I
AE
S
I
nte
rna
t
io
na
l J
o
urna
l o
f
Art
if
icia
l In
t
ellig
ence
(
I
J
-
AI
)
Vo
l.
7
,
No
.
1
,
Ma
r
ch
2
0
1
8
,
p
p
.
1
~
10
I
SS
N:
2252
-
8938
,
DOI
: 1
0
.
1
1
5
9
1
/i
j
ai.
v
7
.
i1
.
p
p
1
-
1
0
1
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ia
e
s
co
r
e.
co
m/jo
u
r
n
a
ls
/in
d
ex
.
p
h
p
/
I
JA
I
Perf
o
r
m
a
nce
Ana
ly
sis
of ANN
Mo
del f
o
r Esti
m
a
tion
o
f
Tro
phic
Status Ind
ex
o
f
L
a
k
es
T
us
ha
r
Ant
hw
a
l
1
,
Ak
a
n
ks
ha
Cha
nd
o
la
2
,
M
P
T
ha
pli
y
a
l
3
1
De
p
a
rtme
n
t
o
f
Co
m
p
u
ter S
c
ien
c
e
,
Bh
a
g
wa
n
t
Un
iv
e
rsity
,
In
d
ia
2
Uttara
k
h
a
n
d
S
tate
Co
u
n
c
il
f
o
r
S
c
ien
c
e
a
n
d
T
e
c
h
n
o
l
o
g
y
,
De
h
ra
d
u
n
,
In
d
ia
3
De
p
a
rtme
n
t
o
f
Co
m
p
u
ter S
c
ien
c
e
,
HN
B
G
a
rh
w
a
l
(Ce
n
tral)
Un
iv
e
r
sity
,
S
rin
a
g
a
r
G
a
rh
w
a
l,
In
d
ia
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
No
v
1
0
,
2
0
1
7
R
ev
i
s
ed
J
an
7
,
2
0
1
8
A
cc
ep
ted
Feb
9
,
2
0
1
8
T
h
e
h
e
a
lt
h
o
f
w
a
ter
b
o
d
ies
a
c
ro
ss
th
e
g
lo
b
e
is
o
f
h
ig
h
c
o
n
c
e
rn
a
s
th
e
p
o
ll
u
ti
o
n
is
a
c
c
e
lera
ti
n
g
rig
o
ro
u
sly
.
W
it
h
th
e
in
terv
e
n
ti
o
n
s
o
f
sim
p
le
tec
h
n
o
lo
g
y
,
so
m
e
sig
n
if
ica
n
t
c
h
a
n
g
e
s
c
o
u
ld
b
e
b
o
u
g
h
t
u
p
.
L
a
k
e
s
a
re
d
y
in
g
b
e
c
a
u
se
o
f
h
ig
h
T
ro
p
h
ic
I
n
d
e
x
S
tat
u
s
w
h
ich
sh
o
w
s
th
e
e
u
tro
p
h
ica
ti
o
n
lev
e
l
o
f
w
a
ter
b
o
d
ies
.
T
a
k
in
g
th
is
i
n
to
a
c
c
o
u
n
t,
f
e
e
d
f
o
rw
a
rd
b
a
c
k
p
ro
p
a
g
a
ti
o
n
n
e
u
ra
l
n
e
t
w
o
rk
m
o
d
e
l
is
u
se
d
t
o
e
stim
a
te
th
e
T
ro
p
h
ic
S
tatu
s
In
d
e
x
(T
S
I)
o
f
lak
e
s
wh
ich
c
o
u
l
d
c
o
m
p
u
te
th
e
v
a
lu
e
o
f
T
S
I
w
it
h
t
h
e
g
iv
e
n
p
a
ra
m
e
ters
;
p
H,
te
m
p
e
ra
tu
re
,
d
isso
lv
e
d
o
x
y
g
e
n
,
S
e
c
c
h
i
d
isk
tra
n
sp
a
re
n
c
y
,
c
h
lo
ro
p
h
y
ll
a
n
d
to
tal
p
h
o
s
p
h
a
te.
Tw
o
lea
rn
in
g
a
lg
o
rit
h
m
s;
L
e
v
e
n
b
e
rg
M
a
rq
u
a
rd
t
(L
M
)
a
n
d
Bro
y
d
e
n
–
F
letc
h
e
r
–
G
o
ld
f
a
rb
–
S
h
a
n
n
o
(BF
G
S
)
Qu
a
si
Ne
w
to
n
we
re
u
se
d
to
train
th
e
n
e
tw
o
rk
,
w
h
ich
b
e
lo
n
g
s
to
d
if
f
e
re
n
t
c
l
a
ss
e
s.
T
h
e
r
e
su
lt
s
we
re
a
n
a
l
y
z
e
d
u
sin
g
m
e
a
n
sq
u
a
re
e
rro
r
f
u
n
c
ti
o
n
a
n
d
f
u
rth
e
r
c
h
e
c
k
e
d
f
o
r
th
e
d
e
v
iatio
n
f
ro
m
a
c
tu
a
l
d
a
ta.
Am
o
n
g
b
o
th
th
e
train
i
n
g
a
lg
o
rit
h
m
;
L
M
d
e
m
o
n
stra
ted
b
e
tt
e
r
p
e
rf
o
rm
a
n
c
e
w
it
h
0
.
0
0
0
7
a
v
e
ra
g
e
m
e
a
n
sq
u
a
re
e
rro
r
f
o
r
b
e
st
v
a
li
d
a
ti
o
n
p
e
rf
o
r
m
a
n
c
e
a
n
d
BF
G
S
Qu
a
si Ne
w
to
n
sh
o
w
s th
e
a
v
e
ra
g
e
m
e
a
n
sq
u
a
re
e
rro
r
o
f
1
.
0
7
.
K
ey
w
o
r
d
:
A
r
ti
f
icial
n
e
u
r
al
n
et
w
o
r
k
E
u
tr
o
p
h
icatio
n
Feed
f
o
r
w
ar
d
b
ac
k
p
r
o
p
ag
atio
n
T
r
o
p
h
ic
Statu
s
I
n
d
ex
W
ater
q
u
alit
y
Co
p
y
rig
h
t
©
2
0
1
8
In
stit
u
te o
f
A
d
v
a
n
c
e
d
E
n
g
i
n
e
e
rin
g
a
n
d
S
c
ien
c
e
.
Al
l
rig
h
ts re
se
rv
e
d
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
T
u
s
h
ar
An
t
h
w
al,
Dep
ar
t
m
en
t o
f
C
o
m
p
u
ter
A
p
p
licatio
n
s
,
B
h
ag
w
a
n
t U
n
iv
er
s
it
y
,
A
j
m
er
,
I
n
d
ia
E
m
ail: a
n
t
h
w
al.
tu
s
h
ar
@
g
m
ail.
co
m
1.
I
NT
RO
D
UCT
I
O
N
W
ater
b
o
d
ies
all
ac
r
o
s
s
th
e
w
o
r
ld
ar
e
f
ac
in
g
m
an
y
c
h
allen
g
e
s
an
d
g
etti
n
g
d
eter
io
r
ated
d
ay
b
y
d
a
y
an
d
s
o
m
e
o
f
th
e
r
ea
s
o
n
s
w
er
e
p
o
l
lu
tio
n
,
en
cr
o
ac
h
m
e
n
t,
illeg
a
l
co
n
s
tr
u
ct
io
n
,
w
a
ter
q
u
alit
y
is
s
u
e
s
,
w
ater
s
u
p
p
l
y
m
an
a
g
e
m
e
n
t,
g
lo
b
al
w
ar
m
in
g
an
d
m
u
ch
m
o
r
e.
I
t
is
an
u
r
g
e
n
t
n
ee
d
to
h
av
e
a
r
ea
l
tim
e
m
o
n
ito
r
in
g
s
y
s
te
m
o
n
w
ater
b
o
d
ies.
R
ec
e
n
tl
y
th
e
Nain
ital
lak
e
o
f
Ku
m
a
u
n
R
e
g
io
n
o
f
Uttar
ak
h
a
n
d
w
a
s
i
n
n
e
w
s
f
o
r
its
d
r
asti
c
d
i
m
in
u
tio
n
in
w
ater
lev
el
w
h
ic
h
is
ag
ai
n
an
alar
m
i
n
g
n
o
te.
T
h
ese
w
ater
b
o
d
ies
r
eq
u
ir
e
a
co
n
tin
u
o
u
s
m
o
n
ito
r
in
g
s
y
s
te
m
th
a
t
w
o
u
ld
alar
m
o
n
a
n
y
d
is
as
tr
o
u
s
s
it
u
atio
n
o
r
f
o
r
ec
as
t
o
v
er
t
h
eir
h
ea
l
th
w
it
h
les
s
er
m
an
u
al
e
f
f
o
r
ts
a
n
d
u
s
e
c
u
r
r
en
t c
o
m
p
u
ti
n
g
tec
h
n
iq
u
es.
W
ith
t
h
e
s
cie
n
ti
f
ic
a
n
d
s
tatis
t
i
ca
l
in
ter
v
en
tio
n
,
t
h
ese
p
r
o
b
lem
s
co
u
ld
b
e
ad
d
r
ess
ed
an
d
ar
ti
f
icial
n
e
u
r
al
n
et
w
o
r
k
m
a
k
es
it
p
o
s
s
ib
le
to
d
ev
elo
p
s
u
ch
m
o
n
ito
r
in
g
s
y
s
te
m
s
th
a
t
co
u
ld
p
r
ed
ict
t
h
e
co
n
d
itio
n
s
o
f
t
h
ese
w
ater
b
o
d
ies.
I
n
th
is
r
e
s
ea
r
ch
w
o
r
k
,
th
e
s
a
m
p
le
d
ata
o
n
p
h
y
s
io
ch
e
m
ical
p
r
o
p
er
ties
o
f
la
k
es
w
er
e
co
llected
all
ac
r
o
s
s
th
e
g
lo
b
e
an
d
th
e
s
e
d
ata
w
er
e
an
al
y
ze
d
to
id
en
ti
f
y
th
e
le
v
el
o
f
e
u
tr
o
p
h
icatio
n
o
n
t
h
e
m
.
A
r
ti
f
icial
N
e
u
r
al
Net
w
o
r
k
o
f
t
y
p
e
Feed
Fo
r
w
ar
d
Net
w
o
r
k
m
o
d
el
w
a
s
d
ev
elo
p
ed
w
i
th
b
ac
k
p
r
o
p
ag
atio
n
le
ar
n
in
g
al
g
o
r
ith
m
to
tr
ain
th
e
n
et
w
o
r
k
w
i
th
p
h
y
s
io
c
h
e
m
ical
p
ar
a
m
eter
s
(
li
k
e;
te
m
p
er
atu
r
e,
p
H,
alk
alin
it
y
,
Di
s
s
o
lv
ed
Ox
y
g
en
(
DO)
,
Secc
h
i
d
is
k
tr
a
n
s
p
ar
en
c
y
,
c
h
lo
r
o
p
h
y
l
l
an
d
p
h
o
s
p
h
ate)
.
T
r
o
p
h
ic
I
n
d
ex
S
tatu
s
(
T
SI)
o
f
th
e
la
k
e
is
f
u
r
th
er
co
m
p
u
ted
u
s
in
g
t
h
e
n
e
u
r
al
n
et
w
o
r
k
a
n
d
lev
el
o
f
e
u
tr
o
p
h
icati
o
n
o
f
w
ater
b
o
d
y
is
p
r
ed
icted
.
T
h
e
en
tire
r
esear
ch
w
o
r
k
is
co
m
p
iled
in
to
f
o
llo
w
in
g
s
ec
tio
n
s
;
w
h
er
e
s
ec
tio
n
I
I
b
r
ief
s
ab
o
u
t
w
ater
q
u
ali
t
y
as
s
ess
m
e
n
t
i.e
.
b
asic c
o
n
ce
p
ts
o
f
lak
e
w
ater
e
u
tr
o
p
h
i
ca
tio
n
p
r
o
b
le
m
,
ex
p
lai
n
s
t
h
e
r
o
le
o
f
T
SI
v
al
u
es
in
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8938
IJ
-
AI
Vo
l.
7
,
No
.
1
,
Ma
r
ch
2
0
1
8
:
1
–
10
2
th
e
esti
m
atio
n
o
f
la
k
e
h
ea
lt
h
an
d
h
o
w
th
e
A
N
N
co
u
ld
s
er
v
e
th
e
p
u
r
p
o
s
e
o
f
class
i
f
icatio
n
o
f
w
ater
b
o
d
ies
o
n
b
asis
o
f
ab
o
v
e
-
m
e
n
tio
n
ed
p
h
y
s
io
ch
e
m
ical
p
r
o
p
er
ties
.
Sectio
n
I
I
I
;
co
v
er
s
th
e
s
tate
-
of
-
th
e
-
ar
t
in
th
is
f
ield
w
h
er
e
t
y
p
es
o
f
A
NN
w
er
e
u
s
ed
in
wate
r
q
u
alit
y
as
s
ess
m
e
n
ts
at
d
i
f
f
er
en
t
lev
el
s
.
Sect
io
n
I
V;
co
v
e
r
s
th
e
m
et
h
o
d
o
lo
g
y
o
n
d
ev
elo
p
m
e
n
t
o
f
t
h
e
A
NN
;
s
tar
tin
g
o
f
f
w
it
h
d
ata
co
llectio
n
o
n
p
h
y
s
io
ch
e
m
i
ca
l
p
r
o
p
er
ti
es
o
f
w
a
ter
b
o
d
ies.
I
t
also
in
cl
u
d
es
th
e
d
etail
s
p
ec
i
f
icatio
n
s
o
f
Ne
u
r
al
Net
w
o
r
k
m
o
d
el
tr
ain
in
g
f
u
n
ctio
n
s
,
ad
ap
ted
lear
n
in
g
f
u
n
ctio
n
,
n
u
m
b
er
o
f
la
y
er
s
,
n
e
u
r
o
n
s
,
a
n
d
p
er
f
o
r
m
a
n
ce
f
u
n
ctio
n
.
Sec
tio
n
V;
g
i
v
e
s
a
d
etailed
p
er
f
o
r
m
an
ce
an
al
y
s
is
o
f
d
ev
elo
p
ed
n
eu
r
al
n
et
w
o
r
k
an
d
co
m
p
ar
ati
v
e
r
es
u
lt
s
o
n
L
M
an
d
B
FG
S
Q
u
asi
-
Ne
w
to
n
lear
n
i
n
g
alg
o
r
it
h
m
s
o
u
tco
m
es.
T
h
e
last
s
ec
tio
n
VI
;
co
n
clu
d
e
o
n
th
e
o
u
tco
m
e
s
an
d
also
th
r
e
w
li
g
h
t o
n
t
h
e
f
u
t
u
r
e
s
co
p
e
o
f
r
esear
ch
.
2.
WAT
E
R
Q
UA
L
I
T
Y
ASS
E
S
SM
E
NT
W
ater
q
u
alit
y
as
s
es
s
m
e
n
t
i
s
t
h
e
m
a
in
p
r
o
b
le
m
t
h
at
w
e
h
ad
ad
d
r
ess
ed
in
th
i
s
r
ese
ar
ch
w
o
r
k
.
W
ater
p
lay
s
a
v
ital
r
o
le
i
n
t
h
e
lif
e
o
f
all
liv
in
g
b
ein
g
s
.
Hu
m
an
s
b
ei
n
g
,
f
lo
r
a,
f
a
u
n
a,
v
er
t
eb
r
ate,
in
v
er
teb
r
ate;
eith
er
o
n
e
in
f
lu
e
n
ce
s
t
h
e
w
ate
r
q
u
alit
y
o
r
b
ein
g
in
f
l
u
e
n
ce
d
b
y
it
o
r
b
o
th
.
I
n
th
is
h
ig
h
p
ac
e
o
f
tr
an
s
itio
n
an
d
ev
o
lu
tio
n
,
w
ater
b
o
d
ies
n
ee
d
to
b
e
ev
alu
ated
in
f
r
eq
u
e
n
t
g
ap
s
.
T
h
is
is
b
ec
au
s
e
o
f
th
e
h
i
g
h
i
n
t
er
v
en
tio
n
o
f
h
u
m
an
co
n
tr
ib
u
ti
n
g
to
p
o
llu
tio
n
,
ad
v
er
s
e
ef
f
ec
t
s
o
f
cli
m
ate
ch
a
n
g
e
,
m
i
s
s
m
a
n
a
g
e
m
e
n
t
o
f
r
eso
u
r
c
es
etc.
T
h
is
s
ec
tio
n
d
ef
in
e
s
ab
o
u
t
t
h
e
p
r
o
b
lem
ar
o
u
n
d
w
h
ic
h
w
h
o
le
o
f
o
u
r
r
esear
ch
r
ev
o
l
v
es
ar
o
u
n
d
an
d
b
r
ief
s
a
b
o
u
t
E
u
tr
o
p
h
icatio
n
an
d
co
n
ce
p
t o
n
it
s
co
m
p
u
tin
g
tech
n
iq
u
e
i.e
.
T
SI
an
d
also
b
r
ief
ab
o
u
t t
h
e
p
o
ten
tial o
f
A
NN
in
p
r
o
b
le
m
-
s
o
lv
i
n
g
b
y
i
n
tell
ig
e
n
t
m
ea
n
s
.
2
.
1
.
E
utr
o
ph
ica
t
io
n
o
f
L
a
k
es
E
u
tr
o
p
h
icatio
n
o
f
la
k
e
is
a
n
e
m
er
g
i
n
g
p
r
o
b
le
m
w
h
ich
f
allo
u
t
to
h
ig
h
p
o
llu
t
io
n
co
n
te
n
t,
i
n
cr
ea
s
e
i
n
n
u
tr
ie
n
t
co
n
ten
t
o
f
la
k
e
w
ater
th
at
r
es
u
lt
s
in
to
d
eter
io
r
atio
n
o
f
w
ater
q
u
alit
y
.
