I
nd
o
ne
s
ia
n J
o
urna
l o
f
E
lect
rica
l En
g
ineering
a
nd
Co
m
p
u
t
er
Science
Vo
l.
21
,
No
.
2
,
Feb
r
u
ar
y
2
0
2
1
,
p
p
.
1
1
5
1
~1
1
59
I
SS
N:
2
5
02
-
4
7
5
2
,
DOI
: 1
0
.
1
1
5
9
1
/i
j
ee
cs.
v
2
1
.
i2
.
p
p
1
1
5
1
-
11
59
1151
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ee
cs.ia
esco
r
e.
co
m
K
no
w
ledg
e
d
iscovery
f
ro
m
g
ene
e
x
press
io
n
d
a
tas
et
u
sing
ba
g
g
ing
la
ss
o
deci
sio
n t
ree
U
m
u Sa
’
a
da
h
,
M
a
s
it
ho
h Ye
s
s
i R
o
cha
y
a
ni
,
Ani B
ud
i A
s
t
u
t
i
F
a
c
u
lt
y
o
f
M
a
th
e
m
a
ti
c
s an
d
Na
tu
ra
l
S
c
ien
c
e
s,
Un
iv
e
rsitas
Bra
w
ij
a
y
a
,
In
d
o
n
e
sia
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
u
n
1
8
,
2
0
2
0
R
ev
i
s
ed
A
u
g
1
1
,
2
0
2
0
A
cc
ep
ted
Sep
7
,
2
0
2
0
Clas
sify
in
g
h
ig
h
-
d
im
e
n
sio
n
a
l
d
a
ta
a
re
a
c
h
a
ll
e
n
g
in
g
tas
k
in
d
a
t
a
m
in
in
g
.
G
e
n
e
e
x
p
re
ss
io
n
d
a
ta i
s a t
y
p
e
o
f
h
ig
h
-
d
im
e
n
sio
n
a
l
d
a
ta t
h
a
t
h
a
s th
o
u
sa
n
d
s o
f
f
e
a
tu
re
s.
T
h
e
stu
d
y
w
a
s
p
ro
p
o
sin
g
a
m
e
th
o
d
to
e
x
trac
t
k
n
o
w
led
g
e
f
ro
m
h
ig
h
-
d
im
e
n
sio
n
a
l
g
e
n
e
e
x
p
re
ss
io
n
d
a
t
a
b
y
s
e
lec
ti
n
g
f
e
a
tu
re
s
a
n
d
c
las
sify
in
g
.
L
a
ss
o
w
a
s
u
se
d
f
o
r
s
e
lec
ti
n
g
f
e
a
tu
re
s
a
n
d
th
e
c
las
sif
ica
ti
o
n
a
n
d
re
g
r
e
ss
io
n
tree
(CA
R
T
)
a
l
g
o
rit
h
m
w
a
s
u
se
d
to
c
o
n
str
u
c
t
th
e
d
e
c
isio
n
tree
m
o
d
e
l.
T
o
e
x
a
m
in
e
th
e
sta
b
il
it
y
o
f
th
e
las
so
d
e
c
isio
n
tree
,
w
e
p
e
rf
o
r
m
e
d
b
o
o
t
stra
p
a
g
g
r
e
g
a
ti
n
g
(B
a
g
g
in
g
)
w
it
h
5
0
re
p
li
c
a
ti
o
n
s.
T
h
e
g
e
n
e
e
x
p
re
ss
io
n
d
a
ta
u
se
d
w
a
s
a
n
o
v
a
rian
t
u
m
o
r
d
a
tas
e
t
t
h
a
t
h
a
s
1
,
5
4
5
o
b
se
rv
a
ti
o
n
s,
1
0
,
9
3
5
g
e
n
e
f
e
a
tu
re
s,
a
n
d
b
in
a
ry
c
las
s.
T
h
e
f
in
d
in
g
s
o
f
th
is
re
se
a
rc
h
sh
o
w
e
d
th
a
t
th
e
las
so
d
e
c
isio
n
tree
c
o
u
ld
p
r
o
d
u
c
e
a
n
in
terp
re
tab
le
m
o
d
e
l
th
a
t
th
e
o
re
ti
c
a
ll
y
c
o
rre
c
t
a
n
d
h
a
d
a
n
a
c
c
u
ra
c
y
o
f
8
9
.
3
2
%
.
M
e
a
n
w
h
il
e
,
th
e
m
o
d
e
l
o
b
tain
e
d
f
ro
m
th
e
m
a
jo
rit
y
v
o
te
g
a
v
e
a
n
a
c
c
u
ra
c
y
o
f
9
0
.
2
9
%
w
h
ich
sh
o
w
e
d
a
n
i
n
c
re
a
se
in
a
c
c
u
ra
c
y
o
f
1
%
f
ro
m
th
e
sin
g
le
las
so
d
e
c
isio
n
tree
m
o
d
e
l.
T
h
e
sli
g
h
tl
y
in
c
re
a
sin
g
a
c
c
u
ra
c
y
sh
o
w
s th
a
t
th
e
las
so
d
e
c
isio
n
tree
c
las
sif
ier
is
s
tab
le.
K
ey
w
o
r
d
s
:
B
ag
g
i
n
g
Dec
is
io
n
t
r
ee
Featu
r
e
s
elec
t
io
n
Gen
e
ex
p
r
es
s
io
n
Hig
h
-
d
i
m
e
n
s
io
n
al
T
h
is
is
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
:
U
m
u
Sa
‟
ad
ah
Dep
ar
t
m
en
t o
f
Stati
s
tics
Un
i
v
er
s
ita
s
B
r
a
w
ij
a
y
a
J
alan
Vete
r
an
,
Ma
lan
g
,
I
n
d
o
n
esia
E
m
ail:
u
.
s
aa
d
ah
@
u
b
.
ac
.
id
1.
I
NT
RO
D
UCT
I
O
N
Gen
e
ex
p
r
es
s
io
n
d
ata
h
a
v
e
b
ee
n
u
s
ed
to
s
tu
d
y
th
e
d
i
f
f
e
r
en
ce
s
in
g
en
e
c
h
ar
ac
ter
is
tic
s
b
et
w
ee
n
p
atien
ts
w
it
h
ce
r
tai
n
d
i
s
ea
s
es
an
d
n
o
r
m
al
p
eo
p
le.
T
h
e
m
aj
o
r
ch
alle
n
g
e
to
a
n
al
y
ze
g
en
e
e
x
p
r
ess
io
n
d
ata
i
s
it
h
as
m
a
n
y
p
r
ed
icto
r
s
(
g
e
n
e
s)
,
b
u
t
th
e
s
a
m
p
le
is
m
u
ch
les
s
.
Gen
e
e
x
p
r
ess
io
n
d
ata
is
a
t
y
p
e
o
f
h
i
g
h
-
d
i
m
e
n
s
io
n
al
d
ata
th
at
co
n
s
is
t
o
f
t
h
o
u
s
an
d
s
,
ev
en
ten
s
o
f
t
h
o
u
s
a
n
d
s
o
f
g
e
n
e
f
ea
tu
r
es,
b
u
t
t
h
e
s
a
m
p
le
s
iz
e
is
o
n
l
y
h
u
n
d
r
ed
s
.
T
h
er
ef
o
r
e,
a
ce
r
tain
s
tr
ateg
y
i
s
n
ee
d
ed
to
d
ea
l w
it
h
d
i
m
en
s
io
n
al
p
r
o
b
lem
s
in
g
e
n
e
ex
p
r
ess
i
o
n
d
ata.
On
e
o
f
t
h
e
s
tr
ateg
ie
s
i
n
th
e
c
lass
i
f
icatio
n
o
f
h
ig
h
-
d
i
m
e
n
s
io
n
al
d
ata
i
s
b
y
r
ed
u
ci
n
g
t
h
e
d
i
m
e
n
s
io
n
.
T
h
er
e
ar
e
tw
o
ap
p
r
o
ac
h
es
in
d
i
m
e
n
s
io
n
r
ed
u
ctio
n
n
a
m
e
l
y
f
ea
tu
r
e
e
x
tr
ac
tio
n
a
n
d
f
ea
tu
r
e
s
elec
tio
n
.
T
h
e
co
m
m
o
n
d
i
m
e
n
s
io
n
r
ed
u
ctio
n
ap
p
r
o
ac
h
in
g
en
e
ex
p
r
es
s
i
o
n
d
ata
is
f
ea
tu
r
e
s
elec
tio
n
.
Featu
r
e
s
elec
tio
n
eli
m
i
n
ate
s
ir
r
elev
a
n
t
a
n
d
r
ed
u
n
d
an
t
f
ea
t
u
r
es.
R
e
s
ea
r
ch
[
1
]
i
n
v
e
s
ti
g
ated
th
e
in
f
l
u
e
n
ce
o
f
f
ea
tu
r
e
s
elec
tio
n
o
n
th
e
ac
cu
r
ac
y
o
f
th
e
clas
s
i
f
icat
io
n
o
f
g
e
n
e
ex
p
r
ess
io
n
d
ata.
T
h
e
r
esu
lt
o
f
th
e
s
t
u
d
y
w
a
s
f
ea
tu
r
e
s
elec
tio
n
ca
n
in
cr
ea
s
e
ac
cu
r
ac
y
u
p
to
9
%.
Sev
er
al
m
et
h
o
d
s
t
h
at
co
m
b
i
n
e
f
ea
t
u
r
e
s
elec
tio
n
a
n
d
clas
s
i
f
icatio
n
h
av
e
b
ee
n
i
m
p
le
m
e
n
ted
in
th
e
class
i
f
icatio
n
o
f
g
e
n
e
e
x
p
r
ess
io
n
d
ata.
A
s
s
a
w
a
m
ak
i
n
et.
al.
[
2
]
u
s
ed
re
cu
r
s
iv
e
f
e
at
u
re
e
li
m
in
at
ion
(
R
FE)
to
s
elec
t
g
e
n
es
a
n
d
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
e
(
SVM)
to
clas
s
i
f
y
s
e
v
er
al
g
e
n
e
e
x
p
r
ess
io
n
d
ata.
Kan
g
e
t.a
l.
[
3
]
p
r
o
p
o
s
ed
a
h
y
b
r
id
m
eth
o
d
o
f
R
elax
ed
L
a
s
s
o
a
n
d
Gen
er
aliz
ed
SVM
f
o
r
t
h
e
m
u
lticla
s
s
cl
ass
i
f
icatio
n
o
f
g
e
n
e
ex
p
r
ess
io
n
d
ata.
I
n
th
e
p
ap
er
,
Kan
g
et.
al.
m
e
n
tio
n
ed
th
e
s
e
lecte
d
g
e
n
es,
b
u
t
t
h
ese
r
e
s
u
l
t
s
ar
e
n
o
t
v
al
id
ated
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4752
I
n
d
o
n
esia
n
J
E
lec
E
n
g
&
C
o
m
p
Sci,
Vo
l.
21
,
No
.
2
,
Feb
r
u
ar
y
2
0
2
1
:
1
1
51
-
11
59
1152
b
ased
o
n
th
eo
r
y
.
Gao
et.
al.
[
4
]
u
s
ed
in
f
o
r
m
atio
n
g
ain
,
g
ain
r
atio
,
r
elief
,
an
d
co
r
r
elatio
n
to
s
elec
t
g
en
es
a
n
d
SVM
f
o
r
class
if
y
i
n
g
ca
n
ce
r
.
H
o
w
e
v
er
,
th
o
s
e
s
t
u
d
ies o
n
l
y
f
o
cu
s
ed
o
n
t
h
e
ac
cu
r
ac
y
o
f
th
e
m
o
d
el
o
b
tain
ed
.
A
cc
u
r
ac
y
,
in
ter
p
r
etab
ilit
y
,
a
n
d
m
o
d
el
s
ize
ar
e
t
h
r
ee
i
m
p
o
r
t
an
t
a
s
p
ec
ts
o
f
m
ac
h
in
e
lear
n
i
n
g
[
5
]
,
an
d
m
an
y
a
u
t
h
o
r
s
a
g
r
ee
t
h
at
m
o
d
el
in
ter
p
r
etab
ilit
y
is
th
e
m
ain
co
n
ce
r
n
[
6
]
.
