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.
15
,
No
.
4
,
A
u
g
u
s
t 2
0
2
6
,
p
p
.
3
2
6
9
~
3
2
8
5
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
2
6
9
-
3
2
8
5
3269
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
i
.
ia
esco
r
e.
co
m
Sy
stema
tic
revi
e
w of fra
ud de
tecti
o
n using
AI
a
nd
ML
with
a
n
empha
sis
on
tele
c
o
mm
unica
tion in
dustry
So
ly
M
a
t
hew
B
ij
u,
Sin
di Rr
y
t
a
S
c
h
o
o
l
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
U
n
i
v
e
r
s
i
t
y
o
f
W
o
l
l
o
n
g
o
n
g
i
n
D
u
b
a
i
,
D
u
b
a
i
,
U
n
i
t
e
d
A
r
a
b
Em
i
r
a
t
e
s
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Ma
y
2
2
,
2
0
2
5
R
ev
is
ed
J
u
n
3
,
2
0
2
6
Acc
ep
ted
J
u
l 9
,
2
0
2
6
Th
e
tele
c
o
m
m
u
n
ica
ti
o
n
s
i
n
d
u
str
y
is
o
n
e
o
f
t
h
e
to
p
i
n
d
u
stries
a
ffe
c
ted
b
y
fra
u
d
u
le
n
t
a
c
ti
v
i
ti
e
s.
G
iv
e
n
th
e
fin
a
n
c
ial
imp
a
c
t,
o
n
to
p
o
f
c
o
n
f
id
e
n
ti
a
li
t
y
b
re
a
c
h
e
s,
se
c
u
rit
y
c
o
n
c
e
rn
s,
a
n
d
r
e
d
u
c
e
d
se
rv
ice
q
u
a
li
ty
a
s
we
ll
a
s
c
o
n
su
m
e
r
d
issa
ti
sfa
c
ti
o
n
,
th
e
re
is
a
n
imm
e
d
iate
n
e
e
d
t
o
im
p
lem
e
n
t
e
ffe
c
ti
v
e
fra
u
d
d
e
tec
ti
o
n
a
p
p
ro
a
c
h
e
s.
W
h
il
e
th
e
re
h
a
v
e
b
e
e
n
d
iffere
n
t
fra
u
d
d
e
tec
ti
o
n
sy
ste
m
s
imp
lem
e
n
ted
,
tec
h
n
o
l
o
g
i
c
a
l
a
d
v
a
n
c
e
m
e
n
ts
a
s
we
ll
a
s
th
e
imp
ro
v
e
d
tec
h
n
iq
u
e
s
o
f
fra
u
d
ste
rs
h
a
v
e
m
a
d
e
th
e
trad
it
io
n
a
l
a
p
p
r
o
a
c
h
e
s
n
o
l
o
n
g
e
r
e
fficie
n
t.
Th
is
p
a
p
e
r
a
ims
to
fu
rth
e
r
i
n
v
e
sti
g
a
te
th
e
u
se
o
f
a
rti
ficia
l
in
telli
g
e
n
c
e
(
AI
)
a
n
d
m
a
c
h
in
e
lea
rn
in
g
(
ML
)
t
o
c
re
a
te
e
ffi
c
ien
t
a
n
d
a
d
v
a
n
c
e
d
fra
u
d
d
e
tec
ti
o
n
m
o
d
e
ls b
a
se
d
o
n
t
h
e
stra
teg
y
u
se
d
,
m
o
d
e
l
s
a
p
p
li
e
d
,
a
c
c
u
ra
c
y
o
f
th
e
sy
ste
m
,
a
s
we
ll
a
s
fu
tu
re
re
se
a
rc
h
wo
rk
su
g
g
e
ste
d
.
A
sy
ste
m
a
ti
c
re
v
iew
o
f
5
0
p
a
p
e
rs
wa
s
c
o
n
d
u
c
ted
.
Th
e
m
o
st
p
re
v
a
le
n
t
stra
teg
y
wa
s
su
p
e
rv
ise
d
o
n
e
,
m
a
jo
rit
y
o
f
p
a
p
e
rs
u
se
d
so
ftwa
re
in
ste
a
d
o
f
h
a
rd
wa
re
,
a
n
d
th
e
m
o
st
c
o
m
m
o
n
M
L
m
o
d
e
ls
we
re
a
rti
ficia
l
n
e
u
ra
l
n
e
two
r
k
(
ANN
)
,
su
p
p
o
rt
v
e
c
to
r
m
a
c
h
in
e
(
S
VM
)
,
a
n
d
d
e
c
isio
n
tree
.
K
ey
w
o
r
d
s
:
Ar
tific
ial
in
tellig
en
ce
B
eh
av
io
r
al
p
atter
n
Fra
u
d
d
etec
tio
n
Ma
ch
in
e
lear
n
in
g
Per
f
o
r
m
an
ce
m
etr
ics
T
elec
o
m
m
u
n
icatio
n
in
d
u
s
tr
y
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
:
So
ly
Ma
th
ew
B
iju
Sch
o
o
l o
f
C
o
m
p
u
ter
Scien
ce
,
Un
iv
er
s
ity
o
f
W
o
llo
n
g
o
n
g
in
Du
b
ai
Of
f
ice
2
0
1
,
B
lo
ck
1
5
,
Kn
o
wle
d
g
e
Par
k
,
PO Bo
x
2
0
1
8
3
,
Du
b
ai,
Un
ited
Ar
ab
E
m
ir
ates
E
m
ail:
s
o
ly
m
ath
ewb
iju
@
u
o
wd
u
b
ai.
ac
.
ae
1.
I
NT
RO
D
UCT
I
O
N
I
n
a
tech
n
o
l
o
g
ically
a
d
v
an
ce
d
wo
r
ld
,
th
e
p
o
te
n
tial
th
r
ea
t
f
r
o
m
ec
o
n
o
m
ic
cr
im
es
an
d
f
r
au
d
h
as
in
cr
ea
s
ed
d
r
asti
ca
lly
,
cr
ea
tin
g
n
ew
s
ec
u
r
ity
co
n
ce
r
n
s
f
o
r
b
u
s
in
ess
es
an
d
co
n
s
u
m
er
s
am
o
n
g
all
ty
p
es
o
f
in
d
u
s
tr
ies.
Fra
u
d
is
d
ef
in
ed
as
th
e
ac
t
o
f
g
ettin
g
m
o
n
ey
b
y
l
y
in
g
to
p
eo
p
le
[
1
]
.
I
n
o
th
er
wo
r
d
s
,
it
in
cl
u
d
es
an
y
o
p
p
o
r
tu
n
is
tic
b
eh
av
io
r
th
r
o
u
g
h
wh
ich
o
n
e
p
ar
ty
ex
p
lo
its
th
e
o
th
er
o
n
e
f
o
r
m
o
n
etar
y
b
en
ef
its
.
Statis
tic
s
s
u
g
g
est th
at
in
2
0
2
5
f
r
au
d
ac
c
o
u
n
ts
f
o
r
m
o
r
e
t
h
an
$
1
tr
illi
o
n
lo
s
s
es wo
r
ld
wid
e
in
all
ty
p
es o
f
in
d
u
s
tr
ies,
o
u
t o
f
wh
ich
o
n
l
y
4
%
o
f
af
f
ec
ted
b
u
s
in
ess
es
d
o
r
etr
iev
e
th
eir
as
s
ets
b
ac
k
[
2
]
.
Acc
o
r
d
in
g
to
th
e
Ass
o
ciatio
n
o
f
C
er
tifie
d
Fra
u
d
E
x
am
in
er
s
,
te
leco
m
m
u
n
icatio
n
is
o
n
e
o
f
th
e
to
p
af
f
ec
ted
in
d
u
s
tr
ies
f
r
o
m
f
r
au
d
[
3
]
.
Su
c
h
a
th
in
g
ca
n
b
e
ex
p
lain
e
d
b
y
m
a
n
y
f
ac
to
r
s
.
Firstl
y
,
th
e
co
m
p
lex
n
atu
r
e
o
f
th
e
telec
o
m
m
u
n
ica
tio
n
n
etwo
r
k
wh
ich
in
clu
d
es
m
an
y
p
r
o
ce
s
s
es
in
cr
e
ases
th
e
o
p
p
o
r
tu
n
ity
f
o
r
p
r
o
f
it
.
Mo
r
eo
v
er
,
th
e
h
ig
h
n
u
m
b
er
o
f
p
a
r
ties
in
v
o
lv
e
d
wh
ich
aim
to
m
ax
im
ize
t
h
eir
i
n
d
iv
id
u
al
b
en
ef
it
b
esid
es
th
e
c
o
m
p
an
y
’
s
o
n
e
,
ad
d
s
to
s
u
ch
c
o
m
p
lex
ity
.
An
o
th
er
im
p
o
r
tan
t
r
ea
s
o
n
is
th
e
ex
is
ten
ce
o
f
d
if
f
er
e
n
t
leg
al
an
d
r
eg
u
lato
r
y
f
r
am
ewo
r
k
s
u
p
o
n
wh
i
ch
ag
r
ee
m
e
n
ts
an
d
s
ettlem
en
ts
ar
e
b
u
ilt.
L
astl
y
,
o
p
p
o
r
t
u
n
ity
f
o
r
h
ig
h
p
r
o
f
it
m
ar
g
in
s
d
u
e
to
d
i
f
f
er
en
ce
s
i
n
ter
m
in
atio
n
r
ates
am
o
n
g
d
if
f
er
e
n
t c
o
u
n
tr
ies m
ak
es tele
co
m
a
h
ig
h
l
y
f
r
a
u
d
p
r
o
n
e
in
d
u
s
tr
y
.
G
i
v
e
n
t
h
e
s
e
v
e
r
i
t
y
,
f
i
n
a
n
c
i
al
a
s
w
e
l
l
a
s
s
ec
u
r
i
t
y
c
o
n
c
e
r
n
s
o
f
f
r
a
u
d
a
m
o
n
g
t
e
l
e
c
o
m
m
u
n
i
c
a
t
i
o
n
i
n
d
u
s
t
r
y
,
d
if
f
er
en
t
ap
p
r
o
ac
h
es
r
e
g
ar
d
i
n
g
f
r
a
u
d
d
etec
tio
n
h
av
e
b
ee
n
p
r
o
p
o
s
ed
.
Acc
o
r
d
in
g
t
o
Yeh
y
a
an
d
Salh
ab
[
4
]
,
th
e
m
ain
f
r
au
d
d
etec
tio
n
tech
n
i
q
u
es
ar
e
ca
ll
d
ata
r
ec
o
r
d
(
C
DR
)
an
aly
s
i
s
,
p
atter
n
r
ec
o
g
n
itio
n
,
r
ea
l
-
tim
e
m
o
n
ito
r
in
g
,
r
u
le
-
b
ased
s
y
s
tem
,
an
d
m
ac
h
in
e
lear
n
in
g
(
ML
)
a
lg
o
r
ith
m
s
.
C
DR
an
aly
s
is
co
n
s
is
ts
o
f
d
ata
r
ec
o
r
d
s
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.
15
,
No
.
4
,
Au
g
u
s
t 2
0
2
6
:
3
2
6
9
-
3
2
8
5
3270
f
o
r
ea
c
h
ca
ll
s
u
ch
as
d
u
r
atio
n
,
co
u
n
t,
o
r
ig
i
n
,
d
esti
n
atio
n
,
r
e
p
etitio
n
an
d
o
th
er
r
elate
d
d
etails.
T
h
is
an
aly
s
is
is
cr
u
cial
in
u
n
d
er
s
tan
d
in
g
p
atter
n
s
an
d
id
en
tif
y
in
g
an
o
m
alies
i
n
u
s
er
b
eh
a
v
io
r
.
Mo
r
e
o
v
er
,
p
a
tter
n
r
ec
o
g
n
itio
n
is
ap
p
lied
to
d
if
f
er
en
tiate
b
etw
ee
n
n
o
r
m
al
an
d
s
u
s
p
icio
u
s
b
eh
av
io
r
s
,
wh
ic
h
d
ep
e
n
d
s
o
n
th
e
r
elatio
n
s
h
ip
b
etwe
en
d
ata
p
o
in
ts
an
d
d
ata
u
s
ag
e.
Ma
n
y
m
o
b
ile
o
p
er
at
o
r
s
u
tili
ze
s
tat
is
tical
m
o
d
els
to
co
n
d
u
ct
r
ec
o
g
n
itio
n
p
atter
n
a
n
aly
s
is
.
R
eg
ar
d
in
g
t
h
e
r
ea
l
-
tim
e
m
o
n
ito
r
in
g
,
it
is
a
cr
u
cial
co
m
p
o
n
en
t
c
o
n
s
id
er
in
g
t
h
e
n
at
u
r
e
o
f
f
r
au
d
,
wh
ich
n
ee
d
s
to
b
e
d
et
ec
ted
as
s
o
o
n
as
th
e
f
r
au
d
h
as
b
ee
n
co
n
d
u
cted
.
Usu
ally
,
f
r
au
d
m
an
a
g
em
en
t
s
y
s
tem
s
u
s
e
aler
ts
an
d
r
ea
l
tim
e
m
o
n
ito
r
in
g
to
m
an
ag
e
f
r
a
u
d
ca
s
es.
A
r
u
le
-
b
ased
s
y
s
tem
is
a
f
r
a
u
d
d
etec
tio
n
m
eth
o
d
th
at
f
u
n
ctio
n
s
u
p
o
n
p
r
e
-
d
ef
i
n
ed
r
u
les
th
at
m
atch
f
r
au
d
p
atter
n
s
.
Su
ch
r
u
les
ar
e
d
ef
in
ed
b
ased
o
n
h
is
to
r
ical
d
ata
an
d
ar
e
ap
p
lied
to
th
e
liv
e
tr
af
f
ic,
to
g
e
n
er
ate
aler
ts
in
ca
s
e
s
u
ch
r
u
les
ar
e
b
r
ea
ch
ed
.
Ho
we
v
er
,
th
is
tech
n
iq
u
e
n
ee
d
s
co
n
tin
u
o
u
s
r
u
le
u
p
d
ate
an
d
en
r
ich
m
en
t
to
e
n
s
u
r
e
ac
c
u
r
ac
y
.
L
astl
y
,
ML
alg
o
r
ith
m
s
ar
e
a
v
er
y
e
f
f
icien
t
m
eth
o
d
co
n
s
id
er
in
g
th
eir
ab
ilit
y
to
h
a
n
d
le
a
lar
g
e
am
o
u
n
t
o
f
C
DR
tr
af
f
ic
an
d
p
r
o
v
id
e
g
r
ea
t
p
r
ed
ictio
n
r
ates.
Su
ch
m
o
d
u
le
s
ca
n
co
n
tin
u
o
u
s
ly
b
e
u
p
d
ated
b
y
lear
n
in
g
th
r
o
u
g
h
n
ew
d
ata
,
h
en
ce
,
im
p
r
o
v
in
g
th
eir
p
er
f
o
r
m
an
ce
.
T
h
e
aim
o
f
th
is
p
ap
er
is
to
p
r
o
v
id
e
a
d
etailed
r
ev
iew
o
f
th
e
ex
is
tin
g
liter
atu
r
e
o
n
f
r
a
u
d
d
etec
tio
n
m
eth
o
d
s
th
at
u
s
e
ar
tific
ial
in
te
llig
en
ce
(
AI
)
an
d
ML
m
o
d
els
in
th
e
telec
o
m
m
u
n
icatio
n
s
in
d
u
s
tr
y
as
in
T
ab
le
1
(
s
ee
in
Ap
p
en
d
i
x
)
.
Fo
r
ea
c
h
r
ev
iewe
d
s
tu
d
y
,
t
h
is
r
ev
iew
p
r
esen
ts
th
e
aim
o
f
th
e
s
tu
d
y
an
d
th
e
ap
p
lied
s
tr
ateg
y
,
in
clu
d
in
g
s
u
p
er
v
is
e
d
,
u
n
s
u
p
e
r
v
is
ed
,
s
em
i
-
s
u
p
er
v
is
ed
,
k
n
o
wled
g
e
-
b
ased
,
an
d
v
is
u
aliza
tio
n
-
b
ased
ap
p
r
o
ac
h
es.
I
t
also
d
escr
ib
e
s
th
e
u
tili
ze
d
h
ar
d
war
e/so
f
twar
e,
ap
p
lied
alg
o
r
ith
m
s
,
s
am
p
le
s
i
ze
,
ac
h
iev
ed
m
o
d
el
ac
cu
r
ac
y
,
u
s
ed
d
atab
ase,
an
d
s
u
g
g
ested
f
u
t
u
r
e
s
tu
d
y
.
2.
M
E
T
H
O
D
T
h
is
s
y
s
tem
atic
liter
atu
r
e
r
ev
i
ew
f
o
cu
s
es
o
n
t
h
e
ex
is
tin
g
r
es
ea
r
ch
wo
r
k
o
n
th
e
d
etec
tio
n
o
f
f
r
au
d
in
th
e
telec
o
m
m
u
n
icatio
n
s
in
d
u
s
tr
y
b
y
u
s
in
g
AI
an
d
ML
m
o
d
els.
D
if
f
er
en
t
m
o
d
els
an
d
ap
p
r
o
ac
h
es
h
a
v
e
b
ee
n
p
r
o
p
o
s
ed
b
y
d
if
f
er
en
t
s
ch
o
l
ar
s
,
g
iv
en
th
e
c
h
an
g
in
g
n
at
u
r
e
o
f
f
r
a
u
d
wh
ile
ad
o
p
tin
g
th
e
tech
n
o
lo
g
ical
ad
v
an
ce
m
e
n
ts
o
f
ea
ch
tim
e
.
T
h
is
s
tu
d
y
aim
s
t
o
p
r
o
v
i
d
e
a
n
o
v
er
v
iew
o
f
t
h
e
ex
is
tin
g
r
e
s
ea
r
ch
wo
r
k
wh
ile
p
o
in
tin
g
o
u
t
an
y
ex
is
tin
g
g
a
p
s
r
eg
ar
d
i
n
g
th
e
u
s
e
o
f
AI
a
n
d
ML
in
f
r
au
d
d
etec
tio
n
.
T
h
e
f
o
l
lo
win
g
ar
e
th
e
m
ain
elem
en
ts
u
p
o
n
wh
ich
t
h
is
liter
atu
r
e
r
ev
iew
is
b
u
ilt:
s
tr
ateg
y
(
s
u
p
er
v
is
ed
,
u
n
s
u
p
er
v
is
ed
,
a
n
d
s
em
i
-
s
u
p
er
v
is
ed
)
,
alg
o
r
ith
m
s
u
s
ed
,
h
ar
d
war
e
o
r
s
o
f
twar
e
o
r
b
o
th
,
ac
c
u
r
ac
y
o
f
m
o
d
el,
d
ata
s
am
p
le
s
ize
,
a
n
d
f
u
tu
r
e
s
u
g
g
esti
o
n
s
.
T
h
is
p
ap
er
d
etec
ted
r
elev
a
n
t
s
tu
d
ies
b
y
u
s
in
g
th
e
k
ey
wo
r
d
s
ea
r
ch
m
eth
o
d
.
Sev
er
al
k
ey
wo
r
d
s
wer
e
d
is
co
v
er
ed
o
n
Go
o
g
le
Sch
o
lar
,
R
esear
ch
Gate
,
I
E
E
E
,
Aca
d
e
m
ia,
E
ls
ev
ier
,
a
n
d
Sp
r
in
g
e
r
.
S
u
ch
k
ey
wo
r
d
s
wer
e
‘
Fra
u
d
d
etec
tio
n
’
,
‘
T
elec
o
m
m
u
n
icatio
n
n
etwo
r
k
’
,
an
d
“AI
an
d
ML
”.
T
h
e
p
r
i
m
ar
y
o
b
jectiv
e
o
f
th
is
s
tu
d
y
is
to
p
r
o
v
id
e
an
o
v
er
v
iew
o
f
t
h
e
ex
is
tin
g
liter
atu
r
e
r
eg
a
r
d
in
g
f
r
au
d
d
etec
tio
n
in
telec
o
m
m
u
n
icatio
n
s
n
etwo
r
k
s
wh
ile
u
s
in
g
AI
an
d
ML
m
o
d
e
ls
,
wh
ich
ca
n
p
r
o
v
id
e
ad
d
e
d
v
alu
e
with
in
th
e
telec
o
m
in
d
u
s
t
r
y
.
Af
ter
s
ea
r
c
h
in
g
in
d
if
f
er
e
n
t
jo
u
r
n
als
an
d
cr
ed
ib
le
web
s
ites
f
o
r
r
esear
ch
p
u
b
licatio
n
s
as
m
en
tio
n
ed
ab
o
v
e,
5
0
ar
ticles
wer
e
ca
r
ef
u
lly
s
elec
ted
.
E
ac
h
p
a
p
e
r
co
n
tain
s
th
r
ee
m
ain
elem
en
ts
,
s
u
ch
as
aim
o
f
th
e
s
tu
d
y
,
m
eth
o
d
o
lo
g
y
,
a
n
d
ac
cu
r
ac
y
.
Please r
ef
er
t
o
th
e
r
e
s
u
lts
an
d
d
is
cu
s
s
io
n
s
ec
tio
n
f
o
r
m
o
r
e
d
etails.
3.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
A
l
B
o
u
g
h
a
[
5
]
p
r
o
p
o
s
e
s
t
h
e
u
s
e
o
f
d
i
f
f
e
r
e
n
t
d
a
t
a
m
i
n
i
n
g
c
l
a
s
s
i
f
ic
a
t
i
o
n
al
g
o
r
i
t
h
m
s
i
n
d
e
t
e
c
t
i
n
g
a
p
a
r
t
i
c
u
l
a
r
t
e
l
ec
o
m
f
r
a
u
d
t
y
p
e
,
s
u
c
h
a
s
s
u
b
s
c
r
i
b
e
r
i
d
e
n
t
it
y
m
o
d
u
l
e
(
S
I
M
)
box
.
T
h
i
s
p
a
p
e
r
c
o
m
p
a
r
e
s
t
h
e
p
e
r
f
o
r
m
a
n
c
e
o
f
f
o
u
r
d
i
f
f
e
r
e
n
t
M
L
m
o
d
e
l
s
s
u
c
h
a
s
b
o
o
s
te
d
tr
e
e
s
c
l
ass
i
f
i
e
r
s
,
s
u
p
p
o
r
t
v
e
c
t
o
r
m
a
c
h
i
n
e
s
(
SV
M
)
,
l
o
g
i
s
ti
c
cl
a
s
s
i
f
ie
r
s
,
a
n
d
n
e
u
r
a
l
n
e
t
w
o
r
k
s
(
N
N
)
.
F
i
n
d
i
n
g
s
o
f
th
e
s
t
u
d
y
s
u
g
g
e
s
t
t
h
a
t
t
h
e
b
o
o
s
t
e
d
t
r
e
e
s
a
n
d
l
o
g
ic
c
l
a
s
s
i
f
ie
r
s
p
e
r
f
o
r
m
e
d
t
h
e
b
es
t
w
h
e
n
c
o
m
p
a
r
e
d
t
o
t
h
e
o
t
h
e
r
m
o
d
e
l
s
w
i
t
h
a
n
a
c
c
u
r
ac
y
o
f
m
o
r
e
t
h
a
n
9
1
%
a
n
d
a
f
a
l
s
e
p
o
s
i
ti
v
e
r
a
t
e
s
m
a
ll
e
r
t
h
a
n
1
%
.
