I
AE
S
I
nte
rna
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
.
417
~
4
23
I
SS
N:
2252
-
8938
,
DOI
: 1
0
.
1
1
5
9
1
/i
j
ai.
v
9
.i
3
.
p
p
417
-
4
23
417
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
i
.
ia
esco
r
e.
co
m
Lo
ng
-
ter
m
loa
d f
o
recas
ting usin
g
grey
w
o
lf
o
pti
m
i
z
e
r
-
lea
st
-
squa
res suppo
rt
v
ector
m
a
chine
Z
.
M
.
Ya
s
in
1
,
N.
A.
Sa
li
m
2
,
N.
F
.
A.
Aziz
3
,
Y.
M
.
Ali
4
,
H
.
M
o
ha
m
a
d
5
1
,
2,
4,
5
F
a
c
u
lt
y
o
f
El
e
c
tri
c
a
l
En
g
in
e
e
rin
g
,
Un
iv
e
rsiti
T
e
k
n
o
lo
g
i
M
A
RA
,
M
a
la
y
si
a
3
F
a
c
u
lt
y
o
f
El
e
c
tri
c
a
l
En
g
in
e
e
rin
g
,
Un
iv
e
rsiti
T
e
n
a
g
a
Na
sio
n
a
l,
M
a
la
y
sia
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Feb
2
7
,
2
0
2
0
R
ev
i
s
ed
A
p
r
2
0
,
2
0
2
0
A
cc
ep
ted
Ma
y
1
7
,
2
0
2
0
L
o
n
g
ter
m
lo
a
d
f
o
re
c
a
stin
g
d
a
ta
is
im
p
o
rtan
t
f
o
r
g
rid
e
x
p
a
n
sio
n
a
n
d
p
o
w
e
r
s
y
ste
m
o
p
e
ra
ti
o
n
.
Be
sid
e
s,
it
a
lso
i
m
p
o
rtan
t
to
e
n
su
re
th
e
g
e
n
e
ra
ti
o
n
c
a
p
a
c
it
y
m
e
e
t
e
lec
tri
c
it
y
d
e
m
a
n
d
a
t
a
ll
ti
m
e
s.
In
th
is
p
a
p
e
r,
L
e
a
st
-
S
q
u
a
re
S
u
p
p
o
r
t
V
e
c
to
r
M
a
c
h
i
n
e
(L
S
S
VM)
is
u
se
d
to
p
re
d
ict
t
h
e
lo
n
g
-
term
lo
a
d
d
e
m
a
n
d
.
F
o
u
r
in
p
u
ts
a
re
c
o
n
sid
e
re
d
w
h
ich
a
re
p
e
a
k
lo
a
d
d
e
m
a
n
d
,
a
m
b
ien
t
tem
p
e
r
a
tu
re
,
h
u
m
id
it
y
a
n
d
w
in
d
s
p
e
e
d
.
T
o
tal
lo
a
d
d
e
m
a
n
d
is se
t
a
s
th
e
o
u
t
p
u
t
o
f
p
re
d
ictio
n
in
L
S
S
V
M
.
In
o
rd
e
r
to
im
p
ro
v
e
th
e
a
c
c
u
ra
c
y
o
f
th
e
L
S
S
V
M
,
G
re
y
W
o
lf
Op
ti
m
i
z
e
r
(
GW
O)
is
h
y
b
rid
ize
d
to
o
b
tai
n
t
h
e
o
p
ti
m
a
l
p
a
ra
m
e
ter
s
o
f
L
S
S
V
M
n
a
m
e
l
y
GW
O
-
L
S
S
V
M
.
M
e
a
n
A
b
so
lu
te
P
e
rc
e
n
t
a
g
e
Err
o
r
(M
A
P
E)
is
u
se
d
a
s
th
e
q
u
a
n
ti
f
y
m
e
a
su
re
m
e
n
t
o
f
th
e
p
re
d
icti
o
n
m
o
d
e
l.
T
h
e
o
b
jec
ti
v
e
o
f
th
e
o
p
ti
m
iza
ti
o
n
is
t
o
m
in
im
ize
th
e
v
a
lu
e
o
f
M
A
P
E.
T
h
e
p
e
r
f
o
rm
a
n
c
e
o
f
GW
O
-
L
S
S
VM
is
c
o
m
p
a
re
d
w
it
h
o
th
e
r
m
e
th
o
d
s
su
c
h
a
s
L
S
S
V
M
a
n
d
A
n
t
L
io
n
Op
t
im
ize
r
–
L
e
a
st
-
S
q
u
a
re
S
u
p
p
o
r
t
V
e
c
to
r
M
a
c
h
in
e
(AL
O
-
L
S
S
V
M
).
F
r
o
m
th
e
re
s
u
lt
s
o
b
tain
e
d
,
i
t
c
a
n
b
e
c
o
n
c
lu
d
e
d
t
h
a
t
GW
O
-
L
S
S
V
M
p
ro
v
i
d
e
lo
w
e
r
M
A
P
E
v
a
lu
e
w
h
ich
is
0
.
1
3
%
a
s
c
o
m
p
a
re
d
to
o
th
e
r
m
e
th
o
d
s.
K
ey
w
o
r
d
s
:
Gr
e
y
w
o
l
f
o
p
ti
m
izer
L
o
ad
f
o
r
ec
asti
n
g
L
S
SVM
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
:
Z
u
h
ai
la
Ma
t
Yasi
n
,
Facu
lt
y
o
f
E
lectr
ical
E
n
g
in
ee
r
in
g
,
Un
i
v
er
s
iti T
ek
n
o
lo
g
i M
A
R
A
,
4
0
4
5
0
Sh
ah
A
la
m
,
Sela
n
g
o
r
,
Ma
la
y
s
ia
.
E
m
ail: z
u
h
ai
la
m
y
7
4
@
g
m
ail.
c
o
m
1.
I
NT
RO
D
UCT
I
O
N
P
lan
n
i
n
g
in
p
o
w
er
s
y
s
te
m
s
is
a
p
r
o
j
ec
tio
n
o
f
h
o
w
th
e
s
y
s
te
m
s
h
o
u
ld
e
v
o
lv
e
o
v
er
a
s
p
ec
if
ic
ti
m
e
-
f
r
a
m
e.
I
t
h
a
s
b
ec
o
m
e
m
o
r
e
d
if
f
icu
l
t
to
p
lan
p
o
w
er
s
y
s
te
m
s
in
d
ev
e
lo
p
in
g
co
u
n
tr
ie
s
.
P
lan
n
in
g
m
u
s
t
b
e
co
m
p
leted
d
esp
ite
n
u
m
er
o
u
s
v
u
l
n
er
ab
ilit
ie
s
s
u
c
h
a
s
f
u
tu
r
e
lo
ad
d
esig
n
s
,
p
o
p
u
latio
n
g
r
o
w
t
h
a
n
d
ec
o
n
o
m
ic
g
r
o
w
t
h
r
ep
r
esen
ti
n
g
d
ev
elo
p
in
g
co
u
n
tr
ies,
tec
h
n
i
ca
l,
ec
o
n
o
m
ic
an
d
en
v
ir
o
n
m
e
n
t
al
co
n
s
tr
ain
t
s
[
1
]
.
L
o
ad
f
o
r
ec
asti
n
g
p
la
y
s
a
n
i
m
p
o
r
tan
t
r
o
le
in
p
lan
n
i
n
g
p
o
w
e
r
s
y
s
te
m
s
to
en
s
u
r
e
u
n
i
n
ter
r
u
p
ted
an
d
ec
o
n
o
m
ical
p
o
w
er
g
en
er
atio
n
an
d
d
is
tr
ib
u
tio
n
to
co
n
s
u
m
er
s
.
P
r
ec
is
e
elec
tr
ical
lo
ad
f
o
r
ec
asti
n
g
i
s
a
m
aj
o
r
is
s
u
e
in
th
e
p
lan
n
i
n
g
an
d
m
a
n
a
g
e
m
en
t o
f
elec
tr
icit
y
g
en
er
ati
n
g
u
t
ilit
ie
s
.
T
h
e
lo
ad
f
o
r
ec
ast
ca
n
b
e
d
i
v
id
ed
in
to
th
r
ee
cla
s
s
i
f
icatio
n
s
.
S
h
o
r
t
-
ter
m
lo
ad
f
o
r
ec
asti
n
g
(
ST
L
F)
co
v
er
s
an
h
o
u
r
to
a
w
ee
k
ti
m
e
[
2
]
.
Sh
o
r
t
ter
m
lo
ad
f
o
r
ec
as
tin
g
co
n
tr
ib
u
te
s
to
th
e
d
ec
is
io
n
-
m
a
k
i
n
g
p
r
o
ce
s
s
,
in
cl
u
d
in
g
u
n
it
co
m
m
it
m
e
n
t
a
n
d
p
r
o
ce
s
s
f
o
r
i
m
p
r
o
v
in
g
s
e
cu
r
it
y
[
3
]
.
