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
.
117
~
125
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
117
-
1
2
5
117
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
i
.
ia
esco
r
e.
co
m
O
pti
m
i
z
a
tion o
f
a
rtif
icia
l
neura
l ne
tw
o
rk
t
o
po
lo
g
y
f
o
r
m
e
m
b
ra
ne biorea
ctor f
il
tration
usi
ng
respo
nse surfa
ce
m
e
thodo
lo
g
y
Sy
a
hira
I
bra
hi
m
,
No
rha
liza
Abdu
l
Wa
ha
b
,
F
a
t
i
m
a
h S
ha
m
I
s
m
a
i
l,
Ya
ha
y
a
M
d Sa
m
S
c
h
o
o
l
o
f
El
e
c
tri
c
a
l
En
g
in
e
e
ri
n
g
,
F
a
c
u
lt
y
o
f
En
g
in
e
e
rin
g
,
Un
iv
e
rsiti
T
e
k
n
o
lo
g
i
M
a
lay
si
a
,
8
1
3
1
0
S
k
u
d
a
i,
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
9
,
2
0
1
9
R
ev
i
s
ed
J
an
5
,
2
0
20
A
cc
ep
ted
J
an
2
0
,
2
0
20
T
h
e
o
p
ti
m
iza
ti
o
n
o
f
a
rti
f
icia
l
n
e
u
ra
l
n
e
tw
o
rk
s
(
A
NN
)
to
p
o
l
o
g
y
f
o
r
p
re
d
icti
n
g
p
e
rm
e
a
te
f
lu
x
o
f
p
a
l
m
o
il
m
il
l
e
ff
lu
e
n
t
(P
OME
)
in
m
e
m
b
ra
n
e
b
io
re
a
c
to
r
(M
BR)
f
il
tratio
n
h
a
s
b
e
e
n
in
v
e
st
ig
a
ted
u
sin
g
re
sp
o
n
se
su
rf
a
c
e
m
e
th
o
d
o
l
o
g
y
(RS
M
).
A
ra
d
ial
b
a
sis
f
u
n
c
ti
o
n
n
e
u
r
a
l
n
e
tw
o
rk
(RBF
NN
)
m
o
d
e
l,
train
e
d
b
y
g
ra
d
ien
t
d
e
sc
e
n
t
w
it
h
m
o
m
e
n
tu
m
(
G
DM)
a
lg
o
rit
h
m
s
wa
s
d
e
v
e
lo
p
e
d
to
c
o
rre
late
o
u
t
p
u
t
(p
e
rm
e
a
te
f
lu
x
)
to
t
h
e
f
o
u
r
e
x
o
g
e
n
o
u
s
in
p
u
t
v
a
riab
les
(a
irf
lo
w
ra
te,
tran
s
m
e
m
b
ra
n
e
p
re
ss
u
re
,
p
e
rm
e
a
te
p
u
m
p
a
n
d
a
e
ra
ti
o
n
p
u
m
p
).
A
s
e
c
o
n
d
-
o
r
d
e
r
p
o
ly
n
o
m
ial
m
o
d
e
l
wa
s
d
e
v
e
lo
p
e
d
f
ro
m
train
in
g
re
su
lt
s
f
o
r
n
a
tu
ra
l
l
o
g
m
e
a
n
sq
u
a
re
e
rro
r
o
f
5
0
d
e
v
e
lo
p
e
d
A
NN
s
to
g
e
n
e
ra
te
3
D
re
sp
o
n
se
su
rf
a
c
e
s.
T
h
e
o
p
ti
m
u
m
A
N
N
to
p
o
lo
g
y
h
a
d
m
in
im
u
m
ln
M
S
E
w
h
e
n
th
e
n
u
m
b
e
r
o
f
h
id
d
e
n
n
e
u
r
o
n
s,
s
p
re
a
d
,
m
o
m
e
n
tu
m
c
o
e
ff
icie
n
t,
le
a
rn
in
g
ra
te
a
n
d
n
u
m
b
e
r
o
f
e
p
o
c
h
s
w
e
r
e
1
6
,
1
.
4
,
0
.
2
8
,
0
.
3
a
n
d
1
8
5
2
,
re
sp
e
c
ti
v
e
l
y
.
T
h
e
M
S
E
a
n
d
re
g
re
ss
io
n
c
o
e
f
fc
ien
t
o
f
th
e
A
NN
m
o
d
e
l
w
e
r
e
d
e
te
rm
in
e
d
a
s
0
.
0
0
2
2
a
n
d
0
.
9
9
0
6
f
o
r
train
i
n
g
,
0
.
0
0
5
2
a
n
d
0
.
9
8
3
9
f
o
r
tes
ti
n
g
a
n
d
0
.
0
2
1
7
a
n
d
0
.
9
7
0
7
f
o
r
v
a
li
d
a
ti
o
n
d
a
ta
se
ts.
T
h
e
se
re
su
lt
s
c
o
n
f
irm
e
d
th
a
t
c
o
m
b
in
in
g
RS
M
a
n
d
A
NN
wa
s
p
re
c
ise
f
o
r
p
re
d
ictin
g
p
e
rm
e
a
tes
f
lu
x
o
f
P
OME
o
n
M
BR
s
y
ste
m
.
T
h
is
d
e
v
e
lo
p
m
e
n
t
m
a
y
h
a
v
e
sig
n
if
i
c
a
n
t
p
o
ten
ti
a
l
to
im
p
ro
v
e
m
o
d
e
l
a
c
c
u
ra
c
y
a
n
d
re
d
u
c
e
c
o
m
p
u
tatio
n
a
l
ti
m
e
.
K
ey
w
o
r
d
s
:
A
r
ti
f
icial
n
e
u
r
al
n
et
w
o
r
k
Me
m
b
r
an
e
b
io
r
ea
cto
r
P
alm
o
il
m
ill e
f
f
lu
e
n
t
R
esp
o
n
s
e
s
u
r
f
ac
e
m
eth
o
d
o
lo
g
y
T
o
p
o
lo
g
y
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
:
No
r
h
aliza
A
b
d
u
l W
ah
ab
,
C
o
n
tr
o
l &
Me
c
h
atr
o
n
ic
s
E
n
g
i
n
ee
r
in
g
,
Sc
h
o
o
l o
f
E
lectr
ical
E
n
g
i
n
ee
r
i
n
g
,
Facu
lt
y
o
f
E
n
g
i
n
ee
r
i
n
g
,
Un
i
v
er
s
iti T
ek
n
o
lo
g
i M
ala
y
s
ia
,
J
o
h
o
r
,
Ma
lay
s
ia.
E
m
ail: a
liza
@
f
k
e.
u
t
m
.
m
y
1.
I
NT
RO
D
UCT
I
O
N
Ma
la
y
s
ia
is
o
n
e
o
f
t
h
e
w
o
r
ld
’
s
lead
in
g
p
r
o
d
u
ce
r
s
i
n
p
al
m
o
il
in
d
u
s
tr
ie
s
[
1
]
.
A
lo
n
g
w
it
h
t
h
e
i
n
cr
ea
s
e
o
f
p
r
o
d
u
ctio
n
ca
p
ac
it
y
o
f
p
al
m
o
il
ev
er
y
y
ea
r
,
a
lar
g
e
a
m
o
u
n
t
o
f
w
as
te
w
ater
w
as
al
s
o
b
ein
g
g
e
n
er
ated
.
T
h
ese
u
n
co
n
tr
o
lled
d
is
ch
ar
g
es
o
f
u
n
tr
ea
ted
p
alm
o
il
m
ill
e
f
f
lu
e
n
t
(
P
OM
E
)
m
a
y
ca
u
s
e
p
o
llu
tio
n
to
t
h
e
w
ater
w
a
y
s
[
2
]
.
I
n
co
m
p
ar
is
o
n
w
it
h
co
n
v
e
n
tio
n
al
ac
ti
v
ated
s
y
s
te
m
,
m
e
m
b
r
an
e
s
y
s
te
m
i
s
p
r
ef
er
ab
le
to
t
r
ea
t
P
OM
E
d
u
e
to
its
s
i
m
p
le
o
p
er
atio
n
,
ea
s
y
t
o
s
ca
le
-
u
p
,
less
w
ei
g
h
t
a
n
d
s
p
ac
e
r
eq
u
ir
e
m
en
t
s
an
d
h
i
g
h
ef
f
icie
n
c
y
[
3
]
.
Me
m
b
r
an
e
b
io
r
ea
cto
r
(
MB
R
)
h
as
b
ee
n
p
r
o
v
en
as
a
r
eliab
le
tech
n
o
l
o
g
y
i
n
tr
ea
tin
g
a
w
id
e
r
an
g
e
o
f
w
ater
s
u
c
h
as
w
a
s
te
w
ater
,
g
r
o
u
n
d
w
ater
an
d
s
u
r
f
ac
e
w
ater
.
Ho
w
e
v
er
,
f
o
u
lin
g
p
h
en
o
m
en
a
i
s
th
e
m
ain
d
r
a
w
b
ac
k
o
f
MB
R
s
y
s
te
m
w
h
ich
co
n
tr
ib
u
te
to
h
i
g
h
e
n
er
g
y
co
n
s
u
m
p
t
io
n
a
n
d
m
ai
n
ten
a
n
ce
co
s
t
[
4
]
.
A
cc
o
r
d
in
g
to
[5
-
7]
,
f
o
u
li
n
g
m
a
y
v
ar
ies
w
i
th
ti
m
e
d
u
r
i
n
g
o
p
er
atio
n
an
d
th
is
v
ar
iatio
n
ca
n
b
e
m
in
i
m
ized
b
y
co
n
tr
o
llin
g
t
h
e
f
o
u
li
n
g
v
ar
iab
les [
8
]
.
Fo
u
li
n
g
ca
n
b
e
co
n
tr
o
lled
an
d
r
ed
u
ce
d
u
s
in
g
s
ev
er
al
h
y
d
r
o
d
y
n
a
m
ic
co
n
d
itio
n
tech
n
iq
u
e
s
s
u
ch
as
air
b
u
b
b
le
(
ae
r
atio
n
)
co
n
tr
o
l,
r
ela
x
atio
n
,
b
ac
k
w
as
h
in
g
,
an
d
c
h
e
m
ical
clea
n
i
n
g
[8
-
10]
.
I
t
w
a
s
f
o
u
n
d
th
at
s
till
li
ttle
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
:
117
–
1
2
5
118
w
o
r
k
co
n
d
u
c
ted
o
n
th
e
d
ev
el
o
p
m
e
n
t
o
f
m
o
d
elli
n
g
an
d
o
p
t
i
m
izatio
n
f
o
r
t
h
e
o
p
er
atio
n
c
o
n
d
itio
n
o
f
P
OM
E
u
s
i
n
g
MB
R
.
Mo
s
t
o
f
t
h
e
w
o
r
k
s
f
o
cu
s
ed
m
ain
l
y
o
n
b
io
lo
g
i
ca
l
r
ed
u
ctio
n
o
f
P
OM
E
u
s
i
n
g
MB
R
f
iltra
tio
n
[9
-
12]
.
M
o
d
elin
g
o
f
m
e
m
b
r
an
e
p
r
o
ce
s
s
,
in
v
o
l
v
i
n
g
w
i
th
lar
g
e
n
u
m
b
er
o
f
p
ar
a
m
eter
s
t
h
at
n
ee
d
s
to
b
e
c
o
n
s
id
er
ed
is
n
o
t a
n
ea
s
y
ta
s
k
.
