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
.
1
,
Ma
r
ch
2
0
2
0
,
p
p
.
33
~
39
I
SS
N:
2
2
5
2
-
8938
,
DOI
: 1
0
.
1
1
5
9
1
/i
j
ai.
v
9
.i
1
.
p
p
33
-
39
33
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
i
.
ia
esco
r
e.
co
m
Artif
i
cia
l neural
netw
o
rk
f
o
recas
ti
ng
perf
o
r
m
a
nce
w
ith
m
iss
ing
v
a
lue i
m
p
utatio
ns
Nur
H
a
izu
m
Abd
Ra
h
m
a
n
1
,
M
uh
a
mm
a
d H
is
y
a
m
L
ee
2
1
D
e
p
a
rtme
n
t
o
f
M
a
th
e
m
a
ti
c
s,
F
a
c
u
lt
y
o
f
S
c
ien
c
e
,
Un
iv
e
rsiti
P
u
tra
M
a
la
y
sia
,
4
3
4
0
0
,
S
e
rd
a
n
g
,
S
e
lan
g
o
r
,
M
a
lay
sia
2
De
p
a
rtme
n
t
o
f
M
a
th
e
m
a
ti
c
a
l
S
c
ien
c
e
s,
F
a
c
u
lt
y
o
f
S
c
ien
c
e
,
Un
iv
e
r
siti
T
e
k
n
o
lo
g
i
M
a
lay
sia
,
8
1
3
1
0
,
S
k
u
d
a
i,
Jo
h
o
r
Ba
h
ru
,
Jo
h
o
r
,
M
a
lay
sia
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
No
v
1
2
,
2
0
1
9
R
ev
i
s
ed
J
an
2
5
,
2
0
20
A
cc
ep
ted
Feb
2
,
2
0
20
T
h
is
p
a
p
e
r
p
re
se
n
ts
ti
m
e
s
e
ries
fo
re
c
a
stin
g
m
e
th
o
d
in
o
rd
e
r
t
o
a
c
h
iev
e
h
ig
h
a
c
c
u
ra
c
y
p
e
rf
o
r
m
a
n
c
e
.
In
th
is
stu
d
y
,
th
e
m
o
d
e
rn
ti
m
e
se
ries
a
p
p
ro
a
c
h
w
it
h
th
e
p
re
se
n
c
e
o
f
m
issin
g
v
a
lu
e
s
p
ro
b
lem
is
d
e
v
e
lo
p
e
d
.
T
h
e
a
rti
f
i
c
ial
n
e
u
ra
l
n
e
tw
o
rk
s
(
A
NN
s)
is
u
se
d
to
f
o
re
c
a
st
th
e
f
u
tu
re
v
a
lu
e
s
w
it
h
th
e
m
issin
g
v
a
lu
e
im
p
u
tatio
n
s
m
e
th
o
d
s
u
se
d
k
n
o
w
n
a
s
a
v
e
ra
g
e
,
n
o
rm
a
l
ra
ti
o
a
n
d
a
lso
th
e
m
o
d
if
ied
m
e
th
o
d
.
T
h
e
re
su
lt
s
a
re
v
a
li
d
a
ted
b
y
u
sin
g
m
e
a
n
a
b
so
lu
te
e
rro
r
(M
A
E)
a
n
d
ro
o
t
m
e
a
n
sq
u
a
re
e
rro
r
(RM
S
E).
T
h
e
re
su
lt
sh
o
w
n
th
a
t
b
y
c
o
n
sid
e
ri
n
g
th
e
rig
h
t
m
e
th
o
d
i
n
m
issin
g
v
a
lu
e
s
p
ro
b
lem
s
c
a
n
im
p
ro
v
e
d
a
rti
f
icia
l
n
e
u
ra
l
n
e
tw
o
rk
f
o
re
c
a
st
a
c
c
u
ra
c
y
.
It
is
p
ro
v
e
n
in
b
o
th
M
A
E
a
n
d
RM
S
E
m
e
a
su
re
m
e
n
ts
a
s
f
o
re
c
a
st
im
p
ro
v
e
d
f
ro
m
8
.
7
5
t
o
4
.
5
6
a
n
d
f
ro
m
1
0
.
5
7
to
5
.
8
5
re
sp
e
c
ti
v
e
ly
.
T
h
u
s,
th
is
st
u
d
y
su
g
g
e
sts
b
y
u
n
d
e
rsta
n
d
in
g
t
h
e
p
ro
b
lem
in
ti
m
e
se
ri
e
s
d
a
ta
c
a
n
p
ro
d
u
c
e
a
c
c
u
ra
te
f
o
re
c
a
st
a
n
d
th
e
c
o
rre
c
t
d
e
c
isio
n
m
a
k
in
g
c
a
n
b
e
p
ro
d
u
c
e
d
.
K
ey
w
o
r
d
s
:
A
ir
p
o
llu
ta
n
t i
n
d
ex
E
r
r
o
r
m
ea
s
u
r
e
m
e
n
t
s
A
r
ti
f
icial
n
e
u
r
al
n
et
w
o
r
k
I
m
p
u
tatio
n
s
Fo
r
ec
atin
g
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
:
Nu
r
Haiz
u
m
A
b
d
R
a
h
m
an
,
Dep
ar
t
m
en
t o
f
Ma
th
e
m
at
ics,
Facu
lt
y
o
f
Sc
ien
ce
,
U
n
i
v
er
s
iti
P
u
tr
a
Ma
la
y
s
ia,
4
3
4
0
0
,
Ser
d
an
g
,
Selan
g
o
r
,
Ma
la
y
s
ia.
E
m
ail:
n
u
r
h
aiz
u
m
_
ar
@
u
p
m
.
e
d
u
.
m
y
1.
I
NT
RO
D
UCT
I
O
N
Data
ca
n
b
e
o
b
tain
ac
co
r
d
in
g
to
ti
m
e
e
ith
er
in
h
o
u
r
l
y
,
d
ail
y
o
r
y
ea
r
l
y
.
T
h
is
f
o
r
m
o
f
r
ec
o
r
d
ed
d
ata
is
k
n
o
w
a
s
ti
m
e
s
er
ie
s
.
B
y
an
a
l
y
s
i
s
t
h
e
ti
m
e
s
er
ie
s
d
ata,
th
e
s
tr
u
ct
u
r
e
o
f
t
h
e
d
ata
ca
n
b
e
t
ak
en
in
b
u
ild
in
g
t
h
e
m
o
d
el
[
1
]
.
Ma
n
y
f
o
r
ec
asti
n
g
t
ec
h
n
iq
u
es
h
av
e
b
ee
n
r
ep
o
r
ted
in
th
e
liter
atu
r
e
[
2
]
.
I
n
g
e
n
er
al,
th
ese
tec
h
n
iq
u
e
s
ca
n
b
e
class
i
f
ied
r
o
u
g
h
l
y
i
n
to
th
r
ee
g
r
o
u
p
s
,
i.e
.
s
tatis
t
ical
m
o
d
i
f
ied
-
clas
s
ic
al
m
et
h
o
d
s
,
tech
n
iq
u
es
b
ased
o
n
ar
tif
icial
i
n
tell
ig
e
n
ce
s
,
a
n
d
ad
v
an
ce
d
tec
h
n
iq
u
es.
T
h
e
m
o
s
t
i
m
p
o
r
tan
t
clas
s
ical
m
et
h
o
d
is
th
e
au
to
r
eg
r
ess
i
v
e
in
te
g
r
ated
m
o
v
i
n
g
av
er
a
g
e
(
AR
I
M
A
)
m
o
d
el
b
ec
a
u
s
e
o
f
it
s
f
lex
ib
ilit
y
in
m
o
d
elli
n
g
d
i
f
f
er
e
n
t
t
y
p
e
s
o
f
d
ataset
[
1
,
3
]
.
T
h
is
m
o
d
el
is
ex
ec
u
ted
f
r
o
m
t
h
e
au
to
r
eg
r
es
s
iv
e
m
o
d
el
(
AR
)
,
t
h
e
m
o
v
i
n
g
a
v
er
ag
e
m
o
d
el
(
M
A
)
a
n
d
t
h
e
co
m
b
in
at
io
n
o
f
AR
a
n
d
M
A
m
o
d
els
,
w
h
ic
h
is
k
n
o
w
n
as
AR
M
A
m
o
d
els.
I
n
ad
d
itio
n
,
if
th
er
e
i
s
an
ex
is
te
n
ce
o
f
s
ea
s
o
n
al
co
m
p
o
n
en
t
i
n
t
h
e
s
er
ie
s
,
th
en
th
e
m
o
d
el
i
s
k
n
o
w
n
as
s
e
aso
n
al
AR
I
M
A
(
S
AR
I
M
A
)
m
o
d
el
[
3
]
.
T
h
u
s
,
th
e
f
le
x
ib
ilit
y
o
f
th
is
m
o
d
el
m
a
k
e
it
co
m
p
etiti
v
e
w
it
h
th
e
r
ec
en
tl
y
d
e
v
elo
p
ed
m
eth
o
d
s
.
Ho
w
e
v
er
,
th
e
m
aj
o
r
lim
itatio
n
o
f
th
i
s
m
o
d
el
is
th
at
it
ca
n
o
n
l
y
ca
p
tu
r
e
t
h
e
l
in
ea
r
f
o
r
m
o
f
ti
m
e
s
er
ies
d
ata
a
n
d
th
e
p
r
eli
m
i
n
ar
y
an
al
y
s
is
s
ta
g
es
b
ec
o
m
e
t
h
e
co
n
s
tr
ai
n
t
i
n
b
u
ild
in
g
th
is
m
o
d
el
[
4
]
.