T
h
is
i
n
cr
ea
s
e
in
n
u
tr
ie
n
t
co
n
ten
t
r
esu
lt
s
i
n
to
alg
al
b
lo
o
m
in
g
t
h
at
r
ed
u
ce
s
t
h
e
a
m
o
u
n
t
o
f
d
i
s
s
o
l
v
ed
o
x
y
g
e
n
i
n
t
h
e
w
ater
b
o
d
y
.
I
n
t
h
is
p
r
o
ce
s
s
w
ater
b
o
d
ies
g
et
au
g
m
e
n
ted
w
ith
n
u
tr
ien
ts
t
h
u
s
in
cr
ea
s
i
n
g
in
th
e
p
r
o
d
u
ctio
n
o
f
en
tr
en
c
h
ed
aq
u
atic
p
lan
ts
an
d
alg
ae
.
W
ith
e
u
tr
o
p
h
icatio
n
t
h
e
f
o
llo
w
i
n
g
t
y
p
e
s
o
f
v
eg
e
tatio
n
s
tar
t
s
g
r
o
w
i
n
g
i
n
s
id
e
th
e
s
u
r
f
ac
e
o
f
th
e
w
ater
b
o
d
ies,
w
h
ic
h
is
d
ep
icted
in
t
h
e
Fi
g
u
r
e
1
,
g
iv
en
b
elo
w
;
F
ig
u
re
1
.
T
y
p
e
o
f
P
lan
t
G
ro
w
th
in
Eu
tr
o
p
h
ica
ti
o
n
A
cc
o
r
d
in
g
to
th
e
r
ep
o
r
t
o
n
Na
in
ital
la
k
e
(
Si
n
g
h
&
B
r
ij
,
2
0
0
2
)
,
it
is
m
en
tio
n
ed
t
h
at
ab
o
u
t
f
e
w
f
ac
to
r
s
lik
e;
o
p
en
s
e
w
er
s
d
i
s
p
o
s
in
g
o
f
lar
g
e
q
u
an
t
ities
s
e
w
a
g
e
is
ca
u
s
i
n
g
d
elete
r
io
u
s
ef
f
ec
t
o
n
t
h
e
w
ater
q
u
a
lit
y
.
Ho
w
e
v
er
,
th
at
ac
tiv
i
ties
h
ad
b
ee
n
ch
ec
k
ed
f
r
o
m
last
o
n
e
d
ec
ad
e,
b
u
t
s
till
th
e
d
eter
io
r
atio
n
o
f
lak
e
is
n
o
t
y
e
t
s
to
p
p
ed
.
Oth
er
an
th
r
o
p
o
g
en
ic
ac
tiv
itie
s
s
u
ch
a
s
illeg
a
l
co
n
s
tr
u
ctio
n
s
,
litt
er
in
g
,
d
o
m
es
tic
d
is
ch
ar
g
e
o
f
w
aste
,
an
d
r
ec
r
ea
tio
n
al
ac
tiv
itie
s
o
n
w
ater
b
o
d
ies
h
ad
n
o
w
b
ec
a
m
e
m
aj
o
r
elem
e
n
t
s
f
o
r
tr
ig
g
er
i
n
g
th
e
eu
tr
o
p
h
icatio
n
o
f
lak
es (
S
h
ar
m
a,
2
0
1
4
)
.
As
g
i
v
e
n
in
t
h
e
p
ap
er
(
A
n
t
h
wal
&
P
an
d
e
y
,
2
0
1
7
)
,
th
e
s
tatis
t
ical
d
a
ta
p
r
ed
icts
th
at
in
th
e
n
ex
t
d
ec
ad
e,
th
e
p
er
ce
n
tag
e
o
f
lak
e
s
w
it
h
o
lig
o
tr
o
p
h
ic
s
tat
u
s
w
o
u
ld
d
ec
r
e
ase
b
y
ap
p
r
o
x
i
m
ate
3
%
-
0
.
5
%
an
d
eu
tr
o
p
h
icatio
n
in
cr
ea
s
es
b
y
5
%
-
5
5
%
(
An
t
h
wal
&
P
an
d
e
y
,
2
0
1
7
)
.
B
y
th
e
y
ea
r
2
0
0
8
,
ab
o
u
t
6
0
%
o
f
lak
es
in
C
h
i
n
a
w
er
e
in
eu
tr
o
p
h
ic
an
d
h
y
p
er
tr
o
p
h
ic
co
n
d
itio
n
an
d
f
u
r
t
h
er
p
r
ed
ic
ted
th
at
b
y
2
0
3
0
all
u
r
b
an
lak
es
w
o
u
ld
s
h
ar
e
th
e
s
a
m
e
s
tat
u
s
.
2
.
2
.
T
ro
ph
ic
Sta
t
us
I
nd
e
x
(
T
SI
)
Dif
f
er
en
t
t
y
p
es
o
f
m
et
h
o
d
s
h
a
v
e
b
ee
n
ad
o
p
ted
f
o
r
lak
e
w
ater
class
i
f
icatio
n
w
it
h
m
a
in
ai
m
f
o
r
ac
ce
s
s
i
n
g
th
e
w
ater
q
u
ali
t
y
.
Am
o
n
g
s
t
all
m
o
s
t
co
m
m
o
n
an
d
w
id
el
y
u
s
ed
m
et
h
o
d
is
T
r
o
p
h
ic
State
I
n
d
ex
f
o
r
m
u
lated
b
y
C
ar
ls
o
n
(
C
ar
l
s
o
n
,
1
9
7
7
)
,
w
h
ic
h
is
b
ased
o
n
id
en
ti
f
y
in
g
t
h
e
b
io
m
as
s
in
t
h
e
lak
e
w
ater
u
s
i
n
g
th
r
ee
p
ar
am
eter
s
;
Secc
h
i
D
is
k
T
r
an
s
p
ar
en
c
y
(
m
ea
s
u
r
e
o
f
tr
an
s
p
ar
en
c
y
o
r
t
u
r
b
id
ity
)
,
C
h
lo
r
o
p
h
y
ll
-
a
co
n
te
n
t
(
m
ea
s
u
r
e
o
f
al
g
al
b
io
m
a
s
s
)
a
n
d
T
o
tal
P
h
o
s
p
h
ate
(
in
d
icate
s
t
h
e
n
u
tr
ien
t
i
n
d
ir
ec
t
s
u
p
p
l
y
f
o
r
alg
al
g
r
o
w
th
)
.
T
SI
is
v
er
y
p
o
w
er
f
u
l
tech
n
iq
u
e
to
id
en
tify
t
h
e
ac
t
u
al
s
tat
u
s
o
f
w
a
ter
b
o
d
ies,
w
h
er
e
th
r
ee
m
aj
o
r
c
o
m
p
o
n
e
n
ts
p
la
y
v
i
tal
r
o
le,
ch
lo
r
o
p
h
y
ll c
o
n
ten
t,
Secc
h
i D
i
s
k
tr
a
n
s
p
ar
en
c
y
m
ai
n
l
y
a
m
ea
s
u
r
e
o
f
tu
r
b
id
it
y
in
w
ater
a
n
d
p
h
o
s
p
h
ate.
H
I
G
H
E
R
P
L
AN
T
S
(
Ma
cr
o
p
h
y
te
s
)
ATTA
CH
ED ALG
E
A
&
AQ
UA
TIC
PLANT
(
P
er
ip
h
y
to
n
)
O
P
E
N
WAT
E
R
AL
G
E
A
(
P
h
y
to
p
lan
k
to
n
)
E
UTROP
HICATI
ON
A
s
tate
o
f
w
ater
b
o
d
ies
w
ith
e
x
ce
s
s
o
f
f
er
ti
lit
y
r
e
s
u
l
ts
in
to
ex
ce
s
s
i
v
e
p
lan
t
g
r
o
w
th
Evaluation Warning : The document was created with Spire.PDF for Python.
IJ
-
AI
I
SS
N:
2252
-
8938
P
erfo
r
ma
n
ce
A
n
a
lysi
s
o
f A
N
N
Mo
d
el
fo
r
E
s
tima
tio
n
o
f..
.
(
Tu
s
h
a
r
A
n
th
w
a
l
)
3
T
h
is
p
r
o
ce
s
s
o
f
est
i
m
a
tio
n
o
f
T
SI
is
d
ata
b
ased
s
y
s
te
m
an
d
r
eq
u
ir
es
m
i
n
i
m
u
m
s
ets
o
f
d
ata
f
o
r
co
m
p
u
tatio
n
o
f
alg
al
b
io
m
as
s
an
d
f
u
r
th
er
p
r
o
v
id
e
a
s
in
g
le
q
u
an
tita
tiv
e
i
n
d
ex
i
n
g
f
o
r
class
if
y
i
n
g
an
d
r
an
k
i
n
g
p
r
o
ce
s
s
o
f
w
ater
b
o
d
ies.
T
h
e
r
an
g
e
o
f
th
e
i
n
d
ex
i
s
f
r
o
m
ap
p
r
o
x
i
m
ate
l
y
0
–
1
0
0
an
d
is
b
ased
o
n
th
e
r
elatio
n
s
h
ip
a
m
o
n
g
t
h
r
ee
p
h
y
s
io
ch
e
m
ical
p
r
o
p
er
ties
Secc
h
i
Dis
k
T
r
an
s
p
ar
en
c
y
,
C
h
lo
r
o
p
h
y
ll
-
a,
T
o
tal
P
h
o
s
p
h
ate
o
f
w
ater
b
o
d
y
.
T
h
e
T
ab
le
1
g
iv
e
n
b
el
o
w
s
u
m
m
ar
izes
th
e
r
elatio
n
o
f
t
h
e
T
SI
v
al
u
es
in
ass
o
ciati
o
n
w
it
h
t
h
e
f
ac
to
r
s
i
m
p
ac
ti
n
g
w
ater
q
u
a
lit
y
a
n
d
o
v
er
all
h
ea
l
th
o
f
w
a
ter
b
o
d
y
.
T
ab
le
1
.
B
r
ief
d
escr
ip
tio
n
o
v
er
T
SI
v
alu
e
an
d
W
ater
q
u
alit
y
T
S
I
(
C
a
r
l
so
n
1
9
7
7
)
T
r
o
p
h
i
c
S
t
a
t
u
s
I
n
d
e
x
a
n
d
W
a
t
e
r
Q
u
a
l
i
t
y
P
a
r
a
me
t
e
r
s
<
4
0
O
l
i
g
o
t
r
o
p
h
i
c
;
c
l
e
a
r
w
a
t
e
r
;
su
f
f
i
c
i
e
n
t
a
mo
u
n
t
o
f
d
i
sso
l
v
e
d
o
x
y
g
e
n
t
h
r
o
u
g
h
o
u
t
y
e
a
r
40
-
50
M
e
so
t
r
o
p
h
i
c
;
M
o
d
e
r
a
t
e
c
l
e
a
r
w
a
t
e
r
;
su
p
p
o
r
t
i
v
e
f
o
r
sw
i
mm
i
n
g
a
n
d
o
t
h
e
r
w
a
t
e
r
a
c
t
i
v
i
t
i
e
s
50
-
70
Eu
t
r
o
p
h
i
c
;
l
o
w
e
r
c
l
a
ssi
c
e
u
t
r
o
p
h
i
c
w
i
t
h
d
e
c
r
e
a
se
i
n
t
r
a
n
s
p
a
r
e
n
c
y
,
ma
c
r
o
p
h
y
t
e
p
r
o
b
l
e
m
e
v
i
d
e
n
t
.
U
p
p
e
r
r
a
n
g
e
w
i
t
h
d
o
m
i
n
a
n
c
e
o
f
b
l
u
e
-
g
r
e
e
n
a
l
g
a
l
scu
ms
a
n
d
e
x
t
e
n
si
v
e
ma
c
r
o
p
h
y
t
e
p
r
o
b
l
e
m
>
7
0
H
y
p
e
r
e
u
t
r
o
p
h
i
c
;
h
e
a
v
y
a
l
g
a
l
b
l
o
o
ms t
h
r
o
u
g
h
o
u
t
s
u
mm
e
r
a
n
d
d
e
n
se
mac
r
o
p
h
y
t
e
b
e
d
s
A
b
o
v
e
8
0
T
S
I
v
a
l
u
e
;
a
l
g
a
l
scu
ms,
su
mm
e
r
f
i
sh
k
i
l
l
s,
f
e
w
mac
r
o
p
h
y
t
e
s a
n
d
d
o
mi
n
a
n
c
e
o
f
r
o
u
g
h
f
i
s
h
T
h
e
th
r
ee
p
ar
am
eter
s
i.e
.
Secc
h
i
Dis
k
T
r
an
s
p
ar
en
c
y
,
C
h
lo
r
o
p
h
y
ll
-
a,
T
o
tal
P
h
o
s
p
h
ate
alo
n
g
w
it
h
te
m
p
er
at
u
r
e,
p
H,
d
is
s
o
lv
ed
o
x
y
g
en
w
er
e
o
b
tain
ed
an
d
th
ese
v
ar
iab
le
s
ar
e
in
teg
r
ated
b
y
t
h
e
li
n
ea
r
r
eg
r
ess
io
n
m
o
d
el,
r
esu
lti
n
g
in
th
e
cla
s
s
i
f
icatio
n
o
f
th
e
w
ater
b
o
d
y
.
C
h
lo
r
o
p
h
y
l
l
-
a
i
s
g
i
v
en
h
ig
h
es
t
p
r
io
r
ity
f
o
r
class
i
f
icatio
n
b
ec
au
s
e
th
i
s
v
ar
iab
le
is
th
e
m
o
s
t
ac
cu
r
ate
p
ar
a
m
eter
f
o
r
co
m
p
u
tatio
n
o
f
T
SI
b
ec
au
s
e
o
f
its
d
ir
ec
t
in
d
ic
atio
n
an
d
p
r
ed
ictio
n
ca
p
ab
ilit
y
o
n
t
h
e
al
g
al
b
io
m
as
s
co
n
te
n
t.
T
h
e
f
o
r
m
u
la
s
f
o
r
ca
lcu
la
tin
g
t
h
e
T
SI
(
T
r
o
p
h
ic
State
I
n
d
ex
)
a
r
e
s
tated
as f
o
llo
w
s
;
1.
T
SI
(
T
r
o
p
h
ic
State
I
n
d
ex
)
f
o
r
C
h
lo
r
o
p
h
y
ll a
(
C
H
L
)
T
SI
(
C
HL
)
=
9
.
8
1
I
n
C
h
lo
r
o
p
h
y
l
l a
(
µg
/l
-
1
)
+
3
0
.
6
(
1
)
2.
T
SI
(
T
r
o
p
h
ic
State
I
n
d
ex
)
f
o
r
Secc
h
i
D
i
sk
T
r
a
nspa
r
en
cy
(
SD
T
)
T
SI
(
SD
T
)
=
6
0
-
1
4
.
4
1
I
n
Secc
h
i D
is
k
T
r
an
s
p
ar
en
c
y
(
m
eter
s
)
(
2
)
3.
T
SI
(
T
r
o
p
h
ic
State
I
n
d
ex
)
f
o
r
T
o
tal
P
h
o
s
p
h
ate
(
T
P
)
T
SI
(
T
P
)
=
1
4
.
4
2
I
n
T
o
tal
P
h
o
s
p
h
ate
(
µg
/l)
+
4
.
1
5
(
3
)
T
SI
(
A
v
g
)
=
(
T
SI
(
C
HL
)
+
T
S
I
(
SDT
)
+
T
SI
(
T
P
)
)
/
3
(
4
)
w
h
er
e,
T
SI
is
C
ar
ls
o
n
T
r
o
p
h
ic
State
I
n
d
ex
a
n
d
I
n
is
n
a
tu
r
al
l
o
g
ar
ith
m
.
2
.
3
.
Ro
le
o
f
Art
if
icia
l N
eura
l N
et
w
o
rk
in Wa
t
er
Q
ua
lity
I
n
th
ese
co
n
d
itio
n
s
,
it
is
q
u
ite
im
p
o
r
ta
n
t
to
co
n
tin
u
o
u
s
l
y
m
o
n
i
to
r
th
e
h
ea
lth
o
f
th
e
lak
e
s
.
I
n
s
u
ch
ca
s
e
s
,
n
e
w
co
m
p
u
ti
n
g
tech
n
iq
u
es a
n
d
to
o
ls
co
u
ld
b
e
h
i
g
h
l
y
u
s
e
f
u
l
an
d
b
y
m
ea
n
s
o
f
th
e
s
e
to
o
ls
,
a
n
o
v
er
all
p
r
ed
ictio
n
an
d
ea
r
ly
w
ar
n
i
n
g
o
n
th
e
h
ea
lt
h
o
f
th
e
w
ater
b
o
d
y
is
p
o
s
s
ib
l
e.
A
NN
in
p
ast
an
d
p
r
esen
t
s
ce
n
ar
io
p
r
o
v
ed
to
b
e
an
ap
p
r
o
p
r
iate
t
o
o
l
f
o
r
d
eter
m
i
n
is
t
ic
m
o
d
eli
n
g
o
f
s
y
s
te
m
b
eh
av
io
r
.
Ho
w
ev
er
,
i
f
p
r
o
ce
s
s
ed
w
ith
w
r
o
n
g
d
ata
th
er
e
is
a
h
ig
h
ch
a
n
ce
f
o
r
f
ai
lu
r
e
o
n
its
class
i
f
icatio
n
a
n
d
p
r
e
d
ictio
n
ca
p
ab
ilit
y
.
