Enfoque	76	72	106	80	595	842	1
UTE,	108	72	127	80	595	842	1
V.9-N.4,	129	72	158	80	595	842	1
Dec.2018,	160	72	197	80	595	842	1
pp.	199	72	210	80	595	842	1
168	213	72	226	80	595	842	1
-	228	72	231	80	595	842	1
179	233	72	246	80	595	842	1
http://ingenieria.ute.edu.ec/enfoqueute/	76	82	215	89	595	842	1
e-ISSN:	76	91	104	99	595	842	1
1390‐6542	107	91	145	99	595	842	1
/	147	91	149	99	595	842	1
p-ISSN:	152	91	180	99	595	842	1
1390-9363	182	91	220	99	595	842	1
Recibido	388	72	420	80	595	842	1
(Received):	422	72	463	80	595	842	1
2018/09/09	465	72	505	80	595	842	1
Aceptado	386	82	420	89	595	842	1
(Accepted):	422	82	463	89	595	842	1
2018/11/29	465	82	505	89	595	842	1
CC	467	91	479	98	595	842	1
BY	481	91	492	98	595	842	1
4.0	494	91	505	98	595	842	1
Descubriendo	75	111	169	124	595	842	1
patrones	172	111	231	124	595	842	1
de	235	111	252	124	595	842	1
comportamiento	256	111	365	124	595	842	1
entre	369	111	403	124	595	842	1
contaminantes	407	111	506	124	595	842	1
del	156	127	176	140	595	842	1
aire:	180	127	209	140	595	842	1
Un	213	127	232	140	595	842	1
enfoque	236	127	290	140	595	842	1
de	294	127	310	140	595	842	1
minería	314	127	364	140	595	842	1
de	368	127	384	140	595	842	1
datos	388	127	425	140	595	842	1
(Discovering	85	156	170	169	595	842	1
behavioral	174	156	244	169	595	842	1
patterns	248	156	303	169	595	842	1
among	307	156	353	169	595	842	1
air	357	156	374	169	595	842	1
pollutants:	378	156	449	169	595	842	1
A	453	156	463	169	595	842	1
data	467	156	496	169	595	842	1
mining	232	172	278	185	595	842	1
approach)	282	172	350	185	595	842	1
Diana	75	200	104	210	595	842	1
Arce	107	200	129	210	595	842	1
1	129	199	133	205	595	842	1
,	133	200	136	210	595	842	1
Fernando	139	200	187	210	595	842	1
Lima	189	200	213	210	595	842	1
1	213	199	217	205	595	842	1
,	217	200	220	210	595	842	1
Marcos	223	200	259	210	595	842	1
Orellana	262	200	304	210	595	842	1
1	304	199	308	205	595	842	1
,	308	200	311	210	595	842	1
John	314	200	338	210	595	842	1
Ortega	341	200	375	210	595	842	1
1	375	199	379	205	595	842	1
,	379	200	382	210	595	842	1
Chester	385	200	423	210	595	842	1
Sellers	426	200	460	210	595	842	1
1	460	199	464	205	595	842	1
,	464	200	467	210	595	842	1
Patricia	470	200	506	210	595	842	1
Ortega	272	213	306	223	595	842	1
1	306	212	309	218	595	842	1
Palabras	99	444	141	454	595	842	1
clave:	150	444	178	454	595	842	1
contaminación	187	444	251	454	595	842	1
atmosférica,	260	444	314	454	595	842	1
conocimiento,	323	444	384	454	595	842	1
minería	393	444	426	454	595	842	1
de	435	444	446	454	595	842	1
datos,	454	444	482	454	595	842	1
correlaciones.	99	456	161	465	595	842	1
Keywords:	99	663	150	672	595	842	1
air	153	663	164	672	595	842	1
pollutant;	167	663	207	672	595	842	1
knowledge;	210	663	261	672	595	842	1
data	264	663	283	672	595	842	1
mining;	286	663	318	672	595	842	1
correlation.	321	663	370	672	595	842	1
1	71	750	75	756	595	842	1
Universidad	78	750	131	759	595	842	1
del	136	750	150	759	595	842	1
Azuay	155	750	183	759	595	842	1
(UDA),	187	750	218	759	595	842	1
Cuenca,	223	750	260	759	595	842	1
Ecuador	266	750	303	759	595	842	1
(	308	750	311	759	595	842	1
{darce,	316	750	347	759	595	842	1
flima,	353	750	377	759	595	842	1
marore,	382	750	416	759	595	842	1
ua069259,	421	750	468	759	595	842	1
csellers,	474	750	510	759	595	842	1
portega}	71	762	108	771	595	842	1
@uazuay.edu.ec).	111	762	192	771	595	842	1
169	511	72	525	80	595	842	2
1.	85	87	94	97	595	842	2
Introduction	97	87	162	97	595	842	2
In	113	113	123	123	595	842	2
recent	126	113	156	123	595	842	2
years,	159	113	189	123	595	842	2
different	192	113	232	123	595	842	2
air	235	113	247	123	595	842	2
pollution	250	113	291	123	595	842	2
events	294	113	327	123	595	842	2
have	330	113	353	123	595	842	2
called	356	113	385	123	595	842	2
the	388	113	404	123	595	842	2
attention	407	113	449	123	595	842	2
of	452	113	461	123	595	842	2
experts	464	113	500	123	595	842	2
both	503	113	525	123	595	842	2
in	85	125	94	136	595	842	2
academic	97	125	145	136	595	842	2
and	148	125	167	136	595	842	2
government	170	125	229	136	595	842	2
(Li,	232	125	248	136	595	842	2
Fan,	251	125	273	136	595	842	2
and	277	125	296	136	595	842	2
Mao,	299	125	324	136	595	842	2
2016).	328	125	359	136	595	842	2
According	363	125	411	136	595	842	2
to	415	125	424	136	595	842	2
Kim,	428	125	450	136	595	842	2
Choi,	454	125	480	136	595	842	2
and	484	125	502	136	595	842	2
Kim	506	125	525	136	595	842	2
(2005)	85	138	117	148	595	842	2
in	120	138	128	148	595	842	2
mega-scale	131	138	188	148	595	842	2
metropolitan	190	138	251	148	595	842	2
cities,	253	138	281	148	595	842	2
the	284	138	299	148	595	842	2
ozone	302	138	332	148	595	842	2
pollution,	334	138	378	148	595	842	2
has	381	138	399	148	595	842	2
often	401	138	426	148	595	842	2
been	428	138	453	148	595	842	2
treated	455	138	490	148	595	842	2
as	492	138	504	148	595	842	2
one	506	138	524	148	595	842	2
of	85	151	94	161	595	842	2
the	98	151	114	161	595	842	2
most	118	151	142	161	595	842	2
serious	146	151	181	161	595	842	2
socio-economic	185	151	262	161	595	842	2
issues	266	151	297	161	595	842	2
due	301	151	319	161	595	842	2
to	323	151	332	161	595	842	2
the	336	151	352	161	595	842	2
frequent	356	151	396	161	595	842	2
warnings	400	151	444	161	595	842	2
associated	448	151	501	161	595	842	2
with	505	151	524	161	595	842	2
ozone	85	163	115	173	595	842	2
exceedance	120	163	179	173	595	842	2
episodes.	185	163	232	173	595	842	2
Air	237	163	250	173	595	842	2
pollution	256	163	296	173	595	842	2
is	302	163	310	173	595	842	2
any	315	163	332	173	595	842	2
substance	337	163	388	173	595	842	2
which	393	163	421	173	595	842	2
may	426	163	447	173	595	842	2
harm	452	163	477	173	595	842	2
humans,	482	163	524	173	595	842	2
animals,	85	176	126	186	595	842	2
vegetation	131	176	182	186	595	842	2
or	187	176	197	186	595	842	2
material	201	176	240	186	595	842	2
(Kampa	246	176	284	186	595	842	2
and	289	176	307	186	595	842	2
Castanas,	312	176	362	186	595	842	2
2008).	367	176	398	186	595	842	2
Air	403	176	416	186	595	842	2
pollutant	421	176	463	186	595	842	2
levels	468	176	496	186	595	842	2
even	501	176	524	186	595	842	2
below	85	189	114	199	595	842	2
standard	120	189	163	199	595	842	2
concentrations	170	189	241	199	595	842	2
are	248	189	263	199	595	842	2
known	270	189	302	199	595	842	2
to	308	189	317	199	595	842	2
affect	324	189	350	199	595	842	2
human	357	189	391	199	595	842	2
health,	397	189	430	199	595	842	2
with	436	189	456	199	595	842	2
increases	462	189	510	199	595	842	2
in	516	189	525	199	595	842	2
respiratory	85	201	137	211	595	842	2
symptoms,	142	201	195	211	595	842	2
chronic	200	201	236	211	595	842	2
cough,	241	201	274	211	595	842	2
bronchitis	279	201	326	211	595	842	2
and	331	201	349	211	595	842	2
chest	354	201	381	211	595	842	2
illness,	386	201	419	211	595	842	2
also	424	201	445	211	595	842	2
deterioration	450	201	511	211	595	842	2
in	516	201	524	211	595	842	2
pulmonary	85	214	137	224	595	842	2
function	142	214	180	224	595	842	2
(Fukuda,	186	214	229	224	595	842	2
2007).	235	214	266	224	595	842	2
According	271	214	320	224	595	842	2
to	325	214	334	224	595	842	2
Katz	340	214	362	224	595	842	2
(1970),	367	214	402	224	595	842	2
a	407	214	413	224	595	842	2
significant	419	214	467	224	595	842	2
number	473	214	510	224	595	842	2
of	515	214	524	224	595	842	2
industrial	85	226	129	237	595	842	2
activities	133	226	175	237	595	842	2
contribute	179	226	227	237	595	842	2
thousands	231	226	281	237	595	842	2
of	285	226	294	237	595	842	2
pollutants	298	226	345	237	595	842	2
to	349	226	358	237	595	842	2
the	362	226	377	237	595	842	2
atmosphere.	381	226	442	237	595	842	2
For	446	226	462	237	595	842	2
that	466	226	484	237	595	842	2
reason,	488	226	524	237	595	842	2
environmental	85	239	154	249	595	842	2
managers	158	239	207	249	595	842	2
and	210	239	229	249	595	842	2
urban	232	239	261	249	595	842	2
planners	264	239	306	249	595	842	2
seek	310	239	333	249	595	842	2
monitoring	337	239	388	249	595	842	2
and	392	239	410	249	595	842	2
controlling	414	239	464	249	595	842	2
levels	468	239	496	249	595	842	2
of	499	239	508	249	595	842	2
air	512	239	524	249	595	842	2
pollution	85	252	126	262	595	842	2
to	129	252	138	262	595	842	2
make	141	252	168	262	595	842	2
the	171	252	187	262	595	842	2
decisions	190	252	235	262	595	842	2
which	239	252	267	262	595	842	2
allow	270	252	295	262	595	842	2
mitigating	298	252	345	262	595	842	2
health	348	252	378	262	595	842	2
and	381	252	399	262	595	842	2
the	402	252	418	262	595	842	2
environmental	421	252	490	262	595	842	2
risks.	493	252	519	262	595	842	2
Air	113	264	127	275	595	842	2
pollution	135	264	176	275	595	842	2
is	184	264	192	275	595	842	2
monitored	200	264	249	275	595	842	2
by	257	264	269	275	595	842	2
stations,	277	264	318	275	595	842	2
which	326	264	354	275	595	842	2
generate	362	264	406	275	595	842	2
data	414	264	435	275	595	842	2
about	443	264	471	275	595	842	2
levels	479	264	507	275	595	842	2
of	515	264	524	275	595	842	2
concentration	85	277	151	287	595	842	2
of	154	277	163	287	595	842	2
air	167	277	179	287	595	842	2
pollutants	182	277	229	287	595	842	2
in	232	277	240	287	595	842	2
a	243	277	250	287	595	842	2
specific	253	277	289	287	595	842	2
zone.	292	277	319	287	595	842	2
Data	322	277	346	287	595	842	2
that	349	277	367	287	595	842	2
is	370	277	378	287	595	842	2
essential	381	277	425	287	595	842	2
to	428	277	437	287	595	842	2
properly	440	277	480	287	595	842	2
evaluate	483	277	524	287	595	842	2
air	85	290	97	300	595	842	2
quality.	102	290	137	300	595	842	2
However,	143	290	189	300	595	842	2
many	194	290	221	300	595	842	2
of	226	290	235	300	595	842	2
these	240	290	267	300	595	842	2
pollutants	273	290	320	300	595	842	2
react	325	290	349	300	595	842	2
chemically	354	290	406	300	595	842	2
or	411	290	421	300	595	842	2
photo	426	290	453	300	595	842	2
chemically	459	290	510	300	595	842	2
to	515	290	524	300	595	842	2
produce	85	302	125	313	595	842	2
new	129	302	149	313	595	842	2
reactants	152	302	197	313	595	842	2
(Katz,	201	302	230	313	595	842	2
1970).	233	302	264	313	595	842	2
Therefore,	268	302	319	313	595	842	2
information	322	302	377	313	595	842	2
about	380	302	408	313	595	842	2
associations	411	302	472	313	595	842	2
among	475	302	509	313	595	842	2
air	512	302	524	313	595	842	2
pollutants,	85	315	135	325	595	842	2
it	139	315	144	325	595	842	2
is	148	315	156	325	595	842	2
of	159	315	169	325	595	842	2
great	172	315	197	325	595	842	2
importance	201	315	255	325	595	842	2
to	259	315	268	325	595	842	2
support	271	315	308	325	595	842	2
such	312	315	335	325	595	842	2
evaluation.	339	315	392	325	595	842	2
In	395	315	405	325	595	842	2
this	408	315	425	325	595	842	2
context,	429	315	467	325	595	842	2
Cagliero	471	315	512	325	595	842	2
et	515	315	525	325	595	842	2
al.,	85	328	100	338	595	842	2
(2016)	103	328	135	338	595	842	2
claim	138	328	164	338	595	842	2
that	167	328	185	338	595	842	2
the	189	328	204	338	595	842	2
combinations	207	328	272	338	595	842	2
of	276	328	285	338	595	842	2
pollutants	288	328	335	338	595	842	2
that	338	328	357	338	595	842	2
simultaneously	360	328	433	338	595	842	2
exceed	436	328	471	338	595	842	2
the	475	328	490	338	595	842	2
critical	493	328	524	338	595	842	2
level,	85	340	111	351	595	842	2
in	114	340	122	351	595	842	2
most	125	340	149	351	595	842	2
cases	152	340	181	351	595	842	2
are	184	340	200	351	595	842	2
particularly	203	340	256	351	595	842	2
interesting.	259	340	313	351	595	842	2
Efficient	113	353	153	363	595	842	2
air	160	353	172	363	595	842	2
pollution	180	353	221	363	595	842	2
evaluation	229	353	279	363	595	842	2
requires	286	353	326	363	595	842	2
well-grounded	334	353	403	363	595	842	2
knowledge	410	353	463	363	595	842	2
concerning	471	353	524	363	595	842	2
emission	85	366	129	376	595	842	2
sources,	131	366	173	376	595	842	2
associations	176	366	236	376	595	842	2
among	239	366	273	376	595	842	2
air	275	366	287	376	595	842	2
pollutants	290	366	337	376	595	842	2
and	340	366	358	376	595	842	2
their	361	366	382	376	595	842	2
effect	385	366	412	376	595	842	2
on	415	366	427	376	595	842	2
air	430	366	442	376	595	842	2
quality	445	366	476	376	595	842	2
(Wagner,	479	366	525	376	595	842	2
1994).	85	378	116	388	595	842	2
For	120	378	137	388	595	842	2
this	141	378	158	388	595	842	2
reason,	162	378	199	388	595	842	2
countries	203	378	248	388	595	842	2
from	252	378	274	388	595	842	2
the	278	378	293	388	595	842	2
Andean	297	378	335	388	595	842	2
region	339	378	370	388	595	842	2
as	374	378	386	388	595	842	2
Ecuador	390	378	431	388	595	842	2
search	435	378	468	388	595	842	2
to	472	378	481	388	595	842	2
improve	485	378	525	388	595	842	2
their	85	391	107	401	595	842	2
process	110	391	148	401	595	842	2
to	151	391	161	401	595	842	2
the	164	391	179	401	595	842	2
discovery	182	391	228	401	595	842	2
of	232	391	241	401	595	842	2
knowledge	244	391	297	401	595	842	2
in	300	391	308	401	595	842	2
their	312	391	333	401	595	842	2
environmental	336	391	405	401	595	842	2
context.	408	391	447	401	595	842	2
Hence,	450	391	485	401	595	842	2
there	488	391	513	401	595	842	2
is	516	391	524	401	595	842	2
a	85	404	91	414	595	842	2
need	95	404	120	414	595	842	2
for	124	404	137	414	595	842	2
effective	141	404	182	414	595	842	2
strategies	186	404	234	414	595	842	2
to	238	404	247	414	595	842	2
extract	251	404	284	414	595	842	2
knowledge	288	404	341	414	595	842	2
from	345	404	367	414	595	842	2
these	371	404	398	414	595	842	2
datasets.	402	404	447	414	595	842	2
In	451	404	460	414	595	842	2
this	464	404	482	414	595	842	2
context,	486	404	524	414	595	842	2
data	85	416	107	426	595	842	2
mining	110	416	142	426	595	842	2
is	145	416	153	426	595	842	2
a	157	416	163	426	595	842	2
useful	166	416	195	426	595	842	2
and	198	416	217	426	595	842	2
flexible	220	416	254	426	595	842	2
tool	257	416	275	426	595	842	2
for	278	416	291	426	595	842	2
knowledge	294	416	347	426	595	842	2
discovery	351	416	397	426	595	842	2
in	400	416	409	426	595	842	2
environmental	412	416	481	426	595	842	2
systems	484	416	524	426	595	842	2
(Fukuda,	85	429	129	439	595	842	2
2007).	132	429	163	439	595	842	2
Thus,	167	429	194	439	595	842	2
we	198	429	211	439	595	842	2
raise	215	429	239	439	595	842	2
the	242	429	258	439	595	842	2
following	261	429	303	439	595	842	2
problem:	307	429	350	439	595	842	2
how	353	429	374	439	595	842	2
to	377	429	386	439	595	842	2
use	389	429	407	439	595	842	2
data	411	429	432	439	595	842	2
mining	435	429	468	439	595	842	2
techniques	471	429	524	439	595	842	2
to	85	442	94	452	595	842	2
identify	98	442	133	452	595	842	2
associations	137	442	198	452	595	842	2
patterns	202	442	242	452	595	842	2
among	246	442	279	452	595	842	2
common	283	442	325	452	595	842	2
air	329	442	342	452	595	842	2
pollutants	346	442	393	452	595	842	2
in	397	442	406	452	595	842	2
the	409	442	425	452	595	842	2
Andean	429	442	467	452	595	842	2
region?	471	442	507	452	595	842	2
To	511	442	524	452	595	842	2
address	85	454	124	464	595	842	2
this	128	454	145	464	595	842	2
problem,	149	454	192	464	595	842	2
we	195	454	209	464	595	842	2
developed	213	454	264	464	595	842	2
an	267	454	280	464	595	842	2
approach	283	454	329	464	595	842	2
for	333	454	346	464	595	842	2
discovery	350	454	396	464	595	842	2
of	400	454	409	464	595	842	2
knowledge	413	454	465	464	595	842	2
based	469	454	499	464	595	842	2
on	503	454	515	464	595	842	2
a	518	454	524	464	595	842	2
Cross-industry	85	467	156	477	595	842	2
Standard	159	467	203	477	595	842	2
Process	206	467	246	477	595	842	2
for	248	467	261	477	595	842	2
Data	264	467	287	477	595	842	2
Mining	290	467	322	477	595	842	2
(CRISP-DM)	325	467	386	477	595	842	2
(Wirth,	389	467	421	477	595	842	2
2000).	424	467	455	477	595	842	2
Unlike	458	467	488	477	595	842	2
related	491	467	525	477	595	842	2
jobs	85	479	105	490	595	842	2
(Cagliero	110	479	154	490	595	842	2
et	159	479	168	490	595	842	2
al.,	171	479	186	490	595	842	2
2016;	190	479	217	490	595	842	2
Doreswamy	222	479	280	490	595	842	2
and	284	479	302	490	595	842	2
Manjaunath,	307	479	367	490	595	842	2
2015;	372	479	399	490	595	842	2
Kingsy	403	479	436	490	595	842	2
et	441	479	450	490	595	842	2
al.,	453	479	468	490	595	842	2
2016),	472	479	503	490	595	842	2
this	507	479	525	490	595	842	2
approach	85	492	131	502	595	842	2
by	133	492	145	502	595	842	2
means	147	492	180	502	595	842	2
of	183	492	192	502	595	842	2
Time	194	492	219	502	595	842	2
Rolling	221	492	254	502	595	842	2
Correlations	257	492	316	502	595	842	2
leads	318	492	345	502	595	842	2
to	347	492	356	502	595	842	2
identify	359	492	394	502	595	842	2
behavioral	396	492	446	502	595	842	2
patterns	449	492	489	502	595	842	2
among	491	492	524	502	595	842	2
five	85	505	102	515	595	842	2
harmful	105	505	142	515	595	842	2
air	144	505	156	515	595	842	2
pollutants	159	505	206	515	595	842	2
in	209	505	217	515	595	842	2
the	220	505	235	515	595	842	2
Andean	238	505	276	515	595	842	2
region:	278	505	312	515	595	842	2
Ozone	314	505	347	515	595	842	2
(O3),	349	505	375	515	595	842	2
Carbon	377	505	413	515	595	842	2
monoxide	416	505	463	515	595	842	2
(CO),	466	505	493	515	595	842	2
Sulfur	496	505	524	515	595	842	2
dioxide	85	518	120	528	595	842	2
(SO2),	123	518	156	528	595	842	2
Nitrogen	159	518	201	528	595	842	2
dioxide	204	518	239	528	595	842	2
(NO2)	242	518	272	528	595	842	2
and,	275	518	297	528	595	842	2
Particle	300	518	337	528	595	842	2
material	340	518	379	528	595	842	2
(PM2.5).	382	518	425	528	595	842	2
Correlations	428	518	487	528	595	842	2
among	491	518	524	528	595	842	2
pollutant	85	530	127	540	595	842	2
levels	130	530	158	540	595	842	2
can	161	530	178	540	595	842	2
vary	181	530	202	540	595	842	2
over	205	530	226	540	595	842	2
time	229	530	250	540	595	842	2
and	253	530	271	540	595	842	2
space.	274	530	307	540	595	842	2
Therefore,	310	530	360	540	595	842	2
users	363	530	390	540	595	842	2
are	393	530	409	540	595	842	2
commonly	412	530	462	540	595	842	2
interested	465	530	513	540	595	842	2
in	516	530	525	540	595	842	2
monitoring	85	543	136	553	595	842	2
their	140	543	161	553	595	842	2
temporal	164	543	207	553	595	842	2
and	210	543	228	553	595	842	2
spatial	231	543	263	553	595	842	2
evolution	266	543	310	553	595	842	2
(Cagliero	313	543	358	553	595	842	2
et	361	543	370	553	595	842	2
al.,	373	543	388	553	595	842	2
2016).	391	543	422	553	595	842	2
This	113	555	134	566	595	842	2
approach	137	555	183	566	595	842	2
aims	186	555	209	566	595	842	2
to	212	555	221	566	595	842	2
support	224	555	260	566	595	842	2
environmental	263	555	332	566	595	842	2
managers	335	555	384	566	595	842	2
to	387	555	396	566	595	842	2
produce	399	555	439	566	595	842	2
knowledge	441	555	494	566	595	842	2
about	497	555	525	566	595	842	2
air	85	568	97	578	595	842	2
pollution.	101	568	145	578	595	842	2
We	149	568	166	578	595	842	2
claim	169	568	195	578	595	842	2
that	199	568	217	578	595	842	2
the	221	568	236	578	595	842	2
proposed	240	568	286	578	595	842	2
approach	290	568	335	578	595	842	2
generates	339	568	388	578	595	842	2
correlations	392	568	449	578	595	842	2
that	452	568	471	578	595	842	2
can	475	568	492	578	595	842	2
assist	496	568	524	578	595	842	2
environmental	85	581	154	591	595	842	2
managers	162	581	211	591	595	842	2
and	219	581	238	591	595	842	2
urban	246	581	274	591	595	842	2
planners	282	581	324	591	595	842	2
in	332	581	341	591	595	842	2
the	349	581	364	591	595	842	2
decision-making	372	581	452	591	595	842	2
process	460	581	498	591	595	842	2
and	506	581	525	591	595	842	2
consequently	85	593	150	604	595	842	2
improve	153	593	192	604	595	842	2
the	194	593	209	604	595	842	2
life	212	593	226	604	595	842	2
quality	229	593	260	604	595	842	2
of	263	593	272	604	595	842	2
citizens.	275	593	314	604	595	842	2
We	317	593	334	604	595	842	2
hypothesize	336	593	395	604	595	842	2
that	397	593	416	604	595	842	2
the	418	593	434	604	595	842	2
application	436	593	489	604	595	842	2
of	491	593	501	604	595	842	2
data	503	593	525	604	595	842	2
mining	85	606	118	616	595	842	2
techniques	121	606	175	616	595	842	2
helps	179	606	205	616	595	842	2
