I
nte
rna
t
io
na
l J
o
urna
l o
f
E
lect
rica
l a
nd
Co
m
p
ute
r
E
ng
in
ee
ring
(
I
J
E
CE
)
Vo
l.
9
,
No
.
4
,
A
u
g
u
s
t
201
9
,
p
p
.
2
7
6
0
~2
7
7
0
I
SS
N:
2
0
8
8
-
8708
,
DOI
: 1
0
.
1
1
5
9
1
/
i
j
ec
e
.
v9
i
4
.
p
p
2
7
6
0
-
2770
2760
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ia
e
s
co
r
e
.
co
m/
jo
u
r
n
a
ls
/in
d
ex
.
p
h
p
/
I
JE
C
E
Fu
zz
y
clus
tering
a
nd
Fu
zz
y
C
-
M
ea
ns
pa
rtit
io
n
cluste
r ana
ly
sis
a
nd va
lida
tion stu
dies o
n a sub
set
of
Cit
eS
co
re
da
tas
et
K
.
Va
ra
da
Ra
j
k
u
m
a
r
1
,
A
di
m
u
la
m
Yes
ub
a
b
u
2
,
K
.
Su
bra
h
m
a
ny
a
m
3
1,
3
De
p
a
rtm
e
n
t
o
f
Co
m
p
u
ter S
c
ien
c
e
a
n
d
En
g
in
e
e
ri
n
g
,
Ko
n
e
r
u
L
a
k
s
h
m
a
iah
Ed
u
c
a
ti
o
n
F
o
u
n
d
a
ti
o
n
,
I
n
d
ia
2
De
p
a
rtme
n
t
o
f
Co
m
p
u
ter S
c
ien
c
e
a
n
d
E
n
g
in
e
e
rin
g
,
S
ir
C
R
Re
d
d
y
Co
ll
e
g
e
o
f
En
g
in
e
e
rin
g
,
In
d
ia
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
A
p
r
2
3
,
2
0
1
8
R
ev
i
s
e
d
Dec
1
5
,
2
0
1
8
A
cc
ep
ted
Ma
r
1
1
,
2
0
1
9
A
h
a
rd
p
a
rti
ti
o
n
c
l
u
ste
rin
g
a
lg
o
r
i
th
m
a
ss
i
g
n
s
e
q
u
a
ll
y
d
istan
t
p
o
i
n
t
s
to
o
n
e
o
f
th
e
c
lu
ste
rs,
w
h
e
re
e
a
c
h
d
a
tu
m
h
a
s
th
e
p
ro
b
a
b
il
i
ty
to
a
p
p
e
a
r
in
si
m
u
lt
a
n
e
o
u
s
a
ss
ig
n
m
e
n
t
to
f
u
rth
e
r
c
lu
ste
rs.
T
h
e
f
u
z
z
y
c
lu
ste
r
a
n
a
l
y
sis a
ss
i
g
n
s
m
e
m
b
e
rsh
ip
c
o
e
ff
icie
n
ts
o
f
d
a
ta
p
o
in
ts
w
h
ich
a
re
e
q
u
id
istan
t
b
e
tw
e
e
n
t
w
o
c
lu
ste
rs
so
th
e
in
f
o
rm
a
ti
o
n
d
irec
ts
h
a
v
e
a
p
lac
e
to
wa
rd
in
e
x
c
e
ss
o
f
o
n
e
c
lu
ste
r
in
th
e
m
e
a
n
ti
m
e
.
F
o
r
a
s
u
b
se
t
o
f
Cit
e
S
c
o
re
d
a
tas
e
t,
f
u
z
z
y
c
lu
ste
rin
g
(f
a
n
n
y
)
a
n
d
f
u
z
z
y
c
-
m
e
a
n
s
(fc
m
)
a
l
g
o
rit
h
m
s
w
e
r
e
i
m
p
le
m
e
n
ted
to
stu
d
y
th
e
d
a
ta
p
o
in
ts
th
a
t
li
e
e
q
u
a
l
ly
d
istan
t
f
ro
m
e
a
c
h
o
th
e
r.
Be
f
o
re
a
n
a
l
y
sis,
c
lu
ste
ra
b
i
li
ty
o
f
th
e
d
a
tas
e
t
w
a
s
e
v
a
lu
a
ted
w
it
h
Ho
p
k
in
s
sta
ti
stic.
T
h
e
o
p
ti
m
a
l
c
lu
ste
rs
w
e
r
e
d
e
term
in
e
d
u
sin
g
Nb
Cl
u
st
p
a
c
k
a
g
e
,
w
h
e
re
it
is
e
v
id
e
n
c
e
d
th
a
t
9
v
a
rio
u
s
in
d
ice
s
p
r
o
p
o
se
d
3
c
lu
ste
r
s
o
lu
t
io
n
s
a
s
b
e
st
c
lu
ste
rs.
F
u
rth
e
r,
a
p
p
r
o
p
riate
v
a
lu
e
o
f
f
u
z
z
in
e
ss
p
a
ra
m
e
ter
m
w
a
s e
v
a
lu
a
ted
to
d
e
term
in
e
th
e
d
ist
rib
u
ti
o
n
o
f
m
e
m
b
e
rsh
ip
v
a
lu
e
s
w
it
h
v
a
riatio
n
in
m
f
ro
m
1
to
2
.
Co
e
f
f
icie
n
t
o
f
v
a
riatio
n
(CV
),
a
lso
k
n
o
w
n
a
s
re
lativ
e
v
a
ri
a
b
il
it
y
w
a
s
e
v
a
lu
a
ted
to
stu
d
y
th
e
sp
re
a
d
o
f
d
a
ta.
T
h
e
ti
m
e
c
o
m
p
lex
it
y
o
f
f
u
z
z
y
c
lu
ste
rin
g
(
f
a
n
n
y
)
a
n
d
f
u
z
z
y
c
-
m
e
a
n
s
a
lg
o
rit
h
m
s
w
e
r
e
e
v
a
lu
a
ted
b
y
k
e
e
p
in
g
d
a
ta
p
o
in
ts
c
o
n
sta
n
t
a
n
d
v
a
ry
in
g
n
u
m
b
e
r
o
f
c
lu
ste
rs.
K
ey
w
o
r
d
s
:
C
ites
co
r
e
C
lu
s
ter
i
n
g
Fan
n
y
FC
M
Fu
zz
y
c
-
m
ea
n
s
T
im
e
co
m
p
le
x
it
y
Valid
atio
n
Co
p
y
rig
h
t
©
2
0
1
x
In
sti
tu
te
o
f
A
d
v
a
n
c
e
d
E
n
g
i
n
e
e
rin
g
a
n
d
S
c
ien
c
e
.
Al
l
rig
h
ts re
se
rv
e
d
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
K
.
Var
ad
a
R
aj
k
u
m
ar
,
Dep
ar
t
m
en
t o
f
C
o
m
p
u
ter
Scie
n
ce
an
d
E
n
g
i
n
ee
r
in
g
,
Sir
C
R
R
ed
d
y
C
o
lle
g
e
o
f
E
n
g
in
e
er
in
g
,
E
lu
r
u
,
W
G
Dt,
A
n
d
h
r
a
P
r
ad
e
s
h
,
I
n
d
ia
.
Em
ail:
v
ar
ad
ar
aj
k
u
m
ar
1
8
@
g
m
ail.
co
m
1.
I
NT
RO
D
UCT
I
O
N
C
lu
s
ter
a
n
al
y
s
i
s
is
a
m
et
h
o
d
to
class
i
f
y
d
ata
f
o
r
a
g
iv
e
n
s
et
o
f
o
b
j
ec
ts
in
to
a
s
et
o
f
cla
s
s
e
s
b
ased
o
n
s
i
m
ilar
f
ea
tu
r
e
s
.
Ma
i
n
g
o
al
o
f
clu
s
ter
in
g
is
to
d
iv
id
e
t
h
e
d
a
taset
i
n
to
cl
u
s
ter
s
r
ep
r
esen
t
in
g
s
i
m
ilar
o
b
j
ec
ts
to
s
a
m
e
cl
u
s
ter
an
d
d
is
s
i
m
ilar
o
b
j
ec
ts
to
d
if
f
er
en
t
cl
u
s
ter
s
[
1
]
.
Gr
ad
u
al
m
e
m
b
er
s
h
ip
s
o
f
d
at
a
p
o
in
ts
to
clu
s
ter
s
in
[
0
,
1
]
ca
n
b
e
as
s
i
g
n
ed
b
y
u
s
i
n
g
F
u
zz
y
cl
u
s
ter
a
n
al
y
s
i
s
.
Ass
i
g
n
in
g
d
ata
p
o
i
n
ts
w
h
ic
h
b
elo
n
g
to
m
a
n
y
n
u
m
b
er
s
o
f
cl
u
s
ter
s
(
m
o
r
e
th
a
n
o
n
e
at
a
g
i
v
en
p
o
in
t
o
f
ti
m
e)
w
ill
b
e
th
e
r
es
u
lt.
T
r
ad
itio
n
al
clu
s
ter
i
n
g
m
et
h
o
d
s
d
o
n
o
t
r
ep
r
esen
t
th
i
s
f
ea
t
u
r
e
an
d
en
tail
s
u
n
ce
r
tain
t
y
w
it
h
wh
ich
d
ata
ar
e
ass
i
g
n
ed
to
th
e
d
if
f
er
e
n
t
clas
s
es
[
2
]
.
A
s
u
p
er
io
r
s
tan
d
ar
d
o
f
d
etail
o
f
t
h
e
d
ata
m
o
d
el
i
s
o
f
f
er
ed
b
y
th
i
s
cl
u
s
ter
a
n
al
y
s
i
s
.
T
h
e
d
etai
ls
ab
o
u
t
th
e
ex
te
n
t
o
f
a
m
b
i
g
u
it
y
o
r
d
e
f
in
i
ten
e
s
s
to
w
h
ic
h
a
d
ata
p
o
in
t
b
elo
n
g
s
to
a
clu
s
ter
ca
n
b
e
f
o
u
n
d
b
y
t
h
i
s
m
e
m
b
er
s
h
ip
d
eg
r
ee
.
Su
c
h
m
e
m
b
er
s
h
ip
d
eg
r
ee
s
ar
e
v
alid
ated
b
y
th
e
i
n
ter
p
r
etatio
n
o
f
f
u
zz
y
s
ets
[
3
]
to
t
h
e
d
i
f
f
er
e
n
t
clu
s
ter
s
ar
e
ca
lcu
lated
f
r
o
m
th
e
s
ep
ar
atio
n
o
f
th
e
d
ata
p
o
in
t
w
it
h
th
e
clu
s
ter
ce
n
tr
e,
th
e
d
eg
r
ee
s
o
f
m
e
m
b
er
s
h
ip
to
w
h
ic
h
a
g
i
v
en
d
ata
p
o
in
t
b
elo
n
g
s
to
ca
n
b
e
ca
lc
u
lated
.
(
i.e
.
,
th
e
d
i
s
s
i
m
ilar
it
y
)
w
i
th
r
esp
ec
t
to
th
e
s
h
ap
e
an
d
th
e
s
ize
o
f
t
h
e
cl
u
s
ter
[
4
]
.
Fu
zz
y
s
ets
ar
e
f
u
n
c
tio
n
s
.
T
h
e
y
ar
e
th
e
g
en
er
aliza
tio
n
o
f
ch
ar
ac
ter
is
tic
f
u
n
cti
o
n
s
o
f
cla
s
s
ical
s
et
s
.
T
h
e
m
e
m
b
er
s
h
ip
o
f
t
h
e
s
et
with
‘
1
’
i
s
g
i
v
en
b
y
th
e
ch
ar
ac
t
er
is
tic
f
u
n
ctio
n
s
,
an
d
t
h
e
m
e
m
b
er
s
h
ip
o
f
th
e
s
et
is
0
,
if
th
e
g
iv
e
n
ele
m
en
t
i
s
’
n
t
a
m
e
m
b
er
o
f
th
e
s
et.
A
l
l
th
e
d
eg
r
ee
s
b
elo
n
g
in
g
in
[
0
,
1
]
ar
e
allo
w
ed
b
y
th
e
Fu
zz
y
s
et
s
.
T
h
e
f
ir
s
t
r
esear
c
h
er
w
h
o
p
r
o
p
o
s
ed
f
u
zz
y
s
et
s
u
s
ag
e
i
n
c
lu
s
ter
in
g
i
s
R
u
s
p
in
i
[
5
]
in
t
h
e
y
ea
r
1
9
6
9
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2
0
8
8
-
8708
F
u
z
z
y
clu
s
teri
n
g
a
n
d
F
u
z
z
y
C
-
Mea
n
s
p
a
r
titi
o
n
clu
s
ter
a
n
a
lysi
s
a
n
d
va
lid
a
tio
n
…
(
K
.
V
a
r
a
d
a
R
a
jku
ma
r
)
2761
I
n
th
e
f
o
r
m
o
f
an
o
b
j
ec
tiv
e
f
u
n
ct
io
n
,
a
b
etter
f
u
zz
y
p
ar
titi
o
n
an
d
clu
s
ter
p
r
o
to
ty
p
e
s
ar
e
f
o
u
n
d
b
y
t
h
e
Fu
zz
y
m
et
h
o
d
s
.
T
h
e
clu
s
ter
i
n
g
w
o
r
k
ca
n
th
e
n
b
e
r
ef
o
r
m
u
la
ted
as
a
f
u
n
ctio
n
o
p
ti
m
izatio
n
p
r
o
b
lem
.
