I
nd
o
ne
s
ia
n J
o
urna
l o
f
E
lect
rica
l En
g
ineering
a
nd
Co
m
p
u
t
er
Science
Vo
l.
21
,
No
.
1
,
J
an
u
ar
y
2
0
2
1
,
p
p
.
5
5
8
~5
6
5
I
SS
N:
2
5
02
-
4
7
5
2
,
DOI
: 1
0
.
1
1
5
9
1
/i
j
ee
cs.v
2
1
.i
1
.
pp
5
5
8
-
565
558
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ee
cs.ia
esco
r
e.
co
m
Liv
e t
o
lea
rn:
l
ea
rning
rules
-
ba
sed a
rtif
icia
l
neura
l
netw
o
rk
Aseel
Sh
a
k
ir
I
.
H
ila
iw
a
h
1
,
H
a
na
n Abed
Alw
a
lly
Abed A
ll
a
h
2
,
B
a
s
i
m
A
kh
ud
ir
Abba
s
3
,
T
o
le
Su
t
i
kn
o
4
1
Im
a
m
A
'
a
d
h
u
m
Un
iv
e
rsit
y
Co
ll
e
g
e
,
De
p
a
rt
m
e
n
t
o
f
F
u
n
d
a
m
e
n
tals o
f
Re
leg
io
n
s F
o
r
G
irl
s,
Ba
g
h
d
a
d
,
Ira
q
2,
3
M
u
sta
n
sir
iy
a
Un
iv
e
rsit
y
,
Co
ll
e
g
e
o
f
S
c
ien
c
e
,
De
p
a
rt
m
e
n
t
o
f
Co
m
p
u
ter S
c
ien
c
e
,
Ba
g
h
d
a
d
,
Ira
q
4
De
p
a
rtme
n
t
o
f
El
e
c
tri
c
a
l
En
g
in
e
e
rin
g
,
Un
iv
e
rsitas
A
h
m
a
d
Da
h
lan
,
Yo
g
y
a
k
a
rta,
In
d
o
n
e
sia
4
Em
b
e
d
d
e
d
S
y
ste
m
a
n
d
P
o
w
e
r
El
e
c
tro
n
ics
Re
se
a
rc
h
G
ro
u
p
(ES
P
E
RG
),
Yo
g
y
a
k
a
rt
a
,
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
Ma
y
5
,
2
0
20
R
ev
i
s
ed
A
u
g
9
,
2
0
20
A
cc
ep
ted
A
u
g
3
0
,
2
0
2
0
A
n
e
x
ten
siv
e
re
v
ie
w
o
f
th
e
a
rti
f
i
c
ial
n
e
u
ra
l
n
e
tw
o
rk
(A
N
N)
is
p
r
e
se
n
ted
in
th
is
p
a
p
e
r.
P
re
v
io
u
s
stu
d
ies
re
v
ie
w
th
e
a
rti
f
icia
l
n
e
u
ra
l
n
e
tw
o
rk
(
AN
N)
b
a
se
d
o
n
th
e
a
p
p
r
o
a
c
h
e
s
(a
lg
o
rit
h
m
s)
u
se
d
o
r
b
a
se
d
o
n
th
e
ty
p
e
s
o
f
th
e
a
rti
f
icia
l
n
e
u
ra
l
n
e
tw
o
rk
(A
NN
).
T
h
e
p
re
s
e
n
ted
p
a
p
e
r
re
v
ie
w
s
th
e
A
NN
b
a
se
d
o
n
th
e
g
o
a
l
o
f
th
e
A
NN
(
m
e
th
o
d
s,
a
n
d
l
a
y
e
rs),
w
h
ich
b
e
c
o
m
e
th
e
m
a
in
o
b
jec
ti
v
e
o
f
th
is
p
a
p
e
r.
A
s
a
f
a
m
o
u
s
a
rti
f
icia
l
in
telli
g
e
n
t
m
o
d
e
l,
A
NN
m
i
m
ics
th
e
h
u
m
a
n
n
e
rv
o
u
s
sy
ste
m
in
h
a
n
d
li
n
g
th
e
in
f
o
rm
a
ti
o
n
tran
sm
it
e
d
b
y
d
iffer
e
n
t
n
o
d
e
s
(a
lso
k
n
o
w
n
a
s
n
e
u
r
o
n
s)
in
th
is
m
o
d
e
l.
T
h
e
se
n
o
d
e
s
a
re
sta
c
k
e
d
in
lay
e
rs
a
n
d
w
o
rk
c
o
ll
e
c
ti
v
e
l
y
to
b
rin
g
a
b
o
u
t
so
lu
t
io
n
t
o
c
o
m
p
lex
p
ro
b
lem
s.
Nu
m
e
ro
u
s
stru
c
tu
re
s
e
x
ist
f
o
r
A
N
N
a
n
d
e
a
c
h
o
f
th
e
se
stru
c
tu
re
s
is
d
e
sig
n
e
d
t
o
a
d
d
re
ss
a
a
sp
e
c
i
f
ic
tas
k
.
Ba
sic
a
ll
y
,
th
e
AN
N
a
rc
h
it
e
c
tu
re
is
c
o
m
p
rise
d
o
f
3
d
if
f
e
r
e
n
t
la
y
e
rs
w
h
e
re
in
th
e
f
irst
la
y
e
r
rp
re
se
n
ts
th
e
in
p
u
t
lay
e
r
th
a
t
c
o
n
sist
o
f
se
v
e
ra
l
in
p
u
t
n
o
d
e
s th
a
t
re
p
re
se
n
t
th
e
in
p
u
t
p
a
ra
m
e
t
e
r
f
o
r
th
e
m
o
d
e
l.
T
h
e
h
i
d
d
e
n
lay
e
r
is
te
s
e
c
o
n
d
la
y
e
r
a
n
d
c
o
n
sists
o
f
a
h
i
d
d
e
n
la
y
e
r
o
f
n
e
u
ro
n
s.
T
h
e
n
e
u
ro
n
s
in
th
is
lay
e
r
a
re
d
irec
tl
y
c
o
n
n
e
c
ted
t
o
t
h
e
n
e
u
ro
n
s
in
t
h
e
o
u
tp
u
t
lay
e
r.
T
h
e
th
ir
d
la
y
e
r
is
th
e
o
u
t
p
u
t
lay
e
r
w
h
ich
is
th
e
m
o
d
e
ls’
re
sp
o
n
se
lay
e
r.
T
h
e
o
u
tp
u
t
la
y
e
r
n
e
u
ro
n
s
h
a
v
e
th
e
a
c
ti
v
a
ti
o
n
f
u
n
c
ti
o
n
s
f
o
r
th
e
c
a
lcu
latio
n
o
f
th
e
AN
N
f
in
a
l
o
u
tp
u
t.
T
h
e
c
o
n
n
e
c
ti
o
n
b
e
twe
e
n
th
e
n
o
d
e
s
in
t
h
e
A
NN
m
o
d
e
l
is
m
e
d
iate
d
b
y
s
y
n
a
p
ti
c
w
e
ig
h
ts.
Th
is
p
a
p
e
r
is
a
c
o
m
p
re
h
e
n
siv
e
stu
d
y
o
f
A
NN
m
o
d
e
ls
a
n
d
th
e
ir
lay
e
rs.
K
ey
w
o
r
d
s
:
A
r
ti
f
icial
n
e
u
r
al
n
et
w
o
r
k
s
Dee
p
lear
n
in
g
L
i
f
elo
n
g
lear
n
in
g
Neu
r
al
n
et
w
o
r
k
T
r
ain
in
g
T
h
is
is
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
Aseel
S
h
ak
ir
I
.
Hilai
w
a
h
Dep
ar
t
m
en
t o
f
Fu
n
d
a
m
en
ta
ls
o
f
R
ele
g
io
n
s
Fo
r
Gir
ls
I
m
a
m
A'
ad
h
u
m
U
n
i
v
er
s
it
y
C
o
lleg
e
,
B
ag
h
d
ad
,
I
r
aq
E
m
ail:
as
s
h
ak
ir
1
9
8
3
@
g
m
ail.
c
o
m
1.
I
NT
RO
D
UCT
I
O
N
T
h
e
cu
r
r
en
tl
y
ex
is
ti
n
g
A
NN
m
o
d
el
s
allo
w
s
f
o
r
ad
j
u
t
m
en
t
o
f
th
e
n
et
w
o
r
k
b
eh
a
v
io
r
b
y
a
lter
in
g
t
h
e
w
ei
g
h
ts
th
a
t
co
n
n
ec
ts
t
h
e
n
e
u
r
o
n
s
to
ea
ch
o
th
er
;
h
o
w
e
v
er
,
th
e
n
et
w
o
r
k
d
esi
g
n
er
i
s
e
x
p
ec
ted
to
s
et
u
p
t
h
e
n
u
m
b
er
o
f
n
e
u
r
o
n
s
,
as
w
ell
as
th
e
s
tr
u
c
tu
r
al
i
n
ter
-
n
e
u
r
o
n
r
elatio
n
s
h
ip
w
h
ich
o
n
ce
d
es
ig
n
ed
,
i
s
s
u
s
ta
in
ed
th
r
o
u
g
h
o
u
t
th
e
li
f
eti
m
e
o
f
t
h
e
n
et
w
o
r
k
[
1
,
2
]
.
T
h
is
i
s
o
n
e
o
f
t
h
e
co
n
s
tr
ai
n
ts
o
n
th
e
ap
p
licab
ilit
y
o
f
A
NN
s
.
“
T
h
e
ab
ilit
y
to
in
ter
p
r
et
an
d
m
an
ip
u
late
i
n
ter
n
al
w
o
r
k
i
n
g
s
o
f
n
e
u
r
al
n
et
w
o
r
k
is
a
m
aj
o
r
b
r
ea
k
th
r
o
u
g
h
i
n
t
h
e
A
N
N
t
h
eo
r
etica
l
r
esear
ch
[
3
-
6
]
.
T
h
e
au
t
h
o
r
p
r
o
v
es
t
h
at
m
o
s
t
o
f
t
h
e
f
ee
d
f
o
r
w
ar
d
n
eu
r
al
n
et
w
o
r
k
s
ar
e
f
u
n
ctio
n
s
;
t
h
u
s
,
m
a
n
y
p
r
o
p
er
ties
o
f
f
u
n
cti
o
n
s
ca
n
b
e
ap
p
lied
to
an
aly
ze
n
eu
r
al
n
et
w
o
r
k
b
eh
av
io
r
s
.
B
ased
o
n
th
is
p
r
o
o
f
,
a
g
r
ap
h
ical
m
ap
p
i
n
g
tec
h
n
iq
u
e
w
a
s
p
r
o
p
o
s
ed
to
in
ter
p
r
et
t
h
e
in
ter
n
al
ac
tiv
i
tie
s
o
f
A
NN
m
o
d
els.
W
ith
t
h
is
tech
n
iq
u
e,
it
i
s
p
o
s
s
ib
le
to
s
t
u
d
y
th
e
i
m
p
ac
t
o
f
n
o
is
y
d
ata
o
n
ANN
m
o
d
elin
g
,
an
d
s
ev
er
al
k
e
y
f
ea
t
u
r
es
o
f
ANN
m
o
d
els
s
u
ch
as
m
e
m
o
r
izatio
n
,
r
o
b
u
s
t
n
ess
,
an
d
s
en
s
iti
v
it
y
f
r
o
m
t
h
e
p
er
s
p
ec
tiv
e
o
f
ar
ti
f
icia
l
n
eu
r
o
n
s
an
d
t
h
eir
co
n
n
ec
tio
n
w
ei
g
h
ts
[
7
-
10
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esia
n
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2502
-
4752
Live
to
lea
r
n
:
Lea
r
n
in
g
r
u
les
-
b
a
s
ed
a
r
tifi
cia
l n
eu
r
a
l n
etw
o
r
k
(
A
s
ee
l S
h
a
kir I
.
