T
E
L
K
O
M
N
I
K
A
T
elec
o
m
m
un
ica
t
io
n,
Co
m
pu
t
ing
,
E
lect
ro
nics
a
nd
Co
ntr
o
l
Vo
l.
18
,
No
.
5
,
Octo
b
er
2
0
2
0
,
p
p
.
2
4
2
1
~
2
4
3
1
I
SS
N:
1
6
9
3
-
6
9
3
0
,
ac
cr
ed
ited
First Gr
ad
e
b
y
Kem
en
r
is
tek
d
i
k
ti,
Dec
r
ee
No
:
2
1
/E/KPT
/2
0
1
8
DOI
: 1
0
.
1
2
9
2
8
/TE
L
KOM
NI
K
A.
v
1
8
i5
.
1
5
1
4
1
2421
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//jo
u
r
n
a
l.u
a
d
.
a
c.
id
/in
d
ex
.
p
h
p
/TELK
OM
N
I
K
A
Integ
ra
ted
bio
-
se
a
rch appro
a
ches
with
mu
lti
-
o
b
ject
i
v
e
a
lg
o
rithms for op
ti
mi
za
tion a
nd
cla
ss
ificatio
n proble
m
M
o
ha
m
m
a
d Aiz
a
t
B
a
s
ir
1
,
M
o
ha
m
ed
Sa
if
ull
a
h H
us
s
in
2
,
Y
uh
a
nis
Y
us
o
f
3
1,
2
F
a
c
u
lt
y
o
f
Oc
e
a
n
En
g
in
e
e
rin
g
Tec
h
n
o
l
o
g
y
a
n
d
In
f
o
rm
a
ti
c
s,
Un
i
v
e
rsiti
M
a
lay
sia
Tere
n
g
g
a
n
u
,
M
a
l
a
y
sia
3
S
c
h
o
o
l
o
f
Co
m
p
u
ti
n
g
,
Un
i
v
e
rsiti
Uta
ra
M
a
lay
sia
,
M
a
lay
sia
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
an
2
,
2
0
2
0
R
ev
is
ed
Ma
y
1
,
2
0
2
0
Acc
ep
ted
May
1
4
,
2
0
2
0
Op
ti
m
a
l
se
lec
ti
o
n
o
f
fe
a
tu
re
s
is
v
e
ry
d
iffi
c
u
lt
a
n
d
c
r
u
c
ial
t
o
a
c
h
iev
e
,
p
a
rti
c
u
larly
f
o
r
t
h
e
tas
k
o
f
c
las
sifi
c
a
ti
o
n
.
It
is d
u
e
t
o
th
e
tra
d
it
io
n
a
l
m
e
th
o
d
o
f
se
lec
ti
n
g
fe
a
tu
re
s
th
a
t
fu
n
c
ti
o
n
in
d
e
p
e
n
d
e
n
tl
y
a
n
d
g
e
n
e
ra
ted
t
h
e
c
o
ll
e
c
ti
o
n
o
f
irrele
v
a
n
t
fe
a
tu
re
s,
wh
ich
th
e
re
f
o
re
a
ffe
c
ts
th
e
q
u
a
li
ty
o
f
t
h
e
a
c
c
u
ra
c
y
o
f
th
e
c
las
sifica
ti
o
n
.
Th
e
g
o
a
l
o
f
th
is
p
a
p
e
r
is
t
o
le
v
e
ra
g
e
th
e
p
o
ten
ti
a
l
o
f
b
io
-
i
n
sp
ire
d
se
a
rc
h
a
lg
o
r
it
h
m
s,
to
g
e
t
h
e
r
with
wra
p
p
e
r,
in
o
p
ti
m
izi
n
g
m
u
lt
i
-
o
b
jec
ti
v
e
a
l
g
o
rit
h
m
s,
n
a
m
e
ly
ENORA
a
n
d
NSG
A
-
II
t
o
g
e
n
e
ra
te
a
n
o
p
ti
m
a
l
se
t
o
f
fe
a
tu
re
s.
T
h
e
m
a
in
ste
p
s
a
re
to
id
e
a
li
z
e
th
e
c
o
m
b
in
a
ti
o
n
o
f
ENORA
a
n
d
NSG
A
-
II
with
s
u
it
a
b
le
b
i
o
-
se
a
rc
h
a
lg
o
r
it
h
m
s
wh
e
r
e
m
u
lt
ip
le
su
b
se
t
g
e
n
e
ra
ti
o
n
h
a
s
b
e
e
n
im
p
lem
e
n
ted
.
Th
e
n
e
x
t
ste
p
is
t
o
v
a
li
d
a
te
th
e
o
p
ti
m
u
m
fe
a
tu
re
se
t
b
y
c
o
n
d
u
c
ti
n
g
a
su
b
se
t
e
v
a
lu
a
t
io
n
.
Ei
g
h
t
(8
)
c
o
m
p
a
riso
n
d
a
tas
e
ts
o
f
v
a
ri
o
u
s
s
ize
s
h
a
v
e
b
e
e
n
d
e
li
b
e
ra
tely
se
le
c
ted
to
b
e
c
h
e
c
k
e
d
.
Re
su
lt
s
s
h
o
w
n
t
h
a
t
t
h
e
id
e
a
l
c
o
m
b
in
a
ti
o
n
o
f
m
u
lt
i
-
o
b
jec
ti
v
e
a
lg
o
rit
h
m
s,
n
a
m
e
ly
ENORA
a
n
d
NSG
A
-
II,
with
t
h
e
se
lec
ted
b
io
-
in
s
p
ired
se
a
rc
h
a
lg
o
rit
h
m
is
p
ro
m
isi
n
g
to
a
c
h
iev
e
a
b
e
tt
e
r
o
p
ti
m
a
l
so
l
u
ti
o
n
(i.
e
.
a
b
e
st
fe
a
tu
re
s
with
h
ig
h
e
r
c
las
sifica
ti
o
n
a
c
c
u
ra
c
y
)
f
o
r
t
h
e
se
lec
ted
d
a
tas
e
ts.
Th
is d
isc
o
v
e
ry
imp
li
e
s
t
h
a
t
th
e
a
b
il
it
y
o
f
b
io
-
i
n
sp
ire
d
wra
p
p
e
r/fi
lt
e
r
e
d
sy
ste
m
a
lg
o
rit
h
m
s
will
b
o
o
st
th
e
e
fficie
n
c
y
o
f
ENORA
a
n
d
NSG
A
-
II
fo
r
t
h
e
tas
k
o
f
se
lec
ti
n
g
a
n
d
c
las
sify
in
g
fe
a
tu
re
s.
K
ey
w
o
r
d
s
:
B
io
-
in
s
p
ir
ed
C
las
s
if
icatio
n
E
NORA
Featu
r
e
s
elec
tio
n
NSGA
-
II
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
:
Mo
h
am
m
ad
Aiza
t Bas
ir
,
Facu
lty
o
f
Oce
an
E
n
g
in
ee
r
i
n
g
T
ec
h
n
o
lo
g
y
an
d
I
n
f
o
r
m
atics,
Un
iv
er
s
iti Ma
lay
s
ia
T
er
en
g
g
a
n
u
(
UM
T
)
,
2
1
0
3
0
Ku
ala
Ner
u
s
,
T
e
r
en
g
g
a
n
u
,
Ma
lay
s
ia
.
E
m
ail:
aiza
t@
u
m
t.e
d
u
.
m
y
1.
I
NT
RO
D
UCT
I
O
N
E
n
o
r
m
o
u
s
d
ataset
n
o
r
m
ally
co
n
s
is
ts
o
f
a
lar
g
e
n
u
m
b
er
o
f
attr
ib
u
tes.
T
h
ese
attr
ib
u
tes
ar
e
r
ep
etitiv
e/ir
r
elev
an
t
o
n
a
r
eg
u
lar
b
asis
an
d
in
f
lu
en
ce
th
e
d
ata
m
in
in
g
m
o
d
el.
I
n
ca
s
es
wh
er
e
th
e
r
u
le
h
as
s
o
m
an
y
co
n
s
tr
ain
ts
,
with
a
wid
e
n
u
m
b
e
r
o
f
ch
ar
ac
ter
is
tics
,
t
h
e
r
u
le
b
ec
o
m
es
m
o
r
e
co
m
p
licat
ed
an
d
d
if
f
icu
lt
to
u
n
d
er
s
tan
d
.
B
y
u
n
d
e
r
s
tan
d
in
g
th
is
p
r
o
b
lem
,
it
is
im
p
o
r
ta
n
t
to
th
e
th
e
n
u
m
b
e
r
o
f
f
ea
t
u
r
es
to
b
e
u
s
ed
i
n
th
e
cr
ea
tio
n
o
f
in
f
o
r
m
atio
n
m
i
n
in
g
m
o
d
els.
I
n
r
ea
lis
tic
s
itu
ati
o
n
s
,
it
is
p
r
o
p
o
s
ed
th
at
th
e
o
b
s
o
lete
a
n
d
r
ed
u
n
d
an
t
m
ea
s
u
r
em
en
ts
s
h
o
u
ld
b
e
r
e
m
o
v
ed
in
o
r
d
er
to
m
in
i
m
ize
p
r
o
ce
s
s
in
g
tim
e
an
d
lab
o
r
co
s
ts
.
I
n
[
1
]
claim
ed
th
at
a
d
atas
et
with
a
lar
g
e
n
u
m
b
er
o
f
attr
ib
u
tes
is
k
n
o
wn
as
a
d
ataset
with
a
h
ig
h
d
im
en
s
io
n
a
lity
.
T
h
is
co
n
d
itio
n
wo
u
ld
lead
to
th
e
cu
r
s
e
o
f
th
e
d
im
en
s
io
n
ality
th
eo
r
em
,
wh
er
e
th
e
tim
e
o
f
m
ea
s
u
r
em
en
t
is
th
e
ex
p
o
n
en
tial
f
u
n
ctio
n
o
f
th
e
n
u
m
b
er
o
f
d
im
en
s
io
n
s
.
I
n
a
d
d
itio
n
,
th
e
h
ig
h
d
im
e
n
s
io
n
o
f
s
p
ac
e
s
ea
r
ch
in
g
lead
s
to
th
e
r
ed
u
n
d
a
n
cy
o
f
f
ea
tu
r
es
i
n
th
e
m
o
d
el.
T
h
e
u
ltima
te
s
o
lu
tio
n
is
to
r
ed
u
ce
th
e
s
ea
r
ch
d
im
e
n
s
io
n
w
h
ile
p
r
ev
en
tin
g
th
e
lo
s
s
o
f
v
ital
in
f
o
r
m
atio
n
i
n
th
e
r
esu
lts
.
L
ar
g
e
n
u
m
b
er
o
f
attr
ib
u
tes
in
ea
ch
p
o
ten
tial
r
u
le
ca
n
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6
9
3
0
T
E
L
KOM
NI
KA
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l Co
n
tr
o
l
,
Vo
l.
18
,
No
.
5
,
Octo
b
e
r
2
0
2
0
:
2
4
2
1
-
2431
2422
cr
ea
te
am
b
ig
u
o
u
s
r
ep
r
esen
tat
io
n
,
m
ak
in
g
it
d
if
f
icu
lt
to
u
n
d
er
s
tan
d
,
u
s
e,
an
d
ex
er
cise.
T
h
e
co
m
p
lex
ity
o
f
th
e
attr
ib
u
te
ca
n
th
en
b
e
m
i
n
im
ized
b
y
r
ed
u
ci
n
g
th
e
n
u
m
b
er
o
f
attr
ib
u
tes
an
d
r
em
o
v
in
g
ir
r
el
ev
an
t
attr
ib
u
tes
th
at
will in
cr
ea
s
e
p
r
o
ce
s
s
in
g
tim
e
an
d
b
o
o
s
t sto
r
ag
e
p
er
f
o
r
m
an
c
e.
Featu
r
e
s
elec
tio
n
(
FS
)
is
d
e
f
in
ed
in
[
2
]
as th
e
p
r
o
ce
s
s
o
f
r
em
o
v
in
g
f
ea
tu
r
es
f
r
o
m
th
e
d
atab
a
s
e
th
at
ar
e
ir
r
elev
an
t
to
th
e
task
to
b
e
p
er
f
o
r
m
ed
.
Featu
r
e
s
elec
tio
n
p
r
o
m
o
tes
d
ata
co
m
p
r
e
h
en
s
io
n
,
r
e
d
u
ce
s
ca
lcu
latio
n
an
d
s
to
r
ag
e
r
e
q
u
ir
em
en
ts
,
r
ed
u
ce
s
co
m
p
u
tatio
n
al
p
r
o
ce
s
s
tim
e,
a
n
d
r
ed
u
ce
s
t
h
e
s
ize
o
f
t
h
e
d
ata
co
llectio
n
,
m
ak
in
g
m
o
d
el
lear
n
in
g
ea
s
ier
.
FS
h
a
s
b
ec
o
m
e
in
cr
ea
s
in
g
ly
p
o
p
u
l
ar
in
ap
p
licatio
n
s
in
g
en
o
m
i
cs,
h
ea
l
th
s
cien
ce
s
,
ec
o
n
o
m
ics,
b
a
n
k
in
g
,
am
o
n
g
o
t
h
er
s
[3
-
5]
as we
ll a
s
in
p
s
y
ch
o
lo
g
y
an
d
s
o
cial
s
cien
ce
s
[
6
,
7
]
.
Fe
atu
r
e
s
elec
tio
n
alg
o
r
ith
m
s
ca
teg
o
r
ized
in
to
2
m
ain
g
r
o
u
p
:
s
u
p
er
v
is
ed
,
u
n
s
u
p
er
v
is
ed
an
d
s
em
i
-
s
u
p
er
v
is
ed
;
th
is
r
elies u
p
o
n
wh
et
h
er
t
h
e
tr
ain
i
n
g
s
et
is
,
o
r
n
o
t,
la
b
elled
.
F
ea
tu
r
e
s
elec
tio
n
m
o
d
els ar
e
also
ca
teg
o
r
ized
in
to
f
ilter
,
wr
a
p
p
er
an
d
em
b
ed
d
e
d
m
o
d
els.