As
an
alter
n
ativ
e
to
t
h
e
cla
s
s
ical
m
et
h
o
d
s
,
f
o
r
ec
aster
s
h
av
e
d
ev
elo
p
ed
n
e
w
m
et
h
o
d
s
th
at
ca
n
o
v
er
co
m
e
th
e
li
m
itat
io
n
o
f
cl
ass
ical
m
eth
o
d
s
,
s
u
c
h
as
ar
tific
ial
n
eu
r
al
n
et
w
o
r
k
(
A
NN)
an
d
t
h
e
f
u
zz
y
t
i
m
e
s
er
ies
(
FT
S).
A
NN
h
as
b
ee
n
w
id
el
y
u
s
ed
as
a
f
o
r
ec
asti
n
g
m
o
d
el
i
n
m
a
n
y
ap
p
licatio
n
s
[
5
]
.
T
h
is
in
cl
u
d
e
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
.
1
,
Ma
r
ch
20
20
:
33
–
39
34
w
ea
t
h
er
f
o
r
ec
ast
[
6
]
,
elec
tr
icit
y
p
r
ice
[
7
]
,
air
lin
e
d
ata
[
8
]
,
a
n
d
m
a
n
y
o
t
h
er
s
.
I
t
i
s
b
ec
au
s
e
A
N
N
is
f
lex
ib
le
in
f
o
r
ec
asti
n
g
ap
p
licatio
n
s
s
in
c
e
it
ca
n
m
o
d
el
b
o
th
lin
ea
r
an
d
n
o
n
-
li
n
ea
r
p
r
o
ce
s
s
es
[
9
]
.
I
n
d
ev
elo
p
in
g
A
N
N
m
o
d
el,
th
e
p
r
e
-
p
r
o
ce
s
s
i
n
g
o
f
d
ata
is
i
m
p
o
r
tan
t
b
ef
o
r
e
an
al
y
s
i
s
w
h
er
e
t
h
e
d
ata
tr
an
s
f
o
r
m
atio
n
n
o
r
m
all
y
u
s
ed
is
[
-
1
,
1
]
o
r
[
0
,
1
]
[
1
0
-
1
2
]
.
Dif
f
er
e
n
t
f
r
o
m
p
r
ev
io
u
s
s
t
u
d
y
,
t
h
e
an
al
y
s
i
s
tak
e
t
h
e
i
m
p
o
r
tan
t
o
f
i
m
p
u
tatio
n
th
e
m
is
s
in
g
d
ata
in
to
A
NN
f
o
r
ec
ast.
T
h
e
f
o
r
ec
ast
p
er
f
o
r
m
an
ce
w
ill
b
e
v
alid
ate
u
s
i
n
g
r
e
al
d
ata
w
h
ich
i
s
air
q
u
alit
y
d
ata.
A
ir
p
o
llu
tio
n
is
t
h
e
m
aj
o
r
p
o
ll
u
tio
n
p
r
o
b
le
m
in
t
h
e
w
o
r
ld
[
1
3
]
.
P
o
w
er
p
r
o
d
u
ctio
n
f
r
o
m
p
o
w
er
p
lan
t
,
v
eh
ic
les
f
u
el
b
u
r
n
in
g
,
in
d
u
s
tr
ial
p
r
o
ce
s
s
es
an
d
n
atu
r
al
f
ac
t
o
r
s
lik
e
v
o
lcan
o
er
u
p
tio
n
m
a
k
e
th
e
air
q
u
alit
y
w
o
r
s
e
n
.
T
h
e
is
s
u
es
o
f
air
q
u
alit
y
n
o
w
b
ec
o
m
e
a
m
aj
o
r
c
o
n
ce
r
n
w
o
r
ld
w
id
e
as
its
ef
f
e
cts
ar
e
d
iv
er
s
e
a
n
d
n
u
m
er
o
u
s
[
1
4
]
.
T
h
e
p
o
llu
tio
n
n
o
t
o
n
l
y
ef
f
ec
t
h
u
m
an
h
e
alth
b
u
t
al
s
o
to
w
ar
d
s
t
h
e
f
o
r
ests
,
w
ater
s
an
d
w
h
o
le
ec
o
s
y
s
te
m
.
I
n
Ma
la
y
s
ia,
a
s
er
ie
s
o
f
h
az
e
ep
is
o
d
es
w
er
e
r
ep
o
r
ted
s
in
ce
th
e
1
9
8
0
s
[
1
5
]
.
Ma
s
s
iv
e
lan
d
an
d
f
o
r
est
f
ir
es i
n
Su
m
atr
a
a
n
d
Kali
m
a
n
t
an
,
I
n
d
o
n
e
s
ia
h
as b
ee
n
th
e
m
a
in
r
ea
s
o
n
o
f
h
az
e
ep
is
o
d
es o
cc
u
r
r
en
ce
.
T
h
e
w
i
n
d
s
h
as
m
ad
e
i
t
ea
s
ier
f
o
r
t
h
e
h
ea
v
y
h
az
e
to
b
e
tr
an
s
p
o
r
ted
.
A
c
co
r
d
in
g
to
DOE
r
ep
o
r
t
[
1
6
]
,
f
o
r
th
e
f
ir
s
t
ti
m
e
i
n
Ma
la
y
s
ia
’
s
h
i
s
to
r
y
,
3
4
s
tatio
n
s
in
t
h
is
co
u
n
tr
y
r
ec
o
r
d
ed
u
n
h
ea
lt
h
y
air
q
u
ali
t
y
s
tatu
s
w
h
i
ch
h
ap
p
en
ed
o
n
1
5
Sep
te
m
b
er
2
0
1
5
.
B
esid
es
Ma
la
y
s
ia,
h
az
e
also
r
ea
c
h
es
a
n
o
th
er
So
u
th
ea
s
t
Asi
a
co
u
n
tr
y
s
u
c
h
a
s
Si
n
g
ap
o
r
e,
T
h
ailan
d
an
d
B
r
u
n
ei
[
1
7
]
.
T
h
e
Dep
ar
tm
e
n
t o
f
E
n
v
ir
o
n
m
en
t (
DOE
)
i
s
a
g
o
v
er
n
m
e
n
t
ag
en
c
y
w
h
ich
is
r
e
s
p
o
n
s
ib
le
to
m
o
n
ito
r
a
n
d
m
an
a
g
e
Ma
la
y
s
ia
’
s
air
q
u
alit
y
.
T
h
u
s
,
to
id
en
ti
f
y
a
n
d
g
iv
e
i
n
f
o
r
m
atio
n
o
n
t
h
e
s
e
v
er
it
y
o
f
air
p
o
llu
tio
n
to
th
e
p
u
b
lic,
th
e
a
m
b
ie
n
t
air
q
u
ali
t
y
m
ea
s
u
r
e
m
e
n
t
i
n
Ma
la
y
s
ia
is
d
escr
ib
ed
in
ter
m
s
o
f
A
ir
P
o
llu
ta
n
t
I
n
d
ex
(
A
P
I
)
.
B
ased
o
n
th
e
av
er
ag
e
o
f
m
ai
n
p
o
llu
ta
n
ts
n
a
m
e
l
y
s
u
lp
h
u
r
d
io
x
id
e
(
SO2
)
,
n
itro
g
en
d
io
x
id
e
(
NO2
)
,
ca
r
b
o
n
m
o
n
o
x
id
e
(
C
O)
,
o
zo
n
e
(
O2
)
,
p
ar
ticu
late
m
atter
d
ia
m
eter
2
.
5
(
P
M
2
.
5
)
an
d
p
ar
ticu
late
m
atter
d
ia
m
eter
1
0
(
P
M1
0
)
,
th
e
A
P
I
v
alu
e
is
m
ea
s
u
r
ed
.
T
h
e
h
ig
h
est
p
o
ll
u
ta
n
t
’
s
co
n
ce
n
tr
atio
n
w
ill
d
eter
m
i
n
e
th
e
A
P
I
v
al
u
e.
Usu
al
l
y
,
P
M2
.
5
is
th
e
h
i
g
h
est
co
n
ce
n
tr
atio
n
r
ec
o
r
d
ed
c
o
m
p
a
r
ed
to
o
th
er
p
o
llu
tan
t
s
.
A
ir
q
u
alit
y
d
ata
h
a
s
b
ee
n
r
ec
o
r
d
ed
in
Ma
la
y
s
ia
s
in
ce
1
9
9
6
an
d
t
h
e
h
u
g
e
a
m
o
u
n
t
o
f
d
ata
u
s
u
al
l
y
p
r
esen
ted
in
t
h
e
f
o
r
m
o
f
te
x
t
i
n
f
o
r
m
atio
n
.
T
h
u
s
,
air
q
u
alit
y
in
f
o
r
m
at
io
n
ar
e
d
i
f
f
icu
l
t
to
b
e
r
ev
ie
w
ed
,
esp
ec
iall
y
f
o
r
th
e
p
u
b
lic
u
n
d
e
r
s
tan
d
in
g
.
Mo
r
eo
v
er
,
t
h
e
p
u
b
l
ic,
esp
ec
iall
y
th
o
s
e
in
h
i
g
h
r
i
s
k
g
r
o
u
p
s
s
u
c
h
a
s
asth
m
atic
i
n
d
iv
id
u
als,
c
h
ild
r
e
n
,
a
n
d
eld
er
l
y
,
n
ee
d
to
b
e
aler
t
ed
b
ef
o
r
eh
an
d
ab
o
u
t
th
e
ca
s
e
s
o
f
p
o
o
r
air
q
u
alit
y
.