identifying	209	606	258	616	595	842	2
patterns	262	606	302	616	595	842	2
of	306	606	315	616	595	842	2
behavior	319	606	361	616	595	842	2
among	365	606	398	616	595	842	2
several	402	606	438	616	595	842	2
air	441	606	454	616	595	842	2
pollutants.	458	606	508	616	595	842	2
To	511	606	524	616	595	842	2
evaluate	85	619	127	629	595	842	2
the	129	619	145	629	595	842	2
implementation	148	619	222	629	595	842	2
of	225	619	234	629	595	842	2
the	237	619	253	629	595	842	2
proposed	255	619	301	629	595	842	2
approach,	304	619	353	629	595	842	2
we	356	619	370	629	595	842	2
use	373	619	390	629	595	842	2
a	393	619	399	629	595	842	2
real	402	619	421	629	595	842	2
air	424	619	436	629	595	842	2
pollutants	439	619	486	629	595	842	2
dataset	488	619	524	629	595	842	2
from	85	631	107	641	595	842	2
Cuenca,	111	631	152	641	595	842	2
Ecuador.	156	631	200	641	595	842	2
This	204	631	225	641	595	842	2
paper	229	631	257	641	595	842	2
is	261	631	269	641	595	842	2
organized	273	631	321	641	595	842	2
as	325	631	337	641	595	842	2
follows:	341	631	377	641	595	842	2
in	382	631	390	641	595	842	2
Section	394	631	431	641	595	842	2
2,	435	631	444	641	595	842	2
we	448	631	462	641	595	842	2
address	466	631	505	641	595	842	2
the	509	631	525	641	595	842	2
general	85	644	122	654	595	842	2
background	125	644	182	654	595	842	2
and	185	644	204	654	595	842	2
related	207	644	240	654	595	842	2
work.	244	644	270	654	595	842	2
In	273	644	282	654	595	842	2
Section	285	644	322	654	595	842	2
3,	325	644	334	654	595	842	2
we	337	644	351	654	595	842	2
detail	354	644	381	654	595	842	2
the	383	644	399	654	595	842	2
Data	402	644	425	654	595	842	2
Mining	428	644	460	654	595	842	2
approach.	464	644	512	654	595	842	2
In	515	644	525	654	595	842	2
Section	85	657	122	667	595	842	2
4,	126	657	135	667	595	842	2
we	139	657	153	667	595	842	2
describe	157	657	199	667	595	842	2
the	203	657	218	667	595	842	2
proposed	222	657	268	667	595	842	2
approach	272	657	318	667	595	842	2
experimentation.	322	657	404	667	595	842	2
Finally,	408	657	442	667	595	842	2
in	447	657	455	667	595	842	2
Section	460	657	496	667	595	842	2
5	500	657	506	667	595	842	2
we	511	657	525	667	595	842	2
present	85	669	122	680	595	842	2
conclusions	125	669	182	680	595	842	2
and	186	669	204	680	595	842	2
future	207	669	235	680	595	842	2
work.	238	669	264	680	595	842	2
2.	85	694	94	705	595	842	2
Background	97	694	162	705	595	842	2
and	165	694	184	705	595	842	2
related	187	694	223	705	595	842	2
work	226	694	252	705	595	842	2
According	100	719	148	730	595	842	2
to	151	719	161	730	595	842	2
Wirth	163	719	189	730	595	842	2
(2000),	192	719	227	730	595	842	2
data	230	719	251	730	595	842	2
mining	254	719	287	730	595	842	2
process	290	719	328	730	595	842	2
needs	331	719	361	730	595	842	2
a	364	719	370	730	595	842	2
standard	373	719	416	730	595	842	2
approach	418	719	464	730	595	842	2
which	467	719	495	730	595	842	2
helps	498	719	524	730	595	842	2
to	85	731	94	742	595	842	2
translate	104	731	146	742	595	842	2
business	155	731	198	742	595	842	2
problems	208	731	253	742	595	842	2
in	262	731	271	742	595	842	2
data	280	731	301	742	595	842	2
mining	311	731	343	742	595	842	2
tasks,	352	731	381	742	595	842	2
suggest	390	731	429	742	595	842	2
appropriate	438	731	494	742	595	842	2
data	503	731	524	742	595	842	2
transformations	85	743	161	754	595	842	2
and	168	743	186	754	595	842	2
data	194	743	215	754	595	842	2
mining	222	743	254	754	595	842	2
techniques	262	743	315	754	595	842	2
and,	322	743	343	754	595	842	2
provide	351	743	387	754	595	842	2
significate	394	743	443	754	595	842	2
information	450	743	504	754	595	842	2
for	511	743	524	754	595	842	2
evaluating	85	755	135	766	595	842	2
the	138	755	154	766	595	842	2
effectiveness	156	755	220	766	595	842	2
of	223	755	232	766	595	842	2
the	235	755	251	766	595	842	2
results	253	755	286	766	595	842	2
and	289	755	307	766	595	842	2
documenting	310	755	373	766	595	842	2
the	376	755	391	766	595	842	2
experience.	394	755	451	766	595	842	2
Cross-industry	454	755	525	766	595	842	2
Enfoque	220	778	250	785	595	842	2
UTE,	252	778	270	785	595	842	2
V.9-N.4,	272	778	302	785	595	842	2
Dec.2018,	304	778	340	785	595	842	2
pp.	343	778	354	785	595	842	2
168	356	778	369	785	595	842	2
-	372	778	374	785	595	842	2
179	377	778	390	785	595	842	2
170	71	72	84	80	595	842	3
standard	71	87	114	97	595	842	3
process	119	87	157	97	595	842	3
for	162	87	175	97	595	842	3
data	180	87	201	97	595	842	3
mining	206	87	239	97	595	842	3
(CRISP-DM)	244	87	306	97	595	842	3
is	311	87	318	97	595	842	3
a	323	87	330	97	595	842	3
model	334	87	365	97	595	842	3
for	369	87	382	97	595	842	3
carrying	387	87	426	97	595	842	3
out	431	87	447	97	595	842	3
data	452	87	473	97	595	842	3
mining	478	87	510	97	595	842	3
projects	71	99	110	109	595	842	3
processes	113	99	163	109	595	842	3
(Wirth,	166	99	198	109	595	842	3
2000).	201	99	232	109	595	842	3
CRISP-DM	85	111	140	121	595	842	3
is	144	111	152	121	595	842	3
composed	157	111	208	121	595	842	3
of	212	111	221	121	595	842	3
six	226	111	239	121	595	842	3
phases:	243	111	282	121	595	842	3
business	286	111	330	121	595	842	3
understanding,	334	111	407	121	595	842	3
data	412	111	433	121	595	842	3
understanding,	438	111	510	121	595	842	3
data	71	123	92	133	595	842	3
preparation,	98	123	157	133	595	842	3
modeling,	162	123	210	133	595	842	3
evaluation	215	123	266	133	595	842	3
and,	271	123	292	133	595	842	3
deployment.	298	123	358	133	595	842	3
These	363	123	394	133	595	842	3
phases	399	123	435	133	595	842	3
have	441	123	465	133	595	842	3
frequent	470	123	510	133	595	842	3
dependence	71	135	132	145	595	842	3
since	137	135	162	145	595	842	3
they	167	135	188	145	595	842	3
represent	193	135	240	145	595	842	3
the	245	135	260	145	595	842	3
life	265	135	279	145	595	842	3
cycle	284	135	309	145	595	842	3
of	314	135	323	145	595	842	3
a	328	135	335	145	595	842	3
data	340	135	361	145	595	842	3
mining	366	135	398	145	595	842	3
project	403	135	436	145	595	842	3
(Wirth,	442	135	474	145	595	842	3
2000).	479	135	510	145	595	842	3
Business	71	147	116	157	595	842	3
understanding	118	147	188	157	595	842	3
is	191	147	199	157	595	842	3
focused	201	147	240	157	595	842	3
on	243	147	255	157	595	842	3
project	258	147	291	157	595	842	3
objectives	294	147	342	157	595	842	3
and	345	147	363	157	595	842	3
requirements	366	147	430	157	595	842	3
from	433	147	455	157	595	842	3
a	458	147	464	157	595	842	3
business	467	147	510	157	595	842	3
perspective.	71	159	130	169	595	842	3
Data	136	159	159	169	595	842	3
understanding	164	159	234	169	595	842	3
is	240	159	248	169	595	842	3
focused	253	159	291	169	595	842	3
on	297	159	309	169	595	842	3
knowing	314	159	355	169	595	842	3
the	360	159	376	169	595	842	3
data	381	159	402	169	595	842	3
and	408	159	426	169	595	842	3
to	432	159	441	169	595	842	3
discover	446	159	487	169	595	842	3
first	493	159	510	169	595	842	3
insights.	71	171	111	181	595	842	3
Data	117	171	140	181	595	842	3
preparation	146	171	202	181	595	842	3
covers	208	171	240	181	595	842	3
activities	246	171	288	181	595	842	3
to	294	171	303	181	595	842	3
construct	309	171	354	181	595	842	3
the	360	171	375	181	595	842	3
final	381	171	401	181	595	842	3
dataset.	407	171	446	181	595	842	3
Modeling	452	171	497	181	595	842	3
is	502	171	510	181	595	842	3
focused	71	183	110	193	595	842	3
on	112	183	125	193	595	842	3
select	127	183	156	193	595	842	3
and	159	183	177	193	595	842	3
applied	180	183	215	193	595	842	3
modeling	218	183	263	193	595	842	3
techniques.	266	183	322	193	595	842	3
Some	325	183	354	193	595	842	3
techniques	356	183	410	193	595	842	3
require	412	183	447	193	595	842	3
specific	449	183	486	193	595	842	3
data	489	183	510	193	595	842	3
formats.	71	195	111	205	595	842	3
Therefore,	113	195	164	205	595	842	3
there	166	195	191	205	595	842	3
is	193	195	201	205	595	842	3
a	204	195	210	205	595	842	3
close	212	195	238	205	595	842	3
link	240	195	257	205	595	842	3
between	259	195	301	205	595	842	3
Data	303	195	326	205	595	842	3
Preparation	329	195	386	205	595	842	3
and	388	195	406	205	595	842	3
Modeling.	409	195	456	205	595	842	3
Evaluation	459	195	510	205	595	842	3
is	71	207	79	217	595	842	3
focused	82	207	121	217	595	842	3
in	125	207	133	217	595	842	3
reviewing	137	207	183	217	595	842	3
the	187	207	202	217	595	842	3
steps	205	207	232	217	595	842	3
executed	235	207	280	217	595	842	3
to	283	207	293	217	595	842	3
construct	296	207	341	217	595	842	3
the	344	207	360	217	595	842	3
model	363	207	393	217	595	842	3
to	397	207	406	217	595	842	3
be	409	207	421	217	595	842	3
certain	425	207	458	217	595	842	3
it	461	207	467	217	595	842	3
properly	470	207	510	217	595	842	3
achieves	71	219	114	229	595	842	3
the	118	219	134	229	595	842	3
business	138	219	181	229	595	842	3
objectives.	185	219	237	229	595	842	3
Deployment	241	219	299	229	595	842	3
is	303	219	311	229	595	842	3
focused	315	219	353	229	595	842	3
on	357	219	370	229	595	842	3
representing	373	219	435	229	595	842	3
the	438	219	454	229	595	842	3
knowledge	458	219	510	229	595	842	3
gained	71	231	104	241	595	842	3
in	107	231	115	241	595	842	3
a	118	231	125	241	595	842	3
way	127	231	147	241	595	842	3
that	150	231	168	241	595	842	3
the	171	231	186	241	595	842	3
customer	189	231	234	241	595	842	3
can	237	231	255	241	595	842	3
use	258	231	276	241	595	842	3
it	279	231	284	241	595	842	3
(Wirth,	287	231	319	241	595	842	3
2000).	322	231	353	241	595	842	3
This	356	231	377	241	595	842	3
paper	380	231	408	241	595	842	3
summarizes	411	231	470	241	595	842	3
CRISP-	473	231	511	241	595	842	3
DM	71	243	88	253	595	842	3
phases	91	243	127	253	595	842	3
in	130	243	138	253	595	842	3
four	141	243	160	253	595	842	3
stages.	163	243	199	253	595	842	3
In	99	256	109	266	595	842	3
the	111	256	126	266	595	842	3
modeling	129	256	173	266	595	842	3
phase,	176	256	209	266	595	842	3
there	211	256	236	266	595	842	3
are	239	256	255	266	595	842	3
several	257	256	293	266	595	842	3
data	295	256	317	266	595	842	3
mining	319	256	351	266	595	842	3
techniques	354	256	407	266	595	842	3
which	409	256	437	266	595	842	3
can	440	256	458	266	595	842	3
be	460	256	472	266	595	842	3
applied	475	256	510	266	595	842	3
according	71	268	119	278	595	842	3
to	122	268	131	278	595	842	3
the	135	268	150	278	595	842	3
problem	153	268	193	278	595	842	3
to	197	268	206	278	595	842	3
solve.	209	268	238	278	595	842	3
Algorithms	241	268	293	278	595	842	3
oriented	297	268	337	278	595	842	3
to	340	268	349	278	595	842	3
the	353	268	368	278	595	842	3
grouping	371	268	414	278	595	842	3
of	418	268	427	278	595	842	3
data,	430	268	455	278	595	842	3
as	458	268	470	278	595	842	3
well	473	268	492	278	595	842	3
as,	495	268	510	278	595	842	3
techniques	71	281	124	291	595	842	3
for	127	281	140	291	595	842	3
treatment	143	281	190	291	595	842	3
and	193	281	211	291	595	842	3
data	214	281	236	291	595	842	3
analysis	239	281	278	291	595	842	3
are	281	281	297	291	595	842	3
addressed	300	281	352	291	595	842	3
to	355	281	364	291	595	842	3
follow.	367	281	398	291	595	842	3
Grouping	71	306	121	316	595	842	3
algorithms	124	306	180	316	595	842	3
Data	85	324	109	334	595	842	3
grouping	111	324	154	334	595	842	3
algorithms	157	324	208	334	595	842	3
divide	211	324	239	334	595	842	3
a	242	324	248	334	595	842	3
dataset	251	324	287	334	595	842	3
into	290	324	308	334	595	842	3
groups	311	324	344	334	595	842	3
with	347	324	367	334	595	842	3
some	369	324	396	334	595	842	3
similarity.	399	324	445	334	595	842	3
This	448	324	469	334	595	842	3
paper	471	324	500	334	595	842	3
is	502	324	510	334	595	842	3
focused	71	336	110	346	595	842	3
on	115	336	127	346	595	842	3
the	133	336	148	346	595	842	3
use	154	336	172	346	595	842	3
of	177	336	186	346	595	842	3
X-means	191	336	236	346	595	842	3
and	241	336	260	346	595	842	3
K-means	265	336	309	346	595	842	3
grouping	315	336	357	346	595	842	3
algorithms.	363	336	417	346	595	842	3
X-means	422	336	467	346	595	842	3
find	472	336	490	346	595	842	3
the	495	336	510	346	595	842	3
efficient	71	348	109	358	595	842	3
cluster	112	348	145	358	595	842	3
number	148	348	186	358	595	842	3
among	189	348	223	358	595	842	3
a	226	348	232	358	595	842	3
maximum	236	348	283	358	595	842	3
and	287	348	305	358	595	842	3
minimum	309	348	353	358	595	842	3
values.	357	348	392	358	595	842	3
This	395	348	416	358	595	842	3
algorithm	419	348	465	358	595	842	3
uses	468	348	491	358	595	842	3
the	495	348	510	358	595	842	3
KD-Tree	71	360	113	370	595	842	3
technique	117	360	164	370	595	842	3
to	168	360	177	370	595	842	3
improve	181	360	220	370	595	842	3
the	224	360	240	370	595	842	3
speed.	244	360	277	370	595	842	3
X-means	281	360	325	370	595	842	3
is	329	360	337	370	595	842	3
composed	341	360	392	370	595	842	3
of	395	360	405	370	595	842	3
two	409	360	426	370	595	842	3
steps:	430	360	459	370	595	842	3
Improve-	467	360	510	370	595	842	3
Params	71	372	109	382	595	842	3
and	113	372	132	382	595	842	3
Improve-Structure.	136	372	227	382	595	842	3
The	232	372	251	382	595	842	3
first	255	372	273	382	595	842	3
step	278	372	298	382	595	842	3
applies	303	372	338	382	595	842	3
K-means	342	372	386	382	595	842	3
to	391	372	400	382	595	842	3
get	405	372	420	382	595	842	3
k	424	372	430	382	595	842	3
in	434	372	443	382	595	842	3
convergence	447	372	510	382	595	842	3
starting	71	384	107	394	595	842	3
from	111	384	133	394	595	842	3
k	136	384	142	394	595	842	3
equal	146	384	173	394	595	842	3
to	176	384	185	394	595	842	3
the	189	384	204	394	595	842	3
minimum	208	384	253	394	595	842	3
value	256	384	282	394	595	842	3
provided.	286	384	331	394	595	842	3
The	335	384	354	394	595	842	3
second	358	384	393	394	595	842	3
step	397	384	418	394	595	842	3
begins	421	384	454	394	595	842	3
by	458	384	469	394	595	842	3
dividing	473	384	510	394	595	842	3
the	71	396	86	406	595	842	3
center	89	396	120	406	595	842	3
of	123	396	132	406	595	842	3
each	135	396	159	406	595	842	3
group	162	396	190	406	595	842	3
into	193	396	211	406	595	842	3
two	214	396	231	406	595	842	3
branches	234	396	279	406	595	842	3
in	282	396	291	406	595	842	3
opposite	294	396	336	406	595	842	3
directions	339	396	386	406	595	842	3
along	389	396	416	406	595	842	3
a	419	396	425	406	595	842	3
random	428	396	465	406	595	842	3
vector.	468	396	501	406	595	842	3
If	504	396	510	406	595	842	3
k>	71	408	83	418	595	842	3
=	87	408	93	418	595	842	3
k	97	408	102	418	595	842	3
(max)	106	408	134	418	595	842	3
the	138	408	153	418	595	842	3
processing	157	408	210	418	595	842	3
ends	214	408	238	418	595	842	3
and	242	408	260	418	595	842	3
reports	264	408	298	418	595	842	3
the	302	408	317	418	595	842	3
model	321	408	351	418	595	842	3
with	354	408	374	418	595	842	3
the	378	408	393	418	595	842	3
best	397	408	418	418	595	842	3
score	421	408	448	418	595	842	3
(Kumar	452	408	488	418	595	842	3
and	492	408	510	418	595	842	3
Wasan,	71	420	108	430	595	842	3
2010).	111	420	143	430	595	842	3
K-means	85	432	130	442	595	842	3
takes	132	432	159	442	595	842	3
a	162	432	168	442	595	842	3
dataset	171	432	207	442	595	842	3
X	210	432	217	442	595	842	3
=	220	432	227	442	595	842	3
{x1,	230	432	248	442	595	842	3
x2,	251	432	266	442	595	842	3
...,	269	432	281	442	595	842	3
xm}	284	432	302	442	595	842	3
and	305	432	324	442	595	842	3
divides	327	432	361	442	595	842	3
it	364	432	369	442	595	842	3
into	373	432	390	442	595	842	3
k	393	432	399	442	595	842	3
groups.	402	432	439	442	595	842	3
This	441	432	462	442	595	842	3
algorithm	465	432	510	442	595	842	3
groups	71	444	105	454	595	842	3
data	108	444	129	454	595	842	3
points	132	444	161	454	595	842	3
by	164	444	176	454	595	842	3
average	179	444	219	454	595	842	3
values	222	444	254	454	595	842	3
using	257	444	283	454	595	842	3
the	286	444	301	454	595	842	3
Euclidean	304	444	352	454	595	842	3
Distance	356	444	398	454	595	842	3
obtaining	401	444	446	454	595	842	3
C	449	444	457	454	595	842	3
=	460	444	467	454	595	842	3
{C1,	470	444	490	454	595	842	3
C2,	493	444	510	454	595	842	3
...,	71	456	83	466	595	842	3
Ck},	86	456	106	466	595	842	3
so	110	456	121	466	595	842	3
that,	124	456	146	466	595	842	3
while	149	456	174	466	595	842	3
less	177	456	196	466	595	842	3
similarity	200	456	242	466	595	842	3
there	245	456	271	466	595	842	3
are	274	456	289	466	595	842	3
among	293	456	326	466	595	842	3
the	329	456	345	466	595	842	3
classes,	348	456	387	466	595	842	3
greater	390	456	425	466	595	842	3
similarity	428	456	471	466	595	842	3
there	474	456	499	466	595	842	3
is	502	456	510	466	595	842	3
among	71	468	105	478	595	842	3
elements	108	468	152	478	595	842	3
from	155	468	177	478	595	842	3
a	180	468	187	478	595	842	3
class	189	468	215	478	595	842	3
(Zhang,	218	468	256	478	595	842	3
Deng,	259	468	288	478	595	842	3
and	291	468	310	478	595	842	3
Li,	313	468	324	478	595	842	3
2017).	327	468	358	478	595	842	3
Techniques	71	486	132	496	595	842	3
for	135	486	150	496	595	842	3
treatment	153	486	203	496	595	842	3
and	206	486	226	496	595	842	3
data	229	486	251	496	595	842	3
analysis	254	486	298	496	595	842	3
Techniques	85	504	142	514	595	842	3
for	147	504	160	514	595	842	3
processing	164	504	217	514	595	842	3
and	222	504	240	514	595	842	3
data	245	504	266	514	595	842	3
analysis	271	504	310	514	595	842	3
allows	315	504	345	514	595	842	3
us	350	504	361	514	595	842	3
to	366	504	375	514	595	842	3
work	380	504	403	514	595	842	3
with	407	504	427	514	595	842	3
data	431	504	453	514	595	842	3
in	457	504	466	514	595	842	3
different	470	504	510	514	595	842	3
measurement	71	516	138	526	595	842	3
units.	142	516	168	526	595	842	3
This	175	516	196	526	595	842	3
paper	199	516	227	526	595	842	3
is	231	516	239	526	595	842	3
focused	242	516	281	526	595	842	3
on	284	516	296	526	595	842	3
the	300	516	315	526	595	842	3
use	319	516	336	526	595	842	3
of	340	516	349	526	595	842	3
the	352	516	368	526	595	842	3
correlation	371	516	423	526	595	842	3
matrix	426	516	456	526	595	842	3
technique,	459	516	510	526	595	842	3
as	71	528	83	538	595	842	3
well	86	528	105	538	595	842	3
as,	109	528	124	538	595	842	3
on	127	528	140	538	595	842	3
the	143	528	159	538	595	842	3
process	162	528	201	538	595	842	3
"Select	204	528	239	538	595	842	3
by	243	528	254	538	595	842	3
weights"	258	528	299	538	595	842	3
use.	303	528	324	538	595	842	3
The	327	528	346	538	595	842	3
correlation	350	528	401	538	595	842	3
matrix	405	528	435	538	595	842	3
is	438	528	446	538	595	842	3
a	450	528	456	538	595	842	3
symmetric	460	528	510	538	595	842	3
matrix	71	540	101	550	595	842	3
of	104	540	113	550	595	842	3
n	116	540	122	550	595	842	3
x	125	540	130	550	595	842	3
n	133	540	139	550	595	842	3
for	142	540	155	550	595	842	3
n	158	540	164	550	595	842	3
vectors.	167	540	205	550	595	842	3
Each	208	540	233	550	595	842	3
position	236	540	274	550	595	842	3
(i,	277	540	286	550	595	842	3
j)	289	540	295	550	595	842	3
has	298	540	315	550	595	842	3
a	318	540	324	550	595	842	3
value	327	540	353	550	595	842	3
between	356	540	398	550	595	842	3
1	401	540	407	550	595	842	3
and	410	540	428	550	595	842	3
-1,	431	540	444	550	595	842	3
such	447	540	470	550	595	842	3
position	473	540	510	550	595	842	3
is	71	552	79	562	595	842	3
the	81	552	97	562	595	842	3
value	99	552	125	562	595	842	3
of	128	552	137	562	595	842	3
the	139	552	155	562	595	842	3
Pearson	157	552	198	562	595	842	3
correlation	201	552	252	562	595	842	3
(Gao,	254	552	282	562	595	842	3
Tung,	284	552	313	562	595	842	3
and	315	552	333	562	595	842	3
Yang,	336	552	365	562	595	842	3
2017).	367	552	398	562	595	842	3
The	401	552	420	562	595	842	3
"Select	422	552	457	562	595	842	3
by	459	552	471	562	595	842	3
weights	473	552	510	562	595	842	3
process"	71	564	114	574	595	842	3
selects	117	564	152	574	595	842	3
only	156	564	176	574	595	842	3
those	180	564	207	574	595	842	3
attributes	211	564	256	574	595	842	3
from	260	564	282	574	595	842	3
an	286	564	299	574	595	842	3
input	303	564	327	574	595	842	3
"ExampleSet"	331	564	398	574	595	842	3
whose	402	564	434	574	595	842	3
weights	438	564	475	574	595	842	3
satisfy	479	564	510	574	595	842	3
the	71	576	86	586	595	842	3
specified	91	576	134	586	595	842	3
