I
AE
S In
t
er
na
t
io
na
l J
o
urna
l o
f
Art
if
icia
l In
t
ellig
ence
(
I
J
-
AI)
Vo
l.
9
,
No
.
1
,
Ma
r
ch
2
0
2
0
,
p
p
.
73
~
80
I
SS
N:
2
2
5
2
-
8938
,
DOI
: 1
0
.
1
1
5
9
1
/i
j
ai.
v
9
.i
1
.
p
p
73
-
80
73
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
i
.
ia
esco
r
e.
co
m
M
a
chine learning
appro
a
ch f
o
r
floo
d risk
s
predi
ctio
n
Na
zi
m
Ra
za
li
1
,
S
hu
ha
ida
I
s
m
a
il
2
,
Ai
da
M
us
t
a
ph
a
3
1,
3
F
a
c
u
lt
y
o
f
Co
m
p
u
ter S
c
ien
c
e
a
n
d
I
n
f
o
rm
a
ti
o
n
T
e
c
h
n
o
lo
g
y
,
Un
iv
e
rsiti
T
u
n
Hu
ss
e
in
O
n
n
M
a
lay
si
a
,
8
6
4
0
0
Ba
tu
P
a
h
a
t
,
Jo
h
o
r
,
M
a
lay
sia
.
2
F
a
c
u
lt
y
o
f
A
p
p
li
e
d
S
c
ien
c
e
s a
n
d
T
e
c
h
n
o
lo
g
y
,
Un
iv
e
rsiti
T
u
n
Hu
ss
e
in
On
n
M
a
lay
sia
,
8
6
4
0
0
Ba
tu
P
a
h
a
t,
Jo
h
o
r,
M
a
lay
sia
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Au
g
2
0
,
2
0
1
9
R
ev
i
s
ed
No
v
2
8
,
2
0
1
9
A
cc
ep
ted
De
c
1
6
,
2
0
1
9
F
lo
o
d
is
o
n
e
o
f
m
a
in
n
a
tu
ra
l
d
isa
ste
r
th
a
t
h
a
p
p
e
n
s a
ll
a
r
o
u
n
d
t
h
e
g
lo
b
e
c
a
u
se
d
law
o
f
n
a
tu
re
.
It
h
a
s
c
a
u
se
d
v
a
st
d
e
stru
c
ti
o
n
o
f
h
u
g
e
a
m
o
u
n
t
o
f
p
ro
p
e
rti
e
s,
li
v
e
sto
c
k
a
n
d
e
v
e
n
lo
ss
o
f
li
f
e
.
T
h
e
re
f
o
re
,
th
e
n
e
e
d
s
t
o
d
e
v
e
lo
p
a
n
a
c
c
u
ra
te
a
n
d
e
ff
icie
n
t
f
lo
o
d
risk
p
re
d
ictio
n
a
s
a
n
e
a
rl
y
w
a
rn
in
g
s
y
ste
m
is
h
ig
h
ly
e
ss
e
n
ti
a
l.
T
h
is
stu
d
y
a
i
m
s
to
d
e
v
e
lo
p
a
p
re
d
ictiv
e
m
o
d
e
ll
in
g
f
o
ll
o
w
Cro
ss
-
In
d
u
stry
S
tan
d
a
rd
P
r
o
c
e
ss
f
o
r
Da
ta
M
in
i
n
g
(CRIS
P
-
DM)
m
e
th
o
d
o
l
o
g
y
b
y
u
sin
g
Ba
y
e
sia
n
n
e
tw
o
rk
(BN)
a
n
d
o
t
h
e
r
M
a
c
h
in
e
L
e
a
rn
in
g
(M
L
)
tec
h
n
iq
u
e
s
su
c
h
a
s
De
c
isio
n
T
re
e
(DT
),
k
-
Ne
a
re
st
Ne
i
g
h
b
o
u
rs
(k
NN
)
a
n
d
S
u
p
p
o
rt
V
e
c
to
r
M
a
c
h
i
n
e
(S
V
M
)
f
o
r
f
lo
o
d
risk
s
p
re
d
ictio
n
i
n
Ku
a
la
Kra
i,
Ke
lan
tan
,
M
a
la
y
sia
.
T
h
e
d
a
ta
is
so
u
rc
e
d
f
r
o
m
5
-
y
e
a
r
p
e
rio
d
b
e
tw
e
e
n
2
0
1
2
u
n
ti
l
2
0
1
6
c
o
n
sistin
g
1
,
8
2
7
o
b
se
rv
a
ti
o
n
s.
T
h
e
p
e
r
f
o
r
m
a
n
c
e
o
f
e
a
c
h
m
o
d
e
ls
we
re
c
o
m
p
a
re
d
in
ter
m
s
o
f
a
c
c
u
ra
c
y
,
p
re
c
isio
n
,
re
c
a
ll
a
n
d
f
-
m
e
a
su
re
.
T
h
e
re
su
lt
s
sh
o
w
e
d
th
a
t
DT
w
it
h
S
M
OT
E
m
e
th
o
d
p
e
rf
o
r
m
e
d
th
e
b
e
st
c
o
m
p
a
re
d
t
o
o
th
e
rs
b
y
a
c
h
iev
in
g
9
9
.
9
2
%
a
c
c
u
ra
c
y
.
A
lso
,
S
M
OT
E
m
e
th
o
d
is
f
o
u
n
d
h
ig
h
ly
e
ffe
c
ti
v
e
in
d
e
a
li
n
g
w
it
h
i
m
b
a
lan
c
e
d
a
tas
e
t.
T
h
u
s,
it
is
h
o
p
e
d
th
a
t
th
e
f
in
d
in
g
o
f
th
is
re
se
a
r
c
h
m
a
y
a
s
sist
th
e
n
o
n
-
g
o
v
e
rn
m
e
n
t
o
r
g
o
v
e
rn
m
e
n
t
o
rg
a
n
iza
ti
o
n
to
tak
e
p
re
v
e
n
ti
v
e
a
c
ti
o
n
o
n
f
lo
o
d
p
h
e
n
o
m
e
n
o
n
th
a
t
c
o
m
m
o
n
ly
o
c
c
u
rs
in
M
a
la
y
sia
d
u
e
to
t
h
e
w
e
t
c
li
m
a
te.
K
ey
w
o
r
d
s
:
B
ay
e
s
ian
Net
w
o
r
k
Dec
is
io
n
T
r
ee
Flo
o
d
Pre
d
ictio
n
k
-
Nea
r
est Ne
i
g
h
b
o
u
r
Su
p
p
o
r
t V
ec
to
r
Ma
ch
in
e
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
:
Sh
u
h
a
id
a
I
s
m
ail
,
Facu
lt
y
o
f
A
p
p
lied
Scie
n
ce
s
a
n
d
T
ec
h
n
o
lo
g
y
,
Un
i
v
e
r
s
iti T
u
n
H
u
s
s
ei
n
On
n
Ma
la
y
s
ia,
8
6
4
0
0
B
atu
P
ah
at,
J
o
h
o
r
,
Ma
l
a
y
s
ia
.
E
m
ail: s
h
u
h
a
id
a@
u
t
h
m
.
ed
u
.
m
y
1.
I
NT
RO
D
UCT
I
O
N
Flo
o
d
is
th
e
m
o
s
t
co
m
m
o
n
n
atu
r
al
d
i
s
aster
t
h
at
h
ap
p
en
e
d
all
ar
o
u
n
d
th
e
w
o
r
ld
.
C
o
u
n
tr
ies
t
h
a
t
ex
p
er
ien
ce
f
lo
o
d
ev
en
t
n
ee
d
to
f
ac
e
a
h
u
g
e
o
f
a
m
o
u
n
t
p
r
o
p
er
ty
d
estru
ctio
n
,
en
v
ir
o
n
m
en
tal
an
d
f
i
n
a
n
cial
lo
s
s
es
ev
e
n
f
ata
liti
es
a
m
o
n
g
th
e
citizen
.
T
h
er
e
ar
e
s
ev
er
al
i
m
p
ac
tf
u
l
f
ac
to
r
s
th
at
af
f
ec
t
t
h
e
in
co
n
s
is
te
n
c
y
f
lo
o
d
o
cc
u
r
r
en
ce
s
u
ch
a
s
te
m
p
er
at
u
r
e,
h
u
m
id
it
y
,
d
e
w
p
o
i
n
t
te
m
p
er
atu
r
e,
w
in
d
s
p
ee
d
,
s
tr
ea
m
f
l
o
w
v
o
lu
m
e,
w
ater
lev
el
a
n
d
r
ain
f
a
ll
v
o
l
u
m
e.
T
h
e
s
tr
ea
m
f
lo
w
v
o
l
u
m
e
in
d
icate
s
th
e
r
iv
er
ca
p
ab
ilit
y
in
h
o
ld
in
g
th
e
w
ater
in
o
r
d
er
to
s
u
s
tai
n
t
h
e
r
ai
n
f
all
v
o
l
u
m
e.
Hig
h
er
te
m
p
er
at
u
r
e
an
d
w
i
n
d
s
p
ee
d
r
esu
lted
in
f
aster
w
at
er
p
ar
ticles
m
o
v
e
s
,
th
u
s
ea
s
ier
to
ev
ap
o
r
ate
in
to
th
e
at
m
o
s
p
h
er
e.
Hu
m
id
it
y
also
af
f
ec
ts
t
h
e
w
ater
p
ar
ticle
in
th
e
air
to
b
e
co
n
d
en
s
ed
o
u
t
o
f
th
e
at
m
o
s
p
h
er
e.
P
r
ev
io
u
s
s
t
u
d
ies
o
n
f
lo
o
d
d
etec
tio
n
h
as
b
ee
n
d
o
n
e
d
u
e
to
d
is
astro
u
s
ev
e
n
t
o
cc
u
r
in
co
u
n
tr
ies
s
u
c
h
as
in
d
is
astro
u
s
f
lo
o
d
ev
en
t
o
cc
u
r
in
Au
s
tr
alia
[
1
]
,
Ma
lay
s
ia
[
2
]
,
I
n
d
ia
[
3
]
an
d
m
a
n
y
o
th
er
co
u
n
tr
ies.
Ma
la
y
s
ia
lo
ca
ted
g
eo
g
r
ap
h
ica
ll
y
n
ea
r
th
e
eq
u
ato
r
ial
li
n
e
w
h
ic
h
p
r
ev
e
n
t
it
f
r
o
m
s
e
v
er
e
n
atu
r
a
l
d
is
aster
p
h
en
o
m
en
o
n
s
u
c
h
as
ea
r
th
q
u
ak
e
s
,
v
o
lcan
ic
er
u
p
ti
o
n
an
d
t
y
p
h
o
o
n
s
.
Ho
w
e
v
er
,
Ma
la
y
s
ia
e
x
p
er
ien
ce
h
o
t
an
d
h
u
m
id
w
ea
t
h
er
b
y
a
v
er
ag
e
d
ail
y
te
m
p
er
at
u
r
e
o
f
2
1
°C
to
3
2
°
C
t
h
r
o
u
g
h
o
u
t
t
h
e
y
ea
r
s
in
ce
Ma
la
y
s
ia
in
f
lu
e
n
ce
d
h
ea
v
il
y
b
y
eq
u
ato
r
i
al
lin
e.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8938
I
n
t J
A
r
ti
f
I
n
tell
,
Vo
l.
9
,
No
.
1
,
Ma
r
ch
20
20
:
73
–
80
74
I
n
ad
d
itio
n
,
Ma
lay
s
ia
also
ex
p
er
ien
ce
cli
m
ate
ch
a
n
g
es
o
f
n
o
r
th
ea
s
ter
n
m
o
n
s
o
o
n
f
r
o
m
No
v
e
m
b
er
to
Ma
r
ch
an
d
t
h
e
w
ester
n
m
o
n
s
o
o
n
f
r
o
m
J
u
n
e
to
Octo
b
er
.
Du
e
to
m
o
n
s
o
o
n
s
ea
s
o
n
,
t
h
e
an
n
u
al
r
ai
n
f
al
l
in
Ma
la
y
s
ia
q
u
ite
h
i
g
h
b
y
2
5
0
0
m
m
i
n
P
en
i
n
s
u
lar
Ma
la
y
s
ia,
2
3
0
0
m
m
i
n
Sar
a
w
a
k
a
n
d
3
3
0
0
m
m
i
n
Sab
ah
.
Ho
w
e
v
er
,
m
o
n
s
o
o
n
i
n
P
en
i
n
s
u
lar
Ma
la
y
s
ia
co
n
tr
ib
u
te
s
8
6
%
f
r
o
m
th
e
a
n
n
u
a
l r
ain
f
all
in
ea
s
t c
o
ast o
f
Ma
la
y
s
ia
co
n
s
is
tin
g
t
h
e
s
tates
o
f
Kela
n
tan
,
T
er
en
g
g
a
n
u
a
n
d
P
ah
a
n
g
.
T
h
e
h
ea
v
y
r
ai
n
f
all
u
s
u
all
y
r
es
u
l
ted
in
f
lo
o
d
w
h
et
h
er
n
atu
r
al
o
r
f
las
h
f
lo
o
d
.
Fo
r
in
s
tan
ce
,
t
h
e
d
is
astro
u
s
f
lo
o
d
ev
en
t
o
cc
u
r
r
ed
in
2
0
1
4
[
4
]
h
as
g
iv
e
n
i
m
p
o
r
tan
t
le
s
s
o
n
o
f
h
a
v
i
n
g
f
lo
o
d
p
r
ed
ictio
n
s
y
s
te
m
to
m
o
n
it
o
r
,
p
r
ed
ict,
an
d
d
etec
t
th
e
f
lo
o
d
ev
en
t.
I
n
o
r
d
er
to
r
ed
u
ce
s
u
c
h
d
a
m
a
g
es,
an
ea
r
l
y
is
s
u
ed
f
lo
o
d
w
ar
n
i
n
g
i
s
ess
e
n
tial.
