I
AE
S
I
nte
rna
t
io
na
l J
o
urna
l o
f
Art
if
icia
l In
t
ellig
ence
(
I
J
-
AI
)
Vo
l.
9
,
No
.
4
,
Dec
em
b
er
2020
,
p
p
.
609
~
61
5
I
SS
N:
2252
-
8938
,
DOI
: 1
0
.
1
1
5
9
1
/ij
ai.
v
9
.i
4
.
p
p
609
-
61
5
609
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
i
.
ia
esco
r
e.
co
m
A
m
ultilay
er
perc
eptron a
rti
ficial n
eura
l net
w
o
rk
ap
pro
a
ch f
o
r
i
m
pro
v
ing
t
he acc
ura
cy
of int
r
usio
n det
ec
tion sy
ste
m
s
Abdu
lra
h
m
a
n J
a
s
s
a
m
M
o
ha
mm
e
d
1
,
M
uh
a
na
d H
a
m
ee
d
Arif
2
,
a
nd
Ali A
dil
Al
i
3
1,
2
Dire
c
to
ra
te o
f
Ed
u
c
a
ti
o
n
i
n
Diy
a
la,
M
in
istry
o
f
Ed
u
c
a
ti
o
n
,
Diy
a
l
a
,
Ira
q
3
M
in
istry
o
f
Hig
h
e
r
Ed
u
c
a
ti
o
n
a
n
d
S
c
ien
ti
f
ic Res
e
a
rc
h
,
Ba
g
h
d
a
d
,
I
ra
q
Art
icle
I
nfo
AB
ST
RAC
T
A
r
tic
le
his
to
r
y:
R
ec
eiv
ed
A
pr
2
2
,
2
0
20
R
ev
i
s
ed
J
ul
2
7
,
2
0
20
A
cc
ep
ted
S
ep
2
0
,
2
0
20
M
a
ss
iv
e
in
f
o
rm
a
ti
o
n
h
a
s
b
e
e
n
tran
sm
it
ted
th
ro
u
g
h
c
o
m
p
li
c
a
te
d
n
e
tw
o
rk
c
o
n
n
e
c
ti
o
n
s
a
ro
u
n
d
t
h
e
w
o
rld
.
T
h
u
s,
p
ro
v
id
i
n
g
a
p
ro
tec
ted
i
n
f
o
rm
a
ti
o
n
s
y
ste
m
h
a
s
f
u
ll
y
c
o
n
sid
e
ra
ti
o
n
o
f
m
a
n
y
p
riv
a
te an
d
g
o
v
e
rn
m
e
n
tal
i
n
stit
u
tes
t
o
p
re
v
e
n
t
th
e
a
tt
a
c
k
e
rs.
T
h
e
a
tt
a
c
k
e
rs
b
lo
c
k
th
e
u
se
rs
to
a
c
c
e
s
s
a
p
a
rti
c
u
lar
n
e
tw
o
rk
s
e
rv
ic
e
b
y
s
e
n
d
in
g
a
lar
g
e
a
m
o
u
n
t
o
f
fa
k
e
tra
ff
ics
.
T
h
e
r
e
f
o
re
,
th
is
a
rti
c
le
d
e
m
o
n
stra
tes
t
w
o
-
c
las
si
f
ica
ti
o
n
m
o
d
e
ls
f
o
r
a
c
c
u
ra
te
in
tr
u
sio
n
d
e
tec
ti
o
n
sy
ste
m
(IDS)
.
T
h
e
f
irst
m
o
d
e
l
d
e
v
e
lo
p
s th
e
a
rti
f
icia
l
n
e
u
ra
l
n
e
tw
o
rk
(A
NN
)
o
f
m
u
lt
il
a
y
e
r
p
e
rc
e
p
tro
n
(
M
L
P
)
w
it
h
o
n
e
h
i
d
d
e
n
lay
e
r
(M
LP
1
)
b
a
se
d
o
n
d
istr
ib
u
ted
d
e
n
ial
o
f
se
rv
ice
(
DD
o
S
).
T
h
e
M
L
P
1
h
a
s
3
8
i
n
p
u
t
n
o
d
e
s,
1
1
h
id
d
e
n
n
o
d
e
s,
a
n
d
5
o
u
tp
u
t
n
o
d
e
s
.
T
h
e
train
in
g
o
f
th
e
M
L
P
1
m
o
d
e
l
is
im
p
le
m
e
n
ted
w
it
h
NSL
-
KD
D
d
a
tas
e
t
th
a
t
h
a
s
3
8
f
e
a
tu
re
s
a
n
d
f
iv
e
t
y
p
e
s
o
f
re
q
u
e
sts.
T
h
e
M
L
P
1
a
c
h
iev
e
s
d
e
tec
ti
o
n
a
c
c
u
ra
c
y
o
f
9
5
.
6
%
.
T
h
e
se
c
o
n
d
m
o
d
e
l
M
L
P
2
h
a
s
tw
o
h
i
d
d
e
n
la
y
e
rs.
T
h
e
im
p
ro
v
e
d
M
L
P
2
m
o
d
e
l
w
it
h
th
e
sa
m
e
s
e
tu
p
a
c
h
iev
e
s
a
n
a
c
c
u
ra
c
y
o
f
2
.
2
%
h
ig
h
e
r
th
a
n
th
e
M
L
P
1
m
o
d
e
l.
T
h
e
stu
d
y
sh
o
w
s
th
a
t
th
e
M
L
P
2
m
o
d
e
l
p
ro
v
id
e
s
h
ig
h
c
las
sif
ic
a
ti
o
n
a
c
c
u
ra
c
y
o
f
d
if
fe
re
n
t
re
q
u
e
st t
y
p
e
s.
K
ey
w
o
r
d
s
:
A
r
ti
f
icia
l n
e
u
r
al
n
et
w
o
r
k
DDo
S a
ttack
s
I
n
tr
u
s
io
n
d
etec
tio
n
s
y
s
te
m
Mu
ltil
a
y
er
p
er
ce
p
tr
o
n
T
h
is i
s
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
:
A
b
d
u
lr
a
h
m
an
J
ass
a
m
Mo
h
a
m
m
ed
Dir
e
cto
r
ate
o
f
E
d
u
ca
tio
n
i
n
Di
y
ala
Min
i
s
tr
y
o
f
E
d
u
ca
tio
n
3
2
0
0
1
,
Diy
ala,
I
r
aq
E
m
ail: a
b
id
1
9
7
4
j
s
@
g
m
ail.
co
m
1.
I
NT
RO
D
UCT
I
O
N
T
h
e
h
ac
k
er
s
h
a
v
e
attac
k
ed
th
e
c
y
b
er
s
y
s
te
m
ev
er
y
s
in
g
le
d
a
y
.
T
h
e
i
n
f
o
r
m
atio
n
s
ec
u
r
it
y
co
m
p
an
ie
s
an
d
g
o
v
er
n
m
e
n
ts
o
f
v
ar
io
u
s
co
u
n
tr
ie
s
h
av
e
s
p
a
n
n
ed
a
s
ig
n
i
f
ican
t
co
n
s
id
er
atio
n
to
p
r
ev
en
t
th
e
d
is
tr
ib
u
ted
d
en
ial
o
f
s
er
v
ice
(
DDo
S)
att
ac
k
s
[
1
-
2
]
.
T
h
e
co
n
v
e
n
tio
n
al
c
y
b
er
a
t
h
r
ea
t
o
n
th
e
w
eb
s
er
v
es
a
s
t
h
e
DDo
S
.
Ho
w
e
v
er
,
s
e
n
d
in
g
h
u
g
e
p
ac
k
e
ts
to
t
h
e
w
eb
s
er
v
er
s
f
r
o
m
att
ac
k
er
s
’
to
o
ls
i
s
ca
lled
t
h
e
d
en
i
al
o
f
s
er
v
ice
attac
k
(
Do
S).
A
ll o
f
t
h
e
n
et
w
o
r
k
,
tr
an
s
p
o
r
t,
an
d
ap
p
licatio
n
la
y
er
s
h
av
e
b
ee
n
u
s
ed
I
C
MP
,
T
C
P
,
a
n
d
HT
T
P p
r
o
to
co
ls
,
r
esp
ec
tiv
el
y
,
to
p
r
ev
e
n
t
t
h
e
Do
S
attac
k
s
[
3
-
4
]
.
Fi
g
u
r
e
1
illu
s
tr
ates
a
b
asic
to
p
o
lo
g
y
o
f
t
h
e
DDo
S
at
tack
.
W
h
er
ea
s
,
th
e
attac
k
er
co
n
tr
o
ls
a
lar
g
e
n
u
m
b
er
o
f
s
er
v
er
s
th
at
s
en
d
in
g
t
h
e
p
ac
k
ets
to
th
e
v
icti
m
.
T
h
e
attac
k
s
atte
m
p
t
to
b
lo
ck
leg
it
i
m
a
te
u
s
e
r
s
to
ac
ce
s
s
a
p
ar
ticu
lar
n
et
w
o
r
k
s
er
v
ice
b
y
s
e
n
d
in
g
a
lar
g
e
a
m
o
u
n
t
o
f
f
a
k
e
tr
af
f
ics
to
t
h
e
v
icti
m
n
e
t
w
o
r
k
co
n
tin
u
o
u
s
l
y
[
5
]
.
