I
nd
o
ne
s
ia
n J
o
urna
l o
f
E
lect
rica
l En
g
ineering
a
nd
Co
m
pu
t
er
Science
Vo
l.
4
3
,
No
.
3
,
Sep
tem
b
er
202
6
,
p
p
.
8
8
0
~
89
7
I
SS
N:
2
5
0
2
-
4
7
5
2
,
DOI
: 1
0
.
1
1
5
9
1
/ijeecs.v
4
3
.
i
3
.
pp
880
-
89
7
880
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ee
cs.ia
esco
r
e.
co
m
Explain
a
ble hybr
id ma
chine l
ea
rni
ng
f
o
r predi
cting
student
enro
llment
decisi
o
ns in Indo
nesia
n
priva
te
univ
ersi
ti
es
Dy
bio
Do
m
pu
H
o
t
Asi
h
1
,
Alex
P
er
ma
t
a
nta
K
a
ro
-
K
a
ro
2
,
E
rbin
Sito
rus
3
1
I
n
f
o
r
mat
i
o
n
S
y
st
e
ms St
u
d
y
P
r
o
g
r
a
m
,
U
n
i
v
e
r
s
i
t
a
s W
i
r
a
h
u
s
a
d
a
M
e
d
a
n
,
M
e
d
a
n
,
I
n
d
o
n
e
si
a
2
M
a
n
a
g
e
me
n
t
S
t
u
d
y
P
r
o
g
r
a
m,
U
n
i
v
e
r
si
t
as
W
i
r
a
h
u
sa
d
a
M
e
d
a
n
,
M
e
d
a
n
,
I
n
d
o
n
e
si
a
3
I
n
f
o
r
mat
i
o
n
S
y
st
e
ms St
u
d
y
P
r
o
g
r
a
m
,
U
n
i
v
e
r
s
i
t
a
s A
u
d
i
I
n
d
o
n
e
si
a
,
M
e
d
a
n
,
I
n
d
o
n
e
s
i
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
u
n
2
7
,
2
0
2
6
R
ev
is
ed
Au
g
1
0
,
2
0
2
6
Acc
ep
ted
Au
g
1
9
,
2
0
2
6
S
tu
d
e
n
t
e
n
r
o
ll
m
e
n
t
p
re
d
icti
o
n
is
a
n
imp
o
r
tan
t
c
h
a
ll
e
n
g
e
f
o
r
p
ri
v
a
te
u
n
i
v
e
rsiti
e
s
b
e
c
a
u
se
a
c
c
u
ra
te
p
re
d
ictio
n
s
c
a
n
su
p
p
o
r
t
a
d
m
issio
n
p
lan
n
in
g
,
m
a
rk
e
ti
n
g
stra
teg
ies
,
a
n
d
in
st
it
u
ti
o
n
a
l
re
so
u
rc
e
a
ll
o
c
a
ti
o
n
.
Ho
we
v
e
r,
c
o
n
v
e
n
ti
o
n
a
l
p
re
d
icti
o
n
m
o
d
e
ls
m
a
y
p
ro
v
id
e
h
i
g
h
p
re
d
icti
v
e
p
e
r
fo
rm
a
n
c
e
wh
il
e
o
ffe
ri
n
g
li
m
it
e
d
i
n
sig
h
t
in
t
o
th
e
fa
c
to
rs
u
n
d
e
rl
y
in
g
stu
d
e
n
ts’
e
n
ro
ll
m
e
n
t
d
e
c
isio
n
s.
T
h
is
stu
d
y
p
ro
p
o
se
s
a
n
e
x
p
lain
a
b
le
h
y
b
rid
m
a
c
h
in
e
lea
rn
in
g
a
p
p
ro
a
c
h
t
o
p
re
d
ict
st
u
d
e
n
t
e
n
r
o
ll
m
e
n
t
d
e
c
isio
n
s
in
I
n
d
o
n
e
sia
n
p
riv
a
te
u
n
i
v
e
rsi
ti
e
s
a
n
d
t
o
id
e
n
ti
f
y
th
e
fa
c
to
rs
t
h
a
t
c
o
n
tri
b
u
te
m
o
s
t
stro
n
g
ly
to
th
e
p
re
d
ictio
n
o
u
tco
m
e
s.
Th
e
stu
d
y
u
se
s
stu
d
e
n
t
e
n
ro
l
lme
n
t
d
a
ta
c
o
ll
e
c
ted
fro
m
Un
i
v
e
rsitas
Wi
ra
h
u
sa
d
a
M
e
d
a
n
a
n
d
Un
iv
e
rsitas
Au
d
i
I
n
d
o
n
e
sia
.
Th
e
p
ro
p
o
se
d
a
p
p
ro
a
c
h
c
o
m
b
in
e
s
fe
a
tu
re
se
lec
ti
o
n
a
n
d
m
u
lt
i
p
le
m
a
c
h
in
e
lea
rn
in
g
a
lg
o
rit
h
m
s
to
imp
ro
v
e
p
re
d
ictiv
e
p
e
rf
o
rm
a
n
c
e
,
wh
il
e
e
x
p
lai
n
a
b
le
a
rti
ficia
l
in
telli
g
e
n
c
e
(XA
I)
u
si
n
g
S
h
a
p
ley
a
d
d
i
ti
v
e
e
x
p
lan
a
ti
o
n
s
(
S
HA
P
)
is
e
m
p
lo
y
e
d
t
o
i
n
terp
re
t
t
h
e
m
o
d
e
l
p
re
d
ictio
n
s.
Th
e
m
o
d
e
ls
a
re
e
v
a
l
u
a
ted
u
si
n
g
a
c
c
u
ra
c
y
,
p
re
c
isio
n
,
re
c
a
ll
,
F
1
-
sc
o
re
,
re
c
e
iv
e
r
o
p
e
ra
ti
n
g
c
h
a
ra
c
teristic
-
a
re
a
u
n
d
e
r
t
h
e
c
u
r
v
e
(
ROC
-
AUC
)
,
a
n
d
c
o
n
f
u
sio
n
m
a
tri
x
a
n
a
ly
sis.
T
h
e
re
su
lt
s
sh
o
w
th
a
t
t
h
e
p
ro
p
o
se
d
h
y
b
ri
d
a
p
p
ro
a
c
h
p
ro
v
id
e
s
c
o
m
p
e
ti
ti
v
e
p
re
d
ictiv
e
p
e
rfo
rm
a
n
c
e
c
o
m
p
a
re
d
with
in
d
iv
id
u
a
l
m
a
c
h
in
e
lea
rn
in
g
m
o
d
e
ls.
T
h
e
S
HA
P
a
n
a
ly
sis
fu
rth
e
r
re
v
e
a
ls
th
e
re
lat
iv
e
c
o
n
tri
b
u
t
io
n
a
n
d
d
irec
ti
o
n
o
f
t
h
e
m
o
st
in
flu
e
n
ti
a
l
fe
a
tu
re
s
in
d
e
term
in
in
g
stu
d
e
n
t
e
n
ro
ll
m
e
n
t
d
e
c
isio
n
s.
T
h
e
se
fin
d
i
n
g
s
d
e
m
o
n
stra
te
t
h
a
t
c
o
m
b
in
in
g
p
re
d
ictiv
e
m
o
d
e
li
n
g
wit
h
e
x
p
l
a
in
a
b
il
it
y
c
a
n
p
ro
v
id
e
n
o
t
o
n
ly
a
c
c
u
ra
te
e
n
ro
ll
m
e
n
t
p
re
d
icti
o
n
s
b
u
t
a
lso
a
c
ti
o
n
a
b
le
in
sig
h
ts
f
o
r
u
n
iv
e
rsity
d
e
c
isio
n
-
m
a
k
e
rs
.
Th
e
st
u
d
y
c
o
n
tr
ib
u
tes
a
n
in
te
g
ra
ted
a
p
p
ro
a
c
h
fo
r
su
p
p
o
rti
n
g
d
a
ta
-
d
r
iv
e
n
st
u
d
e
n
t
re
c
ru
it
m
e
n
t
a
n
d
e
n
ro
ll
m
e
n
t
m
a
n
a
g
e
m
e
n
t
in
In
d
o
n
e
sia
n
p
riv
a
t
e
h
ig
h
e
r
e
d
u
c
a
ti
o
n
i
n
stit
u
ti
o
n
s.
K
ey
w
o
r
d
s
:
E
x
p
lain
ab
le
ar
tific
ial
in
tellig
en
ce
Hy
b
r
id
m
ac
h
in
e
lear
n
i
n
g
Ma
ch
in
e
lear
n
in
g
Priv
ate
u
n
iv
er
s
ities
Stu
d
en
t e
n
r
o
llm
e
n
t p
r
e
d
ictio
n
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
Dy
b
io
Do
m
p
u
Ho
t A
s
ih
I
n
f
o
r
m
atio
n
Sy
s
tem
s
Stu
d
y
Pr
o
g
r
am
,
U
n
iv
er
s
itas
W
ir
ah
u
s
ad
a
Me
d
an
J
l.
T
B
.
Sima
tu
p
an
g
No
.
1
4
8
,
Kel.
Su
n
g
g
al,
Kec
.
Me
d
an
Su
n
g
g
al,
Ko
ta
Me
d
an
2
0
1
2
8
,
I
n
d
o
n
esia
E
m
ail:
d
y
b
io
.
d
o
m
p
u
@
g
m
ail.
co
m
1.
I
NT
RO
D
UCT
I
O
N
T
h
e
r
ap
id
d
ev
el
o
p
m
en
t
o
f
d
ig
i
tal
tech
n
o
lo
g
y
h
as
ch
an
g
ed
h
o
w
h
ig
h
er
ed
u
ca
tio
n
in
s
titu
tio
n
s
m
an
ag
e
s
tu
d
en
t
r
ec
r
u
itm
en
t,
a
d
m
is
s
io
n
,
an
d
e
n
r
o
llm
en
t.
Un
iv
er
s
itie
s
in
cr
ea
s
in
g
ly
r
ely
o
n
d
ata
-
d
r
iv
en
ap
p
r
o
ac
h
es
to
u
n
d
er
s
tan
d
p
r
o
s
p
ec
tiv
e
s
tu
d
e
n
ts
an
d
im
p
r
o
v
e
r
ec
r
u
it
m
en
t stra
teg
ies,
p
ar
ticu
lar
ly
as c
o
m
p
et
itio
n
am
o
n
g
h
ig
h
e
r
ed
u
ca
tio
n
i
n
s
titu
tio
n
s
co
n
tin
u
es
to
in
cr
ea
s
e.
I
n
I
n
d
o
n
esia,
th
is
ch
allen
g
e
is
esp
ec
ially
r
elev
an
t
f
o
r
p
r
iv
ate
u
n
iv
er
s
ities
,
wh
er
e
s
tu
d
en
ts
’
e
n
r
o
llm
en
t d
ec
is
io
n
s
m
a
y
b
e
in
f
lu
en
ce
d
b
y
in
s
titu
tio
n
al
r
e
p
u
t
atio
n
,
af
f
o
r
d
ab
ilit
y
,
f
in
an
cial
s
u
p
p
o
r
t,
ac
ad
e
m
ic
co
n
s
id
er
atio
n
s
,
an
d
en
g
ag
em
en
t
with
u
n
iv
er
s
ity
in
f
o
r
m
ati
o
n
ch
an
n
els.
As
a
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2502
-
4
7
5
2
E
xp
la
in
a
b
le
h
yb
r
id
ma
c
h
in
e
l
ea
r
n
in
g
fo
r
p
r
ed
ictin
g
s
tu
d
en
t
en
r
o
llmen
t d
ec
is
io
n
s
…
(
Dyb
io
Do
mp
u
Ho
t
A
s
ih
)
881
r
esu
lt,
in
s
titu
tio
n
s
n
ee
d
an
al
y
tical
ap
p
r
o
ac
h
es
th
at
ca
n
i
d
en
tify
ap
p
lican
ts
with
a
h
ig
h
p
r
o
b
ab
ilit
y
o
f
en
r
o
llm
en
t w
h
ile
p
r
o
v
i
d
in
g
u
s
ef
u
l in
f
o
r
m
atio
n
f
o
r
r
ec
r
u
itm
e
n
t p
lan
n
in
g
an
d
r
eso
u
r
ce
allo
c
atio
n
[
1
]
–
[
4
]
.
E
d
u
ca
tio
n
al
d
ata
m
in
i
n
g
a
n
d
lear
n
in
g
an
aly
tics
h
a
v
e
e
x
p
an
d
ed
t
h
e
u
s
e
o
f
m
ac
h
in
e
le
ar
n
in
g
f
o
r
p
r
ed
ictin
g
d
if
f
e
r
en
t
e
d
u
ca
tio
n
al
o
u
tco
m
es,
in
clu
d
i
n
g
s
tu
d
e
n
t
p
er
f
o
r
m
a
n
ce
,
d
r
o
p
o
u
t
r
is
k
,
lear
n
in
g
b
e
h
av
io
r
,
an
d
en
r
o
llm
en
t
-
r
elate
d
d
ec
is
io
n
s
.
A
s
y
s
tem
atic
r
ev
iew
o
f
p
r
ed
ictiv
e
lear
n
in
g
an
aly
tics
s
h
o
ws
th
at
r
esear
ch
in
th
is
ar
ea
h
as
in
c
r
ea
s
in
g
ly
m
o
v
ed
f
r
o
m
co
n
v
en
tio
n
al
s
tatis
ti
ca
l
ap
p
r
o
ac
h
es
to
war
d
m
ac
h
in
e
lear
n
in
g
m
et
h
o
d
s
ca
p
ab
le
o
f
h
an
d
lin
g
c
o
m
p
lex
r
elatio
n
s
h
ip
s
wi
th
in
ed
u
ca
tio
n
al
d
ata
[
1
]
.
Un
lik
e
tr
a
d
itio
n
al
m
o
d
els
th
at
o
f
ten
d
ep
en
d
o
n
p
r
ed
ef
in
e
d
r
elatio
n
s
h
ip
s
b
etwe
en
v
a
r
iab
les,
m
ac
h
in
e
lear
n
in
g
ca
n
ca
p
tu
r
e
n
o
n
lin
ea
r
p
atter
n
s
an
d
in
ter
ac
tio
n
s
th
at
m
ay
b
e
d
if
f
i
cu
lt
to
i
d
en
tify
u
s
in
g
co
n
v
en
tio
n
al
an
aly
tical
ap
p
r
o
ac
h
e
s
.
T
h
ese
ca
p
ab
ilit
ies
m
ak
e
m
ac
h
in
e
lear
n
in
g
in
cr
ea
s
in
g
ly
r
elev
an
t f
o
r
e
d
u
ca
tio
n
al
d
ec
is
io
n
-
s
u
p
p
o
r
t a
p
p
licatio
n
s
[
1
]
,
[
5
]
.
Stu
d
en
t
en
r
o
llm
en
t
p
r
ed
ictio
n
is
p
ar
ticu
lar
ly
im
p
o
r
ta
n
t
b
ec
au
s
e
th
e
n
u
m
b
er
o
f
ap
p
lican
ts
wh
o
u
ltima
tely
r
eg
is
ter
ca
n
d
ir
ec
t
ly
af
f
ec
t
in
s
titu
tio
n
al
p
lan
n
in
g
an
d
s
u
s
tain
ab
ilit
y
.
Un
iv
er
s
ities
n
ee
d
r
eliab
le
esti
m
ates
o
f
en
r
o
llm
en
t
o
u
tco
m
es
to
s
u
p
p
o
r
t
r
ec
r
u
itm
e
n
t
ca
m
p
aig
n
s
,
s
ch
o
lar
s
h
ip
a
llo
ca
tio
n
,
ac
ad
em
ic
ca
p
ac
ity
p
lan
n
in
g
,
f
ac
u
lty
wo
r
k
lo
a
d
m
an
ag
em
e
n
t,
a
n
d
f
i
n
an
cial
p
lan
n
in
g
.
Pre
v
io
u
s
s
tu
d
ies
h
a
v
e
d
em
o
n
s
tr
ated
th
e
f
ea
s
ib
ilit
y
o
f
u
s
in
g
m
ac
h
i
n
e
lear
n
in
g
to
p
r
ed
ict
en
r
o
llm
en
t
am
o
n
g
a
d
m
itted
ap
p
lican
ts
,
s
u
g
g
esti
n
g
th
at
p
r
ed
ictiv
e
m
o
d
els
ca
n
p
r
o
v
id
e
u
s
ef
u
l
in
f
o
r
m
atio
n
b
ef
o
r
e
f
in
al
r
eg
is
tr
atio
n
d
ec
is
io
n
s
ar
e
m
ad
e
[
6
]
–
[
8
]
.