T
h
er
ef
o
r
e,
th
is
s
tu
d
y
u
s
e
ti
m
e
s
er
ies
ap
p
r
o
ac
h
b
y
u
s
i
n
g
clas
s
ical
an
d
m
o
d
er
n
m
et
h
o
d
s
w
h
i
ch
ar
e
B
o
x
-
J
en
k
i
n
s
an
d
A
NN
to
s
o
lv
e
f
o
r
ec
ast
ac
c
u
r
ac
y
i
s
s
u
ed
w
it
h
th
e
p
r
ese
n
ce
o
f
m
is
s
i
n
g
d
ata.
I
t is
i
m
p
o
r
tan
t to
i
m
p
le
m
e
n
t a
ir
q
u
alit
y
m
a
n
ag
e
m
e
n
t a
n
d
p
u
b
lic
w
ar
n
in
g
s
tr
ate
g
ies
f
o
r
p
o
llu
ti
o
n
lev
el
s
th
at
ar
e
ac
ce
p
tab
le
to
th
e
p
u
b
lic.
2.
RE
S
E
ARCH
M
E
T
H
O
D
2
.
1
.
B
o
x
-
J
enk
in
s
m
et
ho
d
B
o
x
-
J
en
k
in
s
m
et
h
o
d
o
r
au
to
r
eg
r
ess
i
v
e
in
te
g
r
ated
m
o
v
i
n
g
av
er
ag
e
(
AR
I
M
A
)
m
et
h
o
d
w
as
f
ir
s
t
in
tr
o
d
u
ce
d
b
y
B
o
x
a
n
d
J
en
k
i
n
s
[
1
8
]
.
Or
ig
i
n
ated
f
r
o
m
t
h
e
au
to
r
eg
r
es
s
iv
e
m
o
d
el
(
A
R
)
,
t
h
e
m
o
v
i
n
g
av
er
a
g
e
m
o
d
el
(
M
A
)
an
d
d
if
f
er
en
ci
n
g
o
r
d
er
o
f
d
k
n
o
w
n
as
th
e
i
n
t
eg
r
ated
(
I
)
m
o
d
el.
T
h
e
s
ea
s
o
n
al
A
R
I
M
A
m
o
d
el
(
SAR
I
M
A
)
i
s
u
s
ed
w
h
e
n
t
h
e
s
ea
s
o
n
al
co
m
p
o
n
en
ts
ar
e
in
cl
u
d
ed
in
t
h
i
s
A
R
I
M
A
m
o
d
el.
T
h
e
g
e
n
er
alize
d
f
o
r
m
o
f
S
A
R
I
M
A
(
p
,
d
,
q
)
(
P
,
D,
Q)
S
m
o
d
el
ca
n
b
e
w
r
itte
n
as
:
(
)
(
)
(
1
−
)
(
1
−
)
=
(
)
(
)
(
1
)
w
h
er
e
(
)
=
1
−
1
−
2
2
−
⋯
−
(
)
=
1
−
1
−
2
2
−
⋯
−
(
)
=
1
−
1
−
2
2
−
⋯
−
(
)
=
1
−
1
−
2
2
−
⋯
−
B
is
d
en
o
ted
as
t
h
e
b
ac
k
w
ar
d
s
h
i
f
t
o
p
er
ato
r
,
d
an
d
D
ar
e
d
e
n
o
ted
as
th
e
n
o
n
-
s
ea
s
o
n
a
l
an
d
s
ea
s
o
n
al
o
r
d
er
s
o
f
d
if
f
er
e
n
ce
r
esp
ec
tiv
el
y
.
B
o
x
-
J
en
k
i
n
s
p
r
o
ce
d
u
r
e
co
n
tain
s
t
h
r
ee
m
ai
n
s
ta
g
e
s
to
b
u
ild
an
AR
I
M
A
m
o
d
el,
i.e
.
m
o
d
el
id
e
n
ti
f
icatio
n
,
m
o
d
el
es
ti
m
atio
n
an
d
m
o
d
el
c
h
ec
k
i
n
g
.
2
.
2
.
Art
if
ici
a
l neura
l net
w
o
r
k
A
r
ti
f
icial
n
e
u
r
al
n
et
w
o
r
k
(
A
N
N)
is
o
n
e
o
f
t
h
e
ar
tific
ial
i
n
tel
lig
e
n
ce
ap
p
r
o
ac
h
es.
I
t
is
o
n
e
o
f
t
h
e
m
o
s
t
ac
cu
r
ate
an
d
w
id
el
y
u
s
ed
f
o
r
ec
asti
n
g
m
et
h
o
d
s
.
Mu
lti
-
la
y
e
r
p
er
ce
p
tr
o
n
(
ML
P
)
o
r
also
k
n
o
w
n
as
t
h
e
f
ee
d
-
f
o
r
w
ar
d
n
e
u
r
al
n
et
w
o
r
k
(
FF
N
N)
is
b
r
o
ad
ly
u
s
ed
as
A
NN
ap
p
r
o
ac
h
[
1
0
,
1
2
]
.
T
h
e
ter
m
p
er
ce
p
tr
o
n
r
ef
er
s
to
th
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
A
r
tifi
cia
l n
eu
r
a
l n
etw
o
r
k
fo
r
ec
a
s
tin
g
p
erfo
r
ma
n
ce
w
ith
... (
N
u
r
Ha
iz
u
m
A
b
d
R
a
h
ma
n
)
35
s
i
m
p
le
s
t
f
o
r
m
o
f
a
n
e
u
r
al
n
et
w
o
r
k
u
s
ed
f
o
r
th
e
cla
s
s
i
f
icatio
n
.
Gen
er
all
y
,
th
e
co
m
p
o
n
e
n
t
s
o
f
ANN
ar
e
n
eu
r
o
n
,
la
y
er
,
ac
tiv
atio
n
f
u
n
ctio
n
a
n
d
w
ei
g
h
t.
ML
P
co
n
s
is
ts
o
f
t
h
r
ee
la
y
e
r
s
i.e
.
in
p
u
t
la
y
er
,
h
id
d
en
l
a
y
e
r
an
d
o
u
tp
u
t
la
y
er
[
1
9
]
as
s
h
o
w
n
in
F
ig
u
r
e
1
.
E
ac
h
in
p
u
t
n
o
d
e
i
n
t
h
e
i
n
p
u
t
la
y
er
w
ill
b
e
f
o
r
war
d
ed
to
th
e
n
e
u
r
o
n
s
w
it
h
t
h
e
ar
r
iv
al
o
f
a
ce
r
tain
w
ei
g
h
t
[
2
0
]
.
I
n
p
u
t
w
i
ll
b
e
p
r
o
ce
s
s
ed
b
y
a
b
ac
k
p
r
o
p
ag
atio
n
f
u
n
ct
io
n
w
h
ich
w
ill
ad
d
u
p
th
e
v
alu
e
s
o
f
all
w
ei
g
h
ts
.
T
h
i
s
s
u
m
w
i
ll
b
e
co
m
p
ar
ed
w
i
th
a
t
h
r
es
h
o
ld
v
al
u
e
g
i
v
e
n
b
y
th
e
ac
ti
v
atio
n
f
u
n
c
tio
n
o
f
ea
ch
n
e
u
r
o
n
.
C
o
m
m
o
n
l
y
,
in
t
h
e
h
id
d
en
la
y
er
,
th
e
ac
ti
v
atio
n
f
u
n
ctio
n
u
s
ed
is
th
e
lo
g
is
tic
f
u
n
ctio
n
,
(
)
=
1
/
(
1
−
e
xp
(
−
)
)
,
m
ea
n
w
h
ile
t
h
e
li
n
ea
r
f
u
n
ctio
n
,
(
)
=
,
is
u
s
ed
at
th
e
o
u
tp
u
t
s
tag
e
.
I
f
th
e
in
p
u
t
is
p
ass
ed
a
ce
r
tai
n
t
h
r
es
h
o
ld
,
th
en
th
e
n
e
u
r
o
n
w
i
ll
b
e
ac
ti
v
a
ted
.
W
h
en
t
h
e
n
eu
r
o
n
s
ar
e
ac
t
iv
ated
,
th
e
n
e
u
r
o
n
w
il
l
tr
an
s
m
i
t
o
u
tp
u
t
v
ia
th
e
o
u
tp
u
t
w
ei
g
h
ts
to
all
n
e
u
r
o
n
s
ass
o
ciate
d
w
it
h
it.
T
h
er
e
ar
e
co
n
s
tan
t
s
o
r
b
ias
(
in
NN
j
ar
g
o
n
)
co
n
n
ec
ted
to
ea
ch
n
e
u
r
o
n
s
a
n
d
o
u
tp
u
t,
d
en
o
t
ed
as o
n
e
[
8
]
.
Fig
u
r
e
1
.
Neu
r
al
n
et
w
o
r
k
ar
c
h
itectu
r
e
ex
a
m
p
le
w
it
h
t
w
o
in
p
u
ts
a
n
d
t
w
o
n
eu
r
o
n
s
.
ML
P
is
tr
ain
ed
b
y
b
ac
k
p
r
o
p
ag
atio
n
lear
n
i
n
g
w
h
ic
h
is
ca
p
ab
le
to
s
o
lv
e
m
o
r
e
co
m
p
le
x
p
r
o
b
lem
s
co
m
p
ar
ed
w
ith
s
in
g
le
la
y
er
n
ets
an
d
o
u
tlier
s
[
2
1
]
.
T
h
is
p
r
o
ce
d
u
r
e
r
ep
ea
ted
ly
m
o
d
if
ie
s
th
e
w
e
ig
h
t
s
o
n
th
e
co
n
n
ec
tio
n
li
n
k
s
i
n
a
NN
s
o
th
at
it
m
i
n
i
m
izes
th
e
d
i
f
f
er
en
c
e
b
et
w
ee
n
ac
t
u
al
o
u
tp
u
t
an
d
t
h
e
d
esire
d
o
u
tp
u
t
.
