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.
1
6
,
No
.
1
,
Feb
r
u
ar
y
20
2
6
,
p
p
.
311
~
320
I
SS
N:
2088
-
8
7
0
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijece.
v
1
6
i
1
.
pp
3
1
1
-
3
2
0
311
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ec
e.
ia
esco
r
e.
co
m
Credit
card
fraud
data
ana
ly
sis
using
propo
sed sa
mpling
a
lg
o
rithm and de
ep ensembl
e learn
ing
Ay
e
Ay
e
K
hin
e,
Z
in T
hu
T
h
u M
y
int
I
n
f
o
r
mat
i
o
n
a
n
d
C
o
mm
u
n
i
c
a
t
i
o
n
Te
c
h
n
o
l
o
g
y
R
e
se
a
r
c
h
C
e
n
t
r
e
,
Y
a
n
g
o
n
,
M
y
a
n
mar
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
u
n
3
,
2
0
2
5
R
ev
is
ed
Sep
2
6
,
2
0
2
5
Acc
ep
ted
No
v
2
3
,
2
0
2
5
Cre
d
it
c
a
rd
fra
u
d
d
e
tec
ti
o
n
is
c
h
a
ll
e
n
g
i
n
g
d
u
e
t
o
t
h
e
se
v
e
re
imb
a
lan
c
e
b
e
twe
e
n
leg
it
ima
te
a
n
d
fra
u
d
u
len
t
tran
sa
c
ti
o
n
s,
wh
ich
h
i
n
d
e
rs
a
c
c
u
ra
te
fra
u
d
id
e
n
ti
fica
ti
o
n
.
T
o
a
d
d
re
ss
t
h
is,
w
e
p
ro
p
o
se
a
d
e
e
p
lea
rn
i
n
g
-
b
a
se
d
e
n
se
m
b
le
m
o
d
e
l
i
n
teg
ra
ted
wit
h
a
p
ro
p
o
s
e
d
sa
m
p
li
n
g
a
l
g
o
ri
th
m
b
a
se
d
o
n
ra
n
d
o
m
o
v
e
rsa
m
p
li
n
g
.
Un
li
k
e
tra
d
it
i
o
n
a
l
m
e
th
o
d
s,
th
e
p
r
o
p
o
se
d
sa
m
p
li
n
g
a
lg
o
rit
h
m
a
d
d
re
ss
e
s
th
e
o
v
e
rsig
h
t
o
f
p
a
ra
m
e
ter
se
lec
ti
o
n
a
n
d
m
a
n
a
g
e
s
c
las
s
imb
a
lan
c
e
with
o
u
t
e
li
m
in
a
t
in
g
a
n
y
leg
i
ti
m
a
te
sa
m
p
les
.
Th
e
e
n
se
m
b
le
fra
m
e
wo
rk
c
o
m
b
i
n
e
s
t
h
e
stre
n
g
t
h
s
o
f
c
o
n
v
o
l
u
ti
o
n
a
l
n
e
u
ra
l
n
e
two
r
k
s
(CNN
)
f
o
r
sp
a
t
ial
fe
a
tu
re
e
x
trac
ti
o
n
,
lo
n
g
sh
o
rt
-
ter
m
m
e
m
o
ry
(LS
TM
)
n
e
two
r
k
s
fo
r
c
a
p
tu
ri
n
g
se
q
u
e
n
ti
a
l
p
a
tt
e
r
n
s,
a
n
d
m
u
lt
i
lay
e
r
p
e
rc
e
p
tro
n
s
(
M
LP
)
f
o
r
e
fficie
n
t
c
las
sifica
ti
o
n
.
T
h
re
e
e
n
se
m
b
le
s
trate
g
ies
—
Weig
h
ted
a
v
e
ra
g
e
,
u
n
we
ig
h
te
d
a
v
e
ra
g
e
,
a
n
d
u
n
we
i
g
h
ted
m
a
jo
rit
y
v
o
ti
n
g
—
a
re
e
m
p
l
o
y
e
d
to
a
g
g
re
g
a
te
p
re
d
ictio
n
s.
Ex
p
e
rime
n
tal
re
su
lt
s
sh
o
w
th
a
t
a
ll
e
n
se
m
b
le
m
e
th
o
d
s
a
c
h
iev
e
p
e
rfe
c
t
sc
o
re
s
(1
.
0
0
)
in
p
re
c
isio
n
,
re
c
a
ll
,
a
n
d
F
1
-
sc
o
re
fo
r
b
o
t
h
fra
u
d
a
n
d
non
-
fra
u
d
c
las
se
s.
Th
is
stu
d
y
d
e
m
o
n
stra
tes
th
e
e
ffe
c
ti
v
e
n
e
ss
o
f
e
n
se
m
b
le
m
o
d
e
l
wit
h
o
p
ti
m
ize
d
sa
m
p
li
n
g
a
p
p
r
o
a
c
h
f
o
r
ro
b
u
st
a
n
d
a
c
c
u
r
a
te
fra
u
d
d
e
tec
ti
o
n
.
K
ey
w
o
r
d
s
:
C
o
n
v
o
lu
tio
n
al
n
eu
r
al
n
etwo
r
k
lo
n
g
s
h
o
r
t
-
ter
m
m
em
o
r
y
C
r
ed
it c
ar
d
f
r
au
d
d
etec
tio
n
E
n
s
em
b
le
m
o
d
el
Mu
ltil
ay
er
p
er
ce
p
tr
o
n
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
Ay
e
Ay
e
Kh
in
e
I
n
f
o
r
m
atio
n
an
d
C
o
m
m
u
n
icatio
n
T
ec
h
n
o
lo
g
y
R
esear
ch
C
en
tr
e
Yan
g
o
n
,
M
y
an
m
ar
E
m
ail: a
y
ek
h
in
e
7
1
@
g
m
ail.
co
m
1.
I
NT
RO
D
UCT
I
O
N
C
r
ed
it
ca
r
d
f
r
au
d
is
an
escalatin
g
th
r
ea
t
in
t
h
e
f
i
n
an
cial
s
ec
to
r
,
ex
ac
e
r
b
ated
b
y
th
e
r
ap
id
g
r
o
wth
o
f
e
-
co
m
m
er
ce
a
n
d
d
i
g
ital
p
ay
m
en
ts
.
T
h
is
n
o
t
o
n
l
y
ca
u
s
es
s
u
b
s
tan
tial
f
in
an
cial
lo
s
s
es
b
u
t
also
u
n
d
e
r
m
in
es
co
n
s
u
m
er
tr
u
s
t
an
d
p
o
s
es
s
er
io
u
s
ch
allen
g
es
to
f
in
an
cial
in
s
titu
tio
n
s
.
E
x
is
tin
g
f
r
au
d
d
etec
tio
n
s
y
s
tem
s
p
r
ed
o
m
i
n
an
tly
r
ely
o
n
r
u
le
-
b
ased
tech
n
iq
u
es,
wh
ich
,
w
h
ile
s
im
p
le
an
d
i
n
ter
p
r
etab
l
e,
ar
e
in
c
r
ea
s
in
g
ly
in
ad
eq
u
ate
ag
ain
s
t
th
e
ev
o
lv
i
n
g
s
tr
ateg
ies
o
f
s
o
p
h
is
ticated
f
r
au
d
s
ter
s
.
T
h
ese
s
y
s
tem
s
s
u
f
f
er
f
r
o
m
h
ig
h
f
alse
p
o
s
itiv
e
r
ates a
n
d
lack
th
e
ad
a
p
tab
ilit
y
r
eq
u
ir
e
d
f
o
r
r
ea
l
-
tim
e
an
d
d
y
n
am
ic
f
r
au
d
d
etec
tio
n
[
1
]
–
[
3
]
.
T
o
o
v
er
co
m
e
th
ese
lim
itatio
n
s
,
th
e
r
esear
ch
c
o
m
m
u
n
ity
h
as
in
cr
ea
s
in
g
ly
tu
r
n
ed
to
m
ac
h
in
e
lear
n
in
g
an
d
d
ee
p
lear
n
in
g
tech
n
iq
u
es.
Sev
er
al
s
tate
-
of
-
th
e
-
a
r
t
a
p
p
r
o
a
ch
es
h
av
e
ex
p
l
o
r
ed
th
e
u
s
e
o
f
in
d
iv
id
u
al
m
o
d
el
s
s
u
ch
as
co
n
v
o
l
u
tio
n
al
n
e
u
r
al
n
etwo
r
k
s
(
C
NNs),
lo
n
g
s
h
o
r
t
-
ter
m
m
em
o
r
y
n
etwo
r
k
s
(
L
STM
s
)
,
an
d
m
u
ltil
ay
er
p
er
ce
p
tr
o
n
s
(
ML
Ps
)
,
ea
ch
o
f
f
er
in
g
d
is
tin
ct
ad
v
an
ta
g
es:
C
NNs
ar
e
ca
p
ab
le
o
f
lear
n
in
g
s
p
atial
p
atter
n
s
with
in
tr
an
s
ac
tio
n
d
ata,
L
STM
s
ex
ce
l
at
m
o
d
elin
g
tem
p
o
r
al
s
eq
u
e
n
ce
s
to
d
etec
t
ir
r
eg
u
lar
ities
o
v
er
tim
e,
an
d
ML
Ps
ar
e
ef
f
ec
tiv
e
at
lea
r
n
in
g
co
m
p
lex
n
o
n
-
lin
ea
r
r
elatio
n
s
h
ip
s
.
Ho
wev
er
,
p
r
i
o
r
s
tu
d
ies
h
av
e
lar
g
ely
f
o
c
u
s
ed
o
n
s
tan
d
alo
n
e
m
o
d
els
o
r
co
n
v
e
n
tio
n
al
en
s
em
b
le
ap
p
r
o
ac
h
es,
o
f
ten
with
o
u
t
ad
eq
u
ately
a
d
d
r
ess
in
g
a
cr
itical
ch
allen
g
e
in
th
is
d
o
m
ain
-
class
im
b
alan
ce
,
wh
er
e
f
r
au
d
u
len
t
tr
an
s
ac
tio
n
s
r
ep
r
esen
t
a
tin
y
f
r
ac
tio
n
o
f
th
e
o
v
er
all
d
ataset
[
4
]
,
[
5
]
.
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.
1
6
,
No
.
1
,
Feb
r
u
ar
y
20
2
6
:
3
1
1
-
320
312
T
h
is
s
tu
d
y
aim
s
to
ad
v
a
n
ce
th
e
s
tate
o
f
th
e
ar
t
b
y
p
r
o
p
o
s
in
g
a
h
y
b
r
id
en
s
em
b
le
f
r
am
ewo
r
k
co
m
b
in
in
g
C
NN,
L
STM
,
an
d
ML
P
ar
ch
itectu
r
es
in
teg
r
at
ed
with
a
n
o
v
el
s
am
p
lin
g
al
g
o
r
ith
m
s
p
ec
if
ically
d
esig
n
ed
to
b
alan
ce
th
e
d
ataset
an
d
im
p
r
o
v
e
f
r
au
d
d
etec
tio
n
p
er
f
o
r
m
an
ce
.
Un
lik
e
p
r
io
r
wo
r
k
s
,
o
u
r
ap
p
r
o
ac
h
lev
er
ag
es
th
e
co
m
p
lem
e
n
tar
y
s
tr
en
g
th
s
o
f
ea
c
h
m
o
d
el
ty
p
e
wh
ile
m
itig
atin
g
th
e
a
d
v
er
s
e
ef
f
ec
ts
o
f
s
k
ewe
d
class
d
is
tr
ib
u
tio
n
s
.
Ou
r
m
o
d
e
l
is
co
m
p
ar
ed
ag
ain
s
t
b
aselin
e
s
in
g
le
-
ar
ch
itectu
r
e
m
o
d
els
as
well
as
ex
is
tin
g
en
s
em
b
le
s
tr
ateg
ies to
d
em
o
n
s
tr
ate
its
s
u
p
er
io
r
ity
in
d
etec
tin
g
b
o
th
f
r
eq
u
en
t a
n
d
r
ar
e
f
r
au
d
p
atter
n
s
.
T
h
ese
co
n
tr
ib
u
tio
n
s
a
r
e
d
e
m
o
n
s
tr
ated
in
th
e
ex
p
er
im
e
n
tal
s
ec
tio
n
s
th
at
f
o
llo
w,
o
f
f
er
i
n
g
f
in
a
n
cial
in
s
titu
tio
n
s
a
s
ca
lab
le
an
d
i
n
tellig
en
t
s
o
lu
tio
n
to
d
etec
t
f
r
au
d
m
o
r
e
ac
cu
r
ately
a
n
d
r
eliab
ly
.
