I
nte
rna
t
io
na
l J
o
urna
l o
f
E
lect
rica
l a
nd
Co
m
p
ute
r
E
ng
in
ee
ring
(
I
J
E
CE
)
Vo
l.
10
,
No
.
5
,
Octo
b
er
2
0
2
0
,
p
p
.
5
4
9
7
~
5
5
0
6
I
SS
N:
2
0
8
8
-
8708
,
DOI
: 1
0
.
1
1
5
9
1
/
i
j
ec
e
.
v
1
0
i
5
.
pp
5
4
9
7
-
5
5
0
6
5497
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ec
e.
ia
esco
r
e.
co
m/in
d
ex
.
p
h
p
/I
JE
C
E
The ef
fect o
f
g
a
mm
a
va
lue on su
pp
o
rt
v
ector
m
a
chi
n
e
perfor
m
a
nce w
ith
diff
erent
k
ern
els
I
ntisa
r
Sh
a
dee
d Al
-
M
ej
ibl
i
1
,
J
w
a
n K
.
Alwa
n
2
,
Dha
f
a
r
H
a
m
e
d Abd
3
1,
2
Un
iv
e
rsity
o
f
In
f
o
rm
a
ti
o
n
T
e
c
h
n
o
l
o
g
y
a
n
d
Co
m
m
u
n
ica
ti
o
n
s,
Ira
q
3
Al
-
M
a
a
r
e
f
Un
iv
e
rsit
y
Co
ll
e
g
e
,
Ira
q
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Oct
2
3
,
2
0
1
9
R
ev
i
s
ed
A
p
r
8
,
2
0
2
0
A
cc
ep
ted
A
p
r
2
5
,
2
0
2
0
Cu
rre
n
tl
y
,
th
e
su
p
p
o
rt
v
e
c
to
r
m
a
c
h
in
e
(S
V
M
)
re
g
a
rd
e
d
a
s
o
n
e
o
f
su
p
e
rv
ise
d
m
a
c
h
in
e
lea
rn
in
g
a
lg
o
rit
h
m
th
a
t
p
ro
v
id
e
s
a
n
a
ly
sis
o
f
d
a
ta
f
o
r
c
las
sif
ic
a
ti
o
n
a
n
d
re
g
re
ss
io
n
.
T
h
is
tec
h
n
iq
u
e
is
im
p
le
m
e
n
ted
in
m
a
n
y
f
ield
s
su
c
h
a
s
b
i
o
i
n
f
o
r
m
a
t
i
c
s
,
f
a
c
e
r
e
c
o
g
n
i
t
i
o
n
,
t
e
x
t
a
n
d
h
y
p
e
r
t
e
x
t
c
a
t
e
g
o
r
i
z
a
t
i
o
n
,
g
e
n
e
r
a
l
i
z
e
d
p
r
e
d
i
c
t
i
v
e
c
o
n
t
r
o
l
a
n
d
m
a
n
y
o
t
h
e
r
d
i
f
f
e
r
e
n
t
a
r
e
a
s
.
T
h
e
p
e
r
f
o
r
m
a
n
c
e
o
f
S
V
M
is
a
ffe
c
ted
b
y
so
m
e
p
a
ra
m
e
ters
,
w
h
ich
a
re
u
se
d
in
t
h
e
train
i
n
g
p
h
a
se
,
a
n
d
th
e
se
tt
in
g
s
o
f
p
a
ra
m
e
ters
c
a
n
h
a
v
e
a
p
ro
f
o
u
n
d
im
p
a
c
t
o
n
th
e
re
su
lt
in
g
e
n
g
in
e
’s
im
p
le
m
e
n
tatio
n
.
T
h
is
p
a
p
e
r
in
v
e
sti
g
a
ted
th
e
S
V
M
p
e
rf
o
r
m
a
n
c
e
b
a
se
d
o
n
v
a
lu
e
o
f
g
a
m
m
a
p
a
ra
m
e
ter
w
it
h
u
se
d
k
e
rn
e
l
s
.
It
stu
d
ied
th
e
im
p
a
c
t
o
f
g
a
m
m
a
v
a
lu
e
o
n
(S
V
M
)
e
f
f
i
c
ien
c
y
c
las
si
f
ier
u
sin
g
d
if
f
e
r
e
n
t
k
e
rn
e
ls
o
n
v
a
rio
u
s
d
a
tas
e
ts
d
e
sc
rip
ti
o
n
s.
S
VM
c
las
sif
ier
h
a
s
b
e
e
n
i
m
p
le
m
e
n
te
d
b
y
u
sin
g
P
y
th
o
n
.
T
h
e
k
e
rn
e
l
f
u
n
c
ti
o
n
s
t
h
a
t
h
a
v
e
b
e
e
n
in
v
e
stig
a
ted
a
re
p
o
ly
n
o
m
ials,
r
a
d
ial
b
a
se
d
f
u
n
c
ti
o
n
(RBF
)
a
n
d
s
ig
m
o
id
.
UC
irv
in
e
m
a
c
h
in
e
lea
rn
in
g
re
p
o
sito
ry
is
th
e
s
o
u
rc
e
o
f
a
ll
th
e
u
se
d
d
a
tas
e
ts.
G
e
n
e
ra
ll
y
,
th
e
re
su
lt
s
sh
o
w
u
n
e
v
e
n
e
ff
e
c
t
o
n
th
e
c
las
sif
ica
ti
o
n
a
c
c
u
r
a
c
y
o
f
th
re
e
k
e
rn
e
ls
o
n
u
se
d
d
a
tas
e
ts.
T
h
e
c
h
a
n
g
in
g
o
f
th
e
g
a
m
m
a
v
a
l
u
e
tak
in
g
o
n
c
o
n
sid
e
ra
ti
o
n
t
h
e
u
se
d
d
a
t
a
se
t
in
f
lu
e
n
c
e
s
p
o
ly
n
o
m
ial
a
n
d
s
ig
m
o
id
k
e
rn
e
ls.
W
h
il
e
th
e
p
e
rf
o
r
m
a
n
c
e
o
f
RBF
k
e
rn
e
l
f
u
n
c
ti
o
n
is
m
o
re
sta
b
le
w
it
h
d
if
f
e
re
n
t
v
a
lu
e
s
o
f
g
a
m
m
a
a
s
it
s
a
c
c
u
ra
c
y
is
slig
h
tl
y
c
h
a
n
g
e
d
.
K
ey
w
o
r
d
s
:
Ga
m
m
a
Ker
n
els
Su
p
p
o
r
t v
ec
to
r
m
ac
h
i
n
e
Co
p
y
rig
h
t
©
2
0
2
0
In
stit
u
te o
f
A
d
v
a
n
c
e
d
E
n
g
i
n
e
e
rin
g
a
n
d
S
c
ien
c
e
.
Al
l
rig
h
ts re
se
rv
e
d
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
I
n
tis
ar
S
h
ad
ee
d
A
l
-
Me
j
ib
li
,
S
m
ar
t C
ities
,
Un
i
v
er
s
it
y
o
f
I
n
f
o
r
m
at
io
n
T
ec
h
n
o
lo
g
y
a
n
d
C
o
m
m
u
n
icatio
n
s
,
Al
-
Nid
h
al
Stre
et,
B
ag
h
d
ad
,
I
r
aq
.
E
m
ail:
d
r
.
in
ti
s
ar
.
al
m
ej
ib
li@
g
m
ail.
co
m
; d
r
.
in
tis
ar
.
al
m
ej
ib
li@
u
o
itc.
ed
u
.
iq
1.
I
NT
RO
D
UCT
I
O
N
Su
p
p
o
r
t
v
ec
to
r
m
ac
h
in
e
(
SV
M)
is
s
u
p
er
v
is
ed
lear
n
i
n
g
m
o
d
els
h
av
e
th
e
ab
ilit
y
o
f
a
n
al
y
z
in
g
d
ata
f
o
r
class
i
f
icatio
n
an
d
r
eg
r
e
s
s
io
n
p
u
r
p
o
s
es.
SVM
i
s
s
u
p
p
o
r
ted
b
y
th
e
t
h
eo
r
y
o
f
s
tatis
t
ical
lear
n
i
n
g
,
w
h
ic
h
h
a
s
b
ee
n
d
ev
elo
p
ed
b
y
Vlad
i
m
ir
[
1
]
.
Alth
o
u
g
h
,
th
er
e
ar
e
m
a
n
y
al
g
o
r
ith
m
s
w
er
e
d
ev
elo
p
ed
f
o
r
p
atter
n
clas
s
i
f
icatio
n
,
SVM
is
th
e
b
est
o
n
e
a
m
o
n
g
th
e
m
.
Ne
w
l
y
,
i
t
w
as
ad
ap
ted
f
o
r
n
e
w
i
m
p
le
m
e
n
tatio
n
s
f
o
r
i
n
s
ta
n
ce
d
is
co
v
er
i
n
g
r
eg
r
ess
io
n
an
d
d
is
tr
ib
u
tio
n
e
s
ti
m
atio
n
.
C
u
r
r
en
tl
y
,
SVM
i
s
i
m
p
le
m
e
n
ted
in
d
i
f
f
er
en
t
f
ield
s
f
o
r
in
s
ta
n
ce
b
io
in
f
o
r
m
at
ics,
w
h
ich
i
s
r
eg
ar
d
ed
as
so
p
r
o
m
i
s
ed
r
esear
ch
ar
ea
f
o
r
r
esear
ch
e
rs
.
I
n
g
en
er
al,
th
e
p
r
o
ce
s
s
o
f
an
al
y
zi
n
g
d
ataset
ite
m
s
i
s
b
ased
o
n
th
eir
lab
els,
w
h
er
e
lab
els
r
ep
r
esen
t
th
e
clas
s
in
f
o
r
m
at
io
n
o
f
r
elate
d
d
ata
in
clu
d
ed
in
th
at
d
atase
t
s
.
C
las
s
i
f
icatio
n
m
et
h
o
d
s
is
i
m
p
le
m
e
n
ted
i
n
m
a
n
y
r
esear
ch
f
ield
s
s
u
c
h
as
b
io
s
cie
n
ce
,
ec
o
n
o
m
y
a
n
d
f
o
r
ec
asti
n
g
b
ec
au
s
e
it
s
u
p
p
o
r
t
th
e
d
ec
is
io
n
s
a
n
d
s
u
m
m
ar
ize
th
e
co
llected
d
ata
to
b
e
an
al
y
z
ed
ef
f
icie
n
tl
y
.
T
h
u
s
,
c
lass
if
ica
tio
n
m
e
th
o
d
s
ar
e
u
s
ed
b
y
ec
o
n
o
m
is
t
s
an
d
d
o
cto
r
s
an
d
n
o
t
li
m
ited
to
co
m
p
u
ter
s
cien
ti
s
t
s
[2
-
4
]
.
T
h
e
SVM
p
r
in
cip
le
is
b
u
ilt
o
n
b
in
ar
y
cla
s
s
i
f
i
ca
tio
n
w
h
er
e
i
t
u
s
e
s
s
tr
aig
h
t
l
in
e
to
s
ep
ar
ates
d
ata
p
o
in
ts
to
cla
s
s
i
f
y
t
h
e
cla
s
s
la
b
el.
N
ev
er
th
ele
s
s
,
s
tr
ai
g
h
t
li
n
e
ca
n
n
o
t
b
e
u
s
ed
to
s
ep
ar
ate
th
e
d
ata
p
o
in
ts
i
n
s
o
m
e
d
atasets
,
s
o
th
e
Ker
n
el
f
u
n
ctio
n
s
w
er
e
r
ev
ea
led
to
tack
le
th
is
i
s
s
u
e
[
5
,
6
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
10
,
No
.
5
,
Octo
b
e
r
2
0
2
0
:
5
4
9
7
-
5506
5498
T
h
e
SVM
p
er
f
o
r
m
a
n
ce
is
b
as
ed
o
n
b
o
th
its
p
ar
a
m
eter
s
an
d
u
s
ed
k
er
n
el
f
u
n
ctio
n
,
w
h
ic
h
i
s
ass
i
s
ti
n
g
SVM
to
ac
h
ie
v
e
t
h
e
b
est
s
ep
ar
atin
g
h
y
p
er
p
la
n
e
i
n
t
h
e
f
ea
t
u
r
e
s
p
ac
e.
On
e
o
f
t
h
e
s
ig
n
i
f
ica
n
t
SV
M
p
ar
a
m
eter
s
is
Ga
m
m
a.
T
h
e
u
n
i
n
t
u
iti
v
e
p
ar
a
m
eter
g
a
m
m
a
d
e
f
in
e
s
th
e
w
id
t
h
o
r
s
lo
p
e
o
f
k
er
n
el
f
u
n
c
tio
n
.
T
h
e
‘
cu
r
v
e
’
o
f
th
e
d
ec
is
io
n
b
o
u
n
d
ar
y
b
ec
o
m
es
v
er
y
lo
w
w
h
e
n
g
a
m
m
a
v
al
u
e
is
lo
w
m
a
k
in
g
t
h
e
d
ec
is
io
n
r
eg
io
n
v
er
y
b
r
o
ad
.
T
h
e
‘
cu
r
v
e
’
o
f
t
h
e
d
ec
is
io
n
b
o
u
n
d
ar
y
b
ec
o
m
es
h
ig
h
wh
en
g
a
m
m
a
i
s
h
i
g
h
,
w
h
ic
h
cr
ea
tes
is
lan
d
s
o
f
d
ec
is
io
n
-
b
o
u
n
d
ar
ies
ar
o
u
n
d
d
ata
p
o
in
ts
.
Ga
m
m
a
p
ar
a
m
e
te
r
is
i
n
v
o
l
v
ed
w
h
en
th
e
p
o
l
y
n
o
m
ial
s
,
R
B
F
an
d
Sig
m
o
id
ar
e
s
et
[
7
]
.
T
h
is
p
ap
e
r
ai
m
s
to
i
n
v
esti
g
ate
t
h
e
i
m
p
a
ct
o
f
g
a
m
m
a
v
alu
e
w
it
h
k
er
n
el
o
n
p
er
f
o
r
m
an
ce
o
f
SVM
u
s
i
n
g
d
i
f
f
er
en
t
d
ataset
s
,
in
o
r
d
er
to
s
p
ec
if
y
t
h
e
b
est
p
er
f
o
r
m
a
n
ce
o
f
SVM
.
