall	recall	NOUN
easat-11054	22	24	,	,	PUNCT
easat-11054	22	25	and	and	CCONJ
easat-11054	22	26	f1	f1	ADJ
easat-11054	22	27	score	score	NOUN
easat-11054	23	1	[	[	X
easat-11054	23	2	4	4	NUM
easat-11054	23	3	]	]	PUNCT
easat-11054	23	4	.	.	PUNCT
easat-11054	24	1	these	these	DET
easat-11054	24	2	problems	problem	NOUN
easat-11054	24	3	reduce	reduce	VERB
easat-11054	24	4	the	the	DET
easat-11054	24	5	effectiveness	effectiveness	NOUN
easat-11054	24	6	of	of	ADP
easat-11054	24	7	securing	secure	VERB
easat-11054	24	8	iiot	iiot	NOUN
easat-11054	24	9	networks	network	NOUN
easat-11054	24	10	[	[	X
easat-11054	24	11	5	5	NUM
easat-11054	24	12	]	]	PUNCT
easat-11054	24	13	.	.	PUNCT
easat-11054	25	1	in	in	ADP
easat-11054	25	2	recent	recent	ADJ
easat-11054	25	3	years	year	NOUN
easat-11054	25	4	,	,	PUNCT
easat-11054	25	5	actor	actor	NOUN
easat-11054	25	6	-	-	PUNCT
easat-11054	25	7	critic	critic	NOUN
easat-11054	25	8	algorithms	algorithm	NOUN
easat-11054	25	9	have	have	AUX
easat-11054	25	10	emerged	emerge	VERB
easat-11054	25	11	to	to	PART
easat-11054	25	12	address	address	VERB
easat-11054	25	13	these	these	DET
easat-11054	25	14	challenges	challenge	NOUN
easat-11054	25	15	.	.	PUNCT
easat-11054	26	1	they	they	PRON
easat-11054	26	2	have	have	AUX
easat-11054	26	3	been	be	AUX
easat-11054	26	4	developed	develop	VERB
easat-11054	26	5	with	with	ADP
easat-11054	26	6	adaptive	adaptive	ADJ
easat-11054	26	7	learning	learn	VERB
easat-11054	26	8	capabilities	capability	NOUN
easat-11054	26	9	to	to	PART
easat-11054	26	10	improve	improve	VERB
easat-11054	26	11	idss	idss	ADV
easat-11054	27	1	[	[	X
easat-11054	27	2	6	6	NUM
easat-11054	27	3	]	]	PUNCT
easat-11054	27	4	.	.	PUNCT
easat-11054	28	1	actor	actor	NOUN
easat-11054	28	2	-	-	PUNCT
easat-11054	28	3	critic	critic	NOUN
easat-11054	28	4	algorithms	algorithm	NOUN
easat-11054	28	5	utilize	utilize	VERB
easat-11054	28	6	neural	neural	ADJ
easat-11054	28	7	network	network	NOUN
easat-11054	28	8	(	(	PUNCT
easat-11054	28	9	nn	nn	NOUN
easat-11054	28	10	)	)	PUNCT
easat-11054	28	11	layers	layer	NOUN
easat-11054	28	12	within	within	ADP
easat-11054	28	13	workers	worker	NOUN
easat-11054	28	14	.	.	PUNCT
easat-11054	29	1	however	however	ADV
easat-11054	29	2	,	,	PUNCT
easat-11054	29	3	these	these	DET
easat-11054	29	4	algorithms	algorithm	NOUN
easat-11054	29	5	face	face	VERB
easat-11054	29	6	computational	computational	ADJ
easat-11054	29	7	challenges	challenge	NOUN
easat-11054	29	8	when	when	SCONJ
easat-11054	29	9	processing	process	VERB
easat-11054	29	10	large	large	ADJ
easat-11054	29	11	-	-	PUNCT
easat-11054	29	12	scale	scale	NOUN
easat-11054	29	13	traffic	traffic	NOUN
easat-11054	29	14	data	datum	NOUN
easat-11054	29	15	or	or	CCONJ
easat-11054	29	16	high	high	ADJ
easat-11054	29	17	-	-	PUNCT
easat-11054	29	18	volume	volume	NOUN
easat-11054	29	19	network	network	NOUN
easat-11054	29	20	activities	activity	NOUN
easat-11054	29	21	[	[	X
easat-11054	29	22	7	7	NUM
easat-11054	29	23	]	]	PUNCT
easat-11054	29	24	.	.	PUNCT
easat-11054	30	1	furthermore	furthermore	ADV
easat-11054	30	2	,	,	PUNCT
easat-11054	30	3	the	the	DET
easat-11054	30	4	intrusion	intrusion	NOUN
easat-11054	30	5	detection	detection	NOUN
easat-11054	30	6	capability	capability	NOUN
easat-11054	30	7	of	of	ADP
easat-11054	30	8	the	the	DET
easat-11054	30	9	iiot	iiot	PROPN
easat-11054	30	10	environment	environment	NOUN
easat-11054	30	11	,	,	PUNCT
easat-11054	30	12	which	which	PRON
easat-11054	30	13	is	be	AUX
easat-11054	30	14	a	a	DET
easat-11054	30	15	mixture	mixture	NOUN
easat-11054	30	16	of	of	ADP
easat-11054	30	17	stochastic	stochastic	ADJ
easat-11054	30	18	and	and	CCONJ
easat-11054	30	19	deterministic	deterministic	ADJ
easat-11054	30	20	environments	environment	NOUN
easat-11054	30	21	,	,	PUNCT
easat-11054	30	22	remains	remain	VERB
easat-11054	30	23	limited	limited	ADJ
easat-11054	30	24	.	.	PUNCT
easat-11054	31	1	this	this	PRON
easat-11054	31	2	poses	pose	VERB
easat-11054	31	3	additional	additional	ADJ
easat-11054	31	4	obstacles	obstacle	NOUN
easat-11054	31	5	during	during	ADP
easat-11054	31	6	the	the	DET
easat-11054	31	7	training	training	NOUN
easat-11054	31	8	and	and	CCONJ
easat-11054	31	9	testing	testing	NOUN
easat-11054	31	10	phases	phase	NOUN
easat-11054	31	11	[	[	X
easat-11054	31	12	8	8	NUM
easat-11054	31	13	]	]	PUNCT
easat-11054	31	14	.	.	PUNCT
easat-11054	32	1	to	to	PART
easat-11054	32	2	overcome	overcome	VERB
easat-11054	32	3	these	these	DET
easat-11054	32	4	challenges	challenge	NOUN
easat-11054	32	5	,	,	PUNCT
easat-11054	32	6	researchers	researcher	NOUN
easat-11054	32	7	have	have	AUX
easat-11054	32	8	proposed	propose	VERB
easat-11054	32	9	various	various	ADJ
easat-11054	32	10	enhancements	enhancement	NOUN
easat-11054	32	11	.	.	PUNCT
easat-11054	33	1	for	for	ADP
easat-11054	33	2	instance	instance	NOUN
easat-11054	33	3	,	,	PUNCT
easat-11054	33	4	the	the	DET
easat-11054	33	5	a3c	a3c	NUM
easat-11054	33	6	algorithm	algorithm	NOUN
easat-11054	33	7	was	be	AUX
easat-11054	33	8	proposed	propose	VERB
easat-11054	33	9	to	to	PART
easat-11054	33	10	use	use	VERB
easat-11054	33	11	cnn	cnn	PROPN
easat-11054	33	12	inside	inside	ADP
easat-11054	33	13	its	its	PRON
easat-11054	33	14	workers	worker	NOUN
easat-11054	33	15	for	for	ADP
easat-11054	33	16	anomaly	anomaly	NOUN
easat-11054	33	17	detection	detection	NOUN
easat-11054	34	1	[	[	X
easat-11054	34	2	9	9	NUM
easat-11054	34	3	]	]	PUNCT
easat-11054	34	4	.	.	PUNCT
easat-11054	35	1	similarly	similarly	ADV
easat-11054	35	2	,	,	PUNCT
easat-11054	35	3	in	in	ADP
easat-11054	35	4	optimized	optimize	VERB
easat-11054	35	5	a3c	a3c	NUM
easat-11054	35	6	,	,	PUNCT
easat-11054	35	7	policies	policy	NOUN
easat-11054	35	8	and	and	CCONJ
easat-11054	35	9	value	value	NOUN
easat-11054	35	10	function	function	NOUN
easat-11054	35	11	parameters	parameter	NOUN
easat-11054	35	12	were	be	AUX
easat-11054	35	13	1054	1054	NUM
easat-11054	35	14	edelweiss	edelweiss	PROPN
easat-11054	35	15	applied	apply	VERB
easat-11054	35	16	science	science	NOUN
easat-11054	35	17	and	and	CCONJ
easat-11054	35	18	technology	technology	NOUN
easat-11054	35	19	issn	issn	PROPN
easat-11054	35	20	:	:	PUNCT
easat-11054	35	21	2576	2576	NUM
easat-11054	35	22	-	-	SYM
easat-11054	35	23	8484	8484	NUM
easat-11054	35	24	vol	vol	NOUN
easat-11054	35	25	.	.	PROPN
easat-11054	36	1	9	9	NUM
easat-11054	36	2	,	,	PUNCT
easat-11054	36	3	no	no	INTJ
easat-11054	36	4	.	.	NOUN
easat-11054	36	5	11	11	NUM
easat-11054	36	6	:	:	PUNCT
easat-11054	36	7	1053	1053	NUM
easat-11054	36	8	-	-	SYM
easat-11054	36	9	1075	1075	NUM
easat-11054	36	10	,	,	PUNCT
easat-11054	36	11	2025	2025	NUM
easat-11054	36	12	doi	doi	NOUN
easat-11054	36	13	:	:	PUNCT
easat-11054	36	14	10.55214/2576	10.55214/2576	NUM
easat-11054	36	15	-	-	SYM
easat-11054	36	16	8484.v9i11.11054	8484.v9i11.11054	NOUN
easat-11054	37	1	©	©	PROPN
easat-11054	37	2	2025	2025	NUM
easat-11054	37	3	by	by	ADP
easat-11054	37	4	the	the	DET
easat-11054	37	5	authors	author	NOUN
easat-11054	37	6	;	;	PUNCT
easat-11054	37	7	licensee	licensee	PROPN
easat-11054	37	8	learning	learning	PROPN
easat-11054	37	9	gate	gate	PROPN
easat-11054	37	10	proposed	propose	VERB
easat-11054	37	11	to	to	PART
easat-11054	37	12	enhance	enhance	VERB
easat-11054	37	13	security	security	NOUN
easat-11054	37	14	in	in	ADP
easat-11054	37	15	cognitive	cognitive	ADJ
easat-11054	37	16	networks	network	NOUN
easat-11054	37	17	[	[	X
easat-11054	37	18	10	10	NUM
easat-11054	37	19	]	]	PUNCT
easat-11054	37	20	.	.	PUNCT
easat-11054	38	1	another	another	DET
easat-11054	38	2	study	study	NOUN
easat-11054	38	3	,	,	PUNCT
easat-11054	38	4	lim	lim	PROPN
easat-11054	38	5	and	and	CCONJ
easat-11054	38	6	joe	joe	PROPN
easat-11054	39	1	[	[	X
easat-11054	39	2	11	11	NUM
easat-11054	39	3	]	]	PUNCT
easat-11054	39	4	,	,	PUNCT
easat-11054	39	5	the	the	DET
easat-11054	39	6	authors	author	NOUN
easat-11054	39	7	explored	explore	VERB
easat-11054	39	8	off	off	ADP
easat-11054	39	9	-	-	PUNCT
easat-11054	39	10	policy	policy	NOUN
easat-11054	39	11	actor	actor	NOUN
easat-11054	39	12	-	-	PUNCT
easat-11054	39	13	critic	critic	NOUN
easat-11054	39	14	drl	drl	PROPN
easat-11054	39	15	techniques	technique	NOUN
easat-11054	39	16	for	for	ADP
easat-11054	39	17	prioritizing	prioritize	VERB
easat-11054	39	18	alerts	alert	NOUN
easat-11054	39	19	in	in	ADP
easat-11054	39	20	ids	id	NOUN
easat-11054	39	21	environments	environment	NOUN
easat-11054	39	22	by	by	ADP
easat-11054	39	23	describing	describe	VERB
easat-11054	39	24	the	the	DET
easat-11054	39	25	effectiveness	effectiveness	NOUN
easat-11054	39	26	of	of	ADP
easat-11054	39	27	drl	drl	PROPN
easat-11054	39	28	in	in	ADP
easat-11054	39	29	complex	complex	ADJ
easat-11054	39	30	security	security	NOUN
easat-11054	39	31	scenarios	scenario	NOUN
easat-11054	39	32	.	.	PUNCT
easat-11054	40	1	despite	despite	SCONJ
easat-11054	40	2	these	these	DET
easat-11054	40	3	improvements	improvement	NOUN
easat-11054	40	4	,	,	PUNCT
easat-11054	40	5	there	there	PRON
easat-11054	40	6	must	must	AUX
easat-11054	40	7	be	be	AUX
easat-11054	40	8	a	a	DET
easat-11054	40	9	balance	balance	NOUN
easat-11054	40	10	between	between	ADP
easat-11054	40	11	exploration	exploration	NOUN
easat-11054	40	12	and	and	CCONJ
easat-11054	40	13	exploitation	exploitation	NOUN
easat-11054	40	14	trade	trade	NOUN
easat-11054	40	15	-	-	PUNCT
easat-11054	40	16	offs	off	NOUN
easat-11054	40	17	in	in	ADP
easat-11054	40	18	idss	idss	ADV
easat-11054	40	19	based	base	VERB
easat-11054	40	20	on	on	ADP
easat-11054	40	21	actor	actor	NOUN
easat-11054	40	22	-	-	PUNCT
easat-11054	40	23	critic	critic	NOUN
easat-11054	40	24	algorithms	algorithm	NOUN
easat-11054	40	25	such	such	ADJ
easat-11054	40	26	as	as	ADP
easat-11054	40	27	a3c	a3c	NUM
easat-11054	40	28	,	,	PUNCT
easat-11054	40	29	which	which	PRON
easat-11054	40	30	remains	remain	VERB
easat-11054	40	31	a	a	DET
easat-11054	40	32	major	major	ADJ
easat-11054	40	33	concern	concern	NOUN
easat-11054	40	34	,	,	PUNCT
easat-11054	40	35	especially	especially	ADV
easat-11054	40	36	in	in	ADP
easat-11054	40	37	complex	complex	ADJ
easat-11054	40	38	environments	environment	NOUN
easat-11054	40	39	such	such	ADJ
easat-11054	40	40	as	as	ADP
easat-11054	40	41	iiot	iiot	ADJ
easat-11054	40	42	environments	environment	NOUN
easat-11054	40	43	that	that	PRON
easat-11054	40	44	have	have	VERB
easat-11054	40	45	continuous	continuous	ADJ
easat-11054	40	46	and	and	CCONJ
easat-11054	40	47	discrete	discrete	ADJ
easat-11054	40	48	intrusion	intrusion	NOUN
easat-11054	40	49	patterns	pattern	NOUN
easat-11054	40	50	[	[	X
easat-11054	40	51	12	12	NUM
easat-11054	40	52	]	]	PUNCT
easat-11054	40	53	.	.	PUNCT
easat-11054	41	1	to	to	PART
easat-11054	41	2	address	address	VERB
easat-11054	41	3	this	this	PRON
easat-11054	41	4	,	,	PUNCT
easat-11054	41	5	the	the	DET
easat-11054	41	6	article	article	NOUN
easat-11054	41	7	proposes	propose	VERB
easat-11054	41	8	an	an	DET
easat-11054	41	9	hadt	hadt	NOUN
easat-11054	41	10	algorithm	algorithm	NOUN
easat-11054	41	11	,	,	PUNCT
easat-11054	41	12	which	which	PRON
easat-11054	41	13	is	be	AUX
easat-11054	41	14	a	a	DET
easat-11054	41	15	hybrid	hybrid	ADJ
easat-11054	41	16	integrating	integrate	VERB
easat-11054	41	17	an	an	DET
easat-11054	41	18	a3c	a3c	ADJ
easat-11054	41	19	algorithm	algorithm	NOUN
easat-11054	41	20	with	with	ADP
easat-11054	41	21	a	a	DET
easat-11054	41	22	dt	dt	NOUN
easat-11054	41	23	algorithm	algorithm	NOUN
easat-11054	41	24	.	.	PUNCT
easat-11054	42	1	this	this	DET
easat-11054	42	2	integration	integration	NOUN
easat-11054	42	3	aims	aim	VERB
easat-11054	42	4	to	to	PART
easat-11054	42	5	mitigate	mitigate	VERB
easat-11054	42	6	the	the	DET
easat-11054	42	7	exploration	exploration	NOUN
easat-11054	42	8	and	and	CCONJ
easat-11054	42	9	exploitation	exploitation	NOUN
easat-11054	42	10	problem	problem	NOUN
easat-11054	42	11	while	while	SCONJ
easat-11054	42	12	improving	improve	VERB
easat-11054	42	13	classification	classification	NOUN
easat-11054	42	14	performance	performance	NOUN
easat-11054	42	15	.	.	PUNCT
easat-11054	43	1	using	use	VERB
easat-11054	43	2	hadt	hadt	NOUN
easat-11054	43	3	algorithms	algorithm	NOUN
easat-11054	43	4	for	for	ADP
easat-11054	43	5	stochastic	stochastic	ADJ
easat-11054	43	6	and	and	CCONJ
easat-11054	43	7	deterministic	deterministic	ADJ
easat-11054	43	8	environments	environment	NOUN
easat-11054	43	9	is	be	AUX
easat-11054	43	10	more	more	ADV
easat-11054	43	11	effective	effective	ADJ
easat-11054	43	12	in	in	ADP
easat-11054	43	13	dealing	deal	VERB
easat-11054	43	14	with	with	ADP
easat-11054	43	15	intrusion	intrusion	NOUN
easat-11054	43	16	patterns	pattern	NOUN
easat-11054	43	17	that	that	PRON
easat-11054	43	18	rely	rely	VERB
easat-11054	43	19	on	on	ADP
easat-11054	43	20	continuous	continuous	ADJ
easat-11054	43	21	and	and	CCONJ
easat-11054	43	22	discrete	discrete	ADJ
easat-11054	43	23	features	feature	NOUN
easat-11054	43	24	,	,	PUNCT
easat-11054	43	25	respectively	respectively	ADV
easat-11054	43	26	,	,	PUNCT
easat-11054	43	27	with	with	SCONJ
easat-11054	43	28	fewer	few	ADJ
easat-11054	43	29	neural	neural	ADJ
easat-11054	43	30	network	network	NOUN
easat-11054	43	31	layers	layer	NOUN
easat-11054	43	32	to	to	PART
easat-11054	43	33	reduce	reduce	VERB
easat-11054	43	34	processing	processing	NOUN
easat-11054	43	35	operations	operation	NOUN
easat-11054	43	36	,	,	PUNCT
easat-11054	43	37	which	which	PRON
easat-11054	43	38	contributes	contribute	VERB
easat-11054	43	39	to	to	ADP
easat-11054	43	40	improving	improve	VERB
easat-11054	43	41	processing	processing	NOUN
easat-11054	43	42	time	time	NOUN
easat-11054	43	43	and	and	CCONJ
easat-11054	43	44	power	power	NOUN
easat-11054	43	45	consumption	consumption	NOUN
easat-11054	43	46	.	.	PUNCT
easat-11054	44	1	on	on	ADP
easat-11054	44	2	the	the	DET
easat-11054	44	3	other	other	ADJ
easat-11054	44	4	hand	hand	NOUN
easat-11054	44	5	,	,	PUNCT
easat-11054	44	6	to	to	PART
easat-11054	44	7	benefit	benefit	VERB
easat-11054	44	8	from	from	ADP
easat-11054	44	9	the	the	DET
easat-11054	44	10	effectiveness	effectiveness	NOUN
easat-11054	44	11	of	of	ADP
easat-11054	44	12	the	the	DET
easat-11054	44	13	proposed	propose	VERB
easat-11054	44	14	algorithm	algorithm	NOUN
easat-11054	44	15	,	,	PUNCT
easat-11054	44	16	it	it	PRON
easat-11054	44	17	requires	require	VERB
easat-11054	44	18	preprocessing	preprocesse	VERB
easat-11054	44	19	the	the	DET
easat-11054	44	20	dataset	dataset	NOUN
easat-11054	44	21	before	before	ADP
easat-11054	44	22	the	the	DET
easat-11054	44	23	training	training	NOUN
easat-11054	44	24	process	process	NOUN
easat-11054	44	25	to	to	PART
easat-11054	44	26	ensure	ensure	VERB
easat-11054	44	27	the	the	DET
easat-11054	44	28	efficiency	efficiency	NOUN
easat-11054	44	29	of	of	ADP
easat-11054	44	30	feature	feature	NOUN
easat-11054	44	31	extraction	extraction	NOUN
easat-11054	44	32	.	.	PUNCT
easat-11054	45	1	the	the	DET
easat-11054	45	2	results	result	NOUN
easat-11054	45	3	of	of	ADP
easat-11054	45	4	the	the	DET
easat-11054	45	5	hadt	hadt	NOUN
easat-11054	45	6	algorithm	algorithm	NOUN
easat-11054	45	7	confirm	confirm	VERB
easat-11054	45	8	enhanced	enhanced	ADJ
easat-11054	45	9	intrusion	intrusion	NOUN
easat-11054	45	10	detection	detection	NOUN
easat-11054	45	11	accuracy	accuracy	NOUN
easat-11054	45	12	,	,	PUNCT
easat-11054	45	13	precision	precision	NOUN
easat-11054	45	14	,	,	PUNCT
easat-11054	45	15	recall	recall	NOUN
easat-11054	45	16	,	,	PUNCT
easat-11054	45	17	and	and	CCONJ
easat-11054	45	18	f1score	f1score	NOUN
easat-11054	45	19	,	,	PUNCT
easat-11054	45	20	which	which	PRON
easat-11054	45	21	make	make	VERB
easat-11054	45	22	it	it	PRON
easat-11054	45	23	a	a	DET
easat-11054	45	24	promising	promising	ADJ
easat-11054	45	25	solution	solution	NOUN
easat-11054	45	26	for	for	ADP
easat-11054	45	27	securing	secure	VERB
easat-11054	45	28	complex	complex	ADJ
easat-11054	45	29	iiot	iiot	NOUN
easat-11054	45	30	networks	network	NOUN
easat-11054	45	31	from	from	ADP
easat-11054	45	32	modern	modern	ADJ
easat-11054	45	33	cyber	cyber	NOUN
easat-11054	45	34	intrusions	intrusion	NOUN
easat-11054	45	35	.	.	PUNCT
easat-11054	46	1	1.1	1.1	NUM
easat-11054	46	2	.	.	PUNCT
easat-11054	47	1	problem	problem	NOUN
easat-11054	47	2	statement	statement	NOUN
easat-11054	47	3	the	the	DET
easat-11054	47	4	iiot	iiot	PROPN
easat-11054	47	5	technologies	technology	NOUN
easat-11054	47	6	have	have	AUX
easat-11054	47	7	been	be	AUX
easat-11054	47	8	developed	develop	VERB
easat-11054	47	9	to	to	PART
easat-11054	47	10	enable	enable	VERB
easat-11054	47	11	seamless	seamless	ADJ
easat-11054	47	12	connectivity	connectivity	NOUN
easat-11054	47	13	and	and	CCONJ
easat-11054	47	14	automation	automation	NOUN
easat-11054	47	15	in	in	ADP
easat-11054	47	16	industries	industry	NOUN
easat-11054	47	17	.	.	PUNCT
easat-11054	48	1	however	however	ADV
easat-11054	48	2	,	,	PUNCT
easat-11054	48	3	this	this	DET
easat-11054	48	4	advancement	advancement	NOUN
easat-11054	48	5	has	have	AUX
easat-11054	48	6	also	also	ADV
easat-11054	48	7	introduced	introduce	VERB
easat-11054	48	8	modern	modern	ADJ
easat-11054	48	9	cyber	cyber	NOUN
easat-11054	48	10	threats	threat	NOUN
easat-11054	48	11	,	,	PUNCT
easat-11054	48	12	exposing	expose	VERB
easat-11054	48	13	iiot	iiot	NOUN
easat-11054	48	14	networks	network	NOUN
easat-11054	48	15	to	to	ADP
easat-11054	48	16	intrusions	intrusion	NOUN
easat-11054	48	17	that	that	PRON
easat-11054	48	18	are	be	AUX
easat-11054	48	19	most	most	ADV
easat-11054	48	20	relevant	relevant	ADJ
easat-11054	48	21	to	to	ADP
easat-11054	48	22	continuous	continuous	ADJ
easat-11054	48	23	features	feature	NOUN
easat-11054	48	24	like	like	ADP
easat-11054	48	25	distributed	distribute	VERB
easat-11054	48	26	denial	denial	NOUN
easat-11054	48	27	of	of	ADP
easat-11054	48	28	service	service	NOUN
easat-11054	48	29	(	(	PUNCT
easat-11054	48	30	ddos	ddos	NOUN
easat-11054	48	31	)	)	PUNCT
easat-11054	48	32	and	and	CCONJ
easat-11054	48	33	reconnaissance	reconnaissance	NOUN
easat-11054	49	1	[	[	X
easat-11054	49	2	13	13	NUM
easat-11054	49	3	]	]	PUNCT
easat-11054	49	4	as	as	ADV
easat-11054	49	5	well	well	ADV
easat-11054	49	6	as	as	ADP
easat-11054	49	7	intrusions	intrusion	NOUN
easat-11054	49	8	that	that	PRON
easat-11054	49	9	are	be	AUX
easat-11054	49	10	most	most	ADV
easat-11054	49	11	relevant	relevant	ADJ
easat-11054	49	12	to	to	PART
easat-11054	49	13	discrete	discrete	VERB
easat-11054	49	14	features	feature	NOUN
easat-11054	49	15	such	such	ADJ
easat-11054	49	16	as	as	ADP
easat-11054	49	17	data	datum	NOUN
easat-11054	49	18	theft	theft	NOUN
easat-11054	49	19	and	and	CCONJ
easat-11054	49	20	malware	malware	NOUN
easat-11054	49	21	exploits	exploit	NOUN
easat-11054	49	22	[	[	X
easat-11054	49	23	14	14	NUM
easat-11054	49	24	]	]	PUNCT
easat-11054	49	25	.	.	PUNCT
easat-11054	50	1	since	since	SCONJ
easat-11054	50	2	the	the	DET
easat-11054	50	3	a3c	a3c	PRON
easat-11054	50	4	can	can	AUX
easat-11054	50	5	be	be	AUX
easat-11054	50	6	used	use	VERB
easat-11054	50	7	with	with	ADP
easat-11054	50	8	discrete	discrete	ADJ
easat-11054	50	9	features	feature	NOUN
easat-11054	50	10	,	,	PUNCT
easat-11054	50	11	it	it	PRON
easat-11054	50	12	suffers	suffer	VERB
easat-11054	50	13	from	from	ADP
easat-11054	50	14	a	a	DET
easat-11054	50	15	lack	lack	NOUN
easat-11054	50	16	of	of	ADP
easat-11054	50	17	smoothness	smoothness	ADJ
easat-11054	50	18	and	and	CCONJ
easat-11054	50	19	poor	poor	ADJ
easat-11054	50	20	generalization	generalization	NOUN
easat-11054	50	21	because	because	SCONJ
easat-11054	50	22	this	this	DET
easat-11054	50	23	algorithm	algorithm	NOUN
easat-11054	50	24	provides	provide	VERB
easat-11054	50	25	smooth	smooth	ADJ
easat-11054	50	26	numeric	numeric	ADJ
easat-11054	50	27	input	input	NOUN
easat-11054	50	28	to	to	PART
easat-11054	50	29	stabilize	stabilize	VERB
easat-11054	50	30	learning	learn	VERB
easat-11054	50	31	.	.	PUNCT
easat-11054	51	1	on	on	ADP
easat-11054	51	2	the	the	DET
easat-11054	51	3	other	other	ADJ
easat-11054	51	4	hand	hand	NOUN
easat-11054	51	5	,	,	PUNCT
easat-11054	51	6	using	use	VERB
easat-11054	51	7	the	the	DET
easat-11054	51	8	dt	dt	PROPN
easat-11054	51	9	algorithm	algorithm	NOUN
easat-11054	51	10	is	be	AUX
easat-11054	51	11	more	more	ADV
easat-11054	51	12	suitable	suitable	ADJ
easat-11054	51	13	for	for	ADP
easat-11054	51	14	these	these	DET
easat-11054	51	15	intrusions	intrusion	NOUN
easat-11054	51	16	.	.	PUNCT
easat-11054	52	1	although	although	SCONJ
easat-11054	52	2	the	the	DET
easat-11054	52	3	dt	dt	NOUN
easat-11054	52	4	is	be	AUX
easat-11054	52	5	also	also	ADV
easat-11054	52	6	used	use	VERB
easat-11054	52	7	to	to	PART
easat-11054	52	8	deal	deal	VERB
easat-11054	52	9	with	with	ADP
easat-11054	52	10	continuous	continuous	ADJ
easat-11054	52	11	features	feature	NOUN
easat-11054	52	12	in	in	ADP
easat-11054	52	13	a	a	DET
easat-11054	52	14	dataset	dataset	NOUN
easat-11054	52	15	,	,	PUNCT
easat-11054	52	16	it	it	PRON
easat-11054	52	17	requires	require	VERB
easat-11054	52	18	specifying	specify	VERB
easat-11054	52	19	the	the	DET
easat-11054	52	20	threshold	threshold	NOUN
easat-11054	52	21	in	in	ADP
easat-11054	52	22	advance	advance	NOUN
easat-11054	52	23	to	to	PART
easat-11054	52	24	split	split	VERB
easat-11054	52	25	features	feature	NOUN
easat-11054	52	26	at	at	ADP
easat-11054	52	27	certain	certain	ADJ
easat-11054	52	28	values	value	NOUN
easat-11054	52	29	.	.	PUNCT
easat-11054	53	1	inaccurate	inaccurate	ADJ
easat-11054	53	2	threshold	threshold	NOUN
easat-11054	53	3	specification	specification	NOUN
easat-11054	53	4	reduces	reduce	VERB
easat-11054	53	5	attack	attack	NOUN
easat-11054	53	6	classification	classification	NOUN
easat-11054	53	7	performance	performance	NOUN
easat-11054	53	8	.	.	PUNCT
easat-11054	54	1	1.2	1.2	NUM
easat-11054	54	2	.	.	PUNCT
easat-11054	54	3	motivation	motivation	NOUN
easat-11054	54	4	this	this	DET
easat-11054	54	5	article	article	NOUN
easat-11054	54	6	focuses	focus	VERB
easat-11054	54	7	on	on	ADP
easat-11054	54	8	increasing	increase	VERB
easat-11054	54	9	the	the	DET
easat-11054	54	10	efficiency	efficiency	NOUN
easat-11054	54	11	of	of	ADP
easat-11054	54	12	intrusion	intrusion	NOUN
easat-11054	54	13	detection	detection	NOUN
easat-11054	54	14	in	in	ADP
easat-11054	54	15	the	the	DET
easat-11054	54	16	iiot	iiot	NOUN
easat-11054	54	17	network	network	NOUN
easat-11054	54	18	against	against	ADP
easat-11054	54	19	hackers	hacker	NOUN
easat-11054	54	20	,	,	PUNCT
easat-11054	54	21	as	as	SCONJ
easat-11054	54	22	traditional	traditional	ADJ
easat-11054	54	23	idss	idss	NOUN
easat-11054	54	24	can	can	AUX
easat-11054	54	25	not	not	PART
easat-11054	54	26	handle	handle	VERB
easat-11054	54	27	these	these	DET
easat-11054	54	28	threats	threat	NOUN
easat-11054	54	29	effectively	effectively	ADV
easat-11054	54	30	due	due	ADP
easat-11054	54	31	to	to	ADP
easat-11054	54	32	the	the	DET
easat-11054	54	33	increasing	increase	VERB
easat-11054	54	34	complexity	complexity	NOUN
easat-11054	54	35	of	of	ADP
easat-11054	54	36	the	the	DET
easat-11054	54	37	expansion	expansion	NOUN
easat-11054	54	38	and	and	CCONJ
easat-11054	54	39	the	the	DET
easat-11054	54	40	architecture	architecture	NOUN
easat-11054	54	41	of	of	ADP
easat-11054	54	42	iiot	iiot	NOUN
easat-11054	54	43	networks	network	NOUN
easat-11054	54	44	in	in	ADP
easat-11054	54	45	industrial	industrial	ADJ
easat-11054	54	46	fields	field	NOUN
easat-11054	54	47	.	.	PUNCT
easat-11054	55	1	hence	hence	ADV
easat-11054	55	2	,	,	PUNCT
easat-11054	55	3	the	the	DET
easat-11054	55	4	intrusion	intrusion	NOUN
easat-11054	55	5	detection	detection	NOUN
easat-11054	55	6	mechanism	mechanism	NOUN
easat-11054	55	7	must	must	AUX
easat-11054	55	8	be	be	AUX
easat-11054	55	9	robust	robust	ADJ
easat-11054	55	10	and	and	CCONJ
easat-11054	55	11	adaptable	adaptable	ADJ
easat-11054	55	12	.	.	PUNCT
easat-11054	56	1	therefore	therefore	ADV
easat-11054	56	2	,	,	PUNCT
easat-11054	56	3	the	the	DET
easat-11054	56	4	use	use	NOUN
easat-11054	56	5	of	of	ADP
easat-11054	56	6	hadt	hadt	NOUN
easat-11054	56	7	can	can	AUX
easat-11054	56	8	overcome	overcome	VERB
easat-11054	56	9	the	the	DET
easat-11054	56	10	shortcomings	shortcoming	NOUN
easat-11054	56	11	of	of	ADP
easat-11054	56	12	traditional	traditional	ADJ
easat-11054	56	13	idss	idss	NOUN
easat-11054	56	14	through	through	ADP
easat-11054	56	15	policy	policy	NOUN
easat-11054	56	16	-	-	PUNCT
easat-11054	56	17	based	base	VERB
easat-11054	56	18	learning	learning	NOUN
easat-11054	56	19	and	and	CCONJ
easat-11054	56	20	improved	improve	VERB
easat-11054	56	21	decision	decision	NOUN
easat-11054	56	22	rules	rule	NOUN
easat-11054	56	23	.	.	PUNCT
easat-11054	57	1	by	by	ADP
easat-11054	57	2	utilizing	utilize	VERB
easat-11054	57	3	the	the	DET
easat-11054	57	4	efficient	efficient	ADJ
easat-11054	57	5	hybrid	hybrid	ADJ
easat-11054	57	6	algorithm	algorithm	NOUN
easat-11054	57	7	and	and	CCONJ
easat-11054	57	8	the	the	DET
easat-11054	57	9	x	x	ADJ
easat-11054	57	10	-	-	ADJ
easat-11054	57	11	iiotid	iiotid	ADJ
easat-11054	57	12	dataset	dataset	NOUN
easat-11054	57	13	,	,	PUNCT
easat-11054	57	14	this	this	DET
easat-11054	57	15	article	article	NOUN
easat-11054	57	16	seeks	seek	VERB
easat-11054	57	17	to	to	PART
easat-11054	57	18	create	create	VERB
easat-11054	57	19	a	a	DET
easat-11054	57	20	high	high	ADJ
easat-11054	57	21	-	-	PUNCT
easat-11054	57	22	quality	quality	NOUN
easat-11054	57	23	ids	id	NOUN
easat-11054	57	24	that	that	PRON
easat-11054	57	25	can	can	AUX
easat-11054	57	26	enhance	enhance	VERB
easat-11054	57	27	intrusion	intrusion	NOUN
easat-11054	57	28	detection	detection	NOUN
easat-11054	57	29	accuracy	accuracy	NOUN
easat-11054	57	30	,	,	PUNCT
easat-11054	57	31	precision	precision	NOUN
easat-11054	57	32	,	,	PUNCT
easat-11054	57	33	f1	f1	NOUN
easat-11054	57	34	-	-	PUNCT
easat-11054	57	35	score	score	NOUN
easat-11054	57	36	,	,	PUNCT
easat-11054	57	37	and	and	CCONJ
easat-11054	57	38	recall	recall	VERB
easat-11054	57	39	in	in	ADP
easat-11054	57	40	iiot	iiot	ADJ
easat-11054	57	41	environments	environment	NOUN
easat-11054	57	42	.	.	PUNCT
easat-11054	58	1	1.3	1.3	NUM
easat-11054	58	2	.	.	PUNCT
easat-11054	58	3	contribution	contribution	NOUN
easat-11054	58	4	this	this	DET
easat-11054	58	5	article	article	NOUN
easat-11054	58	6	proposes	propose	VERB
easat-11054	58	7	the	the	DET
easat-11054	58	8	hadt	hadt	NOUN
easat-11054	58	9	algorithm	algorithm	NOUN
easat-11054	58	10	with	with	ADP
easat-11054	58	11	fewer	few	ADJ
easat-11054	58	12	neural	neural	ADJ
easat-11054	58	13	network	network	NOUN
easat-11054	58	14	layers	layer	NOUN
easat-11054	58	15	,	,	PUNCT
easat-11054	58	16	which	which	PRON
easat-11054	58	17	integrates	integrate	VERB
easat-11054	58	18	a	a	DET
easat-11054	58	19	simple	simple	NOUN
easat-11054	58	20	a3c	a3c	VERB
easat-11054	58	21	with	with	ADP
easat-11054	58	22	decision	decision	NOUN
easat-11054	58	23	trees	tree	NOUN
easat-11054	58	24	.	.	PUNCT
easat-11054	59	1	this	this	DET
easat-11054	59	2	integration	integration	NOUN
easat-11054	59	3	addresses	address	VERB
easat-11054	59	4	the	the	DET
easat-11054	59	5	shortcomings	shortcoming	NOUN
easat-11054	59	6	of	of	ADP
easat-11054	59	7	traditional	traditional	ADJ
easat-11054	59	8	a3c	a3c	VERB
easat-11054	59	9	for	for	ADP
easat-11054	59	10	intrusion	intrusion	NOUN
easat-11054	59	11	detection	detection	NOUN
easat-11054	59	12	systems	system	NOUN
easat-11054	59	13	(	(	PUNCT
easat-11054	59	14	idss	idss	ADV
easat-11054	59	15	)	)	PUNCT
easat-11054	59	16	.	.	PUNCT
easat-11054	60	1	the	the	DET
easat-11054	60	2	article	article	NOUN
easat-11054	60	3	combines	combine	VERB
easat-11054	60	4	different	different	ADJ
easat-11054	60	5	complementary	complementary	ADJ
easat-11054	60	6	machine	machine	NOUN
easat-11054	60	7	learning	learn	VERB
easat-11054	60	8	techniques	technique	NOUN
easat-11054	60	9	and	and	CCONJ
easat-11054	60	10	employs	employ	VERB
easat-11054	60	11	a	a	DET
easat-11054	60	12	contemporary	contemporary	ADJ
easat-11054	60	13	dataset	dataset	NOUN
easat-11054	60	14	to	to	PART
easat-11054	60	15	develop	develop	VERB
easat-11054	60	16	specialized	specialized	ADJ
easat-11054	60	17	approaches	approach	NOUN
easat-11054	60	18	for	for	ADP
easat-11054	60	19	industrial	industrial	ADJ
easat-11054	60	20	internet	internet	NOUN
easat-11054	60	21	of	of	ADP
easat-11054	60	22	things	thing	NOUN
easat-11054	60	23	(	(	PUNCT
easat-11054	60	24	iiot	iiot	NOUN
easat-11054	60	25	)	)	PUNCT
easat-11054	60	26	networks	network	NOUN
easat-11054	60	27	.	.	PUNCT
easat-11054	61	1	the	the	DET
easat-11054	61	2	main	main	ADJ
easat-11054	61	3	highlights	highlight	NOUN
easat-11054	61	4	of	of	ADP
easat-11054	61	5	this	this	DET
easat-11054	61	6	article	article	NOUN
easat-11054	61	7	are	be	AUX
easat-11054	61	8	:	:	PUNCT
easat-11054	61	9	•	•	ADV
easat-11054	61	10	using	use	VERB
easat-11054	61	11	the	the	DET
easat-11054	61	12	x	x	ADJ
easat-11054	61	13	-	-	ADJ
easat-11054	61	14	iiotid	iiotid	ADJ
easat-11054	61	15	dataset	dataset	NOUN
easat-11054	61	16	to	to	PART
easat-11054	61	17	address	address	VERB
easat-11054	61	18	real	real	ADJ
easat-11054	61	19	-	-	PUNCT
easat-11054	61	20	world	world	NOUN
easat-11054	61	21	iiot	iiot	NOUN
easat-11054	61	22	environments	environment	NOUN
easat-11054	61	23	,	,	PUNCT
easat-11054	61	24	which	which	PRON
easat-11054	61	25	contains	contain	VERB
easat-11054	61	26	two	two	NUM
easat-11054	61	27	types	type	NOUN
easat-11054	61	28	of	of	ADP
easat-11054	61	29	features	feature	NOUN
easat-11054	61	30	continuous	continuous	ADJ
easat-11054	61	31	and	and	CCONJ
easat-11054	61	32	discrete	discrete	ADJ
easat-11054	61	33	representing	represent	VERB
easat-11054	61	34	a	a	DET
easat-11054	61	35	general	general	ADJ
easat-11054	61	36	and	and	CCONJ
easat-11054	61	37	realistic	realistic	ADJ
easat-11054	61	38	view	view	NOUN
easat-11054	61	39	of	of	ADP
easat-11054	61	40	data	datum	NOUN
easat-11054	61	41	traffic	traffic	NOUN
easat-11054	61	42	in	in	ADP
easat-11054	61	43	the	the	DET
easat-11054	61	44	iiot	iiot	NOUN
easat-11054	61	45	network	network	NOUN
easat-11054	61	46	.	.	PUNCT
easat-11054	62	1	1055	1055	NUM
easat-11054	62	2	edelweiss	edelweiss	PROPN
easat-11054	62	3	applied	apply	VERB
easat-11054	62	4	science	science	NOUN
easat-11054	62	5	and	and	CCONJ
easat-11054	62	6	technology	technology	NOUN
easat-11054	62	7	issn	issn	PROPN
easat-11054	62	8	:	:	PUNCT
easat-11054	62	9	2576	2576	NUM
easat-11054	62	10	-	-	SYM
easat-11054	62	11	8484	8484	NUM
easat-11054	62	12	vol	vol	NOUN
easat-11054	62	13	.	.	PROPN
easat-11054	63	1	9	9	NUM
easat-11054	63	2	,	,	PUNCT
easat-11054	63	3	no	no	INTJ
easat-11054	63	4	.	.	NOUN
easat-11054	63	5	11	11	NUM
easat-11054	63	6	:	:	PUNCT
easat-11054	63	7	1053	1053	NUM
easat-11054	63	8	-	-	SYM
easat-11054	63	9	1075	1075	NUM
easat-11054	63	10	,	,	PUNCT
easat-11054	63	11	2025	2025	NUM
easat-11054	63	12	doi	doi	NOUN
easat-11054	63	13	:	:	PUNCT
easat-11054	63	14	10.55214/2576	10.55214/2576	NUM
easat-11054	63	15	-	-	SYM
easat-11054	63	16	8484.v9i11.11054	8484.v9i11.11054	NOUN
easat-11054	64	1	©	©	PROPN
easat-11054	64	2	2025	2025	NUM
easat-11054	64	3	by	by	ADP
easat-11054	64	4	the	the	DET
easat-11054	64	5	authors	author	NOUN
easat-11054	64	6	;	;	PUNCT
easat-11054	64	7	licensee	licensee	PROPN
easat-11054	64	8	learning	learning	NOUN
easat-11054	64	9	gate	gate	VERB
easat-11054	64	10	•	•	NOUN
easat-11054	64	11	hybridizing	hybridize	VERB
easat-11054	64	12	simple	simple	NOUN
easat-11054	64	13	a3c	a3c	VERB
easat-11054	64	14	with	with	ADP
easat-11054	64	15	dt	dt	NOUN
easat-11054	64	16	using	use	VERB
easat-11054	64	17	fewer	few	ADJ
easat-11054	64	18	nn	nn	PROPN
easat-11054	64	19	layers	layer	NOUN
easat-11054	64	20	helps	help	VERB
easat-11054	64	21	reduce	reduce	VERB
easat-11054	64	22	processing	processing	NOUN
easat-11054	64	23	operations	operation	NOUN
easat-11054	64	24	,	,	PUNCT
easat-11054	64	25	thereby	thereby	ADV
easat-11054	64	26	improving	improve	VERB
easat-11054	64	27	processing	processing	NOUN
easat-11054	64	28	time	time	NOUN
easat-11054	64	29	and	and	CCONJ
easat-11054	64	30	power	power	NOUN
easat-11054	64	31	consumption	consumption	NOUN
easat-11054	64	32	.	.	PUNCT
easat-11054	65	1	this	this	DET
easat-11054	65	2	approach	approach	NOUN
easat-11054	65	3	integrates	integrate	VERB
easat-11054	65	4	the	the	DET
easat-11054	65	5	advantages	advantage	NOUN
easat-11054	65	6	of	of	ADP
easat-11054	65	7	a3c	a3c	VERB
easat-11054	65	8	with	with	ADP
easat-11054	65	9	dt	dt	NOUN
easat-11054	65	10	into	into	ADP
easat-11054	65	11	another	another	DET
easat-11054	65	12	ffnn	ffnn	NOUN
easat-11054	65	13	layer	layer	NOUN
easat-11054	65	14	,	,	PUNCT
easat-11054	65	15	which	which	PRON
easat-11054	65	16	helps	help	VERB
easat-11054	65	17	to	to	PART
easat-11054	65	18	get	get	VERB
easat-11054	65	19	the	the	DET
easat-11054	65	20	advantages	advantage	NOUN
easat-11054	65	21	of	of	ADP
easat-11054	65	22	both	both	DET
easat-11054	65	23	algorithms	algorithm	NOUN
easat-11054	65	24	in	in	ADP
easat-11054	65	25	detecting	detect	VERB
easat-11054	65	26	intrusions	intrusion	NOUN
easat-11054	65	27	.	.	PUNCT
easat-11054	66	1	•	•	NOUN
easat-11054	66	2	combining	combine	VERB
easat-11054	66	3	improved	improved	ADJ
easat-11054	66	4	policy	policy	NOUN
easat-11054	66	5	learning	learning	NOUN
easat-11054	66	6	with	with	ADP
easat-11054	66	7	decision	decision	NOUN
easat-11054	66	8	rules	rule	NOUN
easat-11054	66	9	significantly	significantly	ADV
easat-11054	66	10	increases	increase	VERB
easat-11054	66	11	the	the	DET
easat-11054	66	12	efficiency	efficiency	NOUN
easat-11054	66	13	of	of	ADP
easat-11054	66	14	intrusion	intrusion	NOUN
easat-11054	66	15	detection	detection	NOUN
easat-11054	66	16	under	under	ADP
easat-11054	66	17	different	different	ADJ
easat-11054	66	18	intrusions	intrusion	NOUN
easat-11054	66	19	.	.	PUNCT
easat-11054	67	1	this	this	PRON
easat-11054	67	2	is	be	AUX
easat-11054	67	3	shown	show	VERB
easat-11054	67	4	through	through	ADP
easat-11054	67	5	iterative	iterative	NOUN
easat-11054	67	6	testing	testing	NOUN
easat-11054	67	7	and	and	CCONJ
easat-11054	67	8	evaluation	evaluation	NOUN
easat-11054	67	9	of	of	ADP
easat-11054	67	10	the	the	DET
easat-11054	67	11	proposed	propose	VERB
easat-11054	67	12	algorithm	algorithm	NOUN
easat-11054	67	13	using	use	VERB
easat-11054	67	14	accuracy	accuracy	NOUN
easat-11054	67	15	,	,	PUNCT
easat-11054	67	16	precision	precision	NOUN
easat-11054	67	17	,	,	PUNCT
easat-11054	67	18	recall	recall	NOUN
easat-11054	67	19	,	,	PUNCT
easat-11054	67	20	and	and	CCONJ
easat-11054	67	21	f1	f1	ADJ
easat-11054	67	22	-	-	PUNCT
easat-11054	67	23	score	score	NOUN
easat-11054	67	24	metrics	metric	NOUN
easat-11054	67	25	,	,	PUNCT
easat-11054	67	26	and	and	CCONJ
easat-11054	67	27	by	by	ADP
easat-11054	67	28	comparing	compare	VERB
easat-11054	67	29	the	the	DET
easat-11054	67	30	results	result	NOUN
easat-11054	67	31	with	with	ADP
easat-11054	67	32	those	those	PRON
easat-11054	67	33	of	of	ADP
easat-11054	67	34	existing	exist	VERB
easat-11054	67	35	a3c	a3c	NUM
easat-11054	67	36	algorithms	algorithm	NOUN
easat-11054	67	37	.	.	PUNCT
easat-11054	68	1	through	through	ADP
easat-11054	68	2	these	these	DET
easat-11054	68	3	contributions	contribution	NOUN
easat-11054	68	4	,	,	PUNCT
easat-11054	68	5	the	the	DET
easat-11054	68	6	proposed	propose	VERB
easat-11054	68	7	algorithm	algorithm	NOUN
easat-11054	68	8	ensures	ensure	VERB
easat-11054	68	9	effective	effective	ADJ
easat-11054	68	10	identification	identification	NOUN
easat-11054	68	11	and	and	CCONJ
easat-11054	68	12	detection	detection	NOUN
easat-11054	68	13	of	of	ADP
easat-11054	68	14	iiot	iiot	NOUN
easat-11054	68	15	intrusions	intrusion	NOUN
easat-11054	68	16	.	.	PUNCT
easat-11054	69	1	2	2	X
easat-11054	69	2	.	.	X
easat-11054	69	3	related	relate	VERB
easat-11054	69	4	work	work	NOUN
easat-11054	69	5	with	with	ADP
easat-11054	69	6	the	the	DET
easat-11054	69	7	rapid	rapid	ADJ
easat-11054	69	8	development	development	NOUN
easat-11054	69	9	of	of	ADP
easat-11054	69	10	iiot	iiot	NOUN
easat-11054	69	11	networks	network	NOUN
easat-11054	69	12	,	,	PUNCT
easat-11054	69	13	actor	actor	NOUN
easat-11054	69	14	-	-	PUNCT
easat-11054	69	15	critic	critic	NOUN
easat-11054	69	16	algorithms	algorithm	NOUN
easat-11054	69	17	have	have	AUX
easat-11054	69	18	been	be	AUX
easat-11054	69	19	developed	develop	VERB
easat-11054	69	20	to	to	PART
easat-11054	69	21	support	support	VERB
easat-11054	69	22	robust	robust	ADJ
easat-11054	69	23	idss	idss	NOUN
easat-11054	69	24	,	,	PUNCT
easat-11054	69	25	especially	especially	ADV
easat-11054	69	26	amid	amid	ADP
easat-11054	69	27	the	the	DET
easat-11054	69	28	emergence	emergence	NOUN
easat-11054	69	29	of	of	ADP
easat-11054	69	30	new	new	ADJ
easat-11054	69	31	intrusion	intrusion	NOUN
easat-11054	69	32	patterns	pattern	NOUN
easat-11054	69	33	that	that	PRON
easat-11054	69	34	are	be	AUX
easat-11054	69	35	increasingly	increasingly	ADV
easat-11054	69	36	difficult	difficult	ADJ
easat-11054	69	37	to	to	PART
easat-11054	69	38	detect	detect	VERB
easat-11054	69	39	.	.	PUNCT
easat-11054	70	1	being	be	AUX
easat-11054	70	2	one	one	NUM
easat-11054	70	3	of	of	ADP
easat-11054	70	4	the	the	DET
easat-11054	70	5	actor	actor	NOUN
easat-11054	70	6	-	-	PUNCT
easat-11054	70	7	critic	critic	NOUN
easat-11054	70	8	algorithms	algorithm	NOUN
easat-11054	70	9	,	,	PUNCT
easat-11054	70	10	a3c	a3c	VERB
easat-11054	70	11	contributes	contribute	VERB
easat-11054	70	12	significantly	significantly	ADV
easat-11054	70	13	to	to	ADP
easat-11054	70	14	current	current	ADJ
easat-11054	70	15	intrusion	intrusion	NOUN
easat-11054	70	16	detection	detection	NOUN
easat-11054	70	17	.	.	PUNCT
easat-11054	71	1	however	however	ADV
easat-11054	71	2	,	,	PUNCT
easat-11054	71	3	its	its	PRON
easat-11054	71	4	traditional	traditional	ADJ
easat-11054	71	5	forms	form	NOUN
easat-11054	71	6	have	have	AUX
easat-11054	71	7	become	become	VERB
easat-11054	71	8	ineffective	ineffective	ADJ
easat-11054	71	9	due	due	ADP
easat-11054	71	10	to	to	ADP
easat-11054	71	11	the	the	DET
easat-11054	71	12	complexity	complexity	NOUN
easat-11054	71	13	of	of	ADP
easat-11054	71	14	iiot	iiot	ADJ
easat-11054	71	15	environments	environment	NOUN
easat-11054	71	16	that	that	PRON
easat-11054	71	17	incorporate	incorporate	VERB
easat-11054	71	18	both	both	CCONJ
easat-11054	71	19	stochastic	stochastic	ADJ
easat-11054	71	20	and	and	CCONJ
easat-11054	71	21	deterministic	deterministic	ADJ
easat-11054	71	22	intrusion	intrusion	NOUN
easat-11054	71	23	patterns	pattern	NOUN
easat-11054	71	24	in	in	ADP
easat-11054	71	25	the	the	DET
easat-11054	71	26	iiot	iiot	NOUN
easat-11054	71	27	network	network	NOUN
easat-11054	71	28	[	[	X
easat-11054	71	29	15	15	NUM
easat-11054	71	30	]	]	PUNCT
easat-11054	71	31	.	.	PUNCT
easat-11054	72	1	a	a	DET
easat-11054	72	2	study	study	NOUN
easat-11054	72	3	by	by	ADP
easat-11054	72	4	muhati	muhati	PROPN
easat-11054	72	5	and	and	CCONJ
easat-11054	72	6	rawat	rawat	PROPN
easat-11054	72	7	[	[	X
easat-11054	72	8	16	16	NUM
easat-11054	72	9	]	]	PUNCT
easat-11054	72	10	suggested	suggest	VERB
easat-11054	72	11	an	an	DET
easat-11054	72	12	a3c	a3c	ADV
easat-11054	72	13	-	-	PUNCT
easat-11054	72	14	based	base	VERB
easat-11054	72	15	cognitive	cognitive	ADJ
easat-11054	72	16	network	network	NOUN
easat-11054	72	17	intrusion	intrusion	NOUN
easat-11054	72	18	detection	detection	NOUN
easat-11054	72	19	system	system	NOUN
easat-11054	72	20	(	(	PUNCT
easat-11054	72	21	nids	nid	NOUN
easat-11054	72	22	)	)	PUNCT
easat-11054	72	23	using	use	VERB
easat-11054	72	24	deep	deep	ADJ
easat-11054	72	25	reinforcement	reinforcement	NOUN
easat-11054	72	26	learning	learning	NOUN
easat-11054	72	27	(	(	PUNCT
easat-11054	72	28	drl	drl	PROPN
easat-11054	72	29	)	)	PUNCT
easat-11054	72	30	to	to	PART
easat-11054	72	31	adaptively	adaptively	ADV
easat-11054	72	32	analyze	analyze	VERB
easat-11054	72	33	network	network	NOUN
easat-11054	72	34	traffic	traffic	NOUN
easat-11054	72	35	.	.	PUNCT
easat-11054	73	1	utilizing	utilize	VERB
easat-11054	73	2	open	open	ADJ
easat-11054	73	3	-	-	PUNCT
easat-11054	73	4	source	source	NOUN
easat-11054	73	5	service	service	NOUN
easat-11054	73	6	discovery	discovery	NOUN
easat-11054	73	7	tools	tool	NOUN
easat-11054	73	8	,	,	PUNCT
easat-11054	73	9	it	it	PRON
easat-11054	73	10	has	have	AUX
easat-11054	73	11	improved	improve	VERB
easat-11054	73	12	intrusion	intrusion	NOUN
easat-11054	73	13	detection	detection	NOUN
easat-11054	73	14	in	in	ADP
easat-11054	73	15	network	network	NOUN
easat-11054	73	16	attacks	attack	NOUN
easat-11054	73	17	,	,	PUNCT
easat-11054	73	18	achieving	achieve	VERB
easat-11054	73	19	high	high	ADJ
easat-11054	73	20	accuracy	accuracy	NOUN
easat-11054	73	21	on	on	ADP
easat-11054	73	22	various	various	ADJ
easat-11054	73	23	datasets	dataset	NOUN
easat-11054	73	24	,	,	PUNCT
easat-11054	73	25	including	include	VERB
easat-11054	73	26	unsw	unsw	NOUN
easat-11054	73	27	-	-	PUNCT
easat-11054	73	28	nb15	nb15	PROPN
easat-11054	73	29	,	,	PUNCT
easat-11054	73	30	awid	awid	NOUN
easat-11054	73	31	,	,	PUNCT
easat-11054	73	32	and	and	CCONJ
easat-11054	73	33	nsl	nsl	PROPN
easat-11054	73	34	-	-	PUNCT
easat-11054	73	35	kdd	kdd	PROPN
easat-11054	73	36	.	.	PROPN
easat-11054	73	37	similarly	similarly	ADV
easat-11054	73	38	,	,	PUNCT
easat-11054	73	39	zhou	zhou	PROPN
easat-11054	73	40	et	et	PROPN
easat-11054	73	41	al	al	PROPN
easat-11054	73	42	.	.	PUNCT
easat-11054	74	1	[	[	X
easat-11054	74	2	17	17	NUM
easat-11054	74	3	]	]	PUNCT
easat-11054	74	4	examined	examine	VERB
easat-11054	74	5	an	an	DET
easat-11054	74	6	improved	improved	ADJ
easat-11054	74	7	a3c	a3c	NUM
easat-11054	74	8	algorithm	algorithm	NOUN
easat-11054	74	9	for	for	ADP
easat-11054	74	10	intrusion	intrusion	NOUN
easat-11054	74	11	detection	detection	NOUN
easat-11054	74	12	,	,	PUNCT
easat-11054	74	13	incorporating	incorporate	VERB
easat-11054	74	14	attention	attention	NOUN
easat-11054	74	15	mechanisms	mechanism	NOUN
easat-11054	74	16	and	and	CCONJ
easat-11054	74	17	cnn	cnn	PROPN
easat-11054	74	18	networks	network	NOUN
easat-11054	74	19	.	.	PUNCT
easat-11054	75	1	this	this	DET
easat-11054	75	2	improvement	improvement	NOUN
easat-11054	75	3	demonstrates	demonstrate	VERB
easat-11054	75	4	that	that	SCONJ
easat-11054	75	5	rl	rl	PROPN
easat-11054	75	6	surpasses	surpass	VERB
easat-11054	75	7	traditional	traditional	ADJ
easat-11054	75	8	machine	machine	NOUN
easat-11054	75	9	learning	learning	NOUN
easat-11054	75	10	methods	method	NOUN
easat-11054	75	11	in	in	ADP
easat-11054	75	12	intrusion	intrusion	NOUN
easat-11054	75	13	detection	detection	NOUN
easat-11054	75	14	,	,	PUNCT
easat-11054	75	15	with	with	ADP
easat-11054	75	16	improved	improved	ADJ
easat-11054	75	17	accuracy	accuracy	NOUN
easat-11054	75	18	and	and	CCONJ
easat-11054	75	19	recall	recall	NOUN
easat-11054	75	20	among	among	ADP
easat-11054	75	21	participants	participant	NOUN
easat-11054	75	22	as	as	ADV
easat-11054	75	23	well	well	ADV
easat-11054	75	24	as	as	ADP
easat-11054	75	25	f1	f1	ADJ
easat-11054	75	26	score	score	NOUN
easat-11054	75	27	for	for	ADP
easat-11054	75	28	high	high	ADJ
easat-11054	75	29	precision	precision	NOUN
easat-11054	75	30	across	across	ADP
easat-11054	75	31	public	public	ADJ
easat-11054	75	32	benchmark	benchmark	NOUN
easat-11054	75	33	datasets	dataset	NOUN
easat-11054	75	34	(	(	PUNCT
easat-11054	75	35	nsl	nsl	NOUN
easat-11054	75	36	-	-	PUNCT
easat-11054	75	37	kdd	kdd	NOUN
easat-11054	75	38	,	,	PUNCT
easat-11054	75	39	awid	awid	NOUN
easat-11054	75	40	cicids-2017	cicids-2017	ADV
easat-11054	75	41	,	,	PUNCT
easat-11054	75	42	and	and	CCONJ
easat-11054	75	43	dohbrw2020	dohbrw2020	NOUN
easat-11054	75	44	)	)	PUNCT
easat-11054	75	45	.	.	PUNCT
easat-11054	76	1	integrating	integrate	VERB
easat-11054	76	2	hybrid	hybrid	NOUN
easat-11054	76	3	models	model	NOUN
easat-11054	76	4	into	into	ADP
easat-11054	76	5	rl	rl	NOUN
easat-11054	76	6	-	-	PUNCT
easat-11054	76	7	based	base	VERB
easat-11054	76	8	idss	idss	ADV
easat-11054	76	9	enhances	enhance	VERB
easat-11054	76	10	the	the	DET
easat-11054	76	11	system	system	NOUN
easat-11054	76	12	's	's	PART
easat-11054	76	13	ability	ability	NOUN
easat-11054	76	14	to	to	PART
easat-11054	76	15	adapt	adapt	VERB
easat-11054	76	16	to	to	ADP
easat-11054	76	17	both	both	CCONJ
easat-11054	76	18	stochastic	stochastic	ADJ
easat-11054	76	19	and	and	CCONJ
easat-11054	76	20	deterministic	deterministic	ADJ
easat-11054	76	21	network	network	NOUN
easat-11054	76	22	environments	environment	NOUN
easat-11054	76	23	.	.	PUNCT
easat-11054	77	1	shen	shen	PROPN
easat-11054	77	2	et	et	PROPN
easat-11054	77	3	al	al	PROPN
easat-11054	77	4	.	.	PUNCT
easat-11054	78	1	[	[	X
easat-11054	78	2	18	18	NUM
easat-11054	78	3	]	]	PUNCT
easat-11054	78	4	suggested	suggest	VERB
easat-11054	78	5	a	a	DET
easat-11054	78	6	mean	mean	ADJ
easat-11054	78	7	-	-	PUNCT
easat-11054	78	8	field	field	NOUN
easat-11054	78	9	gamebased	gamebase	VERB
easat-11054	78	10	asynchronous	asynchronous	ADJ
easat-11054	78	11	advantage	advantage	NOUN
easat-11054	78	12	actor	actor	NOUN
easat-11054	78	13	-	-	PUNCT
easat-11054	78	14	critic	critic	NOUN
easat-11054	78	15	(	(	PUNCT
easat-11054	78	16	mfga3c	mfga3c	NUM
easat-11054	78	17	)	)	PUNCT
easat-11054	78	18	model	model	NOUN
easat-11054	78	19	to	to	PART
easat-11054	78	20	counter	counter	VERB
easat-11054	78	21	malware	malware	NOUN
easat-11054	78	22	propagation	propagation	NOUN
easat-11054	78	23	in	in	ADP
easat-11054	78	24	edge	edge	NOUN
easat-11054	78	25	intelligence	intelligence	NOUN
easat-11054	78	26	-	-	PUNCT
easat-11054	78	27	based	base	VERB
easat-11054	78	28	iot	iot	NOUN
easat-11054	78	29	(	(	PUNCT
easat-11054	78	30	eiot	eiot	NOUN
easat-11054	78	31	)	)	PUNCT
easat-11054	78	32	networks	network	NOUN
easat-11054	78	33	.	.	PUNCT
easat-11054	79	1	their	their	PRON
easat-11054	79	2	model	model	NOUN
easat-11054	79	3	improves	improve	VERB
easat-11054	79	4	multi	multi	ADJ
easat-11054	79	5	-	-	ADJ
easat-11054	79	6	worker	worker	ADJ
easat-11054	79	7	reinforcement	reinforcement	NOUN
easat-11054	79	8	learning	learning	NOUN
easat-11054	79	9	(	(	PUNCT
easat-11054	79	10	marl	marl	NOUN
easat-11054	79	11	)	)	PUNCT
easat-11054	79	12	to	to	PART
easat-11054	79	13	advance	advance	VERB
easat-11054	79	14	optimal	optimal	ADJ
easat-11054	79	15	defense	defense	NOUN
easat-11054	79	16	policies	policy	NOUN
easat-11054	79	17	against	against	ADP
easat-11054	79	18	adversarial	adversarial	ADJ
easat-11054	79	19	attacks	attack	NOUN
easat-11054	79	20	,	,	PUNCT
easat-11054	79	21	demonstrating	demonstrate	VERB
easat-11054	79	22	improved	improved	ADJ
easat-11054	79	23	resilience	resilience	NOUN
easat-11054	79	24	in	in	ADP
easat-11054	79	25	malware	malware	NOUN
easat-11054	79	26	mitigation	mitigation	NOUN
easat-11054	79	27	.	.	PUNCT
easat-11054	80	1	a	a	DET
easat-11054	80	2	dynamic	dynamic	ADJ
easat-11054	80	3	threshold	threshold	NOUN
easat-11054	80	4	mechanism	mechanism	NOUN
easat-11054	80	5	is	be	AUX
easat-11054	80	6	used	use	VERB
easat-11054	80	7	by	by	ADP
easat-11054	80	8	the	the	DET
easat-11054	80	9	bayesian	bayesian	NOUN
easat-11054	80	10	drl	drl	PROPN
easat-11054	80	11	framework	framework	NOUN
easat-11054	80	12	of	of	ADP
easat-11054	80	13	watts	watts	PROPN
easat-11054	80	14	et	et	PROPN
easat-11054	80	15	al	al	PROPN
easat-11054	80	16	.	.	PUNCT
easat-11054	81	1	[	[	X
easat-11054	81	2	9	9	NUM
easat-11054	81	3	]	]	PUNCT
easat-11054	81	4	,	,	PUNCT
easat-11054	81	5	who	who	PRON
easat-11054	81	6	developed	develop	VERB
easat-11054	81	7	specialized	specialized	ADJ
easat-11054	81	8	techniques	technique	NOUN
easat-11054	81	9	for	for	ADP
easat-11054	81	10	real	real	ADJ
easat-11054	81	11	-	-	PUNCT
easat-11054	81	12	time	time	NOUN
easat-11054	81	13	intrusion	intrusion	NOUN
easat-11054	81	14	classification	classification	NOUN
easat-11054	81	15	.	.	PUNCT
easat-11054	82	1	they	they	PRON
easat-11054	82	2	combine	combine	VERB
easat-11054	82	3	a	a	DET
easat-11054	82	4	cnn	cnn	NOUN
easat-11054	82	5	with	with	ADP
easat-11054	82	6	essentially	essentially	ADV
easat-11054	82	7	any	any	DET
easat-11054	82	8	partially	partially	ADV
easat-11054	82	9	observable	observable	ADJ
easat-11054	82	10	markov	markov	NOUN
easat-11054	82	11	decision	decision	NOUN
easat-11054	82	12	process	process	NOUN
easat-11054	82	13	(	(	PUNCT
easat-11054	82	14	pomdp	pomdp	NOUN
easat-11054	82	15	)	)	PUNCT
easat-11054	82	16	,	,	PUNCT
easat-11054	82	17	which	which	PRON
easat-11054	82	18	allows	allow	VERB
easat-11054	82	19	for	for	ADP
easat-11054	82	20	very	very	ADV
easat-11054	82	21	robust	robust	ADJ
easat-11054	82	22	intrusion	intrusion	NOUN
easat-11054	82	23	detection	detection	NOUN
easat-11054	82	24	in	in	ADP
easat-11054	82	25	automated	automate	VERB
easat-11054	82	26	systems	system	NOUN
easat-11054	82	27	.	.	PUNCT
easat-11054	83	1	to	to	PART
easat-11054	83	2	improve	improve	VERB
easat-11054	83	3	detection	detection	NOUN
easat-11054	83	4	accuracy	accuracy	NOUN
easat-11054	83	5	,	,	PUNCT
easat-11054	83	6	effective	effective	ADJ
easat-11054	83	7	ids	id	NOUN
easat-11054	83	8	frameworks	framework	NOUN
easat-11054	83	9	of	of	ADP
easat-11054	83	10	a3c	a3c	ADJ
easat-11054	83	11	algorithms	algorithm	NOUN
easat-11054	83	12	require	require	VERB
easat-11054	83	13	extensive	extensive	ADJ
easat-11054	83	14	datasets	dataset	NOUN
easat-11054	83	15	to	to	PART
easat-11054	83	16	train	train	VERB
easat-11054	83	17	and	and	CCONJ
easat-11054	83	18	validate	validate	VERB
easat-11054	83	19	their	their	PRON
easat-11054	83	20	algorithms	algorithm	NOUN
easat-11054	83	21	.	.	PUNCT
easat-11054	84	1	as	as	SCONJ
easat-11054	84	2	stated	state	VERB
easat-11054	84	3	by	by	ADP
easat-11054	84	4	dong	dong	PROPN
easat-11054	84	5	et	et	PROPN
easat-11054	84	6	al	al	PROPN
easat-11054	84	7	.	.	PUNCT
easat-11054	85	1	[	[	X
easat-11054	85	2	19	19	NUM
easat-11054	85	3	]	]	PUNCT
easat-11054	85	4	,	,	PUNCT
easat-11054	85	5	they	they	PRON
easat-11054	85	6	demonstrated	demonstrate	VERB
easat-11054	85	7	that	that	SCONJ
easat-11054	85	8	feature	feature	NOUN
easat-11054	85	9	selection	selection	NOUN
easat-11054	85	10	and	and	CCONJ
easat-11054	85	11	data	datum	NOUN
easat-11054	85	12	preprocessing	preprocessing	NOUN
easat-11054	85	13	are	be	AUX
easat-11054	85	14	crucial	crucial	ADJ
easat-11054	85	15	factors	factor	NOUN
easat-11054	85	16	for	for	ADP
easat-11054	85	17	systematically	systematically	ADV
easat-11054	85	18	refining	refine	VERB
easat-11054	85	19	high	high	ADJ
easat-11054	85	20	-	-	PUNCT
easat-11054	85	21	dimensional	dimensional	ADJ
easat-11054	85	22	network	network	NOUN
easat-11054	85	23	traffic	traffic	NOUN
easat-11054	85	24	data	datum	NOUN
easat-11054	85	25	in	in	ADP
easat-11054	85	26	the	the	DET
easat-11054	85	27	optimization	optimization	NOUN
easat-11054	85	28	of	of	ADP
easat-11054	85	29	deep	deep	ADJ
easat-11054	85	30	rl	rl	NOUN
easat-11054	85	31	-	-	PUNCT
easat-11054	85	32	based	base	VERB
easat-11054	85	33	idss	idss	NOUN
easat-11054	85	34	.	.	PUNCT
easat-11054	86	1	in	in	ADP
easat-11054	86	2	a	a	DET
easat-11054	86	3	previous	previous	ADJ
easat-11054	86	4	research	research	NOUN
easat-11054	86	5	article	article	NOUN
easat-11054	86	6	presented	present	VERB
easat-11054	86	7	by	by	ADP
easat-11054	86	8	the	the	DET
easat-11054	86	9	authors	author	NOUN
