I
A
E
S
I
n
t
e
r
n
at
io
n
al
Jou
r
n
al
of
A
r
t
if
ic
ia
l
I
n
t
e
ll
ig
e
n
c
e
(
I
J
-
A
I
)
V
ol
.
10
, N
o.
2
,
J
une
2021
, pp.
365
~
373
I
S
S
N
:
2252
-
8938
,
D
O
I
:
10.11591/
ij
a
i.
v
10
.i
2
.pp
365
-
373
365
Jou
r
n
al
h
om
e
page
:
ht
tp
:
//
ij
ai
.
ia
e
s
c
or
e
.c
om
S
p
at
i
al
an
a
l
ysi
s
m
od
e
l
f
or
t
r
af
f
i
c
a
c
c
i
d
e
n
t
-
p
r
on
e
r
oad
s
c
l
ass
i
f
i
c
at
i
on
:
a p
r
op
ose
d
f
r
am
e
w
or
k
A
n
ik
V
e
ga V
it
ia
n
in
gs
ih
1
, N
an
n
a
S
u
r
yan
a
2
,
Z
ah
r
ia
h
O
t
h
m
an
3
1
Departmen
t of I
nformatics, Universitas Dr. Soetomo
, Surabaya, Indonesia
1,2,3
Faculty
of In
forma
tion an
d Communi
cation
Tech
nology,
Univer
siti Tek
nikal M
alaysi
a Mela
ka, Me
laka,
Malays
ia
A
r
t
ic
le
I
n
f
o
A
B
S
T
R
A
C
T
A
r
ti
c
le
h
is
to
r
y
:
R
e
c
e
iv
e
d
F
e
b
2
5
, 20
20
R
e
vi
s
e
d
D
e
c
1
0, 20
20
A
c
c
e
pt
e
d
A
pr
2
, 20
2
1
The
classification
method
in
the
spatial
analysis
modeling
based
on
the
multi
-
criteria
parameter
is
currently
widely
used
to
manage
geo
graphic
information
systems
(GIS)
software
engineer
ing.
The
accura
cy
of
the
proposed
model
will
play
an
essential
role
in
the
successful
s
oftware
development
of
GIS.
This
is
related
to
th
e
nature
of
GIS
used
for
m
apping
through
spatial
analysis.
This
paper
aims
to
propose
a
framewo
rk
of
spatial
analysis
using
a
hybrid
estimat
ion
model
-
based
on
a
combination
of
multi
-
criteria
decision
-
making
(MCDM)
and
artificial
neural
networks
(
ANNs)
(MCDM
-
ANNs)
classifi
cation
.
The
proposed
framework
is
based
on
the
compariso
n
of
existin
g
frameworks
through
the
concept
of
a
lit
erature
review.
The
model
in
the
proposed
framework
will
be
used
for
futur
e
work
on
the
traffic
accident
-
prone
road
classific
ation
through
testing
with
a
private
or public spatial dataset. Model validation testing
on the proposed fra
mework
uses
metaheuristic
optimization
techniques.
Policymakers
can
use
the
results
of
t
he
model
on
the
proposed
framework
for
initial
planning
developi
ng
GIS
software en
gineering through spa
tial analysis models
.
K
e
y
w
o
r
d
s
:
G
I
S
s
of
twa
r
e
e
ngi
ne
e
r
in
g
H
ybr
id
e
s
ti
m
a
ti
on mode
l
-
ba
s
e
d
M
C
D
M
-
A
N
N
s
P
r
opos
e
d f
r
a
m
e
w
or
k
S
pa
ti
a
l
a
na
ly
s
is
m
ode
l
T
r
a
f
f
ic
a
c
c
id
e
nt
-
pr
one
r
oa
ds
This is an
open
acce
ss artic
le unde
r the
CC BY
-
SA
license.
C
or
r
e
s
pon
di
n
g A
u
th
or
:
A
ni
k V
e
ga
V
it
ia
ni
ngs
ih
D
e
pa
r
tm
e
nt
of
I
nf
or
m
a
ti
c
s
U
ni
ve
r
s
it
a
s
D
r
. S
oe
to
m
o
J
a
la
n S
e
m
ol
ow
a
r
u 84 S
ur
a
ba
ya
, 60118, S
ur
a
ba
ya
, I
ndone
s
ia
E
m
a
il
:
ve
ga
@
uni
to
m
o.a
c
.i
d
1.
I
N
T
R
O
D
U
C
T
I
O
N
M
ode
l
a
c
c
ur
a
c
y
pr
e
di
c
ti
on
in
th
e
de
ve
lo
pm
e
nt
of
f
r
a
m
e
w
or
ks
on
G
I
S
s
of
twa
r
e
is
th
e
f
ir
s
t
s
te
p
in
e
f
f
or
ts
to
im
pr
ove
th
e
qua
li
ty
of
G
I
S
s
of
twa
r
e
de
ve
lo
pe
d
a
nd
i
s
pa
r
t
of
qua
li
ty
c
ont
r
ol
a
nd
qua
li
ty
a
s
s
ur
a
nc
e
[
1]
.
Q
ua
li
ty
c
ont
r
ol
w
il
l
de
t
e
r
m
in
e
th
e
m
e
th
od
of
s
pa
ti
a
l
a
na
ly
s
is
to
te
s
t
qu
a
li
ty
s
ta
nda
r
ds
[
1]
.
A
s
p
a
ti
a
l
a
na
ly
s
is
m
ode
li
ng
i
s
a
pr
oc
e
s
s
to
bui
ld
a
n
a
r
ti
f
ic
ia
l
in
te
ll
ig
e
nc
e
(
A
I
)
m
ode
l
th
a
t
is
c
om
bi
ne
d
w
it
h
tr
ia
ls
on
s
pa
ti
a
l
da
ta
s
e
t
s
[
2]
,
ga
th
e
r
in
g
s
pa
ti
a
l
knowle
dge
th
r
ough
s
pa
ti
a
l
da
ta
s
e
ts
a
nd
pr
ovi
di
ng
knowle
dge
of
m
ode
ls
in
th
e
f
r
a
m
e
w
or
k
th
r
ough
A
I
m
e
th
ods
f
r
om
va
r
io
us
s
our
c
e
s
.
T
he
pur
pos
e
of
th
e
s
pa
ti
a
l
a
na
ly
s
is
m
ode
l
is
to
m
a
ke
a
de
s
c
r
ip
ti
on
of
th
e
G
I
S
s
of
twa
r
e
th
a
t
w
i
ll
be
de
ve
lo
pe
d,
c
onduc
t
s
im
ul
a
ti
ons
to
te
s
t
s
pa
ti
a
l
da
ta
s
e
t
s
th
r
ough
m
ode
ls
on
th
e
A
I
m
e
th
od
u
s
e
d
on
th
e
pr
opo
s
e
d
f
r
a
m
e
w
or
k
th
a
t
ha
s
a
lr
e
a
dy
b
e
e
n
d
e
s
c
r
ib
e
d.
S
pa
ti
a
l
da
ta
s
e
ts
in
G
I
S
r
e
la
te
to
how
pr
im
a
r
y
a
nd
s
e
c
onda
r
y
da
ta
a
r
e
obt
a
in
e
d
th
r
ough
th
e
c
ol
le
c
ti
on
pr
oc
e
s
s
,
a
nd
th
e
n
how
th
e
da
ta
is
pr
oc
e
s
s
e
d
th
r
ough
s
pa
ti
a
l
a
na
ly
s
is
to
be
i
nf
or
m
a
ti
on
in
th
e
de
c
is
io
n
s
uppor
t
s
ys
te
m
[
3]
.
V
is
ua
li
z
a
ti
on
of
s
p
a
ti
a
l
da
ta
c
a
n
b
e
done
w
it
h
c
lo
ud
-
te
r
m
in
a
l
i
nt
e
gr
a
ti
on
G
I
S
to
pr
ovi
de
c
onve
ni
e
nc
e
in
th
e
pr
oc
e
s
s
of
s
pa
ti
a
l
a
na
ly
s
is
on
a
la
r
ge
num
be
r
of
s
pa
ti
a
l
da
t
a
s
e
ts
[
4]
,
a
ggr
e
ga
ti
on
-
ba
s
e
d
s
pa
ti
a
l
da
ta
s
e
ts
in
f
or
m
a
ti
on
r
e
t
r
ie
va
l
s
ys
te
m
[
5]
.
S
pa
ti
a
l
da
ta
s
e
ts
a
s
th
e
ke
y
to
th
e
va
lu
e
of
bi
g
da
ta
in
s
pa
ti
a
l
da
ta
m
in
in
g
(
S
D
M
)
th
a
t
r
e
f
e
r
s
to
th
e
de
s
c
r
ip
ti
on
of
a
tt
r
ib
ut
e
da
ta
r
e
qui
r
e
m
e
nt
s
,
how
th
e
da
ta
is
obt
a
in
e
d,
a
nd
w
h
a
t
A
I
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
,
V
ol
.
10
, N
o.
2, J
une
20
21
:
365
–
373
366
m
e
th
od
is
us
e
d
to
pe
r
f
or
m
s
pa
ti
a
l
a
na
ly
s
is
of
th
e
da
ta
[
6]
,
[
4]
.
S
pa
ti
a
l
da
ta
s
e
ts
be
c
om
e
th
e
ba
s
ic
s
tr
uc
tu
r
e
in
G
I
S
f
o
r
th
e
pr
oc
e
s
s
of
s
pa
ti
a
l
a
na
ly
s
is
a
lg
or
it
hm
s
,
a
na
ly
z
in
g
a
lg
or
it
hm
pr
in
c
ip
le
s
,
or
a
da
pt
in
g
e
xi
s
ti
ng
a
lg
or
it
hm
s
[
7]
.
T
he
c
la
s
s
if
ic
a
ti
on
m
ode
l
in
m
a
c
hi
ne
le
a
r
ni
ng
is
pr
e
va
le
nt
[
8]
to
be
us
e
d
r
e
s
e
a
r
c
h
in
th
e
f
ie
ld
of
s
pa
ti
a
l
a
na
ly
s
is
of
G
I
S
.
H
ow
e
ve
r
,
th
e
r
e
is
no
c
onc
r
e
te
s
ta
te
m
e
nt
r
e
ga
r
di
ng
w
hi
c
h
c
la
s
s
if
ic
a
ti
on
a
lg
or
it
h
m
is
be
s
t
to
us
e
w
it
h c
e
r
ta
in
ty
be
c
a
u
s
e
t
he
a
c
c
ur
a
c
y, pr
e
c
is
io
n, a
n
d r
e
c
a
ll
(
A
P
R
)
t
e
s
ts
i
n e
a
c
h s
tu
dy us
e
di
f
f
e
r
e
nt
s
a
m
pl
e
da
ta
. I
t
is
a
l
s
o ba
s
e
d on th
e
f
ie
ld
of
s
tu
dy, whic
h i
s
a
lwa
ys
ot
he
r
on t
he
obj
e
c
t
of
r
e
s
e
a
r
c
h c
ondu
c
te
d.
P
r
e
vi
ous
r
e
s
e
a
r
c
h
pr
opos
e
d
a
f
r
a
m
e
w
or
k
us
in
g
th
e
C
A
R
T
m
ode
l
(
c
la
s
s
if
ic
a
ti
on
a
nd
r
e
gr
e
s
s
io
n
tr
e
e
s
)
,
w
hi
c
h
r
e
por
te
d
a
10
-
f
ol
d
in
c
r
e
a
s
e
in
th
e
be
s
t
va
lu
e
f
o
r
c
r
a
s
h
s
e
ve
r
it
y
pr
e
di
c
ti
on
[
9]
.
H
ow
e
ve
r
,
th
e
C
A
R
T
m
ode
l
h
a
s
a
w
e
a
kne
s
s
in
th
e
num
be
r
of
tr
a
in
in
g
da
ta
s
a
m
pl
e
s
be
c
a
us
e
c
ha
ng
e
s
in
tr
a
in
in
g
a
nd
te
s
ti
ng
da
ta
s
a
m
pl
e
s
a
f
f
e
c
t
th
e
r
e
s
ul
t
s
of
s
pa
ti
a
l
a
n
a
ly
s
is
[
10]
.
S
pa
ti
a
l
a
na
ly
s
is
m
ode
l
us
in
g
d
a
ta
m
in
in
g
de
c
is
io
n
tr
e
e
(
J
48,
I
D
3,
a
nd
C
A
R
T
)
a
nd
na
ïv
e
b
a
ye
s
c
la
s
s
if
ie
r
s
[
11]
S
ta
te
s
th
a
t
th
e
a
c
c
ur
a
c
y
v
a
lu
e
of
96.30%
on
th
e
J
48
m
e
th
od
is
hi
ghe
r
th
a
n
I
D
3,
C
A
R
T
,
a
nd
na
ïv
e
b
a
ye
s
,
w
he
r
e
t
he
na
ïv
e
ba
ye
s
ha
v
e
be
tt
e
r
pe
r
f
or
m
a
nc
e
e
ve
n
th
ough
th
e
a
c
c
ur
a
c
y
va
lu
e
is
s
m
a
ll
.
D
if
f
e
r
e
nt
s
tu
di
e
s
s
ugg
e
s
t
t
ha
t
th
e
a
c
c
ur
a
c
y
of
pr
e
di
c
ti
on
of
c
la
s
s
if
ic
a
ti
on
m
ode
ls
w
it
h
th
e
de
c
is
io
n
tr
e
e
a
ppr
oa
c
h
to
r
e
a
c
h
84.1%
[
12]
.