criterion	138	576	177	586	595	842	3
for	182	576	194	586	595	842	3
the	199	576	214	586	595	842	3
input	218	576	242	586	595	842	3
weights.	246	576	287	586	595	842	3
Input	291	576	316	586	595	842	3
weights	320	576	357	586	595	842	3
are	361	576	377	586	595	842	3
provided	382	576	424	586	595	842	3
through	428	576	465	586	595	842	3
an	470	576	482	586	595	842	3
input	486	576	510	586	595	842	3
port.	71	588	93	598	595	842	3
The	98	588	117	598	595	842	3
criterion	122	588	161	598	595	842	3
for	166	588	179	598	595	842	3
selecting	184	588	227	598	595	842	3
attributes	233	588	278	598	595	842	3
by	283	588	294	598	595	842	3
weight	299	588	331	598	595	842	3
is	336	588	344	598	595	842	3
determined	349	588	404	598	595	842	3
by	409	588	421	598	595	842	3
a	426	588	432	598	595	842	3
weight	437	588	469	598	595	842	3
relation	474	588	510	598	595	842	3
parameter	71	600	121	610	595	842	3
(«Select	124	600	165	610	595	842	3
by	168	600	179	610	595	842	3
Weights	182	600	222	610	595	842	3
-	225	600	229	610	595	842	3
RapidMiner	232	600	288	610	595	842	3
Documentation»,	291	600	374	610	595	842	3
s.	377	600	386	610	595	842	3
f.).	389	600	402	610	595	842	3
In	85	612	95	622	595	842	3
the	100	612	115	622	595	842	3
air	120	612	133	622	595	842	3
pollution	138	612	179	622	595	842	3
context,	184	612	222	622	595	842	3
both	228	612	249	622	595	842	3
algorithms	254	612	305	622	595	842	3
and	310	612	329	622	595	842	3
data	334	612	355	622	595	842	3
mining	360	612	393	622	595	842	3
techniques	398	612	451	622	595	842	3
have	457	612	480	622	595	842	3
been	486	612	510	622	595	842	3
examined	71	624	119	634	595	842	3
in	124	624	132	634	595	842	3
the	137	624	153	634	595	842	3
literature	158	624	200	634	595	842	3
from	205	624	228	634	595	842	3
three	233	624	258	634	595	842	3
perspectives:	263	624	327	634	595	842	3
pollutants	332	624	380	634	595	842	3
predictions,	385	624	441	634	595	842	3
infrastructure	446	624	510	634	595	842	3
improvement,	71	636	138	646	595	842	3
and	144	636	162	646	595	842	3
correlation	169	636	220	646	595	842	3
analysis.	227	636	269	646	595	842	3
Du	276	636	290	646	595	842	3
and	296	636	315	646	595	842	3
Varde	321	636	350	646	595	842	3
(2016,	357	636	388	646	595	842	3
p.	394	636	404	646	595	842	3
2)	410	636	420	646	595	842	3
use	426	636	444	646	595	842	3
data	450	636	472	646	595	842	3
mining	478	636	510	646	595	842	3
algorithms	71	648	122	658	595	842	3
for	124	648	137	658	595	842	3
association,	140	648	198	658	595	842	3
grouping	200	648	243	658	595	842	3
and,	245	648	267	658	595	842	3
classification	269	648	332	658	595	842	3
to	334	648	343	658	595	842	3
identify	346	648	381	658	595	842	3
correlations	383	648	440	658	595	842	3
among	442	648	476	658	595	842	3
PM2.5	479	648	510	658	595	842	3
pollution	71	660	112	670	595	842	3
and	117	660	135	670	595	842	3
traffic,	141	660	170	670	595	842	3
in	176	660	184	670	595	842	3
order	189	660	215	670	595	842	3
to	220	660	229	670	595	842	3
make	234	660	261	670	595	842	3
predictions	266	660	319	670	595	842	3
of	325	660	334	670	595	842	3
PM2.5.	339	660	374	670	595	842	3
Souza	379	660	410	670	595	842	3
and	415	660	434	670	595	842	3
Rabelo	439	660	474	670	595	842	3
(2015)	479	660	510	670	595	842	3
applied	71	672	106	682	595	842	3
association	109	672	164	682	595	842	3
rules	167	672	191	682	595	842	3
to	193	672	203	682	595	842	3
find	205	672	223	682	595	842	3
correlations	226	672	283	682	595	842	3
among	285	672	319	682	595	842	3
air	322	672	334	682	595	842	3
pollutants	337	672	384	682	595	842	3
and	386	672	405	682	595	842	3
respiratory	407	672	459	682	595	842	3
problems.	462	672	510	682	595	842	3
Shazan	71	684	108	694	595	842	3
et	113	684	122	694	595	842	3
al.,	127	684	141	694	595	842	3
(2017)	146	684	178	694	595	842	3
use	183	684	200	694	595	842	3
a	205	684	211	694	595	842	3
spatial	216	684	247	694	595	842	3
data	252	684	274	694	595	842	3
mining	278	684	310	694	595	842	3
technique	315	684	363	694	595	842	3
to	367	684	377	694	595	842	3
find	381	684	399	694	595	842	3
associations	404	684	464	694	595	842	3
between	469	684	510	694	595	842	3
spatial	71	696	103	707	595	842	3
features.	107	696	150	707	595	842	3
Such	155	696	180	707	595	842	3
research	185	696	228	707	595	842	3
seeks	232	696	261	707	595	842	3
to	266	696	275	707	595	842	3
analyze	280	696	318	707	595	842	3
whether	322	696	362	707	595	842	3
the	366	696	382	707	595	842	3
maternal	386	696	429	707	595	842	3
exposure	434	696	479	707	595	842	3
to	484	696	493	707	595	842	3
air	498	696	510	707	595	842	3
pollutants	71	708	118	719	595	842	3
during	122	708	153	719	595	842	3
pregnancy	157	708	208	719	595	842	3
could	212	708	239	719	595	842	3
be	243	708	255	719	595	842	3
potentially	259	708	309	719	595	842	3
associated	313	708	366	719	595	842	3
with	370	708	389	719	595	842	3
adverse	394	708	433	719	595	842	3
birth	437	708	458	719	595	842	3
outcomes	463	708	510	719	595	842	3
(Shazan	71	720	112	731	595	842	3
et	115	720	124	731	595	842	3
al.,	127	720	142	731	595	842	3
2017).	145	720	176	731	595	842	3
The	85	732	104	743	595	842	3
literature	108	732	150	743	595	842	3
reflects	154	732	189	743	595	842	3
the	192	732	208	743	595	842	3
use	211	732	229	743	595	842	3
of	232	732	241	743	595	842	3
data	245	732	266	743	595	842	3
mining	269	732	301	743	595	842	3
in	305	732	313	743	595	842	3
air	317	732	329	743	595	842	3
pollution	332	732	373	743	595	842	3
studies.	376	732	414	743	595	842	3
However,	418	732	464	743	595	842	3
research	467	732	510	743	595	842	3
focused	71	744	110	755	595	842	3
on	112	744	124	755	595	842	3
data	127	744	148	755	595	842	3
mining	151	744	183	755	595	842	3
to	186	744	195	755	595	842	3
analyze	197	744	235	755	595	842	3
the	238	744	253	755	595	842	3
correlation	256	744	307	755	595	842	3
between	310	744	351	755	595	842	3
several	354	744	389	755	595	842	3
air	392	744	404	755	595	842	3
pollutants	407	744	454	755	595	842	3
as	456	744	468	755	595	842	3
O3,	470	744	488	755	595	842	3
CO,	491	744	510	755	595	842	3
Enfoque	206	778	236	785	595	842	3
UTE,	238	778	256	785	595	842	3
V.9-N.4,	258	778	288	785	595	842	3
Dec.2018,	290	778	326	785	595	842	3
pp.	328	778	339	785	595	842	3
168	342	778	355	785	595	842	3
-	357	778	360	785	595	842	3
179	362	778	376	785	595	842	3
171	511	72	525	80	595	842	4
SO2,	85	87	110	97	595	842	4
NO2,	113	87	139	97	595	842	4
and	141	87	160	97	595	842	4
PM2.5	162	87	194	97	595	842	4
is	197	87	205	97	595	842	4
scarce.	207	87	243	97	595	842	4
Such	246	87	271	97	595	842	4
analysis	273	87	313	97	595	842	4
is	316	87	324	97	595	842	4
essential	326	87	370	97	595	842	4
to	372	87	381	97	595	842	4
provide	384	87	420	97	595	842	4
complete	423	87	468	97	595	842	4
information	470	87	525	97	595	842	4
to	85	99	94	109	595	842	4
support	97	99	134	109	595	842	4
the	137	99	153	109	595	842	4
decision-making	156	99	235	109	595	842	4
process	238	99	277	109	595	842	4
of	280	99	289	109	595	842	4
environmental	292	99	361	109	595	842	4
managers	364	99	413	109	595	842	4
and	416	99	434	109	595	842	4
urban	437	99	465	109	595	842	4
planners.	468	99	514	109	595	842	4
3.	85	117	94	127	595	842	4
Data	97	117	121	127	595	842	4
mining	124	117	160	127	595	842	4
approach	163	117	213	127	595	842	4
to	216	117	226	127	595	842	4
identify	229	117	269	127	595	842	4
behavioral	271	117	327	127	595	842	4
patterns	330	117	373	127	595	842	4
among	376	117	412	127	595	842	4
air	415	117	429	127	595	842	4
pollutants	432	117	484	127	595	842	4
Based	113	143	145	153	595	842	4
on	148	143	160	153	595	842	4
CRISP-DM,	163	143	220	153	595	842	4
the	223	143	238	153	595	842	4
approach	241	143	287	153	595	842	4
proposed	290	143	335	153	595	842	4
in	338	143	347	153	595	842	4
this	349	143	366	153	595	842	4
study	369	143	396	153	595	842	4
is	399	143	406	153	595	842	4
divided	409	143	444	153	595	842	4
into	447	143	465	153	595	842	4
four	467	143	486	153	595	842	4
stages:	489	143	525	153	595	842	4
contextualization,	85	155	170	166	595	842	4
data	174	155	196	166	595	842	4
preparation,	200	155	258	166	595	842	4
modeling	262	155	307	166	595	842	4
and,	311	155	332	166	595	842	4
results.	336	155	372	166	595	842	4
Figure	376	155	407	166	595	842	4
1,	411	155	420	166	595	842	4
shows	424	155	455	166	595	842	4
the	459	155	475	166	595	842	4
proposed	479	155	525	166	595	842	4
approach	85	168	131	178	595	842	4
flowchart.	134	168	181	178	595	842	4
Figure	148	319	178	328	595	842	4
1:	181	319	190	328	595	842	4
Approach	193	319	235	328	595	842	4
to	238	319	247	328	595	842	4
identify	249	319	281	328	595	842	4
behavioral	284	319	330	328	595	842	4
patterns	333	319	369	328	595	842	4
among	372	319	402	328	595	842	4
air	405	319	416	328	595	842	4
pollutants	419	319	461	328	595	842	4
The	113	342	133	352	595	842	4
stages	140	342	173	352	595	842	4
are	180	342	196	352	595	842	4
composed	204	342	255	352	595	842	4
of	262	342	271	352	595	842	4
procedures	279	342	334	352	595	842	4
which	342	342	370	352	595	842	4
lead	378	342	399	352	595	842	4
to	406	342	416	352	595	842	4
the	423	342	439	352	595	842	4
identification	446	342	508	352	595	842	4
of	515	342	524	352	595	842	4
associations	85	355	146	365	595	842	4
among	154	355	187	365	595	842	4
air	195	355	207	365	595	842	4
pollutants.	215	355	265	365	595	842	4
Results	273	355	310	365	595	842	4
are	318	355	334	365	595	842	4
presented	342	355	391	365	595	842	4
considering	399	355	455	365	595	842	4
elements	463	355	507	365	595	842	4
of	515	355	524	365	595	842	4
temporality.	85	367	142	377	595	842	4
The	145	367	164	377	595	842	4
following	167	367	210	377	595	842	4
sections	213	367	253	377	595	842	4
present	256	367	293	377	595	842	4
in	296	367	305	377	595	842	4
detail	308	367	334	377	595	842	4
every	337	367	364	377	595	842	4
stage.	367	367	397	377	595	842	4
Contextualization	85	392	177	403	595	842	4
stage	180	392	209	403	595	842	4
Generally,	120	418	171	428	595	842	4
data	173	418	195	428	595	842	4
is	198	418	205	428	595	842	4
captured	208	418	251	428	595	842	4
by	254	418	265	428	595	842	4
time	268	418	289	428	595	842	4
intervals	292	418	332	428	595	842	4
in	335	418	344	428	595	842	4
seconds	347	418	388	428	595	842	4
or	390	418	400	428	595	842	4
minutes,	403	418	444	428	595	842	4
according	447	418	495	428	595	842	4
to	497	418	507	428	595	842	4
the	509	418	525	428	595	842	4
sensor	85	430	118	441	595	842	4
configuration,	121	430	187	441	595	842	4
and	190	430	208	441	595	842	4
stored	211	430	241	441	595	842	4
in	244	430	252	441	595	842	4
databases.	255	430	309	441	595	842	4
However,	312	430	358	441	595	842	4
cities	361	430	386	441	595	842	4
have	388	430	412	441	595	842	4
specific	415	430	451	441	595	842	4
characteristics	454	430	524	441	595	842	4
according	85	443	133	453	595	842	4
to	137	443	146	453	595	842	4
its	150	443	161	453	595	842	4
localization,	165	443	222	453	595	842	4
so	226	443	238	453	595	842	4
that	242	443	260	453	595	842	4
the	264	443	279	453	595	842	4
monitoring	283	443	334	453	595	842	4
of	338	443	347	453	595	842	4
air	351	443	363	453	595	842	4
pollutants	367	443	415	453	595	842	4
can	418	443	436	453	595	842	4
change	440	443	476	453	595	842	4
from	480	443	502	453	595	842	4
one	506	443	524	453	595	842	4
country	85	456	121	466	595	842	4
to	124	456	134	466	595	842	4
another.	137	456	177	466	595	842	4
Consequently,	184	456	254	466	595	842	4
this	257	456	274	466	595	842	4
stage	277	456	304	466	595	842	4
aims	308	456	331	466	595	842	4
to	334	456	343	466	595	842	4
promote	346	456	387	466	595	842	4
a	390	456	396	466	595	842	4
first	399	456	417	466	595	842	4
contact	420	456	455	466	595	842	4
with	459	456	478	466	595	842	4
the	481	456	497	466	595	842	4
data.	500	456	524	466	595	842	4
In	85	468	94	479	595	842	4
this	97	468	115	479	595	842	4
paper,	118	468	149	479	595	842	4
the	152	468	168	479	595	842	4
proposed	171	468	216	479	595	842	4
approach	220	468	266	479	595	842	4
focuses	268	468	307	479	595	842	4
on	310	468	322	479	595	842	4
five	325	468	342	479	595	842	4
air	345	468	358	479	595	842	4
pollution	361	468	402	479	595	842	4
elements:	405	468	452	479	595	842	4
O3,	456	468	473	479	595	842	4
CO,	477	468	496	479	595	842	4
SO2,	499	468	524	479	595	842	4
NO2	85	481	108	491	595	842	4
and,	111	481	132	491	595	842	4
PM2.5.	135	481	170	491	595	842	4
Data	85	506	109	517	595	842	4
preparation	112	506	173	517	595	842	4
stage	176	506	204	517	595	842	4
The	113	530	133	542	595	842	4
data	136	530	158	542	595	842	4
preparation	161	530	217	542	595	842	4
stage	220	530	247	542	595	842	4
focuses	251	530	289	542	595	842	4
on	292	530	305	542	595	842	4
two	308	530	325	542	595	842	4
procedures	329	530	384	542	595	842	4
on	388	530	400	542	595	842	4
a	404	530	410	542	595	842	4
dataset	413	530	449	542	595	842	4
to	453	530	462	542	595	842	4
get	466	530	481	542	595	842	4
a	485	530	491	542	595	842	4
“valid”	494	530	524	542	595	842	4
data	85	544	107	555	595	842	4
input	113	544	137	555	595	842	4
before	143	544	174	555	595	842	4
the	180	544	195	555	595	842	4
analysis	201	544	241	555	595	842	4
process:	247	544	288	555	595	842	4
data	294	544	316	555	595	842	4
correction	322	544	370	555	595	842	4
and	376	544	395	555	595	842	4
data	401	544	422	555	595	842	4
cleaning.	428	544	472	555	595	842	4
The	478	544	497	555	595	842	4
data	503	544	524	555	595	842	4
correction	85	557	133	567	595	842	4
procedure	138	557	188	567	595	842	4
seeks	193	557	222	567	595	842	4
to	227	557	236	567	595	842	4
get	241	557	256	567	595	842	4
real	261	557	280	567	595	842	4
data	285	557	306	567	595	842	4
values	311	557	343	567	595	842	4
from	348	557	370	567	595	842	4
location.	375	557	416	567	595	842	4
In	426	557	435	567	595	842	4
the	440	557	455	567	595	842	4
case	460	557	484	567	595	842	4
of	489	557	498	567	595	842	4
data	503	557	524	567	595	842	4
correction,	85	570	136	580	595	842	4
the	139	570	155	580	595	842	4
literature	157	570	200	580	595	842	4
suggests	203	570	247	580	595	842	4
using	249	570	275	580	595	842	4
the	278	570	294	580	595	842	4
barometric	296	570	348	580	595	842	4
pressure	351	570	394	580	595	842	4
and	396	570	415	580	595	842	4
the	417	570	433	580	595	842	4
temperature	435	570	495	580	595	842	4
of	497	570	506	580	595	842	4
the	509	570	524	580	595	842	4
locality	85	582	119	593	595	842	4
to	122	582	131	593	595	842	4
be	135	582	147	593	595	842	4
analyzed.	150	582	197	593	595	842	4
For	201	582	217	593	595	842	4
that,	221	582	242	593	595	842	4
generally,	246	582	293	593	595	842	4
it	297	582	302	593	595	842	4
is	306	582	314	593	595	842	4
applied	317	582	353	593	595	842	4
the	356	582	371	593	595	842	4
formula	375	582	412	593	595	842	4
in	415	582	424	593	595	842	4
Equation	427	582	471	593	595	842	4
1	474	582	480	593	595	842	4
(Walden	484	582	525	593	595	842	4
and	85	595	103	605	595	842	4
Andrew,	107	595	147	605	595	842	4
2013).	150	595	181	605	595	842	4
	338	601	344	632	595	842	4
	406	601	412	632	595	842	4
o	411	610	415	619	595	842	4
760	255	616	280	630	595	842	4
mmHh	280	616	325	630	595	842	4
273	343	616	367	630	595	842	4
	370	611	379	630	595	842	4
t	381	616	386	630	595	842	4
o	388	613	393	622	595	842	4
C	394	616	405	630	595	842	4
K	418	616	430	630	595	842	4
Cc	187	626	205	641	595	842	4
	209	621	218	641	595	842	4
Co	222	626	241	641	595	842	4
	243	621	252	641	595	842	4
	327	621	335	641	595	842	4
(	434	626	440	641	595	842	4
1	438	626	447	641	595	842	4
)	445	626	451	641	595	842	4
PblmmHg	256	639	323	653	595	842	4
298	364	639	389	653	595	842	4
o	389	636	394	645	595	842	4
k	397	639	404	653	595	842	4
Where:	113	670	149	680	595	842	4
Cc:	151	670	168	680	595	842	4
Corrected	170	670	219	680	595	842	4
concentration,	221	670	290	680	595	842	4
Co:	293	670	310	680	595	842	4
Observed	312	670	360	680	595	842	4
concentration,	362	670	432	680	595	842	4
Plb:	434	670	453	680	595	842	4
Pressure	456	670	500	680	595	842	4
local	502	670	525	680	595	842	4
barometric	85	682	137	693	595	842	4
(millimeters	140	682	196	693	595	842	4
per	199	682	215	693	595	842	4
mercury),	217	682	264	693	595	842	4
tC:	267	682	281	693	595	842	4
Local	284	682	310	693	595	842	4
temperature	312	682	372	693	595	842	4
(degrees	374	682	418	693	595	842	4
centigrade).	421	682	479	693	595	842	4
The	481	682	501	693	595	842	4
data	503	682	525	693	595	842	4
cleaning	85	695	126	705	595	842	4
procedure	131	695	181	705	595	842	4
seeks	186	695	215	705	595	842	4
to	220	695	229	705	595	842	4
obtain	234	695	264	705	595	842	4
a	269	695	275	705	595	842	4
clean	281	695	307	705	595	842	4
sample	312	695	348	705	595	842	4
for	353	695	366	705	595	842	4
the	371	695	386	705	595	842	4
study.	391	695	420	705	595	842	4
For	426	695	442	705	595	842	4
that,	447	695	469	705	595	842	4
this	474	695	491	705	595	842	4
paper	496	695	524	705	595	842	4
proposes	85	708	130	718	595	842	4
to	134	708	143	718	595	842	4
manage	146	708	186	718	595	842	4
the	189	708	204	718	595	842	4
data	208	708	229	718	595	842	4
through	232	708	270	718	595	842	4
the	273	708	288	718	595	842	4
next	291	708	312	718	595	842	4
steps.	315	708	345	718	595	842	4
First,	348	708	372	718	595	842	4
obtain	376	708	406	718	595	842	4
a	409	708	415	718	595	842	4
row	419	708	436	718	595	842	4
for	439	708	452	718	595	842	4
a	456	708	462	718	595	842	4
time	465	708	486	718	595	842	4
interval	489	708	524	718	595	842	4
for	85	720	98	731	595	842	4
all	102	720	113	731	595	842	4
pollutants.	118	720	168	731	595	842	4
Second,	172	720	212	731	595	842	4
control	217	720	250	731	595	842	4
the	254	720	269	731	595	842	4
null	274	720	291	731	595	842	4
values	295	720	327	731	595	842	4
normally	331	720	373	731	595	842	4
generated	377	720	426	731	595	842	4
by	431	720	442	731	595	842	4
problems	447	720	492	731	595	842	4
in	496	720	505	731	595	842	4
the	509	720	524	731	595	842	4
calibration	85	733	135	743	595	842	4
of	141	733	150	743	595	842	4
the	155	733	171	743	595	842	4
sensors,	176	733	217	743	595	842	4
the	223	733	238	743	595	842	4
process	243	733	282	743	595	842	4
may	287	733	308	743	595	842	4
result	313	733	340	743	595	842	4
in:	346	733	357	743	595	842	4
impute	363	733	396	743	595	842	4
with	401	733	420	743	595	842	4
learning	426	733	465	743	595	842	4
algorithms,	470	733	524	743	595	842	4
replace	85	746	121	756	595	842	4
with	125	746	144	756	595	842	4
means	148	746	181	756	595	842	4
if	184	746	190	756	595	842	4
the	193	746	209	756	595	842	4
dataset	212	746	248	756	595	842	4
has	252	746	270	756	595	842	4
a	273	746	279	756	595	842	4
normal	283	746	316	756	595	842	4
distribution	320	746	373	756	595	842	4
or	377	746	386	756	595	842	4
delete	390	746	420	756	595	842	4
null	424	746	441	756	595	842	4
values	444	746	476	756	595	842	4
if	479	746	485	756	595	842	4
the	489	746	504	756	595	842	4
null	507	746	524	756	595	842	4
values	85	758	117	769	595	842	4
subset	119	758	152	769	595	842	4
is	154	758	162	769	595	842	4
short.	165	758	192	769	595	842	4
Third,	195	758	223	769	595	842	4
obtain	226	758	256	769	595	842	4
samples	258	758	299	769	595	842	4
with	302	758	321	769	595	842	4
all-time	324	758	359	769	595	842	4
interval	362	758	397	769	595	842	4
records,	400	758	439	769	595	842	4
this	442	758	459	769	595	842	4
step	462	758	483	769	595	842	4
consists	485	758	525	769	595	842	4
Enfoque	220	778	250	785	595	842	4
UTE,	252	778	270	785	595	842	4
V.9-N.4,	272	778	302	785	595	842	4
Dec.2018,	304	778	340	785	595	842	4
pp.	343	778	354	785	595	842	4
168	356	778	369	785	595	842	4
-	372	778	374	785	595	842	4
179	377	778	390	785	595	842	4
172	71	72	84	80	595	842	5
of	71	87	80	98	595	842	5
obtaining	83	87	128	98	595	842	5
samples	131	87	172	98	595	842	5
of	175	87	184	98	595	842	5
days	188	87	211	98	595	842	5
with	214	87	233	98	595	842	5
complete	237	87	281	98	595	842	5
records.	284	87	324	98	595	842	5
We	327	87	344	98	595	842	5
suggest	347	87	386	98	595	842	5
filtering	389	87	424	98	595	842	5
the	427	87	443	98	595	842	5
days	446	87	469	98	595	842	5
with	472	87	492	98	595	842	5
the	495	87	510	98	595	842	5
total	71	100	92	110	595	842	5
of	95	100	104	110	595	842	5
records	108	100	145	110	595	842	5
greater	148	100	183	110	595	842	5