easat-11054	86	10	anbazhagan	anbazhagan	VERB
easat-11054	86	11	and	and	CCONJ
easat-11054	86	12	mugelan	mugelan	VERB
easat-11054	86	13	[	[	X
easat-11054	86	14	20	20	NUM
easat-11054	86	15	]	]	PUNCT
easat-11054	86	16	,	,	PUNCT
easat-11054	86	17	the	the	DET
easat-11054	86	18	enhanced	enhanced	ADJ
easat-11054	86	19	a3c	a3c	NUM
easat-11054	86	20	(	(	PUNCT
easat-11054	86	21	ea3c	ea3c	NOUN
easat-11054	86	22	)	)	PUNCT
easat-11054	86	23	algorithm	algorithm	NOUN
easat-11054	86	24	was	be	AUX
easat-11054	86	25	compared	compare	VERB
easat-11054	86	26	with	with	ADP
easat-11054	86	27	several	several	ADJ
easat-11054	86	28	recent	recent	ADJ
easat-11054	86	29	algorithms	algorithm	NOUN
easat-11054	86	30	such	such	ADJ
easat-11054	86	31	as	as	ADP
easat-11054	86	32	dt	dt	PROPN
easat-11054	86	33	,	,	PUNCT
easat-11054	86	34	attentionenhanced	attentionenhanced	ADJ
easat-11054	86	35	reinforcement	reinforcement	NOUN
easat-11054	86	36	learning	learning	NOUN
easat-11054	86	37	(	(	PUNCT
easat-11054	86	38	aerl	aerl	NOUN
easat-11054	86	39	)	)	PUNCT
easat-11054	86	40	,	,	PUNCT
easat-11054	86	41	double	double	VERB
easat-11054	86	42	deep	deep	ADJ
easat-11054	86	43	q	q	NOUN
easat-11054	86	44	-	-	PUNCT
easat-11054	86	45	network	network	NOUN
easat-11054	86	46	(	(	PUNCT
easat-11054	86	47	ddqn	ddqn	NOUN
easat-11054	86	48	)	)	PUNCT
easat-11054	86	49	,	,	PUNCT
easat-11054	86	50	and	and	CCONJ
easat-11054	86	51	traditional	traditional	ADJ
easat-11054	86	52	a3c	a3c	NUM
easat-11054	86	53	.	.	PUNCT
easat-11054	87	1	the	the	DET
easat-11054	87	2	experiments	experiment	NOUN
easat-11054	87	3	indicated	indicate	VERB
easat-11054	87	4	that	that	SCONJ
easat-11054	87	5	the	the	DET
easat-11054	87	6	ea3c	ea3c	PROPN
easat-11054	87	7	algorithm	algorithm	NOUN
easat-11054	87	8	outperformed	outperform	VERB
easat-11054	87	9	these	these	DET
easat-11054	87	10	algorithms	algorithm	NOUN
easat-11054	87	11	by	by	ADP
easat-11054	87	12	1.3–1.7	1.3–1.7	NUM
easat-11054	87	13	%	%	NOUN
easat-11054	87	14	for	for	ADP
easat-11054	87	15	accuracy	accuracy	NOUN
easat-11054	87	16	,	,	PUNCT
easat-11054	87	17	precision	precision	NOUN
easat-11054	87	18	,	,	PUNCT
easat-11054	87	19	recall	recall	NOUN
easat-11054	87	20	,	,	PUNCT
easat-11054	87	21	and	and	CCONJ
easat-11054	87	22	f1	f1	NOUN
easat-11054	87	23	-	-	PUNCT
easat-11054	87	24	score	score	NOUN
easat-11054	87	25	,	,	PUNCT
easat-11054	87	26	particularly	particularly	ADV
easat-11054	87	27	on	on	ADP
easat-11054	87	28	handling	handle	VERB
easat-11054	87	29	low	low	ADJ
easat-11054	87	30	-	-	PUNCT
easat-11054	87	31	sample	sample	NOUN
easat-11054	87	32	-	-	PUNCT
easat-11054	87	33	size	size	NOUN
easat-11054	87	34	attacks	attack	NOUN
easat-11054	87	35	and	and	CCONJ
easat-11054	87	36	mixed	mixed	ADJ
easat-11054	87	37	feature	feature	NOUN
easat-11054	87	38	-	-	PUNCT
easat-11054	87	39	type	type	NOUN
easat-11054	87	40	environments	environment	NOUN
easat-11054	87	41	,	,	PUNCT
easat-11054	87	42	as	as	SCONJ
easat-11054	87	43	empirically	empirically	ADV
easat-11054	87	44	verified	verify	VERB
easat-11054	87	45	on	on	ADP
easat-11054	87	46	the	the	DET
easat-11054	87	47	x	x	ADJ
easat-11054	87	48	-	-	ADJ
easat-11054	87	49	iiotid	iiotid	ADJ
easat-11054	87	50	dataset	dataset	NOUN
easat-11054	87	51	.	.	PUNCT
easat-11054	88	1	all	all	PRON
easat-11054	88	2	of	of	ADP
easat-11054	88	3	these	these	DET
easat-11054	88	4	improvements	improvement	NOUN
easat-11054	88	5	are	be	AUX
easat-11054	88	6	due	due	ADJ
easat-11054	88	7	to	to	ADP
easat-11054	88	8	the	the	DET
easat-11054	88	9	improved	improved	ADJ
easat-11054	88	10	cnn	cnn	NOUN
easat-11054	88	11	architecture	architecture	NOUN
easat-11054	88	12	of	of	ADP
easat-11054	88	13	the	the	DET
easat-11054	88	14	ea3c	ea3c	PROPN
easat-11054	88	15	worker	worker	NOUN
easat-11054	88	16	,	,	PUNCT
easat-11054	88	17	the	the	DET
easat-11054	88	18	training	training	NOUN
easat-11054	88	19	stability	stability	NOUN
easat-11054	88	20	achieved	achieve	VERB
easat-11054	88	21	by	by	ADP
easat-11054	88	22	gradient	gradient	NOUN
easat-11054	88	23	clipping	clipping	NOUN
easat-11054	88	24	,	,	PUNCT
easat-11054	88	25	and	and	CCONJ
easat-11054	88	26	the	the	DET
easat-11054	88	27	improved	improve	VERB
easat-11054	88	28	parameter	parameter	NOUN
easat-11054	88	29	synchronization	synchronization	NOUN
easat-11054	88	30	mechanism	mechanism	NOUN
easat-11054	88	31	.	.	PUNCT
easat-11054	89	1	however	however	ADV
easat-11054	89	2	,	,	PUNCT
easat-11054	89	3	it	it	PRON
easat-11054	89	4	suffers	suffer	VERB
easat-11054	89	5	from	from	ADP
easat-11054	89	6	the	the	DET
easat-11054	89	7	1056	1056	NUM
easat-11054	89	8	edelweiss	edelweiss	PROPN
easat-11054	89	9	applied	apply	VERB
easat-11054	89	10	science	science	NOUN
easat-11054	89	11	and	and	CCONJ
easat-11054	89	12	technology	technology	NOUN
easat-11054	89	13	issn	issn	PROPN
easat-11054	89	14	:	:	PUNCT
easat-11054	89	15	2576	2576	NUM
easat-11054	89	16	-	-	SYM
easat-11054	89	17	8484	8484	NUM
easat-11054	89	18	vol	vol	NOUN
easat-11054	89	19	.	.	PROPN
easat-11054	90	1	9	9	NUM
easat-11054	90	2	,	,	PUNCT
easat-11054	90	3	no	no	INTJ
easat-11054	90	4	.	.	NOUN
easat-11054	90	5	11	11	NUM
easat-11054	90	6	:	:	PUNCT
easat-11054	90	7	1053	1053	NUM
easat-11054	90	8	-	-	SYM
easat-11054	90	9	1075	1075	NUM
easat-11054	90	10	,	,	PUNCT
easat-11054	90	11	2025	2025	NUM
easat-11054	90	12	doi	doi	NOUN
easat-11054	90	13	:	:	PUNCT
easat-11054	90	14	10.55214/2576	10.55214/2576	NUM
easat-11054	90	15	-	-	SYM
easat-11054	90	16	8484.v9i11.11054	8484.v9i11.11054	NOUN
easat-11054	91	1	©	©	PROPN
easat-11054	91	2	2025	2025	NUM
easat-11054	91	3	by	by	ADP
easat-11054	91	4	the	the	DET
easat-11054	91	5	authors	author	NOUN
easat-11054	91	6	;	;	PUNCT
easat-11054	91	7	licensee	licensee	PROPN
easat-11054	91	8	learning	learning	NOUN
easat-11054	91	9	gate	gate	PROPN
easat-11054	91	10	fact	fact	NOUN
easat-11054	91	11	that	that	SCONJ
easat-11054	91	12	it	it	PRON
easat-11054	91	13	requires	require	VERB
easat-11054	91	14	a	a	DET
easat-11054	91	15	long	long	ADJ
easat-11054	91	16	time	time	NOUN
easat-11054	91	17	to	to	PART
easat-11054	91	18	train	train	VERB
easat-11054	91	19	,	,	PUNCT
easat-11054	91	20	in	in	ADP
easat-11054	91	21	addition	addition	NOUN
easat-11054	91	22	to	to	ADP
easat-11054	91	23	the	the	DET
easat-11054	91	24	fact	fact	NOUN
easat-11054	91	25	that	that	SCONJ
easat-11054	91	26	it	it	PRON
easat-11054	91	27	requires	require	VERB
easat-11054	91	28	complex	complex	ADJ
easat-11054	91	29	calculations	calculation	NOUN
easat-11054	91	30	,	,	PUNCT
easat-11054	91	31	which	which	PRON
easat-11054	91	32	causes	cause	VERB
easat-11054	91	33	a	a	DET
easat-11054	91	34	large	large	ADJ
easat-11054	91	35	consumption	consumption	NOUN
easat-11054	91	36	of	of	ADP
easat-11054	91	37	power	power	NOUN
easat-11054	91	38	.	.	PUNCT
easat-11054	92	1	this	this	PRON
easat-11054	92	2	leads	lead	VERB
easat-11054	92	3	to	to	ADP
easat-11054	92	4	proposing	propose	VERB
easat-11054	92	5	the	the	DET
easat-11054	92	6	hadt	hadt	NOUN
easat-11054	92	7	algorithm	algorithm	NOUN
easat-11054	92	8	,	,	PUNCT
easat-11054	92	9	which	which	PRON
easat-11054	92	10	enhances	enhance	VERB
easat-11054	92	11	detection	detection	NOUN
easat-11054	92	12	performance	performance	NOUN
easat-11054	92	13	through	through	ADP
easat-11054	92	14	effective	effective	ADJ
easat-11054	92	15	training	training	NOUN
easat-11054	92	16	of	of	ADP
easat-11054	92	17	iiot	iiot	NOUN
easat-11054	92	18	network	network	NOUN
easat-11054	92	19	data	datum	NOUN
easat-11054	92	20	,	,	PUNCT
easat-11054	92	21	especially	especially	ADV
easat-11054	92	22	in	in	ADP
easat-11054	92	23	low	low	ADJ
easat-11054	92	24	-	-	PUNCT
easat-11054	92	25	sample	sample	NOUN
easat-11054	92	26	-	-	PUNCT
easat-11054	92	27	size	size	NOUN
easat-11054	92	28	attacks	attack	NOUN
easat-11054	92	29	.	.	PUNCT
easat-11054	93	1	moreover	moreover	ADV
easat-11054	93	2	,	,	PUNCT
easat-11054	93	3	the	the	DET
easat-11054	93	4	use	use	NOUN
easat-11054	93	5	of	of	ADP
easat-11054	93	6	the	the	DET
easat-11054	93	7	xiiotid	xiiotid	NOUN
easat-11054	93	8	dataset	dataset	VERB
easat-11054	93	9	in	in	ADP
easat-11054	93	10	the	the	DET
easat-11054	93	11	hadt	hadt	NOUN
easat-11054	93	12	algorithm	algorithm	NOUN
easat-11054	93	13	provides	provide	VERB
easat-11054	93	14	an	an	DET
easat-11054	93	15	opportunity	opportunity	NOUN
easat-11054	93	16	to	to	PART
easat-11054	93	17	overcome	overcome	VERB
easat-11054	93	18	existing	exist	VERB
easat-11054	93	19	limitations	limitation	NOUN
easat-11054	93	20	in	in	ADP
easat-11054	93	21	ids	id	NOUN
easat-11054	93	22	datasets	dataset	NOUN
easat-11054	93	23	.	.	PUNCT
easat-11054	94	1	unlike	unlike	ADP
easat-11054	94	2	previous	previous	ADJ
easat-11054	94	3	studies	study	NOUN
easat-11054	94	4	that	that	PRON
easat-11054	94	5	relied	rely	VERB
easat-11054	94	6	heavily	heavily	ADV
easat-11054	94	7	on	on	ADP
easat-11054	94	8	specific	specific	ADJ
easat-11054	94	9	datasets	dataset	NOUN
easat-11054	94	10	such	such	ADJ
easat-11054	94	11	as	as	ADP
easat-11054	94	12	nsl	nsl	NOUN
easat-11054	94	13	-	-	PUNCT
easat-11054	94	14	kdd	kdd	NOUN
easat-11054	94	15	,	,	PUNCT
easat-11054	94	16	awid	awid	NOUN
easat-11054	94	17	,	,	PUNCT
easat-11054	94	18	and	and	CCONJ
easat-11054	94	19	cicids-2017	cicids-2017	ADV
easat-11054	94	20	,	,	PUNCT
easat-11054	94	21	this	this	DET
easat-11054	94	22	study	study	NOUN
easat-11054	94	23	uses	use	VERB
easat-11054	94	24	the	the	DET
easat-11054	94	25	x	x	ADJ
easat-11054	94	26	-	-	ADJ
easat-11054	94	27	iiotid	iiotid	ADJ
easat-11054	94	28	dataset	dataset	NOUN
easat-11054	94	29	,	,	PUNCT
easat-11054	94	30	allowing	allow	VERB
easat-11054	94	31	for	for	ADP
easat-11054	94	32	testing	testing	NOUN
easat-11054	94	33	algorithms	algorithm	NOUN
easat-11054	94	34	in	in	ADP
easat-11054	94	35	real	real	ADJ
easat-11054	94	36	iiot	iiot	NOUN
easat-11054	94	37	environments	environment	NOUN
easat-11054	94	38	.	.	PUNCT
easat-11054	95	1	therefore	therefore	ADV
easat-11054	95	2	,	,	PUNCT
easat-11054	95	3	the	the	DET
easat-11054	95	4	hadt	hadt	NOUN
easat-11054	95	5	algorithm	algorithm	NOUN
easat-11054	95	6	,	,	PUNCT
easat-11054	95	7	integrating	integrate	VERB
easat-11054	95	8	simple	simple	ADJ
easat-11054	95	9	a3c	a3c	PUNCT
easat-11054	95	10	and	and	CCONJ
easat-11054	95	11	dt	dt	ADP
easat-11054	95	12	algorithms	algorithm	NOUN
easat-11054	95	13	,	,	PUNCT
easat-11054	95	14	was	be	AUX
easat-11054	95	15	proposed	propose	VERB
easat-11054	95	16	to	to	PART
easat-11054	95	17	handle	handle	VERB
easat-11054	95	18	intrusions	intrusion	NOUN
easat-11054	95	19	in	in	ADP
easat-11054	95	20	both	both	CCONJ
easat-11054	95	21	continuous	continuous	ADJ
easat-11054	95	22	and	and	CCONJ
easat-11054	95	23	discrete	discrete	ADJ
easat-11054	95	24	features	feature	NOUN
easat-11054	95	25	and	and	CCONJ
easat-11054	95	26	to	to	PART
easat-11054	95	27	improve	improve	VERB
easat-11054	95	28	the	the	DET
easat-11054	95	29	detection	detection	NOUN
easat-11054	95	30	of	of	ADP
easat-11054	95	31	various	various	ADJ
easat-11054	95	32	classes	class	NOUN
easat-11054	95	33	of	of	ADP
easat-11054	95	34	intrusions	intrusion	NOUN
easat-11054	95	35	.	.	PUNCT
easat-11054	96	1	by	by	ADP
easat-11054	96	2	utilizing	utilize	VERB
easat-11054	96	3	advanced	advanced	ADJ
easat-11054	96	4	dataset	dataset	ADJ
easat-11054	96	5	preprocessing	preprocessing	NOUN
easat-11054	96	6	strategies	strategy	NOUN
easat-11054	96	7	and	and	CCONJ
easat-11054	96	8	sophisticated	sophisticated	ADJ
easat-11054	96	9	algorithmic	algorithmic	ADJ
easat-11054	96	10	approaches	approach	NOUN
easat-11054	96	11	,	,	PUNCT
easat-11054	96	12	the	the	DET
easat-11054	96	13	proposed	propose	VERB
easat-11054	96	14	work	work	NOUN
easat-11054	96	15	facilitates	facilitate	VERB
easat-11054	96	16	the	the	DET
easat-11054	96	17	development	development	NOUN
easat-11054	96	18	of	of	ADP
easat-11054	96	19	effective	effective	ADJ
easat-11054	96	20	and	and	CCONJ
easat-11054	96	21	adaptive	adaptive	ADJ
easat-11054	96	22	ids	id	NOUN
easat-11054	96	23	solutions	solution	NOUN
easat-11054	96	24	for	for	ADP
easat-11054	96	25	iiot	iiot	NOUN
easat-11054	96	26	security	security	NOUN
easat-11054	96	27	.	.	PUNCT
easat-11054	97	1	in	in	ADP
easat-11054	97	2	table	table	NOUN
easat-11054	97	3	1	1	NUM
easat-11054	97	4	,	,	PUNCT
easat-11054	97	5	a	a	DET
easat-11054	97	6	thorough	thorough	ADJ
easat-11054	97	7	comparison	comparison	NOUN
easat-11054	97	8	of	of	ADP
easat-11054	97	9	a3c	a3c	VERB
easat-11054	97	10	-	-	PUNCT
easat-11054	97	11	based	base	VERB
easat-11054	97	12	network	network	NOUN
easat-11054	97	13	intrusion	intrusion	NOUN
easat-11054	97	14	detection	detection	NOUN
easat-11054	97	15	methods	method	NOUN
easat-11054	97	16	is	be	AUX
easat-11054	97	17	presented	present	VERB
easat-11054	97	18	,	,	PUNCT
easat-11054	97	19	emphasizing	emphasize	VERB
easat-11054	97	20	their	their	PRON
easat-11054	97	21	study	study	NOUN
easat-11054	97	22	algorithms	algorithm	NOUN
easat-11054	97	23	,	,	PUNCT
easat-11054	97	24	nn	nn	PROPN
easat-11054	97	25	architectures	architecture	NOUN
easat-11054	97	26	,	,	PUNCT
easat-11054	97	27	datasets	dataset	NOUN
easat-11054	97	28	,	,	PUNCT
easat-11054	97	29	strengths	strength	NOUN
easat-11054	97	30	,	,	PUNCT
easat-11054	97	31	and	and	CCONJ
easat-11054	97	32	limitations	limitation	NOUN
easat-11054	97	33	.	.	PUNCT
easat-11054	98	1	these	these	DET
easat-11054	98	2	related	relate	VERB
easat-11054	98	3	works	work	NOUN
easat-11054	98	4	are	be	AUX
easat-11054	98	5	summarized	summarize	VERB
easat-11054	98	6	below	below	ADP
easat-11054	98	7	:	:	PUNCT
easat-11054	98	8	table	table	NOUN
easat-11054	98	9	1	1	NUM
easat-11054	98	10	.	.	PUNCT
easat-11054	98	11	related	relate	VERB
easat-11054	98	12	work	work	NOUN
easat-11054	98	13	summary	summary	NOUN
easat-11054	98	14	.	.	PUNCT
easat-11054	99	1	ref	ref	NOUN
easat-11054	99	2	.	.	PUNCT
easat-11054	100	1	publisher	publisher	NOUN
easat-11054	100	2	study	study	NOUN
easat-11054	100	3	methods	method	NOUN
easat-11054	100	4	type	type	NOUN
easat-11054	100	5	of	of	ADP
easat-11054	100	6	nn	nn	PROPN
easat-11054	100	7	in	in	ADP
easat-11054	100	8	worker	worker	NOUN
easat-11054	100	9	dataset	dataset	NOUN
easat-11054	100	10	strength	strength	NOUN
easat-11054	100	11	of	of	ADP
easat-11054	100	12	methods	method	NOUN
easat-11054	100	13	limitations	limitation	NOUN
easat-11054	100	14	of	of	ADP
easat-11054	100	15	methods	method	NOUN
easat-11054	100	16	muhati	muhati	PROPN
easat-11054	100	17	and	and	CCONJ
easat-11054	100	18	rawat	rawat	PROPN
easat-11054	100	19	[	[	X
easat-11054	100	20	16	16	NUM
easat-11054	100	21	]	]	X
easat-11054	100	22	muhati	muhati	PROPN
easat-11054	100	23	and	and	CCONJ
easat-11054	100	24	rawat	rawat	PROPN
easat-11054	101	1	[	[	X
easat-11054	101	2	16	16	NUM
easat-11054	101	3	]	]	PUNCT
easat-11054	101	4	a3c	a3c	NUM
easat-11054	101	5	-	-	PUNCT
easat-11054	101	6	based	base	VERB
easat-11054	101	7	cognitive	cognitive	ADJ
easat-11054	101	8	nids	nid	NOUN
easat-11054	101	9	with	with	ADP
easat-11054	101	10	drl	drl	PROPN
easat-11054	101	11	ffnn	ffnn	NOUN
easat-11054	101	12	unswnb15	unswnb15	NOUN
easat-11054	101	13	,	,	PUNCT
easat-11054	101	14	awid	awid	NOUN
easat-11054	101	15	,	,	PUNCT
easat-11054	101	16	nslkdd	nslkdd	ADV
easat-11054	101	17	adaptive	adaptive	ADJ
easat-11054	101	18	to	to	ADP
easat-11054	101	19	evolving	evolve	VERB
easat-11054	101	20	attack	attack	NOUN
easat-11054	101	21	patterns	pattern	NOUN
easat-11054	101	22	,	,	PUNCT
easat-11054	101	23	improving	improve	VERB
easat-11054	101	24	anomaly	anomaly	NOUN
easat-11054	101	25	detection	detection	NOUN
easat-11054	101	26	accuracy	accuracy	NOUN
easat-11054	101	27	.	.	PUNCT
easat-11054	102	1	limited	limited	ADJ
easat-11054	102	2	adaptability	adaptability	NOUN
easat-11054	102	3	to	to	ADP
easat-11054	102	4	new	new	ADJ
easat-11054	102	5	attack	attack	NOUN
easat-11054	102	6	patterns	pattern	NOUN
easat-11054	102	7	due	due	ADP
easat-11054	102	8	to	to	ADP
easat-11054	102	9	static	static	ADJ
easat-11054	102	10	feature	feature	NOUN
easat-11054	102	11	selection	selection	NOUN
easat-11054	102	12	.	.	PUNCT
easat-11054	103	1	zhou	zhou	PROPN
easat-11054	103	2	et	et	PROPN
easat-11054	103	3	al	al	PROPN
easat-11054	103	4	.	.	PUNCT
easat-11054	104	1	[	[	X
easat-11054	104	2	17	17	NUM
easat-11054	104	3	]	]	X
easat-11054	104	4	zhou	zhou	PROPN
easat-11054	104	5	et	et	PROPN
easat-11054	104	6	al	al	PROPN
easat-11054	104	7	.	.	PUNCT
easat-11054	105	1	[	[	X
easat-11054	105	2	17	17	NUM
easat-11054	105	3	]	]	PUNCT
easat-11054	105	4	a3c	a3c	NUM
easat-11054	105	5	algorithm	algorithm	NOUN
easat-11054	105	6	with	with	ADP
easat-11054	105	7	attention	attention	NOUN
easat-11054	105	8	&	&	CCONJ
easat-11054	105	9	cnn	cnn	PROPN
easat-11054	105	10	cnn	cnn	PROPN
easat-11054	106	1	+	+	CCONJ
easat-11054	106	2	attention	attention	NOUN
easat-11054	106	3	nsl	nsl	PROPN
easat-11054	106	4	-	-	PUNCT
easat-11054	106	5	kdd	kdd	PROPN
easat-11054	106	6	,	,	PUNCT
easat-11054	106	7	awid	awid	NOUN
easat-11054	106	8	,	,	PUNCT
easat-11054	106	9	cicids2017	cicids2017	PROPN
easat-11054	106	10	,	,	PUNCT
easat-11054	106	11	dohbrw2020	dohbrw2020	VERB
easat-11054	106	12	high	high	ADJ
easat-11054	106	13	precision	precision	NOUN
easat-11054	106	14	,	,	PUNCT
easat-11054	106	15	recall	recall	NOUN
easat-11054	106	16	,	,	PUNCT
easat-11054	106	17	and	and	CCONJ
easat-11054	106	18	f1score	f1score	NOUN
easat-11054	106	19	for	for	ADP
easat-11054	106	20	anomaly	anomaly	NOUN
easat-11054	106	21	detection	detection	NOUN
easat-11054	106	22	high	high	ADJ
easat-11054	106	23	computational	computational	ADJ
easat-11054	106	24	costs	cost	NOUN
easat-11054	106	25	limit	limit	VERB
easat-11054	106	26	real	real	ADJ
easat-11054	106	27	-	-	PUNCT
easat-11054	106	28	time	time	NOUN
easat-11054	106	29	deployment	deployment	NOUN
easat-11054	106	30	in	in	ADP
easat-11054	106	31	iiot	iiot	PROPN
easat-11054	106	32	.	.	PUNCT
easat-11054	107	1	shen	shen	PROPN
easat-11054	107	2	et	et	PROPN
easat-11054	107	3	al	al	PROPN
easat-11054	107	4	.	.	PUNCT
easat-11054	108	1	[	[	X
easat-11054	108	2	18	18	NUM
easat-11054	108	3	]	]	X
easat-11054	108	4	shen	shen	NOUN
easat-11054	108	5	et	et	PROPN
easat-11054	108	6	al	al	PROPN
easat-11054	108	7	.	.	PUNCT
easat-11054	109	1	[	[	X
easat-11054	109	2	18	18	NUM
easat-11054	109	3	]	]	SYM
easat-11054	109	4	mean	mean	ADJ
easat-11054	109	5	-	-	PUNCT
easat-11054	109	6	field	field	NOUN
easat-11054	109	7	gamebased	gamebase	VERB
easat-11054	109	8	a3c	a3c	NUM
easat-11054	109	9	(	(	PUNCT
easat-11054	109	10	mfga3c	mfga3c	NUM
easat-11054	109	11	)	)	PUNCT
easat-11054	109	12	ffnn	ffnn	NOUN
easat-11054	109	13	not	not	PART
easat-11054	109	14	specified	specify	VERB
easat-11054	109	15	effective	effective	ADJ
easat-11054	109	16	against	against	ADP
easat-11054	109	17	adversarial	adversarial	ADJ
easat-11054	109	18	attacks	attack	NOUN
easat-11054	109	19	,	,	PUNCT
easat-11054	109	20	enhancing	enhance	VERB
easat-11054	109	21	malware	malware	NOUN
easat-11054	109	22	mitigation	mitigation	NOUN
easat-11054	109	23	.	.	PUNCT
easat-11054	110	1	scalability	scalability	NOUN
easat-11054	110	2	issues	issue	NOUN
easat-11054	110	3	in	in	ADP
easat-11054	110	4	large	large	ADJ
easat-11054	110	5	iiot	iiot	NOUN
easat-11054	110	6	networks	network	NOUN
easat-11054	110	7	with	with	ADP
easat-11054	110	8	high	high	ADJ
easat-11054	110	9	traffic	traffic	NOUN
easat-11054	110	10	loads	load	NOUN
easat-11054	110	11	watts	watt	VERB
easat-11054	110	12	et	et	PROPN
easat-11054	110	13	al	al	PROPN
easat-11054	110	14	.	.	PUNCT
easat-11054	111	1	[	[	X
easat-11054	111	2	9	9	NUM
easat-11054	111	3	]	]	PUNCT
easat-11054	111	4	watts	watts	PROPN
easat-11054	111	5	et	et	PROPN
easat-11054	111	6	al	al	PROPN
easat-11054	111	7	.	.	PUNCT
easat-11054	112	1	[	[	X
easat-11054	112	2	9	9	NUM
easat-11054	112	3	]	]	PUNCT
easat-11054	112	4	bayesian	bayesian	NOUN
easat-11054	112	5	drl	drl	PROPN
easat-11054	112	6	with	with	ADP
easat-11054	112	7	cnn	cnn	PROPN
easat-11054	112	8	&	&	CCONJ
easat-11054	112	9	pomdp	pomdp	PROPN
easat-11054	112	10	cnn	cnn	PROPN
easat-11054	112	11	+	+	CCONJ
easat-11054	112	12	pomdp	pomdp	NOUN
easat-11054	112	13	not	not	PART
easat-11054	112	14	specified	specify	VERB
easat-11054	112	15	robust	robust	ADJ
easat-11054	112	16	real	real	ADJ
easat-11054	112	17	-	-	PUNCT
easat-11054	112	18	time	time	NOUN
easat-11054	112	19	anomaly	anomaly	NOUN
easat-11054	112	20	detection	detection	NOUN
easat-11054	112	21	with	with	ADP
easat-11054	112	22	adjustable	adjustable	ADJ
easat-11054	112	23	thresholds	threshold	NOUN
easat-11054	112	24	.	.	PUNCT
easat-11054	113	1	requires	require	VERB
easat-11054	113	2	extensive	extensive	ADJ
easat-11054	113	3	training	training	NOUN
easat-11054	113	4	data	datum	NOUN
easat-11054	113	5	;	;	PUNCT
easat-11054	113	6	struggles	struggle	NOUN
easat-11054	113	7	with	with	ADP
easat-11054	113	8	low	low	ADJ
easat-11054	113	9	sample	sample	NOUN
easat-11054	113	10	size	size	NOUN
easat-11054	113	11	attacks	attack	NOUN
easat-11054	113	12	in	in	ADP
easat-11054	113	13	the	the	DET
easat-11054	113	14	dataset	dataset	NOUN
easat-11054	113	15	.	.	PUNCT
easat-11054	114	1	dong	dong	PROPN
easat-11054	114	2	et	et	PROPN
easat-11054	114	3	al	al	PROPN
easat-11054	114	4	.	.	PUNCT
easat-11054	115	1	[	[	X
easat-11054	115	2	19	19	NUM
easat-11054	115	3	]	]	X
easat-11054	115	4	dong	dong	PROPN
easat-11054	115	5	et	et	PROPN
easat-11054	115	6	al	al	PROPN
easat-11054	115	7	.	.	PUNCT
easat-11054	116	1	[	[	X
easat-11054	116	2	19	19	NUM
easat-11054	116	3	]	]	PUNCT
easat-11054	116	4	feature	feature	NOUN
easat-11054	116	5	selection	selection	NOUN
easat-11054	116	6	&	&	CCONJ
easat-11054	116	7	drl	drl	PROPN
easat-11054	116	8	-	-	PUNCT
easat-11054	116	9	based	base	VERB
easat-11054	116	10	ids	id	NOUN
easat-11054	116	11	ffnn	ffnn	NOUN
easat-11054	116	12	not	not	PART
easat-11054	116	13	specified	specify	VERB
easat-11054	116	14	optimizing	optimize	VERB
easat-11054	116	15	detection	detection	NOUN
easat-11054	116	16	accuracy	accuracy	NOUN
easat-11054	116	17	through	through	ADP
easat-11054	116	18	feature	feature	NOUN
easat-11054	116	19	selection	selection	NOUN
easat-11054	116	20	.	.	PUNCT
easat-11054	117	1	high	high	ADJ
easat-11054	117	2	-	-	PUNCT
easat-11054	117	3	dimensional	dimensional	ADJ
easat-11054	117	4	feature	feature	NOUN
easat-11054	117	5	space	space	NOUN
easat-11054	117	6	increases	increase	VERB
easat-11054	117	7	training	train	VERB
easat-11054	117	8	complexity	complexity	NOUN
easat-11054	117	9	.	.	PUNCT
easat-11054	118	1	khako	khako	PROPN
easat-11054	118	2	and	and	CCONJ
easat-11054	118	3	alhomdy	alhomdy	VERB
easat-11054	119	1	[	[	X
easat-11054	119	2	21	21	NUM
easat-11054	119	3	]	]	X
easat-11054	119	4	khako	khako	PROPN
easat-11054	119	5	and	and	CCONJ
easat-11054	119	6	alhomdy	alhomdy	VERB
easat-11054	120	1	[	[	X
easat-11054	120	2	21	21	NUM
easat-11054	120	3	]	]	X
easat-11054	120	4	ea3c	ea3c	VERB
easat-11054	120	5	with	with	ADP
easat-11054	120	6	gradient	gradient	NOUN
easat-11054	120	7	clipping	clipping	NOUN
easat-11054	120	8	,	,	PUNCT
easat-11054	120	9	and	and	CCONJ
easat-11054	120	10	an	an	DET
easat-11054	120	11	improved	improved	ADJ
easat-11054	120	12	parameter	parameter	NOUN
easat-11054	120	13	synchronization	synchronization	NOUN
easat-11054	120	14	mechanism	mechanism	NOUN
easat-11054	120	15	cnn	cnn	PROPN
easat-11054	120	16	x	x	PROPN
easat-11054	120	17	-	-	PROPN
easat-11054	120	18	iiotid	iiotid	ADJ
easat-11054	120	19	handling	handling	NOUN
easat-11054	120	20	lowsample	lowsample	ADJ
easat-11054	120	21	-	-	PUNCT
easat-11054	120	22	size	size	NOUN
easat-11054	120	23	attacks	attack	NOUN
easat-11054	120	24	and	and	CCONJ
easat-11054	120	25	mixed	mixed	ADJ
easat-11054	120	26	featuretype	featuretype	NOUN
easat-11054	120	27	environments	environment	NOUN
easat-11054	120	28	it	it	PRON
easat-11054	120	29	requires	require	VERB
easat-11054	120	30	a	a	DET
easat-11054	120	31	long	long	ADJ
easat-11054	120	32	time	time	NOUN
easat-11054	120	33	to	to	PART
easat-11054	120	34	train	train	VERB
easat-11054	120	35	and	and	CCONJ
easat-11054	120	36	complex	complex	ADJ
easat-11054	120	37	calculations	calculation	NOUN
easat-11054	120	38	.	.	PUNCT
easat-11054	121	1	3	3	X
easat-11054	121	2	.	.	NUM
easat-11054	121	3	proposed	propose	VERB
easat-11054	121	4	hadt	hadt	NOUN
easat-11054	121	5	the	the	DET
easat-11054	121	6	proposed	propose	VERB
easat-11054	121	7	hadt	hadt	NOUN
easat-11054	121	8	is	be	AUX
easat-11054	121	9	a	a	DET
easat-11054	121	10	combination	combination	NOUN
easat-11054	121	11	of	of	ADP
easat-11054	121	12	two	two	NUM
easat-11054	121	13	algorithms	algorithm	NOUN
easat-11054	121	14	,	,	PUNCT
easat-11054	121	15	which	which	PRON
easat-11054	121	16	are	be	AUX
easat-11054	121	17	the	the	DET
easat-11054	121	18	simple	simple	ADJ
easat-11054	121	19	a3c	a3c	NUM
easat-11054	121	20	and	and	CCONJ
easat-11054	121	21	dt	dt	NOUN
easat-11054	121	22	.	.	PUNCT
easat-11054	122	1	in	in	ADP
easat-11054	122	2	a	a	DET
easat-11054	122	3	continuous	continuous	ADJ
easat-11054	122	4	feature	feature	NOUN
easat-11054	122	5	,	,	PUNCT
easat-11054	122	6	a3c	a3c	NUM
easat-11054	122	7	can	can	AUX
easat-11054	122	8	be	be	AUX
easat-11054	122	9	used	use	VERB
easat-11054	122	10	to	to	PART
easat-11054	122	11	efficiently	efficiently	ADV
easat-11054	122	12	calculate	calculate	VERB
easat-11054	122	13	policy	policy	NOUN
easat-11054	122	14	functions	function	NOUN
easat-11054	122	15	and	and	CCONJ
easat-11054	122	16	predict	predict	VERB
easat-11054	122	17	intrusion	intrusion	NOUN
easat-11054	122	18	actions	action	NOUN
easat-11054	122	19	.	.	PUNCT
easat-11054	123	1	adding	add	VERB
easat-11054	123	2	a	a	DET
easat-11054	123	3	dt	dt	PROPN
easat-11054	123	4	unit	unit	NOUN
easat-11054	123	5	was	be	AUX
easat-11054	123	6	necessary	necessary	ADJ
easat-11054	123	7	to	to	PART
easat-11054	123	8	address	address	VERB
easat-11054	123	9	features	feature	NOUN
easat-11054	123	10	of	of	ADP
easat-11054	123	11	the	the	DET
easat-11054	123	12	deterministic	deterministic	ADJ
easat-11054	123	13	environment	environment	NOUN
easat-11054	123	14	and	and	CCONJ
easat-11054	123	15	improve	improve	VERB
easat-11054	123	16	intrusion	intrusion	NOUN
easat-11054	123	17	detection	detection	NOUN
easat-11054	123	18	performance	performance	NOUN
easat-11054	123	19	by	by	ADP
easat-11054	123	20	selecting	select	VERB
easat-11054	123	21	optimal	optimal	ADJ
easat-11054	123	22	actions	action	NOUN
easat-11054	123	23	for	for	ADP
easat-11054	123	24	detecting	detect	VERB
easat-11054	123	25	attacks	attack	NOUN
easat-11054	123	26	.	.	PUNCT
easat-11054	124	1	to	to	PART
easat-11054	124	2	achieve	achieve	VERB
easat-11054	124	3	this	this	PRON
easat-11054	124	4	,	,	PUNCT
easat-11054	124	5	an	an	DET
easat-11054	124	6	1057	1057	NUM
easat-11054	124	7	edelweiss	edelweiss	PROPN
easat-11054	124	8	applied	apply	VERB
easat-11054	124	9	science	science	NOUN
easat-11054	124	10	and	and	CCONJ
easat-11054	124	11	technology	technology	NOUN
easat-11054	124	12	issn	issn	PROPN
easat-11054	124	13	:	:	PUNCT
easat-11054	124	14	2576	2576	NUM
easat-11054	124	15	-	-	SYM
easat-11054	124	16	8484	8484	NUM
easat-11054	124	17	vol	vol	NOUN
easat-11054	124	18	.	.	PROPN
easat-11054	125	1	9	9	NUM
easat-11054	125	2	,	,	PUNCT
easat-11054	125	3	no	no	INTJ
easat-11054	125	4	.	.	NOUN
easat-11054	125	5	11	11	NUM
easat-11054	125	6	:	:	PUNCT
easat-11054	125	7	1053	1053	NUM
easat-11054	125	8	-	-	SYM
easat-11054	125	9	1075	1075	NUM
easat-11054	125	10	,	,	PUNCT
easat-11054	125	11	2025	2025	NUM
easat-11054	125	12	doi	doi	NOUN
easat-11054	125	13	:	:	PUNCT
easat-11054	125	14	10.55214/2576	10.55214/2576	NUM
easat-11054	125	15	-	-	SYM
easat-11054	125	16	8484.v9i11.11054	8484.v9i11.11054	NOUN
easat-11054	126	1	©	©	PROPN
easat-11054	126	2	2025	2025	NUM
easat-11054	126	3	by	by	ADP
easat-11054	126	4	the	the	DET
easat-11054	126	5	authors	author	NOUN
easat-11054	126	6	;	;	PUNCT
easat-11054	126	7	licensee	licensee	PROPN
easat-11054	126	8	learning	learn	VERB
easat-11054	126	9	gate	gate	NOUN
easat-11054	126	10	ffnn	ffnn	NOUN
easat-11054	126	11	layer	layer	NOUN
easat-11054	126	12	is	be	AUX
easat-11054	126	13	proposed	propose	VERB
easat-11054	126	14	to	to	PART
easat-11054	126	15	hybridize	hybridize	VERB
easat-11054	126	16	the	the	DET
easat-11054	126	17	a3c	a3c	NUM
easat-11054	126	18	-	-	PUNCT
easat-11054	126	19	ffnn	ffnn	NOUN
easat-11054	126	20	and	and	CCONJ
easat-11054	126	21	dt	dt	NOUN
easat-11054	126	22	outputs	output	NOUN
easat-11054	126	23	in	in	ADP
easat-11054	126	24	the	the	DET
easat-11054	126	25	hadt	hadt	NOUN
easat-11054	126	26	algorithm	algorithm	NOUN
easat-11054	126	27	,	,	PUNCT
easat-11054	126	28	as	as	SCONJ
easat-11054	126	29	illustrated	illustrate	VERB
easat-11054	126	30	in	in	ADP
easat-11054	126	31	figure	figure	NOUN
easat-11054	126	32	1	1	NUM
easat-11054	126	33	.	.	PUNCT
easat-11054	127	1	as	as	SCONJ
easat-11054	127	2	described	describe	VERB
easat-11054	127	3	,	,	PUNCT
easat-11054	127	4	the	the	DET
easat-11054	127	5	hadt	hadt	NOUN
easat-11054	127	6	algorithm	algorithm	NOUN
easat-11054	127	7	integrates	integrate	VERB
easat-11054	127	8	multiple	multiple	ADJ
easat-11054	127	9	learning	learn	VERB
easat-11054	127	10	algorithms	algorithm	NOUN
easat-11054	127	11	to	to	PART
easat-11054	127	12	enhance	enhance	VERB
easat-11054	127	13	intrusion	intrusion	NOUN
easat-11054	127	14	detection	detection	NOUN
easat-11054	127	15	in	in	ADP
easat-11054	127	16	iiot	iiot	ADJ
easat-11054	127	17	environments	environment	NOUN
easat-11054	127	18	as	as	SCONJ
easat-11054	127	19	follows	follow	VERB
easat-11054	127	20	:	:	PUNCT
easat-11054	127	21	1	1	NUM
easat-11054	127	22	)	)	PUNCT
easat-11054	127	23	a3c	a3c	NUM
easat-11054	127	24	-	-	PUNCT
easat-11054	127	25	ffnn	ffnn	NOUN
easat-11054	127	26	whose	whose	DET
easat-11054	127	27	worker	worker	NOUN
easat-11054	127	28	has	have	VERB
easat-11054	127	29	a	a	DET
easat-11054	127	30	simple	simple	ADJ
easat-11054	127	31	ffnn	ffnn	NOUN
easat-11054	127	32	layer	layer	NOUN
easat-11054	127	33	that	that	PRON
easat-11054	127	34	produces	produce	VERB
easat-11054	127	35	initial	initial	ADJ
easat-11054	127	36	policy	policy	NOUN
easat-11054	127	37	and	and	CCONJ
easat-11054	127	38	value	value	NOUN
easat-11054	127	39	functions	function	NOUN
easat-11054	127	40	.	.	PUNCT
easat-11054	128	1	it	it	PRON
easat-11054	128	2	consists	consist	VERB
easat-11054	128	3	of	of	ADP
easat-11054	128	4	a	a	DET
easat-11054	128	5	two	two	NUM
easat-11054	128	6	-	-	PUNCT
easat-11054	128	7	layer	layer	NOUN
easat-11054	128	8	neural	neural	ADJ
easat-11054	128	9	network	network	NOUN
easat-11054	128	10	connected	connect	VERB
easat-11054	128	11	to	to	ADP
easat-11054	128	12	an	an	DET
easat-11054	128	13	input	input	NOUN
easat-11054	128	14	layer	layer	NOUN
easat-11054	128	15	,	,	PUNCT
easat-11054	128	16	followed	follow	VERB
easat-11054	128	17	by	by	ADP
easat-11054	128	18	a	a	DET
easat-11054	128	19	dense	dense	ADJ
easat-11054	128	20	layer	layer	NOUN
easat-11054	128	21	with	with	ADP
easat-11054	128	22	relu	relu	NOUN
easat-11054	128	23	activation	activation	NOUN
easat-11054	128	24	and	and	CCONJ
easat-11054	128	25	a	a	DET
easat-11054	128	26	dropout	dropout	NOUN
easat-11054	128	27	layer	layer	NOUN
easat-11054	128	28	for	for	ADP
easat-11054	128	29	regularization	regularization	NOUN
easat-11054	128	30	.	.	PUNCT
easat-11054	129	1	the	the	DET
easat-11054	129	2	algorithm	algorithm	NOUN
easat-11054	129	3	has	have	VERB
easat-11054	129	4	two	two	NUM
easat-11054	129	5	output	output	NOUN
easat-11054	129	6	layers	layer	NOUN
easat-11054	129	7	:	:	PUNCT
easat-11054	129	8	one	one	NUM
easat-11054	129	9	for	for	ADP
easat-11054	129	10	the	the	DET
easat-11054	129	11	policy	policy	NOUN
easat-11054	129	12	function	function	NOUN
easat-11054	129	13	activated	activate	VERB
easat-11054	129	14	with	with	ADP
easat-11054	129	15	a	a	DET
easat-11054	129	16	softmax	softmax	NOUN
easat-11054	129	17	function	function	NOUN
easat-11054	129	18	,	,	PUNCT
easat-11054	129	19	and	and	CCONJ
easat-11054	129	20	the	the	DET
easat-11054	129	21	value	value	NOUN
easat-11054	129	22	function	function	NOUN
easat-11054	129	23	produced	produce	VERB
easat-11054	129	24	with	with	ADP
easat-11054	129	25	a	a	DET
easat-11054	129	26	linear	linear	ADJ
easat-11054	129	27	function	function	NOUN
easat-11054	129	28	.	.	PUNCT
easat-11054	130	1	the	the	DET
easat-11054	130	2	ffnn	ffnn	NOUN
easat-11054	130	3	layer	layer	NOUN
easat-11054	130	4	in	in	ADP
easat-11054	130	5	a3c	a3c	VERB
easat-11054	130	6	-	-	PUNCT
easat-11054	130	7	ffnn	ffnn	NOUN
easat-11054	130	8	is	be	AUX
easat-11054	130	9	compiled	compile	VERB
easat-11054	130	10	using	use	VERB
easat-11054	130	11	the	the	DET
easat-11054	130	12	adam	adam	PROPN
easat-11054	130	13	optimizer	optimizer	NOUN
easat-11054	130	14	.	.	PUNCT
easat-11054	131	1	2	2	X
easat-11054	131	2	)	)	PUNCT
easat-11054	131	3	the	the	DET
easat-11054	131	4	dt	dt	PROPN
easat-11054	131	5	algorithm	algorithm	PROPN
easat-11054	131	6	is	be	AUX
easat-11054	131	7	used	use	VERB
easat-11054	131	8	to	to	PART
easat-11054	131	9	make	make	VERB
easat-11054	131	10	action	action	NOUN
easat-11054	131	11	predictions	prediction	NOUN
easat-11054	131	12	in	in	ADP
easat-11054	131	13	parallel	parallel	NOUN
easat-11054	131	14	with	with	ADP
easat-11054	131	15	the	the	DET
easat-11054	131	16	a3c	a3c	NUM
easat-11054	131	17	-	-	PUNCT
easat-11054	131	18	ffnn	ffnn	NOUN
easat-11054	131	19	.	.	PUNCT
easat-11054	132	1	this	this	DET
easat-11054	132	2	algorithm	algorithm	NOUN
easat-11054	132	3	has	have	VERB
easat-11054	132	4	automatic	automatic	ADJ
easat-11054	132	5	feature	feature	NOUN
easat-11054	132	6	selection	selection	NOUN
easat-11054	132	7	and	and	CCONJ
easat-11054	132	8	interactions	interaction	NOUN
easat-11054	132	9	and	and	CCONJ
easat-11054	132	10	the	the	DET
easat-11054	132	11	capability	capability	NOUN
easat-11054	132	12	to	to	PART
easat-11054	132	13	offer	offer	VERB
easat-11054	132	14	explainable	explainable	ADJ
easat-11054	132	15	decision	decision	NOUN
easat-11054	132	16	rules	rule	NOUN
easat-11054	132	17	,	,	PUNCT
easat-11054	132	18	leading	lead	VERB
easat-11054	132	19	to	to	ADP
easat-11054	132	20	reduced	reduce	VERB
easat-11054	132	21	overfitting	overfitting	NOUN
easat-11054	132	22	,	,	PUNCT
easat-11054	132	23	increased	increase	VERB
easat-11054	132	24	accuracy	accuracy	NOUN
easat-11054	132	25	,	,	PUNCT
easat-11054	132	26	and	and	CCONJ
easat-11054	132	27	more	more	ADV
easat-11054	132	28	reliable	reliable	ADJ
easat-11054	132	29	estimates	estimate	NOUN
easat-11054	132	30	of	of	ADP
easat-11054	132	31	probability	probability	NOUN
easat-11054	132	32	for	for	ADP
easat-11054	132	33	intrusion	intrusion	NOUN
easat-11054	132	34	classification	classification	NOUN
easat-11054	132	35	.	.	PUNCT
easat-11054	133	1	3	3	X
easat-11054	133	2	)	)	PUNCT
easat-11054	133	3	the	the	DET
easat-11054	133	4	ffnn	ffnn	NOUN
easat-11054	133	5	layer	layer	NOUN
easat-11054	133	6	hybridizes	hybridize	VERB
easat-11054	133	7	the	the	DET
easat-11054	133	8	predictions	prediction	NOUN
easat-11054	133	9	of	of	ADP
easat-11054	133	10	both	both	CCONJ
easat-11054	133	11	the	the	DET
easat-11054	133	12	dt	dt	NOUN
easat-11054	133	13	and	and	CCONJ
easat-11054	133	14	the	the	DET
easat-11054	133	15	a3c	a3c	VERB
easat-11054	133	16	-	-	PUNCT
easat-11054	133	17	ffnn	ffnn	NOUN
easat-11054	133	18	as	as	ADP
easat-11054	133	19	inputs	input	NOUN
easat-11054	133	20	.	.	PUNCT
easat-11054	134	1	concatenation	concatenation	NOUN
easat-11054	134	2	of	of	ADP
easat-11054	134	3	policy	policy	NOUN
easat-11054	134	4	predictions	prediction	NOUN
easat-11054	134	5	is	be	AUX
easat-11054	134	6	achieved	achieve	VERB
easat-11054	134	7	through	through	ADP
easat-11054	134	8	a	a	DET
easat-11054	134	9	hidden	hide	VERB
easat-11054	134	10	layer	layer	NOUN
easat-11054	134	11	that	that	PRON
easat-11054	134	12	is	be	AUX
easat-11054	134	13	activated	activate	VERB
easat-11054	134	14	with	with	ADP
easat-11054	134	15	relu	relu	NOUN
easat-11054	134	16	,	,	PUNCT
easat-11054	134	17	and	and	CCONJ
easat-11054	134	18	regularization	regularization	NOUN
easat-11054	134	19	is	be	AUX
easat-11054	134	20	accomplished	accomplish	VERB
easat-11054	134	21	through	through	ADP
easat-11054	134	22	the	the	DET
easat-11054	134	23	use	use	NOUN
easat-11054	134	24	of	of	ADP
easat-11054	134	25	a	a	DET
easat-11054	134	26	dropout	dropout	NOUN
easat-11054	134	27	layer	layer	NOUN
easat-11054	134	28	.	.	PUNCT
easat-11054	135	1	compiling	compile	VERB
easat-11054	135	2	the	the	DET
easat-11054	135	3	output	output	NOUN
easat-11054	135	4	policy	policy	NOUN
easat-11054	135	5	function	function	NOUN
easat-11054	135	6	with	with	ADP
easat-11054	135	7	the	the	DET
easat-11054	135	8	adam	adam	PROPN
easat-11054	135	9	optimizer	optimizer	NOUN
easat-11054	135	10	and	and	CCONJ
easat-11054	135	11	sparse	sparse	VERB
easat-11054	135	12	categorical	categorical	ADJ
easat-11054	135	13	cross	cross	ADJ
easat-11054	135	14	-	-	ADJ
easat-11054	135	15	entropy	entropy	ADJ
easat-11054	135	16	loss	loss	NOUN
easat-11054	135	17	is	be	AUX
easat-11054	135	18	necessary	necessary	ADJ
easat-11054	135	19	for	for	SCONJ
easat-11054	135	20	the	the	DET
easat-11054	135	21	hybrid	hybrid	ADJ
easat-11054	135	22	ffnn	ffnn	NOUN
easat-11054	135	23	output	output	NOUN
easat-11054	135	24	layer	layer	NOUN
easat-11054	135	25	,	,	PUNCT
easat-11054	135	26	which	which	PRON
easat-11054	135	27	uses	use	VERB
easat-11054	135	28	the	the	DET
easat-11054	135	29	softmax	softmax	NOUN
easat-11054	135	30	activation	activation	NOUN
easat-11054	135	31	function	function	NOUN
easat-11054	135	32	.	.	PUNCT
easat-11054	136	1	the	the	DET
easat-11054	136	2	convergence	convergence	NOUN
easat-11054	136	3	effect	effect	NOUN
easat-11054	136	4	of	of	ADP
easat-11054	136	5	the	the	DET
easat-11054	136	6	network	network	NOUN
easat-11054	136	7	is	be	AUX
easat-11054	136	8	maintained	maintain	VERB
easat-11054	136	9	by	by	ADP
easat-11054	136	10	adjusting	adjust	VERB
easat-11054	136	11	its	its	PRON
easat-11054	136	12	weights	weight	NOUN
easat-11054	136	13	and	and	CCONJ
easat-11054	136	14	biases	bias	NOUN
easat-11054	136	15	repeatedly	repeatedly	ADV
easat-11054	136	16	,	,	PUNCT
easat-11054	136	17	which	which	PRON
easat-11054	136	18	requires	require	VERB
easat-11054	136	19	frequent	frequent	ADJ
easat-11054	136	20	training	training	NOUN
easat-11054	136	21	.	.	PUNCT
easat-11054	137	1	this	this	DET
easat-11054	137	2	layer	layer	NOUN
easat-11054	137	3	is	be	AUX
easat-11054	137	4	added	add	VERB
easat-11054	137	5	to	to	PART
easat-11054	137	6	leverage	leverage	VERB
easat-11054	137	7	the	the	DET
easat-11054	137	8	strengths	strength	NOUN
easat-11054	137	9	of	of	ADP
easat-11054	137	10	both	both	CCONJ
easat-11054	137	11	the	the	DET
easat-11054	137	12	a3c	a3c	NUM
easat-11054	137	13	-	-	PUNCT
easat-11054	137	14	ffnn	ffnn	NOUN
easat-11054	137	15	and	and	CCONJ
easat-11054	137	16	dt	dt	ADP
easat-11054	137	17	algorithms	algorithm	NOUN
easat-11054	137	18	,	,	PUNCT
easat-11054	137	19	potentially	potentially	ADV
easat-11054	137	20	leading	lead	VERB
easat-11054	137	21	to	to	ADP
easat-11054	137	22	a	a	DET
easat-11054	137	23	generalized	generalized	ADJ
easat-11054	137	24	improvement	improvement	NOUN
easat-11054	137	25	in	in	ADP
easat-11054	137	26	intrusion	intrusion	NOUN
easat-11054	137	27	detection	detection	NOUN
easat-11054	137	28	prediction	prediction	NOUN
easat-11054	137	29	performance	performance	NOUN
easat-11054	137	30	due	due	ADP
easat-11054	137	31	to	to	ADP
easat-11054	137	32	its	its	PRON
easat-11054	137	33	ability	ability	NOUN
easat-11054	137	34	to	to	PART
easat-11054	137	35	capture	capture	VERB
easat-11054	137	36	complex	complex	ADJ
easat-11054	137	37	interactions	interaction	NOUN
easat-11054	137	38	between	between	ADP
easat-11054	137	39	dataset	dataset	NOUN
easat-11054	137	40	features	feature	NOUN
easat-11054	137	41	.	.	PUNCT
easat-11054	138	1	the	the	DET
easat-11054	138	2	hadt	hadt	NOUN
easat-11054	138	3	algorithm	algorithm	NOUN
easat-11054	138	4	inherently	inherently	ADV
easat-11054	138	5	uses	use	VERB
easat-11054	138	6	features	feature	NOUN
easat-11054	138	7	based	base	VERB
easat-11054	138	8	on	on	ADP
easat-11054	138	9	their	their	PRON
easat-11054	138	10	statistical	statistical	ADJ
easat-11054	138	11	characteristics	characteristic	NOUN
easat-11054	138	12	.	.	PUNCT
easat-11054	139	1	they	they	PRON
easat-11054	139	2	are	be	AUX
easat-11054	139	3	separated	separate	VERB
easat-11054	139	4	by	by	ADP
easat-11054	139	5	correlation	correlation	NOUN
easat-11054	139	6	analysis	analysis	NOUN
easat-11054	139	7	and	and	CCONJ
easat-11054	139	8	skewness	skewness	NOUN
easat-11054	139	9	testing	testing	NOUN
easat-11054	139	10	and	and	CCONJ
easat-11054	139	11	are	be	AUX
easat-11054	139	12	fed	feed	VERB
easat-11054	139	13	to	to	ADP
easat-11054	139	14	a3c	a3c	ADJ
easat-11054	139	15	-	-	PUNCT
easat-11054	139	16	ffnn	ffnn	NOUN
easat-11054	139	17	for	for	ADP
easat-11054	139	18	the	the	DET
easat-11054	139	19	continuously	continuously	ADV
easat-11054	139	20	distributed	distribute	VERB
easat-11054	139	21	and	and	CCONJ
easat-11054	139	22	high	high	ADJ
easat-11054	139	23	-	-	PUNCT
easat-11054	139	24	variance	variance	NOUN
easat-11054	139	25	features	feature	NOUN
easat-11054	139	26	and	and	CCONJ
easat-11054	139	27	to	to	PART
easat-11054	139	28	dt	dt	VERB
easat-11054	139	29	for	for	ADP
easat-11054	139	30	the	the	DET
easat-11054	139	31	discrete	discrete	ADJ
easat-11054	139	32	or	or	CCONJ
easat-11054	139	33	categorically	categorically	ADV
easat-11054	139	34	distributed	distribute	VERB
easat-11054	139	35	features	feature	NOUN
easat-11054	139	36	.	.	PUNCT
easat-11054	140	1	this	this	DET
easat-11054	140	2	separation	separation	NOUN
easat-11054	140	3	is	be	AUX
easat-11054	140	4	realized	realize	VERB
easat-11054	140	5	in	in	ADP
easat-11054	140	6	algorithm	algorithm	NOUN
easat-11054	140	7	1	1	NUM
easat-11054	140	8	to	to	PART
easat-11054	140	9	realize	realize	VERB
easat-11054	140	10	maximum	maximum	ADJ
easat-11054	140	11	featuretype	featuretype	NOUN
easat-11054	140	12	allocation	allocation	NOUN
easat-11054	140	13	for	for	ADP
easat-11054	140	14	enhanced	enhanced	ADJ
easat-11054	140	15	detection	detection	NOUN
easat-11054	140	16	accuracy	accuracy	NOUN
easat-11054	140	17	.	.	PUNCT
easat-11054	141	1	input	input	NOUN
easat-11054	141	2	:	:	PUNCT
easat-11054	141	3	feature	feature	NOUN
easat-11054	141	4	set	set	NOUN
easat-11054	141	5	(	(	PUNCT
easat-11054	141	6	f	f	PROPN
easat-11054	141	7	)	)	PUNCT
easat-11054	141	8	,	,	PUNCT
easat-11054	141	9	threshold	threshold	NOUN
easat-11054	141	10	(	(	PUNCT
easat-11054	141	11	τ	τ	PROPN
easat-11054	141	12	=	=	SYM
easat-11054	141	13	0.5	0.5	NUM
easat-11054	141	14	)	)	PUNCT
easat-11054	141	15	output	output	NOUN
easat-11054	141	16	:	:	PUNCT
easat-11054	141	17	continuous	continuous	ADJ
easat-11054	141	18	features	feature	NOUN
easat-11054	141	19	(	(	PUNCT
easat-11054	141	20	f_c	f_c	NUM
easat-11054	141	21	)	)	PUNCT
easat-11054	141	22	,	,	PUNCT
easat-11054	141	23	discrete	discrete	ADJ
easat-11054	141	24	features	feature	NOUN
easat-11054	141	25	(	(	PUNCT
easat-11054	141	26	f_d	f_d	SYM
easat-11054	141	27	)	)	PUNCT
easat-11054	142	1	1	1	X
easat-11054	142	2	.	.	X
easat-11054	143	1	for	for	ADP
easat-11054	143	2	each	each	DET
easat-11054	143	3	feature	feature	NOUN
easat-11054	143	4	(	(	PUNCT
easat-11054	143	5	f_i	f_i	X
easat-11054	143	6	)	)	PUNCT
easat-11054	143	7	in	in	ADP
easat-11054	143	8	(	(	PUNCT
easat-11054	143	9	f	f	PROPN
easat-11054	143	10	):	):	PUNCT
easat-11054	143	11	2	2	X
easat-11054	143	12	.	.	X
easat-11054	143	13	compute	compute	NOUN
easat-11054	143	14	skewness	skewness	NOUN
easat-11054	143	15	(	(	PUNCT
easat-11054	143	16	s_i	s_i	ADV
easat-11054	143	17	)	)	PUNCT
easat-11054	144	1	and	and	CCONJ
easat-11054	144	2	correlation	correlation	NOUN
easat-11054	144	3	(	(	PUNCT
easat-11054	144	4	c_i	c_i	SYM
easat-11054	144	5	)	)	PUNCT
easat-11054	145	1	3	3	X
easat-11054	145	2	..	..	PUNCT
easat-11054	145	3	if	if	SCONJ
easat-11054	145	4	(	(	PUNCT
easat-11054	145	5	|s_i|	|s_i|	NOUN
easat-11054	145	6	<	<	X
easat-11054	145	7	τ	τ	PROPN
easat-11054	145	8	)	)	PUNCT
easat-11054	145	9	and	and	CCONJ
easat-11054	145	10	(	(	PUNCT
easat-11054	145	11	c_i	c_i	X
easat-11054	145	12	>	>	X
easat-11054	145	13	τ	τ	PROPN
easat-11054	145	14	):	):	PUNCT
easat-11054	145	15	4	4	NUM
easat-11054	145	16	..	..	PUNCT
easat-11054	145	17	assign	assign	VERB
easat-11054	145	18	(	(	PUNCT
easat-11054	145	19	f_i	f_i	PROPN
easat-11054	145	20	)	)	PUNCT
easat-11054	145	21	to	to	ADP
easat-11054	145	22	(	(	PUNCT
easat-11054	145	23	f_c	f_c	PROPN
easat-11054	145	24	)	)	PUNCT
easat-11054	145	25	5	5	NUM
easat-11054	145	26	.	.	X
easat-11054	146	1	else	else	ADV
easat-11054	146	2	:	:	PUNCT
easat-11054	146	3	6	6	X
easat-11054	146	4	.	.	X
easat-11054	147	1	assign	assign	VERB
easat-11054	147	2	(	(	PUNCT
easat-11054	147	3	f_i	f_i	PROPN
easat-11054	147	4	)	)	PUNCT
easat-11054	147	5	to	to	ADP
easat-11054	147	6	(	(	PUNCT
easat-11054	147	7	f_d	f_d	SYM
easat-11054	147	8	)	)	PUNCT
easat-11054	147	9	7	7	X
easat-11054	147	10	.	.	X
easat-11054	147	11	return	return	NOUN
easat-11054	147	12	(	(	PUNCT
easat-11054	147	13	f_c	f_c	ADJ
easat-11054	147	14	,	,	PUNCT
easat-11054	147	15	f_d	f_d	NUM
easat-11054	147	16	)	)	PUNCT
easat-11054	147	17	algorithm	algorithm	NOUN
easat-11054	147	18	1	1	NUM
easat-11054	147	19	:	:	PUNCT
easat-11054	147	20	automatic	automatic	ADJ
easat-11054	147	21	feature	feature	NOUN
easat-11054	147	22	-	-	PUNCT
easat-11054	147	23	type	type	NOUN
easat-11054	147	24	separation	separation	NOUN
easat-11054	147	25	for	for	ADP
easat-11054	147	26	hadt	hadt	NOUN
easat-11054	147	27	1058	1058	NUM
easat-11054	147	28	edelweiss	edelweiss	PROPN
easat-11054	147	29	applied	apply	VERB
easat-11054	147	30	science	science	NOUN
easat-11054	147	31	and	and	CCONJ
easat-11054	147	32	technology	technology	NOUN
easat-11054	147	33	issn	issn	PROPN
easat-11054	147	34	:	:	PUNCT
easat-11054	147	35	2576	2576	NUM
easat-11054	147	36	-	-	SYM
easat-11054	147	37	8484	8484	NUM
easat-11054	147	38	vol	vol	NOUN
easat-11054	147	39	.	.	PROPN
easat-11054	148	1	9	9	NUM
easat-11054	148	2	,	,	PUNCT
easat-11054	148	3	no	no	INTJ
easat-11054	148	4	.	.	NOUN
easat-11054	148	5	11	11	NUM
easat-11054	148	6	:	:	PUNCT
easat-11054	148	7	1053	1053	NUM
easat-11054	148	8	-	-	SYM
easat-11054	148	9	1075	1075	NUM
easat-11054	148	10	,	,	PUNCT
easat-11054	148	11	2025	2025	NUM
easat-11054	148	12	doi	doi	NOUN
easat-11054	148	13	:	:	PUNCT
easat-11054	148	14	10.55214/2576	10.55214/2576	NUM
easat-11054	148	15	-	-	SYM
easat-11054	148	16	8484.v9i11.11054	8484.v9i11.11054	NOUN
easat-11054	149	1	©	©	PROPN
easat-11054	149	2	2025	2025	NUM
easat-11054	149	3	by	by	ADP
easat-11054	149	4	the	the	DET
easat-11054	149	5	authors	author	NOUN
easat-11054	149	6	;	;	PUNCT
easat-11054	149	7	licensee	licensee	PROPN
easat-11054	149	8	learning	learn	VERB
easat-11054	149	9	gate	gate	NOUN
easat-11054	149	10	figure	figure	NOUN
easat-11054	149	11	1	1	NUM
easat-11054	149	12	.	.	PUNCT
easat-11054	150	1	the	the	DET
easat-11054	150	2	structure	structure	NOUN
easat-11054	150	3	of	of	ADP
easat-11054	150	4	the	the	DET
easat-11054	150	5	hadt	hadt	NOUN
easat-11054	150	6	algorithm	algorithm	NOUN
easat-11054	150	7	.	.	PUNCT
easat-11054	151	1	the	the	DET
easat-11054	151	2	hadt	hadt	NOUN
easat-11054	151	3	algorithm	algorithm	NOUN
easat-11054	151	4	employs	employ	VERB
easat-11054	151	5	a	a	DET
easat-11054	151	6	limited	limited	ADJ
easat-11054	151	7	number	number	NOUN
easat-11054	151	8	of	of	ADP
easat-11054	151	9	neural	neural	ADJ
easat-11054	151	10	network	network	NOUN
easat-11054	151	11	layers	layer	NOUN
easat-11054	151	12	to	to	PART
easat-11054	151	13	strike	strike	VERB
easat-11054	151	14	a	a	DET
easat-11054	151	15	balance	balance	NOUN
easat-11054	151	16	between	between	ADP
easat-11054	151	17	detection	detection	NOUN
easat-11054	151	18	performance	performance	NOUN
easat-11054	151	19	and	and	CCONJ
easat-11054	151	20	computational	computational	ADJ
easat-11054	151	21	efficiency	efficiency	NOUN
easat-11054	151	22	.	.	PUNCT
easat-11054	152	1	this	this	DET
easat-11054	152	2	design	design	NOUN
easat-11054	152	3	choice	choice	NOUN
easat-11054	152	4	is	be	AUX
easat-11054	152	5	based	base	VERB
easat-11054	152	6	on	on	ADP
easat-11054	152	7	empirical	empirical	ADJ
easat-11054	152	8	analysis	analysis	NOUN
easat-11054	152	9	,	,	PUNCT
easat-11054	152	10	which	which	PRON
easat-11054	152	11	showed	show	VERB
easat-11054	152	12	that	that	SCONJ
easat-11054	152	13	increasing	increase	VERB
easat-11054	152	14	layers	layer	NOUN
easat-11054	152	15	beyond	beyond	ADP
easat-11054	152	16	this	this	DET
easat-11054	152	17	point	point	NOUN
easat-11054	152	18	yielded	yield	VERB
easat-11054	152	19	diminishing	diminishing	NOUN
easat-11054	152	20	returns	return	NOUN
easat-11054	152	21	in	in	ADP
easat-11054	152	22	accuracy	accuracy	NOUN
easat-11054	152	23	while	while	SCONJ
easat-11054	152	24	significantly	significantly	ADV
easat-11054	152	25	raising	raise	VERB
easat-11054	152	26	power	power	NOUN
easat-11054	152	27	consumption	consumption	NOUN
easat-11054	152	28	and	and	CCONJ
easat-11054	152	29	inference	inference	NOUN
easat-11054	152	30	time	time	NOUN
easat-11054	152	31	.	.	PUNCT
easat-11054	153	1	thus	thus	ADV
easat-11054	153	2	,	,	PUNCT
easat-11054	153	3	the	the	DET
easat-11054	153	4	selected	select	VERB
easat-11054	153	5	architecture	architecture	NOUN
easat-11054	153	6	is	be	AUX
easat-11054	153	7	optimal	optimal	ADJ
easat-11054	153	8	for	for	ADP
easat-11054	153	9	real	real	ADJ
easat-11054	153	10	-	-	PUNCT
easat-11054	153	11	time	time	NOUN
easat-11054	153	12	iiot	iiot	NOUN
easat-11054	153	13	intrusion	intrusion	NOUN
easat-11054	153	14	detection	detection	NOUN
easat-11054	153	15	at	at	ADP
easat-11054	153	16	the	the	DET
easat-11054	153	17	edge	edge	NOUN
easat-11054	153	18	.	.	PUNCT
easat-11054	154	1	the	the	DET
easat-11054	154	2	detailed	detailed	ADJ
easat-11054	154	3	structure	structure	NOUN
easat-11054	154	4	of	of	ADP
easat-11054	154	5	the	the	DET
easat-11054	154	6	hadt	hadt	NOUN
easat-11054	154	7	algorithm	algorithm	NOUN
easat-11054	154	8	is	be	AUX
easat-11054	154	9	shown	show	VERB
easat-11054	154	10	in	in	ADP
easat-11054	154	11	figure	figure	NOUN
easat-11054	154	12	2	2	NUM
easat-11054	154	13	,	,	PUNCT
easat-11054	154	14	which	which	PRON
easat-11054	154	15	helps	help	VERB
easat-11054	154	16	to	to	PART
easat-11054	154	17	improve	improve	VERB
easat-11054	154	18	the	the	DET
easat-11054	154	19	results	result	NOUN
easat-11054	154	20	of	of	ADP
easat-11054	154	21	the	the	DET
easat-11054	154	22	predicted	predict	VERB
easat-11054	154	23	intrusion	intrusion	NOUN
easat-11054	154	24	class	class	NOUN
easat-11054	154	25	‘	'	PUNCT
easat-11054	154	26	actions	action	NOUN
easat-11054	154	27	’	'	PUNCT
easat-11054	154	28	and	and	CCONJ
easat-11054	154	29	to	to	PART
easat-11054	154	30	obtain	obtain	VERB
easat-11054	154	31	the	the	DET
easat-11054	154	32	best	good	ADJ
easat-11054	154	33	accuracy	accuracy	NOUN
easat-11054	154	34	.	.	PUNCT
easat-11054	155	1	figure	figure	NOUN
easat-11054	155	2	2	2	NUM
easat-11054	155	3	.	.	PUNCT
easat-11054	156	1	the	the	DET
easat-11054	156	2	layers	layer	NOUN