T
h
e
h
ac
k
er
s
h
a
v
e
u
tili
ze
d
t
h
e
B
o
tn
ets
to
r
ea
lize
th
e
ai
m
o
f
DDo
S
attac
k
s
.
A
n
y
n
et
w
o
r
k
is
cr
ea
ted
b
y
a
h
o
s
t
co
m
p
u
te
r
ar
e
B
o
tn
ets,
w
h
ile
t
h
at
i
s
m
a
n
ag
ed
b
y
s
o
m
e
attac
k
er
s
ar
e
th
e
n
et
w
o
r
k
f
o
r
m
ed
b
y
ca
lled
b
o
t
m
a
s
ter
s
[
6
]
.
I
t
is
n
o
ted
th
at
th
e
attac
k
er
s
s
en
d
a
lar
g
e
n
u
m
b
er
o
f
r
eq
u
ests
to
a
s
y
s
te
m
d
u
r
i
n
g
a
s
h
o
r
t
tim
e,
w
h
ic
h
m
a
k
es
th
at
s
y
s
te
m
h
a
n
g
i
n
g
.
T
h
u
s
,
t
h
e
DDo
S
attac
k
tak
e
s
a
s
h
o
r
ter
ti
m
e
th
a
n
t
h
e
Do
S.
T
h
er
ef
o
r
e,
an
ac
cu
r
ate
an
d
f
ast
in
tr
u
s
io
n
d
etec
tio
n
s
y
s
te
m
(
I
DS)
is
s
tr
o
n
g
l
y
en
q
u
ir
ed
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8938
I
n
t J
A
r
ti
f
I
n
tell
,
Vo
l.
9
,
No
.
4
,
Dec
e
m
b
er
20
20
:
6
0
9
–
61
5
610
Fig
u
r
e
1
.
T
h
e
ar
ch
itectu
r
e
o
f
DDo
S a
ttack
T
h
e
m
ec
h
a
n
is
m
o
f
t
h
e
I
DS
i
s
in
te
g
r
ated
b
ased
o
n
th
e
cla
s
s
i
f
icatio
n
o
f
th
e
h
o
s
t
o
r
n
et
w
o
r
k
.
T
h
e
p
r
ec
is
io
n
o
f
clas
s
i
f
icatio
n
t
h
e
n
o
r
m
al
a
n
d
a
n
o
m
a
l
y
r
eq
u
est
s
g
i
v
es t
h
e
I
D
S a
cc
u
r
ac
y
.
T
h
e
c
lass
i
f
icatio
n
s
y
s
te
m
ca
n
r
ec
o
g
n
ize
t
h
e
D
Do
S
atta
ck
b
ased
o
n
tr
af
f
ic
s
b
eh
a
v
io
r
.
Ma
n
y
clas
s
if
icatio
n
tech
n
i
q
u
es
w
er
e
u
s
ed
to
i
m
p
r
o
v
e
t
h
e
I
DS
’
s
ac
cu
r
ac
y
.
Ma
ch
i
n
e
-
lear
n
i
n
g
al
g
o
r
ith
m
s
h
a
v
e
b
ee
n
w
id
el
y
u
t
ilized
f
o
r
I
DSs
.
Var
io
u
s
ar
tif
icial
in
te
lli
g
en
ce
d
etec
t
io
n
m
eth
o
d
s
,
n
a
m
el
y
,
b
a
y
esia
n
n
et
w
o
r
k
s
,
[
7
]
f
u
zz
y
lo
g
ic,
[
8
]
g
en
e
tic
al
g
o
r
ith
m
s
,
[
9
]
clu
s
ter
in
g
,
m
u
ltil
a
y
er
p
er
ce
p
tr
o
n
(
ML
P
)
,
[
1
0
]
ar
tif
icial
n
e
u
r
al
n
et
w
o
r
k
s
(
ANN)
,
[
1
1
]
s
o
f
t
w
ar
e
ag
e
n
t
tech
n
o
lo
g
y
,
a
n
d
s
u
p
p
o
r
t v
ec
to
r
m
ac
h
i
n
es
(
SV
M)
,
[
1
2
]
a
r
e
i
m
p
le
m
en
ted
f
o
r
I
DS a
p
p
licatio
n
[
2
]
.
T
h
e
A
N
N
i
s
o
n
e
o
f
t
h
e
i
m
p
le
m
e
n
ted
tec
h
n
iq
u
es
to
p
r
ev
e
n
t
DDo
S
attac
k
s
.
T
h
is
s
ec
tio
n
a
d
d
r
ess
es
a
s
u
r
v
e
y
o
f
t
h
e
ar
ticle
s
u
s
ed
A
N
N
f
o
r
I
DS.
T
h
e
ar
tif
icia
l
n
eu
r
al
n
et
w
o
r
k
cla
s
s
i
f
icat
i
o
n
-
b
ased
I
DS
w
as
in
tr
o
d
u
ce
d
.
[
1
3
]
T
h
e
A
NN
m
o
d
el
u
s
ed
NSL
-
KDD
d
ataset
w
i
th
2
9
f
ea
tu
r
es.
T
h
e
m
o
d
el
a
ch
ie
v
ed
an
ac
c
u
r
ac
y
o
f
8
1
.
2
%.
T
h
en
,
Ma
n
e
p
r
o
p
o
s
ed
A
NN
to
d
etec
t
th
e
at
ta
ck
tr
af
f
ic
s
w
h
ich
is
m
i
s
lead
i
n
g
w
it
h
th
e
n
o
r
m
a
l
tr
af
f
ics
[
1
4
]
.
On
l
y
2
0
0
0
0
s
am
p
les
o
f
KDD9
9
d
ataset
an
d
(
1
7
o
f
4
1
)
f
ea
tu
r
es
w
er
e
u
s
ed
to
ev
alu
ate
th
e
p
er
f
o
r
m
a
n
ce
o
f
th
e
p
r
o
p
o
s
ed
m
o
d
el.
T
h
e
s
y
s
te
m
r
ea
lized
th
e
ac
cu
r
ac
y
o
f
9
8
%.
T
h
e
s
elec
ti
n
g
ef
f
ec
t
iv
e
f
ea
t
u
r
es
f
o
r
th
e
AN
N
m
o
d
el
t
h
at
i
m
p
r
o
v
ed
t
h
e
s
y
s
te
m
ac
c
u
r
ac
y
.
T
s
ai
in
[
1
5
]
p
r
o
p
o
s
ed
a
ti
m
e
-
d
ela
y
n
eu
r
al
n
et
w
o
r
k
t
h
at
ca
n
b
e
u
s
ed
as
a
n
ea
r
l
y
d
etec
tio
n
s
y
s
te
m
o
f
D
Do
S
attac
k
s
.
T
h
is
m
et
h
o
d
is
i
m
p
le
m
en
ted
b
ased
o
n
th
e
ti
m
e
p
ar
a
m
eter
o
f
ea
ch
r
eq
u
est.
T
h
e
s
y
s
te
m
is
te
s
ted
a
m
a
n
u
a
ll
y
g
en
er
ated
d
ataset.
T
h
e
s
y
s
te
m
o
u
tco
m
e
d
is
p
la
y
ed
ac
cu
r
ac
y
o
f
8
2
%.
I
n
ad
d
itio
n
,
it c
an
d
ete
ct
a
f
e
w
t
y
p
e
s
o
f
attac
k
s
.
On
t
h
e
o
th
er
h
an
d
,
t
h
e
M
L
P
h
as
b
ee
n
d
e
v
elo
p
ed
in
s
e
v
er
al
s
tu
d
ies
f
o
r
I
DS
ap
p
licatio
n
s
.
T
h
e
ML
P
in
cl
u
d
es
m
u
ltip
l
e
h
id
d
en
la
y
er
s
to
i
m
p
r
o
v
e
th
e
lear
n
i
n
g
r
ate
o
f
th
e
n
e
u
r
al
n
et
w
o
r
k
.
S
in
g
h
d
ev
elo
p
ed
th
e
M
L
P
w
i
th
a
g
e
n
etic
al
g
o
r
ith
m
to
i
m
p
r
o
v
e
t
h
e
d
etec
tio
n
e
f
f
ic
ien
c
y
o
f
th
e
I
DS
m
o
d
el
[
1
6
]
.
T
h
e
liter
at
u
r
e
s
h
o
ws
d
if
f
er
e
n
t
t
y
p
es
o
f
t
h
e
d
ataset
t
h
at
u
s
ed
to
e
v
al
u
ate
t
h
e
I
D
S
s
y
s
te
m
,
s
u
c
h
a
s
C
I
D
A
,
D
A
R
B
A
FIF
A
w
o
r
ld
c
u
p
an
d
KDD
-
N
SL
[
2
]
.
T
h
e
s
elec
t
io
n
o
f
t
h
e
e
f
f
ec
ti
v
e
f
ea
t
u
r
es
o
f
th
e
C
A
I
D
A
2
0
0
7
d
ataset
w
i
t
h
g
en
e
tic
al
g
o
r
it
h
m
w
a
s
i
m
p
r
o
v
ed
t
h
e
s
y
s
te
m
p
er
f
o
r
m
a
n
ce
i
n
ter
m
s
o
f
DDo
S
cla
s
s
i
f
icatio
n
.
