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.
10
,
No
.
3
,
Sep
tem
b
er
202
1
,
p
p
.
6
1
4
~
6
2
2
I
SS
N:
2252
-
8
9
3
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijai.v
10
.i
3
.
p
p
6
1
4
-
6
2
2
614
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
i
.
ia
esco
r
e.
co
m
Fish surv
iv
a
l pre
diction in
an a
qu
a
tic
env
iro
nmen
t
using
ra
ndo
m forest mo
del
M
d.
M
o
n
irul I
s
la
m
1
,
M
o
ha
m
m
o
d Abul K
a
s
hem
2
,
J
ia
Uddi
n
3
1
,2
De
p
a
rtme
n
t
o
f
C
o
m
p
u
ter S
c
ien
c
e
a
n
d
En
g
in
e
e
ri
n
g
,
Dh
a
k
a
Un
iv
e
rsity
o
f
En
g
i
n
e
e
rin
g
a
n
d
Tec
h
n
o
l
o
g
y
,
G
a
z
ip
u
r,
Ba
n
g
lad
e
sh
3
Tec
h
n
o
l
o
g
y
S
t
u
d
ies
De
p
a
rtme
n
t,
En
d
ico
tt
C
o
ll
e
g
e
,
Wo
o
so
n
g
U
n
iv
e
rsity
,
Da
e
jeo
n
,
S
o
u
th
Ko
re
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Oct
3
1
,
2
0
2
0
R
ev
is
ed
May
5
,
2
0
2
1
Acc
ep
ted
May
20
,
2
0
2
1
In
th
e
re
a
l
wo
rld
,
it
is
v
e
ry
d
iffi
c
u
lt
fo
r
fish
fa
rm
e
rs
to
se
lec
t
th
e
p
e
rfe
c
t
fish
sp
e
c
ies
fo
r
a
q
u
a
c
u
lt
u
re
in
a
sp
e
c
ifi
c
a
q
u
a
ti
c
e
n
v
ir
o
n
m
e
n
t.
T
h
e
m
a
in
g
o
a
l
o
f
th
is
re
se
a
rc
h
is
to
b
u
il
d
a
m
a
c
h
i
n
e
lea
rn
i
n
g
th
a
t
c
a
n
p
re
d
ict
th
e
p
e
rfe
c
t
fis
h
sp
e
c
ies
in
a
n
a
q
u
a
ti
c
e
n
v
ir
o
n
m
e
n
t.
I
n
t
h
is
p
a
p
e
r,
we
h
a
v
e
u
ti
li
z
e
d
a
m
o
d
e
l
u
sin
g
ra
n
d
o
m
fo
re
st
(RF
)
.
T
o
v
a
li
d
a
te
th
e
m
o
d
e
l,
we
h
a
v
e
u
se
d
a
d
a
tas
e
t
o
f
a
q
u
a
ti
c
e
n
v
ir
o
n
m
e
n
t
fo
r
1
1
d
iffe
re
n
t
fish
e
s.
To
p
re
d
ict
th
e
fish
s
p
e
c
ies
,
we
u
ti
li
z
e
d
t
h
e
d
iffere
n
t
c
h
a
ra
c
teristics
o
f
a
q
u
a
ti
c
e
n
v
ir
o
n
m
e
n
t
i
n
c
lu
d
i
n
g
p
H,
tem
p
e
ra
tu
re
,
a
n
d
tu
r
b
id
it
y
.
As
a
p
e
rfo
rm
a
n
c
e
m
e
tri
c
s,
we
m
e
a
su
re
d
a
c
c
u
ra
c
y
,
tru
e
p
o
si
ti
v
e
(
TP
)
ra
te,
a
n
d
k
a
p
p
a
sta
ti
stics
.
Ex
p
e
rime
n
tal
re
su
lt
s
d
e
m
o
n
stra
te
th
a
t
t
h
e
p
ro
p
o
se
d
RF
-
b
a
se
d
p
re
d
i
c
ti
o
n
m
o
d
e
l
sh
o
ws
a
c
c
u
ra
c
y
8
8
.
4
8
%
,
k
a
p
p
a
sta
ti
stic
8
7
.
1
1
%
a
n
d
TP
ra
te
8
8
.
5
%
fo
r
th
e
tes
ted
d
a
tas
e
t.
I
n
a
d
d
it
i
o
n
,
we
c
o
m
p
a
re
th
e
p
ro
p
o
se
d
m
o
d
e
l
wit
h
th
e
sta
te
-
of
-
a
rt
m
o
d
e
ls
J4
8
,
RF
,
k
-
n
e
a
re
st
n
e
ig
h
b
o
r
(
k
-
NN
)
,
a
n
d
c
las
sifica
ti
o
n
a
n
d
re
g
re
ss
io
n
tree
s
(CART).
Th
e
p
ro
p
o
se
d
m
o
d
e
l
o
u
tp
e
rf
o
rm
s
th
a
n
th
e
e
x
isti
n
g
m
o
d
e
ls
b
y
e
x
h
ib
it
in
g
th
e
h
i
g
h
e
r
a
c
c
u
ra
c
y
sc
o
re
,
TP
ra
te an
d
k
a
p
p
a
sta
ti
stics
.
K
ey
w
o
r
d
s
:
Acc
u
r
ac
y
p
r
ed
ictio
n
Aq
u
ac
u
ltu
r
e
Fis
h
s
u
r
v
iv
al
R
an
d
o
m
f
o
r
est m
o
d
el
Su
p
er
v
is
ed
m
ac
h
i
n
e
lear
n
in
g
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r
th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
J
ia
Ud
d
in
T
ec
h
n
o
lo
g
y
Stu
d
ies De
p
ar
tm
e
n
t
E
n
d
ico
tt C
o
lleg
e,
W
o
o
s
o
n
g
Un
iv
er
s
ity
So
u
th
Ko
r
ea
E
m
ail:
jia.
u
d
d
in
@
wsu
.
ac
.
k
r
1.
I
NT
RO
D
UCT
I
O
N
Aq
u
ac
u
ltu
r
e
r
e
f
er
s
to
th
e
f
ar
m
in
g
o
f
aq
u
atic
an
im
als
o
r
p
lan
ts
p
r
im
ar
ily
f
o
r
f
o
o
d
.
I
t
c
o
n
tain
s
th
e
b
r
ee
d
in
g
,
n
u
r
tu
r
e
,
an
d
r
ea
p
in
g
o
f
f
is
h
,
m
o
llu
s
k
s
,
cr
u
s
tace
an
s
,
an
d
p
lan
ts
in
f
r
esh
an
d
s
altwate
r
en
v
ir
o
n
m
e
n
ts
.
T
h
e
p
r
ac
tice
was
in
itiated
in
C
h
in
a
ab
o
u
t
4
,
0
0
0
y
ea
r
s
ag
o
a
n
d
g
lo
b
al
p
r
o
d
u
ctio
n
r
e
m
ain
s
to
b
e
s
u
b
ju
g
ated
b
y
C
h
in
a
an
d
o
th
er
Asi
an
co
u
n
tr
ies.
Aq
u
ac
u
ltu
r
e
is
u
s
ed
to
h
ar
v
est
f
o
o
d
b
y
s
o
m
e
o
f
th
e
d
ep
r
iv
ed
co
m
m
u
n
ities
ev
er
y
wh
er
e
o
n
th
e
g
lo
b
e
as
we
ll
as
b
y
k
ey
co
r
p
o
r
atio
n
s
.
Glo
b
ally
,
aq
u
ac
u
ltu
r
e
b
y
n
o
w
s
u
p
p
lies
m
o
r
e
th
an
h
alf
o
f
all
s
ea
f
o
o
d
u
s
ed
u
p
b
y
h
u
m
a
n
s
,
a
p
r
o
p
o
r
tio
n
th
at
co
n
tin
u
es
to
r
is
e
as
th
e
w
o
r
ld
p
o
p
u
latio
n
p
r
o
d
u
ce
s
.
Acc
o
r
d
in
g
to
th
e
Fo
o
d
an
d
A
g
r
icu
ltu
r
al
O
r
g
an
izatio
n
(
FA
O)
[
1
]
,
3
m
illi
o
n
to
n
s
o
f
f
o
o
d
wer
e
p
r
o
d
u
ce
d
b
y
aq
u
ac
u
ltu
r
e
i
n
th
e
1
9
7
0
s
,
a
f
ig
u
r
e
th
at
r
o
s
e
s
tead
ily
to
o
v
er
8
0
m
illi
o
n
to
n
s
in
2
0
1
7
.
Ma
n
u
all
y
f
is
h
class
i
f
i
ca
t
io
n
i
s
a
v
e
r
y
co
m
p
le
x
a
n
d
te
d
i
o
u
s
ass
i
g
n
m
e
n
t
f
o
r
t
h
ese
w
h
o
a
r
e
n
o
w
n
o
t
s
p
e
cia
lis
ts
.
F
is
h
s
p
e
cies
a
r
e
c
o
n
ce
r
n
e
d
i
n
m
a
n
y
in
d
u
s
t
r
i
al
an
d
a
g
r
ic
u
l
tu
r
a
l
in
d
u
s
t
r
ies
,
as
n
i
ce
l
y
as
t
h
e
m
a
n
u
f
ac
t
u
r
e
o
f
f
o
o
d
s
t
u
f
f
s
an
d
u
s
e
d
as
f
o
o
d
t
h
at
is
v
e
r
y
v
ita
l
t
o
h
u
m
a
n
s
[
2
]
.
As
m
a
r
i
n
e
b
i
o
lo
g
is
ts
class
i
f
y
f
is
h
f
r
o
m
t
h
ei
r
t
r
a
its
a
n
d
a
ls
o
u
s
e
d
th
e
cl
ass
if
ic
ati
o
n
t
r
ee
in
th
e
c
l
ass
i
f
ica
ti
o
n
o
f
f
is
h
,
w
h
i
c
h
le
d
th
e
m
t
o
u
s
e
l
a
p
t
o
p
s
g
ai
n
i
n
g
k
n
o
wl
e
d
g
e
o
f
a
n
d
s
t
r
u
ctu
r
es
i
n
t
h
e
d
a
ta
,
w
h
i
c
h
s
av
e
d
ti
m
e
,
ef
f
o
r
t,
a
n
d
v
el
o
ci
ty
i
n
th
e
class
i
f
i
ca
ti
o
n
o
f
f
is
h
[
3
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
F
is
h
s
u
r
viva
l p
r
ed
ictio
n
in
a
n
a
q
u
a
tic
e
n
viro
n
men
t u
s
in
g
r
a
n
d
o
m
fo
r
est mo
d
el
(
Md
.
Mo
n
ir
u
l I
s
la
m
)
615
Fis
h
cl
ass
i
f
ic
ati
o
n
ca
n
b
e
th
e
i
d
e
n
ti
f
i
ca
t
io
n
o
f
f
is
h
s
p
e
cies
,
d
e
p
en
d
i
n
g
o
n
th
ei
r
p
h
y
s
i
o
g
n
o
m
i
es
o
r
s
im
il
ar
iti
es.
Als
o
,
it
ca
n
b
e
d
e
s
cr
i
b
ed
as
th
e
t
ec
h
n
i
q
u
e
o
f
d
e
ter
m
i
n
i
n
g
t
h
e
t
y
p
es
o
f
f
is
h
[
4
]
.
C
lass
i
f
i
ca
t
io
n
o
f
f
is
h
is
cr
iti
ca
l
f
o
r
n
u
m
e
r
o
u
s
r
ea
s
o
n
s
,
i
n
c
lu
s
iv
e
o
f
s
a
m
p
l
e
an
d
s
u
b
s
is
te
n
c
e
m
a
tc
h
i
n
g
ex
t
r
ac
ti
o
n
f
ea
t
u
r
e
,
id
e
n
t
if
ica
ti
o
n
o
f
p
h
y
s
i
ca
l
o
r
b
e
h
a
v
i
o
r
al
ch
ar
ac
te
r
is
ti
cs,
s
t
atis
ti
ca
l
c
o
n
t
r
o
l
a
n
d
h
i
g
h
-
q
u
alit
y
u
t
iliz
ed
t
o
f
is
h
o
f
a
ll
k
i
n
d
s
[
5
]
.
