id	sid	tid	token	lemma	pos
flr-107	1	1	frontline	frontline	NOUN
flr-107	1	2	learning	learn	VERB
flr-107	1	3	research	research	NOUN
flr-107	1	4	3	3	NUM
flr-107	1	5	(	(	PUNCT
flr-107	1	6	2014	2014	NUM
flr-107	1	7	)	)	PUNCT
flr-107	1	8	78	78	NUM
flr-107	1	9	-	-	SYM
flr-107	1	10	82	82	NUM
flr-107	1	11	issn	issn	PROPN
flr-107	1	12	2295	2295	NUM
flr-107	1	13	-	-	SYM
flr-107	1	14	3159	3159	NUM
flr-107	1	15	corresponding	corresponding	ADJ
flr-107	1	16	author	author	NOUN
flr-107	1	17	:	:	PUNCT
flr-107	1	18	tomi	tomi	PROPN
flr-107	1	19	silander	silander	NOUN
flr-107	1	20	,	,	PUNCT
flr-107	1	21	xerox	xerox	PROPN
flr-107	1	22	research	research	PROPN
flr-107	1	23	centre	centre	PROPN
flr-107	1	24	europe	europe	PROPN
flr-107	1	25	,	,	PUNCT
flr-107	1	26	www.xrce.xerox.com	www.xrce.xerox.com	NOUN
flr-107	1	27	,	,	PUNCT
flr-107	1	28	tomi.silander@xrce.xerox.com	tomi.silander@xrce.xerox.com	X
flr-107	1	29	;	;	PUNCT
flr-107	1	30	petri	petri	ADJ
flr-107	1	31	nokelainen	nokelainen	NOUN
flr-107	1	32	,	,	PUNCT
flr-107	1	33	university	university	NOUN
flr-107	1	34	of	of	ADP
flr-107	1	35	tampere	tampere	PROPN
flr-107	1	36	,	,	PUNCT
flr-107	1	37	www.uta.fi/edu	www.uta.fi/edu	PROPN
flr-107	1	38	,	,	PUNCT
flr-107	1	39	petri.nokelainen@uta.fi	petri.nokelainen@uta.fi	NOUN
flr-107	1	40	.	.	X
flr-107	1	41	http://dx.doi.org/10.14786/flr.v2i1.107	http://dx.doi.org/10.14786/flr.v2i1.107	VERB
flr-107	1	42	78	78	NUM
flr-107	1	43	|	|	NOUN
flr-107	1	44	f	f	NOUN
flr-107	1	45	l	l	NOUN
flr-107	1	46	r	r	NOUN
flr-107	1	47	using	use	VERB
flr-107	1	48	new	new	ADJ
flr-107	1	49	models	model	NOUN
flr-107	1	50	to	to	PART
flr-107	1	51	analyze	analyze	VERB
flr-107	1	52	complex	complex	ADJ
flr-107	1	53	regularities	regularity	NOUN
flr-107	1	54	of	of	ADP
flr-107	1	55	the	the	DET
flr-107	1	56	world	world	NOUN
flr-107	1	57	:	:	PUNCT
flr-107	1	58	commentary	commentary	NOUN
flr-107	1	59	on	on	ADP
flr-107	1	60	musso	musso	PROPN
flr-107	1	61	et	et	PROPN
flr-107	1	62	al	al	PROPN
flr-107	1	63	.	.	PROPN
flr-107	2	1	(	(	PUNCT
flr-107	2	2	2013	2013	NUM
flr-107	2	3	)	)	PUNCT
flr-107	2	4	petri	petri	ADJ
flr-107	2	5	nokelainen	nokelainen	NOUN
flr-107	2	6	a	a	PRON
flr-107	2	7	,	,	PUNCT
flr-107	2	8	tomi	tomi	PROPN
flr-107	2	9	silander	silander	PROPN
flr-107	3	1	b	b	PROPN
flr-107	3	2	a	a	DET
flr-107	3	3	university	university	NOUN
flr-107	3	4	of	of	ADP
flr-107	3	5	tampere	tampere	PROPN
flr-107	3	6	,	,	PUNCT
flr-107	3	7	finland	finland	PROPN
flr-107	3	8	b	b	PROPN
flr-107	3	9	xerox	xerox	PROPN
flr-107	3	10	research	research	PROPN
flr-107	3	11	centre	centre	PROPN
flr-107	3	12	europe	europe	PROPN
flr-107	3	13	,	,	PUNCT
flr-107	3	14	france	france	PROPN
flr-107	3	15	abstract	abstract	ADV
flr-107	3	16	this	this	DET
flr-107	3	17	commentary	commentary	NOUN
flr-107	3	18	to	to	ADP
flr-107	3	19	the	the	DET
flr-107	3	20	recent	recent	ADJ
flr-107	3	21	article	article	NOUN
flr-107	3	22	by	by	ADP
flr-107	3	23	musso	musso	PROPN
flr-107	3	24	et	et	PROPN
flr-107	3	25	al	al	PROPN
flr-107	3	26	.	.	PROPN
flr-107	4	1	(	(	PUNCT
flr-107	4	2	2013	2013	NUM
flr-107	4	3	)	)	PUNCT
flr-107	4	4	discusses	discuss	VERB
flr-107	4	5	issues	issue	NOUN
flr-107	4	6	related	relate	VERB
flr-107	4	7	to	to	ADP
flr-107	4	8	model	model	NOUN
flr-107	4	9	fitting	fitting	ADJ
flr-107	4	10	,	,	PUNCT
flr-107	4	11	comparison	comparison	NOUN
flr-107	4	12	of	of	ADP
flr-107	4	13	classification	classification	NOUN
flr-107	4	14	accuracy	accuracy	NOUN
flr-107	4	15	of	of	ADP
flr-107	4	16	generative	generative	ADJ
flr-107	4	17	and	and	CCONJ
flr-107	4	18	discriminative	discriminative	NOUN
flr-107	4	19	models	model	NOUN
flr-107	4	20	,	,	PUNCT
flr-107	4	21	and	and	CCONJ
flr-107	4	22	two	two	NUM
flr-107	4	23	(	(	PUNCT
flr-107	4	24	or	or	CCONJ
flr-107	4	25	more	more	ADJ
flr-107	4	26	)	)	PUNCT
flr-107	4	27	cultures	culture	NOUN
flr-107	4	28	of	of	ADP
flr-107	4	29	data	datum	NOUN
flr-107	4	30	modeling	modeling	NOUN
flr-107	4	31	.	.	PUNCT
flr-107	5	1	we	we	PRON
flr-107	5	2	start	start	VERB
flr-107	5	3	by	by	ADP
flr-107	5	4	questioning	question	VERB
flr-107	5	5	the	the	DET
flr-107	5	6	extremely	extremely	ADV
flr-107	5	7	high	high	ADJ
flr-107	5	8	classification	classification	NOUN
flr-107	5	9	accuracy	accuracy	NOUN
flr-107	5	10	with	with	ADP
flr-107	5	11	an	an	DET
flr-107	5	12	empirical	empirical	ADJ
flr-107	5	13	data	datum	NOUN
flr-107	5	14	from	from	ADP
flr-107	5	15	a	a	DET
flr-107	5	16	complex	complex	ADJ
flr-107	5	17	domain	domain	NOUN
flr-107	5	18	.	.	PUNCT
flr-107	6	1	there	there	PRON
flr-107	6	2	is	be	VERB
flr-107	6	3	a	a	DET
flr-107	6	4	risk	risk	NOUN
flr-107	6	5	that	that	PRON
flr-107	6	6	we	we	PRON
flr-107	6	7	model	model	VERB
flr-107	6	8	perfect	perfect	ADJ
flr-107	6	9	nonsense	nonsense	NOUN
flr-107	6	10	perfectly	perfectly	ADV
flr-107	6	11	.	.	PUNCT
flr-107	7	1	our	our	PRON
flr-107	7	2	second	second	ADJ
flr-107	7	3	concern	concern	NOUN
flr-107	7	4	is	be	AUX
flr-107	7	5	related	relate	VERB
flr-107	7	6	to	to	ADP
flr-107	7	7	the	the	DET
flr-107	7	8	relevance	relevance	NOUN
flr-107	7	9	of	of	ADP
flr-107	7	10	comparing	compare	VERB
flr-107	7	11	multilayer	multilayer	ADJ
flr-107	7	12	perceptron	perceptron	PROPN
flr-107	7	13	neural	neural	ADJ
flr-107	7	14	networks	network	NOUN
flr-107	7	15	and	and	CCONJ
flr-107	7	16	linear	linear	ADJ
flr-107	7	17	discriminant	discriminant	ADJ
flr-107	7	18	analysis	analysis	NOUN
flr-107	7	19	classification	classification	NOUN
flr-107	7	20	accuracy	accuracy	NOUN
flr-107	7	21	indices	index	NOUN
flr-107	7	22	.	.	PUNCT
flr-107	8	1	we	we	PRON
flr-107	8	2	find	find	VERB
flr-107	8	3	this	this	DET
flr-107	8	4	problematic	problematic	ADJ
flr-107	8	5	,	,	PUNCT
flr-107	8	6	as	as	SCONJ
flr-107	8	7	it	it	PRON
flr-107	8	8	is	be	AUX
flr-107	8	9	like	like	ADP
flr-107	8	10	comparing	compare	VERB
flr-107	8	11	apples	apple	NOUN
flr-107	8	12	and	and	CCONJ
flr-107	8	13	oranges	orange	NOUN
flr-107	8	14	.	.	PUNCT
flr-107	9	1	it	it	PRON
flr-107	9	2	would	would	AUX
flr-107	9	3	have	have	AUX
flr-107	9	4	been	be	AUX
flr-107	9	5	easier	easy	ADJ
flr-107	9	6	to	to	PART
flr-107	9	7	interpret	interpret	VERB
flr-107	9	8	the	the	DET
flr-107	9	9	model	model	NOUN
flr-107	9	10	and	and	CCONJ
flr-107	9	11	the	the	DET
flr-107	9	12	variable	variable	ADJ
flr-107	9	13	(	(	PUNCT
flr-107	9	14	group	group	NOUN
flr-107	9	15	)	)	PUNCT
flr-107	9	16	importance	importance	NOUN
flr-107	9	17	’s	’	VERB
flr-107	10	1	if	if	SCONJ
flr-107	10	2	the	the	DET
flr-107	10	3	authors	author	NOUN
flr-107	10	4	would	would	AUX
flr-107	10	5	have	have	AUX
flr-107	10	6	compared	compare	VERB
flr-107	10	7	mlp	mlp	PROPN
flr-107	10	8	to	to	ADP
flr-107	10	9	some	some	DET
flr-107	10	10	discriminative	discriminative	NOUN
flr-107	10	11	classifier	classifier	NOUN
flr-107	10	12	,	,	PUNCT
flr-107	10	13	such	such	ADJ
flr-107	10	14	as	as	ADP
flr-107	10	15	group	group	NOUN
flr-107	10	16	lasso	lasso	NOUN
flr-107	10	17	logistic	logistic	ADJ
flr-107	10	18	regression	regression	NOUN
flr-107	10	19	.	.	PUNCT
flr-107	11	1	finally	finally	ADV
flr-107	11	2	,	,	PUNCT
flr-107	11	3	we	we	PRON
flr-107	11	4	conclude	conclude	VERB
flr-107	11	5	our	our	PRON
flr-107	11	6	commentary	commentary	NOUN
flr-107	11	7	with	with	ADP
flr-107	11	8	a	a	DET
flr-107	11	9	discussion	discussion	NOUN
flr-107	11	10	about	about	ADP
flr-107	11	11	the	the	DET
flr-107	11	12	predictive	predictive	ADJ
flr-107	11	13	properties	property	NOUN
flr-107	11	14	of	of	ADP
flr-107	11	15	the	the	DET
flr-107	11	16	adopted	adopt	VERB
flr-107	11	17	data	datum	NOUN
flr-107	11	18	modeling	modeling	NOUN
flr-107	11	19	approach	approach	NOUN
flr-107	11	20	.	.	PUNCT
flr-107	12	1	keywords	keyword	NOUN
flr-107	12	2	:	:	PUNCT
flr-107	12	3	artificial	artificial	ADJ
flr-107	12	4	neural	neural	ADJ
flr-107	12	5	networks	network	NOUN
flr-107	12	6	;	;	PUNCT
flr-107	12	7	commentary	commentary	NOUN
flr-107	12	8	;	;	PUNCT
flr-107	12	9	model	model	NOUN
flr-107	12	10	-	-	PUNCT
flr-107	12	11	fit	fit	ADJ
flr-107	12	12	;	;	PUNCT
flr-107	12	13	generative	generative	ADJ
flr-107	12	14	and	and	CCONJ
flr-107	12	15	discriminative	discriminative	NOUN
flr-107	12	16	models	model	NOUN
flr-107	12	17	;	;	PUNCT
flr-107	12	18	algorithmic	algorithmic	ADJ
flr-107	12	19	data	datum	NOUN
flr-107	12	20	modeling	model	VERB
flr-107	12	21	1	1	NUM
flr-107	12	22	.	.	PUNCT
flr-107	13	1	introduction	introduction	NOUN
flr-107	13	2	statistical	statistical	ADJ
flr-107	13	3	methods	method	NOUN
flr-107	13	4	are	be	AUX
flr-107	13	5	constantly	constantly	ADV
flr-107	13	6	developed	develop	VERB
flr-107	13	7	,	,	PUNCT
flr-107	13	8	not	not	PART
flr-107	13	9	only	only	ADV
flr-107	13	10	within	within	ADP
flr-107	13	11	statistics	statistic	NOUN
flr-107	13	12	,	,	PUNCT
flr-107	13	13	but	but	CCONJ
flr-107	13	14	also	also	ADV
flr-107	13	15	in	in	ADP
flr-107	13	16	other	other	ADJ
flr-107	13	17	disciplines	discipline	NOUN
flr-107	13	18	,	,	PUNCT
flr-107	13	19	such	such	ADJ
flr-107	13	20	as	as	ADP
flr-107	13	21	,	,	PUNCT
flr-107	13	22	physics	physics	NOUN
flr-107	13	23	,	,	PUNCT
flr-107	13	24	economics	economic	NOUN
flr-107	13	25	,	,	PUNCT
flr-107	13	26	bioinformatics	bioinformatics	NOUN
flr-107	13	27	,	,	PUNCT
flr-107	13	28	linguistics	linguistic	NOUN
flr-107	13	29	,	,	PUNCT
flr-107	13	30	and	and	CCONJ
flr-107	13	31	computer	computer	NOUN
flr-107	13	32	science	science	NOUN
flr-107	13	33	.	.	PUNCT
flr-107	14	1	we	we	PRON
flr-107	14	2	therefore	therefore	ADV
flr-107	14	3	are	be	AUX
flr-107	14	4	very	very	ADV
flr-107	14	5	sympathetic	sympathetic	ADJ
flr-107	14	6	to	to	ADP
flr-107	14	7	the	the	DET
flr-107	14	8	attempts	attempt	NOUN
flr-107	14	9	to	to	PART
flr-107	14	10	promote	promote	VERB
flr-107	14	11	new	new	ADJ
flr-107	14	12	methods	method	NOUN
flr-107	14	13	for	for	ADP
flr-107	14	14	analyzing	analyze	VERB
flr-107	14	15	educational	educational	ADJ
flr-107	14	16	data	datum	NOUN
flr-107	14	17	.	.	PUNCT
flr-107	15	1	however	however	ADV
flr-107	15	2	,	,	PUNCT
flr-107	15	3	also	also	ADV
flr-107	15	4	in	in	ADP
flr-107	15	5	this	this	DET
flr-107	15	6	research	research	NOUN
flr-107	15	7	field	field	NOUN
flr-107	15	8	,	,	PUNCT
flr-107	15	9	for	for	ADP
flr-107	15	10	years	year	NOUN
flr-107	15	11	there	there	PRON
flr-107	15	12	has	have	AUX
flr-107	15	13	been	be	AUX
flr-107	15	14	an	an	DET
flr-107	15	15	emphasis	emphasis	NOUN
flr-107	15	16	on	on	ADP
flr-107	15	17	the	the	DET
flr-107	15	18	predictive	predictive	ADJ
flr-107	15	19	modeling	modeling	NOUN
flr-107	15	20	,	,	PUNCT
flr-107	15	21	for	for	ADP
flr-107	15	22	example	example	NOUN
flr-107	15	23	,	,	PUNCT
flr-107	15	24	to	to	PART
flr-107	15	25	learn	learn	VERB
flr-107	15	26	structures	structure	NOUN
flr-107	15	27	from	from	ADP
flr-107	15	28	the	the	DET
flr-107	15	29	data	datum	NOUN
flr-107	15	30	(	(	PUNCT
flr-107	15	31	nokelainen	nokelainen	ADJ
flr-107	15	32	,	,	PUNCT
flr-107	15	33	silander	silander	NOUN
flr-107	15	34	,	,	PUNCT
flr-107	15	35	ruohotie	ruohotie	NOUN
flr-107	15	36	,	,	PUNCT
flr-107	15	37	&	&	CCONJ
flr-107	15	38	tirri	tirri	NOUN
flr-107	15	39	,	,	PUNCT
flr-107	15	40	2007	2007	NUM
flr-107	15	41	;	;	PUNCT
flr-107	15	42	tirri	tirri	NOUN
flr-107	15	43	,	,	PUNCT
flr-107	15	44	nokelainen	nokelainen	NOUN
flr-107	15	45	,	,	PUNCT
flr-107	15	46	&	&	CCONJ
flr-107	15	47	komulainen	komulainen	PROPN
flr-107	15	48	,	,	PUNCT
flr-107	15	49	2013	2013	NUM
flr-107	15	50	)	)	PUNCT
flr-107	15	51	and	and	CCONJ
flr-107	15	52	to	to	PART
flr-107	15	53	predict	predict	VERB
flr-107	15	54	class	class	NOUN
flr-107	15	55	membership	membership	NOUN
flr-107	15	56	(	(	PUNCT
flr-107	15	57	nokelainen	nokelainen	NOUN
flr-107	15	58	&	&	CCONJ
flr-107	15	59	ruohotie	ruohotie	PROPN
flr-107	15	60	,	,	PUNCT
flr-107	15	61	2009	2009	NUM
flr-107	15	62	;	;	PUNCT
flr-107	15	63	nokelainen	nokelainen	NOUN
flr-107	15	64	,	,	PUNCT
flr-107	15	65	tirri	tirri	NOUN
flr-107	15	66	,	,	PUNCT
flr-107	15	67	campbell	campbell	PROPN
flr-107	15	68	,	,	PUNCT
flr-107	15	69	&	&	CCONJ
flr-107	15	70	walberg	walberg	PROPN
flr-107	15	71	,	,	PUNCT
flr-107	15	72	2007	2007	NUM
flr-107	15	73	;	;	PUNCT
flr-107	15	74	villaverde	villaverde	NOUN
flr-107	15	75	,	,	PUNCT
flr-107	15	76	p.	p.	NOUN
flr-107	15	77	nokelainen	nokelainen	PROPN
flr-107	15	78	&	&	CCONJ
flr-107	15	79	t.	t.	PROPN
