I
AE
S In
t
er
na
t
io
na
l J
o
urna
l o
f
Art
if
icia
l In
t
ellig
ence
(
I
J
-
AI
)
Vo
l.
14
,
No
.
6
,
Dec
em
b
er
2
0
2
5
,
p
p
.
4
8
7
8
~
4
8
9
0
I
SS
N:
2
2
5
2
-
8
9
3
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijai.v
14
.i
6
.
p
p
4
8
7
8
-
4
8
9
0
4878
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
i
.
ia
esco
r
e.
co
m
Suppo
rt
v
ec
tor m
a
chine per
forma
nce:
sim
ula
tion
and rice
pheno
lo
g
y
appli
ca
tion
H
eng
k
i Mu
ra
di
1,
2
,
Asep Sa
ef
ud
di
n
1
,
I
M
a
de
S
um
er
t
a
j
a
y
a
1
,
Ag
us
M
o
ha
m
a
d So
leh
1
,
Dede
Dirg
a
ha
y
u Do
m
iri
3
1
D
e
p
a
r
t
me
n
t
o
f
S
t
a
t
i
s
t
i
c
s
a
n
d
D
a
t
a
S
c
i
e
n
c
e
,
S
c
h
o
o
l
o
f
D
a
t
a
S
c
i
e
n
c
e
,
M
a
t
h
e
mat
i
c
s,
a
n
d
I
n
f
o
r
mat
i
c
s
,
I
P
B
U
n
i
v
e
r
si
t
y
,
B
o
g
o
r
,
I
n
d
o
n
e
si
a
2
R
e
s
e
a
r
c
h
C
e
n
t
e
r
f
o
r
D
a
t
a
S
c
i
e
n
c
e
a
n
d
I
n
f
o
r
ma
t
i
o
n
,
N
a
t
i
o
n
a
l
R
e
se
a
r
c
h
a
n
d
I
n
n
o
v
a
t
i
o
n
A
g
e
n
c
y
(
B
R
I
N
)
,
B
a
n
d
u
n
g
,
I
n
d
o
n
e
si
a
3
R
e
s
e
a
r
c
h
C
e
n
t
e
r
f
o
r
G
e
o
i
n
f
o
r
ma
t
i
c
s,
N
a
t
i
o
n
a
l
R
e
se
a
r
c
h
a
n
d
I
n
n
o
v
a
t
i
o
n
A
g
e
n
c
y
(
B
R
I
N
)
,
B
a
n
d
u
n
g
,
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
May
27
,
2
0
2
4
R
ev
is
ed
Oct
28
,
2
0
2
5
Acc
ep
ted
No
v
08
,
2
0
2
5
In
t
h
e
c
a
se
o
f
c
las
sifica
ti
o
n
,
m
o
d
e
l
a
c
c
u
ra
c
y
is
e
x
p
e
c
ted
to
re
su
lt
in
c
o
rre
c
t
p
re
d
ictio
n
s.
Th
is
stu
d
y
a
ims
t
o
a
n
a
ly
z
e
th
e
p
e
rfo
rm
a
n
c
e
o
f
two
k
i
n
d
s
o
f
su
p
p
o
rt
v
e
c
to
r
m
a
c
h
in
e
(S
VM)
m
e
th
o
d
s:
t
h
e
su
p
p
o
r
t
v
e
c
to
r
m
a
c
h
in
e
o
n
e
v
e
rsu
s
o
n
e
(S
VM
Ov
O)
m
e
th
o
d
a
n
d
th
e
g
e
n
e
ra
li
z
e
d
m
u
lt
icla
s
s
s
u
p
p
o
rt
v
e
c
to
r
m
a
c
h
in
e
(G
e
n
S
VM)
m
e
th
o
d
.
T
h
is
m
e
th
o
d
will
c
o
m
p
a
re
to
t
h
e
g
e
n
e
ra
li
z
e
d
li
n
e
a
r
m
o
d
e
l,
n
a
m
e
ly
th
e
m
u
l
ti
n
o
m
ial
lo
g
isti
c
re
g
re
ss
io
n
(
M
LR)
m
e
th
o
d
.
S
imu
latio
n
s
we
re
c
o
n
d
u
c
ted
u
si
n
g
S
V
M
Ov
O
a
n
d
Ge
n
S
VM
m
e
th
o
d
s
to
g
e
t
a
n
o
v
e
r
v
iew
o
f
th
e
p
a
ra
m
e
ters
a
ff
e
c
ti
n
g
b
o
t
h
m
e
th
o
d
s'
p
e
rfo
rm
a
n
c
e
.
F
u
rth
e
rm
o
re
,
th
e
th
re
e
c
las
sifica
ti
o
n
m
e
th
o
d
s
a
re
im
p
lem
e
n
ted
in
t
h
e
c
a
se
o
f
m
o
d
e
ll
in
g
th
e
ri
c
e
p
h
e
n
o
lo
g
y
a
n
d
tes
ted
fo
r
p
e
r
fo
rm
a
n
c
e
.
S
imu
latio
n
re
su
lt
s
sh
o
w
th
a
t,
h
o
we
v
e
r,
th
e
S
VM
Ov
O
a
n
d
G
e
n
S
VM
m
a
c
h
in
e
lea
rn
in
g
m
e
th
o
d
s
a
re
se
n
siti
v
e
to
th
e
c
h
o
ice
o
f
m
o
d
e
l
p
a
ra
m
e
ters
.
Th
e
e
m
p
iri
c
a
l
stu
d
y
re
su
lt
s
s
h
o
w
th
a
t
th
e
S
V
M
Ov
O
a
n
d
G
e
n
S
VM
m
e
th
o
d
s
c
a
n
p
ro
d
u
c
e
sa
ti
sfa
c
to
ry
m
o
d
e
l
a
c
c
u
ra
c
y
a
n
d
a
re
c
o
m
p
a
ra
b
le
to
th
e
M
LR
m
e
th
o
d
.
T
h
e
b
e
st
rice
p
h
e
n
o
l
o
g
y
m
o
d
e
l
a
c
c
u
ra
c
y
wa
s
o
b
tain
e
d
fro
m
th
e
S
VM
Ov
O
m
o
d
e
l,
wh
e
re
7
9
.
2
0
±
0
.
2
1
o
v
e
ra
ll
a
c
c
u
ra
c
y
a
n
d
7
0
.
6
9
±
0
.
2
9
k
a
p
p
a
we
re
o
b
tain
e
d
.
T
h
is
re
se
a
rc
h
c
a
n
b
e
c
o
n
ti
n
u
e
d
b
y
h
a
n
d
l
in
g
sa
m
p
les
,
e
sp
e
c
ially
wh
e
n
c
las
s
m
e
m
b
e
rs
a
re
a
m
in
o
rit
y
,
a
n
d
c
a
n
a
lso
a
d
d
ra
n
d
o
m
e
ffe
c
ts t
o
th
e
S
VM
m
o
d
e
l.
K
ey
w
o
r
d
s
:
Gen
SVM
M
u
ltin
o
m
ial
lo
g
is
tic
r
eg
r
ess
io
n
O
n
e
v
er
s
u
s
o
n
e
Pad
d
y
S
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
e
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
Hen
g
k
i M
u
r
a
d
i
Dep
ar
tm
en
t o
f
Statis
tics
an
d
Data
Scien
ce
,
Sch
o
o
l o
f
Data
Scien
ce
,
Ma
th
em
atics,
an
d
I
n
f
o
r
m
atics
I
PB
Un
iv
er
s
ity
B
o
g
o
r
,
I
n
d
o
n
esia
E
m
ail:
h
en
g
k
im
u
r
ad
is
3
@
g
m
ai
l.c
o
m
1.
I
NT
RO
D
UCT
I
O
N
T
h
e
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
e
(
SVM)
is
a
s
u
p
er
v
is
ed
,
n
o
n
-
p
ar
am
etr
ic
class
if
icatio
n
m
eth
o
d
th
at
h
as
d
em
o
n
s
tr
ated
s
tr
o
n
g
p
er
f
o
r
m
an
ce
in
p
r
o
d
u
cin
g
h
i
g
h
-
ac
cu
r
ac
y
m
o
d
els
.
It
ef
f
ec
tiv
ely
a
d
d
r
ess
in
g
co
m
m
o
n
m
o
d
ellin
g
ch
allen
g
es
s
u
ch
as
m
u
ltico
llin
ea
r
ity
,
n
o
n
lin
ea
r
it
y
,
an
d
o
v
er
f
itti
n
g
[
1
]
,
[
2
]
.
D
u
e
to
its
r
o
b
u
s
tn
ess
an
d
v
er
s
atility
,
SVM
h
as
b
ee
n
wid
ely
ap
p
lied
ac
r
o
s
s
d
iv
er
s
e
s
cien
tific
f
ield
s
,
in
clu
d
in
g
p
atter
n
r
ec
o
g
n
itio
n
[
3
]
,
r
e
m
o
te
s
en
s
in
g
[
4
]
,
[
5
]
,
as
well
as m
ed
ical
ap
p
licatio
n
s
s
u
ch
as c
an
ce
r
a
n
d
tu
m
o
r
d
iag
n
o
s
is
[
6
]
,
[
7
]
.
T
h
e
SVM
m
eth
o
d
was
in
itially
d
ev
elo
p
ed
f
o
r
b
i
n
ar
y
class
if
icatio
n
ca
s
es.
Dev
elo
p
in
g
SVMs
f
o
r
m
u
lticlas
s
ca
s
e
s
is
d
if
f
icu
lt
b
ec
au
s
e
th
e
o
u
tp
u
ts
ar
e
n
o
t
o
n
a
ca
lib
r
ated
s
ca
le
an
d
ar
e
d
if
f
ic
u
lt
to
co
m
p
a
r
e
[
8
]
.
I
n
th
e
ca
s
e
o
f
m
u
lticlas
s
class
if
icatio
n
,
th
e
SVM
m
eth
o
d
w
as
d
ev
elo
p
e
d
u
s
in
g
p
r
im
ar
y
b
i
n
ar
y
class
if
icatio
n
,
s
u
ch
as
th
e
o
n
e
v
er
s
u
s
o
n
e
(
Ov
O)
,
d
ir
ec
te
d
ac
y
clic
g
r
ap
h
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
e
s
(
DAGSVM
)
,
an
d
o
n
e
v
e
r
s
u
s
all
(
Ov
A)
m
eth
o
d
s
.
T
h
e
Ov
O
m
eth
o
d
is
s
o
m
eti
m
es
b
etter
th
an
th
e
DAGSVM
an
d
Ov
A
m
eth
o
d
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
S
u
p
p
o
r
t v
ec
to
r
ma
ch
in
e
p
erfo
r
ma
n
ce
:
s
imu
la
tio
n
a
n
d
r
ice
p
h
en
o
lo
g
y
a
p
p
lica
tio
n
(
He
n
g
ki
Mu
r
a
d
i
)
4879
[
9
]
,
[
1
0
]
.
I
n
th
e
m
u
lticlas
s
cl
ass
if
icatio
n
o
f
co
m
p
lex
r
e
m
o
te
s
en
s
in
g
d
ata,
th
e
Ov
O
SVM
m
eth
o
d
r
em
ai
n
s
co
m
p
ar
ab
le
to
th
e
q
u
an
tu
m
m
u
lticlas
s
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
e
(
QM
SVM)
a
n
d
Ov
A
m
eth
o
d
s
[
1
1
]
.
A
m
u
lticlas
s
class
if
icatio
n
m
eth
o
d
t
h
at
is
n
o
t
b
ased
o
n
b
i
n
ar
y
class
if
icatio
n
was
also
d
ev
elo
p
e
d
in
[
1
2
]
,
n
am
ely
a
m
u
lticlas
s
clas
s
if
icat
io
n
m
eth
o
d
u
s
in
g
a
s
im
p
lex
ap
p
r
o
ac
h
.
T
h
e
ad
v
a
n
tag
e
o
f
th
is
s
im
p
lex
m
eth
o
d
is
th
at
it
ca
n
p
r
o
d
u
ce
class
if
icatio
n
s
with
o
u
t
a
m
b
ig
u
ity
i
n
th
e
p
r
e
d
ictio
n
s
p
ac
e
an
d
allo
ws
g
eo
m
et
r
ic
in
ter
p
r
etatio
n
.
