I
AE
S In
t
er
na
t
io
na
l J
o
urna
l o
f
Art
if
icia
l In
t
ellig
ence
(
I
J
-
AI)
Vo
l.
9
,
No
.
3
,
Sep
tem
b
er
2020
,
p
p
.
402
~
408
I
SS
N:
2
2
5
2
-
8938
,
DOI
: 1
0
.
1
1
5
9
1
/i
j
ai.
v
9
.i
3
.
p
p
4
0
2
-
408
402
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
i
.
ia
esco
r
e.
co
m
Diff
e
rent
m
utatio
n and
cros
so
v
er s
et
o
f
g
en
etic
prog
ra
m
m
ing
in
a
n auto
m
a
te
d
m
a
chine learni
ng
Su
r
a
y
a
M
a
s
ro
m
1
,
M
a
s
ura
h M
o
ha
m
a
d
2
,
Sh
a
hira
h M
o
ha
m
e
d H
a
t
i
m
3
,
No
rha
y
a
t
i Ba
ha
run
4
,
Na
s
iro
h
O
m
a
r
5
,
Ab
du
lla
h Sa
n
i A
bd
.
Ra
h
m
a
n
6
1,
2,
3,
4
F
a
c
u
lt
y
o
f
Co
m
p
u
ter an
d
M
a
th
e
m
a
ti
c
a
l
S
c
ien
c
e
s,
Un
iv
e
rsiti
Tek
n
o
lo
g
i
M
A
RA
,
P
e
ra
k
Bra
n
c
h
T
a
p
a
h
Ca
m
p
u
s,
M
a
la
y
sia
5
F
a
c
u
lt
y
o
f
Co
m
p
u
ter an
d
M
a
th
e
m
a
ti
c
a
l
S
c
ien
c
e
s,
Un
iv
e
rsiti
T
e
k
n
o
lo
g
i
M
A
RA
,
S
h
a
h
A
la
m
,
M
a
la
y
sia
6
F
a
c
u
lt
y
o
f
S
c
ien
c
e
a
n
d
In
f
o
rm
a
ti
o
n
T
e
c
h
n
o
l
o
g
y
,
Un
iv
e
rsiti
T
e
k
n
o
lo
g
i
P
ET
RON
A
S
,
P
e
ra
k
,
M
a
la
y
sia
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
F
eb
1
2
,
2
0
2
0
R
ev
i
s
ed
A
pr
17
,
2
0
2
0
A
cc
ep
ted
M
a
y
1
0
,
2
0
2
0
A
u
to
m
a
ted
m
a
c
h
in
e
lea
rn
in
g
is
a
p
ro
m
isin
g
a
p
p
ro
a
c
h
w
id
e
ly
u
se
d
to
s
o
lv
e
c
las
si
f
ica
ti
o
n
a
n
d
p
re
d
icti
o
n
p
ro
b
lem
s,
w
h
ich
c
u
rre
n
tl
y
r
e
c
e
i
v
e
s
m
u
c
h
a
tt
e
n
ti
o
n
f
o
r
m
o
d
if
ica
ti
o
n
a
n
d
i
m
p
ro
v
e
m
e
n
t.
On
e
o
f
th
e
p
r
o
g
re
ss
in
g
w
o
rk
s
f
o
r
a
u
to
m
a
ted
m
a
c
h
in
e
lea
rn
in
g
i
m
p
ro
v
e
m
e
n
t
is t
h
e
in
c
lu
sio
n
o
f
e
v
o
lu
ti
o
n
a
ry
a
lg
o
rit
h
m
su
c
h
a
s
Ge
n
e
t
ic
P
ro
g
ra
m
m
in
g
.
T
h
e
f
u
n
c
ti
o
n
o
f
Ge
n
e
ti
c
P
r
o
g
ra
m
m
in
g
is
to
o
p
ti
m
ize
th
e
b
e
st
c
o
m
b
in
a
ti
o
n
o
f
so
lu
t
io
n
s
f
ro
m
th
e
p
o
ss
ib
le
p
i
p
e
li
n
e
s
o
f
m
a
c
h
in
e
lea
rn
in
g
m
o
d
e
ll
in
g
,
in
c
l
u
d
i
n
g
se
lec
ti
o
n
o
f
a
lg
o
rit
h
m
s
a
n
d
p
a
ra
m
e
t
e
rs
o
p
ti
m
iz
a
ti
o
n
o
f
th
e
se
lec
t
e
d
a
lg
o
rit
h
m
.
A
s
a
fa
m
il
y
o
f
e
v
o
lu
ti
o
n
a
ry
b
a
s
e
d
a
lg
o
rit
h
m
,
th
e
e
ff
e
c
ti
v
e
n
e
ss
o
f
G
e
n
e
ti
c
P
r
o
g
ra
m
m
in
g
in
p
ro
v
id
in
g
th
e
b
e
st
m
a
c
h
in
e
lea
rn
in
g
p
ip
e
li
n
e
s
f
o
r
a
g
iv
e
n
p
ro
b
lem
o
r
d
a
tas
e
t
is
su
b
sta
n
ti
a
ll
y
d
e
p
e
n
d
in
g
o
n
th
e
a
lg
o
rit
h
m
p
a
ra
m
e
teriz
a
ti
o
n
s
in
c
lu
d
in
g
t
h
e
m
u
tatio
n
a
n
d
c
ro
ss
o
v
e
r
ra
tes
.
T
h
is
p
a
p
e
r
p
re
se
n
ts
th
e
e
ff
e
c
t
o
f
d
iff
e
re
n
t
p
a
irs
o
f
m
u
tatio
n
a
n
d
c
ro
ss
o
v
e
r
ra
tes
o
n
th
e
a
u
to
m
a
ted
m
a
c
h
in
e
lea
rn
in
g
p
e
rfo
rm
a
n
c
e
s
th
a
t
tes
ted
o
n
d
if
f
e
r
e
n
t
ty
p
e
s
o
f
d
a
tas
e
ts.
T
h
e
f
in
d
in
g
c
a
n
b
e
u
se
d
to
s
u
p
p
o
rt
th
e
th
e
o
ry
th
a
t
h
ig
h
e
r
c
ro
ss
o
v
e
r
ra
tes
u
se
d
to
im
p
ro
v
e
th
e
a
lg
o
ri
th
m
a
c
c
u
ra
c
y
sc
o
re
w
h
il
e
lo
w
e
r
c
ro
ss
o
v
e
r
ra
tes
m
a
y
c
a
u
se
th
e
a
lg
o
rit
h
m
to
c
o
n
v
e
rg
e
a
t
e
a
rli
e
r
sta
g
e
.
K
ey
w
o
r
d
s
:
Au
to
m
a
ted
m
ac
h
in
e
lear
n
i
n
g
C
las
s
i
f
icatio
n
C
r
o
s
s
o
v
er
Gen
etic
p
r
o
g
r
a
m
m
i
n
g
Mu
tatio
n
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
Su
r
a
y
a
Ma
s
r
o
m
,
Facu
lt
y
o
f
C
o
m
p
u
ter
an
d
Ma
t
h
e
m
a
tical
Scie
n
ce
s
,
Un
i
v
er
s
iti T
ek
n
o
lo
g
i M
A
R
A
,
P
er
ak
B
r
an
ch
,
T
a
p
ah
C
a
m
p
u
s
,
P
er
ak
,
Ma
lay
s
ia
.
E
m
ail: s
u
r
a
y
0
7
8
@
u
it
m
.
ed
u
.
m
y
1.
I
NT
RO
D
UCT
I
O
N
Desp
ite
th
e
n
o
tab
le
i
m
p
ac
t
o
f
m
ac
h
i
n
e
lear
n
i
n
g
tech
n
iq
u
e
s
in
a
w
id
e
ar
ea
o
f
ap
p
licatio
n
d
o
m
ai
n
s
[1
-
4]
,
o
n
e
m
aj
o
r
is
s
u
e
th
at
e
m
er
g
ed
s
i
n
ce
th
e
b
eg
i
n
n
in
g
o
f
t
h
e
ap
p
licatio
n
d
ep
lo
y
m
en
t
is
co
m
p
le
x
it
y
o
f
d
esig
n
[5
-
7]
.
I
n
o
r
d
er
to
ac
h
iev
e
s
tate
-
of
-
t
h
e
-
ar
t
p
er
f
o
r
m
a
n
ce
,
m
ac
h
i
n
e
lear
n
i
n
g
t
ec
h
n
iq
u
es
r
eq
u
ir
e
ca
u
tio
u
s
s
tep
s
o
f
d
ata
p
r
e
-
p
r
o
ce
s
s
i
n
g
an
d
h
y
p
er
-
p
ar
a
m
eter
t
u
n
i
n
g
.
Mo
r
eo
v
er
,
g
iv
e
n
w
i
th
t
h
e
e
v
er
-
i
n
cr
ea
s
i
n
g
n
u
m
b
er
o
f
m
ac
h
i
n
e
lear
n
i
n
g
m
o
d
el
s
b
ei
n
g
d
ev
elo
p
ed
,
m
o
d
el
s
elec
tio
n
is
b
ec
o
m
i
n
g
i
n
c
r
ea
s
in
g
l
y
i
m
p
o
r
tan
t.
W
ith
t
h
e
d
e
m
a
n
d
s
f
r
o
m
i
n
e
x
p
er
ien
ce
d
d
ata
s
cie
n
ti
s
t
to
u
s
e
m
ac
h
in
e
lear
n
i
n
g
m
o
d
el
s
,
au
to
m
ati
n
g
th
e
alg
o
r
ith
m
s
elec
t
io
n
an
d
o
p
ti
m
izi
n
g
th
e
h
y
p
er
-
p
ar
a
m
eter
s
tu
n
in
g
h
a
ve
b
ee
n
a
m
o
n
g
o
f
th
e
p
r
ef
er
r
ed
ap
p
r
o
ac
h
es
o
f
th
e
d
ata
s
cien
ce
s
co
m
m
u
n
it
y
.