Hila
iw
a
h
)
559
T
h
e
d
ec
is
io
n
-
m
a
k
i
n
g
ca
p
ab
ilit
y
o
f
ANNs
is
r
elia
n
t o
n
a
s
et
o
f
w
ei
g
h
ts
a
s
t
h
e
w
ei
g
h
ts
s
er
v
e
as a
s
to
r
e
f
o
r
i
m
p
o
r
tan
t
n
et
w
o
r
k
i
n
f
o
r
m
atio
n
.
T
h
u
s
,
it
i
s
i
m
p
o
r
tan
t
to
en
s
u
r
e
a
ca
r
e
f
u
l
s
elec
tio
n
o
f
t
h
e
n
et
w
o
r
k
w
eig
h
t
s
b
ec
au
s
e
th
e
ac
cu
r
ac
y
o
f
th
e
m
o
d
el
d
ec
is
io
n
is
d
ep
en
d
en
t
o
n
th
e
d
etec
tio
n
o
f
th
e
o
p
ti
m
a
l
w
eig
h
t
[
11
-
20
]
.
T
h
e
ai
m
o
f
t
h
e
tr
ai
n
i
n
g
p
h
ase
i
s
to
d
eter
m
in
e
t
h
e
o
p
ti
m
a
l
w
ei
g
h
t
f
o
r
th
e
w
h
o
le
n
et
w
o
r
k
an
d
t
h
e
tr
ain
i
n
g
p
r
o
ce
s
s
i
s
r
elian
t
o
n
a
s
e
t
o
f
r
u
les.
Se
v
e
r
al
alg
o
r
ith
m
s
h
a
v
e
b
ee
n
b
u
il
t
f
o
r
ANN
w
ei
g
h
ts
ad
j
u
s
t
m
e
n
t
b
ased
o
n
s
p
ec
if
ic
m
ea
s
u
r
es
[
21
,
22
]
.
Du
r
in
g
th
e
tr
ain
i
n
g
p
h
a
s
e,
th
e
w
eig
h
t
s
ar
e
f
ir
s
t
ass
i
g
n
ed
r
an
d
o
m
l
y
b
ef
o
r
e
th
e
esti
m
atio
n
o
f
ea
ch
n
e
u
r
o
n
s
o
u
tp
u
t
b
ased
o
n
s
p
ec
if
ied
r
u
les.
T
h
en
,
th
e
w
ei
g
h
t
s
ar
e
ad
j
u
s
ted
b
y
r
ep
etitiv
e
l
y
m
atc
in
g
w
it
h
th
e
ex
p
ec
ted
o
u
tp
u
t
u
n
til
th
e
o
p
ti
m
al
w
ei
g
h
t
s
ar
e
r
ea
ch
ed
.
T
h
e
lear
n
in
g
f
r
a
m
e
w
o
r
k
s
ca
n
eith
er
b
e
u
n
s
u
p
er
v
i
s
ed
,
s
u
p
er
v
i
s
ed
,
o
r
r
ein
f
o
r
ce
m
e
n
t a
lg
o
r
ith
m
[
23
-
25
]
.
I
n
t
h
is
s
t
u
d
y
,
o
n
e
o
f
t
h
e
k
e
y
c
o
m
p
o
n
en
t
s
i
s
t
h
e
s
e
n
s
iti
v
it
y
a
n
al
y
s
i
s
t
h
at
w
il
l
b
e
co
n
d
u
c
ted
u
s
in
g
t
h
e
A
N
N
m
o
d
els.
T
h
e
ai
m
o
f
t
h
e
s
en
s
i
tiv
it
y
a
n
al
y
s
i
s
is
to
ex
a
m
i
n
e
th
e
co
n
s
is
ten
c
y
o
f
t
h
e
m
o
d
el
s
’
p
er
f
o
r
m
an
ce
w
it
h
k
n
o
w
n
o
il sa
n
d
s
e
x
tr
ac
ti
o
n
b
e
h
a
v
io
r
; it
b
asical
l
y
p
o
r
tr
a
y
s
t
h
e
lev
el
o
f
tr
u
s
t
in
t
h
e
ANN
m
o
d
el.
T
h
e
A
NN
m
o
d
el
s
h
a
v
e
b
ee
n
p
r
o
v
en
s
u
c
ce
s
s
f
u
l
i
n
ar
ea
s
w
h
er
e
s
ta
n
d
ar
d
s
tatis
tical
li
n
ea
r
r
eg
r
es
s
io
n
m
o
d
el
s
h
a
v
e
f
ailed
to
id
en
tify
t
h
e
ex
i
s
ten
ce
o
f
a
n
y
f
o
r
m
o
f
r
elat
io
n
s
h
ip
b
et
wee
n
th
e
k
e
y
in
p
u
t
p
ar
a
m
eter
s
f
r
o
m
t
h
e
ex
tr
ac
tio
n
d
atab
ase
[
26
-
28
]
.
T
h
is
p
ap
er
co
n
tr
ib
u
tes th
e
f
o
ll
o
w
i
n
g
:
a)
R
ev
ie
w
o
f
p
r
ev
io
u
s
s
t
u
d
ies o
n
A
NN
s
.
b)
R
ev
ie
w
o
f
A
NN
s
b
ased
o
n
th
e
g
o
al
an
d
m
et
h
o
d
s
o
f
t
h
e
A
NN
m
eth
o
d
s
.
c)
B
en
ch
m
ar
k
ı
n
g
o
f
d
ı
m
e
n
s
ıo
n
alıt
y
r
ed
u
c
tıo
n
m
et
h
o
d
s
f
o
r
m
al
w
ar
e
cla
s
s
ı
f
ıcat
ıo
n
b
ase
d
o
n
n
et
w
o
r
k
b
eh
av
io
u
r
.
1)
Su
p
er
v
i
s
ed
m
et
h
o
d
T
h
e
n
eu
r
al
n
et
w
o
r
k
s
y
s
te
m
i
n
a
s
u
p
er
v
i
s
ed
m
et
h
o
d
r
eq
u
ir
es
an
ex
ter
n
al
tr
ai
n
in
g
p
h
a
s
e;
t
h
e
n
et
w
o
r
k
is
p
r
esen
ted
w
it
h
t
h
e
e
x
p
ec
ted
o
u
tp
u
t
f
o
r
ea
c
h
i
n
p
u
t
a
n
d
t
h
e
n
et
w
o
r
k
w
ei
g
h
ts
ar
e
ad
j
u
s
ted
b
y
co
m
p
ar
i
n
g
t
h
e
ac
tu
al
o
u
tp
u
t
w
it
h
t
h
e
ex
p
e
cted
o
u
tp
u
t
f
o
r
ea
ch
n
o
d
e
[
29
-
31
]
.
T
h
e
o
u
tp
u
t
er
r
o
r
r
at
e
is
p
r
o
g
r
ess
i
v
el
y
d
ec
r
ea
s
ed
; th
is
p
r
o
ce
s
s
i
s
r
ep
ea
ted
u
n
til t
h
e
ANN
o
u
tp
u
t
is
c
lo
s
e
to
t
h
e
ex
p
ec
ted
o
u
tp
u
t.
B
ac
k
p
r
o
p
ag
atio
n
a
n
d
Delta
r
u
le
ar
e
th
e
m
o
s
tl
y
u
s
ed
tr
ain
i
n
g
i
n
s
ta
n
ce
s
f
o
r
s
u
p
er
v
is
ed
lear
n
in
g
m
et
h
o
d
[
32
]
.
2)
Un
s
u
p
er
v
i
s
ed
m
et
h
o
d
Fo
r
th
is
n
et
w
o
r
k
,
th
e
lear
n
i
n
g
p
h
aseis
n
o
t
r
elia
n
t
o
n
an
y
e
x
t
er
n
al
tr
ain
in
g
p
h
a
s
e
i
n
t
h
is
m
e
th
o
d
.
T
h
e
A
N
N
m
o
d
el
lead
s
t
h
e
lear
n
i
n
g
p
r
o
ce
s
s
f
o
r
t
h
e
w
h
o
le
m
o
d
el
b
ased
o
n
s
p
ec
if
ic
cr
iter
ia
a
n
d
s
u
ch
cr
iter
ia
i
s
f
o
r
m
ed
w
i
t
h
in
t
h
e
n
eu
r
al
n
et
wo
r
k
w
it
h
o
u
t a
n
y
ex
ter
n
al
as
s
is
t
an
ce
[
32
]
.
3)
R
ein
f
o
r
ce
m
e
n
t le
ar
n
in
g
m
et
h
o
d
Her
e,
th
e
NN
ca
n
lear
n
i
ts
b
e
h
av
io
u
r
s
f
r
o
m
t
h
e
e
x
ter
n
al
wo
r
ld
.
I
t
r
esem
b
les
t
h
e
s
u
p
er
v
is
ed
m
eth
o
d
o
f
lear
n
i
n
g
j
u
s
t
th
at
t
h
e
r
ein
f
o
r
ce
m
e
n
t
lear
n
in
g
m
eth
o
d
is
d
ep
en
d
en
t
o
n
less
p
r
e
-
in
f
o
r
m
atio
n
an
d
d
o
es
r
eies
less
o
n
e
x
ac
tn
e
s
s
o
f
th
e
o
u
tp
u
t
.
Ho
w
ev
er
,
t
h
e
m
o
d
els
’
o
u
tp
u
t
is
co
n
s
id
er
ed
as tr
u
e
o
r
f
alse
[
33
]
.
4)
A
r
ti
f
icial
Ne
u
r
al
Net
w
o
r
k
Mo
d
els
T
h
e
A
NN
m
o
d
els
co
n
s
i
s
t
o
f
t
w
o
m
o
d
el
s
t
h
at
f
o
r
m
t
h
e
m
ai
n
i
n
f
r
a
s
tr
u
ct
u
r
e
o
f
ANN.
Fi
g
u
r
e
1
s
h
o
w
s
th
e
A
NN
m
o
d
els.
Fig
u
r
e
1
.
T
h
e
A
NN
m
o
d
els
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2502
-
475
2
I
n
d
o
n
esia
n
J
E
lec
E
n
g
&
C
o
m
p
Sci,
Vo
l.
21
,
No
.
1
,
J
an
u
ar
y
2
0
2
1
:
558
-
565
560
1
.
1
.
Sin
g
le
la
y
er
perc
ept
ro
n (
SL
P
)
T
h
e
b
asic
A
NN
m
o
d
el
is
r
ep
r
esen
ted
a
s
a
f
ee
d
f
o
r
w
ar
d
m
o
d
el.
Sin
g
le
la
y
er
p
er
ce
p
tr
o
n
m
o
d
el
s
ar
e
co
m
p
r
is
ed
o
f
o
n
l
y
t
h
e
in
p
u
t
&
o
u
tp
u
t
la
y
er
s
as
t
h
e
y
lack
th
e
f
ee
d
b
ac
k
p
r
o
ce
s
s
.
A
d
d
iti
o
n
all
y
,
t
h
e
y
lack
a
h
id
d
en
la
y
er
in
t
h
eir
s
tr
u
ct
u
r
e
b
u
t
r
elies
o
n
s
u
p
er
v
is
ed
lear
n
in
g
to
d
eter
m
in
e
t
h
e
n
et
w
o
r
k
o
u
tp
u
t.
T
h
e
y
ar
e
t
y
p
icall
y
u
s
ed
i
n
s
o
l
v
i
n
g
s
i
m
p
le
p
r
o
b
lem
s
th
at
in
v
o
l
v
e
s
i
m
p
le
co
m
p
u
tatio
n
s
.
T
h
e
S
L
P
m
o
d
el
s
tr
u
ct
u
r
e
i
s
d
ep
icted
in
Fig
u
r
e
2
[
34
-
37
]
.
Fig
u
r
e
2
.
T
h
e
s
tr
u
ctu
r
e
o
f
S
L
P
m
o
d
el
1
.
2
.
M
ultila
y
er
p
er
ce
pt
ro
n (
M
L
P
)
T
h
e
d
if
f
er
en
ce
b
et
w
ee
n
M
L
P
an
d
SL
P
is
th
e
p
r
esen
ce
o
f
a
h
id
d
en
la
y
er
in
ML
P
;
h
e
n
ce
,
th
e
ML
P
h
as t
h
r
ee
b
asic la
y
er
s
w
h
ic
h
ar
e
th
e
i
n
p
u
t,
h
id
d
en
,
a
n
d
o
u
tp
u
t
la
y
er
s
.
