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.
1
4
,
No
.
6
,
Dec
em
b
er
2
0
2
5
,
p
p
.
4
4
6
1
~
4
4
7
3
I
SS
N:
2
2
5
2
-
8
9
3
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijai.v
14
.i
6
.
p
p
4
4
6
1
-
4
4
7
3
4461
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
i
.
ia
esco
r
e.
co
m
Predic
ting th
e sev
erity of roa
d
traff
ic acciden
ts Mo
ro
cco: a
superv
ised ma
chi
ne learning
appro
a
ch
H
a
lim
a
Driss
i To
uza
ni
1
,
Sa
n
a
a
F
a
qu
ir
1,
2
,
Ali Y
a
hy
a
o
uy
1
1
LI
S
A
C
La
b
o
r
a
t
o
r
y
,
F
a
c
u
l
t
y
o
f
S
c
i
e
n
c
e
D
h
a
r
M
e
h
r
a
z
,
S
i
d
i
M
o
h
a
me
d
B
e
n
A
b
d
e
l
l
a
h
U
n
i
v
e
r
si
t
y
,
F
e
s
,
M
o
r
o
c
c
o
2
D
e
p
a
r
t
me
n
t
o
f
E
n
g
i
n
e
e
r
i
n
g
a
n
d
T
e
c
h
n
o
l
o
g
y
,
V
a
u
g
h
n
C
o
l
l
e
g
e
o
f
A
e
r
o
n
a
u
t
i
c
s a
n
d
T
e
c
h
n
o
l
o
g
y
,
N
e
w
Y
o
r
k
,
U
n
i
t
e
d
S
t
a
t
e
s
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Ap
r
3
,
2
0
2
5
R
ev
is
ed
Au
g
1
7
,
2
0
2
5
Acc
ep
ted
Sep
7
,
2
0
2
5
Early
p
re
d
icti
o
n
o
f
r
o
a
d
a
c
c
id
e
n
t
s
fa
talit
y
a
n
d
i
n
j
u
ries
se
v
e
rit
y
is
o
n
e
o
f
th
e
imp
o
rtan
t
s
u
b
jec
ts
t
o
ro
a
d
sa
fe
ty
e
m
p
h
a
siz
in
g
th
e
c
rit
ica
l
n
e
e
d
to
p
re
v
e
n
t
se
rio
u
s
c
o
n
se
q
u
e
n
c
e
s
t
o
re
d
u
c
e
i
n
ju
ries
a
n
d
fa
tali
ti
e
s.
T
h
is
stu
d
y
u
se
s
re
a
l
ro
a
d
a
c
c
id
e
n
ts
d
a
ta
se
t
in
M
o
ro
c
c
o
.
It
re
p
re
se
n
ts
t
h
e
i
n
ters
e
c
ti
o
n
b
e
twe
e
n
ro
a
d
sa
fe
ty
a
n
d
d
a
ta
sc
ien
c
e
,
a
imin
g
t
o
e
m
p
lo
y
m
a
c
h
in
e
lea
rn
i
n
g
t
e
c
h
n
iq
u
e
s
to
p
ro
v
id
e
v
a
lu
a
b
le
in
si
g
h
ts
in
a
c
c
id
e
n
t’s
se
v
e
rit
y
p
re
v
e
n
ti
o
n
.
Th
e
p
u
r
p
o
se
o
f
th
is
p
a
p
e
r
is
to
stu
d
y
r
o
a
d
a
c
c
id
e
n
ts
d
a
ta
i
n
t
h
e
c
o
u
n
try
a
n
d
c
o
m
b
in
e
re
su
lt
s
fro
m
sta
ti
stica
l
m
e
th
o
d
s
,
sp
a
ti
a
l
a
n
a
ly
sis,
a
n
d
m
a
c
h
in
e
lea
rn
in
g
m
o
d
e
ls
t
o
d
e
term
in
e
wh
ich
fa
c
t
o
rs
will
m
o
stl
y
c
o
n
tri
b
u
te
t
o
i
n
c
re
a
se
th
e
a
c
c
id
e
n
t’
se
v
e
rit
y
in
th
e
c
o
u
n
try
.
A
c
o
m
p
a
riso
n
o
f
re
su
lt
s
o
b
tai
n
e
d
wa
s
a
lso
c
o
n
d
u
c
ted
in
t
h
is
p
a
p
e
r
u
si
n
g
d
if
fe
re
n
t
m
e
tri
c
s
to
e
v
a
lu
a
te
th
e
e
ff
e
c
ti
v
e
n
e
ss
o
f
e
a
c
h
m
e
th
o
d
a
n
d
d
e
term
in
e
t
h
e
m
o
st
imp
o
rtan
t
fa
c
to
rs
t
h
a
t
c
o
n
tri
b
u
te
t
o
in
c
re
a
se
th
e
fa
talit
y
o
r
in
ju
ries
se
v
e
rit
y
in
t
h
e
sp
e
c
ifi
c
c
o
n
tex
t
o
f
a
c
c
id
e
n
ts.
Th
e
b
e
st
p
re
d
ictio
n
m
o
d
e
l
wa
s
th
e
n
i
n
jec
ted
i
n
to
a
p
r
o
p
o
se
d
a
lg
o
rit
h
m
wh
e
re
m
o
re
i
n
telli
g
e
n
t
tec
h
n
i
q
u
e
s
a
re
in
c
l
u
d
e
d
t
o
b
e
im
p
lem
e
n
ted
i
n
a
c
a
r
e
n
g
in
e
t
o
p
e
rf
o
rm
a
n
e
a
rly
d
e
t
e
c
ti
o
n
o
f
se
v
e
re
a
c
c
id
e
n
ts
a
n
d
th
e
re
fo
re
p
re
v
e
n
ti
n
g
c
ra
sh
e
s fr
o
m
h
a
p
p
e
n
i
n
g
.
K
ey
w
o
r
d
s
:
Acc
id
en
ts
s
ev
er
ity
p
r
ed
ictio
n
Data
an
aly
tics
Hu
m
an
f
ac
to
r
s
R
o
ad
ac
cid
en
ts
Su
p
er
v
is
ed
m
ac
h
i
n
e
lear
n
in
g
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
Halim
a
Dr
is
s
i T
o
u
za
n
i
L
I
SAC
L
ab
o
r
ato
r
y
,
Facu
lt
y
o
f
Scien
ce
Dh
ar
Me
h
r
az
,
Sid
i M
o
h
am
ed
B
en
Ab
d
ellah
Un
iv
e
r
s
ity
Ap
p
t 5
im
m
1
0
2
E
r
r
ac
h
i
d
ia
2
R
o
u
te
Ain
C
h
k
ef
,
Fes
,
Mo
r
o
c
co
E
m
ail:
h
alim
a.
d
r
is
s
ito
u
za
n
iwalali@
u
s
m
b
a.
ac
.
m
a
1.