R
ec
en
t
l
y
,
m
o
d
eli
n
g
o
f
m
e
m
b
r
an
e
p
r
o
ce
s
s
u
s
i
n
g
n
e
u
r
al
n
et
w
o
r
k
h
as
r
ec
ei
v
ed
en
o
r
m
o
u
s
atten
t
io
n
b
ec
au
s
e
o
f
t
h
eir
ab
ilit
y
i
n
m
o
d
elin
g
an
d
p
r
ed
ictio
n
o
f
co
m
p
l
ex
p
r
o
ce
s
s
es.
A
NN
h
a
s
b
ee
n
s
u
cc
es
s
f
u
ll
y
ap
p
lied
to
p
r
ed
ict
o
il
y
w
a
s
te
w
ater
[
1
3
-
14]
,
p
er
m
ea
te
f
lu
x
o
f
alb
u
m
i
n
f
r
o
m
s
er
u
m
b
o
v
i
n
e
[
1
5
]
an
d
p
al
m
o
il
m
ill
w
a
s
te
w
ater
[
1
6
-
17]
.
I
n
ad
d
iti
o
n
,
a
g
o
o
d
u
n
d
er
s
tan
d
i
n
g
o
f
f
ac
to
r
s
t
h
at
af
f
ec
t
A
NN
s
m
o
d
el
p
er
f
o
r
m
an
ce
is
cr
u
cial
to
p
r
ed
ict
th
e
o
p
ti
m
u
m
v
al
u
e
o
f
th
e
n
u
m
b
er
o
f
iter
atio
n
s
,
lear
n
i
n
g
r
ate,
m
o
m
en
tu
m
co
ef
f
icie
n
t,
n
u
m
b
er
o
f
h
id
d
en
la
y
er
s
an
d
n
u
m
b
er
o
f
h
id
d
en
n
e
u
r
o
n
s
.
T
h
e
p
ar
a
m
eter
s
ar
e
v
ar
ied
u
n
t
i
l
th
eir
o
p
ti
m
al
v
al
u
e
ar
e
d
eter
m
i
n
ed
[
1
8
]
.
D
e
ter
m
i
n
atio
n
o
f
t
h
e
b
est
A
N
N
to
p
o
lo
g
y
i
s
i
m
p
o
r
ta
n
t
b
ec
au
s
e
it
a
f
f
ec
ts
th
e
w
e
ig
h
t
a
n
d
b
ias.
Us
u
all
y
it
p
er
f
o
r
m
ed
b
y
tr
ial
a
n
d
er
r
o
r
[
1
9
-
20]
o
r
o
n
e
-
v
ar
iab
le
-
at
-
ti
m
e
(
OV
A
T
)
[
2
1
-
22]
w
h
er
e
th
i
s
p
r
o
ce
d
u
r
e
is
v
er
y
ti
m
e
-
co
n
s
u
m
i
n
g
a
n
d
m
o
n
o
to
n
o
u
s
ta
s
k
.
A
cc
o
r
d
in
g
to
[
2
3
]
f
o
r
th
r
ee
d
if
f
er
en
t
le
v
el
o
f
ea
c
h
A
N
N
v
ar
iab
le
s
,
ab
o
u
t
2
4
5
(
=
3
5
)
d
if
f
er
e
n
t
co
n
f
i
g
u
r
atio
n
o
f
A
N
N
w
o
u
ld
b
e
r
eq
u
ir
ed
.
T
h
er
e
is
n
o
s
p
ec
i
f
ic
r
u
l
e
u
s
ed
i
n
s
elec
ti
n
g
t
h
e
v
alu
e
o
f
v
ar
iab
les
in
A
NN.
I
t
is
d
ep
e
n
d
en
t
o
n
t
h
e
co
m
p
lex
i
t
y
o
f
t
h
e
m
o
d
eled
s
y
s
te
m
.
T
h
u
s
,
it is
o
f
i
m
p
o
r
tan
ce
f
o
r
r
e
s
e
ar
ch
er
s
i
n
o
r
d
er
to
f
in
d
a
s
ta
n
d
ar
d
tech
n
iq
u
e
to
s
o
lv
e
th
e
p
r
o
b
lem
s
ass
o
ciate
d
w
it
h
t
h
e
A
NN
d
ev
e
lo
p
m
en
t.
R
esp
o
n
s
e
s
u
r
f
ac
e
m
eth
o
d
o
lo
g
y
(
R
SM)
as
a
co
llectio
n
o
f
s
ta
tis
tical
a
n
d
m
at
h
e
m
atica
l
tec
h
n
iq
u
e
s
h
a
s
a
ca
p
ab
ilit
y
f
o
r
o
p
ti
m
izin
g
o
b
j
ec
tiv
e
f
u
n
ctio
n
s
.
I
t
i
s
a
p
o
w
er
f
u
l
o
p
ti
m
u
m
d
esig
n
to
o
l
i
n
m
an
y
e
n
g
i
n
ee
r
i
n
g
ap
p
licatio
n
s
an
d
ca
n
p
r
o
v
id
e
a
cc
u
r
ate
m
o
d
els.
R
SM
tech
n
iq
u
e
h
a
s
b
ee
n
u
s
ed
to
d
eter
m
in
e
th
e
A
N
N
to
p
o
lo
g
y
ap
p
lied
f
o
r
m
u
lti
-
la
y
er
f
ee
d
f
o
r
w
ar
d
w
it
h
b
ac
k
p
r
o
p
ag
atio
n
n
eu
r
al
n
e
t
w
o
r
k
[
2
3
-
24]
.
I
t
i
s
a
ls
o
u
s
ed
to
f
i
n
d
t
h
e
o
p
tim
u
m
v
al
u
e
o
f
n
eu
r
o
n
n
u
m
b
er
in
f
ir
s
t
an
d
s
ec
o
n
d
h
id
d
en
la
y
er
s
[
1
8
]
.
T
h
is
p
ap
er
aim
s
f
o
r
th
e
d
ev
elo
p
m
e
n
t
o
f
r
ad
ial
b
asis
f
u
n
ctio
n
n
e
u
r
al
n
et
w
o
r
k
(
R
B
FNN)
m
o
d
el
s
f
o
r
p
r
ed
ictio
n
o
f
p
er
m
ea
te
f
lu
x
d
u
r
i
n
g
MB
R
f
iltra
tio
n
o
f
P
OM
E
w
aste
w
at
er
.
I
n
th
is
ca
s
e,
th
e
R
SM
i
s
p
r
o
p
o
s
ed
to
f
in
d
th
e
o
p
tim
u
m
ANN
to
p
o
lo
g
y
to
ac
h
iev
e
m
i
n
i
m
u
m
m
ea
n
s
q
u
ar
e
er
r
o
r
t
o
im
p
r
o
v
e
t
h
e
p
er
f
o
r
m
an
ce
o
f
t
h
e
m
o
d
el
2.
RE
S
E
ARCH
M
E
T
H
O
D
2
.
1
.
Da
t
a
c
o
llect
io
n
T
h
e
ex
p
er
im
e
n
ts
w
er
e
ca
r
r
ied
o
u
t
u
s
in
g
m
e
m
b
r
an
e
b
io
r
ea
cto
r
f
o
r
p
alm
o
il
m
i
ll
ef
f
l
u
en
t
(
P
OM
E
)
w
it
h
w
o
r
k
i
n
g
v
o
lu
m
e
o
f
2
0
L
.
T
h
e
s
a
m
p
le
o
f
P
OM
E
w
as
t
ak
en
f
r
o
m
Sed
en
a
k
P
al
m
Oil
Mill
Sd
n
.
B
h
d
.
in
J
o
h
o
r
,
Ma
lay
s
ia
w
i
th
t
h
e
w
o
r
k
in
g
te
m
p
er
at
u
r
e
at
2
7
±
1
°
C
.
T
h
er
e
ar
e
f
o
u
r
i
n
p
u
t
v
ar
ia
b
les
f
o
r
t
h
e
P
OM
E
m
o
d
el
in
cl
u
d
i
n
g
tr
a
n
s
m
e
m
b
r
an
e
p
r
ess
u
r
e
(
T
MP)
,
air
f
lo
w
r
ate,
p
er
m
ea
te
s
p
u
m
p
an
d
a
er
atio
n
p
u
m
p
.
T
h
e
o
u
tp
u
t
v
ar
iab
le
i
s
p
er
m
ea
te
f
l
u
x
.
T
h
e
an
al
y
s
is
o
f
r
eq
u
ir
ed
d
ata
w
a
s
ca
r
r
ied
o
u
t
b
y
u
s
in
g
M
A
T
L
A
B
R
2
0
1
4
a
an
d
Desi
g
n
E
x
p
er
t
v
er
s
io
n
7
.
1
.
6
to
o
b
tain
th
e
r
esp
o
n
s
e
s
u
r
f
ac
e
an
d
t
h
e
co
n
to
u
r
s
p
lo
t.
T
h
e
to
tal
o
f
1
6
0
2
d
ata
f
o
r
ea
ch
p
ar
a
m
eter
w
er
e
co
llected
f
r
o
m
th
e
e
x
p
er
i
m
e
n
t i
n
cl
u
d
in
g
air
f
lo
w
r
ate,
T
MP
,
p
er
m
ea
te
p
u
m
p
,
ae
r
a
tio
n
p
u
m
p
a
n
d
p
er
m
ea
te
f
lu
x
.
Fi
g
u
r
e
1
s
h
o
w
s
t
h
e
f
l
u
x
w
a
s
r
ap
id
ly
d
ec
r
ea
s
ed
a
f
ter
t
h
e
air
f
lo
w
r
ate
w
as
d
ec
r
ea
s
ed
f
r
o
m
8
S
L
P
M
to
5
SL
P
M.
Fig
u
r
e
1
.
Data
f
r
o
m
MB
R
f
iltr
atio
n
ex
p
er
i
m
en
t
0
200
400
600
800
1000
1200
1400
1600
1800
0
5
10
D
a
t
a
A
i
r
f
l
o
w
(
S
L
P
M
)
0
200
400
600
800
1000
1200
1400
1600
1800
0
500
D
a
t
a
T
M
P
(
m
b
a
r
)
0
200
400
600
800
1000
1200
1400
1600
1800
0
1
2
D
a
t
a
P
u
m
p
(
V
o
l
t
)
0
200
400
600
800
1000
1200
1400
1600
1800
1
1
.
5
2
D
a
t
a
A
e
r
a
t
i
o
n
v
a
l
v
e
(
v
o
l
t
)
0
200
400
600
800
1000
1200
1400
1600
1800
0
20
40
D
a
t
a
F
l
u
x
(
L
/
m
2
h
)
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
Op
timiz
a
tio
n
o
f a
r
tifi
cia
l n
eu
r
a
l n
etw
o
r
k
to
p
o
lo
g
y
fo
r
mem
b
r
a
n
e
b
io
r
ea
cto
r
…
(
S
ya
h
ir
a
I
b
r
a
h
im
)
119
2
.
2
.
M
o
del dev
elo
p
m
ent
I
n
t
h
is
w
o
r
k
,
t
h
e
R
B
F
NN
m
o
d
el
w
as
u
s
ed
to
p
r
ed
ict
th
e
p
er
m
ea
te
f
lu
x
o
f
P
OM
E
m
e
m
b
r
an
e
b
io
r
ea
cto
r
.