T
h
e
f
ir
s
t
o
n
es
ap
p
ly
s
tatis
tical
m
ea
s
u
r
es
to
ass
ig
n
a
s
co
r
e
to
ea
c
h
f
ea
tu
r
e;
f
ea
tu
r
es
ar
e
r
an
k
ed
b
y
th
ei
r
s
co
r
e,
an
d
eith
e
r
s
elec
ted
to
b
e
k
e
p
t
o
r
r
em
o
v
ed
f
r
o
m
th
e
d
ata
s
et.
Fil
ter
m
o
d
els
d
o
n
o
t
in
ter
ac
t
with
lear
n
in
g
alg
o
r
ith
m
s
,
an
d
th
ey
ca
n
b
e
u
n
i
v
ar
i
ate
(
wh
en
f
ea
t
u
r
es
ar
e
ev
alu
ated
o
n
e
b
y
o
n
e)
o
r
m
u
ltiv
ar
iate
(
wh
en
t
h
ey
ar
e
ev
alu
ated
in
s
u
b
s
ets).
W
r
ap
p
er
m
eth
o
d
s
d
ef
in
e
th
e
s
elec
tio
n
o
f
a
s
et
o
f
f
ea
tu
r
e
s
as
a
s
ea
r
ch
p
r
o
b
lem
,
w
h
er
e
d
if
f
er
en
t
co
m
b
in
atio
n
s
ar
e
p
r
ep
ar
ed
,
ev
alu
ate
d
an
d
co
m
p
ar
ed
to
o
th
e
r
co
m
b
in
atio
n
s
.
Fin
ally
,
th
e
u
n
d
er
ly
in
g
id
e
a
o
f
em
b
e
d
d
ed
m
o
d
els
is
lear
n
in
g
wh
ich
f
ea
tu
r
es
b
est co
n
tr
ib
u
te
to
th
ac
cu
r
ac
y
o
f
th
e
m
o
d
el
wh
ile
th
e
m
o
d
el
i
s
b
ein
g
cr
ea
ted
.
Featu
r
e
s
elec
tio
n
c
o
n
s
is
ts
o
f
f
o
u
r
s
tag
es,
ty
p
ically
r
ef
er
r
ed
t
o
as
s
u
b
s
et
c
r
ea
tio
n
,
s
u
b
s
et
e
v
alu
atio
n
,
s
to
p
cr
iter
io
n
,
an
d
r
esu
lt
v
alid
atio
n
.
Du
r
in
g
th
e
p
h
ase
o
f
s
u
b
s
et
ev
alu
atio
n
th
e
g
o
o
d
n
ess
o
f
a
s
u
b
s
et
p
r
o
d
u
ce
d
b
y
a
g
iv
en
s
u
b
s
et
g
en
er
atio
n
p
r
o
ce
d
u
r
e
is
m
ea
s
u
r
ed
.
E
x
am
p
les
o
f
s
u
b
s
et
ev
alu
atio
n
m
ea
s
u
r
es
f
o
r
m
u
ltiv
a
r
iate
f
ilter
m
eth
o
d
s
ar
e
th
e
d
i
s
tan
ce
[
8
]
,
th
e
u
n
ce
r
tain
ty
[
9
]
,
th
e
d
ep
en
d
en
c
e
[
1
0
]
,
an
d
th
e
c
o
n
s
is
ten
cy
[
4
]
,
w
h
ile
wr
ap
p
er
m
eth
o
d
s
m
o
s
tly
u
s
e
t
h
e
ac
cu
r
ac
y
[
1
1
]
.
T
h
e
s
to
p
p
in
g
cr
iter
io
n
estab
lis
h
es
wh
en
t
h
e
f
ea
tu
r
e
s
elec
tio
n
p
r
o
ce
s
s
m
u
s
t
f
in
is
h
;
it
ca
n
b
e
d
ef
in
ed
as
a
co
n
tr
o
l
p
r
o
ce
d
u
r
e
th
at
en
s
u
r
es
th
at
n
o
f
u
r
th
er
ad
d
itio
n
o
r
d
eletio
n
o
f
f
ea
tu
r
es
d
o
es
p
r
o
d
u
ce
a
b
etter
s
u
b
s
et,
o
r
it
ca
n
b
e
as
s
im
p
le
as
a
co
u
n
ter
o
f
iter
atio
n
s
.
Fin
a
lly
,
in
th
e
p
h
ase
o
f
r
esu
lt v
alid
atio
n
th
e
v
alid
ity
o
f
th
e
s
elec
ted
s
u
b
s
et
is
test
ed
.
A
r
ec
en
t
o
v
er
v
iew,
ca
teg
o
r
izatio
n
an
d
c
o
m
p
ar
is
o
n
o
f
ex
is
tin
g
m
eth
o
d
s
f
o
r
s
elec
tin
g
f
ea
tu
r
es
is
s
h
o
wn
in
[
1
2
]
.
A
s
ig
n
if
ican
t
d
o
wn
s
id
e
to
th
ese
tech
n
iq
u
es
is
th
at
th
ey
o
n
ly
c
o
n
s
id
er
a
s
in
g
le
cr
iter
io
n
wh
en
lo
o
k
i
n
g
f
o
r
a
s
u
b
s
et,
an
d
d
o
n
o
t
s
ee
k
to
lim
it
th
e
n
u
m
b
er
t
o
attr
ib
u
tes
ch
o
s
en
;
th
e
y
ca
n
th
en
b
e
r
e
f
er
r
ed
to
as
s
in
g
le
-
o
b
jectiv
e
f
ea
t
u
r
e
s
elec
t
io
n
m
eth
o
d
s
.
Ho
wev
e
r
,
th
e
s
in
g
le
m
ec
h
a
n
is
m
s
d
o
n
o
t
s
u
f
f
i
ce
wh
en
t
h
e
n
u
m
b
er
o
f
f
ea
t
u
r
es
is
p
ar
ticu
la
r
ly
h
ig
h
,
an
d
a
s
ep
ar
ate
f
ea
tu
r
e
s
elec
tio
n
p
r
o
ce
s
s
d
o
es
im
p
r
o
v
e
t
h
e
p
er
f
o
r
m
a
n
ce
s
o
f
th
e
lear
n
ed
m
o
d
el.
E
v
o
lu
tio
n
a
r
y
(
o
r
g
e
n
etic)
co
m
p
u
tatio
n
u
s
es a
s
im
p
le
ev
o
lu
ti
o
n
ar
y
m
etap
h
o
r
.
T
h
e
p
r
o
b
lem
,
ac
co
r
d
in
g
to
th
is
m
etap
h
o
r
,
p
lay
s
th
e
f
u
n
ctio
n
o
f
a
n
atm
o
s
p
h
er
e
in
wh
ich
a
p
o
p
u
latio
n
o
f
in
d
iv
i
d
u
als
r
esid
es,
ea
ch
r
ep
r
esen
tin
g
a
p
o
ten
tial
s
o
lu
t
io
n
to
th
e
p
r
o
b
lem
.
T
h
e
d
eg
r
ee
o
f
ad
a
p
tatio
n
o
f
ea
c
h
p
e
r
s
o
n
to
h
is
o
r
h
e
r
en
v
ir
o
n
m
en
t
is
e
x
p
r
ess
ed
b
y
a
m
ea
s
u
r
e
o
f
ad
eq
u
ac
y
k
n
o
wn
as
f
itn
ess
f
u
n
ctio
n
.
Un
lik
e
ev
o
lu
tio
n
in
n
atu
r
e,
ev
o
lu
tio
n
ar
y
alg
o
r
ith
m
s
h
av
e
th
e
ab
ilit
y
to
s
lo
wly
ev
o
lv
e
s
o
lu
tio
n
s
to
th
e
p
r
o
b
lem
.
Alg
o
r
ith
m
s
b
eg
in
with
an
in
itial
p
o
p
u
latio
n
o
f
r
an
d
o
m
s
o
lu
tio
n
s
an
d
,
in
ea
ch
iter
atio
n
,
th
e
b
est
in
d
iv
id
u
als
ar
e
s
elec
ted
an
d
co
m
b
in
e
d
u
s
in
g
v
ar
iatio
n
o
p
er
at
o
r
s
,
s
u
ch
as
cr
o
s
s
o
v
er
s
an
d
m
u
tatio
n
s
,
to
cr
ea
te
th
e
n
e
x
t
g
en
e
r
atio
n
.
T
h
e
cy
cle
is
r
ep
ea
te
d
u
n
til
ea
ch
o
f
th
e
s
to
p
cr
iter
ia
is
m
et.
So
m
e
p
r
o
b
le
m
s
in
v
o
lv
e
m
u
lti
-
o
b
jectiv
e
o
p
tim
izatio
n
(
MO
)
in
p
a
r
ticu
lar
wh
er
e
th
er
e
is
an
im
p
licit
ten
s
io
n
b
etwe
en
two
o
r
m
o
r
e
p
r
o
b
lem
o
b
jectiv
es;
th
e
s
elec
tio
n
f
u
n
ctio
n
,
in
wh
ic
h
o
n
e
m
u
s
t
o
p
tim
ize
th
e
ac
cu
r
a
cy
o
f
th
e
class
if
ier
an
d
r
ed
u
c
e
th
e
n
u
m
b
er
o
f
f
ea
t
u
r
es,
is
an
ex
am
p
le
o
f
s
u
ch
a
p
r
o
b
lem
.
Mu
lti
-
o
b
jectiv
e
ev
o
lu
tio
n
a
r
y
a
lg
o
r
ith
m
s
[
1
3
,
1
4
]
h
av
e
p
r
o
v
e
n
to
b
e
v
er
y
s
u
cc
ess
f
u
l
in
f
in
d
in
g
o
p
tim
al
s
o
lu
tio
n
s
to
m
u
ltip
le
o
b
jectiv
e
p
r
o
b
le
m
s
.
Mu
lti
-
o
b
jectiv
e
e
v
o
lu
tio
n
ar
y
alg
o
r
ith
m
s
ar
e
es
p
ec
ially
s
u
itab
le
f
o
r
m
u
lti
-
o
b
jectiv
e
o
p
tim
izatio
n
b
ec
au
s
e
th
ey
lo
o
k
f
o
r
m
u
ltip
le
o
p
tim
al
s
o
lu
tio
n
s
in
p
ar
allel
a
n
d
ar
e
a
b
le
to
f
in
d
a
s
et
o
f
o
p
tim
al
s
o
lu
tio
n
s
in
th
eir
f
in
al
p
o
p
u
latio
n
in
a
s
in
g
le
s
p
r
in
t.
W
h
en
an
o
p
tim
al
s
o
lu
t
io
n
s
et
is
av
ailab
le,
th
e
m
o
s
t
s
u
itab
le
s
o
lu
tio
n
ca
n
b
e
ch
o
s
en
b
y
ap
p
l
y
in
g
a
p
r
ef
e
r
en
ce
cr
iter
io
n
.
T
h
e
g
o
al
o
f
a
m
u
lti
-
o
b
jectiv
e
s
ea
r
ch
alg
o
r
ith
m
,
t
h
er
ef
o
r
e,
is
to
d
is
co
v
er
a
f
am
ily
o
f
s
o
lu
tio
n
s
th
at
ar
e
a
g
o
o
d
a
p
p
r
o
x
im
atio
n
to
th
e
Par
eto
f
r
o
n
t.
I
n
th
e
ca
s
e
o
f
m
u
lti
-
o
b
jectiv
e
f
ea
tu
r
e
s
elec
tio
n
,
ea
c
h
f
r
o
n
t
-
e
n
d
s
o
lu
tio
n
m
ay
r
ep
r
esen
t
a
s
u
b
s
et
o
f
f
ea
tu
r
es
with
an
r
elate
d
tr
ad
e
-
o
f
f
b
etwe
en
,
f
o
r
ex
am
p
le,
ac
cu
r
ac
y
an
d
m
o
d
el
co
m
p
lex
ity
.
I
n
m
u
lti
-
o
b
jectiv
e
f
ea
tu
r
e
s
el
ec
tio
n
m
eth
o
d
s
,
two
co
m
m
o
n
m
eth
o
d
s
ar
e
k
n
o
wn
as
E
N
OR
A
an
d
NSGA
-
I
I
.
E
NORA
(
ev
o
lu
tio
n
ar
y
n
o
n
-
d
o
m
i
n
ated
r
ad
ial
s
lo
ts
b
ased
alg
o
r
ith
m
)
is
o
n
e
o
f
t
h
e
m
u
lti
-
o
b
jectiv
e
ev
o
lu
tio
n
ar
y
alg
o
r
ith
m
s
elec
tio
n
tech
n
iq
u
es
f
o
r
r
an
d
o
m
s
ea
r
ch
[
1
5
,
1
6
]
with
t
h
e
f
o
llo
win
g
two
o
b
jectiv
es:
m
in
im
izin
g
th
e
n
u
m
b
er
o
f
s
elec
ted
f
ea
tu
r
es
an
d
m
in
im
i
zin
g
th
e
r
o
o
t
m
ea
n
s
q
u
a
r
ed
er
r
o
r
(
R
MSE
)
o
f
th
e
R
an
d
o
m
Fo
r
est
(
R
F)
m
o
d
el,
a
well
-
k
n
o
wn
r
eg
r
ess
io
n
m
o
d
el
lear
n
in
g
alg
o
r
ith
m
[
1
7
]
.
I
n
ad
d
itio
n
,
th
e
m
u
lti
-
o
b
jectiv
e
ev
o
lu
tio
n
ar
y
alg
o
r
ith
m
k
n
o
wn
as
th
e
NSGA
-
I
I
(
non
-
d
o
m
in
at
ed
s
o
r
ted
g
en
etic
alg
o
r
ith
m
)
[
1
8
]
is
co
n
s
id
er
e
d
a
n
o
r
m
in
th
e
m
u
lti
-
o
b
jectiv
e
ev
o
lu
tio
n
ar
y
co
m
p
u
tin
g
co
m
m
u
n
ity
,
b
o
th
i
n
ter
m
s
o
f
th
e
h
y
p
er
v
o
lu
m
e
s
tatis
tics
o
f
t
h
e
last
p
o
p
u
latio
n
an
d
in
ter
m
s
o
f
th
e
R
MSE
o
f
t
h
e
ch
o
s
en
p
er
s
o
n
.