flr-107	15	80	silander	silander	NOUN
flr-107	16	1	79	79	NUM
flr-107	17	1	|	|	NOUN
flr-107	17	2	f	f	NOUN
flr-107	17	3	l	l	NOUN
flr-107	17	4	r	r	NOUN
flr-107	17	5	godoy	godoy	PROPN
flr-107	17	6	,	,	PUNCT
flr-107	17	7	&	&	CCONJ
flr-107	17	8	amandi	amandi	PROPN
flr-107	17	9	,	,	PUNCT
flr-107	17	10	2006	2006	NUM
flr-107	17	11	)	)	PUNCT
flr-107	17	12	.	.	PUNCT
flr-107	18	1	the	the	DET
flr-107	18	2	recent	recent	ADJ
flr-107	18	3	boom	boom	NOUN
flr-107	18	4	of	of	ADP
flr-107	18	5	data	datum	NOUN
flr-107	18	6	analytics	analytic	NOUN
flr-107	18	7	has	have	AUX
flr-107	18	8	further	far	ADV
flr-107	18	9	increased	increase	VERB
flr-107	18	10	the	the	DET
flr-107	18	11	efforts	effort	NOUN
flr-107	18	12	in	in	ADP
flr-107	18	13	this	this	DET
flr-107	18	14	front	front	NOUN
flr-107	18	15	.	.	PUNCT
flr-107	19	1	one	one	NUM
flr-107	19	2	methodological	methodological	ADJ
flr-107	19	3	rationale	rationale	NOUN
flr-107	19	4	behind	behind	ADP
flr-107	19	5	this	this	DET
flr-107	19	6	development	development	NOUN
flr-107	19	7	is	be	AUX
flr-107	19	8	that	that	SCONJ
flr-107	19	9	predictiveness	predictiveness	NOUN
flr-107	19	10	guards	guard	NOUN
flr-107	19	11	against	against	ADP
flr-107	19	12	over	over	ADV
flr-107	19	13	-	-	PUNCT
flr-107	19	14	fitting	fit	VERB
flr-107	19	15	and	and	CCONJ
flr-107	19	16	serves	serve	VERB
flr-107	19	17	as	as	ADP
flr-107	19	18	a	a	DET
flr-107	19	19	natural	natural	ADJ
flr-107	19	20	criterion	criterion	NOUN
flr-107	19	21	for	for	ADP
flr-107	19	22	the	the	DET
flr-107	19	23	quality	quality	NOUN
flr-107	19	24	of	of	ADP
flr-107	19	25	the	the	DET
flr-107	19	26	model	model	NOUN
flr-107	19	27	.	.	PUNCT
flr-107	20	1	classical	classical	ADJ
flr-107	20	2	statistical	statistical	ADJ
flr-107	20	3	literature	literature	NOUN
flr-107	20	4	was	be	AUX
flr-107	20	5	not	not	PART
flr-107	20	6	emphasizing	emphasize	VERB
flr-107	20	7	this	this	DET
flr-107	20	8	aspect	aspect	NOUN
flr-107	20	9	,	,	PUNCT
flr-107	20	10	since	since	SCONJ
flr-107	20	11	models	model	NOUN
flr-107	20	12	were	be	AUX
flr-107	20	13	kept	keep	VERB
flr-107	20	14	relatively	relatively	ADV
flr-107	20	15	simple	simple	ADJ
flr-107	20	16	to	to	PART
flr-107	20	17	avoid	avoid	VERB
flr-107	20	18	over	over	ADV
flr-107	20	19	-	-	PUNCT
flr-107	20	20	fitting	fit	VERB
flr-107	20	21	and	and	CCONJ
flr-107	20	22	to	to	PART
flr-107	20	23	keep	keep	VERB
flr-107	20	24	the	the	DET
flr-107	20	25	calculations	calculation	NOUN
flr-107	20	26	reasonable	reasonable	ADJ
flr-107	20	27	.	.	PUNCT
flr-107	21	1	in	in	ADP
flr-107	21	2	addition	addition	NOUN
flr-107	21	3	,	,	PUNCT
flr-107	21	4	much	much	ADJ
flr-107	21	5	of	of	ADP
flr-107	21	6	the	the	DET
flr-107	21	7	theory	theory	NOUN
flr-107	21	8	concerned	concern	VERB
flr-107	21	9	asymptotic	asymptotic	ADJ
flr-107	21	10	behavior	behavior	NOUN
flr-107	21	11	in	in	ADP
flr-107	21	12	which	which	DET
flr-107	21	13	case	case	NOUN
flr-107	21	14	over	over	ADP
flr-107	21	15	-	-	PUNCT
flr-107	21	16	fitting	fitting	NOUN
flr-107	21	17	is	be	AUX
flr-107	21	18	usually	usually	ADV
flr-107	21	19	not	not	PART
flr-107	21	20	an	an	DET
flr-107	21	21	issue	issue	NOUN
flr-107	21	22	.	.	PUNCT
flr-107	22	1	increased	increase	VERB
flr-107	22	2	computing	computing	NOUN
flr-107	22	3	power	power	NOUN
flr-107	22	4	now	now	ADV
flr-107	22	5	allows	allow	VERB
flr-107	22	6	more	more	ADV
flr-107	22	7	complicated	complicated	ADJ
flr-107	22	8	models	model	NOUN
flr-107	22	9	,	,	PUNCT
flr-107	22	10	such	such	ADJ
flr-107	22	11	as	as	ADP
flr-107	22	12	bayesian	bayesian	NOUN
flr-107	22	13	,	,	PUNCT
flr-107	22	14	fuzzy	fuzzy	ADJ
flr-107	22	15	and	and	CCONJ
flr-107	22	16	neural	neural	ADJ
flr-107	22	17	networks	network	NOUN
flr-107	22	18	,	,	PUNCT
flr-107	22	19	to	to	PART
flr-107	22	20	be	be	AUX
flr-107	22	21	used	use	VERB
flr-107	22	22	.	.	PUNCT
flr-107	23	1	while	while	SCONJ
flr-107	23	2	the	the	DET
flr-107	23	3	increased	increase	VERB
flr-107	23	4	flexibility	flexibility	NOUN
flr-107	23	5	brings	bring	VERB
flr-107	23	6	benefits	benefit	NOUN
flr-107	23	7	,	,	PUNCT
flr-107	23	8	there	there	PRON
flr-107	23	9	are	be	VERB
flr-107	23	10	also	also	ADV
flr-107	23	11	possibilities	possibility	NOUN
flr-107	23	12	to	to	PART
flr-107	23	13	make	make	VERB
flr-107	23	14	new	new	ADJ
flr-107	23	15	kind	kind	NOUN
flr-107	23	16	of	of	ADP
flr-107	23	17	errors	error	NOUN
flr-107	23	18	in	in	ADP
flr-107	23	19	the	the	DET
flr-107	23	20	analysis	analysis	NOUN
flr-107	23	21	.	.	PUNCT
flr-107	24	1	since	since	SCONJ
flr-107	24	2	we	we	PRON
flr-107	24	3	share	share	VERB
flr-107	24	4	the	the	DET
flr-107	24	5	enthusiasm	enthusiasm	NOUN
flr-107	24	6	to	to	PART
flr-107	24	7	promote	promote	VERB
flr-107	24	8	new	new	ADJ
flr-107	24	9	methods	method	NOUN
flr-107	24	10	,	,	PUNCT
flr-107	24	11	we	we	PRON
flr-107	24	12	also	also	ADV
flr-107	24	13	feel	feel	VERB
flr-107	24	14	that	that	PRON
flr-107	24	15	is	be	AUX
flr-107	24	16	of	of	ADP
flr-107	24	17	utmost	utmost	ADJ
flr-107	24	18	importance	importance	NOUN
flr-107	24	19	to	to	PART
flr-107	24	20	perform	perform	VERB
flr-107	24	21	the	the	DET
flr-107	24	22	analyses	analysis	NOUN
flr-107	24	23	with	with	ADP
flr-107	24	24	these	these	DET
flr-107	24	25	new	new	ADJ
flr-107	24	26	methods	method	NOUN
flr-107	24	27	using	use	VERB
flr-107	24	28	extremely	extremely	ADV
flr-107	24	29	high	high	ADJ
flr-107	24	30	methodological	methodological	ADJ
flr-107	24	31	standards	standard	NOUN
flr-107	24	32	.	.	PUNCT
flr-107	25	1	in	in	ADP
flr-107	25	2	this	this	DET
flr-107	25	3	respect	respect	NOUN
flr-107	25	4	,	,	PUNCT
flr-107	25	5	we	we	PRON
flr-107	25	6	find	find	VERB
flr-107	25	7	some	some	PRON
flr-107	25	8	of	of	ADP
flr-107	25	9	the	the	DET
flr-107	25	10	procedures	procedure	NOUN
flr-107	25	11	followed	follow	VERB
flr-107	25	12	in	in	ADP
flr-107	25	13	the	the	DET
flr-107	25	14	recent	recent	ADJ
flr-107	25	15	article	article	NOUN
flr-107	25	16	by	by	ADP
flr-107	25	17	musso	musso	PROPN
flr-107	25	18	,	,	PUNCT
flr-107	25	19	kyndt	kyndt	PROPN
flr-107	25	20	,	,	PUNCT
flr-107	25	21	cascallar	cascallar	ADJ
flr-107	25	22	and	and	CCONJ
flr-107	25	23	dochy	dochy	ADJ
flr-107	25	24	(	(	PUNCT
flr-107	25	25	2013	2013	NUM
flr-107	25	26	)	)	PUNCT
flr-107	25	27	problematic	problematic	NOUN
flr-107	25	28	.	.	PUNCT
flr-107	26	1	before	before	ADP
flr-107	26	2	discussing	discuss	VERB
flr-107	26	3	about	about	ADP
flr-107	26	4	these	these	DET
flr-107	26	5	issues	issue	NOUN
flr-107	26	6	in	in	ADP
flr-107	26	7	detail	detail	NOUN
flr-107	26	8	,	,	PUNCT
flr-107	26	9	we	we	PRON
flr-107	26	10	wish	wish	VERB
flr-107	26	11	to	to	PART
flr-107	26	12	indicate	indicate	VERB
flr-107	26	13	that	that	SCONJ
flr-107	26	14	we	we	PRON
flr-107	26	15	agree	agree	VERB
flr-107	26	16	with	with	ADP
flr-107	26	17	edelsbrunner	edelsbrunner	NOUN
flr-107	26	18	and	and	CCONJ
flr-107	26	19	schneider	schneider	PROPN
flr-107	26	20	’s	’s	PART
flr-107	26	21	(	(	PUNCT
flr-107	26	22	2013	2013	NUM
flr-107	26	23	)	)	PUNCT
flr-107	26	24	previous	previous	ADJ
flr-107	26	25	commentary	commentary	NOUN
flr-107	26	26	on	on	ADP
flr-107	26	27	this	this	DET
flr-107	26	28	article	article	NOUN
flr-107	26	29	where	where	SCONJ
flr-107	26	30	they	they	PRON
flr-107	26	31	state	state	VERB
flr-107	26	32	that	that	SCONJ
flr-107	26	33	there	there	PRON
flr-107	26	34	are	be	VERB
flr-107	26	35	other	other	ADJ
flr-107	26	36	data	datum	NOUN
flr-107	26	37	analysis	analysis	NOUN
flr-107	26	38	techniques	technique	NOUN
flr-107	26	39	with	with	ADP
flr-107	26	40	similar	similar	ADJ
flr-107	26	41	properties	property	NOUN
flr-107	26	42	than	than	ADP
flr-107	26	43	anns	ann	NOUN
flr-107	26	44	,	,	PUNCT
flr-107	26	45	but	but	CCONJ
flr-107	26	46	without	without	ADP
flr-107	26	47	the	the	DET
flr-107	26	48	drawbacks	drawback	NOUN
flr-107	26	49	.	.	PUNCT
flr-107	27	1	2	2	X
flr-107	27	2	.	.	NOUN
flr-107	27	3	fitting	fit	VERB
flr-107	27	4	to	to	ADP
flr-107	27	5	the	the	DET
flr-107	27	6	test	test	NOUN
flr-107	27	7	data	datum	NOUN
flr-107	27	8	our	our	PRON
flr-107	27	9	first	first	ADJ
flr-107	27	10	concern	concern	NOUN
flr-107	27	11	is	be	AUX
flr-107	27	12	the	the	DET
flr-107	27	13	reported	reported	ADJ
flr-107	27	14	100	100	NUM
flr-107	27	15	%	%	NOUN
flr-107	27	16	classification	classification	NOUN
flr-107	27	17	accuracy	accuracy	NOUN
flr-107	27	18	in	in	ADP
flr-107	27	19	such	such	DET
flr-107	27	20	a	a	DET
flr-107	27	21	complex	complex	ADJ
flr-107	27	22	domain	domain	NOUN
flr-107	27	23	,	,	PUNCT
flr-107	27	24	and	and	CCONJ
flr-107	27	25	the	the	DET
flr-107	27	26	lack	lack	NOUN
flr-107	27	27	of	of	ADP
flr-107	27	28	thorough	thorough	ADJ
flr-107	27	29	discussion	discussion	NOUN
flr-107	27	30	of	of	ADP
flr-107	27	31	this	this	DET
flr-107	27	32	issue	issue	NOUN
flr-107	27	33	.	.	PUNCT
flr-107	28	1	multilayer	multilayer	PROPN
flr-107	28	2	perceptron	perceptron	PROPN
flr-107	28	3	(	(	PUNCT
flr-107	28	4	mlp	mlp	NOUN
flr-107	28	5	)	)	PUNCT
flr-107	28	6	neural	neural	ADJ
flr-107	28	7	networks	network	NOUN
flr-107	28	8	are	be	AUX
flr-107	28	9	universal	universal	ADJ
flr-107	28	10	function	function	NOUN
flr-107	28	11	approximators	approximator	NOUN
flr-107	28	12	(	(	PUNCT
flr-107	28	13	lek	lek	PROPN
flr-107	28	14	&	&	CCONJ
flr-107	28	15	guegan	guegan	PROPN
flr-107	28	16	,	,	PUNCT
flr-107	28	17	1999	1999	NUM
flr-107	28	18	)	)	PUNCT
flr-107	28	19	.	.	PUNCT
flr-107	29	1	with	with	ADP
flr-107	29	2	enough	enough	ADJ
flr-107	29	3	twisting	twisting	NOUN
flr-107	29	4	of	of	ADP
flr-107	29	5	the	the	DET
flr-107	29	6	parameters	parameter	NOUN
flr-107	29	7	,	,	PUNCT
flr-107	29	8	one	one	PRON
flr-107	29	9	can	can	AUX
flr-107	29	10	use	use	VERB
flr-107	29	11	them	they	PRON
flr-107	29	12	to	to	PART
flr-107	29	13	implement	implement	VERB
flr-107	29	14	any	any	DET
flr-107	29	15	classification	classification	NOUN
flr-107	29	16	rule	rule	NOUN
flr-107	29	17	(	(	PUNCT
flr-107	29	18	schittenkopf	schittenkopf	ADV
flr-107	29	19	,	,	PUNCT
flr-107	29	20	deco	deco	NOUN
flr-107	29	21	,	,	PUNCT
flr-107	29	22	&	&	CCONJ
flr-107	29	23	brauer	brauer	PROPN
flr-107	29	24	,	,	PUNCT
flr-107	29	25	1997	1997	NUM
flr-107	29	26	)	)	PUNCT
flr-107	29	27	.	.	PUNCT
flr-107	30	1	consequently	consequently	ADV
flr-107	30	2	,	,	PUNCT
flr-107	30	3	the	the	DET
flr-107	30	4	networks	network	NOUN
flr-107	30	5	could	could	AUX
flr-107	30	6	in	in	ADP
flr-107	30	7	theory	theory	NOUN
flr-107	30	8	also	also	ADV
flr-107	30	9	be	be	AUX
flr-107	30	10	designed	design	VERB
flr-107	30	11	to	to	PART
flr-107	30	12	explain	explain	VERB
flr-107	30	13	the	the	DET
flr-107	30	14	version	version	NOUN
flr-107	30	15	of	of	ADP
flr-107	30	16	the	the	DET
flr-107	30	17	dataset	dataset	NOUN
flr-107	30	18	in	in	ADP
flr-107	30	19	which	which	PRON
flr-107	30	20	the	the	DET
flr-107	30	21	gpa	gpa	PROPN
flr-107	30	22	scores	score	NOUN
flr-107	30	23	would	would	AUX
flr-107	30	24	be	be	AUX
flr-107	30	25	randomly	randomly	ADV
flr-107	30	26	assigned	assign	VERB
flr-107	30	27	to	to	ADP
flr-107	30	28	the	the	DET
flr-107	30	29	students	student	NOUN
flr-107	30	30	.	.	PUNCT
flr-107	31	1	what	what	PRON
flr-107	31	2	knowledge	knowledge	NOUN
flr-107	31	3	does	do	AUX
flr-107	31	4	such	such	DET
flr-107	31	5	a	a	DET
flr-107	31	6	model	model	NOUN
flr-107	31	7	(	(	PUNCT
flr-107	31	8	that	that	PRON
flr-107	31	9	can	can	AUX
flr-107	31	10	explain	explain	VERB
flr-107	31	11	anything	anything	PRON
flr-107	31	12	)	)	PUNCT
flr-107	31	13	extract	extract	VERB
flr-107	31	14	from	from	ADP
flr-107	31	15	the	the	DET
flr-107	31	16	real	real	ADJ
flr-107	31	17	world	world	NOUN
flr-107	31	18	?	?	PUNCT
flr-107	32	1	one	one	NUM
flr-107	32	2	persuasive	persuasive	ADJ
flr-107	32	3	answer	answer	NOUN
flr-107	32	4	does	do	AUX
flr-107	32	5	indeed	indeed	ADV
flr-107	32	6	lie	lie	VERB
flr-107	32	7	in	in	ADP
flr-107	32	8	prediction	prediction	NOUN
flr-107	32	9	.	.	PUNCT
flr-107	33	1	only	only	ADV
flr-107	33	2	the	the	DET
flr-107	33	3	regularities	regularity	NOUN
flr-107	33	4	help	help	VERB
flr-107	33	5	one	one	NOUN
flr-107	33	6	to	to	PART
flr-107	33	7	generalize	generalize	VERB
flr-107	33	8	beyond	beyond	ADP
flr-107	33	9	the	the	DET
flr-107	33	10	training	training	NOUN
flr-107	33	11	sample	sample	NOUN
flr-107	33	12	,	,	PUNCT
flr-107	33	13	that	that	ADV
flr-107	33	14	is	is	ADV