ML
P
m
o
d
el
in
s
tati
s
tics
m
o
d
ellin
g
f
o
r
ti
m
e
s
er
ies
f
o
r
ec
as
tin
g
ca
n
b
e
co
n
s
id
er
ed
as
a
n
o
n
-
li
n
ea
r
au
to
r
eg
r
es
s
iv
e
(
AR
)
m
o
d
el.
I
n
ti
m
e
s
er
ies
f
o
r
ec
asti
n
g
,
t
h
e
in
p
u
t
n
o
d
e
is
th
e
la
g
(
s
)
o
f
av
a
il
ab
le
h
is
to
r
ical
d
ata
d
eter
m
in
ed
b
ased
o
n
th
e
a
u
to
r
eg
r
ess
i
v
e
o
r
d
er
in
th
e
B
o
x
-
J
en
k
in
s
m
o
d
el
[
8
,
5
]
.
B
ased
o
n
Fig
u
r
e
1
,
th
e
ML
P
r
elatio
n
s
h
ip
b
et
w
ee
n
th
e
o
u
tp
u
t
an
d
th
e
in
p
u
ts
,
−
1
,
−
2
,
.
.
.
,
−
h
as t
h
e
f
o
llo
w
i
n
g
m
at
h
e
m
atic
al
r
ep
r
esen
tatio
n
:
=
0
+
∑
=
1
(
2
)
=
0
+
∑
(
0
+
=
1
∑
−
+
=
1
)
(
3
)
w
h
er
e
;
(
=
1
,
2
,
.
.
,
)
an
d
(
=
1
,
2
,
.
.
.
,
;
=
1
,
2
,
.
.
.
,
)
ar
e
th
e
m
o
d
el
p
ar
a
m
eter
s
t
h
at
ar
e
o
f
te
n
ca
lled
as th
e
co
n
n
ec
tio
n
w
ei
g
h
ts
; p
is
th
e
n
u
m
b
er
o
f
i
n
p
u
t
n
o
d
es a
n
d
q
is
th
e
n
u
m
b
er
o
f
h
id
d
en
n
o
d
es.
2
.
3
.
M
is
s
ing
v
a
lues
i
m
p
uta
t
io
n
2
.
3
.
1
.
Dec
o
m
po
s
it
io
n
m
et
ho
d
T
h
e
b
asic
id
ea
f
o
r
th
e
d
ec
o
m
p
o
s
itio
n
m
et
h
o
d
is
to
d
ec
o
m
p
o
s
e
th
e
p
r
o
b
lem
in
to
s
u
b
p
r
o
b
lem
s
,
w
h
ic
h
is
u
s
ed
as
th
e
s
o
lu
tio
n
f
o
r
v
ar
io
u
s
p
r
o
b
le
m
s
an
d
al
g
o
r
ith
m
s
.
T
h
e
s
u
b
p
r
o
b
le
m
s
ar
e
tr
e
n
d
,
s
ea
s
o
n
a
l,
c
y
cl
ical
an
d
ir
r
eg
u
lar
(
er
r
o
r
)
.
T
h
e
esti
m
ates
f
r
o
m
th
e
s
e
f
ac
to
r
s
ar
e
u
s
ed
to
d
escr
ib
e
th
e
s
er
ies
a
n
d
ca
n
b
e
u
s
ed
to
co
m
p
u
te
p
o
in
t
f
o
r
ec
asts
.
T
h
i
s
m
et
h
o
d
ca
n
b
e
p
r
esen
ted
i
n
to
t
w
o
f
o
r
m
s
;
an
ad
d
iti
v
e
d
ec
o
m
p
o
s
itio
n
an
d
m
u
ltip
licati
v
e
d
ec
o
m
p
o
s
itio
n
.
T
h
e
eq
u
atio
n
o
f
b
o
th
f
o
r
m
s
w
i
th
th
e
f
ac
to
r
s
ca
n
b
e
p
r
esen
ted
as b
elo
w
.
A
d
d
i
ti
v
e
d
ec
o
m
p
o
s
itio
n
:
=
+
+
+
(
4
)
Mu
ltip
licat
iv
e
d
ec
o
m
p
o
s
itio
n
:
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
.
1
,
Ma
r
ch
20
20
:
33
–
39
36
=
×
×
×
(
5
)
w
h
er
e
,
,
an
d
ar
e
tr
en
d
,
s
ea
s
o
n
al,
c
y
clic
a
n
d
ir
r
eg
u
lar
at
ti
m
e
r
esp
ec
tiv
el
y
.
2
.
3
.
2
.
Sp
a
t
ia
l
w
eig
hting
m
et
ho
d
T
h
e
an
al
y
s
is
o
f
m
i
s
s
i
n
g
v
alu
e
s
u
s
i
n
g
th
e
s
p
atial
w
ei
g
h
t
in
g
m
et
h
o
d
s
w
ill
in
v
o
l
v
e
a
tar
g
et
s
tatio
n
w
it
h
s
elec
ted
n
ei
g
h
b
o
r
in
g
s
ta
tio
n
s
.
Gen
er
all
y
,
t
h
e
w
ei
g
h
ti
n
g
m
et
h
o
d
f
o
r
m
u
la
is
g
i
v
en
a
s
f
o
llo
w
s
:
̂
=
∑
=
1
≠
(
6
)
w
h
er
e
is
th
e
e
s
ti
m
ated
v
al
u
e
o
f
th
e
m
is
s
in
g
d
ata
at
th
e
tar
g
et
s
tatio
n
,
(
≠
)
,
N
is
th
e
n
u
m
b
er
o
f
n
eig
h
b
o
r
in
g
s
tatio
n
s
,
is
th
e
o
b
s
er
v
atio
n
at
th
eit
h
n
eig
h
b
o
r
in
g
s
tat
io
n
an
d
is
th
e
w
ei
g
h
t
o
f
th
e
it
h
n
eig
h
b
o
r
in
g
s
tatio
n
w
it
h
co
n
s
t
r
ain
t
=
1
.
T
h
e
ar
ith
m
etic
av
er
a
g
e
(
AA
)
m
et
h
o
d
is
t
h
e
clas
s
icall
y
w
a
y
to
id
en
tify
w
e
ig
h
t.
I
t
co
n
s
id
er
ed
eq
u
al
w
ei
g
h
t f
o
r
ea
ch
s
elec
ted
n
ei
g
h
b
o
r
in
g
s
tatio
n
.
I
t c
an
b
e
d
ef
in
e
d
as:
=
1
(
7
)
T
h
e
s
ec
o
n
d
m
et
h
o
d
is
th
e
n
o
r
m
al
r
atio
(
NR
)
m
e
th
o
d
.
T
h
e
NR
m
eth
o
d
w
as
f
ir
s
tl
y
p
r
o
p
o
s
ed
b
y
[
2
2
]
.
T
h
e
m
et
h
o
d
is
b
ased
o
n
t
h
e
m
ea
n
r
atio
o
f
a
v
ailab
le
d
ata
b
et
w
ee
n
t
h
e
tar
g
e
t
s
tat
io
n
,
a
n
d
th
e
it
h
n
ei
g
h
b
o
r
in
g
s
tatio
n
s
.
T
h
e
m
eth
o
d
is
g
iv
e
n
as f
o
llo
w
s
:
=
1
∑
=
1
(
8
)
w
h
er
e
µ
an
d
µ
ar
e
th
e
s
a
m
p
le
m
ea
n
o
f
th
e
a
v
ailab
le
d
ata
at
th
e
tar
g
et
s
tatio
n
,
an
d
th
e
it
h
n
ei
g
h
b
o
r
in
g
s
tatio
n
s
r
esp
ec
tiv
el
y
.
I
n
1
9
9
2
,
Yo
u
n
g
p
r
o
p
o
s
ed
to
u
s
e
t
h
e
co
r
r
elatio
n
b
et
w
ee
n
th
e
tar
g
et
s
tatio
n
a
n
d
th
e
n
eig
h
b
o
r
in
g
s
tatio
n
as
th
e
w
ei
g
h
tin
g
f
ac
to
r
s
[
2
3
]
.
T
h
e
w
ei
g
h
t
k
n
o
w
n
a
s
th
e
m
o
d
i
f
ied
n
o
r
m
al
r
atio
b
ased
o
n
co
r
r
elatio
n
(
MN
R
)
is
g
i
v
e
n
as
f
o
llo
w
s
:
=
(
−
2
)
2
(
1
−
2
)
−
1
∑
(
−
2
)
2
(
1
−
2
)
−
1
=
1
(
9
)
w
h
er
e
is
th
e
co
r
r
elatio
n
co
e
f
f
icien
t
o
f
th
e
d
ail
y
ti
m
e
s
er
i
es
d
ata
b
etw
ee
n
th
e
tar
g
e
t
s
t
atio
n
an
d
th
e
ith
n
eig
h
b
o
r
in
g
s
tatio
n
s
,
is
th
e
le
n
g
t
h
o
f
d
ata
s
er
ie
s
th
at
ar
e
u
s
e
d
to
co
m
p
u
te
t
h
e
co
r
r
elatio
n
co
ef
f
icie
n
t.
2
.
4
.
E
rr
o
r
m
ea
s
ure
m
ent
L
et
b
e
th
e
ac
tu
al
v
alu
es,
̂
is
th
e
f
o
r
ec
ast
v
alu
e
s
an
d
is
ti
m
e.
