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
.
2
,
A
p
r
il
20
26
,
p
p
.
1
4
0
2
~
1
4
0
8
I
SS
N:
2
2
5
2
-
8
9
3
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijai.v
15
.i
2
.
p
p
1
4
0
2
-
1
4
0
8
1402
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
i
.
ia
esco
r
e.
co
m
The eff
ects
o
f
d
a
t
a
i
mb
a
l
a
nce
o
n
f
r
a
ud
detect
io
n m
o
del
a
ccu
ra
cy
Rus
m
a
Anieza
Rus
la
n
1
,
N
ureize
Arba
iy
1
,
P
ei
-
Chun
L
i
n
2
1
F
a
c
u
l
t
y
o
f
C
o
mp
u
t
e
r
S
c
i
e
n
c
e
a
n
d
I
n
f
o
r
mat
i
o
n
Te
c
h
n
o
l
o
g
y
,
U
n
i
v
e
r
si
t
i
Tu
n
H
u
sse
i
n
O
n
n
M
a
l
a
y
si
a
,
P
a
r
i
t
R
a
j
a
,
M
a
l
a
y
si
a
2
D
e
p
a
r
t
me
n
t
o
f
I
n
f
o
r
mat
i
o
n
En
g
i
n
e
e
r
i
n
g
a
n
d
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
F
e
n
g
C
h
i
a
U
n
i
v
e
r
si
t
y
,
T
a
i
c
h
u
n
g
,
T
a
i
w
a
n
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Au
g
27
,
2
0
2
4
R
ev
is
ed
Feb
9
,
2
0
2
6
Acc
ep
ted
Mar
5
,
2
0
2
6
M
a
c
h
in
e
lea
rn
in
g
(M
L)
m
o
d
e
l
p
e
rfo
rm
a
n
c
e
is
o
ften
a
ss
e
ss
e
d
b
y
a
c
c
u
ra
c
y
,
b
u
t
t
h
e
q
u
a
li
ty
a
n
d
b
a
lan
c
e
o
f
d
a
ta
a
lso
p
la
y
c
ru
c
ial
ro
les
.
I
m
b
a
lan
c
e
d
d
a
tas
e
ts,
wh
e
re
th
e
m
in
o
ri
ty
c
las
s
h
a
s
fe
we
r
sa
m
p
les
th
a
n
th
e
m
a
jo
rit
y
c
las
s,
c
a
n
lea
d
t
o
b
ias
e
d
p
re
d
icti
o
n
s
fa
v
o
r
in
g
t
h
e
m
a
jo
rit
y
c
las
s.
Th
is
st
u
d
y
a
d
d
re
ss
e
s
th
e
issu
e
o
f
c
las
s
i
m
b
a
lan
c
e
th
ro
u
g
h
re
sa
m
p
li
n
g
t
e
c
h
n
iq
u
e
s,
in
c
lu
d
in
g
ra
n
d
o
m
u
n
d
e
rsa
m
p
li
n
g
(RUS)
a
n
d
ra
n
d
o
m
o
v
e
rsa
m
p
li
n
g
(ROS)
,
sp
e
c
ifi
c
a
ll
y
a
p
p
li
e
d
to
a
fra
u
d
d
e
tec
ti
o
n
d
a
tas
e
t.
We
c
las
sify
th
e
r
e
sa
m
p
led
d
a
tas
e
ts
u
sin
g
ra
n
d
o
m
f
o
re
st
(RF
)
a
n
d
g
ra
d
ie
n
t
b
o
o
stin
g
(G
B)
m
o
d
e
ls.
Ou
r
fin
d
i
n
g
s
in
d
ica
te
t
h
a
t
t
h
e
RF
m
o
d
e
l,
w
h
e
n
c
o
m
b
in
e
d
wit
h
ROS,
a
c
h
iev
e
s
a
n
a
c
c
u
ra
c
y
o
f
9
7
.
4
%
,
s
u
rp
a
ss
in
g
t
h
e
9
6
.
1
%
a
c
c
u
ra
c
y
o
f
th
e
G
B
m
o
d
e
l
with
RUS.
Th
is
a
p
p
r
o
a
c
h
d
e
m
o
n
str
a
tes
th
e
imp
o
rta
n
c
e
o
f
a
d
d
re
ss
in
g
c
las
s
imb
a
lan
c
e
to
imp
r
o
v
e
p
re
d
icti
o
n
a
c
c
u
ra
c
y
in
M
L.
K
ey
w
o
r
d
s
:
Data
au
g
m
en
tatio
n
I
m
b
alan
ce
d
d
ataset
Ma
ch
in
e
lear
n
in
g
R
esam
p
lin
g
SMOT
E
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
Nu
r
eize
Ar
b
aiy
Facu
lty
o
f
C
o
m
p
u
ter
Scien
ce
an
d
I
n
f
o
r
m
atio
n
T
ec
h
n
o
lo
g
y
,
Un
iv
er
s
iti T
u
n
Hu
s
s
ein
On
n
Ma
lay
s
ia
Par
it R
aja,
B
atu
Pah
at,
Ma
lay
s
ia
E
m
ail:
n
u
r
eize
@
u
th
m
.
ed
u
.
m
y
1.
I
NT
RO
D
UCT
I
O
N
I
n
ar
tific
ial
in
tellig
en
ce
(
AI
)
,
m
ac
h
in
e
lear
n
in
g
(
ML
)
f
o
cu
s
es
o
n
p
e
r
f
o
r
m
in
g
task
s
with
o
u
t
ex
p
licit
h
u
m
an
p
r
o
g
r
am
m
in
g
b
y
d
e
v
elo
p
in
g
m
o
d
els
th
at
id
en
tify
p
atter
n
s
in
d
ata,
e
n
ab
li
n
g
p
r
e
d
ictio
n
s
an
d
d
ec
is
io
n
-
m
ak
in
g
[
1
]
.
ML
e
n
h
an
ce
s
co
m
p
u
tatio
n
al
ca
p
a
b
ilit
ies,
allo
win
g
f
o
r
f
aster
an
d
m
o
r
e
ac
cu
r
ate
p
r
o
ce
s
s
in
g
o
f
lar
g
e
d
atasets
.
Key
ca
te
g
o
r
ies
o
f
ML
in
cl
u
d
e
s
u
p
er
v
is
ed
lea
r
n
in
g
,
u
n
s
u
p
er
v
is
ed
lear
n
in
g
,
s
em
i
-
s
u
p
er
v
is
ed
lear
n
in
g
,
an
d
r
ein
f
o
r
ce
m
en
t
lea
r
n
in
g
[
2
]
.
Su
p
er
v
is
ed
lear
n
in
g
u
s
es
lab
eled
d
atasets
f
o
r
tr
ain
in
g
,
en
ab
lin
g
m
o
d
els
to
p
r
e
d
ict
o
u
tco
m
es
f
o
r
u
n
s
ee
n
d
ata
[
3
]
.
T
h
is
ca
teg
o
r
y
f
u
r
th
er
d
iv
id
es
in
to
r
eg
r
ess
io
n
an
d
class
if
icatio
n
task
s
.
C
o
n
v
er
s
ely
,
u
n
s
u
p
er
v
is
ed
lear
n
in
g
w
o
r
k
s
with
u
n
lab
eled
d
atasets
,
allo
win
g
m
o
d
els to
f
in
d
p
atter
n
s
in
d
ep
en
d
en
tly
,
em
p
lo
y
in
g
t
ec
h
n
iq
u
es lik
e
clu
s
ter
in
g
an
d
ass
o
ciatio
n
[
4
]
.
A
r
o
b
u
s
t
d
ataset
is
ess
en
tial
f
o
r
ef
f
ec
tiv
e
ML
,
co
m
p
r
is
in
g
s
tr
u
ctu
r
ed
an
d
u
n
s
tr
u
ctu
r
ed
d
ata
th
at
s
er
v
es
as
in
p
u
t
f
o
r
tr
ai
n
in
g
m
o
d
els
[
5
]
.
Data
s
ets
ca
n
b
e
cla
s
s
if
ied
as
b
alan
ce
d
o
r
im
b
ala
n
ce
d
b
ased
o
n
class
r
ep
r
esen
tatio
n
[
6
]
.
B
alan
ce
d
d
atasets
h
av
e
eq
u
al
r
ep
r
esen
tati
o
n
ac
r
o
s
s
class
es,
wh
ile
im
b
al
an
ce
d
d
atasets
m
ay
lead
to
u
n
d
er
p
er
f
o
r
m
i
n
g
m
o
d
els
o
n
m
in
o
r
ity
class
es.
A
d
d
r
ess
in
g
th
is
im
b
alan
ce
is
v
ital
f
o
r
ac
cu
r
ate
p
r
ed
ictio
n
s
,
with
m
eth
o
d
s
s
u
c
h
as
r
esam
p
lin
g
,
w
h
er
e
m
in
o
r
ity
class
es
ar
e
o
v
er
s
am
p
led
,
a
n
d
m
ajo
r
ity
class
es
u
n
d
er
s
am
p
le
d
[
7
]
.
T
h
is
s
tu
d
y
ev
alu
ates
class
im
b
alan
ce
m
itig
atio
n
tech
n
i
q
u
es
in
cr
e
d
it
ca
r
d
f
r
a
u
d
d
etec
tio
n
.
I
t
an
aly
ze
s
p
er
f
o
r
m
an
ce
m
et
r
ics
—
ac
cu
r
ac
y
,
p
r
ec
is
io
n
,
r
e
ca
ll,
an
d
F1
-
s
co
r
e
—
ac
r
o
s
s
v
ar
io
u
s
r
esam
p
lin
g
r
atio
s
,
wh
ile
ad
d
r
ess
in
g
th
e
r
is
k
s
o
f
o
v
er
f
itti
n
g
ass
o
ciate
d
with
s
y
n
th
etic
o
v
er
s
am
p
li
n
g
.
T
h
e
g
o
al
is
to
im
p
r
o
v
e
ML
m
o
d
el
r
eliab
ilit
y
in
im
b
alan
ce
d
s
ce
n
a
r
io
s
,
p
ar
ti
cu
lar
ly
in
f
r
au
d
d
etec
tio
n
a
p
p
l
icatio
n
s
.
T
h
e
s
tr
u
ctu
r
e
o
f
th
is
ar
ticle
is
as
f
o
llo
ws
.
Sectio
n
1
p
r
esen
ts
b
ac
k
g
r
o
u
n
d
o
n
im
b
alan
ce
d
d
atasets
.
Sectio
n
2
co
v
er
s
f
o
u
n
d
atio
n
a
l
co
n
ce
p
ts
in
ML.
Sectio
n
3
r
ev
iews
s
tu
d
ies
o
n
ad
d
r
ess
in
g
class
im
b
alan
ce
s
.
Sectio
n
4
p
r
esen
ts
ex
p
er
im
en
tal
r
esu
lts
o
f
th
e
class
if
icatio
n
m
o
d
els
u
n
d
er
d
if
f
e
r
en
t
r
esa
m
p
lin
g
tech
n
iq
u
es
.
L
astl
y
,
s
ec
tio
n
5
p
r
o
v
id
es a
co
n
clu
s
io
n
.
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
Th
e
effec
ts
o
f d
a
ta
imb
a
la
n
ce
o
n
fr
a
u
d
d
etec
tio
n
mo
d
el
a
cc
u
r
a
cy
(
R
u
s
ma
A
n
ieza
R
u
s
la
n
)
1403
2.
RE
L
AT
E
D
WO
RK
2
.
1
.
Da
t
a
s
et
s
in m
a
chine le
a
rning
I
n
M
L
,
d
a
tase
ts
ar
e
s
tr
u
ct
u
r
e
d
c
o
ll
ec
ti
o
n
s
o
f
i
n
f
o
r
m
ati
o
n
u
s
e
d
f
o
r
test
i
n
g
,
v
a
li
d
at
in
g
,
a
n
d
t
r
a
i
n
i
n
g
m
o
d
els.
T
h
e
y
e
n
a
b
l
e
ML
al
g
o
r
it
h
m
s
t
o
i
d
en
tif
y
p
a
tte
r
n
s
,
l
ea
d
i
n
g
to
ac
c
u
r
a
te
p
r
ed
ict
io
n
s
.
T
y
p
ic
all
y
,
d
atas
ets
ar
e
o
r
g
a
n
iz
e
d
i
n
ta
b
u
la
r
f
o
r
m
at
,
w
h
e
r
e
r
o
ws
r
e
p
r
ese
n
t
i
n
d
i
v
i
d
u
al
s
a
m
p
les
an
d
c
o
l
u
m
n
s
s
i
g
n
if
y
a
tt
r
i
b
u
tes
ass
o
cia
te
d
w
it
h
t
h
o
s
e
s
am
p
l
es
[
8
]
.
T
h
e
q
u
al
it
y
o
f
d
at
asets
i
s
cr
u
cia
l
f
o
r
p
e
r
f
o
r
m
a
n
c
e
an
d
g
e
n
er
ali
za
b
il
it
y
o
f
ML
m
o
d
els;
w
ell
-
s
t
r
u
ct
u
r
e
d
d
a
tasets
a
ll
o
w
f
o
r
m
e
an
in
g
f
u
l
p
a
t
ter
n
e
x
t
r
a
cti
o
n
an
d
r
eli
a
b
le
p
r
e
d
ic
ti
o
n
s
[
9
]
,
[
1
0
]
.