T
h
e
cl
u
s
te
r
p
r
o
to
ty
p
es
an
d
t
h
e
m
e
m
b
er
s
h
ip
s
o
f
d
ata
p
o
in
ts
ar
e
t
h
e
b
asis
f
o
r
th
e
o
b
j
e
ctiv
e
f
u
n
c
t
io
n
to
th
e
clu
s
ter
s
.
Ma
n
y
f
u
zz
y
cl
u
s
ter
in
g
w
a
y
s
h
av
e
b
ee
n
p
r
o
p
o
s
ed
d
u
e
to
m
o
d
if
icatio
n
s
o
f
t
h
e
o
b
j
ec
tiv
e
f
u
n
ctio
n
s
,
w
h
ic
h
ai
m
s
at
i
m
p
r
o
v
i
n
g
t
h
e
r
es
u
lt
w
it
h
r
esp
ec
t
to
n
o
is
e,
o
u
tlier
s
et
c
[
6
]
.
Fu
zz
y
C
-
Me
an
s
(
FC
M)
clu
s
ter
i
n
g
i
s
an
u
n
s
u
p
er
v
is
ed
tech
n
iq
u
e
th
at
class
i
f
ies
t
h
e
i
m
ag
e
b
y
g
r
o
u
p
in
g
s
i
m
ilar
d
ata
p
o
in
ts
in
t
h
e
f
ea
tu
r
e
s
p
ac
e
in
to
clu
s
ter
s
[
7
]
.
Fu
zz
y
c
-
m
ea
n
s
cl
u
s
ter
i
n
g
to
f
r
a
m
e
th
e
cl
u
s
ter
s
i
n
th
e
d
ata
s
et.
T
h
e
o
p
tim
a
l
an
d
eq
u
alize
d
clu
s
ter
s
ar
e
g
en
er
ated
b
y
th
e
F
C
M
alg
o
r
ith
m
[
8
]
.
I
n
th
is
task
,
f
u
zz
y
clu
s
ter
i
n
g
alg
o
r
ith
m
f
o
llo
w
ed
b
y
f
u
zz
y
c
-
m
ea
n
s
(
FC
M)
alg
o
r
ith
m
ar
e
i
m
p
le
m
en
ted
o
n
a
s
u
b
s
et
o
f
d
ata
w
h
ich
is
d
er
iv
ed
f
r
o
m
C
i
teSco
r
e
d
atab
ase.
A
n
o
te
s
p
ec
if
y
in
g
t
h
e
n
u
m
b
er
o
f
cl
u
s
t
er
s
to
s
ea
r
ch
f
o
r
th
e
p
r
io
r
an
al
y
s
i
s
h
a
s
b
ee
n
p
r
esen
ted
.
2.
RE
S
E
ARCH
M
E
T
H
O
D
a.
Data
s
et
A
d
ataset
o
f
6
7
0
j
o
u
r
n
als
w
er
e
s
elec
ted
a
s
d
a
taset
in
t
h
is
s
t
u
d
y
.
T
h
e
d
atase
t
i
n
cl
u
d
ed
in
a
n
al
y
s
i
s
h
a
s
d
ata
r
eg
ar
d
in
g
C
iteSco
r
e,
p
er
ce
n
tile,
ci
tatio
n
co
u
n
t,
p
er
ce
n
t
cited
,
SNI
P
,
SJ
R
r
esp
ec
ti
v
el
y
.
A
p
ar
t
f
r
o
m
th
i
s
,
p
u
b
lis
h
er
d
ata
w
as
al
s
o
in
cl
u
d
ed
.
B
ef
o
r
e
f
in
alizi
n
g
d
ataset,
a
f
ilter
ed
s
ea
r
ch
w
a
s
p
er
f
o
r
m
ed
o
n
2
2
,
6
1
8
j
o
u
r
n
al
titl
es
f
o
r
w
h
ic
h
C
iteSco
r
e
w
a
s
ev
al
u
ated
.
A
p
r
eli
m
i
n
ar
y
b
r
o
ad
s
ea
r
ch
e
m
p
lo
y
i
n
g
ter
m
s
u
ch
a
s
“
C
o
m
p
u
ter
Scien
ce
”,
w
it
h
at
least
“
2
5
0
d
o
cu
m
en
ts
”
i
n
an
y
j
o
u
r
n
al
w
i
th
in
c
lu
s
io
n
o
f
o
p
en
ac
ce
s
s
ti
tles
r
esu
l
ted
in
6
7
0
j
o
u
r
n
als.
b.
C
iteSco
r
e
T
h
e
u
s
e
o
f
C
ite
Sco
r
e
is
to
m
ea
s
u
r
e
t
h
e
r
ef
er
e
n
ce
(
citati
o
n
)
i
m
p
ac
t
o
f
t
h
e
s
er
ial
t
itles
s
u
c
h
as
i
n
j
o
u
r
n
als.
Ser
ial
tit
les
p
u
b
li
s
h
a
t
u
n
if
o
r
m
i
n
ter
v
al
s
(
i.e
.
o
n
e
o
r
m
o
r
e
n
u
m
b
er
o
f
v
o
l
u
m
es
p
er
y
ea
r
)
.
C
iteSco
r
e
is
u
s
ed
to
ca
lc
u
late
t
h
e
a
v
er
ag
e
n
u
m
b
er
o
f
r
e
f
er
en
ce
s
(
citatio
n
s
)
i
n
t
h
e
j
o
u
r
n
al
p
er
y
ea
r
b
y
co
m
p
ar
in
g
it
w
ith
th
o
s
e
o
f
p
r
ec
ed
in
g
t
h
r
ee
y
ea
r
C
iteSco
r
e.
C
ite
Sco
r
e
m
etr
ic
s
ar
e
ca
lcu
lated
f
r
o
m
Sco
p
u
s
D
ata.
T
r
a
d
e
jo
u
r
n
als,
j
o
u
r
n
als,
b
o
o
k
s
er
ies
an
d
co
n
f
er
en
ce
p
r
o
ce
ed
in
g
s
ar
e
th
e
k
in
d
s
o
f
s
er
ial
d
ata
av
ailab
le
.
Ho
w
e
v
er
,
o
n
l
y
j
o
u
r
n
al
titl
es a
r
e
s
elec
ted
f
o
r
th
e
s
t
u
d
y
.
c.
Ass
es
s
i
n
g
t
h
e
cl
u
s
ter
ab
ilit
y
-
Ho
p
k
in
s
s
tati
s
tic
T
h
e
f
u
n
ctio
n
g
et_
clu
s
t_
ten
d
e
n
c
y
(
)
i
n
f
ac
to
e
x
tr
a
s
h
all
b
e
u
s
ed
to
ass
ess
w
h
et
h
er
th
e
d
at
aset
ca
n
b
e
clu
s
ter
ed
.
C
o
m
p
u
ti
n
g
Ho
p
k
i
n
s
s
tati
s
tic
i
s
a
w
a
y
t
o
ac
h
ie
v
e
th
is
.
T
h
e
‘
clu
s
ter
i
n
g
te
n
d
en
c
y
’
o
f
a
d
ataset
i
s
ass
es
s
ed
b
y
Ho
p
k
in
s
ta
tis
t
ics.
W
e
ca
n
r
ej
ec
t
t
h
e
n
u
ll
h
y
p
o
t
h
esi
s
i
f
th
e
v
al
u
e
o
f
Ho
p
k
in
s
tatis
t
ics
is
a
l
m
o
s
t
ze
r
o
.
W
e
ca
n
co
n
clu
d
e
th
at
th
e
d
ataset
is
s
i
g
n
if
ica
n
tl
y
a
cl
u
s
ter
ab
le
d
ata.
d.
Dete
r
m
i
n
i
n
g
o
p
ti
m
al
clu
s
ter
s
A
cl
u
s
ter
v
alid
it
y
p
ac
k
ag
e
n
a
m
el
y
Nb
C
l
u
s
t
w
as
u
s
ed
to
e
s
ti
m
ate
t
h
e
n
u
m
b
er
o
f
cl
u
s
ter
s
i
n
a
d
ataset.
Nb
C
lu
s
t
co
n
tai
n
s
n
ea
r
l
y
3
0
v
ar
io
u
s
v
alid
it
y
in
d
ice
s
.
T
h
e
aim
o
f
Nb
C
lu
s
t
is
to
g
ath
er
all
th
e
av
ailab
le
in
d
ices
.
Dif
f
er
en
t
clu
s
ter
s
o
f
d
ata
a
r
e
p
r
o
d
u
ce
d
b
y
d
i
f
f
er
e
n
t
cl
u
s
te
r
in
g
a
lg
o
r
it
h
m
s
.
So
m
eti
m
es,
th
e
f
i
n
al
cl
u
s
ter
i
n
g
p
ar
titi
o
n
s
m
a
y
v
ar
y
g
r
ea
tl
y
e
v
en
w
it
h
th
e
s
a
m
e
alg
o
r
it
h
m
h
av
in
g
a
v
ar
iatio
n
i
n
t
h
e
p
r
esen
tatio
n
o
r
d
er
o
f
th
e
d
ata
o
r
p
ar
am
eter
s
.
T
h
er
e
ca
n
b
e
an
y
n
u
m
b
er
o
f
c
lu
s
ter
s
in
t
h
e
r
an
g
e
2
-
8.
T
h
e
d
is
tan
ce
m
etr
ic
i
s
s
et
to
"
eu
clid
ea
n
"
.
So
m
e
o
t
h
er
d
is
t
an
ce
m
etr
ic
s
ar
e
"
m
i
n
k
o
w
s
k
i"
,
"
m
a
x
i
m
u
m
"
,
"
ca
m
b
e
r
r
a"
,
"
m
a
n
h
at
tan
"
,
an
d
"
b
in
ar
y
"
.
T
h
e
ag
g
lo
m
er
atio
n
m
et
h
o
d
f
o
r
h
ier
ar
c
h
ical
cl
u
s
te
r
in
g
is
s
et
to
"
w
ar
d
.
D2
"
.
Oth
er
m
et
h
o
d
s
s
u
ch
a
s
"
w
ar
d
.
D"
,
"
co
m
p
lete"
,
"
s
in
g
le
"
,
"
m
cq
u
itt
y
"
,
"
av
er
ag
e"
,
"
m
e
d
ian
"
o
r
"
ce
n
tr
o
id
"
ca
n
also
b
e
s
elec
ted
.
e.
So
f
t c
l
u
s
ter
i
n
g
So
f
t
p
ar
titi
o
n
s
o
f
d
ata
ar
e
n
atu
r
all
y
g
i
v
en
o
u
t
as
o
u
tp
u
t
b
y
s
o
m
e
clu
s
ter
in
g
alg
o
r
it
h
m
s
li
k
e
as
E
M
[
9
]
an
d
f
u
zz
y
c
m
ea
n
s
[
1
0
]
.
T
h
e
d
eg
r
ee
o
f
r
elatio
n
(
ass
o
c
iatio
n
)
o
f
ea
ch
in
s
tan
ce
to
ea
ch
o
u
tp
u
t
cl
u
s
ter
ca
n
b
e
ass
i
g
n
ed
a
v
a
lu
e
b
y
s
o
f
t
p
a
r
titi
o
n
.
P
ed
r
y
cz
cr
ea
ted
a
m
iles
to
n
e
o
n
“
co
llab
o
r
ativ
e”
f
u
zz
y
cl
u
s
ter
in
g
b
y
w
o
r
k
i
n
g
o
n
it
[
1
1
]
.
P
ed
r
y
cz
co
n
s
id
er
ed
a
v
er
tical
p
ar
titi
o
n
in
g
s
ce
n
ar
io
.
He
th
en
r
ep
r
esen
ted
th
e
a
s
s
o
ciatio
n
o
f
m
u
ltip
le
p
ar
titi
o
n
i
n
g
v
ia
p
air
w
i
s
e
in
ter
ac
tio
n
co
ef
f
icie
n
t
s
.
T
h
is
r
esu
lted
i
n
th
e
i
n
cr
ea
s
e
i
n
th
e
co
s
t
f
u
n
c
t
io
n
to
in
cl
u
d
e
th
e
e
f
f
ec
t
o
f
ass
o
ciatio
n
in
o
p
ti
m
izatio
n
p
r
o
ce
s
s
.
f.
Fan
n
y
F
A
NNY,
d
er
iv
es i
ts
n
a
m
e
f
r
o
m
F
u
zz
y
ANa
lYs
i
s
.
Fa
n
n
y
ai
m
s
to
m
i
n
i
m
ize
th
e
o
b
j
ec
tiv
e
f
u
n
ct
io
n
.
S
UM_
[
v=1
.
.
k
]
(
S
UM_
(
i,j)
u
(
i,
v)
^r
u
(
j,v
)
^
r
d
(
i,j
)
)
/ (
2
S
UM_
j
u
(
j,v
)
^
r
)
W
h
er
e
n
is
t
h
e
n
u
m
b
er
o
f
o
b
s
e
r
v
atio
n
s
,
k
is
t
h
e
n
u
m
b
er
o
f
cl
u
s
ter
s
,
r
is
th
e
m
e
m
b
er
s
h
ip
ex
p
o
n
e
n
t
m
e
m
b
.
e
x
p
d
(
i,j)
is
th
e
d
is
s
i
m
ilar
it
y
b
et
w
e
en
o
b
s
er
v
atio
n
s
i
an
d
j.
I
f
r
-
> 1
g
i
v
es i
n
cr
ea
s
i
n
g
l
y
cr
i
s
p
er
clu
s
ter
i
n
g
’
s
w
h
er
ea
s
r
-
> ∞
lead
s
to
co
m
p
lete
f
u
zz
i
n
es
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
9
,
No
.
