I
nte
rna
t
io
na
l J
o
urna
l o
f
E
lect
rica
l a
nd
Co
m
pu
t
er
E
ng
ineering
(
I
J
E
CE
)
Vo
l.
15
,
No
.
4
,
A
u
g
u
s
t
20
25
,
p
p
.
3
9
3
8
~
3
9
4
8
I
SS
N:
2088
-
8
7
0
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijece.
v
15
i
4
.
pp
3
9
3
8
-
3
9
4
8
3938
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ec
e.
ia
esco
r
e.
co
m
M
ultilevel a
nd m
ultiso
urce da
ta fu
sio
n app
ro
a
ch f
o
r
net
wo
rk
intrusio
n de
tec
tion sy
stem using
m
a
chine learning
te
chniques
H
a
rsh
it
ha
So
m
a
s
hek
a
r,
P
ra
m
o
d H
a
lebid
u B
a
s
a
v
a
ra
j
u
D
e
p
a
r
t
me
n
t
o
f
I
n
f
o
r
m
a
t
i
o
n
S
c
i
e
n
c
e
a
n
d
E
n
g
i
n
e
e
r
i
n
g
,
A
d
i
c
h
u
n
c
h
a
n
a
g
i
r
i
I
n
s
t
i
t
u
t
e
o
f
T
e
c
h
n
o
l
o
g
y
a
f
f
i
l
i
a
t
e
d
t
o
V
i
sv
e
sv
a
r
a
y
a
T
e
c
h
n
o
l
o
g
i
c
a
l
U
n
i
v
e
r
si
t
y
,
B
e
l
a
g
a
v
i
,
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
Oct
7
,
2
0
2
4
R
ev
is
ed
Ma
r
2
6
,
2
0
2
5
Acc
ep
ted
Ma
y
2
4
,
2
0
2
5
To
e
n
h
a
n
c
e
th
e
p
e
rfo
rm
a
n
c
e
o
f
n
e
two
rk
i
n
tru
si
o
n
d
e
tec
ti
o
n
s
y
ste
m
s
(NID
S
)
,
th
is
p
a
p
e
r
p
r
o
p
o
se
s
a
n
o
v
e
l
m
u
lt
il
e
v
e
l
a
n
d
m
u
lt
iso
u
rc
e
d
a
ta
fu
sio
n
a
p
p
ro
a
c
h
,
a
p
p
li
e
d
to
NSL
-
KD
D
a
n
d
UN
S
W
-
NB1
5
d
a
tas
e
ts.
T
h
e
p
r
o
p
o
se
d
a
p
p
r
o
a
c
h
in
c
lu
d
e
s
th
re
e
v
a
rio
u
s
lev
e
ls
o
f
o
p
e
ra
ti
o
n
s,
wh
ich
a
re
fe
a
tu
re
lev
e
l
f
u
sio
n
,
d
ime
n
sio
n
a
l
it
y
re
d
u
c
ti
o
n
,
a
n
d
p
re
d
ictio
n
lev
e
l
f
u
sio
n
.
In
th
e
f
irst
sta
g
e
fe
a
tu
re
s o
f
NSL
-
KD
D an
d
UN
S
W
-
NB1
5
b
o
t
h
d
a
tas
e
ts are
fu
se
d
b
y
a
p
p
ly
i
n
g
th
e
in
n
e
r
jo
i
n
jo
i
n
t
o
p
e
ra
ti
o
n
b
y
se
lec
ti
n
g
c
o
m
m
o
n
fe
a
tu
re
s
li
k
e
p
ro
t
o
c
o
l,
se
rv
ice
a
n
d
lab
e
l.
On
c
e
th
e
d
a
ta
se
ts
a
r
e
fu
se
d
in
th
e
first
lev
e
l,
l
in
e
a
r
d
isc
rimin
a
n
t
a
n
a
ly
sis
is
a
p
p
li
e
d
f
o
r
1
2
fe
a
tu
re
c
o
l
u
m
n
s
wh
ich
is
re
d
u
c
e
d
to
a
sin
g
le
fe
a
tu
re
c
o
l
u
m
n
lea
d
i
n
g
t
o
d
ime
n
si
o
n
a
li
t
y
re
d
u
c
ti
o
n
a
t
t
h
e
se
c
o
n
d
lev
e
l.
F
i
n
a
ll
y
,
in
th
e
t
h
ird
lev
e
l
,
th
e
p
re
d
ictio
n
le
v
e
l
fu
si
o
n
tec
h
n
i
q
u
e
is
a
p
p
li
e
d
t
o
two
n
e
u
ra
l
n
e
two
r
k
m
o
d
e
ls,
wh
e
re
o
n
e
n
e
u
ra
l
n
e
two
r
k
m
o
d
e
l
h
a
s
a
sin
g
le
in
p
u
t
n
o
d
e
,
tw
o
h
i
d
d
e
n
n
o
d
e
s,
a
n
d
tw
o
o
u
tp
u
t
n
o
d
e
s,
a
n
d
a
n
o
t
h
e
r
m
o
d
e
l
h
a
v
i
n
g
a
sin
g
le
i
n
p
u
t
n
o
d
e
,
th
re
e
h
id
d
e
n
n
o
d
e
s,
a
n
d
tw
o
o
u
tp
u
t
n
o
d
e
s.
Th
e
o
u
tp
u
ts
o
b
tain
e
d
fro
m
t
h
e
se
two
m
o
d
e
ls
a
re
th
e
n
fu
se
d
u
sin
g
a
p
re
d
ictio
n
fu
si
o
n
tec
h
n
i
q
u
e
.
T
h
e
p
ro
p
o
se
d
a
p
p
ro
a
c
h
a
c
h
iev
e
s
a
c
las
sifica
ti
o
n
a
c
c
u
ra
c
y
o
f
9
7
.
5
%
.
K
ey
w
o
r
d
s
:
Data
f
u
s
io
n
Dim
en
s
io
n
ality
r
ed
u
ctio
n
Netwo
r
k
in
tr
u
s
io
n
d
etec
tio
n
Neu
r
al
n
etwo
r
k
Pre
d
ictio
n
lev
el
f
u
s
i
on
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
:
Har
s
h
ith
a
So
m
ash
ek
ar
Dep
ar
tm
en
t o
f
I
n
f
o
r
m
atio
n
Scien
ce
an
d
E
n
g
in
ee
r
i
n
g
,
Ad
ic
h
u
n
ch
an
a
g
ir
i I
n
s
titu
te
o
f
T
ec
h
n
o
lo
g
y
,
af
f
iliated
to
Vis
v
esv
ar
ay
a
T
ec
h
n
o
lo
g
ica
l U
n
iv
er
s
ity
B
elag
av
i 5
9
0
0
1
8
,
Kar
n
atak
a,
I
n
d
ia
E
m
ail: sh
@
m
ce
h
ass
an
.
ac
.
in
1.
I
NT
RO
D
UCT
I
O
N
Netwo
r
k
in
tr
u
s
io
n
d
etec
tio
n
s
y
s
tem
s
(
NI
D
S)
h
av
e
b
ec
o
m
e
o
n
e
o
f
th
e
p
r
im
e
m
ec
h
an
is
m
s
f
o
r
s
h
ield
in
g
o
r
p
r
o
tectin
g
co
m
p
u
ter
n
etwo
r
k
s
f
r
o
m
v
a
r
io
u
s
ty
p
es
o
f
m
alicio
u
s
attac
k
s
[
1
]
.
B
a
s
ically
,
it
m
o
n
ito
r
s
n
etwo
r
k
tr
af
f
ic
f
o
r
an
y
s
u
s
p
ici
o
u
s
ac
tiv
ities
f
r
o
m
th
e
h
ac
k
er
s
.
I
n
ca
s
e
o
f
an
y
r
em
a
r
k
ab
le
o
r
u
n
r
elate
d
ev
e
n
t,
it
in
f
o
r
m
s
o
r
g
iv
es
th
e
s
ig
n
al
t
o
th
e
ad
m
in
is
tr
ato
r
s
.
T
r
ad
itio
n
ally
,
in
tr
u
s
io
n
d
etec
tio
n
s
y
s
t
em
s
h
av
e
r
elied
o
n
s
in
g
le
-
s
o
u
r
ce
d
ata,
an
d
lik
e
n
etwo
r
k
tr
af
f
ic
lo
g
s
o
r
s
y
s
tem
ev
en
t
lo
g
s
.
Alth
o
u
g
h
it
was
v
er
y
ef
f
ec
tiv
e
in
ce
r
tain
asp
ec
ts
,
th
ese
m
eth
o
d
s
u
s
u
ally
f
ail
to
d
etec
t
h
ig
h
ly
s
o
p
h
is
ticated
an
d
co
n
tin
u
o
u
s
ly
ev
o
lv
in
g
th
r
ea
ts
du
e
to
co
m
p
lex
n
etwo
r
k
s
tr
u
ct
u
r
es,
esp
ec
ially
th
e
o
n
es th
at
i
n
clu
d
e
s
ev
er
al
s
tag
es o
r
d
if
f
er
en
t so
u
r
ce
s
[
2
]
.
Fo
r
ex
am
p
le,
a
n
attac
k
m
a
y
co
n
s
is
t
o
f
s
o
m
e
u
n
u
s
u
al
tr
af
f
ic
p
att
er
n
s
,
u
n
a
u
th
o
r
ize
d
ac
ce
s
s
attem
p
ts
,
o
r
a
n
o
m
alo
u
s
s
y
s
tem
b
eh
av
io
u
r
.
T
h
is
wo
u
ld
b
e
p
r
etty
h
ar
d
to
d
etec
t in
a
s
in
g
le
lo
g
s
o
u
r
ce
o
f
d
ata.
T
h
ese
lim
itat
io
n
s
h
o
wev
er
co
n
tr
o
lled
th
r
o
u
g
h
d
ata
f
u
s
io
n
t
ec
h
n
iq
u
es
b
y
th
e
r
esear
c
h
er
s
wh
ich
ar
e
d
is
cu
s
s
ed
in
th
e
liter
atu
r
e.
Da
ta
f
u
s
io
n
r
ef
er
s
to
a
tech
n
iq
u
e
f
o
r
m
er
g
in
g
in
f
o
r
m
atio
n
f
r
o
m
s
ev
er
al
s
o
u
r
ce
s
in
o
r
d
er
to
im
p
r
o
v
e
th
e
ac
c
u
r
ac
y
an
d
r
o
b
u
s
tn
ess
o
f
in
tr
u
s
io
n
d
et
ec
tio
n
.
I
t
b
ec
o
m
es
p
o
s
s
ib
le
to
cr
ea
te
a
m
o
r
e
co
m
p
lete
p
ictu
r
e
o
f
n
etwo
r
k
a
ctiv
ities
b
y
in
teg
r
atin
g
d
ata
f
r
o
m
v
ar
io
u
s
lo
g
s
an
d
s
en
s
o
r
s
lik
e
n
etwo
r
k
tr
af
f
ic,
s
y
s
tem
ev
en
ts
,
an
d
ap
p
licatio
n
lo
g
s
.
Su
ch
a
h
o
lis
tic
ap
p
r
o
ac
h
ca
n
also
b
o
o
s
t
an
im
p
r
o
v
ed
id
en
tific
at
io
n
o
f
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
Mu
ltil
ev
el
a
n
d
mu
ltis
o
u
r
ce
d
a
ta
fu
s
io
n
a
p
p
r
o
a
c
h
fo
r
n
etw
o
r
k
in
tr
u
s
io
n
…
(
Ha
r
s
h
ith
a
S
o
ma
s
h
ek
a
r
)
3939
k
n
o
wn
a
n
d
u
n
k
n
o
wn
th
r
ea
ts
in
th
e
n
etwo
r
k
b
y
e
n
ab
lin
g
th
e
co
r
r
elatio
n
o
f
d
is
p
ar
ate
d
a
ta
o
f
ev
id
e
n
ce
th
at
m
ig
h
t
o
t
h
er
wis
e
g
o
u
n
n
o
tice
d
.
NI
DSs
ar
e
o
n
e
o
f
th
e
k
e
y
asp
ec
ts
o
f
a
co
m
p
u
ter
n
et
wo
r
k
f
o
r
p
r
o
v
id
in
g
s
ec
u
r
ity
an
d
in
teg
r
ity
,
s
o
NI
DSs
s
h
o
u
ld
b
e
v
er
y
ac
cu
r
ate
as
ad
v
an
ce
d
d
etec
tio
n
m
ec
h
a
n
is
m
s
ag
ain
s
t
ev
er
-
in
cr
ea
s
in
g
cy
b
er
th
r
ea
ts
.
R
esear
ch
er
s
wo
r
k
ed
o
n
d
if
f
er
e
n
t
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
s
f
o
r
b
u
ild
in
g
ef
f
ec
tiv
e
in
tr
u
s
io
n
d
etec
tio
n
class
if
ier
s
,
wh
ich
a
r
e
n
e
u
r
al
n
etwo
r
k
s
,
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
es
(
SVMs)
,
an
d
k
-
n
ea
r
est
n
eig
h
b
o
r
s
(
KNN)
all
h
av
in
g
d
i
s
tin
ct
ad
v
an
tag
es a
n
d
c
h
allen
g
es.
I
n
m
an
y
ty
p
ical
in
s
tan
ce
s
n
eu
r
al
n
etwo
r
k
will b
e
ab
le
to
lear
n
v
er
y
co
m
p
lex
p
atter
n
s
,
h
en
c
e
it is
ab
le
to
f
in
d
an
o
m
alies
wh
ich
co
u
ld
b
e
an
in
tr
u
s
io
n
.
Sin
ce
n
e
u
r
a
l
n
etwo
r
k
m
o
d
els
m
ak
e
u
s
e
o
f
a
lar
g
e
am
o
u
n
t
o
f
tr
ain
in
g
d
ata
a
n
d
h
ig
h
co
m
p
u
tatio
n
al
r
eso
u
r
ce
s
t
o
an
aly
s
es
th
e
s
eq
u
e
n
ce
,
it
will
b
e
a
d
r
awb
ac
k
f
o
r
th
e
d
etec
tio
n
o
f
in
tr
u
s
io
n
s
.
An
o
t
h
er
p
o
p
u
lar
m
eth
o
d
in
f
in
d
in
g
in
tr
u
s
io
n
d
etec
tio
n
is
SVM,
it
is
m
o
r
e
p
o
p
u
lar
b
ec
au
s
e
th
ey
ar
e
ca
p
a
b
le
o
f
h
an
d
lin
g
h
ig
h
-
d
im
e
n
s
io
n
al
d
a
ta
an
d
p
er
f
o
r
m
well
ev
e
n
wit
h
a
s
m
all
t
r
ain
in
g
s
am
p
le.
I
n
s
o
m
e
ca
s
es,
SVMs
m
ay
b
e
m
o
r
e
s
en
s
itiv
e
wh
ile
ch
o
o
s
in
g
th
e
h
y
p
er
p
a
r
am
eter
s
,
wh
ich
h
a
v
e
to
b
e
ca
r
ef
u
lly
ad
ju
s
ted
,
else
m
o
d
el
will
p
r
ed
ict
u
s
ef
u
l
d
ata
as
an
in
tr
u
s
io
n
.
