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
.
3
,
J
u
n
e
20
25
,
p
p
.
3095
~
3
1
0
6
I
SS
N:
2088
-
8
7
0
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijece.
v
15
i
3
.
pp
3
0
9
5
-
3
1
0
6
3095
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ec
e.
ia
esco
r
e.
co
m
Exploring
t
h
e
e
ff
e
ctiveness
of m
ulti
cla
ss
decisio
n j
un
g
le f
o
r
interne
t
o
f
things
security
Sm
it
ha
Ra
j
a
g
o
pa
l
1
,
Abhi
k
Sa
rk
a
r
2
,
Venk
a
t
N
a
ra
y
a
na
n M
a
njuna
t
h
1
1
A
l
l
i
a
n
c
e
S
c
h
o
o
l
o
f
A
d
v
a
n
c
e
d
C
o
m
p
u
t
i
n
g
,
A
l
l
i
a
n
c
e
U
n
i
v
e
r
si
t
y
,
B
e
n
g
a
l
u
r
u
,
I
n
d
i
a
2
A
l
l
i
a
n
c
e
S
c
h
o
o
l
o
f
Li
b
e
r
a
l
A
r
t
s,
A
l
l
i
a
n
c
e
U
n
i
v
e
r
s
i
t
y
,
B
e
n
g
a
l
u
r
u
,
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
Au
g
2
5
,
2
0
2
4
R
ev
is
ed
Feb
4
,
2
0
2
5
Acc
ep
ted
Ma
r
3
,
2
0
2
5
Ne
two
rk
in
tr
u
sio
n
d
e
tec
ti
o
n
sy
ste
m
s
(NID
S
)
a
re
v
it
a
l
in
p
ro
tec
ti
n
g
c
o
m
p
u
ter
n
e
two
rk
s
a
g
a
i
n
st
c
y
b
e
r
se
c
u
rit
y
in
c
id
e
n
ts.
Th
e
re
latio
n
sh
i
p
b
e
tw
e
e
n
NID
S
a
n
d
in
tern
e
t
o
f
t
h
in
g
s
(Io
T)
se
c
u
r
it
y
is
p
i
v
o
tal
a
n
d
NID
S
p
lay
s
a
sig
n
ifi
c
a
n
t
ro
le
in
e
n
su
ri
n
g
th
e
se
c
u
rit
y
a
n
d
re
li
a
b
il
it
y
o
f
I
o
T
e
c
o
s
y
ste
m
s.
En
su
rin
g
th
e
se
c
u
rit
y
o
f
Io
T
d
e
v
ice
s
is
c
rit
ica
l
fo
r
se
v
e
ra
l
re
a
so
n
s.
It
h
e
lp
s
sa
fe
g
u
a
rd
se
n
siti
v
e
i
n
fo
rm
a
ti
o
n
,
g
u
a
ra
n
tee
s
th
e
d
e
p
e
n
d
a
b
il
i
ty
o
f
c
ru
c
ial
i
n
fr
a
stru
c
tu
re
,
m
e
e
ts
re
g
u
lato
ry
o
b
li
g
a
ti
o
n
s,
a
n
d
fo
ste
rs
u
se
r
c
o
n
fid
e
n
c
e
.
As
th
e
Io
T
e
c
o
sy
ste
m
e
x
p
a
n
d
s,
p
rio
ri
ti
z
in
g
se
c
u
rit
y
is
e
ss
e
n
ti
a
l
t
o
m
in
imiz
e
risk
s
a
n
d
m
a
x
imiz
e
th
e
b
e
n
e
fit
s o
f
c
o
n
n
e
c
ted
d
e
v
ice
s.
G
iv
e
n
t
h
e
e
v
e
r
-
e
x
p
a
n
d
in
g
c
y
b
e
r
th
re
a
t
lan
d
sc
a
p
e
,
t
h
e
m
u
lt
icla
ss
c
las
sifica
ti
o
n
tas
k
is
e
ss
e
n
ti
a
l
to
e
m
p
o
we
r
th
e
NID
S
with
a
n
a
b
il
it
y
to
d
ist
in
g
u
ish
b
e
twe
e
n
v
a
ri
o
u
s
a
tt
a
c
k
p
a
tt
e
rn
s
in
les
s
c
o
m
p
u
tatio
n
a
l
ti
m
e
.
Th
e
m
u
lt
icla
ss
d
e
c
isio
n
j
u
n
g
le
a
lg
o
rit
h
m
is
in
v
e
stig
a
te
d
to
o
p
ti
m
ize
th
e
p
e
rfo
rm
a
n
c
e
o
f
NID
S
.
Th
e
re
s
e
a
rc
h
h
a
s
c
o
n
sid
e
re
d
p
e
rm
u
tati
o
n
fe
a
tu
re
imp
o
rtan
c
e
t
o
in
c
l
u
d
e
o
n
ly
t
h
e
re
lev
a
n
t
fe
a
tu
re
s
fro
m
th
e
d
a
ta.
Us
in
g
a
c
o
n
tem
p
o
ra
ry
d
a
tas
e
t
su
c
h
a
s
CICIOT
2
0
2
3
,
th
e
stu
d
y
h
a
s
d
e
m
o
n
stra
ted
a
n
i
m
p
re
ss
iv
e
a
tt
a
c
k
d
e
tec
ti
o
n
ra
te
o
f
o
v
e
r
9
0
%
fo
r
2
0
m
o
d
e
rn
a
tt
a
c
k
t
y
p
e
s.
T
h
is
re
se
a
rc
h
h
a
s
in
v
e
stig
a
ted
th
e
e
ff
e
c
ti
v
e
n
e
ss
o
f
Io
T
se
c
u
rit
y
m
e
a
su
re
s
a
n
d
i
t
s
p
ro
s
p
e
c
ti
v
e
c
o
n
tri
b
u
ti
o
n
s
to
th
e
field
o
f
c
y
b
e
r
se
c
u
rit
y
.
K
ey
w
o
r
d
s
:
C
I
C
I
OT
2
0
2
3
Dec
is
io
n
ju
n
g
le
I
n
ter
n
et
o
f
th
in
g
s
s
ec
u
r
ity
Netwo
r
k
in
tr
u
s
io
n
d
etec
tio
n
Per
m
u
tatio
n
f
ea
tu
r
e
im
p
o
r
tan
c
e
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
:
Sm
ith
a
R
ajag
o
p
al
Allian
ce
Sch
o
o
l o
f
Ad
v
an
ce
d
C
o
m
p
u
tin
g
,
Allian
ce
Un
iv
e
r
s
ity
B
en
g
alu
r
u
,
I
n
d
ia
E
m
ail:
s
m
ith
a.
r
esear
ch
1
0
1
2
@
g
m
ail.
co
m
1.
I
NT
RO
D
UCT
I
O
N
T
h
e
n
etwo
r
k
lan
d
s
ca
p
e
is
h
ig
h
ly
v
u
ln
er
ab
le
to
s
ec
u
r
ity
th
r
ea
ts
.
T
h
e
co
n
s
tan
t
em
er
g
e
n
ce
o
f
n
ewe
r
th
r
ea
ts
is
p
ar
am
o
u
n
t
in
d
ev
elo
p
in
g
r
eliab
le
an
d
a
d
ap
tiv
e
s
ec
u
r
ity
m
ea
s
u
r
es.
T
h
e
m
eth
o
d
b
y
wh
ich
an
attac
k
er
g
ain
s
ac
ce
s
s
to
a
n
etwo
r
k
,
th
er
eb
y
s
en
d
i
n
g
m
alicio
u
s
p
ac
k
ets
to
a
u
s
er
’
s
s
y
s
tem
to
s
teal,
m
o
d
if
y
,
o
r
r
a
v
ag
e
co
n
f
id
en
tial
d
ata
is
ter
m
ed
in
tr
u
s
io
n
[
1
]
.
E
x
is
tin
g
s
y
s
tem
v
u
ln
er
ab
ilit
ies,
s
u
ch
as
h
u
m
an
e
r
r
o
r
s
,
m
is
co
n
f
ig
u
r
atio
n
,
o
r
s
o
f
twar
e
b
u
g
s
ca
n
lead
to
a
n
attac
k
o
n
th
e
s
er
v
er
o
r
d
e
v
ice.
As
a
n
d
wh
en
h
u
m
u
n
g
o
u
s
d
ata
is
b
ein
g
ex
c
h
an
g
e
d
o
n
t
h
e
I
n
ter
n
et,
n
etwo
r
k
s
b
ec
o
m
e
in
cr
ea
s
in
g
ly
s
u
s
ce
p
tib
le
to
attac
k
s
.
T
h
er
ef
o
r
e,
in
tr
u
s
io
n
d
etec
tio
n
s
y
s
tem
s
(
I
DS)
ar
e
r
eq
u
ir
e
d
to
s
af
eg
u
ar
d
o
r
g
an
izatio
n
al
n
etwo
r
k
s
.
T
h
e
q
u
a
n
tity
o
f
in
ter
n
et
o
f
th
i
n
g
s
(
I
o
T
)
co
n
n
ec
tio
n
s
h
as
im
p
r
o
v
ed
o
v
e
r
th
e
p
ast
d
ec
ad
e.
T
h
is
tr
en
d
is
an
ticip
ated
to
m
ai
n
tain
in
n
u
m
er
o
u
s
s
ec
to
r
s
in
t
h
e
co
m
i
n
g
y
ea
r
s
,
h
o
wev
er
d
em
an
d
in
g
s
i
tu
atio
n
s
co
n
tin
u
e
to
b
e.
I
t
o
u
g
h
t
to
b
e
f
ix
ed
t
o
m
ak
e
ce
r
tain
th
at
I
o
T
g
a
d
g
ets
o
p
er
ate
ap
p
r
o
p
r
iately
a
n
d
ef
f
ec
tiv
ely
[
2
]
.
I
m
p
lem
en
tin
g
s
ec
u
r
ity
f
ea
tu
r
es
u
n
d
er
I
o
T
is
m
o
r
e
c
o
m
p
licated
th
an
im
p
lem
en
tin
g
th
em
o
n
tr
ad
itio
n
al
n
etwo
r
k
s
b
ec
au
s
e
o
f
th
e
e
x
ten
s
iv
e
f
o
r
m
o
f
p
r
o
t
o
co
ls
s
u
ch
as
all
ap
p
licab
le
n
o
d
e
q
u
an
t
ities
[
3
]
.
T
h
r
o
u
g
h
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
.
3
,
J
u
n
e
20
25
:
3
0
9
5
-
3
1
0
6
3096
d
is
tr
ib
u
ted
d
en
ial
-
of
-
ca
r
r
ier
(
DDo
S)
ass
au
lt
s
,
tar
g
etin
g
I
o
T
g
ad
g
ets
s
u
ch
as
v
ir
tu
al
ca
m
er
as
an
d
d
ig
ital
v
id
eo
r
ec
o
r
d
er
s
(
DVRs
)
,
is
an
in
s
tan
ce
o
f
a
cy
b
er
h
az
ar
d
th
at
co
n
s
t
an
tly
h
ap
p
en
s
in
th
e
I
o
T
lan
d
s
ca
p
e.
I
n
ter
n
et
o
f
th
i
n
g
s
(
I
o
T
)
n
et
wo
r
k
s
h
av
e
u
n
i
q
u
e
c
h
ar
ac
ter
is
tics
th
at
s
et
th
em
ap
ar
t
f
r
o
m
wir
eless
s
en
s
o
r
n
etwo
r
k
s
(
W
SN)
an
d
cy
b
er
p
h
y
s
ical
s
y
s
tem
s
(
C
P
S
)
.
T
h
is
d
is
tin
ctio
n
ar
is
es
f
r
o
m
th
e
h
eter
o
g
en
o
u
s
lay
er
s
o
f
p
r
o
to
c
o
ls
em
p
lo
y
e
d
in
I
o
T
n
etwo
r
k
s
.
Ad
d
itio
n
ally
,
th
e
c
h
allen
g
es
en
co
u
n
ter
ed
d
u
r
in
g
th
e
d
ep
lo
y
m
e
n
t
o
f
v
ar
i
o
u
s
I
o
T
u
s
e
ca
s
es
v
ar
y
s
ig
n
if
ican
tly
f
r
o
m
th
o
s
e
ass
o
ciate
d
with
W
SN
n
etwo
r
k
s
,
d
u
e
to
th
e
s
p
ec
if
ic
co
n
tex
t
an
d
r
eq
u
ir
em
en
ts
o
f
ea
c
h
ap
p
licatio
n
[
4
]
,
[
5
]
.
T
o
en
h
an
ce
s
ec
u
r
ity
in
an
I
o
T
en
v
ir
o
n
m
en
t,
it
is
cr
u
cial
to
d
esig
n
a
d
ata
-
o
r
ien
ted
s
ec
u
r
ity
m
ec
h
an
is
m
.
T
h
is
m
ec
h
an
is
m
s
h
o
u
ld
p
r
io
r
itize
d
ata
co
n
f
id
en
tiality
an
d
in
teg
r
ity
to
m
itig
ate
th
r
ea
ts
.
T
r
ad
itio
n
al
cr
y
p
to
g
r
ap
h
y
-
b
ased
s
ec
u
r
ity
m
ea
s
u
r
es
m
ay
n
o
t
b
e
id
ea
l
f
o
r
I
o
T
d
u
e
to
th
e
v
ast am
o
u
n
t o
f
d
ata
in
v
o
lv
e
d
[
6
]
.