S
al
l
e
h
u
d
d
i
n
e
t
a
l
.
[
6
]
s
u
g
g
e
s
t
t
h
e
u
s
e
o
f
9
f
e
at
u
r
e
s
t
o
i
d
e
n
t
i
f
y
S
I
M
b
o
x
f
r
a
u
d
w
h
i
l
e
u
s
i
n
g
a
r
ti
f
i
c
ia
l
n
e
u
r
a
l
n
et
w
o
r
k
(
A
N
N
)
a
n
d
S
V
M
m
o
d
el
s
.
A
ll
c
o
m
b
i
n
a
t
i
o
n
s
o
f
t
h
e
p
a
r
a
m
e
t
e
r
s
h
a
v
e
b
e
e
n
t
e
s
te
d
t
h
o
u
g
h
a
1
0
c
r
o
s
s
v
a
l
i
d
a
t
i
o
n
t
o
c
h
e
c
k
t
h
e
p
e
r
f
o
r
m
a
n
c
e
o
f
t
h
e
s
e
t
w
o
M
L
m
o
d
e
ls
.
T
h
e
b
e
s
t
m
o
d
e
l
s
we
r
e
c
h
o
s
e
n
o
n
t
h
e
c
r
i
t
e
r
i
a
o
f
a
c
c
u
r
a
c
y
,
t
i
m
e
,
g
e
n
e
r
a
l
i
z
a
ti
o
n
e
r
r
o
r
a
n
d
p
r
e
c
i
s
i
o
n
f
r
o
m
8
0
A
N
N
a
n
d
4
0
S
V
M
m
o
d
e
l
s
c
r
e
a
t
e
d
.
T
h
e
s
e
t
w
o
b
es
t
m
o
d
e
ls
w
e
r
e
t
h
e
n
c
o
m
p
a
r
e
d
w
i
t
h
ea
ch
o
t
h
e
r
b
y
a
p
p
l
y
i
n
g
d
i
f
f
e
r
e
n
t
p
a
r
t
i
t
i
o
n
s
o
f
t
r
a
i
n
i
n
g
a
n
d
t
e
s
t
i
n
g
.
S
u
c
h
p
a
p
e
r
s
u
g
g
e
s
t
s
t
h
a
t
SV
M
h
a
s
a
h
i
g
h
e
r
a
c
c
u
r
a
c
y
o
f
9
9
.
0
6
%
t
h
a
n
A
N
N
w
i
t
h
9
8
.
6
9
%
.
M
o
r
e
o
v
e
r
,
t
h
e
f
o
r
m
e
r
p
e
r
f
o
r
m
s
b
e
t
t
e
r
i
n
t
e
r
m
s
o
f
f
a
l
s
e
n
e
g
a
ti
v
e
s
a
n
d
t
ak
e
s
t
h
r
e
e
t
i
m
es
l
es
s
p
r
o
c
es
s
i
n
g
t
i
m
e
t
h
a
n
t
h
e
l
a
t
te
r
w
h
e
n
d
e
t
e
c
t
i
n
g
S
I
M
b
o
x
.
K
r
e
n
k
e
r
e
t
a
l
.
[
7
]
s
h
o
w
c
as
e
u
s
e
o
f
b
i
d
i
r
e
c
t
i
o
n
a
l
a
r
t
i
f
i
ci
a
l
n
e
u
r
al
n
e
t
w
o
r
k
(
B
i
-
A
N
N
)
t
o
d
e
t
e
c
t
m
o
b
i
l
e
p
h
o
n
e
f
r
a
u
d
b
y
p
r
e
d
i
c
t
i
n
g
i
n
d
i
v
i
d
u
a
l
b
e
h
a
v
i
o
r
s
.
T
h
i
s
s
y
s
t
e
m
m
a
n
a
g
e
s
t
o
d
e
te
c
t
f
r
a
u
d
i
n
o
f
f
l
i
n
e
p
r
o
c
e
s
s
i
n
g
a
n
d
r
e
a
l
t
i
m
e
C
DR
s
.
M
o
r
e
o
v
e
r
,
s
u
c
h
m
o
d
e
l
s
w
e
r
e
a
b
l
e
t
o
p
r
e
d
i
c
t
c
o
m
p
l
e
x
t
i
m
e
s
e
r
i
e
s
s
u
c
h
a
s
c
al
l
d
u
r
a
t
i
o
n
,
b
u
t
w
it
h
a
b
et
t
e
r
ac
cu
r
a
c
y
o
f
9
0
%
w
h
e
n
c
o
m
p
a
r
e
d
t
o
o
t
h
e
r
m
o
d
e
ls
.
I
t
w
as
al
s
o
c
o
n
c
l
u
d
e
d
t
h
a
t
t
h
e
l
o
n
g
e
r
i
n
p
u
t
t
i
m
e
s
e
r
i
es
a
n
d
s
h
o
r
t
e
r
p
r
e
d
i
c
t
e
d
t
i
m
e
s
e
r
i
es
y
ie
l
d
e
d
b
e
t
t
e
r
r
e
s
u
l
ts
,
w
h
il
e
t
h
e
n
u
m
b
e
r
o
f
h
i
d
d
e
n
l
a
y
e
r
s
i
n
Bi
-
AN
N
w
a
s
c
r
u
ci
a
l
in
i
m
p
r
o
v
i
n
g
t
h
e
p
r
e
d
i
c
t
i
o
n
r
e
s
u
l
t
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
S
ystema
tic
r
ev
iew
o
f fra
u
d
d
etec
tio
n
u
s
in
g
A
I
a
n
d
ML
w
ith
a
n
emp
h
a
s
is
o
n
…
(
S
o
ly
Ma
th
e
w
B
iju
)
3271
Sar
av
an
an
et
a
l.
[
8
]
ad
o
p
t
a
p
r
o
b
ab
ilit
y
-
b
ased
m
o
d
el
s
u
ch
a
s
n
aïv
e
B
ay
esian
to
p
r
e
d
ict
s
u
b
s
cr
ip
tio
n
f
r
au
d
wh
ile
also
u
s
in
g
s
ix
attr
ib
u
tes
to
cr
ea
te
th
e
f
r
au
d
p
r
o
f
iles
s
u
ch
as
d
u
r
atio
n
,
d
ay
tim
e,
n
ig
h
ttime
,
wee
k
d
ay
s
,
wee
k
en
d
s
,
an
d
to
t
al
ca
ll
co
u
n
t.
T
o
r
eso
lv
e
th
e
m
is
class
if
icatio
n
o
f
f
r
au
d
u
len
t
ca
s
es
d
u
e
to
th
e
th
r
esh
o
ld
-
b
ased
n
atu
r
e
o
f
th
e
m
o
d
el,
th
is
s
tu
d
y
u
tili
ze
s
n
aïv
e
B
ay
esian
f
o
r
co
n
tin
u
o
u
s
v
al
u
es
an
d
an
ad
ju
s
ted
v
er
s
io
n
o
f
Ku
llb
ac
k
–
L
eib
ler
(
KL
)
d
iv
er
g
e
n
ce
,
wh
ich
s
u
p
p
o
r
ts
in
f
in
d
in
g
s
th
e
d
if
f
e
r
en
ce
b
etwe
en
two
p
r
o
b
a
b
ilis
tic
q
u
an
titi
es
an
d
r
e
s
o
lv
es
th
e
n
ee
d
o
f
a
b
ig
tr
ain
i
n
g
d
ataset.
T
h
e
s
tu
d
y
s
h
o
ws
th
at
th
e
C
DR
s
th
at
h
av
e
b
ig
d
iv
er
g
en
ce
v
alu
e
id
e
n
tify
th
e
f
r
au
d
u
len
t
cu
s
to
m
er
s
.
I
n
th
e
s
am
e
lin
e
o
f
th
o
u
g
h
t,
W
as
s
ie
[
9
]
s
u
g
g
ests
th
e
u
s
e
o
f
a
p
r
ed
ictiv
e
m
o
d
el
u
s
in
g
d
ee
p
au
to
en
co
d
er
in
d
etec
tin
g
s
u
b
s
cr
ip
tio
n
f
r
au
d
.
T
h
is
m
o
d
el
p
er
f
o
r
m
s
with
9
8
.
9
5
%
ac
cu
r
ac
y
o
n
d
etec
tin
g
r
ea
l
f
r
au
d
ca
s
es
b
y
co
m
b
in
in
g
d
ee
p
n
eu
r
al
au
to
en
c
o
d
er
,
co
n
v
o
l
u
tio
n
al
n
eu
r
al
n
etwo
r
k
(
C
NN)
,
an
d
lo
n
g
s
h
o
r
t
-
te
r
m
m
em
o
r
y
(
L
T
SM)
au
to
en
co
d
er
al
g
o
r
ith
m
s
with
th
r
esh
o
ld
s
o
f
0
.
0
1
0
3
a
n
d
0
.
0
1
7
3
,
r
esp
ec
tiv
e
ly
.
Gr
o
s
s
er
et
a
l.
[
1
0
]
also
s
u
g
g
est
th
e
u
s
e
o
f
NN
in
cr
ea
tin
g
m
o
b
ile
f
r
au
d
u
len
t
u
s
er
s
’
p
r
o
f
iles
b
ased
o
n
s
ev
e
r
al
p
ar
am
eter
s
s
u
c
h
as
in
ter
n
atio
n
al
m
o
b
ile
s
u
b
s
cr
ib
er
i
d
e
n
tity
(
I
MSI
)
,
d
ate,
tim
e,
d
u
r
ati
o
n
,
an
d
ca
ll
ty
p
e
(
l
o
ca
l,
n
atio
n
al,
an
d
in
te
r
n
atio
n
al)
.
On
ce
s
u
c
h
p
r
o
f
iles
ar
e
cr
e
ated
,
th
e
NN
m
o
d
el
is
tr
ain
ed
to
d
etec
t
s
u
ch
f
r
au
d
u
len
t
b
e
h
av
io
r
al
p
atter
n
s
,
w
h
ich
m
ak
e
s
u
r
e
t
h
e
m
o
d
el
p
er
f
o
r
m
s
with
h
ig
h
ac
cu
r
ac
y
in
d
etec
tin
g
r
ea
l
f
u
tu
r
e
f
r
au
d
ca
s
es
an
d
p
r
o
v
id
e
s
m
o
r
e
in
s
ig
h
ts
in
to
th
e
ch
an
g
e
o
f
s
u
ch
u
s
er
s
’
b
eh
av
io
r
o
v
er
tim
e.
E
lm
i
et
a
l
.
[
1
1
]
s
u
p
p
o
r
t
th
e
u
s
e
o
f
a
n
ANN
u
tili
zin
g
a
m
u
lti
-
lay
er
p
er
ce
p
tr
o
n
(
MLP
)
class
if
ier
in
d
etec
tin
g
SIM
b
o
x
f
r
a
u
d
.
T
h
is
s
tu
d
y
cr
ea
te
d
2
4
0
NN
m
o
d
els
wh
ich
wer
e
ev
alu
ated
r
e
g
ar
d
in
g
th
eir
ac
cu
r
ac
y
,
g
e
n
er
aliza
tio
n
er
r
o
r
,
b
u
ild
in
g
tim
e,
p
r
ec
i
s
io
n
,
an
d
r
ec
o
r
d
.
R
esu
lts
s
h
o
wed
th
at
th
e
b
est
p
er
f
o
r
m
in
g
ac
c
u
r
ac
y
was
ac
h
i
ev
ed
wh
e
n
two
h
id
d
en
lay
er
s
with
a
lear
n
in
g
r
ate
o
f
0
.
6
we
r
e
ap
p
lied
,
m
ea
n
in
g
a
lo
wer
m
o
m
en
tu
m
an
d
a
h
ig
h
er
lear
n
in
g
r
ate
co
m
b
in
atio
n
.
T
h
e
b
est
p
er
f
o
r
m
in
g
NN
m
o
d
el
h
ad
an
ac
cu
r
ac
y
o
f
9
8
.
7
1
% with
a
f
alse n
e
g
ativ
e
r
ate
o
f
2
.
9
%.
San
v
er
an
d
Kar
a
h
o
ca
[
1
2
]
co
n
s
is
t
o
f
a
co
m
p
ar
ativ
e
s
tu
d
y
b
etwe
en
d
if
f
er
en
t
d
ata
m
in
i
n
g
t
ec
h
n
iq
u
es
to
d
eter
m
in
e
t
h
e
b
est
m
o
d
el
f
o
r
th
e
telec
o
m
f
r
au
d
p
r
ed
ictio
n
.
T
h
is
m
o
d
el
u
s
es
s
ev
er
al
d
a
ta
m
in
in
g
m
eth
o
d
s
s
u
ch
as
J
R
ip
,
R
id
o
r
,
PAR
T
,
On
eR,
Nn
g
e,
d
ec
is
io
n
tab
le
,
co
n
ju
n
ctio
n
r
u
les
,
AD
tr
ee
,
I
B
1
,
B
ay
esn
et,
an
d
ad
ap
tiv
e
n
e
u
r
o
f
u
zz
y
in
f
er
en
c
e
(
ANFI
S)
.
Fin
d
in
g
s
o
f
th
e
s
tu
d
y
s
u
g
g
est th
at
th
e
ANFI
S m
e
th
o
d
is
th
e
b
est o
n
e
with
an
ac
cu
r
ac
y
o
f
9
8
.
3
3
%,
s
en
s
itiv
ity
o
f
9
7
,
a
n
d
s
p
ec
if
icit
y
o
f
9
9
%.
On
e
o
f
th
e
m
ai
n
ad
v
an
tag
es
o
f
ANFI
S
m
eth
o
d
is
th
at
it
co
m
b
in
es
th
e
p
r
ec
is
io
n
o
f
f
u
zz
y
lo
g
ic
class
if
icatio
n
an
d
th
e
a
d
ap
tatio
n
o
f
NN
s
,
p
r
o
v
id
in
g
a
g
r
ea
t
p
er
f
o
r
m
an
ce
.
T
h
ad
d
e
u
s
et
a
l.
[
1
3
]
d
o
s
u
p
p
o
r
t
th
e
u
s
e
o
f
th
e
ANN
m
o
d
el
in
p
r
e
d
ictin
g
SIM
b
o
x
.
A
class
o
f
ANN
m
o
d
el,
n
am
el
y
ML
P,
was
u
s
ed
to
an
aly
ze
th
e
C
DR
s
b
y
b
ein
g
tr
ain
ed
5
0
t
im
es
in
a
d
ata
s
et
co
n
tain
in
g
tr
u
e
d
ata
class
if
ied
as
tr
u
e
an
d
n
o
t
tr
u
e
SIM
b
o
x
ca
s
es
an
d
u
s
in
g
s
p
ec
if
ic
p
ar
a
m
eter
s
s
u
ch
as
ca
ll
d
u
r
atio
n
,
ca
ll
f
r
e
q
u
en
c
y
,
tim
e,
o
r
ig
in
atin
g
,
an
d
ter
m
in
atin
g
n
u
m
b
er
.
T
h
e
ac
c
u
r
ac
y
o
f
t
h
e
A
NN
im
p
r
o
v
es
f
r
o
m
5
7
.
0
4
%
to
8
0
.
0
1
%
af
ter
5
0
it
er
atio
n
s
with
a
to
p
lay
e
r
o
f
1
6
n
o
d
es.
Hilas
an
d
Ma
s
to
r
o
c
o
s
tas
[
1
4
]
co
m
p
ar
e
s
ev
er
al
ML
m
o
d
els,
s
u
ch
as
r
an
d
o
m
f
o
r
est
(
R
F)
,
lo
g
is
tic
r
eg
r
ess
io
n
,
SVM,
g
r
ad
ie
n
t
b
o
o
s
tin
g
,
an
d
n
aiv
e
B
ay
e
s
,
to
p
r
ed
ict
th
e
I
n
ter
n
a
tio
n
al
R
ev
en
u
e
Sh
ar
e
Fra
u
d
(
I
R
SF
)
.
T
h
e
ev
alu
atio
n
o
f
th
e
b
est
m
o
d
el
was
co
n
d
u
cte
d
o
n
s
ev
er
al
cr
iter
ia
s
u
ch
as a
cc
u
r
ac
y
,
p
r
ec
is
io
n
,
r
ec
all,
an
d
F1
-
s
co
r
e.
Fin
d
in
g
s
o
f
th
e
s
tu
d
y
co
n
clu
d
e
th
at
th
e
b
est
m
o
d
el
f
o
r
I
R
SF
p
r
ed
ictio
n
is
R
F
with
9
9
%
ac
cu
r
ac
y
wh
ile
also
co
n
v
ey
i
n
g
th
a
t
ML
m
o
d
els
ar
e
a
p
o
wer
f
u
l
to
o
l
th
at
s
h
o
u
ld
b
e
c
o
n
s
id
er
ed
f
o
r
telec
o
m
f
r
au
d
d
etec
tio
n
an
d
m
itig
atio
n
.
Mo
r
e
o
v
er
,
L
i
et
a
l
.
[
1
5
]
also
s
u
g
g
est
th
e
u
s
e
o
f
d
if
f
er
en
t
ML
m
o
d
els
in
d
etec
tin
g
m
alicio
u
s
p
h
o
n
e
ca
lls
.
T
h
e
s
tu
d
y
co
n
s
id
er
s
2
9
p
ar
am
eter
s
o
f
C
DR
s
an
d
th
e
h
is
to
r
ical
d
ata
to
o
b
tain
t
h
e
av
er
ag
e
f
ir
s
t
p
r
ed
ictio
n
(
AFP)
m
etr
ic,
wh
ich
p
r
esen
ts
th
e
am
o
u
n
ts
o
f
ca
ll
s
a
f
r
au
d
u
len
t
p
e
r
s
o
n
m
u
s
t
m
ak
e
b
ef
o
r
e
b
ein
g
p
r
ed
icted
an
d
b
lo
ck
ed
in
ad
v
an
ce
.
Ou
t o
f
all
th
e
m
o
d
els,
R
F is
th
e
b
est o
n
e
with
9
0
% a
cc
u
r
ac
y
.
E
l
m
i
et
a
l
.
[
1
6
]
s
u
g
g
e
s
t
t
h
e
u
s
e
o
f
S
V
M
a
n
d
AN
N
i
n
d
e
t
ec
t
i
n
g
S
I
M
b
o
x
i
n
t
h
e
t
el
e
c
o
m
m
u
n
i
c
a
t
i
o
n
i
n
d
u
s
t
r
y
.
F
o
r
t
h
e
d
e
v
e
l
o
p
m
e
n
t
o
f
t
h
e
A
N
N
m
o
d
e
l
,
2
4
0
m
o
d
e
l
s
w
it
h
a
l
l
t
h
e
p
o
s
s
i
b
le
p
a
r
a
m
e
t
e
r
c
o
m
b
i
n
a
t
i
o
n
s
w
e
r
e
c
r
e
a
te
d
a
n
d
t
h
e
m
o
d
e
ls
’
e
v
a
l
u
a
t
i
o
n
wa
s
c
o
n
d
u
c
t
e
d
b
as
e
d
o
n
p
r
e
d
i
c
t
i
o
n
ac
c
u
r
a
c
y
,
g
e
n
e
r
a
l
i
z
a
ti
o
n
e
r
r
o
r
,
r
u
n
n
i
n
g
t
i
m
e
,
p
r
e
c
is
i
o
n
,
a
n
d
r
ec
a
l
l
.
S
i
m
il
a
r
l
y
,
f
o
r
t
h
e
S
VM
m
o
d
e
l
,
4
0
m
o
d
e
l
s
w
e
r
e
c
r
e
at
e
d
,
w
h
i
l
e
t
h
e
e
v
a
l
u
a
ti
o
n
c
r
i
t
e
r
i
a
w
e
r
e
p
r
e
d
i
c
t
i
o
n
a
c
c
u
r
ac
y
,
r
u
n
n
i
n
g
t
i
m
e
,
p
r
e
c
i
s
i
o
n
,
a
n
d
r
e
c
a
l
l
r
a
t
e
.
F
i
n
d
i
n
g
s
o
f
t
h
e
p
a
p
e
r
s
u
g
g
e
s
t
t
h
a
t
S
V
M
m
o
d
e
l
p
e
r
f
o
r
m
s
b
e
t
te
r
th
a
n
t
h
e
A
N
N
w
it
h
9
8
.
9
%
a
c
c
u
r
a
c
y
,
9
9
.
0
6
4
%
c
l
a
s
s
i
f
i
c
a
ti
o
n
a
c
c
u
r
a
c
y
o
f
f
a
l
s
e
p
o
s
i
t
i
v
es
a
n
d
n
e
g
a
ti
v
e
s
a
n
d
5
.
5
5
s
e
c
o
n
d
s
r
u
n
n
i
n
g
ti
m
e
.
B
ar
s
o
n
e
t
a
l
.
[
1
7
]
e
x
p
l
o
r
e
t
h
e
u
s
e
o
f
a
s
u
p
e
r
v
is
e
d
a
p
p
r
o
a
c
h
u
t
i
l
i
z
i
n
g
t
h
e
M
L
P
n
et
w
o
r
k
i
n
d
e
t
e
c
ti
n
g
m
o
b
i
l
e
p
h
o
n
e
f
r
a
u
d
s
w
i
t
h
9
2
.
5
%
ac
c
u
r
a
c
y
.
F
o
r
e
a
c
h
u
s
e
r
,
t
h
e
d
e
t
a
i
ls
f
r
o
m
t
h
e
C
DR
h
e
l
p
c
r
ea
t
e
a
u
s
e
r
p
r
o
f
i
l
e
,
w
h
i
c
h
c
a
n
b
e
l
o
n
g
t
o
h
i
s
t
o
r
i
c
a
l
o
r
r
e
ce
n
t
t
im
e
s
.
B
y
u
s
i
n
g
a
M
L
P
n
e
t
w
o
r
k
,
w
it
h
1
8
i
n
p
u
ts
a
n
d
t
w
o
o
u
t
p
u
ts
(
v
a
li
d
o
r
f
r
a
u
d
)
,
t
h
e
m
o
d
e
l
i
s
t
es
t
e
d
a
n
d
t
r
ai
n
e
d
c
o
n
t
i
n
u
o
u
s
l
y
t
o
i
m
p
r
o
v
e
t
h
e
d
e
t
e
c
t
i
o
n
,
h
o
w
e
v
e
r
,
t
h
e
s
i
z
e
o
f
t
h
e
ti
m
e
w
i
n
d
o
w
f
r
o
m
w
h
i
c
h
t
h
e
r
e
c
e
n
t
p
a
t
ter
n
b
e
h
a
v
i
o
r
i
s
t
a
k
e
n
a
n
d
t
h
e
a
g
i
n
g
o
f
t
h
e
d
at
a
u
s
e
d
m
a
y
i
m
p
a
c
t
t
h
e
a
c
c
u
r
a
c
y
.
M
u
r
y
n
e
t
s
et
a
l
.
