I
n
d
on
e
s
ian
Jou
r
n
al
o
f
E
lec
t
r
ica
l
E
n
gin
e
e
r
in
g
a
n
d
Com
p
u
t
e
r
S
c
ience
Vo
l
.
25
,
N
o
.
1
,
J
a
n
ua
r
y
2022
,
pp.
569
~
579
I
S
S
N:
2502
-
4752,
DO
I
:
10
.
11591/i
j
e
e
c
s
.
v
25
.i
1
.
pp
569
-
579
569
Jou
r
n
al
h
o
m
e
page
:
ht
tp:
//
ij
e
e
c
s
.
iaes
c
or
e
.
c
om
M
ax
st
ab
le
se
t
p
r
ob
le
m
to
f
ou
n
d
t
h
e
in
itial
c
e
n
t
r
oi
d
s
i
n
c
lu
st
e
r
in
g
p
r
ob
le
m
Aw
at
if
K
ar
im
1
,
Chak
ir
L
oq
m
an
1
,
Yous
s
e
f
Ham
i
2
,
Jaou
ad
B
o
u
m
h
id
i
1
1
L
I
S
A
C
L
a
bor
a
to
r
y
,
F
a
c
ul
t
y
of
S
c
i
e
n
c
e
D
ha
r
El
M
e
h
r
a
z
,
U
ni
ve
r
s
it
y
S
id
i
M
o
ha
m
e
d
B
e
n
A
bd
e
ll
a
h,
F
e
z
,
M
o
r
oc
c
o
2
R
T
M
C
S
A
Re
s
e
a
r
c
h
T
e
a
m,
S
c
h
oo
l
N
a
ti
o
na
l
of
A
ppl
i
e
d
S
c
i
e
nc
e
s
,
U
ni
ve
r
s
it
y
A
bd
El
M
a
le
kE
s
s
a
a
di
,
T
a
ngi
e
r
,
M
o
r
oc
c
o
Ar
t
ic
l
e
I
n
f
o
AB
S
T
RA
CT
A
r
ti
c
le
h
is
tor
y
:
R
e
c
e
i
ve
d
F
e
b
9
,
2021
R
e
vi
s
e
d
No
v
5
,
2021
A
c
c
e
pt
e
d
N
o
v
26
,
2021
I
n
th
i
s
p
ap
e
r,
w
e
p
ro
p
o
s
e
a
n
e
w
ap
p
ro
ach
to
s
o
l
v
e
th
e
d
o
c
u
me
nt
-
cl
u
s
t
e
r
i
ng
us
i
ng
th
e
K
-
Me
an
s
al
g
o
r
i
th
m
.
Th
e
l
at
t
e
r
i
s
s
e
ns
i
t
i
v
e
to
th
e
ran
d
o
m
s
e
l
e
ct
i
on
of
th
e
k
c
l
u
s
t
e
r
c
e
n
t
ro
i
ds
i
n
th
e
i
n
i
t
i
al
i
zat
i
on
p
h
as
e
.
To
e
v
a
l
u
at
e
th
e
q
u
al
i
ty
of
K
-
Me
an
s
cl
u
s
t
e
r
i
ng
w
e
p
ro
p
o
s
e
to
m
od
e
l
th
e
t
e
xt
d
o
c
u
me
nt
cl
u
s
t
e
r
i
ng
p
ro
bl
em
as
th
e
m
ax
s
t
a
b
l
e
s
e
t
p
ro
b
l
em
(
M
SSP)
an
d
us
e
c
o
n
t
i
n
u
o
u
s
H
o
p
f
ie
ld
n
e
t
w
o
r
k
to
s
o
l
v
e
th
e
M
SSP
p
ro
b
l
em
to
h
a
v
e
i
n
i
t
i
al
c
e
n
t
ro
i
d
s
.
Th
e
i
d
e
a
i
s
i
n
s
p
i
r
e
d
by
th
e
fa
c
t
t
h
at
M
SSP
an
d
cl
u
s
t
e
r
i
ng
s
h
ar
e
th
e
sa
me
pr
i
nc
i
pl
e
,
M
SSP
co
n
s
i
sts
to
f
i
nd
th
e
l
ar
g
e
st
s
e
t
of
n
o
d
e
s
co
m
pl
e
t
e
ly
d
i
s
co
n
n
e
ct
e
d
i
n
a
g
rap
h
,
an
d
i
n
cl
u
s
t
e
r
i
n
g
,
al
l
o
b
j
e
c
t
s
ar
e
d
i
v
i
d
e
d
i
nto
d
i
sjo
i
nt
cl
u
s
t
e
rs
.
S
im
u
l
at
i
on
r
e
s
u
l
t
s
d
em
o
n
s
t
rat
e
t
h
at
th
e
p
ro
p
o
s
e
d
K
-
Me
an
s
im
p
ro
v
e
d
by
M
SSP
(K
M
_
M
SSP)
i
s
e
ff
i
c
ie
nt
of
l
ar
g
e
d
at
a
s
e
ts,
i
s
m
u
ch
o
p
t
i
m
i
z
e
d
i
n
t
e
r
m
s
of
t
i
me
,
a
n
d
p
ro
v
i
d
e
s
b
e
tt
e
r
q
u
al
i
ty
of
c
l
u
s
t
e
r
i
ng
t
h
an
o
t
h
e
r
me
t
h
o
d
s
.
K
e
y
w
o
r
d
s
:
C
o
n
t
i
n
o
us
h
o
p
f
i
e
l
d
n
e
t
wo
r
k
Do
c
um
e
n
t
c
l
us
t
e
r
i
n
g
I
ni
t
i
a
l
c
e
n
t
r
o
i
ds
M
a
xi
m
u
m
s
t
a
bl
e
s
e
t
pr
o
bl
e
m
Th
i
s
is
an
o
p
en
a
c
ces
s
a
r
t
i
c
l
e
u
n
d
e
r
the
CC
BY
-
SA
l
i
cen
s
e.
C
or
r
e
s
pon
din
g
A
u
th
or
:
A
wa
t
i
f
Ka
r
im
L
I
S
A
C
L
a
b
o
r
a
to
r
y
,
F
a
c
u
l
t
y
of
Sc
ie
nc
e
D
h
a
r
E
l
Me
h
r
a
z
,
Un
i
v
e
rs
i
ty
S
i
d
i
M
o
ha
me
d
B
e
n
Ab
d
e
ll
a
h
B
o
x
30003,
F
e
z
,
M
or
o
c
c
o
E
m
a
i
l
:
a
wa
t
i
f
.
ka
r
im
@us
m
ba
.
a
c
.
m
a
1.
I
NT
RODU
C
T
I
ON
T
o
da
y
,
c
l
us
t
e
r
i
n
g
is
an
i
m
po
r
t
a
n
t
too
l
in
da
t
a
m
i
n
i
ng
;
it
is
an
a
ut
o
m
a
t
i
c
l
e
a
r
ni
ng
pr
o
c
e
s
s
w
h
e
r
e
t
he
o
bj
e
c
t
s
a
r
e
c
l
us
t
e
r
e
d
or
gr
o
upe
d
b
a
s
e
d
on
t
h
e
pr
i
n
c
i
p
l
e
of
m
a
xi
mi
z
i
ng
t
h
e
i
n
t
r
a
c
l
a
s
s
s
im
i
l
a
r
i
t
y
a
nd
mi
n
im
i
z
i
ng
t
h
e
i
n
t
e
r
c
l
a
s
s
s
im
il
a
r
i
t
y
w
i
t
h
o
ut
kn
o
w
i
n
g
t
h
e
l
a
b
e
l
s
of
t
h
e
d
a
t
a
.
M
a
ny
c
l
us
t
e
r
i
n
g
s
t
r
a
teg
i
e
s
a
r
e
a
v
a
il
a
bl
e
in
t
h
e
l
i
t
e
r
a
t
ur
e
as
[
1]
,
[
2]
,
a
m
o
n
g
t
h
e
m
t
wo
m
a
i
n
c
a
t
e
go
r
i
e
s
k
n
o
wn
as
hi
e
r
a
r
c
hi
c
a
l
a
n
d
pa
r
t
i
t
i
o
ni
n
g
c
l
u
s
t
e
r
i
n
g.
A
t
r
e
e
s
t
r
uc
t
ur
e
c
r
e
a
t
e
d
in
t
h
e
pr
o
c
e
s
s
of
h
i
e
r
a
r
c
hi
c
a
l
c
l
us
t
e
r
i
n
g
a
n
d
s
h
o
ws
h
o
w
o
bjec
t
s
a
r
e
gr
o
upe
d
(
i
n
an
a
gg
l
o
m
e
r
a
t
i
v
e
m
e
t
h
o
d)
or
pa
r
t
i
t
i
o
ne
d
(
i
n
a
d
i
vi
s
i
ve
m
e
t
h
o
d)
.
Dum
o
n
d
[
3]
h
a
ve
a
r
gue
d
t
h
a
t
t
h
e
a
gg
l
o
m
e
r
a
t
i
v
e
hi
e
r
a
r
c
hi
c
a
l
c
l
us
t
e
r
i
n
g
a
l
go
r
i
t
hm
s
t
a
r
t
s
wi
t
h
each
o
bj
e
c
t
f
o
r
m
i
ng
a
s
e
pa
r
a
t
e
gr
o
up
.
I
n
pa
r
t
i
t
i
o
ni
ng
m
e
t
h
o
ds
,
a
da
t
a
s
e
t
is
de
c
o
m
po
s
e
d
i
n
t
o
k
pa
r
t
i
t
i
o
ns
.
A
c
l
u
s
t
e
r
is
a
s
e
t
of
e
l
e
m
e
n
t
s
r
e
pr
e
s
e
n
t
e
d
by
t
h
e
c
e
n
t
r
o
i
d
(
o
r
pr
otot
y
pe
)
of
t
h
e
c
l
us
t
e
r
.
T
hi
s
l
a
t
t
e
r
is
f
o
r
m
e
d
in
s
u
c
h
a
wa
y
t
h
a
t
it
is
c
l
o
s
e
l
y
r
e
l
a
t
e
d
(
i
n
t
e
r
m
s
of
s
im
il
a
r
i
t
y
f
u
nc
t
i
o
n
)
t
o
a
l
l
o
bjec
t
s
in
t
h
a
t
c
l
u
s
t
e
r
.
T
h
e
b
e
s
t
kn
o
wn
pa
r
t
i
t
i
o
ni
ng
m
e
t
h
o
ds
i
nc
l
ud
i
ng,
K
-
M
e
a
ns
a
n
d
i
t
s
va
r
i
a
n
t
s
[
4]
,
f
uz
z
y
c
-
m
e
a
n
s
,
po
s
s
i
b
il
i
s
t
i
c
c
-
m
e
a
n
s
,
h
a
r
d
c
-
m
e
a
ns
(
HC
M
)
a
n
d
m
e
a
n
s
hi
f
t
[
5]
,
K
-
m
e
do
i
d
s
[
6]
,
pa
r
t
i
t
i
o
ni
ng
a
r
o
un
d
m
e
do
i
ds
(
P
A
M
)
,
C
L
AR
A
[
7]
a
n
d
C
L
AR
A
NS
[
8]
.
M
o
s
t
of
t
h
e
s
e
a
l
go
r
i
t
hm
s
s
e
l
e
c
t
r
a
n
do
m
ly
t
h
e
i
n
i
t
i
a
l
c
e
n
t
e
r
s
in
a
d
v
a
nc
e
,
whi
c
h
c
o
n
s
i
de
r
a
bly
a
f
f
e
c
t
t
h
e
qua
l
i
t
y
of
c
l
u
s
t
e
r
i
n
g
r
e
s
u
l
t
s
.
In
t
h
e
s
a
m
e
c
o
n
t
e
x
t
,
K
-
M
e
a
n
s
is
an
u
n
s
upe
r
vi
s
e
d
l
e
a
r
ni
ng
m
e
t
h
o
d
[
9]
,
whi
c
h
is
w
i
d
e
l
y
us
e
d
in
t
e
x
t
c
l
us
t
e
r
i
n
g.
K
-
M
e
a
n
s
is
an
a
l
go
r
i
t
hm
w
h
o
s
e
c
l
u
s
t
e
r
n
u
m
b
e
r
is
g
i
ve
n
at
t
h
e
s
t
a
r
t
a
n
d
we
ll
c
o
n
s
t
r
uc
t
e
d
as
we
l
l
.
B
ut
,
t
h
e
di
f
f
i
c
u
l
t
y
is
t
h
a
t
t
h
e
r
e
s
u
l
t
s
of
c
l
us
t
e
r
i
n
g
de
pe
n
ds
on
t
h
e
f
i
xe
d
n
u
m
be
r
of
c
l
a
s
s
e
s
a
n
d
on
t
h
e
r
a
n
do
m
c
h
o
i
c
e
of
i
n
i
t
i
a
l
c
l
u
s
t
e
r
s
,
e
s
pe
c
i
a
ll
y
w
he
n
t
h
e
da
t
a
s
et
is
l
a
r
ge
a
n
d
we
don'
t
h
a
v
e
a
s
s
u
m
pt
i
o
n
s
a
b
o
ut
t
h
e
da
t
a
[
10]
.
H
o
we
v
e
r
,
it
c
a
n
be
s
t
uc
k
in
a
l
o
c
a
l
mi
n
im
u
m
a
n
d
c
a
u
s
e
an
u
n
s
t
a
bl
e
r
e
s
u
l
t
(
if
we
r
e
i
ni
t
i
a
li
z
e
t
h
e
a
l
go
r
i
t
hm
w
i
t
h
o
t
h
e
r
v
a
l
ue
s
,
it
m
a
y
c
o
nv
e
r
ge
to
a
n
ot
h
e
r
l
o
c
a
l
s
o
l
ut
i
o
n
)
[
11
]
-
[
14
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2502
-
4752
I
n
do
n
e
s
i
a
n
J
E
l
e
c
E
n
g
&
C
o
m
p
S
c
i
,
Vo
l
.
25
,
N
o
.
1
,
J
a
n
ua
r
y
20
22
:
569
-
579
570
To
o
v
e
r
c
o
m
e
t
he
a
b
o
v
e
de
f
i
c
i
e
n
c
i
e
s
,
m
o
s
t
r
e
s
e
a
r
c
h
in
c
l
us
t
e
r
i
n
g
a
n
a
ly
s
i
s
h
a
s
b
e
e
n
f
o
c
us
e
d
on
t
h
e
“
a
ut
o
m
a
t
i
c
c
l
us
t
e
r
i
n
g
a
l
go
r
i
t
hm
”
.