T
h
er
ef
o
r
e
d
ata
an
al
y
s
is
r
es
u
lt
s
p
u
r
el
y
d
ep
en
d
o
n
th
e
co
r
r
ec
tn
ess
an
d
a
m
o
u
n
t o
f
d
ata
u
s
ed
to
tr
ain
th
e
n
et
w
o
r
k
.
I
n
t
h
is
w
o
r
k
,
t
h
e
p
h
y
s
io
ch
e
m
i
ca
l
p
ar
a
m
eter
s
;
Secc
h
i
d
is
k
tr
an
s
p
ar
en
c
y
(
SDT
)
,
p
H
v
al
u
e,
d
is
s
o
lv
ed
o
x
y
g
en
,
te
m
p
er
atu
r
e,
p
h
o
s
p
h
a
te
an
d
ch
lo
r
o
p
h
y
ll
(
C
h
l
-
a)
w
e
r
e
co
llected
an
d
o
v
er
all
an
al
y
s
is
o
f
th
e
co
n
d
itio
n
o
f
th
e
lak
e
is
p
r
ep
ar
ed
w
ith
t
h
e
m
ai
n
f
o
c
u
s
o
n
th
e
T
SI
v
al
u
e.
T
SI
is
ca
lcu
lated
f
r
o
m
ab
o
v
e
-
g
i
v
en
d
ata
o
n
p
h
y
s
io
ch
e
m
ical
p
ar
a
m
eter
s
a
n
d
s
tat
u
s
o
f
t
h
e
la
k
e
is
b
ei
n
g
p
r
o
d
u
ce
d
.
T
h
e
lak
e
s
tat
u
s
w
a
s
m
o
d
eled
u
s
i
n
g
A
r
ti
f
icial
Ne
u
r
al
Net
w
o
r
k
w
it
h
ab
o
v
e
-
m
en
tio
n
ed
i
n
d
ices
a
s
in
p
u
t
to
o
ls
a
n
d
o
n
b
a
s
is
o
f
t
h
ese
p
ar
a
m
eter
s
th
e
lak
e
tr
o
p
h
ic
co
n
d
it
io
n
w
a
s
g
en
er
ated
.
T
h
ese
m
o
d
els
ar
e
o
f
h
ig
h
i
m
p
o
r
tan
ce
i
n
t
h
e
p
r
ed
ictio
n
o
f
lak
e
w
a
ter
s
tatu
s
f
o
r
d
if
f
er
en
t
s
ea
s
o
n
a
n
d
also
ti
m
e
s
er
ie
s
an
al
y
s
i
s
f
o
r
f
o
r
ec
asti
n
g
th
e
co
n
d
itio
n
s
i
n
an
d
af
ter
s
ev
er
al
y
ea
r
s
.
3.
ST
A
T
E
O
F
T
H
E
ART
3
.
1
.
Wa
t
er
Q
ua
lity
Ana
ly
s
is
Sy
s
t
e
m
s
W
ater
q
u
alit
y
an
al
y
s
is
w
as a
l
w
a
y
s
b
ee
n
at
p
r
i
m
e
f
o
cu
s
f
o
r
a
n
y
ci
v
iliza
tio
n
to
s
u
r
v
i
v
e,
as p
r
o
v
ed
w
it
h
h
is
to
r
ical
an
d
s
cie
n
ti
f
ic
ev
id
en
ce
s
.
An
al
y
s
i
s
o
f
th
e
w
ater
b
o
d
y
i
s
o
f
cr
u
cial
i
m
p
o
r
tan
ce
,
n
o
t
j
u
s
t
in
ter
m
s
o
f
it
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8938
IJ
-
AI
Vo
l.
7
,
No
.
1
,
Ma
r
ch
2
0
1
8
:
1
–
10
4
w
ater
p
u
r
it
y
b
u
t
t
h
e
o
v
er
all
h
ea
lth
o
f
w
ater
r
eso
u
r
ce
m
atte
r
s
.
As
t
h
e
d
e
m
a
n
d
f
o
r
w
ater
i
s
r
is
i
n
g
d
r
ast
icall
y
,
b
y
a
n
d
lar
g
e,
its
m
a
n
ag
e
m
e
n
t i
s
b
ec
o
m
in
g
v
er
y
i
m
p
o
r
ta
n
t a
n
d
s
o
its
q
u
alit
y
.
A
g
en
er
al
m
o
d
el
o
u
t
lin
ed
i
n
(
T
w
i
g
t,
R
e
g
o
,
T
y
r
r
ell,
&
T
r
o
o
s
t,
2
0
1
1
)
p
r
o
p
o
s
ed
a
f
r
a
m
e
w
o
r
k
o
f
w
ater
q
u
alit
y
f
o
r
e
ca
s
t
in
g
s
y
s
te
m
h
a
v
in
g
k
e
y
co
m
p
o
n
e
n
t
s
li
k
e;
r
ea
l
-
ti
m
e
d
ata
ac
q
u
is
it
io
n
m
o
d
u
l
e,
d
ata
m
an
a
g
e
m
e
n
t
m
o
d
u
le,
f
o
r
ec
asti
n
g
m
o
d
el,
p
r
o
ce
s
s
in
g
u
n
it,
an
d
r
es
u
lt
d
is
s
e
m
in
at
io
n
u
n
i
t.
T
h
e
ap
p
licatio
n
o
f
g
e
n
er
ic
f
o
r
ec
asti
n
g
s
y
s
te
m
t
h
e
Del
f
t
-
FEW
S
w
a
s
i
m
p
le
m
en
ted
f
o
r
b
ath
in
g
w
ater
q
u
a
lit
y
an
d
h
ar
m
f
u
l
a
lg
ae
b
lo
o
m
s
f
o
r
ec
asti
n
g
.
T
h
is
t
y
p
e
o
f
s
y
s
t
e
m
s
s
er
v
e
s
as
b
u
ild
in
g
b
lo
ck
f
o
r
th
e
d
ev
elo
p
m
en
t
o
f
ap
p
li
ca
tio
n
s
th
at
w
o
u
l
d
f
ac
ilit
ate
to
w
ar
d
s
ea
r
l
y
alar
m
i
n
g
a
n
d
h
ea
lt
h
m
o
n
ito
r
in
g
o
f
wate
r
r
eso
u
r
ce
s
.
W
ith
co
n
ti
n
u
o
u
s
r
esear
ch
th
er
e
h
a
d
b
ee
n
s
o
m
e
in
te
g
r
ated
s
y
s
te
m
f
o
r
w
ater
q
u
alit
y
as
s
ess
m
e
n
t h
o
w
e
v
er
,
th
er
e
lack
s
a
s
i
n
g
le
in
t
eg
r
ated
s
y
s
te
m
t
h
at
co
u
ld
b
e
d
ep
lo
y
ed
to
an
y
r
an
d
o
m
w
ater
b
o
d
y
an
d
f
i
n
d
o
u
t
th
e
ac
cu
r
ac
y
o
f
q
u
alit
y
p
ar
a
m
ete
r
s
.
T
h
u
s
as
th
e
w
ater
p
o
llu
tio
n
is
in
cr
ea
s
in
g
d
a
y
af
t
er
d
ay
,
a
s
u
itab
le
in
teg
r
ated
s
y
s
te
m
n
ee
d
s
to
b
e
d
ev
elo
p
e
d
w
it
h
th
e
f
ea
tu
r
e
s
o
n
d
ata
ac
q
u
is
itio
n
,
ass
e
s
s
m
e
n
t,
p
r
o
ce
s
s
in
g
,
a
n
d
f
o
r
ec
asti
n
g
o
f
w
ater
q
u
alit
y
.
C
u
r
r
en
tl
y
w
i
th
th
is
r
esear
ch
w
o
r
k
,
w
e
ai
m
at
d
ev
e
lo
p
in
g
a
n
A
N
N
m
o
d
el
w
h
ic
h
i
s
tr
ain
ed
w
it
h
th
e
d
ata
ac
r
o
s
s
t
h
e
g
lo
b
e
an
d
is
ab
le
to
ass
es
s
t
h
e
w
ater
p
ar
a
m
eter
s
w
i
th
m
u
ch
d
iv
er
s
i
f
icatio
n
.
W
h
ile
f
o
c
u
s
i
n
g
o
n
o
n
l
y
e
u
tr
o
p
h
icatio
n
m
a
n
ag
e
m
e
n
t
a
n
d
co
n
tr
o
l
o
f
w
ater
b
o
d
y
,
t
h
er
e
ar
e
v
ar
io
u
s
w
a
y
s
lik
e;
ch
e
m
ica
l
m
o
n
ito
r
in
g
,
b
io
-
ass
e
s
s
m
en
t
an
d
esti
m
ati
o
n
tech
n
iq
u
es
a
s
p
u
b
lis
h
ed
b
y
(
“
P
lan
n
in
g
a
n
d
M
a
n
a
g
e
m
e
n
t,
”
2
0
1
7
)
,
h
o
w
ev
er
,
co
u
ld
s
h
o
w
s
o
m
e
v
ar
iatio
n
s
i
n
r
esu
lt
s
d
u
e
to
t
h
e
f
ac
to
r
s
o
n
s
p
atial
d
ata,
i
m
a
g
er
y
d
ata
co
r
r
ec
tio
n
,
class
if
icatio
n
an
d
o
th
er
tech
n
iq
u
es
u
s
e
d
f
o
r
co
m
p
u
t
a
tio
n
(
A
n
th
w
a
l
&
P
an
d
e
y
,
2
0
1
7
)
.
T
h
er
e
co
m
es
a
n
o
th
er
ca
teg
o
r
y
o
f
w
ater
q
u
ali
t
y
as
s
es
s
m
e
n
t
s
y
s
te
m
s
w
h
ic
h
ai
m
s
to
id
en
ti
f
y
t
h
e
t
y
p
e
o
f
p
o
llu
tio
n
s
,
s
o
u
r
ce
o
f
p
o
llu
tio
n
s
,
s
u
s
p
en
d
ed
p
ar
ticu
late
m
at
te
r
,
w
ater
s
h
ed
m
an
a
g
e
m
en
t
o
r
o
v
er
all
h
ea
lt
h
etc.
(
W
ah
aa
b
&
B
ad
a
w
y
2
0
0
4
,
Nea
m
t
u
et
al.
,
2
0
0
9
,
C
ar
r
eó
n
-
P
alau
et
al.
,
2
017,
B
asti
d
as e
t a
l.,
2
0
1
7
)
.
A
l
f
er
es,
T
ik
,
C
o
p
p
,
&
Van
r
o
ll
eg
h
e
m
,
(
2
0
1
3
)
p
u
t
em
p
h
asi
s
o
n
th
e
co
n
tin
u
o
u
s
m
o
n
ito
r
in
g
s
y
s
te
m
f
o
r
w
ater
b
o
d
ies
w
h
er
e
in
s
it
u
d
ata
s
h
o
u
ld
b
e
r
eg
u
lar
l
y
e
v
al
u
ated
an
d
v
alid
ate
d
f
o
r
ef
f
ec
tiv
e
in
ter
p
r
etatio
n
.
T
h
er
eb
y
p
r
o
p
o
s
ed
s
o
f
t
w
ar
e
t
o
o
ls
f
o
r
r
ev
ie
w
o
v
er
w
a
ter
q
u
alit
y
d
ata
u
s
in
g
p
r
in
cip
al
c
o
m
p
o
n
en
t
a
n
al
y
s
i
s
,
s
tatis
t
ical
m
etr
ic
an
d
d
ev
elo
p
ed
alg
o
r
ith
m
ill
u
s
tr
ated
w
it
h
au
to
m
ated
m
o
n
ito
r
i
n
g
s
y
s
te
m
s
w
er
e
m
o
u
n
ted
in
a
t
th
e
in
let
o
f
w
as
te
w
ater
tr
ea
t
m
en
t p
lan
t i
n
an
u
r
b
an
r
iv
er
.
An
o
th
er
w
o
r
k
p
r
esen
ted
b
y
C
h
an
g
,
B
ai,
&
C
h
en
,
(
2
0
1
7
)
al
s
o
m
en
tio
n
ed
ab
o
u
t
th
e
n
ee
d
f
o
r
w
ater
q
u
alit
y
m
o
n
ito
r
in
g
a
n
d
d
ev
elo
p
ed
an
in
teg
r
ated
m
u
lti
-
s
e
n
s
o
r
r
e
m
o
te
s
e
n
s
in
g
d
ata
m
o
d
elin
g
s
y
s
te
m
w
i
th
i
m
a
g
e
p
r
o
ce
s
s
in
g
al
g
o
r
ith
m
s
in
co
ll
ab
o
r
atio
n
w
it
h
m
ac
h
i
n
e
lear
n
in
g
r
es
u
lt
i
n
g
in
to
an
a
u
to
m
ated
w
ater
q
u
alit
y
m
o
n
ito
r
i
n
g
s
y
s
te
m
.
T
h
is
s
y
s
te
m
w
as
ter
m
ed
as
"
cr
o
s
s
-
m
is
s
i
o
n
d
ata
m
er
g
in
g
an
d
i
m
a
g
e
r
ec
o
n
s
tr
u
ct
io
n
w
it
h
m
ac
h
in
e
lear
n
i
n
g
"
(
C
DM
I
M)
,
an
d
test
ed
b
y
p
r
ed
ictin
g
th
e
w
ater
q
u
al
it
y
m
ai
n
l
y
n
u
tr
ie
n
ts
,
ch
lo
r
o
p
h
y
ll
-
a
an
d
Secc
h
i d
is
k
tr
an
s
p
ar
en
c
y
o
f
L
ak
e
Nica
r
ag
u
a,
p
r
o
v
id
in
g
an
e
f
f
icien
t to
o
l f
o
r
lak
e
w
ater
s
h
e
d
m
a
n
a
g
e
m
e
n
t.
I
n
th
i
s
d
ig
ital
er
a,
co
m
p
u
ter
s
ar
e
eq
u
ip
p
ed
w
ith
e
n
d
less
p
o
s
s
ib
ilit
ie
s
an
d
in
(
Xiao
-
k
ai,
Z
h
en
g
-
y
a,
Qi
-
li,
&
Da
-
q
u
a
n
,
2
0
1
0
)
,
au
th
o
r
s
p
r
o
p
o
s
ed
an
in
f
o
r
m
atio
n
s
y
s
te
m
f
o
r
m
o
n
ito
r
in
g
p
o
llu
tio
n
co
n
te
n
t
i
n
w
ater
b
o
d
ies
th
at
o
f
f
er
s
d
ata
an
al
y
s
is
an
d
p
r
ed
ictio
n
f
ea
t
u
r
es
f
o
r
s
cien
tific
d
ec
is
io
n
m
a
k
i
n
g
.
T
h
e
d
ata
s
ets
ar
e
co
llected
f
r
o
m
lab
o
r
ato
r
ies
a
n
d
w
ater
b
o
d
y
q
u
ali
t
y
is
d
ia
g
n
o
s
ed
,
h
o
w
ev
er
,
d
u
e
to
t
h
e
li
m
itatio
n
o
f
t
h
ese
pr
o
ce
s
s
es,
th
e
s
y
s
te
m
co
u
ld
b
e
r
estra
in
ed
f
r
o
m
d
eli
v
er
in
g
h
ig
h
-
q
u
alit
y
en
d
r
es
u
lts
.
3
.
2
.
Neura
l N
et
w
o
rk
M
o
dels
f
o
r
Wa
t
er
Q
ua
lity
Ass
ess
m
e
nt
L
i
m
ited
w
ater
q
u
a
lit
y
d
ata,
in
s
itu
tr
ad
itio
n
al
w
a
y
s
o
n
w
ater
q
u
alit
y
an
al
y
s
is
,
t
h
e
co
s
tl
y
w
a
ter
q
u
alit
y
m
o
n
ito
r
i
n
g
a
n
d
ac
cu
r
ac
y
o
v
e
r
p
h
y
s
io
c
h
e
m
ica
l
p
ar
a
m
eter
s
r
ec
o
r
d
e
d
etc.
all
th
ese
f
ac
to
r
s
o
f
ten
p
o
s
e
s
er
io
u
s
p
r
o
b
lem
s
f
o
r
p
r
o
ce
s
s
-
b
ased
m
o
d
elin
g
ap
p
r
o
ac
h
es
to
m
o
d
u
la
te
th
e
m
o
n
i
to
r
in
g
s
y
s
te
m
o
f
wate
r
b
o
d
ies
f
o
r
tim
e
s
er
ies
f
o
r
ec
ast.
Ho
w
ev
er
,
i
n
d
u
e
co
u
r
s
e
o
f
ti
m
e,
ar
tifi
cial
i
n
tel
lig
e
n
ce
tec
h
n
iq
u
es
s
a
y
;
k
n
o
w
l
ed
g
e
-
b
ased
s
y
s
te
m
,
g
en
et
ic
al
g
o
r
ith
m
,
ar
ti
f
icial
n
eu
r
al
n
e
t
w
o
r
k
,
a
n
d
f
u
zz
y
in
f
e
r
en
ce
s
y
s
te
m
p
r
o
v
ed
to
p
r
o
v
i
d
e
b
etter
en
d
r
esu
lt
s
o
v
er
in
s
it
u
co
m
p
u
tat
io
n
s
a
n
d
also
s
h
o
w
ed
th
e
in
n
o
v
a
tiv
e
wa
y
to
w
ar
d
s
b
etter
an
d
ef
f
icie
n
t
m
o
n
ito
r
in
g
s
y
s
te
m
f
o
r
w
ater
q
u
alit
y
m
o
d
eli
n
g
(
C
h
au
,
2
0
0
9
)
.