B
ef
o
r
e
th
at,
all
d
at
a
n
ee
d
to
u
n
d
er
g
o
d
ata
p
r
e
-
p
r
o
ce
s
s
in
g
s
ta
g
e
s
o
ca
lled
n
o
r
m
aliza
tio
n
.
Si
n
ce
th
e
in
p
u
t
d
ata
f
o
r
t
h
is
s
y
s
te
m
i
n
v
o
lv
ed
w
ith
d
if
f
er
en
t
m
a
g
n
it
u
d
e
v
al
u
e
an
d
s
ca
le,
all
d
ata
w
er
e
n
o
r
m
alize
d
i
n
to
a
m
i
n
i
m
u
m
o
f
+0
an
d
m
a
x
i
m
u
m
o
f
+1
.
T
h
is
p
r
o
ce
d
u
r
e
p
r
ev
en
ts
t
h
e
tr
an
s
f
er
f
u
n
ctio
n
m
o
d
el
f
r
o
m
b
ec
o
m
in
g
s
atu
r
ated
[
2
5
]
.
E
q
u
atio
n
(
1
)
u
s
ed
f
o
r
n
o
r
m
a
lizatio
n
g
i
v
en
a
s
:
′
=
−
−
(
1
)
w
h
er
e
′
is
th
e
s
ca
le
v
alu
e,
is
th
e
s
a
m
p
le
v
al
u
e
w
h
ile
an
d
ar
e
m
in
i
m
u
m
an
d
m
a
x
i
m
u
m
v
al
u
e
o
f
d
ata.
T
h
e
p
er
m
ea
te
f
l
u
x
w
a
s
d
eter
m
i
n
ed
as g
i
v
en
i
n
(
2
)
:
=
(
2
)
w
h
er
e
is
th
e
p
er
m
ea
te
f
l
u
x
i
n
(
−
2
ℎ
−
1
)
,
is
t
h
e
v
o
lu
m
e
f
lo
w
r
ate
in
li
ter
,
is
m
e
m
b
r
a
n
e
s
u
r
f
ac
e
ar
ea
(
2
)
an
d
is
th
e
ti
m
e
(
ℎ
)
.
T
o
in
v
esti
g
a
te
th
e
f
ea
s
ib
ilit
y
o
f
th
e
p
r
ed
ictiv
e
m
o
d
el,
th
e
co
llect
ed
d
ata
w
er
e
s
ep
ar
ated
in
to
th
r
ee
d
ata
s
et
s
.
Fro
m
th
e
to
tal,
6
5
1
f
o
r
tr
ain
in
g
d
ata
s
et,
w
h
er
e
t
h
is
d
ata
i
n
cl
u
d
ed
th
e
tr
an
s
itio
n
b
et
w
ee
n
h
ig
h
a
n
d
lo
w
air
f
lo
w
r
ate.
T
h
e
5
0
0
f
o
r
test
in
g
d
at
a
s
et
w
as
tak
e
n
f
r
o
m
th
e
h
i
g
h
air
f
lo
w
an
d
f
i
n
all
y
,
4
5
1
f
o
r
v
alid
atio
n
d
ata
s
et
w
a
s
tak
e
n
f
r
o
m
t
h
e
lo
w
air
f
lo
w
r
ate.
T
h
e
tr
ain
in
g
d
ata
w
a
s
u
s
ed
to
co
m
p
u
te
th
e
n
et
w
o
r
k
p
ar
a
m
eter
s
.
T
h
e
test
i
n
g
d
ata
w
as
u
s
ed
to
ass
e
s
s
th
e
p
r
ed
ictiv
e
ab
ilit
y
o
f
th
e
g
e
n
er
ated
m
o
d
el,
w
h
ile
th
e
r
e
m
ai
n
in
g
v
al
id
atio
n
d
ata
w
as
s
u
b
s
eq
u
e
n
t
y
u
s
ed
to
e
n
s
u
r
e
r
o
b
u
s
t
n
es
s
o
f
t
h
e
n
et
w
o
r
k
p
ar
a
m
eter
s
a
n
d
t
o
av
o
id
o
v
er
-
tr
ai
n
i
n
g
[
2
6
]
.
T
h
e
a
m
o
u
n
t
o
f
tr
ain
in
g
d
ata
s
e
t
m
u
s
t
b
e
eq
u
al
o
r
lar
g
er
t
h
an
th
e
a
m
o
u
n
t
o
f
tes
tin
g
an
d
v
alid
atio
n
d
ata
s
et
to
a
v
o
id
ex
tr
ap
o
latio
n
p
r
o
b
l
em
[
2
7
]
.
I
n
t
h
is
p
ap
er
,
th
r
ee
la
y
er
s
o
f
R
B
FNN
w
h
ic
h
ar
e
i
n
p
u
t,
o
u
tp
u
t
a
n
d
h
id
d
en
w
er
e
u
s
ed
.
T
h
e
n
o
n
-
li
n
ea
r
tr
an
s
f
er
f
u
n
ctio
n
o
f
h
y
p
er
b
o
li
c
tan
g
en
t
s
i
g
m
o
id
w
as
u
s
ed
i
n
t
h
e
h
id
d
en
la
y
er
an
d
t
h
e
l
in
ea
r
tr
an
s
f
er
f
u
n
ctio
n
o
f
p
u
r
eli
n
w
a
s
c
h
o
s
en
f
o
r
t
h
e
o
u
tp
u
t
la
y
er
to
p
r
o
d
u
ce
a
co
n
tin
u
o
u
s
o
u
tp
u
t.
T
h
e
R
SM
is
u
s
ed
to
f
i
n
d
th
e
o
p
tim
a
l
v
al
u
e
f
o
r
ea
ch
lear
n
i
n
g
p
ar
a
m
eter
s
o
f
R
B
FN
N
m
o
d
el.
5
0
d
if
f
er
en
t
ex
p
er
i
m
e
n
t
s
o
f
ce
n
tr
al
co
m
p
o
s
it
e
d
esig
n
(
C
C
D)
f
o
r
f
i
v
e
n
u
m
er
ical
f
ac
to
r
s
(
n
u
m
b
er
o
f
n
e
u
r
o
n
s
,
n
u
m
b
er
o
f
s
p
r
ea
d
,
lear
n
i
n
g
r
ate,
m
o
m
en
t
u
m
r
ate
an
d
n
u
m
b
er
o
f
ep
o
ch
)
w
it
h
ei
g
h
t
r
ep
etitio
n
at
ce
n
t
er
p
o
in
t
w
er
e
u
s
ed
.
Fiv
e
n
u
m
er
ical
f
a
cto
r
s
an
d
s
i
m
u
lat
io
n
r
an
k
s
f
o
r
R
B
FNN
ar
e
s
h
o
w
n
in
T
ab
le
1
.
T
h
e
ex
p
er
im
e
n
tal
r
e
s
u
l
ts
o
f
t
h
e
C
C
D
w
er
e
f
itted
w
i
th
a
s
ec
o
n
d
-
o
r
d
er
p
o
ly
n
o
m
ial
eq
u
atio
n
b
y
a
m
u
ltip
le
r
eg
r
e
s
s
io
n
tec
h
n
iq
u
e.
Fo
r
p
r
ed
ictin
g
t
h
e
o
p
tia
p
o
in
t,
t
h
e
q
u
ad
r
atic
m
o
d
el
is
ex
p
r
ess
ed
b
y
(
3
)
:
Y=
0
+
∑
=
1
+
∑
=
1
+
∑
∑
=
+
1
+
−
1
=
1
(
3
)
w
h
er
e
is
th
e
r
e
s
p
o
n
s
e
l
n
(
MS
E
)
,
0
,
,
an
d
ar
e
r
eg
r
ess
io
n
co
ef
f
i
cien
ts
f
o
r
i
n
ter
ce
p
t,
lin
ea
r
q
u
a
d
r
atic
an
d
in
ter
ac
tio
n
co
ef
f
icien
ts
,
r
esp
ec
tiv
el
y
an
d
an
d
ar
e
in
d
ep
en
d
en
t
v
ar
iab
les an
d
k
is
a
n
u
m
b
er
o
f
f
ac
to
r
s
.
T
ab
le
1
.
T
h
e
r
an
g
e
o
f
tr
ain
i
n
g
p
ar
am
eter
s
R
B
F
N
N
P
a
r
a
me
t
e
r
L
o
w
H
i
g
h
1
:
N
o
.
o
f
n
e
u
r
o
n
s
1
20
2
:
S
p
r
e
a
d
0
.
1
2
3
:
L
e
a
r
n
i
n
g
r
a
t
e
0
.
0
1
0
.
4
4
:
M
o
me
n
t
u
m ra
t
e
0
.
0
1
0
.
9
5
:
N
u
mb
e
r
o
f
Ep
o
c
h
10
3
0
0
0
A
ll
A
N
N
to
p
o
lo
g
ies
w
er
e
d
esig
n
ed
an
d
tr
ain
ed
u
s
i
n
g
R
S
M.
T
h
e
o
b
tain
ed
q
u
ad
r
atic
e
q
u
atio
n
w
a
s
s
o
lv
ed
u
s
i
n
g
r
esp
o
n
s
e
o
p
ti
m
i
ze
r
o
f
R
SM
u
n
til
t
h
e
o
p
ti
m
u
m
co
n
d
itio
n
to
m
in
i
m
ize
M
S
E
(
r
esp
o
n
s
e
v
ar
iab
le)
d
ata
s
et
w
as
f
o
u
n
d
.
T
h
e
MSE
w
er
e
tr
a
n
s
f
o
r
m
ed
i
n
to
n
at
u
r
al
lo
g
f
u
n
ctio
n
(
l
n
(
MSE
)
)
w
it
h
α
eq
u
al
to
1
.
I
n
th
is
ca
s
e,
th
e
d
is
tr
ib
u
tio
n
o
f
t
h
e
r
es
p
o
n
s
e
v
ar
iab
le
b
ec
o
m
e
clo
s
er
to
th
e
n
o
r
m
al
d
is
tr
ib
u
tio
n
[
2
4
]
.
2
.3
.
P
er
f
o
rm
a
nce
ev
a
lua
t
io
n
=
1
∑
(
−
)
2
=
1
(
4
)
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
:
117
–
1
2
5
120
=
√
1
∑
(
−
)
2
=
1
(
5
)
2
=
1
−
∑
(
−
)
2
=
1
∑
(
−
̅
)
2
=
1
(
6
)
w
h
er
e
is
t
h
e
p
r
ed
icted
o
u
tp
u
t
f
r
o
m
o
b
s
er
v
a
tio
n
i
,
is
th
e
ex
p
er
i
m
en
ta
l
o
r
ac
tu
al
o
u
tp
u
t
f
o
r
m
o
b
s
er
v
atio
n
i
,
̅
is
t
h
e
av
er
a
g
e
v
alu
e
o
f
t
h
e
e
x
p
er
i
m
e
n
tal
o
u
t
p
u
t
an
d
N
is
th
e
n
u
m
b
er
o
f
d
ata.
S
m
al
ler
v
al
u
es
o
f
MSE
an
d
R
MSE
m
ea
n
a
b
etter
p
er
f
o
r
m
a
n
ce
o
f
t
h
e
m
o
d
el.
Fo
r
R
2
eq
u
al
to
1
r
ev
ea
ls
th
at
th
e
r
eg
r
es
s
io
n
l
i
n
e
p
er
f
ec
tl
y
f
it t
h
e
d
ata
[
2
6
]
.