T
h
e
n
u
m
b
er
o
f
h
id
d
en
la
y
er
s
i
n
th
e
M
L
P
v
ar
ies
.
Fi
g
u
r
e
3
d
ep
icts
th
e
ML
P
s
tr
u
ct
u
r
e
[
34
]
.
Fig
u
r
e
3
.
T
h
e
s
tr
u
ctu
r
e
o
f
m
u
l
tila
y
er
p
er
ce
p
tr
o
n
(
ML
P
)
As
a
f
ee
d
f
o
r
w
ar
d
n
eu
r
al
n
e
t
w
o
r
k
,
t
h
e
ML
P
is
co
m
m
o
n
l
y
u
s
ed
to
s
o
lv
e
n
o
n
l
in
ea
r
p
r
o
b
lem
s
.
I
t
allo
w
s
o
n
e
-
w
a
y
in
f
o
r
m
a
tio
n
f
lo
w
t
h
r
o
u
g
h
th
e
n
et
w
o
r
k
w
it
h
o
u
t
f
ee
d
b
ac
k
,
m
ea
n
i
n
g
t
h
at
th
er
e
is
n
o
lo
o
p
o
r
c
y
cle.
A
t
y
p
ic
a
l
FF
NN
is
m
a
d
e
u
p
o
f
3
la
y
er
s
a
n
d
ea
ch
o
f
t
h
ese
la
y
er
s
co
n
ta
in
s
s
o
m
e
n
o
d
es
b
ased
o
n
th
e
co
n
s
id
er
ed
p
r
o
b
lem
t
y
p
e.
T
h
e
in
p
u
t
d
ata
i
s
f
ed
in
to
t
h
e
m
o
d
el
v
ia
t
h
e
i
n
p
u
t
la
y
er
a
n
d
f
o
r
w
a
r
d
ed
to
th
e
h
id
d
en
la
y
er
f
o
r
o
n
w
ar
d
p
r
o
ce
s
s
i
n
g
.
Af
ter
p
r
o
ce
s
s
i
n
g
t
h
e
d
ata
b
y
th
e
h
id
d
en
la
y
er
,
th
e
o
u
tp
u
t
u
s
f
o
r
w
ar
d
ed
to
t
h
e
o
u
tp
u
t
la
y
er
to
d
eter
m
i
n
e
th
e
f
i
n
al
m
o
d
el
r
esu
lt.
T
h
e
ML
P
m
o
d
el
is
n
o
r
m
all
y
tr
ain
ed
u
s
i
n
g
th
e
b
ac
k
p
r
o
p
ag
atio
n
alg
o
r
ith
m
[
38
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esia
n
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2502
-
4752
Live
to
lea
r
n
:
Lea
r
n
in
g
r
u
les
-
b
a
s
ed
a
r
tifi
cia
l n
eu
r
a
l n
etw
o
r
k
(
A
s
ee
l S
h
a
kir I
.
Hila
iw
a
h
)
561
2.
T
RAI
NIN
G
P
RO
C
E
SS
T
h
e
f
o
llo
w
i
n
g
p
r
o
ce
s
s
es
ar
e
in
v
o
l
v
ed
in
th
e
o
v
er
all
tr
ai
n
in
g
o
f
th
e
A
NN
m
o
d
el;
t
h
e
in
p
u
t
d
ata
ar
e
w
ei
g
h
ed
an
d
f
ed
in
to
th
e
m
o
d
el
v
ia
th
e
in
p
u
t
la
y
er
;
t
h
e
d
ata
is
f
o
r
w
ar
d
ed
to
th
e
h
id
d
en
la
y
er
f
o
r
t
h
e
h
id
d
en
la
y
er
n
e
u
r
o
n
s
to
p
r
o
d
u
ce
th
e
o
u
tp
u
ts
b
y
ap
p
l
y
i
n
g
a
n
ac
tiv
atio
n
f
u
n
ct
io
n
to
th
e
s
u
m
o
f
th
e
w
eig
h
ted
in
p
u
t
v
alu
e
s
.
T
h
en
,
co
n
s
id
er
in
g
th
e
h
id
d
en
la
y
er
-
o
u
tp
u
t
la
y
er
co
n
n
ec
tio
n
s
,
th
e
r
esu
lti
n
g
o
u
tp
u
ts
ar
e
w
ei
g
h
ted
[
39
]
.
T
h
e
o
u
tp
u
t
la
y
er
g
e
n
er
ates
t
h
e
ex
p
ec
ted
m
o
d
el
r
es
u
lts
[
40
]
.
T
h
e
ex
p
ec
ted
lear
n
in
g
i
s
ac
h
ie
v
ed
b
y
co
n
t
in
o
u
s
l
y
ad
j
u
s
tin
g
th
e
i
n
ter
co
n
n
ec
tio
n
w
ei
g
h
ts
o
f
th
e
n
et
w
o
r
k
u
n
til
th
e
r
ea
l
n
e
u
r
o
n
o
u
tp
u
t
m
atc
h
es
clo
s
el
y
w
it
h
th
e
tar
g
et
o
u
tp
u
t
n
e
u
r
o
n
b
ased
o
n
th
e
tr
ain
i
n
g
d
ata.
T
h
e
v
ar
iato
n
b
ew
ee
n
th
ac
t
u
al
an
d
p
r
ed
icted
o
u
tp
u
ts
is
ca
lled
th
e
er
r
o
r
v
alu
e.
I
t
is
a
d
if
f
icu
l
t
task
to
d
eter
m
i
n
e
w
h
et
h
er
a
n
et
w
o
r
k
h
as
b
ee
n
s
u
f
f
icien
t
l
y
tr
ain
ed
in
o
r
d
er
to
ter
m
i
n
ate
t
h
e
tr
ain
i
n
g
p
r
o
ce
s
s
;
th
i
s
p
r
o
b
le
m
ca
n
b
e
ad
d
r
ess
e
d
a
n
et
w
o
r
k
ca
lib
r
atio
n
a
n
d
t
h
is
ca
n
b
e
d
o
n
e
i
n
2
w
a
y
s
:
(
i)
C
h
ec
k
in
g
th
e
n
u
m
b
e
r
o
f
ev
e
n
t
s
t
h
at
h
as
o
cc
u
r
r
ed
s
in
ce
t
h
e
o
cc
u
r
r
en
ce
o
f
t
h
e
m
i
n
i
m
u
m
er
r
o
r
f
ac
to
r
(
u
s
u
all
y
b
et
w
ee
n
2
0
,
0
0
0
–
4
0
,
0
0
0
f
o
r
B
P
N)
;
(
ii)
b
y
ca
lc
u
la
tin
g
t
h
e
ca
lib
r
atio
n
te
s
t
i
n
ter
v
al
(
it
r
eg
u
lates
t
h
e
co
n
v
er
g
e
n
ce
r
ate
o
f
t
h
e
iter
at
io
n
p
r
o
ce
s
s
es).
T
h
e
test
s
et
is
in
d
ir
ec
tl
y
i
n
v
o
lv
ed
in
t
h
e
d
e
ter
m
i
n
atio
n
o
f
th
e
ap
p
r
o
p
r
iate
tim
e
to
ter
m
in
a
te
n
et
w
o
r
k
tr
ai
n
in
g
[
41
,
42
]
.
2
.
1
.
K
o
ho
nen neura
l net
w
o
r
k
T
h
is
n
et
w
o
r
k
i
s
a
s
el
f
-
o
r
g
an
iz
in
g
m
ap
n
e
t
w
o
r
k
t
h
at
ca
n
lear
n
w
it
h
o
u
t
t
h
e
n
ee
d
f
o
r
an
o
u
tp
u
t
d
ata.
I
t
ex
p
lo
its
t
h
e
clu
s
ter
i
n
g
p
r
in
cip
le
to
s
ep
ar
ate
d
ata
in
to
id
en
ti
ca
l
ca
teg
o
r
ies.
I
t
co
n
s
is
t
s
o
f
o
n
l
y
an
in
p
u
t
la
y
er
an
d
an
o
u
tp
u
t la
y
er
[
43
]
.
2
.
2
.
P
r
o
ba
bil
is
t
ic
neura
l net
w
o
rk
T
h
is
n
et
w
o
r
k
ca
n
tr
ain
o
n
f
e
w
d
ata
s
ets;
its
tr
ain
in
g
p
h
ase
i
s
s
o
f
a
s
t
t
h
at
it
ca
n
b
e
s
u
f
f
icien
tl
y
tr
ai
n
ed
in
j
u
s
t
o
n
e
p
ar
t
o
f
th
e
tr
ain
i
n
g
s
et.
T
h
e
P
NN
also
clu
s
ter
s
d
ata
in
to
s
p
ec
if
ic
n
u
m
b
er
o
f
o
u
tp
u
t
ca
teg
o
r
ies
a
s
s
h
o
w
n
in
F
ig
u
r
e
4
.
Fig
u
r
e
4
.
Stru
ct
u
r
e
o
f
b
ac
k
p
r
o
p
ag
atio
n
n
et
w
o
r
k
(
B
P
N)
w
it
h
3
h
id
d
en
la
y
er
s
2
.
3
.
L
ea
rning
rules
Var
io
u
s
r
u
les
ar
e
av
ailab
le
f
o
r
lear
n
in
g
ANN
m
o
d
els;
t
h
e
ai
m
o
f
u
s
i
n
g
t
h
ese
r
u
le
s
i
s
t
o
d
etec
t
a
g
iv
e
n
s
et
o
f
n
et
w
o
r
k
w
ei
g
h
t
s
.
So
m
e
r
esear
c
h
er
s
r
el
y
o
n
s
e
v
er
al
lear
n
i
n
g
r
u
le
s
(
s
u
c
h
as
b
io
lo
g
ical
lear
n
in
g
)
w
h
ile
o
th
er
s
f
o
cu
s
o
n
t
h
e
ev
alu
atio
n
o
f
th
eir
p
er
ce
p
tio
n
o
f
tr
ain
i
n
g
.
I
t
is
s
till
m
o
r
e
ted
io
u
s
lear
n
i
n
g
A
NN
co
m
p
ar
ed
to
th
e
f
ac
ilit
atio
n
p
r
o
v
id
ed
b
y
th
e
lear
n
i
n
g
r
u
les.
Am
o
n
g
t
h
e
av
ailab
le
co
m
m
o
n
lear
n
i
n
g
r
u
les
ar
e
T
h
e
Delta
R
u
le,
Heb
b
r
u
le,
H
o
p
f
ield
r
u
le,
&
Ko
h
o
n
en
's lear
n
in
g
r
u
le
[
43
]
.
a)
Heb
b
’
s
R
u
le:
T
h
is
i
s
a
p
o
p
u
l
ar
lear
n
in
g
r
u
le
p
r
o
p
o
s
ed
b
y
Heb
b
in
1
9
4
9
.
T
h
e
Heb
b
r
u
le
p
r
o
v
id
es
th
at
w
h
e
n
a
n
o
d
e
is
f
ed
b
y
a
n
y
o
th
er
n
o
d
e
w
i
th
i
n
a
n
e
t
w
o
r
k
,
a
n
d
all
th
e
o
th
er
n
o
d
es
ar
e
ac
t
iv
e,
th
ese
o
th
er
n
o
d
es a
r
e
ar
e
co
n
s
id
er
ed
to
h
av
e
r
ein
f
o
r
ce
d
w
ei
g
h
ts
[
44
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2502
-
475
2
I
n
d
o
n
esia
n
J
E
lec
E
n
g
&
C
o
m
p
Sci,
Vo
l.
21
,
No
.
1
,
J
an
u
ar
y
2
0
2
1
:
558
-
565
562
b)
Ho
p
f
ield
R
u
le:
T
h
is
r
u
le
d
eter
m
i
n
es
th
e
lev
e
l
o
f
s
tr
e
n
g
th
s
a
n
d
w
ea
k
n
e
s
s
es.