T
h
e
ac
cu
r
ate
lo
ad
f
o
r
ec
asti
n
g
h
elp
s
en
h
a
n
ci
n
g
th
e
u
n
it
co
m
m
it
m
en
t
s
c
h
e
d
u
li
n
g
an
d
t
h
u
s
ca
n
s
a
v
e
lar
g
e
a
m
o
u
n
t
o
f
co
s
t
p
er
y
ea
r
[
4
]
.
Me
d
iu
m
lo
ad
f
o
r
ec
a
s
t
in
g
(
MT
L
F)
co
v
er
s
a
w
ee
k
-
to
-
y
ea
r
p
er
io
d
.
I
t
is
ess
en
tial
f
o
r
f
u
e
l
s
u
p
p
l
y
an
d
m
ai
n
ten
a
n
ce
o
p
er
atio
n
s
to
b
e
p
lan
n
ed
[
5
]
.
L
o
n
g
-
t
er
m
lo
ad
f
o
r
ec
ast
(
L
T
L
F)
f
o
r
ec
ast
f
o
r
m
o
r
e
th
an
o
n
e
y
ea
r
[
6
]
.
L
o
n
g
ter
m
l
o
ad
f
o
r
ec
asti
n
g
f
ac
ilit
ate
s
th
e
d
ec
is
io
n
-
m
a
k
i
n
g
o
f
a
n
elec
tr
ic
al
u
tili
t
y
in
cl
u
d
i
n
g
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
1
7
–
4
23
418
th
e
p
u
r
ch
a
s
e
an
d
g
en
er
atio
n
o
f
elec
tr
icit
y
,
lo
ad
s
w
itc
h
in
g
an
d
in
f
r
astr
u
ctu
r
e
d
ev
elo
p
m
e
n
t.
T
h
e
im
p
o
r
tan
ce
o
f
lo
ad
p
r
ed
ictio
n
tu
r
n
s
o
u
t
to
b
e
m
o
r
e
n
o
te
w
o
r
t
h
y
in
h
i
g
h
-
g
r
o
w
t
h
d
ev
e
lo
p
in
g
co
u
n
tr
ies.
Un
d
er
s
tan
d
i
n
g
an
d
f
o
r
ec
asti
n
g
o
f
lo
ad
ch
ar
ac
ter
is
tics
w
as
co
m
p
lex
d
u
e
to
it
s
d
ep
en
d
en
ce
o
n
a
w
id
e
r
a
n
g
e
o
f
f
ac
to
r
s
i
n
f
lu
e
n
ci
n
g
th
e
w
ea
th
er
,
g
eo
g
r
ap
h
ical
d
iv
er
s
it
y
,
s
u
n
r
i
s
e/s
u
n
s
et
ti
m
e,
s
e
aso
n
al
d
iv
er
s
it
y
,
etc.
[
7
]
.
Des
p
ite
th
e
in
cr
ea
s
e
in
elec
tr
icit
y
d
e
m
a
n
d
an
n
u
all
y
,
t
h
e
p
atter
n
o
f
t
h
e
lo
ad
p
r
o
f
ile
m
a
y
al
s
o
ch
a
n
g
e
[8
-
9
]
.
An
al
y
s
i
s
o
f
lo
n
g
-
ter
m
lo
ad
f
o
r
ec
asti
n
g
h
as
r
ec
ei
v
ed
litt
le
atten
tio
n
.
R
e
f
er
e
n
ce
[
10
]
p
r
esen
ted
a
lo
n
g
ter
m
lo
ad
f
o
r
ec
ast
u
s
i
n
g
F
u
zz
y
L
o
g
ic
m
o
d
el
.
Fu
zz
y
lo
g
ic
m
o
d
el
w
a
s
d
ev
e
lo
p
ed
b
ased
o
n
th
e
w
ea
t
h
er
p
ar
am
eter
s
w
h
ic
h
ar
e
te
m
p
er
atu
r
e
a
n
d
h
u
m
id
it
y
.
T
w
o
y
ea
r
s
h
i
s
to
r
ical
lo
ad
d
ata
w
er
e
u
tili
ze
d
to
p
r
ed
ict
a
y
ea
r
-
a
h
ea
d
lo
ad
d
e
m
a
n
d
.
T
h
e
v
al
u
e
o
f
Me
a
n
Av
er
a
g
e
P
er
ce
n
tag
e
E
r
r
o
r
(
MA
P
E
)
o
b
tain
ed
w
as
6
.
9
%.
T
h
e
tech
n
iq
u
e
i
s
t
h
en
i
m
p
r
o
v
ed
in
[1
1
]
u
s
i
n
g
F
u
zz
y
-
Neu
r
o
.
Fu
zz
y
-
Ne
u
r
o
is
t
h
e
co
m
b
i
n
at
io
n
o
f
Fu
zz
y
L
o
g
ic
an
d
A
r
ti
f
ic
ial
Neu
r
al
Ne
t
w
o
r
k
(
A
NN)
.
I
n
F
u
zz
y
-
Neu
r
o
,
t
h
e
o
u
tp
u
t
o
f
t
h
e
f
u
zz
y
lo
g
ic
s
y
s
te
m
a
r
e
f
ed
to
A
NN.
Fu
zz
y
lo
g
ic
s
y
s
te
m
s
u
s
ed
f
o
r
m
ak
in
g
d
ec
is
io
n
b
ased
o
n
th
e
r
u
les,
m
ea
n
w
h
ile,
A
NN
is
u
s
ed
f
o
r
tr
ain
i
n
g
an
d
test
i
n
g
p
r
o
ce
s
s
.
T
h
e
ch
ar
ac
ter
is
tic
o
f
ANN
ar
e
h
ig
h
lear
n
in
g
ca
p
ab
ilit
ie
s
,
ad
ap
tab
ilit
y
an
d
g
en
er
aliza
t
io
n
.
B
y
u
s
i
n
g
F
u
zz
y
-
Ne
u
r
o
,
MA
P
E
h
as
i
m
p
r
o
v
ed
to
1
.
2
2
%.
R
ec
en
t
l
y
,
v
ar
io
u
s
A
r
ti
f
ic
ial
I
n
telli
g
e
n
t
(
A
I
)
tec
h
n
iq
u
e
s
ca
n
b
e
i
m
p
le
m
e
n
ted
f
o
r
f
o
r
ec
asti
n
g
.
On
e
o
f
t
h
e
m
is
ANN.
R
esear
ch
i
n
A
N
N
h
a
s
r
ec
ei
v
ed
co
n
s
id
er
ab
le
atten
tio
n
.
ANN
h
as
m
an
y
ad
v
an
tag
e
s
a
s
co
m
p
ar
ed
to
t
h
e
co
n
v
e
n
tio
n
al
co
m
p
u
tatio
n
al
s
y
s
te
m
s
s
u
c
h
a
s
co
m
p
u
tatio
n
s
p
e
ed
an
d
r
o
b
u
s
t
n
ess
.
T
h
e
m
o
s
t
p
r
o
f
o
u
n
d
a
n
d
i
m
p
o
r
tan
t
c
h
ar
ac
ter
is
ti
c
o
f
A
N
N
i
s
th
e
ab
ili
t
y
to
m
e
m
o
r
i
z
e
w
h
ic
h
i
n
v
o
l
v
es
a
lar
g
e
n
u
m
b
er
o
f
p
r
o
ce
s
s
o
r
s
o
p
er
atin
g
in
p
ar
allel
.
E
ac
h
p
r
o
ce
s
s
o
r
s
co
n
s
is
t
s
o
f
o
w
n
k
n
o
w
led
g
e
an
d
ca
n
ac
ce
s
s
t
h
e
d
ata
in
its
lo
ca
l
m
e
m
o
r
y
.
T
h
e
ab
ilit
y
o
f
A
NN
to
g
en
er
ali
z
e
o
n
u
n
s
ee
n
d
ata
an
d
lear
n
f
r
o
m
n
o
i
s
y
d
ata
m
ak
e
it
a
v
er
y
p
o
w
er
f
u
l
m
ac
h
i
n
e
lear
n
in
g
al
g
o
r
ith
m
[
1
2
]
.
A
p
ar
t
f
r
o
m
ANN,
th
er
e
i
s
an
o
th
er
i
n
t
ellig
e
n
t
s
y
s
te
m
t
h
a
t
h
as b
ee
n
g
i
v
en
atte
n
tio
n
r
ec
en
tl
y
n
a
m
el
y
S
u
p
p
o
r
t V
ec
to
r
Ma
ch
in
e
(
SVM)
.
T
h
e
ad
v
a
n
ta
g
es
o
f
SVM
i
n
clu
d
i
n
g
f
ast
co
n
v
er
g
e
n
ce
r
ate
,
g
o
o
d
g
en
er
ali
z
atio
n
ca
p
ab
ilit
y
an
d
a
u
to
m
at
ic
d
eter
m
i
n
atio
n
o
f
h
i
d
d
en
n
e
u
r
o
n
s
[
1
3
]
.
