T
E
L
K
O
MN
I
K
A
T
el
eco
m
m
u
n
i
ca
t
i
o
n
,
C
o
m
p
u
t
i
n
g
,
E
l
ect
ro
n
i
cs
a
n
d
C
o
n
t
ro
l
Vo
l
.
19
, N
o
.
5
,
O
ct
o
b
er
2021
, p
p
.
1605
~
1
611
I
S
S
N
:
1693
-
6930,
a
c
c
r
e
di
t
e
d F
i
r
s
t
G
r
a
de
by K
e
m
e
nr
i
s
t
e
kdi
kt
i
,
D
e
c
r
e
e
N
o:
21/
E
/
K
P
T
/
2018
D
O
I
:
10.
12928/
T
E
L
K
O
M
N
I
K
A
.
v19i
5.
20406
1605
Jou
r
n
al
h
om
e
page
:
ht
t
p:
/
/
j
our
nal
.
uad
.
ac
.
i
d/
i
nde
x
.
php/
T
E
L
K
O
M
N
I
K
A
C
erv
i
ca
l
ca
n
ce
r
cl
a
s
s
i
f
i
ca
t
i
o
n
us
i
ng
co
nv
o
l
ut
i
o
na
l
ne
ura
l
net
w
o
r
k
-
s
u
ppo
rt
v
ec
t
o
r m
a
ch
i
ne
Jan
e
E
va A
u
r
e
l
i
a
,
Z
uhe
r
m
a
n R
us
t
a
m
,
I
l
s
ya Wi
r
as
at
i
De
pa
r
tm
e
n
t of
M
a
t
he
m
a
t
ic
s,
F
a
c
u
lt
y of
M
a
t
he
m
a
t
ic
s a
n
d Na
t
ur
a
l S
c
ie
nc
e
s,
Un
ive
r
si
ta
s I
n
do
ne
s
ia
,
De
po
k,
I
nd
on
e
sia
A
rt
i
cl
e I
n
f
o
AB
S
T
RACT
A
r
tic
le
h
is
to
r
y
:
R
ecei
v
ed
J
ul
22,
2020
R
ev
i
s
ed
J
ul
16,
2021
A
ccep
t
ed
A
ug 1,
2021
C
e
r
vic
a
l c
a
nc
e
r
is the
se
c
on
d m
os
t c
om
m
o
n c
a
nc
e
r
i
n wom
e
n
w
or
l
dw
ide
,
a
nd oc
c
ur
s w
he
n t
he
r
e
a
r
e
pr
e
se
nc
e
s of
a
bn
or
m
a
l c
e
ll
s
in the
c
e
r
v
ix,
wh
ic
h
c
on
ti
nue
t
o gr
ow
unc
on
tr
o
lla
bl
y.
I
n
the
e
a
r
ly
sta
ge
s,
c
e
r
v
ic
a
l c
a
nc
e
r
ind
ic
a
ti
on
s
a
r
e
not pe
r
c
e
p
ti
ble
;
ho
we
ve
r
,
i
t is
e
a
s
il
y
de
te
c
te
d
wi
th
dif
f
e
r
e
n
t
f
or
m
s of
m
a
c
hi
ne
le
a
r
n
in
g m
e
tho
ds,
s
uc
h a
s the
c
on
vo
lu
ti
ona
l ne
ur
a
l
ne
t
wor
k (
C
NN)
.
T
hi
s i
s a
po
pu
la
r
m
e
t
ho
d wi
th a
w
id
e
r
a
nge
of
a
pp
lic
a
ti
on
s
a
nd
kn
ow
n f
or
it
s h
ig
h a
c
c
ur
a
c
y va
lue
.
M
or
e
o
ve
r
,
the
r
e
is a
su
pp
or
t
ve
c
t
or
m
a
c
hi
ne
(
S
VM
)
wi
th se
ve
r
a
l
ke
r
ne
l f
unc
ti
on
s tha
t i
s c
om
m
on
ly
use
d in t
he
c
la
s
sifi
c
a
ti
on
of
d
ise
a
se
s,
a
n
d
a
ls
o
kn
ow
n
f
or
it
s
h
igh
a
c
c
ur
a
c
y
va
lue
.
The
r
e
f
or
e
,
t
he
c
om
bi
na
t
io
n
of
C
N
N
–
S
VM
wi
th
se
ve
r
a
l
li
ne
a
r
ke
r
ne
ls
f
unc
ti
on
s
a
s
c
la
ss
if
ie
r
f
or
the
c
a
te
go
r
iz
a
ti
on
of
c
e
r
vic
a
l
c
a
nc
e
r
.
Ke
y
wo
r
d
s
:
C
er
v
i
cal
can
cer
C
la
s
s
if
ic
a
ti
on
C
onvol
ut
i
ona
l
ne
ur
a
l
ne
t
w
or
k
M
ach
i
n
e l
ear
n
i
n
g
S
uppor
t
ve
c
t
or
m
a
c
hi
ne
T
his
is
a
n
o
pe
n
ac
c
e
s
s
ar
tic
le
u
nde
r
the
CC
B
Y
-
SA
lic
e
n
se
.
C
or
r
e
s
pon
di
n
g A
u
t
h
or
:
J
an
e E
v
a A
u
r
el
i
a
D
ep
ar
t
m
en
t
o
f
M
at
h
em
at
i
cs
U
ni
ve
r
s
i
t
y
of
I
ndone
s
i
a
P
r
of
.
D
R
.
S
udj
ono D
.
P
us
pone
gor
o
S
t.
,
P
ondok C
i
na
,
D
e
pok,
Wes
t
J
av
a
16424,
I
ndone
s
i
a
E
ma
il:
co
r
r
es
p
-
a
ut
hor
@
m
a
i
l
.
c
om
1.
I
NT
RO
DUC
T
I
O
N
M
a
ny c
ount
r
i
e
s
r
a
nke
d c
a
nc
e
r
a
s
t
he
s
e
c
ond m
os
t
c
om
m
on he
a
l
t
h
i
s
s
ue
s
,
of
w
hi
c
h
cer
v
i
cal
can
cer
cau
s
es
t
h
e m
o
s
t
o
f
d
eat
h
s
r
eco
r
d
ed
[
1
]
.
C
er
v
i
cal
can
cer
o
ccu
r
s
w
h
en
t
h
er
e
ar
e
ab
n
o
r
m
al
cel
l
s
i
n
t
h
e cer
v
i
x
,
w
hi
c
h c
ont
i
nue
s
t
o
gr
ow
unc
ont
r
ol
l
a
bl
y,
a
nd
r
e
s
ul
t
s
i
n be
ni
gn
t
um
o
r
s
,
w
hi
c
h
l
a
t
e
r
de
ve
l
ops
i
nt
o
c
e
r
vi
c
a
l
c
a
nc
e
r
c
e
l
l
s
t
ha
t
s
pr
e
a
d t
o ot
he
r
body pa
r
t
s
[
2]
.
T
hi
s
c
a
nc
e
r
i
s
one
of
t
he
m
os
t
c
om
m
on
di
s
e
a
s
e
i
n
w
om
e
n t
hr
oughout
t
he
w
o
r
l
d w
i
t
h
ne
a
r
l
y
500
,
000
w
om
e
n de
ve
l
opi
ng t
he
di
s
e
a
s
e
e
a
c
h ye
a
r
,
a
nd r
a
nke
d t
he
f
o
ur
t
h m
os
t
c
om
m
una
l
m
a
l
i
gna
nt
di
s
e
a
s
e
w
or
l
dw
i
de
[
3
]
,
[
4
]
.
I
n
t
h
e i
n
i
t
i
al
s
t
ag
es
,
ear
l
y
cer
v
i
cal
can
cer
an
d
p
r
e
-
c
a
nc
e
r
do not
e
xpe
r
i
e
nc
e
s
ym
pt
om
s
,
s
i
nc
e
t
he
y do
not
s
how
s
ym
pt
om
s
unt
i
l
t
he
t
u
m
or
i
s
f
o
r
m
e
d.
M
os
t
c
a
s
e
s
a
r
e
r
e
c
or
de
d
i
n
l
e
s
s
d
e
ve
l
ope
d
c
ount
r
i
e
s
w
ith
u
n
a
v
a
ila
b
ility
o
f
e
f
f
e
c
tiv
e
s
c
r
e
e
n
in
g
s
y
s
te
ms
[
4
]
.
A
l
m
o
s
t
al
l
cas
es
ar
e cau
s
ed
b
y
hu
m
a
n pa
pi
l
l
om
a
vi
r
us
(
H
P
V
)
a
nd
t
he
r
i
s
k
f
a
c
t
or
s
i
nc
l
ude
e
xpos
ur
e
t
o s
m
oki
ng,
a
n
d
immu
n
e
-
s
ys
t
e
m
dys
f
unc
t
i
on [
4
]
.
T
he
r
e
a
r
e
m
o
r
e
t
ha
n
one
hundr
e
d
t
ype
s
o
f
H
P
V
;
how
e
ve
r
,
o
ne
of
a
bout
15
ge
not
ype
s
of
c
a
r
c
i
noge
ni
c
H
P
V
i
s
ve
r
y c
om
m
on
a
m
ong young
w
om
e
n i
n
t
he
i
r
f
i
r
s
t
s
e
xua
l
a
c
t
i
vi
t
y [
5]
.