A
ls
o,
in
di
c
a
te
th
a
t
th
e
e
nha
nc
e
d
e
m
pi
r
ic
a
l
ba
ye
s
ia
n
(
E
B
)
m
e
th
od
is
a
s
pa
ti
a
l
a
na
ly
s
is
a
ppr
oa
c
h
th
a
t
i
s
pr
e
f
e
r
r
e
d
f
or
pr
e
di
c
ti
on
of
th
e
num
be
r
of
acci
de
nt
s
in
r
oa
d
s
e
gm
e
nt
s
[
13]
.
M
a
xi
m
iz
e
s
th
e
a
c
c
ur
a
c
y
va
lu
e
of
t
he
m
ode
l
f
or
G
e
o
-
s
pa
ti
a
l
da
ta
us
in
g
th
e
a
da
pt
iv
e
k
-
ne
a
r
e
s
t
ne
ig
hbor
(
kN
N
)
c
la
s
s
if
ie
r
,
i.
e
.,
by
dyn
a
m
ic
a
ll
y
s
e
le
c
ti
ng
k
f
or
e
a
c
h
in
s
ta
n
c
e
,
th
e
va
lu
e
be
in
g
c
la
s
s
if
ie
d
r
e
a
c
h
e
s
a
R
O
C
A
U
C
s
c
or
e
of
0,9. T
he
f
uz
z
y
d
e
e
p
-
le
a
r
ni
ng
a
ppr
oa
c
h
m
ode
l
is
u
s
e
d
to
r
e
duc
e
th
e
unc
e
r
ta
in
ty
of
da
ta
in
th
e
pr
e
di
c
ti
on
of
tr
a
f
f
ic
f
lo
w
s
th
a
t
a
f
f
e
c
t
r
oa
d
tr
a
f
f
ic
a
c
c
id
e
nt
r
a
te
s
[
14]
.
C
onvolut
io
na
l
lo
ng
s
hor
t
-
te
r
m
m
e
m
or
y
(
C
onv
L
S
T
M
)
ne
ur
a
l
ne
twor
k
m
ode
l
[
15]
s
ta
te
s
th
a
t
th
e
pr
opos
e
d
f
r
a
m
e
w
or
k
is
s
uf
f
ic
ie
nt
ly
a
c
c
ur
a
te
a
nd
s
ig
ni
f
ic
a
nt
to
im
pr
ov
e
a
c
c
ur
a
c
y
in
tr
a
f
f
ic
a
c
c
id
e
nt
pr
e
di
c
ti
on
f
or
he
te
r
oge
ne
ous
da
ta
.
T
he
r
oa
d
a
c
c
id
e
nt
c
la
s
s
if
ic
a
ti
on
m
ode
l
u
s
in
g
r
a
ndom
f
o
r
e
s
ts
a
nd
boos
te
d
tr
e
e
s
w
or
ks
e
qua
ll
y w
e
ll
w
it
h a
n a
ve
r
a
ge
va
lu
e
of
80%
a
c
c
ur
a
c
y a
nd a
s
e
ns
i
ti
vi
ty
va
lu
e
of
50%
[
16]
.
T
he
di
s
c
us
s
io
n
in
th
is
pa
p
e
r
e
m
pha
s
i
z
e
s
th
e
c
om
pa
r
is
on
in
m
ode
li
ng
s
pa
ti
a
l
a
n
a
ly
s
is
us
in
g
c
la
s
s
if
ic
a
ti
on
m
e
th
ods
f
or
hybr
id
m
ode
ls
th
r
ough
th
e
p
r
opos
e
d
f
r
a
m
e
w
or
k.
T
he
ge
ne
r
a
l
c
ont
r
ib
ut
io
n
o
f
th
is
pr
opos
e
d
f
r
a
m
e
w
or
k
w
il
l
be
us
e
d
f
or
f
ut
u
r
e
w
or
k
is
in
te
gr
a
te
d
th
r
ough
th
e
G
I
S
-
pl
a
t
f
or
m
f
o
r
th
e
s
a
f
e
m
a
na
ge
m
e
nt
a
nd
r
is
k
a
s
s
e
s
s
m
e
nt
[
17]
,
[
18]
of
t
r
a
f
f
ic
a
c
c
id
e
nt
-
pr
one
r
oa
ds
c
la
s
s
if
ic
a
ti
on,
to
a
na
ly
z
e
m
ul
ti
-
c
r
it
e
r
ia
pa
r
a
m
e
te
r
s
th
a
t
in
f
lu
e
nc
e
th
e
r
e
s
ul
ts
on
th
e
tr
a
f
f
ic
a
c
c
id
e
nt
-
pr
one
r
oa
d
c
la
s
s
if
ic
a
ti
on,
to
pur
pos
e
ne
w
pa
r
a
m
e
te
r
s
of
s
pa
ti
a
l
da
ta
s
e
t
s
,
to
e
nha
nc
e
a
f
r
a
m
e
w
or
k
of
s
pa
t
ia
l
a
na
ly
s
is
us
in
g
a
hybr
id
e
s
ti
m
a
ti
on
m
ode
l
-
ba
s
e
d
on
a
c
om
bi
na
ti
on
of
M
C
D
M
-
A
N
N
s
,
a
nd
to
e
v
a
lu
a
te
th
e
e
nha
nc
e
m
e
nt
of
th
e
n
e
w
m
ode
l
th
r
ough
th
e
hybr
id
.
M
ode
l
e
va
lu
a
ti
on
ne
e
ds
to
be
done
to
pr
ovi
de
be
s
t
p
r
a
c
ti
c
e
s
f
or
th
e
r
e
s
ul
ti
ng
m
ode
l
[
19]
.
M
ode
l
pe
r
f
or
m
a
nc
e
a
s
s
e
s
s
m
e
nt
is
in
f
lu
e
nc
e
d
by
ba
la
nc
e
d
d
a
ta
to
de
s
c
r
ib
e
th
e
qua
li
ty
of
th
e
r
e
s
ul
ti
ng
m
ode
l,
s
o
a
s
not
to
le
a
d
to
m
is
le
a
di
ng
c
onc
lu
s
io
ns
[
16]
.
T
he
pr
opos
e
d
f
r
a
m
e
w
or
k
of
c
la
s
s
if
ic
a
ti
on
m
ode
ls
w
it
h
M
C
D
M
-
A
N
N
s
hybr
id
to
th
e
im
pl
e
m
e
nt
a
ti
on
of
pr
one
-
r
oa
ds
tr
a
f
f
ic
a
c
c
id
e
nt
c
la
s
s
if
ic
a
ti
on
a
nd
it
s
di
f
f
e
r
e
nc
e
s
w
it
h
e
xi
s
ti
ng
f
r
a
m
e
w
or
ks
a
r
e
pr
e
s
e
nt
e
d
of
c
l
a
s
s
if
ic
a
ti
on
m
od
e
ls
.
T
h
e
s
e
l
e
c
ti
on
of
a
m
ode
l
-
ba
s
e
d
hybr
id
e
s
ti
m
a
ti
on
on
a
c
om
bi
na
ti
on
of
M
C
D
M
-
A
N
N
s
c
la
s
s
if
ic
a
ti
on
in
th
is
pr
opos
e
d
f
r
a
m
e
w
or
k
s
tu
dy
is
ba
s
e
d
on
a
li
te
r
a
tu
r
e
r
e
vi
e
w
.
T
he
c
ol
le
c
ti
on
of
da
ta
s
e
t
m
ul
ti
-
c
r
it
e
r
ia
pa
r
a
m
e
te
r
f
or
pr
one
-
r
oa
ds
tr
a
f
f
ic
a
c
c
id
e
nt
c
la
s
s
if
ic
a
ti
on
w
hi
c
h
ha
s
be
e
n
us
e
d
in
th
e
pa
pe
r
a
r
ti
c
le
s
obt
a
in
e
d
to
e
va
lu
a
t
e
th
e
pr
opos
e
d
f
r
a
m
e
w
or
k
of
c
la
s
s
if
ic
a
ti
on
m
ode
ls
,
e
xpl
a
in
s
a
ls
o
th
e
va
li
da
ti
on
a
nd
e
va
lu
a
ti
on
te
c
hni
que
s
of
th
e
pr
opos
e
d
m
ode
l.
M
ode
li
ng
of
gr
oup
a
na
ly
ti
c
hi
e
r
a
r
c
hy
pr
oc
e
s
s
(
G
A
H
P
)
te
c
hni
que
to
de
v
e
lo
p
w
e
ig
ht
in
g
t
e
c
hni
que
on
m
ul
ti
-
pa
r
a
m
e
te
r
c
r
it
e
r
ia
a
ppl
ie
d
to
M
C
D
M
M
e
th
ods
w
hi
c
h
s
ti
ll
us
e
a
r
e
a
hum
a
n
a
s
s
um
pt
io
n
in
w
e
ig
ht
in
g,
pr
ovi
ng
th
r
ough
th
e
s
e
ns
it
iv
it
y
a
nd
s
ta
bi
li
ty
te
s
t
of
G
A
H
P
te
c
hni
que
m
ode
li
ng
to
M
C
D
M
m
e
th
ods
by
c
om
pa
r
in
g
th
e
w
e
ig
h
t
w
a
s
gi
ve
n
th
e
hum
a
n by ma
nua
l
a
s
s
um
pt
io
n.
M
ul
ti
-
c
r
it
e
r
ia
de
c
is
io
n
m
a
ki
ng
(
M
C
D
M
)
m
e
th
ods
a
r
e
u
s
e
d
in
th
is
s
tu
dy
to
pr
oc
e
s
s
th
e
d
e
te
r
m
in
a
nt
pa
r
a
m
e
te
r
da
ta
in
th
e
c
la
s
s
if
ic
a
ti
on
of
a
c
c
id
e
nt
-
pr
one
a
r
e
a
s
th
a
t
in
c
lu
de
r
oa
d
c
ondi
ti
ons
,
tr
a
f
f
ic
vol
um
e
,
a
c
c
id
e
nt
r
a
te
[
20]
,
[
21]
,
a
s
s
ig
n
w
e
ig
ht
in
g
va
lu
e
s
to
e
a
c
h
f
a
c
t
or
ba
s
e
d
on
li
te
r
a
tu
r
e
a
nd
s
ur
ve
y
s
to
e
xpe
r
t
s
our
c
e
s
[
22]
.
F
r
om
th
e
c
la
s
s
if
ic
a
ti
on
of
th
e
a
c
c
id
e
nt
-
pr
one
a
r
e
a
s
,
it
be
c
om
e
s
c
r
uc
ia
l
to
pr
ovi
de
r
e
c
om
m
e
nda
ti
ons
to
th
e
r
oa
d
a
udi
to
r
to
c
onduc
t
a
tr
a
f
f
ic
s
a
f
e
ty
a
udi
t
to
obt
a
in
a
s
s
e
s
s
m
e
nt
c
r
it
e
r
ia
,
im
pl
e
m
e
nt
a
ti
on
e
xpe
ns
e
s
,
th
e
num
be
r
of
in
vol
ve
d
tr
a
f
f
ic
pa
r
t
ic
ip
a
nt
s
,
th
e
e
f
f
e
c
t
of
r
oa
d
s
a
f
e
ty
,
pr
ot
e
c
ti
ve
e
f
f
e
c
t,
a
nd
s
oc
ia
l
f
a
c
to
r
s
pr
e
s
e
nt
in
g
di
f
f
ic
ul
ti
e
s
[
23]
.
T
he
tr
a
f
f
ic
s
a
f
e
ty
a
udi
t
is
c
a
r
r
ie
d
out
by
th
e
a
dm
in
is
tr
a
ti
on
of
th
e
r
oa
d
a
udi
to
r
by
c
o
nduc
ti
ng
a
f
e
a
s
ib
il
it
y
s
tu
dy
of
th
e
ne
twor
k
of
a
c
c
id
e
nt
-
pr
one
r
oa
d
c
a
te
gor
ie
s
[
24]
.
M
C
D
M
m
e
th
ods
ha
ve
be
e
n
us
e
d
f
or
a
na
ly
s
is
w
it
h
s
im
pl
e
a
ddi
ti
ve
w
e
ig
ht
(
S
A
W
)
,
a
na
ly
ti
c
a
l
hi
e
r
a
r
c
hy pr
oc
e
s
s
(
A
H
P
)
, a
nd f
uz
z
y A
H
P
m
e
th
od, us
e
d f
or
r
oa
d s
a
f
e
ty
a
na
ly
s
i
s
(
R
S
A
)
t
ha
t
c
a
n he
lp
de
c
i
s
io
ns
pr
oc
e
s
s
in
n
d
e
te
r
m
in
in
g
th
e
pr
io
r
it
y
of
r
oa
d
m
a
n
a
ge
m
e
nt
a
n
d
pr
ovi
de
m
it
ig
a
ti
ng
a
c
ti
ons
a
ga
in
s
t
th
e
m
os
t
vul
ne
r
a
bl
e
to
a
c
c
id
e
nt
s
[
25]
.