T
h
is
r
u
le
p
r
o
v
id
es
th
a
t
i
f
t
h
e
in
p
u
t
&
o
u
tp
u
t
f
o
r
n
o
d
es
i
n
ter
lin
k
ed
to
a
co
m
m
o
n
w
ei
g
h
t
h
av
e
a
s
i
m
ilar
s
i
g
n
(
ac
ti
v
e/p
ass
iv
e)
,
th
e
n
,
t
h
e
w
ei
g
h
t s
h
o
u
ld
b
e
in
cr
ea
s
ed
/d
e
cr
ea
s
ed
b
y
u
s
i
n
g
t
h
e
lear
n
in
g
r
ate
[
44
]
.
c)
Delta
L
a
w
:
Als
o
a
m
o
n
g
t
h
e
c
o
m
m
o
n
est
r
u
le
s
u
s
ed
i
n
lear
n
in
g
p
r
o
ce
s
s
es;
it
i
s
al
s
o
r
ef
er
r
ed
to
as
leas
t
m
ea
n
s
q
u
ar
e
(
L
MS)
lear
n
i
n
g
r
u
le.
T
h
e
ai
m
o
f
t
h
i
s
la
w
is
to
r
ed
u
ce
th
e
r
ate
o
f
er
r
o
r
(
d
elta)
b
et
w
ee
n
t
h
e
ex
p
ec
ted
an
d
ac
tu
al
o
u
tp
u
ts
f
o
r
ea
ch
o
f
th
e
n
et
w
o
r
k
s
’
o
u
tp
u
t
la
y
er
n
e
u
r
o
n
s
[
44
,
45
]
.
d)
Gr
ad
ien
t
Descen
t
L
a
w
:
T
h
is
i
s
th
e
co
m
m
o
n
e
s
t
r
u
le
u
s
ed
f
o
r
A
NN
lear
n
i
n
g
;
it
is
s
i
m
ilar
to
th
e
d
elta
la
w
b
u
t
in
th
i
s
r
u
le,
th
e
s
et
o
f
s
y
n
a
p
tic
w
ei
g
h
t
s
is
u
p
d
ated
b
y
tr
a
n
s
f
er
r
in
g
th
e
er
r
o
r
r
ate
th
r
o
u
g
h
th
e
n
et
w
o
r
k
.
T
h
is
r
u
le
u
s
e
s
p
r
o
p
o
r
tio
n
al
co
n
s
ta
n
t
i
n
co
n
s
id
er
atio
n
o
f
t
h
e
lear
n
in
g
r
ate
th
at
i
s
ef
f
ec
ti
v
e
o
n
th
e
s
et
o
f
w
ei
g
h
ts
[
44
]
.
So
m
eti
m
es,
tr
ain
i
n
g
p
er
f
o
r
m
an
ce
is
i
m
p
r
o
v
ed
b
y
u
s
i
n
g
d
i
f
f
er
en
t
lear
n
in
g
r
ates
f
o
r
d
if
f
er
e
n
t
n
et
w
o
r
k
la
y
er
s
.
T
h
e
s
p
ec
if
ied
lear
n
i
n
g
r
ate
f
o
r
la
y
er
s
n
ea
r
th
e
o
u
tp
u
t
la
y
er
is
n
o
r
m
all
y
s
m
a
ller
in
s
o
m
e
n
e
u
r
al
n
et
w
o
r
k
m
o
d
el
s
co
m
p
ar
ed
to
th
e
s
p
ec
if
ied
le
ar
n
in
g
r
ate
f
o
r
la
y
er
n
ea
r
th
e
i
n
p
u
t
[
44
]
.
e)
Ko
h
o
n
e
n
L
ea
r
n
in
g
R
u
le:
T
h
is
r
u
le
w
as
ad
o
p
ted
b
y
T
eu
v
o
Ko
h
o
n
en
as
a
b
io
lo
g
ical
in
s
p
ir
ed
ap
p
r
o
ac
h
.
I
t
is
also
ca
lled
a
s
el
f
-
o
r
g
a
n
izi
n
g
to
p
o
lo
g
y
;
it
s
co
n
ce
p
t
is
b
as
ed
o
n
th
e
f
ac
t
t
h
at
t
h
e
p
r
o
ce
s
s
in
g
ele
m
en
ts
co
m
p
ete
to
i
m
p
r
o
v
e
t
h
eir
w
e
i
g
h
t
s
,
w
i
th
t
h
e
b
est
p
er
f
o
r
m
i
n
g
ele
m
e
n
t
i
n
ter
m
s
o
f
t
h
e
b
es
t
o
u
tp
u
t
v
alu
e
b
ein
g
co
n
s
id
er
ed
th
e
lear
n
er
th
at
th
e
clo
s
e
ele
m
e
n
t
s
w
il
l
l
ea
r
n
f
r
o
m
.
Ho
w
e
v
er
,
t
h
is
lea
r
n
er
s
tr
i
v
es
to
k
ee
p
th
e
co
m
p
etito
r
s
f
r
o
m
i
m
p
r
o
v
in
g
th
eir
o
w
n
w
e
ig
h
t
s
.
T
h
e
n
u
m
b
er
o
f
p
r
o
ce
s
s
in
g
ele
m
e
n
ts
w
it
h
in
t
h
e
n
eig
h
b
o
u
r
h
o
o
d
is
v
ar
ied
d
u
r
in
g
th
e
lear
n
i
n
g
p
h
a
s
e
an
d
is
u
s
u
all
y
k
ep
t lo
w
[
44
]
.
3.
NNIDS
-
N
E
URA
L
N
E
T
WO
RK
B
ASE
D
I
N
T
RUS
I
O
N
D
E
T
E
C
T
I
O
N
SY
ST
E
M
I
n
th
is
w
o
r
k
,
a
s
y
s
te
m
t
h
at
d
e
p
en
d
s
o
n
a
n
o
v
el
ap
p
r
o
ac
h
f
o
r
r
ea
l
-
ti
m
e
tr
af
f
ic
a
n
al
y
s
is
w
a
s
ev
elo
p
ed
f
o
r
th
e
d
etec
tio
n
a
n
d
class
i
f
ic
atio
n
o
f
m
al
w
ar
e.
T
h
e
s
y
s
te
m
w
as
n
o
d
ev
elo
p
ed
to
an
al
y
ze
th
e
tr
af
f
ic
it
s
el
f
b
u
t
to
ex
tr
ac
t
t
h
e
s
ec
o
n
d
ar
y
d
ata
f
ea
t
u
r
es
t
h
at
c
o
n
ta
in
s
m
u
c
h
i
n
f
o
r
m
atio
n
ab
o
u
t
t
h
e
g
en
er
al
s
t
ate
o
f
t
h
e
n
e
t
w
o
r
k
[
46
]
.
T
h
e
cu
r
r
en
t
s
y
s
te
m
w
as
b
u
ilt
o
n
a
d
ata
s
et
t
h
at
h
as
9
0
d
i
m
en
s
io
n
s
w
h
ic
h
i
s
m
o
r
e
t
h
a
n
t
h
e
n
u
m
b
er
u
s
ed
f
o
r
DNN
class
i
f
ier
s
;
h
e
n
ce
,
th
e
ac
cu
r
a
c
y
o
f
th
e
an
al
y
s
i
s
is
en
s
u
r
ed
.
T
h
e
u
s
e
o
f
th
i
s
lar
g
e
m
al
w
ar
e
d
ataset
to
tr
ain
t
h
e
D
NN
m
a
k
es
i
p
o
s
ib
le
to
d
etec
t
m
al
w
ar
e
i
n
th
e
d
ataset
a
n
d
m
al
w
ar
e
t
h
at
e
x
h
i
b
it
s
i
m
ilar
n
et
w
o
r
k
b
eh
av
io
r
.
T
h
is
ap
p
r
o
ac
h
u
s
ed
t
w
o
s
m
aller
d
i
m
en
s
io
n
a
l
D
NN
s
y
s
te
m
s
to
d
etec
t
m
al
w
a
r
e;
th
is
r
ed
u
ce
d
t
h
e
co
m
p
u
tatio
n
al
p
o
w
er
w
h
e
n
c
o
m
p
ar
ed
to
o
n
e
lar
g
er
d
i
m
en
s
io
n
al
DNN.
T
h
er
e
f
o
r
e,
a
a
s
y
s
te
m
f
o
r
r
ea
l
-
ti
m
e
m
al
w
ar
e
m
o
n
ito
r
in
g
ca
n
b
e
d
ev
elo
p
ed
.
T
h
e
ac
u
r
ac
y
a
n
d
p
er
f
o
r
m
an
ce
o
f
f
o
u
r
m
ac
h
i
n
e
l
ea
r
n
in
g
cla
s
s
i
f
ier
s
w
er
e
also
an
a
l
y
ze
d
an
d
d
eter
m
i
n
e
d
; a
d
d
itio
n
all
y
,
a
h
eu
r
i
s
ti
c
p
r
ed
ictio
n
ev
alu
atio
n
w
as p
e
r
f
o
r
m
ed
[
47
]
.
3
.
1
.
Arc
hıte
ct
ure
Fig
u
r
e
5
s
h
o
w
s
t
h
e
o
v
er
all
s
tr
u
ctu
r
e
o
f
t
h
e
p
r
o
p
o
s
ed
s
y
s
te
m
.
T
h
e
d
ata
p
ac
k
ets
ar
e
ca
p
tu
r
ed
b
y
a
d
ata
p
ac
k
et
s
n
i
f
f
ier
an
d
tr
an
s
f
er
r
ed
in
th
e
p
r
o
tecte
d
n
et
w
o
r
k
s
.
T
h
e
o
b
tain
ed
d
ata
f
r
o
m
t
h
e
s
n
if
f
ier
is
f
ir
s
t
p
r
ep
r
o
ce
s
s
ed
w
it
h
t
h
e
an
al
y
z
er
to
ex
tr
ac
ts
th
e
i
m
p
o
r
tan
t
d
ata
f
ea
t
u
r
es.
T
h
e
DNN
class
i
f
ier
is
t
h
en
u
s
ed
to
an
al
y
ze
t
h
e
s
et
o
f
s
a
v
ed
f
ea
t
u
r
es;
t
h
e
c
h
a
n
ce
s
o
f
t
h
e
m
ali
cio
u
s
n
ess
o
f
th
e
d
ata
s
tr
ea
m
in
th
e
n
et
w
o
r
k
i
s
co
n
s
id
er
ed
th
e
o
u
tp
u
t o
f
th
e
s
y
s
te
m
.
Fig
u
r
e
5
.
S
y
s
te
m
o
v
er
v
ie
w
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esia
n
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2502
-
4752
Live
to
lea
r
n
:
Lea
r
n
in
g
r
u
les
-
b
a
s
ed
a
r
tifi
cia
l n
eu
r
a
l n
etw
o
r
k
(
A
s
ee
l S
h
a
kir I
.
Hila
iw
a
h
)
563
3
.
2
.
Dua
l dnn
m
a
lw
a
re
det
ec
t
ıo
n a
nd
cla
s
s
ıf
ıca
t
ıo
n
Af
ter
e
x
tr
ac
t
n
g
th
e
9
0
d
ata
f
ea
tu
r
es
f
r
o
m
th
e
P
C
A
P
f
i
le,
h
e
y
w
er
e
f
ed
i
n
to
t
h
e
DN
N
clas
s
i
f
ier
w
h
ic
h
w
a
s
b
u
ilt
u
s
i
n
g
th
e
Ker
as
f
r
am
e
w
o
r
k
[
4
8
]
.
T
h
is
w
o
r
k
p
r
o
p
o
s
ed
th
e
u
s
e
o
f
t
w
o
DNN
cla
s
s
if
ier
s
;
th
e
f
ir
s
t
D
NN
class
i
f
ier
o
n
l
y
d
eter
m
i
n
e
s
t
h
e
s
tat
u
s
o
f
th
e
tr
a
f
ic
(
n
o
r
m
al
o
r
s
u
s
p
icio
u
s
)
a
n
d
w
h
e
a
s
u
s
p
icio
u
s
ac
ti
v
it
y
i
s
d
etec
ted
,
th
e
s
ec
o
n
d
DN
N
cl
ass
i
f
ier
w
ill
b
e
r
eq
u
ir
ed
to
f
u
r
th
er
cla
s
s
i
f
y
t
h
e
tr
af
ic
b
as
ed
o
n
th
e
m
al
w
ar
e
d
ataset.