easat-11054	156	3	of	of	ADP
easat-11054	156	4	the	the	DET
easat-11054	156	5	proposed	propose	VERB
easat-11054	156	6	hadt	hadt	NOUN
easat-11054	156	7	algorithm	algorithm	NOUN
easat-11054	156	8	.	.	PUNCT
easat-11054	157	1	the	the	DET
easat-11054	157	2	actor	actor	NOUN
easat-11054	157	3	in	in	ADP
easat-11054	157	4	a3c	a3c	VERB
easat-11054	157	5	-	-	PUNCT
easat-11054	157	6	ffnn	ffnn	NOUN
easat-11054	157	7	seeks	seek	VERB
easat-11054	157	8	to	to	PART
easat-11054	157	9	predict	predict	VERB
easat-11054	157	10	a	a	DET
easat-11054	157	11	policy	policy	NOUN
easat-11054	157	12	π(s	π(s	NOUN
easat-11054	157	13	)	)	PUNCT
easat-11054	157	14	by	by	ADP
easat-11054	157	15	getting	get	VERB
easat-11054	157	16	criticism	criticism	NOUN
easat-11054	157	17	from	from	ADP
easat-11054	157	18	an	an	DET
easat-11054	157	19	advantage	advantage	NOUN
easat-11054	157	20	function	function	NOUN
easat-11054	157	21	.	.	PUNCT
easat-11054	158	1	the	the	DET
easat-11054	158	2	expected	expect	VERB
easat-11054	158	3	return	return	NOUN
easat-11054	158	4	in	in	ADP
easat-11054	158	5	rewards	reward	NOUN
easat-11054	158	6	,	,	PUNCT
easat-11054	158	7	or	or	CCONJ
easat-11054	158	8	value	value	NOUN
easat-11054	158	9	function	function	NOUN
easat-11054	158	10	v(s	v(s	PROPN
easat-11054	158	11	)	)	PUNCT
easat-11054	158	12	,	,	PUNCT
easat-11054	158	13	is	be	AUX
easat-11054	158	14	taught	teach	VERB
easat-11054	158	15	to	to	ADP
easat-11054	158	16	the	the	DET
easat-11054	158	17	critic	critic	NOUN
easat-11054	158	18	and	and	CCONJ
easat-11054	158	19	is	be	AUX
easat-11054	158	20	used	use	VERB
easat-11054	158	21	to	to	PART
easat-11054	158	22	calculate	calculate	VERB
easat-11054	158	23	the	the	DET
easat-11054	158	24	relative	relative	ADJ
easat-11054	158	25	advantages	advantage	NOUN
easat-11054	158	26	[	[	X
easat-11054	158	27	22	22	NUM
easat-11054	158	28	]	]	PUNCT
easat-11054	158	29	.	.	PUNCT
easat-11054	159	1	if	if	SCONJ
easat-11054	159	2	s	s	PROPN
easat-11054	159	3	is	be	AUX
easat-11054	159	4	the	the	DET
easat-11054	159	5	current	current	ADJ
easat-11054	159	6	state	state	NOUN
easat-11054	159	7	,	,	PUNCT
easat-11054	159	8	s	s	AUX
easat-11054	159	9	'	'	PUNCT
easat-11054	159	10	is	be	AUX
easat-11054	159	11	the	the	DET
easat-11054	159	12	next	next	ADJ
easat-11054	159	13	state	state	NOUN
easat-11054	159	14	,	,	PUNCT
easat-11054	159	15	r	r	NOUN
easat-11054	159	16	is	be	AUX
easat-11054	159	17	the	the	DET
easat-11054	159	18	current	current	ADJ
easat-11054	159	19	reward	reward	NOUN
easat-11054	159	20	,	,	PUNCT
easat-11054	159	21	γ	γ	X
easat-11054	159	22	is	be	AUX
easat-11054	159	23	the	the	DET
easat-11054	159	24	discount	discount	NOUN
easat-11054	159	25	factor	factor	NOUN
easat-11054	159	26	,	,	PUNCT
easat-11054	159	27	and	and	CCONJ
easat-11054	159	28	v(s	v(s	NUM
easat-11054	159	29	)	)	PUNCT
easat-11054	159	30	is	be	AUX
easat-11054	159	31	the	the	DET
easat-11054	159	32	value	value	NOUN
easat-11054	159	33	function	function	NOUN
easat-11054	159	34	,	,	PUNCT
easat-11054	159	35	then	then	ADV
easat-11054	159	36	the	the	DET
easat-11054	159	37	advantage	advantage	NOUN
easat-11054	159	38	is	be	AUX
easat-11054	159	39	defined	define	VERB
easat-11054	159	40	as	as	ADP
easat-11054	159	41	a(s	a(	NOUN
easat-11054	159	42	)	)	PUNCT
easat-11054	159	43	=	=	PUNCT
easat-11054	159	44	r+γv(s	r+γv(s	NOUN
easat-11054	159	45	’	'	PUNCT
easat-11054	159	46	)	)	PUNCT
easat-11054	159	47	v(s	v(s	PROPN
easat-11054	159	48	)	)	PUNCT
easat-11054	159	49	.	.	PUNCT
easat-11054	160	1	the	the	DET
easat-11054	160	2	expected	expect	VERB
easat-11054	160	3	value	value	NOUN
easat-11054	160	4	or	or	CCONJ
easat-11054	160	5	reward	reward	NOUN
easat-11054	160	6	associated	associate	VERB
easat-11054	160	7	with	with	ADP
easat-11054	160	8	a	a	DET
easat-11054	160	9	specific	specific	ADJ
easat-11054	160	10	condition	condition	NOUN
easat-11054	160	11	is	be	AUX
easat-11054	160	12	predicted	predict	VERB
easat-11054	160	13	using	use	VERB
easat-11054	160	14	the	the	DET
easat-11054	160	15	value	value	NOUN
easat-11054	160	16	1059	1059	NUM
easat-11054	160	17	edelweiss	edelweiss	PROPN
easat-11054	160	18	applied	apply	VERB
easat-11054	160	19	science	science	NOUN
easat-11054	160	20	and	and	CCONJ
easat-11054	160	21	technology	technology	NOUN
easat-11054	160	22	issn	issn	PROPN
easat-11054	160	23	:	:	PUNCT
easat-11054	160	24	2576	2576	NUM
easat-11054	160	25	-	-	SYM
easat-11054	160	26	8484	8484	NUM
easat-11054	160	27	vol	vol	NOUN
easat-11054	160	28	.	.	PROPN
easat-11054	161	1	9	9	NUM
easat-11054	161	2	,	,	PUNCT
easat-11054	161	3	no	no	INTJ
easat-11054	161	4	.	.	NOUN
easat-11054	161	5	11	11	NUM
easat-11054	161	6	:	:	PUNCT
easat-11054	161	7	1053	1053	NUM
easat-11054	161	8	-	-	SYM
easat-11054	161	9	1075	1075	NUM
easat-11054	161	10	,	,	PUNCT
easat-11054	161	11	2025	2025	NUM
easat-11054	161	12	doi	doi	NOUN
easat-11054	161	13	:	:	PUNCT
easat-11054	161	14	10.55214/2576	10.55214/2576	NUM
easat-11054	161	15	-	-	SYM
easat-11054	161	16	8484.v9i11.11054	8484.v9i11.11054	NOUN
easat-11054	162	1	©	©	PROPN
easat-11054	162	2	2025	2025	NUM
easat-11054	162	3	by	by	ADP
easat-11054	162	4	the	the	DET
easat-11054	162	5	authors	author	NOUN
easat-11054	162	6	;	;	PUNCT
easat-11054	162	7	licensee	licensee	PROPN
easat-11054	162	8	learning	learn	VERB
easat-11054	162	9	gate	gate	NOUN
easat-11054	162	10	function	function	PROPN
easat-11054	162	11	.	.	PUNCT
easat-11054	163	1	a	a	DET
easat-11054	163	2	linear	linear	ADJ
easat-11054	163	3	activation	activation	NOUN
easat-11054	163	4	function	function	NOUN
easat-11054	163	5	is	be	AUX
easat-11054	163	6	employed	employ	VERB
easat-11054	163	7	to	to	PART
easat-11054	163	8	obtain	obtain	VERB
easat-11054	163	9	the	the	DET
easat-11054	163	10	expected	expect	VERB
easat-11054	163	11	value	value	NOUN
easat-11054	163	12	,	,	PUNCT
easat-11054	163	13	as	as	SCONJ
easat-11054	163	14	shown	show	VERB
easat-11054	163	15	in	in	ADP
easat-11054	163	16	equation	equation	NOUN
easat-11054	163	17	(	(	PUNCT
easat-11054	163	18	1	1	NUM
easat-11054	163	19	)	)	PUNCT
easat-11054	163	20	.	.	PUNCT
easat-11054	164	1	(	(	PUNCT
easat-11054	164	2	)	)	PUNCT
easat-11054	164	3	(	(	PUNCT
easat-11054	164	4	)	)	PUNCT
easat-11054	164	5	;	;	PUNCT
easat-11054	164	6	·	·	PUNCT
easat-11054	164	7	v	v	X
easat-11054	164	8	s	s	X
easat-11054	164	9	w	w	PROPN
easat-11054	164	10	s	s	PROPN
easat-11054	164	11	=	=	PROPN
easat-11054	164	12	(	(	PUNCT
easat-11054	164	13	1	1	NUM
easat-11054	164	14	)	)	PUNCT
easat-11054	164	15	where	where	SCONJ
easat-11054	164	16	w	w	NOUN
easat-11054	164	17	is	be	AUX
easat-11054	164	18	the	the	DET
easat-11054	164	19	weight	weight	NOUN
easat-11054	164	20	matrix	matrix	NOUN
easat-11054	164	21	connecting	connect	VERB
easat-11054	164	22	the	the	DET
easat-11054	164	23	hidden	hidden	ADJ
easat-11054	164	24	layer	layer	NOUN
easat-11054	164	25	to	to	ADP
easat-11054	164	26	the	the	DET
easat-11054	164	27	value	value	NOUN
easat-11054	164	28	output	output	NOUN
easat-11054	164	29	,	,	PUNCT
easat-11054	164	30	and	and	CCONJ
easat-11054	164	31	v	v	NOUN
easat-11054	164	32	is	be	AUX
easat-11054	164	33	the	the	DET
easat-11054	164	34	predicted	predict	VERB
easat-11054	164	35	value	value	NOUN
easat-11054	164	36	of	of	ADP
easat-11054	164	37	states	state	NOUN
easat-11054	164	38	s	s	AUX
easat-11054	164	39	given	give	VERB
easat-11054	164	40	the	the	DET
easat-11054	164	41	ffnn	ffnn	NOUN
easat-11054	164	42	layer	layer	NOUN
easat-11054	164	43	's	's	PART
easat-11054	164	44	parameters	parameter	NOUN
easat-11054	164	45	θ	θ	PROPN
easat-11054	164	46	.	.	PUNCT
easat-11054	164	47	given	give	VERB
easat-11054	164	48	a	a	DET
easat-11054	164	49	state	state	NOUN
easat-11054	164	50	,	,	PUNCT
easat-11054	164	51	the	the	DET
easat-11054	164	52	policy	policy	NOUN
easat-11054	164	53	function	function	NOUN
easat-11054	164	54	predicts	predict	VERB
easat-11054	164	55	the	the	DET
easat-11054	164	56	probability	probability	NOUN
easat-11054	164	57	distribution	distribution	NOUN
easat-11054	164	58	across	across	ADP
easat-11054	164	59	possible	possible	ADJ
easat-11054	164	60	actions	action	NOUN
easat-11054	164	61	.	.	PUNCT
easat-11054	165	1	ordinarily	ordinarily	ADV
easat-11054	165	2	,	,	PUNCT
easat-11054	165	3	a	a	DET
easat-11054	165	4	softmax	softmax	NOUN
easat-11054	165	5	activation	activation	NOUN
easat-11054	165	6	function	function	NOUN
easat-11054	165	7	is	be	AUX
easat-11054	165	8	used	use	VERB
easat-11054	165	9	in	in	ADP
easat-11054	165	10	equation	equation	NOUN
easat-11054	165	11	(	(	PUNCT
easat-11054	165	12	2	2	NUM
easat-11054	165	13	)	)	PUNCT
easat-11054	165	14	to	to	PART
easat-11054	165	15	calculate	calculate	VERB
easat-11054	165	16	the	the	DET
easat-11054	165	17	probabilities	probability	NOUN
easat-11054	165	18	of	of	ADP
easat-11054	165	19	actions	action	NOUN
easat-11054	165	20	.	.	PUNCT
easat-11054	166	1	(	(	PUNCT
easat-11054	166	2	)	)	PUNCT
easat-11054	166	3	(	(	PUNCT
easat-11054	166	4	)	)	PUNCT
easat-11054	166	5	(	(	PUNCT
easat-11054	166	6	)	)	PUNCT
easat-11054	166	7	;	;	PUNCT
easat-11054	166	8	|	|	ADV
easat-11054	166	9	·	·	PUNCT
easat-11054	166	10	a	a	DET
easat-11054	166	11	s	s	X
easat-11054	166	12	softmax	softmax	NOUN
easat-11054	166	13	w	w	PROPN
easat-11054	166	14	s	s	PROPN
easat-11054	166	15			PROPN
easat-11054	166	16	=	=	NUM
easat-11054	167	1	(	(	PUNCT
easat-11054	167	2	2	2	NUM
easat-11054	167	3	)	)	PUNCT
easat-11054	167	4	here	here	ADV
easat-11054	167	5	,	,	PUNCT
easat-11054	167	6	π(a|s;θ	π(a|s;θ	PROPN
easat-11054	167	7	)	)	PUNCT
easat-11054	167	8	is	be	AUX
easat-11054	167	9	the	the	DET
easat-11054	167	10	probability	probability	NOUN
easat-11054	167	11	of	of	ADP
easat-11054	167	12	taking	take	VERB
easat-11054	167	13	action	action	NOUN
easat-11054	167	14	a	a	PRON
easat-11054	167	15	and	and	CCONJ
easat-11054	167	16	φ(s	φ(s	NOUN
easat-11054	167	17	)	)	PUNCT
easat-11054	167	18	is	be	AUX
easat-11054	167	19	the	the	DET
easat-11054	167	20	hidden	hide	VERB
easat-11054	167	21	layer	layer	NOUN
easat-11054	167	22	's	's	PART
easat-11054	167	23	features	feature	NOUN
easat-11054	167	24	,	,	PUNCT
easat-11054	167	25	or	or	CCONJ
easat-11054	167	26	output	output	NOUN
easat-11054	167	27	for	for	ADP
easat-11054	167	28	state	state	NOUN
easat-11054	167	29	s.	s.	PROPN
easat-11054	167	30	in	in	ADP
easat-11054	167	31	equation	equation	NOUN
easat-11054	167	32	(	(	PUNCT
easat-11054	167	33	3	3	NUM
easat-11054	167	34	)	)	PUNCT
easat-11054	167	35	,	,	PUNCT
easat-11054	167	36	the	the	DET
easat-11054	167	37	loss	loss	NOUN
easat-11054	167	38	function	function	NOUN
easat-11054	167	39	is	be	AUX
easat-11054	167	40	formulated	formulate	VERB
easat-11054	167	41	based	base	VERB
easat-11054	167	42	on	on	ADP
easat-11054	167	43	the	the	DET
easat-11054	167	44	difference	difference	NOUN
easat-11054	167	45	between	between	ADP
easat-11054	167	46	the	the	DET
easat-11054	167	47	actual	actual	ADJ
easat-11054	167	48	and	and	CCONJ
easat-11054	167	49	predicted	predict	VERB
easat-11054	167	50	values	value	NOUN
easat-11054	167	51	[	[	X
easat-11054	167	52	23	23	NUM
easat-11054	167	53	]	]	PUNCT
easat-11054	167	54	.	.	PUNCT
easat-11054	168	1	specifically	specifically	ADV
easat-11054	168	2	,	,	PUNCT
easat-11054	168	3	ri	ri	PROPN
easat-11054	168	4	represents	represent	VERB
easat-11054	168	5	the	the	DET
easat-11054	168	6	actual	actual	ADJ
easat-11054	168	7	return	return	NOUN
easat-11054	168	8	,	,	PUNCT
easat-11054	168	9	which	which	PRON
easat-11054	168	10	is	be	AUX
easat-11054	168	11	computed	compute	VERB
easat-11054	168	12	from	from	ADP
easat-11054	168	13	the	the	DET
easat-11054	168	14	observed	observe	VERB
easat-11054	168	15	rewards	reward	NOUN
easat-11054	168	16	in	in	ADP
easat-11054	168	17	the	the	DET
easat-11054	168	18	iiot	iiot	NOUN
easat-11054	168	19	environment	environment	NOUN
easat-11054	168	20	.	.	PUNCT
easat-11054	169	1	the	the	DET
easat-11054	169	2	predicted	predict	VERB
easat-11054	169	3	state	state	NOUN
easat-11054	169	4	-	-	PUNCT
easat-11054	169	5	value	value	NOUN
easat-11054	169	6	function	function	NOUN
easat-11054	169	7	,	,	PUNCT
easat-11054	169	8	denoted	denote	VERB
easat-11054	169	9	as	as	ADP
easat-11054	169	10	v(si;θ	v(si;θ	NOUN
easat-11054	169	11	)	)	PUNCT
easat-11054	169	12	,	,	PUNCT
easat-11054	169	13	estimates	estimate	VERB
easat-11054	169	14	the	the	DET
easat-11054	169	15	expected	expect	VERB
easat-11054	169	16	return	return	NOUN
easat-11054	169	17	given	give	VERB
easat-11054	169	18	si	si	PROPN
easat-11054	169	19	and	and	CCONJ
easat-11054	169	20	θ	θ	PROPN
easat-11054	169	21	.	.	PUNCT
easat-11054	170	1	the	the	DET
easat-11054	170	2	loss	loss	NOUN
easat-11054	170	3	is	be	AUX
easat-11054	170	4	computed	compute	VERB
easat-11054	170	5	over	over	ADP
easat-11054	170	6	a	a	DET
easat-11054	170	7	batch	batch	NOUN
easat-11054	170	8	of	of	ADP
easat-11054	170	9	n	n	DET
easat-11054	170	10	samples	sample	NOUN
easat-11054	170	11	,	,	PUNCT
easat-11054	170	12	where	where	SCONJ
easat-11054	170	13	n	n	PRON
easat-11054	170	14	represents	represent	VERB
easat-11054	170	15	the	the	DET
easat-11054	170	16	number	number	NOUN
easat-11054	170	17	of	of	ADP
easat-11054	170	18	s	s	PRON
easat-11054	170	19	used	use	VERB
easat-11054	170	20	in	in	ADP
easat-11054	170	21	each	each	DET
easat-11054	170	22	training	training	NOUN
easat-11054	170	23	iteration	iteration	NOUN
easat-11054	170	24	.	.	PUNCT
easat-11054	171	1	this	this	DET
easat-11054	171	2	formulation	formulation	NOUN
easat-11054	171	3	ensures	ensure	VERB
easat-11054	171	4	that	that	SCONJ
easat-11054	171	5	the	the	DET
easat-11054	171	6	ffnn	ffnn	NOUN
easat-11054	171	7	layer	layer	NOUN
easat-11054	171	8	in	in	ADP
easat-11054	171	9	a3cffnn	a3cffnn	NOUN
easat-11054	171	10	effectively	effectively	ADV
easat-11054	171	11	learns	learn	VERB
easat-11054	171	12	to	to	PART
easat-11054	171	13	minimize	minimize	VERB
easat-11054	171	14	the	the	DET
easat-11054	171	15	divergence	divergence	NOUN
easat-11054	171	16	between	between	ADP
easat-11054	171	17	predicted	predict	VERB
easat-11054	171	18	and	and	CCONJ
easat-11054	171	19	actual	actual	ADJ
easat-11054	171	20	returns	return	NOUN
easat-11054	171	21	,	,	PUNCT
easat-11054	171	22	enhancing	enhance	VERB
easat-11054	171	23	the	the	DET
easat-11054	171	24	accuracy	accuracy	NOUN
easat-11054	171	25	of	of	ADP
easat-11054	171	26	the	the	DET
easat-11054	171	27	rl	rl	NOUN
easat-11054	171	28	process	process	NOUN
easat-11054	171	29	.	.	PUNCT
easat-11054	172	1	(	(	PUNCT
easat-11054	172	2	)	)	PUNCT
easat-11054	172	3	(	(	PUNCT
easat-11054	172	4	)	)	PUNCT
easat-11054	172	5	1	1	NUM
easat-11054	172	6	1	1	NUM
easat-11054	172	7	;	;	PUNCT
easat-11054	172	8	²	²	NUM
easat-11054	172	9	n	n	NUM
easat-11054	172	10	v	v	VERB
easat-11054	173	1	i	i	PRON
easat-11054	173	2	i	i	PRON
easat-11054	174	1	i	i	VERB
easat-11054	174	2	l	l	NOUN
easat-11054	175	1	r	r	NOUN
easat-11054	175	2	v	v	NOUN
easat-11054	175	3	s	s	NOUN
easat-11054	175	4	n	n	NOUN
easat-11054	175	5			X
easat-11054	176	1	=	=	PUNCT
easat-11054	176	2	=	=	SYM
easat-11054	176	3	−	−	X
easat-11054	176	4	(	(	PUNCT
easat-11054	176	5	3	3	NUM
easat-11054	176	6	)	)	PUNCT
easat-11054	176	7	from	from	ADP
easat-11054	176	8	equation	equation	NOUN
easat-11054	176	9	(	(	PUNCT
easat-11054	176	10	4	4	NUM
easat-11054	176	11	)	)	PUNCT
easat-11054	176	12	,	,	PUNCT
easat-11054	176	13	the	the	DET
easat-11054	176	14	policy	policy	NOUN
easat-11054	176	15	loss	loss	NOUN
easat-11054	176	16	function	function	NOUN
easat-11054	176	17	is	be	AUX
easat-11054	176	18	a	a	DET
easat-11054	176	19	key	key	ADJ
easat-11054	176	20	component	component	NOUN
easat-11054	176	21	in	in	ADP
easat-11054	176	22	policy	policy	NOUN
easat-11054	176	23	gradient	gradient	NOUN
easat-11054	176	24	methods	method	NOUN
easat-11054	176	25	.	.	PUNCT
easat-11054	177	1	this	this	DET
easat-11054	177	2	equation	equation	NOUN
easat-11054	177	3	measures	measure	VERB
easat-11054	177	4	how	how	SCONJ
easat-11054	177	5	well	well	ADV
easat-11054	177	6	the	the	DET
easat-11054	177	7	policy	policy	NOUN
easat-11054	177	8	π	π	NOUN
easat-11054	177	9	selects	select	VERB
easat-11054	177	10	actions	action	NOUN
easat-11054	177	11	that	that	PRON
easat-11054	177	12	lead	lead	VERB
easat-11054	177	13	to	to	ADP
easat-11054	177	14	higher	high	ADJ
easat-11054	177	15	rewards	reward	NOUN
easat-11054	177	16	[	[	X
easat-11054	177	17	24	24	NUM
easat-11054	177	18	]	]	PUNCT
easat-11054	177	19	.	.	PUNCT
easat-11054	178	1	the	the	DET
easat-11054	178	2	expectation	expectation	NOUN
easat-11054	178	3	operator	operator	NOUN
easat-11054	178	4	e	e	NOUN
easat-11054	178	5	represents	represent	VERB
easat-11054	178	6	the	the	DET
easat-11054	178	7	average	average	NOUN
easat-11054	178	8	over	over	ADP
easat-11054	178	9	multiple	multiple	ADJ
easat-11054	178	10	sampled	sample	VERB
easat-11054	178	11	states	state	NOUN
easat-11054	178	12	and	and	CCONJ
easat-11054	178	13	actions	action	NOUN
easat-11054	178	14	.	.	PUNCT
easat-11054	179	1	the	the	DET
easat-11054	179	2	term	term	NOUN
easat-11054	179	3	a(s	a(s	PROPN
easat-11054	179	4	,	,	PUNCT
easat-11054	179	5	a	a	PRON
easat-11054	179	6	)	)	PUNCT
easat-11054	179	7	is	be	AUX
easat-11054	179	8	the	the	DET
easat-11054	179	9	advantage	advantage	NOUN
easat-11054	179	10	function	function	NOUN
easat-11054	179	11	,	,	PUNCT
easat-11054	179	12	which	which	PRON
easat-11054	179	13	indicates	indicate	VERB
easat-11054	179	14	how	how	SCONJ
easat-11054	179	15	much	much	ADV
easat-11054	179	16	better	well	ADJ
easat-11054	179	17	or	or	CCONJ
easat-11054	179	18	worse	bad	ADJ
easat-11054	179	19	a	a	DET
easat-11054	179	20	specific	specific	ADJ
easat-11054	179	21	action	action	NOUN
easat-11054	179	22	a	a	PRON
easat-11054	179	23	is	be	AUX
easat-11054	179	24	compared	compare	VERB
easat-11054	179	25	to	to	ADP
easat-11054	179	26	the	the	DET
easat-11054	179	27	average	average	NOUN
easat-11054	179	28	a	a	PRON
easat-11054	179	29	at	at	ADP
easat-11054	179	30	s.	s.	PROPN
easat-11054	179	31	the	the	DET
easat-11054	179	32	logarithm	logarithm	NOUN
easat-11054	179	33	of	of	ADP
easat-11054	179	34	the	the	DET
easat-11054	179	35	policy	policy	NOUN
easat-11054	179	36	function	function	NOUN
easat-11054	179	37	ensures	ensure	VERB
easat-11054	179	38	that	that	SCONJ
easat-11054	179	39	the	the	DET
easat-11054	179	40	ffnn	ffnn	NOUN
easat-11054	179	41	layer	layer	NOUN
easat-11054	179	42	in	in	ADP
easat-11054	179	43	the	the	DET
easat-11054	179	44	a3c	a3c	NUM
easat-11054	179	45	worker	worker	NOUN
easat-11054	179	46	updates	update	VERB
easat-11054	179	47	in	in	ADP
easat-11054	179	48	proportion	proportion	NOUN
easat-11054	179	49	to	to	ADP
easat-11054	179	50	how	how	SCONJ
easat-11054	179	51	confident	confident	ADJ
easat-11054	179	52	it	it	PRON
easat-11054	179	53	is	be	AUX
easat-11054	179	54	in	in	ADP
easat-11054	179	55	selecting	select	VERB
easat-11054	179	56	a.	a.	NOUN
easat-11054	179	57	the	the	DET
easat-11054	179	58	negative	negative	ADJ
easat-11054	179	59	sign	sign	NOUN
easat-11054	179	60	in	in	ADP
easat-11054	179	61	the	the	DET
easat-11054	179	62	function	function	NOUN
easat-11054	179	63	is	be	AUX
easat-11054	179	64	necessary	necessary	ADJ
easat-11054	179	65	because	because	SCONJ
easat-11054	179	66	optimization	optimization	NOUN
easat-11054	179	67	algorithms	algorithm	NOUN
easat-11054	179	68	typically	typically	ADV
easat-11054	179	69	perform	perform	VERB
easat-11054	179	70	gradient	gradient	ADJ
easat-11054	179	71	descent	descent	NOUN
easat-11054	179	72	,	,	PUNCT
easat-11054	179	73	which	which	PRON
easat-11054	179	74	minimizes	minimize	VERB
easat-11054	179	75	loss	loss	NOUN
easat-11054	179	76	functions	function	NOUN
easat-11054	179	77	.	.	PUNCT
easat-11054	180	1	by	by	ADP
easat-11054	180	2	minimizing	minimize	VERB
easat-11054	180	3	this	this	DET
easat-11054	180	4	loss	loss	NOUN
easat-11054	180	5	,	,	PUNCT
easat-11054	180	6	the	the	DET
easat-11054	180	7	algorithm	algorithm	NOUN
easat-11054	180	8	maximizes	maximize	VERB
easat-11054	180	9	the	the	DET
easat-11054	180	10	probability	probability	NOUN
easat-11054	180	11	of	of	ADP
easat-11054	180	12	choosing	choose	VERB
easat-11054	180	13	advantageous	advantageous	ADJ
easat-11054	180	14	actions	action	NOUN
easat-11054	180	15	while	while	SCONJ
easat-11054	180	16	reducing	reduce	VERB
easat-11054	180	17	the	the	DET
easat-11054	180	18	probability	probability	NOUN
easat-11054	180	19	of	of	ADP
easat-11054	180	20	selecting	select	VERB
easat-11054	180	21	suboptimal	suboptimal	ADJ
easat-11054	180	22	ones	one	NOUN
easat-11054	180	23	,	,	PUNCT
easat-11054	180	24	leading	lead	VERB
easat-11054	180	25	to	to	ADP
easat-11054	180	26	an	an	DET
easat-11054	180	27	improved	improved	ADJ
easat-11054	180	28	policy	policy	NOUN
easat-11054	180	29	over	over	ADP
easat-11054	180	30	time	time	NOUN
easat-11054	180	31	.	.	PUNCT
easat-11054	181	1	(	(	PUNCT
easat-11054	181	2	)	)	PUNCT
easat-11054	181	3	(	(	PUNCT
easat-11054	181	4	)	)	PUNCT
easat-11054	181	5	(	(	PUNCT
easat-11054	181	6	)	)	PUNCT
easat-11054	181	7	;	;	PUNCT
easat-11054	181	8	|	|	ADV
easat-11054	181	9	)	)	PUNCT
easat-11054	181	10	(	(	PUNCT
easat-11054	181	11	pl	pl	PROPN
easat-11054	181	12	e	e	PROPN
easat-11054	181	13	a	a	DET
easat-11054	181	14	s	s	X
easat-11054	181	15	log	log	NOUN
easat-11054	181	16	a	a	DET
easat-11054	181	17	s	s	NOUN
easat-11054	181	18	=	=	NOUN
easat-11054	181	19	−	−	PROPN
easat-11054	181	20	(	(	PUNCT
easat-11054	181	21	4	4	NUM
easat-11054	181	22	)	)	PUNCT
easat-11054	181	23	thus	thus	ADV
easat-11054	181	24	,	,	PUNCT
easat-11054	181	25	it	it	PRON
easat-11054	181	26	is	be	AUX
easat-11054	181	27	crucial	crucial	ADJ
easat-11054	181	28	to	to	PART
easat-11054	181	29	include	include	VERB
easat-11054	181	30	the	the	DET
easat-11054	181	31	entropy	entropy	NOUN
easat-11054	181	32	h(π	h(π	PROPN
easat-11054	181	33	)	)	PUNCT
easat-11054	181	34	in	in	ADP
easat-11054	181	35	the	the	DET
easat-11054	181	36	loss	loss	NOUN
easat-11054	181	37	function	function	NOUN
easat-11054	181	38	equations	equation	NOUN
easat-11054	181	39	as	as	SCONJ
easat-11054	181	40	expressed	express	VERB
easat-11054	181	41	in	in	ADP
easat-11054	181	42	equation	equation	NOUN
easat-11054	181	43	(	(	PUNCT
easat-11054	181	44	5	5	NUM
easat-11054	181	45	)	)	PUNCT
easat-11054	181	46	.	.	PUNCT
easat-11054	182	1	(	(	PUNCT
easat-11054	182	2	)	)	PUNCT
easat-11054	182	3	(	(	PUNCT
easat-11054	182	4	)	)	PUNCT
easat-11054	182	5	(	(	PUNCT
easat-11054	182	6	)	)	PUNCT
easat-11054	182	7	(	(	PUNCT
easat-11054	182	8	)	)	PUNCT
easat-11054	182	9	;	;	PUNCT
easat-11054	182	10	|	|	ADV
easat-11054	182	11	|	|	ADV
easat-11054	182	12	;	;	PUNCT
easat-11054	182	13	h	h	PRON
easat-11054	183	1	a	a	DET
easat-11054	183	2	s	s	X
easat-11054	183	3	log	log	NOUN
easat-11054	183	4	a	a	DET
easat-11054	183	5	s	s	PROPN
easat-11054	183	6			PROPN
easat-11054	183	7			X
easat-11054	183	8			X
easat-11054	183	9	=	=	PROPN
easat-11054	183	10	−	−	X
easat-11054	183	11			PROPN
easat-11054	183	12			ADJ
easat-11054	183	13			PROPN
easat-11054	183	14			PROPN
easat-11054	183	15	(	(	PUNCT
easat-11054	183	16	5	5	NUM
easat-11054	183	17	)	)	PUNCT
easat-11054	183	18	a	a	DET
easat-11054	183	19	measure	measure	NOUN
easat-11054	183	20	of	of	ADP
easat-11054	183	21	the	the	DET
easat-11054	183	22	dispersion	dispersion	NOUN
easat-11054	183	23	of	of	ADP
easat-11054	183	24	probability	probability	NOUN
easat-11054	183	25	is	be	AUX
easat-11054	183	26	called	call	VERB
easat-11054	183	27	entropy	entropy	NOUN
easat-11054	183	28	.	.	PUNCT
easat-11054	184	1	the	the	PRON
easat-11054	184	2	higher	high	ADJ
easat-11054	184	3	the	the	DET
easat-11054	184	4	entropy	entropy	NOUN
easat-11054	184	5	,	,	PUNCT
easat-11054	184	6	the	the	DET
easat-11054	184	7	more	more	ADV
easat-11054	184	8	similar	similar	ADJ
easat-11054	184	9	the	the	DET
easat-11054	184	10	likelihood	likelihood	NOUN
easat-11054	184	11	of	of	ADP
easat-11054	184	12	any	any	DET
easat-11054	184	13	a	a	NOUN
easat-11054	184	14	and	and	CCONJ
easat-11054	184	15	the	the	DET
easat-11054	184	16	more	more	ADV
easat-11054	184	17	uncertain	uncertain	ADJ
easat-11054	184	18	the	the	DET
easat-11054	184	19	worker	worker	NOUN
easat-11054	184	20	will	will	AUX
easat-11054	184	21	be	be	AUX
easat-11054	184	22	about	about	ADP
easat-11054	184	23	which	which	DET
easat-11054	184	24	course	course	NOUN
easat-11054	184	25	of	of	ADP
easat-11054	184	26	a	a	PRON
easat-11054	184	27	to	to	PART
easat-11054	184	28	take	take	VERB
easat-11054	184	29	[	[	NOUN
easat-11054	184	30	20	20	NUM
easat-11054	184	31	]	]	PUNCT
easat-11054	184	32	.	.	PUNCT
easat-11054	185	1	adding	add	VERB
easat-11054	185	2	entropy	entropy	NOUN
easat-11054	185	3	to	to	ADP
easat-11054	185	4	the	the	DET
easat-11054	185	5	loss	loss	NOUN
easat-11054	185	6	function	function	NOUN
easat-11054	185	7	,	,	PUNCT
easat-11054	185	8	as	as	SCONJ
easat-11054	185	9	expressed	express	VERB
easat-11054	185	10	in	in	ADP
easat-11054	185	11	equation	equation	NOUN
easat-11054	185	12	(	(	PUNCT
easat-11054	185	13	6	6	NUM
easat-11054	185	14	)	)	PUNCT
easat-11054	185	15	,	,	PUNCT
easat-11054	185	16	will	will	AUX
easat-11054	185	17	then	then	ADV
easat-11054	185	18	encourage	encourage	VERB
easat-11054	185	19	exploration	exploration	NOUN
easat-11054	185	20	by	by	ADP
easat-11054	185	21	preventing	prevent	VERB
easat-11054	185	22	the	the	DET
easat-11054	185	23	worker	worker	NOUN
easat-11054	185	24	from	from	ADP
easat-11054	185	25	being	be	AUX
easat-11054	185	26	very	very	ADV
easat-11054	185	27	definite	definite	ADJ
easat-11054	185	28	and	and	CCONJ
easat-11054	185	29	converging	converge	VERB
easat-11054	185	30	to	to	ADP
easat-11054	185	31	a	a	DET
easat-11054	185	32	local	local	ADJ
easat-11054	185	33	optimum	optimum	NOUN
easat-11054	185	34	.	.	PUNCT
easat-11054	186	1	(	(	PUNCT
easat-11054	186	2	)	)	PUNCT
easat-11054	186	3	(	(	PUNCT
easat-11054	186	4	)	)	PUNCT
easat-11054	186	5	(	(	PUNCT
easat-11054	186	6	)	)	PUNCT
easat-11054	186	7	(	(	PUNCT
easat-11054	186	8	)	)	PUNCT
easat-11054	186	9	)	)	PUNCT
easat-11054	186	10	(	(	PUNCT
easat-11054	186	11	;	;	PUNCT
easat-11054	186	12	*	*	PUNCT
easat-11054	186	13	|	|	NOUN
easat-11054	186	14	pl	pl	X
easat-11054	186	15	e	e	NOUN
easat-11054	186	16	a	a	DET
easat-11054	186	17	s	s	X
easat-11054	186	18	log	log	NOUN
easat-11054	186	19	a	a	DET
easat-11054	186	20	s	s	NOUN
easat-11054	186	21	h	h	NOUN
easat-11054	186	22			PROPN
easat-11054	186	23			PROPN
easat-11054	186	24	=	=	NOUN
easat-11054	186	25	−	−	PROPN
easat-11054	186	26	−	−	PROPN
easat-11054	186	27	(	(	PUNCT
easat-11054	186	28	6	6	NUM
easat-11054	186	29	)	)	PUNCT
easat-11054	186	30	the	the	DET
easat-11054	186	31	loss	loss	NOUN
easat-11054	186	32	function	function	NOUN
easat-11054	186	33	for	for	ADP
easat-11054	186	34	the	the	DET
easat-11054	186	35	entire	entire	ADJ
easat-11054	186	36	layer	layer	NOUN
easat-11054	186	37	is	be	AUX
easat-11054	186	38	obtained	obtain	VERB
easat-11054	186	39	by	by	ADP
easat-11054	186	40	combining	combine	VERB
easat-11054	186	41	the	the	DET
easat-11054	186	42	two	two	NUM
easat-11054	186	43	loss	loss	NOUN
easat-11054	186	44	functions	function	NOUN
easat-11054	186	45	as	as	ADP
easat-11054	186	46	in	in	ADP
easat-11054	186	47	equation	equation	NOUN
easat-11054	186	48	(	(	PUNCT
easat-11054	186	49	7	7	X
easat-11054	186	50	)	)	PUNCT
easat-11054	186	51	lv	lv	PROPN
easat-11054	186	52	pl	pl	X
easat-11054	186	53	l	l	NOUN
easat-11054	187	1	=	=	X
easat-11054	188	1	+	+	CCONJ
easat-11054	188	2	(	(	PUNCT
easat-11054	188	3	7	7	NUM
easat-11054	188	4	)	)	PUNCT
easat-11054	188	5	1060	1060	NUM
easat-11054	188	6	edelweiss	edelweiss	PROPN
easat-11054	188	7	applied	apply	VERB
easat-11054	188	8	science	science	NOUN
easat-11054	188	9	and	and	CCONJ
easat-11054	188	10	technology	technology	NOUN
easat-11054	188	11	issn	issn	PROPN
easat-11054	188	12	:	:	PUNCT
easat-11054	188	13	2576	2576	NUM
easat-11054	188	14	-	-	SYM
easat-11054	188	15	8484	8484	NUM
easat-11054	188	16	vol	vol	NOUN
easat-11054	188	17	.	.	PROPN
easat-11054	189	1	9	9	NUM
easat-11054	189	2	,	,	PUNCT
easat-11054	189	3	no	no	INTJ
easat-11054	189	4	.	.	NOUN
easat-11054	189	5	11	11	NUM
easat-11054	189	6	:	:	PUNCT
easat-11054	189	7	1053	1053	NUM
easat-11054	189	8	-	-	SYM
easat-11054	189	9	1075	1075	NUM
easat-11054	189	10	,	,	PUNCT
easat-11054	189	11	2025	2025	NUM
easat-11054	189	12	doi	doi	NOUN
easat-11054	189	13	:	:	PUNCT
easat-11054	189	14	10.55214/2576	10.55214/2576	NUM
easat-11054	189	15	-	-	SYM
easat-11054	189	16	8484.v9i11.11054	8484.v9i11.11054	NOUN
easat-11054	190	1	©	©	PROPN
easat-11054	190	2	2025	2025	NUM
easat-11054	190	3	by	by	ADP
easat-11054	190	4	the	the	DET
easat-11054	190	5	authors	author	NOUN
easat-11054	190	6	;	;	PUNCT
easat-11054	190	7	licensee	licensee	PROPN
easat-11054	190	8	learning	learn	VERB
easat-11054	190	9	gate	gate	NOUN
easat-11054	190	10	ensemble	ensemble	ADJ
easat-11054	190	11	usage	usage	NOUN
easat-11054	190	12	of	of	ADP
easat-11054	190	13	the	the	DET
easat-11054	190	14	dt	dt	NOUN
easat-11054	190	15	technique	technique	NOUN
easat-11054	190	16	will	will	AUX
easat-11054	190	17	output	output	VERB
easat-11054	190	18	class	class	NOUN
easat-11054	190	19	probabilities	probability	NOUN
easat-11054	190	20	.	.	PUNCT
easat-11054	191	1	these	these	DET
easat-11054	191	2	probabilities	probability	NOUN
easat-11054	191	3	are	be	AUX
easat-11054	191	4	computed	compute	VERB
easat-11054	191	5	as	as	ADP
easat-11054	191	6	in	in	ADP
easat-11054	191	7	equation	equation	NOUN
easat-11054	191	8	(	(	PUNCT
easat-11054	191	9	8)	8)	NUM
easat-11054	191	10	from	from	ADP
easat-11054	191	11	the	the	DET
easat-11054	191	12	percentage	percentage	NOUN
easat-11054	191	13	of	of	ADP
easat-11054	191	14	samples	sample	NOUN
easat-11054	191	15	within	within	ADP
easat-11054	191	16	the	the	DET
easat-11054	191	17	leaf	leaf	NOUN
easat-11054	191	18	node	node	NOUN
easat-11054	191	19	that	that	PRON
easat-11054	191	20	fall	fall	VERB
easat-11054	191	21	into	into	ADP
easat-11054	191	22	a	a	DET
easat-11054	191	23	specific	specific	ADJ
easat-11054	191	24	class	class	NOUN
easat-11054	191	25	.	.	PUNCT
easat-11054	192	1	(	(	PUNCT
easat-11054	192	2	)	)	PUNCT
easat-11054	192	3	|	|	ADV
easat-11054	192	4	an	an	DET
easat-11054	192	5	p	p	NOUN
easat-11054	192	6	a	a	DET
easat-11054	192	7	s	s	NOUN
easat-11054	192	8	n	n	NOUN
easat-11054	192	9	=	=	SYM
easat-11054	192	10	(	(	PUNCT
easat-11054	192	11	8)	8)	NUM
easat-11054	192	12	where	where	SCONJ
easat-11054	192	13	na	na	NOUN
easat-11054	192	14	is	be	AUX
easat-11054	192	15	the	the	DET
easat-11054	192	16	number	number	NOUN
easat-11054	192	17	of	of	ADP
easat-11054	192	18	samples	sample	NOUN
easat-11054	192	19	associated	associate	VERB
easat-11054	192	20	with	with	ADP
easat-11054	192	21	action	action	NOUN
easat-11054	192	22	'	'	PUNCT
easat-11054	192	23	a	a	PRON
easat-11054	192	24	'	'	PUNCT
easat-11054	192	25	in	in	ADP
easat-11054	192	26	the	the	DET
easat-11054	192	27	leaf	leaf	NOUN
easat-11054	192	28	node	node	NOUN
easat-11054	192	29	,	,	PUNCT
easat-11054	192	30	n	n	X
easat-11054	192	31	is	be	AUX
easat-11054	192	32	the	the	DET
easat-11054	192	33	total	total	ADJ
easat-11054	192	34	number	number	NOUN
easat-11054	192	35	of	of	ADP
easat-11054	192	36	samples	sample	NOUN
easat-11054	192	37	in	in	ADP
easat-11054	192	38	the	the	DET
easat-11054	192	39	leaf	leaf	NOUN
easat-11054	192	40	node	node	NOUN
easat-11054	192	41	,	,	PUNCT
easat-11054	192	42	and	and	CCONJ
easat-11054	192	43	p	p	NOUN
easat-11054	192	44	is	be	AUX
easat-11054	192	45	the	the	DET
easat-11054	192	46	probability	probability	NOUN
easat-11054	192	47	of	of	ADP
easat-11054	192	48	executing	execute	VERB
easat-11054	192	49	a	a	DET
easat-11054	192	50	given	give	VERB
easat-11054	192	51	state	state	NOUN
easat-11054	192	52	s.	s.	PROPN
easat-11054	192	53	final	final	ADJ
easat-11054	192	54	probabilities	probability	NOUN
easat-11054	192	55	are	be	AUX
easat-11054	192	56	then	then	ADV
easat-11054	192	57	obtained	obtain	VERB
easat-11054	192	58	by	by	ADP
easat-11054	192	59	averaging	average	VERB
easat-11054	192	60	all	all	DET
easat-11054	192	61	the	the	DET
easat-11054	192	62	predictions	prediction	NOUN
easat-11054	192	63	made	make	VERB
easat-11054	192	64	by	by	ADP
easat-11054	192	65	the	the	DET
easat-11054	192	66	dts	dts	NOUN
easat-11054	192	67	in	in	ADP
easat-11054	192	68	the	the	DET
easat-11054	192	69	ensemble	ensemble	NOUN
easat-11054	192	70	.	.	PUNCT
easat-11054	193	1	each	each	DET
easat-11054	193	2	dt	dt	PUNCT
easat-11054	193	3	predicts	predict	VERB
easat-11054	193	4	the	the	DET
easat-11054	193	5	probability	probability	NOUN
easat-11054	193	6	of	of	ADP
easat-11054	193	7	a	a	DET
easat-11054	193	8	class	class	NOUN
easat-11054	193	9	independently	independently	ADV
easat-11054	193	10	.	.	PUNCT
easat-11054	194	1	this	this	DET
easat-11054	194	2	process	process	NOUN
easat-11054	194	3	of	of	ADP
easat-11054	194	4	averaging	average	VERB
easat-11054	194	5	gives	give	VERB
easat-11054	194	6	robustness	robustness	NOUN
easat-11054	194	7	and	and	CCONJ
easat-11054	194	8	stability	stability	NOUN
easat-11054	194	9	to	to	ADP
easat-11054	194	10	the	the	DET
easat-11054	194	11	predictions	prediction	NOUN
easat-11054	194	12	.	.	PUNCT
easat-11054	195	1	the	the	DET
easat-11054	195	2	dt	dt	PROPN
easat-11054	195	3	algorithm	algorithm	PROPN
easat-11054	195	4	is	be	AUX
easat-11054	195	5	based	base	VERB
easat-11054	195	6	on	on	ADP
easat-11054	195	7	the	the	DET
easat-11054	195	8	gini	gini	NOUN
easat-11054	195	9	impurity	impurity	NOUN
easat-11054	195	10	,	,	PUNCT
easat-11054	195	11	which	which	PRON
easat-11054	195	12	serves	serve	VERB
easat-11054	195	13	as	as	ADP
easat-11054	195	14	a	a	DET
easat-11054	195	15	loss	loss	NOUN
easat-11054	195	16	function	function	NOUN
easat-11054	195	17	to	to	PART
easat-11054	195	18	measure	measure	VERB
easat-11054	195	19	the	the	DET
easat-11054	195	20	impurity	impurity	NOUN
easat-11054	195	21	of	of	ADP
easat-11054	195	22	a	a	DET
easat-11054	195	23	node	node	NOUN
easat-11054	195	24	,	,	PUNCT
easat-11054	195	25	as	as	SCONJ
easat-11054	195	26	shown	show	VERB
easat-11054	195	27	in	in	ADP
easat-11054	195	28	equation	equation	NOUN
easat-11054	195	29	(	(	PUNCT
easat-11054	195	30	9	9	NUM
easat-11054	195	31	)	)	PUNCT
easat-11054	195	32	,	,	PUNCT
easat-11054	195	33	estimating	estimate	VERB
easat-11054	195	34	the	the	DET
easat-11054	195	35	likelihood	likelihood	NOUN
easat-11054	195	36	that	that	SCONJ
easat-11054	195	37	a	a	DET
easat-11054	195	38	randomly	randomly	ADV
easat-11054	195	39	selected	select	VERB
easat-11054	195	40	state	state	NOUN
easat-11054	195	41	from	from	ADP
easat-11054	195	42	the	the	DET
easat-11054	195	43	dataset	dataset	NOUN
easat-11054	195	44	in	in	ADP
easat-11054	195	45	a	a	DET
easat-11054	195	46	node	node	NOUN
easat-11054	195	47	would	would	AUX
easat-11054	195	48	be	be	AUX
easat-11054	195	49	incorrectly	incorrectly	ADV
easat-11054	195	50	classified	classify	VERB
easat-11054	195	51	.	.	PUNCT
easat-11054	196	1	2	2	NUM
easat-11054	196	2	(	(	PUNCT
easat-11054	196	3	)	)	PUNCT
easat-11054	196	4	1	1	NUM
easat-11054	196	5	(	(	PUNCT
easat-11054	196	6	)	)	PUNCT
easat-11054	196	7	igl	igl	PROPN
easat-11054	196	8	p	p	PROPN
easat-11054	196	9	p=	p=	NOUN
easat-11054	196	10	−	−	NOUN
easat-11054	196	11	(	(	PUNCT
easat-11054	196	12	9	9	NUM
easat-11054	196	13	)	)	PUNCT
easat-11054	196	14	where	where	SCONJ
easat-11054	196	15	lg	lg	NOUN
easat-11054	196	16	is	be	AUX
easat-11054	196	17	the	the	DET
easat-11054	196	18	gini	gini	PROPN
easat-11054	196	19	impurity	impurity	NOUN
easat-11054	196	20	of	of	ADP
easat-11054	196	21	a	a	DET
easat-11054	196	22	node	node	NOUN
easat-11054	196	23	and	and	CCONJ
easat-11054	196	24	pi	pi	NOUN
easat-11054	196	25	is	be	AUX
easat-11054	196	26	the	the	DET
easat-11054	196	27	proportion	proportion	NOUN
easat-11054	196	28	of	of	ADP
easat-11054	196	29	samples	sample	NOUN
easat-11054	196	30	in	in	ADP
easat-11054	196	31	that	that	DET
easat-11054	196	32	node	node	NOUN
easat-11054	196	33	that	that	PRON
easat-11054	196	34	belong	belong	VERB
easat-11054	196	35	to	to	ADP
easat-11054	196	36	class	class	NOUN
easat-11054	196	37	i.	i.	NOUN
easat-11054	196	38	the	the	DET
easat-11054	196	39	training	training	NOUN
easat-11054	196	40	predictions	prediction	NOUN
easat-11054	196	41	from	from	ADP
easat-11054	196	42	the	the	DET
easat-11054	196	43	a3c	a3c	NUM
easat-11054	196	44	-	-	PUNCT
easat-11054	196	45	ffnn	ffnn	NOUN
easat-11054	196	46	algorithm	algorithm	NOUN
easat-11054	196	47	and	and	CCONJ
easat-11054	196	48	the	the	DET
easat-11054	196	49	dt	dt	PROPN
easat-11054	196	50	algorithm	algorithm	PROPN
easat-11054	196	51	are	be	AUX
easat-11054	196	52	combined	combine	VERB
easat-11054	196	53	in	in	ADP
easat-11054	196	54	the	the	DET
easat-11054	196	55	ffnn	ffnn	NOUN
easat-11054	196	56	layer	layer	NOUN
easat-11054	196	57	.	.	PUNCT
easat-11054	197	1	the	the	DET
easat-11054	197	2	outputs	output	NOUN
easat-11054	197	3	of	of	ADP
easat-11054	197	4	the	the	DET
easat-11054	197	5	predictions	prediction	NOUN
easat-11054	197	6	from	from	ADP
easat-11054	197	7	the	the	DET
easat-11054	197	8	a3c	a3c	NUM
easat-11054	197	9	-	-	PUNCT
easat-11054	197	10	ffnn	ffnn	NOUN
easat-11054	197	11	and	and	CCONJ
easat-11054	197	12	the	the	DET
easat-11054	197	13	dt	dt	PROPN
easat-11054	197	14	algorithms	algorithm	NOUN
easat-11054	197	15	define	define	VERB
easat-11054	197	16	the	the	DET
easat-11054	197	17	input	input	NOUN
easat-11054	197	18	layers	layer	NOUN
easat-11054	197	19	of	of	ADP
easat-11054	197	20	the	the	DET
easat-11054	197	21	ffnn	ffnn	NOUN
easat-11054	197	22	layer	layer	NOUN
easat-11054	197	23	.	.	PUNCT
easat-11054	198	1	the	the	DET
easat-11054	198	2	following	follow	VERB
easat-11054	198	3	equation	equation	NOUN
easat-11054	198	4	(	(	PUNCT
easat-11054	198	5	10	10	NUM
easat-11054	198	6	)	)	PUNCT
easat-11054	198	7	represents	represent	VERB
easat-11054	198	8	the	the	DET
easat-11054	198	9	results	result	NOUN
easat-11054	198	10	of	of	ADP
easat-11054	198	11	the	the	DET
easat-11054	198	12	ffnn	ffnn	NOUN
easat-11054	198	13	layer	layer	NOUN
easat-11054	198	14	f	f	PROPN
easat-11054	198	15	from	from	ADP
easat-11054	198	16	concatenating	concatenate	VERB
easat-11054	198	17	these	these	DET
easat-11054	198	18	predictions	prediction	NOUN
easat-11054	198	19	,	,	PUNCT
easat-11054	198	20	represented	represent	VERB
easat-11054	198	21	by	by	ADP
easat-11054	198	22	π(a|s;θ	π(a|s;θ	PROPN
easat-11054	198	23	)	)	PUNCT
easat-11054	198	24	with	with	ADP
easat-11054	198	25	p(a|s	p(a|s	NUM
easat-11054	198	26	)	)	PUNCT
easat-11054	198	27	.	.	PUNCT
easat-11054	199	1	(	(	PUNCT
easat-11054	199	2	)	)	PUNCT
easat-11054	199	3	(	(	PUNCT
easat-11054	199	4	)	)	PUNCT
easat-11054	199	5	(	(	PUNCT
easat-11054	199	6	)	)	PUNCT
easat-11054	199	7	;	;	PUNCT
easat-11054	199	8	|	|	ADV
easat-11054	199	9	|	|	ADV
easat-11054	199	10	|a	|a	PRON
easat-11054	199	11	s	s	VERB
easat-11054	199	12	a	a	DET
easat-11054	199	13	s	s	X
easat-11054	199	14	p	p	NOUN
easat-11054	199	15	a	a	DET
easat-11054	199	16	sf	sf	X
easat-11054	199	17			X
easat-11054	199	18	=	=	X
easat-11054	199	19	(	(	PUNCT
easat-11054	199	20	10	10	NUM
easat-11054	199	21	)	)	PUNCT
easat-11054	199	22	so	so	ADV
easat-11054	199	23	,	,	PUNCT
easat-11054	199	24	the	the	DET
easat-11054	199	25	total	total	ADJ
easat-11054	199	26	loss	loss	NOUN
easat-11054	199	27	function	function	NOUN
easat-11054	199	28	from	from	ADP
easat-11054	199	29	ffnn	ffnn	NOUN
easat-11054	199	30	and	and	CCONJ
easat-11054	199	31	dt	dt	ADP
easat-11054	199	32	algorithms	algorithm	NOUN
easat-11054	199	33	will	will	AUX
easat-11054	199	34	be	be	AUX
easat-11054	199	35	calculated	calculate	VERB
easat-11054	199	36	as	as	ADP
easat-11054	199	37	in	in	ADP
easat-11054	199	38	equation	equation	NOUN
easat-11054	199	39	(	(	PUNCT
easat-11054	199	40	11	11	NUM
easat-11054	199	41	)	)	PUNCT
easat-11054	199	42	:	:	PUNCT
easat-11054	199	43	(	(	PUNCT
easat-11054	199	44	)	)	PUNCT
easat-11054	199	45	(	(	PUNCT
easat-11054	199	46	)	)	PUNCT
easat-11054	199	47	(	(	PUNCT
easat-11054	199	48	)	)	PUNCT
easat-11054	199	49	(	(	PUNCT
easat-11054	199	50	)	)	PUNCT
easat-11054	199	51	(	(	PUNCT
easat-11054	199	52	)	)	PUNCT
easat-11054	199	53	(	(	PUNCT
easat-11054	199	54	)	)	PUNCT
easat-11054	199	55	(	(	PUNCT
easat-11054	199	56	)	)	PUNCT
easat-11054	199	57	'	'	PART
easat-11054	199	58	0.5	0.5	NUM
easat-11054	199	59	*	*	PUNCT
easat-11054	199	60	—	—	PUNCT
easat-11054	199	61	;	;	PUNCT
easat-11054	199	62	²	²	NUM
easat-11054	199	63	|	|	NOUN
easat-11054	199	64	*	*	PUNCT
easat-11054	199	65	;	;	PUNCT
easat-11054	199	66	*	*	PUNCT
easat-11054	199	67	p	p	NOUN
easat-11054	199	68	gl	gl	NOUN
easat-11054	199	69	r	r	NOUN
easat-11054	199	70	v	v	NOUN
easat-11054	199	71	s	s	NOUN
easat-11054	199	72	log	log	NOUN
easat-11054	199	73	a	a	DET
easat-11054	199	74	s	s	NOUN
easat-11054	199	75	a	a	DET
easat-11054	199	76	s	s	NOUN
easat-11054	199	77	h	h	NOUN
easat-11054	199	78	l	l	NOUN
easat-11054	199	79	p	p	PROPN
easat-11054	200	1			ADJ
easat-11054	200	2			PROPN
easat-11054	200	3			PROPN
easat-11054	200	4			PROPN
easat-11054	200	5	+	+	PROPN
easat-11054	200	6	=	=	PROPN
easat-11054	200	7	−	−	PROPN
easat-11054	200	8	−	−	PROPN
easat-11054	200	9	(	(	PUNCT
easat-11054	200	10	11	11	NUM
easat-11054	200	11	)	)	PUNCT
easat-11054	200	12	as	as	ADP
easat-11054	200	13	training	training	NOUN
easat-11054	200	14	,	,	PUNCT
easat-11054	200	15	the	the	DET
easat-11054	200	16	ltotal	ltotal	ADJ
easat-11054	200	17	total	total	ADJ
easat-11054	200	18	loss	loss	NOUN
easat-11054	200	19	function	function	NOUN
easat-11054	200	20	is	be	AUX
easat-11054	200	21	a	a	DET
easat-11054	200	22	combination	combination	NOUN
easat-11054	200	23	of	of	ADP
easat-11054	200	24	the	the	DET
easat-11054	200	25	loss	loss	NOUN
easat-11054	200	26	from	from	ADP
easat-11054	200	27	the	the	DET
easat-11054	200	28	a3c	a3c	NUM
easat-11054	200	29	-	-	PUNCT
easat-11054	200	30	ffnn	ffnn	NOUN
easat-11054	200	31	and	and	CCONJ
easat-11054	200	32	dt	dt	ADP
easat-11054	200	33	algorithms	algorithm	NOUN
easat-11054	200	34	.	.	PUNCT
easat-11054	201	1	the	the	DET
easat-11054	201	2	compound	compound	NOUN
easat-11054	201	3	loss	loss	NOUN
easat-11054	201	4	is	be	AUX
easat-11054	201	5	optimized	optimize	VERB
easat-11054	201	6	using	use	VERB
easat-11054	201	7	the	the	DET
easat-11054	201	8	adam	adam	PROPN
easat-11054	201	9	optimizer	optimizer	NOUN
easat-11054	201	10	,	,	PUNCT
easat-11054	201	11	which	which	PRON
easat-11054	201	12	automatically	automatically	ADV
easat-11054	201	13	updates	update	VERB
easat-11054	201	14	the	the	DET
easat-11054	201	15	layer	layer	NOUN
easat-11054	201	16	weights	weight	VERB
easat-11054	201	17	via	via	ADP
easat-11054	201	18	backpropagation	backpropagation	NOUN
easat-11054	201	19	.	.	PUNCT
easat-11054	202	1	the	the	DET
easat-11054	202	2	hadt	hadt	NOUN
easat-11054	202	3	algorithm	algorithm	NOUN
easat-11054	202	4	optimizes	optimize	VERB
easat-11054	202	5	the	the	DET
easat-11054	202	6	policy	policy	NOUN
easat-11054	202	7	,	,	PUNCT
easat-11054	202	8	value	value	NOUN
easat-11054	202	9	,	,	PUNCT
easat-11054	202	10	and	and	CCONJ
easat-11054	202	11	decision	decision	NOUN
easat-11054	202	12	tree	tree	NOUN
easat-11054	202	13	components	component	NOUN
easat-11054	202	14	simultaneously	simultaneously	ADV
easat-11054	202	15	with	with	ADP
easat-11054	202	16	sustained	sustained	ADJ
easat-11054	202	17	generalization	generalization	NOUN
easat-11054	202	18	,	,	PUNCT
easat-11054	202	19	detection	detection	NOUN
easat-11054	202	20	accuracy	accuracy	NOUN
easat-11054	202	21	,	,	PUNCT
easat-11054	202	22	and	and	CCONJ
easat-11054	202	23	convergence	convergence	NOUN
easat-11054	202	24	over	over	ADP
easat-11054	202	25	various	various	ADJ
easat-11054	202	26	intrusion	intrusion	NOUN
easat-11054	202	27	patterns	pattern	NOUN
easat-11054	202	28	.	.	PUNCT
easat-11054	203	1	on	on	ADP
easat-11054	203	2	the	the	DET
easat-11054	203	3	other	other	ADJ
easat-11054	203	4	hand	hand	NOUN
easat-11054	203	5	,	,	PUNCT
easat-11054	203	6	a	a	DET
easat-11054	203	7	hidden	hidden	ADJ
easat-11054	203	8	layer	layer	NOUN
easat-11054	203	9	with	with	ADP
easat-11054	203	10	256	256	NUM
easat-11054	203	11	units	unit	NOUN
easat-11054	203	12	and	and	CCONJ
easat-11054	203	13	relu	relu	NOUN
easat-11054	203	14	activation	activation	NOUN
easat-11054	203	15	,	,	PUNCT
easat-11054	203	16	followed	follow	VERB
easat-11054	203	17	by	by	ADP
easat-11054	203	18	dropout	dropout	NOUN
easat-11054	203	19	regularization	regularization	NOUN
easat-11054	203	20	,	,	PUNCT
easat-11054	203	21	is	be	AUX
easat-11054	203	22	applied	apply	VERB
easat-11054	203	23	to	to	PART
easat-11054	203	24	reduce	reduce	VERB
easat-11054	203	25	overfitting	overfitting	NOUN
easat-11054	203	26	.	.	PUNCT
easat-11054	204	1	the	the	DET
easat-11054	204	2	output	output	NOUN
easat-11054	204	3	layer	layer	NOUN
easat-11054	204	4	then	then	ADV
easat-11054	204	5	provides	provide	VERB
easat-11054	204	6	class	class	NOUN
easat-11054	204	7	probabilities	probability	NOUN
easat-11054	204	8	using	use	VERB
easat-11054	204	9	the	the	DET
easat-11054	204	10	softmax	softmax	ADJ
easat-11054	204	11	activation	activation	NOUN
easat-11054	204	12	function	function	NOUN
easat-11054	204	13	.	.	PUNCT
easat-11054	205	1	therefore	therefore	ADV
easat-11054	205	2	,	,	PUNCT
easat-11054	205	3	the	the	DET
easat-11054	205	4	improved	improved	ADJ
easat-11054	205	5	police	police	NOUN
easat-11054	205	6	output	output	NOUN
easat-11054	205	7	,	,	PUNCT
easat-11054	205	8	in	in	ADP
easat-11054	205	9	terms	term	NOUN
easat-11054	205	10	of	of	ADP
easat-11054	205	11	the	the	DET
easat-11054	205	12	police	police	NOUN
easat-11054	205	13	output	output	NOUN
easat-11054	205	14	before	before	ADP
easat-11054	205	15	improvement	improvement	NOUN
easat-11054	205	16	,	,	PUNCT
easat-11054	205	17	will	will	AUX
easat-11054	205	18	be	be	AUX
easat-11054	205	19	as	as	ADP
easat-11054	205	20	in	in	ADP
easat-11054	205	21	equation	equation	NOUN
easat-11054	205	22	(	(	PUNCT
easat-11054	205	23	12	12	NUM
easat-11054	205	24	)	)	PUNCT
easat-11054	205	25	:	:	PUNCT
easat-11054	205	26	(	(	PUNCT
easat-11054	205	27	)	)	PUNCT
easat-11054	205	28	(	(	PUNCT
easat-11054	205	29	)	)	PUNCT
easat-11054	205	30	(	(	PUNCT
easat-11054	205	31	)	)	PUNCT
easat-11054	205	32	(	(	PUNCT
easat-11054	205	33	)	)	PUNCT
easat-11054	205	34	(	(	PUNCT
easat-11054	205	35	)	)	PUNCT
easat-11054	206	1	|	|	ADV
easat-11054	206	2	|	|	ADV
easat-11054	206	3	)	)	PUNCT
easat-11054	206	4	*	*	PUNCT
easat-11054	206	5	)	)	PUNCT
easat-11054	206	6	.	.	PUNCT
easat-11054	206	7	)	)	PUNCT
easat-11054	207	1	|	|	ADV
easat-11054	207	2	|	|	ADV
easat-11054	207	3	)	)	PUNCT
easat-11054	207	4	(	(	PUNCT
easat-11054	207	5	max(0	max(0	NOUN
easat-11054	207	6	,	,	PUNCT
easat-11054	207	7	;	;	PUNCT
easat-11054	207	8	(	(	PUNCT
easat-11054	207	9	max(0	max(0	NOUN
easat-11054	207	10	,	,	PUNCT
easat-11054	207	11	;	;	PUNCT
easat-11054	207	12	1	1	NUM
easat-11054	207	13	*	*	PUNCT
easat-11054	207	14	)	)	PUNCT
easat-11054	207	15	.	.	PUNCT
easat-11054	207	16	)	)	PUNCT
easat-11054	208	1	|	|	ADV
easat-11054	208	2	i	i	PRON
easat-11054	208	3	j	j	PROPN
easat-11054	209	1	a	a	PRON
easat-11054	209	2	s	s	X
easat-11054	209	3	p	p	NOUN
easat-11054	209	4	a	a	DET
easat-11054	209	5	s	s	NOUN
easat-11054	209	6	n	n	PRON
easat-11054	209	7	a	a	DET
easat-11054	209	8	s	s	X
easat-11054	209	9	j	j	NOUN
easat-11054	209	10	m	m	PROPN
easat-11054	209	11	p	p	NOUN
easat-11054	209	12	w	w	PROPN
easat-11054	209	13	b	b	PROPN
easat-11054	209	14	ms	ms	PROPN
easat-11054	209	15	ba	ba	PROPN
easat-11054	209	16	w	w	PROPN
easat-11054	209	17	q	q	PROPN
easat-11054	210	1	e	e	X
easat-11054	210	2	e	e	X
easat-11054	210	3	a	a	PRON
easat-11054	210	4	s	s	X
easat-11054	210	5			NOUN
easat-11054	210	6			X
easat-11054	210	7			NOUN
easat-11054	210	8			X
easat-11054	210	9	+	+	PUNCT
easat-11054	210	10	+	+	PUNCT
easat-11054	210	11	=	=	SYM
easat-11054	210	12	=	=	SYM
easat-11054	210	13			X
easat-11054	210	14	(	(	PUNCT
easat-11054	210	15	12	12	NUM
easat-11054	210	16	)	)	PUNCT
easat-11054	210	17	with	with	ADP
easat-11054	210	18	the	the	DET
easat-11054	210	19	adam	adam	PROPN
easat-11054	210	20	optimizer	optimizer	NOUN
easat-11054	210	21	,	,	PUNCT
easat-11054	210	22	a	a	DET
easat-11054	210	23	batch	batch	NOUN
easat-11054	210	24	size	size	NOUN
easat-11054	210	25	of	of	ADP
easat-11054	210	26	512	512	NUM
easat-11054	210	27	,	,	PUNCT
easat-11054	210	28	and	and	CCONJ
easat-11054	210	29	the	the	DET
easat-11054	210	30	sparse	sparse	ADJ
easat-11054	210	31	categorical	categorical	ADJ
easat-11054	210	32	cross	cross	ADJ
easat-11054	210	33	-	-	ADJ
easat-11054	210	34	entropy	entropy	ADJ
easat-11054	210	35	loss	loss	NOUN
easat-11054	210	36	function	function	NOUN
easat-11054	210	37	,	,	PUNCT
easat-11054	210	38	the	the	DET
easat-11054	210	39	resulting	result	VERB
easat-11054	210	40	equation	equation	NOUN
easat-11054	210	41	builds	build	VERB
easat-11054	210	42	an	an	DET
easat-11054	210	43	ffnn	ffnn	NOUN
easat-11054	210	44	layer	layer	NOUN
easat-11054	210	45	across	across	ADP
easat-11054	210	46	training	train	VERB
easat-11054	210	47	iterations	iteration	NOUN
easat-11054	210	48	for	for	ADP
easat-11054	210	49	up	up	ADP
easat-11054	210	50	to	to	PART
easat-11054	210	51	50	50	NUM
easat-11054	210	52	epochs	epoch	NOUN
easat-11054	210	53	with	with	ADP
easat-11054	210	54	early	early	ADJ
easat-11054	210	55	stopping	stopping	NOUN
easat-11054	210	56	,	,	PUNCT
easat-11054	210	57	as	as	SCONJ
easat-11054	210	58	shown	show	VERB
easat-11054	210	59	in	in	ADP
easat-11054	210	60	algorithm	algorithm	NOUN
easat-11054	210	61	2	2	NUM
easat-11054	210	62	,	,	PUNCT
easat-11054	210	63	if	if	SCONJ
easat-11054	210	64	the	the	DET
easat-11054	210	65	validation	validation	NOUN
easat-11054	210	66	loss	loss	NOUN
easat-11054	210	67	decays	decay	VERB
easat-11054	210	68	with	with	ADP
easat-11054	210	69	a	a	DET
easat-11054	210	70	learning	learn	VERB
easat-11054	210	71	rate	rate	NOUN
easat-11054	210	72	of	of	ADP
easat-11054	210	73	0.001	0.001	NUM
easat-11054	210	74	.	.	PUNCT
easat-11054	211	1	the	the	DET
easat-11054	211	2	a3c	a3c	NUM
easat-11054	211	3	-	-	PUNCT
easat-11054	211	4	ffnn	ffnn	NOUN
easat-11054	211	5	and	and	CCONJ
easat-11054	211	6	dt	dt	ADP
easat-11054	211	7	algorithms	algorithm	NOUN
easat-11054	211	8	'	'	PART
easat-11054	211	9	predictions	prediction	NOUN
easat-11054	211	10	use	use	VERB
easat-11054	211	11	validation	validation	NOUN
easat-11054	211	12	data	datum	NOUN
easat-11054	211	13	to	to	PART
easat-11054	211	14	assess	assess	VERB
easat-11054	211	15	an	an	DET
easat-11054	211	16	algorithm	algorithm	NOUN
easat-11054	211	17	's	's	PART
easat-11054	211	18	performance	performance	NOUN
easat-11054	211	19	during	during	ADP
easat-11054	211	20	the	the	DET
easat-11054	211	21	training	training	NOUN
easat-11054	211	22	of	of	ADP
easat-11054	211	23	the	the	DET
easat-11054	211	24	ffnn	ffnn	NOUN
easat-11054	211	25	layer	layer	NOUN
easat-11054	211	26	.	.	PUNCT
easat-11054	212	1	algorithm2	algorithm2	NOUN
easat-11054	212	2	:	:	PUNCT
easat-11054	212	3	hybrid	hybrid	ADJ
easat-11054	212	4	hadt	hadt	NOUN
easat-11054	212	5	algorithm	algorithm	NOUN
easat-11054	212	6	1	1	NUM
easat-11054	212	7	2	2	NUM
easat-11054	212	8	3	3	NUM
easat-11054	212	9	4	4	NUM
easat-11054	212	10	//compute	//compute	SYM
easat-11054	212	11	a3c	a3c	NUM
easat-11054	212	12	and	and	CCONJ