T
h
e
KNN
is
a
v
er
y
s
im
p
le
alg
o
r
ith
m
th
at
will
wo
r
k
in
s
o
m
e
ca
s
es
b
u
t
th
ey
o
f
ten
lead
to
a
h
i
g
h
f
alse
-
p
o
s
itiv
e
r
ate,
m
o
s
tly
wh
e
n
ev
er
t
h
e
d
ata
h
av
e
m
o
r
e
n
o
is
e.
T
h
e
m
ajo
r
d
r
aw
b
ac
k
o
f
u
s
in
g
tr
ad
itio
n
al
m
ac
h
i
n
e
lear
n
i
n
g
m
eth
o
d
s
ar
e
th
ese
m
o
d
els
alwa
y
s
lead
s
to
a
h
ig
h
n
u
m
b
er
o
f
f
alse
p
o
s
itiv
e
r
ates
an
d
f
alse
alar
m
s
.
I
n
o
r
d
er
to
o
v
er
c
o
m
e
th
ese
is
s
u
es,
r
esear
ch
er
s
ar
e
f
o
cu
s
in
g
to
cr
ea
te
b
etter
u
p
d
at
ed
m
eth
o
d
s
wh
ich
ca
n
im
p
r
o
v
e
d
etec
tio
n
r
ate.
I
n
o
r
d
e
r
to
cr
ea
te
b
etter
m
o
d
els
r
esear
ch
er
s
ar
e
f
o
cu
s
in
g
o
n
m
u
lti
-
s
o
u
r
ce
d
ata
f
u
s
io
n
a
p
p
r
o
a
ch
es
f
o
r
n
etwo
r
k
in
tr
u
s
io
n
d
et
ec
tio
n
.
B
y
s
tu
d
y
i
n
g
th
e
r
ec
en
t
liter
atu
r
e
in
m
u
lti
-
s
o
u
r
ce
d
ata,
it
is
co
n
clu
d
e
d
th
a
t
th
e
k
ey
co
n
ce
p
ts
,
tech
n
i
q
u
es
an
d
ap
p
licatio
n
s
o
f
th
ese
n
ewly
d
ev
elo
p
ed
ap
p
r
o
ac
h
es
s
h
o
wed
b
etter
r
esu
lts
wh
en
co
m
p
ar
e
d
to
tr
a
d
itio
n
a
l
m
ac
h
in
e
lear
n
in
g
(
ML
)
m
o
d
els.
I
n
o
r
d
er
to
b
u
ild
a
s
tr
o
n
g
in
tr
u
s
io
n
d
etec
tio
n
s
y
s
tem
,
th
e
i
n
tr
u
s
io
n
d
etec
tio
n
s
y
s
tem
s
(
I
DS)
s
h
o
u
ld
b
e
tr
ain
ed
o
n
v
ar
iety
o
f
n
etwo
r
k
in
tr
u
s
io
n
s
.
T
h
ese
in
tr
u
s
io
n
s
h
av
e
lim
ited
s
co
p
e
in
in
d
iv
id
u
al
ty
p
e
o
f
d
ata
s
ets.
Fo
r
ex
am
p
le,
t
h
e
NSL
-
KDD
d
ataset
wh
ich
we
co
n
s
id
er
ed
in
th
is
wo
r
k
f
o
cu
s
es
o
n
tr
ad
itio
n
al
n
etwo
r
k
attac
k
s
lik
e
Do
S,
Pro
b
e,
R
2
L
,
a
n
d
U2
R
,
wh
ile
an
o
th
er
o
n
e
UNSW
-
NB
1
5
d
ataset
h
as
co
n
s
id
er
e
d
m
o
d
er
n
th
r
ea
ts
s
u
ch
as
b
o
tn
ets,
wo
r
m
s
,
an
d
a
d
v
a
n
ce
d
p
er
s
is
ten
t
th
r
ea
ts
(
APTs)
.
B
y
in
teg
r
atin
g
b
o
th
NSDL
-
KDD
an
d
UNSW
-
NB
1
5
d
ata
s
ets
a
n
ew
f
u
s
ed
d
ata
s
ets
ca
n
b
e
g
en
er
ated
f
o
r
t
r
ain
in
g
,
b
y
th
is
we
ca
n
tr
ain
o
u
r
m
o
d
el
with
th
e
v
ar
iety
o
f
i
n
tr
u
s
io
n
s
ce
n
ar
i
o
s
,
wh
ich
m
ak
es
th
e
s
y
s
tem
m
o
r
e
r
o
b
u
s
t
an
d
ab
ilit
y
to
d
etec
t
v
ar
io
u
s
in
tr
u
s
io
n
s
ef
f
ec
tiv
ely
.
Ho
wev
er
,
th
e
f
u
s
i
o
n
p
r
o
ce
s
s
r
esu
lts
in
a
lar
g
e
f
ea
tu
r
e
s
et
o
f
d
ata
wh
ich
co
n
ta
in
s
b
o
th
NSL
-
KDD
an
d
UNSW
-
NB
1
5
f
ea
tu
r
es,
wh
ich
in
cr
ea
s
es
co
m
p
u
tatio
n
al
co
m
p
lex
ity
.
R
ed
u
cin
g
th
e
d
im
en
s
io
n
ality
o
f
f
ea
tu
r
es
p
lay
s
a
v
ital
r
o
le,
w
h
ich
ca
n
also
co
n
tr
ib
u
te
to
o
p
tim
ize
m
o
d
el
p
er
f
o
r
m
a
n
ce
.
Fu
r
th
er
,
h
a
n
d
lin
g
a
v
ar
iety
o
f
in
p
u
t
d
ata
s
ets
r
em
a
in
s
a
s
ig
n
if
ican
t
ch
allen
g
e
,
wh
ich
ca
n
b
e
h
an
d
led
b
y
ad
o
p
tin
g
p
r
e
d
ictio
n
-
lev
el
-
f
u
s
io
n
th
r
o
u
g
h
d
iv
er
s
e
n
eu
r
a
l
n
etwo
r
k
ar
c
h
itectu
r
es.
T
h
e
s
y
s
tem
'
s
r
eliab
ilit
y
ca
n
b
e
i
m
p
r
o
v
e
d
a
n
d
also
en
s
u
r
es m
o
r
e
ac
c
u
r
ate
d
etec
ti
o
n
o
f
n
ew
in
tr
u
s
io
n
s
.
Th
e
p
r
o
p
o
s
ed
m
o
d
el
ac
h
iev
e
d
9
7
.
5
%
ac
cu
r
ac
y
wh
ich
o
u
t
p
er
f
o
r
m
ed
s
ev
er
al
ex
is
tin
g
m
eth
o
d
s
.
All
p
r
ev
io
u
s
ML
-
b
ased
ap
p
r
o
ac
h
e
s
ar
e
ap
p
lied
to
s
in
g
le
ty
p
es
o
f
d
ata
s
ets.
I
n
th
is
p
a
p
er
,
we
p
r
o
p
o
s
ed
a
s
tate
-
of
-
ar
t
wo
r
k
,
th
at
is
th
e
f
u
s
io
n
o
f
d
ata
s
ets
with
a
m
u
lti
-
le
v
el
f
u
s
io
n
ap
p
r
o
ac
h
.
T
h
e
e
x
is
tin
g
m
o
d
els
lik
e
SVM,
r
an
d
o
m
f
o
r
est
(
R
F)
,
an
d
e
n
s
em
b
le
m
o
d
els
r
ep
o
r
ted
9
0
%
–
9
5
%
ac
cu
r
ac
y
.
Dee
p
lear
n
in
g
-
b
ased
m
eth
o
d
s
,
s
u
ch
as
co
n
v
o
lu
tio
n
al
n
eu
r
al
n
et
wo
r
k
s
(
C
NNs
)
an
d
r
ec
u
r
r
en
t
n
eu
r
al
n
etwo
r
k
s
(
R
NNs
)
ac
h
iev
ed
9
2
%
–
9
6
%
ac
cu
r
ac
y
,
wh
ic
h
also
r
eq
u
ir
e
d
h
ig
h
co
m
p
u
tatio
n
al
r
eso
u
r
ce
s
f
o
r
ex
ec
u
tio
n
.
I
n
s
o
m
e
wo
r
k
,
th
e
y
u
tili
ze
d
s
tack
in
g
m
o
d
els
wh
ich
r
ep
o
r
t
ed
9
2
%
–
9
5
%
ac
cu
r
ac
y
,
b
u
t
t
h
ese
d
id
n
o
t
e
f
f
ec
tiv
ely
in
co
r
p
o
r
ate
m
u
ltis
o
u
r
ce
d
ata
f
u
s
io
n
.
I
n
o
u
r
p
r
o
p
o
s
ed
wo
r
k
f
ea
tu
r
e
-
lev
el
f
u
s
io
n
tech
n
i
q
u
e
im
p
r
o
v
e
d
in
s
tr
u
ctio
n
d
e
tectio
n
ac
cu
r
ac
y
b
y
in
teg
r
atin
g
two
v
ar
ieties
o
f
d
atasets
(
NSL
-
KDD
an
d
UNSW
-
NB
1
5
)
at
th
e
in
itial
s
tag
e,
in
th
e
s
ec
o
n
d
s
tag
e
th
e
lin
ea
r
d
is
cr
im
in
an
t
an
al
y
s
is
(
L
DA)
-
b
ased
tech
n
iq
u
e
was
u
s
ed
f
o
r
d
im
e
n
s
io
n
ality
r
ed
u
cti
o
n
wh
ic
h
o
p
tim
i
ze
d
th
e
co
m
p
u
tatio
n
al
ef
f
icien
cy
.
Ad
d
itio
n
ally
,
in
th
e
last
s
tag
e,
a
p
r
e
d
ictio
n
-
l
ev
el
f
u
s
io
n
o
f
tw
o
d
if
f
er
en
t
n
eu
r
al
n
etwo
r
k
m
o
d
els
was
u
s
ed
to
f
u
r
th
er
en
h
an
ce
th
e
p
er
f
o
r
m
an
ce
.
T
h
e
o
b
tain
ed
r
esu
lts
co
n
f
ir
m
e
d
th
e
ef
f
ec
tiv
en
ess
o
f
u
s
in
g
a
m
u
ltil
ev
el
f
u
s
io
n
a
p
p
r
o
ac
h
in
cy
b
er
s
ec
u
r
ity
t
o
id
en
tify
in
tr
u
s
io
n
attac
k
s
.
B
y
h
ig
h
lig
h
tin
g
its
p
o
ten
tial
f
o
r
r
ea
l
-
tim
e
i
n
tr
u
s
io
n
d
etec
tio
n
,
th
e
m
o
d
el
o
v
er
co
m
es
th
e
co
m
p
u
tatio
n
al
lim
itatio
n
s
o
f
tr
ad
itio
n
al
ML
an
d
d
ee
p
lea
r
n
in
g
m
eth
o
d
s
.
T
h
is
s
tu
d
y
alig
n
s
with
o
n
g
o
in
g
r
esear
ch
in
cy
b
er
s
ec
u
r
it
y
,
m
ac
h
i
n
e
lear
n
in
g
,
a
n
d
d
a
ta
f
u
s
io
n
,
m
ak
in
g
it
h
i
g
h
ly
r
ele
v
an
t
to
th
e
s
co
p
e
o
f
co
m
p
u
ter
en
g
i
n
ee
r
in
g
o
r
n
etwo
r
k
s
ec
u
r
ity
.
B
y
ad
d
r
ess
in
g
th
e
lim
itatio
n
s
o
f
tr
ad
itio
n
al
I
DS
m
o
d
els
an
d
in
tr
o
d
u
cin
g
a
n
o
v
el
m
u
ltil
ev
el
f
u
s
io
n
s
tr
at
eg
y
,
th
is
p
r
ese
n
ted
wo
r
k
p
r
o
v
id
es
s
ig
n
i
f
ican
t
co
n
tr
ib
u
tio
n
s
to
th
e
f
ield
o
f
in
tellig
en
t
n
etwo
r
k
s
ec
u
r
ity
.
T
h
e
p
r
i
m
ar
y
co
n
tr
ib
u
tio
n
s
o
f
th
is
r
esear
ch
a
r
e:
i
)
a
n
o
v
el
m
u
ltil
ev
el
f
u
s
io
n
s
tr
a
teg
y
th
at
im
p
r
o
v
es
I
DS
d
e
tectio
n
ac
cu
r
ac
y
,
ii
)
th
e
in
teg
r
atio
n
o
f
two
d
iv
er
s
e
d
atase
ts
(
NSL
-
KDD
an
d
UNSW
-
N
B
1
5
)
to
co
v
er
a
w
id
e
r
an
g
e
o
f
c
y
b
er
th
r
ea
ts
,
an
d
iii
)
a
h
y
b
r
id
n
eu
r
al
n
etwo
r
k
-
b
ased
p
r
ed
i
ctio
n
-
lev
el
f
u
s
io
n
th
at
e
n
h
a
n
ce
s
class
if
icatio
n
p
er
f
o
r
m
an
ce
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
15
,
No
.
4
,
Au
g
u
s
t
20
25
:
3
9
3
8
-
3948
3940
2.
L
I
T
E
R
AT
U
RE
SU
RVE
Y
Fro
m
r
ec
en
t
y
ea
r
s
r
esear
ch
o
n
u
s
in
g
m
u
ltis
en
s
o
r
y
d
ata
f
u
s
io
n
in
in
tr
u
s
io
n
d
etec
tio
n
s
y
s
tem
s
h
as
b
ee
n
g
r
o
win
g
r
ec
en
tly
.
Mu
ltis
en
s
o
r
y
d
ata
f
u
s
io
n
co
m
b
in
es
in
f
o
r
m
atio
n
f
r
o
m
m
u
ltip
le
s
en
s
o
r
s
o
r
d
ata
s
o
u
r
ce
s
to
o
b
tain
m
o
r
e
a
cc
u
r
ate
an
d
c
o
m
p
lete
in
f
o
r
m
atio
n
th
an
u
s
in
g
a
s
in
g
le
s
o
u
r
ce
o
f
d
ata.
I
n
in
tr
u
s
io
n
d
etec
tio
n
,
th
e
d
ata
f
u
s
io
n
m
eth
o
d
ca
n
u
t
ilize
v
ar
io
u
s
d
ata
s
o
u
r
ce
s
s
u
ch
as
n
etwo
r
k
tr
af
f
ic
l
o
g
s
,
s
y
s
tem
ev
en
t
lo
g
s
,
an
d
u
s
er
ac
tiv
ity
lo
g
s
.
Hall
an
d
L
li
n
as
[
3
]
in
th
eir
r
esear
ch
wo
r
k
s
h
o
wed
a
n
o
v
er
v
iew
o
f
m
u
lti
s
en
s
o
r
y
d
ata
f
u
s
io
n
ap
p
r
o
ac
h
b
y
c
o
v
er
in
g
k
ey
p
r
o
ce
s
s
m
o
d
els
an
d
tech
n
iq
u
es.
