I
AE
S In
t
er
na
t
io
na
l J
o
urna
l o
f
Art
if
icia
l In
t
ellig
ence
(
I
J
-
AI
)
Vo
l.
1
5
,
No
.
4
,
A
u
g
u
s
t 2
0
2
6
,
p
p
.
3
7
1
2
~
3
7
2
1
I
SS
N:
2
2
5
2
-
8
9
3
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijai.v
15
.i
4
.
p
p
3
7
1
2
-
3
7
2
1
3712
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
i
.
ia
esco
r
e.
co
m
Feature
-
g
uided
tr
a
nsfo
rmer app
ro
a
ch f
o
r de
tecting
distribut
ed
denia
l of servi
ce
attacks
L
o
k
esh
wa
ra
n K
a
na
g
a
ra
j
1
,
R
a
g
ura
ma
n P
urus
ho
t
ha
m
a
n
2
,
Sa
t
hy
a
Su
bra
m
a
nia
n
3
,
Durg
a
Dev
i Sa
ra
v
a
na
n
4
,
K
o
m
a
l K
um
a
r
Na
pa
5
,
Co
rnelius
K
a
runa
k
a
ra
n
6
,
B
illa
M
a
nin
dh
a
r
7
1
D
e
p
a
r
t
me
n
t
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
a
n
d
E
n
g
i
n
e
e
r
i
n
g
,
S
c
h
o
o
l
o
f
C
o
m
p
u
t
i
n
g
,
V
e
l
T
e
c
h
R
a
n
g
a
r
a
j
a
n
D
r
.
S
a
g
u
n
t
h
a
l
a
R
&D
I
n
st
i
t
u
t
e
o
f
S
c
i
e
n
c
e
a
n
d
T
e
c
h
n
o
l
o
g
y
,
A
v
a
d
i
,
I
n
d
i
a
2
D
e
p
a
r
t
me
n
t
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
&
En
g
i
n
e
e
r
i
n
g
(
A
r
t
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
)
,
M
a
d
a
n
a
p
a
l
l
e
I
n
st
i
t
u
t
e
o
f
Te
c
h
n
o
l
o
g
y
& Sci
e
n
c
e
,
M
a
d
a
n
a
p
a
l
l
e
,
I
n
d
i
a
3
D
e
p
a
r
t
me
n
t
o
f
El
e
c
t
r
o
n
i
c
s a
n
d
C
o
m
mu
n
i
c
a
t
i
o
n
En
g
i
n
e
e
r
i
n
g
,
S
.
A
.
E
n
g
i
n
e
e
r
i
n
g
C
o
l
l
e
g
e
,
C
h
e
n
n
a
i
,
I
n
d
i
a
4
D
e
p
a
r
t
me
n
t
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
a
n
d
E
n
g
i
n
e
e
r
i
n
g
,
V
e
l
T
e
c
h
H
i
g
h
T
e
c
h
D
r
.
R
a
n
g
a
r
a
j
a
n
D
r
.
S
a
k
u
n
t
h
a
l
a
E
n
g
i
n
e
e
r
i
n
g
C
o
l
l
e
g
e
,
A
v
a
d
i
,
I
n
d
i
a
5
D
e
p
a
r
t
me
n
t
o
f
A
r
t
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
a
n
d
D
a
t
a
S
c
i
e
n
c
e
,
S
a
v
e
e
t
h
a
E
n
g
i
n
e
e
r
i
n
g
C
o
l
l
e
g
e
,
C
h
e
n
n
a
i
,
I
n
d
i
a
.
6
D
e
p
a
r
t
me
n
t
o
f
A
r
t
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
a
n
d
D
a
t
a
S
c
i
e
n
c
e
,
R
M
K
C
o
l
l
e
g
e
o
f
E
n
g
i
n
e
e
r
i
n
g
a
n
d
Te
c
h
n
o
l
o
g
y
,
Th
i
r
u
v
a
l
l
u
r
,
I
n
d
i
a
7
D
e
p
a
r
t
me
n
t
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
a
n
d
E
n
g
i
n
e
e
r
i
n
g
,
K
o
n
e
r
u
L
a
k
sh
m
a
i
a
h
E
d
u
c
a
t
i
o
n
F
o
u
n
d
a
t
i
o
n
,
V
a
d
d
e
sw
a
r
a
m
,
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
Sep
1
1
,
2
0
2
5
R
ev
is
ed
J
u
l 1
1
,
2
0
2
6
Acc
ep
ted
J
u
l 2
1
,
2
0
2
6
Distrib
u
te
d
d
e
n
ial
o
f
se
r
v
ice
(DD
o
S
)
a
tt
a
c
k
s
c
o
n
ti
n
u
e
s
to
p
o
se
se
rio
u
s
risk
s
to
m
o
d
e
r
n
n
e
two
r
k
s,
with
t
h
e
ir
g
ro
wi
n
g
i
n
ten
sit
y
m
a
k
in
g
e
a
rly
d
e
tec
ti
o
n
b
o
t
h
c
rit
ica
l
a
n
d
c
h
a
ll
e
n
g
in
g
.
C
o
n
v
e
n
t
io
n
a
l
m
a
c
h
in
e
lea
rn
in
g
(M
L)
m
o
d
e
ls
o
ften
stru
g
g
le
with
t
h
e
n
o
n
li
n
e
a
r
a
n
d
h
i
g
h
l
y
d
y
n
a
m
ic
n
a
t
u
re
o
f
a
tt
a
c
k
traffic,
wh
ich
m
o
ti
v
a
tes
th
e
u
se
o
f
a
d
v
a
n
c
e
d
a
rc
h
it
e
c
tu
re
s.
In
th
is stu
d
y
i
n
v
e
stig
a
te a
tran
sfo
rm
e
r
-
b
a
se
d
c
las
sifier
f
o
r
DD
o
S
d
e
tec
ti
o
n
o
n
th
e
CIC
-
DD
o
S
2
0
1
9
d
a
tas
e
t.
Th
e
wo
r
k
flo
w
in
c
l
u
d
e
d
p
re
p
ro
c
e
ss
in
g
,
fe
a
tu
re
sc
a
li
n
g
,
a
n
d
d
o
m
a
in
-
g
u
i
d
e
d
fe
a
tu
re
se
lec
ti
o
n
.
Lo
g
is
ti
c
re
g
re
ss
io
n
(LR)
wa
s
e
m
p
lo
y
e
d
a
s
a
b
a
se
li
n
e
,
a
c
h
iev
i
n
g
9
2
.
1
%
a
c
c
u
ra
c
y
a
n
d
F
1
-
sc
o
re
o
f
0
.
9
0
,
th
e
re
b
y
re
v
e
a
li
n
g
th
e
li
m
it
a
ti
o
n
s
o
f
li
n
e
a
r
m
o
d
e
l
s.
Th
e
tran
sf
o
rm
e
r
,
a
fter
h
y
p
e
r
p
a
ra
m
e
ter
tu
n
i
n
g
a
n
d
5
-
fo
l
d
c
r
o
ss
-
v
a
li
d
a
ti
o
n
,
re
a
c
h
e
d
a
n
a
v
e
ra
g
e
a
c
c
u
ra
c
y
o
f
9
9
.
9
5
%
with
p
re
c
isio
n
,
re
c
a
ll
,
a
n
d
F
1
-
sc
o
re
s
a
ll
a
b
o
v
e
9
9
.
9
%
.
T
h
e
m
o
d
e
l
d
e
m
o
n
stra
ted
sta
b
le
c
o
n
v
e
r
g
e
n
c
e
a
n
d
g
e
n
e
ra
li
z
a
ti
o
n
a
c
ro
ss
f
o
l
d
s.
T
h
e
se
re
su
lt
s
h
ig
h
li
g
h
t
t
h
e
stre
n
g
t
h
o
f
a
tt
e
n
ti
o
n
m
e
c
h
a
n
ism
s
in
c
a
p
tu
ri
n
g
fe
a
tu
re
d
e
p
e
n
d
e
n
c
ies
,
wh
il
e
a
lso
p
o
in
ti
n
g
t
o
fu
t
u
re
d
irec
ti
o
n
s
su
c
h
a
s
re
a
l
-
ti
m
e
d
e
p
lo
y
m
e
n
t,
e
x
p
lai
n
a
b
il
it
y
,
a
n
d
r
e
sili
e
n
c
e
to
z
e
ro
-
d
a
y
a
tt
a
c
k
s.
K
ey
w
o
r
d
s
:
Dee
p
lear
n
in
g
f
o
r
c
y
b
er
s
ec
u
r
ity
Dis
tr
ib
u
ted
d
en
ial
o
f
s
er
v
ice
I
n
tr
u
s
io
n
d
etec
tio
n
s
y
s
tem
s
Netwo
r
k
tr
af
f
ic
class
if
icatio
n
T
r
an
s
f
o
r
m
e
r
m
o
d
el
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
:
L
o
k
esh
war
an
Kan
a
g
ar
aj
Dep
ar
tm
en
t o
f
C
o
m
p
u
ter
Scie
n
ce
an
d
E
n
g
in
ee
r
in
g
,
Sch
o
o
l o
f
C
o
m
p
u
tin
g
Vel
T
ec
h
R
an
g
ar
ajan
Dr
.
Sag
u
n
th
ala
R
&
D
I
n
s
titu
te
o
f
Scie
n
ce
an
d
T
ec
h
n
o
l
o
g
y
No
.
4
2
,
Av
a
d
i
-
Vel
T
ec
h
R
o
ad
,
Vel
Nag
ar
,
Av
ad
i,
C
h
en
n
ai,
T
am
il Na
d
u
6
0
0
0
6
2
,
I
n
d
ia
E
m
ail: k
.
lo
k
esh
war
an
@
g
m
ail.
co
m
1.
I
NT
RO
D
UCT
I
O
N
Dis
tr
ib
u
ted
d
en
ial
o
f
s
er
v
ice
(
DDo
S)
attac
k
s
r
em
ain
o
n
e
o
f
th
e
m
o
s
t
d
is
r
u
p
tiv
e
th
r
ea
ts
t
o
m
o
d
er
n
n
etwo
r
k
ed
s
y
s
tem
s
,
ca
p
ab
le
o
f
o
v
er
wh
elm
in
g
tar
g
et
r
eso
u
r
c
es
an
d
ca
u
s
in
g
s
ev
er
e
f
in
an
cia
l,
r
ep
u
tatio
n
al,
an
d
o
p
er
atio
n
al
d
am
a
g
e.
T
h
e
wid
esp
r
ea
d
av
ailab
ilit
y
o
f
b
o
tn
et
s
,
r
ef
lectio
n
an
d
am
p
lific
atio
n
tech
n
iq
u
es,
an
d
attac
k
-
as
-
a
-
s
er
v
ice
p
latf
o
r
m
s
h
as
s
ig
n
if
ican
tly
lo
wer
ed
th
e
b
ar
r
ier
to
lau
n
ch
in
g
lar
g
e
-
s
ca
le
attac
k
s
,
m
ak
in
g
tim
ely
d
etec
tio
n
in
cr
ea
s
in
g
ly
ch
allen
g
in
g
[
1
]
,
[
2
]
.
T
r
ad
itio
n
al
d
ef
en
s
e
m
ec
h
a
n
is
m
s
s
u
ch
as
s
ig
n
atu
r
e
-
b
ased
in
tr
u
s
io
n
d
etec
tio
n
s
y
s
tem
s
,
r
u
le
-
b
ased
f
i
r
ewa
lls
,
an
d
t
h
r
esh
o
ld
-
d
r
iv
en
an
o
m
aly
d
et
ec
to
r
s
ar
e
ef
f
ec
tiv
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
F
ea
tu
r
e
-
g
u
id
e
d
tr
a
n
s
fo
r
mer a
p
p
r
o
a
ch
f
o
r
d
etec
tin
g
d
is
tr
ib
u
ted
d
en
ia
l
o
f
…
(
Lo
ke
s
h
w
a
r
a
n
K
a
n
a
g
a
r
a
j
)
3713
ag
ain
s
t
k
n
o
wn
attac
k
p
atter
n
s
b
u
t
s
u
f
f
e
r
f
r
o
m
lim
ited
ad
a
p
t
ab
ilit
y
an
d
p
o
o
r
s
ca
lab
ilit
y
in
d
y
n
am
ic
n
etwo
r
k
en
v
ir
o
n
m
en
ts
,
o
f
ten
lead
in
g
t
o
h
ig
h
f
alse a
lar
m
r
ates o
r
m
is
s
ed
d
etec
tio
n
s
[
3
]
,
[
4
]
.
T
o
ad
d
r
ess
th
ese
lim
itatio
n
s
,
m
ac
h
in
e
lear
n
in
g
(
ML
)
tech
n
i
q
u
es h
av
e
b
ee
n
wid
ely
ad
o
p
te
d
f
o
r
DDo
S
d
etec
tio
n
.
C
lass
ical
m
o
d
els,
in
clu
d
in
g
l
o
g
is
tic
r
eg
r
ess
io
n
(
L
R
)
,
d
ec
is
io
n
tr
ee
s
(
DT
)
,
r
an
d
o
m
f
o
r
ests
(
R
F)
,
an
d
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
es
(
SV
M)
,
ca
n
lear
n
d
is
cr
im
in
ativ
e
t
r
af
f
ic
p
atter
n
s
b
u
t
r
el
y
h
ea
v
il
y
o
n
m
an
u
al
f
ea
tu
r
e
en
g
in
ee
r
in
g
an
d
s
h
allo
w
d
ec
is
io
n
b
o
u
n
d
ar
ies,
lim
itin
g
th
eir
ab
ilit
y
to
ca
p
tu
r
e
co
m
p
le
x
n
o
n
lin
ea
r
r
elatio
n
s
h
ip
s
in
lar
g
e
-
s
ca
le
attac
k
tr
af
f
ic
[
5
]
,
[
6
]
.