easat-11054	212	13	dt	dt	NOUN
easat-11054	212	14	probabilities	probability	NOUN
easat-11054	212	15	.	.	PUNCT
easat-11054	213	1	pdt:=	pdt:=	NOUN
easat-11054	213	2	dt_predict_proba(x′	dt_predict_proba(x′	VERB
easat-11054	213	3	)	)	PUNCT
easat-11054	213	4	pa3c:=	pa3c:=	ADJ
easat-11054	213	5	a3c_predict_proba	a3c_predict_proba	NOUN
easat-11054	213	6	(	(	PUNCT
easat-11054	213	7	x′	x′	NUM
easat-11054	213	8	)	)	PUNCT
easat-11054	213	9	for	for	ADP
easat-11054	213	10	e=1	e=1	PROPN
easat-11054	213	11	to	to	ADP
easat-11054	213	12	50	50	NUM
easat-11054	213	13	do	do	NOUN
easat-11054	213	14	//	//	NUM
easat-11054	213	15	loop	loop	NOUN
easat-11054	213	16	for	for	ADP
easat-11054	213	17	training	training	NOUN
easat-11054	213	18	over	over	ADP
easat-11054	213	19	epochs	epoch	NOUN
easat-11054	213	20	.	.	PUNCT
easat-11054	214	1	1061	1061	NUM
easat-11054	214	2	edelweiss	edelweiss	PROPN
easat-11054	214	3	applied	apply	VERB
easat-11054	214	4	science	science	NOUN
easat-11054	214	5	and	and	CCONJ
easat-11054	214	6	technology	technology	NOUN
easat-11054	214	7	issn	issn	PROPN
easat-11054	214	8	:	:	PUNCT
easat-11054	214	9	2576	2576	NUM
easat-11054	214	10	-	-	SYM
easat-11054	214	11	8484	8484	NUM
easat-11054	214	12	vol	vol	NOUN
easat-11054	214	13	.	.	PROPN
easat-11054	215	1	9	9	NUM
easat-11054	215	2	,	,	PUNCT
easat-11054	215	3	no	no	INTJ
easat-11054	215	4	.	.	NOUN
easat-11054	215	5	11	11	NUM
easat-11054	215	6	:	:	PUNCT
easat-11054	215	7	1053	1053	NUM
easat-11054	215	8	-	-	SYM
easat-11054	215	9	1075	1075	NUM
easat-11054	215	10	,	,	PUNCT
easat-11054	215	11	2025	2025	NUM
easat-11054	215	12	doi	doi	NOUN
easat-11054	215	13	:	:	PUNCT
easat-11054	215	14	10.55214/2576	10.55214/2576	NUM
easat-11054	215	15	-	-	SYM
easat-11054	215	16	8484.v9i11.11054	8484.v9i11.11054	NOUN
easat-11054	216	1	©	©	PROPN
easat-11054	216	2	2025	2025	NUM
easat-11054	216	3	by	by	ADP
easat-11054	216	4	the	the	DET
easat-11054	216	5	authors	author	NOUN
easat-11054	216	6	;	;	PUNCT
easat-11054	216	7	licensee	licensee	PROPN
easat-11054	216	8	learning	learning	NOUN
easat-11054	216	9	gate	gate	NOUN
easat-11054	216	10	5	5	NUM
easat-11054	216	11	6	6	NUM
easat-11054	216	12	7	7	NUM
easat-11054	216	13	8	8	NUM
easat-11054	216	14	9	9	NUM
easat-11054	216	15	1	1	NUM
easat-11054	216	16	0	0	NUM
easat-11054	216	17	1	1	NUM
easat-11054	216	18	1	1	NUM
easat-11054	216	19	1	1	NUM
easat-11054	216	20	2	2	NUM
easat-11054	216	21	1	1	NUM
easat-11054	216	22	3	3	NUM
easat-11054	216	23	1	1	NUM
easat-11054	216	24	4	4	NUM
easat-11054	216	25	1	1	NUM
easat-11054	216	26	5	5	NUM
easat-11054	216	27	1	1	NUM
easat-11054	216	28	6	6	NUM
easat-11054	216	29	f:=	f:=	NOUN
easat-11054	216	30	pdt,||pa3c	pdt,||pa3c	VERB
easat-11054	216	31	)	)	PUNCT
easat-11054	216	32	//concatenate	//concatenate	PUNCT
easat-11054	217	1	features	feature	NOUN
easat-11054	217	2	from	from	ADP
easat-11054	217	3	both	both	DET
easat-11054	217	4	algorithms	algorithm	NOUN
easat-11054	217	5	h:=	h:=	NOUN
easat-11054	217	6	σ(w1f+b1	σ(w1f+b1	NOUN
easat-11054	217	7	)	)	PUNCT
easat-11054	217	8	//apply	//apply	PUNCT
easat-11054	217	9	fully	fully	ADV
easat-11054	217	10	connected	connect	VERB
easat-11054	217	11	layer	layer	NOUN
easat-11054	217	12	.	.	PUNCT
easat-11054	218	1	h′:=	h′:=	PROPN
easat-11054	218	2	dropout(h	dropout(h	NOUN
easat-11054	218	3	,	,	PUNCT
easat-11054	218	4	p=0.5	p=0.5	NOUN
easat-11054	218	5	)	)	PUNCT
easat-11054	218	6	//apply	//apply	VERB
easat-11054	218	7	dropout	dropout	NOUN
easat-11054	218	8	for	for	ADP
easat-11054	218	9	regularization	regularization	NOUN
easat-11054	218	10	.	.	PUNCT
easat-11054	219	1	q:=	q:=	PROPN
easat-11054	220	1	softmax(w2h′+b2	softmax(w2h′+b2	PROPN
easat-11054	220	2	)	)	PUNCT
easat-11054	220	3	//compute	//compute	PUNCT
easat-11054	221	1	output	output	NOUN
easat-11054	221	2	layer	layer	NOUN
easat-11054	221	3	.	.	PUNCT
easat-11054	222	1	l:=	l:=	NOUN
easat-11054	222	2	−∑ylog	−∑ylog	PROPN
easat-11054	222	3	(	(	PUNCT
easat-11054	222	4	q	q	NOUN
easat-11054	222	5	)	)	PUNCT
easat-11054	222	6	//	//	NOUN
easat-11054	222	7	loss	loss	NOUN
easat-11054	222	8	function	function	NOUN
easat-11054	222	9	(	(	PUNCT
easat-11054	222	10	sparse	sparse	ADJ
easat-11054	222	11	categorical	categorical	ADJ
easat-11054	222	12	cross	cross	NOUN
easat-11054	222	13	-	-	NOUN
easat-11054	222	14	entropy	entropy	ADJ
easat-11054	222	15	)	)	PUNCT
easat-11054	222	16	.	.	PUNCT
easat-11054	222	17	θ←θ−α∇l	θ←θ−α∇l	PUNCT
easat-11054	223	1	//update	//update	PUNCT
easat-11054	223	2	layer	layer	NOUN
easat-11054	223	3	weights	weight	NOUN
easat-11054	223	4	using	use	VERB
easat-11054	223	5	adam	adam	PROPN
easat-11054	223	6	optimizer	optimizer	NOUN
easat-11054	223	7	.	.	PUNCT
easat-11054	224	1	m(e	m(e	VERB
easat-11054	224	2	)	)	PUNCT
easat-11054	224	3	:	:	PUNCT
easat-11054	225	1	=	=	PUNCT
easat-11054	225	2	train(pdt	train(pdt	ADJ
easat-11054	225	3	,	,	PUNCT
easat-11054	225	4	pa3c	pa3c	ADJ
easat-11054	225	5	,	,	PUNCT
easat-11054	225	6	y,512	y,512	NUM
easat-11054	225	7	)	)	PUNCT
easat-11054	225	8	//	//	NOUN
easat-11054	225	9	output	output	NOUN
easat-11054	225	10	trained	train	VERB
easat-11054	225	11	hybrid	hybrid	NOUN
easat-11054	225	12	layer	layer	NOUN
easat-11054	225	13	//early	//early	ADV
easat-11054	225	14	stopping	stop	VERB
easat-11054	225	15	if	if	SCONJ
easat-11054	225	16	validation	validation	NOUN
easat-11054	225	17	loss	loss	NOUN
easat-11054	225	18	does	do	AUX
easat-11054	225	19	not	not	PART
easat-11054	225	20	improve	improve	VERB
easat-11054	225	21	for	for	ADP
easat-11054	225	22	5	5	NUM
easat-11054	225	23	if	if	SCONJ
easat-11054	225	24	lval(e)≥lval(e−5	lval(e)≥lval(e−5	NUM
easat-11054	225	25	)	)	PUNCT
easat-11054	225	26	then	then	ADV
easat-11054	225	27	break	break	VERB
easat-11054	225	28	else	else	ADV
easat-11054	225	29	continue	continue	VERB
easat-11054	225	30	end	end	NOUN
easat-11054	225	31	for	for	ADP
easat-11054	225	32	mfinal:=	mfinal:=	ADJ
easat-11054	225	33	q	q	NOUN
easat-11054	225	34	//output	//output	PUNCT
easat-11054	225	35	trained	train	VERB
easat-11054	225	36	hybrid	hybrid	ADJ
easat-11054	225	37	layer	layer	NOUN
easat-11054	225	38	.	.	PUNCT
easat-11054	226	1	4	4	X
easat-11054	226	2	.	.	NUM
easat-11054	226	3	proposed	propose	VERB
easat-11054	226	4	methodology	methodology	NOUN
easat-11054	226	5	the	the	DET
easat-11054	226	6	article	article	NOUN
easat-11054	226	7	methodology	methodology	NOUN
easat-11054	226	8	relies	rely	VERB
easat-11054	226	9	on	on	ADP
easat-11054	226	10	comprehensive	comprehensive	ADJ
easat-11054	226	11	pre	pre	NOUN
easat-11054	226	12	-	-	NOUN
easat-11054	226	13	processing	processing	NOUN
easat-11054	226	14	of	of	ADP
easat-11054	226	15	the	the	DET
easat-11054	226	16	x	x	ADJ
easat-11054	226	17	-	-	ADJ
easat-11054	226	18	iiotid	iiotid	ADJ
easat-11054	226	19	dataset	dataset	NOUN
easat-11054	226	20	and	and	CCONJ
easat-11054	226	21	training	training	NOUN
easat-11054	226	22	layers	layer	NOUN
easat-11054	226	23	of	of	ADP
easat-11054	226	24	the	the	DET
easat-11054	226	25	hadt	hadt	NOUN
easat-11054	226	26	algorithm	algorithm	NOUN
easat-11054	226	27	,	,	PUNCT
easat-11054	226	28	which	which	PRON
easat-11054	226	29	achieves	achieve	VERB
easat-11054	226	30	performance	performance	NOUN
easat-11054	226	31	improvements	improvement	NOUN
easat-11054	226	32	,	,	PUNCT
easat-11054	226	33	including	include	VERB
easat-11054	226	34	accuracy	accuracy	NOUN
easat-11054	226	35	,	,	PUNCT
easat-11054	226	36	precision	precision	NOUN
easat-11054	226	37	,	,	PUNCT
easat-11054	226	38	recall	recall	NOUN
easat-11054	226	39	,	,	PUNCT
easat-11054	226	40	and	and	CCONJ
easat-11054	226	41	f1	f1	PROPN
easat-11054	226	42	score	score	NOUN
easat-11054	226	43	.	.	PUNCT
easat-11054	227	1	the	the	DET
easat-11054	227	2	phases	phase	NOUN
easat-11054	227	3	of	of	ADP
easat-11054	227	4	this	this	DET
easat-11054	227	5	methodology	methodology	NOUN
easat-11054	227	6	,	,	PUNCT
easat-11054	227	7	as	as	SCONJ
easat-11054	227	8	described	describe	VERB
easat-11054	227	9	in	in	ADP
easat-11054	227	10	figure	figure	NOUN
easat-11054	227	11	3	3	NUM
easat-11054	227	12	,	,	PUNCT
easat-11054	227	13	are	be	AUX
easat-11054	227	14	an	an	DET
easat-11054	227	15	effective	effective	ADJ
easat-11054	227	16	tool	tool	NOUN
easat-11054	227	17	for	for	ADP
easat-11054	227	18	identifying	identify	VERB
easat-11054	227	19	different	different	ADJ
easat-11054	227	20	imbalanced	imbalanced	ADJ
easat-11054	227	21	classes	class	NOUN
easat-11054	227	22	of	of	ADP
easat-11054	227	23	attacks	attack	NOUN
easat-11054	227	24	that	that	PRON
easat-11054	227	25	belong	belong	VERB
easat-11054	227	26	to	to	ADP
easat-11054	227	27	a	a	DET
easat-11054	227	28	mixed	mixed	ADJ
easat-11054	227	29	environment	environment	NOUN
easat-11054	227	30	,	,	PUNCT
easat-11054	227	31	which	which	PRON
easat-11054	227	32	is	be	AUX
easat-11054	227	33	stochastic	stochastic	ADJ
easat-11054	227	34	or	or	CCONJ
easat-11054	227	35	deterministic	deterministic	ADJ
easat-11054	227	36	.	.	PUNCT
easat-11054	228	1	finally	finally	ADV
easat-11054	228	2	,	,	PUNCT
easat-11054	228	3	the	the	DET
easat-11054	228	4	testing	testing	NOUN
easat-11054	228	5	results	result	NOUN
easat-11054	228	6	of	of	ADP
easat-11054	228	7	the	the	DET
easat-11054	228	8	hadt	hadt	NOUN
easat-11054	228	9	algorithm	algorithm	NOUN
easat-11054	228	10	are	be	AUX
easat-11054	228	11	evaluated	evaluate	VERB
easat-11054	228	12	by	by	ADP
easat-11054	228	13	comparing	compare	VERB
easat-11054	228	14	them	they	PRON
easat-11054	228	15	with	with	ADP
easat-11054	228	16	existing	exist	VERB
easat-11054	228	17	a3c	a3c	NUM
easat-11054	228	18	algorithms	algorithm	NOUN
easat-11054	228	19	.	.	PUNCT
easat-11054	229	1	figure	figure	NOUN
easat-11054	229	2	3	3	NUM
easat-11054	229	3	.	.	PUNCT
easat-11054	229	4	methodology	methodology	NOUN
easat-11054	229	5	phases	phase	NOUN
easat-11054	229	6	.	.	PUNCT
easat-11054	230	1	4.1	4.1	NUM
easat-11054	230	2	.	.	PUNCT
easat-11054	230	3	phase	phase	NOUN
easat-11054	230	4	1	1	NUM
easat-11054	230	5	.	.	PUNCT
easat-11054	230	6	dataset	dataset	ADJ
easat-11054	230	7	preprocessing	preprocessing	NOUN
easat-11054	230	8	in	in	ADP
easat-11054	230	9	this	this	DET
easat-11054	230	10	phase	phase	NOUN
easat-11054	230	11	,	,	PUNCT
easat-11054	230	12	preprocessing	preprocessing	NOUN
easat-11054	230	13	is	be	AUX
easat-11054	230	14	essential	essential	ADJ
easat-11054	230	15	to	to	PART
easat-11054	230	16	prepare	prepare	VERB
easat-11054	230	17	the	the	DET
easat-11054	230	18	x	x	ADJ
easat-11054	230	19	-	-	ADJ
easat-11054	230	20	iiotid	iiotid	ADJ
easat-11054	230	21	dataset	dataset	NOUN
easat-11054	230	22	for	for	ADP
easat-11054	230	23	use	use	NOUN
easat-11054	230	24	in	in	ADP
easat-11054	230	25	training	training	NOUN
easat-11054	230	26	and	and	CCONJ
easat-11054	230	27	testing	testing	NOUN
easat-11054	230	28	processes	process	NOUN
easat-11054	230	29	for	for	ADP
easat-11054	230	30	intrusion	intrusion	NOUN
easat-11054	230	31	attack	attack	NOUN
easat-11054	230	32	classes	class	NOUN
easat-11054	230	33	.	.	PUNCT
easat-11054	231	1	each	each	PRON
easat-11054	231	2	of	of	ADP
easat-11054	231	3	these	these	DET
easat-11054	231	4	attacks	attack	NOUN
easat-11054	231	5	has	have	VERB
easat-11054	231	6	different	different	ADJ
easat-11054	231	7	data	datum	NOUN
easat-11054	231	8	traffic	traffic	NOUN
easat-11054	231	9	features	feature	NOUN
easat-11054	231	10	.	.	PUNCT
easat-11054	232	1	this	this	DET
easat-11054	232	2	phase	phase	NOUN
easat-11054	232	3	consists	consist	VERB
easat-11054	232	4	of	of	ADP
easat-11054	232	5	three	three	NUM
easat-11054	232	6	steps	step	NOUN
easat-11054	232	7	as	as	SCONJ
easat-11054	232	8	follows	follow	VERB
easat-11054	232	9	:	:	PUNCT
easat-11054	232	10	1062	1062	NUM
easat-11054	232	11	edelweiss	edelweiss	PROPN
easat-11054	232	12	applied	apply	VERB
easat-11054	232	13	science	science	NOUN
easat-11054	232	14	and	and	CCONJ
easat-11054	232	15	technology	technology	NOUN
easat-11054	232	16	issn	issn	PROPN
easat-11054	232	17	:	:	PUNCT
easat-11054	232	18	2576	2576	NUM
easat-11054	232	19	-	-	SYM
easat-11054	232	20	8484	8484	NUM
easat-11054	232	21	vol	vol	NOUN
easat-11054	232	22	.	.	PROPN
easat-11054	233	1	9	9	NUM
easat-11054	233	2	,	,	PUNCT
easat-11054	233	3	no	no	INTJ
easat-11054	233	4	.	.	NOUN
easat-11054	233	5	11	11	NUM
easat-11054	233	6	:	:	PUNCT
easat-11054	233	7	1053	1053	NUM
easat-11054	233	8	-	-	SYM
easat-11054	233	9	1075	1075	NUM
easat-11054	233	10	,	,	PUNCT
easat-11054	233	11	2025	2025	NUM
easat-11054	233	12	doi	doi	NOUN
easat-11054	233	13	:	:	PUNCT
easat-11054	233	14	10.55214/2576	10.55214/2576	NUM
easat-11054	233	15	-	-	SYM
easat-11054	233	16	8484.v9i11.11054	8484.v9i11.11054	NOUN
easat-11054	234	1	©	©	PROPN
easat-11054	234	2	2025	2025	NUM
easat-11054	234	3	by	by	ADP
easat-11054	234	4	the	the	DET
easat-11054	234	5	authors	author	NOUN
easat-11054	234	6	;	;	PUNCT
easat-11054	234	7	licensee	licensee	PROPN
easat-11054	234	8	learning	learn	VERB
easat-11054	234	9	gate	gate	NOUN
easat-11054	234	10	-step	-step	PROPN
easat-11054	234	11	1	1	NUM
easat-11054	234	12	:	:	PUNCT
easat-11054	234	13	class	class	NOUN
easat-11054	234	14	label	label	NOUN
easat-11054	234	15	encoding	encoding	NOUN
easat-11054	234	16	:	:	PUNCT
easat-11054	234	17	the	the	DET
easat-11054	234	18	dataset	dataset	NOUN
easat-11054	234	19	was	be	AUX
easat-11054	234	20	split	split	VERB
easat-11054	234	21	twice	twice	ADV
easat-11054	234	22	.	.	PUNCT
easat-11054	235	1	the	the	DET
easat-11054	235	2	first	first	ADJ
easat-11054	235	3	time	time	NOUN
easat-11054	235	4	,	,	PUNCT
easat-11054	235	5	it	it	PRON
easat-11054	235	6	was	be	AUX
easat-11054	235	7	divided	divide	VERB
easat-11054	235	8	into	into	ADP
easat-11054	235	9	two	two	NUM
easat-11054	235	10	main	main	ADJ
easat-11054	235	11	categories	category	NOUN
easat-11054	235	12	:	:	PUNCT
easat-11054	235	13	normal	normal	ADJ
easat-11054	235	14	traffic	traffic	NOUN
easat-11054	235	15	and	and	CCONJ
easat-11054	235	16	malicious	malicious	ADJ
easat-11054	235	17	data	datum	NOUN
easat-11054	235	18	.	.	PUNCT
easat-11054	236	1	this	this	PRON
easat-11054	236	2	is	be	AUX
easat-11054	236	3	called	call	VERB
easat-11054	236	4	binary	binary	ADJ
easat-11054	236	5	classification	classification	NOUN
easat-11054	236	6	.	.	PUNCT
easat-11054	237	1	the	the	DET
easat-11054	237	2	second	second	ADJ
easat-11054	237	3	time	time	NOUN
easat-11054	237	4	,	,	PUNCT
easat-11054	237	5	the	the	DET
easat-11054	237	6	classification	classification	NOUN
easat-11054	237	7	was	be	AUX
easat-11054	237	8	based	base	VERB
easat-11054	237	9	on	on	ADP
easat-11054	237	10	normal	normal	ADJ
easat-11054	237	11	traffic	traffic	NOUN
easat-11054	237	12	and	and	CCONJ
easat-11054	237	13	18	18	NUM
easat-11054	237	14	intrusion	intrusion	NOUN
easat-11054	237	15	data	datum	NOUN
easat-11054	237	16	classes	class	NOUN
easat-11054	237	17	.	.	PUNCT
easat-11054	238	1	normal	normal	ADJ
easat-11054	238	2	traffic	traffic	NOUN
easat-11054	238	3	represents	represent	VERB
easat-11054	238	4	legitimate	legitimate	ADJ
easat-11054	238	5	network	network	NOUN
easat-11054	238	6	activity	activity	NOUN
easat-11054	238	7	.	.	PUNCT
easat-11054	239	1	rdos	rdo	NOUN
easat-11054	239	2	(	(	PUNCT
easat-11054	239	3	resource	resource	NOUN
easat-11054	239	4	depletion	depletion	NOUN
easat-11054	239	5	denial	denial	NOUN
easat-11054	239	6	-	-	PUNCT
easat-11054	239	7	of	of	ADP
easat-11054	239	8	-	-	PUNCT
easat-11054	239	9	service	service	NOUN
easat-11054	239	10	)	)	PUNCT
easat-11054	239	11	and	and	CCONJ
easat-11054	239	12	brute	brute	ADJ
easat-11054	239	13	-	-	PUNCT
easat-11054	239	14	force	force	NOUN
easat-11054	239	15	attacks	attack	NOUN
easat-11054	239	16	aim	aim	VERB
easat-11054	239	17	to	to	PART
easat-11054	239	18	overwhelm	overwhelm	VERB
easat-11054	239	19	or	or	CCONJ
easat-11054	239	20	gain	gain	VERB
easat-11054	239	21	unauthorized	unauthorized	ADJ
easat-11054	239	22	access	access	NOUN
easat-11054	239	23	to	to	ADP
easat-11054	239	24	systems	system	NOUN
easat-11054	239	25	.	.	PUNCT
easat-11054	240	1	scanning	scan	VERB
easat-11054	240	2	vulnerability	vulnerability	NOUN
easat-11054	240	3	and	and	CCONJ
easat-11054	240	4	generic	generic	ADJ
easat-11054	240	5	scanning	scanning	NOUN
easat-11054	240	6	involve	involve	VERB
easat-11054	240	7	probing	probe	VERB
easat-11054	240	8	networks	network	NOUN
easat-11054	240	9	for	for	ADP
easat-11054	240	10	weaknesses	weakness	NOUN
easat-11054	240	11	.	.	PUNCT
easat-11054	241	1	mqtt	mqtt	PROPN
easat-11054	241	2	cloud	cloud	PROPN
easat-11054	241	3	broker	broker	PROPN
easat-11054	241	4	subscription	subscription	NOUN
easat-11054	241	5	and	and	CCONJ
easat-11054	241	6	resource	resource	NOUN
easat-11054	241	7	discovery	discovery	NOUN
easat-11054	241	8	target	target	NOUN
easat-11054	241	9	iot	iot	ADJ
easat-11054	241	10	communication	communication	NOUN
easat-11054	241	11	protocols	protocol	NOUN
easat-11054	241	12	and	and	CCONJ
easat-11054	241	13	resource	resource	NOUN
easat-11054	241	14	identification	identification	NOUN
easat-11054	241	15	.	.	PUNCT
easat-11054	242	1	exfiltration	exfiltration	NOUN
easat-11054	242	2	refers	refer	VERB
easat-11054	242	3	to	to	ADP
easat-11054	242	4	unauthorized	unauthorized	ADJ
easat-11054	242	5	data	datum	NOUN
easat-11054	242	6	extraction	extraction	NOUN
easat-11054	242	7	,	,	PUNCT
easat-11054	242	8	while	while	SCONJ
easat-11054	242	9	insider	insider	NOUN
easat-11054	242	10	malicious	malicious	ADJ
easat-11054	242	11	threats	threat	NOUN
easat-11054	242	12	originate	originate	VERB
easat-11054	242	13	from	from	ADP
easat-11054	242	14	within	within	ADP
easat-11054	242	15	the	the	DET
easat-11054	242	16	organization	organization	NOUN
easat-11054	242	17	.	.	PUNCT
easat-11054	243	1	modbus	modbus	PROPN
easat-11054	243	2	register	register	PROPN
easat-11054	243	3	reading	reading	NOUN
easat-11054	243	4	and	and	CCONJ
easat-11054	243	5	false	false	ADJ
easat-11054	243	6	data	datum	NOUN
easat-11054	243	7	injection	injection	NOUN
easat-11054	243	8	can	can	AUX
easat-11054	243	9	manipulate	manipulate	VERB
easat-11054	243	10	industrial	industrial	ADJ
easat-11054	243	11	control	control	NOUN
easat-11054	243	12	systems	system	NOUN
easat-11054	243	13	.	.	PUNCT
easat-11054	244	1	c&c	c&c	NOUN
easat-11054	244	2	(	(	PUNCT
easat-11054	244	3	command	command	NOUN
easat-11054	244	4	and	and	CCONJ
easat-11054	244	5	control	control	NOUN
easat-11054	244	6	)	)	PUNCT
easat-11054	244	7	attacks	attack	NOUN
easat-11054	244	8	enable	enable	VERB
easat-11054	244	9	remote	remote	ADJ
easat-11054	244	10	control	control	NOUN
easat-11054	244	11	of	of	ADP
easat-11054	244	12	compromised	compromise	VERB
easat-11054	244	13	devices	device	NOUN
easat-11054	244	14	.	.	PUNCT
easat-11054	245	1	dictionary	dictionary	ADJ
easat-11054	245	2	and	and	CCONJ
easat-11054	245	3	tcp	tcp	VERB
easat-11054	245	4	relay	relay	NOUN
easat-11054	245	5	attacks	attack	NOUN
easat-11054	245	6	focus	focus	VERB
easat-11054	245	7	on	on	ADP
easat-11054	245	8	credential	credential	NOUN
easat-11054	245	9	-	-	PUNCT
easat-11054	245	10	based	base	VERB
easat-11054	245	11	intrusions	intrusion	NOUN
easat-11054	245	12	and	and	CCONJ
easat-11054	245	13	traffic	traffic	NOUN
easat-11054	245	14	redirection	redirection	NOUN
easat-11054	245	15	.	.	PUNCT
easat-11054	246	1	fuzzing	fuzze	VERB
easat-11054	246	2	tests	test	NOUN
easat-11054	246	3	system	system	NOUN
easat-11054	246	4	vulnerabilities	vulnerability	NOUN
easat-11054	246	5	by	by	ADP
easat-11054	246	6	injecting	inject	VERB
easat-11054	246	7	malformed	malformed	ADJ
easat-11054	246	8	data	datum	NOUN
easat-11054	246	9	.	.	PUNCT
easat-11054	247	1	a	a	DET
easat-11054	247	2	reverse	reverse	ADJ
easat-11054	247	3	shell	shell	NOUN
easat-11054	247	4	provides	provide	VERB
easat-11054	247	5	attackers	attacker	NOUN
easat-11054	247	6	with	with	ADP
easat-11054	247	7	remote	remote	ADJ
easat-11054	247	8	system	system	NOUN
easat-11054	247	9	access	access	NOUN
easat-11054	247	10	.	.	PUNCT
easat-11054	248	1	crypto	crypto	NOUN
easat-11054	248	2	-	-	ADJ
easat-11054	248	3	ransomware	ransomware	ADJ
easat-11054	248	4	encrypts	encrypt	NOUN
easat-11054	248	5	critical	critical	ADJ
easat-11054	248	6	data	datum	NOUN
easat-11054	248	7	for	for	ADP
easat-11054	248	8	ransom	ransom	NOUN
easat-11054	248	9	.	.	PUNCT
easat-11054	249	1	man	man	NOUN
easat-11054	249	2	-	-	PUNCT
easat-11054	249	3	in	in	ADP
easat-11054	249	4	-	-	PUNCT
easat-11054	249	5	the	the	DET
easat-11054	249	6	-	-	PUNCT
easat-11054	249	7	middle	middle	ADJ
easat-11054	249	8	(	(	PUNCT
easat-11054	249	9	mitm	mitm	NOUN
easat-11054	249	10	)	)	PUNCT
easat-11054	249	11	attacks	attack	NOUN
easat-11054	249	12	intercept	intercept	VERB
easat-11054	249	13	communication	communication	NOUN
easat-11054	249	14	,	,	PUNCT
easat-11054	249	15	while	while	SCONJ
easat-11054	249	16	fake	fake	ADJ
easat-11054	249	17	notifications	notification	NOUN
easat-11054	249	18	deceive	deceive	VERB
easat-11054	249	19	users	user	NOUN
easat-11054	249	20	with	with	ADP
easat-11054	249	21	fraudulent	fraudulent	ADJ
easat-11054	249	22	alerts	alert	NOUN
easat-11054	249	23	.	.	PUNCT
easat-11054	250	1	these	these	DET
easat-11054	250	2	attack	attack	NOUN
easat-11054	250	3	categories	category	NOUN
easat-11054	250	4	highlight	highlight	VERB
easat-11054	250	5	the	the	DET
easat-11054	250	6	diverse	diverse	ADJ
easat-11054	250	7	threats	threat	NOUN
easat-11054	250	8	facing	face	VERB
easat-11054	250	9	iiot	iiot	NOUN
easat-11054	250	10	networks	network	NOUN
easat-11054	250	11	and	and	CCONJ
easat-11054	250	12	emphasize	emphasize	VERB
easat-11054	250	13	the	the	DET
easat-11054	250	14	need	need	NOUN
easat-11054	250	15	for	for	ADP
easat-11054	250	16	robust	robust	ADJ
easat-11054	250	17	intrusion	intrusion	NOUN
easat-11054	250	18	detection	detection	NOUN
easat-11054	250	19	systems	system	NOUN
easat-11054	250	20	(	(	PUNCT
easat-11054	250	21	idss	idss	ADV
easat-11054	250	22	)	)	PUNCT
easat-11054	250	23	.	.	PUNCT
easat-11054	251	1	-step	-step	NOUN
easat-11054	251	2	2	2	NUM
easat-11054	251	3	:	:	PUNCT
easat-11054	251	4	feature	feature	NOUN
easat-11054	251	5	encoding	encoding	NOUN
easat-11054	251	6	:	:	PUNCT
easat-11054	251	7	in	in	ADP
easat-11054	251	8	this	this	DET
easat-11054	251	9	step	step	NOUN
easat-11054	251	10	,	,	PUNCT
easat-11054	251	11	categorical	categorical	ADJ
easat-11054	251	12	features	feature	NOUN
easat-11054	251	13	in	in	ADP
easat-11054	251	14	the	the	DET
easat-11054	251	15	dataset	dataset	NOUN
easat-11054	251	16	,	,	PUNCT
easat-11054	251	17	such	such	ADJ
easat-11054	251	18	as	as	ADP
easat-11054	251	19	service	service	NOUN
easat-11054	251	20	and	and	CCONJ
easat-11054	251	21	protocol	protocol	NOUN
easat-11054	251	22	,	,	PUNCT
easat-11054	251	23	are	be	AUX
easat-11054	251	24	converted	convert	VERB
easat-11054	251	25	to	to	ADP
easat-11054	251	26	numeric	numeric	ADJ
easat-11054	251	27	values	value	NOUN
easat-11054	251	28	.	.	PUNCT
easat-11054	252	1	ipv4	ipv4	PROPN
easat-11054	252	2	and	and	CCONJ
easat-11054	252	3	ipv6	ipv6	PROPN
easat-11054	252	4	addresses	address	NOUN
easat-11054	252	5	are	be	AUX
easat-11054	252	6	also	also	ADV
easat-11054	252	7	converted	convert	VERB
easat-11054	252	8	to	to	ADP
easat-11054	252	9	numeric	numeric	ADJ
easat-11054	252	10	values	value	NOUN
easat-11054	252	11	to	to	PART
easat-11054	252	12	facilitate	facilitate	VERB
easat-11054	252	13	the	the	DET
easat-11054	252	14	processing	processing	NOUN
easat-11054	252	15	and	and	CCONJ
easat-11054	252	16	analysis	analysis	NOUN
easat-11054	252	17	of	of	ADP
easat-11054	252	18	network	network	NOUN
easat-11054	252	19	traffic	traffic	NOUN
easat-11054	252	20	data	datum	NOUN
easat-11054	252	21	.	.	PUNCT
easat-11054	253	1	in	in	ADP
easat-11054	253	2	addition	addition	NOUN
easat-11054	253	3	,	,	PUNCT
easat-11054	253	4	missing	miss	VERB
easat-11054	253	5	values	value	NOUN
easat-11054	253	6	are	be	AUX
easat-11054	253	7	replaced	replace	VERB
easat-11054	253	8	with	with	ADP
easat-11054	253	9	-1	-1	PUNCT
easat-11054	253	10	,	,	PUNCT
easat-11054	253	11	while	while	SCONJ
easat-11054	253	12	nan	nan	PROPN
easat-11054	253	13	values	value	NOUN
easat-11054	253	14	are	be	AUX
easat-11054	253	15	pre	pre	ADJ
easat-11054	253	16	-	-	VERB
easat-11054	253	17	filled	filled	ADJ
easat-11054	253	18	to	to	PART
easat-11054	253	19	prevent	prevent	VERB
easat-11054	253	20	data	datum	NOUN
easat-11054	253	21	inconsistencies	inconsistency	NOUN
easat-11054	253	22	during	during	ADP
easat-11054	253	23	training	training	NOUN
easat-11054	253	24	and	and	CCONJ
easat-11054	253	25	testing	testing	NOUN
easat-11054	253	26	.	.	PUNCT
easat-11054	254	1	by	by	ADP
easat-11054	254	2	using	use	VERB
easat-11054	254	3	uniform	uniform	ADJ
easat-11054	254	4	processing	processing	NOUN
easat-11054	254	5	,	,	PUNCT
easat-11054	254	6	feature	feature	NOUN
easat-11054	254	7	values	value	NOUN
easat-11054	254	8	are	be	AUX
easat-11054	254	9	bounded	bound	VERB
easat-11054	254	10	between	between	ADP
easat-11054	254	11	[	[	X
easat-11054	254	12	0	0	NUM
easat-11054	254	13	,	,	PUNCT
easat-11054	254	14	1	1	NUM
easat-11054	254	15	]	]	PUNCT
easat-11054	254	16	to	to	PART
easat-11054	254	17	simplify	simplify	VERB
easat-11054	254	18	calculations	calculation	NOUN
easat-11054	254	19	and	and	CCONJ
easat-11054	254	20	remove	remove	VERB
easat-11054	254	21	dimensionality	dimensionality	NOUN
easat-11054	254	22	,	,	PUNCT
easat-11054	254	23	as	as	SCONJ
easat-11054	254	24	shown	show	VERB
easat-11054	254	25	in	in	ADP
easat-11054	254	26	equation	equation	NOUN
easat-11054	254	27	(	(	PUNCT
easat-11054	254	28	13	13	NUM
easat-11054	254	29	)	)	PUNCT
easat-11054	254	30	.	.	PUNCT
easat-11054	255	1	min	min	PROPN
easat-11054	255	2	max	max	PROPN
easat-11054	255	3	min	min	PROPN
easat-11054	255	4	n	n	PROPN
easat-11054	255	5	v	v	NOUN
easat-11054	255	6	v	v	NUM
easat-11054	255	7	v	v	NUM
easat-11054	255	8	v	v	NOUN
easat-11054	255	9	v	v	NOUN
easat-11054	255	10	−	−	NOUN
easat-11054	255	11	=	=	SYM
easat-11054	255	12	−	−	PROPN
easat-11054	255	13	(	(	PUNCT
easat-11054	255	14	13	13	NUM
easat-11054	255	15	)	)	PUNCT
easat-11054	255	16	where	where	SCONJ
easat-11054	255	17	vmax	vmax	PROPN
easat-11054	255	18	and	and	CCONJ
easat-11054	255	19	vmin	vmin	PROPN
easat-11054	255	20	represent	represent	VERB
easat-11054	255	21	the	the	DET
easat-11054	255	22	highest	high	ADJ
easat-11054	255	23	and	and	CCONJ
easat-11054	255	24	lowest	low	ADJ
easat-11054	255	25	feature	feature	NOUN
easat-11054	255	26	values	value	NOUN
easat-11054	255	27	,	,	PUNCT
easat-11054	255	28	respectively	respectively	ADV
easat-11054	255	29	.	.	PUNCT
easat-11054	256	1	step	step	NOUN
easat-11054	256	2	3	3	NUM
easat-11054	256	3	:	:	PUNCT
easat-11054	256	4	train	train	NOUN
easat-11054	256	5	-	-	PUNCT
easat-11054	256	6	test	test	NOUN
easat-11054	256	7	split	split	NOUN
easat-11054	256	8	:	:	PUNCT
easat-11054	256	9	the	the	DET
easat-11054	256	10	dataset	dataset	NOUN
easat-11054	256	11	is	be	AUX
easat-11054	256	12	divided	divide	VERB
easat-11054	256	13	into	into	ADP
easat-11054	256	14	a	a	DET
easat-11054	256	15	training	training	NOUN
easat-11054	256	16	set	set	NOUN
easat-11054	256	17	(	(	PUNCT
easat-11054	256	18	70	70	NUM
easat-11054	256	19	%	%	NOUN
easat-11054	256	20	)	)	PUNCT
easat-11054	256	21	,	,	PUNCT
easat-11054	256	22	of	of	ADP
easat-11054	256	23	which	which	PRON
easat-11054	256	24	20	20	NUM
easat-11054	256	25	%	%	NOUN
easat-11054	256	26	is	be	AUX
easat-11054	256	27	allocated	allocate	VERB
easat-11054	256	28	for	for	ADP
easat-11054	256	29	validation	validation	NOUN
easat-11054	256	30	during	during	ADP
easat-11054	256	31	the	the	DET
easat-11054	256	32	training	training	NOUN
easat-11054	256	33	process	process	NOUN
easat-11054	256	34	,	,	PUNCT
easat-11054	256	35	and	and	CCONJ
easat-11054	256	36	a	a	DET
easat-11054	256	37	test	test	NOUN
easat-11054	256	38	set	set	NOUN
easat-11054	256	39	(	(	PUNCT
easat-11054	256	40	30	30	NUM
easat-11054	256	41	%	%	NOUN
easat-11054	256	42	)	)	PUNCT
easat-11054	256	43	,	,	PUNCT
easat-11054	256	44	as	as	SCONJ
easat-11054	256	45	shown	show	VERB
easat-11054	256	46	in	in	ADP
easat-11054	256	47	table	table	NOUN
easat-11054	256	48	2	2	NUM
easat-11054	256	49	,	,	PUNCT
easat-11054	256	50	to	to	PART
easat-11054	256	51	determine	determine	VERB
easat-11054	256	52	the	the	DET
easat-11054	256	53	algorithm	algorithm	NOUN
easat-11054	256	54	's	's	PART
easat-11054	256	55	performance	performance	NOUN
easat-11054	256	56	when	when	SCONJ
easat-11054	256	57	using	use	VERB
easat-11054	256	58	the	the	DET
easat-11054	256	59	x	x	ADJ
easat-11054	256	60	-	-	ADJ
easat-11054	256	61	iiotid	iiotid	ADJ
easat-11054	256	62	dataset	dataset	NOUN
easat-11054	256	63	.	.	PUNCT
easat-11054	257	1	learning	learn	VERB
easat-11054	257	2	is	be	AUX
easat-11054	257	3	therefore	therefore	ADV
easat-11054	257	4	performed	perform	VERB
easat-11054	257	5	on	on	ADP
easat-11054	257	6	the	the	DET
easat-11054	257	7	training	training	NOUN
easat-11054	257	8	set	set	NOUN
easat-11054	257	9	.	.	PUNCT
easat-11054	258	1	the	the	DET
easat-11054	258	2	algorithm	algorithm	NOUN
easat-11054	258	3	is	be	AUX
easat-11054	258	4	evaluated	evaluate	VERB
easat-11054	258	5	on	on	ADP
easat-11054	258	6	a	a	DET
easat-11054	258	7	separate	separate	ADJ
easat-11054	258	8	test	test	NOUN
easat-11054	258	9	set	set	VERB
easat-11054	258	10	to	to	PART
easat-11054	258	11	determine	determine	VERB
easat-11054	258	12	its	its	PRON
easat-11054	258	13	detection	detection	NOUN
easat-11054	258	14	ability	ability	NOUN
easat-11054	258	15	.	.	PUNCT
easat-11054	259	1	1063	1063	NUM
easat-11054	259	2	edelweiss	edelweiss	PROPN
easat-11054	259	3	applied	apply	VERB
easat-11054	259	4	science	science	NOUN
easat-11054	259	5	and	and	CCONJ
easat-11054	259	6	technology	technology	NOUN
easat-11054	259	7	issn	issn	PROPN
easat-11054	259	8	:	:	PUNCT
easat-11054	259	9	2576	2576	NUM
easat-11054	259	10	-	-	SYM
easat-11054	259	11	8484	8484	NUM
easat-11054	259	12	vol	vol	NOUN
easat-11054	259	13	.	.	PROPN
easat-11054	260	1	9	9	NUM
easat-11054	260	2	,	,	PUNCT
easat-11054	260	3	no	no	INTJ
easat-11054	260	4	.	.	NOUN
easat-11054	260	5	11	11	NUM
easat-11054	260	6	:	:	PUNCT
easat-11054	260	7	1053	1053	NUM
easat-11054	260	8	-	-	SYM
easat-11054	260	9	1075	1075	NUM
easat-11054	260	10	,	,	PUNCT
easat-11054	260	11	2025	2025	NUM
easat-11054	260	12	doi	doi	NOUN
easat-11054	260	13	:	:	PUNCT
easat-11054	260	14	10.55214/2576	10.55214/2576	NUM
easat-11054	260	15	-	-	SYM
easat-11054	260	16	8484.v9i11.11054	8484.v9i11.11054	NOUN
easat-11054	261	1	©	©	PROPN
easat-11054	261	2	2025	2025	NUM
easat-11054	261	3	by	by	ADP
easat-11054	261	4	the	the	DET
easat-11054	261	5	authors	author	NOUN
easat-11054	261	6	;	;	PUNCT
easat-11054	261	7	licensee	licensee	PROPN
easat-11054	261	8	learning	learn	VERB
easat-11054	261	9	gate	gate	NOUN
easat-11054	261	10	table	table	NOUN
easat-11054	261	11	2	2	NUM
easat-11054	261	12	.	.	PUNCT
easat-11054	261	13	train	train	NOUN
easat-11054	261	14	-	-	PUNCT
easat-11054	261	15	test	test	NOUN
easat-11054	261	16	split	split	NOUN
easat-11054	261	17	for	for	ADP
easat-11054	261	18	multi	multi	ADJ
easat-11054	261	19	-	-	ADJ
easat-11054	261	20	class	class	ADJ
easat-11054	261	21	and	and	CCONJ
easat-11054	261	22	binary	binary	ADJ
easat-11054	261	23	classification	classification	NOUN
easat-11054	261	24	.	.	PUNCT
easat-11054	262	1	class	class	NOUN
easat-11054	262	2	i	i	PROPN
easat-11054	262	3	d	d	PROPN
easat-11054	262	4	class	class	NOUN
easat-11054	262	5	type	type	NOUN
easat-11054	262	6	19	19	NUM
easat-11054	262	7	classes	class	NOUN
easat-11054	262	8	train	train	NOUN
easat-11054	262	9	count	count	NOUN
easat-11054	262	10	(	(	PUNCT
easat-11054	262	11	70	70	NUM
easat-11054	262	12	%	%	NOUN
easat-11054	262	13	)	)	PUNCT
easat-11054	262	14	test	test	NOUN
easat-11054	262	15	count	count	NOUN
easat-11054	262	16	(	(	PUNCT
easat-11054	262	17	30	30	NUM
easat-11054	262	18	%	%	NOUN
easat-11054	262	19	)	)	PUNCT
easat-11054	262	20	binary	binary	PROPN
easat-11054	262	21	classes	class	NOUN
easat-11054	262	22	train	train	NOUN
easat-11054	262	23	count	count	NOUN
easat-11054	262	24	(	(	PUNCT
easat-11054	262	25	70	70	NUM
easat-11054	262	26	%	%	NOUN
easat-11054	262	27	)	)	PUNCT
easat-11054	262	28	test	test	NOUN
easat-11054	262	29	count	count	NOUN
easat-11054	262	30	(	(	PUNCT
easat-11054	262	31	30	30	NUM
easat-11054	262	32	%	%	NOUN
easat-11054	262	33	)	)	PUNCT
easat-11054	262	34	0	0	PUNCT
easat-11054	263	1	normal	normal	ADJ
easat-11054	263	2	300.000	300.000	NUM
easat-11054	263	3	210.504	210.504	NUM
easat-11054	263	4	89.496	89.496	NUM
easat-11054	263	5	300.000	300.000	NUM
easat-11054	263	6	210.504	210.504	NUM
easat-11054	263	7	89.496	89.496	NUM
easat-11054	263	8	1	1	NUM
easat-11054	263	9	rdos	rdo	VERB
easat-11054	263	10	141.261	141.261	NUM
easat-11054	263	11	98.753	98.753	NUM
easat-11054	263	12	42.508	42.508	NUM
easat-11054	263	13	301.498	301.498	NUM
easat-11054	263	14	181.168	181.168	NUM
easat-11054	263	15	120.330	120.330	NUM
easat-11054	263	16	2	2	NUM
easat-11054	263	17	scanning_vulnerability	scanning_vulnerability	NOUN
easat-11054	263	18	52.852	52.852	NUM
easat-11054	263	19	36.923	36.923	NUM
easat-11054	263	20	15.929	15.929	NUM
easat-11054	263	21	3	3	NUM
easat-11054	263	22	generic_scanning	generic_scanne	VERB
easat-11054	263	23	50.277	50.277	NUM
easat-11054	263	24	35.165	35.165	NUM
easat-11054	263	25	15.112	15.112	NUM
easat-11054	263	26	4	4	NUM
easat-11054	263	27	bruteforce	bruteforce	NOUN
easat-11054	263	28	47.241	47.241	NUM
easat-11054	263	29	33.037	33.037	NUM
easat-11054	263	30	14.204	14.204	NUM
easat-11054	263	31	5	5	NUM
easat-11054	263	32	mqtt_cloud_broker_subscription	mqtt_cloud_broker_subscription	NOUN
easat-11054	263	33	23.524	23.524	NUM
easat-11054	263	34	16.415	16.415	NUM
easat-11054	263	35	7.109	7.109	NUM
easat-11054	263	36	6	6	NUM
easat-11054	263	37	discovering_resources	discovering_resource	NOUN
easat-11054	263	38	23.148	23.148	NUM
easat-11054	263	39	16.258	16.258	NUM
easat-11054	263	40	6.890	6.890	NUM
easat-11054	263	41	7	7	NUM
easat-11054	263	42	exfiltration	exfiltration	NOUN
easat-11054	263	43	22.134	22.134	NUM
easat-11054	263	44	15.429	15.429	NUM
easat-11054	263	45	6.705	6.705	NUM
easat-11054	263	46	8	8	NUM
easat-11054	263	47	insider_malicious	insider_malicious	ADJ
easat-11054	263	48	17.447	17.447	NUM
easat-11054	263	49	12.083	12.083	NUM
easat-11054	263	50	5.364	5.364	NUM
easat-11054	263	51	9	9	NUM
easat-11054	263	52	modbus_register_reading	modbus_register_reade	VERB
easat-11054	263	53	5.953	5.953	NUM
easat-11054	263	54	4.129	4.129	NUM
easat-11054	263	55	1.824	1.824	NUM
easat-11054	263	56	10	10	NUM
easat-11054	263	57	false_data_injection	false_data_injection	NOUN
easat-11054	263	58	5.094	5.094	NUM
easat-11054	263	59	3.565	3.565	NUM
easat-11054	263	60	1.529	1.529	NUM
easat-11054	263	61	11	11	NUM
easat-11054	263	62	c&c	c&c	NOUN
easat-11054	263	63	2.863	2.863	NUM
easat-11054	263	64	2.002	2.002	NUM
easat-11054	263	65	861	861	NUM
easat-11054	263	66	12	12	NUM
easat-11054	263	67	dictionary	dictionary	NOUN
easat-11054	263	68	2.572	2.572	NUM
easat-11054	263	69	1.773	1.773	NUM
easat-11054	263	70	799	799	NUM
easat-11054	263	71	13	13	NUM
easat-11054	263	72	tcp	tcp	AUX
easat-11054	263	73	relay	relay	VERB
easat-11054	263	74	2.119	2.119	NUM
easat-11054	263	75	1.507	1.507	NUM
easat-11054	263	76	612	612	NUM
easat-11054	263	77	14	14	NUM
easat-11054	263	78	fuzzing	fuzze	VERB
easat-11054	263	79	1.313	1.313	NUM
easat-11054	263	80	902	902	NUM
easat-11054	263	81	411	411	NUM
easat-11054	263	82	15	15	NUM
easat-11054	263	83	reverse_shell	reverse_shell	NUM
easat-11054	263	84	1.016	1.016	NUM
easat-11054	263	85	722	722	NUM
easat-11054	263	86	294	294	NUM
easat-11054	263	87	16	16	NUM
easat-11054	263	88	crypto	crypto	NOUN
easat-11054	263	89	-	-	NOUN
easat-11054	263	90	ransomware	ransomware	NOUN
easat-11054	263	91	458	458	NUM
easat-11054	263	92	322	322	NUM
easat-11054	263	93	136	136	NUM
easat-11054	263	94	17	17	NUM
easat-11054	263	95	mitm	mitm	NOUN
easat-11054	263	96	117	117	NUM
easat-11054	263	97	83	83	NUM
easat-11054	263	98	34	34	NUM
easat-11054	263	99	18	18	NUM
easat-11054	263	100	fake_notification	fake_notification	NUM
easat-11054	263	101	28	28	NUM
easat-11054	263	102	19	19	NUM
easat-11054	263	103	9	9	NUM
easat-11054	263	104	#	#	SYM
easat-11054	263	105	#	#	NOUN
easat-11054	263	106	total	total	NOUN
easat-11054	263	107	601.498	601.498	NUM
easat-11054	263	108	391.672	391.672	NUM
easat-11054	263	109	209.826	209.826	NUM
easat-11054	263	110	601.498	601.498	NUM
easat-11054	263	111	391.672	391.672	NUM
easat-11054	263	112	209.826	209.826	NUM
easat-11054	263	113	4.2	4.2	NUM
easat-11054	263	114	.	.	PUNCT
easat-11054	263	115	phase	phase	NOUN
easat-11054	263	116	2	2	NUM
easat-11054	263	117	:	:	PUNCT
easat-11054	263	118	training	training	NOUN
easat-11054	263	119	and	and	CCONJ
easat-11054	263	120	testing	testing	NOUN
easat-11054	263	121	to	to	PART
easat-11054	263	122	address	address	VERB
easat-11054	263	123	the	the	DET
easat-11054	263	124	hybrid	hybrid	ADJ
easat-11054	263	125	nature	nature	NOUN
easat-11054	263	126	of	of	ADP
easat-11054	263	127	iiot	iiot	NOUN
easat-11054	263	128	networks	network	NOUN
easat-11054	263	129	,	,	PUNCT
easat-11054	263	130	the	the	DET
easat-11054	263	131	proposed	propose	VERB
easat-11054	263	132	hadt	hadt	NOUN
easat-11054	263	133	algorithm	algorithm	NOUN
easat-11054	263	134	incorporates	incorporate	VERB
easat-11054	263	135	the	the	DET
easat-11054	263	136	strengths	strength	NOUN
easat-11054	263	137	of	of	ADP
easat-11054	263	138	both	both	DET
easat-11054	263	139	a3c	a3c	VERB
easat-11054	263	140	-	-	PUNCT
easat-11054	263	141	ffnn	ffnn	NOUN
easat-11054	263	142	and	and	CCONJ
easat-11054	263	143	dt	dt	X
easat-11054	263	144	with	with	ADP
easat-11054	263	145	ease	ease	NOUN
easat-11054	263	146	.	.	PUNCT
easat-11054	264	1	feature	feature	NOUN
easat-11054	264	2	extraction	extraction	NOUN
easat-11054	264	3	from	from	ADP
easat-11054	264	4	the	the	DET
easat-11054	264	5	input	input	NOUN
easat-11054	264	6	layer	layer	NOUN
easat-11054	264	7	involves	involve	VERB
easat-11054	264	8	preprocessing	preprocesse	VERB
easat-11054	264	9	network	network	NOUN
easat-11054	264	10	traffic	traffic	NOUN
easat-11054	264	11	features	feature	NOUN
easat-11054	264	12	,	,	PUNCT
easat-11054	264	13	which	which	PRON
easat-11054	264	14	are	be	AUX
easat-11054	264	15	then	then	ADV
easat-11054	264	16	transmitted	transmit	VERB
easat-11054	264	17	to	to	ADP
easat-11054	264	18	a	a	DET
easat-11054	264	19	dense	dense	ADJ
easat-11054	264	20	layer	layer	NOUN
easat-11054	264	21	with	with	ADP
easat-11054	264	22	64	64	NUM
easat-11054	264	23	neurons	neuron	NOUN
easat-11054	264	24	and	and	CCONJ
easat-11054	264	25	relu	relu	NOUN
easat-11054	264	26	activation	activation	NOUN
easat-11054	264	27	.	.	PUNCT
easat-11054	265	1	multi	multi	ADJ
easat-11054	265	2	-	-	ADJ
easat-11054	265	3	class	class	ADJ
easat-11054	265	4	classification	classification	NOUN
easat-11054	265	5	(	(	PUNCT
easat-11054	265	6	19	19	NUM
easat-11054	265	7	classes	class	NOUN
easat-11054	265	8	for	for	ADP
easat-11054	265	9	class1	class1	NOUN
easat-11054	265	10	)	)	PUNCT
easat-11054	265	11	is	be	AUX
easat-11054	265	12	achieved	achieve	VERB
easat-11054	265	13	through	through	ADP
easat-11054	265	14	a	a	DET
easat-11054	265	15	softmax	softmax	ADJ
easat-11054	265	16	activation	activation	NOUN
easat-11054	265	17	in	in	ADP
easat-11054	265	18	the	the	DET
easat-11054	265	19	policy	policy	NOUN
easat-11054	265	20	output	output	NOUN
easat-11054	265	21	,	,	PUNCT
easat-11054	265	22	and	and	CCONJ
easat-11054	265	23	the	the	DET
easat-11054	265	24	value	value	NOUN
easat-11054	265	25	output	output	NOUN
easat-11054	265	26	provides	provide	VERB
easat-11054	265	27	an	an	DET
easat-11054	265	28	estimated	estimated	ADJ
easat-11054	265	29	state	state	NOUN
easat-11054	265	30	evaluation	evaluation	NOUN
easat-11054	265	31	in	in	ADP
easat-11054	265	32	the	the	DET
easat-11054	265	33	a3c	a3c	NUM
easat-11054	265	34	-	-	PUNCT
easat-11054	265	35	ffnn	ffnn	NOUN
easat-11054	265	36	algorithm	algorithm	NOUN
easat-11054	265	37	.	.	PUNCT
easat-11054	266	1	the	the	DET
easat-11054	266	2	dt	dt	PROPN
easat-11054	266	3	algorithm	algorithm	PROPN
easat-11054	266	4	provides	provide	VERB
easat-11054	266	5	deterministic	deterministic	ADJ
easat-11054	266	6	decision	decision	NOUN
easat-11054	266	7	-	-	PUNCT
easat-11054	266	8	making	making	NOUN
easat-11054	266	9	of	of	ADP
easat-11054	266	10	the	the	DET
easat-11054	266	11	probabilities	probability	NOUN
easat-11054	266	12	of	of	ADP
easat-11054	266	13	intrusion	intrusion	NOUN
easat-11054	266	14	classes	class	NOUN
easat-11054	266	15	.	.	PUNCT
easat-11054	267	1	finally	finally	ADV
easat-11054	267	2	,	,	PUNCT
easat-11054	267	3	the	the	DET
easat-11054	267	4	final	final	ADJ
easat-11054	267	5	predictions	prediction	NOUN
easat-11054	267	6	are	be	AUX
easat-11054	267	7	generated	generate	VERB
easat-11054	267	8	by	by	ADP
easat-11054	267	9	combining	combine	VERB
easat-11054	267	10	outputs	output	NOUN
easat-11054	267	11	from	from	ADP
easat-11054	267	12	both	both	CCONJ
easat-11054	267	13	algorithms	algorithm	NOUN
easat-11054	267	14	and	and	CCONJ
easat-11054	267	15	applying	apply	VERB
easat-11054	267	16	additional	additional	ADJ
easat-11054	267	17	dense	dense	ADJ
easat-11054	267	18	layers	layer	NOUN
easat-11054	267	19	.	.	PUNCT
easat-11054	268	1	this	this	DET
easat-11054	268	2	training	training	NOUN
easat-11054	268	3	process	process	NOUN
easat-11054	268	4	uses	use	VERB
easat-11054	268	5	an	an	DET
easat-11054	268	6	adam	adam	PROPN
easat-11054	268	7	optimizer	optimizer	NOUN
easat-11054	268	8	with	with	ADP
easat-11054	268	9	a	a	DET
easat-11054	268	10	0.001	0.001	NUM
easat-11054	268	11	learning	learning	NOUN
easat-11054	268	12	rate	rate	NOUN
easat-11054	268	13	,	,	PUNCT
easat-11054	268	14	takes	take	VERB
easat-11054	268	15	early	early	ADJ
easat-11054	268	16	stops	stop	NOUN
easat-11054	268	17	,	,	PUNCT
easat-11054	268	18	and	and	CCONJ
easat-11054	268	19	requires	require	VERB
easat-11054	268	20	at	at	ADV
easat-11054	268	21	least	least	ADJ
easat-11054	268	22	35	35	NUM
easat-11054	268	23	epochs	epoch	NOUN
easat-11054	268	24	to	to	PART
easat-11054	268	25	reduce	reduce	VERB
easat-11054	268	26	overfitting	overfitte	VERB
easat-11054	268	27	.	.	PUNCT
easat-11054	269	1	in	in	ADP
easat-11054	269	2	addition	addition	NOUN
easat-11054	269	3	,	,	PUNCT
easat-11054	269	4	the	the	DET
easat-11054	269	5	algorithm	algorithm	NOUN
easat-11054	269	6	is	be	AUX
easat-11054	269	7	made	make	VERB
easat-11054	269	8	more	more	ADV
easat-11054	269	9	robust	robust	ADJ
easat-11054	269	10	and	and	CCONJ
easat-11054	269	11	can	can	AUX
easat-11054	269	12	be	be	AUX
easat-11054	269	13	generalized	generalize	VERB
easat-11054	269	14	by	by	ADP
easat-11054	269	15	training	training	NOUN
easat-11054	269	16	on	on	ADP
easat-11054	269	17	the	the	DET
easat-11054	269	18	dataset	dataset	NOUN
easat-11054	269	19	to	to	PART
easat-11054	269	20	ensure	ensure	VERB
easat-11054	269	21	fair	fair	ADJ
easat-11054	269	22	representation	representation	NOUN
easat-11054	269	23	across	across	ADP
easat-11054	269	24	all	all	DET
easat-11054	269	25	attack	attack	NOUN
easat-11054	269	26	classes	class	NOUN
easat-11054	269	27	.	.	PUNCT
easat-11054	270	1	4.3	4.3	NUM
easat-11054	270	2	.	.	PUNCT
easat-11054	270	3	phase	phase	NOUN
easat-11054	270	4	3	3	NUM
easat-11054	270	5	:	:	PUNCT
easat-11054	270	6	performance	performance	NOUN
easat-11054	270	7	evaluation	evaluation	NOUN
easat-11054	270	8	.	.	PUNCT
easat-11054	271	1	standard	standard	ADJ
easat-11054	271	2	metrics	metric	NOUN
easat-11054	271	3	such	such	ADJ
easat-11054	271	4	as	as	ADP
easat-11054	271	5	accuracy	accuracy	NOUN
easat-11054	271	6	,	,	PUNCT
easat-11054	271	7	precision	precision	NOUN
easat-11054	271	8	,	,	PUNCT
easat-11054	271	9	recall	recall	NOUN
easat-11054	271	10	,	,	PUNCT
easat-11054	271	11	and	and	CCONJ
easat-11054	271	12	f1	f1	NOUN
easat-11054	271	13	-	-	PUNCT
easat-11054	271	14	score	score	NOUN
easat-11054	271	15	are	be	AUX
easat-11054	271	16	utilized	utilize	VERB
easat-11054	271	17	to	to	PART
easat-11054	271	18	measure	measure	VERB
easat-11054	271	19	the	the	DET
easat-11054	271	20	performance	performance	NOUN
easat-11054	271	21	of	of	ADP
easat-11054	271	22	the	the	DET
easat-11054	271	23	proposed	propose	VERB
easat-11054	271	24	hadt	hadt	NOUN
easat-11054	271	25	algorithm	algorithm	NOUN
easat-11054	271	26	.	.	PUNCT
easat-11054	272	1	the	the	DET
easat-11054	272	2	test	test	NOUN
easat-11054	272	3	set	set	NOUN
easat-11054	272	4	is	be	AUX
easat-11054	272	5	subjected	subject	VERB
easat-11054	272	6	to	to	ADP
easat-11054	272	7	an	an	DET
easat-11054	272	8	evaluation	evaluation	NOUN
easat-11054	272	9	,	,	PUNCT
easat-11054	272	10	and	and	CCONJ
easat-11054	272	11	the	the	DET
easat-11054	272	12	outcomes	outcome	NOUN
easat-11054	272	13	are	be	AUX
easat-11054	272	14	compared	compare	VERB
easat-11054	272	15	with	with	ADP
easat-11054	272	16	baseline	baseline	ADJ
easat-11054	272	17	algorithms	algorithm	NOUN
easat-11054	272	18	like	like	ADP
easat-11054	272	19	the	the	DET
easat-11054	272	20	a3c	a3c	NUM
easat-11054	272	21	-	-	PUNCT
easat-11054	272	22	cnn	cnn	PROPN
easat-11054	272	23	and	and	CCONJ
easat-11054	272	24	a3c	a3c	VERB
easat-11054	272	25	-	-	PUNCT
easat-11054	272	26	ffnn	ffnn	NOUN
easat-11054	272	27	.	.	PUNCT
easat-11054	273	1	accuracy	accuracy	NOUN
easat-11054	273	2	measures	measure	NOUN
easat-11054	273	3	the	the	DET
easat-11054	273	4	algorithm	algorithm	NOUN
easat-11054	273	5	's	's	PART
easat-11054	273	6	overall	overall	ADJ
easat-11054	273	7	correctness	correctness	NOUN
easat-11054	273	8	,	,	PUNCT
easat-11054	273	9	while	while	SCONJ
easat-11054	273	10	precision	precision	NOUN
easat-11054	273	11	denotes	denote	VERB
easat-11054	273	12	the	the	DET
easat-11054	273	13	percentage	percentage	NOUN
easat-11054	273	14	of	of	ADP
easat-11054	273	15	correctly	correctly	ADV
easat-11054	273	16	identified	identify	VERB
easat-11054	273	17	attacks	attack	NOUN
easat-11054	273	18	based	base	VERB
easat-11054	273	19	on	on	ADP
easat-11054	273	20	all	all	DET
easat-11054	273	21	predicted	predict	VERB
easat-11054	273	22	attacks	attack	NOUN
easat-11054	273	23	.	.	PUNCT
easat-11054	274	1	the	the	DET
easat-11054	274	2	ability	ability	NOUN
easat-11054	274	3	of	of	ADP
easat-11054	274	4	the	the	DET
easat-11054	274	5	proposed	propose	VERB
easat-11054	274	6	hadt	hadt	NOUN
easat-11054	274	7	algorithm	algorithm	NOUN
easat-11054	274	8	to	to	PART
easat-11054	274	9	identify	identify	VERB
easat-11054	274	10	every	every	DET
easat-11054	274	11	attack	attack	NOUN
easat-11054	274	12	is	be	AUX
easat-11054	274	13	assessed	assess	VERB
easat-11054	274	14	by	by	ADP
easat-11054	274	15	recall	recall	NOUN
easat-11054	274	16	,	,	PUNCT
easat-11054	274	17	and	and	CCONJ
easat-11054	274	18	the	the	DET
easat-11054	274	19	f1	f1	ADJ
easat-11054	274	20	-	-	PUNCT
easat-11054	274	21	score	score	NOUN
easat-11054	274	22	balances	balance	NOUN
easat-11054	274	23	both	both	DET
easat-11054	274	24	accuracy	accuracy	NOUN
easat-11054	274	25	and	and	CCONJ
easat-11054	274	26	recall	recall	VERB
easat-11054	274	27	to	to	PART
easat-11054	274	28	provide	provide	VERB
easat-11054	274	29	a	a	DET
easat-11054	274	30	comprehensive	comprehensive	ADJ
easat-11054	274	31	performance	performance	NOUN
easat-11054	274	32	score	score	NOUN
easat-11054	274	33	.	.	PUNCT
easat-11054	275	1	5	5	X
easat-11054	275	2	.	.	X
easat-11054	275	3	experiments	experiment	NOUN
easat-11054	275	4	and	and	CCONJ
easat-11054	275	5	results	result	NOUN
easat-11054	275	6	in	in	ADP
easat-11054	275	7	this	this	DET
easat-11054	275	8	section	section	NOUN
easat-11054	275	9	,	,	PUNCT
easat-11054	275	10	the	the	DET
easat-11054	275	11	experiment	experiment	NOUN
easat-11054	275	12	and	and	CCONJ
easat-11054	275	13	results	result	NOUN
easat-11054	275	14	of	of	ADP
easat-11054	275	15	the	the	DET
easat-11054	275	16	proposed	propose	VERB
easat-11054	275	17	hadt	hadt	NOUN
easat-11054	275	18	algorithm	algorithm	NOUN
easat-11054	275	19	for	for	ADP
easat-11054	275	20	detecting	detect	VERB
easat-11054	275	21	intrusions	intrusion	NOUN
easat-11054	275	22	in	in	ADP
easat-11054	275	23	the	the	DET
easat-11054	275	24	iiot	iiot	NOUN
easat-11054	275	25	environment	environment	NOUN
easat-11054	275	26	will	will	AUX
easat-11054	275	27	be	be	AUX
easat-11054	275	28	explained	explain	VERB
easat-11054	275	29	using	use	VERB
easat-11054	275	30	multiple	multiple	ADJ
easat-11054	275	31	experiments	experiment	NOUN
easat-11054	275	32	on	on	ADP
easat-11054	275	33	the	the	DET
easat-11054	275	34	x	x	ADJ
easat-11054	275	35	-	-	ADJ
easat-11054	275	36	iiotid	iiotid	ADJ
easat-11054	275	37	dataset	dataset	NOUN
easat-11054	275	38	.	.	PUNCT
easat-11054	276	1	this	this	DET
easat-11054	276	2	algorithm	algorithm	NOUN
easat-11054	276	3	was	be	AUX
easat-11054	276	4	evaluated	evaluate	VERB
easat-11054	276	5	in	in	ADP
easat-11054	276	6	comparison	comparison	NOUN
easat-11054	276	7	to	to	ADP
easat-11054	276	8	recent	recent	ADJ
easat-11054	276	9	a3c	a3c	NUM
easat-11054	276	10	models	model	NOUN
easat-11054	276	11	based	base	VERB
easat-11054	276	12	on	on	ADP
easat-11054	276	13	cnn	cnn	PROPN
easat-11054	276	14	and	and	CCONJ
easat-11054	276	15	ffnn	ffnn	NOUN
easat-11054	276	16	layers	layer	NOUN
easat-11054	276	17	,	,	PUNCT
easat-11054	276	18	to	to	PART
easat-11054	276	19	demonstrate	demonstrate	VERB
easat-11054	276	20	the	the	DET
easat-11054	276	21	effectiveness	effectiveness	NOUN
easat-11054	276	22	of	of	ADP
easat-11054	276	23	detecting	detect	VERB
easat-11054	276	24	a	a	DET
easat-11054	276	25	wide	wide	ADJ
easat-11054	276	26	range	range	NOUN
easat-11054	276	27	of	of	ADP
easat-11054	276	28	cyberattack	cyberattack	NOUN
easat-11054	276	29	classes	class	NOUN
easat-11054	276	30	with	with	ADP
easat-11054	276	31	high	high	ADJ
easat-11054	276	32	accuracy	accuracy	NOUN
easat-11054	276	33	.	.	PUNCT
easat-11054	277	1	1064	1064	NUM
easat-11054	277	2	edelweiss	edelweiss	PROPN
easat-11054	277	3	applied	apply	VERB
easat-11054	277	4	science	science	NOUN
easat-11054	277	5	and	and	CCONJ
easat-11054	277	6	technology	technology	NOUN
easat-11054	277	7	issn	issn	PROPN
easat-11054	277	8	:	:	PUNCT
easat-11054	277	9	2576	2576	NUM
easat-11054	277	10	-	-	SYM
easat-11054	277	11	8484	8484	NUM
easat-11054	277	12	vol	vol	NOUN
easat-11054	277	13	.	.	PROPN
easat-11054	278	1	9	9	NUM
easat-11054	278	2	,	,	PUNCT
easat-11054	278	3	no	no	INTJ
easat-11054	278	4	.	.	NOUN
easat-11054	278	5	11	11	NUM
easat-11054	278	6	:	:	PUNCT
easat-11054	278	7	1053	1053	NUM
easat-11054	278	8	-	-	SYM
easat-11054	278	9	1075	1075	NUM
easat-11054	278	10	,	,	PUNCT
easat-11054	278	11	2025	2025	NUM
easat-11054	278	12	doi	doi	NOUN
easat-11054	278	13	:	:	PUNCT
easat-11054	278	14	10.55214/2576	10.55214/2576	NUM
easat-11054	278	15	-	-	SYM
easat-11054	278	16	8484.v9i11.11054	8484.v9i11.11054	NOUN
easat-11054	279	1	©	©	PROPN
easat-11054	279	2	2025	2025	NUM
easat-11054	279	3	by	by	ADP
easat-11054	279	4	the	the	DET
easat-11054	279	5	authors	author	NOUN
easat-11054	279	6	;	;	PUNCT
easat-11054	279	7	licensee	licensee	PROPN
easat-11054	279	8	learning	learning	NOUN
easat-11054	279	9	gate	gate	NOUN
easat-11054	279	10	5.1	5.1	NUM