T
h
eir
r
esear
ch
d
e
m
o
n
s
tr
ated
th
e
u
s
e
o
f
d
ata
f
u
s
io
n
in
p
r
ac
tical
a
p
p
licatio
n
s
lik
e
au
to
m
ated
tar
g
et
r
ec
o
g
n
itio
n
,
b
attlef
ield
s
u
r
v
eillan
ce
,
an
d
co
m
p
lex
m
ac
h
in
er
y
m
o
n
ito
r
in
g
,
wh
ich
a
r
e
clo
s
ely
r
elate
d
to
n
etwo
r
k
in
tr
u
s
io
n
d
e
tectio
n
.
T
h
e
lim
itatio
n
s
o
f
tr
a
d
itio
n
al
in
tr
u
s
io
n
d
etec
tio
n
m
eth
o
d
s
c
an
b
e
o
v
er
co
m
e
u
s
in
g
m
u
lti
-
s
en
s
o
r
d
ata
f
u
s
io
n
ap
p
r
o
ac
h
es,
s
in
ce
tr
a
d
itio
n
al
m
eth
o
d
s
r
ely
o
n
s
in
g
le
s
o
u
r
ce
o
f
i
n
f
o
r
m
atio
n
.
R
ah
u
l
-
Vig
n
eswar
an
et
a
l
.
[
4
]
s
u
g
g
ested
t
h
at
m
a
n
y
ex
is
tin
g
in
t
r
u
s
io
n
d
etec
tio
n
s
y
s
tem
s
s
u
f
f
er
s
fr
om
h
ig
h
f
alse
d
ete
ctio
n
r
ates
a
n
d
lack
in
d
ec
is
io
n
s
u
p
p
o
r
t
f
o
r
th
e
i
n
c
id
en
t.
B
y
co
m
b
in
in
g
d
ata
f
r
o
m
d
if
f
er
e
n
t
s
o
u
r
ce
s
,
we
ca
n
i
n
cr
ea
s
e
th
e
s
tr
en
g
th
o
f
d
ata
s
ets an
d
also
f
ea
tu
r
es o
f
in
tr
u
s
io
n
s
f
o
r
wid
e
r
a
n
g
e
o
f
attac
k
s
ca
n
b
e
co
n
s
id
er
e
d
.
I
n
tr
u
s
io
n
d
etec
tio
n
s
y
s
tem
s
(
I
DS)
p
lay
a
k
e
y
r
o
le
i
n
s
af
eg
u
ar
d
in
g
m
o
d
er
n
n
etwo
r
k
s
a
g
a
in
s
t
ev
er
y
d
ay
ev
o
lv
in
g
v
ar
io
u
s
ty
p
es
o
f
cy
b
er
th
r
ea
ts
.
T
r
ad
itio
n
al
tech
n
iq
u
es
f
ail
to
m
an
ag
e
th
e
to
d
ay
’
s
co
m
p
lex
n
etwo
r
k
attac
k
s
.
T
o
ad
d
r
ess
th
ese
is
s
u
es,
r
esear
ch
er
s
s
tar
ted
u
s
in
g
ad
v
an
ce
d
m
ac
h
in
e
lear
n
in
g
m
o
d
els,
wh
ich
h
av
e
s
h
o
wed
p
r
o
m
is
in
g
r
esu
lt
s
b
y
en
h
a
n
cin
g
d
etec
tio
n
r
ates
[
5
]
,
[
6
]
.
On
e
o
f
th
e
m
o
s
t
p
r
o
m
is
in
g
tech
n
iq
u
es
is
to
ad
o
p
t
d
ata
f
u
s
io
n
ap
p
r
o
ac
h
es
in
NI
DS.
So
m
e
wo
r
k
s
ar
e
ex
p
lo
r
ed
in
th
is
ar
ea
[
7
]
,
s
h
o
wed
s
o
m
e
p
r
o
m
is
in
g
r
esu
lts
.
Hen
ce
,
r
esear
ch
er
s
s
tar
ted
u
s
in
g
d
ata
f
u
s
io
n
a
p
p
r
o
ac
h
es
to
o
v
e
r
co
m
e
lim
itatio
n
s
o
f
tr
ad
itio
n
al
m
eth
o
d
s
.
T
h
es
e
m
o
d
els
ex
ce
l
at
h
an
d
lin
g
lar
g
e
-
s
ca
le
d
ata
a
n
d
cr
ea
tin
g
m
o
r
e
e
f
f
ec
tiv
e
r
ep
r
esen
tatio
n
s
wh
ich
s
ig
n
if
ican
tly
im
p
r
o
v
e
in
tr
u
s
io
n
d
etec
tio
n
r
ate
[
7
]
,
[
8
]
.
T
h
e
p
r
im
ar
y
g
o
al
o
f
an
I
DS
is
to
id
en
tify
an
d
p
r
ev
en
t
b
o
t
h
k
n
o
wn
an
d
u
n
k
n
o
wn
cy
b
e
r
th
r
ea
ts
in
n
etwo
r
k
en
v
i
r
o
n
m
e
n
ts
.
T
r
ad
itio
n
al
m
eth
o
d
s
s
tr
u
g
g
le
to
k
ee
p
u
p
with
th
e
f
ast
-
ev
o
lv
in
g
th
r
e
at
lan
d
s
ca
p
e,
o
f
ten
r
esu
ltin
g
in
d
ec
r
ea
s
ed
ac
cu
r
ac
y
as
d
ata
s
ets
g
r
o
w
v
er
y
f
ast.
B
y
r
ec
o
g
n
izin
g
th
ese
s
h
o
r
tco
m
in
g
s
,
r
esear
ch
er
s
h
av
e
f
o
cu
s
ed
o
n
m
ac
h
i
n
e
lea
r
n
in
g
al
g
o
r
ith
m
s
lik
e
n
eu
r
al
n
etwo
r
k
s
,
SVMs
,
an
d
KNN
t
o
im
p
r
o
v
e
I
DS
[
9
]
.
Ho
wev
er
,
th
ese
m
et
h
o
d
s
also
f
ac
e
ch
allen
g
es
wit
h
h
ig
h
f
a
ls
e
p
o
s
itiv
e
an
d
alar
m
r
ates.
T
o
o
v
e
r
co
m
e
th
ese
is
s
u
es,
r
esear
ch
er
s
h
av
e
ex
p
lo
r
ed
d
ee
p
lear
n
in
g
m
o
d
els,
p
ar
ticu
lar
ly
lo
n
g
s
h
o
r
t
-
ter
m
m
em
o
r
y
(
L
STM
)
n
etwo
r
k
s
[
1
0
]
,
wh
ich
ar
e
b
ette
r
at
p
r
o
ce
s
s
in
g
a
n
d
class
if
y
in
g
co
m
p
lex
d
ata
p
atter
n
s
.
I
n
teg
r
atin
g
d
ata
f
u
s
io
n
tech
n
i
q
u
es
with
n
eu
r
al
n
etwo
r
k
lear
n
in
g
m
o
d
els
ca
n
f
u
r
th
er
en
h
an
ce
I
DS
p
er
f
o
r
m
an
ce
.
B
y
f
u
s
in
g
m
u
lt
ip
l
e
d
ata
s
o
u
r
ce
s
an
d
m
a
n
ip
u
latin
g
th
e
Neu
r
al
n
etwo
r
k
m
o
d
el
f
o
r
p
o
wer
f
u
l
f
ea
tu
r
e
ex
tr
ac
tio
n
,
I
DS
ca
n
m
o
r
e
ef
f
ec
tiv
el
y
d
etec
t
an
d
m
iti
g
ate
a
wid
e
r
an
g
e
o
f
n
etwo
r
k
th
r
ea
ts
[
2
]
,
[
5
]
,
[
7
]
,
[
8
]
,
[
1
1
]
.
Un
li
k
e
th
e
p
r
e
v
io
u
s
s
tu
d
ies
th
at
r
elied
o
n
s
in
g
le
-
s
o
u
r
ce
in
tr
u
s
io
n
d
etec
tio
n
o
r
tr
ad
itio
n
al
m
ac
h
in
e
lear
n
in
g
m
o
d
els,
th
is
r
esear
c
h
wo
r
k
d
em
o
n
s
tr
ates
th
at
th
e
m
u
ltis
o
u
r
ce
d
ata
f
u
s
io
n
a
p
p
r
o
ac
h
s
ig
n
if
ica
n
tly
en
h
an
ce
s
in
tr
u
s
io
n
d
etec
tio
n
a
cc
u
r
ac
y
.
T
h
e
p
r
o
p
o
s
ed
m
o
d
el
b
r
id
g
es
th
e
g
ap
b
etwe
en
f
ea
tu
r
e
-
lev
el
f
u
s
io
n
an
d
d
ee
p
lear
n
in
g
-
b
ased
p
r
ed
ictio
n
f
u
s
io
n
,
s
ettin
g
a
n
ew
b
e
n
ch
m
ar
k
f
o
r
NI
DS p
er
f
o
r
m
a
n
ce
.
3.
M
E
T
H
O
DO
L
O
G
Y
E
n
g
in
ee
r
e
d
ap
p
r
o
ac
h
es
an
d
t
ec
h
n
iq
u
es
in
th
e
f
ield
o
f
d
at
a
f
u
s
io
n
f
o
cu
s
o
n
ef
f
ec
tiv
el
y
m
er
g
in
g
in
f
o
r
m
atio
n
f
r
o
m
m
u
ltip
le
s
o
u
r
ce
s
o
f
d
ata.
T
h
e
k
ey
f
u
s
io
n
ar
ch
itectu
r
es
in
clu
d
e
ce
n
tr
ali
ze
d
,
d
ec
en
t
r
alize
d
,
an
d
d
is
tr
ib
u
ted
m
o
d
els,
ea
ch
in
v
o
lv
in
g
v
ar
io
u
s
lev
els
o
f
ab
s
tr
ac
tio
n
i
n
th
e
f
u
s
io
n
p
r
o
ce
s
s
lik
e
r
aw
m
ea
s
u
r
em
en
ts
,
s
ig
n
als
an
d
c
h
ar
ac
ter
is
tics
,
o
r
d
ec
is
io
n
s
.
At
lo
wer
ab
s
tr
ac
tio
n
lev
els,
th
e
f
u
s
io
n
tech
n
i
q
u
es
m
ay
wo
r
k
with
r
aw
s
en
s
o
r
m
e
asu
r
em
en
ts
o
r
s
ig
n
al
-
lev
el
d
at
a,
u
s
in
g
s
tatis
tical
m
eth
o
d
s
.
T
h
is
p
ap
er
d
escr
ib
es
a
m
u
ltil
ev
el
an
d
m
u
ltis
o
u
r
ce
d
ata
f
u
s
io
n
ap
p
r
o
ac
h
d
esig
n
ed
to
en
h
an
ce
p
er
f
o
r
m
an
ce
in
n
etwo
r
k
in
tr
u
s
io
n
d
etec
tio
n
s
y
s
tem
s
an
d
ad
d
r
ess
th
ese
ch
allen
g
es.
I
n
th
is
p
ap
er
,
a
n
o
v
e
l
m
eth
o
d
was
p
r
o
p
o
s
ed
wh
e
r
e
d
ata
f
u
s
io
n
was
a
p
p
lied
at
v
ar
i
o
u
s
s
tag
es
an
d
also
d
ata
d
im
e
n
s
io
n
a
lity
r
ed
u
ctio
n
was
ap
p
lied
to
im
p
r
o
v
e
th
e
p
er
f
o
r
m
an
ce
o
f
class
if
icatio
n
ac
cu
r
ac
y
.
T
h
e
p
r
o
p
o
s
ed
m
o
d
el
with
m
u
lti
-
lev
el
m
u
lti
-
s
o
u
r
ce
d
ata
f
u
s
io
n
is
s
h
o
wn
in
Fig
u
r
e
1
.
T
h
e
m
o
d
el
we
b
u
ilt
u
s
ed
d
ata
f
u
s
io
n
ap
p
r
o
ac
h
es,
b
ec
au
s
e
th
e
tr
ad
itio
n
al
I
DS
s
tr
u
g
g
le
with
l
im
ited
d
ata
s
o
u
r
ce
s
,
h
en
ce
co
m
b
i
n
in
g
m
u
ltip
le
s
o
u
r
ce
s
o
f
d
ata
im
p
r
o
v
es
d
etec
tio
n
ac
cu
r
ac
y
.
W
e
u
s
ed
th
e
d
im
en
s
io
n
ality
r
ed
u
ctio
n
tech
n
iq
u
e
s
in
ce
h
ig
h
-
d
im
en
s
io
n
al
d
ata
lead
s
to
co
m
p
u
tatio
n
al
in
ef
f
icien
cies,
s
o
L
DA
o
p
tim
izes
f
ea
t
u
r
e
s
el
ec
tio
n
an
d
r
e
d
u
ce
s
th
e
d
ata
s
ize.
W
e
ap
p
lied
a
p
r
ed
ictio
n
-
le
v
el
f
u
s
io
n
ap
p
r
o
a
ch
b
ec
au
s
e
Sin
g
le
-
m
o
d
el
a
p
p
r
o
ac
h
es
h
av
e
lim
itatio
n
s
,
b
u
t
c
o
m
b
in
in
g
m
u
ltip
le
n
eu
r
al
n
etwo
r
k
o
u
tp
u
ts
im
p
r
o
v
es
r
o
b
u
s
tn
ess
.
T
h
is
p
ap
e
r
p
r
o
v
id
es
m
ath
em
atica
l
f
o
r
m
u
lati
o
n
s
(
1
)
-
(
8
)
f
o
r
ea
ch
ex
ec
u
tio
n
s
tep
an
d
also
th
e
A
lg
o
r
ith
m
ic
d
escr
ip
tio
n
s
p
r
o
v
i
d
ed
f
o
r
th
e
p
r
o
p
o
s
ed
m
u
ltil
ev
el
an
d
m
u
ltis
o
u
r
ce
d
ata
f
u
s
io
n
a
p
p
r
o
ac
h
to
e
n
s
u
r
e
th
at
r
esear
ch
er
s
ca
n
r
ep
licate
th
e
m
eth
o
d
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
Mu
ltil
ev
el
a
n
d
mu
ltis
o
u
r
ce
d
a
ta
fu
s
io
n
a
p
p
r
o
a
c
h
fo
r
n
etw
o
r
k
in
tr
u
s
io
n
…
(
Ha
r
s
h
ith
a
S
o
ma
s
h
ek
a
r
)
3941
3
.
1
.
L
ev
el
1
:
F
ea
t
ure
-
lev
el
f
us
io
n
I
n
th
e
f
ir
s
t
lev
el,
f
ea
tu
r
es
f
r
o
m
d
atasets
NSL
-
KDD
an
d
U
NSW
-
N
B
1
5
ar
e
f
u
s
ed
u
s
in
g
a
n
in
n
er
j
o
in
tech
n
iq
u
e
b
ased
o
n
c
o
m
m
o
n
f
ea
tu
r
es
s
u
ch
as
p
r
o
to
c
o
l,
s
er
v
ice,
an
d
lab
el.