I
NT
RO
D
UCT
I
O
N
A
r
o
ad
ac
cid
en
t
is
an
ev
e
n
t
th
at
in
v
o
lv
es
o
n
e
o
r
m
o
r
e
v
eh
icl
es
o
n
a
p
u
b
lic
r
o
ad
r
esu
ltin
g
i
n
m
ater
ial
d
am
ag
e
r
elate
d
o
n
ly
to
o
b
jec
ts
o
r
p
h
y
s
ical
d
am
ag
e
th
at
c
an
b
e
s
er
io
u
s
b
y
in
v
o
l
v
in
g
a
t
least
o
n
e
in
ju
r
ed
p
er
s
o
n
o
r
ev
e
n
f
atal
with
at
least
o
n
e
p
er
s
o
n
k
illed
.
E
v
er
y
y
ea
r
,
r
o
ad
s
af
ety
o
f
f
icial
s
s
et
u
p
awa
r
en
es
s
ca
m
p
aig
n
s
o
n
th
is
s
u
b
ject
an
d
p
u
b
lic
p
o
licies
tak
e
n
ew
m
ea
s
u
r
es
s
u
ch
as
lo
wer
in
g
th
e
s
p
ee
d
lim
it
to
8
0
o
n
r
o
ad
s
.
Desp
ite
th
is
wo
r
k
,
r
o
ad
f
atalities
r
em
ain
to
o
h
ig
h
all
ar
o
u
n
d
th
e
wo
r
ld
.
I
n
Fra
n
ce
,
t
h
e
2
0
2
2
r
e
p
o
r
t
o
n
r
o
ad
ac
cid
e
n
ts
s
tated
th
at
3
,
5
4
1
p
eo
p
le
wer
e
d
ea
d
o
n
th
e
m
et
r
o
p
o
litan
r
o
ad
s
o
f
Fra
n
ce
o
r
o
v
er
s
ea
s
.
I
t w
as a
ls
o
r
ep
o
r
ted
th
at
th
e
n
u
m
b
e
r
o
f
f
atal
ac
cid
en
ts
in
cr
ea
s
ed
b
y
1
.
3
%
co
m
p
ar
e
d
to
2
0
1
9
[
1
]
.
I
n
th
e
co
u
n
t
r
y
o
f
Slo
v
ak
ia,
r
o
ad
tr
a
f
f
ic
ac
ci
d
en
t
s
h
av
e
ar
is
en
as
a
cr
itical
p
u
b
lic
h
ea
lth
m
atter
,
d
em
a
n
d
in
g
a
m
u
ltid
is
cip
lin
ar
y
ap
p
r
o
ac
h
f
o
r
ef
f
ec
tiv
e
r
eso
lu
ti
o
n
.
Mo
r
e
th
an
4
0
,
0
0
0
d
ea
th
e
v
er
y
y
ea
r
o
n
th
e
r
o
ad
s
ev
en
wh
en
r
o
a
d
f
atalities
ar
e
d
ec
r
ea
s
in
g
.
So
m
e
p
r
ev
e
n
tio
n
s
wer
e
ap
p
lied
s
u
ch
as
in
-
v
eh
icle
s
af
ety
an
d
d
r
iv
er
ass
is
tan
ce
s
y
s
tem
s
to
in
ter
f
er
e
b
ef
o
r
e
th
e
ac
cid
en
t
h
ap
p
en
s
[
2
]
.
A
c
c
o
r
d
i
n
g
t
o
E
n
n
aji
h
et
a
l
.
[
3
]
,
t
h
e
r
e
h
as
b
e
e
n
a
n
o
tic
ea
b
l
e
i
n
c
r
e
ase
in
r
o
ad
ac
c
id
e
n
t
-
r
ela
te
d
f
ata
lit
ies
s
in
ce
1
9
6
8
.
T
h
e
t
r
a
f
f
ic
a
c
cid
e
n
ts
h
a
v
e
b
ec
o
m
e
t
h
e
l
ea
d
in
g
ca
u
s
e
o
f
d
ea
t
h
am
o
n
g
y
o
u
n
g
p
e
o
p
le
a
g
ed
1
7
t
o
2
9
y
e
ar
s
.
O
v
e
r
9
0
%
o
f
d
e
ath
s
an
d
i
n
j
u
r
ies
d
u
e
t
o
t
r
a
f
f
ic
a
cc
i
d
en
ts
o
cc
u
r
in
lo
w
-
a
n
d
m
i
d
d
le
-
i
n
c
o
m
e
c
o
u
n
t
r
ies
.
D
u
r
i
n
g
t
h
e
l
ast
t
h
r
e
e
y
ea
r
s
,
t
r
af
f
i
c
ac
ci
d
e
n
ts
h
av
e
b
e
en
in
cr
ea
s
i
n
g
f
r
o
m
Sep
te
m
b
er
t
o
De
ce
m
b
e
r
y
ea
r
ly
i
n
T
a
iwa
n
Pr
o
v
i
n
c
e
o
f
C
h
in
a
ac
c
o
r
d
i
n
g
t
o
b
ig
d
a
ta
a
n
aly
s
is
o
f
h
is
to
r
i
ca
l
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2252
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
1
4
,
No
.
6
,
Dec
em
b
er
2
0
2
5
:
4
4
6
1
-
4
4
7
3
4462
ac
c
id
en
t
d
ata
.
T
h
e
lat
est
s
ta
ti
s
tics
f
r
o
m
th
e
r
o
a
d
s
a
f
e
ty
in
f
o
r
m
a
ti
o
n
p
l
at
f
o
r
m
o
f
t
h
e
M
i
n
is
t
r
y
o
f
T
r
a
n
s
p
o
r
t
s
h
o
we
d
a
s
li
g
h
t
d
e
cr
ea
s
e
i
n
t
r
af
f
ic
ac
ci
d
e
n
t
d
ea
th
s
d
u
e
t
o
s
o
m
e
s
a
f
et
y
m
ea
s
u
r
es
t
ak
en
af
t
er
v
a
r
i
o
u
s
ac
ci
d
e
n
t
p
r
e
v
e
n
ti
o
n
wo
r
k
[
4
]
.
P
o
r
tu
g
a
l
a
ls
o
is
f
ac
in
g
a
s
e
r
i
o
u
s
p
r
o
b
le
m
wit
h
f
at
al
ac
ci
d
e
n
ts
.
I
n
2
0
2
0
,
t
h
e
c
o
u
n
t
r
y
r
e
v
e
ale
d
a
to
tal
n
u
m
b
e
r
o
f
2
7
,
7
2
5
a
cc
i
d
e
n
ts
wi
th
v
ict
im
s
f
r
o
m
w
h
i
ch
5
3
6
v
icti
m
s
d
ea
d
.
T
h
ese
n
u
m
b
er
s
p
u
t
t
h
e
co
u
n
t
r
y
a
t t
h
e
n
i
n
t
h
-
h
ig
h
est
p
o
s
iti
o
n
wi
th
t
h
e
m
o
s
t
f
at
ali
ties
p
er
m
illi
o
n
in
h
ab
ita
n
ts
i
n
th
e
E
u
r
o
p
ea
n
Un
io
n
[
5
]
.
I
n
th
e
co
u
n
tr
y
o
f
Mo
r
o
cc
o
a
s
well,
ca
r
ac
cid
en
ts
b
ec
am
e
s
o
u
n
s
ig
n
if
ican
t
s
in
ce
th
er
e
ar
e
p
eo
p
le
p
ass
in
g
awa
y
ev
er
y
d
a
y
.