T
h
u
s
,
th
e
er
r
o
r
b
e
d
ef
in
ed
a
s
,
=
−
̂
.
T
h
e
m
ea
s
u
r
e
m
e
n
ts
u
s
ed
in
t
h
is
s
t
u
d
y
ar
e
m
ea
n
ab
s
o
lu
te
er
r
o
r
(
MA
E
)
an
d
r
o
o
t
m
ea
n
s
q
u
ar
e
er
r
o
r
(
R
MSE
)
.
T
h
e
eq
u
atio
n
f
o
r
b
o
th
m
ea
s
u
r
e
m
e
n
ts
as
f
o
llo
w
:
MA
E
=
∑
|
−
̂
|
=
1
(
1
0
)
R
MSE
=
√
∑
(
−
̂
)
2
=
1
(
1
1
)
T
h
e
MA
E
a
n
d
R
MSE
ar
e
s
ca
le
d
ep
en
d
en
t
m
ea
s
u
r
e
wh
er
e
b
o
th
n
o
t
s
u
itab
le
to
co
m
p
ar
e
t
h
e
f
o
r
ec
ast
w
it
h
d
if
f
er
en
t
s
ca
le.
B
o
th
o
f
th
ese
m
ea
s
u
r
e
m
en
t
s
ar
e
ea
s
y
to
in
ter
p
r
et
s
in
ce
th
e
er
r
o
r
ca
n
b
e
co
m
p
u
ted
d
ir
ec
tl
y
f
r
o
m
t
h
e
a
ctu
al
a
n
d
f
o
r
ec
ast
v
al
u
es
w
it
h
o
u
t
in
v
o
l
v
in
g
a
n
y
u
n
k
n
o
w
n
p
ar
a
m
eter
th
a
t
n
ee
d
s
to
b
e
esti
m
ated
[
2
4
]
.
3.
RE
SU
L
T
S
A
ND
D
I
SCU
SS
I
O
N
T
h
e
an
al
y
s
es
o
f
A
P
I
d
ata
ar
e
p
r
esen
ted
i
n
t
h
is
s
ec
tio
n
.
S
tatio
n
lo
ca
ted
i
n
J
o
h
o
r
B
ah
r
u
cit
y
w
a
s
ch
o
s
en
f
o
r
th
is
s
tu
d
y
s
i
n
ce
it
is
th
e
ca
p
ital
o
f
J
o
h
o
r
s
tate
an
d
th
e
s
ec
o
n
d
lar
g
e
s
t
m
e
tr
o
p
o
litan
in
Ma
la
y
s
ia.
T
h
u
s
,
it
is
h
o
m
e
to
a
lar
g
e
n
u
m
b
er
o
f
t
h
e
r
eg
io
n
’
s
i
n
d
u
s
tr
ie
s
,
r
esid
en
tial,
an
d
co
m
m
er
cial
h
o
ts
p
o
ts
.
T
h
e
s
tu
d
y
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
A
r
tifi
cia
l n
eu
r
a
l n
etw
o
r
k
fo
r
ec
a
s
tin
g
p
erfo
r
ma
n
ce
w
ith
... (
N
u
r
Ha
iz
u
m
A
b
d
R
a
h
ma
n
)
37
u
s
ed
d
ail
y
d
ata
s
et
f
o
r
s
ev
e
n
y
ea
r
s
,
f
r
o
m
y
ea
r
2
0
0
5
u
n
til
2
0
1
1
.
T
h
e
d
ata
w
er
e
d
iv
id
ed
in
to
t
w
o
d
ata
s
et
s
:
(
1
)
a
tr
ain
i
n
g
s
et
f
r
o
m
2
0
0
5
u
n
til
2
0
1
0
w
it
h
to
tal
o
f
2
1
9
1
o
b
s
er
v
atio
n
s
to
f
in
d
t
h
e
s
u
ita
b
le
A
P
I
m
o
d
el,
an
d
(
2
)
a
test
d
ata
s
et
in
y
ea
r
2
0
1
1
w
it
h
a
to
tal
o
f
3
6
5
o
b
s
er
v
atio
n
s
to
ch
ec
k
t
h
e
m
o
d
el
p
er
f
o
r
m
a
n
ce
.
T
h
e
m
is
s
i
n
g
d
ata
w
er
e
i
n
itial
l
y
e
s
ti
m
ated
b
y
u
s
in
g
t
h
e
d
ec
o
m
p
o
s
itio
n
m
et
h
o
d
.
Fig
u
r
e
2
s
h
o
w
th
e
t
i
m
e
s
er
ies
p
lo
t
f
o
r
b
o
th
tr
ain
i
n
g
a
n
d
test
i
n
g
d
ata
s
et.
Fig
u
r
e
2
.
T
im
e
s
er
ies p
lo
t f
o
r
d
ail
y
A
P
I
in
J
o
h
o
r
B
ah
r
u
Fro
m
Fi
g
u
r
e
2
,
th
e
A
P
I
s
er
i
es
co
n
s
i
s
t
y
ea
r
l
y
s
ea
s
o
n
alit
y
w
h
er
e
t
h
e
d
ata
s
tar
t
to
i
n
cr
ea
s
e
in
th
e
m
id
d
le
o
f
th
e
y
e
a
r
.
T
h
e
s
e
aso
n
alit
y
i
n
d
icate
n
o
n
s
tatio
n
ar
y
d
ata.
T
h
u
s
,
t
h
e
d
ata
tr
a
n
s
f
o
r
m
atio
n
an
d
d
if
f
er
e
n
ci
n
g
i
n
b
o
th
s
ea
s
o
n
al
a
n
d
n
o
n
s
ea
s
o
n
a
l
w
er
e
ca
r
r
ied
o
u
t to
o
b
tain
s
tatio
n
ar
y
s
er
ies.
T
h
er
e
w
er
e
t
w
o
p
o
s
s
ib
le
m
o
d
els
s
h
o
w
ed
s
i
g
n
i
f
ican
t
r
es
u
lt
in
b
o
th
p
ar
a
m
eter
a
n
d
L
-
j
u
n
g
B
o
x
s
tatis
t
ics,
w
h
ic
h
w
er
e
SAR
I
M
A
(
3
,
1
,
3
)
(
0
,
1
,
1)
365
an
d
SA
R
I
MA
(
3
,
1
,
3
)
(
2
,
1
,
0)
365
.
C
o
m
p
a
r
in
g
b
o
th
m
o
d
els
in
ter
m
s
of
R
M
SE,
th
e
S
AR
I
MA
(
3
,
1
,
3
)
(
0
,
1
,
1)
365
w
a
s
t
h
e
b
est
m
o
d
el
to
f
o
r
ec
ast
th
e
d
ail
y
d
ata
b
ec
au
s
e
of
t
h
e
lo
w
es
t
R
MSE
v
al
u
e,
1
0
.
5
2
.
T
h
e
d
ata
p
r
e
-
p
r
o
ce
s
s
in
g
f
o
r
A
N
N
m
e
th
o
d
u
s
ed
in
J
o
h
o
r
B
ah
r
u
d
ail
y
A
P
I
d
ata
in
clu
d
ed
th
e
n
o
r
m
alize
d
ata
tr
an
s
f
o
r
m
atio
n
an
d
s
ca
led
i
n
ter
v
al
tr
an
s
f
o
r
m
atio
n
o
f
[
0
,
1
]
an
d
[
-
1
,
1
]
.
I
n
in
p
u
t
la
y
er
,
t
h
e
i
n
p
u
t
n
o
d
es
w
er
e
id
en
ti
f
ied
b
ased
o
n
all
la
g
s
f
r
o
m
t
h
e
b
est
S
AR
I
MA
m
o
d
el
(
S
AR
I
M
A
(
3
,
1
,
3
)
(
0
,
1
,
1)
365
),
s
ea
s
o
n
al
lag
s
w
h
ich
ar
e
365
an
d
7
3
0
a
n
d
last
l
y
la
g
1
w
it
h
b
ef
o
r
e
an
d
af
ter
s
ea
s
o
n
al
lag
s
(
1
,
3
6
4
,
3
6
5
,
3
6
6
,
7
2
9
,
7
3
0
an
d
7
3
1
)
.
T
a
b
l
e
1
s
h
o
w
ed
th
a
t
th
e
s
m
allest
R
MSE
w
it
h
v
al
u
e
o
f
1
0
.
5
7
w
as
b
y
u
s
i
n
g
all
i
n
p
u
t
la
g
s
f
r
o
m
th
e
b
est S
A
R
I
M
A
m
o
d
el
w
ith
t
h
e
d
ata
tr
an
s
f
o
r
m
atio
n
o
f
[
-
1
,
1
]
.
Mis
s
i
n
g
d
ata
or
i
n
co
m
p
lete
d
ata
m
a
tr
ices
is
a
p
r
o
b
lem
t
h
at
is
r
ep
ea
ted
l
y
e
n
co
u
n
ter
ed
in
m
an
y
ar
ea
s
,
in
cl
u
d
in
g
t
h
e
e
n
v
ir
o
n
m
en
ta
l
r
esear
ch
.
T
h
is
is
co
m
m
o
n
a
n
d
u
n
a
v
o
id
ab
le
p
r
o
b
lem
ca
u
s
ed
by
u
n
s
y
s
te
m
at
ic
d
ata
s
to
r
in
g
,
in
s
tr
u
m
en
t
m
al
f
u
n
cti
o
n
s
,
a
n
d
s
tatio
n
s
r
elo
ca
tio
n
[
2
5
]
.