Au
to
m
a
ted
Ma
ch
in
e
L
ea
r
n
i
n
g
(
A
M
L
)
is
th
e
cu
r
r
en
t
ap
p
r
o
ac
h
p
r
o
v
ed
b
en
ef
icial
to
e
n
e
x
p
er
ien
ce
d
d
ata
s
cie
n
tis
t
an
d
i
n
s
o
m
e
ca
s
es
o
f
f
er
ed
b
etter
p
er
f
o
r
m
a
n
ce
s
th
a
n
t
h
e
co
n
v
e
n
tio
n
al
o
r
m
a
n
u
al
ap
p
r
o
ac
h
[4
-
5]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
A
r
ti
f
I
n
tell
I
SS
N:
2252
-
8938
Differ
en
t m
u
ta
tio
n
a
n
d
cro
s
s
o
ve
r
s
et
o
f g
en
etic
p
r
o
g
r
a
mmin
g
in
a
n
... (
S
u
r
a
ya
Ma
s
r
o
m
)
403
C
u
r
r
en
tl
y
,
A
M
L
h
a
s
b
ee
n
g
i
v
en
a
w
id
e
atte
n
tio
n
b
y
r
eseac
h
er
s
an
d
m
aj
o
r
o
f
w
o
r
k
s
w
er
e
d
ir
ec
ted
to
u
s
e
m
eta
-
h
eu
r
i
s
tic
s
ap
p
r
o
ac
h
to
o
p
ti
m
ize
t
h
e
m
ac
h
i
n
e
lear
n
in
g
p
ip
elin
e
s
[7
-
10]
.
P
ip
elin
e
s
is
a
co
llectio
n
o
f
m
ac
h
in
e
lear
n
i
n
g
m
o
d
el
s
p
o
o
led
f
r
o
m
th
e
p
r
e
-
d
ata
p
r
o
ce
s
s
in
g
to
t
h
e
h
y
p
er
-
p
ar
am
eter
s
t
u
n
i
n
g
.
E
x
a
m
p
le
s
o
f
m
eta
-
h
er
i
s
tics
s
ea
r
ch
alg
o
r
it
h
m
s
in
AM
L
ar
e
P
ar
ticle
S
w
ar
m
Op
ti
m
izatio
n
[
1
1
]
,
Gen
etic
P
r
o
g
r
a
m
m
in
g
(
GP
)
[1
2]
an
d
B
ay
esia
n
[
1
3
-
1
5
]
.
Dep
en
d
in
g
o
n
t
h
e
co
m
p
u
t
atio
n
al
av
ailab
ili
t
y
,
th
e
p
r
o
ce
s
s
to
co
m
p
lete
t
h
e
m
i
n
i
m
u
m
f
u
n
ctio
n
o
f
A
M
L
o
p
ti
m
izatio
n
is
co
n
s
id
er
ab
le
co
s
tl
y
.
T
h
is
is
d
u
e
to
th
e
co
m
p
le
x
p
r
o
ce
s
es
o
f
m
ac
h
i
n
e
lear
n
in
g
al
g
o
r
it
h
m
s
t
h
at
r
eq
u
i
r
es
a
m
o
d
el
tr
ain
ed
a
n
d
p
er
f
o
r
m
an
ce
s
e
v
al
u
atio
n
w
it
h
th
e
s
et
o
f
h
y
p
er
-
p
ar
a
m
eter
s
.
T
h
e
ev
al
u
atio
n
ti
m
e
w
il
l
b
e
e
x
ten
d
ed
w
h
e
n
t
h
e
d
ataset
is
n
o
t
w
ell
co
n
s
tr
u
cted
in
s
u
c
h
o
f
e
x
c
ess
i
v
e
f
ea
t
u
r
es
s
elec
tio
n
s
o
r
in
ap
p
r
o
p
r
iate
i
m
p
u
ta
tio
n
o
f
m
is
s
i
n
g
v
a
lu
e
s
.
Nev
er
th
e
less
,
b
y
u
n
d
er
s
ta
n
d
i
n
g
h
o
w
th
e
th
e
h
y
p
er
-
p
ar
a
m
eter
s
w
o
r
k
s
in
an
A
M
L
,
t
h
e
co
m
p
lex
i
ties
a
n
d
co
m
p
let
io
n
ti
m
e
to
co
m
p
le
te
m
i
g
h
t b
e
e
f
f
ec
t
iv
el
y
g
ain
ed
.
T
h
is
p
ap
er
p
r
esen
ts
th
e
d
escr
ip
tio
n
o
f
A
M
L
t
h
at
u
s
ed
p
ip
elin
es
o
p
ti
m
izatio
n
s
th
r
o
u
g
h
its
f
o
r
m
a
l
d
ef
in
i
tio
n
a
n
d
co
n
ce
p
tu
al
e
x
p
lan
atio
n
.
A
d
d
itio
n
a
ll
y
,
b
y
f
o
c
u
s
i
n
g
o
n
GP
b
ased
A
M
L
,
th
e
r
o
les
o
f
s
o
m
e
GP
h
y
p
er
-
p
ar
a
m
eter
s
n
a
m
e
l
y
m
u
t
atio
n
an
d
cr
o
s
s
o
v
er
r
ates
w
er
e
also
p
r
esen
ted
in
t
h
is
p
ap
er
s
u
p
p
o
r
ted
w
it
h
th
e
f
i
n
d
in
g
o
f
e
m
p
ir
ical
ex
p
er
i
m
e
n
ts
t
h
at
test
ed
o
n
v
ar
iatio
n
s
iz
es
o
f
d
atas
et.
T
o
th
e
b
est
o
f
o
u
r
k
n
o
w
led
g
e,
th
er
e
is
v
er
y
li
m
ited
liter
at
u
r
e
ca
n
b
e
f
o
u
n
d
o
n
d
is
c
u
s
s
i
n
g
an
d
e
x
p
lain
i
n
g
d
etail
o
f
GP
h
y
p
er
-
p
ar
am
eter
s
in
AM
L
.
Mo
r
eo
v
er
,
f
e
w
clai
m
s
w
er
e
r
ep
o
r
ted
o
n
th
e
s
etti
n
g
o
f
t
h
e
t
w
o
i
m
p
o
r
tan
t
o
f
co
n
tr
o
l
p
ar
a
m
eter
s
(
m
u
ta
tio
n
an
d
c
r
o
s
s
o
v
er
)
i
n
th
e
GP
o
p
ti
m
izatio
n
o
f
a
u
to
m
ated
m
ac
h
in
e
lear
n
in
g
.
T
h
e
f
i
n
d
i
n
g
p
r
o
v
id
es
ad
d
itio
n
al
f
u
n
d
a
m
e
n
tal
k
n
o
wled
g
e
to
th
e
d
es
ig
n
a
n
d
d
ev
elo
p
m
e
n
t
o
f
d
if
f
er
en
t
s
ett
in
g
o
f
m
u
tat
io
n
a
n
d
cr
o
s
s
o
v
er
r
ates f
o
r
th
e
A
M
L
b
ased
GP
.
2.
RE
S
E
ARCH
B
ACK
G
RO
U
ND
2
.
1
.
F
o
r
m
a
l def
ini
t
io
n o
f
m
a
chine le
a
rning
pi
pelin
e
o
pti
m
i
za
t
io
n in AM
L
T
h
e
f
o
r
m
a
l d
ef
i
n
itio
n
o
f
A
M
L
w
it
h
s
ea
r
c
h
o
p
ti
m
izatio
n
is
as
th
e
f
o
llo
w
i
n
g
:
1.
T
h
e
o
b
j
ec
tiv
e
is
to
m
i
n
i
m
ize
s
s
o
m
e
p
r
ed
ef
i
n
ed
lo
s
s
f
u
n
c
t
io
n
ℒ
(
(
)
;
ℳ
)
,
w
h
er
e
ℳ
is
th
e
tr
ai
n
i
n
g
m
o
d
el
te
s
ted
o
n
th
e
g
i
v
en
v
alid
atio
n
d
ataset
(
)
w
it
h
lear
n
i
n
g
al
g
o
r
ith
m
Α
th
at
lear
n
t
h
e
tr
ain
i
n
g
d
ataset
(
)
p
atter
n
.
2.
T
h
e
tr
ain
in
g
an
d
v
alid
atio
n
d
ataset
(
)
,
(
)
co
n
s
is
t
s
a
s
et
o
f
f
ea
tu
r
e
s
v
ec
to
r
to
b
e
s
e
lecte
d
,
p
r
e
-
p
r
o
ce
s
s
ed
an
d
w
ell
-
co
n
s
tr
u
cted
b
ased
o
n
s
o
m
e
f
ea
tu
r
iza
tio
n
m
ec
h
a
n
is
m
s
.
3.
T
h
e
lear
n
in
g
alg
o
r
it
h
m
Α
ca
n
b
e
a
s
in
g
le
m
ac
h
i
n
e
lear
n
i
n
g
o
r
a
h
y
b
r
id
izatio
n
d
if
f
er
en
t
al
g
o
r
ith
m
t
h
at
n
ee
d
to
b
e
co
n
f
ig
u
r
ed
w
i
th
a
s
et
o
f
h
y
p
er
p
ar
a
m
eter
s
Η
[
=
1
.
.
]
in
s
u
c
h
in
th
e
f
o
llo
w
i
n
g
(
1
)
.
ℳ
=
(
Α
=
0
(
(
)
;
Η
),
whe
r
e
Α
=
(
(
)
;
Η
);
Α
)
(
1
)
4.
T
h
er
ef
o
r
e,
th
e
g
o
al
o
f
t
h
e
A
M
L
s
ea
r
c
h
is
to
f
in
d
a
s
e
t o
f
h
y
p
er
p
ar
am
eter
s
Α
∗
th
at
y
ield
an
o
p
t
i
m
al
m
o
d
el
ℳ
∗
w
h
ic
h
m
i
n
i
m
ize
s
th
e
ℒ
(
(
)
;
ℳ
)
f
u
n
ctio
n
,
f
o
r
m
u
lated
as t
h
e
f
o
llo
w
i
n
g
(
2
)
.
