T
E
L
K
O
MN
I
K
A
T
el
eco
m
m
u
n
i
ca
t
i
o
n
,
C
o
m
p
u
t
i
n
g
,
E
l
ect
ro
n
i
cs
a
n
d
C
o
n
t
ro
l
Vo
l
.
19
, N
o
.
5
,
O
ct
o
b
er
2021
, p
p
.
1498
~
1506
I
S
S
N
:
1693
-
6930
,
a
c
c
r
e
di
t
e
d F
i
r
s
t
G
r
a
de
by K
e
m
e
nr
i
s
t
e
kdi
kt
i
,
D
e
c
r
e
e
N
o:
21/
E
/
K
P
T
/
2018
D
O
I
:
10.
12928/
T
E
L
K
O
M
N
I
K
A
.
v19i
5.
18965
1498
Jou
r
n
al
h
om
e
page
:
ht
t
p:
/
/
j
our
nal
.
uad
.
ac
.
i
d/
i
nde
x
.
php/
T
E
L
K
O
M
N
I
K
A
H
y
bri
d
cl
us
t
eri
n
g
ba
s
e
d
o
n
m
ul
t
i
-
c
ri
t
e
ri
a
s
eg
m
en
t
a
t
i
o
n
f
o
r
hi
g
he
r
e
du
ca
t
i
o
n
m
a
r
ke
t
i
ng
H
a
r
di
ka
K
hus
nul
i
a
w
a
t
i
1
,
D
hi
a
n
R
i
s
ki
a
na
P
ut
r
i
2
1
De
pa
r
tm
e
n
t of
I
nf
or
m
a
t
ic
s,
U
ni
ve
r
s
ita
s S
a
hi
d S
ur
a
ka
r
ta
,
S
u
r
a
k
a
r
t
a
,
I
nd
on
e
si
a
2
De
pa
r
tm
e
n
t of
P
s
yc
h
ol
og
y,
Un
ive
r
si
ta
s S
a
h
id S
ur
a
ka
r
ta
,
S
u
r
a
k
a
r
t
a
,
I
nd
on
e
si
a
A
rt
i
cl
e I
n
f
o
AB
S
T
RACT
A
r
tic
le
h
is
to
r
y
:
R
ecei
v
ed
D
e
c
9,
2020
R
ev
i
s
ed
A
pr
14,
2021
A
ccep
t
ed
A
pr
22,
2021
M
a
r
ke
t se
gm
e
nta
ti
on in
h
ig
he
r
e
d
uc
a
t
io
n in
st
it
ut
io
ns i
s
s
ti
ll
r
a
r
e
ly
a
pp
li
e
d
a
lt
ho
ug
h i
t c
a
n
a
s
si
st
in
de
f
in
in
g t
he
r
i
gh
t
str
a
te
g
ie
s
a
nd a
c
ti
on
s f
or
t
he
ta
r
ge
te
d
m
a
r
ke
t.
T
he
pr
ob
le
m
tha
t
of
te
n
a
r
ise
s
in
m
a
r
ke
t
se
gm
e
nta
ti
on
i
s
ho
w
to e
x
pl
oi
t the
pr
e
f
e
r
e
nc
e
s of
st
ude
nt
s a
s c
u
st
om
e
r
s.
T
o ove
r
c
om
e
th
is,
t
he
c
om
b
ina
ti
on of
h
ybr
i
d c
lu
ste
r
in
g m
e
th
od
wi
th m
u
lt
ipl
e
c
r
ite
r
ia
w
il
l be
a
pp
lie
d to th
e
c
a
se
of
the
m
a
r
ke
t se
gm
e
nta
ti
on f
or
stude
nt
s in h
ig
he
r
e
duc
a
ti
on i
ns
ti
tu
ti
on
s.
The
in
te
gr
a
ti
on
of
ge
ogr
a
phi
c
,
de
m
o
gr
a
p
hic
,
psy
c
h
ogr
a
ph
ic
,
a
nd be
ha
v
ior
a
l c
r
i
te
r
ia
f
r
om
s
tu
de
nt
s is u
se
d t
o ge
t
mo
re
ins
ig
htf
ul
inf
or
m
a
t
io
n a
bo
ut
st
ude
nt pr
e
f
e
r
e
nc
e
.
Da
ta
r
e
sul
t of
t
he
i
nte
gr
a
t
io
n
wil
l be
pr
oc
e
s
se
d u
si
ng h
ybr
id c
l
us
te
r
i
ng
us
in
g K
-
m
e
a
n
s
a
nd
se
lf
or
ga
niz
in
g
map
(SO
M)
a
lgor
it
hm
.
The
hy
br
i
d c
lu
ste
r
in
g c
on
duc
t
e
d to ge
t pr
om
i
si
ng
c
lu
ste
r
in
g r
e
su
lt a
lo
ng
wi
th t
he
v
is
ua
l
iz
a
t
io
n of
se
gm
e
nta
ti
on.
T
hi
s st
ud
y
suc
c
e
ssf
ul
ly pr
od
uc
e
s f
i
ve
st
ude
nt se
gm
e
n
ts.
I
t r
e
c
e
i
ve
d
1,
386 a
s t
he
Da
vie
s
-
B
oul
di
n in
de
x (
DB
I
)
va
lue
a
n
d 2,
7
52 a
s the
q
ua
n
tiz
a
t
ion e
r
r
o
r
(
QE)
va
l
ue
whi
c
h
in
dic
a
te
s
a
go
od
c
lu
ste
r
in
g
r
e
su
lt
f
or
m
a
r
ke
t
se
g
m
e
nta
ti
on.
I
n a
ddi
ti
on,
the
v
is
ua
l
iz
a
t
io
n of
t
he
c
l
us
te
r
i
ng r
e
s
ul
t c
a
n b
e
se
e
n
in a
he
xa
go
na
l m
a
p.
Ke
y
wo
r
d
s
:
H
ybr
i
d c
l
us
t
e
r
i
ng
M
u
lti
-
c
r
ite
r
ia
s
e
g
me
n
ta
tio
n
K
-
m
ean
s
S
el
f
o
r
g
an
i
zi
n
g
m
ap
M
ar
k
et
s
eg
m
en
t
at
i
o
n
T
his
is
a
n
o
pe
n
ac
c
e
s
s
ar
tic
le
u
nde
r
the
CC
B
Y
-
SA
lic
e
n
se
.
C
or
r
e
s
pon
di
n
g A
u
t
h
or
:
H
a
r
di
ka
K
hus
nul
i
a
w
a
t
i
D
ep
ar
t
m
en
t
o
f
I
n
f
o
r
m
at
i
cs
U
ni
ve
r
s
i
t
a
s
S
a
hi
d S
ur
a
ka
r
t
a
A
di
S
uc
i
pt
o
S
t
.
N
o.
154,
J
a
j
a
r
,
L
a
w
e
ya
n,
t
he
c
i
t
y o
f
S
ur
a
ka
r
t
a
C
e
nt
r
a
l
J
a
va
pr
ovi
nc
e
57144,
I
ndone
s
i
a
E
ma
il:
ha
r
di
ka
.
khus
nul
i
a
@
gm
a
i
l
.
c
om
1.
I
NT
RO
DUC
T
I
O
N
M
a
r
ke
t
s
e
gm
e
nt
a
t
i
on i
s
i
nc
r
e
a
s
i
ngl
y be
i
ng i
m
pl
e
m
e
nt
e
d i
n
H
i
ghe
r
E
duc
a
t
i
on
I
n
s
titu
tio
n
s
(
H
E
I
s
).
It
i
s
us
e
d t
o
f
i
nd
t
he
m
ode
l
t
ha
t
he
l
p
de
f
i
ne
t
he
r
i
ght
s
t
r
a
t
e
gi
e
s
a
nd a
c
t
i
ons
f
or
t
he
t
a
r
ge
t
e
d
m
a
r
ke
t
[
1]
.
B
a
s
e
d
on
t
he
r
e
s
ul
t
i
ng
m
a
r
ke
t
s
e
gm
e
nt
,
H
E
I
s
c
a
n a
l
s
o
ga
i
n
i
ns
i
ght
s
i
nt
o po
t
e
nt
i
a
l
ba
s
e
s
f
o
r
s
e
gm
e
nt
a
t
i
on,
pos
i
t
i
oni
ng
an
d
co
m
m
u
n
i
cat
i
o
n
s
t
r
at
eg
i
es
t
ar
g
et
i
n
g
t
h
e
s
p
eci
f
i
c ch
ar
act
er
i
s
t
i
cs
o
f
each
s
eg
m
en
t
[
2]
.
T
he
ke
y poi
nt
of
m
a
r
ke
t
s
e
gm
e
nt
a
t
i
on pr
obl
e
m
i
s
how
t
o e
xpl
oi
t
t
he
pr
e
f
e
r
e
nc
e
s
of
c
us
t
om
e
r
s
,
be
c
a
us
e
i
t
pr
ovi
de
s
a
ba
s
i
s
f
or
t
he
f
ol
l
ow
i
ng
s
e
gm
e
nt
a
t
i
on de
c
i
s
i
on
[
3]
.
C
u
s
t
o
m
er
p
r
ef
e
r
en
ces
can
b
e
obt
a
i
ne
d by de
t
e
r
m
i
ni
ng
s
e
gm
e
nt
a
t
i
on c
r
i
t
e
r
i
a
i
n
or
de
r
t
o obt
a
i
n
i
nf
o
r
m
a
t
i
on t
ha
t
w
i
l
l
be
us
e
d f
o
r
m
a
r
ke
t
s
e
gm
e
nt
a
t
i
on
[
4]
.
T
h
es
e cr
ite
r
ia
c
a
n
b
e
a
s
in
g
le
c
r
ite
r
io
n
o
r
a
c
o
mb
in
a
tio
n
o
f
s
e
v
e
r
a
l c
r
ite
r
ia
c
a
lle
d
mu
ltip
le
c
r
ite
r
ia
.
M
a
r
k
e
t s
e
g
me
n
ta
tio
n
in
v
o
lv
in
g
mu
ltip
le
c
r
ite
r
ia
h
a
s
b
e
e
n
c
a
r
r
ie
d
o
u
t
to
a
n
a
ly
z
e
th
e
r
e
a
d
in
e
s
s
o
f
th
e
c
r
i
s
i
s
pe
r
i
ods
i
n t
e
r
m
s
of
f
ood s
e
c
ur
i
t
y us
i
ng t
he
vol
um
e
o
f
a
gr
i
c
ul
t
u
r
a
l
pr
oduc
t
i
on
,
t
he
ba
l
a
nc
e
o
f
f
or
e
i
gn
t
r
ad
e o
f
ag
r
i
-
f
ood p
r
oduc
t
s
a
s
w
e
l
l
a
s
t
he
s
t
r
uc
t
ur
e
of
t
ot
a
l
a
gr
i
-
f
ood
p
r
oduc
t
t
r
a
de
[
5]
.
I
t
a
l
s
o
us
e
d
t
o di
vi
de
t
he
t
our
i
s
t
s
e
gm
e
nt
i
n c
hoos
i
ng out
bound de
s
t
i
na
t
i
ons
[
6]
.
T
he
c
om
bi
na
t
i
on o
f
de
m
ogr
a
phi
c
,
ps
yc
hogr
a
phi
c
,
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
KA
T
el
eco
m
m
u
n
C
om
put
E
l
C
ont
r
o
l
H
y
br
i
d c
l
us
t
e
r
i
ng bas
e
d on m
ul
t
i
-
c
r
i
t
e
r
i
a s
e
gm
e
nt
at
i
on f
or
…
(
H
ar
di
k
a K
hus
nul
i
aw
at
i
)
1499
a
nd be
ha
vi
or
a
l
c
r
i
t
e
r
i
a
ha
s
a
l
s
o be
e
n us
e
d f
or
s
e
gm
e
nt
a
t
i
on of
t
he
w
i
ne
m
a
r
ke
t
w
he
r
e
t
he
r
e
s
ul
t
s
s
how
t
he
us
e
of
t
he
s
e
c
om
bi
na
t
i
on i
s
m
o
r
e
pr
om
i
s
i
ng
[
7]
.
T
he
i
nt
e
gr
a
t
i
on be
t
w
e
e
n ps
yc
hogr
a
phi
c
,
de
m
og
r
a
phi
c
,
a
nd
be
ha
vi
or
a
l
c
r
i
t
e
r
i
a
o
f
m
a
r
ke
t
s
e
gm
e
nt
s
i
n
f
i
na
nc
i
a
l
s
e
r
vi
c
e
s
pr
ovi
de
s
us
e
f
ul
i
ns
i
ght
i
nt
o
t
he
i
r
uni
que
c
us
t
om
e
r
be
ha
vi
or
[
8]
.
T
he
i
nvol
ve
m
e
nt
o
f
m
ul
t
i
c
r
i
t
e
r
i
a
i
n m
a
r
ke
t
s
e
gm
e
nt
a
t
i
on i
s
pr
ove
n
t
o be
a
bl
e
t
o a
dd us
e
f
ul
i
nf
or
m
a
t
i
on a
bout
c
us
t
om
e
r
s
i
n t
he
r
e
s
e
a
r
c
he
d f
i
e
l
d.