B
e
s
i
de
s
c
a
r
c
i
noge
ni
c
r
i
s
ks
t
ha
t
a
r
e
l
i
nke
d t
o e
vol
ut
i
ona
r
y s
pe
c
i
e
s
,
e
a
c
h ge
not
ype
a
c
t
s
a
s
a
n
i
nde
pe
nde
nt
i
nf
e
c
t
i
on [
5
]
.
I
n w
om
e
n's
bodi
e
s
,
t
h
i
s
vi
r
us
pr
oduc
e
s
2 t
ype
s
o
f
pr
ot
e
i
ns
,
na
m
e
l
y E
6
a
nd
E
7.
B
ot
h of
t
he
m
a
r
e
da
nge
r
ous
,
s
i
nc
e
t
he
y de
a
c
t
i
va
t
e
c
e
r
t
a
i
n
ge
ne
s
t
h
a
t
pl
a
y c
r
uc
i
a
l
r
ol
e
i
n
s
t
oppi
ng t
um
o
r
de
ve
l
opm
e
nt
.
T
he
s
e
t
w
o pr
ot
e
i
ns
a
l
s
o
a
ggr
e
s
s
i
ve
l
y t
r
i
gge
r
t
he
gr
ow
t
h
of
ut
e
r
i
ne
c
e
l
l
w
a
l
l
.
T
hi
s
unna
t
ur
a
l
c
e
l
l
g
r
ow
t
h e
v
e
nt
ua
l
l
y
cau
s
es
g
en
e ch
an
g
e
s
o
r
g
en
e m
u
t
at
i
o
n
s
,
w
h
i
ch
t
h
e
n
b
eco
m
e t
h
e cau
s
e o
f
cer
v
i
cal
can
cer
t
h
at
d
ev
el
o
p
s
i
n
t
h
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SSN
:
1693
-
6930
T
E
L
KOM
NI
KA
T
el
eco
m
m
u
n
C
om
put
E
l
C
ont
r
o
l
,
Vo
l
.
19
, N
o
.
5
,
O
c
t
obe
r
2021:
1605
-
161
1
1606
body.
T
he
s
ym
pt
om
s
t
ha
t
c
ha
r
a
c
t
e
r
i
z
e
d t
he
di
s
e
a
s
e
a
r
e
a
s
f
ol
l
ow
s
,
unus
ua
l
bl
e
e
di
ng f
r
om
t
he
va
gi
na
,
i
r
r
eg
u
l
ar
m
en
s
t
r
ua
l
c
yc
l
e
s
,
pa
i
n i
n t
he
hi
p,
l
ow
ba
c
k pa
i
n,
body w
e
a
kne
s
s
a
nd t
i
r
e
dne
s
s
,
w
e
i
ght
l
os
s
w
he
n not
on a
di
e
t
,
l
os
s
of
a
ppe
t
i
t
e
,
a
bnor
m
a
l
va
gi
na
l
f
l
ui
d,
a
nd l
e
g i
nf
l
a
m
m
a
t
i
on
.
C
er
v
i
cal
can
cer
s
ar
e m
o
s
t
l
y
as
s
o
ci
at
ed
w
i
t
h
t
h
e l
o
w
an
d
m
i
d
d
l
e
-
in
c
o
me
c
o
u
n
tr
ie
s
w
ith
a
ppr
oxi
m
a
t
e
l
y 90%
H
P
V
va
c
c
i
na
t
i
on pr
og
r
a
m
s
a
nd uor
ga
ni
z
e
d s
c
r
e
e
ni
ng
[
6]
.
T
he
t
r
e
a
t
m
e
nt
de
pe
n
ds
on t
he
l
ev
el
o
f
t
h
e d
i
s
eas
e w
i
t
h
r
es
p
ect
t
o
t
h
e av
ai
l
ab
l
e r
es
o
u
r
ces
an
d
d
i
ag
n
o
s
i
s
m
ad
e i
n
t
h
e ear
l
y
-
st
a
g
e
s.
F
e
r
tility
-
pr
e
s
e
r
vi
ng s
ur
gi
c
a
l
p
r
oc
e
dur
e
s
ha
ve
be
e
n
t
h
e car
e
s
t
an
d
ar
d
f
o
r
w
o
m
en
w
i
t
h
l
o
w
-
r
i
s
k
.
T
h
e
o
v
er
al
l
pr
ognos
i
s
r
e
m
a
i
ns
poor
f
o
r
w
om
e
n
w
i
t
h
m
e
t
a
s
t
a
t
i
c
or
r
e
c
ur
r
e
nt
di
s
e
a
s
e
.
Y
e
t
,
t
he
pe
r
i
od
of
s
ur
vi
va
l
i
s
l
e
s
s
t
ha
n 12
m
ont
hs
,
how
e
ve
r
,
t
he
i
nc
or
por
a
t
i
on
of
t
he
an
t
i
–
va
s
c
ul
a
r
e
ndot
he
l
i
a
l
gr
ow
t
h
f
a
c
t
or
(
VE
GF
)
a
ge
nt
ha
s
be
e
n a
bl
e
t
o e
xt
e
nd i
t
.
2.
R
ES
EA
R
C
H
M
ETH
O
D
2
.1
.
C
on
vol
u
t
i
on
al
n
eu
ra
l
n
e
t
w
o
rk
C
onvol
ut
i
ona
l
ne
ur
a
l
ne
t
w
or
k
(C
N
N
) i
s
a
t
y
p
e
o
f
de
e
p ne
ur
a
l
ne
t
w
or
ks
as
a r
es
u
l
t
o
f
t
h
e
mu
ltila
y
e
r
pe
r
c
e
pt
r
on (
M
L
P
)
[
7
]
,
[
8]
.
T
he
d
i
f
f
e
r
e
nc
e
be
t
w
e
e
n C
N
N
s
a
nd M
L
P
i
s
t
he
i
r
a
bi
l
i
t
y
o
f
be
i
ng
us
e
d i
n
t
he
de
t
e
c
t
i
on a
nd r
e
c
ogni
t
i
on of
obj
e
c
t
s
i
n i
m
a
ge
f
or
m
s
.
C
N
N
s
gi
ve
s
be
t
t
e
r
r
e
s
ul
t
s
t
ha
n ne
ur
a
l
ne
t
w
or
ks
(
N
N
s
)
,
due
t
o t
he
a
ddi
t
i
on
of
one
l
a
ye
r
t
o
C
N
N
s
,
w
hi
c
h
i
s
know
n a
s
t
he
c
onvol
u
t
i
ona
l
l
a
ye
r
a
nd
c
ons
i
s
t
of
ne
ur
ons
w
i
t
h a
c
t
i
va
t
i
on
f
unc
t
i
ons
,
bi
a
s
,
a
nd
w
e
ig
h
t
[
7
]
.
C
N
N
s
is
c
la
s
s
if
ie
d
in
to
tw
o
imp
o
r
ta
n
t
p
a
r
ts
w
h
ic
h
a
r
e
,
f
eat
u
r
e
e
xt
r
a
c
t
i
on a
nd f
ul
l
y
-
co
n
n
ect
ed
l
ay
er
[
8
]
,
[
9
]
.
I
l
l
us
t
r
a
t
i
on of
C
N
N
s
i
s
s
how
n i
n
F
i
gur
e
1 [
1
0]
.
2.
1.
1.
F
ea
t
u
re
ex
t
ra
ct
i
o
n
l
a
y
er
F
eat
u
r
e ex
t
r
act
i
o
n
l
ay
er
"
en
co
d
es
"
an
i
m
ag
e i
n
t
h
e f
o
r
m
o
f
t
h
e o
b
j
ect
r
ep
r
es
en
t
ed
(
f
eat
u
r
e
ex
t
r
act
i
o
n
)
[
7
]
.
H
en
ce,
C
N
N
s
i
s
t
ech
n
i
cal
l
y
an
a
r
ch
i
t
ect
u
r
e en
co
m
p
as
s
i
n
g
s
ev
er
al
s
t
ag
es
,
an
d
each
i
n
p
u
t
an
d
out
put
pr
oc
e
s
s
,
f
e
a
t
ur
e
s
m
a
ps
a
nd num
e
r
ous
a
r
r
a
y
s
,
w
hi
l
e
t
he
e
xt
r
a
c
t
i
on l
a
ye
r
i
ndi
vi
dua
l
l
y c
om
pr
i
s
e
s
of
t
w
o
pa
r
t
s
,
a
s
f
ol
l
ow
s
[
11
]
-
[
13]
.
−
C
onvol
ut
i
ona
l
l
ay
er
C
onvol
ut
e
d
l
ay
er
i
s
t
he
m
a
i
n s
t
r
uc
t
u
r
e
of
a
c
onvol
ut
i
ona
l
ne
ur
a
l
ne
t
w
or
k
(
C
N
N
)
.
T
h
is
la
y
e
r
is
u
tiliz
e
d
in
th
e
tr
a
n
s
f
o
r
ma
ti
on
of
i
nput
s
i
nt
o
a
f
or
m
t
ha
t
i
s
e
a
s
i
l
y
pr
oc
e
s
s
e
d
by
goi
ng
t
hr
ough
a
fi
l
t
e
r
or
ke
r
ne
l
of
a
fi
xe
d
s
i
z
e
w
i
t
hout
l
os
i
ng
e
s
s
e
nt
i
a
l
c
onvul
a
t
e
d
f
e
a
t
ur
e
s
[
14
]
.