Am
o
n
g
SVM
an
d
A
NN,
later
p
r
o
v
id
e
r
ea
s
o
n
ab
le
i
m
p
le
m
e
n
tat
io
n
o
p
tio
n
s
,
b
ec
au
s
e
th
e
y
ar
e
co
m
p
u
tat
io
n
all
y
v
er
y
f
as
t
an
d
r
eq
u
ir
e
m
an
y
f
e
w
er
in
p
u
t
s
p
ar
a
m
eter
an
d
in
p
u
t
s
co
n
d
itio
n
s
t
h
a
n
d
eter
m
i
n
is
tic
m
o
d
el
s
.
Sala
m
i,
Salar
i,
E
h
te
s
h
a
m
i,
B
id
o
k
h
ti
an
d
Gh
ad
i
m
i
(
2
0
1
6
)
s
h
o
w
ed
th
e
ap
p
licatio
n
o
f
A
N
N
an
d
m
at
h
e
m
a
tical
m
o
d
elin
g
to
p
r
ed
ict
w
ater
q
u
alit
y
o
f
r
iv
er
f
r
o
m
t
h
e
s
o
u
t
h
w
es
t
r
eg
io
n
o
f
I
r
an
an
d
tak
in
g
in
to
ac
co
u
n
t
f
iv
e
i
n
d
icato
r
s
(
DO,
T
DS,
SA
R
,
B
OD5
,
HC
O3
)
an
d
a
s
s
o
ciate
th
e
m
w
i
th
p
ar
a
m
et
er
s
E
C
,
te
m
p
er
atu
r
e
an
d
p
H.
T
h
e
f
ee
d
f
o
r
w
ar
d
m
o
d
el
w
a
s
ad
o
p
ted
an
d
r
esu
lts
s
h
o
w
s
u
p
er
io
r
ac
cu
r
ac
y
o
v
er
t
h
e
tr
ad
itio
n
al
w
ater
q
u
alit
y
m
ea
s
u
r
i
n
g
m
et
h
o
d
s
.
I
n
a
s
i
m
ilar
w
o
r
k
,
Seo
,
Yu
n
an
d
C
h
o
i
(
2
0
1
6
)
s
tu
d
ied
th
e
w
ater
q
u
a
lit
y
p
ar
a
m
e
ter
s
;
T
em
p
er
atu
r
e,
DO,
p
H,
E
C
,
T
N,
T
P
,
T
u
r
b
id
ity
,
an
d
C
h
lo
r
o
p
h
y
ll
-
a,
f
r
o
m
d
o
w
n
s
tr
ea
m
o
f
C
h
eo
n
g
p
y
eo
n
g
D
a
m
an
d
p
r
ed
icted
th
e
v
alu
e
s
u
s
in
g
A
NN
w
h
ic
h
s
h
o
w
s
s
a
tis
f
ac
to
r
y
r
esu
lts
.
I
n
an
o
th
er
s
t
u
d
y
,
K
a
n
d
a
,
Kip
k
o
r
ir
a
n
d
Ko
sg
e
i
(2
0
1
6
)
m
ea
s
u
r
ed
th
e
ab
ilit
y
o
f
m
u
lti
la
y
er
p
er
ce
p
tr
o
n
b
ased
n
eu
r
al
n
e
t
w
o
r
k
to
p
r
ed
ict
d
is
s
o
lv
ed
o
x
y
g
en
w
h
ile
tak
in
g
in
ac
co
u
n
t
te
m
p
er
at
u
r
e,
tu
r
b
id
it
y
,
p
H
an
d
elec
tr
ic
co
n
d
u
cti
v
it
y
a
s
th
e
i
n
p
u
t
v
ar
i
ab
le.
T
h
e
r
esu
lts
w
er
e
an
al
y
z
ed
u
s
in
g
r
o
o
t
m
ea
n
s
q
u
ar
e
er
r
o
r
v
alu
es
f
o
r
th
e
Evaluation Warning : The document was created with Spire.PDF for Python.
IJ
-
AI
I
SS
N:
2252
-
8938
P
erfo
r
ma
n
ce
A
n
a
lysi
s
o
f A
N
N
Mo
d
el
fo
r
E
s
tima
tio
n
o
f..
.
(
Tu
s
h
a
r
A
n
th
w
a
l
)
5
tr
ain
i
n
g
d
ata
f
r
o
m
R
iv
er
Nzo
ia
an
d
co
r
r
elatio
n
am
o
n
g
t
h
e
in
p
u
t
a
n
d
o
u
tp
u
t
v
ar
iab
les
wer
e
d
ev
elo
p
ed
th
u
s
s
h
o
w
i
n
g
t
h
e
co
m
p
u
ti
n
g
p
o
w
er
o
f
th
e
n
e
u
r
al
n
e
t
w
o
r
k
.
L
i
u
an
d
L
u
,
(
2
0
1
4
)
im
p
le
m
e
n
t
ed
SVM
f
o
r
w
ater
q
u
ali
t
y
f
o
r
ec
asti
n
g
i
n
an
a
g
r
icu
lt
u
r
al
n
o
n
p
o
in
t
s
o
u
r
ce
p
o
llu
ted
r
i
v
er
b
y
esti
m
atio
n
o
v
er
to
tal
n
itro
g
en
an
d
to
tal
p
h
o
s
p
h
ate
co
n
ten
t
a
n
d
co
m
p
ar
ed
it
w
it
h
A
NN
r
es
u
lts
,
w
h
er
e
later
tech
n
iq
u
e
d
eliv
er
e
d
s
u
p
er
io
r
o
u
tco
m
e
s
.
On
th
e
o
th
er
h
an
d
,
to
b
eh
av
e
co
g
n
iti
v
el
y
A
N
N
d
o
es
r
e
q
u
ir
e
a
lar
g
e
p
o
o
l
o
f
r
ep
r
esen
tativ
e
d
ata
s
ets f
o
r
tr
ain
in
g
an
d
ap
p
r
o
p
r
iate
lear
n
in
g
alg
o
r
it
h
m
s
.
I
n
an
o
th
er
ex
p
er
i
m
e
n
t
b
y
Sa
ttar
i,
J
o
u
d
i
an
d
Ku
s
ia
k
(
2
0
1
6
)
,
elec
tr
ical
co
n
d
u
ctiv
it
y
(
E
C
)
an
d
to
tal
d
is
s
o
lv
ed
s
o
lid
s
(
T
DS)
w
a
s
esti
m
ated
u
s
i
n
g
k
-
n
ea
r
est
n
eig
h
b
o
r
s
(
k
-
NN)
alg
o
r
ith
m
a
n
d
s
u
p
p
o
r
t
v
ec
to
r
r
eg
r
ess
io
n
m
o
d
e
l
in
o
r
d
er
to
i
d
en
tify
w
a
ter
q
u
alit
y
o
f
L
ig
h
v
an
C
h
a
y
R
iv
er
an
d
f
o
u
n
d
to
d
eliv
er
ac
cu
r
ate
p
r
ed
ictio
n
s
o
v
er
th
e
t
w
o
v
al
u
e
s
.
Ha
y
k
in
(
1
9
9
1
)
,
d
escr
ib
e
d
a
n
eu
r
al
n
et
w
o
r
k
as
a
p
r
o
ce
s
s
o
r
w
it
h
p
ar
allel
-
d
is
tr
ib
u
ti
n
g
ar
ch
itect
u
r
e
h
av
i
n
g
ac
ce
p
ted
ab
ilit
y
o
f
e
f
f
ic
ien
t
co
m
p
u
ti
n
g
,
p
r
o
ce
s
s
i
n
g
,
f
o
r
ec
asti
n
g
,
an
d
clu
s
ter
i
n
g
,
a
n
d
h
av
e
th
e
ad
v
a
n
ta
g
es
o
f
n
o
n
l
in
ea
r
it
y
,
in
p
u
t
-
o
u
tp
u
t
m
ap
p
in
g
,
ad
ap
tiv
e
l
y
,
g
en
er
aliza
tio
n
an
d
f
au
l
t
to
ler
an
ce
.
A
NN
co
n
s
tit
u
te
s
in
telli
g
e
n
t
b
io
n
ic
m
o
d
els
an
d
th
e
n
o
n
l
in
ea
r
,
lar
g
e
-
s
ca
le,
ad
ap
tiv
e
d
y
n
a
m
ics
s
y
s
te
m
s
w
h
i
ch
co
n
s
is
t
o
f
m
a
n
y
in
ter
co
n
n
ec
ted
n
e
u
r
o
n
s
(
Xiao
-
k
a
i
et
al.
,
2
0
1
0
)
.
A
NN
m
o
d
els
h
a
v
e
b
ee
n
w
id
el
y
ap
p
lied
to
th
e
w
ater
q
u
alit
y
an
al
y
s
is
u
s
i
n
g
Ho
p
f
ield
n
eu
r
a
l
n
et
w
o
r
k
(
HNN)
m
o
d
el
(
Xia
o
-
k
ai
e
t
al.
,
2
0
1
0
)
;
b
ac
k
p
r
o
p
ag
atio
n
m
u
lti
-
la
y
er
p
er
ce
p
tr
o
n
(
ML
P
)
(
Hey
d
ar
i,
Ol
y
aie,
Mo
h
eb
za
d
eh
,
&
Ki
s
i
,
2
0
1
3
)
;
a
d
ap
tiv
e
n
eu
r
o
f
u
zz
y
i
n
f
er
en
ce
s
y
s
te
m
(
A
NFI
S)
(
T
h
air
,
Ham
ee
d
,
&
Ay
ad
,
2
0
1
4
)
an
d
Mu
lti
v
ar
iate
L
i
n
ea
r
R
e
g
r
ess
io
n
(
C
h
u
,
L
u
,
&
Z
h
a
n
g
2
0
1
3
)
.
W
ith
th
e
m
ai
n
ai
m
o
n
t
h
e
co
n
s
er
v
atio
n
o
f
w
ater
r
eso
u
r
ce
s
,
an
ap
p
licatio
n
b
ased
o
n
Ho
p
f
i
eld
n
eu
r
a
l
n
et
w
o
r
k
m
e
th
o
d
an
d
f
ac
to
r
an
al
y
s
i
s
tech
n
iq
u
e
w
a
s
in
tr
o
d
u
c
ed
an
d
o
u
tco
m
e
s
h
o
w
ed
th
at
b
io
ch
e
m
ical
o
x
y
g
e
n
d
em
a
n
d
,
lead
,
zin
c,
m
er
c
u
r
y
,
am
m
o
n
iu
m
n
itra
te
an
d
p
er
m
an
g
a
n
ate
in
d
e
x
ar
e
v
er
y
cr
u
cial
w
ater
q
u
alit
y
in
d
icato
r
s
(
C
h
u
et
al.
,
2
0
1
3
)
.
W
ith
HNN,
th
e
clas
s
i
f
icatio
n
w
a
s
d
o
n
e
o
n
b
asis
o
f
ab
o
v
e
-
m
en
tio
n
ed
p
ar
a
m
e
ter
s
,
w
ater
s
a
m
p
les
w
er
e
ca
teg
o
r
iz
ed
an
d
r
ea
s
o
n
ab
le
r
esu
lts
w
er
e
p
r
o
d
u
ce
d
.
Hey
d
ar
i
et
al.
(
2
0
1
3
)
p
r
esen
ted
A
NN
m
o
d
el
w
it
h
b
ac
k
p
r
o
p
ag
atio
n
m
u
lti
-
la
y
er
p
er
ce
p
tr
o
n
w
it
h
co
n
f
i
g
u
r
atio
n
4
-
5
-
1
an
d
4
-
6
-
1
w
h
er
e
5
an
d
6
ar
e
n
eu
r
o
n
s
i
n
h
id
d
en
la
y
er
,
wo
r
k
in
g
o
n
t
w
o
i
n
p
u
t
v
ar
iab
les
d
is
s
o
l
v
ed
o
x
y
g
en
a
n
d
s
p
ec
if
ic
co
n
d
u
ctan
ce
,
f
o
r
Dela
w
ar
e
R
i
v
er
,
P
en
n
s
y
lv
an
ia
U.
S.
t
h
e
p
er
f
o
r
m
an
ce
w
a
s
ev
al
u
ated
w
it
h
s
tati
s
tical
cr
it
er
ia
i.e
.
co
r
r
elatio
n
co
ef
f
icie
n
t,
r
o
o
t m
ea
n
s
q
u
ar
e
an
d
m
ea
n
ab
s
o
lu
te
er
r
o
r
,
th
er
eb
y
s
o
w
in
g
t
h
e
p
o
ten
tia
l
o
f
tr
ai
n
ed
A
NN
m
o
d
el
f
o
r
w
ater
q
u
al
it
y
m
ea
s
u
r
e
m
en
t
s
.
An
o
th
er
w
ater
q
u
alit
y
m
o
n
ito
r
in
g
a
n
d
ass
ess
m
e
n
t s
y
s
te
m
f
o
r
E
u
p
h
r
ates Ri
v
er
w
as p
r
o
p
o
s
e
d
b
y
T
h
air
et
al.
(
2
0
1
4
)
.
T
h
is
A
NN
m
o
d
el
ai
m
s
to
p
r
ed
ictio
n
an
d
f
o
r
ec
asti
n
g
to
tal
d
is
s
o
l
v
ed
s
o
lid
(
T
DS)
w
it
h
m
u
lt
ip
le
r
eg
r
ess
io
n
(
ML
R
)
m
o
d
el
an
d
f
i
n
d
s
o
u
t
to
b
e
m
o
r
e
co
n
v
en
ie
n
t
to
co
n
v
en
tio
n
al
s
ta
tis
tica
l
tech
n
iq
u
es.
A
w
ater
q
u
alit
y
a
n
al
y
s
is
s
y
s
te
m
p
r
o
p
o
s
ed
b
y
So
ltan
i,
Ker
ac
h
ia
n
an
d
Sh
ir
a
n
g
i
(
2010)
w
as
b
ased
o
n
g
e
n
etic
al
g
o
r
it
h
m
s
i
m
u
lated
o
n
p
h
y
s
io
-
c
h
e
m
ic
al
p
ar
am
eter
s
u
s
i
n
g
M
L
R
,
R
B
FN,
an
d
A
N
FIS,
w
it
h
t
h
e
ai
m
to
id
en
ti
f
y
eu
tr
o
p
h
icatio
n
o
f
r
eser
v
o
ir
s
.
W
ith
t
h
e
i
n
s
p
ir
atio
n
,
C
h
en
a
n
d
L
i
u
(
2
0
1
5
)
in
th
eir
s
tu
d
y
co
m
p
ar
ed
t
w
o
ar
ti
f
icia
l
n
eu
r
al
n
et
w
o
r
k
m
o
d
el
i.e
.
R
B
NF
an
d
A
N
FIS,
an
d
M
L
R
m
o
d
el.
I
n
th
e
co
m
p
u
tatio
n
DO,
T
P
,
C
h
l
-
a,
a
n
d
SD
v
alu
e
s
w
er
e
m
ea
s
u
r
ed
an
d
A
N
FIS
m
o
d
el
f
o
u
n
d
to
b
e
s
u
p
er
io
r
to
o
th
er
s
.
Si
m
ilar
e
f
f
o
r
ts
ar
e
d
o
n
e
b
y
A
r
ee
r
ac
h
a
k
u
l
an
d
San
g
u
a
n
s
in
tu
k
u
l
(
2
0
0
9
)
w
h
er
e
a
m
u
ltil
a
y
er
p
er
ce
p
tio
n
(
ML
P
)
n
eu
r
a
l
n
et
w
o
r
k
u
s
in
g
th
e
L
e
v
en
b
er
g
Ma
r
q
u
ar
d
t
(
L
M)
a
lg
o
r
ith
m
e
x
ec
u
ted
to
clas
s
i
f
y
w
ater
q
u
alit
y
o
f
ca
n
a
ls
i
n
B
a
n
g
k
o
k
an
d
f
o
u
n
d
to
p
o
s
s
ess
9
9
.
3
4
% a
cc
u
r
ac
y
w
it
h
less
er
co
s
t a
n
d
m
o
r
e
ef
f
ic
ien
c
y
.
I
n
o
u
r
r
esear
ch
w
o
r
k
,
w
e
ar
e
also
co
m
p
u
ti
n
g
o
n
p
h
y
s
io
c
h
e
m
i
ca
l
p
r
o
p
e
r
ti
es
o
f
r
eser
v
o
ir
s
m
a
in
l
y
la
k
e
s
,
to
d
ev
elo
p
a
n
eu
r
al
n
et
w
o
r
k
m
o
d
el
f
o
r
id
en
tif
icat
io
n
o
f
eu
tr
o
p
h
icatio
n
s
tatu
s
o
f
w
ate
r
b
o
d
ies.
W
e
h
ad
co
n
s
id
er
ed
th
e
d
ata
ac
r
o
s
s
th
e
g
lo
b
e
to
d
ev
elo
p
a
g
e
n
er
ic
m
o
d
el
w
h
ic
h
co
u
ld
co
v
er
u
p
a
b
r
o
ad
er
r
an
g
e
o
f
p
r
o
p
er
ties
.
