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
.
3
,
J
u
n
e
201
9
,
p
p
.
2
1
0
3
~
2
1
1
1
I
SS
N:
2
0
8
8
-
8708
,
DOI
: 1
0
.
1
1
5
9
1
/
i
j
ec
e
.
v9
i
3
.
p
p
2
1
0
3
-
2111
2103
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
Co
m
ple
te
a
g
g
lo
m
erativ
e
h
iera
rchy
d
o
cu
m
e
nt’
s
clust
e
ring
bas
ed
o
n f
u
zz
y
L
uh
n’s g
ibbs l
a
tent
d
iric
hl
et
a
l
lo
ca
tion
P.
M
.
P
riha
t
ini
1
,
I.
K.
G.
D
.
P
utr
a
2
,
I.
A.
D.
G
iria
nta
ri
3
,
M
.
Su
da
r
m
a
4
1
S
tu
d
y
P
r
o
g
ra
m
o
f
Do
c
to
ra
l
En
g
in
e
e
rin
g
S
c
ien
c
e
,
F
a
c
u
lt
y
o
f
En
g
in
e
e
rin
g
,
U
d
a
y
a
n
a
Un
iv
e
rsit
y
,
In
d
o
n
e
sia
2
S
tu
d
y
P
r
o
g
ra
m
o
f
In
f
o
rm
a
ti
o
n
T
e
c
h
n
o
l
o
g
y
,
F
a
c
u
lt
y
o
f
En
g
in
e
e
rin
g
,
Ud
a
y
a
n
a
Un
iv
e
rsit
y
,
In
d
o
n
e
sia
3.
4
S
tu
d
y
P
ro
g
ra
m
o
f
El
e
c
tri
c
a
l
En
g
in
e
e
rin
g
,
F
a
c
u
lt
y
o
f
En
g
in
e
e
rin
g
,
Ud
a
y
a
n
a
Un
iv
e
rsit
y
,
In
d
o
n
e
sia
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
A
p
r
2
4
,
2
0
1
8
R
ev
i
s
ed
No
v
18
,
2
0
1
8
A
cc
ep
ted
Dec
1
1
,
2
0
1
8
Ag
g
lo
m
e
ra
ti
v
e
h
iera
rc
h
ica
l
is
a
b
o
tt
o
m
u
p
c
lu
ste
rin
g
m
e
th
o
d
,
w
h
e
re
th
e
d
istan
c
e
s
b
e
tw
e
e
n
d
o
c
u
m
e
n
ts
c
a
n
b
e
re
tri
e
v
e
d
b
y
e
x
trac
ti
n
g
fe
a
t
u
re
v
a
lu
e
s
u
sin
g
a
t
o
p
ic
-
b
a
se
d
late
n
t
d
iri
c
h
l
e
t
a
l
l
o
c
a
ti
o
n
m
e
th
o
d
.
T
o
re
d
u
c
e
t
h
e
n
u
m
b
e
r
o
f
fe
a
tu
re
s,
ter
m
se
lec
ti
o
n
c
a
n
b
e
d
o
n
e
u
sin
g
L
u
h
n
’s
Id
e
a
.
T
h
o
s
e
m
e
th
o
d
s
c
a
n
b
e
u
se
d
to
b
u
il
d
th
e
b
e
tt
e
r
c
lu
ste
rs
f
o
r
d
o
c
u
m
e
n
t.
Bu
t,
th
e
re
is
l
e
ss
re
se
a
rc
h
d
isc
u
ss
e
s
it
.
T
h
e
re
f
o
re
,
in
th
is
re
se
a
rc
h
,
th
e
term
w
e
ig
h
ti
n
g
c
a
lcu
latio
n
u
se
s
L
u
h
n
’s
Id
e
a
to
s
e
lec
t
th
e
ter
m
s
b
y
d
e
f
in
in
g
u
p
p
e
r
a
n
d
lo
w
e
r
c
u
t
-
o
f
f
,
a
n
d
th
e
n
e
x
trac
ts
th
e
f
e
a
tu
re
o
f
ter
m
s
u
sin
g
g
ib
b
s
sa
m
p
li
n
g
late
n
t
d
iri
c
h
l
e
t
a
l
lo
c
a
ti
o
n
c
o
m
b
in
e
d
w
it
h
term
f
re
q
u
e
n
c
y
a
n
d
f
u
z
z
y
S
u
g
e
n
o
m
e
th
o
d
.
T
h
e
fe
a
tu
re
v
a
lu
e
s
u
se
d
to
b
e
th
e
d
istan
c
e
b
e
tw
e
e
n
d
o
c
u
m
e
n
ts,
a
n
d
c
lu
ste
re
d
w
it
h
sin
g
le,
c
o
m
p
lete
a
n
d
a
v
e
ra
g
e
li
n
k
a
lg
o
rit
h
m
.
T
h
e
e
v
a
lu
a
ti
o
n
s
sh
o
w
th
e
f
e
a
tu
re
e
x
trac
ti
o
n
w
it
h
a
n
d
w
it
h
o
u
t
l
o
w
e
r
c
u
t
-
o
f
f
h
a
v
e
les
s
d
if
fe
re
n
c
e
.
Bu
t,
t
h
e
to
p
ic
d
e
term
in
a
ti
o
n
o
f
e
a
c
h
term
b
a
se
d
o
n
term
f
re
q
u
e
n
c
y
a
n
d
f
u
z
z
y
S
u
g
e
n
o
m
e
th
o
d
is
b
e
tt
e
r
th
a
n
T
su
k
a
m
o
to
m
e
th
o
d
in
f
in
d
in
g
m
o
re
re
lev
a
n
t
d
o
c
u
m
e
n
ts.
T
h
e
u
se
d
o
f
lo
w
e
r
c
u
t
-
o
f
f
a
n
d
f
u
z
z
y
S
u
g
e
n
o
g
ib
b
s
late
n
t
d
iri
c
h
l
e
t
al
lo
c
a
ti
o
n
f
o
r
c
o
m
p
lete
a
g
g
lo
m
e
r
a
ti
v
e
h
iera
rc
h
ica
l
c
lu
ste
rin
g
h
a
v
e
c
o
n
siste
n
t
m
e
tri
c
v
a
lu
e
s.
T
h
is
c
lu
ste
rin
g
m
e
th
o
d
su
g
g
e
ste
d
a
s
a
b
e
tt
e
r
m
e
th
o
d
i
n
c
lu
ste
rin
g
d
o
c
u
m
e
n
ts t
h
a
t
is m
o
re
re
lev
a
n
t
to
it
s g
o
l
d
sta
n
d
a
rd
.
K
ey
w
o
r
d
s
:
F
u
zz
y
s
u
g
e
n
o
G
ib
b
s
s
a
m
p
li
n
g
H
ier
ar
ch
ical
c
lu
s
ter
i
n
g
L
ate
n
t
d
ir
ich
l
et
a
l
lo
ca
tio
n
L
u
h
n
’
s
id
ea
C
o
p
y
rig
h
t
©
2
0
1
9
In
stit
u
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
:
P.
M.
P
r
ih
atin
i,
Stu
d
y
P
r
o
g
r
a
m
o
f
Do
cto
r
al
E
n
g
in
ee
r
i
n
g
Scie
n
ce
,
Facu
lt
y
o
f
E
n
g
i
n
ee
r
i
n
g
,
Ud
a
y
an
a
U
n
i
v
er
s
it
y
,
Ka
m
p
u
s
UNUD
B
u
k
it J
i
m
b
ar
an
Ku
ta
Selata
n
,
B
ad
u
n
g
,
B
ali,
I
n
d
o
n
esia,
8
0
3
6
1
.
E
m
ail:
m
a
n
ik
p
r
i
h
ati
n
i@
p
n
b
.
ac
.
id
1.
I
NT
RO
D
UCT
I
O
N
C
lu
s
ter
i
n
g
is
o
n
e
o
f
th
e
tas
k
s
i
n
d
ata
m
i
n
in
g
to
an
al
y
ze
lar
g
e
a
m
o
u
n
t
s
o
f
d
ata
an
d
i
s
ab
le
t
o
g
en
er
ate
h
id
d
en
in
f
o
r
m
at
io
n
th
a
t
is
v
er
y
u
s
e
f
u
l
f
o
r
d
ec
is
io
n
m
a
k
i
n
g
.
C
lu
s
ter
in
g
is
a
n
u
n
s
u
p
er
v
is
ed
clas
s
if
icatio
n
tech
n
iq
u
e
th
at
g
r
o
u
p
in
g
d
ata
w
it
h
s
i
m
ilar
itie
s
in
to
a
clu
s
ter
[
1
]
.
T
h
e
tech
n
iq
u
e
s
in
clu
s
ter
i
n
g
in
cl
u
d
e
p
ar
titi
o
n
in
g
,
h
ier
ar
ch
ical,
g
r
i
d
-
b
ased
,
an
d
d
en
s
i
t
y
-
b
ased
m
et
h
o
d
[
2
]
.
T
h
ese
m
et
h
o
d
s
h
av
e
b
ee
n
u
s
ed
i
n
v
ar
io
u
s
ap
p
licatio
n
s
s
u
c
h
as
K
-
Me
a
n
s
f
o
r
SME
R
is
k
An
al
y
s
i
s
Do
cu
m
en
t
s
[
3
]
,
KPr
o
to
ty
p
e
f
o
r
clu
s
ter
in
g
b
ig
d
ata
b
ased
o
n
Ma
p
R
ed
u
ce
[
4
]
,
K
-
Me
an
s
f
o
r
ea
r
th
q
u
ak
e
clu
s
ter
an
al
y
s
is
[
5
]
,
W
ar
d
’
s
l
in
k
a
g
e
m
e
th
o
d
f
o
r
class
i
f
y
in
g
th
e
l
a
n
g
u
a
g
es
[
6
]
,
g
r
id
an
d
d
e
n
s
it
y
-
b
ased
f
o
r
tr
aj
ec
to
r
y
cl
u
s
ter
i
n
g
[
7
]
,
an
d
DB
SC
AN
f
o
r
ca
teg
o
r
izin
g
d
is
tr
ict
s
[
8
]
.
T
h
e
h
ier
ar
ch
ical
cl
u
s
ter
i
n
g
m
et
h
o
d
u
s
es
a
tr
ee
co
n
ce
p
t,
w
h
ich
is
d
iv
id
ed
in
to
ag
g
lo
m
er
ati
v
e
an
d
d
iv
i
s
i
v
e
ap
p
r
o
ac
h
es
[
9
]
.
A
g
g
lo
m
er
ati
v
e
i
s
k
n
o
w
n
a
s
b
o
tto
m
u
p
m
et
h
o
d
,
w
h
ile
d
i
v
is
iv
e
i
s
k
n
o
w
n
as
to
p
d
o
w
n
m
et
h
o
d
[
1
0
]
.
I
n
g
en
er
al,
ag
g
lo
m
er
ati
v
e
al
g
o
r
ith
m
h
av
e
t
h
r
ee
c
h
ar
ac
ter
is
tics
s
u
c
h
a
s
s
in
g
le
lin
k
,
co
m
p
lete
l
in
k
,
a
n
d
av
er
a
g
e
l
in
k
[
1
1
]
.
T
h
e
d
if
f
er
en
ce
b
et
w
ee
n
th
e
s
e
al
g
o
r
ith
m
s
is
h
o
w
to
d
eter
m
in
e
th
e
d
is
ta
n
ce
b
et
w
ee
n
d
ata
t
h
at
w
i
ll b
e
m
er
g
ed
.
T
h
e
d
ata
th
at
w
ill
b
e
m
er
g
ed
h
a
s
d
iv
er
s
e
f
o
r
m
s
s
u
ch
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
.
3
,
J
u
n
e
2
0
1
9
:
2
1
0
3
-
2
1
1
1
2104
as
tex
t,
i
m
ag
e
s
,
s
o
u
n
d
o
r
v
id
eo
.
Fo
r
tex
t
-
s
h
ap
ed
d
ata,
th
e
tex
t
is
p
r
o
ce
s
s
ed
f
ir
s
t
t
h
r
o
u
g
h
s
ev
er
al
s
tep
s
s
u
ch
a
s
to
k
en
izat
io
n
,
f
ilter
in
g
,
le
m
m
at
izatio
n
o
r
s
te
m
m
i
n
g
[
1
2
]
.
T
h
e
r
esu
lts
o
f
tex
t
p
r
o
ce
s
s
i
n
g
w
ill
b
e
u
s
ed
to
g
en
er
ate
te
r
m
s
o
f
i
n
d
ex
i
n
g
,
w
h
ic
h
is
a
v
o
ca
b
u
lar
y
ex
tr
ac
ted
in
th
e
co
llectio
n
o
f
t
ex
ts
,
an
d
d
eter
m
i
n
e
a
w
eig
h
t
f
o
r
ea
ch
ter
m
[
1
3
]
.
T
h
e
ter
m
s
an
d
its
w
eig
h
t
s
w
ill
b
e
u
s
ed
to
d
eter
m
i
n
e
t
h
e
d
is
ta
n
ce
b
et
w
ee
n
d
ata
to
b
e
m
er
g
e
d
in
ag
g
lo
m
er
ati
v
e
al
g
o
r
ith
m
.
T
h
er
e
ar
e
s
ev
er
al
m
et
h
o
d
s
o
f
ter
m
w
ei
g
h
t
in
g
i
n
tex
t
p
r
o
ce
s
s
in
g
[
1
4
]
.
Fo
r
v
ec
t
o
r
s
p
ac
e
m
o
d
el,
t
h
er
e
is
a
co
m
m
o
n
l
y
u
s
ed
T
er
m
Fre
q
u
en
c
y
-
I
n
v
er
s
e
Do
cu
m
en
t
Fre
q
u
en
c
y
(
T
F
-
I
DF)
[
1
5
]
,
[
1
6
]
.
T
o
o
v
er
co
m
e
th
e
w
ea
k
n
ess
o
f
T
F
-
I
DF
i
n
ad
d
r
ess
in
g
s
y
n
o
n
y
m
a
n
d
p
o
ly
s
e
m
y
in
n
at
u
r
al
la
n
g
u
a
g
e,
L
ate
n
t
Se
m
a
n
tic
I
n
d
ex
i
n
g
(
L
SI)
w
a
s
d
ev
elo
p
ed
.