Data
s
ets
ca
n
b
e
ca
teg
o
r
ized
as
lab
eled
o
r
u
n
lab
eled
.
L
a
b
e
led
d
atasets
co
n
tain
in
s
tan
ce
s
lin
k
ed
to
tar
g
et
v
alu
es,
wh
ich
ar
e
es
s
en
tial
in
s
u
p
er
v
is
ed
lear
n
i
n
g
f
o
r
class
if
y
in
g
d
ata
a
n
d
g
en
er
atin
g
p
r
ec
is
e
p
r
ed
ictio
n
s
[
1
1
]
.
I
n
c
o
n
tr
ast,
u
n
lab
eled
d
atasets
lack
p
r
ed
e
f
in
ed
lab
els,
allo
win
g
m
o
d
els
to
lear
n
p
atter
n
s
in
d
ep
en
d
en
tly
[
1
2
]
.
A
d
d
itio
n
a
lly
,
d
atasets
ca
n
b
e
s
tr
u
ctu
r
ed
,
o
r
g
an
ize
d
in
tab
les
(
e.
g
.
,
cu
s
t
o
m
er
I
D,
a
g
e,
an
d
g
en
d
er
)
,
o
r
u
n
s
tr
u
ctu
r
ed
,
i
n
clu
d
in
g
f
o
r
m
ats lik
e
tex
t,
im
a
g
es,
an
d
au
d
io
.
Data
s
ets
ar
e
f
u
r
th
er
d
iv
id
ed
i
n
to
th
r
ee
m
ain
ca
te
g
o
r
ies:
tr
ain
in
g
,
v
alid
atio
n
,
an
d
test
in
g
s
ets
[
1
3
]
.
T
h
e
tr
ain
in
g
s
et
co
n
tain
s
la
b
eled
s
am
p
les
to
ed
u
ca
te
th
e
m
o
d
el,
w
h
ile
th
e
v
alid
atio
n
s
et
h
elp
s
ad
ju
s
t
p
ar
am
eter
s
d
u
r
in
g
tr
ai
n
in
g
,
a
n
d
th
e
test
in
g
s
et
ev
al
u
ates
th
e
m
o
d
el'
s
p
er
f
o
r
m
a
n
ce
o
n
u
n
s
ee
n
d
ata.
C
o
m
m
o
n
m
etr
ics
f
o
r
ass
ess
m
en
t
in
clu
d
e
ac
cu
r
ac
y
a
n
d
er
r
o
r
r
ate
[
1
4
]
.
Pre
p
r
o
ce
s
s
in
g
is
v
ital
f
o
r
p
r
ep
ar
i
n
g
d
atasets
b
ef
o
r
e
m
o
d
el
tr
ain
i
n
g
.
T
h
is
in
clu
d
es
n
o
r
m
aliza
tio
n
,
tr
an
s
f
o
r
m
atio
n
,
a
n
d
clea
n
in
g
to
ad
d
r
e
s
s
o
u
tlier
s
,
m
is
s
in
g
v
alu
es,
o
r
n
o
is
e,
all
o
f
wh
ic
h
ca
n
n
eg
ativ
ely
im
p
ac
t
m
o
d
el
p
er
f
o
r
m
an
c
e
[
1
5
]
.
E
n
s
u
r
i
n
g
d
ata
in
teg
r
ity
is
ess
en
tial f
o
r
ac
cu
r
ate
an
d
ef
f
ic
ien
t a
n
aly
s
is
an
d
m
o
d
elin
g
.
2
.
2
.
I
m
ba
l
a
nced
da
t
a
s
et
s
I
m
b
alan
ce
s
in
d
atasets
r
ef
er
to
v
ar
iatio
n
s
in
th
e
n
u
m
b
er
o
f
in
p
u
t
s
am
p
les
ac
r
o
s
s
d
if
f
er
en
t
o
u
tp
u
t
class
es.
An
im
b
alan
ce
d
d
ataset
ex
h
ib
its
a
s
k
ewe
d
d
is
tr
ib
u
t
io
n
o
f
class
es,
p
o
s
in
g
s
ig
n
if
i
ca
n
t
ch
allen
g
es
f
o
r
ML
.
W
h
en
th
e
n
u
m
b
e
r
o
f
in
s
t
an
ce
s
in
th
e
m
i
n
o
r
ity
class
is
s
u
b
s
tan
tially
lo
wer
th
an
i
n
th
e
m
ajo
r
ity
class
,
th
e
m
o
d
el
ten
d
s
to
b
ias
p
r
ed
ictio
n
s
to
war
d
t
h
e
m
ajo
r
ity
class
.
T
h
is
ca
n
lead
to
i
n
ac
cu
r
ate
p
r
ed
ictio
n
s
an
d
p
o
o
r
p
er
f
o
r
m
an
ce
f
o
r
th
e
m
in
o
r
ity
class
[
1
6
]
,
wh
ich
is
cr
itical
in
r
ea
l
-
wo
r
ld
a
p
p
licatio
n
s
lik
e
f
r
au
d
d
etec
tio
n
an
d
m
ed
ical
d
iag
n
o
s
tics
.
Fo
r
e
x
am
p
le,
f
r
au
d
u
len
t
tr
a
n
s
ac
tio
n
s
o
f
ten
co
n
s
titu
te
a
s
m
all
f
r
ac
tio
n
o
f
to
tal
tr
an
s
ac
tio
n
s
,
m
ak
in
g
ac
cu
r
ate
id
en
tific
atio
n
ess
en
tial
to
m
itig
ate
f
in
an
cial
lo
s
s
es.
Similar
ly
,
d
iag
n
o
s
in
g
r
ar
e
m
ed
ical
co
n
d
itio
n
s
r
eq
u
ir
es
p
r
ec
is
io
n
to
en
s
u
r
e
ef
f
ec
tiv
e
p
atien
t
ca
r
e,
em
p
h
asizin
g
t
h
e
n
ee
d
to
ad
d
r
ess
d
ataset
im
b
alan
ce
s
to
av
o
id
a
d
v
er
s
e
o
u
tco
m
es.
I
m
b
alan
ce
d
d
atasets
ar
e
c
h
ar
ac
ter
ized
b
y
u
n
eq
u
al
class
r
ep
r
esen
tatio
n
,
r
esu
ltin
g
in
s
k
ewe
d
p
r
ed
ictio
n
s
b
y
al
g
o
r
ith
m
s
.
T
o
ad
d
r
ess
th
ese
d
is
p
ar
ities
,
v
ar
io
u
s
tech
n
iq
u
es
h
av
e
b
ee
n
d
e
v
elo
p
ed
,
i
n
clu
d
in
g
r
esam
p
lin
g
m
eth
o
d
s
,
a
d
v
an
ce
d
en
s
em
b
le
a
p
p
r
o
ac
h
es
an
d
c
o
s
t
-
s
en
s
itiv
e
lear
n
in
g
s
tr
ateg
ie
s
.
T
h
ese
tech
n
iq
u
es
ar
e
in
s
tr
u
m
en
tal
in
im
p
r
o
v
in
g
class
if
icatio
n
p
er
f
o
r
m
a
n
ce
in
s
ce
n
ar
io
s
wh
er
e
class
d
is
t
r
ib
u
tio
n
is
s
k
ewe
d
,
en
s
u
r
in
g
th
at
m
o
d
els ar
e
tr
ain
ed
ef
f
ec
tiv
ely
e
v
en
with
lim
ite
d
r
ep
r
esen
tatio
n
o
f
m
in
o
r
ity
cl
ass
es.
2
.
2
.
1
.
Resa
m
pli
ng
m
et
ho
d
R
esam
p
lin
g
is
a
tech
n
iq
u
e
u
s
ed
to
ad
ju
s
t
th
e
n
u
m
b
er
o
f
in
s
ta
n
ce
s
in
m
ajo
r
ity
a
n
d
m
i
n
o
r
ity
class
es
to
ad
d
r
ess
d
ata
im
b
alan
ce
s
in
ML
[
1
7
]
.
T
h
e
r
e
ar
e
two
p
r
im
ar
y
r
esam
p
lin
g
m
eth
o
d
s
:
o
v
er
s
am
p
lin
g
an
d
u
n
d
er
s
am
p
lin
g
.
O
v
er
s
am
p
lin
g
in
cr
ea
s
es
th
e
r
ep
r
esen
tatio
n
o
f
th
e
m
in
o
r
ity
class
b
y
d
u
p
licatin
g
in
s
tan
ce
s
.
W
h
ile
th
is
m
eth
o
d
is
u
s
ef
u
l,
it
m
ay
lead
to
o
v
er
f
itti
n
g
as
it
d
o
es
n
o
t
in
tr
o
d
u
ce
n
ew
in
f
o
r
m
atio
n
.
Alter
n
ativ
es
to
s
im
p
le
o
v
er
s
am
p
lin
g
in
clu
d
e
d
ata
au
g
m
en
tatio
n
,
a
d
ap
ti
v
e
s
y
n
th
etic
s
am
p
lin
g
(
ADA
SYN)
,
an
d
s
y
n
th
etic
m
in
o
r
ity
o
v
e
r
s
am
p
lin
g
tech
n
iq
u
e
(
SMOT
E
)
.
SMO
T
E
g
en
e
r
ates
n
ew
s
y
n
t
h
eti
c
s
am
p
l
es
f
o
r
th
e
m
i
n
o
r
i
ty
cl
ass
,
en
s
u
r
i
n
g
a
d
is
t
r
i
b
u
ti
o
n
a
cr
o
s
s
t
h
e
f
e
at
u
r
e
s
p
a
ce
;
h
o
w
ev
e
r
,
it c
an
cr
ea
te
d
e
n
s
e
cl
u
s
t
er
s
o
f
s
y
n
t
h
et
ic
s
am
p
l
es
[
1
8
]
.
C
o
n
v
er
s
ely
,
ADASYN
s
p
ec
if
i
ca
lly
g
en
er
ates sy
n
th
etic
s
am
p
les f
o
r
h
ar
d
-
to
-
class
if
y
m
in
o
r
ity
in
s
tan
ce
s
,
wh
ich
h
elp
s
en
h
an
ce
d
ec
is
io
n
-
m
ak
i
n
g
b
y
s
itu
atin
g
n
ew
s
am
p
les clo
s
er
to
th
e
d
ec
is
io
n
b
o
u
n
d
ar
y
[
1
9
]
.
Un
d
er
s
am
p
lin
g
r
e
d
u
ce
s
th
e
n
u
m
b
er
o
f
in
s
tan
ce
s
in
th
e
m
ajo
r
ity
class
,
wh
ich
ca
n
p
o
s
e
a
r
is
k
o
f
lo
s
in
g
v
alu
ab
le
d
ata.
T
ec
h
n
iq
u
es
lik
e
n
ea
r
-
m
is
s
u
n
d
er
s
am
p
l
in
g
s
elec
t
m
ajo
r
ity
class
in
s
tan
ce
s
b
ased
o
n
t
h
eir
p
r
o
x
im
ity
to
m
in
o
r
ity
class
ex
am
p
les,
u
s
in
g
E
u
clid
ea
n
d
is
tan
ce
to
im
p
r
o
v
e
th
e
b
alan
ce
.
R
esam
p
lin
g
tech
n
iq
u
es
h
a
v
e
b
ee
n
ef
f
ec
t
iv
ely
ap
p
lied
ac
r
o
s
s
v
ar
io
u
s
d
o
m
ain
s
,
p
ar
ticu
lar
ly
in
class
if
icatio
n
task
s
,
d
em
o
n
s
tr
atin
g
th
eir
ca
p
ab
ilit
y
to
en
h
an
ce
class
if
icatio
n
o
u
t
co
m
es.
T
ab
le
1
s
u
m
m
ar
izes
n
o
tab
le
ap
p
licatio
n
s
o
f
r
esam
p
lin
g
tech
n
iq
u
es,
h
ig
h
lig
h
tin
g
th
eir
co
n
tr
i
b
u
tio
n
s
a
n
d
k
ey
f
in
d
i
n
g
s
.
2
.
2
.
2
.
E
ns
em
ble m
et
ho
ds
E
n
s
em
b
l
e
m
e
th
o
d
s
p
r
o
v
id
e
e
f
f
e
cti
v
e
s
t
r
at
e
g
y
f
o
r
a
d
d
r
ess
i
n
g
cl
ass
im
b
a
la
n
c
e
b
y
c
o
m
b
i
n
i
n
g
m
u
lt
ip
le
b
as
eli
n
e
m
o
d
els
t
o
c
r
e
at
e
m
o
r
e
r
o
b
u
s
t
o
v
e
r
al
l
m
o
d
e
l
[
2
0
]
.
T
h
is
ap
p
r
o
ac
h
ca
n
e
n
h
an
ce
cl
as
s
if
i
ca
t
io
n
ac
cu
r
a
c
y
,
esp
ec
i
all
y
in
d
atas
ets
wit
h
i
m
b
ala
n
ce
d
class
es.