W
an
g
u
s
ed
t
h
e
M
L
P
to
o
v
er
co
m
e
t
h
e
DD
OS
attac
k
p
r
o
b
le
m
[
1
7
]
.
W
h
er
e
th
e
M
L
P
w
as
co
m
b
in
ed
w
it
h
d
y
n
a
m
ic
f
ea
t
u
r
es
s
e
l
ec
tio
n
.
Du
r
i
n
g
t
h
e
tr
ain
i
n
g
p
r
o
ce
s
s
,
t
h
e
m
o
d
el
c
h
o
s
e
t
h
e
o
p
ti
m
al
f
ea
t
u
r
es
ac
c
o
r
d
in
g
to
th
e
f
ee
d
b
ac
k
c
h
ec
k
i
n
g
o
f
t
h
e
er
r
o
r
s
o
f
ea
ch
ep
o
ch
[
1
8
]
.
T
h
er
ef
o
r
e,
an
an
a
l
y
s
es
i
n
v
esti
g
atio
n
is
r
ec
o
m
m
e
n
d
ed
to
h
i
g
h
li
g
h
t
th
e
ef
f
ec
ti
v
e
d
etec
tio
n
m
et
h
o
d
b
ased
o
n
n
eu
r
al
n
e
t
w
o
r
k
alg
o
r
it
h
m
s
.
T
h
is
ar
ticle
p
r
o
v
id
es
t
w
o
M
L
P
m
o
d
els
o
f
A
NN
f
o
r
cla
s
s
i
f
i
ca
tio
n
-
b
a
s
ed
I
DS.
T
h
e
M
L
P
b
ased
I
DS
s
y
s
te
m
d
esig
n
w
i
th
o
n
e
h
id
d
en
la
y
er
(
M
L
P
1
)
an
d
t
w
o
h
id
d
en
la
y
er
s
(
M
L
P
2
)
to
im
p
r
o
v
e
class
i
f
icatio
n
ac
cu
r
ac
y
.
T
h
e
ar
ticle
is
o
r
g
a
n
i
ze
d
in
to
s
ev
er
al
s
ec
t
io
n
s
a
s
f
o
llo
w
s
.
I
n
s
ec
tio
n
I
I
,
th
e
d
esig
n
o
f
I
DS
m
o
d
el
s
is
ad
d
r
ess
ed
an
d
d
iv
id
ed
i
n
to
t
h
r
ee
s
u
b
s
ec
t
io
n
s
,
w
h
ic
h
ar
e
th
e
m
et
h
o
d
s
a
n
d
m
a
ter
ials
s
ec
ti
o
n
(
d
ataset,
M
L
P
1
,
an
d
M
L
P
2
)
.
Af
ter
t
h
at,
t
h
e
s
i
m
u
latio
n
o
f
t
h
e
t
w
o
p
r
o
p
o
s
ed
m
o
d
els
a
n
d
t
h
e
e
v
al
u
atio
n
o
f
t
h
e
r
e
s
u
lts
ar
e
ad
d
r
ess
ed
in
Sectio
n
3
.
Fin
a
ll
y
,
Sectio
n
4
co
n
cl
u
d
es t
h
e
s
t
u
d
y
’
s
ac
h
ie
v
e
m
en
ts
.
2.
M
E
T
H
O
DS A
ND
M
AT
E
RI
AL
S
I
n
th
is
s
ec
t
io
n
,
b
r
ief
b
ac
k
g
r
o
u
n
d
s
ab
o
u
t
th
e
test
ed
d
ataset
th
e
s
elec
ted
d
ataset
’
s
f
e
atu
r
es
ar
e
d
e
m
o
n
s
tr
ated
.
Mo
r
eo
v
er
,
th
e
i
n
v
e
s
ti
g
ated
A
NN
-
b
ased
class
i
f
icatio
n
I
DS
m
e
th
o
d
s
ar
e
ad
d
r
ess
ed
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
A
r
ti
f
I
n
tell
I
SS
N:
2252
-
8938
A
mu
ltil
a
ye
r
p
ercep
tr
o
n
a
r
tifi
c
ia
l n
eu
r
a
l n
etw
o
r
k
a
p
p
r
o
a
ch
fo
r
…
(
A
b
d
u
lr
a
h
ma
n
Ja
s
s
a
m
Mo
h
a
mme
d
)
611
2
.
1
.
T
esting
da
t
a
s
et
Am
o
n
g
th
e
d
if
f
er
en
t
t
y
p
e
s
o
f
d
atasets
,
t
h
e
N
S
L
-
KDD
d
ata
s
et
is
s
elec
ted
d
u
e
to
it
s
v
ar
iet
y
o
f
f
ea
t
u
r
es
th
at
co
u
ld
b
e
s
u
itab
le
to
e
v
al
u
ate
th
e
p
er
f
o
r
m
a
n
ce
o
f
o
u
r
p
r
o
p
o
s
ed
m
o
d
el.
T
h
e
av
ailab
le
N
SL
-
KD
D
d
ataset
i
n
[
1
9
-
2
1
]
w
a
s
u
t
ilized
to
ev
al
u
a
te
th
e
A
N
N
m
o
d
el
i
n
t
h
is
s
t
u
d
y
.
T
h
e
NS
L
-
KDD
Data
s
et
is
th
e
u
p
d
ate
v
er
s
io
n
o
f
9
9
KDD
d
ataset
[
2
2
-
2
4
]
.
T
h
e
NSL
-
K
DD
w
as
r
ec
o
r
d
e
d
ac
co
r
d
in
g
to
a
p
r
ac
tical
1
2
5
9
7
3
r
e
q
u
ests
o
f
a
w
eb
s
i
te
in
2
0
1
3
.
T
h
ese
r
eq
u
ests
w
er
e
m
o
n
ito
r
ed
a
n
d
r
ec
o
r
d
ed
w
it
h
s
e
v
er
al
f
ea
t
u
r
es.
T
h
e
NS
L
-
KDD
d
ataset
in
cl
u
d
es
4
2
f
ea
tu
r
e
s
[
2
]
.
Du
r
in
g
t
h
e
p
r
e
-
p
r
o
ce
s
s
i
n
g
,
s
o
m
e
o
f
th
ese
f
ea
t
u
r
es
w
er
e
ex
c
lu
d
ed
to
m
ee
t
th
e
a
p
p
licatio
n
r
eq
u
ir
e
m
e
n
ts
.
W
h
er
e
th
e
r
eq
u
est
t
y
p
e
f
ea
tu
r
e
s
w
er
e
e
x
tr
ad
ited
,
an
d
th
e
n
it
i
s
s
et
as
t
h
e
tar
g
e
t
o
f
th
e
tr
ai
n
i
n
g
I
DS
m
o
d
el.
I
t
is
n
o
ted
th
at
t
h
e
NS
L
-
KDD
d
atas
et
in
c
lu
d
es
f
iv
e
t
y
p
e
s
o
f
r
eq
u
e
s
ts
,
w
h
ic
h
ar
e
DOS
o
f
4
5
9
2
7
,
Pro
b
e
o
f
5
2
,
R
2
L
o
f
9
9
5
,
U
2
R
o
f
1
1
6
5
6
,
an
d
No
r
m
al
o
f
6
7
3
4
3
,
as d
is
p
lay
ed
in
Fig
u
r
e
2
.
Fig
u
r
e
2
.
T
h
e
r
eq
u
est t
y
p
es o
f
th
e
NS
L
-
KDD
d
atase
t
So
m
e
o
f
th
e
u
n
w
a
n
ted
f
ea
tu
r
es,
s
u
c
h
as
Fl
g
,
T
C
P
an
d
HT
T
P
w
er
e
r
e
m
o
v
ed
f
r
o
m
t
h
e
d
ataset.
T
h
er
ef
o
r
e,
th
e
r
e
m
ai
n
ed
3
8
f
ea
t
u
r
es
o
f
th
e
s
elec
ted
d
ata
s
et
w
er
e
i
m
p
o
r
ted
as
i
n
p
u
t
s
o
f
th
e
I
DS
m
o
d
el.
Sp
ec
if
icall
y
,
8
8
1
8
1
s
am
p
le
s
o
f
th
e
d
ata
s
et
ar
e
u
s
ed
f
o
r
tr
ain
i
n
g
w
h
ile
3
7
7
9
1
2
s
am
p
le
s
ar
e
u
s
ed
f
o
r
test
in
g
an
d
v
alid
atio
n
.
2
.
2
.
T
he
I
DS
f
ra
m
ew
o
rk
T
h
e
A
NN
is
m
o
d
elled
w
i
th
o
n
e
h
id
d
en
la
y
er
(
M
L
P
1
)
an
d
t
w
o
h
id
d
en
la
y
er
s
(
ML
P
2
)
to
i
m
p
r
o
v
e
class
i
f
icatio
n
ac
cu
r
ac
y
.
W
h
er
e
p
r
e
-
p
r
ess
in
g
h
a
s
s
elec
ted
th
e
in
p
u
t
a
n
d
tar
g
et
f
ea
t
u
r
es
f
r
o
m
th
e
N
S
L
-
KDD
da
taset.