T
he
M
C
D
M
m
e
th
od
w
it
h
te
c
hni
que
f
or
or
de
r
pr
e
f
e
r
e
nc
e
by
s
im
il
a
r
it
y
to
id
e
a
l
s
ol
ut
i
on
(
T
O
P
S
I
S
)
m
e
th
od
is
us
e
d
in
th
e
m
a
na
ge
m
e
nt
of
r
oa
d
s
a
f
e
ty
,
a
nd
r
oa
d
s
a
f
e
ty
is
one
of
th
e
f
a
c
to
r
s
to
r
e
duc
e
th
e
num
be
r
of
tr
a
f
f
ic
a
c
c
id
e
nt
s
by
knowing
th
e
pos
it
io
n
of
a
r
oa
d
s
a
f
e
ty
s
tu
dy
in
B
us
he
hr
pr
ovi
nc
e
B
us
he
hr
-
B
or
a
z
ja
n
r
oa
ds
a
nd
B
or
a
z
ja
n
-
G
e
na
ve
h
ba
s
e
d
on
va
r
i
ous
qua
nt
it
a
ti
ve
a
nd
qua
li
ta
ti
ve
c
r
it
e
r
ia
[
26]
.
T
he
M
C
D
M
m
ode
l
i
s
one
of
th
e
r
ig
ht
a
ppr
oa
c
h
m
ode
ls
to
d
e
a
l
w
it
h
th
e
pr
obl
e
m
of
a
c
c
id
e
nt
-
pr
one
r
oa
d
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
Spat
ia
l
anal
y
s
is
m
ode
l
fo
r
t
r
af
fi
c
a
c
c
id
e
nt
-
pr
one
r
oads
c
la
s
s
if
ic
at
io
n
…
(
A
ni
k
V
e
ga V
it
ia
ni
ngs
ih
)
367
s
e
c
ti
on
(
A
P
R
S
)
be
c
a
u
s
e
it
u
s
e
s
s
e
v
e
r
a
l
r
oa
d
a
nd
e
nvi
r
onm
e
nt
a
l
c
r
it
e
r
ia
,
bot
h
qua
nt
it
a
ti
ve
or
qua
li
ta
ti
ve
;
M
C
D
M
i
s
r
e
la
te
d t
o t
he
r
e
s
ul
ts
of
de
c
is
io
n m
a
ki
ng f
or
pl
a
nni
ng t
ha
t
in
vol
ve
s
s
ta
ke
hol
de
r
s
[
27]
. A
f
r
a
m
e
w
or
k
to
be
pr
opos
e
d
th
r
ough
th
e
pr
oc
e
s
s
of
a
li
te
r
a
tu
r
e
r
e
vi
e
w
f
r
o
m
s
e
ve
r
a
l
s
tu
di
e
s
th
a
t
ha
ve
be
e
n
don
e
be
f
or
e
.
T
hi
s
pr
os
e
s
to
e
va
lu
a
te
th
e
be
ne
f
it
s
of
r
e
s
e
a
r
c
h
th
a
t
ha
s
be
e
n
done
,
to
know
th
e
li
m
it
a
ti
ons
of
th
e
m
e
th
od
us
e
d,
to
id
e
nt
if
y
r
e
s
e
a
r
c
h
ga
ps
th
a
t
ha
v
e
be
e
n
c
onduc
t
e
d,
a
nd
to
a
dvi
s
e
de
ve
lo
pm
e
nt
f
or
f
ur
th
e
r
r
e
s
e
a
r
c
h
to
ge
t
th
e
r
ig
ht
f
r
a
m
e
w
or
k
i
n t
he
r
e
s
e
a
r
c
h t
he
ne
w
[
28]
.
T
he
r
e
s
e
a
r
c
h que
s
ti
ons
i
n r
e
s
e
a
r
c
h a
r
e
i
nt
e
nde
d t
o
f
oc
us
on
th
e
s
ubj
e
c
t
a
r
e
a
of
th
e
s
tu
dy
by
id
e
nt
if
yi
ng
a
nd
c
la
s
s
if
yi
ng
th
e
s
pa
ti
a
l
a
na
ly
s
is
f
r
a
m
e
w
or
k
f
or
a
c
c
id
e
nt
-
pr
one
t
r
a
f
f
ic
r
oa
ds
t
o be
done
[
29]
.
2.
R
E
S
E
A
R
C
H
M
E
T
H
O
D
T
he
s
pa
ti
a
l
a
na
ly
s
is
m
ode
l
us
in
g
M
C
D
M
is
a
m
ul
ti
-
c
r
it
e
r
ia
s
pa
ti
a
l
de
c
is
io
n
s
uppor
t
s
y
s
te
m
(
M
C
-
S
D
S
S
)
de
ve
lo
pe
d
in
G
I
S
te
c
hnol
ogy
by
in
te
gr
a
ti
ng
M
C
D
M
a
s
a
m
e
th
od
to
de
te
r
m
in
e
th
e
be
s
t
a
lt
e
r
na
ti
ve
f
r
om
th
e
m
a
ny
c
hoi
c
e
s
a
va
il
a
bl
e
ba
s
e
d
on
th
e
s
pa
ti
a
l
da
ta
s
e
t
s
de
s
c
r
ib
e
d
[
30]
.
A
N
N
s
c
la
s
s
if
ic
a
ti
on
is
a
da
ta
m
in
in
g
te
c
hni
que
in
m
a
c
hi
ne
le
a
r
ni
ng,
m
a
ppi
ng
va
r
io
us
a
tt
r
ib
ut
e
s
a
s
in
put
la
ye
r
in
a
nod
e
,
a
ddi
ng
th
e
hi
dde
n
la
ye
r
,
w
hi
c
h
is
th
e
n
u
s
e
d
to
ge
t
th
e
th
r
e
s
hol
d
to
th
e
non
-
li
ne
a
r
out
put
la
ye
r
[
31]
.
T
he
pr
opos
e
d
f
r
a
m
e
w
or
k
w
it
h t
he
s
te
ps
i
n F
ig
ur
e
1.
T
he
in
it
ia
l
s
ta
ge
a
pr
opos
e
d
f
r
a
m
e
w
or
k
in
F
ig
ur
e
1
is
to
p
la
n
to
pi
c
s
a
nd
r
e
s
e
a
r
c
h
tr
e
nds
w
it
h
id
e
nt
if
yi
ng
in
r
e
s
e
a
r
c
h
ne
e
ds
f
or
th
e
li
te
r
a
tu
r
e
r
e
vi
e
w
p
r
oc
e
s
s
th
r
ough
s
ta
te
-
of
-
th
e
-
a
r
t
f
r
a
m
e
w
or
ks
,
m
e
th
ods
,
da
ta
s
e
ts
r
e
qui
r
e
m
e
nt
s
,
a
nd
ga
p
a
na
ly
s
is
of
e
xi
s
ti
ng
m
e
th
od
s
a
nd
f
r
a
m
e
w
or
ks
.
A
c
ti
on
a
da
pt
in
g,
im
pr
ovi
ng,
a
nd hybr
id
i
m
pl
e
m
e
nt
a
ti
on t
o m
ode
l
a
c
c
ur
a
c
y pr
e
di
c
ti
on i
n t
he
de
ve
lo
pm
e
nt
of
f
r
a
m
e
w
or
ks
. T
he
s
ta
te
-
of
-
th
e
-
a
r
t
f
r
om
t
he
l
it
e
r
a
tu
r
e
r
e
vi
e
w
w
it
hi
n t
he
pr
im
a
r
y s
tu
dy i
s
di
s
pl
a
ye
d i
n T
a
bl
e
1.
P
l
a
n
n
i
n
g
:
R
e
s
e
a
rc
h
T
o
p
i
c
s
a
n
d
T
re
n
d
s
S
t
a
t
e
-
of
-
t
h
e
-
a
rt
D
a
t
a
s
et
s
S
t
a
t
e
-
of
-
t
h
e
-
a
rt
Met
h
o
d
s
G
a
p
A
n
a
l
y
s
i
s
o
f
t
h
e
E
x
i
s
t
i
n
g
M
et
h
o
d
s
a
n
d
F
ra
mew
o
rk
s
S
t
a
t
e
-
of
-
t
h
e
-
a
rt
F
ra
m
ew
o
rk
s
A
c
t
i
o
n
:
A
d
a
p
t
i
n
g
,
Im
p
ro
v
i
n
g
,
a
n
d
H
y
b
ri
d
Im
p
l
emen
t
a
t
i
o
n
t
o
mo
d
el
a
c
c
u
ra
c
y
p
red
i
c
t
i
o
n
i
n
t
h
e
d
evel
o
p
men
t
o
f
f
ra
mew
o
rk
s
F
ig
ur
e
1.
R
e
s
e
a
r
c
h m
e
th
od s
t
e
ps
T
he
li
te
r
a
tu
r
e
r
e
vi
e
w
is
in
T
a
bl
e
1.
T
h
e
r
e
s
e
a
r
c
h
[
32]
not
s
how
n
th
e
c
om
pa
r
is
on
of
th
e
a
c
c
ur
a
c
y
a
nd
c
ons
is
te
nc
y
of
e
a
c
h
m
e
th
od
us
e
d
w
i
th
th
e
c
onf
us
io
n
m
a
tr
ix
.
T
he
m
e
a
ni
ng
of
e
m
pi
r
ic
a
l
ba
ye
s
ha
s
th
e
be
s
t
a
c
c
ur
a
c
y
a
nd
c
ons
i
s
te
nc
y
va
lu
e
th
a
t
is
not
r
e
a
ll
y
vi
s
ib
le
. T
he
s
t
a
nda
r
d
de
vi
a
ti
on
of
th
e
da
ta
di
s
tr
ib
ut
io
n
va
lu
e
in
t
he
s
a
m
pl
e
da
ta
i
s
onl
y us
e
d t
o c
a
lc
ul
a
te
t
he
di
s
a
s
te
r
-
pr
one
t
r
a
f
f
ic
a
c
c
id
e
nt
r
a
te
, a
nd t
he
r
e
i
s
no pr
oof
of
t
he
tr
ut
h
of
th
e
m
ode
l
us
e
d
[
33]
.
D
is
c
us
s
io
n
[
34]
is
s
ti
ll
li
m
it
e
d
to
th
e
us
e
of
a
n
e
xi
s
ti
ng
m
e
th
od,
a
nd
knowle
dge
c
om
bi
na
ti
on
ha
s
not
be
e
n
done
a
s
a
hybr
id
m
ode
l
a
ppr
oa
c
h.
T
he
r
e
s
ul
ts
of
th
e
c
om
pa
r
is
on
of
th
e
two
m
e
th
ods
a
r
e
s
ta
te
d
to
be
m
or
e
a
c
c
ur
a
te
,
but
no
pr
e
c
is
e
a
c
c
ur
a
c
y
va
lu
e
is
gi
ve
n
ba
s
e
d
on
th
e
va
lu
e
of
th
e
c
onf
us
io
n
m
a
tr
ix
[
35]
.
O
n
r
e
s
e
a
r
c
h
[
36]
ha
ve
not
c
ons
id
e
r
e
d
th
e
ty
pe
of
r
oa
d
ty
pe
de
s
ig
n,
f
or
e
xa
m
pl
e
,
a
r
te
r
ia
l
r
oa
ds
,
c
ol
le
c
to
r
r
oa
ds
,
or
r
oa
ds
b
a
s
e
d
on
th
e
ir
na
tu
r
e
(
ge
om
e
tr
ic
r
oa
d)
,
th
e
r
e
a
r
e
no
s
tu
di
e
s
on
a
da
pt
iv
e
m
ode
ls
th
a
t
c
a
n
e
xpa
nd
m
a
c
hi
ne
le
a
r
ni
ng
th
r
ough
a
c
om
bi
na
ti
on
of
onl
in
e
le
a
r
ni
ng
a
nd
de
e
p
le
a
r
ni
ng
[
37]
.
P
a
pe
r
di
s
c
us
s
io
n
[
38]
is
s
ti
ll
li
m
it
e
d
to
th
e
u
s
e
o
f
a
n
e
xi
s
ti
ng
m
e
th
od;
knowle
dg
e
c
om
bi
na
ti
on
ha
s
not
be
e
n
done
a
s
a
hybr
id
m
ode
l
a
ppr
oa
c
h.
T
h
e
D
T
R
m
ode
l
in
c
onduc
ti
ng
th
e
pr
e
di
c
ti
on
a
c
c
ur
a
c
y
in
th
is
s
tu
dy i
s
s
ti
ll
a
m
a
c
r
o
-
le
ve
l
c
r
a
s
h c
ount
[
39]
.
T
he
m
ode
l
ha
s
w
e
a
kne
s
s
e
s
i
n t
e
r
m
s
of
da
ta
s
im
ul
a
ti
on be
c
a
us
e
i
t
r
e
qui
r
e
s
a
c
c
id
e
nt
da
ta
a
t
th
e
be
gi
nni
ng
of
th
e
c
a
lc
ul
a
ti
on
[
27
]
.
M
a
th
e
m
a
ti
c
a
l
m
od
e
li
ng
in
th
e
c
om
pa
r
is
on
a
lg
or
it
hm
doe
s
not
e
xi
s
t,
s
o
th
e
c
om
pa
r
is
on
of
r
e
s
ul
ts
is
di
f
f
ic
ul
t
[
40]
;
th
e
r
e
is
no
e
va
lu
a
ti
on
of
th
e
m
ode
ls
of
f
e
r
e
d
be
c
a
us
e
th
e
te
s
t
da
ta
c
ol
le
c
t
e
d
doe
s
not
ha
v
e
a
lo
n
g
-
ti
m
e
s
pa
n
[
16]
.