T
h
is
m
et
h
o
d
is
ca
lled
th
e
g
e
n
er
alize
d
m
u
lticlas
s
s
u
p
p
o
r
t
v
ec
t
o
r
m
ac
h
in
e
(
Gen
SVM)
m
eth
o
d
.
T
h
e
Gen
SVM
m
eth
o
d
is
clai
m
ed
to
h
av
e
q
u
ite
a
co
m
p
eti
tiv
e
p
er
f
o
r
m
a
n
ce
co
m
p
ar
e
d
to
th
e
SVM
Ov
O,
SVM
Ov
A,
DAGSVM
class
if
icatio
n
m
eth
o
d
s
an
d
s
ev
er
a
l
o
th
er
m
u
lticlas
s
class
if
icati
o
n
m
eth
o
d
s
[
1
2
]
.
Mu
lticlas
s
cla
s
s
if
icat
io
n
ca
n
also
b
e
d
o
n
e
u
s
in
g
th
e
m
u
ltin
o
m
ial
lo
g
is
tic
r
eg
r
ess
io
n
(
ML
R
)
m
eth
o
d
.
T
h
is
m
eth
o
d
is
b
ased
o
n
th
e
g
en
er
alize
d
lin
ea
r
m
o
d
el
(
GL
M)
[
1
3
]
.
T
h
e
ML
R
m
et
h
o
d
c
an
ac
cu
r
ately
d
e
f
in
e
th
e
r
elatio
n
s
h
ip
b
etwe
en
g
r
o
u
p
s
o
f
ex
p
lan
at
o
r
y
v
ar
iab
les
an
d
r
esp
o
n
s
e
v
ar
ia
b
les,
id
en
ti
f
y
th
e
i
n
f
lu
en
ce
o
f
ea
ch
v
ar
iab
le,
a
n
d
p
r
ed
ict
th
e
class
if
icatio
n
o
f
ea
ch
ca
s
e
[
1
4
]
.
T
h
is
s
tu
d
y
aim
s
to
an
aly
ze
t
h
e
p
er
f
o
r
m
an
ce
o
f
two
ty
p
es
o
f
SVM
m
eth
o
d
s
:
th
e
SVM
Ov
O
m
eth
o
d
b
ased
o
n
b
in
ar
y
class
if
icatio
n
an
d
th
e
Ge
n
SVM
m
eth
o
d
b
as
ed
o
n
n
o
n
-
b
in
a
r
y
class
if
icatio
n
.
T
h
is
m
eth
o
d
will
co
m
p
ar
e
to
th
e
GL
M
,
n
am
ely
th
e
ML
R
m
eth
o
d
.
Fu
r
t
h
er
m
o
r
e
,
th
e
th
r
ee
class
if
ic
atio
n
m
eth
o
d
s
ar
e
im
p
lem
en
ted
i
n
th
e
ca
s
e
o
f
m
o
d
ellin
g
r
ice
p
h
en
o
lo
g
y
an
d
t
ested
f
o
r
p
er
f
o
r
m
an
ce
.
T
h
e
S
VM
m
eth
o
d
ca
n
b
e
ap
p
lied
f
o
r
m
o
d
ellin
g
r
ice
p
h
en
o
lo
g
y
[
4
]
,
[
5
]
,
as
ca
n
th
e
ML
R
m
eth
o
d
[
1
5
]
,
b
u
t
s
p
ec
if
ically
,
th
er
e
h
as
b
ee
n
n
o
r
esear
ch
th
at
ap
p
lies
th
e
Gen
SVM
m
eth
o
d
f
o
r
m
o
d
ellin
g
o
f
r
ice
p
h
e
n
o
lo
g
y
an
d
c
o
m
p
ar
in
g
th
e
th
r
e
e
m
eth
o
d
s
to
g
et
t
h
e
b
est m
o
d
el.
T
h
is
s
tu
d
y
also
d
e
v
elo
p
s
th
e
i
n
p
u
t m
o
d
el
wh
er
e,
in
p
r
e
v
io
u
s
s
tu
d
ies,
s
o
m
e
r
esear
ch
er
s
o
n
l
y
u
s
ed
o
n
e
in
p
u
t
v
a
r
iab
le,
s
u
c
h
as
a
s
in
g
le
VH
p
o
lar
izatio
n
[
4
]
,
[
1
6
]
,
[
1
7
]
an
d
VH/VV
p
o
lar
izatio
n
in
d
ex
[
1
8
]
,
[
1
9
]
.
I
n
th
is
s
tu
d
y
,
we
will
u
s
e
b
o
th
VV
an
d
VH
p
o
lar
izatio
n
,
as
well
as
p
o
lar
izatio
n
in
d
ices
s
u
ch
as
r
atio
p
o
lar
izatio
n
in
d
ex
(
R
PI)
,
n
o
r
m
alize
d
d
if
f
er
e
n
t
p
o
lar
izatio
n
in
d
ex
(
NDPI
)
,
a
n
d
a
v
er
ag
e
p
o
lar
izatio
n
in
d
ex
(
API
)
[
2
0
]
.
I
n
ad
d
itio
n
,
a
r
e
-
c
lass
if
icatio
n
s
ce
n
ar
io
o
n
th
e
r
ice
p
h
en
o
lo
g
y
class
was
al
s
o
t
ested
to
o
b
tain
th
e
b
est r
ice
p
h
en
o
lo
g
y
m
o
d
el.
2.
M
UL
T
I
CL
A
SS
C
L
AS
SI
F
I
C
AT
I
O
N
M
E
T
H
O
D
2
.
1
.
Su
pp
o
rt
v
ec
t
o
r
ma
chine
I
n
th
e
ca
s
e
o
f
b
in
a
r
y
class
if
icatio
n
,
let
∈
{
−
1
,
1
}
an
d
a
s
et
o
f
p
r
e
d
i
cto
r
s
{
1
,
2
,
…
,
}
,
∈
ℝ
.
T
h
e
m
o
s
t
o
p
tim
al
b
ar
r
ier
is
n
ee
d
ed
to
s
ep
ar
ate
th
e
n
e
g
ativ
e
an
d
p
o
s
itiv
e
class
es,
c
alled
a
h
y
p
er
p
la
n
e.
T
h
e
h
y
p
er
p
lan
e
eq
u
atio
n
ca
n
b
e
s
tated
in
(
1
)
[
2
1
]
,
[
2
2
]
:
0
+
1
1
+
⋯
+
=
0
(
1
)
I
n
m
an
y
r
ea
l
-
w
o
r
ld
a
p
p
licatio
n
s
,
th
e
r
elatio
n
s
h
ip
s
b
etwe
en
v
ar
iab
les
ar
e
n
o
n
-
lin
ea
r
.
T
h
e
m
ain
f
ea
tu
r
e
o
f
SVM
is
its
ab
ili
ty
to
m
ap
p
r
o
b
lem
s
in
to
a
h
ig
h
e
r
d
im
en
s
io
n
al
s
p
ac
e
u
s
in
g
a
p
r
o
ce
s
s
k
n
o
wn
as
th
e
k
er
n
el
tr
ic
k
s
o
th
at
n
o
n
-
li
n
ea
r
r
elatio
n
s
h
ip
s
b
ec
o
m
e
lin
ea
r
[
1
]
.
I
n
(
2
)
ca
n
b
e
tr
a
n
s
f
o
r
m
e
d
b
y
th
e
f
u
n
ctio
n
(
.
)
to
b
ec
o
m
e
[
3
]
:
≥
0
,
(
(
)
+
0
)
≥
1
−
,
=
1
,
2
,
…
,
(
2
)
Sin
ce
th
e
d
esire
d
s
p
ac
e
(
.
)
is
u
n
k
n
o
wn
,
s
o
lv
in
g
th
e
p
r
o
b
lem
(
2
)
s
u
b
ject
to
co
n
s
tr
ain
t
(
2
)
b
e
co
m
es
m
o
r
e
co
m
p
licated
.
T
o
o
v
e
r
co
m
e
th
is
p
r
o
b
lem
,
th
e
d
u
al
o
f
S
VM
is
p
r
esen
ted
as (
3
)
[
3
]
,
[
2
3
]
:
min
∑
=
1
−
1
2
∑
∑
=
1
(
,
)
=
1
∑
=
1
=
0
0
≤
≤
,
=
1
,
…
,
(
3
)
W
h
er
e
=
(
1
,
2
,
…
,
)
is
th
e
L
ag
r
an
g
e
m
u
lt
ip
lier
an
d
(
,
)
is
a
s
y
m
m
etr
ic
k
er
n
el
f
u
n
ctio
n
with
a
n
o
n
n
e
g
ativ
e
v
alu
e.
Ker
n
el
f
u
n
ctio
n
s
co
n
s
is
t
o
f
lin
ea
r
,
p
o
ly
n
o
m
ial,
an
d
r
ad
ial
b
asis
f
u
n
ctio
n
(
R
B
F)
k
er
n
el
f
u
n
ctio
n
s
[
8
]
.
Fo
r
th
e
ca
s
e
o
f
m
u
ltin
o
m
ial
cl
ass
if
icatio
n
,
SVM
m
eth
o
d
s
h
a
v
e
b
ee
n
u
s
ed
,
s
u
ch
as
th
e
O
v
O
an
d
Ov
A
m
eth
o
d
s
.
B
o
th
o
f
th
ese
m
eth
o
d
s
ar
e
b
ased
o
n
b
in
a
r
y
class
if
icatio
n
,
an
d
i
n
s
ev
er
al
test
s
,
th
e
O
v
O
m
eth
o
d
is
m
o
r
e
co
m
p
etitiv
e
an
d
ea
s
ier
to
ap
p
ly
th
an
th
e
Ov
A
m
eth
o
d
.
T
h
e
o
n
e
-
ag
ain
s
t
-
o
n
e
m
eth
o
d
b
u
ild
s
(
+
1
)
/
2
class
if
ier
s
,
wh
ich
ar
e
tr
ain
ed
o
n
e
b
y
o
n
e
f
o
r
t
h
e
two
class
es.
Fo
r
tr
ain
in
g
d
ata
f
r
o
m
class
−
an
d
class
−
,
th
e
b
est s
o
lu
tio
n
is
th
e
s
o
lu
tio
n
to
th
e
p
r
o
b
lem
as in
(
4
)
[
9
]
:
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
14
,
No
.
6
,
Dec
em
b
er
2
0
2
5
:
4
8
7
8
-
4
8
9
0
4880
min
,
0
,
1
2
(
)
+
∑
(
)
(
)
+
≥
1
−
,
jik
a
=
(
)
(
)
+
≤
−
1
+
,
jik
a
≠
I
≥
0
(
4
)
Sev
er
al
m
eth
o
d
s
f
o
r
f
u
r
th
e
r
test
in
g
ar
e
u
s
ed
af
ter
all
(
+
1
)
/
2
class
if
ier
s
h
av
e
b
ee
n
b
u
ilt.
On
e
way
is
to
u
s
e
a
v
o
tin
g
s
tr
ateg
y
p
r
o
p
o
s
ed
b
y
Frie
d
m
a
n
(
1
9
9
6
)
,
wh
ich
is
ca
lled
th
e
"
Ma
x
w
in
s
"
s
tr
ateg
y
[
9
]
,
[
2
4
]
.
I
n
th
e
'
'
m
ax
win
''
al
g
o
r
ith
m
,
ea
ch
class
if
ier
g
iv
es
o
n
e
v
o
te
f
o
r
th
e
class
o
f
its
c
h
o
ice,
an
d
th
e
f
in
al
r
esu
lt
is
th
e
class
with
th
e
m
o
s
t
v
o
tes
[
2
4
]
.
T
h
e
p
er
f
o
r
m
a
n
ce
o
f
th
e
SVM
m
u
lticlas
s
cl
ass
if
icatio
n
m
eth
o
d
also
d
ep
en
d
s
o
n
k
er
n
el
s
elec
tio
n
.
T
h
e
m
u
lticlas
s
SVM
cla
s
s
if
icatio
n
m
eth
o
d
u
s
es
th
e
Ov
O
m
eth
o
d
[
2
5
]
with
R
B
F
k
er
n
el
[
3
]
,
[
5
]
.