I
n
ad
d
itio
n
,
7
0
%
o
f
t
h
e
r
eq
u
ests
ar
e
i
m
p
o
r
ted
to
t
h
e
ANN
f
o
r
th
e
tr
ain
in
g
p
r
o
ce
s
s
w
h
i
le
th
e
o
th
er
3
0
%
o
f
th
e
r
eq
u
e
s
ts
ar
e
u
s
ed
f
o
r
test
i
n
g
.
T
h
is
tr
ain
i
n
g
is
i
m
p
le
m
en
ted
w
ith
s
ev
er
al
s
to
p
p
in
g
cr
iter
ia.
Fo
r
ex
a
m
p
le,
t
h
e
KDD
d
ataset
i
s
tr
ain
ed
w
ith
1
0
0
ep
o
ch
s
.
Als
o
,
th
e
m
ea
n
s
q
u
ar
e
er
r
o
r
o
f
1
0
-
6
is
s
et
to
s
to
p
th
e
s
y
s
te
m
tr
ai
n
in
g
.
Af
ter
t
h
at,
th
e
b
ias
an
d
w
eig
h
t
s
o
f
th
e
A
N
N
ar
e
s
a
v
ed
f
o
r
te
s
tes
p
r
o
ce
s
s
.
T
h
e
p
r
o
p
o
s
ed
A
NN
f
o
r
I
DS
s
y
s
te
m
i
s
i
m
p
le
m
en
te
d
b
y
u
s
i
n
g
M
A
T
L
A
B
s
o
f
t
w
ar
e.
T
h
e
I
DS
f
r
a
m
e
w
o
r
k
w
as
d
esig
n
ed
to
cla
s
s
i
f
y
th
e
r
eq
u
est t
y
p
e
ac
co
r
d
in
g
to
t
h
e
s
elec
ted
f
ea
tu
r
e
s
,
as s
h
o
w
n
in
Fi
g
u
r
e
3
.
Fig
u
r
e
3
.
T
h
e
I
DS
f
r
a
m
e
w
o
r
k
N
o
r
mal
,
67343
D
o
S
,
45927
Pr
o
b
i
n
g
,
52
U
2
R
,
11656
R
2
L
,
995
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8938
I
n
t J
A
r
ti
f
I
n
tell
,
Vo
l.
9
,
No
.
4
,
Dec
e
m
b
er
20
20
:
6
0
9
–
61
5
612
2
.
3
.
T
he
M
L
P
1
m
o
del
T
h
e
A
NN
h
as
b
ee
n
d
ev
elo
p
ed
b
y
u
s
in
g
a
co
m
p
u
ter
m
o
d
el
o
f
m
u
ltil
a
y
er
p
er
ce
p
tr
o
n
w
i
th
o
n
e
h
id
d
en
la
y
er
(
M
L
P
1
)
.
T
h
e
ML
P
1
ar
ch
itect
u
r
e
i
s
b
u
i
ld
b
ased
o
n
t
h
e
b
io
lo
g
ical
n
eu
r
o
n
.
W
h
er
e
t
h
e
in
f
o
r
m
at
io
n
f
lo
w
s
in
t
h
e
n
eu
r
o
n
s
c
h
an
g
e
s
t
h
e
le
ar
n
in
g
o
f
t
h
e
m
o
d
el
b
ased
o
n
its
s
tr
u
ct
u
r
e.
T
h
e
o
u
tp
u
t
o
f
t
h
e
M
L
P
1
is
u
s
ed
to
ch
ec
k
t
h
e
er
r
o
r
an
d
u
p
d
ate
t
h
e
m
o
d
el
w
e
ig
h
t.
T
h
e
s
i
m
p
le
ML
P
1
co
n
f
ig
u
r
atio
n
h
as
th
r
e
e
la
y
er
s
,
w
h
ich
ar
e
in
p
u
t
la
y
er
,
h
id
d
en
la
y
er
an
d
o
u
tp
u
t
la
y
er
.
T
h
e
in
f
o
r
m
atio
n
f
lo
w
s
t
h
r
o
u
g
h
l
in
k
a
g
es
a
m
o
n
g
t
h
e
la
y
er
s
.
T
h
e
o
u
tp
u
t
la
y
er
v
al
u
es
ar
e
co
m
p
u
ted
b
y
m
u
ltip
l
y
i
n
g
t
h
e
i
n
p
u
t
i
n
f
o
r
m
atio
n
w
it
h
th
e
w
e
ig
h
t
o
f
t
h
e
l
in
k
a
g
e
ac
co
r
d
in
g
to
t
h
e
u
s
ed
ac
ti
v
ati
o
n
f
u
n
ctio
n
.
T
h
e
er
r
o
r
ca
n
b
e
ca
lcu
lated
b
y
co
m
p
ar
in
g
t
h
e
o
u
tp
u
t
o
f
M
L
P
1
w
it
h
th
e
d
ata
ta
r
g
et.
T
h
is
er
r
o
r
is
u
tili
ze
d
to
u
p
d
ate
t
h
e
M
L
P
1
w
ei
g
h
t
b
y
u
s
i
n
g
b
ac
k
p
r
o
p
ag
atio
n
ar
ch
itect
u
r
e
.
r
n
hI
ei
tn
i
n
i
e
eeiI
n
t
n
i
tneI
intn
n
h
ee
n
n
i
e
n
e n
r
et
i
i
n i
n
I
e
e
he
ni
e
eI
I
n
I
.
T
o
d
esig
n
I
DS,
t
h
e
M
L
P
1
is
i
m
p
le
m
en
ted
f
o
r
th
e
au
r
ea
te
cl
ass
i
f
icatio
n
s
y
s
te
m
.
T
h
e
p
r
o
p
o
s
ed
ML
P
1
is
d
ev
elo
p
ed
an
d
s
i
m
u
lated
b
y
u
s
i
n
g
M
A
T
L
A
B
.
T
h
e
p
r
o
p
o
s
ed
ML
P
1
f
o
r
I
DS
is
d
es
ig
n
e
d
w
it
h
t
h
r
ee
l
a
y
er
s
,
as
d
is
p
la
y
ed
in
Fi
g
u
r
e
4
(
a)
.
T
h
e
s
elec
ted
3
8
f
ea
tu
r
es
o
f
NS
L
-
KDD
d
ata
ar
e
o
r
g
an
ized
an
d
u
s
ed
as
in
p
u
ts
f
o
r
th
e
M
L
P
1
.
T
h
er
ef
o
r
e,
th
e
in
p
u
t
la
y
er
in
cl
u
d
es
3
8
in
p
u
t
n
o
d
es
o
n
e
n
o
d
e
f
o
r
ea
ch
in
p
u
t
f
ea
tu
r
e.
T
h
e
d
ataset
in
cl
u
d
es
s
o
m
a
n
y
r
eq
u
est
s
.
E
ac
h
r
eq
u
est
h
as
3
8
f
ea
t
u
r
es
t
h
at
i
m
p
o
r
ted
to
th
e
d
esig
n
M
L
P
1
as
in
p
u
t.
W
h
er
ea
s
ev
er
y
s
i
n
g
le
f
ea
tu
r
e
o
f
a
r
eq
u
est
is
i
m
p
o
r
ted
to
a
n
i
n
p
u
t
n
o
d
e.
T
h
e
f
ea
tu
r
es
o
f
a
ll
t
h
e
r
eq
u
est
s
i
n
t
h
e
d
atas
et
ar
e
u
s
ed
to
tr
ai
n
t
h
e
M
L
P
1
r
esp
ec
tiv
el
y
.
Mo
r
eo
v
er
,
a
s
i
n
g
l
e
h
id
d
en
la
y
er
i
s
co
n
f
i
g
u
r
ed
w
i
th
1
1
n
o
d
es.
T
h
e
h
id
d
en
la
y
er
is
ac
t
iv
ated
b
y
u
s
in
g
t
h
e
s
i
n
g
le
-
b
ias
o
f
ea
ch
n
o
d
e.
Hen
ce
,
a
s
ig
m
o
id
ac
ti
v
at
io
n
f
u
n
ctio
n
is
u
s
ed
to
ac
tiv
ate
t
h
e
n
e
u
r
o
n
s
.
T
h
e
s
ig
m
o
id
i
s
v
er
ticall
y
n
o
r
m
al
i
ze
d
th
e
tr
ain
i
n
g
f
ea
t
u
r
es
b
et
w
ee
n
0
a
n
d
1
.
T
h
e
h
id
d
en
la
y
er
is
co
n
n
ec
ted
w
ith
th
e
o
u
tp
u
t
la
y
er
o
f
t
h
e
A
NN.
Sin
ce
t
h
e
r
eq
u
est
s
o
f
t
h
e
d
atas
et
ar
e
lab
elled
in
to
f
i
v
e
d
if
f
er
en
t
t
y
p
e
s
,
s
o
th
e
o
u
tp
u
t
la
y
er
is
co
n
f
i
g
u
r
ed
w
it
h
5
n
o
d
es.
T
h
er
ef
o
r
e,
th
e
p
r
o
p
o
s
ed
ML
P
1
f
o
r
I
DS
ca
n
b
e
class
i
f
ied
as
r
eq
u
e
s
ts
in
to
f
i
v
e
t
y
p
es
as
s
h
o
w
n
i
n
Fig
u
r
e
4
(
a)
.