4
,
A
u
g
u
s
t 2
0
1
9
:
2
7
6
0
-
2770
2762
C
o
m
p
ar
ed
to
o
th
er
f
u
zz
y
clu
s
t
er
in
g
m
et
h
o
d
s
,
f
a
n
n
y
h
a
s
th
e
f
o
llo
w
i
n
g
f
ea
t
u
r
es:
(
a)
I
t a
ls
o
ac
ce
p
ts
a
d
is
s
i
m
ilar
i
t
y
m
a
tr
ix
;
(
b
)
I
t is m
o
r
e
r
o
b
u
s
t t
o
th
e
s
p
h
er
ical
clu
s
ter
as
s
u
m
p
tio
n
;
(
c)
I
t p
r
o
v
id
es a
n
o
v
el
g
r
ap
h
ic
al
d
is
p
la
y
,
th
e
s
ilh
o
u
ette
p
lo
t.
g.
Valid
atio
n
T
o
ass
ess
w
h
et
h
er
f
a
n
n
y
r
e
s
u
lted
i
n
b
etter
clu
s
ter
s
,
v
al
id
atio
n
o
n
cl
u
s
ter
s
s
h
o
u
ld
b
e
c
ar
r
ied
o
u
t.
Hen
ce
,
clVa
lid
p
ac
k
a
g
e
w
a
s
u
s
ed
w
h
er
e
it
o
f
f
e
r
s
i
n
ter
n
al
an
d
s
tab
ilit
y
v
alid
atio
n
p
r
o
ce
d
u
r
es.
On
l
y
th
e
clu
s
ter
i
n
g
p
ar
titi
o
n
an
d
t
h
e
d
a
taset
ar
e
co
n
s
id
er
ed
as
i
n
p
u
t
b
y
i
n
ter
n
al
v
a
lid
atio
n
.
I
n
tr
in
s
ic
in
f
o
r
m
atio
n
i
n
t
h
e
d
ata
is
u
s
ed
to
ass
ess
th
e
q
u
ali
t
y
o
f
clu
s
ter
in
g
.
T
h
e
s
tab
ilit
y
m
ea
s
u
r
es
ev
al
u
ate
t
h
e
co
n
s
is
t
en
c
y
o
f
a
clu
s
ter
in
g
r
esu
lt.
I
t
is
d
o
n
e
b
y
co
m
p
ar
i
n
g
cl
u
s
ter
i
n
g
r
es
u
lt
w
it
h
t
h
e
clu
s
ter
s
o
b
tain
e
d
a
f
ter
ea
ch
c
o
lu
m
n
is
r
e
m
o
v
ed
,
o
n
e
at
a
ti
m
e,
a
leav
e
-
o
n
e
-
o
u
t
m
et
h
o
d
.
h.
I
n
ter
n
al
v
alid
atio
n
I
n
ter
n
al
v
alid
atio
n
is
a
m
ea
s
u
r
e
o
f
th
e
co
n
n
ec
ted
n
e
s
s
co
m
p
ac
tn
e
s
s
a
n
d
s
ep
ar
atio
n
o
f
th
e
clu
s
ter
p
ar
titi
o
n
s
.
T
h
e
e
x
te
n
t
to
w
h
ic
h
,
o
b
s
er
v
atio
n
s
ar
e
p
lace
d
i
n
t
h
e
s
a
m
e
cl
u
s
ter
a
s
t
h
eir
n
ea
r
e
s
t
n
ei
g
h
b
o
r
s
i
n
th
e
d
ata
s
p
ac
e
is
r
ef
er
r
ed
to
as
co
n
n
ec
ted
n
e
s
s
[
1
2
]
.
C
o
m
p
ac
t
n
es
s
,
r
ef
er
r
ed
b
y
Du
n
n
i
n
d
ex
[
1
3
]
,
ass
ess
es
t
h
e
h
o
m
o
g
en
eit
y
o
f
a
clu
s
ter
.
T
h
is
is
u
s
u
all
y
d
o
n
e
b
y
lo
o
k
in
g
at
th
e
i
n
tr
a
-
clu
s
ter
v
ar
ian
ce
.
Sep
ar
atio
n
,
r
ef
er
r
ed
b
y
s
ilh
o
u
ette
w
id
t
h
[
1
4
]
,
m
ea
s
u
r
es
th
e
d
eg
r
ee
o
f
s
ep
ar
atio
n
b
et
w
ee
n
th
e
cl
u
s
ter
s
.
T
h
is
is
u
s
u
al
l
y
d
o
n
e
b
y
m
ea
s
u
r
in
g
t
h
e
d
is
ta
n
ce
b
et
w
ee
n
th
e
ce
n
tr
o
id
s
o
f
t
h
e
clu
s
ter
s
.
i.
Stab
ilit
y
v
alid
atio
n
L
et
t
h
e
to
tal
n
u
m
b
er
o
f
o
b
s
er
v
atio
n
s
(ro
w
s
)
in
a
d
ataset
b
e
N
an
d
th
e
to
ta
l
n
u
m
b
er
o
f
co
lu
m
n
s
b
e
M.
T
h
en
th
e
cl
u
s
ter
in
g
r
es
u
lt
s
will
b
e
co
m
p
ar
ed
b
y
t
h
e
s
tab
il
it
y
m
ea
s
u
r
es
b
ased
o
n
th
e
f
u
ll
d
ata
to
p
er
f
o
r
m
clu
s
ter
i
n
g
v
e
r
s
u
s
b
ased
o
n
lea
v
i
n
g
o
n
e
co
l
u
m
n
,
o
n
e
at
a
ti
m
e
[
1
5
]
.
T
h
e
av
er
ag
e
d
is
ta
n
ce
b
et
w
ee
n
t
h
e
m
ea
n
s
(
A
DM
)
,
t
h
e
av
er
a
g
e
p
r
o
p
o
r
tio
n
o
f
n
o
n
-
o
v
er
lap
(
A
P
N)
,
an
d
th
e
f
ig
u
r
e
o
f
m
er
it
(
F
OM
)
,
th
e
av
er
ag
e
d
is
tan
c
e
(
A
D)
,
ar
e
th
e
s
tab
ili
t
y
m
ea
s
u
r
es.
j.
Fu
zz
y
C
-
Me
a
n
s
(
F
C
M)
Fu
zz
y
C
-
i
s
a
c
lu
s
ter
in
g
m
et
h
o
d
w
h
ich
allo
w
s
o
n
e
p
o
in
t
to
b
elo
n
g
to
o
n
e
o
r
m
o
r
e
t
h
a
n
o
n
e
cl
u
s
ter
s
.
Du
n
n
d
e
v
elo
p
ed
th
is
m
et
h
o
d
in
1
9
7
3
an
d
B
ez
d
ek
i
m
p
r
o
v
e
d
it
in
t
h
e
y
ea
r
1
9
8
1
.
I
t
is
o
f
ten
u
s
ed
i
n
p
atter
n
r
ec
o
g
n
itio
n
.
A
f
i
n
ite
co
llect
io
n
o
f
p
o
in
ts
is
p
ar
titi
o
n
ed
i
n
to
a
co
llectio
n
o
f
C
f
u
zz
y
c
lu
s
t
er
s
o
n
t
h
e
b
asi
s
o
f
g
iv
e
n
cr
iter
ia
is
d
o
n
e
b
y
t
h
e
FC
M
al
g
o
r
it
h
m
.
T
h
u
s
,
t
h
e
p
o
in
ts
l
y
in
g
o
n
t
h
e
ed
g
e
o
f
a
cl
u
s
ter
,
m
a
y
b
e
i
n
t
h
e
clu
s
ter
to
a
less
er
d
eg
r
ee
th
a
n
p
o
in
ts
l
y
i
n
g
i
n
th
e
ce
n
tr
e
o
f
clu
s
ter
.
B
ased
o
n
m
i
n
i
m
izatio
n
o
f
th
e
f
o
llo
w
in
g
o
b
j
ec
tiv
e
f
u
n
c
tio
n
t
h
e
FC
M
al
g
o
r
ith
m
is
d
er
iv
ed
:
1
,
µ
a
µ,
2
1
1
m
a
x
J
i
j
c
i
n
j
m
ij
m
Fu
zz
y
c
-
m
ea
n
s
(
F
C
M)
is
t
h
e
b
es
t
f
u
zz
y
cl
u
s
ter
i
n
g
m
et
h
o
d
k
n
o
w
n
[
1
6
].
Fu
zz
y
lo
g
ic
i
s
u
s
ed
s
o
th
at
ea
ch
i
n
s
ta
n
ce
is
a
s
s
o
ciate
d
n
o
t
o
n
l
y
w
it
h
a
s
i
n
g
le
cl
u
s
ter
b
u
t
al
s
o
h
a
s
a
ce
r
ta
in
d
eg
r
ee
o
f
m
e
m
b
er
s
h
ip
w
it
h
ea
ch
o
f
t
h
e
ce
n
tr
o
id
s
th
at
e
x
is
t.
T
h
e
f
u
zz
in
e
s
s
o
f
th
e
s
o
lu
tio
n
b
ec
o
m
e
s
clo
s
er
to
1
as
m
g
o
es
c
lo
s
er
to
in
f
in
it
y
(
∞).
T
h
e
s
o
l
u
tio
n
b
ec
o
m
e
s
i
n
cr
ea
s
i
n
g
l
y
s
i
m
ila
r
to
th
e
c
lu
s
ter
in
g
o
f
b
i
n
ar
y
k
-
m
ea
n
s
[
1
7
]
.
Sett
in
g
m=
2
.
0
is
a
g
o
o
d
ch
o
ice
[
1
8
].
R
u
s
p
in
i
F
u
zz
y
cl
u
s
ter
in
g
t
h
eo
r
y
is
t
h
e
b
ase
f
o
r
F
C
M
cl
u
s
te
r
in
g
a
lg
o
r
it
h
m
.
On
th
e
b
a
s
is
o
f
d
is
ta
n
ce
b
et
w
ee
n
v
ar
io
u
s
in
p
u
t
d
ata
p
o
in
ts
,
a
n
al
y
s
is
i
s
ca
r
r
ied
o
u
t
b
y
th
is
al
g
o
r
ith
m
.
B
asin
g
o
n
th
e
d
is
tan
ce
b
et
w
ee
n
d
ata
p
o
in
ts
an
d
t
h
e
cl
u
s
ter
ce
n
tr
es
t
h
e
cl
u
s
ter
s
ar
e
f
o
r
m
ed
.
Gr
o
u
p
in
g
o
f
d
ataset
in
to
n
cl
u
s
ter
s
w
it
h
ea
c
h
a
n
d
ev
er
y
d
ata
p
o
in
t
in
th
e
d
atas
et
b
ein
g
r
elate
d
to
ev
er
y
clu
s
ter
is
d
o
n
e
b
y
u
s
in
g
th
e
F
C
M
d
ata
clu
s
ter
i
n
g
tech
n
iq
u
e
[
1
9
]
.
T
h
e
d
ata
p
o
in
t
w
ill
h
a
v
e
a
h
i
g
h
lo
w
d
e
g
r
ee
o
f
co
n
n
ec
tio
n
w
i
th
th
e
cl
u
s
ter
w
h
o
s
e
ce
n
tr
e
o
f
t
h
e
clu
s
ter
i
s
l
y
i
n
g
f
ar
a
w
a
y
f
r
o
m
it a
n
d
w
ill h
a
v
e
h
ig
h
d
eg
r
ee
o
f
co
n
n
ec
tio
n
i
f
t
h
e
cl
u
s
ter
ce
n
tr
e
is
clo
s
er
to
it.
k.
Valid
atio
n
: F
u
zz
y
C
-
Me
an
s
(
F
C
M)
C
lu
s
te
r
v
alid
it
y
i
n
d
ices
ar
e
o
b
j
ec
tiv
e
f
u
n
ctio
n
s
.
T
h
e
clu
s
ter
v
alid
it
y
i
n
d
ices
ar
e
u
s
ed
to
o
b
tain
clu
s
ter
i
n
g
in
d
ices,
an
d
cl
u
s
t
er
in
g
s
o
l
u
tio
n
[
2
0
]
.
Valid
atio
n
m
ea
s
u
r
es
s
u
ch
as
"
p
ar
ti
tio
n
.
co
ef
f
icie
n
t"
,
"
g
ath
.
g
ev
a"
,
"
f
u
k
u
y
a
m
a.
s
u
g
en
o
"
"
x
ie.
b
en
i"
,
“
p
ar
titi
o
n
en
tr
o
p
y
”,
"
s
ep
ar
atio
n
.
in
d
ex
"
w
er
e
o
b
tain
ed
b
y
r
u
n
n
i
n
g
e1
0
7
1
clu
s
ter
p
ac
k
ag
e.
Gath
an
d
Gev
a
[
2
1
]
ad
o
p
te
d
th
r
ee
m
ea
s
u
r
es
s
u
ch
as
a
v
er
ag
e
p
ar
titi
o
n
d
en
s
i
t
y
(
A
P
D)
,
f
u
zz
y
h
y
p
er
v
o
lu
m
e
(
F
HV)
an
d
p
ar
titi
o
n
d
en
s
it
y
(
P
D)
.
T
h
e
Xie
-
B
en
i
i
n
d
ex
(
XB
)
[
2
2
]
d
ef
in
ed
in
ter
-
clu
s
ter
s
ep
ar
atio
n
as t
h
e
s
q
u
ar
e
o
f
th
e
m
i
n
i
m
u
m
d
is
ta
n
ce
b
et
w
ee
n
t
h
e
ce
n
ter
s
o
f
th
e
cl
u
s
ter
s
.
T
h
ey
d
e
f
i
n
ed
th
e
in
tr
a
-
c
lu
s
ter
co
m
p
ac
t
n
es
s
as
t
h
e
s
q
u
ar
e
o
f
t
h
e
m
ea
n
g
ap
(
d
is
tan
ce
)
b
et
w
ee
n
ea
c
h
d
ata
o
b
j
ec
t
an
d
its
clu
s
te
r
ce
n
ter
.