Dee
p
lear
n
i
n
g
m
o
d
els
s
u
ch
as
co
n
v
o
l
u
tio
n
al
an
d
r
ec
u
r
r
en
t
n
e
u
r
al
n
etwo
r
k
s
(
R
NNs)
im
p
r
o
v
e
r
ep
r
esen
tatio
n
lear
n
in
g
an
d
tem
p
o
r
al
m
o
d
elin
g
b
u
t
ar
e
o
f
t
en
co
m
p
u
tatio
n
ally
ex
p
en
s
iv
e,
s
en
s
itiv
e
to
h
y
p
er
p
ar
am
eter
ch
o
ices,
an
d
p
r
o
n
e
to
o
v
e
r
f
itti
n
g
wh
e
n
ap
p
lie
d
t
o
h
ig
h
-
d
im
en
s
io
n
al
tab
u
lar
f
lo
w
d
ata
[
7
]
,
[
8
]
.
Atten
tio
n
m
ec
h
an
is
m
s
,
p
ar
ticu
lar
ly
tr
an
s
f
o
r
m
e
r
ar
ch
itectu
r
es
,
p
r
o
v
id
e
a
p
r
o
m
is
in
g
alter
n
ativ
e
b
y
s
elec
tiv
ely
f
o
cu
s
in
g
o
n
in
f
o
r
m
ativ
e
f
ea
tu
r
e
in
ter
ac
tio
n
s
t
h
r
o
u
g
h
s
elf
-
atten
tio
n
,
with
o
u
t
r
e
ly
in
g
o
n
r
ec
u
r
r
e
n
ce
o
r
co
n
v
o
lu
tio
n
.
T
h
is
ca
p
ab
ilit
y
en
ab
les
ef
f
ec
tiv
e
m
o
d
elin
g
o
f
co
m
p
lex
d
e
p
en
d
e
n
cies
am
o
n
g
h
eter
o
g
e
n
eo
u
s
tr
af
f
ic
f
ea
t
u
r
es
an
d
h
as
g
ain
e
d
in
cr
ea
s
in
g
atten
tio
n
in
cy
b
e
r
s
ec
u
r
ity
r
esear
c
h
[
9
]
.
T
h
e
ef
f
ec
tiv
en
ess
o
f
s
u
ch
lear
n
in
g
-
b
ased
ap
p
r
o
ac
h
es,
h
o
wev
er
,
d
e
p
en
d
s
cr
itically
o
n
th
e
av
ailab
ilit
y
o
f
r
ea
lis
tic
an
d
d
iv
er
s
e
d
atasets
.
T
h
e
C
I
C
-
DDo
S2
0
1
9
d
ataset
ad
d
r
ess
es
m
an
y
lim
itatio
n
s
o
f
ea
r
lier
b
e
n
ch
m
a
r
k
s
b
y
in
co
r
p
o
r
atin
g
m
u
ltip
l
e
co
n
tem
p
o
r
ar
y
attac
k
f
am
ilies
,
r
ea
lis
tic
b
en
ig
n
tr
af
f
ic,
an
d
co
m
p
r
eh
e
n
s
iv
e
f
lo
w
-
lev
el
f
ea
tu
r
es,
m
ak
in
g
it
a
wid
ely
ad
o
p
te
d
b
en
c
h
m
ar
k
f
o
r
DDo
S d
etec
tio
n
[
1
0
]
.
Mo
t
iv
ate
d
b
y
t
h
es
e
d
e
v
el
o
p
m
en
ts
,
t
h
is
wo
r
k
p
r
o
p
o
s
es
a
f
ea
t
u
r
e
-
g
u
i
d
e
d
tr
an
s
f
o
r
m
e
r
-
b
ase
d
f
r
am
ew
o
r
k
f
o
r
DD
o
S
d
ete
cti
o
n
.
B
y
in
t
eg
r
a
ti
n
g
d
o
m
a
in
-
i
n
f
o
r
m
e
d
f
e
atu
r
e
s
ele
cti
o
n
wit
h
a
c
o
m
p
ac
t
at
te
n
ti
o
n
-
b
ase
d
ar
ch
ite
ct
u
r
e,
t
h
e
p
r
o
p
o
s
e
d
a
p
p
r
o
ac
h
ai
m
s
t
o
ac
h
i
ev
e
h
ig
h
d
e
tect
io
n
a
cc
u
r
ac
y
.
I
t
als
o
m
a
in
t
ain
s
co
m
p
u
tat
io
n
al
ef
f
ic
ie
n
c
y
a
n
d
r
o
b
u
s
t
n
ess
,
s
u
p
p
o
r
t
e
d
b
y
r
e
p
r
o
d
u
c
ib
le
e
v
a
lu
at
io
n
a
n
d
f
ai
r
b
asel
in
e
co
m
p
a
r
is
o
n
.
2.
RE
L
AT
E
D
WO
RK
T
h
is
s
ec
tio
n
r
ev
iews
ex
is
tin
g
r
esear
ch
o
n
DDo
S
d
etec
tio
n
,
co
v
er
in
g
s
u
r
v
e
y
s
tu
d
ies,
class
ical
an
d
d
ee
p
lear
n
in
g
ap
p
r
o
ac
h
es,
f
e
atu
r
e
s
elec
tio
n
tech
n
iq
u
es,
a
n
d
r
ec
en
t
t
r
a
n
s
f
o
r
m
e
r
-
b
ased
in
tr
u
s
io
n
d
etec
tio
n
s
y
s
tem
s
.
Sev
er
al
s
u
r
v
ey
s
p
r
o
v
id
e
co
m
p
r
eh
e
n
s
iv
e
o
v
e
r
v
iews
o
f
DDo
S
attac
k
t
ax
o
n
o
m
ies,
tr
af
f
ic
ch
ar
ac
ter
is
tics
,
an
d
m
itig
atio
n
s
tr
ateg
ies
.
T
h
ese
s
u
r
v
ey
s
h
ig
h
lig
h
t
th
e
ev
o
lu
tio
n
f
r
o
m
v
o
lu
m
etr
ic
f
lo
o
d
s
to
s
o
p
h
is
ticated
m
u
lti
-
v
ec
to
r
an
d
ap
p
licatio
n
-
lay
er
attac
k
s
a
n
d
u
n
d
e
r
s
co
r
in
g
th
e
n
ee
d
f
o
r
ad
ap
tiv
e
d
etec
tio
n
m
ec
h
an
is
m
s
[
1
1
]
–
[
1
3
]
.
E
ar
ly
lear
n
in
g
-
b
ased
DDo
S
d
etec
tio
n
m
eth
o
d
s
p
r
im
ar
ily
e
m
p
lo
y
ed
class
ical
ML
m
o
d
els
s
u
ch
as
LR
,
SVM
,
an
d
RF
tr
ain
ed
o
n
h
an
d
cr
af
ted
f
lo
w
-
lev
el
f
ea
tu
r
e
s
.
W
h
ile
th
ese
ap
p
r
o
ac
h
es
im
p
r
o
v
e
d
d
etec
tio
n
ac
cu
r
ac
y
o
v
er
r
u
le
-
b
ased
s
y
s
tem
s
,
th
eir
d
ep
en
d
en
ce
o
n
m
an
u
al
f
ea
tu
r
e
en
g
in
ee
r
in
g
a
n
d
lim
ited
ab
ilit
y
to
m
o
d
el
n
o
n
lin
ea
r
tr
af
f
ic
b
e
h
a
v
io
r
r
estricte
d
g
en
e
r
aliza
tio
n
u
n
d
er
d
y
n
a
m
ic
n
etwo
r
k
c
o
n
d
itio
n
s
[
1
4
]
–
[
1
6
]
.
Hy
b
r
id
an
d
en
s
em
b
le
-
b
ased
m
o
d
els
wer
e
s
u
b
s
eq
u
en
tly
in
tr
o
d
u
ce
d
to
e
n
h
an
ce
r
o
b
u
s
tn
ess
b
y
co
m
b
in
i
n
g
m
u
ltip
le
class
if
ier
s
o
r
s
tatis
tical
co
m
p
o
n
en
ts
;
h
o
wev
er
,
th
es
e
m
eth
o
d
s
o
f
te
n
in
c
u
r
h
ig
h
e
r
co
m
p
u
tatio
n
al
co
s
ts
an
d
r
ed
u
ce
d
in
te
r
p
r
etab
ilit
y
,
li
m
itin
g
r
ea
l
-
tim
e
ap
p
licab
ilit
y
[
1
7
]
,
[
1
8
]
.
W
ith
th
e
ad
o
p
tio
n
o
f
d
ee
p
lear
n
in
g
,
co
n
v
o
lu
tio
n
al
an
d
R
N
N
s
wer
e
ap
p
lied
to
i
n
tr
u
s
io
n
d
etec
tio
n
to
ca
p
tu
r
e
s
p
atial
a
n
d
tem
p
o
r
al
d
ep
en
d
en
cies
in
n
etwo
r
k
tr
af
f
ic.
Alth
o
u
g
h
C
NN
-
an
d
R
NN
-
b
ased
m
o
d
els
d
em
o
n
s
tr
ated
im
p
r
o
v
ed
p
er
f
o
r
m
an
ce
,
th
ey
ar
e
f
r
eq
u
en
tl
y
s
en
s
itiv
e
to
d
ata
r
e
p
r
esen
t
atio
n
an
d
p
r
o
n
e
to
o
v
er
f
itti
n
g
wh
en
ap
p
lied
to
h
ig
h
-
d
im
en
s
io
n
al
tab
u
lar
f
e
atu
r
es
with
o
u
t
s
u
f
f
icien
t
r
e
g
u
lar
izatio
n
o
r
d
ata
v
o
lu
m
e
[
1
9
]
–
[
2
1
]
.
C
o
n
s
eq
u
en
tly
,
f
ea
tu
r
e
s
elec
tio
n
h
as
em
er
g
ed
as
a
cr
itical
co
m
p
o
n
en
t,
with
r
ec
en
t
s
tu
d
ies
ex
p
lo
r
in
g
f
ilter
-
b
ased
,
wr
ap
p
e
r
-
b
ased
,
an
d
h
y
b
r
i
d
tech
n
iq
u
e
s
to
r
ed
u
ce
r
ed
u
n
d
an
cy
,
im
p
r
o
v
e
g
e
n
e
r
a
l
i
z
a
t
i
o
n
,
a
n
d
e
n
h
a
n
c
e
c
o
m
p
u
t
a
ti
o
n
a
l
e
f
f
i
c
i
e
n
c
y
f
o
r
b
o
t
h
c
l
as
s
i
c
a
l
a
n
d
d
e
e
p
l
e
a
r
n
i
n
g
m
o
d
e
l
s
[
2
2
]
–
[
2
4
]
.
Mo
r
e
r
ec
en
tly
,
atten
tio
n
-
b
ased
m
o
d
els
an
d
tr
an
s
f
o
r
m
er
ar
ch
itectu
r
es
h
a
v
e
g
ain
ed
p
r
o
m
in
e
n
ce
in
n
etwo
r
k
s
ec
u
r
ity
r
esear
c
h
.
B
y
lev
er
ag
in
g
s
elf
-
atten
tio
n
m
ec
h
an
is
m
s
,
tr
an
s
f
o
r
m
er
s
ca
n
ef
f
ec
tiv
ely
m
o
d
el
co
m
p
lex
f
ea
tu
r
e
in
ter
ac
tio
n
s
an
d
h
a
v
e
ac
h
iev
ed
n
ea
r
s
tate
-
of
-
th
e
-
ar
t
p
e
r
f
o
r
m
an
ce
o
n
b
en
ch
m
ar
k
in
tr
u
s
io
n
d
etec
tio
n
d
atasets
[
2
5
]
–
[
2
8
]
.
Ho
wev
er
,
m
an
y
ex
is
tin
g
ap
p
r
o
ac
h
es
r
ely
o
n
h
ig
h
-
d
im
en
s
io
n
al
f
ea
tu
r
e
s
ets
an
d
d
ee
p
a
r
ch
itectu
r
es,
in
cr
ea
s
in
g
o
v
er
f
itti
n
g
r
is
k
an
d
lim
itin
g
d
ep
lo
y
a
b
ilit
y
.
Mo
r
e
o
v
er
,
ev
alu
atio
n
s
ar
e
o
f
ten
r
estricte
d
to
s
in
g
le
d
atasets
with
o
u
t
ex
p
licitly
ad
d
r
ess
in
g
f
ea
tu
r
e
r
ed
u
n
d
an
cy
o
r
m
o
d
el
co
m
p
ac
tn
ess
.
T
h
ese
lim
ita
tio
n
s
m
o
tiv
ate
th
e
d
ev
elo
p
m
e
n
t
o
f
f
ea
tu
r
e
-
g
u
id
e
d
,
lig
h
tweig
h
t
t
r
an
s
f
o
r
m
er
-
b
ased
f
r
am
ew
o
r
k
s
th
at
b
alan
ce
a
cc
u
r
ac
y
,
ef
f
icien
cy
,
an
d
g
e
n
er
aliza
tio
n
,
f
o
r
m
in
g
th
e
b
asis
o
f
th
e
ap
p
r
o
ac
h
p
r
o
p
o
s
ed
in
th
is
wo
r
k
[
2
9
]
,
[
3
0
]
.
3.
M
E
T
H
O
D
T
h
e
m
eth
o
d
o
lo
g
y
o
f
th
is
r
ese
ar
ch
h
as
b
ee
n
o
r
g
an
ized
in
a
s
tep
-
by
-
s
tep
m
an
n
er
s
o
t
h
at
e
ac
h
s
tag
e
ca
n
b
e
u
n
d
er
s
to
o
d
in
co
n
tin
u
ity
with
th
e
o
th
e
r
s
.
T
h
e
s
tr
u
ctu
r
e
b
r
o
ad
l
y
f
o
llo
ws
s
ev
e
n
p
h
ases
:
d
ataset
ac
q
u
is
itio
n
,
d
ata
p
r
e
p
r
o
ce
s
s
in
g
,
f
ea
tu
r
e
s
elec
tio
n
,
b
aselin
e
m
o
d
el
co
n
s
tr
u
ctio
n
,
tr
an
s
f
o
r
m
er
m
o
d
el
d
esig
n
,
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
1
5
,
No
.