Ho
wev
e
r
,
p
r
e
d
ictio
n
at
th
e
in
d
iv
id
u
al
ap
p
lican
t
lev
el
r
em
ain
s
ch
allen
g
in
g
b
ec
au
s
e
en
r
o
llm
en
t
d
ec
is
io
n
s
ar
e
af
f
ec
ted
b
y
a
c
o
m
b
in
atio
n
o
f
ac
ad
em
ic
,
f
in
an
c
ial,
in
s
titu
tio
n
al,
an
d
b
e
h
av
io
r
al
f
ac
to
r
s
.
E
x
is
tin
g
r
esear
ch
h
as
em
p
lo
y
ed
v
ar
io
u
s
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
s
f
o
r
ed
u
ca
tio
n
al
p
r
ed
ictio
n
,
in
clu
d
in
g
d
ec
is
io
n
tr
ee
s
,
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
i
n
es,
r
an
d
o
m
f
o
r
est,
g
r
a
d
ien
t
b
o
o
s
tin
g
,
a
n
d
o
th
er
class
if
icatio
n
tech
n
iq
u
es.
Fo
r
ex
am
p
le,
r
ese
ar
ch
o
n
u
n
d
er
g
r
ad
u
ate
ap
p
lica
n
t
en
r
o
llm
e
n
t
h
as
d
em
o
n
s
tr
ate
d
th
e
u
s
ef
u
ln
ess
o
f
b
in
ar
y
class
if
icatio
n
m
o
d
els
f
o
r
p
r
ed
ictin
g
wh
eth
e
r
ad
m
itted
ap
p
lican
ts
will
ev
en
tu
ally
e
n
r
o
ll
[
6
]
,
wh
ile
o
th
e
r
s
tu
d
ies
h
av
e
in
v
esti
g
ated
m
ac
h
in
e
lear
n
i
n
g
m
o
d
els
f
o
r
e
n
r
o
llm
en
t
an
d
attr
itio
n
p
r
ed
ictio
n
[
8
]
.
Mu
lti
-
m
o
d
el
ap
p
r
o
ac
h
es
h
av
e
also
b
ee
n
in
v
esti
g
ated
f
o
r
h
ig
h
e
r
ed
u
ca
tio
n
ad
m
is
s
io
n
p
r
ed
ictio
n
,
in
d
icatin
g
th
at
co
m
b
in
in
g
d
if
f
er
en
t
class
if
ier
s
ca
n
im
p
r
o
v
e
p
r
ed
ictiv
e
ca
p
ab
ilit
y
c
o
m
p
ar
ed
with
r
ely
i
n
g
o
n
a
s
i
n
g
le
m
o
d
el
[
9
]
,
[
1
0
]
.
Nev
er
th
eless
,
m
u
ch
o
f
th
is
r
esear
ch
em
p
h
asizes
p
r
ed
ictiv
e
p
er
f
o
r
m
an
ce
,
wh
er
ea
s
t
h
e
r
ea
s
o
n
s
u
n
d
er
ly
in
g
in
d
iv
id
u
al
p
r
ed
ictio
n
s
ar
e
o
f
te
n
less
th
o
r
o
u
g
h
ly
ex
a
m
in
ed
.
T
h
is
lim
itatio
n
b
ec
o
m
es
in
cr
e
asin
g
ly
i
m
p
o
r
ta
n
t
as
m
ac
h
in
e
lear
n
in
g
is
u
s
ed
t
o
s
u
p
p
o
r
t
in
s
titu
tio
n
al
d
ec
is
io
n
s
.
A
m
o
d
el
th
at
ac
h
ie
v
es
h
ig
h
p
r
ed
ictiv
e
ac
cu
r
ac
y
m
ay
s
till
h
av
e
lim
ited
p
r
ac
tica
l
v
alu
e
if
a
d
m
is
s
io
n
o
f
f
icer
s
an
d
u
n
i
v
er
s
ity
m
an
a
g
er
s
ca
n
n
o
t
u
n
d
er
s
tan
d
wh
y
a
p
ar
ticu
lar
a
p
p
lican
t
is
cla
s
s
if
ied
as
lik
ely
o
r
u
n
lik
ely
t
o
e
n
r
o
ll.
E
x
p
lain
a
b
le
ar
tific
ial
in
tellig
en
ce
(
XAI
)
ad
d
r
ess
es
th
is
is
s
u
e
b
y
p
r
o
v
id
in
g
in
te
r
p
r
etab
l
e
in
f
o
r
m
atio
n
ab
o
u
t
h
o
w
in
p
u
t
v
ar
iab
les
co
n
tr
ib
u
te
t
o
m
o
d
el
p
r
ed
ictio
n
s
.
T
h
e
b
r
o
ad
er
XAI
liter
atu
r
e
em
p
h
asizes
th
at
ex
p
lain
ab
il
ity
ca
n
s
tr
en
g
th
en
tr
an
s
p
ar
e
n
cy
,
ac
co
u
n
tab
ilit
y
,
an
d
tr
u
s
t
in
ar
tific
ial
in
tellig
en
ce
s
y
s
tem
s
,
p
ar
ticu
lar
ly
wh
e
n
p
r
ed
ictio
n
s
in
f
lu
en
ce
co
n
s
eq
u
en
tial
d
ec
is
io
n
s
[
1
1
]
.
I
n
e
d
u
ca
tio
n
al
s
ettin
g
s
,
ex
p
lain
ab
le
m
ac
h
in
e
lear
n
in
g
h
as
also
b
ee
n
s
h
o
wn
t
o
h
el
p
p
r
ac
titi
o
n
er
s
u
n
d
er
s
tan
d
t
h
e
f
ac
to
r
s
u
n
d
er
ly
in
g
p
r
ed
ictio
n
s
an
d
ass
ess
wh
eth
e
r
m
o
d
el
o
u
tp
u
ts
ar
e
c
o
n
s
is
ten
t w
ith
d
o
m
ain
k
n
o
wled
g
e
[5
]
,
[
1
2
]
,
[
1
3
]
.
R
ec
en
t
ed
u
ca
tio
n
al
s
tu
d
ies
p
r
o
v
id
e
ev
i
d
en
ce
o
f
th
e
v
alu
e
o
f
ex
p
lain
ab
ilit
y
,
alth
o
u
g
h
th
eir
ap
p
licatio
n
s
h
av
e
lar
g
ely
f
o
c
u
s
ed
o
n
o
u
tco
m
es
o
th
er
th
an
en
r
o
llm
en
t.
E
x
p
lain
a
b
le
m
ac
h
in
e
lear
n
in
g
h
as
b
ee
n
ap
p
lied
to
s
tu
d
en
t
p
er
f
o
r
m
a
n
c
e
p
r
ed
ictio
n
,
d
r
o
p
o
u
t
p
r
e
d
ictio
n
,
s
tu
d
en
t
ad
ap
ta
b
ilit
y
,
an
d
ea
r
ly
id
en
tific
atio
n
o
f
at
-
r
is
k
s
tu
d
en
ts
[5
]
,
[
12
]
,
[
1
4
]
–
[
1
7
]
.
T
h
ese
s
tu
d
ies
d
em
o
n
s
tr
ate
th
at
ex
p
lain
ab
ilit
y
ca
n
co
m
p
lem
en
t
p
r
e
d
ictiv
e
p
er
f
o
r
m
an
ce
b
y
r
ev
ea
lin
g
in
f
lu
en
tial
v
ar
iab
les
an
d
s
u
p
p
o
r
tin
g
m
o
r
e
in
f
o
r
m
ed
in
ter
v
en
t
io
n
s
.
Fo
r
ex
am
p
le,
ex
p
lain
ab
le
d
r
o
p
o
u
t
p
r
ed
icti
o
n
h
as
b
ee
n
u
s
ed
to
id
e
n
tify
f
ac
to
r
s
ass
o
ciate
d
with
s
tu
d
en
t
r
is
k
,
w
h
ile
ex
p
lain
ab
le
lear
n
i
n
g
an
al
y
tics
h
as
h
ig
h
lig
h
ted
th
e
im
p
o
r
tan
ce
o
f
ex
am
in
i
n
g
th
e
s
tab
ilit
y
a
n
d
in
ter
p
r
etatio
n
o
f
p
r
ed
ictiv
e
m
o
d
els
r
ath
er
th
a
n
r
ely
in
g
s
o
lely
o
n
th
ei
r
ac
cu
r
ac
y
[
1
3
]
-
[
15
]
.
T
h
ese
f
in
d
in
g
s
s
u
g
g
est
th
at
ex
p
lain
ab
ilit
y
s
h
o
u
ld
b
e
co
n
s
id
er
ed
as
p
ar
t
o
f
th
e
p
r
ed
ic
tiv
e
f
r
am
ewo
r
k
its
elf
r
at
h
er
th
an
tr
ea
ted
as
an
ad
d
itio
n
al
f
ea
tu
r
e
af
ter
m
o
d
el
d
ev
elo
p
m
e
n
t.
T
h
e
liter
atu
r
e
o
n
s
tu
d
en
t
ad
m
is
s
io
n
an
d
e
n
r
o
llm
en
t
p
r
o
v
i
d
es
a
m
o
r
e
d
ir
ec
t
b
asis
f
o
r
t
h
e
p
r
esen
t
s
tu
d
y
.
R
ec
en
t
r
esear
ch
h
as
d
e
m
o
n
s
tr
ated
th
at
m
ac
h
in
e
lear
n
in
g
ca
n
p
r
e
d
ict
en
r
o
llm
e
n
t
am
o
n
g
u
n
d
er
g
r
ad
u
ate
ap
p
lican
ts
an
d
ad
m
itted
s
tu
d
e
n
ts
[6
]
,
[
7
]
.
Oth
er
s
tu
d
ie
s
h
a
v
e
in
v
esti
g
ated
th
e
s
elec
tio
n
o
f
in
co
m
in
g
f
r
esh
m
en
u
s
in
g
m
ac
h
in
e
lear
n
in
g
an
d
h
av
e
s
h
o
wn
th
at
p
r
ed
ictiv
e
m
o
d
els
ca
n
ass
i
s
t
in
s
titu
tio
n
al
a
d
m
is
s
io
n
p
r
o
ce
s
s
es
[
1
8
]
.
Similar
ly
,
en
r
o
llm
en
t
p
r
ed
ictio
n
h
as
b
ee
n
ex
am
in
e
d
u
s
in
g
m
ac
h
in
e
l
ea
r
n
in
g
in
u
n
i
v
er
s
ity
ed
u
ca
tio
n
al
en
v
ir
o
n
m
en
ts
,
co
n
f
ir
m
in
g
t
h
e
p
o
ten
tial
o
f
p
r
ed
ictiv
e
an
aly
ti
cs
f
o
r
in
s
titu
tio
n
al
p
la
n
n
in
g
[
1
9
]
.
Ho
we
v
er
,
th
ese
s
tu
d
ies
p
r
im
ar
ily
co
n
ce
n
tr
ate
o
n
p
r
e
d
ictio
n
o
u
tco
m
es
an
d
d
o
n
o
t
f
u
lly
in
te
g
r
ate
p
r
e
d
ic
tiv
e
m
o
d
elin
g
with
m
ec
h
an
is
m
s
f
o
r
e
x
p
lain
in
g
th
e
co
n
tr
ib
u
tio
n
o
f
in
d
iv
i
d
u
al
ap
p
lican
t c
h
ar
ac
ter
is
tics
.
E
n
s
em
b
le
lear
n
in
g
r
e
p
r
esen
t
s
an
o
th
er
im
p
o
r
ta
n
t
d
ev
elo
p
m
en
t
in
th
is
r
esear
ch
ar
ea
.
C
o
m
b
in
in
g
m
u
ltip
le
class
if
ier
s
ca
n
ex
p
l
o
it
co
m
p
lem
en
tar
y
lear
n
in
g
p
at
ter
n
s
an
d
p
o
ten
tially
im
p
r
o
v
e
g
en
e
r
aliza
tio
n
co
m
p
ar
ed
with
an
in
d
iv
id
u
al
c
lass
if
ier
.
I
n
h
ig
h
er
ed
u
ca
tio
n
a
d
m
is
s
io
n
p
r
ed
ictio
n
,
Z
u
b
et
a
l
.
d
ev
elo
p
e
d
a
two
-
s
tag
e
PNN
–
SVM
en
s
em
b
le
an
d
d
em
o
n
s
tr
ated
th
e
p
o
ten
t
ial
o
f
m
u
lti
-
m
o
d
el
a
r
ch
itectu
r
es
f
o
r
im
p
r
o
v
in
g
ad
m
is
s
io
n
p
r
ed
ictio
n
p
er
f
o
r
m
an
ce
[
1
0
]
.
L
i
et
a
l
.
[
9
]
lik
ew
is
e
p
r
o
p
o
s
ed
a
m
u
lti
-
m
o
d
el
an
aly
tics
s
y
s
tem
to
s
u
p
p
o
r
t
s
tu
d
e
n
t
ad
m
is
s
io
n
d
ec
is
io
n
-
m
ak
in
g
.
T
h
ese
s
tu
d
ies
p
r
o
v
id
e
e
v
id
en
ce
th
at
m
u
lti
-
m
o
d
el
ap
p
r
o
ac
h
es
ar
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
5
2
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
4
3
,
No
.
3
,
Sep
tem
b
er
20
2
6
:
8
8
0
-
89
7
882
p
r
o
m
is
in
g
f
o
r
h
ig
h
e
r
ed
u
ca
tio
n
an
aly
tics
.
Ho
wev
e
r
,
th
e
y
d
o
n
o
t
f
u
lly
ad
d
r
ess
th
e
in
teg
r
atio
n
o
f
h
y
b
r
id
f
ea
tu
r
e
s
elec
tio
n
,
ad
ap
tiv
e
en
s
em
b
le
l
ea
r
n
in
g
,
a
n
d
e
x
p
lain
ab
le
p
r
ed
i
ctio
n
with
in
o
n
e
f
r
am
ewo
r
k
.
An
o
th
er
s
tr
ea
m
o
f
r
esear
ch
h
a
s
f
o
cu
s
ed
o
n
co
u
r
s
e
en
r
o
llm
e
n
t
an
d
b
r
o
ad
e
r
en
r
o
llm
en
t
m
a
n
ag
em
en
t.
Ma
ch
in
e
lear
n
i
n
g
h
as
b
ee
n
u
s
ed
to
p
r
ed
ict
c
o
u
r
s
e
e
n
r
o
ll
m
en
t
p
atter
n
s
to
s
u
p
p
o
r
t
ac
a
d
em
ic
p
lan
n
in
g
an
d
r
eso
u
r
ce
allo
ca
tio
n
[
2
0
]
,
[
2
1
]
.
Oth
er
wo
r
k
h
as
ex
am
in
e
d
s
tu
d
en
t
en
r
o
llm
en
t
an
d
attr
itio
n
s
im
u
ltan
e
o
u
s
ly
,
d
em
o
n
s
tr
atin
g
th
at
p
r
ed
ictiv
e
an
aly
tics
ca
n
s
u
p
p
o
r
t
b
o
th
r
ec
r
u
itm
en
t
an
d
r
et
en
tio
n
p
la
n
n
in
g
[
8
]
.
Alth
o
u
g
h
th
ese
s
tu
d
ies
ar
e
r
elev
an
t
to
en
r
o
llm
en
t
m
a
n
ag
em
e
n
t,
co
u
r
s
e
-
lev
el
en
r
o
llm
en
t
an
d
in
s
titu
tio
n
al
-
lev
el
f
o
r
ec
asti
n
g
ar
e
co
n
ce
p
tu
ally
d
if
f
er
en
t
f
r
o
m
p
r
ed
ictin
g
an
in
d
iv
id
u
al
ap
p
lican
t'
s
d
ec
i
s
io
n
to
en
r
o
ll
af
ter
r
ec
eiv
in
g
a
n
ad
m
is
s
io
n
o
p
p
o
r
tu
n
ity
.