Fig
u
r
e
1
illu
s
tr
ates th
e
I
o
T
s
ec
u
r
ity
lan
d
s
ca
p
e.
Fig
u
r
e
1
.
I
o
T
s
ec
u
r
ity
la
n
d
s
ca
p
e
I
DS
ca
n
b
e
d
iv
i
d
ed
i
n
to
d
if
f
er
en
t
ty
p
es.
Fo
r
ex
a
m
p
le,
t
h
ey
c
an
b
e
g
r
o
u
p
e
d
as
h
o
s
t
-
b
ased
o
r
n
etwo
r
k
-
b
ased
.
Netwo
r
k
-
b
ased
s
y
s
tem
s
watc
h
o
v
er
n
etwo
r
k
tr
af
f
ic
a
n
d
d
ev
ices
f
o
r
an
y
u
n
u
s
u
al
ac
tiv
ities
,
wh
ile
h
o
s
t
-
b
ased
s
y
s
tem
s
ch
ec
k
in
d
iv
i
d
u
al
d
ev
ices
f
o
r
ch
an
g
es
in
f
i
les,
lo
g
s
,
o
r
b
e
h
av
io
r
s
th
at
d
o
n
’
t
m
atch
wh
at’
s
ex
p
ec
ted
[
7
]
,
[
8
]
.
T
y
p
ically
,
I
DS
ca
n
b
e
a
n
aly
ze
d
u
s
in
g
tw
o
m
ain
m
eth
o
d
s
:
s
ig
n
atu
r
e
-
b
a
s
ed
an
d
a
n
o
m
aly
-
b
ased
s
y
s
tem
s
.
Sig
n
atu
r
e
-
b
ased
I
DS,
lik
e
SID
S
an
d
AI
D
S,
d
etec
t
attac
k
s
b
y
lo
o
k
in
g
f
o
r
k
n
o
wn
p
atter
n
s
s
to
r
ed
in
a
d
ata
b
ase.
SID
S
s
p
ec
if
ically
s
ea
r
ch
es
f
o
r
t
h
ese
p
r
ed
e
f
in
ed
p
atter
n
s
,
wh
ile
A
I
DS
k
ee
p
s
tr
ac
k
o
f
n
o
r
m
al
s
y
s
tem
b
eh
av
io
r
an
d
s
en
d
s
aler
ts
wh
en
th
er
e
ar
e
b
ig
d
if
f
er
en
ce
s
f
r
o
m
wh
at’
s
co
n
s
id
er
ed
n
o
r
m
al
[
9
]
,
[
1
0
]
.
T
h
is
r
esear
ch
aim
s
to
s
tr
en
g
th
en
s
ec
u
r
ity
in
th
e
f
ield
o
f
I
o
T
b
y
cr
ea
tin
g
an
d
e
v
alu
atin
g
an
en
h
an
ce
d
in
tr
u
s
io
n
d
etec
tio
n
s
y
s
tem
(
I
DS)
th
at
u
s
es
m
u
lticlas
s
d
ec
is
io
n
ju
n
g
le
(
MD
J
)
alg
o
r
ith
m
[
1
1
]
.
B
y
em
p
l
o
y
in
g
a
s
o
p
h
is
ticated
alg
o
r
ith
m
lik
e
MD
J
,
th
e
I
DS
a
im
s
to
d
etec
t
a
n
d
g
u
ar
d
ag
ain
s
t
a
b
r
o
ad
s
p
ec
t
r
u
m
o
f
cy
b
er
th
r
ea
ts
th
at
tar
g
et
I
o
T
p
r
o
t
o
co
ls
.
T
h
e
I
DS
f
r
am
ewo
r
k
in
co
r
p
o
r
ates
th
e
MD
J
co
m
p
lem
en
ted
b
y
f
ea
tu
r
e
en
g
in
ee
r
in
g
an
d
d
ata
clea
n
s
in
g
tec
h
n
iq
u
es
to
im
p
r
o
v
e
class
if
icatio
n
ac
c
u
r
ac
y
.
T
h
r
o
u
g
h
ex
h
au
s
tiv
e
p
e
r
f
o
r
m
a
n
ce
test
in
g
,
th
is
r
esear
ch
in
ten
d
s
to
o
f
f
er
cu
ttin
g
-
ed
g
e
i
n
s
ig
h
ts
in
to
th
e
f
ield
o
f
I
o
T
s
ec
u
r
ity
.
T
h
e
r
ati
o
n
ale
b
eh
in
d
u
s
in
g
m
u
lticlas
s
d
ec
is
io
n
ju
n
g
le
f
o
r
m
u
lticlas
s
clas
s
if
icatio
n
is
d
eli
n
ea
ted
b
elo
w:
a.
Netwo
r
k
in
tr
u
s
io
n
d
etec
tio
n
s
y
s
tem
s
(
NI
DS)
n
ee
d
to
class
if
y
n
etwo
r
k
tr
af
f
ic
in
t
o
d
if
f
er
e
n
t
ca
teg
o
r
ies,
n
o
t
lim
itin
g
to
o
n
ly
b
en
ig
n
o
r
n
e
f
ar
io
u
s
.
T
h
e
m
u
lticlas
s
d
ec
is
i
o
n
ju
n
g
le
is
r
o
b
u
s
t
en
o
u
g
h
to
d
etec
t
v
ar
io
u
s
ty
p
es o
f
in
tr
u
s
io
n
s
.
b.
T
h
e
co
n
s
tan
tly
ch
a
n
g
in
g
p
att
er
n
s
o
f
n
etwo
r
k
tr
af
f
ic
a
n
d
t
h
e
g
r
o
win
g
n
u
m
b
er
o
f
cy
b
er
th
r
ea
ts
r
eq
u
ir
e
f
ea
s
ib
le
s
o
lu
tio
n
s
.
Dec
is
io
n
ju
n
g
les,
wh
ich
u
s
e
ap
p
r
o
p
r
ia
te
d
ec
is
io
n
-
m
ak
in
g
p
r
o
ce
s
s
es
an
d
b
ein
g
a
n
en
s
em
b
le
o
f
DAGs c
an
h
a
n
d
le
th
ese
ch
an
g
es b
etter
th
a
n
s
im
p
ler
m
eth
o
d
s
.
c.
T
h
e
d
ec
is
io
n
ju
n
g
le
alg
o
r
ith
m
is
lig
h
t
-
weig
h
t
an
d
ca
n
ea
s
ily
s
ca
le
to
h
an
d
le
m
o
r
e
d
ata,
m
ak
in
g
it
a
g
o
o
d
f
it
f
o
r
s
y
s
tem
s
with
lim
ited
r
eso
u
r
ce
s
,
s
u
ch
as
I
o
T
d
ev
ices.
I
t
ca
n
q
u
ick
ly
m
an
ag
e
lar
g
e
a
m
o
u
n
ts
o
f
d
ata
with
o
u
t c
o
m
p
r
o
m
is
in
g
o
n
th
e
ef
f
ec
tiv
en
ess
.
d.
Fals
e
p
o
s
itiv
es
an
d
f
alse
n
e
g
ativ
es
p
o
s
e
a
co
n
s
id
er
ab
le
ch
allen
g
e
f
o
r
NI
DS,
as
th
ey
ca
n
in
u
n
d
ate
ad
m
in
is
tr
ato
r
s
an
d
u
n
d
er
m
i
n
e
th
eir
co
n
f
id
en
ce
in
th
e
s
y
s
tem
.
Dec
is
io
n
ju
n
g
les
o
p
tim
ize
th
e
p
ar
am
eter
s
to
r
ed
u
ce
f
alse a
ler
ts
wh
ile
ef
f
ici
en
tly
d
etec
tin
g
th
e
m
ajo
r
ity
o
f
th
r
ea
ts
.
e.
T
h
e
m
u
lticlas
s
d
ec
is
io
n
ju
n
g
l
e
alg
o
r
ith
m
d
em
o
n
s
tr
ates
ex
c
ep
tio
n
al
p
r
o
f
icien
c
y
in
m
an
a
g
in
g
im
b
ala
n
ce
d
d
atasets
,
wh
ich
is
a
p
r
ev
alen
t
ch
allen
g
e
in
th
e
n
etwo
r
k
in
tr
u
s
io
n
d
etec
tio
n
s
tu
d
y
wh
e
r
e
in
s
tan
ce
s
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
E
xp
lo
r
in
g
th
e
effec
tiven
ess
o
f
mu
lticla
s
s
d
ec
is
io
n
ju
n
g
le
fo
r
in
tern
et
o
f th
in
g
s
s
ec
u
r
ity
(
S
mith
a
R
a
ja
g
o
p
a
l
)
3097
attac
k
s
ar
e
s
u
b
s
tan
tially
o
u
tn
u
m
b
er
ed
b
y
r
e
g
u
lar
tr
a
f
f
ic
p
att
er
n
s
.
B
y
u
tili
zin
g
its
en
s
em
b
le
m
ec
h
an
is
m
,
a
d
ec
is
io
n
ju
n
g
le
s
ig
n
if
ican
tly
i
m
p
r
o
v
es
d
etec
tio
n
ac
c
u
r
ac
y
f
o
r
b
o
th
f
r
eq
u
e
n
tly
o
cc
u
r
r
in
g
an
d
r
ar
e
attac
k
ca
teg
o
r
ies.
Netwo
r
k
in
tr
u
s
io
n
d
etec
tio
n
s
y
s
tem
s
f
ac
e
ch
allen
g
es
in
d
etec
tin
g
an
o
m
alies
ef
f
icien
tly
.
T
h
e
f
ir
s
t
ch
allen
g
e
p
e
r
tain
s
to
th
e
f
ac
t
t
h
at
m
alicio
u
s
attac
k
s
an
d
th
r
e
ats
ar
e
co
n
s
tan
tly
e
v
o
lv
in
g
,
m
ak
in
g
it d
if
f
icu
lt
f
o
r
ex
is
tin
g
in
tr
u
s
io
n
d
etec
tio
n
s
y
s
tem
s
to
co
n
ten
d
with
.
T
h
is
o
f
ten
lead
s
to
lo
w
d
etec
tio
n
r
ates
an
d
h
ig
h
f
alse
alar
m
s
.
T
h
e
s
ec
o
n
d
c
h
allen
g
e
is
th
at
th
e
tr
ad
itio
n
al
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
s
ap
p
lied
in
th
e
f
ield
o
f
n
etwo
r
k
in
t
r
u
s
io
n
d
etec
tio
n
s
u
f
f
er
f
r
o
m
p
r
o
b
lem
s
s
u
ch
as
o
v
er
f
itti
n
g
,
wh
e
r
e
th
e
m
o
d
e
l
p
er
f
o
r
m
s
well
o
n
tr
ain
in
g
d
ata
b
u
t
d
o
es
n
o
t
p
er
f
o
r
m
well
o
n
u
n
k
n
o
wn
in
s
tan
c
es
an
d
th
e
p
r
esen
ce
o
f
h
ig
h
b
ias
d
u
e
to
ir
r
elev
an
t
f
ea
tu
r
es.
Ad
d
itio
n
ally
,
th
e
cla
s
s
d
is
tr
ib
u
tio
n
o
f
n
etwo
r
k
tr
af
f
ic
is
o
f
te
n
u
n
b
alan
ce
d
,
with
n
o
r
m
al
tr
af
f
ic
b
ein
g
m
u
ch
m
o
r
e
co
m
m
o
n
th
a
n
n
ef
ar
io
u
s
tr
af
f
ic
d
u
e
to
wh
ich
th
e
m
ac
h
in
e
lear
n
in
g
m
o
d
el
m
ay
f
ail
to
lear
n
th
e
n
etwo
r
k
p
atter
n
s
ef
f
ec
tiv
ely
[
1
2
]
,
[
1
3
]
.
2.
L
I
T
E
R
AT
U
RE
R
E
VI
E
W
T
h
e
r
esear
ch
co
m
m
u
n
it
y
is
s
tr
iv
in
g
h
ar
d
to
ex
p
l
o
r
e
th
e
a
r
ea
s
o
f
t
h
e
I
o
T
an
d
in
tr
u
s
io
n
d
etec
tio
n
s
y
s
tem
s
(
I
DS)
to
d
ev
elo
p
n
ew
way
s
to
im
p
r
o
v
e
s
ec
u
r
ity
.
T
h
is
s
ec
tio
n
f
o
cu
s
es
o
n
s
o
m
e
o
f
th
e
latest
I
DS
id
ea
s
th
at
h
av
e
b
ee
n
s
u
g
g
ested
.
A
k
ey
r
eso
u
r
ce
in
th
is
ar
ea
is
th
e
E
d
g
e
-
I
I
o
T
s
et
d
ataset
[
1
4
]
,
wh
ich
is
a
d
etailed
co
llectio
n
o
f
d
ata
ab
o
u
t
c
y
b
er
s
ec
u
r
ity
f
o
r
I
o
T
th
at
ca
n
b
e
u
s
ed
b
y
m
ac
h
in
e
lear
n
i
n
g
(
ML
)
alg
o
r
ith
m
s
.
E
x
p
er
ts
h
av
e
ex
p
e
r
im
en
ted
with
d
if
f
er
en
t
tech
n
iq
u
es
th
at
f
u
s
e
v
ar
io
u
s
alg
o
r
ith
m
s
to
d
etec
t
n
etwo
r
k
in
tr
u
s
io
n
s
ef
f
ec
tiv
ely
an
d
im
p
r
o
v
e
h
o
w
well
th
ey
wo
r
k
.