[
1
8
]
s
u
g
g
es
t
u
s
e
o
f
d
i
f
f
e
r
e
n
t
cl
as
s
i
f
i
e
r
s
t
o
d
e
t
e
ct
S
I
M
b
o
x
c
as
es
wi
t
h
h
i
g
h
a
c
c
u
r
a
c
y
.
Na
m
e
l
y
,
t
h
i
s
s
tu
d
y
u
s
e
s
a
l
t
e
r
n
at
i
n
g
d
e
ci
s
i
o
n
tr
e
e
,
f
u
n
c
t
i
o
n
a
l
t
r
e
e
,
a
n
d
RF
a
n
d
e
v
a
l
u
a
t
es
t
h
e
i
r
p
e
r
f
o
r
m
a
n
c
e
o
n
t
h
e
i
r
o
w
n
w
i
t
h
a
n
a
c
c
u
r
a
c
y
o
f
9
9
.
1
%
,
9
9
.
9
0
%
,
a
n
d
9
9
.
9
3
%
,
r
e
s
p
e
c
t
i
v
el
y
.
T
h
e
s
t
u
d
y
d
o
e
s
al
s
o
c
o
m
b
i
n
e
a
l
l
t
h
r
e
e
cl
as
s
i
f
i
e
r
s
i
n
t
o
a
c
l
as
s
i
f
i
c
a
ti
o
n
r
u
l
e
,
w
h
i
c
h
p
e
r
f
o
r
m
s
b
e
t
t
e
r
t
h
a
n
a
n
y
s
t
a
n
d
a
l
o
n
e
cl
a
s
s
i
f
ie
r
w
i
t
h
9
9
.
9
5
%
p
r
e
d
i
c
ti
o
n
a
c
c
u
r
ac
y
o
f
t
h
e
S
I
M
b
o
x
c
a
s
es
.
Ho
wev
er
,
Ma
r
a
h
et
a
l.
[
1
9
]
s
u
g
g
est
th
e
u
s
e
o
f
f
u
zz
y
lo
g
ic
an
d
p
r
o
f
ilin
g
t
h
e
f
r
a
u
d
u
l
en
t
u
s
er
s
’
b
eh
av
io
r
to
d
etec
t
SIM
b
o
x
.
T
h
e
s
tu
d
y
u
tili
ze
s
r
ea
l
-
life
d
ata,
o
u
t
o
f
wh
ich
f
iv
e
p
ar
am
eter
s
s
u
ch
as
s
u
b
s
cr
ib
er
’
s
m
o
b
ilit
y
,
in
co
m
i
n
g
to
o
u
tg
o
in
g
ca
lls
r
atio
,
s
u
s
p
icio
u
s
ce
ll
ac
tiv
ity
,
ir
r
eg
u
la
r
ca
lls
,
an
d
s
er
v
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.
15
,
No
.
4
,
Au
g
u
s
t 2
0
2
6
:
3
2
6
9
-
3
2
8
5
3272
ty
p
e,
ar
e
u
s
ed
to
b
u
ild
th
e
m
e
m
b
er
s
h
ip
f
u
n
ctio
n
(
MF)
an
d
t
h
e
f
u
zz
y
lo
g
ic
r
u
les.
E
ac
h
r
u
l
e
h
as
in
p
u
ts
wh
ich
ar
e
th
e
ch
o
s
en
p
a
r
am
eter
s
,
a
n
o
u
tp
u
t
wh
ic
h
ca
n
b
e
f
r
au
d
o
r
n
o
t
f
r
au
d
,
as
well
as
a
d
ef
in
ed
th
r
esh
o
ld
.
Mo
r
eo
v
er
,
ea
ch
r
u
le
h
as
th
e
m
in
im
u
m
an
d
m
ax
im
u
m
v
al
u
es
d
ef
in
ed
.
Su
c
h
v
alu
es
d
o
af
f
ec
t
h
o
w
th
e
m
o
d
el
p
er
f
o
r
m
s
an
d
ca
n
b
e
ea
s
ily
ad
ju
s
ted
in
b
u
ild
in
g
th
e
v
is
u
aliza
tio
n
m
o
d
el.
T
h
e
m
o
d
el
o
b
tain
s
th
e
I
MSI
an
d
th
e
th
r
esh
o
ld
b
y
th
e
u
s
er
a
n
d
p
r
o
v
id
es
th
e
f
r
au
d
s
co
r
e
MF.
I
f
t
h
e
latter
b
r
ea
ch
es
th
e
t
h
r
esh
o
ld
v
alu
e,
ca
s
es
ar
e
co
n
s
id
er
ed
as
f
r
a
u
d
u
len
t.
H
o
wev
er
,
th
e
m
o
d
el
n
ee
d
s
to
b
e
tr
ain
ed
o
n
t
r
u
s
ted
d
ata,
m
ea
n
in
g
d
ata
th
at
is
co
n
f
ir
m
e
d
as
f
r
au
d
u
len
t
f
o
r
ac
cu
r
ac
y
im
p
r
o
v
em
e
n
t.
Fo
llo
win
g
th
e
s
am
e
lin
e
o
f
th
o
u
g
h
t,
Og
wu
elek
a
[
2
0
]
also
s
u
g
g
ests
th
e
u
s
e
o
f
u
s
er
p
r
o
f
il
in
g
an
d
class
if
icatio
n
tech
n
iq
u
es
to
d
etec
t
f
r
au
d
in
m
o
b
ile
telec
o
m
m
u
n
icatio
n
n
etwo
r
k
s
.
T
h
e
f
ir
s
t
s
tep
o
f
th
e
s
tu
d
y
co
n
s
is
ts
o
f
lear
n
in
g
th
e
u
s
er
p
r
o
f
ilin
g
f
r
o
m
la
b
eled
d
ata,
wh
ich
is
th
e
n
later
u
s
ed
to
d
etec
t
f
r
a
u
d
.
B
o
t
h
p
r
o
b
ab
ilis
tic
an
d
NN
s
wer
e
tr
ain
ed
at
u
s
er
lev
el,
u
s
er
p
r
o
f
ile
lev
el
,
an
d
class
lev
el,
o
u
t
o
f
wh
ic
h
th
e
s
ec
o
n
d
p
er
f
o
r
m
e
d
b
etter
.
T
h
is
s
tu
d
y
u
s
es
th
e
r
ec
eiv
e
r
o
p
er
atin
g
c
h
ar
ac
ter
is
tic
(
R
OC
)
cu
r
v
e
f
o
r
b
o
th
tr
ai
n
in
g
an
d
test
in
g
s
ets,
wh
ich
r
esu
lted
in
less
th
an
3
%
o
f
f
alse
p
o
s
itiv
es,
en
tailin
g
m
o
r
e
th
a
n
9
7
%
ac
cu
r
ac
y
.
Ar
an
u
wa
[
2
1
]
p
r
o
v
id
es
a
n
i
n
ter
esti
n
g
h
y
b
r
id
m
o
d
el
th
at
m
i
x
es
d
if
f
e
r
en
tial
an
d
ab
s
o
lu
te
an
aly
s
is
to
m
o
d
el
th
e
n
etwo
r
k
m
ec
h
a
n
is
m
wh
ile
u
tili
zin
g
th
e
NN
s
to
lear
n
th
e
s
u
b
s
cr
ib
er
s
’
ca
ll
d
ata.
T
h
e
ab
s
o
lu
te
ap
p
r
o
ac
h
p
r
o
v
id
es
w
h
eth
er
a
u
s
er
is
f
r
au
d
u
len
t o
r
n
o
t
b
ased
u
p
o
n
th
e
ca
ll p
atter
n
s
tatis
t
ics
,
wh
ile
th
e
d
if
f
er
en
tial
o
n
e
f
o
c
u
s
es
o
n
c
h
an
g
es
o
f
b
eh
av
i
o
r
p
atter
n
s
wh
en
m
o
v
in
g
f
r
o
m
a
n
o
r
m
al
u
s
er
to
a
f
r
au
d
o
n
e.
R
esu
lts
o
f
th
e
s
tu
d
y
p
r
o
v
id
e
an
8
9
%
ac
cu
r
ac
y
a
n
d
2
.
7
1
m
ea
n
ab
s
o
lu
te
er
r
o
r
(
MA
E
)
in
d
etec
tin
g
f
r
a
u
d
o
f
p
o
s
tp
aid
m
o
b
ile
co
m
m
u
n
icatio
n
n
etwo
r
k
s
.
R
eb
ah
i
et
a
l
.
[
2
2
]
p
r
o
p
o
s
e
a
f
u
ll
f
r
am
ewo
r
k
o
f
im
p
lem
en
tin
g
a
p
latf
o
r
m
th
at
m
itig
ates
f
r
au
d
o
n
v
o
ice
o
v
er
in
ter
n
et
p
r
o
to
co
l
(
Vo
I
P)
en
v
ir
o
n
m
e
n
ts
.
T
h
is
f
r
a
m
ewo
r
k
is
b
u
ilt
o
n
u
n
s
u
p
e
r
v
i
s
ed
m
eth
o
d
s
s
u
ch
as
s
ig
n
atu
r
e
an
d
cl
u
s
ter
in
g
-
b
ase
d
tech
n
iq
u
es.
T
h
e
m
ain
id
ea
o
f
th
e
f
r
a
m
ewo
r
k
is
to
u
s
e
clu
s
ter
in
g
m
eth
o
d
s
to
tr
ac
k
ch
a
n
g
es
i
n
u
s
er
b
e
h
av
io
r
b
y
f
ir
s
t
tr
ain
in
g
th
e
m
o
d
el
wit
h
d
ata
th
at
is
ass
u
m
ed
to
b
e
cl
ea
n
d
ata
o
f
n
o
r
m
al
u
s
er
s
an
d
test
in
g
to
en
s
u
r
e
th
at
th
e
u
s
er
b
eh
av
io
r
d
o
es
m
atch
with
th
e
n
o
r
m
al
u
s
er
b
e
h
a
v
io
r
lear
n
ed
i
n
th
e
f
o
r
m
er
p
h
ase.
T
h
e
m
ain
p
ar
a
m
eter
s
th
at
ar
e
u
s
ed
f
o
r
clu
s
ter
in
g
ar
e
A
-
n
u
m
b
er
,
B
-
n
u
m
b
er
,
s
tar
t
tim
e,
d
u
r
atio
n
,
s
tatu
s
,
A
-
I
P,
B
-
I
P,
g
r
o
u
p
,
an
d
f
lag
s
.
W
h
ile
th
is
f
r
am
ewo
r
k
is
s
till
u
n
d
er
d
ev
elo
p
m
en
t,
it
en
v
is
io
n
s
a
3
6
0
s
o
lu
tio
n
,
f
r
o
m
o
b
tain
in
g
Vo
I
P
C
D
R
s
d
ata
to
m
o
d
elin
g
,
an
aly
s
is
,
an
d
au
to
m
atio
n
o
f
th
e
alar
m
s
wh
en
ev
er
s
o
m
e
ch
an
g
es
i
n
th
e
cl
u
s
ter
in
g
m
o
d
els
ar
e
in
d
icativ
e
o
f
f
r
a
u
d
u
len
t
ac
tiv
ities
.
Ho
llm
én
[
2
3
]
also
p
r
o
p
o
s
es
th
e
u
s
e
o
f
an
u
n
s
u
p
er
v
is
ed
ap
p
r
o
ac
h
in
d
etec
tin
g
m
o
b
ile
c
o
m
m
u
n
icatio
n
s
n
etwo
r
k
f
r
au
d
u
s
in
g
NN
s
an
d
p
r
o
b
a
b
ilis
tic
m
o
d
els
.
Ho
wev
e
r
,
th
e
ad
d
e
d
v
al
u
e
o
f
th
is
wo
r
k
co
n
s
is
ts
o
f
t
h
e
f
ac
t
th
at
it
le
ar
n
s
h
o
w
to
d
etec
t
f
r
au
d
b
eh
a
v
io
r
f
r
o
m
p
ar
tially
l
ab
eled
d
ata
with
an
u
n
k
n
o
wn
m
ix
in
g
m
ec
h
an
is
m
,
b
ec
au
s
e
t
h
e
d
ata
s
et
d
o
es
n
o
t
d
if
f
er
en
tiate
th
e
f
r
au
d
an
d
n
o
n
f
r
au
d
ca
s
es.
Mo
r
eo
v
er
,
th
e
r
ep
r
esen
tatio
n
o
f
d
ata
is
im
p
o
r
tan
t
f
o
r
th
e
p
r
ed
ictio
n
’
s
ac
cu
r
ac
y
b
u
t
th
e
tr
ad
e
-
o
f
f
b
etwe
en
t
h
e
laten
c
y
in
d
etec
tio
n
tim
e
an
d
r
ich
n
ess
o
f
d
escr
ip
tio
n
s
h
o
u
ld
b
e
co
n
s
id
er
ed
.
T
h
e
m
o
d
els
p
er
f
o
r
m
r
elativ
ely
w
ell
with
a
r
atio
o
f
2
%
–
3
%
o
f
f
alse
p
o
s
itiv
es.
Mo
r
ea
u
et
a
l.
[
2
4
]
s
u
g
g
ested
s
ev
er
al
f
r
au
d
d
etec
tio
n
s
tr
ateg
i
es in
m
o
b
ile
telec
o
m
m
u
n
icati
o
n
n
etwo
r
k
s
s
u
ch
as
s
u
p
er
v
is
ed
NN
,
u
n
s
u
p
er
v
is
ed
n
etwo
r
k
a
n
d
r
u
le
b
ased
.
Acc
o
r
d
in
g
t
o
th
is
p
a
p
er
,
t
h
e
p
ar
a
m
eter
s
th
at
s
h
o
u
ld
b
e
co
n
s
id
er
ed
as
th
e
m
o
s
t
f
r
au
d
-
r
elev
an
t
o
n
es
ar
e
ch
ar
g
ed
I
M
SI,
f
ir
s
t
ce
ll
I
D,
ch
ar
g
ea
b
le
d
u
r
atio
n
,
B
n
u
m
b
er
ty
p
e,
an
d
ca
lle
d
n
u
m
b
er
.
R
e
g
ar
d
in
g
t
h
e
r
u
le
-
b
ased
ap
p
r
o
ac
h
,
b
o
th
a
b
s
o
lu
te
o
r
d
if
f
er
e
n
tial
u
s
ag
e
ca
n
b
e
ap
p
lied
to
p
r
o
v
id
e
a
b
etter
v
i
ew
o
f
wh
eth
e
r
a
u
s
er
is
f
r
au
d
u
len
t
o
r
n
o
t
an
d
wh
at
ch
a
n
g
es
ar
e
o
b
s
er
v
ed
in
h
is
/h
er
b
eh
a
v
io
r
.
Mo
r
eo
v
er
,
th
e
n
ee
d
f
o
r
n
ea
r
r
ea
l
tim
e
d
ata
is
also
em
p
h
asized
f
o
r
th
is
ap
p
r
o
ac
h
.
I
n
s
u
p
er
v
is
ed
NN
ap
p
r
o
ac
h
,
th
e
m
o
d
el
will
g
r
o
u
p
s
im
ilar
p
atter
n
s
in
clu
s
ter
s
,
h
o
wev
er
;
th
e
u
s
er
m
u
s
t
ass
ig
n
wh
ich
class
is
ass
o
ciate
d
with
ea
ch
clu
s
ter
.
Ho
wev
er
,
f
o
r
th
e
s
u
p
er
v
is
ed
NN
ap
p
r
o
ac
h
th
e
d
ata
m
u
s
t
b
e
p
r
e
-
lab
elled
b
ef
o
r
e
lear
n
in
g
o
f
th
e
m
o
d
el.
T
an
ig
u
ch
i
et
a
l.
[
2
5
]
ex
p
lo
r
e
th
e
d
etec
tio
n
o
f
telec
o
m
f
r
a
u
d
b
y
u
s
in
g
NN,
Gau
s
s
ian
m
o
d
el,
an
d
B
ay
esian
n
etwo
r
k
wh
ile
u
s
in
g
th
e
R
OC
cu
r
v
e
t
o
ass
ess
th
eir
r
esp
ec
tiv
e
p
er
f
o
r
m
an
ce
.
T
h
e
f
o
r
m
er
m
o
d
el
u
s
es
s
ev
er
al
p
ar
am
eter
s
s
u
ch
as
d
u
r
atio
n
av
er
a
g
e,
d
u
r
atio
n
s
tan
d
ar
d
d
ev
iatio
n
,
ca
ll
n
u
m
b
er
d
u
r
in
g
t
h
e
d
a
y
,
m
ax
im
u
m
d
u
r
atio
n
,
an
d
ca
ll
n
u
m
b
er
p
e
r
d
a
y
d
u
r
in
g
t
h
e
o
b
s
er
v
ed
tim
e
win
d
o
w.
T
h
e
R
OC
f
o
r
th
is
m
o
d
e
l
in
d
icate
d
8
5
%
ac
cu
r
ac
y
with
o
u
t
an
y
f
alse
p
o
s
itiv
es.
T
h
e
Gau
s
s
ian
p
r
o
b
ab
ilis
tic
m
o
d
el
ca
lcu
lates
th
e
p
r
o
b
a
b
ilit
y
d
en
s
ity
f
u
n
ctio
n
o
f
th
e
p
ast
a
n
d
c
u
r
r
en
t
s
u
b
s
cr
ib
er
’
s
b
eh
a
v
io
r
.
Par
am
eter
s
u
s
ed
in
th
is
ap
p
r
o
ac
h
wer
e
d
aily
ca
lls
co
u
n
t
an
d
ca
l
l
d
u
r
atio
n
d
u
r
in
g
o
n
-
p
ea
k
a
n
d
o
f
f
-
p
ea
k
h
o
u
r
s
.
T
h
e
R
OC
cu
r
v
e
s
u
g
g
ests
a
7
0
%
ac
cu
r
ac
y
o
f
s
u
ch
m
o
d
el
with
o
u
t
an
y
f
alse
p
o
s
itiv
es.
L
ast
ly
,
f
o
r
th
e
B
ay
esian
n
etwo
r
k
,
two
m
o
d
els
wer
e
b
u
ilt,
o
n
e
th
at
m
o
d
eled
th
e
b
eh
a
v
io
r
o
f
f
r
a
u
d
u
le
n
t
u
s
er
s
an
d
a
n
o
th
er
f
o
r
le
g
itima
te
u
s
er
s
to
ca
lcu
late
wh
at
wo
u
ld
b
e
th
e
f
r
au
d
p
r
o
b
ab
ilit
y
f
o
r
th
e
c
o
n
s
id
er
ed
u
s
er
.
T
h
e
R
OC
cu
r
v
e
s
h
o
ws
8
5
%
ac
cu
r
ac
y
in
p
r
ed
ictio
n
.
I
n
an
o
t
h
er
ar
ticle,
Ho
llm
én
a
n
d
T
r
esp
[
2
6
]
s
u
g
g
est
a
h
ier
ar
ch
ical
r
e
g
im
e
-
s
witch
in
g
m
o
d
el
to
d
et
ec
t
ca
ll
-
b
ased
f
r
au
d
.
I
n
f
er
en
ce
r
u
les
o
f
th
e
m
o
d
el
ar
e
r
etr
iev
ed
f
r
o
m
th
e
ju
n
ctio
n
tr
ee
alg
o
r
ith
m
wh
ile
th
e
m
o
d
el
is
tr
ain
ed
u
s
in
g
th
e
ex
p
ec
tatio
n
-
m
ax
im
izatio
n
(
EM
)
an
d
d
is
cr
im
in
ativ
e
alg
o
r
ith
m
s
.
L
astl
y
,
th
e
m
o
d
el
is
test
ed
th
r
o
u
g
h
o
n
-
lin
e
d
etec
tio
n
m
o
d
e
a
n
d
r
etr
o
s
p
e
ctiv
e
class
if
icatio
n
.
Fin
d
in
g
s
o
f
t
h
e
ar
ticle
s
u
g
g
est
th
at
with
a
f
alse
alar
m
p
r
o
b
a
b
ilit
y
o
f
0
.
3
%
th
e
o
n
-
li
n
e
d
etec
tio
n
an
d
r
etr
o
s
p
ec
tiv
e
class
if
icatio
n
p
er
f
o
r
m
well
with
ac
cu
r
ac
y
o
f
9
7
.
4
%
an
d
9
3
.
4
%,
r
esp
ec
tiv
ely
.
Ho
wev
er
,
wh
en
t
h
e
f
al
s
e
alar
m
p
r
o
b
ab
ilit
y
is
in
cr
ea
s
ed
to
2
%,
th
eir
p
er
f
o
r
m
an
ce
d
r
o
p
s
to
9
2
.
8
%
a
n
d
9
2
.
1
%
,
r
esp
ec
tiv
ely
.
Ho
llm
én
et
a
l.
[
2
7
]
s
u
g
g
est
th
e
u
s
e
o
f
a
s
elf
-
o
r
g
an
izin
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
S
ystema
tic
r
ev
iew
o
f fra
u
d
d
etec
tio
n
u
s
in
g
A
I
a
n
d
ML
w
ith
a
n
emp
h
a
s
is
o
n
…
(
S
o
ly
Ma
th
e
w
B
iju
)
3273
m
ap
(
SOM)
t
o
clu
s
ter
p
r
o
b
a
b
ilis
tic
m
o
d
els
th
at
ca
n
h
elp
d
etec
t
f
r
au
d
u
len
t
ca
s
es
in
t
elec
o
m
m
u
n
icatio
n
s
n
etwo
r
k
s
.
T
h
e
p
ap
e
r
tr
ain
s
a
S
OM
o
f
8
0
u
n
its
,
with
a
r
ec
tan
g
u
lar
to
p
o
l
o
g
y
(
8
×
1
0
m
a
p
u
n
i
ts
)
,
o
n
th
e
s
tatis
t
ics
o
f
th
e
tim
e
s
er
ies.
T
r
ain
in
g
o
f
th
e
m
ap
was
co
n
d
u
cted
o
n
two
s
tep
s
,
th
e
co
ar
s
e
o
r
d
er
in
g
an
d
f
in
e
tu
n
in
g
.
R
esu
lts
o
f
th
e
m
ap
t
o
class
-
s
p
ec
if
ic
d
is
tr
ib
u
tio
n
s
s
h
o
w
th
a
t
th
e
u
p
p
er
p
ar
ts
o
f
th
e
m
a
p
ar
e
f
r
au
d
u
len
t
u
s
er
s
wh
ile
th
e
lo
wer
p
ar
ts
ar
e
le
g
itima
te
o
n
es.
Keh
elwa
la
[
2
8
]
p
r
o
p
o
s
es
a
n
ew
ap
p
r
o
ac
h
wh
ich
u
s
es
co
m
p
lex
ev
en
t
p
r
o
ce
s
s
in
g
(
C
E
P),
b
u
s
in
ess
ac
tiv
ity
m
o
n
ito
r
in
g
(
B
AM
)
to
en
s
u
r
e
r
ea
l
-
tim
e
d
etec
tio
n
o
f
o
f
f
-
n
et
an
d
o
n
-
n
et
SIM
b
o
x
ca
s
es.