R
esear
ch
e
s
o
n
te
x
t
clu
s
ter
in
g
w
it
h
Ag
g
lo
m
er
ati
v
e
Hier
ar
ch
ical
C
l
u
s
ter
in
g
(
A
HC
)
al
g
o
r
ith
m
h
a
s
b
ee
n
d
o
n
e
w
it
h
t
h
o
s
e
ter
m
w
ei
g
h
t
in
g
s
c
h
e
m
es.
T
h
e
A
HC
a
lg
o
r
it
h
m
w
it
h
T
F
-
I
DF
h
as
b
ee
n
u
s
ed
to
clu
s
ter
th
e
w
eb
p
ag
e
s
[
1
7
]
,
co
n
s
tr
u
ct
tax
o
n
o
m
ies
f
r
o
m
a
co
r
p
u
s
o
f
tex
t
d
o
cu
m
e
n
ts
[
1
8
]
,
co
n
s
tr
u
ct
m
u
lti
-
k
e
y
w
o
r
d
r
an
k
ed
s
ea
r
ch
s
ch
e
m
e
[
1
9
]
,
co
n
tex
t
a
w
ar
e
d
o
cu
m
e
n
t
clu
s
ter
i
n
g
[
2
0
]
,
au
to
m
atic
ta
x
o
n
o
m
y
co
n
s
tr
u
ctio
n
f
r
o
m
k
e
y
w
o
r
d
s
[
2
1
]
.
T
h
e
A
H
C
al
g
o
r
ith
m
h
a
s
al
s
o
b
ee
n
d
ev
e
lo
p
ed
w
it
h
L
SI
f
o
r
d
o
cu
m
e
n
t
cl
u
s
ter
in
g
[
2
2
]
,
clu
s
ter
i
n
g
o
f
n
e
ws
ar
ticles
[
2
3
]
,
in
f
o
r
m
atio
n
r
etr
iev
al
[
2
4
]
.
T
h
e
w
ea
k
n
es
s
o
f
L
SI
is
o
v
er
co
m
e
b
y
d
e
v
elo
p
in
g
a
to
p
ic
-
b
ased
w
ei
g
h
ti
n
g
ter
m
ca
lled
L
aten
t
Dir
ich
l
et
al
lo
ca
tio
n
(
L
D
A
)
.
L
D
A
i
s
a
g
e
n
er
ati
v
e
p
r
o
b
a
b
ilis
tic
m
o
d
el
o
f
a
co
r
p
u
s
,
w
h
ic
h
d
o
cu
m
e
n
ts
ar
e
r
ep
r
esen
ted
as
r
a
n
d
o
m
m
ix
t
u
r
es
o
v
er
late
n
t
to
p
ics,
a
n
d
ea
ch
to
p
ic
i
s
ch
ar
ac
ter
ized
b
y
a
d
is
tr
ib
u
tio
n
o
v
er
w
o
r
d
s
[
2
5
]
.
A
d
o
cu
m
en
t
in
a
co
r
p
u
s
is
n
o
t
o
n
l
y
id
en
ti
f
ied
a
s
a
s
i
n
g
l
e
to
p
ic,
b
u
t c
an
b
e
id
en
tif
ied
as
s
ev
er
al
to
p
ics
w
it
h
t
h
eir
r
esp
ec
tiv
e
p
r
o
b
ab
ilit
ies [
2
6
]
-
[
2
8
]
.
L
D
A
h
a
s
b
ee
n
d
ev
e
lo
p
ed
b
ased
o
n
a
h
ier
ar
ch
y
,
k
n
o
w
n
as
h
L
D
A
[
2
9
]
,
b
u
t
t
h
is
m
eth
o
d
i
s
n
o
t
ab
le
t
o
ca
p
tu
r
e
th
e
h
ier
ar
ch
ical
r
elati
o
n
s
h
ip
th
at
is
f
o
r
m
ed
[
3
0
]
.
T
h
er
ef
o
r
e,
r
esear
ch
n
ee
d
s
to
b
e
d
o
n
e
to
class
if
y
d
o
cu
m
en
ts
h
ier
ar
ch
icall
y
b
y
u
s
in
g
h
ier
ar
ch
ica
l
cl
u
s
ter
i
n
g
m
eth
o
d
an
d
L
D
A
f
o
r
w
ei
g
h
t
in
g
ter
m
.
T
h
e
r
esear
c
h
th
at
i
n
te
g
r
ates
L
D
A
i
n
to
t
h
e
h
ier
ar
ch
ical
cl
u
s
ter
i
n
g
m
et
h
o
d
,
esp
ec
iall
y
ag
g
lo
m
er
ativ
e,
h
as
alr
ea
d
y
b
ee
n
d
o
n
e.
X.
L
i,
H.
W
an
g
,
G.
Yin
,
T
.
W
an
g
,
C
.
Yan
g
,
Y.
Yu
,
D.
T
an
g
[
3
1
]
u
s
ed
L
DA
f
o
r
in
d
u
ci
n
g
t
ax
o
n
o
m
y
f
r
o
m
ta
g
s
b
ased
o
n
w
o
r
d
clu
s
ter
in
g
.
A
H
C
f
r
a
m
e
w
o
r
k
is
u
s
ed
to
d
eter
m
i
n
e
h
o
w
s
i
m
ilar
ev
er
y
t
w
o
ta
g
s
,
an
d
th
e
n
L
D
A
is
u
s
ed
to
ca
p
tu
r
e
th
e
m
atic
co
r
r
elati
o
n
s
a
m
o
n
g
tag
s
t
h
at
r
esu
l
ted
b
y
AHC.
D.
T
u
,
L
.
C
h
en
,
G.
C
h
en
[
3
2
]
u
s
ed
L
D
A
to
ex
tr
ac
t
t
h
e
m
o
s
t
t
y
p
ic
al
w
o
r
d
s
in
e
v
er
y
late
n
t
to
p
ic
an
d
ap
p
ly
a
m
u
l
ti
-
w
a
y
h
ier
ar
ch
ical
ag
g
lo
m
er
ativ
e
clu
s
ter
i
n
g
alg
o
r
it
h
m
(
A
H
C
an
d
W
o
r
d
Net)
to
clu
s
ter
ca
n
d
id
ate
co
n
ce
p
t
w
o
r
d
s
.
T
h
e
p
r
o
b
l
e
m
is
t
h
o
s
e
p
ap
er
s
d
is
cu
s
s
ed
ab
o
u
t
E
n
g
l
is
h
tex
t.
Un
til
n
o
w
,
t
h
e
p
er
f
o
r
m
an
ce
o
f
u
s
i
n
g
t
h
e
L
D
A
m
eth
o
d
an
d
ag
g
lo
m
er
ati
v
e
h
ier
ar
ch
ical
cl
u
s
ter
in
g
in
I
n
d
o
n
esia
n
tex
t
h
as
n
ev
er
b
ee
n
p
u
b
lis
h
ed
.
I
f
b
o
th
o
f
t
h
ese
m
et
h
o
d
s
ar
e
p
r
o
v
en
to
h
av
e
g
o
o
d
p
er
f
o
r
m
a
n
ce
in
cl
u
s
ter
i
n
g
I
n
d
o
n
esia
n
tex
ts
,
th
e
n
it
ca
n
also
b
e
u
s
ed
o
n
o
th
e
r
tex
t
m
i
n
i
n
g
ta
s
k
s
,
f
o
r
ex
a
m
p
le
f
o
r
d
o
cu
m
e
n
t s
u
m
m
ar
izatio
n
.
T
o
o
v
er
co
m
e
th
is
p
r
o
b
le
m
,
in
t
h
i
s
r
esear
ch
,
A
H
C
an
d
L
D
A
ar
e
u
s
ed
to
cl
u
s
ter
d
o
cu
m
en
ts
,
w
h
er
e
L
D
A
is
n
o
t
u
s
ed
f
o
r
cl
u
s
ter
i
n
g
,
b
u
t
u
s
ed
to
g
e
n
er
ate
th
e
w
eig
h
t
o
f
te
r
m
s
co
n
tai
n
e
d
in
d
o
cu
m
e
n
t
te
x
t.
T
h
is
r
esear
ch
h
as
d
i
f
f
er
en
ce
s
w
i
th
o
t
h
er
r
elate
d
r
esear
ch
e
s
f
o
r
I
n
d
o
n
e
s
ia
n
tex
t.
First,
t
h
e
ter
m
w
ei
g
h
t
in
g
ca
lcu
latio
n
u
s
ed
L
u
h
n
’
s
I
d
ea
to
s
elec
t
th
e
ter
m
s
o
f
te
x
t
b
y
d
ef
i
n
in
g
u
p
p
er
cu
t
-
o
f
f
a
n
d
lo
w
er
cu
t
-
o
f
f
,
a
n
d
th
e
n
ex
tr
ac
ts
t
h
e
f
ea
t
u
r
e
o
f
ter
m
s
u
s
in
g
Gib
b
s
Sa
m
p
li
n
g
L
D
A
co
m
b
in
ed
w
it
h
th
e
ter
m
f
r
eq
u
e
n
c
y
v
al
u
es
an
d
f
u
zz
y
Su
g
en
o
lo
g
ic.
W
h
ile,
i
n
o
t
h
er
r
esear
ch
,
P
.
M.
P
r
ih
atin
i,
I
.
K.
G.
D.
P
u
tr
a,
I
.
A
.
D.
Gir
ia
n
tar
i,
M.
Su
d
ar
m
a
[
2
6
]
u
s
ed
o
n
l
y
T
F
-
I
DF
f
o
r
ter
m
w
e
ig
h
tin
g
ca
lcu
latio
n
.
Seco
n
d
,
th
e
ca
lcu
latio
n
o
f
th
e
d
is
ta
n
ce
b
et
w
ee
n
d
o
cu
m
e
n
t
s
f
o
r
A
H
C
is
to
p
ic
-
b
ased
b
ec
au
s
e
it
u
s
es
t
h
e
v
al
u
e
t
h
at
r
esu
l
ted
b
y
F
u
zz
y
L
u
h
n
’
s
Gib
b
s
L
D
A
.
T
h
ir
d
,
th
e
d
o
cu
m
en
t
cl
u
s
ter
i
n
g
w
i
th
A
H
C
u
s
e
s
t
h
r
ee
c
h
ar
a
cter
is
tics
:
s
i
n
g
le
lin
k
,
co
m
p
let
e
lin
k
an
d
a
v
er
a
g
e
lin
k
b
ased
o
n
F
u
zz
y
L
u
h
n
’
s
Gib
b
s
L
D
A
,
an
d
th
e
n
co
m
p
ar
es
t
h
e
b
est
AHC
ch
ar
ac
ter
is
tic
s
w
it
h
m
ea
s
u
r
e
m
en
t
m
etr
ics.
W
h
ile,
i
n
o
th
er
r
esear
ch
es,
Y
u
h
e
f
izar
,
B
.
San
to
s
a,
I
.
K.
E
d
d
y
,
Y.
K.
S
u
p
r
ap
to
[
3
3
]
u
s
ed
E
u
c
lid
ea
n
f
o
r
d
is
tan
ce
ca
lc
u
la
tio
n
an
d
s
in
g
l
e
lin
k
ag
e
f
o
r
d
o
cu
m
e
n
t
cl
u
s
te
r
in
g
.
M.
A
.
A
.
R
i
y
ad
i,
D.
S.
P
r
ati
w
i,
A
.
R
.
I
r
a
w
an
,
K.
Fit
h
r
iasar
i
[
3
4
]
u
s
ed
s
in
g
le
lin
k
,
co
m
p
lete
lin
k
an
d
W
ar
d
’
s
lin
k
b
as
ed
o
n
au
to
c
o
r
r
elatio
n
d
is
tan
ce
.
T
h
e
d
is
cu
s
s
io
n
i
n
t
h
is
r
e
s
ea
r
ch
i
s
d
i
v
id
ed
as
f
o
llo
w
s
.
Sect
i
o
n
2
d
is
cu
s
s
es
ab
o
u
t
r
esear
ch
m
et
h
o
d
.
Sectio
n
3
d
is
cu
s
s
es a
b
o
u
t t
h
e
r
es
u
lts
a
n
d
its
an
al
y
s
i
s
.
Sectio
n
4
d
is
cu
s
s
es a
b
o
u
t th
e
co
n
cl
u
s
io
n
o
f
t
h
is
r
esear
ch
.
2.
RE
S
E
ARCH
M
E
T
H
O
D
T
h
is
r
esear
ch
co
n
s
i
s
ts
o
f
s
e
v
er
al
s
tep
s
,
s
u
c
h
as
d
o
cu
m
en
t
tex
t
p
r
o
ce
s
s
in
g
,
ter
m
w
eig
h
t
in
g
w
it
h
Fu
zz
y
L
u
h
n
’
s
Gib
b
s
L
D
A
,
d
o
cu
m
e
n
t
s
cl
u
s
ter
in
g
w
i
th
F
u
zz
y
L
u
h
n
’
s
Gib
b
s
L
D
A
,
a
n
d
e
v
al
u
atio
n
,
as
s
h
o
w
n
in
Fig
u
r
e
1
.
2
.
1
.
Do
cum
e
nt
t
ex
t
p
ro
ce
s
s
i
ng
I
n
th
is
r
esear
ch
,
t
h
e
d
o
cu
m
e
n
ts
u
s
ed
ar
e
n
e
w
s
tex
t
f
i
les
o
b
tain
ed
f
r
o
m
I
n
d
o
n
e
s
ia
n
n
e
w
s
w
eb
s
ite.
E
ac
h
f
i
le
i
s
d
eli
m
ited
in
to
a
c
o
llectio
n
o
f
ter
m
s
i
n
t
h
e
to
k
e
n
izatio
n
p
r
o
ce
s
s
.
I
n
th
e
f
i
lter
in
g
p
r
o
ce
s
s
,
ea
ch
ter
m
is
f
il
ter
ed
u
s
i
n
g
a
s
to
p
w
o
r
d
s
lis
t
r
es
u
lti
n
g
i
n
a
m
ea
n
i
n
g
f
u
l
s
et
o
f
ter
m
s
.
T
h
e
ter
m
s
g
e
n
er
ated
b
y
th
e
f
ilter
i
n
g
p
r
o
ce
s
s
,
s
o
m
e
ar
e
alr
ea
d
y
i
n
th
e
f
o
r
m
o
f
b
asic
w
o
r
d
,
s
o
m
e
ar
e
s
till
h
a
v
e
af
f
i
x
es.