T
h
e
co
n
tr
i
b
u
tio
n
s
o
f
th
is
p
ap
er
ar
e
as f
o
llo
ws:
−
T
h
e
d
esig
n
o
f
a
r
o
b
u
s
t d
ee
p
le
ar
n
in
g
-
b
ased
en
s
em
b
le
th
at
f
u
s
es C
N
N,
L
STM
,
an
d
ML
P c
o
m
p
o
n
e
n
ts
.
−
T
h
e
in
teg
r
atio
n
o
f
a
p
r
o
p
o
s
ed
s
am
p
lin
g
m
eth
o
d
to
ef
f
ec
tiv
el
y
h
an
d
le
class
im
b
alan
ce
.
−
A
co
m
p
r
eh
e
n
s
iv
e
ev
alu
atio
n
o
f
th
e
p
r
o
p
o
s
ed
s
y
s
tem
ag
ai
n
s
t stan
d
ar
d
b
e
n
ch
m
ar
k
s
an
d
p
r
io
r
m
eth
o
d
s
.
−
A
d
etailed
an
aly
s
is
s
h
o
win
g
s
ig
n
if
ican
t
im
p
r
o
v
em
e
n
ts
in
ac
cu
r
ac
y
,
p
r
ec
is
io
n
,
a
n
d
r
e
d
u
ctio
n
o
f
f
alse
p
o
s
itiv
es.
T
h
e
s
tr
u
ctu
r
e
o
f
th
e
p
ap
e
r
u
n
f
o
ld
s
as
f
o
llo
ws:
s
ec
tio
n
2
p
r
o
v
id
es
a
co
m
p
r
e
h
en
s
iv
e
r
e
v
iew
o
f
ex
is
tin
g
liter
atu
r
e,
en
c
o
m
p
ass
in
g
c
r
ed
i
t
f
r
au
d
d
etec
tio
n
m
eth
o
d
o
lo
g
i
es,
im
b
alan
ce
d
d
atasets
,
an
d
s
am
p
lin
g
tec
h
n
iq
u
es.
Sectio
n
3
d
etails
th
e
m
eth
o
d
o
l
o
g
y
,
o
u
tlin
in
g
th
e
p
r
o
p
o
s
ed
s
am
p
lin
g
alg
o
r
ith
m
,
d
ataset
ch
ar
ac
ter
is
tics
,
an
d
th
e
ar
ch
itectu
r
al
d
etails
o
f
th
e
in
teg
r
ated
d
ee
p
lear
n
in
g
m
o
d
el
s
.
Sectio
n
4
p
r
esen
ts
ex
p
er
i
m
en
tal
r
esu
lts
an
d
an
aly
s
es.
Fin
ally
,
s
ec
tio
n
5
co
n
clu
d
es
th
e
p
ap
er
b
y
s
u
m
m
ar
izin
g
f
in
d
in
g
s
,
d
is
cu
s
s
in
g
th
e
s
y
s
tem
’
s
im
p
licatio
n
s
,
an
d
s
u
g
g
esti
n
g
a
v
en
u
es f
o
r
f
u
tu
r
e
r
esear
ch
.
2.
L
I
T
E
R
AT
U
RE
R
E
VI
E
W
S
Stu
d
ies
[
6
]
–
[
1
0
]
e
x
p
lo
r
e
d
ad
v
an
ce
d
m
ac
h
i
n
e
lear
n
in
g
an
d
d
ee
p
lear
n
i
n
g
tech
n
iq
u
es
f
o
r
ad
d
r
ess
in
g
class
im
b
alan
ce
in
cr
ed
it
ca
r
d
f
r
au
d
d
etec
tio
n
,
in
clu
d
in
g
f
ed
er
ated
lear
n
in
g
with
h
y
b
r
id
r
esam
p
lin
g
[
6
]
,
T
r
an
s
f
o
r
m
e
r
-
b
ased
m
o
d
els
[
7
]
,
AE
-
Net
with
ex
tr
e
m
e
g
r
ad
ie
n
t
b
o
o
s
tin
g
[
8
]
,
C
NNs
f
o
r
p
att
er
n
r
ec
o
g
n
itio
n
[
9
]
,
an
d
AFLCS
u
s
in
g
Ap
p
r
o
x
-
S
MO
T
E
with
C
NN
o
p
tim
izati
o
n
[
1
0
]
.
R
esear
ch
i
n
[
1
1
]
–
[
1
5
]
h
ig
h
lig
h
te
d
th
e
ef
f
ec
tiv
en
ess
o
f
g
r
ad
ien
t
b
o
o
s
tin
g
an
d
r
a
n
d
o
m
f
o
r
ests
[
1
1
]
,
o
p
tim
ized
ANN
m
o
d
els
f
o
r
d
ata
s
p
ar
s
ity
[
1
2
]
,
C
NN
-
L
STM
h
y
b
r
id
s
with
d
at
a
au
g
m
en
tatio
n
[
1
3
]
,
p
r
o
b
ab
ili
ty
-
b
ased
K
-
n
ea
r
est
n
eig
h
b
o
r
s
(
KNN)
f
o
r
e
f
f
icien
t
class
if
icatio
n
[
1
4
]
,
an
d
C
NNs
co
m
b
i
n
ed
with
d
ee
p
au
to
e
n
c
o
d
er
s
f
o
r
im
p
r
o
v
ed
f
r
au
d
d
etec
tio
n
[
1
5
]
.
Fu
r
th
er
ad
v
an
ce
m
e
n
ts
[
1
6
]
–
[
2
0
]
in
clu
d
e
a
p
ar
ticle
s
war
m
o
p
tim
izatio
n
(
PS
O)
-
o
p
tim
ized
s
tack
in
g
en
s
em
b
le
with
h
i
g
h
s
ca
lab
ilit
y
[
1
6
]
,
an
aly
s
is
o
f
DL
p
ar
am
eter
s
with
R
an
d
o
m
Fo
r
est
ac
h
iev
in
g
9
9
.
5
%
ac
cu
r
ac
y
[
1
7
]
,
im
p
ac
t
ass
es
s
m
en
t
o
f
th
e
“
T
im
e
”
f
ea
tu
r
e
ac
r
o
s
s
m
u
ltip
le
m
o
d
e
ls
[
1
8
]
,
s
u
p
er
io
r
g
r
ap
h
n
eu
r
al
n
etwo
r
k
(
GNN)
p
er
f
o
r
m
an
ce
i
n
g
r
ap
h
-
b
ased
an
o
m
aly
d
etec
tio
n
[
1
9
]
,
a
n
d
a
co
m
p
a
r
ativ
e
e
v
alu
atio
n
o
f
s
u
p
er
v
is
ed
v
s
.
d
ee
p
lear
n
in
g
m
eth
o
d
s
u
s
in
g
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
es
(
SVM)
,
K
-
n
ea
r
est
n
eig
h
b
o
r
s
(
KNN)
,
an
d
ar
tific
ial
n
eu
r
al
n
etwo
r
k
(
ANN)
[
2
0
]
.
Stu
d
ies
[
1
5
]
em
p
h
asized
h
y
b
r
i
d
d
ee
p
lear
n
in
g
ar
ch
itectu
r
es
s
u
ch
as
au
to
en
co
d
e
r
-
d
ee
p
n
eu
r
al
n
etwo
r
k
s
m
o
d
els
f
o
r
r
ea
l
-
tim
e
f
r
au
d
d
etec
tio
n
[
1
5
]
,
C
NNs
with
f
u
lly
co
n
n
ec
ted
l
ay
er
s
f
o
r
en
h
an
ce
d
r
ec
all
[
2
1
]
,
a
n
d
ML
Ps
with
d
r
o
p
o
u
t
an
d
au
g
m
en
tatio
n
to
h
a
n
d
le
o
v
er
f
itti
n
g
an
d
im
b
alan
c
e
[
2
2
]
.
C
o
llectiv
ely
,
th
ese
wo
r
k
s
d
em
o
n
s
tr
ate
th
e
g
r
o
win
g
e
f
f
ec
tiv
en
ess
o
f
h
y
b
r
id
,
e
n
s
em
b
le,
an
d
o
p
tim
i
za
tio
n
-
b
ased
d
ee
p
lear
n
in
g
s
tr
ateg
ies in
b
u
ild
in
g
r
o
b
u
s
t a
n
d
s
ca
lab
le
f
r
au
d
d
ete
ctio
n
s
y
s
tem
s
.
3.
M
E
T
H
O
DO
L
O
G
Y
T
h
e
p
r
o
p
o
s
ed
s
y
s
tem
p
er
f
o
r
m
s
a
co
m
p
ar
ativ
e
an
aly
s
is
o
f
cr
ed
it
ca
r
d
f
r
a
u
d
d
etec
tio
n
u
s
in
g
C
NN,
L
STM
,
ML
P
m
o
d
els
[
2
3
]
,
an
d
th
eir
en
s
em
b
le,
ap
p
lied
in
d
iv
id
u
ally
an
d
co
llectiv
el
y
to
ev
alu
ate
th
eir
ef
f
ec
tiv
en
ess
in
h
an
d
lin
g
a
h
ig
h
ly
im
b
alan
ce
d
d
ataset.
An
en
h
a
n
ce
d
s
am
p
lin
g
al
g
o
r
ith
m
is
in
tr
o
d
u
ce
d
to
b
alan
ce
th
e
d
ataset,
m
itig
atin
g
th
e
b
ias
f
r
o
m
n
o
n
-
f
r
au
d
u
len
t
tr
an
s
ac
tio
n
s
an
d
ex
p
o
s
in
g
t
h
e
m
o
d
els
to
a
m
o
r
e
r
ep
r
esen
tativ
e
class
d
is
tr
ib
u
tio
n
.
C
NN
ca
p
tu
r
es
s
p
atial
p
atter
n
s
,
L
STM
f
o
cu
s
es
o
n
tem
p
o
r
al
d
ep
en
d
en
cies
in
tr
an
s
ac
tio
n
s
eq
u
en
ce
s
,
an
d
M
L
P
s
er
v
es
as
a
b
aselin
e
class
i
f
ier
.
T
h
e
e
n
s
em
b
le
m
o
d
el
co
m
b
in
es
th
e
s
tr
en
g
th
s
o
f
th
ese
in
d
iv
id
u
al
m
o
d
els,
lev
er
ag
in
g
C
NN'
s
f
ea
tu
r
e
ex
tr
ac
tio
n
,
L
STM
'
s
s
eq
u
en
tial
lear
n
in
g
,
a
n
d
ML
P'
s
class
if
icatio
n
ef
f
icien
cy
.
Fig
u
r
e
1
illu
s
tr
ates th
e
p
r
o
p
o
s
ed
s
y
s
tem
'
s
f
lo
w
d
iag
r
am
,
em
p
h
asizi
n
g
th
e
im
p
o
r
ta
n
ce
o
f
b
alan
ce
d
d
ata
in
i
m
p
r
o
v
in
g
m
o
d
el
ac
cu
r
ac
y
an
d
r
eliab
ilit
y
.
3
.
1
.
Da
t
a
c
o
llect
io
n
T
h
e
d
ataset
u
s
ed
in
th
is
s
t
u
d
y
[
2
4
]
co
m
p
r
is
es
2
8
4
,
8
0
7
cr
ed
it
ca
r
d
tr
an
s
ac
tio
n
s
b
y
E
u
r
o
p
ea
n
ca
r
d
h
o
ld
e
r
s
o
v
er
two
d
a
y
s
in
Sep
tem
b
er
2
0
1
3
,
o
f
wh
ic
h
o
n
ly
4
9
2
(
0
.
1
7
2
%)
ar
e
f
r
a
u
d
u
l
en
t,
h
ig
h
lig
h
tin
g
a
s
ev
er
e
class
im
b
alan
ce
.
I
t
in
clu
d
es
o
n
ly
n
u
m
e
r
ical
f
ea
tu
r
es
tr
an
s
f
o
r
m
ed
v
ia
p
r
in
cip
al
co
m
p
o
n
en
t
an
aly
s
is
(
PC
A)
,
lab
eled
V1
to
V2
8
,
alo
n
g
with
t
r
an
s
ac
tio
n
am
o
u
n
t
an
d
tim
estam
p
,
allo
win
g
f
o
r
th
e
ca
p
tu
r
e
o
f
co
m
p
lex
b
e
h
av
io
r
al
p
atter
n
s
.