T
h
e
m
o
d
el
was
ab
le
to
ca
p
tu
r
e
ce
r
tain
f
r
au
d
f
ea
tu
r
es
th
at
ca
n
n
o
t
b
e
m
an
i
p
u
lated
b
y
f
r
au
d
s
ter
s
.
T
h
ese
f
ea
tu
r
es
en
ab
le
th
e
m
o
d
el
to
d
etec
t
g
r
ay
ca
ll
in
s
t
an
ce
s
with
a
9
9
.
9
%
ac
cu
r
ac
y
in
b
o
th
o
n
-
n
et
a
n
d
o
f
f
-
n
et
SI
M
b
o
x
ca
s
es
wh
ile
also
in
cr
ea
s
in
g
th
e
d
etec
tio
n
s
p
ee
d
.
Mo
r
eo
v
e
r
,
th
e
m
o
d
el
n
o
t
o
n
ly
d
etec
ts
f
r
au
d
u
len
t
ca
s
es
b
u
t
also
m
a
n
ag
es
to
p
r
ed
ict
9
.
9
3
ca
ll
attem
p
ts
ea
r
lier
th
an
o
th
er
o
p
er
ato
r
s
’
ex
is
tin
g
s
o
lu
tio
n
s
.
Kash
ir
a
n
d
B
ash
ir
[
2
9
]
tr
y
d
if
f
er
en
t
ML
tech
n
iq
u
es
to
p
r
o
v
id
e
t
h
e
b
est
SIM
b
o
x
p
r
ed
i
ctio
n
ac
cu
r
a
cy
f
o
r
ANN
an
d
SVM
m
o
d
els.
W
h
en
b
u
ild
in
g
ANN
m
o
d
el,
2
5
attr
ib
u
tes
o
f
th
e
u
s
u
al
a
n
d
f
r
au
d
u
len
t
u
s
er
s
a
r
e
u
s
ed
as
in
p
u
ts
wh
ile
th
e
m
o
d
el’
s
p
er
f
o
r
m
an
ce
is
ev
alu
ated
b
as
ed
u
p
o
n
co
n
f
u
s
io
n
m
atr
ix
,
R
OC
,
an
d
v
ar
ian
ts
o
f
ANN.
As
p
er
th
e
co
n
f
u
s
io
n
m
atr
ix
,
th
e
ANN
p
er
f
o
r
m
s
with
an
ac
cu
r
ac
y
o
f
1
0
0
%
in
p
r
ed
ictin
g
SIM
b
o
x
ca
s
es
an
d
9
9
.
3
%
p
r
ed
ictin
g
leg
itima
te
o
n
es.
Mo
r
eo
v
er
,
b
a
s
ed
o
n
th
e
R
OC
,
co
n
tin
u
o
u
s
tr
ain
in
g
o
f
th
e
m
o
d
el
im
p
r
o
v
ed
th
e
r
atio
o
f
th
e
tr
u
e
p
o
s
itiv
es
o
v
er
th
e
f
alse
p
o
s
itiv
es.
L
astl
y
,
5
d
if
f
er
en
t
NN
alg
o
r
ith
m
s
wer
e
u
s
ed
to
p
r
e
d
ict
SIM
b
o
x
ca
s
es
an
d
r
esu
lts
s
u
g
g
est
th
at
B
ay
esian
r
eg
u
lar
izatio
n
p
e
r
f
o
r
m
s
th
e
b
es
t
wh
ile
r
esil
ien
t
b
ac
k
p
r
o
p
ag
at
io
n
th
e
wo
r
s
t
with
ac
cu
r
ac
y
9
9
.
8
7
%
an
d
9
9
.
5
3
5
%,
r
esp
ec
tiv
ely
.
As
f
o
r
th
e
SVM
m
o
d
el,
th
e
p
er
f
o
r
m
an
c
e
was
ev
alu
ated
o
n
d
if
f
er
en
t
SVM
k
er
n
els,
an
d
f
i
n
d
in
g
s
s
u
g
g
est
th
at
u
s
e
o
f
p
o
l
y
n
o
m
ial,
r
a
d
ial
b
asis
an
d
s
ig
m
o
id
k
er
n
els
p
e
r
f
o
r
m
b
etter
with
an
ac
cu
r
ac
y
o
f
9
9
.
2
4
%
an
d
a
r
e
g
r
ess
io
n
o
f
3
%
wh
ile
p
r
e
-
co
m
p
u
ted
k
e
r
n
e
l
an
d
lin
er
k
er
n
el
p
er
f
o
r
m
wo
r
s
t w
ith
an
ac
c
u
r
ac
y
an
d
r
eg
r
ess
io
n
o
f
9
5
.
7
8
% a
n
d
1
7
%,
an
d
9
5
.
1
8
% a
n
d
1
9
%
,
r
esp
ec
tiv
ely
.
L
i
et
a
l
[
3
0
]
p
r
o
p
o
s
e
a
n
o
v
e
l
n
atu
r
al
lan
g
u
a
g
e
p
r
o
ce
s
s
in
g
(
NL
P)
b
ased
m
o
d
el
to
d
etec
t
telec
o
m
f
r
au
d
.
T
h
e
s
tu
d
y
ad
o
p
ts
an
ex
t
en
s
iv
e
f
iv
e
ca
teg
o
r
y
d
ata
s
et
s
u
ch
as th
e
im
p
er
s
o
n
atio
n
o
f
cu
s
to
m
er
s
er
v
ice,
th
e
im
p
er
s
o
n
atio
n
o
f
lead
e
r
s
h
ip
ac
q
u
ain
tan
ce
s
,
lo
an
s
,
p
u
b
lic
s
ec
u
r
ity
f
r
au
d
,
an
d
n
o
r
m
al
tex
ts
.
Mo
r
eo
v
er
,
it
also
co
m
b
in
es
th
e
in
co
n
s
is
ten
cy
lo
s
s
f
u
n
ctio
n
with
cr
o
s
s
en
tr
o
p
y
f
u
n
ctio
n
,
wh
ich
cr
ea
ted
a
n
ew
lo
s
s
f
u
n
ctio
n
n
etwo
r
k
,
th
at
h
as
im
p
r
o
v
ed
f
r
au
d
d
etec
tio
n
p
e
r
f
o
r
m
an
ce
.
L
astl
y
,
a
n
ew
d
ata
m
o
d
el
f
o
r
telec
o
m
f
r
au
d
d
etec
tio
n
is
cr
ea
ted
b
ased
o
n
a
r
o
b
u
s
tly
o
p
tim
ized
b
id
ir
ec
t
io
n
al
en
co
d
e
r
r
ep
r
esen
tatio
n
s
f
r
o
m
tr
a
n
s
f
o
r
m
er
s
p
r
etr
ain
in
g
a
p
p
r
o
ac
h
(
R
o
b
E
R
T
a
)
,
wh
ich
is
im
p
r
o
v
ed
with
m
u
lti
-
h
ea
d
atten
tio
n
an
d
r
esid
u
al
co
n
n
ec
tio
n
s
(
MH
AR
C
)
.
W
h
en
co
m
p
ar
ed
with
f
iv
e
o
th
e
r
m
o
d
els
th
at
wer
e
test
ed
in
th
e
s
am
e
th
r
ee
d
ata
s
ets,
R
o
B
E
R
T
a
-
MH
A
R
C
,
s
u
g
g
ested
b
y
th
is
s
tu
d
y
,
p
e
r
f
o
r
m
ed
th
e
b
est
with
an
ac
cu
r
ac
y
o
f
9
3
.
6
9
%.
T
h
is
co
n
f
ir
m
s
th
at
th
e
u
s
e
o
f
p
r
e
-
tr
ain
ed
m
o
d
els
ca
n
p
o
s
itiv
ely
im
p
r
o
v
e
p
er
f
o
r
m
an
ce
in
m
u
lti
-
cl
ass
class
if
ica
tio
n
s
.
Mo
r
eo
v
er
,
th
e
u
s
e
o
f
m
u
lti
-
h
e
ad
atten
tio
n
en
ab
les th
e
m
o
d
el
to
f
o
cu
s
o
n
d
if
f
er
en
t p
a
r
ts
o
f
th
e
in
p
u
t,
ca
p
tu
r
i
n
g
in
f
o
r
m
atio
n
a
b
o
u
t
s
p
ec
if
ic
s
e
m
an
tics
,
m
ak
in
g
th
e
m
o
d
el
b
etter
u
n
d
er
s
tan
d
r
elatio
n
s
h
ip
s
b
etwe
en
v
ar
iab
les.
An
o
th
er
in
ter
esti
n
g
n
ew
ap
p
r
o
ac
h
is
p
r
o
v
id
ed
b
y
Z
h
ao
et
a
l
.
[
3
1
]
,
wh
ic
h
u
s
es
s
p
ee
ch
r
ec
o
g
n
itio
n
th
r
o
u
g
h
n
atu
r
al
lan
g
u
ag
e
p
r
o
g
r
am
m
in
g
to
d
etec
t
k
ey
wo
r
d
s
f
r
o
m
f
r
au
d
u
len
t
ca
lls
an
d
later
u
s
e
t
h
ese
k
ey
wo
r
d
s
as
p
ar
am
eter
s
in
th
e
m
o
d
els
th
at
will
b
e
cr
ea
ted
,
n
am
ely
lo
g
is
tic
r
eg
r
ess
io
n
,
NN,
an
d
d
ec
is
io
n
t
r
ee
.
Mo
r
e
o
v
er
,
th
e
au
th
o
r
s
o
f
th
is
p
ap
e
r
h
a
v
e
cr
ea
ted
a
m
o
b
ile
ap
p
licatio
n
th
at
p
er
f
o
r
m
s
r
ea
l
tim
e
f
r
au
d
u
len
t
d
etec
tio
n
th
r
o
u
g
h
v
o
ice
r
ec
o
g
n
itio
n
o
f
th
e
k
ey
wo
r
d
s
,
wh
ich
a
r
e
r
eg
is
ter
ed
th
r
o
u
g
h
an
o
p
tio
n
with
in
th
e
ap
p
licatio
n
,
co
n
v
er
ted
in
to
tex
t,
an
d
co
m
p
ar
ed
to
th
e
ex
is
tin
g
r
u
les
o
b
tain
ed
b
y
tr
ain
in
g
an
d
test
in
g
th
e
ML
m
o
d
els,
an
d
in
ca
s
e
th
e
ca
ll
is
d
eter
m
in
ed
as
f
r
au
d
u
len
t,
an
alar
m
will
b
e
s
en
t
to
m
o
b
ile
ap
p
licatio
n
to
n
o
tif
y
th
e
u
s
er
.
Fin
d
in
g
s
o
f
th
e
s
tu
d
y
s
u
g
g
est th
at
th
e
m
o
d
els p
er
f
o
r
m
with
a
n
ac
cu
r
ac
y
h
ig
h
e
r
th
an
9
8
.
5
3
%
.
Su
b
u
d
h
i
an
d
Pan
ig
r
ah
i
[
3
2
]
s
u
g
g
est
th
e
u
s
e
o
f
p
r
o
b
ab
ilis
tic
f
u
zz
y
C
-
m
ea
n
s
clu
s
ter
in
g
(
PF
C
M)
to
d
etec
t
telec
o
m
f
r
au
d
.
T
h
e
s
t
u
d
y
co
n
s
id
er
s
f
o
u
r
m
ain
p
a
r
am
eter
s
o
f
a
C
DR
s
u
ch
as
i
n
ter
n
atio
n
al
m
o
b
ile
eq
u
ip
m
en
t
id
en
tity
(
I
ME
I
)
,
ti
m
e
o
f
th
e
ca
ll,
d
u
r
atio
n
o
f
t
h
e
ca
ll,
an
d
ty
p
e
o
f
th
e
ca
ll.
On
ce
th
e
r
eq
u
ir
ed
p
ar
am
eter
s
ar
e
o
b
tain
ed
,
s
ev
er
al
test
s
ar
e
h
an
d
led
to
f
in
d
th
e
o
p
tim
al
n
u
m
b
er
o
f
clu
s
ter
s
n
ee
d
ed
,
wh
ic
h
r
esu
lts
in
two
clu
s
ter
s
.
Mo
r
eo
v
er
,
th
e
PF
C
M
clu
s
ter
in
g
is
r
u
n
s
ev
er
al
tim
es
with
d
if
f
e
r
en
t p
ar
am
eter
v
al
u
es
to
o
b
s
er
v
e
w
h
ich
th
r
esh
o
ld
is
r
e
q
u
ir
ed
f
o
r
ef
f
icie
n
t
f
r
a
u
d
d
ete
ctio
n
.
R
esu
lts
o
f
th
e
s
tu
d
y
s
u
g
g
est
th
at
a
t
h
r
esh
o
ld
o
f
0
.
0
0
3
e
n
s
u
r
es
an
ac
cu
r
ac
y
o
f
9
3
.
0
9
%
wh
ic
h
is
th
e
h
ig
h
est
am
o
n
g
all
th
e
co
m
b
in
ati
o
n
s
.
Mo
r
e
o
v
er
,
th
e
s
tu
d
y
also
co
m
p
a
r
es
th
e
PF
C
M
with
o
th
er
clu
s
ter
in
g
tec
h
n
i
q
u
es.
Su
ch
an
al
y
s
is
s
u
g
g
ests
t
h
at
th
e
f
o
r
m
er
is
th
e
m
o
s
t
ac
cu
r
ate
o
n
e
(
9
3
.
0
3
%)
wh
ile
m
ain
tain
in
g
th
e
h
ig
h
est
r
atio
o
f
tr
u
e
p
o
s
itiv
es
(
9
5
.
0
7
%)
an
d
th
e
lo
west
r
atio
o
f
f
alse
p
o
s
itiv
es
(
9
.
2
5
%).
Dan
iel
[
3
3
]
ad
o
p
ts
an
u
n
s
u
p
er
v
is
ed
m
o
d
el
th
at
c
o
m
b
in
es
ev
id
en
ce
f
r
o
m
m
u
ltip
le
s
o
u
r
ce
s
f
o
r
f
r
au
d
d
ete
ctio
n
,
s
u
ch
as th
e
Dem
p
s
ter
-
S
h
af
er
m
o
d
el
an
d
co
m
p
a
r
es it to
th
e
B
ay
esian
o
n
e.
T
h
e
s
tu
d
y
s
tar
ts
b
y
ch
ec
k
i
n
g
th
e
p
er
f
o
r
m
an
ce
o
f
ea
ch
m
o
d
el
s
ep
ar
ately
;
th
e
f
o
r
m
er
p
er
f
o
r
m
s
b
etter
with
an
ac
cu
r
ac
y
o
f
8
5
.
7
1
% wh
en
co
m
p
ar
ed
to
th
e
latter
with
an
ac
cu
r
ac
y
o
f
5
0
.
4
%.
Ho
wev
er
,
th
e
aim
o
f
th
e
s
tu
d
y
is
to
co
m
b
in
e
p
r
o
b
ab
ilit
ies
o
f
s
ev
er
al
in
d
iv
id
u
al
f
r
a
u
d
u
le
n
t
r
u
les
in
to
a
g
lo
b
al
b
elief
f
o
r
a
g
iv
en
h
y
p
o
th
esis
,
wh
ich
is
d
o
n
e
th
r
o
u
g
h
th
e
Dem
p
s
ter
-
Sh
af
er
m
o
d
el.
T
h
e
m
ain
tak
ea
way
o
f
th
e
s
tu
d
y
is
th
at
th
is
m
o
d
el
co
n
s
id
er
s
th
e
lev
el
o
f
u
n
ce
r
tain
ty
,
p
r
o
v
id
in
g
an
i
n
ter
v
al
in
s
t
ea
d
o
f
a
s
in
g
le
p
o
in
t
f
o
r
th
e
tr
u
e
p
r
o
b
ab
ilit
y
o
f
a
ca
s
e
b
ein
g
f
r
au
d
u
len
t.
T
h
e
r
ef
o
r
e,
th
e
m
o
d
el
ca
n
p
r
o
v
id
e
b
ett
er
ac
cu
r
ac
y
o
f
f
r
au
d
p
r
ed
ictio
n
ca
s
es.
B
o
u
k
er
c
h
e
an
d
No
tar
e
[
3
4
]
s
u
g
g
est
th
e
u
s
e
o
f
NN
with
in
a
n
in
tr
u
s
io
n
d
etec
tio
n
s
y
s
tem
to
ca
p
tu
r
e
th
e
clo
n
in
g
m
o
b
ile
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.
15
,
No
.
4
,
Au
g
u
s
t 2
0
2
6
:
3
2
6
9
-
3
2
8
5
3274
p
h
o
n
es.
W
h
ile
th
er
e
ar
e
f
ew
in
tr
u
s
io
n
d
etec
tio
n
s
y
s
tem
s
,
m
an
y
o
f
th
em
d
o
ig
n
o
r
e
th
e
p
h
o
n
e
clo
n
in
g
f
r
au
d
,
wh
ich
ca
n
h
av
e
h
u
g
e
co
n
s
eq
u
en
ce
s
if
n
o
t
tack
le
d
.
T
h
e
s
tu
d
y
s
u
g
g
ests
th
e
u
s
e
o
f
NN
to
e
n
s
u
r
e
d
etec
tio
n
o
f
an
y
ca
ll
th
at
d
o
es
n
o
t
h
av
e
a
n
o
r
m
al
cu
s
to
m
er
’
s
b
eh
av
i
o
r
a
l
p
atter
n
,
co
n
tr
o
l
th
e
im
p
o
s
ter
s
,
an
d
p
r
o
tects
th
e
r
ev
en
u
e
o
f
th
e
ca
r
r
ier
b
y
d
etec
tin
g
th
e
clo
n
ed
p
h
o
n
e
with
in
th
e
s
am
e
d
a
y
o
f
o
cc
u
r
r
en
ce
,
r
at
h
er
th
an
at
th
e
en
d
o
f
th
e
m
o
n
th
wh
e
n
th
e
b
illi
n
g
cy
cle
is
clo
s
ed
.
A
l
g
h
a
w
i
[
3
5
]
u
s
es
a
c
l
ass
i
f
i
c
ati
o
n
a
l
g
o
r
i
t
h
m
t
o
p
r
e
d
i
ct
S
I
M
b
o
x
c
a
s
es
.
S
u
c
h
c
la
s
s
i
f
ic
a
t
i
o
n
a
lg
o
r
i
t
h
m
i
s
o
b
t
a
i
n
e
d
a
s
a
l
i
n
e
a
r
c
o
m
b
i
n
at
io
n
o
f
d
e
c
i
s
i
o
n
t
r
e
e
,
f
u
n
c
t
i
o
n
a
l
t
r
e
e
,
a
n
d
RF
.
T
h
e
e
v
a
l
u
at
i
o
n
c
r
i
t
e
r
i
a
o
f
s
u
c
h
a
n
a
l
g
o
r
i
t
h
m
a
r
e
t
h
e
r
at
i
o
o
f
f
al
s
e
p
o
s
i
t
i
v
es
a
n
d
f
a
ls
e
n
e
g
a
tiv
e
s
.
R
e
s
u
lt
s
o
f
t
h
e
s
t
u
d
y
s
h
o
w
t
h
a
t
w
h
e
n
r
a
n
i
n
d
i
v
i
d
u
a
l
l
y
RF
a
l
g
o
r
i
t
h
m
h
a
d
t
h
e
l
o
w
es
t
f
a
ls
e
p
o
s
i
t
i
v
es
o
f
0
.
0
1
%
w
h
i
l
e
p
r
o
v
i
d
i
n
g
h
i
g
h
e
r
f
a
l
s
e
n
e
g
a
t
i
v
es
o
f
1
6
%
.
H
o
w
e
v
e
r
,
w
h
e
n
c
o
m
p
a
r
e
d
t
o
t
h
e
l
a
t
t
e
r
,
t
h
e
f
u
n
c
t
i
o
n
a
l
t
r
e
e
a
l
g
o
r
i
t
h
m
y
i
e
l
d
e
d
t
h
e
l
o
w
e
s
t
f
a
l
s
e
n
e
g
a
t
i
v
es
o
f
7
%
w
h
il
e
p
r
o
v
i
d
i
n
g
o
n
l
y
0
.
0
7
%
f
a
ls
e
p
o
s
i
ti
v
e
s
.
T
o
f
u
r
t
h
er
i
m
p
r
o
v
e
t
h
e
ac
c
u
r
a
c
y
o
f
t
h
e
m
o
d
e
l
,
a
m
u
l
t
i
p
l
e
r
e
g
r
e
s
s
i
o
n
t
e
c
h
n
i
q
u
e
w
a
s
a
p
p
l
i
e
d
w
h
i
c
h
c
o
n
s
i
d
e
r
e
d
t
h
e
p
r
e
d
i
c
t
i
o
n
o
u
t
p
u
t
o
f
t
h
e
t
h
r
e
e
c
l
a
s
s
i
f
i
e
r
s
a
s
p
r
e
d
i
c
t
o
r
v
a
l
u
e
s
a
n
d
i
t
s
l
i
n
ea
r
c
o
m
b
i
n
a
t
i
o
n
a
s
a
c
r
i
t
e
r
i
o
n
v
a
r
i
a
b
l
e
.
F
i
n
d
i
n
g
s
o
f
t
h
e
r
e
g
r
e
s
s
i
o
n
m
o
d
e
l
s
u
g
g
e
s
t
e
d
9
9
.
9
5
%
a
c
c
u
r
a
c
y
w
h
i
l
e
m
i
n
i
m
i
z
i
n
g
t
h
e
f
a
l
s
e
p
o
s
it
i
v
e
s
t
o
0
.
0
1
%
a
n
d
f
a
l
s
e
n
e
g
a
ti
v
e
s
t
o
9
%
.
B
h
o
w
m
ik
[
3
6
]
s
u
g
g
e
s
t
s
t
h
e
u
s
e
o
f
n
a
ï
v
e
B
a
y
e
s
i
a
n
c
l
as
s
i
f
i
er
a
s
a
r
e
l
i
a
b
l
e
m
e
t
r
i
c
w
h
e
n
d
e
te
c
t
i
n
g
f
r
a
u
d
.
T
h
e
c
l
a
s
s
i
f
i
e
r
is
in
i
t
i
a
ll
y
t
r
a
i
n
e
d
u
s
i
n
g
s
i
m
u
l
at
e
d
d
a
t
a
,
a
n
d
b
a
s
e
d
o
n
t
h
e
f
o
r
m
e
r
p
r
e
d
i
c
t
i
o
n
s
,
f
r
a
u
d
o
r
l
e
g
a
l
i
n
s
t
a
n
c
e
is
o
b
t
a
i
n
e
d
.
R
es
u
l
ts
o
f
t
h
e
s
t
u
d
y
s
u
g
g
e
s
t
t
h
at
s
u
c
h
c
la
s
s
i
f
ie
r
p
e
r
f
o
r
m
s
w
e
l
l
w
i
t
h
a
n
a
c
c
u
r
a
c
y
o
f
8
5
.
1
%
i
n
p
r
e
d
i
c
t
e
d
f
r
a
u
d
c
a
s
e
s
.