M
o
r
e
o
v
e
r
,
f
is
h
cl
as
s
if
i
ca
t
io
n
is
r
e
g
a
r
d
e
d
as
a
v
it
a
l
v
e
n
t
u
r
e
f
o
r
f
is
h
i
n
g
a
n
d
p
o
p
u
lati
o
n
ass
ess
m
e
n
ts
[
6
]
.
On
th
e
o
th
e
r
h
a
n
d
,
c
o
m
p
u
ter
ized
f
is
h
class
if
ic
atio
n
ca
n
s
p
e
ed
u
p
th
e
tech
n
iq
u
e
a
n
d
ca
n
i
m
p
r
o
v
e
th
e
ac
cu
r
ac
y
o
f
class
if
icatio
n
o
r
i
d
en
tific
atio
n
o
f
f
is
h
s
p
ec
ies.
Sev
er
al
tactics
ar
e
in
tr
o
d
u
ce
d
in
th
e
liter
atu
r
e
f
o
r
co
m
p
u
ter
ized
f
is
h
s
p
ec
ies
id
en
tific
atio
n
.
I
n
th
is
p
ap
er
,
we
d
id
class
if
icatio
n
u
s
in
g
m
ac
h
in
e
lear
n
in
g
m
o
d
el
in
clu
d
in
g
d
ec
is
io
n
tr
ee
class
if
ier
(
J
4
8
)
,
r
a
n
d
o
m
f
o
r
est
(
R
F),
k
-
n
ea
r
est
n
ei
g
h
b
o
r
(
k
-
NN
)
,
a
n
d
class
if
icatio
n
an
d
r
eg
r
ess
io
n
tr
ee
(
C
AR
T
)
.
C
las
s
if
icatio
n
h
as u
s
ed
f
o
r
p
r
ed
ictio
n
p
u
r
p
o
s
es; tr
ad
itio
n
al
r
u
le
-
b
a
s
ed
alg
o
r
ith
m
d
o
es
n
o
t
p
r
o
v
id
e
a
n
y
p
r
ed
ictio
n
f
ea
tu
r
e
f
o
r
th
e
u
n
k
n
o
wn
d
ataset.
C
o
n
f
u
s
io
n
m
atr
ix
p
r
o
v
id
es
v
a
r
io
u
s
m
ea
s
u
r
em
en
t
o
f
ac
cu
r
ac
y
in
p
r
e
d
ictio
n
,
w
h
e
r
e
r
u
le
-
b
ased
alg
o
r
ith
m
ca
n
n
o
t
p
er
f
o
r
m
th
is
[
7
]
.
C
NN
is
a
d
ee
p
lear
n
in
g
m
o
d
el
wh
er
e
co
m
p
u
tatio
n
c
o
m
p
le
x
ity
is
h
ig
h
e
r
th
an
m
ac
h
in
e
lea
r
n
in
g
m
o
d
els.
I
n
th
is
p
ap
er
,
we
h
av
e
co
n
s
id
er
ed
t
h
e
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
s
o
n
ly
d
u
e
t
o
its
less
co
m
p
u
ta
tio
n
al
co
m
p
lex
ity
.
I
n
t
h
e
C
NN,
we
n
ee
d
m
u
c
h
tr
ain
in
g
tim
e
th
a
n
tr
ad
itio
n
al
m
ac
h
in
e
lear
n
in
g
m
o
d
els.
I
n
th
is
p
ap
er
,
we
p
r
o
p
o
s
ed
a
f
is
h
s
u
r
v
iv
al
p
r
ed
i
ctio
n
i
n
a
n
aq
u
atic
en
v
ir
o
n
m
e
n
t
b
ased
o
n
th
e
RF
m
o
d
el.
Fo
r
th
e
r
est
o
f
t
h
e
p
ap
er
,
we
o
r
g
an
ize
as
s
h
o
w
n
in
s
ec
tio
n
2
s
tates
th
e
lit
er
atu
r
e
r
ev
iew.
I
n
s
ec
tio
n
3
,
th
e
p
r
o
p
o
s
ed
m
o
d
el
is
d
is
cu
s
s
ed
.
Sectio
n
4
d
ep
icts
th
e
ex
p
er
im
en
tal
s
etu
p
an
d
r
esu
lt
f
r
o
m
th
e
a
n
aly
s
is
.
Fin
ally
,
th
e
f
in
d
in
g
s
o
f
th
is
p
ap
er
ar
e
d
is
cu
s
s
ed
in
s
ec
tio
n
5
.
2.
L
I
T
E
R
AT
U
RE
R
E
VI
E
W
T
h
e
liter
atu
r
e
s
tates
a
p
o
r
tio
n
o
f
ac
tiv
ities
r
elate
d
to
d
ec
is
io
n
s
u
p
p
o
r
t
s
y
s
tem
s
in
aq
u
ac
u
lt
u
r
e
g
ar
d
en
o
p
er
atio
n
s
.
Sev
er
al
d
ec
is
io
n
s
u
p
p
o
r
t
s
y
s
tem
s
h
av
e
b
ee
n
d
ev
elo
p
ed
.
So
m
e
o
f
th
em
u
s
e
m
ac
h
in
e
lear
n
in
g
m
eth
o
d
s
a
n
d
o
th
er
s
d
o
n
o
t.
A
n
au
t
o
m
atic
f
is
h
id
en
tific
atio
n
is
p
r
o
p
o
s
ed
wh
er
e
s
h
ad
e
an
d
tex
tu
r
e
f
ea
tu
r
es
ar
e
ex
tr
ac
ted
f
r
o
m
th
e
f
is
h
im
ag
e
s
[
8
]
.
A
s
tr
u
ctu
r
e
is
in
tr
o
d
u
ce
d
u
s
in
g
th
e
r
ea
l
-
tim
e
wate
r
q
u
ality
in
d
ica
to
r
s
an
d
o
p
er
atio
n
al
in
f
o
r
m
atio
n
,
wh
e
r
e
im
p
ac
t
o
n
s
u
r
v
iv
al
r
ate,
b
io
m
ass
,
an
d
p
r
o
d
u
ctio
n
f
ail
u
r
e
o
f
aq
u
ac
u
ltu
r
e
s
p
ec
ies
ar
e
ev
alu
ated
[
9
]
.
A
p
r
ed
ictio
n
m
o
d
el
u
s
in
g
o
n
e
f
ea
tu
r
e
o
f
wate
r
ca
lled
DO
is
p
r
esen
ted
f
o
r
t
h
e
aq
u
atic
cr
ea
tu
r
e
[
1
0
]
.
A
h
ar
d
war
e
is
m
ad
e
f
o
r
m
o
n
ito
r
in
g
wate
r
q
u
ality
f
ac
to
r
s
in
clu
d
i
n
g
p
H,
tem
p
er
atu
r
e,
an
d
d
is
s
o
lv
ed
o
x
y
g
en
[
1
1
]
.
An
I
o
T
d
ev
ice
is
p
r
o
p
o
s
ed
f
o
r
d
etec
tin
g
an
d
co
n
tr
o
llin
g
th
e
wate
r
f
ac
to
r
s
in
clu
d
in
g
p
H,
tem
p
er
atu
r
e;
h
o
wev
er
,
t
h
ey
d
id
n
o
t
an
aly
ze
th
e
d
ata
[
1
2
]
.
A
r
eg
r
ess
io
n
m
o
d
e
l
is
u
tili
ze
d
f
o
r
p
r
ed
ictin
g
wate
r
q
u
ality
o
f
c
u
ltiv
atin
g
f
is
h
;
h
o
wev
er
,
t
h
ey
d
id
n
o
t
co
n
s
id
er
t
h
e
p
r
e
d
ictio
n
ac
cu
r
ac
y
[
1
3
]
.
A
n
au
to
m
ated
s
tr
ateg
y
is
d
ev
elo
p
ed
f
o
r
f
is
h
id
en
tific
atio
n
p
r
im
ar
ily
b
ased
o
n
th
e
u
s
e
o
f
aid
v
ec
to
r
d
esk
to
p
an
d
k
-
m
ea
n
s
clu
s
ter
in
g
alg
o
r
ith
m
[
1
4
]
.
A
co
m
p
u
ter
ized
r
o
b
u
s
t
Nile
-
T
ilap
ia
f
is
h
class
if
icatio
n
ap
p
r
o
ac
h
is
p
r
o
p
o
s
ed
in
[
1
5
]
,
w
h
er
e
t
h
e
s
ca
le
-
in
v
a
r
ian
t
ch
ar
ac
ter
is
tics
o
f
f
is
h
’
s
ch
an
g
e
ar
e
ex
tr
ac
te
d
.
T
h
en
,
th
ese
p
o
in
ts
a
r
e
u
s
ed
t
o
f
ee
d
th
e
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
i
n
e.
Ma
n
ag
in
g
h
atch
er
y
p
r
o
d
u
ctio
n
is
f
o
c
u
s
ed
u
s
in
g
r
u
les
a
n
d
ca
lcu
latio
n
s
o
f
p
h
y
s
ical,
c
h
e
m
ical,
an
d
b
io
lo
g
ical
p
r
o
ce
s
s
es
[
1
6
]
.
A
s
cien
tific
m
o
d
el
is
d
ev
elo
p
ed
t
o
ev
alu
ate
en
v
ir
o
n
m
en
tal
im
p
ac
t
[
1
7
]
.
A
r
u
le
is
h
an
d
-
c
r
af
ted
b
y
d
o
m
ain
ex
p
e
r
ts
[
1
8
]
.
A
m
ac
h
in
e
lear
n
in
g
m
eth
o
d
is
p
r
e
s
en
ted
to
o
b
tain
a
b
alan
ce
b
etwe
en
th
e
f
ar
m
cl
o
s
u
r
e
an
d
th
e
f
ar
m
o
p
en
in
g
ev
e
n
ts
[
1
9
]
.
A
f
ea
tu
r
e
r
an
k
i
n
g
alg
o
r
ith
m
is
d
is
p
lay
ed
to
i
d
en
tify
th
e
m
o
s
t
in
f
lu
en
tial
ca
u
s
e
o
f
th
e
clo
s
u
r
e
[
2
0
]
.
T
im
e
s
er
ies
m
ac
h
in
e
lear
n
in
g
ap
p
r
o
ac
h
es
is
ad
o
p
ted
lik
e
p
r
in
cip
al
co
m
p
o
n
en
t
an
aly
s
is
(
PC
A)
a
n
d
au
to
co
r
r
elatio
n
f
u
n
ctio
n
(
AC
F)
to
p
r
ed
ict
th
e
clo
s
u
r
e
e
v
en
t
[
2
1
]
.
A
s
et
o
f
r
u
les
ar
e
ex
tr
ac
ted
f
r
o
m
d
at
a
g
ath
er
ed
b
y
s
en
s
o
r
n
etwo
r
k
s
to
f
in
d
ass
o
ciatio
n
s
b
etwe
en
en
v
ir
o
n
m
e
n
tal
v
ar
iab
les
an
d
alg
ae
g
r
o
wth
[
2
2
]
.
A
n
en
s
em
b
le
m
eth
o
d
i
s
d
esig
n
ed
to
f
in
d
th
e
r
ele
v
an
t
en
v
ir
o
n
m
en
tal
v
ar
iab
les
r
esp
o
n
s
ib
le
f
o
r
alg
ae
g
r
o
wth
an
d
th
e
g
r
o
wth
p
r
ed
ictio
n
[
2
3
]
.
A
m
ac
h
in
e
l
ea
r
n
in
g
m
eth
o
d
is
d
ev
elo
p
e
d
to
p
r
ed
ict
th
e
p
r
o
p
a
g
atio
n
o
f
alg
ae
p
atch
es a
lo
n
g
t
h
e
wate
r
way
[
2
4
]
.
3.
P
RO
P
O
SE
D
M
O
D
E
L
Fig
u
r
e
1
s
h
o
ws
a
d
etailed
b
lo
ck
d
iag
r
am
o
f
th
e
p
r
o
p
o
s
ed
m
o
d
el.