T
h
e
NSGA
-
I
I
wr
ap
p
e
r
s
o
lu
tio
n
is
in
tr
o
d
u
ce
d
f
o
r
th
e
id
e
n
tific
atio
n
o
f
d
esig
n
ated
p
e
r
s
o
n
s
in
[
1
9
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
KA
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l Co
n
tr
o
l
I
n
teg
r
a
ted
b
i
o
-
s
ea
r
ch
a
p
p
r
o
a
c
h
es w
ith
mu
lti
-
o
b
jective
a
lg
o
r
ith
ms fo
r
… (
Mo
h
a
mma
d
A
iz
a
t
B
a
s
ir
)
2423
A
ch
an
g
e
in
th
e
d
o
m
in
an
t
r
ela
tio
n
s
h
ip
is
im
p
lem
en
ted
in
[
2
0
]
to
co
n
s
id
er
an
ar
b
itra
r
y
lar
g
e
n
u
m
b
e
r
o
f
g
o
als
an
d
is
u
s
ed
in
a
co
m
b
i
n
atio
n
o
f
NSGA
-
I
I
,
lo
g
is
tic
r
eg
r
ess
io
n
,
an
d
n
ai
v
e
B
ay
es
with
L
ap
lace
co
r
r
ec
tio
n
as
class
if
icatio
n
alg
o
r
ith
m
s
.
I
n
[
8
]
,
th
e
s
elec
tio
n
o
f
a
m
u
lti
-
o
b
jectiv
e
f
u
n
ctio
n
is
ap
p
lied
t
o
a
d
iag
n
o
s
tic
is
s
u
e
in
th
e
m
ed
icin
e.
Fo
r
an
a
p
p
l
icatio
n
in
en
g
in
ee
r
i
n
g
,
a
m
u
l
ti
-
o
b
jectiv
e
alg
o
r
ith
m
th
at
m
in
im
izes
th
e
er
r
o
r
id
en
tific
at
io
n
r
ate,
u
n
d
etec
ted
id
en
tific
atio
n
r
ate
an
d
th
e
n
u
m
b
er
o
f
s
elec
ted
f
ea
tu
r
es
is
p
r
o
p
o
s
ed
in
[
9
]
.
I
n
[
2
1
]
a
m
u
lti
-
o
b
j
ec
tiv
e
B
ay
esian
ar
tific
ial
im
m
u
n
e
s
y
s
tem
is
u
s
ed
f
o
r
th
e
s
elec
tio
n
o
f
f
ea
tu
r
es
in
class
if
icatio
n
p
r
o
b
lem
s
,
with
th
e
g
o
al
o
f
r
ed
u
cin
g
b
o
th
th
e
class
if
icatio
n
er
r
o
r
an
d
th
e
ca
r
d
in
ality
o
f
th
e
s
u
b
s
et
o
f
f
ea
tu
r
es.
I
n
[
1
0
]
a
wr
ap
p
er
ap
p
r
o
ac
h
is
p
r
o
p
o
s
ed
to
o
p
tim
ize
th
e
d
ata
m
in
in
g
alg
o
r
ith
m
er
r
o
r
r
ate
an
d
th
e
m
o
d
el
s
ize
o
f
th
e
lear
n
in
g
alg
o
r
it
h
m
u
s
in
g
NSGA
an
d
NSGA
-
I
I
.
A
m
u
lti
-
o
b
jectiv
e
esti
m
atio
n
o
f
th
e
d
i
s
tr
ib
u
tio
n
alg
o
r
ith
m
is
p
r
o
p
o
s
ed
in
[
1
1
]
f
o
r
th
e
s
elec
tio
n
o
f
a
f
u
n
ctio
n
s
u
b
s
et
b
ased
o
n
a
c
o
m
m
o
n
m
o
d
elin
g
o
f
o
b
jectiv
es
a
n
d
v
ar
iab
les.
Fig
u
r
e
1
s
h
o
ws th
e
co
m
p
lete
f
lo
w
o
f
E
NORA/NS
GAI
I
ad
ap
ted
f
r
o
m
[
2
2
]
.
Fig
u
r
e
1
.
Flo
w
ch
a
r
t o
f
an
E
N
OR
A/N
SGA
-
I
I
ad
ap
ted
f
r
o
m
[
2
2
]
A
m
u
lti
-
o
b
jectiv
e
ap
p
r
o
ac
h
t
o
th
e
co
llectio
n
o
f
f
u
n
ctio
n
s
u
b
s
ets
u
s
in
g
AC
O
an
d
f
u
zz
y
h
as
b
ee
n
p
r
o
p
o
s
ed
[
2
3
]
.
AC
O
was
u
s
ed
in
r
esear
c
h
to
ef
f
ec
tiv
ely
s
o
lv
e
th
e
f
u
zz
y
m
u
lti
-
o
b
jectiv
e
p
r
o
b
lem
.
T
h
eir
w
o
r
k
s
h
o
ws
th
at
th
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
ca
n
p
r
o
d
u
ce
b
etter
s
u
b
s
ets
an
d
ac
h
iev
e
h
ig
h
er
class
if
icatio
n
ac
cu
r
ac
y
.
AC
O
was
also
u
s
ed
with
a
g
en
etic
a
lg
o
r
ith
m
to
p
ic
k
a
f
u
n
ctio
n
f
o
r
p
atter
n
r
ec
o
g
n
itio
n
in
[
2
4
]
.
T
h
e
m
eth
o
d
c
o
n
s
is
ts
o
f
two
in
ter
esti
n
g
m
o
d
els,
t
h
e
v
is
ib
ilit
y
d
e
n
s
ity
m
o
d
el
(
VM
B
A
C
O)
an
d
th
e
p
h
e
r
o
m
o
n
e
d
en
s
ity
m
o
d
el
(
PMB
A
C
O)
f
o
r
th
e
o
p
tim
al
s
o
lu
tio
n
f
o
r
s
elec
tin
g
an
d
d
e
-
s
elec
tin
g
f
ea
tu
r
es.
Pro
m
is
in
g
r
esu
lts
h
av
e
b
ee
n
o
b
tain
ed
wh
er
e
th
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
d
em
o
n
s
tr
ates
r
o
b
u
s
tn
ess
an
d
ad
ap
tiv
e
ef
f
icien
cy
r
elativ
e
to
o
th
er
ap
p
r
o
ac
h
es.
Similar
ly
,
th
e
an
t
co
lo
n
y
o
p
tim
izatio
n
(
AC
O)
al
g
o
r
ith
m
was
u
s
ed
in
th
e
m
e
d
i
ca
l
f
ield
to
id
en
tify
im
p
o
r
tan
t
f
ea
tu
r
es
f
o
r
th
e
d
iag
n
o
s
is
o
f
R
am
an
-
b
ased
b
r
ea
s
t
c
an
ce
r
[
2
5
]
.
E
x
p
er
im
en
tal
r
esu
l
ts
d
em
o
n
s
tr
ated
th
at
AC
O
h
as
th
e
ca
p
ab
ilit
y
to
b
o
o
s
t
th
e
d
iag
n
o
s
tic
ac
cu
r
ac
y
o
f
R
am
an
-
b
ased
d
iag
n
o
s
tic
m
o
d
els.
Similar
ly
,
A
C
O
was
u
s
ed
in
th
e
ar
ea
o
f
n
etwo
r
k
s
ec
u
r
it
y
to
d
etec
t
in
tr
u
s
io
n
[
2
6
]
.
Fig
u
r
e
2
p
r
esen
ts
b
a
s
ic
p
s
eu
d
o
-
co
d
e
o
f
an
an
t a
lg
o
r
ith
m
.
New
m
eta
-
h
eu
r
is
tic
alg
o
r
ith
m
ar
tific
ial
b
ee
co
lo
n
y
(
AB
C
)
[
2
7
]
h
as
b
ee
n
u
s
ed
f
o
r
th
e
co
l
lectio
n
o
f
f
ea
tu
r
es
in
co
m
p
u
ted
to
m
o
g
r
a
p
h
y
(
C
T
Scan
)
im
ag
es
o
f
ce
r
v
i
ca
l
ca
n
ce
r
th
at
h
elp
to
r
ec
o
g
n
ize
ex
is
tin
g
ca
n
ce
r
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6
9
3
0
T
E
L
KOM
NI
KA
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l Co
n
tr
o
l
,
Vo
l.
18
,
No
.
5
,
Octo
b
e
r
2
0
2
0
:
2
4
2
1
-
2431
2424
F
o
r
th
e
h
an
d
lin
g
o
f
h
ig
h
d
im
e
n
s
io
n
al
p
r
o
b
lem
s
,
[
2
8
]
s
u
g
g
est
ed
a
n
ew
m
eth
o
d
o
f
s
elec
tio
n
o
f
f
ea
t
u
r
es
b
ased
o
n
AB
C
with
g
r
ad
ien
t
-
b
o
o
s
tin
g
d
ec
is
io
n
tr
ee
.
T
h
e
r
esear
ch
r
esu
lt
h
as
s
h
o
wn
th
at
t
h
e
p
r
o
p
o
s
ed
m
eth
o
d
ef
f
ec
tiv
ely
r
ed
u
ce
s
th
e
s
ize
o
f
th
e
d
atase
t
an
d
ac
h
iev
es
s
u
p
er
io
r
class
i
f
icatio
n
ac
cu
r
ac
y
b
y
u
s
in
g
th
e
s
elec
ted
f
ea
tu
r
es.
Simila
r
ly
,
th
e
h
y
b
r
id
ap
p
r
o
ac
h
[
2
9
]
u
s
ed
th
e
AB
C
alg
o
r
ith
m
with
a
d
if
f
e
r
en
tial e
v
o
lu
tio
n
alg
o
r
ith
m
to
a
d
d
r
ess
th
e
h
ig
h
d
im
en
s
io
n
al
p
r
o
b
lem
.
T
h
e
d
ev
elo
p
e
d
h
y
b
r
id
ap
p
r
o
ac
h
d
em
o
n
s
tr
ates
th
e
ab
ilit
y
t
o
p
ick
g
o
o
d
f
ea
t
u
r
es
f
o
r
th
e
class
if
icatio
n
task
s
an
d
th
u
s
in
cr
ea
s
es
th
e
r
u
n
-
t
im
e
ef
f
icien
cy
an
d
ac
cu
r
ac
y
o
f
th
e
class
if
ier
.
A
m
u
lti
-
o
b
jectiv
e
ar
tific
ial
b
e
e
co
lo
n
y
(
MO
AB
C
)
m
o
d
el
h
a
s
b
ee
n
d
ev
elo
p
e
d
[
3
0
]
.
T
h
e
d
ev
elo
p
ed
alg
o
r
ith
m
was
in
co
r
p
o
r
ate
d
with
a
f
u
zz
y
ap
p
r
o
ac
h
to
ev
al
u
atin
g
t
h
e
r
elev
an
ce
o
f
th
e
f
u
n
ctio
n
s
u
b
s
ets.
E
x
p
er
im
en
tal
f
in
d
in
g
s
in
d
icate
a
s
u
b
s
tan
tial
co
n
tr
ib
u
tio
n
to
s
ee
k
in
g
a
s
u
c
ce
s
s
f
u
l
s
u
b
s
et
o
f
f
ea
tu
r
es.
F
ig
u
r
e
3
d
em
o
n
t
r
ates
b
asic p
s
eu
d
o
-
co
d
e
o
f
b
ee
alg
o
r
ith
m
.
Fig
u
r
e
2
.
T
h
e
b
asic
p
s
eu
d
o
-
co
d
e
o
f
a
n
a
nt
alg
o
r
ith
m
Fig
u
r
e
3
.
T
h
e
p
s
eu
d
o
-
co
d
e
o
f
a
b
ee
alg
o
r
ith
m
B
at
alg
o
r
ith
m
h
as
b
ee
n
u
s
ed
ef
f
ec
tiv
ely
in
e
n
g
in
ee
r
i
n
g
[
3
1
]
.
Mu
lti
-
o
b
jectiv
e
b
in
ar
y
b
at
alg
o
r
ith
m
(
MBB
A)
p
r
o
p
o
s
ed
b
y
[
3
2
]
m
o
d
if
ied
b
at
p
o
s
itio
n
u
p
d
ate
s
tr
ateg
y
th
at
wo
r
k
s
b
etter
with
b
in
ar
y
p
r
o
b
lem
s
an
d
also
im
p
lem
en
ted
m
u
tatio
n
o
p
er
ato
r
t
o
b
o
o
s
t
lo
ca
l
s
ea
r
ch
c
ap
ab
ilit
y
an
d
s
u
p
p
o
r
t
th
e
d
iv
e
r
s
ity
o
f
alg
o
r
ith
m
s
.
T
h
e
ex
p
er
im
e
n
tal
r
esu
lts
s
h
o
w
th
at
th
e
p
r
o
p
o
s
e
d
MBB
A
is
a
co
m
p
etitiv
e
m
u
lti
-
o
b
jectiv
e
alg
o
r
ith
m
th
at
o
u
tp
er
f
o
r
m
s
NSGA
-
I
I
.
B
at
alg
o
r
ith
m
h
as
also
b
ee
n
u
s
ed
in
th
e
ar
ea
o
f
r
en
ewa
b
le
en
e
r
g
y
[
3
3
]
,
wh
ich
h
as
g
r
ea
t
p
o
ten
tial
f
o
r
ap
p
licatio
n
o
f
th
e
p
r
o
p
o
s
ed
alg
o
r
ith
m
to
th
e
win
d
p
o
wer
n
etwo
r
k
.