Α
∗
=
Α
ℒ
(
(
)
;
(
Α
=
0
(
(
)
;
Η
)
;
Α
=
(
(
)
;
Η
)
)
;
Α
)
=
Α
min
ℱ
(
Α
;
(
Α
(
)
=
0
(
(
)
;
Η
)
;
Α
=
(
(
)
;
Η
)
)
,
ℒ
)
(
2
)
T
h
e
o
p
tim
ized
p
ip
elin
e
i
s
a
s
et
o
f
m
ac
h
in
e
lear
n
i
n
g
m
o
d
el
w
i
th
t
h
e
b
est
s
elec
tio
n
o
f
m
ac
h
in
e
lear
n
in
g
al
g
o
r
ith
m
an
d
t
h
e
b
es
t o
p
tim
a
l o
f
h
y
p
er
-
p
ar
a
m
e
ter
s
s
etti
n
g
.
2
.
2
.
G
enet
ic
p
ro
g
ra
mm
ing
b
a
s
ed
AM
L
A
M
L
w
it
h
GP
i
s
t
h
e
i
n
ter
est
o
f
s
t
u
d
y
r
ep
o
r
ted
in
t
h
is
p
ap
er
,
h
en
ce
A
M
L
T
P
O
T
P
y
t
h
o
n
lib
r
ar
y
h
a
s
b
ee
n
u
s
ed
a
s
th
e
e
x
p
er
i
m
e
n
tal
to
o
l
[
1
2
,
16]
.
A
ML
T
P
O
T
h
as
ca
p
ab
ilit
y
to
p
r
o
v
id
e
th
e
b
es
t
m
ac
h
in
e
lear
n
i
n
g
m
o
d
el
o
n
a
tr
ain
in
g
d
at
a
s
et
f
r
o
m
a
s
er
ies
o
f
p
r
o
ce
s
s
e
s
ca
lle
d
f
ea
t
u
r
izatio
n
,
al
g
o
r
it
h
m
s
ele
ctio
n
a
n
d
p
ar
a
m
eter
o
p
tim
izatio
n
a
s
p
r
esen
ted
in
F
ig
u
r
e
1
.
Featu
r
izatio
n
i
s
t
h
e
p
r
o
ce
s
s
o
f
s
elec
ti
n
g
th
e
b
est
p
o
s
s
ib
le
o
f
f
ea
tu
r
e
s
f
r
o
m
t
h
e
d
ata
s
et
t
h
r
o
u
g
h
th
e
P
r
in
cip
le
C
o
m
p
o
n
e
n
t
An
al
y
s
is
(
P
C
A
)
an
d
p
o
l
y
n
o
m
ial
a
p
p
r
o
ac
h
es.
I
t
also
in
co
r
p
o
r
ated
w
i
th
a
s
et
o
f
i
m
p
u
ta
tio
n
s
tr
ate
g
ies
i
n
h
an
d
l
in
g
t
h
e
m
i
s
s
i
n
g
v
al
u
es
o
f
th
e
tr
ain
i
n
g
d
ataset.
T
h
e
b
e
s
t
o
p
ti
m
ized
f
ea
t
u
r
es
w
ill
b
e
p
ass
ed
to
m
o
d
el
s
elec
t
io
n
s
m
o
d
u
les
t
h
at
s
u
p
p
o
r
ted
w
it
h
v
ar
io
u
s
k
in
d
s
o
f
m
ac
h
i
n
e
l
ea
r
n
in
g
al
g
o
r
it
h
m
s
(
m
o
r
e
th
an
f
i
f
t
y
o
f
cla
s
s
i
f
ier
s
an
d
r
eg
r
ess
o
r
s
)
.
T
h
e
s
elec
ted
alg
o
r
ith
m
o
r
ca
n
b
e
a
h
y
b
r
id
ized
m
o
d
el
w
ill
b
e
p
ass
ed
th
r
o
u
g
h
f
o
r
th
e
b
est
o
p
ti
m
ized
p
ar
a
m
eter
s
co
n
f
i
g
u
r
atio
n
s
.
A
M
L
T
P
O
T
u
s
ed
Gen
etic
P
r
o
g
r
a
m
m
in
g
(
GP
)
o
p
tim
izatio
n
ap
p
r
o
ac
h
in
s
elec
t
in
g
t
h
e
b
est
m
o
d
el
g
ai
n
ed
f
r
o
m
t
h
e
co
m
b
i
n
atio
n
s
o
f
p
ip
eli
n
es
(
f
ea
tu
r
izat
io
n
,
m
o
d
el
s
elec
t
io
n
s
,
p
ar
am
eter
s
o
p
ti
m
izatio
n
)
in
th
e
alg
o
r
it
h
m
s
elec
tio
n
p
o
o
l.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8938
I
n
t J
A
r
ti
f
I
n
tell
,
Vo
l.
9
,
No
.
3
,
Sep
te
m
b
er
20
20
:
4
0
2
–
408
404
Fea
tur
izat
ion
Fea
tur
es
selection
Fea
tur
es
pre-pr
ocessing
Fea
tur
es
constructio
n
Mo
del
sel
ecti
on
Mo
del
para
me
ter
optimization
Fig
u
r
e
1
.
A
M
L
T
P
O
T
m
ai
n
f
u
n
ctio
n
s
As a
f
a
m
il
y
o
f
ev
o
l
u
tio
n
ar
y
al
g
o
r
ith
m
,
t
h
e
p
er
f
o
r
m
a
n
ce
o
f
G
P
is
h
i
g
h
l
y
d
ep
en
d
in
g
o
n
th
e
p
ar
a
m
eter
s
co
n
f
i
g
u
r
atio
n
m
ai
n
l
y
p
o
p
u
lat
io
n
s
izi
n
g
,
m
u
tatio
n
a
n
d
cr
o
s
s
o
v
er
r
ate
s
.
Mu
tatio
n
a
n
d
cr
o
s
s
o
v
er
ar
e
t
w
o
ess
e
n
tial
ele
m
en
t
s
in
a
n
ev
o
lu
tio
n
ar
y
b
ased
alg
o
r
it
h
m
[
1
7
-
19]
lik
e
GP
th
at
ca
n
in
f
l
u
en
ce
t
h
e
s
ea
r
c
h
o
p
tim
izatio
n
lan
d
s
ca
p
e
b
y
m
ea
n
s
o
f
e
x
p
lo
itatio
n
an
d
e
x
p
l
o
r
atio
n
.
T
h
e
o
p
ti
m
izatio
n
o
f
b
est
o
p
ti
m
al
r
esu
lt
n
ee
d
s
a
b
alan
ce
b
et
w
ee
n
b
o
th
ex
p
lo
itatio
n
a
n
d
ex
p
lo
r
atio
n
[
2
0
]
.
E
x
p
lo
r
atio
n
ca
n
eli
m
in
ate
th
e
s
ea
r
c
h
to
co
n
v
er
g
e
p
r
e
m
at
u
r
el
y
b
y
a
v
o
i
d
in
g
th
e
s
o
lu
tio
n
f
r
o
m
tr
ap
p
in
g
i
n
lo
ca
l
o
p
ti
m
a
w
h
ile
e
x
p
lo
i
tatio
n
f
o
cu
s
es
to
a
n
in
te
n
s
it
y
s
ea
r
ch
d
ir
ec
tio
n
s
o
th
at
th
e
b
est
p
o
te
n
tial
s
o
l
u
tio
n
ca
n
b
e
co
n
v
er
g
ed
at
th
e
g
l
o
b
al
o
p
tim
a
[
2
1
]
.
Mu
tatio
n
is
h
i
g
h
l
y
r
elate
d
to
th
e
e
x
p
lo
r
atio
n
asp
ec
t
to
co
n
t
r
o
l
th
e
s
ea
r
ch
d
i
v
er
s
i
t
y
,
o
f
te
n
b
e
d
ef
in
ed
w
it
h
a
v
er
y
s
m
all
p
r
o
b
ab
ilit
y
r
ate
to
p
r
o
d
u
ce
a
litt
le
m
o
d
if
icatio
n
t
o
an
in
d
iv
id
u
al
(
ch
r
o
m
o
s
o
m
e)
.
On
th
e
o
th
er
h
a
n
d
,
cr
o
s
s
o
v
er
is
u
s
ed
to
id
e
n
ti
f
y
th
e
s
tar
tin
g
p
o
in
t
o
f
t
h
e
b
est
o
p
tim
a
l
s
o
l
u
tio
n
a
t
g
lo
b
al
o
p
ti
m
a
a
n
d
f
u
r
t
h
er
l
y
d
r
iv
es
t
h
e
s
ea
r
c
h
d
ir
ec
tio
n
in
ac
h
iev
i
n
g
th
e
b
est
s
o
l
u
tio
n
a
t
lo
ca
l
o
p
tim
a.
T
h
er
ef
o
r
e,
cr
o
s
s
o
v
er
is
g
en
er
all
y
th
o
u
g
h
t
to
en
ab
le
t
h
e
e
x
p
lo
itatio
n
asp
ec
t.
I
n
s
i
m
p
ler
w
o
r
d
s
,
th
e
p
u
r
p
o
s
e
o
f
m
u
tatio
n
is
to
s
ea
r
ch
lo
ca
ll
y
ar
o
u
n
d
th
e
p
r
o
m
is
in
g
ar
ea
s
f
o
u
n
d
b
y
t
h
e
cr
o
s
s
o
v
e
r
o
p
er
ato
r
.
Gen
er
all
y
,
t
h
e
m
o
s
t
ap
p
r
o
p
r
iate
m
u
ta
tio
n
an
d
cr
o
s
s
o
v
er
s
etti
n
g
s
r
eq
u
ir
ed
b
y
an
ev
o
lu
tio
n
ar
y
alg
o
r
ith
m
li
k
e
GP
to
s
o
lv
e
d
if
f
er
en
t
o
p
ti
m
izatio
n
p
r
o
b
le
m
s
ca
n
b
e
d
if
f
er
en
t
[
2
2
-
24]
.
I
n
ad
d
itio
n
,
ev
id
en
ce
s
s
h
o
w
th
a
t
ev
e
n
f
o
r
o
n
e
s
p
ec
if
i
c
o
p
tim
izat
io
n
p
r
o
b
lem
,
t
h
e
r
eq
u
ir
ed
b
est
s
tr
ateg
ies
f
o
r
th
e
b
o
th
o
p
er
ato
r
s
m
a
y
v
ar
y
d
u
r
i
n
g
th
e
e
v
o
lu
tio
n
ar
y
p
r
o
ce
s
s
[
2
5
-
26]
.