At
f
ir
s
t,
we
im
p
o
r
t
o
u
r
d
ataset.
I
n
th
e
p
r
e
p
r
o
ce
s
s
in
g
s
ec
tio
n
,
we
f
ilter
an
d
r
esam
p
le
f
o
r
o
u
r
d
at
aset.
T
h
en
we
s
elec
t
o
u
r
m
o
d
el
as
RF
class
if
ier
s
in
th
e
class
if
icatio
n
s
ec
tio
n
.
W
e
class
if
y
o
u
r
v
a
r
io
u
s
m
ac
h
in
e
lear
n
i
n
g
m
o
d
els
h
e
r
e.
Af
ter
class
if
icatio
n
,
class
if
ier
o
u
tp
u
t is p
r
ed
icted
.
3
.
1
.
Descript
io
n o
f
da
t
a
s
et
T
h
e
d
ata
u
s
ed
in
th
is
s
tu
d
y
in
v
o
lv
in
g
p
ar
a
m
eter
s
o
f
an
a
q
u
atic
en
v
ir
o
n
m
en
t
f
o
r
f
is
h
f
a
r
m
in
g
tak
e
n
f
r
o
m
th
e
Un
iv
er
s
ity
o
f
Dh
a
k
a,
Facu
lt
y
o
f
Fis
h
er
ies
,
Dh
ak
a,
B
an
g
lad
esh
.
T
h
er
e
ar
e
1
9
1
in
s
tan
ce
s
o
f
4
attr
ib
u
tes.
Attr
ib
u
tes
ar
e
p
H,
t
em
p
er
atu
r
e,
tu
r
b
i
d
ity
,
an
d
f
is
h
.
W
e
ch
o
o
s
e
p
H,
tem
p
er
atu
r
e,
tu
r
b
id
ity
as
f
ea
t
u
r
e
attr
ib
u
tes
an
d
f
is
h
as
tar
g
et
a
ttrib
u
te.
T
h
e
d
ataset
is
p
ar
titi
o
n
ed
in
t
o
two
p
a
r
ts
.
On
e
is
a
q
u
atic
en
v
ir
o
n
m
en
t
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
10
,
N
o
.
3
,
Sep
tem
b
er
202
1
:
6
1
4
-
622
616
ch
ar
ac
ter
is
tics
an
d
an
o
th
er
is
f
is
h
s
p
ec
ies.
T
h
e
d
etailed
o
f
tar
g
et
attr
ib
u
te
is
o
f
1
1
f
is
h
s
p
e
cies
in
clu
d
in
g
k
atla
1
4
im
ag
es,
s
h
in
g
1
7
im
ag
es,
p
r
awn
1
4
im
ag
es,
r
u
i
1
9
im
a
g
es,
k
o
i
1
5
im
ag
es,
p
an
g
as
2
2
im
ag
es,
tilap
ia
2
5
im
ag
es,
s
ilv
er
ca
r
p
7
im
ag
es,
k
ar
p
io
3
3
im
ag
es,
m
ag
u
r
1
1
i
m
ag
es a
n
d
s
h
r
im
p
1
4
im
ag
es.
Aq
u
atic
en
v
ir
o
n
m
en
t
ch
ar
ac
ter
is
tics
:
W
e
u
tili
ze
d
p
H,
tem
p
er
atu
r
e,
an
d
tu
r
b
i
d
ity
a
s
aq
u
atic
en
v
ir
o
n
m
en
t p
a
r
am
eter
s
in
o
u
r
s
tu
d
y
.
−
p
H:
p
H
is
n
ec
ess
ar
y
f
o
r
aq
u
ac
u
ltu
r
e
as
a
m
ea
s
u
r
e
o
f
th
e
ac
i
d
ity
o
f
th
e
wate
r
o
r
s
o
il.
T
h
e
o
p
tim
al
p
H
f
o
r
f
is
h
is
b
etwe
en
6
.
5
an
d
9
.
F
is
h
will
g
r
o
w
p
o
o
r
ly
,
an
d
r
e
p
r
o
d
u
ctio
n
will
b
e
af
f
ec
ted
a
t
co
n
s
is
ten
tly
g
r
ea
ter
o
r
lo
wer
p
H
tier
s
[
2
5
]
.
T
h
e
p
H
lev
el
f
o
r
war
m
-
wate
r
p
o
n
d
f
is
h
is
4
f
o
r
ac
id
d
ea
th
p
o
in
t,
4
to
5
f
o
r
n
o
r
e
p
r
o
d
u
ctio
n
,
5
to
6
.
5
f
o
r
s
lo
w
g
r
o
wth
,
6
.
5
to
8
.
5
f
o
r
d
esira
b
le
r
an
g
es,
9
to
1
0
f
o
r
s
lo
w
g
r
o
wth
,
an
d
≥
1
1
f
o
r
alk
alin
e
d
ea
th
p
o
in
t.
−
T
em
p
er
atu
r
e:
T
h
e
in
cr
ea
s
e
a
n
d
en
d
ea
v
o
r
o
f
t
h
e
f
is
h
r
ely
o
n
th
eir
p
h
y
s
iq
u
e
tem
p
er
at
u
r
e.
T
h
e
b
o
d
y
tem
p
er
atu
r
e
o
f
th
e
f
is
h
is
ab
o
u
t th
e
s
am
e
as th
e
wate
r
tem
p
er
atu
r
e
an
d
v
a
r
ies with
it.
E
ac
h
f
is
h
s
p
ec
ies i
s
tailo
r
ed
to
d
ev
elo
p
an
d
r
e
p
r
o
d
u
ce
in
s
id
e
well
-
d
ef
i
n
ed
s
tag
es
o
f
wate
r
tem
p
er
at
u
r
es,
b
u
t
th
e
m
o
s
t
u
s
ef
u
l
b
o
o
m
an
d
r
e
p
lica
tak
e
ar
ea
with
in
n
a
r
r
o
wer
tier
s
o
f
tem
p
e
r
at
u
r
e.
I
t
is
im
p
o
r
tan
t,
th
er
ef
o
r
e,
to
u
n
d
er
s
tan
d
th
e
wate
r
tem
p
er
atu
r
es
r
ea
ch
ab
le
at
y
o
u
r
f
is
h
f
ar
m
n
icely
to
p
ick
o
u
t
th
e
r
i
g
h
t
s
p
ec
ies
o
f
f
is
h
an
d
t
o
g
r
ap
h
its
m
an
a
g
em
en
t a
s
a
r
es
u
lt.
T
ab
le
1
s
h
o
ws th
e
th
er
m
al
r
an
g
e
o
f
s
o
m
e
c
o
m
m
o
n
f
is
h
s
p
ec
ies [
2
6
]
.
−
T
u
r
b
id
ity
:
T
h
e
a
b
ilit
y
o
f
wate
r
to
tr
a
n
s
m
it
th
e
lig
h
t
th
at
r
estricts
lig
h
t
p
en
etr
a
tio
n
an
d
lim
it
p
h
o
to
s
y
n
t
h
esis
is
ter
m
ed
as
tu
r
b
id
ity
a
n
d
is
th
e
r
esu
ltan
t
im
p
ac
t
o
f
s
ev
er
al
elem
en
ts
s
u
c
h
as
s
u
s
p
en
d
ed
clay
p
ar
ticles,
d
is
p
er
s
io
n
o
f
p
l
an
k
to
n
o
r
g
an
is
m
s
,
p
ar
ticu
la
te
n
atu
r
al
th
in
g
s
an
d
also
th
e
p
ig
m
en
ts
ca
u
s
ed
with
th
e
aid
o
f
th
e
d
ec
o
m
p
o
s
itio
n
o
f
o
r
g
an
ic
m
atter
.
Acc
ep
tab
le
tu
r
b
id
ity
v
ar
ies
f
r
o
m
3
0
-
8
0
cm
is
p
r
o
p
er
l
y
f
o
r
f
is
h
h
ea
lth
[
2
7
]
.
−
Fis
h
s
p
ec
ies:
I
n
o
u
r
d
ataset,
w
e
u
tili
ze
d
a
to
tal
o
f
1
1
f
is
h
s
p
ec
ies
as
th
e
tar
g
et
v
a
r
iab
le.
T
h
e
f
is
h
s
p
ec
ies
in
o
u
r
d
ataset
ar
e
p
r
esen
ted
in
Fig
u
r
e
2
;
w
h
er
e
ca
r
p
io
f
is
h
is
s
h
o
wn
in
Fig
u
r
e
2
(
a)
,
k
atla
f
is
h
is
i
n
Fig
u
r
e
2
(
b
)
,
r
u
i
f
is
h
is
in
Fig
u
r
e
2
(
c)
,
k
o
i
f
is
h
is
in
Fig
u
r
e
2
(
d
)
,
m
ag
u
r
f
is
h
is
in
Fig
u
r
e
2
(
e
)
,
p
an
g
as
f
is
h
is
in
Fig
u
r
e
2
(
f
)
,
p
r
awn
f
is
h
is
in
Fig
u
r
e
2
(
g
)
,
s
ilv
er
ca
r
p
f
i
s
h
is
in
Fig
u
r
e
2
(
h
)
,
tilap
ia
f
is
h
is
in
Fig
u
r
e
2
(
i)
,
an
d
s
h
in
g
f
is
h
is
in
Fig
u
r
e
2
(
j)
.
Fig
u
r
e
1
.
B
lo
ck
d
iag
r
am
o
f
p
r
o
p
o
s
ed
m
o
d
el
T
ab
le
1
.
T
h
er
m
al
r
a
n
g
e
o
f
s
o
m
e
co
m
m
o
n
f
is
h
s
p
ec
ies (
in
○
C)
F
i
sh
sp
e
c
i
e
s
D
a
n
g
e
r
o
u
s
p
o
n
d
-
w
a
t
e
r
t
e
m
p
e
r
a
t
u
r
e
l
o
w
e
r
-
u
p
p
e
r
l
i
mi
t
O
p
t
i
mu
m
t
h
e
r
m
a
l
r
a
n
g
e
f
o
r
a
d
u
l
t
s
Th
e
r
m
a
l
r
a
n
g
e
f
o
r
sp
a
w
n
i
n
g
C
a
r
p
i
o
2
36
23
-
2
6
(
2
5
)
A
b
o
v
e
1
8
K
a
t
l
a
15
34
26
-
29
22
-
28
B
i
g
h
e
a
d
c
a
r
p
5
37
23
-
31
17
-
30
Fig
u
r
e
2
.
Sam
p
le
f
is
h
es: (
a)
ca
r
p
io
f
is
h
,
(
b
)
k
atla
f
is
h
,
(
c)
r
u
i
f
is
h
,
(
d
)
k
o
i f
is
h
,
(
e)
m
ag
u
r
f
is
h
,
(
f
)
p
an
g
as f
is
h
,
(
g
)
p
r
awn
f
is
h
,
(
h
)
s
ilv
er
ca
r
p
f
is
h
,
(
i)
tilap
ia
f
is
h
,
an
d
(
j)
s
h
i
n
g
f
is
h
(
a)
(
b
)
(
c)
(
d
)
(
e)
(f)
(
g
)
(
h
)
(
i)
(
j)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
F
is
h
s
u
r
viva
l p
r
ed
ictio
n
in
a
n
a
q
u
a
tic
e
n
viro
n
men
t u
s
in
g
r
a
n
d
o
m
fo
r
est mo
d
el
(
Md
.
Mo
n
ir
u
l I
s
la
m
)
617
3
.
2
.
P
re
pro
ce
s
s
ing
I
n
th
e
p
r
ep
r
o
ce
s
s
in
g
s
tep
,
we
f
ilter
ed
o
u
r
d
ataset
u
s
in
g
a
r
esa
m
p
lin
g
o
p
tio
n
f
o
r
o
b
s
er
v
in
g
th
e
cu
r
r
e
n
t
r
elatio
n
o
f
in
s
tan
ce
s
an
d
attr
ib
u
tes o
f
th
e
d
ataset.