I
n
t
h
e
cla
s
s
if
ica
tio
n
o
f
g
en
e
ex
p
r
ess
io
n
d
ata,
i
n
ter
p
r
etatio
n
is
v
er
y
i
m
p
o
r
ta
n
t
b
esid
es
t
h
e
m
o
d
el
ac
cu
r
ac
y
.
C
las
s
i
f
icati
o
n
an
d
R
e
g
r
es
s
io
n
T
r
ee
(
C
A
R
T
)
,
p
r
o
p
o
s
ed
b
y
B
r
ei
m
a
n
et.
al.
[
7
]
,
is
a
Dec
i
s
io
n
T
r
ee
alg
o
r
ith
m
t
h
at
i
s
p
r
o
m
o
t
ed
to
b
e
ea
s
il
y
u
s
ed
in
i
n
ter
p
r
etatio
n
to
ex
p
lo
r
e
k
n
o
w
led
g
e
f
r
o
m
th
e
d
ata
[
8
]
.
T
h
e
C
A
R
T
alg
o
r
it
h
m
is
o
f
te
n
u
s
ed
i
n
d
ec
is
io
n
an
al
y
s
is
to
v
is
u
alize
d
ec
is
io
n
m
a
k
in
g
.
B
esid
e
s
t
h
at,
t
h
e
C
AR
T
alg
o
r
ith
m
a
ls
o
ca
n
b
e
u
s
ed
f
o
r
f
ea
tu
r
e
s
elec
tio
n
.
B
u
t
C
AR
T
r
eq
u
ir
es
h
i
g
h
co
m
p
u
tat
io
n
al
co
s
t
s
to
w
o
r
k
o
n
v
er
y
lar
g
e
d
ata
s
u
ch
as
g
e
n
e
e
x
p
r
ess
io
n
d
ata.
T
h
er
ef
o
r
e,
ap
p
ly
i
n
g
t
h
e
f
ea
tu
r
e
s
elec
tio
n
b
ef
o
r
e
ap
p
ly
i
n
g
C
AR
T
is
v
er
y
u
s
ef
u
l.
R
o
ch
a
y
an
i,
et.
al.
[
9
]
co
m
b
in
e
d
L
as
s
o
r
eg
u
lar
izatio
n
a
n
d
D
ec
is
io
n
T
r
ee
to
s
elec
t
g
e
n
es
a
n
d
clas
s
i
f
y
g
en
e
ex
p
r
es
s
io
n
d
ata.
T
h
e
L
ass
o
r
eg
u
lar
izatio
n
h
as
a
lo
w
co
m
p
u
ta
tio
n
al
co
s
t.
Mo
d
els
o
b
tain
ed
f
r
o
m
t
h
e
L
as
s
o
Dec
is
io
n
T
r
ee
ar
e
also
ea
s
y
to
in
ter
p
r
et
an
d
th
eo
r
etic
all
y
co
r
r
ec
t.
I
n
th
is
s
tu
d
y
,
w
e
ar
e
in
ter
ested
in
ex
a
m
in
i
n
g
t
h
e
s
tab
ilit
y
o
f
t
h
e
L
as
s
o
Dec
i
s
io
n
T
r
ee
.
A
s
tab
le
class
if
ier
is
a
cla
s
s
i
f
ier
f
o
r
wh
ich
t
h
e
p
r
ed
ictio
n
d
o
es n
o
t c
h
an
g
e
m
u
c
h
w
h
en
t
h
e
s
li
g
h
t
m
o
d
i
f
icatio
n
o
cc
u
r
s
i
n
th
e
tr
ai
n
i
n
g
s
et.
B
ag
g
i
n
g
,
i
n
tr
o
d
u
ce
d
b
y
B
r
eim
an
[
1
0
]
,
is
o
n
e
o
f
th
e
en
s
e
m
b
le
m
et
h
o
d
s
th
a
t
h
as
b
ee
n
w
i
d
ely
u
s
ed
t
o
in
cr
ea
s
e
th
e
ac
cu
r
ac
y
o
f
p
r
ed
ictio
n
m
o
d
els
[
1
1
]
,
im
p
r
o
v
e
th
e
r
o
b
u
s
tn
ess
a
n
d
s
tab
ilit
y
o
f
t
h
e
m
o
d
el
[
1
2
,
1
3
]
an
d
h
a
n
d
le
u
n
b
alan
ce
d
clas
s
p
r
o
b
lem
s
[
1
4
]
.
A
cc
o
r
d
in
g
to
B
r
eim
a
n
[
1
0
]
,
im
p
r
o
v
i
n
g
th
e
ac
cu
r
ac
y
o
f
B
ag
g
i
n
g
d
ep
en
d
s
o
n
th
e
s
tab
ilit
y
o
f
t
h
e
class
i
f
ier
.
B
ag
g
in
g
u
s
u
all
y
co
u
ld
i
m
p
r
o
v
e
ac
c
u
r
ac
y
o
n
u
n
s
t
ab
le
class
i
f
ier
s
b
u
t
ca
n
n
o
t
o
n
s
tab
le
class
i
f
ier
s
.
As
an
en
s
e
m
b
le
m
et
h
o
d
,
B
ag
g
i
n
g
co
m
b
i
n
es
s
ev
er
al
s
i
n
g
le
m
o
d
els
in
to
o
n
e
f
i
n
al
m
o
d
el
b
ased
o
n
th
e
m
aj
o
r
ity
v
o
tes.
C
o
m
m
o
n
l
y
,
t
h
e
s
i
n
g
le
class
i
f
ier
f
o
r
B
ag
g
i
n
g
is
t
h
e
D
ec
is
io
n
T
r
ee
.
B
u
t
r
ec
en
tl
y
B
ag
g
i
n
g
h
a
s
b
ee
n
u
s
ed
in
o
th
er
m
ac
h
i
n
e
lear
n
i
n
g
alg
o
r
ith
m
s
s
u
c
h
as
B
ag
g
i
n
g
C
o
n
v
o
l
u
tio
n
al
Ne
u
r
al
Net
w
o
r
k
[
1
5
]
an
d
B
ag
g
in
g
Ne
ar
est Ne
ig
h
b
o
r
Su
p
p
o
r
t V
ec
to
r
Ma
ch
in
e
[
1
6
]
.
W
e
p
er
f
o
r
m
ed
th
e
B
a
g
g
i
n
g
L
ass
o
Dec
is
io
n
T
r
ee
to
ex
a
m
in
e
th
e
s
tab
ili
t
y
o
f
th
e
L
a
s
s
o
De
cisi
o
n
T
r
ee
in
m
o
d
elin
g
t
h
e
g
en
e
e
x
p
r
ess
i
o
n
d
ataset.
T
h
e
d
ataset
u
s
ed
w
a
s
t
h
e
o
v
ar
ia
n
t
u
m
o
r
d
atase
t
.
I
t
is
i
n
ter
esti
n
g
to
u
s
e
t
h
is
d
atase
t
s
i
n
ce
o
v
ar
ian
ca
n
ce
r
is
o
n
e
o
f
t
h
e
m
o
s
t
f
ata
l
d
is
ea
s
es
i
n
w
o
m
e
n
w
h
ich
i
s
f
r
eq
u
en
t
l
y
s
tr
ik
e
s
p
o
s
t
-
m
e
n
o
p
au
s
al
w
o
m
e
n
[
1
7
,
18]
.
T
h
e
s
u
cc
ess
o
f
th
e
t
w
o
m
et
h
o
d
s
w
i
ll
b
e
v
er
y
u
s
ef
u
l
i
n
m
ed
ical
r
esear
ch
,
esp
ec
iall
y
to
d
is
co
v
er
n
e
w
k
n
o
w
led
g
e
f
r
o
m
a
d
is
ea
s
e.
2.
T
H
E
O
R
E
T
I
CA
L
B
ACK
G
R
O
UND
2
.
1
.
L
o
g
is
t
ic
re
g
re
s
s
io
n
B
in
ar
y
lo
g
is
tic
r
e
g
r
ess
io
n
is
u
s
ed
to
m
o
d
el
d
ataset
w
it
h
a
b
in
ar
y
r
esp
o
n
s
e
v
ar
iab
le.
T
h
e
s
tan
d
ar
d
b
in
o
m
ial
lo
g
is
t
ic
r
eg
r
ess
io
n
m
o
d
el
is
s
tated
as a
lo
g
o
f
o
d
d
s
:
(
(
)
(
)
)
∑
(
1
)
w
h
er
e
(
)
(
∑
)
,
is
th
e
o
b
s
er
v
atio
n
in
d
ex
,
is
th
e
i
n
d
ex
o
f
p
r
ed
icto
r
v
ar
iab
le,
is
th
e
in
ter
ce
p
t,
an
d
is
th
e
r
eg
r
ess
i
o
n
co
ef
f
icie
n
t
o
f
th
p
r
e
d
icto
r
v
ar
iab
le.
T
h
e
esti
m
atio
n
i
n
lo
g
i
s
tic
r
eg
r
es
s
io
n
p
ar
a
m
eter
s
w
as c
o
n
d
u
cted
b
y
m
a
x
i
m
izi
n
g
th
e
lo
g
-
l
ik
el
ih
o
o
d
f
u
n
ct
io
n
:
(
)
∑
*
(
∑
)
.
∑
/
+
(
2
)
No
te
th
at
th
e
f
ir
s
t
ter
m
o
f
(
2
)
,
∑
,
is
a
f
o
r
m
o
f
a
f
f
in
e
f
u
n
ctio
n
,
th
er
ef
o
r
e
it
is
co
n
ca
v
e
.
An
d
th
e
s
ec
o
n
d
ter
m
,
.
∑
/
,
is
al
s
o
co
n
ca
v
e.
Si
n
ce
t
h
e
s
u
m
o
f
co
n
ca
v
e
f
u
n
ctio
n
s
i
s
also
co
n
ca
v
e,
h
en
ce
(
)
is
a
co
n
ca
v
e
f
u
n
ct
io
n
a
n
d
it
i
m
p
lies
th
a
t
th
e
n
e
g
ati
v
e
lo
g
-
li
k
eli
h
o
o
d
,
i.e
.
(
)
,
is
a
c
o
n
v
ex
f
u
n
ctio
n
.
T
h
e
n
eg
ati
v
e
lo
g
-
l
ik
el
ih
o
o
d
is
also
ca
lled
th
e
o
b
j
ec
tiv
e
f
u
n
c
tio
n
o
f
lo
g
is
tic
r
eg
r
es
s
io
n
.
T
h
e
ad
v
a
n
tag
e
o
f
co
n
v
ex
it
y
o
f
n
e
g
ati
v
e
lo
g
-
l
ik
el
ih
o
o
d
f
u
n
ctio
n
g
u
ar
an
tees
th
at
t
h
e
lo
ca
l
o
p
tim
u
m
i
s
al
s
o
a
g
lo
b
al
o
p
ti
m
u
m
.
A
n
o
p
ti
m
izat
io
n
al
g
o
r
ith
m
f
o
r
a
co
n
v
ex
f
u
n
ctio
n
s
u
ch
as
t
h
e
Ne
w
to
n
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esia
n
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2502
-
4752
K
n
o
w
led
g
e
d
is
co
ve
r
y
fr
o
m
g
en
e
ex
p
r
ess
io
n
d
a
ta
s
et
u
s
in
g
b
a
g
g
in
g
l
a
s
s
o
d
ec
is
io
n
tr
ee
(
Um
u
S
a
’
a
d
a
h
)
1153
m
et
h
o
d
or
th
e
g
r
ad
ien
t
d
esce
n
t
ca
n
b
e
i
m
p
le
m
en
ted
to
s
o
l
v
e
t
h
e
lo
g
is
t
ic
r
eg
r
e
s
s
io
n
p
r
o
b
le
m
.