B
ased
o
n
s
tatis
tics
r
elea
s
ed
b
y
th
e
Min
is
tr
y
o
f
E
q
u
ip
m
en
t
an
d
T
r
an
s
p
o
r
t
i
n
2
0
1
2
,
th
er
e
wer
e
4
4
,
9
0
2
ca
r
ac
cid
e
n
ts
,
r
esu
ltin
g
in
t
h
e
lo
s
s
o
f
1
0
liv
es
a
n
d
leav
in
g
1
4
0
in
ju
r
ed
p
e
r
d
ay
.
T
h
ese
in
cid
en
ts
n
o
t
o
n
ly
d
ev
astate
e
n
tire
f
am
ilies
b
u
t
also
lea
d
to
th
e
cr
ea
tio
n
o
f
o
r
p
h
an
s
an
d
wid
o
ws
[
6
]
.
I
n
lin
e
with
th
e
d
ata
o
f
th
e
Natio
n
a
l
R
o
ad
Saf
ety
Ag
en
cy
(
NARS
A)
,
th
e
n
ewe
s
t
s
tatis
tics
s
h
o
w
th
at
Mo
r
o
cc
o
r
ec
o
r
d
e
d
3
,
4
9
9
r
o
ad
d
ea
th
s
i
n
2
0
2
2
,
m
a
r
k
in
g
a
5
%
r
e
g
r
es
s
io
n
co
m
p
ar
e
d
to
2
0
2
1
.
Nea
r
ly
th
r
ee
-
q
u
ar
ter
s
o
f
th
ese
f
atalities
in
v
o
lv
e
v
u
ln
er
ab
le
r
o
a
d
u
s
er
s
[
7
]
.
Fig
u
r
e
1
s
h
o
ws
th
at
Mo
r
o
cc
o
h
as
th
e
h
ig
h
est
f
atality
r
is
k
5
.
6
r
o
ad
d
ea
t
h
s
o
f
p
e
r
1
0
0
,
0
0
0
r
eg
is
ter
ed
v
eh
icles in
2
0
2
2
co
m
p
ar
ed
to
o
th
er
co
u
n
tr
ies.
Fig
u
r
e
1
.
R
o
ad
f
atalities
p
er
1
0
,
0
0
0
r
eg
is
ter
ed
v
eh
icles in
M
o
r
o
cc
o
c
o
m
p
a
r
ed
to
o
th
er
c
o
u
n
tr
ies in
2
0
2
2
Af
ter
all
th
ese
alar
m
in
g
s
tatis
tics
,
it
is
v
er
y
im
p
o
r
tan
t
to
ap
p
ly
d
i
f
f
er
en
t
tech
n
iq
u
es
f
o
r
ef
f
ec
tiv
e
p
r
ev
en
tio
n
th
at
ca
n
h
elp
i
n
r
ed
u
cin
g
h
u
m
an
lo
s
s
es
an
d
i
r
r
ep
ar
ab
le
p
h
y
s
ical
an
d
p
s
y
c
h
o
lo
g
ical
d
a
m
ag
e.
Ho
wev
er
,
b
u
ild
in
g
p
r
e
d
ictiv
e
m
o
d
els
r
e
q
u
ir
es
d
etec
tin
g
f
ac
to
r
s
r
esp
o
n
s
ib
le
f
o
r
th
e
o
c
cu
r
r
en
ce
o
f
th
ese
ac
cid
en
ts
an
d
th
eir
d
a
n
g
er
o
u
s
n
ess
.
Dif
f
er
en
t
p
r
ed
ictiv
e
m
o
d
els
h
av
e
b
ee
n
p
r
esen
ted
in
th
e
liter
atu
r
e.
R
esu
lts
d
ep
en
d
lar
g
el
y
o
n
th
e
v
ar
ia
b
les
an
d
ty
p
e
o
f
th
e
d
ata
s
am
p
le
an
al
y
ze
d
.
Dif
f
er
en
t
an
aly
tic
m
eth
o
d
s
an
d
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
s
u
s
ed
to
f
in
d
v
ar
io
u
s
ac
cid
en
t r
i
s
k
f
ac
to
r
s
an
d
d
e
v
elo
p
p
r
ed
ict
io
n
s
av
ailab
le
wer
e
lis
ted
[
8
]
.
I
n
g
e
n
er
al,
m
o
s
t
s
tu
d
ies
o
n
tr
af
f
ic
ac
cid
en
t
f
o
r
ec
a
s
tin
g
f
o
cu
s
ed
o
n
two
im
p
o
r
tan
t
r
esear
ch
m
eth
o
d
s
:
s
tatis
t
ical
m
eth
o
d
s
an
d
n
eu
r
al
n
etwo
r
k
s
.
Statis
tical
tech
n
iq
u
es
s
u
ch
as
k
-
n
ea
r
est
n
eig
h
b
o
r
s
(
KNN)
,
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
e
(
SVM)
,
a
n
d
l
o
g
is
tic
r
eg
r
ess
io
n
(
L
R
)
wer
e
em
p
lo
y
ed
to
f
o
r
ec
ast
th
e
f
r
eq
u
en
cy
o
f
th
e
tr
af
f
ic
ac
cid
en
ts
wh
ile
ar
tific
ial
n
e
u
r
al
n
etwo
r
k
s
(
ANN)
wer
e
a
p
p
li
ed
to
i
m
p
lem
en
t
th
e
f
le
x
ib
ilit
y
,
g
en
e
r
aliza
tio
n
,
o
r
m
o
r
e
r
o
b
u
s
t
p
r
e
d
ictio
n
s
y
s
tem
[
9
]
.
Dete
ctin
g
ac
cid
en
ts
b
l
ac
k
s
p
o
ts
was
also
an
im
p
o
r
t
an
t
is
s
u
e
to
im
p
r
o
v
e
r
o
ad
tr
a
f
f
ic
s
af
ety
a
n
d
r
ed
u
c
e
th
e
tr
af
f
ic
ac
cid
en
ts
s
ev
er
it
y
.
Var
io
u
s
s
tu
d
ies
ass
ess
ed
th
e
ef
f
icac
y
o
f
th
e
em
p
ir
ical
B
ay
esian
tech
n
iq
u
e
o
n
ac
cid
en
ts
b
lack
s
p
o
t
class
i
f
icatio
n
as
it
wa
s
p
r
o
p
o
s
ed
in
[
1
0
]
to
co
n
s
tr
u
ct
a
b
lack
s
p
o
t
id
en
tific
atio
n
m
o
d
e
l
th
at
co
u
ld
id
en
tify
a
b
lack
s
p
o
t.
T
h
e
m
o
d
el
g
av
e
th
e
b
est
ac
cu
r
ac
y
co
m
p
ar
e
d
to
th
e
I
D
3
d
ec
is
io
n
tr
ee
,
LR
,
an
d
SVM
[
1
1
]
.