E
ac
h
n
o
d
e
o
f
th
e
h
id
d
en
la
y
er
i
s
co
n
n
ec
ted
w
it
h
all
th
e
o
u
tp
u
t
n
o
d
es
b
y
u
s
i
n
g
f
i
v
e
n
e
u
r
o
n
s
.
Mo
r
e
d
etails
ab
o
u
t
th
e
i
m
p
le
m
en
ta
tio
n
ar
e
ad
d
r
ess
ed
in
th
e
n
e
x
t p
ar
ag
r
ap
h
.
2
.
4
.
T
he
M
L
P
2
m
o
del
Gen
er
all
y
,
t
h
e
M
L
P
2
h
as
th
e
s
a
m
e
s
tr
u
ctu
r
e
o
f
t
h
e
ML
P
1
ex
ce
p
t
th
at
m
u
ltip
le
h
id
d
en
la
y
er
s
ar
e
co
n
n
ec
ted
in
s
er
ies.
T
h
e
ML
P
2
in
clu
d
es
a
lar
g
e
n
u
m
b
er
o
f
i
n
n
er
co
n
n
ec
tio
n
s
o
f
n
eu
r
o
n
s
t
o
s
o
lv
e
a
p
r
o
b
lem
.
Sev
er
al
h
id
d
en
la
y
er
s
ar
e
lin
k
ed
b
et
w
ee
n
t
h
e
i
n
p
u
t la
y
er
an
d
th
e
o
u
tp
u
t la
y
er
,
as d
is
p
la
y
ed
in
Fi
g
u
r
e
4
(
b
)
.
T
h
e
in
p
u
t
o
f
th
e
n
ex
t
la
y
er
is
g
e
n
e
r
ated
b
y
ap
p
l
y
i
n
g
a
n
ac
ti
v
atio
n
f
u
n
ctio
n
to
t
h
e
o
u
tp
u
t
o
f
th
e
p
r
ev
io
u
s
la
y
er
.
T
h
e
s
ig
m
o
id
ac
tiv
at
i
o
n
f
u
n
ctio
n
h
as
w
id
el
y
u
s
ed
in
th
e
f
o
r
w
ar
d
-
p
r
o
p
ag
atio
n
n
e
u
r
al
n
et
w
o
r
k
tr
ain
in
g
.
T
h
e
er
r
o
r
is
o
b
tain
ed
b
y
co
m
p
ar
i
n
g
th
e
n
et
w
o
r
k
o
u
tp
u
t
w
i
th
t
h
e
d
ata
tar
g
et.
T
h
e
b
ac
k
-
p
r
o
p
ag
atio
n
ad
j
u
s
ts
th
e
M
L
P
2
w
ei
g
h
t
v
ec
to
r
.
T
h
e
d
is
cu
s
s
ed
p
r
o
ce
s
s
f
o
r
all
th
e
i
n
p
u
t
v
ec
to
r
s
ca
lls
tr
ain
i
n
g
ep
o
ch
,
w
h
i
ch
is
r
ep
ea
ted
f
o
r
s
ev
er
al
ep
o
ch
s
u
n
til
m
ee
ti
n
g
t
h
e
s
to
p
cr
iter
ia.
(
a)
T
h
e
ML
P
1
co
n
f
i
g
u
r
atio
n
o
f
th
r
ee
-
la
y
er
s
(
b
)
T
h
e
ML
P
2
co
n
f
ig
u
r
atio
n
o
f
f
o
u
r
-
la
y
er
s
Fig
u
r
e
4
.
T
h
e
ML
P
to
p
o
l
o
g
y
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
A
r
ti
f
I
n
tell
I
SS
N:
2252
-
8938
A
mu
ltil
a
ye
r
p
ercep
tr
o
n
a
r
tifi
c
ia
l n
eu
r
a
l n
etw
o
r
k
a
p
p
r
o
a
ch
fo
r
…
(
A
b
d
u
lr
a
h
ma
n
Ja
s
s
a
m
Mo
h
a
mme
d
)
613
T
h
e
ML
P
2
is
d
esig
n
ed
in
t
h
i
s
s
ec
tio
n
to
class
if
y
t
h
e
NS
L
-
KDD
da
taset
f
o
r
I
DS
ap
p
licatio
n
.
T
h
e
p
r
o
p
o
s
ed
P
L
M
is
d
esig
n
ed
w
i
th
f
o
u
r
la
y
er
s
(
1
in
p
u
t,
2
h
id
d
en
,
an
d
1
o
u
tp
u
t
la
y
er
s
)
.
T
h
e
ML
P
2
la
y
er
s
’
h
a
v
e
th
e
s
a
m
e
n
u
m
b
er
o
f
n
o
d
es
in
th
e
p
r
o
p
o
s
ed
A
NN
m
o
d
el.
W
h
er
e
th
e
in
p
u
t
la
y
er
h
a
s
3
8
n
o
d
es,
th
e
h
id
d
en
la
y
er
s
h
a
v
e
1
1
n
o
d
es
a
n
d
t
h
e
o
u
tp
u
t
la
y
er
h
as
5
n
o
d
es.
F
ig
u
r
e
4
illu
s
tr
ates
th
e
MP
L
ar
c
h
it
ec
tu
r
e
w
it
h
i
ts
f
o
u
r
la
y
er
s
co
n
n
ec
t
io
n
.
Feed
-
f
o
r
w
ar
d
b
ac
k
-
p
r
o
p
ag
atio
n
h
as
co
n
n
ec
ted
th
e
n
e
u
r
o
n
s
a
m
o
n
g
th
e
la
y
er
s
’
n
o
d
es.
I
n
ad
d
itio
n
,
th
e
s
i
g
m
o
id
ac
tiv
atio
n
f
u
n
ctio
n
i
s
u
s
ed
w
ith
n
o
r
m
al
izatio
n
t
h
e
i
n
p
u
t
f
ea
t
u
r
es
v
al
u
es
b
et
w
ee
n
0
an
d
1
.
Sa
m
e
t
h
e
p
r
e
-
p
r
o
ce
s
s
i
n
g
t
h
at
u
s
ed
f
o
r
t
h
e
ANN
m
o
d
el
is
al
s
o
i
m
p
le
m
e
n
ted
f
o
r
th
i
s
M
L
P
2
m
o
d
el.
Mo
r
eo
v
er
,
th
e
s
o
p
p
in
g
cr
iter
i
a
(
n
o
o
f
e
p
o
ch
s
=
1
0
0
an
d
m
e
an
s
q
u
ar
e
er
r
o
r
o
f
1
0
-
6
)
ar
e
u
s
ed
f
o
r
t
h
e
p
r
o
p
o
s
ed
ML
P
m
o
d
el.
T
h
e
r
eq
u
est
t
y
p
e
o
f
th
e
tes
ted
d
ataset
is
s
e
t
as
a
tar
g
et
w
h
ile
t
h
e
s
e
lect
ed
3
8
f
ea
tu
r
es
ar
e
i
m
p
o
r
ted
as
i
n
p
u
ts
o
f
th
e
M
L
P
I
DS
m
o
d
el.
T
h
e
m
o
d
el
i
s
i
m
p
le
m
en
ted
b
y
u
s
in
g
M
A
T
L
A
B
f
o
r
1
0
0
ep
o
ch
s
tr
ain
i
n
g
.
T
h
e
w
eig
h
t
s
an
d
b
ias
es o
f
th
e
M
L
P
2
ar
e
r
ec
o
r
d
ed
f
o
r
th
e
ev
alu
a
tio
n
p
r
o
ce
s
s
.
3.
SI
M
UL
AT
I
O
N
E
VA
L
UA
T
I
O
N
AND
RE
SU
L
T
S
T
h
e
ML
P
1
an
d
M
L
P
2
m
o
d
els
ar
e
i
m
p
le
m
e
n
ted
b
y
u
s
in
g
M
A
T
L
A
B
s
o
f
t
w
ar
e,
an
d
6
4
b
it
W
in
d
o
w
s
8
.
T
h
e
u
s
ed
co
m
p
u
t
er
h
as
s
p
ec
if
icatio
n
s
o
f
I
n
tel
C
P
U
co
r
e
i7
@
2
.
1
0
GHz
w
it
h
R
A
M
o
f
4
GB
.
T
h
e
p
er
f
o
r
m
a
n
ce
o
f
t
h
e
s
y
s
te
m
i
s
c
o
m
p
u
ted
b
y
an
a
l
y
zi
n
g
th
e
r
es
u
lts
o
f
t
h
e
test
ed
d
ataset.
T
h
e
last
3
0
%
o
f
t
h
e
NS
L
-
KD
D
d
ataset
is
u
s
ed
to
test
b
o
th
t
h
e
p
r
o
p
o
s
ed
m
o
d
el.
Firs
tl
y
,
th
e
p
r
o
p
o
s
ed
ML
P
1
w
ith
th
e
r
ec
o
r
d
ed
w
eig
h
ts
is
i
m
p
le
m
e
n
ted
w
it
h
t
h
e
te
s
tin
g
d
ata
s
et.