On
e
o
f
th
ese
tech
n
iq
u
es
u
s
es
a
co
m
b
in
atio
n
o
f
a
n
e
u
tr
o
s
o
p
h
ic
lo
g
ic
class
if
ier
(
a
m
o
r
e
a
d
v
an
ce
d
v
er
s
io
n
o
f
f
u
zz
y
l
o
g
ic)
a
n
d
a
g
en
etic
alg
o
r
ith
m
to
cr
ea
te
r
u
les.
T
h
is
m
eth
o
d
h
as
b
ee
n
s
u
cc
ess
f
u
l
in
lo
wer
in
g
th
e
r
ate
o
f
f
alse
alar
m
s
t
o
ju
s
t
3
.
1
9
%,
w
h
ich
is
b
ette
r
th
an
m
a
n
y
o
th
er
m
eth
o
d
s
[
1
5
]
.
I
n
a
s
t
u
d
y
b
y
P
r
az
e
r
e
s
N
,
C
o
s
t
a
R
,
S
a
n
t
o
s
L
,
a
n
d
t
h
e
i
r
t
e
a
m
,
t
h
e
y
p
r
o
p
o
s
e
d
a
d
is
t
r
i
b
u
t
e
d
I
D
S
wi
t
h
a
s
t
r
o
n
g
a
n
d
d
e
p
e
n
d
a
b
l
e
d
es
i
g
n
f
o
r
f
o
g
c
o
m
p
u
t
i
n
g
.
T
h
e
y
u
s
e
d
t
w
o
d
a
t
as
e
ts
,
I
o
T
-
2
3
a
n
d
M
QT
T
-
I
o
T
-
I
D
S
2
0
2
0
,
t
o
c
r
e
a
t
e
I
D
S
m
o
d
e
ls
.
F
o
r
t
h
e
I
o
T
-
2
3
d
a
t
a
s
e
t
,
t
h
e
y
u
s
e
d
r
a
n
d
o
m
f
o
r
e
s
t
(
R
F
)
a
n
d
n
a
i
v
e
B
a
y
e
s
(
N
B
)
m
o
d
e
l
s
,
a
n
d
f
o
r
t
h
e
M
Q
T
T
-
I
o
T
-
I
DS
2
0
2
0
d
a
t
as
e
t
,
t
h
e
y
u
s
e
d
l
o
g
is
t
i
c
r
e
g
r
e
s
s
i
o
n
(
L
R
)
a
n
d
d
ec
is
i
o
n
t
r
e
e
(
D
T
)
m
o
d
e
l
s
.
T
h
e
s
e
m
o
d
e
l
s
we
r
e
t
h
e
n
c
o
m
p
a
r
e
d
t
o
t
h
r
e
e
o
t
h
e
r
I
D
S
a
r
c
h
i
t
e
ct
u
r
e
s
d
e
s
i
g
n
e
d
f
o
r
I
o
T
t
r
a
f
f
i
c
.
T
h
e
c
o
m
p
a
r
i
s
o
n
w
as
b
as
e
d
o
n
p
e
r
f
o
r
m
a
n
c
e
m
e
t
r
i
c
s
li
k
e
a
cc
u
r
a
c
y
,
p
r
e
c
i
s
i
o
n
,
r
e
c
al
l
,
a
n
d
F1
-
s
c
o
r
e
[
1
6
]
.
R
esear
ch
er
s
u
tili
ze
d
th
e
UNSW
-
N
B
1
5
n
etwo
r
k
tr
af
f
ic
d
ataset
to
ev
alu
ate
ML
m
o
d
e
ls
f
o
r
I
DS
en
h
an
ce
m
e
n
t
in
Mo
r
e
et
al.
Sp
ec
if
ically
,
th
ey
em
p
l
o
y
ed
l
o
g
is
tic
r
eg
r
ess
io
n
(
L
R
)
,
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
es
(
SVM)
,
d
ec
is
io
n
tr
ee
s
(
DT
)
,
an
d
r
an
d
o
m
f
o
r
est
(
R
F).
T
o
o
p
tim
ize
th
e
p
er
f
o
r
m
a
n
ce
o
f
L
R
,
DT
,
an
d
lin
ea
r
SVM,
th
e
s
tu
d
y
co
n
d
u
cted
h
y
p
er
p
ar
am
eter
tu
n
in
g
th
at
led
to
a
g
o
o
d
p
er
f
o
r
m
an
ce
[
1
7
]
.
I
n
a
s
tu
d
y
p
u
b
lis
h
e
d
in
[
1
8
]
,
r
esear
ch
e
r
s
ev
alu
ated
th
e
ef
f
ec
tiv
en
ess
o
f
eig
h
t
d
if
f
er
en
t
tr
ee
-
b
ased
class
if
icatio
n
alg
o
r
ith
m
s
f
o
r
p
r
ed
ictin
g
n
etwo
r
k
ev
en
ts
u
s
in
g
th
e
NSL
-
KDD
d
ata
s
et.
Am
o
n
g
s
t
th
ese
alg
o
r
ith
m
s
,
th
e
d
ec
is
io
n
tr
ee
alg
o
r
ith
m
was
em
p
lo
y
e
d
f
o
r
f
ea
tu
r
e
s
elec
tio
n
,
wh
ile
a
r
an
d
o
m
f
o
r
est
alg
o
r
ith
m
was
u
tili
ze
d
as
th
e
p
r
im
ar
y
class
if
ier
.
Aam
ir
an
d
Z
aid
i
[
1
9
]
p
r
o
p
o
s
e
d
an
i
n
tr
u
s
io
n
d
etec
tio
n
m
o
d
e
l
th
at
u
s
ed
p
r
in
cip
al
c
o
m
p
o
n
e
n
t
an
aly
s
is
(
PC
A)
an
d
a
s
u
b
s
et
o
f
th
e
b
e
n
ch
m
ar
k
d
ataset
co
n
tain
in
g
n
o
v
el
attac
k
v
ec
to
r
s
.
T
h
ey
u
s
ed
clu
s
ter
in
g
to
la
b
el
th
e
d
ata
an
d
id
en
tif
y
attac
k
er
class
es
b
ased
o
n
n
o
r
m
al
tr
af
f
ic
p
atter
n
s
.
Af
ter
lab
elin
g
,
th
ey
em
p
lo
y
ed
th
r
ee
m
ac
h
in
e
lear
n
in
g
al
g
o
r
ith
m
s
(
SVM,
k
-
n
ea
r
est
n
eig
h
b
o
r
(
K
-
NN)
,
an
d
R
F)
to
ac
h
iev
e
d
etec
tio
n
ac
cu
r
ac
y
lev
els o
f
9
2
%,
9
5
%,
an
d
9
6
.
6
6
%,
r
esp
ec
tiv
ely
.
Alq
ah
tan
i
an
d
h
is
team
[
2
0
]
u
s
ed
p
o
p
u
la
r
m
ac
h
in
e
lear
n
in
g
m
eth
o
d
s
lik
e
B
ay
esian
n
etwo
r
k
,
n
aiv
e
B
ay
es,
d
ec
is
io
n
tr
ee
s
,
r
an
d
o
m
d
ec
is
io
n
f
o
r
est
,
r
an
d
o
m
tr
e
e,
d
ec
is
io
n
tab
le,
a
n
d
ar
tific
ia
l
n
eu
r
al
n
etwo
r
k
to
d
etec
t
n
etwo
r
k
in
tr
u
s
io
n
s
.
T
h
ey
test
ed
th
ese
m
eth
o
d
s
u
s
in
g
th
e
KDD
cu
p
9
9
d
ataset.
I
n
an
o
th
er
s
tu
d
y
[
2
1
]
,
r
esear
ch
er
s
lo
o
k
ed
in
to
d
etec
tin
g
n
etwo
r
k
in
tr
u
s
io
n
s
u
s
in
g
th
e
UNS
W
-
NB
1
5
an
d
C
I
C
I
DS
2
0
1
7
d
atasets
.
T
h
ey
in
co
r
p
o
r
ated
g
r
a
d
ien
t
b
o
o
s
tin
g
tr
ee
f
o
r
b
in
a
r
y
class
if
icatio
n
an
d
d
ee
p
n
e
u
r
al
n
et
wo
r
k
f
o
r
m
u
lticlas
s
class
if
icatio
n
.
Del
-
I
o
T
[
2
2
]
,
a
n
e
n
s
e
m
b
l
e
le
a
r
n
i
n
g
m
o
d
e
l
f
o
r
a
n
o
m
a
l
y
d
e
t
e
c
t
i
o
n
u
s
i
n
g
s
o
f
t
w
a
r
e
d
e
f
i
n
e
d
n
e
t
w
o
r
k
s
(
SD
N
)
wa
s
p
r
o
p
o
s
ed
t
o
i
d
e
n
t
i
f
y
t
h
e
at
t
a
c
k
s
c
e
n
a
r
i
o
s
o
n
t
h
e
d
y
n
a
m
i
c
c
y
b
e
r
t
h
r
e
a
t
l
a
n
d
s
ca
p
e
.
D
e
ep
f
e
a
t
u
r
e
e
x
t
r
a
c
t
i
o
n
w
as
u
s
e
d
t
o
in
p
u
t
t
h
e
f
e
a
t
u
r
e
s
t
o
a
p
r
o
b
a
b
i
l
i
s
t
i
c
n
e
u
r
a
l
n
e
tw
o
r
k
(
P
N
N
)
i
n
o
r
d
e
r
t
o
o
p
t
i
m
i
z
e
t
h
e
p
e
r
f
o
r
m
a
n
c
e
.
T
a
b
l
e
1
p
r
o
v
i
d
e
s
a
s
u
m
m
a
r
y
o
f
d
a
t
a
s
et
s
w
i
t
h
a
n
d
w
it
h
o
u
t
I
O
T
t
r
a
c
es
a
v
ai
l
a
b
l
e
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
.
I
n
th
e
p
r
o
p
o
s
ed
s
tu
d
y
,
C
I
C
I
o
T
2
0
2
3
d
ataset
h
as
b
ee
n
u
s
ed
to
f
o
r
ex
p
er
im
e
n
tatio
n
.
T
h
e
C
I
C
I
o
T
2
0
2
3
d
ataset
is
p
r
im
ar
ily
u
s
ed
f
o
r
test
in
g
I
o
T
s
y
s
tem
s
b
ec
au
s
e
it
s
im
u
lates
r
ea
l
-
wo
r
ld
I
o
T
en
v
ir
o
n
m
en
ts
with
a
wid
e
r
an
g
e
o
f
d
e
v
ices.
T
h
es
e
d
ev
ices
ca
n
b
eh
a
v
e
as
b
o
th
tar
g
ets
o
f
attac
k
s
o
r
s
o
u
r
ce
s
o
f
attac
k
s
.
W
ith
th
e
g
r
o
win
g
p
r
e
v
alen
ce
o
f
t
h
e
I
o
T
,
it
h
as
ess
ay
ed
a
c
r
itical
r
o
le
in
o
u
r
d
aily
liv
es.
As
a
r
esu
lt,
en
s
u
r
in
g
its
s
ec
u
r
ity
h
as b
ec
o
m
e
p
ar
am
o
u
n
t to
f
ac
ilit
ate
its
s
ea
m
less
,
s
e
cu
r
e,
an
d
r
eliab
le
o
p
er
atio
n
.
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
.
3
,
J
u
n
e
20
25
:
3
0
9
5
-
3
1
0
6
3098
T
ab
le
1
.
Su
m
m
a
r
y
o
f
r
elev
a
n
t
n
etwo
r
k
in
tr
u
s
io
n
d
etec
tio
n
d
a
tasets
S
l
.
N
o
D
a
t
a
s
e
t
D
e
scri
p
t
i
o
n
N
o
o
f
F
e
a
t
u
r
e
s
W
h
e
t
h
e
r
a
n
I
O
T
d
a
t
a
s
e
t
Y
e
s/
N
o
1.
Ed
g
e
-
I
I
O
T
[
1
4
]
R
e
se
a
r
c
h
e
r
s
p
r
o
p
o
s
e
d
t
h
e
e
d
g
e
-
i
n
d
u
s
t
r
i
a
l
i
n
t
e
r
n
e
t
o
f
t
h
i
n
g
s
(
I
I
o
T)
,
a
c
o
m
p
r
e
h
e
n
si
v
e
d
a
t
a
s
e
t
t
o
e
n
h
a
n
c
e
i
n
t
r
u
si
o
n
d
e
t
e
c
t
i
o
n
s
y
st
e
ms
f
o
r
I
o
T
a
n
d
I
I
o
T
a
p
p
l
i
c
a
t
i
o
n
s.
I
t
e
n
a
b
l
e
s
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
-
b
a
s
e
d
d
e
t
e
c
t
i
o
n
i
n
t
w
o
mo
d
e
s:
c
e
n
t
r
a
l
i
z
e
d
a
n
d
f
e
d
e
r
a
t
e
d
l
e
a
r
n
i
n
g
.
