I
nd
o
ne
s
ia
n J
o
urna
l o
f
E
lect
rica
l En
g
ineering
a
nd
Co
m
pu
t
er
Science
Vo
l.
36
,
No
.
2
,
N
o
v
em
b
e
r
20
24
,
p
p
.
1
1
8
2
~1
188
I
SS
N:
2
502
-
4
7
52
,
DOI
: 1
0
.
1
1
5
9
1
/ijee
cs
.v
36.
i
2
.
p
p
1
1
8
2
-
1
1
8
8
1182
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ee
cs
.
ia
esco
r
e.
co
m
Enha
ncing
phish
i
ng
URL
de
tec
tion throug
h com
pr
e
hensiv
e
feature
select
io
n
:
a
com
pa
ra
tive a
na
ly
sis
acro
ss
dive
rse da
tas
ets
P
re
et
i,
P
rit
i Sha
rm
a
D
e
p
a
r
t
me
n
t
o
f
C
o
mp
u
t
e
r
S
c
i
e
n
c
e
a
n
d
A
p
p
l
i
c
a
t
i
o
n
s,
M
a
h
a
r
s
h
i
D
a
y
a
n
a
n
d
U
n
i
v
e
r
s
i
t
y
,
R
o
h
t
a
k
H
a
r
y
a
n
a
,
I
n
d
i
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Mar
14
,
2
0
2
4
R
ev
is
ed
J
u
l
30
,
2
0
2
4
Acc
ep
ted
A
ug
5
,
2
0
2
4
M
a
l
i
c
i
o
u
s
a
t
ta
c
k
s
h
a
v
e
d
e
v
e
l
o
p
e
d
a
p
r
o
m
i
n
e
n
t
r
i
s
k
t
o
t
h
e
s
a
fe
t
y
o
f
o
n
l
i
n
e
u
s
e
r
s
,
w
i
t
h
a
t
t
a
c
k
e
r
s
e
m
p
l
o
y
i
n
g
i
n
c
r
e
a
s
i
n
g
l
y
s
o
p
h
i
s
t
i
c
a
t
e
d
s
y
s
t
e
m
s
t
o
d
e
c
e
i
v
e
u
n
s
u
s
p
e
c
t
i
n
g
v
i
c
t
i
m
s
.
T
h
i
s
r
e
s
e
a
rc
h
f
o
c
u
s
e
s
o
n
t
h
e
c
r
i
t
ic
a
l
a
s
p
e
c
t
o
f
f
e
a
t
u
r
e
s
e
l
e
c
t
i
o
n
i
n
o
p
t
i
m
i
z
i
n
g
p
h
i
s
h
i
n
g
u
n
i
f
o
r
m
r
e
s
o
u
r
c
e
l
o
c
a
t
o
r
(
U
R
L
)
d
e
t
e
c
t
i
o
n
s
y
s
t
e
m
.
F
e
a
t
u
re
s
e
l
e
c
t
i
o
n
b
o
o
s
t
s
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
(
M
L
)
a
n
d
d
e
e
p
l
e
a
r
n
i
n
g
(
D
L
)
b
y
p
i
c
k
i
n
g
v
i
t
a
l
a
t
t
r
i
b
u
t
e
s
e
f
f
i
c
ie
n
t
l
y
.
T
h
i
s
r
e
se
a
rc
h
p
a
p
e
r
p
r
o
v
i
d
e
s
a
c
o
m
p
r
e
h
e
n
s
i
v
e
e
x
a
m
i
n
a
t
i
o
n
o
f
f
e
a
t
u
r
e
se
l
e
c
t
i
o
n
t
e
c
h
n
i
q
u
e
s
u
s
i
n
g
f
iv
e
d
i
v
e
r
s
e
d
a
t
a
s
e
ts
.
V
a
r
i
o
u
s
m
e
t
h
o
d
s
,
i
n
c
l
u
d
i
n
g
r
a
n
d
o
m
f
o
r
e
s
t
(
R
F
)
s
e
l
e
c
t
f
r
o
m
m
o
d
e
l
,
S
e
lec
tKBe
st
with
c
h
i
-
sq
u
a
re
sta
t
isti
c
,
p
ri
n
c
ip
a
l
c
o
m
p
o
n
e
n
t
a
n
a
ly
sis
(P
CA)
a
n
d
re
c
u
rsiv
e
fe
a
tu
re
e
li
m
in
a
ti
o
n
(RF
E)
,
we
re
e
m
p
lo
y
e
d
.
T
h
e
e
x
p
e
rime
n
ts,
with
a
p
a
rti
c
u
lar
e
m
p
h
a
sis
o
n
P
C
A
a
n
d
fo
u
r
th
d
a
tas
e
t,
re
v
e
a
led
th
a
t
a
ll
fo
u
r
m
o
d
e
ls
RF
,
d
e
c
isi
o
n
tree
s
(DT
s)
,
XG
Bo
o
st,
a
n
d
m
u
lt
i
lay
e
r
p
e
rc
e
p
tro
n
)
a
c
h
iev
e
d
1
0
0
%
a
c
c
u
ra
c
y
i
n
d
e
tec
ti
n
g
p
h
ish
i
n
g
URL
a
tt
a
c
k
s.
T
h
i
s
u
n
d
e
rsc
o
re
s
t
h
e
e
ffica
c
y
o
f
fe
a
t
u
re
se
lec
ti
o
n
m
e
th
o
d
s
in
e
n
h
a
n
c
in
g
t
o
a
d
e
e
p
e
r
u
n
d
e
rsta
n
d
i
n
g
o
f
fe
a
tu
re
se
lec
ti
o
n
’s
r
o
le
in
b
o
lst
e
rin
g
t
h
e
e
ffe
c
ti
v
e
n
e
ss
o
f
p
h
ish
i
n
g
d
e
t
e
c
ti
o
n
sy
ste
m
a
c
ro
ss
d
i
v
e
rse
d
a
tas
e
ts,
h
ig
h
li
g
h
ti
n
g
t
h
e
imp
o
r
tan
c
e
o
f
lev
e
ra
g
in
g
tec
h
n
i
q
u
e
s
su
c
h
a
s
P
CA
fo
r
o
p
ti
m
a
l
re
su
lt
s
.
K
ey
w
o
r
d
s
:
Dec
is
io
n
tr
ee
Featu
r
e
s
elec
tio
n
tech
n
iq
u
e
Ph
is
h
in
g
attac
k
R
an
d
o
m
f
o
r
est
UR
L
clas
s
if
icatio
n
XGBo
o
t a
n
d
MLP
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
:
Pre
eti
Dep
ar
tm
en
t o
f
C
o
m
p
u
ter
Scie
n
ce
an
d
Ap
p
licatio
n
s
,
Ma
h
ar
s
h
i D
ay
an
an
d
Un
iv
er
s
ity
R
o
h
tak
Har
y
an
a,
I
n
d
ia
E
m
ail: m
is
k
h
o
k
h
ar
1
2
1
@
g
m
ai
l.c
o
m
1.
I
NT
RO
D
UCT
I
O
N
P
h
i
s
h
i
n
g
a
t
t
a
c
k
s
,
u
ti
l
i
zi
n
g
d
e
ce
p
t
i
v
e
u
n
i
f
o
r
m
r
e
s
o
u
r
c
e
l
o
c
a
t
o
r
s
(
UR
L
s
)
,
e
p
i
t
o
m
iz
e
s
o
p
h
i
s
ti
ca
t
e
d
c
y
b
e
r
d
e
c
e
p
t
i
o
n
,
e
m
p
l
o
y
i
n
g
i
m
i
t
a
t
i
o
n
w
e
b
s
e
r
v
e
r
s
t
o
i
l
l
i
ci
t
l
y
a
c
q
u
ir
e
s
e
n
s
it
i
v
e
i
n
f
o
r
m
a
t
i
o
n
T
h
i
s
s
t
u
d
y
e
x
a
m
i
n
e
d
h
o
w
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
(
M
L
)
a
l
g
o
r
ith
m
i
m
p
a
c
t
t
h
e
d
e
t
ec
t
i
o
n
o
f
p
h
is
h
i
n
g
a
t
t
ac
k
s
[
1
]
.
A
lt
h
o
u
g
h
p
r
e
v
i
o
u
s
r
e
s
ea
r
c
h
h
a
s
l
o
o
k
e
d
i
n
t
o
t
r
a
d
i
t
i
o
n
a
l
d
e
t
ec
t
io
n
m
e
t
h
o
d
s
,
i
t
h
a
s
n
o
t
s
p
e
ci
f
ic
a
l
l
y
a
n
a
l
y
z
e
d
h
o
w
a
d
v
a
n
c
e
d
ML
t
e
c
h
n
i
q
u
e
s
c
a
n
e
n
h
a
n
c
e
t
h
e
a
c
c
u
r
a
c
y
a
n
d
e
f
f
i
c
i
e
n
c
y
o
f
p
h
i
s
h
i
n
g
p
r
e
v
e
n
t
i
o
n
.
T
h
e
f
i
n
d
i
n
g
r
e
v
e
a
l
s
t
h
a
t
in
t
e
g
r
a
t
i
n
g
ML
c
a
n
s
i
g
n
i
f
i
c
a
n
t
l
y
i
m
p
r
o
v
e
t
h
e
e
f
f
e
cti
v
e
n
e
s
s
o
f
d
e
t
e
c
ti
n
g
a
n
d
p
r
e
v
e
n
t
i
n
g
p
h
i
s
h
i
n
g
a
t
te
m
p
t
s
.
T
h
is
r
esear
ch
d
ev
elo
p
s
a
co
s
t
-
ef
f
ec
tiv
e
p
h
is
h
in
g
d
etec
t
io
n
s
en
s
o
r
u
s
in
g
d
ee
p
lear
n
in
g
(
DL
)
tech
n
iq
u
es
[
2
]
.
T
r
a
d
itio
n
al
m
eth
o
d
s
r
ely
o
n
u
s
er
r
ep
o
r
t
s
,
b
u
t
r
ec
en
t
ad
v
an
ce
m
en
ts
in
DL
h
av
e
led
to
im
p
r
o
v
e
d
d
etec
tio
n
m
eth
o
d
s
.
T
h
is
p
ap
er
in
tr
o
d
u
ce
s
a
lig
h
tw
eig
h
t
DL
alg
o
r
ith
m
f
o
r
r
ea
l
-
tim
e,
en
er
g
y
-
ef
f
icien
t
p
h
is
h
in
g
d
etec
tio
n
,
d
em
o
n
s
tr
atin
g
en
h
an
ce
d
ac
cu
r
ac
y
a
n
d
p
r
ac
tical
v
iab
ilit
y
o
n
e
m
b
ed
d
e
d
s
y
s
tem
s
.
Mic
r
o
s
o
f
t’
s
atten
tio
n
-
m
ec
h
a
n
is
m
-
b
ased
m
eth
o
d
ex
ce
ls
in
d
etec
tio
n
p
h
is
h
in
g
attac
k
s
b
y
d
y
n
am
ically
weig
h
tin
g
UR
L
co
m
p
o
n
en
t,
E
m
p
h
asizin
g
n
u
an
ce
d
a
n
aly
s
is
[
3
]
.