W
h
en
t
h
e
m
in
i
m
u
m
o
f
XB
is
f
o
u
n
d
th
e
o
p
ti
m
a
l c
lu
s
t
e
r
n
u
m
b
er
i
s
s
aid
to
b
e
r
ea
ch
ed
.
T
w
o
p
r
o
p
er
tie
s
o
f
clu
s
ter
in
g
ar
e
co
m
p
u
ted
b
y
u
s
in
g
F
u
k
u
y
a
m
a
-
S
u
g
e
n
o
in
d
e
x
(
FS
)
.
T
h
ey
ar
e
co
h
esio
n
an
d
s
ep
ar
atio
n
.
T
h
e
ai
m
o
f
en
h
a
n
cin
g
o
f
t
h
i
s
o
b
j
ec
tiv
e
f
u
n
ctio
n
i
s
to
d
ec
r
ea
s
e
t
h
e
s
ep
ar
atio
n
b
et
w
ee
n
clu
s
ter
s
o
f
a
d
ata
s
et
an
d
in
cr
ea
s
e
th
e
co
m
p
ac
t
n
es
s
o
f
a
ce
r
tain
clu
s
ter
.
A
g
o
o
d
p
a
r
titi
o
n
is
p
r
o
p
o
s
ed
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2
0
8
8
-
8708
F
u
z
z
y
clu
s
teri
n
g
a
n
d
F
u
z
z
y
C
-
Mea
n
s
p
a
r
titi
o
n
clu
s
ter
a
n
a
lysi
s
a
n
d
va
lid
a
tio
n
…
(
K
.
V
a
r
a
d
a
R
a
jku
ma
r
)
2763
b
y
t
h
e
m
in
i
m
u
m
v
al
u
es
o
f
t
h
is
in
d
ex
.
T
h
e
f
u
zz
i
n
es
s
o
f
th
e
p
ar
titi
o
n
is
m
ea
s
u
r
ed
b
y
p
ar
titi
o
n
co
ef
f
icie
n
t
i
n
d
ex
b
u
t
w
ith
o
u
t
tak
i
n
g
in
to
co
n
s
i
d
er
atio
n
th
e
d
ataset
its
el
f
.
I
t
is
a
k
in
d
o
f
tr
ail
an
d
er
r
o
r
m
et
h
o
d
s
in
ce
it
h
a
s
n
o
co
n
n
ec
tio
n
to
a
n
y
p
r
o
p
er
t
y
o
f
th
e
d
ata.
A
g
o
o
d
p
ar
titi
o
n
is
i
m
p
lied
b
y
m
ax
i
m
u
m
v
al
u
es
i
n
t
h
e
m
ea
n
i
n
g
o
f
a
least
f
u
zz
y
cl
u
s
ter
i
n
g
.
I
n
f
o
r
m
atio
n
ab
o
u
t
t
h
e
m
e
m
b
er
s
h
i
p
m
atr
ix
is
p
r
o
v
id
ed
b
y
p
ar
titi
o
n
e
n
tr
o
p
y
in
d
e
x
,
w
it
h
o
u
t
also
co
n
s
id
er
i
n
g
t
h
e
d
ata
its
el
f
.
A
g
o
o
d
p
ar
titi
o
n
is
im
p
lied
b
y
th
e
m
i
n
i
m
u
m
v
a
lu
e
s
in
t
h
e
m
ea
n
i
n
g
o
f
a
m
o
r
e
cr
is
p
p
ar
titi
o
n
.
Sep
ar
atio
n
in
d
e
x
(
C
S
I
n
d
ex
)
id
en
t
if
ies
u
n
iq
u
e
cl
u
s
ter
s
tr
u
ct
u
r
e
w
it
h
w
ell
-
d
e
f
in
ed
f
ea
t
u
r
es
t
h
at
d
ep
en
d
o
n
th
e
d
ata
an
d
a
m
ea
s
u
r
e
o
f
s
ep
er
ati
o
n
.
I
n
f
o
r
m
atio
n
ab
o
u
t
th
e
m
e
m
b
er
s
h
ip
m
a
tr
ix
is
p
r
o
v
id
ed
b
y
p
ar
titi
o
n
en
tr
o
p
y
in
d
ex
,
w
it
h
o
u
t
also
co
n
s
id
er
in
g
t
h
e
d
ata
its
e
lf
.
A
g
o
o
d
p
ar
titi
o
n
is
i
m
p
lied
b
y
th
e
m
i
n
i
m
u
m
v
al
u
es
i
n
t
h
e
m
ea
n
in
g
o
f
a
m
o
r
e
cr
is
p
p
ar
titi
o
n
.
Sep
ar
atio
n
i
n
d
ex
(
C
S
I
n
d
ex
)
id
en
ti
f
ie
s
u
n
iq
u
e
clu
s
ter
s
tr
u
ct
u
r
e
w
it
h
w
ell
-
d
ef
in
ed
f
ea
t
u
r
es
t
h
at
d
ep
en
d
o
n
t
h
e
d
ata
an
d
a
m
ea
s
u
r
e
o
f
s
ep
er
atio
n
.
3.
RE
SU
L
T
S
A
ND
AN
AL
Y
SI
S
I
f
a
f
u
zz
y
cl
u
s
ter
i
n
g
tec
h
n
iq
u
e
i
s
to
b
e
e
m
p
lo
y
ed
in
an
a
p
p
licatio
n
o
r
ien
ted
ta
s
k
to
s
i
g
n
i
f
ica
n
tl
y
clu
s
ter
g
r
o
u
p
s
o
f
o
b
j
ec
ts
,
th
en
t
h
e
f
ir
s
t
a
s
p
ec
t
to
en
co
u
n
ter
i
s
ass
e
s
s
i
n
g
cl
u
s
ter
i
n
g
ten
d
e
n
c
y
.
T
o
ac
h
iev
e
b
etter
r
esu
lt
s
i
n
cl
u
s
ter
an
a
l
y
s
is
,
d
a
ta
h
a
s
b
ee
n
i
n
v
e
s
ti
g
ated
f
o
r
clu
s
ter
ab
ilit
y
.
He
n
ce
,
a
m
eth
o
d
p
u
t
f
o
r
w
ar
d
b
y
B
an
er
j
ee
et
al
[
2
3
]
s
u
ch
a
s
H
o
p
k
in
s
s
tati
s
tic
a
s
s
e
s
s
es
cl
u
s
t
er
in
g
te
n
d
e
n
c
y
o
f
a
d
ata
s
et.
T
h
is
is
m
ea
s
u
r
ed
b
y
ca
lcu
lati
n
g
th
e
p
r
o
b
ab
ilit
y
o
f
u
n
i
f
o
r
m
d
ata
d
is
tr
ib
u
tio
n
o
r
o
th
er
w
i
s
e
th
e
‘
s
p
atial
r
an
d
o
m
n
e
s
s
’
o
f
t
h
e
d
ata.
A
v
al
u
e
o
f
Ho
p
k
i
n
s
s
tati
s
tic
clo
s
e
to
ze
r
o
r
ep
r
esen
ts
t
h
e
n
u
l
l
h
y
p
o
th
es
is
b
ein
g
r
ej
ec
ted
an
d
th
e
d
ataset
re
p
r
esen
ts
s
i
g
n
if
ican
t
cl
u
s
ter
ab
ilit
y
.
T
h
e
citesco
r
e
d
ataset
co
n
s
id
er
ed
in
th
e
an
al
y
s
i
s
r
esu
lted
i
n
Ho
p
k
in
s
s
tatis
t
ic
o
f
0
.
4
3
7
1
,
a
v
alu
e
<
0
.
5
,
in
d
icatin
g
t
h
at
t
h
e
d
ata
is
h
ig
h
l
y
cl
u
s
ter
ab
le.
a.
NB
C
lu
s
t
Fro
m
Nb
C
l
u
s
t
an
a
l
y
s
is
,
t
h
e
r
esu
lt
g
i
v
en
b
elo
w
s
ig
n
i
f
ie
s
th
at
ar
o
u
n
d
9
v
ar
io
u
s
in
d
ices
p
r
o
p
o
s
ed
3
clu
s
ter
s
o
lu
t
io
n
s
a
s
b
est
c
lu
s
t
er
s
an
d
4
p
r
o
p
o
s
ed
2
s
o
lu
tio
n
s
a
n
d
8
v
alid
it
y
i
n
d
ices
s
u
g
g
ested
1
0
clu
s
ter
s
.
A
cc
o
r
d
in
g
to
th
e
m
aj
o
r
ity
r
u
le
,
th
e
b
est n
u
m
b
er
o
f
cl
u
s
ter
s
is
3
clu
s
ter
g
r
o
u
p
s
.
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
Am
o
n
g
all
in
d
ice
s
:
*
4
p
r
o
p
o
s
ed
2
as th
e
b
est n
u
m
b
er
o
f
cl
u
s
ter
s
*
9
p
r
o
p
o
s
ed
3
as th
e
b
est n
u
m
b
er
o
f
cl
u
s
ter
s
*
1
p
r
o
p
o
s
ed
4
as th
e
b
est n
u
m
b
er
o
f
cl
u
s
ter
s
*
1
p
r
o
p
o
s
ed
5
as th
e
b
est n
u
m
b
er
o
f
cl
u
s
ter
s
*
2
p
r
o
p
o
s
ed
7
as th
e
b
est n
u
m
b
er
o
f
cl
u
s
ter
s
*
1
p
r
o
p
o
s
ed
8
as th
e
b
est n
u
m
b
er
o
f
cl
u
s
ter
s
*
1
p
r
o
p
o
s
ed
9
as th
e
b
est n
u
m
b
er
o
f
cl
u
s
ter
s
*
8
p
r
o
p
o
s
ed
1
0
as th
e
b
est n
u
m
b
er
o
f
cl
u
s
ter
s
*
A
cc
o
r
d
in
g
to
t
h
e
m
aj
o
r
it
y
r
u
le,
th
e
b
est n
u
m
b
er
o
f
cl
u
s
ter
s
is
3
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
b.
Hu
b
er
t in
d
ex
T
h
e
Hu
b
er
t
in
d
ex
is
a
g
r
ap
h
ic
al
m
et
h
o
d
o
f
d
eter
m
in
i
n
g
th
e
n
u
m
b
er
o
f
clu
s
ter
s
.
I
n
t
h
e
p
lo
t
o
f
Hu
b
er
t
in
d
ex
,
a
s
i
g
n
if
ican
t
k
n
ee
w
as
o
b
s
er
v
ed
at
3
r
d
clu
s
ter
th
at
co
r
r
esp
o
n
d
s
to
a
s
ig
n
i
f
ica
n
t
i
n
cr
ea
s
e
o
f
th
e
v
al
u
e
o
f
th
e
m
ea
s
u
r
e
i.e
th
e
s
i
g
n
if
ican
t
p
ea
k
in
H
u
b
er
t in
d
ex
s
ec
o
n
d
d
if
f
er
en
ce
s
p
lo
t a
s
s
h
o
w
n
in
Fi
g
u
r
e
1.
Fig
u
r
e
1
.
Hu
b
er
t i
n
d
ex
c.
D
-
i
n
d
ex
T
h
e
D
in
d
ex
i
s
a
g
r
ap
h
ical
m
eth
o
d
o
f
d
eter
m
i
n
i
n
g
t
h
e
n
u
m
b
er
o
f
cl
u
s
ter
s
.
I
n
t
h
e
p
lo
t
o
f
D
i
n
d
ex
,
a
s
i
g
n
i
f
ica
n
t
k
n
ee
(
th
e
s
i
g
n
i
f
ican
t
p
ea
k
i
n
Di
n
d
ex
s
ec
o
n
d
d
if
f
er
en
ce
s
p
lo
t)
o
b
s
er
v
ed
at
3
r
d
clu
s
ter
t
h
at
co
r
r
esp
o
n
d
s
to
a
s
ig
n
i
f
ica
n
t
i
n
cr
ea
s
e
o
f
th
e
v
al
u
e
o
f
t
h
e
m
ea
s
u
r
e
as
s
h
o
w
n
in
Fi
g
u
r
e
2.
Fro
m
t
h
e
Fi
g
u
r
e
3
an
d
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
9
,
No
.
4
,
A
u
g
u
s
t 2
0
1
9
:
2
7
6
0
-
2770
2764
Nb
C
lu
s
t o
u
tp
u
t
in
T
ab
le
1
,
it c
an
b
e
co
n
c
lu
d
ed
t
h
at
t
h
e
o
p
ti
m
al
n
u
m
b
er
o
f
cl
u
s
ter
s
,
f
o
r
th
e
d
ataset
co
m
p
r
i
s
in
g
o
f
cite
s
co
r
e
b
ib
lio
m
etr
ic
d
atas
et
w
as
f
o
u
n
d
to
b
e
3
cl
u
s
ter
s
o
lu
tio
n
s
.
T
h
er
ef
o
r
e,
in
itial
v
al
u
e
o
f
k
=3
w
as
u
s
ed
to
p
er
f
o
r
m
f
u
zz
y
cl
u
s
ter
i
n
g
a
n
al
y
s
i
s
.
Fig
u
r
e
2
.
D
-
i
n
d
ex
Fig
u
r
e
3
.
C
o
n
s
e
n
s
u
s
o
n
o
p
ti
m
al
n
u
m
b
er
o
f
clu
s
ter
s
o
b
tain
ed
f
r
o
m
Nb
C
l
u
s
t p
ac
k
a
g
e
T
ab
le
1
.
NB
C
L
UST
Ou
tp
u
t
d.
