T
E
L
K
O
M
N
I
K
A
T
elec
o
m
m
un
ica
t
io
n,
C
o
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
9
8
~
2
5
0
4
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
4
6
6
5
2498
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
Predic
tion o
f
ra
in
fall usin
g
impro
v
ed deep
learning
with
pa
r
ticle
s
wa
rm o
ptimiza
tion
I
m
a
m
Cho
lis
s
o
din
,
Su
t
ris
no
F
a
c
u
lt
y
o
f
C
o
m
p
u
ter
S
c
ien
c
e
,
Un
iv
e
rsitas
Bra
wijay
a
,
In
d
o
n
e
sia
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
No
v
2
4
,
2
0
1
9
R
ev
is
ed
Ap
r
8
,
2
0
2
0
Acc
ep
ted
May
1
,
2
0
2
0
Ra
in
fa
ll
is
a
n
a
t
u
ra
l
fa
c
to
r
t
h
a
t
is
v
e
ry
imp
o
rta
n
t
fo
r
fa
rm
e
rs
o
r
c
e
rtain
in
stit
u
ti
o
n
s
to
p
re
d
ict
th
e
p
lan
ti
n
g
p
e
rio
d
o
f
a
p
lan
t.
Th
e
p
ro
b
lem
is
t
h
a
t
ra
in
fa
ll
is
v
e
ry
d
iff
icu
lt
t
o
p
re
d
ict.
Tri
a
ls
to
g
e
t
o
p
ti
m
a
l
ra
in
fa
ll
p
re
d
icti
o
n
h
a
v
e
b
e
e
n
c
a
rried
o
u
t
b
y
BM
KG
th
ro
u
g
h
r
e
se
a
rc
h
with
v
a
riety
o
f
m
e
th
o
d
s
in
v
a
ri
o
u
s
field
s,
in
c
l
u
d
i
n
g
m
e
teo
r
o
lo
g
y
,
c
li
m
a
to
lo
g
y
a
n
d
g
e
o
p
h
y
sic
s.
Th
e
re
su
lt
s
o
f
th
e
stu
d
y
u
n
fo
rt
u
n
a
tel
y
o
b
tain
e
d
a
les
s
o
p
ti
m
a
l
su
c
c
e
ss
ra
t
e
in
p
re
d
ictin
g
ra
in
fa
ll
.
T
o
d
a
y
,
th
e
re
a
re
m
a
n
y
n
e
w
m
e
th
o
d
s
f
o
r
p
re
d
ictin
g
e
v
e
n
ts.
T
h
e
se
m
e
th
o
d
s
i
n
c
lu
d
e
d
e
e
p
lea
rn
i
n
g
(
DL)
a
n
d
P
a
rti
c
le
sw
a
rm
o
p
t
imiz
a
ti
o
n
(
P
S
O).
Th
e
u
se
o
f
th
e
d
e
e
p
lea
rn
in
g
m
e
t
h
o
d
is
v
e
ry
su
sc
e
p
ti
b
le
t
o
i
n
it
ial
we
ig
h
ts
t
h
a
t
a
re
les
s
th
a
n
o
p
ti
m
a
l,
so
it
r
e
q
u
ires
a
p
ro
c
e
ss
o
f
o
p
ti
m
iza
t
io
n
u
sin
g
a
m
e
ta
h
e
u
risti
c
tec
h
n
iq
u
e
,
w
h
ich
i
s
th
e
P
S
O
a
l
g
o
ri
th
m
,
b
e
c
a
u
se
t
h
is
a
lg
o
rit
h
m
h
a
s
a
lev
e
l
o
f
c
o
m
p
lex
it
y
t
h
a
t
is
m
u
c
h
lo
we
r
t
h
a
n
g
e
n
e
ti
c
a
lg
o
r
it
h
m
s.
In
th
is
stu
d
y
,
t
h
is
m
e
th
o
d
is
u
ti
li
z
e
d
to
p
re
d
ict
ra
i
n
fa
ll
b
y
d
e
term
in
in
g
th
e
e
x
a
c
t
re
g
re
ss
io
n
e
q
u
a
ti
o
n
m
o
d
e
l
a
c
c
o
rd
in
g
to
t
h
e
n
u
m
b
e
r
o
f
lay
e
rs
i
n
h
i
d
d
e
n
n
o
d
e
s
b
a
se
d
o
n
t
h
e
siz
e
o
f
th
e
k
e
r
n
e
l
a
n
d
th
e
we
ig
h
t
b
e
twe
e
n
th
e
lay
e
rs.
T
h
is
re
se
a
rc
h
is ap
p
ro
v
e
d
a
c
h
iev
e
d
g
e
t
m
o
re
o
p
ti
m
a
l
ra
in
fa
ll
p
re
d
ictio
n
re
su
lt
s th
a
t
th
o
se
o
f
p
re
v
io
u
s res
e
a
rc
h
th
a
t
wit
h
o
u
t
o
p
t
imiz
a
ti
o
n
wit
h
P
S
O
.
K
ey
w
o
r
d
s
:
Dee
p
P
SO
Fo
r
ec
asti
n
g
I
m
p
r
o
v
ed
d
ee
p
lear
n
i
n
g
Par
tic
le
s
war
m
o
p
tim
izatio
n
R
ain
f
all
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
:
I
m
am
C
h
o
lis
s
o
d
in
,
Facu
lty
o
f
C
o
m
p
u
ter
Scien
ce
,
Un
iv
er
s
itas
B
r
awijay
a,
8
Vete
r
an
R
o
ad
,
Ma
lan
g
6
5
1
4
5
,
I
n
d
o
n
esia.
E
m
ail:
im
am
cs@
u
b
.
ac
.
id
1.
I
NT
RO
D
UCT
I
O
N
Ma
lan
g
R
eg
en
cy
E
ast
J
av
a
Pro
v
in
ce
is
a
n
a
r
ea
th
at
h
a
s
a
q
u
ite
h
ig
h
lev
el
o
f
p
r
o
d
u
ctio
n
in
th
e
Ag
r
icu
ltu
r
e
an
d
Plan
tatio
n
s
ec
to
r
[
1
]
.
T
h
e
p
r
o
d
u
ctio
n
will
d
ec
r
ea
s
e
if
it
h
ar
v
est
f
ailu
r
es
o
cc
u
r
wh
en
en
ter
in
g
th
e
r
ain
y
s
ea
s
o
n
with
h
ig
h
r
a
in
f
all
(
ab
o
v
e
3
0
0
m
m
p
er
m
o
n
th
)
an
d
wh
en
e
n
ter
in
g
th
e
d
r
y
s
ea
s
o
n
with
lo
w
r
ain
f
all
[
2
]
.
On
e
o
f
th
e
e
f
f
o
r
ts
th
at
h
av
e
b
ee
n
m
ad
e
b
y
f
ar
m
er
s
to
o
v
e
r
co
m
e
t
h
is
p
r
o
b
lem
is
b
y
h
a
r
v
esti
n
g
ea
r
ly
.
T
h
ese
ef
f
o
r
ts
ar
e
in
d
ee
d
e
f
f
ec
tiv
e
en
o
u
g
h
to
r
e
d
u
ce
t
h
e
am
o
u
n
t
o
f
lo
s
s
.
I
t
wo
u
ld
b
e
b
etter
,
h
o
wev
er
,
to
m
ak
e
a
p
r
o
ac
tiv
e
e
f
f
o
r
t t
o
av
o
i
d
cr
o
p
f
ailu
r
e
[
3
]
.
T
h
e
p
r
o
ac
tiv
e
ef
f
o
r
ts
o
f
th
e
f
a
r
m
er
s
to
d
ate
ar
e
b
y
ex
am
in
i
n
g
th
e
ca
len
d
ar
to
d
eter
m
i
n
e
th
e
b
est
s
tar
t
o
f
th
e
g
r
o
win
g
s
ea
s
o
n
,
lik
e
th
e
o
n
e
th
e
I
n
d
o
n
esian
Ag
en
cy
f
o
r
Ag
r
icu
ltu
r
al
R
esear
ch
an
d
Dev
elo
p
m
en
t
(
B
alitb
an
g
tan
)
o
f
th
e
Min
is
tr
y
o
f
Ag
r
icu
ltu
r
e
d
o
es,
wh
ich
is
twice
a
y
ea
r
.
T
h
e
p
lan
tin
g
p
er
i
o
d
is
d
eter
m
in
ed
b
y
u
s
in
g
’
d
a
s
a
r
ia
n
’
(
1
0
d
ay
s
)
r
ai
n
f
all
f
o
r
ec
asti
n
g
d
ata
t
o
s
ee
th
e
b
eg
in
n
in
g
o
f
th
e
en
t
r
y
a
n
d
e
n
d
o
f
t
h
e
r
ain
y
s
ea
s
o
n
o
r
th
e
d
r
y
s
ea
s
o
n
f
r
o
m
t
h
e
M
eteo
r
o
lo
g
y
,
C
lim
ato
lo
g
y
a
n
d
Geo
p
h
y
s
ical
Ag
en
cy
/B
MK
G
[
4
]
.