T
h
is
s
tu
d
y
in
v
esti
g
ated
th
e
e
f
f
ec
ts
o
f
u
s
in
g
a
d
o
u
b
le
d
ee
p
Q
-
Netwo
r
k
(
DDQN
)
f
o
r
we
b
p
h
is
h
in
g
d
etec
tio
n
[
4
]
.
W
h
ile
p
r
ev
i
o
u
s
r
esear
ch
h
as
e
x
a
m
in
e
d
DL
tech
n
iq
u
es
f
o
r
th
is
task
,
it
h
as
n
o
t
s
p
ec
if
ically
ad
d
r
ess
ed
th
e
in
teg
r
atio
n
o
f
d
ee
p
r
ein
f
o
r
ce
m
en
t
lear
n
in
g
wi
th
an
im
b
alan
ce
d
class
if
icatio
n
Ma
r
k
o
v
d
ec
is
io
n
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2
5
0
2
-
4
7
52
E
n
h
a
n
ci
n
g
p
h
is
h
in
g
UR
L d
ete
ctio
n
th
r
o
u
g
h
co
mp
r
eh
en
s
ive
f
ea
tu
r
e
s
elec
tio
n
:
a
co
mp
a
r
a
tive
… (
P
r
ee
ti
)
1183
p
r
o
ce
s
s
(
I
C
MD
P).
T
h
e
s
tu
d
y
d
em
o
n
s
tr
ates
th
at
th
e
DDQN
-
b
ased
class
if
ier
s
ig
n
if
ican
tly
o
u
tp
e
r
f
o
r
m
s
tr
ad
itio
n
al
m
eth
o
d
s
in
h
an
d
lin
g
d
ata
im
b
alan
ce
an
d
im
p
r
o
v
i
n
g
d
etec
tio
n
ac
cu
r
ac
y
.
T
h
is
r
esear
ch
ex
p
lo
r
ed
XGBo
o
s
t
f
o
r
p
h
is
h
in
g
d
etec
tio
n
u
s
in
g
2
2
,
0
0
0
UR
L
s
an
d
2
2
f
ea
tu
r
es
with
NL
P
ev
alu
atio
n
.
W
h
ile
p
r
ev
i
o
u
s
Stu
d
ies
ex
am
in
ed
v
ar
io
u
s
tech
n
iq
u
es,
th
ey
d
i
d
n
’
t
f
o
c
u
s
o
n
tex
t
-
as
-
im
ag
e
r
ep
r
esen
tatio
n
’
s
im
p
ac
t.
T
h
e
m
o
d
el
ac
h
iev
ed
9
4
%
ac
cu
r
ac
y
an
d
9
1
%
p
r
ec
is
io
n
[
5
]
.
T
h
e
p
ap
er
ex
p
lo
r
es
ML
f
o
r
p
h
is
h
in
g
d
o
m
ain
d
etec
tio
n
,
d
ev
el
o
p
in
g
an
d
c
o
m
p
ar
i
n
g
m
o
d
els
b
ased
o
n
s
u
p
p
o
r
t
v
ec
t
o
r
m
ac
h
in
e
(
SVM
)
,
DT
,
ANN
,
an
d
r
an
d
o
m
f
o
r
est
(
RF
)
.
Usi
n
g
th
e
U
C
I
p
h
is
h
in
g
d
o
m
ain
s
d
ataset,
th
e
f
in
d
in
g
r
ev
ea
l
th
at
th
e
RF
m
o
d
el
o
u
tp
er
f
o
r
m
s
o
th
er
s
,
p
r
o
v
in
g
its
s
u
p
er
io
r
ity
o
v
er
ex
is
tin
g
s
o
lu
tio
n
s
[
6
]
.
Pre
s
en
ts
a
in
n
o
v
ativ
e
ap
p
r
o
ac
h
f
o
r
p
h
is
h
in
g
web
s
ites
class
if
icatio
n
u
s
in
g
co
n
v
o
lu
tio
n
a
l
n
eu
r
al
n
etwo
r
k
s
(
C
NN
)
with
UR
L
-
b
ased
f
ea
tu
r
es.
T
h
e
C
NNs,
em
p
lo
y
in
g
e
n
tr
o
p
y
lo
s
s
f
u
n
ctio
n
an
d
R
eL
U
to
ad
d
r
ess
v
an
is
h
in
g
g
r
ad
ien
t,
ac
h
iev
e
a
n
8
6
.
5
%
class
if
icatio
n
ac
cu
r
ac
y
o
n
a
d
ataset
o
f
1
,
3
5
3
UR
L
s
[
7
]
.
T
h
is
s
u
r
v
e
y
ev
alu
ated
m
u
ltip
le
ML
alg
o
r
i
th
m
s
f
o
r
p
h
i
s
h
in
g
d
etec
tio
n
,
s
u
ch
as
SVM,
Naïv
e
B
ay
es
(
NB
)
,
d
ec
is
io
n
tr
ee
(
DT
)
an
d
RF
[
8
]
.
Alth
o
u
g
h
p
r
ev
io
u
s
s
tu
d
ies
em
p
lo
y
ed
t
h
ese
m
eth
o
d
s
,
th
ey
d
id
n
o
t
s
p
ec
if
ically
co
v
er
n
ewe
r
s
y
s
tem
s
lik
e
Ph
is
h
Sco
r
e
an
d
Ph
is
h
C
h
ec
k
er
.
T
h
e
r
esear
ch
s
u
m
m
ar
ized
ef
f
ec
tiv
e
tech
n
i
q
u
es
an
d
em
p
h
asized
u
p
d
atin
g
f
ea
tu
r
es to
a
d
d
r
ess
e
m
er
g
in
g
p
h
is
h
in
g
th
r
ea
ts
.
E
x
p
lo
r
es
m
o
d
er
n
ML
tec
h
n
i
q
u
e
f
o
r
d
etec
tio
n
p
h
is
h
in
g
a
ttack
s
with
h
ig
h
ac
cu
r
ac
y
.
Utilizin
g
a
Kag
g
le
d
ataset
o
f
o
v
er
1
1
,
0
0
0
web
s
ites
,
th
e
s
tu
d
y
ass
ess
e
s
3
0
web
s
ite
f
ea
tu
r
es,
u
s
in
g
n
eu
r
al
ne
two
r
k
,
NB
,
an
d
Ad
a
B
o
o
s
t
m
o
d
els.
R
esu
lt
s
h
o
w
ac
cu
r
ac
ies
o
f
9
0
.
2
3
%,
9
2
.
9
7
%,
an
d
9
5
.
4
3
%,
co
r
r
esp
o
n
d
i
n
g
ly
[
9
]
.
Pre
s
en
ted
a
r
ec
o
g
n
itio
n
way
u
tili
zin
g
n
in
e
lex
ical
f
ea
tu
r
es
,
r
ea
ch
in
g
a
9
9
.
5
7
%
ac
cu
r
ac
y
u
tili
zin
g
th
e
R
F
m
o
d
el
o
n
th
e
I
SC
XURL
-
2
0
1
6
d
ataset
co
m
p
r
is
in
g
1
1
,
9
6
4
in
s
tan
ce
s
[
1
0
]
.
Used
ML
tec
h
n
iq
u
es
(
XGBo
o
s
t,
DT
,
lo
g
is
tic
r
eg
r
ess
io
n
(
L
R
)
,
RF
,
an
d
SVM)
,
th
e
r
esea
r
ch
i
d
en
tifie
s
p
h
is
h
in
g
web
s
ites
.
RF
ac
h
iev
es
9
8
.
9
0
%
an
d
9
7
.
8
7
% a
cc
u
r
ac
y
o
n
Ph
is
h
-
T
an
k
an
d
UC
I
d
atasets
[
1
1
]
.
I
n
th
e
co
n
d
u
cted
s
tu
d
y
,
a
p
i
o
n
ee
r
in
g
p
h
is
h
in
g
d
etec
tio
n
m
eth
o
d
ac
h
ie
v
ed
ac
cu
r
ac
y
,
s
u
r
p
ass
in
g
p
r
ev
io
u
s
m
o
d
els.
A
d
ataset
o
f
1
0
,
0
0
0
m
alicio
u
s
UR
L
s
a
n
d
leg
itima
te
s
ites
wa
s
u
tili
ze
d
,
an
d
o
u
r
in
teg
r
ated
C
NN
-
b
ased
m
o
d
el
ex
ce
lled
w
ith
a
9
8
.
7
7
%
ac
cu
r
ac
y
,
b
en
e
f
itin
g
f
r
o
m
d
ee
p
e
r
lay
er
a
n
d
ad
d
itio
n
al
f
ea
tu
r
es
f
o
r
en
h
an
ce
d
p
er
f
o
r
m
an
ce
[
1
2
]
.
Pre
s
en
ted
em
p
h
asizin
g
tec
h
n
iq
u
es
f
o
r
p
r
ev
e
n
tio
n
r
at
h
er
th
an
m
itig
atio
n
.
Pro
v
id
in
g
a
g
en
er
al
o
v
er
v
iew,
it
h
ig
h
lig
h
ts
DL
as
a
k
ey
s
tr
ateg
y
in
ef
f
ec
tiv
e
p
h
is
h
in
g
att
ac
k
d
etec
tio
n
[
1
3
]
.
I
n
tr
o
d
u
ce
d
th
r
ee
DL
tech
n
iq
u
es
f
o
r
p
h
is
h
in
g
we
b
s
ite
id
en
tific
atio
n
.
T
h
e
ex
p
e
r
im
en
tal
o
u
tco
m
es
in
d
icate
im
p
r
ess
iv
e
ac
cu
r
ac
y
r
ate
o
f
9
6
.
8
%,
9
9
.
2
%
,
an
d
9
7
.
6
%
f
o
r
lo
n
g
s
h
o
r
t
-
ter
m
m
em
o
r
y
(
L
STM
)
,
C
NN
an
d
L
STM
-
C
NN
m
o
d
els,
co
r
r
esp
o
n
d
in
g
ly
.
T
h
e
C
NN
-
b
ased
s
y
s
tem
p
r
o
v
es su
p
er
io
r
i
n
p
h
is
h
in
g
d
etec
tio
n
[
1
4
]
.
Pre
s
en
ted
a
“Ph
is
h
Der
b
y
”
c
o
m
p
etitio
n
at
U.
S.
u
n
iv
e
r
s
ity
to
g
am
if
y
p
h
is
h
in
g
s
ec
u
r
ity
awa
r
en
ess
tr
ain
in
g
.
Fin
d
in
g
h
ig
h
lig
h
ted
r
elatio
n
s
h
ip
b
etwe
en
d
em
o
g
r
ap
h
ics,
p
er
s
o
n
ality
tr
aits
,
g
o
al
o
r
ien
tatio
n
,
an
d
p
h
is
h
in
g
d
etec
tio
n
p
er
f
o
r
m
an
ce
.