E
ve
n
t
hough t
he
us
e
of
m
u
l
t
i
pl
e
c
r
i
t
e
r
i
a
ha
s
be
e
n done
i
n
m
a
r
ke
t
s
e
gm
e
nt
a
t
i
on a
nd ot
he
r
f
i
e
l
ds
,
but
t
hi
s
ha
s
not
be
e
n w
i
de
l
y
i
m
pl
e
m
e
nt
e
d i
n
m
a
r
ke
t
s
e
gm
e
nt
a
t
i
on f
or
H
E
I
s
.
F
or
m
a
r
ke
t
s
e
gm
e
nt
a
t
i
on on H
E
I
s
,
t
he
c
r
i
t
e
r
i
a
us
e
d a
r
e
a
s
i
n t
he
f
ol
l
ow
i
ng s
t
udi
e
s
.
D
av
ar
i
e
t a
l.
[
1]
,
so
c
i
o
-
de
m
ogr
a
phi
c
a
nd ge
ogr
a
phi
c
i
nf
or
m
a
t
i
on i
s
us
e
d a
s
s
e
gm
e
n
t
a
t
i
on c
r
i
t
e
r
i
a
f
or
t
he
pur
pos
e
o
f
m
a
r
ke
t
s
e
gm
e
nt
a
t
i
on i
n t
he
e
duc
a
t
i
on i
ndus
t
r
y
.
A
not
h
e
r
c
r
i
t
e
r
i
on
us
e
d i
n m
a
r
ke
t
s
e
gm
e
nt
a
t
i
on f
o
r
H
E
I
s
i
s
ps
yc
hogr
a
phi
c
i
nf
or
m
a
t
i
on
[
2]
,
[
9]
w
hi
c
h p
r
ovi
de
s
a
de
e
pe
r
unde
r
s
t
a
ndi
ng o
f
s
t
ude
nt
s
a
s
c
us
t
om
e
r
s
.
W
hi
l
e
t
he
be
ha
vi
or
a
l
c
r
i
t
e
r
i
on i
s
a
ppl
i
e
d a
s
t
he
ba
s
i
s
f
o
r
m
a
r
ke
t
s
e
gm
e
nt
a
t
i
on t
o he
l
p pr
i
va
t
e
H
E
I
s
r
eco
g
n
i
ze
t
h
e
be
ha
vi
or
of
p
r
os
pe
c
t
i
ve
s
t
ude
nt
s
w
ho w
i
l
l
c
hoos
e
t
he
i
r
pl
a
c
e
t
o s
t
udy
[
1
0]
.
T
o
obt
a
i
n m
or
e
us
e
f
ul
i
nf
or
m
a
t
i
on
r
e
g
a
r
d
in
g
c
u
s
to
me
r
s
,
th
is
s
tu
d
y
w
ill a
p
p
ly
ma
r
k
e
t
s
e
g
me
n
ta
tio
n
w
ith
mu
ltip
le
c
r
ite
r
ia
.
T
h
e
c
r
ite
r
ia
th
a
t w
ill
be
i
nvol
ve
d a
r
e
c
r
i
t
e
r
i
a
t
ha
t
ha
ve
be
e
n w
i
de
l
y u
s
e
d i
n pr
e
vi
ous
s
t
udi
e
s
i
nc
l
ude
ge
ogr
a
phi
c
,
de
m
ogr
a
phi
c
,
ps
yc
hogr
a
phi
c
,
a
nd be
ha
vi
or
a
l
.
T
he
n,
i
n
f
or
m
a
t
i
on
obt
a
i
ne
d ba
s
e
d on m
ul
t
i
pl
e
c
r
i
t
e
r
i
a
be
c
om
e
s
t
he
i
nput
f
o
r
pa
r
t
i
t
i
oni
ng m
e
t
hod
i
n
m
a
r
ke
t
s
e
gm
e
nt
a
t
i
on.
T
he
m
os
t
popul
a
r
pa
r
t
i
t
i
on
i
n
g m
e
t
hod
i
s
K
-
m
ean
s
cl
u
s
t
er
i
n
g
[
4]
.
T
hi
s
m
e
t
hod
h
as
b
een
i
m
pl
e
m
e
nt
e
d f
or
s
e
gm
e
nt
a
t
i
on of
t
he
or
ga
ni
c
f
ood
m
a
r
ke
t
[
11]
,
m
obi
l
e
s
e
r
vi
c
e
pr
ovi
de
r
s
c
us
t
om
e
r
s
[
12]
,
a
n
d
i
n
t
er
n
et
ba
nki
ng
c
us
t
om
e
r
s
[
1
3]
.
M
ean
w
h
i
l
e,
f
o
r
m
ar
k
et
s
e
gm
e
nt
a
t
i
on
i
n H
E
I
,
K
-
me
a
n
s
c
lu
s
te
r
in
g
is
c
onduc
t
e
d i
n
t
he
s
e
c
ond s
t
a
ge
a
f
t
e
r
i
nt
e
r
na
l
va
l
i
da
t
i
on w
i
t
h
t
h
r
e
e
c
l
us
t
e
r
s
a
s
t
he
be
s
t
r
e
s
ul
t
[
2]
.
I
n
a
d
d
itio
n
,
it
i
s
a
l
s
o us
e
d i
n t
he
hi
ghe
r
e
duc
a
t
i
on m
a
r
ke
t
s
e
gm
e
nt
a
t
i
on t
o de
t
e
r
m
in
e
th
e
r
ig
h
t me
d
ia
to
ta
r
g
e
t
p
r
o
s
p
e
c
tiv
e
s
t
ude
nt
s
[
9]
.
A
l
t
hough K
-
m
e
a
ns
pr
ovi
de
s
pr
om
i
s
i
ng s
e
gm
e
nt
a
t
i
on r
e
s
ul
t
s
,
t
he
r
e
i
s
s
t
i
l
l
a
c
ha
nc
e
t
o
i
m
pr
ove
t
he
r
e
s
ul
t
s
by us
i
ng hyb
r
i
d m
e
t
hod
.
E
.
B.
Ba
y
l
a
n
[
14]
,
pr
opos
e
d
t
he
hyb
r
i
d
m
e
t
hod us
i
ng
an
al
y
t
i
c h
i
er
ar
ch
y
p
r
o
c
e
ss (
A
H
P
)
-
t
e
c
hni
que
f
or
or
de
r
pr
e
f
e
r
e
nc
e
by
s
i
m
i
l
a
r
i
t
y t
o
t
he
i
de
a
l
s
ol
ut
i
on
(
T
O
P
S
I
S
)
w
h
ic
h
is
a
p
p
lie
d
t
o t
he
p
r
o
j
e
c
t
r
i
sk
a
sse
ssm
e
n
t
i
s
a
bl
e
t
o pr
ovi
de
a
pl
a
t
f
or
m
t
ha
t
c
a
n be
a
na
l
yz
e
d qua
lita
tiv
e
ly
.
W
h
e
r
e
a
s
in
th
e
c
a
s
e
of
ne
t
w
or
k
s
e
c
ur
i
t
y,
t
he
i
nt
e
gr
a
t
i
on
of
m
e
t
hods
be
t
w
e
e
n
i
nt
e
r
ne
t
of
t
hi
ngs
(
I
oT
)
a
nd
bl
oc
kc
ha
i
n i
s
a
ppl
i
e
d t
o i
ndus
t
r
i
a
l
pr
oc
e
s
s
e
s
f
or
s
ol
vi
ng t
he
s
e
c
ur
i
t
y i
s
s
ue
s
i
n r
e
a
l
-
time
[
15]
.
T
he
hybr
i
d m
e
t
h
o
d
w
as
al
s
o
a
dopt
e
d by e
ne
r
gy c
om
pa
ni
e
s
i
n
t
he
pr
oc
e
s
s
of
i
m
pl
e
m
e
nt
i
ng t
he
s
t
r
a
t
e
gy o
f
a
da
pt
a
t
i
on
[
16]
.
M
en
aw
h
i
l
e,
s
ev
er
al
s
t
u
d
i
es
h
av
e i
n
t
eg
r
at
ed
K
-
m
ean
s
w
i
t
h ot
he
r
c
l
us
t
e
r
i
ng a
l
gor
i
t
hm
s
t
o
i
m
pr
ove
d
c
l
us
t
e
r
i
ng r
e
s
ul
t
.
T
he
vi
s
u
a
liz
a
tio
n
c
a
p
a
b
ilitie
s
o
f
s
el
f
o
r
g
an
i
zi
n
g
m
ap
(
S
O
M
)
c
o
mb
in
e
d
w
ith
K
-
m
ean
s
w
er
e
us
e
d t
o de
t
e
r
m
i
ne
t
he
s
pa
t
i
ot
e
m
por
a
l
pa
t
t
e
r
n of
w
a
t
e
r
qua
l
i
t
y da
t
a
a
nd i
de
nt
i
f
y s
our
c
e
s
of
pol
l
ut
i
on
[
17]
.
T
h
e
a
ppl
i
c
a
t
i
on of
t
he
S
O
M
a
l
gor
i
t
hm
w
i
t
h K
-
m
ean
s
f
or
m
a
r
ke
t
s
e
gm
e
nt
a
t
i
on i
n i
ns
ur
a
nc
e
c
om
pa
ni
e
s
i
s
a
bl
e
t
o
s
uppor
t
vi
s
ua
l
i
z
a
t
i
on a
nd pr
ovi
de
be
t
t
e
r
r
e
s
ul
t
s
w
he
n c
om
pa
r
e
d us
i
ng s
i
ngl
e
a
l
gor
i
t
hm
[
1
8]
.
I
t
can
b
e s
een
t
ha
t
t
he
hybr
i
d m
e
t
hod be
t
w
e
e
n K
-
m
ean
s
a
nd S
O
M
a
l
go
r
i
t
hm
c
a
n he
l
p i
n t
e
r
m
s
of
vi
s
ua
l
i
z
a
t
i
on a
nd a
c
c
ur
a
c
y
of
c
l
us
t
e
r
i
ng r
e
s
ul
t
s
.
H
ow
e
ve
r
,
t
he
us
e
of
t
he
c
om
bi
na
t
i
on of
hybr
i
d c
l
us
t
e
r
i
ng
m
e
t
hod w
i
t
h m
ul
t
i
pl
e
c
r
i
t
e
r
i
a
ha
s
not
be
e
n a
ppl
i
e
d t
o
t
he
c
a
s
e
of
t
he
m
a
r
ke
t
s
e
g
m
e
nt
a
t
i
on f
or
H
E
I
s
.
T
he
r
e
f
or
e
,
t
hi
s
s
t
udy
pr
opos
e
s
hybr
i
d c
l
us
t
e
r
i
ng ba
s
e
d on m
ul
t
i
-
c
r
i
t
e
r
i
a
s
e
gm
e
nt
a
t
i
on f
or
hi
ghe
r
e
duc
a
t
i
on m
a
r
ke
t
i
ng.
T
he
i
n
t
e
gr
a
t
i
on o
f
ge
ogr
a
p
hi
c
,
de
m
ogr
a
phi
c
,
ps
yc
hogr
a
phi
c
,
a
nd
be
ha
vi
or
a
l
c
r
i
t
e
r
i
a
f
r
om
s
t
ude
nt
s
i
s
us
e
d t
o ge
t
m
or
e
i
ns
i
ght
f
ul
i
n
f
or
m
a
t
i
on a
bout
s
t
ude
nt
pr
e
f
e
r
en
ce i
n
ch
o
o
s
i
n
g
H
E
I
s
.
Wh
er
eas
,
hybr
i
d c
l
us
t
e
r
i
ng us
i
ng K
-
m
ean
s
a
nd S
O
M
a
l
gor
i
t
hm
w
i
l
l
be
c
onduc
t
e
d t
o ge
t
pr
om
i
s
i
ng c
l
us
t
e
r
i
ng r
e
s
ul
t
a
l
ong w
i
t
h t
he
v
i
s
ua
l
i
z
a
t
i
on of
s
e
gm
e
nt
a
t
i
on.
2.
RE
S
E
ARCH
M
E
T
H
O
D
T
he
s
t
a
ge
s
of
m
a
r
ke
t
s
e
gm
e
nt
a
t
i
on on H
E
I
s
us
i
ng hybr
i
d c
l
us
t
e
r
i
ng
ba
s
e
d on t
he
m
u
l
t
i
-
c
r
ite
r
ia
s
e
gm
e
nt
a
t
i
on i
s
e
xpl
a
i
ne
d i
n
F
i
gur
e
1
.
T
he
r
e
a
r
e
f
our
s
t
a
ge
s
c
ons
i
s
t
i
ng of
da
t
a
c
ol
l
e
c
t
i
on,
da
t
a
e
xpl
or
a
t
i
on,
e
xt
r
a
c
t
i
ng s
e
gm
e
nt
s
,
a
nd
de
s
c
r
i
bi
ng
s
e
gm
e
nt
s
.
T
he
f
i
r
s
t
s
t
a
ge
i
s
c
ol
l
e
c
t
i
ng da
t
a
f
r
om
hi
ghe
r
e
duc
a
t
i
on
s
t
ude
nt
s
w
i
t
h t
he
s
pe
c
i
f
i
e
d c
r
i
t
e
r
i
a
.