than	186	100	208	110	595	842	5
99%	211	100	233	110	595	842	5
due	237	100	255	110	595	842	5
to	258	100	268	110	595	842	5
the	271	100	286	110	595	842	5
sensors	290	100	328	110	595	842	5
could	332	100	358	110	595	842	5
fail	361	100	375	110	595	842	5
and	379	100	397	110	595	842	5
do	401	100	413	110	595	842	5
not	416	100	432	110	595	842	5
record	435	100	466	110	595	842	5
all	470	100	481	110	595	842	5
times	484	100	510	110	595	842	5
intervals	71	113	112	123	595	842	5
on	115	113	127	123	595	842	5
a	130	113	136	123	595	842	5
day.	140	113	160	123	595	842	5
Modeling	71	138	119	148	595	842	5
stage	122	138	151	148	595	842	5
The	99	163	118	173	595	842	5
purpose	121	163	161	173	595	842	5
of	164	163	173	173	595	842	5
this	177	163	194	173	595	842	5
stage	197	163	224	173	595	842	5
is	227	163	235	173	595	842	5
to	238	163	247	173	595	842	5
create	250	163	281	173	595	842	5
models	284	163	319	173	595	842	5
to	322	163	331	173	595	842	5
group	334	163	363	173	595	842	5
techniques	366	163	419	173	595	842	5
and	422	163	441	173	595	842	5
algorithms	444	163	494	173	595	842	5
for	497	163	510	173	595	842	5
the	71	176	86	186	595	842	5
data	90	176	111	186	595	842	5
mining	115	176	147	186	595	842	5
in	151	176	159	186	595	842	5
air	163	176	175	186	595	842	5
pollution	178	176	219	186	595	842	5
context.	223	176	261	186	595	842	5
Thus,	264	176	292	186	595	842	5
we	296	176	309	186	595	842	5
propose	313	176	353	186	595	842	5
two	356	176	373	186	595	842	5
models	377	176	413	186	595	842	5
for	416	176	429	186	595	842	5
data	432	176	454	186	595	842	5
processing	457	176	510	186	595	842	5
which	71	189	99	199	595	842	5
are	104	189	120	199	595	842	5
composed	125	189	175	199	595	842	5
mainly	180	189	212	199	595	842	5
of	217	189	226	199	595	842	5
two	231	189	248	199	595	842	5
algorithms:	253	189	307	199	595	842	5
X-means	311	189	356	199	595	842	5
and	360	189	379	199	595	842	5
K-means.	384	189	431	199	595	842	5
For	435	189	452	199	595	842	5
the	457	189	472	199	595	842	5
correct	477	189	510	199	595	842	5
operation	71	201	117	211	595	842	5
of	120	201	129	211	595	842	5
these	131	201	158	211	595	842	5
algorithms,	161	201	214	211	595	842	5
the	217	201	232	211	595	842	5
dataset	235	201	271	211	595	842	5
requires	274	201	313	211	595	842	5
only	316	201	336	211	595	842	5
numeric	338	201	378	211	595	842	5
values.	380	201	415	211	595	842	5
Therefore,	418	201	468	211	595	842	5
the	471	201	486	211	595	842	5
data	489	201	510	211	595	842	5
preparation	71	214	127	224	595	842	5
process	130	214	168	224	595	842	5
must	171	214	195	224	595	842	5
be	198	214	211	224	595	842	5
applied	213	214	249	224	595	842	5
beforehand.	252	214	311	224	595	842	5
The	99	226	118	237	595	842	5
purpose	122	226	162	237	595	842	5
of	166	226	175	237	595	842	5
the	179	226	194	237	595	842	5
model	198	226	228	237	595	842	5
1	232	226	238	237	595	842	5
is	242	226	250	237	595	842	5
threefold:	254	226	300	237	595	842	5
(a)	304	226	317	237	595	842	5
to	321	226	330	237	595	842	5
define	334	226	364	237	595	842	5
the	368	226	384	237	595	842	5
number	388	226	425	237	595	842	5
of	429	226	438	237	595	842	5
data	442	226	463	237	595	842	5
grouping	467	226	510	237	595	842	5
clusters,	71	239	112	249	595	842	5
(b)	116	239	130	249	595	842	5
to	134	239	143	249	595	842	5
obtain	147	239	177	249	595	842	5
all	181	239	192	249	595	842	5
correlations	196	239	253	249	595	842	5
datasets	257	239	299	249	595	842	5
between	303	239	344	249	595	842	5
pollutants,	349	239	399	249	595	842	5
and	403	239	421	249	595	842	5
(c)	425	239	438	249	595	842	5
to	442	239	452	249	595	842	5
identify	456	239	491	249	595	842	5
the	495	239	510	249	595	842	5
incidence	71	252	117	262	595	842	5
between	120	252	161	262	595	842	5
pollutants.	164	252	214	262	595	842	5
To	217	252	230	262	595	842	5
this	232	252	249	262	595	842	5
end,	252	252	273	262	595	842	5
we	276	252	290	262	595	842	5
suggest	292	252	331	262	595	842	5
defining	333	252	372	262	595	842	5
the	374	252	390	262	595	842	5
number	392	252	429	262	595	842	5
of	432	252	441	262	595	842	5
data	444	252	465	262	595	842	5
grouping	467	252	510	262	595	842	5
clusters	71	264	109	275	595	842	5
using	113	264	139	275	595	842	5
X-means	144	264	188	275	595	842	5
algorithm,	192	264	240	275	595	842	5
to	244	264	254	275	595	842	5
structure	258	264	301	275	595	842	5
a	305	264	311	275	595	842	5
correlation	315	264	367	275	595	842	5
matrix	371	264	401	275	595	842	5
using	405	264	431	275	595	842	5
the	436	264	451	275	595	842	5
Pearson	455	264	496	275	595	842	5
or	500	264	510	275	595	842	5
Spearman	71	277	122	287	595	842	5
correlation	125	277	177	287	595	842	5
coefficient	180	277	230	287	595	842	5
depending	233	277	285	287	595	842	5
on	289	277	301	287	595	842	5
data	304	277	326	287	595	842	5
distribution,	329	277	386	287	595	842	5
and	389	277	408	287	595	842	5
define	411	277	441	287	595	842	5
the	445	277	460	287	595	842	5
incidence	464	277	510	287	595	842	5
using	71	288	97	300	595	842	5
"Select	102	288	137	300	595	842	5
by	142	288	154	300	595	842	5
Weights	159	288	198	300	595	842	5
process”.	204	288	249	300	595	842	5
Considering	254	288	313	300	595	842	5
the	318	288	333	300	595	842	5
procedures	338	288	393	300	595	842	5
addressed	398	288	450	300	595	842	5
in	455	290	463	300	595	842	5
the	469	290	484	300	595	842	5
data	489	290	510	300	595	842	5
preparation	71	302	127	313	595	842	5
stage,	130	302	160	313	595	842	5
Figure	163	302	194	313	595	842	5
2	197	302	203	313	595	842	5
shows	206	302	237	313	595	842	5
the	240	302	256	313	595	842	5
process	259	302	297	313	595	842	5
to	300	302	310	313	595	842	5
follow	312	302	341	313	595	842	5
in	344	302	352	313	595	842	5
model	355	302	385	313	595	842	5
1.	388	302	398	313	595	842	5
Figure	120	466	151	476	595	842	5
2:	154	466	162	476	595	842	5
Data	165	466	186	476	595	842	5
grouping	189	466	228	476	595	842	5
cluster	231	466	260	476	595	842	5
(X-means)	263	466	310	476	595	842	5
and	313	466	329	476	595	842	5
incidence	332	466	374	476	595	842	5
between	377	466	415	476	595	842	5
pollutants	418	466	460	476	595	842	5
The	99	490	118	500	595	842	5
data	122	490	143	500	595	842	5
analysis	146	490	186	500	595	842	5
is	189	490	197	500	595	842	5
focused	201	490	239	500	595	842	5
on	242	490	255	500	595	842	5
the	258	490	273	500	595	842	5
temporality;	277	490	333	500	595	842	5
consequently,	337	490	405	500	595	842	5
the	408	490	423	500	595	842	5
data	427	490	448	500	595	842	5
represented	451	490	510	500	595	842	5
in	71	502	79	512	595	842	5
ReadCSV	82	502	131	512	595	842	5
must	134	502	158	512	595	842	5
be	161	502	173	512	595	842	5
converted	176	502	224	512	595	842	5
to	227	502	236	512	595	842	5
date	239	502	260	512	595	842	5
type.	263	502	286	512	595	842	5
Such	289	502	314	512	595	842	5
conversion	317	502	370	512	595	842	5
is	373	502	381	512	595	842	5
denoted	384	502	423	512	595	842	5
in	426	502	435	512	595	842	5
Nominal2Date.	437	502	510	512	595	842	5
Also,	71	515	95	525	595	842	5
air	98	515	110	525	595	842	5
pollutants	113	515	160	525	595	842	5
measure	163	515	206	525	595	842	5
units	208	515	232	525	595	842	5
must	234	515	258	525	595	842	5
be	261	515	273	525	595	842	5
homogenized.	276	515	345	525	595	842	5
Thus,	348	515	375	525	595	842	5
all	378	515	389	525	595	842	5
pollutants	392	515	439	525	595	842	5
measure	441	515	484	525	595	842	5
units	487	515	510	525	595	842	5
must	71	527	95	538	595	842	5
be	99	527	111	538	595	842	5
converted	115	527	163	538	595	842	5
into	167	527	185	538	595	842	5
the	189	527	204	538	595	842	5
same	208	527	235	538	595	842	5
measurement	239	527	306	538	595	842	5
unit	310	527	327	538	595	842	5
or	331	527	341	538	595	842	5
normalize	345	527	393	538	595	842	5
by	397	527	408	538	595	842	5
Z-Transformation	412	527	497	538	595	842	5
or	501	527	510	538	595	842	5
Range	71	540	103	550	595	842	5
transformation.	107	540	180	550	595	842	5
These	183	540	214	550	595	842	5
conversions	217	540	276	550	595	842	5
are	279	540	295	550	595	842	5
reflected	298	540	341	550	595	842	5
in	344	540	353	550	595	842	5
MeasureUnits.	356	540	427	550	595	842	5
According	430	540	479	550	595	842	5
to	482	540	492	550	595	842	5
the	495	540	510	550	595	842	5
data	71	553	92	563	595	842	5
loggers	96	553	132	563	595	842	5
reflected	135	553	177	563	595	842	5
in	180	553	188	563	595	842	5
Table	191	553	219	563	595	842	5
1	222	553	228	563	595	842	5
CO	231	553	248	563	595	842	5
is	251	553	259	563	595	842	5
the	262	553	277	563	595	842	5
only	280	553	300	563	595	842	5
pollutant	303	553	345	563	595	842	5
which	348	553	376	563	595	842	5
needs	379	553	409	563	595	842	5
to	412	553	421	563	595	842	5
be	424	553	437	563	595	842	5
converted.	440	553	491	563	595	842	5
Attribute	135	590	176	599	595	842	5
Table	232	578	258	587	595	842	5
1:	261	578	270	587	595	842	5
Stored	273	578	302	587	595	842	5
Pollutants	305	578	349	587	595	842	5
Minimum	212	590	256	599	595	842	5
Maximum	279	590	325	599	595	842	5
Average	352	590	391	599	595	842	5
SD	432	590	446	599	595	842	5
O3	114	608	127	617	595	842	5
(ug/m3)	130	608	166	617	595	842	5
0.000	208	608	233	617	595	842	5
140.274	271	608	307	617	595	842	5
30.074	343	608	374	617	595	842	5
25.835	411	608	441	617	595	842	5
CO	114	626	129	635	595	842	5
(mg/m3)	134	626	173	635	595	842	5
0.263	208	626	233	635	595	842	5
3.698	271	626	296	635	595	842	5
0.873	343	626	368	635	595	842	5
0.403	411	626	436	635	595	842	5
NO2	114	645	134	654	595	842	5
(ug/m3)	137	645	173	654	595	842	5
0.000	208	645	233	654	595	842	5
94.818	271	645	302	654	595	842	5
18.098	343	645	374	654	595	842	5
15.067	411	645	441	654	595	842	5
SO2	114	663	134	672	595	842	5
(ug/m3)	137	663	173	672	595	842	5
0.000	208	663	233	672	595	842	5
88.244	271	663	302	672	595	842	5
8.177	343	663	368	672	595	842	5
8.171	411	663	436	672	595	842	5
PM2.5	114	681	143	690	595	842	5
(ug/m3)	146	681	182	690	595	842	5
0.000	208	681	233	691	595	842	5
72.000	271	681	302	691	595	842	5
14.836	343	681	374	691	595	842	5
13.181	411	681	441	691	595	842	5
In	99	711	109	721	595	842	5
order	113	711	138	721	595	842	5
to	142	711	152	721	595	842	5
achieve	156	711	194	721	595	842	5
model	198	711	228	721	595	842	5
1	232	711	238	721	595	842	5
first	242	711	260	721	595	842	5
goal,	264	711	288	721	595	842	5
only	292	711	312	721	595	842	5
pollutants	316	711	363	721	595	842	5
columns	367	711	408	721	595	842	5
must	412	711	436	721	595	842	5
be	440	711	452	721	595	842	5
selected	456	711	497	721	595	842	5
in	502	711	510	721	595	842	5
SelectByPollutants(1).	71	724	179	734	595	842	5
Next,	187	724	213	734	595	842	5
the	220	724	236	734	595	842	5
X-means	244	724	288	734	595	842	5
algorithm	296	724	341	734	595	842	5
must	349	724	373	734	595	842	5
be	380	724	393	734	595	842	5
executed.	401	724	448	734	595	842	5
That,	456	724	481	734	595	842	5
it	489	724	494	734	595	842	5
is	502	724	510	734	595	842	5
represented	71	737	130	747	595	842	5
in	133	737	141	747	595	842	5
X-means.	144	737	191	747	595	842	5
Enfoque	206	778	236	785	595	842	5
UTE,	238	778	256	785	595	842	5
V.9-N.4,	258	778	288	785	595	842	5
Dec.2018,	290	778	326	785	595	842	5
pp.	328	778	339	785	595	842	5
168	342	778	355	785	595	842	5
-	357	778	360	785	595	842	5
179	362	778	376	785	595	842	5
173	511	72	525	80	595	842	6
To	113	87	126	98	595	842	6
achieve	130	87	168	98	595	842	6
the	171	87	187	98	595	842	6
second	190	87	226	98	595	842	6
goal	229	87	250	98	595	842	6
in	254	87	262	98	595	842	6
model	266	87	296	98	595	842	6
1,	299	87	308	98	595	842	6
it	312	87	317	98	595	842	6
is	321	87	329	98	595	842	6
necessary	333	87	383	98	595	842	6
to	386	87	395	98	595	842	6
select	399	87	428	98	595	842	6
the	431	87	447	98	595	842	6
cluster	450	87	482	98	595	842	6
with	486	87	506	98	595	842	6
the	509	87	524	98	595	842	6
greatest	85	100	125	110	595	842	6
number	129	100	166	110	595	842	6
of	170	100	179	110	595	842	6
elements	184	100	228	110	595	842	6
in	232	100	241	110	595	842	6
BiggestCluster.	245	100	319	110	595	842	6
Once	323	100	350	110	595	842	6
again,	354	100	384	110	595	842	6
only	388	100	408	110	595	842	6
pollutants	412	100	459	110	595	842	6
columns	463	100	504	110	595	842	6
are	508	100	524	110	595	842	6
selected	85	113	126	123	595	842	6
in	135	113	144	123	595	842	6
SelectedPollutants(2).	153	113	261	123	595	842	6
Finally,	270	113	305	123	595	842	6
the	314	113	329	123	595	842	6
correlation	339	113	390	123	595	842	6
matrix	399	113	429	123	595	842	6
is	438	113	446	123	595	842	6
developed	456	113	506	123	595	842	6
in	516	113	524	123	595	842	6
CorrelationMatrix.	85	125	172	136	595	842	6
The	175	125	194	136	595	842	6
correlation	196	125	248	136	595	842	6
matrix	250	125	280	136	595	842	6
values	283	125	315	136	595	842	6
reflect	318	125	348	136	595	842	6
the	350	125	366	136	595	842	6
behavior	368	125	410	136	595	842	6
between	413	125	455	136	595	842	6
two	458	125	475	136	595	842	6
pollutants	477	125	524	136	595	842	6
(«Quick-R:	85	138	138	148	595	842	6
Correlations»,	142	138	210	148	595	842	6
s.	215	138	223	148	595	842	6
f.).	226	138	239	148	595	842	6
There	244	138	272	148	595	842	6
are	277	138	293	148	595	842	6
some	297	138	324	148	595	842	6
factors	328	138	361	148	595	842	6
to	365	138	375	148	595	842	6
consider	379	138	421	148	595	842	6
for	425	138	438	148	595	842	6
its	442	138	453	148	595	842	6
interpretation.	457	138	524	148	595	842	6
First,	85	151	110	161	595	842	6
if	114	151	119	161	595	842	6
the	123	151	138	161	595	842	6
correlation	143	151	194	161	595	842	6
value	198	151	224	161	595	842	6
is	228	151	236	161	595	842	6
less	240	151	260	161	595	842	6
than	264	151	285	161	595	842	6
0,	289	151	298	161	595	842	6
it	303	151	308	161	595	842	6
indicates	312	151	355	161	595	842	6
that	359	151	378	161	595	842	6
the	382	151	397	161	595	842	6
behavior	401	151	443	161	595	842	6
of	447	151	456	161	595	842	6
a	461	151	467	161	595	842	6
pollutant	471	151	512	161	595	842	6
is	516	151	524	161	595	842	6
inversely	85	163	129	173	595	842	6
proportional	132	163	190	173	595	842	6
to	193	163	202	173	595	842	6
another.	206	163	246	173	595	842	6
This	249	163	270	173	595	842	6
means	273	163	307	173	595	842	6
that	310	163	328	173	595	842	6
the	331	163	347	173	595	842	6
highest	350	163	385	173	595	842	6
values	389	163	420	173	595	842	6
in	424	163	432	173	595	842	6
a	436	163	442	173	595	842	6
pollutant	445	163	486	173	595	842	6
are	490	163	506	173	595	842	6
the	509	163	524	173	595	842	6
lowest	85	176	116	186	595	842	6
values	120	176	151	186	595	842	6
of	155	176	164	186	595	842	6
another	167	176	205	186	595	842	6
pollutant.	208	176	253	186	595	842	6
This	256	174	277	186	595	842	6
is	280	174	288	186	595	842	6
also	292	174	312	186	595	842	6
known	315	174	347	186	595	842	6
as	350	174	362	186	595	842	6
“negative	365	174	410	186	595	842	6
correlation”.	414	174	472	186	595	842	6
Second,	475	174	515	186	595	842	6
if	519	174	524	186	595	842	6
the	85	189	100	199	595	842	6
correlation	107	189	158	199	595	842	6
value	164	189	190	199	595	842	6
is	197	189	204	199	595	842	6
greater	211	189	246	199	595	842	6
than	252	189	273	199	595	842	6
0,	279	189	288	199	595	842	6
it	295	189	300	199	595	842	6
indicates	306	189	350	199	595	842	6
that	356	189	374	199	595	842	6
the	380	189	396	199	595	842	6
pollutants	402	189	449	199	595	842	6
have	455	189	479	199	595	842	6
a	485	189	491	199	595	842	6
direct	497	189	524	199	595	842	6
correlation.	85	201	140	211	595	842	6
This	143	201	164	211	595	842	6
means	167	201	200	211	595	842	6
that	203	201	221	211	595	842	6
the	225	201	240	211	595	842	6
highest	243	201	279	211	595	842	6
values	282	201	313	211	595	842	6
in	317	201	325	211	595	842	6
a	328	201	335	211	595	842	6
pollutant	338	201	379	211	595	842	6
are	383	201	399	211	595	842	6
highest	402	201	437	211	595	842	6
values	440	201	472	211	595	842	6
in	475	201	484	211	595	842	6
another	487	201	524	211	595	842	6
pollutant.	85	214	130	224	595	842	6
That,	134	214	159	224	595	842	6
it	163	212	168	224	595	842	6
is	172	212	180	224	595	842	6
known	184	212	215	224	595	842	6
as	219	212	231	224	595	842	6
“positive	235	212	276	224	595	842	6
correlation”.	280	212	338	224	595	842	6
Finally,	342	212	376	224	595	842	6
if	380	212	386	224	595	842	6
the	390	212	405	224	595	842	6
correlation	409	212	460	224	595	842	6
value	464	212	490	224	595	842	6
is	494	212	502	224	595	842	6
0,	506	212	515	224	595	842	6
it	519	212	524	224	595	842	6
indicates	85	226	129	237	595	842	6
that	132	226	150	237	595	842	6
there	154	226	179	237	595	842	6
is	182	226	190	237	595	842	6
no	194	226	206	237	595	842	6
correlation	209	226	261	237	595	842	6
between	264	226	306	237	595	842	6
pollutants.	309	226	359	237	595	842	6
This	363	226	383	237	595	842	6
means	387	226	420	237	595	842	6
that	423	226	441	237	595	842	6
it	445	226	450	237	595	842	6
is	454	226	462	237	595	842	6
not	465	226	480	237	595	842	6
possible	484	226	524	237	595	842	6
to	85	239	94	249	595	842	6
establish	97	239	141	249	595	842	6
senses	144	239	179	249	595	842	6
of	182	239	191	249	595	842	6
covariation.	194	239	250	249	595	842	6
In	113	250	123	262	595	842	6
order	126	250	152	262	595	842	6
to	155	250	165	262	595	842	6
achieve	168	250	206	262	595	842	6
the	209	250	225	262	595	842	6
third	228	250	249	262	595	842	6
goal	253	250	274	262	595	842	6
of	277	250	286	262	595	842	6
model	290	250	320	262	595	842	6
1,	323	250	332	262	595	842	6
“Select	336	250	370	262	595	842	6
by	373	250	385	262	595	842	6
Weights	388	250	428	262	595	842	6
process”	432	250	474	262	595	842	6
is	477	250	485	262	595	842	6
applied	489	250	524	262	595	842	6
to	85	264	94	275	595	842	6
the	99	264	114	275	595	842	6
correlation	119	264	170	275	595	842	6
matrix	175	264	205	275	595	842	6
to	210	264	219	275	595	842	6
identify	224	264	258	275	595	842	6
the	263	264	278	275	595	842	6
incidence	283	264	330	275	595	842	6
between	334	264	376	275	595	842	6
the	381	264	396	275	595	842	6
pollutants	401	264	448	275	595	842	6
represented	452	264	511	275	595	842	6
in	516	264	524	275	595	842	6
SelectbyWeights.	85	277	170	287	595	842	6
It	173	277	179	287	595	842	6
is	182	277	190	287	595	842	6
important	192	277	238	287	595	842	6
to	241	277	250	287	595	842	6
consider	253	277	295	287	595	842	6
that	297	277	316	287	595	842	6
a	318	277	325	287	595	842	6
pollutant	327	277	369	287	595	842	6
with	372	277	391	287	595	842	6
an	394	277	406	287	595	842	6
incidence	409	277	455	287	595	842	6
near	458	277	480	287	595	842	6
to	483	277	492	287	595	842	6
1	495	277	501	287	595	842	6
may	504	277	525	287	595	842	6
affect	85	290	112	300	595	842	6
another	115	290	152	300	595	842	6
pollutant	155	290	197	300	595	842	6
with	200	290	220	300	595	842	6
a	223	290	229	300	595	842	6
lower	232	290	258	300	595	842	6
value.	261	290	290	300	595	842	6
The	120	302	140	313	595	842	6
second	143	302	179	313	595	842	6
model	182	302	212	313	595	842	6
purpose	216	302	255	313	595	842	6
is	259	302	267	313	595	842	6
to	270	302	280	313	595	842	6
obtain	283	302	313	313	595	842	6
the	316	302	332	313	595	842	6
correlations	335	302	392	313	595	842	6
between	396	302	437	313	595	842	6
pollutants	441	302	488	313	595	842	6
in	491	302	500	313	595	842	6
time	503	302	524	313	595	842	6
intervals.	85	313	129	325	595	842	6
We	134	313	151	325	595	842	6
suggest	156	313	194	325	595	842	6
use	199	313	217	325	595	842	6
“Time	222	313	250	325	595	842	6
Rolling	255	313	289	325	595	842	6
Correlations”	294	313	357	325	595	842	6
also	362	313	382	325	595	842	6
denominated	387	313	451	325	595	842	6
“Time	456	313	484	325	595	842	6
Moving	489	313	524	325	595	842	6
Correlations”.	85	326	151	338	595	842	6
In	153	326	163	338	595	842	6
addition,	165	326	207	338	595	842	6
we	209	326	223	338	595	842	6
suggest	225	326	264	338	595	842	6
showing	266	326	307	338	595	842	6
the	309	326	325	338	595	842	6
pollutants	327	326	374	338	595	842	6
behavior	376	326	418	338	595	842	6
using	421	326	447	338	595	842	6
“Simple	449	326	487	338	595	842	6
Moving	489	326	524	338	595	842	6
Average	85	338	126	351	595	842	6
(SMA)”	131	338	166	351	595	842	6
technique	171	338	218	351	595	842	6
to	223	338	233	351	595	842	6