A
m
u
lticlas
s
class
if
icat
io
n
m
e
th
o
d
th
at
is
n
o
t
b
ased
o
n
b
in
a
r
y
class
if
icatio
n
was
also
d
ev
elo
p
ed
b
y
B
u
r
g
an
d
Gr
o
en
en
[
1
2
]
,
n
a
m
ely
th
e
Gen
SVM
m
et
h
o
d
.
T
h
e
Gen
SVM
m
eth
o
d
is
a
f
lex
ib
le
an
d
g
en
e
r
al
m
u
lticlas
s
S
VM
m
eth
o
d
th
at
u
s
es
s
im
p
lex
co
d
in
g
to
f
o
r
m
u
late
th
e
m
u
lticlas
s
SV
M
p
r
o
b
lem
as
a
s
in
g
le
o
p
tim
izatio
n
p
r
o
b
lem
,
wh
ich
r
ed
u
ce
s
to
a
b
in
ar
y
SVM
wh
en
k
=2
.
T
h
e
co
m
p
lete
lo
s
s
f
u
n
ctio
n
o
f
Gen
SVM
is
as (
5
)
[
1
2
]
:
=
1
∑
∑
(
∑
ℎ
(
(
)
)
≠
)
1
/
∈
+
W
′
W
=
1
>
0
,
=
,
∈
,
=
{
∶
=
}
(
5
)
T
h
e
p
r
ed
icte
d
class
lab
els
o
n
ly
co
r
r
esp
o
n
d
to
th
e
clo
s
est
s
im
p
lex
v
er
tices
as
m
ea
s
u
r
ed
b
y
th
e
s
q
u
ar
ed
E
u
clid
ea
n
n
o
r
m
as in
(
6
)
:
̂
+
1
=
‖
+
1
′
−
′
‖
2
,
=
1
,
2
,
…
,
(
6
)
T
h
e
Gen
SVM
alg
o
r
ith
m
is
av
ailab
le
in
th
e
Gen
s
v
m
p
ac
k
ag
e
in
th
e
R
p
r
o
g
r
a
m
.
Gen
SVM
ca
n
b
e
u
s
ed
f
o
r
lin
ea
r
an
d
n
o
n
lin
ea
r
m
u
lticlas
s
SVM
clas
s
if
icatio
n
.
I
n
g
en
er
al,
lin
ea
r
class
if
icatio
n
will
b
e
f
aster
,
b
u
t
d
ep
en
d
i
n
g
o
n
th
e
d
ataset,
h
ig
h
er
class
if
icatio
n
p
er
f
o
r
m
a
n
ce
ca
n
b
e
ac
h
ie
v
ed
u
s
in
g
n
o
n
lin
e
ar
k
er
n
els
[
2
6
]
.
2
.
2
.
M
ultino
m
ia
l lo
g
is
t
ic
re
g
re
s
s
io
n
I
n
th
is
s
tu
d
y
,
we
ca
r
r
ied
o
u
t
c
lass
if
icatio
n
u
s
in
g
th
e
an
aly
tical
class
if
icatio
n
m
eth
o
d
,
n
a
m
ely
ML
R
.
T
h
e
m
eth
o
d
is
a
n
o
n
-
p
ar
am
et
r
ic
class
if
icatio
n
m
eth
o
d
[
6
]
,
p
ar
t
o
f
th
e
f
am
ily
o
f
GL
M
m
eth
o
d
s
.
I
t
is
u
s
ed
wh
en
th
e
r
esp
o
n
s
e
v
a
r
iab
le
h
a
s
m
o
r
e
th
a
n
two
ca
teg
o
r
ies
[
1
3
]
.
Su
p
p
o
s
e
∈
{
1
,
2
,
…
,
}
an
d
a
s
et
o
f
p
r
ed
i
cto
r
s
{
{
1
,
2
,
…
,
}
,
∈
ℝ
,
th
e
ML
R
eq
u
atio
n
in
p
r
o
b
ab
ilit
y
f
o
r
m
ca
n
b
e
e
x
p
r
ess
ed
as (
7
)
:
(
x
)
=
(
=
|
x
)
ex
p
(
+
β
′
x
)
1
+
∑
ex
p
(
ℎ
β
h
′
x
)
−
1
ℎ
−
1
(
7
)
T
h
e
ML
R
m
o
d
el
p
ar
a
m
eter
s
ca
n
b
e
esti
m
ated
u
s
in
g
th
e
m
ax
im
u
m
lik
elih
o
o
d
m
et
h
o
d
.
T
h
e
ML
R
m
eth
o
d
p
er
f
o
r
m
s
well
an
d
is
co
m
p
etiti
v
e
with
th
e
r
an
d
o
m
f
o
r
est cla
s
s
if
icatio
n
m
eth
o
d
[
1
3
]
.
2
.
3
.
M
o
del a
cc
ura
cy
m
ea
s
ur
em
ent
Mo
d
el
ac
cu
r
ac
y
is
m
ea
s
u
r
ed
f
r
o
m
o
v
er
all
ac
cu
r
ac
y
(
OA)
s
tatis
tics
an
d
k
ap
p
a
s
tatis
tic
s
(
κ)
as
s
h
o
wn
in
(
8
)
a
n
d
(
9
)
[
2
7
]
.
=
∑
=
1
|
|
(
8
)
=
−
1
−
(
9
)
W
h
er
e
is
th
e
n
u
m
b
er
o
f
p
o
in
ts
p
r
ed
icted
c
o
r
r
ec
tly
,
=
∑
.
.
/
n
u
m
b
er
o
f
p
o
in
ts
test
ed
,
.
=
∑
,
.
=
∑
an
d
th
e
n
u
m
b
er
o
f
item
s
th
at
h
av
e
to
b
e
class
if
ied
.
OA
p
r
o
v
id
es
a
s
im
p
le
m
ea
s
u
r
e
o
f
th
e
p
r
o
p
o
r
tio
n
o
f
co
r
r
ec
t
c
lass
if
icatio
n
s
,
wh
ile
Kap
p
a
ad
d
s
im
p
o
r
tan
t
in
f
o
r
m
atio
n
b
y
t
ak
in
g
in
to
ac
c
o
u
n
t
th
e
p
o
s
s
ib
ilit
y
o
f
r
an
d
o
m
ag
r
ee
m
en
t,
th
e
r
eb
y
p
r
o
v
id
in
g
a
f
air
er
e
v
alu
atio
n
in
ca
s
es
o
f
im
b
alan
ce
d
class
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
S
u
p
p
o
r
t v
ec
to
r
ma
ch
in
e
p
erfo
r
ma
n
ce
:
s
imu
la
tio
n
a
n
d
r
ice
p
h
en
o
lo
g
y
a
p
p
lica
tio
n
(
He
n
g
ki
Mu
r
a
d
i
)
4881
d
is
tr
ib
u
tio
n
s
[
2
8
]
,
[
2
9
]
.
T
h
e
co
m
b
in
atio
n
o
f
OA
an
d
Kap
p
a
h
as
b
ee
n
s
h
o
wn
to
co
m
p
l
em
en
t
ea
ch
o
th
er
in
im
ag
e
class
if
icatio
n
r
esear
ch
,
th
er
eb
y
im
p
r
o
v
in
g
th
e
r
eliab
ilit
y
o
f
m
o
d
el
ass
ess
m
en
t
[
3
0
]
.
3.
M
E
T
H
O
D
3
.
1
.
Sim
ula
t
i
o
n study
T
h
e
s
im
u
latio
n
s
tep
s
tak
en
ca
n
b
e
s
ee
n
in
Fig
u
r
e
1
.
T
h
e
s
i
m
u
latio
n
s
tu
d
y
is
co
n
d
u
cte
d
t
o
o
b
tain
a
n
o
v
er
v
iew
o
f
th
e
p
ar
am
eter
s
ettin
g
s
in
th
e
SVM
Ov
O
an
d
Gen
SVM
m
eth
o
d
s
.
I
n
th
e
S
VM
Ov
O
m
eth
o
d
,
th
e
co
s
t
(
)
an
d
g
am
m
a
(
)
p
ar
am
eter
s
ar
e
th
o
u
g
h
t
to
in
f
lu
e
n
ce
m
o
d
el
p
er
f
o
r
m
a
n
ce
.
Similar
ly
,
with
th
e
Gen
SVM
m
eth
o
d
,
th
e
p
ar
a
m
e
ter
s
k
ap
p
a
(
)
an
d
lam
b
d
a
(
)
ar
e
th
o
u
g
h
t
to
in
f
lu
e
n
ce
th
e
p
e
r
f
o
r
m
an
ce
o
f
t
h
e
r
esu
ltin
g
m
o
d
el.
I
n
p
ac
k
ag
e
e
1
0
7
1
i
n
th
e
R
p
r
o
g
r
am
m
e,
th
e
p
ar
am
eter
s
an
d
γ
in
t
h
e
SVM
Ov
O
m
eth
o
d
ca
n
b
e
u
s
ed
b
y
d
ef
au
lt
[
2
5
]
,
b
u
t
it
is
also
p
o
s
s
ib
le
to
m
o
d
if
y
th
em
.
T
h
en
,
in
th
e
g
en
s
v
m
p
ac
k
a
g
e
[
2
6
]
,
th
e
p
ar
am
eter
s
κ
a
n
d
λ
ca
n
u
s
e
th
e
d
ef
au
lt
s
ettin
g
s
b
u
t
ca
n
also
b
e
m
o
d
if
ied
.
Me
an
wh
ile,
th
e
ML
R
m
eth
o
d
d
o
es
n
o
t
r
eq
u
ir
e
p
ar
a
m
eter
s
ettin
g
s
b
ec
au
s
e,
in
th
e
m
ax
im
u
m
lik
elih
o
o
d
m
et
h
o
d
,
th
e
ML
R
m
o
d
el
p
ar
am
eter
s
ca
n
b
e
d
eter
m
in
e
d
with
o
u
t in
v
o
lv
i
n
g
th
e
in
itial v
alu
e
o
f
t
h
e
m
o
d
el
p
ar
am
eter
s
.
Fig
u
r
e
1
.
Simu
latio
n
s
tep
s
T
h
e
s
im
u
latio
n
d
ata
u
s
ed
i
n
th
is
r
esear
ch
is
g
lass
d
ata
ac
ce
s
s
ed
f
r
o
m
th
e
k
k
n
n
R
p
ac
k
a
g
e,
wh
ich
co
n
s
is
ts
o
f
n
=1
8
5
,
th
e
n
u
m
b
e
r
o
f
p
r
ed
icto
r
s
is
n
in
e,
an
d
th
e
n
u
m
b
er
o
f
class
es
is
4
[
3
1
]
.
I
n
th
is
s
im
u
latio
n
,
all
p
r
ed
icto
r
s
ar
e
in
clu
d
e
d
in
t
h
e
m
o
d
el.
I
n
th
e
SVM
Ov
O
m
eth
o
d
,
C
V=
1
0
-
f
o
ld
an
d
R
B
F
k
er
n
el
s
ettin
g
s
ar
e
u
s
ed
,
wh
ile
th
e
an
d
p
ar
am
eter
s
u
s
ed
ar
e
s
et
to
th
e
v
alu
e
∈
{
2
0
,
2
1
,
2
2
,
2
4
,
2
6
,
2
8
,
2
10
,
2
12
,
2
14
,
2
16
}
an
d
th
e
v
alu
e
∈
{
2
−
6
,
2
−
4
,
2
−
2
,
2
−
1
,
2
0
,
2
2
,
2
3
,
2
4
,
2
5
,
2
6
}
[
9
]
.
T
h
e
Gen
SV
M
m
eth
o
d
is
s
et
with
p
ar
a
m
eter
v
alu
es
∈
{
−
0
.
9
,
−
0
.
5
,
0
.
5
,
1
.
5
,
2
.
0
,
2
.
5
,
3
.
0
,
4
.
0
,
4
.
5
,
5
.
0
}
,
∈
{
2
0
,
2
1
,
2
2
,
2
4
,
2
6
,
2
8
,
2
10
,
2
12
,
2
14
,
2
16
}
,
an
d
∈
{
1
.
0
,
1
.
1
,
1
.
2
,
1
.
4
,
1
.
5
,
1
.
6
,
1
.
7
,
1
.
8
,
1
.