easat-11054	279	11	.	.	PUNCT
easat-11054	280	1	experimental	experimental	ADJ
easat-11054	280	2	environment	environment	NOUN
easat-11054	280	3	and	and	CCONJ
easat-11054	280	4	hyperparameter	hyperparameter	NOUN
easat-11054	280	5	configuration	configuration	NOUN
easat-11054	280	6	the	the	DET
easat-11054	280	7	hadt	hadt	NOUN
easat-11054	280	8	algorithm	algorithm	NOUN
easat-11054	280	9	was	be	AUX
easat-11054	280	10	created	create	VERB
easat-11054	280	11	using	use	VERB
easat-11054	280	12	tensorflow	tensorflow	NOUN
easat-11054	280	13	,	,	PUNCT
easat-11054	280	14	similar	similar	ADJ
easat-11054	280	15	to	to	ADP
easat-11054	280	16	other	other	ADJ
easat-11054	280	17	drl	drl	PROPN
easat-11054	280	18	experiments	experiment	NOUN
easat-11054	280	19	currently	currently	ADV
easat-11054	280	20	available	available	ADJ
easat-11054	280	21	.	.	PUNCT
easat-11054	281	1	all	all	DET
easat-11054	281	2	experiments	experiment	NOUN
easat-11054	281	3	were	be	AUX
easat-11054	281	4	conducted	conduct	VERB
easat-11054	281	5	using	use	VERB
easat-11054	281	6	tensorflow	tensorflow	NOUN
easat-11054	281	7	with	with	ADP
easat-11054	281	8	gpu	gpu	NOUN
easat-11054	281	9	support	support	NOUN
easat-11054	281	10	.	.	PUNCT
easat-11054	282	1	it	it	PRON
easat-11054	282	2	ran	run	VERB
easat-11054	282	3	on	on	ADP
easat-11054	282	4	a	a	DET
easat-11054	282	5	64	64	NUM
easat-11054	282	6	-	-	PUNCT
easat-11054	282	7	bit	bit	NOUN
easat-11054	282	8	windows	window	NOUN
easat-11054	282	9	10	10	NUM
easat-11054	282	10	laptop	laptop	NOUN
easat-11054	282	11	with	with	ADP
easat-11054	282	12	32	32	NUM
easat-11054	282	13	gb	gb	NOUN
easat-11054	282	14	of	of	ADP
easat-11054	282	15	ram	ram	NOUN
easat-11054	282	16	,	,	PUNCT
easat-11054	282	17	an	an	DET
easat-11054	282	18	intel	intel	PROPN
easat-11054	282	19	core	core	NOUN
easat-11054	282	20	i7	i7	NOUN
easat-11054	282	21	-	-	PUNCT
easat-11054	282	22	8850	8850	NUM
easat-11054	282	23	2.70ghz	2.70ghz	NUM
easat-11054	282	24	cpu	cpu	NOUN
easat-11054	282	25	,	,	PUNCT
easat-11054	282	26	and	and	CCONJ
easat-11054	282	27	an	an	DET
easat-11054	282	28	nvidia	nvidia	PROPN
easat-11054	282	29	quadro	quadro	PROPN
easat-11054	282	30	p3200	p3200	PROPN
easat-11054	282	31	gpu	gpu	PROPN
easat-11054	282	32	.	.	PUNCT
easat-11054	282	33	utilizing	utilize	VERB
easat-11054	282	34	the	the	DET
easat-11054	282	35	x	x	ADJ
easat-11054	282	36	-	-	ADJ
easat-11054	282	37	iiotid	iiotid	ADJ
easat-11054	282	38	dataset	dataset	NOUN
easat-11054	282	39	,	,	PUNCT
easat-11054	282	40	the	the	DET
easat-11054	282	41	assessments	assessment	NOUN
easat-11054	282	42	have	have	AUX
easat-11054	282	43	been	be	AUX
easat-11054	282	44	carried	carry	VERB
easat-11054	282	45	out	out	ADP
easat-11054	282	46	.	.	PUNCT
easat-11054	283	1	this	this	DET
easat-11054	283	2	dataset	dataset	NOUN
easat-11054	283	3	is	be	AUX
easat-11054	283	4	considered	consider	VERB
easat-11054	283	5	a	a	DET
easat-11054	283	6	benchmark	benchmark	NOUN
easat-11054	283	7	in	in	ADP
easat-11054	283	8	the	the	DET
easat-11054	283	9	context	context	NOUN
easat-11054	283	10	of	of	ADP
easat-11054	283	11	the	the	DET
easat-11054	283	12	iiot	iiot	ADJ
easat-11054	283	13	-	-	PUNCT
easat-11054	283	14	ids	ids	NOUN
easat-11054	283	15	environment	environment	NOUN
easat-11054	283	16	.	.	PUNCT
easat-11054	284	1	it	it	PRON
easat-11054	284	2	also	also	ADV
easat-11054	284	3	facilitates	facilitate	VERB
easat-11054	284	4	the	the	DET
easat-11054	284	5	comparison	comparison	NOUN
easat-11054	284	6	of	of	ADP
easat-11054	284	7	results	result	NOUN
easat-11054	284	8	and	and	CCONJ
easat-11054	284	9	supports	support	VERB
easat-11054	284	10	the	the	DET
easat-11054	284	11	evaluation	evaluation	NOUN
easat-11054	284	12	of	of	ADP
easat-11054	284	13	methods	method	NOUN
easat-11054	284	14	against	against	ADP
easat-11054	284	15	previous	previous	ADJ
easat-11054	284	16	studies	study	NOUN
easat-11054	284	17	.	.	PUNCT
easat-11054	285	1	in	in	ADP
easat-11054	285	2	this	this	DET
easat-11054	285	3	article	article	NOUN
easat-11054	285	4	,	,	PUNCT
easat-11054	285	5	the	the	DET
easat-11054	285	6	hyperparameters	hyperparameter	NOUN
easat-11054	285	7	are	be	AUX
easat-11054	285	8	carefully	carefully	ADV
easat-11054	285	9	defined	define	VERB
easat-11054	285	10	and	and	CCONJ
easat-11054	285	11	tuned	tune	VERB
easat-11054	285	12	as	as	SCONJ
easat-11054	285	13	listed	list	VERB
easat-11054	285	14	in	in	ADP
easat-11054	285	15	table	table	NOUN
easat-11054	285	16	3	3	NUM
easat-11054	285	17	to	to	PART
easat-11054	285	18	enhance	enhance	VERB
easat-11054	285	19	the	the	DET
easat-11054	285	20	performance	performance	NOUN
easat-11054	285	21	of	of	ADP
easat-11054	285	22	our	our	PRON
easat-11054	285	23	hadt	hadt	NOUN
easat-11054	285	24	algorithm	algorithm	NOUN
easat-11054	285	25	for	for	ADP
easat-11054	285	26	intrusion	intrusion	NOUN
easat-11054	285	27	detection	detection	NOUN
easat-11054	285	28	.	.	PUNCT
easat-11054	286	1	the	the	DET
easat-11054	286	2	number	number	NOUN
easat-11054	286	3	of	of	ADP
easat-11054	286	4	actions	action	NOUN
easat-11054	286	5	(	(	PUNCT
easat-11054	286	6	num_actions	num_action	NOUN
easat-11054	286	7	=	=	SYM
easat-11054	286	8	19	19	NUM
easat-11054	286	9	)	)	PUNCT
easat-11054	286	10	determines	determine	VERB
easat-11054	286	11	the	the	DET
easat-11054	286	12	possible	possible	ADJ
easat-11054	286	13	decisions	decision	NOUN
easat-11054	286	14	the	the	DET
easat-11054	286	15	algorithm	algorithm	NOUN
easat-11054	286	16	can	can	AUX
easat-11054	286	17	take	take	VERB
easat-11054	286	18	to	to	PART
easat-11054	286	19	identify	identify	VERB
easat-11054	286	20	the	the	DET
easat-11054	286	21	classes	class	NOUN
easat-11054	286	22	of	of	ADP
easat-11054	286	23	attacks	attack	NOUN
easat-11054	286	24	.	.	PUNCT
easat-11054	287	1	different	different	ADJ
easat-11054	287	2	learning	learning	NOUN
easat-11054	287	3	rates	rate	NOUN
easat-11054	287	4	were	be	AUX
easat-11054	287	5	used	use	VERB
easat-11054	287	6	(	(	PUNCT
easat-11054	287	7	0.001	0.001	NUM
easat-11054	287	8	for	for	ADP
easat-11054	287	9	the	the	DET
easat-11054	287	10	a3c	a3c	NUM
easat-11054	287	11	-	-	PUNCT
easat-11054	287	12	ffnn	ffnn	NOUN
easat-11054	287	13	and	and	CCONJ
easat-11054	287	14	0.0005	0.0005	NUM
easat-11054	287	15	for	for	ADP
easat-11054	287	16	the	the	DET
easat-11054	287	17	ffnn	ffnn	NOUN
easat-11054	287	18	)	)	PUNCT
easat-11054	287	19	for	for	ADP
easat-11054	287	20	optimization	optimization	NOUN
easat-11054	287	21	,	,	PUNCT
easat-11054	287	22	which	which	PRON
easat-11054	287	23	helps	help	VERB
easat-11054	287	24	balance	balance	NOUN
easat-11054	287	25	convergence	convergence	NOUN
easat-11054	287	26	speed	speed	NOUN
easat-11054	287	27	and	and	CCONJ
easat-11054	287	28	stability	stability	NOUN
easat-11054	287	29	.	.	PUNCT
easat-11054	288	1	the	the	DET
easat-11054	288	2	neural	neural	ADJ
easat-11054	288	3	network	network	NOUN
easat-11054	288	4	layers	layer	NOUN
easat-11054	288	5	in	in	ADP
easat-11054	288	6	a3c	a3c	ADJ
easat-11054	288	7	workers	worker	NOUN
easat-11054	288	8	were	be	AUX
easat-11054	288	9	trained	train	VERB
easat-11054	288	10	for	for	ADP
easat-11054	288	11	50	50	NUM
easat-11054	288	12	epochs	epoch	NOUN
easat-11054	288	13	with	with	ADP
easat-11054	288	14	a	a	DET
easat-11054	288	15	batch	batch	NOUN
easat-11054	288	16	size	size	NOUN
easat-11054	288	17	of	of	ADP
easat-11054	288	18	512	512	NUM
easat-11054	288	19	to	to	PART
easat-11054	288	20	ensure	ensure	VERB
easat-11054	288	21	the	the	DET
easat-11054	288	22	efficiency	efficiency	NOUN
easat-11054	288	23	of	of	ADP
easat-11054	288	24	learning	learn	VERB
easat-11054	288	25	.	.	PUNCT
easat-11054	289	1	early	early	ADJ
easat-11054	289	2	stopping	stopping	NOUN
easat-11054	289	3	was	be	AUX
easat-11054	289	4	enabled	enable	VERB
easat-11054	289	5	with	with	ADP
easat-11054	289	6	a	a	DET
easat-11054	289	7	patience	patience	NOUN
easat-11054	289	8	value	value	NOUN
easat-11054	289	9	of	of	ADP
easat-11054	289	10	5	5	NUM
easat-11054	289	11	to	to	PART
easat-11054	289	12	prevent	prevent	VERB
easat-11054	289	13	overfitting	overfitting	NOUN
easat-11054	289	14	.	.	PUNCT
easat-11054	290	1	a	a	DET
easat-11054	290	2	dense	dense	ADJ
easat-11054	290	3	layer	layer	NOUN
easat-11054	290	4	with	with	ADP
easat-11054	290	5	64	64	NUM
easat-11054	290	6	hidden	hide	VERB
easat-11054	290	7	units	unit	NOUN
easat-11054	290	8	and	and	CCONJ
easat-11054	290	9	a	a	DET
easat-11054	290	10	dropout	dropout	NOUN
easat-11054	290	11	rate	rate	NOUN
easat-11054	290	12	of	of	ADP
easat-11054	290	13	0.5	0.5	NUM
easat-11054	290	14	was	be	AUX
easat-11054	290	15	then	then	ADV
easat-11054	290	16	used	use	VERB
easat-11054	290	17	for	for	ADP
easat-11054	290	18	regularization	regularization	NOUN
easat-11054	290	19	.	.	PUNCT
easat-11054	291	1	the	the	DET
easat-11054	291	2	a3c	a3c	NUM
easat-11054	291	3	-	-	PUNCT
easat-11054	291	4	ffnn	ffnn	NOUN
easat-11054	291	5	optimizes	optimize	VERB
easat-11054	291	6	the	the	DET
easat-11054	291	7	policy	policy	NOUN
easat-11054	291	8	loss	loss	NOUN
easat-11054	291	9	by	by	ADP
easat-11054	291	10	using	use	VERB
easat-11054	291	11	sparse	sparse	ADJ
easat-11054	291	12	categorical	categorical	ADJ
easat-11054	291	13	cross	cross	NOUN
easat-11054	291	14	-	-	NOUN
easat-11054	291	15	entropy	entropy	NOUN
easat-11054	291	16	and	and	CCONJ
easat-11054	291	17	the	the	DET
easat-11054	291	18	value	value	NOUN
easat-11054	291	19	loss	loss	NOUN
easat-11054	291	20	using	use	VERB
easat-11054	291	21	mean	mean	NOUN
easat-11054	291	22	squared	square	VERB
easat-11054	291	23	error	error	NOUN
easat-11054	291	24	,	,	PUNCT
easat-11054	291	25	respectively	respectively	ADV
easat-11054	291	26	.	.	PUNCT
easat-11054	292	1	additionally	additionally	ADV
easat-11054	292	2	,	,	PUNCT
easat-11054	292	3	l2	l2	NOUN
easat-11054	292	4	regularization	regularization	NOUN
easat-11054	292	5	is	be	AUX
easat-11054	292	6	applied	apply	VERB
easat-11054	292	7	in	in	ADP
easat-11054	292	8	the	the	DET
easat-11054	292	9	ffnn	ffnn	NOUN
easat-11054	292	10	to	to	PART
easat-11054	292	11	prevent	prevent	VERB
easat-11054	292	12	overfitting	overfitte	VERB
easat-11054	292	13	and	and	CCONJ
easat-11054	292	14	improve	improve	VERB
easat-11054	292	15	generalization	generalization	NOUN
easat-11054	292	16	.	.	PUNCT
easat-11054	293	1	the	the	DET
easat-11054	293	2	hyperparameter	hyperparameter	NOUN
easat-11054	293	3	configuration	configuration	NOUN
easat-11054	293	4	of	of	ADP
easat-11054	293	5	the	the	DET
easat-11054	293	6	ffnn	ffnn	NOUN
easat-11054	293	7	in	in	ADP
easat-11054	293	8	a3c	a3c	VERB
easat-11054	293	9	-	-	PUNCT
easat-11054	293	10	ffnn	ffnn	NOUN
easat-11054	293	11	and	and	CCONJ
easat-11054	293	12	ffnn	ffnn	NOUN
easat-11054	293	13	layers	layer	NOUN
easat-11054	293	14	is	be	AUX
easat-11054	293	15	identified	identify	VERB
easat-11054	293	16	in	in	ADP
easat-11054	293	17	table	table	NOUN
easat-11054	293	18	3	3	NUM
easat-11054	293	19	.	.	PUNCT
easat-11054	293	20	table	table	NOUN
easat-11054	293	21	3	3	NUM
easat-11054	293	22	.	.	PUNCT
easat-11054	293	23	hyperparameter	hyperparameter	NOUN
easat-11054	293	24	configuration	configuration	NOUN
easat-11054	293	25	for	for	ADP
easat-11054	293	26	hadt	hadt	NOUN
easat-11054	293	27	algorithm	algorithm	NOUN
easat-11054	293	28	parameter	parameter	NOUN
easat-11054	293	29	description	description	NOUN
easat-11054	293	30	value	value	NOUN
easat-11054	293	31	num	num	NOUN
easat-11054	293	32	of	of	ADP
easat-11054	293	33	actions	action	NOUN
easat-11054	293	34	number	number	NOUN
easat-11054	293	35	of	of	ADP
easat-11054	293	36	possible	possible	ADJ
easat-11054	293	37	actions	action	NOUN
easat-11054	293	38	in	in	ADP
easat-11054	293	39	the	the	DET
easat-11054	293	40	a3c	a3c	NUM
easat-11054	293	41	algorithm	algorithm	NOUN
easat-11054	293	42	2	2	NUM
easat-11054	293	43	-	-	PUNCT
easat-11054	293	44	binary	binary	NOUN
easat-11054	293	45	classifications	classification	NOUN
easat-11054	293	46	19	19	NUM
easat-11054	293	47	-	-	PUNCT
easat-11054	293	48	multiclassification	multiclassification	NOUN
easat-11054	293	49	learning	learning	NOUN
easat-11054	293	50	rate	rate	NOUN
easat-11054	293	51	learning	learning	NOUN
easat-11054	293	52	rate	rate	NOUN
easat-11054	293	53	for	for	ADP
easat-11054	293	54	optimizer	optimizer	NOUN
easat-11054	293	55	0.001	0.001	NUM
easat-11054	293	56	(	(	PUNCT
easat-11054	293	57	ffnn	ffnn	NOUN
easat-11054	293	58	in	in	ADP
easat-11054	293	59	a3c	a3c	NUM
easat-11054	293	60	)	)	PUNCT
easat-11054	293	61	,	,	PUNCT
easat-11054	293	62	0.0005	0.0005	NUM
easat-11054	293	63	(	(	PUNCT
easat-11054	293	64	ffnn	ffnn	NOUN
easat-11054	293	65	)	)	PUNCT
easat-11054	293	66	epochs	epoch	NOUN
easat-11054	293	67	number	number	NOUN
easat-11054	293	68	of	of	ADP
easat-11054	293	69	training	training	NOUN
easat-11054	293	70	epochs	epoch	VERB
easat-11054	293	71	50	50	NUM
easat-11054	293	72	batch	batch	NOUN
easat-11054	293	73	size	size	NOUN
easat-11054	293	74	number	number	NOUN
easat-11054	293	75	of	of	ADP
easat-11054	293	76	samples	sample	NOUN
easat-11054	293	77	per	per	ADP
easat-11054	293	78	batch	batch	NOUN
easat-11054	293	79	512	512	NUM
easat-11054	293	80	patience	patience	NOUN
easat-11054	293	81	early	early	ADV
easat-11054	293	82	stopping	stop	VERB
easat-11054	293	83	patience	patience	NOUN
easat-11054	293	84	5	5	NUM
easat-11054	293	85	dense	dense	ADJ
easat-11054	293	86	units	unit	NOUN
easat-11054	293	87	number	number	NOUN
easat-11054	293	88	of	of	ADP
easat-11054	293	89	neurons	neuron	NOUN
easat-11054	293	90	in	in	ADP
easat-11054	293	91	the	the	DET
easat-11054	293	92	hidden	hide	VERB
easat-11054	293	93	layer	layer	NOUN
easat-11054	293	94	64	64	NUM
easat-11054	293	95	dropout	dropout	NOUN
easat-11054	293	96	rate	rate	NOUN
easat-11054	293	97	dropout	dropout	NOUN
easat-11054	293	98	rate	rate	NOUN
easat-11054	293	99	for	for	ADP
easat-11054	293	100	regularization	regularization	NOUN
easat-11054	293	101	0.5	0.5	NUM
easat-11054	293	102	random	random	ADJ
easat-11054	293	103	state	state	NOUN
easat-11054	293	104	seeds	seed	NOUN
easat-11054	293	105	for	for	ADP
easat-11054	293	106	reproducibility	reproducibility	NOUN
easat-11054	293	107	42	42	NUM
easat-11054	293	108	test	test	NOUN
easat-11054	293	109	size	size	NOUN
easat-11054	293	110	proportion	proportion	NOUN
easat-11054	293	111	of	of	ADP
easat-11054	293	112	the	the	DET
easat-11054	293	113	dataset	dataset	NOUN
easat-11054	293	114	used	use	VERB
easat-11054	293	115	for	for	ADP
easat-11054	293	116	testing	test	VERB
easat-11054	293	117	0.3	0.3	NUM
easat-11054	293	118	policy	policy	NOUN
easat-11054	293	119	loss	loss	NOUN
easat-11054	293	120	loss	loss	NOUN
easat-11054	293	121	function	function	NOUN
easat-11054	293	122	for	for	ADP
easat-11054	293	123	policy	policy	NOUN
easat-11054	293	124	output	output	NOUN
easat-11054	293	125	sparse	sparse	ADJ
easat-11054	293	126	categorical	categorical	ADJ
easat-11054	293	127	crossentropy	crossentropy	NOUN
easat-11054	293	128	value	value	NOUN
easat-11054	293	129	loss	loss	NOUN
easat-11054	293	130	loss	loss	NOUN
easat-11054	293	131	function	function	NOUN
easat-11054	293	132	for	for	ADP
easat-11054	293	133	value	value	NOUN
easat-11054	293	134	output	output	NOUN
easat-11054	293	135	mean	mean	VERB
easat-11054	293	136	squared	square	VERB
easat-11054	293	137	error	error	NOUN
easat-11054	293	138	kernel	kernel	NOUN
easat-11054	293	139	regularizer	regularizer	NOUN
easat-11054	293	140	regularization	regularization	NOUN
easat-11054	293	141	for	for	ADP
easat-11054	293	142	a	a	DET
easat-11054	293	143	hybrid	hybrid	ADJ
easat-11054	293	144	layer	layer	NOUN
easat-11054	293	145	in	in	ADP
easat-11054	293	146	the	the	DET
easat-11054	293	147	worker	worker	NOUN
easat-11054	293	148	l2	l2	NOUN
easat-11054	293	149	5.2	5.2	NUM
easat-11054	293	150	.	.	PUNCT
easat-11054	294	1	evaluation	evaluation	NOUN
easat-11054	294	2	indicators	indicator	NOUN
easat-11054	294	3	this	this	DET
easat-11054	294	4	section	section	NOUN
easat-11054	294	5	will	will	AUX
easat-11054	294	6	make	make	VERB
easat-11054	294	7	use	use	NOUN
easat-11054	294	8	of	of	ADP
easat-11054	294	9	the	the	DET
easat-11054	294	10	following	follow	VERB
easat-11054	294	11	metrics	metric	NOUN
easat-11054	294	12	:	:	PUNCT
easat-11054	294	13	1	1	X
easat-11054	294	14	)	)	PUNCT
easat-11054	294	15	a	a	DET
easat-11054	294	16	true	true	ADJ
easat-11054	294	17	positive	positive	ADJ
easat-11054	294	18	(	(	PUNCT
easat-11054	294	19	tp	tp	NOUN
easat-11054	294	20	)	)	PUNCT
easat-11054	294	21	is	be	AUX
easat-11054	294	22	an	an	DET
easat-11054	294	23	attack	attack	NOUN
easat-11054	294	24	data	data	NOUN
easat-11054	294	25	point	point	NOUN
easat-11054	294	26	that	that	PRON
easat-11054	294	27	is	be	AUX
easat-11054	294	28	correctly	correctly	ADV
easat-11054	294	29	classified	classify	VERB
easat-11054	294	30	as	as	ADP
easat-11054	294	31	an	an	DET
easat-11054	294	32	attack	attack	NOUN
easat-11054	294	33	.	.	PUNCT
easat-11054	295	1	data	datum	NOUN
easat-11054	295	2	that	that	PRON
easat-11054	295	3	is	be	AUX
easat-11054	295	4	seen	see	VERB
easat-11054	295	5	as	as	ADP
easat-11054	295	6	normal	normal	ADJ
easat-11054	295	7	yet	yet	ADV
easat-11054	295	8	is	be	AUX
easat-11054	295	9	incorrectly	incorrectly	ADV
easat-11054	295	10	classified	classify	VERB
easat-11054	295	11	as	as	SCONJ
easat-11054	295	12	an	an	DET
easat-11054	295	13	attack	attack	NOUN
easat-11054	295	14	is	be	AUX
easat-11054	295	15	known	know	VERB
easat-11054	295	16	as	as	ADP
easat-11054	295	17	a	a	DET
easat-11054	295	18	false	false	ADJ
easat-11054	295	19	positive	positive	ADJ
easat-11054	295	20	(	(	PUNCT
easat-11054	295	21	fp	fp	NOUN
easat-11054	295	22	)	)	PUNCT
easat-11054	295	23	.	.	PUNCT
easat-11054	296	1	2	2	X
easat-11054	296	2	)	)	PUNCT
easat-11054	296	3	true	true	ADJ
easat-11054	296	4	negative	negative	ADJ
easat-11054	296	5	(	(	PUNCT
easat-11054	296	6	tn	tn	NOUN
easat-11054	296	7	):	):	PUNCT
easat-11054	296	8	truly	truly	ADV
easat-11054	296	9	classified	classify	VERB
easat-11054	296	10	as	as	ADP
easat-11054	296	11	normal	normal	ADJ
easat-11054	296	12	data	datum	NOUN
easat-11054	296	13	even	even	ADV
easat-11054	296	14	when	when	SCONJ
easat-11054	296	15	it	it	PRON
easat-11054	296	16	is	be	AUX
easat-11054	296	17	not	not	PART
easat-11054	296	18	.	.	PUNCT
easat-11054	297	1	3	3	X
easat-11054	297	2	)	)	PUNCT
easat-11054	297	3	a	a	DET
easat-11054	297	4	false	false	ADJ
easat-11054	297	5	negative	negative	ADJ
easat-11054	297	6	(	(	PUNCT
easat-11054	297	7	fn	fn	NOUN
easat-11054	297	8	)	)	PUNCT
easat-11054	297	9	is	be	AUX
easat-11054	297	10	attack	attack	NOUN
easat-11054	297	11	data	datum	NOUN
easat-11054	297	12	that	that	PRON
easat-11054	297	13	has	have	AUX
easat-11054	297	14	been	be	AUX
easat-11054	297	15	incorrectly	incorrectly	ADV
easat-11054	297	16	classified	classify	VERB
easat-11054	297	17	as	as	ADP
easat-11054	297	18	normal	normal	ADJ
easat-11054	297	19	.	.	PUNCT
easat-11054	298	1	the	the	DET
easat-11054	298	2	effectiveness	effectiveness	NOUN
easat-11054	298	3	of	of	ADP
easat-11054	298	4	this	this	DET
easat-11054	298	5	proposed	propose	VERB
easat-11054	298	6	solution	solution	NOUN
easat-11054	298	7	will	will	AUX
easat-11054	298	8	be	be	AUX
easat-11054	298	9	evaluated	evaluate	VERB
easat-11054	298	10	using	use	VERB
easat-11054	298	11	the	the	DET
easat-11054	298	12	following	follow	VERB
easat-11054	298	13	metrics	metric	NOUN
easat-11054	298	14	:	:	PUNCT
easat-11054	298	15	the	the	DET
easat-11054	298	16	precision	precision	NOUN
easat-11054	298	17	:	:	PUNCT
easat-11054	298	18	measures	measure	VERB
easat-11054	298	19	the	the	DET
easat-11054	298	20	proportion	proportion	NOUN
easat-11054	298	21	of	of	ADP
easat-11054	298	22	correctly	correctly	ADV
easat-11054	298	23	predicted	predict	VERB
easat-11054	298	24	positive	positive	ADJ
easat-11054	298	25	states	state	NOUN
easat-11054	298	26	among	among	ADP
easat-11054	298	27	all	all	ADV
easat-11054	298	28	predicted	predict	VERB
easat-11054	298	29	positive	positive	ADJ
easat-11054	298	30	states	state	NOUN
easat-11054	298	31	.	.	PUNCT
easat-11054	299	1	it	it	PRON
easat-11054	299	2	is	be	AUX
easat-11054	299	3	given	give	VERB
easat-11054	299	4	by	by	ADP
easat-11054	299	5	equation	equation	NOUN
easat-11054	299	6	(	(	PUNCT
easat-11054	299	7	14	14	NUM
easat-11054	299	8	):	):	PUNCT
easat-11054	299	9	precision	precision	NOUN
easat-11054	299	10	=	=	PUNCT
easat-11054	299	11	𝑇𝑃	𝑇𝑃	PROPN
easat-11054	299	12	t	t	NOUN
easat-11054	299	13	p	p	NOUN
easat-11054	300	1	+	+	CCONJ
easat-11054	300	2	f	f	X
easat-11054	300	3	p	p	X
easat-11054	300	4	(	(	PUNCT
easat-11054	300	5	14	14	NUM
easat-11054	300	6	)	)	PUNCT
easat-11054	300	7	the	the	DET
easat-11054	300	8	accuracy	accuracy	NOUN
easat-11054	300	9	determines	determine	VERB
easat-11054	300	10	the	the	DET
easat-11054	300	11	proportion	proportion	NOUN
easat-11054	300	12	of	of	ADP
easat-11054	300	13	all	all	DET
easat-11054	300	14	states	state	NOUN
easat-11054	300	15	that	that	PRON
easat-11054	300	16	are	be	AUX
easat-11054	300	17	correctly	correctly	ADV
easat-11054	300	18	categorized	categorize	VERB
easat-11054	300	19	.	.	PUNCT
easat-11054	301	1	it	it	PRON
easat-11054	301	2	is	be	AUX
easat-11054	301	3	given	give	VERB
easat-11054	301	4	by	by	ADP
easat-11054	301	5	equation	equation	NOUN
easat-11054	301	6	(	(	PUNCT
easat-11054	301	7	15	15	NUM
easat-11054	301	8	):	):	PUNCT
easat-11054	301	9	accuracy	accuracy	NOUN
easat-11054	301	10	=	=	SYM
easat-11054	301	11	𝑇𝑃+𝑇𝑁	𝑇𝑃+𝑇𝑁	NOUN
easat-11054	301	12	t	t	PROPN
easat-11054	301	13	p	p	PROPN
easat-11054	302	1	+	+	PROPN
easat-11054	302	2	t	t	PROPN
easat-11054	302	3	n	n	PROPN
easat-11054	303	1	+	+	CCONJ
easat-11054	303	2	f	f	PROPN
easat-11054	304	1	p	p	X
easat-11054	305	1	+	+	CCONJ
easat-11054	305	2	f	f	PROPN
easat-11054	305	3	n	n	PROPN
easat-11054	305	4	(	(	PUNCT
easat-11054	305	5	15	15	NUM
easat-11054	305	6	)	)	PUNCT
easat-11054	305	7	the	the	DET
easat-11054	305	8	f1	f1	NOUN
easat-11054	305	9	-	-	PUNCT
easat-11054	305	10	score	score	NOUN
easat-11054	305	11	is	be	AUX
easat-11054	305	12	a	a	DET
easat-11054	305	13	derived	derive	VERB
easat-11054	305	14	effectiveness	effectiveness	NOUN
easat-11054	305	15	statistic	statistic	NOUN
easat-11054	305	16	that	that	PRON
easat-11054	305	17	is	be	AUX
easat-11054	305	18	generated	generate	VERB
easat-11054	305	19	from	from	ADP
easat-11054	305	20	the	the	DET
easat-11054	305	21	harmonic	harmonic	ADJ
easat-11054	305	22	mean	mean	NOUN
easat-11054	305	23	of	of	ADP
easat-11054	305	24	recall	recall	NOUN
easat-11054	305	25	and	and	CCONJ
easat-11054	305	26	accuracy	accuracy	NOUN
easat-11054	305	27	.	.	PUNCT
easat-11054	306	1	it	it	PRON
easat-11054	306	2	is	be	AUX
easat-11054	306	3	given	give	VERB
easat-11054	306	4	by	by	ADP
easat-11054	306	5	equation	equation	NOUN
easat-11054	306	6	(	(	PUNCT
easat-11054	306	7	16	16	NUM
easat-11054	306	8	):	):	PUNCT
easat-11054	306	9	f1	f1	NOUN
easat-11054	306	10	-	-	PUNCT
easat-11054	306	11	score	score	NOUN
easat-11054	306	12	=	=	SYM
easat-11054	306	13	2	2	NUM
easat-11054	306	14	·	·	PUNCT
easat-11054	306	15	p	p	NOUN
easat-11054	306	16	recision	recision	NOUN
easat-11054	306	17	·	·	PUNCT
easat-11054	306	18	recall	recall	VERB
easat-11054	306	19	p	p	PROPN
easat-11054	306	20	recision+recall	recision+recall	PROPN
easat-11054	306	21	(	(	PUNCT
easat-11054	306	22	16	16	NUM
easat-11054	306	23	)	)	PUNCT
easat-11054	306	24	1065	1065	NUM
easat-11054	306	25	edelweiss	edelweiss	PROPN
easat-11054	306	26	applied	apply	VERB
easat-11054	306	27	science	science	NOUN
easat-11054	306	28	and	and	CCONJ
easat-11054	306	29	technology	technology	NOUN
easat-11054	306	30	issn	issn	PROPN
easat-11054	306	31	:	:	PUNCT
easat-11054	306	32	2576	2576	NUM
easat-11054	306	33	-	-	SYM
easat-11054	306	34	8484	8484	NUM
easat-11054	306	35	vol	vol	NOUN
easat-11054	306	36	.	.	PROPN
easat-11054	307	1	9	9	NUM
easat-11054	307	2	,	,	PUNCT
easat-11054	307	3	no	no	INTJ
easat-11054	307	4	.	.	NOUN
easat-11054	307	5	11	11	NUM
easat-11054	307	6	:	:	PUNCT
easat-11054	307	7	1053	1053	NUM
easat-11054	307	8	-	-	SYM
easat-11054	307	9	1075	1075	NUM
easat-11054	307	10	,	,	PUNCT
easat-11054	307	11	2025	2025	NUM
easat-11054	307	12	doi	doi	NOUN
easat-11054	307	13	:	:	PUNCT
easat-11054	307	14	10.55214/2576	10.55214/2576	NUM
easat-11054	307	15	-	-	SYM
easat-11054	307	16	8484.v9i11.11054	8484.v9i11.11054	NOUN
easat-11054	308	1	©	©	PROPN
easat-11054	308	2	2025	2025	NUM
easat-11054	308	3	by	by	ADP
easat-11054	308	4	the	the	DET
easat-11054	308	5	authors	author	NOUN
easat-11054	308	6	;	;	PUNCT
easat-11054	308	7	licensee	licensee	PROPN
easat-11054	308	8	learning	learning	NOUN
easat-11054	308	9	gate	gate	VERB
easat-11054	308	10	the	the	DET
easat-11054	308	11	recall	recall	NOUN
easat-11054	308	12	measures	measure	VERB
easat-11054	308	13	the	the	DET
easat-11054	308	14	proportion	proportion	NOUN
easat-11054	308	15	of	of	ADP
easat-11054	308	16	correctly	correctly	ADV
easat-11054	308	17	predicted	predict	VERB
easat-11054	308	18	positive	positive	ADJ
easat-11054	308	19	instances	instance	NOUN
easat-11054	308	20	out	out	ADP
easat-11054	308	21	of	of	ADP
easat-11054	308	22	all	all	DET
easat-11054	308	23	actual	actual	ADJ
easat-11054	308	24	positive	positive	ADJ
easat-11054	308	25	instances	instance	NOUN
easat-11054	308	26	.	.	PUNCT
easat-11054	309	1	it	it	PRON
easat-11054	309	2	is	be	AUX
easat-11054	309	3	given	give	VERB
easat-11054	309	4	by	by	ADP
easat-11054	309	5	equation	equation	NOUN
easat-11054	309	6	(	(	PUNCT
easat-11054	309	7	18	18	NUM
easat-11054	309	8	):	):	PUNCT
easat-11054	309	9	recall	recall	NOUN
easat-11054	309	10	=	=	SYM
easat-11054	309	11	𝑇𝑃	𝑇𝑃	PROPN
easat-11054	309	12	t	t	NOUN
easat-11054	309	13	p	p	NOUN
easat-11054	310	1	+	+	CCONJ
easat-11054	310	2	f	f	PROPN
easat-11054	310	3	n	n	PROPN
easat-11054	310	4	(	(	PUNCT
easat-11054	310	5	17	17	NUM
easat-11054	310	6	)	)	PUNCT
easat-11054	310	7	on	on	ADP
easat-11054	310	8	the	the	DET
easat-11054	310	9	other	other	ADJ
easat-11054	310	10	hand	hand	NOUN
easat-11054	310	11	,	,	PUNCT
easat-11054	310	12	the	the	DET
easat-11054	310	13	x	x	ADJ
easat-11054	310	14	-	-	ADJ
easat-11054	310	15	iiotid	iiotid	ADJ
easat-11054	310	16	dataset	dataset	NOUN
easat-11054	310	17	was	be	AUX
easat-11054	310	18	evaluated	evaluate	VERB
easat-11054	310	19	in	in	ADP
easat-11054	310	20	two	two	NUM
easat-11054	310	21	test	test	NOUN
easat-11054	310	22	scenarios	scenario	NOUN
easat-11054	310	23	:	:	PUNCT
easat-11054	310	24	1	1	X
easat-11054	310	25	)	)	PUNCT
easat-11054	310	26	binary	binary	ADJ
easat-11054	310	27	classification	classification	NOUN
easat-11054	310	28	(	(	PUNCT
easat-11054	310	29	normal	normal	ADJ
easat-11054	310	30	vs.	vs.	ADP
easat-11054	310	31	intrusion	intrusion	NOUN
easat-11054	310	32	):	):	PUNCT
easat-11054	310	33	the	the	DET
easat-11054	310	34	dataset	dataset	NOUN
easat-11054	310	35	was	be	AUX
easat-11054	310	36	divided	divide	VERB
easat-11054	310	37	into	into	ADP
easat-11054	310	38	two	two	NUM
easat-11054	310	39	categories	category	NOUN
easat-11054	310	40	.	.	PUNCT
easat-11054	311	1	the	the	DET
easat-11054	311	2	hadt	hadt	NOUN
easat-11054	311	3	algorithm	algorithm	NOUN
easat-11054	311	4	was	be	AUX
easat-11054	311	5	trained	train	VERB
easat-11054	311	6	and	and	CCONJ
easat-11054	311	7	tested	test	VERB
easat-11054	311	8	to	to	PART
easat-11054	311	9	classify	classify	VERB
easat-11054	311	10	traffic	traffic	NOUN
easat-11054	311	11	as	as	ADP
easat-11054	311	12	either	either	CCONJ
easat-11054	311	13	normal	normal	ADJ
easat-11054	311	14	or	or	CCONJ
easat-11054	311	15	intrusion	intrusion	NOUN
easat-11054	311	16	.	.	PUNCT
easat-11054	312	1	performance	performance	NOUN
easat-11054	312	2	metrics	metric	NOUN
easat-11054	312	3	,	,	PUNCT
easat-11054	312	4	including	include	VERB
easat-11054	312	5	accuracy	accuracy	NOUN
easat-11054	312	6	,	,	PUNCT
easat-11054	312	7	precision	precision	NOUN
easat-11054	312	8	,	,	PUNCT
easat-11054	312	9	recall	recall	NOUN
easat-11054	312	10	,	,	PUNCT
easat-11054	312	11	and	and	CCONJ
easat-11054	312	12	f1	f1	NOUN
easat-11054	312	13	-	-	PUNCT
easat-11054	312	14	score	score	NOUN
easat-11054	312	15	,	,	PUNCT
easat-11054	312	16	were	be	AUX
easat-11054	312	17	computed	compute	VERB
easat-11054	312	18	to	to	PART
easat-11054	312	19	assess	assess	VERB
easat-11054	312	20	the	the	DET
easat-11054	312	21	detection	detection	NOUN
easat-11054	312	22	effectiveness	effectiveness	NOUN
easat-11054	312	23	of	of	ADP
easat-11054	312	24	the	the	DET
easat-11054	312	25	hadt	hadt	NOUN
easat-11054	312	26	algorithm	algorithm	NOUN
easat-11054	312	27	.	.	PUNCT
easat-11054	313	1	the	the	DET
easat-11054	313	2	experiment	experiment	NOUN
easat-11054	313	3	also	also	ADV
easat-11054	313	4	compared	compare	VERB
easat-11054	313	5	the	the	DET
easat-11054	313	6	classification	classification	NOUN
easat-11054	313	7	performance	performance	NOUN
easat-11054	313	8	of	of	ADP
easat-11054	313	9	a3c	a3c	PROPN
easat-11054	313	10	-	-	PUNCT
easat-11054	313	11	cnn	cnn	NOUN
easat-11054	313	12	and	and	CCONJ
easat-11054	313	13	a3c	a3c	VERB
easat-11054	313	14	-	-	PUNCT
easat-11054	313	15	ffnn	ffnn	NOUN
easat-11054	313	16	algorithms	algorithm	NOUN
easat-11054	313	17	.	.	PUNCT
easat-11054	314	1	2	2	X
easat-11054	314	2	)	)	PUNCT
easat-11054	314	3	nineteen	nineteen	NUM
easat-11054	314	4	-	-	PUNCT
easat-11054	314	5	class	class	NOUN
easat-11054	314	6	classification	classification	NOUN
easat-11054	314	7	:	:	PUNCT
easat-11054	314	8	the	the	DET
easat-11054	314	9	dataset	dataset	NOUN
easat-11054	314	10	was	be	AUX
easat-11054	314	11	analyzed	analyze	VERB
easat-11054	314	12	with	with	ADP
easat-11054	314	13	nineteen	nineteen	NUM
easat-11054	314	14	attack	attack	NOUN
easat-11054	314	15	categories	category	NOUN
easat-11054	314	16	,	,	PUNCT
easat-11054	314	17	including	include	VERB
easat-11054	314	18	normal	normal	ADJ
easat-11054	314	19	,	,	PUNCT
easat-11054	314	20	rdos	rdo	NOUN
easat-11054	314	21	,	,	PUNCT
easat-11054	314	22	generic	generic	ADJ
easat-11054	314	23	scanning	scanning	NOUN
easat-11054	314	24	,	,	PUNCT
easat-11054	314	25	brute	brute	ADJ
easat-11054	314	26	force	force	NOUN
easat-11054	314	27	,	,	PUNCT
easat-11054	314	28	standard	standard	ADJ
easat-11054	314	29	scanning	scanning	NOUN
easat-11054	314	30	,	,	PUNCT
easat-11054	314	31	insider	insider	NOUN
easat-11054	314	32	malicious	malicious	ADJ
easat-11054	314	33	,	,	PUNCT
easat-11054	314	34	modbus	modbus	PROPN
easat-11054	314	35	register	register	PROPN
easat-11054	314	36	reading	reading	PROPN
easat-11054	314	37	,	,	PUNCT
easat-11054	314	38	false	false	ADJ
easat-11054	314	39	data	datum	NOUN
easat-11054	314	40	injection	injection	NOUN
easat-11054	314	41	,	,	PUNCT
easat-11054	314	42	c&c	c&c	PROPN
easat-11054	314	43	,	,	PUNCT
easat-11054	314	44	dictionary	dictionary	ADJ
easat-11054	314	45	,	,	PUNCT
easat-11054	314	46	tcp	tcp	ADJ
easat-11054	314	47	relay	relay	NOUN
easat-11054	314	48	,	,	PUNCT
easat-11054	314	49	fuzzing	fuzzing	NOUN
easat-11054	314	50	,	,	PUNCT
easat-11054	314	51	reverse	reverse	ADJ
easat-11054	314	52	shell	shell	NOUN
easat-11054	314	53	,	,	PUNCT
easat-11054	314	54	cryptoransomware	cryptoransomware	NOUN
easat-11054	314	55	,	,	PUNCT
easat-11054	314	56	mitm	mitm	NOUN
easat-11054	314	57	,	,	PUNCT
easat-11054	314	58	and	and	CCONJ
easat-11054	314	59	fake	fake	ADJ
easat-11054	314	60	notification	notification	NOUN
easat-11054	314	61	.	.	PUNCT
easat-11054	315	1	all	all	DET
easat-11054	315	2	these	these	DET
easat-11054	315	3	nineteen	nineteen	NUM
easat-11054	315	4	kinds	kind	NOUN
easat-11054	315	5	have	have	AUX
easat-11054	315	6	been	be	AUX
easat-11054	315	7	identified	identify	VERB
easat-11054	315	8	in	in	ADP
easat-11054	315	9	this	this	DET
easat-11054	315	10	dataset	dataset	NOUN
easat-11054	315	11	,	,	PUNCT
easat-11054	315	12	and	and	CCONJ
easat-11054	315	13	classification	classification	NOUN
easat-11054	315	14	detection	detection	NOUN
easat-11054	315	15	has	have	AUX
easat-11054	315	16	been	be	AUX
easat-11054	315	17	carried	carry	VERB
easat-11054	315	18	out	out	ADP
easat-11054	315	19	.	.	PUNCT
easat-11054	316	1	the	the	DET
easat-11054	316	2	classification	classification	NOUN
easat-11054	316	3	detection	detection	NOUN
easat-11054	316	4	produced	produce	VERB
easat-11054	316	5	a	a	DET
easat-11054	316	6	nineteendimensional	nineteendimensional	ADJ
easat-11054	316	7	confusion	confusion	NOUN
easat-11054	316	8	matrix	matrix	NOUN
easat-11054	316	9	(	(	PUNCT
easat-11054	316	10	detection	detection	NOUN
easat-11054	316	11	result	result	NOUN
easat-11054	316	12	)	)	PUNCT
easat-11054	316	13	based	base	VERB
easat-11054	316	14	on	on	ADP
easat-11054	316	15	which	which	PRON
easat-11054	316	16	the	the	DET
easat-11054	316	17	accuracy	accuracy	NOUN
easat-11054	316	18	,	,	PUNCT
easat-11054	316	19	precision	precision	NOUN
easat-11054	316	20	,	,	PUNCT
easat-11054	316	21	recall	recall	NOUN
easat-11054	316	22	,	,	PUNCT
easat-11054	316	23	and	and	CCONJ
easat-11054	316	24	f1score	f1score	NOUN
easat-11054	316	25	indicators	indicator	NOUN
easat-11054	316	26	of	of	ADP
easat-11054	316	27	the	the	DET
easat-11054	316	28	hadt	hadt	NOUN
easat-11054	316	29	for	for	ADP
easat-11054	316	30	the	the	DET
easat-11054	316	31	eighteen	eighteen	NUM
easat-11054	316	32	assault	assault	NOUN
easat-11054	316	33	types	type	NOUN
easat-11054	316	34	were	be	AUX
easat-11054	316	35	calculated	calculate	VERB
easat-11054	316	36	.	.	PUNCT
easat-11054	317	1	using	use	VERB
easat-11054	317	2	the	the	DET
easat-11054	317	3	x	x	ADJ
easat-11054	317	4	-	-	ADJ
easat-11054	317	5	iiotid	iiotid	ADJ
easat-11054	317	6	dataset	dataset	NOUN
easat-11054	317	7	with	with	ADP
easat-11054	317	8	the	the	DET
easat-11054	317	9	other	other	ADJ
easat-11054	317	10	a3c	a3c	NUM
easat-11054	317	11	networks	network	NOUN
easat-11054	317	12	,	,	PUNCT
easat-11054	317	13	which	which	PRON
easat-11054	317	14	are	be	AUX
easat-11054	317	15	a3c	a3c	VERB
easat-11054	317	16	-	-	PUNCT
easat-11054	317	17	cnn	cnn	NOUN
easat-11054	317	18	and	and	CCONJ
easat-11054	317	19	a3c	a3c	VERB
easat-11054	317	20	-	-	PUNCT
easat-11054	317	21	ffnn	ffnn	NOUN
easat-11054	317	22	,	,	PUNCT
easat-11054	317	23	commonly	commonly	ADV
easat-11054	317	24	used	use	VERB
easat-11054	317	25	in	in	ADP
easat-11054	317	26	intrusion	intrusion	NOUN
easat-11054	317	27	detection	detection	NOUN
easat-11054	317	28	techniques	technique	NOUN
easat-11054	317	29	,	,	PUNCT
easat-11054	317	30	is	be	AUX
easat-11054	317	31	the	the	DET
easat-11054	317	32	second	second	ADJ
easat-11054	317	33	experiment	experiment	NOUN
easat-11054	317	34	.	.	PUNCT
easat-11054	318	1	the	the	DET
easat-11054	318	2	experiments	experiment	NOUN
easat-11054	318	3	carry	carry	VERB
easat-11054	318	4	out	out	ADP
easat-11054	318	5	testing	testing	NOUN
easat-11054	318	6	and	and	CCONJ
easat-11054	318	7	training	training	NOUN
easat-11054	318	8	,	,	PUNCT
easat-11054	318	9	then	then	ADV
easat-11054	318	10	compare	compare	VERB
easat-11054	318	11	the	the	DET
easat-11054	318	12	test	test	NOUN
easat-11054	318	13	results	result	NOUN
easat-11054	318	14	with	with	ADP
easat-11054	318	15	the	the	DET
easat-11054	318	16	hadt	hadt	NOUN
easat-11054	318	17	detection	detection	NOUN
easat-11054	318	18	results	result	NOUN
easat-11054	318	19	.	.	PUNCT
easat-11054	319	1	5.3	5.3	NUM
easat-11054	319	2	.	.	PUNCT
easat-11054	319	3	training	training	NOUN
easat-11054	319	4	algorithms	algorithm	NOUN
easat-11054	319	5	results	result	VERB
easat-11054	319	6	there	there	PRON
easat-11054	319	7	were	be	VERB
easat-11054	319	8	three	three	NUM
easat-11054	319	9	training	training	NOUN
easat-11054	319	10	and	and	CCONJ
easat-11054	319	11	testing	testing	NOUN
easat-11054	319	12	procedures	procedure	NOUN
easat-11054	319	13	in	in	ADP
easat-11054	319	14	the	the	DET
easat-11054	319	15	experiment	experiment	NOUN
easat-11054	319	16	.	.	PUNCT
easat-11054	320	1	figures	figure	NOUN
easat-11054	320	2	4	4	NUM
easat-11054	320	3	to	to	PART
easat-11054	320	4	7	7	NUM
easat-11054	320	5	illustrate	illustrate	VERB
easat-11054	320	6	the	the	DET
easat-11054	320	7	relationship	relationship	NOUN
easat-11054	320	8	between	between	ADP
easat-11054	320	9	training	training	NOUN
easat-11054	320	10	accuracy	accuracy	NOUN
easat-11054	320	11	and	and	CCONJ
easat-11054	320	12	loss	loss	NOUN
easat-11054	320	13	during	during	ADP
easat-11054	320	14	the	the	DET
easat-11054	320	15	hadt	hadt	NOUN
easat-11054	320	16	iterations	iteration	NOUN
easat-11054	320	17	throughout	throughout	ADP
easat-11054	320	18	the	the	DET
easat-11054	320	19	training	training	NOUN
easat-11054	320	20	process	process	NOUN
easat-11054	320	21	.	.	PUNCT
easat-11054	321	1	the	the	DET
easat-11054	321	2	training	training	NOUN
easat-11054	321	3	and	and	CCONJ
easat-11054	321	4	validation	validation	NOUN
easat-11054	321	5	accuracy	accuracy	NOUN
easat-11054	321	6	of	of	ADP
easat-11054	321	7	the	the	DET
easat-11054	321	8	ffnn	ffnn	NOUN
easat-11054	321	9	layer	layer	NOUN
easat-11054	321	10	in	in	ADP
easat-11054	321	11	the	the	DET
easat-11054	321	12	a3c	a3c	NUM
easat-11054	321	13	-	-	PUNCT
easat-11054	321	14	ffnn	ffnn	NOUN
easat-11054	321	15	algorithm	algorithm	NOUN
easat-11054	321	16	displayed	display	VERB
easat-11054	321	17	stability	stability	NOUN
easat-11054	321	18	and	and	CCONJ
easat-11054	321	19	high	high	ADJ
easat-11054	321	20	performance	performance	NOUN
easat-11054	321	21	over	over	ADP
easat-11054	321	22	the	the	DET
easat-11054	321	23	course	course	NOUN
easat-11054	321	24	of	of	ADP
easat-11054	321	25	50	50	NUM
easat-11054	321	26	epochs	epoch	NOUN
easat-11054	321	27	,	,	PUNCT
easat-11054	321	28	with	with	ADP
easat-11054	321	29	both	both	DET
easat-11054	321	30	values	value	NOUN
easat-11054	321	31	close	close	ADV
easat-11054	321	32	to	to	ADP
easat-11054	321	33	98.5	98.5	NUM
easat-11054	321	34	%	%	NOUN
easat-11054	321	35	,	,	PUNCT
easat-11054	321	36	as	as	SCONJ
easat-11054	321	37	shown	show	VERB
easat-11054	321	38	in	in	ADP
easat-11054	321	39	figure	figure	NOUN
easat-11054	321	40	4	4	NUM
easat-11054	321	41	.	.	PUNCT
easat-11054	321	42	figure	figure	VERB
easat-11054	321	43	4	4	NUM
easat-11054	321	44	.	.	PUNCT
easat-11054	321	45	training	training	NOUN
easat-11054	321	46	and	and	CCONJ
easat-11054	321	47	validation	validation	NOUN
easat-11054	321	48	accuracy	accuracy	NOUN
easat-11054	321	49	curves	curve	NOUN
easat-11054	321	50	for	for	ADP
easat-11054	321	51	the	the	DET
easat-11054	321	52	a3c	a3c	NUM
easat-11054	321	53	-	-	PUNCT
easat-11054	321	54	ffnn	ffnn	NOUN
easat-11054	321	55	.	.	PUNCT
easat-11054	322	1	in	in	ADP
easat-11054	322	2	epoch	epoch	PROPN
easat-11054	322	3	35	35	NUM
easat-11054	322	4	,	,	PUNCT
easat-11054	322	5	the	the	DET
easat-11054	322	6	training	training	NOUN
easat-11054	322	7	accuracy	accuracy	NOUN
easat-11054	322	8	increases	increase	VERB
easat-11054	322	9	to	to	ADP
easat-11054	322	10	a	a	DET
easat-11054	322	11	constant	constant	ADJ
easat-11054	322	12	0.9842	0.9842	NUM
easat-11054	322	13	and	and	CCONJ
easat-11054	322	14	then	then	ADV
easat-11054	322	15	reaches	reach	VERB
easat-11054	322	16	at	at	ADP
easat-11054	322	17	least	least	ADJ
easat-11054	322	18	an	an	DET
easat-11054	322	19	average	average	NOUN
easat-11054	322	20	of	of	ADP
easat-11054	322	21	0.9851	0.9851	NUM
easat-11054	322	22	in	in	ADP
easat-11054	322	23	epoch	epoch	PROPN
easat-11054	322	24	50	50	NUM
easat-11054	322	25	,	,	PUNCT
easat-11054	322	26	suggesting	suggest	VERB
easat-11054	322	27	that	that	SCONJ
easat-11054	322	28	this	this	DET
easat-11054	322	29	layer	layer	NOUN
easat-11054	322	30	is	be	AUX
easat-11054	322	31	effectively	effectively	ADV
easat-11054	322	32	learning	learn	VERB
easat-11054	322	33	from	from	ADP
easat-11054	322	34	the	the	DET
easat-11054	322	35	training	training	NOUN
easat-11054	322	36	data	datum	NOUN
easat-11054	322	37	.	.	PUNCT
easat-11054	323	1	in	in	ADP
easat-11054	323	2	the	the	DET
easat-11054	323	3	same	same	ADJ
easat-11054	323	4	manner	manner	NOUN
easat-11054	323	5	,	,	PUNCT
easat-11054	323	6	validation	validation	NOUN
easat-11054	323	7	accuracy	accuracy	NOUN
easat-11054	323	8	remains	remain	VERB
easat-11054	323	9	unchanged	unchanged	ADJ
easat-11054	323	10	,	,	PUNCT
easat-11054	323	11	with	with	ADP
easat-11054	323	12	a	a	DET
easat-11054	323	13	slight	slight	ADJ
easat-11054	323	14	variation	variation	NOUN
easat-11054	323	15	between	between	ADP
easat-11054	323	16	0.9840	0.9840	NUM
easat-11054	323	17	and	and	CCONJ
easat-11054	323	18	0.9847	0.9847	NUM
easat-11054	323	19	,	,	PUNCT
easat-11054	323	20	indicating	indicate	VERB
easat-11054	323	21	that	that	SCONJ
easat-11054	323	22	the	the	DET
easat-11054	323	23	layer	layer	NOUN
easat-11054	323	24	can	can	AUX
easat-11054	323	25	be	be	AUX
easat-11054	323	26	generalized	generalize	VERB
easat-11054	323	27	to	to	ADP
easat-11054	323	28	this	this	DET
easat-11054	323	29	data	datum	NOUN
easat-11054	323	30	effectively	effectively	ADV
easat-11054	323	31	.	.	PUNCT
easat-11054	324	1	as	as	SCONJ
easat-11054	324	2	shown	show	VERB
easat-11054	324	3	in	in	ADP
easat-11054	324	4	figure	figure	NOUN
easat-11054	324	5	5	5	NUM
easat-11054	324	6	,	,	PUNCT
easat-11054	324	7	the	the	DET
easat-11054	324	8	figure	figure	NOUN
easat-11054	324	9	is	be	AUX
easat-11054	324	10	not	not	PART
easat-11054	324	11	the	the	DET
easat-11054	324	12	same	same	ADJ
easat-11054	324	13	when	when	SCONJ
easat-11054	324	14	it	it	PRON
easat-11054	324	15	comes	come	VERB
easat-11054	324	16	to	to	ADP
easat-11054	324	17	loss	loss	NOUN
easat-11054	324	18	values	value	NOUN
easat-11054	324	19	.	.	PUNCT
easat-11054	325	1	at	at	ADP
easat-11054	325	2	0.9763	0.9763	NUM
easat-11054	325	3	,	,	PUNCT
easat-11054	325	4	the	the	DET
easat-11054	325	5	validation	validation	NOUN
easat-11054	325	6	loss	loss	NOUN
easat-11054	325	7	is	be	AUX
easat-11054	325	8	slightly	slightly	ADV
easat-11054	325	9	higher	high	ADJ
easat-11054	325	10	than	than	ADP
easat-11054	325	11	the	the	DET
easat-11054	325	12	training	training	NOUN
easat-11054	325	13	loss	loss	NOUN
easat-11054	325	14	,	,	PUNCT
easat-11054	325	15	which	which	PRON
easat-11054	325	16	remains	remain	VERB
easat-11054	325	17	steady	steady	ADJ
easat-11054	325	18	at	at	ADP
easat-11054	325	19	0.8592	0.8592	NUM
easat-11054	325	20	.	.	PUNCT
easat-11054	326	1	despite	despite	SCONJ
easat-11054	326	2	the	the	DET
easat-11054	326	3	slight	slight	ADJ
easat-11054	326	4	difference	difference	NOUN
easat-11054	326	5	between	between	ADP
easat-11054	326	6	train	train	NOUN
easat-11054	326	7	and	and	CCONJ
easat-11054	326	8	validation	validation	NOUN
easat-11054	326	9	loss	loss	NOUN
easat-11054	326	10	,	,	PUNCT
easat-11054	326	11	it	it	PRON
easat-11054	326	12	appears	appear	VERB
easat-11054	326	13	that	that	SCONJ
easat-11054	326	14	the	the	DET
easat-11054	326	15	layer	layer	NOUN
easat-11054	326	16	is	be	AUX
easat-11054	326	17	well	well	ADV
easat-11054	326	18	-	-	PUNCT
easat-11054	326	19	trained	train	VERB
easat-11054	326	20	and	and	CCONJ
easat-11054	326	21	may	may	AUX
easat-11054	326	22	require	require	VERB
easat-11054	326	23	refinement	refinement	NOUN
easat-11054	326	24	to	to	PART
easat-11054	326	25	minimize	minimize	VERB
easat-11054	326	26	overfitting	overfitte	VERB
easat-11054	326	27	.	.	PUNCT
easat-11054	327	1	the	the	DET
easat-11054	327	2	stable	stable	ADJ
easat-11054	327	3	performance	performance	NOUN
easat-11054	327	4	of	of	ADP
easat-11054	327	5	this	this	DET
easat-11054	327	6	layer	layer	NOUN
easat-11054	327	7	indicates	indicate	VERB
easat-11054	327	8	good	good	ADJ
easat-11054	327	9	tuning	tuning	NOUN
easat-11054	327	10	.	.	PUNCT
easat-11054	328	1	these	these	DET
easat-11054	328	2	1066	1066	NUM
easat-11054	328	3	edelweiss	edelweiss	PROPN
easat-11054	328	4	applied	apply	VERB
easat-11054	328	5	science	science	NOUN
easat-11054	328	6	and	and	CCONJ
easat-11054	328	7	technology	technology	NOUN
easat-11054	328	8	issn	issn	PROPN
easat-11054	328	9	:	:	PUNCT
easat-11054	328	10	2576	2576	NUM
easat-11054	328	11	-	-	SYM
easat-11054	328	12	8484	8484	NUM
easat-11054	328	13	vol	vol	NOUN
easat-11054	328	14	.	.	PROPN
easat-11054	329	1	9	9	NUM
easat-11054	329	2	,	,	PUNCT
easat-11054	329	3	no	no	INTJ
easat-11054	329	4	.	.	NOUN
easat-11054	329	5	11	11	NUM
easat-11054	329	6	:	:	PUNCT
easat-11054	329	7	1053	1053	NUM
easat-11054	329	8	-	-	SYM
easat-11054	329	9	1075	1075	NUM
easat-11054	329	10	,	,	PUNCT
easat-11054	329	11	2025	2025	NUM
easat-11054	329	12	doi	doi	NOUN
easat-11054	329	13	:	:	PUNCT
easat-11054	329	14	10.55214/2576	10.55214/2576	NUM
easat-11054	329	15	-	-	SYM
easat-11054	329	16	8484.v9i11.11054	8484.v9i11.11054	NOUN
easat-11054	330	1	©	©	PROPN
easat-11054	330	2	2025	2025	NUM
easat-11054	330	3	by	by	ADP
easat-11054	330	4	the	the	DET
easat-11054	330	5	authors	author	NOUN
easat-11054	330	6	;	;	PUNCT
easat-11054	330	7	licensee	licensee	PROPN
easat-11054	330	8	learning	learning	NOUN
easat-11054	330	9	gate	gate	NOUN
easat-11054	330	10	gradually	gradually	ADV
easat-11054	330	11	increasing	increase	VERB
easat-11054	330	12	reward	reward	NOUN
easat-11054	330	13	values	value	NOUN
easat-11054	330	14	indicate	indicate	VERB
easat-11054	330	15	a	a	DET
easat-11054	330	16	slight	slight	ADJ
easat-11054	330	17	refinement	refinement	NOUN
easat-11054	330	18	to	to	ADP
easat-11054	330	19	the	the	DET
easat-11054	330	20	layer	layer	NOUN
easat-11054	330	21	in	in	ADP
easat-11054	330	22	making	make	VERB
easat-11054	330	23	decisions	decision	NOUN
easat-11054	330	24	,	,	PUNCT
easat-11054	330	25	although	although	SCONJ
easat-11054	330	26	further	further	ADJ
easat-11054	330	27	alterations	alteration	NOUN
easat-11054	330	28	to	to	ADP
easat-11054	330	29	this	this	DET
easat-11054	330	30	layer	layer	NOUN
easat-11054	330	31	may	may	AUX
easat-11054	330	32	improve	improve	VERB
easat-11054	330	33	efficiency	efficiency	NOUN
easat-11054	330	34	.	.	PUNCT
easat-11054	331	1	figure	figure	NOUN
easat-11054	331	2	5	5	NUM
easat-11054	331	3	.	.	PUNCT
easat-11054	331	4	training	training	NOUN
easat-11054	331	5	and	and	CCONJ
easat-11054	331	6	validation	validation	NOUN
easat-11054	331	7	loss	loss	NOUN
easat-11054	331	8	curves	curve	NOUN
easat-11054	331	9	for	for	ADP
easat-11054	331	10	the	the	DET
easat-11054	331	11	a3c	a3c	NUM
easat-11054	331	12	-	-	PUNCT
easat-11054	331	13	ffnn	ffnn	NOUN
easat-11054	331	14	.	.	PUNCT
easat-11054	332	1	the	the	DET
easat-11054	332	2	ffnn	ffnn	NOUN
easat-11054	332	3	layer	layer	NOUN
easat-11054	332	4	,	,	PUNCT
easat-11054	332	5	whose	whose	DET
easat-11054	332	6	input	input	NOUN
easat-11054	332	7	is	be	AUX
easat-11054	332	8	a	a	DET
easat-11054	332	9	combination	combination	NOUN
easat-11054	332	10	of	of	ADP
easat-11054	332	11	the	the	DET
easat-11054	332	12	dt	dt	PROPN
easat-11054	332	13	output	output	NOUN
easat-11054	332	14	and	and	CCONJ
easat-11054	332	15	the	the	DET
easat-11054	332	16	a3c	a3c	VERB
easat-11054	332	17	-	-	PUNCT
easat-11054	332	18	ffnn	ffnn	NOUN
easat-11054	332	19	output	output	NOUN
easat-11054	332	20	,	,	PUNCT
easat-11054	332	21	shows	show	VERB
easat-11054	332	22	significant	significant	ADJ
easat-11054	332	23	improvements	improvement	NOUN
easat-11054	332	24	in	in	ADP
easat-11054	332	25	both	both	DET
easat-11054	332	26	training	training	NOUN
easat-11054	332	27	and	and	CCONJ
easat-11054	332	28	validation	validation	NOUN
easat-11054	332	29	performance	performance	NOUN
easat-11054	332	30	compared	compare	VERB
easat-11054	332	31	to	to	ADP
easat-11054	332	32	the	the	DET
easat-11054	332	33	a3c	a3c	NUM
easat-11054	332	34	-	-	PUNCT
easat-11054	332	35	ffnn	ffnn	NOUN
easat-11054	332	36	.	.	PUNCT
easat-11054	333	1	across	across	ADP
easat-11054	333	2	multiple	multiple	ADJ
easat-11054	333	3	epochs	epoch	NOUN
easat-11054	333	4	,	,	PUNCT
easat-11054	333	5	as	as	SCONJ
easat-11054	333	6	shown	show	VERB
easat-11054	333	7	in	in	ADP
easat-11054	333	8	figure	figure	NOUN
easat-11054	333	9	6	6	NUM
easat-11054	333	10	,	,	PUNCT
easat-11054	333	11	the	the	DET
easat-11054	333	12	training	training	NOUN
easat-11054	333	13	accuracy	accuracy	NOUN
easat-11054	333	14	of	of	ADP
easat-11054	333	15	approximately	approximately	ADV
easat-11054	333	16	1.0	1.0	NUM
easat-11054	333	17	indicates	indicate	VERB
easat-11054	333	18	that	that	SCONJ
easat-11054	333	19	the	the	DET
easat-11054	333	20	layer	layer	NOUN
easat-11054	333	21	achieved	achieve	VERB
easat-11054	333	22	optimal	optimal	ADJ
easat-11054	333	23	performance	performance	NOUN
easat-11054	333	24	on	on	ADP
easat-11054	333	25	the	the	DET
easat-11054	333	26	training	training	NOUN
easat-11054	333	27	data	datum	NOUN
easat-11054	333	28	,	,	PUNCT
easat-11054	333	29	implying	imply	VERB
easat-11054	333	30	improved	improved	ADJ
easat-11054	333	31	overfitting	overfitting	NOUN
easat-11054	333	32	and	and	CCONJ
easat-11054	333	33	learning	learning	NOUN
easat-11054	333	34	.	.	PUNCT
easat-11054	334	1	the	the	DET
easat-11054	334	2	validation	validation	NOUN
easat-11054	334	3	accuracy	accuracy	NOUN
easat-11054	334	4	also	also	ADV
easat-11054	334	5	remains	remain	VERB
easat-11054	334	6	very	very	ADV
easat-11054	334	7	high	high	ADJ
easat-11054	334	8	,	,	PUNCT
easat-11054	334	9	at	at	ADP
easat-11054	334	10	0.9985	0.9985	NUM
easat-11054	334	11	to	to	ADP
easat-11054	334	12	0.9986	0.9986	NUM
easat-11054	334	13	,	,	PUNCT
easat-11054	334	14	demonstrating	demonstrate	VERB
easat-11054	334	15	excellent	excellent	ADJ
easat-11054	334	16	generalization	generalization	NOUN
easat-11054	334	17	to	to	ADP
easat-11054	334	18	unseen	unseen	ADJ
easat-11054	334	19	data	datum	NOUN
easat-11054	334	20	,	,	PUNCT
easat-11054	334	21	with	with	ADP
easat-11054	334	22	only	only	ADJ
easat-11054	334	23	minor	minor	ADJ
easat-11054	334	24	variations	variation	NOUN
easat-11054	334	25	across	across	ADP
easat-11054	334	26	epochs	epoch	NOUN
easat-11054	334	27	.	.	PUNCT
easat-11054	335	1	compared	compare	VERB
easat-11054	335	2	to	to	ADP
easat-11054	335	3	the	the	DET
easat-11054	335	4	a3c	a3c	NUM
easat-11054	335	5	-	-	PUNCT
easat-11054	335	6	ffnn	ffnn	NOUN
easat-11054	335	7	,	,	PUNCT
easat-11054	335	8	whose	whose	DET
easat-11054	335	9	accuracy	accuracy	NOUN
easat-11054	335	10	reached	reach	VERB
easat-11054	335	11	approximately	approximately	ADV
easat-11054	335	12	99.9	99.9	NUM
easat-11054	335	13	percent	percent	NOUN
easat-11054	335	14	,	,	PUNCT
easat-11054	335	15	this	this	PRON
easat-11054	335	16	confirms	confirm	VERB
easat-11054	335	17	the	the	DET
easat-11054	335	18	positive	positive	ADJ
easat-11054	335	19	contribution	contribution	NOUN
easat-11054	335	20	of	of	ADP
easat-11054	335	21	the	the	DET
easat-11054	335	22	hybrid	hybrid	NOUN
easat-11054	335	23	dt	dt	NOUN
easat-11054	335	24	to	to	ADP
easat-11054	335	25	improving	improve	VERB
easat-11054	335	26	the	the	DET
easat-11054	335	27	ffnn	ffnn	NOUN
easat-11054	335	28	layer	layer	NOUN
easat-11054	335	29	's	's	PART
easat-11054	335	30	ability	ability	NOUN
easat-11054	335	31	to	to	PART
easat-11054	335	32	perform	perform	VERB
easat-11054	335	33	both	both	DET
easat-11054	335	34	training	training	NOUN
easat-11054	335	35	and	and	CCONJ
easat-11054	335	36	validation	validation	NOUN
easat-11054	335	37	tasks	task	NOUN
easat-11054	335	38	with	with	ADP
easat-11054	335	39	high	high	ADJ
easat-11054	335	40	efficiency	efficiency	NOUN
easat-11054	335	41	.	.	PUNCT
easat-11054	336	1	figure	figure	NOUN