Ma
in
l
y
C
h
l
-
a,
S
DT
,
an
d
T
P
w
er
e
ta
k
en
in
t
o
ac
co
u
n
t
to
co
m
p
u
te
t
h
e
T
SI
v
al
u
e,
h
o
w
e
v
er
,
o
th
er
p
ar
am
eter
s
lik
e
p
H,
D
O,
elec
tr
ical
co
n
d
u
cti
v
it
y
a
n
d
te
m
p
er
atu
r
e
w
as
al
s
o
in
cl
u
d
ed
w
h
i
le
tr
ain
i
n
g
o
f
n
et
w
o
r
k
b
ec
au
s
e
all
th
e
s
e
f
ac
t
o
r
s
also
h
av
e
a
d
ir
ec
t
co
n
tr
ib
u
t
io
n
to
w
ar
d
s
al
g
al
b
lo
o
m
in
g
o
f
r
eser
v
o
ir
s
.
T
h
e
T
SI
v
alu
e
s
o
v
er
n
u
m
er
o
u
s
d
ata
av
ailab
le
f
r
o
m
v
ar
io
u
s
s
o
u
r
ce
s
w
er
e
co
llected
,
au
to
-
co
r
r
ec
ted
,
clea
n
ed
an
d
A
N
N
m
o
d
el
b
ased
o
n
f
ee
d
f
o
r
w
ar
d
b
ac
k
p
r
o
p
ag
atio
n
tr
ain
ed
w
it
h
L
M
an
d
B
FGS
Qu
a
s
i
-
Ne
w
t
o
n
alg
o
r
ith
m
w
er
e
d
ev
elo
p
ed
to
co
m
p
u
te
an
d
p
r
ed
ict
th
e
eu
tr
o
p
h
icatio
n
lev
e
l o
f
w
ater
b
o
d
ies.
4.
M
AT
E
RIAL
A
ND
M
E
T
H
O
DS
4
.
1
.
Da
t
a
Co
llect
io
n
a
nd
Ana
ly
s
is
Data
p
la
y
s
an
i
m
p
o
r
tan
t
r
o
le
in
th
e
tr
ai
n
i
n
g
o
f
Neu
r
al
Net
w
o
r
k
a
n
d
th
e
ac
cu
r
ac
y
o
f
t
h
e
p
r
ed
ictiv
e
v
alu
e
s
al
s
o
d
ir
ec
tl
y
r
elies o
n
t
h
e
ac
c
u
r
ac
y
o
f
f
o
r
m
er
.
Fo
r
o
u
r
r
esear
ch
w
o
r
k
th
e
d
ata
o
n
p
h
y
s
ical
an
d
c
h
e
m
ical
f
ield
p
r
o
f
ilin
g
p
ar
a
m
eter
s
o
f
la
k
e
w
er
e
as
s
i
m
ilated
f
r
o
m
v
ar
io
u
s
s
o
u
r
ce
s
ac
r
o
s
s
t
h
e
g
lo
b
e;
w
ater
o
n
w
eb
p
o
r
tal
(
h
ttp
://
www
.
w
a
ter
o
n
th
e
w
eb
.
o
r
g
/d
ata/in
d
e
x
.
h
t
m
l)
,
E
u
r
o
p
ea
n
E
n
v
ir
o
n
m
e
n
t
Ag
en
c
y
(
h
ttp
s
://
www
.
ee
a.
eu
r
o
p
a.
eu
/d
a
ta
-
a
n
d
-
m
ap
s
/d
ata/
w
ater
b
ase
-
la
k
es
-
6
)
,
Natio
n
al
I
n
s
tit
u
te
o
f
H
y
d
r
o
lo
g
y
,
R
o
o
r
k
ee
,
I
n
d
ia
(
h
ttp
://
www
.
n
i
h
r
o
o
r
k
ee
.
g
o
v
.
i
n
/)
an
d
s
e
v
er
al
p
u
b
licatio
n
s
w
it
h
s
tatu
s
o
n
eu
tr
o
p
h
icatio
n
o
f
I
n
d
ia
n
lak
e
s
.
T
h
ese
d
ata
in
clu
d
e;
te
m
p
er
at
u
r
e,
d
is
s
o
lv
ed
o
x
y
g
e
n
lev
el,
p
H
v
alu
e,
elec
tr
ical
co
n
d
u
cti
v
it
y
,
Secc
h
i
d
is
k
tr
an
s
p
ar
e
n
c
y
,
to
tal
p
h
o
s
p
h
ate
an
d
ch
lo
r
o
p
h
y
ll
-
a
co
n
te
n
t.
T
h
e
p
ar
am
eter
tak
e
n
in
to
ac
co
u
n
t
w
er
e
o
f
h
i
g
h
s
ig
n
i
f
ica
n
ce
li
k
e;
T
em
p
er
at
u
r
e
in
⁰C
tel
ls
ab
o
u
t
t
h
e
w
ater
te
m
p
er
atu
r
e,
p
H
v
al
u
e
o
f
w
a
t
er
i
m
p
l
y
ac
id
ic
o
r
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8938
IJ
-
AI
Vo
l.
7
,
No
.
1
,
Ma
r
ch
2
0
1
8
:
1
–
10
6
alk
ali
n
it
y
,
a
m
o
u
n
t
o
f
d
is
s
o
lv
e
d
o
x
y
g
e
n
i
n
w
ater
w
h
ich
is
m
o
s
t
i
m
p
o
r
tan
t
f
ac
to
r
f
o
r
d
ef
in
i
n
g
t
h
e
q
u
a
lit
y
o
f
w
ater
,
Secc
h
i
d
is
k
tr
an
s
p
ar
en
c
y
w
h
ic
h
is
m
ea
s
u
r
ed
ass
ess
th
e
clar
ity
o
f
w
ate
r
b
o
d
y
,
ch
lo
r
o
p
h
y
ll
-
a
th
at
i
s
m
ea
s
u
r
e
o
f
alg
ae
in
t
h
e
w
ater
,
to
tal
p
h
o
s
p
h
ate
w
h
ich
d
ep
ict
s
th
e
n
u
tr
ien
t
th
at
ar
e
f
o
o
d
f
o
r
b
i
o
lo
g
ical
o
r
g
an
is
m
s
liv
i
n
g
i
n
w
ater
an
d
h
av
e
d
ir
ec
t
i
m
p
ac
t
o
n
q
u
alit
y
o
f
w
ater
an
d
elec
tr
ical
co
n
d
u
ctiv
i
t
y
w
h
i
ch
is
r
elate
d
to
th
e
di
s
s
o
lv
ed
ele
m
en
t
s
i
n
an
aq
u
e
o
u
s
s
o
l
u
tio
n
.
A
r
o
u
n
d
4
0
0
0
d
atasets
w
er
e
co
llected
an
d
T
SI
v
alu
e
o
f
ea
ch
d
ataset
w
a
s
co
m
p
u
ted
an
d
Feed
Fo
r
w
ar
d
Neu
r
al
Net
w
o
r
k
m
o
d
el
w
it
h
L
ev
e
n
b
er
g
-
Ma
r
q
u
ar
d
t
an
d
B
FGS
Qu
a
s
i
Ne
w
to
n
B
ac
k
P
r
o
p
ag
atio
n
lear
n
i
n
g
alg
o
r
ith
m
s
w
er
e
o
p
ted
.
Ne
u
r
al
n
et
w
o
r
k
s
y
s
te
m
w
as
s
et
u
p
an
d
tr
ain
ed
w
ith
d
ataset
s
an
d
test
ed
d
ata
w
as
s
i
m
u
lated
to
a
n
al
y
ze
t
h
e
p
er
f
o
r
m
an
ce
o
f
t
h
e
n
et
w
o
r
k
r
e
g
ar
d
in
g
b
o
th
t
h
e
lear
n
in
g
al
g
o
r
ith
m
s
.
T
h
e
ev
o
lu
tio
n
o
f
a
lak
e
f
r
o
m
o
lig
o
tr
o
p
h
ic
to
e
u
t
r
o
p
h
ic
co
u
ld
b
e
co
m
p
u
ted
th
r
o
u
g
h
m
ea
s
u
r
e
m
e
n
t
o
f
to
tal
p
h
o
s
p
h
ate
(
T
P
)
,
SDT
,
an
d
ch
lo
r
o
p
h
y
ll
a
(
C
h
l
-
a)
.
Fig
u
r
e
2
s
h
o
w
s
t
h
e
r
an
g
e
o
f
T
SI
v
alu
e
s
an
d
attr
ib
u
tes
t
h
at
clas
s
if
ie
s
w
ater
b
o
d
ies
in
to
o
lig
o
tr
o
p
h
ic,
m
e
s
o
tr
o
p
h
ic
,
eu
tr
o
p
h
ic,
an
d
h
y
p
er
e
u
tr
o
p
h
i
c.
Fig
u
r
e
2
.
Gr
ap
h
ical
r
ep
r
esen
ta
tio
n
o
f
T
SI,
T
r
an
s
p
ar
en
c
y
,
C
h
l
o
r
o
p
h
y
ll
-
a
a
n
d
T
o
tal
P
h
o
s
p
h
ate
v
alu
e
s
an
d
ass
o
ciate
d
eu
tr
o
p
h
icatio
n
lev
e
l (
h
ttp
://
www
.
la
k
ea
cc
es
s
.
o
r
g
/l
ak
ed
ata/d
atain
f
o
ts
i.
h
t
m
l
)
4.
2
.
Neura
l N
et
w
o
rk
M
o
dellin
g
A
r
ti
f
icia
l
Ne
u
r
al
Net
w
o
r
k
m
o
d
elin
g
i
s
a
v
er
y
cr
u
cial
p
r
o
ce
s
s
w
h
er
e
n
et
w
o
r
k
t
y
p
e,
tr
ain
i
n
g
f
u
n
c
tio
n
,
lear
n
in
g
alg
o
r
it
h
m
,
n
u
m
b
er
o
f
lay
er
s
a
n
d
n
eu
r
o
n
s
,
an
d
p
er
f
o
r
m
an
ce
ev
al
u
atio
n
f
u
n
ct
io
n
h
a
v
e
to
d
ec
id
e
o
n
an
d
w
h
o
le
lear
n
i
n
g
o
f
s
y
s
te
m
is
b
ased
o
n
all
th
ese
ab
o
v
e
f
ac
t
o
r
s
.
Fo
r
th
e
n
et
w
o
r
k
t
y
p
e,
th
e
f
ee
d
-
f
o
r
w
ar
d
b
ac
k
p
r
o
p
ag
atio
n
n
eu
r
al
n
e
t
w
o
r
k
i
s
p
r
ef
er
r
ed
b
ec
au
s
e
o
f
its
s
i
m
p
le
ar
ch
itect
u
r
e
an
d
ea
s
y
c
o
m
p
u
tatio
n
m
et
h
o
d
,
f
o
r
lar
g
e
s
ca
le
d
ata
f
ee
d
f
o
r
w
a
r
d
b
ac
k
p
r
o
p
ag
atio
n
d
eliv
er
b
etter
r
esu
lts
.
4
.
2
.
1
.
F
ee
d f
o
r
w
a
rd
neura
l
net
w
o
rk
w
it
h ba
ck
pro
pa
g
a
t
io
n
A
f
ee
d
-
f
o
r
w
ar
d
n
e
u
r
al
n
et
wo
r
k
is
th
e
s
i
m
p
les
t
n
et
w
o
r
k
th
at
s
i
m
u
lates
t
h
e
f
u
n
ctio
n
a
lit
y
o
f
t
h
e
b
io
lo
g
ical
n
e
u
r
al
s
y
s
te
m
i
n
th
e
h
u
m
a
n
b
o
d
y
a
n
d
co
m
p
r
is
es
o
f
lar
g
e
n
u
m
b
er
o
f
n
e
u
r
o
n
i.e
.
p
r
o
ce
s
s
in
g
u
n
its
o
r
n
o
d
es
ar
r
an
g
ed
i
n
la
y
er
s
.
E
v
er
y
u
n
it
i
n
a
la
y
er
is
co
n
n
ec
te
d
w
it
h
all
th
e
u
n
its
in
th
e
p
r
e
v
io
u
s
la
y
er
a
n
d
all
co
n
n
ec
tio
n
s
m
a
y
b
ea
r
s
a
m
e
o
r
d
if
f
er
e
n
t
w
ei
g
h
ts
.
T
h
e
w
ei
g
h
ts
o
n
th
e
s
e
co
n
n
ec
tio
n
s
ac
t
u
all
y
a
r
e
u
s
ed
to
co
m
p
u
te
an
d
in
ter
p
r
et
th
e
tr
ain
in
g
o
f
n
et
w
o
r
k
t
h
u
s
p
r
ed
eter
m
i
n
in
g
th
e
k
n
o
w
led
g
e
o
f
a
n
et
w
o
r
k
.
T
h
e
en
tr
y
o
f
d
ata
is
d
o
n
e
th
r
o
u
g
h
th
e
i
n
p
u
ts
,
w
h
ic
h
f
u
r
th
er
p
ass
e
s
f
r
o
m
la
y
er
to
la
y
er
th
r
o
u
g
h
t
h
e
n
et
w
o
r
k
a
n
d
f
i
n
all
y
ar
r
iv
e
s
at
t
h
e
o
u
tp
u
ts
.
As
th
er
e
d
o
es
n
o
t
ex
i
s
t
an
y
f
ee
d
b
ac
k
b
et
w
ee
n
la
y
er
s
,
it
r
esu
lts
i
n
to
n
o
n
-
c
y
clic
n
e
t
w
o
r
k
k
n
o
w
n
a
s
th
e
f
ee
d
-
f
o
r
w
ar
d
n
eu
r
al
n
e
t
w
o
r
k
.
Fig
u
r
e
3
.
Feed
Fo
r
w
ar
d
Neu
r
al
Net
w
o
r
k
w
it
h
B
ac
k
P
r
o
p
ag
atio
n
B
ec
au
s
e
o
f
its
s
i
m
p
lic
it
y
,
b
ac
k
p
r
o
p
ag
atio
n
alg
o
r
ith
m
i
s
p
r
ef
er
r
ed
an
d
is
b
ased
o
n
g
r
ad
ien
t
d
escen
t
tech
n
iq
u
es.
I
t
p
o
s
s
ess
e
s
th
e
ca
p
ab
ilit
y
to
m
i
n
i
m
ize
t
h
e
m
ea
n
s
q
u
ar
ed
er
r
o
r
b
y
f
i
n
d
i
n
g
a
lo
ca
l
m
in
i
m
u
m
o
f
th
e
er
r
o
r
f
u
n
ctio
n
.
T
h
e
in
i
tializati
o
n
o
f
th
e
n
et
w
o
r
k
i
s
d
o
n
e
b
y
r
an
d
o
m
w
eig
h
t
w
,
w
h
o
s
e
f
u
n
ctio
n
s
er
v
e
s
as
t
h
e
o
u
tp
u
t
o
f
ea
c
h
n
eu
r
o
n
i
n
t
h
e
o
u
tp
u
t
la
y
er
.
Fo
r
ea
c
h
iter
atio
n
,
t
h
e
g
r
ad
ien
t
o
f
t
h
e
er
r
o
r
f
u
n
ctio
n
is
co
m
p
u
ted
r
ec
u
r
s
iv
e
l
y
an
d
ac
co
r
d
in
g
l
y
weig
h
t
ass
i
g
n
ed
to
th
e
la
y
er
s
is
co
r
r
ec
te
d
.
T
h
e
eq
u
atio
n
f
o
r
m
i
n
i
m
u
m
er
r
o
r
E
is
s
h
o
w
n
b
y
th
e
f
o
llo
w
i
n
g
eq
u
ati
o
n
;
∇
=
(
1
,
2
,
…
…
,
)
(
5
)
Evaluation Warning : The document was created with Spire.PDF for Python.
IJ
-
AI
I
SS
N:
2252
-
8938
P
erfo
r
ma
n
ce
A
n
a
lysi
s
o
f A
N
N
Mo
d
el
fo
r
E
s
tima
tio
n
o
f..
.
(
Tu
s
h
a
r
A
n
th
w
a
l
)
7
E
v
er
y
ti
m
e
w
ei
g
h
t
w
i
n
b
ac
k
p
r
o
p
ag
atio
n
is
u
p
d
ated
u
s
in
g
t
h
e
eq
u
atio
n
;
(
+
1
)
=
(
)
+
∇
(
)
(
6
)
∆
=
−
for
=
1
,
2
,
…
,
n
(
7
)
w
h
er
e
r
ep
r
esen
ts
a
lear
n
i
n
g
co
n
s
ta
n
t,
i.e
.
,
a
p
r
o
p
o
r
tio
n
ality
p
ar
a
m
eter
w
h
ic
h
d
ef
in
e
s
th
e
s
te
p
len
g
t
h
o
f
ea
c
h
iter
atio
n
in
t
h
e
n
eg
ati
v
e
g
r
ad
ien
t d
ir
ec
tio
n
.
T
w
o
tr
ain
i
n
g
f
u
n
ctio
n
s
;
L
ev
e
n
b
er
g
-
Ma
r
q
u
ar
d
t
an
d
B
FGS
Q
u
asi
-
Ne
w
to
n
w
er
e
s
elec
ted
as
b
o
th
b
elo
n
g
to
th
e
d
if
f
er
en
t
clas
s
o
f
o
p
ti
m
i
za
tio
n
.
L
e
v
e
n
b
er
g
-
Ma
r
q
u
ar
d
t
is
m
o
s
t w
id
el
y
u
s
ed
o
p
ti
m
izati
o
n
alg
o
r
it
h
m
a
n
d
i
s
b
ased
o
n
g
r
ad
ie
n
t
d
esce
n
t
a
n
d
Gau
s
s
-
Ne
w
to
n
iter
atio
n
,
w
h
e
r
ea
s
B
FGS
b
elo
n
g
s
to
t
h
e
f
a
m
il
y
o
f
h
ill
cli
m
b
i
n
g
o
p
tim
izatio
n
tec
h
n
iq
u
e
a
n
d
Qu
asi
Ne
w
to
n
is
a
m
e
t
h
o
d
to
f
in
d
th
e
r
o
o
t
o
f
th
e
f
ir
s
t
d
er
iv
a
tiv
e.