I
n
s
ig
h
ts
s
tr
ess
f
o
s
ter
in
g
p
o
s
itiv
e
cy
b
er
b
e
h
av
io
r
s
b
ey
o
n
d
click
r
ates
in
o
r
g
an
izatio
n
al
t
r
ain
in
g
c
u
ltu
r
e
s
[
1
5
]
.
Pro
p
o
s
ed
a
n
o
v
el
clien
t
-
s
id
e
tech
n
iq
u
e
f
o
r
e
f
f
o
r
tles
s
p
h
is
h
in
g
web
s
ite
id
en
tific
atio
n
u
s
in
g
a
r
ed
esig
n
ed
b
r
o
wser
ar
ch
itectu
r
e.
A
RF
class
if
icatio
n
m
o
d
el
an
aly
ze
s
3
0
UR
L
p
r
o
p
er
ties
ex
tr
ac
ted
th
r
o
u
g
h
a
r
u
le
o
f
ex
tr
ac
tio
n
f
r
am
ewo
r
k
.
T
h
e
‘
em
b
ed
d
e
d
p
h
is
h
in
g
d
etec
tio
n
b
r
o
wser
’
(
E
PDB
)
in
teg
r
ates
p
h
is
h
in
g
d
etec
tio
n
with
o
u
t c
o
m
p
r
o
m
is
in
g
u
s
er
ex
p
er
ien
ce
,
ac
h
iev
in
g
a
r
ea
l
-
tim
e
ac
cu
r
ac
y
o
f
9
9
.
3
6
% [
1
6
]
.
T
h
is
s
tu
d
y
in
v
esti
g
ates
th
e
ef
f
icac
y
o
f
h
y
b
r
id
L
STM
an
d
C
NN
DL
m
o
d
els
f
o
r
f
ak
e
web
s
ite
U
R
L
d
etec
tio
n
,
lev
er
a
g
in
g
th
e
s
tr
en
g
th
s
o
f
b
o
th
ap
p
r
o
ac
h
es
[
1
7
]
.
Usi
n
g
two
p
u
b
licly
ex
is
tin
g
d
atasets
,
th
e
h
y
b
r
id
m
o
d
el
ac
h
ie
v
ed
s
ig
n
if
ica
n
tly
h
ig
h
er
ac
cu
r
ac
ies
co
m
p
ar
e
d
to
s
tan
d
alo
n
e
C
NN
an
d
L
S
T
M
m
o
d
els.
T
h
ese
co
n
clu
s
io
n
s
s
u
g
g
est th
e
laten
t
o
f
h
y
b
r
id
DL
tec
h
n
iq
u
es in
m
i
tig
atin
g
lo
s
s
es f
r
o
m
s
p
o
o
f
in
g
attac
k
s
.
T
h
is
ar
ticle
p
r
esen
ts
a
n
ex
p
er
i
m
en
tal
s
tu
d
y
en
h
a
n
cin
g
ML
m
o
d
el
p
er
f
o
r
m
an
ce
f
o
r
m
alici
o
u
s
d
ataset
[
1
8
]
.
I
t
ex
p
l
o
r
es
h
y
p
er
p
ar
a
m
eter
o
p
tim
izatio
n
,
d
ata
b
alan
cin
g
an
d
f
ea
tu
r
e
s
elec
tio
n
,
s
h
o
win
g
s
ig
n
if
ican
t
ac
cu
r
ac
y
im
p
r
o
v
em
en
t.
C
o
m
b
in
in
g
tu
n
e
d
f
ac
to
r
s
en
h
a
n
ce
s
alg
o
r
ith
m
ef
f
icien
cy
,
with
g
r
a
d
ien
t
b
o
o
s
tin
g
a
n
d
ex
tr
em
e
g
r
ad
ie
n
t
b
o
o
s
tin
g
ac
h
iev
in
g
h
ig
h
ac
c
u
r
ac
y
r
ates
f
o
r
b
o
th
d
atasets
.
I
n
tr
o
d
u
c
ed
a
h
y
b
r
id
f
ea
tu
r
e
-
b
ase
d
an
ti
-
p
h
is
h
in
g
s
tr
ateg
y
,
ac
h
iev
e
d
9
9
.
1
7
% d
etec
tio
n
ac
c
u
r
ac
y
u
s
in
g
XGBo
o
s
t o
n
clien
t
-
s
id
e
UR
L
an
d
h
y
p
er
lin
k
d
ata
[
1
9
]
.
T
h
is
p
ap
er
s
h
o
wca
s
es
a
s
y
s
tem
atic
ap
p
r
o
ac
h
t
o
co
n
s
tr
u
ctin
g
d
etec
tio
n
m
o
d
els
em
p
lo
y
in
g
th
r
ee
DL
ar
ch
itectu
r
es
[
2
0
]
.
Utilizin
g
,
f
u
lly
co
n
n
ec
ted
d
ee
p
n
eu
r
al
n
e
two
r
k
s
(
DNNs)
,
C
N
N
s
an
d
L
STM
,
it
ac
h
iev
ed
a
p
ea
k
ac
cu
r
ac
y
o
f
9
7
.
3
7
%
ac
r
o
s
s
f
o
u
r
p
h
is
h
in
g
web
s
ite
d
atasets
.
Ad
d
itio
n
ally
,
co
m
p
ar
is
o
n
o
f
o
p
tim
izatio
n
alg
o
r
ith
m
s
led
to
ac
c
u
r
ac
y
en
h
an
ce
m
en
ts
o
f
0
.
1
%
-
1
%.
Alm
o
m
an
i
et
a
l.
[
2
1
]
e
m
p
lo
y
1
6
ML
m
o
d
els
with
ten
f
ea
tu
r
es
to
d
etec
t
f
ak
e
we
b
p
a
g
e
f
r
o
m
2
d
atasets
.
Gr
ad
ien
t
b
o
o
s
tin
g
cl
ass
if
ier
(
GB
C
)
ac
h
iev
e
h
ig
h
e
s
t
ac
cu
r
ac
y
(
9
7
%)
wh
ile
Gau
s
s
ian
NB
an
d
SGD
class
if
ier
ex
h
ib
it
lo
west
ac
cu
r
ac
y
(
8
4
%
an
d
8
1
%
r
esp
ec
tiv
ely
)
am
o
n
g
class
if
ier
s
.
I
n
tr
o
d
u
ce
s
a
in
n
o
v
ativ
e
m
alicio
u
s
UR
L
d
etec
tio
n
tec
h
n
iq
u
e
co
m
b
in
in
g
DL
an
d
B
E
R
T
f
ea
tu
r
e
e
x
tr
ac
tio
n
.
B
E
R
T
ex
tr
ac
ts
tex
t
f
r
o
m
UR
L
s
,
NL
P a
lg
o
r
ith
m
s
ex
tr
ac
t m
ea
n
in
g
f
u
l f
ea
tu
r
es,
a
n
d
a
C
NN
m
eth
o
d
d
etec
ts
p
h
is
h
in
g
UR
L
s
.
W
ith
9
6
.
6
6
%
ac
cu
r
ac
y
,
it p
r
o
v
es e
f
f
icien
t in
d
etec
tin
g
p
h
is
h
in
g
web
s
ites
’
UR
L
s
,
v
alid
ated
ag
ain
s
t e
x
is
ti
n
g
liter
atu
r
e
[
2
2
]
.
T
h
is
r
esear
ch
f
o
c
u
s
es
o
n
d
et
ec
tin
g
th
r
ee
-
s
tag
e
p
h
is
h
in
g
attac
k
s
v
ia
co
n
ten
t
a
n
aly
s
is
.
I
n
p
u
t
v
alu
es
in
clu
d
e
UR
L
s
,
tr
af
f
ic,
an
d
web
co
n
ten
t
f
ea
tu
r
es
[
2
3
]
.
R
ea
l
p
h
is
h
in
g
ca
s
es
d
ataset
y
ield
s
h
ig
h
ac
c
u
r
ac
y
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
52
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
36
,
No
.
2
,
No
v
em
b
er
20
24
:
1
1
8
2
-
1
1
8
8
1184
(
9
5
.
1
8
%)
with
NN,
o
u
tp
er
f
o
r
m
in
g
SVM
(
8
5
.
4
5
%)
,
an
d
R
F
(
7
8
.
8
9
%).
ML
p
r
o
v
es
ef
f
ec
tiv
e
f
o
r
p
h
is
h
in
g
d
etec
tio
n
.
T
h
is
s
tu
d
y
in
tr
o
d
u
ce
s
a
n
o
v
el
m
eth
o
d
d
e
v
elo
p
ed
b
y
r
esear
ch
er
s
,
in
teg
r
a
tin
g
DL
f
o
r
UR
L
class
if
icatio
n
with
a
g
e
n
etic
al
g
o
r
ith
m
f
o
r
f
ea
tu
r
e
s
elec
tio
n
[
2
4
]
.
T
h
eir
ap
p
r
o
ac
h
s
ig
n
if
ica
n
tly
im
p
r
o
v
es
r
ec
all
in
UR
L
class
if
icatio
n
,
o
u
tp
er
f
o
r
m
in
g
r
e
ce
n
t
DL
m
eth
o
d
s
with
n
o
tab
le
ac
c
u
r
ac
y
an
d
r
e
ca
ll
en
h
an
ce
m
e
n
ts
.
T
h
is
s
tu
d
y
in
tr
o
d
u
ce
s
a
s
u
p
er
v
is
ed
lear
n
in
g
m
eth
o
d
f
o
r
An
d
r
o
id
m
alwa
r
e
d
etec
tio
n
,
lev
e
r
ag
in
g
a
f
u
ll
lab
eled
d
ataset
o
f
o
v
er
1
8
,
0
0
0
s
am
p
le
s
ac
r
o
s
s
5
ca
teg
o
r
ies
[
2
5
]
.
Val
id
atio
n
ag
ai
n
s
t
estab
lis
h
ed
d
at
asets
d
em
o
n
s
tr
ates
o
u
tp
er
f
o
r
m
an
ce
in
s
p
ec
if
ic
p
ar
am
eter
s
,
co
n
tr
i
b
u
tin
g
to
a
d
v
an
ce
d
tec
h
n
iq
u
es
f
o
r
An
d
r
o
id
d
ev
ice
s
ec
u
r
ity
am
id
s
t e
v
o
lv
in
g
t
h
r
ea
ts
.
2.
SUM
M
AR
I
Z
I
NG
K
E
Y
F
I
N
DING
R
ec
en
t
s
tu
d
ies
h
ig
h
lig
h
t
s
ig
n
if
ican
t
ad
v
an
ce
m
en
ts
in
p
h
is
h
in
g
d
etec
tio
n
th
r
o
u
g
h
ML
an
d
DL
tech
n
iq
u
es.
I
n
teg
r
ati
n
g
DL
m
o
d
els,
s
u
ch
as
lig
h
tweig
h
t
alg
o
r
ith
m
s
an
d
h
y
b
r
id
ap
p
r
o
ac
h
e
s
lik
e
L
STM
-
C
NN,
h
as
m
ar
k
e
d
ly
im
p
r
o
v
ed
ac
cu
r
ac
y
a
n
d
p
r
ac
tical
a
p
p
licab
ilit
y
in
r
ea
l
-
tim
e
d
etec
tio
n
.