I
n
d
iv
id
u
al
en
r
o
llm
en
t
d
ec
is
io
n
s
m
ay
d
e
p
en
d
o
n
in
s
titu
tio
n
al
ch
ar
ac
ter
is
tics
,
p
er
ce
iv
ed
v
alu
e,
f
in
an
cial
co
n
s
id
er
atio
n
s
,
an
d
co
m
m
u
n
icatio
n
o
r
en
g
a
g
em
en
t
f
ac
to
r
s
th
at
ar
e
n
o
t n
ec
ess
ar
ily
r
ep
r
esen
te
d
in
co
u
r
s
e
-
lev
el
o
r
ag
g
r
eg
ate
f
o
r
e
ca
s
tin
g
m
o
d
els.
T
h
e
d
is
tin
ctio
n
b
etwe
en
in
d
i
v
id
u
al
p
r
ed
ictio
n
an
d
ag
g
r
eg
ate
f
o
r
ec
asti
n
g
is
th
er
ef
o
r
e
i
m
p
o
r
tan
t.
T
im
e
-
s
er
ies
an
d
o
th
er
f
o
r
ec
as
tin
g
ap
p
r
o
ac
h
es
ca
n
p
r
o
v
id
e
u
s
ef
u
l
esti
m
ates
o
f
f
u
tu
r
e
en
r
o
llm
en
t
tr
en
d
s
,
b
u
t
th
ey
g
e
n
er
ally
o
p
er
ate
at
an
a
g
g
r
eg
ate
in
s
titu
tio
n
al
o
r
tem
p
o
r
al
lev
el.
I
n
co
n
tr
ast,
a
n
ap
p
l
ican
t
-
lev
el
m
ac
h
in
e
lear
n
in
g
f
r
a
m
ewo
r
k
ca
n
u
s
e
m
u
ltid
im
en
s
io
n
al
ch
ar
ac
ter
i
s
tics
to
e
s
tim
ate
th
e
p
r
o
b
ab
ilit
y
th
at
a
s
p
ec
if
ic
p
r
o
s
p
ec
tiv
e
s
tu
d
en
t
will
en
r
o
ll
.
Fo
r
in
s
titu
tio
n
al
r
ec
r
u
itm
en
t,
th
is
d
is
tin
ctio
n
h
as
p
r
ac
tical
i
m
p
o
r
tan
ce
b
ec
au
s
e
d
ec
is
io
n
-
m
ak
er
s
m
ay
n
ee
d
to
id
en
tify
wh
ich
ap
p
lican
ts
s
h
o
u
ld
r
ec
eiv
e
tar
g
eted
c
o
m
m
u
n
icatio
n
,
f
in
an
cial
in
f
o
r
m
atio
n
,
o
r
o
th
er
r
ec
r
u
itm
en
t
in
ter
v
en
tio
n
s
r
ath
er
th
an
s
im
p
ly
esti
m
ate
th
e
to
tal
n
u
m
b
er
o
f
f
u
t
u
r
e
s
tu
d
en
ts
.
B
ey
o
n
d
p
r
e
d
i
ctiv
e
m
o
d
elin
g
,
r
esear
ch
o
n
u
n
iv
er
s
ity
ch
o
ice
s
u
g
g
ests
th
at
en
r
o
llm
en
t
d
ec
is
io
n
s
ar
e
s
h
ap
ed
b
y
s
ev
er
al
in
s
titu
tio
n
al
an
d
s
o
ci
o
ec
o
n
o
m
ic
co
n
s
id
er
atio
n
s
.
R
ec
en
t
ev
id
e
n
ce
f
r
o
m
p
r
iv
ate
h
i
g
h
er
ed
u
ca
tio
n
in
d
icate
s
th
at
in
s
titu
tio
n
al
r
e
p
u
tatio
n
an
d
lo
ca
ti
o
n
ca
n
in
f
l
u
en
ce
s
tu
d
en
ts
’
u
n
i
v
er
s
ity
ch
o
ices
[
2
]
.
R
esear
ch
o
n
en
r
o
llm
en
t
m
a
n
a
g
em
en
t
also
em
p
h
asizes
th
e
im
p
o
r
tan
ce
o
f
u
n
d
er
s
tan
d
in
g
s
tu
d
en
ts
’
p
r
ef
e
r
en
ce
s
an
d
d
ec
is
io
n
-
m
ak
in
g
p
r
o
ce
s
s
es
wh
en
d
ev
elo
p
in
g
i
n
s
titu
tio
n
al
r
ec
r
u
itm
en
t
s
tr
ateg
ie
s
[
3
]
.
Stu
d
ies
o
f
in
ter
n
atio
n
al
a
n
d
d
o
m
esti
c
u
n
iv
er
s
ity
ch
o
ice
f
u
r
t
h
er
d
em
o
n
s
tr
ate
th
at
s
tu
d
en
ts
m
ay
co
n
s
i
d
er
m
u
ltip
le
f
ac
to
r
s
wh
en
s
elec
tin
g
a
h
ig
h
er
ed
u
ca
tio
n
in
s
titu
tio
n
,
wh
ile
f
in
an
cial
s
u
p
p
o
r
t
an
d
af
f
o
r
d
a
b
ilit
y
ca
n
af
f
ec
t
h
ig
h
e
r
ed
u
ca
tio
n
c
h
o
ices
[4
]
,
[
2
2
]
.
T
h
ese
f
in
d
in
g
s
ar
e
im
p
o
r
tan
t
f
o
r
m
ac
h
i
n
e
lear
n
i
n
g
-
b
ased
en
r
o
llm
en
t
p
r
ed
ictio
n
b
ec
au
s
e
th
ey
in
d
icate
th
at
p
r
ed
ictio
n
m
o
d
els
s
h
o
u
ld
co
n
s
id
er
h
eter
o
g
e
n
eo
u
s
f
ac
t
o
r
s
r
ath
er
th
an
r
ely
in
g
ex
clu
s
iv
ely
o
n
ac
a
d
em
ic
v
a
r
iab
les.
T
h
e
im
p
o
r
tan
ce
o
f
th
ese
f
ac
to
r
s
also
r
ein
f
o
r
ce
s
th
e
n
ee
d
f
o
r
in
ter
p
r
etab
le
m
o
d
els.
I
f
an
e
n
r
o
llm
en
t
p
r
ed
ictio
n
m
o
d
el
id
en
tifie
s
in
s
titu
tio
n
al
r
ep
u
tatio
n
,
f
in
an
c
ial
s
u
p
p
o
r
t,
af
f
o
r
d
ab
ilit
y
,
d
ig
ital
en
g
ag
em
en
t,
o
r
ac
ad
em
ic
f
ac
ili
ties
as
in
f
lu
e
n
tial
v
ar
iab
les,
u
n
iv
er
s
ity
m
a
n
a
g
er
s
n
ee
d
to
u
n
d
er
s
tan
d
t
h
e
d
i
r
ec
tio
n
an
d
r
elativ
e
co
n
tr
ib
u
tio
n
o
f
th
o
s
e
v
ar
iab
le
s
b
ef
o
r
e
tr
an
s
latin
g
p
r
e
d
ictio
n
s
in
to
r
ec
r
u
itm
en
t
ac
tio
n
s
.
E
x
p
lain
ab
le
m
o
d
els
ca
n
p
r
o
v
id
e
th
is
a
d
d
itio
n
al
l
ay
er
o
f
d
ec
is
io
n
s
u
p
p
o
r
t
b
y
co
n
n
ec
tin
g
p
r
e
d
ictiv
e
o
u
t
p
u
ts
with
in
ter
p
r
etab
l
e
f
ea
tu
r
e
co
n
tr
ib
u
tio
n
s
[
1
1
]
,
[
13
]
,
[
2
3
]
.
C
o
n
s
eq
u
en
tly
,
a
u
s
e
f
u
l
en
r
o
llm
en
t
p
r
ed
ictio
n
f
r
a
m
ewo
r
k
s
h
o
u
ld
n
o
t
o
n
ly
an
s
wer
wh
eth
e
r
an
a
p
p
l
ican
t
is
lik
ely
to
en
r
o
ll
b
u
t
s
h
o
u
ld
also
p
r
o
v
i
d
e
ev
id
e
n
ce
r
eg
ar
d
in
g
wh
y
t
h
e
m
o
d
el
r
ea
ch
es th
at
p
r
ed
ictio
n
.
Desp
ite
th
ese
ad
v
an
ce
s
,
an
i
m
p
o
r
tan
t
g
a
p
r
em
ain
s
in
th
e
liter
atu
r
e.
First,
p
r
ev
io
u
s
en
r
o
llm
en
t
an
d
ad
m
is
s
io
n
s
tu
d
ies
h
av
e
lar
g
el
y
ev
alu
ated
i
n
d
iv
id
u
al
m
ac
h
i
n
e
lear
n
in
g
m
o
d
els
o
r
co
n
v
e
n
tio
n
al
m
u
lti
-
m
o
d
el
ap
p
r
o
ac
h
es
[6
]
,
[
7
]
,
[
9
]
,
[
10
]
,
[
18
]
,
[
19
]
.
Seco
n
d
,
r
esear
ch
o
n
ex
p
lain
a
b
le
ed
u
ca
tio
n
al
m
ac
h
in
e
lear
n
in
g
h
as
p
r
ed
o
m
i
n
an
tly
f
o
cu
s
e
d
o
n
s
tu
d
en
t
p
e
r
f
o
r
m
an
ce
,
d
r
o
p
o
u
t,
ad
ap
tab
ilit
y
,
o
r
ac
ad
em
ic
r
is
k
r
ath
er
th
a
n
p
o
s
t
-
ad
m
is
s
io
n
en
r
o
llm
en
t
d
ec
is
io
n
s
[5
]
,
[
12
]
,
[
1
4
]
-
[
17
]
.
T
h
ir
d
,
ex
is
tin
g
en
s
em
b
le
ap
p
r
o
ac
h
es
d
o
n
o
t
co
n
s
is
ten
tly
in
teg
r
ate
f
ea
tu
r
e
s
elec
tio
n
an
d
ex
p
lain
a
b
ilit
y
in
to
a
u
n
if
ied
ap
p
lican
t
-
lev
el
e
n
r
o
llm
e
n
t
p
r
e
d
ictio
n
f
r
am
ewo
r
k
[9
]
,
[
1
0
]
.
Fo
u
r
t
h
,
r
elativ
ely
lim
ited
ev
id
en
ce
is
av
ailab
le
o
n
ex
p
lain
ab
le
en
r
o
llm
en
t
p
r
ed
ictio
n
in
I
n
d
o
n
esia
n
p
r
iv
ate
u
n
i
v
er
s
ities
,
d
esp
ite
th
e
co
m
p
etitiv
e
r
ec
r
u
itm
en
t e
n
v
ir
o
n
m
en
t
f
ac
ed
b
y
th
ese
in
s
titu
tio
n
s
.
T
h
e
m
eth
o
d
o
lo
g
ical
g
ap
is
p
ar
ticu
lar
ly
r
elev
an
t
b
ec
au
s
e
co
m
b
in
in
g
co
m
p
lem
en
tar
y
m
ac
h
in
e
lear
n
in
g
m
o
d
els
d
o
es
n
o
t
au
to
m
atica
lly
g
u
ar
an
tee
th
at
th
e
r
esu
ltin
g
s
y
s
tem
will
b
e
in
ter
p
r
etab
le
o
r
ef
f
icien
t.
Featu
r
e
s
elec
tio
n
ca
n
r
ed
u
ce
ir
r
elev
an
t
an
d
r
e
d
u
n
d
an
t
in
f
o
r
m
atio
n
b
ef
o
r
e
m
o
d
el
tr
ain
in
g
,
wh
ile
en
s
em
b
le
lear
n
in
g
ca
n
co
m
b
in
e
th
e
s
tr
en
g
th
s
o
f
d
if
f
er
e
n
t
class
if
ier
s
.
E
x
p
lain
ab
ilit
y
ca
n
th
en
b
e
u
s
ed
to
id
en
tify
th
e
co
n
tr
ib
u
tio
n
o
f
in
d
i
v
id
u
al
p
r
e
d
icto
r
s
to
th
e
f
in
al
m
o
d
el
o
u
tp
u
t.
T
h
e
i
n
teg
r
atio
n
o
f
th
ese
co
m
p
o
n
e
n
ts
p
r
o
v
i
d
es
an
o
p
p
o
r
t
u
n
ity
t
o
m
o
v
e
b
e
y
o
n
d
a
c
o
n
v
e
n
tio
n
al
p
r
ed
ic
tio
n
-
o
n
ly
ap
p
r
o
ac
h
to
war
d
a
d
ec
is
io
n
-
s
u
p
p
o
r
t
f
r
am
ewo
r
k
th
at
s
im
u
ltan
eo
u
s
l
y
ad
d
r
ess
es p
r
ed
ictiv
e
p
er
f
o
r
m
an
ce
an
d
in
ter
p
r
eta
b
ilit
y
[
1
1
]
,
[
2
4
]
–
[
2
8
]
.
Acc
o
r
d
in
g
ly
,
th
is
s
tu
d
y
p
r
o
p
o
s
es
an
ex
p
lain
a
b
le
h
y
b
r
id
m
a
ch
in
e
lear
n
in
g
f
r
am
ewo
r
k
(
E
HM
L
F)
f
o
r
p
r
ed
ictin
g
s
tu
d
e
n
t
en
r
o
llm
en
t
d
ec
is
io
n
s
in
I
n
d
o
n
esian
p
r
iv
ate
u
n
iv
e
r
s
ities
,
u
s
in
g
ad
m
is
s
io
n
d
ata
f
r
o
m
Un
iv
er
s
itas
W
ir
ah
u
s
ad
a
Me
d
an
an
d
Un
i
v
er
s
itas
Au
d
i
I
n
d
o
n
esia.
T
h
e
f
r
a
m
ewo
r
k
co
m
b
in
es
h
y
b
r
id
f
ea
tu
r
e
s
elec
tio
n
,
ad
ap
tiv
e
e
n
s
em
b
le
l
ea
r
n
in
g
,
an
d
XAI
to
i
d
en
tify
p
r
o
s
p
ec
tiv
e
s
tu
d
en
ts
wh
o
ar
e
li
k
ely
to
e
n
r
o
ll
wh
ile
also
r
ev
ea
lin
g
th
e
f
ac
to
r
s
th
at
co
n
tr
ib
u
te
to
th
e
p
r
e
d
ictio
n
s
.
T
h
e
f
r
am
ew
o
r
k
is
d
esig
n
ed
to
co
m
b
in
e
th
e
co
m
p
lem
en
ta
r
y
s
tr
en
g
th
s
o
f
m
u
ltip
le
class
if
ier
s
r
ath
er
th
an
r
ely
in
g
o
n
a
s
in
g
le
p
r
ed
ictiv
e
a
lg
o
r
ith
m
,
w
h
ile
th
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2502
-
4
7
5
2
E
xp
la
in
a
b
le
h
yb
r
id
ma
c
h
in
e
l
ea
r
n
in
g
fo
r
p
r
ed
ictin
g
s
tu
d
en
t
en
r
o
llmen
t d
ec
is
io
n
s
…
(
Dyb
io
Do
mp
u
Ho
t
A
s
ih
)
883
ex
p
lain
ab
ilit
y
co
m
p
o
n
e
n
t
is
i
n
ten
d
ed
to
m
a
k
e
th
e
r
esu
ltin
g
p
r
ed
ictio
n
s
m
o
r
e
tr
an
s
p
ar
e
n
t
f
o
r
in
s
titu
tio
n
al
d
ec
is
io
n
-
m
ak
er
s
.
T
h
is
s
tu
d
y
m
ak
es
th
r
ee
m
ain
co
n
tr
ib
u
tio
n
s
.
First,
it
d
ev
el
o
p
s
an
ap
p
lican
t
-
lev
el
h
y
b
r
i
d
p
r
ed
ictio
n
f
r
am
ewo
r
k
th
at
in
teg
r
ates
f
ea
t
u
r
e
s
elec
tio
n
an
d
m
u
ltip
le
m
a
ch
in
e
lear
n
in
g
class
if
ier
s
f
o
r
s
tu
d
en
t
en
r
o
llm
e
n
t
p
r
ed
ictio
n
,
ex
ten
d
in
g
p
r
e
v
io
u
s
ad
m
is
s
io
n
s
tu
d
ies
th
at
p
r
im
ar
ily
r
ely
o
n
in
d
i
v
id
u
al
o
r
co
n
v
en
tio
n
al
en
s
em
b
le
m
o
d
els
[6
]
,
[
9
]
,
[
10
]
,
[
18
]
,
[
19
]
.