C
lass
if
icatio
n
alg
o
r
ith
m
s
u
ch
as
KNN,
n
aï
v
e
B
ay
es
(
NB
),
an
d
ANN
wer
e
u
tili
ze
d
in
[
1
2
]
t
o
d
ev
elo
p
a
p
r
ed
ictiv
e
m
o
d
el
f
o
r
p
r
ed
ictin
g
o
cc
u
r
r
en
ce
s
o
f
tr
af
f
ic
ac
cid
e
n
ts
.
Ho
wev
er
,
f
r
o
m
all
o
f
th
is
r
ese
ar
ch
,
o
n
ly
a
f
ew
u
s
ed
m
ac
h
i
n
e
lear
n
in
g
m
eth
o
d
s
to
p
r
ed
ict
ca
r
cr
ash
es
’
f
atality
o
r
th
e
s
er
io
u
s
n
ess
o
f
in
ju
r
ies
in
T
u
r
k
e
y
,
in
ju
r
ies
an
d
f
atalities
d
ata
an
aly
s
is
wer
e
an
aly
ze
d
u
s
in
g
n
o
n
-
lin
ea
r
r
eg
r
ess
io
n
an
d
ANN
[
1
3
]
.
T
o
d
eter
m
in
e
t
h
e
m
o
s
t
im
p
o
r
tan
t
f
ac
to
r
s
lead
in
g
to
ca
r
ac
cid
en
ts
,
d
ata
m
in
in
g
tech
n
iq
u
es
s
u
ch
as
d
ec
is
io
n
tr
ee
s
,
n
o
n
-
lin
ea
r
r
e
g
r
ess
io
n
,
an
d
class
if
icatio
n
m
eth
o
d
s
wer
e
u
s
ed
i
n
[
1
4
]
an
d
r
esu
lts
r
ev
ea
led
th
e
v
eh
icle
ty
p
e
f
ac
to
r
as
o
n
e
o
f
th
e
m
o
s
t
d
a
n
g
er
o
u
s
f
ac
to
r
s
r
elate
d
to
ac
cid
e
n
t
s
ev
er
ity
.
A
tr
af
f
ic
cr
ash
r
is
k
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
P
r
ed
ictin
g
th
e
s
ev
erit
y
o
f ro
a
d
tr
a
ffic a
cc
id
en
ts
Mo
r
o
cc
o
:
a
s
u
p
ervis
ed
ma
ch
in
e
…
(
Ha
lim
a
Dri
s
s
i To
u
z
a
n
i)
4463
p
r
ed
ictio
n
m
o
d
el
u
s
in
g
t
h
e
lim
ited
r
an
d
o
m
-
s
y
n
th
etic
m
in
o
r
ity
o
v
er
-
s
am
p
lin
g
tech
n
iq
u
e
(
LR
-
SMOT
E
)
alg
o
r
ith
m
was
u
s
ed
in
[
1
5
]
to
ca
teg
o
r
ize
th
e
s
h
ar
p
ac
c
eler
atio
n
an
d
d
ec
ele
r
atio
n
,
v
o
lu
m
e
,
av
er
ag
e
s
p
ee
d
,
th
e
q
u
o
tien
t
b
etwe
en
f
r
ee
f
lo
w
s
p
ee
d
an
d
cu
r
r
e
n
t
av
er
ag
e
r
o
a
d
s
p
ee
d
an
d
th
e
co
ef
f
icien
t
o
f
v
ar
iatio
n
o
f
s
p
ee
d
as
b
ein
g
th
e
m
ain
attr
ib
u
tes
th
at
af
f
ec
t
th
e
r
is
k
o
f
a
cr
ash
in
an
ac
cid
en
t.
T
h
e
m
o
d
el
u
s
ed
r
ea
l
-
tim
e
tr
af
f
ic
f
lo
w
d
ata
an
d
r
is
k
y
d
r
iv
in
g
b
e
h
av
io
r
d
ata
to
s
tu
d
y
th
e
tr
af
f
ic
cr
ash
r
is
k
o
n
f
r
ee
way
s
.
I
n
th
e
UK
a
s
well,
an
in
-
d
e
p
th
an
aly
s
is
o
f
th
e
cu
r
r
en
t M
L
m
o
d
els wa
s
im
p
lem
en
ted
f
o
r
p
r
e
d
ictin
g
in
ju
r
y
s
ev
er
ity
in
r
o
ad
ac
cid
en
ts
[
1
6
]
.
T
h
e
Had
d
o
n
m
at
r
ix
was
u
s
ed
in
th
is
co
n
tex
t
to
an
aly
ze
th
e
ca
u
s
es
an
d
s
er
io
u
s
n
ess
o
f
th
e
ac
cid
en
ts
in
th
e
co
u
n
tr
y
.
C
h
ak
r
a
b
o
r
ty
et
a
l.
[
1
7
]
c
o
n
d
u
cted
in
I
n
d
i
a
s
o
u
g
h
t
to
d
e
v
elo
p
a
m
o
d
el
f
o
r
p
r
e
d
ictin
g
f
atal
p
ed
estrian
cr
ash
es
a
n
d
to
d
ete
r
m
in
e
th
e
f
ac
to
r
s
th
at
e
x
ac
er
b
ate
th
e
s
ev
er
ity
o
f
tr
an
s
p
o
r
tatio
n
-
r
elate
d
in
ju
r
ies
an
d
d
ea
th
s
.
T
h
e
r
esear
ch
r
e
v
ea
led
th
at
th
e
'
ap
p
r
o
ac
h
in
g
s
p
ee
d
'
o
f
m
o
to
r
ized
v
e
h
icles
was
th
e
m
o
s
t
p
r
o
n
o
u
n
ce
d
in
f
lu
e
n
ce
o
n
f
atal
p
ed
estrian
cr
ash
es
.
T
h
is
p
ap
er
in
clu
d
es
two
d
ata
m
in
in
g
tech
n
iq
u
es:
d
ata
an
aly
tics
an
d
an
d
s
u
p
e
r
v
is
ed
m
ac
h
i
n
e
lear
n
in
g
alg
o
r
ith
m
s
to
p
r
e
d
ict
th
e
s
ev
e
r
ity
o
f
ac
cid
e
n
ts
o
n
Mo
r
o
cc
an
r
o
ad
s
.
An
ac
cid
e
n
t
is
co
n
s
id
er
ed
a
f
atal
ac
cid
en
t
if
at
least
o
n
e
p
er
s
o
n
was
k
il
led
.
W
h
en
n
o
p
er
s
o
n
is
k
illed
in
t
h
e
ac
cid
e
n
t
b
u
t
th
er
e
ar
e
in
ju
r
ies
th
en
th
e
ac
cid
en
t
is
n
o
t
f
atal
.
I
n
th
is
ca
s
e
th
e
in
ju
r
ies
s
er
io
u
s
n
ess
is
m
ea
s
u
r
ed
.
T
h
e
aim
o
f
th
is
r
esear
ch
is
to
d
eter
m
in
e
a
p
r
ed
ictiv
e
m
o
d
el
f
o
r
f
o
r
ec
as
tin
g
in
ad
v
a
n
ce
th
e
f
atality
o
f
in
ju
r
y
s
ev
e
r
ity
o
f
a
n
ac
cid
en
t
to
h
ap
p
en
b
ased
o
n
ac
cid
en
ts
h
eter
o
g
e
n
eo
u
s
c
h
ar
ac
ter
is
tics
.