3.
RE
S
E
ARCH
M
E
T
H
O
D
3
.
1
.
E
x
peri
m
ent
s
a
n
d pa
ra
m
et
er
s
co
nfig
ura
t
io
n
I
n
th
i
s
s
t
u
d
y
,
t
h
e
ex
p
er
i
m
en
t
co
d
es
w
er
e
d
ev
elo
p
ed
w
it
h
P
y
h
to
n
3
.
6
J
u
p
y
ter
No
teb
o
o
k
p
latf
o
r
m
a
n
d
w
er
e
i
m
p
le
m
e
n
ted
i
n
th
e
p
r
o
ce
s
s
o
r
o
f
I
n
tel
i7
7
th
G
e
n
er
atio
n
p
r
o
ce
s
s
o
r
,
1
6
G
B
R
A
M.
T
h
e
A
M
L
T
P
O
T
m
ac
h
i
n
e
lea
r
n
in
g
m
o
d
el
h
a
s
b
ee
n
s
et
to
e
m
p
lo
y
7
5
:2
5
p
er
ce
n
t
r
atio
b
etw
ee
n
t
h
e
tr
ain
i
n
g
a
n
d
cr
o
s
s
v
alid
atio
n
f
o
r
all
th
e
ex
p
er
i
m
en
ts
.
T
P
O
T
u
s
ed
cr
o
s
s
v
alid
atio
n
f
o
r
th
e
m
ac
h
i
n
e
lear
n
in
g
tr
ain
i
n
g
ap
p
r
o
ac
h
an
d
th
e
n
u
m
b
er
o
f
d
ef
a
u
lt
v
a
lu
e
f
o
r
th
e
tr
ain
in
g
an
d
v
alid
atio
n
ch
u
n
k
s
is
d
ef
i
n
ed
as
5
.
T
h
er
ef
o
r
e,
f
o
r
ea
ch
p
o
p
u
latio
n
s
ize
p
,
t
h
e
n
u
m
b
er
o
f
iter
at
io
n
s
f
o
r
s
e
lecti
n
g
th
e
b
est
p
ip
eli
n
es
i
s
eq
u
i
v
alen
t
to
5
*
p
ti
m
es.
I
f
th
e
n
u
m
b
er
o
f
p
o
p
u
latio
n
s
ize
is
1
0
,
th
e
GP
iter
atio
n
to
f
in
d
i
n
g
t
h
e
b
est
p
ip
elin
es
is
50
ti
m
es.
A
lo
n
g
er
ti
m
e
w
ill
b
e
r
eq
u
ir
ed
to
co
m
p
lete
i
f
t
h
e
n
u
m
b
er
o
f
iter
atio
n
s
in
cr
ea
s
ed
ac
co
r
d
in
g
to
t
h
e
p
o
p
u
lat
io
n
s
ize
an
d
g
e
n
er
ati
o
n
s
etti
n
g
.
T
ab
le
1
lis
ts
th
e
p
ar
a
m
eter
s
u
s
ed
in
th
is
r
esear
ch
.
T
h
e
n
u
m
b
er
o
f
iter
atio
n
s
f
o
r
t
h
e
GP
to
s
et
t
h
e
f
i
n
al
o
p
ti
m
ize
d
p
ip
elin
es
is
d
eter
m
i
n
ed
b
y
t
h
e
n
u
m
b
er
o
f
g
e
n
er
atio
n
s
i
n
s
u
ch
th
at
t
h
e
alg
o
r
it
h
m
s
h
o
u
ld
w
o
r
k
b
etter
w
it
h
m
o
r
e
g
en
er
atio
n
s
.
Nev
er
t
h
eles
s
,
lo
n
g
er
ti
m
e
w
o
u
ld
b
e
r
eq
u
ir
ed
to
co
m
p
let
e
w
ith
b
ig
g
er
n
u
m
b
er
o
f
g
e
n
e
r
atio
n
an
d
5
i
s
t
h
e
co
m
m
o
n
s
ettin
g
u
s
ed
i
n
m
a
n
y
ex
p
er
i
m
e
n
ts
.
P
o
p
u
latio
n
s
ize
i
s
t
h
e
n
u
m
b
er
o
f
i
n
d
i
v
id
u
al
s
o
r
p
o
ten
tial
p
ip
eli
n
es
t
h
at
ca
n
b
e
s
to
r
ed
in
th
e
GP
s
elec
tio
n
p
o
o
l.
Ou
r
p
r
ev
io
u
s
ex
p
er
i
m
e
n
ts
f
o
u
n
d
t
h
at
t
h
e
d
ef
au
lt
p
o
p
u
lat
io
n
s
ize
1
0
is
f
it
e
n
o
u
g
h
to
th
e
alg
o
r
ith
m
to
p
r
o
d
u
ce
b
est
o
p
ti
m
al
r
es
u
lt.
T
h
e
o
f
f
s
p
r
i
n
g
_
s
ize
b
y
d
ef
a
u
lt
i
s
n
o
n
e
s
o
th
at
it
w
il
l
b
e
eq
u
al
to
th
e
p
o
p
u
latio
n
_
s
ize.
Mu
tatio
n
r
ate
is
a
s
m
all
v
al
u
e
u
s
ed
to
co
n
tr
o
l
th
e
GP
ex
p
lo
r
atio
n
s
ea
r
ch
b
y
ap
p
l
y
in
g
r
an
d
o
m
ch
an
g
es
to
s
o
m
e
o
f
th
e
p
ip
eli
n
es
i
n
th
e
s
e
lectio
n
p
o
o
l.
T
h
e
cr
o
s
s
o
v
er
f
u
r
th
er
m
o
r
e
is
u
s
ed
to
tell
th
e
GP
h
o
w
m
an
y
p
ip
eli
n
es
to
r
ep
r
o
d
u
ce
d
at
ea
ch
o
f
t
h
e
g
en
er
atio
n
.
T
h
e
v
al
u
es
f
o
r
b
o
th
m
u
s
t
b
e
i
n
b
et
w
ee
n
0
.
1
to
0
.
9
an
d
n
o
t
ex
ce
ed
1
.
0
.
T
h
e
m
u
ta
tio
n
a
n
d
cr
o
s
s
o
v
er
r
ates
s
etti
n
g
ar
e
d
i
v
id
ed
i
n
to
t
h
r
ee
s
e
ts
o
f
p
air
.
F
ir
s
t
p
air
is
MCR
1
a
n
d
MC
R
2
,
w
h
ic
h
t
h
e
r
an
g
e
i
s
b
et
w
ee
n
0
.
1
an
d
0
.
9
,
s
ec
o
n
d
p
air
is
M
C
R
3
an
d
M
C
R
4
w
it
h
i
n
0
.
3
an
d
0
.
7
an
d
last
l
y
i
s
m
o
d
er
ate
m
u
tatio
n
a
n
d
cr
o
s
s
o
v
er
r
ate
s
(
0
.
5
,
0
.
5
)
.
I
t
is
an
ticip
ated
i
n
t
h
i
s
s
t
u
d
y
t
h
at
h
i
g
h
e
r
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
A
r
ti
f
I
n
tell
I
SS
N:
2252
-
8938
Differ
en
t m
u
ta
tio
n
a
n
d
cro
s
s
o
ve
r
s
et
o
f g
en
etic
p
r
o
g
r
a
mmin
g
in
a
n
... (
S
u
r
a
ya
Ma
s
r
o
m
)
405
cr
o
s
s
o
v
er
r
ates
in
MCR
1
an
d
MCR
3
ca
n
h
elp
th
e
al
g
o
r
ith
m
to
ac
h
iev
e
b
etter
o
p
ti
m
al
r
esu
lt
s
w
h
i
le
lo
w
er
cr
o
s
s
o
v
er
r
ates in
M
C
R
2
an
d
MCR
4
m
a
y
a
f
f
ec
t o
n
s
h
o
r
ten
i
n
g
t
h
e
co
m
p
letio
n
ti
m
e.
T
ab
le
1
.
T
h
e
im
p
o
r
tan
t p
ar
a
m
eter
s
P
a
r
a
me
t
e
r
s
C
o
n
f
i
g
u
r
e
d
v
a
l
u
e
g
e
n
e
r
a
t
i
o
n
s
5
p
o
p
u
l
a
t
i
o
n
_
si
z
e
10
o
f
f
sp
r
i
n
g
_
s
i
z
e
n
o
n
e
M
u
t
a
t
i
o
n
&
C
r
o
sso
v
e
r
r
a
t
e
s
D
i
v
i
d
e
d
i
n
t
o
5
g
r
o
u
p
s:
MC
R1
=
M
R
0
.
1
,
C
R
0
.
9
MC
R
2
=
M
R
0
.
9
,
C
R
0
.
1
MC
R
3
=
M
R
0
.
3
,
C
R
0
.
7
MC
R
4
=
M
R
0
.
7
,
C
R
0
.
3
MC
R
5
=
M
R
0
.
5
,
C
R
0
.
5
*
M
C
R
=
m
u
t
a
t
i
o
n
a
n
d
c
r
o
sso
v
e
r
r
a
t
e
s,
M
R
=
mu
t
a
t
i
o
n
r
a
t
e
,
C
R
=
c
r
o
sso
v
e
r
r
a
t
e
3
.
2
.
Da
t
a
s
et
s
T
h
e
ev
al
u
atio
n
o
f
th
is
s
t
u
d
y
h
as
b
ee
n
co
n
d
u
cted
o
n
a
co
ll
ec
tio
n
o
f
8
d
atase
ts
w
it
h
d
i
f
f
er
en
t
s
ize
s
d
o
w
n
lo
ad
ed
f
r
o
m
Ka
g
g
le
w
eb
s
ite
(
www
.
k
a
g
g
le.
co
m
)
as li
s
t
ed
in
th
e
f
o
llo
w
in
g
T
ab
le
2
.