Seco
n
d
,
it
in
co
r
p
o
r
ates
ex
p
lain
ab
ilit
y
d
ir
ec
tly
in
to
th
e
p
r
ed
ictio
n
p
ip
elin
e
,
ad
d
r
ess
in
g
th
e
lim
itatio
n
o
f
p
r
ev
io
u
s
en
r
o
llm
en
t
s
tu
d
ies
th
at
em
p
h
asize
p
r
ed
ictiv
e
ac
cu
r
ac
y
with
o
u
t
s
u
f
f
icien
tly
ex
p
lain
in
g
th
e
d
et
er
m
in
an
ts
o
f
in
d
iv
i
d
u
al
p
r
e
d
ictio
n
s
[
1
1
]
,
[
13
]
,
[
2
3
]
.
T
h
ir
d
,
it
p
r
o
v
id
es
em
p
ir
ical
ev
id
en
ce
f
r
o
m
two
I
n
d
o
n
esia
n
p
r
iv
ate
u
n
iv
e
r
s
ities
an
d
id
en
tifie
s
th
e
r
elativ
e
im
p
o
r
tan
ce
o
f
in
s
titu
tio
n
al,
f
in
an
cial,
ac
a
d
em
i
c,
a
n
d
en
g
a
g
em
en
t
-
r
elate
d
f
ac
to
r
s
in
e
n
r
o
llm
en
t
d
ec
is
io
n
s
,
t
h
er
eb
y
co
n
n
ec
tin
g
p
r
e
d
ictiv
e
an
aly
tics
with
p
r
ac
tical
u
n
iv
er
s
ity
r
ec
r
u
itm
en
t a
n
d
d
ec
is
io
n
-
m
ak
in
g
n
ee
d
s
[2
]
-
[
4
]
,
[
22
]
.
T
h
er
ef
o
r
e,
th
e
c
o
n
tr
ib
u
tio
n
o
f
E
HM
L
F
is
n
o
t
lim
ited
to
o
b
tain
in
g
a
h
ig
h
-
p
e
r
f
o
r
m
in
g
cla
s
s
if
icatio
n
m
o
d
el.
T
h
e
f
r
am
ewo
r
k
s
ee
k
s
to
p
r
o
v
id
e
a
m
o
r
e
co
m
p
lete
d
ec
is
io
n
-
s
u
p
p
o
r
t
m
ec
h
an
is
m
in
wh
ich
p
r
e
d
ictio
n
,
f
ea
tu
r
e
s
elec
tio
n
,
en
s
em
b
le
lear
n
in
g
,
a
n
d
e
x
p
lan
atio
n
ar
e
co
n
s
id
er
ed
t
o
g
eth
er
.
B
y
ex
p
licitly
co
m
p
ar
in
g
p
r
ed
ictiv
e
p
er
f
o
r
m
an
c
e
an
d
in
f
lu
en
tial
f
ac
to
r
s
with
f
in
d
in
g
s
f
r
o
m
p
r
e
v
io
u
s
ad
m
is
s
io
n
,
en
r
o
llm
en
t,
an
d
ex
p
lain
ab
le
e
d
u
ca
tio
n
al
m
ac
h
in
e
lear
n
in
g
s
tu
d
ies,
th
is
r
esear
ch
aim
s
to
s
tr
en
g
th
e
n
th
e
m
eth
o
d
o
l
o
g
ical
an
d
p
r
ac
tical
u
n
d
e
r
s
tan
d
in
g
o
f
ex
p
lain
ab
le
s
tu
d
en
t e
n
r
o
llm
en
t p
r
ed
ictio
n
in
I
n
d
o
n
esian
p
r
iv
ate
h
ig
h
er
e
d
u
ca
tio
n
.
2.
M
E
T
H
O
D
2
.
1
.
Resea
rc
h
d
esig
n
T
h
is
s
tu
d
y
em
p
lo
y
ed
a
q
u
a
n
titativ
e
p
r
ed
ictiv
e
an
aly
tics
d
esig
n
to
d
ev
elo
p
an
E
HM
L
F
f
o
r
p
r
ed
ictin
g
s
tu
d
en
t
en
r
o
llm
e
n
t
d
ec
is
io
n
s
in
I
n
d
o
n
esian
p
r
iv
ate
u
n
iv
e
r
s
ities
.
T
h
e
r
esear
ch
was
d
esi
g
n
ed
n
o
t
o
n
ly
to
m
ax
im
ize
class
if
icatio
n
p
er
f
o
r
m
an
ce
b
u
t
also
to
id
e
n
tify
th
e
v
ar
iab
les
th
at
c
o
n
tr
ib
u
te
to
in
d
iv
id
u
al
en
r
o
llm
en
t
p
r
ed
ictio
n
s
.
T
h
is
ap
p
r
o
ac
h
is
co
n
s
is
ten
t
with
th
e
g
r
o
win
g
u
s
e
o
f
p
r
ed
ictiv
e
lear
n
in
g
a
n
aly
tics
an
d
e
x
p
lain
ab
le
m
ac
h
in
e
lear
n
i
n
g
in
h
ig
h
er
e
d
u
ca
tio
n
,
wh
er
e
p
r
ed
ictiv
e
ac
cu
r
ac
y
n
ee
d
s
to
b
e
ac
co
m
p
a
n
ied
b
y
tr
an
s
p
ar
e
n
t
in
ter
p
r
etatio
n
o
f
m
o
d
el
o
u
t
p
u
t
s
[1
]
,
[
11
]
,
[
13
]
.
T
h
e
p
r
o
p
o
s
ed
E
HM
L
F
in
teg
r
ates
f
iv
e
s
eq
u
en
tial
s
tag
es:
1
)
d
ata
p
r
ep
ar
atio
n
,
2
)
f
ea
tu
r
e
e
n
g
in
ee
r
in
g
an
d
p
r
ep
r
o
ce
s
s
in
g
,
3
)
h
y
b
r
id
f
ea
tu
r
e
s
elec
tio
n
,
4
)
ad
ap
tiv
e
en
s
em
b
le
class
if
ica
tio
n
,
an
d
5
)
ex
p
lain
ab
ilit
y
an
aly
s
is
u
s
in
g
Sh
ap
ley
ad
d
i
tiv
e
ex
p
lan
atio
n
s
(
SHAP).
T
h
e
f
r
am
ewo
r
k
was
d
esig
n
e
d
to
ad
d
r
ess
th
r
ee
m
eth
o
d
o
l
o
g
ical
r
eq
u
ir
em
en
ts
s
im
u
ltan
eo
u
s
ly
:
r
ed
u
cin
g
ir
r
el
ev
an
t
p
r
e
d
icto
r
s
,
im
p
r
o
v
in
g
p
r
ed
ictiv
e
r
o
b
u
s
tn
ess
th
r
o
u
g
h
m
o
d
el
c
o
m
b
in
atio
n
,
an
d
p
r
o
v
id
in
g
in
ter
p
r
etab
le
ev
id
en
ce
r
eg
a
r
d
in
g
th
e
f
ac
t
o
r
s
ass
o
ciate
d
with
s
tu
d
en
t e
n
r
o
llm
e
n
t d
ec
is
io
n
s
[
1
1
]
,
[
2
3
]
,
[
2
8
]
.
Un
lik
e
a
co
n
v
en
tio
n
al
p
r
ed
ictio
n
p
ip
elin
e
in
wh
ich
a
s
in
g
le
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
is
ev
alu
ated
p
r
im
ar
ily
ac
co
r
d
in
g
to
ac
cu
r
a
cy
,
th
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
tr
ea
ts
f
ea
tu
r
e
s
elec
tio
n
,
en
s
em
b
le
p
r
ed
ictio
n
,
an
d
ex
p
lain
ab
ilit
y
as
in
ter
co
n
n
ec
t
ed
co
m
p
o
n
en
ts
.
T
h
is
d
esig
n
is
p
ar
ticu
lar
ly
r
elev
an
t
to
u
n
iv
er
s
ity
ad
m
is
s
io
n
d
ec
is
io
n
-
m
ak
in
g
b
ec
au
s
e
an
a
cc
u
r
ate
p
r
ed
ictio
n
with
o
u
t
an
u
n
d
er
s
tan
d
a
b
le
ex
p
lan
atio
n
m
ay
p
r
o
v
i
d
e
lim
ited
p
r
ac
tical
v
alu
e
to
a
d
m
is
s
io
n
m
an
ag
er
s
.
2
.
2
.
Resea
rc
h o
bje
ct
a
nd
da
t
a
s
et
T
h
e
r
esear
ch
was
co
n
d
u
cted
u
s
in
g
ad
m
is
s
io
n
d
atasets
o
b
tain
ed
f
r
o
m
Un
i
v
er
s
itas
W
ir
ah
u
s
a
d
a
Me
d
an
an
d
Un
iv
er
s
itas
Au
d
i
I
n
d
o
n
e
s
ia.
T
h
e
d
ataset
co
n
s
is
ted
o
f
2
,
0
0
0
u
n
d
er
g
r
ad
u
ate
s
tu
d
en
t
r
ec
o
r
d
s
f
r
o
m
th
e
2025
–
2
0
2
6
ac
a
d
em
ic
y
ea
r
,
w
ith
1
,
0
0
0
r
ec
o
r
d
s
co
n
tr
ib
u
ted
b
y
ea
ch
u
n
i
v
er
s
ity
.
T
h
e
u
s
e
o
f
d
ata
f
r
o
m
two
p
r
iv
ate
u
n
iv
er
s
ities
was
in
t
en
d
ed
to
p
r
o
v
id
e
v
ar
iatio
n
in
in
s
titu
tio
n
al
co
n
te
x
ts
wh
ile
m
ain
tain
in
g
a
co
m
p
ar
ab
le
a
d
m
is
s
io
n
s
ettin
g
.
E
ac
h
o
b
s
er
v
atio
n
r
e
p
r
esen
ts
an
ad
m
itted
u
n
d
e
r
g
r
ad
u
ate
ap
p
l
ican
t a
n
d
co
n
tain
s
in
f
o
r
m
atio
n
r
elate
d
to
d
em
o
g
r
a
p
h
ic
ch
ar
a
cter
is
tics
,
ad
m
is
s
io
n
b
ac
k
g
r
o
u
n
d
,
ac
a
d
em
ic
c
h
ar
ac
ter
is
tics
,
f
in
an
c
ial
co
n
s
id
er
atio
n
s
,
in
s
titu
tio
n
al
p
er
ce
p
tio
n
s
,
u
n
iv
er
s
ity
in
ter
ac
tio
n
,
a
n
d
f
in
al
e
n
r
o
llm
en
t
s
tatu
s
.
T
h
e
tar
g
et
v
ar
iab
le
was
Stu
d
en
t
E
n
r
o
llm
en
t D
ec
is
io
n
,
r
ep
r
esen
ted
as a
b
in
ar
y
class
if
icatio
n
p
r
o
b
lem
:
−
E
n
r
o
lled
(
1
)
: th
e
a
d
m
itted
ap
p
lic
an
t c
o
m
p
leted
th
e
r
eq
u
ir
e
d
en
r
o
llm
en
t o
r
r
eg
is
tr
atio
n
p
r
o
c
ess
.
−
No
t E
n
r
o
lled
(
0
)
: th
e
ad
m
itted
ap
p
lican
t d
id
n
o
t c
o
m
p
lete
th
e
en
r
o
llm
en
t
o
r
r
e
g
is
tr
atio
n
p
r
o
ce
s
s
.
T
h
e
d
is
tin
ctio
n
b
etwe
en
ad
m
is
s
io
n
an
d
en
r
o
llm
en
t
is
im
p
o
r
tan
t
b
ec
a
u
s
e
b
ein
g
ad
m
itte
d
d
o
es
n
o
t
n
ec
ess
ar
ily
m
ea
n
th
at
an
ap
p
lican
t
will
u
ltima
tely
r
eg
is
ter
.
Pre
v
io
u
s
s
tu
d
ies
h
av
e
d
em
o
n
s
tr
ated
th
e
u
s
ef
u
ln
ess
o
f
m
ac
h
in
e
lear
n
in
g
f
o
r
p
r
e
d
ic
tin
g
en
r
o
llm
en
t
am
o
n
g
ad
m
itt
ed
ap
p
lican
ts
,
b
u
t
th
e
ex
p
lain
a
b
ilit
y
o
f
in
d
iv
id
u
al
en
r
o
llm
en
t d
ec
is
io
n
s
r
e
m
ain
s
less
d
e
v
elo
p
ed
[6
]
-
[
8
]
.
T
h
e
d
ataset
th
er
ef
o
r
e
d
if
f
er
s
f
r
o
m
s
tu
d
ies
th
at
p
r
ed
ict
a
ca
d
em
ic
p
er
f
o
r
m
an
ce
o
r
d
r
o
p
o
u
t
af
ter
s
tu
d
en
ts
h
av
e
e
n
ter
ed
u
n
iv
er
s
ity
.
T
h
e
u
n
it
o
f
p
r
e
d
ictio
n
i
n
th
is
s
tu
d
y
is
an
a
d
m
itted
ap
p
lican
t'
s
en
r
o
llm
en
t
d
ec
is
io
n
b
ef
o
r
e
o
r
at
th
e
f
in
al
r
eg
is
tr
atio
n
s
tag
e,
m
a
k
in
g
t
h
e
f
r
am
ewo
r
k
m
o
r
e
d
ir
ec
tly
r
ele
v
an
t
to
r
ec
r
u
itm
en
t
an
d
ad
m
is
s
io
n
m
a
n
ag
em
en
t.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
5
2
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
4
3
,
No
.
3
,
Sep
tem
b
er
20
2
6
:
8
8
0
-
89
7
884
2
.
3
.
Respo
nd
ent
cha
ra
ct
er
is
t
i
cs
T
ab
le
1
p
r
esen
ts
th
e
d
em
o
g
r
ap
h
ic
c
h
ar
ac
ter
is
tics
o
f
th
e
d
ataset.
T
h
e
d
ataset
was
b
alan
ce
d
b
y
in
s
titu
tio
n
,
with
ea
ch
u
n
iv
e
r
s
ity
co
n
tr
ib
u
tin
g
a
n
eq
u
al
n
u
m
b
er
o
f
o
b
s
er
v
atio
n
s
.
T
h
is
s
tr
u
ctu
r
e
allo
ws
th
e
p
r
ed
ictiv
e
f
r
am
ew
o
r
k
to
b
e
d
ev
elo
p
e
d
with
o
u
t
o
n
e
in
s
titu
tio
n
d
o
m
i
n
atin
g
th
e
lear
n
in
g
p
r
o
ce
s
s
.
Fem
ale
ap
p
lican
ts
r
ep
r
esen
ted
5
5
%
o
f
th
e
o
b
s
er
v
atio
n
s
,
wh
ile
th
e
lar
g
est
ag
e
g
r
o
u
p
w
as
1
8
–
1
9
y
ea
r
s
,
f
o
llo
wed
b
y
20
–
2
1
y
ea
r
s
an
d
2
2
y
ea
r
s
.
T
h
e
in
s
titu
tio
n
al
b
alan
ce
is
a
ls
o
r
elev
an
t
to
th
e
in
te
r
p
r
etat
io
n
o
f
t
h
e
r
esu
lts
b
ec
au
s
e
th
e
p
r
o
p
o
s
ed
f
r
am
ewo
r
k
is
in
ten
d
ed
to
c
ap
tu
r
e
en
r
o
llm
e
n
t
p
atter
n
s
th
at
ar
e
n
o
t
r
estricte
d
to
a
s
in
g
le
in
s
titu
tio
n
al
en
v
ir
o
n
m
en
t.
Pre
v
io
u
s
r
esear
ch
h
as
s
h
o
wn
th
at
in
s
tit
u
tio
n
al
ch
ar
ac
ter
is
tics
,
r
ep
u
tatio
n
,
lo
ca
tio
n
,
f
in
an
cial
co
n
s
id
er
atio
n
s
,
an
d
s
tu
d
en
ts
'
d
ec
is
io
n
-
m
ak
in
g
p
r
o
ce
s
s
es
ca
n
in
f
lu
en
ce
u
n
iv
e
r
s
ity
ch
o
ice
an
d
en
r
o
llm
en
t
b
eh
av
io
r
[
2
]
-
[
4
]
,
[
2
2
]
.