T
h
u
s
,
w
ater
le
v
el
f
o
r
ec
a
s
tin
g
is
e
s
s
e
n
tial
to
p
r
ed
ict
f
u
t
u
r
e
f
lo
o
d
o
cc
u
r
r
en
c
e.
W
ater
lev
el
p
r
ed
ictio
n
also
b
en
ef
it
s
o
t
h
er
s
ec
to
r
s
s
u
ch
as
ag
r
icu
l
tu
r
e,
p
la
n
ts
,
d
o
m
e
s
tics
a
n
d
in
d
u
s
tr
ia
l
an
d
co
m
m
er
cial
[
5
]
.
T
h
e
aim
o
f
th
i
s
r
esear
ch
p
ap
er
is
to
d
e
v
elo
p
a
p
r
e
d
ictiv
e
m
o
d
ell
in
g
w
h
ic
h
f
o
llo
w
C
r
o
s
s
-
I
n
d
u
s
tr
y
Sta
n
d
ar
d
P
r
o
ce
s
s
f
o
r
Data
Min
in
g
(
C
R
I
SP
-
DM
)
m
eth
o
d
o
lo
g
y
b
y
u
s
i
n
g
B
ay
e
s
ia
n
n
et
w
o
r
k
(
B
N)
an
d
o
th
er
Ma
ch
in
e
L
ea
r
n
in
g
(
ML
)
tech
n
iq
u
es
s
u
c
h
as
Dec
is
io
n
T
r
ee
(
D
T
)
,
k
-
Nea
r
est
Neig
h
b
o
u
r
s
(
k
NN)
a
n
d
Su
p
p
o
r
t
Vec
to
r
Ma
ch
in
e
(
SVM)
f
o
r
f
lo
o
d
r
is
k
s
p
r
ed
ictio
n
in
Ku
ala
Kr
ai,
Kela
n
ta
n
,
Ma
la
y
s
ia.
T
h
e
r
e
m
a
in
i
n
g
o
f
th
is
p
ap
er
is
o
r
g
an
ize
d
as
f
o
llo
w
s
.
Sectio
n
2
r
ev
ie
ws
all
w
o
r
k
s
r
elate
d
to
tech
n
iq
u
e
s
u
s
ed
f
o
r
f
lo
o
d
r
is
k
p
r
ed
ictio
n
.
Sec
tio
n
3
p
r
ese
n
t
s
t
h
e
d
ata
m
in
i
n
g
m
et
h
o
d
o
lo
g
y
as
w
ell
a
s
d
ataset
p
r
e
-
p
r
o
ce
s
s
i
n
g
,
ex
p
er
i
m
en
tal
s
etu
p
,
a
n
d
t
h
e
e
v
al
u
a
tio
n
m
etr
ics.
Sectio
n
4
p
r
ese
n
ts
t
h
e
r
es
u
lt
s
a
n
d
f
i
n
all
y
Sectio
n
5
co
n
cl
u
d
es
w
i
th
s
o
m
e
d
ir
ec
tio
n
s
f
o
r
f
u
t
u
r
e
w
o
r
k
.
2.
RE
L
AT
E
D
WO
RK
A
p
p
licatio
n
o
f
B
N
an
d
o
th
er
ML
tech
n
iq
u
es
in
p
r
ed
ictio
n
a
n
d
class
i
f
icatio
n
h
as
b
ee
n
u
s
e
d
w
id
el
y
i
n
m
an
y
f
ield
i
n
cl
u
d
in
g
ag
r
ic
u
lt
u
r
e,
ec
o
n
o
m
y
,
a
n
d
etc
in
Ma
la
y
s
ia.
[
6
]
Used
ML
tec
h
n
iq
u
es
s
u
ch
a
s
an
A
r
tific
ial
Neu
r
al
Net
w
o
r
k
(
A
NN)
,
K
-
N
ea
r
est
Nei
g
h
b
o
u
r
s
(
k
NN)
,
De
cisi
o
n
T
ab
le
(
DT
)
an
d
M5
P
T
r
ee
alg
o
r
ith
m
s
i
n
th
eir
r
esear
ch
to
class
i
f
y
h
er
b
s
f
o
r
ag
r
icu
lt
u
r
e
in
d
u
s
tr
y
i
n
Ma
la
y
s
ia
s
in
ce
t
h
is
i
n
d
u
s
tr
y
i
s
cr
u
cial
to
ass
is
t
t
h
e
ec
o
n
o
m
y
d
e
v
elo
p
m
e
n
t
o
f
Ma
la
y
s
ia
a
s
o
n
e
o
f
lead
in
g
ex
p
o
r
ter
o
f
h
er
b
s
.
Me
an
w
h
ile,
[
7
]
h
as
p
r
o
p
o
s
ed
n
e
w
Hala
l
tech
n
o
lo
g
i
es
u
s
in
g
M
L
tech
n
iq
u
e
to
f
ac
ili
tate
th
e
M
u
s
li
m
co
n
s
u
m
er
s
i
n
Ma
la
y
s
ia
to
au
th
e
n
ticate
t
h
e
Hala
l
lo
g
o
i
m
a
g
e
n
o
t
j
u
s
t
lo
ca
ll
y
b
u
t
al
s
o
g
lo
b
all
y
a
s
lo
n
g
it
w
er
e
r
ec
o
g
n
is
ed
b
y
t
h
e
Ma
la
y
s
ia
n
’
s
Dep
ar
t
m
en
t
o
f
I
s
la
m
ic
De
v
elo
p
m
e
n
t
(
J
A
K
I
M)
.
Ho
w
e
v
er
,
r
ec
en
t
w
o
r
k
o
f
ap
p
licatio
n
o
f
B
N
an
d
ML
tech
n
iq
u
es
h
av
e
b
ee
n
u
s
ed
w
id
el
y
f
o
cu
s
i
n
g
o
n
n
at
u
r
al
d
is
aster
d
etec
tio
n
s
u
ch
a
s
f
l
o
o
d
r
is
k
s
d
etec
tio
n
.
I
n
2
0
1
8
,
[
2
]
h
as
ca
r
r
ied
o
u
t
r
esear
ch
o
n
p
r
ed
ictin
g
f
lo
o
d
r
is
k
s
u
s
i
n
g
B
a
y
esia
n
ap
p
r
o
ac
h
es.
T
h
e
y
co
n
d
u
cted
e
x
p
er
i
m
e
n
t
u
s
in
g
t
h
r
ee
B
a
y
e
s
ian
clas
s
i
f
ier
al
g
o
r
ith
m
s
n
a
m
el
y
g
e
n
er
al
B
a
y
e
s
i
an
Net
w
o
r
k
s
,
n
ai
v
e
B
ay
e
s
an
d
T
r
ee
A
u
g
m
e
n
ted
Naiv
e
B
a
y
es
to
p
r
ed
ict
th
e
f
lo
o
d
r
is
k
s
in
Ku
a
la
Kr
ai,
Kela
n
tan
,
Ma
la
y
s
ia
f
o
r
5
-
y
ea
r
s
.
T
h
e
r
esu
lts
s
h
o
w
ed
th
a
t
g
en
er
al
B
a
y
esia
n
Net
w
o
r
k
s
s
u
cc
e
s
s
f
u
ll
y
o
u
tp
er
f
o
r
m
ed
b
o
t
h
Naiv
e
B
a
y
es
a
n
d
tr
ee
au
g
m
e
n
ted
n
aiv
e
B
a
y
es
i
n
ter
m
o
f
ac
c
u
r
ac
y
.
T
h
is
p
ap
er
h
as
b
ee
n
u
s
ed
as
o
u
r
an
c
h
o
r
p
ap
er
to
co
n
d
u
ct
f
u
r
t
h
er
r
esear
c
h
b
y
co
m
p
ar
i
n
g
B
a
y
e
s
ia
n
n
et
w
o
r
k
(
B
N)
w
i
th
o
t
h
er
M
L
tech
n
iq
u
e
s
s
u
c
h
as
Dec
is
io
n
T
r
ee
(
DT
)
,
k
-
Nea
r
e
s
t
Nei
g
h
b
o
u
r
s
(
k
NN
)
a
n
d
S
u
p
p
o
r
t
Vec
to
r
Ma
ch
in
e
(
SVM)
.
[
8
]
P
r
o
p
o
s
ed
an
ea
r
l
y
p
r
ed
ictio
n
s
y
s
te
m
u
s
i
n
g
A
u
to
r
eg
r
ess
i
v
e
Neu
r
al
Net
w
o
r
k
s
w
i
th
E
x
o
g
e
n
o
u
s
I
n
p
u
t
(
NN
AR
X)
f
o
r
5
-
h
o
u
r
ah
ea
d
f
o
r
f
lo
o
d
.
W
ater
lev
el
an
d
r
ain
f
all
f
o
r
v
ar
io
u
s
s
tatio
n
s
lo
ca
ted
in
Kela
n
tan
,
Ma
la
y
s
ia
w
er
e
o
b
s
er
v
e
d
.
T
h
e
p
er
f
o
r
m
an
ce
o
f
p
r
o
p
o
s
ed
NNA
R
X
w
er
e
co
m
p
ar
ed
w
it
h
co
n
v
e
n
tio
n
al
Neu
r
al
Net
w
o
r
k
f
o
r
p
r
ed
ictio
n
p
er
f
o
r
m
a
n
ce
.
T
h
e
r
esu
lt
s
s
h
o
w
ed
t
h
at
NN
AR
X
h
as
s
m
a
lles
t
v
al
u
e
o
f
R
o
o
t
Me
an
Sq
u
ar
e
E
r
r
o
r
(
R
MSE
)
co
m
p
ar
ed
to
co
n
v
e
n
tio
n
al
Ne
u
r
al
Net
w
o
r
k
s
.
B
esid
e
s
,
[
9
]
also
f
o
cu
s
in
g
o
n
f
lo
o
d
ev
en
t
i
n
Kela
n
ta
n
,
Ma
la
y
s
ia.
T
h
e
y
p
r
o
p
o
s
ed
Sp
ik
in
g
Neu
r
al
Net
w
o
r
k
to
p
r
ed
ict
th
e
f
lo
o
d
r
is
k
ev
e
n
t.
Me
a
n
w
h
ile,
[
1
0
]
h
as
i
n
tr
o
d
u
ce
d
a
s
e
m
i
-
s
u
p
er
v
i
s
ed
ML
m
o
d
el,
w
h
ic
h
is
W
ea
k
l
y
L
ab
elled
SVM
(
W
E
L
L
SVM)
to
p
r
ed
ict
u
r
b
an
f
lo
o
d
b
ased
o
n
d
ata
s
a
m
p
l
es
co
llected
f
r
o
m
u
r
b
an
ar
ea
s
in
B
eij
in
g
f
o
r
1
0
-
y
ea
r
b
et
w
ee
n
2
0
0
4
to
2
0
1
4
.
T
h
e
s
a
m
p
les
co
n
s
i
s
ted
o
f
n
in
e
d
o
m
in
a
n
t
f
ac
to
r
s
o
f
m
etr
o
lo
g
ical,
g
eo
g
r
ap
h
ical
a
n
d
an
t
h
r
o
p
o
g
en
ic.
T
h
e
m
o
d
el
w
er
e
t
h
en
e
v
al
u
ated
an
d
co
m
p
ar
ed
w
it
h
o
th
er
t
w
o
m
o
d
el
b
u
ilt
f
r
o
m
L
o
g
i
s
tic
R
e
g
r
ess
io
n
an
d
A
r
ti
f
icial
Neu
r
al
N
et
w
o
r
k
s
i
n
ter
m
s
o
f
ac
cu
r
ac
y
,
p
r
ec
is
io
n
,
r
ec
all
an
d
f
-
s
co
r
e.
T
h
e
r
esu
lts
s
h
o
w
ed
t
h
at
W
E
L
L
S
VM
f
lo
o
d
s
u
cc
e
s
s
f
u
ll
y
o
u
tp
er
f
o
r
m
ed
b
o
th
m
o
d
els b
ec
au
s
e
W
E
L
L
S
VM
h
as t
h
e
ad
v
a
n
ta
g
e
in
u
tili
zin
g
t
h
e
u
n
lab
elled
d
ata.
I
n
s
o
m
e
o
t
h
er
ca
s
e
s
,
[
1
1
]
p
r
o
p
o
s
ed
a
m
o
d
el
b
ased
o
n
e
n
s
e
m
b
le
c
lass
if
ier
f
o
r
m
o
r
e
p
r
ec
is
e
w
ate
r
f
lo
o
d
ed
la
y
er
r
ec
o
g
n
itio
n
w
h
i
ch
m
ea
n
s
th
e
tar
g
et
clas
s
es
ar
e
d
iv
id
ed
i
n
to
f
o
u
r
tar
g
et
cla
s
s
e
s
,
w
h
ich
ar
e
t
h
e
o
i
l
la
y
er
,
w
ea
k
w
ater
f
lo
o
d
ed
,
m
i
d
d
le
w
ater
f
lo
o
d
ed
an
d
s
tr
o
n
g
w
ater
f
lo
o
d
ed
.
I
n
ter
esti
n
g
l
y
,
t
h
is
m
o
d
el
w
as
u
s
ed
to
p
r
ed
ict
th
e
w
ater
f
lo
o
d
ed
lay
er
in
o
il
o
r
g
as
r
eser
v
o
ir
.
T
h
e
en
s
e
m
b
le
clas
s
i
f
ier
w
er
e
m
ad
e
u
p
o
f
th
e
m
o
d
el
-
f
r
ee
class
i
f
icatio
n
(
MFB
C
)
alg
o
r
ith
m
,
th
e
k
-
Nea
r
e
s
t
Neig
h
b
o
u
r
s
(
k
NN)
alg
o
r
ith
m
an
d
t
h
e
Su
p
p
o
r
t
Vec
to
r
M
ac
h
i
n
e
(
SVM)
al
g
o
r
ith
m
w
h
ich
w
er
e
t
h
en
v
alid
ated
an
d
e
v
alu
a
ted
.