T
h
e
p
r
o
b
lem
is
th
at
th
is
f
o
r
ec
asti
n
g
p
r
o
v
id
ed
b
y
B
MK
G
is
o
f
ten
in
ac
cu
r
ate
[
5
]
;
h
en
ce
,
th
e
ac
cu
r
ac
y
o
f
th
e
p
lan
tin
g
ca
len
d
ar
f
r
o
m
B
alitb
an
g
tan
h
as o
n
ly
r
e
ac
h
ed
5
0
% f
o
r
th
e
wh
o
le
ar
ea
o
f
I
n
d
o
n
esia [
6
]
.
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
P
r
ed
ictio
n
o
f ra
in
fa
ll u
s
in
g
im
p
r
o
ve
d
d
ee
p
lea
r
n
in
g
w
ith
p
a
r
ticle
s
w
a
r
m
o
p
timiz
a
tio
n
(
I
ma
m
C
h
o
lis
s
o
d
in
)
2499
T
o
d
ate,
B
MK
G
h
as o
f
ten
u
s
ed
ad
ap
tiv
e
n
eu
r
o
-
f
u
zz
y
in
f
e
r
e
n
ce
s
y
s
tem
(
ANFI
S)
m
eth
o
d
[
7
]
,
wav
elet
tr
an
s
f
o
r
m
atio
n
[
8
]
an
d
a
u
to
r
e
g
r
ess
iv
e
in
teg
r
ated
m
o
v
i
n
g
a
v
er
ag
e
(
AR
I
MA
)
[
9
]
,
to
p
r
e
d
ict
r
ain
f
all.
B
MK
G
ad
m
itted
th
at
th
e
ac
c
u
r
ac
y
o
f
th
e
m
eth
o
d
is
s
till
n
o
t
g
o
o
d
,
wh
ich
is
7
0
%.
C
u
r
r
e
n
tly
,
th
er
e
ar
e
m
a
n
y
o
t
h
er
m
eth
o
d
s
u
s
ed
to
f
o
r
ec
ast
r
ain
f
all.
On
e
o
f
th
e
m
eth
o
d
s
is
d
ee
p
lear
n
in
g
(
DL
)
w
h
ich
is
co
n
tain
ed
in
th
e
n
eu
r
al
ne
two
r
k
(
NN)
.
T
h
e
d
r
awb
ac
k
o
f
th
is
m
eth
o
d
is
th
at
it
is
o
f
t
en
s
tu
ck
at
th
e
l
o
ca
l
o
p
tim
u
m
b
ec
au
s
e
th
e
in
itial
weig
h
ts
ar
e
g
en
er
ated
r
an
d
o
m
ly
.
T
h
er
ef
o
r
e,
it
is
n
ec
ess
a
r
y
to
h
av
e
a
tech
n
iq
u
e
th
at
i
s
ab
le
to
ac
ce
ler
ate
th
e
s
ea
r
ch
f
o
r
weig
h
ts
s
o
th
at
t
h
e
r
esu
lts
o
b
tain
ed
ca
n
b
e
o
p
ti
m
al.
Par
ticle
s
war
m
o
p
tim
izatio
n
(
PS
O)
alg
o
r
ith
m
in
[
1
0
-
1
8
]
,
is
th
e
alg
o
r
ith
m
th
a
t
p
o
s
s
ess
e
s
s
am
e
lev
el
o
f
ef
f
ec
tiv
en
ess
with
alg
o
r
ith
m
g
en
etics
in
th
e
co
m
p
letio
n
o
f
p
r
o
b
lem
s
,
b
u
t
in
ter
m
s
o
f
e
f
f
icien
cy
,
PS
O
alg
o
r
ith
m
is
s
u
p
er
io
r
to
g
e
n
etic
alg
o
r
ith
m
[
1
9
]
.
T
h
e
r
ef
o
r
e,
th
is
r
esear
ch
will
ap
p
ly
th
e
m
eth
o
d
o
f
im
p
r
o
v
ed
d
ee
p
lear
n
i
n
g
u
s
i
n
g
PS
O
alg
o
r
ith
m
.
I
t
is
ex
p
ec
t
ed
th
at
th
is
m
eth
o
d
is
ab
le
to
p
r
ed
ict
r
ain
f
all
in
M
alan
g
R
eg
en
cy
ac
c
u
r
ately
.
2.
RE
S
E
ARCH
M
E
T
H
O
D
2
.
1
.
Ra
inf
a
ll
R
ain
f
all
is
th
e
h
eig
h
t
o
f
r
ain
w
ater
th
at
is
co
llected
o
n
a
f
lat
p
lace
with
o
u
t
ex
p
er
ien
cin
g
ev
ap
o
r
atio
n
,
d
r
ain
ag
e
o
r
in
f
iltra
tio
n
.
O
n
e
m
illi
m
eter
o
f
r
ain
f
all
m
ea
n
s
th
er
e
is
a
wate
r
r
eser
v
o
ir
in
a
f
lat
ar
ea
as h
ig
h
as o
n
e
m
illi
m
eter
o
r
o
n
e
liter
[
2
,
20
,
21]
.
Fig
u
r
e
1
d
is
p
lay
s
th
e
r
ain
f
all
m
ea
s
u
r
ed
at
v
ar
io
u
s
p
er
io
d
s
.
T
h
e
Me
teo
r
o
lo
g
y
Statio
n
will
m
ea
s
u
r
e
th
e
r
ain
f
a
ll
in
a
s
h
o
r
t
p
er
io
d
o
f
tim
e
(
p
er
h
o
u
r
an
d
p
er
d
ay
)
,
wh
ile
t
h
e
C
lim
ato
lo
g
y
Statio
n
will m
ea
s
u
r
e
it in
a
lo
n
g
p
er
io
d
o
f
tim
e
(
p
er
1
0
d
a
y
s
an
d
p
er
m
o
n
th
)
.
Fig
u
r
e
1
.
Ma
p
o
f
r
ain
f
all
in
I
n
d
o
n
esia
[
2
1
]
2
.
2
.
P
re
dict
io
ns
a
nd
cla
s
s
if
ica
t
io
ns
Pre
d
ictio
n
is
a
d
if
f
er
en
t
th
i
n
g
f
r
o
m
class
if
icatio
n
.
T
h
e
m
ac
h
in
e
lear
n
in
g
,
h
o
wev
e
r
,
co
n
s
id
er
s
class
if
icatio
n
as
o
n
e
k
in
d
o
f
p
r
ed
ictio
n
.
I
n
t
h
e
im
p
l
em
en
tatio
n
o
f
d
ataset
p
r
e
p
r
o
s
ess
in
g
p
r
o
ce
s
s
,
th
e
class
if
icatio
n
u
s
es
th
e
T
ec
h
n
ical
An
aly
s
is
an
d
Fu
n
d
am
en
tal
An
aly
s
is
ap
p
r
o
ac
h
es.
T
h
e
p
u
r
p
o
s
e
o
f
class
if
icatio
n
in
th
is
ca
s
e
i
s
to
p
r
ed
ict
th
e
class
o
r
ca
teg
o
r
y
lab
els
[
2
2
-
2
4
]
.
T
h
e
class
if
icatio
n
is
d
iv
id
ed
in
to
two
ty
p
es,
wh
ich
ar
e:
a.
Su
p
er
v
is
ed
class
if
icatio
n
(
c
lass
if
icatio
n
)
an
d
b.
Un
s
u
p
er
v
is
ed
class
if
icatio
n
(
c
lu
s
ter
in
g
)
T
h
e
ex
tr
ac
tio
n
r
esu
lts
o
f
in
iti
al
d
ata
co
m
es
f
r
o
m
,
e.
g
.
d
ata
1
with
3
f
ea
tu
r
es
(
b
y
tech
n
ic
al
an
aly
s
is
ap
p
r
o
ac
h
)
th
at
is
d
is
p
lay
e
d
o
n
T
ab
le
1
an
d
th
e
c
o
n
v
e
r
tio
n
r
esu
lt
is
p
lace
d
in
T
a
b
le
2
.
Me
a
n
wh
ile,
th
e
illu
s
tr
atio
n
o
f
th
e
v
is
u
aliza
tio
n
f
o
r
m
f
r
o
m
T
ab
le
1
ca
n
b
e
o
b
s
er
v
ed
i
n
Fig
u
r
e
2
.
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
9
8
-
2504
2500
T
ab
le
1
.