Fo
r
in
s
tan
ce
,
RF
an
d
XGBo
o
s
t
m
eth
o
d
s
ac
h
iev
ed
h
ig
h
ac
cu
r
ac
y
r
ates,
with
RF
r
ea
ch
in
g
u
p
to
9
9
.
5
7
%
in
i
d
en
tify
in
g
p
h
is
h
in
g
UR
L
s
.
T
h
e
u
s
e
o
f
ad
v
an
ce
d
tech
n
iq
u
es lik
e
B
E
R
T
f
o
r
f
ea
tu
r
e
ex
tr
ac
tio
n
an
d
g
e
n
etic
alg
o
r
i
th
m
f
o
r
o
p
tim
izin
g
DL
m
o
d
els
h
as
f
u
tu
r
e
en
h
an
c
ed
d
etec
tio
n
ca
p
ab
ilit
ies.
T
h
ese
f
in
d
in
g
d
em
o
n
s
tr
ate
th
at
m
o
d
er
n
ML
an
d
DL
m
eth
o
d
s
s
ig
n
if
ican
tly
o
u
tp
er
f
o
r
m
tr
ad
itio
n
al
ap
p
r
o
ac
h
es,
o
f
f
er
in
g
r
o
b
u
s
t
s
o
lu
tio
n
f
o
r
p
r
ev
e
n
tin
g
an
d
d
etec
tin
g
p
h
is
h
in
g
attac
k
s
.
T
h
is
u
n
d
er
s
co
r
es
th
e
p
o
te
n
tial
o
f
th
ese
tech
n
o
lo
g
ies
to
ad
d
r
ess
ev
o
lv
i
n
g
cy
b
er
s
ec
u
r
ity
th
r
ea
ts
ef
f
ec
tiv
e
ly
.
3.
I
NT
E
RP
R
E
T
I
NG
RE
SUL
T
S
Ou
r
m
eth
o
d
o
lo
g
y
e
n
co
m
p
ass
ed
a
ju
d
icio
u
s
s
elec
tio
n
o
f
a
lg
o
r
ith
m
s
an
d
m
eth
o
d
o
l
o
g
ies
aim
ed
at
m
ax
im
izin
g
ac
c
u
r
ac
y
an
d
r
o
b
u
s
tn
ess
.
W
e
lev
er
ag
ed
a
d
i
v
er
s
e
ar
r
ay
o
f
ML
a
n
d
DL
,
in
cl
u
d
in
g
RF
,
XGBo
o
s
t,
m
u
lti
lay
er
p
er
ce
p
tr
o
n
(
ML
P)
an
d
DT
.
T
h
ese
m
o
d
els
wer
e
ch
o
s
en
f
o
r
t
h
eir
v
er
s
atility
an
d
ef
f
ec
tiv
e
n
ess
in
h
an
d
lin
g
class
if
icatio
n
task
s
ac
r
o
s
s
v
ar
io
u
s
d
o
m
ain
s
.
B
y
tr
ain
in
g
th
ese
m
o
d
els
o
n
f
iv
e
d
is
tin
ct
d
atasets
,
we
aim
ed
to
ca
p
tu
r
e
a
b
r
o
ad
s
p
ec
tr
u
m
o
f
p
atter
n
s
a
n
d
in
tr
ica
cies in
h
er
en
t in
p
h
is
h
in
g
UR
L
s
.
Mo
r
eo
v
er
,
f
ea
tu
r
e
s
elec
tio
n
p
lay
ed
a
p
iv
o
tal
r
o
le
in
en
h
an
cin
g
t
h
e
m
o
d
els’
p
er
f
o
r
m
a
n
ce
.
W
e
em
p
lo
y
ed
a
two
-
p
r
o
n
g
e
d
ap
p
r
o
ac
h
,
u
tili
zin
g
RF
s
e
lect
f
r
o
m
m
o
d
el
alo
n
g
s
id
e
a
co
m
b
in
atio
n
o
f
SelectKBe
s
t
with
th
e
C
h
i
-
s
q
u
ar
e
(
χ²)
s
tatis
tic,
r
ec
u
r
s
iv
e
f
ea
tu
r
e
elim
in
atio
n
(
R
FE)
an
d
p
r
in
cip
al
co
m
p
o
n
e
n
t
an
aly
s
is
(
PC
A)
.
T
h
is
m
eticu
lo
u
s
f
ea
tu
r
e
s
elec
tio
n
p
r
o
ce
s
s
ai
m
ed
to
d
is
til
th
e
m
o
s
t
in
f
o
r
m
a
tiv
e
f
ea
tu
r
es
wh
ile
m
itig
atin
g
th
e
r
is
k
o
f
o
v
e
r
f
itti
n
g
an
d
im
p
r
o
v
in
g
m
o
d
el
i
n
ter
p
r
etab
ilit
y
.
T
o
f
ac
ilit
ate
a
c
o
m
p
r
eh
en
s
iv
e
u
n
d
er
s
tan
d
i
n
g
o
f
o
u
r
f
i
n
d
in
g
,
we
o
r
g
a
n
ized
o
u
r
r
esu
lts
in
to
a
s
tr
ictu
r
es
f
o
r
m
at,
as
ex
em
p
lifie
d
in
T
ab
le
1
(
i
n
Ap
p
en
d
ix
)
.
T
h
is
tab
le
p
r
o
v
id
ed
a
s
u
cc
in
ct
y
et
in
f
o
r
m
ati
v
e
o
v
er
v
iew
o
f
th
e
m
o
d
els’
p
e
r
f
o
r
m
an
ce
m
etr
ics
ac
r
o
s
s
d
if
f
er
e
n
t
d
atasets
,
en
a
b
lin
g
r
ea
d
er
s
to
d
is
ce
r
n
tr
en
d
s
an
d
d
r
aw
in
s
ig
h
ts
.
Fu
r
th
er
m
o
r
e
,
we
h
o
n
ed
in
o
n
th
e
an
aly
s
is
o
f
th
e
f
o
u
r
th
d
a
tasets
to
s
h
o
wca
s
e
th
e
ef
f
icac
y
o
f
o
u
r
p
r
o
p
o
s
ed
m
o
d
el.
N
o
te
b
y
,
th
e
ap
p
licatio
n
o
f
PC
A
f
o
r
f
ea
tu
r
e
s
elec
tio
n
o
f
1
0
0
%.
Su
c
h
f
i
n
d
i
n
g
u
n
d
er
s
co
r
ed
t
h
e
r
o
b
u
s
tn
ess
an
d
r
eliab
ilit
y
o
f
o
u
r
ap
p
r
o
ac
h
in
id
en
tif
y
in
g
p
h
i
s
h
in
g
UR
L
s
.
I
n
ad
d
itio
n
to
q
u
an
titativ
e
an
aly
s
is
,
we
s
u
p
p
lem
en
ted
o
u
r
f
in
d
in
g
s
with
v
is
u
al
r
ep
r
e
s
en
t
atio
n
s
,
as
d
ep
icted
Fig
u
r
e
1
a
n
d
Fig
u
r
e
2
.
T
h
ese
f
ig
u
r
es
elu
cid
ate
d
th
e
co
m
p
ar
ativ
e
p
er
f
o
r
m
an
ce
o
f
th
e
m
o
d
els
an
d
h
ig
h
lig
h
ted
th
e
ta
n
g
i
b
le
en
h
an
ce
m
en
ts
o
b
s
er
v
ed
in
b
o
th
ex
is
tin
g
an
d
p
r
o
p
o
s
ed
m
o
d
els.
No
tab
ly
,
o
u
r
p
r
o
p
o
s
ed
m
o
d
els
ex
h
ib
it
ed
a
s
ig
n
if
ica
n
t
im
p
r
o
v
em
e
n
t
,
ac
h
iev
in
g
1
0
0
%
ac
c
u
r
ac
y
a
n
d
af
f
ir
m
in
g
th
eir
s
u
itab
ilit
y
f
o
r
r
ea
l
-
wo
r
ld
d
e
p
lo
y
m
en
t.
Fig
u
r
e
1
p
r
esen
ts
th
e
ac
c
u
r
ac
y
o
f
f
o
u
r
m
o
d
els:
DT
,
RF
,
XGBo
o
s
t,
an
d
MLP
b
y
em
p
l
o
y
i
n
g
PC
A
f
o
r
f
ea
tu
r
e
s
elec
tio
n
.
E
ac
h
m
o
d
el
ac
h
iev
ed
a
p
er
f
ec
t
ac
cu
r
ac
y
o
f
1
0
0
%,
h
ig
h
lig
h
tin
g
th
e
ef
f
ec
tiv
en
ess
o
f
PC
A
in
im
p
r
o
v
in
g
m
o
d
el
p
er
f
o
r
m
an
c
e.
T
h
is
f
ig
u
r
e
d
em
o
n
s
tr
ates
th
e
r
o
b
u
s
tn
ess
o
f
o
u
r
f
ea
tu
r
e
s
e
lectio
n
m
eth
o
d
an
d
co
n
f
ir
m
s
th
e
h
ig
h
p
r
ed
ictiv
e
ca
p
ab
ilit
ies
o
f
th
e
m
o
d
els.
T
h
e
co
n
s
is
ten
tly
h
ig
h
ac
cu
r
ac
y
ac
r
o
s
s
all
m
o
d
els
unde
r
s
co
r
es th
e
i
m
p
o
r
ta
n
ce
o
f
PC
A
in
o
p
tim
izin
g
f
ea
tu
r
e
s
el
ec
tio
n
f
o
r
b
etter
ML
r
esu
lts
.
Fig
u
r
e
2
p
r
esen
ts
a
h
o
r
izo
n
ta
l
b
ar
g
r
ap
h
co
m
p
ar
in
g
th
e
ac
cu
r
ac
y
o
f
v
a
r
io
u
s
m
o
d
els
b
y
d
if
f
er
en
t
au
th
o
r
s
f
r
o
m
2
0
2
2
t
o
2
0
2
4
.
T
h
e
g
r
a
p
h
h
ig
h
lig
h
ts
th
at
th
e
p
r
o
p
o
s
ed
m
o
d
el
ac
h
i
ev
es
t
h
e
h
ig
h
est
ac
cu
r
ac
y
at
1
0
0
.
0
%,
o
u
tp
er
f
o
r
m
i
n
g
o
t
h
er
s
.
No
tab
ly
,
Gu
p
tta
et
a
l.
[
1
9
]
a
ch
iev
ed
9
9
.
1
7
%
an
d
Ujah
-
Og
b
u
ag
u
[
1
7
]
r
ea
ch
e
d
9
8
.
9
0
%.