T
h
e
d
ataset
w
e
n
t
t
h
r
o
u
g
h
o
v
er
s
a
m
p
li
n
g
p
r
o
ce
s
s
u
s
in
g
S
y
n
t
h
etic
Mi
n
o
r
it
y
O
v
er
s
a
m
p
l
in
g
T
ec
h
n
iq
u
e
(
SMOT
E
)
d
u
e
to
i
m
b
ala
n
ce
class
es.
T
h
e
r
es
u
lt
s
s
h
o
w
ed
th
at
t
h
e
en
s
e
m
b
le
cla
s
s
i
f
ier
p
er
f
o
r
m
ed
b
ette
r
as
co
m
p
ar
ed
to
MFB
C
,
KNN
an
d
SVM
f
o
r
b
o
th
UC
I
d
ata
an
d
ch
r
o
m
ato
g
r
a
m
d
ata
w
h
ile
all
th
r
ee
MFB
C
,
KNN
an
d
SVM
w
er
e
s
i
m
ilar
in
ac
c
u
r
ac
y
f
o
r
9
0
%
an
d
7
0
.
5
9
%
r
esp
ec
tiv
el
y
.
Fi
n
all
y
,
th
e
d
estr
u
cti
v
e
f
lo
o
d
ev
en
t
at
Au
s
tr
alia
i
n
2
0
1
1
h
as
u
r
g
ed
[
1
]
to
u
s
e
t
w
o
M
L
ap
p
r
o
ac
h
es;
DT
an
d
SVM
to
ev
alu
a
te
s
p
atial
co
r
r
elatio
n
s
b
et
w
ee
n
th
e
co
n
tr
ib
u
ti
n
g
f
ac
to
r
s
to
f
lo
o
d
an
d
r
ate
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
A
r
ti
f
I
n
tell
I
SS
N:
2252
-
8938
Ma
ch
in
e
lea
r
n
in
g
a
p
p
r
o
a
ch
fo
r
flo
o
d
r
is
ks p
r
ed
ictio
n
(
N
a
z
i
m
R
a
z
a
li
)
75
th
e
f
ac
to
r
s
ac
co
r
d
in
g
to
t
h
eir
i
m
p
o
r
tan
ce
i
n
m
ap
p
in
g
t
h
e
f
l
o
o
d
p
r
o
n
e
ar
ea
s
.
T
h
e
r
esu
lt
s
s
h
o
w
ed
th
at
DT
w
a
s
s
lig
h
tl
y
b
etter
in
ac
cu
r
ac
y
as
co
m
p
ar
ed
to
SV
M
w
h
er
eb
y
DT
ac
h
iev
ed
8
9
%
an
d
SVM
ac
h
iev
ed
8
7
%
in
t
h
e
f
ir
s
t
d
ataset
an
d
lo
w
in
ac
cu
r
ac
y
as
co
m
p
ar
ed
to
SVM
in
th
e
s
ec
o
n
d
d
ataset
b
y
DT
(
8
9
%)
an
d
SVM
(
8
7
%)
r
esp
ec
tiv
el
y
.
3.
RE
S
E
ARCH
M
E
T
H
O
D
T
h
is
p
r
o
j
ec
t
a
d
o
p
te
d
th
e
C
r
o
s
s
-
I
n
d
u
s
tr
y
Sta
n
d
ar
d
Pro
ce
s
s
f
o
r
Data
Min
in
g
(
C
R
I
SP
-
DM
)
[
1
2
]
.
T
h
is
m
et
h
o
d
o
lo
g
y
d
iv
id
ed
d
ata
m
i
n
in
g
ta
s
k
i
n
to
s
i
x
p
h
a
s
es
a
s
s
h
o
w
n
i
n
Fi
g
u
r
e
1
.
T
h
e
s
ix
p
h
ase
s
co
m
p
r
is
es
o
f
b
u
s
i
n
ess
u
n
d
er
s
ta
n
d
i
n
g
,
d
ata
u
n
d
er
s
ta
n
d
in
g
,
d
ata
p
r
ep
ar
ati
o
n
,
m
o
d
ellin
g
,
ev
al
u
atio
n
a
n
d
d
ep
lo
y
m
e
n
t.
T
h
is
C
R
I
SP
-
DM
h
as
b
ec
o
m
e
a
b
en
ch
m
ar
k
o
r
s
tan
d
ar
d
m
et
h
o
d
o
lo
g
y
to
b
e
f
o
llo
w
i
n
d
ata
m
i
n
in
g
tas
k
p
r
o
j
ec
t
co
m
p
let
io
n
.
E
ac
h
o
f
t
h
e
p
h
a
s
es
i
n
C
R
I
SP
-
DM
w
i
ll
p
r
o
d
u
ce
o
u
tp
u
t
th
a
t
b
en
e
f
it
s
t
h
e
f
l
o
w
o
f
d
ata
m
i
n
in
g
p
r
o
j
ec
t
as
w
h
o
le.
I
t
m
a
k
e
t
h
e
p
r
o
j
ec
t
m
o
r
e
f
lex
ib
le
a
n
d
e
f
f
icien
t
to
s
o
lv
e
b
u
s
i
n
es
s
is
s
u
es
u
s
i
n
g
an
a
l
y
t
ics.
T
ab
le
1
d
escr
ib
e
d
in
b
r
ief
th
e
s
ix
m
aj
o
r
s
tep
s
in
C
R
I
SP
-
DM
.
F
ig
u
re
1
.
CRI
S
P
-
DM
m
e
th
o
d
o
l
o
g
y
[
1
2
]
.
T
ab
le
1
.
C
R
I
SP
-
DM
m
et
h
o
d
o
lo
g
y
C
R
I
S
P
-
D
M
st
e
p
s
D
e
scri
p
t
i
o
n
1
.
B
u
si
n
e
ss
U
n
d
e
r
st
a
n
d
i
n
g
F
o
c
u
se
s
o
n
u
n
d
e
r
st
a
n
d
i
n
g
t
h
e
r
e
se
a
r
c
h
o
b
j
e
c
t
i
v
e
s
a
n
d
r
e
q
u
i
r
e
me
n
t
s,
a
n
d
t
h
e
n
c
o
n
v
e
r
t
i
n
g
t
h
i
s
k
n
o
w
l
e
d
g
e
i
n
t
o
a
d
a
t
a
mi
n
i
n
g
p
r
o
b
l
e
m d
e
f
i
n
i
t
i
o
n
.
2
.
D
a
t
a
U
n
d
e
r
st
a
n
d
i
n
g
F
o
c
u
se
s
o
n
d
a
t
a
c
o
l
l
e
c
t
i
o
n
,
a
n
d
p
r
o
c
e
e
d
w
i
t
h
i
n
v
e
st
i
g
a
t
i
n
g
a
n
d
st
u
d
y
i
n
g
t
h
e
d
a
t
a
t
o
i
d
e
n
t
i
f
y
d
a
t
a
q
u
a
l
i
t
y
p
r
o
b
l
e
ms,
t
o
d
i
sco
v
e
r
f
i
r
st
i
n
s
i
g
h
t
s
i
n
t
o
t
h
e
d
a
t
a
,
o
r
t
o
d
e
t
e
c
t
i
n
t
e
r
e
st
i
n
g
su
b
se
t
s
t
o
f
o
r
m h
y
p
o
t
h
e
se
s fo
r
h
i
d
d
e
n
i
n
f
o
r
ma
t
i
o
n
.
3
.
D
a
t
a
P
r
e
p
a
r
a
t
i
o
n
T
h
e
d
a
t
a
p
r
e
p
a
r
a
t
i
o
n
p
h
a
se
o
r
a
l
so
k
n
o
w
n
a
s
d
a
t
a
p
r
e
p
r
o
c
e
ssi
n
g
c
o
v
e
r
s
a
l
l
a
c
t
i
v
i
t
i
e
s
t
o
c
o
n
st
r
u
c
t
t
h
e
f
i
n
a
l
d
a
t
a
se
t
f
r
o
m
t
h
e
i
n
i
t
i
a
l
r
a
w
d
a
t
a
a
n
d
t
o
e
n
su
r
e
t
h
e
d
a
t
a
b
e
u
se
d
i
s
i
mp
r
o
v
e
d
i
n
q
u
a
l
i
t
y
a
n
d
a
c
c
e
p
t
a
b
l
e
b
e
f
o
r
e
mo
d
e
l
l
i
n
g
p
h
a
se
.
I
n
c
l
u
d
e
c
l
e
a
n
si
n
g
,
t
r
a
n
sf
o
r
mat
i
o
n
,
d
i
scre
t
i
z
a
t
i
o
n
,
r
e
d
u
c
t
i
o
n
a
n
d
f
e
a
t
u
r
e
e
n
g
i
n
e
e
r
i
n
g.
4
.
M
o
d
e
l
i
n
g
M
o
d
e
l
l
i
n
g
t
e
c
h
n
i
q
u
e
s
su
c
h
a
s
mac
h
i
n
e
l
e
a
r
n
i
n
g
a
l
g
o
r
i
t
h
ms
a
r
e
se
l
e
c
t
e
d
a
n
d
a
p
p
l
i
e
d
.
C
a
n
b
e
l
o
o
p
b
a
c
k
t
o
d
a
t
a
p
r
e
p
a
r
a
t
i
o
n
p
h
a
se
a
c
c
o
r
d
i
n
g
l
y
t
o
su
i
t
a
b
i
l
i
t
y
o
f
d
a
t
a
se
t
s
w
i
t
h
a
p
p
l
i
e
d
a
l
g
o
r
i
t
h
ms
5
.
Ev
a
l
u
a
t
i
o
n
F
o
c
u
se
s
o
n
e
v
a
l
u
a
t
e
a
n
d
v
a
l
i
d
a
t
e
t
h
e
mo
d
e
l
s
t
h
a
t
h
a
v
e
b
e
e
n
b
u
i
l
t
f
o
r
me
a
s
u
r
i
n
g
t
h
e
q
u
a
l
i
t
y
a
n
d
p
e
r
f
o
r
man
c
e
o
f
t
h
e
mo
d
e
l
s
c
o
n
si
d
e
r
i
n
g
t
h
e
o
b
j
e
c
t
i
v
e
s
a
n
d
r
e
q
u
i
r
e
me
n
t
s
.
T
h
e
m
o
d
e
l
w
h
i
c
h
su
c
c
e
ssf
u
l
l
y
o
b
t
a
i
n
e
d
t
h
e
h
i
g
h
e
st
q
u
a
l
i
t
y
a
n
d
p
e
r
f
o
r
man
c
e
w
i
l
l
b
e
sel
e
c
t
e
d
a
s e
n
d
p
r
o
d
u
c
t
.
6
.
D
e
p
l
o
y
me
n
t
T
h
i
s
l
a
st
p
h
a
se
i
s
t
o
d
e
p
l
o
y
t
h
e
e
n
d
p
r
o
d
u
c
t
t
o
b
e
a
p
p
l
i
e
d
i
n
r
e
a
l
w
o
r
l
d
si
t
u
a
t
i
o
n
.
3
.
1
.
Da
t
a
s
et
A
ll
t
h
e
d
ataset
w
er
e
ex
tr
ac
ted
f
r
o
m
[
1
3
-
1
4
]
w
h
ich
ar
e
5
y
e
ar
p
er
io
d
r
ec
o
r
d
s
o
f
f
lo
o
d
d
ata
in
Ku
al
a
Kr
ai,
Kela
n
ta
n
,
Ma
la
y
s
ia
b
et
w
ee
n
1
s
t
J
an
u
ar
y
2
0
1
2
u
n
t
il
3
1
s
t
Dec
e
m
b
er
2
0
1
6
co
n
s
is
ti
n
g
o
f
1
,
8
2
7
in
s
tan
ce
s
an
d
8
f
ea
tu
r
es
in
cl
u
d
in
g
d
ate,
r
ain
f
all
m
o
n
t
h
l
y
,
r
ain
f
all
d
a
il
y
,
w
ater
lev
el,
h
u
m
id
it
y
,
w
i
n
d
an
d
th
e
b
in
ar
y
tar
g
et
clas
s
w
h
eth
er
f
lo
o
d
o
r
n
o
t
w
h
ich
ar
e
co
r
r
esp
o
n
d
f
ea
tu
r
es
f
o
r
f
lo
o
d
r
is
k
s
p
r
ed
ictio
n
.
No
te
th
at
’
d
ate
’
f
ea
t
u
r
e
w
er
e
n
o
t
i
n
clu
d
ed
in
t
h
e
ex
p
er
i
m
e
n
tal
p
r
o
ce
s
s
s
i
n
ce
it
co
n
tain
u
n
iq
u
e
v
al
u
e
f
o
r
ea
ch
in
s
tan
ce
s
w
h
ic
h
n
o
t
g
a
v
e
a
n
y
s
i
g
n
if
ican
t
i
m
p
a
ct
to
lear
n
i
n
g
p
r
o
ce
s
s
.
T
h
e
p
a
r
t
o
f
s
a
m
p
le
d
ata
ac
co
r
d
in
g
l
y
to
th
e
f
ea
t
u
r
es
ar
e
s
h
o
w
n
in
T
ab
le
2
.
T
ab
le
2
.
E
x
ce
r
p
t o
f
k
u
ala
k
r
ai
f
lo
o
d
d
ata
f
o
r
5
y
ea
r
p
er
io
d
D
a
t
e
L
e
v
e
l
(
c
m)
R
F
M
o
n
t
h
(
mm
)
R
F
D
a
i
l
y
(
mm
)
T
e
mp
e
r
a
t
u
r
e
(
?
C
)
H
u
mi
d
i
t
y
(
%)
W
i
n
d
(
m
/
s)
c
l
a
ss
0
1
/
0
1
/
1
2
1
8
7
1
1
0
5
7
45
2
4
.
2
9
2
.
8
0
.
7
N
O
F
L
O
O
D
0
1
/
0
2
/
1
2
1
9
1
1
1
0
5
8
1
2
4
.