T
he
m
os
t
c
om
m
on s
e
gm
e
nt
a
t
i
on c
r
i
t
e
r
i
a
a
r
e
ge
ogr
a
phi
c
,
de
m
ogr
a
phi
c
,
ps
yc
hogr
a
phi
c
a
nd be
ha
vi
our
a
l
[
4]
.
D
e
m
ogr
a
phi
c
a
nd ge
ogr
a
phi
c
i
nf
o
r
m
a
t
i
on i
s
obt
a
i
ne
d f
r
om
i
n
t
e
r
na
l
da
t
a
s
t
or
a
ge
ow
ne
d by H
E
I
s
a
s
r
e
s
e
a
r
c
h s
i
t
e
s
.
S
u
r
ve
y d
a
t
a
w
e
r
e
c
ol
l
e
c
t
e
d t
o
obt
a
i
n ps
yc
hogr
a
phi
c
a
nd
be
ha
vi
or
a
l
cr
i
t
er
i
a.
P
r
ep
r
o
ces
s
i
n
g
i
s
p
er
f
o
r
m
ed
at
t
h
e d
at
a
ex
p
l
o
r
at
i
o
n
s
t
ag
e.
T
h
i
s
s
t
ag
e i
n
cl
u
d
es
d
at
a cl
ean
i
n
g
,
d
at
a
i
nt
e
gr
a
t
i
on,
a
nd
da
t
a
t
r
a
ns
f
or
m
a
t
i
on
[
1
9]
,
[
2
0]
.
A
l
l
da
t
a
obt
a
i
ne
d f
r
om
t
he
pr
e
vi
ous
s
t
a
ge
w
i
l
l
be
pr
o
c
e
s
s
e
d a
t
t
he
e
xpl
or
a
t
i
on s
t
a
ge
.
T
he
r
e
s
ul
t
s
of
t
he
pr
e
p
r
o
c
e
s
s
e
d da
t
a
a
r
e
t
he
da
t
a
s
e
t
w
i
t
hout
m
i
s
s
i
ng
v
al
u
es
an
d
in
c
o
n
s
is
te
n
c
ie
s
w
ith
th
e
tr
a
n
s
f
o
r
ma
tio
n
in
to
n
u
me
r
ic
v
a
lu
e
s
.
A
t th
e
ex
t
r
act
i
n
g
s
eg
m
en
t
s
s
t
a
ge
,
s
t
ude
nt
s
a
r
e
c
l
us
t
e
r
e
d ba
s
e
d on t
he
da
t
a
obt
a
i
ne
d f
r
om
t
he
e
xpl
or
a
t
i
on s
t
a
ge
.
H
ybr
i
d c
l
us
t
e
r
i
ng us
i
ng K
-
m
ean
s
a
nd S
O
M
a
l
gor
i
t
hm
i
s
pr
oc
e
s
s
e
d f
ol
l
ow
i
n
g
F
i
gur
e
2
.
T
h
e
f
ir
s
t c
lu
s
te
r
in
g
a
lg
o
r
ith
m th
a
t imp
le
me
n
te
d
to
d
a
ta
is
S
O
M
a
lg
o
r
ith
m.
T
h
e
ma
in
s
te
p
s
o
f
th
is
c
lu
s
te
r
in
g
m
e
t
hod a
r
e
a
s
f
ol
l
ow
s
[
21]
:
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SSN
:
1693
-
6930
T
E
L
KOM
NI
KA
T
el
eco
m
m
u
n
C
om
put
E
l
C
ont
r
o
l
,
Vo
l
.
19
, N
o
.
5
,
O
c
t
obe
r
2021:
1498
-
1506
1500
−
I
ni
t
i
a
l
i
z
e
t
he
num
be
r
of
ne
ur
ons
d.
a
nd
t
he
s
i
z
e
of
a
pl
a
ne
g
r
i
d
m
a
p o
f
ne
ur
ons
m
.
T
h
e
i
n
p
u
t
l
ay
er
accep
t
t
r
a
i
ni
ng da
t
a
∈
w
i
t
h d ne
ur
ons
w
hi
l
e
t
he
out
put
l
a
ye
r
i
s
of
t
e
n l
a
i
d out
a
s
a
pl
a
ne
gr
i
d m
a
p w
i
t
h
=
×
ne
ur
ons
.
−
T
r
a
i
ne
d
S
O
M
m
ode
l
i
n
a
n
i
t
e
r
a
t
i
ve
pr
oc
e
s
s
i
nc
l
udi
ng c
om
pe
t
i
t
i
on
a
nd
c
onve
r
ge
nc
e
.
t
he
w
e
i
ght
s
ke
e
p
c
onne
c
t
i
on f
r
om
e
a
c
h
i
nput
ne
ur
on
t
o
e
a
c
h out
put
ne
u
r
on a
nd
c
a
n
be
de
not
e
d a
s
=
{
|
∈
,
=
1
,
…
,
}
.
−
F
ound t
he
w
i
nne
r
i
n t
he
c
om
pe
t
i
t
i
on,
t
ha
t
i
s
,
t
h
e
c
l
os
e
s
t
ne
ur
on c
t
o t
he
i
nput
s
a
m
pl
e
x(
t
)
i
n t
h
e
t
t
h
i
t
er
at
i
v
e s
t
ep
.
T
h
e v
al
u
e o
f
c can
b
e
o
b
t
ai
n
ed
b
y
(
1)
.
=
ar
g
m
in
{
‖
(
)
−
(
)
‖
}
(
1
)
−
T
he
w
e
i
ght
ve
c
t
or
s
upda
t
e
ba
s
e
d on t
he
ne
i
ghbor
h
ood r
e
l
a
t
i
ons
hi
ps
w
i
t
h t
he
w
i
nne
r
ne
ur
on
by
(
2)
.
(
+
1
)
=
(
)
+
ℎ
(
)
(
(
)
−
(
)
)
(
2
)
w
h
er
e
ℎ
(
)
i
s
t
he
ne
i
ghbor
hood f
unc
t
i
on t
ha
t
de
t
e
r
m
i
ne
s
t
he
ne
i
ghbor
upda
t
e
s
c
he
m
e
f
or
t
opol
ogy
-
pr
e
s
e
r
vi
ng na
t
ur
e
of
S
O
M
w
hi
c
h f
ol
l
ow
s
(
3
)
i
n t
he
f
or
m
of
G
a
us
s
i
a
n
f
unc
t
i
on.
ℎ
(
)
=
(
)
−
(
,
)
2
2
(
)
(
3
)
w
h
er
e
(
)
is
th
e
le
a
r
n
in
g
r
a
te
th
a
t
mo
n
o
to
n
ic
a
lly
d
e
c
r
e
a
s
e
s
w
ith
s
te
p
t
,
(
,
)
i
s
t
he
s
qua
r
e
of
di
s
t
a
nc
e
be
t
w
e
e
n ne
ur
on
c
a
nd ne
u
r
on
i
on
t
he
pl
a
ne
g
r
i
d
m
a
p
a
nd
σ
(
t
)
i
s
t
he
ke
r
ne
l
r
a
di
us
t
ha
t
de
t
e
r
m
i
ne
s
t
he
r
a
nge
of
ne
i
ghbor
hood
r
e
l
a
t
i
ons
hi
ps
.
T
he
S
O
M
t
r
a
i
n
i
ng
m
ode
l
c
a
n be
r
e
pr
e
s
e
nt
e
d us
i
ng U
m
a
t
r
i
x,
hi
s
t
ogr
a
m
c
ol
l
i
s
i
on
a
nd ot
he
r
i
nde
xe
s
to
v
is
u
a
l th
e
d
a
ta
s
e
ts
[
22]
.
I
n
th
is
p
a
p
e
r
,
g
r
a
y
s
c
a
le
o
f
U
ma
tr
ix
a
n
d
h
it
h
is
to
g
r
a
m
is
u
s
e
d
le
v
e
l
to
d
e
te
r
min
e
th
e
r
e
la
tio
n
s
h
ip
s
b
e
tw
e
e
n
th
e
mu
ltip
le
c
r
i
t
e
r
i
a
o
f
hi
ghe
r
e
duc
a
t
i
on
s
t
ude
nt
s
a
nd
t
o
r
e
duc
e
t
he
m
t
o a
f
e
w
i
nde
pe
nde
nt
f
e
a
t
ur
e
s
.
I
t
a
l
s
o pr
oduc
e
s
vi
s
ua
l
i
m
a
ge
s
of
e
a
c
h c
l
us
t
e
r
[
2
3]
.
F
i
gur
e
1
.
T
he
s
t
a
ge
s
of
t
he
p
r
opos
e
d r
e
s
e
a
r
c
h
m
e
t
hodol
ogy
F
i
gur
e
2
.
H
ybr
i
d
c
l
us
t
e
r
i
ng us
i
ng K
-
m
ean
s
a
nd S
O
M
a
l
gor
i
t
hm
T
he
n,
K
-
m
ean
s
i
s
us
e
d t
o r
e
-
c
lu
s
te
r
th
e
f
e
a
tu
r
e
s
r
e
s
u
lte
d
f
r
o
m S
O
M
.
T
h
e
s
te
p
s
o
f
c
lu
s
te
r
in
g
w
ith
K
-
m
ean
s
m
e
t
hod a
r
e
a
s
f
ol
l
ow
s
[
24]
:
−
S
el
ect
t
h
e n
u
m
b
er
o
f
cl
u
s
t
er
s
k
an
d
i
n
i
t
i
al
i
ze
t
h
e cl
u
s
t
er
cen
t
er
k
.
C
l
u
s
t
er
cen
t
er
s
ar
e as
s
i
g
n
ed
i
n
i
t
i
al
va
l
ue
s
w
i
t
h r
a
ndom
num
be
r
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
KA
T
el
eco
m
m
u
n
C
om
put
E
l
C
ont
r
o
l
H
y
br
i
d c
l
us
t
e
r
i
ng bas
e
d on m
ul
t
i
-
c
r
i
t
e
r
i
a s
e
gm
e
nt
at
i
on f
or
…
(
H
ar
di
k
a K
hus
nul
i
aw
at
i
)
1501
−
A
llo
c
a
te
a
ll d
a
ta
o
b
je
c
ts
to
th
e
n
e
a
r
e
s
t c
lu
s
te
r
.
T
h
e
p
r
o
x
imity
o
f
a
d
a
ta
to
a
p
a
r
tic
u
la
r
c
lu
s
te
r
is
d
e
te
r
min
e
d
th
e
d
is
ta
n
ce
b
et
w
een
t
h
e
d
at
a
w
i
t
h
t
h
e cl
u
s
t
er
cen
t
er
.
T
o
d
i
s
t
an
ce al
l
d
at
a
t
o
each
c
l
us
t
e
r
c
e
nt
e
r
poi
nt
c
a
n us
e
E
uc
l
i
de
a
n di
s
t
a
nc
e
t
he
or
y f
o
r
m
ul
a
t
e
d a
s
f
ol
l
ow
s
(
4)
:
=
1
−
1
2
+
2
−
2
2
+
⋯
+
−
2
(
4
)
−
R
ecal
cu
l
at
e cl
u
s
t
er
cen
t
er
w
i
t
h
cu
r
r
en
t
cl
u
s
t
er
m
em
b
er
s
h
i
p
.
T
h
e n
ew
cl
u
s
t
er
cen
t
er
i
s
t
h
e av
er
ag
e o
f
al
l
d
at
a i
n
a p
ar
t
i
cu
l
ar
cl
u
s
t
er
.
−
R
eas
s
i
g
n
each
o
b
j
ect
u
s
i
n
g
t
h
e
n
ew
cl
u
s
t
er
cen
t
er
.
I
f
t
h
e
cl
u
s
t
er
cen
t
er
d
o
es
n
o
t
ch
an
g
e ag
ai
n
t
h
en
t
h
e
cl
u
s
t
er
i
n
g
p
r
o
ces
s
i
s
co
m
p
l
et
e.
−
O
t
h
er
w
i
s
e,
r
e
t
ur
n
t
o s
t
e
p num
be
r
2 un
t
i
l
t
he
c
e
nt
e
r
of
t
he
c
l
us
t
e
r
doe
s
not
c
ha
nge
a
nym
or
e
.
T
he
e
va
l
ua
t
i
on m
e
t
hod of
c
l
us
t
e
r
r
e
s
ul
t
s
us
e
s
t
he
D
a
vi
e
s
-
B
oul
di
n
i
nde
x
(
D
B
I
)
a
nd qua
nt
i
z
a
t
i
on
e
r
r
or
.
D
B
I
r
e
pr
e
s
e
nt
s
how
w
e
l
l
c
l
us
t
e
r
i
ng ha
s
b
e
e
n done
by c
a
l
c
ul
a
t
i
ng t
he
qua
nt
i
t
y a
nd t
he
de
r
i
va
t
i
ve
f
eat
u
r
es
o
f
t
h
e d
at
as
et
[
2
5]
.