I
n
t
h
e
attr
i
b
u
te
s
elec
tio
n
win
d
o
w,
we
ca
n
ch
ec
k
th
e
m
is
s
in
g
,
u
n
iq
u
e,
an
d
d
is
tin
ct
v
alu
e
o
f
ea
ch
attr
ib
u
te.
All
attr
ib
u
tes
s
h
o
w
0
%
m
is
s
in
g
an
d
p
H
h
as
2
8
u
n
iq
u
e
v
alu
es,
tem
p
er
atu
r
e
h
as 2
2
u
n
iq
u
e
v
al
u
es,
tu
r
b
id
ity
h
as 5
6
u
n
iq
u
e
an
d
f
is
h
h
as 1
1
d
is
tin
ct
v
alu
es.
3
.
3
.
Cla
s
s
if
ica
t
io
n
I
n
th
e
class
if
icatio
n
s
ec
tio
n
,
we
class
if
ied
o
u
r
d
ataset
u
s
in
g
5
v
ar
io
u
s
class
if
ier
s
m
o
d
el.
R
F
o
u
tp
er
f
o
r
m
s
th
e
o
th
er
d
escr
ib
e
d
m
o
d
el.
3
.
3
.
1
.
Ra
nd
o
m
f
o
re
s
t
RF
is
a
s
u
p
er
v
is
ed
lear
n
in
g
m
eth
o
d
th
at
is
a
d
ec
is
io
n
tr
ee
-
b
ased
alg
o
r
ith
m
.
As
th
e
n
am
e
p
r
o
p
o
s
es
as
f
o
r
est
th
e
RF
cla
s
s
if
ier
i
s
an
en
s
em
b
le
o
f
d
ec
is
io
n
tr
ee
s
wh
er
ev
er
a
r
an
d
o
m
v
ec
to
r
s
am
p
le
p
r
o
d
u
ce
ea
ch
class
if
ier
f
r
o
m
th
e
i
n
p
u
t
v
ec
to
r
[
2
8
]
an
d
ev
e
r
y
tr
ee
ca
s
t
a
u
n
it
v
o
te
f
o
r
t
h
e
m
o
s
t
p
o
p
u
la
r
class
to
class
if
y
an
in
p
u
t v
ec
to
r
,
n
ea
r
ly
all
o
f
t
h
e
t
im
e
tr
ain
ed
with
a
b
a
g
g
in
g
m
e
th
o
d
.
T
h
e
p
r
ep
a
r
atio
n
ca
lcu
latio
n
f
o
r
RF
ap
p
lies
th
e
o
v
er
all
s
tr
ateg
y
o
f
b
o
o
ts
tr
ap
c
o
llectin
g
,
o
r
p
ac
k
in
g
,
t
o
tr
ee
s
tu
d
en
ts
.
Giv
en
a
p
r
ep
a
r
a
tio
n
s
et
X
=
x
1
,
.
.
.
,
x
n
with
r
ea
ctio
n
s
Y
=
y
1
,
.
.
.
,
y
n
,
s
to
win
g
m
o
r
e
th
a
n
o
n
ce
(
A
tim
es)
ch
o
o
s
es
an
ir
r
e
g
u
lar
ex
am
p
le
with
s
u
b
s
titu
tio
n
o
f
t
h
e
p
r
e
p
ar
atio
n
s
et
an
d
f
its
tr
ee
s
to
th
ese
e
x
am
p
les.
Fo
r
a=
1
,
……,
A:
−
T
est,
with
s
u
b
s
titu
tio
n
,
n
p
r
e
p
ar
in
g
m
o
d
els f
r
o
m
X,
Y;
ca
ll th
ese
X
a
, Y
a
.
−
T
r
ain
a
ch
ar
ac
te
r
izatio
n
o
r
r
el
ap
s
e
tr
ee
f
a
o
n
X
a
, Y
a
.
Af
ter
p
r
e
p
ar
in
g
,
ex
p
ec
tatio
n
s
f
o
r
co
n
ce
aled
e
x
am
p
les
x
'
ca
n
b
e
m
ad
e
b
y
av
e
r
ag
in
g
th
e
f
o
r
ec
asts
f
r
o
m
all
th
e
i
n
d
iv
id
u
al
r
elap
s
e
tr
ee
s
o
n
x
'
:
̂
=
1
∑
(
x′
)
−
1
(
1
)
also
,
a
g
au
g
e
o
f
th
e
v
u
ln
e
r
ab
il
ity
o
f
th
e
f
o
r
ec
ast ca
n
b
e
m
ad
e
as th
e
s
tan
d
ar
d
d
e
v
iatio
n
o
f
th
e
ex
p
ec
tatio
n
s
f
r
o
m
all
th
e
i
n
d
iv
id
u
al
r
elap
s
e
tr
ee
s
o
n
x
'
:
=
√
∑
(
(
x′
)
−
̂
)
2
−
1
−
1
(
2
)
T
h
e
u
n
i
v
er
s
al
th
o
u
g
h
t
o
f
t
h
e
b
ag
g
in
g
m
eth
o
d
is
th
at
th
e
co
m
p
o
s
in
g
o
f
th
e
lear
n
in
g
m
eth
o
d
in
cr
ea
s
es
th
e
o
v
er
all
r
esu
lt.
T
h
e
RF
is
less
s
en
s
itiv
e
th
an
o
th
er
s
tr
ea
m
lin
e
m
ac
h
in
e
lear
n
in
g
class
i
f
ier
s
to
o
v
er
f
itti
n
g
an
d
to
th
e
q
u
ality
o
f
tr
ain
i
n
g
s
am
p
les
[
2
9
]
.
Fig
u
r
e
3
s
h
o
ws
th
e
co
n
ce
p
t
o
f
RF
m
o
d
el.
T
r
ee
1
an
d
T
r
e
e
2
b
el
o
n
g
to
C
lass
A
.
S
o
,
p
r
e
d
i
cte
d
o
u
t
p
u
t
will
b
e
C
lass
A
.
Ma
j
o
r
it
y
v
o
t
e
is
C
l
ass
A
i
n
Fi
g
u
r
e
3
.
Fig
u
r
e
3
.
RF
m
o
d
el
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
10
,
N
o
.
3
,
Sep
tem
b
er
202
1
:
6
1
4
-
622
618
3
.
4
.
Cla
s
s
if
ier
o
utput
I
n
th
e
class
if
ier
s
ec
tio
n
,
we
ca
n
s
ee
th
e
r
esu
lt
p
er
f
o
r
m
an
ce
o
f
o
u
r
m
o
d
el
an
d
o
th
er
s
tate
-
of
-
ar
t
m
o
d
els.
B
y
ch
o
o
s
in
g
o
u
r
d
esc
r
ib
ed
m
o
d
el,
we
ca
n
c
h
ec
k
r
esu
lts
.
I
n
th
is
s
ec
tio
n
,
we
ca
n
s
e
e
d
etailed
ac
cu
r
ac
y
b
y
class
.
Fig
u
r
e
4
s
h
o
ws
th
ese
p
er
f
o
r
m
an
ce
r
esu
lts
.
W
e
d
id
n
o
t
f
in
d
a
n
y
m
ac
h
i
n
e
lear
n
in
g
m
o
d
el
f
o
r
f
is
h
en
v
ir
o
n
m
en
t
m
o
n
ito
r
in
g
u
s
in
g
R
F.
T
h
e
d
ataset
we
h
av
e
u
s
ed
in
o
u
r
o
wn
d
ataset.
Fig
u
r
e
4
p
r
esen
ts
av
er
ag
e
tr
u
e
p
o
s
itiv
e
(
TP
)
r
ate
as
0
.
8
8
5
,
FP
r
ate
as
0
.
0
1
3
,
p
r
ec
is
io
n
as
0
.
8
9
0
,
r
ec
all
as
0
.
8
8
5
,
F
-
m
ea
s
u
r
e
as
0
.
8
7
9
,
MCC
as
0
.
8
7
1
,
R
OC
a
r
ea
as
0
.
9
8
1
,
PR
C
Ar
ea
as
0
.
9
2
9
,
C
o
r
r
ec
tly
C
lass
if
ied
I
n
s
tan
ce
s
as 8
8
.
4
8
%,
I
n
c
o
r
r
ec
tly
class
if
ied
in
s
tan
ce
s
as
1
1
.
5
2
%
,
Kap
p
a
s
tatis
tics
as
0
.
8
7
,
m
ea
n
ab
s
o
lu
te
e
r
r
o
r
as
0
.
0
4
,
r
o
o
t
m
ea
n
s
q
u
ar
e
d
er
r
o
r
as 0
.
1
3
,
r
elativ
e
ab
s
o
lu
te
er
r
o
r
as 2
4
.
5
3
%,
R
o
o
t
r
elativ
e
s
q
u
a
r
ed
er
r
o
r
as 4
5
.
4
6
%.
Fig
u
r
e
4
.
C
lass
if
ier
o
u
tp
u
t
o
f
o
u
r
m
o
d
el
4.
E
XP
E
R
I
M
E
N
T
A
L
SE
T
UP
AND
RE
SUL
T
ANA
L
YS
I
S
As
d
ata
an
aly
s
is
,
we
h
av
e
u
s
ed
W
E
KA
to
o
l
f
o
r
class
if
y
in
g
th
e
p
r
o
p
o
s
ed
m
o
d
el
a
n
d
d
escr
ib
ed
o
th
e
r
m
o
d
els.
T
h
e
to
o
l
is
v
er
y
h
elp
f
u
l
to
a
n
aly
ze
a
n
d
h
as
v
a
r
io
u
s
tech
n
iq
u
es
em
b
ed
d
e
d
in
it.
W
e
h
av
e
u
s
ed
1
0
%
im
ag
es f
o
r
test
in
g
an
d
9
0
% im
ag
es f
o
r
tr
ain
i
n
g
in
ea
c
h
s
p
ec
i
es f
o
r
all
d
escr
ib
ed
m
o
d
el.
4
.
1
.
P
er
f
o
r
m
a
nce
m
e
t
rics
Per
f
o
r
m
an
ce
p
ar
am
ete
r
s
ar
e
t
h
e
m
o
s
t
im
p
o
r
tan
t
m
et
r
ics
to
co
m
p
ar
e
a
m
o
n
g
class
if
ier
m
et
h
o
d
s
to
g
et
th
e
b
est
class
if
ier
.
W
e
h
av
e
ap
p
lied
3
p
er
f
o
r
m
a
n
ce
p
a
r
am
ete
r
s
wh
ich
ar
e
ac
cu
r
ac
y
,
tr
u
e
p
o
s
itiv
e
(
TP
)
r
ate
a
n
d
k
ap
p
a
s
tatis
tics
.
T
h
e
p
ar
am
et
er
is
ca
lcu
lated
f
r
o
m
a
co
n
f
u
s
io
n
m
atr
ix
wh
ich
is
s
itu
ate
d
in
ev
er
y
s
tep
o
f
class
if
icatio
n
.
Acc
u
r
ac
y
is
m
e
asu
r
ed
b
y
d
i
v
id
in
g
th
e
to
tal
n
u
m
b
er
o
f
c
o
r
r
ec
tly
class
if
ied
in
s
tan
ce
s
b
y
th
e
to
tal
n
u
m
b
er
o
f
in
s
tan
ce
s
an
d
also
i
t
is
m
ea
s
u
r
ed
b
y
co
n
f
u
s
io
n
m
atr
ix
wh
ich
is
m
ath
em
atica
lly
co
u
n
ted
b
y
(
4
)
.
T
P
r
ate
is
an
o
th
er
p
e
r
f
o
r
m
an
ce
m
etr
ic
o
f
o
u
r
s
tu
d
y
an
d
it
is
ca
lcu
lated
b
y
(
3
)
.
A
n
d
k
a
p
p
a
s
tatis
tic
i
s
th
e
last
m
etr
ic
o
f
o
u
r
p
ap
er
wh
ich
is
c
o
m
p
u
ted
b
y
(
5
)
.
T
h
e
h
ig
h
e
r
t
h
e
k
ap
p
a
s
tatis
tics
,
th
e
b
etter
th
e
m
o
d
el
ac
c
u
r
ac
y
lev
el.