4
,
Au
g
u
s
t 2
0
2
6
:
3
7
1
2
-
3
7
2
1
3714
tr
ain
in
g
with
h
y
p
e
r
p
ar
am
eter
tu
n
in
g
,
an
d
ev
alu
atio
n
with
ap
p
r
o
p
r
iate
p
er
f
o
r
m
a
n
ce
m
etr
ics.
E
ac
h
p
h
ase
i
s
d
escr
ib
ed
in
d
etail
as f
o
llo
ws
.
3
.
1
.
Da
t
a
s
et
des
cr
iptio
n a
nd
prepro
ce
s
s
i
ng
T
h
e
C
I
C
-
DDo
S2
0
1
9
d
ataset
was
u
s
ed
f
o
r
e
v
alu
atio
n
d
u
e
t
o
its
wid
e
a
d
o
p
ti
o
n
a
n
d
co
m
p
r
eh
en
s
iv
e
co
v
er
ag
e
o
f
DDo
S
s
ce
n
ar
io
s
.
I
n
tr
o
d
u
ce
d
b
y
Sh
ar
af
ald
i
n
et
a
l.
[
1
0
]
,
in
clu
d
es
b
en
ig
n
tr
a
f
f
i
c
an
d
attac
k
s
f
r
o
m
1
1
DDo
S
f
am
ilies
s
u
ch
as
s
y
n
ch
r
o
n
ize
(
SYN
)
fl
o
o
d
,
d
o
m
ain
n
am
e
s
y
s
tem
(
DNS
)
,
n
et
wo
r
k
tim
e
p
r
o
to
co
l
(
NT
P
)
,
s
im
p
le
s
er
v
ice
d
is
co
v
er
y
p
r
o
to
co
l
(
SS
DP
)
,
an
d
W
eb
DDo
S,
g
en
e
r
ated
in
an
is
o
lated
test
b
ed
with
r
ea
lis
tic
b
ac
k
g
r
o
u
n
d
tr
af
f
ic.
E
ig
h
ty
f
lo
w
-
le
v
el
s
tatis
tical
f
ea
tu
r
es
wer
e
ex
tr
ac
ted
u
s
in
g
C
I
C
Flo
wM
eter
.
Pre
p
r
o
ce
s
s
in
g
in
v
o
l
v
ed
r
em
o
v
in
g
d
u
p
licates,
im
p
u
tin
g
m
is
s
in
g
o
r
ze
r
o
v
alu
es
with
th
e
m
e
d
ian
,
an
d
ap
p
ly
i
n
g
m
in
–
m
ax
n
o
r
m
aliza
tio
n
.
Stra
tifie
d
s
am
p
lin
g
was u
s
ed
to
s
p
li
t th
e
d
ata
in
to
8
0
% tr
ain
in
g
an
d
2
0
% testi
n
g
s
ets,
with
5
-
f
o
ld
cr
o
s
s
-
v
alid
atio
n
e
m
p
lo
y
ed
f
o
r
r
o
b
u
s
t e
v
alu
atio
n
.
3
.
2
.
F
e
a
t
ure
s
elec
t
io
n a
nd
ex
plo
ra
t
io
n
F
e
a
t
u
r
e
s
e
l
e
c
t
io
n
i
s
c
r
i
t
ic
a
l
f
o
r
e
f
f
ic
i
e
n
t
i
n
tr
u
s
i
o
n
d
e
t
e
c
t
io
n
,
a
s
u
s
i
n
g
a
l
l
av
a
i
l
ab
l
e
a
tt
r
ib
u
t
e
s
o
f
t
en
i
n
t
r
o
d
u
c
e
s
r
e
d
u
n
d
a
n
cy
a
n
d
n
o
i
s
e
.
I
n
t
h
i
s
s
tu
d
y
,
d
o
m
a
i
n
-
g
u
i
d
ed
r
e
c
o
m
m
en
d
a
t
io
n
s
f
r
o
m
S
h
a
r
a
f
a
ld
i
n
e
t
a
l.
[
1
0
]
w
er
e
c
o
m
b
in
e
d
w
i
t
h
e
x
p
l
o
r
a
t
o
r
y
a
n
a
ly
s
i
s
o
f
th
e
C
I
C
-
D
D
o
S
2
0
1
9
d
a
ta
s
e
t
to
id
e
n
t
i
f
y
d
i
s
c
r
im
i
n
a
t
iv
e
f
e
a
t
u
r
e
s
.
A
s
s
h
o
wn
in
F
ig
u
r
e
1
,
a
l
th
o
u
g
h
D
Do
S
s
am
p
l
e
s
d
o
m
i
n
a
t
e,
b
o
th
c
l
a
s
s
e
s
a
r
e
s
u
f
f
i
c
i
en
t
l
y
r
e
p
r
e
s
e
n
t
ed
.
S
t
a
t
i
s
t
i
c
a
l
an
a
l
y
s
i
s
h
i
g
h
l
ig
h
te
d
p
a
ck
e
t
l
e
n
g
t
h
–
b
a
s
ed
a
t
tr
i
b
u
t
e
s
a
s
s
tr
o
n
g
in
d
i
c
a
to
r
s
o
f
a
t
t
a
c
k
b
e
h
av
i
o
r
,
wh
i
c
h
w
a
s
f
u
r
th
e
r
co
n
f
ir
m
ed
th
r
o
u
g
h
f
e
a
t
u
r
e
im
p
o
r
t
an
c
e
an
a
l
y
s
i
s
,
a
s
s
h
o
wn
in
T
a
b
l
e
1
.
B
a
s
ed
o
n
t
h
e
s
e
f
i
n
d
i
n
g
s
,
a
r
e
f
in
e
d
s
u
b
s
e
t
o
f
1
1
f
ea
t
u
r
e
s
w
a
s
s
e
l
e
c
te
d
an
d
u
s
e
d
c
o
n
s
i
s
t
e
n
t
ly
f
o
r
b
o
t
h
b
a
s
e
l
in
e
a
n
d
t
r
a
n
s
f
o
r
m
e
r
m
o
d
e
l
s
.
Fig
u
r
e
1
.
C
lass
d
is
tr
ib
u
tio
n
T
ab
le
1
.
Featu
r
e
s
et
(
o
n
ly
s
ele
cted
)
R
a
n
k
F
e
a
t
u
r
e
n
a
me
C
a
t
e
g
o
r
y
I
mp
o
r
t
a
n
c
e
(
r
e
l
a
t
i
v
e
)
D
e
c
i
s
i
o
n
1
B
w
d
p
a
c
k
e
t
l
e
n
g
t
h
ma
x
P
a
c
k
e
t
l
e
n
g
t
h
s
t
a
t
i
s
t
i
c
s
H
i
g
h
S
e
l
e
c
t
e
d
2
F
w
d
p
a
c
k
e
t
l
e
n
g
t
h
mi
n
P
a
c
k
e
t
l
e
n
g
t
h
s
t
a
t
i
s
t
i
c
s
H
i
g
h
S
e
l
e
c
t
e
d
3
B
w
d
p
a
c
k
e
t
s
l
e
n
g
t
h
t
o
t
a
l
P
a
c
k
e
t
l
e
n
g
t
h
s
t
a
t
i
s
t
i
c
s
H
i
g
h
S
e
l
e
c
t
e
d
4
F
w
d
p
a
c
k
e
t
s
l
e
n
g
t
h
t
o
t
a
l
P
a
c
k
e
t
l
e
n
g
t
h
s
t
a
t
i
s
t
i
c
s
H
i
g
h
S
e
l
e
c
t
e
d
5
F
l
o
w
p
a
c
k
e
t
s
/
s
Tr
a
f
f
i
c
i
n
t
e
n
s
i
t
y
H
i
g
h
S
e
l
e
c
t
e
d
6
To
t
a
l
F
w
d
p
a
c
k
e
ts
Tr
a
f
f
i
c
i
n
t
e
n
s
i
t
y
H
i
g
h
S
e
l
e
c
t
e
d
7
A
C
K
f
l
a
g
c
o
u
n
t
H
e
a
d
e
r
/
f
l
a
g
f
i
e
l
d
H
i
g
h
S
e
l
e
c
t
e
d
8
I
d
l
e
me
a
n
F
l
o
w
t
i
m
i
n
g
M
e
d
i
u
m
S
e
l
e
c
t
e
d
9
P
a
c
k
e
t
l
e
n
g
t
h
m
e
a
n
P
a
c
k
e
t
l
e
n
g
t
h
s
t
a
t
i
s
t
i
c
s
M
e
d
i
u
m
S
e
l
e
c
t
e
d
10
B
w
d
p
a
c
k
e
t
l
e
n
g
t
h
mi
n
P
a
c
k
e
t
l
e
n
g
t
h
s
t
a
t
i
s
t
i
c
s
M
e
d
i
u
m
S
e
l
e
c
t
e
d
11
F
l
o
w
d
u
r
a
t
i
o
n
F
l
o
w
t
i
m
i
n
g
M
e
d
i
u
m
S
e
l
e
c
t
e
d
Her
e,
f
ea
tu
r
e
-
g
u
id
ed
d
etec
tio
n
r
ef
er
s
to
d
o
m
ain
-
in
f
o
r
m
e
d
f
ea
tu
r
e
s
elec
tio
n
b
ef
o
r
e
tr
ain
i
n
g
,
r
ath
e
r
th
an
n
ai
v
ely
s
u
p
p
ly
in
g
all
ex
t
r
ac
ted
f
ea
t
u
r
es.
W
h
ile
s
elf
-
att
en
tio
n
ca
n
in
te
r
n
ally
r
ewe
ig
h
t
in
p
u
ts
,
p
r
o
v
id
in
g
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
F
ea
tu
r
e
-
g
u
id
e
d
tr
a
n
s
fo
r
mer a
p
p
r
o
a
ch
f
o
r
d
etec
tin
g
d
is
tr
ib
u
ted
d
en
ia
l
o
f
…
(
Lo
ke
s
h
w
a
r
a
n
K
a
n
a
g
a
r
a
j
)
3715
all
f
ea
tu
r
es
in
cr
ea
s
es
lear
n
in
g
co
m
p
lex
ity
an
d
o
v
er
f
itti
n
g
r
i
s
k
.
B
y
r
etain
in
g
o
n
ly
in
f
o
r
m
a
tiv
e
attr
ib
u
tes,
th
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
r
ed
u
ce
s
n
o
is
e
an
d
e
n
ab
les
th
e
tr
a
n
s
f
o
r
m
er
to
f
o
cu
s
o
n
m
ea
n
in
g
f
u
l
f
ea
tu
r
e
in
ter
ac
tio
n
s
,
lead
in
g
to
im
p
r
o
v
e
d
ac
cu
r
ac
y
an
d
g
e
n
er
aliza
tio
n
,
as d
em
o
n
s
tr
ated
b
y
th
e
ab
latio
n
s
tu
d
y
.
3
.
3
.
B
a
s
eline
m
o
dels
A
L
R
cla
s
s
if
ier
wa
s
im
p
lem
e
n
ted
as
th
e
b
aselin
e.
L
R
is
w
i
d
ely
u
s
ed
in
in
tr
u
s
io
n
d
etec
tio
n
r
esear
ch
d
u
e
to
its
s
im
p
lici
ty
an
d
in
ter
p
r
etab
ilit
y
.
I
t
ass
u
m
es
a
lin
ea
r
d
ec
is
io
n
b
o
u
n
d
ar
y
,
m
ak
in
g
it
a
n
atu
r
al
b
en
ch
m
ar
k
a
g
ain
s
t
wh
ich
to
co
m
p
ar
e
m
o
r
e
ad
v
a
n
ce
d
a
r
ch
itectu
r
es.
I
n
th
is
s
tu
d
y
,
L
R
w
as
tr
ain
ed
with
th
e
lib
lin
ea
r
s
o
lv
er
an
d
b
alan
ce
d
class
weig
h
ts
.
T
h
e
m
o
d
el
p
r
o
v
id
e
d
in
s
ig
h
t
in
to
h
o
w
well
s
im
p
le
lin
ea
r
s
ep
ar
atio
n
co
u
l
d
p
er
f
o
r
m
o
n
C
I
C
-
DDo
S2
0
1
9
.
3
.
4
.
T
ra
ns
f
o
rm
er
mo
del des
ig
n
T
h
e
p
r
o
p
o
s
ed
tr
an
s
f
o
r
m
er
wa
s
d
esig
n
ed
f
o
r
DDo
S
d
etec
ti
o
n
o
n
ta
b
u
lar
n
etwo
r
k
f
l
o
w
d
ata.
E
ac
h
f
lo
w
was
r
ep
r
esen
ted
as
a
f
ix
e
d
-
len
g
th
s
eq
u
en
ce
o
f
1
0
-
tim
e
s
tep
s
with
1
6
f
ea
tu
r
es
p
er
s
tep
an
d
p
r
o
jecte
d
in
to
a
2
5
6
-
d
im
en
s
io
n
al
em
b
e
d
d
in
g
s
p
ac
e.
A
s
in
g
le
tr
an
s
f
o
r
m
e
r
b
lo
ck
with
two
-
h
ea
d
s
elf
-
att
en
tio
n
was
u
s
ed
to
ca
p
tu
r
e
f
ea
t
u
r
e
d
ep
en
d
en
cies,
f
o
llo
wed
b
y
lay
e
r
n
o
r
m
aliza
tio
n
,
r
esid
u
al
co
n
n
ec
tio
n
s
,
a
co
m
p
ac
t
f
ee
d
-
f
o
r
war
d
lay
er
(
d
im
e
n
s
io
n
2
)
,
an
d
d
r
o
p
o
u
t (
0
.
3
)
f
o
r
r
e
g
u
lar
izatio
n
.
Glo
b
al
av
er
a
g
e
p
o
o
lin
g
was
ap
p
lied
ac
r
o
s
s
th
e
s
eq
u
en
c
e
d
im
en
s
io
n
to
o
b
tain
a
f
i
x
ed
-
len
g
t
h
r
ep
r
esen
tatio
n
,
wh
ich
was
t
h
en
p
ass
ed
to
a
lig
h
tweig
h
t
m
u
ltil
ay
er
p
er
ce
p
tr
o
n
(
M
L
P
)
with
1
2
8
a
n
d
6
4
u
n
its
an
d
a
s
ig
m
o
id
o
u
tp
u
t f
o
r
b
i
n
ar
y
class
if
icatio
n
.