As
a
m
u
lticlas
s
clas
s
if
ier
,
th
e
s
ec
o
n
d
DNN
class
i
f
ier
s
tr
iv
es
to
ass
i
g
n
th
e
n
a
m
e
o
f
alr
ea
d
y
k
n
o
w
n
m
al
w
ar
e
to
t
h
e
i
n
p
u
t
d
ata
b
ased
o
n
t
h
e
d
ataset
it
h
as
b
ee
n
tr
ai
n
ed
o
n
.
T
h
e
ch
a
n
ce
s
o
f
m
al
w
ar
e
tr
a
f
f
ic
b
elo
n
g
i
n
g
to
an
y
o
f
t
h
e
k
n
o
w
n
m
al
w
ar
e
ar
e
g
en
er
ated
an
d
t
h
e
y
ar
e
h
elp
f
u
l
f
o
r
h
eu
r
i
s
tic
d
et
ec
tio
n
[
48
,
49
]
.
3
.
3
.
B
ench
m
a
r
kın
g
dı
m
e
ns
ıo
na
lıty
re
du
ct
ıo
n
m
et
ho
ds
f
o
r
m
a
lw
a
re
cla
s
s
ıf
ıca
t
ıo
n
ba
s
ed
o
n
net
w
o
rk
beha
v
ıo
r
M
a
lw
a
r
e
c
l
a
s
s
i
f
i
c
a
t
i
o
n
b
a
s
e
d
o
n
n
e
t
w
o
r
k
b
e
h
a
v
i
o
r
i
s
a
n
i
m
p
o
r
t
a
n
t
s
t
e
p
i
n
t
h
e
a
n
a
l
y
s
i
s
f
o
r
m
a
l
w
a
r
e
d
e
t
e
c
t
i
o
n
.
T
h
e
e
f
f
i
c
i
e
n
c
y
o
f
t
h
i
s
s
t
e
p
i
s
c
r
u
c
i
a
l
f
o
r
b
e
t
t
e
r
p
e
r
f
o
r
m
a
n
c
e
o
f
t
h
e
I
D
S
.
t
h
e
d
e
e
p
n
e
u
r
a
l
n
e
t
w
o
r
k
(
D
N
N
)
w
a
s
t
h
e
b
e
s
t
o
p
t
i
o
n
i
n
t
e
r
m
s
o
f
a
c
c
u
r
a
c
y
,
b
u
t
n
o
t
i
n
t
e
r
m
s
o
f
p
e
r
f
o
r
m
a
n
c
e
.
W
e
p
r
o
p
o
s
e
tw
o
w
a
y
s
o
f
i
n
c
r
e
a
s
i
n
g
p
e
r
f
o
r
m
a
n
c
e
b
y
r
e
d
u
c
i
n
g
t
h
e
d
i
m
e
n
s
i
o
n
a
l
i
t
y
o
f
t
h
e
d
a
t
a
s
e
t
s
a
n
d
s
e
l
e
c
t
i
n
g
t
h
e
o
p
t
i
m
a
l
f
e
a
t
u
r
e
s
[
50
,
51
]
.
4.
ADVA
N
T
A
G
E
S O
F
ANN
T
h
e
r
o
b
u
s
tn
e
s
s
o
f
th
e
A
N
N
a
a
p
r
ed
ictio
n
to
o
l is attr
ib
u
ted
to
its
f
o
llo
w
in
g
ca
p
ab
ilit
ie
s
;
a)
A
N
N
m
a
k
es a
r
ep
id
an
d
co
n
f
i
d
en
t p
r
ed
ictio
n
o
n
ce
a
n
e
w
d
at
aset is p
r
esen
ted
to
th
e
co
n
s
tr
u
cted
m
o
d
el.
b)
As d
ata
-
d
r
i
v
e
n
m
o
d
els,
ANNs
d
o
n
o
t n
ee
d
a
p
r
e
-
k
n
o
w
led
g
e
o
f
th
e
d
ata
th
a
t th
e
y
w
ill b
e
ap
p
lied
.
c)
A
N
N
lear
n
s
t
h
e
d
ata
b
eh
a
v
i
o
r
b
y
s
el
f
-
t
u
n
i
n
g
its
p
ar
a
m
et
er
s
in
a
m
an
n
er
t
h
at
t
h
e
tr
ai
n
ed
A
NN
w
ill
ac
cu
r
atel
y
m
atc
h
th
e
e
m
p
lo
y
e
d
d
ata.
d)
A
N
N
ca
n
estab
li
s
h
th
e
h
id
d
en
n
o
n
–
lin
ea
r
i
n
p
u
t
-
o
u
tp
u
t
d
ata
r
elatio
n
s
h
ip
a
n
d
th
is
is
s
u
itab
le
f
o
r
h
eter
o
g
e
n
eo
u
s
s
ce
n
ar
io
s
as c
o
m
m
o
n
l
y
ex
p
er
ie
n
ce
d
in
o
il &
g
as r
eser
v
o
ir
s
.
e)
A
N
N
ca
n
ac
c
u
r
atel
y
g
en
er
aliz
e
o
v
er
a
r
an
g
e
o
f
i
n
p
u
t d
ata
o
w
i
n
g
to
its
s
el
f
-
ad
ap
tab
ilit
y
.
f)
A
N
Ns
ca
n
r
aid
l
y
p
r
o
ce
s
s
d
ata
an
d
s
er
v
e
a
s
a
n
ea
s
y
w
a
y
o
f
ap
p
ly
i
n
g
a
n
alr
ea
d
y
ex
i
s
ti
n
g
m
o
d
el
to
a
n
e
w
s
y
s
te
m
.
5.
CO
NCLU
SI
O
N
A
N
N
h
a
s
v
ar
io
u
s
s
tr
u
ct
u
r
es
a
n
d
ea
ch
o
f
t
h
e
s
e
s
tr
u
c
tu
r
es
is
d
esig
n
ed
to
ad
d
r
ess
a
s
p
ec
if
ic
task
.
T
h
er
e
ar
e
th
r
ee
la
y
er
s
in
t
h
e
b
asic
ar
ch
itect
u
r
e
o
f
A
NN
(
in
p
u
t
la
y
e
r
,
h
id
d
en
la
y
er
,
an
d
o
u
tp
u
t
la
y
er
)
an
d
ea
ch
la
y
er
h
as
a
d
f
f
er
e
n
t
f
u
n
ct
io
n
i
n
t
h
e
n
et
w
o
r
k
s
y
s
t
e
m
.
T
h
e
f
n
al
o
u
tp
u
t
o
f
th
e
A
NN
i
s
ca
lc
u
l
ated
b
y
th
e
n
e
u
r
o
n
s
co
n
tain
ed
i
n
th
e
o
u
tp
u
t
la
y
er
as
th
e
y
h
a
v
e
th
e
ac
ti
v
atio
n
f
u
n
ctio
n
s
r
eq
u
ir
ed
f
o
r
th
i
s
r
o
le.
T
h
e
co
n
n
ec
tio
n
b
et
w
ee
n
th
e
n
o
d
es
in
a
n
A
N
N
m
o
d
el
is
m
ed
iated
b
y
th
e
s
y
n
ap
tic
w
e
ig
h
ts
w
h
ic
h
ar
e
r
an
d
o
m
l
y
as
s
i
g
n
ed
an
d
u
p
d
ated
d
u
r
in
g
t
h
e
lear
n
i
n
g
p
er
io
d
u
s
i
n
g
an
y
o
f
th
e
ex
i
s
ti
n
g
tr
ain
i
n
g
al
g
o
r
ith
m
s
.
T
h
is
s
t
u
d
y
p
r
i
m
ar
il
y
ai
m
s
to
p
er
f
o
r
m
s
en
s
iti
v
it
y
an
al
y
s
is
u
s
in
g
t
h
e
ANN
m
o
d
els
f
o
r
th
e
s
ak
e
o
f
e
x
a
m
in
i
n
g
th
e
co
n
s
is
t
en
c
y
o
f
t
h
e
m
o
d
el
’
s
p
er
f
o
r
m
a
n
ce
w
i
th
k
n
o
w
n
o
i
l
s
an
d
s
e
x
tr
ac
tio
n
b
e
h
av
io
r
.
I
t
also
p
r
o
v
id
es
a
lev
el
o
f
co
n
f
id
e
n
ce
in
t
h
e
p
er
f
o
r
m
a
n
ce
o
f
th
e
A
NN
ap
p
r
o
ac
h
.
Mo
r
e
o
v
er
,
A
NN
m
o
d
els
h
av
e
b
ee
n
p
r
o
v
en
s
u
cc
e
s
s
f
u
l
i
n
ap
p
licatio
n
s
w
er
e
t
h
e
s
ta
n
d
ar
d
s
tatis
tical
l
in
ea
r
r
eg
r
ess
io
n
m
e
th
o
d
s
h
ad
n
o
t
b
ee
n
ab
le
to
id
en
ti
f
y
t
h
e
ex
is
te
n
ce
o
f
t
h
e
ex
p
ec
ted
r
elatio
n
s
h
ip
b
et
w
ee
n
th
e
m
aj
o
r
in
p
u
t p
ar
a
m
eter
s
f
r
o
m
t
h
e
e
x
tr
ac
tio
n
d
atab
ase.
RE
F
E
R
E
NC
E
S
[1
]
D.
J.
Am
it
a
n
d
D.
J.
Am
it
,
M
o
d
e
li
n
g
b
ra
i
n
fu
n
c
ti
o
n
:
T
h
e
wo
rl
d
o
f
a
tt
ra
c
to
r
n
e
u
ra
l
n
e
tw
o
rk
s
,
Ca
m
b
r
id
g
e
u
n
iv
e
rsity
p
re
ss
,
1
9
9
2
.
[2
]
H.
M
.
T
.
A
l
-
Hilf
i,
M
.
T
.
M
.
A
l
Ha
sh
im
i,
a
n
d
I.
M
a
rg
h
e
sc
u
,
"
W
iM
A
X
a
n
d
W
iF
i
n
e
tw
o
rk
in
teg
ra
ti
o
n
a
s
b
a
c
k
h
a
u
l
o
n
e
m
e
rg
e
n
c
y
c
o
m
m
u
n
ica
ti
o
n
s,"
in
2
0
1
8
1
4
th
I
n
ter
n
a
t
io
n
a
l
Co
n
fer
e
n
c
e
o
n
Ad
v
a
n
c
e
d
T
re
n
d
s
in
R
a
d
io
e
lec
rtro
n
ics
,
T
e
lec
o
mm
u
n
ica
ti
o
n
s
a
n
d
C
o
mp
u
t
e
r E
n
g
i
n
e
e
rin
g
(
T
C
S
ET
)
,
p
p
.
1
0
6
3
-
1
0
6
8
,
2
0
1
8
.
[3
]
K.
Ba
y
o
u
d
h
,
"
F
ro
m
M
a
c
h
in
e
L
e
a
rn
in
g
T
o
De
e
p
L
e
a
rn
in
g
,
"
e
d
:
Eb
o
o
k
,
2
0
1
7
.
[4
]
R.
A
.
Ha
s
a
n
,
M
.
A
.
M
o
h
a
m
m
e
d
,
Z.
H.
S
a
li
h
,
M
.
A
.
B.
Am
e
e
d
e
e
n
,
N.
Ţ
ă
p
u
ş,
a
n
d
M
.
N.
M
o
h
a
m
m
e
d
,
"
HSO:
A
H
y
b
rid
S
w
a
r
m
Op
ti
m
iza
ti
o
n
A
l
g
o
rit
h
m
f
o
r
R
e
d
u
c
in
g
En
e
rg
y
Co
n
su
m
p
ti
o
n
in
th
e
Clo
u
d
lets,"
T
e
lko
mn
ika
,
v
o
l.
1
6
,
p
p
.
2
1
4
4
-
2
1
5
4
,
2
0
1
8
.
[5
]
R.
A
.
H
a
sa
n
,
M
.
A
.
M
o
h
a
m
m
e
d
,
N.