T
h
e
o
u
tco
m
e
v
ar
iab
le,
r
ef
e
r
r
ed
to
as
"Cl
as
s
,
"
is
b
in
ar
y
—
wh
er
e
a
v
alu
e
o
f
0
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
C
r
ed
it c
a
r
d
fr
a
u
d
d
a
ta
a
n
a
lysi
s
u
s
in
g
p
r
o
p
o
s
ed
s
a
mp
lin
g
a
lg
o
r
ith
m
…
(
A
ye
A
ye
K
h
in
e
)
313
in
d
icate
s
a
g
en
u
i
n
e
tr
an
s
ac
tio
n
an
d
a
v
alu
e
o
f
1
f
la
g
a
tr
a
n
s
ac
tio
n
as
f
r
au
d
u
len
t.
T
h
is
r
ich
an
d
m
u
ltifa
ce
ted
d
ataset
s
tr
u
ctu
r
e
en
ab
les
d
etailed
an
aly
s
is
an
d
m
o
d
elin
g
f
o
r
d
etec
tin
g
an
d
p
r
e
v
en
tin
g
f
r
au
d
u
len
t
f
i
n
an
cial
ac
tiv
ities
.
Fig
u
r
e
2
d
escr
ib
es th
e
cr
ed
it c
ar
d
f
r
a
u
d
d
ataset.
Fig
u
r
e
1
.
Flo
w
d
ia
g
r
am
f
o
r
p
r
o
p
o
s
ed
cr
e
d
it c
ar
d
f
r
au
d
d
ata
an
aly
s
is
s
y
s
tem
Fig
u
r
e
2
.
Data
s
et
f
o
r
cr
ed
it c
ar
d
f
r
au
d
d
ata
3
.
2
.
Da
t
a
prepro
ce
s
s
ing
Data
n
o
r
m
aliza
tio
n
is
a
cr
u
cial
p
r
ep
r
o
ce
s
s
in
g
s
tep
in
m
ac
h
in
e
lear
n
in
g
th
at
en
s
u
r
es
all
f
ea
tu
r
es
ar
e
o
n
a
co
m
p
ar
a
b
le
s
ca
le,
p
r
ev
e
n
tin
g
f
ea
tu
r
es
with
lar
g
er
r
an
g
es
f
r
o
m
d
o
m
in
ati
n
g
th
e
lear
n
in
g
p
r
o
ce
s
s
.
I
n
th
is
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.
1
6
,
No
.
1
,
Feb
r
u
ar
y
20
2
6
:
3
1
1
-
320
314
s
y
s
tem
,
m
in
-
m
ax
n
o
r
m
aliza
ti
o
n
is
ap
p
lied
t
o
s
ca
le
d
ata
w
ith
in
th
e
r
an
g
e
o
f
0
t
o
1
,
s
tan
d
ar
d
izin
g
f
ea
tu
r
e
v
alu
es
f
o
r
im
p
r
o
v
ed
m
o
d
el
p
e
r
f
o
r
m
a
n
ce
an
d
co
m
p
u
tatio
n
al
ef
f
icien
cy
[
2
5
]
.
T
h
is
co
n
s
is
ten
t
s
ca
lin
g
n
o
t
o
n
ly
ac
ce
ler
ates
m
o
d
el
co
n
v
er
g
en
c
e
d
u
r
i
n
g
tr
ai
n
in
g
b
u
t
also
h
el
p
s
m
ain
tain
th
e
in
teg
r
ity
o
f
f
ea
tu
r
e
r
elatio
n
s
h
ip
s
ac
r
o
s
s
d
if
f
er
en
t le
ar
n
in
g
alg
o
r
ith
m
s
.
T
h
e
n
o
r
m
aliza
tio
n
f
o
r
m
u
la
u
s
ed
is
:
′
=
−
−
(
_
−
_
)
+
_
(
1
)
T
h
is
n
o
r
m
aliza
tio
n
p
r
o
ce
s
s
ad
ju
s
ts
ea
ch
d
ata
p
o
in
t
v
to
a
n
e
w
v
alu
e
v′
th
at
f
its
with
in
th
e
n
o
r
m
alize
d
r
an
g
e.
Fo
llo
win
g
d
ata
n
o
r
m
a
lizatio
n
,
th
e
s
y
s
tem
ap
p
lies
a
p
r
o
p
o
s
ed
s
am
p
lin
g
tech
n
iq
u
e
to
ad
d
r
ess
d
ata
im
b
alan
ce
.
T
h
is
s
am
p
lin
g
m
e
th
o
d
e
n
s
u
r
es
th
at
th
e
d
ataset
u
s
ed
f
o
r
m
o
d
el
tr
ai
n
in
g
is
m
o
r
e
r
e
p
r
esen
tativ
e,
en
h
an
cin
g
th
e
p
e
r
f
o
r
m
an
ce
an
d
ac
cu
r
ac
y
o
f
t
h
e
f
r
a
u
d
d
etec
ti
o
n
m
o
d
els.
3
.
3
.
P
r
o
po
s
ed
s
a
m
pli
ng
a
lg
o
rit
hm
C
las
s
im
b
alan
ce
in
th
e
C
r
ed
it
C
ar
d
d
ataset
ca
n
lead
to
b
iase
d
m
ac
h
in
e
lear
n
in
g
m
o
d
els
th
at
s
tr
u
g
g
le
to
co
r
r
ec
tly
class
if
y
m
in
o
r
ity
class
in
s
tan
ce
s
,
r
esu
lt
in
g
in
h
ig
h
f
alse
-
n
eg
ativ
e
r
ates
an
d
r
ed
u
ce
d
s
en
s
itiv
ity
.
T
h
is
s
y
s
tem
tack
les
class
im
b
alan
ce
u
s
in
g
th
e
p
r
o
p
o
s
ed
s
am
p
lin
g
ap
p
r
o
ac
h
,
av
o
id
in
g
th
e
o
v
e
r
f
itti
n
g
is
s
u
es
o
f
o
v
er
s
am
p
lin
g
an
d
th
e
d
ata
l
o
s
s
r
is
k
s
o
f
u
n
d
er
s
am
p
lin
g
.
I
n
Alg
o
r
ith
m
1
,
th
e
p
r
o
p
o
s
ed
d
ata
b
alan
cin
g
alg
o
r
ith
m
is
s
h
o
wn
.
T
h
e
p
r
o
p
o
s
ed
s
y
s
tem
in
tr
o
d
u
ce
s
a
p
r
o
p
o
s
ed
s
am
p
lin
g
alg
o
r
ith
m
to
b
a
lan
ce
th
e
d
ataset
b
y
g
en
er
atin
g
s
y
n
th
etic
m
in
o
r
ity
class
in
s
tan
ce
s
,
en
s
u
r
in
g
eq
u
al
r
ep
r
esen
tatio
n
o
f
n
o
r
m
al
an
d
f
r
au
d
u
len
t
tr
an
s
ac
tio
n
s
.
B
y
r
em
o
v
in
g
t
h
e
lab
el
f
r
o
m
th
e
f
ea
tu
r
e
v
ec
to
r
,
th
e
s
y
s
tem
p
r
e
v
en
ts
d
ata
leak
ag
e
an
d
m
ain
tain
s
m
o
d
el
in
teg
r
ity
wh
ile
a
p
p
ly
in
g
th
e
s
am
p
lin
g
tech
n
iq
u
e
ef
f
e
ctiv
ely
.
Un
lik
e
tr
a
d
itio
n
al
o
v
e
r
s
am
p
lin
g
m
eth
o
d
s
,
th
is
ap
p
r
o
ac
h
cr
ea
tes
n
ew,
s
y
n
th
etic
s
am
p
les
r
ath
er
th
an
d
u
p
licatin
g
ex
is
tin
g
o
n
es,
r
ed
u
cin
g
o
v
e
r
f
itti
n
g
an
d
im
p
r
o
v
in
g
m
o
d
el
p
er
f
o
r
m
a
n
ce
.
Alg
o
r
ith
m
1
.
Pro
p
o
s
ed
d
ata
b
a
lan
cin
g
alg
o
r
ith
m
Step 1. Computes the quantity of normal and anomalous entries in the training dataset.
Step
2.
If
th
e
co
un
t
of
th
e
no
rm
al
cl
as
s
is
gr
ea
te
r
th
an
th
e
co
un
t
of
th
e
ab
no
rm
al
cl
as
s,
from step 3 to step 9 is performed.
Step 3. Calculates t
he difference (num) between the counts of the two classes
.
Step 4. For each iteration (from 0 to num),
Step 5. Extracts a feature vector from the abnormal class.
Step 6. Eliminates the last component of the feature vector that was extracted.
Step 7. Acquires a feature value from the neighboring anomalous data point.
Step 8. Integrates this feature into the updated feature vector.
Step
9.
Th
e
up
da
te
d
fe
at
ur
e
ve
ct
or
is
ap
pe
nd
ed
to
th
e
ab
no
rm
al
da
ta
se
t
,
al
on
g
wi
th
th
e
label 1.
Step
10.
Wh
en
a
cl
as
s
im
ba
la
nc
e
is
de
te
ct
ed
,
wi
t
h
th
e
ab
no
rm
al
cl
as
s
be
in
g
pr
ed
om
in
an
t,
steps 11 to 17 are performed.
Step 11. Computes the numerical difference (num) between the class counts.
Step 12. For each iteration (from 0 to num),
Step 13. Extracts a feature vector from the normal class.
Step 14. The extracted feature vector is modified by removing its last entry.
Step 15. A feature value is obtained from a nearby normal instance.
Step 16. Integrates this feature into the updated feature vector.
Step 17. The updated feature vector and label 0 are added to the normal dataset.
A
k
ey
s
tep
in
b
alan
cin
g
cr
ed
i
t
ca
r
d
tr
an
s
ac
tio
n
d
atasets
in
v
o
lv
es
r
em
o
v
in
g
th
e
f
in
al
elem
en
t
o
f
th
e
f
ea
tu
r
e
v
ec
to
r
.
T
h
is
last
ele
m
en
t
ty
p
ically
r
ep
r
esen
ts
th
e
lab
el,
s
u
c
h
as
"
f
r
au
d
u
len
t"
o
r
"n
o
n
-
f
r
au
d
u
len
t,"
wh
ich
in
d
icate
s
wh
eth
er
a
tr
an
s
ac
tio
n
is
f
r
au
d
u
len
t o
r
n
o
t.
B
y
ex
clu
d
in
g
th
is
lab
el
f
r
o
m
th
e
f
ea
tu
r
e
v
ec
to
r
,
th
e
p
r
o
p
o
s
ed
s
am
p
lin
g
tec
h
n
iq
u
e
s
ca
n
b
e
ap
p
lied
m
o
r
e
ef
f
ec
t
iv
ely
to
ad
d
r
ess
class
im
b
alan
ce
.
R
em
o
v
in
g
th
e
lab
el
f
r
o
m
th
e
f
ea
tu
r
e
v
ec
to
r
is
cr
u
cial
f
o
r
s
ev
e
r
al
r
ea
s
o
n
s
.
Firstl
y
,
it
p
r
ev
e
n
ts
d
ata
lea
k
ag
e,
en
s
u
r
in
g
th
at
th
e
m
o
d
el
is
tr
ain
e
d
o
n
f
ea
tu
r
es
al
o
n
e
with
o
u
t
th
e
in
f
lu
e
n
ce
o
f
t
h
e
tar
g
et
v
ar
iab
le.
T
h
is
p
r
ac
tice
ad
h
er
es
t
o
m
o
d
el
in
p
u
t
r
eq
u
ir
em
e
n
ts
,
m
ain
tain
i
n
g
th
e
in
teg
r
ity
o
f
t
h
e
d
ata
an
d
en
s
u
r
i
n
g
th
at
th
e
m
o
d
el
lea
r
n
s
f
r
o
m
t
h
e
ac
tu
a
l
f
ea
tu
r
es
r
ath
er
th
an
b
ein
g
b
i
ased
b
y
th
e
lab
els.
Seco
n
d
ly
,
it
en
s
u
r
es
th
at
s
y
n
th
etic
s
a
m
p
les
ar
e
g
en
er
ated
s
o
lely
f
r
o
m
g
en
u
in
e
tr
an
s
ac
ti
o
n
ch
a
r
ac
ter
is
tics
r
ath
er
th
an
b
ein
g
in
f
lu
en
ce
d
b
y
class
id
en
tifie
r
s
.
T
h
ir
d
ly
,
it
p
r
eser
v
es
a
clea
r
s
ep
ar
atio
n
b
etwe
en
f
ea
tu
r
es
an
d
lab
els,
w
h
ich
is
es
s
en
tial
f
o
r
r
ep
r
o
d
u
c
ib
ilit
y
an
d
co
r
r
ec
t
im
p
lem
en
tatio
n
o
f
th
e
s
am
p
lin
g
alg
o
r
ith
m
.