T
o
m
ak
e
all
ter
m
s
in
a
u
n
i
f
o
r
m
s
h
ap
e,
all
ter
m
s
ar
e
p
ar
s
in
g
i
n
to
b
asic
w
o
r
d
s
th
r
o
u
g
h
th
e
s
te
m
m
i
n
g
p
r
o
ce
s
s
.
I
n
th
i
s
r
esear
ch
,
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
C
o
mp
lete
a
g
g
lo
mera
tive
h
iera
r
ch
y
d
o
cu
men
t’
s
clu
s
teri
n
g
b
a
s
ed
o
n
fu
z
z
y
L
u
h
n
’
s
g
i
b
b
s
…
(
P
.
M.
P
r
ih
a
tin
i
)
2105
s
te
m
m
i
n
g
u
s
es
t
h
e
d
eletio
n
o
f
af
f
ix
e
s
m
eth
o
d
b
ased
o
n
r
u
le
s
an
d
b
asic
d
ictio
n
ar
y
.
T
h
e
s
t
e
m
m
in
g
al
g
o
r
ith
m
u
s
ed
is
a
m
o
d
i
f
icatio
n
o
f
Naz
i
ef
-
A
d
r
ia
n
i [
3
5
]
.
Fig
u
r
e
1
.
T
h
e
r
esear
ch
m
e
th
o
d
d
esig
n
2
.
2
.
T
er
m
w
eig
t
hin
g
w
it
h
F
uzzy
L
uh
n’
s
G
i
bb
s
L
DA
I
n
th
i
s
r
esear
ch
,
ter
m
w
eig
h
t
in
g
is
d
o
n
e
th
r
o
u
g
h
ter
m
s
el
ec
tio
n
an
d
f
ea
tu
r
e
ex
tr
ac
tio
n
.
T
h
e
ter
m
s
elec
tio
n
is
b
ased
o
n
t
h
e
co
n
c
ep
t o
f
L
u
h
n
'
s
I
d
ea
,
w
h
er
e
ea
ch
ter
m
i
s
ca
lc
u
lated
b
ased
o
n
it
s
r
elati
v
e
f
r
eq
u
en
c
y
ag
ain
s
t
all
ter
m
s
in
th
e
d
o
cu
m
e
n
t
te
x
t
[
3
6
]
.
L
u
h
n
d
es
cr
ib
es
th
e
r
elatio
n
s
h
ip
b
et
w
ee
n
th
e
o
cc
u
r
r
en
ce
f
r
eq
u
en
c
ies
o
f
a
ter
m
(
ter
m
f
r
eq
u
en
c
y
)
w
it
h
t
h
e
i
m
p
o
r
tan
ce
o
f
a
ter
m
i
n
t
h
e
d
o
cu
m
en
t.
T
h
e
ter
m
w
i
th
m
ed
iu
m
-
f
r
eq
u
en
c
y
is
m
o
r
e
i
m
p
o
r
tan
t
th
an
h
i
g
h
o
r
lo
w
f
r
eq
u
en
c
y
ter
m
s
.
L
o
w
f
r
eq
u
e
n
c
y
te
r
m
s
ar
e
in
cl
u
d
ed
in
th
e
lo
w
er
c
u
t
-
o
f
f
,
w
h
ile
h
i
g
h
f
r
eq
u
en
c
y
ter
m
s
ar
e
i
n
cl
u
d
ed
i
n
t
h
e
u
p
p
er
cu
t
-
o
f
f
.
Me
d
i
u
m
-
f
r
eq
u
en
c
y
ter
m
s
ca
n
b
e
o
b
tain
ed
b
y
c
u
tti
n
g
t
h
e
u
p
p
er
an
d
lo
w
er
c
u
t
-
o
f
f
.
T
o
elim
i
n
ate
ter
m
s
i
n
t
h
e
u
p
p
er
cu
t
-
o
f
f
ca
n
b
e
d
o
n
e
b
y
f
ilter
i
n
g
ter
m
s
b
ased
o
n
s
to
p
w
o
r
d
s
lis
t.
Ho
w
e
v
er
,
to
eli
m
i
n
ate
ter
m
s
in
t
h
e
lo
w
er
cu
t
-
o
f
f
,
s
o
f
ar
th
er
e
h
a
s
b
ee
n
n
o
r
esear
ch
t
h
at
ca
n
d
ete
r
m
in
e
e
f
f
ec
ti
v
e
w
a
y
s
to
d
eter
m
i
n
e
th
e
lo
w
er
cu
t
-
o
f
f
l
i
m
its
.
I
n
t
h
is
r
esear
ch
,
t
h
e
eli
m
i
n
ati
o
n
o
f
ter
m
s
in
t
h
e
u
p
p
er
cu
t
-
o
f
f
li
m
it
is
d
o
n
e
t
w
ice.
Fir
s
t,
it
d
o
n
e
b
y
f
ilter
i
n
g
o
f
ter
m
s
b
ased
o
n
s
t
o
p
w
o
r
d
lis
t.
Seco
n
d
,
th
e
f
ilt
er
in
g
r
e
s
u
l
ts
ar
e
f
il
ter
ed
a
g
ai
n
t
h
r
o
u
g
h
s
te
m
m
in
g
p
r
o
cc
ess
.
Fo
r
th
e
eli
m
i
n
atio
n
o
f
ter
m
s
i
n
t
h
e
lo
w
er
c
u
t
-
o
f
f
li
m
it
i
s
b
ased
o
n
th
e
s
te
m
m
in
g
r
es
u
lt,
w
it
h
d
if
f
er
e
n
t
p
er
ce
n
ta
g
e
r
e
m
o
v
al
v
alu
e
s
f
o
r
ea
ch
tex
t
d
o
cu
m
e
n
t,
as
in
(
1
)
.
T
h
is
is
b
ased
o
n
th
e
id
ea
th
at
ea
c
h
d
o
cu
m
en
t
h
as
d
i
f
f
er
e
n
t
tex
t
len
g
t
h
s
,
s
o
th
at
n
o
s
i
n
g
le
co
n
s
ta
n
t
v
al
u
e
ca
n
b
e
tak
e
n
f
o
r
all
d
o
cu
m
e
n
ts
.
Var
iab
le
lco
d
(
lo
w
er
cu
t
-
o
f
f
d
o
cu
m
e
n
t
)
r
ef
er
s
to
th
e
lo
w
er
-
c
u
t
-
o
f
f
co
n
s
ta
n
t
v
a
lu
e
f
o
r
d
o
cu
m
e
n
t
d
(
in
th
e
f
o
r
m
o
f
a
p
o
s
itiv
e
i
n
te
g
er
)
.
Var
iab
le
fs
d
(
f
al
s
e
-
s
te
m
m
i
n
g
d
o
cu
m
e
n
t)
r
ef
e
r
s
to
t
h
e
n
u
m
b
er
o
f
u
n
s
u
cc
es
s
f
u
l
ter
m
s
te
m
m
i
n
g
in
d
o
cu
m
en
t
d
.
Var
iab
le
fr
d
(
f
ilter
in
g
r
es
u
lt
d
o
cu
m
en
t)
r
ef
er
s
to
th
e
n
u
m
b
er
o
f
ter
m
s
i
n
d
o
cu
m
e
n
t
d
u
s
ed
f
o
r
th
e
s
te
m
m
i
n
g
p
r
o
ce
s
s
.
Var
iab
le
ts
d
(
tr
u
e
-
s
te
m
m
in
g
d
o
cu
m
e
n
t)
r
e
f
er
s
to
th
e
n
u
m
b
er
o
f
s
u
cc
es
s
f
u
l
ter
m
s
te
m
m
i
n
g
i
n
d
o
cu
m
en
t
d
.
=
(
)
|
|
(
1
)
T
h
e
ter
m
s
elec
tio
n
r
es
u
lt
i
s
a
co
llectio
n
o
f
s
elec
ted
ter
m
s
o
f
ea
ch
d
o
cu
m
e
n
t
t
h
at
h
a
v
e
i
m
p
o
r
ta
n
t
m
ea
n
in
g
s
to
b
e
p
r
o
ce
s
s
ed
at
th
e
f
ea
tu
r
e
e
x
tr
ac
tio
n
.
I
n
t
h
is
r
esear
ch
,
f
ea
tu
r
e
e
x
tr
ac
tio
n
is
d
o
n
e
b
y
to
p
ic
-
b
ased
L
D
A
m
et
h
o
d
.
L
D
A
h
as
s
o
m
e
r
ea
s
o
n
in
g
alg
o
r
ith
m
s
,
o
n
e
o
f
w
h
ic
h
is
Gib
b
s
Sa
m
p
li
n
g
th
at
h
av
e
p
r
o
v
en
ef
f
ec
tiv
e
i
n
co
n
d
u
ctin
g
th
e
to
p
ic
s
a
m
p
li
n
g
p
r
o
ce
s
s
[
2
8
]
.
I
n
g
en
er
al,
i
n
t
h
e
in
i
tializatio
n
p
r
o
ce
s
s
,
Gib
b
s
Sa
m
p
li
n
g
as
s
ig
n
s
t
h
e
t
o
p
i
c
o
f
ea
ch
ter
m
r
an
d
o
m
l
y
u
s
i
n
g
a
m
u
l
tin
o
m
ial
r
a
n
d
o
m
f
u
n
ctio
n
.
Ho
w
ev
er
,
th
e
u
s
e
o
f
th
i
s
f
u
n
ctio
n
ca
n
n
o
t
r
ep
r
esen
t
th
e
ex
i
s
ten
ce
o
f
ea
ch
ter
m
i
n
th
e
to
p
ic.
T
h
er
ef
o
r
e,
in
th
is
r
esear
c
h
,
th
e
d
eter
m
i
n
atio
n
o
f
to
p
ic
f
o
r
ea
ch
ter
m
i
n
t
h
e
in
it
ializatio
n
p
r
o
ce
s
s
is
d
o
n
e
b
ased
o
n
th
e
h
ig
h
e
s
t
o
cc
u
r
r
en
c
e
f
r
eq
u
en
c
y
(
tf
)
o
f
t
h
e
ter
m
in
a
ll
to
p
ics,
as
in
(
2
)
.
Var
iab
le
z
t,
k
s
i
m
ilar
w
it
h
k
r
ef
er
s
to
t
h
e
t
o
p
ic.
Var
iab
le
tf
t,
k
r
ef
er
s
to
tf
v
al
u
e
o
f
ter
m
t
o
n
to
p
ic
k
.
T
o
ca
lcu
late
th
e
p
r
o
b
ab
ilit
y
o
f
ea
c
h
ter
m
i
n
t
h
e
s
a
m
p
lin
g
p
r
o
ce
s
s
u
s
ed
th
e
f
o
r
m
u
la
as
i
n
(
3
)
.
Var
iab
le
p
t,
k
r
ef
er
s
to
p
r
o
b
a
b
ilit
y
v
al
u
e
o
f
s
a
m
p
lin
g
f
o
r
ter
m
t
o
n
to
p
ic
k
.
Var
iab
le
n
kw
-
1
r
ef
er
s
to
th
e
v
alu
e
o
f
th
e
to
p
ic
-
ter
m
m
atr
i
x
b
y
i
g
n
o
r
in
g
t
h
e
cu
r
r
en
t
ter
m
v
alu
e.
Var
ia
b
le
V
is
th
e
u
n
iq
u
e
n
u
m
b
er
o
f
ter
m
s
i
n
all
d
o
cu
m
en
ts
.
Var
iab
le
ndk
-
1
r
ef
er
s
to
v
alu
e
o
f
th
e
d
o
cu
m
en
t
-
to
p
i
c
m
a
tr
ix
b
y
i
g
n
o
r
in
g
th
e
cu
r
r
en
t
ter
m
v
al
u
e.
Var
iab
le
β
d
eter
m
i
n
e
th
e
m
i
x
in
g
p
r
o
p
o
r
tio
n
o
f
d
o
cu
m
e
n
t
s
o
n
th
e
to
p
ics,
w
h
ile
α
d
eter
m
in
e
s
th
e
m
i
x
t
u
r
e
co
m
p
o
n
en
t
s
o
f
w
o
r
d
s
o
n
t
h
e
to
p
ics [
3
7
]
.
Var
iab
le
K
is
th
e
n
u
m
b
er
o
f
to
p
ic.
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
.
3
,
J
u
n
e
2
0
1
9
:
2
1
0
3
-
2
1
1
1
2106
,
=
~
ma
x
(
,
)
(
2
)
,
=
−
1
(
∑
−
1
−
1
)
+
(
)
−
1
∝
(
∑
−
1
−
1
)
+
(
∝
)
(
3
)
I
n
g
e
n
er
al,
Gib
b
s
Sa
m
p
lin
g
i
n
t
h
e
L
D
A
r
eq
u
ir
es
s
ev
er
al
t
i
m
es
i
ter
atio
n
s
f
o
r
t
h
e
s
a
m
p
li
n
g
p
r
o
ce
s
s
u
n
t
il
it
r
ea
ch
e
s
co
n
v
er
g
e
n
t
co
n
d
itio
n
s
.
T
h
is
ta
k
es
ti
m
e
an
d
h
i
g
h
co
m
p
lex
it
y
.
T
h
e
ad
d
itio
n
o
f
f
u
zz
y
T
s
u
k
a
m
o
to
lo
g
ic
i
n
to
t
h
e
s
a
m
p
lin
g
p
r
o
ce
s
s
ca
n
ac
ce
ler
at
e
th
e
ac
h
ie
v
e
m
en
t
o
f
co
n
v
er
g
en
t
co
n
d
itio
n
s
w
it
h
g
o
o
d
m
ea
s
u
r
e
m
en
t
v
a
lu
e
s
[
2
6
]
.
T
h
e
f
u
zz
y
lo
g
ic
co
n
ce
p
t
t
h
a
t
u
s
ed
in
t
h
at
r
esear
c
h
w
i
ll
b
e
i
m
p
r
o
v
ed
t
h
r
o
u
g
h
th
is
r
esear
ch
b
y
u
s
in
g
S
u
g
e
n
o
m
et
h
o
d
to
i
n
cr
ea
s
e
th
e
ac
cu
r
ac
y
v
al
u
e,
co
n
s
id
er
in
g
t
h
e
o
u
tp
u
t
o
f
f
u
zz
y
lo
g
ic
w
h
ic
h
n
ee
d
ed
f
o
r
s
a
m
p
li
n
g
is
a
co
n
s
ta
n
t v
a
lu
e.