9
,
2
.
0
}
.
3
.
2
.
Sim
ula
t
i
o
n r
esu
lt
s
T
h
e
s
im
u
latio
n
r
esu
lts
s
h
o
w
d
if
f
er
en
ce
s
in
th
e
ac
c
u
r
ac
y
o
f
th
e
SVM
Ov
O
m
o
d
el
in
ea
ch
c
o
s
t
p
ar
am
eter
ex
p
er
im
e
n
t
a
n
d
th
e
g
am
m
a
p
ar
am
eter
ex
p
er
im
en
t.
T
h
e
s
im
u
latio
n
r
esu
lts
s
h
o
w
t
h
at
th
e
SVM
Ov
O
m
eth
o
d
'
s
p
er
f
o
r
m
a
n
ce
d
e
p
en
d
s
o
n
th
e
c
o
s
t
v
alu
e
an
d
g
a
m
m
a
p
ar
am
eter
s
.
T
h
e
ef
f
ec
t
o
f
th
e
co
s
t
p
ar
a
m
eter
v
alu
e
o
n
th
e
p
er
f
o
r
m
an
ce
o
f
t
h
e
SVM
Ov
O
m
eth
o
d
is
p
r
ese
n
ted
in
Fig
u
r
e
2
.
Me
an
wh
ile,
th
e
ef
f
ec
t o
f
g
am
m
a
p
ar
am
eter
s
is
p
r
esen
ted
in
Fig
u
r
e
3
.
I
n
th
e
ca
s
e
o
f
g
lass
d
at
a
class
if
icatio
n
,
it
s
h
o
ws
th
at
th
e
ac
cu
r
ac
y
o
f
th
e
SVM
Ov
O
m
o
d
el
r
ea
ch
es
1
0
0
%
wh
en
co
s
t
=
2
14
.
T
h
er
ef
o
r
e,
r
esear
ch
er
s
ca
n
ad
ju
s
t
th
e
co
s
t
p
ar
am
eter
v
alu
es.
B
y
d
ef
au
lt,
p
ac
k
ag
e
e
1
0
7
1
p
r
o
v
i
d
es a
p
ar
am
eter
v
alu
e
o
f
co
s
t=1
.
Fig
u
r
e
2
.
Simu
latio
n
r
esu
lts
o
f
th
e
SVM
Ov
O
m
eth
o
d
: c
o
s
t p
ar
am
eter
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
14
,
No
.
6
,
Dec
em
b
er
2
0
2
5
:
4
8
7
8
-
4
8
9
0
4882
Fig
u
r
e
3
.
Simu
latio
n
r
esu
lts
o
f
th
e
SVM
Ov
O
m
eth
o
d
:
g
am
m
a
p
ar
am
eter
T
h
e
s
im
u
latio
n
r
esu
lts
o
f
s
ev
er
al
g
am
m
a
p
ar
am
eter
s
ettin
g
s
also
s
h
o
w
th
at
th
e
ac
cu
r
ac
y
o
f
th
e
Ov
O
SVM
m
o
d
el
in
cr
ea
s
es
as
th
e
g
am
m
a
p
ar
a
m
eter
in
cr
ea
s
es.
T
h
e
ac
cu
r
ac
y
o
f
th
e
O
v
O
SVM
m
o
d
el
r
ea
ch
es
o
p
tim
al
wh
en
g
am
m
a
=
2
4
.
B
ased
o
n
t
h
is
s
im
u
latio
n
,
it
ca
n
b
e
s
ee
n
th
at
th
e
p
e
r
f
o
r
m
an
ce
o
f
th
e
SVM
Ov
O
class
if
icatio
n
m
eth
o
d
is
v
er
y
d
ep
en
d
e
n
t
o
n
th
e
c
o
s
t
an
d
g
a
m
m
a
p
ar
am
ete
r
s
,
wh
er
ein
th
e
s
im
u
latio
n
ca
r
r
ied
o
u
t,
th
e
SVM
Ov
O
m
o
d
el
ac
h
iev
ed
th
e
b
est
ac
cu
r
ac
y
w
h
e
n
s
ettin
g
th
e
co
s
t
p
ar
a
m
eter
=
2
14
an
d
th
e
g
a
m
m
a
p
ar
am
eter
=
2
4
.
T
h
e
s
im
u
l
ati
o
n
r
esu
l
ts
s
h
o
w
t
h
at
t
h
e
k
a
p
p
a
,
la
m
b
d
a,
a
n
d
p
p
ar
am
ete
r
s
i
n
f
l
u
e
n
c
e
th
e
Ge
n
SV
M
m
o
d
el.
T
h
e
e
f
f
ec
t
o
f
t
h
e
k
a
p
p
a
p
a
r
a
m
et
e
r
o
n
t
h
e
p
e
r
f
o
r
m
a
n
c
e
o
f
t
h
e
Ge
n
SVM
m
e
th
o
d
is
p
r
e
s
en
t
e
d
in
F
ig
u
r
e
4
,
th
e
ef
f
ec
t
o
f
th
e
lam
b
d
a
p
ar
a
m
eter
o
n
th
e
p
er
f
o
r
m
a
n
ce
o
f
t
h
e
Gen
SVM
m
eth
o
d
is
p
r
esen
ted
in
Fig
u
r
e
5
,
an
d
th
e
ef
f
ec
t o
f
th
e
p
p
ar
am
eter
on
th
e
p
er
f
o
r
m
a
n
ce
o
f
th
e
Gen
SV
M
m
eth
o
d
ca
n
b
e
s
ee
n
in
F
ig
u
r
e
6
.
Fig
u
r
e
4
.
Simu
latio
n
r
esu
lts
o
f
th
e
Gen
SVM
m
eth
o
d
:
k
ap
p
a
Fig
u
r
e
5
.
Simu
latio
n
r
esu
lts
o
f
th
e
Gen
SVM
m
eth
o
d
: la
m
b
d
a
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
S
u
p
p
o
r
t v
ec
to
r
ma
ch
in
e
p
erfo
r
ma
n
ce
:
s
imu
la
tio
n
a
n
d
r
ice
p
h
en
o
lo
g
y
a
p
p
lica
tio
n
(
He
n
g
ki
Mu
r
a
d
i
)
4883
Fig
u
r
e
6
.
Simu
latio
n
r
esu
lts
o
f
th
e
Gen
SVM
m
eth
o
d
:
p
ar
am
e
ter
p
T
h
e
p
e
r
f
o
r
m
a
n
c
e
o
f
t
h
e
G
e
n
S
V
M
m
e
t
h
o
d
f
l
u
c
t
u
a
t
e
s
w
i
t
h
s
e
v
e
r
a
l
v
a
r
i
a
t
i
o
n
s
o
f
t
h
e
k
a
p
p
a
p
a
r
a
m
e
t
e
r
a
n
d
r
e
a
c
h
e
s
o
p
t
i
m
a
l
p
e
r
f
o
r
m
a
n
c
e
w
h
e
n
t
h
e
k
a
p
p
a
p
a
r
a
m
e
t
e
r
i
s
s
e
t
t
o
2
.
5
o
r
4
.
0
.
T
h
e
p
e
r
f
o
r
m
a
n
c
e
o
f
t
h
e
G
e
n
S
V
M
m
e
t
h
o
d
b
e
c
o
m
e
s
m
o
r
e
o
p
t
i
m
a
l
w
h
e
n
t
h
e
l
a
m
b
d
a
p
a
r
a
m
e
t
e
r
i
s
i
n
c
r
e
a
s
i
n
g
l
y
r
e
d
u
c
e
d
a
n
d
w
i
l
l
r
e
a
c
h
o
p
t
i
m
a
l
w
h
e
n
t
h
e
l
a
m
b
d
a
p
a
r
a
m
e
t
e
r
i
s
s
e
t
t
o
2
−
16
.
M
e
a
n
w
h
i
l
e
,
t
h
e
p
e
r
f
o
r
m
a
n
c
e
o
f
G
e
n
S
V
M
f
l
u
c
t
u
a
t
e
s
w
h
e
n
s
i
m
u
l
a
t
i
n
g
s
e
v
e
r
a
l
p
a
r
a
m
e
t
e
r
v
a
l
u
e
s
,
w
h
e
r
e
t
h
i
s
m
e
t
h
o
d
w
i
l
l
a
c
h
i
e
v
e
o
p
t
i
m
a
l
p
e
r
f
o
r
m
a
n
c
e
w
h
e
n
s
e
t
t
i
n
g
t
h
e
p
a
r
a
m
e
t
e
r
p
=
1
.
7
.
I
n
th
is
s
im
u
latio
n
,
s
ettin
g
th
e
p
ar
am
eter
s
o
f
t
h
e
SVM
Ov
O
an
d
Gen
SVM
m
o
d
els
g
r
ea
tly
i
m
p
ac
ts
th
e
p
er
f
o
r
m
an
ce
o
f
b
o
th
m
o
d
els.
Ho
wev
er
,
s
ettin
g
m
o
d
el
p
a
r
am
eter
s
s
u
ch
as
co
s
t
an
d
g
am
m
a
s
h
o
u
ld
also
co
n
s
id
er
th
e
im
p
ac
t
on
th
e
p
o
s
s
ib
ilit
y
o
f
o
v
e
r
f
itti
n
g
.
Fo
r
e
x
am
p
le,
wh
e
n
th
e
co
s
t=2
1
6
a
n
d
g
am
m
a
p
ar
a
m
eter
s
ar
e
d
ef
a
u
lt,
th
e
ac
cu
r
ac
y
o
f
th
e
SVM
Ov
O
m
o
d
el
o
n
tr
a
in
in
g
d
ata
in
c
r
ea
s
es
to
1
0
0
%.
Ho
wev
er
,
wh
en
p
r
ed
icted
o
n
test
in
g
d
ata,
th
e
ac
cu
r
ac
y
o
f
SVM
Ov
O
d
r
o
p
s
to
6
4
.
29
%.
Similar
l
y
,
wh
en
th
e
g
am
m
a
p
ar
am
eter
is
2
4
an
d
th
e
c
o
s
t
p
ar
am
eter
is
s
et
b
y
d
ef
au
lt,
th
e
m
o
d
el
'
s
ac
cu
r
ac
y
in
th
e
tr
ain
in
g
d
ata
r
ea
ch
es
1
0
0
%.
Ho
wev
er
,
wh
en
p
r
ed
icted
o
n
t
h
e
test
in
g
d
ata,
th
e
ac
c
u
r
ac
y
o
f
SVM
Ov
O
d
ec
r
ea
s
es to
5
3
.
57
%.
S
i
m
i
la
r
l
y
,
i
n
t
h
e
G
e
n
S
V
M
m
e
th
o
d
,
w
h
e
n
t
h
e
k
a
p
p
a
p
a
r
a
m
e
t
e
r
=
2
.
5
,
t
h
e
m
o
d
e
l
a
c
c
u
r
a
c
y
i
n
t
h
e
t
r
a
i
n
i
n
g
d
a
t
a
r
e
a
c
h
es
9
9
.
2
2
%
,
b
u
t
w
h
en
p
r
e
d
i
c
t
i
o
n
s
a
r
e
m
a
d
e
o
n
t
h
e
t
e
s
ti
n
g
d
a
t
a
,
t
h
e
a
cc
u
r
a
c
y
o
f
th
e
G
e
n
SV
M
m
o
d
el
d
r
o
p
s
t
o
5
5
.
3
6
%
.
T
h
i
s
s
h
o
w
s
th
a
t
t
h
e
SV
M
O
v
O
m
e
t
h
o
d
i
s
s
e
n
s
i
ti
v
e
t
o
c
h
a
n
g
e
s
i
n
c
o
s
t
a
n
d
g
a
m
m
a
p
a
r
a
m
e
t
e
r
v
a
l
u
e
s
.
T
h
e
G
e
n
S
V
M
m
e
t
h
o
d
a
l
s
o
d
e
p
e
n
d
s
o
n
s
et
t
i
n
g
k
a
p
p
a
,
l
a
m
b
d
a
,
a
n
d
p
p
a
r
a
m
e
t
e
r
s
.