Ţ
ă
p
u
ş,
a
n
d
O.
A
.
Ha
m
m
o
o
d
,
"
A
c
o
m
p
re
h
e
n
siv
e
stu
d
y
:
A
n
t
Co
lo
n
y
Op
ti
m
iza
ti
o
n
(A
C
O)
f
o
r
f
a
c
il
it
y
la
y
o
u
t
p
ro
b
lem
,
"
in
2
0
1
7
1
6
t
h
R
o
Ed
u
Ne
t
Co
n
fer
e
n
c
e
:
Ne
two
rk
in
g
in
Ed
u
c
a
ti
o
n
a
n
d
Res
e
a
rc
h
(
Ro
Ed
u
Ne
t)
,
2
0
1
7
,
p
p
.
1
-
8.
[6
]
R.
A
.
Ha
sa
n
a
n
d
M
.
N.
M
o
h
a
m
m
e
d
,
"
A
k
ril
l
h
e
rd
b
e
h
a
v
io
u
r
i
n
sp
ir
e
d
lo
a
d
b
a
lan
c
in
g
o
f
tas
k
s
in
c
lo
u
d
c
o
m
p
u
ti
n
g
,
"
S
tu
d
ies
in
In
f
o
rm
a
ti
c
s a
n
d
Co
n
tro
l,
v
o
l.
2
6
,
p
p
.
4
1
3
-
4
2
4
,
2
0
1
7
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2502
-
475
2
I
n
d
o
n
esia
n
J
E
lec
E
n
g
&
C
o
m
p
Sci,
Vo
l.
21
,
No
.
1
,
J
an
u
ar
y
2
0
2
1
:
558
-
565
564
[7
]
M
.
A
.
M
o
h
a
m
m
e
d
a
n
d
R.
A
.
Ha
s
a
n
,
"
P
a
rti
c
le
sw
a
r
m
o
p
ti
m
iz
a
ti
o
n
f
o
r
f
a
c
il
it
y
la
y
o
u
t
p
r
o
b
lem
s
F
L
P
-
A
c
o
m
p
re
h
e
n
siv
e
stu
d
y
,
"
in
2
0
1
7
1
3
th
I
EE
E
I
n
ter
n
a
t
io
n
a
l
C
o
n
fer
e
n
c
e
o
n
In
tell
ig
e
n
t
Co
mp
u
ter
Co
m
mu
n
ica
t
io
n
a
n
d
Pro
c
e
ss
in
g
(
ICCP)
,
2
0
1
7
,
p
p
.
9
3
-
99.
[8
]
M
.
A
.
M
o
h
a
m
m
e
d
,
R.
A
.
Ha
s
a
n
,
M
.
A
.
A
h
m
e
d
,
N.
T
a
p
u
s,
M
.
A
.
S
h
a
n
a
n
,
M
.
K
.
Kh
a
lee
l
,
e
t
a
l
.
,
"
A
F
o
c
a
l
lo
a
d
b
a
lan
c
e
r
b
a
se
d
a
lg
o
rit
h
m
f
o
r
tas
k
a
ss
i
g
n
m
e
n
t
in
c
lo
u
d
e
n
v
iro
n
m
e
n
t,
"
in
2
0
1
8
1
0
th
I
n
ter
n
a
t
io
n
a
l
Co
n
fer
e
n
c
e
o
n
El
e
c
tro
n
ics
,
Co
m
p
u
ter
s a
n
d
Arti
fi
c
ia
l
In
tell
ig
e
n
c
e
(
ECA
I)
,
2
0
1
8
,
p
p
.
1
-
4.
[9
]
M
.
A
.
M
o
h
a
m
m
e
d
,
A
.
A
.
Ka
m
il
,
R.
A
.
Ha
sa
n
,
a
n
d
N.
T
a
p
u
s,
"
A
n
Ef
fe
c
ti
v
e
Co
n
tex
t
S
e
n
siti
v
e
Off
l
o
a
d
in
g
S
y
ste
m
f
o
r
M
o
b
il
e
Clo
u
d
En
v
iro
n
m
e
n
ts
u
sin
g
S
u
p
p
o
rt
V
a
lu
e
-
b
a
se
d
Clas
sif
ic
a
ti
o
n
,
"
S
c
a
la
b
le
Co
mp
u
ti
n
g
:
Pra
c
ti
c
e
a
n
d
Exp
e
rie
n
c
e
,
v
o
l.
2
0
,
p
p
.
6
8
7
-
6
9
8
,
2
0
1
9
.
[1
0
]
M
.
A
.
M
o
h
a
m
m
e
d
,
I.
A
.
M
o
h
a
m
m
e
d
,
R.
A
.
Ha
sa
n
,
N.
Ţ
ă
p
u
ş,
A
.
H.
A
li
,
a
n
d
O.
A
.
Ha
m
m
o
o
d
,
"
G
r
e
e
n
En
e
rg
y
S
o
u
rc
e
s:
Iss
u
e
s
a
n
d
C
h
a
ll
e
n
g
e
s,"
in
2
0
1
9
1
8
th
Ro
E
d
u
Ne
t
Co
n
fe
re
n
c
e
:
Ne
two
rk
in
g
in
Ed
u
c
a
ti
o
n
a
n
d
Res
e
a
rc
h
(
Ro
Ed
u
Ne
t)
,
2
0
1
9
,
p
p
.
1
-
8.
[1
1
]
M
.
A
.
A
h
m
e
d
,
R.
A
.
Ha
sa
n
,
A
.
H.
A
li
,
a
n
d
M
.
A
.
M
o
h
a
m
m
e
d
,
"
T
h
e
c
l
a
ss
i
f
ica
ti
o
n
o
f
th
e
m
o
d
e
rn
a
ra
b
ic
p
o
e
try
u
sin
g
m
a
c
h
in
e
lea
rn
in
g
,
"
T
e
lko
mn
ika
,
v
o
l.
1
7
,
p
p
.
2
6
6
7
-
2
6
7
4
,
2
0
1
9
.
[1
2
]
A
.
H.
A
li
,
"
A
S
u
rv
e
y
o
n
V
e
rti
c
a
l
a
n
d
Ho
rizo
n
tal
S
c
a
li
n
g
P
latf
o
rm
s
f
o
r
Big
Da
ta
A
n
a
l
y
ti
c
s,
"
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
In
teg
ra
ted
En
g
i
n
e
e
rin
g
,
v
o
l.
1
1
,
p
p
.
1
3
8
-
1
5
0
,
2
0
1
9
.
[1
3
]
O.
A
.
Ha
m
m
o
o
d
,
M
.
N.
M
.
Ka
h
a
r,
W
.
A
.
Ha
m
m
o
o
d
,
R.
A
.
Ha
s
a
n
,
M
.
A
.
M
o
h
a
m
m
e
d
,
A
.
A
.
Y
o
o
b
,
e
t
a
l.
,
"
A
n
e
ffe
c
ti
v
e
tran
s
m
it
p
a
c
k
e
t
c
o
d
in
g
w
it
h
tru
st
-
b
a
se
d
re
la
y
n
o
d
e
s
in
VA
NE
T
s,
"
Bu
ll
e
ti
n
o
f
El
e
c
trica
l
E
n
g
i
n
e
e
rin
g
a
n
d
In
fo
rm
a
t
ics
,
v
o
l.
9
,
p
p
.
6
8
5
-
6
9
7
,
2
0
2
0
.
[1
4
]
O.
A
.
Ha
m
m
o
o
d
,
N.
Niz
a
m
,
M
.
Na
fa
a
,
a
n
d
W
.
A
.
Ha
m
m
o
o
d
,
"
R
ES
P
:
Re
lay
S
u
it
a
b
il
it
y
-
b
a
se
d
Ro
u
ti
n
g
P
r
o
to
c
o
l
f
o
r
V
id
e
o
S
trea
m
in
g
in
V
e
h
icu
lar
A
d
Ho
c
Ne
t
w
o
rk
s,"
In
ter
n
a
ti
o
n
a
l
J
o
u
r
n
a
l
o
f
C
o
mp
u
ter
s,
Co
m
mu
n
ica
t
io
n
s
&
Co
n
tro
l,
v
o
l.
1
4
,
2
0
1
9
.
[1
5
]
M
.
A
.
M
o
h
a
m
m
e
d
,
Z.
H.
S
a
li
h
,
N.
Ţ
ă
p
u
ş,
a
n
d
R.
A
.
K.
Ha
sa
n
,
"
S
e
c
u
rit
y
a
n
d
a
c
c
o
u
n
tab
il
i
ty
f
o
r
sh
a
rin
g
th
e
d
a
ta
sto
re
d
i
n
t
h
e
c
lo
u
d
,
"
in
2
0
1
6
1
5
t
h
Ro
Ed
u
Ne
t
Co
n
fer
e
n
c
e
:
Ne
two
rk
in
g
i
n
E
d
u
c
a
ti
o
n
a
n
d
Res
e
a
rc
h
,
2
0
1
6
,
p
p
.
1
-
5.
[1
6
]
M
.
A
.
M
o
h
a
m
m
e
d
a
n
d
N.
ŢĂ
P
UŞ,
"
A
No
v
e
l
A
p
p
ro
a
c
h
o
f
Re
d
u
c
i
n
g
En
e
rg
y
Co
n
su
m
p
ti
o
n
b
y
Util
iz
in
g
En
th
a
l
p
y
in
M
o
b
i
le Cl
o
u
d
C
o
m
p
u
ti
n
g
,
"
S
t
u
d
i
e
s in
In
f
o
rm
a
ti
c
s
a
n
d
Co
n
tro
l
,
v
o
l
.
2
6
,
p
p
.
4
2
5
-
4
3
4
,
2
0
1
7
.
[1
7
]
N.
Q.
M
o
h
a
m
m
e
d
,
M
.
S
.
A
h
m
e
d
,
M
.
A
.
M
o
h
a
m
m
e
d
,
O.
A
.
H
a
m
m
o
o
d
,
H.
A
.
N.
A
lsh
a
ra
,
a
n
d
A
.
A
.
Ka
m
il
,
"
Co
m
p
a
r
a
ti
v
e
A
n
a
l
y
sis
b
e
t
w
e
e
n
S
o
lar
a
n
d
W
in
d
T
u
rb
in
e
En
e
r
g
y
S
o
u
rc
e
s
in
Io
T
Ba
s
e
d
o
n
Eco
n
o
m
ica
l
a
n
d
Eff
icie
n
c
y
Co
n
sid
e
ra
ti
o
n
s,"
in
2
0
1
9
2
2
n
d
I
n
ter
n
a
ti
o
n
a
l
C
o
n
fer
e
n
c
e
o
n
Co
n
tro
l
S
y
ste
ms
a
n
d
Co
mp
u
ter
S
c
ien
c
e
(
CS
CS
)
,
2
0
1
9
,
p
p
.
4
4
8
-
4
5
2
.
[1
8
]
M.
A
.
A
.
Ro
y
id
a
A
.
Ib
ra
h
e
m
A
l
h
a
y
a
li
,
Ya
s
m
in
M
a
k
k
i
M
o
h
iald
e
n
,
A
h
m
e
d
H.
A
li
,
"
E
ff
icie
n
t
m
e
t
h
o
d
f
o
r
b
re
a
st
c
a
n
c
e
r
c
las
sif
i
c
a
ti
o
n
b
a
se
d
o
n
e
n
se
m
b
le
h
o
f
fe
d
in
g
tree
a
n
d
n
a
ïv
e
Ba
y
e
s,
"
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
8
,
p
p
.
1
0
74
-
1
0
8
0
,
2
0
2
0
.
[1
9
]
Z.
H.
S
a
li
h
,
G
.
T
.
H
a
sa
n
,
a
n
d
M
.
A
.
M
o
h
a
m
m
e
d
,
"
In
v
e
stig
a
te
a
n
d
a
n
a
l
y
z
e
th
e
le
v
e
ls
o
f
e
lec
tro
m
a
g
n
e
ti
c
ra
d
iatio
n
s
e
m
it
ted
f
ro
m
u
n
d
e
rg
ro
u
n
d
p
o
w
e
r
c
a
b
les
e
x
ten
d
e
d
in
m
o
d
e
rn
c
it
ies
,
"
in
2
0
1
7
9
t
h
In
ter
n
a
t
io
n
a
l
Co
n
fer
e
n
c
e
o
n
El
e
c
tro
n
ics
,
Co
m
p
u
ter
s
a
n
d
Arti
fi
c
ia
l
In
tell
ig
e
n
c
e
(
ECA
I)
,
2
0
1
7
.