T
h
e
esti
m
ated
p
ar
am
eter
s
o
f
lo
g
i
s
tic
r
eg
r
es
s
i
o
n
ar
e
s
tated
in
(
3
)
.
̂
[
(
)
]
(
3
)
2
.
2
.
Reg
ula
riza
t
io
n
R
eg
u
lar
izatio
n
is
a
m
et
h
o
d
to
av
o
id
o
v
er
f
itti
n
g
b
y
ad
d
in
g
co
n
s
tr
ai
n
ts
w
h
ile
s
o
l
v
i
n
g
t
h
e
o
p
ti
m
izatio
n
p
r
o
b
lem
.
L
et
t
h
e
r
eg
u
lar
izati
o
n
f
u
n
ctio
n
is
d
en
o
ted
b
y
(
)
,
w
h
er
e
(
)
.
I
n
th
e
r
eg
u
lar
ized
lo
g
is
tic
r
eg
r
es
s
io
n
,
th
e
co
n
s
t
r
ain
t
w
a
s
ad
d
ed
to
t
h
e
lo
g
is
t
ic
r
eg
r
es
s
io
n
lo
s
s
f
u
n
ctio
n
.
T
h
e
s
o
lu
tio
n
to
t
h
e
o
p
tim
izatio
n
p
r
o
b
le
m
s
s
tated
in
(
4
)
.
̂
[
(
)
]
,
s
u
ch
t
h
at
(
)
.
(
4
)
T
h
e
L
ag
r
an
g
e
f
o
r
m
o
f
t
h
e
o
p
tim
izatio
n
p
r
o
b
lem
i
n
(
4
)
is
s
tat
ed
in
(
5
)
.
̂
[
(
)
(
)
]
(
5
)
W
h
er
e
is
a
r
eg
u
lar
izatio
n
p
ar
a
m
eter
an
d
is
a
p
o
s
itiv
e
r
e
al
n
u
m
b
er
.
C
o
n
s
id
er
th
e
o
p
tim
izatio
n
p
r
o
b
lem
in
(
5
)
.
W
h
en
is
s
et
t
o
ze
r
o
,
th
en
th
e
s
o
l
u
tio
n
s
̂
ar
e
th
e
s
a
m
e
as
t
h
e
s
o
l
u
tio
n
o
f
t
h
e
u
n
co
n
s
tr
ai
n
ed
p
r
o
b
lem
(
f
u
ll
m
o
d
el)
s
tated
in
(
3
)
.
Ho
w
e
v
er
,
f
o
r
h
i
g
h
-
d
i
m
en
s
io
n
al
p
r
o
b
lem
s
(
p
>n
)
,
th
e
ca
n
n
o
t b
e
s
et
to
ze
r
o
b
ec
au
s
e
th
e
s
atu
r
ated
lo
g
is
tic
r
eg
r
ess
io
n
f
it is
u
n
d
e
f
in
ed
[
1
9
]
.
L
ea
s
t
A
b
s
o
lu
te
Sh
r
i
n
k
ag
e
Sel
ec
tio
n
Op
er
ato
r
(
L
ass
o
)
is
a
p
o
p
u
lar
r
eg
u
lar
izatio
n
m
e
th
o
d
in
tr
o
d
u
ce
d
b
y
T
ib
s
h
ir
an
i
[
2
0
]
.
L
ass
o
w
o
r
k
s
b
y
ad
d
in
g
t
h
e
L
1
p
en
alt
y
ter
m
,
d
ef
i
n
ed
as
‖
‖
∑
|
|
,
to
s
h
r
in
k
t
h
e
co
ef
f
icie
n
t
s
o
f
p
ar
ticu
lar
v
ar
ia
b
les
to
be
ze
r
o
.
T
h
er
ef
o
r
e,
L
ass
o
ca
n
b
e
u
s
ed
f
o
r
v
ar
iab
le
s
elec
ti
o
n
.
T
h
e
v
ec
to
r
o
f
esti
m
ated
co
ef
f
icie
n
ts
o
f
t
h
e
L
as
s
o
is
s
tated
as:
̂
[
(
)
‖
‖
]
(
6
)
Sin
ce
t
h
e
co
ef
f
icie
n
ts
o
f
th
e
r
eg
u
lar
izatio
n
p
r
o
b
lem
ar
e
co
n
tr
o
lled
b
y
t
h
e
r
eg
u
lar
izatio
n
p
ar
am
eter
(
)
,
th
en
th
e
o
p
ti
m
u
m
s
h
o
u
ld
b
e
esti
m
ated
.
K
-
f
o
ld
cr
o
s
s
-
v
al
id
atio
n
is
u
s
ed
to
e
s
ti
m
ate
t
h
e
o
p
ti
m
u
m
.
T
h
e
o
p
tim
u
m
λ
i
s
th
e
o
n
e
w
i
th
t
h
e
s
m
al
lest
a
v
er
ag
e
b
in
o
m
ial
d
ev
ian
ce
f
r
o
m
th
e
cr
o
s
s
-
v
alid
atio
n
r
esu
lts
,
i.e
.
̂
*
+
(
)
(
7
)
w
h
er
e
(
)
∑
(
)
(
8
)
2
.
3
.
Dec
is
io
n t
re
e
On
e
o
f
t
h
e
D
ec
i
s
io
n
T
r
ee
alg
o
r
ith
m
s
i
s
t
h
e
C
las
s
i
f
icatio
n
a
n
d
R
e
g
r
es
s
io
n
T
r
ee
(
C
A
R
T
)
t
h
at
ca
n
b
e
u
s
ed
w
h
en
t
h
e
p
r
ed
icto
r
v
ar
iab
les
ar
e
ca
teg
o
r
ical
v
al
u
e
o
r
co
n
tin
u
o
u
s
v
al
u
e
.
T
h
e
C
AR
T
a
lg
o
r
ith
m
u
tili
ze
d
t
h
e
Gin
i i
m
p
u
r
it
y
as t
h
e
s
p
lit
tin
g
c
r
ite
r
ia.
T
h
e
Gin
i i
m
p
u
r
it
y
m
ea
s
u
r
e
at
n
o
d
e
t
w
as d
ef
i
n
ed
as:
(
)
∑
(
)
(
9
)
W
h
er
e
(
)
w
as
th
e
Gi
n
i
i
m
p
u
r
it
y
a
n
d
(
)
w
a
s
t
h
e
p
r
o
p
o
r
tio
n
o
f
cla
s
s
i
n
t
h
e
n
o
d
e
.
T
h
e
Dec
is
io
n
T
r
ee
co
n
s
tr
u
ctio
n
b
eg
in
s
w
i
th
s
p
litt
i
n
g
b
i
n
ar
y
t
h
e
r
o
o
t
n
o
d
e
w
h
ich
co
n
tai
n
s
all
o
b
s
er
v
atio
n
s
o
f
t
h
e
tr
ain
i
n
g
s
et.
T
h
e
cr
iter
ia
u
s
ed
to
d
eter
m
in
e
t
h
e
s
p
litt
in
g
o
f
th
e
r
o
o
t
n
o
d
e
is
th
e
g
o
o
d
n
ess
o
f
s
p
lit,
d
en
o
ted
b
y
(
)
,
an
d
d
ef
in
ed
as:
(
)
(
)
,
(
)
-
,
(
)
-
(
10
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4752
I
n
d
o
n
esia
n
J
E
lec
E
n
g
&
C
o
m
p
Sci,
Vo
l.
21
,
No
.
2
,
Feb
r
u
ar
y
2
0
2
1
:
1
1
51
-
11
59
1154
w
h
er
e:
: a
s
p
lit,
: p
r
o
p
o
r
tio
n
o
f
o
b
s
er
v
atio
n
s
at
n
o
d
e
t
th
at
g
o
in
to
th
e
le
f
t c
h
i
ld
n
o
d
e
: p
r
o
p
o
r
tio
n
o
f
o
b
s
er
v
atio
n
s
at
n
o
d
e
t
th
at
g
o
in
to
th
e
r
ig
h
t c
h
ild
n
o
d
e
,
(
)
:
i
m
p
u
r
it
y
o
f
t
h
e
le
f
t c
h
i
ld
n
o
d
e,
an
d
(
)
: i
m
p
u
r
it
y
o
f
t
h
e
r
ig
h
t c
h
ild
n
o
d
e.
T
o
o
b
tain
th
e
p
o
s
s
ib
le
s
p
lit
-
p
o
in
t
f
o
r
co
n
t
in
u
o
u
s
-
v
al
u
ed
f
e
atu
r
es
,
t
h
e
o
b
s
er
v
atio
n
v
a
lu
e
s
o
f
th
e
p
r
ed
icto
r
,
i.e
.
{
}
,
ar
e
s
o
r
te
d
in
in
cr
ea
s
i
n
g
o
r
d
er
an
d
o
r
d
e
r
ed
o
b
s
er
v
atio
n
v
al
u
es
{
(
)
}
ar
e
o
b
tain
ed
.
T
h
en
,
th
e
m
id
p
o
in
t
o
f
t
w
o
ad
j
ac
en
t
v
al
u
es
is
co
n
s
id
er
ed
as
a
p
o
s
s
ib
le
s
p
lit
-
p
o
i
n
t
[
2
1
]
.
T
h
e
p
o
in
t
th
at
h
ad
m
a
x
i
m
u
m
g
o
o
d
n
ess
o
f
s
p
lit,
i.e
.
{
(
)
}
is
s
elec
ted
as
th
e
s
p
lit
-
p
o
in
t
o
f
th
e
n
o
d
e
[
2
2
]
.
T
h
e
s
ize
o
f
th
e
tr
ee
is
n
o
t
li
m
ited
.
T
h
er
ef
o
r
e,
s
p
litt
i
n
g
ca
n
b
e
ca
r
r
ied
o
u
t
to
o
b
tain
m
o
r
e
lea
f
n
o
d
es.
Ho
w
e
v
er
,
lar
g
e
tr
ee
s
ten
d
to
o
v
er
f
it.
T
h
e
w
a
y
to
s
i
m
p
lify
t
h
e
tr
ee
is
to
p
r
u
n
e.
T
h
e
p
ar
a
m
eter
to
m
ea
s
u
r
e
t
h
e
co
m
p
le
x
it
y
o
f
a
tr
ee
is
ca
lled
th
e
co
m
p
le
x
it
y
p
ar
a
m
ete
r
(
C
P
)
.
T
o
g
et
th
e
C
P
v
alu
e,
f
ir
s
t
,
ca
lc
u
late
r
esu
b
s
tit
u
ti
o
n
est
i
m
a
te
an
d
r
el
ativ
e
er
r
o
r
.
R
es
u
b
s
tit
u
tio
n
e
s
ti
m
ate
i
s
t
h
e
p
r
o
p
o
r
tio
n
o
f
m
is
c
lass
i
f
icatio
n
i
n
th
e
tr
ain
i
n
g
s
e
t
[
7
]
.
Su
p
p
o
s
e
r
ep
r
esen
ts
t
h
e
s
u
b
tr
ee
o
f
th
e
m
ax
i
m
u
m
tr
ee
(
)
.
T
h
en
th
e
r
esu
b
s
tit
u
tio
n
esti
m
ate
o
f
d
en
o
ted
b
y
(
)
is
ca
lc
u
lated
u
s
in
g
t
h
e
f
o
r
m
u
la
in
(
11
).
(
)
∑
(
(
)
)
(
1
1
)
W
h
er
e,
is
an
in
d
icato
r
f
u
n
c
tio
n
th
at
h
a
s
v
alu
e
1
if
(
)
an
d
v
alu
e
0
if
(
)
,
an
d
is
th
e
n
u
m
b
er
o
f
o
b
s
er
v
atio
n
s
.
Nex
t,
th
e
r
elativ
e
er
r
o
r
is
o
b
tain
ed
f
r
o
m
t
h
e
r
atio
o
f
th
e
r
esu
b
s
titu
tio
n
esti
m
ate
o
f
th
e
s
u
b
tr
ee
an
d
th
e
r
es
u
b
s
tit
u
tio
n
e
s
ti
m
ate
at
t
h
e
r
o
o
t
n
o
d
e
o
r
.