T
h
e
o
u
tp
u
t
clas
s
if
icatio
n
r
es
u
lt
s
o
f
th
e
test
ed
M
L
P
1
is
r
ec
o
r
d
ed
an
d
co
m
p
a
r
ed
w
it
h
th
e
tar
g
et
(
r
eq
u
est
t
y
p
e
f
ea
t
u
r
e)
.
T
h
e
co
n
f
u
s
io
n
m
a
tr
ix
i
s
u
s
ed
to
ca
m
p
o
u
t
th
e
c
lass
if
ica
tio
n
ac
cu
r
ac
y
o
f
t
h
e
M
L
P
1
m
o
d
el,
as
illu
s
tr
ated
i
n
Fi
g
u
r
e
5
.
T
h
e
ML
P
1
m
o
d
el
ac
h
ie
v
es t
h
e
b
est ac
cu
r
ac
y
o
f
9
5
.
6
2
% u
s
in
g
(
1
)
[
2
5
-
3
1
]
af
ter
1
0
0
e
p
o
ch
s
.
(
1
)
w
h
er
e
T
N,
T
P
,
FN,
FP
d
en
o
tes
tr
u
e
n
e
g
ati
v
e,
tr
u
e
p
o
s
i
tiv
e,
f
a
ls
e
n
e
g
ati
v
e
an
d
f
als
e
p
o
s
itiv
e,
r
esp
ec
tiv
el
y
.
Seco
n
d
l
y
,
t
h
e
M
L
P
2
m
o
d
el
with
t
h
e
r
ec
o
r
d
ed
w
ei
g
h
ts
i
s
i
m
p
le
m
en
ted
b
y
u
s
i
n
g
th
e
test
i
n
g
d
ataset.
T
h
e
o
u
tp
u
ts
o
f
M
L
P
2
in
I
DS
ar
e
r
ec
o
r
d
ed
an
d
v
alid
a
ted
w
it
h
th
e
tar
g
et.
T
h
e
s
y
s
te
m
ac
c
u
r
ac
y
is
ca
lc
u
lated
b
y
u
s
i
n
g
(
1
)
.
Fig
u
r
e
6
s
h
o
w
s
t
h
e
co
n
f
u
s
io
n
m
atr
ix
o
f
t
h
e
p
r
o
p
o
s
ed
s
y
s
te
m
.
I
t
ca
n
b
e
n
o
t
ed
th
at
t
h
e
M
L
P
2
r
ea
lized
th
e
b
est ac
cu
r
ac
y
o
f
9
7
.
8
2
% a
f
ter
1
0
0
ep
o
ch
s
.
Fig
u
r
e
5
.
T
h
e
co
n
f
u
s
io
n
m
atr
i
x
an
d
clas
s
i
f
icatio
n
p
er
f
o
r
m
a
n
ce
o
f
t
h
e
M
L
P
1
Fig
u
r
e
6
.
T
h
e
co
n
f
u
s
io
n
m
atr
i
x
an
d
clas
s
i
f
icatio
n
p
er
f
o
r
m
a
n
ce
o
f
t
h
e
M
L
P
2
Gen
er
all
y
,
t
h
e
p
r
o
p
o
s
ed
ML
P
2
m
o
d
el
d
e
m
o
n
s
tr
ates
t
h
e
ac
c
u
r
ac
y
o
f
9
7
.
8
2
%,
w
h
ic
h
i
s
o
b
s
er
v
ed
b
y
2
.
2
%
h
ig
h
er
th
a
n
th
e
M
L
P
1
m
o
d
el.
Fig
u
r
e
7
s
h
o
w
s
th
e
ev
al
u
atio
n
ac
cu
r
ac
y
r
es
u
lts
f
o
r
b
o
th
ML
P
1
an
d
ML
P
2
m
o
d
els.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8938
I
n
t J
A
r
ti
f
I
n
tell
,
Vo
l.
9
,
No
.
4
,
Dec
e
m
b
er
20
20
:
6
0
9
–
61
5
614
Fig
u
r
e
7
.
T
h
e
ac
cu
r
ac
y
e
v
al
u
a
tio
n
b
et
w
ee
n
t
h
e
M
L
P
1
an
d
ML
P
2
f
o
r
1
0
0
e
p
o
ch
s
Fro
m
t
h
e
r
es
u
lts
,
i
t
ca
n
b
e
ar
g
u
ed
t
h
at
t
h
e
ad
d
itio
n
a
l
h
id
d
en
la
y
er
in
cr
ea
s
ed
th
e
l
ea
r
n
i
n
g
r
ate
m
o
r
e
th
a
n
th
e
s
i
n
g
le
h
id
d
en
la
y
er
.
I
n
ad
d
itio
n
,
b
o
th
o
f
th
e
M
L
P
1
an
d
ML
P
2
ar
e
f
ailed
to
d
etec
t th
e
P
r
o
b
in
g
r
eq
u
est
t
y
p
e.
T
h
er
e
ar
e
1
2
P
r
o
b
in
g
r
eq
u
ests
ar
e
i
m
p
le
m
en
ted
in
t
h
e
test
i
n
g
d
atase
t.
T
h
er
ef
o
r
e,
a
f
u
r
th
er
i
n
v
e
s
ti
g
atio
n
o
f
th
e
s
e
lecte
d
f
ea
t
u
r
es a
n
d
t
h
e
d
ev
elo
p
ed
I
DS m
o
d
els is
r
eq
u
ir
ed
.
4.
CO
NCLUS
I
O
N
I
n
th
is
ar
ticle,
t
w
o
t
y
p
e
s
o
f
ML
P
ar
e
d
esig
n
ed
a
n
d
m
o
d
el
led
f
o
r
I
DS
ap
p
licatio
n
s
.
T
h
e
NS
L
-
KDD
d
ataset
is
u
s
ed
f
o
r
b
o
th
o
f
th
e
tr
ain
i
n
g
a
n
d
test
i
n
g
s
ta
g
es.
O
n
l
y
3
8
o
f
4
2
f
ea
tu
r
e
s
ar
e
u
s
ed
f
o
r
th
e
class
i
f
icat
io
n
s
tag
e.
Sa
m
e
s
et
u
p
an
d
p
ar
am
eter
s
ar
e
u
s
ed
o
f
b
o
th
ML
P
s
m
o
d
els.
T
h
e
ev
alu
atio
n
s
t
u
d
y
is
d
em
o
n
s
tr
ated
i
n
th
is
p
ap
er
b
y
t
h
e
M
L
P
2
m
o
d
el
w
it
h
t
w
o
h
id
d
en
la
y
er
s
ac
h
ie
v
es
t
h
e
ac
cu
r
ac
y
o
f
9
7
.
8
2
%
w
h
ile
t
h
e
M
L
P
1
m
o
d
el
ac
h
iev
e
s
th
e
ac
c
u
r
ac
y
o
f
9
5
.
6
2
%.
Su
b
s
eq
u
en
tl
y
,
it
i
s
o
b
s
er
v
ed
th
at
in
cr
ea
s
i
n
g
th
e
n
u
m
b
er
o
f
h
i
d
d
en
la
y
er
s
m
a
y
i
m
p
r
o
v
e
th
e
lear
n
i
n
g
r
ate
o
f
A
NNs.
Uti
lit
y
,
t
h
e
p
r
o
p
o
s
ed
ML
P
-
b
ased
I
DS
d
em
o
n
s
tr
ate
a
p
o
w
er
f
u
l
to
o
l f
o
r
s
ec
u
r
it
y
ap
p
licatio
n
s
.
Fu
r
t
h
er
in
v
esti
g
atio
n
is
r
eq
u
ir
ed
in
th
e
f
ea
tu
r
e
s
s
elec
tio
n
.
RE
F
E
R
E
NC
E
S
[1
]
K.
K.
Ch
o
o
,
"
T
h
e
C
y
b
e
r
T
h
re
a
t
L
a
n
d
sc
a
p
e
:
Ch
a
ll
e
n
g
e
s
a
n
d
F
u
tu
re
Re
se
a
r
c
h
Dire
c
ti
o
n
s,"
Co
mp
u
t
e
rs
&
S
e
c
u
rity,
v
o
l.
3
0
,
p
p
.
7
1
9
-
7
3
1
,
2
0
1
1
.
[2
]
B.
A
.
Kh
a
la
f
,
e
t
a
l.
,
"
Co
m
p
re
h
e
n
siv
e
Re
v
ie
w
o
f
A
rti
f
icia
l
In
telli
g
e
n
c
e
a
n
d
S
tatisti
c
a
l
A
p
p
ro
a
c
h
e
s
in
Distrib
u
ted
De
n
ial
o
f
S
e
rv
ice
A
tt
a
c
k
a
n
d
De
fe
n
se
M
e
th
o
d
s,"
I
EE
E
Acc
e
ss
,
v
o
l.
7
,
p
p
.
5
1
6
9
1
-
5
1
7
1
3
,
2
0
1
9
.
[3
]
G
.
A
.
Ja
a
f
a
r,
e
t
a
l.
,
"
Re
v
ie
w
o
f
R
e
c
e
n
t
De
tec
ti
o
n
M
e
th
o
d
s
f
o
r
Http
DD
o
S
A
tt
a
c
k
,
"
J
o
u
rn
a
l
o
f
Co
mp
u
ter
Ne
two
rk
s
a
n
d
Co
mm
u
n
ica
t
io
n
s
,
ID 1
2
8
3
4
7
2
,
p
p
.