1
9
2
.
8
0
.
6
N
O
F
L
O
O
D
0
1
/
0
3
/
1
2
1
7
9
9
1
0
6
4
6
2
4
.
7
9
1
.
2
0
.
7
N
O
F
L
O
O
D
0
1
/
0
4
/
1
2
1
7
6
3
1
0
6
4
0
25
8
2
.
8
0
.
9
N
O
F
L
O
O
D
0
1
/
0
5
/
1
2
1
7
3
8
1
0
6
4
0
2
4
.
3
8
3
.
7
1
N
O
F
L
O
O
D
0
1
/
0
6
/
1
2
1
7
2
1
1
0
6
4
0
2
4
.
6
8
0
.
5
0
.
9
N
O
F
L
O
O
D
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8938
I
n
t J
A
r
ti
f
I
n
tell
,
Vo
l.
9
,
No
.
1
,
Ma
r
ch
20
20
:
73
–
80
76
T
ab
le
2
.
E
x
ce
r
p
t o
f
k
u
ala
k
r
ai
f
lo
o
d
d
ata
f
o
r
5
y
ea
r
p
er
io
d
(
C
o
n
tin
u
e)
D
a
t
e
L
e
v
e
l
(
c
m)
R
F
M
o
n
t
h
(
mm
)
R
F
D
a
i
l
y
(
mm
)
T
e
mp
e
r
a
t
u
r
e
(
?
C
)
H
u
mi
d
i
t
y
(
%)
W
i
n
d
(
m
/
s)
c
l
a
ss
0
1
/
0
7
/
1
2
1
7
1
1
1
0
6
4
0
2
4
.
3
8
2
.
1
1
.
1
N
O
F
L
O
O
D
0
1
/
0
8
/
1
2
1
7
0
3
1
0
7
1
7
2
4
.
6
8
5
.
5
0
.
6
N
O
F
L
O
O
D
0
1
/
0
9
/
1
2
1
7
0
3
1
0
7
1
0
2
4
.
2
89
0
.
6
N
O
F
L
O
O
D
0
1
/
1
0
/
1
2
1
7
5
5
1
0
7
8
7
2
4
.
9
9
2
.
1
1
.
1
N
O
F
L
O
O
D
0
1
/
1
1
/
1
2
1
8
1
8
1
0
8
4
6
2
4
.
5
8
9
.
9
0
.
9
N
O
F
L
O
O
D
0
1
/
1
2
/
1
2
2
0
8
2
1
1
0
3
18
2
4
.
4
9
2
.
5
1
.
4
N
O
F
L
O
O
D
0
1
/
1
3
/
1
2
2
5
0
1
1
1
4
3
33
2
3
.
6
9
5
.
9
0
.
4
F
L
O
O
D
0
1
/
1
4
/
1
2
2
5
4
3
1
1
5
3
9
24
9
4
.
8
0
.
6
F
L
O
O
D
0
1
/
1
5
/
1
2
2
2
3
9
1
1
5
3
0
2
6
.
2
8
6
.
5
0
.
6
F
L
O
O
D
0
1
/
1
6
/
1
2
1
9
5
5
1
1
5
6
3
2
6
.
4
8
7
.
4
1
N
O
F
L
O
O
D
0
1
/
1
7
/
1
2
1
8
6
3
1
1
5
6
0
2
6
.
6
8
4
.
6
0
.
9
N
O
F
L
O
O
D
0
1
/
1
8
/
1
2
1
9
2
3
1
1
6
8
12
2
5
.
7
8
8
.
8
0
.
8
N
O
F
L
O
O
D
0
1
/
1
9
/
1
2
1
9
1
6
1
1
7
4
6
2
6
.
2
8
8
.
1
1
N
O
F
L
O
O
D
0
1
/
2
0
/
1
2
1
9
9
1
1
2
7
5
1
0
1
2
5
.
4
8
8
.
3
0
.
7
N
O
F
L
O
O
D
0
1
/
2
1
/
1
2
1
9
0
4
1
2
7
5
0
2
6
.
1
8
4
.
9
1
N
O
F
L
O
O
D
0
1
/
2
2
/
1
2
1
8
0
6
1
2
7
5
0
26
8
4
.
1
1
N
O
F
L
O
O
D
0
1
/
2
3
/
1
2
1
8
0
6
1
2
7
5
0
2
6
.
2
8
0
.
8
0
.
9
N
O
F
L
O
O
D
3
.
2
.
E
x
peri
m
ent
a
l s
et
up
A
ll
t
h
e
B
a
y
e
s
ian
Ne
t
w
o
r
k
s
a
n
d
m
ac
h
in
e
lear
n
i
n
g
(
M
L
)
al
g
o
r
ith
m
s
u
s
ed
i
n
th
i
s
r
esear
c
h
s
u
ch
a
s
Dec
is
io
n
T
r
ee
s
(
DT
)
,
k
-
Nea
r
est
Nei
g
h
b
o
u
r
s
(
k
N
N
)
an
d
S
u
p
p
o
r
t
Vec
to
r
Ma
ch
in
e
(
SV
M)
alg
o
r
ith
m
s
f
u
ll
y
av
ailab
le
in
t
h
e
W
aik
ato
E
n
v
i
r
o
n
m
e
n
t
f
o
r
Kn
o
w
led
g
e
An
al
y
s
i
s
(
W
E
KA
)
[
1
5
]
.
T
h
e
W
ek
a
s
o
f
t
w
ar
e
r
u
n
s
o
n
I
n
tel(
R
)
C
o
r
e
(
T
M)
i5
-
4
2
0
0
M
C
P
U
in
W
in
d
o
w
8
(
6
4
-
b
it)
o
p
er
atin
g
s
y
s
te
m
w
i
th
8
G
B
o
f
r
an
d
o
m
ac
ce
s
s
m
e
m
o
r
y
(
R
A
M)
.
T
h
e
1
0
-
f
o
ld
cr
o
s
s
v
alid
atio
n
w
as
ap
p
lied
f
o
r
v
alid
atin
g
t
h
e
p
er
f
o
r
m
a
n
ce
f
o
r
ea
ch
al
g
o
r
ith
m
in
ter
m
o
f
ac
c
u
r
ac
y
,
p
r
ec
is
io
n
,
r
ec
all
a
n
d
f
-
m
ea
s
u
r
e.
Fi
v
e
y
ea
r
p
er
io
d
o
f
s
a
m
p
le
f
lo
o
d
d
ata
ar
e
s
elec
ted
to
o
b
s
er
v
e
th
e
s
tab
ili
t
y
o
f
p
er
f
o
r
m
an
ce
f
o
r
ea
ch
B
a
y
es
ian
Net
w
o
r
k
s
an
d
t
h
r
ee
o
th
er
class
i
f
ie
r
s
.
3
.
3
.
P
re
-
pro
ce
s
s
ing
T
h
e
p
r
e
-
p
r
o
ce
s
s
in
g
p
er
f
o
r
m
ed
in
t
h
e
d
ata
p
r
ep
ar
atio
n
p
h
ase
is
i
m
p
er
at
iv
e
b
e
f
o
r
e
b
u
ild
in
g
th
e
f
lo
o
d
r
is
k
s
p
r
ed
ictio
n
m
o
d
el
u
s
in
g
t
h
e
f
o
u
r
clas
s
i
f
icatio
n
al
g
o
r
ith
m
s
,
w
h
ich
ar
e
B
N,
k
NN
,
DT
an
d
SVM.
T
h
e
d
ata
f
ir
s
t
r
eq
u
ir
ed
u
n
d
er
g
o
i
n
g
r
esa
m
p
lin
g
p
r
o
ce
s
s
s
in
ce
th
e
d
ata
s
et
is
i
m
b
a
lan
ce
.
I
m
b
a
lan
ce
d
at
a
is
a
clas
s
i
f
icatio
n
p
r
o
b
lem
t
h
at
o
cc
u
r
w
h
e
n
t
h
e
tar
g
et
cla
s
s
e
s
ar
e
n
o
t
eq
u
all
y
d
is
tr
ib
u
ted
.
Fo
r
e
x
a
m
p
le,
th
e
d
ataset
f
o
r
f
lo
o
d
i
n
Ku
ala
Kr
ai
co
n
tain
in
g
ab
o
u
t
1
,
7
9
5
in
s
tan
ce
s
o
f
‘
n
o
f
lo
o
d
’
class
as
co
m
p
ar
ed
to
r
em
ai
n
in
g
3
2
i
n
s
ta
n
ce
s
o
f
‘
f
lo
o
d
’
clas
s
.
A
cc
o
r
d
in
g
to
r
ev
ie
w
co
n
d
u
cted
b
y
[
1
6
]
,
m
an
y
r
ea
l
w
o
r
ld
d
o
m
ain
s
h
as
i
m
b
alan
ce
d
ata
p
r
o
b
le
m
an
d
it
is
cr
u
cial
to
co
m
b
at
i
m
b
alan
ce
d
ata
b
ec
au
s
e
it
w
il
l
n
eg
ati
v
el
y
a
f
f
ec
t
th
e
m
ac
h
i
n
e
l
ea
r
n
in
g
p
r
o
ce
s
s
an
d
d
r
iv
en
er
r
o
r
i
n
cla
s
s
i
f
icatio
n
o
r
p
r
ed
ictio
n
.
R
e
s
a
m
p
li
n
g
i
s
o
n
e
o
f
th
e
m
et
h
o
d
to
co
m
b
at
i
m
b
ala
n
ce
d
ata.
R
esa
m
p
lin
g
p
r
o
ce
s
s
co
n
s
is
t
s
o
f
o
v
er
s
a
m
p
l
in
g
a
n
d
u
n
d
er
-
s
a
m
p
lin
g
.
O
v
er
s
a
m
p
lin
g
i
s
a
p
r
o
ce
s
s
to
ad
d
co
p
ies
o
r
s
y
n
t
h
etic
in
s
ta
n
ce
s
to
u
n
d
er
-
r
ep
r
esen
ted
clas
s
w
h
ile
u
n
d
er
-
s
a
m
p
li
n
g
is
a
p
r
o
ce
s
s
to
d
elete
th
e
i
n
s
ta
n
ce
s
f
r
o
m
o
v
er
r
ep
r
esen
ted
cla
s
s
.
I
n
o
t
h
er
w
o
r
d
,
th
e
o
v
e
r
s
a
m
p
li
n
g
m
eth
o
d
ca
lled
S
y
n
t
h
etic
Mi
n
o
r
it
y
Ov
er
s
a
m
p
lin
g
T
ec
h
n
iq
u
e
(
SM
OT
E
)
h
as b
ee
n
ap
p
lied
t
o
u
n
d
er
-
r
ep
r
esen
ted
d
ata
class
w
h
ic
h
ar
e
f
lo
o
d
class
b
y
ad
d
in
g
s
y
n
t
h
etic
i
n
s
tan
ce
s
to
m
ak
e
t
h
e
d
ata
clas
s
b
alan
ce
.
[
1
7
]
P
r
o
p
o
s
ed
th
e
SMOT
E
tech
n
iq
u
e
to
co
m
b
a
t
i
m
b
alan
ce
d
ata
b
y
cr
ea
tin
g
e
x
t
r
a
tr
ain
i
n
g
d
ata
ca
lled
s
y
n
t
h
eti
c
d
ata.
T
h
e
s
y
n
t
h
etic
d
ata
w
er
e
cr
e
ated
b
y
ta
k
i
n
g
th
e
d
if
f
er
en
ce
b
et
w
ee
n
t
w
o
p
o
in
t
o
r
n
eig
h
b
o
u
r
s
f
r
o
m
r
ea
l
s
a
m
p
le
d
ata.
A
s
a
r
es
u
lt,
th
er
e
n
e
w
d
ataset
w
ill
b
e
cr
ea
ted
r
an
d
o
m
l
y
alo
n
g
t
h
e
li
n
e
s
e
g
m
en
t
b
et
w
ee
n
t
w
o
s
p
ec
if
ic
f
ea
t
u
r
es
o
f
r
ea
l
d
ata.
T
h
u
s
,
1
,
7
9
5
in
s
tan
ce
s
o
f
’
n
o
f
lo
o
d
’
clas
s
an
d
2
,
0
4
8
in
s
ta
n
ce
s
o
f
’
f
lo
o
d
’
cla
s
s
ar
e
p
r
o
d
u
ce
d
af
ter
ap
p
ly
in
g
S
MO
T
E
to
u
n
d
er
-
r
ep
r
esen
ted
’
f
lo
o
d
’
class
.
3
.
4
.
M
o
dellin
g
T
h
is
p
ap
er
is
s
et
to
in
v
esti
g
at
e
th
e
p
er
f
o
r
m
a
n
ce
o
f
B
a
y
esia
n
Net
w
o
r
k
s
a
n
d
o
th
er
m
ac
h
i
n
e
lear
n
in
g
tech
n
iq
u
es
w
h
ic
h
ar
e
DT
,
k
N
N
an
d
SVM
in
p
r
ed
ictin
g
f
lo
o
d
r
is
k
s
b
ased
o
n
a
C
R
I
SP
-
DM
m
et
h
o
d
o
lo
g
y
.
T
h
e
B
ay
e
s
ian
ap
p
r
o
ac
h
is
a
m
o
n
g
o
f
w
ell
-
k
n
o
w
n
tech
n
iq
u
es
to
b
e
u
s
ed
b
y
r
esear
ch
er
s
f
o
r
co
n
s
tr
u
cti
n
g
p
r
ed
ictio
n
m
o
d
el
a
s
w
ell
as
t
h
r
ee
o
th
er
ML
tec
h
n
iq
u
es.
Fo
u
r
class
i
f
ie
r
s
tech
n
iq
u
es
w
h
ic
h
ar
e
B
ay
e
s
ian
Net
w
o
r
k
s
(
B
N)
,
DT
,
k
NN
an
d
SVM
ar
e
w
ell
s
u
p
p
o
r
ted
b
y
d
ata
m
i
n
in
g
to
o
ls
,
W
E
KA
f
o
r
ex
ec
u
ti
n
g
o
f
ex
p
e
r
i
m
en
t [
1
8
]
.