Ke
y
en
s
e
m
b
le
tec
h
n
i
q
u
es
in
cl
u
d
e
b
o
o
s
tin
g
an
d
b
ag
g
i
n
g
.
B
o
o
s
ti
n
g
i
n
v
o
l
v
es
t
r
a
in
in
g
w
e
ak
m
o
d
els
s
e
q
u
e
n
ti
all
y
,
wit
h
e
ac
h
s
u
cc
ess
iv
e
m
o
d
el
f
o
c
u
s
i
n
g
o
n
t
h
e
e
r
r
o
r
s
m
a
d
e
b
y
its
p
r
e
d
e
ce
s
s
o
r
.
A
n
ex
a
m
p
l
e
o
f
t
h
is
te
ch
n
i
q
u
e
is
A
d
aB
o
o
s
t,
w
h
i
ch
a
d
j
u
s
ts
w
ei
g
h
ts
o
f
m
is
class
i
f
i
ed
i
n
s
ta
n
c
es
to
i
m
p
r
o
v
e
f
u
t
u
r
e
p
r
ed
ict
io
n
s
[
2
1
]
.
B
ag
g
i
n
g
,
g
en
e
r
at
es
m
u
l
ti
p
le
s
u
b
s
e
ts
f
o
r
t
h
e
t
r
ai
n
i
n
g
d
at
a
b
y
s
a
m
p
li
n
g
wi
th
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
.
2
,
Ap
r
il
20
26
:
1
4
0
2
-
1
4
0
8
1404
r
e
p
l
ac
e
m
e
n
t
,
w
h
ic
h
h
el
p
s
r
e
d
u
c
e
p
r
e
d
ic
ti
o
n
v
a
r
i
an
ce
.
T
h
is
t
ec
h
n
iq
u
e
ag
g
r
e
g
a
tes
p
r
e
d
i
cti
o
n
s
th
r
o
u
g
h
m
aj
o
r
i
ty
v
o
ti
n
g
t
o
p
r
o
d
u
ce
a
f
i
n
al
o
u
tp
u
t
[
2
2
]
.
A
n
ex
am
p
l
e
o
f
e
n
s
e
m
b
le
m
et
h
o
d
s
i
n
a
cti
o
n
is
p
r
ese
n
te
d
b
y
u
til
izi
n
g
a
n
en
s
em
b
l
e
c
o
n
v
o
lu
ti
o
n
al
n
e
u
r
a
l
n
etw
o
r
k
(
E
n
C
NN
)
a
p
p
r
o
ac
h
f
o
r
i
n
te
lli
g
e
n
t
f
au
lt
d
i
ag
n
o
s
is
in
m
ac
h
i
n
es
u
n
d
e
r
im
b
al
an
ce
d
c
o
n
d
it
io
n
s
.
T
h
is
m
et
h
o
d
c
o
m
b
i
n
es
s
e
v
e
r
al
class
i
f
i
er
s
tr
ai
n
e
d
o
n
b
ala
n
c
ed
s
u
b
s
ets
o
f
t
h
e
im
b
al
an
ce
d
d
at
aset
,
em
p
l
o
y
i
n
g
we
ig
h
t
ed
v
o
ti
n
g
f
o
r
f
in
al
p
r
ed
icti
o
n
s
to
i
m
p
r
o
v
e
ac
c
u
r
ac
y
a
n
d
r
o
b
u
s
t
n
ess
[
2
3
]
.
2
.
2
.
3
.
Co
s
t
-
s
ens
it
iv
e
lea
rning
C
o
s
t
-
s
en
s
it
iv
e
lear
n
in
g
tech
n
iq
u
es
ad
ju
s
t
th
e
co
s
ts
ass
o
cia
ted
with
m
is
cla
s
s
if
y
in
g
in
s
ta
n
ce
s
f
r
o
m
m
ajo
r
ity
an
d
m
in
o
r
ity
class
es
to
a
d
d
r
ess
class
im
b
alan
ce
s
ef
f
ec
tiv
ely
[
2
4
]
.
I
n
th
is
f
r
am
e
wo
r
k
,
t
h
e
m
ajo
r
ity
class
u
s
u
ally
in
cu
r
s
lo
wer
c
o
s
ts
,
wh
ile
th
e
m
in
o
r
ity
class
is
ass
o
ciate
d
with
h
ig
h
e
r
co
s
ts
.
T
h
is
ap
p
r
o
ac
h
c
an
b
e
ca
teg
o
r
ized
in
t
o
f
o
u
r
ty
p
es
:
i)
f
alse
p
o
s
itiv
e
(
FP
)
:
t
h
e
c
o
s
t
o
f
in
co
r
r
ec
tly
class
if
y
in
g
a
p
o
s
itiv
e
in
s
tan
ce
as
n
eg
ativ
e
,
ii)
f
alse
n
eg
ativ
e
(
FN)
:
t
h
e
co
s
t
o
f
in
co
r
r
ec
tly
cla
s
s
if
y
in
g
a
n
eg
ativ
e
in
s
tan
ce
a
s
p
o
s
itiv
e
,
iii)
t
r
u
e
p
o
s
itiv
e
(
T
P):
t
h
e
ac
c
u
r
ate
c
lass
if
icatio
n
o
f
a
p
o
s
itiv
e
in
s
tan
ce
,
an
d
i
v
)
t
r
u
e
n
eg
ativ
e
(
T
N)
:
t
h
e
ac
cu
r
ate
class
if
icatio
n
o
f
a
n
eg
ativ
e
in
s
tan
ce
.
I
n
f
r
a
u
d
d
etec
tio
n
,
f
o
r
in
s
tan
ce
,
m
is
lab
elin
g
leg
itima
te
tr
a
n
s
ac
tio
n
s
as
f
r
au
d
u
le
n
t
(
FP
)
ca
n
lead
to
f
in
an
cial
lo
s
s
es,
wh
ile
m
is
cl
ass
if
y
in
g
f
r
au
d
u
len
t
tr
an
s
ac
ti
o
n
s
as
leg
itima
te
(
F
N)
ca
n
r
esu
lt
in
cu
s
to
m
er
d
is
s
atis
f
ac
tio
n
.
T
h
e
ty
p
ically
s
m
all
p
r
o
p
o
r
tio
n
o
f
f
r
a
u
d
u
le
n
t tr
an
s
ac
tio
n
s
co
n
tr
ib
u
tes to
h
ig
h
ly
s
k
ewe
d
d
ataset.
An
illu
s
tr
ativ
e
ex
am
p
le
o
f
c
o
s
t
-
s
en
s
itiv
e
lear
n
in
g
is
th
e
d
ev
elo
p
m
e
n
t
o
f
co
s
t
-
s
en
s
itiv
e
f
ea
tu
r
e
s
elec
tio
n
g
en
er
al
v
ec
to
r
m
ac
h
in
e
(
C
FGVM)
alg
o
r
ith
m
.
T
h
is
alg
o
r
ith
m
in
teg
r
ates
g
en
er
al
v
ec
to
r
m
ac
h
in
e
with
b
in
ar
y
an
t
lio
n
o
p
tim
izer
to
e
n
h
an
ce
th
e
p
er
f
o
r
m
a
n
ce
o
f
im
b
alan
ce
d
class
if
icatio
n
task
s
.
T
h
is
ap
p
r
o
ac
h
u
n
d
er
s
co
r
es
th
e
im
p
o
r
tan
ce
o
f
a
d
ju
s
tin
g
co
s
ts
to
im
p
r
o
v
e
p
r
ed
ictiv
e
ac
cu
r
ac
y
in
s
ce
n
ar
io
s
with
im
b
alan
ce
d
d
ata
[
2
5
]
.
T
ab
le
1
.
Ap
p
licatio
n
o
f
r
esam
p
lin
g
tech
n
i
q
u
es
R
e
f
e
r
e
n
c
e
C
o
n
t
r
i
b
u
t
i
o
n
F
i
n
d
i
n
g
s
[
2
6
]
A
d
d
r
e
ss
e
s
c
l
a
ss i
mb
a
l
a
n
c
e
i
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
,
i
mp
a
c
t
i
n
g
t
h
e
e
f
f
e
c
t
i
v
e
n
e
ss
o
f
c
l
a
ss
i
f
i
c
a
t
i
o
n
t
e
c
h
n
i
q
u
e
s
C
l
a
s
si
f
i
c
a
t
i
o
n
m
o
d
e
l
s
d
e
mo
n
s
t
r
a
t
e
d
st
a
t
i
s
t
i
c
a
l
l
y
s
i
g
n
i
f
i
c
a
n
t
i
mp
r
o
v
e
me
n
t
s
o
v
e
r
t
h
e
i
n
i
t
i
a
l
i
m
b
a
l
a
n
c
e
d
d
a
t
a
se
t
[
2
7
]
I
n
t
r
o
d
u
c
e
s a
u
n
i
q
u
e
a
p
p
r
o
a
c
h
c
o
mb
i
n
i
n
g
u
n
d
e
r
sam
p
l
i
n
g
a
n
d
o
v
e
r
sam
p
l
i
n
g
me
t
h
o
d
s
S
i
g
n
i
f
i
c
a
n
t
i
m
p
r
o
v
e
me
n
t
i
n
c
l
a
s
si
f
i
e
r
p
e
r
f
o
r
ma
n
c
e
,
p
a
r
t
i
c
u
l
a
r
l
y
i
n
d
e
t
e
c
t
i
n
g
r
a
r
e
e
v
e
n
t
s
[
2
8
]
D
e
v
e
l
o
p
s
mo
d
e
l
s
f
o
r
a
u
t
o
ma
t
i
c
d
e
t
e
r
mi
n
a
t
i
o
n
o
f
e
f
f
e
c
t
i
v
e
r
e
sam
p
l
i
n
g
t
e
c
h
n
i
q
u
e
s
b
a
se
d
o
n
d
a
t
a
s
e
t
p
r
o
p
e
r
t
i
e
s
Th
e
e
f
f
i
c
i
e
n
c
y
o
f
o
v
e
r
sa
mp
l
i
n
g
a
n
d
u
n
d
e
r
sam
p
l
i
n
g
met
h
o
d
s
v
a
r
i
e
s
b
a
se
d
o
n
t
h
e
i
mb
a
l
a
n
c
e
r
a
t
i
o
a
n
d
t
h
e
d
a
t
a
se
t
c
h
a
r
a
c
t
e
r
i
st
i
c
s
[
2
9
]
S
t
u
d
i
e
s
t
h
e
i
mp
a
c
t
o
f
r
e
s
a
mp
l
i
n
g
o
n
A
N
N
c
l
a
ss
i
f
i
e
r
p
e
r
f
o
r
m
a
n
c
e
i
n
n
e
t
w
o
r
k
i
n
t
r
u
si
o
n
d
e
t
e
c
t
i
o
n
En
h
a
n
c
e
d
i
d
e
n
t
i
f
i
c
a
t
i
o
n
o
f
mi
n
o
r
i
t
y
-
c
l
a
ss
d
a
t
a
t
h
r
o
u
g
h
r
e
samp
l
i
n
g
[
3
0
]
I
n
v
e
st
i
g
a
t
e
s
d
i
s
t
i
n
c
t
i
o
n
s
i
n
c
l
a
ssi
f
i
c
a
t
i
o
n
e
f
f
i
c
a
c
y
u
s
i
n
g
v
a
r
i
o
u
s res
a
m
p
l
i
n
g
s
t
r
a
t
e
g
i
e
s
RF
c
l
a
ssi
f
i
e
r
o
u
t
p
e
r
f
o
r
me
d
o
t
h
e
r
s
w
h
e
n
e
n
h
a
n
c
e
d
w
i
t
h
t
h
e
S
V
M
-
S
M
O
TE
me
t
h
o
d
2
.
3
.
M
a
chine le
a
rning
m
o
de
ls
C
lass
if
ie
r
s
a
r
e
m
o
d
els
t
h
at
ass
ig
n
s
p
ec
if
ic
class
if
ica
ti
o
n
s
t
o
i
n
p
u
t
d
at
a,
w
it
h
a
p
p
li
ca
t
io
n
s
in
ar
ea
s
li
k
e
f
r
a
u
d
d
e
te
cti
o
n
.
St
a
n
d
ar
d
class
if
i
er
s
o
f
te
n
s
t
r
u
g
g
l
e
wi
th
im
b
a
lan
ce
d
d
at
ase
ts
,
p
r
o
m
p
ti
n
g
r
es
ea
r
c
h
e
r
s
t
o
d
e
v
el
o
p
tec
h
n
i
q
u
es
to
en
h
an
ce
m
o
d
e
l
r
o
b
u
s
t
n
ess
a
n
d
ac
c
u
r
ac
y
.
T
h
is
r
ese
ar
ch
f
o
c
u
s
es
o
n
t
wo
e
f
f
e
cti
v
e
c
lass
i
f
i
ca
ti
o
n
tec
h
n
i
q
u
es:
r
a
n
d
o
m
f
o
r
est
(
R
F
)
a
n
d
g
r
a
d
ie
n
t
b
o
o
s
ti
n
g
(
GB
)
.
B
o
t
h
ar
e
a
d
e
p
t
a
t
h
an
d
l
in
g
c
o
m
p
le
x
d
atas
ets
a
n
d
im
b
al
an
ce
d
cl
ass
d
is
t
r
i
b
u
ti
o
n
s
.