flr-107	33	15	,	,	PUNCT
flr-107	33	16	to	to	PART
flr-107	33	17	predict	predict	VERB
flr-107	33	18	.	.	PUNCT
flr-107	34	1	but	but	CCONJ
flr-107	34	2	here	here	ADV
flr-107	34	3	one	one	NUM
flr-107	34	4	needs	need	VERB
flr-107	34	5	to	to	PART
flr-107	34	6	be	be	AUX
flr-107	34	7	very	very	ADV
flr-107	34	8	careful	careful	ADJ
flr-107	34	9	.	.	PUNCT
flr-107	35	1	to	to	PART
flr-107	35	2	do	do	VERB
flr-107	35	3	this	this	PRON
flr-107	35	4	right	right	NOUN
flr-107	35	5	,	,	PUNCT
flr-107	35	6	the	the	DET
flr-107	35	7	data	datum	NOUN
flr-107	35	8	must	must	AUX
flr-107	35	9	first	first	ADV
flr-107	35	10	be	be	AUX
flr-107	35	11	split	split	VERB
flr-107	35	12	into	into	ADP
flr-107	35	13	two	two	NUM
flr-107	35	14	parts	part	NOUN
flr-107	35	15	and	and	CCONJ
flr-107	35	16	then	then	ADV
flr-107	35	17	the	the	DET
flr-107	35	18	model	model	NOUN
flr-107	35	19	must	must	AUX
flr-107	35	20	be	be	AUX
flr-107	35	21	built	build	VERB
flr-107	35	22	using	use	VERB
flr-107	35	23	only	only	ADV
flr-107	35	24	the	the	DET
flr-107	35	25	first	first	ADJ
flr-107	35	26	part	part	NOUN
flr-107	35	27	.	.	PUNCT
flr-107	36	1	the	the	DET
flr-107	36	2	testing	testing	NOUN
flr-107	36	3	should	should	AUX
flr-107	36	4	be	be	AUX
flr-107	36	5	done	do	VERB
flr-107	36	6	with	with	ADP
flr-107	36	7	the	the	DET
flr-107	36	8	second	second	ADJ
flr-107	36	9	part	part	NOUN
flr-107	36	10	of	of	ADP
flr-107	36	11	the	the	DET
flr-107	36	12	data	datum	NOUN
flr-107	36	13	–	–	PUNCT
flr-107	36	14	the	the	DET
flr-107	36	15	part	part	NOUN
flr-107	36	16	that	that	PRON
flr-107	36	17	was	be	AUX
flr-107	36	18	not	not	PART
flr-107	36	19	used	use	VERB
flr-107	36	20	in	in	ADP
flr-107	36	21	the	the	DET
flr-107	36	22	model	model	NOUN
flr-107	36	23	building	building	NOUN
flr-107	36	24	process	process	NOUN
flr-107	36	25	at	at	ADV
flr-107	36	26	all	all	ADV
flr-107	36	27	.	.	PUNCT
flr-107	37	1	the	the	DET
flr-107	37	2	big	big	ADJ
flr-107	37	3	question	question	NOUN
flr-107	37	4	is	be	AUX
flr-107	37	5	:	:	PUNCT
flr-107	37	6	can	can	AUX
flr-107	37	7	we	we	PRON
flr-107	37	8	trust	trust	VERB
flr-107	37	9	us	we	PRON
flr-107	37	10	to	to	PART
flr-107	37	11	be	be	AUX
flr-107	37	12	able	able	ADJ
flr-107	37	13	to	to	PART
flr-107	37	14	refrain	refrain	VERB
flr-107	37	15	from	from	ADP
flr-107	37	16	“	"	PUNCT
flr-107	37	17	cheating	cheat	VERB
flr-107	37	18	”	"	PUNCT
flr-107	37	19	(	(	PUNCT
flr-107	37	20	using	use	VERB
flr-107	37	21	the	the	DET
flr-107	37	22	test	test	NOUN
flr-107	37	23	data	datum	NOUN
flr-107	37	24	)	)	PUNCT
flr-107	37	25	?	?	PUNCT
flr-107	38	1	in	in	ADP
flr-107	38	2	order	order	NOUN
flr-107	38	3	to	to	PART
flr-107	38	4	avoid	avoid	VERB
flr-107	38	5	this	this	PRON
flr-107	38	6	,	,	PUNCT
flr-107	38	7	it	it	PRON
flr-107	38	8	would	would	AUX
flr-107	38	9	be	be	AUX
flr-107	38	10	best	good	ADJ
flr-107	38	11	to	to	PART
flr-107	38	12	gather	gather	VERB
flr-107	38	13	the	the	DET
flr-107	38	14	test	test	NOUN
flr-107	38	15	data	datum	NOUN
flr-107	38	16	after	after	ADP
flr-107	38	17	building	build	VERB
flr-107	38	18	the	the	DET
flr-107	38	19	model	model	NOUN
flr-107	38	20	,	,	PUNCT
flr-107	38	21	or	or	CCONJ
flr-107	38	22	to	to	PART
flr-107	38	23	separate	separate	VERB
flr-107	38	24	it	it	PRON
flr-107	38	25	from	from	ADP
flr-107	38	26	the	the	DET
flr-107	38	27	training	training	NOUN
flr-107	38	28	data	datum	NOUN
flr-107	38	29	in	in	ADP
flr-107	38	30	the	the	DET
flr-107	38	31	very	very	ADJ
flr-107	38	32	beginning	beginning	NOUN
flr-107	38	33	,	,	PUNCT
flr-107	38	34	and	and	CCONJ
flr-107	38	35	give	give	VERB
flr-107	38	36	it	it	PRON
flr-107	38	37	to	to	ADP
flr-107	38	38	somebody	somebody	PRON
flr-107	38	39	else	else	ADV
flr-107	38	40	who	who	PRON
flr-107	38	41	will	will	AUX
flr-107	38	42	then	then	ADV
flr-107	38	43	,	,	PUNCT
flr-107	38	44	after	after	SCONJ
flr-107	38	45	the	the	DET
flr-107	38	46	model	model	NOUN
flr-107	38	47	has	have	AUX
flr-107	38	48	been	be	AUX
flr-107	38	49	built	build	VERB
flr-107	38	50	,	,	PUNCT
flr-107	38	51	test	test	VERB
flr-107	38	52	the	the	DET
flr-107	38	53	accuracy	accuracy	NOUN
flr-107	38	54	of	of	ADP
flr-107	38	55	the	the	DET
flr-107	38	56	model	model	NOUN
flr-107	38	57	–	–	PUNCT
flr-107	38	58	once	once	ADV
flr-107	38	59	and	and	CCONJ
flr-107	38	60	for	for	ADP
flr-107	38	61	all	all	PRON
flr-107	38	62	!	!	PUNCT
flr-107	39	1	the	the	DET
flr-107	39	2	paper	paper	NOUN
flr-107	39	3	by	by	ADP
flr-107	39	4	musso	musso	PROPN
flr-107	39	5	and	and	CCONJ
flr-107	39	6	her	her	PRON
flr-107	39	7	colleagues	colleague	NOUN
flr-107	39	8	(	(	PUNCT
flr-107	39	9	2013	2013	NUM
flr-107	39	10	)	)	PUNCT
flr-107	39	11	practically	practically	ADV
flr-107	39	12	acknowledges	acknowledge	VERB
flr-107	39	13	that	that	SCONJ
flr-107	39	14	such	such	DET
flr-107	39	15	a	a	DET
flr-107	39	16	discipline	discipline	NOUN
flr-107	39	17	was	be	AUX
flr-107	39	18	not	not	PART
flr-107	39	19	rigorously	rigorously	ADV
flr-107	39	20	followed	follow	VERB
flr-107	39	21	.	.	PUNCT
flr-107	40	1	the	the	DET
flr-107	40	2	network	network	NOUN
flr-107	40	3	structure	structure	NOUN
flr-107	40	4	and	and	CCONJ
flr-107	40	5	learning	learn	VERB
flr-107	40	6	parameters	parameter	NOUN
flr-107	40	7	were	be	AUX
flr-107	40	8	adjusted	adjust	VERB
flr-107	40	9	to	to	PART
flr-107	40	10	maximize	maximize	VERB
flr-107	40	11	the	the	DET
flr-107	40	12	accuracy	accuracy	NOUN
flr-107	40	13	in	in	ADP
flr-107	40	14	test	test	NOUN
flr-107	40	15	data	datum	NOUN
flr-107	40	16	.	.	PUNCT
flr-107	41	1	many	many	ADJ
flr-107	41	2	models	model	NOUN
flr-107	41	3	were	be	AUX
flr-107	41	4	tested	test	VERB
flr-107	41	5	to	to	PART
flr-107	41	6	achieve	achieve	VERB
flr-107	41	7	this	this	PRON
flr-107	41	8	.	.	PUNCT
flr-107	42	1	even	even	ADV
flr-107	42	2	the	the	DET
flr-107	42	3	division	division	NOUN
flr-107	42	4	of	of	ADP
flr-107	42	5	the	the	DET
flr-107	42	6	data	datum	NOUN
flr-107	42	7	into	into	ADP
flr-107	42	8	training	training	NOUN
flr-107	42	9	and	and	CCONJ
flr-107	42	10	test	test	NOUN
flr-107	42	11	samples	sample	NOUN
flr-107	42	12	was	be	AUX
flr-107	42	13	manipulated	manipulate	VERB
flr-107	42	14	in	in	ADP
flr-107	42	15	order	order	NOUN
flr-107	42	16	to	to	PART
flr-107	42	17	“	"	PUNCT
flr-107	42	18	…	…	PUNCT
flr-107	42	19	maximize	maximize	VERB
flr-107	42	20	the	the	DET
flr-107	42	21	training	training	NOUN
flr-107	42	22	sample	sample	NOUN
flr-107	42	23	while	while	SCONJ
flr-107	42	24	preserving	preserve	VERB
flr-107	42	25	the	the	DET
flr-107	42	26	appearance	appearance	NOUN
flr-107	42	27	of	of	ADP
flr-107	42	28	all	all	DET
flr-107	42	29	detected	detect	VERB
flr-107	42	30	patterns	pattern	NOUN
flr-107	42	31	in	in	ADP
flr-107	42	32	the	the	DET
flr-107	42	33	testing	testing	NOUN
flr-107	42	34	sample	sample	NOUN
flr-107	42	35	…	…	PUNCT
flr-107	42	36	”	"	PUNCT
flr-107	42	37	(	(	PUNCT
flr-107	42	38	musso	musso	PROPN
flr-107	42	39	et	et	PROPN
flr-107	42	40	al	al	PROPN
flr-107	42	41	.	.	PROPN
flr-107	42	42	,	,	PUNCT
flr-107	42	43	2013	2013	NUM
flr-107	42	44	,	,	PUNCT
flr-107	42	45	60	60	NUM
flr-107	42	46	)	)	PUNCT
flr-107	42	47	.	.	PUNCT
flr-107	43	1	now	now	ADV
flr-107	43	2	,	,	PUNCT
flr-107	43	3	one	one	PRON
flr-107	43	4	can	can	AUX
flr-107	43	5	not	not	PART
flr-107	43	6	totally	totally	ADV
flr-107	43	7	exclude	exclude	VERB
flr-107	43	8	the	the	DET
flr-107	43	9	possibility	possibility	NOUN
flr-107	43	10	that	that	SCONJ
flr-107	43	11	the	the	DET
flr-107	43	12	authors	author	NOUN
flr-107	43	13	actually	actually	ADV
flr-107	43	14	promote	promote	VERB
flr-107	43	15	this	this	DET
flr-107	43	16	methodology	methodology	NOUN
flr-107	43	17	as	as	ADP
flr-107	43	18	a	a	DET
flr-107	43	19	sound	sound	ADJ
flr-107	43	20	one	one	NUM
flr-107	43	21	.	.	PUNCT
flr-107	44	1	take	take	VERB
flr-107	44	2	a	a	DET
flr-107	44	3	maximally	maximally	ADV
flr-107	44	4	flexible	flexible	ADJ
flr-107	44	5	model	model	NOUN
flr-107	44	6	family	family	NOUN
flr-107	44	7	,	,	PUNCT
flr-107	44	8	find	find	VERB
flr-107	44	9	the	the	DET
flr-107	44	10	most	most	ADV
flr-107	44	11	parsimonious	parsimonious	ADJ
flr-107	44	12	model	model	NOUN
flr-107	44	13	that	that	PRON
flr-107	44	14	fits	fit	VERB
flr-107	44	15	100	100	NUM
flr-107	44	16	%	%	NOUN
flr-107	44	17	to	to	ADP
flr-107	44	18	the	the	DET
flr-107	44	19	data	datum	NOUN
flr-107	44	20	,	,	PUNCT
flr-107	44	21	and	and	CCONJ
flr-107	44	22	then	then	ADV
flr-107	44	23	analyze	analyze	VERB
flr-107	44	24	the	the	DET
flr-107	44	25	model	model	NOUN
flr-107	44	26	.	.	PUNCT
flr-107	45	1	but	but	CCONJ
flr-107	45	2	if	if	SCONJ
flr-107	45	3	that	that	PRON
flr-107	45	4	were	be	AUX
flr-107	45	5	the	the	DET
flr-107	45	6	case	case	NOUN
flr-107	45	7	,	,	PUNCT
flr-107	45	8	why	why	SCONJ
flr-107	45	9	torture	torture	VERB
flr-107	45	10	oneself	oneself	PRON
flr-107	45	11	with	with	ADP
flr-107	45	12	the	the	DET
flr-107	45	13	tedious	tedious	ADJ
flr-107	45	14	manual	manual	ADJ
flr-107	45	15	work	work	NOUN
flr-107	45	16	to	to	PART
flr-107	45	17	find	find	VERB
flr-107	45	18	100	100	NUM
flr-107	45	19	%	%	NOUN
flr-107	45	20	fit	fit	ADJ
flr-107	45	21	(	(	PUNCT
flr-107	45	22	yes	yes	INTJ
flr-107	45	23	fit	fit	ADJ
flr-107	45	24	,	,	PUNCT
flr-107	45	25	not	not	PART
flr-107	45	26	generalization	generalization	NOUN
flr-107	45	27	)	)	PUNCT
flr-107	45	28	to	to	ADP
flr-107	45	29	the	the	DET
flr-107	45	30	test	test	NOUN
flr-107	45	31	data	datum	NOUN
flr-107	45	32	?	?	PUNCT
flr-107	46	1	it	it	PRON
flr-107	46	2	would	would	AUX
flr-107	46	3	be	be	AUX
flr-107	46	4	easier	easy	ADJ
flr-107	46	5	to	to	PART
flr-107	46	6	just	just	ADV
flr-107	46	7	fit	fit	VERB
flr-107	46	8	to	to	ADP
flr-107	46	9	the	the	DET
flr-107	46	10	whole	whole	ADJ
flr-107	46	11	data	datum	NOUN
flr-107	46	12	set	set	VERB
flr-107	46	13	–	–	PUNCT
flr-107	46	14	but	but	CCONJ
flr-107	46	15	that	that	PRON
flr-107	46	16	would	would	AUX
flr-107	46	17	break	break	VERB
flr-107	46	18	the	the	DET
flr-107	46	19	illusion	illusion	NOUN
flr-107	46	20	of	of	ADP
flr-107	46	21	prediction	prediction	NOUN
flr-107	46	22	.	.	PUNCT
flr-107	47	1	3	3	X
flr-107	47	2	.	.	X
flr-107	47	3	comparison	comparison	NOUN
flr-107	47	4	with	with	ADP
flr-107	47	5	the	the	DET
flr-107	47	6	linear	linear	ADJ
flr-107	47	7	discriminant	discriminant	NOUN
flr-107	47	8	analysis	analysis	NOUN
flr-107	47	9	we	we	PRON
flr-107	47	10	find	find	VERB
flr-107	47	11	that	that	SCONJ
flr-107	47	12	the	the	DET
flr-107	47	13	authors	author	NOUN
flr-107	47	14	’	’	PART
flr-107	47	15	decision	decision	NOUN
flr-107	47	16	to	to	PART
flr-107	47	17	compare	compare	VERB
flr-107	47	18	the	the	DET
flr-107	47	19	model	model	NOUN
flr-107	47	20	to	to	ADP
flr-107	47	21	the	the	DET
flr-107	47	22	other	other	ADJ
flr-107	47	23	models	model	NOUN
flr-107	47	24	sets	set	VERB
flr-107	47	25	a	a	DET
flr-107	47	26	very	very	ADV
flr-107	47	27	good	good	ADJ
flr-107	47	28	example	example	NOUN
flr-107	47	29	that	that	PRON
flr-107	47	30	should	should	AUX
flr-107	47	31	more	more	ADV
flr-107	47	32	often	often	ADV
flr-107	47	33	be	be	AUX
flr-107	47	34	followed	follow	VERB
flr-107	47	35	in	in	ADP
flr-107	47	36	the	the	DET
flr-107	47	37	educational	educational	ADJ
flr-107	47	38	research	research	NOUN
flr-107	47	39	.	.	PUNCT
flr-107	48	1	such	such	ADJ
flr-107	48	2	comparisons	comparison	NOUN
flr-107	48	3	are	be	AUX
flr-107	48	4	widely	widely	ADV
flr-107	48	5	used	use	VERB
flr-107	48	6	in	in	ADP
flr-107	48	7	machine	machine	NOUN
flr-107	48	8	learning	learning	NOUN
flr-107	48	9	(	(	PUNCT
flr-107	48	10	e.g.	e.g.	ADV
flr-107	48	11	,	,	PUNCT
flr-107	48	12	demšar	demšar	ADJ
flr-107	48	13	,	,	PUNCT
flr-107	48	14	2006	2006	NUM
flr-107	48	15	)	)	PUNCT
flr-107	48	16	.	.	PUNCT
flr-107	49	1	however	however	ADV
flr-107	49	2	,	,	PUNCT
flr-107	49	3	comparing	compare	VERB
flr-107	49	4	the	the	DET
flr-107	49	5	multilayer	multilayer	ADJ
flr-107	49	6	perceptron	perceptron	NOUN
flr-107	49	7	and	and	CCONJ
flr-107	49	8	the	the	DET
flr-107	49	9	discriminant	discriminant	ADJ
flr-107	49	10	analysis	analysis	NOUN
flr-107	49	11	raises	raise	VERB
flr-107	49	12	some	some	DET
flr-107	49	13	questions	question	NOUN
flr-107	49	14	.	.	PUNCT
flr-107	50	1	behind	behind	ADP
flr-107	50	2	the	the	DET
flr-107	50	3	linear	linear	ADJ
flr-107	50	4	discriminant	discriminant	NOUN
flr-107	50	5	analysis	analysis	NOUN