T
h
e
r
eg
r
ess
io
n
p
r
o
b
le
m
i
n
SV
M
is
f
o
r
m
u
la
ted
b
ased
o
n
co
n
v
ex
q
u
ad
r
atic
p
r
o
g
r
a
m
m
in
g
p
r
o
b
lem
w
h
er
e
it
u
s
e
lin
ea
r
r
eg
r
es
s
o
r
[
1
4
]
.
I
n
lin
ea
r
r
eg
r
ess
o
r
,
th
e
r
e
g
r
ess
o
r
w
ill
m
ap
s
t
h
e
i
n
p
u
t
s
i
n
to
a
h
i
g
h
er
d
i
m
en
s
io
n
al
f
ea
t
u
r
e
s
p
ac
e
to
m
i
n
i
m
ize
th
e
co
s
t
f
u
n
ctio
n
.
T
h
e
m
aj
o
r
s
h
o
r
tco
m
i
n
g
o
f
S
VM
is
th
e
co
m
p
u
tatio
n
al
b
u
r
d
en
f
o
r
o
p
tim
izatio
n
p
r
o
g
r
a
m
m
i
n
g
co
n
s
tr
ai
n
ed
.
T
h
is
w
ea
k
n
es
s
h
a
s
b
ee
n
o
v
er
co
m
ed
in
L
ea
s
t
-
Sq
u
ar
e
Su
p
p
o
r
t
Vec
to
r
Ma
ch
i
n
es
(
L
S
SVM)
,
w
h
ic
h
s
u
b
s
ti
tu
te
a
q
u
ad
r
atic
p
r
o
g
r
am
m
i
n
g
w
it
h
li
n
ea
r
eq
u
atio
n
s
[
1
5
-
1
8
]
.
T
h
e
ac
cu
r
ac
y
o
f
th
e
p
r
ed
ictio
n
o
f
L
S
SVM
d
ep
en
d
s
o
n
t
h
e
h
y
p
er
p
ar
am
e
ter
s
v
al
u
e
s
etti
n
g
.
T
h
er
ef
o
r
e,
Gr
e
y
W
o
lf
Op
ti
m
izer
-
L
ea
s
t
Sq
u
ar
e
Su
p
p
o
r
t
Vec
to
r
Ma
ch
in
e(
G
W
O
-
L
SS
VM
)
i
s
p
r
o
p
o
s
ed
to
s
o
lv
e
lo
n
g
-
ter
m
lo
ad
f
o
r
ec
asti
n
g
p
r
o
b
lem
.
GW
O
i
s
in
te
g
r
ated
w
it
h
L
SS
VM
f
o
r
d
eter
m
i
n
in
g
th
e
o
p
ti
m
al
h
y
p
er
p
ar
a
m
eter
s
.
GW
O
i
s
in
s
p
ir
ed
b
y
h
ier
ar
ch
y
s
y
s
te
m
an
d
h
u
n
ti
n
g
m
ec
h
a
n
is
m
o
f
g
r
e
y
w
o
l
v
es
i
n
n
at
u
r
e.
GW
O
h
as
b
ee
n
e
f
f
ec
ti
v
el
y
i
m
p
le
m
en
ted
f
o
r
s
o
lv
i
n
g
v
ar
io
u
s
p
o
w
er
s
y
s
te
m
p
r
o
b
le
m
s
[
1
9
-
2
3
]
.
2.
RE
S
E
ARCH
M
E
T
H
O
D
I
n
o
r
d
er
to
p
r
ed
ict
to
tal
lo
ad
d
e
m
an
d
,
a
to
tal
o
f
1
0
3
5
s
ets
o
f
d
ata
ar
e
co
llected
in
th
e
ar
ea
o
f
Da
y
to
n
,
Oh
io
,
US
f
r
o
m
J
an
u
ar
y
2
0
1
6
u
n
t
il
Ma
y
2
0
1
8
.
Fo
r
tr
ain
i
n
g
,
668
d
ata
s
ets
ar
e
u
s
ed
a
n
d
th
e
r
e
m
ain
in
g
367
d
ata
s
ets
ar
e
u
s
ed
f
o
r
te
s
ti
n
g
.
Mo
r
eo
v
er
,
th
e
te
m
p
er
at
u
r
e,
w
i
n
d
s
p
ee
d
an
d
h
u
m
id
it
y
ar
e
t
h
e
in
p
u
ts
,
w
h
ile
th
e
to
tal
o
u
tp
u
t
p
o
w
er
i
s
th
e
o
u
tp
u
t
o
f
th
e
m
o
d
el.
Fi
g
u
r
e
1
p
r
esen
ts
o
v
er
all
d
ata
u
s
ed
in
t
h
i
s
p
ap
er
.
I
t
ca
n
b
e
o
b
s
er
v
ed
th
at
t
h
e
to
t
al
lo
ad
d
em
a
n
d
f
lu
ct
u
ated
.
T
h
er
ef
o
r
e,
an
ac
cu
r
ate
p
r
ed
ictio
n
tech
n
iq
u
e
i
s
r
eq
u
ir
ed
to
en
s
u
r
e
th
e
elec
tr
icit
y
g
e
n
er
ated
ca
p
ac
it
y
m
ee
t
elec
tr
icit
y
d
e
m
a
n
d
at
all
ti
m
e
s
.
I
n
t
h
is
p
ap
er
,
GW
O
-
L
SS
V
M
is
d
ev
elo
p
ed
to
p
r
ed
ict
th
e
to
tal
lo
ad
d
em
a
n
d
i
n
lo
n
g
-
ter
m
d
u
r
atio
n
.
T
h
e
ef
f
icac
y
o
f
t
h
e
p
r
o
p
o
s
ed
alg
o
r
ith
m
i
s
ev
alu
a
ted
th
r
o
u
g
h
co
m
p
ar
i
s
o
n
w
it
h
o
th
er
tec
h
n
iq
u
es
s
u
c
h
as
L
S
SVM
an
d
An
t
L
io
n
Op
ti
m
izer
-
L
ea
s
t
-
Sq
u
ar
e
Su
p
p
o
r
t V
ec
to
r
Ma
ch
in
e
(
AL
O
-
L
S
SVM)
.
Fig
u
r
e
1
.
T
o
tal
lo
ad
d
em
a
n
d
d
ata
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
Lo
n
g
-
term lo
a
d
fo
r
ec
a
s
tin
g
u
s
in
g
g
r
ey
w
o
lf o
p
timiz
er
-
lea
s
t
-
s
q
u
a
r
es
...
(
Z.
M.
Ya
s
in
)
419
2
.
1
.
P
re
dict
io
n
m
ea
s
ure
m
en
t
Me
an
A
b
s
o
l
u
te
P
er
ce
n
ta
g
e
E
r
r
o
r
(
MA
P
E
)
an
d
co
ef
f
icie
n
t
o
f
d
eter
m
in
at
io
n
(
R
2
)
w
er
e
u
s
e
d
f
o
r
th
e
ev
alu
a
tio
n
p
u
r
p
o
s
es
a
s
s
h
o
w
n
in
(
1
)
an
d
(
2
)
.
B
o
th
o
f
t
h
e
m
ac
t
as
i
n
d
icato
r
s
w
h
ic
h
a
s
s
e
s
s
th
e
p
er
f
o
r
m
a
n
ce
o
f
s
u
g
g
e
s
ted
tec
h
n
iq
u
e.
M
A
P
E
i
s
a
m
ea
s
u
r
e
o
f
p
r
ed
ictio
n
p
r
ec
is
io
n
o
f
a
f
o
r
ec
asti
n
g
tech
n
iq
u
e
in
m
ea
s
u
r
e
m
en
ts
.
I
t
u
s
u
all
y
u
s
es
p
er
ce
n
tag
e
to
s
h
o
w
s
t
h
e
ac
c
u
r
ac
y
.
Fo
r
M
A
P
E
,
th
e
lo
w
er
th
e
v
a
l
u
e,
t
h
e
b
et
ter
th
e
p
er
f
o
r
m
a
n
ce
.
I
n
th
e
s
i
m
u
latio
n
,
M
A
P
E
is
u
t
ili
ze
d
as o
b
j
ec
tiv
e
f
u
n
ctio
n
o
f
GW
O
-
L
SS
VM
.
T
h
e
co
ef
f
ic
ie
n
t o
f
d
eter
m
i
n
atio
n
,
R
2
is
a
p
ar
t
o
f
th
e
o
u
tp
u
t
v
ar
ia
b
le
v
ar
ian
ce
th
a
t
ca
n
b
e
p
r
ed
icted
f
r
o
m
th
e
i
n
p
u
t
v
ar
iab
le.
I
t
g
iv
e
s
a
m
ea
s
u
r
e
o
f
h
o
w
s
i
m
i
lar
t
h
e
o
u
tco
m
es
ar
e
r
ep
licated
b
y
t
h
e
m
o
d
el,
i
n
v
ie
w
o
f
th
e
ex
ten
t
o
f
to
tal
v
ar
iatio
n
o
f
r
e
s
u
l
ts
clar
if
ied
b
y
th
e
m
o
d
el.