T
h
e
f
u
s
io
n
o
f
f
ea
tu
r
es
f
r
o
m
th
ese
s
o
u
r
ce
s
m
ak
es
a
d
etailed
p
er
c
ep
tio
n
o
f
d
ata
h
ap
p
e
n
in
g
in
th
e
n
etwo
r
k
u
n
d
er
o
b
s
er
v
atio
n
,
wh
ich
will
en
h
an
ce
b
et
ter
d
etec
tio
n
o
f
b
o
th
k
n
o
wn
an
d
u
n
k
n
o
wn
attac
k
s
o
r
th
r
ea
ts
.
T
h
e
p
r
o
p
o
s
ed
m
et
h
o
d
s
f
o
r
f
ea
tu
r
e
-
le
v
el
f
u
s
io
n
in
lev
el
1
ar
e
s
h
o
wn
in
m
ath
em
atica
l f
o
r
m
.
a.
Featu
r
e
s
elec
tio
n
o
n
th
e
NSL
-
KDD
d
ataset:
F
N
S
L
−
K
D
D
=
Selec
tFeatu
r
es
(
D
N
S
L
−
KDD
,
15
)
(
1
)
w
h
er
e
D
N
S
L
−
KDD
i
s
t
h
e
NSL
-
KDD
d
ataset,
an
d
F
NS
L
−
KD
D
ar
e
th
e
s
elec
ted
f
ea
tu
r
e.
b.
Selectin
g
f
ea
tu
r
es f
r
o
m
th
e
U
NSW
-
N
B
1
5
d
ataset:
F
U
N
S
W
−
NB
15
=
SelectFea
tu
r
es
(
D
U
N
S
W
−
NB
15
,
15
)
(
2
)
w
h
er
e
D
UN
S
W−
N
B1
5
is
th
e
U
NSW
-
N
B
1
5
d
ataset,
an
d
F
U
N
S
W−
N
B
1
5
ar
e
th
e
s
elec
ted
f
ea
tu
r
es.
c.
Data
s
et
f
u
s
io
n
in
n
er
jo
in
th
e
s
elec
ted
f
ea
tu
r
es:
F
f
u
s
ed
=
F
N
SL
−
K
D
D
∩
F
U
N
S
W
−
NB
1
5
(
3
)
w
h
er
e
th
e
in
ter
s
ec
tio
n
is
p
er
f
o
r
m
ed
o
n
th
e
co
l
u
m
n
s
p
r
o
to
co
l,
s
er
v
ice,
an
d
la
b
el.
Fig
u
r
e
1
.
T
h
e
p
r
o
p
o
s
ed
a
p
p
r
o
ac
h
in
v
o
l
v
es th
r
ee
lev
els o
f
d
a
ta
f
u
s
io
n
3
.
2
.
L
ev
el
2
:
Dim
ens
io
na
lity
r
educt
io
n
I
n
th
e
p
r
o
p
o
s
ed
m
o
d
el
th
e
s
ec
o
n
d
lev
el
o
f
t
h
e
ap
p
r
o
ac
h
in
v
o
lv
es
d
im
en
s
io
n
ality
r
ed
u
ctio
n
[
1
2
]
.
T
h
e
1
2
f
ea
tu
r
e
c
o
lu
m
n
s
f
r
o
m
th
e
f
u
s
ed
d
ataset
ar
e
r
ed
u
ce
d
to
a
s
in
g
le
f
ea
tu
r
e
co
lu
m
n
u
s
in
g
L
DA.
T
h
is
d
im
en
s
io
n
ality
r
ed
u
ctio
n
s
tep
h
elp
s
to
m
itig
ate
th
e
cu
r
s
e
o
f
d
im
en
s
io
n
ality
an
d
im
p
r
o
v
e
t
h
e
ef
f
icien
cy
o
f
t
h
e
s
u
b
s
eq
u
en
t
p
r
e
d
ictio
n
m
o
d
els.
T
h
e
p
r
o
p
o
s
ed
m
eth
o
d
s
f
o
r
d
i
m
en
s
io
n
ality
r
ed
u
ctio
n
in
lev
e
l
2
in
m
ath
em
atica
l
f
o
r
m
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
15
,
No
.
4
,
Au
g
u
s
t
20
25
:
3
9
3
8
-
3948
3942
a.
Dim
en
s
io
n
ality
r
ed
u
ctio
n
u
s
in
g
L
DA
a
p
p
ly
L
DA:
X
L
D
A
=
LDA(F
f
us
ed
,
n
=
1
2
,
m
=
1
)
(
4
)
3
.
3
.
L
ev
el
3
:
P
re
dict
io
n
-
lev
e
l f
us
io
n
T
h
e
last
lev
el
in
th
e
p
r
o
p
o
s
e
d
ap
p
r
o
ac
h
is
p
r
e
d
ictio
n
-
lev
e
l
f
u
s
io
n
[
1
3
]
.
Her
e,
two
n
eu
r
al
n
etwo
r
k
m
o
d
els
h
av
e
b
ee
n
u
s
ed
,
o
u
t
o
f
wh
ich
o
n
e
h
as
a
s
in
g
le
in
p
u
t
n
o
d
e,
two
h
id
d
e
n
n
o
d
es,
an
d
two
o
u
tp
u
t
n
o
d
es,
wh
ile
th
e
s
ec
o
n
d
m
o
d
el
h
as
a
s
in
g
le
in
p
u
t
n
o
d
e,
t
h
r
ee
h
i
d
d
en
n
o
d
es,
an
d
two
o
u
t
p
u
t
n
o
d
es.
T
h
ese
o
u
t
p
u
ts
f
r
o
m
th
e
two
n
e
u
r
al
n
etwo
r
k
m
o
d
els
ar
e
th
e
n
f
u
s
ed
u
s
in
g
a
p
r
ed
ictio
n
f
u
s
io
n
tec
h
n
iq
u
e.
I
n
p
r
ed
ictio
n
f
u
s
io
n
,
m
u
ltip
le
m
o
d
els
f
u
s
e
to
u
p
g
r
ad
e
th
e
s
tr
en
g
th
s
o
f
th
e
d
if
f
er
en
t
n
e
u
r
al
m
o
d
els
f
o
r
o
v
er
all
im
p
r
o
v
ed
p
er
f
o
r
m
an
ce
in
class
if
icatio
n
ac
cu
r
ac
y
.
T
h
e
l
ev
el
3
p
r
ed
icti
o
n
lev
el
f
u
s
io
n
is
s
h
o
wn
b
elo
w
in
m
ath
em
atica
l
f
o
r
m
.
a.
Neu
r
al
n
etwo
r
k
m
o
d
el
1
O
1
=
NN
1
(
X
L
D
A
,
[
1
]
,
[
2
]
,
[
2
]
,
s
ig
m
o
id
)
(
5
)
w
h
er
e
O1
is
th
e
o
u
t
p
u
t o
f
th
e
f
ir
s
t n
eu
r
al
n
etwo
r
k
m
o
d
el.
b.
Neu
r
al
n
etwo
r
k
m
o
d
el
2
O
2
=
NN
2
(
X
L
D
A
,
[
1
]
,
[
3
]
,
[
2
]
,
s
o
f
tp
lu
s
)
(
6
)
w
h
er
e
O2
is
th
e
o
u
t
p
u
t o
f
th
e
s
ec
o
n
d
n
eu
r
al
n
etwo
r
k
m
o
d
el.
c.
Ou
tp
u
t
f
u
s
io
n
: Fu
s
e
o
u
tp
u
ts
f
r
o
m
th
e
two
m
o
d
els:
O
f
us
e
d
=
O
1
+
O
2
2
(7
)
Ap
p
ly
th
e
d
en
s
e
n
etwo
r
k
th
r
es
h
o
ld
:
O
f
i
n
a
l
=
{
1
if
O
f
u
s
ed
>
T
,
O
o
th
er
wis
e
(
8
)
w
h
er
e
T
is
th
e
d
en
s
e
n
etwo
r
k
t
h
r
esh
o
ld
.
d.
E
v
alu
atio
n
o
f
th
e
f
u
s
ed
o
u
tp
u
t
:
R
=E
va
lu
a
te
(
O
f
i
n
a
l
)
w
h
er
e
R
is
th
e
ev
alu
atio
n
r
esu
lt.
E
q
u
atio
n
(
1
)
to
(
8
)
g
iv
es th
e
f
o
r
m
atio
n
o
f
d
ata
at
ea
c
h
lev
el
in
th
e
m
ath
em
atica
l f
o
r
m
f
o
r
a
n
aly
s
is
.
4.
P
RO
P
O
SE
D
AL
G
O
R
I
T
H
M
:
M
UL
T
I
L
E
V
E
L
AND
M
U
L
T
I
SO
URC
E
DA
T
A
F
US
I
O
N
(
M
M
DF
)
AP
P
RO
ACH
F
O
R
NIDS
T
h
e
MM
DF
m
eth
o
d
o
lo
g
y
w
as
im
p
lem
en
ted
in
a
NI
DS
u
tili
zin
g
th
e
f
o
llo
win
g
alg
o
r
ith
m
.
T
h
e
m
eth
o
d
its
elf
is
b
r
o
k
en
d
o
wn
in
to
th
r
ee
m
ain
s
tag
es
in
ten
d
e
d
to
im
p
r
o
v
e
d
etec
tio
n
r
elev
a
n
cy
an
d
ca
lcu
latio
n
tim
e.
I
n
th
e
in
itial
s
tag
e,
d
ata
p
r
ep
r
o
ce
s
s
in
g
an
d
f
ea
tu
r
e
s
elec
tio
n
ar
e
th
e
f
o
cu
s
.
Fo
r
th
is
ex
am
p
le,
we
lo
ad
th
e
NSL
-
KDD
an
d
UNSW
-
N
B
1
5
d
atasets
th
at
r
ep
r
esen
t
d
if
f
e
r
e
n
t
ty
p
es
o
f
n
etwo
r
k
tr
af
f
ic
f
l
o
ws.
T
o
o
b
tain
a
r
ich
an
d
in
f
o
r
m
ativ
e
f
ea
tu
r
e
s
et,
1
5
f
ea
tu
r
es
ar
e
ex
tr
ac
ted
f
r
o
m
e
ac
h
d
ataset.
T
h
e
r
esu
ltan
t
f
ea
tu
r
es
ar
e
th
en
jo
in
ed
u
s
in
g
an
in
n
er
jo
i
n
b
ased
o
n
co
m
m
o
n
attr
ib
u
tes
(
i.e
.
,
p
r
o
t
o
co
l,
s
er
v
ice,
an
d
lab
el)
.
I
t
i
s
an
ess
en
tial
s
tep
in
p
r
ep
ar
in
g
th
e
f
i
n
al
d
ataset,
r
et
ain
in
g
o
n
ly
r
elev
a
n
t d
ata
f
r
o
m
b
o
t
h
o
r
ig
in
al
d
atasets
.
T
h
is
s
ec
o
n
d
s
tag
e
r
elate
s
to
f
ea
tu
r
e
tr
an
s
f
o
r
m
atio
n
d
esig
n
e
d
to
m
a
k
e
ca
lcu
latio
n
s
m
o
r
e
e
f
f
icien
t,
as
well
as
to
im
p
r
o
v
e
m
o
d
el
ac
c
u
r
ac
y
.
At
th
is
s
tag
e,
t
h
e
m
e
r
g
ed
d
ataset
is
s
p
lit
in
to
in
d
e
p
e
n
d
en
t
v
ar
iab
les
(
X)
an
d
tar
g
et
v
ar
ia
b
les
(
y
)
,
en
s
u
r
in
g
s
tr
u
ctu
r
ed
d
ata
o
r
g
an
izat
io
n
f
o
r
tr
ai
n
in
g
.
Sin
ce
th
e
f
e
atu
r
e
s
p
ac
e
is
h
i
g
h
d
im
en
s
io
n
al,
th
is
u
s
es
L
DA
t
o
co
n
d
en
s
e
m
u
ltip
le
f
ea
tu
r
es
in
to
a
s
in
g
le,
o
p
tim
ized
r
ep
r
e
s
en
tatio
n
.
T
h
is
s
tep
r
ed
u
ce
s
d
im
en
s
io
n
s
b
u
t r
etain
s
th
e
n
ec
ess
ar
y
in
f
o
r
m
atio
n
f
o
r
m
a
in
tain
in
g
s
ep
a
r
ab
ilit
y
am
o
n
g
d
if
f
e
r
en
t c
lass
es
wh
ile
co
n
s
id
er
ab
ly
r
ed
u
cin
g
co
m
p
lex
ity
,
th
u
s
m
ak
in
g
th
e
d
ataset
m
o
r
e
am
en
ab
le
f
o
r
class
if
icatio
n
u
s
in
g
m
ac
h
in
e
lear
n
in
g
-
b
ased
tec
h
n
i
q
u
es.
T
h
e
last
s
tag
e
is
f
o
cu
s
ed
o
n
t
h
e
d
ev
elo
p
m
en
t,
tr
ai
n
in
g
&
e
v
alu
atio
n
o
f
th
e
m
o
d
el.
W
e
d
ef
in
e
two
d
if
f
er
en
t
n
eu
r
al
n
etwo
r
k
s
,
with
d
if
f
er
e
n
t
co
m
p
lex
ities
.
I
n
th
e
f
ir
s
t
m
o
d
el,
we
h
a
v
e
o
n
e
in
p
u
t
lay
er
-
two
h
id
d
en
lay
er
s
-
a
n
d
two
o
u
t
p
u
t
n
o
d
es
u
s
in
g
a
s
ig
m
o
id
ac
ti
v
atio
n
f
u
n
ctio
n
.
T
h
e
s
ec
o
n
d
m
o
d
el
h
as
o
n
e
in
p
u
t
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
Mu
ltil
ev
el
a
n
d
mu
ltis
o
u
r
ce
d
a
ta
fu
s
io
n
a
p
p
r
o
a
c
h
fo
r
n
etw
o
r
k
in
tr
u
s
io
n
…
(
Ha
r
s
h
ith
a
S
o
ma
s
h
ek
a
r
)
3943
lay
er
,
th
r
ee
h
id
d
en
lay
e
r
s
,
two
o
u
tp
u
t n
o
d
es,
an
d
a
s
o
f
t
p
l
u
s
ac
tiv
atio
n
f
u
n
ctio
n
.
W
e
th
e
n
s
p
lit
o
u
r
d
ataset
in
t
o
tr
ain
an
d
test
d
atasets
an
d
tr
ain
b
o
th
m
o
d
els
s
ep
ar
ately
.
Fo
llo
win
g
th
at,
th
ei
r
p
r
ed
ictio
n
s
o
n
test
d
ata
ar
e
en
s
em
b
led
u
s
in
g
a
n
av
er
a
g
e
t
ec
h
n
iq
u
e
with
a
d
ef
in
e
d
th
r
es
h
o
ld
to
im
p
r
o
v
e
th
e
class
if
icatio
n
r
atio
.