3.
RE
SU
L
T
S
A
ND
D
I
SCU
SS
I
O
N
T
h
e
r
elatio
n
s
h
ip
b
et
w
ee
n
th
e
p
er
m
ea
te
f
lu
x
a
n
d
t
h
e
i
n
d
ep
en
d
en
t
p
ar
a
m
eter
s
,
n
a
m
el
y
n
u
m
b
er
o
f
n
eu
r
o
n
(
1
)
,
s
p
r
ea
d
(
2
)
,
lear
n
in
g
r
ate
(
3
)
,
m
o
m
e
n
t
u
m
co
e
f
f
icien
t
(
4
)
an
d
n
u
m
b
er
o
f
ep
o
ch
(
5
)
g
iv
en
as
f
o
llo
w
s
:
ln
(
MSE
)
=
-
5
.
7
6
-
1
.
1
8
1
−
0
.
68
2
−
0
.
075
3
−
0
.
041
4
−
0
.
15
5
+
0
.
18
1
2
+
0
.
077
1
3
+
0
.
043
1
4
+
0
.
15
1
5
−
0
.
051
2
3
−
0
.
043
2
4
−
0
.
12
2
5
−
0
.
043
3
4
+
0
.
040
3
5
+
0
.
041
4
5
+
0
.
81
1
2
+
0
.
44
2
2
+
0
.
010
3
2
−
0
.
024
4
2
+
0
.
081
5
2
(
7
)
T
h
e
f
itn
e
s
s
o
f
th
e
m
o
d
el
is
d
e
ter
m
i
n
ed
b
y
an
al
y
s
is
o
f
v
ar
ia
n
ce
(
A
NOV
A
)
w
h
ic
h
co
n
s
is
t
s
o
f
s
u
m
o
f
s
q
u
ar
e
(
SS
)
,
d
eg
r
ee
o
f
f
r
ee
d
o
m
(
d
f
)
,
m
ea
n
s
q
u
ar
e
(
MS)
,
F
-
v
a
lu
e
s
an
d
P
-
v
alu
e
s
as
s
h
o
wn
i
n
T
ab
le
2
.
T
h
e
s
ig
n
i
f
ica
n
ce
o
f
ea
c
h
co
ef
f
icie
n
t
w
as
d
eter
m
i
n
ed
b
y
t
h
e
F
-
t
est
an
d
P
-
v
al
u
e.
T
h
e
s
ig
n
i
f
i
ca
n
t
o
f
co
r
r
esp
o
n
d
in
g
v
ar
iab
les
w
o
u
ld
b
e
in
cr
ea
s
e
i
f
th
e
ab
s
o
l
u
te
F
-
v
al
u
e
b
ec
o
m
e
s
g
r
ea
ter
a
n
d
t
h
e
P
-
v
alu
e
b
ec
o
m
e
s
s
m
a
ller
.
Fro
m
T
ab
le
2
,
th
e
m
o
d
el
g
iv
e
s
F
-
v
a
lu
e
o
f
8
1
.
2
5
an
d
v
er
y
lo
w
P
-
v
alu
e
(
<
0
.
0
0
0
1
)
.
P
-
v
alu
es
<
0
.
0
5
r
ev
ea
l
th
at
th
e
m
o
d
el
ter
m
s
w
er
e
s
ig
n
i
f
ica
n
t.
T
h
e
n
u
m
b
er
o
f
n
e
u
r
o
n
h
ad
t
h
e
h
ig
h
es
t
ef
f
ec
t
o
n
l
n
(
MSE
)
r
es
p
o
n
s
e
f
o
llo
w
ed
b
y
n
u
m
b
er
o
f
s
p
r
ea
d
a
n
d
n
u
m
b
er
o
f
ep
o
ch
.
T
h
e
lear
n
i
n
g
r
at
e
an
d
m
o
m
en
t
u
m
co
e
f
f
icien
t
h
ad
n
o
s
ig
n
i
f
ica
n
t
ef
f
ec
t
o
n
th
e
r
esp
o
n
s
es.
T
h
e
p
r
ed
ictio
n
2
o
f
0
.
9
8
2
5
is
in
r
e
aso
n
ab
le
ag
r
ee
m
e
n
t
w
it
h
ad
j
u
s
ted
2
,
0
.
9
7
0
4
.
T
h
e
lo
w
v
al
u
e
o
f
co
ef
f
icie
n
t
o
f
v
ar
ia
n
ce
(
C
V=
4
.
6
2
%)
w
h
i
ch
i
s
les
s
t
h
a
n
1
0
s
h
o
w
ed
t
h
at
th
e
e
x
p
er
i
m
en
t
s
co
n
d
u
cted
w
er
e
p
r
ec
is
e
a
n
d
r
eliab
le.
T
ab
le
2
.
A
NOV
A
f
o
r
p
r
ed
icted
R
SM
m
o
d
el
S
o
u
r
c
e
SS
df
MS
F
-
v
a
l
u
e
P
-
V
a
l
u
e
P
r
o
b
>
F
M
o
d
e
l
8
4
.
4
5
20
4
.
2
2
8
1
.
2
5
<
0
.
0
0
0
1
S
i
g
n
i
f
i
c
a
n
t
1
-
N
u
m
b
e
r
o
f
n
e
u
r
o
n
4
4
.
6
2
1
4
4
.
6
2
8
5
8
.
6
4
<
0
.
0
0
0
1
2
-
S
p
re
a
d
1
4
.
7
4
1
1
4
.
7
4
2
8
3
.
5
3
<
0
.
0
0
0
1
3
-
L
e
a
r
n
i
n
g
r
a
t
e
0
.
1
8
1
0
.
1
8
3
.
5
0
0
.
0
7
1
4
4
-
Mo
m
e
n
t
u
m
c
o
e
f
f
i
c
e
n
t
0
.
0
5
5
1
0
.
0
5
5
1
.
0
5
0
.
3
1
3
7
5
-
N
u
m
b
e
r
o
f
e
p
o
c
h
0
.
6
9
1
0
.
6
9
1
3
.
2
1
0
.
0
0
1
1
R
e
si
d
u
a
l
1
.
5
1
29
0
.
0
5
2
L
a
c
k
o
f
F
i
t
1
.
5
1
21
0
.
0
7
2
Pu
re
Err
o
r
0
.
0
0
0
8
0
.
0
0
0
C
o
r
T
o
t
a
l
8
5
.
9
6
49
M
o
d
e
l
st
a
t
i
s
t
i
c
s
S
t
d
.
D
e
v
.
0
.
2
3
R
-
S
q
u
a
r
e
d
0
.
9
8
2
5
M
e
a
n
-
4
.
9
3
A
d
j
R
-
S
q
u
a
r
e
d
0
.
9
7
0
4
C
.
V
.
%
4
.
6
2
P
r
e
d
R
-
S
q
u
a
r
e
d
0
.
9
2
6
1
3
.
1
.
Respo
ns
e
s
urf
a
ce
plo
t
re
s
ults
T
h
e
p
lo
t
o
f
r
esp
o
n
s
e
s
u
r
f
ac
e
r
esu
lt
s
is
p
r
esen
ted
i
n
Fig
u
r
e
2
.
E
ac
h
g
r
ap
h
r
ep
r
esen
ted
a
co
m
b
in
at
io
n
o
f
t
w
o
f
ac
to
r
s
at
th
e
ti
m
e
an
d
h
o
ld
in
g
all
o
th
er
f
ac
to
r
s
at
th
e
m
id
d
le
lev
e
l.
Fig
u
r
e
2
(
a)
s
h
o
w
s
t
h
e
r
esp
o
n
s
e
s
u
r
f
ac
e
ln
(
MSE
)
v
er
s
u
s
th
e
n
u
m
b
er
o
f
n
e
u
r
o
n
an
d
s
p
r
ea
d
w
h
ile
o
th
er
f
ac
to
r
s
r
e
m
ai
n
ed
co
n
s
ta
n
t
at
ze
r
o
lev
el.
I
t
ca
n
b
e
s
ee
n
f
r
o
m
Fi
g
u
r
e
2
(
a)
th
at
m
i
n
i
m
u
m
v
al
u
e
o
f
ln
(
MSE
)
ca
n
b
e
f
o
u
n
d
b
y
1
5
-
2
0
n
eu
r
o
n
s
a
n
d
1
.
5
-
2
.
0
s
p
r
ea
d
.
Mo
r
eo
v
er
,
th
is
r
an
g
e
w
a
s
o
b
s
er
v
ed
f
o
r
n
eu
r
o
n
n
u
m
b
er
in
r
elat
io
n
w
it
h
ep
o
ch
n
u
m
b
er
a
s
s
h
o
w
n
i
n
Fig
u
r
e
2
(
b
)
.
T
h
e
r
esp
o
n
s
e
s
u
r
f
ac
e
s
h
o
w
th
at
alo
n
g
w
i
th
a
n
i
n
cr
ea
s
e
in
n
u
m
b
er
o
f
ep
o
ch
f
r
o
m
1
0
to
3
0
0
0
an
d
s
p
r
ea
d
f
r
o
m
0
.
1
to
2
.
0
,
th
e
ln
(
MSE
)
d
ec
r
ea
s
ed
to
-
6
.
2
as
s
h
o
w
n
in
Fi
g
u
r
e
2
(
c)
.
T
h
e
o
b
j
ec
tiv
e
f
u
n
ctio
n
v
al
u
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
Op
timiz
a
tio
n
o
f a
r
tifi
cia
l n
eu
r
a
l n
etw
o
r
k
to
p
o
lo
g
y
fo
r
mem
b
r
a
n
e
b
io
r
ea
cto
r
…
(
S
ya
h
ir
a
I
b
r
a
h
im
)
121
f
o
r
ln
(
MSE
)
is
-
6
.
3
4
2
f
o
r
th
e
f
in
al
p
o
in
t
s
as
p
r
esen
ted
in
Fig
u
r
e
2
(
a)
.
T
h
e
o
p
tim
u
m
v
al
u
es
g
iv
e
n
b
y
R
SM
w
er
e
as
f
o
llo
w
s
:
n
u
m
b
er
o
f
n
eu
r
o
n
s
=
1
6
,
s
p
r
ea
d
=
1
.
4
,
lear
n
in
g
r
ate
=
0
.
2
8
,
m
o
m
e
n
t
u
m
r
ate
=
0
.
3
an
d
n
u
m
b
er
o
f
ep
o
ch
s
=
1
8
5
2
.
T
h
ese
o
p
ti
m
u
m
v
al
u
es a
r
e
u
s
ed
i
n
th
e
tr
ai
n
i
n
g
o
f
R
B
FNN.
(
a)
(
b
)
(
c)
Fig
u
r
e
2
.
R
esp
o
n
s
e
s
u
r
f
ac
e
an
d
co
n
to
u
r
p
lo
t o
f
ln
(
MSE
)
f
o
r
(
a)
n
eu
r
o
n
n
u
m
b
er
an
d
s
p
r
ea
d
,
(
b
)
n
eu
r
o
n
n
u
m
b
er
an
d
ep
o
ch
n
u
m
b
er
,
an
d
(
c)
s
p
r
ea
d
an
d
ep
o
ch
n
u
m
b
e
r
w
h
ile
o
th
er
f
ac
to
r
s
r
e
m
ai
n
ed
co
n
s
tan
t.