T
h
e
s
tu
d
y
u
s
ed
r
e
al
r
o
ad
t
r
af
f
ic
ac
cid
e
n
ts
p
r
o
v
i
d
ed
b
y
th
e
NARS
A
f
o
r
th
e
two
y
ea
r
s
2
0
1
5
an
d
2
0
1
6
.
T
h
e
p
r
ed
ictio
n
is
o
b
tain
ed
f
o
r
ea
ch
ca
s
e
g
iv
e
n
d
if
f
e
r
en
t
ac
cid
en
t
attr
ib
u
tes
an
d
co
m
b
in
es
r
esu
lts
f
r
o
m
s
tatis
tical
m
eth
o
d
s
,
s
p
atial
an
al
y
s
is
,
an
d
ar
tific
ial
in
tell
ig
en
ce
m
o
d
els.
Statis
tics
m
eth
o
d
s
f
u
n
ctio
n
well
in
th
e
p
r
o
ce
s
s
in
g
an
d
d
ata
an
aly
s
i
s
,
h
o
wev
er
,
d
i
f
f
er
en
t
s
tu
d
ies
d
em
o
n
s
tr
ated
th
at
ANNs
g
iv
e
m
o
r
e
ac
cu
r
ate
p
r
ed
ictio
n
ca
p
ab
ilit
ies
[
1
8
]
,
[
1
9
]
.
I
n
th
is
r
esear
ch
,
f
iv
e
d
if
f
er
en
t
s
u
p
r
e
v
is
ed
m
ac
h
in
e
lear
n
in
g
m
eth
o
d
s
w
er
e
u
s
ed
:
NB
,
SVM,
ANN,
KNN,
L
R
.
T
h
e
p
er
f
o
r
m
an
ce
o
f
th
ese
m
o
d
els
was
co
m
p
ar
ed
u
s
in
g
d
if
f
er
e
n
t
m
etr
ics
o
n
th
e
p
r
o
v
id
e
d
d
atasets
.
T
h
e
attr
ib
u
tes
in
th
e
d
ataset
wer
e
g
r
o
u
p
e
d
b
y
th
r
ee
f
ac
to
r
s
as
d
escr
ib
ed
i
n
th
e
Had
d
o
n
’
s
m
atr
ix
(
h
u
m
an
f
ac
to
r
,
v
e
h
icle
,
an
d
eq
u
ip
m
en
t
f
ac
t
o
r
a
n
d
en
v
ir
o
n
m
en
t
f
ac
to
r
)
.
T
h
e
r
a
n
d
o
m
f
o
r
est
im
p
o
r
tan
ce
tec
h
n
iq
u
e
was
also
u
s
ed
as
an
em
b
ed
d
ed
f
ea
tu
r
e
s
elec
tio
n
m
eth
o
d
in
m
ac
h
in
e
lear
n
in
g
to
id
e
n
tify
th
e
attr
ib
u
tes
th
at
co
n
tr
i
b
u
te
m
o
r
e
t
o
ac
c
id
en
ts
s
ev
er
ity
.
A
t
th
e
en
d
,
a
p
r
o
p
o
s
ed
alg
o
r
ith
m
was
d
esig
n
ed
t
o
in
clu
d
e
a
co
llectio
n
o
f
d
ata
f
r
o
m
d
if
f
e
r
en
t
d
ev
ices,
th
e
b
est
p
r
ed
ictio
n
m
o
d
el
alo
n
g
with
o
th
er
in
tellig
en
t te
ch
n
iq
u
es su
c
h
as a
d
v
an
ce
d
d
r
iv
er
-
ass
is
tan
ce
s
y
s
tem
s
(
ADA
S).
T
h
is
alg
o
r
ith
m
ca
n
b
e
im
p
lem
en
ted
in
s
em
i
-
au
to
n
o
m
o
u
s
v
e
h
icles
to
en
ab
le
ea
r
ly
d
etec
tio
n
an
d
p
r
ev
en
tio
n
o
f
s
ev
er
e
ac
cid
en
ts
.
2.
M
E
T
H
O
D
R
o
ad
tr
af
f
ic
ac
cid
e
n
ts
p
o
s
e
a
s
ig
n
if
ican
t
ch
allen
g
e
t
o
r
o
a
d
s
af
ety
,
em
p
h
asizin
g
t
h
e
cr
itic
al
n
ee
d
to
p
r
ev
en
t
th
eir
s
er
io
u
s
co
n
s
eq
u
en
ce
s
to
r
ed
u
ce
in
ju
r
ies
an
d
f
atalities.
I
n
th
is
co
n
tex
t,
lev
er
ag
in
g
m
ac
h
in
e
lear
n
in
g
m
o
d
els
s
h
o
ws
p
r
o
m
i
s
e
f
o
r
p
r
ed
ictin
g
ac
ci
d
en
t
s
ev
er
ity
an
d
f
ac
ilit
atin
g
s
wif
t
in
t
er
v
en
tio
n
.
So
lu
tio
n
s
an
d
d
ec
is
io
n
s
ca
n
b
e
f
o
u
n
d
with
o
u
t
d
ata.
Ho
wev
er
,
th
ese
d
ec
is
io
n
s
ar
e
o
n
ly
b
ased
o
n
p
er
s
o
n
al
h
eu
r
is
tics
d
ev
elo
p
e
d
b
y
liv
ed
ex
p
e
r
ien
c
e.
T
o
f
in
d
m
o
r
e
co
n
cr
ete
d
ec
i
s
io
n
s
,
esp
ec
ially
in
cr
itical
d
o
m
ain
s
s
u
ch
as
r
o
ad
tr
af
f
ic
ac
cid
e
n
ts
r
eq
u
ir
es
a
l
o
t
o
f
d
ata
an
d
th
in
k
in
g
.
Du
e
to
t
h
e
d
e
v
elo
p
m
en
t
o
f
i
n
ter
n
et
o
f
th
in
g
s
(
I
o
T
)
d
ev
ices
co
n
n
e
cted
to
th
e
n
etwo
r
k
,
d
ata
co
llectio
n
an
d
tr
a
n
s
m
is
s
io
n
b
ec
am
e
p
r
ac
tical
f
o
r
f
u
r
th
er
an
aly
s
is
.
Hav
in
g
g
o
o
d
d
ata
is
ess
en
tial
f
o
r
m
a
k
in
g
th
e
r
ig
h
t
d
ec
is
io
n
s
.
T
h
ir
ty
y
ea
r
s
ag
o
in
th
e
USA,
W
illi
am
Had
d
o
n
J
r
.
d
escr
ib
ed
r
o
ad
tr
an
s
p
o
r
t
as
a
p
o
o
r
ly
d
esig
n
ed
"m
an
-
m
ac
h
in
e"
s
y
s
tem
r
eq
u
i
r
in
g
co
m
p
r
eh
en
s
iv
e
s
y
s
tem
ic
tr
ea
tm
en
t
[
2
0
]
.