Fin
ally
,
th
e
ac
cu
r
ac
y
_
s
co
r
e
f
u
n
ctio
n
ch
ec
k
s
th
e
f
in
al
p
r
e
d
ictio
n
s
,
p
r
o
v
id
i
n
g
a
co
m
p
r
eh
en
s
iv
e
p
er
f
o
r
m
a
n
ce
ev
alu
atio
n
.
Usi
n
g
th
is
tech
n
iq
u
e,
we
ca
n
m
itig
ate
th
e
wea
k
n
ess
es
o
f
s
in
g
le
m
o
d
els
e
f
f
ec
t
iv
ely
,
a
n
d
im
p
r
o
v
e
th
e
ca
p
ab
ilit
ies
o
f
th
e
in
tr
u
s
io
n
d
etec
tio
n
p
r
o
ce
s
s
in
ter
m
s
o
f
ac
cu
r
ac
y
an
d
r
o
b
u
s
tn
ess
b
y
in
teg
r
atin
g
m
u
ltip
le
m
o
d
els an
d
d
ata
s
o
u
r
ce
s
.
T
h
e
s
tep
-
by
-
s
tep
p
s
eu
d
o
co
d
e
o
f
m
u
ltil
ev
el
an
d
m
u
ltis
o
u
r
c
e
d
ata
f
u
s
io
n
ap
p
r
o
ac
h
f
o
r
NI
DS
is
s
h
o
wn
in
th
e
Fig
u
r
e
2
.
T
h
e
g
e
n
er
al
o
v
e
r
v
ie
w
o
f
ea
ch
s
tep
s
ar
e:
−
At
th
e
b
e
g
in
n
in
g
t
h
e
b
o
th
NSL
-
KDD
an
d
UNSW
-
NB
1
5
d
a
tasets
wer
e
n
o
r
m
alize
d
.
T
h
e
f
ea
tu
r
es
p
r
o
to
co
l,
s
er
v
ice
an
d
lab
els ar
e
u
s
ed
as k
ey
c
o
lu
m
n
s
f
o
r
f
u
s
io
n
.
−
Per
f
o
r
m
ed
an
i
n
n
er
jo
in
o
p
er
atio
n
o
n
s
elec
ted
f
ea
tu
r
es
to
co
m
b
in
e
f
ea
tu
r
es
b
y
ap
p
ly
i
n
g
f
ea
t
u
r
e
lev
el
f
u
s
io
n
tech
n
i
q
u
e.
W
h
ich
h
elp
ed
in
k
ee
p
in
g
all
ty
p
es
o
f
atta
ck
s
f
r
o
m
b
o
th
t
h
e
d
ata
s
ets
in
a
s
in
g
le
f
u
s
ed
d
ata
s
et
.
−
B
y
ap
p
ly
in
g
L
DA
th
e
f
ea
tu
r
e
co
lu
m
n
1
2
is
r
ed
u
ce
d
to
1
,
wh
ich
in
cr
ea
s
es th
e
p
er
f
o
r
m
an
ce
o
f
th
e
m
o
d
el.
−
Neu
r
al
n
etwo
r
k
ar
c
h
itectu
r
es:
Mo
d
el
1
:
1
in
p
u
t
lay
er
,
2
h
id
d
en
,
2
o
u
tp
u
t,
ac
tiv
atio
n
=
s
ig
m
o
id
No
te:
R
ep
lace
d
ac
tiv
atio
n
with
a
less
u
s
ed
ac
tiv
atio
n
f
u
n
ctio
n
s
o
f
t
p
lu
s
.
−
Fu
s
io
n
:
T
h
e
o
u
t
p
u
ts
o
f
b
o
th
m
o
d
els
wer
e
av
er
ag
e
d
f
o
r
ef
f
ec
tiv
e
d
ec
is
io
n
-
m
ak
in
g
.
Mo
d
e
l
p
er
f
o
r
m
a
n
ce
&
E
v
alu
atio
n
:
ac
cu
r
ac
y
,
p
r
ec
is
io
n
,
r
ec
all
an
d
F1
-
s
co
r
es we
r
e
c
alcu
lated
to
co
m
p
ar
e
p
e
r
f
o
r
m
a
n
ce
.
Fig
u
r
e
2
.
Alg
o
r
ith
m
MM
DF
-
m
u
ltil
ev
el
an
d
m
u
ltis
o
u
r
ce
d
at
a
f
u
s
io
n
ap
p
r
o
ac
h
f
o
r
NI
DS
5.
E
XP
E
R
I
M
E
N
T
A
L
RE
SUL
T
S
T
h
e
p
r
o
p
o
s
ed
m
u
ltil
ev
el
an
d
m
u
ltis
o
u
r
ce
d
ata
f
u
s
io
n
ap
p
r
o
ac
h
ac
h
iev
ed
a
class
if
icatio
n
ac
cu
r
ac
y
o
f
9
7
.
5
%.
T
h
e
m
o
s
t
ef
f
icien
t
ac
c
u
r
ac
y
r
esu
lts
ac
h
iev
ed
ar
e
b
ec
au
s
e
o
f
d
if
f
e
r
en
t f
u
s
io
n
tech
n
i
q
u
es
u
s
ed
at
v
ar
io
u
s
s
tag
es.
Featu
r
e
-
lev
el
f
u
s
io
n
co
m
b
in
ed
u
s
ef
u
l f
ea
tu
r
es f
r
o
m
b
o
th
NSL
-
KDD
an
d
UNS
W
-
N
B
1
5
d
atasets
,
wh
ich
p
r
o
v
id
e
d
a
m
o
r
e
co
m
p
lete
v
ie
w
o
f
n
etwo
r
k
ac
tiv
ity
,
w
h
ich
in
tu
r
n
h
elp
e
d
in
im
p
r
o
v
in
g
t
h
e
d
etec
tio
n
o
f
b
o
t
h
k
n
o
wn
an
d
u
n
k
n
o
w
n
attac
k
s
.
T
h
e
L
DA
is
u
s
ed
f
o
r
d
im
e
n
s
io
n
ality
r
e
d
u
ctio
n
,
wh
ich
h
elp
ed
to
o
v
er
c
o
m
e
t
h
e
is
s
u
es
with
lar
g
e
d
atasets
an
d
m
ad
e
th
e
p
r
ed
ictio
n
m
o
d
els
m
o
r
e
e
f
f
icien
t.
A
d
d
itio
n
ally
,
co
m
b
in
in
g
th
e
two
d
if
f
er
en
t
n
e
u
r
al
n
etwo
r
k
m
o
d
e
ls
at
th
e
p
r
ed
ictio
n
lev
el
to
o
k
ad
v
an
tag
e
o
f
th
eir
u
n
iq
u
e
s
tr
e
n
g
th
s
f
o
r
im
p
r
o
v
in
g
class
if
icatio
n
ac
cu
r
ac
y
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
15
,
No
.
4
,
Au
g
u
s
t
20
25
:
3
9
3
8
-
3948
3944
Pre
v
io
u
s
r
esear
ch
h
as
s
h
o
wn
th
e
b
en
ef
its
o
f
u
tili
zin
g
in
f
o
r
m
atio
n
f
u
s
io
n
in
ar
ea
s
lik
e
f
au
lt
d
iag
n
o
s
is
an
d
id
e
n
tify
in
g
f
a
u
lt
d
etec
ti
o
n
.
T
h
is
r
esear
ch
wo
r
k
d
e
m
o
n
s
tr
ates
th
e
ef
f
ec
tiv
en
ess
o
f
a
m
u
ltil
ev
el
an
d
m
u
ltis
o
u
r
ce
d
ata
f
u
s
io
n
a
p
p
r
o
ac
h
f
o
r
n
etwo
r
k
i
n
tr
u
s
io
n
d
etec
tio
n
,
wh
ich
is
p
iv
o
tal
f
o
r
cy
b
er
s
ec
u
r
ity
.
T
h
e
s
im
u
latio
n
m
o
d
el
is
s
h
o
wn
in
Fig
u
r
e
3
.
Fig
u
r
e
3
.
T
h
e
p
r
o
p
o
s
ed
a
p
p
r
o
ac
h
s
im
u
latio
n
m
o
d
el
u
s
in
g
K
NI
ME
T
h
e
KNI
ME
wo
r
k
f
lo
w
in
t
h
e
Fig
u
r
e
3
s
h
o
ws
th
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
Simu
latio
n
m
o
d
el.
T
h
e
s
im
u
latio
n
s
tep
s
ar
e
lis
ted
b
elo
w,
all
b
o
ld
wo
r
d
s
in
s
tep
s
r
ep
r
esen
ts
ea
ch
n
o
d
e
i
n
KNI
ME
.
a.
Data
p
r
ep
r
o
ce
s
s
in
g
:
−
T
wo
tab
le
r
ea
d
er
s
lo
a
d
d
atasets
.
−
C
o
lu
m
n
f
ilter
s
r
em
o
v
e
u
n
n
ec
e
s
s
ar
y
f
ea
tu
r
es.
−
Dim
en
s
io
n
ality
r
ed
u
ctio
n
(
L
D
A)
r
ed
u
ce
s
f
ea
tu
r
e
d
im
en
s
io
n
s
.
−
On
e
to
m
an
y
n
o
d
es sp
lit d
ata
f
o
r
m
u
ltip
le
m
o
d
els.
b.
Neu
r
al
n
etwo
r
k
t
r
ain
in
g
:
−
Ker
as
in
p
u
t la
y
er
s
d
ef
i
n
e
in
p
u
t stru
ctu
r
es.
−
Ker
as
d
en
s
e
lay
er
s
cr
ea
te
h
id
d
en
lay
er
s
with
ac
tiv
atio
n
f
u
n
ct
io
n
s
(
R
eL
U
an
d
Sig
m
o
id
/So
f
t
PLUS)
.
−
Ker
as
n
etwo
r
k
lear
n
e
r
s
tr
ain
m
o
d
els u
s
in
g
b
ac
k
p
r
o
p
ag
atio
n
an
d
Ad
am
o
p
tim
izer
.
c.
Mo
d
el
ex
ec
u
tio
n
an
d
p
r
e
d
ictio
n
:
−
Ker
as
n
etwo
r
k
ex
ec
u
to
r
s
ap
p
l
y
tr
ain
ed
m
o
d
els.
−
C
SV
w
r
iter
s
s
to
r
e
r
esu
lts
.
d.
Pre
d
ictio
n
Pro
ce
s
s
in
g
:
−
R
u
le
en
g
in
es e
x
tr
ac
t
p
r
e
d
ictio
n
s
.
−
C
o
lu
m
n
f
ilter
s
r
ef
in
e
o
u
tp
u
ts
.
−
C
o
lu
m
n
a
p
p
e
n
d
er
s
co
m
b
in
e
p
r
ed
ictio
n
s
.
e.
E
v
alu
atio
n
:
−
Sco
r
er
s
ass
ess
m
o
d
el
ac
cu
r
ac
y
.
−
C
SV
w
r
iter
s
s
to
r
e
ev
alu
atio
n
m
etr
ics.
T
h
e
p
r
o
p
o
s
ed
wo
r
k
f
lo
w
ef
f
e
ctiv
ely
h
an
d
le
s
in
p
u
t
d
ata,
f
ea
tu
r
e
f
u
s
io
n
,
d
ee
p
lear
n
i
n
g
tr
ain
in
g
,
p
r
ed
ictio
n
f
u
s
io
n
an
d
m
o
d
el
v
alid
atio
n
i
n
KNI
ME
.
T
ab
le
1
s
h
o
ws
th
e
r
esu
lts
o
f
e
x
p
er
im
en
tatio
n
,
wh
er
e
th
e
p
r
o
p
o
s
ed
m
o
d
el
u
p
h
o
l
d
s
th
e
ac
c
u
r
ac
y
.
T
h
e
p
r
o
p
o
s
ed
m
o
d
el
is
ab
le
to
f
in
d
th
e
n
etwo
r
k
attac
k
s
o
f
b
o
t
h
d
ata
s
ets
u
s
ed
f
o
r
ex
p
er
im
en
tati
o
n
.
Fig
u
r
e
4
s
h
o
ws
th
e
co
m
p
ar
is
o
n
g
r
a
p
h
,
wh
e
r
e
t
h
e
p
r
o
p
o
s
ed
m
et
h
o
d
s
h
o
wed
a
p
r
o
m
in
e
n
t r
esu
lt.
Sev
er
al
s
tu
d
ies
h
av
e
ex
p
lo
r
ed
m
ac
h
in
e
lear
n
in
g
-
b
ased
i
n
tr
u
s
io
n
d
etec
tio
n
,
with
ex
p
er
im
e
n
t
m
o
d
els
s
u
ch
as
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
i
n
es
(
SVM)
,
k
-
n
ea
r
est
n
ei
g
h
b
o
r
s
(
K
NN)
,
d
ec
is
io
n
tr
ee
s
(
DT
)
,
an
d
r
a
n
d
o
m
f
o
r
ests
(
R
F).
T
ab
le
2
s
h
o
ws
th
e
c
o
m
p
ar
is
o
n
o
f
r
esu
lts
with
th
e
ex
i
s
tin
g
m
o
d
els.
Ou
r
m
u
ltil
ev
el
an
d
m
u
ltis
o
u
r
ce
d
at
a
f
u
s
io
n
ap
p
r
o
ac
h
s
u
r
p
ass
es
d
e
ep
lear
n
in
g
-
b
ased
I
DS
in
th
r
e
e
k
ey
way
s
:
B
etter
ac
cu
r
ac
y
(
9
7
.
5
%)
th
a
n
C
NNs
an
d
R
NNs,
d
u
e
to
m
u
ltis
o
u
r
ce
d
ata
in
teg
r
atio
n
.
L
o
wer
f
alse
p
o
s
itiv
es
(
2
.
8
%),
as
p
r
e
d
ictio
n
-
lev
el
f
u
s
io
n
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
Mu
ltil
ev
el
a
n
d
mu
ltis
o
u
r
ce
d
a
ta
fu
s
io
n
a
p
p
r
o
a
c
h
fo
r
n
etw
o
r
k
in
tr
u
s
io
n
…
(
Ha
r
s
h
ith
a
S
o
ma
s
h
ek
a
r
)
3945
r
ef
in
es
class
if
icatio
n
r
esu
lts
.
R
ed
u
ce
d
d
im
en
s
io
n
ality
,
as
L
DA
d
ec
r
ea
s
es
th
e
n
u
m
b
er
o
f
p
r
ed
icto
r
v
ar
iab
les
,
wh
il
e
d
ee
p
lear
n
in
g
tech
n
iq
u
es
o
f
ten
w
o
r
k
with
s
u
ch
h
i
g
h
-
d
im
en
s
io
n
al
i
n
p
u
t.
Dee
p
lear
n
in
g
-
b
ased
m
eth
o
d
s
lik
e
L
STM
s
an
d
C
NNs
r
eq
u
ir
e
s
ig
n
if
ican
t
co
m
p
u
tatio
n
al
r
eso
u
r
ce
s
.