T
h
e
E
l
b
o
w
M
e
th
o
d
is
o
n
e
of
t
h
e
m
o
s
t
p
o
pul
a
r
m
e
t
h
o
ds
to
de
t
e
r
mi
ne
t
hi
s
o
p
t
i
m
a
l
v
a
l
ue
of
k.
It
c
o
n
s
i
s
t
s
of
r
unni
ng
K
-
M
e
a
n
s
c
l
u
s
t
e
r
i
n
g
of
t
h
e
da
t
a
s
e
t
wi
t
h
a
r
a
n
ge
of
v
a
l
u
e
s
of
k
,
c
a
l
c
u
l
a
t
i
n
g
t
h
e
s
u
m
of
s
qua
r
e
d
e
r
r
o
r
s
f
o
r
each
k
,
a
n
d
p
l
o
t
t
i
n
g
t
h
e
m
in
a
l
i
ne
c
h
a
r
t
.
If
t
h
e
c
h
a
r
t
l
o
o
k
s
l
i
ke
an
a
r
m
,
t
h
e
b
e
s
t
v
a
l
u
e
of
k
w
i
ll
be
on
t
h
e
"
e
l
b
o
w"
[
15]
.
W
i
t
h
t
h
e
h
e
l
p
of
m
e
tah
e
u
r
i
s
t
i
c
s
a
l
g
or
i
t
hm
s
,
it
wa
s
i
n
v
e
s
t
i
ga
t
e
d
as
a
n
e
w
m
e
t
h
od
in
c
l
us
t
e
r
a
n
a
l
y
s
i
s
[
1
6
]
.
O
n
e
of
t
h
e
m
a
j
o
r
m
e
t
h
o
ds
f
o
r
t
hi
s
pr
o
bl
e
m
is
a
c
o
m
bi
na
t
i
o
n
of
s
wa
r
m
i
n
t
e
l
li
g
e
n
c
e
a
l
go
r
i
t
hm
s
w
i
t
h
c
l
u
s
t
e
r
v
a
l
i
d
i
t
y
i
nd
i
c
e
s
as
an
o
bj
e
c
t
i
v
e
f
u
nc
t
i
o
n
[
17
]
,
[
18
]
.
T
h
e
r
e
a
r
e
s
o
m
e
c
r
i
t
e
r
i
a
,
de
r
i
v
e
d
f
r
o
m
d
i
f
f
e
r
e
n
t
a
ppr
o
a
c
h
e
s
,
f
o
r
de
t
e
r
m
i
ni
ng
t
h
e
o
p
t
i
m
a
l
n
u
m
be
r
of
c
l
a
s
s
e
s
.
We
c
i
t
e
t
h
e
c
r
i
t
e
r
i
o
n
of
X
i
e
a
n
d
B
e
ni
[
19]
,
whi
c
h
is
b
a
s
e
d
on
a
m
e
a
s
ur
e
of
s
e
pa
r
a
bil
i
t
y
a
n
d
t
h
e
c
o
m
pa
c
t
n
e
s
s
of
c
l
a
s
s
e
s
.
T
h
e
s
e
t
wo
n
ot
i
o
ns
de
f
i
ne
c
r
i
t
e
r
i
a
f
o
r
e
v
a
l
ua
t
i
n
g
a
c
l
a
s
s
i
f
i
c
a
t
i
o
n
.
X
i
e
a
n
d
B
e
ni
pr
o
po
s
e
to
c
h
o
o
s
e
t
h
e
o
p
t
i
m
a
l
k
w
hi
c
h
m
i
n
im
i
z
e
s
t
h
e
r
e
l
a
t
i
o
n
s
hi
p
b
e
t
we
e
n
s
e
pa
r
a
bil
i
t
y
a
n
d
c
o
m
pa
c
t
n
e
s
s
.
In
[
20]
,
r
e
s
e
a
r
c
h
e
r
s
h
a
v
e
pr
e
s
e
n
t
e
d
a
m
e
t
h
o
d
f
u
z
z
y
s
il
h
o
ue
t
t
e
i
n
de
x
on
d
y
n
a
mi
c
da
t
a
to
f
i
nd
t
h
e
o
p
t
i
m
a
l
n
u
m
be
r
of
c
l
us
t
e
r
s
.
T
h
e
o
p
t
i
m
a
l
n
u
m
be
r
of
c
l
u
s
t
e
r
s
k
is
t
h
e
o
n
e
t
h
a
t
m
a
xim
i
z
e
s
t
h
e
a
v
e
r
a
ge
s
il
h
o
ue
t
t
e
o
v
e
r
a
r
a
n
ge
of
po
s
s
i
b
l
e
v
a
l
ue
s
f
o
r
k,
t
h
e
c
l
us
t
e
r
i
n
g
r
e
s
u
l
t
is
u
n
s
t
a
bl
e
.
In
[
21]
,
t
h
e
gr
e
e
d
y
a
l
go
r
i
t
hm
wa
s
a
pp
l
i
e
d
to
o
b
t
a
i
n
t
h
e
n
u
m
be
r
of
c
e
n
t
e
r
s
a
n
d
m
a
de
t
h
e
f
i
na
l
c
l
us
t
e
r
i
n
g
r
e
s
u
l
t
f
e
a
t
ur
e
s
of
hi
g
h
e
r
a
c
c
ur
a
c
y
r
a
t
e
a
n
d
s
t
a
bil
i
t
y
.
S
h
e
n
-
Yi
et
al.
[
22]
h
a
v
e
i
n
i
t
i
a
l
i
z
e
d
K
-
M
e
a
ns
by
t
he
i
n
i
t
i
a
l
c
e
n
t
e
r
s
pr
o
duc
e
d
by
hi
e
r
a
r
c
hi
c
a
l
c
l
u
s
t
e
r
i
n
g
a
l
go
r
i
t
hm
,
t
h
e
c
l
us
t
e
r
i
n
g
r
e
s
u
l
t
s
f
o
u
n
d
in
t
e
r
m
s
of
FM
a
r
e
n
o
t
v
e
r
y
c
o
nvi
nc
i
ng,
a
n
d
t
h
e
da
t
a
s
e
t
is
f
r
o
m
t
h
e
C
hi
n
e
s
e
t
e
x
t
c
o
r
pus
.
A
c
c
o
r
di
n
g
to
[
23]
,
th
e
F
u
z
z
y
C
-
M
e
a
n
s
is
ge
n
e
r
a
t
e
d
to
pr
o
duc
e
i
ni
t
i
a
l
c
e
n
t
e
r
s
,
a
n
d
pa
r
t
i
c
l
e
s
wa
r
m
o
pt
i
m
i
z
a
t
i
o
n
(
P
S
O)
is
us
e
d
to
m
a
ke
o
pt
i
m
u
m
c
l
us
t
e
r
s
.
S
o
n
g
et
al.
[
24]
i
m
pr
o
v
e
d
Hua
n
g
M
i
n’
s
a
l
go
r
i
t
hm
[
25]
to
s
e
l
e
c
t
i
ni
t
i
a
l
c
l
us
t
e
r
i
n
g
c
e
n
t
e
r
s
f
o
c
us
i
ng
on
t
h
e
d
i
s
t
a
n
c
e
b
e
t
we
e
n
t
h
e
s
a
m
p
l
e
s
t
h
a
t
h
a
v
e
t
he
s
a
m
e
m
a
xim
u
m
de
n
s
i
t
y
p
a
r
a
m
e
t
e
r
s
a
n
d
c
o
m
pa
r
e
it
w
i
t
h
t
h
e
a
v
e
r
a
ge
d
i
s
t
a
n
c
e
of
t
h
e
da
t
a
s
e
t
.
S
h
e
r
ka
t
et
al.
[
26]
pr
o
duc
e
d
i
ni
t
i
a
l
c
e
n
t
e
r
s
b
a
s
e
d
on
a
m
e
t
h
o
d
c
a
l
l
e
d
de
t
e
r
m
i
n
i
s
t
i
c
s
e
e
d
i
ng
K
-
M
e
a
ns
(
DS
K
M
).
T
h
e
ke
y
i
d
e
a
of
t
h
e
pr
o
po
s
e
d
m
e
t
h
o
d
is
to
s
e
l
e
c
t
k
da
t
a
po
i
n
t
s
t
h
a
t
a
r
e
d
i
s
t
a
n
t
f
r
o
m
e
a
c
h
o
t
h
e
r
,
a
n
d
at
t
h
e
s
a
m
e
t
i
m
e
ha
v
e
a
hi
g
h
L1
n
o
r
m
.
T
h
e
s
e
da
t
a
po
i
n
t
s
a
r
e
us
e
d
to
i
n
i
t
i
a
li
z
e
t
h
e
K
-
M
e
a
n
s
a
l
go
r
i
t
hm
,
t
h
e
m
e
t
h
o
d
ha
s
pr
o
duc
e
d
goo
d
r
e
s
u
l
t
s
be
c
a
us
e
it
t
a
ke
s
as
pa
r
a
m
e
t
e
r
(
k)
t
h
e
r
e
a
l
n
u
m
be
r
of
c
l
a
s
s
e
s
,
whi
c
h
m
e
a
ns
t
h
a
t
it
is
s
e
ns
i
t
i
ve
to
t
h
e
n
u
m
be
r
of
c
l
u
s
t
e
r
pa
r
a
m
e
t
e
r
.
In
t
hi
s
pa
pe
r
,
we
a
r
e
i
n
t
e
r
e
s
t
e
d
in
e
v
a
l
u
a
t
i
n
g
t
h
e
a
ppr
o
a
c
h
de
s
c
r
i
be
d
in
o
ur
pr
e
vi
o
u
s
wo
r
k
[
27
],
whi
c
h
a
s
s
u
m
e
d
go
o
d
r
e
s
u
l
t
s
in
de
t
e
r
m
i
ni
ng
t
h
e
n
u
m
be
r
of
c
l
u
s
t
e
r
s
by
u
s
i
ng
M
S
S
P
a
n
d
c
o
m
m
u
ni
t
y
he
a
l
t
h
c
a
r
e
n
e
t
wo
r
k
(
C
HN
)
,
b
ut
it
o
v
e
r
l
o
o
ke
d
t
h
e
s
e
l
e
c
t
i
n
g
i
n
i
t
i
a
l
c
l
us
t
e
r
i
n
g
c
e
n
t
e
r
s
.
In
o
r
de
r
to
e
v
a
l
u
a
t
e
t
h
i
s
a
ppr
o
a
c
h
,
we
c
o
n
s
i
de
r
t
h
e
n
o
de
s
f
o
u
n
d
by
M
S
S
P
as
t
h
e
i
nit
i
a
l
c
l
u
s
t
e
r
c
e
n
t
r
o
i
ds
.
T
h
u
s
,
we
pr
o
p
o
s
e
a
m
e
t
h
o
d
f
o
r
t
h
e
a
uto
m
a
t
i
c
de
t
e
c
t
i
o
n
of
i
ni
t
i
a
l
c
l
u
s
t
e
r
c
e
n
t
r
o
i
d
s
,
whi
c
h
a
r
e
t
h
e
i
nput
pa
r
a
m
e
t
e
r
s
in
s
e
v
e
r
a
l
pa
r
t
i
t
i
o
ni
ng
c
l
u
s
t
e
r
i
n
g
m
e
t
h
o
ds
.
T
h
e
pr
o
po
s
e
d
a
l
go
r
i
t
hm
is
e
x
e
c
ut
e
d
b
e
f
o
r
e
c
l
u
s
t
e
r
i
n
g,
w
hi
c
h
m
e
a
n
s
t
h
a
t
it
is
i
nde
pe
n
de
n
t
of
a
ny
c
l
u
s
t
e
r
i
n
g
m
e
t
h
o
d
t
h
a
t
s
t
a
r
t
s
wi
t
h
k
c
e
n
t
e
r
s
,
it
is
e
f
f
i
c
i
e
n
t
of
l
a
r
ge
da
t
a
s
e
t
s
a
n
d
m
uc
h
o
p
t
i
mi
z
e
d
in
t
e
r
m
s
of
t
i
m
e
.
It
c
o
n
s
i
s
t
s
to
m
o
de
l
i
z
e
t
h
e
t
e
x
t
doc
u
m
e
n
t
c
l
u
s
t
e
r
i
n
g
pr
o
bl
e
m
as
t
h
e
M
S
S
P
,
a
n
d
u
s
e
C
HN
to
s
o
l
ve
t
h
e
M
S
S
P
pr
o
bl
e
m
to
h
a
v
e
do
c
u
m
e
n
t
s
e
e
d
s
.
We
ha
v
e
c
h
o
s
e
n
K
-
M
e
a
n
s
as
a
c
l
us
t
e
r
i
n
g
m
e
t
h
o
d
to
t
e
s
t
o
ur
e
x
pe
r
i
m
e
n
t
a
t
i
o
n
.
T
hi
s
p
a
pe
r
is
s
t
r
uc
t
ur
e
d:
S
e
c
t
i
o
n
2
de
s
c
r
i
be
s
t
h
e
c
o
n
c
e
pt
of
t
e
x
t
c
l
us
t
e
r
i
n
g.
S
e
c
t
i
o
n
3
d
i
s
c
us
s
e
s
M
S
S
P
a
n
d
C
HN
,
w
hi
c
h
a
r
e
t
h
e
m
a
i
n
c
o
m
po
n
e
n
t
s
of
t
he
m
e
t
h
o
d
a
n
d
pr
e
s
e
n
t
s
i
t
s
s
t
e
ps
.
In
s
e
c
t
i
o
n
4,
we
pr
e
s
e
n
t
t
h
e
im
p
l
e
m
e
n
t
a
t
i
o
n
a
n
d
t
he
e
x
pe
r
im
e
n
t
a
l
r
e
s
u
l
t
s
of
t
h
e
K
M
_
M
S
S
P
.
S
e
c
t
i
o
n
5
c
o
n
c
ludes
t
hi
s
wo
r
k
.
2.
F
UN
DA
ME
NT
AL
CONC
E
P
T
S
OF
DOCU
ME
NT
S
CL
UST
E
R
I
NG
A
do
c
u
me
nt
c
l
us
t
e
r
i
ng
i
s
a
d
i
ff
i
c
u
l
t
t
a
s
k
i
n
t
e
xt
mi
n
i
ng,
i
t
c
o
n
s
i
s
t
s
of
gr
o
up
i
ng
s
i
mi
l
a
r
do
c
u
me
n
t
s
.
wh
e
n
th
e
t
o
p
i
cs
do
n’
t
kn
o
w
i
n
a
d
va
n
c
e
.