Fan
n
y
A
s
u
b
s
et
d
ata
co
m
p
r
i
s
in
g
o
f
2
6
o
b
j
ec
ts
w
er
e
s
u
b
j
ec
ted
to
f
u
zz
y
clu
s
ter
i
n
g
(
f
an
n
y
)
a
lg
o
r
it
h
m
an
d
t
h
e
d
ata
w
a
s
s
ca
led
to
co
n
v
er
t
in
t
o
a
d
ata
f
r
a
m
e,
p
r
o
v
id
ed
k
=3
as
d
esire
d
n
u
m
b
er
o
f
cl
u
s
ter
s
.
T
h
e
r
u
n
r
es
u
lted
i
n
clu
s
ter
as
w
ell
as
s
il
h
o
u
ette
p
lo
ts
.
A
s
p
er
R
u
s
p
i
n
i
[
5
]
,
th
e
d
ata
p
o
in
ts
i
n
t
h
e
ce
n
tr
e
o
f
th
e
clu
s
t
er
ca
n
h
a
v
e
a
d
eg
r
ee
eq
u
al
to
1
an
d
ar
e
as
s
o
ciate
d
w
i
th
th
e
m
e
m
b
er
s
h
ip
d
eg
r
ee
s
to
t
h
e
n
ea
r
b
y
cl
u
s
ter
s
.
C
er
tain
p
o
i
n
ts
n
ea
r
th
e
b
o
u
n
d
ar
y
o
f
th
e
cl
u
s
ter
wer
e
id
en
tif
ied
a
m
o
n
g
clu
s
ter
s
1
an
d
2
r
esp
ec
tiv
el
y
.
Fro
m
Fi
g
u
r
e
4
,
p
o
in
t
1
6
o
f
clu
s
ter
1
a
n
d
p
o
in
t
4
o
f
clu
s
t
er
2
h
av
e
a
c
h
a
n
ce
to
ap
p
ea
r
in
eit
h
er
o
f
th
e
clu
s
ter
s
.
Fro
m
t
h
e
m
e
m
b
er
s
h
ip
co
ef
f
icie
n
t
d
ata
g
i
v
en
b
elo
w
,
it
ca
n
b
e
f
o
u
n
d
th
at
o
n
l
y
p
o
in
t
1
6
lies
b
etw
ee
n
clu
s
ter
1
an
d
2
(
w
i
th
co
ef
f
icie
n
t
s
4
3
an
d
3
9
)
s
h
ar
in
g
5
0
%
ch
ar
a
cter
is
tics
w
h
er
ea
s
p
o
in
t
4
h
as
m
o
r
e
cl
u
s
ter
-
2
s
i
g
n
i
f
ican
ce
t
h
an
clu
s
ter
1
w
h
ic
h
is
f
o
u
n
d
f
r
o
m
co
ef
f
icie
n
ts
b
ein
g
7
3
f
o
r
clu
s
ter
-
2
a
n
d
2
0
f
o
r
clu
s
ter
-
1.
Me
m
b
er
s
h
ip
co
ef
f
icie
n
t
s
(
in
%,
r
o
u
n
d
ed
)
as sh
o
w
n
in
T
ab
le
2
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2
0
8
8
-
8708
F
u
z
z
y
clu
s
teri
n
g
a
n
d
F
u
z
z
y
C
-
Mea
n
s
p
a
r
titi
o
n
clu
s
ter
a
n
a
lysi
s
a
n
d
va
lid
a
tio
n
…
(
K
.
V
a
r
a
d
a
R
a
jku
ma
r
)
2765
Fig
u
r
e
4
.
Fan
n
y
o
u
tp
u
t s
h
o
w
i
n
g
3
clu
s
ter
s
T
ab
le
2
.
Me
m
b
er
s
h
ip
c
o
ef
f
ici
en
t
s
[
,
1
]
[
,
2
]
[
,
3
]
[
,
1
]
[
,
2
]
[
,
3
]
[
1
,
]
73
19
8
[
1
4
,
]
60
19
21
[
2
,
]
85
10
5
[
1
5
,
]
11
69
20
[
3
,
]
18
71
10
[
1
6
,
]
43
39
19
[
4
,
]
20
73
7
[
1
7
,
]
10
68
22
[
5
,
]
12
79
9
[
1
8
,
]
6
8
86
[
6
,
]
68
20
12
[
1
9
,
]
2
4
93
[
7
,
]
11
82
8
[
2
0
,
]
7
19
75
[
8
,
]
9
83
8
[
2
1
,
]
5
1
5
80
[
9
,
]
76
15
9
[
2
2
,
]
2
3
95
[
1
0
,
]
7
89
5
[
2
3
,
]
1
3
96
[
1
1
,
]
6
89
5
[
2
4
,
]
2
4
94
[
1
2
,
]
26
57
17
[
2
5
,
]
2
5
93
[
1
3
,
]
10
56
33
[
2
6
,
]
4
6
90
e.
Fan
n
y
o
u
tp
u
t
T
ab
les 3
-
6
p
r
esen
ts
ab
o
u
t Fa
n
n
y
o
u
tp
u
t.
T
ab
le
3
.
Fu
zz
y
clu
s
ter
in
g
o
b
j
e
cts o
f
clas
s
‘
f
a
n
n
y
’
F
u
z
z
y
C
l
u
st
e
r
i
n
g
O
b
j
e
c
t
s
o
f
C
l
a
ss ‘f
a
n
n
y
’
m.sh
i
p
.
e
x
p
o
n
2
O
b
j
e
c
t
i
v
e
1
0
.
1
5
7
8
5
T
o
l
e
r
a
n
c
e
1e
-
15
I
t
e
r
a
t
i
o
n
s
2
9
9
C
o
n
v
e
r
g
e
d
1
M
a
x
i
t
5
0
0
n
26
T
ab
le
4
.
Fu
zz
y
n
e
s
s
c
o
ef
f
icie
n
t
s
d
u
n
n
_
c
o
e
f
f
N
o
r
mal
i
z
e
d
0
.
4
3
2
4
3
6
9
0
.
1
4
8
6
5
5
4
T
ab
le
5
.
A
T
T
R
(
“
SC
AL
E
D:
C
E
NT
E
R
)
C
i
t
e
S
c
o
r
e
C
o
u
n
t
S
N
I
P
S
JR
1
8
.
6
3
1
5
3
8
1
0
0
6
6
.
5
7
6
9
2
3
6
.
3
1
3
3
4
6
1
0
.
7
4
0
3
0
8
T
ab
le
6
.
A
T
T
R
(
,
"
SC
AL
E
D:
S
C
AL
E
")
C
i
t
e
S
c
o
r
e
C
o
u
n
t
S
N
I
P
S
JR
9
.
6
8
7
7
4
6
1
5
6
2
2
.
3
4
4
8
6
9
3
.
3
1
4
1
1
9
7
.
4
7
3
6
1
6
f.
Fu
zz
y
C
-
Me
a
n
s
T
h
e
f
u
zz
y
c
-
m
ea
n
s
p
r
o
g
r
a
m
w
r
itte
n
in
R
w
a
s
r
u
n
u
s
i
n
g
d
ef
au
lt
p
ar
a
m
eter
s
w
it
h
m
a
x
i
m
u
m
n
u
m
b
er
o
f
iter
atio
n
s
f
i
x
ed
to
1
0
0
,
d
is
t
an
ce
m
ea
s
u
r
e
b
ei
n
g
e
u
clid
ea
n
,
k
=3
as
in
it
ial
ce
n
tr
o
id
s
,
an
d
d
ata
f
r
a
m
e
as
i
n
p
u
t.
Ho
w
e
v
er
,
th
e
f
u
zz
in
e
s
s
p
ar
am
eter
,
m
d
ep
en
d
s
o
n
th
e
d
at
aset
an
d
v
ar
iat
io
n
s
o
f
o
b
j
ec
ts
w
it
h
i
n
t
h
e
d
ataset.
I
n
liter
at
u
r
e,
m
a
n
y
w
o
r
k
s
f
i
x
e
d
m
=2
,
w
h
ic
h
allo
w
ed
co
m
p
u
tatio
n
o
f
µ
ij
.
T
h
er
ef
o
r
e,
w
e
in
ten
d
ed
to
ev
al
u
ate
th
e
ap
p
r
o
p
r
iate
v
al
u
e
o
f
m
f
o
r
citesco
r
e
s
u
b
s
et
d
ata
s
et
s
tar
ti
n
g
f
r
o
m
m
=1
,
w
h
ic
h
r
esu
lted
i
n
f
ai
lu
r
e
o
f
FC
M
to
ex
tr
ac
t
a
n
y
cl
u
s
ter
i
n
g
s
o
l
u
ti
o
n
.
F
u
r
th
er
,
g
r
ad
u
al
i
n
cr
ea
s
e
f
r
o
m
m
=1
.
1
til
l
1
.
5
r
esu
l
te
d
in
r
elati
v
el
y
lo
w
m
e
m
b
er
s
h
ip
v
a
lu
e
s
s
u
g
g
es
tin
g
th
at
F
C
M
f
ai
led
to
ass
o
ciate
an
y
cite
s
co
r
e
v
al
u
e
to
an
y
cl
u
s
t
er
w
it
h
k
=3
.
A
p
p
r
o
p
r
ia
te
va
lu
e
o
f m
I
t
w
as
r
ep
o
r
ted
b
y
B
ez
d
ek
,
[
1
7
]
th
at
w
h
en
m
g
o
es
to
in
f
i
n
it
y
,
v
al
u
e
o
f
µi
j
attai
n
s
1
/
K,
w
h
ic
h
s
u
g
g
e
s
ts
th
at
f
o
r
a
g
iv
e
n
d
atas
et,
at
a
p
ar
ticu
lar
v
al
u
e
o
f
m
a
n
d
ab
o
v
e
w
h
ic
h
t
h
e
m
e
m
b
er
s
h
ip
v
al
u
es
o
b
tain
ed
f
r
o
m
f
u
zz
y
c
-
m
ea
n
s
eq
u
als
to
1
/K.
I
t
s
h
o
u
ld
b
e
n
o
ted
t
h
at
t
h
e
m
e
m
b
er
s
h
ip
v
alu
e
s
µ
ij
d
ep
en
d
o
n
th
e
d
is
ta
n
ce
s
b
et
w
ee
n
cite
s
co
r
e
o
b
j
ec
ts
an
d
clu
s
ter
ce
n
tr
o
id
s
.
Fo
r
a
co
m
p
lex
d
ata
s
et,
it
i
s
ap
p
r
o
x
i
m
ated
th
at
t
h
e
cl
u
s
te
r
ce
n
tr
o
id
s
ap
p
ea
r
clo
s
e
to
th
e
o
b
j
ec
ts
o
f
a
d
ataset.
T
h
is
r
aises
p
r
o
b
ab
ilit
y
t
h
at
w
h
en
m
v
ar
ies,
t
h
e
F
C
M
m
e
m
b
er
s
h
ip
v
a
lu
e
s
an
d
co
ef
f
i
cien
t
o
f
v
ar
iatio
n
(
C
V)
o
f
th
e
s
et
o
f
d
is
ta
n
ce
s
b
et
w
ee
n
o
b
j
ec
ts
als
o
v
ar
y
[
2
4
]
.
T
h
er
ef
o
r
e,
FC
M
alg
o
r
ith
m
was
r
u
n
to
d
eter
m
in
e
th
e
d
is
tr
ib
u
tio
n
o
f
m
e
m
b
er
s
h
ip
v
al
u
es
w
it
h
v
ar
iatio
n
i
n
m
f
r
o
m
1
to
2
.
T
h
e
p
r
o
g
r
a
m
i
te
r
ativ
el
y
u
p
d
ates
th
e
clu
s
ter
c
en
tr
es
a
n
d
t
h
e
m
e
m
b
er
s
h
ip
v
alu
es
f
o
r
ea
c
h
d
ata
p
o
in
t,
w
h
ich
m
o
v
es
t
h
e
clu
s
te
r
ce
n
tr
es
to
an
ap
p
r
o
p
r
iate
lo
ca
tio
n
w
ith
i
n
a
d
ata
s
et.
T
h
e
iter
atio
n
m
i
n
i
m
izes
th
e
o
b
j
ec
tiv
e
f
u
n
ctio
n
w
h
ic
h
r
ep
r
esen
ts
th
e
d
is
ta
n
ce
f
r
o
m
a
n
y
g
i
v
e
n
d
a
ta
p
o
in
t
to
th
e
clu
s
ter
ce
n
tr
e.
T
h
e
o
b
tain
ed
d
ata
is
g
i
v
e
n
i
n
T
ab
le
7
an
d
th
e
m
e
m
b
er
s
h
ip
p
lo
ts
ar
e
g
i
v
en
in
Fi
g
u
r
e
5
.
T
h
e
co
ef
f
icie
n
t
o
f
v
ar
iatio
n
(
C
V)
,
also
k
n
o
w
n
a
s
r
elativ
e
v
ar
iab
ilit
y
,
eq
u
als
th
e
s
ta
n
d
ar
d
d
ev
iatio
n
d
iv
id
ed
b
y
t
h
e
m
ea
n
.
I
t
is
a
m
ea
s
u
r
e
o
f
s
p
r
ea
d
th
at
d
escr
ib
es
th
e
a
m
o
u
n
t
o
f
v
ar
iab
ilit
y
r
e
lativ
e
to
t
h
e
m
ea
n
.
I
n
g
e
n
er
al,
th
e
lo
w
er
t
h
e
r
atio
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
9
,
No
.
4
,
A
u
g
u
s
t 2
0
1
9
:
2
7
6
0
-
2770
2766
o
f
s
tan
d
ar
d
d
ev
iatio
n
to
m
ea
n
,
th
e
b
etter
is
th
e
d
ata.