I
n
th
i
s
r
esear
ch
,
th
e
u
p
p
er
an
d
lo
w
er
li
m
i
ts
f
o
r
th
e
f
u
zz
y
c
u
r
v
e
ar
e
d
eter
m
i
n
ed
b
ased
o
n
t
h
e
tf
v
al
u
e
o
f
ea
c
h
ter
m
.
F
u
zz
if
ic
atio
n
u
s
es
a
tr
ian
g
u
lar
cu
r
v
e
w
it
h
t
h
e
p
r
o
b
ab
ilit
y
v
alu
e
o
f
t
h
e
s
a
m
p
li
n
g
r
es
u
lt
f
o
r
ea
ch
ter
m
p
,
as
i
n
(
4
)
.
Var
i
ab
le
u
[
t
]
r
ef
er
s
to
t
h
e
d
eg
r
ee
o
f
m
e
m
b
er
s
h
ip
f
o
r
t
er
m
t
.
Var
iab
le
a
r
ef
er
s
to
t
h
e
lo
w
er
b
o
u
n
d
o
f
th
e
c
u
r
v
e.
Va
r
iab
le
b
r
ef
er
s
to
th
e
p
ea
k
o
f
t
h
e
cu
r
v
e.
Var
iab
le
c
r
ef
er
s
to
th
e
u
p
p
er
b
o
u
n
d
o
f
t
h
e
cu
r
v
e.
T
h
e
i
m
p
licatio
n
f
u
n
ctio
n
u
s
ed
is
OR
b
ec
au
s
e
f
u
zz
y
lo
g
ic
h
er
e
is
u
s
ed
to
d
eter
m
in
e
th
e
p
r
o
b
ab
ilit
y
v
alu
e
o
f
ter
m
f
o
r
o
n
e
to
p
ic,
w
h
ile
a
ll
to
p
ic
w
il
l
b
e
d
eter
m
i
n
ed
in
s
a
m
p
li
n
g
p
r
o
ce
s
s
.
T
h
e
r
u
le
co
m
p
o
s
i
tio
n
g
e
n
er
ates
th
e
α
p
v
al
u
e
b
as
ed
o
n
th
e
m
ax
i
m
u
m
v
al
u
e
o
f
all
u
[
t
]
as
in
(
5
)
,
an
d
th
e
v
al
u
e
o
f
z
o
is
b
ased
o
n
th
e
ter
m
p
r
o
b
ab
ilit
y
v
al
u
e
a
cr
o
s
s
th
e
to
p
ic
w
h
o
s
e
v
al
u
e
is
n
o
t
eq
u
al
to
ze
r
o
,
as
in
(
6
)
.
Var
iab
le
t
r
ef
er
s
to
th
e
ter
m
p
r
o
b
ab
ilit
y
o
f
s
a
m
p
l
in
g
r
es
u
lt.
Var
iab
le
z
o
r
ef
er
s
to
th
e
co
m
p
o
s
it
io
n
o
u
tp
u
t.
Var
iab
le
n
r
ef
er
s
to
th
e
n
u
m
b
er
o
f
to
p
ics
w
h
o
s
e
ter
m
p
r
o
b
ab
ilit
y
i
s
n
o
t
eq
u
al
to
ze
r
o
.
Fo
r
th
e
d
ef
u
zz
i
f
icatio
n
,
t
h
e
f
i
n
al
o
u
tp
u
t
o
f
f
u
zz
y
z
i
s
o
b
tai
n
ed
b
y
ca
lcu
lati
n
g
t
h
e
m
ea
n
v
al
u
e,
as
in
(
7
)
.
T
h
e
v
alu
e
o
f
z
is
u
s
ed
as
th
e
p
r
o
b
ab
ilit
y
v
al
u
e
o
f
ter
m
p
f
o
r
to
p
ic
k
an
d
w
il
l
b
e
u
s
ed
f
o
r
th
e
n
ex
t
s
a
m
p
lin
g
p
r
o
ce
s
s
u
n
til
it
r
ea
ch
es
co
n
v
er
g
e
n
t
co
n
d
itio
n
s
.
Af
ter
co
n
v
er
g
e
n
ce
,
t
h
e
f
i
n
al
v
alu
e
o
f
z
w
il
l
b
e
t
h
e
f
ea
tu
r
e
v
alu
e
f
o
r
ea
c
h
ter
m
an
d
r
ea
d
y
f
o
r
clu
s
ter
i
n
g
.
[
]
=
{
0
,
≤
,
≥
−
−
,
≤
≤
−
−
,
≤
≤
(
4
)
∝
=
ma
x
(
[
1
]
,
[
2
]
,
…
,
[
]
)
(
5
)
=
∑
(
≠
0
)
(
6
)
=
∝
∝
(
7
)
2
.
3
.
Do
cum
e
nts
clu
s
t
er
ing
w
it
h F
uzzy
L
uh
n’
s
G
i
bb
s
L
D
A
T
h
e
f
ea
tu
r
e
v
al
u
es
o
b
tai
n
ed
at
th
e
f
ea
t
u
r
e
e
x
tr
ac
tio
n
ar
e
u
s
ed
to
ca
lc
u
late
t
h
e
d
i
s
tan
c
e
b
et
w
ee
n
d
o
cu
m
en
ts
to
b
e
u
s
ed
in
th
e
clu
s
ter
i
n
g
p
r
o
ce
s
s
.
I
n
th
is
r
esear
ch
,
d
is
ta
n
ce
ca
lc
u
latio
n
s
u
s
i
n
g
t
h
e
C
o
s
i
n
e
Si
m
i
lar
it
y
,
as
in
(
8
)
.
Var
iab
le
|d
i
-
d
j|
r
ef
er
s
to
th
e
d
is
ta
n
ce
b
et
w
ee
n
d
o
cu
m
e
n
ts
i
a
n
d
j
.
Var
iab
le
d
i
r
ef
er
s
to
d
o
cu
m
en
t
i
,
w
h
ile
d
j
r
ef
er
s
to
d
o
cu
m
en
t
j
.
|
−
|
=
√
2
2
(
8)
T
h
e
d
is
tan
ce
b
et
w
ee
n
d
o
cu
m
e
n
ts
i
s
u
s
ed
to
clu
s
ter
d
o
cu
m
en
t
u
s
i
n
g
t
h
r
ee
t
y
p
es
o
f
A
H
C
al
g
o
r
ith
m
s
.
I
n
th
e
Sin
g
le
L
i
n
k
A
H
C
al
g
o
r
ith
m
,
cl
u
s
ter
s
ar
e
b
ased
o
n
th
e
s
m
a
lles
t
d
i
s
tan
ce
b
et
w
ee
n
p
air
s
o
f
t
w
o
d
o
cu
m
en
ts
,
a
s
i
n
(
9
)
.
I
n
t
h
e
C
o
m
p
lete
L
in
k
AHC
a
lg
o
r
i
th
m
,
cl
u
s
ter
s
ar
e
b
ased
o
n
t
h
e
lar
g
e
s
t
d
i
s
tan
ce
b
et
w
ee
n
p
air
s
o
f
t
w
o
d
o
cu
m
e
n
ts
,
a
s
i
n
(
1
0
)
.
I
n
th
e
Av
er
a
g
e
L
i
n
k
A
H
C
al
g
o
r
ith
m
,
cl
u
s
te
r
s
ar
e
b
ased
o
n
t
h
e
av
er
ag
e
d
is
ta
n
ce
b
et
w
ee
n
p
air
s
o
f
t
w
o
d
o
cu
m
e
n
ts
,
as
in
(
1
1
)
.
Var
iab
le
d
ij
r
ef
er
s
to
th
e
s
elec
ted
p
air
o
f
do
cu
m
en
ts
i
a
n
d
j
.
=
min
(
|
−
1
|
,
|
−
2
|
,
…
,
|
−
|
)
(
9
)
=
ma
x
(
|
−
1
|
,
|
−
2
|
,
…
,
|
−
|
)
(
1
0
)
=
a
vg
(
|
−
1
|
,
|
−
2
|
,
…
,
|
−
|
)
(
1
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
C
o
mp
lete
a
g
g
lo
mera
tive
h
iera
r
ch
y
d
o
cu
men
t’
s
clu
s
teri
n
g
b
a
s
ed
o
n
fu
z
z
y
L
u
h
n
’
s
g
i
b
b
s
…
(
P
.
M.
P
r
ih
a
tin
i
)
2107
2
.
4
.
M
et
rics e
v
a
lua
t
io
n
I
n
th
i
s
r
esear
ch
,
e
v
alu
a
tio
n
i
s
d
o
n
e
in
t
w
o
s
tep
s
:
ev
al
u
ati
o
n
o
f
th
e
f
ea
t
u
r
e
ex
tr
ac
tio
n
r
esu
lt
s
an
d
ev
alu
a
tio
n
o
f
th
e
cl
u
s
ter
i
n
g
r
esu
lt
s
.
T
h
e
te
x
t
d
o
cu
m
e
n
t
u
s
ed
in
th
i
s
r
esear
c
h
h
as
b
ee
n
class
i
f
ied
i
n
to
f
i
v
e
ca
teg
o
r
ies b
y
I
n
d
o
n
e
s
ian
n
e
w
s
m
ed
ia
w
eb
s
ites
,
s
o
it c
a
n
b
e
u
s
ed
as g
o
ld
s
tan
d
ar
d
f
o
r
th
e
e
v
alu
atio
n
p
r
o
ce
s
s
.
E
v
alu
a
tio
n
o
f
f
ea
t
u
r
e
ex
tr
ac
ti
o
n
r
esu
lt
s
is
d
o
n
e
b
y
co
m
p
ar
in
g
r
es
u
lt
s
w
it
h
lo
w
er
cu
t
-
o
f
f
a
n
d
w
ith
o
u
t
lo
w
er
cu
t
-
o
f
f
.
A
n
e
v
al
u
atio
n
w
a
s
also
p
er
f
o
r
m
ed
to
co
m
p
ar
e
th
e
f
ea
tu
r
e
e
x
tr
ac
tio
n
r
esu
lts
b
et
w
ee
n
t
h
e
F
u
zz
y
Gib
b
s
L
D
A
m
e
th
o
d
[
2
6
]
an
d
Fu
zz
y
L
u
h
n
's
Gib
b
s
L
D
A
th
at
u
s
ed
in
t
h
is
r
esear
ch
.
T
h
e
ev
alu
atio
n
w
a
s
p
er
f
o
r
m
ed
u
s
i
n
g
t
w
o
m
ea
s
u
r
e
m
en
t
m
etr
ic
s
.
First,
t
h
e
p
er
p
lex
it
y
is
u
s
ed
to
m
ea
s
u
r
e
th
e
a
b
ilit
y
o
f
t
h
e
F
u
zz
y
L
u
h
n
'
s
Gib
b
s
L
D
A
f
ea
tu
r
e
ex
tr
ac
tio
n
m
eth
o
d
to
g
en
er
al
ize
th
e
h
id
d
en
d
ata,
as
in
(
1
2
)
an
d
(
1
3
)
[
2
5
]
.
T
h
e
s
m
aller
v
al
u
e
o
f
t
h
e
p
er
p
lex
it
y
in
d
icate
s
th
e
b
etter
p
er
f
o
r
m
a
n
ce
o
f
t
h
e
m
et
h
o
d
.
Var
iab
le
(
̃
|
)
r
ef
er
s
to
p
er
p
lex
ity
v
al
u
e.
Var
iab
le
M
r
ef
er
s
to
th
e
n
u
m
b
er
o
f
d
o
cu
m
e
n
t
s
.
Var
iab
le
V
is
th
e
u
n
iq
u
e
n
u
m
b
er
o
f
ter
m
s
in
all
d
o
cu
m
en
t
s
.
Var
iab
le
r
ef
er
s
to
th
e
n
u
m
b
er
o
f
o
cc
u
r
r
en
ce
s
o
f
t
h
e
w
o
r
d
t
in
d
o
cu
m
e
n
t
m
.
Var
iab
le
K
is
th
e
n
u
m
b
er
o
f
to
p
ic.
Var
iab
le
,
r
ef
er
s
to
th
e
n
u
m
b
er
o
f
d
o
c
u
m
e
n
t
s
f
o
r
ea
ch
to
p
ic.
Var
iab
le
,
r
ef
er
s
to
th
e
n
u
m
b
er
o
f
w
o
r
d
s
f
o
r
ea
ch
to
p
ic.
(
̃
|
)
=
e
xp
−
∑
l
o
g
(
⃗
⃗
|
)
=
1
∑
=
1
(
1
2
)
l
og
(
⃗
⃗
|
)
=
∑
=
1
l
og
(
∑
,
=
1
.
,
)
(
1
3
)
Seco
n
d
,
P
r
ec
is
io
n
(
P
)
,
R
ec
all
(
R
)
an
d
F
-
Sco
r
e
(
F)
m
etr
ics
ar
e
u
s
ed
to
m
ea
s
u
r
e
t
h
e
ab
ilit
y
o
f
t
h
e
Fu
zz
y
L
u
h
n
's
Gib
b
s
L
D
A
f
e
atu
r
e
ex
tr
ac
tio
n
m
et
h
o
d
in
f
i
n
d
in
g
r
elev
a
n
t
d
o
cu
m
e
n
t
s
ac
co
r
d
in
g
to
th
e
g
o
ld
st
an
d
ar
d
,
as
in
(
1
4
)
-
(
1
6
)
[
1
4
]
.
Var
iab
le
TP
is
tr
u
e
p
o
s
it
iv
e,
r
ef
er
s
to
th
e
n
u
m
b
er
o
f
r
ele
v
a
n
t
ite
m
s
r
etr
ie
v
ed
.
Var
iab
le
FP
is
f
alse
p
o
s
itiv
e
,
r
ef
er
s
to
th
e
n
u
m
b
er
o
f
n
o
n
-
r
ele
v
a
n
t
ite
m
s
r
etr
iev
ed
.
Var
iab
le
FN
is
f
als
e
n
eg
at
iv
e,
r
e
f
er
s
to
th
e
n
u
m
b
er
o
f
r
elev
a
n
t
ite
m
s
t
h
at
ca
n
n
o
t
r
etr
iev
ed
.