R
F
e
m
p
lo
y
s
a
n
en
s
em
b
l
e
ap
p
r
o
ac
h
to
i
m
p
r
o
v
e
ac
cu
r
ac
y
an
d
r
e
d
u
c
e
o
v
e
r
f
itt
in
g
,
wh
i
le
GB
en
h
an
ce
s
p
r
ed
icti
v
e
p
e
r
f
o
r
m
a
n
c
e
t
h
r
o
u
g
h
ite
r
a
ti
v
e
lea
r
n
i
n
g
.
B
y
f
o
c
u
s
i
n
g
o
n
t
h
ese
cl
ass
i
f
ie
r
s
,
w
e
ai
m
to
le
v
e
r
ag
e
th
ei
r
s
t
r
en
g
t
h
s
t
o
ta
ck
l
e
th
e
c
h
al
le
n
g
es
p
o
s
e
d
b
y
s
k
ew
e
d
t
a
r
g
et
cl
ass
d
is
t
r
i
b
u
ti
o
n
.
3.
M
E
T
H
O
D
3
.
1
.
Da
t
a
g
a
t
hering
a
nd
da
t
a
s
pli
t
t
ing
T
h
is
s
tu
d
y
an
aly
ze
s
a
d
ataset
o
f
cr
e
d
it
ca
r
d
tr
an
s
ac
tio
n
s
f
r
o
m
E
u
r
o
p
ea
n
ca
r
d
h
o
ld
er
s
i
n
Sep
tem
b
er
2
0
1
3
,
co
m
p
r
is
in
g
2
8
4
,
8
0
7
tr
an
s
ac
tio
n
s
,
o
f
wh
ich
4
9
2
we
r
e
f
r
a
u
d
u
le
n
t.
Du
e
to
th
e
s
ig
n
if
ican
t
im
b
ala
n
ce
b
etwe
en
n
o
n
-
f
r
a
u
d
u
le
n
t
an
d
f
r
au
d
u
len
t
tr
an
s
ac
tio
n
s
,
th
e
s
tu
d
y
f
o
cu
s
es
o
n
r
esam
p
lin
g
t
h
e
class
es
at
r
atio
s
o
f
2
0
:8
0
,
3
0
:7
0
,
an
d
4
0
:
6
0
.
T
h
ese
r
atio
s
en
ab
le
a
co
m
p
r
eh
en
s
iv
e
ev
alu
atio
n
o
f
h
o
w
d
if
f
er
en
t
le
v
els
o
f
r
esam
p
lin
g
af
f
ec
t
m
o
d
el
p
er
f
o
r
m
an
ce
.
Data
p
r
ep
r
o
ce
s
s
in
g
in
clu
d
e
d
s
ev
er
al
cr
itical
s
tep
s
to
en
s
u
r
e
in
p
u
t
q
u
ality
.
Fea
tu
r
es
wer
e
s
tan
d
ar
d
ized
to
h
av
e
a
m
ea
n
o
f
ze
r
o
an
d
a
s
tan
d
ar
d
d
ev
iatio
n
o
f
o
n
e,
en
h
an
cin
g
t
h
e
p
er
f
o
r
m
an
ce
o
f
alg
o
r
ith
m
s
s
en
s
itiv
e
to
f
ea
tu
r
e
s
ca
lin
g
.
T
em
p
o
r
al
p
atter
n
s
wer
e
ca
p
tu
r
e
d
b
y
en
g
in
ee
r
in
g
ti
m
e
-
r
elate
d
f
ea
tu
r
es,
an
d
an
o
n
y
m
ized
V
-
f
ea
tu
r
es,
cr
ea
ted
th
r
o
u
g
h
p
r
i
n
cip
al
co
m
p
o
n
e
n
t
an
aly
s
is
(
PC
A
)
f
o
r
co
n
f
id
en
tiality
,
wer
e
also
in
clu
d
ed
.
T
h
e
an
aly
s
is
w
as
co
n
d
u
cte
d
u
s
in
g
Py
th
o
n
,
u
t
ilizin
g
th
e
s
cik
it
-
lear
n
lib
r
ar
y
(
v
er
s
io
n
0
.
2
4
.
2
)
f
o
r
im
p
lem
en
tin
g
ML
m
o
d
els an
d
r
esam
p
lin
g
tech
n
iq
u
es.
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
Th
e
effec
ts
o
f d
a
ta
imb
a
la
n
ce
o
n
fr
a
u
d
d
etec
tio
n
mo
d
el
a
cc
u
r
a
cy
(
R
u
s
ma
A
n
ieza
R
u
s
la
n
)
1405
3.
2
.
Resa
m
pli
ng
m
et
ho
do
lo
g
y
a
nd
perf
o
rma
nce
a
s
s
ess
m
ent
T
h
is
s
tu
d
y
em
p
lo
y
ed
th
r
ee
p
r
o
m
in
e
n
t
r
esam
p
lin
g
tech
n
iq
u
es:
r
an
d
o
m
u
n
d
er
s
am
p
li
n
g
(
R
US)
,
r
an
d
o
m
o
v
e
r
s
am
p
lin
g
(
R
OS)
,
an
d
SMOT
E
.
Per
f
o
r
m
a
n
ce
e
v
alu
atio
n
s
wer
e
c
o
n
d
u
cted
f
o
r
ea
ch
tec
h
n
iq
u
e
to
ass
es
s
th
e
ef
f
ec
tiv
e
n
ess
o
f
t
h
e
class
if
icatio
n
m
eth
o
d
s
u
s
ed
,
f
o
cu
s
in
g
o
n
k
ey
class
if
icatio
n
er
r
o
r
s
:
FP
an
d
FN,
as we
ll a
s
T
P a
n
d
T
N.
C
lass
if
i
ca
tio
n
m
o
d
el
ac
c
u
r
ac
y
was c
o
m
p
ar
ed
u
s
in
g
(
1
)
.
=
+
(
+
+
+
)
(
1
)
T
o
p
r
o
v
i
d
e
a
co
m
p
r
e
h
en
s
iv
e
ass
es
s
m
en
t
o
f
m
o
d
el
p
er
f
o
r
m
a
n
ce
,
we
also
ca
lcu
lated
th
e
(
2
)
to
(
4
)
b
ased
o
n
th
e
co
n
f
u
s
io
n
m
atr
ix
.
=
+
(
2
)
=
+
(
3
)
1
−
=
2
×
×
+
(
4
)
W
h
i
l
e
m
e
t
r
i
cs
l
i
k
e
r
e
c
ei
v
e
r
o
p
er
a
t
i
n
g
c
h
a
r
a
c
t
e
r
i
s
t
i
c
(
R
OC
)
-
a
r
e
a
u
n
d
e
r
t
h
e
c
u
r
v
e
(
A
UC
)
a
n
d
M
a
t
t
h
e
ws
c
o
r
r
e
l
a
t
i
o
n
c
o
e
f
f
ic
i
e
n
t
(
M
C
C
)
a
r
e
v
a
l
u
a
b
l
e
,
t
h
i
s
s
t
u
d
y
f
o
c
u
s
ed
o
n
a
c
c
u
r
a
c
y
,
p
r
e
c
is
i
o
n
,
r
e
c
al
l
,
a
n
d
F
1
-
s
c
o
r
e
t
o
p
r
o
v
i
d
e
a
c
l
e
a
r
a
s
s
ess
m
e
n
t
o
f
m
o
d
e
l
p
e
r
f
o
r
m
a
n
c
e
i
n
f
r
a
u
d
d
e
t
e
c
t
i
o
n
.
A
c
c
u
r
a
c
y
o
f
f
e
r
s
a
s
t
r
ai
g
h
t
f
o
r
w
a
r
d
p
e
r
f
o
r
m
a
n
c
e
m
e
a
s
u
r
e
,
w
h
i
l
e
p
r
e
c
i
s
i
o
n
a
n
d
r
e
c
a
l
l
h
e
l
p
u
n
d
e
r
s
t
a
n
d
t
h
e
b
a
l
a
n
c
e
b
e
tw
e
e
n
c
o
r
r
e
c
t
l
y
i
d
e
n
t
i
f
ie
d
f
r
a
u
d
u
l
e
n
t
c
a
s
es
a
n
d
m
i
s
c
l
as
s
i
f
i
c
a
t
i
o
n
s
.
T
h
e
F
1
-
s
c
o
r
e
,
a
s
t
h
e
h
a
r
m
o
n
i
c
m
e
a
n
o
f
p
r
e
c
is
i
o
n
a
n
d
r
e
c
a
l
l
,
e
f
f
e
c
t
i
v
el
y
c
a
p
t
u
r
e
s
t
h
e
t
r
a
d
e
-
o
f
f
b
e
t
w
e
e
n
t
h
e
s
e
tw
o
m
e
t
r
i
cs
,
w
h
i
c
h
i
s
e
s
p
e
c
i
a
l
l
y
r
el
e
v
a
n
t
i
n
c
o
n
t
e
x
t
o
f
im
b
a
l
a
n
c
e
d
d
a
t
as
e
ts
.
4.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
T
h
is
s
t
u
d
y
an
al
y
z
es
th
e
p
er
f
o
r
m
a
n
c
e
o
f
v
a
r
i
o
u
s
r
esa
m
p
li
n
g
t
ec
h
n
iq
u
es
—
R
US,
R
OS
,
an
d
S
MO
T
E
—
ac
r
o
s
s
t
h
r
e
e
r
ati
o
s
:
2
0
:
8
0
,
3
0
:7
0
,
an
d
4
0
:
6
0
.
I
n
2
0
:
8
0
r
a
ti
o
as
s
h
o
wn
in
T
a
b
l
e
2
,
R
F
a
ch
ie
v
e
d
9
7
.
3
6
%
ac
c
u
r
a
cy
b
u
t
h
a
d
l
o
w
p
r
e
cisi
o
n
(
0
.
0
5
7
9
)
a
n
d
F
1
-
s
co
r
e
(
0
.
1
0
9
1
)
,
w
h
ile
GB
h
ad
a
s
i
m
i
la
r
a
cc
u
r
ac
y
o
f
9
6
.
1
5
%
wit
h
a
p
r
ec
is
i
o
n
o
f
0
.
0
4
0
3
.
T
h
e
R
OS
ap
p
r
o
ac
h
s
h
o
w
e
d
s
tr
o
n
g
p
er
f
o
r
m
an
ce
,
wit
h
R
F
a
ch
ie
v
i
n
g
9
9
.
9
6
%
a
cc
u
r
ac
y
a
n
d
9
7
.
4
4
%
p
r
ec
is
io
n
,
t
h
o
u
g
h
its
r
ec
a
ll
d
r
o
p
p
e
d
to
7
7
.
5
5
%
,
r
es
u
l
tin
g
in
a
n
F
1
-
s
c
o
r
e
o
f
0
.
8
6
3
6
;
GB
s
h
o
w
ed
9
9
.
3
4
%
ac
c
u
r
a
cy
b
u
t
lo
w
e
r
p
r
ec
is
i
o
n
(
1
9
.
6
1
%
)
.
W
it
h
SMO
T
E
,
R
F
m
ai
n
t
ai
n
e
d
h
i
g
h
a
cc
u
r
ac
y
(
9
9
.
9
5
%
)
a
n
d
g
o
o
d
p
r
ec
is
i
o
n
(
8
6
.
6
0
%
)
al
o
n
g
s
i
d
e
a
r
e
ca
l
l
o
f
8
5
.
7
1
%
,
le
ad
in
g
t
o
a
n
F1
-
s
co
r
e
o
f
0
.
8
6
1
5
,
w
h
il
e
GB
'
s
p
e
r
f
o
r
m
a
n
ce
was a
d
e
q
u
at
e,
a
ch
ie
v
i
n
g
2
1
.
3
1
%
p
r
e
cisi
o
n
.
I
n
3
0
:
7
0
r
a
ti
o
as sh
o
wn
i
n
T
ab
le
3
,
R
F a
g
ai
n
d
e
m
o
n
s
t
r
a
te
d
9
7
.
4
1
%
ac
c
u
r
a
cy
y
et
l
o
w
p
r
ec
is
i
o
n
(
0
.
0
5
4
5
)
a
n
d
a
n
F
1
-
s
c
o
r
e
o
f
0
.
1
0
2
9
,
wh
ile
GB
'
s
p
r
ec
is
i
o
n
f
e
ll
t
o
0
.
0
3
4
7
.
R
OS
r
esu
lte
d
i
n
R
F
ac
h
i
e
v
i
n
g
9
9
.
9
6
%
ac
c
u
r
ac
y
wit
h
a
p
r
e
ci
s
io
n
o
f
9
4
.
8
3
%
an
d
a
n
F
1
-
s
c
o
r
e
o
f
0
.
8
7
3
0
,
w
h
il
e
GB
p
e
r
f
o
r
m
e
d
l
o
we
r
wit
h
a
p
r
ec
is
i
o
n
o
f
1
5
.
1
4
%.
SM
OT
E
y
i
eld
e
d
9
9
.
9
5
%
a
cc
u
r
ac
y
an
d
8
3
.