A
s
m
all
CV
in
d
icate
s
th
at
t
h
e
d
ata
p
o
i
n
ts
te
n
d
to
b
e
v
er
y
clo
s
e
to
ea
ch
o
th
er
.
A
h
i
g
h
C
V
in
d
icate
s
th
a
t
th
e
d
ata
p
o
in
ts
ar
e
s
p
r
ea
d
o
u
t
f
r
o
m
o
n
e
an
o
t
h
er
.
T
h
is
o
b
s
er
v
atio
n
is
e
v
id
en
ce
d
f
r
o
m
m
e
m
b
er
s
h
ip
p
lo
ts
in
Fig
u
r
e
5.
T
ab
le
6
.
C
o
ef
f
icie
n
t o
f
v
ar
iati
o
n
v
er
s
u
s
c
h
an
g
es i
n
‘
m
’
v
al
u
es
m
S
t
a
n
d
a
r
d
d
e
v
i
a
t
i
o
n
o
f
c
l
u
st
e
r
d
a
t
a
se
t
M
e
a
n
o
f
c
l
u
st
e
r
d
a
t
a
se
t
C
o
e
f
f
i
c
i
e
n
t
o
f
v
a
r
i
a
n
c
e
1
.
0
0
.
7
7
7
1
.
7
3
1
4
4
.
9
2
1
.
1
0
.
7
7
7
1
.
7
3
1
4
4
.
9
2
1
.
2
0
.
7
7
7
1.
7
3
1
4
4
.
9
2
1
.
3
0
.
7
3
2
2
.
2
6
9
4
2
.
9
6
1
.
4
0
.
7
3
2
2
.
1
5
3
3
3
.
9
7
1
.
5
0
.
7
3
2
2
.
1
5
3
3
3
.
9
7
1
.
6
0
.
7
3
2
2
.
1
5
3
3
3
.
9
7
2
.
0
0
.
7
3
2
1
.
8
4
6
3
9
.
6
3
Fig
u
r
e
5
.
Dis
tr
ib
u
tio
n
o
f
m
e
m
b
er
s
h
ip
v
al
u
es
u
p
o
n
v
ar
y
in
g
m
T
h
e
d
is
tr
ib
u
tio
n
o
f
t
h
e
m
e
m
b
er
s
h
ip
v
alu
e
s
v
ar
ies
m
ar
k
ed
l
y
f
r
o
m
o
n
e
m
v
alu
e
to
an
o
t
h
er
.
Fro
m
t
h
e
ab
o
v
e
d
ata,
it
ca
n
b
e
i
n
f
er
r
ed
t
h
at
w
h
e
n
m
n
ea
r
er
to
1
.
0
,
th
e
FC
M
a
lg
o
r
it
h
m
r
es
u
lts
in
les
s
f
u
zz
y
m
e
m
b
er
s
h
ip
v
alu
e
s
a
n
d
h
en
ce
to
o
b
tain
h
i
g
h
an
d
r
ea
s
o
n
ab
le
m
e
m
b
er
s
h
i
p
v
al
u
es
f
o
r
o
b
j
ec
ts
w
it
h
in
a
clu
s
ter
,
a
co
n
ti
n
u
a
l
p
r
o
ce
s
s
o
f
ev
alu
ati
n
g
m
v
a
lu
e
s
r
esu
lted
in
C
V
3
3
.
9
7
f
o
r
m
v
alu
e
s
1
.
3
to
1
.
5
.
W
h
en
h
ig
h
e
r
m
v
a
lu
e
s
ar
e
u
s
ed
,
a
m
o
r
e
s
p
r
ea
d
in
µij
d
is
tr
ib
u
ti
o
n
w
as o
b
s
er
v
ed
.
W
ith
lo
w
er
m
v
al
u
es,
all
o
b
j
ec
ts
ten
d
to
cl
u
s
ter
ed
as
g
r
o
u
p
s
,
a
f
ea
t
u
r
e
ev
id
en
ce
d
w
h
e
n
u
s
in
g
h
ar
d
clu
s
ter
i
n
g
p
r
o
ce
d
u
r
e
s
u
ch
as
k
-
m
ea
n
s
.
Hen
ce
,
m
e
m
b
er
s
h
ip
v
a
lu
e
s
ar
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2
0
8
8
-
8708
F
u
z
z
y
clu
s
teri
n
g
a
n
d
F
u
z
z
y
C
-
Mea
n
s
p
a
r
titi
o
n
clu
s
ter
a
n
a
lysi
s
a
n
d
va
lid
a
tio
n
…
(
K
.
V
a
r
a
d
a
R
a
jku
ma
r
)
2767
p
lo
tted
f
o
r
clu
s
ter
s
o
b
tain
ed
f
r
o
m
F
C
M
al
g
o
r
ith
m
an
d
o
b
s
er
v
atio
n
o
f
p
lo
ts
i
n
Fi
g
.
3
s
u
g
g
es
ts
t
h
at
a
f
in
e
d
is
tr
ib
u
tio
n
o
f
d
ata
w
a
s
e
v
id
en
ce
d
w
h
e
n
m
=1
.
6
an
d
m
=2
.
0
.
T
h
u
s
,
t
h
e
s
elec
ted
v
al
u
e
f
o
r
m
=2
.
0
s
h
o
w
ed
m
o
r
e
d
is
tr
ib
u
tio
n
o
f
d
ata.
Nea
r
l
y
5
r
u
n
s
o
f
F
C
M
al
g
o
r
ith
m
f
o
r
ea
ch
m
w
as
p
er
f
o
r
m
ed
,
a
n
d
th
e
m
e
m
b
er
s
h
ip
v
alu
e
s
an
d
ce
n
tr
o
id
s
co
r
r
esp
o
n
d
in
g
t
o
th
e
lo
w
e
s
t
v
alu
e
w
as
r
ep
o
r
ted
,
w
h
ic
h
allo
w
ed
to
s
ca
n
t
h
e
e
n
tire
d
ata
s
p
ac
e
f
o
r
b
est o
u
tco
m
e.
g.
FC
M
o
u
tp
u
t
Fu
zz
y
c
-
m
ea
n
s
cl
u
s
ter
in
g
w
i
th
3
clu
s
ter
s
as
s
h
o
w
n
in
T
ab
le
8
an
d
9
.
Fro
m
t
h
e
o
u
tp
u
t
o
f
m
e
m
b
er
s
h
i
p
co
ef
f
icie
n
t
s
a
n
d
Fi
g
.
6
,
i
t
i
s
o
b
s
er
v
ed
th
a
t
p
o
in
t
1
3
ap
p
ea
r
s
to
b
e
ass
o
ciate
d
w
i
th
t
h
e
eq
u
i
m
e
m
b
er
s
h
ip
v
ec
to
r
(
0
.
4
2
an
d
0
.
4
9
)
b
etw
ee
n
clu
s
t
er
1
an
d
clu
s
ter
2
r
esp
ec
tiv
el
y
.
C
o
m
p
ar
is
o
n
o
f
o
u
tp
u
t
cl
u
s
te
r
s
f
r
o
m
f
a
n
n
y
an
d
f
u
zz
y
c
-
m
ea
n
s
r
ep
o
r
ted
in
Fi
g
.
4
a
n
d
6
s
u
g
g
es
t
t
h
at
FC
M
o
u
tp
er
f
o
r
m
s
t
h
e
m
eth
o
d
.
I
t
s
h
o
u
ld
b
e
n
o
ted
t
h
at
o
b
j
ec
t 1
6
b
elo
n
g
s
to
clu
s
ter
1
in
f
a
n
n
y
w
h
er
ea
s
it a
p
p
ea
r
ed
u
n
d
er
clu
s
ter
-
2.
T
ab
le
8
.
C
lu
s
ter
c
en
t
r
es
C
i
t
e
sco
r
e
C
o
u
n
t
S
N
I
P
S
JR
1
0
.
4
3
5
1
6
2
9
-
0
.
4
5
9
4
6
0
9
0
.
6
5
8
9
1
3
0
0
.
2
6
0
2
4
9
5
2
1
.
1
2
2
3
9
5
5
1
.
7
1
5
9
5
9
9
0
.
1
3
9
4
1
9
5
0
.
9
1
5
6
1
0
9
3
-
0
.
9
6
7
7
2
7
7
-
0
.
3
8
4
4
4
6
4
-
0
.
8
6
8
1
5
5
4
-
0
.
8
0
7
4
9
4
0
T
ab
le
9
.
Me
m
b
er
s
h
ip
s
1
2
3
1
2
3
[
1
,
]
0
.
0
8
9
7
5
0
2
3
0
.
6
8
0
1
6
9
9
9
4
0
.
2
3
0
0
7
9
7
8
[
1
4
,
]
0
.
2
5
8
2
8
3
59
0
.
5
3
4
0
4
0
2
6
2
0
.
2
0
7
6
7
6
1
5
[
2
,
]
0
.
0
4
0
5
2
1
4
9
0
.
8
7
9
0
5
1
4
4
3
0
.
0
8
0
4
2
7
0
7
[
1
5
,
]
0
.
2
4
0
9
9
7
8
3
0
.
0
8
7
7
3
4
8
3
4
0
.
6
7
1
2
6
7
3
4
[
3
,
]
0
.
1
0
7
6
6
6
2
9
0
.
1
3
5
1
0
5
0
5
7
0
.
7
5
7
2
2
8
6
5
[
1
6
,
]
0
.
2
1
5
3
7
0
9
1
0
.
3
5
0
2
4
4
0
5
1
0
.
4
3
4
3
8
5
0
4
[
4
,
]
0
.
0
6
7
1
3
2
5
9
0
.
1
3
1
5
9
9
9
7
1
0
.
8
0
1
2
6
7
4
4
[
1
7
,
]
0
.
2
6
5
3
4
9
2
3
0
.
0
7
7
2
1
1
8
7
2
0
.
6
5
7
4
3
8
9
0
[
5
,
]
0
.
1
0
1
0
5
0
1
0
0
.
0
8
9
1
9
2
2
6
0
0
.
8
0
9
7
5
7
6
4
[
1
8
,
]
0
.
8
9
7
2
9
4
6
1
0
.
0
3
7
8
4
5
1
4
9
0
.
0
6
4
8
6
0
2
4
[
6
,
]
0
.
1
3
7
8
3
6
8
4
0
.
6
5
4
4
5
1
2
9
9
0
.
2
0
7
7
1
1
8
6
[
1
9
,
]
0
.
9
7
2
6
8
4
7
3
0
.
0
0
7
6
1
8
1
4
9
0
.
0
1
9
6
9
7
1
2
[
7
,
]
0
.
0
9
0
9
4
6
8
3
0
.
0
8
4
6
6
7
8
8
4
0
.
8
2
4
3
8
5
2
9
[
2
0
,
]
0
.
8
2
3
5
7
3
8
4
0
.
0
4
0
0
6
7
8
0
3
0
.
1
3
6
3
5
8
3
6
[
8
,
]
0
.
1
0
5
2
5
9
8
6
0
.
0
7
3
0
7
5
9
3
1
0
.
8
2
1
6
6
4
2
1
[
2
1
,
]
0
.
8
5
1
8
7
4
9
5
0
.
0
3
3
7
5
4
7
4
4
0
.
1
1
4
3
7
0
3
0
[
9
,
]
0
.
1
0
6
3
9
5
6
0
0
.
7
3
2
3
2
2
9
8
2
0
.
1
6
1
2
8
1
4
2
[
2
2
,
]
0
.
9
6
4
8
2
5
6
7
0
.
0
1
1
1
2
4
2
7
3
0
.
0
2
4
0
5
0
0
5
[
1
0
,
]
0
.
0
2
6
8
7
4
9
8
0
.
0
2
3
9
9
6
2
4
8
0
.
9
4
9
1
2
8
7
7
[
2
3
,
]
0
.
9
6
7
8
0
9
7
5
0
.
0
0
9
8
5
5
4
4
6
0
.
0
2
2
3
3
4
8
1
[
1
1
,
]
0
.
0
2
0
7
8
8
4
1
0
.
0
1
6
7
6
0
5
4
9
0
.
9
6
2
4
5
1
0
4
[
2
4
,
]
0
.
9
4
3
9
3
3
7
9
0
.
0
1
9
4
1
1
5
5
3
0
.
0
3
6
6
5
4
6
5
[
1
2
,
]
0
.
1
8
1
4
5
6
1
3
0
.
2
0
5
3
3
2
6
3
5
0
.
6
1
3
2
1
1
2
3
[
2
5
,
]
0
.
9
4
3
3
9
5
1
4
0
.
0
1
7
4
1
8
9
5
2
0
.
0
3
9
1
8
5
9
1
[
1
3
,
]
0
.
4
2
2
7
0
3
2
7
0
.
0
8
1
0
9
1
9
1
8
0
.
4
9
6
2
0
4
8
1
[
2
6
,
]
0
.
9
1
0
6
2
3
7
2
0
.
0
3
3
0
4
3
8
0
2
0
.
0
5
6
3
3
2
4
8
Fig
u
r
e
6
.
I
m
a
g
e
s
h
o
w
i
n
g
F
C
M
o
u
tp
u
t
h.
T
im
e
co
m
p
le
x
it
y
T
h
e
tim
e
co
m
p
le
x
it
y
[
2
5
]
o
f
f
u
zz
y
cl
u
s
ter
i
n
g
(
f
a
n
n
y
)
an
d
f
u
zz
y
c
-
m
ea
n
s
al
g
o
r
ith
m
s
w
er
e
ev
alu
a
ted
b
y
k
ee
p
in
g
d
ata
p
o
in
t
’
s
co
n
s
ta
n
t a
n
d
v
ar
y
i
n
g
n
u
m
b
er
o
f
cl
u
s
ter
s
.