1
6
1
1
1
5
2
0
0
.
1
0
.
6
1
.
1
1
.
5
2
.
0
-
6
.
4
-
5
.
4
-
4
.
4
-
3
.
4
-
2
.
4
L
n
(
M
S
E
)
Numb
e
r
o
f
n
e
u
r
o
n
Sp
r
e
a
d
1
6
11
15
20
0
.
1
0
.
6
1
.
1
1
.
5
2
.
0
Ln(
MS
E
)
N
u
m
b
e
r
o
f
n
e
u
ro
n
S
p
r
e
a
d
-
5
.
7
1
0
-
5
.
0
6
3
-
3
.
7
6
8
-
3
.
1
2
0
-
6
.
3
1
5
-
6
.
3
5
6
-
6
.
1
6
8
9
1
6
1
1
1
5
2
0
1
0
7
5
8
1
5
0
5
2
2
5
3
3
0
0
0
-
6
.
2
-
5
.
4
7
5
-
4
.
7
5
-
4
.
0
2
5
-
3
.
3
L
n
(
M
S
E
)
Numb
e
r
o
f
n
e
u
r
o
n
Numb
e
r
o
f
e
po
c
h
1
6
11
15
20
10
7
5
8
1
5
0
5
2
2
5
3
3
0
0
0
Ln(
MS
E
)
N
u
m
b
e
r
o
f
n
e
u
ro
n
N
u
m
b
e
r
o
f
e
p
o
c
h
-
5
.
7
1
0
-
5
.
0
6
3
-
3
.
7
6
8
-
6
.
1
6
8
9
0
.
1
0
.
6
1
.
1
1
.
5
2
.
0
1
0
7
5
8
1
5
0
5
2
2
5
3
3
0
0
0
-
6
.
2
-
5
.
7
7
5
-
5
.
3
5
-
4
.
9
2
5
-
4
.
5
L
n
(
M
S
E
)
Sp
r
e
a
d
Numb
e
r
o
f
e
po
c
h
0
.
1
0
.
6
1
.
1
1
.
5
2
.
0
10
7
5
8
1
5
0
5
2
2
5
3
3
0
0
0
Ln(
MS
E
)
S
p
re
a
d
N
u
m
b
e
r
o
f
e
p
o
c
h
-
5
.
7
1
0
-
5
.
0
6
3
-
6
.
1
6
8
-
6
.
0
9
3
7
1
9
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
:
117
–
1
2
5
122
3
.
2
.
Neura
l net
w
o
rk
plo
t
re
s
ults
I
n
t
h
is
s
ec
tio
n
,
t
h
e
r
eg
r
e
s
s
io
n
p
lo
ts
o
f
th
e
ex
p
er
i
m
en
tal
d
a
ta
v
er
s
u
s
t
h
e
co
m
p
u
ted
n
eu
r
a
l
n
et
w
o
r
k
d
ata
u
s
i
n
g
t
h
e
o
p
ti
m
u
m
ANN
to
p
o
lo
g
y
ar
e
p
r
ese
n
ted
f
o
r
ea
ch
s
tep
in
c
u
d
i
n
g
tr
ain
i
n
g
,
te
s
tin
g
a
n
d
v
alid
atio
n
n
et
w
o
k
s
.
T
h
e
p
r
ed
icted
m
o
d
els
w
er
e
w
ell
f
it
ted
to
th
e
ex
p
er
i
m
e
n
tal
d
ata
f
o
r
all
s
tep
s
as
d
e
p
icted
in
Fi
g
u
r
e
3
.
T
h
e
co
r
r
elatio
n
co
ef
f
icie
n
t
s
(
R
)
f
o
r
tr
ain
in
g
is
0
.
9
9
0
6
,
f
o
r
t
esti
n
g
is
0
.
9
8
3
9
an
d
f
o
r
v
alid
atio
n
is
0
.
9
7
0
7
.
T
h
e
co
m
p
ar
ati
v
e
v
al
u
es
co
r
r
elatio
n
o
f
d
eter
m
i
n
atio
n
(
R
2
)
,
R
M
SE
an
d
M
SE
w
er
e
g
i
v
en
i
n
T
ab
le
3
.
T
h
e
r
esu
lts
s
h
o
w
ed
th
at
t
h
e
o
p
ti
m
u
m
ANN
m
o
d
el
is
s
u
itab
le
f
o
r
d
escr
ib
in
g
t
h
e
p
er
m
ea
te
f
l
u
x
o
f
P
OM
E
u
s
in
g
MB
R
f
iltra
tio
n
.
T
h
e
o
p
ti
m
al
to
p
o
lo
g
y
o
f
A
N
N
u
s
i
n
g
R
SM
p
r
o
v
id
e
d
g
o
o
d
q
u
alit
y
p
r
ed
ictio
n
f
o
r
t
h
e
f
iv
e
e
x
o
g
e
n
o
u
s
o
u
tp
u
t
s
.
T
h
e
r
esu
lt
s
h
av
e
b
ee
n
co
m
p
ar
ed
w
it
h
t
h
e
co
n
v
e
n
tio
n
al
R
B
FN
N
a
n
d
s
h
o
w
ed
an
i
m
p
r
o
v
ed
A
N
N
m
o
d
el
p
er
f
o
r
m
a
n
ce
as
s
h
o
w
n
in
T
ab
le
3
.
T
h
e
R
B
FNN
-
R
S
M
s
h
o
w
ed
i
ts
s
u
p
er
io
r
it
y
an
d
f
aster
t
h
e
n
tr
ial
-
a
n
d
-
er
r
o
r
m
eth
o
d
s
i
n
f
in
d
i
n
g
t
h
e
o
p
ti
m
u
m
to
p
o
lo
g
y
o
f
A
NN
s
.
(
a)
(
b
)
(
c)
Fig
u
r
e
3
.
R
eg
r
es
s
io
n
p
lo
t f
o
r
p
r
ed
icted
v
er
s
u
s
e
x
p
er
i
m
e
n
tal
f
lu
x
f
o
r
(
a)
tr
ain
in
g
d
ata,
(
b
)
te
s
tin
g
d
ata
an
d
(
c)
v
alid
atio
n
d
ata
T
ab
le
3
.
P
er
f
o
r
m
a
n
ce
ev
al
u
ati
o
n
R
B
F
N
N
-
R
S
M
R
2
M
S
E
R
M
S
E
T
r
a
i
n
i
n
g
0
.
9
8
1
3
0
.
0
0
2
2
0
.
0
4
7
0
T
e
st
i
n
g
0
.
9
6
8
1
0
.
0
0
5
2
0
.
0
7
2
2
V
a
l
i
d
a
t
i
o
n
0
.
9
4
2
3
0
.
0
2
1
7
0
.
1
4
7
3
C
o
n
v
e
n
t
i
o
n
a
l
R
B
F
N
N
R
2
M
S
E
R
M
S
E
T
r
a
i
n
i
n
g
0
.
9
4
2
2
0
.
0
0
7
6
0
.
0
8
7
2
T
e
st
i
n
g
0
.
9
3
7
4
0
.
0
0
9
6
0
.
0
9
8
0
V
a
l
i
d
a
t
i
o
n
0
.
7
9
5
6
0
.
0
2
0
0
0
.
1
4
1
4
0
0
.
2
0
.
4
0
.
6
0
.
8
0
0
.
1
0
.
2
0
.
3
0
.
4
0
.
5
0
.
6
0
.
7
0
.
8
0
.
9
E
x
p
e
r
i
m
e
n
t
a
l
f
l
u
x
(
L
/
m
2
h
)
P
r
e
d
i
c
t
e
d
f
l
u
x
(
L
m
2
h
)
:
R
=
0
.
9
9
0
6
D
a
t
a
P
r
e
d
i
ct
e
d
A
ct
u
a
l
0
0
.
2
0
.
4
0
.
6
0
.
8
1
0
0
.
1
0
.
2
0
.
3
0
.
4
0
.
5
0
.
6
0
.
7
0
.
8
0
.
9
1
E
x
p
e
r
i
m
e
n
t
a
l
f
l
u
x
(
L
/
m
2
h
)
P
r
e
d
i
c
t
e
d
f
l
u
x
(
L
/
m
2
h
)
:
R
=
0
.
9
8
3
8
9
D
a
t
a
P
r
e
d
i
ct
e
d
A
ct
u
a
l
0
0
.
1
0
.
2
0
.
3
0
.
4
0
.
5
0
.
6
0
.
7
0
0
.
1
0
.
2
0
.
3
0
.
4
0
.
5
0
.
6
0
.
7
E
x
p
e
r
i
m
e
n
t
a
l
f
l
u
x
(
L
/
m
2
h
)
P
r
e
d
i
c
t
e
d
f
l
u
x
(
L
/
m
2
h
)
:
R
=
0
.
9
7
0
7
1
D
a
t
a
P
r
e
d
i
ct
e
d
A
ct
u
a
l
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
Op
timiz
a
tio
n
o
f a
r
tifi
cia
l n
eu
r
a
l n
etw
o
r
k
to
p
o
lo
g
y
fo
r
mem
b
r
a
n
e
b
io
r
ea
cto
r
…
(
S
ya
h
ir
a
I
b
r
a
h
im
)
123
Fig
u
r
es
4
(
a)
-
(
c)
s
h
o
w
t
h
e
r
es
p
o
n
s
e
v
ar
iab
le
o
f
p
er
m
ea
te
f
l
u
x
f
o
r
tr
ain
in
g
,
test
i
n
g
an
d
v
alid
atio
n
,
r
esp
ec
tiv
el
y
.
Fo
r
tr
ain
i
n
g
d
ata
w
h
ich
is
t
h
e
tr
an
s
itio
n
b
et
w
ee
n
h
i
g
h
to
lo
w
air
f
lo
w
r
at
e,
th
e
p
er
m
ea
te
f
lu
x
s
tar
ts
to
d
ec
r
ea
s
e
s
lo
w
l
y
f
r
o
m
0
.
8
8
to
0
.
6
0
L
/
m
2
h
.
Fo
r
test
i
n
g
d
ata,
t
h
e
p
er
m
ea
te
f
l
u
x
i
s
a
t
h
i
g
h
air
f
lo
w
r
ate
an
d
it
r
e
m
ain
s
at
0
.
8
L
/
m2
h
.
Fo
r
v
alid
atio
n
d
ata,
th
e
p
er
m
ea
te
f
l
u
x
d
ec
r
ea
s
es
r
ap
id
ly
co
m
p
ar
ed
to
th
e
p
er
m
ea
te
f
l
u
x
at
h
ig
h
air
f
lo
w
.
I
t
ca
n
b
e
s
ee
n
th
at
g
o
o
d
p
r
ed
ictio
n
m
o
d
els
ar
e
o
b
tain
ed
f
o
r
th
e
p
er
m
ea
te
f
l
u
x
f
o
r
all
d
ata
s
et.
(
a)
(
b
)
(
c)
Fig
u
r
e
4
.
P
OM
E
P
er
m
ea
te
f
l
u
x
f
o
r
(
a)
tr
ain
in
g
d
ata,
(
b
)
test
in
g
d
ata
an
d
(
c)
v
alid
atio
n
d
ata
s
et.
0
100
200
300
400
500
600
700
-
0
.
2
0
0
.
2
0
.
4
0
.
6
0
.
8
1
1
.