1
0
%
p
r
ec
is
io
n
f
o
r
R
F,
le
ad
in
g
t
o
a
n
F
1
-
s
c
o
r
e
o
f
0
.
8
4
8
9
;
GB
r
ec
o
r
d
ed
9
9
.
3
5
%
a
cc
u
r
a
cy
w
it
h
l
o
we
r
m
et
r
i
cs.
I
n
4
0
:
6
0
r
ati
o
as
s
h
o
wn
i
n
T
a
b
le
4
,
R
F
r
e
ac
h
ed
9
8
.
0
5
%
ac
c
u
r
a
cy
b
u
t
r
ep
o
r
te
d
l
o
w
p
r
e
cisi
o
n
(
0
.
0
7
2
7
)
an
d
an
F1
-
s
c
o
r
e
o
f
0
.
1
3
4
6
,
a
n
d
GB
s
h
o
w
e
d
s
i
m
il
ar
l
o
w
p
r
e
cisi
o
n
at
0
.
0
4
2
0
.
H
o
we
v
er
,
wit
h
R
OS,
R
F
m
a
in
t
ain
e
d
h
i
g
h
ac
cu
r
a
c
y
(
9
9
.
9
5
%
)
a
n
d
p
r
ec
is
io
n
(
9
3
.
7
5
%)
,
a
c
h
ie
v
i
n
g
a
n
F
1
-
s
c
o
r
e
o
f
0
.
8
5
4
7
,
w
h
il
e
GB
'
s
p
e
r
f
o
r
m
an
ce
l
a
g
g
ed
.
W
h
e
n
u
s
i
n
g
SM
OT
E
,
R
F
ac
h
i
ev
e
d
9
9
.
9
5
%
ac
cu
r
ac
y
wit
h
8
4
.
4
6
%
p
r
ec
is
i
o
n
,
c
u
lm
in
ati
n
g
i
n
an
F
1
-
s
c
o
r
e
o
f
0
.
8
4
9
0
.
T
ab
le
2.
C
o
m
p
a
r
is
o
n
p
er
f
o
r
m
an
ce
r
esu
lts
o
f
2
0
:8
0
r
atio
S
a
mp
l
i
n
g
t
e
c
h
n
i
q
u
e
M
o
d
e
l
A
c
c
u
r
a
c
y
P
r
e
c
i
s
i
o
n
R
e
c
a
l
l
F1
-
sc
o
r
e
R
a
n
d
o
m
u
n
d
e
r
sam
p
l
i
n
g
R
a
n
d
o
m f
o
r
e
s
t
0
.
9
7
3
6
0
.
0
5
7
9
0
.
9
3
8
8
0
.
1
0
9
1
G
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
0
.
9
6
1
5
0
.
0
4
0
3
0
.
9
3
8
8
0
.
0
7
7
4
R
a
n
d
o
m
o
v
e
r
sa
mp
l
i
n
g
R
a
n
d
o
m f
o
r
e
s
t
0
.
9
9
9
6
0
.
9
7
4
4
0
.
7
7
5
5
0
.
8
6
3
6
G
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
0
.
9
9
3
4
0
.
1
9
6
1
0
.
9
1
8
4
0
.
3
2
3
2
S
M
O
TE
R
a
n
d
o
m f
o
r
e
s
t
0
.
9
9
9
5
0
.
8
6
6
0
0
.
8
5
7
1
0
.
8
6
1
5
G
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
0
.
9
9
4
1
0
.
2
1
3
1
0
.
8
9
8
0
0
.
3
4
4
4
T
ab
le
3
.
C
o
m
p
a
r
is
o
n
p
er
f
o
r
m
an
ce
r
esu
lts
o
f
3
0
:7
0
r
atio
S
a
mp
l
i
n
g
t
e
c
h
n
i
q
u
e
M
o
d
e
l
A
c
c
u
r
a
c
y
P
r
e
c
i
s
i
o
n
R
e
c
a
l
l
F1
-
sc
o
r
e
R
a
n
d
o
m
u
n
d
e
r
sam
p
l
i
n
g
R
a
n
d
o
m f
o
r
e
s
t
0
.
9
7
4
1
0
.
0
5
4
5
0
.
9
3
3
8
0
.
1
0
2
9
G
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
0
.
9
5
8
6
0
.
0
3
4
7
0
.
9
3
3
8
0
.
0
6
7
0
R
a
n
d
o
m
o
v
e
r
sa
mp
l
i
n
g
R
a
n
d
o
m f
o
r
e
s
t
0
.
9
9
9
6
0
.
9
4
8
3
0
.
8
0
8
8
0
.
8
7
3
0
G
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
0
.
9
9
1
6
0
.
1
5
1
4
0
.
9
2
6
5
0
.
2
6
0
3
S
M
O
TE
R
a
n
d
o
m f
o
r
e
s
t
0
.
9
9
9
5
0
.
8
3
1
0
0
.
8
6
7
6
0
.
8
4
8
9
G
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
0
.
9
9
3
5
0
.
1
8
5
6
0
.
9
1
1
8
0
.
3
0
8
5
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
.
2
,
Ap
r
il
20
26
:
1
4
0
2
-
1
4
0
8
1406
T
ab
le
4.
C
o
m
p
a
r
is
o
n
p
er
f
o
r
m
an
ce
r
esu
lts
o
f
4
0
:6
0
r
atio
S
a
mp
l
i
n
g
t
e
c
h
n
i
q
u
e
M
o
d
e
l
A
c
c
u
r
a
c
y
P
r
e
c
i
s
i
o
n
R
e
c
a
l
l
F1
-
sc
o
r
e
R
a
n
d
o
m
u
n
d
e
r
sam
p
l
i
n
g
R
a
n
d
o
m f
o
r
e
s
t
0
.
9
8
0
5
0
.
0
7
2
7
0
.
9
0
5
8
0
.
1
3
4
6
G
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
0
.
9
6
4
8
0
.
0
4
2
0
0
.
9
1
6
2
0
.
0
8
0
2
R
a
n
d
o
m
o
v
e
r
sa
mp
l
i
n
g
R
a
n
d
o
m f
o
r
e
s
t
0
.
9
9
9
5
0
.
9
3
7
5
0
.
7
8
5
3
0
.
8
5
4
7
G
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
0
.
9
9
3
9
0
.
2
0
1
2
0
.
8
9
0
1
0
.
3
2
8
2
S
M
O
TE
R
a
n
d
o
m f
o
r
e
s
t
0
.
9
9
9
5
0
.
8
4
4
6
0
.
8
5
3
4
0
.
8
4
9
0
G
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
0
.
9
9
4
3
0
.
2
1
2
7
0
.
8
7
9
6
0
.
3
4
2
5
R
F
co
n
s
is
ten
tly
o
u
tp
er
f
o
r
m
ed
GB
ac
r
o
s
s
all
m
etr
ics
an
d
r
at
io
s
,
p
ar
ticu
lar
ly
with
R
OS
an
d
SMOT
E
,
wh
ich
s
ig
n
if
ican
tly
e
n
h
an
ce
d
R
F
'
s
ab
ilit
y
to
ac
cu
r
ately
class
if
y
f
r
au
d
u
len
t
t
r
an
s
ac
tio
n
s
.
W
h
ile
SMOT
E
o
f
f
er
s
ad
v
an
tag
es,
it
also
ca
r
r
ies
r
is
k
s
s
u
ch
as
o
v
er
f
itti
n
g
,
esp
ec
ially
with
lim
ited
d
iv
er
s
it
y
in
th
e
m
in
o
r
ity
class
[
3
1
]
.
T
h
er
ef
o
r
e
,
m
itig
atio
n
s
tr
ateg
ies,
in
clu
d
in
g
ap
p
l
y
in
g
SMOT
E
o
n
ly
to
th
e
tr
a
in
in
g
s
et
o
r
u
s
in
g
m
o
d
if
ied
v
e
r
s
io
n
s
lik
e
B
o
r
d
e
r
lin
e
-
SMOT
E
,
ca
n
h
elp
im
p
r
o
v
e
m
o
d
el
r
o
b
u
s
tn
ess
.
W
h
ile
R
F
wi
th
R
OS
an
d
SMOT
E
d
em
o
n
s
tr
ated
s
tr
o
n
g
p
er
f
o
r
m
an
ce
,
e
x
p
lo
r
i
n
g
alter
n
ativ
e
m
eth
o
d
s
an
d
tu
n
in
g
f
o
r
GB
m
ay
also
y
ield
f
av
o
r
a
b
le
r
esu
lts
,
em
p
h
asizin
g
th
e
n
ee
d
to
tailo
r
th
e
ch
o
ice
o
f
r
esam
p
lin
g
tech
n
i
q
u
e
a
n
d
class
if
ier
to
th
e
s
p
ec
if
ic
d
ataset
an
d
co
n
te
x
t f
o
r
o
p
tim
al
o
u
tco
m
es.
W
h
ile
th
is
s
tu
d
y
d
em
o
n
s
tr
at
es
th
e
ef
f
ec
tiv
e
n
ess
o
f
R
F
with
R
OS
an
d
SMOT
E
,
o
t
h
er
r
esear
ch
h
as
ex
p
lo
r
ed
d
if
f
er
en
t
ap
p
r
o
ac
h
es.
Fo
r
in
s
tan
ce
,
s
o
m
e
s
t
u
d
ies
h
av
e
f
o
u
n
d
GB
to
b
e
m
o
r
e
ef
f
ec
tiv
e
f
o
r
h
an
d
lin
g
im
b
alan
ce
d
d
ata,
p
a
r
ticu
lar
ly
wh
en
ca
r
ef
u
lly
t
u
n
e
d
.
T
h
is
is
b
ec
au
s
e
GB
f
o
cu
s
e
s
m
o
r
e
o
n
d
if
f
icu
lt
ex
am
p
les,
p
o
ten
tially
o
u
tp
e
r
f
o
r
m
in
g
R
F
in
ce
r
tain
s
ce
n
ar
io
s
.
Ad
d
itio
n
ally
,
r
esear
ch
in
d
icat
es
th
at
co
m
b
in
in
g
SMOT
E
with
RF
ca
n
ac
h
iev
e
h
ig
h
ac
cu
r
ac
y
a
n
d
F1
-
s
co
r
es
in
cr
ed
it
ca
r
d
f
r
au
d
d
etec
tio
n
[
3
2
]
,
[
3
3
]
.
Oth
er
s
tu
d
ies
s
u
g
g
est
B
o
r
d
er
lin
e
-
SMOT
E
ca
n
f
u
r
th
er
en
h
an
ce
ac
cu
r
ac
y
co
m
p
ar
e
d
to
o
th
er
o
v
er
s
am
p
lin
g
m
eth
o
d
s
[
3
4
]
,
[
3
5
]
.
T
h
ese
v
a
r
y
in
g
r
esu
lts
h
ig
h
lig
h
t
th
e
im
p
o
r
tan
ce
o
f
co
n
s
id
er
i
n
g
t
h
e
s
p
ec
if
ic
d
ataset
an
d
p
r
o
b
lem
c
o
n
tex
t w
h
e
n
s
elec
tin
g
th
e
m
o
s
t a
p
p
r
o
p
r
iate
r
esam
p
lin
g
tech
n
iq
u
e
an
d
class
if
icati
o
n
alg
o
r
ith
m
.
5.
CO
NCLU
SI
O
N
T
h
is
s
tu
d
y
h
ig
h
lig
h
ts
th
e
cr
itical
is
s
u
e
o
f
class
im
b
alan
ce
in
ML
an
d
its
im
p
ac
t
o
n
m
o
d
el
p
er
f
o
r
m
an
ce
.
T
h
e
R
F c
lass
if
ie
r
co
n
s
is
ten
tly
ac
h
iev
ed
th
e
h
ig
h
est ac
cu
r
ac
y
ac
r
o
s
s
th
r
ee
r
esam
p
lin
g
tech
n
iq
u
es
(
R
US,
R
OS
an
d
SMOT
E
)
.
No
tab
ly
,
wh
ile
SMOT
E
d
em
o
n
s
tr
ated
s
tr
o
n
g
ef
f
icac
y
,
R
OS
y
ield
ed
ev
e
n
m
o
r
e
co
m
p
ellin
g
r
esu
lts
.
T
h
ese
f
in
d
in
g
s
em
p
h
asize
th
e
n
ec
ess
ity
o
f
ca
r
e
f
u
lly
s
elec
tin
g
r
esam
p
lin
g
tech
n
iq
u
es
an
d
alg
o
r
ith
m
s
to
o
p
tim
ize
p
er
f
o
r
m
an
ce
,
p
ar
ticu
lar
ly
in
a
p
p
lica
tio
n
s
lik
e
f
r
au
d
d
etec
tio
n
an
d
m
ed
ical
d
iag
n
o
s
is
.
Desp
ite
th
e
p
r
o
m
is
in
g
r
esu
lts
,
th
er
e
ar
e
lim
itatio
n
s
to
th
is
s
t
u
d
y
.
T
h
e
p
er
f
o
r
m
an
ce
o
f
th
e
class
if
ier
s
m
ay
v
ar
y
with
d
if
f
er
en
t
d
atasets
o
r
in
r
ea
l
-
wo
r
ld
s
ce
n
ar
io
s
o
u
ts
id
e
o
f
th
e
s
co
p
e
o
f
th
is
r
esear
ch
.