M
L
P
is
m
or
e
a
c
c
ur
a
te
f
or
a
va
il
a
bl
e
s
pa
ti
a
l
da
ta
s
e
ts
but
be
c
om
e
s
ve
r
y
vul
ne
r
a
bl
e
w
he
n
th
e
r
e
is
da
ta
noi
s
e
th
a
t
c
a
n
c
a
us
e
e
r
r
or
s
in
pr
e
di
c
ti
ons
[
34]
.
P
N
N
ha
s
pr
oba
bi
li
s
ti
c
out
put
s
w
it
h
m
ul
ti
la
ye
r
pe
r
c
e
pt
r
on
ne
twor
ks
,
pr
oduc
in
g
f
a
ir
ly
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
,
V
ol
.
10
, N
o.
2, J
une
20
21
:
365
–
373
368
a
c
c
ur
a
te
p
r
e
di
c
ti
ons
[
34]
.
R
B
F
is
v
e
r
y
w
e
a
k
in
m
a
ki
ng
pr
e
di
c
ti
ons
[
34]
.
V
K
T
pa
r
a
m
e
te
r
s
pr
ove
d
to
be
th
e
m
os
t
in
f
lu
e
nt
ia
l
in
r
oa
d
tr
a
f
f
ic
a
c
c
id
e
nt
s
,
th
e
n
th
e
V
/C
va
r
ia
bl
e
a
nd
dr
iv
e
r
s
pe
e
d
ba
s
e
d
on
th
e
R
R
e
li
e
f
F
a
lg
or
it
hm
c
a
lc
ul
a
ti
on
m
e
th
od
[
34]
.
T
he
e
va
lu
a
ti
on
to
pe
r
f
o
r
m
t
he
te
c
hni
que
,
T
he
s
it
e
c
on
s
is
te
n
c
y
te
s
t
(
S
C
T
)
,
T
he
m
e
th
od
c
ons
i
s
te
nc
y
te
s
t
(
M
C
T
)
,
T
he
to
ta
l
r
a
nk
di
f
f
e
r
e
nc
e
s
te
s
t
(
T
R
D
T
)
,
a
nd
T
he
to
ta
l
s
c
or
e
te
s
t
(
T
S
T
)
[
38]
.
T
a
bl
e
1. L
it
e
r
a
tu
r
e
r
e
vi
e
w
s
a
f
r
a
m
e
w
or
k c
om
pa
r
is
on
F
r
a
m
e
w
or
k
M
ode
l
a
nd
m
e
t
hod
S
pa
t
i
a
l
d
a
t
a
s
e
t
s
R
e
s
ul
t
s
[
32]
M
ode
l
-
ba
s
e
d
s
pa
t
i
a
l
s
t
a
t
i
s
t
i
c
a
l
m
e
t
hods
:
P
oi
s
s
on r
e
gr
e
s
s
i
on,
N
e
ga
t
i
ve
B
i
nom
i
a
l
r
e
gr
e
s
s
i
on,
E
m
pi
r
i
c
a
l
B
a
ye
s
i
a
n.
T
he
a
c
c
i
de
nt
s
, i
nj
ur
i
e
s
,
a
nd de
a
t
hs
by ye
a
r
s
I
n
t
hi
s
s
t
udy
c
om
pa
r
i
ng
a
l
l
m
e
t
hods
u
s
e
d,
w
he
r
e
E
m
pi
r
i
c
a
l
B
a
ye
s
ha
s
t
he
be
s
t
a
c
c
ur
a
c
y
a
nd
c
ons
i
s
t
e
nc
y,
r
e
c
om
m
e
nd
e
d
by
t
he
H
i
ghw
a
y
S
a
f
e
t
y
M
a
nua
l
(
H
S
M
)
a
nd t
he
E
ur
ope
a
n U
ni
on A
c
qui
s
[
33]
M
ode
l
-
ba
s
e
d
s
pa
t
i
a
l
s
t
a
t
i
s
t
i
c
a
l
m
e
t
hods
:
K
e
r
ne
l
de
ns
i
t
y
a
na
l
ys
i
s
, N
e
a
r
e
s
t
ne
i
ghbor
, K
-
f
unc
t
i
on
I
nt
e
r
c
i
t
y a
c
c
i
de
nt
s
,
a
c
c
i
de
nt
s
l
e
a
d
i
ng t
o
i
nj
ur
y, a
c
c
i
de
nt
s
l
e
a
di
ng
t
o de
a
t
h, a
nd a
c
c
i
de
nt
s
l
e
a
di
ng t
o da
m
a
ge
s
T
he
obs
e
r
ve
d
va
l
ue
c
ur
ve
on
t
he
s
pa
t
i
a
l
a
na
l
ys
i
s
pr
oc
e
s
s
,
t
he
va
l
ue
of
s
pa
t
i
a
l
da
t
a
s
e
t
s
i
s
a
bove
t
he
5%
c
onf
i
de
nc
e
i
nt
e
r
va
l
[
36]
S
pa
t
i
a
l
a
na
l
ys
i
s
t
e
c
hni
que
s
:
N
e
a
r
e
s
t
N
e
i
ghbor
hood
H
i
e
r
a
r
c
hi
c
a
l
(
N
N
H
)
C
l
us
t
e
r
i
ng,
S
pa
t
i
a
l
-
T
e
m
por
a
l
C
l
us
t
e
r
i
ng
A
na
l
ys
i
s
(
S
T
A
C
)
R
oa
d a
c
c
i
de
nt
s
i
nvol
vi
ng a
l
l
t
ype
s
of
ve
hi
c
l
e
s
T
he
r
e
s
ul
t
s
of
t
he
s
pa
t
i
a
l
a
na
l
ys
i
s
va
r
y
a
c
c
or
di
ng
t
o
t
he
pa
r
a
m
e
t
e
r
va
l
ue
s
i
n
t
he
s
pa
t
i
a
l
da
t
a
s
e
t
s
,
w
h
e
r
e
i
s
S
T
A
C
ha
s
a
461,57 hi
ghe
r
P
r
e
di
c
t
i
on A
c
c
u
r
a
c
y I
nde
x
(
P
A
I
)
c
om
pa
r
e
d t
o N
N
H
163,69.
[
34]
A
N
N
s
t
e
c
hni
que
s
:
E
xt
r
e
m
e
l
e
a
r
ni
ng m
a
c
hi
ne
(
E
L
M
)
,
P
r
oba
bi
l
i
s
t
i
c
ne
u
r
a
l
ne
t
w
or
k
(
P
N
N
)
, R
a
di
a
l
ba
s
i
s
f
unc
t
i
on
(
R
B
F
)
, a
nd M
ul
t
i
l
a
ye
r
pe
r
c
e
pt
r
on (
M
L
P
)
.
V
/
C
, s
pe
e
d, ve
hi
c
l
e
ki
l
om
e
t
e
r
t
r
a
ve
l
e
d
(
V
K
T
)
, r
oa
dw
a
y w
i
dt
h,
t
he
e
xi
s
t
e
nc
e
of
m
e
di
a
n,
a
nd a
l
l
ow
a
bl
e
/
not
‐
a
l
l
ow
a
bl
e
pa
r
ki
ng
E
va
l
ua
t
i
on m
e
t
hod us
i
ng N
a
s
h
–
S
ut
c
l
i
f
f
e
(
N
S
)
, m
e
a
n
a
bs
ol
ut
e
e
r
r
or
(
M
A
E
)
,
a
nd
r
oot
m
e
a
ns
s
qu
a
r
e
e
r
r
or
(
R
M
S
E
)
.
E
L
M
,
a
s
a
f
e
e
d
-
f
or
w
a
r
d
ne
ur
a
l
ne
t
w
or
k,
be
c
om
e
s
t
he
a
l
gor
i
t
hm
t
ha
t
ha
s
t
he
be
s
t
pe
r
f
or
m
a
nc
e
a
nd
t
he
m
o
s
t
a
c
c
ur
a
t
e
pr
e
di
c
t
i
on
r
e
s
ul
t
s
(
R
M
S
E
=3,576;
N
S
=0,81;
M
A
E
=2,5062)
by
r
a
ndom
l
y
s
e
l
e
c
t
i
ng hi
dde
n
node
s
u
s
i
ng r
a
ndom
w
e
i
ght
s
.
[
35]
H
ot
s
pot
a
na
l
ys
i
s
(
G
e
t
i
s
-
O
r
d
G
i
*)
:
N
e
t
w
or
k s
pa
t
i
a
l
w
e
i
ght
s
,
K
e
r
ne
l
D
e
ns
i
t
y m
e
t
hod
T
he
t
r
a
f
f
i
c
a
c
c
i
de
nt
)
H
ot
s
pot
a
na
l
ys
i
s
gi
ve
s
be
t
t
e
r
r
e
s
ul
t
s
be
c
a
u
s
e
i
t
i
s
done
by c
ons
i
de
r
i
ng t
he
w
e
i
ght
of
s
pa
t
i
a
l
da
t
a
s
e
t
s
[
37]
T
he
s
uppor
t
ve
c
t
or
m
a
c
hi
ne
c
om
bi
ne
s
t
he
t
e
c
hni
que
s
of
s
t
a
t
i
s
t
i
c
a
l
l
e
a
r
ni
ng, m
a
c
hi
ne
l
e
a
r
ni
ng, t
he
ne
ur
a
l
ne
t
w
or
ks
ba
s
e
d:
S
uppor
t
ve
c
t
or
m
a
c
hi
ne
,
D
e
e
p ne
ur
a
l
ne
t
w
or
k
A
c
c
i
de
nt
, pe
r
s
on,
ve
hi
c
l
e
, r
oa
d, a
nd
e
nvi
r
onm
e
nt
da
t
a
T
he
y
pr
opos
e
d
a
r
e
a
l
-
t
i
m
e
onl
i
ne
de
e
p
l
e
a
r
ni
ng
f
r
a
m
e
w
or
k
B
a
s
e
d
on
t
r
a
f
f
i
c
a
c
c
i
de
nt
bl
a
c
k
s
pot
s
.
S
V
M
a
l
gor
i
t
hm
i
n
m
a
c
hi
ne
l
e
a
r
ni
ng
ha
s
63%
pr
e
c
i
s
i
on
a
nd
a
61%
r
e
c
a
l
l
r
a
t
e
i
n
a
na
l
yz
i
ng
t
he
bl
a
c
k
s
pot
s
of
t
r
a
f
f
i
c
a
c
c
i
de
nt
s
.
I
f
t
he
t
r
a
i
ni
ng
da
t
a
pe
r
i
od
i
s
a
dde
d,
t
he
S
V
M
a
nd
de
e
p
ne
ur
a
l
ne
t
w
or
k
va
l
ue
s
i
nc
r
e
a
s
e
by
95%
a
nd
89
%
a
c
c
ur
a
c
y,
69%
,
a
nd
79%
r
e
c
a
l
l
r
a
t
e
s
.
[
38]
B
l
a
c
k s
pot
i
de
nt
i
f
i
c
a
t
i
on (
B
S
I
D
)
m
e
t
hod a
nd S
e
gm
e
nt
a
t
i
on
m
e
t
hod:
E
m
pi
r
i
c
a
l
B
a
ye
s
i
a
n
(
E
B
)
, E
xc
e
s
s
E
m
pi
r
i
c
a
l
B
a
y
e
s
i
a
n (
E
E
B
)
, A
c
c
i
de
nt
F
r
e
que
nc
y (
A
F
)
, A
c
c
i
de
nt
R
a
t
e
(
A
R
)
.
T
he
t
r
a
f
f
i
c
a
c
c
i
de
nt
A
F
m
e
t
hod
ha
s
t
he
be
s
t
pe
r
f
or
m
a
nc
e
w
i
t
h
a
c
ons
i
s
t
e
nc
y
of
93.1%
c
om
p
a
r
e
d
t
o
E
B
92.2%
,
a
nd
E
E
B
77.4%
.
T
he
pe
r
f
or
m
a
nc
e
of
t
he
E
E
B
a
nd
A
R
m
e
t
hods
i
s
t
he
w
e
a
ke
s
t
i
n
t
he
c
a
s
e
of
s
e
gm
e
nt
a
t
i
on
i
n
m
os
t
c
a
s
e
s
of
s
e
gm
e
nt
a
t
i
on.
[
39]
M
a
c
hi
ne
l
e
a
r
ni
ng t
e
c
hni
que
s
t
o
pr
e
di
c
t
i
on m
ode
l
:
D
e
c
i
s
i
on t
r
e
e
r
e
gr
e
s
s
i
on (
D
T
R
)
m
e
t
hods
R
e
gr
e
s
s
i
on t
r
e
e
f
r
a
m
e
w
or
k,
E
ns
e
m
bl
e
t
e
c
hni
que
s
. M
od
e
l
a
s
s
e
s
s
m
e
nt
:
A
ve
r
a
ge
S
qua
r
e
d
E
r
r
or
(
A
S
E
)
, S
t
a
nda
r
d
D
e
vi
a
t
i
on of
E
r
r
o
r
(
S
D
E
)
S
t
a
t
e
w
i
de
T
r
a
f
f
i
c
A
na
l
ys
i
s
Z
one
(
S
T
A
Z
)
T
he
D
T
R
m
ode
l
t
o
p
r
e
di
c
t
i
on
a
c
c
ur
a
c
y
w
or
ks
be
t
t
e
r
t
ha
n
t
he
s
pa
t
i
a
l
D
T
R
m
ode
l
.