T
h
e
r
elativ
e
er
r
o
r
v
alu
e
is
d
ef
i
n
ed
b
y
(
1
2
)
.
(
)
(
)
(
)
(
1
2
)
w
h
er
e
(
)
:
r
elativ
e
er
r
o
r
o
f
th
e
s
u
b
tr
ee
(
)
:
r
esu
b
s
tit
u
tio
n
esti
m
ate
o
f
th
e
s
u
b
tr
ee
(
)
:
r
esu
b
s
tit
u
tio
n
esti
m
ate
o
f
th
e
f
ir
s
t su
b
tr
ee
(
s
u
b
tr
ee
th
at
o
n
l
y
co
n
s
is
ts
o
f
a
r
o
o
t n
o
d
e)
t
he
co
m
p
le
x
it
y
p
ar
a
m
eter
is
d
ef
i
n
ed
b
y
(
13
).
(
)
(
)
(
)
(
)
(
1
3
)
w
h
er
e
:
:
t
h
e
co
m
p
le
x
it
y
p
ar
a
m
eter
s
o
f
s
u
b
tr
ee
(
)
:
th
e
r
elati
v
e
er
r
o
r
of
s
u
b
tr
ee
(
)
:
th
e
n
u
m
b
er
o
f
t
h
e
s
p
li
tti
n
g
o
f
s
u
b
tr
ee
A
v
al
u
e
o
f
C
P
:
0
in
d
icate
s
n
o
p
r
u
n
in
g
w
h
ich
m
ea
n
s
t
h
e
s
u
b
tr
ee
is
a
m
a
x
i
m
u
m
tr
ee
.
A
cc
o
r
d
in
g
to
[
7
]
,
to
o
b
tain
th
e
o
p
ti
m
u
m
Dec
is
io
n
T
r
ee
,
th
e
o
n
e
s
tan
d
ar
d
er
r
o
r
r
u
le
(
1
S
E
r
u
le)
i
s
u
s
ed
.
T
h
e
1
SE
r
u
le
s
elec
ts
a
m
o
d
el
w
i
th
a
r
elati
v
e
er
r
o
r
o
f
th
e
cr
o
s
s
-
v
alid
atio
n
r
esu
lt
(
(
)
)
s
m
aller
o
r
eq
u
al
to
(
)
m
i
n
i
m
u
m
p
l
u
s
o
n
e
s
tan
d
ar
d
d
ev
iatio
n
(
(
)
)
.
T
h
e
r
elativ
e
er
r
o
r
o
f
th
e
cr
o
s
s
-
v
alid
ati
o
n
r
esu
lt
is
ca
lc
u
lated
u
s
i
n
g
th
e
f
o
r
m
u
la
in
(
14
)
.
Me
an
w
h
ile
,
th
e
s
tan
d
ar
d
d
ev
iatio
n
an
d
s
tan
d
ar
d
er
r
o
r
in
th
e
s
u
b
tr
ee
ar
e
ca
lcu
lated
u
s
i
n
g
(
15
)
an
d
(
16
).
(
)
∑
(
(
)
)
(
1
4
)
(
)
√
(
(
(
)
)
)
(
(
)
)
)
(
1
5
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esia
n
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2502
-
4752
K
n
o
w
led
g
e
d
is
co
ve
r
y
fr
o
m
g
en
e
ex
p
r
ess
io
n
d
a
ta
s
et
u
s
in
g
b
a
g
g
in
g
l
a
s
s
o
d
ec
is
io
n
tr
ee
(
Um
u
S
a
’
a
d
a
h
)
1155
(
)
(
)
√
(
1
6
)
t
h
e
o
n
e
s
ta
n
d
ar
d
er
r
o
r
r
u
le
is
p
r
esen
ted
in
t
h
e
(
1
7
)
.
(
)
(
̂
)
(
̂
)
(
1
7
)
w
h
er
e
̂
*
+
(
)
(
1
8
)
2
.
4
.
B
o
o
t
s
t
ra
p
a
g
g
re
g
a
t
ing
(
ba
g
g
ing
)
T
h
e
b
asic
id
ea
o
f
B
ag
g
in
g
is
t
o
u
s
e
b
o
o
ts
tr
ap
r
esam
p
li
n
g
to
g
et
co
m
b
in
ed
p
r
ed
ictio
n
s
.
T
h
e
f
ir
s
t
s
te
p
o
f
B
ag
g
in
g
i
s
b
o
o
t
st
r
ap
r
esa
m
p
lin
g
th
a
t ta
k
es
s
a
m
p
le
s
f
r
o
m
th
e
tr
ai
n
i
n
g
s
et
w
it
h
r
ep
lace
m
en
t to
g
et
t
h
e
n
e
w
tr
ain
i
n
g
s
et
.
F
u
r
th
er
m
o
r
e,
class
if
ica
tio
n
m
o
d
eli
n
g
i
s
p
er
f
o
r
m
ed
in
b
o
o
ts
tr
ap
r
ep
licates.
T
h
e
s
ec
o
n
d
s
tep
is
ag
g
r
e
g
ati
n
g
,
w
h
ic
h
is
co
m
b
i
n
i
n
g
t
h
e
est
i
m
a
tio
n
r
es
u
lt
s
in
to
a
s
in
g
le
est
i
m
ated
v
al
u
e.
T
h
e
ag
g
r
e
g
ati
n
g
p
r
o
ce
s
s
is
d
o
n
e
b
y
m
aj
o
r
it
y
v
o
te.
3.
RE
S
E
ARCH
M
E
T
H
O
D
W
e
im
p
le
m
e
n
ted
th
e
m
et
h
o
d
in
th
e
OV
A
O
v
ar
y
g
en
e
ex
p
r
es
s
io
n
d
ata
av
ailab
le
o
n
h
ttp
://
o
p
en
m
l.o
r
g
.
T
h
e
d
atas
et
co
n
s
is
ts
o
f
1
,
5
4
5
tu
m
o
r
t
is
s
u
e
o
b
s
er
v
atio
n
s
,
in
w
h
ic
h
10,
9
3
5
g
en
es
ar
e
o
b
s
er
v
ed
o
n
ea
c
h
o
f
th
o
s
e
t
u
m
o
r
tis
s
u
e
s
.
Ob
s
er
v
ed
t
u
m
o
r
t
is
s
u
e
s
ar
e
lab
eled
w
it
h
“
O
v
ar
y
”
a
n
d
“
Ot
h
er
”.
T
h
e
“
O
v
ar
y
”
cla
s
s
is
t
h
e
clas
s
o
f
o
v
ar
ian
t
u
m
o
r
ti
s
s
u
e;
w
h
ile
th
e
“
Ot
h
er
”
class
i
s
th
e
cla
s
s
o
f
o
th
er
tu
m
o
r
tis
s
u
e
s
,
in
cl
u
d
in
g
co
lo
n
,
b
r
ea
s
t,
e
n
d
o
m
etr
ial,
k
id
n
e
y
,
l
u
n
g
,
o
m
e
n
ta
l,
p
r
o
s
tate,
an
d
u
ter
u
s
t
u
m
o
r
.
T
h
e
d
ata
an
al
y
s
i
s
s
tep
s
ar
e
d
escr
ib
ed
as f
o
llo
w
s
.
3
.
1
.
Ste
ps
f
o
r
predict
o
r
v
a
ri
a
bles
s
elec
t
io
n u
s
ing
la
s
s
o
a)
Sp
lit
th
e
o
r
ig
i
n
al
d
atase
t
in
to
th
e
tr
ain
in
g
an
d
te
s
ti
n
g
s
et.
W
e
u
s
ed
th
e
r
atio
o
f
8
0
%
f
o
r
th
e
tr
ain
i
n
g
s
et
an
d
2
0
% f
o
r
th
e
test
i
n
g
s
et.
b)
Stan
d
ar
d
ize
th
e
tr
ai
n
in
g
s
et
a
n
d
r
u
n
th
e
L
as
s
o
r
eg
u
lar
izatio
n
w
it
h
1
0
0
iter
atio
n
s
.
c)
Dete
r
m
i
n
e
th
e
o
p
ti
m
u
m
r
eg
u
l
ar
izatio
n
p
ar
a
m
eter
(
)
o
f
th
e
L
as
s
o
u
s
i
n
g
1
0
-
f
o
ld
cr
o
s
s
-
v
al
id
ati
o
n
d)
E
x
tr
ac
t th
e
v
ec
to
r
o
f
co
ef
f
icie
n
ts
(
)
at
th
e
o
p
ti
m
u
m
3
.
2
.
Ste
ps
f
o
r
m
o
deli
ng
a
decisi
o
n t
re
e
us
i
ng
t
he
CAR
T
a
lg
o
rit
h
m
a)
Dete
r
m
i
n
e
all
p
o
s
s
ib
le
s
p
lit
-
p
o
in
t f
o
r
ea
ch
p
r
ed
icto
r
v
ar
iab
le.
b)
C
alcu
late
th
e
g
o
o
d
n
es
s
o
f
s
p
l
it
(
)
u
s
i
n
g
th
e
f
o
r
m
u
la
(
7
)
to
g
et
th
e
b
est
s
p
li
t
-
p
o
in
t.
T
h
e
b
e
s
t
s
p
lit
-
p
o
in
t is
{
(
)
}
.
c)
Use
t
h
e
s
p
lit
-
p
o
in
t
to
s
p
lit
t
h
e
r
o
o
t
n
o
d
e
b
in
ar
y
s
o
th
at
t
h
e
le
f
t
c
h
ild
n
o
d
e
an
d
r
i
g
h
t
c
h
ild
n
o
d
e
ar
e
o
b
tain
ed
.
d)
Sp
lit
th
e
t
w
o
ch
ild
n
o
d
es u
n
til
a
m
ax
i
m
u
m
D
ec
is
io
n
T
r
ee
is
f
o
r
m
ed
.
e)
P
r
u
n
e
th
e
m
a
x
i
m
u
m
tr
ee
b
as
ed
o
n
th
e
r
u
le
s
tated
i
n
t
h
e
(
1
7
)
s
o
th
at
th
e
m
a
x
i
m
u
m
D
e
cisi
o
n
T
r
ee
i
s
o
b
tain
ed
3
.
3
.
Ste
ps
f
o
r
ba
g
g
ing
la
s
s
o
decisi
o
n t
re
e
a)
P
er
f
o
r
m
s
a
m
p
li
n
g
w
it
h
r
ep
lace
m
en
t
ti
m
es i
n
t
h
e
tr
ain
in
g
s
e
t,
w
h
er
e
s
tates th
e
s
ize
o
f
o
b
s
er
v
atio
n
s
i
n
th
e
tr
ain
in
g
s
et
.
b)
R
u
n
t
h
e
L
a
s
s
o
o
n
t
h
e
tr
ain
i
n
g
s
et
to
s
elec
t p
r
ed
icto
r
v
ar
iab
les.
c)
C
o
n
s
tr
u
ct
a
Dec
i
s
io
n
T
r
ee
u
s
in
g
t
h
e
C
A
R
T
alg
o
r
it
h
m
o
n
th
e
tr
ai
n
i
n
g
s
et
t
h
at
h
as
b
ee
n
s
elec
ted
f
o
r
p
r
ed
icto
r
v
ar
iab
les.
d)
R
ep
ea
t step
1
to
s
tep
3
ti
m
es t
o
o
b
tain
Dec
is
io
n
T
r
ee
s
.
e)
C
o
m
b
i
n
e
D
ec
is
io
n
T
r
ee
u
s
i
n
g
m
aj
o
r
ity
v
o
tes.
T
h
e
d
ata
an
al
y
s
i
s
p
r
o
ce
s
s
wa
s
co
n
d
u
cted
u
s
i
n
g
R
.