D
B
I
is
th
e
in
te
r
n
a
l
v
a
l
id
ity
,
w
h
ic
h
if
it
is
c
lo
s
e
r
to
n
o
n
-
ne
ga
t
i
ve
z
e
r
o
t
he
n s
how
s
t
h
e t
h
e b
et
t
er
cl
u
s
t
er
r
es
u
l
t
s
.
T
h
e q
u
an
t
i
zat
i
o
n
er
r
o
r
(
Q
E
)
m
eas
u
r
es
t
h
e av
er
ag
e s
q
u
ar
ed
d
i
s
t
an
ce f
r
o
m
each
poi
nt
t
o t
he
c
e
nt
r
oi
d o
f
t
he
c
l
us
t
e
r
w
he
r
e
i
t
be
l
on
gs
,
a
c
c
or
di
ng t
o
t
he
pa
r
t
i
t
i
on
i
nt
o
num
be
r
o
f
s
ubs
e
t
s
[
26]
.
L
o
w
er
v
al
u
es
f
o
r
Q
E
ar
e b
et
t
er
.
F
o
r
t
h
e l
as
t
s
t
ag
e,
t
h
e c
l
u
s
t
er
l
ab
el
s
w
hi
c
h
ar
e
t
he
r
e
s
ul
t
of
K
-
m
ean
s
c
lu
s
te
r
in
g
ar
e m
ap
p
ed
t
o
S
O
M
m
ode
l
s
t
ha
t
ha
ve
be
e
n obt
a
i
ne
d f
r
om
t
he
pr
e
vi
o
u
s
cl
u
s
t
er
p
r
o
ces
s
.
T
h
e
v
is
u
a
liz
a
tio
n
o
f
c
lu
s
te
r
r
e
s
u
lts
f
r
o
m
S
O
M
a
l
gor
i
t
hm
i
n
t
he
f
or
m
o
f
U
M
a
t
r
i
x
c
a
n
he
l
p i
n
unde
r
s
t
a
ndi
ng
t
he
s
t
r
uc
t
ur
e
o
f
e
a
c
h
s
t
ude
nt
s
e
gm
e
nt
obt
a
i
ne
d f
r
om
t
he
c
l
us
t
e
r
i
ng
r
e
s
ul
t
s
[
27]
.
3.
R
ES
U
LTS
A
ND ANAL
YS
I
S
T
hi
s
r
e
s
e
a
r
c
h w
a
s
a
ppl
i
e
d t
o
230
s
t
ude
nt
s
f
r
om
t
he
S
a
hi
d U
ni
ve
r
s
i
t
y
of
S
ur
a
ka
r
t
a
a
s
one
o
f
t
he
p
r
i
v
at
e H
E
I
s
i
n
t
h
e r
eg
i
o
n
o
f
C
en
t
r
al
J
av
a,
I
n
d
o
n
es
i
a.
T
h
e r
es
u
l
t
i
n
g
d
at
as
et
f
r
o
m
t
h
e d
at
a co
l
l
ect
i
o
n
s
t
ag
e
c
ons
i
s
t
s
of
230
obj
e
c
t
s
w
i
t
h
a
t
ot
a
l
o
f
4
8
f
eat
u
r
es
.
T
h
e
f
eat
u
r
es
o
f
t
h
e
d
at
as
et
r
ep
r
es
en
t
t
h
e
cr
i
t
e
r
i
a
t
h
at
h
av
e
be
e
n de
t
e
r
m
i
ne
d t
o be
us
e
d i
n t
hi
s
s
t
udy
(
ge
ogr
a
ph
i
c
,
de
m
ogr
a
phi
c
,
ps
yc
hogr
a
phi
c
,
a
nd
be
ha
vi
or
a
l
)
.
A
f
eat
u
r
e r
ep
r
es
en
t
i
n
g
g
eo
g
r
ap
h
i
c cr
i
t
er
i
a i
s
an
a
d
d
r
es
s
.
M
ean
w
h
i
l
e,
ag
e,
r
el
i
g
i
o
n
a
nd ge
nde
r
a
r
e
f
eat
u
r
es
t
h
at
r
ep
r
es
en
t
d
em
o
g
r
ap
h
i
c cr
i
t
er
i
a.
F
o
r
f
e
at
u
r
es
t
h
at
r
ep
r
es
en
t
p
s
y
ch
o
g
r
ap
h
i
c an
d
b
eh
av
i
o
r
al
cr
i
t
er
i
a
a
r
e
t
he
a
ns
w
e
r
s
of
que
s
t
i
onna
i
r
e
s
t
o
s
t
ude
nt
s
.
E
a
c
h que
s
t
i
onna
i
r
e
i
t
e
m
r
e
pr
e
s
e
nt
s
one
f
e
a
t
ur
e
.
T
he
r
e
a
r
e
36
ite
ms
th
a
t b
e
c
o
me
a
f
e
a
t
ur
e
t
o
r
e
pr
e
s
e
nt
t
he
ps
yc
hogr
a
phi
c
c
r
i
t
e
r
i
a
.
F
or
be
ha
vi
or
a
l
c
r
i
t
e
r
i
a
a
r
e
ba
s
e
d on t
he
a
ns
w
e
r
s
of
7 i
t
e
m
s
i
n
t
he
que
s
t
i
onna
i
r
e
.
A
l
l
i
t
e
m
s
i
n t
he
que
s
t
i
onna
i
r
e
w
e
r
e
a
r
r
a
nge
d on
5
-
poi
nt
lik
e
r
t
-
t
ype
s
cal
es
(
1
=co
m
p
l
et
el
y
d
i
s
ag
r
ee t
o
5
=co
m
p
l
et
el
y
ag
r
ee)
[
28]
.
3.
1.
Hy
b
r
id
cl
u
s
t
eri
n
g
s
eg
m
en
t
a
t
i
o
n
D
a
ta
f
e
a
tu
r
e
s
w
ith
c
a
te
g
o
r
ic
a
l ty
p
e
s
is
tr
a
n
s
f
o
r
me
d
in
to
n
u
me
r
ic
v
a
lu
e
s
o
th
a
t
it c
a
n
b
e
ma
x
ima
lly
pr
oc
e
s
s
e
d by t
he
c
l
us
t
e
r
i
ng m
e
t
hod.
A
l
l
f
e
a
t
u
r
e
s
a
r
e
pr
oc
e
s
s
e
d w
i
t
h
S
O
M
c
l
us
t
e
r
i
ng i
n or
de
r
t
o
obt
a
i
n
a
S
O
M
m
ode
l
t
ha
t
c
a
n be
v
i
s
ua
l
i
z
e
d i
n t
he
f
or
m
of
U
m
a
t
ri
x
.
F
i
gur
e
3
s
how
s
t
he
f
e
a
t
ur
e
m
a
ps
f
o
r
e
ve
r
y c
l
us
t
e
r
a
nd
f
or
a
l
l
i
nput
a
t
t
r
i
bu
t
e
s
.
T
he
S
O
M
m
a
ps
w
e
r
e
f
or
m
e
d us
i
ng
a
n
8
×
9
he
xa
gona
l
a
r
c
hi
t
e
c
t
ur
e
w
i
t
h
a
G
a
us
s
i
a
n ne
i
ghbour
hood
f
unc
t
i
on ba
s
e
d on
t
he
opt
i
m
um
num
be
r
of
ne
ur
ons
(
M
=
76
)
.
T
he
opt
i
m
u
m
num
be
r
o
f
ne
ur
ons
i
s
c
a
l
c
ul
a
t
e
d
(
5
)
.
≈
5
√
(
5
)
w
h
er
e
M
i
s
t
he
num
be
r
of
ne
ur
ons
,
w
hi
c
h i
s
a
n i
nt
e
ge
r
c
l
os
e
t
o t
he
r
e
s
ul
t
of
t
he
r
i
gh
t
ha
nd s
i
de
of
t
he
e
qua
t
i
on
a
nd
N
i
s
t
he
num
be
r
of
obj
e
c
t
s
[
29]
.
R
e
l
a
t
i
ons
hi
ps
be
t
w
e
e
n f
e
a
t
ur
e
s
c
oul
d be
i
ns
pe
c
t
e
d by vi
s
ua
l
l
y c
om
pa
r
i
ng
t
he
pa
t
t
e
r
n
o
f
s
h
ad
ed
p
i
x
el
s
f
o
r
each
m
ap
[
23]
.
T
h
e
f
eat
u
r
es
w
i
t
h
s
i
m
i
l
a
r
p
at
t
e
r
n
s
ar
e
co
n
s
i
d
er
ed
h
i
g
h
l
y
co
r
r
el
at
ed
an
d
b
e
r
em
o
v
ed
f
r
o
m
t
h
e f
eat
u
r
e s
et
[
3
0]
.
T
he
r
e
f
or
e
,
ba
s
e
d on
F
i
gu
e
3
,
a
de
t
a
i
l
e
d i
ns
pe
c
t
i
on of
th
e
s
imila
r
f
e
a
tu
r
e
p
a
tte
r
n
is
car
r
i
ed
o
u
t
.
F
o
r
ex
am
p
l
e,
s
ev
er
al
f
eat
u
r
es
s
u
ch
as
C
3
an
d
C
4
h
av
e t
h
e s
am
e p
at
t
er
n
as
C
2
.
T
h
u
s
,
C
3
an
d
C
4
c
a
n be
r
e
m
ove
d f
r
om
t
he
f
e
a
t
ur
e
s
e
t
.
F
i
na
l
l
y
,
w
e
g
ot
26 f
e
a
t
u
r
e
s
a
s
ne
w
i
nput
s
f
or
t
he
ne
xt
c
l
us
t
e
r
i
ng
pr
oc
e
s
s
.
T
h
i
s r
e
s
e
a
r
c
h us
e
s
e
l
bow
m
e
t
hod t
o de
t
e
r
m
i
ne
t
he
opt
i
m
a
l
num
be
r
of
c
l
us
t
e
r
s
[
31]
.
F
r
om
t
he
g
r
a
ph
s
how
n i
n F
i
gur
e
4
,
t
he
opt
i
m
a
l
num
be
r
o
f
c
l
us
t
e
r
s
i
s
f
i
ve
.
T
he
va
l
ue
o
f
t
he
c
l
us
t
e
r
k
=
4
t
o
k
=
5,
t
he
n f
r
om
k
=5
t
o
k
=
6
,
it s
h
o
w
s
a
d
r
a
s
tic
de
c
r
e
a
s
e
t
o f
o
r
m
t
he
e
l
bow
a
t
poi
nt
k
=
5 t
he
n
t
he
i
de
a
l
num
be
r
of
c
l
us
t
e
r
k
i
s
5.
T
he
n,
K
-
m
e
a
ns
a
l
gor
i
t
hm
i
s
us
e
d t
o di
vi
de
t
he
da
t
a
of
s
t
ude
nt
s
w
i
t
h
26 f
e
a
t
ur
e
s
i
n
t
o 5
c
l
us
t
e
r
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SSN
:
1693
-
6930
T
E
L
KOM
NI
KA
T
el
eco
m
m
u
n
C
om
put
E
l
C
ont
r
o
l
,
Vo
l
.
19
, N
o
.
5
,
O
c
t
obe
r
2021:
1498
-
1506
1502
F
i
gur
e
3
.
F
e
a
t
ur
e
m
a
p of
S
O
M
m
ode
l
ba
s
e
d on
m
u
l
t
i
-
cr
i
t
er
i
a
s
e
g
me
n
ta
tio
n
F
i
gur
e
4
.
T
he
gr
a
ph o
f
e
l
bow
m
e
t
hod f
o
r
opt
i
m
a
l
n
um
be
r
of
c
l
us
t
e
r
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
KA
T
el
eco
m
m
u
n
C
om
put
E
l
C
ont
r
o
l
H
y
br
i
d c
l
us
t
e
r
i
ng bas
e
d on m
ul
t
i
-
c
r
i
t
e
r
i
a s
e
gm
e
nt
at
i
on f
or
…
(
H
ar
di
k
a K
hus
nul
i
aw
at
i
)
1503
3.
2.
Ma
rk
et
s
e
gm
e
n
t
at
i
on
e
val
u
at
i
on
T
he
e
va
l
ua
t
i
on of
m
a
r
ke
t
s
e
gm
e
nt
a
t
i
on w
i
t
h
c
l
us
t
e
r
a
na
l
ys
i
s
i
s
done
by f
i
ndi
ng
t
he
D
B
I
va
l
ue
a
nd
qua
nt
i
z
a
t
i
on e
r
r
or
(
Q
E
)
.
T
he
r
e
s
ul
t
s
f
r
om
T
a
bl
e
1
s
how
a
s
m
a
l
l
va
l
ue
of
D
B
I
a
nd Q
E
,
s
o i
t
c
a
n be
c
onc
l
ude
d
th
a
t c
lu
s
te
r
in
g
b
a
s
e
d
o
n
mu
lti
-
cr
i
t
er
i
a s
eg
m
en
t
at
i
o
n
p
e
r
f
o
r
ms
a
n
o
p
tima
l
r
e
s
u
lt.