A
g
en
er
al
v
iew
o
f
th
e
c
o
n
f
u
s
io
n
m
atr
ix
is
illu
s
tr
ated
in
T
ab
le
2
.
T
ab
le
2
.
C
o
n
f
u
s
io
n
m
atr
ix
P
r
e
d
i
c
t
e
d
Y
e
s
P
r
e
d
i
c
t
e
d
N
o
A
c
t
u
a
l
Y
e
s
TP
FN
A
c
t
u
a
l
N
o
FP
TN
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
F
is
h
s
u
r
viva
l p
r
ed
ictio
n
in
a
n
a
q
u
a
tic
e
n
viro
n
men
t u
s
in
g
r
a
n
d
o
m
fo
r
est mo
d
el
(
Md
.
Mo
n
ir
u
l I
s
la
m
)
619
Her
e,
T
P si
g
n
if
ies th
e
n
u
m
b
er
o
f
p
r
o
p
er
ly
class
if
ied
p
o
s
itiv
e
o
cc
u
r
r
e
n
ce
s
.
=
+
(
3
)
I
t is also
k
n
o
wn
as th
e
r
ec
all.
I
t te
lls
u
s
wh
at
p
er
ce
n
tag
e
o
f
p
o
s
itiv
e
in
s
tan
ce
s
h
av
e
b
ee
n
c
o
r
r
ec
tly
id
en
tifie
d
.
−
FP
s
ig
n
if
ies th
e
n
u
m
b
er
o
f
m
i
s
class
if
ied
p
o
s
itiv
e
o
cc
u
r
r
e
n
ce
s
.
−
FN sig
n
if
ies th
e
n
u
m
b
er
o
f
m
i
s
class
if
ied
n
eg
ativ
e
o
cc
u
r
r
en
c
es.
−
T
N
s
ig
n
if
ies th
e
n
u
m
b
e
r
o
f
p
r
o
p
er
ly
class
if
ied
n
eg
ativ
e
o
cc
u
r
r
en
ce
s
.
=
+
+
+
+
(
4
)
Acc
u
r
ac
y
is
also
r
ep
r
esen
ted
b
y
to
tal
ac
cu
r
ac
y
.
=
−
1
−
(
5
)
wh
er
e
=
(
+
)
×
(
+
)
+
(
+
)
×
(
+
)
(
+
+
+
)
×
(
+
+
+
)
(
6
)
W
e
h
av
e
u
s
ed
W
aik
ato
en
v
ir
o
n
m
en
t
f
o
r
k
n
o
wled
g
e
an
al
y
s
is
(
W
E
K
A)
f
o
r
p
r
o
ce
s
s
in
g
d
ata.
T
h
e
p
r
o
p
o
s
ed
m
o
d
el,
R
F
s
h
o
ws
th
e
ac
cu
r
ac
y
as
th
e
v
alu
e
8
8
.
4
8
1
7
%,
th
e
av
er
a
g
e
T
P
r
ate
as
t
h
e
weig
h
t
o
f
8
8
.
5
%
an
d
k
ap
p
a
s
tatis
tic
as
th
e
s
tan
d
ar
d
o
f
8
7
.
1
1
%.
W
e
ca
n
s
ay
,
th
ese
th
r
ee
m
etr
ics
g
iv
e
a
b
ett
er
r
esu
lt.
W
e
h
av
e
co
m
p
ar
ed
t
h
e
p
er
f
o
r
m
an
ce
m
e
tr
ics with
o
u
r
p
r
o
p
o
s
ed
m
o
d
el
an
d
o
th
er
s
tate
-
ar
t
-
m
o
d
els.
W
e
u
tili
ze
d
5
m
o
d
els
in
o
u
r
ex
p
er
im
en
tal
wo
r
k
.
T
h
ey
ar
e
RF
,
J
4
8
,
N
aïv
e
B
ay
es,
k
-
NN
,
an
d
C
AR
T
.
T
ab
le
3
d
ep
icted
a
d
etailed
co
m
p
ar
is
o
n
with
all
m
o
d
el
ea
c
h
o
th
er
.
T
ab
le
3
s
h
o
ws
th
at
R
F
g
iv
es
t
h
e
h
ig
h
est
s
co
r
e
o
f
ev
e
r
y
m
et
r
ic
as
ac
cu
r
ac
y
8
8
.
4
8
%,
k
ap
p
a
s
tatis
t
ic
a
s
8
7
.
1
1
%
,
an
d
T
P
r
ate
as
8
8
.
5
%
.
T
h
e
s
ec
o
n
d
h
i
g
h
est
s
co
r
e
b
el
o
n
g
s
to
t
h
e
k
-
NN
m
o
d
el
wh
ic
h
tells
ac
cu
r
ac
y
as
8
5
.
7
9
%,
k
ap
p
a
s
tatis
tic
as
8
4
.
0
5
%
an
d
T
P
r
ate
as
8
5
.
8
%.
J
4
8
ac
q
u
i
r
es
3
r
d
h
ig
h
est
p
o
s
iti
o
n
b
y
ac
h
iev
in
g
an
ac
cu
r
ac
y
as
7
3
.
1
6
%,
k
ap
p
a
s
tatis
t
ic
as
6
9
.
8
8
%
an
d
T
P
r
ate
as
7
3
.
2
%.
C
AR
T
h
as
4
th
p
lace
in
s
co
r
in
g
p
er
f
o
r
m
an
ce
m
etr
ics b
y
g
ettin
g
ac
cu
r
ac
y
as 6
4
.
2
1
%,
k
ap
p
a
s
tatis
tic
as 5
9
.
8
0
an
d
T
P r
ate
as
6
4
.
2
%.
Naïv
e
B
ay
es
(
NB
)
g
iv
es
th
e
lo
west
s
co
r
e
b
y
ac
q
u
ir
in
g
ac
cu
r
ac
y
as
5
6
.
8
4
%,
k
ap
p
a
s
tatis
t
ic
as
5
1
.
6
0
%
an
d
T
P
r
ate
as
5
6
.
8
8
%.
NB
p
r
o
v
id
es
th
e
l
o
west
p
e
r
f
o
r
m
a
n
ce
.
B
ec
au
s
e
NB
class
if
ies
o
n
ly
1
0
8
im
a
g
es
co
r
r
ec
tly
a
m
o
n
g
1
9
1
im
a
g
es
an
d
ca
n
n
o
t
class
if
y
in
s
ilv
er
cu
p
f
is
h
.
W
e
k
n
o
w,
NB
is
p
r
o
b
ab
ilis
tic
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
an
d
it
s
tu
d
ies
th
at
th
e
f
ea
tu
r
es
ar
e
f
r
e
e
o
f
ea
ch
o
th
e
r
.
I
t
also
g
iv
es
lo
wer
ac
cu
r
ac
y
th
an
o
th
er
class
if
ier
m
o
d
els.
Ho
we
v
er
,
in
r
ea
l
wo
r
ld
,
f
ea
tu
r
es
d
e
p
en
d
o
n
ea
ch
o
th
er
.
I
f
we
ad
d
m
u
ltip
le
class
if
ier
s
in
th
e
m
o
d
el,
th
e
co
m
p
u
tatio
n
al
co
m
p
lex
ity
will
b
e
h
ig
h
e
r
an
d
f
o
r
o
u
r
test
ed
d
ataset,
we
alr
ea
d
y
h
av
e
a
s
ig
n
if
ican
t r
esu
lt f
o
r
o
u
r
m
o
d
e
l.
T
ab
le
3
.
C
o
m
p
a
r
is
o
n
am
o
n
g
c
lass
if
icatio
n
m
o
d
el
b
ased
o
n
p
er
f
o
r
m
a
n
ce
m
etr
ics
S
.
L.
N
o
.
M
a
c
h
i
n
e
l
e
a
r
n
i
n
g
m
o
d
e
l
A
c
c
u
r
a
c
y
(
%)
K
a
p
p
a
s
t
a
t
i
s
t
i
c
(
K
S
)
(
%)
A
v
g
.
TP r
a
t
e
(
%)
R
e
mar
k
s
1
R
F
(
P
r
o
p
o
s
e
d
M
o
d
e
l
)
8
8
.
4
8
8
7
.
1
1
8
8
.
5
H
i
g
h
e
s
t
2
J4
8
[
3
0
]
7
3
.
1
6
6
9
.
8
8
7
3
.
2
3
rd
H
i
g
h
e
s
t
3
N
B
[
3
1
]
5
6
.
8
4
5
1
.
6
0
5
6
.
8
Lo
w
e
s
t
5
k
-
N
N
[
3
2
]
8
5
.
7
9
8
4
.
0
5
8
5
.
8
2
nd
H
i
g
h
e
s
t
6
C
A
R
T
[
3
3
]
6
4
.
2
1
5
9
.
8
0
6
4
.
2
4
t
h
H
i
g
h
e
st
T
h
ese
p
er
f
o
r
m
an
ce
m
etr
ics
ar
e
s
h
o
wn
in
Fig
u
r
e
5
g
r
ap
h
ically
.
W
e
m
ar
k
ed
t
h
r
ee
c
o
lo
r
ed
cu
r
v
es
f
o
r
th
r
ee
p
er
f
o
r
m
a
n
ce
m
etr
ics.
T
h
e
b
lu
e
cu
r
v
e
is
m
ar
k
ed
as
a
n
ac
cu
r
ac
y
m
etr
ic.
T
h
e
m
id
d
l
e
cu
r
v
e
is
id
en
tifie
d
f
o
r
k
ap
p
a
s
tatis
tic
wh
ich
is
m
a
r
o
o
n
co
lo
r
an
d
t
h
e
g
r
ee
n
cu
r
v
e
is
n
o
ticed
f
o
r
T
P
r
ate.
W
e
ca
n
s
ee
f
r
o
m
th
is
,
th
e
p
r
o
p
o
s
ed
m
o
d
e
l,
R
F
g
iv
es
th
e
h
ig
h
est
s
co
r
e
in
all
ca
teg
o
r
ies
o
f
p
er
f
o
r
m
an
ce
m
etr
ics.
All
cir
cle
p
o
in
t
f
o
r
R
F
m
o
d
el
h
as th
e
to
p
p
o
s
itio
n
in
p
er
f
o
r
m
an
ce
m
etr
ics as a
cc
u
r
a
cy
8
8
.
4
8
%,
KS a
s
8
7
.
1
1
% a
n
d
T
P r
ate
as 8
8
.
5
%.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
10
,
N
o
.
3
,
Sep
tem
b
er
202
1
:
6
1
4
-
622
620
Fig
u
r
e
5
.
Acc
u
r
ac
y
a
n
aly
s
is
5.
CO
NCLU
SI
O
N
W
e
co
n
d
u
cted
th
is
r
esear
ch
to
f
in
d
o
u
t
th
e
b
est
p
r
ed
ictio
n
m
o
d
el
f
o
r
f
is
h
f
ar
m
er
s
in
an
aq
u
atic
en
v
ir
o
n
m
en
t
u
s
in
g
v
ar
io
u
s
aq
u
atic
p
ar
am
eter
s
.
W
e
u
s
ed
p
H
,
tem
p
er
atu
r
e
,
tu
r
b
i
d
ity
,
an
d
f
i
s
h
as
p
ar
am
eter
s
o
f
th
e
d
ataset
wh
er
e
we
m
ar
k
ed
tem
p
er
atu
r
e,
p
H,
tu
r
b
i
d
i
ty
as
f
ea
tu
r
e
v
ar
iab
les
an
d
f
is
h
as
th
e
tar
g
et
v
ar
iab
le.
W
e
u
s
ed
to
tal
1
1
t
y
p
es
o
f
f
is
h
.
T
h
e
y
a
r
e
k
atla,
s
h
in
g
,
p
r
a
wn
,
r
u
i,
k
o
i,
p
an
g
as,
tilap
ia,
s
ilv
er
ca
r
p
,
ca
r
p
io
,
m
ag
u
r
an
d
s
h
r
im
p
.
W
e
f
in
d
o
u
t
t
h
e
ac
cu
r
ac
y
,
k
ap
p
a
s
tatis
tic
an
d
T
P
r
ate
as
p
er
f
o
r
m
an
ce
m
e
tr
ics.