T
h
e
o
v
er
all
ar
ch
itectu
r
e
is
illu
s
tr
ated
in
Fig
u
r
e
2
.
T
h
is
co
m
p
ac
t
d
esig
n
b
alan
ce
s
ex
p
r
ess
iv
en
ess
an
d
e
f
f
icien
cy
,
en
ab
lin
g
ac
c
u
r
ate
DDo
S
d
etec
tio
n
with
lo
w
co
m
p
u
tatio
n
al
o
v
er
h
ea
d
.
Fig
u
r
e
2
.
Pro
p
o
s
ed
tr
a
n
s
f
o
r
m
e
r
m
o
d
el
a
r
ch
itectu
r
e
3
.
5
.
H
y
perpa
ra
m
e
t
er
t
un
ing
Hy
p
er
p
ar
a
m
eter
tu
n
in
g
was
p
er
f
o
r
m
e
d
u
s
in
g
Ker
as
t
u
n
er
w
ith
r
an
d
o
m
s
ea
r
ch
,
o
p
tim
izin
g
v
alid
atio
n
ac
cu
r
ac
y
an
d
v
alid
ated
v
ia
5
-
f
o
ld
cr
o
s
s
-
v
alid
atio
n
.
T
h
e
s
ea
r
ch
co
v
er
ed
atten
tio
n
h
ea
d
s
ize
,
n
u
m
b
er
o
f
h
ea
d
s
,
f
ee
d
-
f
o
r
war
d
d
im
en
s
io
n
,
tr
an
s
f
o
r
m
er
b
l
o
ck
s
,
d
r
o
p
o
u
t,
a
n
d
ML
P
u
n
its
.
T
h
e
c
h
o
s
en
c
o
n
f
ig
u
r
atio
n
(
T
ab
le
2
)
h
ea
d
s
ize
2
5
6
,
two
h
ea
d
s
,
f
e
ed
-
f
o
r
war
d
d
im
e
n
s
io
n
2
,
o
n
e
tr
an
s
f
o
r
m
er
b
lo
ck
,
ML
P
[
1
2
8
,
6
4
]
,
an
d
d
r
o
p
o
u
t
0
.
3
—
r
ef
lects
a
co
m
p
ac
t
d
esi
g
n
tailo
r
e
d
f
o
r
tab
u
lar
f
lo
w
f
ea
tu
r
es.
Dee
p
er
o
r
wid
er
m
o
d
els
s
h
o
wed
n
o
s
ig
n
if
ican
t
g
ain
s
an
d
in
cr
ea
s
e
d
o
v
er
f
itti
n
g
r
is
k
,
m
ak
in
g
th
e
s
h
allo
w,
r
eg
u
lar
ized
ar
c
h
itectu
r
e
an
ef
f
icien
t
a
n
d
g
en
er
aliza
b
le
ch
o
ice.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
1
5
,
No
.
4
,
Au
g
u
s
t 2
0
2
6
:
3
7
1
2
-
3
7
2
1
3716
T
ab
le
2
.
Hy
p
er
p
ar
a
m
eter
co
n
f
ig
u
r
atio
n
H
y
p
e
r
p
a
r
a
me
t
e
r
S
e
l
e
c
t
e
d
v
a
l
u
e
R
a
t
i
o
n
a
l
e
H
e
a
d
si
z
e
2
5
6
La
r
g
e
e
m
b
e
d
d
i
n
g
d
i
m
e
n
s
i
o
n
f
o
r
f
e
a
t
u
r
e
i
n
t
e
r
a
c
t
i
o
n
s
N
u
m
h
e
a
d
s
2
C
a
p
t
u
r
e
s s
u
f
f
i
c
i
e
n
t
a
t
t
e
n
t
i
o
n
d
i
v
e
r
si
t
y
w
i
t
h
o
u
t
h
i
g
h
c
o
st
F
e
e
d
-
f
o
r
w
a
r
d
d
i
me
n
si
o
n
2
C
o
m
p
a
c
t
l
a
y
e
r
,
r
e
d
u
c
e
s
o
v
e
r
f
i
t
t
i
n
g
r
i
s
k
N
u
m
t
r
a
n
sf
o
r
mer
b
l
o
c
k
s
1
S
h
a
l
l
o
w
b
u
t
e
f
f
e
c
t
i
v
e
,
e
n
a
b
l
e
d
f
a
st
e
r
c
o
n
v
e
r
g
e
n
c
e
M
LP
u
n
i
ts
[
1
2
8
,
6
4
]
B
a
l
a
n
c
e
d
d
e
n
se
l
a
y
e
r
s fo
r
c
l
a
ssi
f
i
c
a
t
i
o
n
D
r
o
p
o
u
t
0
.
3
R
e
g
u
l
a
r
i
z
a
t
i
o
n
t
o
i
m
p
r
o
v
e
g
e
n
e
r
a
l
i
z
a
t
i
o
n
3
.
6
.
T
ra
ini
ng
a
nd
ev
a
lua
t
io
n pro
t
o
co
l
T
h
e
tr
an
s
f
o
r
m
e
r
an
d
LR
m
o
d
els
wer
e
tr
ain
ed
u
s
in
g
s
u
p
e
r
v
is
ed
lear
n
in
g
o
n
t
h
e
C
I
C
-
DDo
S2
0
1
9
d
ataset
with
an
id
en
tical
f
ea
tu
r
e
s
et
to
en
s
u
r
e
f
air
co
m
p
ar
i
s
o
n
.
T
h
e
tr
an
s
f
o
r
m
er
was
o
p
t
im
ized
us
in
g
Ad
am
with
a
lear
n
in
g
r
ate
o
f
0
.
0
0
1
an
d
b
i
n
ar
y
cr
o
s
s
-
en
tr
o
p
y
lo
s
s
,
tr
ain
ed
with
a
b
atch
s
ize
o
f
6
4
f
o
r
u
p
to
5
0
e
p
o
ch
s
,
an
d
r
eg
u
lar
ized
u
s
in
g
ea
r
ly
s
to
p
p
in
g
b
ased
o
n
v
alid
atio
n
lo
s
s
.
Mo
d
el
r
o
b
u
s
tn
ess
was
en
s
u
r
ed
th
r
o
u
g
h
5
-
f
o
ld
c
r
o
s
s
-
v
alid
atio
n
,
with
f
in
al
r
esu
lts
r
ep
o
r
ted
a
s
th
e
av
er
ag
e
ac
r
o
s
s
all
f
o
ld
s
.
Per
f
o
r
m
an
ce
was
ev
alu
ate
d
u
s
in
g
ac
cu
r
ac
y
,
p
r
ec
is
io
n
,
r
e
ca
ll,
an
d
F1
-
s
co
r
e
to
ac
c
o
u
n
t
f
o
r
class
im
b
alan
ce
an
d
asy
m
m
et
r
ic
er
r
o
r
co
s
ts
.
C
o
n
f
u
s
io
n
m
at
r
ices,
r
ec
eiv
er
o
p
er
atin
g
ch
ar
ac
ter
is
tic
(
R
O
C
)
-
ar
ea
u
n
d
er
th
e
cu
r
v
e
(
AUC
)
,
an
d
p
r
ec
is
io
n
–
r
ec
all
cu
r
v
es
wer
e
al
s
o
an
aly
ze
d
to
ass
ess
m
is
clas
s
if
icatio
n
b
e
h
av
io
r
.
T
h
ese
m
etr
ics f
u
r
th
er
d
em
o
n
s
tr
ate
th
e
d
is
cr
im
in
ativ
e
ca
p
a
b
ilit
y
u
n
d
er
im
b
alan
ce
d
tr
af
f
ic
c
o
n
d
itio
n
s
.
4.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
4
.
1
.
B
a
s
eline
m
o
del per
f
o
r
ma
nce
L
R
was
ad
o
p
ted
as
th
e
b
aselin
e
class
if
ier
.
I
t
d
eliv
er
ed
r
ea
s
o
n
ab
le
p
er
f
o
r
m
a
n
ce
,
ac
h
iev
in
g
an
ac
cu
r
ac
y
o
f
9
2
.
1
%,
a
n
F1
-
s
co
r
e
o
f
0
.
9
0
,
an
d
a
R
OC
-
AUC
o
f
0
.
9
1
.
T
h
ese
r
esu
lts
in
d
icate
th
at
ce
r
tain
attac
k
p
atter
n
s
in
C
I
C
-
DDo
S2
0
1
9
ar
e
lin
ea
r
ly
s
ep
a
r
ab
le.
H
o
wev
er
,
an
aly
s
is
o
f
t
h
e
co
n
f
u
s
io
n
m
at
r
ix
s
h
o
wed
th
at
L
R
s
tr
u
g
g
led
to
d
etec
t
co
m
p
lex
tr
af
f
ic,
p
ar
ticu
lar
ly
MSSQL
am
p
lific
atio
n
an
d
NetBIOS
f
lo
o
d
s
,
wh
er
e
th
e
s
ep
ar
atio
n
b
etwe
en
b
en
ig
n
an
d
m
alicio
u
s
f
lo
ws is
in
h
er
en
tl
y
n
o
n
li
n
ea
r
.
4
.
2
.
F
e
a
t
ure
s
elec
t
io
n a
bla
t
i
o
n
An
ab
latio
n
s
tu
d
y
was p
er
f
o
r
m
ed
to
ass
ess
th
e
im
p
ac
t o
f
f
e
atu
r
e
s
elec
tio
n
u
s
in
g
th
r
ee
c
o
n
f
ig
u
r
atio
n
s
:
all
f
ea
tu
r
es,
th
e
p
r
o
p
o
s
ed
f
ea
tu
r
e
-
g
u
i
d
ed
s
u
b
s
et,
an
d
a
r
an
d
o
m
ly
s
elec
ted
s
u
b
s
et
o
f
eq
u
al
s
ize.
All
ex
p
er
im
en
ts
u
s
ed
t
h
e
s
am
e
t
r
an
s
f
o
r
m
er
ar
ch
itectu
r
e
an
d
5
-
f
o
ld
cr
o
s
s
-
v
alid
atio
n
.
As
s
h
o
wn
in
T
ab
le
3
,
th
e
f
ea
tu
r
e
-
g
u
i
d
ed
s
u
b
s
et
ac
h
iev
es
th
e
h
ig
h
est
ac
cu
r
ac
y
an
d
F1
-
s
co
r
e
with
s
u
b
s
tan
tially
f
ewe
r
f
ea
tu
r
es.
T
h
e
all
-
f
ea
tu
r
e
co
n
f
ig
u
r
atio
n
s
h
o
ws
s
lig
h
tly
wea
k
er
g
e
n
er
aliza
tio
n
d
u
e
to
r
ed
u
n
d
an
c
y
,
wh
il
e
th
e
r
a
n
d
o
m
s
u
b
s
et
p
er
f
o
r
m
s
n
o
tab
ly
wo
r
s
e,
co
n
f
i
r
m
in
g
th
e
b
en
ef
it
o
f
in
f
o
r
m
ed
f
ea
tu
r
e
s
elec
tio
n
.
T
ab
le
3
.
Featu
r
e
s
elec
tio
n
a
b
la
tio
n
r
esu
lts
F
e
a
t
u
r
e
s
e
t
F
e
a
t
u
r
e
s
A
c
c
u
r
a
c
y
(
%)
F1
-
sc
o
r
e
(
%)
A
l
l
f
e
a
t
u
r
e
s
80
9
9
.
7
2
9
9
.
7
0
F
e
a
t
u
r
e
-
g
u
i
d
e
d
(
p
r
o
p
o
se
d
)
11
9
9
.
9
5
9
9
.
9
6
R
a
n
d
o
m s
u
b
se
t
11
9
8
.
4
0
9
8
.
1
0
4
.
3
.
T
ra
ns
f
o
rm
er
mo
del per
f
o
rm
a
nce
T
h
e
p
r
o
p
o
s
ed
tr
an
s
f
o
r
m
er
cl
ass
if
ier
,
in
co
r
p
o
r
atin
g
d
en
s
e
p
r
o
jectio
n
s
,
m
u
lti
-
h
ea
d
s
elf
-
a
tten
tio
n
,
a
co
m
p
ac
t
f
ee
d
-
f
o
r
war
d
la
y
er
,
d
r
o
p
o
u
t,
an
d
g
lo
b
al
av
er
a
g
e
p
o
o
lin
g
,
s
ig
n
if
ican
tly
o
u
t
p
er
f
o
r
m
ed
th
e
LR
b
aselin
e.
Usi
n
g
th
e
tu
n
ed
h
y
p
er
p
ar
am
eter
s
,
th
e
m
o
d
el
ac
h
ie
v
ed
co
n
s
is
ten
tly
h
ig
h
p
er
f
o
r
m
an
ce
u
n
d
er
5
-
f
o
ld
cr
o
s
s
-
v
alid
atio
n
,
with
an
a
v
er
a
g
e
ac
cu
r
ac
y
o
f
9
9
.
9
5
%,
p
r
ec
is
io
n
o
f
9
9
.
9
7
%,
r
ec
all
o
f
9
9
.
9
5
%,
an
d
F1
-
s
co
r
e
o
f
9
9
.
9
6
%,
in
d
icatin
g
s
tr
o
n
g
g
en
er
aliza
tio
n
ac
r
o
s
s
f
o
ld
s
.
T
h
e
co
n
f
u
s
io
n
m
atr
i
x
in
Fig
u
r
e
3
s
h
o
ws
th
at
m
is
class
if
icatio
n
s
wer
e
n
eg
lig
ib
le
f
o
r
b
o
th
b
en
i
g
n
an
d
attac
k
f
lo
ws.
I
n
ad
d
itio
n
,
th
e
tr
ain
in
g
cu
r
v
e
an
d
v
alid
atio
n
cu
r
v
e
as
in
Fig
u
r
e
4
d
em
o
n
s
tr
ate
s
tab
le
co
n
v
er
g
e
n
ce
with
clo
s
ely
alig
n
ed
ac
c
u
r
a
cy
an
d
l
o
s
s
tr
en
d
s
,
co
n
f
ir
m
in
g
th
at
th
e
c
o
m
p
ac
t tr
an
s
f
o
r
m
er
d
esig
n
ac
h
iev
es
n
ea
r
-
p
er
f
ec
t
d
etec
tio
n
with
o
u
t o
v
er
f
itti
n
g
.