T
h
e
m
eth
o
d
o
lo
g
y
o
f
th
i
s
s
tu
d
y
b
ased
o
n
u
s
i
n
g
d
i
f
f
er
en
t
v
al
u
es
o
f
g
a
m
m
a
p
ar
a
m
e
ter
w
it
h
d
if
f
er
e
n
t
k
er
n
el
f
u
n
ctio
n
s
to
ev
al
u
ate
th
e
p
er
f
o
r
m
a
n
ce
o
f
S
VM
m
et
h
o
d
.
Var
io
u
s
d
atase
ts
d
escr
ip
tio
n
s
i
s
u
s
ed
i
n
t
h
i
s
s
tu
d
y
,
w
h
ich
g
i
v
e
th
e
e
v
al
u
atio
n
tas
k
an
o
t
h
er
ax
i
s
.
I
n
g
e
n
er
al,
t
h
i
s
r
esear
ch
s
t
u
d
ies
t
h
e
i
m
p
ac
t
o
f
g
a
m
m
a
v
al
u
e,
k
er
n
el
f
u
n
ctio
n
s
a
n
d
d
atasets
d
escr
ip
tio
n
s
o
n
SVM
p
er
f
o
r
m
an
ce
.
T
h
r
ee
k
er
n
el
f
u
n
ctio
n
s
w
er
e
s
t
u
d
ied
w
it
h
f
i
v
e
d
i
f
f
er
en
t
d
ataset
s
w
h
ic
h
h
av
i
n
g
v
ar
io
u
s
clas
s
es
i
n
s
tan
c
es.
P
o
ly
n
o
m
ials
,
r
ad
ial
b
ased
f
u
n
ctio
n
(
R
B
F)
a
n
d
Si
g
m
o
id
ar
e
th
e
i
n
v
esti
g
ated
k
er
n
el
f
u
n
ctio
n
s
.
T
h
e
m
ai
n
o
b
j
ec
tiv
e
o
f
th
is
s
tu
d
y
i
s
r
ev
ea
l
t
h
e
p
er
f
o
r
m
an
ce
o
f
SVM
u
s
in
g
d
if
f
er
en
t
v
al
u
es
o
f
Ga
m
m
a
p
ar
a
m
eter
w
it
h
th
e
u
s
ed
k
er
n
el
f
u
n
ctio
n
s
an
d
d
atase
ts
.
T
h
e
r
est
p
ar
t
o
f
th
is
r
esear
c
h
is
o
r
g
an
ized
as
f
o
llo
w
i
n
g
:
Se
ctio
n
2
p
r
esen
ts
r
elate
d
w
o
r
k
.
S
ec
tio
n
3
ex
p
lain
e
s
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
e
s
(
SVM)
.
Data
s
et
s
is
d
escr
ib
ed
in
s
ec
tio
n
4
.
Sectio
n
5
d
is
cu
s
s
e
s
th
e
r
esu
lts
.
C
o
n
cl
u
s
io
n
is
ill
u
s
tr
ates
i
n
s
ec
tio
n
6
.
2.
RE
L
AT
E
D
WO
RK
T
h
e
p
er
f
o
r
m
a
n
ce
o
f
SVM
r
el
ated
to
g
a
m
m
a
p
ar
a
m
eter
v
al
u
es
i
n
k
er
n
el
f
u
n
ctio
n
s
h
as
b
ee
n
s
t
u
d
ied
b
y
f
e
w
r
esear
c
h
e
r
s
.
T
h
e
f
o
llo
w
i
n
g
ab
s
tr
ac
ted
s
o
m
e
o
f
t
h
e
p
u
b
lis
h
ed
s
t
u
d
ies
:
T
h
e
au
th
o
r
s
in
[
8
]
p
r
esen
ted
a
r
ev
ie
w
in
t
h
e
ef
f
icac
y
o
f
t
h
e
ch
o
ice
o
f
k
er
n
el
s
f
o
r
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
e
s
a
n
d
t
w
o
d
if
f
er
en
t
d
i
m
e
n
s
io
n
a
lit
y
tec
h
n
iq
u
es,
w
h
ic
h
ar
e
f
ix
ed
s
lo
p
e
r
eg
r
e
s
s
io
n
an
d
p
r
in
c
i
p
al
co
m
p
o
n
e
n
t
a
n
al
y
s
i
s
.
A
lea
v
e
–
o
n
e
-
o
u
t
cla
s
s
i
f
ier
u
s
ed
to
d
eter
m
i
n
e
b
o
th
o
f
d
i
m
e
n
s
io
n
alit
y
r
ed
u
c
tio
n
co
m
b
i
n
atio
n
a
n
d
ac
cu
r
ac
y
.
A
h
ig
h
-
d
i
m
en
s
io
n
al
b
io
-
i
m
a
g
in
g
d
ataset
i
s
u
s
ed
a
n
d
t
h
e
r
es
u
lt
s
s
h
o
w
ed
t
h
at
R
B
F
k
er
n
el
ac
h
ie
v
ed
th
e
b
est ac
cu
r
ac
y
.
In
[
9
]
th
e
au
t
h
o
r
s
co
m
p
ar
e
th
e
p
er
f
o
r
m
a
n
ce
o
f
SVM
clas
s
i
f
i
ca
tio
n
alg
o
r
it
h
m
f
o
r
all
d
if
f
er
e
n
t
k
er
n
el
s
(
lin
ea
r
,
p
o
l
y
n
o
m
ial,
R
B
F
a
n
d
s
i
g
m
o
id
)
w
it
h
d
i
f
f
er
en
t
p
ar
am
eter
s
(
C
-
SV
C
,
NU
-
SV
C
)
to
f
il
ter
j
u
n
k
o
r
s
p
a
m
E
-
m
a
ils
.
A
s
p
a
m
-
b
ase
d
ata
s
et
f
r
o
m
p
u
b
lic
d
o
m
ai
n
web
s
ite
(
UC
I
r
esp
ir
ato
r
y
)
u
s
ed
to
im
p
le
m
e
n
t
th
e
co
m
p
ar
i
s
o
n
.
T
h
e
SVM
o
b
tain
ed
b
est
ac
c
u
r
ac
y
w
h
e
n
l
in
ea
r
k
er
n
el
f
u
n
c
tio
n
a
n
d
SV
C
ar
e
i
m
p
le
m
en
ted
a
s
th
e
r
esu
lts
s
h
o
w
ed
.
P
ap
er
in
[
10
]
im
p
le
m
e
n
ted
SVM
b
y
u
s
i
n
g
elev
e
n
m
at
h
e
m
a
tical
f
u
n
ctio
n
s
ca
ll
ed
k
er
n
els
.
T
h
is
r
esear
ch
w
as
i
m
p
le
m
e
n
te
d
o
n
i
m
ag
e
s
o
f
A
r
ab
ic
alp
h
ab
etic
ch
ar
ac
ter
.
T
h
e
d
ataset
u
s
e
d
in
t
h
i
s
p
ap
er
w
as
a
d
atab
ase
cr
ea
ted
f
r
o
m
c
h
ar
ac
ter
i
m
ag
e
s
.
T
h
e
s
alt
an
d
p
ep
p
er
n
o
is
e
f
r
o
m
0
.
1
to
0
.
9
ar
e
u
s
ed
to
co
r
r
u
p
t
ch
ar
ac
ter
s
.
T
h
e
r
es
u
lt
s
s
h
o
w
ed
th
e
r
elatio
n
b
et
w
ee
n
n
o
is
e
lev
el
an
d
ef
f
icie
n
c
y
.
W
h
e
n
th
e
n
o
i
s
e
le
v
el
i
s
in
cr
ea
s
ed
th
e
e
f
f
icie
n
c
y
ar
e
r
eg
u
lar
l
y
d
ec
r
ea
s
ed
f
o
r
b
o
th
P
o
l
y
n
o
m
ial
a
n
d
lin
ea
r
k
er
n
els.
T
h
e
p
ap
e
r
[
11
]
p
r
esen
ted
a
co
m
p
ar
at
iv
e
s
tu
d
y
f
o
r
SVM
i
m
p
le
m
en
ta
tio
n
w
it
h
f
o
u
r
d
if
f
er
en
t
k
er
n
el
s
(
lin
ea
r
,
r
ad
ial
b
asis
,
p
o
ly
n
o
m
ial
,
an
d
s
i
g
m
o
id
)
.
T
h
is
co
m
p
ar
at
iv
e
s
t
u
d
y
u
s
ed
m
a
n
y
d
if
f
er
en
t
d
ataset
s
.
T
h
e
lin
ea
r
k
er
n
el
o
b
tain
ed
b
es
t
r
esu
lt
as
it
o
b
tai
n
ed
8
8
.
2
0
%
in
ac
c
u
r
ac
y
a
n
d
also
s
h
o
w
ed
th
at
t
h
e
p
r
ed
ictio
n
ti
m
e
w
a
s
f
a
s
ter
ab
o
u
t
4
.
0
7
8
s
e
c.
I
n
th
i
s
p
ap
er
[
12
]
an
o
th
er
co
m
p
ar
at
iv
e
s
tu
d
y
h
a
v
e
b
ee
n
d
o
n
e
to
s
t
u
d
y
t
h
e
i
m
p
le
m
e
n
tatio
n
o
f
li
n
ea
r
an
d
R
B
F
SVM
f
o
r
f
M
R
I
clas
s
if
ica
tio
n
w
i
th
v
o
x
el
s
elec
tio
n
s
ch
e
m
es
.
T
h
is
s
t
u
d
y
co
m
p
ar
e
d
th
e
class
if
icatio
n
ac
cu
r
ac
y
an
d
ti
m
e
-
co
n
s
u
m
in
g
to
d
ec
ip
h
er
b
r
ain
p
atter
n
s
f
r
o
m
f
u
n
ctio
n
al
M
R
I
(
f
M
R
I
)
d
ata.
Si
x
d
i
f
f
er
en
t
m
et
h
o
d
s
o
f
v
o
x
el
s
elec
tio
n
h
a
v
e
b
ee
n
e
m
p
lo
y
ed
to
s
p
ec
if
y
,
w
h
ic
h
v
o
x
els
o
f
f
M
R
I
d
ata
w
o
u
ld
b
e
in
clu
d
ed
i
n
SVM
class
if
ier
s
w
ith
li
n
ea
r
an
d
R
B
F
k
er
n
els
i
n
clas
s
if
y
i
n
g
4
-
ca
te
g
o
r
y
o
b
j
ec
ts
.
R
esu
lts
r
ev
ea
led
th
at
th
e
R
B
F
SVM
w
as
ou
tp
er
f
o
r
m
li
n
ea
r
SVM
w
h
e
n
i
m
p
le
m
e
n
ti
n
g
a
f
ea
tu
r
e
s
p
ac
e
o
f
a
r
elati
v
e
lo
w
d
i
m
e
n
s
io
n
al.
W
h
er
ea
s
,
lin
ea
r
SVM
o
u
tp
er
f
o
r
m
ed
R
B
F
SV
M
w
h
e
n
i
m
p
le
m
en
t
in
g
a
f
ea
tu
r
e
s
p
ac
e
o
f
a
r
elativ
e
h
ig
h
d
i
m
en
s
io
n
al.
In
[
13
]
th
e
a
u
t
h
o
r
s
in
v
esti
g
ated
t
h
e
SVM
i
m
p
le
m
en
ta
tio
n
w
it
h
li
n
ea
r
,
p
o
ly
n
o
m
ial
a
n
d
R
ad
ial
B
asic
F
u
n
ct
io
n
k
er
n
el
s
u
s
in
g
th
r
ee
d
ata
s
ets
f
o
r
t
h
e
p
u
r
p
o
s
e
o
f
tex
t
i
n
d
ep
en
d
en
t
s
p
ea
k
er
.
T
h
e
SVM
w
it
h
p
o
ly
n
o
m
ia
l k
er
n
el
ac
h
ie
v
ed
th
e
b
est p
er
f
o
r
m
a
n
ce
as t
h
e
r
es
u
lts
s
h
o
w
ed
.
T
h
e
au
th
o
r
s
in
[
14
]
p
r
esen
ted
th
e
i
m
p
le
m
en
tat
io
n
o
f
SVM
w
it
h
f
o
u
r
k
er
n
el
s
:
L
in
ea
r
,
R
a
d
ial
B
asis
,
P
o
ly
n
o
m
ial
an
d
Q
u
ad
r
atic
to
s
o
lv
e
t
h
e
s
p
ea
k
er
id
en
ti
f
icatio
n
tas
k
u
s
in
g
T
I
M
I
T
cor
pus
.
Fo
u
r
d
ataset
s
w
er
e
u
s
e
d
t
o
im
p
l
em
e
n
t
t
h
e
e
x
p
e
r
i
m
e
n
t
.
T
h
e
Q
u
a
d
r
a
t
i
c
k
e
r
n
e
l
o
u
t
p
e
r
f
o
r
m
t
h
e
o
t
h
e
r
t
h
r
e
e
k
e
r
n
e
l
s
a
s
t
h
e
r
e
s
u
l
t
s
s
h
o
w
e
d.
In
[
15
]
th
e
au
t
h
o
r
s
an
al
y
ze
d
t
h
e
i
m
p
le
m
e
n
tatio
n
o
f
d
i
f
f
er
en
t
k
er
n
el
f
u
n
c
tio
n
s
in
SVM
s
u
c
h
as
R
ad
ial
B
asis
,
L
in
ea
r
,
Q
u
ad
r
atic
a
n
d
P
o
ly
n
o
m
ial
in
r
ec
o
g
n
iz
in
g
t
h
e
e
m
o
tio
n
s
o
f
f
ac
e
.
A
d
atab
a
s
e
is
cr
ea
ted
u
s
i
n
g
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2
0
8
8
-
8708
Th
e
effec
t o
f g
a
mma
va
l
u
e
o
n
s
u
p
p
o
r
t v
ec
to
r
ma
ch
in
e
…
(
I
n
t
is
a
r
S
h
a
d
ee
d
A
l
-
Mejib
li
)
5499
d
ig
ital
ca
m
er
a
an
d
b
ased
o
n
5
1
p
er
s
o
n
s
,
w
h
ic
h
i
n
clu
d
e
s
7
1
4
im
a
g
es
o
f
f
ac
e
e
m
o
tio
n
w
it
h
s
e
v
e
n
f
ac
ial
ex
p
r
ess
io
n
s
.
T
h
e
Qu
ad
r
atic
k
e
r
n
el
o
b
tain
ed
b
est
ac
cu
r
ac
y
.