W
e
an
aly
ze
d
a
t
o
tal
o
f
f
iv
e
s
u
p
er
v
is
ed
m
ac
h
in
e
m
o
d
els.
T
h
e
y
ar
e
R
F,
NB
,
k
-
NN,
C
AR
T
an
d
J
4
8
.
Am
o
n
g
t
h
ese
m
o
d
els,
o
u
r
p
r
o
p
o
s
ed
m
o
d
el,
RF
s
h
o
ws
th
e
b
est
ac
cu
r
ac
y
,
k
ap
p
a
s
tatis
tic
an
d
T
P
r
ate
as
p
er
f
o
r
m
a
n
ce
m
etr
ics
th
at
ca
n
p
r
ed
ict
t
h
e
m
o
s
t
f
is
h
s
p
ec
ies
in
an
a
q
u
atic
e
n
v
ir
o
n
m
en
t.
RF
p
r
o
v
id
es
ac
cu
r
ac
y
8
8
.
4
8
%,
KS
8
7
.
1
1
%
an
d
T
P
r
ate
8
8
.
5
%.
Fu
r
th
er
,
t
h
e
r
esear
ch
s
co
p
e
ca
n
b
e
d
ef
i
n
ed
b
y
en
r
ic
h
in
g
th
e
d
ataset
b
y
m
o
r
e
o
b
s
er
v
atio
n
an
d
test
in
g
with
ar
tific
ial
n
eu
r
al
n
etwo
r
k
.
ACK
NO
WL
E
DG
E
M
E
NT
S
T
h
is
r
esear
ch
is
f
u
n
d
e
d
b
y
W
o
o
s
o
n
g
Un
iv
e
r
s
ity
Aca
d
em
ic
R
esear
ch
in
2
0
2
1
.
RE
F
E
R
E
NC
E
S
[1
]
F
AO
,
Th
e
S
tate
o
f
Wo
rl
d
F
ish
e
ri
e
s
a
n
d
A
q
u
a
c
u
lt
u
re
2
0
2
0
-
s
u
sta
in
a
b
il
it
y
i
n
a
c
ti
o
n
,
2
0
2
0
,
F
AO
,
Ital
y
,
2
0
2
0
,
p
p
.
0
-
2
4
4
,
h
tt
p
:
//
ww
w.fao
.
o
rg
/
3
/ca
9
2
3
1
e
n
/CA9
2
3
1
EN.
p
d
f
.
[2
]
M
.
K.
Alsm
a
d
i,
K.
B.
Om
a
r,
S
.
A.
No
a
h
,
a
n
d
A.
I
.
Alm
a
ra
sh
d
e
h
,
“
F
ish
Re
c
o
g
n
it
i
o
n
Ba
se
d
o
n
R
o
b
u
st
F
e
a
t
u
re
s
Ex
trac
ti
o
n
fr
o
m
S
ize
a
n
d
S
h
a
p
e
M
e
a
su
re
m
e
n
ts
Us
in
g
Ne
u
ra
l
Ne
t
wo
rk
,
”
J
o
u
rn
a
l
o
f
Co
m
p
u
ter
S
c
ie
n
c
e
,
v
o
l.
6
,
n
o
.
1
0
,
p
p
.
1
0
8
8
-
1
0
9
4
,
2
0
1
0
,
d
o
i:
1
0
.
3
8
4
4
/j
c
ss
p
.
2
0
1
0
.
1
0
8
8
.
1
0
9
4.
[3
]
T.
H.
H
o
a
n
g
,
K.
Lo
c
k
,
A.
M
o
u
t
o
n
,
a
n
d
P
.
L.
M
.
G
o
e
th
a
ls,
“
Ap
p
l
i
c
a
ti
o
n
o
f
c
las
sifica
ti
o
n
tree
s
a
n
d
su
p
p
o
rt
v
e
c
to
r
m
a
c
h
in
e
s
to
m
o
d
e
l
th
e
p
re
se
n
c
e
o
f
m
a
c
ro
in
v
e
rteb
ra
tes
in
riv
e
rs
in
Vie
tn
a
m
,
”
Eco
lo
g
ica
l
I
n
fo
rm
a
t
ics
,
v
o
l.
5
,
n
o
.
2
,
2
0
1
0
,
p
p
.
1
4
0
-
1
4
6
,
2
0
1
0
,
d
o
i
:
1
0
.
1
0
1
6
/j
.
e
c
o
in
f.
2
0
0
9
.
1
2
.
0
0
1
.
[4
]
B.
Be
n
so
n
,
J.
C
h
o
,
D.
G
o
sh
o
rn
,
a
n
d
R
.
k
a
st
n
e
r,
“
F
ield
p
r
o
g
ra
m
m
a
b
le
g
a
te
a
rra
y
(
F
P
G
A)
b
a
se
d
fish
d
e
tec
ti
o
n
u
sin
g
h
a
a
r
c
las
sifiers
,
”
Ame
ric
a
n
Aca
d
e
my
o
f
Un
d
e
rwa
ter
S
c
ien
c
e
s
,
G
e
o
rg
ia,
USA,
2
0
0
9
,
p
p
.
1
-
8
.
[5
]
S
.
Be
rm
e
jo
,
“
F
ish
a
g
e
c
las
sifica
ti
o
n
b
a
se
d
o
n
len
g
th
,
we
ig
h
t,
se
x
a
n
d
o
t
o
li
t
h
m
o
r
p
h
o
l
o
g
ica
l
fe
a
tu
re
s,”
Fi
sh
e
rie
s
Res
e
a
rc
h
,
v
o
l.
8
4
,
n
o
.
2
,
p
p
.
2
7
0
-
2
7
4
,
2
0
0
7
,
d
o
i:
1
0
.
1
0
1
6
/
j.
fish
re
s.
2
0
0
6
.
1
2
.
0
0
7
.
[6
]
A.
G
.
Ca
b
re
ira,
M
.
Tri
p
o
d
e
,
a
n
d
A.
M
a
d
iro
las
,
“
Artifi
c
ial
n
e
u
ra
l
n
e
two
rk
s
fo
r
fis
h
-
s
p
e
c
ies
id
e
n
ti
fica
ti
o
n
,
”
ICES
J
o
u
rn
a
l
o
f
M
a
rin
e
S
c
ien
c
e
,
v
o
l.
6
6
,
n
o
.
6
,
p
p
.
1
1
1
9
-
1
1
2
9
,
2
0
0
9
,
d
o
i
:
1
0
.
1
0
9
3
/i
c
e
sjm
s/fsp
0
0
9
.
[7
]
I.
Kh
a
n
,
X.
Z
h
a
n
g
,
M
.
Re
h
m
a
n
,
a
n
d
R.
Ali,
“
A
Li
tera
tu
re
S
u
r
v
e
y
a
n
d
Emp
iri
c
a
l
S
t
u
d
y
o
f
M
e
ta
-
Lea
rn
in
g
f
o
r
Clas
sifier S
e
lec
ti
o
n
,
”
IEE
E
Ac
c
e
ss
,
v
o
l
.
8
,
p
p
.
1
0
2
6
2
-
1
0
2
8
1
,
2
0
2
0
,
d
o
i:
1
0
.
1
1
0
9
/ACCES
S
.
2
0
2
0
.
2
9
6
4
7
2
6
.
[8
]
J.
Hu
,
D.
L
i,
Q.
D
u
a
n
,
Y.
Ha
n
,
G
.
Ch
e
n
,
a
n
d
X.
S
i,
“
F
ish
sp
e
c
ie
s
c
las
sifica
ti
o
n
b
y
c
o
l
o
r,
tex
tu
re
a
n
d
m
u
lt
i
-
c
las
s
su
p
p
o
rt
v
e
c
to
r
m
a
c
h
in
e
u
sin
g
c
o
m
p
u
ter
v
isio
n
,
”
Co
m
p
u
ter
s
a
n
d
El
e
c
tro
n
ic
s
in
Ag
ric
u
lt
u
re
,
v
o
l.
8
8
,
p
p
.
1
3
3
-
1
4
0
,
2
0
1
2
,
d
o
i:
1
0
.
1
0
1
6
/
j.
c
o
m
p
a
g
.
2
0
1
2
.
0
7
.
0
0
8
.
8
8
.
4
8
5
6
.
8
4
8
5
.
7
9
6
4
.
2
1
7
3
.
1
6
8
7
.
1
1
5
1
.
6
8
4
.
0
5
5
9
.
8
6
9
.
8
8
8
8
.
5
5
6
.
8
8
5
.
8
6
4
.
2
7
3
.
2
RF
NB
K
N
N
C
A
R
T
J4
8
P
e
r
f
or
m
an
c
e
M
e
t
r
i
c
s
An
al
y
si
s
A
c
c
u
r
a
c
y
KS
TP
R
a
t
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
F
is
h
s
u
r
viva
l p
r
ed
ictio
n
in
a
n
a
q
u
a
tic
e
n
viro
n
men
t u
s
in
g
r
a
n
d
o
m
fo
r
est mo
d
el
(
Md
.
Mo
n
ir
u
l I
s
la
m
)
621
[9
]
G
.
Bo
u
rk
e
,
F
.
S
tag
n
it
ti
,
a
n
d
B.
M
it
c
h
e
ll
,
“
A
d
e
c
isio
n
s
u
p
p
o
rt
s
y
ste
m
fo
r
a
q
u
a
c
u
lt
u
re
re
se
a
rc
h
a
n
d
m
a
n
a
g
e
m
e
n
t,
”
Aq
u
a
c
u
lt
u
ra
l
En
g
in
e
e
rin
g
,
v
o
l
.
1
2
,
n
o
.
2
,
p
p
.
1
1
1
-
1
2
3
,
1
9
9
3
,
d
o
i
:
1
0
.
1
0
1
6
/0
1
4
4
-
8
6
0
9
(9
3
)9
0
0
2
0
-
C.
[1
0
]
W.
Li
,
H.
Wu
,
N.
Zh
u
,
Y.
Jia
n
g
,
J.
Tan
,
a
n
d
Y.
G
u
o
,
“
P
re
d
icti
o
n
o
f
d
isso
l
v
e
d
o
x
y
g
e
n
i
n
a
fish
e
ry
p
o
n
d
b
a
se
d
o
n
g
a
ted
re
c
u
rre
n
t
u
n
it
(G
RU),”
In
fo
rm
a
t
io
n
Pr
o
c
e
ss
in
g
in
Ag
ric
u
lt
u
re
,
v
o
l.
8
,
n
o
.
1
,
p
p
.
1
8
5
-
1
9
3
,
2
0
2
1
,
d
o
i
:
j.
in
p
a
.
2
0
2
0
.
0
2
.
0
0
2
.
[1
1
]
G
.
A.
De
fe
a
n
d
A.
Z.
C.
An
to
n
io
,
“
M
u
l
ti
-
p
a
ra
m
e
ter
Wate
r
Q
u
a
li
ty
M
o
n
i
to
ri
n
g
De
v
ice
f
o
r
G
ro
u
p
e
r
A
q
u
a
c
u
lt
u
re
,
”
i
n
IEE
E
1
0
th
In
ter
n
a
ti
o
n
a
l
C
o
n
fer
e
n
c
e
o
n
H
u
ma
n
o
i
d
,
N
a
n
o
tec
h
n
o
lo
g
y
,
In
f
o
rm
a
ti
o
n
T
e
c
h
n
o
lo
g
y
,
C
o
mm
u
n
ica
t
io
n
a
n
d
Co
n
tro
l,
E
n
v
iro
n
me
n
t
a
n
d
M
a
n
a
g
e
me
n
t
,
2
0
1
8
,
p
p
.
1
-
5
,
d
o
i:
1
0
.
1
1
0
9
/HNICEM
.
2
0
1
8
.
8
6
6
6
4
1
4
.
[1
2
]
T.
Ab
in
a
y
a
,
J.
Ish
wa
ry
a
,
a
n
d
M
.