T
h
e
g
r
ea
ter
v
a
lu
e
o
f
P
,
R
a
n
d
F
in
d
icate
s
t
h
e
b
etter
p
er
f
o
r
m
an
ce
o
f
th
e
m
eth
o
d
s
.
=
+
(
1
4
)
=
+
(
1
5
)
=
2
.
.
+
(
1
6
)
E
v
alu
a
tio
n
o
f
cl
u
s
ter
i
n
g
r
es
u
l
ts
is
d
o
n
e
b
y
co
m
p
ar
i
n
g
t
h
e
clu
s
ter
i
n
g
r
esu
lt
s
b
et
w
ee
n
Si
n
g
le
L
i
n
k
,
C
o
m
p
lete
L
i
n
k
a
n
d
A
v
er
a
g
e
L
i
n
k
A
H
C
u
s
i
n
g
f
ea
tu
r
e
e
x
tr
a
ctio
n
r
es
u
lt
s
.
I
n
t
h
i
s
r
esear
c
h
,
th
e
e
v
al
u
atio
n
w
a
s
p
er
f
o
r
m
ed
u
s
i
n
g
f
i
v
e
m
ea
s
u
r
e
m
en
t
m
etr
ic
s
.
P
r
ec
is
io
n
,
R
ec
all,
an
d
F
-
Sco
r
e
(
P
R
F)
ar
e
u
s
ed
to
m
ea
s
u
r
e
th
e
ab
ilit
y
o
f
m
e
th
o
d
s
to
clu
s
ter
r
elev
an
t
d
o
cu
m
e
n
ts
ac
co
r
d
in
g
t
o
th
e
g
o
ld
s
tan
d
ar
d
,
as
in
(
1
4
)
-
(
1
6
)
.
T
h
e
f
o
u
r
th
is
No
r
m
a
lized
Mu
tu
al
I
n
f
o
r
m
ati
o
n
(
NM
I
)
,
as
i
n
(
1
7
)
-
(
2
0
)
[
3
8
]
.
Var
iab
le
(
Ω
,
)
r
ef
er
s
to
t
h
e
v
alu
e
o
f
m
u
t
u
a
l
in
f
o
r
m
atio
n
f
o
r
class
(
g
o
ld
s
ta
n
d
ar
d
)
an
d
clu
s
ter
.
Var
iab
le
(
Ω
)
r
ef
er
s
to
th
e
e
n
tr
o
p
y
o
f
class
.
Var
iab
le
(
)
r
ef
er
s
to
th
e
en
tr
o
p
y
o
f
cl
u
s
ter
.
Var
iab
le
r
ef
er
s
to
th
e
n
u
m
b
er
o
f
d
o
cu
m
e
n
t
b
elo
n
g
s
to
cla
s
s
k
.
Var
iab
le
r
ef
er
s
to
th
e
n
u
m
b
er
o
f
d
o
cu
m
en
t b
elo
n
g
s
to
clu
s
ter
j
.
=
2
(
Ω
,
)
(
Ω
)
+
(
)
(
1
7
)
(
Ω
,
)
=
∑
∑
|
∩
|
l
og
|
∩
|
|
|
|
|
(
1
8
)
(
Ω
)
=
−
∑
|
|
l
og
|
|
(
1
9
)
(
)
=
−
∑
|
|
l
og
|
|
(
2
0
)
T
h
e
f
if
t
h
is
A
d
j
u
s
ted
R
an
d
I
n
d
ex
(
AR
I
)
,
as
in
(
2
1
)
-
(
2
4
)
[
3
9
]
.
Var
iab
le
r
ef
er
s
to
th
e
n
u
m
b
er
o
f
d
o
cu
m
en
t
b
elo
n
g
s
to
class
i
an
d
clu
s
ter
j
.
Var
iab
le
r
ef
er
s
to
th
e
n
u
m
b
er
o
f
d
o
c
u
m
e
n
t
b
elo
n
g
s
to
class
i
.
Var
iab
le
′
r
ef
er
s
to
th
e
n
u
m
b
e
r
o
f
d
o
cu
m
en
t
b
elo
n
g
s
to
clu
s
ter
j
.
T
h
e
g
r
ea
ter
v
alu
e
o
f
P
,
R
,
F,
NM
I
an
d
AR
I
in
d
icate
s
th
e
b
etter
p
er
f
o
r
m
an
ce
o
f
th
e
m
et
h
o
d
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
.
3
,
J
u
n
e
2
0
1
9
:
2
1
0
3
-
2
1
1
1
2108
=
∑
∑
(
2
)
−
3
=
1
=
1
1
2
(
1
+
2
)
−
3
(
2
1
)
1
=
∑
(
|
|
2
)
=
1
(
2
2
)
2
=
∑
(
|
′
|
2
)
=
1
(
2
3
)
3
=
2
1
2
(
−
1
)
(
2
4
)
3.
RE
SU
L
T
S
A
ND
AN
AL
Y
SI
S
3
.
1
.
F
uzzy
L
uh
n’s
G
ib
bs
L
DA
T
h
e
ev
alu
atio
n
r
esu
lt
s
o
f
t
h
e
Fu
zz
y
L
u
h
n
's
Gib
b
s
L
D
A
f
e
atu
r
e
ex
tr
ac
tio
n
m
et
h
o
d
ca
n
b
e
s
ee
n
in
T
ab
le
1
.
T
h
e
v
alu
e
in
t
h
e
tab
l
e
s
h
o
w
s
th
e
co
m
p
ar
i
s
o
n
b
et
wee
n
F
u
zz
y
L
u
h
n
's
Gib
b
s
L
D
A
f
ea
t
u
r
e
ex
tr
ac
tio
n
m
et
h
o
d
an
d
Fu
zz
y
Gib
b
s
L
D
A
t
h
at
h
av
e
p
u
b
lis
h
ed
b
y
P
.
M.
P
r
ih
atin
i,
I
.
K.
G.
D.
P
u
tr
a,
I
.
A
.
D.
Gir
ian
tar
i,
M.
Su
d
ar
m
a
[
2
6
]
.
T
h
e
Fu
zz
y
L
u
h
n
'
s
Gib
b
s
L
D
A
f
ea
t
u
r
e
ex
tr
ac
t
io
n
m
et
h
o
d
u
s
es
lo
w
er
cu
t
-
o
f
f
an
d
w
it
h
o
u
t
lo
w
er
cu
t
-
o
f
f
f
o
r
th
e
s
elec
tio
n
f
ea
t
u
r
e
m
eth
o
d
,
w
h
i
le
Fu
zz
y
Gib
b
s
L
D
A
d
id
n
o
t u
s
e
L
u
h
n
’
s
co
n
c
ep
t.
T
ab
le
1
.
C
o
m
p
ar
is
o
n
o
f
Fu
zz
y
L
u
h
n
’
s
Gib
b
s
L
D
A
an
d
F
u
zz
y
Gib
b
s
L
D
A
M
e
t
r
i
c
s
e
v
a
l
u
a
t
i
o
n
s
F
u
z
z
y
L
u
h
n
’
s G
i
b
b
s L
D
A
T
h
e
d
i
f
f
e
r
e
n
c
e
F
u
z
z
y
G
i
b
b
s L
D
A
T
h
e
d
i
f
f
e
r
e
n
c
e
L
o
w
e
r
c
u
t
-
o
f
f
(
1
)
W
i
t
h
o
u
t
l
o
w
e
r
c
u
t
-
o
f
f
(
2
)
(
1
)
&
(
2
)
(
3
)
(
1
)
&
(
3
)
(
2
)
&
(
3
)
P
e
r
p
l
e
x
i
t
y
0
.
0
3
7
5
0
.
0
3
3
9
0
.
0
0
3
6
0
.
0
3
7
6
0
.
0
0
0
1
0
.
0
0
3
7
P
r
e
c
i
si
o
n
0
.
9
4
3
5
0
.
9
5
1
5
0
.
0
0
8
0
0
.
8
9
7
5
0
.
0
460
0
.
0
5
4
0
R
e
c
a
l
l
0
.
9
2
8
0
0
.
9
3
6
0
0
.
0
0
8
0
0
.
8
4
8
6
0
.
0
7
9
4
0
.
0
8
7
4
F
-
S
c
o
r
e
0
.
9
2
9
6
0
.
9
3
8
7
0
.
0
0
9
1
0
.
8
4
2
0
0
.
0
8
7
6
0
.
0
9
6
7
T
h
e
ev
alu
atio
n
r
es
u
lt
s
in
T
ab
l
e
1
s
h
o
w
s
t
h
at
th
e
f
ea
t
u
r
e
ex
tr
ac
tio
n
w
ith
lo
w
er
cu
t
-
o
f
f
u
s
i
n
g
eq
u
atio
n
(
1
)
g
iv
es
t
h
e
ev
al
u
atio
n
v
al
u
e
n
o
t
m
u
ch
d
i
f
f
er
en
t
t
h
a
n
w
it
h
o
u
t
t
h
e
lo
w
er
cu
t
-
o
f
f
.
T
h
e
d
if
f
er
en
ce
o
f
m
etr
ic
m
ea
s
u
r
e
m
e
n
t
v
al
u
es
b
et
w
ee
n
th
e
t
w
o
m
et
h
o
d
s
is
v
er
y
s
m
all
w
ith
th
e
r
a
n
g
e
f
r
o
m
0
.
0
0
3
6
to
0
.
0
0
9
1
.
T
h
is
in
s
i
g
n
i
f
ican
t
d
if
f
er
en
ce
o
cc
u
r
s
b
ec
au
s
e
th
e
f
ea
tu
r
e
s
elec
tio
n
in
th
i
s
r
esear
ch
h
as
b
ee
n
d
o
n
e
th
r
o
u
g
h
t
w
o
s
tep
o
f
th
e
u
p
p
er
cu
t
-
o
f
f
,
w
h
ich
at
f
ilter
i
n
g
s
tep
w
i
th
s
to
p
w
o
r
d
lis
t
an
d
th
e
n
at
s
te
m
m
i
n
g
s
te
p
.
T
h
ese
t
w
o
s
tep
s
h
av
e
f
i
lter
ed
th
e
ter
m
w
it
h
f
r
eq
u
en
c
ies
t
h
at
ap
p
ea
r
f
r
eq
u
en
t
l
y
an
d
r
ar
el
y
ap
p
ea
r
,
s
o
it
r
es
u
lts
a
li
s
t
o
f
m
ea
n
in
g
f
u
l te
r
m
s
f
o
r
t
h
e
f
ea
t
u
r
e
ex
tr
ac
tio
n
.
T
h
e
lo
w
er
cu
t
-
o
f
f
p
r
o
ce
s
s
w
it
h
t
h
e
v
al
u
e
ad
j
u
s
t
ed
to
th
e
len
g
t
h
o
f
th
e
d
o
cu
m
e
n
t
o
n
l
y
r
e
m
o
v
es
a
s
m
all
p
o
r
tio
n
o
f
th
e
m
ea
n
i
n
g
f
u
l
ter
m
in
t
h
e
f
ea
tu
r
e
s
ele
ctio
n
s
o
it
d
o
es
n
o
t
s
ig
n
i
f
ica
n
tl
y
a
f
f
ec
t t
h
e
f
ea
t
u
r
e
ex
tr
ac
tio
n
r
es
u
lt
s
.
T
h
e
e
v
alu
atio
n
r
es
u
lt
s
in
T
a
b
le
1
also
s
h
o
w
s
F
uzzy
G
ib
bs
L
DA
m
et
h
o
d
r
esu
lted
p
er
p
lex
it
y
o
f
0
.
0
3
7
6
,
w
h
ile
F
uzzy
L
u
hn
'
s
G
ib
bs
L
D
A
i
n
t
h
i
s
r
esear
c
h
g
iv
es
t
h
e
v
alu
e
o
f
p
er
p
lex
it
y
0
.
0
375
f
o
r
lo
w
er
cut
-
o
f
f
an
d
0
.
0
3
3
9
w
it
ho
ut
lo
w
e
r
cut
-
o
f
f
.
T
h
is
in
d
icate
s
t
h
at
t
h
e
F
u
z
z
y
L
u
h
n
'
s
G
ib
b
s
L
D
A
alg
o
r
ith
m
p
er
f
o
r
m
s
as
w
e
ll
as
F
u
zz
y
Gib
b
s
L
D
A
i
n
g
e
n
er
ati
n
g
h
id
d
en
d
ata.
B
u
t,
th
e
r
esu
lts
o
f
t
h
e
P
,
R
,
an
d
F
m
etr
ic
i
n
d
icate
th
at
th
e
F
uzzy
L
u
hn
'
s
G
ib
bs
L
D
A
alg
o
r
it
h
m
p
er
f
o
r
m
ed
g
i
v
es
b
etter
r
esu
lt
s
r
an
g
i
n
g
f
r
o
m
0
.
9
2
8
0
t
o
0
.
9
5
1
5
th
an
F
uz
zy
G
ib
bs
L
DA
a
lg
o
r
it
h
m
r
an
g
i
n
g
f
r
o
m
0
.
8
4
2
0
t
o
0
.
8
9
7
5
.
T
h
e
in
cr
ea
s
in
g
v
alu
e
o
f
P
R
F
m
etr
ic
s
h
o
w
s
t
h
at
th
e
to
p
ic
d
eter
m
i
n
atio
n
o
f
ea
ch
ter
m
f
o
r
in
itial
s
a
m
p
li
n
g
t
h
at
p
er
f
o
r
m
ed
b
ased
o
n
th
e
h
ig
h
e
s
t
o
cc
u
r
r
en
ce
f
r
eq
u
en
c
y
(
tf
)
o
f
ter
m
to
all
t
o
p
ics
b
y
u
s
in
g
L
u
h
n
’
s
I
d
ea
a
n
d
th
e
u
s
e
o
f
t
h
e
F
u
zz
y
Su
g
e
n
o
m
et
h
o
d
is
b
etter
ab
le
to
f
in
d
d
o
cu
m
e
n
ts
r
ele
v
a
n
t
to
t
h
e
g
o
ld
s
tan
d
ar
d
.
T
h
is
i
n
d
icate
s
t
h
at
Fu
zz
y
L
u
h
n
'
s
Gi
b
b
s
L
D
A
al
g
o
r
ith
m
is
a
b
etter
ch
o
ice
in
p
er
f
o
r
m
i
n
g
f
ea
t
u
r
e
ex
tr
ac
tio
n
f
o
r
clu
s
ter
in
g
.