T
o
i
m
pr
ove
pr
e
di
c
t
i
on
a
c
c
ur
a
c
y
us
i
ng
e
ns
e
m
bl
e
t
e
c
hni
que
s
(
ba
ggi
ng,
r
a
ndom
f
or
e
s
t
,
a
nd
gr
a
di
e
nt
boos
t
i
ng)
w
i
t
h
s
l
i
ght
l
y
be
t
t
e
r
r
e
s
ul
t
s
,
de
pe
ndi
ng
on
t
he
a
m
ount
of
t
r
a
i
ni
ng
da
t
a
.
[
27]
M
ul
t
i
c
r
i
t
e
r
i
a
de
c
i
s
i
on m
a
ki
ng
(
M
C
D
M
)
m
ode
l
:
W
e
i
ght
e
d
l
i
ne
a
r
c
om
bi
na
t
i
on (
W
L
C
)
m
e
t
hod
T
he
t
r
a
f
f
i
c
a
c
c
i
de
nt
R
e
por
t
s
T
he
m
ode
l
w
a
s
d
e
ve
l
ope
d
t
o
de
t
e
r
m
i
ne
t
he
c
r
i
t
e
r
i
a
w
e
i
ght
s
t
ha
t
ha
v
e
be
e
n
de
t
e
r
m
i
ne
d
by
e
xp
e
r
t
s
w
i
t
h
i
nt
e
r
e
s
t
i
n s
ubj
e
c
t
i
ve
r
e
s
ul
t
s
.
[
40]
P
r
e
di
c
t
i
on m
ode
l
:
D
e
e
p ne
ur
a
l
ne
t
w
or
k m
ode
l
, G
e
ne
e
xpr
e
s
s
i
on pr
ogr
a
m
m
i
ng (
G
E
P
)
,
R
a
ndom
e
f
f
e
c
t
ne
ga
t
i
ve
bi
nom
i
a
l
(
R
E
N
B
)
m
ode
l
s
,
R
e
gul
a
r
ne
ga
t
i
ve
bi
nom
i
a
l
m
ode
l
(
F
E
N
B
)
T
he
r
oa
d ge
om
e
t
r
y,
t
r
a
f
f
i
c
, a
nd r
oa
d
e
nvi
r
onm
e
nt
)
T
he
D
N
N
m
ode
l
e
xpe
r
i
e
nc
e
d
a
n
i
nc
r
e
a
s
e
i
n
r
oa
d
pr
e
di
c
t
i
on
w
i
t
h
0.914
(
R
M
S
E
=7.474)
by
G
E
P
,
a
nd
0.891
(
R
M
S
E
=8.862)
.
G
E
P
w
or
ks
be
t
t
e
r
t
ha
n
R
E
N
B
t
o
m
e
a
s
ur
e
t
he
r
a
nki
ng
o
f
va
r
i
a
bl
e
s
t
ha
t
i
n
f
l
ue
nc
e
a
c
c
i
de
nt
s
.
[
16]
R
a
ndom
e
f
f
e
c
t
s
ne
ga
t
i
ve
bi
nom
i
a
l
m
ode
l
:
H
i
e
r
a
r
c
hi
c
a
l
c
l
us
t
e
r
m
e
t
hod
R
e
a
l
t
i
m
e
-
f
r
e
que
nc
y of
a
c
c
i
de
n
t
da
t
a
a
nd
c
ont
r
i
but
i
ng f
a
c
t
or
s
T
he
m
ode
l
de
ve
l
ope
d
c
a
n
pr
ovi
de
i
nf
o
r
m
a
t
i
on
on
t
he
m
a
i
n c
a
us
e
s
of
a
c
c
i
de
nt
s
a
t
r
oa
d i
nt
e
r
s
e
c
t
i
ons
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
Spat
ia
l
anal
y
s
is
m
ode
l
fo
r
t
r
af
fi
c
a
c
c
id
e
nt
-
pr
one
r
oads
c
la
s
s
if
ic
at
io
n
…
(
A
ni
k
V
e
ga V
it
ia
ni
ngs
ih
)
369
3.
R
E
S
U
L
T
S
A
ND
D
I
S
C
U
S
S
I
O
N
T
he
pr
opos
e
d
f
r
a
m
e
w
or
k
is
ba
s
e
d
on
a
li
te
r
a
tu
r
e
s
tu
dy
by
c
om
pa
r
in
g
th
e
e
xi
s
ti
ng
f
r
a
m
e
w
or
k
to
de
te
r
m
in
e
th
e
pe
r
f
or
m
a
nc
e
of
th
e
s
pa
ti
a
l
a
na
ly
s
i
s
m
ode
l
of
f
e
r
e
d
f
or
tr
a
f
f
ic
a
c
c
id
e
nt
pr
one
r
oa
d
s
in
T
a
bl
e
1
a
nd
th
e
M
C
D
M
-
ba
s
e
d
f
r
a
m
e
w
or
k
[
27]
,
[
41]
-
[
44]
.
T
he
f
r
a
m
e
w
or
k
be
in
g
c
om
pa
r
e
d
in
c
lu
de
s
th
e
m
e
th
od
us
e
d
f
or
pr
im
a
r
y
s
tu
dy
(
P
S
)
s
pa
ti
a
l
a
na
ly
s
is
f
or
a
c
c
id
e
nt
-
pr
one
tr
a
f
f
ic
r
oa
ds
,
th
e
s
p
a
ti
a
l
a
na
ly
s
i
s
m
ode
l
u
s
e
d,
s
pa
ti
a
l
da
ta
s
e
t
s
us
e
d
to
te
s
t
th
e
m
ode
l
th
r
ough
m
e
th
od
s
e
le
c
ti
on,
a
nd
th
e
va
lu
e
of
th
e
m
e
a
s
ur
e
m
e
nt
r
e
s
ul
ts
th
r
ough the
a
s
s
e
s
s
m
e
nt
.
T
he
f
r
a
m
e
w
or
k
th
a
t
ha
s
be
e
n
de
ve
lo
pe
d
by
pr
e
vi
ous
r
e
s
e
a
r
c
h
e
r
s
w
il
l
be
de
s
c
r
ib
e
d
in
th
is
s
e
c
ti
on.
T
he
f
r
a
m
e
w
or
k
m
o
de
l
[
44]
,
w
a
s
de
v
e
lo
pe
d
to
c
r
e
a
te
th
e
M
a
yc
oc
k
a
nd
H
a
ll
’
s
a
c
c
id
e
nt
pr
e
di
c
ti
on
m
ode
l.
T
hi
s
m
ode
l
pr
ovi
de
s
s
e
ns
it
iv
it
y
a
na
ly
s
is
on
m
ode
li
ng
r
e
s
ul
ts
us
in
g
m
ul
ti
-
obj
e
c
ti
ve
opt
im
iz
a
ti
on
(
M
O
O
)
us
in
g
m
ul
ti
-
c
r
it
e
r
ia
de
c
is
io
n
m
a
ki
ng
f
or
th
e
a
na
ly
ti
c
a
l
hi
e
r
a
r
c
hy
pr
o
c
e
s
s
(
A
H
P
m
ode
l)
.
T
he
ne
e
ds
pr
im
a
r
y
s
pa
ti
a
l
da
ta
s
e
ts
in
th
e
r
oa
d
ge
om
e
tr
y
c
a
te
gor
y,
th
e
ne
c
e
s
s
it
y
s
e
c
onda
r
y
s
pa
ti
a
l,
i.
e
.,
th
e
num
be
r
s
a
nd
ty
pe
s
of
tr
a
f
f
ic
a
c
c
id
e
nt
s
,
tr
a
f
f
ic
a
nd
de
m
a
nd
f
or
s
tr
uc
tu
r
a
l
f
lo
w
,
vi
s
ua
l
d
is
ta
nc
e
a
nd
ve
hi
c
le
s
pe
e
d,
r
oa
d
s
ig
ns
,
a
nd
e
qui
pm
e
nt
,
li
ght
in
g,
d
r
iv
e
r
be
ha
vi
or
.
T
he
va
lu
e
of
th
e
m
ul
ti
-
c
r
it
e
r
ia
pa
r
a
m
e
te
r
s
obt
a
in
e
d
w
il
l
be
done
m
a
th
e
m
a
ti
c
s
pa
ti
a
l
da
ta
m
ode
li
ng
to
pr
oduc
e
th
e
s
e
ns
it
iv
it
y
o
f
s
pa
ti
a
l
a
na
ly
s
is
,
th
e
r
e
s
ul
ts
of
m
ul
ti
-
c
r
i
te
r
ia
opt
im
iz
a
ti
on
in
th
e
f
or
m
o
f
tr
a
f
f
ic
e
f
f
ic
ie
nc
y
(
T
S
)
,
a
nd
tr
a
f
f
ic
s
a
f
e
ty
(
T
S
)
to
th
e
pr
e
di
c
te
d
tr
a
f
f
ic
a
c
c
id
e
nt
.
M
O
O
m
ode
l
is
m
e
a
s
ur
e
d
us
in
g
a
c
ons
is
te
nc
y
in
de
x
(
C
I
)
a
nd
c
ons
is
te
nc
y
r
a
ti
o
(
C
R
)
,
th
e
m
ode
l
is
pr
ove
n
t
o
ha
ve
a
good
s
tr
uc
tu
r
e
w
it
h
a
va
lu
e
of
C
R
≤10,
or
th
e
C
R
va
lu
e
is
0.00298;
th
is
s
how
s
th
a
t
th
e
M
O
O
m
ode
l
w
it
h M
C
D
M
on t
he
A
H
P
m
ode
l
ha
s
a
c
on
s
is
te
nt
v
a
lu
e
t
he
good
one
.
T
he
[
41]
f
r
a
m
e
w
or
k
w
a
s
de
ve
lo
pe
d
by
th
e
P
R
O
M
E
T
R
E
E
-
R
S
M
C
D
M
m
ode
l.
M
C
D
M
is
us
e
d
be
c
a
us
e
it
c
a
n
us
e
m
or
e
th
a
n
one
pa
r
a
m
e
te
r
to
ge
t
th
e
be
s
t
r
e
s
u
lt
s
f
r
om
th
e
a
lt
e
r
na
ti
ve
s
pr
oduc
e
d.
T
hi
s
m
ode
l
w
a
s
de
ve
lo
pe
d
to
e
va
lu
a
t
e
th
e
D
E
A
a
nd
T
O
P
S
I
S
m
e
th
ods
in
r
oa
d
s
a
f
e
ty
to
r
e
duc
e
r
is
k
th
e
num
be
r
o
f
a
c
c
id
e
nt
s
on
th
e
r
oa
d
th
r
ough
th
e
r
oa
d
s
a
f
e
ty
in
de
x.
T
he
m
o
de
l
is
te
s
te
d
by
u
s
in
g
th
e
R
obus
tn
e
s
s
of
th
e
c
om
pos
it
e
in
de
x.
T
he
a
ve
r
a
g
e
c
or
r
e
la
ti
on
v
a
lu
e
,
th
e
a
ve
r
a
ge
r
a
nk
va
lu
e
,
a
nd
th
e
c
lu
s
te
r
v
a
r
ia
ti
on
a
ve
r
a
ge
va
lu
e
s
w
il
l
be
e
nt
e
r
e
d
in
t
o
th
e
M
C
D
M
P
R
O
M
E
T
H
E
E
-
R
S
to
te
s
t
th
e
r
e
s
ul
ti
ng
m
ode
l.
M
ul
ti
-
c
r
it
e
r
ia
pa
r
a
m
e
te
r
s
te
s
te
d
in
th
is
m
ode
l,
i.
e
.,
th
e
P
ol
ic
e
D
e
pa
r
tm
e
nt
da
ta
,
f
a
ta
li
ti
e
s
,
s
e
r
io
us
in
ju
r
ie
s
,
num
be
r
of
in
ha
bi
ta
nt
s
,
num
be
r
of
r
e
gi
s
te
r
e
d
ve
hi
c
le
s
,
tr
a
f
f
ic
r
is
k,
a
nd
publ
ic
r
is
k.
T
hi
s
p
a
r
a
m
e
te
r
w
il
l
be
us
e
d
to
m
a
th
e
m
a
ti
c
s
p
a
ti
a
l
da
ta
m
ode
li
ng
th
r
ough
D
E
A
a
nd
T
O
P
S
I
S
,
th
e
pr
oduc
e
opt
im
a
l
c
om
pos
it
e
in
d
e
x
th
r
ough
th
e
va
lu
e
of
f
in
a
l
r
is
k
e
f
f
ic
ie
nc
y.
D
E
A
-
W
R
pr
ovi
de
s
th
e
be
s
t
r
a
nki
ng
r
e
s
ul
ts
c
om
pa
r
e
d
to
th
e
D
E
A
-
ba
s
e
d
c
om
pos
it
e
i
ndi
c
a
to
r
m
ode
l
(
D
E
A
-
C
I
)
.
T
he
[
42]
,
[
43]
f
r
a
m
e
w
or
k
is
a
m
ode
l
bui
lt
us
in
g
M
C
D
M
.
T
he
pur
pos
e
of
th
is
m
ode
l
is
to
c
r
e
a
te
a
knowle
dge
da
ta
m
in
in
g
r
ul
e
de
c
is
io
n
tr
e
e
th
r
ough
F
P
-
gr
ow
th
a
nd
a
pa
c
he
s
pa
r
k
f
r
a
m
e
w
or
k.