2
Perm
eat
e
f
lux
(L/
m
2
h)
D
at
a
T
r
a
i
n
i
n
g
Ac
t
ual
P
redic
t
ed
0
50
100
150
200
250
300
350
400
450
500
-
0
.
2
0
0
.
2
0
.
4
0
.
6
0
.
8
1
1
.
2
Perm
eat
e
f
lux
(L/
m
2
h)
D
at
a
T
e
s
t
i
n
g
Ac
t
ual
P
redic
t
ed
0
50
100
150
200
250
300
350
400
450
500
-
0
.
1
0
0
.
1
0
.
2
0
.
3
0
.
4
0
.
5
0
.
6
0
.
7
0
.
8
Perm
eat
e
f
lux
(L/
m
2
h)
D
at
a
V
a
l
i
d
a
t
i
o
n
Ac
t
ual
P
redic
t
ed
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
:
117
–
1
2
5
124
4.
CO
NCLU
SI
O
N
T
h
e
f
o
lo
w
i
n
g
co
n
clu
s
io
n
ca
n
b
e
d
r
aw
n
f
r
o
m
th
e
in
v
est
ig
at
i
o
n
s
co
n
d
u
cted
i
n
t
h
is
w
o
r
k
:
T
h
e
o
p
ti
m
al
A
N
N
to
p
o
lo
g
y
o
f
R
B
FNN
to
p
o
lo
g
y
w
as
m
o
r
e
p
r
ec
is
e
f
o
r
p
r
ed
ictin
g
p
er
m
ea
te
f
l
u
x
o
f
P
OM
E
u
s
i
n
g
MB
R
w
it
h
lo
w
MSE
(
0
.
0
0
2
2
)
an
d
h
ig
h
co
r
r
elatio
n
co
ef
f
icie
n
t
(
0
.
9
9
0
6
)
.
T
h
e
o
p
ti
m
al
n
e
u
r
al
m
o
d
el
h
ad
m
i
n
i
m
u
m
w
h
e
n
th
e
n
u
m
b
er
o
f
n
eu
r
o
n
s
,
s
p
r
ea
d
,
lear
n
in
g
r
ate,
m
o
m
e
n
t
u
m
e
r
ate
an
d
n
u
m
b
er
o
f
ep
o
c
h
s
r
u
n
s
w
er
e
eq
u
a
l
to
1
6
,
1
.
4
,
0
.
2
8
,
0
.
3
an
d
1
8
5
2
,
r
esp
ec
tiv
el
y
.
T
h
e
r
es
u
lt
s
o
f
te
s
tin
g
an
d
v
alid
atio
n
m
o
d
el
o
n
n
e
w
tr
ials
s
h
o
w
ed
ex
ce
lle
n
t
ag
r
ee
m
e
n
t
b
et
w
ee
n
th
e
ac
t
u
al
an
d
p
r
ed
icted
d
ata
w
it
h
co
r
r
elatio
n
co
e
f
f
ic
ien
t
e
q
u
al
to
0
.
9
8
3
9
an
d
0
.
9
7
0
7
,
r
esp
ec
tiv
el
y
.
T
h
e
ap
p
licatio
n
o
f
in
teg
r
ated
R
B
FN
N
an
d
R
SM
r
ed
u
ce
s
th
e
co
m
p
u
tatio
n
al
co
s
t
an
d
i
m
p
r
o
v
ed
th
e
A
NN
m
o
d
el
p
r
ed
ictio
n
.
ACK
NO
WL
E
D
G
E
M
E
NT
S
T
h
e
au
th
o
r
s
w
o
u
ld
lik
e
to
th
an
k
t
h
e
Min
i
s
tr
y
o
f
E
d
u
ca
tio
n
(
MO
E
)
P
R
GS
v
o
te
R
.
J
1
3
0
0
0
0
.
7
8
5
1
.
4
L
7
0
2
an
d
th
e
Un
i
v
er
s
iti
T
ek
n
o
lo
g
i
Ma
la
y
s
ia
UT
MH
I
R
v
o
te
0
8
G7
4
f
o
r
th
eir
f
in
a
n
cial
a
n
d
f
ac
ilit
y
s
u
p
p
o
r
t o
f
th
is
w
o
r
k
.
RE
F
E
R
E
NC
E
S
[1
]
M
.
F
.
A
lk
h
a
ti
b
,
A
.
A
.
M
a
m
u
n
,
a
n
d
I.
A
k
b
a
r,
“
A
p
p
li
c
a
ti
o
n
o
f
re
sp
o
n
se
su
rf
a
c
e
m
e
th
o
d
o
lo
g
y
(
RS
M
)
f
o
r
o
p
ti
m
iza
ti
o
n
o
f
c
o
lo
r
re
m
o
v
a
l
f
ro
m
P
OME
b
y
g
ra
n
u
lar ac
ti
v
a
ted
c
a
rb
o
n
,
”
In
t
.
J
.
En
v
iro
n
.
S
c
i.
T
e
c
h
n
o
l.
,
v
o
l.
1
2
,
p
p
.
1
2
9
5
–
1
3
0
2
,
2
0
1
5
.
[2
]
W
.
P
.
W
a
h
,
S
.
N
ik
M
e
ria
m
,
M
.
Na
c
h
iap
p
a
n
,
a
n
d
B.
V
a
ra
d
a
ra
j,
“
P
re
-
trea
tm
e
n
t
a
n
d
m
e
m
b
r
a
n
e
u
lt
ra
f
il
tratio
n
u
sin
g
trea
ted
p
a
lm
o
il
m
il
l
e
ff
lu
e
n
t
(P
O
M
E),
”
S
o
n
g
k
la
n
a
k
a
ri
n
J
.
S
c
i.
T
e
c
h
n
o
l.
,
v
o
l.
2
4
,
p
p
.
8
9
1
–
8
9
8
,
2
0
0
2
.
[3
]
A
.
Ca
ss
a
n
o
a
n
d
A
.
Ba
sile,
“
M
e
m
b
ra
n
e
s
f
o
r
in
d
u
strial
m
icro
f
il
tr
a
ti
o
n
a
n
d
u
lt
ra
f
il
tratio
n
,
”
in
A
d
v
a
n
c
e
d
M
e
mb
ra
n
e
S
c
ien
c
e
a
n
d
T
e
c
h
n
o
lo
g
y
fo
r
S
u
s
ta
in
a
b
le
E
n
e
rg
y
a
n
d
En
v
ir
o
n
me
n
ta
l
A
p
p
li
c
a
ti
o
n
s
,
1
st
e
d
.
,
A
.
Ba
sile
a
n
d
S
.
P
.
Nu
n
e
s,
E
d
s.
El
se
v
ier,
2
0
1
1
,
p
p
.
6
4
7
–
6
7
9
.
[4
]
G
.
M
o
h
d
S
y
a
h
m
i
Ha
f
izi,
T
.
Y.
Ha
a
n
,
A
.
W
.
L
u
n
,
N.
A
b
d
u
l
W
a
h
a
b
,
M
o
h
a
m
m
a
d
Ra
h
m
a
t,
a
n
d
M
.
Kh
a
iru
l
M
u
is,
“
F
o
u
li
n
g
a
ss
e
ss
m
e
n
t
o
f
tertiar
y
p
a
lm
o
il
m
il
l
e
ff
lu
e
n
t
(P
OME
)
m
e
m
b
ra
n
e
trea
t
m
e
n
t
f
o
r
wa
ter
r
e
c
la
m
a
ti
o
n
,
”
J
.
W
a
ter
Reu
se
De
sa
li
n
.
,
v
o
l.
8
,
n
o
.
3
,
p
p
.
4
1
2
–
4
2
3
,
2
0
1
8
.
[5
]
T
.
L
e
i
k
n
e
s,
“
Was
te
w
a
ter
T
re
a
t
m
e
n
t
b
y
M
e
m
b
ra
n
e
Bio
re
a
c
to
rs,”
in
M
e
mb
ra
n
e
Op
e
ra
ti
o
n
s:
In
n
o
v
a
t
ive
S
e
p
a
ra
ti
o
n
s
a
n
d
T
r
a
n
sf
o
rm
a
ti
o
n
s
,
E.
Dri
o
li
a
n
d
L
.
G
io
rn
o
,
Ed
s.
Italy
:
W
I
L
E
Y
-
V
CH
V
e
rlag
Gm
b
H
&
Co
.
K
G
a
A
,
2
0
0
9
,
p
p
.
374
–
3
9
1
.
[6
]
H.
L
in
e
t
a
l.
,
“
M
e
m
b
ra
n
e
b
io
re
a
c
to
rs
f
o
r
in
d
u
strial
w
a
ste
wa
ter
tr
e
a
t
m
e
n
t:
A
c
rit
ica
l
re
v
ie
w
,
”
Crit.
Rev
.
En
v
iro
n
.
S
c
i.
T
e
c
h
n
o
l.
,
v
o
l.
4
2
,
n
o
.
7
,
p
p
.
6
7
7
–
7
4
0
,
2
0
1
2
.
[7
]
T
.
Ja
n
u
s,
“
M
o
d
e
ll
i
n
g
a
n
d
S
im
u
latio
n
o
f
M
e
m
b
ra
n
e
Bio
re
a
c
to
rs
f
o
r
W
a
ste
w
a
ter
T
r
e
a
t
m
e
n
t,
”
De
M
o
n
tf
o
rt
Un
iv
e
rsit
y
,
L
e
ic
e
ste
r,
2
0
1
3
.
[8
]
S
.
Ju
d
d
,
“
F
o
u
li
n
g
c
o
n
tr
o
l
in
su
b
m
e
rg
e
d
m
e
m
b
ra
n
e
b
io
re
a
c
to
rs,”
W
a
ter
S
c
i.
T
e
c
h
n
o
l.
,
v
o
l.
5
1
,
n
o
.
6
–
7
,
p
p
.
2
7
–
3
4
,
2
0
0
5
.
[9
]
N.
H.
A
b
d
u
ra
h
m
a
n
,
H.
N.
A
z
h
a
ri,
a
n
d
S
.
Nu
rd
i
n
,
“
A
n
In
teg
ra
ted
Ultras
o
n
ic
M
e
m
b
ra
n
e
A
n
a
e
ro
b
ic
S
y
ste
m
(IU
M
A
S
)
f
o
r
P
a
lm
Oil
M
il
l
Ef
f
lu
e
n
t
(
P
OME
)
T
re
a
t
m
e
n
t,
”
En
e
rg
y
Pro
c
e
d
ia
,
v
o
l.
1
3
8
,
p
p
.
1
0
1
7
–
1
0
2
2
,
2
0
1
7
.
[1
0
]
Z.
A
h
m
a
d
,
M
.
B.
Rid
z
u
a
n
,
a
n
d
Z.
Da
u
d
,
“
M
e
m
b
ra
n
e
Bio
re
a
c
to
r
f
o
r
P
a
lm
Oil
M
il
l
E
ff
lu
e
n
t
a
n
d
Re
so
u
rc
e
Re
c
o
v
e
r
y
,
”
in
In
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
S
u
st
a
in
a
b
le
De
v
e
lo
p
me
n
t
fo
r
W
a
ter
a
n
d
W
a
ste
W
a
ter
T
re
a
tme
n
t.
De
c
e
mb
e
r 2
0
0
9
,
2
0
0
9
,
p
p
.
1
–
8.
[1
1
]
S
.
M
u
h
a
m
m
a
d
,
M
.
A
b
d
u
l
W
a
h
a
b
,
M
.