obtain	237	338	267	351	595	842	6
a	272	338	278	351	595	842	6
better	283	338	311	351	595	842	6
data	316	338	338	351	595	842	6
representation.	343	338	416	351	595	842	6
For	421	338	437	351	595	842	6
this,	442	338	462	351	595	842	6
the	467	338	483	351	595	842	6
k	487	338	493	351	595	842	6
value	498	338	524	351	595	842	6
returned	85	353	126	363	595	842	6
by	131	353	143	363	595	842	6
X-means	148	353	192	363	595	842	6
algorithm	197	351	242	363	595	842	6
in	247	351	255	363	595	842	6
model	260	351	291	363	595	842	6
1	295	351	302	363	595	842	6
as	307	351	318	363	595	842	6
“k”	323	351	336	363	595	842	6
value	341	351	367	363	595	842	6
is	372	351	380	363	595	842	6
used	385	351	409	363	595	842	6
as	414	351	426	363	595	842	6
priori	431	351	455	363	595	842	6
parameter	460	351	510	363	595	842	6
to	515	351	524	363	595	842	6
structure	85	366	128	376	595	842	6
the	133	366	149	376	595	842	6
clusters	154	366	192	376	595	842	6
using	198	366	224	376	595	842	6
the	229	366	245	376	595	842	6
K-means	250	366	294	376	595	842	6
algorithm.	299	366	348	376	595	842	6
Next,	353	366	379	376	595	842	6
data	384	366	406	376	595	842	6
filter	411	366	432	376	595	842	6
and	437	366	455	376	595	842	6
Time	461	366	485	376	595	842	6
Rolling	491	366	524	376	595	842	6
Correlations	85	378	145	388	595	842	6
techniques	152	378	205	388	595	842	6
are	213	378	229	388	595	842	6
applied	237	378	272	388	595	842	6
to	280	378	289	388	595	842	6
get	297	378	312	388	595	842	6
the	320	378	335	388	595	842	6
correlation	343	378	394	388	595	842	6
matrix.	402	378	435	388	595	842	6
Considering	443	378	501	388	595	842	6
the	509	378	524	388	595	842	6
procedures	85	391	140	401	595	842	6
addressed	143	391	195	401	595	842	6
in	198	391	206	401	595	842	6
the	209	391	225	401	595	842	6
model	228	391	258	401	595	842	6
1,	261	391	270	401	595	842	6
Figure	273	391	304	401	595	842	6
3,	307	391	316	401	595	842	6
shows	319	391	350	401	595	842	6
the	353	391	369	401	595	842	6
process	372	391	410	401	595	842	6
to	413	391	423	401	595	842	6
follow	426	391	454	401	595	842	6
in	457	391	465	401	595	842	6
model	468	391	498	401	595	842	6
2.	501	391	511	401	595	842	6
Figure	92	542	122	551	595	842	6
3:	125	542	134	551	595	842	6
Correlations	136	542	190	551	595	842	6
between	193	542	231	551	595	842	6
pollutants	236	542	279	551	595	842	6
in	285	542	292	551	595	842	6
time	298	542	317	551	595	842	6
intervals	323	542	360	551	595	842	6
(K-means)	362	542	409	551	595	842	6
and	412	542	429	551	595	842	6
data	432	542	451	551	595	842	6
representation	454	542	518	551	595	842	6
In	113	565	123	575	595	842	6
order	125	565	151	575	595	842	6
to	153	565	163	575	595	842	6
apply	165	565	191	575	595	842	6
the	194	565	209	575	595	842	6
K-Means	212	565	256	575	595	842	6
algorithm,	258	565	306	575	595	842	6
only	309	565	329	575	595	842	6
the	332	565	347	575	595	842	6
pollutants	349	565	396	575	595	842	6
columns	399	565	440	575	595	842	6
must	442	565	466	575	595	842	6
be	469	565	481	575	595	842	6
selected	483	565	524	575	595	842	6
represented	85	578	144	588	595	842	6
in	146	578	155	588	595	842	6
SelectedPollutants(1).	158	578	265	588	595	842	6
Once	268	578	294	588	595	842	6
applied	297	578	332	588	595	842	6
K-means	335	578	379	588	595	842	6
represented	381	578	440	588	595	842	6
in	442	578	451	588	595	842	6
K-means,	454	578	501	588	595	842	6
data	503	578	525	588	595	842	6
must	85	590	109	600	595	842	6
be	112	590	125	600	595	842	6
filtered	128	590	161	600	595	842	6
by	165	590	176	600	595	842	6
time	180	590	200	600	595	842	6
intervals.	204	590	248	600	595	842	6
It	251	590	257	600	595	842	6
is	261	590	269	600	595	842	6
represented	272	590	331	600	595	842	6
in	334	590	343	600	595	842	6
FilterByHours.	346	590	416	600	595	842	6
On	419	590	434	600	595	842	6
the	437	590	453	600	595	842	6
other	456	590	481	600	595	842	6
hand,	485	590	512	600	595	842	6
in	516	590	524	600	595	842	6
order	85	603	111	613	595	842	6
to	114	603	124	613	595	842	6
facilitate	127	603	168	613	595	842	6
the	171	603	186	613	595	842	6
data	190	603	211	613	595	842	6
representation	215	603	285	613	595	842	6
by	289	603	301	613	595	842	6
time	304	603	325	613	595	842	6
intervals,	329	603	372	613	595	842	6
an	376	603	388	613	595	842	6
identifier	392	603	433	613	595	842	6
is	437	603	445	613	595	842	6
assigned	449	603	493	613	595	842	6
to	496	603	505	613	595	842	6
the	509	603	524	613	595	842	6
rows	85	616	108	626	595	842	6
corresponding	112	616	181	626	595	842	6
to	185	616	194	626	595	842	6
every	197	616	224	626	595	842	6
pollutants	227	616	275	626	595	842	6
column.	278	616	317	626	595	842	6
That,	320	616	345	626	595	842	6
it	348	616	354	626	595	842	6
is	357	616	365	626	595	842	6
represented	368	616	427	626	595	842	6
in	430	616	439	626	595	842	6
SetRowID.	442	616	495	626	595	842	6
After,	498	616	524	626	595	842	6
in	85	628	94	638	595	842	6
order	98	628	124	638	595	842	6
to	129	628	138	638	595	842	6
recover	143	628	179	638	595	842	6
the	184	628	200	638	595	842	6
hour	204	628	226	638	595	842	6
filed,	231	628	254	638	595	842	6
both	259	628	280	638	595	842	6
SetRowID	285	628	335	638	595	842	6
and	339	628	358	638	595	842	6
FilterByHours	363	628	429	638	595	842	6
are	434	628	450	638	595	842	6
linked	454	628	483	638	595	842	6
in	488	628	496	638	595	842	6
Join.	501	628	524	638	595	842	6
Subsequently,	85	641	155	651	595	842	6
it	159	641	165	651	595	842	6
is	170	641	178	651	595	842	6
necessary	182	641	232	651	595	842	6
to	237	641	246	651	595	842	6
select	251	641	279	651	595	842	6
the	284	641	299	651	595	842	6
cluster	304	641	336	651	595	842	6
with	341	641	360	651	595	842	6
the	365	641	381	651	595	842	6
greater	385	641	420	651	595	842	6
elements	425	641	469	651	595	842	6
number	474	641	511	651	595	842	6
in	516	641	524	651	595	842	6
BiggestCluster.	85	653	160	664	595	842	6
Once	163	653	189	664	595	842	6
more,	192	653	221	664	595	842	6
only	224	653	244	664	595	842	6
pollutants	247	653	294	664	595	842	6
columns	298	653	339	664	595	842	6
are	342	653	358	664	595	842	6
selected	361	653	402	664	595	842	6
in	405	653	414	664	595	842	6
SelectedPollutants(2).	417	653	524	664	595	842	6
Finally,	85	664	120	676	595	842	6
in	126	664	135	676	595	842	6
order	141	664	167	676	595	842	6
to	173	664	183	676	595	842	6
generate	189	664	232	676	595	842	6
the	239	664	254	676	595	842	6
correlation	261	664	312	676	595	842	6
matrix	318	664	348	676	595	842	6
between	355	664	396	676	595	842	6
pollutants	403	664	450	676	595	842	6
“Time	456	664	484	676	595	842	6
Rolling	491	664	524	676	595	842	6
Correlations”	85	677	148	689	595	842	6
is	153	677	161	689	595	842	6
applied.	166	677	204	689	595	842	6
That	209	677	231	689	595	842	6
is	236	677	244	689	595	842	6
represented	249	677	308	689	595	842	6
in	312	677	321	689	595	842	6
CorrelationMatrix.	326	677	412	689	595	842	6
Unlike	417	677	448	689	595	842	6
the	453	677	468	689	595	842	6
correlation	473	677	524	689	595	842	6
matrix	85	691	115	702	595	842	6
in	118	691	127	702	595	842	6
model	130	691	160	702	595	842	6
1,	163	691	172	702	595	842	6
model	175	691	205	702	595	842	6
2	208	691	214	702	595	842	6
presents	217	691	260	702	595	842	6
result	263	691	289	702	595	842	6
in	293	691	301	702	595	842	6
Time	304	691	328	702	595	842	6
Rolling	331	691	365	702	595	842	6
Correlations.	368	691	431	702	595	842	6
Results	85	717	125	727	595	842	6
stage	128	717	157	727	595	842	6
Once	113	742	140	752	595	842	6
the	145	742	160	752	595	842	6
models	165	742	201	752	595	842	6
are	206	742	222	752	595	842	6
applied,	227	742	265	752	595	842	6
the	270	742	286	752	595	842	6
knowledge	290	742	343	752	595	842	6
gained	348	742	381	752	595	842	6
needs	386	742	416	752	595	842	6
to	421	742	430	752	595	842	6
be	435	742	448	752	595	842	6
organized	453	742	501	752	595	842	6
and	506	742	524	752	595	842	6
presented	85	755	134	765	595	842	6
in	137	755	146	765	595	842	6
a	149	755	155	765	595	842	6
way	158	755	178	765	595	842	6
that	181	755	199	765	595	842	6
environmental	202	755	271	765	595	842	6
managers	274	755	323	765	595	842	6
or	327	755	336	765	595	842	6
urban	339	755	368	765	595	842	6
planners	371	755	413	765	595	842	6
can	416	755	434	765	595	842	6
use	437	755	454	765	595	842	6
(Wirth,	458	755	490	765	595	842	6
2000).	494	755	525	765	595	842	6
Enfoque	220	778	250	785	595	842	6
UTE,	252	778	270	785	595	842	6
V.9-N.4,	272	778	302	785	595	842	6
Dec.2018,	304	778	340	785	595	842	6
pp.	343	778	354	785	595	842	6
168	356	778	369	785	595	842	6
-	372	778	374	785	595	842	6
179	377	778	390	785	595	842	6
174	71	72	84	80	595	842	7
Statistics	71	87	115	98	595	842	7
plots	119	87	142	98	595	842	7
are	146	87	162	98	595	842	7
useful	165	87	194	98	595	842	7
to	198	87	207	98	595	842	7
reflect	211	87	241	98	595	842	7
the	244	87	260	98	595	842	7
impact	263	87	296	98	595	842	7
of	299	87	308	98	595	842	7
one	312	87	330	98	595	842	7
pollutant	334	87	376	98	595	842	7
on	379	87	392	98	595	842	7
another	395	87	433	98	595	842	7
and	437	87	455	98	595	842	7
help	459	87	479	98	595	842	7
in	483	87	491	98	595	842	7
the	495	87	510	98	595	842	7
identification	71	100	132	110	595	842	7
of	135	100	144	110	595	842	7
the	147	100	163	110	595	842	7
pollutant	166	100	207	110	595	842	7
of	210	100	219	110	595	842	7
major	223	100	250	110	595	842	7
importance.	253	100	311	110	595	842	7
4.	71	125	80	135	595	842	7
Experimentation	83	125	169	135	595	842	7
To	106	151	119	161	595	842	7
evaluate	124	151	165	161	595	842	7
the	170	151	186	161	595	842	7
approach	190	151	236	161	595	842	7
proposed	241	151	287	161	595	842	7
in	291	151	300	161	595	842	7
this	304	151	322	161	595	842	7
paper,	326	151	357	161	595	842	7
an	362	151	374	161	595	842	7
experiment	379	151	433	161	595	842	7
using	438	151	464	161	595	842	7
the	469	151	484	161	595	842	7
data	489	151	510	161	595	842	7
collected	71	163	114	173	595	842	7
at	118	163	127	173	595	842	7
the	131	163	146	173	595	842	7
air	150	163	162	173	595	842	7
pollutants	166	163	213	173	595	842	7
monitoring	216	163	268	173	595	842	7
station	271	163	304	173	595	842	7
in	307	163	316	173	595	842	7
Cuenca,	320	163	361	173	595	842	7
Ecuador	364	163	405	173	595	842	7
was	409	163	428	173	595	842	7
conducted.	432	163	486	173	595	842	7
This	490	163	510	173	595	842	7
approach	71	176	117	186	595	842	7
was	122	176	142	186	595	842	7
evaluated	147	176	195	186	595	842	7
by	200	176	211	186	595	842	7
an	217	176	229	186	595	842	7
expert	234	176	265	186	595	842	7
in	270	176	278	186	595	842	7
environmental	283	176	353	186	595	842	7
management,	358	176	425	186	595	842	7
and	430	176	449	186	595	842	7
a	454	176	460	186	595	842	7
statistical	465	176	510	186	595	842	7
expert	71	189	101	199	595	842	7
along	105	189	132	199	595	842	7
with	136	189	156	199	595	842	7
researchers.	159	189	220	199	595	842	7
The	224	189	243	199	595	842	7
aim	247	189	265	199	595	842	7
of	268	189	278	199	595	842	7
the	281	189	297	199	595	842	7
experiment	300	189	355	199	595	842	7
was	359	189	378	199	595	842	7
to	382	189	391	199	595	842	7
evaluate	395	189	436	199	595	842	7
the	440	189	456	199	595	842	7
veracity	459	189	497	199	595	842	7
of	501	189	510	199	595	842	7
the	71	201	86	211	595	842	7
results	90	201	123	211	595	842	7
obtained	126	201	169	211	595	842	7
through	172	201	210	211	595	842	7
this	214	201	231	211	595	842	7
approach	235	201	281	211	595	842	7
and,	284	201	306	211	595	842	7
its	310	201	321	211	595	842	7
utility	324	201	349	211	595	842	7
for	353	201	366	211	595	842	7
environmental	370	201	439	211	595	842	7
management.	443	201	510	211	595	842	7
This	71	214	92	224	595	842	7
section	96	214	130	224	595	842	7
describes	134	214	181	224	595	842	7
in	185	214	193	224	595	842	7
detail	197	214	224	224	595	842	7
the	227	214	243	224	595	842	7
application	246	214	299	224	595	842	7
of	303	214	312	224	595	842	7
the	315	214	331	224	595	842	7
proposed	335	214	380	224	595	842	7
approach,	384	214	433	224	595	842	7
and	437	214	455	224	595	842	7
the	459	214	474	224	595	842	7
results	478	214	510	224	595	842	7
obtained	71	226	113	237	595	842	7
in	116	226	125	237	595	842	7
the	128	226	143	237	595	842	7
context	146	226	182	237	595	842	7
of	185	226	194	237	595	842	7
the	197	226	212	237	595	842	7
city	215	226	232	237	595	842	7
of	235	226	244	237	595	842	7
Cuenca.	247	226	288	237	595	842	7
Contextualizing	71	252	153	262	595	842	7
the	156	252	172	262	595	842	7
data	176	252	198	262	595	842	7
analysis	201	252	244	262	595	842	7
Cuenca	99	277	137	287	595	842	7
city	141	277	158	287	595	842	7
is	161	277	169	287	595	842	7
located	173	277	209	287	595	842	7
in	212	277	221	287	595	842	7
the	225	277	240	287	595	842	7
Andean	244	277	282	287	595	842	7
alley	286	277	308	287	595	842	7
of	312	277	321	287	595	842	7
South	325	277	354	287	595	842	7
America	357	277	398	287	595	842	7
at	402	277	411	287	595	842	7
2500	415	277	439	287	595	842	7
meters	443	277	477	287	595	842	7
above	481	277	510	287	595	842	7
the	71	290	86	300	595	842	7
sea	91	290	109	300	595	842	7
level.	114	290	140	300	595	842	7
Temperature	144	290	207	300	595	842	7
in	212	290	221	300	595	842	7
Cuenca	226	290	264	300	595	842	7
is	269	290	277	300	595	842	7
14.7°C	282	290	315	300	595	842	7
on	320	290	333	300	595	842	7
average,	338	290	380	300	595	842	7
which	385	290	413	300	595	842	7
is	418	290	426	300	595	842	7
classified	431	290	476	300	595	842	7
as	481	290	493	300	595	842	7
an	498	290	510	300	595	842	7
oceanic	71	302	109	313	595	842	7
climate	116	302	151	313	595	842	7
(Cfb)	158	302	182	313	595	842	7
in	189	302	198	313	595	842	7
the	205	302	220	313	595	842	7
Köppen	227	302	265	313	595	842	7
and	272	302	291	313	595	842	7
Geige	298	302	327	313	595	842	7
weather	334	302	373	313	595	842	7
map	380	302	402	313	595	842	7
(«Clima	416	302	454	313	595	842	7
CUENCA:	461	302	510	313	595	842	7
Temperatura,	71	315	137	325	595	842	7
Climograma	141	315	200	325	595	842	7
y	204	315	209	325	595	842	7
Tabla	213	315	240	325	595	842	7
climática	244	315	287	325	595	842	7
para	290	315	312	325	595	842	7
CUENCA	316	315	362	325	595	842	7
-	366	315	370	325	595	842	7
Climate-Data.org»,	374	315	466	325	595	842	7
s.	469	315	478	325	595	842	7
f.).	481	315	494	325	595	842	7
In	501	315	510	325	595	842	7
the	71	328	86	338	595	842	7
experimentation,	89	328	170	338	595	842	7
a	173	328	179	338	595	842	7
real	182	328	200	338	595	842	7
dataset	203	328	239	338	595	842	7
from	242	328	264	338	595	842	7
Cuenca	266	328	304	338	595	842	7
city	307	328	323	338	595	842	7
was	326	328	345	338	595	842	7
used.	348	328	375	338	595	842	7
The	378	328	397	338	595	842	7
air	400	328	412	338	595	842	7
pollutant	415	328	456	338	595	842	7
monitoring	459	328	510	338	595	842	7
station	71	340	103	351	595	842	7
from	108	340	130	351	595	842	7
Cuenca	134	340	172	351	595	842	7
records	177	340	213	351	595	842	7
variables	218	340	262	351	595	842	7
as	266	340	278	351	595	842	7
O3,	283	340	300	351	595	842	7
CO,	305	340	324	351	595	842	7
SO2,	329	340	354	351	595	842	7
NO2	358	340	381	351	595	842	7
and,	386	340	407	351	595	842	7
PM2.5.	411	340	446	351	595	842	7
Moreover,	451	340	500	351	595	842	7
it	505	340	510	351	595	842	7
reports	71	353	105	363	595	842	7
variables	110	353	154	363	595	842	7
of	159	353	169	363	595	842	7
precipitation;	174	353	236	363	595	842	7
radiation,	241	353	286	363	595	842	7
direction	292	353	333	363	595	842	7
and	338	353	357	363	595	842	7
wind	362	353	384	363	595	842	7
speed;	389	353	422	363	595	842	7
temperature	427	353	487	363	595	842	7
and	492	353	510	363	595	842	7
humidity.	71	366	115	376	595	842	7
Every	118	366	146	376	595	842	7
second,	149	366	188	376	595	842	7
the	191	366	206	376	595	842	7
data	209	366	230	376	595	842	7
is	233	366	241	376	595	842	7
stored	244	366	275	376	595	842	7
in	278	366	286	376	595	842	7
a	289	366	295	376	595	842	7
database	298	366	344	376	595	842	7
registering	347	366	398	376	595	842	7
864.000	401	366	441	376	595	842	7
records	444	366	481	376	595	842	7
every	483	366	510	376	595	842	7
day.	71	378	92	388	595	842	7
Preparing	71	403	122	414	595	842	7
data	125	403	148	414	595	842	7
stage	151	403	180	414	595	842	7
According	85	435	134	445	595	842	7
to	141	435	150	445	595	842	7
the	157	435	172	445	595	842	7
Data	179	435	202	445	595	842	7
preparation	209	435	265	445	595	842	7
stage	271	435	298	445	595	842	7
detailed	305	435	344	445	595	842	7
in	350	435	359	445	595	842	7
Section	366	435	402	445	595	842	7
III	409	435	418	445	595	842	7
the	425	435	440	445	595	842	7
data	447	435	469	445	595	842	7
for	475	435	488	445	595	842	7
the	495	435	510	445	595	842	7
experimentation	71	448	149	458	595	842	7
was	155	448	174	458	595	842	7
prepared	179	448	224	458	595	842	7
and	229	448	247	458	595	842	7
consisted	253	448	299	458	595	842	7
of	304	448	314	458	595	842	7
an	319	448	331	458	595	842	7
input	337	448	360	458	595	842	7
dataset	366	448	402	458	595	842	7
from	407	448	429	458	595	842	7
a	434	448	441	458	595	842	7
month	446	448	477	458	595	842	7
range	482	448	510	458	595	842	7
random	71	460	108	470	595	842	7
of	114	460	123	470	595	842	7
215.436	128	460	168	470	595	842	7
records	173	460	210	470	595	842	7
of	215	460	224	470	595	842	7
30	229	460	242	470	595	842	7
days	247	460	270	470	595	842	7
from	275	460	297	470	595	842	7
September	303	460	356	470	595	842	7
16	362	460	374	470	595	842	7
to	379	460	388	470	595	842	7
October	394	460	433	470	595	842	7
15,	438	460	453	470	595	842	7
2017.	459	460	486	470	595	842	7
The	491	460	510	470	595	842	7
measurements	71	473	144	483	595	842	7
were	148	473	171	483	595	842	7
obtained	175	473	217	483	595	842	7
every	221	473	248	483	595	842	7
minute.	252	473	288	483	595	842	7
The	292	473	311	483	595	842	7
data	315	473	337	483	595	842	7
was	340	473	360	483	595	842	7
corrected	364	473	410	483	595	842	7
through	413	473	451	483	595	842	7
the	455	473	470	483	595	842	7
formula	474	473	510	483	595	842	7
detailed	71	485	109	496	595	842	7
in	113	485	121	496	595	842	7
the	125	485	140	496	595	842	7
data	143	485	164	496	595	842	7
preparation	168	485	223	496	595	842	7
stage	227	485	254	496	595	842	7
in	257	485	265	496	595	842	7
Section	269	485	305	496	595	842	7
III.	308	485	321	496	595	842	7
Also,	324	485	348	496	595	842	7
we	351	485	365	496	595	842	7
seek	369	485	392	496	595	842	7
a	395	485	401	496	595	842	7
clean	405	485	431	496	595	842	7
sample	434	485	470	496	595	842	7
through	473	485	510	496	595	842	7
SQL	71	498	93	508	595	842	7
sentences.	96	498	149	508	595	842	7
Table	152	498	179	508	595	842	7
2	182	498	188	508	595	842	7
shows	191	498	222	508	595	842	7
the	224	498	240	508	595	842	7
record	243	498	274	508	595	842	7
number	276	498	314	508	595	842	7
obtained	316	498	359	508	595	842	7
after	361	498	383	508	595	842	7
every	386	498	413	508	595	842	7
step.	416	498	439	508	595	842	7
Once	442	498	469	508	595	842	7
the	471	498	487	508	595	842	7
data	489	498	510	508	595	842	7
preparation	71	511	127	521	595	842	7
was	130	511	149	521	595	842	7
finished,	152	511	193	521	595	842	7
we	196	511	210	521	595	842	7
obtained	213	511	255	521	595	842	7
a	258	511	264	521	595	842	7
dataset	267	511	303	521	595	842	7
with	306	511	326	521	595	842	7
25.877	329	511	362	521	595	842	7
records	365	511	402	521	595	842	7
of	405	511	414	521	595	842	7
17	417	511	429	521	595	842	7
days	432	511	455	521	595	842	7
from	458	511	480	521	595	842	7
00:00	483	511	510	521	595	842	7
am	71	524	86	534	595	842	7
to	89	524	99	534	595	842	7
24:00	102	524	129	534	595	842	7
pm.	132	524	150	534	595	842	7
Table	154	524	181	534	595	842	7
3	184	524	190	534	595	842	7
shows	193	524	224	534	595	842	7
the	227	524	243	534	595	842	7
initial	246	524	271	534	595	842	7
statistics	274	524	316	534	595	842	7
of	319	524	328	534	595	842	7
this	331	524	349	534	595	842	7
dataset.	352	524	391	534	595	842	7
Table	239	539	266	548	595	842	7
2:	268	539	277	548	595	842	7
Cleaning	280	539	319	548	595	842	7
data	322	539	342	548	595	842	7
Steps	193	551	220	560	595	842	7
Input	345	551	369	560	595	842	7
Output	425	551	458	560	595	842	7
Transformation	104	572	178	581	595	842	7