I
n
itial
d
ata
(
e
x
am
p
le
cr
ea
te
f
ea
tu
r
es u
s
in
g
t
ec
h
n
ical
a
n
aly
s
is
)
No
X
1
(
3
d
a
y
s a
g
o
X
2
(
2
d
a
y
s a
g
o
)
X
3
(
1
d
a
y
a
g
o
)
Y
(
t
a
r
g
e
t
)
1
1
3
3
3
8
1
3
3
5
6
1
3
3
3
2
1
3
3
3
1
2
1
3
3
5
6
1
3
3
3
2
1
3
3
3
1
..
..
1
3
3
3
2
1
3
3
3
1
..
..
..
1
3
3
3
1
..
..
..
..
..
..
..
..
..
..
..
..
..
T
ab
le
2
.
T
h
e
e
x
tr
ac
tio
n
r
esu
lts
f
r
o
m
i
n
itial
d
ata
to
im
ag
e
m
at
r
ix
No
I
mag
e
ma
t
r
i
x
:
a
s
q
u
a
r
e
ma
t
r
i
x
w
i
t
h
si
z
e
[
n
u
m
_
o
f
_
f
e
a
t
u
r
e
s
x
n
u
m
_
o
f
_
f
e
a
t
u
r
e
s]
Y
(
t
a
r
g
e
t
)
1
1
3
3
3
8
1
3
3
5
6
1
3
3
3
2
1
3
3
3
1
1
3
3
3
8
1
3
3
5
6
1
3
3
3
2
1
3
3
3
8
1
3
3
5
6
1
3
3
3
2
..
..
..
..
..
..
..
..
..
..
Fig
u
r
e
2
.
Desig
n
: 1
D
d
ata
(
e.
g
.
r
ain
f
all
in
mm
-
b
ased
tim
e
s
er
ies)
2
.
3
.
P
r
o
po
s
ed
m
et
ho
d:
im
pr
o
v
ed
deep
lea
rning
wit
h P
S
O
T
h
e
d
ee
p
lear
n
in
g
alg
o
r
ith
m
w
ith
PS
O
ca
n
b
e
u
tili
ze
d
to
p
r
ed
ict
r
ain
f
all
in
Ma
lan
g
R
eg
en
c
y
,
wh
er
e
it
wo
r
k
s
b
y
ch
an
g
i
n
g
th
e
f
e
atu
r
e
ex
tr
ac
tio
n
an
d
d
ata
t
r
an
s
f
o
r
m
atio
n
in
to
im
ag
e
f
o
r
m
.
T
h
e
am
o
u
n
t
o
f
co
n
v
o
l
u
tio
n
a
n
d
p
o
o
li
n
g
la
y
er
d
e
p
en
d
s
o
n
th
e
c
o
m
p
lex
it
y
o
f
th
e
ca
s
e.
T
h
e
co
n
v
o
lu
tio
n
lay
er
co
n
s
is
ts
o
f
s
ev
er
al
g
r
o
u
p
s
o
f
f
ea
tu
r
es
an
d
th
e
p
o
o
lin
g
lay
er
c
o
n
s
is
ts
o
f
a
r
e
d
u
ctio
n
o
r
s
u
m
m
ar
y
o
f
s
ev
er
al
g
r
o
u
p
s
o
f
f
ea
tu
r
es [
2
5
-
4
1
]
.
Her
e
ar
e
th
e
d
etailed
s
tep
s
o
f
d
ee
p
lea
r
n
in
g
with
PS
O:
a.
C
r
ea
te
a
r
elev
an
t m
ap
SDL
-
E
L
M
with
PS
O
b
ased
o
n
Fig
u
r
e
3
.
b.
Set th
e
p
ar
am
eter
v
al
u
e.
Fig
u
r
e
3
.
Ma
p
o
f
s
im
p
lifie
d
d
e
ep
lear
n
in
g
C
NN
b
ased
E
L
M
with
PS
O
(
d
ee
p
PS
O)
c.
Dee
p
PS
O
p
r
o
ce
s
s
T
h
is
is
th
e
p
r
o
ce
s
s
wh
er
e
th
e
r
ep
r
esen
tatio
n
o
f
4
d
im
en
s
io
n
al
clu
s
ter
s
o
n
ea
ch
PS
O
p
ar
ticle
in
Hy
b
r
id
PSO
-
DL
NN
(
d
ee
p
PS
O)
ca
n
b
e
v
iewe
d
in
T
ab
le
3
.
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
P
r
ed
ictio
n
o
f ra
in
fa
ll u
s
in
g
im
p
r
o
ve
d
d
ee
p
lea
r
n
in
g
w
ith
p
a
r
ticle
s
w
a
r
m
o
p
timiz
a
tio
n
(
I
ma
m
C
h
o
lis
s
o
d
in
)
2501
−
T
r
ain
in
g
p
r
o
ce
s
s
o
f
SDLCNN
-
ELM
−
T
esti
n
g
p
r
o
ce
s
s
o
f
SDLCNN
-
ELM
wh
er
e,
k
co
n
s
is
ts
o
f
1
d
im
en
s
io
n
=
[
K
min
=1
;
K
ma
x
=5
]
,
wh
en
ca
lcu
l
atin
g
DL
,
k
v
alu
e
will b
e
co
n
v
er
ted
t
o
2
*
k
+
1
.
F
C
1
_
Wjk
co
n
s
is
ts
o
f
1
x
(
5
x
7
)
=
[
-
0
.
5
;0
.
5
]
F
C
2
_
Wjk
co
n
s
is
ts
o
f
1
x
(
7
x
7
)
=
[
-
0
.
5
;0
.
5
]
F
C
3
_
Wjk
co
n
s
is
ts
o
f
1
x
(
4
x
7
)
=
[
-
0
.
5
;0
.
5
]
T
h
er
ef
o
r
e,
th
e
le
n
g
th
o
f
th
e
d
i
m
en
s
io
n
o
f
ea
ch
p
ar
ticle
is
1
1
3
,
th
e
f
itn
ess
v
alu
e
o
f
wh
ich
i
s
th
e
s
am
e
as
th
e
(
1
/(
1
+
MA
D)
)
v
alu
e
o
f
th
e
r
esu
lts
o
f
d
ee
p
lear
n
in
g
test
in
g
p
r
o
ce
s
s
.
Fig
u
r
e
4
is
th
e
s
n
ip
p
et
co
d
e
p
r
o
ject
f
o
r
d
em
o
,
an
d
p
lease
s
ee
f
u
ll
co
d
e
at
o
u
r
web
p
ag
e:
h
ttp
s
:/
/g
ith
u
b
.
co
m
/im
am
cs1
9
/I
m
p
r
o
v
e
-
Deep
-
L
ea
r
n
in
g
-
with
-
PSO
.
T
ab
le
3
.
R
ep
r
esen
tatio
n
o
f
PS
O
p
ar
ticles f
o
r
d
ee
p
lear
n
in
g
x
i
(
t
)
k
FC
1
_
Wj
k
FC
2
_
Wj
k
FC
3
_
Wj
k
..
..
..
..
..
Fig
u
r
e
4
.
Sn
ip
p
et
co
d
e
o
f
im
p
r
o
v
e
d
ee
p
lear
n
in
g
with
PS
O
(
d
ee
p
PS
O)
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
9
8
-
2504
2502
3.
RE
SU
L
T
S
A
ND
AN
AL
Y
SI
S
B
ased
o
n
Fig
u
r
e
5
,
th
e
E
L
M
an
d
DL
C
NNE
L
M
alg
o
r
ith
m
o
n
r
ain
f
all
d
ata
h
a
v
e
th
e
s
am
e
ten
d
en
cy
f
r
o
m
all
test
s
,
b
u
t
a
r
e
s
till
b
etter
th
an
DL
C
NNE
L
M,
alth
o
u
g
h
in
a
f
lu
ctu
atin
g
m
an
n
er
,
E
L
M
is
b
etter
.
T
h
is
is
b
ec
au
s
e
th
e
weig
h
t
ca
r
r
ied
o
u
t
o
n
ea
ch
p
er
s
o
n
is
d
o
n
e
r
an
d
o
m
ly
,
a
n
d
ea
ch
tim
e
d
o
in
g
an
ex
p
e
r
im
en
t
ca
n
b
e
v
er
y
d
if
f
e
r
en
t.
W
h
en
co
m
p
ar
e
d
,
th
e
b
est
r
esu
lt
o
f
th
e
th
ir
d
is
PS
O
DL
C
N
NE
L
M
(
d
ee
p
PS
O)
.
T
h
is
alg
o
r
ith
m
u
s
es
PS
O
o
p
tim
izatio
n
tech
n
i
q
u
es
to
g
et
th
e
s
am
e
r
esu
lts
with
f
ilter
in
g
p
r
o
ce
s
s
es,
n
am
e
ly
in
th
e
p
r
o
ce
s
s
o
f
co
n
v
o
l
u
tio
n
,
f
ilin
g
an
d
p
o
o
lin
g
.