B
ay
e
s
ian
Net
w
o
r
k
s
(
B
N)
o
r
also
k
n
o
w
n
a
s
B
ay
e
s
ia
n
Nets
o
r
B
ay
esia
n
B
elief
Net
w
o
r
k
s
(
B
B
N)
is
a
n
et
w
o
r
k
s
tr
u
c
tu
r
e
m
ad
e
u
p
o
f
Dir
ec
ted
A
c
y
clic
Gr
ap
h
s
(
DAG)
th
a
t
li
n
k
f
ea
tu
r
es
b
ased
o
n
t
h
eir
co
n
d
itio
n
al
p
r
o
b
ab
ilit
ies
w
h
ic
h
th
e
n
ar
e
ca
lcu
lated
u
s
i
n
g
B
ay
es
’
T
h
eo
r
em
[
1
9
]
.
[
1
9
]
A
ls
o
s
tated
th
at
B
N
also
v
er
y
u
s
e
f
u
l
i
n
d
eter
m
i
n
e
,
r
ep
r
esen
t
a
n
d
v
is
u
alize
th
e
r
elatio
n
s
h
ip
a
m
o
n
g
f
ea
t
u
r
es
f
r
o
m
e
m
p
ir
ical
d
ata,
ex
p
er
t
k
n
o
w
led
g
e
o
r
b
o
th
e
m
p
ir
ical
an
d
e
x
p
er
t
k
n
o
w
led
g
e
b
esid
es
d
eter
m
i
n
e
th
e
k
e
y
o
f
u
n
ce
r
tai
n
tie
s
.
A
K2
s
ea
r
c
h
in
g
alg
o
r
it
h
m
w
i
th
B
a
y
e
s
ian
D
ir
ich
let
B
Deu
s
co
r
in
g
m
etr
ic
ad
o
p
ted
f
r
o
m
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
A
r
ti
f
I
n
tell
I
SS
N:
2252
-
8938
Ma
ch
in
e
lea
r
n
in
g
a
p
p
r
o
a
ch
fo
r
flo
o
d
r
is
ks p
r
ed
ictio
n
(
N
a
z
i
m
R
a
z
a
li
)
77
[
2
0
]
h
av
e
b
ee
n
u
s
ed
to
co
n
s
tr
u
ct
B
N
m
o
d
el.
No
te
th
at,
th
e
in
itial
s
tr
u
ct
u
r
e
l
ea
r
n
i
n
g
w
er
e
s
et
to
r
an
d
o
m
s
in
ce
th
e
d
ef
a
u
lt
s
ett
in
g
i
n
W
E
KA
w
ill
co
n
s
tr
u
ct
n
ai
v
e
B
a
y
es
a
s
i
n
it
ial
s
tr
u
ct
u
r
e
lear
n
i
n
g
.
I
n
ad
d
itio
n
,
th
e
n
u
m
b
er
o
f
p
ar
en
t n
o
d
e
is
li
m
it to
f
iv
e
i
n
o
r
d
er
to
av
o
id
co
m
p
lex
i
t
y
a
n
d
h
i
g
h
co
m
p
u
tat
io
n
al
co
s
ts
.
Dec
is
io
n
T
r
ee
(
DT
)
o
r
also
k
n
o
w
n
as
C
la
s
s
i
f
icat
io
n
a
n
d
R
eg
r
ess
io
n
T
r
ee
s
(
C
AR
T
)
is
o
n
e
o
f
p
o
p
u
lar
ML
tec
h
n
iq
u
e
s
.
T
h
e
tar
g
et
cl
ass
i
n
cla
s
s
i
f
icatio
n
tr
ee
s
is
c
ateg
o
r
ical
t
y
p
e
c
lass
w
h
ile
n
u
m
er
ica
l
t
y
p
e
class
f
o
r
r
eg
r
ess
io
n
tr
ee
s
[
2
1
]
.
I
t
is
s
h
o
w
s
t
h
at
tr
ee
s
ar
e
ca
p
ab
le
to
p
r
o
ce
s
s
b
o
th
d
is
cr
ete
an
d
co
n
tin
u
o
u
s
d
ata.
W
E
KA
o
f
f
er
ed
d
ec
is
io
n
tr
ee
in
it
s
s
o
f
t
w
ar
e
h
o
w
e
v
er
i
n
th
i
s
r
esear
ch
,
m
o
r
e
ad
v
an
ce
tr
ee
alg
o
r
ith
m
,
C
4
.
5
alg
o
r
ith
m
o
r
also
k
n
o
w
n
as J
4
8
in
W
E
KA
h
a
v
e
b
ee
n
u
s
ed
to
co
n
s
tr
u
ct
t
h
e
d
ec
is
io
n
tr
e
e
co
m
p
ar
e
to
b
asic
R
E
PT
r
ee
.
T
h
e
s
ettin
g
o
f
n
o
p
r
u
n
in
g
is
s
et
to
f
alse
to
allo
w
th
e
p
r
u
n
in
g
p
r
o
ce
s
s
o
cc
u
r
.
T
h
u
s
,
it
m
a
y
r
ed
u
ce
th
e
co
m
p
le
x
it
y
o
f
tr
e
e
an
d
co
m
p
u
tat
io
n
al
co
s
t
s
b
esid
es
d
ed
u
ce
d
ata
o
v
er
-
f
itt
i
n
g
w
h
ic
h
m
a
y
in
cr
ea
s
e
t
h
e
p
r
ed
ictiv
e
ac
cu
r
a
c
y
.
k
-
n
ea
r
est
n
eig
h
b
o
u
r
s
(
k
N
N)
also
s
u
p
p
o
r
ts
b
o
th
c
lass
if
ica
ti
o
n
an
d
r
eg
r
es
s
io
n
s
a
m
e
lik
e
DT
.
k
NN
i
s
a
s
i
m
p
le
alg
o
r
it
h
m
t
h
at
s
to
r
e
all
tr
ain
i
n
g
d
ataset
an
d
ca
ll
b
ac
k
th
e
d
ata
to
p
r
ed
ict
t
h
e
k
m
o
s
t
s
i
m
ilar
w
i
th
tr
ain
i
n
g
p
atter
n
f
r
o
m
s
to
r
ed
d
ataset.
So
,
k
NN
o
n
l
y
u
s
ed
litt
le
co
m
p
u
tatio
n
a
l
co
s
ts
i
n
o
r
d
er
to
co
m
p
u
te
th
e
d
is
ta
n
ce
b
et
w
ee
n
t
w
o
i
n
s
t
an
ce
s
f
o
r
k
v
al
u
e.
I
n
W
E
K
A
,
k
NN
al
g
o
r
ith
m
w
er
e
p
u
t
u
n
d
er
th
e
‘
laz
y
’
g
r
o
u
p
.
T
h
e
d
ef
au
lt
s
e
ttin
g
w
e
r
e
L
in
ea
r
NN
Sear
ch
w
h
ic
h
u
s
ed
E
u
clid
ea
n
d
is
tan
ce
as
d
i
s
t
an
ce
f
u
n
ctio
n
p
ar
am
eter
to
ca
lc
u
late
t
h
e
d
is
t
an
ce
b
et
w
ee
n
i
n
s
tan
ce
s
.
No
te
th
at,
t
h
e
cr
o
s
s
v
a
lid
ate
p
ar
a
m
e
ter
w
er
e
s
et
to
tr
u
e
i
n
o
r
d
er
to
allo
w
W
E
K
A
d
is
co
v
er
a
g
o
o
d
v
alu
e
f
o
r
k
.
Ho
w
e
v
er
,
t
h
e
v
al
u
e
o
f
k
w
er
e
s
et
to
1
,
3
,
5
an
d
7
to
co
n
tr
o
l
th
e
s
ize
o
f
t
h
e
n
eig
h
b
o
u
r
h
o
o
d
f
o
r
k
NN
i
n
t
h
e
ex
p
er
i
m
en
t
.
T
h
e
b
est
r
esu
lt
p
r
o
d
u
ce
d
b
y
s
elec
ted
k
v
al
u
e
w
il
l
b
e
u
s
ed
as
co
m
p
ar
is
o
n
w
it
h
B
N
an
d
o
th
er
M
L
tec
h
n
iq
u
es.
No
te
th
at
3
as
k
v
al
u
e
p
r
o
d
u
ce
d
m
o
s
t o
p
ti
m
u
m
r
esu
lt
.
Su
p
p
o
r
t
Vec
to
r
Ma
ch
i
n
e
(
SV
M)
w
a
s
ac
t
u
all
y
d
ev
elo
p
ed
f
o
r
s
o
lv
i
n
g
th
e
b
in
ar
y
clas
s
i
f
icat
io
n
p
r
o
b
le
m
s
.
Ho
w
e
v
er
,
th
e
r
esear
ch
er
s
h
av
e
ex
te
n
d
ed
th
e
SVM
to
m
a
k
e
it
s
u
itab
le
to
s
u
p
p
o
r
t
m
u
lti
-
cla
s
s
class
i
f
icatio
n
an
d
r
eg
r
es
s
io
n
p
r
o
b
lem
s
.
SVM
h
as
t
h
e
ca
p
ab
ilit
y
to
h
a
n
d
le
b
o
th
co
n
ti
n
u
o
u
s
a
n
d
d
is
cr
ete
d
ata
as
it
au
to
m
atica
ll
y
n
o
r
m
alize
s
th
e
d
ata
b
ef
o
r
e
th
e
y
ar
e
m
o
d
elled
.
SVM
th
e
n
ca
lcu
l
ates
a
lin
e
th
a
t
b
est
is
o
late
t
h
e
d
ata
i
n
to
t
w
o
g
r
o
u
p
s
a
n
d
o
n
l
y
co
n
s
id
er
th
o
s
e
in
s
ta
n
ce
s
th
at
ar
e
clo
s
e
s
t
to
th
e
s
ep
ar
atin
g
lin
e.
T
h
e
in
s
ta
n
ce
s
ar
e
ca
lled
s
u
p
p
o
r
t
v
ec
to
r
s
,
h
en
ce
t
h
e
n
a
m
e
o
f
t
h
e
tech
n
iq
u
e.
I
n
W
E
KA
,
t
h
e
SVM
alg
o
r
ith
m
is
i
m
p
le
m
e
n
ted
as
t
h
e
Seq
u
e
n
tial
Mi
n
i
m
al
Op
ti
m
izatio
n
(
SMO)
,
w
h
ich
ar
e
th
e
o
p
tim
iza
tio
n
alg
o
r
ith
m
u
s
ed
in
s
id
e
th
e
S
V
M
i
m
p
le
m
en
ta
tio
n
.
3
.
5
.
E
v
a
lua
t
i
o
n
m
et
rics
E
v
er
y
p
r
ed
ictio
n
m
o
d
el
h
a
s
it
s
o
w
n
w
a
y
o
f
v
al
id
atin
g
a
n
d
ev
alu
a
tin
g
i
ts
p
er
f
o
r
m
a
n
ce
.
E
v
alu
atio
n
is
p
er
f
o
r
m
ed
to
co
m
p
ar
e
w
h
et
h
er
th
er
e
ar
e
s
i
m
ilar
it
y
o
r
co
n
s
i
s
te
n
c
y
b
et
w
ee
n
th
e
o
b
s
er
v
ed
r
esu
lt
s
an
d
th
e
p
r
ed
icted
r
esu
lts
o
r
ac
r
o
s
s
a
n
u
m
b
er
o
f
d
if
f
er
e
n
t
m
o
d
els
’
p
r
ed
icted
r
esu
lts
.
Fo
r
th
is
r
esear
ch
,
ac
cu
r
ac
y
,
p
r
ec
is
io
n
,
r
ec
all
an
d
f
-
s
co
r
e
h
as
b
ee
n
u
s
ed
as
ev
al
u
atio
n
m
e
tr
ic
b
ec
au
s
e
it
h
as
b
ee
n
u
s
ed
w
id
el
y
b
y
a
m
aj
o
r
ity
o
f
r
esear
ch
er
s
in
cl
u
d
in
g
[
2
,
22
]
f
o
r
f
lo
o
d
r
is
k
s
p
r
ed
ictio
n
c
o
m
p
ar
e
to
o
th
er
e
v
al
u
atio
n
m
etr
ics
s
u
ch
a
s
R
o
o
t
Me
an
Sq
u
ar
e
E
r
r
o
r
[
2
3
]
an
d
m
o
d
el
co
n
s
tr
u
ctio
n
ti
m
e
s
[
2
4
]
.
A
cc
u
r
ac
y
ca
n
b
e
d
er
iv
ed
f
r
o
m
a
co
n
f
u
s
io
n
m
a
tr
ix
as sh
o
w
n
i
n
T
ab
le
3
.
T
ab
le
3
.
C
o
n
f
u
s
io
n
m
atr
i
x
N
O
(
P
r
e
d
i
c
t
i
o
n
)
Y
ES(
P
r
e
d
i
c
t
i
o
n
)
N
O
(
A
c
t
u
a
l
)
T
r
u
e
N
e
g
a
t
i
v
e
(
T
N
)
F
a
l
s
e
P
o
si
t
i
v
e
(
F
P
)
Y
ES (
A
c
t
u
a
l
)
F
a
l
s
e
N
e
g
a
t
i
v
e
(
F
N
)
T
r
u
e
P
o
si
t
i
v
e
(
T
P
)
T
h
e
co
lu
m
n
s
r
ep
r
ese
n
t
t
h
e
p
r
ed
ictio
n
class
an
d
r
o
w
s
s
h
o
w
th
e
ac
tu
al
clas
s
tar
g
et.
T
h
e
f
lo
o
d
o
u
tco
m
e
is
r
ep
r
esen
ted
w
it
h
th
e
lab
el
YE
S,
an
d
n
o
-
f
lo
o
d
is
r
ep
r
esen
ted
w
it
h
th
e
lab
el
NO.