T
h
e
R
O
C
c
u
r
v
e
(
F
i
g
u
r
e
5
)
f
u
r
t
h
e
r
v
a
l
i
d
at
e
s
t
h
i
s
p
e
r
f
o
r
m
a
n
c
e
,
w
i
th
th
e
A
U
C
a
p
p
r
o
a
ch
i
n
g
1
.
0
.
L
i
k
e
w
i
s
e
,
th
e
p
r
e
c
i
s
io
n
-
r
e
c
al
l
c
u
r
v
e
(
F
ig
u
r
e
6
)
d
e
m
o
n
s
tr
a
t
e
s
t
h
a
t
th
e
tr
a
n
s
f
o
r
m
e
r
s
u
s
t
a
i
n
s
b
o
t
h
h
i
g
h
r
e
c
a
l
l
an
d
h
ig
h
p
r
ec
i
s
i
o
n
,
ev
e
n
u
n
d
e
r
i
m
b
a
l
an
c
ed
c
o
n
d
it
i
o
n
s
.
T
o
g
e
th
e
r
,
t
h
e
s
e
r
e
s
u
l
ts
c
o
n
f
ir
m
t
h
a
t
t
h
e
t
r
a
n
s
f
o
r
m
e
r
o
f
f
e
r
s
h
i
g
h
l
y
r
e
l
ia
b
l
e
d
e
t
e
c
t
io
n
o
f
D
D
o
S
t
r
a
f
f
i
c
,
o
u
t
p
er
f
o
r
m
i
n
g
c
l
a
s
s
i
c
a
l
b
a
s
e
l
i
n
e
s
a
cr
o
s
s
a
l
l
e
v
a
l
u
a
t
i
o
n
p
e
r
s
p
ec
t
i
v
e
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
F
ea
tu
r
e
-
g
u
id
e
d
tr
a
n
s
fo
r
mer a
p
p
r
o
a
ch
f
o
r
d
etec
tin
g
d
is
tr
ib
u
ted
d
en
ia
l
o
f
…
(
Lo
ke
s
h
w
a
r
a
n
K
a
n
a
g
a
r
a
j
)
3717
Fig
u
r
e
3
.
C
o
n
f
u
s
io
n
m
atr
i
x
Fig
u
r
e
4
.
T
r
ain
in
g
a
n
d
v
alid
atio
n
cu
r
v
es
Fig
u
r
e
5
.
R
OC
cu
r
v
e
Fig
u
r
e
6
.
Pre
cisi
o
n
-
r
ec
all
cu
r
v
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
1
5
,
No
.
4
,
Au
g
u
s
t 2
0
2
6
:
3
7
1
2
-
3
7
2
1
3718
4
.
4
.
Co
m
pa
ra
t
iv
e
ev
a
lua
t
io
n
T
h
e
co
m
p
ar
ativ
e
an
al
y
s
is
h
ig
h
lig
h
ts
th
e
clea
r
ad
v
an
tag
e
o
f
th
e
tr
an
s
f
o
r
m
e
r
m
o
d
el
o
v
er
th
e
LR
b
aselin
e
f
o
r
DDo
S
d
etec
tio
n
.
LR
,
tr
ain
ed
o
n
th
e
s
elec
ted
1
1
f
ea
tu
r
es,
ac
h
iev
ed
an
ac
cu
r
ac
y
o
f
9
2
.
1
%
with
an
F1
-
s
co
r
e
o
f
0
.
9
0
,
in
d
icatin
g
lim
ited
ab
ilit
y
to
m
o
d
el
th
e
n
o
n
lin
ea
r
b
o
u
n
d
ar
ies
p
r
esen
t
in
co
m
p
lex
attac
k
s
ce
n
ar
io
s
.
I
n
c
o
n
tr
ast,
th
e
tr
a
n
s
f
o
r
m
er
d
em
o
n
s
tr
ated
co
n
s
is
ten
tly
h
ig
h
an
d
s
tab
le
p
er
f
o
r
m
an
ce
ac
r
o
s
s
5
-
f
o
ld
cr
o
s
s
-
v
alid
atio
n
,
as
s
h
o
wn
b
y
th
e
f
o
ld
-
wis
e
r
esu
lts
in
T
ab
le
4
.
T
h
e
av
er
a
g
e
p
er
f
o
r
m
a
n
ce
s
u
m
m
ar
ized
in
T
ab
le
5
(
ac
c
u
r
ac
y
=
9
9
.
9
5
%,
p
r
ec
is
io
n
=
9
9
.
9
7
%,
r
ec
all
=
9
9
.
9
5
%,
an
d
F1
-
s
co
r
e
=
9
9
.
9
6
%)
d
em
o
n
s
tr
ates
a
s
u
b
s
tan
tial
im
p
r
o
v
em
e
n
t
o
v
e
r
th
e
b
aselin
e
w
h
ile
m
ain
ta
in
in
g
a
well
-
b
alan
ce
d
tr
ad
e
-
o
f
f
b
etwe
en
f
alse
p
o
s
itiv
es
an
d
f
alse
n
e
g
ativ
es.
T
h
ese
co
n
s
is
ten
tly
h
ig
h
an
d
b
alan
ce
d
m
etr
ics
co
n
f
ir
m
t
h
e
e
f
f
ec
tiv
en
ess
o
f
th
e
p
r
o
p
o
s
ed
tr
an
s
f
o
r
m
er
ar
c
h
itectu
r
e,
p
ar
ticu
lar
l
y
its
s
elf
-
atten
tio
n
m
ec
h
a
n
is
m
,
in
c
ap
tu
r
in
g
c
o
m
p
lex
d
ep
en
d
e
n
cies a
n
d
d
is
cr
im
in
ati
v
e
p
atter
n
s
with
in
n
etwo
r
k
tr
a
f
f
ic
f
o
r
r
eliab
le
DDo
S d
etec
tio
n
.
T
ab
le
4
.
5
-
f
o
ld
cr
o
s
s
-
v
alid
atio
n
r
esu
lts
F
o
l
d
A
c
c
u
r
a
c
y
P
r
e
c
i
s
i
o
n
(
D
D
o
S
)
R
e
c
a
l
l
(
D
D
o
S
)
F1
-
sc
o
r
e
(
D
D
o
S
)
1
0
.
9
9
9
4
0
.
9
9
9
8
0
.
9
9
9
2
0
.
9
9
9
5
2
0
.
9
9
9
5
0
.
9
9
9
9
0
.
9
9
9
4
0
.
9
9
9
6
3
0
.
9
9
9
5
0
.
9
9
9
4
0
.
9
9
9
8
0
.
9
9
9
6
4
0
.
9
9
9
5
0
.
9
9
9
8
0
.
9
9
9
4
0
.
9
9
9
6
5
0
.
9
9
9
5
0
.
9
9
9
4
0
.
9
9
9
9
0
.
9
9
9
6
A
v
g
.
0
.
9
9
9
5
0
.
9
9
9
7
0
.
9
9
9
5
0
.
9
9
9
6
T
ab
le
5
.
B
aselin
e
an
d
tr
a
n
s
f
o
r
m
er
m
o
d
el
c
o
m
p
a
r
is
o
n
M
o
d
e
l
A
c
c
u
r
a
c
y
P
r
e
c
i
s
i
o
n
R
e
c
a
l
l
F1
-
s
c
o
r
e
LR
0
.
9
2
1
0
0
.
9
1
0
0
0
.
9
0
0
0
0
.
9
0
0
0
Tr
a
n
sf
o
r
mer (
A
v
g
.
C
V
)
0
.
9
9
9
5
0
.
9
9
9
7
0
.
9
9
9
5
0
.
9
9
9
6
4
.
5
.
Co
m
pu
t
a
t
io
na
l
c
o
m
plex
it
y
a
nd
ef
f
iciency
T
h
e
p
r
o
p
o
s
ed
tr
an
s
f
o
r
m
er
m
o
d
el
was
in
ten
tio
n
ally
d
esig
n
ed
to
b
e
lig
h
tweig
h
t
to
s
u
p
p
o
r
t
p
r
ac
tical
d
ep
lo
y
m
e
n
t.
Du
e
to
its
co
m
p
a
ct
ar
ch
itectu
r
e
(
s
in
g
le
tr
a
n
s
f
o
r
m
er
b
lo
ck
an
d
r
ed
u
ce
d
f
ee
d
-
f
o
r
war
d
d
im
e
n
s
io
n
)
,
tr
ain
in
g
co
n
v
er
g
ed
with
in
a
s
m
all
n
u
m
b
er
o
f
ep
o
c
h
s
u
n
d
er
5
-
f
o
ld
c
r
o
s
s
-
v
alid
atio
n
.
I
n
f
er
en
ce
in
v
o
lv
es
a
s
in
g
le
f
o
r
war
d
p
ass
th
r
o
u
g
h
a
s
h
allo
w
atten
tio
n
m
o
d
u
le
an
d
a
co
m
p
ac
t
ML
P,
r
esu
ltin
g
in
lo
w
laten
cy
an
d
m
o
d
est
m
em
o
r
y
u
s
ag
e.
C
o
m
p
ar
ed
to
d
ee
p
er
tr
an
s
f
o
r
m
er
v
a
r
ian
ts
,
th
e
p
r
o
p
o
s
ed
m
o
d
el
s
ig
n
if
ican
tly
r
ed
u
ce
s
co
m
p
u
tatio
n
al
o
v
er
h
ea
d
wh
ile
m
ain
tain
in
g
h
ig
h
d
etec
tio
n
p
er
f
o
r
m
a
n
ce
,
m
ak
in
g
it
s
u
itab
l
e
f
o
r
n
ea
r
r
ea
l
-
tim
e
in
tr
u
s
io
n
d
etec
tio
n
s
ce
n
ar
io
s
.
4
.
6
.
St
a
t
is
t
ica
l
s
ig
nifica
nce
a
nd
co
nfidence
a
na
ly
s
is
T
o
ass
ess
th
e
s
tab
ilit
y
o
f
th
e
r
ep
o
r
ted
p
er
f
o
r
m
an
ce
,
we
co
m
p
u
ted
9
5
%
co
n
f
id
e
n
ce
in
ter
v
als
f
o
r
th
e
ev
alu
atio
n
m
etr
ics
ac
r
o
s
s
th
e
5
-
f
o
ld
c
r
o
s
s
-
v
alid
atio
n
.
T
h
e
n
ar
r
o
w
co
n
f
id
en
ce
b
o
u
n
d
s
o
b
s
er
v
ed
f
o
r
ac
cu
r
ac
y
an
d
F1
-
s
co
r
e
in
d
icate
lo
w
v
ar
ian
ce
ac
r
o
s
s
f
o
ld
s
,
co
n
f
i
r
m
in
g
th
at
th
e
n
ea
r
-
p
er
f
ec
t
p
er
f
o
r
m
an
ce
o
f
th
e
p
r
o
p
o
s
ed
tr
an
s
f
o
r
m
er
is
co
n
s
is
ten
t
an
d
n
o
t
attr
ib
u
tab
le
to
r
a
n
d
o
m
d
ata
p
a
r
titi
o
n
in
g
.
T
h
is
f
u
r
th
er
s
u
p
p
o
r
ts
th
e
m
o
d
el'
s
g
en
er
aliza
tio
n
ca
p
ab
il
ity
an
d
allev
iates c
o
n
ce
r
n
s
r
eg
ar
d
in
g
o
v
er
f
itti
n
g
.
4
.
7
.
Crit
ica
l
dis
cus
s
io
n a
nd
l
im
it
a
t
io
ns
T
h
e
r
esu
lts
r
ev
ea
l
a
s
u
b
s
tan
ti
al
p
er
f
o
r
m
an
ce
g
a
p
b
etwe
en
LR
an
d
th
e
p
r
o
p
o
s
ed
t
r
an
s
f
o
r
m
er
.
LR
ac
h
iev
ed
9
2
.
1
%
ac
cu
r
ac
y
a
n
d
an
F1
-
s
co
r
e
o
f
0
.
9
0
,
r
e
f
lectin
g
its
lim
itatio
n
s
o
n
c
o
m
p
l
ex
DDo
S
tr
a
f
f
ic.
I
n
co
n
tr
ast,
th
e
tr
an
s
f
o
r
m
er
a
ch
i
ev
ed
9
9
.
9
5
%
av
e
r
ag
e
ac
cu
r
a
cy
with
b
alan
ce
d
p
r
ec
is
io
n
a
n
d
r
ec
all
(
≈
9
9
.
9
%)
u
n
d
er
5
-
f
o
ld
cr
o
s
s
-
v
alid
atio
n
.
L
o
w
v
ar
ian
ce
i
n
d
icate
s
s
tr
o
n
g
g
en
er
aliza
tio
n
d
u
e
to
r
eg
u
lar
i
ze
d
d
esig
n
,
f
ea
t
u
r
e
s
elec
tio
n
,
an
d
s
elf
-
atten
tio
n
.
Desp
ite
s
tr
o
n
g
r
esu
lts
,
ev
alu
atio
n
is
lim
ited
to
C
I
C
-
DDo
S2
0
1
9
,
m
o
tiv
atin
g
f
u
tu
r
e
r
ea
l
-
w
o
r
ld
,
c
r
o
s
s
-
d
ataset,
an
d
ze
r
o
-
d
ay
s
tu
d
ies.
T
h
e
ev
alu
atio
n
is
lim
ited
to
th
e
C
I
C
-
DDo
S2
0
1
9
d
ataset,
ch
o
s
en
f
o
r
its
r
ea
lis
ti
c
tr
af
f
ic
an
d
d
iv
er
s
e
DDo
S
attac
k
s
.