T
ab
le
1
.
Dem
o
g
r
ap
h
ic
ch
ar
ac
t
er
is
tics
o
f
s
tu
d
en
ts
(
n
=
2
,
0
0
0
)
V
a
r
i
a
b
l
e
C
a
t
e
g
o
r
y
F
r
e
q
u
e
n
c
y
P
e
r
c
e
n
t
a
g
e
U
n
i
v
e
r
si
t
y
U
n
i
v
e
r
si
t
a
s
W
i
r
a
h
u
sa
d
a
M
e
d
a
n
1
,
0
0
0
5
0
.
0
%
U
n
i
v
e
r
si
t
a
s
A
u
d
i
I
n
d
o
n
e
si
a
1
,
0
0
0
5
0
.
0
%
G
e
n
d
e
r
M
a
l
e
9
0
0
4
5
.
0
%
F
e
mal
e
1
,
1
0
0
5
5
.
0
%
A
g
e
18
–
1
9
y
e
a
r
s
1
,
1
0
0
5
5
.
0
%
20
–
2
1
y
e
a
r
s
7
0
0
3
5
.
0
%
2
2
y
e
a
r
s
3
0
0
1
5
.
0
%
2
.
4
.
Da
t
a
c
o
llect
io
n a
nd
et
hica
l c
o
ns
idera
t
io
ns
T
h
e
s
tu
d
y
u
s
ed
s
ec
o
n
d
ar
y
ad
m
is
s
io
n
d
ata
o
b
tain
ed
f
r
o
m
th
e
o
f
f
icial
ad
m
is
s
io
n
d
atab
ases
o
f
Un
iv
er
s
itas
W
ir
ah
u
s
ad
a
Me
d
an
an
d
Un
iv
er
s
itas
Au
d
i
I
n
d
o
n
esia
f
o
llo
win
g
in
s
titu
tio
n
al
ap
p
r
o
v
al.
T
h
e
v
ar
iab
les
in
clu
d
ed
g
en
d
e
r
,
ag
e,
p
r
e
v
io
u
s
ed
u
ca
tio
n
al
b
ac
k
g
r
o
u
n
d
,
ad
m
is
s
io
n
p
ath
way
,
s
ch
o
lar
s
h
ip
av
ailab
ilit
y
,
tu
itio
n
af
f
o
r
d
a
b
ilit
y
,
u
n
iv
er
s
ity
r
e
p
u
tatio
n
,
d
ig
ital
m
ar
k
etin
g
en
g
ag
em
e
n
t,
a
ca
d
em
ic
f
ac
ilit
ies,
s
er
v
ice
q
u
a
lity
,
co
m
m
u
n
icatio
n
in
ten
s
ity
,
an
d
f
in
al
en
r
o
llm
e
n
t statu
s
.
B
ef
o
r
e
an
aly
s
is
,
d
ir
ec
t
p
er
s
o
n
al
id
en
tifie
r
s
wer
e
r
em
o
v
e
d
f
r
o
m
th
e
d
ataset.
T
h
e
a
n
aly
s
is
wa
s
co
n
d
u
cte
d
u
s
in
g
d
e
-
id
en
tifie
d
r
ec
o
r
d
s
,
an
d
th
e
d
ata
wer
e
u
s
ed
o
n
ly
f
o
r
r
esear
ch
an
d
p
r
ed
ictiv
e
m
o
d
elin
g
p
u
r
p
o
s
es.
T
h
is
p
r
o
ce
d
u
r
e
was
in
ten
d
ed
to
m
in
im
ize
p
r
iv
ac
y
r
is
k
s
ass
o
ciate
d
with
th
e
u
s
e
o
f
ed
u
ca
tio
n
al
d
ata.
T
h
e
s
elec
tio
n
o
f
p
r
ed
icto
r
s
was
g
u
id
ed
b
y
b
o
t
h
d
ata
av
ailab
ilit
y
an
d
th
eir
r
elev
an
ce
to
s
tu
d
en
t
en
r
o
llm
en
t
d
ec
is
io
n
s
.
Pre
v
io
u
s
r
esear
ch
i
n
d
icate
s
th
at
u
n
iv
er
s
ity
r
ep
u
tatio
n
,
l
o
ca
tio
n
,
f
in
an
cial
co
n
s
id
er
atio
n
s
,
in
s
titu
tio
n
al
s
u
p
p
o
r
t,
an
d
o
th
e
r
ch
ar
ac
ter
is
tics
ca
n
in
f
lu
en
ce
h
ig
h
er
ed
u
ca
tio
n
ch
o
ice
[2
]
-
[
4
]
,
[
22
]
,
[
2
9
]
.
Acc
o
r
d
in
g
ly
,
th
e
p
r
o
p
o
s
ed
f
r
a
m
ewo
r
k
in
co
r
p
o
r
ate
s
m
u
ltid
im
en
s
io
n
al
p
r
ed
icto
r
s
r
ath
er
th
an
r
ely
in
g
ex
clu
s
iv
ely
o
n
ac
a
d
em
ic
ch
a
r
a
cter
is
t
ics.
2
.
5
.
Da
t
a
prepro
ce
s
s
ing
Data
p
r
ep
r
o
ce
s
s
in
g
was
c
o
n
d
u
cted
b
ef
o
r
e
f
ea
tu
r
e
s
elec
tio
n
a
n
d
m
o
d
el
tr
ain
in
g
t
o
im
p
r
o
v
e
d
ata
q
u
ality
an
d
r
e
d
u
ce
p
o
ten
tial
s
o
u
r
ce
s
o
f
m
o
d
elin
g
er
r
o
r
.
T
h
e
p
r
ep
r
o
ce
s
s
in
g
p
r
o
ce
d
u
r
e
co
n
s
is
ted
o
f
th
e
f
o
llo
win
g
s
tep
s
,
−
Data
clea
n
in
g
,
in
clu
d
in
g
c
h
ec
k
in
g
d
ata
t
y
p
es,
in
v
alid
v
alu
es
,
an
d
in
c
o
n
s
is
ten
t r
ec
o
r
d
s
.
−
Du
p
licate
r
em
o
v
al
to
en
s
u
r
e
t
h
at
th
e
s
am
e
ap
p
lican
t
r
ec
o
r
d
was n
o
t r
ep
r
esen
ted
m
o
r
e
th
a
n
o
n
ce
.
−
Miss
in
g
-
v
alu
e
tr
ea
tm
en
t
u
s
in
g
an
ap
p
r
o
p
r
iate
im
p
u
tatio
n
p
r
o
ce
d
u
r
e
b
ased
o
n
th
e
c
h
ar
ac
t
er
is
tics
o
f
th
e
v
ar
iab
le.
−
C
ateg
o
r
ical
en
co
d
in
g
u
s
in
g
o
n
e
-
h
o
t e
n
c
o
d
in
g
f
o
r
ca
teg
o
r
ical
p
r
ed
icto
r
s
.
−
N
u
m
e
r
i
c
a
l
s
c
a
l
i
n
g
u
s
i
n
g
M
i
n
-
M
a
x
n
o
r
m
a
l
i
z
a
t
i
o
n
w
h
e
r
e
s
c
a
l
i
n
g
w
a
s
r
e
q
u
i
r
e
d
b
y
t
h
e
m
o
d
e
l
i
n
g
p
r
o
c
e
d
u
r
e
.
−
Ou
tlier
id
en
tific
atio
n
u
s
in
g
th
e
in
ter
q
u
ar
tile r
a
n
g
e
(
I
QR
)
cr
iter
io
n
.
T
h
ese
p
r
ep
r
o
ce
s
s
in
g
p
r
o
ce
d
u
r
es
wer
e
ap
p
lied
b
e
f
o
r
e
m
o
d
el
d
ev
elo
p
m
e
n
t
to
p
r
o
d
u
ce
a
co
n
s
is
ten
t
an
aly
tical
d
ataset.
Pre
d
ictiv
e
lear
n
in
g
an
al
y
tics
r
esear
ch
em
p
h
asizes
th
e
im
p
o
r
tan
ce
o
f
s
y
s
te
m
atic
p
r
ep
r
o
ce
s
s
in
g
b
ec
au
s
e
in
co
n
s
is
ten
t o
r
r
ed
u
n
d
a
n
t o
b
s
er
v
atio
n
s
ca
n
in
f
lu
en
ce
th
e
s
tab
ilit
y
an
d
g
en
er
aliza
tio
n
o
f
p
r
ed
ictiv
e
m
o
d
els
[
1
]
.
T
o
p
r
ev
en
t
in
f
o
r
m
atio
n
leak
ag
e,
p
r
ep
r
o
ce
s
s
in
g
p
ar
a
m
eter
s
th
at
r
eq
u
ir
e
esti
m
atio
n
f
r
o
m
th
e
d
ata
wer
e
f
itted
u
s
in
g
th
e
tr
ain
in
g
d
ata
an
d
s
u
b
s
eq
u
en
tl
y
ap
p
lie
d
to
v
alid
atio
n
an
d
test
o
b
s
er
v
atio
n
s
.
T
h
e
s
am
e
p
r
in
cip
le
was
ap
p
lied
to
f
ea
t
u
r
e
s
elec
tio
n
.
I
n
o
th
er
wo
r
d
s
,
i
n
f
o
r
m
atio
n
f
r
o
m
th
e
f
in
al
test
s
et
was
n
o
t
u
s
ed
wh
en
s
elec
tin
g
p
r
ed
icto
r
s
o
r
d
eter
m
in
in
g
en
s
em
b
le
weig
h
t
s
.
Af
ter
p
r
ep
r
o
ce
s
s
in
g
,
th
e
d
ataset
was
d
iv
id
ed
u
s
in
g
an
8
0
:
2
0
s
tr
atif
ied
h
o
ld
-
o
u
t
p
r
o
ce
d
u
r
e.
T
h
e
tr
ain
i
n
g
s
et
co
n
tain
ed
1
,
6
0
0
o
b
s
e
r
v
atio
n
s
,
wh
ile
th
e
in
d
ep
en
d
en
t
test
s
et
co
n
tain
ed
4
0
0
o
b
s
er
v
atio
n
s
.
Stra
tific
atio
n
was
u
s
ed
to
p
r
eser
v
e
th
e
p
r
o
p
o
r
tio
n
o
f
th
e
two
en
r
o
llm
en
t c
lass
es in
b
o
th
s
u
b
s
ets.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2502
-
4
7
5
2
E
xp
la
in
a
b
le
h
yb
r
id
ma
c
h
in
e
l
ea
r
n
in
g
fo
r
p
r
ed
ictin
g
s
tu
d
en
t
en
r
o
llmen
t d
ec
is
io
n
s
…
(
Dyb
io
Do
mp
u
Ho
t
A
s
ih
)
885
T
h
e
1
,
6
0
0
tr
ain
in
g
o
b
s
er
v
atio
n
s
wer
e
s
u
b
s
eq
u
en
tly
u
s
ed
f
o
r
m
o
d
el
d
ev
elo
p
m
en
t
an
d
s
tr
atif
ied
1
0
-
f
o
ld
cr
o
s
s
-
v
alid
atio
n
.
T
h
e
4
0
0
o
b
s
er
v
atio
n
s
in
th
e
h
o
ld
-
o
u
t
test
s
et
r
em
ain
ed
u
n
to
u
c
h
ed
u
n
til
th
e
f
in
al
ev
alu
atio
n
.
T
h
is
s
ep
ar
atio
n
p
r
o
v
id
es
a
m
o
r
e
r
eliab
le
esti
m
ate
o
f
o
u
t
-
of
-
s
am
p
le
p
er
f
o
r
m
a
n
ce
b
ec
au
s
e
t
h
e
f
i
n
al
test
d
ata
d
o
n
o
t
p
ar
ticip
ate
in
f
ea
tu
r
e
s
elec
tio
n
,
h
y
p
er
p
ar
am
eter
o
p
tim
izatio
n
,
en
s
em
b
le
weig
h
tin
g
,
o
r
m
o
d
el
f
itti
n
g
.
2
.
6
.
H
y
brid f
ea
t
ure
s
elec
t
io
n
Featu
r
e
s
elec
tio
n
was
p
er
f
o
r
m
ed
to
id
en
tify
p
r
e
d
icto
r
s
th
at
p
r
o
v
id
e
m
ea
n
in
g
f
u
l
in
f
o
r
m
atio
n
ab
o
u
t
s
tu
d
en
t
en
r
o
llm
en
t
d
ec
is
io
n
s
wh
ile
r
ed
u
cin
g
ir
r
elev
an
t
a
n
d
r
ed
u
n
d
an
t
v
ar
ia
b
les.
T
h
e
s
tu
d
y
ad
o
p
ted
a
h
y
b
r
id
f
ea
tu
r
e
s
elec
tio
n
s
tr
ateg
y
c
o
m
b
in
in
g
th
e
B
o
r
u
ta
alg
o
r
ith
m
a
n
d
m
u
t
u
al
in
f
o
r
m
atio
n
(
M
I
)
.
B
o
r
u
ta
is
a
wr
ap
p
er
-
b
ased
f
ea
tu
r
e
s
elec
tio
n
m
et
h
o
d
th
at
ev
alu
ates
th
e
im
p
o
r
tan
ce
o
f
o
r
ig
i
n
al
p
r
ed
icto
r
s
a
g
ain
s
t
r
an
d
o
m
ly
g
en
e
r
ated
s
h
a
d
o
w
f
ea
tu
r
es.
A
p
r
ed
icto
r
is
co
n
s
id
er
e
d
r
elev
an
t
wh
en
its
im
p
o
r
tan
ce
is
co
n
s
is
ten
tly
h
ig
h
er
th
a
n
th
e
im
p
o
r
tan
c
e
o
f
th
e
co
r
r
esp
o
n
d
i
n
g
r
a
n
d
o
m
ized
s
h
a
d
o
w
v
ar
iab
les
[
2
3
]
.
MI
was
s
u
b
s
eq
u
en
tly
u
s
ed
to
ass
es
s
th
e
d
eg
r
ee
o
f
s
tatis
tical
d
ep
en
d
en
cy
b
etwe
en
ea
ch
p
r
ed
icto
r
an
d
th
e
en
r
o
llm
e
n
t
tar
g
et.
U
n
lik
e
m
eth
o
d
s
th
at
f
o
cu
s
o
n
ly
o
n
lin
ea
r
ass
o
ciatio
n
,
MI
ca
n
ca
p
tu
r
e
n
o
n
lin
ea
r
d
ep
en
d
e
n
cy
b
etwe
en
v
a
r
iab
le
s
an
d
th
e
ta
r
g
et
class
.
T
h
is
c
h
ar
ac
ter
is
tic
is
u
s
ef
u
l
f
o
r
ed
u
ca
tio
n
al
d
atasets
in
wh
ich
r
elatio
n
s
h
ip
s
b
etwe
en
a
p
p
lican
t c
h
ar
ac
ter
is
tics
an
d
en
r
o
llm
en
t d
ec
is
io
n
s
m
ay
n
o
t b
e
s
tr
ictly
lin
ea
r
.
T
h
e
h
y
b
r
id
s
elec
tio
n
p
r
o
ce
d
u
r
e
was im
p
lem
en
ted
as f
o
llo
ws
:
−
All c
an
d
id
ate
p
r
ed
ict
o
r
s
wer
e
s
u
p
p
lied
to
th
e
B
o
r
u
ta
p
r
o
ce
d
u
r
e
u
s
in
g
o
n
ly
th
e
tr
ain
in
g
d
at
a.
−
Var
iab
les id
en
tifie
d
as r
elev
an
t b
y
B
o
r
u
ta
wer
e
r
etain
ed
as c
an
d
id
ate
p
r
e
d
icto
r
s
.
−
Mu
tu
al
I
n
f
o
r
m
atio
n
s
co
r
es we
r
e
ca
lcu
lated
f
o
r
th
e
ca
n
d
id
ate
p
r
ed
icto
r
s
ag
ai
n
s
t th
e
en
r
o
llm
en
t ta
r
g
et.
−
Pre
d
icto
r
s
m
ee
tin
g
th
e
p
r
ed
ef
i
n
ed
MI
s
elec
tio
n
cr
iter
io
n
wer
e
r
etain
ed
.