Similar
ly
,
in
th
e
m
ed
ical
s
ec
to
r
,
a
m
o
d
if
ied
b
at
alg
o
r
ith
m
(
MBA)
f
o
r
f
ea
tu
r
e
s
elec
tio
n
d
e
v
elo
p
ed
b
y
[
3
4
]
p
er
f
o
r
m
ed
s
ig
n
if
ic
an
tly
well
to
r
em
o
v
e
u
n
wan
ted
an
d
r
ep
etitiv
e
d
ata
o
n
b
r
ea
s
t
ca
n
ce
r
p
r
io
r
to
d
iag
n
o
s
is
.
I
n
[
3
5
]
,
th
e
h
y
b
r
id
b
i
n
ar
y
b
at
en
h
an
ce
d
p
a
r
ticle
s
war
m
o
p
tim
izatio
n
alg
o
r
ith
m
(
HB
B
E
PS
O)
w
as
d
ev
elo
p
ed
an
d
claim
ed
to
h
a
v
e
th
e
ab
ilit
y
to
s
ca
n
t
h
e
f
ea
tu
r
e
s
p
ac
e
f
o
r
ap
p
r
o
p
r
iate
co
m
b
in
a
tio
n
s
o
f
f
ea
tu
r
es.
Fig
u
r
e
4
o
u
tl
in
e
th
e
b
asic p
s
eu
d
o
-
co
d
e
o
f
b
at
alg
o
r
ith
m
.
A
m
u
lti
-
o
b
jectiv
e
alg
o
r
ith
m
b
ased
o
n
a
cu
ck
o
o
s
ea
r
ch
al
g
o
r
i
th
m
h
as
b
ee
n
ap
p
lied
to
th
e
o
p
tim
izatio
n
p
r
o
b
lem
[
3
6
-
3
8
]
.
I
n
th
e
d
im
e
n
s
io
n
al
r
ed
u
ctio
n
p
r
o
b
lem
,
a
n
ew
m
u
lti
-
o
b
jectiv
e
cu
ck
o
o
s
ea
r
ch
alg
o
r
ith
m
[
3
9
]
h
as
b
ee
n
d
ev
elo
p
e
d
to
s
ea
r
ch
th
e
s
p
ac
e
attr
ib
u
te
with
m
in
im
al
co
r
r
elatio
n
b
etwe
en
th
e
s
elec
ted
attr
ib
u
tes.
E
x
p
er
im
en
tal
f
in
d
in
g
s
h
av
e
s
h
o
wn
th
at
th
e
p
r
o
p
o
s
ed
m
u
lti
-
o
b
jectiv
e
C
S
m
eth
o
d
h
as
s
u
cc
ess
f
u
lly
o
u
tp
e
r
f
o
r
m
ed
p
ar
ticle
s
war
m
o
p
tim
izatio
n
(
P
SO)
an
d
g
en
etic
alg
o
r
ith
m
(
G
A)
o
p
tim
izatio
n
alg
o
r
ith
m
s
.
F
o
r
ex
a
m
p
le,
a
h
y
b
r
id
r
o
u
g
h
s
et
b
ased
o
n
a
m
o
d
i
f
ied
cu
c
k
o
o
s
ea
r
ch
alg
o
r
ith
m
h
a
s
b
ee
n
p
r
o
p
o
s
ed
[
3
9
]
.
T
h
e
al
g
o
r
ith
m
d
ev
el
o
p
e
d
d
em
o
n
s
tr
ates
th
e
ab
ilit
y
t
o
r
e
d
u
ce
th
e
n
u
m
b
e
r
o
f
f
ea
tu
r
es
i
n
th
e
r
ed
u
ctio
n
s
et
with
o
u
t
lo
s
in
g
th
e
ac
cu
r
ac
y
o
f
th
e
class
if
icatio
n
.
I
n
[
4
0
]
also
p
r
o
p
o
s
ed
a
p
r
e
d
ictio
n
alg
o
r
it
h
m
f
o
r
h
ea
r
t
d
is
ea
s
e
b
ased
o
n
th
e
cu
ck
o
o
s
ea
r
ch
s
y
s
tem
.
T
wo
alg
o
r
ith
m
s
,
n
am
ely
cu
ck
o
o
s
ea
r
ch
alg
o
r
ith
m
(
C
SA)
an
d
cu
c
k
o
o
o
p
tim
izatio
n
alg
o
r
ith
m
(
C
OA)
,
h
av
e
b
ee
n
u
s
ed
f
o
r
s
u
b
s
et
g
en
er
atio
n
an
d
th
e
r
esu
lts
s
h
o
w
th
at
b
o
th
alg
o
r
ith
m
s
h
a
v
e
ac
h
ie
v
ed
b
etter
p
r
ed
ictiv
e
ac
cu
r
ac
y
o
n
s
elec
ted
d
atasets
.
Fig
u
r
e
5
s
u
m
m
a
r
is
e
th
e
g
en
er
al
p
s
eu
d
o
-
c
o
d
e
o
f
C
u
ck
o
o
alg
o
r
ith
m
.
Fire
f
ly
alg
o
r
ith
m
h
as
b
ee
n
in
v
en
ted
b
y
Yan
g
[
4
1
]
a
n
d
h
as
b
ee
n
u
s
ed
in
m
a
n
y
ar
ea
s
,
es
p
ec
ially
in
th
e
s
elec
tio
n
o
f
ap
p
s
.
New
f
ir
ef
ly
alg
o
r
ith
m
b
ased
o
n
th
e
A
d
a
-
b
o
o
s
t
m
eth
o
d
h
as
r
ec
e
n
tly
b
ee
n
d
ev
el
o
p
ed
in
th
e
m
ed
ical
f
ield
[
4
2
]
to
d
ia
g
n
o
s
e
liv
er
ca
n
ce
r
.
T
h
e
d
ev
el
o
p
ed
h
y
b
r
id
m
eth
o
d
u
s
ed
b
y
f
ir
ef
ly
alg
o
r
ith
m
t
o
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
KA
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l Co
n
tr
o
l
I
n
teg
r
a
ted
b
i
o
-
s
ea
r
ch
a
p
p
r
o
a
c
h
es w
ith
mu
lti
-
o
b
jective
a
lg
o
r
ith
ms fo
r
… (
Mo
h
a
mma
d
A
iz
a
t
B
a
s
ir
)
2425
im
p
r
o
v
e
t
h
e
r
esu
ltin
g
Ad
a
-
b
o
o
s
t
alg
o
r
ith
m
ca
n
h
elp
p
h
y
s
ician
s
r
ec
o
g
n
ize
an
d
class
if
y
s
af
e
an
d
u
n
h
ea
lth
f
u
l
in
d
iv
id
u
als.
I
t
ca
n
also
b
e
u
s
ed
in
m
ed
ical
ce
n
ter
s
to
im
p
r
o
v
e
ac
c
u
r
ac
y
an
d
s
p
ee
d
an
d
r
ed
u
ce
co
s
ts
.
I
n
ad
d
itio
n
,
[
4
3
]
p
r
o
p
o
s
es
th
e
c
o
llectio
n
o
f
f
ea
tu
r
es
in
t
h
e
Ar
a
b
ic
tex
t
class
if
icatio
n
b
ased
o
n
f
ir
ef
ly
alg
o
r
ith
m
.
T
h
e
p
r
o
p
o
s
ed
alg
o
r
ith
m
h
as
b
ee
n
s
u
cc
ess
f
u
lly
a
p
p
lied
t
o
v
ar
io
u
s
co
m
b
in
ato
r
ial
p
r
o
b
le
m
s
an
d
h
as
ac
h
iev
e
d
h
ig
h
p
r
ec
is
io
n
in
th
e
d
ev
elo
p
m
en
t
o
f
th
e
Ar
ab
ic
tex
t
class
if
icatio
n
.
I
n
th
e
m
u
lti
-
o
b
jectiv
e
q
u
esti
o
n
,
th
e
f
ir
ef
ly
alg
o
r
ith
m
was
s
u
cc
ess
f
u
lly
a
p
p
lied
to
th
e
s
ch
ed
u
lin
g
p
r
o
b
lem
f
ield
,
s
u
ch
as
in
[
4
4
-
46]
.
Fig
u
r
e
6
p
r
esen
ts
th
e
b
asic p
s
eu
d
o
-
c
o
d
e
o
f
f
ir
e
f
ly
alg
o
r
ith
m
.
Fig
u
r
e
4.
T
h
e
p
s
eu
d
o
-
co
d
e
o
f
a
b
at
alg
o
r
ith
m
Fig
u
r
e
5.
T
h
e
p
s
eu
d
o
-
co
d
e
o
f
a
cu
ck
o
o
s
ea
r
ch
alg
o
r
ith
m
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6
9
3
0
T
E
L
KOM
NI
KA
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l Co
n
tr
o
l
,
Vo
l.
18
,
No
.
5
,
Octo
b
e
r
2
0
2
0
:
2
4
2
1
-
2431
2426
Fig
u
r
e
6
.
T
h
e
p
s
eu
d
o
-
co
d
e
o
f
a
f
ir
ef
ly
alg
o
r
ith
m
I
n
th
is
p
ap
er
,
we
s
u
g
g
est
an
o
p
tim
al
co
m
b
in
atio
n
o
f
a
s
ele
ctio
n
m
ec
h
an
is
m
b
ased
o
n
ev
o
lu
tio
n
ar
y
s
u
b
s
et
g
en
er
atio
n
.
W
r
ap
p
er
a
n
d
f
ilter
ed
a
p
p
r
o
ac
h
es
h
av
e
b
ee
n
u
s
ed
.
B
io
-
s
ea
r
ch
al
g
o
r
ith
m
s
h
av
e
b
ee
n
co
m
b
in
ed
with
E
NORA
an
d
NSG
A
-
I
I
to
p
er
f
o
r
m
th
e
o
p
tim
u
m
co
llec
tio
n
o
f
ap
p
s
.
I
n
s
p
ir
ed
b
y
th
e
a
b
ilit
y
o
f
b
io
-
s
ea
r
ch
alg
o
r
ith
m
s
to
s
elec
t
f
ea
tu
r
es,
th
e
p
u
r
p
o
s
e
o
f
th
is
p
ap
er
is
to
p
r
esen
t
o
p
tim
ized
E
NORA
an
d
NSGA
-
I
I
alg
o
r
ith
m
s
b
y
d
e
p
lo
y
in
g
b
io
-
s
ea
r
ch
alg
o
r
ith
m
s
to
o
b
tain
an
o
p
tim
u
m
n
u
m
b
er
o
f
attr
ib
u
tes
f
o
r
s
elec
ted
d
ata
s
ets.
T
h
e
k
ey
co
n
ce
p
t
is
to
in
co
r
p
o
r
ate
in
teg
r
ated
alg
o
r
ith
m
s
b
y
n
u
m
e
r
o
u
s
r
e
d
u
ctio
n
s
b
etwe
en
m
u
lti
-
o
b
jectiv
e
alg
o
r
ith
m
s
an
d
b
io
-
s
ea
r
c
h
alg
o
r
ith
m
s
f
o
r
th
e
co
llectio
n
o
f
f
ea
tu
r
es.
Descr
ip
tio
n
o
f
th
e
ex
ec
u
tio
n
s
tep
s
ar
e
lis
ted
in
th
e
n
ex
t sectio
n
.
2.
RE
S
E
ARC
H
M
E
T
H
O
D
Me
th
o
d
o
lo
g
y
o
f
th
is
p
ap
e
r
is
r
ep
r
esen
ted
in
Fig
u
r
e
7
h
as
b
ee
n
p
r
esen
ted
in
th
e
f
o
r
m
o
f
t
h
e
wo
r
k
f
lo
w
.
I
t c
o
n
s
is
ts
o
f
s
er
ies o
f
s
tep
s
an
d
m
en
tio
n
in
d
etails th
r
o
u
g
h
o
u
t th
is
s
ec
tio
n
.
−
Step
1
Data
co
llectio
n
:
d
atasets
wer
e
s
elec
ted
f
r
o
m
UC
I
Ma
ch
in
e
L
ea
r
n
in
g
R
ep
o
s
ito
r
y
[
4
7
]
(
r
ef
er
T
ab
le
1
f
o
r
p
r
o
f
ile
o
f
t
h
e
s
elec
ted
d
atas
ets
)
.
T
h
ese
d
atasets
co
n
s
is
t
o
f
v
ar
io
u
s
s
izes
an
d
m
ix
d
o
m
ain
s
in
o
r
d
e
r
to
ex
am
in
e
th
e
ca
p
ab
ilit
y
o
f
alg
o
r
ith
m
s
to
p
er
f
o
r
m
attr
ib
u
te
s
elec
tio
n
.
−
Step
2
Da
ta
h
an
d
lin
g
:
m
is
s
in
g
v
alu
es
in
th
e
d
ataset
h
as
b
ee
n
p
r
e
-
p
r
o
ce
s
s
ed
to
b
e
r
ea
d
y
f
o
r
ex
p
er
i
m
en
tatio
n
.
Data
s
et
th
at
h
as m
is
s
in
g
v
alu
e
(
s
y
m
b
o
lized
as
‘
?’
in
o
r
ig
in
al
d
ataset)
s
h
o
u
ld
b
e
r
ep
lace
d
eit
h
er
with
0
o
r
m
ea
n
v
alu
e.
B
o
th
m
eth
o
d
s
h
av
e
b
ee
n
test
ed
an
d
a
r
e
s
u
lt
in
d
icate
s
in
s
ig
n
if
ican
t
d
if
f
er
en
ce
in
ter
m
s
o
f
p
er
f
o
r
m
an
ce
.
T
h
is
r
esear
ch
d
ec
id
ed
u
s
in
g
v
alu
e
o
f
“0
”
to
b
e
r
ep
lace
d
f
o
r
m
is
s
in
g
v
alu
es.
−
Step
3
L
o
ad
clea
n
d
atasets
:
all
d
atase
ts
h
av
e
b
ee
n
tr
ain
ed
an
d
test
ed
u
s
in
g
W
E
KA
s
o
f
twar
e.
W
E
K
A
also
h
as
b
ee
n
u
s
ed
to
d
o
th
e
d
ata
p
r
e
-
p
r
o
ce
s
s
in
g
in
s
tep
2
.