of	181	572	190	581	595	842	7
rows	193	572	216	581	595	842	7
to	219	572	228	581	595	842	7
columns	231	572	272	581	595	842	7
(pivot)	275	572	306	581	595	842	7
215.436	319	572	355	581	595	842	7
43.760	406	572	437	581	595	842	7
Deletion	104	593	144	602	595	842	7
of	147	593	156	602	595	842	7
null	159	593	177	602	595	842	7
values.	179	593	213	602	595	842	7
43.760	319	593	350	602	595	842	7
40.421	406	593	437	602	595	842	7
Days	104	615	128	624	595	842	7
with	131	615	151	624	595	842	7
records	154	615	190	624	595	842	7
mount	193	615	224	624	595	842	7
greater	227	615	260	624	595	842	7
than	263	615	284	624	595	842	7
99%	287	615	307	624	595	842	7
40.421	319	615	350	624	595	842	7
25.877	406	615	437	624	595	842	7
Table	158	640	184	650	595	842	7
3:	187	640	196	650	595	842	7
Summary	199	640	242	650	595	842	7
of	244	640	252	650	595	842	7
descriptive	255	640	303	650	595	842	7
statistic	306	640	339	650	595	842	7
for	342	640	354	650	595	842	7
data	356	640	376	650	595	842	7
used	378	640	400	650	595	842	7
in	403	640	411	650	595	842	7
study	413	640	437	650	595	842	7
Attribute	105	655	146	665	595	842	7
Minimum	207	655	251	665	595	842	7
Maximum	295	655	341	665	595	842	7
Average	381	655	420	665	595	842	7
SD	470	655	484	665	595	842	7
O3	71	674	85	684	595	842	7
(ug/m3)	88	674	124	684	595	842	7
0.000	191	674	216	684	595	842	7
140.274	278	674	314	684	595	842	7
30.074	368	674	399	684	595	842	7
25.835	444	674	475	684	595	842	7
CO	71	693	86	702	595	842	7
(mg/m3)	92	693	131	702	595	842	7
263.327	191	693	227	702	595	842	7
3698.027	278	693	320	702	595	842	7
873.58	368	693	399	702	595	842	7
403.52	444	693	475	702	595	842	7
NO2	71	713	92	722	595	842	7
(ug/m3)	95	713	131	722	595	842	7
0.000	191	713	216	722	595	842	7
94.818	278	713	309	722	595	842	7
18.098	368	713	399	722	595	842	7
15.067	444	713	475	722	595	842	7
SO2	71	732	91	741	595	842	7
(ug/m3)	94	732	130	741	595	842	7
0.000	191	732	216	741	595	842	7
88.244	278	732	309	741	595	842	7
8.177	368	732	393	741	595	842	7
8.171	444	732	469	741	595	842	7
PM2.5	71	751	100	760	595	842	7
(ug/m3)	103	751	139	760	595	842	7
0.000	191	751	216	760	595	842	7
72.000	278	751	309	760	595	842	7
14.836	368	751	399	760	595	842	7
13.181	444	751	475	760	595	842	7
Enfoque	206	778	236	785	595	842	7
UTE,	238	778	256	785	595	842	7
V.9-N.4,	258	778	288	785	595	842	7
Dec.2018,	290	778	326	785	595	842	7
pp.	328	778	339	785	595	842	7
168	342	778	355	785	595	842	7
-	357	778	360	785	595	842	7
179	362	778	376	785	595	842	7
175	511	72	525	80	595	842	8
Applying	85	87	132	97	595	842	8
the	135	87	152	97	595	842	8
model	155	87	187	97	595	842	8
and	190	87	210	97	595	842	8
results	213	87	249	97	595	842	8
stages	252	87	287	97	595	842	8
Based	100	104	131	116	595	842	8
in	135	104	143	116	595	842	8
the	147	104	163	116	595	842	8
first	166	104	184	116	595	842	8
model	188	104	218	116	595	842	8
“Data	222	104	249	116	595	842	8
grouping	253	106	296	116	595	842	8
cluster	299	106	332	116	595	842	8
and	336	106	354	116	595	842	8
incidence	358	106	404	116	595	842	8
between	408	106	450	116	595	842	8
pollutants”,	454	106	508	116	595	842	8
for	512	104	525	116	595	842	8
the	85	119	100	129	595	842	8
X-means	103	119	148	129	595	842	8
algorithm	150	119	196	129	595	842	8
was	199	119	218	129	595	842	8
necessary	221	119	271	129	595	842	8
to	274	119	284	129	595	842	8
define	286	119	317	129	595	842	8
the	319	119	335	129	595	842	8
N	338	119	346	129	595	842	8
max	349	119	370	129	595	842	8
of	373	119	382	129	595	842	8
range	385	119	413	129	595	842	8
values	416	119	447	129	595	842	8
to	451	119	460	129	595	842	8
generate	463	119	506	129	595	842	8
the	509	119	525	129	595	842	8
clusters	85	131	123	142	595	842	8
number	128	131	165	142	595	842	8
p	169	131	176	142	595	842	8
from	180	131	202	142	595	842	8
a	206	131	213	142	595	842	8
range	217	131	245	142	595	842	8
2	250	129	256	142	595	842	8
≤	260	129	266	142	595	842	8
p(min)	271	129	302	142	595	842	8
≤	307	129	313	142	595	842	8
p	317	129	323	142	595	842	8
≤	328	129	334	142	595	842	8
p(max)	338	129	372	142	595	842	8
≤	377	129	383	142	595	842	8
N	388	129	396	142	595	842	8
−1	400	129	413	142	595	842	8
[21].	417	131	439	142	595	842	8
To	443	131	456	142	595	842	8
define	461	131	491	142	595	842	8
N,	495	131	506	142	595	842	8
we	511	131	525	142	595	842	8
consulted	85	144	132	154	595	842	8
an	135	144	148	154	595	842	8
expert	151	144	181	154	595	842	8
in	184	144	193	154	595	842	8
air	196	144	208	154	595	842	8
pollution.	211	144	255	154	595	842	8
Thus,	258	144	286	154	595	842	8
the	288	144	304	154	595	842	8
values	307	144	338	154	595	842	8
were	342	144	365	154	595	842	8
defined	368	144	405	154	595	842	8
as	408	144	419	154	595	842	8
N=8	422	144	443	154	595	842	8
and	446	144	464	154	595	842	8
the	467	144	483	154	595	842	8
result	486	144	512	154	595	842	8
of	515	144	524	154	595	842	8
X-means	85	157	129	167	595	842	8
execution	132	157	179	167	595	842	8
returned	182	157	223	167	595	842	8
p=k=4.	226	157	261	167	595	842	8
Figure	113	169	145	179	595	842	8
4,	148	169	157	179	595	842	8
presents	160	169	202	179	595	842	8
the	205	169	220	179	595	842	8
general	223	169	260	179	595	842	8
behavior	263	169	305	179	595	842	8
of	308	169	317	179	595	842	8
all	320	169	331	179	595	842	8
dataset	334	169	370	179	595	842	8
pollutants.	373	169	423	179	595	842	8
On	426	169	441	179	595	842	8
the	443	169	459	179	595	842	8
diagonal,	462	169	506	179	595	842	8
the	509	169	524	179	595	842	8
plot	85	182	103	192	595	842	8
presents	107	182	149	192	595	842	8
the	153	182	168	192	595	842	8
data	172	182	193	192	595	842	8
distribution.	197	182	253	192	595	842	8
In	257	182	266	192	595	842	8
the	270	182	286	192	595	842	8
lower	289	182	315	192	595	842	8
part	319	182	338	192	595	842	8
of	342	182	351	192	595	842	8
the	355	182	370	192	595	842	8
diagonal,	374	182	419	192	595	842	8
the	422	182	438	192	595	842	8
plot	442	182	459	192	595	842	8
presents	463	182	505	192	595	842	8
the	509	182	524	192	595	842	8
data	85	195	107	205	595	842	8
dispersion.	110	195	163	205	595	842	8
At	166	195	176	205	595	842	8
the	179	195	195	205	595	842	8
top	198	195	213	205	595	842	8
of	216	195	225	205	595	842	8
the	228	195	244	205	595	842	8
diagonal,	247	195	291	205	595	842	8
the	294	195	310	205	595	842	8
plot	313	195	330	205	595	842	8
presents	333	195	376	205	595	842	8
the	379	195	394	205	595	842	8
correlation	397	195	448	205	595	842	8
values.	451	195	486	205	595	842	8
Figure	197	407	228	416	595	842	8
4:	230	407	239	416	595	842	8
Model	242	407	269	416	595	842	8
1-	272	407	281	416	595	842	8
Pollutants	284	407	328	416	595	842	8
correlations	331	407	382	416	595	842	8
matrix	385	407	412	416	595	842	8
In	120	431	130	441	595	842	8
model	135	431	165	441	595	842	8
1,	170	431	179	441	595	842	8
we	185	431	199	441	595	842	8
also	204	431	224	441	595	842	8
seek	230	431	253	441	595	842	8
to	258	431	268	441	595	842	8
identify	273	431	308	441	595	842	8
the	313	431	328	441	595	842	8
most	334	431	357	441	595	842	8
relevant	363	431	402	441	595	842	8
pollutant.	407	431	452	441	595	842	8
For	457	431	474	441	595	842	8
that,	479	431	500	441	595	842	8
it	506	431	511	441	595	842	8
is	517	431	524	441	595	842	8
necessary	85	444	135	454	595	842	8
to	138	444	147	454	595	842	8
know	150	444	176	454	595	842	8
the	179	444	194	454	595	842	8
incidence	197	444	243	454	595	842	8
among	246	444	280	454	595	842	8
pollutants.	282	444	333	454	595	842	8
Table	335	444	363	454	595	842	8
4	366	444	372	454	595	842	8
shows	375	444	406	454	595	842	8
the	409	444	424	454	595	842	8
incidence	427	444	473	454	595	842	8
among	476	444	509	454	595	842	8
air	512	444	525	454	595	842	8
pollutants.	85	456	135	466	595	842	8
Table	205	475	232	484	595	842	8
4:	234	475	243	484	595	842	8
Model	246	475	273	484	595	842	8
1	276	475	282	484	595	842	8
–incidence	284	473	332	484	595	842	8
between	335	475	373	484	595	842	8
pollutants	376	475	418	484	595	842	8
Pollutant	234	490	277	499	595	842	8
O3	249	504	262	514	595	842	8
NO2	245	518	266	527	595	842	8
CO	248	531	263	541	595	842	8
SO2	245	545	265	554	595	842	8
PM2.5	241	559	270	568	595	842	8
Weight	337	490	371	499	595	842	8
1.000	341	504	366	514	595	842	8
0.323	341	518	366	527	595	842	8
0.178	341	531	366	541	595	842	8
0.103	341	545	366	554	595	842	8
0.000	341	559	366	568	595	842	8
According	113	585	162	595	595	842	8
to	165	585	175	595	595	842	8
Table	178	585	205	595	595	842	8
4,	208	585	217	595	595	842	8
O3	220	585	235	595	595	842	8
was	238	585	258	595	595	842	8
the	261	585	276	595	595	842	8
most	279	585	303	595	595	842	8
relevant	306	585	345	595	595	842	8
pollutant	348	585	389	595	595	842	8
in	393	585	401	595	595	842	8
Cuenca	404	585	442	595	595	842	8
city.	445	585	465	595	595	842	8
This	468	585	488	595	595	842	8
means	491	585	524	595	595	842	8
that	85	598	104	608	595	842	8
the	108	598	124	608	595	842	8
O3	128	598	143	608	595	842	8
pollutant	147	598	189	608	595	842	8
had	193	598	212	608	595	842	8
greater	216	598	251	608	595	842	8
repercussion	256	598	319	608	595	842	8
on	323	598	335	608	595	842	8
other	340	598	365	608	595	842	8
pollutants.	370	598	420	608	595	842	8
No	424	598	438	608	595	842	8
pollutant	443	598	485	608	595	842	8
had	489	598	508	608	595	842	8
an	512	598	524	608	595	842	8
incidence	85	610	132	620	595	842	8
on	135	610	148	620	595	842	8
PM2.5	151	610	183	620	595	842	8
In	186	610	196	620	595	842	8
model	199	610	229	620	595	842	8
2,	233	610	242	620	595	842	8
the	246	610	261	620	595	842	8
p-value	265	610	301	620	595	842	8
was	304	610	324	620	595	842	8
used	328	610	351	620	595	842	8
as	355	610	367	620	595	842	8
the	370	610	385	620	595	842	8
K	389	610	396	620	595	842	8
parameter	400	610	450	620	595	842	8
to	454	610	463	620	595	842	8
improve	466	610	505	620	595	842	8
the	509	610	524	620	595	842	8
K-mean	85	623	124	633	595	842	8
efficiency	127	623	173	633	595	842	8
(Kingsy	176	623	213	633	595	842	8
et	216	623	225	633	595	842	8
al.,	228	623	243	633	595	842	8
2016).	246	623	277	633	595	842	8
Once	281	623	307	633	595	842	8
applied	310	623	346	633	595	842	8
the	349	623	364	633	595	842	8
K-means	367	623	412	633	595	842	8
algorithm,	415	623	463	633	595	842	8
we	466	623	480	633	595	842	8
selected	484	623	525	633	595	842	8
the	85	636	100	646	595	842	8
biggest	103	636	139	646	595	842	8
cluster	141	636	174	646	595	842	8
and	177	636	195	646	595	842	8
followed	198	636	238	646	595	842	8
the	241	636	256	646	595	842	8
suggestion	259	636	312	646	595	842	8
of	315	636	324	646	595	842	8
the	327	636	342	646	595	842	8
approach	345	636	391	646	595	842	8
in	394	636	402	646	595	842	8
modelling	405	636	452	646	595	842	8
stage,	455	636	485	646	595	842	8
Section	488	636	524	646	595	842	8
III,	85	648	97	658	595	842	8
we	101	648	115	658	595	842	8
proceeded	119	648	171	658	595	842	8
to	174	648	183	658	595	842	8
analyze	187	648	225	658	595	842	8
the	228	648	244	658	595	842	8
correlation	247	648	299	658	595	842	8
by	302	648	314	658	595	842	8
Time	317	648	342	658	595	842	8
Rolling	346	648	379	658	595	842	8
Correlations.	383	648	445	658	595	842	8
Our	449	648	467	658	595	842	8
aim	471	648	489	658	595	842	8
was	492	648	512	658	595	842	8
to	515	648	525	658	595	842	8
verify	85	661	111	671	595	842	8
the	114	661	130	671	595	842	8
continuous	132	661	185	671	595	842	8
behavior	188	661	230	671	595	842	8
among	233	661	267	671	595	842	8
pollutants	269	661	317	671	595	842	8
throughout	319	661	372	671	595	842	8
the	374	661	390	671	595	842	8
day.	393	661	413	671	595	842	8
Thus,	416	661	444	671	595	842	8
from	446	661	468	671	595	842	8
the	471	661	486	671	595	842	8
original	489	661	524	671	595	842	8
dataset,	85	674	124	684	595	842	8
we	129	674	143	684	595	842	8
analyzed	148	674	192	684	595	842	8
the	197	674	212	684	595	842	8
correlations	217	674	274	684	595	842	8
using	279	674	305	684	595	842	8
Time	310	674	334	684	595	842	8
Rolling	339	674	372	684	595	842	8
Correlations	377	674	437	684	595	842	8
between	441	674	483	684	595	842	8
60-time	488	674	525	684	595	842	8
series	85	686	115	696	595	842	8
where	118	686	148	696	595	842	8
1-time	152	686	183	696	595	842	8
series	186	686	216	696	595	842	8
is	220	686	227	696	595	842	8
equivalent	231	686	281	696	595	842	8
to	285	686	295	696	595	842	8
1	298	686	305	696	595	842	8
minute	308	686	341	696	595	842	8
and	345	686	364	696	595	842	8
Rolling	367	686	401	696	595	842	8
(moving)	405	686	448	696	595	842	8
it	451	686	457	696	595	842	8
by	461	686	472	696	595	842	8
10	476	686	488	696	595	842	8
series.	492	686	524	696	595	842	8
Finally,	85	699	120	709	595	842	8
we	123	699	137	709	595	842	8
show	140	699	166	709	595	842	8
a	168	699	175	709	595	842	8
row	177	699	195	709	595	842	8
per	198	699	214	709	595	842	8
pollutants	217	699	264	709	595	842	8
correlation	267	699	318	709	595	842	8
and	321	699	339	709	595	842	8
time.	342	699	366	709	595	842	8
Table	369	699	397	709	595	842	8
5,	400	699	409	709	595	842	8
show	412	699	437	709	595	842	8
a	440	699	446	709	595	842	8
summary	449	699	494	709	595	842	8
of	497	699	506	709	595	842	8
the	509	699	525	709	595	842	8
correlation	85	711	137	722	595	842	8
matrix	140	711	170	722	595	842	8
among	173	711	206	722	595	842	8
pollutants	209	711	256	722	595	842	8
at	259	711	269	722	595	842	8
the	272	711	287	722	595	842	8
beginning	290	711	338	722	595	842	8
of	341	711	350	722	595	842	8
a	353	711	359	722	595	842	8
day.	362	711	383	722	595	842	8
Enfoque	220	778	250	785	595	842	8
UTE,	252	778	270	785	595	842	8
V.9-N.4,	272	778	302	785	595	842	8
Dec.2018,	304	778	340	785	595	842	8
pp.	343	778	354	785	595	842	8
168	356	778	369	785	595	842	8
-	372	778	374	785	595	842	8
179	377	778	390	785	595	842	8
176	71	72	84	80	595	842	9
Table	120	87	146	96	595	842	9
5:	149	87	158	96	595	842	9
Model	161	87	188	96	595	842	9
2	191	87	196	96	595	842	9
–	199	85	204	96	595	842	9
Correlation	207	87	256	96	595	842	9
matrix	259	87	286	96	595	842	9
by	289	87	300	96	595	842	9
time	302	87	321	96	595	842	9
rolling	324	87	351	96	595	842	9
correlations.	353	87	408	96	595	842	9
Star	410	87	429	96	595	842	9
of	432	87	440	96	595	842	9
the	443	87	457	96	595	842	9
day	459	87	476	96	595	842	9
TIME	116	103	131	113	595	842	9
O3-NO2	145	103	170	113	595	842	9
O3-SO2	183	103	206	113	595	842	9
O3-CO	220	103	241	113	595	842	9
O3-PM2.5	251	103	282	113	595	842	9
NO2-SO2	289	103	317	113	595	842	9
NO2-CO	327	103	352	113	595	842	9
NO2-PM2.5	357	103	393	113	595	842	9
SO2-CO	400	103	423	113	595	842	9
SO2-PM2.5	431	103	465	113	595	842	9
CO-PM2.5	469	103	500	113	595	842	9
00:00:00	111	117	137	127	595	842	9
-0.605	156	117	175	127	595	842	9
0.291	195	117	211	127	595	842	9
-0.392	228	117	247	127	595	842	9
-0.344	265	117	284	127	595	842	9
-0.092	301	117	320	127	595	842	9
0.497	340	117	356	127	595	842	9
0.345	376	117	392	127	595	842	9
0.291	412	117	429	127	595	842	9
-0.245	446	117	465	127	595	842	9
-0.151	482	117	501	127	595	842	9
00:10:00	111	130	137	141	595	842	9
-0.576	156	130	175	141	595	842	9
0.260	195	130	211	141	595	842	9
-0.394	228	130	247	141	595	842	9
-0.313	265	130	284	141	595	842	9
-0.083	301	130	320	141	595	842	9
0.497	340	130	356	141	595	842	9
0.329	376	130	392	141	595	842	9
0.307	412	130	429	141	595	842	9
-0.232	446	130	465	141	595	842	9
-0.107	482	130	501	141	595	842	9
00:20:00	111	144	137	154	595	842	9
-0.567	156	144	175	154	595	842	9
0.214	195	144	211	154	595	842	9
-0.371	228	144	247	154	595	842	9
-0.302	265	144	284	154	595	842	9
-0.104	301	144	320	154	595	842	9
0.464	340	144	356	154	595	842	9
0.299	376	144	392	154	595	842	9
0.320	412	144	429	154	595	842	9
-0.182	446	144	465	154	595	842	9
-0.073	482	144	501	154	595	842	9
00:30:00	111	158	137	168	595	842	9
-0.559	156	158	175	168	595	842	9
-0.344	228	158	247	168	595	842	9
-0.303	265	158	284	168	595	842	9
-0.144	301	158	320	168	595	842	9
0.443	340	158	356	168	595	842	9
0.319	376	158	392	168	595	842	9
0.315	412	158	429	168	595	842	9
-0.083	446	158	465	168	595	842	9
0.022	485	158	501	168	595	842	9
00:40:00	111	172	137	182	595	842	9
-0.580	156	172	175	182	595	842	9
0.085	195	172	211	182	595	842	9
-0.328	228	172	247	182	595	842	9
-0.312	265	172	284	182	595	842	9
-0.099	301	172	320	182	595	842	9
0.443	340	172	356	182	595	842	9
0.329	376	172	392	182	595	842	9
0.352	412	172	429	182	595	842	9
0.054	448	172	465	182	595	842	9
0.132	485	172	501	182	595	842	9
00:50:00	111	185	137	196	595	842	9
-0.582	156	185	175	196	595	842	9
0.042	195	185	211	196	595	842	9
-0.291	228	185	247	196	595	842	9
-0.278	265	185	284	196	595	842	9
-0.094	301	185	320	196	595	842	9
0.403	340	185	356	196	595	842	9
0.261	376	185	392	196	595	842	9
0.382	412	185	429	196	595	842	9
0.150	448	185	465	196	595	842	9
0.123	485	185	501	196	595	842	9
01:00:00	111	199	137	210	595	842	9
-0.572	156	199	175	210	595	842	9
0.062	195	199	211	210	595	842	9
-0.241	228	199	247	210	595	842	9
-0.196	265	199	284	210	595	842	9
-0.077	301	199	320	210	595	842	9
0.367	340	199	356	210	595	842	9
0.143	376	199	392	210	595	842	9
0.324	412	199	429	210	595	842	9
0.173	448	199	465	210	595	842	9
0.079	485	199	501	210	595	842	9
01:10:00	111	213	137	223	595	842	9
-0.583	156	213	175	223	595	842	9
0.086	195	213	211	223	595	842	9
-0.208	228	213	247	223	595	842	9
-0.127	265	213	284	223	595	842	9
-0.009	301	213	320	223	595	842	9
0.360	340	213	356	223	595	842	9
0.083	376	213	392	223	595	842	9
0.296	412	213	429	223	595	842	9
0.156	448	213	465	223	595	842	9
0.124	485	213	501	223	595	842	9
01:20:00	111	227	137	237	595	842	9
-0.600	156	227	175	237	595	842	9
0.129	195	227	211	237	595	842	9
-0.196	228	227	247	237	595	842	9
-0.038	265	227	284	237	595	842	9
0.030	303	227	320	237	595	842	9
0.347	340	227	356	237	595	842	9
0.031	376	227	392	237	595	842	9
0.250	412	227	429	237	595	842	9
0.101	448	227	465	237	595	842	9
0.182	485	227	501	237	595	842	9
01:30:00	111	241	137	251	595	842	9
-0.588	156	241	175	251	595	842	9
0.197	195	241	211	251	595	842	9
-0.171	228	241	247	251	595	842	9
0.023	267	241	284	251	595	842	9
0.069	303	241	320	251	595	842	9
0.295	340	241	356	251	595	842	9
0.009	376	241	392	251	595	842	9
0.189	412	241	429	251	595	842	9
0.035	448	241	465	251	595	842	9
0.231	485	241	501	251	595	842	9
01:40:00	111	254	137	265	595	842	9
-0.558	156	254	175	265	595	842	9
0.284	195	254	211	265	595	842	9
-0.116	228	254	247	265	595	842	9
0.099	267	254	284	265	595	842	9
0.059	303	254	320	265	595	842	9
0.200	340	254	356	265	595	842	9
-0.010	373	254	392	265	595	842	9
0.136	412	254	429	265	595	842	9
-0.027	446	254	465	265	595	842	9
0.265	485	254	501	265	595	842	9
01:50:00	111	268	137	278	595	842	9
-0.544	156	268	175	278	595	842	9
0.352	195	268	211	278	595	842	9
-0.070	228	268	247	278	595	842	9
0.158	267	268	284	278	595	842	9
0.040	303	268	320	278	595	842	9
0.106	340	268	356	278	595	842	9
-0.007	373	268	392	278	595	842	9
0.060	412	268	429	278	595	842	9
-0.069	446	268	465	278	595	842	9
0.325	485	268	501	278	595	842	9
02:00:00	111	282	137	292	595	842	9
-0.530	156	282	175	292	595	842	9
0.375	195	282	211	292	595	842	9
-0.046	228	282	247	292	595	842	9
0.205	267	282	284	292	595	842	9
0.030	303	282	320	292	595	842	9
0.054	340	282	356	292	595	842	9
0.005	376	282	392	292	595	842	9
0.033	412	282	429	292	595	842	9
-0.053	446	282	465	292	595	842	9
0.430	485	282	501	292	595	842	9
Table	99	313	127	323	595	842	9
5	130	313	137	323	595	842	9
shows	140	313	171	323	595	842	9
the	175	313	190	323	595	842	9
behavior	194	313	236	323	595	842	9
with	240	313	259	323	595	842	9
the	263	313	278	323	595	842	9
greatest	282	313	321	323	595	842	9
variation	325	313	367	323	595	842	9
among	370	313	404	323	595	842	9
the	408	313	423	323	595	842	9
correlations.	426	313	486	323	595	842	9
Due	490	313	510	323	595	842	9
to	71	325	80	335	595	842	9
that	83	325	101	335	595	842	9
a	104	325	110	335	595	842	9
value	112	325	139	335	595	842	9
in	141	325	150	335	595	842	9
the	152	325	168	335	595	842	9
correlations	170	325	227	335	595	842	9
is	230	325	237	335	595	842	9
more	240	325	265	335	595	842	9
representative	268	325	337	335	595	842	9
throughout	340	325	393	335	595	842	9
the	395	325	410	335	595	842	9