I
t me
a
n
s
th
a
t th
e
d
is
ta
n
c
e
b
et
w
een
t
h
e cl
u
s
t
er
s
i
s
l
ar
g
e an
d
t
h
e
d
i
s
t
an
ce b
et
w
een
t
h
e o
b
j
ect
s
i
n
t
h
e cl
u
s
t
er
i
s
s
m
al
l
[
3
2]
.
I
n a
ddi
t
i
on,
t
he
D
B
I
va
l
ue
i
s
1,
386
a
nd t
he
Q
E
va
l
ue
i
s
2,
752 w
he
n
us
i
ng hybr
i
d c
l
us
t
e
r
i
ng
.
W
hi
l
e
us
i
ng s
i
ngl
e
c
l
us
t
e
r
i
ng
t
he
D
B
I
va
l
ue
i
s
1,
562
a
n
d t
he
Q
E
va
l
ue
i
s
4
,
033.
F
r
om
t
he
T
a
bl
e
1,
i
t
c
a
n a
l
s
o be
s
e
e
n t
ha
t
t
he
D
B
I
a
nd Q
E
va
l
ue
s
us
i
ng hybr
i
d
c
l
us
t
e
r
i
ng a
r
e
s
m
a
l
l
e
r
t
ha
n us
i
ng s
i
ngl
e
c
l
us
t
e
r
i
ng.
T
hi
s
i
ndi
c
a
t
e
s
t
ha
t
t
he
pe
r
f
or
m
a
nc
e
of
hybr
i
d
c
l
us
t
e
r
i
ng us
i
ng K
-
m
ean
s
an
d
S
O
M
al
g
o
r
i
t
h
m
h
as
ex
ceed
ed
t
h
e
pe
r
f
or
m
a
nc
e
us
i
ng a
s
i
ngl
e
c
l
us
t
e
r
i
ng
m
e
t
hod.
T
ab
l
e 1
.
T
h
e
pe
r
f
or
m
a
nc
e
c
om
pa
r
i
s
on
of
c
l
us
t
e
r
i
n
g us
i
ng
s
i
ngl
e
a
nd hybr
i
d
m
e
t
hod
M
e
t
hod
DB
I
QE
S
i
ngl
e
c
l
us
t
e
r
i
ng us
i
ng K
-
m
ean
s
1,562
4,033
H
ybr
i
d c
l
us
t
e
r
i
ng us
i
ng K
-
m
ean
s
a
nd S
O
M
1,386
2,752
3.
3.
V
i
s
u
al
i
z
at
i
on
of
s
eg
m
en
t
a
t
i
o
n
res
u
l
t
T
he
r
e
s
ul
t
s
of
m
a
r
ke
t
s
e
gm
e
nt
a
t
i
on us
i
ng hybr
i
d c
l
us
t
e
r
i
ng m
e
t
hod ba
s
e
d on m
ul
t
i
-
cr
i
t
er
i
a ar
e
vi
s
ua
l
i
z
e
d i
n
F
i
gur
e
5
.
T
he
f
i
gur
e
i
s
S
O
M
m
ode
l
ge
ne
r
a
t
e
d f
r
o
m th
e
e
x
p
e
r
ime
n
ts
.
A
ll o
b
je
c
ts
f
r
o
m th
e
da
t
a
s
e
t
a
r
e
c
l
a
s
s
i
f
i
e
d i
nt
o e
a
c
h ne
ur
on i
n t
he
f
or
m
of
a
he
xa
gon hi
t
s
m
a
p.
F
r
om
t
he
vi
s
ua
l
i
z
a
t
i
on
r
e
s
ul
t
s
,
i
t
c
a
n be
s
e
e
n t
he
num
be
r
o
f
obj
e
c
t
s
f
o
r
e
a
c
h c
l
us
t
e
r
a
nd t
he
i
r
di
s
t
r
i
but
i
on
on t
he
m
a
p.
T
he
va
r
i
a
t
i
o
n i
n t
h
e
c
ol
or
s
c
he
m
e
f
r
om
l
i
ght
bl
ue
t
o
da
r
k
bl
ue
i
s
s
how
i
ng t
he
a
ve
r
a
ge
E
uc
l
i
de
a
n di
s
t
a
nc
e
be
t
w
e
e
n a
dj
a
c
e
nt
node
s
.
T
he
c
ol
or
de
gr
a
de
be
t
w
e
e
n da
r
k b
l
ue
t
o l
i
ght
bl
ue
s
how
i
ng t
he
r
e
duc
t
i
on i
n di
s
t
a
nc
e
be
t
w
e
e
n node
s
[
18]
.
D
ar
k
b
l
u
e ar
e
a
s
a
r
e
r
e
p
r
e
s
e
n
tin
g
d
is
s
imila
r
ity
(
la
r
g
e
E
u
c
lid
e
a
n
d
is
ta
n
c
e
s
b
e
tw
e
e
n
o
b
je
c
ts
)
.
W
h
ile
th
e
lig
h
t
b
l
u
e ar
eas
ar
e r
ep
r
es
en
t
i
n
g
s
i
m
i
l
a
r
i
t
y
(
s
m
al
l
E
u
cl
i
d
ean
d
i
s
t
an
ces
b
et
w
een
o
b
j
ect
s
)
.
C
l
u
s
t
er
0
2
an
d
C
l
u
s
t
er
0
3
ha
ve
t
he
m
os
t
m
e
m
be
r
s
a
nd oc
c
upy
l
i
ght
bl
ue
a
r
e
a
s
.
M
ean
w
h
i
l
e,
t
h
e d
a
r
k
b
l
u
e
ar
ea i
s
o
ccu
p
i
ed
b
y
a cl
u
s
t
er
w
i
t
h f
e
w
e
r
m
e
m
be
r
s
,
na
m
e
l
y,
C
l
us
t
e
r
00,
C
l
us
t
e
r
0
1,
a
nd C
l
us
t
e
r
04
.
F
i
gur
e
5
.
V
is
u
a
liz
a
tio
n
o
f
c
lu
s
te
r
r
e
s
u
lts
w
ith
th
e
h
its
ma
p
3.
4.
A
n
al
ys
i
s
of
s
e
gm
e
n
t
p
r
o
f
i
l
e
s
T
o s
t
udy t
he
ch
ar
act
er
i
s
t
i
c
o
f
cu
s
t
o
m
er
s
w
i
t
h
i
n
each
cl
u
s
t
er
,
t
h
e f
eat
u
r
e p
l
an
e
of
S
O
M
m
ode
l
(
F
i
gur
e
3
)
a
n
d
th
e
h
its
ma
p
(
F
i
g
ur
e
5
)
a
r
e
us
e
d.
B
y c
om
pa
r
i
ng
t
he
c
l
us
t
e
r
i
ng m
a
p
w
i
t
h
t
he
f
e
a
t
ur
e
p
l
a
ne
s
of
S
O
M
m
ode
l
,
t
he
f
ol
l
ow
i
ng
a
na
l
ys
i
s
of
s
e
gm
e
nt
r
e
s
ul
t
s
c
a
n be
c
onc
l
ude
d:
−
C
l
us
t
e
r
00:
M
e
m
be
r
s
of
t
hi
s
c
l
us
t
e
r
a
r
e
dom
i
na
t
e
d by s
t
ude
nt
s
f
r
om
t
he
K
a
l
i
m
a
nt
a
n
r
e
gi
on,
I
ndon
e
s
i
a
.
T
h
e
c
h
a
r
a
c
te
r
is
tic
s
o
f
th
is
c
lu
s
te
r
a
r
e
s
imila
r
to
C
l
u
s
te
r
0
4
.
T
h
e
a
g
e
s
o
f
s
tu
d
e
n
ts
w
h
o
a
r
e
me
mb
e
r
s
o
f
th
is
cl
u
s
t
er
ar
e 1
6
-
25.
W
he
n vi
e
w
e
d f
r
om
t
he
r
e
a
s
o
ns
f
or
c
hoos
i
ng t
he
i
r
uni
ve
r
s
i
t
y,
t
he
qua
l
i
t
y f
a
c
t
or
of
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SSN
:
1693
-
6930
T
E
L
KOM
NI
KA
T
el
eco
m
m
u
n
C
om
put
E
l
C
ont
r
o
l
,
Vo
l
.
19
, N
o
.
5
,
O
c
t
obe
r
2021:
1498
-
1506
1504
e
duc
a
t
i
on got
a
l
ow
e
r
s
c
or
e
w
he
n c
om
pa
r
e
d t
o
t
he
s
c
or
e
of
C
l
us
t
e
r
04 m
e
m
be
r
s
.
T
he
i
nvo
l
ve
m
e
nt
of
s
tu
d
e
n
ts
f
r
o
m me
mb
e
r
s
o
f
th
is
c
lu
s
te
r
in
c
a
mp
u
s
a
n
d
s
o
c
ia
l me
d
ia
a
c
tiv
itie
s
w
a
s
h
ig
h
e
r
th
a
n
in
c
lu
s
te
r
4.
−
C
l
us
t
e
r
01:
M
e
m
be
r
s
of
t
h
i
s
c
l
us
t
e
r
a
r
e
s
t
ude
n
t
s
f
r
om
P
a
pua
,
I
ndone
s
i
a
,
w
i
t
h
t
he
l
e
a
s
t
num
b
e
r
of
m
e
m
be
r
s
a
nd dom
i
na
t
e
d by f
e
m
a
l
e
s
t
ude
nt
s
.
T
he
a
ge
of
t
he
s
t
ude
nt
s
i
n t
hi
s
c
l
us
t
e
r
i
s
16
-
25 ye
a
r
s
.
A
m
ong
a
ll th
e
c
lu
s
te
r
s
,
f
a
mily
mo
tiv
a
tio
n
is
th
e
h
ig
h
e
s
t
f
a
c
t
or
t
ha
t
e
nc
our
a
ge
s
c
ont
i
nui
ng e
duc
a
t
i
on t
o
hi
ghe
r
e
duc
a
t
i
on.
I
n
a
ddi
t
i
on,
t
he
i
nf
l
ue
nc
e
of
a
dve
r
t
i
s
e
m
e
nt
s
a
nd pr
om
ot
i
ons
,
e
s
pe
c
i
a
l
l
y r
e
c
om
m
e
nda
t
i
ons
f
r
om
f
r
i
e
nds
a
nd f
a
m
i
l
y
,
i
s
t
he
hi
ghe
s
t
f
a
c
t
o
r
f
o
r
c
hoos
i
n
g t
he
i
r
uni
ve
r
s
i
t
y c
om
pa
r
e
d t
o
ot
he
r
c
l
us
t
e
r
s
.
−
C
l
us
t
e
r
02:
T
he
m
e
m
be
r
s
o
f
t
hi
s
c
l
us
t
e
r
a
r
e
dom
i
na
t
e
d by
s
t
ude
nt
s
f
r
om
J
a
va
,
I
ndone
s
i
a
w
i
t
h
t
he
s
e
c
ond
l
ar
g
es
t
n
u
m
b
er
o
f
m
em
b
er
s
o
f
al
l
cl
u
s
t
er
s
.
T
h
e ag
e o
f
t
h
e s
t
u
d
en
t
s
w
h
o
ar
e m
em
b
er
s
o
f
t
h
i
s
cl
u
s
t
er
v
ar
i
es
w
i
t
h s
om
e
s
t
ude
nt
s
a
ge
d
ove
r
36
ye
a
r
s
.
T
he
nu
m
be
r
of
m
a
l
e
a
nd
f
e
m
a
l
e
s
t
ude
nt
s
i
s
e
qua
l
.
C
om
pa
r
e
d t
o
C
lu
s
te
r
0
1
,
me
mb
e
r
s
o
f
th
is
c
lu
s
te
r
h
a
v
e
h
ig
h
e
r
e
c
o
n
o
mic
mo
tiv
a
tio
n
,
e
s
p
e
c
ia
lly
r
e
la
te
d
to
a
f
f
o
r
d
a
b
le
t
ui
t
i
on
f
e
e
s
a
nd s
c
hol
a
r
s
hi
p of
f
e
r
s
.
A
not
he
r
hi
g
he
s
t
f
a
c
t
or
w
he
n
c
om
pa
r
e
d
t
o o
t
he
r
c
l
us
t
e
r
s
i
s
e
a
s
y
tr
a
n
s
p
o
r
ta
tio
n
to
th
e
ir
u
n
iv
e
r
s
itie
s
.
M
e
mb
e
r
s
o
f
t
h
is
c
lu
s
te
r
h
a
v
e
th
e
h
ig
h
e
s
t a
c
tiv
ity
in
p
a
r
tic
ip
a
ti
n
g
in
cam
p
u
s
act
i
v
i
t
i
es
an
d
s
o
ci
al
m
ed
i
a.
−
C
l
us
t
e
r
03:
T
he
m
e
m
be
r
s
of
t
hi
s
c
l
us
t
e
r
a
r
e
dom
i
na
t
e
d by s
t
ude
nt
s
f
r
om
J
a
va
,
I
ndone
s
i
a
w
i
t
h
t
he
l
a
r
ge
s
t
num
be
r
of
c
l
us
t
e
r
m
e
m
be
r
s
.