Ho
w
e
v
er
,
f
o
r
R
2
,
th
e
clo
s
er
th
e
v
alu
e
to
1
,
th
e
b
etter
it
w
ill b
e.
=
∑
|
−
|
|
|
=
1
×
100%
(
1
)
2
=
1
−
∑
(
−
)
2
∑
(
−
̅
)
2
(
2
)
W
h
er
e
t=1
,
2
,
….
,
t
A
t
=
A
ct
u
al
v
alu
e
s
F
t
=
P
r
ed
icted
v
alu
es/
Fo
r
ec
asti
n
g
v
al
u
es
2
.
2
.
Dev
el
o
p
m
ent
o
f
G
WO
-
L
SS
VM
GW
O
is
a
m
eta
h
e
u
r
ictic
-
b
a
s
ed
tech
n
iq
u
e
s
ti
m
u
lated
b
y
th
e
h
u
n
ti
n
g
b
eh
av
io
r
an
d
lead
er
s
h
ip
h
ier
ar
ch
y
o
f
g
r
e
y
w
o
l
v
es
a
s
d
ep
icted
in
Fig
u
r
e
2
.
A
lp
h
a
(
α
)
is
th
e
lead
er
o
f
th
e
w
o
l
v
es.
I
t
is
r
esp
o
n
s
ib
le
to
m
ak
e
a
s
ettle
m
e
n
t
w
h
ic
h
ar
e
r
elate
s
to
h
u
n
ti
n
g
,
w
h
e
n
an
d
wh
er
e
to
s
leep
,
an
d
etc
[
2
4
]
.
B
e
ta
(
β)
is
t
h
e
s
ec
o
n
d
lev
el
w
h
ic
h
is
t
h
e
as
s
is
tan
t
o
f
α
in
m
a
k
in
g
d
ec
is
io
n
.
β
i
s
lik
e
w
i
s
e
t
h
e
b
est
ap
p
lican
t
to
r
ep
lace
α
w
h
e
n
α
g
etti
n
g
o
ld
o
r
d
ie.
T
h
e
lo
w
e
s
t
in
g
r
e
y
w
o
l
v
es
’
s
y
s
te
m
h
ier
ar
ch
y
is
ca
l
led
O
m
eg
a
(
ω
)
.
I
t
ac
ts
as
a
s
ca
p
eg
o
at.
ω
ca
n
f
u
lf
ill
th
e
en
tire
w
h
o
le
g
r
e
y
w
o
l
v
es
g
r
o
u
p
.
T
h
e
th
ir
d
lev
el
in
g
r
e
y
w
o
l
v
es
’
s
y
s
te
m
h
i
er
ar
ch
y
is
Delta
(
δ)
w
h
ic
h
m
u
s
t
r
ep
o
r
t
to
α
an
d
β,
h
o
w
ev
er
t
h
e
y
lead
th
e
ω
.
δ
r
esp
o
n
s
ib
ilit
y
i
s
s
co
u
tin
g
to
s
ec
u
r
e
an
d
en
s
u
r
e
th
e
s
ec
u
r
it
y
o
f
t
h
e
g
r
o
u
p
o
f
w
o
l
v
es.
T
h
e
s
y
s
te
m
h
ier
a
r
ch
y
o
f
g
r
e
y
w
o
lv
e
s
i
s
an
a
s
s
i
g
n
ed
c
h
ar
ac
ter
is
ti
c
u
s
ed
i
n
GW
O.
A
n
o
t
h
er
a
m
az
in
g
c
h
ar
ac
ter
is
tic
o
f
w
o
l
v
es
i
s
t
h
e
g
r
o
u
p
h
u
n
ti
n
g
s
tr
ate
g
y
.
T
h
e
f
itte
s
t
s
o
lu
tio
n
is
co
n
s
id
er
ed
as
th
e
α
.
Me
an
w
h
ile,
th
e
s
ec
o
n
d
an
d
t
h
ir
d
b
est
s
o
lu
tio
n
s
ar
e
co
n
s
id
er
ed
as
β
an
d
d
elta
δ
r
esp
ec
tiv
el
y
.
T
h
e
r
est
o
f
th
e
in
d
i
v
id
u
al
s
o
l
u
tio
n
s
ar
e
co
n
s
id
er
ed
as
ω
.
Du
r
i
n
g
h
u
n
ti
n
g
,
t
h
e
w
o
l
v
es
ten
d
to
en
cir
cle
th
eir
p
r
e
y
.
T
h
e
en
ci
r
clin
g
b
e
h
av
io
u
r
i
s
r
ep
r
esen
te
d
as (
3
)
.
⃗
⃗
=
|
.
⃗
⃗
⃗
⃗
−
(
)
(
+
1
)
=
⃗
⃗
⃗
⃗
(
)
−
.
⃗
⃗
(
3
)
=
2
⃗
.
1
⃗
⃗
⃗
−
=
2
.
2
⃗
⃗
⃗
W
h
er
e
t is th
e
o
n
g
o
i
n
g
iter
atio
n
an
d
ar
e
co
ef
f
icie
n
t v
ec
to
r
s
⃗
⃗
⃗
⃗
is
th
e
p
o
s
itio
n
v
ec
to
r
o
f
th
e
p
r
e
y
,
,
an
d
is
th
e
p
o
s
itio
n
o
f
th
e
s
ea
r
ch
ag
en
t
.
T
h
e
b
est
th
r
ee
s
o
lu
tio
n
s
;
,
,
an
d
ar
e
s
av
ed
in
cl
u
d
in
g
th
e
lates
t
p
o
s
itio
n
s
o
f
ω
ac
co
r
d
in
g
to
th
e
cu
r
r
en
t
b
est
p
o
s
itio
n
.
is
th
e
b
est
p
o
s
itio
n
o
f
α
,
is
th
e
b
est
p
o
s
itio
n
o
f
β,
an
d
is
th
e
b
est
p
o
s
itio
n
o
f
δ.
T
h
e
f
in
a
l
p
o
s
itio
n
X
(
t
+1
)
,
is
d
ef
in
ed
b
y
t
h
e
p
o
s
itio
n
s
o
f
α
,
β,
an
d
δ in
t
h
e
s
ea
r
ch
s
p
ac
e
as sh
o
w
n
in
(
4
)
.
is
th
e
p
o
s
itio
n
o
f
th
e
g
r
e
y
w
o
l
f
⃗
⃗
⃗
⃗
⃗
=
|
1
⃗
⃗
⃗
⃗
.
⃗
⃗
⃗
⃗
−
|
⃗
⃗
⃗
⃗
=
|
3
⃗
⃗
⃗
⃗
.
⃗
⃗
⃗
⃗
−
|
1
⃗
⃗
⃗
⃗
=
⃗
⃗
⃗
⃗
−
1
⃗
⃗
⃗
⃗
.
(
⃗
⃗
⃗
⃗
⃗
)
(
4
)
2
⃗
⃗
⃗
⃗
=
⃗
⃗
⃗
⃗
−
2
⃗
⃗
⃗
⃗
.
(
⃗
⃗
⃗
⃗
)
3
⃗
⃗
⃗
⃗
=
⃗
⃗
⃗
⃗
−
3
⃗
⃗
⃗
⃗
.
(
⃗
⃗
⃗
⃗
)
(
+
1
)
=
1
⃗
⃗
⃗
⃗
⃗
+
2
⃗
⃗
⃗
⃗
⃗
+
3
⃗
⃗
⃗
⃗
⃗
3
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
1
7
–
4
23
420
W
h
er
e
⃗
⃗
is
th
e
co
ef
f
icie
n
t
v
ec
to
r
I
n
L
S
SVM,
th
er
e
ar
e
t
w
o
tu
n
i
n
g
p
ar
a
m
eter
s
th
at
a
s
s
o
ciate
d
w
i
th
R
B
F
k
er
n
el
,
σ
2
an
d
r
eg
u
lar
izatio
n
p
ar
a
m
e
ter
,
γ
.
T
h
ese
p
ar
a
m
eter
s
af
f
ec
t
th
e
L
SS
VM
est
i
m
a
tio
n
m
o
d
el
's
ac
c
u
r
ac
y
.
T
h
ese
p
ar
am
eter
s
ar
e
o
p
ti
m
iz
ed
u
s
in
g
GW
O
-
L
SS
VM
to
m
in
i
m
ize
M
A
P
E
.
T
h
e
f
lo
w
c
h
ar
t
o
f
GW
O
-
L
S
SVM
is
p
r
esen
ted
in
Fig
u
r
e
3
.
Fig
u
r
e
2
.
T
h
e
s
y
s
te
m
h
ier
ar
ch
y
o
f
g
r
e
y
w
o
lv
e
s
Fig
u
r
e
3
.