Als
o
th
e
o
p
tim
al
weig
h
t v
al
u
e
b
etwe
en
lay
er
s
in
th
e
Fu
ll
C
o
n
n
ec
ted
p
r
o
ce
s
s
.
T
h
e
g
r
a
p
h
in
Fig
u
r
e
6
s
h
o
ws
th
e
r
esu
lts
o
f
th
e
c
o
n
v
er
g
en
ce
test
.
T
h
is
co
n
v
er
g
e
n
ce
test
in
g
is
d
o
n
e
to
d
eter
m
in
e
th
e
id
ea
l
iter
atio
n
.
T
h
e
id
ea
l
iter
atio
n
u
s
ed
in
t
h
e
d
ataset
is
1
8
.
W
h
ile
th
e
r
esu
lts
o
f
th
e
s
in
g
le
co
n
v
er
g
en
ce
test
o
f
o
b
tain
e
d
th
e
lo
west
MA
D
v
alu
e
o
f
th
e
PS
ODL
C
N
NE
L
M
(
d
ee
p
PS
O)
is
0
.
3
4
1
8
.
On
th
e
g
r
ap
h
,
o
n
e
th
in
g
th
at
ca
n
in
d
icate
a
g
o
o
d
co
n
v
er
g
en
ce
test
r
esu
lt
is
an
aly
s
is
th
e
m
o
v
em
en
t
th
at
o
b
tain
ed
w
h
en
ea
ch
iter
atio
n
in
c
r
ea
s
es.
I
f
e
v
er
y
iter
atio
n
th
e
r
e
is
a
s
tep
b
y
s
tep
s
ig
n
o
f
im
p
r
o
v
e
m
en
t
m
o
v
em
en
t
o
f
th
e
MA
D
v
alu
e,
g
r
ad
u
ally
an
d
th
en
wh
en
ap
p
r
o
ac
h
in
g
to
war
d
th
e
f
in
al
iter
atio
n
th
er
e
will
a
p
p
ea
r
a
co
n
v
er
g
en
t
s
ig
n
,
ie
b
y
n
o
m
o
r
e
s
ig
n
if
ican
t
ch
an
g
es f
r
o
m
th
e
MA
D
v
al
u
e
,
s
o
th
e
test
ca
n
b
e
s
aid
to
b
e
s
u
cc
ess
f
u
l.
Fig
u
r
e
5
.
C
o
m
p
a
r
is
o
n
b
etwe
e
n
E
L
M,
DL
C
NNE
L
M,
an
d
PS
O
-
DL
C
NNE
L
M
(
d
ee
p
PS
O)
Fig
u
r
e
6
.
C
o
n
v
er
g
e
n
ce
T
est (
5
tim
es)
o
f
PSO
-
DL
C
NNE
L
M
(
d
ee
p
PS
O)
4.
CO
NCLU
SI
O
N
Dee
p
lear
n
in
g
with
PS
O
alg
o
r
ith
m
ca
n
b
e
u
s
ed
to
p
r
ed
ict
r
ai
n
f
all
in
Ma
lan
g
R
eg
en
cy
wh
er
e
it wo
r
k
s
u
s
in
g
f
ea
tu
r
e
ex
tr
ac
tio
n
an
d
tr
an
s
f
o
r
m
in
g
th
e
d
ata
in
to
i
m
ag
e.
T
h
e
f
in
al
r
esu
lt
f
r
o
m
p
r
o
p
o
s
ed
alg
o
r
ith
m
im
p
lem
en
tatio
n
h
as
b
ee
n
s
u
cc
ess
f
u
lly
g
iv
e
a
s
ig
n
if
ican
t
im
p
r
o
v
em
en
t
as
in
th
e
o
b
tain
ed
test
r
esu
lts
,
with
lo
wes
t
av
er
ag
e
v
alu
e
o
f
MA
D
f
r
o
m
PS
ODL
C
N
NE
L
M
0
.
3
4
1
8
.
Fo
r
m
o
r
e
im
p
r
o
v
em
en
t,
th
e
n
ex
t
r
e
s
ea
r
ch
ca
n
u
s
e
m
o
r
e
d
ata
an
d
a
d
d
s
o
m
e
o
p
tim
ized
p
ar
ticle
d
im
en
s
io
n
clu
s
ter
.
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
P
r
ed
ictio
n
o
f ra
in
fa
ll u
s
in
g
im
p
r
o
ve
d
d
ee
p
lea
r
n
in
g
w
ith
p
a
r
ticle
s
w
a
r
m
o
p
timiz
a
tio
n
(
I
ma
m
C
h
o
lis
s
o
d
in
)
2503
RE
F
E
R
E
NC
E
S
[1
]
BP
S
Ja
ti
m
,
“
Eas
t
Ja
v
a
P
ro
v
in
c
e
in
2
0
1
4
(in
Ba
h
a
sa
:
P
ro
v
i
n
si
Ja
wa
Ti
m
u
r
Da
lam
An
g
k
a
2
0
1
4
),
”
2
0
1
4
.
[On
li
n
e
]
Av
a
il
a
b
le:
h
tt
p
:/
/j
a
ti
m
.
b
p
s.g
o
.
i
d
/e
n
/?h
a
l=p
u
b
li
k
a
si_
d
e
ti
l&id
=
5
7
.
[2
]
BM
KG
S
tak
li
m
Ka
ra
n
g
p
lo
so
M
a
lan
g
,
“
Da
sa
rian
III
Atm
o
sp
h
e
ric
a
n
d
S
e
a
Dy
n
a
m
ics
An
a
ly
sis
M
a
rc
h
2
0
1
5
Up
d
a
te
Ap
ri
l
2
,
2
0
1
5
(in
Ba
h
a
sa
:
An
a
li
si
s
Din
a
m
ik
a
Atm
o
sfe
r
Da
n
La
u
t
Da
sa
rian
III
M
a
re
t
2
0
1
5
Up
d
a
te
2
Ap
ri
l
2
0
1
5
),
”
2
0
1
5
.
[On
li
n
e
].
Av
a
il
a
b
le:
h
tt
p
:/
/
k
a
ra
n
g
p
lo
so
.
jatim.b
m
k
g
.
g
o
.
id
/i
n
d
e
x
.
p
h
p
/an
a
li
sis
-
k
o
n
d
isi
-
d
in
a
m
ik
a
-
a
tmo
sfe
r
-
lau
t
-
d
a
sa
rian
/1
5
8
-
a
n
a
li
sis
-
k
o
n
d
isi
-
d
in
a
m
ik
a
-
a
tmo
sfe
r
-
lau
t
-
d
a
sa
rian
-
tah
u
n
-
2
0
1
5
/
3
9
9
-
a
n
a
li
sis
-
d
i
n
a
m
ik
a
-
a
t
m
o
sfe
r
-
d
a
n
-
lau
t
-
d
a
sa
rian
-
iii
-
m
a
re
t
-
2
0
1
5
-
u
p
d
a
te
-
2
-
a
p
r
il
-
2
0
1
5
#
a
x
z
z
3
X8
h
9
y
4
fg
&
g
sc
.
tab
=
0
,
2
0
1
5
.
[
3
]
R
o
q
i
b
M
.
,
“
R
i
c
e
f
i
e
l
d
s
i
n
B
e
n
g
a
w
a
n
S
o
l
o
H
a
r
v
e
s
t
E
a
r
l
y
(
i
n
B
a
h
a
s
a
:
S
a
w
a
h
D
i
B
e
n
g
a
w
a
n
S
o
l
o
P
a
n
e
n
D
i
n
i
)
,
”
2
0
1
5
.
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
:
/
/
w
w
w
.
k
o
r
a
n
-
s
i
n
d
o
.
c
o
m
/
r
e
a
d
/
9
8
5
5
4
4
/
1
5
1
/
s
a
w
a
h
-
di
-
b
e
n
g
a
w
a
n
-
s
o
l
o
-
p
a
n
e
n
-
d
i
n
i
-
142828943
5
.
[4
]
Ek
a
sa
ri
N.
,
“
Wan
t
to
P
lan
t? S
e
e
t
h
e
Ne
w Ve
rsio
n
Ka
tam
(
in
Ba
h
a
sa
:
M
a
u
Tan
a
m
?
Li
h
a
t
Ka
ta
m
Ve
r
si Baru
)
,
”
S
in
a
r
T
a
n
i
,
2
0
1
5
.
[O
n
li
n
e
].
Av
a
il
a
b
le:
h
tt
p
:
//
tab
l
o
id
si
n
a
rtan
i.
c
o
m
/co
n
te
n
t/
re
a
d
/ma
u
-
tan
a
m
-
li
h
a
t
-
k
a
tam
-
v
e
r
si
-
b
a
ru
/,
2
0
1
5
.
[5
]
Uto
m
o
Y.
W
.