flr-107	50	6	is	be	AUX
flr-107	50	7	a	a	DET
flr-107	50	8	linear	linear	ADJ
flr-107	50	9	discriminant	discriminant	NOUN
flr-107	50	10	model	model	NOUN
flr-107	50	11	that	that	PRON
flr-107	50	12	defines	define	VERB
flr-107	50	13	a	a	DET
flr-107	50	14	joint	joint	ADJ
flr-107	50	15	probability	probability	NOUN
flr-107	50	16	distribution	distribution	NOUN
flr-107	50	17	for	for	ADP
flr-107	50	18	the	the	DET
flr-107	50	19	whole	whole	ADJ
flr-107	50	20	19	19	NUM
flr-107	50	21	-	-	PUNCT
flr-107	50	22	variate	variate	NOUN
flr-107	50	23	(	(	PUNCT
flr-107	50	24	18	18	NUM
flr-107	50	25	independent	independent	ADJ
flr-107	50	26	variables	variable	NOUN
flr-107	50	27	+	+	CCONJ
flr-107	50	28	gpa	gpa	PROPN
flr-107	50	29	p.	p.	PROPN
flr-107	50	30	nokelainen	nokelainen	PROPN
flr-107	50	31	&	&	CCONJ
flr-107	50	32	t.	t.	PROPN
flr-107	50	33	silander	silander	PROPN
flr-107	51	1	80	80	NUM
flr-107	52	1	|	|	NOUN
flr-107	52	2	f	f	NOUN
flr-107	52	3	l	l	NOUN
flr-107	52	4	r	r	NOUN
flr-107	52	5	class	class	NOUN
flr-107	52	6	)	)	PUNCT
flr-107	52	7	data	data	NOUN
flr-107	52	8	vector	vector	NOUN
flr-107	52	9	.	.	PUNCT
flr-107	53	1	such	such	ADJ
flr-107	53	2	joint	joint	ADJ
flr-107	53	3	probability	probability	NOUN
flr-107	53	4	distributions	distribution	NOUN
flr-107	53	5	can	can	AUX
flr-107	53	6	be	be	AUX
flr-107	53	7	used	use	VERB
flr-107	53	8	for	for	ADP
flr-107	53	9	classification	classification	NOUN
flr-107	53	10	,	,	PUNCT
flr-107	53	11	since	since	SCONJ
flr-107	53	12	the	the	DET
flr-107	53	13	conditional	conditional	ADJ
flr-107	53	14	probability	probability	NOUN
flr-107	53	15	p(gpa	p(gpa	NOUN
flr-107	53	16	-	-	PUNCT
flr-107	53	17	class	class	NOUN
flr-107	53	18	|	|	NOUN
flr-107	53	19	predictors	predictor	NOUN
flr-107	53	20	)	)	PUNCT
flr-107	53	21	is	be	AUX
flr-107	53	22	proportional	proportional	ADJ
flr-107	53	23	to	to	ADP
flr-107	53	24	the	the	DET
flr-107	53	25	joint	joint	ADJ
flr-107	53	26	distribution	distribution	NOUN
flr-107	53	27	p(gpa	p(gpa	NOUN
flr-107	53	28	-	-	PUNCT
flr-107	53	29	class	class	NOUN
flr-107	53	30	&	&	CCONJ
flr-107	53	31	predictors	predictor	NOUN
flr-107	53	32	)	)	PUNCT
flr-107	53	33	.	.	PUNCT
flr-107	54	1	these	these	DET
flr-107	54	2	kinds	kind	NOUN
flr-107	54	3	of	of	ADP
flr-107	54	4	classifiers	classifier	NOUN
flr-107	54	5	are	be	AUX
flr-107	54	6	usually	usually	ADV
flr-107	54	7	called	call	VERB
flr-107	54	8	generative	generative	ADJ
flr-107	54	9	classifiers	classifier	NOUN
flr-107	54	10	(	(	PUNCT
flr-107	54	11	e.g.	e.g.	ADV
flr-107	54	12	,	,	PUNCT
flr-107	54	13	xue	xue	PROPN
flr-107	54	14	&	&	CCONJ
flr-107	54	15	titterington	titterington	PROPN
flr-107	54	16	,	,	PUNCT
flr-107	54	17	2008	2008	NUM
flr-107	54	18	)	)	PUNCT
flr-107	54	19	,	,	PUNCT
flr-107	54	20	since	since	SCONJ
flr-107	54	21	they	they	PRON
flr-107	54	22	are	be	AUX
flr-107	54	23	based	base	VERB
flr-107	54	24	on	on	ADP
flr-107	54	25	the	the	DET
flr-107	54	26	models	model	NOUN
flr-107	54	27	that	that	PRON
flr-107	54	28	can	can	AUX
flr-107	54	29	be	be	AUX
flr-107	54	30	used	use	VERB
flr-107	54	31	to	to	PART
flr-107	54	32	sample	sample	VERB
flr-107	54	33	the	the	DET
flr-107	54	34	whole	whole	ADJ
flr-107	54	35	(	(	PUNCT
flr-107	54	36	19	19	NUM
flr-107	54	37	-	-	PUNCT
flr-107	54	38	variate	variate	NOUN
flr-107	54	39	)	)	PUNCT
flr-107	54	40	data	datum	NOUN
flr-107	54	41	vectors	vector	NOUN
flr-107	54	42	.	.	PUNCT
flr-107	55	1	mlps	mlp	NOUN
flr-107	55	2	are	be	AUX
flr-107	55	3	not	not	PART
flr-107	55	4	generative	generative	ADJ
flr-107	55	5	classifiers	classifier	NOUN
flr-107	55	6	,	,	PUNCT
flr-107	55	7	but	but	CCONJ
flr-107	55	8	so	so	ADV
flr-107	55	9	called	call	VERB
flr-107	55	10	discriminative	discriminative	NOUN
flr-107	55	11	classifiers	classifier	NOUN
flr-107	55	12	.	.	PUNCT
flr-107	56	1	they	they	PRON
flr-107	56	2	are	be	AUX
flr-107	56	3	built	build	VERB
flr-107	56	4	to	to	PART
flr-107	56	5	directly	directly	ADV
flr-107	56	6	estimate	estimate	VERB
flr-107	56	7	the	the	DET
flr-107	56	8	conditional	conditional	ADJ
flr-107	56	9	distribution	distribution	NOUN
flr-107	56	10	p(gpa	p(gpa	NOUN
flr-107	56	11	-	-	PUNCT
flr-107	56	12	class	class	NOUN
flr-107	57	1	|	|	NOUN
flr-107	57	2	predictors	predictor	NOUN
flr-107	57	3	)	)	PUNCT
flr-107	57	4	without	without	ADP
flr-107	57	5	modeling	model	VERB
flr-107	57	6	the	the	DET
flr-107	57	7	relationships	relationship	NOUN
flr-107	57	8	among	among	ADP
flr-107	57	9	the	the	DET
flr-107	57	10	predictors	predictor	NOUN
flr-107	57	11	.	.	PUNCT
flr-107	58	1	(	(	PUNCT
flr-107	58	2	reading	read	VERB
flr-107	58	3	the	the	DET
flr-107	58	4	paper	paper	NOUN
flr-107	58	5	sometimes	sometimes	ADV
flr-107	58	6	makes	make	VERB
flr-107	58	7	you	you	PRON
flr-107	58	8	feel	feel	VERB
flr-107	58	9	that	that	SCONJ
flr-107	58	10	the	the	DET
flr-107	58	11	authors	author	NOUN
flr-107	58	12	claim	claim	VERB
flr-107	58	13	otherwise	otherwise	ADV
flr-107	58	14	.	.	PUNCT
flr-107	58	15	)	)	PUNCT
flr-107	59	1	while	while	SCONJ
flr-107	59	2	the	the	DET
flr-107	59	3	linear	linear	ADJ
flr-107	59	4	discriminant	discriminant	NOUN
flr-107	59	5	model	model	PROPN
flr-107	59	6	da1	da1	PROPN
flr-107	59	7	used	use	VERB
flr-107	59	8	in	in	ADP
flr-107	59	9	the	the	DET
flr-107	59	10	musso	musso	PROPN
flr-107	59	11	et	et	PROPN
flr-107	59	12	al	al	PROPN
flr-107	59	13	.	.	PROPN
flr-107	60	1	(	(	PUNCT
flr-107	60	2	2013	2013	NUM
flr-107	60	3	)	)	PUNCT
flr-107	60	4	paper	paper	NOUN
flr-107	60	5	has	have	VERB
flr-107	60	6	about	about	ADV
flr-107	60	7	2	2	NUM
flr-107	60	8	*	*	SYM
flr-107	60	9	18	18	NUM
flr-107	60	10	+	+	SYM
flr-107	60	11	18	18	NUM
flr-107	60	12	*	*	SYM
flr-107	60	13	18	18	NUM
flr-107	60	14	=	=	SYM
flr-107	60	15	360	360	NUM
flr-107	60	16	parameters	parameter	NOUN
flr-107	60	17	,	,	PUNCT
flr-107	60	18	the	the	DET
flr-107	60	19	neural	neural	ADJ
flr-107	60	20	network	network	NOUN
flr-107	60	21	model	model	NOUN
flr-107	60	22	has	have	VERB
flr-107	60	23	18	18	NUM
flr-107	60	24	*	*	SYM
flr-107	60	25	15	15	NUM
flr-107	60	26	*	*	SYM
flr-107	60	27	2	2	NUM
flr-107	60	28	=	=	SYM
flr-107	60	29	540	540	NUM
flr-107	60	30	parameters	parameter	NOUN
flr-107	60	31	.	.	PUNCT
flr-107	61	1	the	the	DET
flr-107	61	2	difference	difference	NOUN
flr-107	61	3	in	in	ADP
flr-107	61	4	number	number	NOUN
flr-107	61	5	of	of	ADP
flr-107	61	6	parameters	parameter	NOUN
flr-107	61	7	is	be	AUX
flr-107	61	8	not	not	PART
flr-107	61	9	huge	huge	ADJ
flr-107	61	10	,	,	PUNCT
flr-107	61	11	but	but	CCONJ
flr-107	61	12	all	all	DET
flr-107	61	13	the	the	DET
flr-107	61	14	parameters	parameter	NOUN
flr-107	61	15	of	of	ADP
flr-107	61	16	the	the	DET
flr-107	61	17	mlp	mlp	NOUN
flr-107	61	18	are	be	AUX
flr-107	61	19	used	use	VERB
flr-107	61	20	for	for	ADP
flr-107	61	21	modeling	model	VERB
flr-107	61	22	the	the	DET
flr-107	61	23	conditional	conditional	ADJ
flr-107	61	24	distribution	distribution	NOUN
flr-107	61	25	,	,	PUNCT
flr-107	61	26	while	while	SCONJ
flr-107	61	27	the	the	DET
flr-107	61	28	parameters	parameter	NOUN
flr-107	61	29	in	in	ADP
flr-107	61	30	the	the	DET
flr-107	61	31	linear	linear	ADJ
flr-107	61	32	discriminant	discriminant	NOUN
flr-107	61	33	model	model	NOUN
flr-107	61	34	also	also	ADV
flr-107	61	35	take	take	VERB
flr-107	61	36	care	care	NOUN
flr-107	61	37	of	of	ADP
flr-107	61	38	modeling	model	VERB
flr-107	61	39	the	the	DET
flr-107	61	40	relationships	relationship	NOUN
flr-107	61	41	between	between	ADP
flr-107	61	42	variables	variable	NOUN
flr-107	61	43	.	.	PUNCT
flr-107	62	1	since	since	SCONJ
flr-107	62	2	the	the	DET
flr-107	62	3	linear	linear	ADJ
flr-107	62	4	discriminant	discriminant	NOUN
flr-107	62	5	is	be	AUX
flr-107	62	6	also	also	ADV
flr-107	62	7	a	a	DET
flr-107	62	8	predictive	predictive	ADJ
flr-107	62	9	classifier	classifier	NOUN
flr-107	62	10	,	,	PUNCT
flr-107	62	11	one	one	PRON
flr-107	62	12	can	can	AUX
flr-107	62	13	not	not	PART
flr-107	62	14	but	but	ADV
flr-107	62	15	wonder	wonder	VERB
flr-107	62	16	why	why	SCONJ
flr-107	62	17	the	the	DET
flr-107	62	18	confusion	confusion	NOUN
flr-107	62	19	matrices	matrix	NOUN
flr-107	62	20	for	for	ADP
flr-107	62	21	linear	linear	ADJ
flr-107	62	22	discriminants	discriminant	NOUN
flr-107	62	23	were	be	AUX
flr-107	62	24	not	not	PART
flr-107	62	25	reported	report	VERB
flr-107	62	26	.	.	PUNCT
flr-107	63	1	those	those	DET
flr-107	63	2	numbers	number	NOUN
flr-107	63	3	surely	surely	ADV
flr-107	63	4	would	would	AUX
flr-107	63	5	have	have	AUX
flr-107	63	6	fitted	fit	VERB
flr-107	63	7	to	to	ADP
flr-107	63	8	the	the	DET
flr-107	63	9	same	same	ADJ
flr-107	63	10	space	space	NOUN
flr-107	63	11	without	without	ADP
flr-107	63	12	any	any	DET
flr-107	63	13	problem	problem	NOUN
flr-107	63	14	.	.	PUNCT
flr-107	64	1	on	on	ADP
flr-107	64	2	the	the	DET
flr-107	64	3	other	other	ADJ
flr-107	64	4	hand	hand	NOUN
flr-107	64	5	,	,	PUNCT
flr-107	64	6	it	it	PRON
flr-107	64	7	is	be	AUX
flr-107	64	8	plausible	plausible	ADJ
flr-107	64	9	that	that	SCONJ
flr-107	64	10	any	any	DET
flr-107	64	11	differences	difference	NOUN
flr-107	64	12	found	find	VERB
flr-107	64	13	are	be	AUX
flr-107	64	14	due	due	ADJ
flr-107	64	15	to	to	ADP
flr-107	64	16	the	the	DET
flr-107	64	17	other	other	ADJ
flr-107	64	18	classifier	classifier	NOUN
flr-107	64	19	being	be	AUX
flr-107	64	20	generative	generative	ADJ
flr-107	64	21	and	and	CCONJ
flr-107	64	22	the	the	DET
flr-107	64	23	other	other	ADJ
flr-107	64	24	one	one	NUM
flr-107	64	25	discriminative	discriminative	NOUN
flr-107	64	26	.	.	PUNCT
flr-107	65	1	it	it	PRON
flr-107	65	2	would	would	AUX
flr-107	65	3	have	have	AUX
flr-107	65	4	been	be	AUX
flr-107	65	5	much	much	ADV
flr-107	65	6	more	more	ADV
flr-107	65	7	meaningful	meaningful	ADJ
flr-107	65	8	to	to	PART
flr-107	65	9	compare	compare	VERB
flr-107	65	10	the	the	DET
flr-107	65	11	mlp	mlp	NOUN
flr-107	65	12	to	to	ADP
flr-107	65	13	some	some	DET
flr-107	65	14	discriminative	discriminative	NOUN
flr-107	65	15	classifier	classifier	NOUN
flr-107	65	16	such	such	ADJ
flr-107	65	17	as	as	ADP
flr-107	65	18	a	a	DET
flr-107	65	19	logistic	logistic	ADJ
flr-107	65	20	regression	regression	NOUN
flr-107	65	21	,	,	PUNCT
flr-107	65	22	or	or	CCONJ
flr-107	65	23	better	well	ADV
flr-107	65	24	yet	yet	ADV
flr-107	65	25	,	,	PUNCT
flr-107	65	26	some	some	DET
flr-107	65	27	sparse	sparse	ADJ
flr-107	65	28	version	version	NOUN
flr-107	65	29	of	of	ADP
flr-107	65	30	it	it	PRON
flr-107	65	31	such	such	ADJ
flr-107	65	32	as	as	ADP
flr-107	65	33	the	the	DET
flr-107	65	34	group	group	NOUN
flr-107	65	35	lasso	lasso	NOUN
flr-107	65	36	with	with	ADP
flr-107	65	37	interaction	interaction	NOUN
flr-107	65	38	terms	term	NOUN
flr-107	65	39	(	(	PUNCT
flr-107	65	40	meier	meier	PROPN
flr-107	65	41	,	,	PUNCT
flr-107	65	42	van	van	PROPN
flr-107	65	43	de	de	PROPN
flr-107	65	44	geer	geer	PROPN
flr-107	65	45	,	,	PUNCT
flr-107	65	46	&	&	CCONJ
flr-107	65	47	bühlmann	bühlmann	PROPN
flr-107	65	48	,	,	PUNCT
flr-107	65	49	2008	2008	NUM
flr-107	65	50	)	)	PUNCT
flr-107	65	51	that	that	PRON
flr-107	65	52	would	would	AUX
flr-107	65	53	make	make	VERB
flr-107	65	54	interpreting	interpret	VERB
flr-107	65	55	the	the	DET
flr-107	65	56	model	model	NOUN
flr-107	65	57	and	and	CCONJ
flr-107	65	58	the	the	DET
flr-107	65	59	variable	variable	ADJ
flr-107	65	60	(	(	PUNCT
flr-107	65	61	group	group	NOUN
flr-107	65	62	)	)	PUNCT
flr-107	65	63	importances	importance	VERB
flr-107	65	64	much	much	ADV
flr-107	65	65	easier	easy	ADJ
flr-107	65	66	.	.	PUNCT
flr-107	66	1	furthermore	furthermore	ADV
flr-107	66	2	,	,	PUNCT
flr-107	66	3	the	the	DET
flr-107	66	4	musso	musso	PROPN
flr-107	66	5	et	et	PROPN
flr-107	66	6	al	al	PROPN
flr-107	66	7	.	.	PROPN
flr-107	66	8	(	(	PUNCT
flr-107	66	9	2013	2013	NUM
flr-107	66	10	)	)	PUNCT
flr-107	66	11	paper	paper	NOUN
flr-107	66	12	is	be	AUX
flr-107	66	13	very	very	ADV
flr-107	66	14	unclear	unclear	ADJ
flr-107	66	15	about	about	ADP
flr-107	66	16	how	how	SCONJ
flr-107	66	17	the	the	DET
flr-107	66	18	variable	variable	ADJ
flr-107	66	19	importances	importance	NOUN
flr-107	66	20	have	have	AUX
flr-107	66	21	been	be	AUX
flr-107	66	22	calculated	calculate	VERB
flr-107	66	23	.	.	PUNCT
flr-107	67	1	the	the	DET
flr-107	67	2	attempt	attempt	NOUN
flr-107	67	3	to	to	PART
flr-107	67	4	follow	follow	VERB