I
n
th
is
la
y
e
r
,
th
e
r
e
a
r
e
f
ilte
r
s
(
k
e
r
n
e
ls
)
th
a
t
s
pr
e
a
d t
o
t
he
e
nt
i
r
e
i
npu
t
,
a
nd
e
a
c
h uni
t
r
e
c
e
i
ve
s
i
nput
f
r
om
t
he
pr
e
vi
ous
l
a
ye
r
.
T
he
r
e
f
or
e
,
t
hr
ough
c
onvol
ut
i
on,
t
he
i
nput
m
a
p
i
s
ge
ne
r
a
t
e
d
be
t
w
e
e
n
e
a
c
h
fi
l
t
e
r
,
t
he
n
s
hi
f
t
i
ng
t
he
i
nput
a
nd
us
i
ng
t
he
s
um
of
dot
pr
oduc
t
s
.
−
P
ool
i
ng
P
ool
i
ng i
s
a
t
e
c
hni
que
f
or
r
e
duc
i
ng d
i
m
e
ns
i
ons
w
i
t
h t
he
a
i
d
of
t
w
o
c
om
m
on a
pp
r
oa
c
he
s
na
m
e
l
y t
he
a
ve
r
a
ge
a
nd m
a
xi
m
um
poo
l
i
ng
[
15]
.
T
hi
s
ope
r
a
t
i
o
n i
s
c
a
l
l
e
d t
he
m
a
x
pool
i
ng
w
he
n
i
t
us
e
s
t
he
hi
ghe
s
t
va
l
ue
,
w
h
i
l
e t
h
e av
er
ag
e p
o
o
l
i
n
g
u
s
es
t
h
e m
ed
i
al
v
al
u
e.
A
f
t
er
t
h
i
s
,
t
h
e f
l
at
t
en
i
n
g
p
r
o
ces
s
t
ak
es
p
l
ace,
w
h
i
ch
i
s
t
h
e
r
e
s
ha
pi
ng of
a
pool
e
d
s
t
r
uc
t
ur
e
i
n
t
o a
one
–
di
m
e
ns
i
ona
l
ve
c
t
or
,
t
he
n
pl
a
c
e
d i
nt
o
f
ul
l
y
-
c
onne
c
t
e
d ne
ur
a
l
ne
t
w
or
ks
or
M
L
P
f
or
c
l
a
s
s
i
fi
c
a
t
i
on
[
14]
.
−
M
LP
l
ay
er
s
M
LP
l
ay
er
s
i
s
a
f
u
l
l
y
co
n
n
ect
ed
mu
lti
–
l
a
ye
r
pe
r
c
e
pt
r
on
t
ha
t
pe
r
f
o
r
m
s
t
he
c
l
a
s
s
i
fi
c
a
t
i
on
ope
r
a
t
i
on.
T
h
er
e ar
e
t
h
r
ee
l
ay
er
s
i
n
M
L
P
n
a
m
e
l
y
,
t
he
hi
dde
n l
a
ye
r
s
,
i
nput
a
nd
out
put
l
a
ye
r
s
.
T
he
a
c
t
i
va
t
i
on
f
unc
t
i
on
u
se
s t
h
e
r
ect
i
fi
ed
l
i
n
ear
u
n
i
t
(
R
e
L
U
)
,
w
hi
c
h
i
s
qui
t
e
popul
a
r
i
n
de
e
p l
e
a
r
ni
ng due
t
o
i
t
s
s
i
m
pl
i
c
i
t
y.
2.
1.
2.
F
u
lly
-
co
n
n
ect
ed
l
a
y
er
F
u
lly
-
co
n
n
ect
ed
l
ay
er
f
unc
t
i
ons
ba
s
e
d on t
he
f
eat
u
r
e ex
t
r
act
i
o
n
l
ay
er
,
w
h
ic
h
is
a
mu
ltid
i
me
n
s
io
n
a
l
ar
r
ay
,
w
i
t
h
f
l
at
t
en
(
r
es
h
ap
e)
i
n
t
h
e
v
ect
o
r
f
eat
u
r
e
m
ap
[
1
6
]
,
[
17
]
.
I
n
a
ddi
t
i
on
,
a
l
l
a
c
t
i
ve
ne
ur
ons
f
r
o
m
pr
e
vi
ous
l
a
ye
r
a
r
e
l
i
nke
d
w
i
t
h
t
he
ne
xt
l
a
ye
r
a
s
i
n
ne
ur
a
l
ne
t
w
or
ks
.
T
he
r
e
f
or
e
,
i
n o
r
de
r
t
o
c
onne
c
t
pr
ope
r
l
y,
i
n
di
vi
dua
l
a
c
t
i
va
t
i
on (
of
t
he
p
r
e
vi
ou
s
)
ought
t
o be
c
onve
r
t
e
d
i
nt
o 1
-
D
d
at
a.
T
h
es
e u
s
u
al
l
y
u
s
e M
L
P
t
er
m
,
w
h
i
ch
p
r
o
ces
s
da
t
a
w
i
t
h pr
ope
r
c
l
a
s
s
i
f
i
c
a
t
i
on [
18]
.
M
e
a
nw
hi
l
e
,
t
he
c
ont
r
a
s
t
a
ga
i
ns
t
c
onvol
ut
i
on l
a
ye
r
s
a
r
e
t
he
ne
ur
ons
,
w
h
i
ch
ar
e
co
n
n
ect
ed
t
o
a
s
p
eci
f
i
c
i
n
p
u
t
ar
ea,
w
h
i
l
e f
u
l
l
y
-
c
onne
c
t
e
d oc
c
ur
s
i
n
a
l
m
os
t
a
l
l
pa
r
t
s
.
H
ow
e
ve
r
,
bot
h
c
ont
i
nue
t
o pe
r
f
o
r
m
“
dot
pr
oduc
t
”
ope
r
a
t
i
ons
;
t
he
r
e
f
or
e
,
t
he
i
r
f
unc
t
i
ons
a
r
e
not
s
i
gni
f
i
c
a
nt
l
y di
f
f
e
r
e
nt
.
2.
2.
Su
ppo
r
t
v
ect
o
r
m
a
ch
i
n
e
S
uppo
r
t
v
ect
o
r
m
ach
i
n
e
(
S
V
M
)
h
as
r
ecei
v
ed
m
u
ch
at
t
en
t
i
o
n
i
n
t
h
e
cl
as
s
i
f
i
cat
i
o
n
as
p
ect
[
1
9
]
.
T
h
e
m
a
i
n f
i
e
l
d of
t
h
i
s
s
t
udy i
s
us
e
d t
o de
ve
l
op S
V
M
a
l
gor
i
t
hm
ba
s
e
d on t
he
s
t
a
t
i
s
t
i
c
a
l
l
e
a
r
ni
ng t
he
or
y [
2
0]
.
S
V
M
i
s
a
l
s
o know
n a
s
one
of
t
he
e
f
f
e
c
t
i
ve
m
a
c
hi
ne
l
e
a
r
ni
ng a
nd ha
s
hi
gh c
l
a
s
s
i
f
i
c
a
t
i
on e
f
f
i
c
i
e
nc
y [
20
]
.
I
l
l
us
t
r
a
t
i
on
of
S
V
M
s
i
s
s
how
n i
n
F
i
gur
e
2
[
21]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
KA
T
el
eco
m
m
u
n
C
om
put
E
l
C
ont
r
o
l
C
er
vi
ca
l
ca
n
ce
r
c
l
as
s
i
f
i
c
at
i
on us
i
ng c
onv
ol
ut
i
onal
ne
ur
al
ne
t
w
or
k
–
s
uppor
t
v
e
c
t
or
… (
J
ane
E
v
a A
ur
e
l
i
a
)
1607
F
i
gur
e
1.
I
l
l
us
t
r
a
t
i
on of
c
onvol
ut
i
ona
l
ne
u
r
a
l
ne
t
w
or
k
F
i
gur
e
2.
I
l
l
us
t
r
a
t
i
on of
s
uppor
t
ve
c
t
or
m
ach
i
n
e
S
V
M
i
s
a
m
a
c
hi
ne
l
e
a
r
ni
ng
a
l
gor
i
t
h
m
f
or
c
l
a
s
s
i
f
i
c
a
t
i
on a
nd
r
e
gr
e
s
s
i
on,
w
hi
c
h
w
a
s
i
nt
r
oduc
e
d
by
V
a
pni
k (
1990)
[
22
]
.
T
he
n,
N
e
l
l
o
C
r
i
s
t
i
a
ni
ni
r
e
s
e
a
r
c
he
d a
bout
S
V
M
ba
s
e
d on V
a
pni
k r
e
s
ul
t
s
[
23]
.
S
ubs
e
que
nt
l
y,
B
e
r
nha
r
d S
c
hol
kop
f
de
ve
l
ope
d
S
V
M
t
he
or
y a
nd ke
r
ne
l
f
unc
t
i
on [
24
]
.