I
n
W
E
KA
s
o
f
twar
e,
th
e
d
etailed
p
ar
a
m
eter
s
ettin
g
f
o
r
all
alg
o
r
ith
m
s
h
as b
ee
n
s
et
u
p
to
b
e
f
u
r
t
h
er
ex
p
er
im
en
ted
in
s
tep
(
4
)
a
n
d
s
tep
(
5
)
as sh
o
wn
in
T
ab
le
2
.
−
Step
4
Su
b
s
et
g
en
er
atio
n
(
1
)
:
in
th
is
s
tep
,
two
(
2
)
r
ed
u
ctio
n
p
r
o
c
ess
es
wh
ich
ar
e
E
NO
R
A
an
d
NSGA
-
II
alg
o
r
ith
m
s
with
f
ilter
ed
m
eth
o
d
h
av
e
b
ee
n
ex
ec
u
ted
.
T
h
e
o
u
t
p
u
t
o
f
th
is
f
ir
s
t
s
u
b
s
et
g
en
er
ati
o
n
co
n
s
id
er
ed
n
o
t
an
o
p
tim
al
s
u
b
s
et
an
d
n
ee
d
t
o
b
e
f
u
r
th
e
r
ed
r
e
d
u
ce
d
.
T
h
e
ex
ten
d
ed
r
ed
u
ctio
n
is
n
ee
d
ed
to
g
et
a
n
o
p
tim
al
r
ed
u
ctio
n
w
h
ich
b
ee
n
d
o
n
e
in
s
tep
(
5
)
.
−
Step
5
Su
b
s
et
g
en
er
atio
n
(
2
)
:
in
th
is
s
tep
,
th
e
o
u
tp
u
t
in
s
tep
(
4
)
will
b
e
f
u
r
th
er
ed
r
ed
u
ce
d
with
f
iv
e
(
5
)
b
io
-
s
ea
r
ch
m
eth
o
d
s
(
an
t,
b
at,
b
ee
,
cu
ck
o
o
a
n
d
f
ir
ef
l
y
)
+
wr
ap
p
er
u
s
ed
in
o
r
d
er
to
s
ea
r
ch
f
o
r
th
e
o
p
tim
al
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
KA
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l Co
n
tr
o
l
I
n
teg
r
a
ted
b
i
o
-
s
ea
r
ch
a
p
p
r
o
a
c
h
es w
ith
mu
lti
-
o
b
jective
a
lg
o
r
ith
ms fo
r
… (
Mo
h
a
mma
d
A
iz
a
t
B
a
s
ir
)
2427
att
r
ib
u
tes.
T
h
is
ex
p
er
im
en
t
p
r
o
ce
s
s
r
ef
lects
r
esear
ch
d
o
n
e
i
n
[
4
8
]
wh
ich
claim
ed
th
at
b
alan
ce
o
f
e
x
p
lo
itatio
n
an
d
ex
p
lo
r
atio
n
n
ee
d
to
b
e
ac
co
m
p
lis
h
ed
f
o
r
e
f
f
icien
t
s
p
ac
e
s
ea
r
ch
in
g
.
T
h
is
s
ec
o
n
d
g
en
e
r
atio
n
o
f
th
e
s
u
b
s
et
co
n
s
id
er
ed
a
n
o
p
tim
al
s
u
b
s
et.
−
Step
6
Su
b
s
et
ev
alu
atio
n
:
in
th
is
s
tep
,
th
e
o
u
tp
u
t
o
f
s
u
b
s
et
g
en
er
ati
o
n
(
1
)
a
n
d
s
u
b
s
et
g
en
er
atio
n
(
2
)
will
b
e
ev
alu
ated
th
r
o
u
g
h
class
if
icatio
n
p
er
f
o
r
m
an
ce
.
T
h
is
s
tep
is
to
co
n
f
ir
m
t
h
e
p
er
f
o
r
m
a
n
ce
o
f
s
u
b
s
et
g
en
er
atio
n
with
g
o
o
d
class
if
icatio
n
ac
cu
r
ac
y
i
n
o
r
d
e
r
to
p
r
o
d
u
ce
an
o
p
tim
al
f
ea
tu
r
e
s
elec
tio
n
m
o
d
el.
−
Step
7
Pro
d
u
ctio
n
o
f
o
p
tim
al
f
ea
tu
r
e
s
elec
tio
n
m
o
d
el:
I
n
th
is
f
in
al
s
tep
,
v
a
r
io
u
s
co
m
b
in
atio
n
s
o
f
b
io
-
s
ea
r
ch
m
eth
o
d
s
an
d
r
ed
u
ctio
n
alg
o
r
i
th
m
s
wer
e
ca
r
e
f
u
lly
s
elec
ted
to
p
er
f
o
r
m
a
f
ea
tu
r
e
s
elec
tio
n
m
o
d
el.
Op
tim
al
n
u
m
b
er
s
o
f
r
ed
u
ctio
n
s
with
g
o
o
d
class
if
icatio
n
ac
cu
r
ac
y
a
r
e
th
e
c
r
iter
ia
f
o
r
ch
o
o
s
in
g
th
e
b
e
s
t selec
ted
lis
t.
Fig
u
r
e
7
.
Me
th
o
d
o
lo
g
y
o
f
th
e
r
esear
ch
T
ab
le
1
.
Pro
f
ile
o
f
th
e
s
elec
ted
d
atasets
S
i
z
e
D
a
t
a
s
e
t
#
A
t
t
r
#
I
n
st
#
C
l
a
ss
S
mal
l
B
r
e
a
s
t
c
a
n
c
e
r
9
3
6
7
2
S
mal
l
P
a
r
k
i
n
s
o
n
22
1
9
7
2
S
mal
l
O
z
o
n
e
72
2
5
3
6
2
M
e
d
i
u
m
C
l
e
a
n
1
1
6
6
4
7
6
2
M
e
d
i
u
m
S
e
mei
o
n
2
6
5
1
5
9
3
2
La
r
g
e
Emai
l
s
4
7
0
2
64
2
La
r
g
e
G
i
set
t
e
5
0
0
0
1
3
5
0
0
2
La
r
g
e
A
r
c
e
n
e
1
0
0
0
0
9
0
0
2
T
ab
le
2
.
Deta
ils
p
ar
am
eter
s
ett
in
g
S
e
a
r
c
A
l
g
o
P
o
p
u
l
a
t
i
o
n
S
i
z
e
S
p
e
c
i
f
i
c
se
t
t
i
n
g
A
n
t
20
Ev
a
p
o
r
a
t
i
o
n
r
a
t
e
:
0
.
9
|
P
h
e
r
o
m
o
n
e
r
a
t
e
:
2
.
0
|
H
e
u
r
i
s
t
i
c
r
a
t
e
:
0
.
7
Bat
20
F
r
e
q
u
e
n
c
y
:
0
.
5
|
Lo
u
d
n
e
ss:
0
.
5
Bee
30
R
a
d
i
u
s Da
mp
:
0
.
9
8
|
R
a
d
i
u
s
M
u
t
a
t
i
o
n
:
0
.
8
0
C
u
c
k
o
o
20
P
a
r
a
t
e
:
0
.
2
5
|
S
i
g
m
a
r
a
t
e
:
0
.
7
0
F
i
r
e
f
l
y
20
B
e
t
a
z
e
r
o
:
0
.
3
3
|
A
b
s
o
r
p
t
i
o
n
C
o
e
f
f
i
c
i
e
n
t
:
0
.
0
0
1
EN
O
R
A
1
0
0
G
e
n
e
r
a
t
i
o
n
:
1
0
N
S
G
A
-
II
1
0
0
G
e
n
e
r
a
t
i
o
n
:
1
0
*
F
i
x
e
d
s
e
t
t
i
n
g
f
o
r
a
l
l
b
i
o
-
sea
r
c
h
a
l
g
o
r
i
t
h
ms
:
I
t
e
r
a
t
i
o
n
:
2
0
,
M
u
t
a
t
i
o
n
P
r
o
b
a
b
i
l
i
t
y
:
0
.
0
1
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6
9
3
0
T
E
L
KOM
NI
KA
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l Co
n
tr
o
l
,
Vo
l.
18
,
No
.
5
,
Octo
b
e
r
2
0
2
0
:
2
4
2
1
-
2431
2428
3.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
T
ab
le
3
s
h
o
ws
th
e
co
m
p
ar
is
o
n
o
f
r
e
d
u
ctio
n
p
er
f
o
r
m
an
ce
b
etwe
en
E
NORA
vs
NS
GA
-
II
in
th
e
f
ir
s
t
s
u
b
s
et
g
en
er
atio
n
an
d
s
e
co
n
d
s
u
b
s
et
g
en
er
atio
n
.
I
t
ca
n
b
e
s
ee
n
th
at
E
NORA+f
ilter
ed
m
eth
o
d
m
an
ag
ed
to
r
ed
u
ce
th
e
attr
ib
u
tes
f
o
r
s
ev
e
n
(
7
)
d
at
asets
(
Ozo
n
e,
Par
k
in
s
o
n
,
C
lean
1
,
Sem
eio
n
,
E
m
ails
,
Gis
ette,
Ar
ce
n
e)
ex
ce
p
t
f
o
r
B
r
ea
s
tcan
ce
r
d
atasets
wh
er
e
t
h
e
o
r
ig
in
al
attr
ib
u
tes
r
em
ain
e
d
.
Sem
eio
n
,
E
m
ails
,
Gis
ette
a
n
d
Ar
ce
n
e
d
atasets
ac
h
iev
ed
m
o
r
e
th
a
n
9
5
%
r
ed
u
ctio
n
.
Similar
s
itu
atio
n
with
N
SGA
-
I
I
wh
er
e
th
e
f
ir
s
t
s
u
b
s
et
g
en
er
atio
n
ac
h
iev
ed
m
o
r
e
attr
ib
u
te
r
e
d
u
ctio
n
th
a
n
E
NORA.
E
m
ails
d
an
Gis
ette
d
atasets
h
av
e
r
ea
ch
e
d
alm
o
s
t
1
0
0
%
r
ed
u
ctio
n
wh
i
c
h
is
ex
tr
em
e
ca
s
es
to
b
e
co
n
s
id
er
ed
in
t
h
e
f
ir
s
t
s
u
b
s
et
g
en
er
atio
n
.
Ho
wev
er
,
th
e
m
ass
iv
e
r
ed
u
c
tio
n
u
s
in
g
E
NORA
an
d
NSGA
-
II
o
f
th
ese
attr
ib
u
te
s
with
f
ilter
ed
ap
p
r
o
ac
h
s
till
d
o
es
n
o
t
ap
p
r
o
v
e
t
h
e
o
p
tim
al
s
elec
tio
n
.
E
v
en
th
o
u
g
h
th
e
p
er
f
o
r
m
an
ce
o
f
NSGA
-
I
I
b
ett
er
th
an
E
NORA
in
ter
m
o
f
m
u
ch
less
s
elec
ted
attr
ib
u
tes
i
n
f
ir
s
t
r
ed
u
ctio
n
,
th
is
co
n
d
itio
n
s
till
n
o
t
p
r
o
m
is
in
g
to
g
et
th
e
o
p
tim
al
s
et
o
f
attr
ib
u
tes.
T
h
e
s
ec
o
n
d
s
u
b
s
et
g
en
er
atio
n
n
ee
d
to
b
e
ex
ec
u
ted
to
o
b
tain
ab
s
o
lu
te
o
p
tim
al
r
ed
u
ctio
n
s
et.
E
x
ten
d
ed
ex
p
er
im
en
t
h
as
b
ee
n
co
n
d
u
cted
to
o
p
tim
ize
th
e
E
NORA
an
d
NSGA
-
I
I
alg
o
r
ith
m
s
with
f
i
v
e
(
5
)
b
io
-
s
ea
r
c
h
alg
o
r
ith
m
a
n
d
wr
ap
p
e
r
m
eth
o
d
.
A
r
esu
lt
s
h
o
ws
m
o
r
e
r
ed
u
ctio
n
h
ap
p
en
ed
f
o
r
all
d
atasets
.
E
x
tr
em
e
ca
s
e
h
as
b
ee
n
d
is
co
v
er
ed
b
y
Ozo
n
e
d
ataset
wh
er
e
twelv
e
(
1
2
)
attr
ib
u
te
s
in
th
e
f
ir
s
t
r
ed
u
ctio
n
with
E
NORA
h
av
e
b
ee
n
r
ed
u
ce
d
to
o
n
ly
o
n
e
(
1
)
attr
ib
u
te
in
t
h
e
s
ec
o
n
d
r
ed
u
ctio
n
.
Sam
e
r
esu
lt
also
b
ee
n
ac
h
iev
ed
with
NSGA
-
II.
Fu
r
th
er
ex
p
er
im
en
t
b
ee
n
c
o
n
d
u
cted
to
o
p
tim
ize
th
e
E
NORA
an
d
NSGA
-
I
I
alg
o
r
ith
m
s
with
f
iv
e
(
5
)
b
io
-
s
ea
r
ch
alg
o
r
ith
m
a
n
d
wr
ap
p
er
m
et
h
o
d
.
R
esu
lts
s
h
o
ws
s
u
p
er
io
r
r
e
d
u
ctio
n
f
o
r
all
d
ata
s
ets
f
o
r
E
NORA
an
d
NSGA
-
I
I
.
Ozo
n
e
d
ataset
m
ain
tain
th
e
s
am
e
r
esu
lt
as
all
s
ea
r
ch
in
g
s
p
ac
e
h
as
b
ee
n
f
u
lly
ex
p
lo
r
e
d
.
Ov
er
all,
all
b
io
-
s
ea
r
ch
alg
o
r
ith
m
s
s
u
cc
ee
d
ed
to
ac
q
u
ir
e
n
ea
r
-
o
p
tim
al
s
o
lu
tio
n
s
(
o
p
tim
al
f
ea
tu
r
es)
i
n
s
ec
o
n
d
s
u
b
s
et
g
en
er
atio
n
s
.