2
.
2
.
AH
C
w
it
h F
uzzy
L
uh
n’s G
i
bb
s
L
DA
T
h
e
ev
alu
atio
n
r
es
u
lt
s
o
f
t
h
e
A
H
C
alg
o
r
it
h
m
s
p
er
f
o
r
m
ed
b
ased
o
n
th
e
Fu
zz
y
L
u
h
n
'
s
G
ib
b
s
L
D
A
f
ea
t
u
r
e
ex
tr
ac
tio
n
ca
n
b
e
s
ee
n
in
T
ab
le
2
.
T
h
e
ev
al
u
atio
n
r
esu
lt
s
i
n
T
ab
le
2
s
h
o
w
s
t
h
at
t
h
e
f
ea
t
u
r
e
s
elec
tio
n
s
w
it
h
lo
w
er
cu
t
-
o
f
f
o
r
w
it
h
o
u
t
lo
w
er
cu
t
-
o
f
f
d
o
n
o
t
a
f
f
ec
t
th
e
p
er
f
o
r
m
a
n
ce
o
f
t
h
e
A
H
C
al
g
o
r
ith
m
s
i
n
t
h
e
clu
s
ter
i
n
g
p
r
o
ce
s
s
.
I
t
ca
n
b
e
s
ee
n
f
r
o
m
th
e
m
ea
s
u
r
e
m
en
t
m
etr
ic
v
al
u
es
t
h
at
b
o
th
f
ea
t
u
r
e
s
elec
tio
n
m
et
h
o
d
s
p
r
o
d
u
ce
C
o
m
p
lete
L
i
n
k
A
H
C
alg
o
r
it
h
m
as
t
h
e
AHC
clu
s
t
er
in
g
alg
o
r
it
h
m
w
it
h
t
h
e
b
es
t
m
etr
ic
v
al
u
e.
T
h
e
d
if
f
er
e
n
ce
s
f
o
r
th
e
C
o
m
p
le
te
L
i
n
k
A
HC
alg
o
r
it
h
m
w
it
h
b
o
th
f
ea
tu
r
e
s
elec
tio
n
m
et
h
o
d
s
r
a
n
g
e
s
f
r
o
m
0
.
0
0
0
3
to
0
.
0
2
6
3
.
T
h
is
s
h
o
w
s
t
h
at
b
o
th
f
ea
tu
r
e
s
elec
tio
n
m
et
h
o
d
s
ca
n
b
e
u
s
ed
as
a
g
o
o
d
ch
o
ice
in
clu
s
ter
i
n
g
p
r
o
ce
s
s
w
it
h
A
H
C
.
B
u
t,
in
ter
m
s
o
f
th
e
co
n
s
is
te
n
c
y
o
f
th
e
v
al
u
e
g
en
er
ated
b
y
t
h
e
f
i
v
e
m
et
r
ic
m
ea
s
u
r
e
m
e
n
ts
,
Co
m
plet
e
L
i
n
k
AH
C
a
nd
F
uzzy
L
uh
n's
G
ib
bs
L
DA
w
it
h
lo
w
er
cut
-
o
f
f
h
av
e
co
n
s
is
ten
t
m
etr
ic
v
al
u
e
s
,
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
C
o
mp
lete
a
g
g
lo
mera
tive
h
iera
r
ch
y
d
o
cu
men
t’
s
clu
s
teri
n
g
b
a
s
ed
o
n
fu
z
z
y
L
u
h
n
’
s
g
i
b
b
s
…
(
P
.
M.
P
r
ih
a
tin
i
)
2109
r
an
g
i
n
g
f
r
o
m
0
.
9
1
9
6
t
o
0
.
9
9
8
9
,
w
it
h
d
if
f
er
e
n
ce
s
r
an
g
i
n
g
f
r
o
m
0
.
0
2
1
3
to
0
.
0
7
9
3
;
w
h
ile
Co
m
p
let
e
L
i
n
k
AH
C
a
nd
F
uzzy
L
uh
n
's
G
i
bb
s
L
DA
w
it
ho
ut
lo
w
er
cut
-
o
f
f
h
av
e
v
alu
e
s
r
an
g
i
n
g
f
r
o
m
0
.
8
9
3
3
t
o
0
.
9
9
7
4
,
w
it
h
d
if
f
er
e
n
ce
s
r
an
g
i
n
g
f
r
o
m
0
.
0
2
1
3
to
0
.
0
7
9
3
,
an
d
d
ec
r
ea
s
ed
t
h
e
NM
I
m
etr
ic
v
al
u
e
0
.
0
2
6
3
co
m
p
ar
ed
to
lo
w
er
cu
t
-
o
f
f
.
T
h
e
r
esu
l
ts
o
f
A
H
C
w
i
th
Fu
zz
y
L
u
h
n
’
s
Gib
b
s
L
D
A
co
m
p
ar
ed
w
i
th
th
e
r
es
u
lts
o
f
A
H
C
w
i
th
au
to
co
r
r
elatio
n
d
is
tan
ce
t
h
at
h
av
e
p
u
b
lis
h
ed
b
y
M.
A
.
A
.
R
i
y
a
d
i,
D.
S.
P
r
atiw
i,
A
.
R
.
I
r
a
w
a
n
,
K.
Fit
h
r
ia
s
ar
i [
3
4
]
.
I
n
th
eir
r
e
s
ea
r
c
h
,
Co
m
plet
e
L
in
k
AH
C
w
it
h
Aut
o
co
rr
ela
t
io
n
dis
t
a
nce
r
esu
lted
ac
c
u
r
ac
y
v
al
u
e
o
f
0
.
8
2
3
5
.
T
h
er
ef
o
r
e,
th
e
u
s
e
o
f
C
o
m
p
let
e
L
i
n
k
A
H
C
a
n
d
F
u
zz
y
L
u
h
n
's
Gib
b
s
L
D
A
w
it
h
lo
w
er
c
u
t
-
o
f
f
is
m
o
r
e
r
elev
a
n
t
as a
b
etter
clu
s
ter
i
n
g
m
et
h
o
d
in
clu
s
ter
in
g
d
o
cu
m
en
ts
esp
ec
i
al
l
y
I
n
d
o
n
e
s
ian
te
x
t
n
e
w
s
.
T
ab
le
2
.
E
v
alu
atio
n
R
e
s
u
lts
o
f
A
H
C
w
it
h
F
u
zz
y
L
u
h
n
’
s
Gib
b
s
L
D
A
M
e
t
r
i
c
s
Ev
a
l
u
a
t
i
o
n
s
F
u
z
z
y
L
u
h
n
’
s G
i
b
b
s L
D
A
w
i
t
h
L
o
w
e
r
c
u
t
-
o
f
f
F
u
z
z
y
L
u
h
n
’
s G
i
b
b
s L
D
A
w
i
t
h
o
u
t
L
o
w
e
r
c
u
t
-
o
f
f
T
h
e
d
i
f
f
e
r
e
n
c
e
S
i
n
g
l
e
L
i
n
k
A
H
C
C
o
mp
l
e
t
e
L
i
n
k
A
H
C
(
1
)
A
v
e
r
a
g
e
L
i
n
k
A
H
C
S
i
n
g
l
e
L
i
n
k
A
H
C
C
o
mp
l
e
t
e
L
i
n
k
A
H
C
(
2
)
A
v
e
r
a
g
e
L
i
n
k
A
H
C
(
1
)
(
2
)
P
r
e
c
i
si
o
n
0
.
8
2
5
5
0
.
9
5
4
9
0
.
9
1
6
9
0
.
8
6
4
2
0
.
9
5
8
3
0
.
9
3
4
0
0
.
0
0
3
4
R
e
c
a
l
l
0
.
6
0
2
1
0
.
9
2
7
3
0
.
8
1
7
9
0
.
6
7
1
4
0
.
9
2
4
7
0
.
8
2
0
1
0
.
0
0
2
6
F
-
S
c
o
r
e
0
.
6
9
6
3
0
.
9
4
0
9
0
.
8
6
4
6
0
.
7
5
5
7
0
.
9
4
1
2
0
.
8
7
3
3
0
.
0
0
0
3
N
M
I
0
.
5
8
2
7
0
.
9
1
9
6
0
.
7
2
0
8
0
.
6
0
7
5
0
.
8
9
3
3
0
.
6
4
7
4
0
.
0
2
6
3
A
R
I
0
.
9
5
2
3
0
.
9
9
8
9
0
.
9
1
2
8
0
.
9
5
3
4
0
.
9
9
7
4
0
.
9
0
1
7
0
.
0
0
1
5
4.
CO
NCLU
SI
O
N
C
o
m
p
lete
L
in
k
A
H
C
an
d
Fu
z
z
y
L
u
h
n
's
Gib
b
s
L
D
A
w
it
h
l
o
w
er
cu
t
-
o
f
f
alg
o
r
it
h
m
t
h
at
h
as
b
u
ilt
in
th
is
r
esear
ch
ca
n
i
m
p
r
o
v
e
t
h
e
q
u
alit
y
o
f
cl
u
s
ter
s
g
en
er
at
io
n
f
o
r
d
o
cu
m
e
n
t
cl
u
s
ter
in
g
e
s
p
ec
iall
y
f
o
r
I
n
d
o
n
e
s
ia
n
tex
t
n
e
w
s
.
T
h
i
s
i
s
s
h
o
w
n
b
y
t
h
e
v
al
u
e
o
f
ev
a
lu
at
io
n
m
etr
ic
s
,
w
h
ic
h
ar
e
P
r
ec
is
io
n
,
R
ec
all,
F
-
Sco
r
e,
P
er
p
lex
it
y
,
No
r
m
a
lized
Mu
t
u
al
I
n
f
o
r
m
ati
o
n
,
an
d
A
d
j
u
s
ted
R
a
n
d
I
n
d
ex
.
T
h
e
v
alu
es
o
f
P
r
ec
is
io
n
,
R
ec
all
an
d
F
-
Sco
r
e
f
o
r
lo
w
er
c
u
t
-
o
f
f
h
av
e
les
s
d
if
f
er
e
n
ce
t
h
an
w
it
h
o
u
t
th
e
lo
w
er
cu
t
-
o
f
f
,
w
h
ic
h
m
ea
n
s
b
o
th
m
e
th
o
d
s
ca
n
b
e
u
s
ed
i
n
ter
m
s
elec
t
io
n
p
r
o
ce
s
s
.
T
h
e
v
a
lu
es
o
f
P
er
p
lex
it
y
,
P
r
ec
is
io
n
,
R
ec
all
an
d
F
-
Sco
r
e
f
o
r
Fu
zz
y
L
u
h
n
'
s
Gib
b
s
L
D
A
alg
o
r
ith
m
w
as
i
n
cr
ea
s
ed
,
w
h
i
ch
m
ea
n
s
it
p
er
f
o
r
m
ed
b
etter
th
an
F
u
zz
y
Gib
b
s
L
D
A
.
T
h
e
v
alu
e
o
f
P
r
ec
i
s
io
n
,
R
ec
all,
F
-
Sco
r
e,
P
er
p
lex
it
y
,
No
r
m
a
lized
Mu
tu
al
I
n
f
o
r
m
at
io
n
,
an
d
A
d
j
u
s
ted
R
a
n
d
I
n
d
ex
s
h
o
w
ed
t
h
at
t
h
e
C
o
m
p
lete
L
i
n
k
AHC
an
d
Fu
z
z
y
L
u
h
n
'
s
Gib
b
s
L
D
A
al
g
o
r
it
h
m
as
t
h
e
b
est
A
HC
cl
u
s
ter
i
n
g
al
g
o
r
ith
m
,
w
ith
o
r
w
it
h
o
u
t
lo
w
er
cu
t
-
o
f
f
.
B
u
t,
th
e
C
o
m
p
lete
L
i
n
k
A
H
C
alg
o
r
it
h
m
a
n
d
Fu
zz
y
L
u
h
n
'
s
Gib
b
s
L
DA
w
it
h
lo
w
er
cu
t
-
o
f
f
p
r
o
d
u
ce
m
o
r
e
co
n
s
i
s
te
n
c
y
v
alu
e
f
o
r
th
e
f
iv
e
m
etr
ic
m
e
asu
r
e
m
en
t
s
,
w
h
ich
m
ea
n
s
is
m
o
r
e
r
ele
v
an
t
to
it
s
g
o
ld
s
tan
d
ar
d
.
RE
F
E
R
E
NC
E
S
[1
]
Ba
n
sa
l
A
.
,
e
t
a
l
.
,
“
Im
p
ro
v
e
d
K
-
m
e
a
n
c
lu
ste
rin
g
a
lg
o
rit
h
m
f
o
r
p
r
e
d
ictio
n
a
n
a
ly
sis
u
sin
g
c
las
sif
ic
a
ti
o
n
tec
h
n
i
q
u
e
in
d
a
ta m
in
in
g
,
”
In
ter
n
a
t
io
n
a
l
J
o
u
rn
a
l
o
f
Co
m
p
u
ter
A
p
p
l
ica
ti
o
n
s
,
v
o
l.
1
5
7
,
p
p
.
3
5
-
4
0
,
2
0
1
7
.
[2
]
A
ro
ra
P
.
,
e
t
a
l
.,
“
A
n
a
l
y
sis
o
f
K
-
M
e
a
n
s
a
n
d
K
-
M
e
d
o
i
d
s
a
lg
o
rit
h
m
f
o
r
b
ig
d
a
ta,
”
Pro
c
e
d
ia
Co
mp
u
t
e
r
S
c
ien
c
e
,
v
o
l.
7
8
,
p
p
.
50
7
-
1
2
,
2
0
1
6
.
[3
]
W
a
h
y
u
d
in
I.
,
e
t
a
l
.
,
“
Clu
ste
r
a
n
a
ly
sis
f
o
r
S
M
E
risk
a
n
a
ly
sis
d
o
c
u
m
e
n
ts
b
a
se
d
o
n
P
il
lar
K
-
M
e
a
n
s,
”
T
EL
KOM
NIKA
T
e
lec
o
mm
u
n
ica
ti
o
n
Co
mp
u
ti
n
g
E
lctro
n
ics
a
n
d
Co
n
tro
l
,
v
o
l.
1
4
,
p
p
.
6
7
4
-
8
3
,
2
0
1
6
.