T
h
e
r
e
s
e
a
r
c
h
e
r
s
g
e
n
e
r
a
t
e
d
a
n
Ed
g
e
-
I
I
o
T
se
t
u
si
n
g
a
c
u
s
t
o
m
I
o
T/
I
I
o
T
t
e
st
b
e
d
,
i
n
c
o
r
p
o
r
a
t
i
n
g
a
w
i
d
e
r
a
n
g
e
o
f
d
e
v
i
c
e
s,
s
e
n
s
o
r
s,
p
r
o
t
o
c
o
l
s,
a
n
d
c
l
o
u
d
/
e
d
g
e
c
o
n
f
i
g
u
r
a
t
i
o
n
s
t
o
p
r
o
v
i
d
e
a
r
e
a
l
i
s
t
i
c
r
e
p
r
e
se
n
t
a
t
i
o
n
o
f
t
h
e
se
s
y
s
t
e
ms
.
61
Y
e
s
2.
N
-
B
a
I
O
T
[
2
3
]
R
e
se
a
r
c
h
e
r
s
d
e
v
e
l
o
p
e
d
a
n
e
w
d
a
t
a
s
e
t
f
o
r
d
e
t
e
c
t
i
n
g
a
b
n
o
r
ma
l
a
c
t
i
v
i
t
i
e
s
i
n
I
o
T
n
e
t
w
o
r
k
s
c
a
l
l
e
d
n
e
t
w
o
r
k
b
a
s
e
d
i
n
t
e
r
n
e
t
o
f
t
h
i
n
g
s
(
B
a
I
o
T)
.
"
S
n
a
p
s
h
o
t
s"
o
f
t
h
e
n
e
t
w
o
r
k
's
b
e
h
a
v
i
o
r
w
e
r
e
e
x
t
r
a
c
t
e
d
a
n
d
a
d
v
a
n
c
e
d
a
l
g
o
r
i
t
h
ms
c
a
l
l
e
d
d
e
e
p
a
u
t
o
e
n
c
o
d
e
r
s
w
e
r
e
e
m
p
l
o
y
e
d
.
T
h
e
s
e
a
l
g
o
r
i
t
h
m
s
a
n
a
l
y
s
e
t
h
e
s
n
a
p
s
h
o
t
s
a
n
d
i
d
e
n
t
i
f
y
u
n
u
s
u
a
l
t
r
a
f
f
i
c
p
a
t
t
e
r
n
s
t
h
a
t
may
i
n
d
i
c
a
t
e
c
o
mp
r
o
mi
se
d
o
r
mal
i
c
i
o
u
s
I
o
T
d
e
v
i
c
e
s w
i
t
h
i
n
t
h
e
n
e
t
w
o
r
k
.
1
1
5
Y
e
s
3.
B
o
T
-
I
o
T
[
2
4
]
B
o
T
-
I
o
T
d
a
t
a
se
t
,
t
e
r
m
e
d
a
s
a
b
i
g
d
a
t
a
d
a
t
a
se
t
b
y
t
h
e
a
u
t
h
o
r
s,
c
o
n
s
i
st
s
o
f
7
3
mi
l
l
i
o
n
i
n
s
t
a
n
c
e
s.
A
s
d
i
sc
u
sse
d
[
2
4
]
,
t
h
e
u
sa
g
e
o
f
t
h
i
s
d
a
t
a
se
t
i
s
r
e
c
o
mm
e
n
d
e
d
a
f
t
e
r
a
t
h
o
r
o
u
g
h
d
a
t
a
c
l
e
a
n
i
n
g
p
r
o
c
e
d
u
r
e
f
o
l
l
o
w
e
d
b
y
t
h
e
u
sa
g
e
o
f
v
a
l
i
d
f
e
a
t
u
r
e
s
.
43
Y
e
s
4.
M
Q
TT
-
I
o
T
-
I
D
S
2
0
2
0
[
1
6
]
Th
e
M
Q
TT
-
I
o
T
-
I
D
S
2
0
2
0
d
a
t
a
se
t
c
o
mp
r
i
ses
d
a
t
a
g
e
n
e
r
a
t
e
d
b
y
a
s
i
m
u
l
a
t
e
d
M
Q
TT
n
e
t
w
o
r
k
.
I
t
c
o
n
s
i
st
s
o
f
u
n
p
r
o
c
e
sse
d
p
c
a
p
f
i
l
e
s,
a
l
o
n
g
w
i
t
h
u
n
i
d
i
r
e
c
t
i
o
n
a
l
a
n
d
b
i
d
i
r
e
c
t
i
o
n
a
l
f
l
o
w
f
e
a
t
u
r
e
s
.
T
h
e
p
r
o
sp
e
c
t
i
v
e
a
r
c
h
i
t
e
c
t
u
r
e
s
,
w
h
e
n
p
r
o
p
o
se
d
u
si
n
g
t
h
i
s
d
a
t
a
s
e
t
m
a
y
a
l
l
o
w
f
o
r
e
f
f
e
c
t
i
v
e
f
e
a
t
u
r
e
e
n
g
i
n
e
e
r
i
n
g
t
o
p
r
o
d
u
c
e
b
e
t
t
e
r
r
e
su
l
t
s
32
Y
e
s
5.
U
N
S
W
NB
-
1
5
[
2
5
]
Th
e
U
N
S
W
-
N
B
1
5
d
a
t
a
s
e
t
a
i
ms
t
o
c
r
e
a
t
e
r
e
a
l
i
s
t
i
c
n
e
t
w
o
r
k
e
n
v
i
r
o
n
me
n
t
s
b
y
i
n
c
l
u
d
i
n
g
c
o
mm
o
n
l
o
w
-
p
r
o
f
i
l
e
c
y
b
e
r
a
t
t
a
c
k
s.
I
t
f
e
a
t
u
r
e
s
t
e
n
a
t
t
a
c
k
t
y
p
e
s.
T
h
e
a
r
t
i
c
l
e
[
2
5
]
p
r
o
v
i
d
e
s
a
d
e
t
a
i
l
e
d
b
r
e
a
k
d
o
w
n
o
f
t
h
e
n
u
m
b
e
r
o
f
r
e
c
o
r
d
s
p
e
r
a
t
t
a
c
k
t
y
p
e
a
n
d
t
h
e
i
r
d
i
st
r
i
b
u
t
i
o
n
i
n
t
h
e
t
r
a
i
n
i
n
g
a
n
d
t
e
st
i
n
g
s
e
t
s
.
49
No
6.
C
I
C
I
D
S
2
0
1
7
[
2
6
]
Th
e
d
a
t
a
set
i
n
c
l
u
d
e
s
m
o
d
e
r
n
a
t
t
a
c
k
s
t
h
a
t
a
c
c
u
r
a
t
e
l
y
r
e
s
e
mb
l
e
a
c
t
u
a
l
r
e
a
l
-
w
o
r
l
d
d
a
t
a
.
I
t
a
l
s
o
i
n
c
l
u
d
e
s
t
h
e
f
i
n
d
i
n
g
s
o
f
n
e
t
w
o
r
k
t
r
a
f
f
i
c
a
n
a
l
y
s
i
s
c
o
n
d
u
c
t
e
d
b
y
t
h
e
C
I
C
-
F
l
o
w
M
e
t
e
r
,
w
h
i
c
h
l
a
b
e
l
s
d
a
t
a
f
l
o
w
s
b
a
se
d
o
n
f
a
c
t
o
r
s l
i
k
e
t
i
me
st
a
mp
,
so
u
r
c
e
a
n
d
d
e
st
i
n
a
t
i
o
n
I
P
a
d
d
r
e
sses
,
so
u
r
c
e
a
n
d
d
e
s
t
i
n
a
t
i
o
n
p
o
r
t
s,
p
r
o
t
o
c
o
l
s,
a
n
d
t
y
p
e
s
o
f
a
t
t
a
c
k
s s
t
o
r
e
d
i
n
C
S
V
f
i
l
e
s.
17
No
T
h
e
C
I
C
I
o
T
2
0
2
3
d
ataset
was
cr
ea
ted
u
s
in
g
1
0
5
d
ev
ices
a
n
d
3
3
r
ea
l
-
wo
r
ld
attac
k
s
.
T
h
e
s
e
attac
k
s
co
v
er
s
ev
en
ca
teg
o
r
ies,
s
u
ch
as
DDo
S,
Do
S,
R
ec
o
n
,
web
-
b
ased
attac
k
s
,
b
r
u
te
f
o
r
ce
,
s
p
o
o
f
in
g
,
an
d
Mir
ai.
I
n
all
ca
s
es,
h
ar
m
f
u
l
I
o
T
d
e
v
ices
attac
k
o
th
er
I
o
T
d
ev
ices.
T
h
i
s
d
ataset
in
clu
d
es
n
ew
ty
p
es
o
f
attac
k
s
th
at
ar
en
'
t
f
o
u
n
d
in
o
th
er
I
o
T
d
atasets
.
I
t
h
elp
s
I
o
T
ex
p
er
ts
b
u
ild
n
e
w
s
ec
u
r
ity
to
o
ls
b
y
p
r
o
v
id
in
g
d
ata
i
n
d
if
f
er
en
t
f
o
r
m
ats.
T
h
e
C
I
C
I
o
T
2
0
2
3
d
at
aset
ad
v
an
ce
s
th
e
f
ield
o
f
I
o
T
s
ec
u
r
ity
b
y
i
n
tr
o
d
u
cin
g
a
co
m
p
r
eh
en
s
iv
e
n
etwo
r
k
to
p
o
lo
g
y
f
ea
tu
r
i
n
g
d
iv
e
r
s
e
I
o
T
d
ev
ices.
I
t
in
co
r
p
o
r
ates
m
u
ltip
le
cy
b
er
attac
k
s
n
o
t
p
r
e
v
io
u
s
ly
in
clu
d
ed
in
a
s
in
g
le
I
o
T
s
ec
u
r
ity
d
ataset.
Ass
ess
in
g
th
e
p
er
f
o
r
m
an
ce
o
f
p
o
p
u
lar
m
ac
h
in
e
lear
n
in
g
m
eth
o
d
s
ag
ain
s
t
d
if
f
er
en
t
attac
k
ty
p
es p
r
esen
t in
th
e
C
I
C
I
o
T
2
0
2
3
d
ataset
p
r
o
v
id
es a
s
ig
n
if
ican
t a
d
v
an
ta
g
e
co
m
p
ar
e
d
to
o
th
e
r
s
tu
d
ies.
3.
M
E
T
H
O
DO
L
O
G
Y
T
h
is
s
ec
tio
n
p
r
o
v
id
es
a
s
u
m
m
ar
y
o
f
t
h
e
ex
p
e
r
im
en
tal
f
in
d
in
g
s
.
I
t
p
r
o
p
o
s
es
a
m
eth
o
d
f
o
r
c
ar
r
y
in
g
o
u
t
class
if
icatio
n
an
d
d
etec
tio
n
t
ask
s
u
s
in
g
m
u
lticlas
s
d
ec
is
io
n
ju
n
g
le.
T
h
e
class
if
icatio
n
f
r
am
ewo
r
k
o
f
th
e
p
r
o
p
o
s
ed
wo
r
k
is
d
ep
icted
in
Fig
u
r
e
2
.
Dec
is
io
n
ju
n
g
les
ar
e
an
im
p
r
o
v
e
d
alg
o
r
ith
m
in
t
h
e
f
ield
o
f
m
ac
h
in
e
lear
n
in
g
th
at
b
u
ild
o
n
d
ec
is
io
n
tr
ee
s
an
d
d
ec
is
io
n
f
o
r
ests
.
I
n
s
tead
o
f
h
av
in
g
a
s
tr
ict
b
r
a
n
ch
in
g
s
tr
u
ctu
r
e
lik
e
tr
ee
s
,
d
ec
is
io
n
ju
n
g
les
u
s
e
d
ir
ec
ted
ac
y
clic
g
r
ap
h
s
(
DAGs)
.
T
h
is
allo
ws
d
ec
is
io
n
p
o
in
ts
(
r
ep
r
esen
ted
b
y
n
o
d
es)
to
b
e
u
s
ed
in
m
u
ltip
le
b
r
an
ch
es.
B
y
d
o
in
g
th
is
,
d
ec
is
io
n
ju
n
g
les
ca
n
b
e
m
o
r
e
ef
f
ici
en
t
an
d
ac
cu
r
ate
f
o
r
m
ak
in
g
p
r
ed
ictio
n
s
.
T
h
e
n
u
m
b
er
o
f
i
n
s
tan
ce
s
co
n
s
id
er
e
d
f
o
r
tr
ain
i
n
g
an
d
test
in
g
in
th
e
r
atio
8
0
:2
0
is
en
u
m
er
ated
i
n
T
ab
le
2
.
I
n
th
e
p
r
o
p
o
s
ed
w
o
r
k
,
we
h
a
v
e
u
s
ed
Azu
r
e
m
ac
h
in
e
lear
n
i
n
g
as
a
p
latf
o
r
m
to
c
r
ea
te
th
e
m
ac
h
in
e
-
lear
n
in
g
p
ip
elin
e
.
Azu
r
e
m
ac
h
in
e
lear
n
in
g
,
a
m
ac
h
in
e
-
lear
n
in
g
p
latf
o
r
m
h
o
s
ted
o
n
th
e
clo
u
d
,
allo
ws
u
s
er
s
to
cr
ea
te
m
ac
h
in
e
-
lear
n
in
g
m
o
d
els
th
at
ca
n
b
e
ad
ju
s
ted
to
v
a
r
io
u
s
wo
r
k
lo
a
d
s
,
s
h
ar
ed
ac
r
o
s
s
m
u
ltip
le
d
ev
ices,
an
d
d
ep
lo
y
ed
d
ir
ec
tly
o
n
to
th
e
clo
u
d
[
2
7
]
.