T
h
is
C
h
ar
t
ef
f
ec
tiv
el
y
h
ig
h
li
g
h
ts
ad
v
a
n
ce
m
en
ts
in
m
o
d
el
p
e
r
f
o
r
m
an
ce
a
n
d
illu
s
t
r
ates
th
e
v
ar
iatio
n
s
in
ac
cu
r
ac
y
ac
r
o
s
s
r
ec
en
t stu
d
i
es.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2
5
0
2
-
4
7
52
E
n
h
a
n
ci
n
g
p
h
is
h
in
g
UR
L d
ete
ctio
n
th
r
o
u
g
h
co
mp
r
eh
en
s
ive
f
ea
tu
r
e
s
elec
tio
n
:
a
co
mp
a
r
a
tive
… (
P
r
ee
ti
)
1185
Fig
u
r
e
1
.
C
o
m
p
a
r
ativ
e
an
aly
s
i
s
o
f
m
o
d
el
p
er
f
o
r
m
an
ce
ac
r
o
s
s
f
o
u
r
d
atasets
Fig
u
r
e
2
.
Acc
u
r
ac
y
c
o
m
p
ar
is
o
n
s
b
etwe
en
ex
is
tin
g
an
d
p
r
o
p
o
s
ed
m
o
d
els f
o
r
p
h
is
h
in
g
UR
L
attac
k
d
etec
tio
n
C
r
u
c
i
a
ll
y
,
o
u
r
e
x
p
e
r
i
m
e
n
t
at
i
o
n
w
a
s
u
n
d
e
r
p
i
n
n
e
d
b
y
t
h
e
u
t
i
liz
a
t
i
o
n
o
f
f
i
v
e
m
u
lt
i
d
i
m
e
n
s
i
o
n
a
l
d
a
t
as
e
ts
,
e
n
s
u
r
i
n
g
t
h
e
m
o
d
e
l
s
’
e
x
p
o
s
u
r
e
t
o
d
i
v
e
r
s
e
a
n
d
r
e
p
r
e
s
e
n
t
at
i
v
e
d
a
t
a
s
a
m
p
l
es
.
T
h
is
s
t
r
a
t
e
g
i
c
a
p
p
r
o
a
c
h
b
o
l
s
t
e
r
e
d
t
h
e
m
o
d
e
l
s
’
g
e
n
e
r
a
li
z
a
ti
o
n
c
a
p
a
b
i
li
t
y
.
T
h
is
s
t
r
a
t
e
g
i
c
a
p
p
r
o
a
c
h
b
o
l
s
t
e
r
e
d
t
h
e
m
o
d
e
l
s
’
g
e
n
e
r
a
li
z
a
ti
o
n
c
a
p
a
b
i
l
it
y
a
n
d
e
n
g
e
n
d
e
r
e
d
c
o
n
f
i
d
e
n
c
e
i
n
t
h
e
i
r
r
e
a
l
-
w
o
r
l
d
a
p
p
li
c
a
b
i
li
t
y
.
I
n
ess
e
n
c
e
,
o
u
r
r
es
e
a
r
c
h
e
n
d
e
a
v
o
r
s
a
i
m
e
d
t
o
c
o
n
t
r
i
b
u
te
v
a
l
u
a
b
l
e
i
n
s
i
g
h
t
s
i
n
t
o
t
h
e
o
p
t
i
m
i
z
at
i
o
n
o
f
M
L
a
n
d
D
L
ap
p
r
o
a
c
h
e
s
f
o
r
d
e
t
e
c
t
i
n
g
p
h
i
s
h
i
n
g
U
R
L
a
tt
a
c
k
s
.
B
y
m
e
t
i
c
u
l
o
u
s
l
y
s
e
l
e
ct
m
o
d
e
ls
,
e
m
p
l
o
y
i
n
g
a
d
v
a
n
c
e
d
f
e
a
t
u
r
e
s
s
el
e
c
ti
o
n
t
e
c
h
n
i
q
u
e
s
,
a
n
d
co
n
d
u
c
t
i
n
g
r
i
g
o
r
o
u
s
e
x
p
e
r
i
m
e
n
t
a
t
i
o
n
,
we
s
t
r
i
v
e
d
t
o
p
u
s
h
t
h
e
b
o
u
n
d
a
r
i
e
s
o
f
a
c
c
u
r
a
cy
a
n
d
e
f
f
i
c
a
c
y
i
n
c
o
m
b
a
t
i
n
g
c
y
b
e
r
t
h
r
e
a
t
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
52
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
36
,
No
.
2
,
No
v
em
b
er
20
24
:
1
1
8
2
-
1
1
8
8
1186
4.
ADDR
E
SS
I
NG
L
I
M
I
T
A
T
I
O
NS
W
h
ile
o
u
r
r
esear
ch
d
em
o
n
s
t
r
ates
p
r
o
m
is
in
g
r
esu
lts
,
s
ev
e
r
al
lim
itatio
n
s
m
u
s
t
b
e
ac
k
n
o
wled
g
ed
.
Firstl
y
,
ac
h
iev
in
g
1
0
0
%
ac
cu
r
a
cy
ac
r
o
s
s
all
m
o
d
els
m
ay
in
d
icate
o
v
er
f
itti
n
g
,
p
ar
ticu
lar
l
y
with
th
e
d
ataset
u
s
ed
.
T
h
is
p
er
f
o
r
m
an
ce
m
i
g
h
t
n
o
t
g
en
e
r
alize
to
o
th
er
d
atasets
o
r
r
ea
l
-
wo
r
ld
s
ce
n
ar
io
s
.
Seco
n
d
ar
y
,
th
e
d
atasets
u
s
ed
m
a
y
n
o
t
f
u
lly
ca
p
tu
r
e
th
e
d
iv
e
r
s
ity
an
d
ev
o
lv
in
g
n
atu
r
e
o
f
p
h
i
s
h
in
g
UR
L
s
.
Mo
r
eo
v
e
r
,
o
u
r
s
tu
d
y
f
o
cu
s
ed
o
n
a
lim
ited
n
u
m
b
er
o
f
f
ea
tu
r
e
s
elec
tio
n
tech
n
iq
u
es;
ex
p
lo
r
in
g
a
d
d
itio
n
al
m
eth
o
d
s
co
u
l
d
y
ield
f
u
r
th
er
i
n
s
ig
h
ts
.
L
astl
y
,
th
e
co
m
p
u
tatio
n
al
e
f
f
icien
cy
an
d
s
ca
lab
ilit
y
o
f
th
e
m
o
d
els
wer
e
n
o
t
e
x
ten
s
iv
ely
ev
alu
ated
,
wh
ich
is
cr
u
cial
f
o
r
r
ea
l
-
tim
e
b
r
o
wser
in
te
g
r
atio
n
an
d
p
r
ac
tical
d
e
p
lo
y
m
en
t.
5.
I
M
P
L
I
CA
T
I
O
N
S F
O
R
F
UT
URE R
E
S
E
ARCH
B
u
ild
in
g
o
n
o
u
r
f
in
d
in
g
,
f
u
tu
r
e
r
esear
ch
s
h
o
u
ld
f
o
cu
s
o
n
v
alid
atin
g
th
e
m
o
d
els
with
m
o
r
e
d
iv
er
s
e
an
d
d
y
n
am
ic
d
atasets
to
en
s
u
r
e
r
o
b
u
s
tn
ess
an
d
g
e
n
er
aliz
ab
ilit
y
.
I
n
v
esti
g
atin
g
ad
d
itio
n
al
f
ea
tu
r
e
s
elec
tio
n
tech
n
iq
u
es
an
d
h
y
b
r
id
ap
p
r
o
ac
h
es
co
u
ld
en
h
an
ce
m
o
d
el
p
er
f
o
r
m
an
ce
an
d
ad
ap
tab
ilit
y
.
Fu
r
th
er
m
o
r
e,
ad
d
r
ess
in
g
co
m
p
u
tatio
n
al
ef
f
i
cien
cy
an
d
s
ca
lab
ilit
y
will b
e
ess
en
tial f
o
r
r
ea
l
-
tim
e
ap
p
licat
io
n
s
an
d
in
teg
r
atio
n
in
to
web
b
r
o
wser
s
.
E
x
p
lo
r
i
n
g
th
e
m
o
d
els’
ef
f
ec
ti
v
en
ess
ag
ain
s
t
em
er
g
in
g
p
h
is
h
in
g
tactic
will
also
b
e
cr
itical.
Fin
ally
,
in
ter
d
is
cip
lin
ar
y
c
o
llab
o
r
atio
n
with
cy
b
e
r
s
ec
u
r
ity
ex
p
er
ts
an
d
p
r
ac
titi
o
n
e
r
s
ca
n
p
r
o
v
id
e
p
r
ac
tical
in
s
ig
h
ts
,
en
s
u
r
in
g
th
e
m
o
d
els’
ef
f
icien
cy
an
d
r
eliab
ilit
y
in
r
e
al
-
wo
r
ld
en
v
ir
o
n
m
e
n
ts
.
6.
CO
NCLU
SI
O
N
I
n
th
is
r
esear
ch
p
ap
er
,
we
co
n
d
u
cted
a
th
o
r
o
u
g
h
an
aly
s
is
o
f
f
ea
tu
r
e
s
elec
tio
n
tech
n
iq
u
es
a
cr
o
s
s
f
iv
e
d
atasets
,
em
p
lo
y
in
g
m
eth
o
d
s
s
u
ch
as
R
F
s
elec
t
f
r
o
m
m
o
d
el
,
Select
KB
est
w
ith
ch
i
-
s
q
u
ar
e
s
tatis
tic,
R
FE
,
PC
A
.
Ou
r
ex
p
er
im
e
n
ts
,
p
ar
ticu
lar
ly
f
o
cu
s
in
g
o
n
PC
A
an
d
th
e
f
o
u
r
t
h
d
ataset,
d
em
o
n
s
tr
ated
th
at
all
f
o
u
r
m
o
d
els
(
DT
s
,
RF
s
,
XG
B
o
o
s
t,
an
d
MLP
)
ac
h
iev
ed
1
0
0
%
ac
cu
r
ac
y
in
d
etec
tin
g
p
h
is
h
in
g
UR
L
s
attac
k
s
.
T
h
is
ex
ce
p
tio
n
al
p
er
f
o
r
m
an
ce
s
u
g
g
ests
th
e
r
o
b
u
s
tn
ess
o
f
o
u
r
p
r
o
p
o
s
ed
m
o
d
els.
Desp
ite
s
o
m
e
lim
itatio
n
s
,
in
clu
d
in
g
p
o
te
n
tial
o
v
e
r
f
itti
n
g
an
d
th
e
n
ee
d
f
o
r
b
r
o
ad
er
d
atas
et
v
alid
atio
n
,
th
ese
f
in
d
in
g
s
in
d
icate
th
e
v
iab
ilit
y
an
d
r
eliab
ilit
y
o
f
o
u
r
ap
p
r
o
ac
h
f
o
r
r
ea
l
-
life
im
p
lem
en
tatio
n
i
n
ad
d
r
ess
in
g
p
h
is
h
in
g
th
r
ea
ts
.
AP
P
E
NDI
X
T
ab
le
1
.