I
n
a
ddi
t
i
on
,
S
V
M
i
s
a
n
in
itia
l f
o
r
m
f
o
r
b
in
a
r
y
c
la
s
s
if
ic
a
tio
n
; h
o
w
e
v
e
r
,
it
is
a
ls
o
f
o
r
mu
lt
ic
la
s
s
c
a
te
g
o
r
iz
a
tio
n
.
S
V
M
doe
s
m
a
ppi
ng f
or
m
s
a
h
i
ghe
r
di
m
e
ns
i
ona
l
s
pa
c
e
f
or
s
uppor
t
i
ng nonl
i
ne
a
r
c
l
a
s
s
i
f
i
c
a
t
i
on,
a
nd
co
n
s
t
r
u
ct
i
n
g
t
h
e m
ax
i
m
al
s
ep
ar
at
i
ng hype
r
pl
a
ne
.
F
or
i
ns
t
a
nc
e
,
t
he
r
e
i
s
a
s
e
t
of
f
i
r
m
s
r
e
pr
e
s
e
nt
e
d by t
he
va
l
ue
o
f
th
e
ir
r
a
tio
s
{
}
,
=
1
,
…
,
an
d
a s
et
o
f
as
s
o
ci
at
ed
l
ab
el
s
∈
{
−
1
,
+
1
}
w
hi
c
h de
s
c
r
i
be
s
r
e
s
ul
t
s
a
s
f
a
i
l
e
d or
he
a
l
t
hy.
T
he
m
a
i
n pu
r
pos
e
of
S
V
M
i
s
t
o
f
i
nd t
he
be
s
t
hype
r
pl
a
ne
th
a
t is
w
r
itte
n
a
s
;
∙
+
=
0
(
1)
T
h
e (
1
)
ab
o
v
e i
s
ab
l
e
t
o
m
ax
i
m
i
ze
t
h
e m
ar
g
i
n
.
T
h
e
o
p
timiz
a
tio
n
p
r
o
b
le
m o
f
S
V
M
is
s
u
mma
r
iz
e
d
a
s
f
o
llo
w
;
M
in
imiz
e
1
2
|
|
|
|
2
(
2)
S
u
b
je
c
t to
;
(
∙
+
)
≥
1
,
∀
=
1
,
…
,
(
3)
T
he
(
2)
f
i
nds
∈
a
nd
∈
w
i
t
h c
ons
t
r
a
i
ns
t
o (
3)
,
a
l
ong
(
w
e
i
ght
s
)
a
nd
(
b
ia
s
)
.
P
r
o
b
le
m in
(
2
)
is
q
u
a
d
r
a
tic
o
p
timiz
a
tio
n
.
T
he
r
e
f
or
e
,
t
he
L
a
gr
a
nge
m
ul
t
i
pl
i
e
r
s
f
or
e
a
c
h of
t
he
c
ons
t
r
a
i
nt
s
i
n
(
2)
i
s
s
how
n by
gi
vi
ng
t
he
f
unc
t
i
on a
s
;
(
,
,
)
=
1
2
|
|
|
|
2
−
∑
{
(
∙
+
)
−
1
}
=
1
(
4)
w
h
er
e
=
(
1
,
2
,
…
,
)
.
Wh
en
a
nd b e
qua
l
t
o
z
e
r
o,
s
e
t
t
i
ng t
he
de
r
i
va
t
i
ve
s
of
(
,
,
)
t
he
e
qua
t
i
ons
obt
a
i
ne
d a
r
e
,
=
−
∑
=
1
=
0
→
=
∑
=
1
(
5)
=
∑
=
1
=
0
→
∑
=
1
=
0
(
6)
T
he
n,
e
l
i
m
i
na
t
i
ng
a
nd b
f
r
om
(
,
,
)
us
i
ng (
5
)
a
nd
(
6)
,
obt
a
i
ne
d t
he
dua
l
f
or
m
a
s
;
(
)
=
m
ax
{
−
1
2
∑
∑
,
=
1
=
1
+
∑
=
1
}
(
7)
∑
=
1
=
0
,
≥
0
(
8)
F
r
om
(
1)
w
hi
c
h i
s
(
)
=
∙
+
,
t
h
e
a
nd
of
r
e
gr
e
s
s
i
on f
unc
t
i
on i
s
f
i
na
l
l
y obt
a
i
ne
d a
s
f
ol
l
ow
s
;
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SSN
:
1693
-
6930
T
E
L
KOM
NI
KA
T
el
eco
m
m
u
n
C
om
put
E
l
C
ont
r
o
l
,
Vo
l
.
19
, N
o
.
5
,
O
c
t
obe
r
2021:
1605
-
161
1
1608
=
∑
=
1
(
9)
=
1
∑
(
−
∑
∈
)
∈
(
10)
I
n t
hi
s
s
t
udy,
l
i
ne
a
r
ke
r
ne
l
s
a
r
e
us
e
d f
or
s
uppor
t
ve
c
t
or
m
a
c
hi
ne
s
(
S
V
M
)
[
25]
.
K
er
n
e
l
f
unc
t
i
on
r
e
s
ol
ve
s
l
i
ne
a
r
di
m
e
ns
i
on pr
obl
e
m
s
a
nd a
l
s
o
f
or
a
l
gor
i
t
hm
s
e
xpr
e
s
s
i
on i
n t
he
i
nne
r
pr
oduc
t
be
t
w
e
e
n t
w
o
ve
c
t
or
s
[
25]
.
T
he
r
e
a
r
e
s
e
ve
r
a
l
ke
r
ne
l
f
unc
t
i
ons
w
i
t
h t
he
i
r
pa
r
a
m
e
t
er
s
i
n
T
ab
l
e
1
.
T
ab
l
e
1
.
K
e
r
ne
l
f
unc
t
i
on
N
a
me
K
e
r
ne
l
F
unc
t
i
on
L
in
ie
r
,
=
⟦
⟧
P
ol
ynom
i
a
l
,
=
+
⟦
⟧
G
a
us
s
i
a
n
R
a
d
ia
l
B
a
si
s
F
unc
t
i
on (
R
B
F
)
,
=
e
xp
(
−
−
2
/
2
)
2.
3.
C
o
nf
us
i
o
n m
a
t
r
i
x
A
ccu
r
acy
i
s
o
n
e o
f
m
ai
n
p
a
r
am
et
er
t
h
at
u
s
ed
t
o
o
b
s
er
v
e a
cl
as
s
i
f
i
cat
i
o
n
’
s
s
u
cces
s
.
R
ef
er
s
t
o
t
h
e
p
er
cen
t
ag
e o
f
co
r
r
ect
an
s
w
er
s
at
t
es
t
i
n
g
s
t
ag
e,
co
n
f
u
s
i
o
n
m
at
r
i
x
u
s
ed
t
o
m
eas
u
r
es
t
h
e accu
r
acy
.
T
h
e
c
onf
us
i
on m
a
t
r
i
x us
e
d
i
s
s
how
n i
n
T
a
bl
e
2
[2
5
].
T
he
f
or
m
ul
a
of
accu
r
acy
i
s
w
r
i
t
t
en
as
:
=
+
+
+
+
(
11)
T
p
:
N
um
be
r
o
f
s
a
m
pl
e
s
ha
vi
ng
cer
v
i
cal
can
cer
an
d
cl
as
s
i
f
i
ed
co
r
r
ect
l
y
.
F
P
:
N
um
be
r
o
f
he
a
l
t
hy
i
ndi
vi
dua
l
s
t
ha
t
a
r
e
i
nc
or
r
e
c
t
l
y c
l
a
s
s
i
f
i
e
d t
o
cer
v
i
cal
can
cer
.
F
N
: N
u
mb
e
r
o
f
s
a
mp
le
s
w
ith
cer
v
i
cal
can
cer
t
h
at
ar
e
i
n
co
r
r
ect
l
y
cl
as
s
i
f
i
ed
as
h
eal
t
h
y
.
T
N
:
N
um
be
r
o
f
he
a
l
t
hy
i
ndi
vi
dua
l
s
c
or
r
e
c
t
l
y
s
pot
t
e
d.
T
ab
l
e 2
.
C
onf
us
i
on m
a
t
r
i
x
A
ct
u
al
P
r
e
d
ic
tio
n
P
o
s
itiv
e
N
e
g
a
tiv
e
P
o
s
itiv
e
T
p
F
P
N
e
g
a
tiv
e
F
N
T
N
3.
R
ES
U
LTS
A
ND ANAL
YS
I
S
3.
1.
Da
t
a
T
h
i
s
p
ap
er
r
ecei
v
ed
d
at
ab
as
e o
f
cer
v
i
cal
can
cer
s
uf
f
e
r
e
r
s
,
w
hi
c
h c
ons
i
s
t
e
d of
652
i
nf
or
m
a
t
i
on
s
w
ith
a
c
t
ua
l
a
m
ount
s
of
607 m
a
j
or
a
nd
45 m
i
nor
da
t
a
.