−
T
h
e
in
ter
s
ec
tio
n
o
f
th
e
B
o
r
u
ta
-
s
elec
ted
an
d
MI
-
s
elec
ted
p
r
e
d
icto
r
s
co
n
s
titu
ted
th
e
f
i
n
al
f
e
atu
r
e
s
et.
−
T
h
e
f
in
al
f
ea
tu
r
e
s
et
was
s
u
b
s
eq
u
en
tly
u
s
ed
b
y
all
class
if
icatio
n
m
o
d
els.
T
h
e
in
ter
s
ec
tio
n
s
tr
ateg
y
was
s
elec
ted
to
p
r
io
r
itize
p
r
e
d
icto
r
s
th
at
d
em
o
n
s
tr
ated
r
ele
v
an
ce
f
r
o
m
two
d
if
f
er
en
t
p
e
r
s
p
ec
tiv
es.
B
o
r
u
ta
ev
alu
ates
f
ea
tu
r
e
im
p
o
r
tan
ce
r
elativ
e
to
r
an
d
o
m
ized
s
h
ad
o
w
v
ar
iab
les,
wh
er
ea
s
MI
ev
alu
ates
d
ep
en
d
en
c
y
with
th
e
tar
g
et.
C
o
m
b
in
in
g
t
h
ese
ap
p
r
o
ac
h
es
p
r
o
v
id
es
a
m
o
r
e
co
n
s
er
v
ativ
e
s
elec
tio
n
m
ec
h
an
is
m
an
d
ca
n
r
ed
u
ce
th
e
r
is
k
o
f
r
etain
i
n
g
p
r
e
d
icto
r
s
th
at
a
p
p
ea
r
u
s
ef
u
l
u
n
d
er
o
n
ly
o
n
e
s
elec
tio
n
cr
iter
io
n
.
I
m
p
o
r
ta
n
tly
,
f
ea
t
u
r
e
s
elec
tio
n
was
p
er
f
o
r
m
ed
with
in
th
e
tr
ain
in
g
p
r
o
ce
s
s
r
ath
e
r
th
a
n
u
s
in
g
th
e
co
m
p
lete
d
ataset.
T
h
is
p
r
o
ce
d
u
r
e
p
r
ev
en
ts
th
e
test
o
b
s
er
v
ati
o
n
s
f
r
o
m
in
f
l
u
en
cin
g
th
e
s
ele
ctio
n
o
f
p
r
e
d
icto
r
s
an
d
th
er
e
f
o
r
e
r
ed
u
ce
s
th
e
p
o
s
s
ib
ilit
y
o
f
o
p
tim
is
tic
p
er
f
o
r
m
a
n
ce
esti
m
ates.
2
.
7
.
E
x
pla
ina
ble
hy
brid m
a
c
hin
e
lea
rning
f
ra
m
ewo
rk
T
h
e
p
r
o
p
o
s
ed
E
HM
L
F c
o
n
s
is
t
s
o
f
f
o
u
r
p
r
i
n
cip
al
co
m
p
u
tatio
n
al
co
m
p
o
n
en
ts
:
−
Data
p
r
ep
r
o
ce
s
s
in
g
a
n
d
f
ea
tu
r
e
en
g
in
ee
r
in
g
;
−
Hy
b
r
id
f
ea
t
u
r
e
s
elec
tio
n
;
−
Ad
ap
tiv
e
en
s
em
b
le
class
if
icatio
n
; a
n
d
−
E
x
p
lain
ab
le
Ar
tific
ial
I
n
tellig
e
n
ce
u
s
in
g
SHAP.
T
h
e
class
if
icatio
n
co
m
p
o
n
e
n
t
co
m
b
in
es
f
o
u
r
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
s
:
R
an
d
o
m
Fo
r
est,
XGBo
o
s
t,
C
atB
o
o
s
t,
an
d
L
ig
h
tGB
M.
T
h
e
u
s
e
o
f
s
ev
er
al
tr
ee
-
b
ased
alg
o
r
ith
m
s
was
m
o
tiv
a
ted
b
y
th
eir
ab
ilit
y
to
m
o
d
el
n
o
n
lin
ea
r
r
elatio
n
s
h
i
p
s
an
d
in
ter
ac
tio
n
s
am
o
n
g
h
et
er
o
g
en
e
o
u
s
p
r
e
d
icto
r
s
.
XGBo
o
s
t
i
s
a
s
ca
lab
le
g
r
ad
ien
t
b
o
o
s
tin
g
f
r
am
ew
o
r
k
d
esig
n
ed
to
co
n
s
tr
u
ct
an
en
s
em
b
le
o
f
d
ec
is
io
n
tr
ee
s
s
eq
u
en
tially
[
2
5
]
.
L
ig
h
t
GB
M
u
s
e
s
a
h
ig
h
ly
ef
f
icien
t
g
r
ad
ien
t
b
o
o
s
tin
g
d
ec
is
io
n
-
t
r
ee
s
tr
ateg
y
an
d
is
d
esig
n
ed
f
o
r
co
m
p
u
tatio
n
ally
ef
f
icien
t
lear
n
in
g
[
2
6
]
.
C
atB
o
o
s
t
is
a
b
o
o
s
tin
g
al
g
o
r
ith
m
d
esig
n
ed
to
p
r
o
v
id
e
r
o
b
u
s
t
lear
n
in
g
,
p
ar
ticu
lar
ly
i
n
d
atasets
co
n
tain
in
g
ca
teg
o
r
i
ca
l
in
f
o
r
m
atio
n
[
2
7
]
.
R
an
d
o
m
f
o
r
est
p
r
o
v
id
es
an
en
s
em
b
le
o
f
r
a
n
d
o
m
ized
d
ec
is
io
n
tr
ee
s
an
d
s
er
v
es a
s
a
co
m
p
lem
en
tar
y
b
a
g
g
in
g
-
b
ased
class
if
ier
.
R
ath
er
th
an
s
elec
tin
g
o
n
e
alg
o
r
ith
m
s
o
lely
ac
co
r
d
i
n
g
to
it
s
in
d
iv
id
u
al
p
er
f
o
r
m
an
ce
,
t
h
e
p
r
o
p
o
s
ed
f
r
am
ewo
r
k
co
m
b
in
es
th
eir
p
r
ed
ictio
n
s
.
T
h
is
d
esig
n
f
o
llo
ws
th
e
p
r
in
cip
le
th
at
m
o
d
els
with
d
if
f
e
r
en
t
lear
n
i
n
g
ch
ar
ac
ter
is
tics
ca
n
ca
p
tu
r
e
c
o
m
p
lem
en
tar
y
p
atter
n
s
in
th
e
s
am
e
d
ataset.
Pre
v
io
u
s
h
ig
h
e
r
ed
u
ca
tio
n
s
tu
d
ies
h
av
e
d
em
o
n
s
tr
ated
th
e
p
o
ten
ti
al
o
f
m
u
lti
-
m
o
d
el
an
d
en
s
em
b
l
e
ap
p
r
o
ac
h
es f
o
r
ad
m
is
s
io
n
p
r
ed
ictio
n
[9
]
,
[
1
0
]
.
2
.
8
.
Ada
ptiv
e
weig
ht
ens
emble
T
h
e
f
in
a
l
p
r
ed
ictio
n
was
g
e
n
er
ated
th
r
o
u
g
h
an
ad
ap
ti
v
e
w
eig
h
ted
v
o
tin
g
m
ec
h
an
is
m
.
L
et
th
e
f
o
u
r
b
ase
class
if
ier
s
b
e
r
ep
r
esen
ted
b
y
,
[
M=
{RF
,
XGB,
C
AT
,
L
GB
M
}]
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
5
2
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
4
3
,
No
.
3
,
Sep
tem
b
er
20
2
6
:
8
8
0
-
89
7
886
Fo
r
an
o
b
s
er
v
atio
n
(
x
_
i)
,
ea
ch
class
if
ier
p
r
o
d
u
ce
s
an
e
n
r
o
llm
en
t p
r
o
b
ab
ilit
y
,
[
p
_
m
(
x
_
i)
=P(
Y=
1
|
x
_
i,m
)
]
wh
er
e
(
m
)
r
ep
r
esen
ts
o
n
e
o
f
th
e
f
o
u
r
b
ase
class
if
ier
s
.
T
h
e
en
s
em
b
le
p
r
o
b
ab
ilit
y
is
ca
lcu
lated
as
,
[
P_
{e
n
s
}(
x
_
i)
=_
{
m
=1
}^{
4
}w
_
m
p
_
m
(
x
_
i
)
]
s
u
b
ject
to
,
[
\
s
u
m
_
{m
=1
}^{
4
}w_
m
=1
,
\
q
q
u
ad
w_
m
\
g
e
q
0
]
wh
er
e
(
w_
m
)
r
ep
r
esen
ts
th
e
c
o
n
tr
ib
u
tio
n
ass
ig
n
ed
to
class
if
ier
(
m
)
.
T
h
e
weig
h
ts
wer
e
d
eter
m
in
ed
u
s
in
g
v
alid
atio
n
p
er
f
o
r
m
an
ce
o
b
tain
e
d
d
u
r
in
g
th
e
tr
ain
in
g
p
r
o
ce
s
s
.
A
class
if
ier
d
em
o
n
s
tr
atin
g
s
tr
o
n
g
er
v
alid
atio
n
p
er
f
o
r
m
a
n
ce
r
e
ce
iv
ed
a
p
r
o
p
o
r
tio
n
ally
g
r
ea
ter
co
n
tr
ib
u
tio
n
to
th
e
f
in
al
e
n
s
em
b
le
p
r
ed
ictio
n
.
T
h
e
r
esu
ltin
g
en
s
em
b
le
th
er
ef
o
r
e
d
o
es
n
o
t
tr
ea
t
all
cla
s
s
if
ier
s
as
eq
u
ally
in
f
o
r
m
ativ
e.
T
h
e
f
i
n
al
b
in
ar
y
p
r
ed
ictio
n
was o
b
tain
ed
u
s
in
g
a
class
if
icatio
n
th
r
esh
o
ld
,
[
\
h
at{Y
}i=
\
b
eg
in
{c
ases
}1
,
&
P{en
s
}(
x
_
i)
\
g
eq
\
tau
\
0
,
&
P_
{
en
s
}(
x
_
i)
<
\
tau
\
en
d
{c
ases
}]
wh
er
e
(
\
tau
)
r
ep
r
esen
ts
th
e
s
elec
ted
d
ec
is
io
n
th
r
esh
o
l
d
.
T
h
is
ad
ap
tiv
e
s
tr
ateg
y
p
r
o
v
i
d
es
a
m
o
r
e
f
lex
ib
le
alter
n
ativ
e
to
s
im
p
le
m
ajo
r
ity
v
o
tin
g
b
e
ca
u
s
e
th
e
co
n
tr
ib
u
tio
n
o
f
ea
c
h
class
if
ier
r
ef
lects
its
em
p
ir
ical
v
alid
atio
n
p
er
f
o
r
m
an
ce
.
T
h
e
ap
p
r
o
ac
h
is
p
ar
ticu
lar
ly
r
elev
an
t
wh
en
th
e
f
o
u
r
alg
o
r
it
h
m
s
p
r
o
d
u
ce
d
if
f
er
en
t
lev
els
o
f
p
r
ed
ictiv
e
p
er
f
o
r
m
a
n
ce
f
o
r
t
h
e
s
am
e
en
r
o
llm
en
t
d
ataset.
2
.
9
.
H
y
perpa
ra
m
e
t
er
o
ptim
i
za
t
io
n
Hy
p
er
p
ar
a
m
eter
o
p
tim
izatio
n
was
p
er
f
o
r
m
e
d
d
u
r
in
g
th
e
m
o
d
el
d
ev
elo
p
m
en
t
s
tag
e
t
o
o
b
ta
in
s
u
itab
le
p
ar
am
eter
c
o
n
f
ig
u
r
atio
n
s
f
o
r
th
e
in
d
iv
id
u
al
class
if
ier
s
.
T
h
e
o
p
tim
izatio
n
p
r
o
ce
s
s
was
co
n
d
u
cted
o
n
ly
u
s
in
g
th
e
tr
ain
in
g
d
ata
s
o
th
at
th
e
in
d
ep
en
d
e
n
t te
s
t set r
em
ain
ed
co
m
p
letely
u
n
s
ee
n
.
T
h
e
o
p
tim
izatio
n
p
r
o
ce
d
u
r
e
co
n
s
id
er
ed
m
o
d
el
-
s
p
ec
if
ic
h
y
p
er
p
a
r
am
eter
s
f
o
r
R
an
d
o
m
Fo
r
est,
XGBo
o
s
t,
C
a
tB
o
o
s
t,
an
d
L
i
g
h
tGB
M.
Op
tu
n
a
was
u
s
ed
as
th
e
h
y
p
er
p
ar
am
ete
r
o
p
tim
izatio
n
f
r
am
ewo
r
k
b
ec
au
s
e
it
p
r
o
v
id
es
an
au
to
m
ated
m
ec
h
a
n
is
m
f
o
r
s
ea
r
ch
in
g
m
o
d
el
co
n
f
ig
u
r
ati
o
n
s
an
d
id
e
n
tify
in
g
p
r
o
m
is
in
g
p
ar
am
eter
co
m
b
in
atio
n
s
[
3
0
]
.
T
h
e
o
p
tim
izatio
n
o
b
jectiv
e
wa
s
b
ased
o
n
v
alid
atio
n
p
er
f
o
r
m
an
ce
.
T
h
e
b
est
-
p
er
f
o
r
m
in
g
co
n
f
ig
u
r
atio
n
f
o
r
ea
ch
class
if
ier
was
r
etain
ed
f
o
r
s
u
b
s
eq
u
en
t
en
s
em
b
le
co
n
s
tr
u
ctio
n
.
T
h
is
p
r
o
ce
d
u
r
e
r
ed
u
ce
s
d
ep
en
d
e
n
ce
o
n
m
an
u
ally
s
elec
ted
h
y
p
er
p
a
r
am
eter
s
an
d
p
r
o
v
id
es a
s
y
s
tem
atic
b
asis
f
o
r
co
m
p
ar
in
g
t
h
e
b
ase
class
if
ier
s
.
2
.
1
0
.
E
x
pla
ina
ble a
rt
if
icia
l int
ellig
ence
T
o
im
p
r
o
v
e
th
e
tr
an
s
p
ar
en
c
y
o
f
th
e
f
in
al
p
r
ed
ictiv
e
m
o
d
e
l,
th
e
s
tu
d
y
em
p
lo
y
ed
SHAP.
SHAP
is
b
ased
o
n
Sh
ap
ley
v
alu
es f
r
o
m
co
o
p
er
ativ
e
g
a
m
e
th
eo
r
y
an
d
q
u
an
tifie
s
th
e
co
n
tr
ib
u
tio
n
o
f
ea
ch
p
r
ed
icto
r
to
a
n
in
d
iv
id
u
al
m
o
d
el
p
r
ed
ictio
n
[
2
8
]
.
Fo
r
a
p
r
ed
ictio
n
(
f
(
x
)
)
,
th
e
SHAP
ex
p
lan
atio
n
ca
n
b
e
e
x
p
r
ess
ed
co
n
ce
p
tu
ally
as
,
[
f
(
x
)
=
\
p
h
i_
0
+
\
s
u
m
_
{j=
1
}^{
p
}
\
p
h
i_
j]
wh
er
e
(
\
p
h
i_
0
)
r
ep
r
esen
ts
th
e
b
aselin
e
p
r
ed
ictio
n
an
d
(
\
p
h
i
_
j)
r
ep
r
esen
ts
th
e
co
n
tr
ib
u
tio
n
o
f
p
r
e
d
icto
r
(
j)
to
th
e
p
r
ed
ict
io
n
.
T
wo
lev
els o
f
i
n
ter
p
r
etatio
n
wer
e
co
n
s
id
er
e
d
.
Glo
b
al
ex
p
lan
atio
n
was
u
s
ed
to
id
en
tify
th
e
p
r
e
d
icto
r
s
th
at
co
n
tr
ib
u
ted
m
o
s
t
s
tr
o
n
g
ly
to
e
n
r
o
llm
en
t
p
r
ed
ictio
n
s
ac
r
o
s
s
th
e
d
ataset.
T
h
e
m
ag
n
itu
d
e
o
f
th
e
m
ea
n
ab
s
o
lu
te
SHAP
v
alu
e
was
u
s
ed
to
r
an
k
p
r
ed
icto
r
im
p
o
r
tan
ce
.