Flo
w
c
h
ar
t o
f
GW
O
-
L
S
SVM
Du
r
in
g
in
i
tializatio
n
,
t
w
o
p
ar
a
m
eter
s
w
h
ich
ar
e
g
a
m
m
a
a
n
d
s
ig
m
a
w
er
e
r
an
d
o
m
l
y
g
e
n
er
at
ed
w
i
th
i
n
lo
w
er
b
o
u
n
d
a
n
d
u
p
p
er
b
o
u
n
d
.
T
h
e
p
ar
am
eter
s
s
e
t
f
o
r
G
W
O
-
L
SS
VM
al
g
o
r
ith
m
ar
e
t
ab
u
lated
i
n
T
ab
le
1
.
T
h
e
n
u
m
b
er
o
f
s
ea
r
c
h
ag
e
n
t
s
is
s
et
to
1
0
.
I
f
th
e
n
u
m
b
er
o
f
s
ea
r
ch
ag
e
n
t
s
is
h
ig
h
,
it
w
ill
c
o
n
s
u
m
e
m
o
r
e
ti
m
e
.
Ho
w
e
v
er
,
s
m
a
ll n
u
m
b
er
o
f
s
ea
r
ch
ag
e
n
ts
m
a
y
n
o
t g
i
v
e
t
h
e
b
est p
r
ed
ictio
n
.
T
ab
le
1
.
P
ar
am
eter
s
o
f
GW
O
P
a
r
a
me
t
e
r
V
a
l
u
e
N
u
mb
e
r
o
f
se
a
r
c
h
a
g
e
n
t
s
10
M
a
x
i
m
u
m
i
t
e
r
a
t
i
o
n
30
U
p
p
e
r
b
o
u
n
d
1
0
0
0
0
0
0
L
o
w
e
r
b
o
u
n
d
0
.
1
T
h
e
p
o
s
itio
n
o
f
g
r
e
y
w
o
l
f
is
th
en
u
p
d
ated
ac
co
r
d
in
g
to
(
4
)
.
α
,
β,
an
d
δ
w
o
l
v
es
d
eter
m
i
n
e
th
e
p
o
s
s
ib
le
p
o
s
itio
n
o
f
t
h
e
p
r
e
y
.
T
h
is
p
r
o
ce
s
s
co
n
tin
u
e
u
n
ti
l
th
e
co
n
v
er
g
e
n
ce
g
o
al
is
m
et
an
d
is
tak
en
a
s
t
h
e
b
est
s
o
lu
t
io
n
.
I
n
th
i
s
al
g
o
r
ith
m
,
M
A
P
E
is
s
et
as
t
h
e
f
it
n
es
s
f
u
n
ct
io
n
.
M
A
P
E
v
al
u
e
i
s
o
b
tain
ed
b
y
ca
lli
n
g
t
h
e
L
S
SVM
alg
o
r
it
h
m
.
T
he
p
o
s
itio
n
o
f
s
ea
r
ch
ag
e
n
ts
ar
e
u
p
d
ated
ac
co
r
d
in
g
to
α
,
β,
an
d
δ
in
s
ea
r
ch
s
p
ac
e
as
s
h
o
w
n
i
n
(
4
)
.
T
h
e
p
r
o
ce
s
s
is
r
ep
ea
ted
u
n
ti
l
t
h
e
s
o
lu
tio
n
co
n
v
er
g
ed
.
T
h
e
s
o
l
u
tio
n
w
ill
co
n
v
er
g
e
if
t
h
e
d
if
f
er
e
n
t
b
et
w
ee
n
m
a
x
i
m
u
m
a
n
d
m
i
n
i
m
u
m
f
it
n
es
s
r
ea
ch
1
x
1
0
-
7
.
I
n
th
is
p
ap
er
,
th
e
r
esu
l
ts
o
f
G
W
O
-
L
SS
VM
i
s
also
co
m
p
ar
ed
to
AL
O
-
L
SS
VM
in
ter
m
s
o
f
ac
cu
r
ac
y
an
d
co
n
v
er
g
en
ce
s
p
ee
d
.
I
n
AL
O
-
L
SS
VM
,
AL
O
i
s
u
tili
ze
d
to
d
eter
m
in
e
t
h
e
o
p
ti
m
al
p
ar
a
m
eter
s
o
f
g
a
m
m
a
a
n
d
s
ig
m
a
.
T
h
e
alg
o
r
it
h
m
r
ep
r
esen
t
s
th
e
h
u
n
t
in
g
b
eh
a
v
io
r
o
f
an
tlio
n
s
i
n
n
atu
r
e.
T
h
er
e
ar
e
f
i
v
e
m
ain
s
tep
s
o
f
t
h
e
al
g
o
r
ith
m
in
c
lu
d
i
n
g
r
an
d
o
m
w
a
lk
o
f
a
n
ts
,
co
n
s
tr
u
c
tin
g
tr
ap
s
,
en
tr
ap
m
e
n
t
o
f
a
n
ts
i
n
tr
ap
s
b
u
ilt
b
y
a
n
tlio
n
s
,
ca
tc
h
i
n
g
an
ts
,
a
n
d
r
e
-
s
et
u
p
t
h
e
tr
ap
s
[
2
5
]
.
T
h
e
n
u
m
b
er
o
f
s
er
ac
h
ag
e
n
t
s
,
m
a
x
i
m
u
m
iter
at
io
n
s
,
lo
w
er
an
d
u
p
p
er
b
o
u
n
d
ar
e
s
et
s
i
m
ilar
w
ith
GW
O
-
L
S
S
VM
.
3.
RE
SU
L
T
S
A
ND
D
I
SCU
SS
I
O
N
Fig
u
r
e
4
p
r
ese
n
ts
th
e
r
es
u
lts
o
f
GW
O
-
L
S
SVM
f
o
r
tr
ai
n
i
n
g
d
ata.
T
h
e
r
ed
li
n
e
is
d
ef
i
n
ed
as
th
e
esti
m
atio
n
o
r
p
r
ed
icted
o
u
tp
u
t
,
w
h
i
le
th
e
b
lu
e
d
o
ts
r
ep
r
esen
t
ed
as
th
e
ac
tu
al
o
u
tp
u
t
.
T
h
e
o
p
ti
m
al
g
a
m
m
a
an
d
s
ig
m
a
v
al
u
e
ar
e
1
0
6
7
8
4
.
3
3
5
a
n
d
0
.
8
5
3
8
1
r
esp
ec
tiv
el
y
.
T
h
e
co
m
p
ar
is
o
n
b
et
w
ee
n
p
r
ed
icted
an
d
ac
tu
al
d
ata
f
o
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
Lo
n
g
-
term lo
a
d
fo
r
ec
a
s
tin
g
u
s
in
g
g
r
ey
w
o
lf o
p
timiz
er
-
lea
s
t
-
s
q
u
a
r
es
...
(
Z.
M.
Ya
s
in
)
421
tr
ain
i
n
g
an
d
test
i
n
g
p
r
o
ce
s
s
u
s
in
g
GW
O
-
L
S
SVM
ar
e
s
h
o
wn
i
n
Fi
g
u
r
e
5
an
d
Fi
g
u
r
e
6
r
es
p
ec
tiv
el
y
.
I
t
ca
n
b
e
o
b
s
er
v
ed
th
at
th
e
p
r
ed
icted
r
esu
lts
al
m
o
s
t
s
i
m
ilar
w
it
h
ac
tu
a
l
d
ata
f
o
r
b
o
th
tr
ain
in
g
an
d
test
in
g
p
r
o
ce
s
s
.
Fig
u
r
e
4
.
R
esu
lts
o
f
GW
O
-
L
S
SVM
Fig
u
r
e
5
.
C
o
m
p
ar
is
o
n
o
f
tr
ai
n
i
n
g
d
ata
an
d
es
ti
m
atio
n
o
u
tp
u
t
f
o
r
GW
O
-
L
SS
VM
Fig
u
r
e
6
.
C
o
m
p
ar
is
o
n
o
f
te
s
ti
n
g
d
ata
an
d
p
r
ed
icted
o
u
tp
u
t f
o
r
GW
O
-
L
SS
VM
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
1
7
–
4
23
422
C
o
m
p
ar
is
o
n
b
et
w
ee
n
GW
O
-
L
SS
VM
,
AL
O
-
L
SS
VM
a
n
d
L
SS
VM
i
s
tab
u
lated
i
n
T
a
b
le
2
.
B
y
r
u
n
n
i
n
g
GW
O
-
L
S
SVM,
th
e
v
al
u
es
o
f
o
p
ti
m
al
p
a
r
a
m
eter
s
o
f
γ
a
n
d
σ
2
ar
e
1
0
6
7
8
4
.
3
3
5
an
d
0
.
8
5
3
8
1
r
esp
ec
tiv
el
y
.
A
s
t
h
e
m
a
x
i
m
u
m
n
u
m
b
er
o
f
iter
ati
o
n
s
r
ea
ch
e
s
3
0
,
MA
P
E
an
d
R
2
ar
e
co
m
p
ar
ed
.