Ho
wev
e
r
,
o
u
r
f
u
s
io
n
-
b
ased
a
p
p
r
o
ac
h
o
f
f
er
s
a
g
o
o
d
b
alan
ce
b
et
wee
n
ac
cu
r
ac
y
an
d
ef
f
icien
cy
,
m
ak
in
g
it
a
p
r
ac
tical
ch
o
ice
f
o
r
r
ea
l
-
tim
e
cy
b
er
s
ec
u
r
ity
a
p
p
licatio
n
s
.
Ke
y
b
en
ef
its
o
f
o
u
r
ap
p
r
o
ac
h
i
n
clu
d
e:
−
Scalab
ilit
y
: By
in
teg
r
atin
g
m
u
ltip
le
d
atasets
,
it e
n
ab
les b
r
o
ad
er
attac
k
d
etec
tio
n
.
−
R
ed
u
ce
d
p
r
o
ce
s
s
in
g
o
v
er
h
ea
d
:
T
ec
h
n
iq
u
es
lik
e
f
ea
tu
r
e
-
le
v
el
f
u
s
io
n
an
d
d
im
en
s
io
n
alit
y
r
e
d
u
ctio
n
h
elp
m
in
im
ize
m
em
o
r
y
an
d
p
r
o
ce
s
s
in
g
p
o
wer
r
e
q
u
ir
em
e
n
ts
.
−
E
n
h
an
ce
d
a
d
ap
tab
ilit
y
: T
h
e
f
u
s
ed
m
o
d
el
ca
n
q
u
ick
ly
ad
ju
s
t t
o
ev
o
lv
i
n
g
cy
b
er
th
r
ea
ts
.
Ov
er
all,
o
u
r
m
eth
o
d
im
p
r
o
v
es
u
p
o
n
p
r
ev
io
u
s
m
o
d
els
b
y
co
m
b
in
in
g
f
ea
t
u
r
e
f
u
s
io
n
,
d
im
e
n
s
io
n
ality
r
ed
u
ctio
n
,
an
d
n
e
u
r
al
n
etwo
r
k
p
r
e
d
ictio
n
f
u
s
io
n
.
T
ab
le
1
.
R
esu
lts
o
f
class
if
icati
o
n
m
o
d
els
S
l
.
N
o
C
l
a
s
si
f
i
c
a
t
i
o
n
m
o
d
e
l
C
l
a
s
si
f
i
c
a
t
i
o
n
a
c
c
u
r
a
c
y
1
ANN
-
1
0
h
i
d
d
e
n
l
a
y
e
r
s
9
5
.
7
4
2
ANN2
-
1
2
h
i
d
d
e
n
l
a
y
e
r
s
8
0
.
4
9
3
ANN
-
2
h
i
d
d
e
n
l
a
y
e
r
w
i
t
h
LD
R
d
a
t
a
9
7
.
0
1
4
ANN
-
3
h
i
d
d
e
n
l
a
y
e
r
w
i
t
h
LD
R
d
a
t
a
9
7
.
0
6
5
Pr
o
p
o
sed
f
u
sed
mo
d
e
l
9
7
.
4
9
Fig
u
r
e
4
.
C
o
m
p
a
r
is
o
n
o
f
class
if
icatio
n
m
o
d
els
T
ab
le
2
.
C
o
m
p
a
r
is
o
n
with
ex
i
s
tin
g
m
eth
o
d
s
M
e
t
h
o
d
D
a
t
a
set
s
u
se
d
/
f
u
se
d
A
c
c
u
r
a
c
y
(
%)
F
a
l
se
p
o
si
t
i
v
e
s (%)
C
o
m
p
u
t
a
t
i
o
n
a
l
c
o
s
t
S
V
M
S
i
n
g
l
e
9
1
.
2
5
.
8
M
e
d
i
u
m
C
N
N
S
i
n
g
l
e
9
3
.
4
4
.
5
H
i
g
h
R
N
N
S
i
n
g
l
e
9
4
.
1
4
.
2
V
e
r
y
H
i
g
h
En
se
mb
l
e
S
i
n
g
l
e
9
5
.
0
5
.
0
V
e
r
y
H
i
g
h
P
r
o
p
o
se
d
M
o
d
e
l
M
u
l
t
i
p
l
e
/
F
u
s
e
d
9
7
.
5
2
.
8
M
e
d
i
u
m
6.
CO
M
P
ARI
SO
N
O
F
P
RO
P
O
SE
D
WO
RK
W
I
T
H
O
T
H
E
R
RE
SE
ARCH
WO
RK
T
h
e
p
r
o
p
o
s
ed
a
p
p
r
o
ac
h
p
er
f
o
r
m
ed
b
etter
t
h
an
o
th
er
r
esear
c
h
wo
r
k
s
o
n
th
e
NSL
-
KDD
an
d
UNSW
-
NB
1
5
d
atasets
.
Pre
v
io
u
s
wo
r
k
s
ac
h
iev
ed
class
if
icatio
n
ac
cu
r
ac
ies r
an
g
in
g
f
r
o
m
9
0
% to
9
5
% o
n
th
ese
d
atasets
[
1
4
]
.
Prio
r
s
tu
d
ies
h
av
e
ac
h
i
ev
ed
p
r
o
m
is
in
g
r
esu
lts
with
m
o
d
els
em
p
lo
y
in
g
tr
ad
itio
n
al
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
s
s
u
ch
as
en
s
em
b
l
e
m
o
d
els
[
1
5
]
,
SVM,
an
d
R
Fs
,
b
u
t
f
r
eq
u
e
n
tly
u
n
d
e
r
p
e
r
f
o
r
m
e
d
d
u
e
to
th
e
co
m
p
lex
ity
a
n
d
d
iv
er
s
ity
o
f
n
etwo
r
k
d
ata
[
1
6
]
.
I
n
[
1
7
]
,
a
f
ea
tu
r
e
s
elec
tio
n
m
et
h
o
d
f
o
r
n
etwo
r
k
i
n
tr
u
s
io
n
d
etec
tio
n
ac
h
iev
e
d
8
8
t
o
9
1
%
ac
cu
r
ac
y
.
T
h
e
r
esear
ch
er
s
s
h
o
wed
th
at
in
s
o
m
e
ca
s
es
th
e
d
ee
p
lear
n
in
g
ap
p
r
o
ac
h
es,
C
NN
an
d
R
NN
ac
h
iev
ed
ac
c
u
r
ac
y
9
2
%
to
9
6
%
[
5
]
,
[
1
8
]
.
B
u
t
th
ese
ap
p
r
o
a
ch
es
r
eq
u
ir
e
d
m
o
r
e
co
m
p
u
tatio
n
tim
e
an
d
s
p
ac
e
p
ar
ticu
lar
ly
d
u
r
in
g
th
e
im
p
le
m
en
tatio
n
o
f
m
u
ltis
o
u
r
ce
d
ata
[
1
9
]
–
[
2
1
]
.
I
n
th
e
r
ec
en
t
ap
p
r
o
ac
h
,
I
DS
tak
es
o
n
B
ig
d
ata
b
y
ap
p
l
y
in
g
ML
m
e
th
o
d
s
[
2
2
]
.
So
m
e
wo
r
k
s
h
o
we
d
th
at
f
ea
tu
r
e
-
le
v
el
f
u
s
io
n
tech
n
i
q
u
es
ca
n
i
n
teg
r
at
e
co
m
p
lem
en
ta
r
y
in
f
o
r
m
atio
n
f
r
o
m
ea
ch
d
ataset
[
2
3
]
,
an
d
t
h
ey
also
im
p
r
o
v
e
d
th
e
d
etec
tio
n
r
ate
o
f
b
o
th
k
n
o
wn
an
d
u
n
k
n
o
wn
attac
k
s
.
T
h
e
r
esear
ch
er
wo
r
k
ed
o
n
MA
NE
T
b
y
u
s
in
g
KDD
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
15
,
No
.
4
,
Au
g
u
s
t
20
25
:
3
9
3
8
-
3948
3946
d
ata
s
ets
wh
ich
s
h
o
wed
an
ac
c
u
r
ac
y
o
f
ab
o
u
t
7
5
%
[
2
4
]
.
B
y
c
o
n
s
id
er
in
g
all
th
e
r
ec
en
t
ap
p
r
o
ac
h
es
,
in
o
u
r
wo
r
k
we
u
s
ed
f
ea
tu
r
e
-
lev
el
f
u
s
io
n
f
o
llo
wed
b
y
d
im
en
s
io
n
ality
r
ed
u
ctio
n
a
n
d
ad
d
itio
n
ally
,
p
r
ed
ictio
n
-
lev
el
f
u
s
io
n
en
h
an
ce
s
th
e
p
e
r
f
o
r
m
an
ce
o
f
n
eu
r
al
n
etwo
r
k
m
o
d
els
b
y
lev
er
ag
in
g
th
eir
s
tr
en
g
th
s
.
T
h
e
e
x
p
er
im
en
tal
r
esu
lts
s
h
o
w
th
at
n
etwo
r
k
in
tr
u
s
io
n
d
etec
tio
n
is
a
v
alu
ab
le
a
p
p
licatio
n
in
c
y
b
er
s
ec
u
r
ity
,
p
r
o
v
id
in
g
ef
f
ec
tiv
e
s
o
lu
tio
n
s
an
d
in
s
ig
h
ts
f
o
r
f
u
tu
r
e
im
p
r
o
v
em
en
ts
.
T
h
is
ap
p
r
o
ac
h
s
er
v
es
as
a
s
tr
o
n
g
f
o
u
n
d
atio
n
f
o
r
f
u
r
th
er
r
esear
ch
,
a
s
m
u
ltil
ev
el
an
d
m
u
ltis
o
u
r
ce
m
eth
o
d
s
g
en
er
ally
o
u
t
p
er
f
o
r
m
tr
a
d
itio
n
al
s
in
g
le
-
s
o
u
r
ce
tech
n
iq
u
es
i
n
cy
b
er
s
ec
u
r
ity
.
T
h
e
p
r
o
p
o
s
ed
m
eth
o
d
ac
h
iev
e
d
a
class
if
icati
o
n
ac
c
u
r
ac
y
o
f
9
7
.
5
%.
T
h
er
e
i
s
n
o
m
o
d
el
wh
ich
is
test
ed
o
n
b
o
th
ty
p
e
o
f
d
atasets
[
2
5
]
.
C
o
m
p
ar
is
o
n
with
p
r
ev
io
u
s
wo
r
k
:
C
o
n
v
en
tio
n
al
ML
m
o
d
els
(
SVM
an
d
RF
)
g
av
e
9
0
-
9
5
%
ac
cu
r
ac
y
with
an
in
cr
ea
s
ed
f
alse
p
o
s
itiv
e
r
atio
.
Sev
er
al
d
ee
p
lear
n
i
n
g
-
b
ased
I
DS
m
o
d
els,
s
u
ch
as
C
NN
an
d
R
NN
ac
h
iev
ed
9
2
%
to
9
6
%
ac
cu
r
ac
y
b
u
t
n
ee
d
ed
s
u
b
s
tan
tial
co
m
p
u
tatio
n
al
r
eso
u
r
ce
s
.
T
h
e
p
r
o
p
o
s
ed
m
u
ltil
ev
el
d
at
a
f
u
s
io
n
m
o
d
el
ta
k
es
ad
v
a
n
tag
e
o
f
m
u
ltip
le
f
ea
tu
r
es
an
d
ac
h
iev
es
9
7
.
5
%
ac
cu
r
ac
y
,
wh
ic
h
is
b
etter
th
an
th
e
p
r
ev
io
u
s
m
o
d
els.
T
h
e
d
atasets
NSL
-
KDD
&
UNS
W
-
NB
1
5
f
o
r
attac
k
d
etec
tio
n
s
h
o
we
d
v
ar
io
u
s
f
o
r
m
s
o
f
attac
k
s
,
wh
ich
ar
e
in
te
g
r
ated
with
m
u
ltip
le
th
r
ea
t
ty
p
es
th
r
o
u
g
h
f
ea
tu
r
e
-
lev
el
f
u
s
io
n
.
Dim
en
s
io
n
ality
r
ed
u
ctio
n
tech
n
iq
u
e
L
DA
r
ed
u
ce
s
co
m
p
u
tatio
n
al
co
m
p
lex
ity
b
y
f
o
cu
s
in
g
o
n
r
ele
v
an
t
f
ea
tu
r
es.
At
p
r
ed
ictio
n
-
lev
e
l
f
u
s
io
n
,
th
e
f
u
s
io
n
o
f
two
n
e
u
r
a
l n
ets in
cr
ea
s
es th
e
r
o
b
u
s
tn
ess
an
d
ac
cu
r
ac
y
o
f
class
if
icatio
n
.
E
v
en
with
th
e
ef
f
ec
tiv
e
o
u
tco
m
e,
th
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
f
a
ce
s
k
ey
ch
allen
g
es.
I
ts
r
ec
u
r
s
iv
e
f
u
s
io
n
s
tr
u
ctu
r
e
in
cr
ea
s
es
d
ata
d
im
en
s
io
n
ality
p
r
o
p
o
r
tio
n
ally
a
n
d
r
eq
u
ir
e
p
o
wer
f
u
l
L
DA
f
o
r
d
ata
r
ed
u
ctio
n
,
s
in
ce
i
n
r
ea
l
-
tim
e
d
ep
l
o
y
m
en
t
L
DA
d
em
an
d
s
h
ig
h
p
r
o
ce
s
s
in
g
p
o
w
er
.
T
h
e
m
o
d
el
test
ed
o
n
NSL
-
K
DD
an
d
UNSW
-
NB
1
5
d
atasets
n
ee
d
s
f
u
r
t
h
er
v
alid
atio
n
f
o
r
r
ea
l
-
w
o
r
ld
n
etwo
r
k
s
,
ex
am
p
le
en
ter
p
r
is
e
an
d
cl
o
u
d
e
n
v
ir
o
n
m
en
ts
.
Ad
d
itio
n
ally
,
it
r
em
ain
s
v
u
ln
e
r
ab
le
to
ad
v
er
s
ar
ial
attac
k
s
s
u
ch
as
d
ata
p
o
is
o
n
in
g
an
d
e
v
asio
n
.
T
o
im
p
r
o
v
e
it
s
ap
p
licab
ilit
y
,
th
e
f
u
tu
r
e
r
esear
ch
m
u
s
t
f
o
cu
s
o
n
test
in
g
th
e
m
o
d
el
with
liv
e
n
etwo
r
k
tr
af
f
ic
in
r
ea
l
-
tim
e
c
y
b
e
r
s
ec
u
r
ity
en
v
ir
o
n
m
en
t.
I
m
p
lem
en
t
alg
o
r
ith
m
s
th
at
lear
n
an
d
r
ef
in
e
th
em
s
elv
es
as
n
ew
p
atter
n
s
o
f
m
alicio
u
s
ac
tiv
ity
em
er
g
e.
I
m
p
r
o
v
e
p
r
ed
ictio
n
-
lev
el
f
u
s
io
n
-
i
n
teg
r
ate
m
u
ltip
l
e
AI
m
o
d
els
s
u
ch
as
g
r
a
p
h
n
eu
r
al
n
etwo
r
k
s
(
GNNs)
o
r
L
STM
n
etwo
r
k
s
f
o
r
s
eq
u
e
n
tially
attac
k
in
g
an
al
y
s
is
.