easat-11054	336	2	6	6	NUM
easat-11054	336	3	.	.	PUNCT
easat-11054	336	4	training	training	NOUN
easat-11054	336	5	and	and	CCONJ
easat-11054	336	6	validation	validation	NOUN
easat-11054	336	7	accuracy	accuracy	NOUN
easat-11054	336	8	curves	curve	NOUN
easat-11054	336	9	for	for	ADP
easat-11054	336	10	the	the	DET
easat-11054	336	11	ffnn	ffnn	NOUN
easat-11054	336	12	layer	layer	NOUN
easat-11054	336	13	.	.	PUNCT
easat-11054	337	1	1067	1067	NUM
easat-11054	337	2	edelweiss	edelweiss	PROPN
easat-11054	337	3	applied	apply	VERB
easat-11054	337	4	science	science	NOUN
easat-11054	337	5	and	and	CCONJ
easat-11054	337	6	technology	technology	NOUN
easat-11054	337	7	issn	issn	PROPN
easat-11054	337	8	:	:	PUNCT
easat-11054	337	9	2576	2576	NUM
easat-11054	337	10	-	-	SYM
easat-11054	337	11	8484	8484	NUM
easat-11054	337	12	vol	vol	NOUN
easat-11054	337	13	.	.	PROPN
easat-11054	338	1	9	9	NUM
easat-11054	338	2	,	,	PUNCT
easat-11054	338	3	no	no	INTJ
easat-11054	338	4	.	.	NOUN
easat-11054	338	5	11	11	NUM
easat-11054	338	6	:	:	PUNCT
easat-11054	338	7	1053	1053	NUM
easat-11054	338	8	-	-	SYM
easat-11054	338	9	1075	1075	NUM
easat-11054	338	10	,	,	PUNCT
easat-11054	338	11	2025	2025	NUM
easat-11054	338	12	doi	doi	NOUN
easat-11054	338	13	:	:	PUNCT
easat-11054	338	14	10.55214/2576	10.55214/2576	NUM
easat-11054	338	15	-	-	SYM
easat-11054	338	16	8484.v9i11.11054	8484.v9i11.11054	NOUN
easat-11054	339	1	©	©	PROPN
easat-11054	339	2	2025	2025	NUM
easat-11054	339	3	by	by	ADP
easat-11054	339	4	the	the	DET
easat-11054	339	5	authors	author	NOUN
easat-11054	339	6	;	;	PUNCT
easat-11054	339	7	licensee	licensee	PROPN
easat-11054	339	8	learning	learn	VERB
easat-11054	339	9	gate	gate	NOUN
easat-11054	339	10	on	on	ADP
easat-11054	339	11	the	the	DET
easat-11054	339	12	other	other	ADJ
easat-11054	339	13	hand	hand	NOUN
easat-11054	339	14	,	,	PUNCT
easat-11054	339	15	by	by	ADP
easat-11054	339	16	epoch	epoch	NOUN
easat-11054	339	17	25	25	NUM
easat-11054	339	18	,	,	PUNCT
easat-11054	339	19	as	as	SCONJ
easat-11054	339	20	shown	show	VERB
easat-11054	339	21	in	in	ADP
easat-11054	339	22	figure	figure	NOUN
easat-11054	339	23	7	7	NUM
easat-11054	339	24	,	,	PUNCT
easat-11054	339	25	the	the	DET
easat-11054	339	26	ffnn	ffnn	NOUN
easat-11054	339	27	layer	layer	NOUN
easat-11054	339	28	demonstrates	demonstrate	VERB
easat-11054	339	29	that	that	SCONJ
easat-11054	339	30	the	the	DET
easat-11054	339	31	training	training	NOUN
easat-11054	339	32	loss	loss	NOUN
easat-11054	339	33	steadily	steadily	ADV
easat-11054	339	34	decreases	decrease	VERB
easat-11054	339	35	from	from	ADP
easat-11054	339	36	approximately	approximately	ADV
easat-11054	339	37	0.0020	0.0020	NUM
easat-11054	339	38	to	to	ADP
easat-11054	339	39	0.0009	0.0009	NUM
easat-11054	339	40	on	on	ADP
easat-11054	339	41	the	the	DET
easat-11054	339	42	loss	loss	NOUN
easat-11054	339	43	curve	curve	NOUN
easat-11054	339	44	.	.	PUNCT
easat-11054	340	1	this	this	PRON
easat-11054	340	2	contrasts	contrast	VERB
easat-11054	340	3	sharply	sharply	ADV
easat-11054	340	4	with	with	ADP
easat-11054	340	5	the	the	DET
easat-11054	340	6	training	training	NOUN
easat-11054	340	7	loss	loss	NOUN
easat-11054	340	8	of	of	ADP
easat-11054	340	9	this	this	DET
easat-11054	340	10	layer	layer	NOUN
easat-11054	340	11	,	,	PUNCT
easat-11054	340	12	which	which	PRON
easat-11054	340	13	was	be	AUX
easat-11054	340	14	stable	stable	ADJ
easat-11054	340	15	at	at	ADP
easat-11054	340	16	0.8592	0.8592	NUM
easat-11054	340	17	.	.	PUNCT
easat-11054	341	1	meanwhile	meanwhile	ADV
easat-11054	341	2	,	,	PUNCT
easat-11054	341	3	the	the	DET
easat-11054	341	4	validation	validation	NOUN
easat-11054	341	5	loss	loss	NOUN
easat-11054	341	6	of	of	ADP
easat-11054	341	7	ffnn	ffnn	NOUN
easat-11054	341	8	is	be	AUX
easat-11054	341	9	slightly	slightly	ADV
easat-11054	341	10	variable	variable	ADJ
easat-11054	341	11	,	,	PUNCT
easat-11054	341	12	ranging	range	VERB
easat-11054	341	13	from	from	ADP
easat-11054	341	14	0.0075	0.0075	NUM
easat-11054	341	15	to	to	ADP
easat-11054	341	16	0.0008	0.0008	NUM
easat-11054	341	17	,	,	PUNCT
easat-11054	341	18	which	which	PRON
easat-11054	341	19	is	be	AUX
easat-11054	341	20	significantly	significantly	ADV
easat-11054	341	21	lower	low	ADJ
easat-11054	341	22	than	than	ADP
easat-11054	341	23	the	the	DET
easat-11054	341	24	0.9737	0.9737	NUM
easat-11054	341	25	loss	loss	NOUN
easat-11054	341	26	achieved	achieve	VERB
easat-11054	341	27	by	by	ADP
easat-11054	341	28	the	the	DET
easat-11054	341	29	validated	validate	VERB
easat-11054	341	30	a3c	a3c	VERB
easat-11054	341	31	-	-	PUNCT
easat-11054	341	32	ffnn	ffnn	NOUN
easat-11054	341	33	.	.	PUNCT
easat-11054	342	1	as	as	ADP
easat-11054	342	2	a	a	DET
easat-11054	342	3	result	result	NOUN
easat-11054	342	4	,	,	PUNCT
easat-11054	342	5	the	the	DET
easat-11054	342	6	ffnn	ffnn	NOUN
easat-11054	342	7	layer	layer	NOUN
easat-11054	342	8	improved	improve	VERB
easat-11054	342	9	not	not	PART
easat-11054	342	10	only	only	ADV
easat-11054	342	11	on	on	ADP
easat-11054	342	12	its	its	PRON
easat-11054	342	13	training	training	NOUN
easat-11054	342	14	data	datum	NOUN
easat-11054	342	15	but	but	CCONJ
easat-11054	342	16	also	also	ADV
easat-11054	342	17	in	in	ADP
easat-11054	342	18	generalization	generalization	NOUN
easat-11054	342	19	to	to	ADP
easat-11054	342	20	the	the	DET
easat-11054	342	21	validation	validation	NOUN
easat-11054	342	22	set	set	NOUN
easat-11054	342	23	.	.	PUNCT
easat-11054	343	1	overall	overall	ADJ
easat-11054	343	2	,	,	PUNCT
easat-11054	343	3	by	by	ADP
easat-11054	343	4	a	a	DET
easat-11054	343	5	reasonable	reasonable	ADJ
easat-11054	343	6	margin	margin	NOUN
easat-11054	343	7	,	,	PUNCT
easat-11054	343	8	the	the	DET
easat-11054	343	9	hybrid	hybrid	ADJ
easat-11054	343	10	approach	approach	NOUN
easat-11054	343	11	with	with	ADP
easat-11054	343	12	the	the	DET
easat-11054	343	13	dt	dt	PROPN
easat-11054	343	14	appears	appear	VERB
easat-11054	343	15	to	to	PART
easat-11054	343	16	enhance	enhance	VERB
easat-11054	343	17	the	the	DET
easat-11054	343	18	layer	layer	NOUN
easat-11054	343	19	's	's	PART
easat-11054	343	20	accuracy	accuracy	NOUN
easat-11054	343	21	and	and	CCONJ
easat-11054	343	22	generalization	generalization	NOUN
easat-11054	343	23	ability	ability	NOUN
easat-11054	343	24	,	,	PUNCT
easat-11054	343	25	significantly	significantly	ADV
easat-11054	343	26	outperforming	outperform	VERB
easat-11054	343	27	the	the	DET
easat-11054	343	28	ffnn	ffnn	NOUN
easat-11054	343	29	layer	layer	NOUN
easat-11054	343	30	in	in	ADP
easat-11054	343	31	a3c	a3c	VERB
easat-11054	343	32	-	-	PUNCT
easat-11054	343	33	ffnn	ffnn	NOUN
easat-11054	343	34	.	.	PUNCT
easat-11054	344	1	figure	figure	NOUN
easat-11054	344	2	7	7	NUM
easat-11054	344	3	.	.	PUNCT
easat-11054	344	4	training	training	NOUN
easat-11054	344	5	and	and	CCONJ
easat-11054	344	6	validation	validation	NOUN
easat-11054	344	7	loss	loss	NOUN
easat-11054	344	8	curves	curve	NOUN
easat-11054	344	9	for	for	ADP
easat-11054	344	10	the	the	DET
easat-11054	344	11	ffnn	ffnn	NOUN
easat-11054	344	12	layer	layer	NOUN
easat-11054	344	13	.	.	PUNCT
easat-11054	345	1	5.4	5.4	NUM
easat-11054	345	2	.	.	PUNCT
easat-11054	345	3	classification	classification	NOUN
easat-11054	345	4	testing	testing	NOUN
easat-11054	345	5	and	and	CCONJ
easat-11054	345	6	results	result	VERB
easat-11054	345	7	this	this	DET
easat-11054	345	8	section	section	NOUN
easat-11054	345	9	presents	present	VERB
easat-11054	345	10	the	the	DET
easat-11054	345	11	classification	classification	NOUN
easat-11054	345	12	performance	performance	NOUN
easat-11054	345	13	of	of	ADP
easat-11054	345	14	the	the	DET
easat-11054	345	15	proposed	propose	VERB
easat-11054	345	16	hadt	hadt	NOUN
easat-11054	345	17	algorithm	algorithm	NOUN
easat-11054	345	18	in	in	ADP
easat-11054	345	19	both	both	CCONJ
easat-11054	345	20	binary	binary	ADJ
easat-11054	345	21	and	and	CCONJ
easat-11054	345	22	multi	multi	ADJ
easat-11054	345	23	-	-	ADJ
easat-11054	345	24	class	class	ADJ
easat-11054	345	25	scenarios	scenario	NOUN
easat-11054	345	26	comprising	comprise	VERB
easat-11054	345	27	19	19	NUM
easat-11054	345	28	classes	class	NOUN
easat-11054	345	29	.	.	PUNCT
easat-11054	346	1	the	the	DET
easat-11054	346	2	results	result	NOUN
easat-11054	346	3	are	be	AUX
easat-11054	346	4	evaluated	evaluate	VERB
easat-11054	346	5	using	use	VERB
easat-11054	346	6	accuracy	accuracy	NOUN
easat-11054	346	7	(	(	PUNCT
easat-11054	346	8	a	a	NOUN
easat-11054	346	9	)	)	PUNCT
easat-11054	346	10	,	,	PUNCT
easat-11054	346	11	precision	precision	NOUN
easat-11054	346	12	(	(	PUNCT
easat-11054	346	13	p	p	NOUN
easat-11054	346	14	)	)	PUNCT
easat-11054	346	15	,	,	PUNCT
easat-11054	346	16	recall	recall	INTJ
easat-11054	346	17	(	(	PUNCT
easat-11054	346	18	r	r	NOUN
easat-11054	346	19	)	)	PUNCT
easat-11054	346	20	,	,	PUNCT
easat-11054	346	21	and	and	CCONJ
easat-11054	346	22	f1	f1	NOUN
easat-11054	346	23	-	-	PUNCT
easat-11054	346	24	score	score	NOUN
easat-11054	346	25	(	(	PUNCT
easat-11054	346	26	f	f	X
easat-11054	346	27	)	)	PUNCT
easat-11054	346	28	to	to	PART
easat-11054	346	29	assess	assess	VERB
easat-11054	346	30	the	the	DET
easat-11054	346	31	algorithm	algorithm	NOUN
easat-11054	346	32	's	's	PART
easat-11054	346	33	effectiveness	effectiveness	NOUN
easat-11054	346	34	in	in	ADP
easat-11054	346	35	detecting	detect	VERB
easat-11054	346	36	various	various	ADJ
easat-11054	346	37	classes	class	NOUN
easat-11054	346	38	of	of	ADP
easat-11054	346	39	iiot	iiot	NOUN
easat-11054	346	40	intrusions	intrusion	NOUN
easat-11054	346	41	.	.	PUNCT
easat-11054	347	1	5.4.1	5.4.1	X
easat-11054	347	2	.	.	PUNCT
easat-11054	347	3	binary	binary	ADJ
easat-11054	347	4	classification	classification	NOUN
easat-11054	347	5	results	result	VERB
easat-11054	347	6	analysis	analysis	NOUN
easat-11054	347	7	the	the	DET
easat-11054	347	8	results	result	NOUN
easat-11054	347	9	in	in	ADP
easat-11054	347	10	binary	binary	ADJ
easat-11054	347	11	classification	classification	NOUN
easat-11054	347	12	show	show	NOUN
easat-11054	347	13	that	that	SCONJ
easat-11054	347	14	the	the	DET
easat-11054	347	15	hadt	hadt	NOUN
easat-11054	347	16	algorithm	algorithm	NOUN
easat-11054	347	17	performs	perform	VERB
easat-11054	347	18	significantly	significantly	ADV
easat-11054	347	19	better	well	ADJ
easat-11054	347	20	than	than	ADP
easat-11054	347	21	both	both	DET
easat-11054	347	22	competing	compete	VERB
easat-11054	347	23	algorithms	algorithm	NOUN
easat-11054	347	24	,	,	PUNCT
easat-11054	347	25	a3c	a3c	PROPN
easat-11054	347	26	-	-	PUNCT
easat-11054	347	27	cnn	cnn	NOUN
easat-11054	347	28	and	and	CCONJ
easat-11054	347	29	a3c	a3c	NOUN
easat-11054	347	30	-	-	PUNCT
easat-11054	347	31	ffnn	ffnn	NOUN
easat-11054	347	32	,	,	PUNCT
easat-11054	347	33	with	with	ADP
easat-11054	347	34	an	an	DET
easat-11054	347	35	accuracy	accuracy	NOUN
easat-11054	347	36	(	(	PUNCT
easat-11054	347	37	a	a	NOUN
easat-11054	347	38	)	)	PUNCT
easat-11054	347	39	,	,	PUNCT
easat-11054	347	40	precision	precision	NOUN
easat-11054	347	41	(	(	PUNCT
easat-11054	347	42	p	p	NOUN
easat-11054	347	43	)	)	PUNCT
easat-11054	347	44	,	,	PUNCT
easat-11054	347	45	recall	recall	INTJ
easat-11054	347	46	(	(	PUNCT
easat-11054	347	47	r	r	NOUN
easat-11054	347	48	)	)	PUNCT
easat-11054	347	49	,	,	PUNCT
easat-11054	347	50	and	and	CCONJ
easat-11054	347	51	f1	f1	NOUN
easat-11054	347	52	score	score	NOUN
easat-11054	347	53	(	(	PUNCT
easat-11054	347	54	f	f	X
easat-11054	347	55	)	)	PUNCT
easat-11054	347	56	,	,	PUNCT
easat-11054	347	57	averaging	average	VERB
easat-11054	347	58	99.84	99.84	NUM
easat-11054	347	59	%	%	NOUN
easat-11054	347	60	,	,	PUNCT
easat-11054	347	61	as	as	SCONJ
easat-11054	347	62	shown	show	VERB
easat-11054	347	63	in	in	ADP
easat-11054	347	64	table	table	NOUN
easat-11054	347	65	4	4	NUM
easat-11054	347	66	.	.	PUNCT
easat-11054	347	67	a3c	a3c	VERB
easat-11054	347	68	-	-	PUNCT
easat-11054	347	69	cnn	cnn	PROPN
easat-11054	347	70	outperforms	outperform	VERB
easat-11054	347	71	a3c	a3c	VERB
easat-11054	347	72	-	-	PUNCT
easat-11054	347	73	ffnn	ffnn	NOUN
easat-11054	347	74	on	on	ADP
easat-11054	347	75	the	the	DET
easat-11054	347	76	same	same	ADJ
easat-11054	347	77	metrics	metric	NOUN
easat-11054	347	78	.	.	PUNCT
easat-11054	348	1	this	this	PRON
easat-11054	348	2	is	be	AUX
easat-11054	348	3	due	due	ADJ
easat-11054	348	4	to	to	ADP
easat-11054	348	5	cnn	cnn	PROPN
easat-11054	348	6	's	's	PART
easat-11054	348	7	superior	superior	ADJ
easat-11054	348	8	ability	ability	NOUN
easat-11054	348	9	to	to	PART
easat-11054	348	10	capture	capture	VERB
easat-11054	348	11	spatial	spatial	ADJ
easat-11054	348	12	features	feature	NOUN
easat-11054	348	13	in	in	ADP
easat-11054	348	14	iiot	iiot	NOUN
easat-11054	348	15	network	network	NOUN
easat-11054	348	16	traffic	traffic	NOUN
easat-11054	348	17	data	datum	NOUN
easat-11054	348	18	,	,	PUNCT
easat-11054	348	19	which	which	PRON
easat-11054	348	20	improves	improve	VERB
easat-11054	348	21	generalization	generalization	NOUN
easat-11054	348	22	and	and	CCONJ
easat-11054	348	23	detection	detection	NOUN
easat-11054	348	24	accuracy	accuracy	NOUN
easat-11054	348	25	.	.	PUNCT
easat-11054	349	1	although	although	SCONJ
easat-11054	349	2	a3c	a3c	ADJ
easat-11054	349	3	-	-	PUNCT
easat-11054	349	4	ffnn	ffnn	NOUN
easat-11054	349	5	has	have	VERB
easat-11054	349	6	high	high	ADJ
easat-11054	349	7	performance	performance	NOUN
easat-11054	349	8	,	,	PUNCT
easat-11054	349	9	it	it	PRON
easat-11054	349	10	is	be	AUX
easat-11054	349	11	not	not	PART
easat-11054	349	12	as	as	ADV
easat-11054	349	13	effective	effective	ADJ
easat-11054	349	14	as	as	ADP
easat-11054	349	15	a3c	a3c	PROPN
easat-11054	349	16	-	-	PUNCT
easat-11054	349	17	cnn	cnn	NOUN
easat-11054	349	18	or	or	CCONJ
easat-11054	349	19	the	the	DET
easat-11054	349	20	hybrid	hybrid	ADJ
easat-11054	349	21	approach	approach	NOUN
easat-11054	349	22	.	.	PUNCT
easat-11054	350	1	table	table	NOUN
easat-11054	350	2	4	4	NUM
easat-11054	350	3	.	.	PUNCT
easat-11054	351	1	the	the	DET
easat-11054	351	2	performance	performance	NOUN
easat-11054	351	3	comparison	comparison	NOUN
easat-11054	351	4	of	of	ADP
easat-11054	351	5	binary	binary	ADJ
easat-11054	351	6	classification	classification	NOUN
easat-11054	351	7	.	.	PUNCT
easat-11054	352	1	algorithm	algorithm	PROPN
easat-11054	352	2	binary	binary	ADJ
easat-11054	352	3	classification	classification	PROPN
easat-11054	352	4	a	a	DET
easat-11054	352	5	p	p	NOUN
easat-11054	352	6	r	r	NOUN
easat-11054	352	7	f1	f1	NOUN
easat-11054	352	8	a3c	a3c	PUNCT
easat-11054	352	9	-	-	PUNCT
easat-11054	352	10	ffnn	ffnn	NOUN
easat-11054	352	11	98.0123	98.0123	NUM
easat-11054	352	12	98.0654	98.0654	NUM
easat-11054	352	13	98.0136	98.0136	NUM
easat-11054	352	14	98.0353	98.0353	NUM
easat-11054	352	15	a3c	a3c	NUM
easat-11054	352	16	-	-	PUNCT
easat-11054	352	17	cnn	cnn	NOUN
easat-11054	352	18	98.5136	98.5136	NUM
easat-11054	352	19	98.5345	98.5345	NUM
easat-11054	352	20	98.5121	98.5121	NUM
easat-11054	352	21	98.5112	98.5112	NUM
easat-11054	352	22	hadt	hadt	NOUN
easat-11054	352	23	99.8401	99.8401	NUM
easat-11054	352	24	99.8408	99.8408	NUM
easat-11054	352	25	99.8421	99.8421	NUM
easat-11054	352	26	99.8411	99.8411	NUM
easat-11054	352	27	1068	1068	NUM
easat-11054	352	28	edelweiss	edelweiss	PROPN
easat-11054	352	29	applied	apply	VERB
easat-11054	352	30	science	science	NOUN
easat-11054	352	31	and	and	CCONJ
easat-11054	352	32	technology	technology	NOUN
easat-11054	352	33	issn	issn	PROPN
easat-11054	352	34	:	:	PUNCT
easat-11054	352	35	2576	2576	NUM
easat-11054	352	36	-	-	SYM
easat-11054	352	37	8484	8484	NUM
easat-11054	352	38	vol	vol	NOUN
easat-11054	352	39	.	.	PROPN
easat-11054	353	1	9	9	NUM
easat-11054	353	2	,	,	PUNCT
easat-11054	353	3	no	no	INTJ
easat-11054	353	4	.	.	NOUN
easat-11054	353	5	11	11	NUM
easat-11054	353	6	:	:	PUNCT
easat-11054	353	7	1053	1053	NUM
easat-11054	353	8	-	-	SYM
easat-11054	353	9	1075	1075	NUM
easat-11054	353	10	,	,	PUNCT
easat-11054	353	11	2025	2025	NUM
easat-11054	353	12	doi	doi	NOUN
easat-11054	353	13	:	:	PUNCT
easat-11054	353	14	10.55214/2576	10.55214/2576	NUM
easat-11054	353	15	-	-	SYM
easat-11054	353	16	8484.v9i11.11054	8484.v9i11.11054	NOUN
easat-11054	354	1	©	©	PROPN
easat-11054	354	2	2025	2025	NUM
easat-11054	354	3	by	by	ADP
easat-11054	354	4	the	the	DET
easat-11054	354	5	authors	author	NOUN
easat-11054	354	6	;	;	PUNCT
easat-11054	354	7	licensee	licensee	NOUN
easat-11054	354	8	learning	learning	NOUN
easat-11054	354	9	gate	gate	NOUN
easat-11054	354	10	for	for	ADP
easat-11054	354	11	applied	apply	VERB
easat-11054	354	12	intrusion	intrusion	NOUN
easat-11054	354	13	detection	detection	NOUN
easat-11054	354	14	,	,	PUNCT
easat-11054	354	15	the	the	DET
easat-11054	354	16	confusion	confusion	NOUN
easat-11054	354	17	matrix	matrix	NOUN
easat-11054	354	18	in	in	ADP
easat-11054	354	19	figure	figure	NOUN
easat-11054	354	20	8	8	NUM
easat-11054	354	21	is	be	AUX
easat-11054	354	22	crucial	crucial	ADJ
easat-11054	354	23	for	for	ADP
easat-11054	354	24	achieving	achieve	VERB
easat-11054	354	25	high	high	ADJ
easat-11054	354	26	reliability	reliability	NOUN
easat-11054	354	27	and	and	CCONJ
easat-11054	354	28	minimal	minimal	ADJ
easat-11054	354	29	misclassification	misclassification	NOUN
easat-11054	354	30	by	by	ADP
easat-11054	354	31	accurately	accurately	ADV
easat-11054	354	32	detecting	detect	VERB
easat-11054	354	33	these	these	DET
easat-11054	354	34	results	result	NOUN
easat-11054	354	35	,	,	PUNCT
easat-11054	354	36	which	which	PRON
easat-11054	354	37	are	be	AUX
easat-11054	354	38	supported	support	VERB
easat-11054	354	39	by	by	ADP
easat-11054	354	40	very	very	ADV
easat-11054	354	41	low	low	ADJ
easat-11054	354	42	false	false	ADJ
easat-11054	354	43	positives	positive	NOUN
easat-11054	354	44	(	(	PUNCT
easat-11054	354	45	97	97	NUM
easat-11054	354	46	)	)	PUNCT
easat-11054	354	47	and	and	CCONJ
easat-11054	354	48	false	false	ADJ
easat-11054	354	49	negatives	negative	NOUN
easat-11054	354	50	(	(	PUNCT
easat-11054	354	51	238	238	NUM
easat-11054	354	52	)	)	PUNCT
easat-11054	354	53	.	.	PUNCT
easat-11054	355	1	figure	figure	VERB
easat-11054	355	2	8	8	NUM
easat-11054	355	3	.	.	PUNCT
easat-11054	356	1	the	the	DET
easat-11054	356	2	confusion	confusion	NOUN
easat-11054	356	3	matrix	matrix	NOUN
easat-11054	356	4	of	of	ADP
easat-11054	356	5	hadt	hadt	NOUN
easat-11054	356	6	in	in	ADP
easat-11054	356	7	binary	binary	ADJ
easat-11054	356	8	classification	classification	NOUN
easat-11054	356	9	.	.	PUNCT
easat-11054	357	1	5.4.2	5.4.2	X
easat-11054	357	2	.	.	PUNCT
easat-11054	357	3	multi	multi	ADJ
easat-11054	357	4	-	-	ADJ
easat-11054	357	5	classification	classification	ADJ
easat-11054	357	6	results	result	VERB
easat-11054	357	7	analysis	analysis	NOUN
easat-11054	357	8	by	by	ADP
easat-11054	357	9	studying	study	VERB
easat-11054	357	10	the	the	DET
easat-11054	357	11	performance	performance	NOUN
easat-11054	357	12	of	of	ADP
easat-11054	357	13	hadt	hadt	NOUN
easat-11054	357	14	in	in	ADP
easat-11054	357	15	the	the	DET
easat-11054	357	16	multi	multi	ADJ
easat-11054	357	17	-	-	ADJ
easat-11054	357	18	classification	classification	ADJ
easat-11054	357	19	case	case	NOUN
easat-11054	357	20	,	,	PUNCT
easat-11054	357	21	as	as	SCONJ
easat-11054	357	22	shown	show	VERB
easat-11054	357	23	in	in	ADP
easat-11054	357	24	table	table	NOUN
easat-11054	357	25	5	5	NUM
easat-11054	357	26	,	,	PUNCT
easat-11054	357	27	this	this	DET
easat-11054	357	28	algorithm	algorithm	NOUN
easat-11054	357	29	has	have	AUX
easat-11054	357	30	demonstrated	demonstrate	VERB
easat-11054	357	31	high	high	ADJ
easat-11054	357	32	accuracy	accuracy	NOUN
easat-11054	357	33	across	across	ADP
easat-11054	357	34	most	most	ADJ
easat-11054	357	35	attack	attack	NOUN
easat-11054	357	36	classes	class	NOUN
easat-11054	357	37	,	,	PUNCT
easat-11054	357	38	based	base	VERB
easat-11054	357	39	on	on	ADP
easat-11054	357	40	the	the	DET
easat-11054	357	41	metrics	metric	NOUN
easat-11054	357	42	of	of	ADP
easat-11054	357	43	precision	precision	NOUN
easat-11054	357	44	,	,	PUNCT
easat-11054	357	45	recall	recall	NOUN
easat-11054	357	46	,	,	PUNCT
easat-11054	357	47	accuracy	accuracy	NOUN
easat-11054	357	48	,	,	PUNCT
easat-11054	357	49	and	and	CCONJ
easat-11054	357	50	f1	f1	NOUN
easat-11054	357	51	-	-	PUNCT
easat-11054	357	52	score	score	NOUN
easat-11054	357	53	,	,	PUNCT
easat-11054	357	54	confirming	confirm	VERB
easat-11054	357	55	its	its	PRON
easat-11054	357	56	ability	ability	NOUN
easat-11054	357	57	to	to	PART
easat-11054	357	58	detect	detect	VERB
easat-11054	357	59	various	various	ADJ
easat-11054	357	60	iiot	iiot	NOUN
easat-11054	357	61	attacks	attack	NOUN
easat-11054	357	62	.	.	PUNCT
easat-11054	358	1	for	for	ADP
easat-11054	358	2	common	common	ADJ
easat-11054	358	3	intrusion	intrusion	NOUN
easat-11054	358	4	attacks	attack	NOUN
easat-11054	358	5	,	,	PUNCT
easat-11054	358	6	all	all	DET
easat-11054	358	7	metrics	metric	NOUN
easat-11054	358	8	were	be	AUX
easat-11054	358	9	observed	observe	VERB
easat-11054	358	10	to	to	PART
easat-11054	358	11	be	be	AUX
easat-11054	358	12	above	above	ADP
easat-11054	358	13	99.8	99.8	NUM
easat-11054	358	14	%	%	NOUN
easat-11054	358	15	,	,	PUNCT
easat-11054	358	16	such	such	ADJ
easat-11054	358	17	as	as	ADP
easat-11054	358	18	normal	normal	ADJ
easat-11054	358	19	(	(	PUNCT
easat-11054	358	20	89,496	89,496	NUM
easat-11054	358	21	samples	sample	NOUN
easat-11054	358	22	)	)	PUNCT
easat-11054	358	23	,	,	PUNCT
easat-11054	358	24	rdos	rdo	NOUN
easat-11054	358	25	(	(	PUNCT
easat-11054	358	26	42,508	42,508	NUM
easat-11054	358	27	samples	sample	NOUN
easat-11054	358	28	)	)	PUNCT
easat-11054	358	29	,	,	PUNCT
easat-11054	358	30	and	and	CCONJ
easat-11054	358	31	scanning	scan	VERB
easat-11054	358	32	vulnerability	vulnerability	NOUN
easat-11054	358	33	(	(	PUNCT
easat-11054	358	34	15,929	15,929	NUM
easat-11054	358	35	samples	sample	NOUN
easat-11054	358	36	)	)	PUNCT
easat-11054	358	37	,	,	PUNCT
easat-11054	358	38	which	which	PRON
easat-11054	358	39	is	be	AUX
easat-11054	358	40	considered	consider	VERB
easat-11054	358	41	almost	almost	ADV
easat-11054	358	42	ideal	ideal	ADJ
easat-11054	358	43	detection	detection	NOUN
easat-11054	358	44	performance	performance	NOUN
easat-11054	358	45	.	.	PUNCT
easat-11054	359	1	for	for	ADP
easat-11054	359	2	intrusion	intrusion	NOUN
easat-11054	359	3	attack	attack	NOUN
easat-11054	359	4	classes	class	NOUN
easat-11054	359	5	such	such	ADJ
easat-11054	359	6	as	as	ADP
easat-11054	359	7	tcp	tcp	ADJ
easat-11054	359	8	relay	relay	NOUN
easat-11054	359	9	(	(	PUNCT
easat-11054	359	10	612	612	NUM
easat-11054	359	11	samples	sample	NOUN
easat-11054	359	12	)	)	PUNCT
easat-11054	359	13	,	,	PUNCT
easat-11054	359	14	fuzzing	fuzze	VERB
easat-11054	359	15	(	(	PUNCT
easat-11054	359	16	411	411	NUM
easat-11054	359	17	samples	sample	NOUN
easat-11054	359	18	)	)	PUNCT
easat-11054	359	19	,	,	PUNCT
easat-11054	359	20	reverse	reverse	ADJ
easat-11054	359	21	shell	shell	NOUN
easat-11054	359	22	(	(	PUNCT
easat-11054	359	23	294	294	NUM
easat-11054	359	24	samples	sample	NOUN
easat-11054	359	25	)	)	PUNCT
easat-11054	359	26	,	,	PUNCT
easat-11054	359	27	crypto	crypto	NOUN
easat-11054	359	28	-	-	NOUN
easat-11054	359	29	ransomware	ransomware	NOUN
easat-11054	359	30	(	(	PUNCT
easat-11054	359	31	136	136	NUM
easat-11054	359	32	samples	sample	NOUN
easat-11054	359	33	)	)	PUNCT
easat-11054	359	34	,	,	PUNCT
easat-11054	359	35	and	and	CCONJ
easat-11054	359	36	mitm	mitm	NOUN
easat-11054	359	37	(	(	PUNCT
easat-11054	359	38	34	34	NUM
easat-11054	359	39	samples	sample	NOUN
easat-11054	359	40	)	)	PUNCT
easat-11054	359	41	,	,	PUNCT
easat-11054	359	42	it	it	PRON
easat-11054	359	43	was	be	AUX
easat-11054	359	44	observed	observe	VERB
easat-11054	359	45	that	that	SCONJ
easat-11054	359	46	the	the	DET
easat-11054	359	47	precision	precision	NOUN
easat-11054	359	48	,	,	PUNCT
easat-11054	359	49	recall	recall	NOUN
easat-11054	359	50	,	,	PUNCT
easat-11054	359	51	and	and	CCONJ
easat-11054	359	52	f1	f1	NOUN
easat-11054	359	53	-	-	PUNCT
easat-11054	359	54	score	score	NOUN
easat-11054	359	55	were	be	AUX
easat-11054	359	56	approximately	approximately	ADV
easat-11054	359	57	equal	equal	ADJ
easat-11054	359	58	to	to	ADP
easat-11054	359	59	91.17	91.17	NUM
easat-11054	359	60	%	%	NOUN
easat-11054	359	61	.	.	PUNCT
easat-11054	360	1	precision	precision	NOUN
easat-11054	360	2	and	and	CCONJ
easat-11054	360	3	full	full	ADJ
easat-11054	360	4	recall	recall	NOUN
easat-11054	360	5	are	be	AUX
easat-11054	360	6	slightly	slightly	ADV
easat-11054	360	7	lower	low	ADJ
easat-11054	360	8	,	,	PUNCT
easat-11054	360	9	with	with	ADP
easat-11054	360	10	97.1	97.1	NUM
easat-11054	360	11	%	%	NOUN
easat-11054	360	12	precision	precision	NOUN
easat-11054	360	13	and	and	CCONJ
easat-11054	360	14	100	100	NUM
easat-11054	360	15	%	%	NOUN
easat-11054	360	16	recall	recall	NOUN
easat-11054	360	17	.	.	PUNCT
easat-11054	361	1	overall	overall	ADV
easat-11054	361	2	,	,	PUNCT
easat-11054	361	3	hadt	hadt	NOUN
easat-11054	361	4	is	be	AUX
easat-11054	361	5	a	a	DET
easat-11054	361	6	reliable	reliable	ADJ
easat-11054	361	7	solution	solution	NOUN
easat-11054	361	8	for	for	ADP
easat-11054	361	9	iiot	iiot	ADJ
easat-11054	361	10	security	security	NOUN
easat-11054	361	11	threats	threat	NOUN
easat-11054	361	12	,	,	PUNCT
easat-11054	361	13	with	with	ADP
easat-11054	361	14	high	high	ADJ
easat-11054	361	15	precision	precision	NOUN
easat-11054	361	16	and	and	CCONJ
easat-11054	361	17	generalizability	generalizability	NOUN
easat-11054	361	18	across	across	ADP
easat-11054	361	19	multiple	multiple	ADJ
easat-11054	361	20	classes	class	NOUN
easat-11054	361	21	.	.	PUNCT
easat-11054	362	1	1069	1069	NUM
easat-11054	362	2	edelweiss	edelweiss	PROPN
easat-11054	362	3	applied	apply	VERB
easat-11054	362	4	science	science	NOUN
easat-11054	362	5	and	and	CCONJ
easat-11054	362	6	technology	technology	NOUN
easat-11054	362	7	issn	issn	PROPN
easat-11054	362	8	:	:	PUNCT
easat-11054	362	9	2576	2576	NUM
easat-11054	362	10	-	-	SYM
easat-11054	362	11	8484	8484	NUM
easat-11054	362	12	vol	vol	NOUN
easat-11054	362	13	.	.	PROPN
easat-11054	363	1	9	9	NUM
easat-11054	363	2	,	,	PUNCT
easat-11054	363	3	no	no	INTJ
easat-11054	363	4	.	.	NOUN
easat-11054	363	5	11	11	NUM
easat-11054	363	6	:	:	PUNCT
easat-11054	363	7	1053	1053	NUM
easat-11054	363	8	-	-	SYM
easat-11054	363	9	1075	1075	NUM
easat-11054	363	10	,	,	PUNCT
easat-11054	363	11	2025	2025	NUM
easat-11054	363	12	doi	doi	NOUN
easat-11054	363	13	:	:	PUNCT
easat-11054	363	14	10.55214/2576	10.55214/2576	NUM
easat-11054	363	15	-	-	SYM
easat-11054	363	16	8484.v9i11.11054	8484.v9i11.11054	NOUN
easat-11054	364	1	©	©	PROPN
easat-11054	364	2	2025	2025	NUM
easat-11054	364	3	by	by	ADP
easat-11054	364	4	the	the	DET
easat-11054	364	5	authors	author	NOUN
easat-11054	364	6	;	;	PUNCT
easat-11054	364	7	licensee	licensee	PROPN
easat-11054	364	8	learning	learn	VERB
easat-11054	364	9	gate	gate	NOUN
easat-11054	364	10	table	table	NOUN
easat-11054	364	11	5	5	NUM
easat-11054	364	12	.	.	PUNCT
easat-11054	365	1	the	the	DET
easat-11054	365	2	performance	performance	NOUN
easat-11054	365	3	of	of	ADP
easat-11054	365	4	hadt	hadt	NOUN
easat-11054	365	5	per	per	ADP
easat-11054	365	6	class	class	NOUN
easat-11054	365	7	.	.	PUNCT
easat-11054	366	1	class	class	NOUN
easat-11054	366	2	a	a	DET
easat-11054	366	3	p	p	NOUN
easat-11054	366	4	r	r	NOUN
easat-11054	366	5	f	f	PROPN
easat-11054	366	6	tot	tot	NOUN
easat-11054	366	7	.	.	PUNCT
easat-11054	367	1	normal	normal	ADJ
easat-11054	367	2	0.998365	0.998365	NUM
easat-11054	367	3	0.997289	0.997289	NUM
easat-11054	367	4	0.998883	0.998883	NUM
easat-11054	367	5	0.998085	0.998085	NUM
easat-11054	367	6	89496	89496	NUM
easat-11054	367	7	rdos	rdo	VERB
easat-11054	367	8	0.999976	0.999976	NUM
easat-11054	367	9	1.000000	1.000000	NUM
easat-11054	367	10	0.999882	0.999882	NUM
easat-11054	367	11	0.999941	0.999941	NUM
easat-11054	367	12	42508	42508	NUM
easat-11054	367	13	scanning_vulnerability	scanning_vulnerability	NOUN
easat-11054	367	14	0.999938	0.999938	NUM
easat-11054	367	15	0.999874	0.999874	NUM
easat-11054	367	16	0.999309	0.999309	NUM
easat-11054	367	17	0.999592	0.999592	NUM
easat-11054	367	18	15929	15929	NUM
easat-11054	367	19	generic_scanning	generic_scanne	VERB
easat-11054	367	20	0.999957	0.999957	NUM
easat-11054	367	21	0.999735	0.999735	NUM
easat-11054	367	22	0.999669	0.999669	NUM
easat-11054	367	23	0.999702	0.999702	NUM
easat-11054	367	24	15112	15112	NUM
easat-11054	367	25	bruteforce	bruteforce	NOUN
easat-11054	367	26	1.000000	1.000000	NUM
easat-11054	367	27	1.000000	1.000000	NUM
easat-11054	367	28	1.000000	1.000000	NUM
easat-11054	367	29	1.000000	1.000000	NUM
easat-11054	367	30	14204	14204	NUM
easat-11054	367	31	mqtt_cloud_broker_subscription	mqtt_cloud_broker_subscription	NOUN
easat-11054	367	32	0.999971	0.999971	NUM
easat-11054	367	33	0.999437	0.999437	NUM
easat-11054	367	34	0.999719	0.999719	NUM
easat-11054	367	35	0.999578	0.999578	NUM
easat-11054	367	36	7109	7109	NUM
easat-11054	367	37	discovering_resources	discovering_resource	NOUN
easat-11054	367	38	0.999176	0.999176	NUM
easat-11054	368	1	0.992521	0.992521	NUM
easat-11054	368	2	0.982293	0.982293	NUM
easat-11054	368	3	0.987381	0.987381	NUM
easat-11054	368	4	6890	6890	NUM
easat-11054	368	5	exfiltration	exfiltration	NOUN
easat-11054	368	6	0.999962	0.999962	NUM
easat-11054	368	7	0.999851	0.999851	NUM
easat-11054	368	8	0.998956	0.998956	NUM
easat-11054	368	9	0.999403	0.999403	NUM
easat-11054	368	10	6705	6705	NUM
easat-11054	368	11	insider_malicious	insider_malicious	ADJ
easat-11054	368	12	1.000000	1.000000	NUM
easat-11054	368	13	1.000000	1.000000	NUM
easat-11054	368	14	1.000000	1.000000	NUM
easat-11054	368	15	1.000000	1.000000	NUM
easat-11054	368	16	5364	5364	NUM
easat-11054	368	17	modbus_register_reading	modbus_register_reade	VERB
easat-11054	368	18	0.999990	0.999990	NUM
easat-11054	368	19	1.000000	1.000000	NUM
easat-11054	368	20	0.998904	0.998904	NUM
easat-11054	368	21	0.999451	0.999451	NUM
easat-11054	368	22	1824	1824	NUM
easat-11054	368	23	false_data_injection	false_data_injection	NOUN
easat-11054	368	24	0.999905	0.999905	VERB
easat-11054	368	25	0.992172	0.992172	NUM
easat-11054	368	26	0.994768	0.994768	NUM
easat-11054	368	27	0.993468	0.993468	NUM
easat-11054	368	28	1529	1529	NUM
easat-11054	368	29	c&c	c&c	NOUN
easat-11054	368	30	0.999938	0.999938	NUM
easat-11054	368	31	0.990741	0.990741	NUM
easat-11054	368	32	0.994193	0.994193	NUM
easat-11054	368	33	0.992464	0.992464	NUM
easat-11054	368	34	861	861	NUM
easat-11054	368	35	dictionary	dictionary	ADJ
easat-11054	368	36	1.000000	1.000000	NUM
easat-11054	368	37	1.000000	1.000000	NUM
easat-11054	368	38	1.000000	1.000000	NUM
easat-11054	368	39	1.000000	1.000000	NUM
easat-11054	368	40	799	799	NUM
easat-11054	368	41	tcp	tcp	NOUN
easat-11054	368	42	relay	relay	NOUN
easat-11054	368	43	0.999890	0.999890	NUM
easat-11054	368	44	0.975767	0.975767	NUM
easat-11054	368	45	0.986928	0.986928	NUM
easat-11054	368	46	0.981316	0.981316	NUM
easat-11054	368	47	612	612	NUM
easat-11054	368	48	fuzzing	fuzze	VERB
easat-11054	368	49	0.999628	0.999628	NUM
easat-11054	368	50	0.982609	0.982609	NUM
easat-11054	368	51	0.824818	0.824818	NUM
easat-11054	368	52	0.896825	0.896825	NUM
easat-11054	369	1	411	411	NUM
easat-11054	370	1	reverse_shell	reverse_shell	NUM
easat-11054	370	2	0.999957	0.999957	NUM
easat-11054	370	3	0.983051	0.983051	NUM
easat-11054	370	4	0.986395	0.986395	NUM
easat-11054	370	5	0.984720	0.984720	NUM
easat-11054	370	6	294	294	NUM
easat-11054	370	7	crypto	crypto	NOUN
easat-11054	370	8	-	-	NOUN
easat-11054	370	9	ransomware	ransomware	NOUN
easat-11054	370	10	0.999995	0.999995	NUM
easat-11054	370	11	1.000000	1.000000	NUM
easat-11054	370	12	0.992647	0.992647	NUM
easat-11054	370	13	0.996310	0.996310	NUM
easat-11054	370	14	136	136	NUM
easat-11054	370	15	mitm	mitm	NOUN
easat-11054	370	16	0.999971	0.999971	NUM
easat-11054	370	17	0.911765	0.911765	NUM
easat-11054	370	18	0.911765	0.911765	NUM
easat-11054	370	19	0.911765	0.911765	NUM
easat-11054	370	20	34	34	NUM
easat-11054	370	21	fake_notification	fake_notification	NOUN
easat-11054	370	22	0.999995	0.999995	NUM
easat-11054	370	23	0.900000	0.900000	NUM
easat-11054	370	24	1.000000	1.000000	NUM
easat-11054	370	25	0.947368	0.947368	NUM
easat-11054	370	26	9	9	NUM
easat-11054	370	27	as	as	SCONJ
easat-11054	370	28	shown	show	VERB
easat-11054	370	29	in	in	ADP
easat-11054	370	30	figure	figure	NOUN
easat-11054	370	31	9	9	NUM
easat-11054	371	1	,	,	PUNCT
easat-11054	371	2	the	the	DET
easat-11054	371	3	confusion	confusion	NOUN
easat-11054	371	4	matrix	matrix	NOUN
easat-11054	371	5	provides	provide	VERB
easat-11054	371	6	a	a	DET
easat-11054	371	7	detailed	detailed	ADJ
easat-11054	371	8	evaluation	evaluation	NOUN
easat-11054	371	9	of	of	ADP
easat-11054	371	10	the	the	DET
easat-11054	371	11	proposed	propose	VERB
easat-11054	371	12	hadt	hadt	NOUN
easat-11054	371	13	algorithm	algorithm	NOUN
easat-11054	371	14	's	's	PART
easat-11054	371	15	classification	classification	NOUN
easat-11054	371	16	performance	performance	NOUN
easat-11054	371	17	for	for	ADP
easat-11054	371	18	19	19	NUM
easat-11054	371	19	attack	attack	NOUN
easat-11054	371	20	classes	class	NOUN
easat-11054	371	21	in	in	ADP
easat-11054	371	22	iiot	iiot	ADJ
easat-11054	371	23	environments	environment	NOUN
easat-11054	371	24	.	.	PUNCT
easat-11054	372	1	this	this	DET
easat-11054	372	2	algorithm	algorithm	NOUN
easat-11054	372	3	demonstrates	demonstrate	VERB
easat-11054	372	4	strong	strong	ADJ
easat-11054	372	5	detection	detection	NOUN
easat-11054	372	6	capabilities	capability	NOUN
easat-11054	372	7	,	,	PUNCT
easat-11054	372	8	with	with	ADP
easat-11054	372	9	high	high	ADJ
easat-11054	372	10	diagonal	diagonal	ADJ
easat-11054	372	11	values	value	NOUN
easat-11054	372	12	indicating	indicate	VERB
easat-11054	372	13	a	a	DET
easat-11054	372	14	high	high	ADJ
easat-11054	372	15	rate	rate	NOUN
easat-11054	372	16	of	of	ADP
easat-11054	372	17	correct	correct	ADJ
easat-11054	372	18	classifications	classification	NOUN
easat-11054	372	19	.	.	PUNCT
easat-11054	373	1	major	major	ADJ
easat-11054	373	2	attack	attack	NOUN
easat-11054	373	3	classes	class	NOUN
easat-11054	373	4	such	such	ADJ
easat-11054	373	5	as	as	ADP
easat-11054	373	6	bruteforce	bruteforce	NOUN
easat-11054	373	7	(	(	PUNCT
easat-11054	373	8	42,503	42,503	NUM
easat-11054	373	9	)	)	PUNCT
easat-11054	373	10	,	,	PUNCT
easat-11054	373	11	mqtt_cloud_broker_subscription	mqtt_cloud_broker_subscription	NOUN
easat-11054	373	12	(	(	PUNCT
easat-11054	373	13	15,918	15,918	NUM
easat-11054	373	14	)	)	PUNCT
easat-11054	373	15	,	,	PUNCT
easat-11054	373	16	and	and	CCONJ
easat-11054	373	17	modbus_register_reading	modbus_register_reade	VERB
easat-11054	373	18	(	(	PUNCT
easat-11054	373	19	15,107	15,107	NUM
easat-11054	373	20	)	)	PUNCT
easat-11054	373	21	are	be	AUX
easat-11054	373	22	classified	classify	VERB
easat-11054	373	23	with	with	ADP
easat-11054	373	24	near	near	ADV
easat-11054	373	25	-	-	PUNCT
easat-11054	373	26	perfect	perfect	ADJ
easat-11054	373	27	accuracy	accuracy	NOUN
easat-11054	373	28	.	.	PUNCT
easat-11054	374	1	the	the	DET
easat-11054	374	2	classification	classification	NOUN
easat-11054	374	3	of	of	ADP
easat-11054	374	4	normal	normal	ADJ
easat-11054	374	5	traffic	traffic	NOUN
easat-11054	374	6	(	(	PUNCT
easat-11054	374	7	89,396	89,396	NUM
easat-11054	374	8	instances	instance	NOUN
easat-11054	374	9	)	)	PUNCT
easat-11054	374	10	also	also	ADV
easat-11054	374	11	demonstrates	demonstrate	VERB
easat-11054	374	12	the	the	DET
easat-11054	374	13	hadt	hadt	NOUN
easat-11054	374	14	algorithm	algorithm	NOUN
easat-11054	374	15	's	's	PART
easat-11054	374	16	effectiveness	effectiveness	NOUN
easat-11054	374	17	in	in	ADP
easat-11054	374	18	distinguishing	distinguish	VERB
easat-11054	374	19	between	between	ADP
easat-11054	374	20	normal	normal	ADJ
easat-11054	374	21	activity	activity	NOUN
easat-11054	374	22	and	and	CCONJ
easat-11054	374	23	intrusion	intrusion	NOUN
easat-11054	374	24	attacks	attack	NOUN
easat-11054	374	25	.	.	PUNCT
easat-11054	375	1	however	however	ADV
easat-11054	375	2	,	,	PUNCT
easat-11054	375	3	minor	minor	ADJ
easat-11054	375	4	shortcomings	shortcoming	NOUN
easat-11054	375	5	were	be	AUX
easat-11054	375	6	found	find	VERB
easat-11054	375	7	in	in	ADP
easat-11054	375	8	minor	minor	ADJ
easat-11054	375	9	attack	attack	NOUN
easat-11054	375	10	classes	class	NOUN
easat-11054	375	11	,	,	PUNCT
easat-11054	375	12	with	with	ADP
easat-11054	375	13	some	some	DET
easat-11054	375	14	misclassifications	misclassification	NOUN
easat-11054	375	15	.	.	PUNCT
easat-11054	376	1	for	for	ADP
easat-11054	376	2	example	example	NOUN
easat-11054	376	3	,	,	PUNCT
easat-11054	376	4	fake	fake	ADJ
easat-11054	376	5	notification	notification	NOUN
easat-11054	376	6	(	(	PUNCT
easat-11054	376	7	9	9	NUM
easat-11054	376	8	cases	case	NOUN
easat-11054	376	9	)	)	PUNCT
easat-11054	376	10	and	and	CCONJ
easat-11054	376	11	mitm	mitm	NOUN
easat-11054	376	12	(	(	PUNCT
easat-11054	376	13	31	31	NUM
easat-11054	376	14	cases	case	NOUN
easat-11054	376	15	)	)	PUNCT
easat-11054	376	16	show	show	VERB
easat-11054	376	17	minor	minor	ADJ
easat-11054	376	18	errors	error	NOUN
easat-11054	376	19	,	,	PUNCT
easat-11054	376	20	as	as	ADV
easat-11054	376	21	well	well	ADV
easat-11054	376	22	as	as	ADP
easat-11054	376	23	tcp	tcp	ADP
easat-11054	376	24	relay	relay	NOUN
easat-11054	376	25	(	(	PUNCT
easat-11054	376	26	612	612	NUM
easat-11054	376	27	cases	case	NOUN
easat-11054	376	28	)	)	PUNCT
easat-11054	376	29	.	.	PUNCT
easat-11054	377	1	17	17	NUM
easat-11054	377	2	cases	case	NOUN
easat-11054	377	3	are	be	AUX
easat-11054	377	4	misclassified	misclassifie	VERB
easat-11054	377	5	,	,	PUNCT
easat-11054	377	6	so	so	ADV
easat-11054	377	7	most	most	ADV
easat-11054	377	8	likely	likely	ADJ
easat-11054	377	9	these	these	DET
easat-11054	377	10	errors	error	NOUN
easat-11054	377	11	could	could	AUX
easat-11054	377	12	be	be	AUX
easat-11054	377	13	caused	cause	VERB
easat-11054	377	14	by	by	ADP
easat-11054	377	15	interference	interference	NOUN
easat-11054	377	16	with	with	ADP
easat-11054	377	17	the	the	DET
easat-11054	377	18	distribution	distribution	NOUN
easat-11054	377	19	of	of	ADP
easat-11054	377	20	features	feature	NOUN
easat-11054	377	21	in	in	ADP
easat-11054	377	22	this	this	DET
easat-11054	377	23	dataset	dataset	NOUN
easat-11054	377	24	due	due	ADP
easat-11054	377	25	to	to	ADP
easat-11054	377	26	the	the	DET
easat-11054	377	27	possible	possible	ADJ
easat-11054	377	28	similarity	similarity	NOUN
easat-11054	377	29	of	of	ADP
easat-11054	377	30	some	some	PRON
easat-11054	377	31	of	of	ADP
easat-11054	377	32	these	these	DET
easat-11054	377	33	features	feature	NOUN
easat-11054	377	34	with	with	ADP
easat-11054	377	35	other	other	ADJ
easat-11054	377	36	types	type	NOUN
easat-11054	377	37	of	of	ADP
easat-11054	377	38	attacks	attack	NOUN
easat-11054	377	39	.	.	PUNCT
easat-11054	378	1	these	these	DET
easat-11054	378	2	minor	minor	ADJ
easat-11054	378	3	errors	error	NOUN
easat-11054	378	4	can	can	AUX
easat-11054	378	5	be	be	AUX
easat-11054	378	6	reduced	reduce	VERB
easat-11054	378	7	by	by	ADP
easat-11054	378	8	adding	add	VERB
easat-11054	378	9	improvements	improvement	NOUN
easat-11054	378	10	to	to	ADP
easat-11054	378	11	the	the	DET
easat-11054	378	12	datasets	dataset	NOUN
easat-11054	378	13	,	,	PUNCT
easat-11054	378	14	such	such	ADJ
easat-11054	378	15	as	as	ADP
easat-11054	378	16	data	datum	NOUN
easat-11054	378	17	augmentation	augmentation	NOUN
easat-11054	378	18	.	.	PUNCT
easat-11054	379	1	1070	1070	NUM
easat-11054	379	2	edelweiss	edelweiss	PROPN
easat-11054	379	3	applied	apply	VERB
easat-11054	379	4	science	science	NOUN
easat-11054	379	5	and	and	CCONJ
easat-11054	379	6	technology	technology	NOUN
easat-11054	379	7	issn	issn	PROPN
easat-11054	379	8	:	:	PUNCT
easat-11054	379	9	2576	2576	NUM
easat-11054	379	10	-	-	SYM
easat-11054	379	11	8484	8484	NUM
easat-11054	379	12	vol	vol	NOUN
easat-11054	379	13	.	.	PROPN
easat-11054	380	1	9	9	NUM
easat-11054	380	2	,	,	PUNCT
easat-11054	380	3	no	no	INTJ
easat-11054	380	4	.	.	NOUN
easat-11054	380	5	11	11	NUM
easat-11054	380	6	:	:	PUNCT
easat-11054	380	7	1053	1053	NUM
easat-11054	380	8	-	-	SYM
easat-11054	380	9	1075	1075	NUM
easat-11054	380	10	,	,	PUNCT
easat-11054	380	11	2025	2025	NUM
easat-11054	380	12	doi	doi	NOUN
easat-11054	380	13	:	:	PUNCT
easat-11054	380	14	10.55214/2576	10.55214/2576	NUM
easat-11054	380	15	-	-	SYM
easat-11054	380	16	8484.v9i11.11054	8484.v9i11.11054	NOUN
easat-11054	381	1	©	©	PROPN
easat-11054	381	2	2025	2025	NUM
easat-11054	381	3	by	by	ADP
easat-11054	381	4	the	the	DET
easat-11054	381	5	authors	author	NOUN
easat-11054	381	6	;	;	PUNCT
easat-11054	381	7	licensee	licensee	PROPN
easat-11054	381	8	learning	learn	VERB
easat-11054	381	9	gate	gate	NOUN
easat-11054	381	10	figure	figure	NOUN
easat-11054	381	11	9	9	NUM
easat-11054	381	12	.	.	PUNCT
easat-11054	382	1	the	the	DET
easat-11054	382	2	confusion	confusion	NOUN
easat-11054	382	3	matrix	matrix	NOUN
easat-11054	382	4	of	of	ADP
easat-11054	382	5	hadt	hadt	NOUN
easat-11054	382	6	in	in	ADP
easat-11054	382	7	multi	multi	ADJ
easat-11054	382	8	-	-	NOUN
easat-11054	382	9	classification	classification	NOUN
easat-11054	382	10	.	.	PUNCT
easat-11054	383	1	the	the	DET
easat-11054	383	2	receiver	receiver	NOUN
easat-11054	383	3	operating	operate	VERB
easat-11054	383	4	characteristic	characteristic	NOUN
easat-11054	383	5	(	(	PUNCT
easat-11054	383	6	roc	roc	PROPN
easat-11054	383	7	)	)	PUNCT
easat-11054	383	8	curve	curve	NOUN
easat-11054	383	9	and	and	CCONJ
easat-11054	383	10	area	area	NOUN
easat-11054	383	11	under	under	ADP
easat-11054	383	12	the	the	DET
easat-11054	383	13	curve	curve	NOUN
easat-11054	383	14	(	(	PUNCT
easat-11054	383	15	auc	auc	NOUN
easat-11054	383	16	)	)	PUNCT
easat-11054	383	17	metrics	metric	NOUN
easat-11054	383	18	are	be	AUX
easat-11054	383	19	examined	examine	VERB
easat-11054	383	20	to	to	PART
easat-11054	383	21	evaluate	evaluate	VERB
easat-11054	383	22	classification	classification	NOUN
easat-11054	383	23	performance	performance	NOUN
easat-11054	383	24	.	.	PUNCT
easat-11054	384	1	by	by	ADP
easat-11054	384	2	plotting	plot	VERB
easat-11054	384	3	the	the	DET
easat-11054	384	4	true	true	ADJ
easat-11054	384	5	positive	positive	ADJ
easat-11054	384	6	rate	rate	NOUN
easat-11054	384	7	(	(	PUNCT
easat-11054	384	8	tpr	tpr	NOUN
easat-11054	384	9	)	)	PUNCT
easat-11054	384	10	and	and	CCONJ
easat-11054	384	11	false	false	ADJ
easat-11054	384	12	positive	positive	ADJ
easat-11054	384	13	rate	rate	NOUN
easat-11054	384	14	(	(	PUNCT
easat-11054	384	15	fpr	fpr	NOUN
easat-11054	384	16	)	)	PUNCT
easat-11054	384	17	,	,	PUNCT
easat-11054	384	18	the	the	DET
easat-11054	384	19	roc	roc	PROPN
easat-11054	384	20	curve	curve	NOUN
easat-11054	384	21	provides	provide	VERB
easat-11054	384	22	a	a	DET
easat-11054	384	23	visual	visual	ADJ
easat-11054	384	24	representation	representation	NOUN
easat-11054	384	25	of	of	ADP
easat-11054	384	26	the	the	DET
easat-11054	384	27	hadt	hadt	NOUN
easat-11054	384	28	algorithm	algorithm	NOUN
easat-11054	384	29	's	's	PART
easat-11054	384	30	performance	performance	NOUN
easat-11054	384	31	across	across	ADP
easat-11054	384	32	decision	decision	NOUN
easat-11054	384	33	thresholds	threshold	NOUN
easat-11054	384	34	.	.	PUNCT
easat-11054	385	1	auc	auc	NOUN
easat-11054	385	2	values	value	NOUN
easat-11054	385	3	as	as	ADV
easat-11054	385	4	low	low	ADJ
easat-11054	385	5	as	as	ADP
easat-11054	385	6	1.0	1.0	NUM
easat-11054	385	7	are	be	AUX
easat-11054	385	8	indicative	indicative	ADJ
easat-11054	385	9	of	of	ADP
easat-11054	385	10	strong	strong	ADJ
easat-11054	385	11	classification	classification	NOUN
easat-11054	385	12	ability	ability	NOUN
easat-11054	385	13	,	,	PUNCT
easat-11054	385	14	while	while	SCONJ
easat-11054	385	15	those	those	PRON
easat-11054	385	16	nearing	near	VERB
easat-11054	385	17	it	it	PRON
easat-11054	385	18	have	have	VERB
easat-11054	385	19	higher	high	ADJ
easat-11054	385	20	auc	auc	NOUN
easat-11054	385	21	.	.	PUNCT
easat-11054	386	1	the	the	DET
easat-11054	386	2	roc	roc	PROPN
easat-11054	386	3	curves	curve	NOUN
easat-11054	386	4	for	for	ADP
easat-11054	386	5	fpr	fpr	PRON
easat-11054	386	6	and	and	CCONJ
easat-11054	386	7	tpr	tpr	NOUN
easat-11054	386	8	were	be	AUX
easat-11054	386	9	generated	generate	VERB
easat-11054	386	10	using	use	VERB
easat-11054	386	11	sample	sample	NOUN
easat-11054	386	12	data	datum	NOUN
easat-11054	386	13	from	from	ADP
easat-11054	386	14	this	this	DET
easat-11054	386	15	earlier	early	ADV
easat-11054	386	16	categorized	categorize	VERB
easat-11054	386	17	test	test	NOUN
easat-11054	386	18	from	from	ADP
easat-11054	386	19	the	the	DET
easat-11054	386	20	x	x	ADJ
easat-11054	386	21	-	-	ADJ
easat-11054	386	22	iiotid	iiotid	ADJ
easat-11054	386	23	dataset	dataset	NOUN
easat-11054	386	24	.	.	PUNCT
easat-11054	387	1	roc	roc	PROPN
easat-11054	387	2	curves	curve	NOUN
easat-11054	387	3	are	be	AUX
easat-11054	387	4	shown	show	VERB
easat-11054	387	5	as	as	SCONJ
easat-11054	387	6	illustrated	illustrate	VERB
easat-11054	387	7	in	in	ADP
easat-11054	387	8	figure	figure	NOUN
easat-11054	387	9	10	10	NUM
easat-11054	387	10	;	;	PUNCT
easat-11054	387	11	the	the	DET
easat-11054	387	12	roc	roc	PROPN
easat-11054	387	13	curve	curve	NOUN
easat-11054	387	14	fitting	fitting	ADJ
easat-11054	387	15	degree	degree	NOUN
easat-11054	387	16	of	of	ADP
easat-11054	387	17	the	the	DET
easat-11054	387	18	hadt	hadt	NOUN
easat-11054	387	19	algorithm	algorithm	NOUN
easat-11054	387	20	for	for	ADP
easat-11054	387	21	the	the	DET
easat-11054	387	22	nineteenclassification	nineteenclassification	NOUN
easat-11054	387	23	test	test	NOUN
easat-11054	387	24	is	be	AUX
easat-11054	387	25	faultless	faultless	ADJ
easat-11054	387	26	when	when	SCONJ
easat-11054	387	27	compared	compare	VERB
easat-11054	387	28	to	to	ADP
easat-11054	387	29	prior	prior	ADJ
easat-11054	387	30	studies	study	NOUN
easat-11054	387	31	'	'	PART
easat-11054	387	32	results	result	NOUN
easat-11054	387	33	.	.	PUNCT
easat-11054	388	1	auc	auc	NOUN
easat-11054	388	2	values	value	NOUN
easat-11054	388	3	of	of	ADP
easat-11054	388	4	x	x	NOUN
easat-11054	388	5	-	-	NOUN
easat-11054	388	6	iiotid	iiotid	NOUN
easat-11054	388	7	,	,	PUNCT
easat-11054	388	8	which	which	PRON
easat-11054	388	9	are	be	AUX
easat-11054	388	10	between	between	ADP
easat-11054	388	11	0.9982	0.9982	NUM
easat-11054	388	12	and	and	CCONJ
easat-11054	388	13	1.0000	1.0000	NUM
easat-11054	388	14	,	,	PUNCT
easat-11054	388	15	have	have	AUX
easat-11054	388	16	been	be	AUX
easat-11054	388	17	obtained	obtain	VERB
easat-11054	388	18	through	through	ADP
easat-11054	388	19	computation	computation	NOUN
easat-11054	388	20	,	,	PUNCT
easat-11054	388	21	indicating	indicate	VERB
easat-11054	388	22	that	that	SCONJ
easat-11054	388	23	the	the	DET
easat-11054	388	24	classification	classification	NOUN
easat-11054	388	25	algorithm	algorithm	NOUN
easat-11054	388	26	trained	train	VERB
easat-11054	388	27	on	on	ADP
easat-11054	388	28	hadt	hadt	NOUN
easat-11054	388	29	is	be	AUX
easat-11054	388	30	highly	highly	ADV
easat-11054	388	31	effective	effective	ADJ
easat-11054	388	32	in	in	ADP
easat-11054	388	33	detecting	detect	VERB
easat-11054	388	34	many	many	ADJ
easat-11054	388	35	intrusion	intrusion	NOUN
easat-11054	388	36	types	type	NOUN
easat-11054	388	37	.	.	PUNCT
easat-11054	389	1	1071	1071	NUM
easat-11054	389	2	edelweiss	edelweiss	PROPN
easat-11054	389	3	applied	apply	VERB
easat-11054	389	4	science	science	NOUN
easat-11054	389	5	and	and	CCONJ
easat-11054	389	6	technology	technology	NOUN
easat-11054	389	7	issn	issn	PROPN
easat-11054	389	8	:	:	PUNCT
easat-11054	389	9	2576	2576	NUM
easat-11054	389	10	-	-	SYM
easat-11054	389	11	8484	8484	NUM
easat-11054	389	12	vol	vol	NOUN
easat-11054	389	13	.	.	PROPN
easat-11054	390	1	9	9	NUM
easat-11054	390	2	,	,	PUNCT
easat-11054	390	3	no	no	INTJ
easat-11054	390	4	.	.	NOUN
easat-11054	390	5	11	11	NUM
easat-11054	390	6	:	:	PUNCT
easat-11054	390	7	1053	1053	NUM
easat-11054	390	8	-	-	SYM
easat-11054	390	9	1075	1075	NUM
easat-11054	390	10	,	,	PUNCT
easat-11054	390	11	2025	2025	NUM
easat-11054	390	12	doi	doi	NOUN
easat-11054	390	13	:	:	PUNCT
easat-11054	390	14	10.55214/2576	10.55214/2576	NUM
easat-11054	390	15	-	-	SYM
easat-11054	390	16	8484.v9i11.11054	8484.v9i11.11054	NOUN
easat-11054	391	1	©	©	PROPN
easat-11054	391	2	2025	2025	NUM
easat-11054	391	3	by	by	ADP
easat-11054	391	4	the	the	DET
easat-11054	391	5	authors	author	NOUN
easat-11054	391	6	;	;	PUNCT
easat-11054	391	7	licensee	licensee	PROPN
easat-11054	391	8	learning	learn	VERB
easat-11054	391	9	gate	gate	PROPN
easat-11054	391	10	figure	figure	NOUN
easat-11054	391	11	10	10	NUM
easat-11054	391	12	.	.	PUNCT
easat-11054	392	1	the	the	DET
easat-11054	392	2	hadt	hadt	ADJ
easat-11054	392	3	nineteen	nineteen	NUM
easat-11054	392	4	-	-	PUNCT
easat-11054	392	5	classification	classification	NOUN
easat-11054	392	6	test	test	NOUN
easat-11054	392	7	's	's	PART
easat-11054	392	8	roc	roc	PROPN
easat-11054	392	9	curve	curve	NOUN
easat-11054	392	10	.	.	PUNCT
easat-11054	393	1	for	for	ADP
easat-11054	393	2	binary	binary	ADJ
easat-11054	393	3	classification	classification	NOUN
easat-11054	393	4	,	,	PUNCT
easat-11054	393	5	the	the	DET
easat-11054	393	6	hadt	hadt	NOUN
easat-11054	393	7	and	and	CCONJ
easat-11054	393	8	current	current	ADJ
easat-11054	393	9	a3c	a3c	VERB
easat-11054	393	10	algorithms	algorithm	NOUN
easat-11054	393	11	were	be	AUX
easat-11054	393	12	used	use	VERB
easat-11054	393	13	for	for	ADP
easat-11054	393	14	nineteen	nineteen	NUM
easat-11054	393	15	classification	classification	NOUN
easat-11054	393	16	test	test	NOUN
easat-11054	393	17	training	training	NOUN
easat-11054	393	18	and	and	CCONJ
easat-11054	393	19	detection	detection	NOUN
easat-11054	393	20	tests	test	NOUN
easat-11054	393	21	.	.	PUNCT
easat-11054	394	1	when	when	SCONJ
easat-11054	394	2	comparing	compare	VERB
easat-11054	394	3	the	the	DET
easat-11054	394	4	test	test	NOUN
easat-11054	394	5	results	result	NOUN
easat-11054	394	6	of	of	ADP
easat-11054	394	7	table	table	NOUN
easat-11054	394	8	6	6	NUM
easat-11054	394	9	to	to	ADP
easat-11054	394	10	the	the	DET
easat-11054	394	11	other	other	ADJ
easat-11054	394	12	two	two	NUM
easat-11054	394	13	algorithms	algorithm	NOUN
easat-11054	394	14	,	,	PUNCT
easat-11054	394	15	it	it	PRON
easat-11054	394	16	is	be	AUX
easat-11054	394	17	evident	evident	ADJ
easat-11054	394	18	that	that	SCONJ
easat-11054	394	19	hadt	hadt	NOUN
easat-11054	394	20	detection	detection	NOUN
easat-11054	394	21	has	have	VERB
easat-11054	394	22	superior	superior	ADJ
easat-11054	394	23	accuracy	accuracy	NOUN
easat-11054	394	24	,	,	PUNCT
easat-11054	394	25	precision	precision	NOUN
easat-11054	394	26	,	,	PUNCT
easat-11054	394	27	recall	recall	NOUN
easat-11054	394	28	,	,	PUNCT
easat-11054	394	29	and	and	CCONJ
easat-11054	394	30	f1	f1	NOUN
easat-11054	394	31	-	-	PUNCT
easat-11054	394	32	score	score	NOUN
easat-11054	394	33	compared	compare	VERB
easat-11054	394	34	to	to	ADP
easat-11054	394	35	a3c	a3c	VERB
easat-11054	394	36	-	-	PUNCT
easat-11054	394	37	ffnn	ffnn	NOUN
easat-11054	394	38	and	and	CCONJ