I
ts
go
a
l
i
s
to
or
ga
n
i
z
e
a
c
o
l
l
e
ct
i
on
of
do
c
u
me
n
t
s
a
c
c
o
r
d
i
ng
to
th
ei
r
to
p
i
cs
a
n
d
h
e
lp
us
e
r
s
to
h
a
v
e
i
nf
o
r
m
at
i
on
acc
e
ss,
i
ts
pr
o
c
e
ss
c
o
n
s
i
s
t
s
of
th
e
f
o
l
l
o
w
i
ng
st
e
ps
[
28]
2.
1.
T
e
x
t
p
r
e
-
p
r
oc
e
s
s
in
g
E
a
c
h
t
e
xt
i
s
r
e
pr
e
s
e
nt
e
d
i
n
th
e
v
e
c
t
or
s
pa
c
e
m
od
e
l
(
VS
M
)
wh
i
ch
i
s
an
a
l
g
e
b
r
a
i
c
m
od
e
l
f
o
r
r
e
pr
e
s
e
nt
i
ng
t
e
xt
do
c
u
me
n
t
s
as
v
e
c
t
o
r
s
of
t
e
r
m
s.
Th
e
n,
th
e
s
t
o
p
w
o
r
ds
l
i
st
a
n
d
pu
n
c
t
ua
t
i
on
m
a
r
ks
m
us
t
b
e
r
em
ov
e
d.
St
e
m
mi
ng
wh
i
ch
i
s
th
e
pr
o
c
e
ss
of
r
e
du
c
i
ng
wo
r
ds
to
th
ei
r
st
em
or
r
oot
f
o
r
m
i
s
p
e
r
f
o
r
me
d,
a
n
d
a
pr
e
-
pr
o
c
e
ss
i
ng
f
i
lt
e
r
ha
s
a
l
r
e
a
d
y
b
ee
n
a
pp
l
ie
d
to
th
e
da
t
a
to
e
l
imi
na
t
e
t
e
r
m
s
t
h
a
t
h
a
v
e
a
l
o
w
fr
e
qu
e
n
c
y
(
c
o
un
t
<
3)
,
l
e
ad
i
ng
to
s
i
gn
i
f
i
c
a
n
t
d
ime
ns
i
o
n
a
l
i
ty
r
e
duc
t
i
on
w
i
t
h
o
ut
l
o
s
s
of
c
l
u
s
t
e
r
i
ng
p
e
r
f
o
r
m
a
n
c
e
.
2.
2.
T
e
r
m
we
igh
t
in
g
To
r
e
pr
e
s
e
nt
qua
n
t
i
t
a
t
i
v
e
ly
th
e
t
e
r
m
i
n
a
do
c
u
me
n
t
,
w
e
us
e
th
e
t
e
r
m
fr
e
qu
e
n
c
y
-
i
nv
e
rs
e
do
c
u
me
nt
fr
e
qu
e
n
c
y
a
l
go
r
i
th
m
(
T
F
-
I
DF
)
.
I
t
t
a
k
e
s
i
n
t
o
a
c
c
o
u
n
t
b
o
t
h
th
e
r
e
l
a
t
i
v
e
fr
e
qu
e
nc
y
of
a
st
em
i
n
a
do
c
u
me
nt
a
n
d
th
e
fr
e
qu
e
nc
y
of
th
e
st
em
w
i
th
i
n
th
e
c
o
r
pus
.
A
co
m
po
s
i
t
e
w
ei
g
h
t
f
or
e
ach
t
e
r
m
i
n
e
a
c
h
do
c
u
me
nt
i
s
ij
,
i
t
i
s
g
i
v
e
n
by
,
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
do
n
e
s
i
a
n
J
E
l
e
c
E
n
g
&
C
o
m
p
S
c
i
I
S
S
N:
2502
-
4752
M
ax
Stabl
e
s
e
t
pr
oblem
to
f
ound
the
ini
ti
al
c
e
ntr
oids
in
c
lus
ter
ing
pr
oblem
(
A
w
ati
f
K
ar
im)
571
=
×
2.
3.
S
im
il
ar
it
y
m
e
a
s
u
r
e
Th
e
r
e
a
r
e
s
e
v
e
r
a
l
me
a
s
ur
e
s
of
s
imi
l
a
r
i
ty
b
e
tw
ee
n
do
c
u
me
n
t
s
i
n
th
e
l
i
t
e
r
a
t
ur
e
.
I
n
pa
r
t
i
c
u
l
a
r
,
w
e
f
i
nd
th
e
E
uc
l
i
d
e
a
n
,
M
a
nh
a
t
t
a
n
a
n
d
C
o
s
i
n
e
d
i
s
t
a
nc
e
t
h
a
t
w
e
w
i
ll
b
e
us
e
d
i
n
o
ur
e
x
pe
r
i
m
e
n
t
a
t
i
o
n
,
i
t
i
s
c
o
m
put
e
d.
(
,
)
=
1
−
(
,
)
W
h
e
r
e
(
,
)
i
s
c
o
s
i
n
e
s
i
mi
l
a
r
i
t
y
o
f
t
w
o
t
e
x
t
v
e
c
t
or
s
.
C
os
i
n
e
v
a
l
u
e
i
s
1
wh
e
n
th
e
do
c
u
me
n
t
s
ar
e
si
mi
l
a
r
a
n
d
0
wh
e
n
th
e
y
ar
e
d
i
s
s
im
i
l
a
r
.
2.
4.
Val
id
it
y
i
n
d
e
x
A
c
l
us
t
e
r
i
n
g
e
v
a
l
ua
t
i
o
n
d
e
m
a
n
ds
an
i
nde
pe
n
de
nt
a
n
d
r
e
l
i
a
bl
e
m
e
a
s
ur
e
f
o
r
t
h
e
a
s
s
e
s
s
m
e
n
t
a
n
d
c
o
m
pa
r
i
s
o
n
of
c
l
us
t
e
r
i
n
g
e
x
pe
r
im
e
n
t
s
a
n
d
r
e
s
u
l
t
s
.
In
t
hi
s
r
e
s
e
a
r
c
h
,
we
f
o
c
us
on
si
x
v
a
li
da
t
i
o
n
m
e
a
s
ur
e
s
:
F
-
m
e
a
s
ur
e
,
pur
i
t
y
,
e
n
t
r
o
py
,
n
o
r
m
a
li
z
e
d
m
ut
ua
l
i
n
f
o
r
m
a
t
i
o
n
,
X
i
e
-
B
e
ni
I
n
d
e
x
a
n
d
F
uku
y
a
m
a
-
S
uge
n
o
i
nde
x
[
29
]
.
T
h
e
c
o
m
m
o
n
ba
s
i
s
of
t
h
e
i
nde
xe
s
is
t
h
a
t
t
h
e
i
r
c
o
m
put
a
t
i
o
n
s
a
r
e
a
ll
ba
s
e
d
on
a
c
o
n
t
i
n
g
e
n
c
y
t
a
ble,
whi
c
h
de
f
i
ne
s
t
h
e
a
s
s
o
c
i
a
t
i
o
n
b
e
t
we
e
n
t
w
o
c
l
a
s
s
i
f
i
c
a
t
i
o
n
s
on
o
n
e
sa
me
s
e
t
of
i
n
d
i
v
i
dua
l
s
.
3.
TH
E
P
ROP
OS
E
D
ME
T
HO
D
I
n
th
e
f
o
l
l
o
w
i
ng
,
w
e
d
e
s
c
r
i
b
e
th
e
m
ax
im
u
m
s
t
a
bl
e
s
e
t
pr
o
bl
em
a
n
d
w
e
i
n
t
r
o
duc
e
th
e
c
o
n
t
i
n
uo
us
Ho
pf
ie
ld
n
e
t
wor
k,
wh
i
ch
ar
e
th
e
m
a
i
n
co
m
po
n
e
n
t
s
of
o
ur
a
ppr
o
a
c
h
.
Al
s
o
,
t
h
e
C
HN
a
n
d
M
S
S
P
pr
obl
e
m
a
r
e
c
o
m
bi
ne
d
to
f
i
n
d
t
h
e
i
n
i
t
i
al
c
e
n
t
r
o
i
ds
or
do
c
um
e
n
t
s
s
e
e
d
s
.
W
e
h
a
v
e
c
h
o
s
e
n
K
-
Me
a
n
s
as
a
c
l
u
s
t
e
r
i
ng
me
t
h
o
d
to
t
e
st
o
ur
a
ppr
o
a
c
h
.
3.
1.
M
ax
im
u
m
s
t
ab
l
e
s
e
t
p
r
ob
l
e
m
W
e
de
n
o
t
e
by
=
(
,
)
an
un
d
i
r
e
ct
e
d
gr
a
ph
.
i
s
th
e
s
e
t
of
n
o
d
e
s
a
n
d
i
s
th
e
s
e
t
of
e
dg
e
s.
A
s
t
a
bl
e
s
e
t
o
f
a
gr
a
ph
i
s
th
e
s
ub
s
e
t
of
w
i
th
th
e
pr
o
p
e
r
t
y
t
h
a
t
e
ach
pa
i
r
of
i
s
n
o
t
c
o
n
n
e
ct
e
d
v
i
a
an
e
dg
e
i
n
gr
a
p
h
[3
0
].
Th
e
M
SSP
c
o
n
s
i
s
t
s
to
f
i
nd
th
e
s
t
a
bl
e
s
e
t
i
n
a
gr
a
ph
of
m
ax
im
u
m
c
a
r
d
i
na
l
i
t
y
.
Th
e
go
a
l
i
n
th
i
s
pa
p
e
r
i
s
to
c
o
n
s
i
d
e
r
th
e
da
t
a
mi
n
i
ng
a
pp
l
i
cat
i
o
ns
sp
e
c
i
f
i
c
a
ll
y
th
e
c
l
u
s
t
e
r
i
ng
of
t
e
xt
do
c
u
me
n
t
s
th
e
m
ot
i
v
a
t
i
ng
a
pp
l
i
cat
i
on
of
th
e
m
ax
im
u
m
s
t
a
bl
e
s
e
t.
Th
e
MS
SP
i
s
a
w
e
ll
-
k
n
o
wn
NP
-
h
a
r
d
pr
o
bl
em
i
n
co
m
b
i
na
t
o
r
i
al
o
pt
i
mi
zat
i
o
n
,
wh
i
ch
can
b
e
f
o
r
m
u
l
a
t
e
d
as
a
qua
dr
a
t
i
c
0
-
1
pr
o
gr
a
mmi
ng.
To
s
o
l
v
e
th
i
s
l
a
t
t
e
r
pr
o
b
l
e
m
,
w
e
us
e
th
e
C
HN
.
3.
2.
Cont
in
u
ou
s
h
o
p
f
iel
d
n
e
t
wor
k
At
th
e
b
e
g
i
nn
i
ng
of
1980,
Ho
pf
ie
ld
a
n
d
T
a
n
k
[3
1
]
i
n
t
r
o
duc
e
a
H
o
pf
ie
ld
n
e
ur
a
l
n
e
t
wo
r
k
.
I
t
h
a
s
b
ee
n
e
xt
e
ns
i
v
e
ly
s
t
ud
ie
d
of
n
e
ur
a
l
n
e
t
wo
r
ks
a
n
d
i
s
t
r
a
i
n
e
d
e
ff
i
c
ie
n
t
l
y
to
s
o
l
v
e
d
i
ff
i
c
u
l
t
pr
o
bl
em
s
[
3
2
]
s
uc
h
as
pa
tt
e
rn
r
e
c
o
gn
i
t
i
o
n
,
m
od
e
l
i
d
e
nt
i
f
i
c
a
t
i
o
n
,
a
n
d
o
pt
i
mi
zat
i
o
n
.
Th
e
m
a
j
o
r
a
d
v
a
n
t
a
g
e
of
th
e
c
o
n
t
i
n
uo
us
Ho
pf
ie
ld
n
e
t
wo
r
k
i
s
i
n
i
ts
s
t
r
uc
t
ur
e
wh
i
ch
c
a
n
b
e
r
e
al
i
z
e
d
on
an
e
n
e
r
g
y
f
u
n
c
t
i
on
a
ppr
o
a
c
h
by
a
dd
i
ng
th
e
o
bj
e
ct
i
v
e
a
n
d
p
e
n
a
l
i
z
i
n
g
t
h
e
c
o
n
s
tr
a
i
n
t
s
in
or
de
r
to
s
o
l
v
e
s
o
m
e
c
l
a
s
s
i
f
i
c
a
t
i
o
n
a
n
d
op
t
i
m
i
z
a
t
i
o
n
p
r
o
bl
e
m
s
.
T
a
l
a
v
n
a
n
d
Y
e
z
[
33
]
s
h
o
w
e
d
t
h
a
t
wh
e
n
th
e
f
u
n
c
t
i
on
r
e
a
c
h
e
s
a
st
e
a
d
y
s
t
a
t
e
of
a
d
i
ff
e
r
e
nt
i
al
e
qua
t
i
on
s
y
s
t
em
a
s
s
o
c
i
at
e
d
w
i
th
th
e
C
HN
,
an
a
ppr
o
x
im
at
e
s
o
l
ut
i
on
of
s
e
v
e
r
a
l
o
pt
i
mi
zat
i
on
pr
o
bl
em
s
i
s
o
b
t
a
i
n
e
d.
Th
ei
r
r
e
s
u
l
t
s
e
n
c
o
ur
a
g
e
d
a
nu
m
b
e
r
of
r
e
s
e
a
r
c
h
e
rs
to
a
ppl
y
th
i
s
n
e
t
wor
k
to
d
i
ff
e
r
e
nt
pr
o
bl
em
s.
Th
e
C
HN
of
s
i
z
e
i
s
a
f
u
ll
y
c
o
nn
e
ct
e
d
n
e
ur
a
l
n
e
t
wo
r
k
w
i
th
c
o
n
t
i
n
uo
us
-
va
l
u
e
d
un
i
ts
(
o
r
n
e
ur
o
n
s
)
.