Var
iatio
n
s
in
f
ea
tu
r
es,
t
r
af
f
ic
p
atter
n
s
,
an
d
lab
elin
g
ac
r
o
s
s
d
atasets
s
u
ch
as
C
I
C
-
I
DS2
0
1
7
,
UNSW
-
NB
1
5
,
an
d
B
o
t
-
I
o
T
m
a
y
ca
u
s
e
d
ataset
s
h
if
t
an
d
af
f
ec
t
g
en
er
aliza
tio
n
.
L
ea
v
e
-
one
-
f
a
m
ily
-
o
u
t
test
in
g
,
a
s
tr
o
n
g
ze
r
o
-
d
ay
ev
al
u
atio
n
,
r
eq
u
ir
es
ex
ten
s
iv
e
r
etr
ain
in
g
an
d
is
b
ey
o
n
d
t
h
is
r
ev
is
io
n
’
s
s
co
p
e
.
5.
CO
NCLU
SI
O
N
T
h
is
wo
r
k
p
r
esen
ted
a
tr
a
n
s
f
o
r
m
er
-
b
ased
f
r
am
ew
o
r
k
f
o
r
DDo
S
attac
k
d
etec
tio
n
u
s
in
g
th
e
C
I
C
-
DDo
S2
0
1
9
d
ataset,
ad
d
r
ess
in
g
th
e
lim
itatio
n
s
o
f
tr
a
d
itio
n
al
class
if
ier
s
in
m
o
d
eli
n
g
co
m
p
lex
attac
k
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
F
ea
tu
r
e
-
g
u
id
e
d
tr
a
n
s
fo
r
mer a
p
p
r
o
a
ch
f
o
r
d
etec
tin
g
d
is
tr
ib
u
ted
d
en
ia
l
o
f
…
(
Lo
ke
s
h
w
a
r
a
n
K
a
n
a
g
a
r
a
j
)
3719
b
eh
av
io
r
s
.
A
s
tr
u
ctu
r
ed
p
ip
eli
n
e
in
v
o
lv
in
g
p
r
ep
r
o
ce
s
s
in
g
,
s
c
alin
g
,
an
d
f
ea
tu
r
e
s
elec
tio
n
wa
s
ad
o
p
ted
,
with
LR
s
er
v
in
g
as
a
b
aselin
e.
T
h
e
p
r
o
p
o
s
ed
m
o
d
el
in
teg
r
ated
d
en
s
e
p
r
o
jectio
n
s
,
m
u
lti
-
h
ea
d
atte
n
tio
n
,
f
ee
d
-
f
o
r
war
d
lay
er
s
,
d
r
o
p
o
u
t,
a
n
d
g
lo
b
al
av
er
ag
e
p
o
o
lin
g
.
E
x
p
er
im
en
t
al
r
esu
lts
d
em
o
n
s
tr
ated
s
u
p
e
r
io
r
p
er
f
o
r
m
a
n
ce
,
ac
h
iev
in
g
9
9
.
9
5
%
av
e
r
ag
e
ac
cu
r
ac
y
with
b
alan
ce
d
p
r
ec
is
io
n
an
d
r
ec
all,
s
u
p
p
o
r
ted
b
y
c
o
n
f
u
s
io
n
m
atr
ix
a
n
d
R
OC
/
p
r
ec
is
io
n
–
r
ec
all
an
aly
s
es.
Featu
r
e
s
elec
t
io
n
f
u
r
th
er
i
m
p
r
o
v
e
d
ac
cu
r
ac
y
wh
ile
r
ed
u
cin
g
co
m
p
le
x
ity
,
h
ig
h
lig
h
tin
g
th
e
ef
f
icien
cy
o
f
co
m
p
ac
t
tr
an
s
f
o
r
m
er
d
esig
n
s
.
Alth
o
u
g
h
ev
alu
ati
o
n
was
lim
ited
to
o
f
f
lin
e
ex
p
er
im
en
ts
o
n
a
s
in
g
le
d
ataset,
th
e
f
ea
tu
r
e
-
g
u
id
ed
t
r
an
s
f
o
r
m
er
s
h
o
ws
s
tr
o
n
g
p
o
ten
tial
f
o
r
cr
o
s
s
-
d
ataset
an
d
ze
r
o
-
d
a
y
d
etec
tio
n
,
wh
ich
will b
e
in
v
esti
g
ated
in
f
u
tu
r
e
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
L
o
k
e
s
h
w
a
r
a
n
K
a
n
a
g
a
r
aj
✓
✓
✓
✓
✓
✓
✓
✓
✓
R
ag
u
r
am
an
Pu
r
u
s
h
o
th
am
an
✓
✓
✓
✓
✓
✓
Sath
y
a
Su
b
r
am
an
ia
n
✓
✓
✓
✓
✓
✓
✓
Du
r
g
a
Dev
i Sar
av
a
n
an
✓
✓
✓
✓
✓
✓
✓
Ko
m
al
Ku
m
ar
Nap
a
✓
✓
✓
✓
✓
✓
✓
✓
✓
C
o
r
n
eliu
s
Kar
u
n
ak
ar
a
n
✓
✓
✓
✓
✓
B
illa M
an
in
d
h
ar
✓
✓
✓
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
T
h
e
au
t
h
o
r
d
ec
lar
es
t
h
at
th
e
r
e
ar
e
n
o
k
n
o
wn
co
n
f
licts
o
f
in
ter
est
ass
o
ciate
d
with
th
is
p
u
b
licatio
n
.
T
h
er
e
ar
e
n
o
f
in
a
n
cial
o
r
p
e
r
s
o
n
al
r
elatio
n
s
h
ip
s
th
at
co
u
ld
in
ap
p
r
o
p
r
iately
in
f
l
u
en
ce
o
r
b
ias
th
e
co
n
ten
t
o
f
th
is
wo
r
k
.
DATA AV
AI
L
AB
I
L
I
T
Y
T
h
e
ex
p
e
r
im
en
ts
in
th
is
s
tu
d
y
wer
e
ca
r
r
ied
o
u
t
u
s
in
g
th
e
C
I
C
-
DDo
S2
0
1
9
d
ataset,
wh
ic
h
is
o
p
en
ly
av
ailab
le
o
n
Kag
g
le
at
h
ttp
s
://www.
k
ag
g
le.
co
m
/d
atasets
/d
h
o
o
g
la/cicd
d
o
s
2
0
1
9
.
RE
F
E
R
E
NC
E
S
[
1
]
T.
M
a
h
j
a
b
i
n
,
Y
.
X
i
a
o
,
G
.
S
u
n
,
a
n
d
W
.
J
i
a
n
g
,
“
A
su
r
v
e
y
o
f
d
i
s
t
r
i
b
u
t
e
d
d
e
n
i
a
l
-
of
-
ser
v
i
c
e
a
t
t
a
c
k
,
p
r
e
v
e
n
t
i
o
n
,
a
n
d
mi
t
i
g
a
t
i
o
n
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
r
n
a
l
o
f
D
i
st
ri
b
u
t
e
d
S
e
n
s
o
r
N
e
t
w
o
r
k
s
,
v
o
l
.
1
3
,
n
o
.
1
2
,
2
0
1
7
,
d
o
i
:
1
0
.
1
1
7
7
/
1
5
5
0
1
4
7
7
1
7
7
4
1
4
6
3
.
[
2
]
O
.
O
sa
n
a
i
y
e
,
K
.
-
K
.
R
.
C
h
o
o
,
a
n
d
M
.
D
l
o
d
l
o
,
“
D
i
s
t
r
i
b
u
t
e
d
d
e
n
i
a
l
o
f
ser
v
i
c
e
(
D
D
o
S
)
r
e
s
i
l
i
e
n
c
e
i
n
c
l
o
u
d
:
r
e
v
i
e
w
a
n
d
c
o
n
c
e
p
t
u
a
l
c
l
o
u
d
D
D
o
S
mi
t
i
g
a
t
i
o
n
f
r
a
m
e
w
o
r
k
,
”
J
o
u
r
n
a
l
o
f
N
e
t
w
o
rk
a
n
d
C
o
m
p
u
t
e
r
A
p
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
6
7
,
p
p
.
1
4
7
–
1
6
5
,
M
a
y
2
0
1
6
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
j
n
c
a
.
2
0
1
6
.
0
1
.
0
0
1
.
[
3
]
W
.
K
h
r
e
i
c
h
,
B
.
K
h
o
sr
a
v
i
f
a
r
,
A
.
H
.
-
L
h
a
d
j
,
a
n
d
C
.
T
a
l
h
i
,
“
A
n
a
n
o
ma
l
y
d
e
t
e
c
t
i
o
n
sy
st
e
m
b
a
s
e
d
o
n
v
a
r
i
a
b
l
e
N
-
g
r
a
m
f
e
a
t
u
r
e
s
a
n
d
o
n
e
-
c
l
a
ss SV
M
,
”
I
n
f
o
rm
a
t
i
o
n
a
n
d
S
o
f
t
w
a
re
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
9
1
,
p
p
.
1
8
6
–
1
9
7
,
2
0
1
7
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
i
n
f
so
f
.
2
0
1
7
.
0
7
.
0
0
9
.
[
4
]
P
.
D
.
B
o
j
o
v
i
ć
,
I
.
B
a
ši
č
e
v
i
ć
,
S
.
O
c
o
v
a
j
,
a
n
d
M
.
P
o
p
o
v
i
ć
,
“
A
p
r
a
c
t
i
c
a
l
a
p
p
r
o
a
c
h
t
o
d
e
t
e
c
t
i
o
n
o
f
d
i
st
r
i
b
u
t
e
d
d
e
n
i
a
l
-
of
-
serv
i
c
e
a
t
t
a
c
k
s
u
si
n
g
a
h
y
b
r
i
d
d
e
t
e
c
t
i
o
n
m
e
t
h
o
d
,
”
C
o
m
p
u
t
e
rs
a
n
d
El
e
c
t
ri
c
a
l
E
n
g
i
n
e
e
r
i
n
g
,
v
o
l
.
7
3
,
n
o
.
1
4
,
p
p
.
8
4
–
9
6
,
2
0
1
9
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
c
o
m
p
e
l
e
c
e
n
g
.
2
0
1
8
.
1
1
.
0
0
4
.
[
5
]
Q
.
N
i
y
a
z
,
W
.
S
u
n
,
a
n
d
A
.
Y
.
Ja
v
a
i
d
,
“
A
d
e
e
p
l
e
a
r
n
i
n
g
-
b
a
se
d
D
D
o
S
d
e
t
e
c
t
i
o
n
s
y
s
t
e
m
i
n
s
o
f
t
w
a
r
e
-
d
e
f
i
n
e
d
n
e
t
w
o
r
k
i
n
g
(
S
D
N
)
,
”
I
C
S
T
T
r
a
n
sa
c
t
i
o
n
s
o
n
S
e
c
u
r
i
t
y
a
n
d
S
a
f
e
t
y
,
v
o
l
.
4
,
n
o
.
1
2
,
2
0
1
7
,
d
o
i
:
1
0
.
4
1
0
8
/
e
a
i
.
2
8
-
12
-
2
0
1
7
.
1
5
3
5
1
5
.
[
6
]
M
.
N
a
j
a
f
i
me
h
r
,
S
.
Z
a
r
i
f
z
a
d
e
h
,
a
n
d
S
.
M
o
s
t
a
f
a
v
i
,
“
A
h
y
b
r
i
d
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
a
p
p
r
o
a
c
h
f
o
r
d
e
t
e
c
t
i
n
g
u
n
p
r
e
c
e
d
e
n
t
e
d
D
D
o
S
a
t
t
a
c
k
s,
”
J
o
u
rn
a
l
o
f
S
u
p
e
rc
o
m
p
u
t
i
n
g
,
v
o
l
.
7
8
,
n
o
.
6
,
p
p
.
8
1
0
6
–
8
1
3
6
,
2
0
2
2
,
d
o
i
:
1
0
.
1
0
0
7
/
s1
1
2
2
7
-
0
2
1
-
0
4
2
5
3
-
x.
[
7
]
A
.
A
l
d
w
e
e
sh
,
A
.
D
e
r
h
a
b
,
a
n
d
A
.
Z.
Emam,
“
D
e
e
p
l
e
a
r
n
i
n
g
a
p
p
r
o
a
c
h
e
s
f
o
r
a
n
o
m
a
l
y
-
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
ms:
a
s
u
r
v
e
y
,
t
a
x
o
n
o
my
,
a
n
d
o
p
e
n
i
ss
u
e
s,”
K
n
o
w
l
e
d
g
e
-
B
a
se
d
S
y
s
t
e
m
s
,
v
o
l
.
1
8
9
,
2
0
2
0
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
k
n
o
s
y
s
.
2
0
1
9
.
1
0
5
1
2
4
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
1
5
,
No
.
4
,
Au
g
u
s
t 2
0
2
6
:
3
7
1
2
-
3
7
2
1
3720
[
8
]
C
.
M
a
,
X
.
D
u
,
a
n
d
L.
C
a
o
,
“
A
n
a
l
y
s
i
s
o
f
mu
l
t
i
-
t
y
p
e
s
o
f
f
l
o
w
f
e
a
t
u
r
e
s
b
a
sed
o
n
h
y
b
r
i
d
n
e
u
r
a
l
n
e
t
w
o
r
k
f
o
r
i
m
p
r
o
v
i
n
g
n
e
t
w
o
r
k
a
n
o
ma
l
y
d
e
t
e
c
t
i
o
n
,
”
I
EEE
Ac
c
e
ss
,
v
o
l
.
7
,
p
p
.
1
4
8
3
6
3
–
1
4
8
3
8
0
,
2
0
1
9
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
1
9
.
2
9
4
6
7
0
8
.
[
9
]
A
.
V
a
sw
a
n
i
e
t
a
l
.
,
“
A
t
t
e
n
t
i
o
n
i
s
a
l
l
y
o
u
n
e
e
d
,
”
A
d
v
a
n
c
e
s
i
n
N
e
u
ra
l
I
n
f
o
rm
a
t
i
o
n
Pr
o
c
e
ss
i
n
g
S
y
st
e
m
s
,
v
o
l
.
2
0
1
7
,
p
p
.
5
9
9
9
–
6
0
0
9
,
2
0
1
7
,
d
o
i
:
1
0
.
1
2
0
1
/
9
7
8
1
0
0
3
5
6
1
4
6
0
-
1
9
.