By
u
s
i
n
g
L
S
SVM
to
m
o
d
el
th
e
to
tal
o
u
tp
u
t
p
o
w
er
,
th
e
v
al
u
es
o
f
M
A
P
E
w
as
0.
2570
%
.
Ho
w
e
v
er
,
b
y
u
s
i
n
g
AL
O
-
L
S
SVM,
it
h
e
lp
s
i
n
r
ed
u
cin
g
th
e
M
A
P
E
v
al
u
es
to
0
.
1860
%.
Fo
r
GW
O
-
L
S
SVM,
t
h
e
M
A
P
E
is
m
u
c
h
b
etter
th
an
th
e
o
t
h
er
t
w
o
m
e
th
o
d
s
w
h
ic
h
is
0.
1265
%.
Fo
r
th
e
co
ef
f
ic
ien
t
o
f
d
eter
m
i
n
atio
n
,
R
2
,
L
SS
V
M
g
i
v
e
s
th
e
v
al
u
e
o
f
0
.
9
9
8
9
.
Ho
w
ev
er
,
th
e
o
p
ti
m
al
v
al
u
es
o
f
p
ar
a
m
eter
s
γ
an
d
σ
2
b
y
u
s
in
g
h
y
b
r
id
tech
n
iq
u
e
s
GW
O
-
L
SS
VM
an
d
AL
O
-
L
S
SVM
,
th
e
R
2
i
n
cr
ea
s
ed
to
0
.
9
998
an
d
0
.
9
9
9
1
r
esp
ec
tiv
ely
.
T
h
u
s
,
it
ca
n
b
e
co
n
clu
d
ed
th
at
GW
O
-
L
S
S
VM
p
r
o
v
id
e
m
o
r
e
ac
cu
r
ate
p
r
ed
ictio
n
f
o
r
lo
n
g
-
ter
m
lo
ad
f
o
r
ec
asti
n
g
as
co
m
p
ar
ed
to
o
th
er
tech
n
i
q
u
es.
T
h
e
co
m
p
ar
is
o
n
o
f
ite
r
ativ
e
co
n
v
er
g
e
n
ce
b
et
w
ee
n
GW
O
-
L
SS
VM
an
d
AL
O
-
L
S
SVM
i
s
p
r
esen
ted
i
n
Fi
g
u
r
e
7
.
GW
O
-
L
S
SVM
co
n
v
er
g
ed
at
7
th
iter
atio
n
wh
ile
AL
O
-
L
S
SV
M
co
n
v
er
g
ed
at
1
2
th
iter
atio
n
.
I
t
ca
n
b
e
co
n
cl
u
d
ed
th
at
GW
O
-
L
S
SVM
h
as
a
p
r
ec
is
e
p
r
ed
ic
ti
o
n
an
d
t
h
e
s
p
ee
d
o
f
o
p
tim
izatio
n
i
s
f
a
s
ter
.
T
ab
le
2
.
C
o
m
p
ar
is
o
n
b
et
w
ee
n
L
S
SVM,
GW
O
-
L
SS
VM
a
n
d
AL
O
-
L
S
SVM
G
W
O
-
L
S
S
V
M
A
L
O
-
L
S
S
V
M
LS
-
S
V
M
G
a
mm
a
(
γ
)
1
0
6
7
8
4
.
3
3
5
1
0
7
9
2
8
6
.
5
4
5
3
6
1
9
2
6
3
5
.
8
6
9
6
S
i
g
ma
(
σ
2
)
0
.
8
5
3
8
1
1
8
1
.
7
4
8
7
3
1
1
5
6
.
7
8
6
9
M
A
P
E
0
.
1
2
6
5
0
.
1
8
6
0
0
.
2
5
7
0
R
2
0
.
9
9
9
8
0
.
9
9
9
1
0
.
9
9
8
9
Fig
u
r
e
7
.
C
o
n
v
er
g
en
ce
g
r
ap
h
f
o
r
GW
O
-
L
S
SVM
an
d
AL
O
-
L
SS
VM
4.
CO
NCLU
SI
O
N
A
h
y
b
r
id
p
r
ed
ictio
n
tec
h
n
iq
u
e
n
a
m
el
y
Gr
e
y
W
o
lf
Op
ti
m
izer
–
L
ea
s
t
Sq
u
ar
e
S
u
p
p
o
r
t
Vec
to
r
Ma
ch
i
n
e
(
GW
O
-
L
S
SVM
)
h
a
d
b
ee
n
d
ev
elo
p
to
f
o
r
ec
ast
lo
n
g
-
ter
m
lo
ad
d
e
m
an
d
,
w
it
h
t
h
e
o
b
j
ec
tiv
e
f
u
n
c
tio
n
to
m
i
n
i
m
ize
th
e
er
r
o
r
.
L
S
S
VM
is
o
n
e
o
f
th
e
g
o
o
d
tec
h
n
iq
u
es
to
s
o
l
v
e
n
o
n
lin
ea
r
p
r
o
b
lem
s
.
Ho
w
e
v
er
,
th
e
ac
cu
r
ac
y
o
f
t
h
e
p
r
ed
ictio
n
d
ep
en
d
s
o
n
th
e
s
elec
tio
n
o
f
th
e
R
B
F
p
ar
am
e
ter
s
.
GW
O
h
as
b
ee
n
u
t
ilized
to
o
p
tim
ized
th
e
v
al
u
e
o
f
t
h
ese
p
ar
a
m
eter
s
.
GW
O
m
i
m
ics
th
e
h
ier
ar
ch
y
o
f
lead
er
s
h
ip
an
d
h
u
n
tin
g
m
ec
h
an
i
s
m
o
f
g
r
e
y
w
o
l
f
i
n
n
at
u
r
e.
Fo
u
r
i
n
p
u
ts
w
er
e
co
n
s
id
er
ed
f
o
r
th
e
p
r
e
d
ictio
n
;
p
ea
k
lo
ad
d
e
m
a
n
d
,
te
m
p
er
at
u
r
e,
h
u
m
id
it
y
an
d
w
in
d
s
p
ee
d
.
A
f
ter
co
n
d
u
ctin
g
th
e
s
i
m
u
latio
n
f
o
r
L
SS
VM
,
GW
O
-
L
SS
VM
an
d
AL
O
-
L
S
SVM,
it
ca
n
b
e
co
n
clu
d
ed
th
at
GW
O
-
L
SS
V
M
h
as
b
etter
p
r
ed
ictio
n
ac
cu
r
ac
y
an
d
h
i
g
h
er
o
p
ti
m
izatio
n
s
p
ee
d
to
p
r
e
d
ict
th
e
lo
n
g
-
ter
m
lo
ad
f
o
r
ec
asti
n
g
as i
t.
ACK
NO
WL
E
D
G
E
M
E
NT
S
T
h
is
r
esear
ch
is
f
u
ll
y
s
u
p
p
o
r
ted
b
y
Un
i
v
er
s
it
i
T
ek
n
o
lo
g
i
MA
R
A
Ma
la
y
s
ia
an
d
th
e
M
in
is
tr
y
o
f
Hig
h
er
E
d
u
ca
tio
n
(
MO
HE
)
M
ala
y
s
ia
t
h
r
o
u
g
h
r
esear
c
h
g
r
a
n
t
6
0
0
-
I
R
MI
/FR
G
S 5
/3
(
1
5
7
/
2
0
1
9
)
.
RE
F
E
R
E
NC
E
S
[1
]
A
.
M
.
A
l
-
S
h
a
a
lan
,
“
Esse
n
ti
a
l
a
sp
e
c
ts
o
f
p
o
w
e
r
s
y
ste
m
p
lan
n
in
g
in
d
e
v
e
lo
p
i
n
g
c
o
u
n
tri
e
s,”
J
u
rn
a
l
o
f
Kin
g
S
a
u
d
Un
ive
rs
it
y
,
v
o
l.
2
3
,
n
o
.
1
,
p
p
.
2
7
-
32,
2
0
1
1
.
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
Lo
n
g
-
term lo
a
d
fo
r
ec
a
s
tin
g
u
s
in
g
g
r
ey
w
o
lf o
p
timiz
er
-
lea
s
t
-
s
q
u
a
r
es
...
(
Z.
M.
Ya
s
in
)
423
[2
]
Zi
z
i
Zh
a
n
g
,
Ho
n
g
L
i,
Y
a
n
g
Zh
a
o
,
a
n
d
X
iao
b
o
Hu
,
“
S
h
o
rt
-
term
lo
a
d
f
o
re
c
a
stin
g
b
a
se
d
o
n
th
e
g
rid
m
e
th
o
d
a
n
d
th
e
ti
m
e
s
e
ries
f
u
z
z
y
lo
a
d
f
o
re
c
a
stin
g
m
e
th
o
d
,
”
in
I
n
t.
C
o
n
fer
e
n
c
e
o
n
R
e
n
e
wa
b
le P
o
we
r Ge
n
e
ra
ti
o
n
(
RP
G 2
0
1
5
)
,
2
0
1
5
.