1
-
1
0
,
2
0
1
9
.
[4
]
B.
A
.
Kh
a
la
f
,
e
t
a
l.
,
"
A
n
A
d
a
p
ti
v
e
M
o
d
e
l
f
o
r
De
tec
ti
o
n
a
n
d
P
re
v
e
n
t
io
n
o
f
Dd
o
s a
n
d
F
las
h
Cro
w
d
F
l
o
o
d
i
n
g
A
tt
a
c
k
s,"
IEE
E
In
ter
n
a
ti
o
n
a
l
S
y
mp
o
si
u
m o
n
Ag
e
n
t,
M
u
l
ti
-
Ag
e
n
t
S
y
ste
ms
a
n
d
Ro
b
o
ti
c
s (
IS
AM
S
R)
,
p
p
.
1
-
6
,
2
0
1
8
.
[5
]
S
.
T
.
Zarg
a
r,
e
t
a
l.
,
"
S
u
rv
e
y
o
f
De
fe
n
se
M
e
c
h
a
n
ism
s
a
g
a
in
st
Di
strib
u
ted
De
n
ial
o
f
S
e
rv
ice
(D
Do
S
)
F
l
o
o
d
in
g
A
tt
a
c
k
s,
"
IEE
E
c
o
mm
u
n
ica
ti
o
n
s
su
rv
e
y
s
&
tu
to
ria
ls
,
v
o
l.
1
5
,
p
p
.
2
0
4
6
-
2
0
6
9
,
2
0
1
3
.
[6
]
S
.
S
.
S
il
v
a
,
e
t
a
l
.
,
"
Bo
t
n
e
ts:
A
S
u
r
v
e
y
,
"
Co
mp
u
ter
Ne
two
rk
s
,
v
o
l.
5
7
,
p
p
.
3
7
8
-
4
0
3
,
2
0
1
3
.
[7
]
A
.
M
e
h
m
o
o
d
,
e
t
a
l
.
,
"
Nb
c
-
M
a
id
s:
Na
ïv
e
Ba
y
e
sia
n
Clas
si
f
ica
ti
o
n
T
e
c
h
n
iq
u
e
i
n
M
u
lt
i
-
A
g
e
n
t
S
y
ste
m
-
En
rich
e
d
Id
s
f
o
r
S
e
c
u
rin
g
Io
t
a
g
a
in
st Dd
o
s A
tt
a
c
k
s,
"
T
h
e
J
o
u
rn
a
l
o
f
S
u
p
e
rc
o
mp
u
ti
n
g
,
v
o
l
.
7
4
,
p
p
.
5
1
5
6
-
5
1
7
0
,
2
0
1
8
.
[8
]
S
.
A
.
M
o
sta
f
a
,
e
t
a
l.
,
"
A
f
u
z
z
y
lo
g
ic
c
o
n
tro
l
in
a
d
ju
sta
b
le
a
u
t
o
n
o
m
y
o
f
a
m
u
lt
i
-
a
g
e
n
t
s
y
ste
m
f
o
r
a
n
a
u
to
m
a
ted
e
ld
e
rly
m
o
v
e
m
e
n
t
m
o
n
it
o
ri
n
g
a
p
p
li
c
a
ti
o
n
,
"
In
ter
n
a
ti
o
n
a
l
j
o
u
r
n
a
l
o
f
me
d
ica
l
i
n
fo
rm
a
ti
c
s
,
1
1
2
,
1
7
3
-
1
8
4
,
2
0
1
8
.
[9
]
A
.
K.
M
a
lh
i
a
n
d
S
.
Ba
tra,
"
G
e
n
e
ti
c
‐Ba
se
d
F
ra
m
e
w
o
rk
f
o
r
P
re
v
e
n
ti
o
n
o
f
M
a
sq
u
e
ra
d
e
a
n
d
DD
o
S
A
tt
a
c
k
s
in
V
e
h
icu
lar A
d
‐Ho
c
n
e
tw
o
rk
s,"
S
e
c
u
rity a
n
d
C
o
mm
u
n
ica
ti
o
n
Ne
tw
o
rk
s
,
v
o
l.
9
,
p
p
.
2
6
1
2
-
2
6
2
6
,
2
0
1
6
.
[1
0
]
M
.
W
a
n
g
,
e
t
a
l.
,
"
A
D
y
n
a
m
i
c
M
l
p
-
Ba
se
d
Dd
o
s
A
tt
a
c
k
D
e
tec
ti
o
n
M
e
th
o
d
Us
in
g
F
e
a
tu
re
S
e
lec
ti
o
n
a
n
d
F
e
e
d
b
a
c
k
,
"
Co
mp
u
ter
s &
S
e
c
u
rity
,
v
o
l.
8
8
,
p
.
1
0
1
6
4
5
,
2
0
2
0
.
[1
1
]
B.
G
u
p
ta,
a
n
d
O.P
.
Ba
d
v
e
,
"
G
a
r
c
h
a
n
d
A
n
n
-
Ba
se
d
Dd
o
s
De
tec
ti
o
n
a
n
d
F
il
terin
g
in
Cl
o
u
d
C
o
m
p
u
ti
n
g
En
v
iro
n
m
e
n
t,
"
In
ter
n
a
ti
o
n
a
l
J
o
u
r
n
a
l
o
f
Em
b
e
d
d
e
d
S
y
ste
ms
,
v
o
l.
9
,
p
p
.
3
9
1
-
4
0
0
,
2
0
1
7
.
[1
2
]
J.
Ye
,
e
t
a
l
.
,
"
A
DD
o
S
A
tt
a
c
k
De
tec
ti
o
n
M
e
th
o
d
Ba
se
d
o
n
S
VM
in
S
o
f
tw
a
re
De
f
in
e
d
Ne
t
w
o
rk
,
"
S
e
c
u
rity
a
n
d
Co
mm
u
n
ic
a
ti
o
n
Ne
two
rk
s
,
ID
9
8
0
4
0
6
1
p
p
.
1
-
8
,
2
0
1
8
.
[1
3
]
B.
In
g
re
a
n
d
A
.
Ya
d
a
v
,
"
P
e
rf
o
r
m
a
n
c
e
A
n
a
l
y
sis
o
f
Ns
l
-
Kd
d
Da
tas
e
t
Us
in
g
A
n
n
,
"
IEE
E
In
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
S
ig
n
a
l
Pro
c
e
ss
in
g
a
n
d
C
o
mm
u
n
ica
ti
o
n
E
n
g
i
n
e
e
rin
g
S
y
ste
ms
,
p
p
.
9
2
-
9
6
,
2
0
1
5
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
A
r
ti
f
I
n
tell
I
SS
N:
2252
-
8938
A
mu
ltil
a
ye
r
p
ercep
tr
o
n
a
r
tifi
c
ia
l n
eu
r
a
l n
etw
o
r
k
a
p
p
r
o
a
ch
fo
r
…
(
A
b
d
u
lr
a
h
ma
n
Ja
s
s
a
m
Mo
h
a
mme
d
)
615
[1
4
]
V
.
D.
M
a
n
e
a
n
d
S
.
P
a
w
a
r,
"
A
n
o
m
a
l
y
Ba
s
e
d
Id
s
Us
in
g
Ba
c
k
p
ro
p
a
g
a
ti
o
n
Ne
u
ra
l
Ne
tw
o
rk
,
"
In
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.
1
3
6
,
p
p
.
2
9
-
3
4
,
2
0
1
6
.
[1
5
]
C.
-
L
.
T
s
a
i,
e
t
a
l.
,
"
Early
Warn
in
g
S
y
ste
m
f
o
r
Dd
o
s
A
tt
a
c
k
in
g
B
a
se
d
o
n
M
u
lt
il
a
y
e
r
De
p
lo
y
m
e
n
t
o
f
T
i
m
e
De
la
y
Ne
u
ra
l
Ne
t
w
o
rk
,
"
IEE
E
In
ter
n
a
ti
o
n
a
l
C
o
n
fer
e
n
c
e
o
n
I
n
telli
g
e
n
t
In
f
o
rm
a
ti
o
n
Hid
i
n
g
a
n
d
M
u
lt
ime
d
ia
S
ig
n
a
l
Pro
c
e
ss
in
g
,
p
p
.
7
0
4
-
7
0
7
,
2
0
1
0
.
[1
6
]
K.
J.
S
in
g
h
a
n
d
T
.
De
,
"
M
L
P
-
GA
Ba
se
d
A
lg
o
rit
h
m
to
De
te
c
t
A
p
p
li
c
a
ti
o
n
L
a
y
e
r
DD
o
S
A
tt
a
c
k
,
"
J
o
u
rn
a
l
o
f
in
fo
rm
a
ti
o
n
se
c
u
rity a
n
d
a
p
p
li
c
a
t
io
n
s
,
v
o
l.
3
6
,
p
p
.
1
4
5
-
1
5
3
,
2
0
1
7
.
[1
7
]
NSL
-
KD
D d
a
tas
e
t
[
o
n
li
n
e
]
a
v
a
il
a
b
le:
h
t
tp
s://
w
ww
.
u
n
b
.
c
a
/cic
/d
a
tas
e
ts/n
sl.h
tm
l,
a
c
c
e
ss
e
d
Da
te
A
c
c
e
s
se
d
o
n
:
2
0
0
9
.