M
o
r
e
o
v
e
r
,
n
a
ï
v
e
B
a
y
e
s
a
p
p
r
o
a
c
h
is
e
as
y
a
s
i
t
c
a
n
h
a
n
d
l
e
a
n
y
m
is
s
i
n
g
v
al
u
es
b
y
o
m
i
t
ti
n
g
t
h
e
i
r
p
r
o
b
a
b
i
li
ty
a
n
d
i
s
al
s
o
t
i
m
e
e
f
f
i
c
i
e
n
t
,
as
i
t
o
n
l
y
m
u
s
t
r
u
n
o
n
c
e
.
Y
e
h
y
a
a
n
d
S
a
l
h
a
b
[
4
]
p
r
o
v
i
d
e
a
c
o
m
p
a
r
a
t
i
v
e
v
i
e
w
b
e
t
w
e
e
n
d
i
f
f
e
r
e
n
t
M
L
m
o
d
e
l
s
t
o
d
et
e
c
t
I
R
SF
f
r
a
u
d
.
T
h
e
d
a
t
a
w
as
o
b
t
ai
n
e
d
f
r
o
m
p
u
b
l
i
c
t
es
t
a
cc
o
u
n
t
s
,
a
n
d
t
h
e
f
o
ll
o
w
i
n
g
m
o
d
e
l
s
w
e
r
e
c
o
m
p
a
r
e
d
n
a
m
e
l
y
:
R
F
,
l
o
g
i
s
t
ic
r
e
g
r
e
s
s
i
o
n
,
S
VM
,
g
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
,
a
n
d
n
a
ï
v
e
B
a
y
e
s
.
A
l
l
t
h
e
c
l
a
s
s
i
f
i
e
r
s
w
e
r
e
t
r
a
i
n
e
d
a
n
d
t
es
t
e
d
u
s
i
n
g
t
h
e
s
a
m
e
d
a
t
a
a
n
d
t
h
e
r
e
s
u
lt
s
s
h
o
w
t
h
a
t
t
h
e
w
o
r
s
t
p
e
r
f
o
r
m
i
n
g
m
o
d
e
l
s
i
n
te
r
m
s
o
f
a
c
c
u
r
a
c
y
a
r
e
l
o
g
i
s
t
i
c
r
e
g
r
es
s
io
n
(
0
%
)
a
n
d
n
a
ï
v
e
B
a
y
e
s
(
6
4
%
)
w
h
i
l
e
t
h
e
o
t
h
e
r
t
h
r
e
e
d
o
p
e
r
f
o
r
m
v
e
r
y
w
e
l
l
(
1
0
0
%
)
.
A
f
t
e
r
t
h
i
s
,
t
h
e
t
h
r
e
e
cl
a
s
s
i
f
i
e
r
s
w
e
r
e
e
v
a
l
u
at
e
d
b
a
s
ed
o
n
t
h
e
c
o
n
f
u
s
i
o
n
m
a
t
r
i
x
w
h
e
r
e
S
V
M
p
e
r
f
o
r
m
e
d
p
o
o
r
l
y
i
n
t
e
r
m
s
o
f
t
r
u
e
n
e
g
a
t
i
v
e
s
a
n
d
f
a
l
s
e
p
o
s
i
ti
v
e
s
.
T
o
d
e
c
id
e
w
h
i
c
h
i
s
t
h
e
b
es
t
p
e
r
f
o
r
m
i
n
g
m
o
d
e
l
b
e
t
w
e
e
n
R
F
a
n
d
g
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
t
h
e
f
e
a
t
u
r
e
i
m
p
o
r
t
a
n
c
e
a
n
a
l
y
s
i
s
wa
s
a
p
p
l
i
e
d
.
F
i
n
d
i
n
g
s
o
f
t
h
e
s
t
u
d
y
s
h
o
w
t
h
a
t
R
F
c
a
n
p
r
o
v
i
d
e
w
i
d
e
r
i
n
f
o
r
m
a
t
i
o
n
,
w
i
t
h
t
h
e
m
i
n
u
t
e
s
as
a
p
a
r
a
m
e
te
r
b
e
i
n
g
t
h
e
t
o
p
o
n
e
,
w
h
i
c
h
i
s
c
r
u
c
i
al
i
n
I
R
S
F
d
e
t
e
c
t
i
o
n
.
T
h
e
r
e
f
o
r
e
,
t
h
e
b
es
t
p
e
r
f
o
r
m
i
n
g
m
o
d
e
l
f
o
r
t
h
i
s
f
r
a
u
d
t
y
p
e
i
s
R
F
.
E
s
tév
ez
et
a
l.
[
3
7
]
s
u
g
g
est
th
e
u
s
e
o
f
f
u
zz
y
lo
g
ic
an
d
NN
in
d
etec
tin
g
s
u
b
s
cr
ip
tio
n
f
r
au
d
.
T
h
e
f
r
au
d
ty
p
es
wer
e
d
iv
id
e
d
in
to
s
u
b
s
cr
ip
tio
n
f
r
au
d
u
len
t,
o
t
h
er
wis
e
f
r
au
d
u
len
t,
i
n
s
o
lv
en
t
an
d
n
o
r
m
al.
As
p
er
th
e
s
y
s
tem
ar
ch
itectu
r
e,
th
is
s
tu
d
y
is
b
u
ilt
u
p
o
n
o
n
e
class
if
icatio
n
m
o
d
el
wh
ich
ca
teg
o
r
izes
th
e
s
u
b
s
cr
ib
er
s
to
o
n
e
o
f
th
e
f
o
u
r
a
b
o
v
e
-
m
en
tio
n
ed
ca
teg
o
r
ies
b
ased
o
n
h
is
to
r
ical
d
ata,
an
d
o
n
e
p
r
ed
ictio
n
m
o
d
u
le
wh
ich
u
s
es
as
in
p
u
t
n
ew
in
f
o
r
m
atio
n
ab
o
u
t
s
u
b
s
cr
ib
er
s
an
d
d
etec
ts
f
r
au
d
u
len
t
u
s
er
s
.
T
h
e
latter
m
o
d
el
was
tr
ain
ed
an
d
test
ed
,
an
d
its
p
er
f
o
r
m
an
ce
was
ev
alu
ated
u
p
o
n
th
e
R
OC
an
aly
s
is
.
R
e
s
u
lts
o
f
th
e
s
t
u
d
y
s
u
g
g
est
th
at
th
e
clu
s
ter
m
o
d
el
p
er
f
o
r
m
ed
g
r
ea
t
ly
with
a
1
0
0
%
ac
c
u
r
ac
y
r
ate
wh
ile
th
e
p
r
e
d
ictio
n
o
n
e
h
ad
an
ac
cu
r
ac
y
r
ate
o
f
5
6
.
2
%.
Qay
y
u
m
et
a
l.
[
3
8
]
a
p
p
r
o
v
e
th
e
u
s
e
o
f
NN
to
d
ete
ct
s
u
b
s
cr
ip
tio
n
telec
o
m
f
r
au
d
.
Af
ter
tr
ain
in
g
,
th
e
d
ev
elo
p
e
d
NN
co
n
s
is
ts
o
f
1
4
in
p
u
t
n
e
u
r
o
n
s
,
5
h
id
d
en
la
y
er
n
eu
r
o
n
s
an
d
1
o
u
tp
u
t
lay
e
r
n
eu
r
o
n
.
Fin
d
in
g
s
o
f
th
e
s
tu
d
y
s
u
g
g
ested
th
at
th
e
m
o
d
els
p
er
f
o
r
m
e
d
with
ap
p
r
o
x
im
ately
6
0
%
ac
c
u
r
ac
y
.
Mo
r
eo
v
er
,
Xio
n
g
[
3
9
]
s
u
g
g
ests
th
e
u
s
e
o
f
tem
p
o
r
al
m
eth
o
d
to
d
etec
t
o
n
ly
f
r
au
d
u
l
en
t
ca
s
es
b
y
u
tili
zin
g
a
g
r
ap
h
-
b
ased
m
o
d
el
w
h
ile
u
s
in
g
is
o
latio
n
f
o
r
est
with
ad
d
itio
n
al
g
r
o
s
s
d
o
m
esti
c
p
r
o
d
u
ct
(
GDP
)
f
ea
tu
r
es
ap
p
lied
to
d
etec
t
I
R
SF
.
8
0
%
o
f
th
e
d
ata
h
as
b
ee
n
u
s
ed
f
o
r
tr
ai
n
in
g
an
d
2
0
%
f
o
r
test
in
g
th
e
m
o
d
els.
R
eg
ar
d
in
g
th
e
f
ir
s
t
tem
p
o
r
al
m
o
d
el,
two
d
ef
in
itio
n
s
ar
e
estab
lis
h
ed
,
n
am
ely
T
1
d
etec
ts
tim
e
in
ter
v
als
wh
er
e
f
r
au
d
co
u
ld
h
a
v
e
h
ap
p
e
n
ed
an
d
T
2
class
if
ies
if
a
s
in
g
le
r
ec
o
r
d
is
f
r
au
d
o
r
n
o
t.
Stu
d
y
s
u
g
g
ests
th
at
T
1
d
etec
ts
th
e
s
u
s
p
icio
u
s
tim
e
s
lo
t
th
r
o
u
g
h
h
is
to
r
ical
tr
af
f
ic
an
d
p
er
f
o
r
m
s
b
etter
th
an
th
e
n
aïv
e
te
m
p
o
r
al
m
o
d
els.
T
h
en
th
e
te
m
p
o
r
al
m
o
d
el
with
d
ef
in
itio
n
T
2
is
co
m
p
ar
e
d
with
th
e
is
o
latio
n
f
o
r
est
f
o
r
d
etec
t
io
n
ac
cu
r
ac
y
an
d
s
u
ch
ev
alu
at
io
n
is
d
o
n
e
th
r
o
u
g
h
th
e
R
OC
an
aly
s
is
.
Fin
d
in
g
s
o
f
th
e
s
tu
d
y
s
u
g
g
est
th
at
th
e
c
o
m
b
in
atio
n
o
f
th
ese
two
m
o
d
els
p
er
f
o
r
m
s
with
a
h
ig
h
ac
cu
r
ac
y
r
ate
o
f
9
6
.
1
%.
J
ay
asin
g
h
an
d
Swain
[
4
0
]
s
u
g
g
est
th
e
u
s
e
o
f
NN
in
d
ete
ctin
g
f
r
a
u
d
ca
s
es.
Af
ter
t
h
e
m
o
d
el
was
tr
ain
ed
,
th
e
p
e
r
f
o
r
m
an
ce
o
f
th
e
r
u
le
d
iag
n
o
s
is
was
cr
ea
ted
,
wh
ich
s
h
o
wca
s
ed
th
e
r
elati
o
n
s
h
ip
b
etwe
en
th
e
s
h
ar
e
an
d
th
e
g
en
er
aliza
tio
n
lev
el.
R
esu
lts
s
h
o
w
th
at
t
h
er
e
is
s
o
m
e
tr
a
d
eo
f
f
b
etwe
e
n
th
e
r
u
le
n
u
m
b
er
co
n
s
id
er
ed
in
t
h
e
NN
m
o
d
el
an
d
th
e
co
n
f
id
e
n
ce
lev
el.
T
h
e
m
o
r
e
r
u
les
in
clu
d
e
d
,
th
e
b
e
tter
th
e
m
o
d
el
will
p
er
f
o
r
m
,
h
o
wev
er
t
h
e
less
g
en
er
al
th
e
r
u
les
ar
e,
th
e
m
o
r
e
th
e
m
o
d
el
p
e
r
f
o
r
m
an
ce
w
ill
d
ep
en
d
o
n
t
h
e
s
tatis
t
ical
v
ar
iatio
n
s
o
f
th
e
f
r
a
u
d
d
ata.
Fin
d
in
g
s
o
f
th
e
s
tu
d
y
s
u
g
g
est
th
at
t
h
is
m
o
d
el
p
er
f
o
r
m
s
with
7
5
.
1
7
%
ac
cu
r
ac
y
r
ate
af
ter
d
ec
r
ea
s
in
g
th
e
n
u
m
b
er
o
f
r
u
les
f
r
o
m
7
4
7
to
5
1
0
.
Mo
r
eo
v
er
,
Ab
id
o
g
u
n
[
4
1
]
p
r
o
p
o
s
es
th
e
u
s
e
o
f
SOM
a
n
d
L
STM
r
ec
u
r
r
en
t
NN
m
o
d
els
to
p
r
ed
ict
th
e
ca
ll
p
atter
n
b
eh
av
io
r
s
in
a
n
u
n
s
u
p
er
v
is
ed
lear
n
i
n
g
ap
p
r
o
ac
h
.
T
h
e
L
STM
R
NN
m
o
d
els
ar
e
s
u
p
e
r
io
r
t
o
th
e
t
r
a
d
itio
n
al
R
NN,
as
th
ey
o
v
er
co
m
e
th
e
d
e
p
en
d
e
n
cy
is
s
u
e.
On
th
e
o
th
er
s
id
e,
SO
M
ar
e
NN
m
o
d
els
th
at
en
s
u
r
e
v
is
u
aliza
tio
n
o
f
h
ig
h
d
im
e
n
s
io
n
al
d
ata
w
h
ile
m
ain
tain
in
g
th
e
im
p
o
r
tan
t
m
e
tr
ic
r
elatio
n
s
h
ip
s
o
f
th
e
p
r
im
a
r
y
d
ata
elem
en
ts
.
Fin
d
in
g
s
o
f
th
e
s
tu
d
y
s
u
g
g
est
th
at
wh
ile
u
s
in
g
th
e
SOM
m
o
d
el
,
i
n
f
o
r
m
atio
n
f
o
r
th
e
t
em
p
o
r
al
n
atu
r
e
o
f
ca
lls
was
lo
s
t
in
th
e
p
r
o
ce
s
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
S
ystema
tic
r
ev
iew
o
f fra
u
d
d
etec
tio
n
u
s
in
g
A
I
a
n
d
ML
w
ith
a
n
emp
h
a
s
is
o
n
…
(
S
o
ly
Ma
th
e
w
B
iju
)
3275
Mo
r
eo
v
er
,
d
if
f
e
r
en
t
m
ap
s
izes
f
o
r
d
if
f
e
r
en
t
s
u
b
s
cr
ib
er
s
ad
d
a
lay
er
o
f
co
m
p
lex
ity
wh
en
it
co
m
es
to
co
m
p
ar
in
g
th
em
,
h
o
wev
er
,
o
v
er
all,
th
e
m
o
d
el
ca
n
b
e
u
s
ed
t
o
p
r
o
v
id
e
in
s
ig
h
ts
f
o
r
th
e
v
ar
i
ab
les
d
ep
en
d
en
cies.
On
th
e
o
th
er
h
a
n
d
,
th
e
L
ST
M
R
NN
wa
s
tr
ain
ed
1
0
tim
e
s
u
s
in
g
th
e
n
o
n
p
ar
am
etr
ic
en
tr
o
p
y
o
p
tim
izatio
n
(
NE
O
)
alg
o
r
ith
m
,
u
n
til
th
er
e
was
n
o
f
u
r
th
er
i
m
p
r
o
v
em
en
t
i
n
th
e
r
esu
lt,
with
an
ac
cu
r
ac
y
o
f
9
0
%.
T
h
er
ef
o
r
e,
th
e
s
tu
d
y
s
u
g
g
ests
th
at
L
ST
M
is
a
b
etter
clu
s
ter
in
g
o
p
tio
n
th
an
SOM.
Djo
m
ad
ji
et
a
l.
[
4
2
]
s
u
g
g
est
th
e
u
s
e
o
f
R
F,
SVM
,
an
d
e
x
tr
em
e
g
r
ad
ie
n
t
b
o
o
s
tin
g
(
XG
B
o
o
s
t
)
to
d
etec
t
SIM
b
o
x
ca
s
es.
All th
e
m
o
d
els we
r
e
tr
ain
ed
a
n
d
test
ed
u
s
in
g
th
e
s
am
e
d
atab
ase,
an
d
th
e
ev
alu
atio
n
was
d
o
n
e
b
y
th
e
co
n
f
u
s
io
n
m
atr
ix
.
Fin
d
in
g
s
s
u
g
g
est
th
at
th
e
SVM
m
o
d
el
p
er
f
o
r
m
ed
p
o
o
r
ly
with
6
9
%
ac
cu
r
ac
y
an
d
th
e
p
e
r
f
o
r
m
an
ce
d
id
n
o
t
i
m
p
r
o
v
e
af
ter
tr
ain
in
g
it.
Ho
w
ev
er
,
t
h
e
R
F
p
er
f
o
r
m
e
d
with
9
2
%
ac
cu
r
ac
y
a
f
ter
b
ein
g
t
r
ain
ed
ag
ain
wh
ile
t
h
e
XGBo
o
s
t
m
o
d
el
p
e
r
f
o
r
m
ed
with
8
1
%
ac
cu
r
ac
y
.
T
h
er
ef
o
r
e,
a
m
o
n
g
th
e
th
r
ee
ML
m
o
d
els,
th
e
R
F
is
th
e
b
est
p
er
f
o
r
m
in
g
o
n
e.
Mo
r
eo
v
e
r
,
Olu
s
h
o
la
an
d
Ma
r
t
[
4
3
]
co
n
s
id
er
t
h
e
u
s
e
o
f
th
r
ee
ML
m
o
d
els to
d
etec
t f
r
au
d
ca
s
es,
s
u
ch
as R
F,
XGBo
o
s
t
,
an
d
lo
g
is
tic
r
eg
r
ess
io
n
.
Fin
d
in
g
s
o
f
th
e
s
tu
d
y
s
u
g
g
est th
at
d
if
f
er
en
t
m
o
d
els
o
u
tp
er
f
o
r
m
t
h
e
o
th
er
s
in
d
if
f
er
e
n
t
ar
ea
s
.
Fo
r
in
s
tan
ce
,
co
m
b
in
in
g
R
F
an
d
XGBo
o
s
t
p
r
o
v
id
es
a
m
o
d
el
with
b
etter
ac
cu
r
ac
y
.
Ho
wev
er
,
lo
g
is
tic
r
eg
r
ess
io
n
o
n
its
o
wn
is
m
o
r
e
ef
f
ec
tiv
e
at
id
en
tify
in
g
ce
r
tain
f
r
au
d
ty
p
es.
Mo
r
eo
v
e
r
,
f
ea
t
u
r
e
en
g
in
ee
r
in
g
is
cr
u
cial
to
im
p
r
o
v
e
th
e
m
o
d
el’
s
p
er
f
o
r
m
an
ce
s
in
ce
th
e
b
eg
in
n
in
g
;
s
u
ch
a
th
in
g
r
eq
u
i
r
es
to
h
av
e
a
d
ee
p
b
u
s
in
ess
u
n
d
er
s
tan
d
in
g
t
o
co
n
v
er
t
th
e
f
r
a
u
d
ca
s
es
in
to
u
s
ef
u
l
f
ea
tu
r
es
f
o
r
ML
m
o
d
els.
L
astl
y
,
co
n
s
id
er
in
g
th
e
n
atu
r
e
o
f
f
r
au
d
,
it
is
a
m
u
s
t
to
en
s
u
r
e
r
e
al
tim
e
m
o
n
ito
r
in
g
an
d
p
r
e
v
en
tio
n
o
f
f
r
au
d
ca
s
e
s
.
Sah
in
[
4
4
]
c
o
n
d
u
cts
a
s
tu
d
y
u
s
in
g
test
ca
lls
to
d
etec
t
o
v
er
-
th
e
-
to
p
(
OT
T
)
b
y
p
ass
f
r
au
d
.
T
wo
s
ep
ar
ate
t
est
ca
ll
ca
m
p
aig
n
s
wer
e
h
eld
wh
er
e
t
h
e
f
ir
s
t
u
tili
ze
s
a
co
m
m
er
cial
test
ca
ll
g
en
er
atio
n
(
T
C
G)
p
latf
o
r
m
with
b
ig
n
etwo
r
k
co
v
e
r
ag
e
wh
i
le
th
e
s
ec
o
n
d
u
s
e
s
a
s
m
aller
d
ed
icate
d
ca
ll
test
p
latf
o
r
m
b
u
ilt
f
o
r
t
h
is
p
u
r
p
o
s
e.
I
n
th
e
f
o
r
m
er
,
1
0
1
6
test
ca
lls
wer
e
m
ad
e
us
in
g
T
C
G
p
latf
o
r
m
f
r
o
m
d
if
f
er
en
t
n
etwo
r
k
o
p
e
r
ato
r
s
in
5
0
co
u
n
tr
ies.
OT
T
f
r
au
d
ca
s
es
wer
e
d
etec
ted
in
9
0
%
o
f
th
e
co
u
n
tr
ies
an
d
6
2
%
o
f
o
p
er
ato
r
s
,
wh
ile
4
0
% o
f
th
e
te
s
t c
alls
wer
e
b
y
p
ass
ed
,
wh
ich
m
ea
n
s
th
at
OT
T
b
y
p
ass
is
h
ig
h
to
war
d
s
th
is
T
C
G
p
latf
o
r
m
.
Fin
d
in
g
s
o
f
th
e
s
tu
d
y
s
u
g
g
est
th
at
with
9
5
%
co
n
f
id
en
ce
,
th
e
OT
T
b
y
p
ass
r
ate
d
o
es
n
o
t
ch
a
n
g
e
r
eg
ar
d
less
o
f
w
h
eth
er
t
h
e
u
s
er
is
m
ak
in
g
d
o
m
esti
c
(
3
6
%)
o
r
r
o
am
in
g
ca
lls
(
3
7
%).
Mo
r
e
o
v
er
,
th
e
s
tu
d
y
u
s
es
th
e
in
teg
r
ate
d
s
er
v
ices
d
i
g
ital
n
etwo
r
k
u
s
er
p
a
r
t
(
I
SUP)
lo
g
s
f
r
o
m
th
e
m
o
b
ile
o
p
er
ato
r
to
c
r
ea
te
a
p
ar
tial
m
ap
o
f
r
o
u
tes
s
tar
tin
g
f
r
o
m
7
m
o
b
ile
o
p
er
ato
r
s
in
Un
ited
Kin
g
d
o
m
wh
ich
h
elp
s
in
id
en
tif
y
in
g
th
e
b
y
p
ass
in
g
o
p
er
ato
r
s
,
h
o
wev
er
,
co
n
s
id
er
i
n
g
th
e
co
s
ts
o
f
T
C
G
p
latf
o
r
m
s
u
ch
an
aly
s
is
was
n
o
t
co
n
d
u
cted
.
Ho
wev
er
,
th
e
o
th
er
test
ca
m
p
aig
n
is
b
u
ilt
o
n
An
d
r
o
id
p
h
o
n
es
m
o
n
ito
r
e
d
v
ia
W
i
-
Fi
co
n
n
ec
tio
n
.
Nev
er
th
e
less
,
co
llectin
g
in
f
o
r
m
atio
n
f
r
o
m
b
o
t
h
th
e
ca
l
lin
g
an
d
ca
lled
p
ar
t
y
.
E
a
ch
p
h
o
n
e
was
p
lace
d
in
o
n
e
E
u
r
o
p
ea
n
c
o
u
n
tr
y
wh
e
r
e
th
e
o
r
ig
in
atin
g
o
p
er
ato
r
wo
u
l
d
b
e
a
b
ig
s
ized
m
o
b
ile
n
etwo
r
k
o
p
er
at
o
r
(
MN
O)
.