B
y
ad
d
r
ess
in
g
th
ese
a
r
e
as,
m
u
ltis
o
u
r
ce
d
ata
f
u
s
io
n
-
b
ased
NI
DS
ca
n
ev
o
lv
e
in
to
a
m
o
r
e
r
o
b
u
s
t,
ad
ap
tiv
e,
an
d
s
ca
lab
le
s
o
lu
tio
n
f
o
r
m
o
d
er
n
cy
b
er
s
ec
u
r
it
y
ch
a
llen
g
es.
7.
CO
NCLU
SI
O
N
T
h
is
s
tu
d
y
d
em
o
n
s
tr
ates
th
e
ef
f
ec
tiv
en
ess
o
f
a
u
s
in
g
m
u
ltil
ev
el
an
d
m
u
ltis
o
u
r
ce
d
ata
f
u
s
io
n
ap
p
r
o
ac
h
in
en
h
a
n
cin
g
n
etwo
r
k
in
tr
u
s
i
o
n
d
etec
tio
n
s
y
s
tem
s
.
T
h
e
p
r
o
p
o
s
ed
m
et
h
o
d
o
lo
g
y
co
m
b
in
es
f
ea
tu
r
es
f
r
o
m
th
e
NSL
-
KDD
an
d
UNS
W
-
NB
1
5
d
atasets
an
d
r
ed
u
ce
s
th
e
d
im
en
s
io
n
ality
o
f
th
e
d
ata
s
ets
th
r
o
u
g
h
lin
ea
r
d
is
cr
im
in
an
t
an
aly
s
is
tech
n
iq
u
e,
an
d
f
u
r
th
e
r
p
r
ed
ictio
n
-
lev
el
f
u
s
io
n
o
f
two
n
eu
r
al
n
etwo
r
k
m
o
d
els
is
ap
p
lie
d
to
ac
h
iev
e
a
class
if
icat
io
n
ac
cu
r
ac
y
o
f
9
7
.
5
%.
T
h
e
s
u
cc
ess
o
f
th
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
is
d
u
e
to
th
e
ex
ten
s
iv
e
r
ep
r
esen
tatio
n
o
f
n
etwo
r
k
ac
ti
v
ities
th
r
o
u
g
h
f
ea
tu
r
e
-
lev
el
f
u
s
io
n
in
t
h
e
f
i
r
s
t
s
tag
e,
th
e
ef
f
ic
ien
cy
o
b
tain
ed
f
r
o
m
d
im
en
s
io
n
ality
r
ed
u
ctio
n
in
t
h
e
s
ec
o
n
d
s
tag
e,
an
d
f
u
r
th
er
im
p
r
o
v
e
d
p
er
f
o
r
m
an
ce
r
esu
ltin
g
f
r
o
m
p
r
ed
ictio
n
-
lev
el
f
u
s
io
n
at
th
e
last
s
tag
e.
T
h
e
r
esu
lts
clea
r
ly
s
h
o
w
th
e
a
d
v
an
tag
es
o
f
a
p
p
ly
in
g
d
ata
f
u
s
io
n
tech
n
iq
u
es
an
d
h
ig
h
lig
h
t
th
eir
p
o
ten
tial
to
im
p
r
o
v
e
cy
b
er
s
ec
u
r
ity
s
ch
em
es
ag
ain
s
t
k
n
o
wn
a
n
d
u
n
k
n
o
wn
t
y
p
es
o
f
th
r
ea
ts
.
T
h
e
p
r
o
p
o
s
ed
m
u
ltil
ev
el
m
u
ltis
o
u
r
ce
f
u
s
io
n
ap
p
r
o
ac
h
h
o
ld
s
p
r
o
m
is
e
f
o
r
f
u
tu
r
e
r
esear
ch
an
d
d
ev
elo
p
m
en
t in
NI
DS.
T
h
is
r
esear
ch
p
r
esen
ts
a
n
o
v
e
l
d
ata
f
u
s
io
n
ap
p
r
o
ac
h
th
at
s
i
g
n
if
ican
tly
e
n
h
an
ce
s
n
etwo
r
k
in
tr
u
s
io
n
d
etec
tio
n
ac
cu
r
ac
y
.
Giv
en
its
c
o
n
tr
ib
u
ti
o
n
s
to
c
y
b
er
s
ec
u
r
ity
an
d
ar
tif
icial
in
tellig
en
ce
,
th
is
s
tu
d
y
i
s
h
ig
h
ly
r
elev
an
t
t
o
th
e
s
co
p
e
o
f
t
h
e
co
m
p
u
ter
s
cien
ce
f
ield
an
d
ca
n
s
er
v
e
as
a
f
o
u
n
d
atio
n
f
o
r
f
u
tu
r
e
r
esear
c
h
in
in
tellig
en
t
th
r
ea
t
d
etec
tio
n
.
F
UNDING
I
NF
O
R
M
A
T
I
O
N
Au
th
o
r
s
s
tate
n
o
f
u
n
d
in
g
in
v
o
lv
ed
.
AUTHO
R
CO
NT
RI
B
UT
I
O
NS ST
A
T
E
M
E
N
T
T
h
is
jo
u
r
n
al
u
s
es
th
e
C
o
n
tr
ib
u
to
r
R
o
les
T
ax
o
n
o
m
y
(
C
R
ed
iT)
to
r
ec
o
g
n
ize
in
d
iv
id
u
al
au
th
o
r
co
n
tr
ib
u
tio
n
s
,
r
ed
u
ce
au
th
o
r
s
h
ip
d
is
p
u
tes,
an
d
f
ac
ilit
ate
co
llab
o
r
atio
n
.
Na
m
e
o
f
Aut
ho
r
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
Har
s
h
ith
a
So
m
ash
ek
ar
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
Pra
m
o
d
Hale
b
id
u
B
asav
ar
aju
✓
✓
✓
✓
✓
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
Mu
ltil
ev
el
a
n
d
mu
ltis
o
u
r
ce
d
a
ta
fu
s
io
n
a
p
p
r
o
a
c
h
fo
r
n
etw
o
r
k
in
tr
u
s
io
n
…
(
Ha
r
s
h
ith
a
S
o
ma
s
h
ek
a
r
)
3947
C
:
C
o
n
c
e
p
t
u
a
l
i
z
a
t
i
o
n
M
:
M
e
t
h
o
d
o
l
o
g
y
So
:
So
f
t
w
a
r
e
Va
:
Va
l
i
d
a
t
i
o
n
Fo
:
Fo
r
mal
a
n
a
l
y
s
i
s
I
:
I
n
v
e
s
t
i
g
a
t
i
o
n
R
:
R
e
so
u
r
c
e
s
D
:
D
a
t
a
C
u
r
a
t
i
o
n
O
:
W
r
i
t
i
n
g
-
O
r
i
g
i
n
a
l
D
r
a
f
t
E
:
W
r
i
t
i
n
g
-
R
e
v
i
e
w
&
E
d
i
t
i
n
g
Vi
:
Vi
su
a
l
i
z
a
t
i
o
n
Su
:
Su
p
e
r
v
i
s
i
o
n
P
:
P
r
o
j
e
c
t
a
d
mi
n
i
st
r
a
t
i
o
n
Fu
:
Fu
n
d
i
n
g
a
c
q
u
i
si
t
i
o
n
CO
NF
L
I
C
T
O
F
I
N
T
E
R
E
S
T
ST
A
T
E
M
E
NT
Au
th
o
r
s
s
tate
n
o
co
n
f
lict o
f
in
t
er
est.
RE
F
E
R
E
NC
E
S
[
1
]
T.
R
.
D
e
v
i
a
n
d
S
.
B
a
d
u
g
u
,
“
A
r
e
v
i
e
w
o
n
n
e
t
w
o
r
k
i
n
t
r
u
si
o
n
d
e
t
e
c
t
i
o
n
sy
s
t
e
m
u
si
n
g
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
,
”
i
n
A
d
v
a
n
c
e
s
i
n
D
e
c
i
s
i
o
n
S
c
i
e
n
c
e
s
,
I
m
a
g
e
Pr
o
c
e
ssi
n
g
,
S
e
c
u
r
i
t
y
a
n
d
C
o
m
p
u
t
e
r V
i
si
o
n
,
2
0
2
0
,
p
p
.
5
9
8
–
6
0
7
.
[
2
]
N
.
S
.
B
h
a
t
i
,
M
.
K
h
a
r
i
,
V
.
G
a
r
c
í
a
-
D
í
a
z
,
a
n
d
E.
V
e
r
d
ú
,
“
A
r
e
v
i
e
w
o
n
i
n
t
r
u
si
o
n
d
e
t
e
c
t
i
o
n
sy
s
t
e
m
s
a
n
d
t
e
c
h
n
i
q
u
e
s,”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
f
U
n
c
e
rt
a
i
n
t
y
,
F
u
zzi
n
e
ss
a
n
d
K
n
o
w
l
e
d
g
e
-
B
a
s
e
d
S
y
st
e
m
s
,
v
o
l
.
2
8
,
p
p
.
6
5
–
9
1
,
2
0
2
0
,
d
o
i
:
1
0
.
1
1
4
2
/
S
0
2
1
8
4
8
8
5
2
0
4
0
0
1
4
0
.
[
3
]
D
.
L.
H
a
l
l
a
n
d
J.
L
l
i
n
a
s
,
“
A
n
i
n
t
r
o
d
u
c
t
i
o
n
t
o
m
u
l
t
i
s
e
n
s
o
r
d
a
t
a
f
u
s
i
o
n
,
”
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
I
EEE
,
v
o
l
.
8
5
,
n
o
.
1
,
p
p
.
6
–
2
3
,
1
9
9
7
,
d
o
i
:
1
0
.
1
1
0
9
/
5
.
5
5
4
2
0
5
.
[
4
]
K
u
n
a
l
a
n
d
M
.
D
u
a
,
“
M
a
c
h
i
n
e
l
e
a
r
n
i
n
g
a
p
p
r
o
a
c
h
t
o
I
D
S
:
A
c
o
m
p
r
e
h
e
n
si
v
e
r
e
v
i
e
w
,
”
i
n
2
0
1
9
3
r
d
I
n
t
e
rn
a
t
i
o
n
a
l
c
o
n
f
e
r
e
n
c
e
o
n
El
e
c
t
r
o
n
i
c
s
,
C
o
m
m
u
n
i
c
a
t
i
o
n
a
n
d
Ae
r
o
sp
a
c
e
T
e
c
h
n
o
l
o
g
y
(
I
C
EC
A)
,
2
0
1
9
,
p
p
.
1
1
7
–
1
2
1
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
EC
A
.
2
0
1
9
.
8
8
2
2
1
2
0
.
[
5
]
L.
A
s
h
i
k
u
a
n
d
C
.
D
a
g
l
i
,
“
N
e
t
w
o
r
k
i
n
t
r
u
si
o
n
d
e
t
e
c
t
i
o
n
s
y
s
t
e
m
u
s
i
n
g
d
e
e
p
l
e
a
r
n
i
n
g
,
”
Pr
o
c
e
d
i
a
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
v
o
l
.
1
8
5
,
p
p
.
2
3
9
–
2
4
7
,
2
0
2
1
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
p
r
o
c
s
.
2
0
2
1
.
0
5
.
0
2
5
.
[
6
]
B
.
P
.
L.
La
u
e
t
a
l
.
,
“
A
s
u
r
v
e
y
o
f
d
a
t
a
f
u
s
i
o
n
i
n
smar
t
c
i
t
y
a
p
p
l
i
c
a
t
i
o
n
s
,
”
I
n
f
o
rm
a
t
i
o
n
F
u
s
i
o
n
,
v
o
l
.
5
2
,
n
o
.
Ja
n
u
a
r
y
,
p
p
.
3
5
7
–
3
7
4
,
2
0
1
9
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
i
n
f
f
u
s.
2
0
1
9
.
0
5
.
0
0
4
.
[
7
]
S
.
M
o
h
a
mm
a
d
i
,
H
.
M
i
r
v
a
z
i
r
i
,
M
.
G
h
a
z
i
z
a
d
e
h
-
A
h
sa
e
e
,
a
n
d
H
.
K
a
r
i
m
i
p
o
u
r
,
“
C
y
b
e
r
i
n
t
r
u
s
i
o
n
d
e
t
e
c
t
i
o
n
b
y
c
o
m
b
i
n
e
d
f
e
a
t
u
r
e
sel
e
c
t
i
o
n
a
l
g
o
r
i
t
h
m,
”
J
o
u
r
n
a
l
o
f
I
n
f
o
r
m
a
t
i
o
n
S
e
c
u
r
i
t
y
a
n
d
Ap
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
4
4
,
p
p
.
8
0
–
8
8
,
2
0
1
9
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
j
i
sa.
2
0
1
8
.
1
1
.
0
0
7
.
[
8
]
R
.
A
h
ma
d
a
n
d
I
.
A
l
sma
d
i
,
“
D
a
t
a
f
u
si
o
n
a
n
d
n
e
t
w
o
r
k
i
n
t
r
u
s
i
o
n
d
e
t
e
c
t
i
o
n
s
y
st
e
ms,
”
C
l
u
s
t
e
r
C
o
m
p
u
t
i
n
g
,
v
o
l
.
2
7
,
n
o
.
6
,
p
p
.
7
4
9
3
–
7
5
1
9
,
S
e
p
.
2
0
2
4
,
d
o
i
:
1
0
.
1
0
0
7
/
s1
0
5
8
6
-
0
2
4
-
0
4
3
6
5
-
y.
[
9
]
I
.
F
.
K
i
l
i
n
c
e
r
,
F
.
Er
t
a
m,
a
n
d
A
.
S
e
n
g
u
r
,
“
M
a
c
h
i
n
e
l
e
a
r
n
i
n
g
m
e
t
h
o
d
s
f
o
r
c
y
b
e
r
se
c
u
r
i
t
y
i
n
t
r
u
s
i
o
n
d
e
t
e
c
t
i
o
n
:
D
a
t
a
s
e
t
s
a
n
d
c
o
m
p
a
r
a
t
i
v
e
st
u
d
y
,
”
C
o
m
p
u
t
e
r
N
e
t
w
o
rks
,
v
o
l
.
1
8
8
,
n
o
.
J
a
n
u
a
r
y
,
p
.
1
0
7
8
4
0
,
2
0
2
1
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
c
o
mn
e
t
.
2
0
2
1
.
1
0
7
8
4
0
.
[
1
0
]
F
.
La
g
h
r
i
ss
i
,
S
.
D
o
u
z
i
,
K
.
D
o
u
z
i
,
a
n
d
B
.
H
ss
i
n
a
,
“
I
n
t
r
u
s
i
o
n
d
e
t
e
c
t
i
o
n
sy
st
e
ms
u
s
i
n
g
l
o
n
g
sh
o
r
t
-
t
e
r
m
m
e
m
o
r
y
(
L
S
TM
)
,
”
J
o
u
rn
a
l
o
f
Bi
g
D
a
t
a
,
v
o
l
.