Azu
r
e
m
ac
h
in
e
lear
n
in
g
p
latf
o
r
m
[
2
8
]
was
u
s
ed
to
co
m
p
ar
e
th
e
p
er
f
o
r
m
an
ce
o
f
v
ar
i
o
u
s
class
if
ier
s
.
T
h
e
s
tu
d
y
u
s
es
m
u
lticlas
s
d
ec
is
io
n
ju
n
g
le,
an
alg
o
r
ith
m
s
u
it
ab
le
f
o
r
task
s
in
v
o
lv
in
g
class
if
y
in
g
d
ata
in
to
m
u
ltip
le
attac
k
ca
teg
o
r
ies.
T
h
is
is
p
ar
ticu
lar
ly
a
p
p
r
o
p
r
iate
f
o
r
t
h
e
task
o
f
id
e
n
tify
in
g
d
if
f
e
r
en
t
ty
p
es
o
f
attac
k
s
,
as
th
er
e
ar
e
3
4
d
is
tin
ct
attac
k
ty
p
es
to
d
is
tin
g
u
is
h
b
etwe
en
in
th
is
ca
s
e.
Dec
i
s
io
n
ju
n
g
les
h
av
e
a
s
tr
u
ctu
r
ed
h
ie
r
ar
ch
y
,
p
r
o
v
id
i
n
g
b
etter
o
v
er
all
p
er
f
o
r
m
an
ce
an
d
s
tab
ilit
y
.
On
t
h
e
o
th
e
r
h
a
n
d
,
d
e
cisi
o
n
f
o
r
ests
p
r
io
r
itize
v
a
r
iety
b
y
u
s
in
g
b
a
g
g
in
g
an
d
s
elec
tin
g
f
ea
tu
r
es
r
an
d
o
m
ly
.
T
h
e
d
ec
is
io
n
j
u
n
g
le
m
o
d
el
h
as
a
m
o
r
e
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
E
xp
lo
r
in
g
th
e
effec
tiven
ess
o
f
mu
lticla
s
s
d
ec
is
io
n
ju
n
g
le
fo
r
in
tern
et
o
f th
in
g
s
s
ec
u
r
ity
(
S
mith
a
R
a
ja
g
o
p
a
l
)
3099
in
tr
icate
s
tr
u
ctu
r
e
o
r
g
a
n
ized
in
a
h
ier
ar
ch
ical
m
an
n
er
.
T
h
is
d
esig
n
allo
ws
f
o
r
en
h
an
ce
d
id
en
tific
atio
n
o
f
p
atter
n
s
an
d
p
o
ten
tially
im
p
r
o
v
es th
e
m
o
d
el'
s
p
r
ec
is
io
n
an
d
r
esil
ien
ce
.
Fig
u
r
e
2
.
Me
th
o
d
o
lo
g
y
o
f
th
e
p
r
o
p
o
s
ed
wo
r
k
T
ab
le
2
.
T
r
ai
n
in
g
a
n
d
test
in
g
s
am
p
les co
n
s
id
er
ed
f
o
r
e
x
p
er
i
m
en
tatio
n
S
l
.
N
o
A
t
t
a
c
k
C
a
t
e
g
o
r
y
Tr
a
i
n
i
n
g
se
t
Te
st
i
n
g
s
e
t
1
B
a
c
k
d
o
o
r
_
M
a
l
w
a
r
e
61
15
2
B
e
n
i
g
n
Tr
a
f
f
i
c
1
9
5
8
1
4
8
9
5
3
B
r
o
w
serH
i
j
a
c
k
i
n
g
1
1
2
28
4
C
o
mm
a
n
d
I
n
j
e
c
t
i
o
n
84
21
5
D
D
o
S
-
A
C
K
_
F
r
a
g
me
n
t
a
t
i
o
n
5
1
4
5
1
2
8
6
6
D
D
o
S
-
H
TTP_
F
l
o
o
d
5
0
1
1
2
5
7
D
D
o
S
-
I
C
M
P
_
F
l
o
o
d
1
2
9
0
2
5
3
2
2
5
6
8
D
D
o
S
-
I
C
M
P
_
F
r
a
g
me
n
t
a
t
i
o
n
8
1
7
8
2
0
4
5
9
D
D
o
S
-
P
S
H
A
C
K
_
F
l
o
o
d
7
3
9
1
6
1
8
4
7
9
10
D
D
o
S
-
R
S
TFI
N
F
l
o
o
d
7
2
6
5
8
1
8
1
6
5
11
D
D
o
S
-
S
l
o
w
L
o
r
i
s
3
9
4
99
12
D
D
o
S
-
S
Y
N
_
F
l
o
o
d
7
3
3
1
5
1
8
3
2
9
13
D
D
o
S
-
S
y
n
o
n
y
m
o
u
sI
P
_
F
l
o
o
d
6
4
5
4
4
1
6
1
3
6
14
D
D
o
S
-
TC
P
_
F
l
o
o
d
8
1
0
3
4
2
0
2
5
9
15
D
D
o
S
-
U
D
P
_
F
l
o
o
d
9
6
9
6
4
2
4
2
4
1
16
D
D
o
S
-
U
D
P
_
F
r
a
g
m
e
n
t
a
t
i
o
n
5
1
4
5
1
2
8
6
17
D
i
c
t
i
o
n
a
r
y
B
r
u
t
e
F
o
r
c
e
2
5
9
65
18
D
N
S
_
S
p
o
o
f
i
n
g
3
2
2
7
8
0
7
19
D
o
S
-
H
TTP_
F
l
o
o
d
1
8
4
5
4
6
1
20
D
o
S
-
S
Y
N
_
F
l
o
o
d
1
0
9
4
8
1
2
7
3
7
0
21
D
o
S
-
TC
P
_
F
l
o
o
d
1
2
8
8
8
0
3
2
2
2
0
22
D
o
S
-
U
D
P
_
F
l
o
o
d
1
5
6
7
9
4
3
9
1
9
8
23
M
i
r
a
i
-
g
r
e
e
t
h
_
f
l
o
o
d
1
7
6
9
2
4
4
2
3
24
M
i
r
a
i
-
g
r
e
i
p
_
f
l
o
o
d
1
3
5
6
2
3
3
9
0
25
M
i
r
a
i
-
u
d
p
p
l
a
i
n
1
6
1
3
3
4
0
3
3
26
M
I
TM
-
A
r
p
S
p
o
o
f
i
n
g
5
6
1
5
1
4
0
4
27
R
e
c
o
n
-
H
o
s
t
D
i
sc
o
v
e
r
y
2
4
0
6
6
0
1
28
R
e
c
o
n
-
O
S
S
c
a
n
1
7
8
0
4
4
5
29
R
e
c
o
n
-
P
i
n
g
S
w
e
e
p
33
10
30
R
e
c
o
n
-
P
o
r
t
S
c
a
n
1
4
9
0
3
7
3
31
S
q
l
I
n
j
e
c
t
i
o
n
98
24
32
U
p
l
o
a
d
i
n
g
_
A
t
t
a
c
k
18
5
33
V
u
l
n
e
r
a
b
i
l
i
t
y
S
c
a
n
6
4
7
1
7
4
34
X
S
S
58
14
DAGs
ty
p
ically
p
r
o
d
u
ce
d
e
cisi
o
n
s
th
at
lead
to
r
elativ
el
y
m
in
im
u
m
d
ata
s
to
r
ag
e,
r
e
s
u
ltin
g
in
ex
ce
llen
t
o
v
er
all
p
e
r
f
o
r
m
an
ce
.
T
h
e
m
u
lticlas
s
asp
ec
t
o
f
d
ec
i
s
io
n
ju
n
g
le
d
o
es
n
o
t
r
ely
o
n
s
p
ec
if
ic
d
is
tr
ib
u
tio
n
ass
u
m
p
tio
n
s
,
en
ab
lin
g
it
t
o
c
ap
tu
r
e
n
o
n
lin
ea
r
b
o
u
n
d
ar
ies
b
etwe
en
class
es
ef
f
ec
tiv
ely
.
Dec
is
io
n
J
u
n
g
le
ca
n
also
id
en
tify
r
elev
an
t f
ea
tu
r
es
an
d
class
if
y
d
ata
wh
ile
b
ein
g
r
esis
tan
t to
n
o
is
y
f
ea
tu
r
es d
u
r
in
g
tr
ain
in
g
,
m
ak
i
n
g
it r
o
b
u
s
t.
As
co
m
p
ar
ed
to
th
e
tr
ad
itio
n
al
tr
ee
s
,
ea
ch
n
o
d
e
in
th
e
d
ec
is
io
n
ju
n
g
le
ca
n
h
av
e
m
u
lti
p
le
p
ath
s
lead
in
g
to
th
e
r
ed
u
ctio
n
in
t
h
e
n
u
m
b
er
o
f
n
o
d
es.
T
h
e
d
e
cisi
o
n
ju
n
g
le
m
o
d
el
en
h
a
n
ce
s
th
e
r
an
d
o
m
f
o
r
est
m
o
d
el
b
y
s
tr
u
ctu
r
in
g
tr
ee
s
in
d
is
tin
ct
lay
er
s
,
ak
in
to
a
lay
e
r
ed
f
o
r
est.
E
ac
h
lay
er
is
tr
ea
te
d
as
an
in
d
iv
i
d
u
al
f
o
r
est,
an
d
th
e
p
r
e
d
ictio
n
s
f
r
o
m
o
n
e
lay
er
b
ec
o
m
e
th
e
in
p
u
t
s
f
o
r
th
e
s
u
b
s
eq
u
en
t
lay
er
,
en
a
b
lin
g
a
h
ier
ar
ch
ical
d
ec
is
io
n
-
m
ak
in
g
p
r
o
ce
s
s
.
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
.
3
,
J
u
n
e
20
25
:
3
0
9
5
-
3
1
0
6
3100
Ma
th
em
atica
lly
,
w
ith
in
th
e
ex
p
an
s
e
o
f
t
h
e
d
ec
is
io
n
ju
n
g
le,
t
h
er
e
ex
is
t
n
u
m
er
o
u
s
d
ec
is
io
n
t
r
ee
s
,
ea
ch
d
en
o
ted
as
T
1
,
T
2
,..,…
T
w
h
er
e
‘
’
r
ep
r
esen
ts
th
e
t
o
tal
n
u
m
b
er
o
f
tr
ee
s
p
r
esen
t.
T
h
e
f
in
al
p
r
ed
ictio
n
is
d
eter
m
in
ed
b
y
co
m
b
in
in
g
th
e
p
r
ed
ictio
n
s
m
ad
e
b
y
ea
c
h
in
d
i
v
id
u
al
d
ec
is
io
n
tr
ee
in
th
e
e
n
s
em
b
le.
=
a
r
g
ma
x
∑
(
(
)
=
=
1
(
1
)
T
h
e
in
d
icat
o
r
f
u
n
ctio
n
(
d
en
o
t
ed
as
"I
")
ass
ig
n
s
a
v
alu
e
o
f
1
to
elem
en
ts
th
at
s
atis
f
y
th
e
s
p
ec
if
ied
co
n
d
itio
n
an
d
0
to
ele
m
en
ts
th
at
d
o
n
o
t
.
I
n
a
m
u
lticlas
s
clas
s
if
icatio
n
task
,
th
e
o
b
jectiv
e
is
to
d
eter
m
in
e
wh
eth
er
a
n
in
s
tan
ce
f
r
o
m
th
e
p
r
o
v
id
ed
d
at
a
co
llectio
n
b
elo
n
g
s
to
a
p
ar
ticu
lar
p
r
e
d
ef
in
ed
g
r
o
u
p
o
r
n
o
t.
3
.
1
.
P
re
-
pro
ce
s
s
ing
Data
clea
n
s
in
g
is
a
ess
en
tia
l st
ep
in
th
e
m
ac
h
in
e
lear
n
in
g
d
o
m
ain
.
I
t in
v
o
lv
es h
an
d
li
n
g
in
c
o
m
p
lete
o
r
wr
o
n
g
r
ec
o
r
d
s
,
as
well
as
d
o
i
n
g
awa
y
with
o
u
tlier
s
an
d
d
u
p
licates.
T
h
e
d
ataset
u
s
ed
in
th
i
s
r
esear
ch
in
clu
d
ed
3
3
s
p
ec
if
ic
ty
p
es
o
f
I
o
T
attac
k
p
atter
n
s
an
d
4
6
n
u
m
e
r
ical
f
ea
tu
r
es.
ch
ar
ac
ter
is
tic
s
ca
lin
g
is
s
ig
n
if
ican
t
f
o
r
alg
o
r
ith
m
s
th
at
d
ep
e
n
d
o
n
d
i
s
tan
ce
-
b
ased
ca
lcu
latio
n
s
.
I
f
th
e
r
ec
o
r
d
s
is
o
n
d
iv
er
s
e
s
ca
les,
s
o
m
e
s
tatis
tic
s
f
ac
to
r
s
m
ig
h
t
a
f
f
ec
t
o
n
th
e
m
o
d
el’
s
p
er
f
o
r
m
a
n
ce
.
T
h
is
p
r
e
-
p
r
o
ce
s
s
in
g
s
tep
allo
ws
alg
o
r
it
h
m
wo
r
k
b
etter
a
n
d
m
o
r
e
r
eliab
ly
.