T
he
m
i
nor
r
e
pr
e
s
e
nt
e
d t
he
c
l
a
s
s
e
s
t
ha
t
i
ndi
c
a
t
e
d t
he
pr
e
s
e
nc
e
o
f
cer
v
i
cal
can
cer
w
i
t
h
l
ab
el
‘
1
’
,
w
h
i
l
e t
h
e
m
aj
o
r
r
ep
r
es
en
t
ed
t
h
e cl
as
s
e
s
t
h
at
d
o
n
o
t
i
n
d
i
cat
e t
h
e p
r
es
en
ce o
f
cer
v
i
cal
can
cer
w
i
t
h
l
ab
el
'
0
'.
T
he
r
e
w
e
r
e
25 f
e
a
t
ur
e
s
us
e
d i
n t
hi
s
s
t
udy,
na
m
e
l
y a
ge
,
num
be
r
of
s
e
xua
l
pa
r
t
ne
r
s
,
f
i
r
s
t
s
e
xua
l
i
nt
e
r
c
our
s
e
,
num
be
r
of
pr
e
gna
nc
i
e
s
,
s
m
oke
s
(
ye
a
r
s
,
pa
c
ks
/
ye
a
r
)
,
h
or
m
ona
l
co
n
t
r
acep
t
i
v
es
(
y
ear
s
)
,
in
tr
a
ut
e
r
i
ne
de
vi
c
e
(y
e
a
rs
),
s
ex
u
al
l
y
t
r
an
s
m
i
t
t
ed
d
i
s
eas
es
(
S
TD
)
(
num
be
r
,
c
ondyl
om
a
t
os
i
s
,
vul
vo
-
pe
r
i
ne
a
l
c
ondyl
om
a
t
os
i
s
,
s
yphi
l
i
s
,
hum
a
n
i
m
m
unode
f
i
c
i
e
nc
y vi
r
us
(
HI
V
)
,
nu
m
b
e
r
of
di
a
gnos
i
s
)
,
d
i
a
gnos
i
s
(
can
cer
,
hum
a
n pa
pi
l
l
om
a
vi
r
us
(
HP
V
)
)
,
hi
ns
e
l
m
a
nn,
s
c
hi
l
l
e
r
,
a
nd
c
i
t
ol
ogy.
3.
2.
R
e
su
l
t
s
F
o
r
th
e
c
la
s
s
if
ic
a
tio
n
me
th
o
d
,
t
hi
s
r
e
s
e
a
r
c
h us
e
d 20%
da
t
a
f
o
r
t
r
a
i
ni
ng a
nd
80%
da
t
a
f
or
t
e
s
t
i
ng.
I
n
t
hi
s
s
t
udy,
1
,
000
a
m
ount
of
e
poc
hs
w
e
r
e
us
e
d f
or
t
he
c
onvol
ut
i
ona
l
ne
ur
a
l
ne
t
w
or
k w
i
t
h
t
he
c
om
bi
n
a
t
i
on of
s
e
ve
r
a
l
ke
r
ne
l
f
unc
t
i
ons
us
e
d f
or
t
he
s
uppor
t
ve
c
t
or
m
a
c
hi
ne
.
T
he
r
e
s
ul
t
s
w
e
r
e
s
how
n i
n
F
i
gur
e
s
3,
4,
a
nd 5.
I
n F
i
gur
e
3
(
a)
,
t
h
er
e w
as
a r
i
s
e i
n
t
h
e accu
r
acy
l
ev
el
o
f
t
h
e m
o
d
el
as
m
an
y
ep
o
ch
s
i
n
cr
eas
e.
T
h
e
b
l
u
e l
i
n
e
w
hi
c
h s
t
a
nds
f
or
t
r
a
i
ni
ng da
t
a
ga
ve
hi
ghe
r
a
c
c
ur
a
c
y of
100%
,
w
hi
l
e
t
he
o
r
a
nge
l
i
ne
f
o
r
t
e
s
t
i
ng da
t
a
g
av
e an
a
c
c
ur
a
c
y va
l
ue
of
93.
67
%
.
F
i
gu
r
e
3
(
b)
s
how
e
d t
ha
t
t
he
num
be
r
of
l
os
s
(
e
r
r
or
)
de
c
r
e
a
s
e
s
a
s
t
he
n
um
be
r
of
e
poc
hs
de
c
r
e
a
s
e
.
T
he
e
r
r
or
f
ound
on t
r
a
i
ni
ng da
t
a
w
a
s
0,
w
hi
l
e
e
r
r
or
on t
e
s
t
da
t
a
w
a
s
0
.
06.
F
i
gur
e
4
(
a)
,
s
h
o
w
ed
t
h
at
t
h
e accu
r
acy
o
f
t
h
e
m
o
d
el
i
n
cr
eas
es
as
m
an
y
ep
o
ch
s
i
n
cr
eas
e,
s
i
n
ce
t
he
bl
ue
l
i
ne
(
t
r
a
i
ni
ng
da
t
a
)
ga
ve
hi
ghe
r
a
c
c
ur
a
c
y t
ha
n
t
he
o
r
a
nge
l
i
ne
(
t
e
s
t
i
ng da
t
a
)
.
T
he
a
c
c
ur
a
c
y of
t
he
t
r
a
i
ni
ng da
t
a
w
a
s
100%
,
w
hi
l
e
f
o
r
t
e
s
t
i
ng da
t
a
i
t
w
a
s
92.
72%
.
F
i
gu
r
e
4
(
b
)
s
how
e
d t
ha
t
t
he
num
be
r
of
l
o
ss
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
KA
T
el
eco
m
m
u
n
C
om
put
E
l
C
ont
r
o
l
C
er
vi
ca
l
ca
n
ce
r
c
l
as
s
i
f
i
c
at
i
on us
i
ng c
onv
ol
ut
i
onal
ne
ur
al
ne
t
w
or
k
–
s
uppor
t
v
e
c
t
or
… (
J
ane
E
v
a A
ur
e
l
i
a
)
1609
(
er
r
o
r
)
d
ecr
eas
es
as
t
h
e n
u
m
b
er
o
f
ep
o
ch
s
d
ecr
eas
e.
T
h
e
er
r
o
r
i
n
t
r
ai
n
i
n
g
d
at
a w
as
0
,
w
h
i
l
e i
n
t
es
t
i
n
g
d
at
a i
t
w
a
s
0.
07.
F
i
gur
e
5
(
a)
s
h
o
w
ed
t
h
at
t
h
e accu
r
acy
o
f
t
h
e m
o
d
el
i
n
cr
eas
es
a
s
m
an
y
ep
o
ch
s
i
n
cr
eas
e.
T
h
en
,
t
h
e
t
r
a
i
ni
ng da
t
a
(
bl
ue
l
i
ne
)
ga
ve
hi
ghe
r
a
c
c
ur
a
c
y t
ha
n t
he
t
e
s
t
i
ng da
t
a
(
or
a
nge
l
i
ne
)
.
T
he
a
c
c
ur
a
c
y
of
t
r
a
i
ni
ng
da
t
a
w
a
s
100%
,
w
hi
l
e
t
he
t
e
s
t
i
ng da
t
a
w
a
s
92.
91%
.
B
e
s
i
de
s
t
hi
s
,
F
i
gur
e
5
(
b)
s
how
e
d t
ha
t
t
he
num
be
r
of
l
os
s
(
er
r
o
r
)
d
ecr
eas
es
a
s
t
h
e n
u
m
b
er
o
f
ep
o
ch
d
ecr
ea
s
e,
w
h
i
ch
w
as
0
f
o
r
t
r
a
i
ni
ng da
t
a
a
nd 0.
07 on t
e
s
t
da
t
a
.
T
ab
l
e 3
s
h
o
w
ed
t
h
e accu
r
acy
o
f
each
m
et
h
o
d
.
T
he
c
om
pa
r
i
s
on of
c
onvol
ut
i
ona
l
ne
ur
a
l
ne
t
w
or
k
–
s
uppor
t
ve
c
t
or
m
a
c
hi
ne
w
i
t
h s
om
e
ke
r
ne
l
s
f
or
t
he
c
la
s
s
if
ic
a
tio
n
o
f
cer
v
i
cal
can
cer
,
w
er
e f
o
u
n
d
t
o
p
r
o
p
er
l
y
an
d
co
r
r
ect
l
y
p
r
ed
i
ct
da
t
a
.
R
e
s
ul
t
s
how
e
d t
ha
t
t
he
c
onvol
ut
i
ona
l
ne
ur
a
l
ne
t
w
or
k
–
s
uppor
t
ve
c
t
or
m
a
c
hi
ne
w
i
t
h l
i
ne
a
r
ke
r
ne
l
ha
d
t
he
be
s
t
a
c
c
ur
a
c
y va
l
ue
of
93.
67%
on t
he
t
e
s
t
da
t
a
.
W
hi
l
e
on t
he
t
r
a
i
ni
ng da
t
a
,
a
l
l
m
e
t
hods
ga
ve
be
s
t
a
c
c
ur
a
c
y f
or
da
t
a
ba
s
e
cat
eg
o
r
i
zat
i
o
n
.
T
h
er
ef
o
r
e,
t
he
be
s
t
m
e
t
hod
f
or
t
he
c
l
a
s
s
i
f
i
c
a
t
i
on of
c
e
r
vi
c
a
l
c
a
nc
e
r
i
s
t
he
c
onvol
ut
i
ona
l
ne
ur
a
l
ne
t
w
or
k
–
s
uppor
t
ve
c
t
or
m
a
c
hi
ne
w
i
t
h
l
i
ne
a
r
ke
r
ne
l
.