A
lt
h
o
u
g
h
t
h
e
d
ataset
s
izes
w
i
ll
n
o
t
t
h
e
m
aj
o
r
f
ac
to
r
o
f
m
a
ch
in
e
lear
n
i
n
g
p
er
f
o
r
m
a
n
ce
s
,
in
ter
esti
n
g
in
f
o
r
m
atio
n
m
i
g
h
t
b
e
i
n
f
er
r
e
d
w
h
en
u
s
in
g
d
ataset
v
ar
iati
o
n
s
o
n
te
s
ti
n
g
th
e
AM
L
T
POT
w
i
th
d
i
f
f
er
en
t
p
ar
am
eter
izat
io
n
s
o
f
m
u
tatio
n
an
d
cr
o
s
s
o
v
er
.
T
ab
le
2
.
Data
s
ets
D
a
t
a
se
t
A
b
b
r
e
v
i
a
t
i
o
n
S
i
z
e
o
f
r
e
c
o
r
d
s
D
o
t
a
2
g
a
me
D
o
t
a
9
0
0
0
0
H
u
ma
n
_
R
e
so
u
r
c
e
s_
Em
p
l
o
y
e
e
_
A
t
t
r
i
t
i
o
n
H
R
EA
1
4
9
9
9
O
r
a
n
g
e
_
T
e
l
e
c
o
m_
C
h
u
r
n
_
D
a
t
a
O
T
C
D
5
0
0
0
H
e
a
r
t
a
t
t
a
c
k
H
e
a
r
t
3
0
0
F
r
a
u
d
t
r
a
n
s
a
c
t
i
o
n
FT
1
5
0
8
7
0
A
d
u
l
t
I
n
c
o
me
AI
3
2
0
0
0
D
i
a
b
e
t
i
c
D
i
a
b
e
t
i
c
1
1
5
1
B
r
e
a
st
c
a
n
c
e
r
B
r
e
a
st
c
a
n
c
e
r
9
0
0
4.
RE
SU
L
T
S
A
ND
D
I
SCU
SS
I
O
N
I
n
th
is
s
ec
tio
n
,
th
e
r
es
u
lt
o
f
ac
cu
r
ac
y
s
co
r
e
an
d
ti
m
e
to
co
m
p
lete
(
T
OC
)
o
f
th
e
alg
o
r
ith
m
w
it
h
th
e
f
i
v
e
m
u
tatio
n
an
d
cr
o
s
s
o
v
er
r
ates
is
ex
p
lai
n
ed
ac
co
r
d
in
g
to
th
e
d
ata
s
izes
:
s
m
all
an
d
b
ig
.
Hea
r
t,
b
r
ea
s
tcan
ce
r
,
d
iab
etic
an
d
OT
C
D
w
er
e
g
r
o
u
p
ed
as
a
s
m
a
ll
d
ataset
wh
ile
o
n
t
h
e
B
r
ea
s
tca
n
ce
r
an
d
OT
C
D
d
atasets
,
th
e
d
i
f
f
er
en
t
s
etti
n
g
o
f
m
u
tati
o
n
an
d
cr
o
s
s
o
v
er
r
ate
s
d
id
n
o
t
p
r
esen
t
a
m
u
c
h
d
if
f
er
en
ce
to
th
e
ac
c
u
r
ac
y
s
co
r
e.
Dif
f
er
en
t
s
h
ap
es
o
f
r
es
u
lts
c
an
b
e
s
ee
n
o
n
th
e
Hea
r
t
a
n
d
Diab
etics
d
atase
t.
T
h
e
h
i
g
h
est
s
co
r
e
o
n
Hea
r
t
p
r
o
v
id
ed
b
y
MC
R
2
(
0
.
9
,
0
.
1
)
w
h
ile
M
C
R
5
(
0
.
3
,
0
.
7
)
is
th
e
b
est s
etti
n
g
o
n
Diab
etic.
4
.
1
.
Acc
ura
cy
s
co
re
Fig
u
r
e
2
p
r
esen
t
s
t
h
e
r
es
u
lts
o
f
ac
c
u
r
ac
y
s
co
r
e
f
o
r
th
e
AM
L
T
P
O
T
test
ed
o
n
th
e
d
atasets
w
it
h
s
izes
l
ess
t
h
an
1
0
0
0
0
.
T
h
e
d
ef
au
lt
c
o
n
f
i
g
u
r
atio
n
M
C
R
1
(
0
.
1
,
0
.
9
)
o
f
th
e
A
M
L
T
P
O
T
r
em
ai
n
as
th
e
o
p
ti
m
al
s
etti
n
g
f
o
r
t
w
o
d
atasets
(
Hea
r
t,
Diab
etic)
b
u
t
n
o
t
m
u
c
h
e
f
f
ec
t
o
n
t
h
e
B
r
ea
s
tcan
ce
r
an
d
OT
C
D
d
ata
s
ets.
B
y
lo
o
k
in
g
at
ea
ch
p
air
o
f
m
u
tatio
n
an
d
cr
o
s
s
o
v
er
s
etti
n
g
,
th
er
e
h
av
e
b
ee
n
s
o
m
e
p
er
f
o
r
m
a
n
ce
s
ef
f
ec
t
t
o
th
e
ac
cu
r
ac
y
s
co
r
e
to
th
e
th
r
ee
o
f
d
ata
s
ets
(
Hea
r
t,
B
r
ea
s
tcan
ce
r
,
Diab
etic)
.
Sli
g
h
tl
y
i
m
p
r
o
v
e
m
e
n
t
also
ca
n
b
e
s
ee
n
o
n
t
h
e
OT
C
D
d
ataset.
B
o
th
h
i
g
h
er
cr
o
s
s
o
v
er
r
ates
in
MC
R
1
(
cr
o
s
s
o
v
er
r
ate
0
.
9
)
an
d
MCR
3
(
cr
o
s
s
o
v
er
r
ate
0
.
7
)
h
av
e
ac
h
iev
ed
h
ig
h
er
ac
cu
r
ac
y
s
co
r
e
co
m
p
ar
ed
to
th
e
lo
w
er
cr
o
s
s
o
v
er
r
ate
in
MC
R
2
(
cr
o
s
s
o
v
er
r
ate
0
.
1
)
an
d
MCR
4
(
cr
o
s
s
o
v
er
r
ate
0
.
3
)
.
W
ith
t
h
e
m
o
d
er
ate
s
et
tin
g
o
f
MC
R
5
(
0
.
5
,
0
.
5
)
,
n
o
n
e
o
f
s
i
n
g
le
p
atter
n
ca
n
b
e
o
b
s
er
v
ed
f
r
o
m
all
th
e
d
ata
s
ets.
Fig
u
r
e
3
p
r
esen
ts
th
e
ac
cu
r
ac
y
s
co
r
es
o
f
AM
L
T
P
O
T
w
i
th
d
if
f
er
e
n
t
s
et
s
o
f
m
u
tatio
n
an
d
cr
o
s
s
o
v
er
o
n
th
e
b
ig
g
er
s
ize
o
f
d
ata
s
ets.
Sli
g
h
t
l
y
d
if
f
er
en
t
w
it
h
t
h
e
s
m
all
d
atasets
,
h
i
g
h
er
cr
o
s
s
o
v
er
r
ates
in
MC
R
1
a
n
d
MCR
3
d
id
n
o
t
p
r
ese
n
t
a
g
r
ea
t
i
m
p
r
o
v
e
m
e
n
t
to
t
h
e
ac
c
u
r
a
cy
s
co
r
es
o
f
all
d
ata
s
ets.
I
n
f
a
ct,
s
li
g
h
t
l
y
h
i
g
h
er
ac
cu
r
ac
y
s
co
r
e
ca
n
b
e
s
ee
n
o
n
H
R
E
A
,
A
I
a
n
d
Do
ta
w
it
h
MRC
4
(
0
.
3
cr
o
s
s
o
v
er
r
ate)
.
N
ev
er
th
ele
s
s
,
w
i
th
th
e
v
er
y
s
m
all
i
m
p
r
o
v
e
m
e
n
t,
t
h
e
d
ef
au
lt
s
et
tin
g
o
f
M
C
R
1
(
0
.
1
,
0
.
9
)
ca
n
b
e
co
n
s
id
er
ed
as
t
h
e
b
est
s
ett
in
g
f
o
r
t
h
e
A
M
L
T
P
O
T
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8938
I
n
t J
A
r
ti
f
I
n
tell
,
Vo
l.
9
,
No
.
3
,
Sep
te
m
b
er
20
20
:
4
0
2
–
408
40
6
Fig
u
r
e
2
.
A
cc
u
r
ac
y
s
co
r
e
f
o
r
s
m
all
s
ize
d
atasets
Fig
u
r
e
3
.
A
cc
u
r
ac
y
s
co
r
e
f
o
r
l
ar
g
e
d
atasizes
B
ased
o
n
th
e
r
esu
lt
s
p
r
esen
ted
f
r
o
m
th
e
ab
o
v
e
Fi
g
u
r
e
2
an
d
Fig
u
r
e
3
,
th
e
g
en
er
al
co
n
c
lu
s
i
o
n
th
at
ca
n
b
e
m
ad
e
i
s
t
h
at
cr
o
s
s
o
v
er
w
i
t
h
t
h
e
e
x
p
lo
itatio
n
asp
ec
t
ca
n
b
e
s
et
to
a
h
ig
h
er
v
al
u
e
t
h
an
m
u
tatio
n
r
ates
s
o
t
h
at
th
e
s
ea
r
ch
d
ir
ec
tio
n
ca
n
g
o
d
ee
p
f
o
r
b
etter
in
ten
s
it
y
i
n
f
i
n
d
i
n
g
t
h
e
b
est
m
ac
h
i
n
e
lear
n
in
g
p
ip
elin
e
s
.
Ho
w
e
v
er
,
f
o
r
th
e
ca
s
e
o
f
b
ig
d
atasizes,
f
u
th
er
e
m
p
ir
ical
e
x
p
er
im
e
n
t
s
w
it
h
d
if
f
er
e
n
t
f
ac
to
r
s
m
ig
h
t
b
e
ess
en
tial
to
b
e
co
n
d
u
cted
.
4
.
2
.