,
“
BM
KG
Ad
m
it
s
F
o
re
c
a
st
Th
e
Wea
th
e
r
Is
S
ti
ll
N
o
t
Ac
c
u
ra
te
(
in
Ba
h
a
sa
:
BM
KG
Ak
u
i
P
ra
k
iraa
n
Cu
a
c
a
n
y
a
M
a
sih
Ku
ra
n
g
Ak
u
ra
t
)
,
”
Ko
mp
a
s
,
2
0
1
4
.
[On
l
in
e
].
Av
a
i
lab
le:
h
tt
p
:/
/sa
in
s.
k
o
m
p
a
s.c
o
m
/re
a
d
/2
0
1
4
/0
1
/3
0
/
1
6
2
8
2
7
5
/BM
KG
.
Ak
u
i.
P
ra
k
iraa
n
.
Cu
a
c
a
n
y
a
.
M
a
sih
.
K
u
ra
n
g
.
Ak
u
ra
t
,
2
0
1
4
.
[6
]
Dia
n
in
g
ty
a
s
T.
,
“
KA
TAM
'
s
a
c
c
u
ra
c
y
is
stil
l
lo
w
(
in
Ba
h
a
sa
:
Ak
u
r
a
si
KA
TAM
M
a
sih
Re
n
d
a
h
)
,
”
S
i
n
a
r
T
a
n
i
,
2
0
1
4
.
[On
li
n
e
].
A
v
a
il
a
b
le:
h
tt
p
:/
/t
a
b
lo
i
d
sin
a
rtan
i.
c
o
m
/co
n
ten
t/
re
a
d
/a
k
u
ra
s
i
-
k
a
tam
-
m
a
sih
-
re
n
d
a
h
.
[7
]
In
g
ra
g
u
sta
ri,
“
Ra
in
fa
ll
P
re
d
ict
io
n
Us
in
g
AN
F
IS
(
i
n
Ba
h
a
sa
:
P
re
d
ik
s
i
Cu
ra
h
Hu
jan
De
n
g
a
n
M
e
n
g
g
u
n
a
k
a
n
AN
F
IS
)
,
”
L
o
k
a
k
a
ry
a
N
a
sio
n
a
l
Fo
r
u
m P
ra
k
i
ra
a
n
,
Eva
l
u
a
si
Da
n
Va
l
id
a
si B
M
G
,
2
0
0
5
.
[
8
]
I
n
g
r
a
g
u
s
t
a
r
i
,
“
R
a
i
n
f
a
l
l
P
r
e
d
i
c
t
i
o
n
U
s
i
n
g
W
a
v
e
l
e
t
T
r
a
n
s
f
o
r
m
s
(
i
n
B
a
h
a
s
a
:
P
r
e
d
i
k
s
i
C
u
r
a
h
H
u
j
a
n
D
e
n
g
a
n
M
e
n
g
g
u
n
a
k
a
n
T
r
a
n
s
f
o
r
m
a
s
i
W
a
v
e
l
e
t
)
,
”
P
r
o
s
i
d
i
n
g
L
o
k
a
k
a
r
y
a
N
a
s
i
o
n
a
l
F
o
r
u
m
P
r
a
k
i
r
a
a
n
,
E
v
a
l
u
a
s
i
D
a
n
V
a
l
i
d
a
s
i
B
M
G
,
2005
.
[
9
]
N
u
r
y
a
d
i
,
“
M
o
d
e
l
V
a
l
i
d
a
t
i
o
n
o
f
L
o
n
g
-
T
e
r
m
F
o
r
e
c
a
s
t
U
s
i
n
g
t
h
e
A
r
i
m
a
M
o
d
e
l
(
i
n
B
a
h
a
s
a
:
V
a
l
i
d
a
s
i
M
o
d
e
l
P
r
a
k
i
r
a
a
n
J
a
n
g
k
a
P
a
n
j
a
n
g
M
e
n
g
g
u
n
a
k
a
n
M
o
d
e
l
A
r
i
m
a
),
”
L
o
k
a
k
a
r
y
a
N
a
s
i
o
n
a
l
F
o
r
u
m
P
r
a
k
i
r
a
a
n
,
E
v
a
l
u
a
s
i
D
a
n
V
a
l
i
d
a
s
i
B
M
G
,
2
0
0
5
.
[1
0
]
Ch
e
n
H.,
Ya
n
g
B.
,
a
n
d
Wan
g
G
.
,
“
A
No
v
e
l
Ba
n
k
ru
p
tcy
P
re
d
icti
o
n
M
o
d
e
l
Ba
se
d
o
n
a
n
Ad
a
p
ti
v
e
F
u
z
z
y
K
-
Ne
a
re
st
Ne
ig
h
b
o
r
M
e
t
h
o
d
,”
Kn
o
wled
g
e
-
B
a
se
d
S
y
ste
ms
,
v
o
l.
2
4
,
n
o
.
8
,
p
p
.
1
3
4
8
-
5
9
,
2
0
1
1
.
[1
1
]
En
g
e
l
b
re
c
h
t
A.
P
.
,
“
Co
m
p
u
tati
o
n
a
l
In
telli
g
e
n
c
e
An
In
tro
d
u
c
ti
o
n
,”
En
g
la
n
d
:
J
o
h
n
Wi
ley
&
S
o
n
s L
td
,
2
0
0
7
.
[1
2
]
M
a
ri
n
i
F
.
&
Walc
z
a
k
B.
,
“
P
a
rti
c
le
S
wa
rm
Op
ti
m
iza
ti
o
n
(
P
S
O).
A
t
u
to
rial
,”
C
h
e
mo
me
trics
a
n
d
In
telli
g
e
n
t
L
a
b
o
ra
t
o
ry
S
y
ste
ms
,
v
o
l.
1
4
9
,
P
a
rt
B,
p
p
.
1
5
3
-
6
5
,
2
0
1
5
.
[1
3
]
S
h
a
y
e
g
h
i
H.
,
a
n
d
G
h
a
se
m
i
A.,
“
Ap
p
li
c
a
ti
o
n
Of P
S
O
-
TVAC
to
Im
p
ro
v
e
L
o
w F
re
q
u
e
n
c
y
Os
c
il
latio
n
s
,”
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
n
T
e
c
h
n
ica
l
a
n
d
Ph
y
sic
a
l
Pro
b
lem
s o
f
En
g
in
e
e
rin
g
(IJ
T
P
E)
,
v
o
l.
3
,
n
o
.
4
,
p
p
.
3
6
-
4
4
,
2
0
1
1
.
[1
4
]
Ti
a
n
D.
P
.
,
“
A
Re
v
iew
o
f
C
o
n
v
e
r
g
e
n
c
e
An
a
ly
sis
o
f
P
a
rti
c
le
S
wa
rm
Op
ti
m
iza
ti
o
n
,”
I
n
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
Gr
id
a
n
d
Distrib
u
ted
Co
m
p
u
ti
n
g
,
v
o
l.
6
,
n
o
.
6
,
p
p
.
1
1
7
-
1
2
8
,
2
0
1
3
.
[1
5
]
Va
sa
n
th
i
F
.
S
.
,
a
n
d
Ba
b
u
lal
S
.
C.
,
“
P
S
O
with
Ti
m
e
Va
ry
in
g
Ac
c
e
lera
ti
o
n
Co
e
fficie
n
ts
F
o
r
S
o
l
v
in
g
Op
ti
m
a
l
P
o
we
r
F
lo
w P
r
o
b
lem
,”
J
o
u
rn
a
l
o
f
El
e
c
tr
ica
l
En
g
i
n
e
e
rin
g
,
v
o
l.
1
,
n
o
.
3
,
p
p
.
1
-
1
0
,
2
0
1
6
.
[1
6
]
Wu
P
.
,
G
a
o
L.
,
Li
,
S
.
,
“
An
Im
p
ro
v
e
d
P
a
rti
c
le
S
wa
rm
Op
ti
m
iza
ti
o
n
Al
g
o
rit
h
m
fo
r
Re
li
a
b
i
li
ty
P
ro
b
lem
s
,”
IS
A
T
ra
n
sa
c
ti
o
n
s
,
v
ol
.
5
0
,
n
o
.
1
,
p
p
.
7
1
-
8
1
,
2
0
1
1
.
[1
7
]
Ya
n
X.,
Wu
Q.,
Li
u
H.
,
Hu
a
n
g
W.
,
“
An
Im
p
ro
v
e
d
P
a
rti
c
le
S
wa
r
m
Op
ti
m
iza
ti
o
n
Alg
o
rit
h
m
a
n
d
It
s
Ap
p
li
c
a
ti
o
n
,”
IJ
CS
I
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
Co
mp
u
ter
S
c
ien
c
e
Iss
u
e
s
,
v
o
l
.
1
0
,
n
o
.
1
,
2
0
1
3
.
[1
8
]
Yo
n
g
h
e
L.
,
M
in
g
h
u
i
L.
,
Zey
u
a
n
Y.