T
h
is
r
esu
lt
co
n
f
ir
m
ed
th
e
a
d
ap
tiv
e
b
eh
a
v
io
r
o
f
b
io
-
s
ea
r
ch
alg
o
r
ith
m
with
wr
a
p
p
er
m
eth
o
d
s
to
p
er
f
o
r
m
o
p
ti
m
al
f
ea
tu
r
es
s
elec
tio
n
f
o
r
E
N
OR
A
an
d
NSGA
-
I
I
alg
o
r
ith
m
s
.
Als
o
,
th
e
ab
ilit
y
o
f
r
a
n
d
o
m
s
ea
r
ch
f
u
n
ctio
n
th
at
ex
is
ts
in
th
e
b
io
-
s
ea
r
c
h
alg
o
r
ith
m
s
g
iv
es
m
o
r
e
ad
v
an
tag
es
to
s
elec
t
th
e
b
est
o
p
tim
u
m
f
ea
tu
r
es.
Fo
r
r
e
d
u
cti
o
n
p
u
r
p
o
s
es,
it
ca
n
b
e
c
o
n
clu
d
ed
th
at
b
io
-
s
ea
r
c
h
alg
o
r
ith
m
s
with
wr
ap
p
e
r
m
e
th
o
d
ca
n
b
e
u
s
ed
to
r
ed
u
ce
attr
ib
u
tes f
r
o
m
all
s
izes o
f
d
ata.
T
ab
le
3
.
C
o
m
p
a
r
is
o
n
o
f
attr
ib
u
te
r
ed
u
ctio
n
: E
NORA v
s
E
NORA +
B
io
-
Sear
ch
D
a
t
a
s
e
t
#
A
t
t
r
Ori
S
u
b
s
e
t
G
e
n
e
r
a
t
i
o
n
(
1
)
S
u
b
s
e
t
G
e
n
e
r
a
t
i
o
n
(
2
)
#
A
t
t
r
#
A
t
t
r
[
EN
O
R
A
+
(
W
r
a
p
p
e
r
+
B
i
o
S
e
a
r
c
h
)
]
#
A
t
t
r
[
N
S
G
A
-
I
I
+
(
W
r
a
p
p
e
r
+
B
i
o
S
e
a
r
c
h
)
]
EN
O
R
A
+
F
i
l
t
e
r
e
d
N
S
G
A
-
I
I
+
F
i
l
t
e
r
e
d
A
n
t
Bat
Bee
C
u
c
F
l
y
A
n
t
Bat
Bee
C
u
c
F
l
y
B
r
e
a
s
t
c
a
n
c
e
r
9
9
(
0
.
0
%)*
9
(
0
.
0
%)*
7
7
7
6
7
7
7
7
6
7
P
a
r
k
i
n
s
o
n
22
9
(
5
9
.
1
%)*
7
(
6
8
.
2
%)*
5
6
6
7
6
5
3
5
5
5
O
z
o
n
e
72
1
2
(
8
5
.
7
%)*
6
(
9
1
.
7
%)*
1
1
1
1
1
1
1
1
1
1
C
l
e
a
n
1
1
6
6
2
2
(
8
6
.
7
%)*
1
9
(
8
8
.
6
%)*
14
13
14
14
14
15
17
15
15
14
S
e
mei
o
n
2
6
5
5
(
9
8
.
1
%)*
7
(
9
7
.
4
%)*
4
4
4
4
4
4
6
6
6
4
Emai
l
s
4
7
0
2
7
9
(
9
8
.
3
%)*
4
0
(
9
9
.
1
%)*
18
24
11
13
34
8
11
4
7
14
G
i
set
t
e
5
0
0
0
6
6
(
9
8
.
7
%)*
4
9
(
9
9
.
0
%)*
23
28
18
15
31
14
20
13
18
20
A
r
c
e
n
e
1
0
0
0
0
3
9
1
(
9
6
.
1
%)*
2
1
6
(
9
7
.
8
%)*
1
0
1
37
36
37
1
3
3
93
86
56
80
84
*
%
o
f
r
e
d
u
c
t
i
o
n
f
r
o
m
o
r
i
g
i
n
a
l
a
t
t
r
i
b
u
t
e
s
T
ab
le
s
u
b
s
et
g
en
er
atio
n
(
2
)
s
h
o
ws
th
e
co
m
p
ar
is
o
n
o
f
class
if
icatio
n
ac
cu
r
ac
y
o
f
E
NORA
wi
th
v
ar
io
u
s
class
if
ier
s
f
o
r
class
if
icatio
n
p
e
r
f
o
m
an
ce
.
Su
r
p
r
is
in
g
ly
th
at
at
tr
ib
u
tes
s
elec
ted
f
r
o
m
all
d
ata
s
ets
b
y
E
NORA
in
th
e
f
ir
s
t r
e
d
u
ctio
n
d
o
es n
o
t
im
p
r
o
v
e
th
e
class
if
icatio
n
ac
cu
r
a
cy
wh
ich
m
ai
n
tain
e
d
th
e
s
am
e
ac
cu
r
ac
y
r
esu
lts
o
f
th
e
o
r
ig
in
al
d
atasets
(
r
ef
er
to
T
ab
le
4
)
.
C
lear
ly
,
attr
ib
u
tes
s
elec
ted
in
s
ec
o
n
d
s
u
b
s
et
g
en
er
a
tio
n
b
y
E
NORA
an
d
b
io
-
s
ea
r
ch
alg
o
r
ith
m
s
with
wr
ap
p
er
m
eth
o
d
s
u
cc
ess
f
u
lly
in
cr
ea
s
ed
th
e
class
if
icatio
n
ac
cu
r
ac
y
.
All
f
iv
e
(
5
)
alg
o
r
ith
m
s
(
a
n
t,
b
at,
b
ee
,
cu
ck
o
o
an
d
f
i
r
ef
ly
)
p
r
o
v
ed
to
h
av
e
g
o
o
d
class
if
icatio
n
ac
cu
r
ac
y
f
o
r
all
d
atasets
ex
ce
p
t
Gis
ette
d
ataset.
B
u
t it
i
s
s
till
c
o
n
s
id
er
ed
ac
ce
p
tab
le
s
in
ce
th
e
p
er
ce
n
tag
e
o
f
r
ed
u
ctio
n
ac
h
ie
v
ed
m
o
r
e
th
an
5
0
%
(
r
ef
er
t
o
T
a
b
le
4
)
th
en
s
till
m
ain
tain
in
g
g
o
o
d
class
if
icatio
n
ac
cu
r
ac
y
f
o
r
Gis
ette
d
ataset.
Gen
er
ally
,
it
ca
n
b
e
s
ee
n
all
b
io
-
s
ea
r
ch
alg
o
r
ith
m
p
er
f
o
r
m
ed
well
to
ac
h
iev
e
b
etter
class
if
icatio
n
ac
cu
r
ac
y
with
v
ar
io
u
s
class
if
ier
s
.
T
h
e
h
ig
h
lig
h
ted
c
o
lu
m
n
i
n
T
a
b
le
5
s
h
o
ws
th
e
s
elec
ted
b
est
p
er
f
o
r
m
a
n
ce
o
f
class
if
icatio
n
r
esu
lts
wh
ich
r
ef
lects
th
e
m
o
d
el
to
b
e
d
e
v
elo
p
e
d
(
r
ef
er
to
T
ab
le
6
).
T
ab
le
7
s
h
o
ws
th
e
co
m
p
ar
is
o
n
o
f
class
if
icatio
n
ac
cu
r
ac
y
o
f
NSGA
-
I
I
with
v
ar
io
u
s
class
if
ier
s
f
o
r
class
if
icatio
n
p
er
f
o
m
an
ce
.
I
n
t
er
esti
n
g
ly
to
h
ig
h
li
g
h
t
th
at
att
r
ib
u
tes
s
elec
ted
f
r
o
m
all
d
atas
ets
b
y
NSGA
-
I
I
in
th
e
f
ir
s
t
s
u
b
s
et
g
en
er
atio
n
s
h
o
w
in
co
n
s
is
ten
t
r
esu
lts
wh
ich
i
m
p
r
o
v
e
d
t
h
e
ac
c
u
r
ac
y
f
o
r
th
e
h
alf
o
f
th
e
d
atasets
(
r
ef
er
to
T
ab
le
4
)
.
An
o
th
er
h
alf
s
h
o
ws
d
ec
r
em
en
t
o
f
class
if
icatio
n
ac
cu
r
ac
y
.
Ob
v
io
u
s
ly
,
th
e
f
ir
s
t
s
u
b
s
et
g
en
er
atio
n
r
esu
lts
b
y
NSGA
-
I
I
alg
o
r
ith
m
n
ee
d
t
o
b
e
o
p
tim
i
ze
d
in
o
r
d
er
to
g
et
b
etter
clas
s
if
icatio
n
ac
cu
r
ac
y
.
I
n
s
ec
o
n
d
s
u
b
s
et
g
e
n
er
atio
n
,
NSGA
-
I
I
an
d
b
io
-
s
ea
r
ch
alg
o
r
ith
m
s
with
wr
ap
p
er
m
eth
o
d
s
h
o
ws
s
ig
n
if
ican
t
in
cr
em
en
t
f
o
r
all
d
atasets
.
T
h
e
h
ig
h
lig
h
ted
co
l
u
m
n
in
T
a
b
le
5
s
h
o
ws
th
e
s
e
lecte
d
b
est
p
er
f
o
r
m
an
ce
o
f
class
if
icatio
n
r
esu
lts
wh
ich
r
ef
lects
th
e
m
o
d
el
to
b
e
d
e
v
elo
p
ed
(
r
e
f
er
to
T
ab
le
6
)
.
T
ab
le
6
s
h
o
ws
th
e
I
DE
AL
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
KA
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l Co
n
tr
o
l
I
n
teg
r
a
ted
b
i
o
-
s
ea
r
ch
a
p
p
r
o
a
c
h
es w
ith
mu
lti
-
o
b
jective
a
lg
o
r
ith
ms fo
r
… (
Mo
h
a
mma
d
A
iz
a
t
B
a
s
ir
)
2429
f
ea
tu
r
e
s
elec
tio
n
m
o
d
el
o
n
v
ar
i
o
u
s
s
izes
o
f
d
ataset
s
.
T
h
is
m
o
d
el
wh
ich
co
n
s
is
t
o
f
c
o
m
b
in
atio
n
lis
t
o
f
alg
o
r
ith
m
s
ca
n
b
e
a
g
u
id
elin
e
f
o
r
s
ea
r
ch
i
n
g
o
p
tim
al
n
u
m
b
er
o
f
attr
ib
u
tes b
ased
o
n
d
ataset
s
ize.
T
ab
le
4
.
Su
b
s
et
g
e
n
er
atio
n
(
1
)
class
if
icatio
n
ac
cu
r
ac
y
:
o
r
ig
i
n
al
d
ata,
E
NORA,
NSG
A
-
I
I
u
s
in
g
DT
,
NB
an
d
k
-
NN
D
a
t
a
s
e
t
N
o
R
e
d
u
c
t
i
o
n
1
st
R
e
d
u
c
t
i
o
n
1
st
R
e
d
u
c
t
i
o
n
A
c
c
u
r
a
c
y
(
%)
[
EN
O
R
A
+
F
i
l
t
e
r
e
d
]
A
c
c
u
r
a
c
y
(
%)
[NSGA
-
I
I
+
F
i
l
t
e
r
e
d
]
A
c
c
u
r
a
c
y
(
%)
DT
NB
k
-
NN
DT
NB
k
-
NN
DT
NB
k
-
NN
B
r
e
a
s
t
c
a
n
c
e
r
9
6
.
2
9
6
.
2
9
5
.
8
9
6
.
2
9
6
.
2
9
5
.
4
9
6
.
2
9
6
.
2
9
5
.
4
P
a
r
k
i
n
s
o
n
8
4
.
8
8
3
.
3
9
2
.
4
8
9
.
4
9
0
.
9
9
2
.
4
8
7
.
9
9
0
.
9
8
7
.
9
O
z
o
n
e
9
3
.
3
7
1
.
9
9
2
.
4
9
3
.
7
8
0
.
5
9
2
.
2
9
3
.
9
8
9
.
4
9
3
.
5
C
l
e
a
n
1
8
5
.
8
8
5
.
2
8
3
.
3
7
5
.
9
8
2
.
1
8
0
.
2
8
2
.
1
8
6
.
4
8
6
.
4
S
e
mei
o
n
9
4
.
5
9
3
.
0
9
7
.
6
9
2
.
4
9
1
.
3
9
2
.
4
9
3
.
4
9
0
.
6
9
3
.
4
Emai
l
s
7
2
.
7
8
6
.
4
7
2
.
7
7
7
.
3
8
6
.
4
8
6
.
4
7
7
.
3
7
7
.
3
7
7
.
3
G
i
set
t
e
9
1
.
5
9
1
.
5
9
2
.
9
8
8
.
2
8
5
.
0
8
5
.
6
8
6
.
8
8
6
.
2
8
5
.
9
A
r
c
e
n
e
7
0
.
6
7
6
.
5
9
1
.
2
8
5
.
3
7
9
.
4
8
8
.
2
8
5
.
3
7
9
.
4
8
8
.
2
T
ab
le
5
.
Su
b
s
et
g
e
n
er
atio
n
(
2
)
class
if
icatio
n
ac
cu
r
ac
y
: E
NORA u
s
in
g
DT
,
NB
an
d
k
-
NN
D
a
t
a
s
e
t
2
nd
R
e
d
u
c
t
i
o
n
[
EN
O
R
A
+
(
W
r
a
p
p
e
r
+
B
i
o
S
e
a
r
c
h
)
]
A
c
c
u
r
a
c
y
(
%)
D
e
c
i
s
i
o
n
Tr
e
e
(
D
T)
N
a
ï
v
e
B
a
y
e
s
(
N
B
)
k
-
N
e
a
r
e
s
t
N
e
i
g
h
b
o
u
r
(
k
-
NN)
A
n
t
Bat
Bee
C
u
c
F
l
y
A
n
t
Bat
Bee
C
u
c
F
l
y
A
n
t
Bat
Bee
C
u
c
F
l
y
B
r
e
a
s
t
c
a
n
c
e
r
9
6
.