T
h
e
d
etails
w
er
e
p
r
esen
te
d
in
T
ab
le
1
0
.
T
ab
le
1
0
.
T
im
e
co
m
p
lex
i
t
y
a
n
a
l
y
s
i
s
o
f
f
an
n
y
an
d
F
C
M
a
lg
o
r
ith
m
s
A
l
g
o
r
i
t
h
m
N
u
mb
e
r
o
f
c
l
u
st
e
r
s (k
)
Ex
e
c
u
t
e
d
t
i
me
i
n
se
c
s
A
l
g
o
r
i
t
h
m
N
u
mb
e
r
o
f
c
l
u
st
e
r
s (k
)
Ex
e
c
u
t
e
d
t
i
me
i
n
se
c
s
F
a
n
n
y
1
0
.
7
4
F
C
M
1
0
.
7
2
2
0
.
7
9
2
0
.
7
6
3
0
.
8
1
3
0
.
7
6
4
0
.
8
2
4
0
.
7
9
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
9
,
No
.
4
,
A
u
g
u
s
t 2
0
1
9
:
2
7
6
0
-
2770
2768
i.
Ass
es
s
i
n
g
q
u
alit
y
o
f
cl
u
s
ter
s
O
n
t
h
e
o
t
h
er
h
a
n
d
,
v
a
lid
it
y
o
f
f
u
zz
y
cl
u
s
ter
i
n
g
w
as
p
r
o
p
o
s
ed
b
y
Gat
h
a
n
d
Gev
a
(
1
9
8
9
)
[
2
1
]
in
ap
p
licatio
n
s
to
m
i
x
tu
r
e
s
o
f
n
o
r
m
al
d
is
tr
ib
u
tio
n
s
.
Di
s
tr
ib
u
ted
h
ier
ar
ch
ical
cl
u
s
ter
v
alid
it
y
was p
r
o
p
o
s
ed
b
y
Xie
an
d
B
en
i
[
2
6
].
I
n
g
e
n
er
al,
t
h
e
clu
s
ter
s
ar
e
n
o
t
cr
is
p
i
n
n
at
u
r
e,
f
o
r
e
x
a
m
p
le,
a
n
o
b
j
ec
t
in
o
n
e
cl
u
s
ter
m
a
y
n
o
t
ab
s
o
lu
tel
y
b
elo
n
g
to
a
p
ar
tic
u
lar
cl
u
s
ter
a
n
d
h
o
w
e
v
er
,
it
m
a
y
co
n
tai
n
th
e
p
r
o
p
er
t
y
o
f
o
th
er
clu
s
ter
.
He
n
ce
,
in
tr
o
d
u
ctio
n
o
f
f
u
zz
i
n
e
s
s
co
n
c
ep
t
w
o
u
ld
g
e
n
er
ate
s
o
f
t
clu
s
te
r
in
g
o
f
o
b
j
ec
ts
r
ath
er
e
m
p
lo
y
i
n
g
h
ar
d
clu
s
ter
i
n
g
m
et
h
o
d
s
s
u
ch
a
s
k
-
m
ea
n
s
,
P
A
M
etc.
Her
e,
w
e
e
m
p
lo
y
ed
t
w
o
f
u
zz
y
cl
u
s
ter
i
n
g
m
o
d
els
s
u
c
h
as
f
a
n
n
y
[
2
7
]
,
f
u
zz
y
c
-
m
ea
n
s
(
FC
M)
[
2
8
]
tech
n
iq
u
es to
cl
u
s
ter
an
d
f
in
d
o
u
t
th
e
v
ar
iab
ilit
y
a
m
o
n
g
th
e
s
e
m
eth
o
d
s
.
j.
Valid
atio
n
1.
Fan
n
y
–
i
n
ter
n
al
v
alid
atio
n
I
n
ter
n
al
v
alid
atio
n
m
ea
s
u
r
es
in
cl
u
d
e,
co
m
p
ac
t
n
es
s
,
co
n
n
ec
ted
n
ess
a
n
d
s
ep
ar
atio
n
o
f
th
e
cl
u
s
ter
p
ar
titi
o
n
s
.
T
h
e
y
ar
e
o
th
er
w
i
s
e
r
ef
er
r
ed
as
C
o
n
n
ec
ti
v
it
y
,
S
ilh
o
u
et
te
w
id
th
a
n
d
D
u
n
n
i
n
d
ex
.
T
h
e
p
r
o
g
r
am
r
esu
lted
in
t
h
e
f
o
llo
w
in
g
o
u
t
p
u
t
w
h
er
e
all
th
e
th
r
ee
m
ea
s
u
r
es
a
r
e
w
it
h
i
n
th
e
s
p
ec
if
ied
li
m
its
as
s
h
o
w
n
i
n
T
ab
le
1
1
.
T
ab
le
1
1
.
I
n
ter
n
al
v
alid
atio
n
d
ata
an
d
th
eir
v
al
u
es o
u
tp
u
t
f
r
o
m
t
h
e
p
r
o
g
r
a
m
C
L
v
alid
I
n
t
e
r
n
a
l
V
a
l
i
d
a
t
i
o
n
I
n
d
e
x
e
s
V
a
l
u
e
C
o
n
n
e
c
t
i
v
i
t
y
1
2
.
6
2
2
2
D
u
n
n
0
.
1
8
4
3
S
i
l
h
o
u
e
t
t
e
0
.
4
2
1
0
2.
Fan
n
y
–
s
tab
ili
t
y
v
a
ld
atio
n
T
h
e
s
tab
ilit
y
m
ea
s
u
r
es
ar
e
ca
lcu
lated
u
s
i
n
g
A
P
N,
A
D,
AD
M,
an
d
FOM.
A
ll
m
ea
s
u
r
e
s
s
h
o
u
ld
r
esu
l
t
in
m
i
n
i
m
u
m
v
al
u
e
s
.
T
h
ese
m
et
h
o
d
s
ta
k
e
co
m
p
u
tatio
n
ti
m
e
d
u
e
to
t
h
e
f
ac
t
t
h
at
th
e
p
r
o
g
r
a
m
r
e
-
c
lu
s
ter
s
a
d
ataset
b
y
o
m
itt
in
g
s
i
n
g
le
co
lu
m
n
.
T
h
e
o
u
tp
u
t
o
f
th
e
p
r
o
g
r
a
m
is
g
i
v
en
i
n
T
ab
le
1
2
.
I
t
ca
n
b
e
o
b
s
er
v
ed
f
r
o
m
th
e
o
u
tp
u
t d
ata
as sh
o
w
n
i
n
T
a
b
les 1
1
an
d
1
2
th
at
all
in
ter
n
al
an
d
s
tab
ilit
y
m
ea
s
u
r
es a
r
e
w
it
h
in
t
h
e
li
m
it
s
.
T
ab
le
1
2
.
Stab
ilit
y
v
alid
atio
n
m
et
h
o
d
s
an
d
r
esp
ec
tiv
e
s
co
r
es o
f
ea
c
h
m
et
h
o
d
s
ig
n
i
f
y
i
n
g
ac
ce
p
tab
le
li
m
it
s
V
a
l
i
d
a
t
i
o
n
me
t
h
o
d
S
c
o
r
e
A
P
N
0
.
1
1
8
2
9
8
4
AD
1
.
5
0
4
8
4
4
4
A
D
M
0
.
3
8
3
4
0
5
6
F
O
M
0
.
8
0
6
9
5
7
5
3.
Valid
atio
n
:
f
u
zz
y
c
-
m
ea
n
s
(
F
C
M)
T
h
e
v
alu
e
s
o
f
t
h
e
i
n
d
ex
e
s
ca
n
b
e
i
n
d
ep
en
d
en
t
l
y
u
s
ed
i
n
o
r
d
er
to
ev
alu
ate
an
d
co
m
p
ar
e
clu
s
ter
in
g
p
ar
titi
o
n
s
o
r
ev
e
n
to
d
ete
r
m
in
e
th
e
n
u
m
b
er
o
f
c
lu
s
ter
s
e
x
is
t
i
n
g
in
a
d
ata
s
et
a
s
s
h
o
w
n
i
n
T
ab
les
1
3
an
d
1
4
.
I
t is th
er
ef
o
r
e
u
n
d
er
s
to
o
d
th
at
all
v
alid
atio
n
p
ar
a
m
e
ter
s
ar
e
w
it
h
i
n
t
h
e
li
m
its
.
T
ab
le
1
3
.
E
s
ti
m
ated
G
A
T
H
an
d
GE
VA
cl
u
s
ter
v
a
lid
it
y
m
ea
s
u
r
es o
f
citesco
r
e
s
u
b
s
et
d
ata
s
et
wi
th
K=
2
,
3
,
4
an
d
M=
2
C
l
u
st
e
r
si
z
e
F
H
V
D
P
A
DP
2
0
.
7
4
5
5
1
5
.
7
4
4
1
1
.
1
6
5
3
1
.
0
9
4
3
1
5
.
8
0
5
6
.
8
1
9
4
1
.
3
2
6
1
4
.
0
0
3
6
.
5
7
8
T
ab
le
1
4
.
E
s
ti
m
ated
clu
s
ter
v
a
lid
it
y
i
n
d
ex
es
of
citesco
r
e
s
u
b
s
et
d
ata
s
et
w
i
th
K=
2
,
3
,
4
an
d
M=
2
C
l
u
st
e
r
si
z
e
XB
FS
PC
PE
SI
2
0
.
0
1
1
-
2
3
.
8
9
4
0
.
7
1
9
0
.
4
3
7
0
.
1
9
3
3
0
.
0
0
6
-
4
4
.
1
9
4
0
.
6
8
7
0
.
5
7
0
0
.
2
2
1
4
0
.
0
3
3
-
4
0
.
1
1
6
0
.
5
6
8
0
.
8
1
4
0
.
2
2
1
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2
0
8
8
-
8708
F
u
z
z
y
clu
s
teri
n
g
a
n
d
F
u
z
z
y
C
-
Mea
n
s
p
a
r
titi
o
n
clu
s
ter
a
n
a
lysi
s
a
n
d
va
lid
a
tio
n
…
(
K
.
V
a
r
a
d
a
R
a
jku
ma
r
)
2769
4.
CO
NCLU
SI
O
N
C
lu
s
ter
a
n
al
y
s
i
s
i
n
itia
ted
o
n
a
s
u
b
s
et
o
f
cite
s
co
r
e
d
ataset
u
s
i
n
g
f
u
zz
y
clu
s
ter
i
n
g
an
d
f
u
zz
y
c
-
m
ea
n
s
alg
o
r
ith
m
r
es
u
lted
in
ap
p
ea
r
an
ce
o
f
eq
u
id
is
ta
n
t
p
o
in
t
s
in
t
h
e
b
o
u
n
d
ar
y
o
f
cl
u
s
ter
s
w
h
ic
h
is
ev
id
en
ce
d
f
r
o
m
m
e
m
b
er
s
h
ip
d
eg
r
ee
s
o
f
ea
c
h
clu
s
ter
d
ata
p
o
in
t.
T
im
e
co
m
p
l
ex
it
y
a
n
al
y
s
i
s
r
ev
ea
led
th
at
F
C
M
p
er
f
o
r
m
s
m
u
c
h
f
aster
th
a
n
f
u
zz
y
m
e
th
o
d
.
F
u
r
th
er
,
all
i
n
ter
n
al
a
n
d
s
tab
il
it
y
m
ea
s
u
r
e
p
r
o
ce
d
u
r
es
o
f
f
u
zz
y
clu
s
ter
in
g
a
n
d
al
l
v
alid
it
y
in
d
e
x
es o
f
F
C
M
ar
e
f
o
u
n
d
to
b
e
w
it
h
i
n
th
e
li
m
i
ts
.
RE
F
E
R
E
NC
E
S
[1
]
Ba
b
u
,
G
.
P
.
,
M
u
rty
,
M
.
N,
“
Clu
ste
r
in
g
w
it
h
e
v
o
lu
ti
o
n
stra
teg
ies
,
”
Pa
tt
e
rn
Rec
o
g
n
it
i
o
n
,
Vo
l.
2
7
,
p
p
.
3
2
1
–
3
2
9
,
1
9
9
4
.
[2
]
Be
z
d
e
k
,
J.C.
,
Ke
ll
e
r,
J.,
Krish
n
a
p
u
ra
m
,
R.
,
P
a
l,
N.R,
“
F
u
z
z
y
M
o
d
e
ls
a
n
d
A
lg
o
rit
h
m
s
f
o
r
P
a
tt
e
rn
Re
c
o
g
n
it
io
n
a
n
d
Im
a
g
e
P
ro
c
e
ss
in
g
,
”
T
h
e
Ha
n
d
b
o
o
k
s o
f
F
u
z
z
y
S
e
ts,
V
o
l.
4
Klu
w
e
r,
Bo
sto
n
a
n
d
L
o
n
d
o
n
,
1
9
9
9
.
[3
]
Zad
e
h
.
L
.
A
,
“
F
u
z
z
y
se
ts,”
In
fo
rm
a
ti
o
n
a
n
d
Co
n
tro
l
,
V
o
l
.
8
,
Iss
.
3
,
p
p
.
3
3
8
–
3
5
3
,
1
9
6
5
.
[4
]
Dö
rin
g
,
C.
,
B
o
rg
e
lt
,
C.
,
Kr
u
se
R,
“
F
u
z
z
y
c
lu
ste
rin
g
o
f
q
u
a
n
ti
t
a
ti
v
e
a
n
d
q
u
a
li
tativ
e
d
a
ta,”
Fu
z
zy
In
fo
rm
a
ti
o
n
,
Pro
c
e
ss
in
g
NAF
IP
S
‘
0
4
,
IEE
E
,
P
P
.