T
h
e
m
atr
ix
Had
d
o
n
'
s
n
in
e
-
ce
ll
m
o
d
el
r
e
p
r
esen
ts
th
e
d
y
n
am
i
c
s
y
s
tem
,
with
ea
ch
ce
ll
o
f
f
e
r
in
g
in
te
r
v
en
tio
n
o
p
tio
n
s
to
r
ed
u
ce
tr
a
f
f
ic
in
j
u
r
y
.
T
h
e
m
atr
ix
h
as
led
to
s
ig
n
if
i
ca
n
t
ad
v
an
ce
s
in
u
n
d
e
r
s
tan
d
i
n
g
th
e
b
e
h
av
io
r
al,
r
o
a
d
an
d
v
eh
icle
f
ac
to
r
s
th
at
in
f
lu
en
ce
th
e
n
u
m
b
e
r
an
d
s
ev
er
ity
o
f
in
ju
r
ies
in
tr
af
f
ic
ac
ci
d
en
ts
.
T
h
e
aim
was
to
ac
h
iev
e
s
p
ec
if
ic
r
ed
u
ctio
n
s
in
th
e
n
u
m
b
er
o
f
r
o
ad
f
ataliti
es
an
d
in
ju
r
ies
all
ar
o
u
n
d
th
e
wo
r
ld
.
Ho
wev
er
,
t
h
e
p
r
ac
tical
ap
p
licatio
n
o
f
th
is
s
y
s
tem
ic
ap
p
r
o
ac
h
r
em
ain
s
th
e
m
ain
c
h
allen
g
e
f
o
r
p
o
licy
m
ak
er
s
an
d
r
o
a
d
s
af
ety
p
r
o
f
ess
io
n
als.
Ou
r
r
esear
ch
f
o
cu
s
es
o
n
f
in
d
i
n
g
s
o
lu
tio
n
s
u
s
in
g
m
ac
h
in
e
lear
n
in
g
tec
h
n
iq
u
es
to
p
r
ev
en
t
ac
cid
en
ts
b
ef
o
r
e
h
ap
p
e
n
in
g
co
n
s
id
er
in
g
th
e
f
a
cto
r
s
lis
ted
in
Had
d
o
n
’
s
m
atr
ix
b
y
d
e
f
in
i
n
g
an
ac
cid
en
t
as
a
s
er
ies
o
f
f
ea
tu
r
es
wh
er
e
d
ata
an
aly
s
is
an
d
p
r
ed
ictio
n
m
eth
o
d
s
ca
n
b
e
ap
p
lied
to
h
elp
r
e
d
u
cin
g
th
e
co
s
t o
f
f
atalities an
d
d
r
iv
er
in
ju
r
ies.
2
.
1
.
Da
t
a
clea
nin
g
a
nd
prepro
ce
s
s
ing
T
h
e
d
ataset
u
s
ed
co
n
tain
ed
two
f
iles
.
On
e
f
ile
co
n
tain
s
in
f
o
r
m
atio
n
o
n
th
e
ac
cid
en
t
’
s
d
r
iv
er
s
,
p
ass
en
g
er
s
in
v
o
lv
ed
as
well
as
th
e
ac
cid
en
t
o
u
tco
m
e
in
d
ica
to
r
s
(
C
OD_
T
UE
f
o
r
f
atalities
an
d
C
OD_
B
L
E
f
o
r
in
ju
r
ies)
wh
ile
th
e
o
t
h
er
f
ile
i
n
clu
d
es
tem
p
o
r
al,
s
p
atial,
v
eh
icle,
an
d
r
o
ad
u
s
er
in
f
o
r
m
atio
n
.
T
h
ese
two
f
iles
wer
e
m
er
g
ed
in
t
o
o
n
e
c
o
m
p
l
ete
d
ataset
to
b
e
ab
le
to
co
n
d
u
ct
th
e
r
esear
c
h
.
9
4
,
8
6
2
a
cc
id
en
t
ca
s
es
wer
e
an
aly
s
ed
af
ter
clea
n
i
n
g
th
e
d
atab
ase
f
r
o
m
in
c
o
m
p
lete
in
f
o
r
m
atio
n
.
T
h
e
d
ataset
h
o
ld
s
es
s
en
tial
in
f
o
r
m
atio
n
,
o
f
f
er
in
g
a
c
o
m
p
r
e
h
en
s
iv
e
u
n
d
er
s
tan
d
in
g
o
f
th
e
cir
c
u
m
s
tan
ce
s
s
u
r
r
o
u
n
d
in
g
ea
ch
ac
cid
en
t:
n
u
m
e
r
ical,
ca
teg
o
r
ical,
an
d
tem
p
o
r
al
d
at
a.
T
h
ese
attr
ib
u
tes
wer
e
clas
s
if
ied
b
ased
o
n
th
e
th
r
e
e
f
ac
t
o
r
s
as
in
d
icate
d
in
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2252
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
1
4
,
No
.
6
,
Dec
em
b
er
2
0
2
5
:
4
4
6
1
-
4
4
7
3
4464
Had
d
o
n
’
s
m
atr
i
x
:
h
u
m
a
n
f
ac
t
o
r
s
as
s
h
o
wn
in
T
ab
le
1
s
u
c
h
as
d
r
iv
er
’
s
p
e
r
s
o
n
al
in
f
o
r
m
atio
n
,
ex
ce
s
s
iv
e
o
r
in
ap
p
r
o
p
r
iate
s
p
ee
d
in
g
,
f
ailu
r
e
to
r
esp
ec
t
s
af
ety
d
is
tan
ce
s
,
v
eh
icles
an
d
eq
u
ip
m
en
t
f
ac
to
r
s
s
u
ch
as
tech
n
ical
p
r
o
b
lem
s
r
elate
d
to
t
h
e
co
n
d
itio
n
o
f
th
e
v
e
h
icle
an
d
en
v
ir
o
n
m
en
tal
f
ac
to
r
s
lik
e
wea
th
er
co
n
d
itio
n
s
,
r
o
ad
co
n
d
itio
n
s
,
an
d
o
th
e
r
s
as st
ate
d
in
d
etails
[
2
1
]
.
T
ab
le
1
.