Ad
d
itio
n
ally
,
th
e
s
tu
d
y
f
o
cu
s
ed
s
o
lely
o
n
R
F
an
d
GB
m
o
d
els,
leav
in
g
o
t
h
er
p
o
ten
tia
lly
ef
f
ec
tiv
e
alg
o
r
ith
m
s
u
n
e
x
p
lo
r
ed
.
Fu
t
u
r
e
wo
r
k
s
h
o
u
ld
e
x
p
lo
r
e
a
wid
er
r
an
g
e
o
f
class
if
icatio
n
alg
o
r
ith
m
s
a
n
d
co
n
s
id
er
a
d
d
itio
n
al
ev
alu
ati
o
n
m
et
r
ics,
s
u
ch
as
R
OC
-
AU
C
an
d
MCC
,
to
p
r
o
v
id
e
a
m
o
r
e
co
m
p
r
eh
e
n
s
iv
e
ass
es
s
m
en
t
o
f
m
o
d
el
p
er
f
o
r
m
an
ce
.
Fu
r
t
h
er
m
o
r
e,
ex
p
an
d
i
n
g
th
e
d
ataset
to
in
c
lu
d
e
m
o
r
e
d
iv
er
s
e
ca
s
es
ca
n
h
elp
v
alid
ate
th
e
r
o
b
u
s
tn
ess
o
f
th
e
p
r
o
p
o
s
ed
m
eth
o
d
o
l
o
g
ies.
C
o
n
tin
u
ed
r
esear
ch
in
to
ef
f
ec
tiv
e
m
eth
o
d
o
l
o
g
ies
f
o
r
m
a
n
ag
in
g
class
im
b
alan
ce
is
v
ital
f
o
r
en
h
an
cin
g
ac
cu
r
ac
y
an
d
r
eliab
ilit
y
in
r
ea
l
-
wo
r
ld
s
ce
n
ar
i
o
s
.
F
UNDING
I
NF
O
R
M
A
T
I
O
N
T
h
is
r
esear
ch
was su
p
p
o
r
ted
b
y
Un
iv
er
s
iti T
u
n
Hu
s
s
ein
On
n
Ma
lay
s
ia
th
r
o
u
g
h
GPPS
(
Vo
t Q
6
6
2
)
.
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
R
u
s
m
a
An
ieza
R
u
s
lan
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
Nu
r
eize
Ar
b
aiy
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
Pei
-
C
h
u
n
L
in
✓
✓
✓
✓
✓
✓
✓
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
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
Th
e
effec
ts
o
f d
a
ta
imb
a
la
n
ce
o
n
fr
a
u
d
d
etec
tio
n
mo
d
el
a
cc
u
r
a
cy
(
R
u
s
ma
A
n
ieza
R
u
s
la
n
)
1407
CO
NF
L
I
C
T
O
F
I
N
T
E
R
E
S
T
ST
A
T
E
M
E
NT
Au
th
o
r
s
s
tate
n
o
co
n
f
lict o
f
in
t
er
est.
DATA AV
AI
L
AB
I
L
I
T
Y
T
h
e
d
ata
th
at
s
u
p
p
o
r
t
th
e
f
in
d
in
g
s
o
f
th
is
s
tu
d
y
ar
e
o
p
en
ly
av
ailab
le
in
Kag
g
le
at
h
ttp
s
://www.
k
ag
g
le.
co
m
/co
d
e/
jan
io
b
ac
h
m
a
n
n
/cr
ed
it
-
f
r
au
d
-
d
ea
lin
g
-
with
-
im
b
alan
ce
d
-
d
atas
ets.
RE
F
E
R
E
NC
E
S
[
1
]
B
.
M
a
h
e
sh
,
“
M
a
c
h
i
n
e
l
e
a
r
n
i
n
g
a
l
g
o
r
i
t
h
ms
-
a
r
e
v
i
e
w
,
”
I
n
t
e
rn
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
f
S
c
i
e
n
c
e
a
n
d
Re
s
e
a
rc
h
,
v
o
l
.
9
,
n
o
.
1
,
p
p
.
3
8
1
–
3
8
6
,
2
0
2
0
,
d
o
i
:
1
0
.
2
1
2
7
5
/
A
R
T
2
0
2
0
3
9
9
5
.
[
2
]
T.
T.
K
h
o
e
i
a
n
d
N
.
K
a
a
b
o
u
c
h
,
“
M
a
c
h
i
n
e
l
e
a
r
n
i
n
g
:
m
o
d
e
l
s
,
c
h
a
l
l
e
n
g
e
s,
a
n
d
r
e
se
a
r
c
h
d
i
r
e
c
t
i
o
n
s
,
”
F
u
t
u
re
I
n
t
e
rn
e
t
,
v
o
l
.
1
5
,
n
o
.
1
0
,
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
f
i
1
5
1
0
0
3
3
2
.
[
3
]
S
.
R
o
y
,
T.
M
e
e
n
a
,
a
n
d
S
.
-
J.
Li
m
,
“
D
e
my
s
t
i
f
y
i
n
g
su
p
e
r
v
i
s
e
d
l
e
a
r
n
i
n
g
i
n
h
e
a
l
t
h
c
a
r
e
4
.
0
:
a
n
e
w
r
e
a
l
i
t
y
o
f
t
r
a
n
sf
o
r
mi
n
g
d
i
a
g
n
o
s
t
i
c
med
i
c
i
n
e
,
”
D
i
a
g
n
o
st
i
c
s
,
v
o
l
.
1
2
,
n
o
.
1
0
,
2
0
2
2
,
d
o
i
:
1
0
.
3
3
9
0
/
d
i
a
g
n
o
st
i
c
s
1
2
1
0
2
5
4
9
.
[
4
]
M
.
T.
A
l
m
u
q
a
t
i
,
F
.
S
i
d
i
,
S
.
N
.
M
.
R
u
m
,
M
.
Zo
l
k
e
p
l
i
,
a
n
d
I
.
I
sh
a
k
,
“
C
h
a
l
l
e
n
g
e
s
i
n
s
u
p
e
r
v
i
se
d
a
n
d
u
n
s
u
p
e
r
v
i
s
e
d
l
e
a
r
n
i
n
g
:
a
c
o
m
p
r
e
h
e
n
si
v
e
o
v
e
r
v
i
e
w
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
n
A
d
v
a
n
c
e
d
S
c
i
e
n
c
e
,
En
g
i
n
e
e
ri
n
g
a
n
d
I
n
f
o
rm
a
t
i
o
n
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
1
4
,
n
o
.
4
,
p
p
.
1
4
4
9
–
1
4
5
5
,
2
0
2
4
,
d
o
i
:
1
0
.
1
8
5
1
7
/
i
j
a
se
i
t
.
1
4
.
4
.
2
0
1
9
1
.
[
5
]
L.
Zh
a
o
,
“
E
v
e
n
t
p
r
e
d
i
c
t
i
o
n
i
n
t
h
e
b
i
g
d
a
t
a
e
r
a
:
a
sy
s
t
e
m
a
t
i
c
s
u
r
v
e
y
,
”
AC
M
C
o
m
p
u
t
i
n
g
S
u
rv
e
y
s
,
v
o
l
.
5
4
,
n
o
.
5
,
p
p
.
1
–
3
7
,
2
0
2
2
,
d
o
i
:
1
0
.
1
1
4
5
/
3
4
5
0
2
8
7
.
[
6
]
T.
P
.
P
a
g
a
n
o
e
t
a
l
.
,
“
B
i
a
s
a
n
d
u
n
f
a
i
r
n
e
ss
i
n
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
m
o
d
e
l
s
:
a
s
y
s
t
e
ma
t
i
c
r
e
v
i
e
w
o
n
d
a
t
a
s
e
t
s
,
t
o
o
l
s,
f
a
i
r
n
e
ss
me
t
r
i
c
s,
a
n
d
i
d
e
n
t
i
f
i
c
a
t
i
o
n
a
n
d
m
i
t
i
g
a
t
i
o
n
m
e
t
h
o
d
s
,
”
Bi
g
D
a
t
a
a
n
d
C
o
g
n
i
t
i
v
e
C
o
m
p
u
t
i
n
g
,
v
o
l
.
7
,
n
o
.
1
,
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
b
d
c
c
7
0
1
0
0
1
5
.
[
7
]
F
.
G
u
r
c
a
n
a
n
d
A
.
S
o
y
l
u
,
“
L
e
a
r
n
i
n
g
f
r
o
m
i
mb
a
l
a
n
c
e
d
d
a
t
a
:
i
n
t
e
g
r
a
t
i
o
n
o
f
a
d
v
a
n
c
e
d
r
e
sam
p
l
i
n
g
t
e
c
h
n
i
q
u
e
s
a
n
d
mac
h
i
n
e
l
e
a
r
n
i
n
g
mo
d
e
l
s
f
o
r
e
n
h
a
n
c
e
d
c
a
n
c
e
r
d
i
a
g
n
o
si
s
a
n
d
p
r
o
g
n
o
si
s,
”
C
a
n
c
e
rs
,
v
o
l
.
1
6
,
n
o
.
1
9
,
2
0
2
4
,
d
o
i
:
1
0
.
3
3
9
0
/
c
a
n
c
e
r
s
1
6
1
9
3
4
1
7
.
[
8
]
N
.
H
o
l
l
m
a
n
n
e
t
a
l
.
,
“
A
c
c
u
r
a
t
e
p
r
e
d
i
c
t
i
o
n
s
o
n
sma
l
l
d
a
t
a
w
i
t
h
a
t
a
b
u
l
a
r
f
o
u
n
d
a
t
i
o
n
mo
d
e
l
,
”
N
a
t
u
r
e
,
v
o
l
.
6
3
7
,
n
o
.
8
0
4
5
,
p
p
.
3
1
9
–
3
2
6
,
2
0
2
5
,
d
o
i
:
1
0
.
1
0
3
8
/
s
4
1
5
8
6
-
0
2
4
-
0
8
3
2
8
-
6.
[
9
]
A
.
O
.
K
e
h
i
n
d
e
,
“
Le
v
e
r
a
g
i
n
g
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
f
o
r
p
r
e
d
i
c
t
i
v
e
m
o
d
e
l
s
i
n
h
e
a
l
t
h
c
a
r
e
t
o
e
n
h
a
n
c
e
p
a
t
i
e
n
t
o
u
t
c
o
me
ma
n
a
g
e
m
e
n
t
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
R
e
se
a
rc
h
J
o
u
rn
a
l
o
f
M
o
d
e
rn
i
z
a
t
i
o
n
i
n
E
n
g
i
n
e
e
ri
n
g
T
e
c
h
n
o
l
o
g
y
a
n
d
S
c
i
e
n
c
e
,
v
o
l
.
7
,
n
o
.
1
,
p
p
.
1
4
6
5
–
1
4
8
2
,
2
0
2
5
,
d
o
i
:
1
0
.
5
6
7
2
6
/
I
R
JM
E
TS6
6
1
9
8
.
[
1
0
]
S
.
F
r
i
e
d
r
i
c
h
a
n
d
T.
F
r
i
e
d
e
,
“
O
n
t
h
e
r
o
l
e
o
f
b
e
n
c
h
mar
k
i
n
g
d
a
t
a
s
e
t
s
a
n
d
si
mu
l
a
t
i
o
n
s
i
n
m
e
t
h
o
d
c
o
mp
a
r
i
s
o
n
st
u
d
i
e
s,
”
B
i
o
m
e
t
r
i
c
a
l
J
o
u
rn
a
l
,
v
o
l
.
6
6
,
n
o
.
1
,
2
0
2
4
,
d
o
i
:
1
0
.
1
0
0
2
/
b
i
m
j
.
2
0
2
2
0
0
2
1
2
.
[
1
1
]
M
.
M
.
A
l
i
,
B
.
K
.
P
a
u
l
,
K
.
A
h
me
d
,
F
.
M
.
B
u
i
,
J
.
M
.
W
.
Q
u
i
n
n
,
a
n
d
M
.
A
.
M
o
n
i
,
“
H
e
a
r
t
d
i
s
e
a
s
e
p
r
e
d
i
c
t
i
o
n
u
si
n
g
su
p
e
r
v
i
s
e
d
mac
h
i
n
e
l
e
a
r
n
i
n
g
a
l
g
o
r
i
t
h
ms:
p
e
r
f
o
r
man
c
e
a
n
a
l
y
si
s
a
n
d
c
o
mp
a
r
i
s
o
n
,
”
C
o
m
p
u
t
e
rs
i
n
Bi
o
l
o
g
y
a
n
d
Me
d
i
c
i
n
e
,
v
o
l
.
1
3
6
,
2
0
2
1
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
c
o
m
p
b
i
o
me
d
.
2
0
2
1
.
1
0
4
6
7
2
.
[
1
2
]
E.
O
l
i
v
e
i
r
a
,
M
.
R
o
d
r
i
g
u
e
s,
J
.
P
.
P
e
r
e
i
r
a
,
A
.
M
.
L
o
p
e
s,
I
.
I
.
M
e
st
r
i
c
,
a
n
d
S
.