8
4
-
8
9
.
[5
]
Ru
sp
in
i,
E
.
H,
“
A
n
e
w
a
p
p
ro
a
c
h
to
c
lu
ste
rin
g
,
”
I
n
fo
rm
a
ti
o
n
a
n
d
Co
n
tro
l
,
Vo
l.
1
5
Iss
u
e
1
,
p
p
.
2
2
–
3
2
,
1
9
6
9
.
[6
]
Da
v
é
,
R,
“
Ch
a
ra
c
teriz
a
ti
o
n
a
n
d
d
e
tec
ti
o
n
o
f
n
o
ise
in
c
lu
ste
rin
g
,
”
Pa
tt
e
rn
Rec
o
g
n
it
i
o
n
L
e
tt
e
rs
,
Vo
l.
1
2
,
p
p
.
6
5
7
–
6
6
4
,
1
9
9
1
.
[7
]
Jo
o
n
s
u
n
g
P
a
rk
,
Do
o
He
o
n
S
o
n
g
,
Ho
su
n
g
Nh
o
,
Hy
u
n
-
M
in
Ch
o
i,
K
y
u
n
g
-
A
e
Ki
m
,
H
y
u
n
Ju
n
P
a
rk
,
Kw
a
n
g
Ba
e
k
Ki
m
,
“
A
u
to
m
a
ti
c
S
e
g
m
e
n
tatio
n
o
f
Bra
c
h
ial
A
rter
y
b
a
se
d
o
n
F
u
z
z
y
C
-
M
e
a
n
s
P
ix
e
l
Clu
ste
ri
n
g
f
ro
m
Ultras
o
u
n
d
Im
a
g
e
s
,
”
In
ter
n
a
t
io
n
a
l
J
o
u
rn
a
l
o
f
E
lec
trica
l
a
n
d
C
o
mp
u
ter
En
g
in
e
e
rin
g
(
IJ
ECE
)
Vo
l.
8
,
N
o
.
2
,
p
p
.
6
3
8
-
6
4
3
,
2
0
1
8
.
[8
]
Am
it
Kr.
K
a
u
sh
ik
,
“
A
H
y
b
rid
A
p
p
ro
a
c
h
o
f
F
u
z
z
y
C
-
M
e
a
n
s
C
lu
ste
rin
g
a
n
d
Ne
u
ra
l
Ne
tw
o
rk
t
o
M
a
k
e
En
e
rg
y
Eff
icie
n
t
He
t
e
ro
g
e
n
e
o
u
s
W
irele
ss
S
e
n
so
r
Ne
tw
o
rk
,
”
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
El
e
c
trica
l
a
n
d
C
o
mp
u
ter
En
g
i
n
e
e
rin
g
(
IJ
ECE
)
V
o
l.
6
,
N
o
.
2
,
p
p
.
6
7
4
-
6
8
1
,
2
0
1
6
.
[9
]
De
m
p
ste
r
A
,
L
a
ird
N
a
n
d
R
u
b
i
n
D,
“
M
a
x
i
m
u
m
li
k
e
li
h
o
o
d
f
ro
m
in
c
o
m
p
lete
d
a
ta
v
ia
th
e
e
m
a
l
g
o
rit
h
m
,
”
J
o
u
rn
a
l
o
f
th
e
Ro
y
a
l
S
ta
ti
st
ica
l
S
o
c
iety
,
V
o
l.
3
9
S
e
ries
B
,
p
p
.
1
–
3
8
,
1
9
7
7
.
[1
0
]
Du
n
n
JC,
“
A
f
u
z
z
y
re
l
a
ti
v
e
o
f
th
e
iso
-
d
a
ta p
ro
c
e
ss
a
n
d
it
s u
se
in
d
e
tec
ti
n
g
c
o
m
p
a
c
t
w
e
ll
-
se
p
a
ra
ted
c
lu
ste
rs,”
J
o
u
rn
a
l
o
f
Cy
b
e
rn
e
ti
c
s
,
V
o
l
-
3
,
p
p
.
3
2
–
5
7
,
1
9
7
3
.
[1
1
]
P
e
d
ry
c
z
W
,
“
Co
ll
a
b
o
ra
ti
v
e
f
u
z
z
y
c
lu
ste
rin
g
,
”
Pa
tt
e
rn
Rec
o
g
n
it
i
o
n
L
e
tt
e
rs
,
V
o
l.
2
3
,
Iss
u
e
1
4
,
p
p
.
1
6
7
5
–
8
6
,
2
0
0
2
.
[1
2
]
Ha
n
d
l
J,
Kn
o
w
les
J,
Ke
ll
DB,
“
Co
m
p
u
tatio
n
a
l
Clu
ste
r
V
a
li
d
a
ti
o
n
in
P
o
st
-
G
e
n
o
m
ic
Da
ta
A
n
a
l
y
sis,
”
Bi
o
in
fo
rm
a
ti
c
s,
V
o
l
.
2
1
Iss
u
e
1
5
,
p
p
.
3
2
0
1
–
1
2
,
2
0
0
5
.
[1
3
]
Du
n
n
JC,
“
W
e
ll
S
e
p
a
ra
ted
Clu
ste
rs an
d
F
u
z
z
y
P
a
rti
ti
o
n
s,”
J
o
u
r
n
a
l
o
n
Cy
b
e
rn
e
ti
c
s
,
Vo
l.
4
,
p
p
.
9
5
–
1
0
4
,
1
9
7
4
.
[1
4
]
Ro
u
ss
e
e
u
w
P
J,
“
S
il
h
o
u
e
tt
e
s:
A
G
ra
p
h
ica
l
A
id
to
th
e
In
terp
re
tati
o
n
a
n
d
V
a
li
d
a
ti
o
n
o
f
Clu
ste
r
A
n
a
l
y
sis,”
J
o
u
rn
a
l
o
f
Co
mp
u
t
a
ti
o
n
a
l
a
n
d
A
p
p
li
e
d
M
a
t
h
e
ma
ti
c
s
,
2
0
,
p
p
.
5
3
–
6
5
,
1
9
8
7
.
[1
5
]
Ye
u
n
g
KY
,
Ha
y
n
o
r
DR,
Ru
z
z
o
WL
,
“
V
a
li
d
a
ti
n
g
Clu
ste
ri
n
g
f
o
r
G
e
n
e
E
x
p
re
ss
io
n
Da
ta,”
Bi
o
in
f
o
r
ma
ti
c
s
,
Vo
l.
1
7
Iss
u
e
4
,
p
p
.
3
0
9
–
1
8
,
2
0
0
1
.
[1
6
]
Zad
e
h
.
L
.
A
,
“
F
u
z
z
y
se
ts an
d
s
y
ste
m
s,”
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
Ge
n
e
ra
l
S
y
ste
ms
,
V
o
l
.
1
7
,
p
p
.
1
2
9
-
1
3
8
,
1
9
6
5
.
[1
7
]
Be
z
d
e
k
.
J.C,
“
P
a
tt
e
rn
Re
c
o
g
n
i
ti
o
n
w
it
h
F
u
z
z
y
Ob
jec
ti
v
e
F
u
n
c
ti
o
n
A
lg
o
rit
h
m
s,”
Pl
e
n
u
m P
re
ss
,
Ne
w
Yo
rk
,
1
9
8
1
.
[1
8
]
Nik
h
il
R.
P
a
l
,
Ja
m
e
s
C.
Be
z
d
e
k
,
a
n
d
Rich
a
r
d
J.
Ha
t
h
a
w
a
y
,
“
S
e
q
u
e
n
t
ial
Co
m
p
e
ti
ti
v
e
L
e
a
rn
in
g
a
n
d
t
h
e
F
u
z
z
y
c
-
M
e
a
n
s Clu
ste
rin
g
A
lg
o
rit
h
m
s,”
Ne
u
ra
l
Ne
two
rk
s
,
V
o
l.
9
,
No
.
5
,
p
p
.
7
8
7
-
7
9
6
,
1
9
9
6
.
[1
9
]
Ch
e
n
.
S
,
Z
h
a
n
g
.
D,
“
Ro
b
u
st
im
a
g
e
se
g
m
e
n
tatio
n
u
si
n
g
F
CM
w
it
h
sp
a
ti
a
l
c
o
n
stra
i
n
ts
b
a
se
d
o
n
n
e
w
k
e
rn
e
l
-
in
d
u
c
e
d
d
istan
c
e
m
e
a
su
re
,
”
IEE
E
T
ra
n
sa
c
ti
o
n
s
o
n
S
y
ste
ms
,
M
a
n
a
n
d
Cy
b
e
r
n
e
ti
c
s
,
Vo
l.
3
4
,
p
p
.
1
9
0
7
-
1
9
1
6
,
1
9
9
8
.
[2
0
]
S
e
n
g
u
p
ta.S
,
S
.
De
,
K
o
n
a
r.
A
a
n
d
Ja
n
a
rth
a
n
a
n
.
R,
“
A
n
im
p
ro
v
e
d
f
u
z
z
y
c
lu
ste
rin
g
m
e
th
o
d
u
si
n
g
m
o
d
if
ied
F
u
k
u
y
a
m
a
-
S
u
g
e
n
o
c
lu
ste
r
v
a
li
d
it
y
in
d
e
x
,
”
I
n
ter
n
a
ti
o
n
a
Co
n
fer
e
n
c
e
o
n
Rec
e
n
t
T
re
n
d
s
in
I
n
fo
rm
a
t
io
n
S
y
ste
ms
,
Ko
lk
a
ta,
p
p
.
269
-
2
7
4
,
2
0
1
1
.
[2
1
]
G
a
th
,
I.
,
G
e
v
a
,
A
.
B,
“
Un
su
p
e
rv
ise
d
o
p
ti
m
a
l
f
u
z
z
y
c
lu
ste
rin
g
,
”
IEE
ET
ra
n
s
a
c
ti
o
n
s
o
n
P
a
tt
e
rn
An
a
lys
is
a
n
d
M
a
c
h
in
e
In
telli
g
e
n
c
e
,
Vo
l.
1
1
,
p
p
.
7
7
3
–
7
8
1
,
1
9
8
9
.
[2
2
]
X
ie.
X
.
L
.
,
Be
n
i.
G
,
“
A
v
a
li
d
it
y
m
e
a
su
re
f
o
r
f
u
z
z
y
c
lu
ste
rin
g
,
”
IEE
E
PA
M
I
,
Vo
l.
1
3
,
No
.
8
,
p
p
.
8
4
1
–
8
4
7
,
1
9
9
1
.
[2
3
]
Ba
n
e
rjee
A
,
Da
v
e
RN,
“
V
a
li
d
a
ti
n
g
c
lu
ste
rs
u
sin
g
th
e
Ho
p
k
in
s
S
t
a
ti
sti
c
,
”
Fu
zz
y
S
y
ste
ms
,
Pro
c
e
e
d
in
g
s
2
0
0
4
IEE
E
in
ter
n
a
t
io
n
a
l
c
o
n
fer
e
n
c
e
,
V
o
l.
1
,
p
p
.
1
4
9
-
1
5
3
,
2
0
0
4
.
[2
4
]
Do
u
lay
e
D
e
m
b
e
l,
P
h
i
li
p
p
e
Ka
stn
e
r,
“
F
u
z
z
y
C
-
m
e
a
n
s
m
e
th
o
d
f
o
r
c
lu
ste
rin
g
m
icro
a
rra
y
d
a
ta,”
Bi
o
in
fo
rm
a
t
is
,
1
9
(8
),
p
p
.
9
7
3
-
9
8
0
,
2
0
0
3
.
[2
5
]
L
i
m
,
A
.
,
T
jh
i,
W
,
“
R
Hig
h
P
e
rf
o
rm
a
n
c
e
P
ro
g
ra
m
m
in
g
,
”
P
a
c
k
e
t
P
u
b
li
sh
i
n
g
L
td
,
2
0
1
5
.
[2
6
]
X
ie.
X
.
L
.
,
Be
n
i.
G
,
“
Distrib
u
ted
h
iera
rc
h
ica
l
d
e
c
isio
n
f
u
sio
n
w
it
h
c
lu
ste
r
v
a
li
d
it
y
,
”
in
Pro
c
e
e
d
in
g
s
IE
EE
In
ter
n
a
t
io
n
a
l
C
o
n
fer
e
n
c
e
,
S
y
st.
M
a
n
Cy
b
e
rn
,
p
p
.
6
8
9
-
6
9
4
,
1
9
9
0
.
[2
7
]
Ka
u
fm
a
n
,
L
.
,
Ro
u
ss
e
e
u
w
,
P
.
J,
“
F
in
d
in
g
G
ro
u
p
s
in
Da
ta:
A
n
In
tr
o
d
u
c
t
io
n
t
o
Clu
ste
r
A
n
a
ly
sis,”
W
il
e
y
,
Ne
w
Yo
rk
,
1
9
9
0
.
[2
8
]
Be
z
d
e
k
,
Ja
m
e
s
C.
,
Ro
b
e
rt
E
h
rli
c
h
,
a
n
d
W
il
li
a
m
F
u
ll
,
“
F
C
M
:
T
h
e
f
u
z
z
y
c
-
m
e
a
n
s
c
lu
ste
rin
g
a
lg
o
rit
h
m
,
”
Co
mp
u
ter
s
&
Ge
o
sc
ien
c
e
s
,
V
o
l.
1
0
,
No
.
2
-
3,
p
p
.
1
9
1
-
2
0
3
,
1
9
8
4
.
Evaluation Warning : The document was created with Spire.PDF for Python.