F
o
l
l
o
w
i
ng
th
e
n
o
t
a
t
i
on
us
e
d
i
n
[3
3
]:
L
e
t
b
e
th
e
s
t
r
e
n
gt
h
of
th
e
c
o
nn
e
ct
i
on
f
r
o
m
n
e
ur
o
n
a
n
d
n
e
ur
o
n
.
Th
e
m
od
e
l
a
s
s
u
me
s
sy
mm
e
tr
i
cal
w
ei
g
h
t
s
(
=
),
i
n
m
o
s
t
c
a
s
e
s,
z
e
ro
s
e
lf
-
c
o
up
l
i
ng
t
e
r
m
s
(
=
0
)
a
n
d
e
ach
n
e
ur
o
n
h
a
s
an
o
f
f
s
e
t
b
i
a
i
s
.
L
e
t
a
n
d
b
e
th
e
c
ur
r
e
nt
s
t
a
t
e
a
n
d
th
e
o
u
t
pu
t
of
th
e
n
e
ur
o
n
,
w
i
th
∈
{
1
,
.
.
,
}
.
Th
e
s
y
s
t
em
of
th
e
C
HN
i
s
d
e
s
c
r
i
b
e
d
by
th
e
d
i
ff
e
r
e
nt
i
al
e
qua
t
i
on
,
=
τ
+
+
w
h
e
r
e
i
s
t
h
e
v
a
l
u
e
of
th
e
t
ime
c
o
ns
t
a
n
t
of
th
e
a
m
pl
i
f
ie
r
s
,
a
n
d
w
i
t
h
o
ut
l
o
s
s
of
g
e
n
e
r
a
l
i
ty
c
a
n
b
e
a
s
s
i
gn
e
d
a
v
a
l
u
e
of
u
ni
t
y
.
Th
e
o
u
t
pu
t
f
u
n
c
t
i
on
i
s
a
hy
p
e
r
b
o
l
i
c
t
a
n
g
e
nt
of
e
a
c
h
n
e
ur
o
n
s
t
a
t
e
.
Th
e
e
x
pr
e
s
s
i
o
n
of
th
e
e
n
e
r
g
y
f
u
n
c
t
i
on
a
s
s
o
c
i
at
e
d
w
i
th
th
e
c
o
n
t
i
n
uo
us
Ho
pf
ie
ld
n
e
t
w
o
r
k
i
s
d
e
f
i
n
e
d
by
,
(
)
=
−
1
2
−
(
)
t
y
p
i
c
a
ll
y
,
i
n
th
e
C
HN
,
th
e
e
n
e
r
g
y
f
u
n
c
t
i
on
i
s
i
n
i
t
i
al
i
z
e
d
to
th
e
o
bj
e
ct
i
v
e
f
u
n
c
t
i
on
of
th
e
o
pt
i
mi
zat
i
on
pr
o
bl
em
a
n
d
f
o
r
e
ach
c
o
n
s
t
r
a
i
nt
a
p
e
n
a
l
t
y
t
e
r
m
i
s
a
dd
e
d.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2502
-
4752
I
n
do
n
e
s
i
a
n
J
E
l
e
c
E
n
g
&
C
o
m
p
S
c
i
,
Vo
l
.
25
,
N
o
.
1
,
J
a
n
ua
r
y
20
22
:
569
-
579
572
3.
3.
M
S
S
P
an
d
CHN
to
f
in
d
in
it
ial
c
e
n
t
r
oid
s
W
e
s
uppo
s
e
t
h
a
t
w
e
h
a
v
e
a
da
t
a
s
e
t
of
do
c
u
me
n
t
s
.
S
o
w
e
w
a
n
t
to
d
i
v
i
d
e
a
da
t
a
s
e
t
i
n
t
o
gr
o
ups
s
uc
h
t
h
a
t
th
e
m
e
m
be
r
s
of
e
a
c
h
gr
o
up
ar
e
as
s
im
i
l
a
r
as
p
o
s
s
i
bl
e
to
on
e
a
n
o
t
h
e
r.
Th
e
ov
e
r
a
ll
f
l
o
w
of
th
e
a
l
g
o
r
i
th
m
i
s
s
h
o
w
n
i
n
F
i
g
ur
e
1.
F
i
gur
e
1
.
M
S
S
P
a
n
d
C
HN
to
f
i
n
d
i
ni
t
i
a
l
c
e
n
t
r
o
i
ds
Th
e
s
p
e
c
i
f
i
c
pr
o
c
e
ss
of
th
e
a
l
go
r
i
th
m
i
s
:
-
Af
t
e
r
t
h
e
pr
e
pr
o
c
e
s
s
i
n
g
s
t
e
p,
we
c
o
m
put
e
t
h
e
t
e
r
m
-
do
c
um
e
n
t
m
a
t
r
i
x
.
-
B
a
s
e
d
on
t
h
e
c
o
s
i
n
e
d
i
s
t
a
n
c
e
be
t
we
e
n
do
c
u
m
e
n
t
s
we
c
r
e
a
t
e
a
s
i
mi
l
a
r
i
t
y
m
a
t
r
i
x
.
T
h
e
r
e
f
o
r
e
t
h
e
n
o
de
s
V
a
n
d
e
dge
s
E
of
t
h
e
gr
a
ph
(
,
)
a
r
e
b
u
i
l
d
i
ng.
-
T
h
e
n
we
r
e
pr
e
s
e
n
t
a
t
e
x
t
d
o
c
um
e
n
t
c
l
us
t
e
r
i
n
g
pr
o
b
l
e
m
as
t
h
e
m
a
x
s
t
a
bl
e
s
e
t
pr
o
bl
e
m
(
M
S
S
P
)
.
-
T
h
e
do
c
u
m
e
n
t
s
a
r
e
c
o
n
s
i
de
r
e
d
as
n
o
de
s
,
To
b
u
i
l
d
t
h
e
e
dge
s
we
c
a
l
c
u
l
a
t
e
t
h
e
s
i
mi
l
a
r
i
t
y
(
c
o
s
i
n
e
d
i
t
a
nc
e
)
b
e
t
we
e
n
d
i
f
f
e
r
e
n
t
n
o
de
s
.
If
s
im
il
a
r
i
t
y
b
e
t
we
e
n
t
o
w
do
c
um
e
n
t
s
bi
j
>
we
h
a
v
e
an
e
dge
b
e
t
we
e
n
t
h
e
m
.
-
Af
t
e
r
t
h
a
t
we
us
e
qua
dr
a
t
i
c
i
n
t
e
g
e
r
pr
o
gr
a
m
mi
ng
f
o
r
m
u
l
a
t
i
o
n
QP
,
to
de
t
e
r
m
i
ne
t
h
e
s
t
a
bl
e
s
e
t
m
a
xim
a
l
S
a
s
s
o
c
i
a
t
e
d
w
i
t
h
t
hi
s
gr
a
ph
.
(
)
{
(
)
=
−
∑
=
1
=
0
∈
{
0
,
1
}
ℎ
∀
∈
{
1
,
…
,
}
,
=
{
1
∈
0
ℎ
C
i
s
an
n
×
n
sy
mm
e
tr
i
c
m
a
t
r
i
x
d
e
f
i
n
e
d
by
:
=
{
1
(
,
)
∈
0
ℎ
-
Th
e
c
o
n
t
i
n
uo
us
Ho
pf
ie
ld
n
e
t
wor
k
i
s
im
pl
e
m
e
nt
e
d
to
s
o
l
v
e
th
e
QP
pr
o
bl
em
.
-
So
th
e
s
o
l
ut
i
on
i
s
d
e
n
o
t
e
d
as
v
e
c
t
or
t
h
a
t
h
a
v
e
v
a
l
u
e
s
0
a
n
d
1.
I
f
th
e
v
a
l
u
e
at
p
o
s
i
t
i
on
i
s
1
t
h
a
t
me
a
n
s
ℎ
do
c
u
me
nt
i
s
s
e
l
e
ct
e
d
as
c
e
n
t
r
o
i
d
ot
h
e
rw
i
s
e
i
t
i
s
n
ot
s
e
l
e
ct
e
d.
-
F
i
na
ll
y
,
w
e
o
b
t
a
i
n
do
c
u
me
n
t
s
f
u
ll
y
d
i
s
c
o
nn
e
ct
e
d
(
do
c
u
me
nt
s
ee
ds
)
.
3.
4.
K
-
M
e
an
s
al
gor
it
h
m
in
it
ial
ize
d
by
M
S
S
P
(
KM
_
M
S
S
P
)
K
-
M
e
a
n
s
i
s
a
co
mm
o
nly
us
e
d
c
l
u
s
t
e
r
i
ng
me
t
h
o
d
i
n
t
e
xt
c
l
u
s
t
e
r
i
n
g,
wh
i
ch
s
e
rv
e
s
c
e
n
t
r
o
i
ds
to
r
e
pr
e
s
e
nt
c
l
u
s
t
e
rs
by
mi
n
imi
z
i
ng
th
e
s
qua
r
e
d
e
r
r
o
r
s
.
Th
e
a
l
go
r
i
th
m
b
e
g
i
ns
w
i
th
a
pr
e
d
e
f
i
n
e
d
s
e
t
of
c
e
n
t
r
o
i
ds
(
wh
i
ch
c
a
n
b
e
pr
o
duc
e
d
ei
th
e
r
r
a
n
do
m
ly
or
by
me
an
s
of
a
ny
o
t
h
e
r
cr
i
t
e
r
i
o
n
,
i
n
th
i
s
s
t
ud
y
by
M
S
S
P
)
.
I
t
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
do
n
e
s
i
a
n
J
E
l
e
c
E
n
g
&
C
o
m
p
S
c
i
I
S
S
N:
2502
-
4752
M
ax
Stabl
e
s
e
t
pr
oblem
to
f
ound
the
ini
ti
al
c
e
ntr
oids
in
c
lus
ter
ing
pr
oblem
(
A
w
ati
f
K
ar
im)
573
ach
ie
v
e
s
s
e
qu
e
nt
i
al
r
e
p
e
t
i
t
i
o
ns
of
th
e
r
e
st
of
th
e
sa
m
pl
e
a
c
c
o
r
d
i
ng
to
th
e
s
i
mi
l
a
r
i
ty
(
us
i
ng
c
o
s
i
n
e
d
i
s
t
a
nc
e
)
w
i
th
th
e
c
e
n
t
r
o
i
ds
s
uc
h
as
e
ach
do
c
u
me
nt
i
s
a
s
s
i
gn
e
d
to
th
e
c
l
u
s
t
e
r
w
i
th
th
e
m
o
s
t
s
i
mi
l
a
r
c
e
n
t
r
o
i
d.
Th
e
n,
th
e
a
l
go
r
i
th
m
i
s
i
t
e
r
a
t
i
v
e
pr
o
c
e
ss
i
ng
a
n
d
a
d
j
us
t
th
e
c
e
nt
e
r
po
s
i
t
i
o
n
,
u
n
t
i
l
no
r
e
a
s
s
i
gn
me
nt
of
pa
t
t
e
r
n
s
to
n
e
w
c
l
u
s
t
e
r
c
e
n
t
r
o
i
ds
or
mi
n
im
al
d
e
cr
e
as
e
i
n
s
qu
a
r
e
e
r
r
o
r
or
th
e
nu
m
b
e
r
of
i
t
e
r
a
t
i
o
n
s
h
a
s
s
ur
pa
s
s
e
d
a
t
h
r
e
s
h
o
l
d.
So
th
i
s
pr
o
duc
e
s
a
s
e
pa
r
a
t
i
on
of
th
e
o
bj
e
c
t
s
i
n
t
o
gr
o
ups
f
r
o
m
wh
i
ch
th
e
me
tr
i
c
to
b
e
mi
n
im
i
z
e
d
can
b
e
c
a
l
c
u
l
a
t
e
d.
Th
e
l
a
s
t
st
e
p
c
o
n
c
e
r
n
s
im
pr
o
v
i
ng
th
e
o
b
t
a
i
n
e
d
s
o
l
ut
i
on
f
o
u
n
d
by
M
SSP
a
n
d
C
HN
as
an
i
n
put
pa
r
a
me
t
e
r
i
n
K
-
Me
a
n
s
a
s
s
h
o
wn
i
n
F
i
g
ur
e
2,
a
nd
e
va
l
ua
t
e
s
th
e
c
l
us
t
e
r
i
ng
r
e
su
l
t
s
b
a
s
e
d
on
so
me
v
a
l
i
d
i
ty
cr
i
t
e
r
i
a.
F
i
gur
e
2.
T
h
e
pr
o
p
o
s
e
d
K
M
_M
S
S
P
a
l
go
r
i
t
hm
4.
R
E
S
UL
T
S
A
ND
D
I
S
CU
S
S
I
ON
I
n
o
r
de
r
to
d
em
o
n
s
t
r
a
t
e
th
e
e
ff
i
c
ie
nc
y
of
o
ur
a
ppr
o
a
c
h
,
w
e
h
a
v
e
a
f
f
e
ct
e
d
a
s
e
r
ie
s
of
e
xp
e
r
im
e
n
t
s
of
i
ns
t
a
nc
e
s
f
r
o
m
th
e
d
a
t
a
s
e
t
b
r
i
t
i
s
h
b
r
o
a
dc
a
s
t
i
n
g
c
o
r
po
r
a
t
i
o
n
(
BBC
)
a
n
d
da
t
a
s
e
t
20Ne
ws
Gr
o
up
.
M
o
s
t
o
f
t
h
e
s
e
i
ns
t
a
nc
e
s
a
r
e
c
r
e
a
t
e
d
by
v
a
r
yi
ng
t
h
e
n
u
m
be
r
o
f
c
l
a
s
s
e
s
a
n
d
t
h
e
n
u
m
be
r
o
f
do
c
u
m
e
n
t
s
.
I
n
t
hi
s
s
e
c
t
i
o
n
we
pr
e
s
e
n
t
t
h
e
da
t
a
s
e
t
de
s
c
r
i
pt
i
o
n
,
t
h
e
r
e
s
u
l
t
s
a
n
d
t
h
e
c
o
m
pa
r
i
s
o
n
o
f
K
M
wh
i
ch
us
e
s
r
a
n
do
m
c
e
n
t
r
o
i
d
s
w
i
th
K
M
_
M
SSP
wh
i
ch
s
t
a
r
t
s
by
c
e
n
t
r
o
i
ds
f
o
un
d
by
M
SSP
.
4.
1.