T
h
er
ef
o
r
e,
d
iag
o
n
al
ele
m
e
n
t
s
(
T
N,
T
P)
in
T
ab
le
2
s
h
o
w
s
t
h
e
tr
u
e
p
r
ed
ictio
n
s
an
d
th
e
o
th
er
elem
e
n
t
s
(
FN,
FP
)
r
ef
lect
t
h
e
f
al
s
e
p
r
ed
ictio
n
s
.
Fo
r
ex
a
m
p
le,
th
er
e
ar
e
t
w
o
o
u
tco
m
e
i
n
t
h
e
f
lo
o
d
p
r
ed
ictio
n
,
w
h
ic
h
ar
e
f
lo
o
d
an
d
n
o
-
f
lo
o
d
.
T
h
e
T
r
u
e
P
o
s
itiv
e
(
T
P
)
m
ea
n
s
co
r
r
ec
t
f
lo
o
d
r
es
u
lt
p
r
ed
ictio
n
an
d
T
r
u
e
Neg
a
ti
v
e
(
T
N)
m
ea
n
s
co
r
r
ec
t
n
o
-
f
lo
o
d
r
esu
lt
p
r
ed
ictio
n
w
h
ile
Fals
e
P
o
s
iti
v
e
(
FP
)
m
e
an
s
i
n
co
r
r
ec
t
f
lo
o
d
r
esu
lt
p
r
ed
ictio
n
an
d
Fa
ls
e
Ne
g
ati
v
e
(
F
N)
m
ea
n
s
i
n
co
r
r
ec
t
no
-
f
lo
o
d
r
esu
lt
p
r
ed
ictio
n
.
I
f
a
tar
g
et
class
i
s
p
r
ed
icted
as
f
lo
o
d
(
YE
S)
ev
en
th
o
u
g
h
it
i
s
a
n
o
-
f
lo
o
d
(
NO)
tar
g
et
cla
s
s
,
t
h
i
s
te
s
t
r
es
u
lt
is
ad
d
ed
to
th
e
FP
i
n
t
h
e
tab
le.
T
h
er
ef
o
r
e,
n
u
m
b
er
FP
is
in
cr
e
m
en
ted
b
y
1
.
T
h
u
s
,
ac
cu
r
ac
y
i
n
th
e
co
n
f
u
s
io
n
m
at
r
ix
is
d
ef
i
n
ed
as i
n
(
1
)
.
A
c
c
ur
a
c
y
=
T
P
T
N
F
P
T
P
T
N
F
N
(
1
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8938
I
n
t J
A
r
ti
f
I
n
tell
,
Vo
l.
9
,
No
.
1
,
Ma
r
ch
20
20
:
73
–
80
78
P
r
ec
is
io
n
(
p
o
s
itiv
e
p
r
ed
ictiv
e
v
alu
e)
ca
n
b
e
d
ef
i
n
ed
a
s
i
n
(
2
)
w
h
er
e
t
h
e
to
tal
n
u
m
b
er
o
f
co
r
r
ec
tly
clas
s
if
ied
p
o
s
itiv
e
s
a
m
p
les ar
e
d
iv
id
ed
b
y
th
e
to
tal
n
u
m
b
er
o
f
ac
t
u
al
p
o
s
iti
v
e
s
a
m
p
le
s
.
P
r
e
c
isi
on
=
TP
F
P
T
P
(
2
)
R
ec
all
(
s
en
s
iti
v
it
y
)
k
n
o
w
ca
n
b
e
d
ef
in
ed
as
in
(
3
)
w
h
er
e
th
e
to
tal
n
u
m
b
er
o
f
co
r
r
ec
tly
class
i
f
ied
p
o
s
itiv
e
s
a
m
p
les d
i
v
id
ed
b
y
t
h
e
to
tal
n
u
m
b
er
o
f
p
r
ed
icted
p
o
s
itiv
e
s
a
m
p
les.
Re
c
a
l
l
=
TP
T
P
F
N
(
3
)
F
-
m
ea
s
u
r
e
(
F1
s
co
r
e
o
r
F sco
r
e)
ca
n
b
e
d
ef
in
ed
as th
e
w
e
ig
h
ted
h
ar
m
o
n
ic
m
ea
n
o
f
t
h
e
p
r
ec
is
io
n
a
n
d
r
ec
all
o
f
th
e
s
a
m
p
les.
2
Pr
e
c
isio
n
Re
c
a
l
l
F
-
M
e
a
su
r
e
=
Pr
e
c
isio
n
+
Re
c
a
l
l
(
4
)
4.
RE
SU
L
T
S
A
ND
D
I
SCU
SS
I
O
N
Fo
llo
w
i
n
g
t
h
e
p
r
ev
io
u
s
ev
al
u
atio
n
m
etr
ic
u
s
ed
i
n
r
esea
r
ch
w
o
r
k
ca
r
r
ied
o
u
t
b
y
[
2
,
2
2
]
,
th
e
ex
p
er
i
m
e
n
tal
r
es
u
lt
s
f
o
r
B
N
an
d
th
r
ee
o
th
er
M
L
tec
h
n
iq
u
es
,
DT
,
k
NN
an
d
SVM
ar
e
co
m
p
ar
ed
i
n
ter
m
s
o
f
ac
cu
r
ac
y
,
p
r
ec
is
io
n
,
r
ec
all
a
n
d
f
-
m
ea
s
u
r
e
p
er
f
o
r
m
a
n
ce
as
th
e
ev
al
u
atio
n
m
etr
ic.
T
ab
le
4
s
h
o
w
s
th
e
ex
p
er
i
m
e
n
tal
r
es
u
lt
s
ac
r
o
s
s
5
y
ea
r
s
p
er
io
d
o
f
f
lo
o
d
d
ata
f
r
o
m
K
u
a
la
Kr
ai,
Kela
n
tan
u
s
in
g
1
0
-
f
o
ld
cr
o
s
s
v
alid
atio
n
.
T
h
e
d
ata
w
er
e
d
i
v
id
ed
in
to
t
w
o
,
n
o
r
m
al
d
ata
an
d
SMOT
E
d
ata.
T
ec
h
n
icall
y
,
th
e
n
o
r
m
al
d
ata
ac
tu
all
y
i
m
b
ala
n
ce
b
ec
au
s
e
t
h
e
tar
g
et
class
w
h
ich
ar
e
f
lo
o
d
an
d
n
o
f
lo
o
d
ar
e
n
o
t
eq
u
all
y
d
is
tr
ib
u
ted
.
I
t
m
a
y
ca
u
s
e
th
e
p
r
ed
ictio
n
r
es
u
lt
s
p
r
o
d
u
ce
w
ill
b
e
b
ia
s
ed
to
m
aj
o
r
ity
tar
g
e
t
cla
s
s
.
So
,
th
e
d
at
a
h
a
v
e
b
ee
n
ap
p
lied
SMOT
E
m
e
th
o
d
i
n
o
r
d
er
to
co
m
b
at
i
m
b
ala
n
ce
d
ata
b
y
ad
d
in
g
th
e
u
n
d
er
-
p
r
ese
n
ted
tar
g
et
class
w
i
th
s
y
n
t
h
eti
c
d
ata
th
at
d
er
iv
ed
f
r
o
m
r
ea
l
d
ata.
Ov
er
all,
t
h
e
r
es
u
lt
s
s
h
o
w
s
th
at
B
N
(
9
9
.
9
4
%)
s
lig
h
tl
y
b
et
ter
th
a
n
o
th
er
t
h
r
ee
ML
tech
n
iq
u
e
s
w
h
ic
h
ar
e
DT
(
9
9
.
8
9
%),
k
NN
(
9
9
.
5
0
%)
an
d
SVM
(
9
8
.
2
3
%)
h
o
w
ev
er
ar
e
co
m
p
ar
ab
le
af
ter
ap
p
licatio
n
o
f
SMOT
E
m
et
h
o
d
to
d
ata
b
y
9
9
.
6
8
%
(
B
N)
,
9
9
.
9
2
%
(
DT
)
,
9
9
.
8
6
%
(
k
NN)
an
d
9
9
.
0
3
%
(
SVM)
as
w
ell
as o
th
er
m
etr
ic
s
u
c
h
as p
r
ec
is
io
n
,
r
ec
all
a
n
d
f
-
m
ea
s
u
r
e
r
esp
ec
tiv
el
y
.
T
h
u
s
,
DT
ac
h
ie
v
e
d
th
e
h
i
g
h
est
v
al
u
e
o
f
p
r
ec
is
io
n
,
r
ec
all
an
d
F
-
m
ea
s
u
r
e
w
h
et
h
er
i
n
n
o
r
m
al
d
ata
a
n
d
SMOT
E
d
ata
b
y
0
.
9
9
9
.
Fig
u
r
e
2
s
h
o
w
s
p
ar
t
o
f
ex
p
er
i
m
e
n
tal
o
u
tp
u
t
f
r
o
m
W
E
KA
u
s
i
n
g
SVM
b
ased
o
n
SMOT
E
d
ata.
T
h
e
ex
p
er
i
m
e
n
tal
o
u
tp
u
t
in
cl
u
d
e
ev
alu
a
tio
n
m
a
tr
ix
s
u
ch
as
ac
c
u
r
ac
y
,
p
r
ec
is
io
n
,
r
ec
all
an
d
f
-
m
ea
s
u
r
e
a
s
w
el
l a
s
co
n
f
u
s
io
n
m
atr
i
x
o
f
tar
g
e
t c
la
s
s
o
f
’
n
o
t f
lo
o
d
’
a
n
d
’
f
lo
o
d
’
.
T
ab
le
5
s
h
o
w
s
th
e
co
n
f
u
s
io
n
m
atr
ix
f
o
r
B
N
an
d
th
r
ee
o
th
er
ML
tech
n
iq
u
es,
DT
,
k
NN
an
d
SV
M
b
ased
o
n
tar
g
et
class
o
f
’
n
o
f
l
o
o
d
’
(
NO)
an
d
’
f
lo
o
d
’
(
YE
S)
u
s
i
n
g
n
o
r
m
al
d
ata
an
d
SMO
T
E
d
ata.
T
h
er
e
ar
e
s
o
m
e
o
b
s
er
v
a
tio
n
h
as
b
ee
n
d
o
n
e
in
t
h
e
e
x
p
er
i
m
e
n
t.
Fo
r
e
x
a
m
p
le,
it
is
o
b
s
er
v
ed
th
at
t
h
e
b
est
k
NN
r
es
u
lt
i
n
ac
cu
r
ac
y
w
er
e
p
r
o
d
u
ce
d
if
th
e
v
alu
e
o
f
k
is
s
et
to
1
w
h
il
e
o
th
er
v
alu
e
o
f
k
s
u
ch
as
3
,
5
an
d
7
p
r
o
d
u
ce
d
s
lig
h
tl
y
w
o
r
s
e
r
es
u
lt
in
ac
c
u
r
ac
y
.
I
n
o
t
h
e
r
w
o
r
d
s
,
w
h
en
th
e
v
al
u
e
o
f
k
i
n
cr
ea
s
ed
,
th
e
ac
cu
r
ac
y
ac
h
ie
v
ed
b
y
k
NN
w
il
l
b
e
d
ec
r
ea
s
ed
.
B
esid
es,
th
e
w
ater
lev
e
l
p
la
y
i
m
p
o
r
tan
t
r
o
les
as
f
ea
tu
r
e
s
f
o
r
f
lo
o
d
s
in
ce
th
e
f
ea
t
u
r
e
b
ec
o
m
e
m
ai
n
r
u
les
f
o
r
DT
(
’
n
o
f
lo
o
d
’
i
f
t
h
e
w
a
ter
lev
e
l
lo
wer
an
d
eq
u
al
to
2
1
8
2
cm
an
d
’
f
lo
o
d
’
i
f
t
h
e
w
a
ter
lev
el
m
o
r
e
t
h
an
2
1
8
2
cm
)
a
n
d
d
ir
ec
tl
y
p
o
in
ted
to
tar
g
et
cl
ass
i
n
B
N.
Me
a
n
w
h
ile,
B
N
s
t
ill
ev
o
l
v
e
o
v
er
ti
m
e
an
d
h
as
p
o
ten
tial
to
b
e
i
m
p
r
o
v
ed
f
u
r
t
h
er
in
f
u
tu
r
e.
[
2
5
]
C
lai
m
ed
th
at
B
N
h
a
v
e
m
a
n
y
ad
v
an
ta
g
e
s
o
v
er
o
th
er
class
i
f
icatio
n
tec
h
n
iq
u
es
to
s
o
lv
e
t
h
e
r
ea
l
w
o
r
ld
p
r
o
b
le
m
s
.
I
n
t
h
eir
s
u
r
v
e
y
,
t
h
e
y
e
x
p
lain
ed
an
d
d
is
c
u
s
s
e
v
er
y
d
is
cr
ete
B
N
clas
s
if
ier
t
h
at
a
v
ailab
le
an
d
ca
te
g
o
r
ize
t
h
e
m
i
n
t
h
r
ee
g
r
o
u
p
b
ased
o
n
f
ac
to
r
izatio
n
.
T
h
e
y
al
s
o
s
tated
th
a
t
B
N
ca
n
b
e
o
r
g
an
i
s
ed
h
ier
ar
ch
ica
ll
y
f
r
o
m
t
h
e
s
i
m
p
lest
alg
o
r
it
h
m
li
k
e
n
aiv
e
B
ay
e
s
to
t
h
e
m
o
s
t
co
m
p
le
x
lik
e
B
a
y
e
s
ian
m
u
ltiu
n
it.