Miss
in
g
d
ata
ca
n
lead
to
in
s
u
f
f
icien
t
d
ata
s
a
m
p
li
n
g
,
er
r
o
r
s
in
m
ea
s
u
r
e
m
en
ts
,
a
n
d
it
g
i
v
es
a
s
i
g
n
if
ica
n
t
e
f
f
ec
t
to
t
h
e
co
n
cl
u
s
io
n
s
t
h
at
co
u
ld
b
e
d
r
a
w
n
f
r
o
m
th
e
d
ata
[
2
6
]
.
I
n
ti
m
e
s
er
ies,
t
h
e
a
n
al
y
s
i
s
r
eq
u
ir
es
t
h
e
d
ata
to
be
co
n
tin
u
o
u
s
l
y
a
v
a
ilab
le.
T
h
u
s
,
an
ap
p
r
o
p
r
iate
s
tatis
tical
m
eth
o
d
is
i
m
p
o
r
tan
t
to
s
o
lv
e
t
h
e
m
is
s
in
g
d
ata
p
r
o
b
lem
.
T
ab
le
1
.
A
NN
f
o
r
ec
asti
n
g
ac
cu
r
ac
y
b
ased
o
n
R
M
SE
f
o
r
d
aily
A
P
I
in
J
o
h
o
r
B
ah
r
u
L
a
g
s
N
o
r
mal
i
z
e
[
0
,
1
]
[
-
1
,
1
]
A
l
l
1
0
.
5
7
1
0
.
6
4
1
0
.
5
7
S
e
a
so
n
a
l
(
3
6
5
,
7
3
0
)
1
0
.
9
9
1
0
.
9
0
1
0
.
9
9
L
ag
1
wi
t
h
se
aso
n
al
l
ag
s
±
1
1
1
.
0
6
1
1
.
0
7
1
1
.
0
6
T
h
e
s
p
atial
w
e
ig
h
ti
n
g
m
et
h
o
d
w
as
u
s
ed
s
in
ce
n
u
m
b
er
s
o
f
m
o
n
ito
r
in
g
s
tatio
n
s
w
er
e
av
aila
b
le
n
ea
r
to
th
e
s
elec
ted
s
tat
io
n
s
.
I
m
p
u
tati
o
n
m
e
th
o
d
s
w
er
e
test
ed
i
n
d
i
f
f
er
en
t
d
is
ta
n
ce
s
,
1
0
0
k
m
,
1
5
0
k
m
,
an
d
2
0
0
k
m
i
n
o
r
d
er
t
o
test
th
e
m
e
th
o
d
’
s
s
e
n
s
itiv
it
y
s
o
t
h
at
o
p
ti
m
al
r
es
u
lt c
a
n
b
e
p
r
o
d
u
ce
d
.
T
o
ch
ec
k
th
e
s
tab
ilit
y
o
f
t
h
e
m
et
h
o
d
,
s
ix
d
if
f
er
en
t
p
er
ce
n
ta
g
e
r
an
g
e
s
f
r
o
m
5
%
to
3
0
%
w
er
e
ch
o
s
en
.
P
er
f
o
r
m
a
n
ce
of
t
h
e
i
m
p
u
tatio
n
m
eth
o
d
s
ar
e
co
m
p
ar
ed
by
u
s
i
n
g
M
A
E
m
ea
s
u
r
e
m
e
n
t.
L
o
w
e
s
t
M
A
E
i
n
d
icate
b
etter
im
p
u
tatio
n
.
As
s
h
o
w
n
in
T
a
b
l
e
2
,
in
1
0
0
k
m
d
is
tan
ce
,
MN
R
w
a
s
t
h
e
b
est
m
et
h
o
d
in
all
p
er
ce
n
tag
e
m
is
s
i
n
g
.
I
n
1
5
0
k
m
d
is
ta
n
ce
,
th
e
r
es
u
lt
v
ar
ied
b
et
w
ee
n
o
l
d
n
o
r
m
al
r
atio
,
NR
a
n
d
m
o
d
if
ied
n
o
r
m
al
r
atio
,
MN
R
.
T
h
e
b
est
i
m
p
u
tatio
n
f
o
r
NR
w
er
e
in
5
%
a
n
d
1
5
%
p
er
ce
n
tag
e
m
is
s
in
g
,
w
h
ile
f
o
r
MN
R
,
th
e
b
est
i
m
p
u
ta
tio
n
w
er
e
in
1
0
%,
2
0
%,
2
5
%,
an
d
3
0
%
p
er
ce
n
ta
g
e
m
is
s
i
n
g
.
C
o
n
s
is
te
n
t
r
esu
lt
also
o
b
tain
ed
in
200
km
as
it
in
d
icate
d
th
at
th
e
MN
R
was
th
e
b
est
i
m
p
u
tatio
n
m
et
h
o
d
.
T
h
u
s
,
as
c
o
n
cl
u
s
io
n
MN
R
was
t
h
e
b
est
m
et
h
o
d
w
h
ile
d
ec
o
m
p
o
s
itio
n
w
as t
h
e
w
o
r
s
t
m
et
h
o
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
.
1
,
Ma
r
ch
20
20
:
33
–
39
38
T
ab
le
2
.
P
er
f
o
r
m
a
n
ce
m
i
s
s
i
n
g
v
alu
e
s
esti
m
at
io
n
b
ased
o
n
M
A
E
M
e
t
h
o
d
s
M
A
E
5%
1
0
%
1
5
%
2
0
%
2
5
%
3
0
%
1
0
0
k
m
AA
0.
3
64
3
0
.
7
6
1
5
1
.
0
9
5
9
1
.
4
8
2
3
1
.
8
4
6
8
2.
2
23
1
NR
0.
3
54
0
0
.
7
3
2
9
1
.
0
5
4
8
1
.
4
3
0
5
1
.
7
8
0
9
2.
1
40
8
M
N
R
D
e
c
o
mp
o
si
t
i
o
n
0
.
3
5
1
8
0
.
4
2
5
6
0
.
7
2
4
7
0
.
8
7
5
8
1
.
0
4
6
2
1
.
2
5
7
3
1
.
4
1
7
3
1
.
7
0
2
5
1
.
7
6
8
7
2
.
1
5
4
0
2
.
1
1
8
2
2
.
5
2
7
6
1
5
0
k
m
AA
0.
3
08
9
0
.
6
1
3
0
0
.
9
2
3
2
1
.
2
4
4
2
1
.
5
4
3
8
1.
8
71
1
NR
0.
3
05
2
0
.
6
0
8
2
0
.
9
1
3
5
1
.
2
2
9
7
1
.
5
3
1
7
1.
8
56
5
M
N
R
D
e
c
o
mp
o
si
t
i
o
n
0
.
3
0
8
2
0
.
4
2
7
8
0
.
6
0
2
5
0
.
8
3
1
3
0
.
9
1
5
3
1
.
2
4
6
5
1
.
2
2
7
3
1
.
7
0
1
5
1
.
5
2
4
7
2
.
1
1
2
4
1
.
8
4
4
3
2
.
5
2
1
9
2
0
0
k
m
AA
0.
3
15
0
0
.
6
1
9
3
0
.
9
2
9
7
1
.
2
9
1
9
1
.
5
7
8
8
1.
8
71
3
NR
0.
3
11
0
0
.
6
1
0
3
0
.
9
2
1
7
1
.
2
8
1
0
1
.
5
6
0
5
1.
8
48
1
M
N
R
D
e
c
o
mp
o
si
t
i
o
n
0
.
2
9
7
0
0
.
4
3
3
8
0
.
5
9
0
6
0
.
8
4
8
8
0
.
8
8
8
8
1
.
2
7
4
3
1
.
2
3
2
8
1
.
7
5
0
6
1
.
5
0
5
9
2
.
1
1
6
8
1
.
7
7
6
5
2
.
5
2
0
6
T
h
e
s
i
m
ilar
p
r
o
ce
s
s
es
ar
e
co
n
d
u
cted
af
ter
m
is
s
in
g
v
a
lu
es
i
m
p
u
tatio
n
p
er
f
o
r
m
ed
.
T
h
e
n
e
w
p
er
f
o
r
m
a
n
ce
ev
a
lu
at
io
n
s
w
er
e
g
i
v
en
in
T
a
b
l
e
3.
T
h
e
r
esu
l
t
s
h
o
w
n
t
h
at
t
h
e
A
NN
m
o
d
el
can
o
u
tp
er
f
o
r
m
ed
S
A
R
I
M
A
m
o
d
el
an
d
i
m
p
r
o
v
ed
th
e
f
o
r
ec
asti
n
g
p
er
f
o
r
m
a
n
ce
.
T
h
e
MA
E
i
m
p
r
o
v
ed
f
r
o
m
8
.
7
5
to
4
.
5
6
an
d
R
MSE
f
r
o
m
1
0
.
5
7
to
5
.
8
5
.
T
a
b
l
e
3
.
P
er
f
o
r
m
a
n
ce
ev
al
u
ati
o
n
f
o
r
d
ail
y
A
P
I
I
mp
u
t
a
t
i
o
n
M
e
t
h
o
d
F
o
r
e
c
a
st
i
n
g
M
e
t
h
o
d
M
A
E
R
M
S
E
D
e
c
o
mp
o
si
t
o
n
S
A
R
I
M
A
8.
3
7
1
0
.
5
2
A
N
N
8.
7
5
1
0
.
5
7
M
N
R
S
A
R
I
M
A
8.
4
4
1
0
.
6
3
A
N
N
4
.
5
6
5
.
8
5
4.