L
ar
g
er
ab
s
o
lu
te
S
HAP
v
alu
es in
d
icate
a
s
tr
o
n
g
e
r
co
n
tr
ib
u
tio
n
to
m
o
d
el
o
u
tp
u
t.
L
o
ca
l
ex
p
lan
atio
n
was
u
s
ed
to
ex
am
in
e
h
o
w
in
d
i
v
id
u
al
p
r
ed
icto
r
s
co
n
tr
ib
u
ted
to
a
p
ar
ticu
lar
ap
p
lican
t'
s
p
r
ed
icted
en
r
o
llm
e
n
t
o
u
tco
m
e.
Po
s
itiv
e
SHAP
v
alu
es
in
d
icate
th
at
a
f
ea
tu
r
e
p
u
s
h
es
th
e
p
r
ed
ictio
n
to
war
d
th
e
e
n
r
o
llm
e
n
t
class
,
wh
er
ea
s
n
eg
ativ
e
SHAP
v
alu
es
in
d
icate
th
at
th
e
f
ea
tu
r
e
p
u
s
h
es
th
e
p
r
ed
ictio
n
to
war
d
th
e
n
o
n
-
en
r
o
llm
en
t
cl
ass
.
T
h
e
u
s
e
o
f
SHAP
is
im
p
o
r
tan
t
b
ec
au
s
e
p
r
e
d
ictiv
e
p
er
f
o
r
m
an
ce
alo
n
e
d
o
es
n
o
t
ex
p
lain
w
h
y
an
a
p
p
lican
t
i
s
class
if
ied
as
lik
ely
o
r
u
n
lik
e
ly
to
en
r
o
ll.
E
x
p
lain
ab
le
AI
r
e
s
ea
r
ch
em
p
h
asizes
th
e
im
p
o
r
ta
n
ce
o
f
tr
an
s
p
ar
e
n
c
y
an
d
in
ter
p
r
eta
b
ilit
y
wh
en
m
ac
h
in
e
lear
n
i
n
g
p
r
ed
ictio
n
s
ar
e
u
s
ed
to
s
u
p
p
o
r
t
co
n
s
eq
u
en
tial
d
ec
is
io
n
s
[
1
1
]
.
I
n
ed
u
ca
tio
n
al
ap
p
li
ca
tio
n
s
,
ex
p
lain
ab
ilit
y
h
as
also
b
ee
n
u
s
ed
to
id
en
tify
in
f
lu
en
tial
f
ac
to
r
s
u
n
d
er
ly
in
g
s
tu
d
en
t
-
r
elate
d
p
r
ed
ictio
n
s
an
d
to
im
p
r
o
v
e
th
e
p
r
ac
tica
l
in
ter
p
r
etatio
n
o
f
p
r
ed
ictiv
e
m
o
d
els
[5
]
,
[
12
]
,
[
13
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2502
-
4
7
5
2
E
xp
la
in
a
b
le
h
yb
r
id
ma
c
h
in
e
l
ea
r
n
in
g
fo
r
p
r
ed
ictin
g
s
tu
d
en
t
en
r
o
llmen
t d
ec
is
io
n
s
…
(
Dyb
io
Do
mp
u
Ho
t
A
s
ih
)
887
2
.
1
1
.
M
o
del e
v
a
lua
t
io
n
Mo
d
el
ev
alu
atio
n
was
co
n
d
u
c
ted
at
two
lev
els:
in
d
iv
id
u
al
class
if
ier
ev
alu
atio
n
an
d
f
in
al
en
s
em
b
le
ev
alu
atio
n
.
T
h
e
f
o
u
r
b
ase
cl
ass
if
ier
s
wer
e
f
ir
s
t
ev
alu
ated
in
d
ep
e
n
d
en
tly
,
f
o
llo
wed
b
y
ev
alu
atio
n
o
f
t
h
e
ad
ap
tiv
e
weig
h
ted
en
s
em
b
le.
Du
r
in
g
m
o
d
el
d
ev
elo
p
m
en
t,
Stra
tifie
d
1
0
-
Fo
ld
C
r
o
s
s
-
Valid
atio
n
was
ap
p
lied
to
th
e
tr
ain
i
n
g
d
ataset.
T
h
e
1
,
6
0
0
tr
ain
i
n
g
o
b
s
er
v
atio
n
s
wer
e
d
i
v
id
ed
in
to
ten
f
o
ld
s
wh
ile
p
r
eser
v
in
g
t
h
e
class
d
is
tr
ib
u
tio
n
.
I
n
ea
c
h
iter
atio
n
,
n
in
e
f
o
ld
s
wer
e
u
s
e
d
f
o
r
tr
ain
in
g
,
an
d
o
n
e
f
o
ld
w
as
u
s
ed
f
o
r
v
alid
atio
n
.
T
h
e
p
r
o
ce
s
s
was
r
ep
ea
ted
u
n
til ev
er
y
f
o
ld
h
ad
s
er
v
e
d
as
th
e
v
alid
atio
n
f
o
ld
.
T
h
e
f
in
al
m
o
d
el
co
n
f
ig
u
r
atio
n
was
th
en
r
etr
ain
ed
u
s
in
g
th
e
co
m
p
lete
tr
ain
in
g
d
ataset
an
d
ev
alu
ated
o
n
ce
o
n
th
e
in
d
e
p
en
d
e
n
t
4
0
0
-
o
b
s
er
v
atio
n
test
s
et.
T
h
is
p
r
o
ce
d
u
r
e
s
ep
a
r
ates
m
o
d
el
d
e
v
elo
p
m
en
t
f
r
o
m
f
i
n
al
p
er
f
o
r
m
an
ce
esti
m
atio
n
an
d
r
e
d
u
ce
s
th
e
r
is
k
o
f
r
ep
o
r
tin
g
o
v
e
r
ly
o
p
tim
is
tic
r
esu
lts
.
Fiv
e
class
if
icatio
n
m
etr
ics we
r
e
u
s
ed
,
−
[
Acc
u
r
ac
y
=
\
f
r
ac
{T
P+T
N}{
T
P
+T
N+
FP
+FN}
]
−
[
Pre
cisi
o
n
=
\
f
r
ac
{T
P}{T
P+FP
}]
−
[
R
ec
all=
\
f
r
ac
{T
P}{T
P+FN}]
−
[
F1
=2
\
tim
es
\
f
r
ac
{Pr
ec
is
io
n
\
tim
es R
ec
all}{
Pre
ci
s
io
n
+Reca
ll
}]
wh
er
e
(
T
P),
(
T
N)
,
(
FP
)
,
an
d
(
FN)
d
en
o
te
tr
u
e
p
o
s
itiv
e,
tr
u
e
n
eg
ativ
e,
f
alse
p
o
s
itiv
e,
an
d
f
alse
n
eg
ativ
e
o
b
s
er
v
atio
n
s
,
r
esp
ec
tiv
ely
.
I
n
ad
d
itio
n
,
th
e
ar
ea
u
n
d
e
r
th
e
r
ec
eiv
er
o
p
er
atin
g
ch
ar
ac
ter
is
tic
cu
r
v
e
(
AUC
-
R
OC
)
was
ca
lcu
lated
to
ev
alu
ate
th
e
ab
ilit
y
o
f
th
e
m
o
d
els to
d
is
tin
g
u
is
h
b
etwe
en
en
r
o
lled
an
d
n
o
n
-
en
r
o
lled
ap
p
lica
n
ts
ac
r
o
s
s
d
if
f
er
en
t
class
if
icatio
n
th
r
esh
o
ld
s
.
Acc
u
r
ac
y
p
r
o
v
id
es
an
o
v
er
all
m
ea
s
u
r
e
o
f
co
r
r
ec
t
class
if
icatio
n
,
wh
ile
Pre
cisi
o
n
,
R
ec
all,
an
d
F1
-
s
co
r
e
p
r
o
v
id
e
co
m
p
lem
en
tar
y
i
n
f
o
r
m
atio
n
r
eg
ar
d
in
g
class
if
icatio
n
er
r
o
r
s
.
AUC
p
r
o
v
id
es
a
th
r
esh
o
ld
-
in
d
ep
en
d
en
t a
s
s
ess
m
en
t o
f
d
is
cr
im
in
ato
r
y
ca
p
ab
i
lity
[
3
1
]
.
T
h
e
ev
alu
atio
n
d
id
n
o
t
r
ely
o
n
ac
cu
r
ac
y
alo
n
e.
T
h
is
is
im
p
o
r
tan
t
b
ec
au
s
e
a
b
in
ar
y
en
r
o
llm
e
n
t
d
ataset
ca
n
p
r
o
d
u
ce
m
is
lead
in
g
ly
h
i
g
h
ac
cu
r
ac
y
wh
en
o
n
e
class
is
m
o
r
e
p
r
ev
alen
t
th
an
th
e
o
th
er
.
T
h
er
ef
o
r
e,
th
e
co
m
b
in
atio
n
o
f
Pre
cisi
o
n
,
R
ec
all,
F1
-
s
co
r
e,
an
d
AUC
was
u
s
ed
to
p
r
o
v
id
e
a
m
o
r
e
co
m
p
r
e
h
en
s
iv
e
ass
ess
m
en
t
o
f
p
r
e
d
ictiv
e
p
er
f
o
r
m
a
n
ce
.
2
.
1
2
.
Co
m
pa
r
a
t
iv
e
e
v
a
lua
t
io
n str
a
t
eg
y
T
o
d
em
o
n
s
tr
ate
th
e
c
o
n
tr
ib
u
tio
n
o
f
th
e
p
r
o
p
o
s
ed
f
r
am
ewo
r
k
,
th
e
E
HM
L
F
was
co
m
p
ar
ed
with
ea
c
h
in
d
iv
id
u
al
b
ase
class
if
ier
:
R
an
d
o
m
f
o
r
est,
XGBo
o
s
t,
C
atB
o
o
s
t,
an
d
L
i
g
h
tGB
M.
T
h
e
co
m
p
a
r
is
o
n
was
co
n
d
u
cte
d
u
s
in
g
th
e
s
am
e
tr
ain
in
g
an
d
in
d
e
p
en
d
e
n
t te
s
t p
ar
t
itio
n
s
.
T
h
e
co
m
p
ar
is
o
n
ad
d
r
ess
ed
t
wo
q
u
esti
o
n
s
.
First,
wh
eth
er
co
m
b
i
n
in
g
th
e
f
o
u
r
class
if
ier
s
th
r
o
u
g
h
ad
ap
tiv
e
weig
h
tin
g
p
r
o
v
id
es
b
etter
p
r
ed
ictiv
e
p
e
r
f
o
r
m
an
ce
th
an
u
s
in
g
an
in
d
iv
id
u
al
class
if
ier
.
Seco
n
d
,
wh
eth
er
th
e
f
in
al
en
s
em
b
le
p
r
o
v
id
es
in
ter
p
r
etab
le
f
ea
tu
r
e
c
o
n
tr
ib
u
tio
n
s
t
h
at
ca
n
b
e
tr
an
s
lated
in
to
m
ea
n
in
g
f
u
l
in
f
o
r
m
atio
n
f
o
r
u
n
i
v
er
s
ity
en
r
o
llm
en
t m
an
ag
e
m
en
t.
T
h
is
co
m
p
ar
ativ
e
d
esig
n
is
im
p
o
r
tan
t
f
o
r
estab
lis
h
in
g
t
h
e
co
n
tr
ib
u
tio
n
o
f
th
e
p
r
o
p
o
s
ed
f
r
am
ewo
r
k
r
ath
er
th
a
n
s
im
p
ly
r
ep
o
r
tin
g
th
e
p
er
f
o
r
m
a
n
ce
o
f
th
e
f
in
al
m
o
d
el.
Pre
v
i
o
u
s
s
tu
d
ies
h
a
v
e
d
em
o
n
s
tr
ated
th
e
ef
f
ec
tiv
en
ess
o
f
in
d
iv
i
d
u
al
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
s
an
d
m
u
lti
-
m
o
d
el
a
p
p
r
o
ac
h
e
s
in
ad
m
is
s
io
n
an
d
en
r
o
llm
en
t
p
r
ed
ictio
n
[6
]
,
[
9
]
,
[
10
]
,
[
18
]
,
[
19
]
.
T
h
er
ef
o
r
e,
th
e
p
r
esen
t
s
tu
d
y
ev
al
u
ates
wh
eth
er
th
e
in
teg
r
ati
o
n
o
f
f
ea
tu
r
e
s
elec
tio
n
,
ad
a
p
tiv
e
en
s
em
b
le
lear
n
in
g
,
a
n
d
SHAP
p
r
o
v
id
es
an
a
d
d
itio
n
al
m
eth
o
d
o
lo
g
ical
ad
v
an
tag
e.
2
.
1
3
.
Resea
rc
h f
ra
m
ew
o
rk
T
h
e
o
v
er
all
r
esear
ch
f
r
a
m
ewo
r
k
is
p
r
esen
ted
in
Fig
u
r
e
1
.
T
h
e
f
r
am
ewo
r
k
illu
s
tr
ates
th
e
s
eq
u
en
tial
r
elatio
n
s
h
ip
b
etwe
en
d
ata
p
r
ep
ar
atio
n
,
f
ea
tu
r
e
s
elec
tio
n
,
p
r
ed
ictiv
e
m
o
d
elin
g
,
en
s
em
b
le
in
teg
r
atio
n
,
a
n
d
ex
p
lain
ab
ilit
y
.
T
h
e
f
in
al
o
u
tp
u
t
co
n
s
is
ts
o
f
b
o
th
a
p
r
e
d
icte
d
en
r
o
llm
e
n
t
d
ec
is
io
n
a
n
d
a
n
ex
p
lan
atio
n
o
f
th
e
v
ar
iab
les co
n
tr
ib
u
tin
g
to
t
h
at
p
r
ed
ictio
n
.
T
h
e
m
eth
o
d
o
lo
g
ical
co
n
tr
ib
u
ti
o
n
o
f
E
HM
L
F
lies
in
th
e
in
teg
r
atio
n
o
f
th
ese
c
o
m
p
o
n
en
ts
in
t
o
a
s
in
g
le
ap
p
lican
t
-
lev
el
p
r
ed
ictio
n
p
ip
elin
e.
Pre
v
io
u
s
s
tu
d
ies
h
av
e
ex
a
m
in
ed
m
ac
h
in
e
lear
n
in
g
f
o
r
ad
m
is
s
io
n
an
d
en
r
o
llm
en
t
p
r
ed
ictio
n
[6
]
-
[
8
]
,
[
18
]
,
[
19
]
,
wh
ile
o
th
e
r
s
tu
d
ies
h
av
e
in
v
esti
g
ated
e
n
s
em
b
le
m
o
d
els
[9
]
,
[
10]
an
d
ex
p
lain
ab
le
m
ac
h
in
e
lear
n
in
g
in
ed
u
ca
tio
n
al
co
n
tex
ts
[
5
]
,
[
11
]
-
[
17
]
,
[
2
3
]
.
T
h
e
p
r
esen
t f
r
a
m
ewo
r
k
b
r
in
g
s
th
ese
m
eth
o
d
o
l
o
g
ical
s
tr
ea
m
s
to
g
eth
er
b
y
co
m
b
in
in
g
h
y
b
r
id
f
ea
tu
r
e
s
elec
tio
n
,
ad
ap
tiv
e
en
s
em
b
l
e
class
if
icatio
n
,
an
d
SHAP
-
b
ased
in
ter
p
r
etatio
n
with
in
th
e
s
am
e
en
r
o
llm
en
t p
r
ed
i
ctio
n
p
r
o
ce
s
s
.
T
h
is
d
esig
n
allo
ws
th
e
s
tu
d
y
to
ev
alu
ate
n
o
t
o
n
ly
wh
eth
e
r
an
ap
p
lican
t
is
lik
ely
to
e
n
r
o
l
l
b
u
t
also
wh
ich
f
ac
to
r
s
co
n
tr
ib
u
te
to
th
e
p
r
ed
ictio
n
an
d
h
o
w
s
tr
o
n
g
ly
th
ey
in
f
lu
e
n
ce
th
e
m
o
d
el
o
u
t
p
u
t.