[
1
0
]
I
.
S
h
a
r
a
f
a
l
d
i
n
,
A
.
H
.
L
a
sh
k
a
r
i
,
S
.
H
a
k
a
k
,
a
n
d
A
.
A
.
G
h
o
r
b
a
n
i
,
“
D
e
v
e
l
o
p
i
n
g
r
e
a
l
i
st
i
c
d
i
st
r
i
b
u
t
e
d
d
e
n
i
a
l
o
f
ser
v
i
c
e
(
D
D
o
S
)
a
t
t
a
c
k
d
a
t
a
se
t
a
n
d
t
a
x
o
n
o
my
,
”
i
n
2
0
1
9
I
n
t
e
rn
a
t
i
o
n
a
l
C
a
rn
a
h
a
n
C
o
n
f
e
re
n
c
e
o
n
S
e
c
u
ri
t
y
T
e
c
h
n
o
l
o
g
y
(
I
C
C
S
T
)
,
2
0
1
9
,
p
p
.
1
–
6
,
d
o
i
:
1
0
.
1
1
0
9
/
C
C
S
T.
2
0
1
9
.
8
8
8
8
4
1
9
.
[
1
1
]
A
.
G
a
r
g
,
S
.
K
a
m
a
l
a
n
a
t
h
a
n
,
P
.
S
u
h
a
s
a
r
i
a
,
a
n
d
A
.
A
g
a
r
w
a
l
,
“
D
i
st
r
i
b
u
t
e
d
d
e
n
i
a
l
o
f
ser
v
i
c
e
a
t
t
a
c
k
s:
a
s
u
r
v
e
y
,
”
I
m
p
e
r
i
a
l
J
o
u
rn
a
l
o
f
I
n
t
e
r
d
i
sci
p
l
i
n
a
r
y
R
e
se
a
rc
h
,
v
o
l
.
3
,
n
o
.
3
,
p
p
.
3
1
–
3
7
,
2
0
1
7
.
[
1
2
]
S
.
H
a
j
j
,
R
.
El
S
i
b
a
i
,
J.
B
.
A
b
d
o
,
J.
D
e
merj
i
a
n
,
A
.
M
a
k
h
o
u
l
,
a
n
d
C
.
G
u
y
e
u
x
,
“
A
n
o
m
a
l
y
-
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
ms:
t
h
e
r
e
q
u
i
r
e
m
e
n
t
s,
me
t
h
o
d
s,
m
e
a
s
u
r
e
m
e
n
t
s,
a
n
d
d
a
t
a
se
t
s,
”
T
r
a
n
s
a
c
t
i
o
n
s
o
n
Em
e
rg
i
n
g
T
e
l
e
c
o
m
m
u
n
i
c
a
t
i
o
n
s
T
e
c
h
n
o
l
o
g
i
e
s
,
v
o
l
.
3
2
,
n
o
.
4
,
2
0
2
1
,
d
o
i
:
1
0
.
1
0
0
2
/
e
t
t
.
4
2
4
0
.
[
1
3
]
A
.
G
a
r
g
a
n
d
P
.
M
a
h
e
s
h
w
a
r
i
,
“
A
h
y
b
r
i
d
i
n
t
r
u
s
i
o
n
d
e
t
e
c
t
i
o
n
s
y
s
t
e
m
:
a
r
e
v
i
e
w
,
”
i
n
1
0
t
h
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
I
n
t
e
l
l
i
g
e
n
t
S
y
s
t
e
m
s
a
n
d
C
o
n
t
r
o
l
,
I
S
C
O
2
0
1
6
,
2
0
1
6
,
p
p
.
1
–
5
,
d
o
i
:
1
0
.
1
1
0
9
/
I
S
C
O
.
2
0
1
6
.
7
7
2
6
9
0
9
.
[
1
4
]
N
.
B
i
n
d
r
a
a
n
d
M
.
S
o
o
d
,
“
E
v
a
l
u
a
t
i
n
g
t
h
e
i
mp
a
c
t
o
f
f
e
a
t
u
r
e
se
l
e
c
t
i
o
n
me
t
h
o
d
s
o
n
t
h
e
p
e
r
f
o
r
ma
n
c
e
o
f
t
h
e
mac
h
i
n
e
l
e
a
r
n
i
n
g
m
o
d
e
l
s
i
n
d
e
t
e
c
t
i
n
g
D
D
o
S
a
t
t
a
c
k
s,
”
R
o
m
a
n
i
a
n
J
o
u
rn
a
l
o
f
I
n
f
o
rm
a
t
i
o
n
S
c
i
e
n
c
e
a
n
d
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
2
3
,
n
o
.
3
,
p
p
.
2
5
0
–
2
6
1
,
2
0
2
0
.
[
1
5
]
K
.
B
o
u
z
o
u
b
a
a
,
Y
.
T
a
h
e
r
,
a
n
d
B
.
N
s
i
r
i
,
“
P
r
e
d
i
c
t
i
n
g
D
O
S
-
D
D
O
S
a
t
t
a
c
k
s:
r
e
v
i
e
w
a
n
d
e
v
a
l
u
a
t
i
o
n
st
u
d
y
o
f
f
e
a
t
u
r
e
se
l
e
c
t
i
o
n
m
e
t
h
o
d
s
b
a
s
e
d
o
n
w
r
a
p
p
e
r
p
r
o
c
e
ss,
”
I
n
t
e
rn
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
f
A
d
v
a
n
c
e
d
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
a
n
d
A
p
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
1
2
,
n
o
.
5
,
p
p
.
1
3
2
–
1
4
5
,
2
0
2
1
,
d
o
i
:
1
0
.
1
4
5
6
9
/
I
JA
C
S
A
.
2
0
2
1
.
0
1
2
0
5
1
7
.
[
1
6
]
O
.
O
san
a
i
y
e
,
H
.
C
a
i
,
K
.
K
.
R
.
C
h
o
o
,
A
.
D
e
h
g
h
a
n
t
a
n
h
a
,
Z.
X
u
,
a
n
d
M
.
D
l
o
d
l
o
,
“
En
s
e
m
b
l
e
-
b
a
s
e
d
mu
l
t
i
-
f
i
l
t
e
r
f
e
a
t
u
r
e
se
l
e
c
t
i
o
n
met
h
o
d
f
o
r
D
D
o
S
d
e
t
e
c
t
i
o
n
i
n
c
l
o
u
d
c
o
mp
u
t
i
n
g
,
”
Eu
r
a
s
i
p
J
o
u
r
n
a
l
o
n
W
i
re
l
e
ss C
o
m
m
u
n
i
c
a
t
i
o
n
s
a
n
d
N
e
t
w
o
rk
i
n
g
,
v
o
l
.
2
0
1
6
,
n
o
.
1
,
2
0
1
6
,
d
o
i
:
1
0
.
1
1
8
6
/
s1
3
6
3
8
-
0
1
6
-
0
6
2
3
-
3.
[
1
7
]
S
.
A
l
m
u
t
a
i
r
i
,
S
.
M
a
h
f
o
u
d
h
,
S
.
A
l
m
u
t
a
i
r
i
,
a
n
d
J
.
S
.
A
l
o
w
i
b
d
i
,
“
H
y
b
r
i
d
b
o
t
n
e
t
d
e
t
e
c
t
i
o
n
b
a
se
d
o
n
h
o
s
t
a
n
d
n
e
t
w
o
r
k
a
n
a
l
y
si
s
,
”
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
C
o
m
m
u
n
i
c
a
t
i
o
n
s
,
v
o
l
.
2
0
2
0
,
2
0
2
0
,
d
o
i
:
1
0
.
1
1
5
5
/
2
0
2
0
/
9
0
2
4
7
2
6
.
[
1
8
]
S
.
S
o
k
k
a
l
i
n
g
a
m
a
n
d
R
.
R
a
ma
k
r
i
s
h
n
a
n
,
“
A
n
i
n
t
e
l
l
i
g
e
n
t
i
n
t
r
u
s
i
o
n
d
e
t
e
c
t
i
o
n
s
y
st
e
m
f
o
r
d
i
st
r
i
b
u
t
e
d
d
e
n
i
a
l
o
f
serv
i
c
e
a
t
t
a
c
k
s:
a
su
p
p
o
r
t
v
e
c
t
o
r
mac
h
i
n
e
w
i
t
h
h
y
b
r
i
d
o
p
t
i
m
i
z
a
t
i
o
n
a
l
g
o
r
i
t
h
m
b
a
se
d
a
p
p
r
o
a
c
h
,
”
C
o
n
c
u
rre
n
c
y
a
n
d
C
o
m
p
u
t
a
t
i
o
n
:
Pr
a
c
t
i
c
e
a
n
d
Ex
p
e
ri
e
n
c
e
,
v
o
l
.
3
4
,
n
o
.
2
7
,
2
0
2
2
,
d
o
i
:
1
0
.
1
0
0
2
/
c
p
e
.
7
3
3
4
.
[
1
9
]
M
.
A
.
H
.
A
z
m
i
,
C
.
F
.
M
.
F
o
o
z
y
,
K
.
A
.
M
.
S
u
k
r
i
,
N
.
A
.
A
b
d
u
l
l
a
h
,
I
.
R
.
A
.
H
a
mi
d
,
a
n
d
H
.
A
m
n
u
r
,
“
F
e
a
t
u
r
e
se
l
e
c
t
i
o
n
a
p
p
r
o
a
c
h
t
o
d
e
t
e
c
t
D
D
o
S
a
t
t
a
c
k
u
s
i
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
rn
a
l
o
n
I
n
f
o
rm
a
t
i
c
s
V
i
s
u
a
l
i
z
a
t
i
o
n
,
v
o
l
.
5
,
n
o
.
4
,
p
p
.
3
9
5
–
4
0
1
,
2
0
2
1
,
d
o
i
:
1
0
.
3
0
6
3
0
/
JO
I
V
.
5
.
4
.
7
3
4
.
[
2
0
]
J.
D
a
v
i
d
a
n
d
C
.
Th
o
m
a
s,
“
D
D
o
S
a
t
t
a
c
k
d
e
t
e
c
t
i
o
n
u
si
n
g
f
a
s
t
e
n
t
r
o
p
y
a
p
p
r
o
a
c
h
o
n
f
l
o
w
-
b
a
s
e
d
n
e
t
w
o
r
k
t
r
a
f
f
i
c
,
”
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
.
5
0
,
p
p
.
3
0
–
3
6
,
2
0
1
5
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
p
r
o
c
s.
2
0
1
5
.
0
4
.
0
0
7
.
[
2
1
]
R
.
B
i
t
i
t
,
A
.
D
e
r
h
a
b
,
M
.
G
u
e
r
r
o
u
m
i
,
a
n
d
F
.
A
.
K
h
a
n
,
“
D
D
o
S
a
t
t
a
c
k
f
o
r
e
c
a
s
t
i
n
g
b
a
se
d
o
n
o
n
l
i
n
e
m
u
l
t
i
p
l
e
c
h
a
n
g
e
p
o
i
n
t
s
d
e
t
e
c
t
i
o
n
a
n
d
t
i
me
s
e
r
i
e
s
a
n
a
l
y
s
i
s,
”
M
u
l
t
i
m
e
d
i
a
T
o
o
l
s
a
n
d
A
p
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
8
3
,
n
o
.
1
8
,
p
p
.
5
3
6
5
5
–
5
3
6
8
5
,
2
0
2
4
,
d
o
i
:
1
0
.
1
0
0
7
/
s1
1
0
4
2
-
023
-
1
7
6
3
7
-
3.
[
2
2
]
L.
H
o
n
g
,
K
.
W
e
h
b
i
,
a
n
d
T
.
H
.
A
l
s
a
l
a
h
,
“
H
y
b
r
i
d
f
e
a
t
u
r
e
se
l
e
c
t
i
o
n
f
o
r
e
f
f
i
c
i
e
n
t
d
e
t
e
c
t
i
o
n
o
f
D
D
o
S
a
t
t
a
c
k
s
i
n
I
o
T,
”
i
n
AC
M
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
Pro
c
e
e
d
i
n
g
S
e
r
i
e
s
,
2
0
2
2
,
p
p
.
1
2
0
–
127
,
d
o
i
:
1
0
.
1
1
4
5
/
3
5
5
6
6
7
7
.
3
5
5
6
6
8
7
.
[
2
3
]
F
.
K
a
ma
l
o
v
,
S
.
M
o
u
ss
a
,
Z
.
E
l
K
h
a
t
i
b
,
a
n
d
A
.
B
e
n
M
n
a
o
u
e
r
,
“
O
r
t
h
o
g
o
n
a
l
v
a
r
i
a
n
c
e
-
b
a
s
e
d
f
e
a
t
u
r
e
sel
e
c
t
i
o
n
f
o
r
i
n
t
r
u
si
o
n
d
e
t
e
c
t
i
o
n
sy
st
e
ms,
”
i
n
2
0
2
1
I
n
t
e
r
n
a
t
i
o
n
a
l
S
y
m
p
o
s
i
u
m
o
n
N
e
t
w
o
rks,
C
o
m
p
u
t
e
rs
a
n
d
C
o
m
m
u
n
i
c
a
t
i
o
n
s
,
I
S
N
C
C
2
0
2
1
,
2
0
2
1
,
p
p
.
1
–
5
,
d
o
i
:
1
0
.
1
1
0
9
/
I
S
N
C
C
5
2
1
7
2
.
2
0
2
1
.
9
6
1
5
6
5
6
.
[
2
4
]
A
.
M
a
s
l
a
n
,
K
.
M
.
B
.
M
o
h
a
m
a
d
,
A
.
H
a
mi
d
,
H
.
P
a
n
g
a
r
i
b
u
a
n
,
a
n
d
S
.
S
i
t
o
h
a
n
g
,
“
F
e
a
t
u
r
e
s
e
l
e
c
t
i
o
n
t
o
e
n
h
a
n
c
e
D
D
o
S
d
e
t
e
c
t
i
o
n
u
s
i
n
g
h
y
b
r
i
d
N
-
g
r
a
m
h
e
u
r
i
st
i
c
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
r
n
a
l
o
n
I
n
f
o
rm
a
t
i
c
s
Vi
s
u
a
l
i
za
t
i
o
n
,
v
o
l
.