A
m
o
d
u
le
k
n
o
wn
as
“
c
o
n
f
ig
u
r
e
n
o
r
m
alize
d
ata”
f
r
o
m
Azu
r
e
s
er
v
ice
was
u
s
ed
to
n
o
r
m
alize
t
h
e
d
ata.
3
.
2
.
F
e
a
t
ure
s
elec
t
io
n
I
n
o
r
d
er
to
s
elec
t
th
e
m
o
s
t
r
el
ev
an
t
f
ea
t
u
r
es
f
r
o
m
th
e
d
ataset,
th
e
c
o
m
p
o
n
en
t
co
n
s
id
er
e
d
f
r
o
m
Azu
r
e
Ma
ch
in
e
L
ea
r
n
in
g
Stu
d
io
is
p
er
m
u
tatio
n
f
ea
tu
r
e
im
p
o
r
tan
ce
(
PF
I
)
,
a
f
ilter
-
b
ased
f
ea
tu
r
e
s
elec
tio
n
m
eth
o
d
.
T
h
is
co
m
p
o
n
e
n
t
ev
alu
ates
th
e
im
p
o
r
tan
ce
o
f
f
ea
tu
r
es
in
a
m
ac
h
in
e
-
lear
n
i
n
g
m
o
d
el.
I
t
s
h
u
f
f
les
f
ea
tu
r
e
v
alu
es
r
an
d
o
m
l
y
f
o
r
ea
ch
co
lu
m
n
,
t
h
en
co
m
p
ar
es
th
e
m
o
d
el'
s
p
er
f
o
r
m
a
n
ce
b
ef
o
r
e
a
n
d
af
te
r
e
ac
h
s
h
u
f
f
le.
Hig
h
er
p
er
f
o
r
m
an
ce
c
h
an
g
es
in
d
icat
e
m
o
r
e
im
p
o
r
tan
t
f
ea
tu
r
es,
a
s
th
ese
ar
e
m
o
r
e
af
f
ec
ted
b
y
th
e
s
h
u
f
f
lin
g
.
T
h
e
twelv
e
f
ea
tu
r
es
th
at
wer
e
co
n
s
id
er
ed
f
o
r
f
u
r
th
e
r
p
r
o
ce
s
s
in
g
ar
e
as
f
o
llo
ws,
Hea
d
er
_
len
g
th
,
r
ate,
d
r
ate,
s
y
n
_
co
u
n
t,
to
ts
u
m
,
I
AT
,
m
a
g
n
itu
d
e,
Av
g
,
s
td
,
co
v
ar
ian
c
e
an
d
r
ad
iu
s
.
I
t
is
n
o
tewo
r
t
h
y
th
at
th
e
ab
o
v
e
m
en
tio
n
ed
twelv
e
f
ea
tu
r
es c
o
n
tr
ib
u
ted
to
war
d
s
th
e
p
r
e
d
ictio
n
o
u
tco
m
e.
3
.
3
.
Dec
is
io
n J
un
g
le
a
s
t
he
c
la
s
s
if
ier
T
h
e
n
u
m
b
er
o
f
o
p
tim
izatio
n
s
tep
s
co
n
f
ig
u
r
e
d
f
o
r
ea
ch
la
y
er
in
t
h
e
d
ec
is
io
n
DAG
s
p
ec
if
ies
h
o
w
m
an
y
s
tep
s
th
e
s
y
s
tem
s
h
o
u
ld
tak
e
to
o
p
tim
ize
th
at
p
ar
tic
u
lar
lay
er
.
2
0
4
8
was
ass
ig
n
ed
th
e
v
alu
e
f
o
r
t
h
e
n
u
m
b
er
o
f
o
p
tim
izatio
n
s
tep
s
.
E
ig
h
t
d
ec
is
io
n
DAGs
wer
e
co
n
f
ig
u
r
ed
in
o
r
d
er
t
o
p
er
f
o
r
m
th
e
class
if
icatio
n
task
.
T
h
e
m
ax
im
u
m
d
e
p
th
o
f
th
e
d
ec
is
io
n
DAGs
was
co
n
s
id
er
ed
as
3
2
.
A
v
al
u
e
o
f
3
2
w
as
ass
ig
n
ed
as
th
e
wid
th
o
f
th
e
DAG.
4.
RE
SU
L
T
S
AND
D
I
SCU
SS
I
O
N
Giv
en
1
,
0
4
8
,
5
7
5
in
s
tan
ce
s
,
8
3
8
,
8
6
0
(
8
0
%)
wer
e
u
s
ed
f
o
r
m
o
d
el
tr
ain
in
g
,
wh
ile
th
e
r
em
ain
in
g
2
0
9
,
7
1
5
(
2
0
%)
wer
e
u
tili
ze
d
f
o
r
test
in
g
.
T
o
en
s
u
r
e
t
h
e
r
eli
ab
ilit
y
o
f
th
e
m
o
d
el,
a
r
ig
o
r
o
u
s
tech
n
iq
u
e
ca
lled
ten
-
f
o
ld
cr
o
s
s
-
v
alid
atio
n
was
ca
r
r
ied
o
u
t.
T
y
p
ically
,
cr
o
s
s
-
v
alid
atio
n
is
o
f
ten
ap
p
lied
to
v
alid
ate
th
e
ef
f
ec
tiv
en
ess
o
f
th
e
p
r
o
p
o
s
ed
m
o
d
el
th
er
eb
y
av
o
id
o
v
er
f
itti
n
g
[
2
9
]
.
T
h
is
p
r
o
ce
s
s
d
iv
id
es
th
e
d
ata
in
to
ten
eq
u
al
p
a
r
ts
an
d
tr
ain
s
th
e
m
o
d
el
o
n
n
in
e
o
f
th
e
p
ar
ts
wh
ile
ev
alu
atin
g
it
o
n
th
e
r
em
ai
n
in
g
p
ar
t.
B
y
r
e
p
ea
tin
g
th
is
p
r
o
ce
s
s
f
o
r
all
ten
co
m
b
in
atio
n
s
,
th
e
alg
o
r
ith
m
is
s
u
b
jec
ted
to
a
v
ar
iety
o
f
d
ata
s
u
b
s
ets,
p
r
o
v
id
in
g
a
m
o
r
e
ac
cu
r
ate
ass
ess
m
en
t
o
f
its
o
v
e
r
all
p
er
f
o
r
m
an
ce
.
T
ab
le
3
s
u
m
m
ar
izes
th
e
p
er
f
o
r
m
a
n
ce
m
etr
ics
o
f
th
e
d
ec
is
io
n
ju
n
g
le
class
if
ier
,
s
h
o
wca
s
in
g
i
ts
ef
f
ec
tiv
en
ess
.
T
h
e
p
er
f
o
r
m
a
n
ce
m
etr
ics,
n
o
r
m
ally
T
ab
le
u
s
ed
f
o
r
a
p
r
e
d
ictiv
e
s
tu
d
y
in
m
ac
h
in
e
lea
r
n
in
g
p
er
tain
in
g
to
m
u
lticlas
s
class
if
icatio
n
ar
e
p
r
ec
is
io
n
,
r
ec
all
an
d
F1
-
s
co
r
e
as sh
o
wn
in
th
e
(
2
)
to
(
4
)
.
=
(
)
(
)
+
(
)
(
2
)
=
(
)
(
)
+
(
)
(
3
)
1
−
=
2
∗
∗
+
(
4
)
T
h
e
p
r
o
p
o
s
ed
m
o
d
el
co
u
ld
s
u
cc
ess
f
u
lly
id
en
tify
b
en
ig
n
p
at
ter
n
s
.
Ho
wev
er
,
th
er
e
h
av
e
b
ee
n
a
f
ew
m
is
class
if
icatio
n
s
wh
er
ein
b
en
ig
n
p
atter
n
s
h
a
v
e
b
ee
n
i
n
co
r
r
ec
tly
class
if
ied
to
b
e
DNS_
s
p
o
o
f
in
g
a
n
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
E
xp
lo
r
in
g
th
e
effec
tiven
ess
o
f
mu
lticla
s
s
d
ec
is
io
n
ju
n
g
le
fo
r
in
tern
et
o
f th
in
g
s
s
ec
u
r
ity
(
S
mith
a
R
a
ja
g
o
p
a
l
)
3101
MI
T
M_
Ar
p
s
p
o
o
f
in
g
.
T
h
e
p
r
o
p
o
s
ed
v
er
s
io
n
s
u
cc
ess
f
u
lly
d
is
co
v
er
ed
in
s
tan
ce
s
o
f
b
r
o
wser
h
ijack
in
g
,
d
etec
tin
g
9
9
%
o
f
th
ese
attac
k
s
.
I
t
ad
d
itio
n
ally
d
etec
ted
m
an
y
d
if
f
er
e
n
t
s
o
r
ts
o
f
attac
k
v
ec
to
r
s
,
in
clu
d
in
g
s
p
ec
if
ic
f
o
r
m
s
o
f
DDo
S
attac
k
s
(
lik
e
AC
K
f
r
ag
m
en
tatio
n
,
HT
T
P
f
lo
o
d
,
I
C
MP
f
lo
o
d
,
I
C
MP
f
r
ag
m
e
n
tatio
n
,
PS
HACK
Flo
o
d
,
R
STFI
N
Flo
o
d
)
,
Do
S
ass
au
lts
(
lik
e
HT
T
P
Flo
o
d
,
SYN
Flo
o
d
,
T
C
P
F
lo
o
d
,
UDP
Flo
o
d
)
,
Mir
ai
b
o
tn
et
ass
au
lts
(
s
u
ch
as g
r
ee
th
f
lo
o
d
,
g
r
ei
p
f
lo
o
d
,
u
d
p
p
lain
f
lo
o
d
)
,
a
n
d
v
u
ln
e
r
ab
ilit
y
s
ca
n
s
.
T
h
e
n
u
m
b
e
r
o
f
s
am
p
les
f
o
u
n
d
in
d
atasets
p
er
tain
in
g
to
d
if
f
er
en
t
ca
teg
o
r
ies
m
ay
n
o
t
b
e
th
e
s
am
e.
T
h
is
is
a
c
o
m
m
o
n
is
s
u
e
th
at
p
eo
p
le
wo
r
k
in
g
with
m
ac
h
i
n
e
lear
n
i
n
g
f
ac
e,
ty
p
ically
k
n
o
wn
as
h
av
i
n
g
an
im
b
alan
ce
d
d
ataset
[
3
0
]
,
[
3
1
]
.
W
h
en
test
in
g
f
o
r
b
ac
k
d
o
o
r
m
alwa
r
e
an
d
co
m
m
a
n
d
in
ject
io
n
,
d
ec
is
io
n
ju
n
g
le
class
if
ier
was
ab
le
to
co
r
r
ec
tl
y
id
en
tify
6
2
.
5
%
o
f
t
h
e
ca
s
es
it
was
s
u
p
p
o
s
ed
to
class
if
y
.
T
h
e
p
r
o
p
o
s
ed
v
er
s
io
n
s
u
cc
ess
f
u
lly
o
b
s
er
v
ed
ca
s
es
o
f
b
r
o
wser
h
ijack
i
n
g
,
d
etec
tin
g
9
9
%
o
f
th
ese
ass
au
lts
.
I
t
also
d
id
f
air
ly
in
d
etec
tin
g
m
an
y
o
th
e
r
f
o
r
m
s
o
f
attac
k
s
,
s
u
ch
as
d
if
f
er
en
t
k
in
d
s
o
f
DDo
S
attac
k
s
(
lik
e
A
C
K
f
r
ag
m
en
tatio
n
,
HT
T
P
f
lo
o
d
,
I
C
MP
f
lo
o
d
,
I
C
MP
f
r
ag
m
en
tatio
n
,
PS
HACK
f
lo
o
d
,
R
STFI
N
f
lo
o
d
)
,
Do
S
a
s
s
au
lts
(
lik
e
HT
T
P
Flo
o
d
,
SYN
Flo
o
d
,
T
C
P
Flo
o
d
,
UDP
Flo
o
d
)
,
Mir
ai
b
o
t
n
et
a
ttack
s
(
s
u
ch
as
g
r
ee
th
f
lo
o
d
,
g
r
eip
f
lo
o
d
,
u
d
p
p
lain
f
lo
o
d
)
,
an
d
v
u
ln
e
r
ab
ilit
y
s
ca
n
s
.
T
ab
le
3
.
Pre
cisi
o
n
,
r
ec
all
an
d
F1
-
s
co
r
es o
f
3
4
attac
k
ty
p
es
S
l
.
N
o
A
t
t
a
c
k
c
a
t
e
g
o
r
y
P
r
e
c
i
s
i
o
n
R
e
c
a
l
l
F1
-
sc
o
r
e
1
B
a
c
k
d
o
o
r
_
M
a
l
w
a
r
e
0
.
6
2
5
0
.
5
8
8
2
3
5
3
0
.
7
5
0
7
0
4
2
2
B
e
n
i
g
n
Tr
a
f
f
i
c
0
.
8
7
9
2
2
4
4
0
.
9
7
6
6
1
5
4
0
.
9
2
5
3
6
4
4
3
B
r
o
w
serH
i
j
a
c
k
i
n
g
1
0
.
3
4
7
8
2
6
1
0
.
5
1
6
1
2
9
4
C
o
mm
a
n
d
I
n
j
e
c
t
i
o
n
0
.
6
2
5
0
.
2
9
4
1
1
7
6
0
.