I
n
o
t
h
e
r
w
o
r
d
s
,
t
h
e
S
V
M
O
v
O
a
n
d
G
e
n
S
V
M
m
a
ch
i
n
e
l
e
a
r
n
i
n
g
m
et
h
o
d
s
a
r
e
s
e
n
s
i
t
i
v
e
t
o
t
h
e
s
el
e
c
ti
o
n
o
f
m
o
d
e
l
p
a
r
a
m
e
t
e
r
s
.
F
o
r
t
h
is
r
e
a
s
o
n
,
t
h
e
h
y
p
e
r
p
a
r
a
m
e
t
e
r
p
r
o
c
e
s
s
is
es
s
e
n
t
i
a
l
.
H
y
p
e
r
p
a
r
a
m
e
t
e
r
s
a
r
e
k
e
y
f
a
c
t
o
r
s
i
n
d
e
v
e
l
o
p
i
n
g
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
m
o
d
e
l
s
,
i
n
cl
u
d
i
n
g
S
VM
a
n
d
G
e
n
S
VM
,
b
e
c
a
u
s
e
t
h
ey
d
e
t
e
r
m
i
n
e
t
h
e
m
o
d
e
l'
s
b
e
h
a
v
i
o
r
b
e
f
o
r
e
t
r
a
i
n
i
n
g
.
S
e
l
e
ct
i
n
g
t
h
e
c
o
r
r
e
c
t
h
y
p
e
r
p
a
r
a
m
e
t
e
r
s
c
a
n
p
r
e
v
e
n
t
o
v
e
r
f
i
t
t
i
n
g
,
i
m
p
r
o
v
e
p
e
r
f
o
r
m
a
n
c
e
o
n
t
e
s
t
d
a
t
a
,
a
n
d
m
a
k
e
t
h
e
m
o
d
e
l
m
o
r
e
s
t
a
b
l
e
.
H
y
p
e
r
p
a
r
am
e
t
e
r
t
u
n
i
n
g
c
a
n
b
e
d
o
n
e
u
s
i
n
g
t
h
e
t
u
n
e
f
u
n
c
t
i
o
n
i
n
t
h
e
e
1
0
7
1
p
a
c
k
a
g
e
[
2
5
]
.
3.
3
.
E
m
p
i
rica
l a
pp
lica
t
io
n
I
n
th
e
em
p
i
r
ical
ap
p
licatio
n
s
tu
d
y
,
th
e
m
o
d
elin
g
o
f
r
ice
p
h
en
o
lo
g
y
u
s
in
g
s
en
tin
el
-
1
s
atellite
im
ag
e
d
ata
was
test
ed
.
T
h
e
s
tep
s
o
f
t
h
e
em
p
ir
ical
s
tu
d
y
ar
e
p
r
esen
t
ed
in
Fig
u
r
e
7
.
I
n
Fig
u
r
e
7
,
it
ca
n
b
e
s
ee
n
t
h
at
in
th
e
in
itial
s
tag
e,
we
ex
tr
ac
ted
s
en
tin
el
-
1
im
ag
e
d
ata
to
p
r
o
d
u
ce
VV
an
d
VH
p
o
lar
izatio
n
an
d
th
eir
d
e
r
iv
ativ
es,
n
am
ely
R
PI,
NDPI
,
an
d
API
.
W
e
ad
ju
s
ted
th
is
d
ata
with
f
ield
s
u
r
v
ey
d
ata
i
n
th
e
f
o
r
m
o
f
r
ice
g
r
o
wth
p
h
ase
in
f
o
r
m
atio
n
to
p
r
o
d
u
ce
tab
u
lat
ed
d
ata.
W
e
ch
ec
k
e
d
th
e
tab
u
l
ated
d
ata
t
o
en
s
u
r
e
n
o
m
is
s
in
g
v
alu
es
o
r
e
x
tr
em
e
o
u
tlier
s
.
W
e
co
n
d
u
cted
a
r
ice
p
h
en
o
l
o
g
y
r
ec
lass
if
icatio
n
s
ce
n
ar
io
to
o
b
tain
th
e
o
p
tim
al
n
u
m
b
er
o
f
r
ice
class
es
in
th
e
r
ice
p
h
e
n
o
lo
g
y
m
o
d
el.
T
en
r
ep
etitio
n
s
o
f
th
e
s
ce
n
ar
i
o
wer
e
p
er
f
o
r
m
ed
o
n
th
e
tr
ai
n
in
g
an
d
test
in
g
d
ata
to
o
b
s
er
v
e
th
e
co
n
s
is
ten
cy
o
f
th
e
SVM
m
o
d
el's
p
er
f
o
r
m
an
ce
.
T
h
e
Ov
O
SVM,
Gen
SVM
,
an
d
ML
R
m
o
d
els
wer
e
tr
ain
ed
u
s
in
g
th
e
tr
ain
in
g
d
ata.
Par
am
eter
an
d
k
er
n
el
t
u
n
in
g
wer
e
also
p
er
f
o
r
m
ed
o
n
th
e
Ov
O
SVM
an
d
Gen
SVM
m
o
d
els
to
o
b
tain
th
e
o
p
tim
al
p
ar
am
eter
s
an
d
k
er
n
els.
T
h
e
m
o
d
el'
s
p
er
f
o
r
m
an
ce
o
n
th
e
tr
ain
in
g
an
d
test
in
g
d
ata
was m
ea
s
u
r
ed
u
s
in
g
OA
an
d
k
ap
p
a
s
tatis
tics
.
3.
3
.
1
.
Resea
rc
h da
t
a
T
h
e
r
esear
ch
d
ata
wer
e
f
r
o
m
th
e
r
ice
p
h
ase
team
at
th
e
r
em
o
te
s
en
s
in
g
r
esear
ch
c
en
ter
o
f
th
e
I
n
d
o
n
esian
Natio
n
al
R
esear
ch
an
d
I
n
n
o
v
atio
n
Ag
en
c
y
.
S
en
tin
el
-
1
s
atellite
d
ata
wer
e
ex
tr
ac
ted
u
s
in
g
th
e
Go
o
g
le
E
ar
th
E
n
g
in
e
p
latf
o
r
m
,
wh
ile
f
ield
d
ata
wer
e
o
b
ta
in
ed
f
r
o
m
s
u
r
v
e
y
ac
tiv
ities
co
n
d
u
cte
d
in
th
e
r
ice
f
ield
s
o
f
PT.
San
g
Hy
an
g
Ser
i
is
in
th
e
ad
m
in
is
tr
ativ
e
ar
ea
o
f
Su
b
an
g
R
eg
en
cy
,
W
est
J
av
a
Pro
v
in
ce
,
I
n
d
o
n
esia
Fig
u
r
e
8
.
Sen
tin
el
-
1
SAR
GR
D
im
ag
e
d
ata
with
ac
q
u
is
itio
n
m
o
d
es
in
ter
f
e
r
o
m
etr
ic
wid
e
s
wath
(
I
W
)
ac
ce
s
s
ed
d
u
r
i
n
g
t
h
e
f
ir
s
t
p
lan
tin
g
s
ea
s
o
n
o
f
2
0
2
1
-
2
0
2
2
f
r
o
m
th
e
Go
o
g
le
E
ar
t
h
E
n
g
in
e
p
latf
o
r
m
.
Sen
tin
el
-
1
d
ata
h
as
g
o
n
e
th
r
o
u
g
h
th
e
p
r
ep
r
o
ce
s
s
in
g
s
tag
e
o
f
th
er
m
al
n
o
is
e
r
em
o
v
al,
r
a
d
io
m
etr
ic
ca
lib
r
atio
n
,
an
d
ter
r
ain
co
r
r
ec
tio
n
to
b
e
r
e
ad
y
f
o
r
u
s
e
[
3
2
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
14
,
No
.
6
,
Dec
em
b
er
2
0
2
5
:
4
8
7
8
-
4
8
9
0
4884
E
x
tr
ac
tio
n
o
f
s
en
tin
el
-
1
im
a
g
e
d
ata
p
r
o
d
u
ce
s
p
o
lar
izatio
n
VV,
VH,
an
d
a
p
o
lar
izat
io
n
in
d
ex
VH/VV,
ca
lled
th
e
R
PI
,
NDPI
,
an
d
API
.
T
h
e
p
o
lar
izatio
n
v
alu
es
VV,
VH,
R
PI,
NDPI
,
an
d
API
p
o
la
r
izatio
n
in
d
ex
ar
e
i
n
ten
s
ities
r
an
g
in
g
f
r
o
m
0
to
1
.
All a
r
e
p
r
ed
icto
r
s
o
f
th
e
r
ice
p
h
en
o
l
o
g
y
m
o
d
el.
Fig
u
r
e
7
.
Step
s
o
f
em
p
ir
ical
s
tu
d
y
o
f
SVM,
Gen
SVM,
an
d
ML
R
m
eth
o
d
s
Fig
u
r
e
8
.
R
esear
ch
a
r
ea
3.
3
.
2
.
Rice
ph
eno
lo
g
y
mo
del
T
h
e
r
i
c
e
p
h
e
n
o
l
o
g
y
c
l
a
s
s
e
s
u
s
ed
i
n
t
h
i
s
s
t
u
d
y
c
o
n
s
i
s
te
d
o
f
7
c
l
a
s
s
e
s
,
n
a
m
e
l
y
t
h
e
w
a
t
e
r
p
h
a
s
e
(
<
0
D
A
P
)
,
ea
r
ly
v
eg
etativ
e
p
h
ase
(
0
-
2
0
DAP)
,
v
eg
etativ
e
-
1
p
h
ase
(
2
1
-
4
0
DAP)
,
v
eg
etativ
e
p
h
a
s
e
-
2
(
4
1
-
6
4
DAP)
,
g
en
er
ativ
e
-
1
p
h
ase
(
6
5
-
9
0
D
AP)
,
g
en
er
ativ
e
-
2
p
h
ase
(
9
1
-
1
2
0
DAP)
,
a
n
d
f
allo
w
p
h
ase
(
>1
2
0
DAP)
.
R
ec
lu
s
e
s
ce
n
ar
io
s
wer
e
also
test
ed
o
n
th
e
m
o
d
el
to
ac
c
o
m
m
o
d
ate
p
o
s
s
ib
le
ch
an
g
es
in
m
o
d
el
ac
cu
r
ac
y
[
1
5
]
.
T
h
e
r
ec
lass
s
ce
n
ar
io
co
n
s
is
t
s
o
f
7
-
class
s
ce
n
ar
io
,
6
-
class
s
ce
n
ar
io
,
5
-
class
s
ce
n
ar
io
,
an
d
4
-
class
s
ce
n
ar
io
T
ab
le
1
.
E
ac
h
s
ce
n
ar
io
r
ep
r
esen
ts
a
d
if
f
er
en
t
lev
el
o
f
ag
g
r
eg
atio
n
,
wh
er
e
class
es
th
at
h
av
e
s
im
ilar
i
ties
o
r
r
elev
an
ce
ar
e
co
m
b
in
ed
to
s
im
p
lify
th
e
class
if
icatio
n
p
r
o
b
lem
.
T
h
is
ex
p
er
im
en
t
aim
s
to
id
en
tify
th
e
n
u
m
b
er
o
f
class
es
th
at
p
r
o
v
id
e
t
h
e
b
est
p
er
f
o
r
m
an
ce
in
th
e
SVM
m
o
d
el,
co
n
s
id
er
in
g
th
e
b
alan
ce
b
etwe
en
m
o
d
el
co
m
p
lex
ity
an
d
g
en
er
aliza
tio
n
a
b
ilit
y
.
Selectin
g
th
e
co
r
r
ec
t
n
u
m
b
er
o
f
class
es
ca
n
h
elp
r
ed
u
ce
o
v
er
f
itti
n
g
a
n
d
im
p
r
o
v
e
m
o
d
el
in
ter
p
r
etab
ilit
y
,
r
esu
ltin
g
in
cl
ass
if
icatio
n
r
esu
lts
th
at
ar
e
m
o
r
e
r
eliab
le
an
d
ea
s
ier
f
o
r
e
n
d
u
s
er
s
to
u
n
d
er
s
tan
d
.
T
h
e
r
ice
p
h
en
o
l
o
g
y
m
o
d
el
was
co
n
s
tr
u
cted
u
s
in
g
VV,
VH,
R
PI,
NDPI
,
an
d
API
p
r
e
d
icto
r
s
.