2
9
6
.
2
9
6
.
2
9
6
.
2
9
6
.
2
9
6
.
6
9
6
.
6
9
6
.
6
9
6
.
6
9
6
.
6
9
5
.
4
9
5
.
4
9
5
.
4
9
6
.
2
9
5
.
4
P
a
r
k
i
n
s
o
n
8
7
.
9
8
9
.
4
8
9
.
4
8
9
.
4
8
9
.
4
9
0
.
9
9
0
.
9
9
0
.
9
9
3
.
9
9
0
.
9
9
0
.
9
9
2
.
4
9
2
.
4
9
2
.
4
9
2
.
4
O
z
o
n
e
9
3
.
9
9
3
.
9
9
3
.
9
9
3
.
9
9
3
.
9
9
3
.
9
9
3
.
9
9
3
.
9
9
3
.
9
9
3
.
9
9
3
.
9
9
3
.
9
9
3
.
9
9
3
.
9
9
3
.
9
C
l
e
a
n
1
8
0
.
9
8
0
.
9
8
0
.
9
8
0
.
9
8
0
.
9
8
4
.
0
8
4
.
0
8
3
.
3
8
4
.
0
8
4
.
0
8
2
.
1
8
2
.
7
8
2
.
7
8
2
.
1
8
2
.
1
S
e
mei
o
n
9
2
.
4
9
2
.
4
9
2
.
4
9
2
.
4
9
2
.
4
9
1
.
3
9
1
.
3
9
1
.
3
9
1
.
3
9
1
.
3
9
2
.
4
9
2
.
4
9
2
.
4
9
2
.
4
9
2
.
4
Emai
l
s
7
7
.
3
7
7
.
3
7
7
.
3
7
2
.
7
7
7
.
3
8
6
.
4
8
6
.
4
8
6
.
4
8
6
.
4
8
6
.
4
8
6
.
4
8
6
.
4
8
6
.
4
8
6
.
4
8
6
.
4
G
i
set
t
e
8
4
.
4
8
7
.
1
8
6
.
2
8
3
.
5
8
7
.
6
8
5
.
3
8
5
.
3
8
6
.
2
8
1
.
8
8
5
.
9
8
3
.
2
8
6
.
2
8
6
.
5
8
2
.
9
8
6
.
8
A
r
c
e
n
e
8
8
.
2
8
8
.
2
7
6
.
5
8
8
.
2
9
4
.
1
8
8
.
2
8
8
.
2
9
1
.
2
8
8
.
2
8
8
.
2
8
5
.
3
8
5
.
3
8
2
.
4
8
8
.
2
9
1
.
2
T
ab
le
6
.
I
d
ea
l f
ea
tu
r
e
s
elec
tio
n
m
o
d
el
Li
st
D
a
t
a
s
e
t
s
i
z
e
M
u
l
t
i
-
o
b
j
e
c
t
i
v
e
a
l
g
o
R
e
d
u
c
t
i
o
n
a
l
g
o
B
i
o
-
sea
r
c
h
a
l
g
o
C
l
a
s
si
f
i
e
r
1.
S
mal
l
EN
O
R
A
W
r
a
p
p
e
r
C
u
c
k
o
o
NB
2.
N
S
G
A
-
II
A
n
t
,
C
u
c
k
o
o
,
F
i
r
e
f
l
y
NB
3.
M
e
d
i
u
m
EN
O
R
A
B
e
e
,
B
a
t
k
-
NN
4.
N
S
G
A
-
II
A
n
t
,
C
u
c
k
o
o
NB
5.
La
r
g
e
EN
O
R
A
F
i
r
e
f
l
y
k
-
NN
6.
N
S
G
A
-
II
Bat
DT
T
ab
le
7
.
Su
b
s
et
g
e
n
er
atio
n
(
2
)
class
if
icatio
n
ac
cu
r
ac
y
: N
SGA
-
I
I
u
s
in
g
DT
,
NB
an
d
k
-
NN
D
a
t
a
s
e
t
2
nd
R
e
d
u
c
t
i
o
n
[NSGA
-
I
I
+
(
W
r
a
p
p
e
r
+
B
i
o
S
e
a
r
c
h
)
]
A
c
c
u
r
a
c
y
(
%)
D
e
c
i
s
i
o
n
Tr
e
e
(
D
T)
N
a
ï
v
e
B
a
y
e
s
(
N
B
)
k
-
N
e
a
r
e
s
t
N
e
i
g
h
b
o
u
r
(
k
-
NN)
A
n
t
Bat
Bee
C
u
c
F
l
y
A
n
t
Bat
Bee
C
u
c
F
l
y
A
n
t
Bat
Bee
C
u
c
F
l
y
B
r
e
a
s
t
c
a
n
c
e
r
9
6
.
2
9
6
.
2
9
6
.
2
9
6
.
2
9
6
.
2
9
6
.
6
9
6
.
6
9
6
.
6
9
6
.
6
9
6
.
6
9
5
.
4
9
5
.
4
9
5
.
4
9
6
.
2
9
5
.
4
P
a
r
k
i
n
s
o
n
8
7
.
9
8
4
.
8
8
7
.
9
8
7
.
9
8
7
.
9
9
3
.
9
8
6
.
4
9
0
.
9
9
3
.
9
9
3
.
9
9
0
.
9
8
6
.
3
9
0
.
9
9
0
.
9
9
0
.
9
O
z
o
n
e
9
3
.
9
9
3
.
9
9
3
.
9
9
3
.
9
9
3
.
9
9
3
.
9
9
3
.
9
9
3
.
9
9
3
.
9
9
3
.
9
9
3
.
9
9
3
.
9
9
3
.
9
9
3
.
9
9
3
.
9
C
l
e
a
n
1
8
4
.
0
8
3
.
3
8
3
.
3
8
3
.
3
8
3
.
3
8
7
.
0
8
4
.
6
8
5
.
8
8
7
.
0
8
4
.
6
8
5
.
2
8
2
.
1
8
4
.
6
8
1
.
5
8
4
.
0
S
e
mei
o
n
9
2
.
6
9
3
.
4
9
3
.
4
9
3
.
4
9
2
.
6
9
3
.
2
9
3
.
2
9
3
.
2
9
3
.
2
9
3
.
2
9
3
.
2
9
3
.
2
9
3
.
2
9
3
.
2
9
3
.
2
Emai
l
s
7
7
.
3
7
7
.
3
7
7
.
3
7
7
.
3
7
7
.
3
7
7
.
3
7
7
.
3
7
7
.
3
7
7
.
3
7
7
.
3
7
7
.
3
7
7
.
3
7
7
.
3
7
7
.
3
7
7
.
3
G
i
set
t
e
8
8
.
2
8
8
.
2
8
8
.
2
8
8
.
8
8
8
.
2
8
6
.
5
8
6
.
2
8
6
.
5
8
6
.
5
8
6
.
5
8
5
.
9
8
5
.
3
8
5
.
9
8
5
.
3
8
5
.
9
A
r
c
e
n
e
8
5
.
3
9
1
.
2
7
6
.
5
7
6
.
5
8
2
.
4
8
5
.
3
8
2
.
4
8
8
.
2
8
5
.
3
8
5
.
3
8
5
.
3
8
8
.
2
8
8
.
2
9
1
.
2
9
1
.
2
4.
CO
NCLU
SI
O
N
AND
F
U
T
U
RE
WO
RK
I
n
s
u
m
m
ar
y
,
th
e
im
p
ac
t
o
f
th
is
p
ap
er
o
n
d
ata
m
in
in
g
ca
n
b
e
s
ee
n
as
lead
in
g
in
p
ar
ticu
la
r
to
alter
n
ativ
e
o
p
tim
izatio
n
tech
n
iq
u
es.
T
h
is
alter
n
ativ
e
tech
n
iq
u
e
p
r
o
v
id
es
a
b
etter
u
n
d
er
s
tan
d
in
g
o
f
th
e
im
p
lem
en
tatio
n
o
f
v
ar
io
u
s
b
i
o
-
in
s
p
ir
ed
alg
o
r
ith
m
s
in
th
e
e
x
p
l
o
r
atio
n
a
n
d
u
tili
za
tio
n
o
f
th
e
s
ea
r
ch
s
p
ac
e,
in
p
ar
ticu
lar
f
o
r
th
e
o
p
tim
izatio
n
o
f
m
u
lti
-
o
b
je
ctiv
e
alg
o
r
ith
m
s
.
T
h
is
p
ap
er
e
x
p
lo
r
es
a
n
ew
id
ea
l
f
ea
tu
r
e
s
elec
tio
n
m
o
d
el
th
a
t
h
as
b
ee
n
c
o
m
p
ar
ed
a
n
d
e
v
alu
ated
o
n
eig
h
t
(
8
)
d
atasets
.
T
h
e
id
ea
l
lis
ts
f
o
r
th
e
s
elec
tio
n
o
f
f
ea
tu
r
es
h
a
v
e
b
e
en
d
eter
m
in
ed
o
n
th
e
b
asis
o
f
th
e
p
r
o
d
u
ce
d
g
o
o
d
class
if
icatio
n
ac
cu
r
ac
y
with
th
e
r
elev
an
t
f
ea
tu
r
es.
Ho
wev
er
,
th
e
lim
itatio
n
o
f
alg
o
r
ith
m
s
n
ee
d
s
to
b
e
ad
d
r
ess
ed
.
On
e
o
f
th
e
lim
itatio
n
s
is
th
e
co
m
p
u
tatio
n
al
co
s
t
(
lo
n
g
er
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6
9
3
0
T
E
L
KOM
NI
KA
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l Co
n
tr
o
l
,
Vo
l.
18
,
No
.
5
,
Octo
b
e
r
2
0
2
0
:
2
4
2
1
-
2431
2430
co
m
p
u
tat
io
n
tim
e)
,
an
d
it
tak
es
tim
e
to
d
is
co
v
er
th
e
f
o
r
m
u
latio
n
o
f
th
e
lis
t.
T
h
e
n
ex
t
r
esear
ch
wo
r
k
to
b
e
ex
p
lo
r
ed
wo
u
ld
th
er
ef
o
r
e
b
e
t
h
e
s
tu
d
y
o
n
d
if
f
e
r
en
t
b
io
-
s
ea
r
ch
alg
o
r
ith
m
s
an
d
th
e
f
o
r
m
u
l
atio
n
o
f
th
e
co
r
r
ec
t
s
ettin
g
o
f
p
ar
am
ete
r
s
f
o
r
n
ew
o
p
tim
izatio
n
tech
n
i
q
u
es.
ACK
NO
WL
E
DG
E
M
E
NT
S
T
h
e
au
th
o
r
s
wo
u
ld
li
k
e
to
r
ec
o
g
n
ize
Un
iv
e
r
s
iti
Ma
lay
s
ia
T
er
en
g
g
an
u
(
UM
T
)
,
Un
i
v
er
s
iti
Utar
a
Ma
lay
s
ia
(
UUM
)
an
d
Kem
en
ter
ian
Pen
d
id
i
k
an
Ma
la
y
s
ia
(
KPM)
f
o
r
th
e
s
u
p
p
o
r
t
o
f
s
er
v
ices
an
d
f
ac
ilit
es.
T
h
is
r
esear
ch
was
s
p
o
n
s
o
r
ed
b
y
th
e
UM
T
an
d
Min
is
tr
y
o
f
E
d
u
ca
tio
n
Ma
lay
s
ia
(
MO
E
)
u
n
d
er
T
APE
-
R
G
r
esear
ch
g
r
a
n
t (
Vo
t N
o
.
5
5
1
3
3
)
.
RE
F
E
R
E
NC
E
S
[1
]
R.
Je
n
se
n
a
n
d
Q.
S
h
e
n
,
"
C
o
m
p
u
tatio
n
a
l
In
tell
ig
e
n
c
e
a
n
d
F
e
a
tu
re
S
e
lec
ti
o
n
:
Ro
u
g
h
a
n
d
F
u
z
z
y
Ap
p
ro
a
c
h
e
s
,"
W
il
e
y
-
IEE
E
Pre
ss
,
2
0
0
8
.
[
2
]
H
.
L
i
u
a
n
d
H
.
M
o
t
o
d
a
,
"
F
e
a
t
u
r
e
S
e
l
e
c
t
i
o
n
f
o
r
K
n
o
w
l
e
d
g
e
D
i
s
c
o
v
e
r
y
a
n
d
D
a
t
a
M
i
n
i
n
g
,"
K
l
u
w
e
r
A
c
a
d
e
m
i
c
P
r
e
s
s
,
1998.
[3
]
R.
Ca
ru
a
n
a
a
n
d
D.
F
re
it
a
g
,
“
G
re
e
d
y
Attr
ib
u
te S
e
lec
ti
o
n
,
”
M
a
c
h
i
n
e
L
e
a
rn
in
g
Pr
o
c
e
e
d
in
g
s
,
1
9
9
4
.
[4
]
A.
Ara
u
z
o
-
Az
o
fra
,
J.
M
.
Be
n
it
e
z
,
a
n
d
J.
L.
Ca
stro
,
“
Co
n
siste
n
c
y
m
e
a
su
re
s
fo
r
fe
a
tu
re
se
lec
ti
o
n
,
”
J
.
I
n
tell.
In
f.
S
y
st.
,
v
o
l.
3
0
,
p
p
.
2
7
3
-
2
9
2
,
2
0
0
8
.
[5
]
X.
Zh
a
n
g
,
Y.
H
u
,
K.
Xie
,
S
.
Wan
g
,
E.
W
.
T.
N
g
a
i,
a
n
d
M
.
Li
u
,
“
A
c
a
u
sa
l
fe
a
tu
re
se
lec
ti
o
n
a
lg
o
r
it
h
m
fo
r
st
o
c
k
p
re
d
ictio
n
m
o
d
e
li
n
g
,
”
Ne
u
r
o
c
o
m
p
u
ti
n
g
,
v
o
l.
1
4
2
,
p
p
.