Dat
a
s
e
t
d
e
s
c
r
ip
t
ion
BBC
D
a
t
a
s
e
t
c
o
n
s
i
s
t
s
of
2225
do
c
u
me
n
t
s
f
r
o
m
th
e
B
B
C
n
e
ws
w
e
bs
i
t
e
c
o
r
r
e
s
po
n
d
i
ng
to
s
tor
ie
s
i
n
f
i
v
e
t
o
p
i
c
a
l
ar
e
as
f
r
o
m
2004
-
2005.
Th
e
d
a
t
a
s
e
t
i
s
c
l
a
s
s
i
f
ie
d
i
n
t
o
f
i
v
e
n
a
t
ur
al
c
l
a
s
s
e
s
s
uc
h
as
b
us
i
n
e
ss,
e
nt
e
r
t
a
i
n
me
n
t
,
p
o
l
i
t
i
c
s
,
s
po
r
t
,
a
n
d
t
e
c
h
n
o
l
o
g
y
.
I
t
i
s
a
v
a
i
l
a
bl
e
i
n
[3
4
].
A
s
u
b
s
e
t
of
th
e
da
t
a
s
e
ts
i
s
t
a
k
e
n
a
n
d
s
e
pa
r
a
t
e
d
i
n
t
o
f
o
ur
c
a
t
e
gor
ie
s,
w
e
us
e
t
h
r
ee
s
ubs
e
ts
of
th
e
do
c
u
me
n
t
s
f
r
o
m
th
e
da
t
a
s
e
t
BBC.
Th
e
f
i
r
s
t
s
ub
s
e
t
c
o
n
t
a
i
ns
5
t
o
p
i
cs
a
n
d
500
do
c
u
me
n
t
s
,
th
e
s
e
c
o
n
d
c
o
n
t
a
i
ns
3
t
o
p
i
cs
a
n
d
1328
do
c
u
me
n
t
s
,
th
e
th
i
rd
c
o
n
t
a
i
ns
f
o
ur
to
p
i
cs
a
n
d
1715
do
c
u
me
n
t
s
a
n
d
e
a
c
h
do
c
u
me
nt
h
a
s
a
s
i
ng
l
e
to
p
i
c
l
a
b
e
l,
as
s
h
o
w
n
i
n
T
a
bl
e
1.
BBC
s
po
r
t
da
t
a
s
e
t
c
o
n
s
i
s
t
s
of
737
do
c
u
me
n
t
s
f
r
o
m
th
e
B
B
C
S
po
r
t
w
e
bs
i
t
e
c
o
r
r
e
s
po
n
d
i
ng
to
s
por
t
s
n
e
w
s
a
r
t
i
cl
e
s
i
n
f
i
v
e
t
o
p
i
c
a
l
ar
e
as
f
r
o
m
2004
-
2005
.
Th
e
da
t
a
s
e
t
i
s
c
l
a
s
s
i
f
ie
d
i
n
t
o
f
i
v
e
n
a
t
ur
a
l
c
l
a
s
s
e
s
s
uc
h
as
a
t
hl
e
t
i
c
s
,
c
r
i
c
ke
t
,
f
o
ot
b
a
l
l
,
r
ug
by
,
t
e
nn
i
s.
I
t
i
s
a
v
a
i
labl
e
i
n
[3
4
].
W
e
us
e
on
e
s
u
bs
e
t
of
th
e
do
c
u
me
n
t
s
f
r
o
m
t
h
e
da
t
a
s
e
t
BBC,
wh
i
ch
c
o
n
t
a
i
ns
3
to
p
i
cs
a
n
d
372
do
c
u
me
n
t
s
.
20Ne
ws
Gr
o
up
:
i
s
a
c
o
l
l
e
ct
i
on
of
a
ppr
o
x
im
at
e
ly
20,
000
n
e
ws
gr
o
up
d
o
c
um
e
n
t
s
,
pa
r
t
i
t
i
on
e
d
i
n
20
gr
o
ups
.
W
e
h
a
v
e
s
e
l
e
ct
e
d
a
s
u
bs
e
t
of
th
i
s
da
t
a
s
e
t
c
o
n
t
a
i
n
i
ng
tot
a
l
400
do
c
u
me
n
t
s
f
r
o
m
ov
e
r
f
o
ur
cat
e
g
o
r
ie
s.
T
h
e
20
n
e
ws
gr
o
ups
c
o
l
l
e
c
t
i
o
n
is
a
po
pul
a
r
da
t
a
s
e
t
f
o
r
e
x
pe
r
im
e
n
t
s
i
n
t
e
x
t
c
l
a
s
s
i
f
i
c
a
t
i
o
n
a
n
d
t
e
x
t
c
l
u
s
ter
i
n
g.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2502
-
4752
I
n
do
n
e
s
i
a
n
J
E
l
e
c
E
n
g
&
C
o
m
p
S
c
i
,
Vo
l
.
25
,
N
o
.
1
,
J
a
n
ua
r
y
20
22
:
569
-
579
574
4.
2.
E
x
p
e
r
im
e
n
t
a
l
r
e
s
u
l
t
s
S
i
m
u
l
a
t
i
ve
e
xp
e
r
im
e
n
t
s
ar
e
e
x
e
c
ut
e
d
to
s
h
o
w
th
e
a
dv
a
n
t
a
g
e
s
a
n
d
dr
a
w
i
ng
sc
ie
nt
i
f
i
c
r
em
a
r
ks
on
th
e
a
l
go
r
i
th
m
a
da
pt
a
b
i
l
i
ty
of
th
e
Im
pr
o
v
e
d
K
M
_
M
S
S
P
.
Th
e
pa
r
a
me
t
e
r
of
s
i
mi
l
a
r
i
ty
i
s
d
e
t
e
r
mi
n
e
d
by
s
e
v
e
r
a
l
t
e
s
t
s
a
n
d
i
s
f
i
x
e
d
as
(
=
1
)
.
As
s
h
o
wn
i
n
T
a
bl
e
1
,
i
n
da
t
a
s
e
t
B
B
C
_2225
w
e
h
a
v
e
2225
d
o
c
u
me
n
t
s
or
ga
n
i
z
e
d
i
n
t
o
5
d
i
ff
e
r
e
nt
c
l
a
s
s
e
s,
i
n
th
e
f
o
ur
t
h
c
o
l
u
m
n
(
Nu
m
b
e
r
of
c
l
u
s
t
e
rs
o
b
t
a
i
n
e
d
by
M
S
S
P
)
,
o
ur
me
t
h
o
d
g
i
v
e
s
6
c
l
u
s
t
e
rs
f
o
r
B
B
C
_2
225
a
n
d
5
c
l
us
t
e
rs
f
o
r
B
B
C
_1715
wh
i
ch
ar
e
v
e
ry
c
l
o
s
e
to
th
e
v
a
l
u
e
of
th
e
th
i
rd
c
o
l
u
m
n
(
r
e
al
nu
m
b
e
r
of
c
l
a
s
s
e
s
)
.
Ho
w
e
v
e
r,
i
n
B
B
C
_500,
B
C
C
_1328,
B
B
C
_S
po
r
t
1
a
n
d
B
B
C
_S
po
r
t
2
th
e
s
o
l
ut
i
on
i
s
e
qua
l
to
th
e
r
e
al
nu
m
b
e
r
of
c
l
a
s
s
e
s.
H
e
nc
e
,
w
e
c
o
n
c
l
ud
e
t
h
a
t
o
u
r
me
t
h
o
d
g
i
v
e
s
f
o
r
e
a
c
h
da
t
a
s
e
t
a
v
e
ry
c
l
o
s
e
or
an
e
qua
l
r
e
s
u
l
t
to
th
e
r
e
al
nu
m
b
e
r
of
c
l
a
s
s
e
s.
I
n
c
o
l
u
m
n
5,
w
e
o
b
s
e
rv
e
t
h
a
t
th
e
i
n
i
t
i
al
c
e
n
t
r
o
i
ds
pr
opo
s
e
d
by
M
SSP
f
o
r
e
ach
i
ns
t
a
n
c
e
of
th
e
da
t
a
s
e
t
ar
e
v
e
ry
d
i
ss
i
mi
l
a
r
a
n
d
b
e
l
o
n
g
to
d
i
ff
e
r
e
nt
gr
o
ups
.
Th
e
e
x
e
c
ut
i
on
t
i
me
i
s
v
e
ry
l
imi
t
e
d;
i
t
v
a
r
ie
s
a
c
c
o
r
d
i
ng
to
th
e
s
i
z
e
of
da
t
a
s
e
t
as
s
h
o
wn
i
n
c
o
l
u
m
n
6.
To
e
x
a
mi
ne
th
e
qua
l
i
ty
of
o
ur
a
ppr
o
a
c
h
,
a
s
t
a
t
i
st
i
c
a
l
s
t
udy
w
a
s
r
e
pr
e
s
e
nt
e
d;
th
i
s
s
t
ud
y
i
s
b
a
s
e
d
on
th
e
c
a
l
c
u
l
a
t
i
on
op
e
r
a
to
r
’
s
p
e
r
f
o
r
m
a
n
c
e
,
=
ℎ
i
n
th
i
s
c
o
n
t
e
x
t
,
th
e
r
a
t
i
o
mi
n
im
u
m
i
s
a
l
wa
y
s
s
up
e
r
i
or
or
e
qua
l
to
1
th
e
n
th
e
K
M
_
M
SSP
g
i
v
e
th
e
upp
e
r
b
o
un
d
of
r
e
al
n
u
m
b
e
r
of
c
l
a
s
s
.
M
or
e
ov
e
r,
i
f
th
e
r
a
t
i
o
mi
n
im
u
m
i
s
e
qu
a
l
s
to
1,
th
e
n
th
e
K
M
_
M
SSP
h
a
s
f
o
un
d
th
e
r
e
al
nu
m
b
e
r
of
c
l
a
s
s
e
s
e
x
i
st
i
ng
i
n
th
e
l
i
t
e
r
a
t
ur
e
.
T
a
bl
e
1.
I
ni
t
i
a
l
c
e
n
t
r
o
i
d
s
o
b
t
a
i
n
e
d
by
KM
_M
S
S
P
out
of
20
r
un
s
D
a
ta
s
e
t
N
umbe
r
of
doc
um
e
nt
s
R
e
a
l
numbe
r
of
cl
a
s
s
e
s
N
umbe
r
of
c
lu
s
t
e
r
s
o
bt
a
in
e
d
R
a
ti
o
I
ni
ti
a
l
c
e
nt
r
oi
ds
o
bt
a
in
e
d
C
P
U
ti
me
(
s
e
c
)
m
ode
M
e
an
m
in
C
lu
s
te
r
s
D
oc
um
e
nt
s
B
B
C
_N
e
w
s
B
B
C
_2225
2225
5
6
1.2
1.19
1.2
S
por
t
po
li
ti
c
s
s
po
r
t
e
nt
e
r
ta
in
m
e
nt
te
c
h
bus
in
e
s
s
199.tx
t
397.tx
t
324.tx
t
243.tx
t
334.tx
t
280.t
xt
4.2
s
B
B
C
_1328
1328
3
3
01
1.1
01
bus
in
e
s
s
po
li
ti
c
s
te
c
h
394.tx
t
397.tx
t
284.tx
t
1.2
s
B
B
C
_500
500
5
5
01
1.1
01
s
po
r
t
bus
in
e
s
s
e
nt
e
r
ta
in
m
e
nt
po
li
ti
c
s
te
c
h
018.tx
t
080.tx
t
092.tx
t
074.tx
t
039.tx
t
0.18
s
B
B
C
_1715
1715
4
5
1.25
1.25
1.25
s
po
r
t
s
po
r
t
po
li
ti
c
s
e
nt
e
r
ta
in
m
e
nt
te
c
h
191.tx
t
324.tx
t
397.tx
t
154.tx
t
284.tx
t
0.06s
B
B
C
_S
por
t
B
B
C
_S
p
or
t1
737
5
5
01
01
01
f
oo
tb
a
ll
a
th
le
ti
c
s
f
oo
tb
a
ll
c
r
i
c
k
e
t
r
ugb
y
263.tx
t
026.tx
t
147.tx
t
077.tx
t
067.tx
t
0.01s
B
B
C
_S
p
or
t2
372
3
3
01
01
01
r
ugb
y
c
r
i
c
k
e
t
a
th
le
ti
c
s
041.tx
t
104.tx
t
019.tx
t
0.002s
20N
e
s
G
r
oup
20N
G
_400
400
4
4
01
01
01
mi
s
c
.
f
o
r
s
a
le
c
o
mp.gr
a
phi
c
s
r
e
c
.a
ut
o
s
r
e
c
.m
o
t
o
r
c
y
c
l
e
s
76027.tx
t
38464.tx
t
103209.tx
t
104297.tx
t
0.16s
4.
3.
Com
p
ar
is
on
b
e
t
we
e
n
c
l
as
s
ic
K
-
M
e
an
s
(
KM
)
an
d
K
-
M
e
an
s
im
p
r
ove
d
(
KM
_
M
S
S
P
)
I
n
or
de
r
to
d
em
o
n
s
t
r
a
t
e
th
e
e
ff
i
c
ie
n
c
y
of
o
ur
a
ppr
o
a
c
h
,
w
e
co
m
pa
r
e
K
M
wh
i
ch
us
e
s
r
a
n
do
m
c
e
n
t
r
o
i
ds
w
i
th
K
M
_
M
SSP
wh
i
ch
s
t
a
r
t
s
by
c
e
n
tr
o
i
ds
f
o
un
d
by
M
SSP,
us
i
ng
s
i
x
v
a
l
i
d
i
ty
me
a
s
u
r
e
s:
F
-
me
a
s
ur
e
,
pur
i
t
y
,
e
n
t
r
o
py
,
X
ie
B
e
n
i
i
nd
e
x,
F
uku
y
a
m
a
S
ug
e
no
i
nd
e
x
a
n
d
n
o
r
m
a
li
z
e
d
m
ut
ua
l
i
nf
o
r
m
a
t
i
o
n
(
N
MI
)
.