T
h
e
au
th
o
r
s
i
n
[
16
]
d
is
cu
s
s
ed
th
e
i
m
p
le
m
e
n
tat
io
n
o
f
k
er
n
e
l f
u
n
ct
io
n
s
w
it
h
th
e
SVM
cla
s
s
i
f
ier
in
f
i
e
l
d
o
f
a
u
t
o
m
a
t
i
c
t
e
x
t
c
l
a
s
s
i
f
i
c
a
t
i
o
n
.
A
f
ew
m
o
d
e
l
s
f
o
r
t
e
x
t
r
e
p
r
e
s
e
n
t
a
t
i
o
n
i
n
c
l
u
d
i
n
g
V
S
M
,
G
V
S
M
,
a
n
d
n
-
g
r
a
m
s
,
s
t
r
i
n
g
s
a
r
e
d
i
s
c
u
s
s
e
d
;
i
n
m
o
s
t
o
f
t
h
e
m
t
e
x
t
i
s
r
e
p
r
e
s
e
n
t
e
d
b
y
a
s
e
t
o
f
v
e
c
t
o
r
s
i
n
a
h
i
g
h
d
i
m
e
n
s
i
o
n
a
l
s
p
a
c
e
.
T
h
e
au
th
o
r
s
in
[
1
7
]
co
m
p
ar
ed
th
e
p
er
f
o
r
m
an
ce
o
f
SVM
w
h
en
s
et
tin
g
it
w
it
h
lin
ea
r
,
p
o
l
y
n
o
m
ial
a
n
d
R
ad
ial
B
asis
Fu
n
ctio
n
(
R
B
F)
k
er
n
el
s
.
T
h
r
ee
d
atasets
ar
e
ap
p
lied
to
test
th
e
th
r
ee
k
er
n
els
,
w
h
ic
h
ar
e
DL
B
C
L
,
b
r
ain
t
u
m
o
r
,
a
n
d
T
u
m
o
r
s
.
T
h
e
test
e
s
w
er
e
co
n
d
u
cti
n
g
w
i
t
h
f
ea
t
u
r
e
s
elec
tio
n
in
SVM
c
lass
i
f
icatio
n
k
er
n
el
f
u
n
ctio
n
an
d
w
it
h
o
u
t
it
f
o
r
co
m
p
ar
i
s
o
n
p
u
r
p
o
s
e.
T
h
e
r
esu
lts
s
h
o
w
ed
th
at
li
n
ea
r
k
er
n
el
f
u
n
ctio
n
o
b
tain
ed
h
ig
h
ac
cu
r
ac
y
f
o
r
th
e
w
h
o
le
d
ataset
w
it
h
o
u
t
f
ea
tu
r
e
s
elec
t
io
n
.
W
h
er
ea
s
,
th
e
d
ataset
th
at
in
cl
u
d
es
in
f
o
r
m
a
tiv
e
g
en
e
s
g
ets
h
i
g
h
ac
c
u
r
ac
y
o
f
9
0
p
er
ce
n
t a
n
d
ab
o
v
e.
I
n
[
1
8
]
th
e
au
t
h
o
r
s
ap
p
lied
d
etailed
ass
es
s
m
en
t
o
f
r
esa
m
p
li
n
g
p
r
o
ce
d
u
r
es
f
o
r
SVM
h
y
p
er
p
ar
am
eter
s
elec
tio
n
as
t
h
e
y
ai
m
ed
to
en
r
ich
liter
at
u
r
e
in
m
ac
h
i
n
e
lear
n
i
n
g
.
T
h
e
co
n
d
u
cted
e
m
p
ir
ical
e
v
alu
a
tio
n
b
ased
o
n
test
i
n
g
1
5
p
r
o
ce
d
u
r
es
o
f
v
ar
i
o
u
s
r
esa
m
p
li
n
g
o
n
1
2
1
b
in
ar
y
clas
s
i
f
icatio
n
d
ataset
s
to
ch
o
o
s
e
th
e
b
est
SV
M
h
y
p
er
p
ar
a
m
eter
s
.
A
n
a
l
y
zi
n
g
th
e
o
b
tai
n
ed
r
esu
l
ts
r
eq
u
ir
ed
u
s
i
n
g
t
h
r
ee
d
if
f
er
en
t
s
tat
is
tical
p
r
o
ce
d
u
r
es,
in
cl
u
d
es
t
h
e
s
tan
d
ar
d
m
u
l
ti
-
cl
ass
i
f
ier
/
m
u
l
tid
ata
s
et
p
r
o
ce
d
u
r
e,
th
e
co
n
f
id
e
n
ce
i
n
ter
v
a
ls
o
n
th
e
e
x
ce
s
s
lo
s
s
o
f
ea
ch
p
r
o
ce
d
u
r
e,
an
d
th
e
B
ay
e
s
f
ac
to
r
an
al
y
s
i
s
.
T
h
e
r
esu
lts
s
h
o
w
ed
th
at
th
e
2
-
f
o
ld
p
r
o
ce
d
u
r
e
is
co
n
v
en
ie
n
t
to
s
elec
t
th
e
h
y
p
er
p
ar
am
e
ter
s
o
f
an
SVM
f
o
r
d
atasets
w
it
h
1
0
0
0
o
r
m
o
r
e
o
f
d
ata
p
o
in
ts
,
w
h
er
ea
s
a
3
-
f
o
ld
p
r
o
ce
d
u
r
e
is
co
n
v
en
ie
n
t
f
o
r
d
atasets
w
it
h
s
m
al
ler
d
ata
p
o
in
ts
.
T
h
e
au
th
o
r
s
i
n
[
1
9
]
p
r
o
p
o
s
ed
a
k
er
n
el
f
u
n
ct
io
n
f
o
r
(
OA
O
)
an
d
(
OAA
)
m
u
lticla
s
s
S
V
M’
s
.
Ma
n
y
m
etr
ics
h
av
e
b
ee
n
m
ea
s
u
r
ed
in
clu
d
i
n
g
ac
cu
r
ac
y
,
s
u
p
p
o
r
t
v
ec
to
r
p
er
ce
n
tag
e,
s
p
ee
d
an
d
cla
s
s
i
f
icatio
n
er
r
o
r
,
f
o
r
OAO
an
d
O
AA
p
er
f
o
r
m
a
n
ce
ev
alu
a
tio
n
.
T
h
e
r
esu
lt
s
r
ev
ea
l
ed
th
e
ab
ilit
y
of
g
e
n
er
aliz
in
g
m
o
r
e
k
er
n
el
s
an
d
it
r
ev
ea
led
th
at
t
h
e
p
o
l
y
n
o
m
ial
i
s
o
u
tp
er
f
o
r
m
t
h
e
o
th
er
k
er
n
el
f
u
n
ct
io
n
s
i
n
SV
M
.
I
n
[
2
0
]
th
e
au
t
h
o
r
s
p
r
esen
ted
an
as
s
ess
m
e
n
t
s
tu
d
y
in
c
lu
d
es
f
i
v
e
k
er
n
el
f
u
n
ctio
n
s
w
it
h
SV
M
in
ter
m
of
s
e
n
s
it
iv
i
t
y
c
lass
if
icatio
n
b
a
s
ed
o
n
P
OS
s
eq
u
en
ce
s
.
T
h
e
e
v
alu
a
ted
k
er
n
el
w
er
e
L
i
n
ea
r
,
Sp
ec
tr
u
m
,
Gau
s
s
ia
n
,
S
m
it
h
-
W
ater
m
a
n
an
d
Mi
s
m
a
t
ch
.
T
h
is
p
ap
er
ex
p
lain
ed
th
at
th
e
co
m
b
in
a
tio
n
o
f
P
OS
s
eq
u
en
ce
cla
s
s
i
f
icatio
n
an
d
te
x
t
clas
s
if
icatio
n
w
a
s
en
h
a
n
ce
t
h
e
e
f
f
ec
tiv
e
n
e
s
s
o
f
s
en
s
iti
v
it
y
clas
s
i
f
icatio
n
.
T
h
e
au
th
o
r
s
i
n
[
2
1
]
p
r
esen
ted
a
s
t
u
d
y
o
f
SVM
p
er
f
o
r
m
a
n
ce
w
it
h
d
i
f
f
er
e
n
t
k
er
n
e
ls
.
T
h
e
r
esu
lt
s
s
h
o
w
ed
th
e
d
ataset
d
escr
ip
tio
n
a
n
d
th
e
i
m
p
le
m
e
n
ted
k
er
n
el
i
m
p
ac
t th
e
ac
cu
r
ac
y
o
f
t
h
e
SVM
cla
s
s
i
f
ier
.
3.
SUPP
O
RT
V
E
CT
O
R
M
ACH
I
NE
S
(
S
VM
)
SVM
i
s
m
ac
h
i
n
e
lear
n
i
n
g
an
d
it is
r
e
g
ar
d
ed
as a
p
ar
t o
f
n
e
u
r
al
n
et
w
o
r
k
s
t
h
at
u
s
ed
in
d
ata
m
i
n
in
g
a
n
d
p
atter
n
r
ec
o
g
n
itio
n
.
T
h
e
m
ai
n
co
n
ce
p
t
o
f
SVM
is
illu
s
tr
ate
d
in
F
i
g
u
r
e
1
.
SVM
tech
n
iq
u
es
ca
n
b
e
ex
p
lain
ed
as
a
p
r
o
ce
s
s
o
f
s
ep
ar
atin
g
t
w
o
d
i
f
f
er
en
t
cla
s
s
e
s
in
f
ea
t
u
r
e
s
p
ac
e
s
u
c
h
as
p
o
s
iti
v
e
an
d
n
e
g
ati
v
e.
Dis
co
v
er
i
n
g
th
e
h
y
p
er
p
la
n
e
t
h
at
s
ep
ar
ates
th
o
s
e
c
lass
e
s
i
s
th
e
m
ai
n
is
s
u
e
in
c
lass
if
ica
ti
o
n
p
r
o
ce
s
s
w
h
er
e
th
is
h
y
p
er
p
lan
e
is
d
ep
en
d
in
g
o
n
m
a
x
i
m
al
m
ar
g
i
n
.
Sev
e
r
al
o
p
ti
m
iza
tio
n
p
r
o
b
le
m
s
ca
n
b
e
tack
led
b
y
SVM
s
u
ch
a
s
r
eg
r
es
s
io
n
i
s
s
u
e
s
an
d
d
ata
class
i
f
icat
io
n
is
s
u
es
.
Dis
tin
g
u
is
h
t
h
e
p
o
s
iti
v
e
an
d
n
eg
at
iv
e
d
ata
p
o
i
n
t
s
r
e
q
u
i
r
e
s
i
d
e
n
t
i
f
y
a
h
y
p
e
r
-
p
l
a
n
e
t
o
s
e
p
a
r
a
t
e
t
h
e
d
a
t
a
p
o
i
n
t
s
a
s
t
w
o
c
l
a
s
s
e
s
.
F
i
g
u
r
e
1
s
h
o
w
s
t
h
e
h
y
p
e
r
p
l
a
n
e
th
a
t
it
ex
p
ec
ted
as
d
ec
is
io
n
b
o
u
n
d
ar
y
f
o
r
lin
ea
r
SVM
.
Fig
u
r
e
1
.
Data
c
lass
i
f
icatio
n
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
10
,
No
.
5
,
Octo
b
e
r
2
0
2
0
:
5
4
9
7
-
5506
5500
SVM
h
a
s
t
h
r
ee
t
y
p
es,
w
h
ic
h
a
r
e
lin
ea
r
SVM
s
,
n
o
n
li
n
ea
r
SV
M
s
an
d
m
u
lticla
s
s
S
VM
s
.
L
i
n
ea
r
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
i
n
es
tac
k
les
th
e
i
s
s
u
e
o
f
b
in
ar
y
cla
s
s
i
f
icatio
n
o
f
tar
g
eted
d
ata
p
o
in
t
s
b
y
d
iv
i
d
in
g
th
e
m
in
to
t
w
o
class
es.
Fig
u
r
e
1
s
h
o
w
ed
p
o
in
ts
o
f
d
ata
ar
e
li
n
ea
r
l
y
s
ep
ar
ab
le
(
t
w
o
-
d
i
m
e
n
s
io
n
al
ca
s
e)
.
W
h
en
t
h
e
t
w
o
clas
s
es
ar
e
n
o
n
-
li
n
ea
r
l
y
s
ep
ar
ab
le,
it
ca
lled
n
o
n
-
s
ep
ar
ab
le
.
T
h
u
s
,
d
eter
m
i
n
in
g
th
e
h
y
p
er
p
lan
e
s
p
ec
if
y
t
h
e
ab
ilit
y
o
f
class
i
f
y
th
e
d
ata
p
o
in
ts
l
in
ea
r
l
y
o
r
n
o
n
li
n
ea
r
l
y
(
ca
u
s
ed
b
y
,
f
o
r
ex
a
m
p
le
n
o
i
s
e)
[
2
1
]
.
W
h
en
li
n
ea
r
SVM
is
n
o
t
p
r
o
p
er
f
o
r
th
e
u
s
ed
d
atase
t
,
t
h
e
No
n
l
in
ea
r
S
VM
s
p
r
o
v
id
e
s
s
o
lu
tio
n
.
T
h
e
co
n
ce
p
t
o
f
li
n
ea
r
SVM
h
as
b
ee
n
e
x
te
n
d
ed
to
tack
le
th
e
n
o
n
l
in
ea
r
ca
s
e.
T
h
e
m
ai
n
c
o
n
ce
p
t
o
f
n
o
n
li
n
ea
r
SVM
is
to
lo
ca
te
an
o
p
ti
m
al
s
ep
ar
atin
g
h
y
p
er
p
la
n
e
in
h
i
g
h
-
d
i
m
e
n
s
io
n
al
f
ea
t
u
r
e
s
p
ac
e
s
i
m
ilar
to
th
e
l
in
ea
r
SVM
i
n
in
p
u
t
s
p
ac
e.
T
h
is
i
m
p
le
m
e
n
t
ed
s
a
m
e
as
in
li
n
ea
r
S
V
M
b
u
t
in
s
tead
o
f
co
m
p
u
ti
n
g
t
h
e
in
n
er
p
r
o
d
u
cts
i
n
f
e
at
u
r
e
s
p
a
c
e
,
w
h
i
ch
w
o
u
l
d
b
e
c
o
m
p
u
t
a
ti
o
n
a
l
ly
ex
p
e
n
s
iv
e
b
e
ca
u
s
e
o
f
i
ts
h
ig
h
d
im
en
s
i
o
n
al
i
ty
.