flr-107	67	5	the	the	DET
flr-107	67	6	references	reference	NOUN
flr-107	67	7	only	only	ADV
flr-107	67	8	lead	lead	VERB
flr-107	67	9	to	to	ADP
flr-107	67	10	the	the	DET
flr-107	67	11	statements	statement	NOUN
flr-107	67	12	like	like	ADP
flr-107	67	13	“	"	PUNCT
flr-107	67	14	this	this	PRON
flr-107	67	15	has	have	AUX
flr-107	67	16	been	be	AUX
flr-107	67	17	implemented	implement	VERB
flr-107	67	18	in	in	ADP
flr-107	67	19	software	software	NOUN
flr-107	67	20	x	x	NOUN
flr-107	67	21	”	"	PUNCT
flr-107	67	22	or	or	CCONJ
flr-107	67	23	to	to	ADP
flr-107	67	24	an	an	DET
flr-107	67	25	unpublished	unpublished	ADJ
flr-107	67	26	technical	technical	ADJ
flr-107	67	27	report	report	NOUN
flr-107	67	28	by	by	ADP
flr-107	67	29	one	one	NUM
flr-107	67	30	of	of	ADP
flr-107	67	31	the	the	DET
flr-107	67	32	authors	author	NOUN
flr-107	67	33	.	.	PUNCT
flr-107	68	1	4	4	X
flr-107	68	2	.	.	X
flr-107	68	3	two	two	NUM
flr-107	68	4	cultures	culture	NOUN
flr-107	68	5	according	accord	VERB
flr-107	68	6	to	to	ADP
flr-107	68	7	breiman	breiman	NOUN
flr-107	68	8	(	(	PUNCT
flr-107	68	9	2001b	2001b	NUM
flr-107	68	10	)	)	PUNCT
flr-107	68	11	,	,	PUNCT
flr-107	68	12	there	there	PRON
flr-107	68	13	are	be	VERB
flr-107	68	14	two	two	NUM
flr-107	68	15	statistical	statistical	ADJ
flr-107	68	16	modeling	modeling	NOUN
flr-107	68	17	cultures	culture	NOUN
flr-107	68	18	.	.	PUNCT
flr-107	69	1	the	the	DET
flr-107	69	2	data	datum	NOUN
flr-107	69	3	modeling	model	VERB
flr-107	69	4	culture	culture	NOUN
flr-107	69	5	assumes	assume	VERB
flr-107	69	6	that	that	SCONJ
flr-107	69	7	the	the	DET
flr-107	69	8	data	datum	NOUN
flr-107	69	9	are	be	AUX
flr-107	69	10	generated	generate	VERB
flr-107	69	11	by	by	ADP
flr-107	69	12	a	a	DET
flr-107	69	13	given	give	VERB
flr-107	69	14	stochastic	stochastic	ADJ
flr-107	69	15	data	datum	NOUN
flr-107	69	16	model	model	NOUN
flr-107	69	17	(	(	PUNCT
flr-107	69	18	such	such	ADJ
flr-107	69	19	as	as	ADP
flr-107	69	20	linear	linear	ADJ
flr-107	69	21	or	or	CCONJ
flr-107	69	22	logistic	logistic	ADJ
flr-107	69	23	regression	regression	NOUN
flr-107	69	24	)	)	PUNCT
flr-107	69	25	.	.	PUNCT
flr-107	70	1	the	the	DET
flr-107	70	2	algorithmic	algorithmic	ADJ
flr-107	70	3	modeling	modeling	NOUN
flr-107	70	4	culture	culture	NOUN
flr-107	70	5	treats	treat	VERB
flr-107	70	6	the	the	DET
flr-107	70	7	data	datum	NOUN
flr-107	70	8	mechanism	mechanism	NOUN
flr-107	70	9	as	as	ADP
flr-107	70	10	unknown	unknown	ADJ
flr-107	70	11	,	,	PUNCT
flr-107	70	12	using	use	VERB
flr-107	70	13	,	,	PUNCT
flr-107	70	14	for	for	ADP
flr-107	70	15	example	example	NOUN
flr-107	70	16	,	,	PUNCT
flr-107	70	17	decision	decision	NOUN
flr-107	70	18	trees	tree	NOUN
flr-107	70	19	and	and	CCONJ
flr-107	70	20	neural	neural	ADJ
flr-107	70	21	networks	network	NOUN
flr-107	70	22	.	.	PUNCT
flr-107	71	1	although	although	SCONJ
flr-107	71	2	the	the	DET
flr-107	71	3	first	first	ADJ
flr-107	71	4	of	of	ADP
flr-107	71	5	these	these	DET
flr-107	71	6	two	two	NUM
flr-107	71	7	cultures	culture	NOUN
flr-107	71	8	,	,	PUNCT
flr-107	71	9	focusing	focus	VERB
flr-107	71	10	on	on	ADP
flr-107	71	11	data	datum	NOUN
flr-107	71	12	models	model	NOUN
flr-107	71	13	,	,	PUNCT
flr-107	71	14	is	be	AUX
flr-107	71	15	still	still	ADV
flr-107	71	16	dominating	dominate	VERB
flr-107	71	17	,	,	PUNCT
flr-107	71	18	many	many	ADJ
flr-107	71	19	fields	field	NOUN
flr-107	71	20	outside	outside	ADP
flr-107	71	21	statistics	statistic	NOUN
flr-107	71	22	are	be	AUX
flr-107	71	23	rapidly	rapidly	ADV
flr-107	71	24	adopting	adopt	VERB
flr-107	71	25	a	a	DET
flr-107	71	26	wide	wide	ADJ
flr-107	71	27	variety	variety	NOUN
flr-107	71	28	of	of	ADP
flr-107	71	29	tools	tool	NOUN
flr-107	71	30	.	.	PUNCT
flr-107	72	1	neural	neural	ADJ
flr-107	72	2	networks	network	NOUN
flr-107	72	3	are	be	AUX
flr-107	72	4	often	often	ADV
flr-107	72	5	considered	consider	VERB
flr-107	72	6	as	as	ADP
flr-107	72	7	black	black	ADJ
flr-107	72	8	-	-	PUNCT
flr-107	72	9	box	box	NOUN
flr-107	72	10	models	model	NOUN
flr-107	72	11	that	that	PRON
flr-107	72	12	do	do	AUX
flr-107	72	13	not	not	PART
flr-107	72	14	offer	offer	VERB
flr-107	72	15	a	a	DET
flr-107	72	16	good	good	ADJ
flr-107	72	17	explanation	explanation	NOUN
flr-107	72	18	and	and	CCONJ
flr-107	72	19	understanding	understanding	NOUN
flr-107	72	20	of	of	ADP
flr-107	72	21	the	the	DET
flr-107	72	22	domain	domain	NOUN
flr-107	72	23	(	(	PUNCT
flr-107	72	24	correa	correa	PROPN
flr-107	72	25	,	,	PUNCT
flr-107	72	26	bielza	bielza	NOUN
flr-107	72	27	,	,	PUNCT
flr-107	72	28	&	&	CCONJ
flr-107	72	29	pamies	pamies	PROPN
flr-107	72	30	-	-	PUNCT
flr-107	72	31	teixeira	teixeira	NOUN
flr-107	72	32	,	,	PUNCT
flr-107	72	33	2009	2009	NUM
flr-107	72	34	)	)	PUNCT
flr-107	72	35	.	.	PUNCT
flr-107	73	1	consequently	consequently	ADV
flr-107	73	2	,	,	PUNCT
flr-107	73	3	such	such	ADJ
flr-107	73	4	models	model	NOUN
flr-107	73	5	are	be	AUX
flr-107	73	6	sometimes	sometimes	ADV
flr-107	73	7	hastily	hastily	ADV
flr-107	73	8	deemed	deem	VERB
flr-107	73	9	as	as	ADP
flr-107	73	10	unsuitable	unsuitable	ADJ
flr-107	73	11	for	for	ADP
flr-107	73	12	much	much	ADJ
flr-107	73	13	of	of	ADP
flr-107	73	14	the	the	DET
flr-107	73	15	science	science	NOUN
flr-107	73	16	.	.	PUNCT
flr-107	74	1	we	we	PRON
flr-107	74	2	would	would	AUX
flr-107	74	3	like	like	VERB
flr-107	74	4	to	to	PART
flr-107	74	5	take	take	VERB
flr-107	74	6	the	the	DET
flr-107	74	7	opportunity	opportunity	NOUN
flr-107	74	8	to	to	PART
flr-107	74	9	say	say	VERB
flr-107	74	10	a	a	DET
flr-107	74	11	word	word	NOUN
flr-107	74	12	for	for	ADP
flr-107	74	13	such	such	ADJ
flr-107	74	14	black	black	ADJ
flr-107	74	15	-	-	PUNCT
flr-107	74	16	box	box	NOUN
flr-107	74	17	models	model	NOUN
flr-107	74	18	along	along	ADP
flr-107	74	19	the	the	DET
flr-107	74	20	lines	line	NOUN
flr-107	74	21	expressed	express	VERB
flr-107	74	22	by	by	ADP
flr-107	74	23	a	a	DET
flr-107	74	24	statistician	statistician	NOUN
flr-107	74	25	,	,	PUNCT
flr-107	74	26	leo	leo	PROPN
flr-107	74	27	breiman	breiman	PROPN
flr-107	74	28	(	(	PUNCT
flr-107	74	29	1928	1928	NUM
flr-107	74	30	-	-	SYM
flr-107	74	31	2005	2005	NUM
flr-107	74	32	)	)	PUNCT
flr-107	74	33	.	.	PUNCT
flr-107	75	1	world	world	NOUN
flr-107	75	2	may	may	AUX
flr-107	75	3	not	not	PART
flr-107	75	4	be	be	AUX
flr-107	75	5	a	a	DET
flr-107	75	6	simple	simple	ADJ
flr-107	75	7	place	place	NOUN
flr-107	75	8	.	.	PUNCT
flr-107	76	1	while	while	SCONJ
flr-107	76	2	among	among	ADP
flr-107	76	3	the	the	DET
flr-107	76	4	simple	simple	ADJ
flr-107	76	5	theories	theory	NOUN
flr-107	76	6	there	there	PRON
flr-107	76	7	are	be	VERB
flr-107	76	8	those	those	PRON
flr-107	76	9	who	who	PRON
flr-107	76	10	most	most	ADV
flr-107	76	11	closely	closely	ADV
flr-107	76	12	approximate	approximate	VERB
flr-107	76	13	the	the	DET
flr-107	76	14	complex	complex	ADJ
flr-107	76	15	reality	reality	NOUN
flr-107	76	16	,	,	PUNCT
flr-107	76	17	it	it	PRON
flr-107	76	18	is	be	AUX
flr-107	76	19	a	a	DET
flr-107	76	20	priori	priori	ADJ
flr-107	76	21	possible	possible	ADJ
flr-107	76	22	that	that	SCONJ
flr-107	76	23	none	none	NOUN
flr-107	76	24	of	of	ADP
flr-107	76	25	those	those	DET
flr-107	76	26	simple	simple	ADJ
flr-107	76	27	theories	theory	NOUN
flr-107	76	28	,	,	PUNCT
flr-107	76	29	even	even	ADV
flr-107	76	30	the	the	DET
flr-107	76	31	best	good	ADJ
flr-107	76	32	of	of	ADP
flr-107	76	33	them	they	PRON
flr-107	76	34	,	,	PUNCT
flr-107	76	35	approximate	approximate	VERB
flr-107	76	36	the	the	DET
flr-107	76	37	situation	situation	NOUN
flr-107	76	38	well	well	ADV
flr-107	76	39	.	.	PUNCT
flr-107	77	1	if	if	SCONJ
flr-107	77	2	the	the	DET
flr-107	77	3	model	model	NOUN
flr-107	77	4	does	do	AUX
flr-107	77	5	not	not	PART
flr-107	77	6	predict	predict	VERB
flr-107	77	7	well	well	ADV
flr-107	77	8	,	,	PUNCT
flr-107	77	9	one	one	PRON
flr-107	77	10	can	can	AUX
flr-107	77	11	argue	argue	VERB
flr-107	77	12	that	that	SCONJ
flr-107	77	13	it	it	PRON
flr-107	77	14	has	have	AUX
flr-107	77	15	not	not	PART
flr-107	77	16	captured	capture	VERB
flr-107	77	17	the	the	DET
flr-107	77	18	regularities	regularity	NOUN
flr-107	77	19	of	of	ADP
flr-107	77	20	the	the	DET
flr-107	77	21	world	world	NOUN
flr-107	77	22	,	,	PUNCT
flr-107	77	23	so	so	ADV
flr-107	77	24	what	what	PRON
flr-107	77	25	insight	insight	NOUN
flr-107	77	26	would	would	AUX
flr-107	77	27	understanding	understand	VERB
flr-107	77	28	and	and	CCONJ
flr-107	77	29	interpreting	interpret	VERB
flr-107	77	30	such	such	DET
flr-107	77	31	a	a	DET
flr-107	77	32	model	model	NOUN
flr-107	77	33	offer	offer	VERB
flr-107	77	34	us	we	PRON
flr-107	77	35	.	.	PUNCT
flr-107	78	1	(	(	PUNCT
flr-107	78	2	breiman	breiman	NOUN
flr-107	78	3	,	,	PUNCT
flr-107	78	4	2001b	2001b	NUM
flr-107	78	5	.	.	PUNCT
flr-107	78	6	)	)	PUNCT
flr-107	79	1	most	most	ADJ
flr-107	79	2	of	of	ADP
flr-107	79	3	the	the	DET
flr-107	79	4	statistical	statistical	ADJ
flr-107	79	5	community	community	NOUN
flr-107	79	6	would	would	AUX
flr-107	79	7	agree	agree	VERB
flr-107	79	8	that	that	SCONJ
flr-107	79	9	only	only	ADV
flr-107	79	10	if	if	SCONJ
flr-107	79	11	the	the	DET
flr-107	79	12	model	model	NOUN
flr-107	79	13	is	be	AUX
flr-107	79	14	reasonably	reasonably	ADV
flr-107	79	15	good	good	ADJ
flr-107	79	16	(	(	PUNCT
flr-107	79	17	and	and	CCONJ
flr-107	79	18	we	we	PRON
flr-107	79	19	mean	mean	VERB
flr-107	79	20	generalization	generalization	NOUN
flr-107	79	21	,	,	PUNCT
flr-107	79	22	not	not	PART
flr-107	79	23	just	just	ADV
flr-107	79	24	fit	fit	ADJ
flr-107	79	25	to	to	ADP
flr-107	79	26	the	the	DET
flr-107	79	27	sample	sample	NOUN
flr-107	79	28	)	)	PUNCT
flr-107	79	29	,	,	PUNCT
flr-107	79	30	interpretation	interpretation	NOUN
flr-107	79	31	makes	make	VERB
flr-107	79	32	sense	sense	NOUN
flr-107	79	33	.	.	PUNCT
flr-107	80	1	edelsbrunner	edelsbrunner	NOUN
flr-107	80	2	and	and	CCONJ
flr-107	80	3	schneider	schneider	NOUN
flr-107	80	4	(	(	PUNCT
flr-107	80	5	2013	2013	NUM
flr-107	80	6	)	)	PUNCT
flr-107	80	7	indicate	indicate	VERB
flr-107	80	8	in	in	ADP
flr-107	80	9	their	their	PRON
flr-107	80	10	commentary	commentary	NOUN
flr-107	80	11	on	on	ADP
flr-107	80	12	this	this	DET
flr-107	80	13	article	article	NOUN
flr-107	80	14	that	that	SCONJ
flr-107	80	15	whenever	whenever	SCONJ
flr-107	80	16	possible	possible	ADJ
flr-107	80	17	,	,	PUNCT
flr-107	80	18	more	more	ADJ
flr-107	80	19	theory	theory	NOUN
flr-107	80	20	-	-	PUNCT
flr-107	80	21	driven	drive	VERB
flr-107	80	22	data	datum	NOUN
flr-107	80	23	modeling	modeling	NOUN
flr-107	80	24	techniques	technique	NOUN
flr-107	80	25	should	should	AUX
flr-107	80	26	be	be	AUX
flr-107	80	27	preferred	prefer	VERB
flr-107	80	28	.	.	PUNCT
flr-107	81	1	however	however	ADV
flr-107	81	2	,	,	PUNCT
flr-107	81	3	if	if	SCONJ
flr-107	81	4	we	we	PRON
flr-107	81	5	limit	limit	VERB
flr-107	81	6	ourselves	ourselves	PRON
flr-107	81	7	to	to	ADP
flr-107	81	8	the	the	DET
flr-107	81	9	models	model	NOUN
flr-107	81	10	that	that	PRON
flr-107	81	11	can	can	AUX
flr-107	81	12	be	be	AUX
flr-107	81	13	easily	easily	ADV
flr-107	81	14	interpreted	interpret	VERB
flr-107	81	15	,	,	PUNCT
flr-107	81	16	we	we	PRON
flr-107	81	17	may	may	AUX
flr-107	81	18	end	end	VERB
flr-107	81	19	up	up	ADP
flr-107	81	20	discarding	discard	VERB
flr-107	81	21	models	model	NOUN
flr-107	81	22	that	that	PRON
flr-107	81	23	truly	truly	ADV
flr-107	81	24	capture	capture	VERB
flr-107	81	25	important	important	ADJ
flr-107	81	26	regularities	regularity	NOUN
flr-107	81	27	of	of	ADP
flr-107	81	28	the	the	DET
flr-107	81	29	domain	domain	NOUN
flr-107	81	30	.	.	PUNCT
flr-107	82	1	there	there	PRON
flr-107	82	2	are	be	VERB
flr-107	82	3	two	two	NUM
flr-107	82	4	different	different	ADJ
flr-107	82	5	strategies	strategy	NOUN
flr-107	82	6	then	then	ADV
flr-107	82	7	to	to	PART
flr-107	82	8	extract	extract	VERB
flr-107	82	9	true	true	ADJ
flr-107	82	10	knowledge	knowledge	NOUN
flr-107	82	11	from	from	ADP
flr-107	82	12	the	the	DET
flr-107	82	13	world	world	NOUN
flr-107	82	14	.	.	PUNCT
flr-107	83	1	the	the	DET
flr-107	83	2	first	first	ADJ
flr-107	83	3	one	one	NOUN
flr-107	83	4	is	be	AUX
flr-107	83	5	a	a	DET
flr-107	83	6	classical	classical	ADJ
flr-107	83	7	one	one	NUM
flr-107	83	8	in	in	ADP
flr-107	83	9	which	which	PRON
flr-107	83	10	we	we	PRON
flr-107	83	11	try	try	VERB
flr-107	83	12	to	to	PART