[2
0
]
Z.
H.
S
a
li
h
,
G
.
T
.
Ha
sa
n
,
M
.
A
.
M
o
h
a
m
m
e
d
,
M
.
A
.
S
.
Kli
b
,
A
.
H.
A
li
,
a
n
d
R.
A
.
Ib
ra
h
im
,
"
S
tu
d
y
th
e
Eff
e
c
t
o
f
In
teg
ra
ti
n
g
th
e
S
o
lar
E
n
e
rg
y
S
o
u
rc
e
o
n
S
tab
il
it
y
o
f
El
e
c
tri
c
a
l
Distrib
u
t
io
n
S
y
ste
m
,
"
in
2
0
1
9
2
2
n
d
In
ter
n
a
t
i
o
n
a
l
Co
n
fer
e
n
c
e
o
n
Co
n
tro
l
S
y
ste
ms
a
n
d
C
o
mp
u
ter
S
c
ien
c
e
(
CS
CS
)
,
2
0
1
9
,
p
p
.
4
4
3
-
4
4
7
.
[2
1
]
Y.
Bo
d
y
a
n
sk
i
y
,
A
.
Do
lo
to
v
,
I.
P
l
iss,
a
n
d
M
.
M
a
ly
a
r,
"
A
fa
st
le
a
rn
in
g
a
lg
o
rit
h
m
o
f
se
l
f
-
lea
rn
in
g
sp
ik
in
g
n
e
u
ra
l
n
e
tw
o
rk
,
"
in
2
0
1
6
IEE
E
Fi
rs
t
I
n
t
e
rn
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
Da
t
a
S
tre
a
m
M
in
in
g
&
Pro
c
e
ss
in
g
(
DS
M
P)
,
2
0
1
6
,
p
p
.
104
-
1
0
7
.
[2
2
]
D.
T
.
Bu
i,
P
.
T
sa
n
g
a
ra
to
s,
V
.
-
T
.
Ng
u
y
e
n
,
N.
V
.
L
iem
,
a
n
d
P
.
T
.
T
rin
h
,
"
Co
m
p
a
rin
g
th
e
p
re
d
ictio
n
p
e
rf
o
r
m
a
n
c
e
o
f
a
De
e
p
Lea
rn
in
g
Ne
u
ra
l
Ne
t
w
o
rk
m
o
d
e
l
w
it
h
c
o
n
v
e
n
ti
o
n
a
l
m
a
c
h
in
e
lea
rn
in
g
m
o
d
e
ls
in
lan
d
slid
e
su
sc
e
p
ti
b
il
it
y
a
ss
e
ss
m
e
n
t,
"
CAT
ENA
,
v
o
l.
1
8
8
,
p
p
.
1
0
4
4
2
6
,
2
0
2
0
.
[2
3
]
C.
Bish
o
p
,
"
Ne
u
ra
l
n
e
tw
o
rk
s f
o
r
p
a
tt
e
rn
re
c
o
g
n
it
i
o
n
Ox
f
o
rd
Un
iv
e
rsity
P
re
ss
Ox
f
o
rd
G
o
o
g
le S
c
h
o
la
r,
"
1
9
9
5
.
[2
4
]
S
.
v
a
n
Cra
n
e
n
b
u
rg
h
a
n
d
A
.
A
l
w
o
sh
e
e
l,
"
A
n
a
rti
f
icia
l
n
e
u
ra
l
n
e
tw
o
rk
b
a
se
d
a
p
p
ro
a
c
h
to
i
n
v
e
stig
a
te
trav
e
ll
e
rs’
d
e
c
isio
n
r
u
les
,
"
T
ra
n
s
p
o
rt
a
ti
o
n
R
e
se
a
rc
h
Pa
rt C:
Eme
rg
i
n
g
T
e
c
h
n
o
lo
g
ies
,
v
o
l.
9
8
,
p
p
.
1
5
2
-
1
6
6
,
2
0
1
9
.
[2
5
]
D.
V
e
n
k
a
tes
a
n
,
K.
Ka
n
n
a
n
,
a
n
d
R.
S
a
ra
v
a
n
a
n
,
"
A
g
e
n
e
ti
c
a
l
g
o
rit
h
m
-
b
a
se
d
a
rti
f
icia
l
n
e
u
ra
l
n
e
tw
o
r
k
m
o
d
e
l
f
o
r
th
e
o
p
ti
m
iza
ti
o
n
o
f
m
a
c
h
in
in
g
p
r
o
c
e
ss
e
s,"
Ne
u
ra
l
Co
mp
u
ti
n
g
a
n
d
Ap
p
l
ica
ti
o
n
s,
v
o
l.
1
8
,
p
p
.
1
3
5
-
1
4
0
,
2
0
0
9
.
[2
6
]
X
.
W
a
n
g
,
X
.
L
in
,
a
n
d
X
.
Da
n
g
,
"
S
u
p
e
rv
ise
d
lea
rn
in
g
in
sp
ik
in
g
n
e
u
ra
l
n
e
tw
o
rk
s:
A
re
v
ie
w
o
f
a
lg
o
rit
h
m
s
a
n
d
e
v
a
lu
a
ti
o
n
s,"
Ne
u
ra
l
Ne
two
rk
s,
v
o
l.
1
2
5
,
p
p
.
2
5
8
-
2
8
0
,
2
0
2
0
.
[2
7
]
Y.
W
a
n
g
,
"
A
M
a
c
h
in
e
-
L
e
a
rn
in
g
M
o
d
e
l
Us
in
g
A
rti
f
icia
l
Ne
u
ra
l
Ne
tw
o
rk
to
F
a
c
il
it
a
te
L
iv
e
r
S
BRT
P
re
sc
rip
ti
o
n
S
e
lec
ti
o
n
By
P
re
d
ictin
g
N
o
rm
a
l
L
iv
e
r
G
e
u
d
Ba
se
d
o
n
G
e
o
m
e
tr
ic
P
ro
p
e
rti
e
s
o
f
L
iv
e
r
a
n
d
P
T
V
,
"
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
R
a
d
i
a
ti
o
n
O
n
c
o
lo
g
y
*
B
io
lo
g
y
*
P
h
y
sic
s,
v
o
l.
1
0
5
,
p
p
.
E1
3
9
,
2
0
1
9
.
[2
8
]
X
.
Ya
o
,
"
A
re
v
ie
w
o
f
e
v
o
lu
ti
o
n
a
ry
a
rti
f
ici
a
l
n
e
u
ra
l
n
e
tw
o
rk
s,"
In
ter
n
a
ti
o
n
a
l
jo
u
rn
a
l
o
f
in
tell
ig
e
n
t
sy
ste
ms
,
v
o
l.
8
,
p
p
.
5
3
9
-
5
6
7
,
1
9
9
3
.
[2
9
]
J.
R.
Bu
se
m
e
y
e
r,
E.
By
u
n
,
E.
L
.
De
lo
sh
,
a
n
d
M
.
A
.
M
c
Da
n
iel,
"
L
e
a
rn
in
g
f
u
n
c
ti
o
n
a
l
re
latio
n
s
b
a
se
d
o
n
e
x
p
e
rien
c
e
w
it
h
in
p
u
t
-
o
u
tp
u
t
p
a
irs
b
y
h
u
m
a
n
s an
d
a
rti
f
icia
l
n
e
u
ra
l
n
e
tw
o
rk
s,"
1
9
9
7
.
[3
0
]
A
.
Ya
stre
b
o
v
,
S
.
G
a
d
,
G
.
S
lo
n
,
A
.
Ku
lak
o
w
sk
i,
a
n
d
E.
V
in
o
g
ra
d
o
v
a
,
"
A
rti
f
ici
a
l
Ne
u
ra
l
Ne
tw
o
rk
s
in
R
u
le
-
b
a
se
d
De
c
isio
n
S
y
ste
m
s,
"
in
EUROCO
N
2
0
0
7
-
T
h
e
In
ter
n
a
ti
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
"
C
o
mp
u
ter
a
s
a
T
o
o
l"
,
2
0
0
7
,
p
p
.
6
8
6
-
6
9
1
.
[3
1
]
A
.
Zh
a
n
g
,
H.
Zh
o
u
,
X.
L
i,
a
n
d
W
.
Zh
u
,
"
F
a
st
a
n
d
ro
b
u
st
lea
rn
in
g
i
n
S
p
ik
in
g
F
e
e
d
-
f
o
rw
a
rd
Ne
u
ra
l
Ne
tw
o
rk
s
b
a
se
d
o
n
I
n
tri
n
sic
P
las
ti
c
it
y
m
e
c
h
a
n
is
m
,
"
Ne
u
ro
c
o
mp
u
ti
n
g
,
v
o
l.
3
6
5
,
p
p
.
1
0
2
-
1
1
2
,
2
0
1
9
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esia
n
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2502
-
4752
Live
to
lea
r
n
:
Lea
r
n
in
g
r
u
les
-
b
a
s
ed
a
r
tifi
cia
l n
eu
r
a
l n
etw
o
r
k
(
A
s
ee
l S
h
a
kir I
.
Hila
iw
a
h
)
565
[3
2
]
T
.
Ej
im
a
,
Y.
Ka
to
,
Y.
T
a
n
,
a
n
d
K
.
Ish
ig
a
k
i,
"
Ch
a
ra
c
ter
Re
c
o
g
n
it
io
n
By
A
Ne
u
ra
l
Ne
t
w
o
rk
W
it
h
F
u
z
z
y
P
a
rti
ti
o
n
i
n
g
Un
it
s
,
"
in
Arti
fi
c
ia
l
Ne
u
ra
l
Ne
tw
o
rk
s
,
T
.
Ko
h
o
n
e
n
,
K.
M
Ä
Kisa
ra
,
O.
S
im
u
la,
a
n
d
J.
Ka
n
g
a
s,
Ed
s.,
e
d
Am
ste
rd
a
m
:
No
rth
-
Ho
ll
a
n
d
,
1
9
9
1
,
p
p
.
1
7
7
7
-
1
7
8
0
.
[3
3
]
R.
N.
Na
g
u
ib
a
n
d
G
.
V
.
S
h
e
rb
e
t,
Arti
fi
c
ia
l
n
e
u
ra
l
n
e
two
rk
s
in
c
a
n
c
e
r
d
ia
g
n
o
sis,
p
ro
g
n
o
si
s,
a
n
d
p
a
ti
e
n
t
ma
n
a
g
e
me
n
t
:
CRC
p
re
ss
,
2
0
0
1
.
[3
4
]
G
.
Ba
k
ırl
ı,
"
A
n
in
c
re
m
e
n
tal
g
e
n
e
ti
c
a
lg
o
rit
h
m
a
n
d
n
e
u
ra
l
n
e
tw
o
rk
fo
r
c
las
sif
ic
a
ti
o
n
a
n
d
se
n
siti
v
it
y
a
n
a
ly
sis
o
f
th
e
ir
p
a
ra
m
e
ters
,
"
2
0
0
9
.
[3
5
]
A
.
H.
A
li
a
n
d
M
.
Z
.
A
b
d
u
l
lah
,
"
Re
c
e
n
t
tren
d
s
in
d
istr
ib
u
ted
o
n
li
n
e
stre
a
m
p
ro
c
e
ss
in
g
p
latf
o
rm
f
o
r
b
ig
d
a
ta:
S
u
rv
e
y
,
"
in
2
0
1
8
1
st A
n
n
u
a
l
In
ter
n
a
ti
o
n
a
l
C
o
n
fer
e
n
c
e
o
n
In
fo
rm
a
t
i
o
n
a
n
d
S
c
ie
n
c
e
s (
Ai
CIS
)
,
2
0
1
8
,
p
p
.
1
4
0
-
1
4
5
.
[3
6
]
A
.
H.