T
h
e
R
-
p
ac
k
ag
e
s
u
s
ed
w
er
e
g
l
m
n
et
f
o
r
g
e
n
e
s
elec
tio
n
u
s
i
n
g
L
as
s
o
r
eg
u
la
r
izatio
n
,
r
p
ar
t
f
o
r
g
en
er
ati
n
g
th
e
class
i
f
icatio
n
m
o
d
el,
an
d
r
p
ar
t
.
p
lo
t
f
o
r
v
is
u
alizi
n
g
th
e
cla
s
s
i
f
icat
io
n
m
o
d
el.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4752
I
n
d
o
n
esia
n
J
E
lec
E
n
g
&
C
o
m
p
Sci,
Vo
l.
21
,
No
.
2
,
Feb
r
u
ar
y
2
0
2
1
:
1
1
51
-
11
59
1156
4.
RE
SU
L
T
S
A
ND
D
IS
CU
SS
I
O
N
T
h
e
d
ataset
w
a
s
f
ir
s
t
s
p
lit
i
n
t
o
th
e
tr
ain
in
g
s
et
an
d
th
e
test
i
n
g
s
et
w
it
h
a
r
atio
o
f
8
0
%:2
0
%
.
F
o
r
th
e
tr
ain
i
n
g
s
et,
1
,
2
3
6
o
b
s
er
v
atio
n
s
w
er
e
o
b
tain
ed
t
h
at
w
o
u
ld
b
e
u
s
ed
to
co
n
s
tr
u
c
t
t
h
e
m
o
d
el
.
Me
an
w
h
ile,
f
o
r
th
e
test
i
n
g
s
et,
3
0
9
o
b
s
er
v
atio
n
s
w
er
e
o
b
tai
n
ed
th
at
w
o
u
ld
b
e
u
s
ed
f
o
r
ev
al
u
ati
n
g
t
h
e
m
o
d
el.
I
n
t
h
e
tr
ai
n
i
n
g
s
et
,
s
tan
d
ar
d
izatio
n
i
s
p
er
f
o
r
m
ed
,
an
d
th
e
n
t
h
e
L
as
s
o
r
eg
u
lar
izat
io
n
is
r
u
n
to
g
et
.
B
ec
au
s
e
d
e
p
en
d
s
o
n
th
e
r
e
g
u
lar
iza
tio
n
p
ar
a
m
ete
r
(
)
,
f
ir
s
t,
t
h
e
o
p
ti
m
u
m
is
esti
m
ated
u
s
i
n
g
1
0
-
f
o
ld
cr
o
s
s
-
v
alid
atio
n
.
Fi
g
u
r
e
1
s
h
o
w
s
t
h
e
r
es
u
lt o
f
1
0
-
f
o
ld
cr
o
s
s
-
v
al
id
atio
n
to
o
b
tain
th
e
o
p
ti
m
u
m
o
f
th
e
L
as
s
o
.
Fig
u
r
e
1
.
E
s
ti
m
at
in
g
t
h
e
o
p
ti
m
u
m
r
e
g
u
lar
izatio
n
p
ar
a
m
eter
T
h
e
lef
t
-
h
an
d
p
ict
u
r
e
in
Fi
g
u
r
e
1
s
h
o
w
s
th
e
r
es
u
lt
o
f
cr
o
s
s
-
v
alid
atio
n
a
n
d
th
e
r
ig
h
t
-
h
an
d
zo
o
m
s
i
n
th
e
f
o
cu
s
ed
ar
ea
.
I
n
b
o
th
p
ict
u
r
es,
th
e
v
er
tical
li
n
e
o
n
t
h
e
le
f
t
s
h
o
w
s
t
h
e
m
in
i
m
u
m
a
v
er
ag
e
b
in
o
m
ial
d
e
v
ia
n
ce
,
w
h
ic
h
is
0
.
0
1
6
9
1
7
.
Me
an
w
h
i
le
,
th
e
o
n
e
o
n
t
h
e
r
i
g
h
t
s
h
o
w
s
th
e
o
n
e
s
ta
n
d
ar
d
er
r
o
r
o
f
th
e
m
i
n
i
m
u
m
,
w
h
ic
h
is
.
On
e
ca
n
u
s
e
th
e
w
i
th
i
n
o
n
e
s
tan
d
ar
d
er
r
o
r
o
f
th
e
m
i
n
i
m
u
m
.
Fo
r
th
is
ca
s
e,
w
e
u
s
e
as
th
e
o
p
tim
u
m
r
e
g
u
lar
iz
atio
n
p
ar
a
m
eter
.
T
h
e
n
u
m
b
er
alo
n
g
t
h
e
to
p
o
f
th
e
p
ictu
r
e
s
t
ate
th
e
n
u
m
b
er
o
f
n
o
n
ze
r
o
co
ef
f
icie
n
t
s
.
T
h
er
ef
o
r
e,
th
e
o
p
ti
m
u
m
th
at
w
e
u
s
e
p
r
o
d
u
ce
d
7
1
p
r
ed
ict
o
r
v
ar
iab
les
w
it
h
a
n
o
n
ze
r
o
co
ef
f
icie
n
t.
Af
ter
th
e
o
p
ti
m
u
m
is
o
b
tain
ed
,
th
e
co
ef
f
icien
t
̂
o
f
th
e
L
a
s
s
o
is
t
h
en
e
x
tr
ac
ted
.
4
.
1
.
Dec
is
io
n t
re
e
m
o
delin
g
T
h
e
s
tep
s
to
b
u
ild
a
D
ec
i
s
io
n
T
r
ee
a
r
e
to
d
eter
m
i
n
e
t
h
e
Gi
n
i
in
d
ex
f
r
o
m
th
e
tr
ain
i
n
g
s
et
,
d
eter
m
in
e
th
e
n
o
d
e
s
p
lit
-
p
o
in
t,
an
d
d
eter
m
in
e
t
h
e
g
o
o
d
n
ess
o
f
s
p
lit
.
Sp
litt
in
g
t
h
e
n
o
d
es
is
d
o
n
e
u
n
til
g
e
tti
n
g
t
h
e
m
ax
i
m
u
m
tr
ee
an
d
th
e
n
p
r
u
n
e
it
to
g
et
a
s
i
m
p
ler
tr
ee
.
Su
p
p
o
s
e
th
at
t
h
e
m
a
x
i
m
u
m
tr
ee
is
d
en
o
ted
b
y
an
d
th
e
s
u
b
tr
ee
o
f
th
e
m
ax
i
m
u
m
t
r
ee
is
d
en
o
ted
b
y
.
T
o
o
b
tain
t
h
e
o
p
ti
m
u
m
tr
ee
,
cr
o
s
s
-
v
alid
a
tio
n
is
u
s
ed
.
T
h
e
o
p
tim
u
m
tr
ee
s
ar
e
s
u
b
tr
ee
s
t
h
at
h
a
v
e
(
)
(
̂
)
(
̂
)
as
s
tated
in
(
1
7
)
.
Fig
u
r
e
2
p
r
esen
ts
th
e
s
u
b
tr
ee
s
o
f
t
h
e
m
ax
i
m
u
m
tr
ee
o
f
th
e
OV
A
_
O
v
ar
y
d
ata
s
et
w
it
h
s
elec
ted
g
e
n
e
s
f
r
o
m
L
a
s
s
o
.
T
h
e
r
esu
lt o
f
cr
o
s
s
-
v
alid
atio
n
i
s
d
is
p
la
y
ed
b
y
t
h
e
p
lo
t o
f
th
e
co
m
p
lex
it
y
p
ar
a
m
e
ter
(
C
P
)
p
r
esen
ted
in
Fi
g
u
r
e
3
.
(
a)
(
b
)
(
c)
(
d
)
Fig
u
r
e
2
.
S
u
b
tr
ee
s
o
f
t
h
e
OV
A_
Ov
ar
y
d
ataset
w
it
h
t
h
e
s
elec
t
ed
g
en
es
f
r
o
m
l
as
s
o
(
a)
,
(
b
)
,
(
c)
,
(
d
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esia
n
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2502
-
4752
K
n
o
w
led
g
e
d
is
co
ve
r
y
fr
o
m
g
en
e
ex
p
r
ess
io
n
d
a
ta
s
et
u
s
in
g
b
a
g
g
in
g
l
a
s
s
o
d
ec
is
io
n
tr
ee
(
Um
u
S
a
’
a
d
a
h
)
1157
In
Fi
g
u
r
e
3
,
th
e
d
o
tted
h
o
r
izo
n
tal
l
in
e
co
r
r
esp
o
n
d
s
to
t
h
e
m
i
n
i
m
u
m
cr
o
s
s
-
v
alid
atio
n
er
r
o
r
p
lu
s
o
n
e
s
tan
d
ar
d
er
r
o
r
o
f
th
e
m
in
i
m
u
m
,
(
̂
)
(
̂
)
,
w
h
ich
is
0
.
5
9
5
.
T
h
e
o
p
ti
m
u
m
tr
ee
is
in
d
icate
d
b
y
a
p
lo
t
b
elo
w
th
e
h
o
r
izo
n
tal
li
n
e
.
T
h
e
n
u
m
b
er
alo
n
g
t
h
e
to
p
o
f
th
e
p
lo
t
s
tates
th
e
n
u
m
b
er
o
f
leaf
n
o
d
es
p
r
o
d
u
ce
d
b
y
th
e
s
u
b
tr
ee
.
R
ef
e
r
r
i
n
g
to
Fi
g
u
r
e
3
,
th
e
o
p
ti
m
u
m
s
u
b
tr
ee
s
ar
e
th
e
s
u
b
tr
ee
th
a
t
p
r
o
d
u
c
e
s
2
,
5
,
an
d
6
lea
f
n
o
d
es
b
ec
au
s
e
th
e
p
lo
t
is
lo
ca
ted
b
el
o
w
th
e
h
o
r
izo
n
ta
l
lin
e
.
Fu
r
t
h
e
r
m
o
r
e,
f
r
o
m
t
h
e
th
r
ee
alter
n
at
iv
e
o
p
ti
m
u
m
tr
ee
s
,
w
e
o
n
l
y
u
s
e
o
n
e
to
b
e
th
e
o
p
ti
m
u
m
.
W
h
e
n
th
e
s
u
b
tr
ee
w
i
th
2
leaf
n
o
d
es
is
u
s
ed
,
th
e
m
o
d
el
w
o
u
ld
p
er
f
o
r
m
p
o
o
r
ly
.
An
d
w
h
e
n
th
e
s
u
b
tr
ee
w
it
h
t
h
e
6
lea
f
n
o
d
es
i
s
u
s
ed
,
th
e
m
o
d
el
w
o
u
ld
te
n
d
to
o
v
er
f
it.
T
h
er
ef
o
r
e,
w
e
u
s
e
t
h
e
s
u
b
tr
ee
w
i
th
5
leaf
n
o
d
es to
b
e
th
e
o
p
ti
m
u
m
tr
ee
as s
h
o
w
n
i
n
Fi
g
u
r
e
4
.
Fig
u
r
e
3
.
T
h
e
p
lo
t
o
f
th
e
co
m
p
lex
it
y
p
ar
a
m
eter
I
n
Fi
g
u
r
e
4
,
t
h
e
lab
el
o
n
t
h
e
n
o
d
e
s
h
o
w
s
th
e
clas
s
th
a
t
h
a
s
th
e
lar
g
est
p
r
o
p
o
r
tio
n
o
f
t
h
at
n
o
d
e.
T
h
e
co
m
p
atib
il
it
y
o
f
t
h
e
m
o
d
el
o
b
tain
ed
f
r
o
m
t
h
e
o
p
ti
m
u
m
D
ec
is
io
n
T
r
ee
is
m
ea
s
u
r
ed
u
s
i
n
g
ac
cu
r
ac
y
.
T
h
e
ac
cu
r
ac
y
o
f
th
e
o
p
ti
m
a
l
D
ec
is
io
n
T
r
ee
is
9
5
.
3
1
%
f
o
r
th
e
tr
ain
in
g
s
et
an
d
8
9
.
3
4
%
f
o
r
t
h
e
test
i
n
g
s
e
t.
T
h
is
ac
cu
r
ac
y
v
alu
e
i
s
h
i
g
h
ev
e
n
t
h
o
u
g
h
th
e
m
o
d
el
o
n
l
y
i
n
v
o
lv
e
s
f
o
u
r
p
r
ed
icto
r
s
.