Fo
r
ea
ch
ca
ll,
th
e
p
ar
am
eter
s
o
b
tain
ed
ar
e
n
etwo
r
k
n
am
e,
ca
ll
s
tar
t
tim
e,
ca
ll
r
in
g
tim
e,
an
d
ca
ll
en
d
tim
e.
I
n
c
o
m
i
n
g
ca
ll
lo
g
s
f
r
o
m
An
d
r
o
id
’
s
d
atab
ase
an
d
OT
T
ar
e
co
llected
wh
ich
p
r
o
v
id
e
p
ar
am
eter
s
s
u
ch
as
in
co
m
in
g
c
all
tim
e,
ca
ll
ty
p
e,
ca
lled
n
u
m
b
er
I
D.
Fin
d
in
g
s
o
f
th
e
s
tu
d
y
s
u
g
g
est
th
at
wh
ile
g
en
er
atin
g
ca
lls
f
o
r
2
m
o
n
th
s
ev
er
y
d
ay
o
n
an
h
o
u
r
ly
b
asis
,
6
o
u
t
o
f
8
co
u
n
tr
ies o
b
s
er
v
ed
OT
T
b
y
p
ass
with
r
ates f
r
o
m
4
2
% to
8
3
%.
Kilin
c
[
4
5
]
s
u
g
g
ests
th
e
u
s
e
o
f
ML
class
if
ier
s
s
u
ch
as
d
ec
is
i
o
n
tr
ee
,
Gau
s
s
ian
n
aïv
e
B
ay
es,
ad
ap
tiv
e
b
o
o
s
tin
g
(
Ad
aBo
o
s
t
)
to
d
etec
t
telec
o
m
f
r
au
d
ca
s
es.
I
n
itially
,
th
e
d
ata
was la
b
eled
u
s
in
g
th
e
lo
ca
l o
u
tlier
f
ac
to
r
m
o
d
el
,
ac
co
r
d
in
g
to
wh
ich
6
.
7
%
o
f
th
e
ca
lls
wer
e
co
n
s
id
er
ed
as
f
r
au
d
u
len
t
u
s
er
s
.
Su
ch
f
r
au
d
u
len
t
d
ata
was
u
s
ed
to
tr
ain
all
th
r
ee
ML
cla
s
s
if
ier
s
.
L
astl
y
,
a
m
ix
ed
m
o
d
el
was
cr
ea
ted
b
y
u
s
in
g
th
e
a
v
er
ag
e
o
f
th
e
th
r
ee
in
d
iv
id
u
al
s
u
p
er
v
is
ed
m
o
d
els.
T
h
e
latter
m
o
d
el
was
tr
ain
e
d
b
y
7
0
%
o
f
th
e
d
ataset
u
s
ed
.
Fin
d
in
g
s
o
f
th
e
s
tu
d
y
s
u
g
g
est
th
at
th
e
b
est
p
e
r
f
o
r
m
i
n
g
m
o
d
el
is
d
ec
is
io
n
tr
ee
with
9
3
%
ac
cu
r
ac
y
r
ate.
Z
h
u
et
a
l.
[
4
6
]
ad
o
p
t
s
ev
er
al
ML
class
if
ier
s
to
d
ev
elo
p
a
s
ca
m
m
ess
ag
e
aler
t.
Data
s
et
wa
s
o
b
tain
ed
f
r
o
m
an
o
p
en
-
s
o
u
r
ce
m
ess
ag
e
d
atase
t
f
r
o
m
GitHu
b
,
o
u
t
o
f
w
h
ich
7
0
%
was
u
s
ed
f
o
r
tr
ai
n
in
g
t
h
e
M
L
m
o
d
es
an
d
th
e
r
em
ain
in
g
3
0
%
f
o
r
test
in
g
.
T
h
e
ch
o
s
en
m
o
d
els
wer
e
d
ec
is
io
n
tr
ee
,
SVM,
lo
g
is
tic
r
eg
r
ess
io
n
,
a
n
d
n
aïv
e
B
ay
es
wh
ich
p
er
f
o
r
m
ed
with
an
ac
cu
r
ac
y
r
ate
o
f
9
8
.
7
2
%,
9
4
.
2
3
%,
9
1
.
0
0
%
,
an
d
9
6
.
2
8
%,
r
es
p
ec
tiv
ely
in
d
etec
tin
g
f
r
au
d
u
len
t m
ess
ag
es.
Ou
t o
f
all
th
ese
m
o
d
els,
th
e
d
ec
is
io
n
tr
e
e
o
n
e
p
er
f
o
r
m
ed
th
e
b
est;
h
o
wev
er
,
its
ac
cu
r
ac
y
in
cr
ea
s
es
with
th
e
in
cr
ea
s
e
o
f
n
_
esti
m
ato
r
s
an
d
th
en
s
tar
ts
to
f
all.
T
h
er
ef
o
r
e,
th
e
n
_
esti
m
ato
r
f
o
r
wh
ich
th
e
m
o
d
el
o
b
tain
ed
th
e
h
ig
h
est
ac
cu
r
ac
y
was
4
0
0
an
d
th
is
was
th
e
ch
o
s
en
n
u
m
b
er
f
o
r
th
e
f
in
al
m
o
d
el.
Mo
r
e
o
v
er
,
E
za
wa
et
a
l.
[
4
7
]
s
u
g
g
es
t
a
g
o
al
-
o
r
ie
n
ted
alg
o
r
ith
m
f
o
r
cr
ea
tin
g
B
ay
esian
n
etwo
r
k
i
n
d
e
tectin
g
u
n
c
o
llectib
le
in
telec
o
m
r
is
k
m
a
n
ag
em
e
n
t
d
ata.
T
h
e
s
tu
d
y
u
s
es
d
if
f
e
r
e
n
t
class
if
ier
s
o
f
B
ay
esian
m
o
d
el,
s
u
ch
as
a
d
v
an
ce
d
p
atte
r
n
r
ec
o
g
n
itio
n
an
d
id
en
tific
atio
n
(
APR
I
)
,
g
o
al
-
o
r
ien
ted
K2
,
in
d
ep
en
d
en
t
a
n
d
g
o
al
-
o
r
ien
ted
c
o
n
d
itio
n
al
in
d
ep
en
d
en
ce
an
d
B
ay
esian
lear
n
in
g
(
CB
)
.
Su
ch
m
o
d
els
ar
e
co
m
p
ar
ed
an
d
ev
alu
ated
ag
ain
s
t
ea
ch
o
th
e
r
b
a
s
ed
u
p
o
n
th
e
R
OC
an
aly
s
is
.
Su
ch
an
aly
s
is
s
u
g
g
ests
th
at
AP
R
I
,
th
e
g
o
al
-
o
r
ien
ted
m
o
d
el,
h
as
a
way
b
etter
p
er
f
o
r
m
an
ce
wh
en
co
m
p
ar
ed
with
all
th
e
o
th
e
r
m
o
d
els,
f
o
llo
wed
b
y
th
e
i
n
d
ep
e
n
d
en
t
m
o
d
el
a
n
d
t
h
en
K
2
an
d
C
B
wh
ich
p
er
f
o
r
m
r
elativ
ely
p
o
o
r
ly
.
Ho
wev
e
r
,
w
h
en
b
u
ild
in
g
th
e
g
o
al
-
o
r
ien
te
d
K2
an
d
g
o
al
-
o
r
ien
ted
C
B
an
d
c
o
m
p
ar
in
g
t
h
em
with
AP
R
I
m
o
d
el,
th
e
R
OC
a
n
aly
s
is
s
u
g
g
ests
th
at
ev
en
th
o
u
g
h
th
e
latter
m
o
d
el
s
till
h
as t
h
e
b
est p
er
f
o
r
m
an
ce
,
th
e
two
o
th
e
r
m
o
d
els’
p
e
r
f
o
r
m
an
ce
h
as
im
p
r
o
v
e
d
s
ig
n
if
ic
an
tly
wh
en
co
m
p
a
r
ed
t
o
b
e
f
o
r
e.
T
h
er
ef
o
r
e,
th
is
s
tu
d
y
s
u
g
g
ests
th
at
wh
en
cr
ea
tin
g
a
B
ay
esian
m
o
d
el,
it
m
u
s
t
b
e
ad
ju
s
ted
as
p
er
th
e
s
p
ec
if
ic
g
o
al
to
ac
h
iev
e
b
etter
p
er
f
o
r
m
an
ce
.
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.
15
,
No
.
4
,
Au
g
u
s
t 2
0
2
6
:
3
2
6
9
-
3
2
8
5
3276
A
s
o
g
w
a
e
t
a
l
.
[
4
8
]
p
r
o
v
i
d
e
a
t
h
e
o
r
et
i
c
a
l
f
r
a
m
e
w
o
r
k
o
f
NN
a
p
p
l
i
c
at
i
o
n
f
o
r
d
e
t
e
c
t
i
o
n
o
f
t
e
l
e
c
o
m
m
u
n
i
c
at
i
o
n
f
r
a
u
d
.
T
h
e
s
t
u
d
y
t
o
u
c
h
e
s
u
p
o
n
t
h
e
d
e
f
i
n
i
t
io
n
o
f
t
e
l
e
c
o
m
f
r
a
u
d
,
i
ts
c
o
n
s
e
q
u
e
n
c
e
s
,
f
r
a
u
d
t
y
p
es
a
n
d
d
e
t
e
c
t
i
o
n
t
ec
h
n
i
q
u
e
s
.
M
o
r
e
o
v
e
r
,
t
h
e
s
t
u
d
y
f
o
c
u
s
es
o
n
t
h
e
u
s
e
o
f
a
r
t
i
f
i
ci
a
l
NN
s
o
n
d
et
e
c
t
i
n
g
t
e
l
e
c
o
m
m
u
n
i
c
at
i
o
n
f
r
a
u
d
f
r
o
m
t
h
e
e
a
r
l
y
1
9
6
0
s
u
p
t
o
d
a
t
e
,
b
y
e
x
p
l
a
i
n
i
n
g
h
o
w
t
o
u
s
e
t
h
e
d
ata
t
o
t
r
a
i
n
a
n
d
t
es
t
a
NN
t
o
e
f
f
ic
i
e
n
tl
y
d
e
t
e
c
t f
r
a
u
d
u
l
e
n
t
a
n
d
n
o
n
-
f
r
a
u
d
u
l
e
n
t
u
s
e
r
s
.
S
u
c
h
a
r
t
i
cl
e
a
n
a
l
y
z
e
s
t
h
e
c
as
e
o
f
a
NN
-
b
a
s
e
d
f
r
a
u
d
i
m
p
l
e
m
e
n
t
a
ti
o
n
t
o
o
l
,
M
i
n
o
t
u
a
r
,
w
h
i
c
h
w
a
s
a
m
i
x
o
f
AI
m
o
d
el
s
a
n
d
t
r
a
d
i
ti
o
n
a
l
c
o
m
p
u
t
i
n
g
t
ec
h
n
i
q
u
e
s
.
F
i
n
d
i
n
g
s
o
f
t
h
e
s
t
u
d
y
s
u
g
g
e
s
t
t
h
a
t
o
n
l
y
in
t
h
r
e
e
m
o
n
t
h
s
o
f
i
m
p
l
e
m
e
n
ta
t
io
n
t
h
e
f
r
a
u
d
l
o
s
s
p
e
r
u
n
i
t
w
a
s
r
e
d
u
c
e
d
b
y
4
0
%
a
n
d
t
h
a
t
c
o
m
b
i
n
a
t
i
o
n
o
f
n
e
u
r
a
l
,
r
u
l
e
b
a
s
e
d
,
a
n
d
c
a
s
e
-
b
as
e
d
t
ec
h
n
o
l
o
g
i
e
s
a
r
e
f
a
r
b
e
t
t
e
r
i
n
f
r
a
u
d
d
e
t
e
c
t
i
o
n
t
h
a
n
t
r
a
d
i
t
i
o
n
a
l
s
y
s
t
e
m
s
.
O
n
u
o
d
u
a
n
d
N
n
a
a
[
4
9
]
p
r
e
s
e
n
t
t
h
e
u
s
e
o
f
f
o
r
w
a
r
d
a
n
d
b
a
c
k
w
a
r
d
r
u
l
e
-
b
a
s
e
d
A
N
N
t
o
i
m
p
r
o
v
e
a
n
e
x
is
t
i
n
g
p
la
t
f
o
r
m
f
o
r
s
m
a
r
t
c
a
r
d
a
n
d
t
e
l
e
c
o
m
m
u
n
i
c
a
t
i
o
n
f
r
a
u
d
d
et
e
c
ti
o
n
i
n
N
i
g
er
i
a
.
T
h
r
o
u
g
h
A
N
N
,
i
m
p
r
o
v
e
m
e
n
t
o
f
a
l
l
t
h
e
c
o
m
p
o
n
e
n
t
s
i
n
t
h
e
e
x
is
t
i
n
g
s
y
s
t
e
m
wa
s
m
a
d
e
i
n
c
l
u
d
i
n
g
t
h
e
f
o
l
l
o
w
in
g
:
i
n
t
e
l
li
g
e
n
t
a
g
e
n
t
,
s
e
c
u
r
i
t
y
m
e
a
s
u
r
es
,
c
u
s
t
o
m
e
r
,
c
u
s
t
o
m
e
r
b
a
n
k
d
a
ta
b
a
s
e
,
c
r
e
d
i
t
c
a
r
d
t
r
a
n
s
ac
t
i
o
n
d
a
ta
b
a
s
e
,
an
d
f
r
a
u
d
d
e
t
e
c
t
i
o
n
d
a
t
a
b
a
s
e
.
M
o
r
e
o
v
e
r
,
u
s
i
n
g
t
h
e
f
o
r
w
a
r
d
a
n
d
b
a
c
k
w
a
r
d
r
u
l
e
-
b
a
s
e
d
A
N
N
,
t
h
e
p
r
o
p
o
s
e
d
s
y
s
te
m
w
a
s
a
b
l
e
t
o
n
o
t
o
n
l
y
d
e
t
e
c
t
i
n
r
e
a
l
t
i
m
e
b
u
t
a
l
s
o
p
r
e
d
i
c
t
i
n
a
d
v
a
n
c
e
a
n
y
f
r
a
u
d
u
l
e
n
t
u
s
e
r
s
o
n
t
h
e
s
m
a
r
t
c
a
r
d
an
d
t
e
l
e
c
o
m
s
e
r
v
i
c
e
s
,
b
l
o
c
k
i
n
g
t
h
e
m
f
r
o
m
u
s
i
n
g
t
h
e
s
e
r
v
i
c
e
.
A
c
o
m
p
a
r
a
t
i
v
e
a
n
a
l
y
s
is
b
e
t
w
e
e
n
t
h
e
o
l
d
a
n
d
n
e
w
l
y
p
r
o
p
o
s
e
d
s
y
s
t
e
m
w
as
c
o
n
d
u
c
t
e
d
,
w
h
e
r
e
t
h
e
e
v
al
u
a
t
io
n
c
r
i
t
e
r
i
a
w
e
r
e
s
p
e
e
d
i
n
p
r
o
ce
s
s
i
n
g
i
n
p
u
t
v
a
l
i
d
at
i
o
n
d
e
t
ai
ls
(
S
P
I
VD
)
,
s
p
e
e
d
i
n
d
e
t
e
c
t
i
n
g
a
n
d
b
l
o
c
k
i
n
g
f
r
a
u
d
u
l
e
n
t
a
t
t
e
m
p
ts
(
S
DB
F
A
)
,
c
r
o
s
s
p
l
a
t
f
o
r
m
c
o
m
p
a
t
i
b
i
l
i
t
y
(
C
P
A)
,
m
o
d
e
l
e
f
f
i
c
i
e
n
c
y
(
M
E
)
,
a
n
d
c
o
s
t
b
e
n
e
f
i
t
a
n
a
l
y
s
i
s
(
C
B
A
)
.
R
e
s
u
l
t
s
o
f
s
u
c
h
a
n
a
l
y
s
is
c
o
n
v
e
y
t
h
a
t
t
h
e
n
e
w
l
y
p
r
o
p
o
s
e
d
m
o
d
e
l
i
s
s
u
p
e
r
i
o
r
i
n
a
l
l
t
h
e
c
r
i
te
r
i
a
a
n
d
p
e
r
f
o
r
m
i
n
g
w
i
t
h
9
4
%
a
c
c
u
r
a
c
y
r
a
t
e
t
o
w
a
r
d
s
6
5
%
o
f
t
h
e
o
l
d
m
o
d
e
l
.
B
h
av
an
i
et
a
l.
[
5
0
]
c
o
m
b
in
e
th
e
u
s
e
o
f
n
aïv
e
B
ay
es
an
d
L
S
T
M
R
NN
to
p
r
o
v
id
e
a
n
ad
v
a
n
ce
d
f
r
a
u
d
p
r
ed
ictio
n
m
o
d
el
f
o
r
m
o
b
ile
f
r
au
d
.
T
h
e
f
o
r
m
er
m
o
d
el
was
tr
ain
ed
to
o
b
tain
s
o
m
e
b
en
ch
m
a
r
k
f
o
r
p
er
f
o
r
m
a
n
ce
ev
alu
atio
n
,
b
ased
o
n
wh
ich
th
e
s
tu
d
y
a
p
p
lied
m
o
r
e
a
d
v
an
ce
d
m
o
d
els
s
u
ch
as
L
STM
.
T
h
e
latter
m
o
d
el
was
d
esig
n
ed
,
tr
ain
e
d
an
d
ad
ju
s
ted
to
d
etec
t
co
m
p
lex
b
eh
av
io
r
p
atter
n
s
o
f
f
r
a
u
d
u
le
n
t
ca
s
es
in
ca
ll
m
ess
ag
es.
T
h
e
s
tu
d
y
co
n
d
u
cts
a
co
m
p
ar
is
o
n
o
f
b
o
th
m
o
d
els
b
ased
o
n
ac
cu
r
ac
y
,
p
r
ec
is
io
n
,
r
ec
all
,
an
d
F1
-
s
co
r
e,
h
ig
h
lig
h
tin
g
th
e
s
tr
en
g
th
s
an
d
wea
k
n
ess
es
o
f
ea
ch
.
L
astl
y
,
L
STM
was
tr
ain
ed
1
0
tim
es
o
n
a
s
ep
ar
at
e
d
ata
s
et
f
r
o
m
th
e
in
itial o
n
e,
p
r
o
v
in
g
its
ab
ilit
y
t
o
lear
n
an
d
g
en
e
r
alize
n
ew
d
at
a
an
d
co
m
p
ar
e
it a
g
ain
s
t th
e
a
ctu
al
lab
eled
ca
s
es.
Fin
d
in
g
s
o
f
th
e
s
tu
d
y
s
u
g
g
es
t
th
at
th
is
m
o
d
el
p
er
f
o
r
m
s
with
9
9
.
8
9
%
ac
cu
r
ac
y
in
th
e
t
r
ain
in
g
s
am
p
le
an
d
9
7
.
2
0
%
ac
cu
r
ac
y
in
th
e
test
in
g
o
n
e.
Hu
et
a
l.
[
5
1
]
s
u
g
g
e
s
t
an
in
ter
esti
n
g
u
s
e
o
f
a
u
g
m
en
ted
g
r
ap
h
n
e
u
r
al
n
etwo
r
k
(
GNN)
b
y
m
ix
in
g
r
e
in
f
o
r
ce
m
e
n
t
an
d
en
s
em
b
le
lea
r
n
in
g
t
o
d
etec
t
m
o
b
ile
f
r
au
d
,
wh
ile
tack
lin
g
two
m
ain
p
r
o
b
lem
s
ass
o
ciate
d
with
it
s
u
ch
as
g
r
ap
h
im
b
ala
n
ce
an
d
o
v
er
-
s
m
o
o
t
h
in
g
.
T
o
tack
l
e
th
e
f
o
r
m
er
is
s
u
e,
r
ein
f
o
r
ce
m
e
n
t
lear
n
in
g
b
ased
-
n
eig
h
b
o
r
s
am
p
ler
is
u
s
ed
wh
il
e
f
o
r
th
e
latter
,
Ad
aBo
o
s
t
f
o
r
en
s
em
b
le
lear
n
in
g
o
f
th
e
GNN
m
o
d
els
is
ap
p
lied
.
All
th
e
tr
ain
in
g
o
f
th
e
m
o
d
el
is
s
u
p
er
v
is
ed
u
tili
zin
g
a
class
b
alan
ce
d
f
o
ca
l
lo
s
s
f
u
n
ctio
n
an
d
th
en
it
is
test
ed
i
n
two
r
e
al
wo
r
ld
telec
o
m
f
r
au
d
d
atasets
to
o
b
s
er
v
e
its
p
e
r
f
o
r
m
an
ce
.
Fin
d
in
g
s
o
f
th
e
s
tu
d
y
s
u
g
g
est
th
at
th
e
au
g
m
en
ted
GNN
m
o
d
el
o
u
tp
er
f
o
r
m
s
o
th
er
ex
is
tin
g
g
r
a
p
h
-
b
a
s
ed
m
o
d
els
in
b
o
th
d
atasets
wh
ile
attain
in
g
a
n
ac
cu
r
ac
y
r
ate
o
f
9
1
.
5
5
%
a
n
d
9
2
.
5
%,
r
esp
ec
tiv
ely
.
Ma
wg
o
u
d
et
a
l.
[
5
2
]
s
u
g
g
est
a
co
m
b
in
atio
n
o
f
class
if
icatio
n
m
eth
o
d
s
an
d
NN
to
d
etec
t
W
an
g
ir
i,
a
s
p
ec
if
ic
telec
o
m
m
u
n
ic
atio
n
s
f
r
au
d
.
Fr
o
m
th
e
d
ataset
u
s
ed
,
7
0
%
ar
e
n
o
r
m
al
u
s
er
s
an
d
3
0
%
f
r
au
d
u
len
t
o
n
es.
I
n
th
is
p
ap
er
,
ANN,
with
th
r
ee
lay
e
r
s
was
ap
p
lied
an
d
r
esu
lts
o
f
th
e
m
o
d
el
wer
e
ev
al
u
ated
u
s
in
g
th
e
R
OC
m
ap
.
Fin
d
in
g
s
o
f
th
e
s
t
u
d
y
s
u
g
g
est
th
at
th
is
m
o
d
el
p
er
f
o
r
m
s
r
elativ
el
y
well
with
a
9
0
.
6
%
ac
cu
r
ac
y
r
ate
o
n
tr
ain
in
g
an
d
8
5
%
o
n
test
in
g
.
L
astl
y
,
W
ah
id
et
a
l.
[
5
3
]
p
r
esen
t
a
n
ew
telec
o
m
f
r
a
u
d
d
etec
tio
n
ap
p
r
o
ac
h
b
y
u
s
in
g
a
n
eu
r
al
f
ac
to
r
izatio
n
au
to
en
co
d
er
,
wh
ich
its
elf
co
m
b
in
es
n
eu
r
al
f
ac
t
o
r
izatio
n
m
ac
h
in
e
(
NFM)
an
d
au
to
e
n
co
d
e
r
s
.