(
a)
(
b)
F
i
gur
e
3.
C
onvol
ut
i
ona
l
ne
u
r
a
l
ne
t
w
or
k
–
s
uppor
t
v
e
c
t
or
m
a
c
hi
ne
w
i
t
h l
i
ne
a
r
ke
r
ne
l
;
(
a
)
m
o
d
el
accu
r
a
cy
of
CN
N
–
S
V
M
l
i
n
ear
k
er
n
el
,
a
nd
(
b)
m
ode
l
l
os
s
o
f
CN
N
–
S
V
M
l
i
n
ear
k
er
n
el
(
a)
(
b)
F
i
gur
e
4.
C
onvol
ut
i
ona
l
ne
u
r
a
l
ne
t
w
or
k
–
s
uppor
t
v
e
c
t
or
m
a
c
hi
ne
w
i
t
h pol
yno
m
i
a
l
ke
r
ne
l
;
(
a
)
m
ode
l
a
c
c
ur
a
c
y
o
f
CN
N
–
S
V
M
pol
ynom
i
a
l
ke
r
ne
l
a
nd (
b
)
m
ode
l
l
o
s
s
of
C
N
N
–
S
V
M
pol
ynom
i
a
l
ke
r
ne
l
(
a)
(
b)
F
i
gur
e
5.
C
onvol
ut
i
ona
l
ne
u
r
a
l
ne
t
w
or
k
–
s
uppor
t
v
e
c
t
or
m
a
c
hi
ne
w
i
t
h G
a
us
s
i
a
n R
B
F
ke
r
ne
l
;
(
a
)
m
od
e
l
accu
r
acy
o
f
C
N
N
–
S
V
M
G
a
us
s
i
a
n R
B
F
ke
r
ne
l
,
a
nd (
b)
m
ode
l
l
os
s
of
C
N
N
–
S
V
M
G
au
s
s
i
an
R
B
F
k
er
n
el
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SSN
:
1693
-
6930
T
E
L
KOM
NI
KA
T
el
eco
m
m
u
n
C
om
put
E
l
C
ont
r
o
l
,
Vo
l
.
19
, N
o
.
5
,
O
c
t
obe
r
2021:
1605
-
161
1
1610
T
ab
l
e 3
. T
h
e
accu
r
acy
o
f
each
m
et
h
o
d
No
M
ode
l
T
r
ai
n
i
n
g
A
ccu
r
acy
T
es
t
i
n
g
A
c
cu
r
acy
1.
C
NN
–
S
V
M
L
i
n
ear
100%
93.67%
2.
C
NN
–
S
V
M
P
ol
ynom
i
a
l
100%
92.72%
3.
C
NN
–
S
V
M
RBF
100%
92.91%
4.
CO
NCL
US
I
O
N
P
r
e
di
c
t
i
ng t
he
p
r
e
s
e
nc
e
of
di
s
e
a
s
e
by di
a
gnos
i
ng w
i
t
h m
a
c
hi
ne
l
e
a
r
ni
ng m
e
t
hod
he
l
p m
e
di
c
a
l
s
t
a
f
f
t
o
c
la
s
s
if
y
a
ilme
n
ts
.
A
n
e
a
r
ly
d
e
te
c
tio
n
o
f
d
is
e
a
s
e
is
imp
o
r
ta
n
t,
s
in
c
e
it ma
k
e
s
th
e
p
a
tie
n
t to
r
e
c
e
iv
e
a
p
r
o
mp
t
r
i
g
h
t
t
r
eat
m
en
t
,
w
h
i
ch
h
el
p
s
t
o
i
n
cr
eas
e t
h
e c
ha
n
c
e
of
s
ur
vi
va
l
a
nd r
e
duc
e
t
he
he
a
l
t
h
r
i
s
k.
T
he
r
e
f
or
e
,
t
hi
s
r
e
s
e
a
r
c
h f
oc
us
e
s
on c
e
r
vi
c
a
l
c
a
nc
e
r
w
hi
c
h i
s
a
c
om
m
on he
a
l
t
h pr
obl
e
m
w
i
t
h 652 da
t
a
c
ol
l
e
c
t
e
d a
nd 25
f
e
a
t
ur
e
s
obs
e
r
ve
d.
T
he
m
e
t
hod
us
e
d w
a
s
t
he
c
o
m
bi
na
t
i
on o
f
c
onvol
u
t
i
ona
l
ne
u
r
a
l
ne
t
w
o
r
k
-
s
uppor
t
v
ect
o
r
m
ach
i
n
e
w
i
t
h s
e
ve
r
a
l
ke
r
ne
l
f
unc
t
i
ons
a
s
c
l
a
s
s
i
f
i
e
r
.
T
he
e
xpe
r
i
m
e
nt
a
l
r
e
s
ul
t
s
s
how
e
d t
ha
t
t
he
m
e
t
ho
ds
us
e
d,
pr
ope
r
l
y a
nd
c
or
r
e
c
t
l
y
pr
e
di
c
t
e
d
t
he
da
t
a
.
B
a
s
e
d
on f
i
ndi
ngs
,
c
onvol
ut
i
ona
l
ne
ur
a
l
ne
t
w
or
k
-
s
uppor
t
ve
c
t
or
m
ach
i
n
e
w
ith
lin
e
a
r
k
e
r
n
e
l is
th
e
b
e
s
t mo
d
e
l
f
o
r
th
e
c
la
s
s
if
ic
a
tio
n
o
f
cer
v
i
cal
can
cer
da
t
a
a
s
s
how
n i
n T
a
bl
e
3
.
T
he
r
e
f
or
e
,
i
n
t
he
f
ut
ur
e
r
e
s
e
a
r
c
h,
t
hi
s
m
e
t
hod
de
ve
l
ops
t
o
gi
ve
hi
ghe
r
a
c
c
ur
a
c
y a
nd
us
e
s
a
l
a
r
ge
r
da
t
a
ba
s
e
,
i
n
or
de
r
t
o
gi
ve
be
t
t
e
r
r
e
s
ul
t
s
f
or
pr
e
di
c
t
i
ng
a
nd c
l
a
s
s
i
f
yi
ng di
f
f
e
r
en
t
d
i
s
eas
es
.
A
C
K
N
O
W
LED
G
EM
EN
TS
T
hi
s
r
e
s
e
a
r
c
h w
a
s
s
uppor
t
e
d f
i
na
nc
i
a
l
l
y
by t
he
I
nd
one
s
i
a
n M
i
ni
s
t
r
y of
R
e
s
e
a
r
c
h T
e
c
hnol
ogy;
w
i
t
h
a
K
E
M
E
N
R
I
S
T
E
K
/
B
R
I
N
P
D
U
P
T
2021 r
e
s
e
a
r
c
h gr
a
nt
s
c
he
m
e
,
I
D
num
be
r
N
K
B
-
2830/
U
N
2.
R
S
T
/
H
K
P
.
05
.
00/
2020.
R
EF
ER
EN
C
ES
[1
]
N.
S
a
lm
i a
nd Z
.
R
u
sta
m
,
“
Na
ïve
B
a
ye
s C
la
s
sif
ie
r
M
od
e
ls f
or
P
r
e
dic
ti
ng the
C
o
lo
n C
a
nc
e
r
,
“
201
9
I
OP C
on
f.
S
e
r.
:
Ma
te
r.
Sc
i.
En
g
.
,
v
ol
. 5
46
,
no
. 5
,
20
19
,
do
i:
10
.
1
08
8/
17
57
-
89
9X
/5
46
/5
/0
52
06
8
.
[2
]
A.
A.
R
a
c
hm
a
n a
n
d Z
.
R
usta
m
,
“
C
a
nc
e
r
C
la
ss
if
ic
a
ti
on
usi
ng F
uz
z
y C
-
M
e
a
n
s wi
th
F
e
a
tur
e
S
e
le
c
ti
on,
”
20
16
12
th
I
nte
rn
at
io
na
l C
on
fe
re
nc
e
o
n Ma
the
ma
ti
c
s,
S
ta
ti
st
ic
s
,
and T
he
ir
Ap
pl
ic
a
ti
on
s (
I
C
MS
A)
,
20
16
,
pp.
3
1
-
3
4
,
doi
:
1
0.
11
09
/I
C
M
S
A.
20
16.
7
95
43
02
.
[3
]
R
.
G
e
e
tha
,
S
.
S
iva
s
ubr
a
m
a
nia
n
,
a
nd M
.
Ka
l
ia
p
pa
n,
“
C
e
r
vic
a
l C
a
nc
e
r
I
de
n
tif
ic
a
t
io
n w
it
h S
y
nt
he
t
ic
M
i
no
r
it
y
Ove
r
sa
m
p
li
ng
Te
c
h
ni
que
a
nd
P
C
A
Ana
ly
si
s u
si
ng
R
a
ndom
F
or
e
st C
la
s
sif
ie
r
,
”
2
01
9
J
M
e
d
S
y
s
t
.
,
v
ol
.
43
,
no
.
9
,
pp.
1
-
19
,
do
i
:
10.
10
07
/s1
09
16
-
01
9
-
14
02
-
6
.
[4
]
S
.
E
.