Fin
ally
,
th
is
ap
p
r
o
ac
h
s
u
p
p
o
r
ts
f
air
m
o
d
el
ev
alu
atio
n
b
y
p
r
ev
en
tin
g
ar
tific
ially
in
f
lated
p
er
f
o
r
m
an
ce
th
at
co
u
l
d
ar
is
e
if
lab
el
in
f
o
r
m
atio
n
leak
ed
in
t
o
th
e
f
ea
tu
r
e
s
p
ac
e
.
T
h
e
p
r
o
p
o
s
ed
s
am
p
lin
g
alg
o
r
ith
m
is
a
d
ap
tab
le
t
o
b
o
th
n
u
m
e
r
ical
an
d
ca
teg
o
r
ical
d
ata.
I
t
allo
ws
f
o
r
ad
ju
s
tm
en
ts
b
ased
o
n
s
p
ec
if
i
c
n
ee
d
s
,
p
ar
ticu
lar
ly
b
y
m
o
d
if
y
in
g
th
e
m
et
h
o
d
u
s
ed
to
g
en
er
ate
s
y
n
t
h
etic
s
am
p
les.
Un
lik
e
tr
ad
itio
n
al
o
v
er
s
am
p
lin
g
m
eth
o
d
s
th
at
s
i
m
p
ly
d
u
p
licate
e
x
is
tin
g
s
am
p
les,
th
is
alg
o
r
ith
m
f
o
cu
s
es
o
n
cr
ea
tin
g
n
ew,
s
y
n
t
h
etic
s
am
p
les.
T
h
is
s
am
p
lin
g
alg
o
r
ith
m
h
elp
s
in
r
ed
u
cin
g
th
e
r
is
k
o
f
o
v
er
f
itti
n
g
b
y
in
tr
o
d
u
cin
g
v
ar
ia
b
ilit
y
in
to
th
e
d
ataset
r
ath
er
th
a
n
m
er
ely
in
cr
ea
s
in
g
th
e
n
u
m
b
er
o
f
i
d
en
t
ical
in
s
tan
ce
s
.
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
C
r
ed
it c
a
r
d
fr
a
u
d
d
a
ta
a
n
a
lysi
s
u
s
in
g
p
r
o
p
o
s
ed
s
a
mp
lin
g
a
lg
o
r
ith
m
…
(
A
ye
A
ye
K
h
in
e
)
315
3
.
4
.
Dee
p e
ns
em
ble le
a
rning
-
ba
s
ed
cr
edit
ca
rd
det
ec
t
io
n
T
o
im
p
r
o
v
e
f
r
au
d
d
etec
tio
n
p
er
f
o
r
m
an
ce
,
t
h
e
d
ataset
is
f
i
r
s
t
b
alan
ce
d
u
s
in
g
a
p
r
o
p
o
s
ed
s
am
p
lin
g
alg
o
r
ith
m
th
at
in
cr
ea
s
es
th
e
r
ep
r
esen
tatio
n
o
f
f
r
a
u
d
u
len
t
tr
an
s
ac
tio
n
s
with
o
u
t
elim
in
atin
g
l
eg
itima
te
s
am
p
les,
ef
f
ec
tiv
ely
ad
d
r
ess
in
g
class
im
b
alan
ce
.
T
h
e
b
ala
n
ce
d
d
ata
is
th
en
d
i
v
id
ed
in
to
tr
ai
n
in
g
an
d
test
in
g
s
ets
to
ass
es
s
th
e
p
er
f
o
r
m
a
n
ce
o
f
t
h
r
ee
in
d
iv
id
u
al
d
ee
p
lear
n
in
g
m
o
d
els:
C
NN,
L
STM
,
an
d
ML
P.
T
h
ese
m
o
d
els
ar
e
ch
o
s
en
f
o
r
th
ei
r
r
esp
ec
tiv
e
s
tr
en
g
th
s
-
C
NN
f
o
r
ca
p
tu
r
i
n
g
s
p
atial
p
atter
n
s
in
tr
an
s
ac
tio
n
f
ea
tu
r
es,
L
STM
f
o
r
m
o
d
elin
g
tem
p
o
r
al
d
ep
e
n
d
en
cies,
an
d
ML
P
f
o
r
h
an
d
lin
g
co
m
p
lex
n
o
n
lin
ea
r
r
elatio
n
s
h
ip
s
.
An
e
n
s
em
b
le
m
o
d
el
is
th
en
co
n
s
tr
u
cted
b
y
in
teg
r
atin
g
th
e
o
u
tp
u
ts
o
f
th
e
s
e
th
r
ee
m
o
d
els
to
en
h
an
ce
o
v
er
all
class
if
icatio
n
ac
cu
r
ac
y
.
T
h
is
en
s
em
b
le
is
ev
alu
ated
u
s
in
g
th
r
ee
ag
g
r
eg
at
io
n
s
tr
ateg
ies:
weig
h
ted
a
v
er
a
g
e,
wh
er
e
s
tr
o
n
g
e
r
m
o
d
els
h
av
e
g
r
ea
te
r
in
f
lu
e
n
ce
o
n
th
e
f
in
al
p
r
ed
ictio
n
;
u
n
weig
h
ted
av
er
a
g
e,
wh
ich
tr
ea
ts
a
ll
m
o
d
els
eq
u
ally
;
an
d
u
n
weig
h
ted
m
ajo
r
ity
v
o
tin
g
,
wh
e
r
e
th
e
f
in
al
class
is
d
eter
m
in
ed
b
y
t
h
e
m
ajo
r
ity
o
f
in
d
iv
id
u
al
m
o
d
el
p
r
ed
ictio
n
s
.
T
h
is
ap
p
r
o
ac
h
en
s
u
r
es r
o
b
u
s
t a
n
d
f
air
d
ec
is
io
n
-
m
ak
in
g
in
f
r
au
d
d
etec
tio
n
.
4.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
T
h
e
p
er
f
o
r
m
an
ce
o
f
C
NN,
L
S
T
M,
ML
P,
an
d
th
eir
en
s
em
b
le
m
o
d
el
is
ev
alu
ated
f
o
r
cr
ed
it c
ar
d
f
r
au
d
d
etec
tio
n
,
with
th
e
d
ataset
s
p
lit
in
to
8
0
%
tr
ain
in
g
an
d
2
0
%
test
in
g
p
o
r
tio
n
s
.
T
h
e
en
s
e
m
b
le
m
o
d
el,
wh
ic
h
co
m
b
in
es
th
e
s
tr
en
g
th
s
o
f
e
ac
h
in
d
i
v
id
u
al
m
o
d
el,
is
ass
ess
ed
u
s
in
g
m
etr
ics
lik
e
p
r
e
cisi
o
n
,
r
ec
all,
a
n
d
F1
-
s
co
r
e,
with
test
s
co
n
d
u
cted
o
n
b
o
th
th
e
o
r
ig
in
al
an
d
b
alan
ce
d
d
atasets
to
ev
alu
ate
th
e
im
p
ac
t
o
f
th
e
s
am
p
lin
g
s
tr
ateg
y
.
T
ab
le
1
d
escr
ib
es
th
e
p
ar
am
eter
co
n
f
i
g
u
r
atio
n
s
o
f
th
e
s
y
s
tem
.
Fig
u
r
e
3
d
escr
ib
es
th
e
r
ec
eiv
er
o
p
er
atin
g
ch
ar
ac
ter
i
s
tic
(
R
OC
)
c
u
r
v
e
o
f
C
NN
with
an
d
with
o
u
t
b
ala
n
cin
g
alg
o
r
ith
m
.
T
ab
le
2
d
escr
ib
es
th
e
r
esu
lts
o
f
C
NN
with
an
d
with
o
u
t
b
ala
n
cin
g
al
g
o
r
ith
m
s
.
Fig
u
r
e
4
d
escr
ib
es
t
h
e
R
OC
C
u
r
v
e
o
f
L
STM
with
an
d
with
o
u
t
b
alan
cin
g
alg
o
r
it
h
m
.
T
ab
le
3
d
escr
ib
es
th
e
r
esu
lts
o
f
L
ST
M
with
an
d
with
o
u
t
b
alan
cin
g
alg
o
r
ith
m
.
Fig
u
r
e
5
d
escr
ib
es
th
e
R
OC
C
u
r
v
e
o
f
ML
P
with
an
d
with
o
u
t
b
alan
cin
g
alg
o
r
ith
m
.
T
ab
le
4
d
escr
ib
es
th
e
r
esu
lts
o
f
ML
P
with
a
n
d
with
o
u
t
b
al
an
cin
g
al
g
o
r
ith
m
.
T
ab
le
5
d
e
s
cr
ib
es
th
e
an
aly
s
is
r
esu
lts
o
f
en
s
em
b
le
with
weig
h
ted
av
er
ag
e,
u
n
weig
h
ted
av
e
r
ag
e
an
d
u
n
weig
h
ted
m
ajo
r
ity
v
o
tin
g
b
y
p
r
o
p
o
s
ed
s
am
p
lin
g
alg
o
r
ith
m
.
T
ab
le
1
.
Par
am
eter
c
o
n
f
i
g
u
r
ati
o
n
s
P
a
r
a
me
t
e
r
s
C
N
N
LSTM
M
LP
N
u
mb
e
r
o
f
F
i
l
t
e
r
s :
1
6
U
n
i
t
s:
5
0
U
n
i
t
s:
1
6
A
c
t
i
v
a
t
i
o
n
f
u
n
c
t
i
o
n
si
g
m
o
i
d
si
g
m
o
i
d
si
g
m
o
i
d
D
r
o
p
o
u
t
r
a
t
e
0
.
5
-
-
O
p
t
i
mi
z
e
r
A
d
a
m
A
d
a
m
A
d
a
m
Ep
o
c
h
s
10
10
10
B
a
t
c
h
si
z
e
64
64
64
K
e
r
n
e
l
si
z
e
3
-
-
P
o
o
l
i
n
g
st
r
a
t
e
g
y
M
a
x
P
o
o
l
i
n
g
(
p
o
o
l
si
z
e
=
2
)
-
-
Le
a
r
n
i
n
g
r
a
t
e
0
.
0
0
1
0
.
0
0
1
0
.
0
0
1
Lo
ss f
u
n
c
t
i
o
n
_
_
_
Wi
tho
ut
prop
os
e
d
s
a
m
pl
in
g
Wi
th
prop
o
s
e
d
s
a
m
pl
in
g
Fig
u
r
e
3
.
R
OC
cu
r
v
e
f
o
r
C
NN
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.
1
6
,
No
.
1
,
Feb
r
u
ar
y
20
2
6
:
3
1
1
-
320
316
T
ab
le
2
.
C
o
m
p
a
r
is
o
n
b
etwe
en
n
o
r
m
al
an
d
f
r
a
u
d
u
s
in
g
C
NN
C
N
N
N
o
r
mal
F
r
a
u
d
P
r
e
c
i
s
i
o
n
R
e
c
a
l
l
F1
-
S
c
o
r
e
P
r
e
c
i
s
i
o
n
R
e
c
a
l
l
F1
-
S
c
o
r
e
W
i
t
h
o
u
t
P
r
o
p
o
se
d
S
a
m
p
l
i
n
g
0
.
8
9
0
.
9
1
0
.
9
0
.
8
8
0
.
7
5
0
.
8
0
W
i
t
h
P
r
o
p
o
se
d
S
a
mp
l
i
n
g
0
.
9
6
0
.
9
7
0
.
9
6
0
.
9
6
0
.
9
7
0
.
9
6
Wi
tho
ut
prop
os
e
d
s
a
m
pl
in
g
Wi
th
prop
o
s
e
d
s
a
m
pl
in
g
Fig
u
r
e
4
.
R
OC
cu
r
v
e
f
o
r
L
ST
M
T
ab
le
3
.
C
o
m
p
a
r
is
o
n
b
etwe
en
n
o
r
m
al
an
d
f
r
a
u
d
u
s
in
g
L
STM
LSTM
N
o
r
mal
F
r
a
u
d
P
r
e
c
i
s
i
o
n
R
e
c
a
l
l
F1
-
S
c
o
r
e
P
r
e
c
i
s
i
o
n
R
e
c
a
l
l
F1
-
S
c
o
r
e
W
i
t
h
o
u
t
P
r
o
p
o
se
d
S
a
m
p
l
i
n
g
0
.
9
1
0
.
7
4
0
.
7
7
0
.
8
8
0
.
7
3
0
.
8
0
W
i
t
h
P
r
o
p
o
se
d
S
a
mp
l
i
n
g
0
.
9
8
0
.
9
0
0
.
9
4
0
.
9
8
0
.
9
1
0
.
9
4
Wi
tho
ut
prop
os
e
d
s
a
m
pl
in
g
Wi
th
prop
o
s
e
d
s
a
m
pl
in
g
Fig
u
r
e
5
.