,
Li
c
h
a
o
,
C.
,
“
Im
p
r
o
v
e
d
P
a
rti
c
le
S
wa
rm
Op
ti
m
iza
ti
o
n
Alg
o
rit
h
m
a
n
d
It
s
Ap
p
li
c
a
ti
o
n
i
n
Tex
t
F
e
a
tu
re
S
e
lec
ti
o
n
,”
Ap
p
li
e
d
S
o
ft
C
o
mp
u
t
in
g
,
v
o
l.
3
5
,
p
p
.
6
2
9
-
6
3
6
,
2
0
1
5
.
[1
9
]
Ha
ss
a
n
R
.
,
e
t
a
l
.
,
“
A
Co
m
p
a
riso
n
Of
P
a
rti
c
le
S
wa
rm
Op
ti
m
iza
ti
o
n
An
d
T
h
e
G
e
n
e
ti
c
Alg
o
rit
h
m
,”
Am
e
ric
a
n
In
sti
tu
te
o
f
Aer
o
n
a
u
ti
c
s a
n
d
Astro
n
a
u
ti
c
s
,
p
p
.
1
-
13,
2
0
0
5
.
[2
0
]
BM
KG
S
tak
li
m
Ka
ra
n
g
p
lo
so
M
a
lan
g
,
“
F
o
re
c
a
st
o
f
Ra
in
y
Ra
in
y
S
e
a
so
n
(in
Ba
h
a
sa
:
P
ra
k
iraa
n
C
u
ra
h
Hu
ja
n
M
u
sim
Hu
jan
),
”
2
0
1
8
.
[On
li
n
e
].
Av
a
i
lab
le:
h
t
tp
s://
k
a
ra
n
g
p
lo
s
o
.
jatim.
b
m
k
g
.
g
o
.
id
/
in
d
e
x
.
p
h
p
/p
ra
k
iraa
n
-
i
k
li
m
/p
ra
k
iraa
n
-
m
u
sim
/p
ra
k
iraa
n
-
m
u
sim
-
h
u
jan
/
p
r
a
k
iraa
n
-
c
u
ra
h
-
h
u
jan
-
m
u
sim
-
h
u
ja
n
,
2
0
1
8
.
[2
1
]
BM
KG
, “
Cli
m
a
te In
fo
rm
a
ti
o
n
,”
2
0
1
9
.
[On
li
n
e
].
A
v
a
il
a
b
le:
h
tt
p
s://
ww
w.b
m
k
g
.
g
o
.
id
/?la
n
g
=
EN
.
[2
2
]
Ch
o
li
ss
o
d
i
n
I.
,
Riy
a
n
d
a
n
i
E.
,
“
Bi
g
Da
ta
An
a
ly
sis
(in
Ba
h
a
sa
:
An
a
li
s
is
Big
Da
ta
)
,”
Fa
k
u
lt
a
s
I
lmu
K
o
m
p
u
ter
(Fi
lk
o
m),
Un
ive
rs
it
a
s B
ra
wij
a
y
a
(UB),
M
a
l
a
n
g
,
2
0
1
6
.
[2
3
]
M
a
d
u
ra
J.
,
“
In
tern
a
ti
o
n
a
l
F
in
a
n
c
i
a
l
M
a
n
a
g
e
m
e
n
t
(1
1
th
e
d
it
i
o
n
)
,”
F
lo
rid
a
At
l
a
n
ti
c
Un
ive
rs
it
y
,
2
0
1
1
.
[2
4
]
Ne
ll
y
C.
J.,
Weller
P
.
A.
,
“
Tec
h
n
ica
l
An
a
l
y
sis
i
n
th
e
F
o
re
i
g
n
E
x
c
h
a
n
g
e
M
a
rk
e
t
:
A
La
y
m
a
n
s’s
G
u
id
e
,”
Res
e
a
rc
h
Div
isio
n
Fed
e
ra
l
Rev
e
rs
e
Ba
n
k
o
f
S
t.
L
o
u
is W
o
rk
in
g
P
a
p
e
r S
e
rie
s
,
v
o
l.
7
9
,
n
o
.
5
,
p
p
.
2
3
-
3
8
,
1
9
9
7
.
[
2
5
]
K
h
e
l
l
a
l
A
.
,
M
a
H
.
F
e
i
,
Q
.
,
“
C
o
n
v
o
l
u
t
i
o
n
a
l
N
e
u
r
a
l
N
e
t
w
o
r
k
B
a
s
e
d
O
n
E
x
t
r
e
m
e
L
e
a
r
n
i
n
g
M
a
c
h
i
n
e
F
o
r
M
a
r
i
t
i
m
e
S
h
i
p
s
R
e
c
o
g
n
i
t
i
o
n
I
n
I
n
f
r
a
r
e
d
I
m
a
g
e
,”
S
e
n
s
o
r
s
2
0
1
8
,
1
8
,
1
4
9
0
;
d
o
i
:
1
0
.
3
3
9
0
/
s
1
8
0
5
1
4
9
0
w
w
w
.
m
d
p
i
.
c
o
m
/
j
o
u
r
n
a
l
/
s
e
n
s
o
r
s
,
2018
.
[2
6
]
P
a
n
g
,
S
.
&
Ya
n
g
,
X.
,
“
De
e
p
Co
n
v
o
l
u
ti
o
n
a
l
Ex
trem
e
Lea
rn
i
n
g
M
a
c
h
in
e
An
d
Its
Ap
p
li
c
a
ti
o
n
In
Ha
n
d
writt
e
n
Di
g
it
Clas
sifica
ti
o
n
,”
Hin
d
a
wi
Co
m
p
u
t
a
ti
o
n
a
l
I
n
telli
g
e
n
c
e
a
n
d
Ne
u
ro
sc
ien
c
e
,
v
o
l.
2
0
1
6
,
Article
ID
3
0
4
9
6
3
2
,
1
0
p
a
g
e
s,
h
tt
p
:
//
d
x
.
d
o
i.
o
rg
/1
0
.
1
1
5
5
/
2
0
1
6
/
3
0
4
9
6
3
2
,
2
0
1
6
.
[2
7
]
Ro
h
re
r,
B.
,
“
Ho
w
d
o
Co
n
v
o
l
u
ti
o
n
a
l
Ne
u
ra
l
Ne
two
rk
s
wo
rk
?
,”
h
tt
p
s://
b
ro
h
re
r.
g
it
h
u
b
.
i
o
/
h
o
w
_
c
o
n
v
o
l
u
ti
o
n
a
l
_
n
e
u
ra
l_
n
e
two
rk
s
_
wo
rk
.
h
tml
,
2
0
1
6
.
[2
8
]
Ch
o
li
ss
o
d
i
n
I.
,
e
t
a
l.
,
“
Op
ti
m
iza
ti
o
n
o
f
th
e
n
u
tri
ti
o
n
a
l
c
o
n
te
n
t
o
f
Et
a
wa
c
ro
ss
b
re
e
d
(P
E)
g
o
a
t
m
il
k
u
sin
g
EL
M
-
P
S
O
in
t
h
e
UPT
o
f
An
ima
l
Bre
e
d
in
g
a
n
d
F
o
ra
g
e
S
in
g
o
sa
ri
-
M
a
lan
g
(i
n
Ba
h
a
sa
:
Op
ti
m
a
si Ka
n
d
u
n
g
a
n
G
iz
i
S
u
s
u
Ka
m
b
in
g
P
e
ra
n
a
k
a
n
E
taw
a
(P
E)
M
e
n
g
g
u
n
a
k
a
n
E
LM
-
P
S
O
d
i
UP
T
P
e
m
b
ib
it
a
n
Tern
a
k
Da
n
Hijau
a
n
M
a
k
a
n
a
n
Tern
a
k
S
in
g
o
sa
ri
-
M
a
lan
g
)
,“
J
u
rn
a
l
T
e
k
n
o
lo
g
i
In
f
o
rm
a
si
d
a
n
Ilmu
K
o
mp
u
t
e
r
(J
T
IIK)
FIL
KOM
UB
,
v
o
l.
4
N
o
.
1
,
3
1
-
36
,
2
0
1
7
.
[2
9
]
Ch
o
li
ss
o
d
i
n
I.
,
De
wi
R.
K
.
,
“
Op
ti
m
iza
ti
o
n
o
f
He
a
lt
h
y
Die
t
M
e
n
u
V
a
riatio
n
u
si
n
g
P
S
O
-
SA
,”
J
o
u
r
n
a
l
o
f
In
f
o
rm
a
ti
o
n
T
e
c
h
n
o
l
o
g
y
a
n
d
C
o
mp
u
ter
S
c
ien
c
e
(J
IT
e
CS
)
,
v
o
l.
2
,
n
o
.
1
,
2
0
1
8
.
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
9
8
-
2504
2504
[3
0
]
Ch
o
li
ss
o
d
i
n
I.