day;	413	325	434	335	595	842	9
Table	436	325	464	335	595	842	9
5	467	325	473	335	595	842	9
reflects	475	325	510	335	595	842	9
more	71	337	96	347	595	842	9
significant	99	337	148	347	595	842	9
data,	150	337	175	347	595	842	9
than	178	337	199	347	595	842	9
data	202	337	223	347	595	842	9
in	226	337	235	347	595	842	9
Figure	237	337	269	347	595	842	9
4.	271	337	281	347	595	842	9
Moreover,	283	337	333	347	595	842	9
these	336	337	363	347	595	842	9
tables	365	337	395	347	595	842	9
reflect	397	337	427	347	595	842	9
that	430	337	448	347	595	842	9
it	451	337	457	347	595	842	9
is	459	337	467	347	595	842	9
possible	470	337	510	347	595	842	9
to	71	349	80	359	595	842	9
obtain	83	349	113	359	595	842	9
a	116	349	122	359	595	842	9
better	126	349	154	359	595	842	9
performance	157	349	218	359	595	842	9
of	221	349	231	359	595	842	9
K-means	234	349	278	359	595	842	9
in	281	349	289	359	595	842	9
comparison	292	349	349	359	595	842	9
with	352	349	372	359	595	842	9
results	375	349	407	359	595	842	9
in	410	349	419	359	595	842	9
Figure	422	349	453	359	595	842	9
4.	456	349	465	359	595	842	9
On	99	361	114	371	595	842	9
the	121	361	136	371	595	842	9
other	143	361	168	371	595	842	9
hand,	175	361	202	371	595	842	9
in	209	361	218	371	595	842	9
order	224	361	250	371	595	842	9
to	257	361	266	371	595	842	9
present	273	361	309	371	595	842	9
the	316	361	331	371	595	842	9
knowledge	338	361	391	371	595	842	9
gained	397	361	431	371	595	842	9
in	437	361	446	371	595	842	9
a	453	361	459	371	595	842	9
way	465	361	485	371	595	842	9
that	492	361	510	371	595	842	9
environmental	71	373	140	383	595	842	9
managers	144	373	193	383	595	842	9
or	197	373	207	383	595	842	9
urban	210	373	239	383	595	842	9
planners	242	373	285	383	595	842	9
can	289	373	306	383	595	842	9
use,	310	373	331	383	595	842	9
we	335	373	349	383	595	842	9
generate	353	373	396	383	595	842	9
ten	400	373	415	383	595	842	9
plots	419	373	442	383	595	842	9
created	446	373	483	383	595	842	9
by	487	373	499	383	595	842	9
R	502	373	510	383	595	842	9
software.	71	385	116	395	595	842	9
Figures	119	385	155	395	595	842	9
5,	158	385	167	395	595	842	9
6	170	385	176	395	595	842	9
and,	179	385	200	395	595	842	9
7	203	385	210	395	595	842	9
are	212	385	228	395	595	842	9
the	231	385	247	395	595	842	9
most	249	385	273	395	595	842	9
relevant	276	385	315	395	595	842	9
plots.	318	385	344	395	595	842	9
These	347	385	378	395	595	842	9
figures	381	385	414	395	595	842	9
show	417	385	442	395	595	842	9
the	445	385	460	395	595	842	9
pollutants	463	385	510	395	595	842	9
dispersion	71	397	121	407	595	842	9
through	124	397	161	407	595	842	9
smoothed	164	397	212	407	595	842	9
curves	215	397	247	407	595	842	9
throughout	250	397	303	407	595	842	9
the	306	397	321	407	595	842	9
day.	324	397	344	407	595	842	9
These	347	397	378	407	595	842	9
figures	380	397	414	407	595	842	9
are	416	397	432	407	595	842	9
divided	435	397	470	407	595	842	9
into	473	397	491	407	595	842	9
two	493	397	510	407	595	842	9
parts:	71	409	98	419	595	842	9
(a)	102	409	115	419	595	842	9
reflect	119	409	149	419	595	842	9
the	152	409	167	419	595	842	9
pollutants	171	409	218	419	595	842	9
behavior	221	409	263	419	595	842	9
and,	267	409	288	419	595	842	9
(b)	291	409	305	419	595	842	9
reflect	308	409	338	419	595	842	9
the	341	409	357	419	595	842	9
correlation	360	409	412	419	595	842	9
between	415	409	456	419	595	842	9
pollutants.	460	409	510	419	595	842	9
In	71	421	80	431	595	842	9
order	84	421	110	431	595	842	9
to	114	421	123	431	595	842	9
help	127	421	147	431	595	842	9
to	151	421	160	431	595	842	9
compare	164	421	207	431	595	842	9
the	211	421	226	431	595	842	9
correlations	230	421	287	431	595	842	9
values,	291	421	325	431	595	842	9
the	329	421	345	431	595	842	9
first	348	421	366	431	595	842	9
part	370	421	389	431	595	842	9
of	393	421	402	431	595	842	9
the	406	421	421	431	595	842	9
figures	425	421	458	431	595	842	9
(a)	462	421	475	431	595	842	9
shows	479	421	510	431	595	842	9
each	71	433	95	443	595	842	9
pollutant	98	433	140	443	595	842	9
behavior	143	433	185	443	595	842	9
using	188	433	214	443	595	842	9
Simple	217	433	251	443	595	842	9
Moving.	254	433	292	443	595	842	9
Figure	211	619	241	628	595	842	9
5:	244	619	253	628	595	842	9
Experimental	256	619	314	628	595	842	9
results	317	619	346	628	595	842	9
NO2	349	619	370	628	595	842	9
-	372	619	376	628	595	842	9
SO2	379	619	399	628	595	842	9
From	99	641	125	651	595	842	9
the	131	641	146	651	595	842	9
results	152	641	185	651	595	842	9
obtained,	191	641	236	651	595	842	9
the	242	641	257	651	595	842	9
evaluators	263	641	314	651	595	842	9
highlighted	320	641	373	651	595	842	9
that	379	641	397	651	595	842	9
the	403	641	418	651	595	842	9
approach	424	641	470	651	595	842	9
returns	476	641	510	651	595	842	9
information	71	653	125	663	595	842	9
close	133	653	159	663	595	842	9
to	167	653	176	663	595	842	9
the	184	653	199	663	595	842	9
reality	207	653	237	663	595	842	9
of	245	653	254	663	595	842	9
Cuenca.	262	653	303	663	595	842	9
The	311	653	330	663	595	842	9
evaluator	338	653	383	663	595	842	9
team	391	653	415	663	595	842	9
consisted	423	653	470	663	595	842	9
in:	477	653	489	663	595	842	9
An	497	653	510	663	595	842	9
Environmental	71	666	141	676	595	842	9
Management	144	666	208	676	595	842	9
Specialist	211	666	258	676	595	842	9
that	261	666	280	676	595	842	9
correlated	283	666	332	676	595	842	9
results	334	666	367	676	595	842	9
generated	370	666	419	676	595	842	9
in	422	666	431	676	595	842	9
this	434	666	451	676	595	842	9
study	454	666	480	676	595	842	9
to	483	666	492	676	595	842	9
the	495	666	510	676	595	842	9
records	71	678	108	689	595	842	9
officially	115	678	154	689	595	842	9
published	161	678	208	689	595	842	9
by	215	678	227	689	595	842	9
the	234	678	250	689	595	842	9
local	257	678	279	689	595	842	9
air	287	678	299	689	595	842	9
monitoring	306	678	357	689	595	842	9
office	365	678	391	689	595	842	9
in	398	678	407	689	595	842	9
their	414	678	435	689	595	842	9
yearly	442	678	472	689	595	842	9
official	479	678	510	689	595	842	9
publications.	71	691	132	701	595	842	9
The	136	691	155	701	595	842	9
Statistics	159	691	203	701	595	842	9
Specialist	206	691	253	701	595	842	9
that	257	691	276	701	595	842	9
assessed	280	691	326	701	595	842	9
and	330	691	348	701	595	842	9
validated	352	691	396	701	595	842	9
the	400	691	415	701	595	842	9
statistical	419	691	464	701	595	842	9
methods	468	691	510	701	595	842	9
used	71	704	95	714	595	842	9
and	98	704	116	714	595	842	9
results	119	704	151	714	595	842	9
obtained.	154	704	200	714	595	842	9
The	205	704	224	714	595	842	9
information	227	704	282	714	595	842	9
returned	285	704	325	714	595	842	9
by	328	704	340	714	595	842	9
the	343	704	358	714	595	842	9
proposed	361	704	407	714	595	842	9
approach	410	704	456	714	595	842	9
was	459	704	478	714	595	842	9
useful	481	704	510	714	595	842	9
to	71	716	80	727	595	842	9
support	84	716	121	727	595	842	9
in	125	716	133	727	595	842	9
the	137	716	153	727	595	842	9
air	157	716	169	727	595	842	9
quality	173	716	205	727	595	842	9
evaluation	209	716	259	727	595	842	9
in	263	716	271	727	595	842	9
Cuenca.	275	716	316	727	595	842	9
Therefore,	320	716	371	727	595	842	9
we	375	716	389	727	595	842	9
claim	393	716	419	727	595	842	9
that	423	716	441	727	595	842	9
the	445	716	460	727	595	842	9
approach	464	716	510	727	595	842	9
can	71	729	89	739	595	842	9
be	92	729	104	739	595	842	9
applied	107	729	143	739	595	842	9
in	146	729	154	739	595	842	9
urban	157	729	185	739	595	842	9
areas	188	729	216	739	595	842	9
with	219	729	239	739	595	842	9
similar	242	729	273	739	595	842	9
characteristics	276	729	347	739	595	842	9
to	350	729	359	739	595	842	9
Cuenca	362	729	400	739	595	842	9
in	403	729	412	739	595	842	9
the	415	729	430	739	595	842	9
Andean	433	729	471	739	595	842	9
region.	474	729	508	739	595	842	9
Enfoque	206	778	236	785	595	842	9
UTE,	238	778	256	785	595	842	9
V.9-N.4,	258	778	288	785	595	842	9
Dec.2018,	290	778	326	785	595	842	9
pp.	328	778	339	785	595	842	9
168	342	778	355	785	595	842	9
-	357	778	360	785	595	842	9
179	362	778	376	785	595	842	9
177	511	72	525	80	595	842	10
Figure	221	261	251	270	595	842	10
6:	254	261	263	270	595	842	10
Experimental	266	261	324	270	595	842	10
results	327	261	356	270	595	842	10
NO2	359	261	379	270	595	842	10
-	382	261	385	270	595	842	10
PM2.5	388	261	417	270	595	842	10
Figure	228	455	259	464	595	842	10
7:	262	455	270	464	595	842	10
Experimental	273	455	332	464	595	842	10
results	334	455	364	464	595	842	10
O3	367	455	380	464	595	842	10
-	383	455	386	464	595	842	10
NO2	389	455	409	464	595	842	10
5.	85	477	94	487	595	842	10
Conclusions	97	477	163	487	595	842	10
and	166	477	186	487	595	842	10
future	189	477	220	487	595	842	10
work	223	477	249	487	595	842	10
This	113	502	134	512	595	842	10
paper	141	502	169	512	595	842	10
presents	176	502	218	512	595	842	10
an	225	502	237	512	595	842	10
approach	243	502	289	512	595	842	10
to	296	502	305	512	595	842	10
discovery	312	502	358	512	595	842	10
of	365	502	374	512	595	842	10
knowledge	381	502	433	512	595	842	10
in	440	502	449	512	595	842	10
environmental	455	502	524	512	595	842	10
management.	85	515	152	525	595	842	10
The	155	515	174	525	595	842	10
approach	177	515	223	525	595	842	10
aim	225	515	243	525	595	842	10
is	246	515	254	525	595	842	10
to	256	515	265	525	595	842	10
support	268	515	305	525	595	842	10
environmental	307	515	376	525	595	842	10
managers	379	515	428	525	595	842	10
and	430	515	449	525	595	842	10
urban	451	515	480	525	595	842	10
planners	482	515	524	525	595	842	10
in	85	527	94	538	595	842	10
the	96	527	112	538	595	842	10
decision-making	115	527	194	538	595	842	10
process.	197	527	239	538	595	842	10
The	241	527	261	538	595	842	10
proposed	263	527	309	538	595	842	10
approach	312	527	358	538	595	842	10
allows	361	527	391	538	595	842	10
identifying	394	527	444	538	595	842	10
correlations	446	527	503	538	595	842	10
and	506	527	524	538	595	842	10
the	85	540	100	550	595	842	10
incidence	103	540	150	550	595	842	10
among	152	540	186	550	595	842	10
five	188	540	205	550	595	842	10
harmful	208	540	245	550	595	842	10
air	247	540	260	550	595	842	10
pollutants	262	540	309	550	595	842	10
on	312	540	324	550	595	842	10
the	327	540	342	550	595	842	10
Andean	345	540	383	550	595	842	10
region.	385	540	419	550	595	842	10
Such	421	540	447	550	595	842	10
correlations	449	540	506	550	595	842	10
are	509	540	524	550	595	842	10
analyzed	85	553	129	563	595	842	10
in	132	553	140	563	595	842	10
Time	143	553	167	563	595	842	10
Moving	170	553	205	563	595	842	10
Correlations.	207	553	270	563	595	842	10
The	274	553	294	563	595	842	10
proposed	296	553	342	563	595	842	10
approach	344	553	390	563	595	842	10
is	393	553	401	563	595	842	10
based	403	553	433	563	595	842	10
mainly	435	553	467	563	595	842	10
in	469	553	478	563	595	842	10
X-means	480	553	525	563	595	842	10
and	85	565	103	576	595	842	10
K-means	109	565	153	576	595	842	10
algorithms.	158	565	212	576	595	842	10
We	217	565	234	576	595	842	10
reported	239	565	280	576	595	842	10
a	285	565	291	576	595	842	10
comprehensible	297	565	374	576	595	842	10
experimentation	380	565	458	576	595	842	10
using	463	565	489	576	595	842	10
a	495	565	501	576	595	842	10
real	506	565	524	576	595	842	10
dataset	85	578	121	588	595	842	10
of	124	578	133	588	595	842	10
monitoring	136	578	188	588	595	842	10
station	191	578	223	588	595	842	10
from	226	578	248	588	595	842	10
Cuenca,	251	578	292	588	595	842	10
Ecuador.	295	578	339	588	595	842	10
In	113	591	123	601	595	842	10
the	127	591	142	601	595	842	10
evaluation,	147	591	200	601	595	842	10
the	204	591	219	601	595	842	10
K-means	224	591	268	601	595	842	10
algorithm	272	591	317	601	595	842	10
reflects	321	591	357	601	595	842	10
a	361	591	367	601	595	842	10
suitable	371	591	409	601	595	842	10
processing	414	591	467	601	595	842	10
time	471	591	492	601	595	842	10
in	496	591	505	601	595	842	10
the	509	591	524	601	595	842	10
grouping	85	603	128	614	595	842	10
according	157	603	204	614	595	842	10
to	209	603	218	614	595	842	10
the	222	603	238	614	595	842	10
used	242	603	266	614	595	842	10
dataset	270	603	306	614	595	842	10
size.	310	603	333	614	595	842	10
As	337	603	350	614	595	842	10
a	354	603	360	614	595	842	10
mentioned	365	603	416	614	595	842	10
in	420	603	429	614	595	842	10
terms	433	603	461	614	595	842	10
of	465	603	474	614	595	842	10
efficiency	478	603	524	614	595	842	10
(Accuracy	85	616	134	626	595	842	10
and	138	616	156	626	595	842	10
Time	160	616	184	626	595	842	10
Execution)	188	616	240	626	595	842	10
K-means	244	616	288	626	595	842	10
Clustering	292	616	341	626	595	842	10
algorithm	345	616	390	626	595	842	10
presents	394	616	436	626	595	842	10
its	439	616	450	626	595	842	10
self	454	616	471	626	595	842	10
as	475	616	487	626	595	842	10
a	490	616	496	626	595	842	10
good	500	616	524	626	595	842	10
choice,	85	629	120	639	595	842	10
similar	125	629	157	639	595	842	10
results	161	629	194	639	595	842	10
are	198	629	214	639	595	842	10
presented	219	629	268	639	595	842	10
in	273	629	281	639	595	842	10
Doreswamy,	286	629	347	639	595	842	10
Ghoneim,	352	629	400	639	595	842	10
and	404	629	423	639	595	842	10
Manjaunath	428	629	485	639	595	842	10
(2015);	490	629	525	639	595	842	10
Kingsy	85	641	118	652	595	842	10
et	121	641	130	652	595	842	10
al.,	134	641	148	652	595	842	10
(2016).	151	641	186	652	595	842	10
The	120	654	140	664	595	842	10
algorithms	146	654	197	664	595	842	10
and	203	654	222	664	595	842	10
techniques	228	654	281	664	595	842	10
applied	288	654	323	664	595	842	10
were	330	654	353	664	595	842	10
evaluated	360	654	408	664	595	842	10
as	414	654	426	664	595	842	10
positive	432	654	469	664	595	842	10
to	476	654	485	664	595	842	10
extract	491	654	524	664	595	842	10
knowledge	85	667	138	677	595	842	10
from	142	667	164	677	595	842	10
urban	168	667	196	677	595	842	10
areas.	200	667	231	677	595	842	10
This	235	667	256	677	595	842	10
corroborates	260	667	322	677	595	842	10
the	326	667	341	677	595	842	10
hypothesis	345	667	398	677	595	842	10
raised:	402	667	435	677	595	842	10
the	439	667	454	677	595	842	10
application	458	667	511	677	595	842	10
of	515	667	524	677	595	842	10
data	85	679	107	690	595	842	10
mining	110	679	142	690	595	842	10
techniques	145	679	198	690	595	842	10
helps	201	679	228	690	595	842	10
identifying	231	679	280	690	595	842	10
patterns	283	679	323	690	595	842	10
of	326	679	335	690	595	842	10
behavior	338	679	381	690	595	842	10
among	384	679	417	690	595	842	10
several	420	679	456	690	595	842	10
air	459	679	471	690	595	842	10
pollutants.	474	679	524	690	595	842	10
However,	85	692	131	702	595	842	10
additional	136	692	183	702	595	842	10
processes	187	692	237	702	595	842	10
are	241	692	257	702	595	842	10
necessary	261	692	311	702	595	842	10
in	315	692	323	702	595	842	10
order	327	692	353	702	595	842	10
to	357	692	366	702	595	842	10
get	370	692	385	702	595	842	10
complete	389	692	434	702	595	842	10
information	438	692	492	702	595	842	10
which	496	692	524	702	595	842	10
effectively	85	705	134	715	595	842	10
supports	140	705	182	715	595	842	10
the	188	705	203	715	595	842	10
decision	209	705	250	715	595	842	10
making.	256	705	294	715	595	842	10
Therefore,	300	705	351	715	595	842	10
some	357	705	384	715	595	842	10
limitations	390	705	439	715	595	842	10
still	444	705	460	715	595	842	10
need	466	705	491	715	595	842	10
to	497	705	506	715	595	842	10
be	512	705	524	715	595	842	10
overcome.	85	717	137	727	595	842	10
First,	140	717	165	727	595	842	10
the	168	717	184	727	595	842	10
proposed	187	717	233	727	595	842	10
approach	237	717	283	727	595	842	10
does	286	717	310	727	595	842	10
not	314	717	329	727	595	842	10
include	333	717	368	727	595	842	10
other	371	717	396	727	595	842	10
important	400	717	446	727	595	842	10
variables	449	717	493	727	595	842	10
in	497	717	506	727	595	842	10
the	509	717	524	727	595	842	10
air	85	730	97	740	595	842	10
pollution	101	730	142	740	595	842	10
analysis	145	730	184	740	595	842	10
as	187	730	199	740	595	842	10
precipitation	202	730	262	740	595	842	10
or	265	730	274	740	595	842	10
wind	278	730	300	740	595	842	10
speed.	303	730	336	740	595	842	10
Second,	339	730	380	740	595	842	10
the	383	730	398	740	595	842	10
approach	401	730	447	740	595	842	10
does	450	730	474	740	595	842	10
not	477	730	493	740	595	842	10
return	496	730	524	740	595	842	10
information	85	743	140	753	595	842	10
about	142	743	170	753	595	842	10
the	172	743	188	753	595	842	10
causality	190	743	233	753	595	842	10
of	236	743	245	753	595	842	10
incidence.	248	743	297	753	595	842	10
As	300	743	313	753	595	842	10
future	315	743	344	753	595	842	10
work,	346	743	372	753	595	842	10
variables	375	743	419	753	595	842	10
such	422	743	445	753	595	842	10
as	448	743	459	753	595	842	10
precipitation,	462	743	524	753	595	842	10
Enfoque	220	778	250	785	595	842	10
UTE,	252	778	270	785	595	842	10
V.9-N.4,	272	778	302	785	595	842	10
Dec.2018,	304	778	340	785	595	842	10
pp.	343	778	354	785	595	842	10
168	356	778	369	785	595	842	10
-	372	778	374	785	595	842	10
179	377	778	390	785	595	842	10
178	71	72	84	80	595	842	11
radiation	71	87	113	98	595	842	11
direction,	116	87	161	98	595	842	11
wind	163	87	186	98	595	842	11
speed,	189	87	222	98	595	842	11
temperature	224	87	284	98	595	842	11
and,	286	87	308	98	595	842	11
the	310	87	326	98	595	842	11
humidity	329	87	369	98	595	842	11
should	372	87	404	98	595	842	11
be	407	87	419	98	595	842	11
included.	422	87	466	98	595	842	11
Also,	469	87	494	98	595	842	11
we	496	87	510	98	595	842	11
will	71	100	86	110	595	842	11
seek	89	100	112	110	595	842	11
to	116	100	125	110	595	842	11
evaluate	128	100	169	110	595	842	11
the	172	100	188	110	595	842	11
approach	191	100	237	110	595	842	11
in	240	100	248	110	595	842	11
other	252	100	277	110	595	842	11
cities	280	100	305	110	595	842	11
in	308	100	316	110	595	842	11
Ecuador.	319	100	363	110	595	842	11
References	71	125	130	135	595	842	11
Cagliero,	71	151	115	161	595	842	11
L.,	118	151	130	161	595	842	11
Cerquitelli,	134	151	186	161	595	842	11
T.,	189	151	202	161	595	842	11
Chiusano,	205	151	255	161	595	842	11
S.,	258	151	271	161	595	842	11
Garza,	275	151	308	161	595	842	11
P.,	311	151	324	161	595	842	11
and	328	151	346	161	595	842	11
Ricupero,	350	151	397	161	595	842	11
G.	400	151	412	161	595	842	11
(2016).	415	151	450	161	595	842	11
Discovering	453	151	510	161	595	842	11
Air	99	163	113	173	595	842	11
Quality	119	163	153	173	595	842	11
Patterns	159	163	200	173	595	842	11
in	206	163	214	173	595	842	11
Urban	220	163	250	173	595	842	11
Environments.	256	163	326	173	595	842	11
En	332	163	346	173	595	842	11
Proceedings	352	163	413	173	595	842	11
of	419	163	428	173	595	842	11
the	434	163	449	173	595	842	11
2016	455	163	480	173	595	842	11
ACM	486	163	510	173	595	842	11
International	99	176	160	186	595	842	11
Joint	163	176	186	186	595	842	11
Conference	189	176	246	186	595	842	11
on	249	176	261	186	595	842	11
Pervasive	264	176	313	186	595	842	11
and	316	176	334	186	595	842	11
Ubiquitous	337	176	389	186	595	842	11
Computing:	392	176	449	186	595	842	11
Adjunct	452	176	488	186	595	842	11
(pp.	491	176	510	186	595	842	11
25–28).	99	189	137	199	595	842	11
New	140	189	162	199	595	842	11
York,	165	189	190	199	595	842	11
NY,	194	189	212	199	595	842	11
USA:	215	189	240	199	595	842	11
ACM.	244	189	271	199	595	842	11
https://doi.org/10.1145/2968219.2971458	274	189	475	199	595	842	11
Clima	71	201	99	211	595	842	11
CUENCA:	104	201	153	211	595	842	11
Temperatura,	158	201	224	211	595	842	11
Climograma	229	201	288	211	595	842	11
y	293	201	298	211	595	842	11
Tabla	303	201	330	211	595	842	11
climática	335	201	378	211	595	842	11
para	383	201	405	211	595	842	11
CUENCA	410	201	456	211	595	842	11
-	461	201	465	211	595	842	11
Climate-	470	201	511	211	595	842	11
Data.org.	99	214	145	224	595	842	11
(s.	155	214	167	224	595	842	11
f.).	170	214	183	224	595	842	11
Recuperado	194	214	254	224	595	842	11
16	264	214	277	224	595	842	11
de	287	214	300	224	595	842	11