T
he
a
ge
of
s
t
ude
nt
s
w
ho
a
r
e
m
e
m
be
r
s
of
t
hi
s
c
l
us
t
e
r
i
s
m
or
e
va
r
i
e
d
t
ha
n
o
t
h
er
cl
u
s
t
er
s
,
n
am
el
y
b
et
w
een
1
6
-
35 ye
a
r
s
.
C
om
p
a
r
e
d t
o a
l
l
c
l
us
t
e
r
s
,
s
t
ude
nt
s
w
ho w
e
r
e
m
e
m
be
r
s
o
f
t
hi
s
c
lu
s
te
r
s
c
o
r
e
d
th
e
lo
w
e
s
t in
te
r
ms
o
f
th
e
f
a
c
to
r
s
t
h
a
t mo
tiv
a
t
e
d t
he
m
t
o
c
hoos
e
t
he
i
r
uni
ve
r
s
i
t
y t
o
s
t
udy.
T
he
f
a
c
t
or
w
i
t
h t
he
l
ow
e
s
t
s
c
or
e
i
s
t
he
i
nf
l
ue
nc
e
of
a
dve
r
t
i
s
i
ng a
nd p
r
om
ot
i
on
.
T
he
i
nvol
ve
m
e
nt
of
s
t
u
d
en
t
s
w
h
o
ar
e m
em
b
er
s
o
f
t
h
i
s
cl
u
s
t
er
i
n
cam
p
u
s
act
i
v
i
t
i
es
an
d
s
o
ci
al
m
ed
i
a h
as
t
h
e l
o
w
es
t
s
co
r
e.
−
C
l
us
t
e
r
04:
M
e
m
be
r
s
of
t
hi
s
c
l
us
t
e
r
a
r
e
dom
i
na
t
e
d by s
t
ude
nt
s
f
r
om
t
he
S
um
a
t
r
a
r
e
gi
on
,
I
ndone
s
i
a
.
T
he
c
h
a
r
a
c
te
r
is
tic
s
o
f
th
is
c
lu
s
te
r
a
r
e
s
imila
r
to
C
lu
s
te
r
0
0
.
T
h
e
a
g
e
o
f
th
e
s
tu
d
e
n
ts
in
th
is
c
lu
s
te
r
is
16
-
25 ye
a
r
s
.
W
he
n vi
e
w
e
d f
r
om
t
he
r
e
a
s
ons
f
or
c
hoos
i
ng t
he
i
r
uni
ve
r
s
i
t
y,
t
he
qua
l
i
t
y f
a
c
t
or
o
f
e
duc
a
t
i
on
ge
t
s
a
hi
ghe
r
s
c
or
e
w
he
n c
om
pa
r
e
d
t
o t
he
s
c
or
e
g
i
ve
n by m
e
m
be
r
s
of
C
l
us
t
e
r
00.
T
he
e
c
onom
i
c
f
a
c
t
or
s
r
e
l
a
t
e
d t
o t
ui
t
i
on
f
e
e
s
a
nd t
he
e
f
f
e
c
t
of
pr
om
o
t
i
on
ha
ve
l
ow
e
r
s
c
or
e
s
w
he
n c
om
pa
r
e
d t
o t
he
s
c
or
e
s
gi
ve
n
by m
e
m
be
r
s
of
C
l
us
t
e
r
00.
B
a
s
e
d on t
he
r
e
s
ul
t
s
of
t
he
s
e
gm
e
nt
a
t
i
on,
i
t
a
ppe
a
r
s
t
ha
t
e
a
c
h c
r
i
t
e
r
i
on
c
a
n
be
l
i
nke
d t
o
one
a
not
he
r
t
o de
s
c
r
i
be
t
he
s
e
gm
e
nt
p
r
of
i
l
e
.
F
or
e
xa
m
pl
e
,
C
lu
s
t
e
r
00 a
nd
C
lu
s
te
r
04 a
r
e
c
l
us
t
e
r
s
t
ha
t
a
r
e
dom
i
na
t
e
d by
s
t
ude
nt
s
f
r
om
out
s
i
de
J
a
va
a
nd
ha
ve
s
i
m
i
l
a
r
pr
of
i
l
e
s
.
T
he
y
a
r
e
s
t
ude
nt
s
f
r
om
S
um
a
t
r
a
a
nd K
a
l
i
m
a
nt
a
n.
S
t
ude
nt
s
f
r
om
S
um
a
t
r
a
a
r
e
m
o
r
e
c
onc
e
r
ne
d w
i
t
h t
he
qua
l
i
t
y of
e
duc
a
t
i
on i
n c
hoos
i
ng hi
ghe
r
e
duc
a
t
i
on t
ha
n
s
tu
d
e
n
ts
f
r
o
m K
a
lima
n
ta
n
.
M
e
a
n
w
h
ile
,
th
e
e
c
o
n
o
mic
f
a
c
to
r
is
a
h
ig
h
e
r
r
e
a
s
on f
o
r
s
t
ude
nt
s
f
r
om
K
a
l
i
m
a
nt
a
n
t
o c
hoos
e
hi
ghe
r
e
duc
a
t
i
on
w
he
n c
om
pa
r
e
d t
o
s
t
ude
nt
s
f
r
om
S
um
a
t
r
a
.
F
or
be
ha
vi
or
c
r
i
t
e
r
i
a
,
s
t
ude
nt
s
f
r
om
K
a
lima
n
ta
n
s
tu
d
e
n
ts
a
r
e
mo
r
e
in
v
o
lv
e
d
in
c
a
mp
u
s
a
c
tiv
itie
s
th
a
n
s
tu
d
e
n
ts
f
r
o
m
S
u
ma
tr
a
.
F
or
C
lu
s
te
r
02 a
nd
C
lu
s
te
r
03
,
t
he
y a
r
e
s
t
ude
nt
s
f
r
om
J
a
va
.
T
he
s
i
gni
f
i
c
a
nt
di
f
f
e
r
e
nc
e
be
t
w
e
e
n t
he
t
w
o s
t
ude
nt
pr
of
i
l
e
s
i
n t
he
t
w
o c
l
us
t
e
r
s
i
s
t
ha
t
C
lu
s
te
r
02 ha
s
m
or
e
m
ot
i
va
t
i
on
t
o pu
r
s
ue
hi
ghe
r
e
duc
a
t
i
on.
I
n
a
d
d
itio
n
,
th
e
y
a
r
e
mo
r
e
o
f
te
n
in
v
o
lv
e
d
w
ith
c
a
mp
u
s
a
c
tiv
itie
s
.
M
e
a
n
w
h
ile
,
C
lu
s
te
r
03 ha
s
t
he
l
ow
e
s
t
m
ot
i
va
t
i
on i
n
pu
r
s
ui
ng hi
ghe
r
e
duc
a
t
i
on.
T
he
y
a
r
e
a
l
s
o
not
m
uc
h
i
nvol
ve
d
i
n c
a
m
pus
a
c
t
i
vi
t
i
e
s
.
S
t
ude
nt
s
f
r
om
P
a
pua
i
nc
l
ude
d
i
n
C
lu
s
te
r
01
.
M
ot
i
va
t
i
on
a
nd r
e
c
om
m
e
nda
t
i
ons
f
r
om
f
a
m
i
l
y
be
c
om
e
s
t
he
hi
ghe
s
t
r
e
a
s
on i
n c
ont
i
nui
ng
hi
ghe
r
e
duc
a
t
i
on.
S
t
ude
nt
s
f
r
om
P
a
pua
a
r
e
m
e
m
be
r
s
of
C
lu
s
te
r
01
.
T
he
m
ot
i
va
t
i
on a
nd
r
e
c
om
m
e
nda
t
i
on f
r
om
t
he
i
r
f
a
m
i
l
i
e
s
a
r
e
t
he
hi
ghe
s
t
r
e
a
s
ons
i
n c
ont
i
nui
ng h
i
ghe
r
e
duc
a
t
i
on.
4.
CO
NCL
US
I
O
N
H
ybr
i
d c
l
us
t
e
r
i
ng us
i
ng K
-
m
ean
s
a
nd S
O
M
a
l
gor
i
t
hm
ba
s
e
d on m
ul
t
i
-
c
r
ite
r
i s
e
gm
e
nt
a
t
i
on pe
r
f
or
m
s
a
n opt
i
m
a
l
c
l
us
t
e
r
i
ng r
e
s
ul
t
.
I
t
i
s
s
how
n by a
s
m
a
l
l
va
l
ue
of
D
B
I
a
nd Q
E
.
W
he
n us
i
ng hybr
i
d c
l
us
t
e
r
i
ng,
t
he
D
B
I
va
l
ue
i
s
1
,
386 a
nd
t
he
Q
E
va
l
ue
i
s
2,
752
.
I
n
a
ddi
t
i
on,
t
he
pe
r
f
or
m
a
nc
e
o
f
hyb
r
i
d c
l
us
t
e
r
i
ng us
i
ng
K
-
m
ean
s
a
nd S
O
M
a
l
gor
i
t
hm
ha
s
e
xc
e
e
de
d t
he
pe
r
f
or
m
a
nc
e
us
i
ng a
s
i
ngl
e
c
l
us
t
e
r
i
ng m
e
t
hod.
T
he
S
O
M
a
l
gor
i
t
hm
i
s
us
e
d t
o obt
a
i
n a
m
ode
l
t
ha
t
he
l
ps
vi
s
ua
l
i
z
e
t
he
f
e
a
t
ur
e
s
of
t
he
da
t
a
a
nd c
l
us
t
e
r
r
e
s
ul
t
s
.
B
a
s
e
d on
o
b
s
er
v
at
i
o
n
s
f
r
o
m
t
h
e f
eat
u
r
e
m
ap
,
u
n
n
eces
s
ar
y
f
eat
u
r
es
can
b
e
r
e
m
ove
d s
o t
ha
t
t
he
qua
l
i
t
y of
t
he
c
l
us
t
e
r
i
m
pr
ove
s
.
T
he
S
O
M
m
ode
l
a
l
s
o he
l
ps
c
r
e
a
t
e
a
vi
s
ua
l
i
z
a
t
i
on of
t
he
c
l
us
t
e
r
l
a
be
l
s
p
r
oduc
e
d by
K
-
m
ean
s
c
lu
s
te
r
in
g
.
S
e
gm
e
nt
a
t
i
on i
n
m
a
r
ke
t
i
ng
t
he
hi
ghe
r
e
duc
a
t
i
o
n s
t
ude
nt
s
s
uc
c
e
s
s
f
ul
l
y done
us
i
ng
t
he
hybr
i
d
c
lu
s
te
r
in
g
m
e
t
hod ba
s
e
d on m
ul
t
i
-
c
r
i
t
e
r
i
a
s
e
g
m
e
nt
a
t
i
on.
T
he
pr
opos
e
d m
e
t
hod pr
oduc
e
s
f
i
v
e
s
t
ude
nt
s
e
gm
e
nt
s
.
T
he
s
e
gm
e
nt
a
t
i
on
c
r
i
t
e
r
i
a
ba
s
e
d on
ge
ogr
a
phy,
de
m
og
r
a
phi
c
s
,
ps
yc
hogr
a
phi
c
s
,
a
nd
be
ha
vi
or
pr
ovi
de
i
nt
e
r
e
s
t
i
ng i
n
f
or
m
a
t
i
on
,
e
s
pe
c
i
a
l
l
y t
he
r
e
l
a
t
i
ons
hi
p be
t
w
e
e
n
t
h
e cr
i
t
er
i
a
u
s
ed
.
T
h
e ch
ar
act
er
i
s
t
i
cs
o
f
e
a
c
h c
l
us
t
e
r
c
a
n be
di
s
t
i
ngui
s
he
d a
c
c
or
di
ng t
o t
he
c
r
i
t
e
r
i
a
,
s
o
t
ha
t
t
he
i
nf
or
m
a
t
i
on
obt
a
i
ne
d i
s
r
i
c
he
r
.
F
ut
ur
e
r
e
s
e
a
r
c
h t
ha
t
ne
e
ds
t
o be
done
i
s
t
o e
xpl
or
e
ot
he
r
c
l
us
t
e
r
i
ng m
e
t
hods
t
ha
t
c
a
n be
i
nt
e
gr
a
t
e
d.
A
c
om
pr
e
he
ns
i
ve
e
va
l
ua
t
i
on a
l
s
o ne
e
ds
t
o
be
i
nc
l
ude
d t
o
c
om
pa
r
e
s
om
e
s
t
a
nda
r
d
t
e
c
hni
que
s
or
p
ar
am
et
er
s
w
i
t
h
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
KA
T
el
eco
m
m
u
n
C
om
put
E
l
C
ont
r
o
l
H
y
br
i
d c
l
us
t
e
r
i
ng bas
e
d on m
ul
t
i
-
c
r
i
t
e
r
i
a s
e
gm
e
nt
at
i
on f
or
…
(
H
ar
di
k
a K
hus
nul
i
aw
at
i
)
1505
t
he
e
xi
s
t
i
ng w
or
k
.