B
e
s
i
d
e
s
e
v
a
l
u
a
t
i
n
g
t
h
e
c
l
a
s
s
i
f
i
c
a
t
i
o
n
p
e
r
f
o
r
m
a
n
c
e
,
w
e
a
l
s
o
i
n
t
e
r
p
r
e
t
t
h
e
m
o
d
e
l
.
B
a
s
e
d
o
n
t
h
e
m
o
d
e
l
i
n
F
i
g
u
r
e
4
,
t
o
d
i
s
t
i
n
g
u
i
s
h
b
e
tw
e
e
n
o
v
a
r
i
a
n
t
u
m
o
r
t
i
s
s
u
e
a
n
d
o
t
h
e
r
t
u
m
o
r
s
,
f
i
r
s
t
,
t
h
e
e
x
p
r
e
s
s
i
o
n
o
f
t
h
e
S
T
A
R
g
e
n
e
i
s
s
e
e
n
.
I
f
t
h
e
S
T
A
R
e
x
p
r
e
s
s
i
o
n
s
h
o
w
s
a
v
a
l
u
e
o
f
m
o
r
e
t
h
a
n
o
r
e
q
u
a
l
t
o
2
2
6
,
t
h
e
n
t
h
e
t
u
m
o
r
t
i
s
s
u
e
i
s
p
r
e
d
i
c
t
e
d
t
o
b
e
o
v
a
r
i
a
n
t
u
m
o
r
t
i
s
s
u
e
w
i
t
h
a
p
r
o
b
a
b
i
l
i
t
y
o
f
8
5
%
.
M
e
a
n
w
h
i
l
e
,
i
f
t
h
e
S
T
A
R
e
x
p
r
e
s
s
i
o
n
v
a
l
u
e
i
s
l
e
s
s
t
h
a
n
2
2
6
,
t
h
e
n
t
h
e
W
T
1
e
x
p
r
e
s
s
i
o
n
i
s
s
e
e
n
.
I
f
W
T
1
e
x
p
r
e
s
s
i
o
n
i
s
l
e
s
s
t
h
a
n
7
,
0
6
7
,
i
t
c
a
n
b
e
e
s
t
i
m
a
t
e
d
t
h
a
t
t
h
e
t
u
m
o
r
t
i
s
s
u
e
i
s
a
n
o
t
h
e
r
t
u
m
o
r
t
i
s
s
u
e
w
i
t
h
a
p
r
o
b
a
b
i
l
i
t
y
o
f
9
7
%
.
I
f
W
T
1
s
h
o
w
s
a
n
e
x
p
r
e
s
s
i
o
n
v
a
l
u
e
g
r
e
a
t
e
r
t
h
a
n
o
r
e
q
u
a
l
t
o
7
,
0
6
7
,
t
h
e
n
t
h
e
C
1
9
o
r
f
5
3
g
e
n
e
e
x
p
r
e
s
s
i
o
n
i
s
t
h
e
n
s
e
e
n
.
I
f
t
h
e
e
x
p
r
e
s
s
i
o
n
o
f
C
1
9
o
r
f
5
3
i
s
m
o
r
e
t
h
a
n
o
r
e
q
u
a
l
t
o
5
,
3
2
3
t
h
e
n
t
h
e
t
u
m
o
r
t
i
s
s
u
e
m
u
s
t
b
e
o
v
a
r
i
a
n
,
b
u
t
i
f
t
h
e
C
1
9
o
r
f
5
3
e
x
p
r
e
s
s
i
o
n
i
s
l
e
s
s
t
h
a
n
5
,
3
2
3
t
h
e
n
t
h
e
e
x
p
r
e
s
s
i
o
n
o
f
t
h
e
M
T
X
2
g
e
n
e
is
a
l
s
o
s
e
e
n
.
I
f
t
h
e
M
T
X
2
g
e
n
e
h
a
s
m
o
r
e
t
h
a
n
o
r
e
q
u
a
l
e
x
p
r
e
s
s
i
o
n
w
i
t
h
2
,
8
6
7
,
t
h
e
n
i
t
i
s
p
r
e
d
i
c
t
e
d
a
s
o
v
a
r
i
a
n
t
u
m
o
r
t
i
s
s
u
e
,
b
u
t
i
f
i
t
i
s
l
e
s
s
t
h
a
n
2
,
8
6
7
,
t
h
e
n
i
t
i
s
p
r
e
d
i
c
t
e
d
a
s
a
n
o
t
h
e
r
t
u
m
o
r
t
i
s
s
u
e
.
T
h
e
D
e
c
i
s
i
o
n
T
r
e
e
m
o
d
e
l
o
b
t
a
i
n
e
d
i
s
t
h
e
o
r
e
t
i
c
a
l
l
y
c
o
r
r
e
c
t
.
A
c
c
o
r
d
i
n
g
t
o
[
2
3
]
,
S
T
A
R
p
l
a
y
s
a
n
i
m
p
o
r
t
a
n
t
r
o
l
e
i
n
t
h
e
p
r
o
d
u
c
t
i
o
n
o
f
s
t
e
r
o
i
d
h
o
r
m
o
n
e
s
.
E
s
t
r
o
g
e
n
i
s
a
s
t
e
r
o
i
d
h
o
r
m
o
n
e
t
h
a
t
f
u
n
c
t
i
o
n
s
a
s
a
f
e
m
a
l
e
s
e
x
h
o
r
m
o
n
e
.
E
x
c
e
s
s
i
v
e
e
s
t
r
o
g
e
n
c
a
n
c
a
u
s
e
o
v
a
r
i
a
n
c
a
n
c
e
r
.
T
h
e
t
h
e
o
r
y
s
u
p
p
o
r
t
s
t
h
e
r
e
s
u
l
t
s
o
f
t
h
i
s
s
t
u
d
y
w
h
e
r
e
o
v
a
r
i
a
n
t
u
m
o
r
p
a
t
i
e
n
t
s
h
a
v
e
h
i
g
h
e
r
S
T
A
R
e
x
p
r
e
s
s
i
o
n
t
h
a
n
o
t
h
e
r
t
u
m
o
r
p
a
t
i
e
n
t
s
.
W
T
1
e
x
p
r
e
s
s
i
o
n
i
n
o
v
a
r
i
a
n
t
u
m
o
r
s
h
a
s
b
e
en
r
e
v
i
ew
e
d
b
y
[
2
4
,
2
5
]
.
W
T
1
h
a
s
a
h
i
g
h
e
x
p
r
e
s
s
i
o
n
i
n
e
p
i
t
h
e
l
i
a
l
o
v
a
r
i
a
n
t
u
m
o
r
p
a
t
i
e
n
t
s
(
i
.
e
.
t
u
m
o
r
s
i
n
t
h
e
o
v
a
r
y
c
o
v
e
r
i
n
g
m
e
m
b
r
a
n
e
)
a
n
d
t
h
e
h
i
g
h
e
r
e
x
p
r
e
s
s
i
o
n
o
f
W
T
1
i
n
d
i
c
a
t
e
s
a
h
i
g
h
t
u
m
o
r
s
t
a
g
e
[
2
4
]
.
M
e
a
n
w
h
i
l
e
,
t
h
e
r
e
s
u
l
t
s
o
f
r
e
s
e
a
r
c
h
[
2
5
]
st
a
t
e
d
t
h
a
t
e
x
c
e
s
s
i
v
e
W
T
1
e
x
p
r
e
s
s
i
o
n
c
a
n
m
a
k
e
c
a
n
c
e
r
c
e
l
l
s
m
o
r
e
a
g
g
r
e
s
s
i
v
e
i
n
o
v
a
r
i
a
n
c
a
n
c
e
r
.
M
e
a
n
w
h
i
l
e
,
t
h
e
e
x
p
r
e
s
s
i
o
n
o
f
C
1
9
o
r
f
5
3
a
n
d
M
T
X
2
g
e
n
e
s
i
n
o
v
a
r
i
a
n
t
u
m
o
r
s
h
a
s
n
o
t
b
e
e
n
o
r
h
a
s
n
o
t
b
e
e
n
s
t
u
d
i
e
d
b
y
r
e
s
e
a
r
c
h
e
r
s
.
H
o
w
e
v
e
r
,
C
1
9
o
r
f
5
3
a
n
d
M
T
X
2
a
r
e
n
o
t
i
m
p
o
r
t
a
n
t
m
a
r
k
e
r
s
b
a
s
e
d
o
n
t
h
e
m
o
d
e
l
.
Fig
u
r
e
4
.
T
h
e
o
p
tim
u
m
tr
ee
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4752
I
n
d
o
n
esia
n
J
E
lec
E
n
g
&
C
o
m
p
Sci,
Vo
l.
21
,
No
.
2
,
Feb
r
u
ar
y
2
0
2
1
:
1
1
51
-
11
59
1158
4
.
2
.
B
a
g
g
ing
l
a
s
s
o
decisi
o
n t
re
e
W
e
cr
ea
ted
b
o
o
ts
tr
ap
r
ep
li
ca
tes
as
d
o
n
e
b
y
[
1
0
]
an
d
in
o
r
d
er
n
o
t
t
o
r
e
q
u
ir
e
a
lo
n
g
co
m
p
u
tatio
n
ti
m
e.
I
n
th
e
5
0
b
o
o
ts
tr
ap
r
e
p
licates,
th
e
L
ass
o
w
as
u
s
ed
to
s
elec
t
t
h
e
p
r
ed
icto
r
v
ar
iab
les,
an
d
th
en
th
e
D
ec
is
io
n
T
r
ee
m
o
d
el
w
a
s
co
n
s
tr
u
cted
u
s
i
n
g
t
h
e
C
AR
T
alg
o
r
ith
m
to
o
b
tain
5
0
s
i
n
g
le
m
o
d
el
s
.
O
f
t
h
e
5
0
s
in
g
le
m
o
d
els
f
r
o
m
t
h
e
b
o
o
ts
tr
ap
r
ep
licates,
a
v
o
te
w
a
s
ca
r
r
ied
o
u
t
to
c
o
m
b
i
n
e
th
e
m
o
d
els
in
to
o
n
e
f
in
a
l
m
o
d
el.
I
n
t
h
e
tr
ai
n
in
g
s
et,
t
h
e
B
ag
g
i
n
g
m
o
d
el
g
av
e
an
ac
c
u
r
ac
y
o
f
9
5
.
2
3
%.
Me
an
w
h
ile,
i
n
th
e
te
s
ti
n
g
s
et,
t
h
e
B
ag
g
i
n
g
m
o
d
el
g
a
v
e
a
n
ac
cu
r
ac
y
o
f
9
0
.
2
9
%.
A
s
u
m
m
ar
y
o
f
t
h
e
ac
cu
r
ac
y
v
alu
e
s
f
r
o
m
t
h
e
s
i
n
g
le
m
o
d
el
an
d
th
e
B
ag
g
in
g
m
o
d
el
i
s
p
r
esen
te
d
in
T
ab
le
1.
T
ab
le
1
.
T
h
e
a
cc
u
r
ac
y
o
f
th
e
s
in
g
le
an
d
th
e
b
a
g
g
i
n
g
M
o
d
e
l
A
c
c
u
r
a
c
y
o
f
t
h
e
t
r
a
i
n
i
n
g
se
t
A
c
c
u
r
a
c
y
o
f
t
h
e
t
e
st
i
n
g
se
t
S
i
n
g
l
e
L
a
sso
D
e
c
i
si
o
n
T
r
e
e
9
5
.
3
1
%
8
9
.
3
2
%
.
B
a
g
g
i
n
g
L
a
sso
D
e
c
i
si
o
n
T
r
e
e
9
5
.
2
3
%
9
0
.
2
9
%
C
o
m
p
ar
in
g
th
e
ac
c
u
r
ac
y
o
f
t
h
e
s
in
g
le
m
o
d
els
an
d
t
h
e
B
ag
g
i
n
g
m
o
d
el
h
as
b
ee
n
co
n
d
u
cted
b
y
[
2
6
]
.