Su
ch
co
m
p
o
n
e
n
ts
en
ab
le
th
e
m
o
d
el
to
r
eg
is
ter
u
s
er
b
eh
a
v
io
r
al
p
atter
n
s
an
d
l
ea
r
n
f
r
o
m
t
h
e
n
ew
r
aw
d
ata,
wh
ich
co
n
tin
u
o
u
s
ly
im
p
r
o
v
e
d
th
e
d
etec
tio
n
o
f
th
e
m
o
d
el.
T
h
e
t
r
ain
ed
m
o
d
el
was
test
ed
in
a
r
ea
l
-
wo
r
ld
d
ata
s
et
en
co
m
p
ass
in
g
a
h
u
g
e
a
m
o
u
n
t
o
f
d
ata.
Mo
r
e
o
v
er
,
th
e
s
tu
d
y
s
u
g
g
ests
th
at
m
em
o
r
y
len
g
th
a
n
d
th
r
esh
o
ld
d
o
im
p
ac
t
th
e
m
o
d
el’
s
p
e
r
f
o
r
m
an
ce
.
I
n
o
th
er
wo
r
d
s
,
t
h
e
b
est
p
e
r
f
o
r
m
an
ce
o
f
th
e
m
o
d
el
is
o
b
tain
e
d
with
a
m
em
o
r
y
len
g
th
o
f
4
,
0
9
6
an
d
an
y
i
n
cr
ea
s
e
af
ter
t
h
at
d
o
es
d
ec
r
ea
s
e
th
e
p
er
f
o
r
m
an
ce
,
wh
ile
f
o
r
th
e
th
r
esh
o
ld
t
h
e
b
est
p
er
f
o
r
m
an
ce
was
r
ea
ch
ed
at
0
.
0
0
0
1
,
an
d
an
y
in
cr
ea
s
e
af
ter
th
at
also
n
eg
ativ
ely
af
f
ec
ted
th
e
m
o
d
e
l.
Fin
d
in
g
s
o
f
th
e
s
tu
d
y
s
u
g
g
est
th
at
th
e
n
o
n
-
d
eter
m
in
is
tic
f
in
ite
au
to
m
ato
n
(
NFA
)
m
o
d
el
with
m
em
o
r
y
len
g
t
h
=
4
,
0
9
6
,
iter
atio
n
s
=
7
5
0
,
an
d
t
h
r
esh
o
ld
=
0
.
0
0
1
p
e
r
f
o
r
m
ed
th
e
b
est with
9
5
.
4
5
% a
cc
u
r
ac
y
r
ate.
4.
CO
NCLU
SI
O
N
T
h
is
p
ap
er
was
f
o
c
u
s
ed
o
n
d
etec
tin
g
telec
o
m
f
r
au
d
u
s
in
g
AI
an
d
ML
.
I
t
p
r
o
v
i
d
ed
a
n
o
v
er
v
iew
o
f
m
ain
s
tr
ateg
ies
an
d
m
eth
o
d
s
u
s
ed
b
y
th
e
au
th
o
r
s
.
6
6
%
o
f
t
h
e
p
ap
er
s
u
s
ed
s
u
p
er
v
is
ed
s
tr
ateg
y
,
u
s
in
g
lab
eled
d
ataset
wh
ile
1
2
%
u
s
ed
u
n
s
u
p
er
v
is
ed
s
tr
ateg
y
,
u
tili
zin
g
u
n
lab
eled
d
ata,
an
d
1
8
%
u
s
ed
b
o
th
s
u
p
er
v
is
ed
an
d
u
n
s
u
p
er
v
is
ed
s
tr
ateg
ies.
Mo
r
eo
v
er
,
o
n
ly
2
%
o
f
t
h
e
p
a
p
er
s
u
s
ed
k
n
o
wled
g
e
-
b
ased
s
tr
a
teg
ies
an
d
th
e
last
r
em
ain
in
g
2
%
u
s
ed
a
co
m
b
in
atio
n
o
f
s
u
p
e
r
v
is
ed
,
u
n
s
u
p
e
r
v
is
ed
an
d
k
n
o
wled
g
e
-
b
ased
s
tr
ateg
ies.
R
eg
ar
d
in
g
th
e
h
ar
d
war
e
an
d
s
o
f
twar
e
u
s
e
,
8
8
%
o
f
th
e
p
ap
er
s
r
ev
iewe
d
u
s
ed
s
o
f
twar
e
an
aly
s
is
wh
ile
t
h
e
r
em
ain
in
g
1
2
%
u
s
ed
b
o
th
s
o
f
twar
e
a
n
d
h
ar
d
w
ar
e.
T
h
e
m
ain
d
if
f
er
e
n
ce
o
n
t
h
e
p
ap
er
s
wer
e
th
e
alg
o
r
ith
m
s
/class
if
ier
s
u
s
ed
f
o
r
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
S
ystema
tic
r
ev
iew
o
f fra
u
d
d
etec
tio
n
u
s
in
g
A
I
a
n
d
ML
w
ith
a
n
emp
h
a
s
is
o
n
…
(
S
o
ly
Ma
th
e
w
B
iju
)
3277
f
r
au
d
d
etec
tio
n
.
Ma
jo
r
ity
o
f
th
e
p
ap
er
s
u
s
ed
o
n
e
o
f
th
e
f
o
l
lo
win
g
m
o
d
els
s
u
ch
as
ANN,
SVM
,
o
r
d
ec
is
io
n
t
r
ee
as
th
e
b
est
o
n
e
f
o
r
telec
o
m
f
r
au
d
p
r
e
d
ictio
n
.
Ho
wev
er
,
o
n
ly
1
0
%
o
f
p
ap
er
s
u
s
ed
v
is
u
al
m
o
d
els
in
th
eir
ap
p
r
o
ac
h
es
,
s
u
ch
as
SOM
an
d
GNN.
As
p
er
th
e
ac
cu
r
a
cy
o
f
th
e
p
ap
er
s
,
7
6
%
o
f
th
em
d
id
p
r
o
v
id
e
th
e
ac
cu
r
ac
y
wh
ile
th
e
r
em
ain
in
g
2
4
% d
id
n
o
t.
ACK
NO
WL
E
DG
M
E
N
T
S
W
e
wan
t
to
th
an
k
th
e
Un
iv
er
s
ity
o
f
W
o
llo
n
g
o
n
g
in
Du
b
ai
(
UOWD)
f
o
r
p
r
o
v
id
in
g
th
e
g
r
a
n
t
f
o
r
th
is
r
esear
ch
wo
r
k
.
F
UNDING
I
NF
O
R
M
A
T
I
O
N
T
h
is
wo
r
k
was
s
u
p
p
o
r
ted
b
y
th
e
Un
iv
er
s
ity
o
f
W
o
llo
n
g
o
n
g
in
Du
b
ai
(
UOWD)
2
0
2
4
in
te
r
n
al
g
r
an
t
n
o
R
C
0
4
8
.
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
So
ly
Ma
th
ew
B
iju
✓
✓
✓
✓
✓
✓
✓
✓
✓
Sin
d
i Rry
ta
✓
✓
✓
✓
✓
✓
✓
✓
C
:
C
o
n
c
e
p
t
u
a
l
i
z
a
t
i
o
n
M
:
M
e
t
h
o
d
o
l
o
g
y
So
:
So
f
t
w
a
r
e
Va
:
Va
l
i
d
a
t
i
o
n
Fo
:
Fo
r
mal
a
n
a
l
y
s
i
s
I
:
I
n
v
e
s
t
i
g
a
t
i
o
n
R
:
R
e
so
u
r
c
e
s
D
:
D
a
t
a
C
u
r
a
t
i
o
n
O
:
W
r
i
t
i
n
g
-
O
r
i
g
i
n
a
l
D
r
a
f
t
E
:
W
r
i
t
i
n
g
-
R
e
v
i
e
w
&
E
d
i
t
i
n
g
Vi
:
Vi
su
a
l
i
z
a
t
i
o
n
Su
:
Su
p
e
r
v
i
s
i
o
n
P
:
P
r
o
j
e
c
t
a
d
mi
n
i
st
r
a
t
i
o
n
Fu
:
Fu
n
d
i
n
g
a
c
q
u
i
si
t
i
o
n
CO
NF
L
I
C
T
O
F
I
N
T
E
R
E
S
T
ST
A
T
E
M
E
NT
Au
th
o
r
s
s
tate
n
o
co
n
f
lict o
f
in
t
er
est.
I
NF
O
RM
E
D
CO
NS
E
N
T
No
p
er
s
o
n
al
in
f
o
r
m
atio
n
o
f
an
y
in
d
iv
id
u
al
h
as b
ee
n
u
s
ed
o
r
d
is
claim
ed
in
th
is
wo
r
k
.
DATA AV
AI
L
AB
I
L
I
T
Y
Data
a
v
aila
b
i
lit
y
is
n
o
t
a
p
p
li
ca
b
le
t
o
t
h
is
p
ap
er
as
n
o
n
e
w
d
a
t
a
w
er
e
cr
ea
te
d
o
r
an
al
y
z
e
d
i
n
t
h
is
s
t
u
d
y
.
RE
F
E
R
E
NC
E
S
[
1
]
C
a
m
b
r
i
d
g
e
D
i
c
t
i
o
n
a
r
y
,
“
M
e
a
n
i
n
g
o
f
f
r
a
u
d
i
n
E
n
g
l
i
sh
,
”
d
i
c
t
i
o
n
a
ry
.
c
a
m
b
ri
d
g
e
.
o
r
g
,
2
0
2
3
.
A
c
c
e
ss
e
d
:
M
a
r
.
2
4
,
2
0
2
5
.
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
s:
/
/
d
i
c
t
i
o
n
a
r
y
.
c
a
m
b
r
i
d
g
e
.
o
r
g
/
d
i
c
t
i
o
n
a
r
y
/
e
n
g
l
i
s
h
/
f
r
a
u
d
[
2
]
D
.
D
a
l
y
,
“
To
p
f
r
a
u
d
t
r
e
n
d
s
a
n
d
p
r
e
d
i
c
t
i
o
n
s
f
o
r
2
0
2
5
–
a
n
d
h
o
w
w
i
l
l
t
h
e
i
n
d
u
s
t
r
y
r
e
s
p
o
n
d
?
,
”
d
a
t
a
v
i
s
o
r.c
o
m
,
2
0
2
5
.
A
c
c
e
sse
d
:
M
a
r
.
2
4
,
2
0
2
5
.
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
s
:
/
/
w
w
w
.
d
a
t
a
v
i
s
o
r
.
c
o
m
/
b
l
o
g
/
f
r
a
u
d
-
t
r
e
n
d
s
-
p
r
e
d
i
c
t
i
o
n
s
-
2
0
2
5
[
3
]
A
sso
c
i
a
t
i
o
n
o
f
C
e
r
t
i
f
i
e
d
F
r
a
u
d
E
x
a
m
i
n
e
r
s,
“
R
e
p
o
r
t
t
o
t
h
e
N
a
t
i
o
n
-
2
0
2
0
g
l
o
b
a
l
st
u
d
y
o
n
o
c
c
u
p
a
t
i
o
n
a
l
f
r
a
u
d
a
n
d
a
b
u
se
,
”
Ass
o
c
i
a
t
i
o
n
o
f
C
e
r
t
i
f
i
e
d
F
ra
u
d
Ex
a
m
i
n
e
rs
,
I
n
c
.
,
2
0
2
0
.
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
s
:
/
/
a
c
f
e
p
u
b
l
i
c
.
s
3
-
us
-
w
e
s
t
-
2
.
a
m
a
z
o
n
a
w
s.c
o
m
/
2
0
2
0
-
R
e
p
o
r
t
-
to
-
t
h
e
-
N
a
t
i
o
n
s.
p
d
f
[
4
]
B
.
A
.
Y
e
h
y
a
a
n
d
N
.
S
a
l
h
a
b
,
“
T
e
l
e
c
o
mm
u
n
i
c
a
t
i
o
n
s
f
r
a
u
d
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
-
b
a
se
d
d
e
t
e
c
t
i
o
n
,
”
i
n
2
0
2
3
4
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
D
a
t
a
A
n
a
l
y
t
i
c
s f
o
r
Bu
s
i
n
e
ss
a
n
d
I
n
d
u
s
t
ry
,
O
c
t
.
2
0
2
3
,
p
p
.
6
5
6
–
6
6
1
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
D
A
B
I
6
0
1
4
5
.
2
0
2
3
.
1
0
6
2
9
6
1
2
.
[
5
]
M
.
R
.
A
l
B
o
u
g
h
a
,
“
C
o
m
p
a
r
i
n
g
d
a
t
a
mi
n
i
n
g
c
l
a
ssi
f
i
c
a
t
i
o
n
a
l
g
o
r
i
t
h
ms
i
n
d
e
t
e
c
t
i
o
n
o
f
S
I
M
b
o
x
f
r
a
u
d
,
”
M.
S
t
h
e
s
i
s
,
D
e
p
a
r
t
me
n
t
o
f
I
n
f
o
r
mat
i
o
n
S
y
st
e
ms,
S
t
.
C
l
o
u
d
S
t
a
t
e
U
n
i
v
e
r
si
t
y
,
M
i
n
n
e
s
o
t
a
,
U
n
i
t
e
d
S
t
a
t
e
s,
2
0
1
6
.
[
6
]
R
.
S
a
l
l
e
h
u
d
d
i
n
,
S
.
I
b
r
a
h
i
m
,
A
.
M
.
Za
i
n
,
a
n
d
A
.
H
.
El
m
i
,
“
D
e
t
e
c
t
i
n
g
S
I
M
b
o
x
f
r
a
u
d
b
y
u
si
n
g
s
u
p
p
o
r
t
v
e
c
t
o
r
m
a
c
h
i
n
e
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
,
”
J
u
r
n
a
l
T
e
k
n
o
l
o
g
i
,
v
o
l
.
7
4
,
n
o
.
1
,
A
p
r
.
2
0
1
5
,
d
o
i
:
1
0
.
1
1
1
1
3
/
j
t
.
v
7
4
.
2
6
4
9
.
[
7
]
A
.
K
r
e
n
k
e
r
,
M
.
V
o
l
k
,
U
.
S
e
d
l
a
r
,
J
.
B
e
št
e
r
,
a
n
d
A
.
K
o
s
,
“
B
i
d
i
r
e
c
t
i
o
n
a
l
a
r
t
i
f
i
c
i
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
s
f
o
r
mo
b
i
l
e
-
p
h
o
n
e
f
r
a
u
d
d
e
t
e
c
t
i
o
n
,
”
ETRI
J
o
u
r
n
a
l
,
v
o
l
.
3
1
,
n
o
.
1
,
p
p
.
9
2
–
9
4
,
F
e
b
.
2
0
0
9
,
d
o
i
:
1
0
.
4
2
1
8
/
e
t
r
i
j
.
0
9
.
0
2
0
8
.
0
2
4
5
.
[
8
]
P
.
S
a
r
a
v
a
n
a
n
,
V
.
S
u
b
r
a
ma
n
i
y
a
sw
a
m
y
,
N
.
S
i
v
a
r
a
m
a
k
r
i
s
h
n
a
n
,
M
.
A
.
P
r
a
k
a
sh
,
a
n
d
T.
A
r
u
n
k
u
m
a
r
,
“
D
a
t
a
m
i
n
i
n
g
a
p
p
r
o
a
c
h
f
o
r
su
b
s
c
r
i
p
t
i
o
n
-
f
r
a
u
d
d
e
t
e
c
t
i
o
n
i
n
t
e
l
e
c
o
mm
u
n
i
c
a
t
i
o
n
se
c
t
o
r
,
”
C
o
n
t
e
m
p
o
r
a
r
y
En
g
i
n
e
e
r
i
n
g
S
c
i
e
n
c
e
s
,
v
o
l
.
7
,
p
p
.
5
1
5
–
5
2
2
,
2
0
1
4
,
d
o
i
:
1
0
.
1
2
9
8
8
/
c
e
s.
2
0
1
4
.
4
4
3
1
.
[
9
]
G
.
W
a
ssi
e
,
“
S
u
b
s
c
r
i
p
t
i
o
n
f
r
a
u
d
d
e
t
e
c
t
i
o
n
u
s
i
n
g
a
u
t
o
e
n
c
o
d
e
r
i
n
c
a
se
o
f
e
t
h
i
o
t
e
l
e
c
o
m,
”
J
o
u
rn
a
l
o
f
Ad
v
a
n
c
e
R
e
se
a
rc
h
i
n
S
o
c
i
a
l
S
c
i
e
n
c
e
a
n
d
H
u
m
a
n
i
t
i
e
s
,
v
o
l
.
1
0
,
n
o
.
7
,
p
p
.
1
1
–
2
3
,
O
c
t
.
2
0
2
4
,
d
o
i
:
1
0
.
6
1
8
4
1
/
t
p
4
k
9
0
1
2
.
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.
15
,
No
.
4
,
Au
g
u
s
t 2
0
2
6
:
3
2
6
9
-
3
2
8
5
3278
[
1
0
]
H
.
G
r
o
sser,
P
.
B
r
i
t
o
s,
a
n
d
R
.
G
.
-
M
a
r
t
í
n
e
z
,
“
D
e
t
e
c
t
i
n
g
f
r
a
u
d
i
n
m
o
b
i
l
e
t
e
l
e
p
h
o
n
y
u
si
n
g
n
e
u
r
a
l
n
e
t
w
o
r
k
s,
”
i
n
I
n
n
o
v
a
t
i
o
n
s
i
n
A
p
p
l
i
e
d
Art
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
,
B
e
r
l
i
n
,
H
e
i
d
e
l
b
e
r
g
:
S
p
r
i
n
g
e
r
,
2
0
0
5
,
p
p
.
6
1
3
–
6
1
5
,
d
o
i
:
1
0
.
1
0
0
7
/
1
1
5
0
4
8
9
4
_
8
5
.
[
1
1
]
A
.
H
.
E
l
mi
,
S
.
I
b
r
a
h
i
m,
a
n
d
R
.
S
a
l
l
e
h
u
d
d
i
n
,
“
D
e
t
e
c
t
i
n
g
S
I
M
b
o
x
f
r
a
u
d
u
s
i
n
g
n
e
u
r
a
l
n
e
t
w
o
r
k
,
”
i
n
I
T
C
o
n
v
e
r
g
e
n
c
e
a
n
d
S
e
c
u
r
i
t
y
2
0
1
2
,
D
o
r
d
r
e
c
h
t
:
S
p
r
i
n
g
e
r
,
2
0
1
3
,
p
p
.
5
7
5
–
5
8
2
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
94
-
007
-
5
8
6
0
-
5
_
6
9
.
[
1
2
]
M
.
S
a
n
v
e
r
a
n
d
A
.
K
a
r
a
h
o
c
a
,
“
F
r
a
u
d
d
e
t
e
c
t
i
o
n
u
si
n
g
a
n
a
d
a
p
t
i
v
e
n
e
u
r
o
-
f
u
z
z
y
i
n
f
e
r
e
n
c
e
sy
s
t
e
m
i
n
m
o
b
i
l
e
t
e
l
e
c
o
mm
u
n
i
c
a
t
i
o
n
n
e
t
w
o
r
k
s,”
J
o
u
r
n
a
l
o
f
Mu
l
t
i
p
l
e
-
v
a
l
u
e
d
L
o
g
i
c
a
n
d
S
o
f
t
C
o
m
p
u
t
i
n
g
,
v
o
l
.
1
5
,
n
o
.
2
,
p
p
.
1
5
5
–
1
7
9
,
2
0
0
9
.
[
1
3
]
B
.
M
.
T
h
a
d
d
e
u
s,
M
.
M
u
h
a
mm
a
d
,
Y
.
M
.
M
a
l
g
w
i
,
a
n
d
E.
T.
M
a
r
t
i
n
,
“
S
i
m
-
b
o
x
i
n
g
f
r
a
u
d
d
e
t
e
c
t
i
o
n
sy
s
t
e
m
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
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
D
e
v
e
l
o
p
m
e
n
t
Ma
t
h
e
m
a
t
i
c
s
,
v
o
l
.
1
,
n
o
.
4
,
p
p
.
2
0
1
–
2
1
3
,
D
e
c
.
2
0
2
4
,
d
o
i
:
1
0
.
6
2
0
5
4
/
i
j
d
m/
0
1
0
4
.
1
6
.
[
1
4
]
C
.
S
.
H
i
l
a
s
a
n
d
P
.
A
.
M
a
st
o
r
o
c
o
s
t
a
s
,
“
A
n
a
p
p
l
i
c
a
t
i
o
n
o
f
s
u
p
e
r
v
i
se
d
a
n
d
u
n
su
p
e
r
v
i
se
d
l
e
a
r
n
i
n
g
a
p
p
r
o
a
c
h
e
s
t
o
t
e
l
e
c
o
mm
u
n
i
c
a
t
i
o
n
s
f
r
a
u
d
d
e
t
e
c
t
i
o
n
,
”
K
n
o
w
l
e
d
g
e
-
B
a
s
e
d
S
y
st
e
m
s
,
v
o
l
.
2
1
,
n
o
.
7
,
p
p
.
7
2
1
–
7
2
6
,
O
c
t
.
2
0
0
8
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
k
n
o
s
y
s.
2
0
0
8
.
0
3
.
0
2
6
.
[
1
5
]
H
.
Li
e
t
a
l
.
,
“
A
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
a
p
p
r
o
a
c
h
t
o
p
r
e
v
e
n
t
ma
l
i
c
i
o
u
s
c
a
l
l
s
o
v
e
r
t
e
l
e
p
h
o
n
y
n
e
t
w
o
r
k
s,
”
i
n
2
0
1
8
I
E
E
E
S
y
m
p
o
si
u
m
o
n
S
e
c
u
r
i
t
y
a
n
d
P
ri
v
a
c
y
,
M
a
y
2
0
1
8
,
p
p
.
5
3
–
6
9
,
d
o
i
:
1
0
.
1
1
0
9
/
S
P
.
2
0
1
8
.
0
0
0
3
4
.
[
1
6
]
A
.
H
.
El
mi
,
R
.
S
a
l
l
e
h
u
d
d
i
n
,
S
.
I
b
r
a
h
i
m,
a
n
d
A
.
M
.
Z
a
i
n
,
“
C
l
a
ssi
f
i
c
a
t
i
o
n
o
f
S
I
M
b
o
x
f
r
a
u
d
d
e
t
e
c
t
i
o
n
u
si
n
g
s
u
p
p
o
r
t
v
e
c
t
o
r
ma
c
h
i
n
e
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
t
e
r
n
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
f
I
n
n
o
v
a
t
i
v
e
C
o
m
p
u
t
i
n
g
,
v
o
l
.
4
,
n
o
.
2
,
p
p
.
1
9
–
2
7
,
2
0
1
4
.
[
1
7
]
P
.
B
a
r
so
n
,
S
.
F
i
e
l
d
,
N
.
D
a
v
e
y
,
G
.
M
c
A
sk
i
e
,
a
n
d
R
.
F
r
a
n
k
,
“
D
e
t
e
c
t
i
o
n
o
f
f
r
a
u
d
i
n
m
o
b
i
l
e
p
h
o
n
e
n
e
t
w
o
r
k
s
,
”
Re
se
a
rch
G
a
t
e
,
1
9
9
6
.