T
i
m
e
t
o
co
m
p
let
e
(
T
T
C
)
T
h
e
T
T
C
s
in
s
ec
o
n
d
t
h
at
t
h
e
alg
o
r
it
m
h
a
s
ta
k
en
f
o
r
ea
ch
m
u
tatio
n
an
d
cr
o
s
s
o
v
er
r
ates
ar
e
p
r
esen
ted
in
Fi
g
u
r
e
4
an
d
Fi
g
u
r
e
5
.
A
s
t
h
e
lo
w
er
cr
o
s
s
o
v
er
r
ates
s
h
o
u
ld
tak
e
s
m
aller
T
T
C
,
it
is
in
te
r
esti
n
g
to
g
e
t
id
ea
s
o
n
h
o
w
t
h
e
d
i
f
f
er
e
n
t
d
ata
s
ets
m
a
y
a
f
f
ec
t
o
n
t
h
e
T
T
C
.
B
ased
o
n
t
h
e
t
h
eo
r
y
,
M
C
R
2
o
r
MCR
4
w
it
h
lo
w
er
cr
o
s
s
o
v
er
r
ates sh
o
u
ld
ab
le
to
co
m
p
lete
f
as
ter
th
a
n
th
e
M
C
R
1
an
d
MCR
3
.
As
p
r
esen
ted
in
Fig
u
r
e
4
,
o
n
ly
B
r
ea
s
tca
n
ce
r
an
d
Diab
etic
d
atasets
th
at
s
u
p
p
o
r
t
th
e
th
eo
r
y
.
MC
R
2
an
d
MCR
4
w
it
h
lo
w
er
cr
o
s
s
o
v
er
r
ates
(
0
.
1
an
d
0
.
3
)
h
as
lo
w
er
T
T
C
th
an
M
C
R
1
an
d
M
C
R
3
(
0
.
9
an
d
0
.
7
)
.
On
t
h
e
Hea
r
t
a
n
d
OT
C
D,
lo
wer
cr
o
s
s
o
v
er
r
ate
in
MC
R
4
f
a
ce
d
th
e
lo
w
er
T
T
C
b
u
t
f
o
u
n
d
to
b
e
s
i
m
ilar
T
T
C
v
alu
e
s
o
n
M
C
R
1
a
n
d
MCR
2
.
Fo
r
all
d
atasets
,
m
o
d
er
ate
cr
o
s
s
o
v
er
r
ate
0
.
5
is
th
e
lo
w
es
t
T
T
C
.
Ma
j
o
r
ef
f
ec
t
o
f
lo
w
er
cr
o
s
s
o
v
er
r
ates to
th
e
T
T
C
s
o
n
all
b
ig
d
atasets
ca
n
b
e
s
ee
n
i
n
Fi
g
u
r
e
5
.
M
CR
1
M
CR
2
M
CR
3
M
CR
4
M
CR
5
Hea
r
t
(
3
0
0
)
0
,
8
5
0
,
8
1
0
,
8
0
0
,
7
9
0
,
7
4
B
r
ea
s
t
c
a
n
c
er
(
6
0
0
)
0
,
9
8
0
,
9
7
0
,
9
8
0
,
9
7
0
,
9
8
D
i
a
b
et
i
c
(
1
1
5
1
)
0
,
7
5
0
,
7
2
0
,
7
0
0
,
7
0
0
,
6
9
O
T
CD
(
5
0
0
0
)
0
,
9
6
0
,
9
6
0
,
9
5
0
,
9
5
0
,
9
5
0
,
6
5
0
,
7
0
0
,
7
5
0
,
8
0
0
,
8
5
0
,
9
0
0
,
9
5
1
,
0
0
A
c
c
ur
ac
y
S
c
o
r
e
M
C
R
1
M
C
R
2
M
C
R
3
M
C
R
4
M
C
R
5
Dot
a
(
9K)
0,
59
0,
59
0,
59
0,
60
0,
60
HR
E
A(
15
K)
0,
99
0,
98
0,
99
0,
99
0,
95
AI
(
32K)
0,
83
0,
83
0,
81
0,
82
0,
83
F
T
(
150
K)
0,
95
0,
96
0,
96
0,
96
0,
94
0,
55
0,
60
0,
65
0,
70
0,
75
0,
80
0,
85
0,
90
0,
95
1,
00
1,
05
A
c
c
u
r
a
c
y
S
c
or
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
A
r
ti
f
I
n
tell
I
SS
N:
2252
-
8938
Differ
en
t m
u
ta
tio
n
a
n
d
cro
s
s
o
ve
r
s
et
o
f g
en
etic
p
r
o
g
r
a
mmin
g
in
a
n
... (
S
u
r
a
ya
Ma
s
r
o
m
)
407
Fig
u
r
e
4
.
T
T
C
f
o
r
s
m
al
l size
d
atasets
Fig
u
r
e
5
.
T
T
C
f
o
r
lar
g
e
s
ize
d
atasizes
On
al
l
lar
g
e
s
ize
d
ataset
s
,
M
C
R
2
(
cr
o
s
s
o
v
er
r
ate
0
.
1
)
an
d
MC
R
4
(
cr
o
s
s
o
v
er
r
ate
0
.
3
)
h
av
e
lo
w
er
T
T
C
s
co
m
p
ar
ed
to
MC
R
1
(
cr
o
s
s
o
v
er
r
ate
0
.
9
)
an
d
MRC
3
(
cr
o
s
s
o
v
er
r
ate
0
.
7
)
.
On
th
e
c
o
n
tr
ar
y
w
it
h
s
m
al
l
d
atasets
,
th
e
m
o
d
er
ate
s
etti
n
g
in
MC
R
5
d
id
n
o
t h
av
e
a
p
atter
n
o
f
e
f
f
ec
t t
h
e
t
h
e
T
T
C
s
.
B
ased
o
n
Fig
u
r
e
4
an
d
Fig
u
r
e
5
,
it
ca
n
b
e
g
en
er
all
y
co
n
cl
u
d
ed
th
at
lo
w
er
cr
o
s
s
o
v
er
r
ates
en
co
u
r
ag
e
th
e
GP
o
p
tim
izatio
n
s
ea
r
ch
to
s
to
p
th
e
p
r
o
ce
s
s
at
ea
r
lier
s
tag
e
m
ain
l
y
o
n
t
h
e
b
ig
d
ata
s
ize.
5.
CO
NCLU
SI
O
N
T
h
is
p
ap
er
p
r
esen
ts
t
h
e
t
h
e
o
r
y
,
d
esi
g
n
o
f
ex
p
er
i
m
en
t
s
an
d
f
i
n
d
in
g
s
r
elate
d
to
th
e
ef
f
ec
t
o
f
p
ar
am
eter
izat
io
n
s
o
f
g
e
n
etic
p
r
o
g
r
a
m
m
in
g
to
an
a
u
to
m
ated
m
ac
h
in
e
lear
n
in
g
.
T
h
e
g
e
n
etic
p
r
o
g
r
am
m
i
n
g
i
s
a
k
in
d
o
f
m
eta
-
h
e
u
r
is
tic
s
u
s
ed
as
th
e
s
ea
r
ch
o
p
ti
m
izatio
n
o
f
f
i
n
d
in
g
th
e
b
est
m
ac
h
i
n
e
lear
n
in
g
p
ip
elin
es.
Desig
n
i
n
g
th
e
b
est
m
ac
h
in
e
le
ar
n
in
g
i
s
a
d
if
f
ic
u
lt
e
n
d
ea
v
o
u
r
an
d
co
s
tl
y
m
a
in
l
y
f
o
r
n
e
w
d
a
ta
s
cien
tis
t
s
.
W
it
h
th
e
au
to
m
ated
m
ac
h
i
n
e
lear
n
i
n
g
,
t
h
e
b
est
s
ett
in
g
o
f
m
ac
h
i
n
e
lear
n
i
n
g
m
o
d
el
ca
n
b
e
ef
f
ec
tiv
el
y
co
n
d
u
cted
.
Ho
w
e
v
er
,
alt
h
o
u
g
h
w
it
h
a
u
to
m
atio
n
s
c
h
e
m
e,
r
esear
c
h
er
s
k
e
ep
o
n
q
u
e
s
tio
n
i
n
g
t
h
e
h
y
p
er
-
p
ar
a
m
eter
s
s
etti
n
g
o
f
au
to
m
ated
m
ac
h
i
n
e
lear
n
i
n
g
,
in
t
h
i
s
ca
s
e
th
e
r
o
le
o
f
cr
o
s
s
o
v
er
an
d
m
u
tatio
n
r
ates.
R
a
th
er
t
h
an
u
s
in
g
t
h
e
d
ef
au
lt
s
etti
n
g
,
it
w
ill
b
e
u
s
e
f
u
l
i
f
t
h
e
r
esear
h
er
s
i
s
p
r
o
v
id
ed
w
it
h
t
h
e
t
h
eo
r
y
a
n
d
r
ea
s
o
n
a
b
le
f
i
n
d
in
g
f
o
r
th
e
p
ar
am
eter
izat
io
n
s
o
f
A
M
L
T
POT
.
A
s
f
o
r
f
u
t
u
r
e
w
o
r
k
s
,
it
w
i
ll
b
e
in
ter
esti
n
g
to
s
tu
d
y
m
o
r
e
ex
ten
s
iv
e
l
y
o
n
t
h
e
n
atu
r
e
o
f
d
ataset
s
,
f
ea
t
u
r
izati
o
n
s
to
th
e
p
er
f
o
r
m
an
ce
s
o
f
A
M
L
T
P
O
T
an
d
th
e
im
p
l
e
m
en
tatio
n
o
f
p
ar
allel
co
m
p
u
ti
n
g
p
lat
f
o
r
m
i
n
ex
ec
u
ti
n
g
t
h
e
b
ig
d
atase
ts
.