,
S
u
tri
sn
o
.
,
“
P
re
d
icti
o
n
o
f
Ra
i
n
fa
ll
u
si
n
g
S
imp
li
fied
De
e
p
Lea
rn
in
g
b
a
se
d
Ex
trem
e
Lea
rn
i
n
g
M
a
c
h
i
n
e
s
,”
J
o
u
rn
a
l
o
f
In
fo
rm
a
ti
o
n
T
e
c
h
n
o
l
o
g
y
a
n
d
Co
mp
u
ter
S
c
ien
c
e
(J
IT
e
CS
)
,
v
o
.
3
,
n
o
.
2
,
2
0
1
8
.
[3
1
]
Ca
o
W.
,
e
t
a
l
.,
“
S
o
m
e
Tri
c
k
s
i
n
P
a
ra
m
e
ter
S
e
lec
ti
o
n
fo
r
E
x
trem
e
Lea
rn
in
g
M
a
c
h
in
e
,”
IOP
C
o
n
f.
S
e
rie
s:
M
a
ter
ia
l
s
S
c
ien
c
e
a
n
d
E
n
g
i
n
e
e
rin
g
,
2
0
1
7
.
[3
2
]
Ert
u
ğ
r
u
l
Ö.
F.
,
Ka
y
a
,
Y.,
“
A
De
tailed
An
a
l
y
sis
o
n
Ex
trem
e
Lea
r
n
in
g
M
a
c
h
in
e
a
n
d
No
v
e
l
Ap
p
ro
a
c
h
e
s
Ba
se
d
o
n
ELM
,”
Ame
ric
a
n
J
o
u
r
n
a
l
o
f
C
o
m
p
u
ter
S
c
ien
c
e
a
n
d
E
n
g
i
n
e
e
rin
g
,
v
o
l.
1
,
n
o
.
5
,
p
p
.
43
-
50
,
2
0
1
5
.
[3
3
]
Hu
a
n
g
,
G
.
Bin
,
Z
h
u
,
Q.Y.
&
S
ie
w,
C.
K.
,
“
Ex
trem
e
Lea
rn
in
g
M
a
c
h
in
e
:
A
Ne
w
Lea
rn
in
g
S
c
h
e
m
e
o
f
F
e
e
d
f
o
rwa
r
d
Ne
u
ra
l
Ne
two
rk
s
,”
IEE
E
In
ter
n
a
t
io
n
a
l
Co
n
fer
e
n
c
e
o
n
Ne
u
ra
l
Ne
two
rk
s
-
Co
n
fer
e
n
c
e
Pro
c
e
e
d
in
g
s
,
p
p
.
9
8
5
-
9
9
0
,
2
0
0
4
.
[3
4
]
Hu
a
n
g
G
.
Bin
,
Z
h
u
Q.
Y.
,
S
iew
C.
K.,
“
Ex
trem
e
Lea
rn
in
g
M
a
c
h
i
n
e
:
T
h
e
o
r
y
a
n
d
Ap
p
li
c
a
ti
o
n
s
,”
N
e
u
ro
c
o
mp
u
ti
n
g
,
v
o
l.
70
,
n
o
.
1
-
3
,
p
p
.
4
8
9
-
5
0
1
,
2
0
0
6
.
[3
5
]
Hu
ix
u
a
n
,
F
.
,
Yu
c
h
a
o
W.
,
Zh
a
n
g
H.,
“
S
h
i
p
Ro
ll
in
g
M
o
ti
o
n
P
re
d
icti
o
n
Ba
se
d
o
n
E
x
trem
e
Lea
rn
in
g
M
a
c
h
in
e
,”
Ch
in
e
se
Co
n
tro
l
Co
n
fer
e
n
c
e
,
p
p
.
3
4
6
8
-
3
4
7
2
,
2
0
1
5
.
[3
6
]
Ism
a
il
,
No
ra
in
i.
,
Oth
m
a
n
Z
.
A.,
S
a
m
su
d
in
N.
A.,
“
Re
g
u
lariz
a
ti
o
n
Ac
ti
v
a
ti
o
n
F
u
n
c
ti
o
n
f
o
r
Ex
t
r
e
m
e
Lea
rn
in
g
M
a
c
h
in
e
,”
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
A
d
v
a
n
c
e
d
C
o
mp
u
ter
S
c
ien
c
e
a
n
d
A
p
p
li
c
a
ti
o
n
s
,
v
o
l.
10
,
n
o
.
3
,
2
0
1
9
.
[3
7
]
S
a
lme
ro
n
J.
L.
,
Ce
lma
A.
R.
,
“
El
l
io
t
a
n
d
S
y
m
m
e
tri
c
El
li
o
t
Ex
trem
e
Lea
rn
in
g
M
a
c
h
in
e
fo
r
G
a
u
ss
ian
No
isy
In
d
u
strial
Th
e
rm
a
l
M
o
d
e
l
li
n
g
,”
E
n
e
rg
ies
,
v
o
l.
12
,
n
o
.
1
,
p
p
.
1
-
19,
2
0
1
8
.
[3
8
]
S
in
g
h
R.
,
Ba
las
u
n
d
a
ra
m
S
.
,
“
A
p
p
li
c
a
ti
o
n
o
f
Ex
trem
e
Lea
rn
in
g
M
a
c
h
in
e
M
e
th
o
d
fo
r
Ti
m
e
S
e
ries
An
a
ly
sis
,”
Pro
c
e
e
d
in
g
s
o
f
W
o
rl
d
Aca
d
e
my
o
f
S
c
ien
c
e
,
E
n
g
i
n
e
e
rin
g
a
n
d
T
e
c
h
n
o
lo
g
y
,
2
0
0
7
.
[3
9
]
S
re
e
k
a
n
th
M
.
S
.
,
Ra
jes
h
R.
,
S
a
th
e
e
sh
k
u
m
a
r
J.,
“
Ex
trem
e
Lea
rn
in
g
M
a
c
h
in
e
fo
r
t
h
e
Clas
sifica
ti
o
n
o
f
Ra
i
n
fa
ll
a
n
d
Th
u
n
d
e
rsto
rm
,”
J
o
u
rn
a
l
o
f
A
p
p
l
ied
S
c
ien
c
e
s
,
v
o
l.
1
,
n
o
.
15
,
p
p
.
1
5
3
-
1
5
6
,
2
0
1
5
.
[4
0
]
S
rimu
a
n
g
W.
,
I
n
tara
so
th
o
n
c
h
u
n
S.,
“
Clas
sifica
ti
o
n
M
o
d
e
l
o
f
Ne
t
wo
rk
In
tr
u
sio
n
u
sin
g
Weig
h
ted
E
x
trem
e
Lea
rn
in
g
M
a
c
h
in
e
,”
In
ter
n
a
ti
o
n
a
l
J
o
in
t
C
o
n
fer
e
n
c
e
o
n
C
o
mp
u
ter
S
c
ien
c
e
a
n
d
S
o
ft
w
a
re
En
g
in
e
e
rin
g
(J
CS
S
E)
,
2
0
1
5
.
[4
1
]
S
a
to
to
,
B.
D.,
e
t
a
l
.,
“
An
Im
p
ro
v
e
m
e
n
t
o
f
G
ra
m
-
n
e
g
a
ti
v
e
Ba
c
teria
I
d
e
n
ti
fica
ti
o
n
u
si
n
g
Co
n
v
o
lu
ti
o
n
a
l
Ne
u
ra
l
Ne
two
r
k
with
F
i
n
e
Tu
n
in
g
,”
T
EL
KO
M
NIKA
T
e
lec
o
mm
u
n
ica
t
io
n
Co
mp
u
ti
n
g
E
lec
tro
n
ics
a
n
d
C
o
n
tro
l
,
v
o
l.
1
8
,
n
o
.
3
,
p
p
.
1
3
9
7
-
1
4
0
5
,
2
0
2
0
.
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
Im
a
m
Cho
li
ss
o
d
i
n
,
b
o
r
n
i
n
Lam
o
n
g
a
n
o
n
Ju
ly
1
9
,
1
9
8
5
,
h
a
s
c
o
m
p
l
e
ted
h
is
M
a
ste
r
in
In
f
o
rm
a
ti
o
n
En
g
i
n
e
e
rin
g
F
TIF
ITS
S
u
ra
b
a
y
a
in
2
0
1
1
.
S
in
c
e
2
0
1
2
,
h
e
h
a
s
b
e
e
n
a
c
ti
v
e
a
s
a
lec
tu
re
r
in
th
e
De
p
a
rtme
n
t
o
f
I
n
fo
rm
a
ti
o
n
T
e
c
h
n
o
l
o
g
y
a
n
d
Co
m
p
u
ter
S
c
ien
c
e
(
a
.
k
.
a
P
TII
K)
t
h
a
t
si
n
c
e
2
0
1
6
h
a
s
b
e
c
o
m
e
th
e
F
a
c
u
l
ty
o
f
C
o
m
p
u
ter
S
c
ien
c
e
(
a
.
k
.
a
F
ILKOM
)
Un
iv
e
rsitas
Bra
wijay
a
(UB)
M
a
lan
g
.