CO
NCLU
SI
O
N
T
h
is
p
ap
er
p
r
esen
ts
th
e
p
r
o
b
le
m
o
f
h
o
w
to
f
o
r
ec
ast
th
e
h
u
g
e
a
m
o
u
n
t
o
f
A
P
I
d
ataset
co
n
ta
m
in
a
ted
w
it
h
d
i
f
f
er
en
t
r
an
g
e
s
o
f
t
h
e
AP
I
s
ca
les
an
d
ter
m
s
th
a
t
u
s
ed
in
as
s
es
s
in
g
a
n
d
d
escr
ib
in
g
th
e
air
q
u
alit
y
s
tat
u
s
o
n
h
u
m
a
n
h
ea
lt
h
.
T
h
e
o
b
j
ec
tiv
e
o
f
t
h
e
s
t
u
d
y
w
a
s
to
ad
ap
t
t
i
m
e
s
er
ies
m
e
th
o
d
,
clas
s
ical
m
et
h
o
d
an
d
m
o
d
er
n
m
et
h
o
d
th
at
w
o
u
ld
y
ield
s
a
tis
f
ac
to
r
y
w
o
r
k
ab
le
r
esu
l
ts
f
o
r
th
e
A
P
I
f
o
r
ec
asts
w
it
h
m
is
s
in
g
d
ata
p
r
o
b
lem
.
Fro
m
th
e
r
es
u
lt,
it
is
e
v
id
en
t
t
h
at
t
h
e
m
o
d
er
n
m
et
h
o
d
,
a
r
tif
icial
n
eu
r
al
n
et
w
o
r
k
(
A
NN)
g
a
v
e
b
etter
f
o
r
ec
asti
n
g
p
er
f
o
r
m
a
n
ce
s
i
n
f
o
r
ec
asti
n
g
co
m
p
ar
ed
to
SAR
I
M
A
m
o
d
el.
B
esid
es,
A
NN
s
h
o
w
e
d
im
p
r
o
v
e
m
e
n
t
i
n
f
o
r
ec
asti
n
g
a
f
ter
h
ig
h
ac
c
u
r
ac
y
d
ata
i
m
p
u
tatio
n
co
n
d
u
cte
d
.
T
h
e
ab
ilit
y
o
f
ANN
to
ca
p
tu
r
e
co
m
p
lex
d
ata
p
atter
n
w
h
ich
co
n
s
i
s
t b
o
th
lin
ea
r
an
d
n
o
n
lin
ea
r
m
a
k
e
t
h
e
r
e
aso
n
f
o
r
th
e
g
o
o
d
p
er
f
o
r
m
a
n
c
e
o
f
A
NN
m
o
d
el
[
9
]
.
Fo
r
f
u
t
u
r
e
r
ec
o
m
m
e
n
d
atio
n
,
th
is
s
tu
d
y
ca
n
b
e
ex
ten
d
w
it
h
in
p
u
t
f
r
o
m
t
h
e
o
th
er
p
o
llu
tan
t
in
tar
g
et
an
d
n
eig
h
b
o
r
in
g
s
tatio
n
s
.
ACK
NO
WL
E
D
G
E
M
E
NT
S
T
h
is
s
tu
d
y
w
a
s
s
u
p
p
o
r
ted
b
y
Un
i
v
er
s
iti
P
u
tr
a
Ma
la
y
s
ia,
M
ala
y
s
ia
u
n
d
er
P
u
tr
a
-
I
P
M
g
r
an
t
,
9
5
8
7
7
0
0
.
W
e
w
o
u
ld
li
k
e
to
th
a
n
k
s
th
e
D
ep
ar
tm
e
n
t o
f
E
n
v
ir
o
n
m
en
t (
D
OE
)
,
Ma
lay
s
ia
f
o
r
p
r
o
v
id
in
g
a
ir
p
o
llu
tan
ts
d
ata.
RE
F
E
R
E
NC
E
S
[1
]
J.
D.
Cry
e
r
a
n
d
K.
S
.
C
h
a
n
,
T
ime
S
e
rie
s
An
a
lys
is:
wit
h
Ap
p
li
c
a
ti
o
n
s
i
n
R
.
Ne
w
Yo
rk
:
S
p
rin
g
e
r
-
Ve
rlag
Ne
w
Yo
rk
In
c
.
,
2
0
1
0
,
p
p
.
1
-
1
4
1
.
[2
]
J.
G
.
D.
G
o
o
ij
e
r
a
n
d
R.
J.
H
y
n
d
m
a
n
,
“
2
5
Ye
a
rs
o
f
T
i
m
e
se
ries
F
o
re
c
a
stin
g
,
”
In
ter
n
a
ti
o
n
a
l
J
o
u
r
n
a
l
o
f
Fo
re
c
a
sti
n
g
,
v
o
l.
2
2
,
Iss
u
e
3
,
p
p
.
4
4
3
-
4
7
3
,
2
0
0
6
.
[3
]
S
.
S
u
h
a
rt
o
n
o
,
“
T
im
e
S
e
ries
F
o
re
c
a
stin
g
b
y
u
sin
g
S
e
a
so
n
a
l
A
u
to
re
g
re
ss
iv
e
In
teg
ra
ted
M
o
v
in
g
Av
e
ra
g
e
:
S
u
b
se
t
,
M
u
lt
i
p
li
c
a
ti
v
e
o
r
A
d
d
it
iv
e
M
o
d
e
l,
”
J
o
u
rn
a
l
o
f
M
a
th
e
ma
ti
c
s a
n
d
S
ta
ti
stics
,
v
o
l.
7
,
p
p
.
2
0
-
2
7
,
2
0
1
1
.
[4
]
M
.
E.
No
r,
e
t
a
l
.
,
“
F
u
z
z
y
T
i
m
e
S
e
ries
a
n
d
S
A
RIM
A
M
o
d
e
l
f
o
r
F
o
r
e
c
a
stin
g
T
o
u
rist
A
rri
v
a
ls
t
o
Ba
li
,
”
J
u
rn
a
l
T
e
k
n
o
lo
g
i
(
S
c
ien
c
e
s a
n
d
E
n
g
i
n
e
e
rin
g
)
,
v
o
l.
5
7
,
p
p
.
6
9
-
8
1
,
2
0
1
2
.
[5
]
G
.
Zh
a
n
g
,
B.
E.
P
a
tu
w
o
a
n
d
M
.
Y.
Hu
,
“
F
o
re
c
a
stin
g
w
it
h
A
rti
f
icia
l
Ne
u
ra
l
N
e
t
w
o
rk
s:
T
h
e
S
ta
te
o
f
th
e
A
rt,
”
In
ter
n
a
t
io
n
a
l
J
o
u
rn
a
l
o
f
F
o
re
c
a
st
in
g
,
v
o
l.
1
4
,
p
p
.
3
5
-
6
2
,
1
9
9
8
.
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
A
r
tifi
cia
l n
eu
r
a
l n
etw
o
r
k
fo
r
ec
a
s
tin
g
p
erfo
r
ma
n
ce
w
ith
... (
N
u
r
Ha
iz
u
m
A
b
d
R
a
h
ma
n
)
39
[6
]
K.
A
b
h
ish
e
k
,
e
t
a
l
.
,
“
W
e
a
th
e
r
F
o
re
c
a
stin
g
M
o
d
e
l
u
sin
g
A
rti
f
ici
a
l
Ne
u
ra
l
Ne
t
w
o
rk
,
”
Pro
c
e
d
ia
T
e
c
h
n
o
l
o
g
y
,
v
o
l.
4
,
p
p
.
3
1
1
-
3
1
8
,
2
0
1
2
.
[7
]
I.
P
.
P
a
n
a
p
a
k
id
is
a
n
d
A
.
S
.
Da
g
o
u
m
a
s,
“
Da
y
-
A
h
e
a
d
El
e
c
tri
c
it
y
P
r
ice
F
o
re
c
a
stin
g
v
ia
th
e
A
p
p
li
c
a
ti
o
n
o
f
A
rt
if
icia
l
Ne
u
ra
l
Ne
tw
o
rk
b
a
se
d
M
o
d
e
ls,”
Ap
p
li
e
d
E
n
e
rg
y
,
v
o
l.
1
7
2
,
p
p
.
1
3
2
-
1
5
1
,
2
0
1
6
.
[8
]
J.
F
a
ra
w
a
y
a
n
d
C.
Ch
a
tf
ield
,
“
T
im
e
S
e
ries
F
o
re
c
a
stin
g
w
it
h
Ne
u
ra
l
Ne
tw
o
rk
s:
A
Co
m
p
a
ra
ti
v
e
S
tu
d
y
u
sin
g
th
e
A
irl
in
e
Da
ta,”
Ap
p
li
e
d
S
t
a
ti
stics
,
v
o
l.
4
7
,
p
p
.
2
3
1
-
2
5
0
,
1
9
9
8
.
[9
]
S.
Ba
rh
m
i
a
n
d
O.
E.
F
a
tn
i,
“
Ho
u
rly
W
in
d
S
p
e
e
d
F
o
re
c
a
stin
g
b
a
se
d
o
n
S
u
p
p
o
rt
V
e
c
to
r
M
a
c
h
i
n
e
a
n
d
A
rti
f
icia
l
Ne
u
ra
l
Ne
tw
o
rk
s,”
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
Arti
fi
c
i
a
l
I
n
telli
g
e
n
c
e
,
v
o
l.
8
,
p
p
.
2
8
6
-
2
9
1
,
2
0
1
9
.
[1
0
]
W
.
S
.
S
a
rle,
“
Ne
u
ra
l
N
e
t
w
o
rk
s
a
n
d
S
tatisti
c
a
l
M
o
d
e
ls,”
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
Nin
e
tee
n
th
A
n
n
u
a
l
S
A
S
Us
e
rs
Gr
o
u
p
In
ter
n
a
t
io
n
a
l
C
o
n
fer
e
n
c
e
,
1
9
9
4
.