[3
]
T
.
Ho
n
g
a
n
d
S
.
F
a
n
,
“
P
ro
b
a
b
i
li
st
ic
e
lec
tri
c
lo
a
d
f
o
re
c
a
stin
g
:
A
tu
to
rial
re
v
iew
,
”
In
t.
J
o
u
rn
a
l
o
f
F
o
re
c
a
st.
,
v
o
l.
3
2
,
n
o
.
3
,
p
p
.
9
1
4
-
9
3
8
,
2
0
1
6
.
[4
]
T
.
S
a
k
so
rn
c
h
a
i,
W
.
J.
L
e
e
,
K.
M
e
th
a
p
ra
y
o
o
n
,
J.
R
.
L
iao
,
a
n
d
R.
J.
Ro
s
s,
“
Im
p
ro
v
e
th
e
u
n
it
c
o
m
m
it
m
e
n
t
sc
h
e
d
u
li
n
g
b
y
u
sin
g
th
e
n
e
u
ra
l
-
n
e
tw
o
rk
-
b
a
s
e
d
sh
o
rt
-
term
lo
a
d
f
o
re
c
a
stin
g
,
”
IEE
E
T
ra
n
s.
In
d
u
stri
a
l
Ap
p
li
c
a
t
io
n
s
,
v
o
l
4
1
,
n
o
.
1
,
p
p
:
1
6
9
-
1
7
9
,
2
0
0
5
.
[5
]
S
.
W
a
n
g
,
S
.
W
a
n
g
,
D.
W
a
n
g
,
"
C
o
m
b
in
e
d
p
ro
b
a
b
il
it
y
d
e
n
sit
y
m
o
d
e
l
f
o
r
m
e
d
iu
m
ter
m
lo
a
d
f
o
re
c
a
stin
g
b
a
se
d
o
n
q
u
a
n
t
il
e
re
g
re
ss
io
n
a
n
d
k
e
rn
e
l
d
e
n
sity
e
sti
m
a
ti
o
n
"
,
in
En
e
rg
y
Pro
c
e
d
ia
,
v
o
l.
1
5
8
,
n
o
.
1
,
p
p
.
6
4
4
6
-
6
4
5
1
,
2
0
1
9
.
[6
]
J.
M
o
ra
l
-
Ca
rc
e
d
o
a
n
d
J.
P
é
re
z
-
Ga
rc
ía,
“
In
teg
ra
ti
n
g
lo
n
g
-
term
e
c
o
n
o
m
ic
sc
e
n
a
rio
s
in
to
p
e
a
k
lo
a
d
f
o
re
c
a
stin
g
:
A
n
a
p
p
li
c
a
ti
o
n
t
o
S
p
a
in
,
”
E
n
e
rg
y
,
v
o
l
.
1
4
0
,
n
o
.
1
,
p
p
.
6
8
2
-
6
9
5
,
2
0
1
7
.
[7
]
S
.
M
.
Ba
d
ra
n
,
“
Ne
u
ra
l
n
e
tw
o
rk
in
teg
ra
ted
w
it
h
re
g
re
ss
io
n
m
e
th
o
d
s
t
o
f
o
re
c
a
st
e
lec
tri
c
a
l
lo
a
d
,
”
in
IE
T
In
t
.
Co
n
fer
e
n
c
e
o
n
De
v
e
lo
p
me
n
ts i
n
Po
we
r S
y
ste
ms
Pro
tec
ti
o
n
,
2
0
1
2
.
[8
]
K.
B.
L
in
d
b
e
rg
,
P
.
S
e
lj
o
m
,
H.
M
a
d
se
n
,
D.
F
isc
h
e
r,
a
n
d
M
.
K
o
rp
å
s,
“
L
o
n
g
-
ter
m
e
le
c
tri
c
it
y
lo
a
d
f
o
re
c
a
stin
g
:
Cu
rre
n
t
a
n
d
f
u
tu
re
tren
d
s,”
Util
i
ti
e
s P
o
l
icy
,
v
o
l.
5
8
,
n
o
.
1
,
p
p
.
1
0
2
-
1
1
9
,
2
0
1
9
.
[9
]
K.
Kh
a
li
d
,
A
.
M
o
h
a
m
e
d
,
R.
M
o
h
a
m
e
d
,
H.
S
h
a
re
e
f
,
"
P
e
rf
o
r
m
a
n
c
e
c
o
m
p
a
riso
n
o
f
A
rti
f
icia
l
In
telli
g
e
n
c
e
tec
h
n
iq
u
e
s
f
o
r
n
o
n
-
i
n
tru
siv
e
e
lec
tri
c
lo
a
d
m
o
n
it
o
ri
n
g
"
,
in
Bu
ll
e
ti
n
o
f
El
e
c
trica
l
En
g
i
n
e
e
rin
g
a
n
d
In
fo
rm
a
ti
c
s
,
v
o
l.
7
,
n
o
.
2
,
p
p
.
143
-
1
5
2
,
2
0
1
8
.
[1
0
]
D.
A
li
,
M
.
Yo
h
a
n
n
a
,
M
.
I.
P
u
w
u
,
a
n
d
B.
M
.
G
a
rk
id
a
,
“
L
o
n
g
-
te
rm
lo
a
d
f
o
re
c
a
st
m
o
d
e
ll
in
g
u
sin
g
a
f
u
z
z
y
lo
g
ic
a
p
p
ro
a
c
h
,
”
in
P
a
c
if
ic
S
c
ien
c
e
Rev
iew A
:
Na
t
u
ra
l
S
c
ien
c
e
a
n
d
En
g
in
e
e
rin
g
,
v
o
l.
1
8
,
n
o
.
2
,
p
p
.
1
2
3
-
1
2
7
,
2
0
1
6
.
[1
1
]
D.
A
li
,
M
.
Yo
h
a
n
n
a
,
P
.
M
.
Ijas
i
n
i,
a
n
d
M
.
B.
G
a
rk
id
a
,
“
A
p
p
li
c
a
ti
o
n
o
f
f
u
z
z
y
–
Ne
u
ro
to
m
o
d
e
l
w
e
a
th
e
r
p
a
ra
m
e
ter
v
a
riab
il
it
y
i
m
p
a
c
ts
o
n
e
le
c
tri
c
a
l
l
o
a
d
b
a
se
d
o
n
lo
n
g
-
term
f
o
re
c
a
sti
n
g
,
”
Al
e
x
a
n
d
ria
En
g
i
n
e
e
rin
g
J
o
u
rn
a
l
,
v
o
l.
5
7
,
n
o
.
1,
2
0
1
8
.
[1
2
]
C.
T
.
Hs
u
,
M
.
S
.
Ka
n
g
a
n
d
C.
S
.
Ch
e
n
,
"
De
sig
n
o
f
A
d
a
p
ti
v
e
L
o
p
a
d
S
h
e
d
d
in
g
b
y
A
rti
f
icia
l
Ne
u
ra
l
Ne
tw
o
rk
s,"
IEE
Pro
c
.
o
f
Ge
n
e
ra
ti
o
n
,
T
ra
n
sm
issio
n
a
n
d
Distri
b
u
ti
o
n
,
v
o
l
.
1
5
2
,
p
p
.
4
1
5
-
4
2
1
,
M
a
y
2
0
0
5
.
[1
3
]
M
.
Niz
a
m
,
A
.
M
o
h
a
m
e
d
,
M
.
Al
-
Da
b
b
a
g
h
a
n
d
A
.
H
u
ss
a
in
,
"
Us
in
g
S
u
p
p
o
rt
v
e
c
to
r
M
a
c
h
in
e
s
f
o
r
De
ter
m
in
in
g
V
o
l
tag
e
Un
sta
b
le
A
re
a
s
in
P
o
w
e
r
S
y
ste
m
s,"
2
n
d
In
t.
Co
n
fer
e
n
c
e
o
n
Po
we
r
a
n
d
En
e
rg
y
Co
n
fer
e
n
c
e
(
PE
Co
n
2
0
0
8
)
,
p
p
.
8
7
8
-
8
8
3
,
2
0
0
8
.
[1
4
]
N.S
.
A
.
Ya
s
m
in
,
N.A
.
W
a
h
a
b
,
A
.
N.
A
n
u
a
r,
M
.
Bo
b
,
"
P
e
rf
o
rm
a
n
c
e
c
o
m
p
a
riso
n
o
f
S
V
M
a
n
d
AN
N
f
o
r
a
e
ro
b
ic
g
ra
n
u
lar slu
d
g
e
"
,
in
Bu
ll
e
ti
n
o
f
El
e
c
trica
l
En
g
in
e
e
rin
g
a
n
d
I
n
f
o
rm
a
ti
c
s
,
v
o
l.
8
,
n
o
.
4
,
p
p
.
1
3
9
2
-
1
4
0
1
,
2
0
1
9
.
[1
5
]
C.
L
iu
,
L
.
T
a
n
g
,
J.