flr-107	83	13	find	find	VERB
flr-107	83	14	a	a	DET
flr-107	83	15	well	well	ADV
flr-107	83	16	predicting	predict	VERB
flr-107	83	17	model	model	NOUN
flr-107	83	18	among	among	ADP
flr-107	83	19	the	the	DET
flr-107	83	20	easily	easily	ADV
flr-107	83	21	interpretable	interpretable	ADJ
flr-107	83	22	ones	one	NOUN
flr-107	83	23	.	.	PUNCT
flr-107	84	1	this	this	PRON
flr-107	84	2	is	be	AUX
flr-107	84	3	the	the	DET
flr-107	84	4	path	path	NOUN
flr-107	84	5	that	that	PRON
flr-107	84	6	should	should	AUX
flr-107	84	7	always	always	ADV
flr-107	84	8	be	be	AUX
flr-107	84	9	attempted	attempt	VERB
flr-107	84	10	.	.	PUNCT
flr-107	85	1	unfortunately	unfortunately	ADV
flr-107	85	2	,	,	PUNCT
flr-107	85	3	we	we	PRON
flr-107	85	4	suspect	suspect	VERB
flr-107	85	5	that	that	SCONJ
flr-107	85	6	it	it	PRON
flr-107	85	7	was	be	AUX
flr-107	85	8	not	not	PART
flr-107	85	9	seriously	seriously	ADV
flr-107	85	10	pursued	pursue	VERB
flr-107	85	11	in	in	ADP
flr-107	85	12	the	the	DET
flr-107	85	13	article	article	NOUN
flr-107	85	14	by	by	ADP
flr-107	85	15	musso	musso	PROPN
flr-107	85	16	et	et	PROPN
flr-107	85	17	al	al	PROPN
flr-107	85	18	.	.	PROPN
flr-107	86	1	(	(	PUNCT
flr-107	86	2	2013	2013	NUM
flr-107	86	3	)	)	PUNCT
flr-107	86	4	.	.	PUNCT
flr-107	87	1	it	it	PRON
flr-107	87	2	is	be	AUX
flr-107	87	3	also	also	ADV
flr-107	87	4	possible	possible	ADJ
flr-107	87	5	to	to	PART
flr-107	87	6	try	try	VERB
flr-107	87	7	to	to	PART
flr-107	87	8	build	build	VERB
flr-107	87	9	a	a	DET
flr-107	87	10	well	well	ADV
flr-107	87	11	predicting	predict	VERB
flr-107	87	12	model	model	NOUN
flr-107	87	13	,	,	PUNCT
flr-107	87	14	even	even	ADV
flr-107	87	15	if	if	SCONJ
flr-107	87	16	it	it	PRON
flr-107	87	17	is	be	AUX
flr-107	87	18	not	not	PART
flr-107	87	19	that	that	ADV
flr-107	87	20	easy	easy	ADJ
flr-107	87	21	to	to	PART
flr-107	87	22	interpret	interpret	VERB
flr-107	87	23	,	,	PUNCT
flr-107	87	24	and	and	CCONJ
flr-107	87	25	then	then	ADV
flr-107	87	26	put	put	VERB
flr-107	87	27	more	more	ADJ
flr-107	87	28	effort	effort	NOUN
flr-107	87	29	to	to	PART
flr-107	87	30	squeeze	squeeze	VERB
flr-107	87	31	out	out	ADP
flr-107	87	32	the	the	DET
flr-107	87	33	knowledge	knowledge	NOUN
flr-107	87	34	from	from	ADP
flr-107	87	35	the	the	DET
flr-107	87	36	model	model	NOUN
flr-107	87	37	.	.	PUNCT
flr-107	88	1	one	one	PRON
flr-107	88	2	could	could	AUX
flr-107	88	3	argue	argue	VERB
flr-107	88	4	that	that	SCONJ
flr-107	88	5	this	this	PRON
flr-107	88	6	is	be	AUX
flr-107	88	7	what	what	PRON
flr-107	88	8	happens	happen	VERB
flr-107	88	9	,	,	PUNCT
flr-107	88	10	when	when	SCONJ
flr-107	88	11	you	you	PRON
flr-107	88	12	ask	ask	VERB
flr-107	88	13	a	a	DET
flr-107	88	14	doctor	doctor	NOUN
flr-107	88	15	why	why	SCONJ
flr-107	88	16	she	she	PRON
flr-107	88	17	made	make	VERB
flr-107	88	18	the	the	DET
flr-107	88	19	diagnosis	diagnosis	NOUN
flr-107	88	20	she	she	PRON
flr-107	88	21	did	do	VERB
flr-107	88	22	.	.	PUNCT
flr-107	89	1	the	the	DET
flr-107	89	2	answer	answer	NOUN
flr-107	89	3	will	will	AUX
flr-107	89	4	(	(	PUNCT
flr-107	89	5	only	only	ADV
flr-107	89	6	)	)	PUNCT
flr-107	89	7	be	be	AUX
flr-107	89	8	some	some	DET
flr-107	89	9	approximation	approximation	NOUN
flr-107	89	10	of	of	ADP
flr-107	89	11	the	the	DET
flr-107	89	12	real	real	ADJ
flr-107	89	13	reason	reason	NOUN
flr-107	89	14	.	.	PUNCT
flr-107	90	1	still	still	ADV
flr-107	90	2	doctors	doctor	NOUN
flr-107	90	3	are	be	AUX
flr-107	90	4	considered	consider	VERB
flr-107	90	5	useful	useful	ADJ
flr-107	90	6	.	.	PUNCT
flr-107	91	1	p.	p.	NOUN
flr-107	91	2	nokelainen	nokelainen	PROPN
flr-107	91	3	&	&	CCONJ
flr-107	91	4	t.	t.	PROPN
flr-107	91	5	silander	silander	PROPN
flr-107	92	1	81	81	NUM
flr-107	93	1	|	|	ADV
flr-107	93	2	f	f	NOUN
flr-107	94	1	l	l	NOUN
flr-107	94	2	r	r	NOUN
flr-107	94	3	we	we	PRON
flr-107	94	4	have	have	VERB
flr-107	94	5	an	an	DET
flr-107	94	6	educated	educate	VERB
flr-107	94	7	guess	guess	NOUN
flr-107	94	8	,	,	PUNCT
flr-107	94	9	that	that	SCONJ
flr-107	94	10	such	such	DET
flr-107	94	11	a	a	DET
flr-107	94	12	procedure	procedure	NOUN
flr-107	94	13	is	be	AUX
flr-107	94	14	behind	behind	ADP
flr-107	94	15	the	the	DET
flr-107	94	16	independent	independent	ADJ
flr-107	94	17	variable	variable	ADJ
flr-107	94	18	importance	importance	NOUN
flr-107	94	19	measures	measure	NOUN
flr-107	94	20	featured	feature	VERB
flr-107	94	21	in	in	ADP
flr-107	94	22	the	the	DET
flr-107	94	23	article	article	NOUN
flr-107	94	24	.	.	PUNCT
flr-107	95	1	naturally	naturally	ADV
flr-107	95	2	,	,	PUNCT
flr-107	95	3	such	such	ADJ
flr-107	95	4	procedures	procedure	NOUN
flr-107	95	5	should	should	AUX
flr-107	95	6	be	be	AUX
flr-107	95	7	carefully	carefully	ADV
flr-107	95	8	documented	document	VERB
flr-107	95	9	in	in	ADP
flr-107	95	10	order	order	NOUN
flr-107	95	11	to	to	PART
flr-107	95	12	understand	understand	VERB
flr-107	95	13	what	what	PRON
flr-107	95	14	kind	kind	NOUN
flr-107	95	15	of	of	ADP
flr-107	95	16	information	information	NOUN
flr-107	95	17	we	we	PRON
flr-107	95	18	have	have	AUX
flr-107	95	19	managed	manage	VERB
flr-107	95	20	to	to	PART
flr-107	95	21	extract	extract	VERB
flr-107	95	22	from	from	ADP
flr-107	95	23	the	the	DET
flr-107	95	24	model	model	NOUN
flr-107	95	25	.	.	PUNCT
flr-107	96	1	the	the	DET
flr-107	96	2	article	article	NOUN
flr-107	96	3	leaves	leave	VERB
flr-107	96	4	the	the	DET
flr-107	96	5	impression	impression	NOUN
flr-107	96	6	of	of	ADP
flr-107	96	7	the	the	DET
flr-107	96	8	claim	claim	NOUN
flr-107	96	9	that	that	SCONJ
flr-107	96	10	artificial	artificial	ADJ
flr-107	96	11	neural	neural	ADJ
flr-107	96	12	networks	network	NOUN
flr-107	96	13	were	be	AUX
flr-107	96	14	somehow	somehow	ADV
flr-107	96	15	especially	especially	ADV
flr-107	96	16	good	good	ADJ
flr-107	96	17	for	for	ADP
flr-107	96	18	inferring	infer	VERB
flr-107	96	19	how	how	SCONJ
flr-107	96	20	different	different	ADJ
flr-107	96	21	complex	complex	ADJ
flr-107	96	22	patterns	pattern	NOUN
flr-107	96	23	of	of	ADP
flr-107	96	24	variables	variable	NOUN
flr-107	96	25	affect	affect	VERB
flr-107	96	26	the	the	DET
flr-107	96	27	outcome	outcome	NOUN
flr-107	96	28	.	.	PUNCT
flr-107	97	1	however	however	ADV
flr-107	97	2	,	,	PUNCT
flr-107	97	3	the	the	DET
flr-107	97	4	presented	present	VERB
flr-107	97	5	results	result	NOUN
flr-107	97	6	list	list	VERB
flr-107	97	7	only	only	ADV
flr-107	97	8	univariate	univariate	ADJ
flr-107	97	9	importance	importance	NOUN
flr-107	97	10	of	of	ADP
flr-107	97	11	variables	variable	NOUN
flr-107	97	12	.	.	PUNCT
flr-107	98	1	how	how	SCONJ
flr-107	98	2	could	could	AUX
flr-107	98	3	that	that	PRON
flr-107	98	4	possibly	possibly	ADV
flr-107	98	5	tell	tell	VERB
flr-107	98	6	us	we	PRON
flr-107	98	7	anything	anything	PRON
flr-107	98	8	relevant	relevant	ADJ
flr-107	98	9	about	about	ADP
flr-107	98	10	complex	complex	ADJ
flr-107	98	11	patterns	pattern	NOUN
flr-107	98	12	?	?	PUNCT
flr-107	99	1	neural	neural	ADJ
flr-107	99	2	networks	network	NOUN
flr-107	99	3	are	be	AUX
flr-107	99	4	by	by	ADP
flr-107	99	5	no	no	DET
flr-107	99	6	means	means	NOUN
flr-107	99	7	the	the	DET
flr-107	99	8	only	only	ADJ
flr-107	99	9	black	black	ADJ
flr-107	99	10	-	-	PUNCT
flr-107	99	11	box	box	NOUN
flr-107	99	12	models	model	NOUN
flr-107	99	13	that	that	PRON
flr-107	99	14	can	can	AUX
flr-107	99	15	be	be	AUX
flr-107	99	16	successful	successful	ADJ
flr-107	99	17	in	in	ADP
flr-107	99	18	the	the	DET
flr-107	99	19	prediction	prediction	NOUN
flr-107	99	20	.	.	PUNCT
flr-107	100	1	many	many	ADJ
flr-107	100	2	ensemble	ensemble	ADJ
flr-107	100	3	learning	learning	NOUN
flr-107	100	4	based	base	VERB
flr-107	100	5	or	or	CCONJ
flr-107	100	6	motivated	motivated	ADJ
flr-107	100	7	methods	method	NOUN
flr-107	100	8	,	,	PUNCT
flr-107	100	9	for	for	ADP
flr-107	100	10	instance	instance	NOUN
flr-107	100	11	random	random	ADJ
flr-107	100	12	decision	decision	NOUN
flr-107	100	13	forests	forest	NOUN
flr-107	100	14	(	(	PUNCT
flr-107	100	15	breiman	breiman	NOUN
flr-107	100	16	,	,	PUNCT
flr-107	100	17	2001a	2001a	NUM
flr-107	100	18	)	)	PUNCT
flr-107	100	19	and	and	CCONJ
flr-107	100	20	bayesian	bayesian	NOUN
flr-107	100	21	additive	additive	ADJ
flr-107	100	22	regression	regression	NOUN
flr-107	100	23	trees	tree	NOUN
flr-107	100	24	(	(	PUNCT
flr-107	100	25	chipman	chipman	PROPN
flr-107	100	26	,	,	PUNCT
flr-107	100	27	george	george	PROPN
flr-107	100	28	,	,	PUNCT
flr-107	100	29	&	&	CCONJ
flr-107	100	30	mcculloch	mcculloch	PROPN
flr-107	100	31	,	,	PUNCT
flr-107	100	32	2010	2010	NUM
flr-107	100	33	)	)	PUNCT
flr-107	100	34	,	,	PUNCT
flr-107	100	35	are	be	AUX
flr-107	100	36	among	among	ADP
flr-107	100	37	such	such	ADJ
flr-107	100	38	models	model	NOUN
flr-107	100	39	.	.	PUNCT
flr-107	101	1	ensemble	ensemble	ADJ
flr-107	101	2	methods	method	NOUN
flr-107	101	3	have	have	AUX
flr-107	101	4	reached	reach	VERB
flr-107	101	5	very	very	ADV
flr-107	101	6	high	high	ADJ
flr-107	101	7	classification	classification	NOUN
flr-107	101	8	accuracies	accuracy	NOUN
flr-107	101	9	by	by	ADP
flr-107	101	10	using	use	VERB
flr-107	101	11	several	several	ADJ
flr-107	101	12	(	(	PUNCT
flr-107	101	13	or	or	CCONJ
flr-107	101	14	growing	grow	VERB
flr-107	101	15	a	a	DET
flr-107	101	16	forest	forest	NOUN
flr-107	101	17	of	of	ADP
flr-107	101	18	)	)	PUNCT
flr-107	101	19	decision	decision	NOUN
flr-107	101	20	trees	tree	NOUN
flr-107	101	21	on	on	ADP
flr-107	101	22	the	the	DET
flr-107	101	23	same	same	ADJ
flr-107	101	24	data	datum	NOUN
flr-107	101	25	instead	instead	ADV
flr-107	101	26	of	of	ADP
flr-107	101	27	a	a	DET
flr-107	101	28	single	single	ADJ
flr-107	101	29	-	-	PUNCT
flr-107	101	30	tree	tree	NOUN
flr-107	101	31	predictor	predictor	NOUN
flr-107	101	32	.	.	PUNCT
flr-107	102	1	ever	ever	ADV
flr-107	102	2	increasing	increase	VERB
flr-107	102	3	data	datum	NOUN
flr-107	102	4	sizes	size	NOUN
flr-107	102	5	(	(	PUNCT
flr-107	102	6	e.g.	e.g.	ADV
flr-107	102	7	,	,	PUNCT
flr-107	102	8	massive	massive	ADJ
flr-107	102	9	open	open	ADJ
flr-107	102	10	online	online	ADJ
flr-107	102	11	courses	course	NOUN
flr-107	102	12	,	,	PUNCT
flr-107	102	13	moocs	moocs	PROPN
flr-107	102	14	,	,	PUNCT
flr-107	102	15	may	may	AUX
flr-107	102	16	have	have	VERB
flr-107	102	17	100	100	NUM
flr-107	102	18	000	000	NUM
flr-107	102	19	students	student	NOUN
flr-107	102	20	with	with	ADP
flr-107	102	21	all	all	DET
flr-107	102	22	their	their	PRON
flr-107	102	23	data	datum	NOUN
flr-107	102	24	gathered	gather	VERB
flr-107	102	25	automatically	automatically	ADV
flr-107	102	26	to	to	ADP
flr-107	102	27	the	the	DET
flr-107	102	28	digital	digital	ADJ
flr-107	102	29	form	form	NOUN
flr-107	102	30	)	)	PUNCT
flr-107	102	31	and	and	CCONJ
flr-107	102	32	increasing	increase	VERB
flr-107	102	33	computer	computer	NOUN
flr-107	102	34	power	power	NOUN
flr-107	102	35	may	may	AUX
flr-107	102	36	well	well	ADV
flr-107	102	37	shift	shift	VERB
flr-107	102	38	focus	focus	NOUN
flr-107	102	39	from	from	ADP
flr-107	102	40	small	small	ADJ
flr-107	102	41	,	,	PUNCT
flr-107	102	42	simple	simple	ADJ
flr-107	102	43	and	and	CCONJ
flr-107	102	44	understandable	understandable	ADJ
flr-107	102	45	models	model	NOUN
flr-107	102	46	,	,	PUNCT
flr-107	102	47	to	to	ADP
flr-107	102	48	the	the	DET
flr-107	102	49	big	big	ADJ
flr-107	102	50	,	,	PUNCT
flr-107	102	51	complex	complex	ADJ
flr-107	102	52	black	black	ADJ
flr-107	102	53	-	-	PUNCT
flr-107	102	54	box	box	NOUN
flr-107	102	55	models	model	NOUN
flr-107	102	56	.	.	PUNCT
flr-107	103	1	but	but	CCONJ
flr-107	103	2	hopefully	hopefully	ADV
flr-107	103	3	some	some	PRON
flr-107	103	4	of	of	ADP
flr-107	103	5	that	that	DET
flr-107	103	6	computing	compute	VERB
flr-107	103	7	power	power	NOUN
flr-107	103	8	can	can	AUX
flr-107	103	9	also	also	ADV
flr-107	103	10	be	be	AUX
flr-107	103	11	used	use	VERB
flr-107	103	12	to	to	PART
flr-107	103	13	extract	extract	VERB
flr-107	103	14	understandable	understandable	ADJ
flr-107	103	15	(	(	PUNCT
flr-107	103	16	even	even	ADV
flr-107	103	17	if	if	SCONJ
flr-107	103	18	not	not	PART
flr-107	103	19	always	always	ADV
flr-107	103	20	very	very	ADV
flr-107	103	21	close	close	ADJ
flr-107	103	22	to	to	ADP
flr-107	103	23	truth	truth	NOUN
flr-107	103	24	)	)	PUNCT
flr-107	103	25	approximations	approximation	NOUN
flr-107	103	26	of	of	ADP
flr-107	103	27	the	the	DET