easat-11054	394	39	a3c	a3c	PROPN
easat-11054	394	40	-	-	PUNCT
easat-11054	394	41	cnn	cnn	NOUN
easat-11054	394	42	.	.	PUNCT
easat-11054	395	1	these	these	DET
easat-11054	395	2	enhanced	enhance	VERB
easat-11054	395	3	results	result	NOUN
easat-11054	395	4	are	be	AUX
easat-11054	395	5	due	due	ADJ
easat-11054	395	6	to	to	ADP
easat-11054	395	7	the	the	DET
easat-11054	395	8	integration	integration	NOUN
easat-11054	395	9	of	of	ADP
easat-11054	395	10	a3c	a3c	ADJ
easat-11054	395	11	and	and	CCONJ
easat-11054	395	12	dt	dt	PROPN
easat-11054	395	13	,	,	PUNCT
easat-11054	395	14	which	which	PRON
easat-11054	395	15	combines	combine	VERB
easat-11054	395	16	strengths	strength	NOUN
easat-11054	395	17	and	and	CCONJ
easat-11054	395	18	reduces	reduce	VERB
easat-11054	395	19	each	each	DET
easat-11054	395	20	algorithm	algorithm	NOUN
easat-11054	395	21	’s	’s	PART
easat-11054	395	22	weaknesses	weakness	NOUN
easat-11054	395	23	.	.	PUNCT
easat-11054	396	1	table	table	NOUN
easat-11054	396	2	6	6	NUM
easat-11054	396	3	.	.	PUNCT
easat-11054	397	1	the	the	DET
easat-11054	397	2	performance	performance	NOUN
easat-11054	397	3	comparison	comparison	NOUN
easat-11054	397	4	of	of	ADP
easat-11054	397	5	multi	multi	ADJ
easat-11054	397	6	-	-	NOUN
easat-11054	397	7	classification	classification	NOUN
easat-11054	397	8	.	.	PUNCT
easat-11054	398	1	algorithm	algorithm	NOUN
easat-11054	398	2	19	19	NUM
easat-11054	398	3	classifications	classification	NOUN
easat-11054	398	4	a	a	DET
easat-11054	398	5	p	p	NOUN
easat-11054	398	6	r	r	NOUN
easat-11054	398	7	f1	f1	NOUN
easat-11054	398	8	a3c	a3c	PUNCT
easat-11054	398	9	-	-	PUNCT
easat-11054	398	10	ffnn	ffnn	NOUN
easat-11054	398	11	98.0193	98.0193	NUM
easat-11054	398	12	98.0678	98.0678	NUM
easat-11054	398	13	98.0124	98.0124	NUM
easat-11054	398	14	98.0309	98.0309	NUM
easat-11054	398	15	a3c	a3c	NUM
easat-11054	398	16	-	-	PUNCT
easat-11054	398	17	cnn	cnn	PROPN
easat-11054	398	18	98.5186	98.5186	NUM
easat-11054	398	19	98.5177	98.5177	NUM
easat-11054	398	20	98.5187	98.5187	NUM
easat-11054	398	21	98.5156	98.5156	NUM
easat-11054	398	22	hadt	hadt	NOUN
easat-11054	398	23	99.8305	99.8305	NUM
easat-11054	398	24	99.8304	99.8304	NUM
easat-11054	398	25	99.8308	99.8308	NUM
easat-11054	399	1	99.8291	99.8291	NUM
easat-11054	400	1	this	this	DET
easat-11054	400	2	table	table	NOUN
easat-11054	400	3	displays	display	VERB
easat-11054	400	4	the	the	DET
easat-11054	400	5	number	number	NOUN
easat-11054	400	6	of	of	ADP
easat-11054	400	7	test	test	NOUN
easat-11054	400	8	samples	sample	NOUN
easat-11054	400	9	from	from	ADP
easat-11054	400	10	the	the	DET
easat-11054	400	11	x	x	ADJ
easat-11054	400	12	-	-	ADJ
easat-11054	400	13	iiotid	iiotid	ADJ
easat-11054	400	14	dataset	dataset	NOUN
easat-11054	400	15	,	,	PUNCT
easat-11054	400	16	which	which	PRON
easat-11054	400	17	ensures	ensure	VERB
easat-11054	400	18	the	the	DET
easat-11054	400	19	equivalence	equivalence	NOUN
easat-11054	400	20	of	of	ADP
easat-11054	400	21	experimental	experimental	ADJ
easat-11054	400	22	results	result	NOUN
easat-11054	400	23	.	.	PUNCT
easat-11054	401	1	this	this	PRON
easat-11054	401	2	confirms	confirm	VERB
easat-11054	401	3	the	the	DET
easat-11054	401	4	superiority	superiority	NOUN
easat-11054	401	5	of	of	ADP
easat-11054	401	6	hadt	hadt	NOUN
easat-11054	401	7	over	over	ADP
easat-11054	401	8	a3c	a3c	VERB
easat-11054	401	9	-	-	PUNCT
easat-11054	401	10	ffnn	ffnn	NOUN
easat-11054	401	11	and	and	CCONJ
easat-11054	401	12	a3ccnn	a3ccnn	VERB
easat-11054	401	13	in	in	ADP
easat-11054	401	14	terms	term	NOUN
easat-11054	401	15	of	of	ADP
easat-11054	401	16	detection	detection	NOUN
easat-11054	401	17	accuracy	accuracy	NOUN
easat-11054	401	18	,	,	PUNCT
easat-11054	401	19	precision	precision	NOUN
easat-11054	401	20	,	,	PUNCT
easat-11054	401	21	recall	recall	NOUN
easat-11054	401	22	,	,	PUNCT
easat-11054	401	23	and	and	CCONJ
easat-11054	401	24	an	an	DET
easat-11054	401	25	f1	f1	NOUN
easat-11054	401	26	-	-	PUNCT
easat-11054	401	27	score	score	NOUN
easat-11054	401	28	of	of	ADP
easat-11054	401	29	approximately	approximately	ADV
easat-11054	401	30	99.83	99.83	NUM
easat-11054	401	31	%	%	NOUN
easat-11054	401	32	.	.	PUNCT
easat-11054	402	1	5.4.3	5.4.3	X
easat-11054	402	2	.	.	PUNCT
easat-11054	402	3	resource	resource	NOUN
easat-11054	402	4	consumption	consumption	NOUN
easat-11054	402	5	analysis	analysis	NOUN
easat-11054	402	6	resource	resource	NOUN
easat-11054	402	7	consumption	consumption	NOUN
easat-11054	402	8	was	be	AUX
easat-11054	402	9	evaluated	evaluate	VERB
easat-11054	402	10	based	base	VERB
easat-11054	402	11	on	on	ADP
easat-11054	402	12	three	three	NUM
easat-11054	402	13	factors	factor	NOUN
easat-11054	402	14	:	:	PUNCT
easat-11054	402	15	training	training	NOUN
easat-11054	402	16	time	time	NOUN
easat-11054	402	17	,	,	PUNCT
easat-11054	402	18	inference	inference	NOUN
easat-11054	402	19	time	time	NOUN
easat-11054	402	20	,	,	PUNCT
easat-11054	402	21	and	and	CCONJ
easat-11054	402	22	memory	memory	NOUN
easat-11054	402	23	usage	usage	NOUN
easat-11054	402	24	.	.	PUNCT
easat-11054	403	1	the	the	DET
easat-11054	403	2	training	training	NOUN
easat-11054	403	3	time	time	NOUN
easat-11054	403	4	for	for	ADP
easat-11054	403	5	the	the	DET
easat-11054	403	6	hadt	hadt	NOUN
easat-11054	403	7	algorithm	algorithm	NOUN
easat-11054	403	8	is	be	AUX
easat-11054	403	9	726.99	726.99	NUM
easat-11054	403	10	seconds	second	NOUN
easat-11054	403	11	,	,	PUNCT
easat-11054	403	12	as	as	SCONJ
easat-11054	403	13	shown	show	VERB
easat-11054	403	14	in	in	ADP
easat-11054	403	15	table	table	NOUN
easat-11054	403	16	7	7	NUM
easat-11054	403	17	.	.	PUNCT
easat-11054	404	1	this	this	PRON
easat-11054	404	2	makes	make	VERB
easat-11054	404	3	it	it	PRON
easat-11054	404	4	more	more	ADV
easat-11054	404	5	efficient	efficient	ADJ
easat-11054	404	6	compared	compare	VERB
easat-11054	404	7	to	to	ADP
easat-11054	404	8	other	other	ADJ
easat-11054	404	9	deeper	deep	ADJ
easat-11054	404	10	architectures	architecture	NOUN
easat-11054	404	11	that	that	PRON
easat-11054	404	12	are	be	AUX
easat-11054	404	13	implemented	implement	VERB
easat-11054	404	14	on	on	ADP
easat-11054	404	15	three	three	NUM
easat-11054	404	16	layers	layer	NOUN
easat-11054	404	17	to	to	PART
easat-11054	404	18	achieve	achieve	VERB
easat-11054	404	19	the	the	DET
easat-11054	404	20	best	good	ADJ
easat-11054	404	21	results	result	NOUN
easat-11054	404	22	,	,	PUNCT
easat-11054	404	23	such	such	ADJ
easat-11054	404	24	as	as	ADP
easat-11054	404	25	a3c	a3c	ADJ
easat-11054	404	26	-	-	PUNCT
easat-11054	404	27	ffnn	ffnn	NOUN
easat-11054	404	28	and	and	CCONJ
easat-11054	404	29	a3c	a3c	PROPN
easat-11054	404	30	-	-	PUNCT
easat-11054	404	31	cnn	cnn	NOUN
easat-11054	404	32	.	.	PUNCT
easat-11054	405	1	a3c	a3c	VERB
easat-11054	405	2	-	-	PUNCT
easat-11054	405	3	ffnn	ffnn	NOUN
easat-11054	405	4	,	,	PUNCT
easat-11054	405	5	which	which	PRON
easat-11054	405	6	consists	consist	VERB
easat-11054	405	7	of	of	ADP
easat-11054	405	8	three	three	NUM
easat-11054	405	9	dense	dense	ADJ
easat-11054	405	10	layers	layer	NOUN
easat-11054	405	11	,	,	PUNCT
easat-11054	405	12	was	be	AUX
easat-11054	405	13	found	find	VERB
easat-11054	405	14	to	to	PART
easat-11054	405	15	require	require	VERB
easat-11054	405	16	1013.34	1013.34	NUM
easat-11054	405	17	seconds	second	NOUN
easat-11054	405	18	,	,	PUNCT
easat-11054	405	19	indicating	indicate	VERB
easat-11054	405	20	that	that	SCONJ
easat-11054	405	21	increasing	increase	VERB
easat-11054	405	22	the	the	DET
easat-11054	405	23	number	number	NOUN
easat-11054	405	24	of	of	ADP
easat-11054	405	25	dense	dense	ADJ
easat-11054	405	26	layers	layer	NOUN
easat-11054	405	27	improves	improve	VERB
easat-11054	405	28	the	the	DET
easat-11054	405	29	results	result	NOUN
easat-11054	405	30	but	but	CCONJ
easat-11054	405	31	significantly	significantly	ADV
easat-11054	405	32	impacts	impact	VERB
easat-11054	405	33	the	the	DET
easat-11054	405	34	time	time	NOUN
easat-11054	405	35	required	require	VERB
easat-11054	405	36	to	to	PART
easat-11054	405	37	train	train	VERB
easat-11054	405	38	the	the	DET
easat-11054	405	39	algorithm	algorithm	NOUN
easat-11054	405	40	.	.	PUNCT
easat-11054	406	1	on	on	ADP
easat-11054	406	2	the	the	DET
easat-11054	406	3	other	other	ADJ
easat-11054	406	4	1072	1072	NUM
easat-11054	406	5	edelweiss	edelweiss	PROPN
easat-11054	406	6	applied	apply	VERB
easat-11054	406	7	science	science	NOUN
easat-11054	406	8	and	and	CCONJ
easat-11054	406	9	technology	technology	NOUN
easat-11054	406	10	issn	issn	PROPN
easat-11054	406	11	:	:	PUNCT
easat-11054	406	12	2576	2576	NUM
easat-11054	406	13	-	-	SYM
easat-11054	406	14	8484	8484	NUM
easat-11054	406	15	vol	vol	NOUN
easat-11054	406	16	.	.	PROPN
easat-11054	407	1	9	9	NUM
easat-11054	407	2	,	,	PUNCT
easat-11054	407	3	no	no	INTJ
easat-11054	407	4	.	.	NOUN
easat-11054	407	5	11	11	NUM
easat-11054	407	6	:	:	PUNCT
easat-11054	407	7	1053	1053	NUM
easat-11054	407	8	-	-	SYM
easat-11054	407	9	1075	1075	NUM
easat-11054	407	10	,	,	PUNCT
easat-11054	407	11	2025	2025	NUM
easat-11054	407	12	doi	doi	NOUN
easat-11054	407	13	:	:	PUNCT
easat-11054	407	14	10.55214/2576	10.55214/2576	NUM
easat-11054	407	15	-	-	SYM
easat-11054	407	16	8484.v9i11.11054	8484.v9i11.11054	NOUN
easat-11054	408	1	©	©	PROPN
easat-11054	408	2	2025	2025	NUM
easat-11054	408	3	by	by	ADP
easat-11054	408	4	the	the	DET
easat-11054	408	5	authors	author	NOUN
easat-11054	408	6	;	;	PUNCT
easat-11054	408	7	licensee	licensee	PROPN
easat-11054	408	8	learning	learn	VERB
easat-11054	408	9	gate	gate	PROPN
easat-11054	408	10	hand	hand	NOUN
easat-11054	408	11	,	,	PUNCT
easat-11054	408	12	a3c	a3c	PROPN
easat-11054	408	13	-	-	PUNCT
easat-11054	408	14	cnn	cnn	NOUN
easat-11054	408	15	,	,	PUNCT
easat-11054	408	16	which	which	PRON
easat-11054	408	17	includes	include	VERB
easat-11054	408	18	three	three	NUM
easat-11054	408	19	convolutional	convolutional	ADJ
easat-11054	408	20	layers	layer	NOUN
easat-11054	408	21	,	,	PUNCT
easat-11054	408	22	had	have	VERB
easat-11054	408	23	the	the	DET
easat-11054	408	24	highest	high	ADJ
easat-11054	408	25	training	training	NOUN
easat-11054	408	26	time	time	NOUN
easat-11054	408	27	of	of	ADP
easat-11054	408	28	1552.82	1552.82	NUM
easat-11054	408	29	seconds	second	NOUN
easat-11054	408	30	,	,	PUNCT
easat-11054	408	31	due	due	ADP
easat-11054	408	32	to	to	ADP
easat-11054	408	33	the	the	DET
easat-11054	408	34	increased	increase	VERB
easat-11054	408	35	computational	computational	ADJ
easat-11054	408	36	complexity	complexity	NOUN
easat-11054	408	37	of	of	ADP
easat-11054	408	38	extracting	extract	VERB
easat-11054	408	39	the	the	DET
easat-11054	408	40	most	most	ADV
easat-11054	408	41	important	important	ADJ
easat-11054	408	42	features	feature	NOUN
easat-11054	408	43	.	.	PUNCT
easat-11054	409	1	inference	inference	NOUN
easat-11054	409	2	time	time	NOUN
easat-11054	409	3	is	be	AUX
easat-11054	409	4	particularly	particularly	ADV
easat-11054	409	5	important	important	ADJ
easat-11054	409	6	in	in	ADP
easat-11054	409	7	real	real	ADJ
easat-11054	409	8	-	-	PUNCT
easat-11054	409	9	time	time	NOUN
easat-11054	409	10	intrusion	intrusion	NOUN
easat-11054	409	11	detection	detection	NOUN
easat-11054	409	12	,	,	PUNCT
easat-11054	409	13	enabling	enable	VERB
easat-11054	409	14	rapid	rapid	ADJ
easat-11054	409	15	decisions	decision	NOUN
easat-11054	409	16	to	to	PART
easat-11054	409	17	prevent	prevent	VERB
easat-11054	409	18	security	security	NOUN
easat-11054	409	19	attacks	attack	NOUN
easat-11054	409	20	.	.	PUNCT
easat-11054	410	1	hadt	hadt	PROPN
easat-11054	410	2	's	's	PART
easat-11054	410	3	inference	inference	NOUN
easat-11054	410	4	time	time	NOUN
easat-11054	410	5	was	be	AUX
easat-11054	410	6	28456.32	28456.32	NUM
easat-11054	410	7	milliseconds	millisecond	NOUN
easat-11054	410	8	,	,	PUNCT
easat-11054	410	9	making	make	VERB
easat-11054	410	10	it	it	PRON
easat-11054	410	11	the	the	DET
easat-11054	410	12	most	most	ADV
easat-11054	410	13	efficient	efficient	ADJ
easat-11054	410	14	algorithm	algorithm	NOUN
easat-11054	410	15	among	among	ADP
easat-11054	410	16	the	the	DET
easat-11054	410	17	evaluated	evaluated	ADJ
easat-11054	410	18	algorithms	algorithm	NOUN
easat-11054	410	19	.	.	PUNCT
easat-11054	411	1	a3c	a3c	VERB
easat-11054	411	2	-	-	PUNCT
easat-11054	411	3	ffnn	ffnn	NOUN
easat-11054	411	4	,	,	PUNCT
easat-11054	411	5	due	due	ADP
easat-11054	411	6	to	to	ADP
easat-11054	411	7	its	its	PRON
easat-11054	411	8	additional	additional	ADJ
easat-11054	411	9	dense	dense	ADJ
easat-11054	411	10	layers	layer	NOUN
easat-11054	411	11	,	,	PUNCT
easat-11054	411	12	required	require	VERB
easat-11054	411	13	41799.95	41799.95	NUM
easat-11054	411	14	milliseconds	millisecond	NOUN
easat-11054	411	15	.	.	PUNCT
easat-11054	412	1	while	while	SCONJ
easat-11054	412	2	the	the	DET
easat-11054	412	3	a3c	a3c	NUM
easat-11054	412	4	-	-	PUNCT
easat-11054	412	5	cnn	cnn	NOUN
easat-11054	412	6	algorithm	algorithm	NOUN
easat-11054	412	7	takes	take	VERB
easat-11054	412	8	the	the	DET
easat-11054	412	9	longest	long	ADJ
easat-11054	412	10	inference	inference	NOUN
easat-11054	412	11	time	time	NOUN
easat-11054	412	12	,	,	PUNCT
easat-11054	412	13	at	at	ADP
easat-11054	412	14	59382.51	59382.51	NUM
easat-11054	412	15	milliseconds	millisecond	NOUN
easat-11054	412	16	,	,	PUNCT
easat-11054	412	17	this	this	PRON
easat-11054	412	18	is	be	AUX
easat-11054	412	19	primarily	primarily	ADV
easat-11054	412	20	due	due	ADJ
easat-11054	412	21	to	to	ADP
easat-11054	412	22	the	the	DET
easat-11054	412	23	computational	computational	ADJ
easat-11054	412	24	requirements	requirement	NOUN
easat-11054	412	25	of	of	ADP
easat-11054	412	26	the	the	DET
easat-11054	412	27	convolutional	convolutional	ADJ
easat-11054	412	28	layers	layer	NOUN
easat-11054	412	29	,	,	PUNCT
easat-11054	412	30	which	which	PRON
easat-11054	412	31	process	process	VERB
easat-11054	412	32	feature	feature	NOUN
easat-11054	412	33	maps	map	NOUN
easat-11054	412	34	across	across	ADP
easat-11054	412	35	multiple	multiple	ADJ
easat-11054	412	36	dimensions	dimension	NOUN
easat-11054	412	37	.	.	PUNCT
easat-11054	413	1	regarding	regard	VERB
easat-11054	413	2	memory	memory	NOUN
easat-11054	413	3	usage	usage	NOUN
easat-11054	413	4	,	,	PUNCT
easat-11054	413	5	hadt	hadt	NOUN
easat-11054	413	6	is	be	AUX
easat-11054	413	7	relatively	relatively	ADV
easat-11054	413	8	easy	easy	ADJ
easat-11054	413	9	to	to	PART
easat-11054	413	10	deploy	deploy	VERB
easat-11054	413	11	,	,	PUNCT
easat-11054	413	12	consuming	consume	VERB
easat-11054	413	13	3344.69	3344.69	NUM
easat-11054	413	14	mb	mb	NOUN
easat-11054	413	15	of	of	ADP
easat-11054	413	16	memory	memory	NOUN
easat-11054	413	17	.	.	PUNCT
easat-11054	414	1	despite	despite	SCONJ
easat-11054	414	2	its	its	PRON
easat-11054	414	3	denser	dense	ADJ
easat-11054	414	4	layers	layer	NOUN
easat-11054	414	5	,	,	PUNCT
easat-11054	414	6	a3c	a3c	VERB
easat-11054	414	7	-	-	PUNCT
easat-11054	414	8	ffnn	ffnn	NOUN
easat-11054	414	9	requires	require	VERB
easat-11054	414	10	5315.73	5315.73	NUM
easat-11054	414	11	mb	mb	ADP
easat-11054	414	12	of	of	ADP
easat-11054	414	13	storage	storage	NOUN
easat-11054	414	14	space	space	NOUN
easat-11054	414	15	,	,	PUNCT
easat-11054	414	16	which	which	PRON
easat-11054	414	17	is	be	AUX
easat-11054	414	18	significantly	significantly	ADV
easat-11054	414	19	more	more	ADJ
easat-11054	414	20	memory	memory	NOUN
easat-11054	414	21	than	than	ADP
easat-11054	414	22	hadt	hadt	NOUN
easat-11054	414	23	.	.	PUNCT
easat-11054	415	1	the	the	DET
easat-11054	415	2	a3c	a3c	NUM
easat-11054	415	3	-	-	PUNCT
easat-11054	415	4	cnn	cnn	NOUN
easat-11054	415	5	algorithm	algorithm	NOUN
easat-11054	415	6	consumes	consume	VERB
easat-11054	415	7	7150.31	7150.31	NUM
easat-11054	415	8	mb	mb	VERB
easat-11054	415	9	because	because	SCONJ
easat-11054	415	10	the	the	DET
easat-11054	415	11	convolutional	convolutional	ADJ
easat-11054	415	12	operations	operation	NOUN
easat-11054	415	13	and	and	CCONJ
easat-11054	415	14	feature	feature	NOUN
easat-11054	415	15	maps	map	NOUN
easat-11054	415	16	are	be	AUX
easat-11054	415	17	memory	memory	NOUN
easat-11054	415	18	-	-	PUNCT
easat-11054	415	19	intensive	intensive	ADJ
easat-11054	415	20	.	.	PUNCT
easat-11054	416	1	hadt	hadt	NOUN
easat-11054	416	2	is	be	AUX
easat-11054	416	3	particularly	particularly	ADV
easat-11054	416	4	suitable	suitable	ADJ
easat-11054	416	5	for	for	ADP
easat-11054	416	6	resourceconstrained	resourceconstraine	VERB
easat-11054	416	7	edge	edge	NOUN
easat-11054	416	8	computing	computing	NOUN
easat-11054	416	9	devices	device	NOUN
easat-11054	416	10	.	.	PUNCT
easat-11054	417	1	table	table	NOUN
easat-11054	417	2	7	7	NUM
easat-11054	417	3	.	.	PUNCT
easat-11054	417	4	resource	resource	NOUN
easat-11054	417	5	consumption	consumption	NOUN
easat-11054	417	6	summary	summary	NOUN
easat-11054	417	7	.	.	PUNCT
easat-11054	418	1	layers	layer	NOUN
easat-11054	418	2	(	(	PUNCT
easat-11054	418	3	algorithm	algorithm	NOUN
easat-11054	418	4	/	/	SYM
easat-11054	418	5	layer	layer	NOUN
easat-11054	418	6	)	)	PUNCT
easat-11054	418	7	training	training	NOUN
easat-11054	418	8	time	time	NOUN
easat-11054	418	9	(	(	PUNCT
easat-11054	418	10	s	s	NOUN
easat-11054	418	11	)	)	PUNCT
easat-11054	418	12	inference	inference	NOUN
easat-11054	418	13	time	time	NOUN
easat-11054	418	14	(	(	PUNCT
easat-11054	418	15	ms	ms	NOUN
easat-11054	418	16	)	)	PUNCT
easat-11054	418	17	memory	memory	NOUN
easat-11054	418	18	utilization	utilization	NOUN
easat-11054	418	19	(	(	PUNCT
easat-11054	418	20	mb	mb	NOUN
easat-11054	418	21	)	)	PUNCT
easat-11054	418	22	dt	dt	X
easat-11054	419	1	172.1529	172.1529	NUM
easat-11054	419	2	2274.9143	2274.9143	NOUN
easat-11054	419	3	762.4727	762.4727	NUM
easat-11054	419	4	ffnn	ffnn	NOUN
easat-11054	419	5	in	in	ADP
easat-11054	419	6	a3c	a3c	VERB
easat-11054	419	7	-	-	PUNCT
easat-11054	419	8	ffnn	ffnn	NOUN
easat-11054	419	9	256.5721	256.5721	NUM
easat-11054	419	10	13954.6807	13954.6807	ADJ
easat-11054	419	11	1135.5586	1135.5586	ADJ
easat-11054	419	12	ffnn	ffnn	NOUN
easat-11054	419	13	297.2711	297.2711	NUM
easat-11054	419	14	14226.7299	14226.7299	NUM
easat-11054	419	15	1446.6602	1446.6602	NUM
easat-11054	419	16	(	(	PUNCT
easat-11054	419	17	dt	dt	X
easat-11054	419	18	+	+	CCONJ
easat-11054	419	19	ffnn	ffnn	NOUN
easat-11054	419	20	in	in	ADP
easat-11054	419	21	a3c	a3c	ADJ
easat-11054	419	22	+	+	NUM
easat-11054	419	23	ffnn	ffnn	NOUN
easat-11054	419	24	)	)	PUNCT
easat-11054	419	25	in	in	ADP
easat-11054	419	26	hadt	hadt	NOUN
easat-11054	419	27	726.9961	726.9961	NUM
easat-11054	419	28	28456.3248	28456.3248	NUM
easat-11054	419	29	3344.6914	3344.6914	NUM
easat-11054	419	30	(	(	PUNCT
easat-11054	419	31	3	3	NUM
easat-11054	419	32	dense	dense	ADJ
easat-11054	419	33	layers	layer	NOUN
easat-11054	419	34	)	)	PUNCT
easat-11054	419	35	in	in	ADP
easat-11054	419	36	a3c	a3c	VERB
easat-11054	419	37	-	-	PUNCT
easat-11054	419	38	ffnn	ffnn	NOUN
easat-11054	419	39	1023.3483	1023.3483	NOUN
easat-11054	419	40	41799.9536	41799.9536	NUM
easat-11054	419	41	5315.7334	5315.7334	NUM
easat-11054	419	42	(	(	PUNCT
easat-11054	419	43	3	3	NUM
easat-11054	419	44	conv	conv	ADJ
easat-11054	419	45	layers	layer	NOUN
easat-11054	419	46	)	)	PUNCT
easat-11054	419	47	in	in	ADP
easat-11054	419	48	a3c	a3c	PROPN
easat-11054	419	49	-	-	PUNCT
easat-11054	419	50	cnn	cnn	NOUN
easat-11054	419	51	1552.8285	1552.8285	NUM
easat-11054	420	1	59382.5156	59382.5156	NUM
easat-11054	420	2	7150.3128	7150.3128	PROPN
easat-11054	420	3	5.4.4	5.4.4	NUM
easat-11054	420	4	.	.	PUNCT
easat-11054	421	1	power	power	NOUN
easat-11054	421	2	consumption	consumption	NOUN
easat-11054	421	3	the	the	DET
easat-11054	421	4	average	average	ADJ
easat-11054	421	5	power	power	NOUN
easat-11054	421	6	consumption	consumption	NOUN
easat-11054	421	7	per	per	ADP
easat-11054	421	8	epoch	epoch	NOUN
easat-11054	421	9	for	for	ADP
easat-11054	421	10	the	the	DET
easat-11054	421	11	proposed	propose	VERB
easat-11054	421	12	hadt	hadt	NOUN
easat-11054	421	13	algorithm	algorithm	NOUN
easat-11054	421	14	in	in	ADP
easat-11054	421	15	an	an	DET
easat-11054	421	16	ids	id	NOUN
easat-11054	421	17	for	for	ADP
easat-11054	421	18	iiot	iiot	NOUN
easat-11054	421	19	at	at	ADP
easat-11054	421	20	the	the	DET
easat-11054	421	21	edge	edge	NOUN
easat-11054	421	22	is	be	AUX
easat-11054	421	23	presented	present	VERB
easat-11054	421	24	in	in	ADP
easat-11054	421	25	table	table	NOUN
easat-11054	421	26	8	8	NUM
easat-11054	421	27	.	.	PUNCT
easat-11054	422	1	the	the	DET
easat-11054	422	2	results	result	NOUN
easat-11054	422	3	highlight	highlight	VERB
easat-11054	422	4	the	the	DET
easat-11054	422	5	key	key	ADJ
easat-11054	422	6	aspects	aspect	NOUN
easat-11054	422	7	of	of	ADP
easat-11054	422	8	processing	processing	NOUN
easat-11054	422	9	and	and	CCONJ
easat-11054	422	10	memory	memory	NOUN
easat-11054	422	11	consumed	consume	VERB
easat-11054	422	12	power	power	NOUN
easat-11054	422	13	.	.	PUNCT
easat-11054	423	1	these	these	DET
easat-11054	423	2	results	result	NOUN
easat-11054	423	3	demonstrate	demonstrate	VERB
easat-11054	423	4	the	the	DET
easat-11054	423	5	feasibility	feasibility	NOUN
easat-11054	423	6	of	of	ADP
easat-11054	423	7	deploying	deploy	VERB
easat-11054	423	8	this	this	DET
easat-11054	423	9	algorithm	algorithm	NOUN
easat-11054	423	10	in	in	ADP
easat-11054	423	11	an	an	DET
easat-11054	423	12	edge	edge	NOUN
easat-11054	423	13	computing	compute	VERB
easat-11054	423	14	environment	environment	NOUN
easat-11054	423	15	.	.	PUNCT
easat-11054	424	1	table	table	NOUN
easat-11054	424	2	8	8	NUM
easat-11054	424	3	.	.	PUNCT
easat-11054	425	1	power	power	NOUN
easat-11054	425	2	consumption	consumption	NOUN
easat-11054	425	3	of	of	ADP
easat-11054	425	4	hadt	hadt	NOUN
easat-11054	425	5	.	.	PUNCT
easat-11054	426	1	processing	processing	NOUN
easat-11054	426	2	power	power	NOUN
easat-11054	426	3	(	(	PUNCT
easat-11054	426	4	dt	dt	NOUN
easat-11054	426	5	)	)	PUNCT
easat-11054	426	6	processing	processing	NOUN
easat-11054	426	7	power	power	NOUN
easat-11054	426	8	(	(	PUNCT
easat-11054	426	9	a3c	a3c	ADJ
easat-11054	426	10	ffnn	ffnn	NOUN
easat-11054	426	11	)	)	PUNCT
easat-11054	426	12	processing	processing	NOUN
easat-11054	426	13	power	power	NOUN
easat-11054	426	14	(	(	PUNCT
easat-11054	426	15	ffnn	ffnn	NOUN
easat-11054	426	16	)	)	PUNCT
easat-11054	426	17	memory	memory	NOUN
easat-11054	426	18	power	power	NOUN
easat-11054	426	19	(	(	PUNCT
easat-11054	426	20	dt	dt	NOUN
easat-11054	426	21	)	)	PUNCT
easat-11054	426	22	memory	memory	NOUN
easat-11054	426	23	power	power	NOUN
easat-11054	426	24	(	(	PUNCT
easat-11054	426	25	ffnn	ffnn	NOUN
easat-11054	426	26	in	in	ADP
easat-11054	426	27	a3c	a3c	ADJ
easat-11054	426	28	)	)	PUNCT
easat-11054	426	29	memory	memory	NOUN
easat-11054	426	30	power	power	NOUN
easat-11054	426	31	(	(	PUNCT
easat-11054	426	32	ffnn	ffnn	NOUN
easat-11054	426	33	)	)	PUNCT
easat-11054	426	34	total	total	NOUN
easat-11054	426	35	of	of	ADP
easat-11054	426	36	average	average	ADJ
easat-11054	426	37	power/	power/	NUM
easat-11054	426	38	epoch	epoch	NOUN
easat-11054	426	39	8.595	8.595	NUM
easat-11054	426	40	8.77	8.77	NUM
easat-11054	426	41	8.385	8.385	NUM
easat-11054	426	42	0.48	0.48	NUM
easat-11054	426	43	0.48	0.48	NUM
easat-11054	426	44	0.48	0.48	NUM
easat-11054	426	45	27.19	27.19	NUM
easat-11054	426	46	w	w	ADP
easat-11054	426	47	the	the	DET
easat-11054	426	48	processing	processing	NOUN
easat-11054	426	49	power	power	NOUN
easat-11054	426	50	of	of	ADP
easat-11054	426	51	hadt	hadt	NOUN
easat-11054	426	52	is	be	AUX
easat-11054	426	53	divided	divide	VERB
easat-11054	426	54	into	into	ADP
easat-11054	426	55	three	three	NUM
easat-11054	426	56	main	main	ADJ
easat-11054	426	57	components	component	NOUN
easat-11054	426	58	.	.	PUNCT
easat-11054	427	1	on	on	ADP
easat-11054	427	2	average	average	ADJ
easat-11054	427	3	,	,	PUNCT
easat-11054	427	4	the	the	DET
easat-11054	427	5	dt	dt	PROPN
easat-11054	427	6	component	component	NOUN
easat-11054	427	7	consumes	consume	VERB
easat-11054	427	8	8.595	8.595	NUM
easat-11054	427	9	watts	watt	NOUN
easat-11054	427	10	.	.	PUNCT
easat-11054	428	1	the	the	DET
easat-11054	428	2	a3c	a3c	NUM
easat-11054	428	3	-	-	PUNCT
easat-11054	428	4	ffnn	ffnn	NOUN
easat-11054	428	5	requires	require	VERB
easat-11054	428	6	slightly	slightly	ADV
easat-11054	428	7	more	more	ADJ
easat-11054	428	8	processing	processing	NOUN
easat-11054	428	9	power	power	NOUN
easat-11054	428	10	at	at	ADP
easat-11054	428	11	8.77	8.77	NUM
easat-11054	428	12	watts	watt	NOUN
easat-11054	428	13	.	.	PUNCT
easat-11054	429	1	the	the	DET
easat-11054	429	2	average	average	ADJ
easat-11054	429	3	power	power	NOUN
easat-11054	429	4	consumption	consumption	NOUN
easat-11054	429	5	of	of	ADP
easat-11054	429	6	the	the	DET
easat-11054	429	7	hybrid	hybrid	ADJ
easat-11054	429	8	processing	processing	NOUN
easat-11054	429	9	layer	layer	NOUN
easat-11054	429	10	(	(	PUNCT
easat-11054	429	11	the	the	DET
easat-11054	429	12	ffnn	ffnn	NOUN
easat-11054	429	13	layer	layer	NOUN
easat-11054	429	14	)	)	PUNCT
easat-11054	429	15	,	,	PUNCT
easat-11054	429	16	which	which	PRON
easat-11054	429	17	receives	receive	VERB
easat-11054	429	18	outputs	output	NOUN
easat-11054	429	19	from	from	ADP
easat-11054	429	20	both	both	DET
easat-11054	429	21	dt	dt	X
easat-11054	429	22	and	and	CCONJ
easat-11054	429	23	a3c	a3c	NUM
easat-11054	429	24	,	,	PUNCT
easat-11054	429	25	is	be	AUX
easat-11054	429	26	8.385	8.385	NUM
easat-11054	429	27	watts	watt	NOUN
easat-11054	429	28	,	,	PUNCT
easat-11054	429	29	indicating	indicate	VERB
easat-11054	429	30	a	a	DET
easat-11054	429	31	better	well	ADJ
easat-11054	429	32	balance	balance	NOUN
easat-11054	429	33	between	between	ADP
easat-11054	429	34	decision	decision	NOUN
easat-11054	429	35	-	-	PUNCT
easat-11054	429	36	making	make	VERB
easat-11054	429	37	and	and	CCONJ
easat-11054	429	38	learning	learn	VERB
easat-11054	429	39	capability	capability	NOUN
easat-11054	429	40	.	.	PUNCT
easat-11054	430	1	furthermore	furthermore	ADV
easat-11054	430	2	,	,	PUNCT
easat-11054	430	3	memory	memory	NOUN
easat-11054	430	4	consumes	consume	VERB
easat-11054	430	5	a	a	DET
easat-11054	430	6	constant	constant	ADJ
easat-11054	430	7	0.48	0.48	NUM
easat-11054	430	8	watts	watt	NOUN
easat-11054	430	9	across	across	ADP
easat-11054	430	10	all	all	DET
easat-11054	430	11	components	component	NOUN
easat-11054	430	12	.	.	PUNCT
easat-11054	431	1	as	as	SCONJ
easat-11054	431	2	shown	show	VERB
easat-11054	431	3	in	in	ADP
easat-11054	431	4	figure	figure	NOUN
easat-11054	431	5	11	11	NUM
easat-11054	431	6	,	,	PUNCT
easat-11054	431	7	the	the	DET
easat-11054	431	8	average	average	ADJ
easat-11054	431	9	total	total	ADJ
easat-11054	431	10	power	power	NOUN
easat-11054	431	11	consumption	consumption	NOUN
easat-11054	431	12	of	of	ADP
easat-11054	431	13	hadt	hadt	NOUN
easat-11054	431	14	per	per	ADP
easat-11054	431	15	epoch	epoch	NOUN
easat-11054	431	16	is	be	AUX
easat-11054	431	17	27.19	27.19	NUM
easat-11054	431	18	watts	watt	NOUN
easat-11054	431	19	,	,	PUNCT
easat-11054	431	20	which	which	PRON
easat-11054	431	21	is	be	AUX
easat-11054	431	22	reasonable	reasonable	ADJ
easat-11054	431	23	given	give	VERB
easat-11054	431	24	the	the	DET
easat-11054	431	25	complexity	complexity	NOUN
easat-11054	431	26	of	of	ADP
easat-11054	431	27	real	real	ADJ
easat-11054	431	28	-	-	PUNCT
easat-11054	431	29	time	time	NOUN
easat-11054	431	30	intrusion	intrusion	NOUN
easat-11054	431	31	detection	detection	NOUN
easat-11054	431	32	at	at	ADP
easat-11054	431	33	the	the	DET
easat-11054	431	34	edge	edge	NOUN
easat-11054	431	35	.	.	PUNCT
easat-11054	432	1	while	while	SCONJ
easat-11054	432	2	a3c	a3c	ADV
easat-11054	432	3	-	-	PUNCT
easat-11054	432	4	based	base	VERB
easat-11054	432	5	cnn	cnn	NOUN
easat-11054	432	6	and	and	CCONJ
easat-11054	432	7	ffnn	ffnn	NOUN
easat-11054	432	8	consume	consume	VERB
easat-11054	432	9	more	more	ADJ
easat-11054	432	10	power	power	NOUN
easat-11054	432	11	,	,	PUNCT
easat-11054	432	12	a3c	a3c	PROPN
easat-11054	432	13	-	-	PUNCT
easat-11054	432	14	cnn	cnn	NOUN
easat-11054	432	15	,	,	PUNCT
easat-11054	432	16	in	in	ADP
easat-11054	432	17	particular	particular	ADJ
easat-11054	432	18	,	,	PUNCT
easat-11054	432	19	consumes	consume	VERB
easat-11054	432	20	the	the	DET
easat-11054	432	21	most	most	ADJ
easat-11054	432	22	(	(	PUNCT
easat-11054	432	23	52.57	52.57	NUM
easat-11054	432	24	watts	watt	NOUN
easat-11054	432	25	)	)	PUNCT
easat-11054	432	26	due	due	ADP
easat-11054	432	27	to	to	ADP
easat-11054	432	28	its	its	PRON
easat-11054	432	29	deep	deep	ADJ
easat-11054	432	30	convolutional	convolutional	ADJ
easat-11054	432	31	layers	layer	NOUN
easat-11054	432	32	and	and	CCONJ
easat-11054	432	33	the	the	DET
easat-11054	432	34	high	high	ADJ
easat-11054	432	35	computational	computational	ADJ
easat-11054	432	36	cost	cost	NOUN
easat-11054	432	37	of	of	ADP
easat-11054	432	38	feature	feature	NOUN
easat-11054	432	39	extraction	extraction	NOUN
easat-11054	432	40	and	and	CCONJ
easat-11054	432	41	training	training	NOUN
easat-11054	432	42	.	.	PUNCT
easat-11054	433	1	1073	1073	NUM
easat-11054	433	2	edelweiss	edelweiss	PROPN
easat-11054	433	3	applied	apply	VERB
easat-11054	433	4	science	science	NOUN
easat-11054	433	5	and	and	CCONJ
easat-11054	433	6	technology	technology	NOUN
easat-11054	433	7	issn	issn	PROPN
easat-11054	433	8	:	:	PUNCT
easat-11054	433	9	2576	2576	NUM
easat-11054	433	10	-	-	SYM
easat-11054	433	11	8484	8484	NUM
easat-11054	433	12	vol	vol	NOUN
easat-11054	433	13	.	.	PROPN
easat-11054	434	1	9	9	NUM
easat-11054	434	2	,	,	PUNCT
easat-11054	434	3	no	no	INTJ
easat-11054	434	4	.	.	NOUN
easat-11054	434	5	11	11	NUM
easat-11054	434	6	:	:	PUNCT
easat-11054	434	7	1053	1053	NUM
easat-11054	434	8	-	-	SYM
easat-11054	434	9	1075	1075	NUM
easat-11054	434	10	,	,	PUNCT
easat-11054	434	11	2025	2025	NUM
easat-11054	434	12	doi	doi	NOUN
easat-11054	434	13	:	:	PUNCT
easat-11054	434	14	10.55214/2576	10.55214/2576	NUM
easat-11054	434	15	-	-	SYM
easat-11054	434	16	8484.v9i11.11054	8484.v9i11.11054	NOUN
easat-11054	435	1	©	©	PROPN
easat-11054	435	2	2025	2025	NUM
easat-11054	435	3	by	by	ADP
easat-11054	435	4	the	the	DET
easat-11054	435	5	authors	author	NOUN
easat-11054	435	6	;	;	PUNCT
easat-11054	435	7	licensee	licensee	PROPN
easat-11054	435	8	learning	learning	NOUN
easat-11054	435	9	gate	gate	PROPN
easat-11054	435	10	.	.	PUNCT
easat-11054	436	1	figure	figure	VERB
easat-11054	436	2	11	11	NUM
easat-11054	436	3	.	.	PUNCT
easat-11054	437	1	a	a	DET
easat-11054	437	2	comparison	comparison	NOUN
easat-11054	437	3	of	of	ADP
easat-11054	437	4	the	the	DET
easat-11054	437	5	power	power	NOUN
easat-11054	437	6	consumption	consumption	NOUN
easat-11054	437	7	of	of	ADP
easat-11054	437	8	the	the	DET
easat-11054	437	9	algorithms	algorithm	NOUN
easat-11054	437	10	.	.	PUNCT
easat-11054	438	1	it	it	PRON
easat-11054	438	2	is	be	AUX
easat-11054	438	3	noted	note	VERB
easat-11054	438	4	that	that	SCONJ
easat-11054	438	5	the	the	DET
easat-11054	438	6	a3c	a3c	NUM
easat-11054	438	7	-	-	PUNCT
easat-11054	438	8	ffnn	ffnn	NOUN
easat-11054	438	9	consumes	consume	VERB
easat-11054	438	10	29.32	29.32	NUM
easat-11054	438	11	watts	watt	NOUN
easat-11054	438	12	of	of	ADP
easat-11054	438	13	power	power	NOUN
easat-11054	438	14	,	,	PUNCT
easat-11054	438	15	which	which	PRON
easat-11054	438	16	is	be	AUX
easat-11054	438	17	moderate	moderate	ADJ
easat-11054	438	18	compared	compare	VERB
easat-11054	438	19	to	to	ADP
easat-11054	438	20	the	the	DET
easat-11054	438	21	a3c	a3c	NUM
easat-11054	438	22	-	-	PUNCT
easat-11054	438	23	cnn	cnn	NOUN
easat-11054	438	24	(	(	PUNCT
easat-11054	438	25	52.52	52.52	NUM
easat-11054	438	26	watts	watt	NOUN
easat-11054	438	27	)	)	PUNCT
easat-11054	438	28	.	.	PUNCT
easat-11054	439	1	this	this	PRON
easat-11054	439	2	is	be	AUX
easat-11054	439	3	due	due	ADJ
easat-11054	439	4	to	to	ADP
easat-11054	439	5	the	the	DET
easat-11054	439	6	fact	fact	NOUN
easat-11054	439	7	that	that	SCONJ
easat-11054	439	8	the	the	DET
easat-11054	439	9	computational	computational	ADJ
easat-11054	439	10	costs	cost	NOUN
easat-11054	439	11	of	of	ADP
easat-11054	439	12	the	the	DET
easat-11054	439	13	a3c	a3c	VERB
easat-11054	439	14	-	-	PUNCT
easat-11054	439	15	ffnn	ffnn	NOUN
easat-11054	439	16	are	be	AUX
easat-11054	439	17	lower	low	ADJ
easat-11054	439	18	than	than	ADP
easat-11054	439	19	those	those	PRON
easat-11054	439	20	of	of	ADP
easat-11054	439	21	the	the	DET
easat-11054	439	22	cnn	cnn	PROPN
easat-11054	439	23	layers	layer	NOUN
easat-11054	439	24	in	in	ADP
easat-11054	439	25	a3c	a3c	PROPN
easat-11054	439	26	-	-	PUNCT
easat-11054	439	27	cnn	cnn	NOUN
easat-11054	439	28	.	.	PUNCT
easat-11054	440	1	however	however	ADV
easat-11054	440	2	,	,	PUNCT
easat-11054	440	3	the	the	DET
easat-11054	440	4	a3c	a3c	NUM
easat-11054	440	5	-	-	PUNCT
easat-11054	440	6	ffnn	ffnn	NOUN
easat-11054	440	7	can	can	AUX
easat-11054	440	8	not	not	PART
easat-11054	440	9	extract	extract	VERB
easat-11054	440	10	spatial	spatial	ADJ
easat-11054	440	11	dependencies	dependency	NOUN
easat-11054	440	12	in	in	ADP
easat-11054	440	13	the	the	DET
easat-11054	440	14	iiot	iiot	NOUN
easat-11054	440	15	network	network	NOUN
easat-11054	440	16	,	,	PUNCT
easat-11054	440	17	which	which	PRON
easat-11054	440	18	can	can	AUX
easat-11054	440	19	affect	affect	VERB
easat-11054	440	20	the	the	DET
easat-11054	440	21	sequence	sequence	NOUN
easat-11054	440	22	of	of	ADP
easat-11054	440	23	intrusion	intrusion	NOUN
easat-11054	440	24	detection	detection	NOUN
easat-11054	440	25	performance	performance	NOUN
easat-11054	440	26	.	.	PUNCT
easat-11054	441	1	on	on	ADP
easat-11054	441	2	the	the	DET
easat-11054	441	3	other	other	ADJ
easat-11054	441	4	hand	hand	NOUN
easat-11054	441	5	,	,	PUNCT
easat-11054	441	6	the	the	DET
easat-11054	441	7	hadt	hadt	NOUN
easat-11054	441	8	consumes	consume	VERB
easat-11054	441	9	27.19	27.19	NUM
easat-11054	441	10	watts	watt	NOUN
easat-11054	441	11	.	.	PUNCT
easat-11054	442	1	therefore	therefore	ADV
easat-11054	442	2	,	,	PUNCT
easat-11054	442	3	this	this	DET
easat-11054	442	4	algorithm	algorithm	NOUN
easat-11054	442	5	can	can	AUX
easat-11054	442	6	provide	provide	VERB
easat-11054	442	7	high	high	ADJ
easat-11054	442	8	accuracy	accuracy	NOUN
easat-11054	442	9	in	in	ADP
easat-11054	442	10	intrusion	intrusion	NOUN
easat-11054	442	11	detection	detection	NOUN
easat-11054	442	12	while	while	SCONJ
easat-11054	442	13	maintaining	maintain	VERB
easat-11054	442	14	acceptable	acceptable	ADJ
easat-11054	442	15	power	power	NOUN
easat-11054	442	16	consumption	consumption	NOUN
easat-11054	442	17	,	,	PUNCT
easat-11054	442	18	making	make	VERB
easat-11054	442	19	it	it	PRON
easat-11054	442	20	suitable	suitable	ADJ
easat-11054	442	21	for	for	ADP
easat-11054	442	22	practical	practical	ADJ
easat-11054	442	23	implementation	implementation	NOUN
easat-11054	442	24	in	in	ADP
easat-11054	442	25	ids	id	NOUN
easat-11054	442	26	.	.	PUNCT
easat-11054	443	1	6	6	NUM
easat-11054	443	2	.	.	X
easat-11054	443	3	conclusion	conclusion	NOUN
easat-11054	443	4	this	this	DET
easat-11054	443	5	article	article	NOUN
easat-11054	443	6	proposes	propose	VERB
easat-11054	443	7	the	the	DET
easat-11054	443	8	hadt	hadt	NOUN
easat-11054	443	9	algorithm	algorithm	NOUN
easat-11054	443	10	,	,	PUNCT
easat-11054	443	11	which	which	PRON
easat-11054	443	12	integrates	integrate	VERB
easat-11054	443	13	two	two	NUM
easat-11054	443	14	hybrid	hybrid	ADJ
easat-11054	443	15	algorithms	algorithm	NOUN
easat-11054	443	16	:	:	PUNCT
easat-11054	443	17	a3c	a3c	VERB
easat-11054	443	18	-	-	PUNCT
easat-11054	443	19	ffnn	ffnn	NOUN
easat-11054	443	20	and	and	CCONJ
easat-11054	443	21	dt	dt	NOUN
easat-11054	443	22	to	to	PART
easat-11054	443	23	improve	improve	VERB
easat-11054	443	24	intrusion	intrusion	NOUN
easat-11054	443	25	detection	detection	NOUN
easat-11054	443	26	in	in	ADP
easat-11054	443	27	the	the	DET
easat-11054	443	28	iiot	iiot	NOUN
easat-11054	443	29	environment	environment	NOUN
easat-11054	443	30	.	.	PUNCT
easat-11054	444	1	this	this	DET
easat-11054	444	2	hadt	hadt	NOUN
easat-11054	444	3	algorithm	algorithm	NOUN
easat-11054	444	4	,	,	PUNCT
easat-11054	444	5	which	which	PRON
easat-11054	444	6	uses	use	VERB
easat-11054	444	7	fewer	few	ADJ
easat-11054	444	8	neural	neural	ADJ
easat-11054	444	9	network	network	NOUN
easat-11054	444	10	layers	layer	NOUN
easat-11054	444	11	to	to	PART
easat-11054	444	12	reduce	reduce	VERB
easat-11054	444	13	computational	computational	ADJ
easat-11054	444	14	overhead	overhead	NOUN
easat-11054	444	15	and	and	CCONJ
easat-11054	444	16	improve	improve	VERB
easat-11054	444	17	processing	processing	NOUN
easat-11054	444	18	time	time	NOUN
easat-11054	444	19	and	and	CCONJ
easat-11054	444	20	energy	energy	NOUN
easat-11054	444	21	efficiency	efficiency	NOUN
easat-11054	444	22	,	,	PUNCT
easat-11054	444	23	addresses	address	VERB
easat-11054	444	24	the	the	DET
easat-11054	444	25	challenges	challenge	NOUN
easat-11054	444	26	that	that	PRON
easat-11054	444	27	exist	exist	VERB
easat-11054	444	28	in	in	ADP
easat-11054	444	29	the	the	DET
easat-11054	444	30	iiot	iiot	NOUN
easat-11054	444	31	environment	environment	NOUN
easat-11054	444	32	,	,	PUNCT
easat-11054	444	33	including	include	VERB
easat-11054	444	34	different	different	ADJ
easat-11054	444	35	attack	attack	NOUN
easat-11054	444	36	patterns	pattern	NOUN
easat-11054	444	37	of	of	ADP
easat-11054	444	38	a	a	DET
easat-11054	444	39	stochastic	stochastic	NOUN
easat-11054	444	40	or	or	CCONJ
easat-11054	444	41	deterministic	deterministic	ADJ
easat-11054	444	42	nature	nature	NOUN
easat-11054	444	43	.	.	PUNCT
easat-11054	445	1	this	this	PRON
easat-11054	445	2	is	be	AUX
easat-11054	445	3	due	due	ADJ
easat-11054	445	4	to	to	ADP
easat-11054	445	5	the	the	DET
easat-11054	445	6	presence	presence	NOUN
easat-11054	445	7	of	of	ADP
easat-11054	445	8	features	feature	NOUN
easat-11054	445	9	with	with	ADP
easat-11054	445	10	continuous	continuous	ADJ
easat-11054	445	11	and	and	CCONJ
easat-11054	445	12	discrete	discrete	ADJ
easat-11054	445	13	values	value	NOUN
easat-11054	445	14	in	in	ADP
easat-11054	445	15	the	the	DET
easat-11054	445	16	x	x	ADJ
easat-11054	445	17	-	-	ADJ
easat-11054	445	18	iiotid	iiotid	ADJ
easat-11054	445	19	dataset	dataset	NOUN
easat-11054	445	20	.	.	PUNCT
easat-11054	446	1	therefore	therefore	ADV
easat-11054	446	2	,	,	PUNCT
easat-11054	446	3	the	the	DET
easat-11054	446	4	proposed	propose	VERB
easat-11054	446	5	algorithm	algorithm	NOUN
easat-11054	446	6	leverages	leverage	VERB
easat-11054	446	7	the	the	DET
easat-11054	446	8	integration	integration	NOUN
easat-11054	446	9	advantages	advantage	NOUN
easat-11054	446	10	of	of	ADP
easat-11054	446	11	dt	dt	PUNCT
easat-11054	446	12	by	by	ADP
easat-11054	446	13	dealing	deal	VERB
easat-11054	446	14	with	with	ADP
easat-11054	446	15	data	datum	NOUN
easat-11054	446	16	from	from	ADP
easat-11054	446	17	a	a	DET
easat-11054	446	18	deterministic	deterministic	ADJ
easat-11054	446	19	environment	environment	NOUN
easat-11054	446	20	.	.	PUNCT
easat-11054	447	1	in	in	ADP
easat-11054	447	2	addition	addition	NOUN
easat-11054	447	3	,	,	PUNCT
easat-11054	447	4	the	the	DET
easat-11054	447	5	simple	simple	ADJ
easat-11054	447	6	a3c	a3c	NUM
easat-11054	447	7	is	be	AUX
easat-11054	447	8	more	more	ADV
easat-11054	447	9	suitable	suitable	ADJ
easat-11054	447	10	to	to	PART
easat-11054	447	11	be	be	AUX
easat-11054	447	12	applied	apply	VERB
easat-11054	447	13	in	in	ADP
easat-11054	447	14	a	a	DET
easat-11054	447	15	stochastic	stochastic	ADJ
easat-11054	447	16	environment	environment	NOUN
easat-11054	447	17	using	use	VERB
easat-11054	447	18	the	the	DET
easat-11054	447	19	ffnn	ffnn	NOUN
easat-11054	447	20	layer	layer	NOUN
easat-11054	447	21	in	in	ADP
easat-11054	447	22	its	its	PRON
easat-11054	447	23	worker	worker	NOUN
easat-11054	447	24	to	to	PART
easat-11054	447	25	reduce	reduce	VERB
easat-11054	447	26	the	the	DET
easat-11054	447	27	consumption	consumption	NOUN
easat-11054	447	28	power	power	NOUN
easat-11054	447	29	due	due	ADP
easat-11054	447	30	to	to	ADP
easat-11054	447	31	its	its	PRON
easat-11054	447	32	simplicity	simplicity	NOUN
easat-11054	447	33	in	in	ADP
easat-11054	447	34	calculating	calculate	VERB
easat-11054	447	35	the	the	DET
easat-11054	447	36	processing	processing	NOUN
easat-11054	447	37	during	during	ADP
easat-11054	447	38	training	training	NOUN
easat-11054	447	39	and	and	CCONJ
easat-11054	447	40	testing	testing	NOUN
easat-11054	447	41	.	.	PUNCT
easat-11054	448	1	the	the	DET
easat-11054	448	2	hadt	hadt	NOUN
easat-11054	448	3	algorithm	algorithm	NOUN
easat-11054	448	4	outperformed	outperform	VERB
easat-11054	448	5	the	the	DET
easat-11054	448	6	current	current	ADJ
easat-11054	448	7	a3c	a3c	NUM
easat-11054	448	8	algorithms	algorithm	NOUN
easat-11054	448	9	,	,	PUNCT
easat-11054	448	10	including	include	VERB
easat-11054	448	11	a3c	a3c	VERB
easat-11054	448	12	-	-	PUNCT
easat-11054	448	13	ffnn	ffnn	NOUN
easat-11054	448	14	and	and	CCONJ
easat-11054	448	15	a3ccnn	a3ccnn	VERB
easat-11054	448	16	,	,	PUNCT
easat-11054	448	17	achieving	achieve	VERB
easat-11054	448	18	better	well	ADJ
easat-11054	448	19	results	result	NOUN
easat-11054	448	20	with	with	ADP
easat-11054	448	21	weighted	weighted	ADJ
easat-11054	448	22	accuracy	accuracy	NOUN
easat-11054	448	23	,	,	PUNCT
easat-11054	448	24	precision	precision	NOUN
easat-11054	448	25	,	,	PUNCT
easat-11054	448	26	recall	recall	NOUN
easat-11054	448	27	,	,	PUNCT
easat-11054	448	28	and	and	CCONJ
easat-11054	448	29	f1	f1	NOUN
easat-11054	448	30	-	-	PUNCT
easat-11054	448	31	scores	score	NOUN
easat-11054	448	32	exceeding	exceed	VERB
easat-11054	448	33	99.8	99.8	NUM
easat-11054	448	34	%	%	NOUN
easat-11054	448	35	across	across	ADP
easat-11054	448	36	both	both	CCONJ
easat-11054	448	37	binary	binary	ADJ
easat-11054	448	38	and	and	CCONJ
easat-11054	448	39	multi	multi	ADJ
easat-11054	448	40	-	-	ADJ
easat-11054	448	41	class	class	ADJ
easat-11054	448	42	classifications	classification	NOUN
easat-11054	448	43	.	.	PUNCT
easat-11054	449	1	in	in	ADP
easat-11054	449	2	addition	addition	NOUN
easat-11054	449	3	,	,	PUNCT
easat-11054	449	4	the	the	DET
easat-11054	449	5	proposed	propose	VERB
easat-11054	449	6	hadt	hadt	NOUN
easat-11054	449	7	displays	display	VERB
easat-11054	449	8	remarkable	remarkable	ADJ
easat-11054	449	9	proficiency	proficiency	NOUN
easat-11054	449	10	in	in	ADP
easat-11054	449	11	identifying	identify	VERB
easat-11054	449	12	low	low	ADJ
easat-11054	449	13	sample	sample	NOUN
easat-11054	449	14	sizes	size	NOUN
easat-11054	449	15	of	of	ADP
easat-11054	449	16	attack	attack	NOUN
easat-11054	449	17	classes	class	NOUN
easat-11054	449	18	,	,	PUNCT
easat-11054	449	19	effectively	effectively	ADV
easat-11054	449	20	alleviating	alleviate	VERB
easat-11054	449	21	a	a	DET
easat-11054	449	22	significant	significant	ADJ
easat-11054	449	23	defect	defect	NOUN
easat-11054	449	24	in	in	ADP
easat-11054	449	25	current	current	ADJ
easat-11054	449	26	systems	system	NOUN
easat-11054	449	27	.	.	PUNCT
easat-11054	450	1	the	the	DET
easat-11054	450	2	results	result	NOUN
easat-11054	450	3	show	show	VERB
easat-11054	450	4	that	that	SCONJ
easat-11054	450	5	the	the	DET
easat-11054	450	6	proposed	propose	VERB
easat-11054	450	7	hadt	hadt	NOUN
easat-11054	450	8	algorithm	algorithm	NOUN
easat-11054	450	9	is	be	AUX
easat-11054	450	10	more	more	ADV
easat-11054	450	11	effective	effective	ADJ
easat-11054	450	12	in	in	ADP
easat-11054	450	13	identifying	identify	VERB
easat-11054	450	14	different	different	ADJ
easat-11054	450	15	classes	class	NOUN
easat-11054	450	16	of	of	ADP
easat-11054	450	17	attacks	attack	NOUN
easat-11054	450	18	,	,	PUNCT
easat-11054	450	19	making	make	VERB
easat-11054	450	20	it	it	PRON
easat-11054	450	21	a	a	DET
easat-11054	450	22	strong	strong	ADJ
easat-11054	450	23	algorithm	algorithm	NOUN
easat-11054	450	24	for	for	ADP
easat-11054	450	25	iiot	iiot	NOUN
easat-11054	450	26	security	security	NOUN
easat-11054	450	27	.	.	PUNCT
easat-11054	451	1	despite	despite	SCONJ
easat-11054	451	2	the	the	DET
easat-11054	451	3	strong	strong	ADJ
easat-11054	451	4	performance	performance	NOUN
easat-11054	451	5	of	of	ADP
easat-11054	451	6	this	this	DET
easat-11054	451	7	algorithm	algorithm	NOUN
easat-11054	451	8	,	,	PUNCT
easat-11054	451	9	it	it	PRON
easat-11054	451	10	faces	face	VERB
easat-11054	451	11	some	some	DET
easat-11054	451	12	limitations	limitation	NOUN
easat-11054	451	13	.	.	PUNCT
easat-11054	452	1	therefore	therefore	ADV
easat-11054	452	2	,	,	PUNCT
easat-11054	452	3	future	future	ADJ
easat-11054	452	4	studies	study	NOUN
easat-11054	452	5	should	should	AUX
easat-11054	452	6	focus	focus	VERB
easat-11054	452	7	on	on	ADP
easat-11054	452	8	improving	improve	VERB
easat-11054	452	9	energy	energy	NOUN
easat-11054	452	10	consumption	consumption	NOUN
easat-11054	452	11	and	and	CCONJ
easat-11054	452	12	adaptive	adaptive	NOUN
easat-11054	452	13	learning	learning	NOUN
easat-11054	452	14	to	to	PART
easat-11054	452	15	increase	increase	VERB
easat-11054	452	16	the	the	DET
easat-11054	452	17	efficiency	efficiency	NOUN
easat-11054	452	18	in	in	ADP
easat-11054	452	19	detecting	detect	VERB
easat-11054	452	20	and	and	CCONJ
easat-11054	452	21	mitigating	mitigate	VERB
easat-11054	452	22	zero	zero	NUM
easat-11054	452	23	-	-	PUNCT
easat-11054	452	24	day	day	NOUN
easat-11054	452	25	attacks	attack	NOUN
easat-11054	452	26	in	in	ADP
easat-11054	452	27	real	real	ADJ
easat-11054	452	28	-	-	PUNCT
easat-11054	452	29	time	time	NOUN
easat-11054	452	30	,	,	PUNCT
easat-11054	452	31	especially	especially	ADV
easat-11054	452	32	with	with	ADP
easat-11054	452	33	the	the	DET
easat-11054	452	34	continuous	continuous	ADJ
easat-11054	452	35	expansion	expansion	NOUN
easat-11054	452	36	of	of	ADP
easat-11054	452	37	iiot	iiot	NOUN
easat-11054	452	38	networks	network	NOUN
easat-11054	452	39	with	with	ADP
easat-11054	452	40	an	an	DET
easat-11054	452	41	increasing	increase	VERB
easat-11054	452	42	number	number	NOUN
easat-11054	452	43	of	of	ADP
easat-11054	452	44	data	datum	NOUN
easat-11054	452	45	pattern	pattern	NOUN
easat-11054	452	46	classes	class	NOUN
easat-11054	452	47	.	.	PUNCT
easat-11054	453	1	overall	overall	ADV
easat-11054	453	2	,	,	PUNCT
easat-11054	453	3	this	this	DET
easat-11054	453	4	work	work	NOUN
easat-11054	453	5	makes	make	VERB
easat-11054	453	6	a	a	DET
easat-11054	453	7	significant	significant	ADJ
easat-11054	453	8	contribution	contribution	NOUN
easat-11054	453	9	to	to	ADP
easat-11054	453	10	the	the	DET
easat-11054	453	11	security	security	NOUN
easat-11054	453	12	of	of	ADP
easat-11054	453	13	iiot	iiot	NOUN
easat-11054	453	14	,	,	PUNCT
easat-11054	453	15	paving	pave	VERB
easat-11054	453	16	the	the	DET
easat-11054	453	17	way	way	NOUN
easat-11054	453	18	for	for	ADP
easat-11054	453	19	idss	idss	NOUN
easat-11054	453	20	in	in	ADP
easat-11054	453	21	edge	edge	NOUN
easat-11054	453	22	computing	compute	VERB
easat-11054	453	23	to	to	PART
easat-11054	453	24	be	be	AUX
easat-11054	453	25	more	more	ADV
easat-11054	453	26	accurate	accurate	ADJ
easat-11054	453	27	and	and	CCONJ
easat-11054	453	28	intelligent	intelligent	ADJ
easat-11054	453	29	in	in	ADP
easat-11054	453	30	this	this	DET
easat-11054	453	31	complex	complex	ADJ
easat-11054	453	32	environment	environment	NOUN
easat-11054	453	33	.	.	PUNCT
easat-11054	454	1	1074	1074	NUM
easat-11054	454	2	edelweiss	edelweiss	PROPN
easat-11054	454	3	applied	apply	VERB
easat-11054	454	4	science	science	NOUN
easat-11054	454	5	and	and	CCONJ
easat-11054	454	6	technology	technology	NOUN
easat-11054	454	7	issn	issn	PROPN
easat-11054	454	8	:	:	PUNCT
easat-11054	454	9	2576	2576	NUM
easat-11054	454	10	-	-	SYM
easat-11054	454	11	8484	8484	NUM
easat-11054	454	12	vol	vol	NOUN
easat-11054	454	13	.	.	PROPN
easat-11054	455	1	9	9	NUM
easat-11054	455	2	,	,	PUNCT
easat-11054	455	3	no	no	INTJ
easat-11054	455	4	.	.	NOUN
easat-11054	455	5	11	11	NUM
easat-11054	455	6	:	:	PUNCT
easat-11054	455	7	1053	1053	NUM
easat-11054	455	8	-	-	SYM
easat-11054	455	9	1075	1075	NUM
easat-11054	455	10	,	,	PUNCT
easat-11054	455	11	2025	2025	NUM
easat-11054	455	12	doi	doi	NOUN
easat-11054	455	13	:	:	PUNCT
easat-11054	455	14	10.55214/2576	10.55214/2576	NUM
easat-11054	455	15	-	-	SYM
easat-11054	455	16	8484.v9i11.11054	8484.v9i11.11054	NOUN
easat-11054	456	1	©	©	PROPN
easat-11054	456	2	2025	2025	NUM
easat-11054	456	3	by	by	ADP
easat-11054	456	4	the	the	DET
easat-11054	456	5	authors	author	NOUN
easat-11054	456	6	;	;	PUNCT
easat-11054	456	7	licensee	licensee	PROPN
easat-11054	456	8	learning	learn	VERB
easat-11054	456	9	gate	gate	NOUN
easat-11054	456	10	transparency	transparency	NOUN
easat-11054	456	11	:	:	PUNCT
easat-11054	456	12	the	the	DET
easat-11054	456	13	authors	author	NOUN
easat-11054	456	14	confirm	confirm	VERB
easat-11054	456	15	that	that	SCONJ
easat-11054	456	16	the	the	DET
easat-11054	456	17	manuscript	manuscript	NOUN
easat-11054	456	18	is	be	AUX
easat-11054	456	19	an	an	DET
easat-11054	456	20	honest	honest	ADJ
easat-11054	456	21	,	,	PUNCT
easat-11054	456	22	accurate	accurate	ADJ
easat-11054	456	23	,	,	PUNCT
easat-11054	456	24	and	and	CCONJ
easat-11054	456	25	transparent	transparent	ADJ
easat-11054	456	26	account	account	NOUN
easat-11054	456	27	of	of	ADP
easat-11054	456	28	the	the	DET
easat-11054	456	29	study	study	NOUN
easat-11054	456	30	;	;	PUNCT
easat-11054	456	31	that	that	SCONJ
easat-11054	456	32	no	no	DET
easat-11054	456	33	vital	vital	ADJ
easat-11054	456	34	features	feature	NOUN
easat-11054	456	35	of	of	ADP
easat-11054	456	36	the	the	DET
easat-11054	456	37	study	study	NOUN
easat-11054	456	38	have	have	AUX
easat-11054	456	39	been	be	AUX
easat-11054	456	40	omitted	omit	VERB
easat-11054	456	41	;	;	PUNCT
easat-11054	456	42	and	and	CCONJ
easat-11054	456	43	that	that	SCONJ
easat-11054	456	44	any	any	DET
easat-11054	456	45	discrepancies	discrepancy	NOUN
easat-11054	456	46	from	from	ADP
easat-11054	456	47	the	the	DET
easat-11054	456	48	study	study	NOUN
easat-11054	456	49	as	as	SCONJ
easat-11054	456	50	planned	plan	VERB
easat-11054	456	51	have	have	AUX
easat-11054	456	52	been	be	AUX
easat-11054	456	53	explained	explain	VERB
easat-11054	456	54	.	.	PUNCT
easat-11054	457	1	this	this	DET
easat-11054	457	2	study	study	NOUN
easat-11054	457	3	followed	follow	VERB
easat-11054	457	4	all	all	DET
easat-11054	457	5	ethical	ethical	ADJ
easat-11054	457	6	practices	practice	NOUN
easat-11054	457	7	during	during	ADP
easat-11054	457	8	writing	writing	NOUN
easat-11054	457	9	.	.	PUNCT
easat-11054	458	1	acknowledgment	acknowledgment	NOUN
easat-11054	458	2	:	:	PUNCT
easat-11054	458	3	the	the	DET
easat-11054	458	4	author	author	NOUN
easat-11054	458	5	would	would	AUX
easat-11054	458	6	like	like	VERB
easat-11054	458	7	to	to	PART
easat-11054	458	8	express	express	VERB
easat-11054	458	9	sincere	sincere	ADJ
easat-11054	458	10	gratitude	gratitude	NOUN
easat-11054	458	11	to	to	ADP
easat-11054	458	12	all	all	DET
easat-11054	458	13	those	those	PRON
easat-11054	458	14	who	who	PRON
easat-11054	458	15	provided	provide	VERB
easat-11054	458	16	support	support	NOUN
easat-11054	458	17	,	,	PUNCT
easat-11054	458	18	guidance	guidance	NOUN
easat-11054	458	19	,	,	PUNCT
easat-11054	458	20	and	and	CCONJ