8
,
n
o
.
1
,
D
e
c
.
2
0
2
1
,
d
o
i
:
1
0
.
1
1
8
6
/
s4
0
5
3
7
-
021
-
0
0
4
4
8
-
4.
[
1
1
]
A
.
K
h
r
a
i
sa
t
a
n
d
A
.
A
l
a
z
a
b
,
“
A
c
r
i
t
i
c
a
l
r
e
v
i
e
w
o
f
i
n
t
r
u
s
i
o
n
d
e
t
e
c
t
i
o
n
s
y
st
e
ms
i
n
t
h
e
i
n
t
e
r
n
e
t
o
f
t
h
i
n
g
s
:
t
e
c
h
n
i
q
u
e
s,
d
e
p
l
o
y
m
e
n
t
st
r
a
t
e
g
y
,
v
a
l
i
d
a
t
i
o
n
st
r
a
t
e
g
y
,
a
t
t
a
c
k
s
,
p
u
b
l
i
c
d
a
t
a
s
e
t
s a
n
d
c
h
a
l
l
e
n
g
e
s
,
”
C
y
b
e
rsec
u
ri
t
y
,
v
o
l
.
4
,
n
o
.
1
,
2
0
2
1
,
d
o
i
:
1
0
.
1
1
8
6
/
s
4
2
4
0
0
-
0
2
1
-
0
0
0
7
7
-
7.
[
1
2
]
S
.
S
u
h
a
n
a
,
S
.
K
a
r
t
h
i
c
,
a
n
d
N
.
Y
u
v
a
r
a
j
,
“
En
sem
b
l
e
b
a
se
d
d
i
me
n
si
o
n
a
l
i
t
y
r
e
d
u
c
t
i
o
n
f
o
r
i
n
t
r
u
s
i
o
n
d
e
t
e
c
t
i
o
n
u
si
n
g
r
a
n
d
o
m
f
o
r
e
s
t
i
n
w
i
r
e
l
e
s
s
n
e
t
w
o
r
k
s,”
i
n
2
0
2
3
5
t
h
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
S
m
a
r
t
S
y
s
t
e
m
s
a
n
d
I
n
v
e
n
t
i
v
e
T
e
c
h
n
o
l
o
g
y
(
I
C
S
S
I
T
)
,
Ja
n
.
2
0
2
3
,
p
p
.
7
0
4
–
7
0
8
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
S
S
I
T5
5
8
1
4
.
2
0
2
3
.
1
0
0
6
0
9
2
9
.
[
1
3
]
H
.
S
o
mas
h
e
k
a
r
a
n
d
R
.
B
o
r
a
i
a
h
,
“
N
e
t
w
o
r
k
i
n
t
r
u
s
i
o
n
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
u
si
n
g
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
p
r
e
d
i
c
t
i
o
n
s
f
u
si
o
n
,
”
I
n
d
o
n
e
si
a
n
J
o
u
r
n
a
l
o
f
El
e
c
t
r
i
c
a
l
E
n
g
i
n
e
e
r
i
n
g
a
n
d
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
v
o
l
.
3
1
,
n
o
.
2
,
p
p
.
1
1
4
7
–
1
1
5
3
,
A
u
g
.
2
0
2
3
,
d
o
i
:
1
0
.
1
1
5
9
1
/
i
j
e
e
c
s.
v
3
1
.
i
2
.
p
p
1
1
4
7
-
1
1
5
3
.
[
1
4
]
S
.
B
h
o
sal
e
a
n
d
S
.
M
.
K
a
ma
l
a
p
u
r
k
a
r
,
“
A
n
o
v
e
l
a
p
p
r
o
a
c
h
f
o
r
n
e
t
w
o
r
k
i
n
t
r
u
s
i
o
n
d
e
t
e
c
t
i
o
n
sy
s
t
e
m
u
si
n
g
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
a
l
g
o
r
i
t
h
ms,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
Ad
v
a
n
c
e
d
R
e
se
a
rc
h
i
n
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
a
n
d
S
o
f
t
w
a
re
En
g
i
n
e
e
ri
n
g
,
v
o
l
.
8
,
n
o
.
4
,
p
p
.
1
–
6
,
2
0
1
8
.
[
1
5
]
M
.
M
.
O
t
o
o
m,
K
.
N
.
A
.
S
a
t
t
a
r
,
a
n
d
M
.
A
l
S
a
d
i
g
,
“
En
sem
b
l
e
mo
d
e
l
f
o
r
n
e
t
w
o
r
k
i
n
t
r
u
s
i
o
n
d
e
t
e
c
t
i
o
n
sy
s
t
e
m b
a
se
d
o
n
b
a
g
g
i
n
g
u
s
i
n
g
J4
8
,
”
Ad
v
a
n
c
e
s
i
n
S
c
i
e
n
c
e
a
n
d
T
e
c
h
n
o
l
o
g
y
R
e
se
a
rc
h
J
o
u
r
n
a
l
,
v
o
l
.
1
7
,
n
o
.
2
,
p
p
.
3
2
2
–
3
2
9
,
2
0
2
3
,
d
o
i
:
1
0
.
1
2
9
1
3
/
2
2
9
9
8
6
2
4
/
1
6
1
8
2
0
.
[
1
6
]
Y
.
C
h
a
n
g
,
W
.
Li
,
a
n
d
Z.
Y
a
n
g
,
“
N
e
t
w
o
r
k
i
n
t
r
u
s
i
o
n
d
e
t
e
c
t
i
o
n
b
a
s
e
d
o
n
r
a
n
d
o
m
f
o
r
e
s
t
a
n
d
s
u
p
p
o
r
t
v
e
c
t
o
r
ma
c
h
i
n
e
,
”
i
n
2
0
1
7
I
EEE
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
C
o
m
p
u
t
a
t
i
o
n
a
l
S
c
i
e
n
c
e
a
n
d
E
n
g
i
n
e
e
r
i
n
g
(
C
S
E)
a
n
d
I
E
EE
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
Em
b
e
d
d
e
d
a
n
d
U
b
i
q
u
i
t
o
u
s
C
o
m
p
u
t
i
n
g
(
EU
C
)
,
2
0
1
7
,
v
o
l
.
1
,
p
p
.
6
3
5
–
6
3
8
,
d
o
i
:
1
0
.
1
1
0
9
/
C
S
E
-
EU
C
.
2
0
1
7
.
1
1
8
.
[
1
7
]
S
.
M
.
K
a
so
n
g
o
a
n
d
Y
.
S
u
n
,
“
P
e
r
f
o
r
m
a
n
c
e
a
n
a
l
y
si
s o
f
i
n
t
r
u
si
o
n
d
e
t
e
c
t
i
o
n
sy
st
e
ms u
si
n
g
a
f
e
a
t
u
r
e
s
e
l
e
c
t
i
o
n
met
h
o
d
o
n
t
h
e
U
N
S
W
-
N
B
1
5
d
a
t
a
se
t
,
”
J
o
u
rn
a
l
o
f
Bi
g
D
a
t
a
,
v
o
l
.
7
,
n
o
.
1
,
2
0
2
0
,
d
o
i
:
1
0
.
1
1
8
6
/
s
4
0
5
3
7
-
0
2
0
-
0
0
3
7
9
-
6.
[
1
8
]
I
.
I
d
r
i
ssi
,
M
.
B
o
u
k
a
b
o
u
s,
M
.
A
z
i
z
i
,
O
.
M
o
u
ss
a
o
u
i
,
a
n
d
H
.
El
F
a
d
i
l
i
,
“
T
o
w
a
r
d
a
d
e
e
p
l
e
a
r
n
i
n
g
-
b
a
se
d
i
n
t
r
u
si
o
n
d
e
t
e
c
t
i
o
n
s
y
st
e
m
f
o
r
I
o
T
a
g
a
i
n
st
b
o
t
n
e
t
a
t
t
a
c
k
s
,
”
I
AE
S
I
n
t
e
rn
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
Art
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
(
I
J
-
AI
)
,
v
o
l
.
1
0
,
n
o
.
1
,
p
p
.
1
1
0
–
1
2
0
,
M
a
r
.
2
0
2
1
,
d
o
i
:
1
0
.
1
1
5
9
1
/
i
j
a
i
.
v
1
0
.
i
1
.
p
p
1
1
0
-
1
2
0
.
[
1
9
]
J.
K
i
m
a
n
d
A
.
B
e
n
t
o
n
,
“
E
v
a
l
u
a
t
i
n
g
d
e
e
p
l
e
a
r
n
i
n
g
a
p
p
r
o
a
c
h
e
s
t
o
i
n
t
r
u
si
o
n
d
e
t
e
c
t
i
o
n
i
n
w
i
r
e
l
e
ss
n
e
t
w
o
r
k
s,
”
i
n
P
ro
c
e
e
d
i
n
g
s
o
f
t
h
e
2
0
1
6
I
EE
E
C
o
n
f
e
re
n
c
e
o
n
N
e
t
w
o
r
k
F
u
n
c
t
i
o
n
Vi
r
t
u
a
l
i
z
a
t
i
o
n
a
n
d
S
o
f
t
w
a
r
e
D
e
f
i
n
e
d
N
e
t
w
o
rk
s,
P
a
l
o
Al
t
o
,
C
A,
U
S
A
,
2
0
1
6
,
p
p
.
1
–
6.
[
2
0
]
H
.
H
.
Ja
z
i
,
H
.
G
o
n
z
a
l
e
z
,
N
.
S
t
a
k
h
a
n
o
v
a
,
a
n
d
A
.
A
.
G
h
o
r
b
a
n
i
,
“
D
e
t
e
c
t
i
n
g
H
T
TP
-
b
a
se
d
a
p
p
l
i
c
a
t
i
o
n
l
a
y
e
r
D
o
S
a
t
t
a
c
k
s
o
n
w
e
b
serv
e
r
s
i
n
t
h
e
p
r
e
se
n
c
e
o
f
sam
p
l
i
n
g
,
”
C
o
m
p
u
t
e
r
N
e
t
w
o
r
k
s
,
v
o
l
.
1
2
1
,
p
p
.
2
5
–
3
6
,
J
u
l
.
2
0
1
7
,
d
o
i
:
1
0
.
1
0
1
6
/
J.C
O
M
N
E
T.
2
0
1
7
.
0
3
.
0
1
8
.
[
2
1
]
M
.
T
a
v
a
l
l
a
e
e
,
E
.
B
a
g
h
e
r
i
,
W
.
L
u
,
a
n
d
A
.
A
.
G
h
o
r
b
a
n
i
,
“
A
d
e
t
a
i
l
e
d
a
n
a
l
y
s
i
s
o
f
t
h
e
K
D
D
C
U
P
9
9
d
a
t
a
s
e
t
,
”
i
n
2
0
0
9
I
EEE
S
y
m
p
o
s
i
u
m
o
n
C
o
m
p
u
t
a
t
i
o
n
a
l
I
n
t
e
l
l
i
g
e
n
c
e
f
o
r
S
e
c
u
ri
t
y
a
n
d
D
e
f
e
n
se
Ap
p
l
i
c
a
t
i
o
n
s
,
J
u
l
.
2
0
0
9
,
p
p
.
1
–
6
,
d
o
i
:
1
0
.
1
1
0
9
/
C
I
S
D
A
.
2
0
0
9
.
5
3
5
6
5
2
8
.
[
2
2
]
M
.
A
.
A
l
a
k
e
t
u
e
t
a
l
.
,
“
C
o
m
p
a
r
a
t
i
v
e
a
n
a
l
y
si
s
o
f
i
n
t
r
u
si
o
n
d
e
t
e
c
t
i
o
n
m
o
d
e
l
s
u
si
n
g
b
i
g
d
a
t
a
a
n
a
l
y
t
i
c
s
a
n
d
mac
h
i
n
e
l
e
a
r
n
i
n
g
t
e
c
h
n
i
q
u
e
s,
”
T
h
e
I
n
t
e
rn
a
t
i
o
n
a
l
Ar
a
b
J
o
u
r
n
a
l
o
f
I
n
f
o
rm
a
t
i
o
n
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
2
1
,
n
o
.
2
,
2
0
2
4
,
d
o
i
:
1
0
.
3
4
0
2
8
/
i
a
j
i
t
/
2
1
/
2
/
1
4
.
[
2
3
]
A
.
A
y
a
n
t
a
y
o
e
t
a
l
.
,
“
N
e
t
w
o
r
k
i
n
t
r
u
si
o
n
d
e
t
e
c
t
i
o
n
u
si
n
g
f
e
a
t
u
r
e
f
u
si
o
n
w
i
t
h
d
e
e
p
l
e
a
r
n
i
n
g
,
”
J
o
u
r
n
a
l
o
f
Bi
g
D
a
t
a
,
v
o
l
.
1
0
,
n
o
.
1
,
D
e
c
.
2
0
2
3
,
d
o
i
:
1
0
.
1
1
8
6
/
s
4
0
5
3
7
-
0
2
3
-
0
0
8
3
4
-
0.
[
2
4
]
M
.
S
a
s
i
k
u
m
a
r
a
n
d
K
.
R
o
h
i
n
i
,
“
E
x
p
e
d
i
e
n
t
i
n
t
r
u
si
o
n
d
e
t
e
c
t
i
o
n
sy
s
t
e
m i
n
M
A
N
ET
u
si
n
g
r
o
b
u
s
t
d
r
a
g
o
n
f
l
y
-
o
p
t
i
mi
z
e
d
e
n
h
a
n
c
e
d
n
a
i
v
e
B
a
y
e
s
(
R
D
O
-
EN
B
)
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
f
C
o
m
p
u
t
e
r
N
e
t
w
o
rks
a
n
d
Ap
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
1
1
,
n
o
.
1
,
p
p
.
4
6
–
6
0
,
2
0
2
4
,
d
o
i
:
1
0
.
2
2
2
4
7
/
i
j
c
n
a
/
2
0
2
4
/
2
2
4
4
3
5
.
[
2
5
]
A
.
R
.
A
l
-
G
h
u
w
a
i
r
i
,
Y
.
S
h
a
r
r
a
b
,
D
.
Al
-
F
r
a
i
h
a
t
,
M
.
A
l
El
a
i
m
a
t
,
A
.
A
l
s
a
r
h
a
n
,
a
n
d
A
.
A
l
g
a
r
n
i
,
“
I
n
t
r
u
si
o
n
d
e
t
e
c
t
i
o
n
i
n
c
l
o
u
d
c
o
m
p
u
t
i
n
g
b
a
se
d
o
n
t
i
m
e
seri
e
s
a
n
o
mal
i
e
s
u
t
i
l
i
z
i
n
g
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
,
”
J
o
u
r
n
a
l
o
f
C
l
o
u
d
C
o
m
p
u
t
i
n
g
,
v
o
l
.
1
2
,
n
o
.
1
,
2
0
2
3
,
d
o
i
:
1
0
.
1
1
8
6
/
s
1
3
6
7
7
-
0
2
3
-
0
0
4
9
1
-
x.
Evaluation Warning : The document was created with Spire.PDF for Python.