flr-107	103	28	true	true	ADJ
flr-107	103	29	complex	complex	ADJ
flr-107	103	30	regularities	regularity	NOUN
flr-107	103	31	of	of	ADP
flr-107	103	32	the	the	DET
flr-107	103	33	world	world	NOUN
flr-107	103	34	.	.	PUNCT
flr-107	104	1	key	key	ADJ
flr-107	104	2	points	point	NOUN
flr-107	104	3	artificial	artificial	ADJ
flr-107	104	4	neural	neural	ADJ
flr-107	104	5	networks	network	NOUN
flr-107	104	6	(	(	PUNCT
flr-107	104	7	ann	ann	PROPN
flr-107	104	8	)	)	PUNCT
flr-107	104	9	certainly	certainly	ADV
flr-107	104	10	provide	provide	VERB
flr-107	104	11	interesting	interesting	ADJ
flr-107	104	12	modeling	modeling	NOUN
flr-107	104	13	possibilities	possibility	NOUN
flr-107	104	14	for	for	ADP
flr-107	104	15	educational	educational	ADJ
flr-107	104	16	scientists	scientist	NOUN
flr-107	104	17	,	,	PUNCT
flr-107	104	18	but	but	CCONJ
flr-107	104	19	they	they	PRON
flr-107	104	20	also	also	ADV
flr-107	104	21	set	set	VERB
flr-107	104	22	certain	certain	ADJ
flr-107	104	23	challenges	challenge	NOUN
flr-107	104	24	for	for	ADP
flr-107	104	25	the	the	DET
flr-107	104	26	design	design	NOUN
flr-107	104	27	of	of	ADP
flr-107	104	28	the	the	DET
flr-107	104	29	study	study	NOUN
flr-107	104	30	and	and	CCONJ
flr-107	104	31	interpretation	interpretation	NOUN
flr-107	104	32	of	of	ADP
flr-107	104	33	the	the	DET
flr-107	104	34	results	result	NOUN
flr-107	104	35	.	.	PUNCT
flr-107	105	1	the	the	DET
flr-107	105	2	article	article	NOUN
flr-107	105	3	shows	show	VERB
flr-107	105	4	a	a	DET
flr-107	105	5	very	very	ADV
flr-107	105	6	good	good	ADJ
flr-107	105	7	example	example	NOUN
flr-107	105	8	by	by	ADP
flr-107	105	9	comparing	compare	VERB
flr-107	105	10	the	the	DET
flr-107	105	11	results	result	NOUN
flr-107	105	12	of	of	ADP
flr-107	105	13	ann	ann	PROPN
flr-107	105	14	to	to	ADP
flr-107	105	15	a	a	DET
flr-107	105	16	conventional	conventional	ADJ
flr-107	105	17	data	datum	NOUN
flr-107	105	18	modeling	modeling	NOUN
flr-107	105	19	approach	approach	NOUN
flr-107	105	20	,	,	PUNCT
flr-107	105	21	but	but	CCONJ
flr-107	105	22	the	the	DET
flr-107	105	23	comparison	comparison	NOUN
flr-107	105	24	should	should	AUX
flr-107	105	25	have	have	AUX
flr-107	105	26	been	be	AUX
flr-107	105	27	made	make	VERB
flr-107	105	28	between	between	ADP
flr-107	105	29	two	two	NUM
flr-107	105	30	discriminative	discriminative	NOUN
flr-107	105	31	classifiers	classifier	NOUN
flr-107	105	32	.	.	PUNCT
flr-107	106	1	ensemble	ensemble	ADJ
flr-107	106	2	methods	method	NOUN
flr-107	106	3	provide	provide	VERB
flr-107	106	4	a	a	DET
flr-107	106	5	modern	modern	ADJ
flr-107	106	6	and	and	CCONJ
flr-107	106	7	powerful	powerful	ADJ
flr-107	106	8	alternative	alternative	NOUN
flr-107	106	9	to	to	ADP
flr-107	106	10	neural	neural	ADJ
flr-107	106	11	networks	network	NOUN
flr-107	106	12	as	as	SCONJ
flr-107	106	13	they	they	PRON
flr-107	106	14	use	use	VERB
flr-107	106	15	predictions	prediction	NOUN
flr-107	106	16	of	of	ADP
flr-107	106	17	several	several	ADJ
flr-107	106	18	models	model	NOUN
flr-107	106	19	built	build	VERB
flr-107	106	20	during	during	ADP
flr-107	106	21	learning	learn	VERB
flr-107	106	22	process	process	NOUN
flr-107	106	23	instead	instead	ADV
flr-107	106	24	of	of	ADP
flr-107	106	25	using	use	VERB
flr-107	106	26	a	a	DET
flr-107	106	27	single	single	ADJ
flr-107	106	28	model	model	NOUN
flr-107	106	29	.	.	PUNCT
flr-107	107	1	references	reference	NOUN
flr-107	107	2	breiman	breiman	PROPN
flr-107	107	3	,	,	PUNCT
flr-107	107	4	l.	l.	PROPN
flr-107	107	5	(	(	PUNCT
flr-107	107	6	2001a	2001a	NUM
flr-107	107	7	)	)	PUNCT
flr-107	107	8	.	.	PUNCT
flr-107	108	1	random	random	ADJ
flr-107	108	2	forests	forest	NOUN
flr-107	108	3	.	.	PUNCT
flr-107	109	1	machine	machine	NOUN
flr-107	109	2	learning	learning	NOUN
flr-107	109	3	,	,	PUNCT
flr-107	109	4	45	45	NUM
flr-107	109	5	,	,	PUNCT
flr-107	109	6	5–32	5–32	NOUN
flr-107	109	7	.	.	PUNCT
flr-107	110	1	doi:10.1023	doi:10.1023	NOUN
flr-107	110	2	/	/	SYM
flr-107	110	3	a:1010933404324	a:1010933404324	PROPN
flr-107	110	4	breiman	breiman	PROPN
flr-107	110	5	,	,	PUNCT
flr-107	110	6	l.	l.	PROPN
flr-107	110	7	(	(	PUNCT
flr-107	110	8	2001b	2001b	NUM
flr-107	110	9	)	)	PUNCT
flr-107	110	10	.	.	PUNCT
flr-107	111	1	statistical	statistical	ADJ
flr-107	111	2	modeling	modeling	NOUN
flr-107	111	3	:	:	PUNCT
flr-107	111	4	the	the	DET
flr-107	111	5	two	two	NUM
flr-107	111	6	cultures	culture	NOUN
flr-107	111	7	.	.	PUNCT
flr-107	112	1	statistical	statistical	ADJ
flr-107	112	2	science	science	NOUN
flr-107	112	3	,	,	PUNCT
flr-107	112	4	16(3	16(3	NOUN
flr-107	112	5	)	)	PUNCT
flr-107	112	6	,	,	PUNCT
flr-107	112	7	199–231	199–231	NUM
flr-107	112	8	.	.	PUNCT
flr-107	113	1	doi:10.1214	doi:10.1214	PROPN
flr-107	113	2	/	/	SYM
flr-107	113	3	ss/1009213726	ss/1009213726	PROPN
flr-107	113	4	chipman	chipman	PROPN
flr-107	113	5	,	,	PUNCT
flr-107	113	6	h.	h.	PROPN
flr-107	113	7	a.	a.	PROPN
flr-107	113	8	,	,	PUNCT
flr-107	113	9	george	george	PROPN
flr-107	113	10	,	,	PUNCT
flr-107	113	11	e.	e.	PROPN
flr-107	113	12	i.	i.	PROPN
flr-107	113	13	,	,	PUNCT
flr-107	113	14	&	&	CCONJ
flr-107	113	15	mcculloch	mcculloch	PROPN
flr-107	113	16	,	,	PUNCT
flr-107	113	17	r.	r.	PROPN
flr-107	113	18	e.	e.	PROPN
flr-107	113	19	(	(	PUNCT
flr-107	113	20	2010	2010	NUM
flr-107	113	21	)	)	PUNCT
flr-107	113	22	.	.	PUNCT
flr-107	114	1	bart	bart	PROPN
flr-107	114	2	:	:	PUNCT
flr-107	115	1	bayesian	bayesian	NOUN
flr-107	115	2	additive	additive	ADJ
flr-107	115	3	regression	regression	NOUN
flr-107	115	4	trees	tree	NOUN
flr-107	115	5	.	.	PUNCT
flr-107	116	1	the	the	DET
flr-107	116	2	annals	annal	NOUN
flr-107	116	3	of	of	ADP
flr-107	116	4	applied	applied	ADJ
flr-107	116	5	statistics	statistic	NOUN
flr-107	116	6	,	,	PUNCT
flr-107	116	7	4(1	4(1	NOUN
flr-107	116	8	)	)	PUNCT
flr-107	116	9	,	,	PUNCT
flr-107	116	10	266–298	266–298	NUM
flr-107	116	11	.	.	PUNCT
flr-107	117	1	doi:10.1214/09	doi:10.1214/09	NOUN
flr-107	117	2	-	-	PUNCT
flr-107	117	3	aoas285	aoas285	PROPN
flr-107	117	4	demšar	demšar	PROPN
flr-107	117	5	,	,	PUNCT
flr-107	117	6	j.	j.	PROPN
flr-107	117	7	(	(	PUNCT
flr-107	117	8	2006	2006	NUM
flr-107	117	9	)	)	PUNCT
flr-107	117	10	.	.	PUNCT
flr-107	118	1	statistical	statistical	ADJ
flr-107	118	2	comparison	comparison	NOUN
flr-107	118	3	of	of	ADP
flr-107	118	4	classifiers	classifier	NOUN
flr-107	118	5	over	over	ADP
flr-107	118	6	multiple	multiple	ADJ
flr-107	118	7	data	datum	NOUN
flr-107	118	8	sets	set	NOUN
flr-107	118	9	.	.	PUNCT
flr-107	119	1	journal	journal	NOUN
flr-107	119	2	of	of	ADP
flr-107	119	3	machine	machine	NOUN
flr-107	119	4	learning	learn	VERB
flr-107	119	5	research	research	NOUN
flr-107	119	6	,	,	PUNCT
flr-107	119	7	7	7	NUM
flr-107	119	8	,	,	PUNCT
flr-107	119	9	1–30	1–30	PROPN
flr-107	119	10	.	.	PUNCT
flr-107	120	1	correa	correa	PROPN
flr-107	120	2	,	,	PUNCT
flr-107	120	3	m.	m.	NOUN
flr-107	120	4	,	,	PUNCT
flr-107	120	5	bielza	bielza	NOUN
flr-107	120	6	,	,	PUNCT
flr-107	120	7	c.	c.	PROPN
flr-107	120	8	,	,	PUNCT
flr-107	120	9	&	&	CCONJ
flr-107	120	10	pamies	pamies	PROPN
flr-107	120	11	-	-	PUNCT
flr-107	120	12	teixeira	teixeira	PROPN
flr-107	120	13	,	,	PUNCT
flr-107	120	14	j.	j.	PROPN
flr-107	120	15	(	(	PUNCT
flr-107	120	16	2009	2009	NUM
flr-107	120	17	)	)	PUNCT
flr-107	120	18	.	.	PUNCT
flr-107	121	1	comparison	comparison	NOUN
flr-107	121	2	of	of	ADP
flr-107	121	3	bayesian	bayesian	NOUN
flr-107	121	4	networks	network	NOUN
flr-107	121	5	and	and	CCONJ
flr-107	121	6	artificial	artificial	ADJ
flr-107	121	7	neural	neural	ADJ
flr-107	121	8	networks	network	NOUN
flr-107	121	9	for	for	ADP
flr-107	121	10	quality	quality	NOUN
flr-107	121	11	detection	detection	NOUN
flr-107	121	12	in	in	ADP
flr-107	121	13	a	a	DET
flr-107	121	14	machining	machining	NOUN
flr-107	121	15	process	process	NOUN
flr-107	121	16	.	.	PUNCT
flr-107	122	1	expert	expert	NOUN
flr-107	122	2	systems	system	NOUN
flr-107	122	3	with	with	ADP
flr-107	122	4	applications	application	NOUN
flr-107	122	5	,	,	PUNCT
flr-107	122	6	36	36	NUM
flr-107	122	7	,	,	PUNCT
flr-107	122	8	7270	7270	NUM
flr-107	122	9	–	–	PUNCT
flr-107	122	10	7279	7279	NUM
flr-107	122	11	.	.	PUNCT
flr-107	123	1	doi:10.1016	doi:10.1016	PROPN
flr-107	123	2	/	/	SYM
flr-107	123	3	j.eswa.2008.09.024	j.eswa.2008.09.024	PROPN
flr-107	123	4	lek	lek	PROPN
flr-107	123	5	,	,	PUNCT
flr-107	123	6	s.	s.	PROPN
flr-107	123	7	,	,	PUNCT
flr-107	123	8	&	&	CCONJ
flr-107	123	9	guegan	guegan	PROPN
flr-107	123	10	,	,	PUNCT
flr-107	123	11	j.	j.	PROPN
flr-107	123	12	f.	f.	PROPN
flr-107	123	13	(	(	PUNCT
flr-107	123	14	1999	1999	NUM
flr-107	123	15	)	)	PUNCT
flr-107	123	16	.	.	PUNCT
flr-107	124	1	artificial	artificial	ADJ
flr-107	124	2	neural	neural	ADJ
flr-107	124	3	networks	network	NOUN
flr-107	124	4	as	as	ADP
flr-107	124	5	a	a	DET
flr-107	124	6	tool	tool	NOUN
flr-107	124	7	in	in	ADP
flr-107	124	8	ecological	ecological	ADJ
flr-107	124	9	modelling	modelling	NOUN
flr-107	124	10	,	,	PUNCT
flr-107	124	11	an	an	DET
flr-107	124	12	introduction	introduction	NOUN
flr-107	124	13	.	.	PUNCT
flr-107	125	1	ecological	ecological	ADJ
flr-107	125	2	modelling	modelling	NOUN
flr-107	125	3	,	,	PUNCT
flr-107	125	4	120	120	NUM
flr-107	125	5	,	,	PUNCT
flr-107	125	6	65–73	65–73	NUM
flr-107	125	7	.	.	PUNCT
flr-107	126	1	doi:10.1016	doi:10.1016	PROPN
flr-107	126	2	/	/	SYM
flr-107	126	3	s0304	s0304	PROPN
flr-107	126	4	-	-	PUNCT
flr-107	126	5	3800(99)00092	3800(99)00092	NUM
flr-107	126	6	-	-	PUNCT
flr-107	126	7	7	7	NUM
flr-107	126	8	meier	meier	PROPN
flr-107	126	9	,	,	PUNCT
flr-107	126	10	l.	l.	PROPN
flr-107	126	11	,	,	PUNCT
flr-107	126	12	van	van	PROPN
flr-107	126	13	de	de	PROPN
flr-107	126	14	geer	geer	PROPN
flr-107	126	15	,	,	PUNCT
flr-107	126	16	s.	s.	PROPN
flr-107	126	17	,	,	PUNCT
flr-107	126	18	&	&	CCONJ
flr-107	126	19	bühlmann	bühlmann	PROPN
flr-107	126	20	,	,	PUNCT
flr-107	126	21	p.	p.	NOUN
flr-107	126	22	(	(	PUNCT
flr-107	126	23	2008	2008	NUM
flr-107	126	24	)	)	PUNCT
flr-107	126	25	.	.	PUNCT
flr-107	127	1	the	the	DET
flr-107	127	2	group	group	NOUN
flr-107	127	3	lasso	lasso	NOUN
flr-107	127	4	for	for	ADP
flr-107	127	5	logistic	logistic	ADJ
flr-107	127	6	regression	regression	NOUN
flr-107	127	7	.	.	PUNCT
flr-107	128	1	journal	journal	NOUN
flr-107	128	2	of	of	ADP
flr-107	128	3	the	the	DET
flr-107	128	4	royal	royal	ADJ
flr-107	128	5	statistical	statistical	ADJ
flr-107	128	6	society	society	NOUN
flr-107	128	7	:	:	PUNCT
flr-107	128	8	series	series	PROPN
flr-107	128	9	b	b	PROPN
flr-107	128	10	,	,	PUNCT
flr-107	128	11	70(part	70(part	NUM
flr-107	128	12	1	1	NUM
flr-107	128	13	)	)	PUNCT
flr-107	128	14	,	,	PUNCT
flr-107	128	15	53	53	NUM
flr-107	128	16	-	-	SYM
flr-107	128	17	71	71	NUM
flr-107	128	18	.	.	PUNCT
flr-107	129	1	doi:10.1111	doi:10.1111	ADJ
flr-107	129	2	/	/	SYM
flr-107	129	3	j.1467	j.1467	PROPN
flr-107	129	4	-	-	PUNCT
flr-107	129	5	9868.2007.00627.x	9868.2007.00627.x	NUM
flr-107	129	6	musso	musso	NOUN
flr-107	129	7	,	,	PUNCT
flr-107	129	8	m.	m.	PROPN
flr-107	129	9	f.	f.	PROPN
flr-107	129	10	,	,	PUNCT
flr-107	129	11	kyndt	kyndt	PROPN
flr-107	129	12	,	,	PUNCT
flr-107	129	13	e.	e.	PROPN
flr-107	129	14	,	,	PUNCT
flr-107	129	15	cascallar	cascallar	PROPN
flr-107	129	16	,	,	PUNCT
flr-107	129	17	e.	e.	PROPN
flr-107	129	18	c.	c.	PROPN
flr-107	129	19	,	,	PUNCT
flr-107	129	20	&	&	CCONJ
flr-107	129	21	dochy	dochy	PROPN
flr-107	129	22	,	,	PUNCT
flr-107	129	23	f.	f.	PROPN
flr-107	129	24	(	(	PUNCT
flr-107	129	25	2013	2013	NUM
flr-107	129	26	)	)	PUNCT
flr-107	129	27	.	.	PUNCT
flr-107	130	1	predicting	predict	VERB
flr-107	130	2	general	general	ADJ
flr-107	130	3	academic	academic	ADJ
flr-107	130	4	performance	performance	NOUN
flr-107	130	5	and	and	CCONJ
flr-107	130	6	identifying	identify	VERB
flr-107	130	7	differential	differential	ADJ
flr-107	130	8	contribution	contribution	NOUN
flr-107	130	9	of	of	ADP
flr-107	130	10	participating	participate	VERB
flr-107	130	11	variables	variable	NOUN
flr-107	130	12	using	use	VERB
flr-107	130	13	artificial	artificial	ADJ
flr-107	130	14	neural	neural	ADJ
flr-107	130	15	networks	network	NOUN
flr-107	130	16	.	.	PUNCT
flr-107	131	1	frontline	frontline	NOUN
flr-107	131	2	learning	learn	VERB
flr-107	131	3	research	research	NOUN
flr-107	131	4	,	,	PUNCT
flr-107	131	5	1	1	NUM
flr-107	131	6	,	,	PUNCT