M
C
R
1
M
C
R
2
M
C
R
3
M
C
R
4
M
C
R
5
He
a
r
t
(
300)
60
60
65
30
15
B
r
e
a
s
tca
nc
e
r
(
600)
73
69
70
45
20
Dia
be
ti
c
(
1151
)
91
44
134
27
82
OT
C
D
(
5000)
180
180
180
120
60
0
20
40
60
80
100
120
140
160
180
200
T
i
m
e
to
C
om
pl
e
te
(
s
e
c
on
d)
M
C
R
1
M
C
R
2
M
C
R
3
M
C
R
4
M
C
R
5
D
ot
a
(
9K
)
360
0
240
0
252
00
900
0
180
00
H
R
E
A
(
15
K
)
900
600
240
180
300
A
I
(
32K
)
174
0
960
108
0
270
540
F
T
(
150
K
)
72
0
0
66
0
0
72
0
0
54
0
0
72
0
0
0
500
0
100
00
150
00
200
00
250
00
300
00
T
i
m
e
to
C
om
pl
e
te
(
s
e
c
on
d)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8938
I
n
t J
A
r
ti
f
I
n
tell
,
Vo
l.
9
,
No
.
3
,
Sep
te
m
b
er
20
20
:
4
0
2
–
408
408
ACK
NO
WL
E
D
G
E
M
E
NT
S
T
h
e
au
th
o
r
s
w
o
u
ld
li
k
e
to
th
a
n
k
U
n
i
v
er
s
it
i
T
ek
n
o
lo
g
i
M
AR
A
an
d
Mi
n
is
tr
y
o
f
E
d
u
ca
t
io
n
Ma
la
y
s
i
a
f
o
r
th
e
f
in
a
n
cial
s
u
p
p
o
r
t o
f
th
i
s
p
r
o
j
e
ct
u
n
d
er
FR
G
S g
r
an
t N
o
6
0
0
-
I
R
MI
/FR
GS 5
/3
(
2
0
8
/2
0
1
9
)
.
RE
F
E
R
E
NC
E
S
[1
]
S
.
Ha
ru
n
a
n
d
M
.
F
.
I
b
ra
h
im
,
“
A
g
e
n
e
ti
c
a
lg
o
rit
h
m
b
a
se
d
tas
k
sc
h
e
d
u
li
n
g
sy
ste
m
f
o
r
lo
g
isti
c
s
se
rv
ic
e
ro
b
o
ts
,
”
Bu
ll
.
El
e
c
tr.
En
g
.
In
fo
rm
a
t
ics
,
v
o
l.
8
,
n
o
.
1
,
p
p
.
2
0
6
-
2
1
3
,
2
0
1
9
.
[2
]
A
.
F
.
J.
A
l
-
G
b
u
ri,
S
.
Na
im
,
a
n
d
A
.
N.
Bo
ra
ik
,
“
H
y
b
rid
iza
ti
o
n
o
f
Ba
t
a
n
d
G
e
n
e
ti
c
A
l
g
o
rit
h
m
to
S
o
lv
e
N
-
Qu
e
e
n
s
P
r
o
b
lem
,
”
Bu
ll
.
El
e
c
tr.
En
g
.
I
n
f
o
rm
a
ti
c
s
,
v
o
l.
7
,
n
o
.
4
,
p
p
.
6
2
6
-
6
3
2
,
2
0
1
8
.
[3
]
M
.
M
o
z
a
a
n
d
S
.
Ku
m
a
r,
“
Ro
u
ti
n
g
in
Ne
tw
o
rk
s
u
sin
g
G
e
n
e
ti
c
A
l
g
o
rit
h
m
,
”
Bu
ll
.
El
e
c
tr.
En
g
.
I
n
fo
rm
a
ti
c
s
,
v
o
l.
6
,
n
o
.
1
,
p
p
.
8
8
-
9
8
,
2
0
1
7
.
[4
]
S
.
S
in
g
h
,
M
.
Ka
u
sh
ik
,
A
.
G
u
p
ta,
a
n
d
A
.
K.
M
a
lv
i
y
a
,
“
W
e
a
th
e
r
F
o
re
c
a
stin
g
u
sin
g
M
a
c
h
in
e
L
e
a
rn
in
g
T
e
c
h
n
iq
u
e
s,”
S
S
RN
El
e
c
tro
n
.
J
.
,
n
o
.
6
,
p
p
.
3
8
-
4
1
,
2
0
1
9
.
[5
]
F
.
R.
Ka
m
a
la,
P
.
R.
J.
T
h
a
n
g
a
iah
,
a
n
d
A
.
In
f
o
,
“
A
n
i
m
p
ro
v
e
d
h
y
b
rid
f
e
a
tu
re
se
lec
ti
o
n
m
e
th
o
d
f
o
r
h
u
g
e
d
ime
n
sio
n
a
l
d
a
tas
e
ts,”
IAE
S
In
t.
J
.
Art
if
.
I
n
tell
.
,
v
o
l.
8
,
n
o
.
1
,
p
p
.
7
7
-
8
6
,
2
0
1
9
.
[6
]
V
.
S
.
P
a
d
a
la,
K.
G
a
n
d
h
i,
a
n
d
D.
V
P
u
s
h
p
a
lath
a
,
“
M
a
c
h
in
e
lea
rn
in
g
:
th
e
n
e
w
lan
g
u
a
g
e
f
o
r
a
p
p
li
c
a
ti
o
n
s,”
IAE
S
In
t.
J
.
Arti
f.
I
n
tell
.
,
v
o
l.
8
,
n
o
.
4
,
p
p
.
4
1
1
-
4
2
1
,
2
0
1
9
.
[7
]
A
.
Za
k
ra
n
i,
M
.
Ha
in
,
a
n
d
A
.
I
d
ri,
“
Im
p
ro
v
in
g
so
f
t
w
a
re
d
e
v
e
l
o
p
m
e
n
t
e
ff
o
rt
e
sti
m
a
ti
o
n
u
s
in
g
su
p
p
o
r
t
v
e
c
to
r
re
g
re
ss
io
n
a
n
d
f
e
a
tu
re
se
lec
ti
o
n
,
”
IAE
S
I
n
t.
J
.
Arti
f.
In
tell
.
,
v
o
l.
8
,
n
o
.
4
,
p
p
.
3
9
9
-
4
1
0
,
2
0
1
9
.
[8
]
R.
S
.
Olso
n
,
N.
Ba
rtl
e
y
,
R.
J.
Ur
b
a
n
o
w
icz
,
a
n
d
J.
H.
M
o
o
re
,
“
Ev
a
lu
a
ti
o
n
o
f
a
tre
e
-
b
a
se
d
p
ip
e
li
n
e
o
p
ti
miza
ti
o
n
t
o
o
l
fo
r
a
u
t
o
ma
t
in
g
d
a
ta
sc
ien
c
e
,”
in
P
ro
c
e
e
d
in
g
s
o
f
th
e
G
e
n
e
ti
c
a
n
d
Ev
o
l
u
ti
o
n
a
ry
Co
m
p
u
tatio
n
C
o
n
f
e
re
n
c
e
2
0
1
6
,
2
0
1
6
,
p
p
.
4
8
5
-
4
9
2
.
[9
]
R.
S
.
Olso
n
,
R.
J.
Urb
a
n
o
w
icz
,
P
.
C.
A
n
d
re
w
s,
N.
A
.
L
a
v
e
n
d
e
r,
J.
H.
M
o
o
re
,
a
n
d
o
th
e
rs,
“
A
u
to
m
a
t
in
g
b
io
m
e
d
ica
l
d
a
ta
sc
ien
c
e
th
ro
u
g
h
tree
-
b
a
se
d
p
i
p
e
l
in
e
o
p
ti
m
iza
ti
o
n
,
”
in
E
u
ro
p
e
a
n
C
o
n
fer
e
n
c
e
o
n
t
h
e
A
p
p
li
c
a
ti
o
n
s
o
f
Evo
lu
ti
o
n
a
ry
Co
mp
u
t
a
ti
o
n
,
2
0
1
6
,
p
p
.
1
2
3
-
1
3
7
.
[1
0
]
M
.
S
u
g
a
n
u
m
a
,
S
.
S
h
irak
a
w
a
,
a
n
d
T
.
Na
g
a
o
,
“
A
Ge
n
e
ti
c
Pro
g
ra
mm
in
g
A
p
p
r
o
a
c
h
to
De
sig
n
in
g
Co
n
v
o
l
u
ti
o
n
a
l
Ne
u
ra
l
Ne
two
rk
Arc
h
it
e
c
tu
re
s
,”
i
n
P
r
o
c
e
e
d
in
g
s
o
f
th
e
G
e
n
e
ti
c
a
n
d
Ev
o
lu
ti
o
n
a
ry
Co
m
p
u
tatio
n
Co
n
f
e
re
n
c
e
,
2
0
1
7
,
p
p
.
4
9
7
-
5
0
4
.
[1
1
]
C.
S
u
i,
M
.
Be
n
n
a
m
o
u
n
,
a
n
d
R.
T
o
g
n
e
ri,
“
De
e
p
f
e
a
tu
re
lea
rn
in
g
f
o
r
d
u
m
m
ies
:
A
si
m
p
le
a
u
to
-
e
n
c
o
d
e
r
trai
n
in
g
m
e
th
o
d
u
sin
g
P
a
rti
c
le S
w
a
rm
Op
t
im
is
a
ti
o
n
,
”
P
a
tt
e
rn
Rec
o
g
n
it
.
L
e
t
t
.
,
v
o
l.
9
4
,
p
p
.
7
5
-
8
0
,
Ju
l
.
2
0
1
7
.
[1
2
]
R.
S
.
Ols
o
n
a
n
d
J.
H.
M
o
o
re
,
“
T
POT
:
A
T
re
e
-
b
a
se
d
Pi
p
e
li
n
e
Op
ti
miza
ti
o
n
T
o
o
l
fo
r
Au
t
o
m
a
ti
n
g
M
a
c
h
in
e
L
e
a
rn
in
g
,
”
i
n
P
r
o
c
e
e
d
in
g
s o
f
th
e
W
o
rk
sh
o
p
o
n
A
u
to
m
a
ti
c
M
a
c
h
in
e
L
e
a
rn
in
g
,
2
0
1
6
,
v
o
l.
6
4
,
p
p
.
6
6
-
7
4
.
[1
3
]
L
.
Ko
tt
h
o
f
f
,
C.
T
h
o
rn
to
n
,
H.
H.
Ho
o
s,
F
.
Hu
tt
e
r,
a
n
d
K.