A
tr
ia
l
m
ode
l
on
r
oa
d
a
c
c
id
e
nt
a
na
ly
s
i
s
,
w
h
e
r
e
th
e
r
e
s
ul
t
s
ha
ve
a
hi
gh
de
gr
e
e
of
a
c
c
ur
a
c
y
a
nd
w
or
k
w
e
ll
to
im
pr
ove
r
oa
d
s
a
f
e
ty
.
T
h
e
m
ul
ti
-
pa
r
a
m
e
te
r
c
r
it
e
r
ia
us
e
d
a
r
e
th
e
r
oa
d
a
c
c
id
e
nt
da
ta
to
de
a
th
a
nd
in
ju
r
ie
s
a
tt
r
ib
ut
e
. T
he
te
s
ti
ng
m
ode
l
f
or
th
e
r
e
le
va
nt
a
s
s
oc
ia
ti
on
r
ul
e
is
done
by
te
s
ti
ng
a
nd
va
li
da
ti
on
by
m
e
a
s
ur
in
g
qua
li
ty
m
e
a
s
ur
e
m
e
nt
.
M
C
D
M
m
ode
l
in
vol
ve
s
m
a
ny
c
r
it
e
r
ia
,
s
o
it
is
s
ui
ta
bl
e
to
o
ve
r
c
om
e
th
e
pr
obl
e
m
of
a
c
c
id
e
nt
-
pr
one
r
oa
d
s
e
c
ti
on
(
A
P
R
S
)
on
th
e
ty
pe
of
hor
iz
ont
a
l
a
li
gnm
e
nt
,
ve
r
ti
c
a
l
a
li
gnm
e
nt
,
in
te
r
s
e
c
ti
ons
,
s
ig
ni
f
ic
a
nt
pl
a
c
e
s
,
a
nd
sh
oul
de
r
w
id
th
s
w
it
h a
n a
c
c
ur
a
c
y va
lu
e
of
0.8830 f
or
t
hr
e
s
hol
d va
lu
e
s
1
[
27]
.
T
he
pr
opos
e
d
f
r
a
m
e
w
or
k
in
pr
e
vi
ous
r
e
s
e
a
r
c
h
w
il
l
be
us
e
d
by
th
e
a
ut
hor
a
s
a
r
e
f
e
r
e
nc
e
in
de
ve
lo
pi
ng f
ur
th
e
r
a
c
ti
vi
ti
e
s
of
t
he
f
r
a
m
e
w
or
k t
ha
t
w
il
l
be
p
r
op
os
e
d. T
he
f
r
a
m
e
w
or
k of
t
he
r
e
s
e
a
r
c
h pr
opos
e
d
in
F
ig
ur
e
2
ha
s
th
e
m
a
in
di
f
f
e
r
e
nc
e
s
f
r
om
th
e
e
xi
s
ti
ng
f
r
a
m
e
w
or
k.
P
r
e
pa
r
e
da
ta
r
e
qui
r
e
m
e
nt
s
f
or
s
pa
ti
a
l
da
ta
s
e
ts
a
s
pr
im
a
r
y
a
nd
s
e
c
onda
r
y
s
pa
ti
a
l
da
ta
s
e
ts
in
de
te
r
m
in
in
g
th
e
r
oa
d
c
a
te
gor
ie
s
to
be
s
tu
di
e
d
(
us
in
g
a
pr
iv
a
te
or
publ
ic
s
pa
ti
a
l
da
t
a
s
e
t
ty
pe
)
.
P
e
r
f
or
m
a
li
te
r
a
tu
r
e
s
tu
d
y
r
e
la
ti
ng
to
m
ul
ti
-
c
r
it
e
r
ia
pa
r
a
m
e
te
r
s
us
e
d
on
e
a
c
h
r
oa
d
c
a
te
gor
y.
M
a
th
e
m
a
ti
c
s
m
ode
li
ng
f
or
s
pa
ti
a
l
a
na
l
ys
is
to
th
e
pr
opos
e
d
f
r
a
m
e
w
or
k
f
or
hybr
i
d
e
s
ti
m
a
ti
on
m
ode
l
-
ba
s
e
d
on
a
c
om
bi
na
ti
on
of
M
C
D
M
-
A
N
N
s
m
ul
ti
-
c
la
s
s
c
la
s
s
if
ic
a
ti
on.
I
n
th
is
c
a
s
e
,
th
e
pr
e
-
pr
oc
e
s
s
in
g
da
ta
pr
oc
e
s
s
w
il
l
r
un
f
or
th
e
c
la
s
s
if
ic
a
ti
on
a
na
ly
s
i
s
pr
oc
e
s
s
.
T
he
r
a
nge
of
c
la
s
s
if
ic
a
ti
on
w
il
l
be
pe
r
f
or
m
e
d
th
r
ough
m
a
th
e
m
a
ti
c
a
l
m
ode
li
ng
u
s
in
g
th
e
G
ut
tm
a
n
m
e
th
od.
T
he
r
e
s
ul
ts
of
th
e
m
ul
ti
-
c
la
s
s
c
la
s
s
if
ic
a
ti
on
w
il
l
be
va
li
da
te
d
w
it
h
S
C
T
,
M
C
T
,
a
nd
A
R
C
va
li
da
ti
on.
F
oc
us
e
s
on
th
e
pr
opos
e
a
c
la
s
s
if
ic
a
ti
on
of
r
oa
ds
pr
one
to
a
c
c
id
e
nt
s
us
in
g
m
ul
ti
pl
e
c
r
it
e
r
ia
pa
r
a
m
e
te
r
s
(
da
ta
s
e
r
ie
s
)
,
m
a
ke
m
ode
li
ng
of
r
oa
d
p
r
one
to
a
c
c
id
e
nt
s
by
c
a
lc
ul
a
ti
ng
th
e
va
lu
e
of
tr
a
f
f
ic
a
c
c
id
e
nt
by
ty
pe
of
e
ve
nt
s
a
nd
th
e
in
de
x
of
th
e
a
c
c
id
e
nt
s
,
th
e
de
ns
it
y
th
a
t
of
r
oa
d
s
tr
a
f
f
ic
a
c
c
id
e
nt
h
a
ppe
ne
d
to
e
a
c
h
z
one
a
n
d
th
e
a
m
ount
of
da
ta
in
e
a
c
h
ye
a
r
,
r
is
k
f
a
c
to
r
s
ba
s
e
d
on
th
e
s
e
ve
r
it
y
of
th
e
a
c
c
id
e
nt
s
, s
e
ve
r
it
y
of
r
oa
ds
tr
a
f
f
ic
a
c
c
id
e
nt
e
v
e
nt
s
,
c
r
a
s
h
pr
e
di
c
ti
on
m
ode
l
s
us
in
g
da
ta
s
e
r
ie
s
,
a
nd
th
e
va
lu
e
of
th
e
s
oc
ie
ta
l
c
os
t
of
e
a
c
h
ty
pe
t
he
a
c
c
id
e
nt
.
T
h
e
A
N
N
s
tr
a
te
gy
h
a
s
th
e
m
os
t
not
e
w
or
th
y
r
a
ti
ng
of
te
c
hni
que
s
th
a
t
a
r
e
r
e
gul
a
r
ly
ut
il
iz
e
d
in
th
e
li
te
r
a
tu
r
e
r
e
vi
e
w
in
e
s
s
e
nt
i
a
l
c
ons
id
e
r
s
.
T
he
e
m
pi
r
ic
a
l
B
a
ye
s
m
e
th
od
a
nd
de
c
is
io
n
tr
e
e
in
da
ta
m
in
in
g
a
r
e
a
l
s
o
br
oa
dl
y
us
e
d
w
it
hi
n
th
e
c
lu
s
te
r
in
g
c
a
te
gor
y
in
s
pa
ti
a
l
in
f
or
m
a
ti
on
m
ode
li
ng
o
f
a
c
c
id
e
nt
-
pr
one
z
one
s
.
T
hi
s
c
ons
id
e
r
s
a
pr
opos
e
d
f
r
a
m
e
w
or
k
of
c
la
s
s
if
ic
a
ti
on
us
e
d
a
hybr
id
e
s
ti
m
a
ti
on
m
ode
l
ba
s
e
d
on
a
c
om
bi
na
ti
on
of
M
C
D
M
-
A
N
N
c
la
s
s
if
ic
a
ti
on.
T
e
s
t
th
e
c
ons
is
te
nc
y
of
th
e
m
e
th
od
f
r
om
th
e
m
ode
l
pr
oduc
e
d
w
it
h
th
e
M
C
T
,
S
C
T
,
a
nd
th
e
va
lu
e
of
A
R
C
m
ode
l
e
va
lu
a
ti
ons
.
A
N
N
s
c
la
s
s
if
ic
a
ti
on
m
e
th
ods
a
r
e
th
e
m
o
s
t
popula
r
da
ta
m
in
in
g
te
c
hni
que
s
in
th
e
f
ie
ld
s
pa
ti
a
l
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
,
V
ol
.
10
, N
o.
2, J
une
20
21
:
365
–
373
370
a
na
ly
s
is
of
a
c
c
id
e
nt
-
pr
one
r
oa
ds
a
nd
th
e
f
a
c
to
r
s
th
a
t
a
f
f
e
c
t
th
e
a
c
c
id
e
nt
r
a
te
,
a
m
ong
ot
h
e
r
s
(
ne
ur
a
l
ne
twor
k
s
,
e
xt
r
e
m
e
l
e
a
r
ni
ng ma
c
hi
ne
s
, k
-
ne
a
r
e
s
t
ne
ig
hbor
, na
iv
e
ba
ye
s
, de
c
is
io
n t
r
e
e
s
)
[
45]
,
[
31]
.
P
r
op
os
e
d
H
yb
r
i
d
of
M
C
D
M
-
A
N
N
s
C
l
as
s
i
f
i
c
at
i
on
F
r
am
e
w
or
k
P
r
e
p
a
r
a
t
i
o
n
S
p
a
t
i
a
l
D
a
t
a
s
e
t
s
M
e
t
h
o
d
C
l
a
s
s
i
fi
c
a
t
i
o
n
fo
r
M
C
D
M
M
o
d
e
l
M
a
t
h
e
m
a
t
i
c
M
o
d
el
l
i
n
g
F
r
a
m
e
w
o
r
k
S
y
s
t
e
m
M
u
l
t
i
-
C
l
a
s
s
C
l
a
s
s
i
fi
c
a
t
i
o
n
T
e
s
t
i
n
g
&
Va
l
i
d
a
t
i
o
n
:
-
P
r
ec
i
s
i
o
n
,
R
ec
a
l
l
,
A
c
c
u
c
a
r
y
(
A
R
C
)
-
M
et
h
o
d
C
o
n
s
i
s
t
en
c
y
T
es
t
(
M
C
T
)
-
S
i
t
e
C
o
n
s
i
s
t
en
c
y
T
es
t
(
S
C
T
)
M
u
l
t
i
-
C
r
i
t
e
r
i
a
P
a
r
a
m
e
t
e
r
t
o
S
p
a
t
i
a
l
D
a
t
a
s
e
t
s
D
a
t
a
P
r
e
-
P
r
o
c
e
s
s
i
n
g
C
l
a
s
s
i
fi
c
a
t
i
o
n
A
n
a
l
y
s
i
s
M
u
l
t
i
-
C
l
a
s
s
C
l
a
s
s
i
fi
c
a
t
i
o
n
(
1
t
o
n
)
R
a
n
g
e
C
l
a
s
s
i
f
i
c
a
t
i
o
n
M
et
h
o
d
t
r
u
e
f
a
l
s
e
M
C
D
M
H
y
b
r
i
d
t
o
A
r
t
i
f
i
c
i
a
l
N
eu
r
a
l
N
e
t
w
o
r
k
(
A
N
N
)
C
l
a
s
s
i
fi
c
a
t
i
o
n
D
et
er
m
i
n
e t
h
e
r
a
n
k
i
n
g
v
a
l
u
e
t
o
c
l
a
s
s
i
fy
P
r
i
ma
r
y
&
S
e
c
o
n
d
a
r
y
R
o
a
d
N
e
t
w
o
r
k
:
-
A
r
t
e
r
i
a
l
r
o
a
d
-
C
o
l
l
e
c
t
o
r
r
o
a
d
-
L
o
c
a
l
r
o
a
d
S
e
c
o
n
d
a
r
y
S
p
a
t
i
a
l
D
a
t
a
s
e
t
s
:
-
t
h
e
g
e
o
m
e
t
r
i
c
r
o
a
d
-
t
h
e
p
a
v
e
m
e
n
t
r
o
a
d
-
t
h
e
e
n
v
i
r
o
n
me
n
t
a
l
r
o
a
d
I
d
e
n
t
i
fy
a
s
s
e
s
s
m
e
n
t
t
o
s
e
c
o
n
d
a
r
y
s
p
a
t
i
a
l
d
a
t
a
s
e
t
s
F
ig
ur
e
2.
P
r
opos
e
d
hybr
id
of
M
C
D
M
-
A
N
N
s
c
la
s
s
if
ic
a
ti
on
f
r
a
m
e
w
or
k
t
o e
va
lu
a
te
a
nd r
a
nk s
p
a
ti
a
l
a
na
ly
s
i
s
m
ode
l
tr
a
f
f
ic
a
c
c
id
e
nt
pr
one
r
oa
ds
4.