L
ik
e
w
is
e,
b
o
th
o
f
th
ese
f
u
n
c
tio
n
d
e
f
in
e
s
m
ea
n
s
q
u
ar
e
er
r
o
r
v
ar
iab
le
as a
p
er
f
o
r
m
an
ce
a
n
al
y
s
i
s
p
ar
a
m
eter
w
h
ich
m
ak
e
s
o
u
r
w
o
r
k
s
tr
aig
h
t
f
o
r
w
ar
d
to
co
r
r
elate
a
n
d
co
m
p
ar
e
b
o
th
th
e
f
u
n
ctio
n
s
.
1.
B
a
ck
pro
pa
g
a
t
io
n w
it
h L
M
a
lg
o
rit
h
m
A
cc
o
r
d
in
g
to
Ha
g
a
n
a
n
d
Me
n
h
aj
(
1
9
9
4
)
,
b
ac
k
p
r
o
p
ag
atio
n
al
g
o
r
ith
m
w
h
e
n
co
m
b
i
n
ed
w
ith
t
h
e
L
e
v
en
b
er
g
Ma
r
q
u
ar
d
t
tr
ain
in
g
f
u
n
ctio
n
,
i
m
p
r
o
v
ed
o
v
er
th
e
tr
ain
i
n
g
ti
m
e
an
d
i
m
p
ar
te
d
m
o
m
en
tu
m
to
it.
An
d
ac
co
r
d
in
g
to
Z
h
o
u
a
n
d
Si
(
1
9
9
8
)
,
th
ey
p
r
o
v
ed
th
a
t
L
M
al
s
o
en
h
a
n
ce
s
t
h
e
p
er
f
o
r
m
a
n
ce
o
f
b
ac
k
p
r
o
p
ag
atio
n
alg
o
r
ith
m
t
h
u
s
p
r
o
v
id
in
g
s
u
p
er
io
r
r
esu
lts
an
d
tr
ain
in
g
ac
cu
r
a
c
y
.
2.
B
a
ck
pro
pa
g
a
t
io
n w
it
h
B
F
G
S qua
s
i
-
new
t
o
n
a
lg
o
rit
h
m
B
FGS
is
also
ap
t
f
o
r
s
o
l
v
i
n
g
lar
g
e
n
o
n
li
n
ea
r
o
p
ti
m
izatio
n
an
d
s
u
ited
f
o
r
th
e
p
r
o
b
lem
with
a
lar
g
e
n
u
m
b
er
o
f
v
ar
iab
les (
F
letch
er
,
1
9
8
7
)
.
4.
3
.
Si
m
ula
t
io
n a
nd
O
utput
T
h
e
s
ec
o
n
d
ar
y
d
ata
is
co
llected
ac
r
o
s
s
v
ar
io
u
s
s
o
u
r
ce
s
is
clea
n
ed
an
d
ab
o
u
t
4
0
0
0
s
am
p
les
w
er
e
f
i
n
alize
d
to
tr
ai
n
,
test
an
d
v
ali
d
ate
th
e
n
et
w
o
r
k
.
Si
m
u
latio
n
i
s
ca
r
r
ied
o
u
t
w
i
th
Ma
t
lab
Ne
u
r
al
Net
w
o
r
k
T
o
o
lb
o
x
an
d
r
esu
l
ts
w
er
e
f
u
r
th
er
a
n
al
y
ze
d
.
Sp
ec
if
icatio
n
s
o
f
th
e
n
e
u
r
al
n
et
w
o
r
k
to
co
m
p
u
te
t
h
e
T
SI
v
al
u
es
o
u
t
f
r
o
m
s
ev
e
n
in
p
u
t p
ar
a
m
eter
s
ar
e
p
r
o
v
id
ed
in
th
e
T
ab
le
2
.
T
ab
le
2
.
Sp
ec
if
icatio
n
o
f
Ne
u
r
al
Net
w
o
r
k
S
.
N
o
.
C
h
a
r
a
c
t
e
r
i
st
i
c
s
V
a
l
u
e
s
1
Ty
p
e
o
f
n
e
t
w
o
r
k
F
e
e
d
F
o
r
w
a
r
d
B
a
c
k
P
r
o
p
a
g
a
t
i
o
n
2
N
o
.
o
f
n
e
u
r
o
n
s
i
n
t
h
e
i
n
p
u
t
l
a
y
e
r
10
3
N
o
.
o
f
h
i
d
d
e
n
l
a
y
e
r
2
4
N
o
.
o
f
n
e
u
r
o
n
s
i
n
t
h
e
o
u
t
p
u
t
l
a
y
e
r
1
5
N
o
.
o
f
I
n
p
u
t
p
a
r
a
me
t
e
r
s
7
6
N
o
.
o
f
O
u
t
p
u
t
p
a
r
a
me
t
e
r
1
7
L
e
n
g
t
h
o
f
d
a
t
a
sam
p
l
e
4
0
0
0
5
P
e
r
f
o
r
man
c
e
f
u
n
c
t
i
o
n
M
S
E
6
T
r
a
i
n
i
n
g
F
u
n
c
t
i
o
n
L
M
a
n
d
B
F
G
S
Q
u
a
si
N
e
w
t
o
n
7
A
d
a
p
t
i
o
n
L
e
a
r
n
i
n
g
F
u
n
c
t
i
o
n
L
E
A
R
N
G
D
M
8
A
c
t
i
v
a
t
i
o
n
f
u
n
c
t
i
o
n
i
n
t
h
e
T
a
n
-
si
g
mo
i
d
h
i
d
d
e
n
l
a
y
e
r
T
a
n
-
si
g
mo
i
d
9
A
c
t
i
v
a
t
i
o
n
f
u
n
c
t
i
o
n
i
n
t
h
e
L
i
n
e
a
r
o
u
t
p
u
t
l
a
y
e
r
L
i
n
e
a
r
10
M
a
x
i
m
u
m
n
o
.
o
f
e
p
o
c
h
s
1
0
0
0
11
M
i
n
i
m
u
m
M
S
E
V
a
l
u
e
0
Am
o
n
g
s
t
4
0
0
0
s
et
o
f
d
ata
is
u
s
ed
7
5
%
o
f
d
ata
is
u
s
ed
to
tr
ain
th
e
n
et
w
o
r
k
,
1
5
%
as
test
i
n
g
s
et
an
d
r
e
m
ain
in
g
1
0
%
f
o
r
v
alid
ate
o
f
th
e
n
et
w
o
r
k
.
I
n
o
r
d
er
to
u
p
d
ate
th
e
w
ei
g
h
t
in
t
h
e
n
et
w
o
r
k
a
n
d
m
a
k
e
th
e
s
y
s
te
m
to
lear
n
,
tr
ain
in
g
d
ata
s
et
p
er
f
o
r
m
s
.
T
h
e
v
alid
at
io
n
s
et
e
s
ti
m
ate
th
e
p
er
f
o
r
m
a
n
ce
o
f
t
h
e
n
et
w
o
r
k
a
n
d
test
d
ata
s
et
co
m
p
u
te
s
th
e
o
v
er
all
p
er
f
o
r
m
a
n
ce
o
f
th
e
n
et
w
o
r
k
.
A
d
a
p
tatio
n
lear
n
in
g
f
u
n
ctio
n
th
at
co
r
r
esp
o
n
d
s
to
th
e
m
o
m
e
n
t
u
m
v
ar
ia
n
t
o
f
B
ac
k
P
r
o
p
ag
atio
n
i.e
.
L
E
A
R
N
GDM
,
t
h
e
g
r
ad
ie
n
t
d
esce
n
t
w
it
h
m
o
m
en
tu
m
w
eig
h
t
a
n
d
b
ias lea
r
n
in
g
f
u
n
ctio
n
w
h
ich
r
etu
r
n
s
th
e
w
ei
g
h
t c
h
a
n
g
e
an
d
a
n
e
w
lear
n
i
n
g
s
tate.
5.
P
E
RF
O
RM
ANCE
ANAL
YS
I
S O
F
NE
UR
AL
N
E
T
WO
R
K
Sev
er
al
p
h
a
s
es
o
f
tr
ai
n
i
n
g
w
er
e
p
er
f
o
r
m
ed
o
n
t
h
e
n
eu
r
al
n
e
t
w
o
r
k
w
i
th
ab
o
v
e
m
en
tio
n
ed
ch
ar
ac
ter
is
tic
s
.
T
h
e
r
esu
lt
s
o
f
th
e
n
e
t
w
o
r
k
p
er
f
o
r
m
a
n
ce
o
n
b
asis
o
f
b
est
v
alid
atio
n
p
er
f
o
r
m
a
n
ce
a
n
d
lo
w
e
s
t
g
r
ad
ien
t
a
n
d
h
i
g
h
e
s
t
r
eg
r
es
s
io
n
r
esu
l
ts
w
er
e
r
ec
o
r
d
ed
.
T
o
ev
alu
ate
t
h
e
p
er
f
o
r
m
a
n
c
e
o
f
th
e
n
e
t
w
o
r
k
,
th
e
co
m
p
ar
ativ
e
an
al
y
s
is
o
f
th
e
r
esu
lt
s
o
f
f
ee
d
f
o
r
w
ar
d
b
ac
k
p
r
o
p
ag
atio
n
n
eu
r
al
n
et
w
o
r
k
w
i
th
B
FGS
Qu
a
s
i
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8938
IJ
-
AI
Vo
l.
7
,
No
.
1
,
Ma
r
ch
2
0
1
8
:
1
–
10
8
Ne
w
to
n
an
d
L
e
v
e
n
b
er
g
-
Ma
r
q
u
ar
d
t
w
er
e
r
ec
o
r
d
ed
.
I
n
th
e
o
v
er
all
an
al
y
s
is
,
t
h
e
later
w
a
s
f
o
u
n
d
to
d
is
p
la
y
m
o
r
e
ac
cu
r
ate
an
d
ef
f
i
cie
n
t o
u
tp
u
ts
o
v
er
th
e
f
o
r
m
er
alg
o
r
ith
m
.
T
h
e
d
etail
r
esu
lts
w
er
e
s
h
o
w
n
in
th
e
T
ab
le
3
.
T
ab
le
3
.
P
er
f
o
r
m
a
n
ce
A
n
al
y
s
i
s
o
f
Feed
Fo
r
w
ar
d
B
ac
k
P
r
o
p
a
g
atio
n
Ne
u
r
al
Net
w
o
r
k
Mo
d
el
b
ased
o
n
B
FGS
Qu
asi Ne
w
to
n
a
n
d
L
e
v
en
b
er
g
-
Ma
r
q
u
ar
d
t tr
ain
i
n
g
alg
o
r
it
h
m
.
To
g
et
t
h
e
p
er
f
o
r
m
a
n
ce
o
f
t
h
e
n
e
u
r
al
n
e
t
w
o
r
k
,
er
r
o
r
v
ec
to
r
is
a
u
to
m
a
ticall
y
co
m
p
u
ted
.
T
h
ese
er
r
o
r
v
ec
to
r
s
d
en
o
te
f
o
r
h
o
w
m
a
n
y
t
i
m
es
a
n
d
u
p
to
w
h
at
ex
te
n
t
t
h
e
n
et
w
o
r
k
w
as
u
n
ab
le
to
p
r
ed
ict
th
e
co
r
r
ec
t
tar
g
et
v
alu
e
a
n
d
u
p
to
w
h
at
e
x
ten
d
t
h
e
er
r
o
r
w
as.
T
h
e
o
v
er
all
n
e
u
r
al
n
et
w
o
r
k
m
ea
n
s
s
q
u
ar
e
er
r
o
r
w
as
also
r
ec
o
r
d
ed
w
h
ic
h
th
e
f
ac
to
r
f
o
r
m
ea
s
u
r
e
m
en
t o
f
er
r
o
r
w
as
m
ad
e
b
y
co
m
p
u
tatio
n
m
o
d
el
t
h
r
o
u
g
h
o
u
t t
h
e
lear
n
in
g
p
h
ase
s
.
T
h
e
p
er
f
o
r
m
a
n
ce
s
o
f
b
o
th
t
h
e
lear
n
in
g
f
u
n
ct
io
n
s
w
er
e
f
u
r
th
e
r
an
al
y
ze
d
b
y
g
i
v
i
n
g
th
e
s
a
m
e
s
et
o
f
d
ata
a
s
in
p
u
t
to
b
o
th
th
e
n
et
w
o
r
k
an
d
co
m
p
u
ti
n
g
th
e
o
u
tco
m
e
o
n
T
SI
v
alu
e.
T
h
e
p
r
e
d
icted
o
u
tp
u
t
f
r
o
m
b
o
th
t
h
e
n
et
w
o
r
k
w
as
e
v
al
u
ated
ag
a
in
s
t
th
e
ac
tu
al
v
al
u
e
s
o
f
T
SI.
A
s
it
co
u
ld
b
e
i
n
ter
p
r
eted
f
r
o
m
F
ig
u
r
e
4
,
t
h
e
o
u
tp
u
t
g
en
er
ated
f
r
o
m
L
M
tr
ain
i
n
g
al
g
o
r
ith
m
s
w
a
s
m
o
s
tl
y
f
o
u
n
d
to
b
e
n
ea
r
er
to
th
e
ac
tu
al
T
SI
v
alu
e.
Err
o
r
i
.e.
dev
i
a
t
i
on
f
r
om
t
h
e
ac
t
ua
l
T
SI
v
al
ue
and
g
ene
r
at
ed
o
ut
pu
t
f
r
om
t
he
n
et
w
or
k
s
w
er
e
al
so
p
l
o
t
t
e
d
t
o
de
pi
c
t
a
be
t
t
e
r
pi
c
t
ur
e
of
t
he
r
es
u
l
t
s.
A
nd
as
t
he
F
i
g
ur
e
5
and
6
g
oes
,
t
he
out
c
om
e
f
r
om
B
FG
D
Q
uas
i
N
ewton
l
ea
r
ni
ng
al
g
or
i
t
hm
di
spl
ay
s
t
h
e
l
ar
g
er
num
ber
o
f
di
s
pl
a
c
em
ent
s
an
d
l
es
se
r
ac
cu
r
ac
y
ov
er
t
h
e
a
ct
u
al
da
t
a.
H
ow
ev
er
,
LM
a
l
g
or
i
t
hm
shows
a
l
es
se
r
num
ber
of
f
l
uct
ua
t
i
on
and
ev
en
up t
o a
l
e
ss
e
r
ex
t
en
t
, w
h
i
ch
shows i
t
s
ac
cu
r
a
cy
and ov
er
al
l
e
f
f
i
c
i
enc
y
ov
er
i
t
s
coun
t
er
p
ar
t
.
Fig
u
r
e
4
.
A
s
s
es
s
m
e
n
t o
n
o
u
tp
u
t g
e
n
er
ated
b
y
Neu
r
al
Net
w
o
r
k
f
o
r
b
o
th
tr
ain
i
n
g
al
g
o
r
it
h
m
B
FGS Qu
a
s
i
Ne
w
to
n
an
d
L
ev
e
n
b
er
g
-
Ma
r
q
u
ar
d
t
Fig
u
r
e
5
.
E
r
r
o
r
p
lo
t f
o
r
Neu
r
al
Net
w
o
r
k
tr
ai
n
ed
w
it
h
B
FGD
Q
u
as
i N
e
w
to
n
lea
r
n
in
g
al
g
o
r
ith
m
Fig
u
r
e
6
.
E
r
r
o
r
p
lo
t f
o
r
Neu
r
al
Net
w
o
r
k
tr
ai
n
ed
w
it
h
L
e
v
en
b
er
g
-
Ma
r
q
u
ar
d
t le
a
r
n
in
g
al
g
o
r
ith
m
EP
OCH
Grad
ien
t
Be
st
Val
id
atio
n
P
erfo
rm
an
ce
(M
ean Sq
u
are
Erro
r)
Train
in
g
Val
id
atio
n
Te
st
Al
l
144
3
.53
1
.29
7
0
.51
5
1
2
0
.98
3
8
3
0
.99
1
9
9
0
.72
8
7
10
0
.86
0
.90
7
6
0
.99
0
6
6
0
.99
1
6
2
0
.98
9
4
5
0
.99
0
5
7
7
2
.26
1
.10
4
3
0
.99
0
3
3
0
.99
1
2
6
0
.99
0
9
3
0
.99
0
5
7
6
0
.91
8
0
.98
9
0
.99
0
8
2
0
.99
1
7
7
0
.98
8
5
5
0
.99
0
6
8
258
0
.56
1
0
.00
1
4
6
0
.99
9
9
9
1
0
.99
9
7
3
0
.99
9
9
4
6
0
.75
9
0
.00
0
7
0
.99
9
9
2
0
.99
9
9
7
0
.99
9
9
9
0
.99
9
9
4
6
0
.25
1
0
.00
0
4
0
.99
9
9
9
0
.99
9
9
9
0
.99
9
7
0
.99
9
9
4
6
1
.35
1
0
.00
0
5
0
.99
9
9
2
0
.99
9
9
7
0
.99
9
9
9
0
.99
9
9
4
Feed Fo
rward
Back
p
ro
p
ag
atio
n
2
-10
-1
M
o
d
el
L
ev
enb
erg-
M
arq
u
ard
t
L
EARN
GD
M
M
ean Sq
u
are
Erro
r
BF
GS
Q
u
asi N
ew
to
n
L
EARN
GD
M
M
ean Sq
u
are
Erro
r
Regressio
n
N
et
wo
rk T
y
p
e
L
earn
in
g
Fu
n
ctio
n
Ad
ap
tat
io
n
L
earn
in
g
Fu
n
ctio
n
P
erfo
rm
an
ce
Fu
n
ctio
n
P
erfo
rm
an
ce
Evaluation Warning : The document was created with Spire.PDF for Python.