He
tea
c
h
e
s
se
v
e
ra
l
su
b
jec
ts
su
c
h
a
s
In
fo
rm
a
ti
o
n
Re
tri
e
v
a
l,
Dig
it
a
l
Im
a
g
e
P
ro
c
e
ss
in
g
,
P
ro
b
a
b
il
it
y
a
n
d
S
tatist
ics
,
Co
m
p
u
ter
G
r
a
p
h
ics
,
De
c
isio
n
S
u
p
p
o
r
t
S
y
ste
m
,
Artifi
c
ial
In
telli
g
e
n
c
e
,
Da
ta
M
in
i
n
g
,
Big
Da
ta
An
a
l
y
sis,
G
P
U
P
ro
g
ra
m
m
in
g
,
Ev
o
l
u
ti
o
n
Al
g
o
rit
h
m
,
S
wa
rm
In
telli
g
e
n
c
e
,
P
a
tt
e
rn
Re
c
o
g
n
it
i
o
n
,
a
n
d
M
o
b
il
e
P
ro
g
ra
m
m
in
g
.
I
n
a
d
d
i
ti
o
n
t
o
tea
c
h
in
g
,
h
e
is
a
lso
a
c
ti
v
e
i
n
th
e
In
telli
g
e
n
t
S
y
ste
m
a
n
d
M
e
d
ia,
G
a
m
e
&
M
o
b
il
e
Tec
h
n
o
lo
g
y
(M
G
M
)
in
th
e
Re
se
a
rc
h
Lab
o
ra
to
r
y
.
F
r
o
m
2
0
1
5
to
2
0
1
9
,
h
e
c
o
n
ti
n
u
e
d
a
re
se
a
rc
h
in
th
e
field
o
f
Big
Da
ta
c
o
ll
a
b
o
ra
ted
wit
h
th
e
field
o
f
Eco
n
o
m
ics
a
lo
n
g
wit
h
a
tea
m
o
f
p
r
o
fe
ss
o
rs
a
n
d
st
u
d
e
n
ts
o
f
th
e
F
a
c
u
lt
y
o
f
Ec
o
n
o
m
ics
a
n
d
B
u
sin
e
ss
(F
EB)
UB
a
n
d
Re
g
io
n
a
l
De
v
e
l
o
p
m
e
n
t
P
lan
n
i
n
g
Ag
e
n
c
y
(BAP
P
EDA)
o
f
Eas
t
Ja
v
a
P
ro
v
i
n
c
e
u
n
d
e
r
t
h
e
t
h
e
m
e
o
f
“
Co
n
stru
c
ti
n
g
th
e
Blu
e
P
ri
n
t
o
f
Op
e
n
Da
ta
Util
iza
ti
o
n
In
it
iatio
n
in
Re
g
io
n
a
l
De
v
e
l
o
p
m
e
n
t
P
lan
n
in
g
”
to
su
p
p
o
rt
i
n
teg
ra
ted
S
m
a
rt
G
o
v
e
rn
a
n
c
e
(i
n
teg
ra
ted
wit
h
a
ll
e
x
isti
n
g
sy
ste
m
s
fro
m
v
a
ri
o
u
s
p
lat
fo
rm
s)
b
a
se
d
o
n
Artifi
c
ial
I
n
telli
g
e
n
c
e
in
th
e
n
e
x
t
fe
w
y
e
a
rs
a
n
d
th
e
d
e
v
e
lo
p
m
e
n
t
o
f
t
h
e
Co
re
En
g
i
n
e
De
e
p
Lea
rn
in
g
a
n
d
Big
Da
ta
a
s
Ge
n
e
ra
l
Li
b
ra
ry
o
r
To
o
l
b
o
x
a
n
d
p
a
c
k
a
g
e
in
sta
ll
e
r,
a
n
d
t
h
e
s
u
p
p
o
rt
f
o
r
p
r
o
g
ra
m
m
in
g
la
n
g
u
a
g
e
s
a
n
d
a
n
y
O
S
p
latfo
rm
s
u
n
d
e
r
th
e
u
m
b
re
ll
a
o
f
In
tell
ig
e
n
t
La
b
o
ra
to
ry
C
o
m
p
u
tat
io
n
F
ILKOM
UB
re
se
a
r
c
h
in
b
a
c
k
e
n
d
a
n
d
fro
n
ten
d
c
o
m
p
u
tati
o
n
o
n
t
h
e
d
e
sk
to
p
,
we
b
a
n
d
m
o
b
il
e
d
e
v
ice
s
o
n
t
h
e
field
o
f
h
e
a
lt
h
,
g
o
v
e
rn
a
n
c
e
a
n
d
th
e
o
th
e
rs
t
h
a
t
a
re
lo
c
a
ll
y
-
b
a
se
d
a
n
d
se
rv
e
rles
s
with
th
e
tec
h
n
o
lo
g
y
o
f
c
lo
u
d
c
o
m
p
u
ti
n
g
in
o
rd
e
r
to
e
sta
b
li
s
h
a
n
d
c
re
a
te
Ad
v
a
n
c
e
d
Tec
h
n
o
l
o
g
y
“
S
m
a
rt
Ap
p
”
p
ro
d
u
c
ts
in
t
h
e
In
d
u
strial
Re
v
o
lu
t
io
n
4
.
0
a
n
d
S
o
c
iety
5
.
0
e
r
a
s
fo
r
H
u
m
a
n
it
y
.
S
u
tr
isn
o
,
b
o
rn
i
n
T
u
lu
n
g
a
g
u
n
g
o
n
M
a
rc
h
2
5
,
1
9
5
7
,
h
a
s
c
o
m
p
lete
d
h
is
u
n
d
e
r
g
ra
d
u
a
te
e
d
u
c
a
ti
o
n
a
t
th
e
El
e
c
tri
c
a
l
En
g
in
e
e
ri
n
g
o
f
Ba
n
d
u
n
g
I
n
stit
u
te
o
f
Tec
h
n
o
lo
g
y
(IT
B)
g
ra
d
u
a
ti
n
g
i
n
1
9
8
2
a
n
d
g
ra
d
u
a
te
e
d
u
c
a
ti
o
n
(
S
2
)
in
th
e
El
e
c
tri
c
a
l
En
g
i
n
e
e
rin
g
M
a
ste
r
S
tu
d
y
P
r
o
g
ra
m
Un
iv
e
rsitas
Bra
wijay
a
(UB)
g
ra
d
u
a
ti
n
g
i
n
2
0
0
8
.
S
i
n
c
e
1
9
8
2
,
h
e
h
a
s
b
e
e
n
a
lec
tu
re
r
in
th
e
De
p
a
rtme
n
t
o
f
El
e
c
tri
c
a
l
En
g
i
n
e
e
rin
g
,
F
a
c
u
lt
y
o
f
En
g
in
e
e
rin
g
,
Un
iv
e
rsitas
Bra
wijay
a
a
n
d
wa
s
th
e
C
h
a
irma
n
o
f
th
e
In
f
o
rm
a
ti
c
s
En
g
in
e
e
rin
g
S
t
u
d
y
P
ro
g
ra
m
fro
m
2
0
0
9
t
o
2
0
1
1
.
In
2
0
1
1
-
2
0
1
6
,
h
e
se
rv
e
d
a
s
th
e
Ch
a
irma
n
o
f
t
h
e
P
ro
g
ra
m
(De
a
n
)
in
t
h
e
In
f
o
rm
a
ti
o
n
Tec
h
n
o
lo
g
y
a
n
d
C
o
m
p
u
ter
S
c
ien
c
e
P
ro
g
ra
m
(P
TII
K))
Un
i
v
e
rsitas
Bra
wijay
a
th
a
t
is
n
o
w
th
e
F
a
c
u
l
ty
o
f
Co
m
p
u
ter
S
c
ien
c
e
(F
ILKOM
)
o
f
U
n
iv
e
rsitas
Bra
wijay
a
.
T
h
e
su
b
jec
ts
th
a
t
h
e
h
a
s
tau
g
h
t
i
n
c
lu
d
e
Dis
tri
b
u
ti
o
n
S
y
ste
m
s,
El
e
c
tro
n
ics
/E
lec
tri
c
a
l
Ne
two
r
k
s,
Ba
sic
P
ro
g
ra
m
m
in
g
,
Ad
v
a
n
c
e
d
P
r
o
g
ra
m
m
in
g
,
Al
g
o
rit
h
m
De
sig
n
a
n
d
An
a
l
y
sis,
Da
ta S
tr
u
c
tu
re
An
a
ly
sis a
n
d
S
o
f
twa
re
De
sig
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.