N.
M
o
h
d
T
u
sirin
,
S
.
A
.
S
it
i
Ro
z
a
ima
h
,
a
n
d
A
.
H.
Ha
ss
i
m
i,
“
I
n
v
e
stig
a
ti
o
n
o
f
T
h
re
e
P
re
-
trea
t
m
e
n
t
M
e
th
o
d
s
P
r
io
r
to
Na
n
o
f
il
tratio
n
M
e
m
b
ra
n
e
f
o
r
P
a
lm
Oil
M
il
l
Eff
lu
e
n
t
T
re
a
tme
n
t,
”
S
a
in
s
M
a
la
y
sia
n
a
,
v
o
l
.
4
4
,
n
o
.
3
,
p
p
.
4
2
1
–
4
2
7
,
2
0
1
5
.
[1
2
]
N.
S
.
A
z
m
i
a
n
d
K.
F
.
M
.
Yu
n
o
s,
“
W
a
st
e
w
a
ter
T
re
a
t
m
e
n
t
o
f
P
a
lm
Oil
M
il
l
Eff
lu
e
n
t
(P
OM
E)
b
y
Ultraf
il
tratio
n
M
e
m
b
ra
n
e
S
e
p
a
ra
ti
o
n
T
e
c
h
n
iq
u
e
Co
u
p
led
w
it
h
A
d
so
rp
ti
o
n
T
re
a
t
m
e
n
t
a
s
P
re
-
trea
t
m
e
n
t,
”
Ag
ric
.
Ag
ric
.
S
c
i
.
Pro
c
e
d
ia
,
v
o
l.
2
,
p
p
.
2
5
7
–
2
6
4
,
2
0
1
4
.
[1
3
]
R.
Ba
d
rn
e
z
h
a
d
a
n
d
B.
M
irza
,
“
M
o
d
e
li
n
g
a
n
d
o
p
ti
m
iza
ti
o
n
o
f
c
ro
ss
-
f
lo
w
u
lt
ra
f
il
tratio
n
u
sin
g
h
y
b
rid
n
e
u
ra
l
n
e
tw
o
rk
-
g
e
n
e
ti
c
a
lg
o
rit
h
m
a
p
p
ro
a
c
h
,
”
J
.
I
n
d
.
En
g
.
C
h
e
m.
,
v
o
l.
2
0
,
n
o
.
2
,
p
p
.
5
2
8
–
5
4
3
,
2
0
1
4
.
[1
4
]
R.
S
o
leim
a
n
i,
N.
A
.
S
h
o
u
sh
tari,
B.
M
irza
,
a
n
d
A
.
S
a
lah
i,
“
Ex
p
e
rime
n
tal
in
v
e
stig
a
ti
o
n
,
m
o
d
e
li
n
g
a
n
d
o
p
ti
m
iza
ti
o
n
o
f
m
e
m
b
ra
n
e
se
p
a
ra
ti
o
n
u
sin
g
a
rti
f
icia
l
n
e
u
ra
l
n
e
t
w
o
rk
a
n
d
m
u
lt
i
-
o
b
jec
ti
v
e
o
p
ti
m
iza
ti
o
n
u
sin
g
g
e
n
e
ti
c
a
lg
o
rit
h
m
,
”
Ch
e
m.
En
g
.
Res
.
De
s.
,
v
o
l.
9
1
,
n
o
.
5
,
p
p
.
8
8
3
–
9
0
3
,
2
0
1
3
.
[1
5
]
S
.
Cu
rc
io
,
V
.
Ca
lab
rò
,
a
n
d
G
.
I
o
rio
,
“
Re
d
u
c
t
io
n
a
n
d
c
o
n
tr
o
l
o
f
f
lu
x
d
e
c
li
n
e
in
c
ro
ss
-
f
lo
w
m
e
m
b
ra
n
e
p
ro
c
e
ss
e
s
m
o
d
e
led
b
y
a
rti
f
icia
l
n
e
u
ra
l
n
e
two
rk
s a
n
d
h
y
b
rid
sy
st
e
m
s,”
De
sa
li
n
a
ti
o
n
,
v
o
l.
2
3
6
,
n
o
.
1
–
3
,
p
p
.
2
3
4
–
2
4
3
,
2
0
0
9
.
[1
6
]
Y.
Zak
a
riah
,
A
.
W
.
No
rh
a
li
z
a
,
S
.
S
h
a
f
ish
u
h
a
iza
,
a
n
d
A
.
H.
A
.
Ra
o
f,
“
P
e
rm
e
a
te F
lu
x
M
e
a
su
re
m
e
n
t
a
n
d
P
re
d
icti
o
n
o
f
S
u
b
m
e
rg
e
d
M
e
m
b
ra
n
e
Bio
re
a
c
to
r
F
il
tratio
n
P
ro
c
e
ss
Us
in
g
In
telli
g
e
n
t
T
e
c
h
n
iq
u
e
s,”
J
.
T
e
k
n
o
l.
UTM
,
v
o
l.
7
3
,
n
o
.
3
,
p
p
.
8
5
–
9
0
,
2
0
1
5
.
[1
7
]
Y.
Zak
a
riah
,
A
.
W
.
No
rh
a
li
z
a
,
a
n
d
S
.
S
a
h
la
n
,
“
M
o
d
e
li
n
g
o
f
su
b
m
e
rg
e
d
m
e
m
b
r
a
n
e
b
io
re
a
c
to
r
f
il
tratio
n
p
ro
c
e
ss
u
sin
g
NA
RX
-
A
NFIS
m
o
d
e
l,
”
in
2
0
1
5
1
0
t
h
Asi
a
n
Co
n
tro
l
Co
n
fer
e
n
c
e
:
Eme
rg
in
g
Co
n
tro
l
T
e
c
h
n
iq
u
e
s
fo
r
a
S
u
sta
in
a
b
le W
o
rl
d
,
A
S
CC
2
0
1
5
,
2
0
1
5
,
p
p
.
1
–
6.
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
Op
timiz
a
tio
n
o
f a
r
tifi
cia
l n
eu
r
a
l n
etw
o
r
k
to
p
o
lo
g
y
fo
r
mem
b
r
a
n
e
b
io
r
ea
cto
r
…
(
S
ya
h
ir
a
I
b
r
a
h
im
)
125
[1
8
]
E.
Ra
z
m
i
-
R
a
d
,
B.
G
h
a
n
b
a
rz
a
d
e
h
,
S
.
M
.
M
o
u
sa
v
i,
Z.
Em
a
m
-
Djo
m
e
h
,
a
n
d
J.
K
h
a
z
a
e
i,
“
P
re
d
icti
o
n
o
f
rh
e
o
l
o
g
ica
l
p
ro
p
e
rti
e
s
o
f
Ira
n
ian
b
re
a
d
d
o
u
g
h
f
ro
m
c
h
e
m
i
c
a
l
c
o
m
p
o
siti
o
n
o
f
wh
e
a
t
f
lo
u
r
b
y
u
sin
g
a
rti
f
icia
l
n
e
u
ra
l
n
e
tw
o
rk
s,”
J
.
Fo
o
d
E
n
g
.
,
v
o
l.
8
1
,
n
o
.
4
,
p
p
.
7
2
8
–
7
3
4
,
2
0
0
7
.
[1
9
]
K.
C.
Ke
o
n
g
,
M
.
M
u
sta
f
a
,
A
.
J.
M
o
h
a
m
m
a
d
,
M
.
H.
S
u
laim
a
n
,
a
n
d
N.
R.
H.
A
b
d
u
ll
a
h
,
“
A
rti
f
icia
l
n
e
u
ra
l
n
e
tw
o
rk
f
lo
o
d
p
re
d
icti
o
n
f
o
r
su
n
g
a
i
isa
p
re
sid
e
n
c
e
,
”
in
2
0
1
6
I
EE
E
I
n
ter
n
a
ti
o
n
a
l
C
o
n
fer
e
n
c
e
o
n
Au
to
ma
t
ic
Co
n
tr
o
l
a
n
d
In
telli
g
e
n
t
S
y
ste
ms
(
I2
CACIS
)
,
2
0
1
6
,
p
p
.
2
3
6
–
2
4
1
.
[2
0
]
D.
Krish
n
a
a
n
d
R.
P
.
S
re
e
,
“
A
rti
f
i
c
ial
Ne
u
ra
l
N
e
t
w
o
rk
a
n
d
Re
sp
o
n
se
S
u
rf
a
c
e
M
e
th
o
d
o
l
o
g
y
A
p
p
ro
a
c
h
f
o
r
M
o
d
e
li
n
g
a
n
d
Op
ti
m
iza
ti
o
n
o
f
Ch
ro
m
iu
m
(
V
I)
A
d
so
rp
ti
o
n
f
ro
m
W
a
ste
Wate
r
u
sin
g
Ra
g
i
Hu
sk
P
o
w
d
e
r,
”
In
d
i
a
n
Ch
e
m.
E
n
g
.
,
v
o
l.
5
5
,
n
o
.
3
,
p
p
.
2
0
0
–
2
2
2
,
2
0
1
3
.
[2
1
]
F
.
Ib
ra
h
im
,
M
.
N.
T
a
ib
,
W
.
A
.
B.
W
a
n
A
b
a
s,
C.
C.
G
u
a
n
,
a
n
d
S
.
S
u
laim
a
n
,
“
A
n
o
v
e
l
d
e
n
g
u
e
fe
v
e
r
(D
F
)
a
n
d
d
e
n
g
u
e
h
a
e
m
o
rrh
a
g
ic
f
e
v
e
r
(DH
F
)
a
n
a
l
y
sis
u
sin
g
a
rti
f
i
c
ial
n
e
u
ra
l
n
e
t
w
o
rk
(
A
NN
),
”
Co
mp
u
t.
M
e
th
o
d
s
P
ro
g
ra
ms
Bi
o
me
d
.
,
v
o
l.
7
9
,
n
o
.
3
,
p
p
.
2
7
3
–
2
8
1
,
2
0
0
5
.
[2
2
]
M
.
M
u
sta
f
a
,
T
.
M
o
h
d
Na
sir,
H.
M
.
Zu
n
a
irah
,
S
.
No
riza
m
,
a
n
d
M
.
A
.
S
it
i
A
r
m
iza
,
“
Clas
si
f
i
c
a
ti
o
n
o
f
EE
G
S
p
e
c
tro
g
ra
m
I
m
a
g
e
w
it
h
A
NN
a
p
p
r
o
a
c
h
f
o
r
Bra
in
w
a
v
e
Ba
lan
c
in
g
A
p
p
li
c
a
ti
o
n
,
”
In
t
.
J
.
S
imu
l.
S
y
st.
S
c
i
.
T
e
c
h
n
o
l.
,
v
o
l.
1
2
,
p
p
.
2
9
–
3
4
,
2
0
1
1
.
[2
3
]
M
.
A
g
h
b
a
sh
lo
,
M
.
H.
Kia
n
m
e
h
r
,
T
.
Na
z
g
h
e
li
c
h
i,
a
n
d
S
.
Ra
f
iee
,
“
Op
ti
m
iz
a
ti
o
n
o
f
a
n
a
rti
f
icia
l
n
e
u
ra
l
n
e
tw
o
rk
to
p
o
lo
g
y
f
o
r
p
re
d
ictin
g
d
ry
in
g
k
i
n
e
ti
c
s
o
f
c
a
rro
t
c
u
b
e
s
u
sin
g
c
o
m
b
in
e
d
re
sp
o
n
se
su
rf
a
c
e
a
n
d
g
e
n
e
ti
c
a
lg
o
ri
th
m
,
”
Dr
y
.