R
OC
c
u
r
v
e
f
o
r
ML
P
T
ab
le
4
.
C
o
m
p
a
r
is
o
n
b
etwe
en
n
o
r
m
al
an
d
f
r
a
u
d
u
s
in
g
ML
P
M
LP
N
o
r
mal
F
r
a
u
d
P
r
e
c
i
s
i
o
n
R
e
c
a
l
l
F1
-
S
c
o
r
e
P
r
e
c
i
s
i
o
n
R
e
c
a
l
l
F1
-
S
c
o
r
e
W
i
t
h
o
u
t
P
r
o
p
o
se
d
S
a
m
p
l
i
n
g
0
.
8
2
0
.
8
3
0
.
8
2
0
.
8
0
0
.
8
3
0
.
8
1
W
i
t
h
P
r
o
p
o
se
d
S
a
mp
l
i
n
g
0
.
9
0
0
.
9
4
0
.
9
2
0
.
9
2
0
.
9
1
0
.
9
1
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
C
r
ed
it c
a
r
d
fr
a
u
d
d
a
ta
a
n
a
lysi
s
u
s
in
g
p
r
o
p
o
s
ed
s
a
mp
lin
g
a
lg
o
r
ith
m
…
(
A
ye
A
ye
K
h
in
e
)
317
T
ab
le
5
.
An
aly
s
is
r
esu
lts
o
f
en
s
em
b
le
m
o
d
el
En
se
mb
l
e
M
o
d
e
l
N
o
r
mal
F
r
a
u
d
P
r
e
c
i
s
i
o
n
R
e
c
a
l
l
F1
-
S
c
o
r
e
P
r
e
c
i
s
i
o
n
R
e
c
a
l
l
F1
-
S
c
o
r
e
W
e
i
g
h
t
e
d
A
v
e
r
a
g
e
1
1
1
1
1
1
U
n
w
e
i
g
h
t
e
d
A
v
e
r
a
g
e
1
1
1
1
1
1
U
n
w
e
i
g
h
t
e
d
M
a
j
o
r
i
t
y
V
o
t
i
n
g
1
1
1
1
1
1
T
h
e
r
esu
lts
s
h
o
w
th
at
all
th
r
e
e
en
s
em
b
le
a
p
p
r
o
a
ch
es
—
W
ei
g
h
ted
a
v
er
ag
e,
u
n
weig
h
ted
av
er
ag
e,
a
n
d
u
n
weig
h
ted
m
ajo
r
ity
v
o
tin
g
—
ac
h
iev
e
p
er
f
ec
t
s
co
r
es
o
f
1
.
0
f
o
r
b
o
th
n
o
r
m
al
an
d
f
r
a
u
d
u
len
t
tr
a
n
s
ac
tio
n
s
,
in
d
icatin
g
h
ig
h
ac
cu
r
ac
y
a
n
d
n
o
m
is
class
if
icatio
n
s
.
T
h
e
en
s
em
b
le
m
o
d
els,
d
esp
i
te
th
eir
d
if
f
er
e
n
t
m
eth
o
d
o
l
o
g
ies,
d
em
o
n
s
tr
ate
r
o
b
u
s
tn
ess
an
d
r
eliab
ilit
y
in
a
cc
u
r
ately
id
en
tify
i
n
g
tr
an
s
ac
ti
o
n
ca
teg
o
r
ies.
T
h
e
p
r
o
p
o
s
ed
s
am
p
lin
g
m
eth
o
d
s
ig
n
if
ican
tly
im
p
r
o
v
es
class
if
icatio
n
m
etr
ics
f
o
r
in
d
iv
id
u
al
m
o
d
els,
in
clu
d
in
g
C
NN,
L
STM
,
an
d
ML
P,
e
n
h
an
cin
g
p
r
ec
is
io
n
,
r
ec
all,
an
d
F1
-
s
co
r
es
f
o
r
b
o
th
n
o
r
m
al
an
d
f
r
au
d
u
len
t
tr
an
s
ac
tio
n
s
.
Ov
er
all,
th
e
e
n
s
em
b
le
m
o
d
els,
s
u
p
p
o
r
te
d
b
y
th
e
s
am
p
lin
g
m
eth
o
d
,
d
eliv
er
o
p
tim
al
p
er
f
o
r
m
an
ce
in
f
r
au
d
d
etec
tio
n
,
ef
f
ec
tiv
ely
ad
d
r
ess
in
g
class
im
b
alan
ce
an
d
en
s
u
r
in
g
p
r
ec
is
e
an
d
r
eliab
le
p
r
ed
ictio
n
s
.
T
ab
le
6
d
ep
icts
th
e
p
e
r
f
o
r
m
an
ce
an
aly
s
is
o
f
en
h
an
ce
d
s
am
p
lin
g
with
o
th
e
r
s
am
p
lin
g
ap
p
r
o
ac
h
es
f
o
r
cr
ed
it
ca
r
d
d
ataset.
T
ab
le
6
p
r
esen
ts
a
co
m
p
ar
ativ
e
p
er
f
o
r
m
an
ce
an
aly
s
is
o
f
v
ar
io
u
s
s
am
p
lin
g
tech
n
iq
u
es
ap
p
lied
to
f
r
au
d
d
etec
tio
n
m
o
d
els,
h
ig
h
lig
h
tin
g
th
eir
im
p
ac
t
o
n
p
r
ec
is
io
n
,
r
ec
all,
an
d
F1
-
S
co
r
e
f
o
r
b
o
th
n
o
r
m
al
an
d
f
r
au
d
u
len
t
tr
an
s
a
ctio
n
s
.
T
h
e
p
r
o
p
o
s
ed
en
h
a
n
ce
d
s
am
p
lin
g
m
eth
o
d
s
ig
n
i
f
ican
tly
o
u
tp
er
f
o
r
m
s
all
o
th
er
s
b
y
ac
h
iev
in
g
p
er
f
ec
t
s
co
r
es
(
1
.
0
0
)
ac
r
o
s
s
all
m
etr
ics
f
o
r
b
o
th
class
es,
ef
f
ec
tiv
ely
ad
d
r
ess
in
g
class
im
b
alan
ce
with
o
u
t
s
ac
r
if
icin
g
leg
itima
te
tr
an
s
ac
tio
n
ac
cu
r
ac
y
.
I
n
co
n
tr
ast,
m
eth
o
d
s
lik
e
u
n
d
er
s
am
p
lin
g
an
d
Nea
r
Miss
s
u
f
f
er
f
r
o
m
e
x
tr
e
m
ely
lo
w
p
r
ec
is
io
n
f
o
r
f
r
a
u
d
d
etec
tio
n
(
0
.
0
4
an
d
0
.
0
5
,
r
esp
ec
tiv
ely
)
,
wh
ile
SMOT
E
,
ADASYN,
an
d
T
o
m
ek
L
in
k
s
y
ield
m
o
d
er
ate
i
m
p
r
o
v
e
m
en
ts
b
u
t
s
till
f
all
s
h
o
r
t
co
m
p
ar
ed
t
o
th
e
en
h
an
ce
d
s
am
p
lin
g
ap
p
r
o
ac
h
.
T
h
e
co
n
s
is
ten
t
1
.
0
0
/1
.
0
0
/1
.
0
0
f
o
r
th
e
No
r
m
al
class
ac
r
o
s
s
m
u
ltip
le
s
am
p
lin
g
m
eth
o
d
s
is
lar
g
el
y
d
u
e
to
th
e
e
x
tr
em
e
class
im
b
alan
ce
in
th
e
UL
B
d
ataset,
wh
er
e
th
e
v
ast m
ajo
r
ity
o
f
s
am
p
les
b
elo
n
g
t
o
th
e
No
r
m
al
class
,
m
ak
in
g
it
ea
s
y
f
o
r
m
o
d
els
to
class
if
y
th
em
co
r
r
ec
tly
.
Ho
wev
er
,
th
is
d
o
es
n
o
t
r
ef
lect
tr
u
e
m
o
d
el
r
o
b
u
s
tn
es
s
,
as
th
e
m
in
o
r
ity
Fra
u
d
cla
s
s
r
em
ain
s
ch
allen
g
in
g
,
a
n
d
in
s
o
m
e
ca
s
es
th
e
im
b
alan
ce
co
m
b
in
ed
with
r
esam
p
lin
g
m
a
y
also
in
tr
o
d
u
ce
h
i
d
d
en
leak
ag
e
o
r
b
ias.
T
ab
le
6
.
Per
f
o
r
m
a
n
ce
an
aly
s
i
s
r
esu
lts
with
v
ar
io
u
s
s
am
p
lin
g
ap
p
r
o
ac
h
es
M
o
d
e
l
s
N
o
r
mal
F
r
a
u
d
P
r
e
c
i
s
i
o
n
R
e
c
a
l
l
F1
-
S
c
o
r
e
P
r
e
c
i
s
i
o
n
R
e
c
a
l
l
F1
-
S
c
o
r
e
O
r
i
g
i
n
a
l
U
n
b
a
l
a
n
c
e
d
1
1
1
0
.
9
4
0
.
8
2
0
.
8
7
O
v
e
r
sam
p
l
i
n
g
1
1
1
0
.
9
5
0
.
7
7
0
.
8
5
U
n
d
e
r
sam
p
l
i
n
g
1
0
.
9
6
0
.
9
8
0
.
0
4
0
.
9
2
0
.
0
8
S
M
O
TE
1
1
1
0
.
8
4
0
.
8
3
0
.
8
3
A
D
A
S
Y
N
1
1
1
0
.
8
2
0
.
8
1
0
.
8
1
N
e
a
r
M
i
ss
1
0
.
9
7
0
.
9
8
0
.
0
5
0
.
9
1
0
.
0
9
To
me
k
Li
n
k
s
1
1
1
0
.
9
3
0
.
8
2
0
.
8
7
En
h
a
n
c
e
d
S
a
m
p
l
i
n
g
1
1
1
1
1
1
T
ab
le
7
p
r
esen
ts
a
co
m
p
ar
ati
v
e
p
er
f
o
r
m
an
ce
a
n
aly
s
is
b
etwe
en
th
e
p
r
o
p
o
s
ed
en
s
em
b
le
m
o
d
el
an
d
o
th
er
s
tate
-
of
-
th
e
-
ar
t
ap
p
r
o
ac
h
es
u
s
in
g
d
if
f
er
e
n
t
d
atasets
.
W
h
ile
th
e
C
NN,
AN
N,
an
d
R
NN
m
o
d
els
r
ep
o
r
ted
h
ig
h
p
r
ec
is
io
n
,
r
ec
all,
an
d
F
1
-
Sco
r
es
r
an
g
in
g
f
r
o
m
0
.
9
1
t
o
0
.
9
9
,
th
e
p
r
o
p
o
s
ed
e
n
s
em
b
le
m
o
d
el
o
f
C
NN,
L
STM
,
an
d
ML
P
o
u
tp
er
f
o
r
m
ed
all
with
p
er
f
ec
t
s
co
r
es
o
f
1
.
0
ac
r
o
s
s
all
m
etr
ics.
T
h
is
d
em
o
n
s
tr
ates
th
e
ef
f
ec
tiv
en
ess
o
f
th
e
en
s
em
b
l
e
ap
p
r
o
ac
h
in
ca
p
tu
r
i
n
g
co
m
p
lex
p
atter
n
s
an
d
im
p
r
o
v
i
n
g
f
r
au
d
d
etec
tio
n
ac
cu
r
ac
y
co
m
p
ar
e
d
to
in
d
iv
id
u
al
d
ee
p
lear
n
in
g
m
o
d
els.
T
ab
le
7
.
C
o
m
p
a
r
is
o
n
with
o
th
er
cu
ttin
g
-
e
d
g
e
tech
n
iq
u
es
R
e
f
.
M
o
d
e
l
D
a
t
a
s
e
t
P
r
e
c
i
s
i
o
n
R
e
c
a
l
l
F1
-
S
c
o
r
e
[
7
]
C
N
N
C
o
mm
e
r
c
i
a
l
b
a
n
k
B
2
C
o
n
l
i
n
e
t
r
a
n
s
a
c
t
i
o
n
d
a
t
a
0
.
9
1
0
.
9
4
0
.
9
3
[
1
3
]
ANN
C
r
e
d
i
t
c
a
r
d
f
r
a
u
d
d
e
t
e
c
t
i
o
n
f
r
o
m
K
a
g
g
l
e
0
.
9
8
0
.
9
7
0
.
9
7
[
1
5
]
R
N
N
B
a
l
a
n
c
e
d
s
h
a
r
e
d
5
5
9
8
5
6
-
r
e
c
o
r
d
K
a
g
g
l
e
r
e
p
o
s
i
t
o
r
y
0
.