I
n
th
at
s
tu
d
y
,
a
s
i
n
g
le
D
ec
is
io
n
T
r
ee
m
o
d
el
w
a
s
co
m
p
ar
ed
w
i
th
a
B
ag
g
i
n
g
D
ec
i
s
io
n
T
r
ee
w
it
h
t
h
r
ee
s
p
litt
in
g
cr
iter
ia,
n
a
m
el
y
t
h
e
g
ain
r
at
i
o
,
in
f
o
r
m
a
tio
n
g
ai
n
,
an
d
Gi
n
i
in
d
ex
.
T
h
e
r
esu
lt
is
o
b
tai
n
e
d
th
at
B
ag
g
in
g
ca
n
in
cr
ea
s
e
ac
cu
r
ac
y
b
y
2
.
9
6
%,
1
.
6
5
%,
an
d
2
.
1
8
%.
I
n
th
is
s
t
u
d
y
,
an
i
n
cr
ea
s
e
i
n
ac
cu
r
ac
y
is
o
n
l
y
1
%
o
f
th
e
s
in
g
le
L
as
s
o
Dec
is
io
n
T
r
ee
m
o
d
el.
T
h
is
m
ea
n
s
t
h
at
t
h
e
L
as
s
o
Dec
is
io
n
T
r
ee
class
if
i
er
is
s
tab
le
o
n
t
h
e
OV
A
_
Ov
ar
y
d
ata
s
et.
5.
CO
NCLU
SI
O
N
T
h
e
B
ag
g
in
g
L
as
s
o
Dec
is
io
n
T
r
ee
w
as
p
er
f
o
r
m
ed
to
ex
am
i
n
e
th
e
s
tab
ilit
y
o
f
th
e
L
as
s
o
Dec
is
io
n
T
r
ee
class
if
ier
.
Fro
m
th
e
O
V
A
Ov
ar
y
d
ataset,
t
h
e
L
as
s
o
Dec
is
io
n
T
r
ee
m
o
d
el
w
as c
o
n
s
tr
u
ct
ed
w
it
h
f
o
u
r
s
p
lit
s
b
u
t
w
a
s
ab
le
to
p
r
o
d
u
ce
an
a
cc
u
r
ac
y
o
f
8
9
.
3
%
f
o
r
th
e
test
in
g
s
et.
Me
an
w
h
ile,
t
h
e
B
ag
g
in
g
m
o
d
el
g
a
v
e
a
n
ac
cu
r
ac
y
o
f
9
0
.
2
9
%.
T
h
e
s
li
g
h
tl
y
i
n
cr
ea
s
i
n
g
ac
cu
r
ac
y
s
h
o
ws
t
h
at
t
h
e
L
a
s
s
o
Dec
i
s
io
n
T
r
ee
class
if
ier
is
s
tab
le.
T
h
e
m
o
d
el
o
b
tain
ed
also
g
av
e
a
r
esu
lt
t
h
at
t
h
eo
r
etica
ll
y
co
r
r
ec
t,
b
eg
in
n
in
g
t
h
at
t
h
e
s
p
lit
-
p
o
in
t
o
f
ST
AR
w
a
s
s
elec
ted
as
a
s
p
lit
-
p
o
in
t
o
f
t
h
e
r
o
o
t
n
o
d
e.
S
T
A
R
w
as
th
e
i
m
p
o
r
ta
n
t
g
e
n
e
in
estro
g
e
n
h
o
r
m
o
n
e
p
r
o
d
u
ctio
n
.
T
h
er
ef
o
r
e,
r
esear
ch
er
s
o
r
p
r
ac
titi
o
n
er
s
co
n
ce
r
n
in
g
o
n
co
g
e
n
o
m
ic
m
a
y
u
s
e
t
h
e
L
a
s
s
o
Dec
i
s
i
o
n
T
r
ee
m
et
h
o
d
to
s
tu
d
y
t
h
e
d
i
f
f
er
en
ce
in
g
en
e
ch
ar
ac
ter
is
tic
s
b
et
w
ee
n
n
o
r
m
al
an
d
d
i
s
ea
s
e
co
n
d
itio
n
s
a
n
d
u
s
e
B
ag
g
i
n
g
L
as
s
o
Dec
is
io
n
T
r
ee
to
in
cr
ea
s
e
th
e
ac
cu
r
ac
y
.
ACK
NO
WL
E
D
G
E
M
E
NT
S
W
e
w
o
u
ld
li
k
e
to
ex
p
r
ess
o
u
r
d
ee
p
g
r
atitu
d
e
to
P
r
o
f
.
W
id
o
d
o
,
S.Si.,
M.
Si.,
P
h
.
D.
Me
d
.
Sc.
,
th
e
h
ea
d
o
f
L
ab
o
r
ato
r
iu
m
Sen
tr
a
l
I
l
m
u
Ha
y
ati
(
L
SIH
)
Un
iv
er
s
itas
B
r
a
w
ij
a
y
a
,
f
o
r
g
iv
i
n
g
u
s
k
n
o
w
le
d
g
e
ab
o
u
t g
en
e
s
.
RE
F
E
R
E
NC
E
S
[1
]
H.
Om
a
ra
,
M
.
L
a
z
a
a
r,
a
n
d
Y.
T
a
b
ii
,
“
Eff
e
c
t
o
f
fe
a
tu
re
se
lec
t
io
n
o
n
g
e
n
e
e
x
p
re
ss
io
n
d
a
tas
e
ts
c
las
si
f
ica
ti
o
n
a
c
c
u
ra
c
y
,
”
In
t.
J
.
El
e
c
tr.
Co
mp
u
t.
En
g
.
,
v
o
l
.
8
,
n
o
.
5
,
p
p
.
3
1
9
4
-
3
2
0
3
,
2
0
1
8
.
[2
]
A
.
As
sa
wa
m
a
k
in
,
S
.
P
r
u
e
k
sa
a
ro
o
n
,
S
.
Ku
law
o
n
g
a
n
u
n
c
h
a
i,
P
.
J.
S
h
a
w
,
V
.
V
a
ra
v
it
h
y
a
,
T
.
Ru
a
n
g
ra
ji
t
p
a
k
o
rn
,
a
n
d
S
.
T
o
n
g
si
m
a
,
“
Bio
m
a
r
k
e
r
s
e
lec
ti
o
n
a
n
d
c
las
sif
ic
a
ti
o
n
o
f
„“
-
o
m
ic
s
”
‟
d
a
ta
u
sin
g
a
t
w
o
-
ste
p
b
a
y
e
s
c
las
si
f
ica
ti
o
n
f
ra
m
e
w
o
rk
,
”
Bi
o
me
d
Res
.
In
t.
,
2
0
1
3
.
[3
]
C.
Ka
n
g
,
Y.
Hu
o
,
L
.
X
i
n
,
B.
T
ian
,
a
n
d
B.
Yu
,
“
F
e
a
tu
re
se
lec
ti
o
n
a
n
d
t
u
m
o
r
c
las
sif
ic
a
ti
o
n
f
o
r
m
icro
a
rra
y
d
a
ta
u
sin
g
re
lax
e
d
las
so
a
n
d
g
e
n
e
ra
li
z
e
d
m
u
lt
i
-
c
las
s su
p
p
o
rt
v
e
c
to
r
m
a
c
h
in
e
,
”
J
.
T
h
e
o
r.
Bi
o
l.
,
v
o
l
.
4
6
3
,
p
p
.
7
7
-
9
1
,
2
0
1
9
.
[4
]
L
.
Ga
o
,
M
.
Ye
,
X
.
L
u
,
a
n
d
D.
H
u
a
n
g
,
“
Hy
b
rid
m
e
th
o
d
b
a
se
d
o
n
i
n
f
o
rm
a
ti
o
n
g
a
in
a
n
d
su
p
p
o
r
t
v
e
c
to
r
m
a
c
h
in
e
f
o
r
g
e
n
e
se
lec
ti
o
n
in
c
a
n
c
e
r
c
las
sif
ica
ti
o
n
,
”
Ge
n
o
mic
s.
Pr
o
teo
mic
s B
i
o
i
n
fo
rm
a
t
ics
,
v
o
l.
1
5
,
n
o
.
6
,
p
p
.
3
8
9
-
3
9
5
,
2
0
1
7
.
[5
]
A
.
B
a
c
k
h
a
u
s
a
n
d
U.
S
e
i
ff
e
rt,
“
Ne
u
ro
c
o
m
p
u
ti
n
g
Clas
sif
ica
ti
o
n
in
h
ig
h
-
d
im
e
n
sio
n
a
l
sp
e
c
tral
d
a
ta :
a
c
c
u
ra
c
y
v
s.
in
terp
re
tab
il
it
y
v
s .
m
o
d
e
l
siz
e
,
”
Ne
u
ro
c
o
mp
u
ti
n
g
,
v
o
l
.
1
3
1
,
p
p
.
1
5
-
2
2
,
2
0
1
4
.
[6
]
A
.
Bib
a
l
a
n
d
B.
F
ré
n
a
y
,
“
In
terp
re
tab
il
it
y
o
f
m
a
c
h
in
e
lea
rn
in
g
m
o
d
e
ls
a
n
d
re
p
re
se
n
tati
o
n
s :
a
n
i
n
tro
d
u
c
ti
o
n
,
”
in
Eu
ro
p
e
a
n
S
y
m
p
o
siu
m
o
n
Arti
f
icia
l
Ne
u
ra
l
Ne
two
rk
s,
Co
mp
u
ta
ti
o
n
a
l
In
telli
g
e
n
c
e
a
n
d
M
a
c
h
i
n
e
L
e
a
rn
in
g
,
p
p
.
7
7
-
82
,
2
0
1
6
.
[7
]
L
.
Bre
i
m
a
n
,
J.
H.
F
ried
m
a
n
,
R.
A
.
Olsh
e
n
,
a
n
d
C.
J.
S
to
n
e
,
Cla
ss
if
ica
ti
o
n
a
n
d
Reg
re
ss
io
n
T
re
e
s
.
Ch
a
p
m
a
n
a
n
d
Ha
ll
,
1
9
8
4
.
[8
]
A
.
A
n
d
rz
e
jak
,
F
.
L
a
n
g
n
e
r,
a
n
d
S
.
Zab
a
la,
“
In
terp
re
tab
le
m
o
d
e
ls
f
ro
m
d
istri
b
u
ted
d
a
ta
v
ia
m
e
r
g
in
g
o
f
d
e
c
isio
n
tree
s
,
”
in
2
0
1
3
IEE
E
S
y
mp
o
siu
m
o
n
C
o
mp
u
ta
t
io
n
a
l
I
n
telli
g
e
n
c
e
a
n
d
Da
t
a
M
in
i
n
g
(
CIDM
)
,
p
p
.
1
-
9
,
2
0
1
3
.
[9
]
M
.
Y.
Ro
c
h
a
y
a
n
i,
U.
S
a
‟a
d
a
h
,
a
n
d
A
.
B.
A
stu
ti
,
“
Tw
o
-
sta
g
e
g
e
n
e
s
e
lec
ti
o
n
a
n
d
c
las
sif
ic
a
ti
o
n
f
o
r
a
h
ig
h
-
d
im
e
n
sio
n
a
l
m
icro
a
rra
y
d
a
ta
,
”
J
.
On
li
n
e
I
n
fo
rm
.
,
v
o
l.
5
,
n
o
.
1
,
p
p
.
9
-
1
8
,
2
0
2
0
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esia
n
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2502
-
4752
K
n
o
w
led
g
e
d
is
co
ve
r
y
fr
o
m
g
en
e
ex
p
r
ess
io
n
d
a
ta
s
et
u
s
in
g
b
a
g
g
in
g
l
a
s
s
o
d
ec
is
io
n
tr
ee
(
Um
u
S
a
’
a
d
a
h
)
1159
[1
0
]
L
.
Bre
i
m
a
n
,
“
Ba
g
g
in
g
P
re
d
ict
o
rs,
”
M
a
c
h
.
L
e
a
rn
.