B
j
e
l
o
g
r
l
i
c
,
“
U
n
l
a
b
e
l
e
d
l
e
a
r
n
i
n
g
a
l
g
o
r
i
t
h
ms
a
n
d
o
p
e
r
a
t
i
o
n
s
:
o
v
e
r
v
i
e
w
a
n
d
f
u
t
u
r
e
t
r
e
n
d
s
i
n
d
e
f
e
n
se
s
e
c
t
o
r
,
”
Art
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
Re
v
i
e
w
,
v
o
l
.
5
7
,
n
o
.
3
,
2
0
2
4
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
0
4
6
2
-
023
-
1
0
6
9
2
-
0.
[
1
3
]
K
.
A
l
o
m
a
r
,
H
.
I
.
A
y
se
l
,
a
n
d
X
.
C
a
i
,
“
D
a
t
a
a
u
g
me
n
t
a
t
i
o
n
i
n
c
l
a
ssi
f
i
c
a
t
i
o
n
a
n
d
s
e
g
me
n
t
a
t
i
o
n
:
a
s
u
r
v
e
y
a
n
d
n
e
w
s
t
r
a
t
e
g
i
e
s,”
J
o
u
r
n
a
l
o
f
I
m
a
g
i
n
g
,
v
o
l
.
9
,
n
o
.
2
,
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
j
i
m
a
g
i
n
g
9
0
2
0
0
4
6
.
[
1
4
]
M
.
O
.
-
A
d
j
e
i
,
J
.
B
.
H
.
-
A
c
q
u
a
h
,
T.
F
r
i
mp
o
n
g
,
a
n
d
G
.
A
.
-
S
a
l
a
a
m,
“
I
mb
a
l
a
n
c
e
d
c
l
a
s
s
d
i
s
t
r
i
b
u
t
i
o
n
a
n
d
p
e
r
f
o
r
man
c
e
e
v
a
l
u
a
t
i
o
n
met
r
i
c
s:
a
s
y
st
e
ma
t
i
c
r
e
v
i
e
w
o
f
p
r
e
d
i
c
t
i
o
n
a
c
c
u
r
a
c
y
f
o
r
d
e
t
e
r
mi
n
i
n
g
m
o
d
e
l
p
e
r
f
o
r
ma
n
c
e
i
n
h
e
a
l
t
h
c
a
r
e
s
y
st
e
ms
,
”
PLO
S
D
i
g
i
t
a
l
H
e
a
l
t
h
,
v
o
l
.
2
,
n
o
.
1
1
,
2
0
2
3
,
d
o
i
:
1
0
.
1
3
7
1
/
j
o
u
r
n
a
l
.
p
d
i
g
.
0
0
0
0
2
9
0
.
[
1
5
]
K
.
M
a
h
a
r
a
n
a
,
S
.
M
o
n
d
a
l
,
a
n
d
B
.
N
e
m
a
d
e
,
“
A
r
e
v
i
e
w
:
d
a
t
a
p
r
e
-
p
r
o
c
e
ss
i
n
g
a
n
d
d
a
t
a
a
u
g
me
n
t
a
t
i
o
n
t
e
c
h
n
i
q
u
e
s,
”
G
l
o
b
a
l
T
r
a
n
si
t
i
o
n
s
Pro
c
e
e
d
i
n
g
s
,
v
o
l
.
3
,
n
o
.
1
,
p
p
.
9
1
–
9
9
,
2
0
2
2
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
g
l
t
p
.
2
0
2
2
.
0
4
.
0
2
0
.
[
1
6
]
M
.
S
.
M
i
l
o
se
v
i
c
a
n
d
V
.
M
.
C
i
r
i
c
,
“
Ex
t
r
e
me
m
i
n
o
r
i
t
y
c
l
a
ss
d
e
t
e
c
t
i
o
n
i
n
i
mb
a
l
a
n
c
e
d
d
a
t
a
f
o
r
n
e
t
w
o
r
k
i
n
t
r
u
s
i
o
n
,
”
C
o
m
p
u
t
e
rs
&
S
e
c
u
r
i
t
y
,
v
o
l
.
1
2
3
,
2
0
2
2
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
c
o
s
e
.
2
0
2
2
.
1
0
2
9
4
0
.
[
1
7
]
M
.
C
a
r
v
a
l
h
o
,
A
.
J
.
P
i
n
h
o
,
a
n
d
S
.
B
r
á
s,
“
R
e
sam
p
l
i
n
g
a
p
p
r
o
a
c
h
e
s
t
o
h
a
n
d
l
e
c
l
a
ss
i
m
b
a
l
a
n
c
e
:
a
r
e
v
i
e
w
f
r
o
m
a
d
a
t
a
p
e
r
sp
e
c
t
i
v
e
,
”
J
o
u
rn
a
l
o
f
Bi
g
D
a
t
a
,
v
o
l
.
1
2
,
n
o
.
1
,
2
0
2
5
,
d
o
i
:
1
0
.
1
1
8
6
/
s
4
0
5
3
7
-
025
-
0
1
1
1
9
-
4.
[
1
8
]
H
.
H
a
i
r
a
n
i
,
T.
W
i
d
i
y
a
n
i
n
g
t
y
a
s,
a
n
d
D
.
D
.
P
r
a
s
e
t
y
a
,
“
A
d
d
r
e
ssi
n
g
c
l
a
ss
i
m
b
a
l
a
n
c
e
o
f
h
e
a
l
t
h
d
a
t
a
:
a
sy
s
t
e
m
a
t
i
c
l
i
t
e
r
a
t
u
r
e
r
e
v
i
e
w
o
n
mo
d
i
f
i
e
d
s
y
n
t
h
e
t
i
c
mi
n
o
r
i
t
y
o
v
e
r
sa
mp
l
i
n
g
t
e
c
h
n
i
q
u
e
(
S
M
O
T
E)
st
r
a
t
e
g
i
e
s,”
J
O
I
V
:
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
n
I
n
f
o
rm
a
t
i
c
s
Vi
su
a
l
i
z
a
t
i
o
n
,
v
o
l
.
8
,
n
o
.
3
,
p
p
.
1
3
1
0
–
1
3
1
8
,
2
0
2
4
,
d
o
i
:
1
0
.
6
2
5
2
7
/
j
o
i
v
.
8
.
3
.
2
2
8
3
.
[
1
9
]
M
.
Z
a
k
a
r
i
a
h
,
S
.
A
.
A
l
Q
a
h
t
a
n
i
,
a
n
d
M
.
S
.
A
l
-
R
a
k
h
a
mi
,
“
M
a
c
h
i
n
e
l
e
a
r
n
i
n
g
-
b
a
s
e
d
a
d
a
p
t
i
v
e
sy
n
t
h
e
t
i
c
sam
p
l
i
n
g
t
e
c
h
n
i
q
u
e
f
o
r
i
n
t
r
u
s
i
o
n
d
e
t
e
c
t
i
o
n
,
”
A
p
p
l
i
e
d
S
c
i
e
n
c
e
s
,
v
o
l
.
1
3
,
n
o
.
1
1
,
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
a
p
p
1
3
1
1
6
5
0
4
.
[
2
0
]
T.
-
T.
-
H
.
Le
,
Y
.
S
h
i
n
,
M
.
K
i
m,
a
n
d
H
.
K
i
m
,
“
T
o
w
a
r
d
s
u
n
b
a
l
a
n
c
e
d
m
u
l
t
i
c
l
a
ss
i
n
t
r
u
s
i
o
n
d
e
t
e
c
t
i
o
n
w
i
t
h
h
y
b
r
i
d
samp
l
i
n
g
m
e
t
h
o
d
s
a
n
d
e
n
s
e
mb
l
e
c
l
a
ssi
f
i
c
a
t
i
o
n
,
”
A
p
p
l
i
e
d
S
o
f
t
C
o
m
p
u
t
i
n
g
,
v
o
l
.
1
5
7
,
2
0
2
4
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
a
s
o
c
.
2
0
2
4
.
1
1
1
5
1
7
.
[
2
1
]
Y
.
D
i
n
g
,
H
.
Zh
u
,
R
.
C
h
e
n
,
a
n
d
R
.
Li
,
“
A
n
e
f
f
i
c
i
e
n
t
A
d
a
B
o
o
st
a
l
g
o
r
i
t
h
m
w
i
t
h
t
h
e
m
u
l
t
i
p
l
e
t
h
r
e
s
h
o
l
d
s
c
l
a
ss
i
f
i
c
a
t
i
o
n
,
”
A
p
p
l
i
e
d
S
c
i
e
n
c
e
s
,
v
o
l
.
1
2
,
n
o
.
1
2
,
2
0
2
2
,
d
o
i
:
1
0
.
3
3
9
0
/
a
p
p
1
2
1
2
5
8
7
2
.
[
2
2
]
P
.
M
a
h
a
j
a
n
,
S
.
U
d
d
i
n
,
F
.
H
a
j
a
t
i
,
a
n
d
M
.
A
.
M
o
n
i
,
“
E
n
sem
b
l
e
l
e
a
r
n
i
n
g
f
o
r
d
i
sea
s
e
p
r
e
d
i
c
t
i
o
n
:
a
r
e
v
i
e
w
,
”
H
e
a
l
t
h
c
a
r
e
,
v
o
l
.
1
1
,
n
o
.
1
2
,
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
h
e
a
l
t
h
c
a
r
e
1
1
1
2
1
8
0
8
.
[
2
3
]
F
.
Ji
a
,
S
.
L
i
,
H
.
Zu
o
,
a
n
d
J
.
S
h
e
n
,
“
D
e
e
p
n
e
u
r
a
l
n
e
t
w
o
r
k
e
n
s
e
m
b
l
e
f
o
r
t
h
e
i
n
t
e
l
l
i
g
e
n
t
f
a
u
l
t
d
i
a
g
n
o
s
i
s o
f
m
a
c
h
i
n
e
s
u
n
d
e
r
i
m
b
a
l
a
n
c
e
d
d
a
t
a
,
”
I
EE
E
A
c
c
e
ss
,
v
o
l
.
8
,
p
p
.
1
2
0
9
7
4
–
1
2
0
9
8
2
,
2
0
2
0
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
E
S
S
.
2
0
2
0
.
3
0
0
6
8
9
5
.
[
2
4
]
O
.
V
o
l
k
a
n
d
G
.
S
i
n
g
e
r
,
“
A
n
a
d
a
p
t
i
v
e
c
o
s
t
-
se
n
si
t
i
v
e
l
e
a
r
n
i
n
g
a
p
p
r
o
a
c
h
i
n
n
e
u
r
a
l
n
e
t
w
o
r
k
s
t
o
m
i
n
i
mi
z
e
l
o
c
a
l
t
r
a
i
n
i
n
g
–
t
e
st
c
l
a
ss
d
i
s
t
r
i
b
u
t
i
o
n
s m
i
sma
t
c
h
,
”
I
n
t
e
l
l
i
g
e
n
t
S
y
st
e
m
s
w
i
t
h
A
p
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
2
1
,
2
0
2
4
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
i
sw
a
.
2
0
2
3
.
2
0
0
3
1
6
.
[
2
5
]
F
.
F
e
n
g
,
K
.
-
C
.
Li
,
J.
S
h
e
n
,
Q
.
Z
h
o
u
,
a
n
d
X
.
Y
a
n
g
,
“
U
si
n
g
c
o
st
-
s
e
n
s
i
t
i
v
e
l
e
a
r
n
i
n
g
a
n
d
f
e
a
t
u
r
e
se
l
e
c
t
i
o
n
a
l
g
o
r
i
t
h
ms
t
o
i
m
p
r
o
v
e
t
h
e
p
e
r
f
o
r
m
a
n
c
e
o
f
i
m
b
a
l
a
n
c
e
d
c
l
a
ssi
f
i
c
a
t
i
o
n
,
”
I
EEE
A
c
c
e
ss
,
v
o
l
.
8
,
p
p
.
6
9
9
7
9
–
6
9
9
9
6
,
2
0
2
0
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
0
.
2
9
8
7
3
6
4
.
[
2
6
]
N
.
F
.
H
o
r
d
r
i
,
S
.
S
o
p
h
i
a
y
a
t
i
,
N
.
F
i
r
d
a
u
s,
a
n
d
S
.
M
a
r
i
y
a
m
,
“
H
a
n
d
l
i
n
g
c
l
a
ss
i
m
b
a
l
a
n
c
e
i
n
c
r
e
d
i
t
c
a
r
d
f
r
a
u
d
u
si
n
g
r
e
sam
p
l
i
n
g
met
h
o
d
s,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
A
d
v
a
n
c
e
d
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
a
n
d
Ap
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
9
,
n
o
.
1
1
,
2
0
1
8
,
d
o
i
:
1
0
.
1
4
5
6
9
/
I
JA
C
S
A
.
2
0
1
8
.
0
9
1
1
5
5
.
[
2
7
]
S
.
C
a
t
e
n
i
,
V
.
C
o
l
l
a
,
a
n
d
M
.