julio	310	214	330	224	595	842	11
de	340	214	352	224	595	842	11
2018,	363	214	391	224	595	842	11
de	401	214	414	224	595	842	11
https://es.climate-	424	214	510	224	595	842	11
data.org/location/875185/	99	226	223	237	595	842	11
Doreswamy,	71	239	132	249	595	842	11
G.	137	239	149	249	595	842	11
O.,	154	239	169	249	595	842	11
and	174	239	192	249	595	842	11
Manjaunath,	197	239	258	249	595	842	11
B.	263	239	273	249	595	842	11
(2015).	278	239	313	249	595	842	11
Air	318	239	332	249	595	842	11
pollution	337	239	378	249	595	842	11
clustering	383	239	430	249	595	842	11
using	435	239	461	249	595	842	11
K-means	466	239	510	249	595	842	11
algorithm	99	252	145	262	595	842	11
in	147	252	156	262	595	842	11
smart	159	252	186	262	595	842	11
city.	189	252	209	262	595	842	11
International	212	252	272	262	595	842	11
Journal	275	252	311	262	595	842	11
of	314	252	323	262	595	842	11
Innovative	326	252	376	262	595	842	11
Research	379	252	426	262	595	842	11
in	429	252	438	262	595	842	11
Computer	441	252	489	262	595	842	11
and	492	252	510	262	595	842	11
Communication	99	264	176	275	595	842	11
Engineering,	179	264	241	275	595	842	11
3,	244	264	253	275	595	842	11
51–57.	256	264	290	275	595	842	11
Doreswamy,	71	277	132	287	595	842	11
Ghoneim,	135	277	183	287	595	842	11
O.,	186	277	200	287	595	842	11
and	204	277	222	287	595	842	11
Manjaunath,	225	277	285	287	595	842	11
B.	289	277	299	287	595	842	11
R.	302	277	313	287	595	842	11
(2015).	316	277	351	287	595	842	11
Air	354	277	367	287	595	842	11
Pollution	370	277	413	287	595	842	11
Clustering	416	277	465	287	595	842	11
Using	468	277	496	287	595	842	11
K-	499	277	510	287	595	842	11
Means	99	290	132	300	595	842	11
Algorithm	136	290	183	300	595	842	11
in	187	290	196	300	595	842	11
Smart	200	290	229	300	595	842	11
City.	234	290	255	300	595	842	11
En	260	290	273	300	595	842	11
International	277	290	338	300	595	842	11
Journal	342	290	379	300	595	842	11
of	383	290	392	300	595	842	11
Innovative	396	290	446	300	595	842	11
Research	450	290	498	300	595	842	11
in	502	290	510	300	595	842	11
Computer	99	302	148	313	595	842	11
and	151	302	169	313	595	842	11
Communication	172	302	249	313	595	842	11
Engineering	252	302	310	313	595	842	11
(Vol.	313	302	336	313	595	842	11
Vol.	339	302	358	313	595	842	11
3,	361	302	370	313	595	842	11
Special	373	302	409	313	595	842	11
Issue	412	302	439	313	595	842	11
7).	442	302	455	313	595	842	11
Du,	71	315	88	325	595	842	11
X.,	91	315	104	325	595	842	11
and	107	315	126	325	595	842	11
Varde,	129	315	161	325	595	842	11
A.	164	315	174	325	595	842	11
S.	177	315	187	325	595	842	11
(2016).	190	315	225	325	595	842	11
Mining	228	315	260	325	595	842	11
PM2.5	263	315	295	325	595	842	11
and	298	315	316	325	595	842	11
traffic	319	315	346	325	595	842	11
conditions	349	315	398	325	595	842	11
for	401	315	414	325	595	842	11
air	417	315	429	325	595	842	11
quality.	432	315	467	325	595	842	11
En	469	315	483	325	595	842	11
2016	486	315	510	325	595	842	11
7th	99	328	115	338	595	842	11
International	119	328	180	338	595	842	11
Conference	185	328	241	338	595	842	11
on	246	328	258	338	595	842	11
Information	263	328	318	338	595	842	11
and	323	328	341	338	595	842	11
Communication	346	328	422	338	595	842	11
Systems	427	328	469	338	595	842	11
(ICICS)	474	328	510	338	595	842	11
(pp.	99	340	118	351	595	842	11
33-38).	121	340	156	351	595	842	11
https://doi.org/10.1109/IACS.2016.7476082	159	340	371	351	595	842	11
Fukuda,	71	353	111	363	595	842	11
K.	116	353	126	363	595	842	11
(2007).	131	353	166	363	595	842	11
Noise	171	353	199	363	595	842	11
Reduction	204	353	253	363	595	842	11
Approach	258	353	305	363	595	842	11
for	310	353	323	363	595	842	11
Decision	328	353	370	363	595	842	11
Tree	375	353	398	363	595	842	11
Construction:	403	353	468	363	595	842	11
A	472	353	480	363	595	842	11
Case	485	353	510	363	595	842	11
Study	99	366	127	376	595	842	11
of	135	366	144	376	595	842	11
Knowledge	151	366	206	376	595	842	11
Discovery	213	366	261	376	595	842	11
on	269	366	281	376	595	842	11
Climate	288	366	326	376	595	842	11
and	333	366	352	376	595	842	11
Air	359	366	372	376	595	842	11
Pollution.	380	366	425	376	595	842	11
En	432	366	446	376	595	842	11
2007	453	366	478	376	595	842	11
IEEE	485	366	510	376	595	842	11
Symposium	99	378	157	388	595	842	11
on	166	378	178	388	595	842	11
Computational	188	378	259	388	595	842	11
Intelligence	268	378	324	388	595	842	11
and	333	378	351	388	595	842	11
Data	361	378	384	388	595	842	11
Mining	393	378	425	388	595	842	11
(pp.	435	378	454	388	595	842	11
697-704).	463	378	510	388	595	842	11
https://doi.org/10.1109/CIDM.2007.368944	99	391	307	401	595	842	11
Gao,	71	404	95	414	595	842	11
B.	97	404	108	414	595	842	11
J.,	110	404	121	414	595	842	11
Tung,	124	404	152	414	595	842	11
R.,	154	404	168	414	595	842	11
and	171	404	189	414	595	842	11
Yang,	191	404	220	414	595	842	11
Y.	222	404	233	414	595	842	11
(2017).	235	404	270	414	595	842	11
Iterative	272	404	311	414	595	842	11
matrix	313	404	343	414	595	842	11
correlation	346	404	397	414	595	842	11
for	399	404	412	414	595	842	11
bisection	414	404	458	414	595	842	11
clustering.	460	404	510	414	595	842	11
En	99	416	113	426	595	842	11
2017	120	416	144	426	595	842	11
IEEE	152	416	177	426	595	842	11
International	184	416	244	426	595	842	11
Conference	251	416	308	426	595	842	11
on	316	416	328	426	595	842	11
Big	335	416	351	426	595	842	11
Data	358	416	381	426	595	842	11
(Big	388	416	408	426	595	842	11
Data)	415	416	442	426	595	842	11
(pp.	449	416	468	426	595	842	11
80-87).	476	416	510	426	595	842	11
https://doi.org/10.1109/BigData.2017.8257914	99	429	324	439	595	842	11
Kampa,	71	442	109	452	595	842	11
M.,	112	442	127	452	595	842	11
and	131	442	149	452	595	842	11
Castanas,	152	442	202	452	595	842	11
E.	205	442	216	452	595	842	11
(2008).	219	442	254	452	595	842	11
Human	257	442	293	452	595	842	11
health	296	442	326	452	595	842	11
effects	329	442	361	452	595	842	11
of	365	442	374	452	595	842	11
air	377	442	389	452	595	842	11
pollution.	393	442	437	452	595	842	11
Environmental	440	442	510	452	595	842	11
Pollution,	99	454	144	464	595	842	11
151(2),	148	454	182	464	595	842	11
362-367.	186	454	229	464	595	842	11
https://doi.org/10.1016/j.envpol.2007.06.012	232	454	447	464	595	842	11
Katz,	71	467	96	477	595	842	11
M.	99	467	111	477	595	842	11
(1970).	114	467	149	477	595	842	11
Photochemical	152	467	224	477	595	842	11
reactions	227	467	272	477	595	842	11
of	274	467	284	477	595	842	11
atmospheric	287	467	346	477	595	842	11
pollutants.	349	467	400	477	595	842	11
The	402	467	421	477	595	842	11
Canadian	424	467	471	477	595	842	11
Journal	474	467	510	477	595	842	11
of	99	479	108	490	595	842	11
Chemical	112	479	157	490	595	842	11
Engineering,	160	479	222	490	595	842	11
48(1),	225	479	254	490	595	842	11
3-11.	257	479	282	490	595	842	11
https://doi.org/10.1002/cjce.5450480102	285	479	481	490	595	842	11
Kim,	71	492	93	502	595	842	11
K.-H.,	98	492	126	502	595	842	11
Choi,	131	492	157	502	595	842	11
Y.-J.,	162	492	188	502	595	842	11
and	193	492	211	502	595	842	11
Kim,	216	492	238	502	595	842	11
M.-Y.	243	492	269	502	595	842	11
(2005).	274	492	309	502	595	842	11
The	314	492	333	502	595	842	11
exceedance	338	492	397	502	595	842	11
patterns	402	492	442	502	595	842	11
of	447	492	456	502	595	842	11
air	461	492	474	502	595	842	11
quality	478	492	510	502	595	842	11
criteria:	99	505	135	515	595	842	11
a	139	505	146	515	595	842	11
case	150	505	173	515	595	842	11
study	177	505	203	515	595	842	11
of	207	505	216	515	595	842	11
ozone	220	505	250	515	595	842	11
and	254	505	272	515	595	842	11
nitrogen	276	505	316	515	595	842	11
dioxide	320	505	355	515	595	842	11
in	359	505	368	515	595	842	11
Seoul,	372	505	403	515	595	842	11
Korea	407	505	436	515	595	842	11
between	440	505	482	515	595	842	11
1990	486	505	510	515	595	842	11
and	99	518	118	528	595	842	11
2000.	173	518	200	528	595	842	11
Chemosphere,	256	518	328	528	595	842	11
60(4),	383	518	412	528	595	842	11
441-452.	467	518	510	528	595	842	11
https://doi.org/10.1016/j.chemosphere.2004.12.067	99	530	348	540	595	842	11
Kingsy,	71	543	107	553	595	842	11
G.	110	543	122	553	595	842	11
R.,	125	543	139	553	595	842	11
Manimegalai,	142	543	207	553	595	842	11
R.,	210	543	224	553	595	842	11
Geetha,	227	543	266	553	595	842	11
D.	269	543	280	553	595	842	11
M.,	283	543	298	553	595	842	11
Rajathi,	301	543	339	553	595	842	11
S.,	342	543	355	553	595	842	11
Usha,	358	543	387	553	595	842	11
K.,	390	543	403	553	595	842	11
and	406	543	425	553	595	842	11
Raabiathul,	428	543	483	553	595	842	11
B.	486	543	497	553	595	842	11
N.	500	543	510	553	595	842	11
(2016).	99	555	134	566	595	842	11
Air	138	555	152	566	595	842	11
pollution	156	555	197	566	595	842	11
analysis	201	555	240	566	595	842	11
using	244	555	271	566	595	842	11
enhanced	275	555	323	566	595	842	11
K-Means	327	555	371	566	595	842	11
clustering	375	555	422	566	595	842	11
algorithm	426	555	471	566	595	842	11
for	475	555	488	566	595	842	11
real	492	555	510	566	595	842	11
time	99	568	120	578	595	842	11
sensor	123	568	156	578	595	842	11
data.	158	568	183	578	595	842	11
En	185	568	199	578	595	842	11
Region	201	568	236	578	595	842	11
10	239	568	251	578	595	842	11
Conference	254	568	310	578	595	842	11
(TENCON),	313	568	370	578	595	842	11
2016	372	568	397	578	595	842	11
IEEE	399	568	424	578	595	842	11
(pp.	427	568	446	578	595	842	11
1945–1949).	448	568	510	578	595	842	11
IEEE.	99	581	127	591	595	842	11
Kumar,	71	593	106	604	595	842	11
P.,	111	593	125	604	595	842	11
and	129	593	148	604	595	842	11
Wasan,	152	593	189	604	595	842	11
S.	194	593	204	604	595	842	11
K.	209	593	220	604	595	842	11
(2010).	224	593	259	604	595	842	11
Analysis	264	593	304	604	595	842	11
of	309	593	318	604	595	842	11
X-means	323	593	367	604	595	842	11
and	372	593	390	604	595	842	11
global	395	593	424	604	595	842	11
k-means	429	593	471	604	595	842	11
USING	476	593	510	604	595	842	11
TUMOR	99	606	140	616	595	842	11
classification.	143	606	208	616	595	842	11
En	212	606	225	616	595	842	11
2010	229	606	253	616	595	842	11
The	256	606	275	616	595	842	11
2nd	279	606	297	616	595	842	11
International	301	606	361	616	595	842	11
Conference	365	606	421	616	595	842	11
on	425	606	437	616	595	842	11
Computer	441	606	489	616	595	842	11
and	492	606	510	616	595	842	11
Automation	99	619	155	629	595	842	11
Engineering	182	619	241	629	595	842	11
(ICCAE)	268	619	308	629	595	842	11
(Vol.	335	619	358	629	595	842	11
5,	385	619	394	629	595	842	11
pp.	421	619	436	629	595	842	11
832-835).	463	619	510	629	595	842	11
https://doi.org/10.1109/ICCAE.2010.5451883	99	631	319	641	595	842	11
Li,	71	644	83	654	595	842	11
H.,	89	644	103	654	595	842	11
Fan,	109	644	131	654	595	842	11
H.,	137	644	151	654	595	842	11
and	157	644	176	654	595	842	11
Mao,	182	644	206	654	595	842	11
F.	213	644	222	654	595	842	11
(2016).	228	644	263	654	595	842	11
A	269	644	277	654	595	842	11
Visualization	283	644	344	654	595	842	11
Approach	351	644	398	654	595	842	11
to	404	644	413	654	595	842	11
Air	419	644	433	654	595	842	11
Pollution	439	644	481	654	595	842	11
Data	487	644	510	654	595	842	11
Exploration—A	99	657	173	667	595	842	11
Case	181	657	207	667	595	842	11
Study	215	657	243	667	595	842	11
of	252	657	261	667	595	842	11
Air	270	657	283	667	595	842	11
Quality	291	657	326	667	595	842	11
Index	334	657	361	667	595	842	11
(PM2.5)	369	657	408	667	595	842	11
in	417	657	425	667	595	842	11
Beijing,	434	657	470	667	595	842	11
China.	479	657	510	667	595	842	11
Atmosphere,	99	669	162	680	595	842	11
7(3),	165	669	187	680	595	842	11
35.	190	669	206	680	595	842	11
https://doi.org/10.3390/atmos7030035	209	669	394	680	595	842	11
Quick-R:	71	682	114	692	595	842	11
Correlations.	131	682	193	692	595	842	11
(s.	209	682	222	692	595	842	11
f.).	225	682	238	692	595	842	11
Recuperado	254	682	314	692	595	842	11
24	331	682	343	692	595	842	11
de	360	682	372	692	595	842	11
julio	389	682	408	692	595	842	11
de	425	682	437	692	595	842	11
2018,	454	682	481	692	595	842	11
de	498	682	510	692	595	842	11
https://www.statmethods.net/stats/correlations.html	99	695	347	705	595	842	11
Select	71	707	102	717	595	842	11
by	104	707	116	717	595	842	11
Weights	119	707	159	717	595	842	11
-	161	707	165	717	595	842	11
RapidMiner	168	707	224	717	595	842	11
Documentation.	227	707	304	717	595	842	11
(s.	307	707	319	717	595	842	11
f.).	322	707	335	717	595	842	11
Recuperado	338	707	398	717	595	842	11
17	400	707	413	717	595	842	11
de	415	707	428	717	595	842	11
julio	430	707	450	717	595	842	11
de	453	707	465	717	595	842	11
2018,	468	707	495	717	595	842	11
de	498	707	510	717	595	842	11
https://docs.rapidminer.com/latest/studio/operators/blending/attributes/selection/sele	99	720	508	730	595	842	11
ct_by_weights.html	99	733	193	743	595	842	11
Shazan,	71	745	111	755	595	842	11
M.,	114	745	129	755	595	842	11
Jabbar,	132	745	169	755	595	842	11
M.,	172	745	187	755	595	842	11
Zaïane,	190	745	227	755	595	842	11
O.	230	745	242	755	595	842	11
R.,	245	745	259	755	595	842	11
and	262	745	280	755	595	842	11
Osornio-Vargas,	283	745	363	755	595	842	11
A.	366	745	376	755	595	842	11
(2017).	379	745	414	755	595	842	11
Discovering	417	745	474	755	595	842	11
Spatial	477	745	510	755	595	842	11
Contrast	99	758	141	768	595	842	11
and	146	758	164	768	595	842	11
Common	170	758	214	768	595	842	11
Sets	219	758	241	768	595	842	11
with	246	758	266	768	595	842	11
Statistically	271	758	326	768	595	842	11
Significant	331	758	382	768	595	842	11
Co-location	387	758	443	768	595	842	11
Patterns.	448	758	492	768	595	842	11
En	497	758	510	768	595	842	11
Enfoque	206	778	236	785	595	842	11
UTE,	238	778	256	785	595	842	11
V.9-N.4,	258	778	288	785	595	842	11
Dec.2018,	290	778	326	785	595	842	11
pp.	328	778	339	785	595	842	11
168	342	778	355	785	595	842	11
-	357	778	360	785	595	842	11
179	362	778	376	785	595	842	11
179	511	72	525	80	595	842	12
Proceedings	113	87	175	98	595	842	12
of	177	87	187	98	595	842	12
the	189	87	205	98	595	842	12
Symposium	208	87	265	98	595	842	12
on	268	87	280	98	595	842	12
Applied	283	87	320	98	595	842	12
Computing	323	87	376	98	595	842	12
(pp.	379	87	397	98	595	842	12
796–803).	400	87	450	98	595	842	12
New	453	87	475	98	595	842	12
York,	478	87	503	98	595	842	12
NY,	506	87	525	98	595	842	12
USA:	113	100	139	110	595	842	12
ACM.	142	100	170	110	595	842	12
https://doi.org/10.1145/3019612.3019665	173	100	374	110	595	842	12
Souza,	85	113	119	123	595	842	12
F.	123	113	133	123	595	842	12
T.,	136	113	149	123	595	842	12
and	153	113	171	123	595	842	12
Rabelo,	175	113	212	123	595	842	12
W.	216	113	230	123	595	842	12
S.	233	113	243	123	595	842	12
(2015).	247	113	282	123	595	842	12
A	285	113	293	123	595	842	12
data	296	113	317	123	595	842	12
mining	321	113	353	123	595	842	12
approach	357	113	403	123	595	842	12
to	406	113	416	123	595	842	12
study	419	113	445	123	595	842	12
the	449	113	464	123	595	842	12
air	468	113	480	123	595	842	12
pollution	484	113	525	123	595	842	12
induced	113	125	152	136	595	842	12
by	155	125	166	136	595	842	12
urban	169	125	197	136	595	842	12
phenomena	200	125	258	136	595	842	12
and	261	125	279	136	595	842	12
the	282	125	297	136	595	842	12
association	300	125	355	136	595	842	12
with	358	125	377	136	595	842	12
respiratory	380	125	432	136	595	842	12
diseases.	435	125	481	136	595	842	12
En	484	125	497	136	595	842	12
2015	500	125	524	136	595	842	12
11th	113	138	135	148	595	842	12
International	141	138	202	148	595	842	12
Conference	208	138	265	148	595	842	12
on	271	138	283	148	595	842	12
Natural	289	138	325	148	595	842	12
Computation	331	138	393	148	595	842	12
(ICNC)	399	138	434	148	595	842	12
(pp.	440	138	459	148	595	842	12
1045-1050).	465	138	525	148	595	842	12
https://doi.org/10.1109/ICNC.2015.7378136	113	151	326	161	595	842	12
Wagner,	85	163	127	173	595	842	12
E.	131	163	141	173	595	842	12
(1994).	145	163	180	173	595	842	12
Impacts	184	163	223	173	595	842	12
on	226	163	239	173	595	842	12
air	243	163	255	173	595	842	12
pollution	259	163	300	173	595	842	12
in	304	163	312	173	595	842	12
urban	316	163	345	173	595	842	12
areas.	349	163	379	173	595	842	12
Environmental	383	163	453	173	595	842	12
Management,	457	163	525	173	595	842	12
18(5),	113	176	142	186	595	842	12
759-765.	145	176	189	186	595	842	12
https://doi.org/10.1007/BF02394638	192	176	367	186	595	842	12
Walden,	85	189	125	199	595	842	12
S.,	129	189	143	199	595	842	12
and	146	189	165	199	595	842	12
Andrew,	168	189	209	199	595	842	12
C.	212	189	223	199	595	842	12
(2013).	227	189	262	199	595	842	12
Publicación	265	189	322	199	595	842	12
de	325	189	338	199	595	842	12
los	341	189	355	199	595	842	12
contaminantes	359	189	430	199	595	842	12
atmosféricos	434	189	496	199	595	842	12
de	500	189	512	199	595	842	12
la	516	189	524	199	595	842	12
estación	113	201	154	211	595	842	12
de	159	201	171	211	595	842	12
monitoreo	175	201	224	211	595	842	12
en	228	201	240	211	595	842	12
tiempo	244	201	277	211	595	842	12
real	281	201	300	211	595	842	12
de	304	201	316	211	595	842	12
la	320	201	329	211	595	842	12
ciudad	333	201	365	211	595	842	12
de	369	201	382	211	595	842	12
Cuenca,	386	201	427	211	595	842	12
utilizando	431	201	478	211	595	842	12
servicios	482	201	524	211	595	842	12
estándares	113	214	168	224	595	842	12
OGC.	171	214	199	224	595	842	12
Recuperado	202	214	262	224	595	842	12
de	265	214	277	224	595	842	12
http://dspace.uazuay.edu.ec/handle/datos/2546	280	214	512	224	595	842	12
Wirth,	85	226	114	237	595	842	12
R.	119	226	130	237	595	842	12
(2000).	136	226	170	237	595	842	12
CRISP-DM:	176	226	233	237	595	842	12
Towards	239	226	281	237	595	842	12
a	286	226	292	237	595	842	12
standard	298	226	341	237	595	842	12
process	346	226	384	237	595	842	12
model	390	226	420	237	595	842	12
for	425	226	438	237	595	842	12
data	443	226	465	237	595	842	12
mining.	470	226	506	237	595	842	12
En	511	226	524	237	595	842	12
Proceedings	113	239	175	249	595	842	12
of	179	239	188	249	595	842	12
the	192	239	208	249	595	842	12
Fourth	212	239	244	249	595	842	12
International	248	239	309	249	595	842	12
Conference	313	239	370	249	595	842	12
on	374	239	387	249	595	842	12
the	391	239	406	249	595	842	12
Practical	411	239	453	249	595	842	12
Application	457	239	511	249	595	842	12
of	515	239	524	249	595	842	12
Knowledge	113	252	168	262	595	842	12
Discovery	171	252	219	262	595	842	12
and	222	252	241	262	595	842	12
Data	244	252	267	262	595	842	12
Mining	270	252	302	262	595	842	12
(pp.	305	252	324	262	595	842	12
29–39).	328	252	365	262	595	842	12
Zhang,	85	264	119	275	595	842	12
L.,	123	264	135	275	595	842	12
Deng,	138	264	167	275	595	842	12
S.,	170	264	184	275	595	842	12
and	187	264	205	275	595	842	12
Li,	208	264	220	275	595	842	12
S.	223	264	234	275	595	842	12
(2017).	237	264	272	275	595	842	12
Analysis	275	264	315	275	595	842	12
of	319	264	328	275	595	842	12
power	331	264	361	275	595	842	12
consumer	364	264	412	275	595	842	12
behavior	415	264	458	275	595	842	12
based	461	264	491	275	595	842	12
on	494	264	506	275	595	842	12
the	509	264	524	275	595	842	12
complementation	113	277	197	287	595	842	12
of	201	277	210	287	595	842	12
K-means	214	277	258	287	595	842	12
and	262	277	281	287	595	842	12
DBSCAN.	285	277	334	287	595	842	12
En	338	277	351	287	595	842	12
2017	355	277	379	287	595	842	12
IEEE	384	277	409	287	595	842	12
Conference	413	277	469	287	595	842	12
on	473	277	485	287	595	842	12
Energy	490	277	524	287	595	842	12
Internet	113	290	151	300	595	842	12
and	175	290	193	300	595	842	12
Energy	217	290	252	300	595	842	12
System	275	290	312	300	595	842	12
Integration	336	290	388	300	595	842	12
(EI2)	412	290	435	300	595	842	12
(pp.	459	290	478	300	595	842	12
1-5).	502	290	525	300	595	842	12
https://doi.org/10.1109/EI2.2017.8245490	113	302	316	313	595	842	12
Enfoque	220	778	250	785	595	842	12
UTE,	252	778	270	785	595	842	12
V.9-N.4,	272	778	302	785	595	842	12
Dec.2018,	304	778	340	785	595	842	12
pp.	343	778	354	785	595	842	12
168	356	778	369	785	595	842	12
-	372	778	374	785	595	842	12
179	377	778	390	785	595	842	12