L
iu
,
“
L
e
a
st S
q
u
a
re
s S
u
p
p
o
rt
V
e
c
to
r
M
a
c
h
i
n
e
w
it
h
se
lf
-
o
rg
a
n
izin
g
m
u
lt
ip
le k
e
rn
e
l
lea
rn
in
g
a
n
d
sp
a
rsity
”
,
in
Ne
u
ro
c
o
m
p
u
t
in
g
,
v
o
l
.
3
3
1
,
n
o
.
1
,
p
p
.
4
9
3
-
5
0
4
,
2
0
1
9
.
[1
6
]
L
.
Ku
m
a
r
,
S
.
K.
S
rip
a
d
a
,
A.
S
u
re
k
a
,
S
.
K.
Ra
th
,
“
Ef
f
e
c
ti
v
e
f
a
u
lt
p
re
d
ictio
n
m
o
d
e
l
d
e
v
e
lo
p
e
d
u
sin
g
L
e
a
st
S
q
u
a
re
S
u
p
p
o
rt
V
e
c
to
r
M
a
c
h
i
n
e
(L
S
S
VM)”,
in
J
o
u
rn
a
l
o
f
S
y
ste
ms
a
n
d
S
o
ft
wa
re
,
v
o
l
.
1
3
,
n
o
.
1
,
p
p
.
6
8
6
-
7
1
2
,
2
0
1
8
.
[1
7
]
A
.
Y
a
n
g
,
W
.
L
i,
X
.
Ya
n
g
,
"
S
h
o
rt
-
term
e
lec
tri
c
it
y
lo
a
d
f
o
re
c
a
sti
n
g
b
a
se
d
o
n
f
e
a
tu
re
se
l
e
c
ti
o
n
a
n
d
L
e
a
st
S
q
u
a
re
S
u
p
p
o
rt
V
e
c
to
r
M
a
c
h
i
n
e
s"
,
in
K
n
o
wled
g
e
-
B
a
se
d
S
y
ste
ms
,
v
o
l.
1
6
,
n
o
.
1
,
p
p
.
1
5
9
-
1
7
3
,
2
0
1
9
.
[1
8
]
M
.
A
.
A
.
A
z
iz,
Z.
M
.
Y
a
sin
,
a
n
d
Z.
Zak
a
ria,
“
P
re
d
ictio
n
o
f
p
h
o
t
o
v
o
lt
a
ic
s
y
st
e
m
o
u
tp
u
t
u
si
n
g
h
y
b
r
id
lea
st
sq
u
a
re
su
p
p
o
rt
v
e
c
to
r
m
a
c
h
in
e
,
”
in
7
th
IEE
E
In
ter
n
a
ti
o
n
a
l
C
o
n
fer
e
n
c
e
o
n
S
y
ste
m
En
g
i
n
e
e
rin
g
a
n
d
T
e
c
h
n
o
lo
g
y
,
p
p
.
1
-
7
,
2
0
1
7
.
[1
8
]
T
.
Ja
y
a
b
a
ra
th
i,
T
.
Ra
g
h
u
n
a
th
a
n
,
B.
R.
A
d
a
rsh
,
a
n
d
P
.
N.
S
u
g
a
n
th
a
n
,
“
Eco
n
o
m
ic
d
isp
a
tch
u
sin
g
h
y
b
rid
g
re
y
w
o
lf
o
p
ti
m
ize
r,
”
En
e
rg
y
,
v
o
l.
1
1
1
,
p
p
.
6
3
0
-
6
4
1
,
2
0
1
6
.
[1
9
]
S
.
M
o
h
a
n
ty
,
B.
S
u
b
u
d
h
i,
a
n
d
P
.
K.
Ra
y
,
“
A
n
e
w
M
P
P
T
d
e
sig
n
u
sin
g
g
re
y
w
o
lf
o
p
ti
m
iza
ti
o
n
tec
h
n
iq
u
e
f
o
r
p
h
o
to
v
o
lt
a
ic
sy
s
te
m
u
n
d
e
r
p
a
rt
ial
sh
a
d
i
n
g
c
o
n
d
it
i
o
n
s,”
IE
EE
T
r
a
n
s.
S
u
st
a
i
n
a
b
il
it
y
E
n
e
rg
y
,
v
o
l
.
7
,
n
o
.
1
,
p
p
.
1
8
1
-
1
8
8
,
2
0
1
6
.
[2
0
]
U.
S
u
lt
a
n
a
,
A
.
B.
Kh
a
iru
d
d
i
n
,
A
.
S
.
M
o
k
h
tar,
N.
Zare
e
n
,
a
n
d
B.
S
u
lt
a
n
a
,
“
G
re
y
w
o
l
f
o
p
ti
m
ize
r
b
a
se
d
p
lac
e
m
e
n
t
a
n
d
siz
in
g
o
f
m
u
lt
ip
le d
istri
b
u
ted
g
e
n
e
ra
ti
o
n
i
n
t
h
e
d
istri
b
u
t
io
n
sy
ste
m
,
”
En
e
rg
y
,
v
o
l.
1
1
1
,
n
o
.
1
,
p
p
.
5
2
5
-
5
3
6
,
2
0
1
6
.
[2
1
]
A
.
L
a
k
u
m
a
n
d
V
.
M
a
h
a
jan
,
“
Op
ti
m
a
l
p
lac
e
m
e
n
t
a
n
d
siz
in
g
o
f
m
u
lt
ip
le
a
c
ti
v
e
p
o
w
e
r
f
il
ters
in
ra
d
ial
d
istri
b
u
ti
o
n
s
y
ste
m
u
sin
g
g
r
e
y
w
o
lf
o
p
ti
m
i
z
e
r
in
p
re
se
n
c
e
o
f
n
o
n
li
n
e
a
r
d
istri
b
u
te
d
g
e
n
e
ra
ti
o
n
,
”
E
lec
tric
Po
we
r
S
y
ste
m
Res
e
a
rc
h
,
v
o
l.
1
7
3
,
n
o
.
1
,
p
p
.
2
8
1
-
2
9
0
,
2
0
1
9
.
[2
2
]
M
.
Z.
M
.
T
u
m
a
ri,
M
.
H.
S
u
i
d
,
M
.
A.
A
h
m
a
d
,
“
P
V
sy
ste
m
re
a
c
ti
v
e
p
o
w
e
r
c
o
o
rd
in
a
ti
o
n
w
it
h
UL
T
C
&
s
h
u
n
t
c
a
p
a
c
it
o
r
s
u
sin
g
g
re
y
w
o
lf
o
p
ti
m
ize
r
a
l
g
o
rit
h
m
”
,
in
In
d
o
n
e
sia
n
J
o
u
rn
a
l
o
f
E
l
e
c
trica
l
En
g
i
n
e
e
rin
g
a
n
d
C
o
mp
u
t
e
r
S
c
ien
c
e
,
v
o
l.
1
9
,
n
o
.
1
,
p
p
.
1
-
1
0
,
2
0
2
0
.
[2
3
]
C.
M
u
r
o
,
R.
Esc
o
b
e
d
o
,
L
.
S
p
e
c
to
r,
a
n
d
R
.
P
.
C
o
p
p
in
g
e
r,
“
W
o
lf
-
p
a
c
k
(Ca
n
is
lu
p
u
s)
h
u
n
ti
n
g
stra
teg
ies
e
m
e
r
g
e
f
ro
m
sim
p
le ru
les
in
c
o
m
p
u
tatio
n
a
l
sim
u
latio
n
s,”
Beh
a
v
io
u
r
a
l
Pr
o
c
e
ss
e
s
,
v
o
l.
8
8
,
n
o
.
3
,
p
p
.
1
9
2
-
1
9
7
,
2
0
1
1
.
[2
4
]
S
.
M
irj
a
li
li
,
S
.
M
.
M
irj
a
li
li
,
a
n
d
A
.
L
e
w
is,
“
G
re
y
W
o
l
f
Op
ti
m
ize
r,
”
Ad
v
a
n
c
e
s
in
En
g
in
e
e
rin
g
S
o
ft
w
a
re
,
v
o
l.
6
9
,
p
p
.
46
-
6
1
,
2
0
1
4
.
[2
5
]
I.
N.
S
a
m
’o
n
,
Z.
M
.
Ya
sin
,
Z.
Za
k
a
ria,
“
A
n
t
L
io
n
Op
ti
m
ize
r
f
o
r
S
o
lv
in
g
Un
it
Co
m
m
it
m
e
n
t
P
ro
b
lem
in
S
m
a
rt
G
rid
S
y
st
e
m
”
,
In
d
o
n
e
si
a
n
J
o
u
rn
a
l
o
f
E
lec
trica
l
En
g
in
e
e
rin
g
a
n
d
Co
mp
u
ter
S
c
ien
c
e
,
v
o
l.
8
,
n
o
.
1
,
p
p
.
1
2
9
-
1
3
6
,
2
0
1
7
.
Evaluation Warning : The document was created with Spire.PDF for Python.