T
h
i
s
c
o
m
p
u
t
a
t
i
o
n
s
a
r
e
p
e
r
f
o
r
m
e
d
b
y
u
s
in
g
a
n
o
n
l
in
e
a
r
k
e
r
n
el
f
u
n
c
t
i
o
n
,
in
in
p
u
t sp
a
c
e
,
w
h
i
ch
h
e
l
p
s
s
p
ee
d
u
p
th
e
c
o
m
p
u
ta
t
i
o
n
s
[
2
2
]
.
Ker
n
el
f
u
n
ctio
n
s
ar
e
i
m
p
le
m
en
ted
w
ith
S
VM
to
s
o
lv
e
m
an
y
i
s
s
u
es
s
u
c
h
as
p
atter
n
a
n
al
y
s
i
s
.
Gen
er
all
y
,
p
atter
n
an
al
y
s
is
ca
n
h
elp
in
d
eter
m
i
n
in
g
an
d
s
t
u
d
y
i
n
g
v
ar
i
o
u
s
t
y
p
es
o
f
r
elatio
n
s
in
d
ataset
s
u
c
h
as
clu
s
ter
s
,
class
i
f
i
ca
tio
n
s
,
p
r
in
cip
al
co
m
p
o
n
en
t
s
,
r
an
k
i
n
g
s
a
n
d
co
r
r
elatio
n
s
.
Mu
lticla
s
s
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
e
s
ar
e
s
er
ies
o
f
b
i
n
ar
y
p
r
o
b
le
m
s
a
n
d
it
h
as
t
w
o
t
y
p
es.
Fi
r
s
t
n
a
m
ed
as
o
n
e
-
v
er
s
u
s
-
r
est
w
h
ich
d
iv
i
d
es
th
e
K
-
clas
s
p
r
o
b
le
m
in
to
K
b
in
ar
y
cla
s
s
i
f
icatio
n
s
u
b
-
p
r
o
b
lem
s
o
f
t
h
e
t
y
p
e
“k
t
h
cla
s
s
”
v
s
.
“n
o
t
k
t
h
cla
s
s
,
”
k
=
1
,
2
,
.
.
.
,
K.
Seco
n
d
ty
p
e
is
ca
lled
o
n
e
-
v
er
s
u
s
-
o
n
e,
w
h
ic
h
d
iv
id
es
th
e
K
-
cla
s
s
p
r
o
b
le
m
i
n
to
co
m
p
ar
is
o
n
s
o
f
all
p
air
s
o
f
clas
s
e
s
.
C
o
n
s
tr
u
c
tin
g
a
tr
u
e
m
u
lticl
ass
S
VM
clas
s
i
f
ier
,
r
eq
u
ir
e
co
n
s
id
er
in
g
all
K
cla
s
s
es
s
i
m
u
ltan
eo
u
s
l
y
.
I
n
ad
d
it
io
n
,
th
e
cla
s
s
i
f
ier
m
u
s
t
r
ed
u
c
e
t
h
e
b
in
ar
y
S
VM
class
i
f
ier
i
f
K=
2
[
22
].
T
h
is
p
ap
er
f
o
cu
s
in
g
o
n
i
n
v
esti
g
ati
n
g
t
h
e
n
o
n
li
n
ea
r
s
u
p
p
o
r
t v
ec
to
r
m
ac
h
in
e
s
w
it
h
k
er
n
e
l
f
u
n
ctio
n
s
.
3
.
1
.
G
ener
a
l
pro
ble
m
f
o
r
m
ula
t
io
n
Gen
er
ati
n
g
a
n
o
n
-
lin
ea
r
s
ep
ar
ato
r
is
th
e
m
ai
n
co
n
ce
p
t
o
f
S
VM
.
T
h
e
task
o
f
m
ap
p
in
g
d
ata
to
h
ig
h
d
i
m
en
s
io
n
al
s
p
ac
e
is
p
er
f
o
r
m
ed
to
f
ac
ilit
ate
th
e
clas
s
i
f
icatio
n
o
f
d
ata
in
lin
ea
r
d
ec
is
io
n
s
u
r
f
ac
es.
Hen
ce
,
r
ef
o
r
m
u
lat
io
n
d
ata
is
p
er
f
o
r
m
ed
to
i
m
p
lici
tl
y
m
ap
p
ed
d
ata
to
th
is
s
p
ac
e.
R
ef
o
r
m
u
lat
io
n
tas
k
is
s
h
o
w
ed
in
F
i
g
u
r
e
2
.
Fig
u
r
e
2
.
Data
class
i
f
icatio
n
u
s
in
g
k
er
n
el
T
h
is
tr
an
s
f
o
r
m
atio
n
is
p
er
f
o
r
m
ed
b
y
k
er
n
el
f
u
n
ctio
n
s
,
s
o
th
e
d
ata
is
tr
an
s
f
o
r
m
ed
in
to
a
s
p
ac
e
o
f
K
(
X,
Y)
w
h
ic
h
is
h
ig
h
er
d
i
m
e
n
s
io
n
al
f
ea
t
u
r
e
b
y
k
er
n
e
l
f
u
n
c
tio
n
to
f
ac
ilit
ate
t
h
e
p
r
o
ce
s
s
o
f
lin
ea
r
s
ep
ar
atio
n
.
SVMs
can
b
e
i
m
p
le
m
e
n
ted
as
b
o
th
lin
ea
r
class
if
ier
a
n
d
n
o
n
-
li
n
ea
r
class
i
f
ier
.
N
on
-
li
n
ea
r
class
if
ier
m
u
s
t
i
m
p
le
m
en
t
s
p
ec
if
ied
k
er
n
el
f
u
n
ct
io
n
to
tr
an
s
f
o
r
m
t
h
e
p
o
in
ts
of
d
atase
t
in
to
h
ig
h
er
d
im
en
s
io
n
a
l
o
f
f
ea
t
u
r
e
s
p
ac
e
[
23
]
.
3
.
2
.
K
er
nel f
un
ct
io
ns
o
f
SVM
SVM
is
a
tech
n
iq
u
e
o
f
d
is
co
v
er
in
g
b
est
b
o
u
n
d
ar
y
to
s
p
ec
if
y
th
e
class
es
o
f
s
o
m
e
f
ea
t
u
r
es
an
d
o
u
tlier
d
etec
tio
n
.
SVM
is
n
a
m
ed
as
id
ea
l
SVM
w
h
e
n
t
h
e
a
n
al
y
s
i
s
r
esu
lt
s
i
n
a
h
y
p
er
-
p
la
n
e
t
h
at
ab
le
to
co
m
p
letel
y
s
ep
ar
ate
th
e
d
ata
p
o
in
ts
in
to
t
w
o
d
if
f
er
en
t
clas
s
es
.
Ker
n
el
f
u
n
ctio
n
s
o
f
f
er
t
h
e
ab
ili
t
y
o
f
co
m
p
u
ti
n
g
d
o
t
p
r
o
d
u
cts
in
h
i
g
h
er
d
i
m
e
n
s
io
n
al
s
p
ac
es
w
it
h
o
u
t
t
h
e
n
ee
d
t
o
ex
p
licitl
y
m
ap
p
i
n
g
i
n
to
th
e
s
e
s
p
ac
es.
Var
io
u
s
m
at
h
e
m
a
tical
p
r
o
p
er
ties
m
u
s
t
b
e
s
atis
f
ied
b
y
a
k
er
n
e
l
f
u
n
c
tio
n
.
On
e
o
f
th
e
i
m
p
o
r
ta
n
t
is
s
u
es
is
d
eter
m
i
n
i
ng
w
h
et
h
er
a
g
iv
e
n
f
u
n
ct
io
n
k
is
a
k
er
n
el
f
u
n
c
tio
n
o
r
n
o
t
?
.
G
i
v
e
n
x
1
,
x
2
,
…
x
m
∈
X
,
t
h
e
(
m
×
m
)
m
a
t
r
i
x
K
w
i
t
h
e
l
e
m
e
n
t
s
K
i
j
=
k
(
x
i
,
x
j
)
i
s
n
a
m
e
d
a
s
t
h
e
G
r
a
m
m
a
t
r
i
x
(
o
r
k
e
r
n
e
l
m
a
t
r
i
x
)
.
B
a
s
e
d
o
n
M
e
r
c
e
r
'
s
T
h
e
o
r
e
m
,
k
i
s
a
k
e
r
n
e
l
i
f
t
h
e
G
r
a
m
m
a
t
r
i
x
i
s
p
o
s
i
t
i
v
e
d
e
f
i
n
i
t
e
[
2
4
]
.
SVM
i
m
p
le
m
en
ts
s
ev
er
al
k
er
n
el
s
to
s
o
l
v
e
d
i
f
f
er
e
n
t
i
s
s
u
es
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2
0
8
8
-
8708
Th
e
effec
t o
f g
a
mma
va
l
u
e
o
n
s
u
p
p
o
r
t v
ec
to
r
ma
ch
in
e
…
(
I
n
t
is
a
r
S
h
a
d
ee
d
A
l
-
Mejib
li
)
5501
T
h
e
g
a
m
m
a
p
ar
a
m
eter
i
s
u
s
ed
in
s
o
m
e
o
f
th
e
s
e
k
er
n
el
s
.
T
h
e
m
o
s
t
co
m
m
o
n
k
er
n
el
s
t
h
at
u
s
ed
th
is
p
ar
a
m
eter
ar
e
p
o
ly
n
o
m
ial,
r
ad
ial
b
asis
f
u
n
ctio
n
(
R
B
F)
a
n
d
s
ig
m
o
id
as e
x
p
lain
ed
b
elo
w
[
2
5
,
2
6
].
Ker
n
el
F
u
n
ct
io
n
s
:
(
γ
X•
Y+
c)
ɗ
,
γ
>
0
Po
ly
n
o
m
i
al
K(
X
,
Y)
=
ex
p
(
-
γ
\
X•
Y
|
2
)
,
γ
>
0
R
B
F
(
1
)
tan
h
(
γ
X•
Y+
c)
Sig
m
o
id
w
h
er
e
X
an
d
Y
act
as
th
e
v
ec
t
o
r
s
in
th
e
in
p
u
t
s
p
ac
e,
γ
r
ep
r
e
s
en
t
s
th
e
g
a
m
m
a
p
ar
a
m
eter
in
th
e
k
er
n
el
f
u
n
ctio
n
an
d
c
is
th
e
b
ias
p
ar
a
m
eter
i
n
th
e
k
er
n
el
f
u
n
ctio
n
.
γ
,
d
,
an
d
c,
ar
e
u
s
er
-
co
n
tr
o
lled
p
ar
a
m
et
er
s
.
Ga
m
m
a
v
al
u
e
is
ca
lcu
lated
b
y
d
iv
id
in
g
1
to
n
u
m
b
er
o
f
f
ea
t
u
r
ed
in
th
e
e
m
p
lo
y
ed
d
ataset.
E
ac
h
k
er
n
el
f
u
n
ct
io
n
h
a
s
s
p
ec
if
ic
p
ar
am
eter
s
th
at
n
ee
d
o
p
ti
m
iz
atio
n
to
ac
h
ie
v
e
th
e
b
es
t
r
es
u
lts
[
2
7
]
.
Her
e
th
ese
p
ar
a
m
e
ter
s
ar
e
n
a
m
ed
as
o
p
tim
ized
p
ar
am
eter
s
.
T
h
e
th
r
ee
in
v
e
s
ti
g
ated
k
er
n
e
ls
ar
e
as f
o
llo
w
s
[
28
-
30
]:
3
.
2
.
1
.
P
o
ly
no
m
ia
l
k
er
nel
So
m
eti
m
es
n
o
is
e
i
n
d
ata
o
r
b
a
d
f
ea
tu
r
e
r
ep
r
esen
tatio
n
r
esu
lt
in
n
o
t p
o
s
s
ib
ili
t
y
o
f
lin
ea
r
s
ep
ar
ab
le
f
o
r
d
ata
.
T
o
tack
le
th
is
i
s
s
u
e
th
e
d
ata
p
o
in
ts
is
m
ap
p
ed
in
to
d
if
f
e
r
en
t sp
ac
e.
P
o
ly
n
o
m
ial
k
er
n
e
l f
u
n
ctio
n
f
o
r
b
o
th
h
ar
d
m
ar
g
i
n
an
d
s
o
f
t
m
ar
g
i
n
a
r
e
to
d
ea
l w
it
h
n
o
n
li
n
ea
r
s
ep
ar
ab
le
p
atter
n
s
.
P
o
ly
n
o
m
ial
k
er
n
el
f
u
n
ctio
n
ap
p
lied
b
y
u
s
in
g
m
ath
e
m
atica
l
f
u
n
ctio
n
as e
x
p
r
e
s
s
ed
in
(
2
)
,
K(
X
,
Y)
=
(
X•
Y+
c)
ɗ
γ
>0
(
2
)
w
h
er
e
c
≥
0
ac
ts
as
a
f
r
ee
p
ar
a
m
eter
in
t
h
e
p
o
l
y
n
o
m
ial
t
h
at
tr
ad
in
g
-
o
f
f
t
h
e
i
m
p
ac
t
o
f
h
ig
h
er
-
o
r
d
er
v
er
s
u
s
lo
w
er
-
o
r
d
er
ter
m
s
[
16
,
23
,
30
]
.
d
,
γ
,
c
an
d
ɗ
ar
e
th
e
o
p
tim
iza
tio
n
p
ar
a
m
eter
s
[
31
,
3
2
].
T
h
e
p
o
ly
n
o
m
ia
l
i
s
r
eg
ar
d
ed
as
n
o
n
-
s
tatio
n
ar
y
k
er
n
el.
I
t
is
ap
p
r
o
p
r
iate
to
b
e
u
s
ed
in
t
h
e
is
s
u
es,
w
h
ic
h
all
o
f
its
tr
ai
n
i
n
g
s
a
m
p
le
s
ar
e
n
o
r
m
alize
d
.
T
h
e
s
lo
p
e
g
a
m
m
a
p
ar
am
eter
s
h
o
u
ld
b
e
s
ettled
w
it
h
t
h
i
s
k
er
n
el.
c
i
s
th
e
co
n
s
tan
t te
r
m
an
d
d
is
t
h
e
p
o
ly
n
o
m
ia
l d
eg
r
e
e
(
w
h
er
e
d
=3
,
c=
0
)
.
3
.
2
.
2
.
RB
F
k
er
nel
T
h
e
R
ad
ial
b
asi
s
f
u
n
ctio
n
al
s
o
ca
lled
th
e
R
B
F
k
er
n
e
l,
o
r
Gau
s
s
ia
n
k
er
n
el
.