4
5
D
D
o
S
-
A
C
K
_
F
r
a
g
me
n
t
a
t
i
o
n
0
.
9
9
7
7
0
9
9
0
.
9
9
9
2
3
5
5
0
.
9
9
8
4
7
2
1
6
D
D
o
S
-
H
TTP_
F
l
o
o
d
0
.
9
9
1
6
6
6
7
0
.
9
4
4
4
4
4
4
0
.
9
6
7
4
7
9
7
7
D
D
o
S
-
I
C
M
P
_
F
l
o
o
d
0
.
9
9
9
9
0
6
5
0
.
9
9
9
9
0
6
5
0
.
9
9
9
9
0
6
5
8
D
D
o
S
-
I
C
M
P
_
F
r
a
g
me
n
t
a
t
i
o
n
0
.
9
9
8
0
0
9
0
.
9
9
8
0
0
9
0
.
9
9
8
0
0
9
9
D
D
o
S
-
P
S
H
A
C
K
_
F
l
o
o
d
0
.
9
9
9
8
3
7
1
0
.
9
9
9
8
3
7
1
0
.
9
9
9
8
3
7
1
10
D
D
o
S
-
R
S
TFI
N
F
l
o
o
d
0
.
9
9
9
9
4
4
7
0
.
9
9
9
9
4
4
7
0
.
9
9
9
9
4
4
7
11
D
D
o
S
-
S
l
o
w
L
o
r
i
s
0
.
9
0
8
1
6
3
3
0
.
9
1
7
5
2
5
8
0
.
9
1
2
8
2
0
5
12
D
D
o
S
-
S
Y
N
_
F
l
o
o
d
0
.
9
9
9
7
3
1
5
0
.
9
9
9
8
3
8
9
0
.
9
9
9
7
8
5
2
13
D
D
o
S
-
S
y
n
o
n
y
m
o
u
sI
P
_
F
l
o
o
d
0
.
9
9
9
6
8
8
6
0
.
9
9
9
8
7
5
4
0
.
9
9
9
7
8
2
14
D
D
o
S
-
TC
P
_
F
l
o
o
d
0
.
9
9
9
8
5
2
4
0
.
9
9
9
8
0
3
2
0
.
9
9
9
8
2
7
8
15
D
D
o
S
-
U
D
P
_
F
l
o
o
d
0
.
9
9
9
6
7
0
5
0
.
9
9
9
7
5
2
8
0
.
9
9
9
7
1
1
6
16
D
D
o
S
-
U
D
P
_
F
r
a
g
m
e
n
t
a
t
i
o
n
0
.
9
9
9
2
4
2
4
0
.
9
9
5
4
7
1
7
0
.
9
9
7
3
5
3
5
17
D
i
c
t
i
o
n
a
r
y
B
r
u
t
e
F
o
r
c
e
0
.
7
1
7
3
9
1
3
0
.
5
5
0
.
6
2
2
6
4
1
5
18
D
N
S
_
S
p
o
o
f
i
n
g
0
.
7
8
4
1
8
2
3
0
.
6
8
7
4
2
6
6
0
.
7
3
2
6
2
3
7
19
D
o
S
-
H
TTP_
F
l
o
o
d
0
.
9
8
7
1
3
8
3
0
.
9
8
7
1
3
8
3
0
.
9
8
7
1
3
8
3
20
D
o
S
-
S
Y
N
_
F
l
o
o
d
0
.
9
9
9
2
2
4
4
0
.
9
9
9
1
1
3
7
0
.
9
9
9
1
6
9
21
D
o
S
-
TC
P
_
F
l
o
o
d
0
.
9
9
9
6
6
4
5
0
.
9
9
9
7
4
8
4
0
.
9
9
9
7
0
6
4
22
D
o
S
-
U
D
P
_
F
l
o
o
d
0
.
9
9
9
7
3
0
.
9
9
9
5
2
7
7
0
.
9
9
9
6
2
8
8
23
M
i
r
a
i
-
g
r
e
e
t
h
_
f
l
o
o
d
0
.
9
9
9
1
0
1
3
0
.
9
9
8
6
5
2
6
0
.
9
9
8
8
7
6
9
24
M
i
r
a
i
-
g
r
e
i
p
_
f
l
o
o
d
0
.
9
9
8
8
3
3
8
0
.
9
9
8
5
4
2
7
0
.
9
9
8
6
8
8
2
25
M
i
r
a
i
-
u
d
p
p
l
a
i
n
0
.
9
9
9
0
3
5
4
0
.
9
9
9
7
5
8
7
0
.
9
9
9
3
9
6
9
26
M
I
TM
-
A
r
p
S
p
o
o
f
i
n
g
0
.
8
8
0
9
1
4
8
0
.
7
9
5
5
8
4
0
.
8
3
6
0
7
7
8
27
R
e
c
o
n
-
H
o
s
t
D
i
sc
o
v
e
r
y
0
.
8
4
2
8
8
3
5
0
.
7
8
6
2
0
6
9
0
.
8
1
3
5
5
9
3
28
R
e
c
o
n
-
O
S
S
c
a
n
0
.
7
7
6
5
0
4
3
0
.
6
0
0
8
8
6
9
0
.
6
7
7
5
29
R
e
c
o
n
-
P
i
n
g
S
w
e
e
p
1
0
.
1
1
1
1
1
1
1
0
.
2
30
R
e
c
o
n
-
P
o
r
t
S
c
a
n
0
.
7
7
5
4
3
8
6
0
.
5
9
4
0
8
6
0
.
6
7
2
7
5
4
9
31
S
q
l
I
n
j
e
c
t
i
o
n
0
.
5
5
5
5
5
5
6
0
.
3
3
3
3
3
3
3
0
.
4
1
6
6
6
6
7
32
U
p
l
o
a
d
i
n
g
_
A
t
t
a
c
k
1
0
.
7
1
4
2
8
5
7
0
.
8
3
3
3
3
3
3
33
V
u
l
n
e
r
a
b
i
l
i
t
y
S
c
a
n
0
.
9
8
3
0
5
0
8
1
0
.
9
9
1
4
5
3
34
X
S
S
0
.
8
3
3
3
3
3
3
0
.
3
5
7
1
4
2
9
0
.
5
I
t
is
n
o
tewo
r
th
y
th
at
2
0
attac
k
ty
p
es
co
n
s
id
er
ed
in
th
e
d
at
aset
h
av
e
b
ee
n
d
etec
ted
with
a
p
r
ec
is
io
n
s
co
r
e
o
f
9
0
%
an
d
ab
o
v
e,
a
f
ea
t
q
u
ite
v
ital
in
a
n
y
m
u
lticlas
s
class
if
icatio
n
task
.
Fig
u
r
e
3
d
em
o
n
s
tr
ates
clea
r
ly
th
e
p
er
f
o
r
m
a
n
ce
e
x
h
ib
ited
b
y
d
ec
is
io
n
ju
n
g
le.
Fig
u
r
e
4
d
ep
i
cts
th
e
p
r
ec
is
io
n
s
co
r
e
p
e
r
tain
in
g
to
th
e
3
4
attac
k
ty
p
es.
Fig
u
r
e
5
r
ep
r
esen
ts
th
e
r
ec
all
s
co
r
e
p
er
tain
in
g
to
th
e
3
4
attac
k
v
ec
to
r
s
.
T
h
e
d
ec
is
io
n
ju
n
g
le
alg
o
r
ith
m
m
ar
k
s
a
s
ig
n
if
ican
t
s
tep
f
o
r
war
d
in
th
e
f
ield
o
f
m
ac
h
in
e
lear
n
in
g
,
esp
ec
ially
wh
en
it
co
m
es
to
h
an
d
lin
g
m
u
lticlas
s
clas
s
if
icatio
n
task
s
.
Pre
cisi
o
n
is
b
ased
o
n
th
e
p
r
o
p
o
r
tio
n
o
f
ac
cu
r
ate
p
r
ed
ictio
n
s
m
ad
e
f
o
r
a
s
p
ec
if
ic
class
o
u
t
o
f
all
th
e
in
s
tan
ce
s
p
r
ed
icted
as
th
at
class
.
Sen
s
it
iv
ity
o
r
tr
u
e
p
o
s
itiv
e
r
ate
(
also
k
n
o
wn
as
r
ec
all)
is
u
s
ed
to
m
ea
s
u
r
e
th
e
p
r
o
p
o
r
tio
n
o
f
ac
tu
al
i
n
s
tan
ce
s
o
f
a
g
i
v
e
n
class
th
at
wer
e
ac
cu
r
ately
p
r
ed
icted
.
T
h
e
d
ec
is
io
n
ju
n
g
le
m
o
d
el
p
r
o
p
o
s
ed
in
th
is
r
esear
ch
wo
r
k
h
as
r
ep
o
r
ted
a
p
r
o
m
is
in
g
p
r
ec
is
io
n
s
co
r
e
f
o
r
t
h
r
ee
attac
k
ty
p
es
n
am
el
y
B
r
o
wser
-
h
ijack
in
g
,
R
ec
o
n
-
Pin
g
s
wee
p
an
d
Up
lo
a
d
in
g
-
attac
k
.
Hijack
ed
b
r
o
wser
s
h
av
e
th
e
ca
p
ab
ilit
y
to
ac
q
u
ir
e
s
en
s
itiv
e
in
f
o
r
m
atio
n
,
in
clu
d
in
g
lo
g
in
cr
e
d
en
tials
,
f
in
an
c
ial
d
ata,
o
r
p
r
iv
at
e
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
.
3
,
J
u
n
e
20
25
:
3
0
9
5
-
3
1
0
6
3102
co
m
m
u
n
icatio
n
s
an
d
th
u
s
ca
n
b
e
a
p
o
ten
tial
r
is
k
to
th
e
p
r
iv
ac
y
a
n
d
s
ec
u
r
ity
o
f
b
o
th
in
d
iv
i
d
u
als
an
d
o
r
g
an
izatio
n
s
.
R
ec
o
n
-
Pin
g
s
wee
p
is
m
o
r
e
o
f
ten
th
e
in
itial st
ep
f
o
r
a
wid
e
s
ca
le
cy
b
er
attac
k
.
C
y
b
er
cr
im
in
als
ca
n
m
ak
e
a
f
o
cu
s
ed
attac
k
,
if
th
e
y
f
in
d
th
e
a
ctiv
e
d
ev
ices
an
d
t
h
eir
I
P
ad
d
r
ess
es.
I
o
T
d
ev
ices
n
o
r
m
ally
s
u
p
p
o
r
t
f
ile
u
p
lo
ad
s
f
o
r
b
o
th
f
ir
m
war
e
u
p
d
ates
an
d
co
n
f
ig
u
r
atio
n
s
.
An
u
p
lo
a
d
in
g
attac
k
wh
ich
is
s
u
cc
ess
f
u
l
ca
n
af
f
ec
t
th
ese
d
ev
ices
an
d
attac
k
er
s
ca
n
ass
u
m
e
th
e
co
n
tr
o
l
o
f
th
em
.
T
h
er
ef
o
r
e,
th
e
in
tr
u
s
io
n
d
etec
tio
n
m
o
d
el
s
h
o
u
ld
p
o
s
s
ess
th
e
ca
p
ab
ilit
y
to
co
r
r
ec
tly
id
en
tif
y
th
ese
attac
k
s
.
A
f
ew
attac
k
ty
p
es
n
am
ely
R
ec
o
n
-
Oss
ca
n
,
s
q
lin
j
ec
tio
n
s
an
d
XSS
wer
e
n
o
t
c
o
r
r
ec
tly
id
e
n
tifie
d
b
y
th
e
p
r
o
p
o
s
ed
m
o
d
el
an
d
in
itiativ
es a
r
e
tak
en
in
th
is
r
eg
ar
d
to
ad
d
r
ess
th
e
is
s
u
es o
f
s
ca
lab
ilit
y
an
d
co
m
p
u
tatio
n
al
c
h
a
llen
g
es.
Fig
u
r
e
3
.
Per
f
o
r
m
an
c
e
o
f
t
h
e
m
o
d
el
co
n
ce
r
n
in
g
p
r
ec
is
io
n
,
r
ec
all
an
d
F1
-
s
co
r
e
Fig
u
r
e
4
.