Ho
w
e
v
er
,
[
2
2
]
s
tated
i
n
t
h
eir
w
o
r
k
w
h
er
e
t
h
e
co
n
tin
u
o
u
s
d
ev
elo
p
m
e
n
t
o
f
m
ac
h
in
e
lear
n
i
n
g
alg
o
r
it
h
m
s
i
n
ti
m
e
m
a
y
ex
p
a
n
d
th
e
m
ac
h
i
n
e
lear
n
i
n
g
ap
p
licatio
n
s
i
n
t
h
e
f
ield
o
f
h
y
d
r
o
lo
g
y
ar
e
b
ec
o
m
i
n
g
m
o
r
e
an
d
m
o
r
e
ex
ten
s
i
v
e
in
t
h
e
f
u
t
u
r
e
esp
ec
ia
ll
y
o
n
f
lo
o
d
r
is
k
ass
e
s
s
m
e
n
t.
T
ab
le
4
.
E
x
p
er
im
e
n
tal
r
esu
lts
T
e
c
h
n
i
q
u
e
s
A
c
c
u
r
a
c
y
P
r
e
c
i
si
o
n
R
e
c
a
l
l
F
-
M
e
a
su
r
e
N
o
r
mal
S
mo
t
e
N
o
r
mal
S
mo
t
e
N
o
r
mal
S
mo
t
e
N
o
r
mal
S
mo
t
e
B
a
y
e
si
a
n
N
e
t
w
o
r
k
9
9
.
9
4
%
9
9
.
6
8
%
0
.
9
9
9
0
.
9
9
7
0
.
9
9
9
0
.
9
9
7
0
.
9
9
9
0
.
9
9
7
D
e
c
i
si
o
n
T
r
e
e
9
9
.
8
9
%
9
9
.
9
2
%
0
.
9
9
9
0
.
9
9
9
0
.
9
9
9
0
.
9
9
9
0
.
9
9
9
0
.
9
9
9
k
-
N
e
a
r
e
st
N
e
i
g
h
b
o
u
r
s
9
9
.
5
0
%
9
9
.
7
3
%
0
.
9
9
5
0
.
9
9
7
0
.
9
9
5
0
.
9
9
7
0
.
9
9
5
0
.
9
9
7
S
u
p
p
o
r
t
V
e
c
t
o
r
M
a
c
h
i
n
e
9
9
.
5
0
%
9
9
.
7
6
%
0
.
9
9
5
0
.
9
9
8
0
.
9
9
5
0
.
9
9
8
0
.
9
9
5
0
.
9
9
8
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
A
r
ti
f
I
n
tell
I
SS
N:
2252
-
8938
Ma
ch
in
e
lea
r
n
in
g
a
p
p
r
o
a
ch
fo
r
flo
o
d
r
is
ks p
r
ed
ictio
n
(
N
a
z
i
m
R
a
z
a
li
)
79
Fig
u
r
e
2
.
E
x
ce
r
p
t o
f
ex
p
er
i
m
e
n
tal
o
u
tp
u
t f
r
o
m
W
E
K
A
T
ab
le
5
.
C
o
n
f
u
s
io
n
m
atr
i
x
f
o
r
all
tech
n
iq
u
e
s
b
ased
o
n
n
o
r
m
a
l d
ata
an
d
s
m
o
te
d
ata
N
O
(
P
r
e
d
i
c
t
i
o
n
)
Y
e
s (P
r
e
d
i
c
t
i
o
n
)
B
a
y
e
si
a
n
N
e
t
w
o
r
k
s w
i
t
h
n
o
r
mal
d
a
t
a
N
O
(
A
c
t
u
a
l
)
1
7
9
5
0
Y
ES (
A
c
t
u
a
l
)
1
31
B
a
y
e
si
a
n
N
e
t
w
o
r
k
s w
i
t
h
S
M
O
T
E
d
a
t
a
N
O
(
A
c
t
u
a
l
)
1
7
9
3
2
Y
ES (
A
c
t
u
a
l
)
10
2
0
3
8
D
e
c
i
si
o
n
T
r
e
e
w
i
t
h
n
o
r
mal
d
a
t
a
N
O
(
A
c
t
u
a
l
)
1
7
9
3
2
Y
ES (
A
c
t
u
a
l
)
0
32
D
e
c
i
si
o
n
T
r
e
e
w
i
t
h
S
M
O
T
E
d
a
t
a
N
O
(
A
c
t
u
a
l
)
1
7
9
2
3
Y
ES (
A
c
t
u
a
l
)
0
2
0
4
8
k
-
N
e
a
r
e
st
N
e
i
g
b
o
r
s w
i
t
h
n
o
r
mal
d
a
t
a
NO
(
A
c
t
u
a
l
)
1
7
9
4
1
Y
ES (
A
c
t
u
a
l
)
8
24
k
-
N
e
a
r
e
st
N
e
i
g
b
o
r
s w
i
t
h
S
M
O
T
E
d
a
t
a
N
O
(
A
c
t
u
a
l
)
1
7
9
1
4
Y
ES (
A
c
t
u
a
l
)
1
2
0
4
7
S
u
p
p
o
r
t
V
e
c
t
o
r
M
a
c
h
i
n
e
w
i
t
h
n
o
r
mal
d
a
t
a
N
O
(
A
c
t
u
a
l
)
1
7
9
5
0
Y
ES (
A
c
t
u
a
l
)
9
23
S
u
p
p
o
r
t
V
e
c
t
o
r
M
a
c
h
i
n
e
w
i
t
h
S
M
O
T
E
d
a
t
a
N
O
(
A
c
t
u
a
l
)
1
7
8
6
9
Y
ES
(
A
c
t
u
a
l
)
0
2
0
4
8
5.
CO
NCLU
SI
O
N
AND
F
U
T
U
RE
WO
RK
I
n
co
n
clu
s
io
n
,
t
h
i
s
r
esear
ch
p
r
esen
ted
a
f
lo
o
d
r
is
k
s
p
r
ed
ictio
n
b
ased
o
n
C
R
I
SP
-
DM
m
e
th
o
d
o
lo
g
y
u
s
i
n
g
B
N
an
d
t
h
r
ee
m
ac
h
i
n
e
lear
n
i
n
g
(
M
L
)
tec
h
n
iq
u
es
k
n
o
w
n
a
s
DT
,
k
NN
a
n
d
SV
M.
T
h
e
ai
m
s
o
f
t
h
i
s
r
esear
ch
is
to
d
ev
elo
p
a
p
r
ed
i
ctiv
e
m
o
d
elli
n
g
f
o
r
f
lo
o
d
r
is
k
s
p
r
ed
ictio
n
in
Ku
ala
Kr
ai,
Kela
n
ta
n
,
Ma
la
y
s
ia.
T
h
e
p
r
ed
ictiv
e
ac
cu
r
ac
y
o
f
al
l
m
o
d
els
w
er
e
co
m
p
ar
ed
an
d
r
esu
lt
s
s
h
o
w
ed
t
h
at
B
N
w
a
s
s
l
ig
h
tl
y
b
etter
u
s
i
n
g
n
o
r
m
al
d
ata.
T
h
e
r
esear
ch
als
o
f
o
u
n
d
t
h
at
SMOT
E
m
eth
o
d
ar
e
h
ig
h
l
y
u
s
e
f
u
l
i
n
co
m
b
ati
n
g
w
i
th
i
m
b
ala
n
ce
d
ataset.
T
h
is
f
i
n
d
in
g
i
s
s
u
p
p
o
r
ted
b
y
t
h
e
r
es
u
lt
s
o
f
t
h
e
m
o
d
e
ls
w
h
e
n
S
MO
T
E
m
et
h
o
d
ar
e
ap
p
lied
.
Oth
er
th
an
th
at,
t
h
e
s
tu
d
y
also
f
o
u
n
d
t
h
at
ea
ch
tec
h
n
iq
u
e
s
as
its
o
wn
ad
v
a
n
ta
g
es
a
n
d
d
is
ad
v
a
n
ta
g
es.
Ho
w
e
v
er
,
s
o
m
e
r
ese
ar
ch
li
k
e
[
1
1
]
s
u
g
g
est
e
n
s
e
m
b
les
cla
s
s
i
f
ier
i
s
b
etter
t
h
a
n
B
N
a
n
d
o
th
er
M
L
tec
h
n
iq
u
es
s
u
ch
as
DT
,
k
N
N
an
d
SVM.
I
t
is
also
en
co
u
r
ag
e
to
m
a
k
e
d
y
n
a
m
ic
s
y
s
te
m
t
h
at
ca
n
in
co
r
p
o
r
ate
w
it
h
ti
m
e
v
ar
iatio
n
f
o
r
f
lo
o
d
p
r
ed
ictio
n
as
t
h
e
f
lo
o
d
ev
e
n
t
i
s
a
r
ac
e
w
it
h
ti
m
e
in
o
r
d
er
to
p
lan
t
h
e
p
r
ev
e
n
ti
v
e
ac
tio
n
t
h
at
m
u
s
t
b
e
ta
k
en
i
n
a
s
h
o
r
t
o
f
ti
m
e
s
i
n
ce
f
lo
o
d
is
d
i
s
aster
t
h
at
n
o
t
j
u
s
t
d
estro
y
h
u
g
e
a
m
o
u
n
t
o
f
p
r
o
p
er
ties
b
u
t
al
s
o
ca
n
ca
u
s
e
lo
s
s
o
f
m
an
y
h
u
m
a
n
li
v
es.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8938
I
n
t J
A
r
ti
f
I
n
tell
,
Vo
l.
9
,
No
.
1
,
Ma
r
ch
20
20
:
73
–
80
80
ACK
NO
WL
E
D
G
E
M
E
NT
S
T
h
is
r
esear
ch
is
s
u
p
p
o
r
ted
b
y
U
n
i
v
er
s
iti
T
u
n
H
u
s
s
ei
n
On
n
Ma
la
y
s
ia
v
ia
t
h
e
T
ier
1
Gr
an
t
Sch
e
m
e
Vo
t H
0
7
3
RE
F
E
R
E
NC
E
S
[1
]
M
.
S
.
T
e
h
ra
n
y
,
S
.
Jo
n
e
s,
a
n
d
F
.
S
h
a
b
a
n
i
,
“
Id
e
n
ti
fy
in
g
th
e
e
ss
e
n
ti
a
l
f
lo
o
d
c
o
n
d
it
io
n
i
n
g
f
a
c
to
rs
f
o
r
flo
o
d
p
ro
n
e
a
re
a
m
a
p
p
in
g
u
si
n
g
m
a
c
h
in
e
lea
rn
in
g
tec
h
n
iq
u
e
s,”
Ca
ten
a
,
v
o
l.
1
7
5
,
n
o
.
A
p
ril
,
p
p
.
1
7
4
–
1
9
2
,
2
0
1
9
.
[2
]
N.
I.
M
.
R
o
slin
,
A
.
M
u
sta
p
h
a
,
N.
A
.
S
a
m
su
d
in
,
a
n
d
N
.
Ra
z
a
li
,
“
A
b
a
y
e
sia
n
a
p
p
ro
a
c
h
t
o
p
re
d
ictio
n
o
f
f
lo
o
d
risk
s,”
In
ter
n
a
t
io
n
a
l
J
o
u
rn
a
l
o
f
E
n
g
in
e
e
rin
g
a
n
d
T
e
c
h
n
o
l
o
g
y
,
v
o
l.
7
,
n
o
.
4
.
3
8
,
p
p
.
1
1
4
2
–
1
1
4
5
,
2
0
1
8
.
[3
]
E.
V
e
n
k
a
tes
a
n
a
n
d
A
.
B.
M
a
h
i
n
d
ra
k
a
r,
“
F
o
re
c
a
stin
g
f
lo
o
d
s
u
sin
g
e
x
tre
m
e
g
ra
d
ien
t
b
o
o
sti
n
g
–
a
n
e
w
a
p
p
ro
a
c
h
,
”
In
ter
n
a
t
io
n
a
l
J
o
u
rn
a
l
o
f
Civil
En
g
in
e
e
rin
g
a
n
d
T
e
c
h
n
o
lo
g
y
,
v
o
l.
1
0
,
n
o
.
2
,
p
p
.
1
3
3
6
–
1
3
4
6
,
2
0
1
9
.
[4
]
M
.
Oth
m
a
n
,
A
.
A
.
L
a
ti
f
,
S
.
S
.
M
a
id
in
,
M
.
F
.
M
.
S
a
a
d
,
a
n
d
M
.
N.
A
h
m
a
d
,
En
g
a
g
e
m
e
n
t
o
f
L
o
c
a
l
He
ro
e
s
in
M
a
n
a
g
in
g
F
lo
o
d
Disa
ste
r:
L
e
s
so
n
s
Lea
rn
t
f
ro
m
th
e
2
0
1
4
F
lo
o
d
o
f
Ke
m
a
m
a
n
,
Tere
n
g
g
a
n
u
,
M
a
lay
sia
.
In
tec
h
Op
e
n
,
2
0
1
8
.
[5
]
V
.
Ya
d
a
v
a
n
d
K.
El
iza
,
“
A
h
y
b
rid
w
a
v
e
l
e
t
-
su
p
p
o
r
t
v
e
c
to
r
m
a
c
h
in
e
m
o
d
e
l
f
o
r
p
re
d
ictio
n
o
f
lak
e
wa
ter
le
v
e
l
f
lu
c
tu
a
ti
o
n
s
u
si
n
g
h
y
d
ro
-
m
e
teo
ro
lo
g
ica
l
d
a
ta,”
J
o
u
rn
a
l
o
f
t
h
e
In
ter
n
a
ti
o
n
a
l
M
e
a
su
re
me
n
t
Co
n
fed
e
ra
ti
o
n
,
v
o
l.
1
0
3
,
p
p
.
2
6
5
5
–
2
6
7
5
,
2
0
1
7
.
[6
]
A
.
D.
A
.
Da
li
,
N.
A
.
O
m
a
r,
a
n
d
A
.
M
u
sta
p
h
a
,
“
Da
ta
m
in
in
g
a
p
p
ro
a
c
h
to
h
e
rb
s
c
las
sif
ica
ti
o
n
,
”
In
d
o
n
e
sia
n
J
o
u
rn
a
l
o
f
El
e
c
trica
l
En
g
in
e
e
rin
g
a
n
d
Co
mp
u
ter
S
c
ien
c
e
,
v
o
l
.