A
li
a
n
d
M
.
Z.
A
b
d
u
ll
a
h
,
"
A
n
o
v
e
l
a
p
p
ro
a
c
h
f
o
r
b
ig
d
a
ta cla
ss
if
ica
ti
o
n
b
a
se
d
o
n
h
y
b
rid
p
a
ra
ll
e
l
d
im
e
n
sio
n
a
li
ty
re
d
u
c
ti
o
n
u
si
n
g
sp
a
rk
c
lu
ste
r,
"
Co
mp
u
ter
S
c
ien
c
e
,
v
o
l
.
2
0
,
2
0
1
9
.
[3
7
]
M
.
A
.
H.
A
li
,
"
A
n
Eff
ici
e
n
t
M
o
d
e
l
f
o
r
D
a
ta
Clas
sif
ic
a
ti
o
n
Ba
se
d
o
n
S
V
M
G
rid
P
a
ra
m
e
ter
Op
ti
m
iz
a
ti
o
n
a
n
d
P
S
O
F
e
a
tu
re
W
e
ig
h
t
S
e
lec
ti
o
n
,
"
In
ter
n
a
ti
o
n
a
l
J
o
u
r
n
a
l
o
f
In
teg
r
a
ted
En
g
in
e
e
rin
g
,
2
0
1
8
.
[3
8
]
A
.
Esp
o
sito
,
S
.
Ra
m
p
o
n
e
,
a
n
d
R.
T
a
g
li
a
fe
rri,
"
A
d
d
it
io
n
a
n
d
su
b
trac
t
io
n
i
n
n
e
u
ra
l
n
e
ts
a
s
re
su
lt
s
o
f
a
le
a
rn
in
g
p
ro
c
e
ss
T
h
is
w
o
rk
wa
s
su
p
p
o
rted
in
p
a
rt
b
y
CNR,
P
ro
g
e
tt
o
F
i
n
a
li
z
z
a
to
“
S
istem
i
In
f
o
r
m
a
ti
c
i
e
Ca
lc
o
lo
P
a
ra
ll
e
lo
”
,
b
y
M
P
I
4
0
%
a
n
d
b
y
II
A
S
S
,
"
in
Arti
fi
c
ia
l
Ne
u
ra
l
Ne
tw
o
rk
s
,
T
.
Ko
h
o
n
e
n
,
K.
M
Ä
Kisa
ra
,
O.
S
im
u
la,
a
n
d
J.
Ka
n
g
a
s,
Ed
s.,
e
d
Am
ste
rd
a
m
:
No
rth
-
Ho
ll
a
n
d
,
1
9
9
1
,
p
p
.
1
7
8
9
-
1
7
9
2
.
[3
9
]
C.
He
,
M
.
M
a
,
a
n
d
P
.
W
a
n
g
,
"
Ex
trac
t
in
terp
re
tab
il
it
y
-
a
c
c
u
ra
c
y
b
a
lan
c
e
d
ru
les
f
ro
m
a
rti
f
icia
l
n
e
u
ra
l
n
e
tw
o
rk
s:
A
re
v
ie
w
,
"
N
e
u
ro
c
o
mp
u
t
in
g
,
v
o
l
.
3
8
7
,
p
p
.
3
4
6
-
3
5
8
,
2
0
2
0
.
[4
0
]
S
.
Ha
n
,
H.
M
a
o
,
a
n
d
W
.
J.
Da
ll
y
,
"
D
e
e
p
c
o
m
p
re
ss
io
n
:
Co
m
p
re
s
sin
g
d
e
e
p
n
e
u
ra
l
n
e
tw
o
rk
s
w
it
h
p
ru
n
in
g
,
train
e
d
q
u
a
n
t
iza
ti
o
n
a
n
d
h
u
f
fm
a
n
c
o
d
in
g
,
"
a
rXiv p
re
p
rin
t
a
rXiv:
1
5
1
0
.
0
0
1
4
9
,
2
0
1
5
.
[4
1
]
M
.
Ish
ik
a
w
a
,
"
N
e
u
ra
l
n
e
tw
o
rk
s
a
p
p
ro
a
c
h
t
o
r
u
le
e
x
trac
ti
o
n
,
"
in
Pr
o
c
e
e
d
in
g
s
1
9
9
5
S
e
c
o
n
d
Ne
w
Z
e
a
la
n
d
In
t
e
rn
a
t
io
n
a
l
T
wo
-
S
tre
a
m Co
n
fer
e
n
c
e
o
n
Arti
fi
c
ia
l
Ne
u
ra
l
Ne
tw
o
rk
s a
n
d
Exp
e
rt
S
y
ste
ms
,
1
9
9
5
,
p
p
.
6
-
9.
[4
2
]
S
.
H.
Kim
,
S
.
G
.
S
h
in
,
S
.
Ha
n
,
M
.
H.
Ki
m
,
a
n
d
C.
H.
P
y
e
o
n
,
"
F
e
a
s
ib
il
it
y
stu
d
y
o
n
a
p
p
li
c
a
ti
o
n
o
f
a
n
a
rti
f
icia
l
n
e
u
ra
l
n
e
tw
o
rk
f
o
r
a
u
to
m
a
ti
c
d
e
sig
n
o
f
a
re
a
c
to
r
c
o
re
a
t
th
e
K
y
o
to
Un
iv
e
rsit
y
Crit
ica
l
A
ss
e
m
b
l
y
,
"
Pro
g
re
ss
in
Nu
c
lea
r
En
e
rg
y
,
v
o
l.
1
1
9
,
p
p
.
1
0
3
1
8
3
,
2
0
2
0
.
[4
3
]
M
.
Nic
k
e
l,
K.
M
u
rp
h
y
,
V
.
T
re
s
p
,
a
n
d
E.
G
a
b
ril
o
v
ich
,
"
A
re
v
ie
w
o
f
re
l
a
ti
o
n
a
l
m
a
c
h
in
e
lea
rn
in
g
f
o
r
k
n
o
w
led
g
e
g
ra
p
h
s,"
Pro
c
e
e
d
in
g
s
o
f
t
h
e
IEE
E
,
v
o
l.
1
0
4
,
p
p
.
1
1
-
3
3
,
2
0
1
5
.
[4
4
]
M
.
BİL
G
İ
L
İ,
"
P
re
d
ictio
n
s
o
f
Wi
n
d
S
p
e
e
d
a
n
d
W
in
d
P
o
w
e
r
P
o
ten
ti
a
l
Us
in
g
A
rti
f
icia
l
Ne
u
ra
l
Ne
tw
o
rk
s,"
e
d
:
Do
k
to
ra
T
e
z
i,
Çu
k
u
ro
v
a
Ün
iv
e
rsites
i,
F
e
n
Bil
im
leri
En
stit
ü
sü
,
2
0
0
7
.
[4
5
]
H.
Og
i,
H.
T
a
n
a
k
a
,
Y.
A
k
i
m
o
to
,
a
n
d
Y.
Iz
u
i,
"
A
P
P
RO
X
IM
A
T
IO
N
CA
P
A
BI
L
I
T
IES
OF
NEURA
L
NETW
ORK
S
USING
S
A
M
P
L
IN
G
F
UN
CT
IO
NS,
"
in
Arti
fi
c
i
a
l
Ne
u
ra
l
Ne
two
r
k
s
,
T
.
Ko
h
o
n
e
n
,
K.
M
Ä
Kisa
ra
,
O
.
S
im
u
la,
a
n
d
J.
Ka
n
g
a
s,
Ed
s.,
e
d
A
m
st
e
rd
a
m
:
No
rth
-
Ho
ll
a
n
d
,
1
9
9
1
,
p
p
.
1
7
8
5
-
1
7
8
8
.
[4
6
]
K.
P
a
tan
a
n
d
M
.
P
a
tan
,
"
Ne
u
ra
l
-
n
e
tw
o
rk
-
b
a
se
d
it
e
ra
ti
v
e
lea
rn
in
g
c
o
n
tro
l
o
f
n
o
n
li
n
e
a
r
sy
st
e
m
s,"
IS
A
T
ra
n
s
a
c
ti
o
n
s,
v
o
l.
9
8
,
p
p
.
4
4
5
-
4
5
3
,
2
0
2
0
.
[4
7
]
J.
S
c
h
m
id
h
u
b
e
r,
"
De
e
p
lea
rn
in
g
i
n
n
e
u
ra
l
n
e
tw
o
rk
s: A
n
o
v
e
rv
ie
w
,
"
Ne
u
ra
l
n
e
two
rk
s,
v
o
l.
6
1
,
p
p
.
8
5
-
1
1
7
,
2
0
1
5
.
[4
8
]
H.
R.
I
b
ra
h
e
e
m
,
Z.
F
.
H
u
ss
a
in
,
S
.
M
.
A
li
,
M
.
A
lj
a
n
a
b
i,
M
.
A
.
M
o
h
a
m
m
e
d
,
a
n
d
T
.
S
u
ti
k
n
o
,
"
A
n
e
w
m
o
d
e
l
f
o
r
larg
e
d
a
tas
e
t
d
im
e
n
sio
n
a
li
ty
re
d
u
c
ti
o
n
b
a
se
d
o
n
tea
c
h
in
g
lea
rn
in
g
-
b
a
se
d
o
p
t
im
iza
ti
o
n
a
n
d
l
o
g
isti
c
re
g
re
s
sio
n
,
"
T
e
lko
mn
ika
,
v
o
l.
1
8
,
2
0
2
0
.
[4
9
]
N.
D.
Zak
i,
N.
Y.
Ha
sh
im
,
Y.
M
.
M
o
h
iald
e
n
,
M
.
A
.
M
o
h
a
m
m
e
d
,
T
.
S
u
ti
k
n
o
,
a
n
d
A
.
H.
A
li
,
"
A
re
a
l
-
ti
m
e
b
ig
d
a
ta
se
n
ti
m
e
n
t
a
n
a
l
y
sis
f
o
r
iraq
i
tw
e
e
ts
u
sin
g
sp
a
rk
stre
a
m
in
g
,
"
Bu
ll
e
ti
n
o
f
El
e
c
trica
l
En
g
in
e
e
rin
g
a
n
d
I
n
fo
rm
a
ti
c
s,
v
o
l.
9
,
p
p
.
1
4
1
1
-
1
4
1
9
,
2
0
2
0
.
[5
0
]
M
.
S
h
a
rif
z
a
d
e
h
,
A
.
S
ik
in
io
ti
-
L
o
c
k
,
a
n
d
N.
S
h
a
h
,
"
M
a
c
h
in
e
-
lea
r
n
in
g
m
e
th
o
d
s
f
o
r
in
teg
ra
ted
re
n
e
wa
b
le
p
o
w
e
r
g
e
n
e
ra
ti
o
n
:
A
c
o
m
p
a
ra
ti
v
e
stu
d
y
o
f
a
rti
f
icia
l
n
e
u
ra
l
n
e
tw
o
rk
s,
su
p
p
o
rt
v
e
c
to
r
re
g
re
ss
io
n
,
a
n
d
G
a
u
ss
ian
P
r
o
c
e
ss
Re
g
re
ss
io
n
,
"
Ren
e
wa
b
le a
n
d
S
u
st
a
in
a
b
le E
n
e
rg
y
Rev
iews
,
v
o
l.
1
0
8
,
p
p
.
5
1
3
-
5
3
8
,
2
0
1
9
.
[5
1
]
M
.
S
h
a
rm
a
a
n
d
R.
G
a
r
g
,
"
A
n
a
rti
f
icia
l
n
e
u
ra
l
n
e
tw
o
rk
b
a
s
e
d
a
p
p
ro
a
c
h
f
o
r
e
n
e
rg
y
e
ff
icie
n
t
tas
k
s
c
h
e
d
u
li
n
g
in
c
l
o
u
d
d
a
ta ce
n
ters
,
"
S
u
sta
i
n
a
b
le C
o
mp
u
ti
n
g
:
In
f
o
rm
a
ti
c
s a
n
d
S
y
ste
ms
,
v
o
l.
2
6
,
p
.
1
0
0
3
7
3
,
2
0
2
0
.
Evaluation Warning : The document was created with Spire.PDF for Python.