[1
8
]
Hu
a
n
g
,
e
t
a
l
.
,
"
A
d
istri
b
u
ted
P
S
O
–
S
V
M
h
y
b
rid
sy
ste
m
w
it
h
fe
a
tu
re
se
lec
ti
o
n
a
n
d
p
a
ra
m
e
ter
o
p
ti
m
iz
a
ti
o
n
,
"
A
p
p
li
e
d
so
ft
c
o
mp
u
ti
n
g
,
v
o
l.
8
,
p
p
.
1
3
8
1
-
1
3
9
1
,
2
0
0
8
.
[1
9
]
M
.
T
a
v
a
ll
a
e
e
,
e
t
a
l.
,
"
A
De
taile
d
A
n
a
ly
sis
o
f
th
e
Kd
d
C
u
p
9
9
Da
ta
S
e
t,
"
IEE
E
S
y
mp
o
siu
m
o
n
Co
mp
u
ta
t
io
n
a
l
In
telli
g
e
n
c
e
fo
r S
e
c
u
rity
a
n
d
De
fen
se
Ap
p
li
c
a
ti
o
n
s
,
p
p
.
1
-
6
,
2
0
0
9
.
[2
0
]
A
.
R.
W
a
n
i,
e
t
a
l.
,
"
A
n
a
l
y
sis
a
n
d
De
tec
ti
o
n
o
f
DD
o
S
A
tt
a
c
k
s
o
n
Clo
u
d
C
o
m
p
u
ti
n
g
E
n
v
iro
n
m
e
n
t
Us
in
g
M
a
c
h
in
e
L
e
a
rn
in
g
Tec
h
n
iq
u
e
s,"
IEE
E
Ami
ty In
ter
n
a
t
io
n
a
l
C
o
n
fer
e
n
c
e
o
n
Ar
ti
fi
c
ia
l
I
n
telli
g
e
n
c
e
(
AICA
I)
,
p
p
.
8
7
0
-
8
7
5
,
2
0
1
9
.
[2
1
]
S
.
A
.
M
o
sta
f
a
,
e
t
a
l.
,
"
Ev
a
lu
a
ti
n
g
th
e
p
e
rf
o
rm
a
n
c
e
o
f
th
re
e
c
la
ss
if
ic
a
ti
o
n
m
e
th
o
d
s
in
d
iag
n
o
sis
o
f
P
a
rk
in
so
n
’s
d
ise
a
se
,
"
In
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
S
o
ft
C
o
mp
u
ti
n
g
a
n
d
Da
ta
M
in
in
g
,
p
p
.
4
3
-
5
2
,
S
p
rin
g
e
r,
Ch
a
m
,
2
0
1
8
.
[2
2
]
J.
A
.
Ju
p
in
,
e
t
a
l
.
,
"
Re
v
ie
w
o
f
th
e
m
a
c
h
in
e
lea
rn
in
g
m
e
th
o
d
s
in
t
h
e
c
las
si
f
ica
ti
o
n
o
f
p
h
ish
in
g
a
tt
a
c
k
,
"
.
Bu
ll
e
ti
n
o
f
El
e
c
trica
l
En
g
in
e
e
rin
g
a
n
d
In
f
o
r
ma
ti
c
s
,
v
o
l.
8
,
n
o
.
4
,
1
5
4
5
-
1
5
5
5
,
2
0
1
9
.
[2
3
]
S
.
A
.
M
o
sta
f
a
,
e
t
a
l.
,
"
S
o
c
ial
n
e
tw
o
rk
in
g
m
o
b
il
e
a
p
p
s
f
ra
m
e
w
o
rk
f
o
r
o
rg
a
n
izin
g
a
n
d
f
a
c
il
it
a
ti
n
g
c
h
a
rit
a
b
le
a
n
d
v
o
lu
n
tary
a
c
ti
v
it
ies
in
M
a
la
y
sia
,
"
Bu
ll
e
ti
n
o
f
El
e
c
trica
l
E
n
g
i
n
e
e
rin
g
a
n
d
I
n
fo
rm
a
t
ics
,
v
o
l.
9
,
n
o
.
2
,
2
0
2
0
.
[2
4
]
M
.
A
.
Na
a
g
a
s,
e
t
a
l.
,
"
De
f
e
n
se
-
th
ro
u
g
h
-
De
c
e
p
ti
o
n
Ne
tw
o
rk
S
e
c
u
rit
y
M
o
d
e
l:
S
e
c
u
rin
g
U
n
iv
e
rsity
C
a
m
p
u
s
Ne
tw
o
rk
f
ro
m
Do
S
/DDo
S
A
tt
a
c
k
,
"
Bu
ll
e
ti
n
o
f
El
e
c
trica
l
E
n
g
i
n
e
e
rin
g
a
n
d
I
n
fo
rm
a
t
ics
,
v
o
l.
7
,
n
o
.
4
,
5
9
3
-
6
0
0
,
2
0
1
8
.
[2
5
]
B.
A
.
Kh
a
la
f
,
e
t
a
l.
,
"
A
S
i
m
u
lati
o
n
S
t
u
d
y
o
f
S
y
n
F
lo
o
d
A
tt
a
c
k
I
n
Clo
u
d
Co
m
p
u
ti
n
g
En
v
iro
n
m
e
n
t,
"
AUS
jo
u
rn
a
l
,
v
o
l.
2
6
,
p
p
.
1
,
2
0
2
0
.
[2
6
]
M
.
A
.
Ju
b
a
ir,
e
t
a
l.
,
"
A
S
u
rv
e
y
o
f
M
u
lt
i
-
a
g
e
n
t
S
y
ste
m
s
a
n
d
Ca
se
-
Ba
se
d
Re
a
so
n
in
g
In
teg
ra
ti
o
n
,
"
In
ter
n
a
ti
o
n
a
l
S
y
mp
o
si
u
m o
n
A
g
e
n
t,
M
u
lt
i
-
A
g
e
n
t
S
y
ste
ms
a
n
d
R
o
b
o
ti
c
s (
IS
AM
S
R)
,
p
p
.
1
-
6
.
IEE
E,
2
0
1
8.
[2
7
]
M
.
A
.
M
o
h
a
m
m
e
d
,
e
t
a
l.
,
"
A
n
a
n
ti
-
sp
a
m
d
e
tec
ti
o
n
m
o
d
e
l
f
o
r
e
m
a
il
s
o
f
m
u
lt
i
-
n
a
tu
ra
l
la
n
g
u
a
g
e
,
"
J
o
u
rn
a
l
o
f
S
o
u
th
we
st Jia
o
t
o
n
g
Un
ive
rs
it
y
5
4
,
n
o
.
3
,
2
0
1
9
.
[2
8
]
M
.
A
.
Ju
b
a
ir,
e
t
a
l.
,
"
Ba
t
Op
ti
m
ize
d
L
in
k
S
tate
Ro
u
ti
n
g
P
r
o
to
c
o
l
f
o
r
En
e
rg
y
-
Aw
a
re
M
o
b
il
e
A
d
-
Ho
c
Ne
t
w
o
rk
s,"
S
y
mm
e
try
1
1
,
n
o
.
1
1
(2
0
1
9
):
1
4
0
9
.
[2
9
]
M
.
A
.
Ju
b
a
ir,
e
t
a
l.
,
"
Co
m
p
e
ti
ti
v
e
A
n
a
l
y
sis
o
f
S
in
g
le
a
n
d
M
u
lt
i
-
P
a
th
R
o
u
ti
n
g
P
r
o
to
c
o
ls
i
n
M
o
b
i
le
A
d
-
Ho
c
Ne
tw
o
rk
,
"
In
d
o
n
e
si
a
n
J
o
u
rn
a
l
o
f
El
e
c
trica
l
En
g
in
e
e
rin
g
a
n
d
Co
mp
u
ter
S
c
ien
c
e
,
1
4
,
n
o
.
2
,
2
0
1
9
.
[3
0
]
M
.
H.
Ha
ss
a
n
,
e
t
a
l.
,
"
A
S
tatisti
c
a
l
Risk
A
s
se
ss
m
e
n
t
M
e
th
o
d
Of
Dy
n
a
m
i
c
En
v
iro
n
m
e
n
ts:
A
Ca
se
S
tu
d
y
O
f
A
ir
P
o
ll
u
ti
o
n
,
"
AUS
jo
u
rn
a
l
,
v
o
l
.
2
6
,
p
p
.
1
,
2
0
2
0
.
[3
1
]
A
.
S
.
A
l
-
Kh
a
le
e
f
a
,
e
t
a
l.
,
"
F
e
a
t
u
re
A
d
a
p
ti
v
e
a
n
d
C
y
c
li
c
D
y
n
a
m
i
c
L
e
a
rn
in
g
Ba
se
d
o
n
In
f
in
it
e
Ter
m
M
e
m
o
r
y
Ex
tre
m
e
L
e
a
rn
in
g
M
a
c
h
in
e
,
"
Ap
p
li
e
d
S
c
ien
c
e
s
,
v
o
l
9
,
n
o
.
5
,
p
p
:
8
9
5
,
2
0
1
9
.
Evaluation Warning : The document was created with Spire.PDF for Python.