C
O
N
C
L
U
S
I
O
N
T
he
pr
opos
e
d
f
r
a
m
e
w
or
k
in
th
is
s
tu
dy
w
il
l
a
c
t
a
s
a
m
ode
l
-
ba
s
e
d
hybr
id
e
s
ti
m
a
ti
on
a
ppr
oa
c
h
on
a
c
om
bi
na
ti
on
of
M
C
D
M
-
A
N
N
s
c
la
s
s
if
ic
a
ti
on
to
s
tr
e
ngt
he
n
da
ta
m
in
in
g
te
c
hni
que
s
in
s
pa
ti
a
l
m
ul
ti
-
c
r
it
e
r
ia
a
na
ly
s
is
in
m
ul
ti
-
c
la
s
s
c
la
s
s
if
ic
a
ti
on
de
c
i
s
io
n
m
a
ki
ng.
I
n
th
e
li
t
e
r
a
tu
r
e
r
e
vi
e
w
on
th
e
p
r
im
a
r
y
s
tu
dy,
th
e
r
e
a
r
e
no
r
e
s
e
a
r
c
h
to
pi
c
s
th
a
t
di
s
c
us
s
on
th
e
tr
a
f
f
ic
a
c
c
id
e
nt
-
pr
one
r
oa
ds
c
la
s
s
if
ic
a
ti
on
on
th
e
a
r
te
r
ia
l
r
oa
d,
c
ol
le
c
to
r
r
oa
d,
a
nd
ty
pe
of
r
oa
d
ba
s
e
d
on
it
s
na
tu
r
e
(
pa
ve
m
e
nt
,
ge
om
e
tr
y,
a
nd
lo
c
a
l
r
oa
d)
c
a
te
go
r
ie
s
.
T
he
s
pa
ti
a
l
a
na
ly
s
is
m
ode
l
u
s
in
g
M
C
D
M
a
m
ong
ot
he
r
s
,
a
n
a
ly
ti
c
hi
e
r
a
r
c
h
y
pr
oc
e
s
s
(
A
H
P
)
,
a
na
ly
ti
c
a
l
ne
twor
k
pr
oc
e
s
s
(
A
N
P
)
, w
e
ig
ht
e
d s
um
m
ode
l
(
W
S
M
)
, w
e
ig
ht
e
d pr
oduc
t
(
W
P
)
,
w
e
ig
ht
pr
oduc
t
m
ode
l
(
W
P
M
)
, s
im
pl
e
a
ddi
ti
ve
w
e
ig
ht
in
g
(
S
A
W
)
,
te
c
hni
que
f
or
or
de
r
pr
e
f
e
r
e
nc
e
by
s
im
il
a
r
it
y
to
id
e
a
l
s
ol
ut
io
n
(
T
O
P
S
I
S
)
,
pr
e
f
e
r
e
nc
e
r
a
nki
ng
or
ga
ni
z
a
ti
on
m
e
th
od
f
or
e
nr
ic
hm
e
nt
of
e
va
lu
a
ti
ons
(
P
R
O
M
E
T
H
E
E
)
,
m
ul
ti
-
a
tt
r
ib
ut
e
ut
il
i
ty
th
e
or
y
(
M
A
U
T
)
,
e
li
m
in
a
ti
on
a
nd
c
hoi
c
e
e
xpr
e
s
s
in
g
r
e
a
li
ty
(
E
L
E
C
T
R
E
)
,
a
nd
vl
s
e
kr
it
e
r
ij
us
ka
opt
im
iz
a
c
ij
a
i
kom
or
om
is
no
r
e
s
e
nj
e
(
V
I
K
O
R
)
.
T
h
e
r
e
s
ul
t
s
of
th
e
be
s
t
m
e
th
ods
th
r
ough
A
P
R
m
e
a
s
ur
e
m
e
nt
w
il
l
be
a
r
e
f
e
r
e
nc
e
in
de
c
is
io
n
m
a
ki
ng
in
r
oa
d
m
a
na
ge
m
e
nt
.
E
xi
s
ti
ng
r
e
s
e
a
r
c
h
is
s
ti
ll
li
m
it
e
d
to
one
ty
pe
of
r
oa
d
us
e
d
a
s
a
n obje
c
t
(
s
pe
c
if
ic
r
e
gi
on)
, a
nd 96%
i
s
us
e
d pr
iv
a
te
s
pa
ti
a
l
da
ta
s
e
ts
.
I
n t
hi
s
s
tu
dy, i
t
w
a
s
us
in
g a
n I
nduc
ti
ve
q
ua
li
ta
ti
ve
a
ppr
oa
c
h i
n t
he
m
ode
li
ng of
r
oa
d pr
one
t
o a
c
c
id
e
nt
s
t
o i
de
nt
if
y t
he
f
in
di
ngs
of
s
c
ie
nc
e
t
ha
t
is
done
dur
in
g
th
e
r
e
s
e
a
r
c
h
pr
oc
e
s
s
.
T
h
e
pr
opos
e
d
a
c
la
s
s
if
ic
a
ti
on
of
r
oa
ds
pr
one
to
a
c
c
id
e
nt
s
us
in
g
m
ul
ti
pl
e
c
r
it
e
r
ia
pa
r
a
m
e
te
r
s
, m
a
ke
a
m
ode
li
ng of
r
oa
d pr
one
t
o a
c
c
id
e
nt
s
c
a
lc
ul
a
ti
ng by the
va
lu
e
of
t
r
a
f
f
ic
a
c
c
id
e
nt
by t
ype
o
f
e
ve
nt
s
a
nd
th
e
in
de
x
of
th
e
a
c
c
id
e
nt
s
,
t
he
va
lu
e
of
th
e
de
ns
it
y
t
ha
t
of
r
oa
ds
t
r
a
f
f
ic
a
c
c
id
e
nt
ha
ppe
ne
d
to
e
a
c
h
z
one
a
nd
th
e
a
m
ount
of
da
ta
in
e
a
c
h
ye
a
r
,
t
he
va
lu
e
of
r
is
k
f
a
c
t
or
s
ba
s
e
d
on
th
e
s
e
ve
r
it
y
of
th
e
a
c
c
id
e
nt
s
,
th
e
va
lu
e
of
s
e
ve
r
it
y
of
r
oa
ds
tr
a
f
f
ic
a
c
c
id
e
nt
e
ve
nt
s
,
t
he
v
a
lu
e
of
c
r
a
s
h
pr
e
di
c
ti
on
m
ode
ls
,
t
he
va
lu
e
of
th
e
s
oc
ie
ta
l
c
o
s
t
of
e
a
c
h t
ype
t
he
a
c
c
id
e
nt
, a
nd t
he
te
s
t
r
e
s
ul
t
is
u
s
in
g t
he
m
e
th
od t
he
S
C
T
, t
he
M
C
T
,
a
nd A
P
R
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
Spat
ia
l
anal
y
s
is
m
ode
l
fo
r
t
r
af
fi
c
a
c
c
id
e
nt
-
pr
one
r
oads
c
la
s
s
if
ic
at
io
n
…
(
A
ni
k
V
e
ga V
it
ia
ni
ngs
ih
)
371
A
C
K
N
O
WL
E
D
G
E
M
E
N
T
S
T
hi
s
r
e
s
e
a
r
c
h
is
s
uppor
te
d
by
U
ni
ve
r
s
it
i
T
e
kni
k
a
l
M
a
la
y
s
ia
M
e
la
ka
,
M
a
la
y
s
ia
,
a
nd
U
ni
ve
r
s
it
a
s
D
r
.
S
oe
to
m
o,
I
ndone
s
ia
.
T
he
de
ve
lo
pm
e
nt
of
r
e
s
ul
ts
of
a
s
tu
dy
f
unde
d
b
y
T
he
D
ir
e
c
to
r
a
te
G
e
ne
r
a
l
of
S
tr
e
ngt
he
ni
ng
R
e
s
e
a
r
c
h
a
nd
D
e
ve
lo
pm
e
nt
of
R
e
s
e
a
r
c
h,
T
e
c
hnol
ogy,
a
nd
H
ig
he
r
E
duc
a
ti
on
M
in
is
tr
y
-
I
ndone
s
ia
i
n 2015
-
2016
.
R
E
F
E
R
E
N
C
E
S
[1]
J.
Albrecht,
“GIS
Project
Management,”
in
Comprehe
nsive
Geograph
ic
Informatio
n
Systems
,
Elsevier
Inc.,
2018,
pp. 446
–
477.
[2]
A.
Banerjee
and
S.
Ray,
“Spatial
models
and
geographic
information
systems,”
Encyclopedia
of
Ecology,
2n
d
Edition
. Elsevier Inc.,
pp. 1
–
10, 2018.
[3]
K.
E.
Brassel
and
R.
Weibel,
“A
review
and
conc
eptual
framewor
k
of
automated
map
generalization,”
Int.
J.
Geogr. Inf. Syst.
, vol. 2, no. 3, pp. 229
–
244, 1988.
[4]
S.
Wang,
Y.
Zhong,
and
E.
Wang,
“An
integrated
GIS
platform
architecture
for
spatiotemporal
big
data,”
Futur
.
Gener. Comput. Syst
.
, vol. 94, no. May, pp. 160
–
172, 2019.
[5]
J.
Lacasta,
F.
J.
Lopez
-
Pellice
r,
B.
Espejo
-
García,
J.
Nogueras
-
Iso,
a
nd
F.
J.
Zarazaga
-
Soria,
“Aggregation
-
base
d
information
retrieva
l
system
for
geospatia
l
data
catalogs,”
Int.
J.
Ge
ogr.
Inf.
Sci.
,
vol.
31,
no.
8,
pp.
1583
–
1605,
2017.
[6]
D.
Li,
S.
Wang,
H.
Yuan,
and
D.
Li,
“Software
and
applications
of
spatial
data
mining,”
Wiley
Interdiscip.
Rev.
Data Min. Kno
wl. Discov.
, vol. 6, no. 3, pp. 84
–
114, 2016.
[7]
L.
Zhao,
L.
Chen,
R.
Ranjan,
K.
K.
R.
Choo,
and
J.
He,
“Geographical
information
system
parallelization
for
spatial big data pro
cessing: a rev
iew,”
Cluster Comp
ut.
, vol. 19, no. 1, pp. 139
–
152, 2016.
[8]
N.
F.
Hordri,
A.
Samar,
S.
S.
Yuhaniz,
and
S.
M.
Shamsudd
in,
“A
sy
stemati
c
litera
ture
revie
w
on
featu
res
of
deep
learning in
big data a
nalytics,”
Int. J. Adv.
Soft Co
mput. it
s Appl.
, vol. 9, no. 1, pp. 32
–
49, 2017.
[9]
M.
Effati
and
A.
Sadeghi
-
Niaraki,
“A
s
emantic
-
based
classification
and
regression
tree
approach
for
modelling
complex
spatial
rules
in
motor
vehicle
crashes
domain,
”
Wiley
Interdiscip.
Rev.
Data
Min.
Knowl.
Discov.
,
vol.
5,
no. 4, pp. 181
–
194, 2015.
[10]
M
.
A.
Raihan,
M.
Hossain,
and
T.
Hasan,
“Data
mining
in
road
crash
analysis:
the
context
of
developing
countries
,”
Int. J. Inj
. Contr.
Saf. Pro
mot.
, vol. 25, no. 1, pp. 41
–
52, 2018.
[11]
T.
K.
Bahiru,
D.
Kumar
Singh,
and
E.
A.
Tessfaw,
“Comparative
Stu
dy
on
Data
Mining
Classification
Algorithm
s
for
Predicting
Road
Traffic
Accident
Severity,”
in
2018
Seco
nd
International
Conference
on
Inventiv
e
Communica
tion and Com
putational Te
chnologie
s (ICICCT)
, 2018, pp. 1655
–
1660.
[12]
Z.
Zheng,
P.
Lu,
and
D.
Tolliver,
“
Decision
Tree
App
roach
to
A
ccident
Prediction
for
Highway
–
Rail
Grade
Crossings,”
Transp. R
es. Re
c. J.
Transp.
Res. B
oard
, vol. 2545, no. 1, pp. 115
–
122, 2016.
[13]
A.
S.
Lee,
W.
H.
Lin,
G.
S.
Gill,
and
W.
Cheng,
“An
enhanced
em
pirical
bayesian
method
for
i
dentifying
roa
d
hotspots
and
predicting
number
of
crashes,”
Journal
of
Transportatio
n
Safety
and
Security
,
vol.
0,
no.
0,
Taylor
&
Franc
is, pp.
1
–
17, 2018.
[14]
W.
Chen
et
al.
,
“A
novel
fuzzy
deep
-
learning
approac
h
to
traffic
flo
w
prediction
with
uncerta
in
spatial
–
temporal
data features,”
Futur. Gener. Comput. Syst.
, vol. 89, no. June, pp. 78
–
88, 2018.
[15]
M.
Kibanov,
M.
Becker,
M.
Atzmueller,
and
A.
Hotho,
“Ada
ptive
kNN
Using
Expected
Accuracy
for
Classific
ation
of
Geo
-
Spatial
Data,
”
in
Proceedings
of
the
33rd
Ann
ual
ACM
Symposium
on
Applied
Computing
Pages 857
-
865
, 2018, pp. 857
–
865.
[16]
M.
Schlögl,
“A
multivariate
analysis
of
environmental
effects
on
road
accident
occurrence
using
a
balanced
bagging approach,”
Accid. Anal. Prev.
, vol. 136, no. March, pp
. 1
–
12, 2020.
[17]
W.
Li
and
S.
Wang,
“PolarGlobe:
A
web
-
wide
virtual
globe
system
fo
r
visualizing
multidimensional,
time
-
varying,
big climate data,”
Int. J. Geogr
. Inf. Sci.