T
e
c
h
n
o
l.
,
v
o
l.
2
9
,
n
o
.
7
,
p
p
.
7
7
0
–
7
7
9
,
2
0
1
1
.
[2
4
]
T
.
Na
z
g
h
e
li
c
h
i,
M
.
A
g
h
b
a
sh
lo
,
a
n
d
M
.
H.
Kia
n
m
e
h
r,
“
Op
ti
m
iza
ti
o
n
o
f
a
n
a
rti
f
ial
n
e
u
ra
l
n
e
tw
o
rk
to
p
o
lo
g
y
u
sin
g
c
o
u
p
le
re
sp
o
n
se
su
rf
a
c
e
m
e
e
th
o
d
o
lo
g
y
a
n
d
g
e
n
e
ti
c
a
lg
o
rit
h
m
f
o
r
f
lu
id
ize
b
e
d
d
ry
in
g
,
”
C
o
mp
u
t
.
El
e
c
tro
n
.
Ag
ric
.
,
v
o
l.
7
5
,
p
p
.
8
4
–
9
1
,
2
0
1
1
.
[2
5
]
I.
S
y
a
h
ira,
“
Am
y
lo
se
c
o
n
ten
t
c
a
li
b
ra
ti
o
n
m
o
d
e
l
f
o
r
th
e
th
re
e
t
y
p
e
s
o
f
se
l
e
c
ted
rice
g
ra
in
s
u
sin
g
v
isib
le
sh
o
rtw
a
v
e
n
e
a
r
in
f
ra
re
d
sp
e
c
tro
sc
o
p
y
,
”
Un
iv
e
rsiti
T
e
k
n
o
lo
g
i
M
a
lay
sia
,
2
0
1
5
.
[2
6
]
H.
No
u
rb
a
k
h
sh
,
Z.
Em
a
m
-
Djo
m
e
h
,
M
.
Om
id
,
H.
M
irsa
e
e
d
g
h
a
z
i
,
a
n
d
S
.
M
o
i
n
i,
“
P
re
d
icti
o
n
o
f
re
d
p
lu
m
ju
ice
p
e
rm
e
a
te
f
lu
x
d
u
rin
g
m
e
m
b
ra
n
e
p
ro
c
e
ss
in
g
w
it
h
A
NN
o
p
ti
m
ize
d
u
sin
g
RS
M
,
”
Co
mp
u
t.
El
e
c
tro
n
.
Ag
ric
.
,
v
o
l
.
1
0
2
,
p
p
.
1
–
9
,
2
0
1
4
.
[2
7
]
K.
Ch
ia,
A
.
R.
He
rli
n
a
,
a
n
d
A
.
R.
Ru
z
a
ir
i,
“
Ne
u
ra
l
n
e
tw
o
rk
a
n
d
p
r
in
c
ip
a
l
c
o
m
p
o
n
e
n
t
re
g
re
ss
io
n
in
n
o
n
-
d
e
stru
c
ti
v
e
so
lu
b
le so
li
d
s c
o
n
ten
t
a
ss
e
ss
m
e
n
t:
a
c
o
m
p
a
riso
n
.
,
”
J
.
Z
h
e
ji
a
n
g
U
n
iv
.
S
c
i.
B
,
v
o
l.
1
3
,
n
o
.
2
,
p
p
.
1
4
5
–
5
1
,
F
e
b
.
2
0
1
2
.
B
I
O
G
RAP
H
I
E
S
O
F
AUTH
O
RS
S
y
a
h
ira
Ib
ra
h
im
re
c
e
i
v
e
d
h
e
r
B.
En
g
.
Ho
n
s
(
El
e
c
tri
c
-
Co
n
tr
o
l
a
n
d
In
str
u
m
e
n
tatio
n
)
a
n
d
M
S
c
.
o
f
En
g
in
e
e
rin
g
(El
e
c
tri
c
a
l)
f
ro
m
Un
i
v
e
rsiti
T
e
k
n
o
lo
g
i
M
a
lay
sia
in
2
0
1
1
a
n
d
2
0
1
5
re
sp
e
c
ti
v
e
l
y
.
S
h
e
is
c
u
rre
n
tl
y
w
o
rk
in
g
to
w
a
rd
h
e
r
P
h
D i
n
p
ro
c
e
ss
c
o
n
tro
l
a
t
S
c
h
o
o
l
o
f
El
e
c
ric
a
l
En
g
in
e
e
rin
g
,
Un
iv
e
rsiti
T
e
k
n
o
lo
g
i
M
a
lay
sia
.
He
r
c
u
rre
n
t
re
se
a
rc
h
in
tere
sts
in
c
lu
d
e
m
o
d
e
li
n
g
o
f
n
e
a
r
-
in
f
ra
re
d
sp
e
c
tro
sc
o
p
y
a
n
d
o
p
ti
m
iza
ti
o
n
o
f
m
e
m
b
ra
n
e
f
il
tratio
n
sy
ste
m
u
sin
g
h
y
b
rid
re
sp
o
n
se
su
rf
a
c
e
m
e
th
o
d
o
l
o
g
y
-
a
rti
f
icia
l
in
telli
g
e
n
t.
Ir.
Dr
No
rh
a
li
z
a
A
b
d
u
l
W
a
h
a
b
is
c
u
rre
n
tl
y
a
n
As
so
c
iate
P
ro
f
e
s
so
r
a
t
Un
iv
e
r
siti
Tek
n
o
lo
g
i
M
a
la
y
sia
(U
T
M
).
S
h
e
is
c
u
rre
n
tl
y
th
e
Dire
c
to
r
f
o
r
Co
n
tro
l
a
n
d
M
e
c
h
a
tro
n
ic
En
g
i
n
e
e
rin
g
a
t
th
e
S
c
h
o
o
l
o
f
El
e
c
tri
c
a
l
En
g
in
e
e
rin
g
,
UT
M
.
S
h
e
c
o
m
p
lete
d
h
e
r
P
h
D
in
El
e
c
tri
c
a
l
En
g
in
e
e
rin
g
m
a
jo
rin
g
in
Co
n
tro
l
i
n
Ju
ly
2
0
0
9
.
S
h
e
is
a
c
ti
v
e
l
y
in
v
o
lv
e
d
in
re
se
a
rc
h
in
g
a
n
d
tea
c
h
in
g
in
t
h
e
f
ield
o
f
in
d
u
strial
p
r
o
c
e
ss
c
o
n
tro
l
.
He
r
e
x
p
e
rti
se
is
in
m
o
d
e
ll
in
g
a
n
d
c
o
n
tro
l
o
f
in
d
u
strial
p
ro
c
e
ss
p
lan
t
.
Re
c
e
n
tl
y
sh
e
h
a
s
w
o
rk
e
d
p
rima
ril
y
o
n
d
iff
e
re
n
t
t
y
p
e
s
o
f
d
o
m
e
stic
a
n
d
in
d
u
strial
w
a
ter
a
n
d
w
a
ste
w
a
ter t
re
a
t
m
e
n
t
tec
h
n
o
l
o
g
y
to
w
a
rd
s o
p
ti
m
iza
ti
o
n
a
n
d
e
n
e
rg
y
sa
v
in
g
s
y
st
e
m
.
F
a
ti
m
a
h
S
h
a
m
I
s
m
a
il
is
c
u
rre
n
tl
y
a
S
e
n
io
r
L
e
c
tu
re
r
a
t
Un
iv
e
rsit
i
T
e
k
n
o
lo
g
i
M
a
la
y
sia
(U
T
M
),
Jo
h
o
r
Ba
h
ru
.
S
h
e
h
a
s
m
o
re
th
a
n
2
0
y
e
a
r
s
e
x
p
e
rien
c
e
in
a
re
a
o
f
Co
n
tro
l
a
n
d
In
str
u
m
e
n
tatio
n
En
g
in
e
e
rin
g
sin
c
e
jo
in
i
n
g
UT
M
i
n
1
9
9
2
.
S
h
e
re
c
e
iv
e
d
th
e
B.
S
c
(Ho
n
s.)
in
P
h
y
sic
s,
1
9
8
9
f
ro
m
Un
iv
e
rsiti
Ke
b
a
n
g
sa
a
n
M
a
la
y
sia
a
n
d
o
b
tai
n
e
d
h
e
r
M
a
ste
r
a
n
d
P
h
.
D
f
ro
m
Un
iv
e
rsit
y
T
e
k
n
o
lo
g
i
M
a
la
y
sia
(U
T
M
)
in
1
9
9
2
a
n
d
2
0
1
1
,
re
sp
e
c
ti
v
e
ly
.
Cu
rre
n
tl
y
,
s
h
e
is
c
o
n
d
u
c
ti
n
g
re
se
a
rc
h
e
s
o
n
d
e
v
e
lo
p
m
e
n
t
o
p
ti
m
iza
ti
o
n
a
lg
o
rit
h
m
f
o
r
m
u
lt
i
-
o
b
jec
ti
v
e
p
ro
b
lem
s,
p
lan
t
o
p
ti
m
iza
ti
o
n
d
e
sig
n
,
a
n
d
f
a
u
lt
d
e
tec
ti
o
n
&
d
iag
n
o
sis.
Ya
h
a
y
a
M
d
.
S
a
m
r
e
c
e
i
v
e
d
th
e
B.
E.
d
e
g
re
e
in
e
lec
tri
c
a
l
e
n
g
in
e
e
rin
g
f
ro
m
Un
iv
e
rsit
y
T
e
c
h
n
o
lo
g
y
o
f
M
a
la
y
sia
in
1
9
8
6
,
M
.
S
c
.
d
e
g
re
e
i
n
c
o
n
tr
o
l
sy
ste
m
s
e
n
g
in
e
e
rin
g
f
ro
m
S
h
e
ff
ield
Un
iv
e
rsit
y
,
Un
it
e
d
Kin
g
d
o
m
,
in
1
9
8
8
,
a
n
d
th
e
P
h
.
D.
d
e
g
re
e
in
c
o
n
tro
l
e
n
g
in
e
e
rin
g
f
ro
m
Un
iv
e
rsit
y
Tec
h
n
o
lo
g
y
o
f
M
a
la
y
sia
in
2
0
0
4
.
He
is cu
rre
n
t
ly
a
P
ro
f
e
ss
o
r
w
it
h
th
e
S
c
h
o
o
l
o
f
El
e
c
tri
c
a
l
En
g
in
e
e
rin
g
,
Un
iv
e
rsit
y
T
e
c
h
n
o
lo
g
y
o
f
M
a
la
y
sia
.
His
r
e
se
a
rc
h
in
tere
sts
in
c
lu
d
e
a
n
o
p
ti
m
a
l
c
o
n
tro
l,
r
o
b
u
st
c
o
n
tro
l
,
c
o
m
p
o
site
n
o
n
li
n
e
a
r
f
e
e
d
b
a
c
k
a
n
d
sli
d
in
g
m
o
d
e
c
o
n
tro
l
a
n
d
a
p
p
li
c
a
ti
o
n
o
f
th
e
se
id
e
a
s
to
th
e
a
u
to
m
o
ti
v
e
s
y
ste
m
s.
Evaluation Warning : The document was created with Spire.PDF for Python.