I
n
a
ddi
t
i
on,
i
t
i
s
ne
c
e
s
s
a
r
y t
o
i
m
pr
ove
t
he
qua
l
i
t
y
of
t
he
da
t
a
s
our
c
e
s
by
i
m
pr
o
vi
ng t
he
que
s
t
i
onna
i
r
e
i
t
e
m
s
a
nd e
nha
nc
i
ng t
he
f
e
a
t
ur
e
s
o
f
ge
ogr
a
phi
c
a
nd de
m
ogr
a
phi
c
c
r
i
t
e
r
i
a
.
A
C
K
N
O
W
LED
G
EM
EN
TS
T
hi
s
s
t
udy w
a
s
s
uppor
t
e
d
by
M
i
ni
s
t
r
y of
E
du
c
a
t
i
on a
nd C
ul
t
ur
e
R
e
publ
i
c
of
I
ndone
s
i
a
a
nd
U
n
i
v
er
s
i
t
as
S
ah
i
d
S
u
r
ak
ar
t
a.
R
EF
ER
EN
C
ES
[1
]
M
.
Da
va
r
i,
P
.
No
ur
s
a
le
hi
a
n
d A.
Ke
r
a
,
"
Da
ta
m
i
ni
ng a
ppr
oa
c
h t
o
pr
of
e
s
si
ona
l
e
d
uc
a
t
io
n m
a
r
ke
t se
gm
e
n
ta
t
io
n
:
a
c
a
se
s
tu
dy,
"
J
ou
rn
al of Ma
rk
e
t
in
g
f
or Hi
ghe
r
Ed
uc
a
ti
on,
vo
l.
29,
no.
2
,
pp.
4
5
-
6
6,
20
19
,
doi
: 1
0.
10
80
/0
88
41
24
1.
2
01
8.
15
45
72
4
.
[2
]
R
.
C
a
sid
y a
nd
W.
W
ym
e
r
,
"
A
ta
x
on
om
y of
pr
e
s
ti
ge
-
se
e
k
in
g un
ive
r
si
ty
st
ude
nt
s:
str
a
te
g
ic
i
nsi
gh
ts f
or
h
ig
he
r
e
duc
a
ti
on,
"
J
ou
rn
al of st
ra
te
g
ic
m
ar
k
e
t
in
g,
vol.
2
6,
no.
2,
pp.
1
40
-
1
55
,
201
8
,
doi
: 1
0.
10
80
/0
96
52
54
X.
2
01
6.
11
82
57
3
.
[3
]
J.
L
iu,
X.
L
ia
o,
W.
Hua
ng a
nd X.
L
ia
o,
"
M
a
r
ke
t se
gm
e
nta
ti
on
: A m
ult
ip
le
c
r
ite
r
ia
a
p
pr
oa
c
h c
om
b
in
in
g pr
e
f
e
r
e
nc
e
a
na
l
ys
is a
n
d se
gm
e
n
ta
t
io
n de
c
i
si
on,
"
Om
e
ga
,
vol.
8
3,
p
p.
1
-
13
,
20
19
,
d
oi
: 10.
10
16
/j.
om
e
ga
.
2
01
8.
01.
00
8
.
[4
]
S
.
Doln
ic
a
r
,
B
.
G
r
ün a
nd F
.
L
e
isc
h,
"
M
a
r
ke
t
se
gm
e
n
ta
t
io
n a
na
l
ys
is
: Un
de
r
s
ta
nd
in
g it,
d
oi
ng i
t,
a
nd m
a
ki
n
g it
use
f
ul,
"
Ma
rk
e
t Se
gm
e
n
ta
ti
on A
na
ly
s
is
,
p.
3
24,
2
01
8
,
do
i: 1
0.
10
07
/9
78
-
98
1
-
10
-
88
18
-
6
_2
.
[5
]
D.
Dok
ić
,
M
.
G
a
vr
a
n,
M
.
G
r
e
gić
a
nd
V.
G
a
ntne
r
,
"
The
im
pa
c
t
of
tr
a
de
ba
la
nc
e
of
a
gr
i
-
f
o
od pr
od
uc
t
s on the
sta
te
’
s
a
bi
li
ty t
o w
it
hs
ta
n
d the
c
r
is
is,
"
Hi
ghT
e
c
h a
nd I
n
nov
at
io
n J
ou
rn
al,
v
ol.
1,
n
o.
3,
pp.
1
07
-
11
1,
20
20
,
doi
: 1
0.
28
99
1/
HI
J
-
20
20
-
01
-
03
-
02
.
[6
]
X.
We
i,
F
.
M
e
n
g a
n
d P
.
Z
ha
ng
,
"
C
hi
ne
se
c
it
iz
e
n
s'
o
ut
b
oun
d de
st
ina
ti
on c
ho
ic
e
:
Ob
je
c
t
ive
a
nd
su
bje
c
ti
ve
f
a
c
t
o
r
s,
"
I
nte
rn
at
io
na
l J
ou
rn
al
of T
o
ur
is
m Re
se
a
rc
h ,
vo
l.
19,
n
o.
1,
pp.
38
-
49,
2
01
7
,
do
i: 10.
1
00
2/
jtr
.
20
82
.
[7
]
J.
B
r
uwe
r
,
B
.
R
oe
di
ge
r
a
nd F
.
He
r
b,
"
Dom
a
in
-
spe
c
i
f
ic
m
a
r
ke
t se
gm
e
n
ta
t
io
n: a
wi
ne
-
r
e
la
te
d lif
e
st
yle
(
W
R
L
)
a
ppr
oa
c
h,
"
Asi
a pac
if
ic
jo
ur
na
l of ma
rk
e
t
in
g an
d lo
gi
st
i
c
s,
vol.
29,
no.
1,
pp.
4
-
2
6,
201
7
,
doi
: 10.
1
10
8/
AP
JM
L
-
10
-
201
5
-
01
61
.
[8
]
R
.
M
a
kgosa
,
T.
M
a
te
n
ge
a
nd
P
.
M
bur
u,
"
Hybr
id se
gm
e
nta
ti
on in
t
he
f
ina
nc
ia
l
se
r
v
ic
e
s
m
a
r
ke
t:
ta
r
ge
t
in
g
sa
v
i
ng
c
on
sum
e
r
s,
"
Fam
ily
an
d C
o
ns
ume
r Sc
ie
nc
e
s
Re
s
e
arc
h J
o
ur
na
l,
vo
l.
4
4,
no.
4,
pp.
4
47
-
46
8,
2
016
,
doi
: 1
0.
11
11
/f
c
sr
.
12
16
8
.
[9
]
K.
G
oodr
ic
h
,
K.
S
wa
n
i a
n
d J.
M
unc
h,
"
Ho
w t
o c
o
nne
c
t
w
it
h
yo
ur
be
st
st
ude
nt
pr
o
spe
c
ts
: S
a
y
in
g t
he
r
i
gh
t t
hi
ng
s
,
to
the
r
ig
ht s
tu
de
n
ts,
i
n the
r
i
gh
t m
e
dia
,
"
J
our
na
l of M
ark
e
ti
ng C
om
mu
ni
c
a
tio
ns,
v
ol.
26,
n
o.
4,
p
p
. 1
-
20,
2
018
,
doi
: 1
0.
10
80
/1
35
27
26
6.
2
01
8.
15
14
31
9
.
[
10]
M
.
B
.
Ha
r
on,
B
.
Abdu
lla
h,
M
.
A.
J.
B
.
Ka
ha
r
a
nd
G
.
M
.
A.
Na
ji,
"
The
C
ha
lle
nge
r
o
f
M
a
r
ke
t S
e
gm
e
n
ts
fo
r
P
ri
v
a
t
e
Hig
he
r
e
duc
a
ti
on I
n M
a
la
y
sia
: Us
in
g B
e
ha
v
ior
a
l S
e
gm
e
n
ta
t
io
n,
"
I
nte
rn
at
io
na
l J
ou
rn
al o
f Bu
si
ne
s
s an
d
Ma
na
ge
me
nt I
nv
e
nt
io
n,
vo
l.
6,
no.
1
2,
pp.
2
4
-
2
8,
20
17.
[
11]
M
.
Tle
i
s,
R
.
C
a
llie
r
is,
a
nd R
.
R
om
a
,
"
S
e
gm
e
n
ti
ng
the
o
r
ga
n
ic
f
oo
d m
a
r
ke
t i
n L
e
ba
n
on
: a
n a
pp
lic
a
ti
on of
k
-
m
ea
n
s
c
lu
ste
r
a
na
ly
si
s,
"
Bri
ti
sh Fo
od
J
o
ur
na
l,
vo
l.
11
9,
no.
7,
201
7
,
do
i: 10.
1
10
8/B
F
J
-
08
-
20
16
-
03
54
.
[
12]
K.
M
.
M
a
ne
r
o ,
R
.
R
im
ir
u a
nd C
.
O
tie
no,
"
C
us
tom
e
r
B
e
ha
v
io
ur
S
e
gm
e
n
ta
t
io
n a
m
o
ng M
o
bi
le
S
e
r
v
ic
e
P
r
ov
ide
r
s i
n
Ke
n
ya
u
si
ng K
-
M
e
a
n
s A
lg
or
i
thm
,
"
I
n
te
r
na
ti
on
al
J
o
u
rna
l
of C
om
pu
te
r
Sc
ie
nc
e
I
ss
ue
s (
I
J
C
SI
)
,
vo
l
.
15
,
no.
5
,
pp.
67
-
7
6,
20
18
,
d
oi
: 10.
52
81
/z
e
n
od
o.
14
67
66
3
.
[
13]
M
.
Ar
yun
i,
E.
D.
M
a
dya
tm
a
dja
a
n
d E.
M
ir
a
nda
,
"
C
us
tom
e
r
se
gm
e
nta
ti
on i
n XYZ
ba
n
k usi
ng K
-
m
e
a
n
s
a
nd K
-
m
e
do
id
s c
l
us
te
r
i
ng,
"
I
n
te
r
na
ti
on
al C
on
fe
re
nc
e
o
n
I
nf
or
mat
io
n
Ma
na
ge
me
nt
an
d T
e
c
h
no
lo
gy
(
I
C
I
MT
e
c
h)
,
201
8
,
pp.
4
12
-
4
16,
d
oi
: 10.
11
09
/I
C
I
M
Te
c
h.
20
18.
85
28
0
86.
[
14]
E
.
B
.
B
a
yla
n,
"
A No
ve
l P
r
o
je
c
t R
i
sk
As
se
s
sm
e
n
t M
e
tho
d De
ve
l
opm
e
nt v
ia
A
HP
-
T
OP
S
I
S
Hy
br
i
d Al
gor
it
h
m
,
"
Eme
rg
in
g Sc
ie
nc
e
J
o
ur
na
l,
vo
l.
4,
no.
5,
p
p.
39
0
-
4
10,
202
0
,
do
i: 10.
2
89
91
/e
sj
-
20
20
-
0
12
39
.
[
15]
A.
I
qba
l,
M
.
Am
ir
,
V.
Kum
a
r
,
A.
Ala
m
a
nd M
.
Um
a
ir
,
"
I
nte
gr
a
t
io
n of
Ne
x
t G
e
ne
r
a
ti
on I
I
o
T wi
th B
l
oc
kc
ha
i
n f
or
the
De
ve
lo
pm
e
n
t of
S
m
a
r
t I
n
du
str
ie
s,
"
E
me
r
gi
ng
Sc
ie
nc
e
J
ou
rn
al ,
v
ol.
4,
pp.
1
-
1
7,
20
20
,
do
i: 10.
2
89
91
/ES
J
-
202
0
-
S
P1
-
01.
[
16]
P
.
F
.
B
or
ow
sk
i,
"
Ne
w te
c
hn
ol
og
ie
s
a
n
d i
nn
ova
ti
ve
so
lu
tio
ns
i
n t
he
de
ve
lo
pm
e
n
t
str
a
te
g
ie
s
of
e
ne
r
g
y e
n
te
r
pr
ise
s,
"
Hig
hT
e
c
h an
d I
n
nov
at
io
n J
ou
rn
al,
v
ol.
1,
n
o.
2,
pp.
3
9
-
58,
20
20
,
d
oi
: 1
0.
28
99
1/
HI
J
-
20
20
-
01
-
02
-
01
.
[
17]
S
.
M
.
S
ha
r
if
,
F
.
M
.
K
us
in,
Z
.
H.
A
sha
’
a
r
i a
nd
A.
Z
.
Ar
i
s
,
"
C
ha
r
a
c
te
r
iz
a
ti
on
of
wa
te
r
qua
li
ty
c
o
nd
it
io
ns
in
the
K
l
a
ng
R
ive
r
B
a
si
n,
M
a
la
ys
ia
u
s
in
g se
lf
or
ga
niz
in
g m
a
p a
nd
K
-
m
e
a
ns a
lg
or
i
thm
,
"
Proc
e
di
a Env
ir
on
me
n
ta
l Sc
ie
nc
e
s,
vol.
3
0,
pp.
7
3
-
78,
20
15
,
do
i: 1
0.
1
01
6/
j.
pr
oe
nv.
20
15.
1
0.
013
.
[
18]
W
.
Qa
da
de
h a
nd S
.
Abda
l
la
h,
"
C
ust
om
e
r
s S
e
gm
e
nta
t
ion i
n the
I
ns
ur
a
nc
e
C
om
pa
n
y (
TI
C
)
Da
ta
se
t,
"
Proc
e
di
a
c
om
pu
te
r sc
ie
nc
e
,
v
ol.