Th
e
co
m
pa
r
i
s
o
n
b
e
tw
ee
n
K
M
a
n
d
K
M
_
M
SSP
h
a
s
to
b
e
im
pl
e
m
e
nt
e
d
on
th
e
sa
me
nu
m
b
e
r
of
c
l
u
s
t
e
rs
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
do
n
e
s
i
a
n
J
E
l
e
c
E
n
g
&
C
o
m
p
S
c
i
I
S
S
N:
2502
-
4752
M
ax
Stabl
e
s
e
t
pr
oblem
to
f
ound
the
ini
ti
al
c
e
ntr
oids
in
c
lus
ter
ing
pr
oblem
(
A
w
ati
f
K
ar
im)
575
fo
u
n
d
e
d
by
M
SSP
a
n
d
u
n
d
e
r
th
e
sa
me
e
nv
i
r
o
n
me
nt
by
me
a
s
ur
i
ng
th
e
c
o
s
i
n
e
d
i
s
t
a
n
c
e
b
e
tw
ee
n
th
e
g
i
v
e
n
e
l
e
m
e
nt
r
e
pr
e
s
e
n
t
a
t
i
on
a
n
d
th
e
c
e
n
t
r
o
i
d
of
th
e
c
l
u
s
t
e
r.
B
u
t
th
e
d
i
ff
e
r
e
nc
e
i
s
t
h
a
t
K
M
i
s
b
a
s
e
d
on
th
e
r
a
n
do
m
c
h
o
i
c
e
of
c
e
nt
e
r
s
,
wh
e
r
e
as
K
M
_
M
SSP
s
t
a
r
t
s
w
i
th
th
e
c
e
n
t
r
o
i
ds
f
o
un
d
by
M
SSP.
4.
3.
1.
F
-
m
e
as
u
r
e
c
om
p
a
r
is
on
of
KM
an
d
t
h
e
p
r
op
os
e
d
KM
_
M
S
S
P
al
gor
it
h
m
As
c
a
n
b
e
s
ee
n
f
r
o
m
F
i
g
ur
e
3
,
th
e
av
e
r
a
g
e
F
-
me
tr
i
c
of
th
e
c
l
us
t
e
r
i
ng
a
l
go
r
i
th
m
of
K
M
_
M
SSP
i
s
94%
on
da
t
a
s
e
t
B
B
C
_1328,
92%
on
da
t
a
s
e
t
B
B
C
_2225,
86%
on
B
B
C
_500,
81
%
on
B
B
C
_1715,
68%
on
B
B
C
_S
po
r
t
1
a
n
d
96%
on
B
B
C
_S
po
r
t
2
.
W
h
i
ch
i
s
h
i
gh
e
r
t
h
a
n
th
e
av
e
r
a
g
e
F
-
me
tr
i
c
of
th
e
K
M
c
l
us
t
e
r
i
ng
of
81%
on
B
B
C
_1328,
75%
on
da
t
a
s
e
t
B
B
C
_2225
a
n
d
73%
on
B
B
C
_500.
S
o
,
w
e
n
ot
i
c
e
t
h
a
t
th
e
r
e
i
s
an
i
nc
r
e
as
e
of
F
-
Me
a
s
ur
e
v
a
l
u
e
s
i
n
K
M
_
M
SSP
co
m
p
a
r
e
d
w
i
th
t
h
o
s
e
of
K
M
i
n
a
ll
a
v
a
i
l
a
bl
e
da
t
a
s
e
t
s
.
F
i
gur
e
3.
F
-
m
e
a
s
ur
e
c
o
m
pa
r
i
s
o
n
of
KM
a
n
d
t
h
e
pr
o
p
o
s
e
d
K
M
_M
S
S
P
a
l
go
r
i
t
hm
4.
3.
2.
P
u
r
it
y
c
om
p
ar
i
s
on
of
KM
an
d
t
h
e
p
r
op
os
e
d
K
M
_
M
S
S
P
al
go
r
it
h
m
A
co
m
pa
r
i
s
o
n
of
pur
i
ty
b
e
tw
ee
n
K
M
a
n
d
K
M
_
M
SSP
i
s
d
e
p
i
ct
e
d
i
n
F
i
g
ur
e
4
.
I
t
i
s
p
i
c
t
ur
e
d
t
h
a
t
K
M
_
M
SSP
o
u
t
p
e
r
f
o
r
m
s
K
M
i
n
t
e
r
m
s
of
pur
i
ty
i
n
a
l
l
a
v
a
i
l
a
bl
e
da
t
a
s
e
t
s
.
Th
e
pur
i
ty
of
K
M
_
M
SSP
i
s
94%
on
da
t
a
s
e
t
B
B
C
_1328,
92%
on
da
t
a
s
e
t
B
B
C
_2225
a
nd
86%
on
B
B
C
_500,
wh
i
ch
i
s
h
i
gh
e
r
t
h
a
n
th
e
pur
i
ty
of
th
e
K
M
of
81%
on
B
B
C
_1328,
72%
on
da
t
a
s
e
t
B
B
C
_2225
a
n
d
66%
on
B
B
C
_500.
F
i
gur
e
4.
P
ur
i
t
y
c
o
m
pa
r
i
s
o
n
of
KM
a
n
d
t
h
e
pr
o
p
o
s
e
d
K
M
_
M
S
S
P
a
l
go
r
i
t
hm
4.
3.
3.
E
n
t
r
op
y
c
om
p
ar
i
s
on
of
KM
a
n
d
t
h
e
p
r
op
os
e
d
KM
_
M
S
S
P
al
gor
it
h
m
Al
s
o
,
we
o
b
s
e
rv
e
t
h
a
t
,
th
e
r
e
ar
e
a
d
e
cl
i
n
e
of
e
n
t
r
o
py
a
s
s
h
o
wn
i
n
F
i
g
ur
e
5
,
v
a
l
u
e
s
i
n
K
M
_
M
SSP
co
m
pa
r
e
d
w
i
th
t
h
o
s
e
of
K
M
.
Th
e
e
n
t
r
o
py
of
th
e
K
M
_
M
SSP
i
s
32%
,
wh
i
ch
i
s
l
o
w
e
r
t
h
a
n
th
e
e
n
t
r
op
y
of
th
e
K
M
c
l
us
t
e
r
i
ng
of
61%
on
th
e
B
B
C
_2225
da
t
a
s
e
t.
T
h
us
K
M
_
M
SSP
o
u
t
p
e
r
f
o
r
m
s
K
M
i
n
t
e
r
m
s
of
e
n
t
r
o
py
i
n
a
l
l
a
v
a
i
l
a
bl
e
da
t
a
s
e
t
s
.
4.
3.
4.
N
M
I
c
om
p
ar
is
on
of
KM
an
d
t
h
e
p
r
op
os
e
d
KM
_M
S
S
P
al
gor
it
h
m
A
co
m
pa
r
i
s
o
n
of
N
M
I
b
e
tw
ee
n
K
M
a
n
d
K
M
_
M
SSP
i
s
d
e
p
i
ct
e
d
i
n
F
i
g
ur
e
6
.
I
t
i
s
p
i
c
t
ur
e
d
t
h
a
t
K
M
_
M
SSP
o
u
t
p
e
r
f
o
r
m
s
K
M
i
n
t
e
r
m
s
of
N
M
I
i
n
a
l
l
a
v
a
i
l
a
bl
e
da
t
a
s
e
t
s
.
Th
e
N
M
I
of
K
M
_
M
SSP
i
s
78%
on
da
t
a
s
e
t
B
B
C
_1328
a
n
d
da
t
a
s
e
t
B
B
C
_2225,
63%
on
B
B
C
_500,
85%
on
B
B
C
_S
po
r
t
2
,
wh
i
ch
i
s
h
i
gh
e
r
t
h
a
n
th
e
N
M
I
of
K
-
Me
a
ns
of
51%
on
B
B
C
_500
a
n
d
da
t
a
s
e
t
B
B
C
_1328,
77%
on
B
B
C
_2225,
64%
on
B
B
C
_S
po
r
t
2.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2502
-
4752
I
n
do
n
e
s
i
a
n
J
E
l
e
c
E
n
g
&
C
o
m
p
S
c
i
,
Vo
l
.
25
,
N
o
.
1
,
J
a
n
ua
r
y
20
22
:
569
-
579
576
F
i
gur
e
5.
E
n
t
r
o
py
c
o
m
pa
r
i
s
o
n
of
KM
a
n
d
t
h
e
pr
o
p
o
s
e
d
K
M
_M
S
S
P
a
l
go
r
i
t
hm
F
i
g
ur
e
6
.
NM
I
c
o
m
pa
r
i
s
o
n
of
KM
a
n
d
t
h
e
pr
o
p
o
s
e
d
K
M
_
M
S
S
P
a
l
go
r
i
t
hm
4.
3.
5.
T
im
e
c
om
p
ar
i
s
on
of
KM
an
d
t
h
e
p
r
op
os
e
d
KM
_M
S
S
P
al
gor
it
h
m
Th
e
c
o
m
pa
r
i
s
o
n
s
of
th
e
C
P
U
t
i
m
e
on
a
l
l
da
t
a
s
e
ts
ar
e
s
h
o
wn
i
n
F
i
g
ur
e
7
.
A
c
c
o
r
d
i
ng
to
th
e
gr
a
ph
,
KM
_M
S
S
P
sp
e
n
d
s
a
v
e
ry
l
i
ttl
e
t
im
e
on
a
l
l
da
t
a
s
e
ts
t
h
a
n
K
M
.
S
o
,
o
ur
a
ppr
o
a
c
h
i
m
pr
o
v
e
s
K
M
i
n
t
e
r
m
s
o
f
C
P
U
t
i
m
e
.
F
i
gur
e
7.
T
i
m
e
c
o
m
pa
r
i
s
o
n
of
KM
a
n
d
t
h
e
pr
o
p
o
s
e
d
K
M
_
M
S
S
P
a
l
go
r
i
t
hm
4.
3.
6.
Com
p
ar
in
g
Xie
-
B
e
n
i
I
n
d
e
x
(
X
B
)
of
KM
a
n
d
t
h
e
p
r
op
os
e
d
KM
_
M
S
S
P
al
gor
it
h
m
A
co
m
pa
r
i
s
o
n
of
th
e
X
ie
-
B
e
n
i
i
nd
e
x
(
XB
)
b
e
tw
ee
n
K
M
a
n
d
K
M
_
M
SSP
i
s
d
e
p
i
ct
e
d
i
n
T
a
bl
e
2
.
I
t
i
s
p
i
c
t
ur
e
d
t
h
a
t
KM
_
M
SSP
o
u
t
p
e
r
f
o
r
m
s
K
M
i
n
t
e
r
m
s
of
th
e
XB
i
nd
e
x
i
n
a
ll
a
v
a
i
l
a
bl
e
da
t
a
s
e
t
s
.
Th
e
XB
of
K
M
_
M
SSP
i
s
3%
on
da
t
a
s
e
t
B
B
C
_1328,
7%
on
da
t
a
s
e
t
B
B
C
_2225
a
n
d
8%
on
B
B
C
_500,
wh
i
ch
i
s
l
o
w
e
r
t
h
a
n
th
e
XB
of
th
e
K
M
of
45%
on
B
B
C
_1328,
47%
on
da
t
a
s
e
t
B
B
C
_2225
a
n
d
52%
on
B
B
C
_500.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
do
n
e
s
i
a
n
J
E
l
e
c
E
n
g
&
C
o
m
p
S
c
i
I
S
S
N:
2502
-
4752
M
ax
Stabl
e
s
e
t
pr
oblem
to
f
ound
the
ini
ti
al
c
e
ntr
oids
in
c
lus
ter
ing
pr
oblem
(
A
w
ati
f
K
ar
im)
577
T
a
bl
e
2.
C
o
m
pa
r
i
n
g
X
i
e
-
B
e
ni
i
n
d
e
x
(
XB
)
of
KM
a
n
d
t
h
e
pr
o
p
o
s
e
d
K
M
_M
S
S
P
a
l
go
r
i
t
hm
D
a
ta
s
e
t
KM
K
M
_M
S
S
P
B
B
C
_500
0.52
0.08
B
B
C
_1328
0.47
0.03
B
B
C
_2225
0.45
0.07
B
B
C
_171
5
0.50
0.07
B
B
C
_S
p
or
t1
0.53
0.09
B
B
C
_S
p
or
t2
0.52
0.06
20N
G
_400
1.08
0.17
4.
3.
7.
Com
p
ar
in
g
F
u
k
u
ya
m
a
-
S
u
ge
n
o
in
d
e
x
of
KM
an
d
t
h
e
p
r
op
os
e
d
KM
_
M
S
S
P
al
gor
it
h
m
F
r
o
m
T
a
bl
e
3
w
e
c
a
n
s
ee
t
h
a
t
th
e
r
e
i
s
a
d
e
cl
i
n
e
of
F
uku
y
a
m
a
-
S
ug
e
no
i
nd
e
x
(
F
S
)
v
a
l
u
e
s
i
n
K
M
_
M
SSP
c
o
m
pa
r
e
d
w
i
th
t
h
o
s
e
of
K
M
.
Th
e
FS
i
nd
e
x
of
th
e
K
M
_
M
SSP
i
s
11%
,
wh
i
ch
i
s
l
o
w
e
r
t
h
a
n
th
e
FS
of
th
e
K
M
c
l
u
s
t
e
r
i
ng
of
17%
on
th
e
da
t
a
s
e
t
_24
.
T
h
us
K
M
_
M
SSP
o
u
t
p
e
r
f
o
r
m
s
K
M
i
n
t
e
r
m
s
of
FS
i
nd
e
x
i
n
a
l
l
a
v
a
i
l
a
bl
e
da
t
a
s
e
t
s
.
T
a
bl
e
3.
C
o
m
pa
r
i
n
g
F
uku
y
a
m
a
-
S
uge
n
o
i
n
de
x
of
KM
a
n
d
t
h
e
pr
o
p
o
s
e
d
K
M
_M
S
S
P
a
l
go
r
i
t
hm
D
a
ta
s
e
t
KM
K
M
_M
S
S
P
B
B
C
_500
0.87
0.42
B
B
C
_1328
0.55
0.13
B
B
C
_2225
0.75
0.4
B
B
C
_1715
0.38
0.21
B
B
C
_S
p
or
t1
0.20
0.16
B
B
C
_S
p
or
t2
0.37
0.24
20N
G
_400
0.53
0.48
I
t
i
s
cl
e
a
r
ly
o
bs
e
rv
e
d
t
h
a
t
th
e
r
e
s
u
l
t
s
of
K
M
_
M
SSP
ar
e
f
a
r
b
e
tt
e
r
f
o
r
a
l
l
da
t
a
s
e
ts
i
n
t
e
r
m
s
of
e
n
t
r
o
py
,
pur
i
t
y
,
F
-
me
a
s
ur
e
,
N
M
I
,
a
n
d
C
P
U
t
i
m
e
.