IJ
-
AI
I
SS
N:
2252
-
8938
P
erfo
r
ma
n
ce
A
n
a
lysi
s
o
f A
N
N
Mo
d
el
fo
r
E
s
tima
tio
n
o
f..
.
(
Tu
s
h
a
r
A
n
th
w
a
l
)
9
6.
DIS
CU
SS
I
O
N
AND
CO
NC
L
US
I
O
N
T
h
e
m
ai
n
ai
m
o
f
th
i
s
r
esear
ch
w
o
r
k
w
as
to
f
o
r
m
u
late
f
ee
d
f
o
r
w
ar
d
b
ac
k
p
r
o
p
ag
atio
n
n
e
u
r
al
n
et
w
o
r
k
f
o
r
co
m
p
u
ti
n
g
eu
tr
o
p
h
icat
io
n
l
ev
el
o
f
w
ater
b
o
d
ies.
Di
f
f
er
e
n
t
p
h
y
s
ico
ch
e
m
ical
p
ar
a
m
eter
s
o
f
w
ater
b
o
d
ies
w
er
e
co
llected
an
d
A
NN
w
as
tr
ain
e
d
w
i
th
th
e
d
ata
u
s
i
n
g
t
w
o
lear
n
in
g
alg
o
r
it
h
m
s
;
L
M
an
d
B
FGD
Qu
asi
Ne
w
to
n
.
Am
o
n
g
th
e
t
w
o
lear
n
in
g
ap
p
r
o
ac
h
es,
L
M
p
r
o
v
ed
to
b
e
ef
f
icie
n
t
as
th
e
m
ea
n
s
q
u
ar
e
er
r
o
r
w
as
v
e
r
y
le
s
s
.
A
l
s
o
w
h
ile
ca
lib
r
atin
g
th
e
d
ev
iatio
n
f
r
o
m
ac
t
u
al
d
ata,
th
e
B
FGD
Qu
as
i
Ne
w
to
n
s
h
o
w
ed
m
u
c
h
h
i
g
h
er
v
al
u
es
an
d
u
n
ev
e
n
d
is
tr
ib
u
tio
n
,
t
h
u
s
d
is
p
la
y
i
n
g
t
h
e
in
f
er
io
r
o
u
t
p
u
t
f
o
r
T
SI
v
a
lu
e
s
f
o
r
t
h
e
g
iv
e
n
s
et
o
f
d
ata.
Feed
-
f
o
r
w
ar
d
b
ac
k
p
r
o
p
ag
atio
n
n
eu
r
al
n
et
w
o
r
k
w
it
h
L
e
v
en
b
er
g
-
Ma
r
q
u
ar
d
t
lear
n
i
n
g
a
lg
o
r
it
h
m
w
ill
b
e
f
u
r
t
h
er
u
s
ed
to
co
m
p
u
te
t
h
e
s
tat
u
s
o
f
lak
es o
n
b
asis
o
f
p
h
y
s
ico
ch
e
m
i
ca
l v
ar
iab
les an
d
p
r
ed
ict
th
e
eu
tr
o
p
h
icatio
n
le
v
el.
Fo
r
th
e
f
u
tu
r
e
s
tu
d
y
,
t
h
e
A
NN
w
o
u
ld
b
e
u
s
ed
to
co
m
p
u
te
th
e
eu
tr
o
p
h
icatio
n
lev
e
l
o
f
a
r
e
g
io
n
al
w
ater
b
o
d
y
w
i
t
h
th
e
h
elp
o
f
s
p
atial
d
ata.
W
it
h
t
h
e
i
m
a
g
er
y
d
ata
o
f
s
e
lecte
d
la
k
e,
m
ain
l
y
c
h
lo
r
o
p
h
y
ll
-
a,
SDT
an
d
to
tal
p
h
o
s
p
h
ate
(
o
n
e
o
r
m
o
r
e
v
ar
iab
le)
w
o
u
ld
b
e
co
m
p
u
ted
a
n
d
u
s
i
n
g
n
e
u
r
al
n
et
w
o
r
k
t
h
e
ac
tu
a
l
tr
o
p
h
ic
s
tat
u
s
i
n
d
ex
o
f
t
h
e
la
k
e
w
o
u
l
d
b
e
ass
ess
ed
.
T
h
e
i
m
a
g
er
y
d
ata
o
f
s
e
lecte
d
w
ater
r
e
s
er
v
o
ir
w
o
u
ld
b
e
u
s
ed
to
co
m
p
u
te
t
h
e
SDT
an
d
C
h
lo
r
o
p
h
y
ll
co
n
te
n
t
at
d
if
f
er
e
n
t
s
ite
s
an
d
o
n
d
i
f
f
er
e
n
t
ti
m
e
s
p
an
.
T
h
e
o
u
tp
u
t
f
r
o
m
th
e
s
p
atial
d
ata
w
o
u
ld
b
e
s
u
b
j
ec
ted
to
an
al
y
s
is
u
s
in
g
s
tat
is
tical
m
o
d
eli
n
g
a
n
d
n
et
o
u
tco
m
e
w
o
u
ld
b
e
p
ass
ed
o
n
t
o
th
e
a
b
o
v
e
tr
ain
ed
n
eu
r
al
n
et
w
o
r
k
to
co
m
p
u
te
T
SI
o
f
th
e
w
ater
b
o
d
y
.
T
h
e
o
v
er
all
e
f
f
icien
c
y
an
d
ac
cu
r
ac
y
o
f
th
e
s
y
s
te
m
w
o
u
ld
b
e
co
m
p
u
ted
th
u
s
g
i
v
i
n
g
a
v
is
io
n
to
w
ar
d
s
au
to
m
ated
an
al
y
ze
r
f
o
r
m
o
n
ito
r
in
g
w
ater
b
o
d
ies.
ACK
NO
WL
E
D
G
E
M
E
NT
S
I
w
o
u
ld
li
k
e
to
ac
k
n
o
w
led
g
e
m
y
m
e
n
to
r
an
d
f
ello
w
r
esear
c
h
er
s
f
o
r
t
h
eir
in
p
u
t in
t
h
is
r
ese
ar
ch
w
o
r
k
.
RE
F
E
R
E
NC
E
S
[1
]
S
in
g
h
,
S
.
,
&
Brij
,
G
.
(2
0
0
2
).
I
n
teg
ra
ted
M
a
n
a
g
e
me
n
t
o
f
W
a
ter
Res
o
u
rc
e
s
o
f
L
a
k
e
Na
i
n
it
a
l
a
n
d
it
s
W
a
ter
sh
e
d
:
A
n
En
v
iro
n
me
n
ta
l
Ec
o
n
o
mic
s A
p
p
ro
a
c
h
.
E
ERC,
I
n
d
ira G
a
n
d
h
i
In
stit
u
te f
o
r
De
v
e
lo
p
m
e
n
tal
Re
se
a
rc
h
,
M
u
m
b
a
i,
In
d
ia.
[2
]
S
h
a
rm
a
,
M
.
W
a
ter
Qu
a
li
t
y
A
ss
e
ss
m
e
n
t
o
f
th
e
Ce
n
tral
Him
a
la
y
a
n
L
a
k
e
,
Na
in
it
a
l.
Ad
v
a
n
c
e
s
i
n
En
v
iro
n
me
n
ta
l
Ch
e
mistry
,
2
0
1
4
,
1
-
5
.
h
tt
p
:/
/d
x
.
d
o
i.
o
rg
/1
0
.
1
1
5
5
/2
0
1
4
/
4
7
3
0
7
4
.
[3
]
An
th
w
a
l,
T
.
,
&
P
a
n
d
e
y
,
M
.
(2
0
1
7
).
A
s
se
ss
in
g
S
tate
o
f
th
e
A
rt
o
n
Artif
icia
l
Ne
u
ra
l
N
e
t
w
o
rk
P
a
ra
d
ig
m
s
f
o
r
Lev
e
l
o
f
Eu
tro
p
h
ica
ti
o
n
Esti
m
a
ti
o
n
o
f
W
a
ter Bo
d
ies
.
IA
ES
I
n
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
Arti
fi
c
ia
l
In
telli
g
e
n
c
e
,
5
(4
),
1
3
5
-
1
4
2
.
[4
]
Tw
ig
t,
D.,
Re
g
o
,
J.,
T
y
rr
e
ll
,
D.,
&
T
ro
o
st,
T
.
(2
0
1
1
).
W
a
ter
Q
u
a
l
it
y
Fo
re
c
a
stin
g
S
y
ste
ms
:
Ad
v
a
n
c
e
d
W
a
rn
i
n
g
o
f
Ha
rm
fu
l
Eve
n
ts
a
n
d
Diss
e
min
a
ti
o
n
o
f
Pu
b
li
c
Al
e
rts
.
I
n
8
th
In
ter
n
a
ti
o
n
a
l
IS
CRA
M
Co
n
f
e
re
n
c
e
.
L
isb
o
n
,
P
o
rtu
g
a
l.
[5
]
Pl
a
n
n
in
g
a
n
d
M
a
n
a
g
e
me
n
t
o
f
L
a
k
e
s
a
n
d
Res
e
rv
o
irs:
An
In
te
g
ra
ted
Ap
p
r
o
a
c
h
to
Eu
tro
p
h
ica
t
io
n
.
(
2
0
1
7
).
U
n
e
p
.
o
r
.
jp
.
[6
]
W
a
h
a
a
b
,
R.
,
&
Ba
d
a
wy
,
M
.
(2
0
0
4
).
W
a
ter
q
u
a
li
ty
a
ss
e
ss
m
e
n
t
o
f
th
e
Riv
e
r
Nile
sy
ste
m
:
a
n
o
v
e
rv
ie
w
.
Bi
o
me
d
ica
l
and
En
v
iro
n
me
n
ta
l
S
c
ien
c
e
s
,
1
7
(1
)
,
8
7
-
1
0
0
.
[7
]
Ne
a
m
tu
,
M
.
,
C
iu
m
a
su
,
I.
,
C
o
stica
,
N.,
C
o
stica
,
M
.
,
B
o
b
u
,
M
.
,
Nic
o
a
ra
,
M
.
,
Ca
tri
n
e
sc
u
,
C.
,
v
a
n
,
S
lo
o
ten
,
K.,
De
,
&
A
len
c
a
stro
,
L
.
(
2
0
0
9
).
Ch
e
m
ica
l,
b
i
o
lo
g
ica
l,
a
n
d
e
c
o
to
x
ico
l
o
g
ica
l
a
ss
e
ss
m
e
n
t
o
f
p
e
sticid
e
s
a
n
d
p
e
rsiste
n
t
o
rg
a
n
i
c
p
o
ll
u
tan
ts
in
t
h
e
Ba
h
l
u
i
Riv
e
r,
Ro
m
a
n
ia.
En
v
iro
n
me
n
ta
l
S
c
ien
c
e
a
n
d
Po
ll
u
t
io
n
Res
e
a
rc
h
,
1
6
(
1
),
76
-
8
5
.
[8
]
Ca
rre
ó
n
-
P
a
lau
,
L
.
,
P
a
rrish
,
C.
,
&
P
é
re
z
-
Esp
a
ñ
a
,
H
.
(
2
0
1
7
).
Urb
a
n
s
e
wa
g
e
li
p
id
s
in
th
e
su
sp
e
n
d
e
d
p
a
r
ti
c
u
late
m
a
tt
e
r
o
f
a
c
o
ra
l
re
e
f
u
n
d
e
r
riv
e
r
in
f
lu
e
n
c
e
in
th
e
S
o
u
th
W
e
st G
u
lf
o
f
M
e
x
ico
.
W
a
ter
Res
e
a
rc
h
,
1
2
3
,
1
9
2
-
2
0
5
.
[9
]
Ba
stid
a
s,
J.,
V
é
lez
,
J.,
Zam
b
ra
n
o
,
J.
&
L
o
n
d
o
ñ
o
,
A
.
(2
0
1
7
).
De
sig
n
o
f
w
a
ter
q
u
a
li
ty
m
o
n
i
to
rin
g
n
e
t
w
o
rk
s
w
it
h
tw
o
in
f
o
rm
a
ti
o
n
sc
e
n
a
rio
s in
tr
o
p
ica
l
A
n
d
e
a
n
b
a
sin
s.
E
n
v
iro
n
me
n
ta
l
S
c
ien
c
e
a
n
d
Po
ll
u
t
io
n
Res
e
a
rc
h
,
1
-
1
5
.
[1
0
]
A
l
f
e
r
e
s,
J.,
T
ik
,
S
.
,
Co
p
p
,
J.,
&
V
a
n
r
o
ll
e
g
h
e
m
,
P
.
(2
0
1
3
).
A
d
v
a
n
c
e
d
m
o
n
it
o
ri
n
g
o
f
w
a
ter
s
y
ste
m
s
u
sin
g
in
situ
m
e
a
su
re
m
e
n
t
sta
ti
o
n
s
:
d
a
ta v
a
li
d
a
ti
o
n
a
n
d
f
a
u
lt
d
e
tec
ti
o
n
.
W
a
ter
S
c
ien
c
e
&
T
e
c
h
n
o
lo
g
y
,
6
8
(5
),
1
0
2
2
-
30.
[1
1
]
Ch
a
n
g
,
N.,
Ba
i,
K.,
&
Ch
e
n
,
C
.
(
2
0
1
7
).
In
teg
ra
ti
n
g
m
u
lt
ise
n
so
r
sa
telli
te
d
a
ta
m
e
rg
in
g
a
n
d
im
a
g
e
re
c
o
n
stru
c
ti
o
n
i
n
su
p
p
o
rt
o
f
m
a
c
h
in
e
lea
rn
i
n
g
f
o
r
b
e
tt
e
r
w
a
ter
q
u
a
li
ty
m
a
n
a
g
e
m
e
n
t.
J
o
u
rn
a
l
o
f
En
v
iro
n
me
n
t
a
l
M
a
n
a
g
e
me
n
t
,
2
0
1
,
2
2
7
-
2
4
0
.
[1
2
]
X
iao
-
k
a
i,
W
.
,
Zh
e
n
g
-
y
a
,
G
.
,
Qi
-
li,
Z.
,
&
Da
-
q
u
a
n
,
W
.
(2
0
1
0
).
In
f
o
rm
a
ti
o
n
S
y
ste
m
De
sig
n
o
f
W
a
ter
Po
ll
u
ti
o
n
.
I
n
In
tern
a
ti
o
n
a
l
Co
n
f
e
re
n
c
e
o
n
Ch
a
ll
e
n
g
e
s
in
En
v
iro
n
m
e
n
tal
S
c
ien
c
e
a
n
d
Co
m
p
u
ter
E
n
g
in
e
e
rin
g
(p
p
.
3
1
5
-
3
1
8
).
W
u
h
a
n
,
Ch
in
a
:
IE
EE
.
[1
3
]
Ch
a
u
,
K
.
,
(
2
0
0
9
)
.
A
Re
v
ie
w
o
n
I
n
teg
ra
ti
o
n
o
f
A
rti
f
icia
l
In
telli
g
e
n
c
e
in
t
o
W
a
ter
Qu
a
li
ty
M
o
d
e
ll
in
g
.
M
a
rin
e
P
o
ll
u
ti
o
n
Bu
ll
e
ti
n
,
5
2
(
7
),
7
2
6
-
7
3
3
.
[1
4
]
S
a
lam
i,
E.
,
S
a
lari,
M
.
,
E
h
tes
h
a
m
i,
M
.
,
B
id
o
k
h
ti
,
N.
,
&
G
h
a
d
im
i,
H.
(2
0
1
6
).
A
p
p
li
c
a
ti
o
n
o
f
a
rti
f
icia
l
n
e
u
ra
l
n
e
tw
o
rk
s
a
n
d
m
a
th
e
m
a
ti
c
a
l
m
o
d
e
li
n
g
f
o
r
th
e
p
re
d
icti
o
n
o
f
w
a
ter
q
u
a
li
ty
v
a
ria
b
les
(c
a
se
stu
d
y
:
so
u
th
w
e
st
o
f
Ira
n
).
De
sa
li
n
a
ti
o
n
a
n
d
W
a
ter
T
re
a
tme
n
t
,
5
7
(5
6
),
2
7
0
7
3
-
2
7
0
8
4
.
[1
5
]
S
e
o
,
I.
,
Yu
n
,
S
.
&
C
h
o
i
,
S
.
(2
0
1
6
).
F
o
re
c
a
stin
g
W
a
ter
Q
u
a
li
ty
P
a
ra
me
ter
s
b
y
ANN
M
o
d
e
l
Us
i
n
g
Pre
-
p
ro
c
e
ss
in
g
T
e
c
h
n
iq
u
e
a
t
th
e
D
o
wn
stre
a
m
o
f
Ch
e
o
n
g
p
y
e
o
n
g
Da
m
.
In
1
2
th
In
tern
a
ti
o
n
a
l
Co
n
f
e
re
n
c
e
o
n
Hy
d
ro
in
f
o
r
m
a
ti
c
s
(HIC
2
0
1
6
)
-
S
m
a
rt
W
a
t
e
r
f
o
r
th
e
F
u
t
u
r
e
,
1
5
4
,
1
1
1
0
-
1
1
1
5
.
[1
6
]
Ka
n
d
a
,
E.
,
Ki
p
k
o
rir
,
E.
,
&
Ko
sg
e
i
,
J.
(
2
0
1
6
)
.