Sev
er
al
m
o
d
ellin
g
s
ch
e
m
es
co
n
s
id
er
e
d
th
e
n
u
m
b
er
o
f
r
ice
p
h
e
n
o
lo
g
y
class
es
an
d
th
e
n
u
m
b
er
o
f
m
o
d
el
p
r
ed
icto
r
s
.
I
n
th
e
r
ice
p
h
en
o
lo
g
y
m
o
d
el
with
o
n
e
p
r
ed
icto
r
,
t
h
e
p
e
r
f
o
r
m
an
ce
o
f
VH
p
o
la
r
izatio
n
[
4
]
,
[
1
7
]
an
d
R
PI
p
o
lar
izatio
n
in
d
ex
was
test
ed
[
2
0
]
.
I
n
class
if
icatio
n
m
o
d
e
ls
with
two
p
r
ed
icto
r
s
,
t
h
e
p
er
f
o
r
m
a
n
ce
o
f
t
h
e
VV+
VH
an
d
VH+
R
PI
p
r
ed
i
cto
r
s
h
as
b
ee
n
test
ed
.
I
n
cl
ass
if
icatio
n
m
o
d
els
with
th
r
ee
p
r
ed
icto
r
s
,
th
e
p
er
f
o
r
m
an
ce
o
f
th
e
VV+
VH+
R
PI
p
r
ed
icto
r
s
h
as b
ee
n
test
ed
[
3
3
]
.
T
h
e
r
is
k
o
f
o
v
e
r
f
itti
n
g
p
r
o
b
l
em
s
in
th
e
m
o
d
el
was
test
ed
b
y
d
iv
i
d
in
g
th
e
s
am
p
le
in
to
7
0
%
f
o
r
tr
ain
in
g
an
d
3
0
%
f
o
r
test
in
g
.
Mo
d
el
ac
cu
r
ac
y
is
m
ea
s
u
r
ed
f
r
o
m
tr
ai
n
in
g
a
n
d
test
in
g
d
ata
to
s
h
o
w
th
e
d
if
f
er
en
ce
b
etwe
en
th
e
esti
m
ated
OA
an
d
k
ap
p
a
p
a
r
am
eter
s
.
T
h
e
s
tab
ilit
y
o
f
th
e
m
o
d
el
is
t
ested
b
y
d
o
i
n
g
ten
r
ep
etitio
n
s
o
f
tr
ain
in
g
an
d
te
n
r
ep
etitio
n
s
o
f
test
in
g
.
Data
p
r
o
ce
s
s
in
g
u
s
es
th
e
p
ac
k
ag
es
ca
r
et
[
3
4
]
,
Ge
n
SVM
p
ac
k
ag
es
[
2
6
]
,
a
n
d
n
n
et
[
3
5
]
i
n
R
p
r
o
g
r
am
m
in
g
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
S
u
p
p
o
r
t v
ec
to
r
ma
ch
in
e
p
erfo
r
ma
n
ce
:
s
imu
la
tio
n
a
n
d
r
ice
p
h
en
o
lo
g
y
a
p
p
lica
tio
n
(
He
n
g
ki
Mu
r
a
d
i
)
4885
T
ab
le
1.
Scen
ar
io
o
f
m
o
d
ellin
g
M
o
d
e
l
l
i
n
g
sce
n
a
r
i
o
P
h
e
n
o
l
o
g
y
c
l
a
s
s (Y
)
D
a
y
s
a
f
t
e
r
p
l
a
n
t
i
n
g
(
D
A
P
)
o
f
p
a
d
d
y
7
-
c
l
a
ss
Y=
{
1
,
W
a
t
e
r
2
,
Ea
r
l
y
-
v
e
g
e
t
a
t
i
v
e
3
,
V
e
g
e
t
a
t
i
v
e
-
1
4
,
V
e
g
e
t
a
t
i
v
e
-
2
5
,
G
e
n
e
r
a
t
i
v
e
-
1
6
,
G
e
n
e
r
a
t
i
v
e
-
2
7
,
F
a
l
l
o
w
1.
<
0
D
A
P
2.
0
-
2
0
D
A
P
3.
21
-
4
0
D
A
P
4.
41
-
6
4
D
A
P
5.
65
-
9
0
D
A
P
6.
91
-
1
2
0
D
A
P
7.
>
1
2
0
D
A
P
6
-
c
l
a
ss
Y=
{
1
,
W
a
t
e
r
2
,
V
e
g
e
t
a
t
i
v
e
-
1
3
,
V
e
g
e
t
a
t
i
v
e
-
2
4
,
G
e
n
e
r
a
t
i
v
e
-
1
5
,
G
e
n
e
r
a
t
i
v
e
-
2
6
,
F
a
l
l
o
w
1.
<
0
D
A
P
2.
0
-
4
0
D
A
P
3.
41
-
6
4
D
A
P
4.
65
-
9
0
D
A
P
5.
91
-
1
2
0
D
A
P
6.
>
1
2
0
D
A
P
5
-
c
l
a
ss
Y=
{
1
,
W
a
t
e
r
2
,
V
e
g
e
t
a
t
i
v
e
-
1
3
,
V
e
g
e
t
a
t
i
v
e
-
2
4
,
G
e
n
e
r
a
t
i
v
e
-
1
5
,
G
e
n
e
r
a
t
i
v
e
-
2
1.
<
0
D
A
P
2.
0
-
4
0
D
A
P
3.
41
-
6
4
D
A
P
4.
65
-
9
0
D
A
P
5.
91
-
1
2
0
D
A
P
4
-
c
l
a
ss
Y=
{
1
,
V
e
g
e
t
a
t
i
v
e
-
1
2
,
V
e
g
e
t
a
t
i
v
e
-
2
3
,
G
e
n
e
r
a
t
i
v
e
-
1
4
,
G
e
n
e
r
a
t
i
v
e
-
2
1.
0
-
4
0
D
A
P
2.
41
-
6
4
D
A
P
3.
65
-
9
0
D
A
P
4.
91
-
1
2
0
D
A
P
4.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
Up
p
er
ex
t
r
em
e
d
ata
is
d
etec
te
d
in
b
o
th
d
atasets
,
wh
ich
o
cc
u
r
r
ed
in
th
e
f
allo
w
p
h
ase.
T
h
is
ex
tr
em
e
d
ata
is
r
ed
u
ce
d
wh
e
n
th
e
d
at
a
is
r
ed
u
ce
d
to
t
h
e
R
PI
an
d
NDPI
p
o
lar
izatio
n
in
d
ices.
F
ig
u
r
e
9
s
h
o
ws
th
e
b
o
x
p
lo
ts
o
f
p
o
lar
izatio
n
p
ar
a
m
eter
s
an
d
in
d
ices
ac
r
o
s
s
r
ice
g
r
o
wth
p
h
ases
.
T
h
is
ex
tr
em
e
d
ata
is
s
till
v
is
ib
le
in
th
e
API
Fig
u
r
e
9
(
a
)
.
I
n
Fig
u
r
e
9
(
b
)
,
it
ca
n
b
e
s
ee
n
th
at
th
e
VV
p
o
lar
izatio
n
f
lu
ctu
ates
f
r
o
m
th
e
wate
r
p
h
ase
to
th
e
f
allo
w
p
h
ase.
T
h
e
s
ca
tter
b
o
x
p
lo
t
s
h
o
ws
th
at
VH
p
o
lar
izatio
n
h
as
a
p
o
s
itiv
e
tr
en
d
f
r
o
m
th
e
ea
r
ly
v
eg
etativ
e
p
h
ase
to
th
e
f
allo
w
p
h
ase
Fig
u
r
e
9
(
c
)
.
T
h
er
ef
o
r
e,
VH
p
o
lar
izatio
n
co
n
s
is
ten
tly
in
cr
ea
s
es
tr
en
d
s
f
r
o
m
t
h
e
ea
r
l
y
v
e
g
etativ
e
p
h
ase
to
th
e
ea
r
ly
r
ip
en
i
n
g
p
h
ase
a
f
ter
th
e
p
lan
t r
ea
ch
es
its
m
atu
r
ity
p
h
ase
[
3
6
]
.
T
h
e
R
PI
p
o
lar
izatio
n
i
n
d
ex
h
as
a
u
n
iq
u
e
p
atter
n
an
d
ca
n
d
escr
i
b
e
th
e
p
atter
n
o
f
r
ice
g
r
o
wth
p
h
ases
Fig
u
r
e
9
(
d
)
.
Ho
wev
er
,
th
e
VH
p
o
lar
izatio
n
h
as
b
etter
ac
cu
r
ac
y
in
d
iv
id
u
ally
th
an
th
e
R
PI
p
o
lar
izatio
n
in
d
ex
.
Me
an
wh
ile,
th
e
NDPI
p
o
lar
izatio
n
i
n
d
ex
h
as
th
e
o
p
p
o
s
ite
p
atter
n
to
R
PI
Fig
u
r
e
9
(
e
)
,
wh
ich
ca
n
b
e
u
s
ed
to
d
escr
ib
e
t
h
e
q
u
an
tity
o
f
wate
r
in
r
ice
f
ield
s
.
Fin
ally
,
th
e
API
p
o
lar
izatio
n
in
d
ex
ap
p
ea
r
s
to
f
lu
ctu
ate,
s
im
ilar
to
th
e
VV
p
o
lar
izatio
n
Fig
u
r
e
9
(
f
)
.
T
h
e
p
er
f
o
r
m
a
n
ce
o
f
th
e
th
r
ee
m
eth
o
d
s
in
c
r
ea
s
es
as
th
e
n
u
m
b
er
o
f
class
es
in
th
e
r
ice
p
h
en
o
lo
g
y
is
r
ed
u
ce
d
a
n
d
r
ea
ch
es
o
p
tim
al
p
er
f
o
r
m
a
n
ce
in
th
e
4
-
class
s
ce
n
ar
io
.
No
n
-
r
ice
class
es,
s
u
ch
as
wate
r
an
d
f
allo
w,
s
till
n
ee
d
to
b
e
class
if
ied
co
r
r
ec
tly
b
y
th
e
th
r
ee
m
eth
o
d
s
.
T
h
is
co
u
ld
b
e
b
ec
au
s
e
th
er
e
a
r
e
v
er
y
f
ew
m
em
b
er
s
o
f
b
o
th
class
es
co
m
p
ar
ed
t
o
m
em
b
er
s
o
f
o
th
e
r
r
ice
clas
s
es.
T
h
er
ef
o
r
e,
ad
d
r
ess
in
g
t
h
e
class
im
b
alan
ce
p
r
o
b
lem
ca
n
im
p
r
o
v
e
th
e
m
o
d
el'
s
ac
cu
r
ac
y
[
3
7
]
,
[
3
8
]
.
T
a
b
le
2
p
r
esen
ts
th
e
r
esu
lts
o
f
m
o
d
ellin
g
th
e
g
r
o
wth
p
h
ase
o
f
r
ice
u
s
in
g
th
e
SVM
Ov
O,
Gen
SVM,
an
d
ML
R
m
e
th
o
d
s
.
T
h
e
SVM
Ov
O
m
eth
o
d
in
th
e
in
itial
s
ce
n
ar
io
,
n
a
m
ely
t
h
e
7
-
class
r
ice
p
h
ase,
ac
h
iev
es
o
p
tim
al
p
er
f
o
r
m
an
ce
a
n
d
,
at
th
e
s
am
e
tim
e,
d
o
es
n
o
t
e
x
p
er
ien
ce
o
v
e
r
f
itti
n
g
p
r
o
b
lem
s
wh
e
n
th
e
p
a
r
am
eter
s
ar
e
s
et
to
co
s
t=2
8
an
d
g
a
m
m
a=
1
/6
.
I
n
t
h
e
4
-
class
r
ice
p
h
en
o
lo
g
y
s
ce
n
ar
io
,
th
e
SVM
Ov
O
m
eth
o
d
ac
h
iev
ed
o
p
tim
al
p
er
f
o
r
m
an
ce
with
th
e
p
ar
a
m
eter
s
ettin
g
s
o
f
co
s
t=2
8
a
n
d
g
am
m
a=
1
/2
.