9
8
1
0
.
9
9
P
r
o
p
o
se
d
En
se
mb
l
e
mo
d
e
l
En
se
mb
l
e
mo
d
e
l
o
f
C
N
N
,
LST
M
,
a
n
d
M
LP
C
r
e
d
i
t
c
a
r
d
f
a
u
l
t
d
e
t
e
c
t
i
o
n
f
r
o
m
K
a
g
g
l
e
1
1
1
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.
1
6
,
No
.
1
,
Feb
r
u
ar
y
20
2
6
:
3
1
1
-
320
318
T
h
e
s
tu
d
y
d
e
m
o
n
s
tr
ates
th
a
t
in
teg
r
atin
g
a
d
ee
p
en
s
em
b
le
lear
n
in
g
m
o
d
el
with
an
en
h
an
ce
d
o
v
er
s
am
p
lin
g
tech
n
iq
u
e
ef
f
e
ctiv
ely
ad
d
r
ess
es
th
e
cr
itical
is
s
u
e
o
f
clas
s
im
b
alan
ce
in
cr
ed
it
ca
r
d
f
r
au
d
d
etec
tio
n
.
I
n
d
i
v
id
u
al
m
o
d
els
(
C
NN,
L
STM
,
an
d
ML
P)
s
h
o
wed
s
u
b
s
tan
tial
im
p
r
o
v
em
e
n
ts
in
F1
-
S
co
r
e
f
o
r
f
r
au
d
d
etec
tio
n
wh
en
t
r
ain
ed
with
th
e
p
r
o
p
o
s
ed
s
am
p
lin
g
m
eth
o
d
—
r
is
in
g
f
r
o
m
0
.
8
0
to
0
.
9
6
f
o
r
C
NN,
0
.
8
0
t
o
0
.
9
4
f
o
r
L
STM
,
an
d
0
.
8
1
to
0
.
9
1
f
o
r
ML
P.
T
h
e
e
n
s
em
b
le
s
tr
ateg
ies
—
weig
h
ted
av
er
ag
e
,
u
n
weig
h
ted
a
v
er
ag
e
,
an
d
m
aj
o
r
ity
v
o
tin
g
—
ac
h
ie
v
ed
p
er
f
ec
t
s
co
r
es
(
1
.
0
0
p
r
ec
i
s
io
n
,
r
ec
all,
a
n
d
F1
-
S
co
r
e)
f
o
r
b
o
th
f
r
au
d
an
d
n
o
r
m
al
class
es,
o
u
tp
e
r
f
o
r
m
in
g
in
d
iv
id
u
al
m
o
d
els.
C
o
m
p
ar
e
d
to
tr
a
d
itio
n
al
s
am
p
lin
g
tech
n
i
q
u
es
lik
e
SMOT
E
,
ADASYN,
an
d
Nea
r
Miss
,
th
e
p
r
o
p
o
s
ed
en
h
an
ce
d
s
am
p
lin
g
alg
o
r
ith
m
d
em
o
n
s
tr
ated
s
u
p
er
i
o
r
ef
f
ec
tiv
en
ess
b
y
ac
h
iev
in
g
p
er
f
ec
t
class
if
icatio
n
m
etr
ics.
T
h
ese
f
in
d
in
g
s
c
o
n
f
ir
m
th
at
th
e
p
r
o
p
o
s
ed
a
p
p
r
o
a
ch
o
f
f
er
s
a
h
ig
h
l
y
ac
cu
r
ate,
b
alan
ce
d
,
an
d
r
eliab
l
e
f
r
au
d
d
etec
tio
n
s
y
s
tem
,
s
u
r
p
ass
in
g
ex
is
tin
g
s
tate
-
of
-
th
e
-
ar
t
tech
n
iq
u
es
.
5.
CO
NCLU
SI
O
N
W
ith
th
e
u
s
e
o
f
th
e
p
r
o
p
o
s
ed
s
am
p
lin
g
s
tr
ateg
y
,
a
co
m
p
a
r
ativ
e
s
tu
d
y
o
f
cr
ed
it
ca
r
d
f
r
au
d
d
etec
tio
n
u
s
in
g
C
NN,
L
STM
,
ML
P,
an
d
th
eir
en
s
em
b
le
m
o
d
el
o
n
th
e
cr
ed
it
ca
r
d
d
ataset
p
r
o
v
id
es
cr
u
cial
in
s
ig
h
ts
in
to
h
o
w
ef
f
ec
tiv
ely
th
ese
m
o
d
el
s
h
an
d
le
u
n
b
alan
ce
d
d
ata.
T
h
e
p
r
o
p
o
s
ed
s
am
p
lin
g
ap
p
r
o
ac
h
s
ig
n
if
ican
tly
m
itig
ates
th
e
is
s
u
e
o
f
class
im
b
alan
ce
,
lea
d
in
g
to
m
ea
s
u
r
a
b
le
g
ain
s
i
n
F1
-
S
co
r
e,
p
r
ec
is
io
n
,
a
n
d
r
ec
all
ac
r
o
s
s
all
m
o
d
els.
Am
o
n
g
th
e
in
d
iv
i
d
u
al
m
o
d
els,
L
STM
ex
ce
lled
in
ca
p
tu
r
in
g
tem
p
o
r
al
p
atter
n
s
with
in
tr
an
s
ac
tio
n
d
ata,
m
ak
in
g
it
p
ar
ticu
lar
l
y
ef
f
ec
tiv
e
f
o
r
f
r
au
d
d
etec
tio
n
,
w
h
ile
C
NN
d
em
o
n
s
tr
ated
s
tr
o
n
g
f
ea
tu
r
e
e
x
tr
ac
tio
n
ca
p
ab
ilit
ies,
an
d
ML
P
p
r
o
v
i
d
ed
a
s
tr
aig
h
tf
o
r
war
d
y
et
e
f
f
ec
tiv
e
ap
p
r
o
ac
h
.
T
h
e
e
n
s
em
b
le
m
o
d
el
f
u
r
th
er
am
p
lifie
d
th
ese
s
tr
en
g
t
h
s
,
ac
h
iev
in
g
s
u
p
er
io
r
p
er
f
o
r
m
an
ce
th
r
o
u
g
h
t
h
e
co
m
b
in
ed
ad
v
a
n
tag
es
o
f
all
th
r
ee
m
o
d
els.
T
h
e
r
esu
lts
h
ig
h
lig
h
t
s
u
b
s
tan
tial
im
p
r
o
v
em
en
ts
,
p
ar
ticu
lar
ly
in
r
ec
all
an
d
F1
-
S
co
r
e
f
o
r
f
r
au
d
d
etec
tio
n
,
wh
en
th
e
p
r
o
p
o
s
ed
s
am
p
lin
g
m
eth
o
d
is
ap
p
lied
.
W
ith
o
u
t
th
is
m
eth
o
d
,
th
e
m
o
d
els
s
tr
u
g
g
led
with
im
b
alan
ce
s
,
esp
ec
ially
i
n
r
ec
all
f
o
r
f
r
a
u
d
u
len
t
ca
s
es.
Ho
w
ev
er
,
t
h
e
p
r
o
p
o
s
ed
s
am
p
lin
g
tech
n
iq
u
e
en
ab
le
d
co
n
s
is
ten
t
an
d
s
ig
n
if
ican
t
p
er
f
o
r
m
an
ce
g
ain
s
ac
r
o
s
s
all
m
o
d
els,
with
th
e
en
s
em
b
le
m
o
d
el
ac
h
iev
in
g
t
h
e
b
est
o
v
er
all
r
esu
lts
d
u
e
to
its
ab
ilit
y
to
lev
er
ag
e
th
e
co
m
p
lem
e
n
tar
y
s
tr
en
g
th
s
o
f
C
NN,
L
STM
,
an
d
ML
P.
T
h
ese
f
in
d
in
g
s
u
n
d
e
r
s
co
r
e
th
e
v
al
u
e
o
f
th
e
e
n
s
em
b
le
a
p
p
r
o
ac
h
a
n
d
th
e
p
r
o
p
o
s
ed
s
am
p
lin
g
s
tr
ateg
y
in
im
p
r
o
v
in
g
class
if
icatio
n
ac
cu
r
ac
y
f
o
r
b
o
th
n
o
r
m
al
an
d
f
r
a
u
d
u
len
t
in
s
tan
ce
s
.
Fu
tu
r
e
r
esear
c
h
co
u
l
d
f
o
cu
s
o
n
r
e
f
in
in
g
th
e
s
am
p
lin
g
alg
o
r
ith
m
f
u
r
th
er
an
d
ex
ten
d
i
n
g
its
ap
p
licatio
n
t
o
o
th
er
ty
p
es
o
f
f
r
a
u
d
a
n
d
d
i
v
er
s
e
d
atasets
.
ACK
NO
WL
E
DG
M
E
N
T
S
W
e
wo
u
ld
lik
e
to
ex
p
r
ess
our
s
in
ce
r
e
g
r
atitu
d
e
to
all
in
s
tr
u
cto
r
s
,
co
lleag
u
es,
an
d
in
s
titu
tio
n
s
wh
o
s
e
g
u
id
an
ce
an
d
s
u
p
p
o
r
t
g
r
ea
tly
co
n
tr
ib
u
ted
to
th
e
s
u
cc
ess
f
u
l
co
m
p
letio
n
o
f
th
is
r
esear
ch
.
W
e
ar
e
also
th
an
k
f
u
l
to
our
f
am
ily
an
d
f
r
ie
n
d
s
f
o
r
th
eir
co
n
tin
u
o
u
s
en
c
o
u
r
a
g
em
en
t th
r
o
u
g
h
o
u
t t
h
e
s
tu
d
y
.
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
Ay
e
Ay
e
Kh
in
e
✓
✓
✓
✓
✓
✓
✓
✓
✓
Z
in
T
h
u
T
h
u
My
i
n
t
✓
✓
✓
✓
✓
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
o
en
s
u
r
e
f
ai
r
an
d
o
b
jectiv
e
d
ec
is
io
n
-
m
ak
in
g
,
au
t
h
o
r
s
m
u
s
t
d
ec
lar
e
an
y
ass
o
ciatio
n
s
t
h
at
p
o
s
e
a
co
n
f
lict
o
f
in
ter
est
(
f
i
n
an
cial
,
p
er
s
o
n
al,
o
r
p
r
o
f
ess
io
n
al)
in
co
n
n
ec
tio
n
with
m
an
u
s
cr
i
p
ts
.
No
n
-
f
in
a
n
cial
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
C
r
ed
it c
a
r
d
fr
a
u
d
d
a
ta
a
n
a
lysi
s
u
s
in
g
p
r
o
p
o
s
ed
s
a
mp
lin
g
a
lg
o
r
ith
m
…
(
A
ye
A
ye
K
h
in
e
)
319
co
m
p
etin
g
in
ter
ests
in
clu
d
e
a
d
ec
lar
atio
n
o
f
p
o
liti
ca
l,
p
er
s
o
n
al,
r
elig
io
u
s
,
id
eo
lo
g
ical,
ac
ad
em
ic,
an
d
in
tellectu
al
co
m
p
etin
g
in
ter
est
s
.
T
h
e
a
u
th
o
r
s
d
ec
lar
e
th
at
t
h
e
y
h
a
v
e
n
o
k
n
o
w
n
c
o
m
p
etin
g
f
i
n
an
cial
in
ter
ests
o
r
p
er
s
o
n
al
r
elatio
n
s
h
ip
s
th
at
co
u
ld
h
av
e
ap
p
ea
r
ed
to
in
f
l
u
en
c
e
th
e
wo
r
k
r
ep
o
r
ted
in
th
is
p
ap
er
.
I
f
th
er
e
ar
e
n
o
co
n
f
licts
o
f
in
ter
est,
p
lease
in
c
lu
d
e
th
e
f
o
llo
win
g
au
t
h
o
r
'
s
s
ta
tem
en
t: Au
th
o
r
s
s
tate
n
o
co
n
f
l
ict
o
f
in
ter
est.
I
NF
O
RM
E
D
CO
NS
E
N
T
No
t A
p
p
licab
le
.
E
T
H
I
CAL AP
P
RO
V
AL
T
h
is
s
tu
d
y
d
id
n
o
t
in
v
o
lv
e
an
y
ex
p
er
im
en
ts
o
n
h
u
m
an
s
o
r
an
im
als,
an
d
all
p
r
o
ce
d
u
r
es
co
m
p
lied
with
in
s
titu
tio
n
al
an
d
i
n
ter
n
atio
n
al
eth
ical
g
u
id
elin
es.