easat-11054	458	21	encouragement	encouragement	NOUN
easat-11054	458	22	throughout	throughout	ADP
easat-11054	458	23	the	the	DET
easat-11054	458	24	completion	completion	NOUN
easat-11054	458	25	of	of	ADP
easat-11054	458	26	this	this	DET
easat-11054	458	27	research	research	NOUN
easat-11054	458	28	.	.	PUNCT
easat-11054	459	1	special	special	ADJ
easat-11054	459	2	appreciation	appreciation	NOUN
easat-11054	459	3	is	be	AUX
easat-11054	459	4	extended	extend	VERB
easat-11054	459	5	to	to	ADP
easat-11054	459	6	al	al	PROPN
easat-11054	459	7	-	-	PUNCT
easat-11054	459	8	ghad	ghad	PROPN
easat-11054	459	9	college	college	NOUN
easat-11054	459	10	of	of	ADP
easat-11054	459	11	technical	technical	PROPN
easat-11054	459	12	health	health	NOUN
easat-11054	459	13	sciences	science	NOUN
easat-11054	459	14	for	for	ADP
easat-11054	459	15	its	its	PRON
easat-11054	459	16	continuous	continuous	ADJ
easat-11054	459	17	support	support	NOUN
easat-11054	459	18	.	.	PUNCT
easat-11054	460	1	copyright	copyright	NOUN
easat-11054	460	2	:	:	PUNCT
easat-11054	460	3	©	©	PROPN
easat-11054	460	4	2025	2025	NUM
easat-11054	460	5	by	by	ADP
easat-11054	460	6	the	the	DET
easat-11054	460	7	authors	author	NOUN
easat-11054	460	8	.	.	PUNCT
easat-11054	461	1	this	this	DET
easat-11054	461	2	article	article	NOUN
easat-11054	461	3	is	be	AUX
easat-11054	461	4	an	an	DET
easat-11054	461	5	open	open	ADJ
easat-11054	461	6	-	-	PUNCT
easat-11054	461	7	access	access	NOUN
easat-11054	461	8	article	article	NOUN
easat-11054	461	9	distributed	distribute	VERB
easat-11054	461	10	under	under	ADP
easat-11054	461	11	the	the	DET
easat-11054	461	12	terms	term	NOUN
easat-11054	461	13	and	and	CCONJ
easat-11054	461	14	conditions	condition	NOUN
easat-11054	461	15	of	of	ADP
easat-11054	461	16	the	the	DET
easat-11054	461	17	creative	creative	ADJ
easat-11054	461	18	commons	common	NOUN
easat-11054	461	19	attribution	attribution	NOUN
easat-11054	461	20	(	(	PUNCT
easat-11054	461	21	cc	cc	NOUN
easat-11054	461	22	by	by	ADP
easat-11054	461	23	)	)	PUNCT
easat-11054	461	24	license	license	NOUN
easat-11054	461	25	(	(	PUNCT
easat-11054	461	26	https://creativecommons.org/licenses/by/4.0/	https://creativecommons.org/licenses/by/4.0/	PROPN
easat-11054	461	27	)	)	PUNCT
easat-11054	461	28	.	.	PUNCT
easat-11054	462	1	references	reference	NOUN
easat-11054	462	2	[	[	X
easat-11054	462	3	1	1	NUM
easat-11054	462	4	]	]	PUNCT
easat-11054	462	5	a.	a.	NOUN
easat-11054	462	6	guezzaz	guezzaz	PROPN
easat-11054	462	7	,	,	PUNCT
easat-11054	462	8	m.	m.	NOUN
easat-11054	462	9	azrour	azrour	PROPN
easat-11054	462	10	,	,	PUNCT
easat-11054	462	11	s.	s.	PROPN
easat-11054	462	12	benkirane	benkirane	PROPN
easat-11054	462	13	,	,	PUNCT
easat-11054	462	14	m.	m.	NOUN
easat-11054	462	15	mohy	mohy	NOUN
easat-11054	462	16	-	-	PUNCT
easat-11054	462	17	eddine	eddine	NOUN
easat-11054	462	18	,	,	PUNCT
easat-11054	462	19	h.	h.	PROPN
easat-11054	462	20	attou	attou	PROPN
easat-11054	462	21	,	,	PUNCT
easat-11054	462	22	and	and	CCONJ
easat-11054	462	23	m.	m.	PROPN
easat-11054	462	24	douiba	douiba	PROPN
easat-11054	462	25	,	,	PUNCT
easat-11054	462	26	"	"	PUNCT
easat-11054	462	27	a	a	DET
easat-11054	462	28	lightweight	lightweight	ADJ
easat-11054	462	29	hybrid	hybrid	NOUN
easat-11054	462	30	intrusion	intrusion	NOUN
easat-11054	462	31	detection	detection	NOUN
easat-11054	462	32	framework	framework	NOUN
easat-11054	462	33	using	use	VERB
easat-11054	462	34	machine	machine	NOUN
easat-11054	462	35	learning	learning	NOUN
easat-11054	462	36	for	for	ADP
easat-11054	462	37	edge	edge	NOUN
easat-11054	462	38	-	-	PUNCT
easat-11054	462	39	based	base	VERB
easat-11054	462	40	iiot	iiot	NOUN
easat-11054	462	41	security	security	NOUN
easat-11054	462	42	,	,	PUNCT
easat-11054	462	43	"	"	PUNCT
easat-11054	462	44	international	international	ADJ
easat-11054	462	45	arab	arab	ADJ
easat-11054	462	46	journal	journal	PROPN
easat-11054	462	47	of	of	ADP
easat-11054	462	48	information	information	NOUN
easat-11054	462	49	technology	technology	NOUN
easat-11054	462	50	,	,	PUNCT
easat-11054	462	51	vol	vol	NOUN
easat-11054	462	52	.	.	PROPN
easat-11054	462	53	19	19	NUM
easat-11054	462	54	,	,	PUNCT
easat-11054	462	55	no	no	INTJ
easat-11054	462	56	.	.	NOUN
easat-11054	462	57	5	5	NUM
easat-11054	462	58	,	,	PUNCT
easat-11054	462	59	pp	pp	ADJ
easat-11054	462	60	.	.	PUNCT
easat-11054	463	1	822–830	822–830	NUM
easat-11054	463	2	,	,	PUNCT
easat-11054	463	3	2022	2022	NUM
easat-11054	463	4	.	.	PUNCT
easat-11054	464	1	https://doi.org/10.34028/iajit/19/5/14	https://doi.org/10.34028/iajit/19/5/14	VERB
easat-11054	464	2	[	[	X
easat-11054	464	3	2	2	NUM
easat-11054	464	4	]	]	PUNCT
easat-11054	464	5	m.	m.	NOUN
easat-11054	464	6	a.	a.	PROPN
easat-11054	464	7	q.	q.	PROPN
easat-11054	464	8	al	al	PROPN
easat-11054	464	9	-	-	PUNCT
easat-11054	464	10	riyashi	riyashi	PROPN
easat-11054	464	11	,	,	PUNCT
easat-11054	464	12	a.	a.	NOUN
easat-11054	464	13	t.	t.	PROPN
easat-11054	464	14	zahary	zahary	PROPN
easat-11054	464	15	,	,	PUNCT
easat-11054	464	16	and	and	CCONJ
easat-11054	464	17	f.	f.	PROPN
easat-11054	464	18	saeed	saeed	PROPN
easat-11054	464	19	,	,	PUNCT
easat-11054	464	20	"	"	PUNCT
easat-11054	464	21	a	a	DET
easat-11054	464	22	review	review	NOUN
easat-11054	464	23	for	for	ADP
easat-11054	464	24	fog	fog	ADJ
easat-11054	464	25	-	-	PUNCT
easat-11054	464	26	cloud	cloud	NOUN
easat-11054	464	27	security	security	NOUN
easat-11054	464	28	:	:	PUNCT
easat-11054	464	29	aspects	aspect	NOUN
easat-11054	464	30	,	,	PUNCT
easat-11054	464	31	attacks	attack	NOUN
easat-11054	464	32	,	,	PUNCT
easat-11054	464	33	solutions	solution	NOUN
easat-11054	464	34	,	,	PUNCT
easat-11054	464	35	and	and	CCONJ
easat-11054	464	36	trends	trend	NOUN
easat-11054	464	37	,	,	PUNCT
easat-11054	464	38	"	"	PUNCT
easat-11054	464	39	sana'a	sana'a	PROPN
easat-11054	464	40	university	university	NOUN
easat-11054	464	41	journal	journal	NOUN
easat-11054	464	42	of	of	ADP
easat-11054	464	43	applied	apply	VERB
easat-11054	464	44	sciences	science	NOUN
easat-11054	464	45	and	and	CCONJ
easat-11054	464	46	technology	technology	NOUN
easat-11054	464	47	,	,	PUNCT
easat-11054	464	48	vol	vol	NOUN
easat-11054	464	49	.	.	PROPN
easat-11054	464	50	2	2	NUM
easat-11054	464	51	,	,	PUNCT
easat-11054	464	52	no	no	INTJ
easat-11054	464	53	.	.	NOUN
easat-11054	464	54	5	5	NUM
easat-11054	464	55	,	,	PUNCT
easat-11054	464	56	pp	pp	ADJ
easat-11054	464	57	.	.	PUNCT
easat-11054	465	1	482	482	NUM
easat-11054	465	2	-	-	SYM
easat-11054	465	3	491	491	NUM
easat-11054	465	4	,	,	PUNCT
easat-11054	465	5	2024	2024	NUM
easat-11054	465	6	.	.	PUNCT
easat-11054	466	1	https://doi.org/10.59628/jast.v2i5.1128	https://doi.org/10.59628/jast.v2i5.1128	X
easat-11054	467	1	[	[	X
easat-11054	467	2	3	3	X
easat-11054	467	3	]	]	PUNCT
easat-11054	467	4	s.	s.	PROPN
easat-11054	467	5	abdelkader	abdelkader	PROPN
easat-11054	467	6	et	et	PROPN
easat-11054	467	7	al	al	PROPN
easat-11054	467	8	.	.	PROPN
easat-11054	467	9	,	,	PUNCT
easat-11054	467	10	"	"	PUNCT
easat-11054	467	11	securing	secure	VERB
easat-11054	467	12	modern	modern	ADJ
easat-11054	467	13	power	power	NOUN
easat-11054	467	14	systems	system	NOUN
easat-11054	467	15	:	:	PUNCT
easat-11054	467	16	implementing	implement	VERB
easat-11054	467	17	comprehensive	comprehensive	ADJ
easat-11054	467	18	strategies	strategy	NOUN
easat-11054	467	19	to	to	PART
easat-11054	467	20	enhance	enhance	VERB
easat-11054	467	21	resilience	resilience	NOUN
easat-11054	467	22	and	and	CCONJ
easat-11054	467	23	reliability	reliability	NOUN
easat-11054	467	24	against	against	ADP
easat-11054	467	25	cyber	cyber	NOUN
easat-11054	467	26	-	-	PUNCT
easat-11054	467	27	attacks	attack	NOUN
easat-11054	467	28	,	,	PUNCT
easat-11054	467	29	"	"	PUNCT
easat-11054	467	30	results	result	NOUN
easat-11054	467	31	in	in	ADP
easat-11054	467	32	engineering	engineering	NOUN
easat-11054	467	33	,	,	PUNCT
easat-11054	467	34	vol	vol	NOUN
easat-11054	467	35	.	.	PROPN
easat-11054	467	36	23	23	NUM
easat-11054	467	37	,	,	PUNCT
easat-11054	467	38	p.	p.	NOUN
easat-11054	467	39	102647	102647	NUM
easat-11054	467	40	,	,	PUNCT
easat-11054	467	41	2024	2024	NUM
easat-11054	467	42	.	.	PUNCT
easat-11054	468	1	https://doi.org/10.1016/j.rineng.2024.102647	https://doi.org/10.1016/j.rineng.2024.102647	ADJ
easat-11054	468	2	[	[	X
easat-11054	468	3	4	4	NUM
easat-11054	468	4	]	]	PUNCT
easat-11054	468	5	a.	a.	NOUN
easat-11054	468	6	y.	y.	PROPN
easat-11054	468	7	abohatem	abohatem	PROPN
easat-11054	468	8	,	,	PUNCT
easat-11054	468	9	f.	f.	PROPN
easat-11054	468	10	m.	m.	PROPN
easat-11054	468	11	m.	m.	PROPN
easat-11054	468	12	ba	ba	PROPN
easat-11054	468	13	-	-	PUNCT
easat-11054	468	14	alwi	alwi	PROPN
easat-11054	468	15	,	,	PUNCT
easat-11054	468	16	and	and	CCONJ
easat-11054	468	17	a.	a.	NOUN
easat-11054	468	18	a.	a.	PROPN
easat-11054	468	19	al	al	PROPN
easat-11054	468	20	-	-	PUNCT
easat-11054	468	21	khulaidi	khulaidi	PROPN
easat-11054	468	22	,	,	PUNCT
easat-11054	468	23	"	"	PUNCT
easat-11054	468	24	suggestion	suggestion	NOUN
easat-11054	468	25	cybersecurity	cybersecurity	NOUN
easat-11054	468	26	framework	framework	NOUN
easat-11054	468	27	(	(	PUNCT
easat-11054	468	28	csf	csf	PROPN
easat-11054	468	29	)	)	PUNCT
easat-11054	468	30	for	for	ADP
easat-11054	468	31	reducing	reduce	VERB
easat-11054	468	32	cyber	cyber	NOUN
easat-11054	468	33	-	-	PUNCT
easat-11054	468	34	attacks	attack	NOUN
easat-11054	468	35	on	on	ADP
easat-11054	468	36	information	information	NOUN
easat-11054	468	37	systems	system	NOUN
easat-11054	468	38	,	,	PUNCT
easat-11054	468	39	"	"	PUNCT
easat-11054	468	40	sana’a	sana’a	NOUN
easat-11054	468	41	university	university	NOUN
easat-11054	468	42	journal	journal	NOUN
easat-11054	468	43	of	of	ADP
easat-11054	468	44	applied	apply	VERB
easat-11054	468	45	sciences	science	NOUN
easat-11054	468	46	and	and	CCONJ
easat-11054	468	47	technology	technology	NOUN
easat-11054	468	48	,	,	PUNCT
easat-11054	468	49	vol	vol	NOUN
easat-11054	468	50	.	.	PROPN
easat-11054	468	51	1	1	NUM
easat-11054	468	52	,	,	PUNCT
easat-11054	468	53	no	no	INTJ
easat-11054	468	54	.	.	NOUN
easat-11054	468	55	3	3	NUM
easat-11054	468	56	,	,	PUNCT
easat-11054	468	57	pp	pp	ADJ
easat-11054	468	58	.	.	PUNCT
easat-11054	469	1	234–252	234–252	NUM
easat-11054	469	2	,	,	PUNCT
easat-11054	469	3	2023	2023	NUM
easat-11054	469	4	.	.	PUNCT
easat-11054	470	1	https://doi.org/10.59628/jast.v1i3.248	https://doi.org/10.59628/jast.v1i3.248	PROPN
easat-11054	470	2	[	[	X
easat-11054	470	3	5	5	X
easat-11054	470	4	]	]	PUNCT
easat-11054	470	5	f.	f.	PROPN
easat-11054	470	6	thabit	thabit	PROPN
easat-11054	470	7	,	,	PUNCT
easat-11054	470	8	s.	s.	PROPN
easat-11054	470	9	alhomdy	alhomdy	PROPN
easat-11054	470	10	,	,	PUNCT
easat-11054	470	11	and	and	CCONJ
easat-11054	470	12	s.	s.	PROPN
easat-11054	470	13	jagtap	jagtap	PROPN
easat-11054	470	14	,	,	PUNCT
easat-11054	470	15	"	"	PUNCT
easat-11054	470	16	a	a	DET
easat-11054	470	17	new	new	ADJ
easat-11054	470	18	data	data	NOUN
easat-11054	470	19	security	security	NOUN
easat-11054	470	20	algorithm	algorithm	NOUN
easat-11054	470	21	for	for	ADP
easat-11054	470	22	the	the	DET
easat-11054	470	23	cloud	cloud	NOUN
easat-11054	470	24	computing	computing	NOUN
easat-11054	470	25	based	base	VERB
easat-11054	470	26	on	on	ADP
easat-11054	470	27	genetics	genetics	NOUN
easat-11054	470	28	techniques	technique	NOUN
easat-11054	470	29	and	and	CCONJ
easat-11054	470	30	logical	logical	ADJ
easat-11054	470	31	-	-	PUNCT
easat-11054	470	32	mathematical	mathematical	ADJ
easat-11054	470	33	functions	function	NOUN
easat-11054	470	34	,	,	PUNCT
easat-11054	470	35	"	"	PUNCT
easat-11054	470	36	international	international	ADJ
easat-11054	470	37	journal	journal	NOUN
easat-11054	470	38	of	of	ADP
easat-11054	470	39	intelligent	intelligent	ADJ
easat-11054	470	40	networks	network	NOUN
easat-11054	470	41	,	,	PUNCT
easat-11054	470	42	vol	vol	NOUN
easat-11054	470	43	.	.	PROPN
easat-11054	471	1	2	2	NUM
easat-11054	471	2	,	,	PUNCT
easat-11054	471	3	pp	pp	ADJ
easat-11054	471	4	.	.	PUNCT
easat-11054	472	1	18	18	NUM
easat-11054	472	2	-	-	SYM
easat-11054	472	3	33	33	NUM
easat-11054	472	4	,	,	PUNCT
easat-11054	472	5	2021	2021	NUM
easat-11054	472	6	.	.	PUNCT
easat-11054	473	1	https://doi.org/10.1016/j.ijin.2021.03.001	https://doi.org/10.1016/j.ijin.2021.03.001	VERB
easat-11054	473	2	[	[	X
easat-11054	473	3	6	6	NUM
easat-11054	473	4	]	]	PUNCT
easat-11054	473	5	h.	h.	PROPN
easat-11054	473	6	kheddar	kheddar	PROPN
easat-11054	473	7	,	,	PUNCT
easat-11054	473	8	d.	d.	PROPN
easat-11054	473	9	w.	w.	PROPN
easat-11054	473	10	dawoud	dawoud	PROPN
easat-11054	473	11	,	,	PUNCT
easat-11054	473	12	a.	a.	PROPN
easat-11054	473	13	i.	i.	PROPN
easat-11054	473	14	awad	awad	PROPN
easat-11054	473	15	,	,	PUNCT
easat-11054	473	16	y.	y.	PROPN
easat-11054	473	17	himeur	himeur	PROPN
easat-11054	473	18	,	,	PUNCT
easat-11054	473	19	and	and	CCONJ
easat-11054	473	20	m.	m.	PROPN
easat-11054	473	21	k.	k.	PROPN
easat-11054	473	22	khan	khan	PROPN
easat-11054	473	23	,	,	PUNCT
easat-11054	473	24	"	"	PUNCT
easat-11054	473	25	reinforcement	reinforcement	NOUN
easat-11054	473	26	-	-	PUNCT
easat-11054	473	27	learning	learning	NOUN
easat-11054	473	28	-	-	PUNCT
easat-11054	473	29	based	base	VERB
easat-11054	473	30	intrusion	intrusion	NOUN
easat-11054	473	31	detection	detection	NOUN
easat-11054	473	32	in	in	ADP
easat-11054	473	33	communication	communication	NOUN
easat-11054	473	34	networks	network	NOUN
easat-11054	473	35	:	:	PUNCT
easat-11054	473	36	a	a	DET
easat-11054	473	37	review	review	NOUN
easat-11054	473	38	,	,	PUNCT
easat-11054	473	39	"	"	PUNCT
easat-11054	473	40	ieee	ieee	NOUN
easat-11054	473	41	communications	communication	NOUN
easat-11054	473	42	surveys	survey	NOUN
easat-11054	473	43	&	&	CCONJ
easat-11054	473	44	tutorials	tutorial	NOUN
easat-11054	473	45	,	,	PUNCT
easat-11054	473	46	vol	vol	NOUN
easat-11054	473	47	.	.	PROPN
easat-11054	473	48	27	27	NUM
easat-11054	473	49	,	,	PUNCT
easat-11054	473	50	no	no	INTJ
easat-11054	473	51	.	.	NOUN
easat-11054	473	52	4	4	NUM
easat-11054	473	53	,	,	PUNCT
easat-11054	473	54	pp	pp	ADJ
easat-11054	473	55	.	.	PUNCT
easat-11054	474	1	2420	2420	NUM
easat-11054	474	2	2469	2469	NUM
easat-11054	474	3	,	,	PUNCT
easat-11054	474	4	2024	2024	NUM
easat-11054	474	5	.	.	PUNCT
easat-11054	475	1	https://doi.org/10.1109/comst.2024.3484491	https://doi.org/10.1109/comst.2024.3484491	PROPN
easat-11054	476	1	[	[	X
easat-11054	476	2	7	7	X
easat-11054	476	3	]	]	X
easat-11054	476	4	f.	f.	PROPN
easat-11054	476	5	zhang	zhang	PROPN
easat-11054	476	6	,	,	PUNCT
easat-11054	476	7	h.	h.	PROPN
easat-11054	476	8	huang	huang	PROPN
easat-11054	476	9	,	,	PUNCT
easat-11054	476	10	h.	h.	PROPN
easat-11054	476	11	lv	lv	PROPN
easat-11054	476	12	,	,	PUNCT
easat-11054	476	13	m.	m.	PROPN
easat-11054	476	14	jia	jia	PROPN
easat-11054	476	15	,	,	PUNCT
easat-11054	476	16	and	and	CCONJ
easat-11054	476	17	j.	j.	PROPN
easat-11054	476	18	liu	liu	PROPN
easat-11054	476	19	,	,	PUNCT
easat-11054	476	20	"	"	PUNCT
easat-11054	476	21	soft	soft	ADJ
easat-11054	476	22	actor	actor	NOUN
easat-11054	476	23	-	-	PUNCT
easat-11054	476	24	critic	critic	NOUN
easat-11054	476	25	based	base	VERB
easat-11054	476	26	anti	anti	ADJ
easat-11054	476	27	-	-	ADJ
easat-11054	476	28	attack	attack	ADJ
easat-11054	476	29	xss	xss	PROPN
easat-11054	476	30	detection	detection	PROPN
easat-11054	476	31	,	,	PUNCT
easat-11054	476	32	"	"	PUNCT
easat-11054	476	33	in	in	ADP
easat-11054	476	34	2024	2024	NUM
easat-11054	476	35	ieee	ieee	NOUN
easat-11054	476	36	24th	24th	ADJ
easat-11054	476	37	international	international	ADJ
easat-11054	476	38	conference	conference	NOUN
easat-11054	476	39	on	on	ADP
easat-11054	476	40	software	software	NOUN
easat-11054	476	41	quality	quality	NOUN
easat-11054	476	42	,	,	PUNCT
easat-11054	476	43	reliability	reliability	NOUN
easat-11054	476	44	,	,	PUNCT
easat-11054	476	45	and	and	CCONJ
easat-11054	476	46	security	security	NOUN
easat-11054	476	47	companion	companion	NOUN
easat-11054	476	48	(	(	PUNCT
easat-11054	476	49	qrs	qrs	PROPN
easat-11054	476	50	-	-	PUNCT
easat-11054	476	51	c	c	NOUN
easat-11054	476	52	)	)	PUNCT
easat-11054	476	53	(	(	PUNCT
easat-11054	476	54	pp	pp	ADJ
easat-11054	476	55	.	.	PUNCT
easat-11054	477	1	591	591	NUM
easat-11054	477	2	-	-	SYM
easat-11054	477	3	600	600	NUM
easat-11054	477	4	)	)	PUNCT
easat-11054	477	5	.	.	PUNCT
easat-11054	478	1	ieee	ieee	NOUN
easat-11054	478	2	,	,	PUNCT
easat-11054	478	3	2024	2024	NUM
easat-11054	478	4	.	.	PUNCT
easat-11054	479	1	[	[	X
easat-11054	479	2	8	8	NUM
easat-11054	479	3	]	]	X
easat-11054	479	4	n.	n.	NOUN
easat-11054	479	5	niknami	niknami	NOUN
easat-11054	479	6	and	and	CCONJ
easat-11054	479	7	j.	j.	PROPN
easat-11054	479	8	wu	wu	PROPN
easat-11054	479	9	,	,	PUNCT
easat-11054	479	10	"	"	PUNCT
easat-11054	479	11	deepidps	deepidps	NOUN
easat-11054	479	12	:	:	PUNCT
easat-11054	479	13	an	an	DET
easat-11054	479	14	adaptive	adaptive	ADJ
easat-11054	479	15	drl	drl	PROPN
easat-11054	479	16	-	-	PUNCT
easat-11054	479	17	based	base	VERB
easat-11054	479	18	intrusion	intrusion	NOUN
easat-11054	479	19	detection	detection	NOUN
easat-11054	479	20	and	and	CCONJ
easat-11054	479	21	prevention	prevention	NOUN
easat-11054	479	22	system	system	NOUN
easat-11054	479	23	for	for	ADP
easat-11054	479	24	sdn	sdn	NOUN
easat-11054	479	25	,	,	PUNCT
easat-11054	479	26	"	"	PUNCT
easat-11054	479	27	in	in	ADP
easat-11054	479	28	icc	icc	PROPN
easat-11054	479	29	2024	2024	NUM
easat-11054	479	30	-	-	PUNCT
easat-11054	479	31	ieee	ieee	NOUN
easat-11054	479	32	international	international	ADJ
easat-11054	479	33	conference	conference	NOUN
easat-11054	479	34	on	on	ADP
easat-11054	479	35	communications	communication	NOUN
easat-11054	479	36	(	(	PUNCT
easat-11054	479	37	pp	pp	ADJ
easat-11054	479	38	.	.	PUNCT
easat-11054	480	1	2040	2040	NUM
easat-11054	480	2	-	-	SYM
easat-11054	480	3	2046	2046	NUM
easat-11054	480	4	)	)	PUNCT
easat-11054	480	5	.	.	PUNCT
easat-11054	481	1	ieee	ieee	NOUN
easat-11054	481	2	,	,	PUNCT
easat-11054	481	3	2024	2024	NUM
easat-11054	481	4	.	.	PUNCT
easat-11054	482	1	[	[	X
easat-11054	482	2	9	9	NUM
easat-11054	482	3	]	]	PUNCT
easat-11054	482	4	j.	j.	PROPN
easat-11054	482	5	watts	watts	PROPN
easat-11054	482	6	,	,	PUNCT
easat-11054	482	7	f.	f.	PROPN
easat-11054	482	8	van	van	PROPN
easat-11054	482	9	wyk	wyk	PROPN
easat-11054	482	10	,	,	PUNCT
easat-11054	482	11	s.	s.	PROPN
easat-11054	482	12	rezaei	rezaei	PROPN
easat-11054	482	13	,	,	PUNCT
easat-11054	482	14	y.	y.	PROPN
easat-11054	482	15	wang	wang	PROPN
easat-11054	482	16	,	,	PUNCT
easat-11054	482	17	n.	n.	PROPN
easat-11054	482	18	masoud	masoud	PROPN
easat-11054	482	19	,	,	PUNCT
easat-11054	482	20	and	and	CCONJ
easat-11054	482	21	a.	a.	PROPN
easat-11054	482	22	khojandi	khojandi	PROPN
easat-11054	482	23	,	,	PUNCT
easat-11054	482	24	"	"	PUNCT
easat-11054	482	25	a	a	DET
easat-11054	482	26	dynamic	dynamic	ADJ
easat-11054	482	27	deep	deep	ADJ
easat-11054	482	28	reinforcement	reinforcement	NOUN
easat-11054	482	29	learningbayesian	learningbayesian	NOUN
easat-11054	482	30	framework	framework	NOUN
easat-11054	482	31	for	for	ADP
easat-11054	482	32	anomaly	anomaly	NOUN
easat-11054	482	33	detection	detection	NOUN
easat-11054	482	34	,	,	PUNCT
easat-11054	482	35	"	"	PUNCT
easat-11054	482	36	ieee	ieee	NOUN
easat-11054	482	37	transactions	transaction	NOUN
easat-11054	482	38	on	on	ADP
easat-11054	482	39	intelligent	intelligent	ADJ
easat-11054	482	40	transportation	transportation	NOUN
easat-11054	482	41	systems	system	NOUN
easat-11054	482	42	,	,	PUNCT
easat-11054	482	43	vol	vol	NOUN
easat-11054	482	44	.	.	PROPN
easat-11054	482	45	23	23	NUM
easat-11054	482	46	,	,	PUNCT
easat-11054	482	47	no	no	INTJ
easat-11054	482	48	.	.	NOUN
easat-11054	482	49	12	12	NUM
easat-11054	482	50	,	,	PUNCT
easat-11054	482	51	pp	pp	ADJ
easat-11054	482	52	.	.	PUNCT
easat-11054	483	1	22884	22884	NUM
easat-11054	483	2	-	-	SYM
easat-11054	483	3	22894	22894	NUM
easat-11054	483	4	,	,	PUNCT
easat-11054	483	5	2022	2022	NUM
easat-11054	483	6	.	.	PUNCT
easat-11054	484	1	https://doi.org/10.1109/tits.2022.3200906	https://doi.org/10.1109/tits.2022.3200906	PROPN
easat-11054	484	2	[	[	X
easat-11054	484	3	10	10	NUM
easat-11054	484	4	]	]	PUNCT
easat-11054	484	5	s.	s.	PROPN
easat-11054	484	6	mishra	mishra	PROPN
easat-11054	484	7	,	,	PUNCT
easat-11054	484	8	s.	s.	PROPN
easat-11054	484	9	chakraborty	chakraborty	PROPN
easat-11054	484	10	,	,	PUNCT
easat-11054	484	11	k.	k.	PROPN
easat-11054	484	12	s.	s.	PROPN
easat-11054	484	13	sahoo	sahoo	PROPN
easat-11054	484	14	,	,	PUNCT
easat-11054	484	15	and	and	CCONJ
easat-11054	484	16	m.	m.	NOUN
easat-11054	484	17	bilal	bilal	PROPN
easat-11054	484	18	,	,	PUNCT
easat-11054	484	19	"	"	PUNCT
easat-11054	484	20	cogni	cogni	PROPN
easat-11054	484	21	-	-	PUNCT
easat-11054	484	22	sec	sec	PROPN
easat-11054	484	23	:	:	PUNCT
easat-11054	484	24	a	a	DET
easat-11054	484	25	secure	secure	ADJ
easat-11054	484	26	cognitive	cognitive	NOUN
easat-11054	484	27	enabled	enable	VERB
easat-11054	484	28	distributed	distribute	VERB
easat-11054	484	29	reinforcement	reinforcement	NOUN
easat-11054	484	30	learning	learning	NOUN
easat-11054	484	31	model	model	NOUN
easat-11054	484	32	for	for	ADP
easat-11054	484	33	medical	medical	ADJ
easat-11054	484	34	cyber	cyber	PROPN
easat-11054	484	35	–	–	PUNCT
easat-11054	484	36	physical	physical	ADJ
easat-11054	484	37	system	system	NOUN
easat-11054	484	38	,	,	PUNCT
easat-11054	484	39	"	"	PUNCT
easat-11054	484	40	internet	internet	NOUN
easat-11054	484	41	of	of	ADP
easat-11054	484	42	things	thing	NOUN
easat-11054	484	43	,	,	PUNCT
easat-11054	484	44	vol	vol	NOUN
easat-11054	484	45	.	.	PROPN
easat-11054	484	46	24	24	NUM
easat-11054	484	47	,	,	PUNCT
easat-11054	484	48	p.	p.	NOUN
easat-11054	484	49	100978	100978	NUM
easat-11054	484	50	,	,	PUNCT
easat-11054	484	51	2023	2023	NUM
easat-11054	484	52	.	.	PUNCT
easat-11054	485	1	https://doi.org/10.1016/j.iot.2023.100978	https://doi.org/10.1016/j.iot.2023.100978	PUNCT
easat-11054	486	1	[	[	X
easat-11054	486	2	11	11	NUM
easat-11054	486	3	]	]	X
easat-11054	486	4	d.	d.	PROPN
easat-11054	486	5	lim	lim	PROPN
easat-11054	486	6	and	and	CCONJ
easat-11054	486	7	i.	i.	PROPN
easat-11054	486	8	joe	joe	PROPN
easat-11054	486	9	,	,	PUNCT
easat-11054	486	10	"	"	PUNCT
easat-11054	486	11	a	a	DET
easat-11054	486	12	drl	drl	PROPN
easat-11054	486	13	-	-	PUNCT
easat-11054	486	14	based	base	VERB
easat-11054	486	15	task	task	NOUN
easat-11054	486	16	offloading	offload	VERB
easat-11054	486	17	scheme	scheme	NOUN
easat-11054	486	18	for	for	ADP
easat-11054	486	19	server	server	NOUN
easat-11054	486	20	decision	decision	NOUN
easat-11054	486	21	-	-	PUNCT
easat-11054	486	22	making	making	NOUN
easat-11054	486	23	in	in	ADP
easat-11054	486	24	multi	multi	ADJ
easat-11054	486	25	-	-	ADJ
easat-11054	486	26	access	access	ADJ
easat-11054	486	27	edge	edge	NOUN
easat-11054	486	28	computing	computing	NOUN
easat-11054	486	29	,	,	PUNCT
easat-11054	486	30	"	"	PUNCT
easat-11054	486	31	electronics	electronic	NOUN
easat-11054	486	32	,	,	PUNCT
easat-11054	486	33	vol	vol	NOUN
easat-11054	486	34	.	.	PROPN
easat-11054	486	35	12	12	NUM
easat-11054	486	36	,	,	PUNCT
easat-11054	486	37	no	no	INTJ
easat-11054	486	38	.	.	NOUN
easat-11054	486	39	18	18	NUM
easat-11054	486	40	,	,	PUNCT
easat-11054	486	41	p.	p.	NOUN
easat-11054	486	42	3882	3882	NUM
easat-11054	486	43	,	,	PUNCT
easat-11054	486	44	2023	2023	NUM
easat-11054	486	45	.	.	PUNCT
easat-11054	487	1	https://doi.org/10.3390/electronics12183882	https://doi.org/10.3390/electronics12183882	NOUN
easat-11054	488	1	[	[	X
easat-11054	488	2	12	12	NUM
easat-11054	488	3	]	]	PUNCT
easat-11054	488	4	a.	a.	PROPN
easat-11054	488	5	s.	s.	PROPN
easat-11054	488	6	rajawat	rajawat	PROPN
easat-11054	488	7	,	,	PUNCT
easat-11054	488	8	s.	s.	PROPN
easat-11054	488	9	goyal	goyal	PROPN
easat-11054	488	10	,	,	PUNCT
easat-11054	488	11	c.	c.	PROPN
easat-11054	488	12	chauhan	chauhan	PROPN
easat-11054	488	13	,	,	PUNCT
easat-11054	488	14	p.	p.	PROPN
easat-11054	488	15	bedi	bedi	PROPN
easat-11054	488	16	,	,	PUNCT
easat-11054	488	17	m.	m.	NOUN
easat-11054	488	18	prasad	prasad	PROPN
easat-11054	488	19	,	,	PUNCT
easat-11054	488	20	and	and	CCONJ
easat-11054	488	21	t.	t.	PROPN
easat-11054	488	22	jan	jan	PROPN
easat-11054	488	23	,	,	PUNCT
easat-11054	488	24	"	"	PUNCT
easat-11054	488	25	cognitive	cognitive	ADJ
easat-11054	488	26	adaptive	adaptive	ADJ
easat-11054	488	27	systems	system	NOUN
easat-11054	488	28	for	for	ADP
easat-11054	488	29	industrial	industrial	ADJ
easat-11054	488	30	internet	internet	NOUN
easat-11054	488	31	of	of	ADP
easat-11054	488	32	things	thing	NOUN
easat-11054	488	33	using	use	VERB
easat-11054	488	34	reinforcement	reinforcement	NOUN
easat-11054	488	35	algorithm	algorithm	NOUN
easat-11054	488	36	,	,	PUNCT
easat-11054	488	37	"	"	PUNCT
easat-11054	488	38	electronics	electronic	NOUN
easat-11054	488	39	,	,	PUNCT
easat-11054	488	40	vol	vol	NOUN
easat-11054	488	41	.	.	PROPN
easat-11054	488	42	12	12	NUM
easat-11054	488	43	,	,	PUNCT
easat-11054	488	44	no	no	INTJ
easat-11054	488	45	.	.	NOUN
easat-11054	488	46	1	1	NUM
easat-11054	488	47	,	,	PUNCT
easat-11054	488	48	p.	p.	NOUN
easat-11054	488	49	217	217	NUM
easat-11054	488	50	,	,	PUNCT
easat-11054	488	51	2023	2023	NUM
easat-11054	488	52	.	.	PUNCT
easat-11054	489	1	https://doi.org/10.3390/electronics12010217	https://doi.org/10.3390/electronics12010217	PROPN
easat-11054	490	1	[	[	X
easat-11054	490	2	13	13	NUM
easat-11054	490	3	]	]	PUNCT
easat-11054	490	4	m.	m.	NOUN
easat-11054	490	5	h.	h.	PROPN
easat-11054	490	6	ali	ali	PROPN
easat-11054	490	7	et	et	PROPN
easat-11054	490	8	al	al	PROPN
easat-11054	490	9	.	.	PROPN
easat-11054	490	10	,	,	PUNCT
easat-11054	490	11	"	"	PUNCT
easat-11054	490	12	threat	threat	NOUN
easat-11054	490	13	analysis	analysis	NOUN
easat-11054	490	14	and	and	CCONJ
easat-11054	490	15	distributed	distribute	VERB
easat-11054	490	16	denial	denial	NOUN
easat-11054	490	17	of	of	ADP
easat-11054	490	18	service	service	NOUN
easat-11054	490	19	(	(	PUNCT
easat-11054	490	20	ddos	ddos	NOUN
easat-11054	490	21	)	)	PUNCT
easat-11054	490	22	attack	attack	NOUN
easat-11054	490	23	recognition	recognition	NOUN
easat-11054	490	24	in	in	ADP
easat-11054	490	25	the	the	DET
easat-11054	490	26	internet	internet	NOUN
easat-11054	490	27	of	of	ADP
easat-11054	490	28	things	thing	NOUN
easat-11054	490	29	(	(	PUNCT
easat-11054	490	30	iot	iot	NOUN
easat-11054	490	31	)	)	PUNCT
easat-11054	490	32	,	,	PUNCT
easat-11054	490	33	"	"	PUNCT
easat-11054	490	34	electronics	electronic	NOUN
easat-11054	490	35	,	,	PUNCT
easat-11054	490	36	vol	vol	NOUN
easat-11054	490	37	.	.	PROPN
easat-11054	490	38	11	11	NUM
easat-11054	490	39	,	,	PUNCT
easat-11054	490	40	no	no	INTJ
easat-11054	490	41	.	.	NOUN
easat-11054	490	42	3	3	NUM
easat-11054	490	43	,	,	PUNCT
easat-11054	490	44	p.	p.	NOUN
easat-11054	490	45	494	494	NUM
easat-11054	490	46	,	,	PUNCT
easat-11054	490	47	2022	2022	NUM
easat-11054	490	48	.	.	PUNCT
easat-11054	491	1	https://doi.org/10.3390/electronics11030494	https://doi.org/10.3390/electronics11030494	NOUN
easat-11054	491	2	[	[	X
easat-11054	491	3	14	14	NUM
easat-11054	491	4	]	]	PUNCT
easat-11054	491	5	a.	a.	NOUN
easat-11054	491	6	m.	m.	NOUN
easat-11054	491	7	alnajim	alnajim	PROPN
easat-11054	491	8	,	,	PUNCT
easat-11054	491	9	s.	s.	PROPN
easat-11054	491	10	habib	habib	PROPN
easat-11054	491	11	,	,	PUNCT
easat-11054	491	12	m.	m.	PROPN
easat-11054	491	13	islam	islam	PROPN
easat-11054	491	14	,	,	PUNCT
easat-11054	491	15	s.	s.	PROPN
easat-11054	491	16	m.	m.	PROPN
easat-11054	491	17	thwin	thwin	PROPN
easat-11054	491	18	,	,	PUNCT
easat-11054	491	19	and	and	CCONJ
easat-11054	491	20	f.	f.	PROPN
easat-11054	491	21	alotaibi	alotaibi	PROPN
easat-11054	491	22	,	,	PUNCT
easat-11054	491	23	"	"	PUNCT
easat-11054	491	24	a	a	DET
easat-11054	491	25	comprehensive	comprehensive	ADJ
easat-11054	491	26	survey	survey	NOUN
easat-11054	491	27	of	of	ADP
easat-11054	491	28	cybersecurity	cybersecurity	NOUN
easat-11054	491	29	threats	threat	NOUN
easat-11054	491	30	,	,	PUNCT
easat-11054	491	31	attacks	attack	NOUN
easat-11054	491	32	,	,	PUNCT
easat-11054	491	33	and	and	CCONJ
easat-11054	491	34	effective	effective	ADJ
easat-11054	491	35	countermeasures	countermeasure	NOUN
easat-11054	491	36	in	in	ADP
easat-11054	491	37	industrial	industrial	ADJ
easat-11054	491	38	internet	internet	NOUN
easat-11054	491	39	of	of	ADP
easat-11054	491	40	things	thing	NOUN
easat-11054	491	41	,	,	PUNCT
easat-11054	491	42	"	"	PUNCT
easat-11054	491	43	technologies	technology	NOUN
easat-11054	491	44	,	,	PUNCT
easat-11054	491	45	vol	vol	NOUN
easat-11054	491	46	.	.	PROPN
easat-11054	491	47	11	11	NUM
easat-11054	491	48	,	,	PUNCT
easat-11054	491	49	no	no	INTJ
easat-11054	491	50	.	.	NOUN
easat-11054	491	51	6	6	NUM
easat-11054	491	52	,	,	PUNCT
easat-11054	491	53	p.	p.	NOUN
easat-11054	491	54	161	161	NUM
easat-11054	491	55	,	,	PUNCT
easat-11054	491	56	2023	2023	NUM
easat-11054	491	57	.	.	PUNCT
easat-11054	492	1	https://doi.org/10.3390/technologies11060161	https://doi.org/10.3390/technologies11060161	ADJ
easat-11054	492	2	[	[	PUNCT
easat-11054	492	3	15	15	NUM
easat-11054	492	4	]	]	X
easat-11054	492	5	m.	m.	NOUN
easat-11054	492	6	nuaimi	nuaimi	PROPN
easat-11054	492	7	,	,	PUNCT
easat-11054	492	8	l.	l.	PROPN
easat-11054	492	9	c.	c.	PROPN
easat-11054	492	10	fourati	fourati	PROPN
easat-11054	492	11	,	,	PUNCT
easat-11054	492	12	and	and	CCONJ
easat-11054	492	13	b.	b.	PROPN
easat-11054	492	14	b.	b.	PROPN
easat-11054	493	1	hamed	hamed	PROPN
easat-11054	493	2	,	,	PUNCT
easat-11054	493	3	"	"	PUNCT
easat-11054	493	4	intelligent	intelligent	ADJ
easat-11054	493	5	approaches	approach	NOUN
easat-11054	493	6	toward	toward	ADP
easat-11054	493	7	intrusion	intrusion	NOUN
easat-11054	493	8	detection	detection	NOUN
easat-11054	493	9	systems	system	NOUN
easat-11054	493	10	for	for	ADP
easat-11054	493	11	industrial	industrial	ADJ
easat-11054	493	12	internet	internet	NOUN
easat-11054	493	13	of	of	ADP
easat-11054	493	14	things	thing	NOUN
easat-11054	493	15	:	:	PUNCT
easat-11054	493	16	a	a	DET
easat-11054	493	17	systematic	systematic	ADJ
easat-11054	493	18	comprehensive	comprehensive	ADJ
easat-11054	493	19	review	review	NOUN
easat-11054	493	20	,	,	PUNCT
easat-11054	493	21	"	"	PUNCT
easat-11054	493	22	journal	journal	NOUN
easat-11054	493	23	of	of	ADP
easat-11054	493	24	network	network	NOUN
easat-11054	493	25	and	and	CCONJ
easat-11054	493	26	computer	computer	NOUN
easat-11054	493	27	applications	application	NOUN
easat-11054	493	28	,	,	PUNCT
easat-11054	493	29	vol	vol	NOUN
easat-11054	493	30	.	.	PROPN
easat-11054	493	31	215	215	NUM
easat-11054	493	32	,	,	PUNCT
easat-11054	493	33	p.	p.	NOUN
easat-11054	493	34	103637	103637	NUM
easat-11054	493	35	,	,	PUNCT
easat-11054	493	36	2023	2023	NUM
easat-11054	493	37	.	.	PUNCT
easat-11054	494	1	https://doi.org/10.1016/j.jnca.2023.103637	https://doi.org/10.1016/j.jnca.2023.103637	PROPN
easat-11054	494	2	https://creativecommons.org/licenses/by/4.0/	https://creativecommons.org/licenses/by/4.0/	PROPN
easat-11054	494	3	https://doi.org/10.34028/iajit/19/5/14	https://doi.org/10.34028/iajit/19/5/14	VERB
easat-11054	494	4	https://doi.org/10.59628/jast.v2i5.1128	https://doi.org/10.59628/jast.v2i5.1128	NOUN
easat-11054	495	1	https://doi.org/10.1016/j.rineng.2024.102647	https://doi.org/10.1016/j.rineng.2024.102647	ADJ
easat-11054	495	2	https://doi.org/10.59628/jast.v1i3.248	https://doi.org/10.59628/jast.v1i3.248	NOUN
easat-11054	495	3	https://doi.org/10.1016/j.ijin.2021.03.001	https://doi.org/10.1016/j.ijin.2021.03.001	VERB
easat-11054	495	4	https://doi.org/10.1109/comst.2024.3484491	https://doi.org/10.1109/comst.2024.3484491	PROPN
easat-11054	495	5	https://doi.org/10.1109/tits.2022.3200906	https://doi.org/10.1109/tits.2022.3200906	PROPN
easat-11054	495	6	https://doi.org/10.1016/j.iot.2023.100978	https://doi.org/10.1016/j.iot.2023.100978	PROPN
easat-11054	495	7	https://doi.org/10.3390/electronics12183882	https://doi.org/10.3390/electronics12183882	NOUN
easat-11054	495	8	https://doi.org/10.3390/electronics12010217	https://doi.org/10.3390/electronics12010217	PROPN
easat-11054	495	9	https://doi.org/10.3390/electronics11030494	https://doi.org/10.3390/electronics11030494	PROPN
easat-11054	495	10	https://doi.org/10.3390/technologies11060161	https://doi.org/10.3390/technologies11060161	ADJ
easat-11054	495	11	https://doi.org/10.1016/j.jnca.2023.103637	https://doi.org/10.1016/j.jnca.2023.103637	PROPN
easat-11054	495	12	1075	1075	NUM
easat-11054	495	13	edelweiss	edelweiss	PROPN
easat-11054	495	14	applied	apply	VERB
easat-11054	495	15	science	science	NOUN
easat-11054	495	16	and	and	CCONJ
easat-11054	495	17	technology	technology	NOUN
easat-11054	495	18	issn	issn	PROPN
easat-11054	495	19	:	:	PUNCT
easat-11054	495	20	2576	2576	NUM
easat-11054	495	21	-	-	SYM
easat-11054	495	22	8484	8484	NUM
easat-11054	495	23	vol	vol	NOUN
easat-11054	495	24	.	.	PROPN
easat-11054	496	1	9	9	NUM
easat-11054	496	2	,	,	PUNCT
easat-11054	496	3	no	no	INTJ
easat-11054	496	4	.	.	NOUN
easat-11054	496	5	11	11	NUM
easat-11054	496	6	:	:	PUNCT
easat-11054	496	7	1053	1053	NUM
easat-11054	496	8	-	-	SYM
easat-11054	496	9	1075	1075	NUM
easat-11054	496	10	,	,	PUNCT
easat-11054	496	11	2025	2025	NUM
easat-11054	496	12	doi	doi	NOUN
easat-11054	496	13	:	:	PUNCT
easat-11054	496	14	10.55214/2576	10.55214/2576	NUM
easat-11054	496	15	-	-	SYM
easat-11054	496	16	8484.v9i11.11054	8484.v9i11.11054	NOUN
easat-11054	497	1	©	©	PROPN
easat-11054	497	2	2025	2025	NUM
easat-11054	497	3	by	by	ADP
easat-11054	497	4	the	the	DET
easat-11054	497	5	authors	author	NOUN
easat-11054	497	6	;	;	PUNCT
easat-11054	497	7	licensee	licensee	PROPN
easat-11054	497	8	learning	learning	NOUN
easat-11054	497	9	gate	gate	NOUN
easat-11054	498	1	[	[	X
easat-11054	498	2	16	16	NUM
easat-11054	498	3	]	]	X
easat-11054	498	4	e.	e.	PROPN
easat-11054	498	5	muhati	muhati	PROPN
easat-11054	498	6	and	and	CCONJ
easat-11054	498	7	d.	d.	PROPN
easat-11054	498	8	b.	b.	PROPN
easat-11054	498	9	rawat	rawat	PROPN
easat-11054	498	10	,	,	PUNCT
easat-11054	498	11	"	"	PUNCT
easat-11054	498	12	asynchronous	asynchronous	ADJ
easat-11054	498	13	advantage	advantage	NOUN
easat-11054	498	14	actor	actor	NOUN
easat-11054	498	15	-	-	PUNCT
easat-11054	498	16	critic	critic	NOUN
easat-11054	498	17	(	(	PUNCT
easat-11054	498	18	a3c	a3c	NUM
easat-11054	498	19	)	)	PUNCT
easat-11054	498	20	learning	learn	VERB
easat-11054	498	21	for	for	ADP
easat-11054	498	22	cognitive	cognitive	ADJ
easat-11054	498	23	network	network	NOUN
easat-11054	498	24	security	security	NOUN
easat-11054	498	25	,	,	PUNCT
easat-11054	498	26	"	"	PUNCT
easat-11054	498	27	in	in	ADP
easat-11054	498	28	2021	2021	NUM
easat-11054	498	29	third	third	ADJ
easat-11054	498	30	ieee	ieee	NOUN
easat-11054	498	31	international	international	ADJ
easat-11054	498	32	conference	conference	NOUN
easat-11054	498	33	on	on	ADP
easat-11054	498	34	trust	trust	PROPN
easat-11054	498	35	,	,	PUNCT
easat-11054	498	36	privacy	privacy	NOUN
easat-11054	498	37	and	and	CCONJ
easat-11054	498	38	security	security	NOUN
easat-11054	498	39	in	in	ADP
easat-11054	498	40	intelligent	intelligent	ADJ
easat-11054	498	41	systems	system	NOUN
easat-11054	498	42	and	and	CCONJ
easat-11054	498	43	applications	application	NOUN
easat-11054	498	44	(	(	PUNCT
easat-11054	498	45	tps	tps	NOUN
easat-11054	498	46	-	-	PUNCT
easat-11054	498	47	isa	isa	NOUN
easat-11054	498	48	)	)	PUNCT
easat-11054	498	49	(	(	PUNCT
easat-11054	498	50	pp	pp	INTJ
easat-11054	498	51	.	.	PUNCT
easat-11054	499	1	106	106	NUM
easat-11054	499	2	-	-	SYM
easat-11054	499	3	113	113	NUM
easat-11054	499	4	)	)	PUNCT
easat-11054	499	5	.	.	PUNCT
easat-11054	500	1	ieee	ieee	NOUN
easat-11054	500	2	,	,	PUNCT
easat-11054	500	3	2021	2021	NUM
easat-11054	500	4	.	.	PUNCT
easat-11054	501	1	[	[	X
easat-11054	501	2	17	17	NUM
easat-11054	501	3	]	]	PUNCT
easat-11054	501	4	k.	k.	PROPN
easat-11054	501	5	zhou	zhou	PROPN
easat-11054	501	6	,	,	PUNCT
easat-11054	501	7	w.	w.	PROPN
easat-11054	501	8	wang	wang	PROPN
easat-11054	501	9	,	,	PUNCT
easat-11054	501	10	t.	t.	PROPN
easat-11054	501	11	hu	hu	PROPN
easat-11054	501	12	,	,	PUNCT
easat-11054	501	13	and	and	CCONJ
easat-11054	501	14	k.	k.	PROPN
easat-11054	501	15	deng	deng	PROPN
easat-11054	501	16	,	,	PUNCT
easat-11054	501	17	"	"	PUNCT
easat-11054	501	18	application	application	NOUN
easat-11054	501	19	of	of	ADP
easat-11054	501	20	improved	improved	ADJ
easat-11054	501	21	asynchronous	asynchronous	ADJ
easat-11054	501	22	advantage	advantage	NOUN
easat-11054	501	23	actor	actor	NOUN
easat-11054	501	24	critic	critic	PROPN
easat-11054	501	25	reinforcement	reinforcement	NOUN
easat-11054	501	26	learning	learning	NOUN
easat-11054	501	27	model	model	NOUN
easat-11054	501	28	on	on	ADP
easat-11054	501	29	anomaly	anomaly	PROPN
easat-11054	501	30	detection	detection	NOUN
easat-11054	501	31	,	,	PUNCT
easat-11054	501	32	"	"	PUNCT
easat-11054	501	33	entropy	entropy	NOUN
easat-11054	501	34	,	,	PUNCT
easat-11054	501	35	vol	vol	NOUN
easat-11054	501	36	.	.	PROPN
easat-11054	501	37	23	23	NUM
easat-11054	501	38	,	,	PUNCT
easat-11054	501	39	no	no	INTJ
easat-11054	501	40	.	.	NOUN
easat-11054	501	41	3	3	NUM
easat-11054	501	42	,	,	PUNCT
easat-11054	501	43	p.	p.	NOUN
easat-11054	501	44	274	274	NUM
easat-11054	501	45	,	,	PUNCT
easat-11054	501	46	2021	2021	NUM
easat-11054	501	47	.	.	PUNCT
easat-11054	502	1	https://doi.org/10.3390/e23030274	https://doi.org/10.3390/e23030274	NUM
easat-11054	502	2	[	[	SYM
easat-11054	502	3	18	18	NUM
easat-11054	502	4	]	]	PUNCT
easat-11054	502	5	s.	s.	PROPN
easat-11054	502	6	shen	shen	PROPN
easat-11054	502	7	,	,	PUNCT
easat-11054	502	8	c.	c.	PROPN
easat-11054	502	9	cai	cai	PROPN
easat-11054	502	10	,	,	PUNCT
easat-11054	502	11	y.	y.	PROPN
easat-11054	502	12	shen	shen	PROPN
easat-11054	502	13	,	,	PUNCT
easat-11054	502	14	x.	x.	PROPN
easat-11054	502	15	wu	wu	PROPN
easat-11054	502	16	,	,	PUNCT
easat-11054	502	17	w.	w.	PROPN
easat-11054	502	18	ke	ke	PROPN
easat-11054	502	19	,	,	PUNCT
easat-11054	502	20	and	and	CCONJ
easat-11054	502	21	s.	s.	PROPN
easat-11054	502	22	yu	yu	PROPN
easat-11054	502	23	,	,	PUNCT
easat-11054	502	24	"	"	PUNCT
easat-11054	502	25	joint	joint	ADJ
easat-11054	502	26	mean	mean	ADJ
easat-11054	502	27	-	-	PUNCT
easat-11054	502	28	field	field	NOUN
easat-11054	502	29	game	game	NOUN
easat-11054	502	30	and	and	CCONJ
easat-11054	502	31	multiagent	multiagent	NOUN
easat-11054	502	32	asynchronous	asynchronous	ADJ
easat-11054	502	33	advantage	advantage	NOUN
easat-11054	502	34	actor	actor	NOUN
easat-11054	502	35	-	-	PUNCT
easat-11054	502	36	critic	critic	NOUN
easat-11054	502	37	for	for	ADP
easat-11054	502	38	edge	edge	NOUN
easat-11054	502	39	intelligence	intelligence	NOUN
easat-11054	502	40	-	-	PUNCT
easat-11054	502	41	based	base	VERB
easat-11054	502	42	iot	iot	PROPN
easat-11054	502	43	malware	malware	NOUN
easat-11054	502	44	propagation	propagation	NOUN
easat-11054	502	45	defense	defense	NOUN
easat-11054	502	46	,	,	PUNCT
easat-11054	502	47	"	"	PUNCT
easat-11054	502	48	ieee	ieee	NOUN
easat-11054	502	49	transactions	transaction	NOUN
easat-11054	502	50	on	on	ADP
easat-11054	502	51	dependable	dependable	ADJ
easat-11054	502	52	and	and	CCONJ
easat-11054	502	53	secure	secure	ADJ
easat-11054	502	54	computing	computing	NOUN
easat-11054	502	55	,	,	PUNCT
easat-11054	502	56	vol	vol	NOUN
easat-11054	502	57	.	.	PROPN
easat-11054	502	58	22	22	NUM
easat-11054	502	59	,	,	PUNCT
easat-11054	502	60	no	no	INTJ
easat-11054	502	61	.	.	NOUN
easat-11054	502	62	4	4	NUM
easat-11054	502	63	,	,	PUNCT
easat-11054	502	64	pp	pp	ADJ
easat-11054	502	65	.	.	PUNCT
easat-11054	503	1	3824	3824	NUM
easat-11054	503	2	3838	3838	NUM
easat-11054	503	3	,	,	PUNCT
easat-11054	503	4	2025	2025	NUM
easat-11054	503	5	.	.	PUNCT
easat-11054	504	1	https://doi.org/10.1109/tdsc.2025.3542104	https://doi.org/10.1109/tdsc.2025.3542104	PUNCT
easat-11054	505	1	[	[	X
easat-11054	505	2	19	19	NUM
easat-11054	505	3	]	]	PUNCT
easat-11054	505	4	s.	s.	PROPN
easat-11054	505	5	dong	dong	PROPN
easat-11054	505	6	,	,	PUNCT
easat-11054	505	7	y.	y.	PROPN
easat-11054	505	8	xia	xia	PROPN
easat-11054	505	9	,	,	PUNCT
easat-11054	505	10	and	and	CCONJ
easat-11054	505	11	t.	t.	PROPN
easat-11054	505	12	wang	wang	PROPN
easat-11054	505	13	,	,	PUNCT
easat-11054	505	14	"	"	PUNCT
easat-11054	505	15	network	network	NOUN
easat-11054	505	16	abnormal	abnormal	ADJ
easat-11054	505	17	traffic	traffic	NOUN
easat-11054	505	18	detection	detection	NOUN
easat-11054	505	19	framework	framework	NOUN
easat-11054	505	20	based	base	VERB
easat-11054	505	21	on	on	ADP
easat-11054	505	22	deep	deep	ADJ
easat-11054	505	23	reinforcement	reinforcement	NOUN
easat-11054	505	24	learning	learning	NOUN
easat-11054	505	25	,	,	PUNCT
easat-11054	505	26	"	"	PUNCT
easat-11054	505	27	ieee	ieee	PROPN
easat-11054	505	28	wireless	wireless	NOUN
easat-11054	505	29	communications	communication	NOUN
easat-11054	505	30	,	,	PUNCT
easat-11054	505	31	vol	vol	NOUN
easat-11054	505	32	.	.	PROPN
easat-11054	505	33	31	31	NUM
easat-11054	505	34	,	,	PUNCT
easat-11054	505	35	no	no	INTJ
easat-11054	505	36	.	.	NOUN
easat-11054	505	37	3	3	NUM
easat-11054	505	38	,	,	PUNCT
easat-11054	505	39	pp	pp	ADJ
easat-11054	505	40	.	.	PUNCT
easat-11054	506	1	185	185	NUM
easat-11054	506	2	-	-	SYM
easat-11054	506	3	193	193	NUM
easat-11054	506	4	,	,	PUNCT
easat-11054	506	5	2024	2024	NUM
easat-11054	506	6	.	.	PUNCT
easat-11054	507	1	https://doi.org/10.1109/mwc.011.2200320	https://doi.org/10.1109/mwc.011.2200320	NOUN
easat-11054	507	2	[	[	X
easat-11054	507	3	20	20	NUM
easat-11054	507	4	]	]	PUNCT
easat-11054	507	5	s.	s.	PROPN
easat-11054	507	6	anbazhagan	anbazhagan	VERB
easat-11054	507	7	and	and	CCONJ
easat-11054	507	8	r.	r.	PROPN
easat-11054	507	9	mugelan	mugelan	PROPN
easat-11054	507	10	,	,	PUNCT
easat-11054	507	11	"	"	PUNCT
easat-11054	507	12	next	next	ADJ
easat-11054	507	13	-	-	PUNCT
easat-11054	507	14	gen	gen	NOUN
easat-11054	507	15	resource	resource	NOUN
easat-11054	507	16	optimization	optimization	NOUN
easat-11054	507	17	in	in	ADP
easat-11054	507	18	nb	nb	PROPN
easat-11054	507	19	-	-	PUNCT
easat-11054	507	20	iot	iot	NOUN
easat-11054	507	21	networks	network	NOUN
easat-11054	507	22	:	:	PUNCT
easat-11054	507	23	harnessing	harness	VERB
easat-11054	507	24	soft	soft	ADJ
easat-11054	507	25	actor	actor	NOUN
easat-11054	507	26	–	–	PUNCT
easat-11054	507	27	critic	critic	NOUN
easat-11054	507	28	reinforcement	reinforcement	NOUN
easat-11054	507	29	learning	learning	NOUN
easat-11054	507	30	,	,	PUNCT
easat-11054	507	31	"	"	PUNCT
easat-11054	507	32	computer	computer	NOUN
easat-11054	507	33	networks	network	NOUN
easat-11054	507	34	,	,	PUNCT
easat-11054	507	35	vol	vol	NOUN
easat-11054	507	36	.	.	PROPN
easat-11054	507	37	252	252	NUM
easat-11054	507	38	,	,	PUNCT
easat-11054	507	39	p.	p.	NOUN
easat-11054	507	40	110670	110670	NUM
easat-11054	507	41	,	,	PUNCT
easat-11054	507	42	2024	2024	NUM
easat-11054	507	43	.	.	PUNCT
easat-11054	508	1	https://doi.org/10.1016/j.comnet.2024.110670	https://doi.org/10.1016/j.comnet.2024.110670	PROPN
easat-11054	508	2	[	[	X
easat-11054	508	3	21	21	NUM
easat-11054	508	4	]	]	PUNCT
easat-11054	508	5	a.	a.	NOUN
easat-11054	508	6	m.	m.	NOUN
easat-11054	508	7	khako	khako	PROPN
easat-11054	508	8	and	and	CCONJ
easat-11054	508	9	s.	s.	PROPN
easat-11054	508	10	a.	a.	PROPN
easat-11054	508	11	alhomdy	alhomdy	PROPN
easat-11054	508	12	,	,	PUNCT
easat-11054	508	13	"	"	PUNCT
easat-11054	508	14	enhanced	enhance	VERB
easat-11054	508	15	a3c	a3c	VERB
easat-11054	508	16	with	with	ADP
easat-11054	508	17	modified	modified	ADJ
easat-11054	508	18	cnn	cnn	PROPN
easat-11054	508	19	for	for	ADP
easat-11054	508	20	intrusion	intrusion	NOUN
easat-11054	508	21	detection	detection	NOUN
easat-11054	508	22	in	in	ADP
easat-11054	508	23	iiot	iiot	ADJ
easat-11054	508	24	environments	environment	NOUN
easat-11054	508	25	,	,	PUNCT
easat-11054	508	26	"	"	PUNCT
easat-11054	508	27	journal	journal	NOUN
easat-11054	508	28	of	of	ADP
easat-11054	508	29	cybersecurity	cybersecurity	NOUN
easat-11054	508	30	and	and	CCONJ
easat-11054	508	31	network	network	NOUN
easat-11054	508	32	protection	protection	NOUN
easat-11054	508	33	,	,	PUNCT
easat-11054	508	34	vol	vol	NOUN
easat-11054	508	35	.	.	PROPN
easat-11054	508	36	5	5	NUM
easat-11054	508	37	,	,	PUNCT
easat-11054	508	38	no	no	INTJ
easat-11054	508	39	.	.	NOUN
easat-11054	508	40	2	2	NUM
easat-11054	508	41	,	,	PUNCT
easat-11054	508	42	pp	pp	ADJ
easat-11054	508	43	.	.	PUNCT
easat-11054	509	1	112	112	NUM
easat-11054	509	2	-	-	SYM
easat-11054	509	3	125	125	NUM
easat-11054	509	4	,	,	PUNCT
easat-11054	509	5	2023	2023	NUM
easat-11054	509	6	.	.	PUNCT
easat-11054	510	1	[	[	X
easat-11054	510	2	22	22	NUM
easat-11054	510	3	]	]	PUNCT
easat-11054	510	4	z.	z.	PROPN
easat-11054	510	5	hu	hu	PROPN
easat-11054	510	6	et	et	PROPN
easat-11054	510	7	al	al	PROPN
easat-11054	510	8	.	.	PROPN
easat-11054	510	9	,	,	PUNCT
easat-11054	510	10	"	"	PUNCT
easat-11054	510	11	a	a	DET
easat-11054	510	12	hybrid	hybrid	ADJ
easat-11054	510	13	data	data	NOUN
easat-11054	510	14	-	-	PUNCT
easat-11054	510	15	driven	drive	VERB
easat-11054	510	16	approach	approach	NOUN
easat-11054	510	17	integrating	integrate	VERB
easat-11054	510	18	temporal	temporal	ADJ
easat-11054	510	19	fusion	fusion	NOUN
easat-11054	510	20	transformer	transformer	NOUN
easat-11054	510	21	and	and	CCONJ
easat-11054	510	22	soft	soft	ADJ
easat-11054	510	23	actor	actor	NOUN
easat-11054	510	24	-	-	PUNCT
easat-11054	510	25	critic	critic	NOUN
easat-11054	510	26	algorithm	algorithm	NOUN
easat-11054	510	27	for	for	ADP
easat-11054	510	28	optimal	optimal	ADJ
easat-11054	510	29	scheduling	scheduling	NOUN
easat-11054	510	30	of	of	ADP
easat-11054	510	31	building	build	VERB
easat-11054	510	32	integrated	integrate	VERB
easat-11054	510	33	energy	energy	NOUN
easat-11054	510	34	systems	system	NOUN
easat-11054	510	35	,	,	PUNCT
easat-11054	510	36	"	"	PUNCT
easat-11054	510	37	journal	journal	NOUN
easat-11054	510	38	of	of	ADP
easat-11054	510	39	modern	modern	ADJ
easat-11054	510	40	power	power	NOUN
easat-11054	510	41	systems	system	NOUN
easat-11054	510	42	and	and	CCONJ
easat-11054	510	43	clean	clean	ADJ
easat-11054	510	44	energy	energy	NOUN
easat-11054	510	45	,	,	PUNCT
easat-11054	510	46	vol	vol	NOUN
easat-11054	510	47	.	.	PROPN
easat-11054	510	48	13	13	NUM
easat-11054	510	49	,	,	PUNCT
easat-11054	510	50	no	no	INTJ
easat-11054	510	51	.	.	NOUN
easat-11054	510	52	3	3	NUM
easat-11054	510	53	,	,	PUNCT
easat-11054	510	54	pp	pp	ADJ
easat-11054	510	55	.	.	PUNCT
easat-11054	510	56	878	878	NUM
easat-11054	510	57	891	891	NUM
easat-11054	510	58	,	,	PUNCT
easat-11054	510	59	2025	2025	NUM
easat-11054	510	60	.	.	PUNCT
easat-11054	511	1	https://doi.org/10.35833/mpce.2024.000909	https://doi.org/10.35833/mpce.2024.000909	ADJ
easat-11054	511	2	[	[	X
easat-11054	511	3	23	23	NUM
easat-11054	511	4	]	]	PUNCT
easat-11054	511	5	h.	h.	PROPN
easat-11054	511	6	kumar	kumar	PROPN
easat-11054	511	7	,	,	PUNCT
easat-11054	511	8	a.	a.	NOUN
easat-11054	511	9	koppel	koppel	PROPN
easat-11054	511	10	,	,	PUNCT
easat-11054	511	11	and	and	CCONJ
easat-11054	511	12	a.	a.	PROPN
easat-11054	511	13	ribeiro	ribeiro	PROPN
easat-11054	511	14	,	,	PUNCT
easat-11054	511	15	"	"	PUNCT
easat-11054	511	16	on	on	ADP
easat-11054	511	17	the	the	DET
easat-11054	511	18	sample	sample	NOUN
easat-11054	511	19	complexity	complexity	NOUN
easat-11054	511	20	of	of	ADP
easat-11054	511	21	actor	actor	NOUN
easat-11054	511	22	-	-	PUNCT
easat-11054	511	23	critic	critic	NOUN
easat-11054	511	24	method	method	NOUN
easat-11054	511	25	for	for	ADP
easat-11054	511	26	reinforcement	reinforcement	NOUN
easat-11054	511	27	learning	learning	NOUN
easat-11054	511	28	with	with	ADP
easat-11054	511	29	function	function	NOUN
easat-11054	511	30	approximation	approximation	NOUN
easat-11054	511	31	,	,	PUNCT
easat-11054	511	32	"	"	PUNCT
easat-11054	511	33	machine	machine	NOUN
easat-11054	511	34	learning	learning	NOUN
easat-11054	511	35	,	,	PUNCT
easat-11054	511	36	vol	vol	NOUN
easat-11054	511	37	.	.	PROPN
easat-11054	512	1	112	112	NUM
easat-11054	512	2	,	,	PUNCT
easat-11054	512	3	pp	pp	ADJ
easat-11054	512	4	.	.	PUNCT
easat-11054	513	1	2433	2433	NUM
easat-11054	513	2	-	-	SYM
easat-11054	513	3	2467	2467	NUM
easat-11054	513	4	,	,	PUNCT
easat-11054	513	5	2023	2023	NUM
easat-11054	513	6	.	.	PUNCT
easat-11054	514	1	https://doi.org/10.1007/s10994023-06303-2	https://doi.org/10.1007/s10994023-06303-2	PROPN
easat-11054	514	2	[	[	X
easat-11054	514	3	24	24	NUM
easat-11054	514	4	]	]	PUNCT
easat-11054	514	5	j.	j.	PROPN
easat-11054	514	6	zheng	zheng	PROPN
easat-11054	514	7	,	,	PUNCT
easat-11054	514	8	m.	m.	PROPN
easat-11054	514	9	n.	n.	PROPN
easat-11054	514	10	kurt	kurt	PROPN
easat-11054	514	11	,	,	PUNCT
easat-11054	514	12	and	and	CCONJ
easat-11054	514	13	x.	x.	NOUN
easat-11054	514	14	wang	wang	PROPN
easat-11054	514	15	,	,	PUNCT
easat-11054	514	16	"	"	PUNCT
easat-11054	514	17	stochastic	stochastic	ADJ
easat-11054	514	18	integrated	integrate	VERB
easat-11054	514	19	actor	actor	NOUN
easat-11054	514	20	–	–	PUNCT
easat-11054	514	21	critic	critic	NOUN
easat-11054	514	22	for	for	ADP
easat-11054	514	23	deep	deep	ADJ
easat-11054	514	24	reinforcement	reinforcement	NOUN
easat-11054	514	25	learning	learning	NOUN
easat-11054	514	26	,	,	PUNCT
easat-11054	514	27	"	"	PUNCT
easat-11054	514	28	ieee	ieee	NOUN
easat-11054	514	29	transactions	transaction	NOUN
easat-11054	514	30	on	on	ADP
easat-11054	514	31	neural	neural	ADJ
easat-11054	514	32	networks	network	NOUN
easat-11054	514	33	and	and	CCONJ
easat-11054	514	34	learning	learning	NOUN
easat-11054	514	35	systems	system	NOUN
easat-11054	514	36	,	,	PUNCT
easat-11054	514	37	vol	vol	NOUN
easat-11054	514	38	.	.	PROPN
easat-11054	514	39	35	35	NUM
easat-11054	514	40	,	,	PUNCT
easat-11054	514	41	no	no	INTJ
easat-11054	514	42	.	.	NOUN
easat-11054	514	43	5	5	NUM
easat-11054	514	44	,	,	PUNCT
easat-11054	514	45	pp	pp	ADJ
easat-11054	514	46	.	.	PUNCT
easat-11054	515	1	6654	6654	NUM
easat-11054	515	2	-	-	SYM
easat-11054	515	3	6666	6666	NUM
easat-11054	515	4	,	,	PUNCT
easat-11054	515	5	2022	2022	NUM
easat-11054	515	6	.	.	PUNCT
easat-11054	516	1	https://doi.org/10.1109/tnnls.2022.3212273	https://doi.org/10.1109/tnnls.2022.3212273	PROPN
easat-11054	516	2	https://doi.org/10.3390/e23030274	https://doi.org/10.3390/e23030274	NUM
easat-11054	516	3	https://doi.org/10.1109/tdsc.2025.3542104	https://doi.org/10.1109/tdsc.2025.3542104	PROPN
easat-11054	516	4	https://doi.org/10.1109/mwc.011.2200320	https://doi.org/10.1109/mwc.011.2200320	NOUN
easat-11054	516	5	https://doi.org/10.1016/j.comnet.2024.110670	https://doi.org/10.1016/j.comnet.2024.110670	PROPN
easat-11054	516	6	https://doi.org/10.35833/mpce.2024.000909	https://doi.org/10.35833/mpce.2024.000909	PROPN
easat-11054	516	7	https://doi.org/10.1007/s10994-023-06303-2	https://doi.org/10.1007/s10994-023-06303-2	NUM
easat-11054	516	8	https://doi.org/10.1007/s10994-023-06303-2	https://doi.org/10.1007/s10994-023-06303-2	NUM
easat-11054	517	1	https://doi.org/10.1109/tnnls.2022.3212273	https://doi.org/10.1109/tnnls.2022.3212273	PROPN