T
h
is
r
esu
lt
in
d
icate
s
th
at
th
e
p
er
f
o
r
m
an
ce
o
f
SVM
Ov
O
is
h
ig
h
ly
d
ep
e
n
d
en
t
o
n
p
ar
am
ete
r
s
elec
tio
n
,
th
u
s
r
eq
u
ir
in
g
ca
r
ef
u
l
tu
n
in
g
in
ea
ch
class
s
ce
n
ar
io
.
T
h
e
Gen
SVM
m
eth
o
d
ac
h
iev
es
o
p
tim
al
p
er
f
o
r
m
a
n
ce
wh
e
n
t
h
e
p
ar
am
eter
s
k
ap
p
a=
0
.
2
,
lam
b
d
a=
2
-
3
6
,
a
n
d
p
=
1
.
T
h
is
m
eth
o
d
d
o
es n
o
t
r
eq
u
ir
e
d
if
f
er
en
t settin
g
s
f
o
r
all
r
ice
p
h
en
o
l
o
g
y
class
s
ce
n
ar
io
s
.
I
n
th
e
c
a
s
e
o
f
t
h
e
r
i
c
e
p
h
e
n
o
l
o
g
y
c
la
s
s
i
f
i
c
a
t
i
o
n
m
o
d
e
l
,
th
e
a
c
cu
r
ac
y
o
f
t
h
e
c
l
a
s
s
i
f
i
c
at
i
o
n
m
o
d
e
l
f
r
o
m
th
e
S
V
M
O
v
O
,
G
en
S
VM
,
an
d
M
L
R
m
e
th
o
d
s
w
i
l
l
b
e
o
p
t
i
m
a
l
wh
e
n
a
l
l
p
r
ed
i
c
t
o
r
s
a
r
e
in
c
l
u
d
ed
in
t
h
e
m
o
d
e
l
.
T
h
e
r
e
s
u
l
t
s
i
n
T
ab
le
2
s
h
o
w
t
h
a
t
t
h
e
O
v
O
SV
M
m
e
t
h
o
d
c
o
n
s
i
s
t
e
n
t
l
y
p
r
o
v
i
d
e
s
t
h
e
h
i
g
h
e
s
t
p
e
r
f
o
r
m
a
n
c
e
co
m
p
a
r
e
d
to
Ge
n
S
V
M
an
d
M
L
R
i
n
a
l
l
c
l
as
s
n
u
m
b
e
r
s
c
e
n
a
r
i
o
s
.
I
n
th
e
4
-
c
l
a
s
s
ca
s
e
,
S
V
M
a
c
h
i
ev
e
d
an
O
A
o
f
7
9
.
2
0
%
o
n
t
h
e
tr
a
i
n
in
g
d
a
t
a
a
n
d
7
8
.
5
7
%
o
n
th
e
t
e
s
t
i
n
g
d
a
t
a
,
w
i
th
k
a
p
p
a
v
a
lu
e
s
o
f
7
0
.
6
9
%
an
d
6
9
.
8
3
%
,
r
e
s
p
e
c
t
iv
e
l
y
.
T
h
e
s
e
a
c
cu
r
ac
i
e
s
a
r
e
m
u
c
h
h
i
g
h
e
r
t
h
an
M
L
R
's
,
w
h
i
ch
o
n
l
y
a
c
h
i
ev
ed
an
O
A
o
f
7
6
.
8
6
%
o
n
t
h
e
tr
a
i
n
in
g
d
a
t
a
a
n
d
7
7
.
2
4
%
o
n
th
e
t
es
t
i
n
g
d
a
t
a.
Me
a
n
wh
i
l
e
,
G
en
S
V
M
s
h
o
we
d
q
u
i
t
e
c
o
m
p
e
t
i
t
i
v
e
p
er
f
o
r
m
a
n
c
e,
an
d
i
n
s
o
m
e
c
a
s
e
s
,
th
e
r
e
s
u
l
t
s
w
e
r
e
c
lo
s
e
t
o
t
h
o
s
e
o
f
S
V
M
,
a
l
t
h
o
u
g
h
g
e
n
e
r
a
l
ly
s
t
i
l
l
s
l
i
g
h
t
l
y
b
e
l
o
w
.
T
h
e
s
e
f
in
d
i
n
g
s
co
n
f
i
r
m
t
h
a
t
t
h
e
S
VM
O
v
O
a
p
p
r
o
a
c
h
c
an
b
e
t
t
e
r
a
c
c
o
m
m
o
d
a
t
e
t
h
e
c
o
m
p
l
e
x
i
ty
o
f
d
a
t
a
w
i
t
h
n
o
n
-
l
i
n
e
a
r
r
e
l
a
t
io
n
s
h
ip
s
,
wh
i
l
e
ML
R
i
s
l
i
m
i
t
e
d
to
l
i
n
ea
r
r
e
l
a
t
io
n
s
h
i
p
s
,
r
e
s
u
l
t
in
g
in
l
o
w
e
r
p
e
r
f
o
r
m
an
c
e.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
14
,
No
.
6
,
Dec
em
b
er
2
0
2
5
:
4
8
7
8
-
4
8
9
0
4886
(
a)
(
b
)
(
c)
(
d
)
(
e)
(f)
Fig
u
r
e
9
.
T
h
e
b
o
x
p
lo
t
o
f
(
a
)
V
V,
VH,
an
d
R
PI,
(
b
)
VV
b
ased
o
n
r
ice
p
h
ase,
(
c
)
VH
b
ased
o
n
r
ice
p
h
ase,
(
d
)
R
PI
b
ased
o
n
r
ice
p
h
ase,
(
e)
NDPI
b
ased
o
n
r
ice
p
h
ase,
an
d
(
f
)
API
b
ased
o
n
r
ice
p
h
ase
T
ab
le
2
.
R
esu
lt o
f
m
o
d
ellin
g
M
o
d
e
l
l
i
n
g
s
c
e
n
a
r
i
o
M
e
t
h
o
d
Tr
a
i
n
i
n
g
Te
st
i
n
g
OA
K
a
p
p
a
OA
K
a
p
p
a
7
-
c
l
a
ss
S
V
M
O
v
O
(
C
V
=
1
0
,
c
o
s
t
=
2
8
,
g
a
mm
a
=
1
/
6
)
6
7
.
3
7
±
0
.
1
5
5
8
.
7
5
±
0
.
33
6
6
.
5
6
±
0
.
37
5
7
.
3
5
±
0
.
77
G
e
n
S
V
M
(
k
a
p
p
a
=
0
.
2
,
l
a
mb
d
a
=
2
-
36
,
p
=
1
)
6
5
.
8
4
±
0
.
17
5
6
.
3
7
±
0
.
19
6
5
.
5
6
±
0
.
41
5
6
.
1
2
±
0
.
31
M
L
R
6
0
.
7
8
±
0
.
1
8
5
0
.
1
2
±
0
.
31
6
0
.
2
0
±
0
.
28
4
9
.
3
6
±
0
.
35
6
-
c
l
a
ss
S
V
M
O
v
O
(
C
V
=
1
0
,
c
o
s
t
=
2
6
,
g
a
mm
a
=
1
)
7
0
.
6
0
±
0
.
09
5
8
.
5
2
±
0
.
24
6
9
.
3
0
±
0
.
22
5
5
.
4
6
±
0
.
62
G
e
n
S
V
M
(
k
a
p
p
a
=
0
.
2
,
l
a
m
b
d
a
=
2
-
36
,
p
=
1
)
7
0
.
3
7
±
0
.
16
5
7
.
3
7
±
0
.
22
6
9
.
8
7
±
0
.
35
5
6
.
2
3
±
0
.
49
M
L
R
6
5
.
4
8
±
0
.
12
5
0
.
2
9
±
0
.
19
6
4
.
9
7
±
0
.
22
4
9
.
1
4
±
0
.
35
5
-
c
l
a
ss
S
V
M
O
v
O
(
C
V
=
1
0
,
c
o
s
t
=
2
2
,
g
a
mm
a
=
1
/
6
)
7
1
.
5
4
±
0
.
08
6
1
.
4
0
±
0
.
12
7
1
.
1
3
±
0
.
18
6
0
.
8
7
±
0
.
23
G
e
n
S
V
M
(
k
a
p
p
a
=
0
.
2
,
l
a
m
b
d
a
=
2
-
36
,
p
=
1
)
7
0
.
9
4
±
0
.
1
4
6
0
.
7
3
±
0
.
20
7
0
.
4
8
±
0
.
23
5
9
.
9
1
±
0
.
32
M
L
R
6
6
.
2
4
±
0
.
0
8
5
2
.
9
8
±
0
.
11
6
5
.
9
8
±
0
.
23
5
2
.
4
5
±
0
.
32
4
-
c
l
a
ss
S
V
M
O
v
O
(
C
V
=
1
0
,
c
o
s
t
=2
8
,
g
a
mm
a
=
1
/
2
)
7
9
.
2
0
±
0
.
21
7
0
.
6
9
±
0
.
2
9
7
8
.
5
7
±
0
.
43
6
9
.
8
3
±
0
.
5
9
G
e
n
S
V
M
(
k
a
p
p
a
=
0
.
2
,
l
a
m
b
d
a
=
2
-
36
,
p
=
1
)
7
8
.
0
1
±
0
.
15
6
8
.
9
6
±
0
.
2
1
7
8
.
0
1
±
0
.
31
6
9
.
1
2
±
0
.
4
6
M
L
R
7
6
.
8
6
±
0
.
1
7
6
7
.
3
4
±
0
.
2
3
7
7
.
2
4
±
0
.
4
0
6
9
.
9
3
±
0
.
5
4
T
h
is
s
tu
d
y
also
o
b
s
er
v
ed
th
e
p
o
s
s
ib
ilit
y
o
f
o
v
er
f
itti
n
g
in
th
e
m
o
d
el.
T
e
n
r
ep
licatio
n
s
wer
e
p
er
f
o
r
m
ed
o
n
b
o
th
th
e
tr
ain
in
g
an
d
test
d
ata
to
ass
ess
th
is
.
T
h
e
an
aly
s
is
s
h
o
wed
th
at
th
e
p
er
f
o
r
m
a
n
ce
d
if
f
er
en
ce
b
etwe
en
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
S
u
p
p
o
r
t v
ec
to
r
ma
ch
in
e
p
erfo
r
ma
n
ce
:
s
imu
la
tio
n
a
n
d
r
ice
p
h
en
o
lo
g
y
a
p
p
lica
tio
n
(
He
n
g
ki
Mu
r
a
d
i
)
4887
th
e
tr
ain
in
g
a
n
d
test
d
ata
wa
s
r
elativ
ely
s
m
all
f
o
r
all
th
r
e
e
m
eth
o
d
s
,
n
a
m
ely
SVM,
Gen
SVM,
an
d
ML
R
.
T
h
is
co
n
d
itio
n
in
d
icate
s
th
at
t
h
e
m
o
d
el
d
id
n
o
t
e
x
p
er
ie
n
ce
s
ig
n
if
ican
t
o
v
er
f
itti
n
g
an
d
co
u
l
d
g
en
e
r
alize
th
e
test
d
ata
well.
I
n
ad
d
itio
n
,
t
h
e
r
elativ
ely
s
m
all
s
tan
d
ar
d
er
r
o
r
v
alu
e
o
f
th
e
a
v
er
ag
e
r
ep
et
itio
n
r
esu
lts
(
<1
%
)
in
d
icate
s
th
e
s
tab
ilit
y
o
f
th
e
m
o
d
el'
s
p
er
f
o
r
m
an
ce
i
n
ten
r
ep
etitio
n
s
.
T
h
is
s
tab
ilit
y
p
r
o
v
id
es
ad
d
itio
n
al
ev
id
en
ce
th
at
th
e
p
er
f
o
r
m
a
n
ce
o
f
SVM,
esp
ec
ially
with
th
e
Ov
O
ap
p
r
o
ac
h
,
is
q
u
ite
co
n
s
is
ten
t
an
d
r
eliab
le
in
m
o
d
ellin
g
h
ig
h
l
y
co
m
p
lex
d
a
ta
s
u
ch
as
th
e
r
ice
g
r
o
wth
p
h
ase.