As
th
e
r
esear
ch
was
co
n
d
u
cted
u
s
in
g
p
u
b
licly
av
ailab
le
d
atasets
,
n
o
ad
d
itio
n
al
eth
ical
ap
p
r
o
v
al
was r
eq
u
ir
e
d
.
DATA AV
AI
L
AB
I
L
I
T
Y
Pu
b
licly
av
ailab
le
d
atasets
we
r
e
an
aly
ze
d
in
th
is
s
tu
d
y
.
T
h
is
d
ata
ca
n
b
e
f
o
u
n
d
h
er
e:
Kag
g
le,
“Cre
d
it
ca
r
d
f
r
au
d
d
etec
tio
n
.
”
2
0
2
5
,
[
On
lin
e]
.
Av
ailab
le:
h
ttp
s
:
//w
w
w
.
ka
g
g
le.
co
m/d
a
ta
s
ets/
mlg
-
u
lb
/cred
itca
r
d
fr
a
u
d
.
[
Acc
ess
ed
: O
ct.
5
,
2
0
2
4
].
RE
F
E
R
E
NC
E
S
[
1
]
M
.
Ja
b
e
e
n
,
S
.
R
a
mz
a
n
,
A
.
R
a
z
a
,
N
.
L.
F
i
t
r
i
y
a
n
i
,
M
.
S
y
a
f
r
u
d
i
n
,
a
n
d
S
.
W
.
Le
e
,
“
E
n
h
a
n
c
e
d
c
r
e
d
i
t
c
a
r
d
f
r
a
u
d
d
e
t
e
c
t
i
o
n
u
si
n
g
d
e
e
p
h
y
b
r
i
d
C
LST
mo
d
e
l
,
”
M
a
t
h
e
m
a
t
i
c
s
,
v
o
l
.
1
3
,
n
o
.
1
2
,
p
.
1
9
5
0
,
2
0
2
5
,
d
o
i
:
1
0
.
3
3
9
0
/
mat
h
1
3
1
2
1
9
5
0
.
[
2
]
F
.
Z.
El
H
l
o
u
l
i
,
J.
R
i
f
f
i
,
M
.
A
.
M
a
h
r
a
z
,
A
.
Y
a
h
y
a
o
u
y
,
K
.
E
l
F
a
z
a
z
y
,
a
n
d
H
.
Ta
i
r
i
,
“
C
r
e
d
i
t
c
a
r
d
f
r
a
u
d
d
e
t
e
c
t
i
o
n
:
A
d
d
r
e
ssi
n
g
i
mb
a
l
a
n
c
e
d
d
a
t
a
se
t
s
w
i
t
h
a
mu
l
t
i
-
p
h
a
se
a
p
p
r
o
a
c
h
,
”
S
N
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
v
o
l
.
5
,
n
o
.
1
,
p
.
1
7
3
,
2
0
2
4
,
d
o
i
:
1
0
.
1
0
0
7
/
s4
2
9
7
9
-
023
-
0
2
5
5
9
-
6.
[
3
]
E.
I
l
e
b
e
r
i
a
n
d
Y
.
S
u
n
,
“
A
h
y
b
r
i
d
d
e
e
p
l
e
a
r
n
i
n
g
e
n
se
mb
l
e
mo
d
e
l
f
o
r
c
r
e
d
i
t
c
a
r
d
f
r
a
u
d
d
e
t
e
c
t
i
o
n
,
”
I
EEE
Ac
c
e
ss
,
v
o
l
.
1
2
,
p
p
.
1
7
5
8
2
9
–
1
7
5
8
3
8
,
2
0
2
4
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
4
.
3
5
0
2
5
4
2
.
[
4
]
I
.
D
.
M
i
e
n
y
e
a
n
d
Y
.
S
u
n
,
“
A
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
me
t
h
o
d
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
f
o
r
i
mp
r
o
v
e
d
c
r
e
d
i
t
c
a
r
d
f
r
a
u
d
d
e
t
e
c
t
i
o
n
,
”
Ap
p
l
i
e
d
S
c
i
e
n
c
e
s
(
S
w
i
t
zer
l
a
n
d
)
,
v
o
l
.
1
3
,
n
o
.
1
2
,
p
.
7
2
5
4
,
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
a
p
p
1
3
1
2
7
2
5
4
.
[
5
]
M
.
S
e
e
r
a
,
C
.
P
.
Li
m
,
A
.
K
u
m
a
r
,
L
.
D
h
a
m
o
t
h
a
r
a
n
,
a
n
d
K
.
H
.
T
a
n
,
“
A
n
i
n
t
e
l
l
i
g
e
n
t
p
a
y
m
e
n
t
c
a
r
d
f
r
a
u
d
d
e
t
e
c
t
i
o
n
s
y
st
e
m,”
A
n
n
a
l
s o
f
O
p
e
r
a
t
i
o
n
s
Re
se
a
r
c
h
,
v
o
l
.
3
3
4
,
n
o
.
1
–
3
,
p
p
.
4
4
5
–
4
6
7
,
M
a
r
.
2
0
2
4
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
0
4
7
9
-
0
2
1
-
0
4
1
4
9
-
2.
[
6
]
M
.
A
b
d
u
l
S
a
l
a
m,
K
.
M
.
F
o
u
a
d
,
D
.
L.
El
b
a
b
l
y
,
a
n
d
S
.
M
.
El
s
a
y
e
d
,
“
F
e
d
e
r
a
t
e
d
l
e
a
r
n
i
n
g
m
o
d
e
l
f
o
r
c
r
e
d
i
t
c
a
r
d
f
r
a
u
d
d
e
t
e
c
t
i
o
n
w
i
t
h
d
a
t
a
b
a
l
a
n
c
i
n
g
t
e
c
h
n
i
q
u
e
s,”
N
e
u
r
a
l
C
o
m
p
u
t
i
n
g
a
n
d
A
p
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
3
6
,
n
o
.
1
1
,
p
p
.
6
2
3
1
–
6
2
5
6
,
2
0
2
4
,
d
o
i
:
1
0
.
1
0
0
7
/
s0
0
5
2
1
-
023
-
0
9
4
1
0
-
2.
[
7
]
C
.
Y
u
,
Y
.
X
u
,
J.
C
a
o
,
Y
.
Zh
a
n
g
,
Y
.
Ji
n
,
a
n
d
M
.
Zh
u
,
“
C
r
e
d
i
t
c
a
r
d
f
r
a
u
d
d
e
t
e
c
t
i
o
n
u
s
i
n
g
a
d
v
a
n
c
e
d
t
r
a
n
sf
o
r
mer
m
o
d
e
l
,
”
i
n
Pro
c
e
e
d
i
n
g
s
-
2
0
2
4
I
EEE
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
Me
t
a
v
e
rs
e
C
o
m
p
u
t
i
n
g
,
N
e
t
w
o
rk
i
n
g
,
a
n
d
A
p
p
l
i
c
a
t
i
o
n
s,
M
e
t
a
C
o
m
2
0
2
4
,
2
0
2
4
,
p
p
.
3
4
3
–
3
5
0
,
d
o
i
:
1
0
.
1
1
0
9
/
M
e
t
a
C
o
m6
2
9
2
0
.
2
0
2
4
.
0
0
0
6
4
.
[
8
]
N
.
Th
a
l
j
i
,
A
.
R
a
z
a
,
M
.
S
.
I
sl
a
m,
N
.
A
.
S
a
mee
,
a
n
d
M
.
M
.
Ja
mj
o
o
m,
“
A
E
-
N
e
t
:
N
o
v
e
l
a
u
t
o
e
n
c
o
d
e
r
-
b
a
se
d
d
e
e
p
f
e
a
t
u
r
e
s
f
o
r
S
Q
L
i
n
j
e
c
t
i
o
n
a
t
t
a
c
k
d
e
t
e
c
t
i
o
n
,
”
I
EE
E
A
c
c
e
ss
,
v
o
l
.
1
1
,
p
p
.
1
3
5
5
0
7
–
1
3
5
5
1
6
,
2
0
2
3
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
3
.
3
3
3
7
6
4
5
.
[
9
]
Y
.
M
u
r
u
g
a
n
,
M
.
V
i
j
a
y
a
l
a
k
sh
m
i
,
L
.
S
e
l
v
a
r
a
j
,
a
n
d
S
.
B
a
l
a
r
a
m
a
n
,
“
C
r
e
d
i
t
c
a
r
d
f
r
a
u
d
d
e
t
e
c
t
i
o
n
sy
s
t
e
m
u
si
n
g
C
N
N
,
”
L
e
c
t
u
re
N
o
t
e
s
i
n
N
e
t
w
o
rks
a
n
d
S
y
st
e
m
s
,
v
o
l
.
3
4
0
LN
N
S
,
n
o
.
3
,
p
p
.
1
9
4
–
2
0
4
,
2
0
2
2
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
3
-
030
-
9
4
5
0
7
-
7
_
1
9
.
[
1
0
]
J.
W
a
n
g
,
W
.
Li
u
,
Y
.
K
o
u
,
D
.
X
i
a
o
,
X
.
W
a
n
g
,
a
n
d
X
.
Ta
n
g
,
“
A
p
p
r
o
x
-
S
M
O
TE
f
e
d
e
r
a
t
e
d
l
e
a
r
n
i
n
g
c
r
e
d
i
t
c
a
r
d
f
r
a
u
d
d
e
t
e
c
t
i
o
n
sy
st
e
m,
”
i
n
Pro
c
e
e
d
i
n
g
s
-
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
m
p
u
t
e
r
S
o
f
t
w
a
re
a
n
d
A
p
p
l
i
c
a
t
i
o
n
s
C
o
n
f
e
r
e
n
c
e
,
2
0
2
3
,
v
o
l
.
2
0
2
3
-
Ju
n
e
,
p
p
.
1
3
7
0
–
1
3
7
5
,
d
o
i
:
1
0
.
1
1
0
9
/
C
O
M
P
S
A
C
5
7
7
0
0
.
2
0
2
3
.
0
0
2
0
8
.
[
1
1
]
I
.
V
e
j
a
l
l
a
,
S
.
P
.
B
a
t
t
u
l
a
,
K
.
K
a
l
l
u
r
i
,
a
n
d
H
.
K
.
K
a
l
l
u
r
i
,
“
C
r
e
d
i
t
c
a
r
d
f
r
a
u
d
d
e
t
e
c
t
i
o
n
u
si
n
g
mac
h
i
n
e
l
e
a
r
n
i
n
g
t
e
c
h
n
i
q
u
e
s,”
i
n
2
0
2
3
2
n
d
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
P
a
ra
d
i
g
m
S
h
i
f
t
s
i
n
C
o
m
m
u
n
i
c
a
t
i
o
n
s
Em
b
e
d
d
e
d
S
y
s
t
e
m
s
,
M
a
c
h
i
n
e
L
e
a
r
n
i
n
g
a
n
d
S
i
g
n
a
l
Pro
c
e
ssi
n
g
,
PC
E
M
S
2
0
2
3
,
2
0
2
3
,
p
p
.
1
–
6
,
d
o
i
:
1
0
.
1
1
0
9
/
P
C
E
M
S
5
8
4
9
1
.
2
0
2
3
.
1
0
1
3
6
0
4
0
.
[
1
2
]
P
.
Y
a
d
l
a
p
a
l
l
i
,
P
.
S
r
i
v
a
t
sa
l
,
N
.
P
o
l
i
m
e
r
a
,
a
n
d
M
.
S
r
i
n
i
v
a
s,
“
C
r
e
d
i
t
c
a
r
d
f
r
a
u
d
d
e
t
e
c
t
i
o
n
u
si
n
g
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
a
l
g
o
r
i
t
h
ms
a
n
d
a
r
t
i
f
i
c
i
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
,
”
i
n
2
0
2
5
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
Art
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
E
n
g
i
n
e
e
ri
n
g
,
AI
D
E
2
0
2
5
-
Pro
c
e
e
d
i
n
g
s
,
2
0
2
5
,
p
p
.
5
3
9
–
5
4
2
,
d
o
i
:
1
0
.
1
1
0
9
/
A
I
D
E6
4
2
2
8
.
2
0
2
5
.
1
0
9
8
7
5
3
7
.
[
1
3
]
P
.
K
.
Z
o
r
i
o
n
,
L
.
S
a
c
h
a
n
,
R
.
C
h
h
a
b
r
a
,
V
.
P
a
n
d
e
y
,
a
n
d
D
.
H
.