M
a
h
e
sw
a
ri,
“
A
No
v
e
l
M
e
t
h
o
d
o
lo
g
y
fo
r
M
o
n
i
to
ri
n
g
a
n
d
C
o
n
tr
o
ll
in
g
o
f
Wate
r
Qu
a
li
ty
in
Aq
u
a
c
u
lt
u
re
u
si
n
g
In
tern
e
t
o
f
Th
i
n
g
s
(Io
T),
”
i
n
2
0
1
9
In
ter
n
a
ti
o
n
a
l
C
o
n
fer
e
n
c
e
o
n
C
o
mp
u
ter
Co
mm
u
n
ica
ti
o
n
a
n
d
In
f
o
rm
a
ti
c
s
,
2
0
1
9
,
p
p
.
1
-
4
,
d
o
i:
1
0
.
1
1
0
9
/ICCC
I.
2
0
1
9
.
8
8
2
1
9
8
8
.
[1
3
]
P.
-
Y.
Ya
n
g
,
J.
-
T.
Tsa
i,
J.
-
H.
C
h
o
u
,
W.
-
H.
Ho
,
a
n
d
Y.
-
Y.
Lai,
“
P
re
d
ictio
n
o
f
wa
ter
q
u
a
li
ty
e
v
a
l
u
a
ti
o
n
f
o
r
f
ish
p
o
n
d
s
o
f
a
q
u
a
c
u
lt
u
re
,
”
i
n
2
0
1
7
5
6
th
A
n
n
u
a
l
C
o
n
fer
e
n
c
e
o
f
t
h
e
S
o
c
iety
o
f
I
n
stru
me
n
t
a
n
d
Co
n
tro
l
En
g
i
n
e
e
rs
o
f
J
a
p
a
n
(S
IC
E)
,
2
0
1
7
,
p
p
.
5
4
5
-
5
4
6
,
d
o
i:
1
0
.
2
3
9
1
9
/
S
ICE.
2
0
1
7
.
8
1
0
5
4
5
5
.
[1
4
]
S
.
O.
O
g
u
n
lan
a
,
O.
Ola
b
o
d
e
,
S
.
Olu
wa
d
a
re
,
a
n
d
G
.
Iwa
so
k
u
n
,
“
F
ish
c
las
sifica
ti
o
n
u
sin
g
su
p
p
o
rt
v
e
c
to
r
m
a
c
h
in
e
,
”
Af
ric
a
n
J
o
u
rn
a
l
o
f
C
o
mp
u
ti
n
g
a
n
d
ICT
,
v
o
l.
8
,
n
o
.
2
,
2
0
0
6
,
p
p
.
1
-
8.
[1
5
]
M
.
M
.
M
.
F
o
u
a
d
,
H.
M
.
Zaw
b
a
a
,
N.
El
-
Be
n
d
a
r
y
,
a
n
d
A.
E.
Ha
ss
a
n
ien
,
“
Au
to
m
a
ti
c
Ni
le
Ti
la
p
ia
fis
h
c
las
sifica
ti
o
n
a
p
p
ro
a
c
h
u
si
n
g
m
a
c
h
in
e
lea
rn
i
n
g
tec
h
n
iq
u
e
s,”
in
1
3
t
h
I
n
ter
n
a
ti
o
n
a
l
C
o
n
fer
e
n
c
e
o
n
Hy
b
rid
In
te
ll
ig
e
n
t
S
y
ste
ms
,
2
0
1
3
,
p
p
.
1
7
3
-
1
7
8
,
d
o
i:
1
0
.
1
1
0
9
/
HIS.
2
0
1
3
.
6
9
2
0
4
7
7
.
[1
6
]
D.
H.
Er
n
st,
J
.
P
.
B
o
lt
e
,
a
n
d
S
.
S
.
Na
th
,
“
Aq
u
a
F
a
rm
:
sim
u
lati
o
n
a
n
d
d
e
c
isio
n
su
p
p
o
rt
fo
r
a
q
u
a
c
u
l
tu
r
e
fa
c
il
it
y
d
e
sig
n
a
n
d
m
a
n
a
g
e
m
e
n
t
p
la
n
n
i
n
g
,
”
A
q
u
a
c
u
lt
u
ra
l
En
g
in
e
e
rin
g
,
v
o
l.
2
3
,
n
o
.
1
-
3
,
p
p
.
1
2
1
-
1
7
9
,
2
0
0
0
,
d
o
i:
1
0
.
1
0
1
6
/S
0
1
4
4
-
8
6
0
9
(
0
0
)
0
0
0
4
5
-
5.
[1
7
]
W.
S
il
v
e
rt,
“
De
c
isio
n
su
p
p
o
rt
sy
s
tem
s fo
r
a
q
u
a
c
u
lt
u
re
li
c
e
n
si
n
g
,
”
J
o
u
rn
a
l
o
f
Ap
p
li
e
d
Ic
h
th
y
o
lo
g
y
,
v
o
l.
1
0
,
n
o
.
4
,
p
p
.
307
-
3
1
1
,
1
9
9
4
,
d
o
i:
1
0
.
1
1
1
1
/j
.
1
4
3
9
-
0
4
2
6
.
1
9
9
4
.
tb
0
0
1
7
0
.
x
.
[1
8
]
H.
Ha
li
d
e
,
A.
S
ti
g
e
b
ra
n
d
t
,
M
.
Re
h
b
e
in
,
a
n
d
A.
D.
M
c
Kin
n
o
n
,
“
De
v
e
lo
p
in
g
a
d
e
c
isio
n
su
p
p
o
rt
sy
ste
m
fo
r
su
sta
in
a
b
le
c
a
g
e
a
q
u
a
c
u
lt
u
re
,
”
En
v
iro
n
me
n
ta
l
M
o
d
e
ll
in
g
&
S
o
f
twa
re
,
v
o
l
.
2
4
,
n
o
.
6
,
p
p
.
6
9
4
-
7
0
2
,
2
0
0
9
,
d
o
i:
1
0
.
1
0
1
6
/
j.
e
n
v
so
ft.
2
0
0
8
.
1
0
.
0
1
3
.
[1
9
]
C.
D’Este
,
A.
Ra
h
m
a
n
,
a
n
d
A.
Tu
rn
b
u
l
l,
“
P
re
d
ictin
g
S
h
e
ll
fish
F
a
rm
Clo
su
re
s
wit
h
Clas
s
Ba
lan
c
in
g
M
e
t
h
o
d
s,
”
L
e
c
tu
re
No
tes
in
C
o
mp
u
ter
S
c
ien
c
e
,
p
p
.
3
9
-
4
8
,
2
0
1
2
,
d
o
i:
1
0
.
1
0
0
7
/
9
7
8
-
3
-
6
4
2
-
3
5
1
0
1
-
3
_
4
.
[2
0
]
A.
Ra
h
m
a
n
,
C.
D’Este
,
a
n
d
J.
M
c
Cu
ll
o
c
h
,
“
E
n
se
m
b
le
F
e
a
tu
r
e
Ra
n
k
in
g
f
o
r
S
h
e
ll
fish
F
a
rm
Clo
su
re
Ca
u
se
Id
e
n
ti
fica
ti
o
n
,
”
in
W
o
rk
sh
o
p
o
n
M
a
c
h
i
n
e
L
e
a
r
n
in
g
fo
r
S
e
n
so
ry
Da
t
a
An
a
lys
is
,
2
0
1
3
,
p
p
.
1
3
-
1
8
,
d
o
i
:
1
0
.
1
1
4
5
/
2
5
4
2
6
5
2
.
2
5
4
2
6
5
5
.
[2
1
]
M
d
.
S
.
S
h
a
h
riar,
A.
Ra
h
m
a
n
,
a
n
d
J.
M
c
Cu
ll
o
c
h
,
“
P
re
d
ict
in
g
sh
e
ll
fi
sh
fa
rm
c
lo
su
re
s
u
sin
g
ti
m
e
se
ries
c
las
sifica
ti
o
n
fo
r
a
q
u
a
c
u
lt
u
re
d
e
c
isi
o
n
su
p
p
o
r
t
,
”
Co
mp
u
ter
s
a
n
d
El
e
c
tro
n
ics
i
n
A
g
ric
u
lt
u
re
,
v
o
l
.
1
0
2
,
p
p
.
8
5
-
9
7
,
2
0
1
4
,
d
o
i:
1
0
.
1
0
1
6
/
j.
c
o
m
p
a
g
.
2
0
1
4
.
0
1
.
0
1
1
.
[2
2
]
M
.
S
.
S
h
a
h
r
iar,
C.
D’Este
,
a
n
d
A.
Ra
h
m
a
n
,
“
On
d
e
tec
ti
n
g
a
n
d
p
re
d
ictin
g
h
a
rm
fu
l
a
lg
a
l
b
lo
o
m
s
in
c
o
a
sta
l
in
fo
rm
a
ti
o
n
sy
ste
m
s,”
in
2
0
1
2
Oc
e
a
n
s
-
Y
e
o
su
,
2
0
1
2
,
p
p
.
1
-
3
,
d
o
i:
1
0
.
1
1
0
9
/OCEANS
-
Ye
o
s
u
.
2
0
1
2
.
6
2
6
3
4
0
6
.
[2
3
]
A.
Ra
h
m
a
n
a
n
d
M
.
S
.
S
h
a
h
riar
,
“
Alg
a
e
g
ro
wth
p
re
d
icti
o
n
t
h
ro
u
g
h
id
e
n
t
ifi
c
a
ti
o
n
o
f
i
n
flu
e
n
ti
a
l
e
n
v
iro
n
m
e
n
tal
v
a
riab
les
:
A
m
a
c
h
in
e
lea
rn
in
g
a
p
p
r
o
a
c
h
,
”
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
Co
mp
u
ta
ti
o
n
a
l
I
n
telli
g
e
n
c
e
a
n
d
Ap
p
li
c
a
ti
o
n
s
,
vol
.
1
2
,
n
o
.
2
,
p
p
.
1
3
5
0
0
0
8
-
1
9
,
2
0
1
3
,
d
o
i:
1
0
.
1
1
4
2
/
S
1
4
6
9
0
2
6
8
1
3
5
0
0
0
8
9
.
[2
4
]
M
.
S
.
S
h
a
h
riar
a
n
d
A.
Ra
h
m
a
n
,
“
S
p
a
ti
a
l
-
tem
p
o
ra
l
p
re
d
ictio
n
o
f
a
lg
a
l
b
lo
o
m
,
”
in
Ni
n
th
I
n
ter
n
a
t
io
n
a
l
Co
n
fer
e
n
c
e
o
n
Na
tu
ra
l
Co
m
p
u
t
a
ti
o
n
,
2
0
1
3
,
p
p
.
9
7
3
-
9
7
7
,
d
o
i:
1
0
.
1
1
0
9
/ICNC.
2
0
1
3
.
6
8
1
8
1
1
7
.
[2
5
]
M
.
M
.
Isla
m
,
M
.
A.
Ka
sh
e
m
,
a
n
d
F
.
I.
Ju
i,
“
Aq
u
a
fish
i
n
g
m
o
n
i
to
rin
g
sy
ste
m
u
sin
g
I
o
T
d
e
v
ice
s,”
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
In
n
o
v
a
ti
v
e
S
c
ien
c
e
,
E
n
g
in
e
e
rin
g
a
n
d
T
e
c
h
n
o
lo
g
y
,
v
o
l.
6
,
n
o
.
1
1
,
2
0
1
9
,
p
p
.
1
0
8
-
1
1
4
.
[2
6
]
[o
n
li
n
e
].
h
tt
p
:/
/www
.
fa
o
.
o
rg
/f
ish
e
ry
/d
o
c
s/CDro
m
/F
AO
_
Trai
n
in
g
/F
AO
_
Trai
n
in
g
/G
e
n
e
ra
l/
x
6
7
0
9
e
/x
6
7
0
9
e
0
2
.
h
t
m
Ac
c
e
ss
e
d
4
Ap
ril
2
0
2
1
.
[2
7
]
A.
Bh
a
tn
a
g
a
r
a
n
d
P
.
De
v
i,
“
Wa
ter
q
u
a
li
t
y
g
u
id
e
li
n
e
s
fo
r
t
h
e
m
a
n
a
g
e
m
e
n
t
o
f
p
o
n
d
fish
c
u
lt
u
re
,
”
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
E
n
v
iro
n
me
n
t
a
l
S
c
ien
c
e
s
,
v
o
l
.