[4
]
Ba
th
la
G
.
,
e
t
a
l
.,
“
A
n
o
v
e
l
a
p
p
ro
a
c
h
f
o
r
c
lu
ste
ri
n
g
b
ig
d
a
ta
b
a
se
d
o
n
M
a
p
Re
d
u
c
e
,
”
In
ter
n
a
ti
o
n
a
l
J
o
u
r
n
a
l
o
f
El
e
c
trica
l
a
n
d
C
o
mp
u
ter
En
g
in
e
e
rin
g
(
IJ
ECE
)
,
v
o
l.
8
,
p
p
.
1
7
1
1
,
2
0
1
8
.
[5
]
Ka
m
a
t
R.
K.,
e
t
a
l
.
,
“
Earth
q
u
a
k
e
c
lu
ste
r
a
n
a
ly
sis:
K
-
M
e
a
n
s
a
p
p
ro
a
c
h
,
”
J
o
u
rn
a
l
o
f
Ch
e
mic
a
l
a
n
d
P
h
a
rm
a
c
e
u
ti
c
a
l
S
c
ien
c
e
s
,
v
o
l.
1
0
,
p
p
.
2
5
0
-
3
,
2
0
1
7
.
[6
]
S
trau
ss
T
.
,
e
t
a
l
.
,
“
G
e
n
e
ra
li
sin
g
wa
rd
'
s
m
e
th
o
d
f
o
r
u
se
w
it
h
m
a
n
h
a
tt
a
n
d
istan
c
e
s,
”
Pl
o
S
o
n
e
,
v
o
l.
1
2
,
2
0
1
7
.
[7
]
M
a
o
Y.,
e
t
a
l
.,
“
A
n
a
d
a
p
ti
v
e
traje
c
to
ry
c
lu
ste
rin
g
m
e
th
o
d
b
a
se
d
o
n
g
rid
a
n
d
d
e
n
s
it
y
in
m
o
b
il
e
p
a
tt
e
rn
a
n
a
ly
sis,
”
S
e
n
so
rs
(
Ba
se
l)
,
v
o
l.
1
7
,
2
0
1
7
.
[8
]
M
a
ju
m
d
a
r
J.,
e
t
a
l
.,
“
A
n
a
l
y
sis
o
f
a
g
ricu
lt
u
re
d
a
ta
u
sin
g
d
a
ta
m
in
in
g
tec
h
n
iq
u
e
s:
a
p
p
li
c
a
ti
o
n
o
f
b
ig
d
a
ta,
”
J
o
u
rn
a
l
o
f
Bi
g
Da
ta
,
v
o
l.
4
,
2
0
1
7
.
[9
]
Ba
lca
n
M
.
F
.
,
e
t
a
l
.,
“
Ro
b
u
st
h
ier
a
rc
h
ica
l
c
lu
ste
rin
g
,
”
J
o
u
rn
a
l
o
f
M
a
c
h
in
e
L
e
a
rn
i
n
g
Res
e
a
rc
h
,
v
o
l.
1
5
,
p
p
.
4
0
1
1
-
51
,
2
0
1
4
.
[1
0
]
G
o
u
las
C.
,
e
t
a
l
.
,
“
HCu
RM
D:
Hie
ra
rc
h
ica
l
c
lu
ste
rin
g
u
sin
g
re
lativ
e
m
in
ima
l
d
istan
c
e
s
,”
i
n
Ch
b
e
ir
R.
M
.
Y
.
,
e
t
a
l
.
,
“
A
rti
f
icia
l
In
telli
g
e
n
c
e
A
p
p
li
c
a
ti
o
n
s
a
n
d
I
n
n
o
v
a
ti
o
n
s,
”
I
F
I
P
A
d
v
a
n
c
e
s
in
In
f
o
rm
a
ti
o
n
a
n
d
C
o
m
m
u
n
ica
ti
o
n
T
e
c
h
n
o
lo
g
y
,
S
p
rin
g
e
r
,
p
p
.
4
4
0
-
7
,
2
0
1
5
.
[1
1
]
M
a
ra
th
e
M
.
,
e
t
a
l
.
,
“
A
su
rv
e
y
o
f
c
lu
ste
rin
g
a
l
g
o
rit
h
m
s
f
o
r
si
m
il
a
rit
y
s
e
a
rc
h
,
”
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
P
u
re
a
n
d
Ap
p
li
e
d
M
a
t
h
e
ma
ti
c
s
,
v
o
l.
1
1
4
,
p
p
.
3
4
3
-
5
1
,
2
0
1
7
.
[1
2
]
A
ll
a
h
y
a
ri
M
.
,
e
t
a
l
.,
“
A
b
rie
f
su
rv
e
y
o
f
tex
t
m
in
in
g
:
c
las
si
f
ica
ti
o
n
,
c
lu
ste
r
in
g
a
n
d
e
x
trac
ti
o
n
tec
h
n
iq
u
e
s,
”
KDD
Big
d
a
s,
2
0
1
7
.
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
.
3
,
J
u
n
e
2
0
1
9
:
2
1
0
3
-
2
1
1
1
2110
[1
3
]
Rib
e
iro
M
.
N.,
e
t
a
l
.
,
“
L
o
c
a
l
f
e
a
t
u
re
se
lec
ti
o
n
i
n
tex
t
c
lu
ste
rin
g
,”
i
n
Kö
p
p
e
n
M
.
K
.
N
.
a
n
d
Co
g
h
il
l
G
.
,
“
A
d
v
a
n
c
e
s
in
Ne
u
ro
-
In
f
o
rm
a
ti
o
n
P
r
o
c
e
ss
in
g
ICON
IP
2
0
0
8
,
”
L
e
c
tu
re
No
tes
in
Co
m
p
u
ter
S
c
ien
c
e
.
Be
rli
n
,
He
i
d
e
lb
e
rg
,
S
p
rin
g
e
r
,
2
0
0
9
.
[1
4
]
M
a
n
n
i
n
g
C.
D.,
e
t
a
l
.
,
“
A
n
In
tro
d
u
c
ti
o
n
to
In
f
o
rm
a
ti
o
n
Re
tri
e
v
a
l
,”
En
g
lan
d
,
Ca
m
b
rid
g
e
Un
iv
e
rsit
y
P
re
ss
,
2
0
0
8
.
[1
5
]
Isla
m
M
.
R.
,
e
t
a
l
.
,
“
T
e
c
h
n
ica
l
a
p
p
ro
a
c
h
i
n
tex
t
m
in
in
g
f
o
r
s
to
c
k
m
a
rk
e
t
p
re
d
ictio
n
:
a
sy
ste
m
a
ti
c
re
v
ie
w
,
”
In
d
o
n
e
sia
n
J
o
u
rn
a
l
o
f
El
e
c
trica
l
En
g
i
n
e
e
rin
g
a
n
d
C
o
mp
u
ter
S
c
ien
c
e
,
v
o
l.
1
0
,
p
p
.
7
7
0
-
7
,
2
0
1
8
.
[1
6
]
Am
o
li
P
.
V
.
,
e
t
a
l
.
,
“
S
c
ien
ti
f
ic
d
o
u
m
e
n
ts
c
lu
ste
ri
n
g
b
a
se
d
o
n
tex
t
su
m
m
a
riz
a
ti
o
n
,
”
I
n
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
in
e
e
rin
g
(
IJ
ECE
)
,
v
o
l.
5
,
p
p
.
7
8
2
-
7
,
2
0
1
5
.
[1
7
]
Ra
m
a
g
e
D.,
e
t
a
l
.,
“
Cl
u
ste
rin
g
th
e
tag
g
e
d
w
e
b
,
”
Pro
c
e
e
d
in
g
s
o
f
t
h
e
S
e
c
o
n
d
ACM
In
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
W
e
b
S
e
a
rc
h
a
n
d
Da
t
a
M
in
i
n
g
,
2
0
0
9
.
[1
8
]
Kn
ij
f
f
J.,
e
t
a
l
.,
“
Do
m
a
in
tax
o
n
o
m
y
lea
rn
in
g
f
ro
m
te
x
t:
th
e
su
b
su
m
p
ti
o
n
m
e
th
o
d
v
e
rsu
s
h
iera
rc
h
ica
l
c
lu
ste
rin
g
,
”
Da
t
a
&
Kn
o
wled
g
e
E
n
g
i
n
e
e
rin
g
,
v
o
l.
8
3
,
p
p
.
5
4
-
6
9
,
2
0
1
3
.
[1
9
]
In
d
h
u
ja
A
.
,
e
t
a
l
.,
“
A
m
u
lt
i
-
k
e
y
w
o
rd
ra
n
k
e
d
se
a
rc
h
sc
h
e
m
e
o
v
e
r
e
n
c
ry
p
ted
b
a
se
d
o
n
h
iera
rc
h
ica
l
c
lu
ste
rin
g
in
d
e
x
,
”
In
ter
n
a
t
io
n
a
l
J
o
u
rn
a
l
O
n
S
m
a
rt S
e
n
sin
g
An
d
In
telli
g
e
n
t
S
y
ste
ms
,
v
o
l.
1
0
,
p
p
.
5
3
9
-
5
9
,
2
0
1
7
.
[2
0
]
V
e
n
k
a
tes
h
k
u
m
a
r
P
.
,
e
t
a
l
.
,
“
Us
in
g
d
a
ta
f
u
sio
n
f
o
r
a
c
o
n
tex
t
a
wa
re
d
o
c
u
m
e
n
t
c
lu
ste
rin
g
,
”
I
n
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
Co
mp
u
ter
A
p
p
li
c
a
ti
o
n
s
,
v
o
l.
7
2
,
p
p
.
1
7
-
2
0
,
2
0
1
3
.
[2
1
]
S
o
n
g
Y.
,
e
t
a
l
.
,
“
A
u
to
m
a
ti
c
ta
x
o
n
o
m
y
c
o
n
stru
c
ti
o
n
f
ro
m
k
e
y
w
o
rd
s
v
i
a
sc
a
lab
le
b
a
y
e
sia
n
ro
se
t
re
e
s,
”
IEE
E
T
ra
n
sa
c
ti
o
n
s
o
n
Kn
o
wled
g
e
a
n
d
Da
ta
E
n
g
in
e
e
rin
g
,
v
o
l.
2
7
,
p
p
.
1
8
6
1
-
7
4
,
2
0
1
5
.
[2
2
]
Ku
ta
M
.
,
e
t
a
l
.
,
“
Co
m
p
a
riso
n
o
f
late
n
t
se
m
a
n
ti
c
a
n
a
l
y
sis
a
n
d
p
ro
b
a
b
il
isti
c
late
n
t
se
m
a
n
ti
c
a
n
a
l
y
sis
f
o
r
d
o
c
u
m
e
n
ts
c
lu
ste
rin
g
,
”
Co
mp
u
ti
n
g
a
n
d
I
n
fo
r
ma
ti
c
s
,
v
o
l.
3
3
,
p
p
.
6
52
–
6
6
,
2
0
1
4
.
[2
3
]
Ro
tt
M
.
,
e
t
a
l
.,
“
In
v
e
stig
a
ti
o
n
o
f
late
n
t
se
m
a
n
ti
c
a
n
a
l
y
sis
f
o
r
c
lu
st
e
rin
g
o
f
Cz
e
c
h
n
e
ws
a
rti
c
les
,
”
2
5
th
In
ter
n
a
ti
o
n
a
l
W
o
rk
sh
o
p
o
n
D
a
ta
b
a
se
a
n
d
Exp
e
rt S
y
ste
ms
Ap
p
li
c
a
ti
o
n
s (
DEXA
)
,
2
0
1
4
.
[2
4
]
P
a
rk
H.,
e
t
a
l
.
,
“
A
g
g
lo
m
e
ra
ti
v
e
h
iera
rc
h
ica
l
c
lu
ste
rin
g
f
o
r
in
f
o
rm
a
t
io
n
re
tri
e
v
a
l
u
si
n
g
late
n
t
se
m
a
n
ti
c
in
d
e
x
,
”
IEE
E
In
ter
n
a
t
io
n
a
l
C
o
n
fer
e
n
c
e
o
n
S
ma
rt Ci
ty/S
o
c
i
a
lCo
m/
S
u
st
a
in
C
o
m (
S
ma
rtCit
y
)
,
p
p
.
4
2
6
-
31
,
2
0
1
5
.
[2
5
]
Blei
D.
M
.
,
e
t
a
l
.,
“
L
a
ten
t
d
iri
c
h
l
e
t
a
l
l
o
c
a
ti
o
n
,
”
J
o
u
r
n
a
l
o
f
M
a
c
h
in
e
L
e
a
rn
in
g
Res
e
a
rc
h
,
v
o
l.
3
,
p
p
.
9
9
3
-
1
0
2
2
,
2
0
0
3
.
[2
6
]
P
ri
h
a
ti
n
i
P
.
M
.
,
e
t
a
l
.,
“
F
u
z
z
y
-
g
ib
b
s
late
n
t
d
iri
c
h
l
e
t
a
l
l
o
c
a
ti
o
n
m
o
d
e
l
f
o
r
f
e
a
tu
re
e
x
tr
a
c
ti
o
n
o
n
In
d
o
n
e
sia
n
d
o
c
u
m
e
n
ts,
”
Co
n
tem
p
o
ra
ry
E
n
g
i
n
e
e
rin
g
S
c
ien
c
e
s
,
v
o
l.
1
0
,
p
p
.
4
0
3
-
2
1
,
2
0
1
7
.
[2
7
]
P
ri
h
a
ti
n
i
P
.
M
.
,
e
t
a
l
.,
“
F
e
a
tu
re
e
x
trac
ti
o
n
f
o
r
d
o
c
u
m
e
n
t
tex
t
u
sin
g
L
a
ten
t
Dirich
l
e
t
a
l
lo
c
a
ti
o
n
,
”
J
o
u
rn
a
l
o
f
P
h
y
sic
s:
Co
n
fer
e
n
c
e
S
e
rie
s
,
v
o
l.
9
5
3
,
p
p
.
0
1
2
0
4
7
,
2
0
1
8
.
[2
8
]
P
ri
h
a
ti
n
i
P
.
M
.