C
o
m
p
r
eh
en
s
iv
e
a
n
aly
s
is
o
f
m
o
d
el
p
er
f
o
r
m
a
n
ce
m
etr
ics ac
r
o
s
s
d
atasets
an
d
ef
f
icac
y
o
f
p
r
o
p
o
s
ed
m
o
d
el
o
n
f
o
u
th
d
ataset
with
PC
A
f
ea
tu
r
e
s
elec
tio
n
(
1
0
0
%)
D
a
t
a
s
e
t
s
F
e
a
t
u
r
e
s
e
l
e
c
t
i
o
n
t
e
c
h
n
i
q
u
e
s
M
L
a
n
d
D
L
m
o
d
e
l
s
A
c
c
u
r
a
c
y
(
%)
1
st
d
a
t
a
s
e
t
s
c
o
n
t
a
i
n
3
9
6
5
6
i
n
s
t
a
n
c
e
s w
i
t
h
1
4
M
u
l
t
i
d
i
me
n
si
o
n
a
l
f
e
a
t
u
r
e
s
R
F
se
l
e
c
t
f
r
o
m
m
o
d
e
l
f
e
a
t
u
r
e
sel
e
c
t
i
o
n
DT
9
4
.
0
2
RF
9
5
.
6
2
X
G
B
o
o
st
9
4
.
3
5
M
LP
7
3
.
0
9
S
e
l
e
c
t
K
B
e
s
t
w
i
t
h
t
h
e
c
h
i
-
sq
u
a
r
e
(
χ
²)
st
a
t
i
st
i
c
DT
9
2
.
3
5
RF
9
3
.
0
2
X
G
B
o
o
st
9
2
.
6
M
LP
8
9
.
9
7
P
C
A
DT
9
3
.
7
1
RF
9
5
.
8
9
X
G
B
o
o
st
9
5
.
0
2
M
LP
4
6
.
2
5
R
F
E
DT
9
4
.
0
4
RF
9
5
.
6
2
X
G
B
o
o
st
9
4
.
4
M
LP
6
7
.
5
2
2
nd
d
a
t
a
s
e
t
s
c
o
n
t
a
i
n
1
1
4
3
0
i
n
st
a
n
c
e
s
w
i
t
h
89
M
u
l
t
i
d
i
m
e
n
s
i
o
n
a
l
f
e
a
t
u
r
e
s
R
F
se
l
e
c
t
f
r
o
m
m
o
d
e
l
f
e
a
t
u
r
e
sel
e
c
t
i
o
n
DT
9
3
.
0
3
RF
9
5
.
6
2
X
G
B
o
o
st
9
6
.
1
2
M
LP
7
7
.
1
9
S
e
l
e
c
t
K
B
e
s
t
w
i
t
h
t
h
e
c
h
i
-
sq
u
a
r
e
(
χ
²)
st
a
t
i
st
i
c
DT
9
2
.
0
0
RF
9
2
.
7
3
X
G
B
o
o
st
9
2
.
9
1
M
LP
9
1
.
8
P
C
A
DT
8
5
.
8
5
RF
9
0
.
9
5
X
G
B
o
o
st
9
0
.
9
3
M
LP
7
0
.
3
1
R
F
E
DT
9
3
.
4
9
RF
9
5
.
1
2
X
G
B
o
o
st
9
5
.
9
4
M
LP
7
6
.
7
8
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2
5
0
2
-
4
7
52
E
n
h
a
n
ci
n
g
p
h
is
h
in
g
UR
L d
ete
ctio
n
th
r
o
u
g
h
co
mp
r
eh
en
s
ive
f
ea
tu
r
e
s
elec
tio
n
:
a
co
mp
a
r
a
tive
… (
P
r
ee
ti
)
1187
T
ab
le
1
.
C
o
m
p
r
eh
en
s
iv
e
a
n
aly
s
is
o
f
m
o
d
el
p
er
f
o
r
m
a
n
ce
m
etr
ics ac
r
o
s
s
d
atasets
an
d
ef
f
icac
y
o
f
p
r
o
p
o
s
ed
m
o
d
el
o
n
f
o
u
th
d
ataset
with
PC
A
f
ea
tu
r
e
s
elec
tio
n
(
1
0
0
%)
(
C
o
n
tin
u
e
d
)
D
a
t
a
s
e
t
s
F
e
a
t
u
r
e
s
e
l
e
c
t
i
o
n
t
e
c
h
n
i
q
u
e
s
M
L
a
n
d
D
L
m
o
d
e
l
s
A
c
c
u
r
a
c
y
(
%)
3
rd
d
a
t
a
s
e
t
s
c
o
n
t
a
i
n
8
8
6
4
6
i
n
st
a
n
c
e
s w
i
t
h
1
1
1
m
u
l
t
i
d
i
me
n
si
o
n
a
l
f
e
a
t
u
r
e
s
R
F
se
l
e
c
t
f
r
o
m
m
o
d
e
l
f
e
a
t
u
r
e
sel
e
c
t
i
o
n
DT
9
5
.
1
9
RF
9
6
.
7
8
X
G
B
o
o
st
9
6
.
7
3
M
LP
9
5
.
7
6
S
e
l
e
c
t
K
B
e
s
t
w
i
t
h
t
h
e
c
h
i
-
sq
u
a
r
e
(
χ
²)
st
a
t
i
st
i
c
DT
8
6
.
5
8
RF
8
6
.
5
8
X
G
B
o
o
st
8
6
.
5
7
M
LP
8
6
.
5
7
P
C
A
DT
9
4
.
0
5
RF
9
6
.
0
2
X
G
B
o
o
st
9
5
.
7
8
M
LP
9
4
.
3
3
R
F
E
DT
9
4
.
7
5
RF
9
6
.
5
2
X
G
B
o
o
st
9
6
.
5
6
M
LP
8
6
.
9
8
4
th
d
a
t
a
set
s c
o
n
t
a
i
n
1
0
0
0
0
i
n
st
a
n
c
e
s
w
i
t
h
5
0
mu
l
t
i
d
i
m
e
n
s
i
o
n
a
l
f
e
a
t
u
r
e
s
R
F
se
l
e
c
t
f
r
o
m
m
o
d
e
l
f
e
a
t
u
r
e
sel
e
c
t
i
o
n
DT
1
0
0
RF
1
0
0
X
G
B
o
o
st
9
9
.
9
6
M
LP
9
5
.
5
6
S
e
l
e
c
t
K
B
e
s
t
w
i
t
h
t
h
e
c
h
i
-
sq
u
a
r
e
(
χ
²)
st
a
t
i
st
i
c
DT
1
0
0
RF
1
0
0
X
G
B
o
o
st
9
9
.
9
6
M
LP
9
9
.
8
6
P
C
A
DT
1
0
0
RF
1
0
0
X
G
B
o
o
st
1
0
0
M
LP
1
0
0
R
F
E
DT
1
0
0
RF
1
0
0
X
G
B
o
o
st
9
9
.
9
6
M
LP
9
4
.
1
5
th
d
a
t
a
set
s c
o
n
t
a
i
n
1
1
0
5
5
i
n
st
a
n
c
e
s
w
i
t
h
3
1
mu
l
t
i
d
i
m
e
n
s
i
o
n
a
l
f
e
a
t
u
r
e
s
R
F
se
l
e
c
t
f
r
o
m
m
o
d
e
l
f
e
a
t
u
r
e
sel
e
c
t
i
o
n
DT
9
3
.
4
8
RF
9
3
.
4
8
X
G
B
o
o
st
9
3
.
4
9
M
LP
9
3
.
5
7
S
e
l
e
c
t
K
B
e
s
t
w
i
t
h
t
h
e
c
h
i
-
sq
u
a
r
e
(
χ
²)
st
a
t
i
st
i
c
DT
9
4
.
1
5
RF
9
4
.
3
3
X
G
B
o
o
st
9
4
.
4
2
M
LP
9
3
.
9
7
P
C
A
DT
9
3
.
9
1
RF
9
5
.
4
1
X
G
B
o
o
st
9
5
.
6
2
M
LP
9
4
.
2
7
R
F
E
DT
9
5
.
1
4
RF
9
5
.
2
X
G
B
o
o
st
9
5
.
1
1
M
LP
9
4
.
9
3
RE
F
E
R
E
NC
E
S
[
1
]
V
.
B
h
a
v
s
a
r
,
A
.
K
a
d
l
a
k
,
a
n
d
S
.
S
h
a
r
ma,
“
S
t
u
d
y
o
n
p
h
i
s
h
i
n
g
a
t
t
a
c
k
s
,
”
I
n
t
e
rn
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
f
C
o
m
p
u
t
e
r
Ap
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
1
8
2
,
n
o
.
3
3
,
p
p
.
2
7
–
2
9
,
D
e
c
.
2
0
1
8
,
d
o
i
:
1
0
.
5
1
2
0
/
i
j
c
a
2
0
1
8
9
1
8
2
8
6
.
[
2
]
B
.
W
e
i
e
t
a
l
.
,
“
A
d
e
e
p
-
l
e
a
r
n
i
n
g
-
d
r
i
v
e
n
l
i
g
h
t
-
w
e
i
g
h
t
p
h
i
s
h
i
n
g
d
e
t
e
c
t
i
o
n
se
n
so
r
,
”
S
e
n
so
rs
(
S
w
i
t
z
e
rl
a
n
d
)
,
v
o
l
.
1
9
,
n
o
.
1
9
,
O
c
t
.
2
0
1
9
,
d
o
i
:
1
0
.
3
3
9
0
/
s
1
9
1
9
4
2
5
8
.
[
3
]
M
.
A
l
a
z
a
b
a
n
d
S
.
F
e
l
l
o
w
,
“
M
a
l
i
c
i
o
u
s
U
R
L
d
e
t
e
c
t
i
o
n
u
si
n
g
d
e
e
p
l
e
a
r
n
i
n
g
,
”
T
e
c
h
Rx
i
v
,
2
0
2
3
,
d
o
i
:
1
0
.
3
6
2
2
7
/
t
e
c
h
r
x
i
v
.
1
1
4
9
2
6
2
2
.
v
1
.
[
4
]
A
.
M
a
c
i
,
A
.
S
a
n
t
o
r
so
l
a
,
A
.
C
o
sc
i
a
,
a
n
d
A
.
I
a
n
n
a
c
o
n
e
,
“
U
n
b
a
l
a
n
c
e
d
w
e
b
p
h
i
s
h
i
n
g
c
l
a
ssi
f
i
c
a
t
i
o
n
t
h
r
o
u
g
h
d
e
e
p
r
e
i
n
f
o
r
c
e
m
e
n
t
l
e
a
r
n
i
n
g
,
”
C
o
m
p
u
t
e
rs
,
v
o
l
.
1
2
,
n
o
.
6
,
J
u
n
.
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
c
o
m
p
u
t
e
r
s
1
2
0
6
0
1
1
8
.
[
5
]
M
.
W
.
S
h
a
u
k
a
t
,
R
.
A
mi
n
,
M
.
M
.
A
.
M
u
s
l
a
m
,
A
.
H
.