K
M
_
M
SSP
b
e
g
i
ns
w
i
th
a
nu
m
b
e
r
of
c
l
us
t
e
rs
(
n
u
m
b
e
r
of
c
l
u
s
t
e
rs
o
b
t
a
i
n
e
d
by
M
SSP)
v
e
ry
c
l
o
s
e
or
e
qua
l
to
th
e
r
e
al
nu
m
b
e
r
of
c
l
a
s
s
e
s
a
n
d
s
t
a
r
t
s
w
i
th
i
n
i
t
i
al
c
l
u
s
t
e
r
c
e
nt
ro
i
ds
s
e
l
e
ct
e
d
by
M
SSP
(
wh
i
ch
gua
r
a
n
t
ee
s
an
i
nd
e
p
e
nd
e
nt
s
e
t
of
do
c
u
me
n
t
s
)
.
B
u
t
K
M
o
b
t
a
i
ns
th
e
i
n
i
t
i
al
c
e
nt
e
rs
by
us
i
ng
a
r
a
n
do
m
m
e
t
h
o
d
wh
i
ch
g
i
v
e
s
u
ns
t
a
bl
e
c
l
us
t
e
r
i
ng
r
e
s
u
l
t
s
a
n
d
too
m
a
ny
i
t
e
r
a
t
i
o
ns
,
i
t
a
f
f
e
c
t
s
th
e
qua
l
i
ty
of
c
l
u
s
t
e
r
i
ng
a
n
d
c
o
s
t
s
a
gr
e
at
a
m
o
un
t
of
t
i
m
e
dur
i
ng
th
e
pr
o
c
e
ss
of
c
l
us
t
e
r
i
n
g.
4.
4.
Com
p
ar
is
on
b
e
t
we
e
n
KM
M
S
S
P
an
d
ot
h
e
r
d
e
t
e
r
m
in
is
t
ic
m
e
t
h
od
I
n
o
r
d
e
r
to
d
em
o
n
s
t
r
a
t
e
th
e
e
ff
e
ct
i
v
e
n
e
ss
of
o
ur
a
ppr
o
a
c
h
,
w
e
co
m
p
a
r
e
o
ur
a
ppr
o
a
c
h
w
i
th
DS
K
M
[
26]
i
n
t
e
r
m
s
of
No
r
m
al
i
z
e
d
M
ut
ua
l
I
nf
o
r
m
at
i
on
N
MI
on
da
t
a
s
e
t
B
B
C
_2225.
F
r
o
m
T
a
bl
e
4
,
th
e
co
m
pa
r
i
s
o
n
b
e
tw
ee
n
K
M
_
M
SSP
a
n
d
DS
K
M
wa
s
do
n
e
on
th
e
sa
m
e
nu
m
b
e
r
of
c
l
us
t
e
rs
f
o
un
d
e
d
by
M
SSP
(6
f
o
r
Da
t
a
s
e
t
B
B
C
_2225
(s
ee
T
a
bl
e
2)
)
.
Th
e
N
M
I
of
K
é
²&
i
s
gr
e
at
e
r
t
h
a
n
th
e
DSK
M
a
l
go
r
i
th
m
.
T
a
bl
e
4.
C
o
m
pa
r
i
n
g
c
l
us
t
e
r
i
n
g
NM
I
s
c
o
r
e
b
e
t
we
e
n
K
M
_
M
S
S
P
a
n
d
DSKM
on
da
t
a
s
e
t
B
B
C
_2225
a
nd
n
u
m
be
r
of
c
l
u
s
t
e
r
s
e
qua
l
to
n
u
m
b
e
r
f
o
un
de
d
by
M
S
S
P
(
k=
6
)
M
e
th
o
ds
N
M
I
D
S
K
M
0.681
K
M
_M
S
S
P
0.78
5.
CONC
L
US
I
ON
T
hi
s
r
e
s
e
a
r
c
h
a
im
s
to
d
e
v
e
l
o
p
a
me
t
h
o
d
f
o
r
d
e
t
e
r
mi
n
i
ng
a
uto
m
at
i
c
a
l
ly
b
o
t
h
,
nu
m
b
e
r
of
cl
us
t
e
rs
a
n
d
i
n
i
t
i
al
c
l
us
t
e
r
c
e
n
t
r
o
i
ds
wh
i
ch
ar
e
th
e
b
a
c
kgr
o
un
d
k
n
o
w
l
e
dg
e
i
n
c
l
a
s
s
i
c
K
-
Me
a
ns
or
ot
h
e
r
me
t
h
o
ds
of
c
l
u
s
t
e
r
i
n
g.
To
ach
ie
v
e
th
i
s
go
a
l
a
n
d
o
b
t
a
i
n
th
e
h
i
gh
e
st
qua
l
i
ty
of
c
l
us
t
e
r
i
n
g,
th
e
M
SSP
i
s
g
e
n
e
r
a
l
i
z
e
d
to
a
pply
to
th
e
c
l
u
s
t
e
r
i
ng
of
t
e
xt
d
o
c
u
me
n
ts
us
i
ng
C
HN
.
Th
e
me
t
h
o
d
i
s
i
nd
e
p
e
nd
e
nt
of
a
ny
c
l
u
s
t
e
r
i
ng
me
t
h
o
d
t
h
a
t
s
t
a
r
t
s
w
i
th
k
c
e
n
t
r
o
i
d
s
.
E
xp
e
r
im
e
n
t
a
l
r
e
s
u
l
t
s
s
h
o
w
t
h
a
t
o
ur
me
t
h
o
d
c
a
n
e
ff
e
ct
i
v
e
ly
f
i
nd
th
e
o
pt
i
m
a
l
nu
m
b
e
r
of
c
l
u
s
t
e
r
s
,
f
i
nd
a
c
o
r
r
e
ct
s
e
t
of
c
e
n
t
r
o
i
ds
,
o
b
t
a
i
n
b
e
tt
e
r
c
l
u
s
t
e
r
i
ng
r
e
s
u
l
t
s
i
n
a
s
h
o
r
t
t
ime
,
a
n
d
a
l
s
o
a
l
a
r
g
e
nu
m
b
e
r
of
do
c
u
me
n
t
s
can
b
e
e
as
i
ly
h
a
n
d
l
e
d.
To
d
em
o
n
s
t
r
a
t
e
th
e
e
ff
i
c
ie
n
c
y
of
o
ur
a
ppr
o
a
c
h
,
w
e
co
m
pa
r
e
c
l
a
s
s
i
c
K
-
Me
a
n
s
wh
i
ch
us
e
s
r
a
n
do
m
c
e
n
t
r
o
i
ds
w
i
t
h
K
M
_
M
SSP
wh
i
ch
s
t
a
r
t
s
by
K
c
e
n
t
r
o
i
ds
f
o
u
n
d
by
M
SSP,
us
i
ng
s
i
x
v
a
l
i
d
i
ty
me
a
s
ur
e
s:
F
-
me
a
s
ur
e
,
pur
i
t
y
,
e
n
t
r
o
py
,
XB
,
FS
a
n
d
N
MI
.
T
h
us
,
K
M
_
M
SSP
o
u
t
p
e
r
f
o
r
m
s
K
M
i
n
a
l
l
a
v
a
i
l
a
bl
e
da
t
a
s
e
t
s
,
f
o
r
e
xa
m
pl
e
,
i
n
da
t
a
s
e
t
B
B
C
_1328
th
e
pur
i
ty
a
n
d
th
e
F
-
me
a
s
ur
e
ar
e
94%
,
N
MI
i
s
78%
,
e
n
t
r
o
py
i
s
22%
a
n
d
t
ime
i
s
1
.
63s
wh
i
ch
i
s
h
i
gh
e
r
t
h
a
n
K
-
Me
a
n
s
of
th
e
pur
i
ty
a
n
d
th
e
F
-
me
a
s
ur
e
ar
e
81%
,
N
MI
i
s
51%
,
e
n
t
r
o
py
i
s
53%
a
n
d
t
ime
i
s
5
.
01s
.
Th
e
f
ie
ld
of
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2502
-
4752
I
n
do
n
e
s
i
a
n
J
E
l
e
c
E
n
g
&
C
o
m
p
S
c
i
,
Vo
l
.
25
,
N
o
.
1
,
J
a
n
ua
r
y
20
22
:
569
-
579
578
M
SSP
i
s
st
i
ll
op
e
n
to
m
a
ny
c
h
a
ll
e
ng
e
s
t
h
a
t
pr
o
v
i
d
e
f
ut
ur
e
s
c
o
p
e
f
o
r
im
pr
o
v
e
m
e
nt
i
n
th
e
do
c
u
me
nt
-
c
l
us
t
e
r
i
n
g
pr
o
bl
em
.
Th
e
f
ut
ur
e
wo
r
k
i
n
c
l
ud
e
s:
(
i
)
a
pp
ly
o
ur
a
pp
r
o
a
c
h
to
ot
h
e
r
a
r
e
a
s
,
(
ii
)
Im
pr
o
v
e
C
HN
by
co
m
b
i
n
i
ng
i
t
w
i
th
a
me
t
a
h
e
ur
i
st
i
c
to
o
b
t
a
i
n
a
g
l
o
b
a
l
mi
n
im
u
m
,
a
nd
(
iii
)
Us
e
a
d
ee
p
l
e
a
r
n
i
ng
c
o
n
c
e
pt
to
i
m
pr
o
v
e
t
h
e
pr
e
-
pr
o
c
e
s
s
i
n
g
s
t
e
p
.
R
E
F
E
R
E
NC
E
S
[
1]
J.
H
a
n,
J.
P
e
i
,
a
nd
M.
K
a
mb
e
r
,
“
D
at
a
M
in
in
g
C
onc
e
pt
s
and
T
e
c
hni
que
s
,
”
M
o
r
ga
n
K
a
u
f
f
ma
n
P
ubl
is
h
e
r
s
,
3r
d
e
d,
U
S
A
,
2012.
[
2]
D.
L
a
r
o
s
e
,
“
D
is
c
ov
e
r
in
g
k
now
le
dge
in
dat
a:
an
in
tr
oduc
ti
on
to
dat
a
m
in
in
g
,”
J
o
hn
W
il
e
y
&
S
o
ns
,
I
nc
.,
H
o
b
o
k
e
n,
N
e
w
J
e
r
s
e
y
,
2014.
[
3]
M
.
D
umo
nt
,
P
.
-
A
.
R
e
ni
nge
r
,
A
.
P
r
y
e
t,
G
.
M
a
r
te
l
e
t,
B
.
A
una
y
,
a
nd
J
.
-
L
.
J
o
in
,
“A
ggl
o
m
e
r
a
ti
ve
hi
e
r
a
r
c
hi
c
a
l
c
lu
s
t
e
r
in
g
of
a
ir
b
o
r
n
e
e
l
e
c
t
r
o
ma
gne
ti
c
da
ta
f
o
r
mul
t
i
-
s
c
a
l
e
g
e
o
l
o
gi
c
a
l
s
tu
di
e
s
,”
J
our
nal
of
A
ppl
ie
d
G
e
ophy
s
ic
s
,
v
o
l.
157,
pp.
1
-
9,
O
c
t.
2
018,
do
i:
10.1016/j
.
ja
ppge
o
.2018.06.020
.
[
4]
S
. B
a
nd
y
a
pa
dh
y
a
y
a
nd
K
. R
. V
a
r
a
da
r
a
ja
n, “
O
n v
a
r
ia
nt
s
of
k
-
m
e
a
ns
c
lu
s
te
r
in
g,”
ar
X
iv
pr
e
pr
in
t
ar
X
iv
:
1512.02985
, 2015.
[
5]
M
d.
A.
B.
S
id
di
que
,
R
.
B
.
A
r
i
f
,
M.
M.
R.
K
ha
n
,
a
nd
Z.
A
s
hr
a
f
i,
“
I
mpl
e
m
e
nt
a
ti
o
n
of
F
u
z
z
y
C
-
M
e
a
ns
a
nd
P
o
s
s
ib
il
is
ti
c
C
-
M
e
a
ns
C
lu
s
te
r
in
g
A
lg
o
r
it
hms
,
C
lu
s
te
r
T
e
nd
e
nc
y
A
na
l
y
s
is
a
nd
C
lu
s
te
r
V
a
li
da
ti
o
n,”
P
r
e
pr
in
ts
2018,
d
o
i:
10.20944/pr
e
pr
in
ts
201811.0581.v
1.
[
6]
H.
S
.
P
a
r
k
a
nd
C.
H
.
J
un,
“A
s
im
pl
e
a
nd
f
a
s
t
a
lg
o
r
it
hm
f
o
r
K
-
me
d
o
i
ds
c
lu
s
te
r
in
g,”
E
x
pe
r
t
Sy
s
tA
ppl
,
v
o
l.
36
,
no
.
2
,
pp.
3336
-
3341,
2009
, do
i:
10.1016/j
.
e
s
w
a
.2008.01.039
.
[
7]
L.
K
a
u
f
ma
n
a
nd
P.
J.
R
o
us
s
e
e
uw
,
“
F
in
di
ng
gr
o
ups
in
da
ta
:
An
in
tr
o
du
c
ti
o
n
to
c
lu
s
te
r
a
na
l
y
s
is
,”
W
il
e
y
,
H
obok
e
n
,
vo
l.
344,
20
08,
do
i:
10.1002/978047031
680
1.
[
8]
R.
T.
Ng
a
nd
J.
H
a
n,
“
C
la
r
a
ns
:
a
me
th
o
d
f
o
r
c
lu
s
t
e
r
in
g
o
bj
e
c
t
s
f
or
s
pa
ti
a
l
da
ta
mi
ni
ng,”
I
E
E
E
T
r
ans
ac
ti
ons
on
K
now
le
dge
and
D
at
a
E
ngi
ne
e
r
in
g
,
vo
l.
14,
n
o
.
5,
pp.
1003
-
1016,
S
e
p.
2002,
d
o
i
:
10.1109/
T
K
D
E
.2002.1033770
.
[
9]
J.
W
u,
“
A
dv
a
nc
e
s
in
K
-
M
e
a
ns
C
lu
s
te
r
in
g
:
A
D
a
ta
M
in
in
g
T
hi
n
ki
ng,”
Spr
in
ge
r
T
he
s
e
s
,
S
pr
in
ge
r
B
e
r
li
n
H
e
id
e
lb
e
r
g,
2012
.
[
10]
W.
A
s
ho
ur
a
nd
C.