I
t
is
ap
p
lied
b
y
u
s
i
n
g
m
at
h
e
m
a
tical
f
u
n
ctio
n
t
h
at
e
x
p
r
ess
ed
in
(
3
)
:
K
(X
,
Y)
=e
x
p
(
-
γ
|
X
-
Y
|
2
)
γ
>0
(
3
)
T
o
s
et
th
e
“
s
p
r
ea
d
”
o
f
th
e
k
e
r
n
el
,
th
e
γ
p
ar
a
m
eter
is
u
s
ed
[
16
,
23
,
33
]
.
T
h
e
ad
j
u
s
tab
le
p
ar
a
m
eter
γ
m
u
s
t
ca
r
ef
u
ll
y
t
u
n
ed
as
it
s
i
g
n
i
f
ica
n
t
i
m
p
ac
t
o
n
t
h
is
k
er
n
el
p
er
f
o
r
m
a
n
ce
.
T
h
e
ex
p
o
n
en
tial
b
eh
a
v
io
r
w
ill
al
m
o
s
t
c
h
an
g
e
to
lin
ea
r
l
y
i
f
th
is
p
ar
a
m
e
ter
is
o
v
er
esti
m
at
ed
.
T
h
is
also
w
ill
m
a
k
e
t
h
e
h
ig
h
er
d
i
m
e
n
s
io
n
a
l
p
r
o
j
ec
tio
n
lo
s
in
g
it
s
n
o
n
-
li
n
ea
r
p
o
w
er
.
R
B
F
k
er
n
el
s
u
f
f
er
s
f
r
o
m
lack
r
eg
u
lar
izatio
n
an
d
t
h
e
d
ec
is
io
n
b
o
u
n
d
ar
ies
te
n
d
to
b
ec
o
m
e
h
i
g
h
l
y
s
en
s
iti
v
e
to
n
o
i
s
e
i
n
tr
ai
n
in
g
d
at
a,
i
f
u
n
d
er
esti
m
ated
.
T
h
u
s
,
th
e
v
al
u
e
o
f
γ
p
ar
am
eter
m
u
s
t b
e
ca
r
ef
u
ll
y
c
h
o
s
en
a
s
it in
f
l
u
e
n
ce
th
e
p
er
f
o
r
m
an
ce
o
f
SVM
[
31
].
3
.
2
.
3
.
Sig
m
o
id
k
er
nel
As
m
e
n
tio
n
ed
p
r
ev
io
u
s
l
y
,
e
ac
h
k
er
n
el
m
u
s
t
b
e
p
o
s
iti
v
e
d
ef
in
i
te
to
s
ati
s
f
y
Me
r
ce
r
’
s
th
eo
r
e
m
.
A
lt
h
o
u
g
h
,
th
e
Sig
m
o
id
k
er
n
el
f
u
n
ctio
n
is
co
m
m
o
n
l
y
i
m
p
le
m
en
ted
,
it
is
n
o
t
p
o
s
iti
v
e
s
e
m
i
-
d
ef
i
n
ite
f
o
r
s
p
ec
i
f
ic
v
alu
e
s
o
f
t
h
is
k
er
n
el
p
ar
a
m
e
t
er
s
.
C
o
n
s
eq
u
e
n
tl
y
,
t
h
e
p
ar
a
m
eter
s
γ
an
d
c
m
u
s
t
b
e
ca
r
ef
u
ll
y
s
e
lecte
d
to
av
o
id
an
y
m
i
s
tak
e
i
n
t
h
e
o
b
tain
ed
r
esu
lt
s
.
Ma
th
e
m
atica
l
f
u
n
c
tio
n
o
f
s
i
g
m
o
id
k
er
n
el
p
r
ese
n
ted
in
(
4
)
:
K(
X
,
Y)
=
tan
h
(
γ
X
•
Y
+
c)
(
4
)
w
h
er
e
γ
ac
ts
as
a
s
ca
li
n
g
p
ar
a
m
eter
o
f
t
h
e
in
p
u
t
d
ata
an
d
c
ac
ts
as
a
s
h
if
t
in
g
p
ar
a
m
eter
,
w
h
ich
is
s
u
p
er
v
i
s
e
th
e
m
ap
p
in
g
th
r
es
h
o
ld
(
h
e
n
ce
c
=0
)
[
1
5
,
1
6
]
.
γ
an
d
c
ar
e
th
e
o
p
t
i
m
izatio
n
p
ar
a
m
eter
s
[
30
].
4.
DATAS
E
T
S
UC
I
r
v
i
n
e
Ma
ch
i
n
e
L
ea
r
n
i
n
g
R
ep
o
s
ito
r
y
[
34
]
is
t
h
e
s
o
u
r
ce
o
f
t
h
e
u
s
ed
d
atase
ts
in
t
h
is
s
t
u
d
y
.
W
h
er
e
f
i
v
e
d
i
f
f
er
en
t
d
ataset
s
h
a
v
e
b
ee
n
tr
an
s
f
o
r
m
ed
to
n
u
m
er
ic
an
d
i
m
p
le
m
e
n
ted
i
n
th
is
i
n
v
e
s
ti
g
atio
n
.
T
ab
le
1
s
h
o
w
s
t
h
e
u
s
ed
d
atase
ts
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
10
,
No
.
5
,
Octo
b
e
r
2
0
2
0
:
5
4
9
7
-
5506
5502
T
h
e
d
o
w
n
lo
ad
ed
d
atasets
ar
e:
Mu
s
h
r
o
o
m
,
C
h
r
o
n
ic
Kid
n
e
y
Dis
ea
s
e,
B
r
ea
s
t
ca
n
ce
r
,
L
u
n
g
ca
n
ce
r
an
d
H
e
a
r
t
s
t
a
t
l
o
g
.
T
h
r
e
e
k
e
r
n
e
l
s
w
i
t
h
S
V
M
w
e
r
e
a
c
h
i
e
v
e
d
f
o
r
a
l
l
t
h
e
a
f
o
r
e
m
e
n
t
i
o
n
e
d
d
a
t
a
s
e
t
s
.
T
a
b
l
e
1
d
e
s
c
r
i
b
e
s
s
o
m
e
f
e
a
t
u
r
e
s
o
f
t
h
e
d
a
t
a
s
e
t
s
u
s
e
d
s
u
c
h
a
s
a
t
t
r
i
b
u
t
e
s
,
i
n
s
t
a
n
c
e
s
,
c
l
a
s
s
e
s
i
n
a
d
d
i
t
i
o
n
t
o
m
i
s
s
i
n
g
v
a
l
u
e
s
(
i
f
e
x
i
s
t
s
)
.
W
h
er
e
th
e
d
escr
ib
ed
f
ea
tu
r
es i
n
T
ab
le
1
w
ill
u
s
ed
in
a
s
s
e
s
s
m
en
t ta
s
k
later
.
T
ab
le
1
.
T
h
e
d
atasets
d
escr
ip
tio
n
N
a
me
N
o
.
o
f
A
t
t
r
i
b
u
t
e
s
N
o
.
o
f
I
n
st
a
n
c
e
s
N
o
.
o
f
C
l
a
ss
M
i
ss
i
n
g
i
n
v
a
l
e
A
t
t
r
i
b
u
t
e
s
1
M
u
s
h
r
o
o
m
22
8
1
2
4
2
Y
e
s
2
C
h
r
o
n
i
c
K
i
d
n
e
y
D
i
se
a
se
25
4
0
0
2
Y
e
s
3
B
r
e
a
st
c
a
n
c
e
r
9
2
8
6
2
Y
e
s
4
L
u
n
g
c
a
n
c
e
r
6
59
2
Y
e
s
5
H
e
a
r
t
st
a
t
l
o
g
13
2
7
0
2
No
5.
RE
SU
L
T
T
o
s
tar
t
th
e
tr
ai
n
in
g
an
d
te
s
ti
n
g
p
h
ase
s
o
f
S
VM
,
ea
ch
o
f
t
h
e
u
s
ed
d
atase
ts
h
a
v
e
b
ee
n
d
iv
id
ed
in
t
o
t
w
o
s
et
s
:
tr
ain
in
g
s
et
a
n
d
test
in
g
s
et
b
ased
o
n
th
eir
n
u
m
b
e
r
o
f
in
s
ta
n
ce
s
.
T
ab
le
2
s
h
o
w
e
d
tr
ain
in
g
s
ets
a
n
d
test
i
n
g
s
e
ts
o
f
u
s
ed
d
ataset
s
.
T
ab
le
2
.
T
h
e
s
iz
e
o
f
tr
ain
in
g
s
et
an
d
test
i
n
g
s
et
D
a
t
a
se
t
T
r
a
i
n
i
n
g
S
i
z
e
T
e
st
i
n
g
S
i
z
e
M
u
s
h
r
o
o
m
5
6
8
6
2
4
3
8
C
h
r
o
n
i
c
K
i
d
n
e
y
2
8
0
1
2
0
B
r
e
a
st
c
a
n
c
e
r
2
0
0
86
L
u
n
g
c
a
n
c
e
r
41
18
H
e
a
r
t
st
a
t
l
o
g
1
8
9
81
As
af
o
r
e
m
en
t
io
n
ed
th
e
g
a
m
m
a
v
alu
e
is
ca
lc
u
lated
b
y
d
iv
id
in
g
1
to
n
u
m
b
er
o
f
f
e
atu
r
ed
i
n
th
e
e
m
p
lo
y
ed
d
ataset.
γ
(
g
a
m
m
a)
=
1
/No
.
o
f
Featu
r
e
s
(
5
)
Fig
u
r
e
3
s
h
o
w
s
t
h
e
f
lo
w
c
h
ar
t
o
f
u
s
ed
al
g
o
r
ith
m
to
as
s
es
s
th
e
i
m
p
le
m
en
ted
k
er
n
el
s
w
i
th
d
if
f
er
e
n
t
v
alu
e
s
o
f
g
a
m
m
a
p
ar
a
m
eter
.
T
h
e
d
ataset
r
ep
r
esen
t
s
t
h
e
in
p
u
t
to
p
r
o
p
o
s
ed
alg
o
r
ith
m
,
w
h
er
e
i
t
w
il
l
b
e
class
i
f
ied
b
y
ap
p
ly
in
g
th
e
S
VM
w
it
h
t
h
r
ee
k
er
n
el
s
.
T
h
en
,
th
e
r
esu
lted
v
alid
it
y
an
d
ac
cu
r
ac
y
ar
e
ch
ec
k
ed
.
W
h
er
e
d
if
f
er
e
n
t v
al
u
e
s
o
f
g
a
m
m
a
p
ar
a
m
eter
ar
e
i
m
p
le
m
en
te
d
w
it
h
k
er
n
els
to
s
p
ec
i
f
y
t
h
e
h
y
p
er
p
la
n
e.
A
p
p
l
y
i
n
g
t
h
e
S
V
M
s
u
c
c
e
s
s
f
u
l
l
y
r
e
q
u
i
r
e
s
e
l
e
c
t
t
h
e
p
r
o
p
e
r
p
a
r
am
e
t
e
r
s
t
h
a
t
S
V
M
a
n
d
k
e
r
n
e
l
f
u
n
c
t
i
o
n
s
n
e
e
d
.
Fo
r
ex
a
m
p
le
t
h
e
k
er
n
el
w
id
t
h
g
a
m
m
a
(
γ
)
,
t
h
e
co
n
s
ta
n
t
p
ar
a
m
eter
c
w
h
ich
is
co
n
s
tan
t
i
n
t
h
e
k
er
n
el
f
u
n
ctio
n
an
d
k
er
n
el
d
e
g
r
ee
(
d
)
.
T
h
e
f
o
llo
w
i
n
g
p
ar
a
m
eter
s
’
v
al
u
es
er
e
ap
p
lied
in
th
r
ee
k
er
n
el
f
u
n
ctio
n
s
:
γ
:
I
ts
v
alu
e
i
s
(
0
.
3
,
0
.
6
,
0
.
9
)
.
c
:
(
c=
0
)
d
: k
er
n
el
(
d
=
3
)
.
T
o
v
alid
ate
all
i
m
p
le
m
en
ted
ex
p
er
i
m
e
n
ts
,
w
e
u
s
ed
1
0
-
f
o
ld
c
r
o
s
s
v
alid
atio
n
.
T
ab
le
3
s
h
o
w
s
t
h
e
ac
h
ie
v
ed
ac
cu
r
ac
y
o
f
Si
g
m
o
id
,
P
o
ly
n
o
m
ial
a
n
d
R
b
f
w
h
en
th
e
g
a
m
m
a
v
alu
e
i
s
ca
lcu
lated
b
ased
o
n
(
5
).
T
h
r
ee
d
if
f
er
en
t
v
al
u
es
o
f
g
a
m
m
a
(
0
.
3
,
0
.
6
,
0
.
9
)
a
r
e
u
s
ed
in
th
e
th
r
ee
k
er
n
el
f
u
n
ctio
n
s
th
at
ar
e
i
m
p
l
e
m
e
n
t
e
d
w
i
t
h
S
V
M
a
p
p
l
i
e
d
t
o
t
h
e
f
i
v
e
d
a
t
a
s
e
t
s
.
T
h
e
s
e
v
a
l
u
e
s
h
a
v
e
b
e
e
n
p
r
o
p
o
s
e
d
,
t
o
b
e
c
l
o
s
e
t
o
t
h
e
g
a
m
m
a
v
a
l
u
e
b
a
s
e
d
o
n
e
q
u
a
t
i
o
n
5
a
n
d
t
o
m
e
a
s
u
r
e
t
h
e
a
c
h
i
e
v
e
d
a
c
c
u
r
a
c
y
w
h
e
n
t
h
e
g
am
m
a
v
a
l
u
e
a
r
e
c
h
a
n
g
e
d
.
T
h
e
r
esu
lt
ed
clas
s
i
f
icatio
n
a
cc
u
r
ac
y
o
f
i
m
p
le
m
e
n
ted
k
er
n
els
o
n
d
ataset
s
ar
e
a
f
f
ec
te
d
b
y
g
a
m
m
a
v
alu
e
ch
an
g
i
n
g
u
n
ev
e
n
l
y
.
T
h
e
F
ig
u
r
e
s
4
-
6
p
r
esen
t
t
h
e
SVM
ac
u
u
r
ac
y
w
it
h
p
o
l
y
n
o
m
ial,
r
ad
ial
b
asis
f
u
n
ctio
n
(
R
B
F)
an
d
s
i
g
m
o
id
r
esp
ec
ti
v
el
y
d
ep
en
d
in
g
o
n
t
h
e
u
s
ed
g
a
m
m
a
v
al
u
e.