A
l
s
h
e
h
r
i
,
a
n
d
J
.
X
i
e
,
“
A
h
y
b
r
i
d
a
p
p
r
o
a
c
h
f
o
r
a
l
l
u
r
i
n
g
a
d
s
p
h
i
s
h
i
n
g
a
t
t
a
c
k
d
e
t
e
c
t
i
o
n
u
si
n
g
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
,
”
S
e
n
so
rs
,
v
o
l
.
2
3
,
n
o
.
1
9
,
O
c
t
.
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
s
2
3
1
9
8
0
7
0
.
[
6
]
S
.
A
l
n
e
mar
i
a
n
d
M
.
A
l
s
h
a
m
mari
,
“
D
e
t
e
c
t
i
n
g
p
h
i
s
h
i
n
g
d
o
ma
i
n
s
u
s
i
n
g
mac
h
i
n
e
l
e
a
r
n
i
n
g
,
”
A
p
p
l
i
e
d
S
c
i
e
n
c
e
s
(
S
w
i
t
zerl
a
n
d
)
,
v
o
l
.
1
3
,
n
o
.
8
,
A
p
r
.
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
a
p
p
1
3
0
8
4
6
4
9
.
[
7
]
A
.
D
.
K
u
l
k
a
r
n
i
,
“
C
o
n
v
o
l
u
t
i
o
n
n
e
u
r
a
l
n
e
t
w
o
r
k
s
f
o
r
p
h
i
sh
i
n
g
d
e
t
e
c
t
i
o
n
c
o
n
v
o
l
u
t
i
o
n
n
e
u
r
a
l
n
e
t
w
o
r
k
s
f
o
r
p
h
i
sh
i
n
g
d
e
t
e
c
t
i
o
n
c
o
n
v
o
l
u
t
i
o
n
n
e
u
r
a
l
n
e
t
w
o
r
k
s f
o
r
p
h
i
s
h
i
n
g
d
e
t
e
c
t
i
o
n
,
”
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
:
/
/
h
d
l
.
h
a
n
d
l
e
.
n
e
t
/
1
0
9
5
0
/
4
2
2
4
w
w
w
.
i
j
a
c
sa
.
t
h
e
s
a
i
.
o
r
g
[
8
]
R
.
K
i
r
u
t
h
i
g
a
a
n
d
D
.
A
k
i
l
a
,
“
P
h
i
s
h
i
n
g
w
e
b
s
i
t
e
s
d
e
t
e
c
t
i
o
n
u
s
i
n
g
mac
h
i
n
e
l
e
a
r
n
i
n
g
,
”
I
n
t
e
rn
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
R
e
c
e
n
t
T
e
c
h
n
o
l
o
g
y
a
n
d
En
g
i
n
e
e
ri
n
g
,
v
o
l
.
8
,
n
o
.
2
S
p
e
c
i
a
l
I
ssu
e
1
1
,
p
p
.
1
1
1
–
1
1
4
,
S
e
p
.
2
0
1
9
,
d
o
i
:
1
0
.
3
5
9
4
0
/
i
j
r
t
e
.
B
1
0
1
8
.
0
9
8
2
S
1
1
1
9
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
52
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
36
,
No
.
2
,
No
v
em
b
er
20
24
:
1
1
8
2
-
1
1
8
8
1188
[
9
]
D
.
T.
M
o
sa
,
M
.
Y
.
S
h
a
ms,
A
.
A
.
A
b
o
h
a
n
y
,
E
.
S
.
M
.
El
-
K
e
n
a
w
y
,
a
n
d
M
.
T
h
a
b
e
t
,
“
M
a
c
h
i
n
e
l
e
a
r
n
i
n
g
t
e
c
h
n
i
q
u
e
s
f
o
r
d
e
t
e
c
t
i
n
g
p
h
i
s
h
i
n
g
U
R
L
a
t
t
a
c
k
s
,
”
C
o
m
p
u
t
e
rs,
Ma
t
e
r
i
a
l
s
a
n
d
C
o
n
t
i
n
u
a
,
v
o
l
.
7
5
,
n
o
.
1
,
p
p
.
1
2
7
1
–
1
2
9
0
,
2
0
2
3
,
d
o
i
:
1
0
.
3
2
6
0
4
/
c
mc.
2
0
2
3
.
0
3
6
4
2
2
.
[
1
0
]
B
.
B
.
G
u
p
t
a
,
K
.
Y
a
d
a
v
,
I
.
R
a
z
z
a
k
,
K
.
P
sa
n
n
i
s,
A
.
C
a
s
t
i
g
l
i
o
n
e
,
a
n
d
X
.
C
h
a
n
g
,
“
A
n
o
v
e
l
a
p
p
r
o
a
c
h
f
o
r
p
h
i
s
h
i
n
g
U
R
Ls
d
e
t
e
c
t
i
o
n
u
si
n
g
l
e
x
i
c
a
l
b
a
s
e
d
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
i
n
a
r
e
a
l
-
t
i
m
e
e
n
v
i
r
o
n
m
e
n
t
,
”
C
o
m
p
u
t
e
r
C
o
m
m
u
n
i
c
a
t
i
o
n
s
,
v
o
l
.
1
7
5
,
p
p
.
4
7
–
5
7
,
J
u
l
.
2
0
2
1
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
c
o
m
c
o
m.
2
0
2
1
.
0
4
.
0
2
3
.
[
1
1
]
T.
C
h
o
u
d
h
a
r
y
,
S
.
M
h
a
p
a
n
k
a
r
,
R
.
B
h
d
d
h
a
,
A
.
K
h
a
r
u
k
,
a
n
d
R
.
P
a
t
i
l
,
“
A
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
a
p
p
r
o
a
c
h
f
o
r
p
h
i
s
h
i
n
g
a
t
t
a
c
k
d
e
t
e
c
t
i
o
n
,
”
J
o
u
rn
a
l
o
f
Art
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
a
n
d
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
3
,
n
o
.
3
,
p
p
.
1
0
8
–
1
1
3
,
J
u
l
.
2
0
2
3
,
d
o
i
:
1
0
.
3
7
9
6
5
/
j
a
i
t
.
2
0
2
3
.
0
1
9
7
.
[
1
2
]
E.
A
.
A
l
d
a
k
h
e
e
l
,
M
.
Za
k
a
r
i
a
h
,
G
.
A
.
G
a
s
h
g
a
r
i
,
F
.
A
.
A
l
m
a
r
sh
a
d
,
a
n
d
A
.
I
.
A
.
A
l
z
a
h
r
a
n
i
,
“
A
d
e
e
p
l
e
a
r
n
i
n
g
-
b
a
se
d
i
n
n
o
v
a
t
i
v
e
t
e
c
h
n
i
q
u
e
f
o
r
p
h
i
s
h
i
n
g
d
e
t
e
c
t
i
o
n
i
n
mo
d
e
r
n
sec
u
r
i
t
y
w
i
t
h
u
n
i
f
o
r
m
r
e
s
o
u
r
c
e
l
o
c
a
t
o
r
s
,
”
S
e
n
so
rs
,
v
o
l
.
2
3
,
n
o
.
9
,
M
a
y
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
s
2
3
0
9
4
4
0
3
.
[
1
3
]
L.
M
.
A
b
d
u
l
r
a
h
m
a
n
,
S
.
H
.
A
h
me
d
,
Z
.
N
.
R
a
s
h
i
d
,
Y
.
S
.
J
g
h
e
f
,
T.
M
.
G
h
a
z
i
,
a
n
d
U
.
H
.
J
a
d
e
r
,
“
W
e
b
p
h
i
sh
i
n
g
d
e
t
e
c
t
i
o
n
u
s
i
n
g
w
e
b
c
r
a
w
l
i
n
g
,
c
l
o
u
d
i
n
f
r
a
s
t
r
u
c
t
u
r
e
a
n
d
d
e
e
p
l
e
a
r
n
i
n
g
f
r
a
mew
o
r
k
,
”
J
o
u
rn
a
l
o
f
A
p
p
l
i
e
d
S
c
i
e
n
c
e
a
n
d
T
e
c
h
n
o
l
o
g
y
T
r
e
n
d
s
,
v
o
l
.
4
,
n
o
.
0
1
,
p
p
.
5
4
–
7
1
,
M
a
r
.
2
0
2
3
,
d
o
i
:
1
0
.
3
8
0
9
4
/
j
a
st
t
4
0
1
1
4
4
.
[
1
4
]
Z.
A
l
sh
i
n
g
i
t
i
,
R
.
A
l
a
q
e
l
,
J
.
A
l
-
M
u
h
t
a
d
i
,
Q
.
E
.
U
.
H
a
q
,
K
.
S
a
l
e
e
m,
a
n
d
M
.
H
.
F
a
h
e
e
m,
“
A
d
e
e
p
l
e
a
r
n
i
n
g
-
b
a
se
d
p
h
i
s
h
i
n
g
d
e
t
e
c
t
i
o
n
sy
st
e
m
u
si
n
g
C
N
N
,
LST
M
,
a
n
d
LST
M
-
C
N
N
,
”
El
e
c
t
ro
n
i
c
s
(
S
w
i
t
zer
l
a
n
d
)
,
v
o
l
.
1
2
,
n
o
.
1
,
Ja
n
.
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
e
l
e
c
t
r
o
n
i
c
s1
2
0
1
0
2
3
2
.
[
1
5
]
M
.
C
a
n
h
a
m,
C
.
P
o
s
e
y
,
a
n
d
M
.
C
o
n
st
a
n
t
i
n
o
,
“
P
h
i
s
h
d
e
r
b
y
:
s
h
o
r
i
n
g
t
h
e
h
u
man
s
h
i
e
l
d
t
h
r
o
u
g
h
g
a
m
i
f
i
e
d
p
h
i
sh
i
n
g
a
t
t
a
c
k
s
,”
Fr
o
n
t
Ed
u
c
(
L
a
u
s
a
n
n
e
)
,
v
o
l
.
6
,
J
a
n
.
2
0
2
2
,
d
o
i
:
1
0
.
3
3
8
9
/
f
e
d
u
c
.
2
0
2
1
.
8
0
7
2
7
7
.
[
1
6
]
M
.
G
.
H
r
,
A
.
M
v
,
S
.
G
.
P
r
a
sa
d
,
a
n
d
S
.
V
i
n
a
y
,
“
D
e
v
e
l
o
p
m
e
n
t
o
f
a
n
t
i
-
p
h
i
s
h
i
n
g
b
r
o
w
s
e
r
b
a
se
d
o
n
r
a
n
d
o
m
f
o
r
e
st
a
n
d
r
u
l
e
o
f
e
x
t
r
a
c
t
i
o
n
f
r
a
mew
o
r
k
,
”
C
y
b
e
rs
e
c
u
r
i
t
y
,
v
o
l
.
3
,
n
o
.
1
,
D
e
c
.
2
0
2
0
,
d
o
i
:
1
0
.
1
1
8
6
/
s
4
2
4
0
0
-
0
2
0
-
0
0
0
5
9
-
1.
[
1
7
]
B
.
C
.