1
2
,
n
o
.
2
,
p
p
.
5
7
0
–
5
7
6
,
2
0
1
8
.
[7
]
S
.
F
.
A
.
Ra
z
a
k
,
C.
P
.
L
e
e
,
K.
M
.
L
i
m
,
a
n
d
P
.
X.
T
e
e
,
“
S
m
a
rt
h
a
lal
re
c
o
g
n
ize
r
f
o
r
m
u
sli
m
c
o
n
su
m
e
r
s,”
In
d
o
n
e
si
a
n
J
o
u
rn
a
l
o
f
El
e
c
trica
l
En
g
in
e
e
rin
g
a
n
d
Co
m
p
u
ter
S
c
ien
c
e
,
v
o
l.
1
4
,
n
o
.
1
,
p
p
.
1
9
3
–
2
0
0
,
2
0
1
9
[8
]
M
.
A
.
S
.
A
n
u
a
r,
R.
Z.
A
.
Ra
h
m
a
n
,
S
.
B.
M
o
h
d
,
A
.
C.
S
o
h
,
a
n
d
Z.
D.
Zu
lk
a
f
li
,
“
Ea
rly
p
re
d
ictio
n
sy
ste
m
u
sin
g
n
e
u
ra
l
n
e
tw
o
rk
in
k
e
la
n
ta
n
riv
e
r,
ma
la
y
sia
,
”
in
P
r
o
c
e
e
d
in
g
s
o
f
th
e
1
5
t
h
IEE
E
S
t
u
d
e
n
t
C
o
n
f
e
re
n
c
e
o
n
Re
se
a
rc
h
a
n
d
De
v
e
lo
p
m
e
n
t:
In
sp
ir
in
g
T
e
c
h
n
o
l
o
g
y
f
o
r
Hu
m
a
n
it
y
,
S
CORe
D 2
0
1
7
,
2
0
1
8
,
p
p
.
1
0
4
–
1
0
9
.
[9
]
M
.
A
b
d
u
ll
a
h
,
M
.
Oth
m
a
n
,
S
.
Ka
si
m
,
a
n
d
S
.
M
o
h
a
m
e
d
,
“
Ev
o
lv
in
g
sp
ik
in
g
n
e
u
ra
l
n
e
tw
o
rk
s
m
e
th
o
d
s
f
o
r
c
las
si
f
ica
ti
o
n
p
r
o
b
lem
:
a
c
a
se
stu
d
y
in
f
lo
o
d
e
v
e
n
ts
risk
a
ss
e
ss
m
e
n
t
,
”
In
d
o
n
e
sia
n
J
o
u
rn
a
l
o
f
El
e
c
trica
l
E
n
g
i
n
e
e
rin
g
a
n
d
Co
m
p
u
ter
S
c
ie
n
c
e
,
v
o
l.
1
6
,
n
o
.
1
,
p
p
.
2
2
2
–
2
2
9
,
2
0
1
9
.
[1
0
]
G
.
Zh
a
o
,
B.
P
a
n
g
,
Z.
Xu
,
D.
P
e
n
g
,
a
n
d
L
.
X
u
,
“
A
ss
e
ss
m
e
n
t
o
f
u
rb
a
n
f
lo
o
d
su
sc
e
p
ti
b
il
it
y
u
sin
g
se
m
i
-
su
p
e
rv
ise
d
m
a
c
h
in
e
lea
rn
in
g
m
o
d
e
l,
”
S
c
ien
c
e
o
f
th
e
T
o
tal
En
v
iro
n
m
e
n
t,
v
o
l.
6
5
9
,
n
o
.
3
,
p
p
.
9
4
0
–
9
4
9
,
2
0
1
9
.
[1
1
]
Z.
G
e
n
g
,
X
.
Hu
,
Q.
Zh
u
,
Y.
Ha
n
,
Y.
X
u
,
a
n
d
Y.
He
,
“
Pa
tt
e
rn
re
c
o
g
n
it
io
n
fo
r
wa
ter
fl
o
o
d
e
d
l
a
y
e
r
b
a
se
d
o
n
e
n
se
mb
le
c
la
ss
if
ier
,
”
in
P
ro
c
e
e
d
i
n
g
s
o
f
th
e
5
th
In
tern
a
ti
o
n
a
l
Co
n
f
e
r
e
n
c
e
o
n
Co
n
tro
l
,
De
c
isio
n
a
n
d
In
f
o
rm
a
ti
o
n
T
e
c
h
n
o
lo
g
ies
,
Co
DIT
2
0
1
8
,
2
0
1
8
,
p
p
.
1
64
–
1
6
9
.
[1
2
]
R.
W
irt
h
a
n
d
J.
Hi
p
p
,
“
Cris
p
-
d
m
:
T
o
wa
r
d
s
a
st
a
n
d
a
rd
p
ro
c
e
ss
mo
d
e
l
f
o
r
d
a
ta
mi
n
in
g
,
”
i
n
P
r
o
c
e
e
d
in
g
s
o
f
th
e
4
th
in
tern
a
ti
o
n
a
l
c
o
n
f
e
re
n
c
e
o
n
th
e
p
ra
c
ti
c
a
l
a
p
p
li
c
a
ti
o
n
s
o
f
k
n
o
w
led
g
e
d
isc
o
v
e
r
y
a
n
d
d
a
ta
m
in
in
g
.
Cit
e
se
e
r,
2
0
0
0
,
p
p
.
29
–
3
9
.
[1
3
]
T
h
e
o
ff
icital
we
b
o
f
p
u
b
li
c
in
f
o
b
a
n
ji
r.
[
O
n
li
n
e
].
A
c
c
e
ss
e
d
o
n
:
No
v
.
2
4
,
2
0
1
9
.
A
v
a
il
a
b
le:
h
tt
p
:
//
p
u
b
l
icin
f
o
b
a
n
j
ir.
w
a
ter.g
o
v
.
m
y
/
[1
4
]
L
a
m
a
n
w
e
b
ra
s
m
i
jab
a
tan
m
e
teo
ro
lo
g
i
M
a
lay
si
a
.
[
On
li
n
e
]
.
A
c
c
e
ss
e
d
o
n
:
No
v
.
2
4
,
2
0
1
9
.
A
v
a
il
a
b
le:
h
tt
p
:
//
ww
w
.
m
e
t.
g
o
v
.
m
y
/
[1
5
]
I.
H.
W
it
ten
,
E.
F
ra
n
k
,
M
.
A
.
H
a
ll
,
a
n
d
C.
J.
P
a
l,
Da
ta
M
i
n
in
g
:
P
ra
c
ti
c
a
l
m
a
c
h
in
e
lea
rn
in
g
to
o
ls
a
n
d
tec
h
n
iq
u
e
s.
M
o
rg
a
n
Ka
u
fm
a
n
n
,
2
0
1
6
.
[1
6
]
H.
A
li
,
M
.
N.
M
.
S
a
ll
e
h
,
R.
S
a
e
d
u
d
i
n
,
K.
H
u
ss
a
in
,
a
n
d
M
.
F
.
M
u
s
h
taq
,
“
Im
b
a
lan
c
e
c
las
s
p
ro
b
lem
s
in
d
a
ta
m
in
in
g
:
A
re
v
ie
w
,
”
In
d
o
n
e
sia
n
J
o
u
r
n
a
l
o
f
E
lec
trica
l
En
g
in
e
e
rin
g
a
n
d
C
o
mp
u
ter
S
c
ien
c
e
,
v
o
l.
1
4
,
n
o
.
3
,
p
p
.
1
5
5
2
–
1
5
6
3
,
2
0
1
9
.
[1
7
]
N.
V
.
C
h
a
w
la,
K.
W
.
Bo
wy
e
r,
L
.
O.
Ha
ll
,
a
n
d
W
.
P
.
Ke
g
e
lme
y
e
r,
“
S
m
o
te:
S
y
n
th
e
ti
c
m
in
o
rit
y
o
v
e
r
-
sa
m
p
li
n
g
tec
h
n
iq
u
e
,
”
J
o
u
r
n
a
l
o
f
Arti
fi
c
ia
l
I
n
telli
g
e
n
c
e
Res
e
a
rc
h
,
v
o
l.
1
6
,
p
p
.
321
–
3
5
7
,
2
0
0
2
.
[1
8
]
J.
Bro
w
n
lee
,
“
Ho
w
to
u
se
c
l
a
ss
i
f
i
c
a
ti
o
n
m
a
c
h
in
e
lea
rn
in
g
a
l
g
o
rit
h
m
s
in
w
e
k
a
”
,
A
u
g
.
2
2
,
2
0
1
9
.
A
c
c
e
ss
e
d
o
n
:
No
v
.
2
4
,
2
0
1
9
.
[
On
li
n
e
]
.
A
v
a
il
a
b
le:
h
tt
p
s:/
/m
a
c
h
in
e
lea
rn
in
g
m
a
ste
r
y
.
c
o
m
/u
se
-
c
las
sif
i
c
a
ti
o
n
-
m
a
c
h
in
e
-
lea
rn
i
n
g
-
a
lg
o
rit
h
m
s
-
w
e
k
a
/
[1
9
]
B.
G
.
M
a
rc
o
t
a
n
d
T
.
D.
P
e
n
m
a
n
,
“
A
d
v
a
n
c
e
s
in
b
a
y
e
sia
n
n
e
tw
o
rk
m
o
d
e
ll
in
g
:
I
n
teg
ra
ti
o
n
o
f
m
o
d
e
ll
in
g
tec
h
n
o
l
o
g
ies
,
”
Kn
o
w
led
g
e
-
Ba
se
d
S
y
st
e
m
s,
v
o
l.
2
2
,
p
p
.
3
8
6
–
3
9
3
,
2
0
1
9
.
[2
0
]
M
.
G
.
M
a
d
d
e
n
,
“
On
th
e
c
las
sif
ica
ti
o
n
p
e
rf
o
rm
a
n
c
e
o
f
tan
a
n
d
g
e
n
e
ra
l
b
a
y
e
sia
n
n
e
tw
o
rk
s,”
En
v
iro
n
m
e
n
tal
M
o
d
e
ll
i
n
g
a
n
d
S
o
f
tw
a
re
,
v
o
l.
2
2
,
n
o
.
2
,
p
p
.
4
8
9
–
4
9
5
,
2
0
0
9
.
[2
1
]
N.
Ya
d
a
v
,
A
.
Ku
m
a
r,
R.
Bh
a
tn
a
g
a
r,
a
n
d
V
.
K.
V
e
rm
a
,
“
Cit
y
c
rime
ma
p
p
i
n
g
u
si
n
g
m
a
c
h
in
e
le
a
rn
i
n
g
tec
h
n
iq
u
e
s
,
”
in
P
r
o
c
e
e
d
in
g
s
o
f
th
e
4
th
In
ter
n
a
ti
o
n
a
l
Co
n
f
e
re
n
c
e
o
n
A
d
v
a
n
c
e
d
M
a
c
h
in
e
L
e
a
rn
in
g
Tec
h
n
o
lo
g
i
e
s
a
n
d
A
p
p
li
c
a
ti
o
n
s,
A
M
L
TA
2
0
1
9
,
v
o
l.
9
2
1
,
2
0
2
0
,
p
p
.
6
5
6
–
6
6
8
.
[2
2
]
X
.
L
i,
D.
Ya
n
,
K.
W
a
n
g
,
B.
W
e
n
g
,
T
.
Qin
,
a
n
d
S
.
L
iu
,
“
F
l
o
o
d
risk
a
ss
e
ss
m
e
n
t
o
f
g
lo
b
a
l
w
a
ters
h
e
d
s
b
a
se
d
o
n
m
u
lt
ip
le m
a
c
h
in
e
lea
rn
in
g
m
o
d
e
ls,” W
a
ter,
v
o
l.
1
1
,
n
o
.
1
6
5
4
,
p
p
.
1
–
1
8
,
2
0
1
9
.
[2
3
]
Y.
W
u
,
W
.
X
u
,
J.
F
e
n
g
t,
S
.
P
a
lai
a
h
n
a
k
o
te,
a
n
d
T
.
L
u
,
“
L
o
c
a
l
a
n
d
g
lo
b
a
l
b
a
y
e
sia
n
n
e
tw
o
rk
b
a
se
d
mo
d
e
l
fo
r
fl
o
o
d
p
re
d
ictio
n
,
”
i
n
P
r
o
c
e
e
d
in
g
s
o
f
th
e
2
4
t
h
I
n
tern
a
ti
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
P
a
tt
e
rn
Re
c
o
g
n
it
io
n
,
IC
P
R
2
0
1
8
,
p
p
.
2
2
5
–
2
3
0
,
2
0
1
8
.
[2
4
]
A
.
H
e
sa
r,
H.
Tab
a
tab
a
e
e
,
a
n
d
M
.
Ja
lali
,
“
S
tru
c
tu
re
lea
rn
in
g
o
f
b
a
y
e
sia
n
n
e
t
w
o
rk
s
u
sin
g
h
e
u
r
isti
c
m
e
th
o
d
s,”
Per
ta
n
ika
J
o
u
rn
a
l
o
f
S
c
ien
c
e
a
n
d
T
e
c
h
n
o
l
o
g
y
,
v
o
l.
4
5
,
p
p
.
2
4
6
–
2
5
0
,
2
0
1
2
.
[2
5
]
C.
Bielz
a
a
n
d
P
.
L
a
rra
˜n
a
g
a
,
“
Dis
c
re
te b
a
y
e
sia
n
n
e
tw
o
rk
c
l
a
ss
i
f
iers
:
A
su
rv
e
y
,
”
A
CM
Co
m
p
u
ti
n
g
S
u
rv
e
y
s,
2
0
1
4
.
Evaluation Warning : The document was created with Spire.PDF for Python.