Ho
wev
er
,
th
e
ac
cu
r
ac
y
o
f
th
e
Ov
O
SVM
m
o
d
el
s
till
n
ee
d
s
to
b
e
im
p
r
o
v
e
d
th
r
o
u
g
h
f
u
r
th
er
o
p
ti
m
izatio
n
s
o
th
at
r
esu
lts
o
f
r
ice
g
r
o
wth
p
h
ase
class
if
icatio
n
b
ec
o
m
e
m
o
r
e
p
r
ec
is
e.
T
h
e
r
esu
lts
o
f
th
is
s
tu
d
y
in
d
icate
th
at
th
e
r
ice
g
r
o
wth
p
h
as
e
m
o
d
el
b
ased
o
n
s
en
tin
el
-
1
i
m
ag
e
d
ata
s
till
p
r
o
d
u
ce
s
m
is
class
if
icat
io
n
.
Misclass
if
icatio
n
m
ain
ly
o
cc
u
r
s
in
th
e
wate
r
an
d
f
o
llo
w
-
u
p
p
h
ases
.
T
h
is
is
b
ec
au
s
e
th
e
n
u
m
b
er
o
f
s
am
p
le
p
o
in
ts
in
th
ese
two
p
h
ases
is
n
o
t c
o
m
p
ar
a
b
le
to
th
e
o
th
er
p
h
ases
.
T
h
is
s
itu
atio
n
ca
n
ca
u
s
e
th
e
m
o
d
el
to
m
ak
e
m
is
class
if
icatio
n
s
,
esp
ec
ially
wh
en
th
e
class
d
is
tr
ib
u
tio
n
is
u
n
b
alan
ce
d
,
s
o
t
h
e
SVM
m
eth
o
d
ca
n
b
e
in
ef
f
ec
ti
v
e
in
d
eter
m
i
n
in
g
class
b
o
u
n
d
ar
ies
[
3
9
]
.
T
o
o
v
e
r
co
m
e
t
h
is
u
n
b
alan
ce
d
p
r
o
b
lem
,
a
s
am
p
lin
g
s
tr
ateg
y
ca
n
b
e
ap
p
lied
to
b
alan
ce
th
e
s
ize
o
f
ea
ch
r
ice
p
h
ase
class
,
s
u
ch
as
th
e
s
y
n
th
etic
m
in
o
r
ity
o
v
er
s
am
p
lin
g
tech
n
iq
u
e
(
S
MO
T
E
)
m
eth
o
d
an
d
its
d
er
iv
ativ
e
m
eth
o
d
s
[
3
7
]
.
T
h
e
ap
p
licatio
n
o
f
class
b
alan
cin
g
m
eth
o
d
s
,
s
u
c
h
as
S
MO
T
E
an
d
its
v
ar
ia
n
ts
,
was
n
o
t
ca
r
r
ied
o
u
t
i
n
th
is
s
tu
d
y
b
e
ca
u
s
e
it
was
o
u
ts
id
e
th
e
s
co
p
e
o
f
th
e
s
tu
d
y
.
Ho
we
v
er
,
th
is
ap
p
r
o
ac
h
h
as
th
e
p
o
t
en
tial
to
b
e
ap
p
lied
in
f
u
tu
r
e
s
tu
d
ies
to
o
v
er
co
m
e
class
im
b
alan
ce
is
s
u
es a
n
d
im
p
r
o
v
e
t
h
e
ac
cu
r
ac
y
o
f
r
ice
g
r
o
wth
p
h
ase
class
if
icatio
n
m
o
d
e
ls
.
Miscla
s
s
if
icatio
n
ca
n
also
o
cc
u
r
d
u
e
to
p
r
o
b
lem
s
o
cc
u
r
r
in
g
in
o
n
e
r
ice
cr
o
p
p
in
g
c
y
cle,
s
u
ch
as
p
r
o
b
lem
s
with
p
est
attac
k
s
,
d
r
o
u
g
h
t,
an
d
f
lo
o
d
s
,
s
o
th
e
r
ice
p
h
en
o
lo
g
y
is
d
is
r
u
p
te
d
.
Dam
ag
ed
r
ice
is
u
s
u
ally
r
ep
l
ac
ed
with
n
ew
r
ice
p
lan
ts
,
s
o
th
e
f
ield
d
ata
u
s
ed
i
n
th
e
s
am
p
le
d
if
f
er
s
f
r
o
m
th
e
l
atest r
ea
l f
ield
d
ata.
Fu
r
th
er
m
o
r
e
,
th
is
r
esear
ch
ca
n
b
e
co
n
tin
u
e
d
b
y
d
ev
elo
p
in
g
a
m
o
d
el
th
at
co
n
s
id
er
s
r
an
d
o
m
ef
f
ec
ts
,
s
u
ch
as
th
e
p
lan
tin
g
s
ea
s
o
n
.
I
n
I
n
d
o
n
esia,
th
er
e
ar
e
two
m
ain
p
lan
tin
g
s
ea
s
o
n
s
,
n
am
ely
th
e
r
ain
y
an
d
d
r
y
s
ea
s
o
n
s
,
wh
ich
ca
n
af
f
ec
t
th
e
d
y
n
am
ics
o
f
r
ice
p
h
e
n
o
lo
g
y
[
4
0
]
.
I
n
I
n
d
o
n
esia,
th
er
e
ar
e
t
wo
m
ain
g
r
o
win
g
s
ea
s
o
n
s
,
n
am
ely
th
e
r
ain
y
an
d
d
r
y
s
ea
s
o
n
s
,
wh
ich
ca
n
af
f
ec
t
th
e
d
y
n
am
ics
o
f
r
ice
g
r
o
wth
.
T
h
e
r
esear
ch
i
n
[
4
1
]
,
[
4
2
]
s
h
o
ws
th
at
ad
d
in
g
r
an
d
o
m
e
f
f
ec
ts
to
m
ac
h
in
e
lear
n
in
g
m
o
d
els
ca
n
im
p
r
o
v
e
p
er
f
o
r
m
an
ce
.
B
y
in
clu
d
in
g
t
h
e
p
lan
tin
g
s
ea
s
o
n
as
a
r
an
d
o
m
e
f
f
ec
t,
th
e
m
o
d
el
is
ex
p
ec
ted
to
b
e
ab
l
e
to
ca
p
tu
r
e
n
atu
r
a
l
v
ar
iatio
n
s
b
etwe
en
s
ea
s
o
n
s
s
o
th
at
r
ice
p
h
en
o
lo
g
y
p
h
ase
p
r
ed
ictio
n
s
b
ec
o
m
e
m
o
r
e
r
o
b
u
s
t
an
d
ac
cu
r
ate.
I
n
s
u
b
s
eq
u
en
t
r
esear
ch
,
ad
d
i
n
g
r
an
d
o
m
e
f
f
ec
ts
to
th
e
m
u
lticlas
s
SVM
cla
s
s
if
icatio
n
m
o
d
el
ca
n
b
e
u
s
ed
as
a
n
alter
n
ativ
e
to
im
p
r
o
v
e
th
e
p
er
f
o
r
m
an
ce
o
f
th
e
r
ice
p
h
e
n
o
lo
g
y
m
o
d
el.
5.
CO
NCLU
SI
O
N
T
h
e
SVM
Ov
O
a
n
d
Gen
SV
M
m
eth
o
d
s
ca
n
p
r
o
d
u
ce
r
ice
p
h
en
o
lo
g
y
m
o
d
el
ac
c
u
r
ac
y
,
wh
ich
is
r
ea
s
o
n
ab
ly
s
atis
f
ac
to
r
y
an
d
co
m
p
ar
ab
le
to
th
e
ML
R
m
eth
o
d
.
I
n
th
e
ca
s
e
o
f
r
ice
g
r
o
wth
p
h
ase
class
if
icatio
n
,
th
e
ac
cu
r
ac
y
o
f
th
e
class
if
icatio
n
m
o
d
el
f
r
o
m
t
h
e
SVM
Ov
O,
Gen
SVM,
an
d
ML
R
m
eth
o
d
s
will
b
e
o
p
tim
al
wh
en
all
p
r
ed
icto
r
s
ar
e
in
clu
d
ed
in
th
e
m
o
d
el.
T
h
e
m
o
d
els p
r
o
d
u
ce
d
b
y
th
e
th
r
ee
m
eth
o
d
s
s
h
o
w
g
o
o
d
s
tab
ilit
y
an
d
n
o
v
is
ib
le
o
v
er
f
itti
n
g
p
r
o
b
lem
s
.
Ho
wev
er
,
it
s
h
o
u
ld
b
e
n
o
ted
th
at
th
e
SVM
Ov
O
m
e
th
o
d
is
s
en
s
itiv
e
to
ch
an
g
es
in
th
e
v
alu
e
o
f
th
e
c
o
s
t
an
d
g
am
m
a
p
ar
a
m
eter
s
.
T
h
e
Gen
SVM
m
eth
o
d
also
d
ep
en
d
s
o
n
th
e
k
ap
p
a,
lam
b
d
a,
an
d
p
p
ar
am
eter
s
ettin
g
s
.
T
h
e
ac
c
u
r
ac
y
o
f
t
h
e
r
ice
p
h
en
o
l
o
g
y
m
o
d
el
u
s
in
g
s
en
tin
el
-
1
s
atellite
im
ag
e
d
ata
is
m
o
s
t
o
p
tim
al
in
th
e
4
-
class
s
ce
n
ar
io
th
r
o
u
g
h
SV
M
Ov
O
m
o
d
ellin
g
.
I
n
f
u
tu
r
e
r
esear
ch
,
s
am
p
lin
g
h
an
d
lin
g
test
s
ca
n
b
e
ca
r
r
ie
d
o
u
t,
f
o
r
ex
am
p
le,
u
s
in
g
th
e
S
MO
T
E
m
eth
o
d
to
in
c
r
ea
s
e
th
e
s
m
all
n
u
m
b
er
o
f
class
m
em
b
er
s
.
I
n
ad
d
itio
n
,
f
u
tu
r
e
r
esear
ch
ca
n
b
e
ca
r
r
ie
d
o
u
t
b
y
ad
d
in
g
r
an
d
o
m
ef
f
ec
ts
to
th
e
SVM
m
o
d
el
th
at
ca
n
ac
co
m
m
o
d
ate
f
ix
ed
ef
f
ec
ts
an
d
r
an
d
o
m
ef
f
ec
ts
in
th
e
m
o
d
el
to
r
ed
u
ce
m
is
class
if
icatio
n
ca
u
s
ed
b
y
f
ix
ed
an
d
r
an
d
o
m
e
f
f
ec
ts
.
ACK
NO
WL
E
DG
M
E
N
T
S
W
e
th
an
k
B
R
I
N
'
s
D
ir
ec
to
r
ate
o
f
T
alen
t
Ma
n
ag
e
m
en
t
f
o
r
s
u
p
p
o
r
tin
g
th
e
s
tu
d
y
.
W
e
also
th
an
k
th
e
r
ice
p
h
ase
r
esear
ch
team
at
t
h
e
B
R
I
N
R
em
o
te
Sen
s
in
g
R
e
s
ea
r
ch
C
en
ter
an
d
th
e
Ma
n
ag
em
en
t
o
f
PT.
San
g
Hy
an
g
Ser
i Su
b
a
n
g
,
W
est J
av
a.
F
UNDING
I
NF
O
R
M
A
T
I
O
N
W
e
wo
u
ld
lik
e
to
th
an
k
th
e
D
ir
ec
to
r
ate
Gen
er
al
o
f
Hig
h
er
E
d
u
ca
tio
n
,
R
esear
ch
,
an
d
T
ec
h
n
o
lo
g
y
o
f
th
e
Min
is
tr
y
o
f
E
d
u
ca
tio
n
,
C
u
ltu
r
e,
R
esear
ch
,
a
n
d
T
ec
h
n
o
lo
g
y
f
o
r
p
r
o
v
id
in
g
a
Do
cto
r
al
d
i
s
s
er
tatio
n
r
esear
ch
g
r
an
t
with
Dec
r
ee
Nu
m
b
er
0
2
7
/E5
/PG.0
2
.
0
0
.
PL/2
0
2
4
an
d
Ag
r
ee
m
e
n
t/C
o
n
tr
ac
t
Nu
m
b
er
2
2
1
1
5
/
I
T
3
.
D1
0
/PT.
0
1
.
0
3
/P/
B
/2
0
2
4
.
Evaluation Warning : The document was created with Spire.PDF for Python.