Ap
p
licatio
n
o
f
Had
d
o
n
'
s
m
atr
ix
o
n
th
e
ac
cid
en
ts
d
at
a
s
et
F
a
c
t
o
r
s
P
h
a
se
H
u
ma
n
V
e
h
i
c
l
e
s a
n
d
e
q
u
i
p
m
e
n
t
En
v
i
r
o
n
m
e
n
t
B
e
f
o
r
e
a
c
c
i
d
e
n
t
A
c
c
i
d
e
n
t
p
r
e
v
e
n
t
i
o
n
D
r
i
v
e
r
’
s
a
g
e
D
r
i
v
e
r
’
s s
e
x
D
r
i
v
e
r
’
s
p
r
o
f
e
ssi
o
n
D
r
i
v
e
r
’
s
t
y
p
e
o
f
d
r
i
v
i
n
g
l
i
c
e
n
se
V
e
h
i
c
l
e
t
y
p
e
U
sag
e
o
f
t
h
e
v
e
h
i
c
l
e
C
o
d
e
p
r
o
v
i
n
c
e
P
a
v
e
me
n
t
c
o
n
d
i
t
i
o
n
S
u
r
f
a
c
e
c
o
n
d
i
t
i
o
n
Li
g
h
t
W
e
a
t
h
e
r
c
o
n
d
i
t
i
o
n
s
A
c
c
i
d
e
n
t
Tr
a
u
m
a
p
r
e
v
e
n
t
i
o
n
i
n
c
a
s
e
o
f
a
c
c
i
d
e
n
t
Er
r
o
r
s
d
u
e
t
o
t
h
e
d
r
i
v
e
r
,
p
h
y
s
i
c
a
l
c
o
n
d
i
t
i
o
n
o
f
t
h
e
d
r
i
v
e
r
O
b
st
a
c
l
e
s
h
u
r
t
R
o
a
d
d
e
t
o
u
r
A
f
t
e
r
a
c
c
i
d
e
n
t
M
a
i
n
t
a
i
n
a
l
i
v
e
K
i
l
l
e
d
/
i
n
j
u
r
e
d
o
r
n
o
p
r
o
b
l
e
m
Up
o
n
in
itial
ex
am
in
atio
n
,
t
h
e
d
atasets
r
ev
ea
led
m
u
ltip
le
ch
allen
g
es.
Sev
er
al
co
lu
m
n
s
h
ad
a
s
ig
n
if
ican
t
p
er
ce
n
tag
e
o
f
m
is
s
in
g
d
ata.
Oth
er
co
lu
m
n
s
co
n
tain
ed
v
er
y
f
ew
n
o
n
-
n
u
ll
v
al
u
es.
T
h
ese
co
lu
m
n
s
ap
p
ea
r
ed
la
r
g
ely
u
n
in
f
o
r
m
ati
v
e.
Ad
d
itio
n
al
c
o
lu
m
n
s
wer
e
r
ed
u
n
d
an
t
wh
ile
f
u
r
th
er
o
n
es
ap
p
ea
r
ed
t
o
b
e
n
o
t
r
elev
an
t
to
th
e
an
aly
s
is
g
o
als,
s
u
ch
as
s
p
ec
if
ic
co
d
es
o
r
id
en
t
if
ier
s
with
n
o
r
elate
d
co
d
if
icat
io
n
o
r
ex
p
lan
atio
n
.
T
h
e
ac
cid
e
n
t
d
ate
f
o
r
m
at
was
also
co
n
v
e
r
ted
to
d
atetim
e
f
o
r
m
at
to
f
ac
ilit
ate
tem
p
o
r
al
m
an
ip
u
latio
n
.
T
h
is
will
allo
w
th
e
m
o
d
el
to
ex
p
lo
it
th
e
tem
p
o
r
al
asp
ec
ts
o
f
ac
ci
d
en
ts
,
s
u
ch
as
th
e
v
ar
iatio
n
in
s
ev
er
ity
lev
els
d
ep
en
d
i
n
g
o
n
th
e
d
ay
s
o
f
t
h
e
wee
k
,
m
o
n
t
h
s
,
o
r
s
ea
s
o
n
s
.
All
th
e
ch
allen
g
es
m
en
tio
n
ed
ab
o
v
e
r
e
q
u
ir
ed
a
th
o
r
o
u
g
h
cle
an
in
g
an
d
p
r
e
p
r
o
ce
s
s
in
g
p
h
ase
to
en
s
u
r
e
th
e
r
eliab
ilit
y
an
d
ac
cu
r
ac
y
o
f
th
e
s
u
b
s
eq
u
en
t
an
aly
s
is
.
T
h
is
p
h
ase
was
p
er
f
o
r
m
ed
b
y
a
p
p
ly
in
g
th
e
f
ilter
s
,
im
p
u
tatio
n
m
eth
o
d
s
an
d
s
elec
tio
n
p
h
ases
av
ailab
le
b
y
t
h
e
R
ap
id
Min
er
w
h
ich
is
a
p
o
wer
f
u
l
d
ata
s
cien
ce
p
latf
o
r
m
t
h
at
o
f
f
er
s
an
in
te
g
r
ated
e
n
v
ir
o
n
m
en
t
f
o
r
d
ata
p
r
ep
a
r
atio
n
,
m
ac
h
in
e
lear
n
i
n
g
,
d
ee
p
lear
n
in
g
,
tex
t
m
in
in
g
,
an
d
p
r
e
d
ictiv
e
an
aly
tics
.
T
o
b
e
ab
le
to
d
o
th
e
a
n
aly
s
is
p
h
ase,
th
e
two
d
atab
ases
m
en
tio
n
ed
ab
o
v
e
wer
e
m
er
g
e
d
t
o
o
n
e
c
o
m
p
lete
d
ataset
u
s
in
g
p
h
y
to
n
.
Py
th
o
n
was
ch
o
s
en
f
o
r
its
f
lex
i
b
ilit
y
an
d
th
e
p
o
wer
f
u
l
d
ata
h
an
d
lin
g
ca
p
ab
ilit
ies
o
f
f
er
ed
b
y
its
lib
r
ar
ies.
T
h
e
N
u
m
Py
an
d
Pan
d
as
lib
r
a
r
ies
wer
e
im
p
o
r
ted
to
u
s
e
class
if
icatio
n
f
u
n
ctio
n
s
.
T
h
e
f
ir
s
t
f
u
n
ctio
n
'
class
if
y
_
in
ju
r
y
_
s
ev
er
ity
'
,
ca
teg
o
r
izes
in
ju
r
ie
s
ac
co
r
d
in
g
t
o
th
e
s
ev
er
ity
co
d
e
(
'
C
OD_
B
L
E
'
)
.
T
h
e
s
ec
o
n
d
,
'
class
if
y
_
m
o
r
tality
'
,
d
eter
m
in
es
th
e
o
v
er
all
s
tatu
s
o
f
th
e
v
ictim
(
'
f
atal
o
r
n
o
n
-
f
atal'
)
b
y
ch
ec
k
i
n
g
wh
eth
er
th
e
r
e
ar
e
an
y
d
ea
d
o
r
in
ju
r
ed
p
eo
p
le
in
th
e
d
ataset.
A
th
ir
d
f
u
n
ctio
n
,
'
cla
s
s
if
y
_
in
ju
r
y
_
g
r
av
ity
'
,
co
n
s
id
er
s
b
o
th
f
atalities
an
d
in
ju
r
i
es
to
class
if
y
th
e
s
ev
er
ity
o
f
t
h
e
in
ju
r
ies
(
'
in
ju
r
y
s
ev
er
ity
'
)
.
R
esu
lt
s
o
f
th
e
class
if
icatio
n
f
u
n
ctio
n
s
wer
e
ad
d
e
d
to
th
e
d
ataset
as
n
ew
co
lu
m
n
s
,
'
f
atal
o
r
n
o
t'
an
d
'
in
ju
r
y
s
ev
er
ity
'
.
T
h
ese
wer
e
th
e
tar
g
et
attr
ib
u
tes
f
o
r
th
is
r
esear
ch
.
Py
th
o
n
was
u
s
ed
in
tan
d
em
with
R
ap
id
Min
er
f
o
r
m
o
r
e
co
m
p
le
x
d
ata
tr
an
s
f
o
r
m
atio
n
s
a
n
d
f
o
r
task
s
r
eq
u
ir
in
g
cu
s
to
m
s
cr
ip
tin
g
.