C
o
n
s
eq
u
e
n
tly
,
th
e
f
r
am
ewo
r
k
is
in
ten
d
ed
to
s
u
p
p
o
r
t
ev
i
d
en
ce
-
b
ased
en
r
o
llm
en
t
m
an
a
g
em
en
t
w
h
ile
m
ain
tain
in
g
a
tr
an
s
p
ar
en
t
c
o
n
n
ec
tio
n
b
etwe
e
n
m
ac
h
i
n
e
lear
n
i
n
g
p
r
ed
ictio
n
s
an
d
th
e
f
ac
to
r
s
th
at
u
n
iv
er
s
it
y
d
ec
is
io
n
-
m
ak
er
s
ca
n
p
o
ten
tially
ac
t
u
p
o
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
5
2
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
4
3
,
No
.
3
,
Sep
tem
b
er
20
2
6
:
8
8
0
-
89
7
888
Fig
u
r
e
1
.
Pro
p
o
s
ed
E
HM
L
F
3.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
3
.
1
.
Da
t
a
s
et
cha
ra
c
t
er
is
t
ics
T
h
e
d
ataset
co
m
p
r
is
ed
2
,
0
0
0
ad
m
is
s
io
n
r
ec
o
r
d
s
co
llected
f
r
o
m
Un
iv
e
r
s
itas
W
ir
ah
u
s
ad
a
Me
d
an
an
d
Un
iv
er
s
itas
Au
d
i
I
n
d
o
n
esia
d
u
r
in
g
th
e
2
0
2
5
–
2
0
2
6
ad
m
is
s
io
n
p
er
io
d
.
E
ac
h
u
n
i
v
er
s
ity
co
n
tr
ib
u
ted
1
,
0
0
0
r
ec
o
r
d
s
,
r
esu
ltin
g
in
an
eq
u
al
in
s
titu
tio
n
al
r
ep
r
esen
tatio
n
o
f
5
0
%
f
o
r
ea
c
h
u
n
iv
er
s
ity
.
T
h
is
b
alan
ce
d
co
m
p
o
s
itio
n
p
r
o
v
id
es
a
s
u
itab
le
b
asis
f
o
r
d
ev
elo
p
in
g
a
p
r
ed
ictiv
e
m
o
d
el
with
o
u
t
allo
win
g
o
n
e
in
s
titu
tio
n
to
d
o
m
in
ate
th
e
lea
r
n
in
g
p
r
o
ce
s
s
.
T
h
e
d
em
o
g
r
ap
h
ic
c
h
ar
ac
ter
is
tics
o
f
th
e
s
tu
d
en
ts
ar
e
s
u
m
m
ar
ized
in
T
ab
les
2
an
d
3
.
Fem
al
e
s
tu
d
en
ts
r
ep
r
esen
ted
th
e
la
r
g
est p
r
o
p
o
r
t
io
n
o
f
t
h
e
d
ataset
(
5
5
.
0
%),
wh
ile
m
ale
s
tu
d
en
ts
ac
co
u
n
ted
f
o
r
4
5
.
0
%.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2502
-
4
7
5
2
E
xp
la
in
a
b
le
h
yb
r
id
ma
c
h
in
e
l
ea
r
n
in
g
fo
r
p
r
ed
ictin
g
s
tu
d
en
t
en
r
o
llmen
t d
ec
is
io
n
s
…
(
Dyb
io
Do
mp
u
Ho
t
A
s
ih
)
889
T
ab
le
2
.
Dis
tr
ib
u
tio
n
o
f
th
e
r
e
s
ea
r
ch
d
ataset
T
ab
le
3
.
Gen
d
er
d
is
tr
ib
u
ti
on
U
n
i
v
e
r
si
t
y
N
u
mb
e
r
o
f
S
t
u
d
e
n
t
s
P
e
r
c
e
n
t
a
g
e
U
n
i
v
e
r
si
t
a
s
W
i
r
a
h
u
sa
d
a
M
e
d
a
n
1
,
0
0
0
5
0
.
0
%
U
n
i
v
e
r
si
t
a
s
A
u
d
i
I
n
d
o
n
e
si
a
1
,
0
0
0
5
0
.
0
%
To
t
a
l
2
,
0
0
0
1
0
0
.
0
%
G
e
n
d
e
r
F
r
e
q
u
e
n
c
y
P
e
r
c
e
n
t
a
g
e
M
a
l
e
9
0
0
4
5
.
0
%
F
e
mal
e
1
,
1
0
0
5
5
.
0
%
To
t
a
l
2
,
0
0
0
1
0
0
.
0
%
T
h
e
ag
e
d
is
tr
ib
u
tio
n
s
h
o
wn
in
T
ab
le
4
in
d
icate
s
th
at
m
o
s
t
ad
m
itted
s
tu
d
en
ts
wer
e
b
etwe
en
1
8
a
n
d
1
9
y
ea
r
s
o
ld
,
r
e
p
r
esen
tin
g
5
5
.
0
%
o
f
th
e
d
ataset.
Stu
d
e
n
ts
ag
ed
2
0
–
2
1
y
ea
r
s
ac
co
u
n
ted
f
o
r
3
5
.
0
%,
wh
ile
s
tu
d
e
n
ts
ag
ed
2
2
y
ea
r
s
r
ep
r
esen
ted
1
5
.
0
%
o
f
th
e
o
b
s
er
v
atio
n
s
.
T
h
e
tar
g
et
v
ar
iab
le
was
s
tu
d
en
t
e
n
r
o
llm
en
t
d
ec
is
io
n
,
d
ef
in
ed
as
a
b
in
ar
y
o
u
tco
m
e
.
Ap
p
lican
ts
wh
o
co
m
p
leted
th
e
r
eg
is
tr
atio
n
p
r
o
ce
s
s
wer
e
cl
ass
if
ied
as
E
n
r
o
lle
d
(
1
)
,
wh
e
r
ea
s
ap
p
lican
ts
wh
o
d
id
n
o
t
c
o
m
p
lete
r
e
g
is
tr
atio
n
a
f
ter
r
ec
eiv
in
g
an
ad
m
is
s
io
n
o
f
f
er
we
r
e
class
if
ied
as n
o
t e
n
r
o
lled
(
0
)
.
T
h
is
f
o
r
m
u
latio
n
is
co
n
s
is
ten
t w
ith
p
r
ev
io
u
s
r
esear
ch
th
at
tr
ea
ts
en
r
o
ll
m
en
t p
r
ed
ictio
n
as a
class
if
icatio
n
p
r
o
b
lem
a
n
d
d
e
m
o
n
s
tr
ates
th
e
p
o
ten
tial
o
f
m
ac
h
in
e
lear
n
i
n
g
f
o
r
i
d
en
tify
in
g
ap
p
lican
ts
wh
o
ar
e
lik
ely
to
p
r
o
ce
ed
w
ith
r
e
g
is
tr
atio
n
[
6
]
-
[
8
]
.
T
h
e
u
s
e
o
f
ap
p
lican
t
-
lev
el
e
n
r
o
llm
en
t
d
ata
also
d
is
tin
g
u
is
h
es
th
is
s
tu
d
y
f
r
o
m
r
esear
c
h
th
at
f
o
cu
s
es
o
n
s
tu
d
en
t
p
er
f
o
r
m
a
n
ce
,
d
r
o
p
o
u
t,
o
r
co
u
r
s
e
en
r
o
llm
en
t.
Pr
ev
io
u
s
ex
p
lain
ab
le
ed
u
ca
tio
n
a
l
m
ac
h
in
e
lear
n
in
g
s
tu
d
ies
h
av
e
p
r
e
d
o
m
in
a
n
tl
y
in
v
esti
g
ated
ac
ad
em
ic
p
e
r
f
o
r
m
a
n
ce
,
ad
ap
ta
b
ilit
y
,
co
u
r
s
e
s
u
cc
ess
,
o
r
d
r
o
p
o
u
t
r
is
k
[
5
]
,
[
1
2
]
-
[
1
7
]
.
I
n
co
n
t
r
ast,
th
e
p
r
esen
t
s
tu
d
y
f
o
cu
s
es
s
p
ec
if
i
ca
lly
o
n
th
e
tr
a
n
s
itio
n
f
r
o
m
a
d
m
is
s
io
n
to
ac
tu
al
en
r
o
llm
en
t,
wh
ic
h
r
ep
r
esen
ts
an
ea
r
lier
in
s
titu
tio
n
al
d
ec
is
io
n
p
o
in
t
an
d
h
as
d
ir
ec
t
im
p
licat
io
n
s
f
o
r
r
ec
r
u
itm
en
t
m
an
ag
em
en
t.
T
ab
le
4
.
Ag
e
d
is
tr
ib
u
tio
n
A
g
e
F
r
e
q
u
e
n
c
y
P
e
r
c
e
n
t
a
g
e
18
–
1
9
y
e
a
r
s
1
,
0
0
0
5
5
.
0
%
20
–
2
1
y
e
a
r
s
7
0
0
3
5
.
0
%
2
2
y
e
a
r
s
3
0
0
1
5
.
0
%
To
t
a
l
2
,
0
0
0
1
0
0
.
0
%
3
.
2
.
H
y
brid f
ea
t
ure
s
elec
t
io
n r
esu
lt
s
T
h
e
in
itial
d
ataset
co
n
tain
ed
1
5
p
r
e
d
icto
r
v
ar
iab
les
(
T
ab
l
e
5
)
co
v
e
r
in
g
d
em
o
g
r
ap
h
ic,
ac
ad
em
ic,
in
s
titu
tio
n
al,
f
in
an
cial,
co
m
m
u
n
icatio
n
,
an
d
d
i
g
ital
en
g
ag
e
m
en
t
ch
ar
ac
ter
is
tics
.
T
h
e
h
y
b
r
id
f
ea
tu
r
e
s
elec
tio
n
p
r
o
ce
d
u
r
e
c
o
m
b
in
e
d
B
o
r
u
ta
an
d
MI
to
i
d
en
tify
p
r
e
d
icto
r
s
th
at
d
em
o
n
s
tr
ated
co
n
s
is
ten
t
r
elev
an
ce
to
th
e
en
r
o
llm
en
t
o
u
tco
m
e.
T
h
e
an
aly
s
is
r
ed
u
ce
d
th
e
o
r
ig
in
al
1
5
v
a
r
iab
les
to
1
0
p
r
e
d
ic
to
r
s
.
Un
iv
er
s
ity
r
ep
u
tatio
n
o
b
tain
ed
th
e
h
ig
h
est
MI
s
co
r
e
(
0
.
8
1
2
)
,
f
o
ll
o
wed
b
y
d
i
g
ital
m
ar
k
etin
g
e
n
g
ag
em
en
t
(
0
.
7
8
4
)
,
s
ch
o
lar
s
h
ip
av
ailab
ilit
y
(
0
.
7
6
3
)
,
tu
itio
n
af
f
o
r
d
ab
ilit
y
(
0
.
7
4
1
)
,
a
n
d
ac
ad
em
ic
f
ac
ilit
ies
(
0
.
7
1
8
)
.
Ser
v
ice
q
u
ality
,
p
r
ev
io
u
s
ac
ad
em
ic
ac
h
iev
em
en
t,
s
o
cial
m
ed
ia
in
ter
ac
tio
n
,
a
d
m
i
s
s
io
n
p
ath
way
,
an
d
co
m
m
u
n
ic
atio
n
in
ten
s
ity
wer
e
also
r
etain
ed
.
T
h
e
r
esu
lts
in
d
icate
(
Fig
u
r
e
2
)
th
at
in
s
titu
tio
n
al
an
d
f
in
an
ci
al
v
ar
iab
les
wer
e
m
o
r
e
in
f
o
r
m
ativ
e
th
an
b
asic
d
em
o
g
r
a
p
h
ic
v
a
r
iab
les.
Un
iv
er
s
ity
R
ep
u
tatio
n
was
th
e
s
tr
o
n
g
est
p
r
ed
icto
r
,
s
u
g
g
esti
n
g
th
at
ap
p
lican
ts
'
p
er
ce
p
tio
n
s
o
f
in
s
titu
tio
n
al
c
r
ed
ib
ilit
y
an
d
q
u
ality
ar
e
cl
o
s
ely
ass
o
ciate
d
with
th
eir
en
r
o
llm
en
t
d
ec
is
io
n
s
.
Dig
ital
Ma
r
k
etin
g
E
n
g
a
g
em
e
n
t
r
an
k
ed
s
ec
o
n
d
,
in
d
icatin
g
t
h
at
in
ter
ac
tio
n
with
d
ig
ital
p
r
o
m
o
tio
n
al
ch
an
n
els
m
ay
p
r
o
v
id
e
u
s
ef
u
l in
f
o
r
m
atio
n
f
o
r
d
is
tin
g
u
is
h
in
g
ap
p
lican
ts
wh
o
ar
e
m
o
r
e
lik
el
y
to
p
r
o
ce
e
d
with
en
r
o
llm
en
t.
I
llu
s
tr
ates th
e
r
elativ
e
im
p
o
r
ta
n
ce
o
f
th
e
s
elec
ted
v
ar
iab
les a
f
te
r
h
y
b
r
id
f
ea
t
u
r
e
s
elec
tio
n
.
T
ab
le
5
.
Hy
b
r
id
f
ea
t
u
r
e
s
elec
tio
n
r
esu
lts
No
P
r
e
d
i
c
t
o
r
v
a
r
i
a
b
l
e
B
o
r
u
t
a
M
u
t
u
a
l
i
n
f
o
r
m
a
t
i
o
n
F
i
n
a
l
d
e
c
i
s
i
o
n
1
U
n
i
v
e
r
si
t
y
R
e
p
u
t
a
t
i
o
n
A
c
c
e
p
t
e
d
0
.
8
1
2
S
e
l
e
c
t
e
d
2
D
i
g
i
t
a
l
M
a
r
k
e
t
i
n
g
E
n
g
a
g
e
me
n
t
A
c
c
e
p
t
e
d
0
.
7
8
4
S
e
l
e
c
t
e
d
3
S
c
h
o
l
a
r
sh
i
p
A
v
a
i
l
a
b
i
l
i
t
y
A
c
c
e
p
t
e
d
0
.
7
6
3
S
e
l
e
c
t
e
d
4
Tu
i
t
i
o
n
A
f
f
o
r
d
a
b
i
l
i
t
y
A
c
c
e
p
t
e
d
0
.
7
4
1
S
e
l
e
c
t
e
d
5
A
c
a
d
e
mi
c
F
a
c
i
l
i
t
i
e
s
A
c
c
e
p
t
e
d
0
.
7
1
8
S
e
l
e
c
t
e
d
6
S
e
r
v
i
c
e
Q
u
a
l
i
t
y
A
c
c
e
p
t
e
d
0
.
6
9
4
S
e
l
e
c
t
e
d
7
P
r
e
v
i
o
u
s Ac
a
d
e
mi
c
A
c
h
i
e
v
e
m
e
n
t
A
c
c
e
p
t
e
d
0
.
6
6
1
S
e
l
e
c
t
e
d
8
S
o
c
i
a
l
M
e
d
i
a
I
n
t
e
r
a
c
t
i
o
n
A
c
c
e
p
t
e
d
0
.
6
3
5
S
e
l
e
c
t
e
d
9
A
d
mi
ss
i
o
n
P
a
t
h
w
a
y
A
c
c
e
p
t
e
d
0
.
5
9
2
S
e
l
e
c
t
e
d
10
C
o
mm
u
n
i
c
a
t
i
o
n
I
n
t
e
n
si
t
y
A
c
c
e
p
t
e
d
0
.
5
6
3
S
e
l
e
c
t
e
d
11
G
e
n
d
e
r
R
e
j
e
c
t
e
d
0
.
2
1
1
R
e
m
o
v
e
d
12
A
g
e
R
e
j
e
c
t
e
d
0
.
1
9
8
R
e
m
o
v
e
d
13
R
e
g
i
o
n
o
f
O
r
i
g
i
n
R
e
j
e
c
t
e
d
0
.
1
8
4
R
e
m
o
v
e
d
14
P
a
r
e
n
t
s’
O
c
c
u
p
a
t
i
o
n
R
e
j
e
c
t
e
d
0
.
1
7
1
R
e
m
o
v
e
d
15
P
a
r
e
n
t
s’
E
d
u
c
a
t
i
o
n
R
e
j
e
c
t
e
d
0
.
1
5
9
R
e
m
o
v
e
d
Evaluation Warning : The document was created with Spire.PDF for Python.