F
y
f
e
,
“
I
mpr
ov
in
g
B
r
e
gma
n
K
-
M
e
a
ns
,”
I
nt
e
r
nat
io
nal
J
our
nal
of
D
at
a
M
in
in
g,
M
ode
ll
in
g
and
M
anage
m
e
nt
,
vo
l.
6,
no
.
1,
pp.
65
-
82,
J
a
n.
2014,
d
o
i:
10.1504/I
J
D
M
M
M
.
2014.059981.
[
11]
L.
I.
K
un
c
h
e
v
a
a
nd
J.
C.
B
e
z
d
e
k,
“
S
e
l
e
c
ti
o
n
of
c
lu
s
t
e
r
pr
ot
ot
y
p
e
s
f
r
o
m
da
ta
by
a
ge
n
e
ti
c
a
lg
o
r
i
th
m,”
in
:
P
r
oc
.
5t
h
E
ur
ope
an
C
ongr
e
s
s
on
I
nt
e
ll
ig
e
nt
T
e
c
hni
que
s
and
So
f
t
C
om
put
in
g
(
E
U
F
I
T
)
,
A
a
c
h
e
n,
G
e
r
ma
n
y
,
vo
l.
18
,
1997
,
pp.
1683
-
1688
.
[
12]
J.
C.
B
e
z
d
e
k
a
nd
R.
J.
H
a
th
a
w
a
y
,
“
O
pt
im
iz
a
ti
o
n
of
f
u
z
z
y
c
lu
s
t
e
r
in
g
c
r
it
e
r
ia
us
in
g
ge
n
e
ti
c
a
lg
or
it
hms
,”
in
P
r
oc
.
F
ir
s
t
I
E
E
E
C
onf
.
E
v
ol
ut
io
nar
y
C
om
put
at
io
n,
P
is
c
at
aw
ay
,
N
J
:
I
E
E
E
P
r
e
s
s
,
1994,
pp.
589
-
594
, do
i
:
10.1109/I
C
E
C
.1994.349993
.
[
13]
B.
K
ö
v
e
s
i,
J.
-
M.
B
o
u
c
h
e
r
,
a
nd
S.
S
a
o
udi
,
“
S
t
oc
ha
s
ti
c
K
-
me
a
ns
a
lg
o
r
it
h
m
f
o
r
ve
c
t
o
r
qua
nt
i
z
a
ti
o
n,”
P
at
te
r
n
R
e
c
ogni
t.
L
e
tt
.,
vo
l.
22,
no
.
6
-
7
,
pp.
603
-
610,
2001,
d
o
i:
10.1016/s
0167
-
8655(
01
)
00021
-
6.
[
14]
M.
S
a
r
ka
r
,
B.
Y
e
gna
na
r
a
y
a
na
,
a
nd
D.
K
h
e
ma
ni
a
l,
“A
c
lu
s
t
e
r
in
g
a
lg
or
it
hm
us
in
g
an
e
vol
ut
i
o
n
a
r
y
-
ba
s
e
d
a
pp
r
o
a
c
h,”
P
at
te
r
n
R
e
c
ogni
ti
on
L
e
tt
e
r
s
,
v
ol
.
18,
no
.
10,
pp.
975
-
986,
1997,
d
o
i:
10.
1016/S
0167
-
8655(
97)
00122
-
0.
[
15]
D.
J.
K
e
t
c
h
e
n
a
nd
C.
l.
S
h
oo
k
,
“
T
he
a
ppl
i
c
a
ti
o
n
of
c
lu
s
te
r
a
na
l
y
s
is
in
s
tr
a
te
gi
c
ma
na
g
e
m
e
nt
r
e
s
e
a
r
c
h:
an
a
na
l
y
s
is
a
nd
c
r
it
iq
u
e
,”
St
r
a
te
gi
c
M
anage
m
e
nt
J
our
nal
,
v
o
l.
17,
n
o
.
6,
pp.
441
-
458
,
J
un
1996,
do
i:
10.1002/
(
s
ic
i
)
1097
-
0266(
199606
)
17:
6<
441:
:a
id
-
s
mj
819>
3.0.c
o
;2
-
g
.
[
16]
A.
J
.
-
G
a
r
c
ia
a
nd
W.
G
.
-
F
l
or
e
s
,
“
A
ut
o
ma
ti
c
c
lu
s
t
e
r
in
g
us
in
g
n
a
tu
r
e
-
in
s
pi
r
e
d
me
ta
h
e
ur
is
ti
c
s
:
A
s
ur
ve
y
,”
A
ppl
ie
d
Sof
t
C
om
put
in
g
,
vo
l.
41,
pp.
192
-
213,
A
p
r
.
2016,
d
o
i:
10.1016/j
.a
s
oc
.2015.12.00
1.
[
17]
S.
D
a
s
,
A.
A
br
a
ha
m
,
a
nd
A.
K
o
na
r
,
“
S
pa
ti
a
l
in
f
o
r
ma
ti
o
n
-
ba
s
e
d
im
a
ge
s
e
gme
nt
a
ti
o
n
us
in
g
a
mo
di
f
ie
d
pa
r
ti
c
l
e
s
w
a
r
m
o
pt
im
iz
a
t
io
n
a
lg
o
r
it
h
m,”
I
SD
A
'
06:
P
r
oc
e
e
di
ngs
of
th
e
Si
x
th
I
nt
e
r
nat
io
nal
C
onf
e
r
e
nc
e
on
I
nt
e
ll
ig
e
nt
S
y
s
te
m
s
D
e
s
ig
n
and
A
ppl
ic
at
io
ns
,
O
c
t.
vo
l.
2,
2006,
pp.
438
-
444,
d
oi
:
10.1109/I
S
D
A
.2006.253877
.
[
18]
R.
K
u
o
,
Y.
S
y
u,
Z.
-
Y.
C
h
e
n
,
a
nd
F.
-
C.
T
i
e
n,
“
I
nt
e
gr
a
ti
o
n
of
pa
r
ti
c
l
e
s
w
a
r
m
o
pt
im
iz
a
ti
o
n
a
nd
g
e
n
e
ti
c
a
lg
o
r
it
h
m
f
o
r
d
y
na
m
ic
c
lu
s
te
r
in
g,”
I
nf
or
m
at
io
n
Sc
ie
nc
e
s
,
v
o
l.
195,
pp.
124
-
140,
J
ul
.
2012,
do
i:
10.1016/j
.
in
s
.2012.01.021.
[
19]
X.
L.
X
i
e
a
nd
G.
B
e
n
i,
“A
v
a
li
di
t
y
m
e
a
s
ur
e
f
o
r
f
u
z
z
y
c
lu
s
t
e
r
in
g,”
I
E
E
E
T
r
ans
.
P
at
te
r
n
A
nal
.
M
ac
h.
I
nt
e
ll
.,
vo
l.
13,
n
o.
8,
pp.
841
-
847,
A
ug.
1991,
do
i:
10.1109
/3
4.85677.
[
20]
C.
S
ubba
la
ks
hmi
,
G.
K
r
is
hna
,
K.
R
a
o
,
a
nd
P.
R
a
o
,
“A
me
th
o
d
to
f
in
d
o
pt
im
um
numbe
r
of
c
lu
s
te
r
s
ba
s
e
d
on
f
u
z
z
y
s
il
h
o
u
e
tt
e
on
d
y
na
mi
c
da
ta
s
e
t,
”
P
r
oc
e
di
a
C
om
put
e
r
Sc
ie
nc
e
,
2015,
v
o
l.
46,
pp.
346
-
353,
do
i
:
10.1016/j
.pr
oc
s
.2015.02.030.
[
21]
X.
T
o
ng,
F.
M
e
ng
,
a
nd
Z.
W
a
ng,
“
O
pt
im
i
z
a
ti
o
n
to
K
-
M
e
a
ns
i
ni
ti
a
l
c
lu
s
t
e
r
c
e
nt
e
r
s
,”
C
om
put
e
r
E
ngi
ne
e
r
in
g
and
D
e
s
ig
n
,
vo
l.
32,
no
.
8,
pp.
2721
-
2723,
2011.
[
22]
Q.
S
.
-
Y
i,
L.
H
.
-
H
ui
,
a
nd
L.
D
.
-
Y
i,
“
R
e
s
e
a
r
c
h
a
nd
A
ppl
ic
a
ti
on
of
I
mp
r
ov
e
d
K
-
M
e
a
ns
A
lg
o
r
it
hm
in
T
e
x
t
C
lu
s
te
r
i
ng,”
D
E
S
te
c
h
T
r
ans
ac
ti
ons
on
C
o
m
put
e
r
Sc
ie
n
c
e
and
E
ngi
ne
e
r
in
g
(
pc
m
m
)
,
J
un.
2018,
do
i:
10.12783/dt
c
s
e
/p
c
mm
2018/23653.
[
23]
S.
K
a
r
ol
a
nd
V.
M
a
nga
t,
“
E
v
a
lu
a
ti
o
n
of
t
e
x
t
d
oc
um
e
nt
c
lu
s
te
r
i
ng
a
ppr
o
a
c
h
ba
s
e
d
on
pa
r
ti
c
le
s
w
a
r
m
o
p
ti
mi
z
a
ti
o
n,”
P
r
oc
e
e
di
ngs
of
C
e
nt
r
al
E
ur
ope
an
J
ou
r
nal
of
C
om
put
e
r
S
c
ie
nc
e
,
v
o
l.
3,
n
o
.
2
, J
un. 2013
,
pp.
69
-
90,
do
i
:
10.2478/s
13537
-
013
-
0104
-
2.
[
24]
J.
S
o
ng,
X.
Li
,
a
nd
Y.
L
iu
,
“
A
n
O
pt
im
i
z
e
d
K
-
M
e
a
ns
A
lg
or
it
h
m
f
o
r
S
e
l
e
c
ti
ng
I
n
it
ia
l
C
lu
s
te
r
in
g
C
e
nt
e
r
s
,”
I
nt
e
r
nat
io
nal
J
our
nal
of
Se
c
ur
it
y
and
I
ts
A
ppl
ic
at
io
ns
,
v
o
l.
9,
n
o
.
10,
pp.
177
-
186,
2015,
do
i:
10.14257/i
js
ia
.2015.9.10.16.
[
25]
M.
H
ua
ng,
Z.
S.
H
e
,
X.
L.
X
in
g,
a
nd
Y.
C
he
n,
“
N
e
w
K
-
M
e
a
ns
c
lu
s
te
r
in
g
c
e
nt
e
r
s
e
le
c
t
a
lg
o
r
it
h
m,”
C
om
put
e
r
E
ngi
ne
e
r
in
g
and
A
ppl
ic
at
io
ns
,
vo
l.
47,
n
o
.
35,
pp.
132
-
134,
2011.
[
26]
E.
S
he
r
ka
t,
J.
V
e
lc
in
,
a
nd
E.
E.
M
il
i
o
s
,
“
F
a
s
t
a
nd
S
im
pl
e
D
e
t
e
r
mi
ni
s
ti
c
S
e
e
di
ng
of
K
-
M
e
a
ns
f
o
r
T
e
x
t
D
oc
um
e
nt
C
lu
s
te
r
in
g,”
9t
h
C
onf
e
r
e
nc
e
and
L
abs
of
th
e
E
v
al
uat
io
n
F
or
um
(
C
L
E
F
)
,
P
r
oc
e
e
di
ngs
,
S
e
p.
2018,
pp.
76
-
88,
d
o
i:
10.1007/978
-
3
-
319
-
98932
-
7_
7.
[
27]
A.
K
a
r
im
,
C.
Lo
qma
n
,
a
nd
J.
B
o
umhi
d
i,
“
D
e
t
e
r
mi
ni
ng
th
e
N
umbe
r
of
C
lu
s
t
e
r
s
us
in
g
N
e
ur
a
l
N
e
tw
o
r
k
a
nd
M
a
x
S
ta
bl
e
S
e
t
P
r
o
bl
e
m,”
T
he
1s
t.
I
nt
.
C
onf
.
On
I
nt
e
ll
ig
e
nt
C
om
put
in
g
in
D
at
a
Sc
ie
nc
e
s
.
P
r
oc
e
di
a
C
om
put
e
r
Sc
ie
nc
e
,
2018,
vo
l.
127,
pp.
16
-
25,
do
i
:
10.1016/j
.p
r
oc
s
.2018.01.0
93.
[
28]
M.
R.
A
li
gul
i
y
e
v
,
“
C
lu
s
te
r
in
g
of
d
oc
um
e
nt
c
o
ll
e
c
t
i
o
n,
A
w
e
ig
ht
in
g
a
ppr
o
a
c
h,”
E
x
pe
r
t
Sy
s
te
m
s
w
it
h
A
ppl
ic
at
io
ns
,
v
o
l.
36,
no.
4,
pp.
7904
-
7916,
2009,
do
i
:
10.1016/j
.
e
s
w
a
.2008.11.017.
[
29]
Y.
F
uku
y
a
ma
a
nd
M.
S
uge
no
,
“A
ne
w
me
th
o
d
of
c
h
oo
s
in
g
th
e
numbe
r
of
cl
us
t
e
r
s
f
or
th
e
f
u
z
z
y
c
-
me
a
ns
me
th
o
d,”
in
:
P
r
oc
.
5t
h
F
uz
z
y
Sy
s
t.
Sy
m
p
.,
J
a
n.
1989,
pp.
247
-
250.
[
30]
Y.
C
hung
a
nd
M.
D
e
ma
nge
,
“
T
h
e
0
-
1
in
ve
r
s
e
ma
x
im
um
s
ta
bl
e
s
e
t
pr
o
bl
e
m,”
D
is
c
r
e
t
e
A
ppl
ie
d
M
at
he
m
at
ic
s
,
v
o
l.
156,
no
.
13,
pp.
2501
-
2516,
2008,
do
i
:
10.1016/j
.da
m.
2008.03.015.
[
31]
J.
H
o
p
f
i
e
ld
a
nd
D.
T
a
nk,
“
N
e
u
r
a
l
c
o
mput
a
ti
o
n
of
d
e
c
is
i
o
ns
in
o
pt
im
iz
a
ti
o
n
p
r
o
bl
e
ms
,”
B
io
lo
gi
c
al
C
y
be
r
ne
ti
c
s
,
v
o
l.
52,
n
o.
3,
pp.
1
-
25,
1985,
do
i
:
10.1007/B
F
00339943.
Evaluation Warning : The document was created with Spire.PDF for Python.