I
n
g
e
n
er
al,
t
h
e
r
es
u
lt
s
ca
n
b
e
d
escr
ib
ed
d
ep
en
d
in
g
o
n
th
r
ee
f
ac
to
r
s
,
w
h
ich
ar
e
th
e
ch
a
n
g
in
g
in
t
h
e
v
alu
e
o
f
g
a
m
m
a
p
ar
am
eter
,
e
m
p
lo
y
ed
d
ataset
an
d
u
s
ed
k
er
n
el
f
u
n
ctio
n
.
T
h
e
P
o
l
y
n
o
m
ial
k
er
n
el
ac
h
iev
ed
b
est
r
esu
lts
w
h
e
n
t
h
e
g
a
m
m
a
v
al
u
e
is
ca
lc
u
lated
as
in
(
5
)
w
it
h
all
t
h
e
e
m
p
lo
y
ed
d
atasets
e
x
ce
p
t
th
e
C
h
r
o
n
ic
Ki
d
n
e
y
d
ataset,
w
h
ic
h
s
h
o
w
s
n
o
ch
an
g
e
i
n
o
b
tain
ed
ac
cu
r
ac
y
.
Fro
m
d
atase
t
p
o
in
t
o
f
v
ie
w
,
th
e
P
o
l
y
n
o
m
ial
Ker
n
el
ac
h
ie
v
e
d
b
est
r
es
u
lt
s
in
ac
cu
r
ac
y
w
h
e
n
th
e
Mu
s
h
r
o
o
m
d
ata
s
et
is
e
m
p
l
o
y
ed
as t
h
e
ac
h
iev
ed
ac
cu
r
ac
y
w
as
n
o
less
t
h
a
n
9
9
.
9
%
.
T
h
e
ac
cu
r
ac
y
o
f
R
B
F
k
er
n
el
f
u
n
ct
io
n
i
s
a
f
f
ec
ted
s
li
g
h
tl
y
w
h
e
n
t
h
e
g
a
m
m
a
v
al
u
e
i
s
c
h
an
g
ed
as
s
h
o
w
n
i
n
F
i
g
u
r
e
5
.
T
h
e
ac
cu
r
ac
y
o
f
R
B
F
k
er
n
el
w
h
e
n
C
h
r
o
n
ic
Kid
n
e
y
a
n
d
L
u
n
g
ca
n
ce
r
d
atasets
ar
e
u
s
ed
s
h
o
w
n
o
i
m
p
ac
t
w
i
th
th
e
ch
a
n
g
in
g
o
f
g
a
m
m
a
v
alu
e
s
.
Sa
m
e
as
P
o
l
y
n
o
m
ial
Ker
n
el,
t
h
e
R
B
F
ac
h
ie
v
ed
h
i
g
h
e
s
t
ac
cu
r
ac
y
w
h
en
t
h
e
M
u
s
h
r
o
o
m
d
ataset
is
e
m
p
lo
y
ed
b
y
n
o
les
s
th
a
n
9
9
.
8
7
6
% w
i
th
all
t
h
e
u
s
ed
g
a
m
m
a
v
al
u
e
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2
0
8
8
-
8708
Th
e
effec
t o
f g
a
mma
va
l
u
e
o
n
s
u
p
p
o
r
t v
ec
to
r
ma
ch
in
e
…
(
I
n
t
is
a
r
S
h
a
d
ee
d
A
l
-
Mejib
li
)
5503
Fig
u
r
e
3
.
Flo
w
c
h
ar
t
o
f
t
h
e
i
m
p
le
m
e
n
ted
alg
o
r
ith
m
T
ab
le
3
.
T
h
e
ac
cu
r
ac
y
o
f
p
o
l
y
n
o
m
ial,
r
b
f
a
n
d
s
i
g
m
o
id
D
a
t
a
se
t
N
u
mb
e
r
o
f
f
e
a
t
u
r
e
s
G
a
mm
a
v
a
l
u
e
P
o
l
y
n
o
mi
a
l
R
b
f
S
i
g
mo
i
d
M
u
s
h
r
o
o
m
22
0
.
0
4
5
4
1
0
0
%
1
0
0
%
2
4
.
4
0
5
%
C
h
r
o
n
i
c
K
i
d
n
e
y
24
0
.
0
4
1
6
8
4
.
1
6
6
%
6
1
.
6
6
6
%
6
1
.
6
6
6
%
B
r
e
a
st
c
a
n
c
e
r
9
0
.
1
1
1
1
6
7
.
4
4
1
%
7
4
.
4
1
8
%
6
2
.
7
9
0
%
L
u
n
g
c
a
n
c
e
r
6
0
.
1
6
6
6
6
1
.
1
1
1
%
3
3
.
3
3
3
%
3
3
.
3
3
3
%
H
e
a
r
t
st
a
t
l
o
g
13
0
.
0
7
6
9
6
5
.
4
3
2
%
4
9
.
3
8
2
%
5
0
.
6
1
7
%
Fig
u
r
e
4
.
SVM
ac
cu
r
ac
y
o
f
p
o
l
y
n
o
m
ial
k
er
n
el
ac
co
r
d
in
g
to
ea
ch
d
atase
t
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
10
,
No
.
5
,
Octo
b
e
r
2
0
2
0
:
5
4
9
7
-
5506
5504
Fig
u
r
e
5
.
SVM
ac
cu
r
a
c
y
o
f
R
B
F
k
er
n
el
ac
co
r
d
in
g
to
ea
c
h
d
atase
t
Fig
u
r
e
6
.
SVM
ac
cu
r
ac
y
o
f
s
i
g
m
o
id
k
er
n
el
ac
co
r
d
in
g
to
ea
c
h
d
ataset
T
h
e
ac
cu
r
ac
y
o
f
Sig
m
o
id
i
s
in
f
lu
e
n
ce
d
b
y
t
h
e
c
h
an
g
i
n
g
o
f
t
h
e
g
a
m
m
a
v
a
lu
e
b
ased
o
n
t
h
e
u
s
e
d
d
ataset
as
F
i
g
u
r
e
5
s
h
o
w
s
.
Usi
n
g
t
h
e
n
e
w
p
r
o
p
o
s
ed
g
a
m
m
a
v
alu
e
s
i
m
p
ac
t
th
e
ac
cu
r
ac
y
o
f
Sig
m
o
id
p
o
s
iti
v
el
y
.
T
h
e
ac
cu
r
ac
y
o
f
SVM
w
it
h
S
ig
m
o
id
is
le
s
s
th
a
n
7
0
%
w
h
e
n
e
m
p
lo
y
i
n
g
all
t
h
e
d
atase
ts
,
w
h
ic
h
is
les
s
t
h
a
n
th
e
p
r
ev
io
u
s
t
w
o
k
er
n
el
s
.
T
h
e
P
o
ly
n
o
m
ial
k
er
n
el
ac
h
iev
ed
h
ig
h
e
s
t
ac
c
u
r
ac
y
th
a
n
R
B
F
a
n
d
Sig
m
o
id
w
it
h
all
th
e
e
m
p
lo
y
ed
d
ataset
s
ex
ce
p
t
th
e
b
r
ea
s
t
ca
n
ce
r
w
h
er
e
R
B
F
o
b
tain
ed
th
e
h
i
g
h
e
s
t
ac
cu
r
a
c
y
r
ate.
P
o
ly
n
o
m
ial
an
d
R
B
F k
er
n
els ac
h
ie
v
ed
h
i
g
h
est ac
c
u
r
ac
y
t
h
a
n
Si
g
m
o
id
w
i
th
Mu
s
h
r
o
o
m
d
ata
s
et.
T
h
e
v
al
u
e
o
f
g
a
m
m
a
v
alu
e
af
f
ec
t
th
e
ac
cu
r
ac
y
o
f
th
e
S
V
M
w
it
h
u
s
ed
k
er
n
el
a
n
d
n
o
t
alw
a
y
s
u
s
i
n
g
(
5
)
in
ca
lcu
lati
n
g
t
h
e
v
al
u
e
o
f
g
a
m
m
a
r
esu
lti
n
g
in
b
est ac
c
u
r
ac
y
.
6.
CO
NCLU
SI
O
N
T
h
is
s
t
u
d
y
ev
al
u
ate
s
t
h
e
p
er
f
o
r
m
a
n
ce
o
f
t
h
r
ee
k
er
n
el
s
w
i
th
SVM
,
w
h
ic
h
ar
e
u
s
ed
f
o
r
cla
s
s
i
f
icatio
n
b
ased
o
n
d
if
f
er
en
t
v
alu
e
s
o
f
g
a
m
m
a
p
ar
a
m
eter
(
γ
)
w
it
h
v
ar
io
u
s
d
atasets
d
escr
ip
tio
n
s
.
SVMs
u
s
es
p
o
p
u
lar
k
er
n
el
s
f
o
r
class
i
f
i
ca
tio
n
p
u
r
p
o
s
es
w
h
er
e
t
h
e
s
e
k
er
n
el
f
u
n
ctio
n
s
(
p
o
l
y
n
o
m
ial,
R
B
F,
a
n
d
s
i
g
m
o
id
)
u
s
ed
γ
p
ar
am
eter
.
T
h
e
ass
ess
m
e
n
t
o
f
ap
p
lied
ex
p
er
im
en
tal
r
ev
e
aled
th
at
th
e
g
a
m
m
a
v
alu
e,
i
m
p
le
m
e
n
ted
k
er
n
el
f
u
n
ctio
n
a
n
d
e
m
p
lo
y
ed
d
ata
s
et
d
escr
ip
ti
o
n
s
af
f
ec
t
t
h
e
S
VM
ac
cu
r
ac
y
.
T
h
e
c
h
an
g
i
n
g
in
g
a
m
m
a
v
al
u
e
u
n
e
v
e
n
l
y
i
n
f
l
u
en
ce
d
t
h
e
clas
s
i
f
icatio
n
ac
cu
r
ac
y
o
f
th
r
ee
k
er
n
els o
n
d
atasets
a
s
s
h
o
w
n
b
y
t
h
e
r
esu
lts
.
T
h
e
P
o
ly
n
o
m
ial
k
er
n
el
o
b
tain
ed
h
i
g
h
ac
cu
r
ac
y
w
h
en
t
h
e
g
a
m
m
a
v
al
u
e
i
s
ca
lc
u
lated
as
(
5
)
in
all
t
h
e
e
m
p
lo
y
ed
d
ata
s
ets
ex
ce
p
t
th
e
C
h
r
o
n
ic
Kid
n
e
y
d
ata
s
et,
w
h
ic
h
s
h
o
w
s
n
o
c
h
a
n
g
e
in
o
b
tain
ed
ac
cu
r
ac
y
.
T
h
e
R
B
F
k
er
n
el
f
u
n
c
tio
n
ef
f
icien
c
y
i
s
s
tab
le
u
s
in
g
v
ar
io
u
s
v
a
lu
e
s
o
f
g
a
m
m
a
as
its
a
cc
u
r
ac
y
is
s
lig
h
tl
y
ch
an
g
ed
.
T
h
e
ch
an
g
i
n
g
o
f
th
e
g
a
m
m
a
v
a
lu
e,
tak
in
g
o
n
co
n
s
i
d
er
atio
n
th
e
u
s
ed
d
ataset,
in
f
lu
en
ce
s
th
e
ac
c
u
r
ac
y
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2
0
8
8
-
8708
Th
e
effec
t o
f g
a
mma
va
l
u
e
o
n
s
u
p
p
o
r
t v
ec
to
r
ma
ch
in
e
…
(
I
n
t
is
a
r
S
h
a
d
ee
d
A
l
-
Mejib
li
)
5505
o
f
Si
g
m
o
id
.
T
h
u
s
,
th
e
u
s
ed
d
ataset
a
f
f
ec
ts
th
e
p
er
f
o
r
m
a
n
c
e
o
f
Sig
m
o
id
i
n
ad
d
itio
n
to
g
a
m
m
a
v
al
u
e
w
h
er
e
s
o
m
e
d
ataset
s
s
h
o
w
c
h
an
g
e
i
n
ac
cu
r
ac
y
w
h
en
t
h
e
g
a
m
m
a
v
alu
e
is
c
h
an
g
ed
s
u
ch
a
s
in
M
u
s
h
r
o
o
m
an
d
B
r
e
ast
ca
n
ce
r
w
h
er
ea
s
t
h
e
ac
cu
r
ac
y
o
f
o
th
er
d
ataset
w
er
e
n
o
t c
h
an
g
ed
.
T
h
e
r
esu
lt
s
h
o
w
s
th
a
t
t
h
e
p
er
f
o
r
m
an
ce
o
f
t
h
e
k
er
n
el
is
af
f
ec
ted
b
y
t
h
e
n
u
m
b
er
o
f
attr
ib
u
te
s
,
in
s
ta
n
ce
s
,
a
n
d
m
i
s
s
i
n
g
v
al
u
e
s
.
I
n
ad
d
itio
n
to
,
g
a
m
m
a
v
alu
e
in
f
lu
e
n
ce
s
th
e
k
er
n
el
p
er
f
o
r
m
an
ce
.
T
h
e
ac
cu
r
ac
y
o
f
SVM
r
es
u
lts
i
n
cr
ea
s
ed
b
y
in
cr
ea
s
i
n
g
o
f
attr
ib
u
te
s
n
u
m
b
er
,
in
s
ta
n
ce
s
n
u
m
b
er
f
o
r
d
atas
et
as
i
n
Mu
s
h
r
o
o
m
an
d
C
h
r
o
n
ic
Kid
n
e
y
.
Fro
m
o
th
er
h
an
d
,
in
cr
ea
s
in
g
th
e
n
u
m
b
er
o
f
m
is
s
in
g
v
a
lu
e
s
af
f
ec
ts
th
e
ac
cu
r
ac
y
n
eg
at
iv
el
y
.
Var
io
u
s
ad
ap
tatio
n
s
,
ex
a
m
i
n
atio
n
s
,
a
n
d
ex
p
er
im
en
ts
b
ased
o
th
er
SVM
o
p
ti
m
izatio
n
p
ar
a
m
eter
s
an
d
f
ea
tu
r
e
s
o
f
d
atasets
ca
n
b
e
in
v
es
tig
a
ted
in
f
u
tu
r
e.