flr-107	131	7	42	42	NUM
flr-107	131	8	-	-	SYM
flr-107	131	9	71	71	NUM
flr-107	131	10	.	.	PUNCT
flr-107	131	11	doi:10.14786	doi:10.14786	NOUN
flr-107	131	12	/	/	SYM
flr-107	131	13	flr.v1i1.13	flr.v1i1.13	PROPN
flr-107	131	14	p.	p.	PROPN
flr-107	131	15	nokelainen	nokelainen	PROPN
flr-107	131	16	&	&	CCONJ
flr-107	131	17	t.	t.	PROPN
flr-107	131	18	silander	silander	PROPN
flr-107	132	1	82	82	NUM
flr-107	133	1	|	|	NOUN
flr-107	133	2	f	f	NOUN
flr-107	133	3	l	l	NOUN
flr-107	133	4	r	r	NOUN
flr-107	133	5	nokelainen	nokelainen	NOUN
flr-107	133	6	,	,	PUNCT
flr-107	133	7	p.	p.	NOUN
flr-107	133	8	,	,	PUNCT
flr-107	133	9	silander	silander	NOUN
flr-107	133	10	,	,	PUNCT
flr-107	133	11	t.	t.	PROPN
flr-107	133	12	,	,	PUNCT
flr-107	133	13	ruohotie	ruohotie	NOUN
flr-107	133	14	,	,	PUNCT
flr-107	133	15	p.	p.	NOUN
flr-107	133	16	,	,	PUNCT
flr-107	133	17	&	&	CCONJ
flr-107	133	18	tirri	tirri	NOUN
flr-107	133	19	,	,	PUNCT
flr-107	133	20	h.	h.	PROPN
flr-107	133	21	(	(	PUNCT
flr-107	133	22	2007	2007	NUM
flr-107	133	23	)	)	PUNCT
flr-107	133	24	.	.	PUNCT
flr-107	134	1	investigating	investigate	VERB
flr-107	134	2	the	the	DET
flr-107	134	3	number	number	NOUN
flr-107	134	4	of	of	ADP
flr-107	134	5	non	non	ADJ
flr-107	134	6	-	-	ADJ
flr-107	134	7	linear	linear	ADJ
flr-107	134	8	and	and	CCONJ
flr-107	134	9	multi	multi	ADJ
flr-107	134	10	-	-	ADJ
flr-107	134	11	modal	modal	ADJ
flr-107	134	12	relationships	relationship	NOUN
flr-107	134	13	between	between	ADP
flr-107	134	14	observed	observe	VERB
flr-107	134	15	variables	variable	NOUN
flr-107	134	16	measuring	measure	VERB
flr-107	134	17	a	a	DET
flr-107	134	18	growth	growth	NOUN
flr-107	134	19	-	-	PUNCT
flr-107	134	20	oriented	orient	VERB
flr-107	134	21	atmosphere	atmosphere	NOUN
flr-107	134	22	.	.	PUNCT
flr-107	135	1	quality	quality	NOUN
flr-107	135	2	&	&	CCONJ
flr-107	135	3	quantity	quantity	NOUN
flr-107	135	4	,	,	PUNCT
flr-107	135	5	41(6	41(6	NOUN
flr-107	135	6	)	)	PUNCT
flr-107	135	7	,	,	PUNCT
flr-107	135	8	869	869	NUM
flr-107	135	9	-	-	SYM
flr-107	135	10	890	890	NUM
flr-107	135	11	.	.	PUNCT
flr-107	135	12	doi:10.1007	doi:10.1007	PROPN
flr-107	135	13	/	/	SYM
flr-107	135	14	s11135	s11135	NOUN
flr-107	135	15	-	-	PUNCT
flr-107	135	16	006	006	NUM
flr-107	135	17	-	-	PUNCT
flr-107	135	18	9030	9030	NUM
flr-107	135	19	-	-	PUNCT
flr-107	135	20	x	x	SYM
flr-107	135	21	nokelainen	nokelainen	NOUN
flr-107	135	22	,	,	PUNCT
flr-107	135	23	p.	p.	NOUN
flr-107	135	24	,	,	PUNCT
flr-107	135	25	&	&	CCONJ
flr-107	135	26	ruohotie	ruohotie	PROPN
flr-107	135	27	,	,	PUNCT
flr-107	135	28	p.	p.	NOUN
flr-107	135	29	(	(	PUNCT
flr-107	135	30	2009	2009	NUM
flr-107	135	31	)	)	PUNCT
flr-107	135	32	.	.	PUNCT
flr-107	136	1	non	non	ADJ
flr-107	136	2	-	-	ADJ
flr-107	136	3	linear	linear	ADJ
flr-107	136	4	modeling	modeling	NOUN
flr-107	136	5	of	of	ADP
flr-107	136	6	growth	growth	NOUN
flr-107	136	7	prerequisites	prerequisite	NOUN
flr-107	136	8	in	in	ADP
flr-107	136	9	a	a	DET
flr-107	136	10	finnish	finnish	ADJ
flr-107	136	11	polytechnic	polytechnic	ADJ
flr-107	136	12	institution	institution	NOUN
flr-107	136	13	of	of	ADP
flr-107	136	14	higher	high	ADJ
flr-107	136	15	education	education	NOUN
flr-107	136	16	.	.	PUNCT
flr-107	137	1	journal	journal	NOUN
flr-107	137	2	of	of	ADP
flr-107	137	3	workplace	workplace	NOUN
flr-107	137	4	learning	learning	NOUN
flr-107	137	5	,	,	PUNCT
flr-107	137	6	21(1	21(1	NUM
flr-107	137	7	)	)	PUNCT
flr-107	137	8	,	,	PUNCT
flr-107	137	9	36	36	NUM
flr-107	137	10	-	-	SYM
flr-107	137	11	57	57	NUM
flr-107	137	12	.	.	PUNCT
flr-107	138	1	doi:10.1108/13665620910924907	doi:10.1108/13665620910924907	PROPN
flr-107	139	1	nokelainen	nokelainen	NOUN
flr-107	139	2	,	,	PUNCT
flr-107	139	3	p.	p.	NOUN
flr-107	139	4	,	,	PUNCT
flr-107	139	5	tirri	tirri	NOUN
flr-107	139	6	,	,	PUNCT
flr-107	139	7	k.	k.	PROPN
flr-107	139	8	,	,	PUNCT
flr-107	139	9	campbell	campbell	PROPN
flr-107	139	10	,	,	PUNCT
flr-107	139	11	j.	j.	PROPN
flr-107	139	12	r.	r.	PROPN
flr-107	139	13	,	,	PUNCT
flr-107	139	14	&	&	CCONJ
flr-107	139	15	walberg	walberg	PROPN
flr-107	139	16	,	,	PUNCT
flr-107	139	17	h.	h.	PROPN
flr-107	139	18	(	(	PUNCT
flr-107	139	19	2007	2007	NUM
flr-107	139	20	)	)	PUNCT
flr-107	139	21	.	.	PUNCT
flr-107	140	1	factors	factor	NOUN
flr-107	140	2	that	that	PRON
flr-107	140	3	contribute	contribute	VERB
flr-107	140	4	or	or	CCONJ
flr-107	140	5	hinder	hinder	VERB
flr-107	140	6	academic	academic	ADJ
flr-107	140	7	productivity	productivity	NOUN
flr-107	140	8	:	:	PUNCT
flr-107	140	9	comparing	compare	VERB
flr-107	140	10	two	two	NUM
flr-107	140	11	groups	group	NOUN
flr-107	140	12	of	of	ADP
flr-107	140	13	most	most	ADV
flr-107	140	14	and	and	CCONJ
flr-107	140	15	least	least	ADV
flr-107	140	16	successful	successful	ADJ
flr-107	140	17	olympians	olympian	NOUN
flr-107	140	18	.	.	PUNCT
flr-107	141	1	educational	educational	ADJ
flr-107	141	2	research	research	NOUN
flr-107	141	3	and	and	CCONJ
flr-107	141	4	evaluation	evaluation	NOUN
flr-107	141	5	,	,	PUNCT
flr-107	141	6	13(6	13(6	PROPN
flr-107	141	7	)	)	PUNCT
flr-107	141	8	,	,	PUNCT
flr-107	141	9	483	483	NUM
flr-107	141	10	-	-	SYM
flr-107	141	11	500	500	NUM
flr-107	141	12	.	.	PUNCT
flr-107	141	13	doi:10.1080/13803610701785931	doi:10.1080/13803610701785931	NOUN
flr-107	141	14	schittenkopf	schittenkopf	ADV
flr-107	141	15	,	,	PUNCT
flr-107	141	16	c.	c.	NOUN
flr-107	141	17	,	,	PUNCT
flr-107	141	18	deco	deco	NOUN
flr-107	141	19	,	,	PUNCT
flr-107	141	20	g.	g.	PROPN
flr-107	141	21	,	,	PUNCT
flr-107	141	22	&	&	CCONJ
flr-107	141	23	brauer	brauer	PROPN
flr-107	141	24	,	,	PUNCT
flr-107	141	25	w.	w.	PROPN
flr-107	141	26	(	(	PUNCT
flr-107	141	27	1997	1997	NUM
flr-107	141	28	)	)	PUNCT
flr-107	141	29	.	.	PUNCT
flr-107	142	1	two	two	NUM
flr-107	142	2	strategies	strategy	NOUN
flr-107	142	3	to	to	PART
flr-107	142	4	avoid	avoid	VERB
flr-107	142	5	overfitting	overfitte	VERB
flr-107	142	6	in	in	ADP
flr-107	142	7	feedforward	feedforward	NOUN
flr-107	142	8	networks	network	NOUN
flr-107	142	9	.	.	PUNCT
flr-107	143	1	neural	neural	ADJ
flr-107	143	2	networks	network	NOUN
flr-107	143	3	,	,	PUNCT
flr-107	143	4	10(3	10(3	NUM
flr-107	143	5	)	)	PUNCT
flr-107	143	6	,	,	PUNCT
flr-107	143	7	505	505	NUM
flr-107	143	8	-	-	SYM
flr-107	143	9	516	516	NUM
flr-107	143	10	.	.	PUNCT
flr-107	144	1	doi:10.1016	doi:10.1016	PROPN
flr-107	144	2	/	/	SYM
flr-107	144	3	s0893	s0893	PROPN
flr-107	144	4	-	-	PUNCT
flr-107	144	5	6080(96)00086	6080(96)00086	NUM
flr-107	144	6	-	-	PUNCT
flr-107	144	7	x	x	NOUN
flr-107	144	8	schneider	schneider	NOUN
flr-107	144	9	,	,	PUNCT
flr-107	144	10	m.	m.	NOUN
flr-107	144	11	,	,	PUNCT
flr-107	144	12	&	&	CCONJ
flr-107	144	13	edelsbrunner	edelsbrunner	VERB
flr-107	144	14	,	,	PUNCT
flr-107	144	15	p.	p.	NOUN
flr-107	144	16	(	(	PUNCT
flr-107	144	17	2013	2013	NUM
flr-107	144	18	)	)	PUNCT
flr-107	144	19	.	.	PUNCT
flr-107	145	1	modelling	model	VERB
flr-107	145	2	for	for	ADP
flr-107	145	3	prediction	prediction	NOUN
flr-107	145	4	vs.	vs.	ADP
flr-107	145	5	modelling	modelling	NOUN
flr-107	145	6	for	for	ADP
flr-107	145	7	understanding	understanding	NOUN
flr-107	145	8	:	:	PUNCT
flr-107	145	9	commentary	commentary	NOUN
flr-107	145	10	on	on	ADP
flr-107	145	11	musso	musso	PROPN
flr-107	145	12	et	et	PROPN
flr-107	145	13	al	al	PROPN
flr-107	145	14	.	.	PROPN
flr-107	146	1	(	(	PUNCT
flr-107	146	2	2013	2013	NUM
flr-107	146	3	)	)	PUNCT
flr-107	146	4	.	.	PUNCT
flr-107	147	1	frontline	frontline	NOUN
flr-107	147	2	learning	learn	VERB
flr-107	147	3	research	research	NOUN
flr-107	147	4	,	,	PUNCT
flr-107	147	5	1(2	1(2	NUM
flr-107	147	6	)	)	PUNCT
flr-107	147	7	,	,	PUNCT
flr-107	147	8	99	99	NUM
flr-107	147	9	-	-	SYM
flr-107	147	10	101	101	NUM
flr-107	147	11	.	.	PUNCT
flr-107	147	12	doi:10.14786	doi:10.14786	NOUN
flr-107	147	13	/	/	SYM
flr-107	147	14	flr.v1i2.74	flr.v1i2.74	PROPN
flr-107	147	15	tirri	tirri	NOUN
flr-107	147	16	,	,	PUNCT
flr-107	147	17	k.	k.	PROPN
flr-107	147	18	,	,	PUNCT
flr-107	147	19	nokelainen	nokelainen	NOUN
flr-107	147	20	,	,	PUNCT
flr-107	147	21	p.	p.	NOUN
flr-107	147	22	,	,	PUNCT
flr-107	147	23	&	&	CCONJ
flr-107	147	24	komulainen	komulainen	PROPN
flr-107	147	25	,	,	PUNCT
flr-107	147	26	e.	e.	PROPN
flr-107	147	27	(	(	PUNCT
flr-107	147	28	2013	2013	NUM
flr-107	147	29	)	)	PUNCT
flr-107	147	30	.	.	PUNCT
flr-107	148	1	multiple	multiple	ADJ
flr-107	148	2	intelligences	intelligence	NOUN
flr-107	148	3	:	:	PUNCT
flr-107	148	4	can	can	AUX
flr-107	148	5	they	they	PRON
flr-107	148	6	be	be	AUX
flr-107	148	7	measured	measure	VERB
flr-107	148	8	?	?	PUNCT
flr-107	149	1	psychological	psychological	ADJ
flr-107	149	2	test	test	NOUN
flr-107	149	3	and	and	CCONJ
flr-107	149	4	assessment	assessment	NOUN
flr-107	149	5	modeling	modeling	NOUN
flr-107	149	6	,	,	PUNCT
flr-107	149	7	55(4	55(4	NUM
flr-107	149	8	)	)	PUNCT
flr-107	149	9	,	,	PUNCT
flr-107	149	10	438	438	NUM
flr-107	149	11	-	-	SYM
flr-107	149	12	461	461	NUM
flr-107	149	13	.	.	PUNCT
flr-107	150	1	doi:10.1007/978	doi:10.1007/978	NOUN
flr-107	150	2	-	-	PUNCT
flr-107	150	3	94	94	NUM
flr-107	150	4	-	-	PUNCT
flr-107	150	5	6091	6091	NUM
flr-107	150	6	-	-	PUNCT
flr-107	150	7	758	758	NUM
flr-107	150	8	-	-	SYM
flr-107	150	9	5_1	5_1	NUM
flr-107	150	10	villaverde	villaverde	NOUN
flr-107	150	11	,	,	PUNCT
flr-107	150	12	j.	j.	PROPN
flr-107	150	13	e.	e.	PROPN
flr-107	150	14	,	,	PUNCT
flr-107	150	15	godoy	godoy	PROPN
flr-107	150	16	,	,	PUNCT
flr-107	150	17	d.	d.	PROPN
flr-107	150	18	,	,	PUNCT
flr-107	150	19	&	&	CCONJ
flr-107	150	20	amandi	amandi	PROPN
flr-107	150	21	,	,	PUNCT
flr-107	150	22	a.	a.	NOUN
flr-107	150	23	(	(	PUNCT
flr-107	150	24	2006	2006	NUM
flr-107	150	25	)	)	PUNCT
flr-107	150	26	.	.	PUNCT
flr-107	151	1	learning	learn	VERB
flr-107	151	2	styles	style	NOUN
flr-107	151	3	’	'	PUNCT
flr-107	151	4	recognition	recognition	NOUN
flr-107	151	5	in	in	ADP
flr-107	151	6	e	e	NOUN
flr-107	151	7	-	-	ADJ
flr-107	151	8	learning	learn	VERB
flr-107	151	9	environments	environment	NOUN
flr-107	151	10	with	with	ADP
flr-107	151	11	feed	feed	NOUN
flr-107	151	12	-	-	PUNCT
flr-107	151	13	forward	forward	ADV
flr-107	151	14	neural	neural	ADJ
flr-107	151	15	networks	network	NOUN
flr-107	151	16	.	.	PUNCT
flr-107	152	1	journal	journal	PROPN
flr-107	152	2	of	of	ADP
flr-107	152	3	computer	computer	NOUN
flr-107	152	4	assisted	assist	VERB
flr-107	152	5	learning	learning	NOUN
flr-107	152	6	,	,	PUNCT
flr-107	152	7	22	22	NUM
flr-107	152	8	,	,	PUNCT
flr-107	152	9	197–206	197–206	NUM
flr-107	152	10	.	.	PUNCT
flr-107	153	1	doi:10.1111	doi:10.1111	ADJ
flr-107	153	2	/	/	SYM
flr-107	153	3	j.1365	j.1365	NOUN
flr-107	153	4	-	-	PUNCT
flr-107	153	5	2729.2006.00169.x	2729.2006.00169.x	NUM
flr-107	153	6	xue	xue	PROPN
flr-107	153	7	,	,	PUNCT
flr-107	153	8	j	j	PROPN
flr-107	153	9	-	-	PROPN
flr-107	153	10	h.	h.	PROPN
flr-107	153	11	,	,	PUNCT
flr-107	153	12	&	&	CCONJ
flr-107	153	13	titterington	titterington	PROPN
flr-107	153	14	,	,	PUNCT
flr-107	153	15	d.	d.	PROPN
flr-107	153	16	m.	m.	PROPN
flr-107	153	17	(	(	PUNCT
flr-107	153	18	2008	2008	NUM
flr-107	153	19	)	)	PUNCT
flr-107	153	20	.	.	PUNCT
flr-107	154	1	comment	comment	NOUN
flr-107	154	2	on	on	ADP
flr-107	154	3	“	"	PUNCT
flr-107	154	4	on	on	ADP
flr-107	154	5	discriminative	discriminative	NOUN
flr-107	154	6	vs.	vs.	ADP
flr-107	154	7	generative	generative	ADJ
flr-107	154	8	classifiers	classifier	NOUN
flr-107	154	9	:	:	PUNCT
flr-107	154	10	a	a	DET
flr-107	154	11	comparison	comparison	NOUN
flr-107	154	12	of	of	ADP
flr-107	154	13	logistic	logistic	ADJ
flr-107	154	14	regression	regression	NOUN
flr-107	154	15	and	and	CCONJ
flr-107	154	16	naive	naive	ADJ
flr-107	154	17	bayes	baye	NOUN
flr-107	154	18	”	"	PUNCT
flr-107	154	19	.	.	PUNCT
flr-107	155	1	neural	neural	ADJ
flr-107	155	2	processing	processing	NOUN
flr-107	155	3	letters	letter	NOUN
flr-107	155	4	,	,	PUNCT
flr-107	155	5	28(3	28(3	NUM
flr-107	155	6	)	)	PUNCT
flr-107	155	7	,	,	PUNCT
flr-107	155	8	169	169	NUM
flr-107	155	9	-	-	SYM
flr-107	155	10	187	187	NUM
flr-107	155	11	.	.	PUNCT
flr-107	156	1	doi:10.1007	doi:10.1007	VERB
flr-107	156	2	/	/	SYM
flr-107	156	3	s11063	s11063	NOUN
flr-107	156	4	-	-	PUNCT
flr-107	156	5	008	008	NUM
flr-107	156	6	-	-	PUNCT
flr-107	156	7	9088	9088	NUM
flr-107	156	8	-	-	SYM
flr-107	156	9	7	7	NUM