L
e
y
to
n
-
Bro
w
n
,
“
A
u
to
-
W
EKA
2
.
0
:
A
u
to
m
a
ti
c
m
o
d
e
l
se
lec
ti
o
n
a
n
d
h
y
p
e
rp
a
ra
m
e
ter o
p
ti
m
iz
a
ti
o
n
in
W
EKA
,
”
J
.
M
a
c
h
.
L
e
a
rn
.
Res
.
,
v
o
l.
1
8
,
n
o
.
1
,
p
p
.
8
2
6
-
8
3
0
,
2
0
1
7
.
[1
4
]
E.
Bro
c
h
u
,
V
.
M
.
Co
ra
,
a
n
d
N.
De
F
re
it
a
s,
“
A
tu
to
rial
o
n
Ba
y
e
si
a
n
o
p
ti
m
iza
ti
o
n
o
f
e
x
p
e
n
siv
e
c
o
st
f
u
n
c
ti
o
n
s,
w
it
h
a
p
p
li
c
a
ti
o
n
t
o
a
c
ti
v
e
u
se
r
m
o
d
e
li
n
g
a
n
d
h
iera
rc
h
ica
l
re
in
f
o
rc
e
m
e
n
t
lea
rn
in
g
,
”
a
rX
iv
P
re
p
r
.
a
rXiv
1
0
1
2
.
2
5
9
9
,
2
0
1
0
.
[1
5
]
M
.
F
e
u
re
r,
A
.
Kle
in
,
K.
Eg
g
e
n
sp
e
rg
e
r,
J.
S
p
rin
g
e
n
b
e
rg
,
M
.
Blu
m
,
a
n
d
F
.
H
u
tt
e
r,
“
Ef
f
icie
n
t
a
n
d
ro
b
u
st
a
u
t
o
m
a
ted
m
a
c
h
in
e
lea
rn
in
g
,
”
in
Ad
v
a
n
c
e
s i
n
n
e
u
r
a
l
i
n
fo
rm
a
ti
o
n
p
ro
c
e
ss
in
g
s
y
ste
ms
,
2
0
1
5
,
p
p
.
2
9
6
2
-
2
9
7
0
.
[1
6
]
P
.
G
ij
sb
e
rs,
J.
V
a
n
sc
h
o
re
n
,
a
n
d
R.
S
.
Ols
o
n
,
“
L
a
y
e
re
d
T
PO
T
:
S
p
e
e
d
i
n
g
u
p
tre
e
-
b
a
se
d
p
ip
e
li
n
e
o
p
t
imiza
ti
o
n
,
”
CEUR
W
o
rk
sh
o
p
P
r
o
c
.
,
v
o
l
.
1
9
9
8
,
2
0
1
7
.
[1
7
]
K.
Ka
m
il
,
K.
H.
Ch
o
n
g
,
H.
Ha
sh
im
,
a
n
d
S
.
A
.
S
h
a
a
y
a
,
“
A
m
u
lt
ip
le
m
it
o
sis
g
e
n
e
ti
c
a
l
g
o
rit
h
m
,
”
IA
ES
In
t.
J
.
Arti
f
.
In
tell
.
,
v
o
l.
8
,
n
o
.
3
,
p
p
.
2
5
2
-
2
5
8
,
2
0
1
9
.
[1
8
]
I.
A
.
Hu
m
ied
,
“
S
o
lv
in
g
N
-
Qu
e
e
n
s
P
ro
b
lem
Us
in
g
S
u
b
p
ro
b
lem
s
b
a
se
d
o
n
G
e
n
e
ti
c
A
lg
o
rit
h
m
,
”
IAE
S
I
n
t.
J
.
Arti
f
.
In
tell
.
,
v
o
l.
7
,
n
o
.
3
,
p
p
.
1
3
0
-
1
3
7
,
2
0
1
8
.
[1
9
]
N.
K.
a
n
d
P
.
N.
R.
,
“
A
m
u
lt
io
b
jec
ti
v
e
g
e
n
e
ti
c
p
ro
g
ra
m
m
in
g
-
b
a
se
d
e
n
se
m
b
le
f
o
r
si
m
u
lt
a
n
e
o
u
s
f
e
a
tu
re
se
lec
ti
o
n
a
n
d
c
las
si
f
ica
ti
o
n
,
”
F.
T
r
a
n
s.
Cy
b
e
rn
.
,
v
o
l.
2
0
1
6
4
6
,
n
o
.
2
,
p
p
.
4
9
9
-
5
1
0
,
2
0
1
6
.
[2
0
]
K.
C.
T
a
n
,
S
.
C.
C
h
iam
,
a
.
a
.
M
a
m
u
n
,
a
n
d
C.
K.
G
o
h
,
“
Ba
lan
c
in
g
e
x
p
lo
ra
ti
o
n
a
n
d
e
x
p
l
o
it
a
ti
o
n
w
it
h
a
d
a
p
ti
v
e
v
a
riatio
n
f
o
r
e
v
o
lu
ti
o
n
a
ry
m
u
lt
i
-
o
b
jec
ti
v
e
o
p
ti
m
iza
ti
o
n
,
”
E
u
r.
J
.
Op
e
r.
Res
.
,
v
o
l.
1
9
7
,
n
o
.
2
,
p
p
.
7
0
1
-
7
1
3
,
S
e
p
.
2
0
0
9
.
[2
1
]
S
.
M
a
sro
m
,
I.
M
o
se
r,
J.
M
o
n
tg
o
m
e
r
y
,
S
.
Z.
Z.
A
b
id
in
,
a
n
d
N.
Om
a
r,
“
Hy
b
rid
iza
ti
o
n
o
f
p
a
rticle
swa
rm
o
p
ti
miza
t
io
n
wit
h
a
d
a
p
ti
v
e
g
e
n
e
ti
c
a
lg
o
rit
h
m
o
p
e
ra
t
o
rs
,
”
in
2
0
1
3
I
n
tern
a
ti
o
n
a
l
Co
n
f
e
re
n
c
e
o
n
In
telli
g
e
n
t
S
y
ste
m
s
De
sig
n
a
n
d
A
p
p
li
c
a
ti
o
n
s,
2
0
1
3
,
p
p
.
1
-
6.
[2
2
]
S
.
M
a
sro
m
,
S
.
Z.
Z.
A
b
id
i
n
,
N.
O
m
a
r,
a
n
d
K.
Na
sir,
“
T
ime
-
V
a
r
y
in
g
M
u
tati
o
n
in
P
a
r
ti
c
le
S
w
a
r
m
Op
ti
m
iza
ti
o
n
,
”
in
In
telli
g
e
n
t
I
n
fo
rm
a
t
io
n
a
n
d
Da
t
a
b
a
se
S
y
ste
ms
-
L
NCS
,
S
p
ri
n
g
e
r
-
V
e
rlag
Be
rli
n
He
id
e
lb
e
rg
,
2
0
1
3
,
p
p
.
3
1
-
4
0
.
[2
3
]
S
.
M
a
sro
m
,
S
.
Z.
Z.
A
b
id
in
,
N.
Om
a
r,
A
.
S
.
A
.
Ra
h
m
a
n
,
a
n
d
Z.
I.
Riz
m
a
n
,
“
D
y
n
a
m
i
c
p
a
ra
m
e
teriz
a
t
io
n
s
o
f
p
a
rti
c
le
sw
a
r
m
o
p
ti
m
iza
ti
o
n
a
n
d
g
e
n
e
ti
c
a
lg
o
rit
h
m
f
o
r
fa
c
il
it
y
la
y
o
u
t
p
r
o
b
lem
,
”
AR
PN
J
.
E
n
g
.
A
p
p
l
.
S
c
i
.
,
v
o
l.
1
2
,
n
o
.
1
0
,
2
0
1
7
.
[2
4
]
N.
Hig
a
sh
i
a
n
d
H.
Ib
a
,
“
P
a
rti
c
l
e
s
w
a
r
m
o
p
ti
m
iz
a
ti
o
n
w
it
h
G
a
u
ss
ian
m
u
tatio
n
,
”
in
IE
EE
S
w
a
r
m
In
telli
g
e
n
c
e
S
y
mp
o
si
u
m
-
S
I
S
’0
3
.
,
2
0
0
3
,
p
p
.
7
2
-
7
9
.
[2
5
]
S
.
M
a
sro
m
,
S
.
Z.
Z.
A
b
id
in
,
N.
O
m
a
r,
K.
Na
sir,
a
n
d
A
.
S
.
A
b
d
Ra
h
m
a
n
,
“
D
y
n
a
m
ic
P
a
ra
m
e
teri
z
a
ti
o
n
o
f
th
e
P
a
rti
c
le
S
w
a
r
m
Op
ti
m
iz
a
ti
o
n
a
n
d
G
e
n
e
ti
c
A
lg
o
rit
h
m
H
y
b
rid
s
f
o
r
V
e
h
icl
e
Ro
u
ti
n
g
P
ro
b
lem
w
it
h
T
i
m
e
W
in
d
o
w
,
”
In
t.
J
.
Hy
b
rid
In
tell
.
S
y
st.,
v
o
l.
1
2
,
n
o
.
1
,
p
p
.
1
3
-
2
5
,
2
0
1
5
.
[
2
6
]
W
.
M
.
Da
h
a
lan
,
H.
M
o
k
h
li
s
,
R.
A
h
m
a
d
,
A
.
H.
A
.
Ba
k
a
r,
a
n
d
I.
M
u
sirin
,
“
S
im
u
lt
a
n
e
o
u
s
n
e
tw
o
rk
re
c
o
n
f
ig
u
ra
ti
o
n
a
n
d
DG
siz
in
g
u
sin
g
e
v
o
lu
ti
o
n
a
r
y
p
ro
g
ra
m
m
in
g
a
n
d
g
e
n
e
ti
c
a
lg
o
rit
h
m
to
m
in
im
iz
e
p
o
w
e
r
lo
ss
e
s,”
Ara
b
.
J
.
S
c
i.
En
g
.
,
v
o
l.
3
9
,
n
o
.
8
,
p
p
.
6
3
2
7
-
6
3
3
8
,
2
0
1
4
.
Evaluation Warning : The document was created with Spire.PDF for Python.