Fu
tu
r
e
w
o
r
k
m
a
y
p
r
ese
n
ts
e
x
ten
s
i
v
e
an
al
y
s
is
o
f
SVM
w
it
h
o
th
er
o
p
ti
m
izatio
n
p
ar
a
m
eter
s
an
d
d
atasets
.
RE
F
E
R
E
NC
E
S
[1
]
V
a
p
n
ik
,
V
.
a
n
d
A
.
Lern
e
r,
"
P
a
tt
e
rn
re
c
o
g
n
it
io
n
u
sin
g
g
e
n
e
ra
li
z
e
d
p
o
rtrait
m
e
th
o
d
,"
Au
to
m
a
ti
o
n
a
n
d
Rem
o
t
e
Co
n
tro
l,
v
o
l.
2
4
,
p
p
.
7
7
4
-
7
8
0
,
1
9
6
3
.
[2
]
Zh
a
n
g
,
D.,
"
M
a
c
h
i
n
e
L
e
a
rn
in
g
in
V
a
lu
e
-
Ba
se
d
S
o
f
tw
a
re
T
e
s
t
Da
ta
Ge
n
e
ra
ti
o
n
,
"
1
8
th
IE
EE
In
ter
n
a
t
io
n
a
l
Co
n
fer
e
n
c
e
o
n
T
o
o
ls wi
th
Art
if
ici
a
l
In
telli
g
e
n
c
e
,
p
p
.
7
3
2
-
7
3
6
,
2
0
0
6
.
[3
]
W
e
rn
ick
,
M
.
N.,
Y
o
n
g
y
i.
Y.
Br
a
n
k
o
v
,
J.
G
.
Yo
u
rg
a
n
o
v
,
G
.
a
n
d
S
tro
th
e
r,
S
.
C.
,
"
M
a
c
h
i
n
e
L
e
a
rn
in
g
i
n
M
e
d
ica
l
Im
a
g
in
g
,
"
IEE
E
S
ig
n
a
l
Pr
o
c
e
ss
in
g
M
a
g
a
zin
e
,
v
o
l.
2
7
,
n
o
.
4
,
p
p
.
25
-
3
8
,
2
0
1
0
.
[4
]
Yo
o
,
P
.
D.,
Kim
,
M
.
H.,
a
n
d
Ja
n
,
T
.
,
"
M
a
c
h
in
e
L
e
a
rn
in
g
T
e
c
h
n
iq
u
e
s
a
n
d
Us
e
o
f
Ev
e
n
t
In
f
o
r
m
a
ti
o
n
f
o
r
S
t
o
c
k
M
a
rk
e
t
P
re
d
icti
o
n
:
A
S
u
rv
e
y
a
n
d
Ev
a
lu
a
ti
o
n
,
"
In
ter
n
a
ti
o
n
a
l
C
o
n
fer
e
n
c
e
o
n
In
telli
g
e
n
t
A
g
e
n
ts,
W
e
b
T
e
c
h
n
o
lo
g
ies
a
n
d
In
ter
n
e
t
Co
mm
e
rc
e
,
v
o
l.
2
,
p
p
.
8
3
5
-
8
4
1
,
2
0
0
5
.
[5
]
S
M
a
ji
,
A
.
C
Be
r
g
.
,
a
n
d
J
M
a
li
k
,
“
Eff
icie
n
t
c
la
ss
i
f
ic
a
ti
o
n
f
o
r
a
d
d
it
iv
e
k
e
rn
e
l
S
V
M
s,”
IEE
E
tra
n
sa
c
ti
o
n
s
o
n
p
a
tt
e
r
n
a
n
a
lys
is a
n
d
ma
c
h
in
e
i
n
te
ll
i
g
e
n
c
e
,
”
v
o
l.
3
5
,
n
o
.
1
,
p
p
.
6
6
-
7
7
,
2
0
1
3
.
[6
]
F
.
G
u
ll
o
,
“
F
ro
m
P
a
tt
e
rn
s
in
Da
ta
to
Kn
o
w
led
g
e
Disc
o
v
e
r
y
:
W
h
a
t
Da
ta
M
in
in
g
Ca
n
Do
,
”
Ph
y
sic
s
Pro
c
e
d
ia
,
v
o
l.
6
2
,
pp
.
1
8
-
2
2
,
2
0
1
5
.
[7
]
Rim
a
h
A
.
,
Do
rra
B.
A
.
,
a
n
d
No
u
re
d
d
i
n
e
E.
,
“
P
ra
c
ti
c
a
l
S
e
lec
ti
o
n
o
f
S
V
M
S
u
p
e
rv
ise
d
P
a
ra
m
e
ters
w
it
h
Di
ff
e
re
n
t
F
e
a
tu
re
Re
p
re
se
n
tatio
n
s
f
o
r
V
o
w
e
l
Re
c
o
g
n
it
io
n
,”
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
Di
g
it
a
l
C
o
n
te
n
t
T
e
c
h
n
o
lo
g
y
a
n
d
it
s
Ap
p
li
c
a
ti
o
n
s
,
v
o
l
.
7
,
n
o
.
9
,
p
p
.
4
1
8
-
4
2
4
,
2
0
1
3
.
[8
]
M
ich
e
ll
e
H
.
,
S
h
r
id
h
a
ra
n
C
.
,
S
u
m
e
d
h
a
S
.,
a
n
d
Ya
n
g
S
,
"
Ke
rn
e
l
S
e
lec
ti
o
n
a
n
d
Dim
e
n
sio
n
a
li
ty
Re
d
u
c
ti
o
n
in
S
VM
Clas
sif
ic
a
ti
o
n
o
f
A
u
ti
sm
S
p
e
c
tru
m
Diso
rd
e
rs
,"
2013.
[9
]
De
e
p
a
k
K.
A
.
,
a
n
d
Ra
h
u
l
K.
,
"
S
p
a
m
F
il
terin
g
u
sin
g
S
V
M
w
it
h
d
i
ff
e
r
e
n
t
Ke
rn
e
l
F
u
n
c
ti
o
n
s
,"
In
ter
n
a
ti
o
n
a
l
J
o
u
r
n
a
l
o
f
Co
m
p
u
ter
A
p
p
l
ica
ti
o
n
s,
v
o
l.
1
3
6
,
n
o
.
5
,
2
0
1
6
.
[1
0
]
S
a
y
e
d
F
.
,
S
a
i
d
G
.
,
M
o
h
a
m
e
d
A
.
,
Hu
ss
e
in
A
.
S
.
,
Am
ira
A
.
,
a
n
d
A
sif
A
.
,
"
In
v
e
stig
a
ti
n
g
th
e
Ef
fe
c
t
o
f
Diff
e
r
e
n
t
Ke
rn
e
l
F
u
n
c
ti
o
n
s
o
n
t
h
e
P
e
rf
o
rm
a
n
c
e
o
f
S
V
M
f
o
r
Re
c
o
g
n
izin
g
A
ra
b
ic
Ch
a
ra
c
ters
,
"
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
Ad
v
a
n
c
e
d
Co
mp
u
ter
S
c
ien
c
e
a
n
d
A
p
p
li
c
a
ti
o
n
s,
v
o
l.
7
,
n
o
.
1
,
p
p
.
4
4
6
-
4
5
0
,
2
0
1
6
.
[1
1
]
S
u
p
riy
a
P
.
,
a
n
d
De
e
p
a
k
S
.
,
"
Co
m
p
a
riso
n
o
f
V
a
rio
u
s
Ke
rn
e
ls
Of
S
u
p
p
o
rt
V
e
c
to
r
M
a
c
h
in
e
,
"
In
ter
n
a
ti
o
n
a
l
J
o
u
r
n
a
l
fo
r
Res
e
a
rc
h
i
n
Ap
p
li
e
d
S
c
ien
c
e
a
n
d
En
g
in
e
e
rin
g
T
e
c
h
n
o
l
o
g
y
,
v
o
l.
3
,
n
o
.
V
II
,
p
p
.
5
3
2
-
5
3
6
,
2
0
1
5
.
[1
2
]
S
u
tao
S
.
,
Zh
ic
h
a
o
Z
.
,
Z
h
iy
in
g
L
.
,
Jia
c
a
i
Z.
,
a
n
d
L
i
Y.
,
"
Co
m
p
a
ra
ti
v
e
S
tu
d
y
o
f
S
VM
M
e
t
h
o
d
s
Co
m
b
i
n
e
d
w
it
h
V
o
x
e
l
S
e
lec
ti
o
n
f
o
r
Ob
jec
t
Ca
teg
o
ry
Cla
ss
if
ic
a
ti
o
n
o
n
f
M
RI
Da
ta,"
PL
o
S
ONE,
v
o
l.
6
,
n
o
.
2
,
p
p
.
1
-
1
1
,
2
0
1
1
.
[1
3
]
IIT
Ka
n
p
u
r,
"
S
u
p
p
o
rt
V
e
c
to
r
M
a
c
h
in
e
s
,"
Pro
g
ra
mm
in
g
C
lu
b
,
M
a
r
.
1
1
,
2
0
1
6
.
[
o
n
li
n
e
]
,
A
v
a
i
la
b
le:
h
tt
p
:
//
h
o
m
e
.
ii
tk
.
a
c
.
in
/~
a
ru
n
o
th
i/
p
c
lu
b
/m
l/
3
.
p
d
f
.
[1
4
]
D.
B.
A
.
M
e
z
g
h
a
n
i,
S
.
Z
.
Bo
u
je
lb
e
n
e
,
a
n
d
N.
El
lo
u
z
e
,
“
Ev
a
lu
a
ti
o
n
o
f
S
V
M
Ke
rn
e
ls
a
n
d
C
o
n
v
e
n
ti
o
n
a
l
M
a
c
h
in
e
L
e
a
rn
in
g
A
lg
o
rit
h
m
s
f
o
r
S
p
e
a
k
e
r
Id
e
n
ti
f
ica
ti
o
n
,
”
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
Hy
b
rid
I
n
f
o
rm
a
ti
o
n
T
e
c
h
n
o
lo
g
y
,
v
o
l.
3
,
n
o.
3
,
p
p
.
2
3
-
3
4
,
2
0
1
0
.
[1
5
]
I.
A
.
A
d
e
y
a
n
ju
,
E.
O.
Om
id
io
r
a
,
a
n
d
O.
F
.
Oy
e
d
o
k
u
n
,
“
P
e
rf
o
rm
a
n
c
e
Ev
a
lu
a
ti
o
n
o
f
Dif
f
e
r
e
n
t
S
u
p
p
o
rt
V
e
c
to
r
M
a
c
h
in
e
Ke
rn
e
ls f
o
r
F
a
c
e
Em
o
ti
o
n
Re
c
o
g
n
it
i
o
n
,
”
S
AI
I
n
telli
g
e
n
t
S
y
ste
ms
Co
n
fer
e
n
c
e
,
p
p
.
8
0
4
-
8
0
6
,
2
0
1
5
.
[1
6
]
Ce
lso
A
.
A
.
K
a
e
stn
e
r,
“
S
u
p
p
o
r
t
V
e
c
to
r
M
a
c
h
i
n
e
s
a
n
d
Ke
rn
e
l
F
u
n
c
ti
o
n
s
f
o
r
T
e
x
t
P
ro
c
e
ss
in
g
,
”
J
o
u
rn
a
l
o
f
RIT
A,
v
o
l.
2
0
,
p
p
.
1
3
0
-
1
5
4
,
2
0
1
3
.
[1
7
]
N.
M
.
Ha
tt
a
a
n
d
Z
.
A
.
.
S
h
a
h
,
“
Ev
a
lu
a
te
th
e
P
e
rf
o
rm
a
n
c
e
o
f
S
VM
Ke
rn
e
l
F
u
n
c
ti
o
n
s
f
o
r
M
u
lt
icla
ss
Ca
n
c
e
r
C
l
a
s
s
i
f
i
c
a
t
i
o
n
,”
U
T
M
C
o
m
p
u
t
i
n
g
P
r
o
c
e
e
d
i
n
g
s
I
n
n
o
v
a
t
i
o
n
s
i
n
C
o
m
p
u
t
i
n
g
T
e
c
h
n
o
l
o
g
y
a
n
d
A
p
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
1
,
p
p
.
1
-
7
,
2016.
[1
8
]
J.
W
a
in
e
r
a
n
d
G
.
Ca
w
le
y
,
“
E
m
p
iri
c
a
l
e
v
a
lu
a
ti
o
n
o
f
re
sa
m
p
li
n
g
p
r
o
c
e
d
u
re
s
f
o
r
o
p
t
im
isin
g
S
V
M
h
y
p
e
rp
a
ra
m
e
t
e
rs
,
”
J
o
u
rn
a
l
o
f
M
a
c
h
i
n
e
L
e
a
rn
in
g
Res
e
a
rc
h
,
v
o
l.
1
8
,
p
p
.
1
-
3
5
,
2
0
1
7
.
[1
9
]
R.
S
a
n
g
e
e
th
a
a
n
d
B.
Ka
lp
a
n
a
,
“
P
e
rf
o
rm
a
n
c
e
E
v
a
lu
a
ti
o
n
o
f
Ke
rn
e
ls
in
M
u
lt
icla
ss
S
u
p
p
o
rt
V
e
c
to
r
M
a
c
h
i
n
e
s
,”
In
ter
n
a
t
io
n
a
l
J
o
u
rn
a
l
o
f
S
o
f
t
Co
m
p
u
ti
n
g
a
n
d
E
n
g
in
e
e
rin
g
,
v
o
l.
1
,
n
o
.
5
,
p
p
.
1
3
8
-
1
4
5
,
2
0
1
1
.
[2
0
]
G
ra
h
a
m
M
c
Do
n
a
ld
,
Nic
o
l´
a
s
Ga
rc
´
ıa
-
P
e
d
ra
jas
,
Cra
ig
M
a
c
d
o
n
a
l
d
a
n
d
Ia
d
h
Ou
n
is,
“
A
S
tu
d
y
o
f
S
V
M
Ke
rn
e
l
F
u
n
c
ti
o
n
s
f
o
r
S
e
n
siti
v
it
y
Clas
si
fica
ti
o
n
E
n
se
m
b
les
w
it
h
P
OS
S
e
q
u
e
n
c
e
s
,”
T
h
e
4
0
th
In
ter
n
a
ti
o
n
a
l
ACM
S
IGI
R
Co
n
fer
e
n
c
e
o
n
Res
e
a
rc
h
a
n
d
De
v
e
lo
p
me
n
t
i
n
I
n
fo
rm
a
ti
o
n
Retrie
v
a
l
,
p
p
.