[
1
8
]
I
.
M
u
r
y
n
e
t
s,
M
.
Z
a
b
a
r
a
n
k
i
n
,
R
.
P
.
J
o
v
e
r
,
a
n
d
A
.
P
a
n
a
g
i
a
,
“
A
n
a
l
y
si
s
a
n
d
d
e
t
e
c
t
i
o
n
o
f
S
I
M
b
o
x
f
r
a
u
d
i
n
m
o
b
i
l
i
t
y
n
e
t
w
o
r
k
s,
”
i
n
I
EEE
C
o
n
f
e
re
n
c
e
o
n
C
o
m
p
u
t
e
r
C
o
m
m
u
n
i
c
a
t
i
o
n
s
,
A
p
r
.
2
0
1
4
,
p
p
.
1
5
1
9
–
1
5
2
6
,
d
o
i
:
1
0
.
1
1
0
9
/
I
N
F
O
C
O
M
.
2
0
1
4
.
6
8
4
8
0
8
7
.
[
1
9
]
H
.
M
.
M
a
r
a
h
,
O
.
M
.
El
r
a
j
u
b
i
,
a
n
d
A
.
A
.
A
b
o
u
d
a
,
“
F
r
a
u
d
d
e
t
e
c
t
i
o
n
i
n
i
n
t
e
r
n
a
t
i
o
n
a
l
c
a
l
l
s
u
s
i
n
g
f
u
z
z
y
l
o
g
i
c
,
”
i
n
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
C
o
m
p
u
t
e
r
Vi
s
i
o
n
a
n
d
I
m
a
g
e
A
n
a
l
y
s
i
s
Ap
p
l
i
c
a
t
i
o
n
s
,
Ja
n
.
2
0
1
5
,
p
p
.
1
–
6
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
C
V
I
A
.
2
0
1
5
.
7
3
5
1
8
9
1
.
[
2
0
]
F
.
N
.
O
g
w
u
e
l
e
k
a
,
“
F
r
a
u
d
d
e
t
e
c
t
i
o
n
i
n
m
o
b
i
l
e
c
o
mm
u
n
i
c
a
t
i
o
n
s
n
e
t
w
o
r
k
s
u
s
i
n
g
u
s
e
r
p
r
o
f
i
l
i
n
g
a
n
d
c
l
a
ss
i
f
i
c
a
t
i
o
n
t
e
c
h
n
i
q
u
e
s,
”
J
o
u
rn
a
l
o
f
S
c
i
e
n
c
e
a
n
d
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
2
9
,
n
o
.
3
,
p
p
.
3
1
–
4
2
,
2
0
0
9
,
d
o
i
:
1
0
.
4
3
1
4
/
j
u
st
.
v
2
9
i
3
.
5
0
0
5
2
.
[
2
1
]
F
.
O
.
A
r
a
n
u
w
a
,
“
H
y
b
r
i
d
i
z
e
d
i
n
t
e
l
l
i
g
e
n
t
d
a
t
a
a
n
a
l
y
s
i
s
m
o
d
e
l
f
o
r
f
r
a
u
d
d
e
t
e
c
t
i
o
n
i
n
m
o
b
i
l
e
c
o
mm
u
n
i
c
a
t
i
o
n
n
e
t
w
o
r
k
s
,
”
A
c
a
d
e
m
i
a
J
o
u
rn
a
l
o
f
S
c
i
e
n
t
i
f
i
c
Re
s
e
a
rc
h
,
v
o
l
.
1
,
n
o
.
5
,
p
p
.
0
8
2
–
0
8
9
,
2
0
1
3
.
[
2
2
]
Y
.
R
e
b
a
h
i
,
R
.
R
u
p
p
e
l
t
,
M
.
N
a
ssar,
a
n
d
O
.
F
e
st
o
r
,
“
S
C
A
M
S
TO
P
:
a
p
l
a
t
f
o
r
m
f
o
r
m
i
t
i
g
a
t
i
n
g
f
r
a
u
d
i
n
V
o
I
P
e
n
v
i
r
o
n
me
n
t
s,”
i
n
N
e
t
w
o
rk
a
n
d
T
r
a
f
f
i
c
En
g
i
n
e
e
ri
n
g
i
n
Em
e
r
g
i
n
g
D
i
st
r
i
b
u
t
e
d
C
o
m
p
u
t
i
n
g
Ap
p
l
i
c
a
t
i
o
n
s
,
R
e
s
e
a
r
c
h
G
a
t
e
,
2
0
1
1
,
p
p
.
3
0
2
–
3
2
5
,
d
o
i
:
1
0
.
4
0
1
8
/
9
7
8
-
1
-
4
6
6
6
-
1
8
8
8
-
6
.
c
h
0
1
2
[
2
3
]
J.
H
o
l
l
mé
n
,
“
U
s
e
r
p
r
o
f
i
l
i
n
g
a
n
d
c
l
a
ssi
f
i
c
a
t
i
o
n
f
o
r
f
r
a
u
d
d
e
t
e
c
t
i
o
n
i
n
m
o
b
i
l
e
c
o
mm
u
n
i
c
a
t
i
o
n
s
n
e
t
w
o
r
k
s
,
”
P
h
.
D
.
d
i
ss
e
rt
a
t
i
o
n
,
D
e
p
a
r
t
me
n
t
o
f
C
o
mp
u
t
e
r
S
c
i
e
n
c
e
a
n
d
En
g
i
n
e
e
r
i
n
g
,
A
a
l
t
o
U
n
i
v
e
r
s
i
t
y
,
Es
p
o
o
,
F
i
n
l
a
n
d
,
2
0
0
0
.
[
2
4
]
Y
.
M
o
r
e
a
u
,
B
.
P
r
e
n
e
e
l
,
P
.
B
u
r
g
e
,
J.
S
.
-
Ta
y
l
o
r
,
C
.
S
t
o
e
r
m
a
n
n
,
a
n
d
C
.
C
o
o
k
e
,
“
N
o
v
e
l
t
e
c
h
n
i
q
u
e
s
f
o
r
f
r
a
u
d
d
e
t
e
c
t
i
o
n
i
n
m
o
b
i
l
e
t
e
l
e
c
o
mm
u
n
i
c
a
t
i
o
n
n
e
t
w
o
r
k
s,
”
R
e
rse
a
rch
G
a
t
e
,
1
9
9
8
.
[
2
5
]
M
.
T
a
n
i
g
u
c
h
i
,
M
.
H
a
f
t
,
J.
H
o
l
l
m
e
n
,
a
n
d
V
.
Tr
e
sp
,
“
F
r
a
u
d
d
e
t
e
c
t
i
o
n
i
n
c
o
mm
u
n
i
c
a
t
i
o
n
n
e
t
w
o
r
k
s
u
si
n
g
n
e
u
r
a
l
a
n
d
p
r
o
b
a
b
i
l
i
s
t
i
c
met
h
o
d
s,
”
i
n
1
9
9
8
I
EE
E
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
A
c
o
u
st
i
c
s
,
S
p
e
e
c
h
a
n
d
S
i
g
n
a
l
Pro
c
e
ss
i
n
g
,
1
9
9
8
,
p
p
.
1
2
4
1
–
1
2
4
4
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
A
S
S
P
.
1
9
9
8
.
6
7
5
4
9
6
.
[
2
6
]
J.
H
o
l
l
mé
n
a
n
d
V
.
Tr
e
s
p
,
“
C
a
l
l
-
b
a
se
d
f
r
a
u
d
d
e
t
e
c
t
i
o
n
i
n
mo
b
i
l
e
c
o
mm
u
n
i
c
a
t
i
o
n
n
e
t
w
o
r
k
s
u
s
i
n
g
a
h
i
e
r
a
r
c
h
i
c
a
l
R
e
g
i
me
-
sw
i
t
c
h
i
n
g
mo
d
e
l
,
”
i
n
1
9
9
8
c
o
n
f
e
re
n
c
e
o
n
Ad
v
a
n
c
e
s
i
n
n
e
u
r
a
l
i
n
f
o
rm
a
t
i
o
n
p
r
o
c
e
ssi
n
g
syst
e
m
s
I
I
,
1
9
9
8
,
p
p
.
8
8
9
-
8
9
5
.
[
2
7
]
J.
H
o
l
l
m
é
n
,
V
.
Tr
e
s
p
,
a
n
d
O
.
S
i
m
u
l
a
,
“
A
sel
f
-
o
r
g
a
n
i
z
i
n
g
ma
p
f
o
r
c
l
u
s
t
e
r
i
n
g
p
r
o
b
a
b
i
l
i
s
t
i
c
m
o
d
e
l
s
,
”
i
n
9
t
h
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
Art
i
f
i
c
i
a
l
N
e
u
r
a
l
N
e
t
w
o
rk
s
,
1
9
9
9
,
p
p
.
9
4
6
–
9
5
1
,
d
o
i
:
1
0
.
1
0
4
9
/
c
p
:
1
9
9
9
1
2
3
4
.
[
2
8
]
K
.
G
.
D
.
C
.
K
e
h
e
l
w
a
l
a
,
“
R
e
a
l
-
t
i
m
e
f
r
a
u
d
d
e
t
e
c
t
i
o
n
i
n
t
e
l
e
c
o
mm
u
n
i
c
a
t
i
o
n
n
e
t
w
o
r
k
u
s
i
n
g
c
a
l
l
p
a
t
t
e
r
n
a
n
a
l
y
si
s
,
”
M.
S
.
t
h
e
s
i
s
,
D
e
p
a
r
t
me
n
t
o
f
C
o
mp
u
t
e
r
S
c
i
e
n
c
e
a
n
d
En
g
i
n
e
e
r
i
n
g
,
U
n
i
v
e
r
s
i
t
y
o
f
M
o
r
a
t
u
w
a
,
M
o
r
a
t
u
w
a
,
S
r
i
L
a
n
k
a
,
2
0
1
7
.
[
2
9
]
M
.
K
a
sh
i
r
a
n
d
S
.
B
a
s
h
i
r
,
“
M
a
c
h
i
n
e
l
e
a
r
n
i
n
g
t
e
c
h
n
i
q
u
e
s
f
o
r
S
I
M
b
o
x
f
r
a
u
d
d
e
t
e
c
t
i
o
n
,
”
i
n
2
0
1
9
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
T
e
c
h
n
o
l
o
g
i
e
s
,
M
a
r
.
2
0
1
9
,
p
p
.
4
–
8
,
d
o
i
:
1
0
.
1
1
0
9
/
C
O
M
TE
C
H
.
2
0
1
9
.
8
7
3
7
8
2
8
.
[
3
0
]
J.
L
i
,
C
.
Z
h
a
n
g
,
a
n
d
L.
Ji
a
n
g
,
“
I
n
n
o
v
a
t
i
v
e
t
e
l
e
c
o
m
f
r
a
u
d
d
e
t
e
c
t
i
o
n
:
a
n
e
w
d
a
t
a
se
t
a
n
d
a
n
a
d
v
a
n
c
e
d
m
o
d
e
l
w
i
t
h
R
o
B
ER
Ta
a
n
d
d
u
a
l
l
o
ss
f
u
n
c
t
i
o
n
s
,
”
Ap
p
l
i
e
d
S
c
i
e
n
c
e
s
,
v
o
l
.
1
4
,
n
o
.
2
4
,
D
e
c
.
2
0
2
4
,
d
o
i
:
1
0
.
3
3
9
0
/
a
p
p
1
4
2
4
1
1
6
2
8
.
[
3
1
]
Q
.
Z
h
a
o
,
K
.
C
h
e
n
,
T
.
L
i
,
Y
.
Y
a
n
g
,
a
n
d
X
.
W
a
n
g
,
“
D
e
t
e
c
t
i
n
g
t
e
l
e
c
o
m
mu
n
i
c
a
t
i
o
n
f
r
a
u
d
b
y
u
n
d
e
r
st
a
n
d
i
n
g
t
h
e
c
o
n
t
e
n
t
s
o
f
a
c
a
l
l
,
”
C
y
b
e
rs
e
c
u
ri
t
y
,
v
o
l
.
1
,
n
o
.
1
,
D
e
c
.
2
0
1
8
,
d
o
i
:
1
0
.
1
1
8
6
/
s4
2
4
0
0
-
0
1
8
-
0
0
0
8
-
5.
[
3
2
]
S
.
S
u
b
u
d
h
i
a
n
d
S
.
P
a
n
i
g
r
a
h
i
,
“
U
se
o
f
p
o
ss
i
b
i
l
i
st
i
c
f
u
z
z
y
C
-
m
e
a
n
s
c
l
u
st
e
r
i
n
g
f
o
r
t
e
l
e
c
o
m
f
r
a
u
d
d
e
t
e
c
t
i
o
n
,
”
i
n
C
o
m
p
u
t
a
t
i
o
n
a
l
I
n
t
e
l
l
i
g
e
n
c
e
i
n
D
a
t
a
Mi
n
i
n
g
,
S
i
n
g
a
p
o
r
e
:
S
p
r
i
n
g
e
r
,
2
0
1
7
,
p
p
.
6
3
3
–
6
4
1
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
9
8
1
-
10
-
3
8
7
4
-
7
_
6
0
.
[
3
3
]
F
.
D
a
n
i
e
l
,
“
B
a
y
e
s
i
a
n
a
n
d
D
e
m
p
st
e
r
-
S
h
a
f
e
r
m
o
d
e
l
s
f
o
r
c
o
m
b
i
n
i
n
g
mu
l
t
i
p
l
e
so
u
r
c
e
s
o
f
e
v
i
d
e
n
c
e
i
n
a
f
r
a
u
d
d
e
t
e
c
t
i
o
n
s
y
s
t
e
m,
”
A
p
r
.
2
0
2
1
,
a
rX
i
v
:
2
1
0
4
.
0
7
4
4
0
.
[
3
4
]
A
.
B
o
u
k
e
r
c
h
e
a
n
d
M
.
S
.
M
.
A
.
N
o
t
a
r
e
,
“
B
e
h
a
v
i
o
r
-
b
a
s
e
d
i
n
t
r
u
s
i
o
n
d
e
t
e
c
t
i
o
n
i
n
m
o
b
i
l
e
p
h
o
n
e
s
y
st
e
ms,
”
J
o
u
r
n
a
l
o
f
P
a
ra
l
l
e
l
a
n
d
D
i
st
ri
b
u
t
e
d
C
o
m
p
u
t
i
n
g
,
v
o
l
.
6
2
,
n
o
.
9
,
p
p
.
1
4
7
6
–
1
4
9
0
,
S
e
p
.
2
0
0
2
,
d
o
i
:
1
0
.
1
0
0
6
/
j
p
d
c
.
2
0
0
2
.
1
8
5
7
.
[
3
5
]
N
.
A
l
g
h
a
w
i
,
“
A
st
u
d
y
o
n
S
I
M
b
o
x
o
r
i
n
t
e
r
c
o
n
n
e
c
t
b
y
p
a
ss
f
r
a
u
d
,
”
M
.
S
.
t
h
e
si
s
,
D
e
p
a
r
t
me
n
t
o
f
I
n
f
o
r
mat
i
c
s,
Th
e
B
r
i
t
i
sh
U
n
i
v
e
r
s
i
t
y
i
n
D
u
b
a
i
,
D
u
b
a
i
,
U
A
E,
2
0
1
9
.
[
3
6
]
R
.
B
h
o
w
mi
k
,
“
D
a
t
a
mi
n
i
n
g
t
e
c
h
n
i
q
u
e
s
i
n
f
r
a
u
d
d
e
t
e
c
t
i
o
n
,
”
J
o
u
r
n
a
l
o
f
D
i
g
i
t
a
l
F
o
re
n
si
c
s,
S
e
c
u
r
i
t
y
a
n
d
L
a
w
,
v
o
l
.
3
,
n
o
.
2
,
2
0
0
8
,
d
o
i
:
1
0
.
1
5
3
9
4
/
j
d
f
s
l
.
2
0
0
8
.
1
0
4
0
.
[
3
7
]
P
.
A
.
Es
t
é
v
e
z
,
C
.
M
.
H
e
l
d
,
a
n
d
C
.
A
.
P
e
r
e
z
,
“
S
u
b
s
c
r
i
p
t
i
o
n
f
r
a
u
d
p
r
e
v
e
n
t
i
o
n
i
n
t
e
l
e
c
o
mm
u
n
i
c
a
t
i
o
n
s
u
si
n
g
f
u
z
z
y
r
u
l
e
s
a
n
d
n
e
u
r
a
l
n
e
t
w
o
r
k
s,”
Ex
p
e
r
t
S
y
s
t
e
m
s
w
i
t
h
A
p
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
3
1
,
n
o
.
2
,
p
p
.
3
3
7
–
3
4
4
,
A
u
g
.
2
0
0
6
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
e
sw
a
.
2
0
0
5
.
0
9
.
0
2
8
.
[
3
8
]
S
.
Q
a
y
y
u
m
,
S
.
M
a
n
s
o
o
r
,
A
.
K
h
a
l
i
d
,
K
h
u
s
h
b
a
k
h
t
,
Z.
H
a
l
i
m
,
a
n
d
A
.
R
.
B
a
i
g
,
“
F
r
a
u
d
u
l
e
n
t
c
a
l
l
d
e
t
e
c
t
i
o
n
f
o
r
mo
b
i
l
e
n
e
t
w
o
r
k
s,”
i
n
2
0
1
0
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
f
o
rm
a
t
i
o
n
a
n
d
Em
e
rg
i
n
g
T
e
c
h
n
o
l
o
g
i
e
s
,
2
0
1
0
,
p
p
.
1
–
5
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
I
ET.
2
0
1
0
.
5
6
2
5
7
1
8
.
[
3
9
]
C
.
X
i
o
n
g
,
“
Te
l
e
c
o
m fr
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
,
”
M
.
S
t
h
e
si
s
,
S
c
h
o
o
l
o
f
E
l
e
c
t
r
i
c
a
l
En
g
i
n
e
e
r
i
n
g
a
n
d
C
o
mp
u
t
e
r
S
c
i
e
n
c
e
,
K
TH
R
o
y
a
l
I
n
st
i
t
u
t
e
o
f
T
e
c
h
n
o
l
o
g
y
,
S
t
o
c
k
h
o
l
m,
2
0
2
2
.
[
4
0
]
S
.
K
.
Ja
y
a
s
i
n
g
h
a
n
d
A
.
K
.
S
w
a
i
n
,
“
N
e
u
r
a
l
n
e
t
w
o
r
k
i
n
f
r
a
u
d
d
e
t
e
c
t
i
o
n
,
”
i
n
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
I
n
d
u
s
t
ri
a
l
A
p
p
l
i
c
a
t
i
o
n
o
f
S
o
f
t
C
o
m
p
u
t
i
n
g
T
e
c
h
n
i
q
u
e
s
,
2
0
1
1
,
p
p
.
1
-
8
.
[
4
1
]
O
.
A
.
A
b
i
d
o
g
u
n
,
“
D
a
t
a
mi
n
i
n
g
,
f
r
a
u
d
d
e
t
e
c
t
i
o
n
a
n
d
m
o
b
i
l
e
t
e
l
e
c
o
mm
u
n
i
c
a
t
i
o
n
s
:
c
a
l
l
p
a
t
t
e
r
n
a
n
a
l
y
si
s
w
i
t
h
u
n
su
p
e
r
v
i
se
d
n
e
u
r
a
l
n
e
t
w
o
r
k
s
,
”
M.
S
t
h
e
s
i
s
,
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
,
U
n
i
v
e
r
si
t
y
o
f
t
h
e
W
e
st
e
r
n
C
a
p
e
,
C
a
p
e
To
w
n
,
2
0
0
5
.
[
4
2
]
E.
M
.
D
.
D
j
o
ma
d
j
i
,
K
.
I
.
B
a
s
i
l
e
,
T.
T.
C
h
r
i
st
i
a
n
,
F
.
V
.
K
.
D
j
o
k
o
,
a
n
d
M
.
E
.
S
o
n
e
,
“
M
a
c
h
i
n
e
l
e
a
r
n
i
n
g
-
b
a
s
e
d
a
p
p
r
o
a
c
h
f
o
r
i
d
e
n
t
i
f
i
c
a
t
i
o
n
o
f
S
I
M
b
o
x
b
y
p
a
ss
f
r
a
u
d
i
n
a
t
e
l
e
c
o
m
n
e
t
w
o
r
k
b
a
se
d
o
n
C
D
R
a
n
a
l
y
s
i
s:
c
a
s
e
o
f
a
f
i
x
e
d
a
n
d
m
o
b
i
l
e
o
p
e
r
a
t
o
r
i
n
C
a
mer
o
o
n
,
”
J
o
u
r
n
a
l
o
f
C
o
m
p
u
t
e
r
a
n
d
C
o
m
m
u
n
i
c
a
t
i
o
n
s
,
v
o
l
.
1
1
,
n
o
.
0
2
,
p
p
.
1
4
2
–
1
5
7
,
2
0
2
3
,
d
o
i
:
1
0
.
4
2
3
6
/
j
c
c
.
2
0
2
3
.
1
1
2
0
1
0
.
[
4
3
]
A
.
O
l
u
sh
o
l
a
a
n
d
J
.
M
a
r
t
,
“
F
r
a
u
d
d
e
t
e
c
t
i
o
n
u
s
i
n
g
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
,
”
S
c
i
e
n
c
e
O
p
e
n
Pre
p
ri
n
t
s
,
Ja
n
.
2
0
2
4
,
d
o
i
:
1
0
.
1
4
2
9
3
/
P
R
2
1
9
9
.
0
0
0
6
4
7
.
v
1
.
[
4
4
]
M
.
S
a
h
i
n
,
“
U
n
d
e
r
st
a
n
d
i
n
g
t
e
l
e
p
h
o
n
y
f
r
a
u
d
a
s
a
n
e
ss
e
n
t
i
a
l
s
t
e
p
t
o
b
e
t
t
e
r
f
i
g
h
t
i
t
,
”
Ph
.
D
.
t
h
e
si
s
,
D
e
p
a
r
t
me
n
t
o
f
D
i
g
i
t
a
l
S
e
c
u
r
i
t
y
,
TELE
C
O
M
P
a
r
i
sT
e
c
h
,
P
a
r
i
s
,
F
r
a
n
c
e
,
2
0
1
7
.
[
4
5
]
H
.
H
.
K
i
l
i
n
c
,
“
A
c
a
s
e
s
t
u
d
y
o
n
f
r
a
u
d
u
l
e
n
t
u
s
e
r
b
e
h
a
v
i
o
r
s
i
n
t
h
e
t
e
l
e
c
o
mm
u
n
i
c
a
t
i
o
n
n
e
t
w
o
r
k
,
”
El
e
c
t
ri
c
a
,
v
o
l
.
2
1
,
n
o
.
1
,
p
p
.
7
4
–
8
4
,
Jan
.
2
0
2
1
,
d
o
i
:
1
0
.
5
1
5
2
/
e
l
e
c
t
r
i
c
a
.
2
0
2
1
.
2
0
0
5
0
.
Evaluation Warning : The document was created with Spire.PDF for Python.