4
8
-
5
9
,
2
0
1
4
.
[6
]
B.
A.
Bles
se
r,
T.
T.
Ku
k
l
in
sk
i,
a
n
d
R.
J.
S
h
il
lma
n
,
“
Emp
iri
c
a
l
tes
ts
fo
r
fe
a
tu
re
se
lec
ti
o
n
b
a
se
d
o
n
a
p
sy
c
h
o
l
o
g
ica
l
th
e
o
ry
o
f
c
h
a
ra
c
ter rec
o
g
n
it
i
o
n
,
”
Pa
tt
e
rn
Rec
o
g
n
it
io
n
,
v
o
l
.
8
,
n
o
.
2
,
p
p
.
7
7
-
8
5
,
1
9
7
6
.
[7
]
J.
Tan
g
a
n
d
H.
Li
u
,
“
F
e
a
tu
re
se
l
e
c
ti
o
n
fo
r
so
c
ial
m
e
d
ia
d
a
ta,”
A
CM
T
ra
n
sa
c
ti
o
n
s
o
n
K
n
o
wle
d
g
e
Disc
o
v
e
ry
fro
m
Da
ta
,
v
o
l.
8
,
n
o
.
4
,
p
p
.
1
-
2
7
,
2
0
1
4
.
[8
]
C.
H.
Ch
e
n
,
“
On
in
f
o
rm
a
ti
o
n
a
n
d
d
istan
c
e
m
e
a
su
re
s,
e
rro
r
b
o
u
n
d
s,
a
n
d
fe
a
t
u
re
se
lec
ti
o
n
,
”
I
n
f
o
rm
a
ti
o
n
S
c
ien
c
e
s
,
v
o
l.
1
0
,
n
o
.
2
,
p
p
.
1
5
9
-
1
7
3
,
1
9
7
6
.
[9
]
L.
S
u
n
,
J.
X
u
,
a
n
d
Y.
Ti
a
n
,
“
F
e
a
tu
r
e
se
lec
ti
o
n
u
si
n
g
ro
u
g
h
e
n
tro
p
y
-
b
a
se
d
u
n
c
e
rtain
ty
m
e
a
su
re
s
in
i
n
c
o
m
p
lete
d
e
c
isio
n
sy
ste
m
s,”
Kn
o
wled
g
e
-
B
a
se
d
S
y
st.
,
v
o
l.
3
6
,
p
p
.
2
0
6
-
2
1
6
,
2
0
1
2
.
[1
0
]
S
.
K.
d
a
s
S
u
b
ra
ta,
“
F
e
a
tu
re
S
e
lec
ti
o
n
with
a
Li
n
e
a
r
De
p
e
n
d
e
n
c
e
M
e
a
su
re
,
”
IEE
E
T
ra
n
sa
c
ti
o
n
s
o
n
Co
m
p
u
ter
s
,
v
o
l.
C
-
2
0
,
n
o
.
9
,
p
p
.
1
1
0
6
-
1
1
0
9,
1
9
7
1
.
[1
1
]
R.
Ko
h
a
v
i
a
n
d
G
.
Jo
h
n
,
“
Wr
a
p
p
e
rs fo
r
fe
a
tu
re
su
b
se
t
se
lec
ti
o
n
,
”
Art
if
.
In
tel
l.
,
v
o
l
.
9
7
,
n
o
.
1
,
p
p
.
2
7
3
-
3
2
4
,
1
9
9
7
.
[1
2
]
V.
Ku
m
a
r,
“
F
e
a
tu
re
S
e
lec
ti
o
n
:
A
li
tera
tu
re
Re
v
iew
,
”
S
ma
rt
Co
mp
u
t.
Rev
.
,
v
o
l.
4
,
n
o
.
3
,
p
p
.
2
1
1
-
2
2
9
,
2
0
1
4
.
[
1
3
]
K
.
D
e
b
,
“
M
u
l
t
i
-
o
b
j
e
c
t
i
v
e
O
p
t
i
m
i
s
a
t
i
o
n
U
s
i
n
g
E
v
o
l
u
t
i
o
n
a
r
y
A
l
g
o
r
i
t
h
m
s
:
A
n
I
n
t
r
o
d
u
c
t
i
o
n
,
”
J
h
o
n
W
i
l
e
y
a
n
d
S
o
n
s
L
t
d
,
2
0
1
1
.
[1
4
]
Co
e
ll
o
C
o
e
ll
o
,
Ca
rl
o
s,
Lam
o
n
t,
G
a
ry
B.
,
v
a
n
Ve
ld
h
u
ize
n
,
Da
v
id
A.
,
"
E
v
o
l
u
ti
o
n
a
ry
Al
g
o
r
it
h
m
s
fo
r
S
o
l
v
in
g
M
u
lt
i
-
O
b
jec
ti
v
e
P
r
o
b
lem
s
,"
Ge
n
e
ti
c
a
n
d
Evo
lu
ti
o
n
a
ry
Co
mp
u
t
a
ti
o
n
,
S
p
rin
g
e
r,
2
0
0
7
.
[1
5
]
P
.
M
.
Na
re
n
d
ra
a
n
d
K.
F
u
k
u
n
a
g
a
,
“
A
Bra
n
c
h
a
n
d
B
o
u
n
d
Alg
o
rit
h
m
f
o
r
F
e
a
tu
re
S
u
b
se
t
S
e
lec
ti
o
n
,
”
IE
E
E
T
ra
n
sa
c
ti
o
n
s
o
n
C
o
mp
u
ter
s
,
v
o
l.
C
-
2
6
,
n
o
.
9
,
p
p
.
9
1
7
-
9
2
2
,
1
9
7
7
.
[1
6
]
P
.
G
u
p
ta,
D.
Do
e
rm
a
n
n
,
a
n
d
D.
De
M
e
n
th
o
n
,
“
Be
a
m
se
a
rc
h
fo
r
fe
a
tu
re
se
lec
ti
o
n
i
n
a
u
t
o
m
a
ti
c
S
VM
d
e
fe
c
t
c
las
sifica
ti
o
n
,
”
Pro
c
.
-
I
n
t.
Co
n
f.
Pa
tt
e
rn
Rec
o
g
n
it
.
,
2
0
0
2
.
[1
7
]
H.
Va
fa
ie
a
n
d
K.
De
Jo
n
g
,
“
G
e
n
e
ti
c
a
lg
o
rit
h
m
s
a
s
a
to
o
l
f
o
r
fe
a
tu
re
se
lec
ti
o
n
in
m
a
c
h
in
e
lea
rn
i
n
g
,
”
Pro
c
e
e
d
in
g
s
In
ter
n
a
t
io
n
a
l
C
o
n
fer
e
n
c
e
o
n
T
o
o
l
s wit
h
Arti
fi
c
ia
l
In
telli
g
e
n
c
e
,
ICT
AI
,
1
9
9
2
.
[
1
8
]
J
.
L
a
n
g
f
o
r
d
e
t
a
l
.
,
“
E
v
o
l
u
t
i
o
n
a
r
y
F
e
a
t
u
r
e
S
e
l
e
c
t
i
o
n
,
”
E
n
c
y
c
l
o
p
e
d
i
a
o
f
M
a
c
h
i
n
e
L
e
a
r
n
i
n
g
,
S
p
r
i
n
g
e
r
U
S
,
p
p
.
3
5
3
-
353
,
2
0
1
1
.
[1
9
]
L.
Ce
rv
a
n
te,
B
.
X
u
e
,
M
.
Zh
a
n
g
,
a
n
d
L.
S
h
a
n
g
,
“
Bin
a
r
y
p
a
rti
c
le
s
wa
rm
o
p
ti
m
isa
ti
o
n
fo
r
fe
a
tu
re
se
l
e
c
ti
o
n
:
A
fi
lt
e
r
b
a
se
d
a
p
p
r
o
a
c
h
,
”
IEE
E
Co
n
g
re
ss
o
n
Ev
o
lu
t
io
n
a
ry
Co
m
p
u
t
a
ti
o
n
,
2
0
1
2
.
[2
0
]
Z.
Yo
n
g
,
G
.
Du
n
-
we
i,
a
n
d
Z.
Wa
n
-
q
i
u
,
“
F
e
a
tu
re
se
lec
ti
o
n
o
f
u
n
re
li
a
b
le
d
a
ta
u
s
in
g
a
n
imp
r
o
v
e
d
m
u
lt
i
-
o
b
jec
ti
v
e
P
S
O
a
lg
o
rit
h
m
,
”
Ne
u
ro
c
o
m
p
u
ti
n
g
,
v
o
l.
1
7
1
,
p
p
.
1
2
8
1
-
12
9
0
,
2
0
1
6
.
[2
1
]
A.
Al
-
An
i
a
n
d
M
.
De
rich
e
,
“
F
e
a
tu
re
se
lec
ti
o
n
u
si
n
g
a
m
u
tu
a
l
i
n
fo
rm
a
ti
o
n
b
a
se
d
m
e
a
su
re
,
”
O
b
j
e
c
t
re
c
o
g
n
it
io
n
su
p
p
o
rte
d
b
y
u
se
r in
ter
a
c
ti
o
n
f
o
r
se
rv
ice
ro
b
o
ts
,
v
o
l.
4
,
p
p
.
8
2
-
85
,
2
0
0
2
.
[2
2
]
F
.
Jim
é
n
e
z
,
G
.
S
á
n
c
h
e
z
,
J.
M
.
G
a
rc
ía,
G
.
S
c
iav
icc
o
,
a
n
d
L.
M
iralles
,
“
M
u
lt
i
-
o
b
jec
ti
v
e
e
v
o
lu
ti
o
n
a
ry
fe
a
tu
re
se
lec
ti
o
n
fo
r
o
n
li
n
e
sa
les
fo
re
c
a
stin
g
,
”
Ne
u
ro
c
o
mp
u
ti
n
g
,
v
o
l
.
2
3
4
,
p
p
.
7
2
-
9
2
,
2
0
1
7
.
[2
3
]
H.
F
a
lag
h
i,
M
.
R.
Ha
g
h
ifam
,
a
n
d
C.
S
i
n
g
h
,
“
An
t
c
o
l
o
n
y
o
p
t
imiz
a
ti
o
n
-
b
a
se
d
m
e
th
o
d
fo
r
p
lac
e
m
e
n
t
o
f
se
c
ti
o
n
a
li
z
i
n
g
sw
it
c
h
e
s in
d
istri
b
u
ti
o
n
n
e
two
r
k
s
u
sin
g
a
f
u
z
z
y
m
u
lt
i
o
b
jec
ti
v
e
a
p
p
r
o
a
c
h
,
”
IEE
E
T
r
a
n
s.
P
o
we
r De
li
v
.
,
v
o
l.
2
4
,
n
o
.
1
,
p
p
.
2
6
8
-
27
7
,
2
0
0
8
.
[2
4
]
Y.
Wan
,
M
.
Wan
g
,
Z.
Ye
,
a
n
d
X.
Lai,
“
A
fe
a
tu
re
s
e
lec
ti
o
n
m
e
th
o
d
b
a
se
d
o
n
m
o
d
ifi
e
d
b
in
a
ry
c
o
d
e
d
a
n
t
c
o
lo
n
y
o
p
ti
m
iza
ti
o
n
a
lg
o
rit
h
m
,
”
Ap
p
l.
S
o
ft
Co
mp
u
t.
J
.
,
v
o
l.
4
9
,
p
p
.
2
4
8
-
2
5
8
,
2
0
1
6
.
[2
5
]
O.
F
a
ll
a
h
z
a
d
e
h
,
Z.
De
h
g
h
a
n
i
-
B
i
d
g
o
li
,
a
n
d
M
.
As
sa
rian
,
“
Ra
m
a
n
sp
e
c
tral
fe
a
t
u
re
se
lec
ti
o
n
u
si
n
g
a
n
t
c
o
l
o
n
y
o
p
ti
m
iza
ti
o
n
f
o
r
b
re
a
st ca
n
c
e
r
d
ia
g
n
o
sis
,
”
L
a
se
rs
M
e
d
.
S
c
i.
,
v
o
l.
3
3
3
,
n
o
.
8
,
p
p
.
1
7
9
9
-
1
8
0
6
,
2
0
1
8
.
[2
6
]
T.
M
e
h
m
o
d
a
n
d
H.
B.
M
.
Ra
is,
“
An
t
c
o
l
o
n
y
o
p
ti
m
iza
ti
o
n
a
n
d
fe
a
t
u
re
se
lec
ti
o
n
fo
r
i
n
tru
si
o
n
d
e
tec
ti
o
n
,
”
Ad
v
a
n
c
e
s in
M
a
c
h
i
n
e
L
e
a
r
n
in
g
a
n
d
S
i
g
n
a
l
Pro
c
e
ss
in
g
,
pp
.
3
0
5
-
3
1
2
,
2
0
1
6
.
[2
7
]
V.
Ag
ra
wa
l
a
n
d
S
.
Ch
a
n
d
ra
,
“
F
e
a
tu
re
se
lec
ti
o
n
u
sin
g
Artifi
c
i
a
l
Be
e
Co
lo
n
y
a
l
g
o
ri
th
m
fo
r
m
e
d
ica
l
ima
g
e
c
las
sifica
ti
o
n
,
”
2
0
1
5
8
t
h
I
n
ter
n
a
t
i
o
n
a
l
Co
n
fer
e
n
c
e
o
n
Co
n
tem
p
o
r
a
ry
Co
mp
u
t
in
g
,
2
0
1
5
.
[2
8
]
H.
Ra
o
e
t
a
l.
,
“
F
e
a
tu
re
se
lec
ti
o
n
b
a
se
d
o
n
a
rti
f
icia
l
b
e
e
c
o
lo
n
y
a
n
d
g
ra
d
ien
t
b
o
o
sti
n
g
d
e
c
isio
n
tr
e
e
,
”
Ap
p
l
.
S
o
ft
Co
mp
u
t
.
J
.
,
v
o
l.
7
4
,
p
p
.
3
4
-
4
2
,
2
0
1
9
.
Evaluation Warning : The document was created with Spire.PDF for Python.