1
0
9
7
-
1
1
0
0
,
2
0
1
6
.
[2
1
]
I.
S
.
Al
-
M
e
ji
b
li
,
D.
H.
A
b
d
,
J.
K
A
l
wa
n
,
a
n
d
A
.
J.
Ra
b
a
sh
,
“
P
e
rfo
rm
a
n
c
e
e
v
a
lu
a
ti
o
n
o
f
k
e
rn
e
ls
in
su
p
p
o
rt
v
e
c
to
r
m
a
c
h
in
e
,
”
1
st
A
n
n
u
a
l
I
n
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
In
f
o
rm
a
ti
o
n
a
n
d
S
c
ien
c
e
s
,
2
0
1
8
.
[2
2
]
Ho
T
u
Ba
o
,
“
Ke
rn
e
l
M
e
t
h
o
d
s a
n
d
S
u
p
p
o
rt
V
e
c
to
r
M
a
c
h
in
e
s
,
”
J
a
p
a
n
Ad
v
a
n
c
e
In
sti
tu
te
o
f
S
c
ien
c
e
a
n
d
T
e
c
h
n
o
l
o
g
y
,
[2
3
]
A
.
N
.
Kh
o
b
ra
g
a
d
e
,
M
.
M
.
Ra
g
h
u
w
a
n
sh
i,
a
n
d
L
.
M
a
li
k
,
“
Ev
a
lu
a
ti
n
g
Ke
rn
e
l
Ef
fe
c
t
o
n
P
e
rf
o
r
m
a
n
c
e
o
f
S
V
M
C
l
a
s
s
i
f
i
c
a
t
i
o
n
u
s
i
n
g
S
a
t
e
l
l
i
t
e
I
m
a
g
e
s
,”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
S
c
i
e
n
t
i
f
i
c
a
n
d
E
n
g
i
n
e
e
r
i
n
g
R
e
s
e
a
r
c
h
,
v
o
l
.
7
,
n
o
.
3
,
p
p
.
7
4
2
-
7
4
8
,
2016.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
10
,
No
.
5
,
Octo
b
e
r
2
0
2
0
:
5
4
9
7
-
5506
5506
[2
4
]
B.
S
c
h
o
lk
o
p
f
a
n
d
A
.
J.
S
m
o
la,
“
L
e
a
rn
in
g
w
it
h
Ke
rn
e
ls
S
u
p
p
o
rt
V
e
c
to
r
M
a
c
h
i
n
e
s,
Re
g
u
lariz
a
ti
o
n
,
”
Op
ti
miza
ti
o
n
,
a
n
d
Be
y
o
n
d
,
”
M
IT
Pre
ss
,
2
0
0
2
.
[2
5
]
A
n
d
re
w
N
g
,
"
S
u
p
p
o
rt
V
e
c
to
r
M
a
c
h
in
e
s
,
"
P
a
rt
V
,
CS
2
2
9
L
e
c
tu
re
n
o
tes
.
,
[
o
n
li
n
e
]
,
A
v
a
i
la
b
le:
h
tt
p
:
//
c
s2
2
9
.
sta
n
f
o
rd
.
e
d
u
/
n
o
tes
/cs
2
2
9
-
n
o
tes
3
.
p
d
f
.
[2
6
]
D.
H.
A
b
d
,
A
.
T
.
S
a
d
iq
,
a
n
d
A
.
R.
A
b
b
a
s,
"
Clas
si
fy
in
g
P
o
li
ti
c
a
l
A
ra
b
ic
A
rti
c
les
U
sin
g
S
u
p
p
o
rt
Ve
c
to
r
M
a
c
h
in
e
w
it
h
Diff
e
r
e
n
t
F
e
a
tu
re
Ex
trac
ti
o
n
,
"
i
n
In
ter
n
a
ti
o
n
a
l
C
o
n
fer
e
n
c
e
o
n
A
p
p
li
e
d
C
o
mp
u
ti
n
g
t
o
S
u
p
p
o
rt
In
d
u
stry
:
In
n
o
v
a
ti
o
n
a
n
d
T
e
c
h
n
o
l
o
g
y
,
S
p
r
i
n
g
e
r
,
p
p
.
7
9
-
9
4
,
2
0
1
9
.
[2
7
]
Z
h
a
n
g
,
Y.,
a
n
d
W
u
,
L
.
,
“
Clas
si
f
ica
ti
o
n
o
f
f
ru
it
s
u
sin
g
c
o
m
p
u
ter
v
isio
n
a
n
d
a
m
u
lt
icla
ss
su
p
p
o
rt
v
e
c
to
r
m
a
c
h
in
e
,
”
S
e
n
so
rs
,
v
o
l
.
1
2
,
p
p
.
1
2
4
8
9
-
1
2
5
0
5
,
2
0
1
2
.
[2
8
]
D.
A
b
d
,
J.
K.
A
l
w
a
n
,
M
.
Ib
ra
h
i
m
a
n
d
M
.
B.
Na
e
e
m
,
"
T
h
e
u
ti
li
s
a
ti
o
n
o
f
m
a
c
h
in
e
lea
rn
in
g
a
p
p
ro
a
c
h
e
s
f
o
r
m
e
d
ica
l
d
a
ta
c
las
sif
ic
a
ti
o
n
a
n
d
p
e
rso
n
a
l
c
a
re
s
y
ste
m
m
a
n
g
e
m
e
n
tf
o
r
sic
k
le
c
e
ll
d
ise
a
se
,
"
An
n
u
a
l
C
o
n
fer
e
n
c
e
o
n
Ne
w
T
re
n
d
s
in
In
fo
rm
a
t
io
n
a
n
d
Co
mm
u
n
ica
ti
o
n
s T
e
c
h
n
o
l
o
g
y
A
p
p
li
c
a
ti
o
n
s,
p
p
.
2
1
3
-
2
1
8
,
2
0
1
7
.
[2
9
]
M
.
Kh
a
laf
e
t
a
l.
,
"
Re
c
u
rre
n
t
Ne
u
ra
l
Ne
t
w
o
rk
A
rc
h
it
e
c
tu
re
s f
o
r
A
n
a
ly
sin
g
Bio
m
e
d
ica
l
D
a
ta S
e
ts,"
1
0
th
In
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
De
v
e
lo
p
me
n
ts i
n
e
S
y
ste
ms
En
g
i
n
e
e
rin
g
,
p
p
.
2
3
2
-
2
3
7
,
2
0
1
7
.
[3
0
]
G
u
a
lt
ieri.
J.
A
,
Ch
e
tt
ri.
S
.
R,
Cr
o
m
p
.
R.
F
a
n
d
Jo
h
n
so
n
.
L
.
F
,
“
S
u
p
p
o
rt
v
e
c
to
r
m
a
c
h
in
e
c
las
si
f
ier
s
a
s
a
p
p
li
e
d
t
o
A
V
IRIS
d
a
ta,
”
in
S
u
mm
a
rie
s 8
th
J
PL
Ai
rb
o
rn
e
E
a
rth
S
c
ien
c
e
W
o
rk
sh
o
p
,
J
PL
,
p
p
.
2
1
7
–
2
2
7
,
1
9
9
9
.
[3
1
]
M
.
A
.
Na
n
d
a
,
K.
B.
S
e
m
in
a
r,
D.
Na
n
d
ik
a
,
a
n
d
A
.
M
a
d
d
u
,
“
A
Co
m
p
a
riso
n
S
t
u
d
y
o
f
Ke
r
n
e
l
F
u
n
c
ti
o
n
s in
th
e
S
u
p
p
o
rt
V
e
c
to
r
M
a
c
h
in
e
a
n
d
Its
A
p
p
li
c
a
ti
o
n
f
o
r
T
e
rm
it
e
D
e
tec
ti
o
n
,
”
In
f
o
rm
a
ti
o
n
,
v
o
l
.
9
,
n
o
.
5
,
p
p
.
1
-
1
4
,
2
0
1
8
.
[3
2
]
M
.
Kh
a
laf
e
t
a
l
.
,
"
A
n
A
p
p
li
c
a
ti
o
n
o
f
Us
in
g
S
u
p
p
o
rt
V
e
c
to
r
M
a
c
h
in
e
Ba
se
d
o
n
Clas
sif
ica
ti
o
n
T
e
c
h
n
iq
u
e
f
o
r
P
re
d
ictin
g
M
e
d
ica
l
Da
ta S
e
ts,"
in
In
ter
n
a
t
io
n
a
l
C
o
n
fer
e
n
c
e
o
n
I
n
tel
li
g
e
n
t
C
o
mp
u
ti
n
g
,
S
p
r
in
g
e
r,
p
p
.
5
8
0
-
5
9
1
,
2
0
1
9
.
[
3
3
]
S
tev
e
R.
G
u
n
n
,
"
S
u
p
p
o
rt
V
e
c
t
o
r
M
a
c
h
i
n
e
s
f
o
r
Clas
sif
ica
ti
o
n
a
n
d
Re
g
re
ss
io
n
,
"
Un
ive
rs
it
y
o
f
S
o
u
t
h
a
m
p
to
n
,
T
e
c
h
n
i
c
a
l
R
e
p
o
r
t
F
a
c
u
l
t
y
o
f
E
n
g
i
n
e
e
r
i
n
g
,
S
c
i
e
n
c
e
a
n
d
M
a
t
h
e
m
a
t
i
c
s
S
c
h
o
o
l
o
f
E
l
e
c
t
r
o
n
i
c
s
a
n
d
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
1998.
[3
4
]
“
UC Irv
in
e
M
a
c
h
in
e
L
e
a
rn
in
g
Re
p
o
sit
o
ry
,
”
UCI
,
[
On
li
n
e
]
,
A
v
a
il
a
b
le:
h
tt
p
:/
/arc
h
iv
e
.
ics
.
u
c
i.
e
d
u
/m
l/
in
d
e
x
.
p
h
p
.
B
I
O
G
RAP
H
I
E
S
O
F
AUTH
O
RS
In
tisa
r
Al
-
M
e
jib
li
,
h
a
s
P
h
D
in
Co
m
p
u
ti
n
g
a
n
d
El
e
c
tro
n
ic
S
y
ste
m
s/
S
c
h
o
o
l
o
f
Co
m
p
u
ter
S
c
ien
c
e
a
n
d
El
e
c
tro
n
ic
En
g
in
e
e
rin
g
/
Un
iv
e
rsit
y
o
f
Es
se
x
/U
K.
Cu
rre
n
tl
y
sh
e
is
w
o
rk
in
g
a
s
As
sista
n
t
P
r
o
f
e
ss
o
r
in
S
m
a
rt
Cit
ies
S
e
c
ti
o
n
/
Un
iv
e
rsity
o
f
In
f
o
r
m
a
ti
o
n
T
e
c
h
n
o
l
o
g
y
a
n
d
Co
m
m
u
n
ica
ti
o
n
s/
B
a
g
h
d
a
d
/
Ira
q
.
T
h
e
i
n
tere
sts
o
f
In
ti
sa
r
i
n
c
lu
d
e
s
W
irele
ss
S
e
n
so
r
Ne
tw
o
rk
s,
S
e
c
u
rit
y
,
(Io
T
)
a
n
d
h
e
a
lt
h
c
a
re
s
y
ste
m
s.
F
u
rth
e
r,
In
t
isa
r
in
tere
sts
in
c
lu
d
e
u
si
n
g
A
rti
f
icia
l
In
telli
g
e
n
t
tec
h
n
iq
u
e
s
t
o
d
e
v
e
lo
p
so
l
u
ti
o
n
s f
o
r
d
if
fe
re
n
t.
J
w
a
n
K
.
Al
w
a
n
h
a
s
M
.
S
c
.
o
f
in
f
o
rm
a
ti
o
n
sy
ste
m
2
0
1
4
f
ro
m
O
s
m
a
n
ia
Un
iv
e
rsit
y
in
In
d
ia
a
n
d
c
u
rre
n
tl
y
sh
e
is
w
o
rk
in
g
a
s
lec
tu
re
r
in
Un
iv
e
rsit
y
o
f
In
f
o
rm
a
ti
o
n
T
e
c
h
n
o
l
o
g
y
a
n
d
Co
m
m
u
n
ica
ti
o
n
s/Ba
g
h
d
a
d
/
Ira
q
.
S
h
e
is
i
n
tere
ste
d
i
n
p
a
tt
e
r
n
c
las
sif
ic
a
ti
o
n
,
m
e
d
ica
l
c
o
m
p
u
ti
n
g
,
m
e
d
ica
l
in
f
o
rm
a
ti
o
n
sy
ste
m
s,
p
a
ti
e
n
t
m
o
n
it
o
ri
n
g
,
su
p
p
o
rt
v
e
c
to
r
m
a
c
h
in
e
s,
lea
rn
i
n
g
(a
rti
f
icia
l
in
telli
g
e
n
c
e
),
o
p
ti
m
iza
ti
o
n
.
Dh
a
f
a
r
H
a
m
e
d
A
b
d
is
w
o
rk
in
g
a
s
lec
tu
re
r
in
A
l
-
M
a
a
re
f
Un
iv
e
rsi
ty
Co
ll
e
g
e
,
Ira
q
.
He
is
in
tere
ste
d
in
d
ise
a
se
s,
p
a
tt
e
rn
c
las
si
f
ic
a
ti
o
n
,
h
e
a
lt
h
c
a
re
,
m
e
d
ica
l
c
o
m
p
u
ti
n
g
,
m
e
d
ica
l
in
f
o
r
m
a
ti
o
n
sy
ste
m
s,
p
a
ti
e
n
t
m
o
n
it
o
rin
g
,
s
u
p
p
o
rt
v
e
c
to
r
m
a
c
h
in
e
s,
Ba
y
e
s
m
e
th
o
d
s,
d
a
ta
a
n
a
ly
sis,
d
e
c
isio
n
tree
s,
e
x
p
e
rt
s
y
ste
m
s,
lea
rn
in
g
(a
rti
f
icia
l
in
telli
g
e
n
c
e
),
o
p
ti
m
iza
ti
o
n
,
p
a
ti
e
n
t
trea
tme
n
t,
p
o
ly
n
o
m
ials,
ra
d
ial
b
a
si
s
f
u
n
c
ti
o
n
n
e
tw
o
rk
s,
re
c
u
rre
n
t
n
e
u
r
a
l
n
e
ts,
sm
a
rt
p
h
o
n
e
s,
a
n
d
s
o
f
twa
r
e
p
e
rf
o
r
m
a
n
c
e
e
v
a
lu
a
ti
o
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.