,
v
o
l
.
2
4
,
n
o
.
2
,
p
p
.
1
2
3
-
1
4
0
,
1
9
9
6
.
[1
1
]
H.
R.
S
a
a
d
,
“
In
d
u
strial
e
n
g
in
e
e
r
in
g
&
m
a
n
a
g
e
m
e
n
t
u
se
b
a
g
g
in
g
a
lg
o
rit
h
m
to
im
p
ro
v
e
p
re
d
ictio
n
a
c
c
u
ra
c
y
f
o
r
e
v
a
lu
a
ti
o
n
o
f
w
o
rk
e
r
p
e
r
f
o
rm
a
n
c
e
s
a
t
a
p
ro
d
u
c
ti
o
n
c
o
m
p
a
n
y
,
”
In
d
En
g
M
a
n
a
g
.
,
v
o
l
.
7
,
n
o
.
2
5
7
,
p
p
.
2
1
6
9
-
0
3
1
6
,
2
0
1
8
.
[1
2
]
P
.
Kim
a
n
d
K.
L
i
m
,
“
V
e
h
icle
t
y
p
e
c
las
si
f
ica
ti
o
n
u
sin
g
b
a
g
g
in
g
a
n
d
c
o
n
v
o
lu
ti
o
n
a
l
n
e
u
ra
l
n
e
tw
o
rk
o
n
m
u
lt
i
v
ie
w
su
rv
e
il
lan
c
e
ima
g
e
,
”
in
IEE
E
C
o
n
fer
e
n
c
e
o
n
Co
mp
u
ter
Vi
si
o
n
a
n
d
Pa
t
ter
n
Rec
o
g
n
i
ti
o
n
W
o
rk
sh
o
p
s
,
p
p
.
41
-
46
,
2
0
1
7
.
[1
3
]
A
.
Ra
m
a
s
wa
m
y
r
e
d
d
y
,
S
.
S
h
iv
a
,
K.
V
Ra
n
g
a
ra
o
,
a
n
d
A
.
S
a
ra
n
y
a
,
“
Eff
icie
n
t
d
a
ta
m
in
in
g
m
o
d
e
l
f
o
r
p
re
d
ictio
n
o
f
c
h
ro
n
ic
k
id
n
e
y
d
ise
a
s
e
u
sin
g
w
r
a
p
p
e
r
m
e
th
o
d
s,”
In
t.
J
.
I
n
fo
rm
a
ti
c
s
Co
mm
u
n
.
T
e
c
h
n
o
l
.
,
v
o
l.
8
,
n
o
.
2
,
p
p
.
6
3
-
7
0
,
2
0
1
9
.
[1
4
]
N.
W
.
S
.
W
a
rd
h
a
n
i,
M
.
Y.
Ro
c
h
a
y
a
n
i,
A
.
Irian
y
,
A
.
D.
S
u
li
sty
o
n
o
,
a
n
d
P
.
L
e
sta
n
t
y
o
,
“
Cro
ss
-
v
a
li
d
a
ti
o
n
m
e
tri
c
s
f
o
r
e
v
a
lu
a
ti
n
g
c
las
sif
i
c
a
ti
o
n
p
e
rf
o
rm
a
n
c
e
o
n
i
m
b
a
lan
c
e
d
d
a
ta
,
”
in
2
0
1
9
In
ter
n
a
t
io
n
a
l
Co
n
fer
e
n
c
e
o
n
Co
mp
u
ter
,
Co
n
tro
l,
I
n
fo
rm
a
ti
c
s a
n
d
i
ts
Ap
p
li
c
a
ti
o
n
s (
IC3
INA)
,
p
p
.
1
4
-
1
8
,
2
0
1
9
.
[1
5
]
H.
L
i,
Y.
L
i,
F
.
P
o
rik
li
,
a
n
d
M
.
W
a
n
g
,
“
Co
n
v
o
lu
ti
o
n
a
l
n
e
u
ra
l
n
e
t
b
a
g
g
in
g
f
o
r
o
n
li
n
e
v
isu
a
l
trac
k
in
g
,
”
Co
mp
u
t.
Vi
s.
Ima
g
e
Un
d
e
rs
t.
,
v
o
l.
1
5
3
,
p
p
.
1
2
0
-
1
2
9
,
2
0
1
6
.
[1
6
]
I.
A
ries
h
a
n
ti
,
Y.
P
u
rw
a
n
a
n
to
,
A
.
Ra
m
a
d
h
a
n
i,
a
n
d
M
.
U.
Nu
h
a
,
“
Co
m
p
a
ra
ti
v
e
stu
d
y
o
f
b
a
n
k
ru
p
tcy
p
re
d
ictio
n
m
o
d
e
ls
,
”
T
EL
KOM
NIKA
(
T
e
le
c
o
mm
u
n
ica
ti
o
n
C
o
mp
u
ti
n
g
El
e
c
tro
n
ics
a
n
d
Co
n
tro
l)
,
v
o
l.
1
1
,
n
o
.
3
,
p
p
.
5
9
1
-
5
9
6
,
2
0
1
3
.
[1
7
]
P
.
M
o
o
rm
a
n
,
B.
Ca
li
n
g
a
e
rt,
R.
P
a
lm
ieri,
E.
Iv
e
rse
n
,
R.
Be
n
tl
e
y
,
S
.
Ha
lab
i,
A
.
Be
rc
h
u
c
k
,
a
n
d
J.
S
c
h
il
d
k
ra
u
t
,
“
Ho
rm
o
n
a
l
risk
f
a
c
to
rs
f
o
r
o
v
a
ria
n
c
a
n
c
e
r
in
p
re
m
e
n
o
p
a
u
sa
l
a
n
d
p
o
stm
e
n
o
p
a
u
sa
l
w
o
m
e
n
,
”
Am.
J
.
Ep
id
e
mi
o
l.
,
v
o
l
.
1
6
7
,
n
o
.
9
,
p
p
.
1
0
5
9
-
1
0
6
9
,
J
u
n
.
2
0
0
8
.
[1
8
]
F
.
S
h
e
n
,
S
.
Ch
e
n
,
Y.
Ga
o
,
X
.
Da
i,
a
n
d
Q.
Ch
e
n
,
“
T
h
e
p
re
v
a
len
c
e
o
f
m
a
li
g
n
a
n
t
a
n
d
b
o
rd
e
rli
n
e
o
v
a
rian
c
a
n
c
e
r
in
p
re
-
a
n
d
p
o
st
-
m
e
n
o
p
a
u
sa
l
c
h
i
n
e
se
w
o
m
e
n
,
”
Bi
o
me
d
Res
.
I
n
t.
,
v
o
l.
8
,
n
o
.
4
6
,
p
p
.
8
0
5
8
9
-
8
0
5
9
4
,
2
0
1
7
.
[1
9
]
T
.
Ha
stie,
R.
T
ib
sh
iran
i,
a
n
d
M
.
W
a
in
w
rig
h
t,
S
ta
ti
s
ti
c
a
l
L
e
a
rn
i
n
g
wit
h
S
p
a
rs
it
y
:
T
h
e
L
a
ss
o
a
n
d
Ge
n
e
ra
li
za
ti
o
n
s
.
Ch
a
p
m
a
n
a
n
d
Ha
ll
,
2
0
1
5
.
[2
0
]
R.
T
ib
sh
iran
i,
“
Re
g
re
ss
io
n
sh
rin
k
a
g
e
a
n
d
se
lec
ti
o
n
v
ia t
h
e
las
so
,
”
J
.
R.
S
ta
t
.
S
o
c
.
S
e
r.
B
,
v
o
l.
5
8
,
n
o
.
1
,
p
p
.
2
6
7
-
2
8
8
,
1
9
9
6
.
[2
1
]
J.
Ha
n
,
M
.
Ka
m
b
e
r,
a
n
d
J.
P
e
i,
D
a
t
a
M
i
n
in
g
Co
n
c
e
p
ts
a
n
d
T
e
c
h
n
iq
u
e
s T
h
ir
d
Ed
it
io
n
.
El
se
v
ier,
2
0
1
2
.
[2
2
]
Y.
Yo
h
a
n
n
e
s
a
n
d
P
.
W
e
b
b
,
“
Clas
sif
ic
a
ti
o
n
a
n
d
re
g
re
ss
io
n
tree
s,
c
a
rt
-
a
u
se
r
m
a
n
u
a
l
f
o
r
id
e
n
ti
fy
in
g
in
d
ica
to
rs
o
f
v
u
ln
e
ra
b
il
it
y
to
f
a
m
in
e
a
n
d
c
h
ro
n
ic
f
o
o
d
in
se
c
u
rit
y
.
W
a
sh
in
to
n
D.
C
,”
In
ter
n
a
ti
o
n
a
l
F
o
o
d
Po
li
c
y
Res
e
a
rc
h
In
stit
u
te
,
1
9
9
9
.
[2
3
]
P
.
R.
M
a
n
n
a
,
C.
L
.
S
tetso
n
,
A
.
T.
S
lo
m
in
sk
i,
a
n
d
K.
P
ru
it
t,
“
Ro
le
o
f
th
e
ste
ro
id
o
g
e
n
ic
a
c
u
te
re
g
u
lato
ry
p
ro
tein
i
n
h
e
a
lt
h
a
n
d
d
ise
a
se
,
”
En
d
o
c
rin
e
,
v
o
l.
5
1
,
n
o
.
1
,
p
p
.
7
-
2
1
,
2
0
1
6
.
[2
4
]
B.
H
y
lan
d
e
r,
E
.
Re
p
a
sk
y
,
P
.
S
h
rik
a
n
t,
M
.
In
te
n
g
a
n
,
A
.
Be
c
k
,
D
.
Drisc
o
ll
,
P
.
S
in
g
h
a
l,
S
.
L
e
le,
a
n
d
K.
Od
u
n
si
,
“
Ex
p
re
ss
io
n
o
f
W
il
m
s
T
u
m
o
r
g
e
n
e
(W
T
1
)
in
e
p
it
h
e
li
a
l
o
v
a
rian
c
a
n
c
e
r,
”
Gy
n
e
c
o
l.
On
c
o
l.
,
v
o
l.
1
0
1
,
n
o
.
1
,
p
p
.
1
2
-
1
7
,
2
0
0
6
.
[2
5
]
Z.
L
iu
,
K.
Ya
m
a
n
o
u
c
h
i
,
T
.
Oh
tao
,
S
.
M
a
tsu
m
u
ra
,
a
n
d
M
.
S
e
in
o
,
“
Hig
h
L
e
v
e
ls
o
f
W
il
m
s‟
Tu
m
o
r
1
(W
T
1
)
e
x
p
re
ss
io
n
we
re
a
ss
o
c
iate
d
w
it
h
a
g
g
r
e
ss
iv
e
c
li
n
ica
l
fe
a
tu
re
s
in
o
v
a
rian
c
a
n
c
e
r
,
”
An
ti
c
a
n
c
e
r
Res
.
,
v
o
l.
3
4
,
n
o
.
5
,
p
p
.
2
3
3
1
-
2
3
4
0
,
2
0
1
4
.
[2
6
]
A
.
Dz
e
li
h
o
d
z
ic an
d
D.
Do
n
k
o
,
“
C
o
m
p
a
riso
n
o
f
e
n
se
m
b
le c
las
si
f
ica
t
io
n
tec
h
n
i
q
u
e
s an
d
si
n
g
le cla
ss
if
i
e
rs co
m
p
a
riso
n
o
f
e
n
se
m
b
le
c
las
si
f
ica
ti
o
n
tec
h
n
i
q
u
e
s
a
n
d
sin
g
le
c
las
sif
ier
s
p
e
r
f
o
rm
a
n
c
e
f
o
r
c
u
sto
m
e
r
c
re
d
it
a
s
se
ss
m
e
n
t
,
”
M
o
d
e
l.
Arti
f.
In
tell.
,
v
o
l.
1
1
,
n
o
.
3
,
p
p
.
1
4
0
-
1
5
0
,
2
0
1
6
.
Evaluation Warning : The document was created with Spire.PDF for Python.