, vol. 31, no. 8, pp. 1562
–
1582, 2017.
[18]
M.
P.
Repetto,
M.
Burlando,
G.
Solari,
P.
De
Gaetano,
M.
Pizzo,
and
M.
Tizzi,
“A
web
-
based
GIS
platform
for
the
safe
management
and
risk
assessment
of
complex
structural
and
infrastructural
systems
exposed
to
wind,”
Adv.
Eng. Softw.
, vol. 117, pp. 29
–
45, 2018.
[19]
H.
Gong,
F.
Wang,
B.
Brenda,
and
S.
Dent,
“Application
of
random
ef
fects
negative
binomial
model
with
clustered
dataset for vehicle cr
ash frequency ana
lysis,”
Int. J. Tr
ansp. Sci.
Technol.
, no. April, pp. 1
–
12, 2020.
[20]
T.
Sipos,
“Spatial
statistical
analysis
of
the
traffi
c
accidents,”
Period
.
Polytech.
Transp.
Eng.
,
vol.
45,
no.
2,
pp.
101
–
105, 2017.
[21]
A.
V.
Vitianingsih
and
D.
Cahyono,
“Geographical
Information
Syst
em
for
Mapping
Road
Using
Multi
-
Attribute
Utility Me
thod,” in
Internati
onal Con
ference on Scien
ce and Technol
ogy
-
Computer
(ICST)
, 2016, pp. 0
–
4.
[22]
R.
Al
-
Ruzouq,
K.
Hamad,
S.
Abu
Dabous,
W.
Zeiada
,
M.
A.
Kha
lil,
and
T.
Voigt,
“Weigh
ted
Multi
-
attribut
e
Framew
ork
to
Identi
fy
Freew
ay
Incide
nt
Hot
Spots
in
a
Spatiote
mpo
ral
Context,
”
Arab.
J.
Sci.
Eng.
,
vol.
44,
no.
10, pp. 8205
–
8223, 2019.
[23]
Á.
Török,
“Statistical
Analysis
of
a
Multi
-
Criteria
Assessme
nt
of
Inte
lligent
Traff
ic
Systems
for
the
Improve
men
t
of Road Safety,”
J. Finan
c. Econ.
, vol. 4, no. 5, pp.
127
–
135, 2016.
[24]
Y.
Huvarinen,
E.
Svatkova,
E.
Oleshchenko,
and
S.
Pushchin
a,
“Roa
d
Safety
Audit,”
in
Transport
ation
Resea
rch
Procedia
, 2017, vol. 20, no. September 2016, pp. 236
–
241.
[25]
S.
Kanuganti,
R.
Agarwala,
B.
Dutta,
P.
N.
Bhanegaonkar,
A.
P.
S
in
gh,
and
A.
K.
Sarkar,
“Road
safety
analysis
using
multi
criteria
approach:
A
case
study
in
India,”
in
Transport
ation
Resea
rch
Proce
dia
,
2017,
vol.
25,
pp.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
,
V
ol
.
10
, N
o.
2, J
une
20
21
:
365
–
373
372
4653
–
5665.
[26]
M.
S.
Fatemeh
Haghighat,
“Application
of
a
Multi
-
Criteria
Approac
h
To
Road
Safety
E
valuation
in
the
Busheh
r
Provinc
e , I
ran,”
Traffic P
lan. Pre
lim. Com
mun.
, vol. 23, no. 5, pp. 341
–
352, 2011.
[27]
F.
Yakar,
“A
multicriteria
decision
making
–
bas
ed
methodology
to
identify
accident
-
prone
road
sections,”
J.
Transp. S
af. Sec
ur.
, pp. 1
–
15, 2019
.
[28]
B.
Kitchenham
and
S.
Charters,
“Guidelines
for
performing
S
ystematic
Literature
Reviews
in
Software
Engineering,”
Engineering
, vol. 2, p. 1051, 2007.
[29]
M.
Vierhauser,
R.
Rabiser,
and
P.
Grünbacher,
“Requirements
monit
oring
frameworks:
A
systemati
c
review,”
Inf.
Softw. Technol.
, vol. 80, pp. 89
–
109, 2016.
[30]
S. M. Gh
avami,
“Multi
-
criteria s
patial d
ecision
support
system
for
ide
ntifyin
g strateg
ic roads
in dis
aster si
tuation
s,”
Int. J. Cr
it. Infr
astruct.
Prot.
, vol. 24, pp. 23
–
36, 2019.
[31]
N.
San
g
and
M.
Aitkenhead,
“Data
Mining,
Machine
Learning
an
d
Spatial
Data
Infrastructures
for
Scenari
o
Modelling,” in
Modelling
Nature
-
based Solutions
, 2020, pp. 276
–
304.
[32]
M. A. Dereli a
nd S. Erdogan, “A
new model for
determining the tra
ffic acc
ident black
spots using GIS
-
aided sp
atial
statistical methods,”
Transp. R
es. Pa
rt A Po
licy P
ract.
, vol. 103, no. September, pp. 106
–
117, 2017.
[33]
G.
A.
Shafabakhsh,
A.
Famili,
and
M.
S.
Bahadori,
“GIS
-
based
spa
tial
analysis
of
urban
traffic
accidents:
Case
study in
Mashhad, Ira
n,”
J. Traf
fic Tr
ansp. E
ng.
, vol. 4, no. 3, pp. 290
–
299, 2017.
[34]
H.
Behbahani
and
A.
Mohamadian,
“Forecasting
accident
frequency
of
an
urban
road
network
:
A
comparison
of
four artificial
neural network
techniques,
”
J. Forecas
t.
, vol. 37, no
. 7, pp. 767
–
780, 2018.
[35]
H.
E.
Colak,
T.
Memisoglu,
Y.
S.
Erbas,
and
S.
Bediroglu,
“Hot
spot
analysis
based
on
network
spatial
weights
to
determine spatial statistics of traffic ac
cidents in Rize, Turkey,”
Arab. J. Geosci.
, vol. 11, no. 151, pp. 1
–
11,
2018.
[36]
S.
S.
R.
Shariff,
H.
A.
Maad,
N.
N.
A.
Halim,
and
Z.
Derasit,
“Determ
ining
hotspots
of
road
accidents
using
spatial
analysis
,”
Indones. J. E
lectr. Eng.
Comput.
Sci.
, vol. 9, no. 1, pp. 146
–
151, 2018.
[37]
Z.
Fan,
C.
Liu,
D.
Cai,
and
S.
Yue,
“Research
on
bl
ack
spot
identifica
tion
of
safety
in
urban
traffic
accidents
based
on machine learning method,”
Saf. Sci.
, vol. 118, no. April, pp. 607
–
616, 2019.
[38]
M.
Ghadi
and
Á.
Török,
“A
comparative
analysis
of
black
spot
identification
methods
and
road
accident
se
gmentation methods,”
Accid. Anal. Prev.
, vol. 128, no. February, pp. 1
–
7, 2019.
[39]
M.
S.
Rahman,
M.
Abdel
-
Aty,
S.
Hasan,
and
Q.
Cai,
“Applying
machine
learning
approaches
to
analyze
the
vulnerable
road
-
users’
crashes
at
statewide
traffic
analysis
zones,”
J.
Safety
Res.
,
vol.
70,
no.
September,
pp.
275
–
288, 2019.
[40]
G.
Singh,
M.
Pal,
Y.
Yadav,
and
T.
Singla,
“Deep
neural
network
-
b
ased
predictive
modeling
of
road
accidents,
”
Neural Comp
ut. Appl.
, vol. 32, no. 16, pp. 12417
–
12426, 2020.
[41]
M.
Rosić,
D.
Pešić,
D.
Kukić,
B.
Antić,
and
M.
Božović,
“Method
f
or
selection
of
optimal
road
safety
composite
index with e
xamples fr
om DEA and
TOPSIS method,”
Accid. Anal. Prev.
, vol. 98, no. January, pp. 277
–
286, 2017.
[42]
A. Ait
-
Mlouk, F. Gharnati, a
nd
T. Agouti, “An improved approach for association rule mining
using a multi
-
criteria
decision
support
system:
a
case
study
in
road
safety,”
Eur.
Transp.
R
es.
Rev.
,
vol.
9:40,
no.
September,
pp.
1
–
13
,
2017.
[43]
A. Ait
-
Mlouk, T. Agouti, and F.
Gharnati, “Mi
ning and
prioritization
of association rules for big data:
multi
-
criteria
decision analysis approach,”
J. Big Da
ta
, vol. 4, no. 1, pp. 1
–
21, 2017.
[44]
H.
Pilko,
S.
Mandžuka,
and
D.
Barić,
“Urban
single
-
lane
roundabo
uts:
A
new
analytical
approach
using
mul
ti
-
criteria
and
simult
aneous
multi
-
objective
optimization
of
geometry
design,
efficiency
and
safety,”
Transp.
Res.
Part C Emerg. Technol.
, vol. 80, no. July, pp. 257
–
271, 2017.
[45]
V.
Rovšek,
M.
Batista,
and
B.
Bogunović,
“Identifying
the
key
risk
factors
of
traffic
accident
injury
severity
on
Slovenia
n roa
ds using
a non
-
parametric classifica
tion tree,”
Transport
, vol. 32, no. 3, pp. 272
–
281, 2017.
B
I
O
G
R
A
P
H
I
E
S
O
F
A
U
T
H
O
R
S
Anik
Vega
Vitianingsih
.
A
bachelor'
s
degree
in
Informatics
Engineer
ing
in
2004
and
a
Maste
r'
s
Degree
in
Game
Tech
was
obtained
in
2011.
The
author
is
a
P
ermanent
Lecturer
in
the
Informatics
Department,
editor
in
chief
of
the
International
Journal
of
Artificial
Intelligen
ce
and
Robotics
Universi
tas
Dr.
Soetomo,
and
students
in
the
Ph.D.
at
the
Faculty
of
Informa
tion
and
Communicat
ion Tec
hnology (
FTMK), Un
iversiti T
eknikal
Malaysia
Melaka
, Malays
ia. The
field
of interest in Spatial Analysis, and Spatial Data Modeling,
Artificial I
n
telligence in Geographical
Information
Systems.
Experiences
in
writing
papers
according
to
t
heir
fields
in
the
Scopus
Journal
include
2019
-
International
Journal
of
Intelligen
t
Engineering
and
Systems,
2019
-
Data
in
Brief,
2018
-
International
Journal
of
Engi
neering
and
Technology
(UAE),
2018
-
Journal
of
Telecommunication,
Electronic
and
Computer
Engin
eering.
The
auth
or
has
been
a
reviewer
for
Taylor
and
Francis
Ltd
publishers,
including
the
Journal
of
Transport
ation
Safety
and
Security,
IETE
Technical
Review
(
Institut
ion
of
Electronics
and
Telecommu
nication
Engineers,
India),
and
International
Journal
of
Injury
Control
and
Safety
Promotion,
as
well
as
a
reviewer
on
the
International
Social S
cience Journal P
ublisher Wi
ley
-
Blackwe
ll Publishing L
td.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
Spat
ia
l
anal
y
s
is
m
ode
l
fo
r
t
r
af
fi
c
a
c
c
id
e
nt
-
pr
one
r
oads
c
la
s
s
if
ic
at
io
n
…
(
A
ni
k
V
e
ga V
it
ia
ni
ngs
ih
)
373
Prof.
Dr.
Nanna
Suryana
Herman
.
Professor
at
th
e
Facu
lty
of
Information
and
Communicat
ion
Technol
ogy
(FTMK),
Universi
ti
Teknika
l
Malaysia
Melaka
,
Malaysia
.
Bachelo
r'
s
degree
in
Soil
and
Water
Enginee
ring
from
Padjadja
ran
Universi
ty,
Bandung,
Indonesia.
Mas
ter'
s
degree
in
Computer
Assisted
Regional
Planning
a
t
the
International
Institute
for
Geoinformatics
and
Earth
Observation
(ITC),
Enschede,
The
Net
herlands.
Doctoral
Degree
in
Department
of
Remote
sensing
and
GIS,
Research
University
of
Wageningen,
Hollan
d.
His
research
interests
are
spatial
data
analytics,
image
processing,
and
sp
atial
modeling,
and
remote
sensing.
Active
as
the
Editorial
Board
of
Intern
ational
Journals,
member
of
The
ASEAN
European
Academic
University
Network
(ASEA
-
UNINET),
and
EU
RAS
-
Eur
asian
Universities
Union.
Zahriah
Othman
.
Lecturer
at
the
Software
Engineering
department, Faculty
of
Information
and
Communicat
ion
Technol
ogy,
Universi
ti
Teknika
l
Malaysia
Mela
ka.
Bachelo
r'
s
degree
in
Information
Technology
from
Universiti
Utara
Mal
aysia
in
2001.
In
2003
he
received
a
master
of
science
degree
in
Software
Engineering
from
the
School
of
In
formatics,
Department
of
Computing,
and
Universi
ty
of
Bradfor
d,
United
Kingdom.
Ph.D.
i
n
Computer
Science
from
Universiti
Teknikal
Malaysia
Melaka,
M
alaysia.
His
areas
of
rese
arch
interest
are
Software
Engineering,
Artificial
Intelligence,
and
information
retrieval,
sp
ecifically
on
terminology
disagreement in retrieving geospatial data
.
Evaluation Warning : The document was created with Spire.PDF for Python.