[1
1
]
J.
J.
S
h
i,
“
Re
d
u
c
i
n
g
P
re
d
ictio
n
E
rro
r
b
y
Tan
sf
o
r
m
in
g
In
p
u
t
Da
ta
f
o
r
Ne
u
ra
l
Ne
t
w
o
rk
s,”
J
o
u
rn
a
l
o
f
Co
mp
u
ti
n
g
in
Civil
En
g
in
e
e
rin
g
,
v
o
l
.
1
4
,
p
p
.
1
0
9
-
1
1
6
,
2
0
0
0
.
[1
2
]
A
.
P
a
lme
r,
J.
J.
M
o
n
tan
o
,
a
n
d
A
.
S
e
se
,
“
De
sig
n
in
g
a
n
A
rti
f
icia
l
Ne
u
ra
l
Ne
t
w
o
rk
f
o
r
F
o
re
c
a
stin
g
T
o
u
rism
T
i
m
e
S
e
ries
,
”
T
o
u
rism
M
a
n
a
g
e
me
n
t
,
v
o
l.
2
7
,
p
p
.
7
8
1
-
7
9
0
,
2
0
0
6
.
[1
3
]
A
.
Ku
rt
a
n
d
A
.
B.
Ok
ta
y
,
“
F
o
re
c
a
stin
g
A
ir
P
o
ll
u
tan
t
I
n
d
ica
t
o
r
L
e
v
e
ls
w
it
h
G
e
o
g
r
a
p
h
ic
M
o
d
e
ls
3
Da
y
s
in
A
d
v
a
n
c
e
u
sin
g
Ne
u
ra
l
Ne
tw
o
rk
s,”
Exp
e
rt S
y
ste
ms
wit
h
Ap
p
li
c
a
ti
o
n
s
,
v
o
l.
3
7
,
Iss
u
e
1
2
,
p
p
.
7
9
8
6
-
7
9
9
2
,
2
0
1
0
.
[1
4
]
Z.
Ya
n
g
a
n
d
J.
W
a
n
g
,
“
A Ne
w
A
i
r
Qu
a
li
ty
M
o
n
it
o
rin
g
a
n
d
Early
W
a
rn
in
g
S
y
ste
m
:
Air
Qu
a
li
t
y
A
ss
e
ss
m
e
n
t
a
n
d
A
i
r
P
o
ll
u
tan
t
C
o
n
c
e
n
trati
o
n
P
re
d
icti
o
n
,
”
En
v
ir
o
n
me
n
t
a
l
Res
e
a
rc
h
,
v
o
l.
1
5
8
,
p
p
.
1
0
5
-
1
1
7
,
2
0
1
7
.
[1
5
]
R.
A
f
ro
z
,
M
.
N.
Ha
ss
a
n
,
a
n
d
N.
A
.
Ib
ra
h
im
“
Re
v
ie
w
o
f
A
ir
P
o
l
lu
ti
o
n
a
n
d
He
a
lt
h
Im
p
a
c
ts
in
M
a
lay
sia
,
”
En
v
iro
n
me
n
ta
l
Res
e
a
rc
h
,
v
o
l.
9
2
,
p
p
.
7
1
-
7
7
,
2
0
0
3
.
[1
6
]
De
p
a
rtme
n
t
o
f
En
v
iro
n
m
e
n
t,
“
Ch
ro
n
o
lo
g
y
o
f
Ha
z
e
Ep
iso
d
e
s in
M
a
la
y
s
ia,” P
u
traja
y
a
:
De
p
a
rt
m
e
n
t
o
f
En
v
iro
n
m
e
n
t.
[1
7
]
N.
H.
A
.
Ra
h
m
a
n
,
M
.
H.
L
e
e
,
S
.
S
u
h
a
rt
o
n
o
a
n
d
M
.
T
.
L
a
ti
f
,
“
E
v
a
l
u
a
ti
o
n
P
e
rf
o
rm
a
n
c
e
o
f
T
i
m
e
S
e
ri
e
s
A
p
p
ro
a
c
h
f
o
r
F
o
re
c
a
stin
g
A
ir
P
o
ll
u
ti
o
n
In
d
e
x
in
Jo
h
o
r,
M
a
lay
sia
,
”
S
a
in
s M
a
la
y
s
ia
n
a
,
v
o
l
.
4
5
,
p
p
.
1
6
2
5
–
1
6
3
3
,
2
0
1
6
.
[1
8
]
M.
Kh
a
sh
e
i
a
n
d
M
.
Bij
a
ri,
“
A
N
o
v
e
l
H
y
b
rid
iza
ti
o
n
o
f
A
rti
f
icia
l
Ne
u
ra
l
Ne
t
w
o
rk
s
a
n
d
A
RIM
A
M
o
d
e
ls
f
o
r
T
i
m
e
S
e
ries
F
o
re
c
a
stin
g
,
”
Ap
p
li
e
d
S
o
ft
Co
mp
u
t
in
g
J
o
u
rn
a
l
,
v
o
l.
1
1
,
p
p
.
2
6
6
4
-
2
6
7
5
,
2
0
1
1
.
[1
9
]
F
.
Yu
m
o
n
o
,
e
t
a
l
.
,
“
A
rti
f
icia
l
Ne
u
ra
l
Ne
tw
o
rk
f
o
r
He
a
lt
h
y
Ch
ick
e
n
M
e
a
t
I
d
e
n
ti
f
ica
ti
o
n
,
”
I
n
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
Arti
fi
c
ia
l
I
n
telli
g
e
n
c
e
,
v
o
l.
7
,
p
p
.
6
3
-
7
0
,
2
0
1
8
.
[2
0
]
K.
C.
Ra
n
i
a
n
d
Y.
P
ra
sa
n
t
h
,
“
A
De
c
isio
n
S
y
ste
m
f
o
r
P
re
d
icti
n
g
Dia
b
e
tes
u
sin
g
Ne
u
ra
l
Ne
tw
o
rk
s,
”
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
Arti
fi
c
i
a
l
I
n
telli
g
e
n
c
e
,
v
o
l.
6
,
p
p
.
5
6
-
65,
2
0
1
7
.
[2
1
]
R.
L
a
w
,
“
Ba
c
k
-
P
ro
p
a
g
a
ti
o
n
L
e
a
rn
in
g
in
Im
p
ro
v
in
g
th
e
A
c
c
u
ra
c
y
o
f
Ne
u
ra
l
Ne
t
w
o
rk
-
Ba
se
d
T
o
u
rism
De
m
a
n
d
F
o
re
c
a
stin
g
,
”
T
o
u
rism
M
a
n
a
g
e
me
n
t
,
v
o
l
.
2
1
,
p
p
.
3
3
1
-
3
4
0
,
2
0
0
0
.
[2
2
]
J.
L
.
H.
P
a
u
lh
u
s a
n
d
M
.
A
.
Ko
h
ler,
“
In
terp
o
lati
o
n
o
f
M
issin
g
P
re
c
ip
it
a
ti
o
n
Re
c
o
rd
s,”
Mo
n
t
h
ly W
e
a
th
e
r R
e
v
iew
,
v
o
l.
8
0
,
p
p
.
1
2
9
-
1
3
3
,
1
9
5
2
.
[2
3
]
K.
C.
Yo
u
n
g
,
“
A
T
h
re
e
-
W
a
y
M
o
d
e
l
f
o
r
In
terp
o
latin
g
f
o
r
M
o
n
t
h
ly
P
re
c
ip
it
a
ti
o
n
V
a
lu
e
s,”
M
o
n
t
h
ly
W
e
a
th
e
r
Rev
iew
,
v
o
l.
1
2
0
,
p
p
.
2
5
6
1
-
2
5
6
9
,
1
9
9
2
.
[2
4
]
G
.
El
li
o
tt
,
I.
Ko
m
u
n
jer
a
n
d
A
.
T
i
m
m
e
r
m
a
n
,
“
Esti
m
a
ti
o
n
a
n
d
T
e
stin
g
o
f
F
o
re
c
a
st
Ra
ti
o
n
a
li
ty
u
n
d
e
r
F
lex
ib
le
L
o
ss
,
”
Rev
iew o
f
Eco
n
o
mic
S
tu
d
ies
,
v
o
l.
7
2
,
p
p
.
1
1
0
7
-
1
1
2
5
,
2
0
0
5
.
[2
5
]
H.
Ju
n
n
i
n
e
n
,
e
t
a
l
.
,
”
M
e
th
o
d
s
f
o
r
I
m
p
u
tatio
n
o
f
M
issin
g
V
a
lu
e
s
in
A
ir
Qu
a
li
t
y
Da
t
a
S
e
t
s,”
At
mo
sp
h
e
ric
En
v
iro
n
me
n
t
,
v
o
l.
3
8
,
p
p
.
2
8
9
5
-
2
9
0
7
,
2
0
0
4
.
[2
6
]
J.
S
u
h
a
il
a
,
M
.
D.
S
a
y
a
n
g
,
a
n
d
A
.
A
.
J
e
m
a
in
,
“
Re
v
ise
d
S
p
a
ti
a
l
W
e
i
g
h
ti
n
g
M
e
th
o
d
s
f
o
r
Esti
m
a
ti
o
n
o
f
M
issin
g
Ra
in
f
a
ll
Da
ta,”
Asia
-
Pa
c
if
ic Jo
u
r
n
a
l
o
f
A
tmo
sp
h
e
ric
S
c
ien
c
e
s
,
v
o
l.
4
4
,
p
p
.
9
3
-
1
0
4
,
2
0
0
8
.
Evaluation Warning : The document was created with Spire.PDF for Python.