1
44,
pp
.
2
77
-
29
0
,
20
18
,
do
i: 10.
1
016
/j.
pr
oc
s.
2
01
8.
10.
52
9
.
[
19]
P
.
D.
Hung,
N.
D.
Ngoc
a
nd T.
D.
Ha
nh,
"
K
-
m
e
a
ns C
lus
te
r
in
g Usi
ng R
A C
a
se
S
tu
dy of
M
a
r
ke
t S
e
gm
e
nta
ti
o
n,
"
Pro
c
e
e
d
in
gs o
f t
he
2
01
9 5
th I
nte
rn
at
io
na
l C
o
nfe
re
n
c
e
on E
-
Bu
si
ne
s
s
a
nd Ap
pl
ic
a
ti
on
s,
20
19
,
pp.
1
00
-
1
04,
doi
: 1
0.
11
45
/3
31
76
14.
33
17
62
6.
[
20]
R
.
P
a
nde
y,
"
Ta
r
ge
t M
a
r
ke
ti
ng
U
si
ng
Da
ta
M
i
ni
ng,
"
I
nte
rn
at
io
na
l
J
o
ur
na
l
of Adv
anc
e
s in
Ele
c
t
ro
ni
c
s
a
nd
C
omp
ut
e
r Sc
ie
nc
e
,
v
ol.
4,
n
o.
6,
pp.
1
6
-
19
,
20
17.
[
21]
J.
Tia
n a
nd M
.
G
u,
"
S
ub
spa
c
e
c
l
us
te
r
i
ng
ba
se
d o
n
se
lf
-
or
ga
niz
in
g m
a
p,
"
i
n
Pr
o
c
e
e
d
in
g of
th
e
24
th
I
nt
e
rn
at
ion
al
C
onf
e
re
nc
e
o
n I
n
du
st
ri
al
En
gi
ne
e
r
in
g a
nd
En
gi
ne
e
r
in
g
Ma
na
ge
me
nt
,
S
in
ga
po
r
e
,
2
01
8
,
do
i: 10.
1
00
7/
97
8
-
98
1
-
13
-
340
2
-
3_
17
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SSN
:
1693
-
6930
T
E
L
KOM
NI
KA
T
el
eco
m
m
u
n
C
om
put
E
l
C
ont
r
o
l
,
Vo
l
.
19
, N
o
.
5
,
O
c
t
obe
r
2021:
1498
-
1506
1506
[
22]
B
.
Z
hong,
J.
Wa
n
g,
H.
W
u,
J.
Z
ho
u a
n
d Q.
J
in,
"
S
O
M
-
ba
se
d v
isu
a
l
iz
a
t
io
n m
o
ni
tor
in
g a
n
d f
a
u
lt d
ia
gn
osi
s
f
or
c
he
m
ic
a
l pr
oc
e
ss,
"
201
6 C
hi
ne
s
e
C
ont
ro
l
and De
c
is
io
n C
onf
e
re
nc
e
(
C
C
DC
)
,
2016
,
doi
:
10.
1
10
9/C
C
DC
.
2
01
6.
75
32
04
3
.
[
23]
M
.
M
.
M
osta
f
a
,
"
S
ha
de
s of
gr
e
e
n
: A p
sy
c
ho
gr
a
p
hic
se
gm
e
n
ta
t
io
n of
the
gr
e
e
n c
o
ns
um
e
r
in K
uwa
it u
si
n
g
se
lf
-
or
ga
niz
in
g m
a
p
s,
"
Ex
pe
rt Sy
s
te
m
s
w
i
th Ap
p
lic
at
io
ns,
v
ol.
36,
n
o.
8,
pp.
1
10
30
-
1
10
38
,
200
9
,
doi
: 1
0.
10
16
/j.
e
swa
.
20
09.
0
2.
0
88.
[
24]
I
.
S
a
r
i,
A.
M
a
se
le
no,
F
.
S
a
tr
ia
a
nd M
.
M
uslih
ud
in,
"
App
lic
a
ti
on
m
o
de
l of
k
-
m
e
a
ns c
lu
st
e
r
i
ng
: I
nsi
gh
ts i
nt
o
pr
om
ot
io
n str
a
te
g
y of
v
oc
a
t
io
na
l
hi
gh
sc
h
oo
l,
"
I
nte
r
nat
io
na
l J
o
ur
na
l o
f En
gi
ne
e
r
in
g &
T
e
c
hn
ol
ogy
,
vo
l.
7,
no.
2.
2
7,
pp.
1
82
-
18
7,
20
18
,
d
oi
: 1
0.
14
41
9/
ije
t.
v
7i
2.
11
4
91
.
[
25]
F
.
Te
m
po
la
a
n
d A.
F
.
A
ssa
ga
f
,
"
C
lu
ste
r
in
g of
P
o
te
n
c
y
of
S
hr
im
p in
I
nd
one
sia
wi
th k
-
m
e
a
n
s
a
l
gor
it
hm
a
nd
va
l
ida
ti
on
of
da
v
ie
s
-
bo
ul
di
n
in
de
x,
"
i
n
I
nte
rn
at
io
na
l C
onf
e
re
nc
e
on
Sc
ie
nc
e
an
d
T
e
c
h
no
lo
gy
(
I
C
ST
2
01
8)
,
20
18
,
doi
: 1
0.
29
91
/i
c
s
t
-
1
8.
20
18.
14
8
.
[
26]
C
.
Or
done
z
a
n
d E.
Om
ie
c
in
sk
i,
"
Ef
f
ic
ie
nt d
is
k
-
ba
s
e
d K
-
m
e
a
n
s c
lu
ste
r
in
g f
or
r
e
la
t
io
na
l da
ta
ba
se
s,
"
I
EE
E
T
ran
sac
ti
on
s o
n K
now
le
d
ge
a
nd D
at
a E
ng
ine
e
ri
ng,
v
ol.
16,
no.
8,
p
p.
90
9
-
92
1,
20
04
,
d
oi
: 10.
11
09
/T
KD
E.
20
04.
25
.
[
27]
M
.
C
ha
tto
pa
dh
ya
y a
n
d P
.
C
ha
r
a
n,
"
S
e
lf
-
Or
ga
n
iz
a
t
io
n
M
a
p
Vi
sua
liz
a
ti
on
f
or
a
C
lu
ste
r
in
g
Ope
r
a
ti
on
S
e
que
nc
e
ba
se
d
C
e
ll F
or
m
a
t
io
n P
r
o
ble
m
,
"
C
on
fe
r
e
nc
e
:
PO
MS
27
th
An
nu
al C
on
fe
re
nc
e
o
f t
he
P
ro
duc
ti
on
an
d O
pe
r
at
ion
s
Ma
na
ge
me
nt S
oc
ie
ty
(
P
OM
S)
,
20
16.
[
28]
T.
Ne
m
o
to a
n
d D.
B
e
g
la
r
,
"
L
ike
r
t
-
sc
a
le
q
ue
s
ti
on
na
ir
e
s,
"
i
n
J
AL
T
2
01
3
C
on
fe
r
e
nc
e
P
roc
e
e
di
ng
s
,
20
14.
[
29]
J.
Tia
n,
M
.
H.
Az
a
r
ia
n a
n
d M
.
P
e
c
ht,
"
Anom
a
l
y de
te
c
ti
on us
in
g se
l
f
-
or
ga
n
iz
i
ng m
a
p
s
-
ba
se
d k
-
ne
a
r
e
s
t ne
i
g
hbo
r
a
lg
or
i
thm
,
"
in
P
roc
e
e
di
ng
s o
f the
Eu
ro
pe
a
n C
o
nfe
re
nc
e
of t
he
P
ro
gn
os
tic
s an
d He
al
th M
an
age
me
n
t S
oc
ie
ty
,
2
0
14
.
[
30]
A.
M
oha
m
e
d,
M
.
S
.
Ha
m
di a
nd S
.
Ta
ha
r
,
"
S
e
lf
-
or
ga
n
iz
i
ng m
a
p
-
ba
se
d f
e
a
t
ur
e
vi
sua
liz
a
ti
on a
n
d se
le
c
t
io
n f
or
de
f
e
c
t
de
p
th e
st
im
a
t
io
n in oi
l a
nd ga
s pip
e
l
ine
s,
"
in
I
nte
r
na
ti
ona
l
C
on
fe
re
nc
e
on I
n
fo
rm
at
io
n Vis
ua
li
sa
ti
on
1
9t
h
,
20
15
,
doi
: 1
0.
11
09
/i
V.
2
01
5.
50
.
[
31]
M.
S
ya
kur
,
B
.
Kh
ot
im
a
h,
E.
R
oc
hm
a
n a
nd B
.
D
.
S
a
to
to,
"
I
nte
gr
a
ti
on
k
-
m
e
a
ns c
lu
ste
r
in
g m
e
t
ho
d a
n
d e
lb
ow m
e
t
ho
d
f
or
ide
n
tif
ic
a
t
io
n of
the
be
s
t c
us
tom
e
r
pr
of
i
le
c
lus
te
r
,
"
I
OP C
on
fe
re
nc
e
Se
r
ie
s:
M
ate
ri
al
s Sc
ie
nc
e
a
nd E
ng
ine
e
r
ing,
vol.
3
36,
n
o.
1,
20
18
,
d
oi
: 1
0.
10
8
8/
17
57
-
89
9X
/3
36
/1/
01
201
7
.
[
32]
M
.
M
ilkha
tu
n,
A.
.
A.
F
.
R
iz
a
l,
N.
W
.
W.
Asth
in
in
gs
i
h,
a
nd A.
J.
L
a
ti,
"
P
e
r
f
or
m
a
nc
e
Asse
ssm
e
nt of
U
ni
ve
r
sit
y
L
e
c
tur
e
r
s
:
A Da
t
a
M
i
ni
ng Ap
pr
oa
c
h,
"
K
ha
za
na
h I
nf
or
m
ati
k
a:
J
ur
na
l I
l
mu
K
o
mp
ute
r d
an I
nf
or
ma
tik
a,
v
ol.
6,
n
o
.
2,
202
0
,
d
oi
: 10.
2
39
17
/k
hi
f
.
v6
i2.
90
69
.
B
I
OGR
A
P
HI
E
S
OF
A
U
T
HOR
S
H
ar
d
ik
a K
h
u
sn
u
l
iaw
at
i
r
e
c
e
i
ve
d t
he
S
.
Kom
.
De
gr
e
e
i
n inf
or
m
a
t
ic
s e
ng
ine
e
r
in
g f
r
om
I
n
st
it
u
t
Te
k
no
lo
gi S
e
p
ul
uh N
ope
m
be
r
,
S
ur
a
ba
ya
,
in 2
01
4,
the
M
.
Kom
.
de
gr
e
e
in i
nf
or
m
a
tic
s
e
ng
ine
e
r
in
g f
r
om
I
nst
it
ut Te
kn
ol
og
i S
e
pu
lu
h Nope
m
be
r
,
S
ur
a
ba
ya
,
in 20
16.
S
he
is c
ur
r
e
nt
ly a
le
c
t
ur
e
r
in the
F
a
c
u
lt
y of
S
a
in
s,
Te
c
h
no
lo
gy,
a
nd
He
a
l
thc
a
r
e
,
a
t Un
ive
r
si
ta
s,
S
ur
a
ka
r
ta
,
I
nd
one
sia
.
He
r
c
ur
r
e
nt r
e
se
a
r
c
h in
te
r
e
s
ts i
nc
l
ude
d
a
ta
m
ini
ng,
im
a
ge
pr
oc
e
ss
in
g,
a
nd s
of
t
wa
r
e
e
ng
ine
e
r
in
g.
Dh
ian
Ri
sk
i
an
a
P
u
t
r
i
r
e
c
e
ive
d th
e
S
.
P
si.
De
gr
e
e
i
n p
syc
ho
lo
gy U
ni
ve
r
s
ita
s S
e
be
la
s M
a
r
e
t
,
S
ur
a
ka
r
ta
,
i
n 2
01
0,
t
he
M
.
A.
de
gr
e
e
i
n p
syc
ho
lo
gy
f
r
om
Un
ive
r
si
ta
s
G
a
dja
h M
a
da
,
Yo
gya
ka
r
t
a
in 20
14.
S
he
is c
ur
r
e
n
tl
y a
le
c
tur
e
r
in t
he
F
a
c
ul
ty of
Ar
t,
S
oc
ia
l,
a
nd Hum
a
ni
tie
s a
t Uni
ve
r
s
ita
s,
S
ur
a
ka
r
ta
,
I
n
do
ne
s
ia
.
He
r
c
ur
r
e
n
t r
e
se
a
r
c
h
in
te
r
e
s
ts
inc
lu
de
E
du
c
a
t
io
na
l P
syc
ho
lo
gy,
P
syc
h
ol
og
y of
l
e
a
r
n
in
g,
B
e
ha
v
ior
m
o
dif
ic
a
t
io
n a
n
d ob
se
r
va
t
io
n.
Evaluation Warning : The document was created with Spire.PDF for Python.