,
e
t
a
l
.,
“
In
d
o
n
e
si
a
n
tex
t
f
e
a
tu
re
e
x
trac
ti
o
n
u
sin
g
g
ib
b
s
sa
m
p
li
n
g
a
n
d
m
e
a
n
v
a
riatio
n
a
l
i
n
f
e
re
n
c
e
late
n
t
d
iri
c
h
l
e
t
a
l
l
o
c
a
ti
o
n
,
”
Qu
a
li
ty
o
f
Res
e
a
rc
h
(
QIR):
In
ter
n
a
t
io
n
a
l
S
y
mp
o
siu
m
o
n
El
e
c
trica
l
a
n
d
C
o
mp
u
ter
En
g
i
n
e
e
rin
g
,
2
0
1
7
1
5
t
h
In
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
,
2
0
1
7
.
[2
9
]
Blei
D.
M
.
,
e
t
a
l
.,
“
Hie
ra
rc
h
ica
l
t
o
p
ic
m
o
d
e
ls
a
n
d
th
e
n
e
s
ted
c
h
in
e
se
re
sta
u
ra
n
t
p
ro
c
e
ss
,
”
NIPS
'
0
3
Pro
c
e
e
d
in
g
s
o
f
th
e
1
6
th
In
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
Ne
u
ra
l
I
n
f
o
rm
a
ti
o
n
Pr
o
c
e
ss
in
g
S
y
ste
ms
,
p
p.
17
-
24
,
2
0
0
3
.
[3
0
]
Ye
re
b
a
k
a
n
H.
Z.
,
e
t
a
l
.,
“
Hie
ra
rc
h
ica
l
late
n
t
w
o
rd
c
lu
ste
rin
g
,
”
B
a
y
e
sia
n
No
n
p
a
ra
me
trics
:
T
h
e
N
e
x
t
Ge
n
e
ra
ti
o
n
NIPS
2
0
1
5
W
o
rk
sh
o
p
,
2
0
1
5
.
[3
1
]
L
i
X
.
,
e
t
a
l
.,
“
In
d
u
c
i
n
g
tax
o
n
o
m
y
f
ro
m
ta
g
s
:
a
n
a
g
g
lo
m
e
ra
ti
v
e
h
iera
rc
h
ica
l
c
lu
ste
rin
g
f
ra
m
e
w
o
rk
,
”
In
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
Ad
v
a
n
c
e
d
Da
t
a
M
in
in
g
a
n
d
Ap
p
li
c
a
ti
o
n
s
ADM
A
2
0
1
2
:
Ad
v
a
n
c
e
d
Da
t
a
M
in
in
g
a
n
d
Ap
p
li
c
a
ti
o
n
s
,
p
p
.
6
4
-
77
,
2
0
1
2
.
[3
2
]
T
u
D.,
e
t
a
l
.,
“
W
o
rd
Ne
t
b
a
se
d
m
u
lt
i
-
w
a
y
c
o
n
c
e
p
t
h
iera
rc
h
y
c
o
n
stru
c
ti
o
n
f
ro
m
tex
t
c
o
rp
u
s
,”
Pro
c
e
e
d
in
g
s
o
f
th
e
T
we
n
ty
-
S
e
v
e
n
th
AA
AI
C
o
n
fer
e
n
c
e
o
n
Arti
fi
c
ia
l
In
telli
g
e
n
c
e
,
p
p
.
1
6
4
7
-
8
,
2
0
1
3
.
[3
3
]
Yu
h
e
f
iza
r
Y.,
e
t
a
l
.
,
“
Co
m
b
in
a
ti
o
n
o
f
Clu
ste
r
M
e
th
o
d
f
o
r
S
e
g
m
e
n
tatio
n
o
f
W
e
b
V
isit
o
rs
,”
T
EL
KOM
NIKA
(
T
e
lec
o
mm
u
n
ica
ti
o
n
Co
mp
u
ti
n
g
El
e
c
tro
n
ics
a
n
d
C
o
n
tro
l)
,
v
o
l.
1
1
,
p
p
.
2
0
7
,
2
0
1
3
.
[3
4
]
R
iy
a
d
i
M
.
A
.
A
.
,
e
t
a
l
.,
“
Clu
ste
rin
g
sta
ti
o
n
a
ry
a
n
d
n
o
n
-
sta
ti
o
n
a
ry
ti
m
e
se
ries
b
a
se
d
o
n
a
u
to
c
o
rre
lat
io
n
d
ista
n
c
e
o
f
h
iera
rc
h
ica
l
a
n
d
k
-
m
e
a
n
s
a
lg
o
rit
h
m
s,
”
In
ter
n
a
ti
o
n
a
l
J
o
u
r
n
a
l
o
f
Ad
v
a
n
c
e
s
in
In
telli
g
e
n
t
In
f
o
rm
a
ti
c
s
,
v
o
l.
3
,
p
p
.
1
5
4
-
6
0
,
2
0
1
7
.
[3
5
]
A
sia
n
J.,
e
t
a
l
.,
“
S
tem
m
in
g
In
d
o
n
e
sia
n
,
”
Pro
c
e
e
d
in
g
s
o
f
t
h
e
T
we
n
ty
-
e
ig
h
t
h
Au
str
a
la
si
a
n
c
o
n
fer
e
n
c
e
o
n
Co
mp
u
ter
S
c
ien
c
e
,
2
0
0
4
.
[3
6
]
Ko
c
a
b
a
s
I.
,
e
t
a
l
.,
“
I
n
v
e
stig
a
ti
o
n
o
f
L
u
h
n
’s
c
laim
o
n
in
f
o
rm
a
ti
o
n
r
e
tri
e
v
a
l,
”
T
u
rk
J
El
e
c
En
g
&
Co
mp
S
c
i
,
v
o
l.
19,
p
p
.
9
9
3
-
1
0
0
4
,
2
0
1
1
.
[3
7
]
He
in
rich
G
.
,
“
P
a
ra
m
e
ter es
ti
m
a
ti
o
n
f
o
r
tex
t
a
n
a
ly
sis
,”
Un
iv
e
r
sit
y
o
f
L
e
ip
z
ig
,
G
e
r
m
a
n
y
,
2
0
0
8
.
[3
8
]
F
re
d
A
.
L
.
N.,
e
t
a
l
.,
“
Ro
b
u
st
Da
ta
Clu
ste
rin
g
,
”
Pro
c
e
e
d
in
g
s
o
f
IE
EE
Co
mp
u
ter
S
o
c
iety
Co
n
fer
e
n
c
e
o
n
Co
m
p
u
te
r
Vi
sio
n
a
n
d
P
a
tt
e
rn
Rec
o
g
n
it
i
o
n
,
v
o
l.
3
,
p
p
.
1
2
8
-
3
6
,
2
0
0
3
.
[3
9
]
Ku
n
c
h
e
v
a
L
.
I.
,
e
t
a
l
.,
“
Us
in
g
d
iv
e
rsit
y
in
c
lu
ste
r
e
n
se
m
b
les
,
”
v
o
l.
2
,
p
p
.
1
2
1
4
-
9
,
2
0
0
4
.
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
Pu
tu
M
a
n
i
k
Prih
a
tin
i
w
a
s
b
o
rn
in
Ba
li
(In
d
o
n
e
sia
)
o
n
M
a
rc
h
1
7
,
1
9
8
0
.
S
h
e
is
e
a
rn
e
d
a
b
a
c
h
e
lo
r'
s
d
e
g
re
e
o
f
in
f
o
r
m
a
ti
c
s
e
n
g
in
e
e
rin
g
a
t
S
e
k
o
lah
T
in
g
g
i
T
e
k
n
o
lo
g
i
T
e
lk
o
m
Ba
n
d
u
n
g
(In
d
o
n
e
sia
)
in
2
0
0
2
,
a
n
d
h
e
r
m
a
ste
r
'
s
d
e
g
r
e
e
a
t
Un
iv
e
rsitas
Ud
a
y
a
n
a
(In
d
o
n
e
sia
)
in
2
0
1
2
.
S
h
e
is
a
lec
tu
re
r
a
t
P
o
li
tek
n
ik
Ne
g
e
ri
Ba
li
(In
d
o
n
e
sia
)
sin
c
e
2
0
0
2
u
n
ti
l
n
o
w
.
Cu
rre
n
tl
y
,
sh
e
is
b
e
in
g
a
d
o
c
t
o
ra
l
stu
d
e
n
t
o
f
Ilm
u
T
e
k
n
ik
a
t
Un
iv
e
rsitas
Ud
a
y
a
n
a
(In
d
o
n
e
sia
).
He
r
re
se
a
rc
h
a
n
d
in
tere
st
is
T
e
x
t
M
in
in
g
,
In
f
o
rm
a
ti
o
n
Re
tri
e
v
a
l
S
y
ste
m
,
a
n
d
S
o
f
t
Co
m
p
u
ti
n
g
.
P
u
t
u
M
a
n
ik
P
ri
h
a
ti
n
i,
S
T
,
M
T
.
E
m
a
il
:
m
a
n
ik
p
ri
h
a
ti
n
i@
p
n
b
.
a
c
.
id
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
C
o
mp
lete
a
g
g
lo
mera
tive
h
iera
r
ch
y
d
o
cu
men
t’
s
clu
s
teri
n
g
b
a
s
ed
o
n
fu
z
z
y
L
u
h
n
’
s
g
i
b
b
s
…
(
P
.
M.
P
r
ih
a
tin
i
)
2111
I
K
e
tu
t
G
e
d
e
D
a
r
m
a
P
u
tr
a
wa
s
b
o
rn
in
Ba
li
(In
d
o
n
e
sia
)
o
n
A
p
ril
2
4
,
1
9
7
4
.
He
is
e
a
rn
e
d
a
b
a
c
h
e
lo
r'
s
d
e
g
re
e
o
f
in
f
o
rm
a
ti
c
s
a
t
In
stit
u
t
T
e
k
n
o
lo
g
i
S
e
p
u
l
u
h
N
o
v
e
m
b
e
r
(IT
S
-
In
d
o
n
e
sia
);
m
a
ste
rs
a
n
d
d
o
c
to
ra
l
d
e
g
re
e
s
a
t
Un
iv
e
r
sitas
G
a
d
jah
M
a
d
a
(U
GM
-
In
d
o
n
e
si
a
).
He
is
a
lec
tu
re
r
a
t
Un
iv
e
rsitas
Ud
a
y
a
n
a
(In
d
o
n
e
sia
)
sin
c
e
1
9
9
9
u
n
ti
l
n
o
w
.
His
re
se
a
rc
h
a
n
d
in
t
e
re
st
is
Da
ta
M
in
in
g
a
n
d
Im
a
g
e
P
r
o
c
e
ss
in
g
.
P
ro
f
.
Dr.
I
Ke
t
u
t
G
e
d
e
Da
r
m
a
P
u
tra,
S
.
Ko
m
.
,
M
T
.
Ema
il
:
ik
g
d
a
r
m
a
p
u
tra@u
n
u
d
.
a
c
.
i
d
Id
a
Ay
u
Dw
i
G
iria
n
ta
ri
w
a
s
b
o
rn
i
n
Ba
li
(I
n
d
o
n
e
sia
)
o
n
De
c
e
m
b
e
r
1
3
,
1
9
6
5
.
S
h
e
is
e
a
rn
e
d
a
b
a
c
h
e
lo
r'
s
d
e
g
re
e
o
f
e
lec
tri
c
a
l
e
n
g
in
e
e
rin
g
a
t
Un
iv
e
rsitas
Ud
a
y
a
n
a
(In
d
o
n
e
sia
);
m
a
ste
r
s
a
n
d
d
o
c
to
ra
l
d
e
g
re
e
s
a
t
T
h
e
Un
iv
e
rsit
y
o
f
Ne
w
S
o
u
th
W
a
les
(
A
u
stra
li
a
).
S
h
e
is a l
e
c
tu
re
r
a
t
Un
iv
e
r
sitas
Ud
a
y
a
n
a
In
d
o
n
e
sia
sin
c
e
1
9
9
1
u
n
ti
l
n
o
w
.
He
r
re
se
a
r
c
h
a
n
d
in
t
e
re
st
is
e
le
c
tri
c
p
o
w
e
r
s
y
ste
m
,
re
n
e
wa
b
le
e
n
e
rg
y
tec
h
n
o
lo
g
y
a
n
d
a
p
p
li
c
a
ti
o
n
,
sm
a
rt
g
rid
a
n
d
c
o
n
tro
l.
P
r
o
f
.
Ir.
Id
a
A
y
u
D
w
i
G
iri
a
n
tari,
M
.
En
g
,
S
c
.
,
P
h
D
.
Em
a
i
l:
d
a
y
u
.
g
iri
a
n
tari
@u
n
u
d
.
a
c
.
id
M
a
d
e
S
u
d
a
r
m
a
wa
s
b
o
rn
in
Ba
li
(In
d
o
n
e
sia
)
o
n
De
c
e
m
b
e
r
3
1
,
1
9
6
5
.
He
is
e
a
rn
e
d
a
b
a
c
h
e
lo
r'
s
d
e
g
re
e
o
f
in
f
o
r
m
a
ti
c
s
a
t
In
stit
u
t
T
e
k
n
o
lo
g
i
S
e
p
u
lu
h
No
p
e
m
b
e
r
(IT
S
-
In
d
o
n
e
sia
);
m
a
ste
r’s
d
e
g
re
e
a
t
S
c
h
o
o
l
o
f
In
f
o
rm
a
ti
o
n
T
e
c
h
n
o
lo
g
y
a
n
d
En
g
in
e
e
rin
g
,
Ottaw
a
Un
iv
e
rsit
y
(C
a
n
a
d
a
)
a
n
d
d
o
c
t
o
ra
l
d
e
g
re
e
s
a
t
Un
iv
e
rsitas
Ud
a
y
a
n
a
(In
d
o
n
e
sia
).
He
is
a
l
e
c
tu
re
r
a
t
Un
iv
e
rsitas
Ud
a
y
a
n
a
In
d
o
n
e
sia
sin
c
e
1
9
9
3
u
n
ti
l
n
o
w
.
His
re
se
a
rc
h
a
n
d
i
n
tere
st
is
Da
ta
M
in
i
n
g
a
n
d
Im
a
g
e
P
ro
c
e
ss
in
g
.
Dr.
Ir.
M
a
d
e
S
u
d
a
rm
a
,
M
.
A
.
S
c
.
Em
a
il
:
m
su
d
a
r
m
a
@u
n
u
d
.
a
c
.
id
Evaluation Warning : The document was created with Spire.PDF for Python.