3
,
n
o
.
6
,
2
0
0
8
,
p
p
.
1
9
8
0
-
2
0
0
9
.
[2
8
]
Z.
F
e
rd
o
u
sh
,
B
.
N.
M
a
h
m
u
d
,
A.
Ch
a
k
ra
b
a
rty
,
a
n
d
J.
Ud
d
i
n
,
“
A
sh
o
rt
-
term
h
y
b
rid
f
o
re
c
a
stin
g
m
o
d
e
l
fo
r
ti
m
e
se
ries
e
lec
tri
c
a
l
-
lo
a
d
d
a
ta
u
sin
g
ra
n
d
o
m
fo
re
st
a
n
d
b
id
irec
ti
o
n
a
l
l
o
n
g
sh
o
rt
-
term
m
e
m
o
ry
,
”
In
ter
n
a
ti
o
n
a
l
J
o
u
r
n
a
l
o
f
El
e
c
trica
l
a
n
d
C
o
mp
u
ter
E
n
g
in
e
e
rin
g
(IJ
ECE
)
,
v
o
l
.
1
1
,
n
o
.
1
,
p
p
.
7
6
3
-
7
7
1
,
F
e
b
.
2
0
2
1
,
d
o
i
:
1
0
.
1
1
5
9
1
/i
jec
e
.
v
1
1
i
1
.
p
p
7
6
3
-
7
7
1
.
[2
9
]
M
.
Be
lg
i
u
a
n
d
L.
Dră
g
u
ţ,
“
Ra
n
d
o
m
fo
re
st
i
n
re
m
o
te
se
n
sin
g
:
A
re
v
iew
o
f
a
p
p
li
c
a
ti
o
n
s
a
n
d
f
u
tu
re
d
irec
ti
o
n
s,”
IS
PR
S
J
o
u
rn
a
l
o
f
Ph
o
to
g
r
a
mm
e
try
a
n
d
Rem
o
te
S
e
n
s
in
g
,
v
o
l.
1
1
4
,
p
p
.
2
4
-
3
1
,
2
0
1
6
,
d
o
i:
1
0
.
1
0
1
6
/
j.
isp
rsj
p
rs.2
0
1
6
.
0
1
.
0
1
1
.
[3
0
]
G
.
Ka
u
r
a
n
d
A.
Ch
h
a
b
ra
,
“
Im
p
r
o
v
e
d
J4
8
Clas
sifica
ti
o
n
Al
g
o
ri
th
m
fo
r
th
e
P
re
d
ictio
n
o
f
Dia
b
e
tes
,
”
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
Co
m
p
u
ter
A
p
p
l
ica
ti
o
n
s
,
v
o
l
.
9
8
,
n
o
.
2
2
,
p
p
.
1
3
-
1
7
,
J
u
l.
2
0
1
4
,
d
o
i:
1
0
.
5
1
2
0
/
1
7
3
1
4
-
7
4
3
3
.
[3
1
]
T
.
Kim
,
B.
D.
C
h
u
n
g
,
a
n
d
J.
-
S
.
Lee
,
“
In
c
o
r
p
o
ra
ti
n
g
re
c
e
iv
e
r
o
p
e
ra
ti
n
g
c
h
a
ra
c
teristics
i
n
to
n
a
iv
e
Ba
y
e
s
fo
r
u
n
b
a
la
n
c
e
d
d
a
ta cla
ss
ifi
c
a
ti
o
n
,
”
C
o
mp
u
t
in
g
,
v
o
l
.
9
9
,
n
o
.
3
,
p
p
.
2
0
3
-
2
1
8
,
2
0
1
6
,
d
o
i:
1
0
.
1
0
0
7
/s
0
0
6
0
7
-
0
1
6
-
0
4
8
3
-
z.
[3
2
]
M
.
M
.
Isla
m
,
J.
Ud
d
in
,
M
.
A.
Ka
sh
e
m
,
F
.
Ra
b
b
i,
a
n
d
M
d
.
W.
Ha
sn
a
t,
“
De
sig
n
a
n
d
Im
p
lem
e
n
tatio
n
o
f
a
n
I
o
T
S
y
ste
m
fo
r
P
re
d
ictin
g
A
q
u
a
F
ish
e
ries
Us
in
g
Ard
u
in
o
a
n
d
KN
N,”
I
n
telli
g
e
n
t
Hu
m
a
n
Co
m
p
u
ter
In
ter
a
c
ti
o
n
,
p
p
.
1
0
8
-
1
1
8
,
2
0
2
1
,
d
o
i:
1
0
.
1
0
0
7
/
9
7
8
-
3
-
0
3
0
-
6
8
4
5
2
-
5
_
1
1
.
[3
3
]
L.
Ru
tk
o
ws
k
i
,
L.
P
ietr
u
c
z
u
k
,
P
.
Du
d
a
,
a
n
d
M
.
Ja
wo
rsk
i,
“
De
c
isio
n
Tree
s
fo
r
M
i
n
in
g
Da
ta
S
trea
m
s
Ba
se
d
o
n
th
e
M
c
Dia
rm
id
’s
Bo
u
n
d
,
”
IEE
E
T
ra
n
sa
c
ti
o
n
s
o
n
Kn
o
wled
g
e
a
n
d
D
a
ta
E
n
g
i
n
e
e
rin
g
,
v
o
l.
2
5
,
n
o
.
6
,
p
p
.
1
2
7
2
-
1
2
7
9
,
2
0
1
3
,
d
o
i:
1
0
.
1
1
0
9
/
TKDE.
2
0
1
2
.
6
6
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
10
,
N
o
.
3
,
Sep
tem
b
er
202
1
:
6
1
4
-
622
622
B
I
O
G
RAP
H
I
E
S
O
F
AUTH
O
RS
Md
.
Mo
n
irul
Is
la
m
is
a
n
M
.
S
c
.
S
tu
d
e
n
t,
De
p
a
rtme
n
t
o
f
Co
m
p
u
t
e
r
S
c
ien
c
e
a
n
d
En
g
i
n
e
e
rin
g
(CS
E),
Dh
a
k
a
Un
i
v
e
rsity
o
f
En
g
i
n
e
e
rin
g
a
n
d
Tec
h
n
o
lo
g
y
(DU
ET
),
G
a
z
ip
u
r
-
1
7
0
0
,
Ba
n
g
lad
e
s
h
.
He
h
a
s
c
o
m
p
lete
d
B.
S
c
.
in
e
n
g
i
n
e
e
rin
g
d
e
g
re
e
in
CS
E
i
n
th
e
y
e
a
r
2
0
1
7
fro
m
P
a
b
n
a
Un
iv
e
rsit
y
o
f
S
c
ien
c
e
a
n
d
Tec
h
n
o
lo
g
y
(P
UST
),
P
a
b
n
a
-
6
6
0
0
,
Ba
n
g
la
d
e
sh
.
Cu
rre
n
t
ly
,
h
e
is
a
t
Uttara
Un
iv
e
rsity
(UU
)
a
s
a
fu
ll
-
ti
m
e
l
e
c
tu
re
r
sin
c
e
3
rd
F
e
b
ru
a
r
y
2
0
2
0
.
F
ield
o
f
i
n
tere
st:
M
a
c
h
i
n
e
lea
rn
in
g
/De
e
p
Lea
rn
in
g
,
Da
ta S
c
i
e
n
c
e
.
Mo
h
a
m
m
o
d
Abu
l
K
a
shem
h
a
s
b
e
e
n
se
rv
i
n
g
a
s
a
p
ro
fe
ss
o
r,
De
p
a
rtme
n
t
o
f
Co
m
p
u
ter
S
c
ien
c
e
a
n
d
En
g
in
e
e
rin
g
(CS
E)
a
n
d
Ad
d
i
ti
o
n
a
l
Dire
c
to
r,
P
la
n
n
i
n
g
&
De
v
e
lo
p
m
e
n
t,
Dh
a
k
a
Un
iv
e
rsity
o
f
E
n
g
i
n
e
e
rin
g
&
Tec
h
n
o
lo
g
y
(
DU
ET
),
G
a
z
ip
u
r
-
1
7
0
0
,
Ba
n
g
la
d
e
sh
.
He
g
o
t
P
o
st
Do
c
t
o
ra
te
(ULL
,
F
ra
n
c
e
),
P
h
D
(LP
NU
,
Uk
ra
in
e
),
M
.
En
g
.
&
B.
E
n
g
.
(S
ULP
,
Uk
ra
in
e
).
F
iel
d
o
f
i
n
tere
st:
In
tern
e
t
o
f
Th
in
g
s,
Cy
b
e
r
P
h
y
sic
a
l
S
y
ste
m
,
T
h
e
Ch
a
ll
e
n
g
e
s,
Ap
p
ro
a
c
h
e
s
a
n
d
Us
e
d
Tec
h
n
i
q
u
e
s
o
f
CP
S
fo
r
M
a
n
u
fa
c
tu
ri
n
g
in
In
d
u
stry
4
.
0
,
CP
S
o
f
c
o
ll
a
b
o
ra
ti
v
e
v
a
lu
e
n
e
two
rk
fo
r
m
a
n
u
fa
c
tu
rin
g
,
m
u
l
ti
a
g
e
n
t
sy
ste
m
fo
r
in
d
u
str
y
a
n
d
c
rit
ica
l
m
issio
n
.
J
ia
U
d
d
in
h
a
s
b
e
e
n
se
r
v
in
g
a
s
a
n
As
sista
n
t
P
ro
fe
ss
o
r,
De
p
a
rtme
n
t
o
f
Tec
h
n
o
lo
g
y
st
u
d
ies
,
En
d
ic
o
tt
C
o
ll
e
g
e
,
W
o
o
s
o
n
g
Un
i
v
e
rsity
,
Da
e
jeo
n
,
S
o
u
th
K
o
re
a
.
He
re
c
e
iv
e
d
P
h
.
D.
in
C
o
m
p
u
ter
En
g
i
n
e
e
rin
g
fr
o
m
Un
i
v
e
rsity
o
f
Ulsa
n
,
S
o
u
th
K
o
re
a
,
M
.
S
c
.
in
El
e
c
tri
c
a
l
En
g
in
e
e
rin
g
(S
p
e
c
ializa
ti
o
n
:
Tele
c
o
m
m
u
n
ica
ti
o
n
s),
Ble
k
in
g
e
In
stit
u
te
o
f
Tec
h
n
o
lo
g
y
,
S
we
d
e
n
,
a
n
d
B
.
S
c
.
i
n
Co
m
p
u
ter
a
n
d
Co
m
m
u
n
ica
ti
o
n
En
g
i
n
e
e
rin
g
,
I
n
tern
a
ti
o
n
a
l
Isla
m
ic
Un
iv
e
rsity
C
h
it
tag
o
n
g
,
Ba
n
g
lad
e
sh
.
He
wa
s a
v
isit
in
g
fa
c
u
lt
y
a
t
S
c
h
o
o
l
o
f
c
o
m
p
u
ti
n
g
,
S
taff
o
rd
sh
i
re
Un
iv
e
rsit
y
,
Un
it
e
d
Kin
g
d
o
m
.
He
is
a
n
As
so
c
iate
P
ro
fe
ss
o
r
(n
o
w
o
n
lea
v
e
),
De
p
a
rtme
n
t
o
f
Co
m
p
u
ter
S
c
ien
c
e
a
n
d
En
g
i
n
e
e
rin
g
,
Bra
c
Un
iv
e
rsity
,
Dh
a
k
a
,
Ba
n
g
la
d
e
sh
.
His
re
se
a
rc
h
in
tere
sts
a
re
In
d
u
strial
F
a
u
lt
Dia
g
n
o
sis,
M
a
c
h
i
n
e
Lea
rn
in
g
/De
e
p
Lea
rn
in
g
b
a
se
d
p
re
d
icti
o
n
a
n
d
d
e
tec
ti
o
n
u
si
n
g
m
u
l
ti
m
e
d
ia
sig
n
a
ls.
Evaluation Warning : The document was created with Spire.PDF for Python.