Diss
o
lv
e
d
o
x
y
g
e
n
m
o
d
e
ll
i
n
g
u
si
n
g
a
rti
f
icia
l
n
e
u
ra
l
n
e
tw
o
rk
:
a
c
a
se
o
f
Riv
e
r
Nz
o
ia,
L
a
k
e
V
icto
ria b
a
si
n
,
Ke
n
y
a
.
J
o
u
rn
a
l
o
f
W
a
ter
S
e
c
u
rit
y
,
2
,
1
-
7.
[1
7
]
L
iu
,
M
.
,
&
L
u
,
J.
(2
0
1
4
).
S
u
p
p
o
rt
v
e
c
to
r
m
a
c
h
in
e
―a
n
a
lt
e
rn
a
ti
v
e
to
a
rti
f
icia
l
n
e
u
ro
n
n
e
tw
o
rk
f
o
r
w
a
ter
q
u
a
li
ty
f
o
re
c
a
stin
g
in
a
n
a
g
ricu
lt
u
ra
l
n
o
n
p
o
in
t
so
u
rc
e
p
o
l
lu
ted
riv
e
r?
En
v
iro
n
me
n
ta
l
S
c
ien
c
e
a
n
d
Po
ll
u
ti
o
n
Re
se
a
rc
h
,
2
1
(
1
8
)
,
1
1
0
3
6
–
1
1
0
5
3
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8938
IJ
-
AI
Vo
l.
7
,
No
.
1
,
Ma
r
ch
2
0
1
8
:
1
–
10
10
[1
8
]
S
a
tt
a
ri,
M
.
,
Jo
u
d
i
,
A
.
,
&
Ku
sia
k
,
A
.
(2
0
1
6
).
Esti
m
a
ti
o
n
o
f
W
a
ter
Qu
a
li
ty
P
a
ra
m
e
ters
w
it
h
Da
ta
-
Driv
e
n
M
o
d
e
l
.
J
o
u
rn
a
l
-
Ame
ric
a
n
W
a
ter
W
o
rk
s A
ss
o
c
ia
ti
o
n
,
1
0
8
(4
),
2
3
2
-
2
3
9
.
[1
9
]
Ha
y
k
in
,
S
.
(1
9
9
9
).
Ne
u
ra
l
n
e
tw
o
r
k
s
:
A
Co
mp
re
h
e
n
siv
e
F
o
u
n
d
a
ti
o
n
.
L
o
n
d
o
n
:
P
re
n
ti
c
e
-
Ha
ll
In
tern
a
ti
o
n
a
l.
[2
0
]
He
y
d
a
ri,
M
.
,
Oly
a
ie,
E.
,
M
o
h
e
b
z
a
d
e
h
,
H.,
&
Kisi,
Ö.
(
2
0
1
3
).
D
e
v
e
lo
p
m
e
n
t
o
f
a
Ne
u
ra
l
Ne
tw
o
rk
T
e
c
h
n
iq
u
e
f
o
r
P
re
d
ictio
n
o
f
W
a
ter
Qu
a
li
t
y
P
a
ra
m
e
ters
in
th
e
De
la
w
a
re
Ri
v
e
r,
P
e
n
n
sy
lv
a
n
ia.
M
id
d
le
-
Ea
st
J
o
u
r
n
a
l
o
f
S
c
ien
ti
fi
c
Res
e
a
rc
h
,
1
3
(1
0
),
1
3
6
7
-
1
3
7
6
.
[2
1
]
T
h
a
ir,
S
.
,
Ha
m
e
e
d
,
M
.
,
&
Ay
a
d
,
S
.
(2
0
1
4
).
P
re
d
icti
o
n
o
f
wa
ter
q
u
a
li
ty
o
f
Eu
p
h
ra
tes
Riv
e
r
b
y
u
sin
g
a
rti
f
icia
l
n
e
u
ra
l
n
e
tw
o
rk
m
o
d
e
l
(sp
a
ti
a
l
a
n
d
tem
p
o
ra
l
stu
d
y
).
In
ter
n
a
ti
o
n
a
l
Res
e
a
rc
h
J
o
u
r
n
a
l
o
f
Na
t
u
ra
l
S
c
ien
c
e
s
,
2
(
3
),
2
5
-
38.
[2
2
]
Ch
u
,
H.,
L
u
,
W
.
&
Zh
a
n
g
,
L
.
(2
0
1
3
).
A
p
p
li
c
a
ti
o
n
o
f
A
rti
f
icia
l
N
e
u
ra
l
Ne
t
w
o
rk
in
En
v
iro
n
m
e
n
tal
W
a
ter
Qu
a
li
t
y
A
s
se
ss
m
e
n
t.
J
o
u
rn
a
l
o
f
A
g
ric
u
lt
u
ra
l
S
c
ien
c
e
a
n
d
T
e
c
h
n
o
l
o
g
y
,
1
5
,
3
4
3
-
3
5
6
.
[2
3
]
S
o
lt
a
n
i
,
F
.
,
Ke
ra
c
h
ian
,
R.
,
&
S
h
i
ra
n
g
i,
E.
(
2
0
1
0
).
De
v
e
lo
p
in
g
o
p
e
ra
ti
n
g
ru
les
f
o
r
re
se
rv
o
irs
c
o
n
sid
e
rin
g
th
e
w
a
ter
q
u
a
li
ty
issu
e
s: ap
p
li
c
a
ti
o
n
o
f
A
N
F
IS
-
b
a
se
d
su
rr
o
g
a
te m
o
d
e
ls.
Exp
e
rt S
y
ste
ms
wit
h
Ap
p
li
c
a
ti
o
n
s
,
3
7
(
9
),
6
6
3
9
-
6
6
4
5
.
[2
4
]
Ch
e
n
, W
.
,
&
L
iu
,
W
.
(2
0
1
5
)
.
W
a
ter Qu
a
li
ty
M
o
d
e
li
n
g
in
Re
se
rv
o
irs
Us
in
g
M
u
lt
iv
a
riate
L
in
e
a
r
R
e
g
r
e
ss
io
n
a
n
d
T
w
o
Ne
u
ra
l
Ne
tw
o
rk
M
o
d
e
ls.
Ad
v
a
n
c
e
s in
Arti
fi
c
i
a
l
Ne
u
r
a
l
S
y
ste
ms
,
1
5
,
1
-
1
2
.
[2
5
]
A
re
e
ra
c
h
a
k
u
l,
S
.
&
S
a
n
g
u
a
n
sin
tu
k
u
l,
S
.
(
2
0
0
9
)
.
W
a
ter
q
u
a
li
ty
c
la
ss
if
ica
ti
o
n
u
si
n
g
n
e
u
r
a
l
n
e
two
rk
s:
Ca
se
stu
d
y
o
f
c
a
n
a
ls
in
B
a
n
g
k
o
k
,
T
h
a
il
a
n
d
.
I
n
I
n
tern
a
ti
o
n
a
l
C
o
n
f
e
re
n
c
e
f
o
r
In
tern
e
t
T
e
c
h
n
o
lo
g
y
a
n
d
S
e
c
u
re
d
T
ra
n
sa
c
ti
o
n
s
,
(ICIT
S
T
),
(p
p
.
1
-
5
),
L
o
n
d
o
n
,
UK
:
IEE
E.
[2
6
]
Ha
g
a
n
,
M
.
,
&
M
e
n
h
a
j,
M
.
(1
9
9
4
).
T
ra
in
in
g
fe
e
d
f
o
r
w
a
rd
n
e
t
w
o
rk
s w
it
h
th
e
M
a
rq
u
a
rd
t
a
lg
o
ri
th
m
.
IEE
E
tra
n
s
a
c
ti
o
n
s
o
n
Ne
u
r
a
l
Ne
two
rk
s
,
5
(6
)
,
9
8
9
-
9
9
3
.
[2
7
]
Zh
o
u
,
G
.
,
&
S
i,
J.
(
1
9
9
8
).
A
d
v
a
n
c
e
d
n
e
u
ra
l
-
n
e
tw
o
rk
train
in
g
a
lg
o
r
it
h
m
w
it
h
re
d
u
c
e
d
c
o
m
p
lex
it
y
b
a
se
d
o
n
Ja
c
o
b
ian
d
e
f
icie
n
c
y
.
IEE
E
T
ra
n
s Ne
u
r
a
l
N
e
two
rk
,
9
(3
)
,
4
4
8
-
4
5
3
.
[2
8
]
B
y
rd
,
H.,
L
u
,
P
.
,
No
c
e
d
a
l,
J.,
&
Z
h
u
,
C.
(1
9
9
5
).
A
L
i
m
it
e
d
M
e
m
o
r
y
A
lg
o
rit
h
m
f
o
r
Bo
u
n
d
Co
n
stra
in
e
d
Op
ti
m
iza
ti
o
n
.
S
IAM
J
o
u
rn
a
l
o
n
S
c
ien
t
if
ic Co
m
p
u
ti
n
g
,
1
6
(
5
),
1
1
9
0
-
1
2
0
8
.
[2
9
]
Ca
rlso
n
,
R.
(1
9
7
7
).
A
tro
p
h
ic sta
t
e
in
d
e
x
f
o
r
lak
e
s.
L
imn
o
l
o
g
y
a
n
d
Oc
e
a
n
o
g
r
a
p
h
y
,
2
2
(2
)
,
3
6
1
-
3
6
9
.
[3
0
]
R.
F
letc
h
e
r
(
1
9
8
7
)
,
Pr
a
c
ti
c
a
l
me
t
h
o
d
s
o
f
o
p
ti
miza
ti
o
n
(
2
n
d
e
d
.
),
N
e
w
Yo
rk
:
Jo
h
n
W
il
e
y
&
S
o
n
s,
I
S
BN
9
7
8
-
0
-
4
7
1
-
9
1
5
4
7
-
8.
BI
O
G
RAP
H
I
E
S
O
F
AUTH
O
RS
T
u
sh
a
r
A
n
th
w
a
l
is
a
n
A
ss
istan
t
P
r
o
f
e
ss
o
r
o
f
Co
m
p
u
ter
A
p
p
li
c
a
ti
o
n
s
a
t
De
v
Bh
o
o
m
i
In
stit
u
te
o
f
T
e
c
h
n
o
lo
g
y
,
De
h
ra
d
u
n
,
a
ff
il
iate
d
to
Uttara
k
h
a
n
d
T
e
c
h
n
ica
l
Un
iv
e
rsity
.
He
re
c
e
i
v
e
d
h
is
M
a
ste
rs
i
n
Co
m
p
u
ter
A
p
p
li
c
a
ti
o
n
f
ro
m
He
m
w
a
ti
N
a
n
d
a
n
G
a
rh
wa
l
Un
iv
e
rsit
y
(A
Ce
n
tral
Un
iv
e
rsit
y
),
S
rin
a
g
a
r
G
a
rh
w
a
l.
His
c
u
rre
n
t
re
se
a
rc
h
in
tere
st
in
c
lu
d
e
s
c
o
m
p
u
ter
g
r
a
p
h
ics
,
a
rti
f
icia
l
n
e
u
ra
l
n
e
tw
o
rk
,
g
e
o
g
ra
p
h
ica
l
in
f
o
rm
a
ti
o
n
sy
ste
m
s
,
a
lg
o
rit
h
m
s
a
n
d
l
o
g
ic
d
e
sig
n
.
He
h
a
d
a
b
o
u
t
1
0
y
e
a
rs
o
f
e
x
p
e
rien
c
e
a
s
f
a
c
u
lt
y
in
c
lu
d
in
g
in
d
u
strial
e
x
p
o
su
re
.
He
h
a
d
se
v
e
ra
l
p
u
b
li
c
a
ti
o
n
s
in
n
a
ti
o
n
a
l
a
n
d
i
n
tern
a
ti
o
n
a
l
jo
u
r
n
a
ls/co
n
f
e
re
n
c
e
s.
Ak
a
n
k
sh
a
Ch
a
n
d
o
la
is
c
u
rre
n
t
ly
w
o
rk
in
g
a
s
S
c
ien
ti
st
a
t
Uttara
k
h
a
n
d
S
tate
Co
u
n
c
il
f
o
r
S
c
ien
c
e
a
n
d
T
e
c
h
n
o
lo
g
y
.
S
h
e
is
M
a
st
e
rs
in
C
o
m
p
u
ter
A
p
p
li
c
a
ti
o
n
w
it
h
h
o
n
o
rs
f
ro
m
Birl
a
In
stit
u
te
o
f
A
p
p
li
e
d
S
c
ien
c
e
s,
a
ff
il
iate
d
to
Ku
m
a
u
n
Un
iv
e
rsit
y
.
S
h
e
h
a
d
g
ra
d
u
a
ti
o
n
d
e
g
re
e
in
S
c
ien
c
e
f
ro
m
He
m
wa
ti
Na
n
d
a
n
Ba
h
u
g
u
n
a
G
a
rh
wa
l
Un
iv
e
rsit
y
.
S
h
e
h
a
d
7
y
e
a
rs
o
f
w
o
r
k
in
g
e
x
p
e
rien
c
e
a
s
a
n
a
c
a
d
e
m
i
c
ian
a
t
p
rim
e
in
stit
u
tes
o
f
Uttara
k
h
a
n
d
.
S
h
e
h
a
d
th
e
v
a
rio
u
s
p
u
b
li
c
a
ti
o
n
s
a
t
n
a
ti
o
n
a
l
a
n
d
in
ter
n
a
ti
o
n
a
l
c
o
n
f
e
re
n
c
e
s
a
n
d
jo
u
rn
a
l.
S
h
e
h
a
d
b
e
e
n
a
re
c
ip
ien
t
o
f
sc
h
o
lars
h
ip
s
f
ro
m
sta
te
g
o
v
t
a
t
h
e
r
p
o
st
g
ra
d
u
a
ti
o
n
lev
e
l
a
n
d
g
o
v
t
o
f
In
d
ia
f
e
ll
o
w
sh
ip
in
In
tellec
tu
a
l
P
ro
p
e
rty
Rig
h
ts.
He
r
su
b
jec
ts
o
f
in
tere
st
a
re
c
lo
u
d
c
o
m
p
u
ti
n
g
,
a
rti
f
icia
l
in
telli
g
e
n
c
e
,
w
e
b
tec
h
n
o
lo
g
y
a
n
d
in
tellec
tu
a
l
p
ro
p
e
rty
rig
h
ts
(m
a
in
l
y
p
a
ten
t
a
n
d
g
e
o
g
ra
p
h
ica
l
in
d
ica
ti
o
n
s).
Dr.
M
.
P
.
T
h
a
p
li
y
a
l
is
w
o
rk
in
g
a
s
P
ro
f
e
ss
o
r
in
t
h
e
De
p
a
rtm
e
n
t
o
f
C
o
m
p
u
ter
S
c
ien
c
e
&
E
n
g
in
e
e
rin
g
,
S
c
h
o
o
l
o
f
En
g
in
e
e
rin
g
&
T
e
c
h
n
o
lo
g
y
,
HN
B
G
a
rh
w
a
l
Un
iv
e
rsit
y
(A
Ce
n
tral
Un
iv
e
rsit
y
),
S
rin
a
g
a
r
(G
a
rh
w
a
l)
Uttara
k
h
a
n
d
,
In
d
ia.
H
e
h
a
s
p
u
b
li
sh
e
d
m
o
re
th
a
n
5
0
re
se
a
rc
h
p
a
p
e
rs
in
th
e
In
tern
a
ti
o
n
a
l
Jo
u
rn
a
ls/Na
ti
o
n
a
l
jo
u
rn
a
ls/C
o
n
f
e
re
n
c
e
s.
He
h
a
s
d
e
li
v
e
r
e
d
m
o
re
th
a
n
3
0
talk
s
a
t
n
a
ti
o
n
a
l
lev
e
l
a
n
d
c
h
a
ired
m
a
n
y
T
e
c
h
n
ica
l
se
ss
io
n
s
a
t
I
n
tern
a
ti
o
n
a
l/
Na
ti
o
n
a
l/
S
y
m
p
o
siu
m
/
W
o
rk
sh
o
p
.
His
m
a
jo
r
re
se
a
rc
h
in
tere
sts
a
re
in
t
h
e
f
ield
o
f
S
o
f
t
w
a
re
En
g
in
e
e
rin
g
,
Hu
m
a
n
-
Co
m
p
u
ter
In
tera
c
ti
o
n
,
E
-
L
e
a
rn
in
g
,
e
d
u
c
a
ti
o
n
a
l
re
se
a
rc
h
a
n
d
th
e
ro
l
e
o
f
In
f
o
r
m
a
ti
o
n
a
n
d
c
o
m
m
u
n
ic
a
ti
o
n
tec
h
n
o
l
o
g
ies
f
o
r
im
p
ro
v
in
g
tea
c
h
in
g
a
n
d
lea
rn
in
g
p
ro
c
e
ss
.
Dr.
T
h
a
p
li
y
a
l
h
a
d
v
isit
e
d
Ch
in
a
,
US
A
,
Itl
a
y
,
F
ra
n
c
e
a
n
d
S
in
g
a
p
o
re
a
s
a
n
e
x
p
e
rt
to
p
re
se
n
t
h
is
p
a
p
e
rs
a
n
d
is
Ed
it
o
rial
Bo
a
rd
M
e
m
b
e
r
o
f
v
a
rio
u
s
In
tern
a
ti
o
n
a
l
Jo
u
rn
a
ls
a
n
d
Ex
p
e
rts
o
f
v
a
rio
u
s In
d
ian
a
n
d
F
o
r
e
ig
n
Un
iv
e
rsiti
e
s.
Evaluation Warning : The document was created with Spire.PDF for Python.