7
,
n
o
.
3
,
p
p
.
8
1
5
–
8
2
2
,
2
0
2
3
,
d
o
i
:
1
0
.
3
0
6
3
0
/
j
o
i
v
.
7
.
3
.
1
5
3
3
.
[
2
5
]
H
.
W
a
n
g
a
n
d
W
.
L
i
,
“
D
D
o
s
TC
:
a
t
r
a
n
sf
o
r
m
e
r
-
b
a
s
e
d
n
e
t
w
o
r
k
a
t
t
a
c
k
d
e
t
e
c
t
i
o
n
h
y
b
r
i
d
mec
h
a
n
i
sm
i
n
S
D
N
,
”
S
e
n
so
rs
,
v
o
l
.
2
1
,
n
o
.
1
5
,
2
0
2
1
,
d
o
i
:
1
0
.
3
3
9
0
/
s2
1
1
5
5
0
4
7
.
[
2
6
]
Y
.
Li
u
a
n
d
L.
W
u
,
“
I
n
t
r
u
s
i
o
n
d
e
t
e
c
t
i
o
n
m
o
d
e
l
b
a
s
e
d
o
n
i
mp
r
o
v
e
d
t
r
a
n
sf
o
r
m
e
r
,
”
A
p
p
l
i
e
d
S
c
i
e
n
c
e
s
,
v
o
l
.
1
3
,
n
o
.
1
0
,
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
a
p
p
1
3
1
0
6
2
5
1
.
[
2
7
]
Z.
L
o
n
g
,
H
.
Y
a
n
,
G
.
S
h
e
n
,
X
.
Z
h
a
n
g
,
H
.
H
e
,
a
n
d
L.
C
h
e
n
g
,
“
A
t
r
a
n
sf
o
r
mer
-
b
a
se
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
a
p
p
r
o
a
c
h
f
o
r
c
l
o
u
d
sec
u
r
i
t
y
,
”
J
o
u
rn
a
l
o
f
C
l
o
u
d
C
o
m
p
u
t
i
n
g
,
v
o
l
.
1
3
,
n
o
.
1
,
2
0
2
4
,
d
o
i
:
1
0
.
1
1
8
6
/
s
1
3
6
7
7
-
0
2
3
-
0
0
5
7
4
-
9.
[
2
8
]
F
.
U
l
l
a
h
,
S
.
U
l
l
a
h
,
G
.
S
r
i
v
a
s
t
a
v
a
,
a
n
d
J
.
C
.
W
.
L
i
n
,
“
I
D
S
-
I
N
T:
i
n
t
r
u
si
o
n
d
e
t
e
c
t
i
o
n
s
y
st
e
m
u
s
i
n
g
t
r
a
n
sf
o
r
m
e
r
-
b
a
s
e
d
t
r
a
n
sf
e
r
l
e
a
r
n
i
n
g
f
o
r
i
mb
a
l
a
n
c
e
d
n
e
t
w
o
r
k
t
r
a
f
f
i
c
,
”
D
i
g
i
t
a
l
C
o
m
m
u
n
i
c
a
t
i
o
n
s
a
n
d
N
e
t
w
o
r
k
s
,
v
o
l
.
1
0
,
n
o
.
1
,
p
p
.
1
9
0
–
2
0
4
,
2
0
2
4
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
d
c
a
n
.
2
0
2
3
.
0
3
.
0
0
8
.
[
2
9
]
M
.
A
.
H
o
ss
a
i
n
a
n
d
M
.
S
.
I
sl
a
m,
“
E
n
h
a
n
c
i
n
g
D
D
o
S
a
t
t
a
c
k
d
e
t
e
c
t
i
o
n
w
i
t
h
h
y
b
r
i
d
f
e
a
t
u
r
e
se
l
e
c
t
i
o
n
a
n
d
e
n
s
e
m
b
l
e
-
b
a
s
e
d
c
l
a
ss
i
f
i
e
r
:
a
p
r
o
m
i
si
n
g
s
o
l
u
t
i
o
n
f
o
r
r
o
b
u
st
c
y
b
e
r
se
c
u
r
i
t
y
,
”
Me
a
s
u
rem
e
n
t
:
S
e
n
so
rs
,
v
o
l
.
3
2
,
2
0
2
4
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
me
a
se
n
.
2
0
2
4
.
1
0
1
0
3
7
.
[
3
0
]
J.
L
i
,
X
.
To
n
g
,
J.
Li
u
,
a
n
d
L
.
C
h
e
n
g
,
“
A
n
e
f
f
i
c
i
e
n
t
f
e
d
e
r
a
t
e
d
l
e
a
r
n
i
n
g
s
y
st
e
m
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
EE
E
S
y
s
t
e
m
s
J
o
u
rn
a
l
,
v
o
l
.
1
7
,
n
o
.
2
,
p
p
.
2
4
5
5
–
2
4
6
4
,
2
0
2
3
,
d
o
i
:
1
0
.
1
1
0
9
/
JS
Y
S
T.
2
0
2
3
.
3
2
3
6
9
9
5
.
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
Dr
.
Lo
k
e
shwa
r
a
n
K
a
n
a
g
a
r
a
j
is
c
u
rre
n
tl
y
w
o
rk
i
n
g
a
s
a
ss
o
c
iate
p
ro
fe
ss
o
r
in
t
h
e
De
p
a
rtme
n
t
o
f
Co
m
p
u
ter
S
c
ien
c
e
a
n
d
En
g
in
e
e
ri
n
g
a
t
V
e
l
T
e
c
h
Ra
n
g
a
ra
jan
Dr
.
S
a
g
u
n
th
a
la
R&
D
In
stit
u
te
o
f
S
c
ien
c
e
a
n
d
T
e
c
h
n
o
lo
g
y
,
A
v
a
d
i,
In
d
ia.
His
re
se
a
rc
h
in
tere
sts
i
n
c
lu
d
e
m
a
c
h
in
e
lea
rn
in
g
,
d
a
ta
m
in
i
n
g
,
se
m
a
n
ti
c
we
b
,
li
n
k
e
d
o
p
e
n
d
a
ta,
a
n
d
c
lo
u
d
c
o
m
p
u
ti
n
g
.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
k
.
lo
k
e
sh
wa
ra
n
@g
m
a
il
.
c
o
m
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
F
ea
tu
r
e
-
g
u
id
e
d
tr
a
n
s
fo
r
mer a
p
p
r
o
a
ch
f
o
r
d
etec
tin
g
d
is
tr
ib
u
ted
d
en
ia
l
o
f
…
(
Lo
ke
s
h
w
a
r
a
n
K
a
n
a
g
a
r
a
j
)
3721
Mr
.
Ra
g
u
r
a
m
a
n
P
u
r
u
sho
t
h
a
m
a
n
is
c
u
rre
n
tl
y
w
o
rk
i
n
g
a
s
a
n
a
ss
istan
t
p
r
o
fe
ss
o
r
i
n
th
e
De
p
a
rtme
n
t
o
f
C
o
m
p
u
ter
S
c
ien
c
e
&
En
g
i
n
e
e
rin
g
(Artif
icia
l
In
telli
g
e
n
c
e
),
M
a
d
a
n
a
p
a
ll
e
In
stit
u
te
o
f
T
e
c
h
n
o
l
o
g
y
&
S
c
ien
c
e
De
e
m
e
d
to
b
e
Un
i
v
e
rsity
,
M
a
d
a
n
a
p
a
ll
e
,
I
n
d
ia
.
His
a
re
a
o
f
in
tere
st
in
c
lu
d
e
s
th
e
o
ry
o
f
c
o
m
p
u
tatio
n
,
d
e
sig
n
a
n
d
a
n
a
ly
sis
o
f
a
lg
o
rit
h
m
s,
ima
g
e
p
r
o
c
e
ss
in
g
,
a
n
d
d
a
ta sc
ien
c
e
.
He
c
a
n
b
e
c
o
n
ta
c
ted
a
t
e
m
a
il
:
y
u
v
a
ra
g
u
.
p
t@
g
m
a
il
.
c
o
m
.
Dr
.
S
a
th
y
a
S
u
b
r
a
m
a
n
i
a
n
is
c
u
rre
n
tl
y
w
o
rk
i
n
g
a
s
a
ss
o
c
iate
p
ro
fe
ss
o
r
in
th
e
De
p
a
rtme
n
t
o
f
El
e
c
tro
n
ics
a
n
d
Co
m
m
u
n
ica
ti
o
n
En
g
i
n
e
e
rin
g
,
S
.
A
.
En
g
in
e
e
rin
g
Co
ll
e
g
e
,
a
f
fil
iate
d
t
o
A
n
n
a
U
n
iv
e
rsity
,
C
h
e
n
n
a
i.
He
r
a
re
a
s
o
f
i
n
tere
st
a
n
d
re
se
a
rc
h
d
o
m
a
in
a
re
ima
g
e
p
ro
c
e
ss
in
g
,
e
m
b
e
d
d
e
d
sy
ste
m
s
,
a
n
d
M
E
M
S
.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
sa
th
y
a
s7
9
7
9
@
g
m
a
il
.
c
o
m
.
Dr
.
Durg
a
De
v
i
S
a
r
a
v
a
n
a
n
i
s
a
ss
o
c
iate
p
ro
fe
ss
o
r
i
n
t
h
e
De
p
a
rtme
n
t
o
f
C
o
m
p
u
ter
S
c
ien
c
e
a
n
d
En
g
in
e
e
rin
g
a
t
V
e
l
T
e
c
h
Hig
h
T
e
c
h
Dr
.
Ra
n
g
a
ra
jan
Dr
.
S
a
k
u
n
th
a
la
En
g
in
e
e
rin
g
Co
ll
e
g
e
.
He
r
a
re
a
o
f
in
tere
st
is
n
e
two
rk
se
c
u
rit
y
,
wire
les
s
sy
st
e
m
s,
a
n
d
I
o
T
.
S
h
e
h
a
s
v
a
st
e
x
p
e
rien
c
e
i
n
a
c
a
d
e
m
ia
a
n
d
a
d
m
in
istratio
n
.
S
h
e
h
a
s
p
u
b
l
ish
e
d
a
n
d
p
re
se
n
ted
p
a
p
e
rs
in
v
a
ri
o
u
s
in
tern
a
ti
o
n
a
l
a
n
d
n
a
ti
o
n
a
l
jo
u
rn
a
l
s
/co
n
fe
re
n
c
e
s.
S
h
e
is
a
n
a
c
ti
v
e
m
e
m
b
e
r
o
f
p
ro
fe
ss
io
n
a
l
b
o
d
ies
CS
I
a
n
d
IS
TE
.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
sd
u
r
g
a
d
e
v
i@v
e
lh
ig
h
t
e
c
h
.
c
o
m
.
Dr
.
K
o
m
a
l
K
u
m
a
r
Na
p
a
is
c
u
rre
n
tl
y
wo
r
k
i
n
g
a
s
a
ss
istan
t
p
ro
fe
ss
o
r
(S
G)
in
th
e
De
p
a
rtme
n
t
o
f
Artifi
c
ial
I
n
telli
g
e
n
c
e
a
n
d
Da
ta
S
c
ien
c
e
a
t
S
a
v
e
e
th
a
En
g
in
e
e
rin
g
Co
ll
e
g
e
,
Ch
e
n
n
a
i,
In
d
ia.
His
re
se
a
rc
h
i
n
tere
sts
in
c
lu
d
e
m
a
c
h
in
e
lea
rn
i
n
g
,
d
a
ta
m
i
n
in
g
,
a
n
d
c
lo
u
d
c
o
m
p
u
ti
n
g
.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
k
o
m
a
lk
u
m
a
rn
a
p
a
@g
m
a
il
.
c
o
m
.
Dr
.
Co
r
n
e
li
u
s
K
a
r
u
n
a
k
a
r
a
n
is
c
u
rre
n
tl
y
wo
rk
i
n
g
a
s
a
ss
o
c
iate
p
ro
fe
ss
o
r
in
th
e
De
p
a
rtme
n
t
o
f
Artifi
c
ial
In
telli
g
e
n
c
e
a
n
d
Da
ta
S
c
ien
c
e
a
t
RM
K
Co
ll
e
g
e
o
f
E
n
g
i
n
e
e
rin
g
a
n
d
Tec
h
n
o
l
o
g
y
,
Ti
r
u
v
a
ll
u
r,
Tam
il
Na
d
u
,
In
d
ia.
His
re
se
a
rc
h
in
tere
sts
in
c
lu
d
e
b
ig
d
a
ta
a
n
a
ly
ti
c
s,
a
rti
ficia
l
in
telli
g
e
n
c
e
,
m
a
c
h
in
e
le
a
rn
in
g
,
a
n
d
c
lo
u
d
c
o
m
p
u
ti
n
g
.
H
e
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
c
o
rn
e
li
u
s
8
5
1
@g
m
a
il
.
c
o
m
.
Mr
.
Bil
la
Ma
n
i
n
d
h
a
r
is
c
u
rre
n
tl
y
w
o
rk
i
n
g
a
s
a
n
a
ss
istan
t
p
r
o
fe
ss
o
r
i
n
th
e
De
p
a
rtme
n
t
o
f
Co
m
p
u
ter
S
c
i
e
n
c
e
a
n
d
En
g
i
n
e
e
rin
g
a
t
Ko
n
e
ru
Lak
sh
m
a
iah
Ed
u
c
a
ti
o
n
F
o
u
n
d
a
ti
o
n
,
V
a
d
d
e
sw
a
ra
m
,
Gu
n
tu
r
,
An
d
h
ra
P
ra
d
e
sh
,
In
d
ia.
Hi
s
re
se
a
r
c
h
in
tere
sts
in
c
lu
d
e
m
a
c
h
in
e
lea
rn
in
g
,
d
a
ta
m
in
i
n
g
,
a
n
d
c
lo
u
d
c
o
m
p
u
t
in
g
.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
b
m
a
n
in
d
h
a
r@k
l
u
n
i
v
e
rsity
.
in
.
Evaluation Warning : The document was created with Spire.PDF for Python.