F
a
t
i
m
a
,
“
C
r
e
d
i
t
c
a
r
d
f
i
n
a
n
c
i
a
l
f
r
a
u
d
d
e
t
e
c
t
i
o
n
u
si
n
g
d
e
e
p
l
e
a
r
n
i
n
g
,
”
S
S
RN
El
e
c
t
ro
n
i
c
J
o
u
r
n
a
l
,
2
0
2
3
,
d
o
i
:
1
0
.
2
1
3
9
/
ssr
n
.
4
6
2
9
0
9
3
.
[
1
4
]
K
.
R
.
V
.
,
S
.
G
a
n
e
s
h
,
M
.
D
,
B
.
G
.
R
.
,
a
n
d
M
.
T
h
i
r
u
v
e
n
g
a
d
a
m,
“
C
r
e
d
i
t
c
a
r
d
f
r
a
u
d
d
a
t
a
a
n
a
l
y
si
s
a
n
d
p
r
e
d
i
c
t
i
o
n
u
s
i
n
g
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
a
l
g
o
r
i
t
h
ms,”
S
e
c
u
ri
t
y
a
n
d
Pr
i
v
a
c
y
,
v
o
l
.
8
,
n
o
.
3
,
p
.
e
7
0
0
4
3
,
2
0
2
5
,
d
o
i
:
1
0
.
1
0
0
2
/
sp
y
2
.
7
0
0
4
3
.
[
1
5
]
E.
-
S
.
M
.
El
-
K
e
n
a
w
y
,
“
C
r
e
d
i
t
c
a
r
d
f
r
a
u
d
d
e
t
e
c
t
i
o
n
b
a
se
d
o
n
d
e
e
p
l
e
a
r
n
i
n
g
mo
d
e
l
s
,
”
Me
s
o
p
o
t
a
m
i
a
n
J
o
u
r
n
a
l
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
v
o
l
.
2
0
2
4
,
p
p
.
2
0
4
–
2
1
3
,
2
0
2
4
,
d
o
i
:
1
0
.
5
8
4
9
6
/
M
JC
S
C
/
2
0
2
4
/
0
1
6
.
[
1
6
]
R
.
K
.
G
u
p
t
a
e
t
a
l
.
,
“
E
n
h
a
n
c
e
d
f
r
a
m
e
w
o
r
k
f
o
r
c
r
e
d
i
t
c
a
r
d
f
r
a
u
d
d
e
t
e
c
t
i
o
n
u
si
n
g
r
o
b
u
st
f
e
a
t
u
r
e
s
e
l
e
c
t
i
o
n
a
n
d
a
s
t
a
c
k
i
n
g
e
n
se
mb
l
e
mo
d
e
l
a
p
p
r
o
a
c
h
,
”
Re
s
u
l
t
s
i
n
E
n
g
i
n
e
e
r
i
n
g
,
v
o
l
.
2
6
,
p
.
1
0
5
0
8
4
,
2
0
2
5
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
r
i
n
e
n
g
.
2
0
2
5
.
1
0
5
0
8
4
.
[
1
7
]
P
.
S
u
n
d
a
r
a
v
a
d
i
v
e
l
,
R
.
A
.
I
saa
c
,
D
.
El
a
n
g
o
v
a
n
,
D
.
K
r
i
sh
n
a
R
a
j
,
V
.
V
.
L.
R
a
h
u
l
,
a
n
d
R
.
R
a
j
a
,
“
O
p
t
i
mi
z
i
n
g
c
r
e
d
i
t
c
a
r
d
f
r
a
u
d
d
e
t
e
c
t
i
o
n
w
i
t
h
r
a
n
d
o
m
f
o
r
e
st
s
a
n
d
S
M
O
TE,
”
S
c
i
e
n
t
i
f
i
c
Re
p
o
rt
s
,
v
o
l
.
1
5
,
n
o
.
1
,
p
.
1
7
8
5
1
,
2
0
2
5
,
d
o
i
:
1
0
.
1
0
3
8
/
s
4
1
5
9
8
-
025
-
0
0
8
7
3
-
y.
[
1
8
]
A
.
P
.
L
o
p
e
s,
S
.
P
a
r
sh
i
o
n
i
k
a
r
,
A
.
K
a
l
e
,
N
.
S
h
a
r
ma
,
a
n
d
A
.
A
.
V
a
r
g
h
e
s
e
,
“
A
c
o
m
p
a
r
a
t
i
v
e
st
u
d
y
o
f
d
e
e
p
l
e
a
r
n
i
n
g
t
e
c
h
n
i
q
u
e
s
f
o
r
c
r
e
d
i
t
c
a
r
d
f
r
a
u
d
d
e
t
e
c
t
i
o
n
,
”
i
n
2
0
2
1
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
A
d
v
a
n
c
e
s
i
n
C
o
m
p
u
t
i
n
g
,
C
o
m
m
u
n
i
c
a
t
i
o
n
,
a
n
d
C
o
n
t
r
o
l
(
I
C
AC
3
)
,
2
0
2
1
,
p
p
.
1
–
5
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
A
C
3
5
3
6
4
2
.
2
0
2
1
.
9
6
9
7
2
0
5
.
[
1
9
]
T.
P
o
u
r
h
a
b
i
b
i
,
K
.
L.
O
n
g
,
B
.
H
.
K
a
m,
a
n
d
Y
.
L
.
B
o
o
,
“
F
r
a
u
d
d
e
t
e
c
t
i
o
n
:
A
sy
st
e
ma
t
i
c
l
i
t
e
r
a
t
u
r
e
r
e
v
i
e
w
o
f
g
r
a
p
h
-
b
a
s
e
d
a
n
o
m
a
l
y
d
e
t
e
c
t
i
o
n
a
p
p
r
o
a
c
h
e
s,
”
D
e
c
i
s
i
o
n
S
u
p
p
o
rt
S
y
st
e
m
s
,
v
o
l
.
1
3
3
,
2
0
2
0
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
d
ss.
2
0
2
0
.
1
1
3
3
0
3
.
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.
1
6
,
No
.
1
,
Feb
r
u
ar
y
20
2
6
:
3
1
1
-
320
320
[
2
0
]
A
.
R
B
a
n
d
S
.
K
.
K
R
,
“
C
r
e
d
i
t
c
a
r
d
f
r
a
u
d
d
e
t
e
c
t
i
o
n
u
s
i
n
g
a
r
t
i
f
i
c
i
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
,
”
G
l
o
b
a
l
T
ra
n
si
t
i
o
n
s
Pro
c
e
e
d
i
n
g
s
,
v
o
l
.
2
,
n
o
.
1
,
p
p
.
3
5
–
4
1
,
2
0
2
1
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
g
l
t
p
.
2
0
2
1
.
0
1
.
0
0
6
.
[
2
1
]
K
.
Y
a
mi
n
i
,
V
.
A
n
i
t
h
a
,
S
.
P
o
l
e
p
a
k
a
,
R
.
C
h
a
u
h
a
n
,
Y
.
V
a
r
s
h
n
e
y
,
a
n
d
M
.
S
i
n
g
h
,
“
A
n
i
n
t
e
l
l
i
g
e
n
t
me
t
h
o
d
f
o
r
c
r
e
d
i
t
c
a
r
d
f
r
a
u
d
d
e
t
e
c
t
i
o
n
u
s
i
n
g
i
mp
r
o
v
e
d
C
N
N
a
n
d
e
x
t
r
e
me
l
e
a
r
n
i
n
g
ma
c
h
i
n
e
,
”
i
n
2
0
2
3
8
t
h
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
C
o
m
m
u
n
i
c
a
t
i
o
n
a
n
d
El
e
c
t
r
o
n
i
c
s
S
y
st
e
m
s (I
C
C
E
S
)
,
2
0
2
3
,
p
p
.
8
1
0
–
8
1
5
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
C
ES
5
7
2
2
4
.
2
0
2
3
.
1
0
1
9
2
7
7
4
.
[
2
2
]
S
.
N
e
g
i
,
S
.
K
.
D
a
s,
a
n
d
R
.
B
o
d
h
,
“
C
r
e
d
i
t
c
a
r
d
f
r
a
u
d
d
e
t
e
c
t
i
o
n
u
s
i
n
g
d
e
e
p
a
n
d
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
,
”
i
n
2
0
2
2
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
A
p
p
l
i
e
d
Ar
t
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
a
n
d
C
o
m
p
u
t
i
n
g
(
I
C
A
AI
C
)
,
2
0
2
2
,
p
p
.
4
5
5
–
4
6
1
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
A
A
I
C
5
3
9
2
9
.
2
0
2
2
.
9
7
9
2
9
4
1
.
[
2
3
]
J.
B
r
o
w
n
l
e
e
,
“
A
g
e
n
t
l
e
i
n
t
r
o
d
u
c
t
i
o
n
t
o
t
r
a
n
sf
e
r
l
e
a
r
n
i
n
g
f
o
r
d
e
e
p
l
e
a
r
n
i
n
g
,
”
Ma
c
h
i
n
e
L
e
a
r
n
i
n
g
M
a
s
t
e
ry
,
2
0
1
7
.
[
2
4
]
M
a
c
h
i
n
e
L
e
a
r
n
i
n
g
G
r
o
u
p
-
U
L
B
,
“
C
r
e
d
i
t
c
a
r
d
f
r
a
u
d
d
e
t
e
c
t
i
o
n
,
”
K
a
g
g
l
e
,
2
0
2
5
.
h
t
t
p
s
:
/
/
w
w
w
.
k
a
g
g
l
e
.
c
o
m
/
d
a
t
a
se
t
s/
m
l
g
-
u
l
b
/
c
r
e
d
i
t
c
a
r
d
f
r
a
u
d
(
a
c
c
e
sse
d
O
c
t
.
0
5
,
2
0
2
4
)
.
[
2
5
]
J.
H
a
n
,
M
.
K
a
m
b
e
r
,
a
n
d
J.
P
e
i
,
D
a
t
a
m
i
n
i
n
g
:
C
o
n
c
e
p
t
s
a
n
d
t
e
c
h
n
i
q
u
e
s
,
3
r
d
Ed
i
t
i
o
.
E
l
se
v
i
e
r
,
2
0
1
1
.
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
Ay
e
Ay
e
K
h
in
e
is
a
n
a
c
c
o
m
p
li
sh
e
d
As
so
c
iate
P
r
o
fe
ss
o
r
a
t
IM
CEIT
S
a
t
In
fo
rm
a
ti
o
n
C
o
m
m
u
n
ica
ti
o
n
Te
c
h
n
o
l
o
g
y
Re
se
a
rc
h
Ce
n
ter
.
S
h
e
h
o
ld
s
a
M
.
I.
S
c
fro
m
t
h
e
Un
iv
e
rsity
o
f
C
o
m
p
u
ter
S
t
u
d
ies
Ya
n
g
o
n
,
wh
e
re
sh
e
s
p
e
c
ialize
d
i
n
a
d
v
a
n
c
e
d
re
se
a
rc
h
with
i
n
h
e
r
field
.
He
r
a
re
a
s
o
f
e
x
p
e
rti
se
i
n
c
lu
d
e
Ja
v
a
p
r
o
g
ra
m
m
in
g
,
Ora
c
le
DB,
d
a
ta
sc
ien
c
e
,
a
n
d
d
e
e
p
lea
rn
in
g
.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
h
e
r
at
e
m
a
il
:
a
y
e
k
h
i
n
e
7
1
@
g
m
a
il
.
c
o
m
.
Zin
Th
u
Th
u
My
in
t
is
a
n
a
c
c
o
m
p
li
sh
e
d
As
so
c
iate
P
ro
fe
ss
o
r
a
t
th
e
F
a
c
u
lt
y
o
f
In
fo
rm
a
ti
o
n
S
c
ien
c
e
a
t
th
e
Un
i
v
e
rsity
o
f
C
o
m
p
u
ter
S
tu
d
ies
,
Ya
n
g
o
n
.
S
h
e
h
o
ld
s
a
P
h
.
D.
fr
o
m
th
e
Un
iv
e
rsit
y
o
f
Tec
h
n
o
lo
g
y
(
Ya
d
a
n
a
p
o
n
Cy
b
e
r
Cit
y
),
wh
e
re
sh
e
sp
e
c
ialize
d
in
a
d
v
a
n
c
e
d
re
se
a
rc
h
with
in
h
e
r
fiel
d
.
He
r
a
re
a
s
o
f
e
x
p
e
rti
se
i
n
c
lu
d
e
se
m
a
n
ti
c
wo
r
d
p
ro
c
e
ss
in
g
,
d
a
ta
sc
ien
c
e
,
a
n
d
c
y
b
e
rse
c
u
rit
y
.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
at
e
m
a
il
:
z
in
t
h
u
t
h
u
m
y
in
t@
u
c
sy
.
e
d
u
.
m
m
.
Evaluation Warning : The document was created with Spire.PDF for Python.