V
a
n
n
u
c
c
i
,
“
A
m
e
t
h
o
d
f
o
r
r
e
s
a
mp
l
i
n
g
i
mb
a
l
a
n
c
e
d
d
a
t
a
s
e
t
s
i
n
b
i
n
a
r
y
c
l
a
ss
i
f
i
c
a
t
i
o
n
t
a
s
k
s
f
o
r
r
e
a
l
-
w
o
r
l
d
p
r
o
b
l
e
ms,
”
N
e
u
ro
c
o
m
p
u
t
i
n
g
,
v
o
l
.
1
3
5
,
p
p
.
3
2
–
4
1
,
2
0
1
4
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
n
e
u
c
o
m.
2
0
1
3
.
0
5
.
0
5
9
.
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
.
2
,
Ap
r
il
20
26
:
1
4
0
2
-
1
4
0
8
1408
[
2
8
]
M
.
S
.
K
r
a
i
e
m
,
F
.
S
.
-
H
e
r
n
á
n
d
e
z
,
a
n
d
M
.
N
.
M
.
-
G
a
r
c
í
a
,
“
S
e
l
e
c
t
i
n
g
t
h
e
su
i
t
a
b
l
e
r
e
sam
p
l
i
n
g
st
r
a
t
e
g
y
f
o
r
i
mb
a
l
a
n
c
e
d
d
a
t
a
c
l
a
ss
i
f
i
c
a
t
i
o
n
r
e
g
a
r
d
i
n
g
d
a
t
a
se
t
p
r
o
p
e
r
t
i
e
s
:
a
n
a
p
p
r
o
a
c
h
b
a
s
e
d
o
n
a
ss
o
c
i
a
t
i
o
n
m
o
d
e
l
s
,
”
A
p
p
l
i
e
d
S
c
i
e
n
c
e
s
,
v
o
l
.
1
1
,
n
o
.
1
8
,
2
0
2
1
,
d
o
i
:
1
0
.
3
3
9
0
/
a
p
p
1
1
1
8
8
5
4
6
.
[
2
9
]
S
.
B
a
g
u
i
a
n
d
K
.
Li
,
“
R
e
sam
p
l
i
n
g
i
m
b
a
l
a
n
c
e
d
d
a
t
a
f
o
r
n
e
t
w
o
r
k
i
n
t
r
u
si
o
n
d
e
t
e
c
t
i
o
n
d
a
t
a
s
e
t
s
,
”
J
o
u
rn
a
l
o
f
B
i
g
D
a
t
a
,
v
o
l
.
8
,
n
o
.
1
,
2
0
2
1
,
d
o
i
:
1
0
.
1
1
8
6
/
s4
0
5
3
7
-
0
2
0
-
0
0
3
9
0
-
x.
[
3
0
]
R
.
G
h
o
r
b
a
n
i
a
n
d
R
.
G
h
o
u
si
,
“
C
o
m
p
a
r
i
n
g
d
i
f
f
e
r
e
n
t
r
e
sam
p
l
i
n
g
me
t
h
o
d
s
i
n
p
r
e
d
i
c
t
i
n
g
st
u
d
e
n
t
s’
p
e
r
f
o
r
ma
n
c
e
u
si
n
g
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,
”
I
EEE
Ac
c
e
ss
,
v
o
l
.
8
,
p
p
.
6
7
8
9
9
–
6
7
9
1
1
,
2
0
2
0
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
0
.
2
9
8
6
8
0
9
.
[
3
1
]
M
.
S
o
p
i
y
a
n
,
F
.
F
a
u
z
i
a
h
,
a
n
d
Y
.
F
.
W
i
j
a
y
a
,
“
F
r
a
u
d
d
e
t
e
c
t
i
o
n
u
s
i
n
g
r
a
n
d
o
m
f
o
r
e
s
t
c
l
a
ssi
f
i
e
r
,
l
o
g
i
s
t
i
c
r
e
g
r
e
ssi
o
n
,
a
n
d
g
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
c
l
a
ssi
f
i
e
r
a
l
g
o
r
i
t
h
ms
o
n
c
r
e
d
i
t
c
a
r
d
s,
”
J
U
I
T
A:
J
u
r
n
a
l
I
n
f
o
rm
a
t
i
k
a
,
v
o
l
.
1
0
,
n
o
.
1
,
p
p
.
7
7
–
8
8
,
2
0
2
2
,
d
o
i
:
1
0
.
3
0
5
9
5
/
j
u
i
t
a
.
v
1
0
i
1
.
1
2
0
5
0
.
[
3
2
]
N
.
S
.
S
.
P
r
a
n
a
v
i
,
T.
K
.
S
.
S
.
S
r
u
t
h
i
,
B
.
J.
N
.
S
i
r
i
s
h
a
,
M
.
S
.
N
a
y
a
k
,
a
n
d
V
.
S
.
G
.
Th
a
d
i
k
e
m
a
l
l
a
,
“
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
mi
n
o
r
i
t
y
o
v
e
r
sam
p
l
i
n
g
a
n
d
r
a
n
d
o
m
f
o
r
e
st
t
e
c
h
n
i
q
u
e
,
”
i
n
2
0
2
2
3
r
d
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
f
o
r
Em
e
r
g
i
n
g
T
e
c
h
n
o
l
o
g
y
,
2
0
2
2
,
p
p
.
1
–
6
,
d
o
i
:
1
0
.
1
1
0
9
/
I
N
C
ET5
4
5
3
1
.
2
0
2
2
.
9
8
2
4
1
4
6
.
[
3
3
]
A
.
M
.
A
b
u
r
b
e
i
a
n
a
n
d
H
.
I
.
A
sh
q
a
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
e
n
h
a
n
c
e
d
r
a
n
d
o
m
f
o
r
e
st
c
l
a
ssi
f
i
e
r
f
o
r
i
m
b
a
l
a
n
c
e
d
d
a
t
a
,
”
i
n
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
2
0
2
3
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
Ad
v
a
n
c
e
s
i
n
C
o
m
p
u
t
i
n
g
Re
s
e
a
r
c
h
,
2
0
2
3
,
p
p
.
6
0
5
–
6
1
6
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
3
-
0
3
1
-
3
3
7
4
3
-
7
_
4
8
.
[
3
4
]
F
.
Y
a
g
i
n
,
I
.
C
i
c
e
k
,
a
n
d
Z
.
K
u
c
u
k
a
k
c
a
l
i
,
“
C
l
a
ssi
f
i
c
a
t
i
o
n
o
f
st
r
o
k
e
w
i
t
h
g
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
t
r
e
e
u
s
i
n
g
s
mo
t
e
-
b
a
s
e
d
o
v
e
r
sa
mp
l
i
n
g
met
h
o
d
,
”
M
e
d
i
c
i
n
e
S
c
i
e
n
c
e
:
I
n
t
e
rn
a
t
i
o
n
a
l
Me
d
i
c
a
l
J
o
u
r
n
a
l
,
v
o
l
.
1
0
,
n
o
.
4
,
p
p
.
1
5
1
0
–
1
5
1
5
,
2
0
2
1
,
d
o
i
:
1
0
.
5
4
5
5
/
me
d
sc
i
e
n
c
e
.
2
0
2
1
.
0
9
.
3
2
2
.
[
3
5
]
Y
.
P
r
i
st
y
a
n
t
o
,
Z
.
M
u
k
a
r
a
b
i
m
a
n
,
a
n
d
A
.
F
.
N
u
g
r
a
h
a
,
“
E
x
t
r
e
me
g
r
a
d
i
e
n
t
b
o
o
st
i
n
g
a
l
g
o
r
i
t
h
m
t
o
i
m
p
r
o
v
e
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
mo
d
e
l
p
e
r
f
o
r
m
a
n
c
e
o
n
m
u
l
t
i
c
l
a
ss
i
mb
a
l
a
n
c
e
d
d
a
t
a
s
e
t
,
”
J
O
I
V
:
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
n
I
n
f
o
rm
a
t
i
c
s
Vi
s
u
a
l
i
za
t
i
o
n
,
v
o
l
.
7
,
n
o
.
3
,
p
p
.
7
1
0
–
7
1
5
,
2
0
2
3
,
d
o
i
:
1
0
.
3
0
6
3
0
/
j
o
i
v
.
7
.
3
.
1
1
0
2
.
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
Rusm
a
An
iez
a
Rusla
n
h
o
l
d
s
a
B.
S
.
d
e
g
re
e
in
In
f
o
rm
a
ti
o
n
Tec
h
n
o
l
o
g
y
fr
o
m
Un
iv
e
rsiti
T
u
n
H
u
ss
e
in
On
n
M
a
l
a
y
sia
,
Jo
h
o
r,
in
2
0
2
4
.
S
h
e
is
p
u
rs
u
in
g
a
n
M
.
S
.
i
n
In
f
o
rm
a
ti
o
n
Tec
h
n
o
l
o
g
y
a
n
d
wo
r
k
in
g
o
n
a
p
ro
jec
t
ti
tl
e
d
"
A
d
d
re
ss
in
g
imb
a
l
a
n
c
e
d
a
tas
e
ts
in
m
a
c
h
in
e
lea
rn
in
g
"
.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
at
e
m
a
il
:
h
i2
3
0
0
2
7
@s
tu
d
e
n
t.
u
th
m
.
e
d
u
.
m
y
.
Nure
ize
Ar
b
a
iy
is
c
u
rre
n
tl
y
in
t
h
e
De
p
a
rtme
n
t
o
f
S
o
f
twa
re
En
g
i
n
e
e
rin
g
a
t
th
e
F
a
c
u
lt
y
o
f
Co
m
p
u
ter
S
c
ien
c
e
a
n
d
In
f
o
rm
a
ti
o
n
Tec
h
n
o
l
o
g
y
,
Un
iv
e
rsiti
Tu
n
Hu
ss
e
in
On
n
M
a
lay
sia
(UTHM
).
He
r
re
se
a
rc
h
in
tere
sts a
re
m
u
lt
i
-
c
rit
e
ria d
e
c
isio
n
-
m
a
k
in
g
,
fu
z
z
y
re
g
re
ss
io
n
,
fu
z
z
y
a
u
to
-
re
g
re
ss
io
n
,
i
n
fo
rm
a
ti
o
n
sy
ste
m
s,
a
n
d
d
e
e
p
lea
rn
in
g
.
S
h
e
h
a
s
su
p
e
rv
ise
d
se
v
e
ra
l
Ph
.
D
.
a
n
d
m
a
ste
r'
s
stu
d
e
n
ts
.
S
h
e
p
u
b
li
sh
e
d
a
rti
c
les
i
n
v
a
rio
u
s
in
tern
a
ti
o
n
a
l
jo
u
rn
a
ls
a
n
d
c
o
n
fe
re
n
c
e
p
ro
c
e
e
d
i
n
g
s.
S
h
e
h
a
s
a
c
ted
a
s
a
re
v
iew
e
r
fo
r
m
u
lt
ip
l
e
jo
u
r
n
a
ls
a
n
d
c
o
n
fe
re
n
c
e
s.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
n
u
re
ize
@u
th
m
.
e
d
u
.
m
y
.
Pei
-
Chu
n
L
in
re
c
e
iv
e
d
a
B
.
S
c
.
d
e
g
re
e
in
th
e
De
p
a
rtme
n
t
o
f
M
a
th
e
m
a
ti
c
s
fro
m
Na
ti
o
n
a
l
Ka
o
h
si
u
n
g
No
rm
a
l
Un
i
v
e
rsity
,
Ka
o
h
siu
n
g
,
Taiwa
n
a
n
d
r
e
c
e
iv
e
d
a
n
M
.
S
c
.
d
e
g
re
e
in
th
e
De
p
a
rtme
n
t
o
f
M
a
th
e
m
a
ti
c
a
l
S
c
ien
c
e
s
fro
m
Na
ti
o
n
a
l
Ch
e
n
g
c
h
i
U
n
iv
e
rsit
y
,
Tai
p
e
i,
Taiwa
n
.
M
o
re
o
v
e
r,
sh
e
re
c
e
iv
e
d
a
P
h
.
D
.
d
e
g
re
e
fro
m
g
ra
d
u
a
te
sc
h
o
o
l
o
f
In
f
o
rm
a
ti
o
n
,
P
ro
d
u
c
ti
o
n
a
n
d
S
y
ste
m
s
,
Was
e
d
a
Un
iv
e
rsity
,
F
u
k
u
o
k
a
,
Ja
p
a
n
.
S
h
e
is
c
u
rre
n
tl
y
a
n
a
ss
istan
t
p
ro
fe
ss
o
r
i
n
t
h
e
De
p
a
rtme
n
t
o
f
I
n
fo
rm
a
ti
o
n
E
n
g
i
n
e
e
rin
g
a
n
d
Co
m
p
u
ter
S
c
ien
c
e
,
F
e
n
g
C
h
ia
Un
iv
e
rsity
.
He
r
re
se
a
rc
h
in
tere
sts
in
c
lu
d
e
so
f
t
c
o
m
p
u
ti
n
g
,
a
rti
ficia
l
in
telli
g
e
n
t,
r
o
b
o
t
c
o
m
p
u
ti
n
g
,
sta
ti
stica
l
m
o
d
e
li
n
g
,
a
n
d
b
ig
d
a
ta
a
n
a
ly
sis.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
p
e
iclin
@fc
u
.
e
d
u
.
tw.
Evaluation Warning : The document was created with Spire.PDF for Python.