W
a
ggone
r
,
“
C
e
r
v
i
c
a
l
C
a
n
c
e
r
,
”
20
03
T
h
e
L
a
n
c
e
t
,
v
ol
.
361,
no
.
937
6,
pp.
22
17
-
22
25
,
2
0
03,
doi
:
10
.
1
01
6/
S
01
40
-
67
36
(
03
)
1
37
78
-
6
.
[5
]
M
.
Sch
i
ffman
,
P
.
E.
C
a
stle
, J
.
Je
r
o
nim
o
,
A
.
C
.
R
odr
ig
ue
z
,
a
n
d
S
.
W
a
c
hol
de
r
,
“
Hum
a
n P
a
pi
ll
om
a
v
ir
u
s a
n
d C
e
r
vi
c
a
l
C
a
n
c
e
r
,
”
T
he
L
anc
e
t
,
v
ol
. 3
7
0
,
no.
9
59
0,
pp.
8
90
-
9
07
,
2
007
,
do
i
:
10
.
1
01
6/
S
01
40
-
67
36
(
07
)
6
14
16
-
0
.
[6
]
P
.
A.
C
ohe
n
,
A
.
J
hi
ngr
a
n
,
A
.
O
a
kn
in
, a
n
d
L
.
De
nn
y
,
“
C
e
r
v
i
c
a
l
C
a
n
c
e
r
,
”
Th
e
L
a
n
c
e
t
,
v
o
l.
3
93,
no.
10
16
7,
pp.
16
9
-
1
82
,
20
19
,
do
i
:
10.
10
16
/S
01
40
-
6
73
6(
18
)
32
47
0
-
X
.
[7
]
K
.
O
’S
h
ea
a
nd
R
.
N
a
s
h
,
“
An I
n
tr
o
duc
ti
on
to C
o
nv
ol
ut
io
na
l
Ne
ur
a
l Ne
t
wor
ks,
”
2
01
5
ar
Xi
v
p
re
pr
in
t
a
rX
iv
:
15
11
.
0
84
58
.
[8
]
S
.
C
.
Tur
a
ga
,
“
C
onv
ol
ut
io
na
l
Ne
t
wor
ks c
a
n le
a
r
n t
o G
e
ne
r
a
te
Af
f
in
it
y
G
r
a
ph
s
f
or
I
m
a
ge
S
e
gm
e
nta
ti
on,
”
N
e
ur
al
c
om
pu
ta
ti
on
,
v
ol
.
22
,
no
.
2,
p
p.
51
1
-
53
8
,
20
10
,
d
oi
:
10
.
11
6
2/
ne
c
o
.
20
09
.
10
-
08
-
88
1
.
[9
]
D
.
Ci
res
a
n
, U
.
M
e
i
e
r
,
a
nd
J
.
S
c
hm
id
hu
be
r
,
“
M
ul
ti
-
C
olum
n De
e
p Ne
ur
a
l
Ne
t
wor
ks f
or
I
m
a
ge
C
la
ss
if
ic
a
ti
on,
”
Pro
c
e
e
d
in
gs of
the
I
E
EE
C
on
fe
re
nc
e
on
C
om
pu
te
r
Vi
si
on an
d
P
at
te
r
n Re
c
o
gn
it
io
n
,
20
12
,
pp
.
36
42
-
3
649
,
doi
:
1
0.
11
09
/C
VP
R
.
20
12.
6
24
81
10.
[
10]
T
. L
.
Nwe
,
T.
H.
Da
t
,
a
n
d B
.
M
a
,
“
C
onv
ol
ut
io
na
l
Ne
ur
a
l Ne
t
wor
k w
it
h M
u
lt
i
-
Ta
sk L
e
a
r
n
in
g S
c
he
m
e
f
or
Ac
o
u
sti
c
S
c
e
n
e
C
l
a
s
s
i
f
i
c
a
t
i
o
n
,
”
As
ia
-
P
ac
i
fic
S
ig
na
l a
nd I
n
fo
rm
at
i
on P
roc
e
ss
in
g A
ss
oc
i
at
io
n An
nu
al
Su
mm
it a
nd C
on
fe
re
nc
e
(
AP
SI
P
A ASC
)
,
20
17
,
pp.
1
34
7
-
13
50
,
d
oi
:
10.
1
10
9/
AP
S
I
P
A.
20
17
.
82
82
24
1
.
[
11]
A
.
Kr
iz
he
vs
ky,
I
.
S
uts
ke
ve
r
,
a
nd
G
.
E
.
Hint
on,
“
I
m
a
ge
Ne
t C
la
s
sif
ic
a
ti
on wi
th De
e
p C
on
vo
lu
ti
ona
l Ne
ur
a
l
Ne
t
wor
ks,
”
Adv
anc
e
s N
e
ur
al I
nf
or
ma
ti
on P
roc
e
ss
i
ng Sy
s
te
m
s
,
v
ol
.
25,
no.
2,
pp.
190
7
-
11
05
,
20
12
,
doi
:
1
0.
11
45
/3
06
53
86
.
[
12]
D
. C
.
Ci
re
s
,
“
Hi
gh P
e
r
f
or
m
a
nc
e
C
o
nv
ol
ut
io
na
l
Ne
ur
a
l
Ne
t
wor
ks f
or
I
m
a
ge
C
la
ss
if
ic
a
ti
on,
”
I
JC
A
I
Pro
c
e
e
d
in
gs
-
I
nte
rn
at
io
na
l
J
oi
nt
C
on
fe
re
nc
e
on
Ar
ti
fic
ia
l
I
nte
ll
ige
nc
e
,
v
ol
.
22,
20
11
,
pp.
1
23
7
-
12
42
.
[
13]
D
. C
.
Ci
res
an
,
U.
M
e
ie
r
,
L.
M
.
G
a
m
ba
r
de
ll
a
,
a
nd J.
S
c
hm
idh
ub
e
r
,
“
C
onv
ol
ut
io
na
l Ne
ur
a
l Ne
tw
or
k C
om
m
it
te
e
s
f
or
Ha
n
dwr
it
te
n C
ha
r
a
c
te
r
C
la
ss
if
ic
a
ti
on,
”
I
EE
E C
on
fe
re
nc
e
on D
oc
u
me
n
t Ana
ly
s
is a
nd Re
c
o
gn
it
io
n
,
20
11
,
pp.
11
35
-
11
39
,
d
oi
:
1
0.
1
10
9/I
C
DAR
.
2
01
1.
22
9
.
[
14]
X.
S
.
Ya
ng,
“
I
n
tr
o
duc
ti
on t
o A
lg
or
i
thm
s f
or
Da
ta
M
in
in
g a
nd M
a
c
hi
ne
L
e
a
r
nin
g
,
”
Ac
a
de
mic
p
re
s
s
,
20
19
.
[
15]
B
.
N
a
v
a
n
e
e
t
h
a
n
d
M
.
S
u
c
h
e
t
h
a
,
“
P
S
O
O
pt
im
iz
e
d 1
-
D
C
NN
-
S
VM
Ar
c
h
ite
c
tur
e
f
or
R
e
a
l
-
T
im
e
De
te
c
t
io
n a
nd
C
la
ss
ifi
c
a
ti
on
A
pp
lic
a
ti
on
s,
”
C
omp
ute
rs i
n Bi
ol
ogy
a
nd Me
dic
in
e
,
v
ol
.
10
8,
pp.
85
-
92
,
20
19
,
doi
:
1
0.
10
16
/j.
c
om
p
bi
om
e
d.
20
19.
0
3.
0
17
.
[
16]
M
.
S
z
a
r
v
a
s
,
A.
Yosh
iz
a
wa
,
M
.
Ya
m
a
m
a
to,
a
n
d J.
Oga
ta
,
“
P
e
de
str
ia
n De
te
c
t
io
n wi
th C
o
nv
ol
ut
io
na
l N
e
ur
a
l
Ne
t
wor
ks,
”
I
n
te
l
li
ge
n
t Ve
h
ic
le
s Sy
mp
os
iu
m I
EE
E
,
20
05
,
pp.
22
4
-
22
9
,
do
i
:
1
0.
11
09
/I
VS
.
2
00
5.
1
50
51
06
.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
KA
T
el
eco
m
m
u
n
C
om
put
E
l
C
ont
r
o
l
C
er
vi
ca
l
ca
n
ce
r
c
l
as
s
i
f
i
c
at
i
on us
i
ng c
onv
ol
ut
i
onal
ne
ur
al
ne
t
w
or
k
–
s
uppor
t
v
e
c
t
or
… (
J
ane
E
v
a A
ur
e
l
i
a
)
1611
[
17]
C
.
S
z
e
g
e
d
y
, A
.
To
sh
e
v
,
a
nd
D
.
Er
ha
n,
“
De
e
p
Ne
ur
a
l Ne
t
wor
ks f
or
O
bje
c
t De
te
c
t
io
n,
”
Adv
anc
e
s i
n N
e
ura
l
I
nf
or
mat
io
n P
roc
e
ss
in
g Sy
ste
ms
,
p
p.
25
53
-
25
61
,
20
13
.
[
18]
F
. H
. C
.
Tiv
ive
a
nd
A
.