U
j
a
h
-
O
g
b
u
a
g
u
,
O
.
N
.
A
k
a
n
d
e
,
a
n
d
E.
O
g
b
u
j
u
,
“
A
h
y
b
r
i
d
d
e
e
p
l
e
a
r
n
i
n
g
t
e
c
h
n
i
q
u
e
f
o
r
sp
o
o
f
i
n
g
w
e
b
si
t
e
U
R
L
d
e
t
e
c
t
i
o
n
i
n
r
e
a
l
-
t
i
me
a
p
p
l
i
c
a
t
i
o
n
s,”
J
o
u
r
n
a
l
o
f
El
e
c
t
ri
c
a
l
S
y
st
e
m
s
a
n
d
I
n
f
o
r
m
a
t
i
o
n
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
1
1
,
n
o
.
1
,
Ja
n
.
2
0
2
4
,
d
o
i
:
1
0
.
1
1
8
6
/
s
4
3
0
6
7
-
023
-
0
0
1
2
8
-
8.
[
1
8
]
S
.
R
.
A
.
S
a
ma
d
e
t
a
l
.
,
“
A
n
a
l
y
si
s
o
f
t
h
e
p
e
r
f
o
r
m
a
n
c
e
i
mp
a
c
t
o
f
f
i
n
e
-
t
u
n
e
d
mac
h
i
n
e
l
e
a
r
n
i
n
g
mo
d
e
l
f
o
r
p
h
i
s
h
i
n
g
U
R
L
d
e
t
e
c
t
i
o
n
,”
El
e
c
t
r
o
n
i
c
s
(
S
w
i
t
zer
l
a
n
d
)
,
v
o
l
.
1
2
,
n
o
.
7
,
A
p
r
.
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
e
l
e
c
t
r
o
n
i
c
s1
2
0
7
1
6
4
2
.
[
1
9
]
S
.
D
.
G
u
p
t
t
a
,
K
.
T.
S
h
a
h
r
i
a
r
,
H
.
A
l
q
a
h
t
a
n
i
,
D
.
A
l
sa
l
ma
n
,
a
n
d
I
.
H
.
S
a
r
k
e
r
,
“
M
o
d
e
l
i
n
g
h
y
b
r
i
d
f
e
a
t
u
r
e
-
b
a
s
e
d
p
h
i
s
h
i
n
g
w
e
b
si
t
e
s
d
e
t
e
c
t
i
o
n
u
s
i
n
g
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
t
e
c
h
n
i
q
u
e
s,
”
An
n
a
l
s
ys
o
f
D
a
t
a
S
c
i
e
n
c
e
,
v
o
l
.
1
1
,
n
o
.
1
,
p
p
.
2
1
7
–
2
4
2
,
F
e
b
.
2
0
2
4
,
d
o
i
:
1
0
.
1
0
0
7
/
s
4
0
7
4
5
-
022
-
0
0
3
7
9
-
8.
[
2
0
]
M
.
A
l
mo
u
s
a
,
T
.
Z
h
a
n
g
,
A
.
S
a
r
r
a
f
z
a
d
e
h
,
a
n
d
M
.
A
n
w
a
r
,
“
P
h
i
s
h
i
n
g
w
e
b
s
i
t
e
d
e
t
e
c
t
i
o
n
:
H
o
w
e
f
f
e
c
t
i
v
e
a
r
e
d
e
e
p
l
e
a
r
n
i
n
g
‐
b
a
s
e
d
mo
d
e
l
s
a
n
d
h
y
p
e
r
p
a
r
a
m
e
t
e
r
o
p
t
i
m
i
z
a
t
i
o
n
?
,”
S
e
c
u
r
i
t
y
a
n
d
Pri
v
a
c
y
,
v
o
l
.
5
,
n
o
.
6
,
N
o
v
.
2
0
2
2
,
d
o
i
:
1
0
.
1
0
0
2
/
sp
y
2
.
2
5
6
.
[
2
1
]
A
.
A
l
mo
m
a
n
i
e
t
a
l
.
,
“
P
h
i
s
h
i
n
g
w
e
b
s
i
t
e
d
e
t
e
c
t
i
o
n
w
i
t
h
s
e
ma
n
t
i
c
f
e
a
t
u
r
e
s
b
a
s
e
d
o
n
mac
h
i
n
e
l
e
a
r
n
i
n
g
c
l
a
ss
i
f
i
e
r
s:
a
c
o
m
p
a
r
a
t
i
v
e
st
u
d
y
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
n
S
e
m
a
n
t
i
c
We
b
a
n
d
I
n
f
o
rm
a
t
i
o
n
S
y
st
e
m
s
(
I
J
S
WIS
)
,
v
o
l
.
1
8
,
n
o
.
1
,
Ja
n
.
2
0
2
2
,
d
o
i
:
1
0
.
4
0
1
8
/
I
JS
W
I
S
.
2
9
7
0
3
2
.
[
2
2
]
M
.
El
sa
d
i
g
e
t
a
l
.
,
“
I
n
t
e
l
l
i
g
e
n
t
d
e
e
p
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
c
y
b
e
r
p
h
i
s
h
i
n
g
U
R
L
d
e
t
e
c
t
i
o
n
b
a
se
d
o
n
B
E
R
T
f
e
a
t
u
r
e
s
e
x
t
r
a
c
t
i
o
n
,
”
El
e
c
t
r
o
n
i
c
s
(
S
w
i
t
zer
l
a
n
d
)
,
v
o
l
.
1
1
,
n
o
.
2
2
,
N
o
v
.
2
0
2
2
,
d
o
i
:
1
0
.
3
3
9
0
/
e
l
e
c
t
r
o
n
i
c
s1
1
2
2
3
6
4
7
.
[
2
3
]
G
.
M
o
h
a
me
d
,
J.
V
i
s
u
m
a
t
h
i
,
M
.
M
a
h
d
a
l
,
J.
A
n
a
n
d
,
a
n
d
M
.
El
a
n
g
o
v
a
n
,
“
A
n
e
f
f
e
c
t
i
v
e
a
n
d
se
c
u
r
e
m
e
c
h
a
n
i
s
m
f
o
r
p
h
i
sh
i
n
g
a
t
t
a
c
k
s
u
si
n
g
a
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
a
p
p
r
o
a
c
h
,
”
Pro
c
e
sses
,
v
o
l
.
1
0
,
n
o
.
7
,
J
u
l
.
2
0
2
2
,
d
o
i
:
1
0
.
3
3
9
0
/
p
r
1
0
0
7
1
3
5
6
.
[
2
4
]
S
.
J.
B
u
a
n
d
H
.
J.
K
i
m,
“
O
p
t
i
mi
z
e
d
U
R
L
f
e
a
t
u
r
e
se
l
e
c
t
i
o
n
b
a
s
e
d
o
n
g
e
n
e
t
i
c
-
a
l
g
o
r
i
t
h
m
-
e
m
b
e
d
d
e
d
d
e
e
p
l
e
a
r
n
i
n
g
f
o
r
p
h
i
s
h
i
n
g
w
e
b
si
t
e
d
e
t
e
c
t
i
o
n
,
”
El
e
c
t
r
o
n
i
c
s (
S
w
i
t
z
e
rl
a
n
d
)
,
v
o
l
.
1
1
,
n
o
.
7
,
A
p
r
.
2
0
2
2
,
d
o
i
:
1
0
.
3
3
9
0
/
e
l
e
c
t
r
o
n
i
c
s
1
1
0
7
1
0
9
0
.
[
2
5
]
A
.
G
ó
m
e
z
a
n
d
A
.
M
u
ñ
o
z
,
“
D
e
e
p
l
e
a
r
n
i
n
g
-
b
a
s
e
d
a
t
t
a
c
k
d
e
t
e
c
t
i
o
n
a
n
d
c
l
a
ss
i
f
i
c
a
t
i
o
n
i
n
a
n
d
r
o
i
d
d
e
v
i
c
e
s
,
”
E
l
e
c
t
r
o
n
i
c
s
(
S
w
i
t
z
e
rl
a
n
d
)
,
v
o
l
.
1
2
,
n
o
.
1
5
,
A
u
g
.
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
e
l
e
c
t
r
o
n
i
c
s
1
2
1
5
3
2
5
3
.
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
Ms.
Pre
e
ti
M
.
Tec
h
.
,
P
h
.
D
.
P
u
rsu
in
g
fr
o
m
De
p
a
rtme
n
t
o
f
Co
m
p
u
ter
S
c
ien
c
e
a
n
d
Ap
p
l
ica
ti
o
n
s,
M
.
D.U
n
iv
e
rsit
y
,
Ro
h
tak
.
S
h
e
h
a
s p
u
b
li
sh
e
d
m
o
re
th
a
n
1
6
p
u
b
li
c
a
ti
o
n
s
in
v
a
rio
u
s
jo
u
r
n
a
ls
/ma
g
a
z
in
e
s
o
f
n
a
ti
o
n
a
l
a
n
d
in
ter
n
a
ti
o
n
a
l
re
p
u
te.
He
r
a
re
a
o
f
re
se
a
r
c
h
in
c
lu
d
e
s
m
a
c
h
in
e
lea
rn
in
g
,
d
e
e
p
lea
rn
in
g
,
a
n
d
c
y
b
e
r
se
c
u
rit
y
.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
m
isk
h
o
k
h
a
r1
2
1
@
g
m
a
il
.
c
o
m
.
Dr
.
Priti
S
h
a
r
m
a
M
CA,
P
h
.
D
.
(Co
m
p
u
te
S
c
ien
c
e
)
is
w
o
rk
i
n
g
a
s
a
n
As
sista
n
t
P
ro
fe
ss
o
r
in
th
e
De
p
a
rtme
n
t
o
f
C
o
m
p
u
ter S
c
ie
n
c
e
a
n
d
Ap
p
li
c
a
ti
o
n
s,
M
.
D.
Un
i
v
e
rsity
,
Ro
h
tak
.
S
h
e
h
a
s
p
u
b
li
sh
e
d
m
o
re
th
a
n
6
0
p
u
b
li
c
a
ti
o
n
s
i
n
v
a
rio
u
s
jo
u
rn
a
ls
/ma
g
a
z
in
e
s
o
f
n
a
ti
o
n
a
l
a
n
d
in
tern
a
ti
o
n
a
l
re
p
u
te.
S
h
e
is
e
n
g
a
g
e
d
i
n
tea
c
h
i
n
g
a
n
d
re
se
a
rc
h
fro
m
th
e
las
t
1
5
y
e
a
rs.
He
r
a
re
a
o
f
re
se
a
rc
h
i
n
c
lu
d
e
s
d
a
ta
m
in
in
g
,
b
i
g
d
a
ta,
s
o
ftwa
re
e
n
g
in
e
e
rin
g
,
m
a
c
h
in
e
lea
rn
in
g
,
a
n
d
d
e
e
p
lea
rn
in
g
.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
P
rit
i@m
d
u
r
o
h
ta
k
.
a
c
.
in
.
Evaluation Warning : The document was created with Spire.PDF for Python.