Pre
cisi
o
n
s
co
r
e
o
f
9
0
% a
n
d
ab
o
v
e
f
o
r
2
0
attac
k
ty
p
es
Ba
c
k
d
o
o
r_
Mal
w
are
Ben
i
g
n
T
raffic
Bro
w
s
e
rHijac
k
i
n
g
Co
m
man
d
In
je
c
t
i
o
n
DD
o
S
-A
CK_
F
r
a
g
me
n
t
a
t
i
o
n
DD
o
S
-H
T
T
P_F
l
o
o
d
DD
o
S
-ICMP
_
F
l
o
o
d
DD
o
S
-ICMP
_
F
rag
me
n
tati
o
n
DD
o
S
-P
S
H
A
CK_
F
l
o
o
d
DD
o
S
-R
S
T
F
IN
F
l
o
o
d
DD
o
S
-S
l
o
w
L
o
ri
s
DD
o
S
-S
Y
N
_
F
l
o
o
d
DD
o
S
-S
yn
o
n
ymo
u
s
IP
_
F
l
o
o
d
DD
o
S
-T
CP
_
F
l
o
o
d
DD
o
S
-UD
P_
F
l
o
o
d
DD
o
S
-UD
P_
F
rag
me
n
tati
o
n
Di
c
ti
o
n
aryBru
te
F
o
rce
DN
S
_
S
p
o
o
fin
g
Do
S
-H
T
T
P_F
l
o
o
d
D
o
S
-S
Y
N
_
F
l
o
o
d
Do
S
-T
CP
_
F
l
o
o
d
Do
S
-UD
P_F
l
o
o
d
Mir
a
i
-g
re
e
th
_
f
l
o
o
d
M
i
r
a
i
-g
r
e
i
p
_
fl
o
o
d
Mi
r
a
i
-u
d
p
p
l
ai
n
MIT
M-A
rp
S
p
o
o
fin
g
R
e
c
o
n
-H
o
s
tD
i
s
c
o
ve
ry
R
e
c
o
n
-O
S
S
c
an
R
e
c
o
n
-P
i
n
g
S
w
e
e
p
R
e
c
o
n
-P
o
rt
S
c
an
S
q
l
In
je
c
t
i
o
n
Up
l
o
ad
i
n
g
_
A
tt
a
c
k
V
u
l
n
e
rab
i
l
i
tyS
c
an
XS
S
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Perf
o
rm
an
ce
E
v
alu
at
ion
Me
trics
Pre
c
i
s
i
o
n
R
e
c
all
F
1-s
c
o
re
0.86
0.88
0.9
0.92
0.94
0.96
0.98
1
1.02
Prec
is
ion
Att
ack
t
y
p
e
s
Perfo
r
man
c
e
o
f the
de
c
i
s
i
o
n
j
u
n
g
l
e
mo
d
e
l
w
r
t
p
r
e
c
i
s
i
o
n
fo
r
2
0
a
tta
c
k
ty
p
e
s
w
i
th
9
0
%
s
c
o
r
e
a
n
d
ab
o
v
e
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
E
xp
lo
r
in
g
th
e
effec
tiven
ess
o
f
mu
lticla
s
s
d
ec
is
io
n
ju
n
g
le
fo
r
in
tern
et
o
f th
in
g
s
s
ec
u
r
ity
(
S
mith
a
R
a
ja
g
o
p
a
l
)
3103
Fig
u
r
e
5
.
R
ec
all
s
co
r
e
o
f
9
0
%
an
d
ab
o
v
e
f
o
r
2
0
attac
k
ty
p
es
T
h
e
F1
-
s
co
r
e
ev
alu
ates
p
er
f
o
r
m
an
ce
f
o
r
ea
ch
ca
teg
o
r
y
s
ep
ar
ately
an
d
co
m
b
in
e
th
ese
in
t
o
a
s
in
g
le
m
etr
ic
to
en
s
u
r
e
th
at
m
in
o
r
it
y
class
es
ar
e
n
o
t
o
v
er
s
h
ad
o
wed
b
y
m
ajo
r
ity
class
es.
T
h
er
ef
o
r
e,
F1
-
s
co
r
e
is
a
cr
itical
ev
alu
atio
n
p
ar
am
eter
.
I
n
th
e
co
n
tex
t
o
f
m
u
lticlas
s
class
if
icatio
n
,
th
e
F1
-
s
co
r
e
g
iv
e
s
a
b
alan
ce
d
way
to
v
alid
ate
th
e
ef
f
ec
tiv
en
ess
o
f
t
h
e
m
o
d
el.
T
h
is
is
esp
ec
ially
u
s
ef
u
l
in
r
ea
l
-
tim
e
s
itu
atio
n
s
wh
er
e
it
is
h
ar
d
to
h
av
e
b
o
th
h
ig
h
p
r
ec
is
io
n
an
d
h
ig
h
r
ec
all
at
th
e
s
am
e
tim
e
[
3
2
]
,
[
3
3
]
.
As
d
is
cu
s
s
ed
in
[
3
4
]
,
F1
-
s
co
r
e
b
alan
ce
s
r
ec
all
an
d
p
r
ec
is
io
n
s
co
r
es
u
s
in
g
th
e
h
ar
m
o
n
ic
m
ea
n
.
T
h
e
p
r
o
p
o
s
ed
d
ec
is
io
n
ju
n
g
le
m
o
d
el
h
as
d
is
p
lay
ed
an
F1
-
s
co
r
e
o
f
0
.
9
a
n
d
ab
o
v
e
f
o
r
2
0
attac
k
ty
p
es a
s
p
r
esen
ted
in
Fig
u
r
e
6
.
Fig
u
r
e
6
.
F1
-
s
co
r
e
o
f
9
0
% a
n
d
ab
o
v
e
f
o
r
2
0
attac
k
t
y
p
es
0.86
0.88
0.9
0.92
0.94
0.96
0.98
1
1.02
Re
call
Att
ack
t
y
p
e
s
Per
for
m
an
c
e
o
f th
e
d
e
c
i
si
o
n
ju
n
g
l
e
m
o
d
e
l
wr
t
r
e
c
al
l
for
20
att
ac
k
typ
e
s w
i
th
90%
s
c
o
r
e
an
d
a
b
o
ve
0.86
0.88
0.9
0.92
0.94
0.96
0.98
1
1.02
F1
-
sco
r
e
A
tt
ac
k
typ
e
s
Per
for
m
a
nce
o
f
t
he
de
ci
sion
j
un
g
l
e
m
o
de
l
w
rt
F
1
-
scor
e
for
2
0
a
t
t
a
ck
t
y
pe
s
w
i
t
h
9
0
%
scor
e
a
nd
a
bo
v
e
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
.
3
,
J
u
n
e
20
25
:
3
0
9
5
-
3
1
0
6
3104
5.
CO
NCLU
SI
O
N
T
h
e
g
o
al
o
f
th
is
s
tu
d
y
was
to
estab
lis
h
th
e
im
p
o
r
tan
ce
o
f
m
u
lticlas
s
class
if
icat
io
n
f
o
r
n
etwo
r
k
in
tr
u
s
io
n
d
etec
tio
n
f
r
o
m
th
e
p
er
s
p
ec
tiv
e
o
f
d
ec
is
io
n
ju
n
g
le
.
T
h
e
p
r
o
p
o
s
ed
wo
r
k
af
f
ir
m
s
th
at
d
ec
is
io
n
ju
n
g
le
is
in
d
ee
d
a
s
o
p
h
is
ticated
alg
o
r
ith
m
to
d
etec
t v
ar
io
u
s
attac
k
p
atter
n
s
th
at
b
elo
n
g
to
d
if
f
e
r
en
t a
ttack
ca
teg
o
r
ies.
T
o
estab
lis
h
th
at
o
u
r
m
eth
o
d
w
o
r
k
s
well,
we
u
s
ed
th
e
C
I
C
I
O
T
2
0
2
3
d
ataset,
wh
ich
co
m
p
r
is
es
I
OT
in
tr
u
s
io
n
p
atter
n
s
.
W
h
en
we
test
ed
it
o
n
Azu
r
e
cl
o
u
d
,
we
f
o
u
n
d
t
h
at
th
e
d
ec
is
io
n
ju
n
g
le
alg
o
r
ith
m
wo
r
k
s
well
f
o
r
d
etec
tin
g
n
etwo
r
k
i
n
tr
u
s
io
n
s
,
esp
ec
ially
f
o
r
class
if
y
in
g
d
i
f
f
er
en
t
ty
p
es
o
f
attac
k
s
.
I
n
s
tead
o
f
u
s
in
g
s
in
g
le
s
y
s
tem
s
,
in
o
u
r
wo
r
k
,
we
m
ain
ly
f
o
cu
s
ed
o
n
h
o
w
f
ast
th
e
s
y
s
tem
r
u
n
s
an
d
p
er
f
o
r
m
ed
th
e
test
s
u
s
in
g
Mic
r
o
s
o
f
t
Azu
r
e
m
ac
h
in
e
le
ar
n
in
g
s
tu
d
i
o
(
MA
ML
S).
T
h
e
m
u
lticlas
s
d
ec
is
io
n
ju
n
g
le
p
r
ed
icto
r
t
o
o
k
o
n
e
m
in
u
te
an
d
f
if
ty
s
ec
o
n
d
s
t
o
ex
ec
u
te
s
u
cc
ess
f
u
lly
.
T
h
e
d
ec
is
io
n
ju
n
g
le
alg
o
r
it
h
m
p
o
s
s
ess
e
s
im
m
en
s
e
ca
p
ab
ilit
ies
to
d
elin
ea
te
an
d
c
o
u
n
ter
a
n
ex
p
an
s
iv
e
a
r
r
ay
o
f
s
ec
u
r
ity
r
is
k
s
.
I
t
also
ex
h
ib
its
a
g
o
o
d
p
e
r
f
o
r
m
an
ce
wh
ile
d
o
in
g
class
if
icatio
n
tas
k
s
th
at
ar
e
s
o
tech
n
ically
ch
allen
g
in
g
,
p
er
s
e.
T
h
e
p
r
o
p
o
s
ed
m
o
d
el
d
ev
el
o
p
e
d
u
s
in
g
d
ec
is
io
n
ju
n
g
le
alg
o
r
ith
m
p
er
f
o
m
ed
at
lo
w
co
m
p
u
tin
g
co
s
ts
h
en
ce
estab
lis
h
in
g
its
v
i
ab
ilit
y
in
r
ea
l
-
tim
e
p
r
o
tectio
n
o
f
I
o
T
a
p
p
licatio
n
s
.
T
h
e
p
r
o
p
o
s
ed
s
tu
d
y
also
s
u
g
g
ests
th
at
th
e
d
ec
is
io
n
ju
n
g
le
alg
o
r
ith
m
en
h
an
ce
s
th
e
ap
p
r
o
p
r
iate
d
etec
tio
n
o
f
attac
k
s
,
b
u
t
ev
en
m
o
r
e
im
p
o
r
tan
t,
it
d
ec
r
ea
s
es
th
e
n
u
m
b
er
o
f
f
alse
p
o
s
itiv
es
p
r
im
ar
ily
n
ee
d
ed
in
cr
itical
ap
p
licatio
n
s
s
u
ch
as n
etwo
r
k
i
n
tr
u
s
io
n
d
etec
tio
n
.
ACK
NO
WL
E
DG
M
E
N
T
S
W
e
th
an
k
Allian
ce
U
n
iv
er
s
ity
f
o
r
all
th
e
r
eso
u
r
ce
s
p
r
o
v
id
e
d
to
u
s
f
o
r
t
h
e
s
u
cc
ess
f
u
l
ex
ec
u
t
io
n
o
f
th
is
r
esear
ch
wo
r
k
.
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
Sm
ith
a
R
ajag
o
p
al
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
Ab
h
ik
Sar
k
ar
✓
✓
✓
✓
✓
✓
✓
✓
✓
Ven
k
at
Nar
ay
an
a
n
Ma
n
ju
n
ath
✓
✓
✓
✓
✓
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.
DATA AV
AI
L
AB
I
L
I
T
Y
T
h
e
d
ata
th
at
s
u
p
p
o
r
t
t
h
e
f
in
d
in
g
s
o
f
th
is
s
tu
d
y
ar
e
o
p
e
n
ly
av
ailab
le
i
n
UNB
a
t
h
ttp
s
://www.
u
n
b
.
ca
/cic/d
atasets
/io
td
ataset
-
2
0
2
3
.
h
tm
l
RE
F
E
R
E
NC
E
S
[
1
]
M
.
A
.
T
a
l
u
k
d
e
r
e
t
a
l
.
,
“
A
d
e
p
e
n
d
a
b
l
e
h
y
b
r
i
d
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
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
,
”
J
o
u
r
n
a
l
o
f
I
n
f
o
rm
a
t
i
o
n
S
e
c
u
r
i
t
y
a
n
d
A
p
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
7
2
,
p
.
1
0
3
4
0
5
,
F
e
b
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
j
i
sa
.
2
0
2
2
.
1
0
3
4
0
5
.
[
2
]
M
.
b
i
n
t
i
M
o
h
a
m
a
d
N
o
o
r
a
n
d
W
.
H
.
H
a
ssa
n
,
“
C
u
r
r
e
n
t
r
e
sea
r
c
h
o
n
i
n
t
e
r
n
e
t
o
f
t
h
i
n
g
s (I
o
T)
sec
u
r
i
t
y
:
a
s
u
r
v
e
y
,
”
C
o
m
p
u
t
e
r
N
e
t
w
o
rks
,
v
o
l
.
1
4
8
,
p
p
.
2
8
3
–
2
9
4
,
J
a
n
.
2
0
1
9
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
c
o
m
n
e
t
.
2
0
1
8
.
1
1
.
0
2
5
.
[
3
]
M
.
W
a
n
g
,
Y
.
L
u
,
a
n
d
J
.
Q
i
n
,
“
A
d
y
n
a
m
i
c
M
LP
-
b
a
se
d
D
D
o
S
a
t
t
a
c
k
d
e
t
e
c
t
i
o
n
me
t
h
o
d
u
si
n
g
f
e
a
t
u
r
e
se
l
e
c
t
i
o
n
a
n
d
f
e
e
d
b
a
c
k
,
”
C
o
m
p
u
t
e
rs
a
n
d
S
e
c
u
r
i
t
y
,
v
o
l
.
8
8
,
p
.
1
0
1
6
4
5
,
Ja
n
.
2
0
2
0
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
c
o
s
e
.
2
0
1
9
.
1
0
1
6
4
5
.
Evaluation Warning : The document was created with Spire.PDF for Python.