B
o
uz
e
r
d
oum
,
“
A Ne
w C
la
ss
of
C
o
nv
ol
ut
io
na
l
Ne
ur
a
l Ne
tw
or
k
s (
S
ic
on
ne
t
s)
a
n
d t
he
i
r
Ap
pli
c
a
t
io
n of
F
a
c
e
De
te
c
ti
on,
”
Pr
oc
e
e
d
in
gs
of T
he
I
n
t
e
rn
at
io
na
l J
o
in
t C
o
nfe
re
nc
e
,
v
ol.
3,
20
03,
pp.
21
57
-
21
62
,
doi
:
1
0.
11
09
/I
JC
NN.
20
03.
1
22
37
42
.
[
19]
Z
.
R
usta
m
a
n
d F
.
Ya
ur
ita
,
“
I
n
so
lve
nc
y P
r
e
dic
ti
on
in I
nsu
r
a
nc
e
C
om
pa
n
ie
s
Us
in
g S
u
pp
or
t
Ve
c
t
or
M
a
c
h
ine
s a
n
d
F
uz
z
y Ke
r
ne
l C
-
M
e
a
n
s
,
”
J
.
Phy
s.
:
C
on
f.
Se
r
.
,
v
ol
.
10
28
,
no
.
1
,
20
18
,
do
i:
10.
10
88
/1
74
2
-
65
96
/1
02
8/
1/
01
21
18
.
[
20]
Z
.
R
usta
m
a
nd
N.
P
.
A.
A.
Ar
ia
nta
r
i,
“
S
up
po
r
t
Ve
c
t
o
r
M
a
c
hi
ne
s f
or
C
la
ss
if
y
in
g P
ol
ic
y
ho
lde
r
s S
a
t
isf
a
c
tor
il
y i
n
Aut
om
o
bi
le
I
n
sur
a
nc
e
,
”
20
18
J
.
P
hy
s.
:
C
o
nf.
Se
r
.
,
vol.
1
028
,
no
.
1
,
do
i:
10.
1
08
8/
17
42
-
6
59
6/
10
28
/1/
01
20
05
.
[
21]
E.
G
.
G
onz
a
lo,
Z
.
F
.
M
uñiz
,
P
.
J.
G
.
Nie
to,
A.
S
.
Bern
ard
o
,
a
nd
M
.
M
.
F
e
r
na
nde
s
,
“
H
a
r
d
-
R
oc
k S
ta
b
il
it
y A
na
l
ysi
s
f
or
S
pa
n De
s
ig
n in E
ntr
y
-
T
ype
E
xc
a
va
ti
on
s wi
th L
e
a
r
nin
g C
la
ss
if
ie
r
s,
”
Ma
te
r
ia
ls
,
vo
l.
9,
no
.
7
,
2016
,
doi
:
1
0.
33
90
/m
a
90
70
53
1
.
[
22]
Z
.
R
usta
m
a
nd
D.
Z
a
hr
a
s,
“
C
om
pa
r
i
so
n be
twe
e
n S
up
p
or
t Ve
c
tor
M
a
c
hi
ne
a
n
d F
uz
z
y C
-
M
e
a
n
s
a
s
C
la
ss
if
ie
r
f
or
I
ntr
us
io
n De
te
c
ti
on,
”
J
.
Phy
s.
:
C
on
f.
Se
r
.
,
v
ol.
10
28
,
n
o.
1,
201
8
,
do
i:
10.
1
08
8/
17
42
-
6
59
6/
10
28
/1/
01
22
27
.
[
23]
H
. K
.
P
e
n
a
w
a
r
a
nd Z
.
R
us
ta
m
,
“
A F
uz
z
y L
ogi
c
M
o
de
l t
o F
or
e
c
a
st S
toc
k M
a
r
ke
t M
om
e
n
tum
in I
nd
one
si
a
’
s
P
r
ope
r
ty
a
nd
R
e
a
l
E
sta
te
S
e
c
tor
,
”
20
16
A
I
P
C
onf
e
r
e
n
c
e
P
r
o
c
e
e
d
i
n
g
s
,
vo
l.
18
62,
no
.
1
,
201
6,
doi
:
1
0.
10
63
/1.
49
91
22
9
.
[
24]
N
.
C
hr
is
tia
ni
ni a
nd
J
. S
.
Ta
yl
or
,
“
An I
n
tr
o
du
c
t
io
n to S
upp
or
t Ve
c
t
or
M
a
c
hi
ne
s a
nd O
the
r
Ke
r
ne
l
ba
se
d L
e
a
r
nin
g
M
e
th
ods,
”
C
am
bri
dg
e
u
ni
v
e
rs
it
y
pr
e
ss
,
2
00
0
.
[
25]
Z
.
R
usta
m
,
D.
A.
U
ta
m
i,
R
.
H
ida
ya
t,
J.
P
a
nde
la
k
i a
n
d
W
.
A.
Nu
gr
o
ho,
“
Hy
br
i
d P
r
e
pr
oc
e
s
si
ng M
e
th
od f
or
S
up
por
t
V
e
c
t
o
r
M
a
c
h
i
n
e
f
o
r
C
la
ss
if
ic
a
t
io
n of
I
m
ba
la
nc
e
d C
e
r
e
br
a
l I
nf
a
r
c
ti
on Da
ta
se
ts,
”
20
19
I
nt
e
rn
at
io
na
l
J
ou
rn
al o
n
Adv
anc
e
d Sc
ie
nc
e
E
ng
ine
e
ri
ng I
nf
orm
at
io
n T
e
c
h
no
lo
gy
,
v
ol
.
9
,
no
.
2
,
do
i:
1
0.
1
85
17
/i
ja
se
it.
9.
2.
8
61
5
.
B
I
OGR
A
P
HI
E
S
OF
A
U
T
HOR
S
Jan
e
E
va Au
r
e
l
ia
wa
s
b
or
n in Ja
ka
r
ta
,
1
9 Ju
ne
1
99
8.
S
he
is
a
f
i
na
l ye
a
r
st
ude
nt in
t
he
De
pa
r
te
m
e
n
t of
M
a
the
m
a
t
ic
s,
Un
ive
r
si
ty of
I
nd
one
sia
.
S
he
is c
ur
r
e
n
tl
y wor
k
in
g on he
r
the
si
s,
whi
c
h i
s f
ir
m
ly a
bo
ut a
pp
lie
d m
a
t
he
m
a
t
ic
s
us
in
g
m
a
c
hi
ne
le
a
r
ni
ng.
A
ls
o,
M
s.
Ja
ne
’
s
spe
c
ia
l
ti
es
in r
e
se
a
r
c
h a
r
e
m
os
tl
y a
bo
ut m
a
c
h
ine
le
a
r
n
ing,
m
a
t
he
m
a
t
ic
a
l m
o
de
l
in
g,
a
nd da
ta
m
ini
ng.
Z
u
h
e
r
man
Ru
st
am
i
s a
n As
soc
ia
te
P
r
of
e
ss
or
a
n
d a
le
c
t
ur
e
r
of
the
i
nte
ll
ige
nc
e
c
om
p
uta
ti
on a
t
the
De
pa
r
tm
e
nt
of
M
a
t
he
m
a
t
ic
s,
U
ni
ve
r
s
it
y of
I
n
do
ne
s
i
a
.
He
obta
ine
d h
is M
a
ste
r
of
S
c
ie
nc
e
in
198
9
in
inf
or
m
a
tic
s,
P
a
r
i
s D
ide
r
ot
U
ni
ve
r
s
it
y,
F
r
e
nc
h,
a
nd c
om
p
le
te
d
hi
s P
h.
D.
in
2
00
6
fro
m
c
om
p
ute
r
sc
ie
nc
e
,
Un
ive
r
si
ty
of
I
nd
one
sia
.
As
soc
.
P
r
of
.
Dr
.
R
usta
m
is a
m
e
m
be
r
of
I
E
EE w
ho
is a
c
t
ive
ly
r
e
se
a
r
c
h
in
g m
a
c
h
ine
le
a
r
n
in
g,
pa
t
te
r
n r
e
c
og
ni
ti
on,
ne
ur
a
l
ne
t
wor
k,
a
r
tif
ic
ia
l
int
e
l
li
ge
nc
e
.
Ils
ya W
ir
a
sat
i
i
s a
f
i
na
l
ye
a
r
s
tu
de
n
t i
n t
he
D
e
p
a
r
t
e
m
e
nt
of
M
a
t
he
m
a
t
ic
s,
U
ni
ve
r
s
it
y of
I
nd
one
sia
,
wh
o i
s c
ur
r
e
n
tl
y w
or
k
in
g on he
r
the
si
s.
He
r
r
e
se
a
r
c
h i
s f
ir
m
ly a
b
ou
t a
p
pl
ie
d
m
a
the
m
a
tic
s us
in
g m
a
c
h
ine
le
a
r
n
in
g i
n m
e
dic
a
l f
ie
ld.
M
s.
I
ls
ya
’
s s
pe
c
ia
lt
ie
s
in r
e
se
a
r
c
h a
r
e
m
ost
ly a
bo
ut m
a
c
hi
ne
le
a
r
n
in
g,
m
a
the
m
a
tic
a
l
m
o
de
l
in
g,
a
nd
da
ta
m
in
in
g
Evaluation Warning : The document was created with Spire.PDF for Python.