I
n
co
n
clu
s
io
n
,
d
ata
p
r
ep
r
o
ce
s
s
in
g
in
v
o
lv
ed
t
h
e
m
eticu
lo
u
s
tr
an
s
f
o
r
m
atio
n
,
s
elec
tio
n
,
an
d
im
p
u
tatio
n
o
f
f
ea
tu
r
es,
as
well
a
s
th
e
ex
p
licit
d
e
f
in
itio
n
o
f
th
e
tar
g
et
v
a
r
iab
le.
T
h
ese
p
r
o
ce
d
u
r
es
wer
e
in
d
is
p
en
s
ab
le
f
o
r
g
u
ar
a
n
teein
g
d
ata
q
u
ality
an
d
th
e
r
o
b
u
s
tn
ess
o
f
p
r
ed
ictiv
e
m
o
d
els.
2
.
2
.
Descript
iv
e
s
t
a
t
is
cics
An
in
itial
d
ata
e
x
p
lo
r
atio
n
to
co
m
p
r
eh
e
n
d
th
e
s
co
p
e
o
f
th
e
av
ailab
le
in
f
o
r
m
atio
n
was
co
n
d
u
cted
o
n
th
e
s
am
e
d
atab
ase
in
[
2
2
]
u
s
in
g
E
x
ce
l
g
r
ap
h
s
.
T
h
e
s
tu
d
y
r
ev
ea
led
s
o
m
e
h
id
d
en
in
f
o
r
m
atio
n
s
u
ch
as
th
e
s
ex
o
f
th
e
d
r
iv
er
th
at
is
co
m
m
o
n
in
th
e
ac
cid
en
ts
,
th
e
m
o
s
t
lik
el
y
tim
in
g
wh
e
n
m
o
s
t
ac
cid
e
n
ts
o
cc
u
r
an
d
o
th
e
r
in
f
o
r
m
atio
n
th
at
ca
n
’
t
b
e
g
r
a
b
b
ed
d
ir
ec
tly
with
th
e
m
ass
iv
e
d
ata
p
r
esen
te
d
.
T
h
is
e
x
p
lo
r
atio
n
f
ac
ilit
ated
th
e
id
en
tific
atio
n
o
f
ch
ar
ac
ter
is
tics
th
at
m
ay
h
av
e
r
elev
an
ce
i
n
p
r
ed
ictin
g
ac
cid
e
n
t sev
er
ity
.
B
ef
o
r
e
co
n
d
u
ct
in
g
th
e
r
es
ea
r
c
h
,
f
ew
te
r
m
s
s
h
o
u
l
d
b
e
ex
p
l
ai
n
e
d
:
a
p
er
s
o
n
i
n
a
n
ac
c
i
d
e
n
t
c
an
b
e
eit
h
e
r
k
il
le
d
o
r
in
ju
r
ed
.
I
f
t
h
e
a
cc
i
d
e
n
t
h
as
at
l
ea
s
t
o
n
e
p
er
s
o
n
k
ill
e
d
,
t
h
e
n
a
cc
i
d
en
t
is
ca
t
e
g
o
r
i
ze
d
as
a
f
at
al
ac
ci
d
e
n
t
(
t
h
e
v
a
lu
e
i
n
t
h
e
f
a
tal
it
y
c
o
l
u
m
n
i
n
t
h
e
d
ata
b
as
e
is
1
i
f
a
t
le
ast
o
n
e
p
e
r
s
o
n
k
i
lle
d
an
d
0
i
f
n
o
p
e
r
s
o
n
was
k
il
le
d
)
.
A
p
e
r
s
o
n
c
a
n
b
e
k
il
le
d
im
m
e
d
iate
ly
o
r
d
i
es
wit
h
i
n
3
0
d
a
y
s
d
u
e
a
f
t
er
t
h
e
ca
r
c
r
as
h
.
I
n
t
h
e
c
ase
w
h
e
n
n
o
p
er
s
o
n
is
k
il
le
d
i
n
th
e
ac
ci
d
e
n
t
b
u
t
t
h
er
e
ar
e
i
n
j
u
r
ies
t
h
e
n
t
h
e
a
cc
i
d
en
t
is
n
o
t
f
at
al
.
I
n
th
is
ca
s
e
th
e
i
n
j
u
r
i
es
s
ev
e
r
it
y
is
m
ea
s
u
r
e
d
.
T
h
is
last
is
c
at
eg
o
r
ize
d
i
n
t
o
th
r
e
e
c
lass
es:
i
)
a
p
e
r
s
o
n
is
s
e
r
i
o
u
s
l
y
i
n
j
u
r
e
d
:
an
y
p
e
r
s
o
n
i
n
j
u
r
ed
i
n
a
r
o
a
d
cr
as
h
r
e
q
u
i
r
i
n
g
h
o
s
p
ita
liz
ati
o
n
f
o
r
s
i
x
d
a
y
s
o
r
m
o
r
e
(
t
h
e
v
al
u
e
i
n
t
h
e
i
n
j
u
r
y
co
lu
m
n
in
th
e
d
a
ta
b
as
e
is
2
i
f
at
leas
t
o
n
e
p
e
r
s
o
n
is
s
e
r
i
o
u
s
ly
i
n
j
u
r
e
d
)
;
i
i)
a
p
e
r
s
o
n
s
l
ig
h
tly
i
n
j
u
r
e
d
:
a
n
y
p
er
s
o
n
in
ju
r
e
d
in
a
r
o
a
d
c
r
as
h
r
e
q
u
i
r
i
n
g
m
e
d
i
ca
l
tr
ea
t
m
e
n
t
o
r
h
o
s
p
it
ali
za
t
io
n
o
f
f
ewe
r
t
h
a
n
s
i
x
d
a
y
s
(
t
h
e
v
al
u
e
i
n
t
h
e
i
n
j
u
r
y
co
lu
m
n
i
n
t
h
e
d
at
a
b
ase
is
1
if
n
o
o
n
e
is
s
e
r
i
o
u
s
ly
in
ju
r
ed
an
d
at
le
ast
o
n
e
p
e
r
s
o
n
is
s
li
g
h
tl
y
i
n
j
u
r
ed
)
;
a
n
d
iii
)
t
h
e
r
e
ar
e
n
o
in
j
u
r
ies
a
t
all
i
n
th
e
ac
ci
d
e
n
t
(
t
h
e
v
al
u
e
i
n
t
h
e
i
n
j
u
r
y
c
o
l
u
m
n
i
n
th
e
d
ata
b
as
e
is
0
if
n
o
o
n
e
is
in
j
u
r
e
d
)
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
P
r
ed
ictin
g
th
e
s
ev
erit
y
o
f ro
a
d
tr
a
ffic a
cc
id
en
ts
Mo
r
o
cc
o
:
a
s
u
p
ervis
ed
ma
ch
in
e
…
(
Ha
lim
a
Dri
s
s
i To
u
z
a
n
i)
4465
Usi
n
g
E
x
ce
l
as
well,
s
p
ec
if
ic
d
ata
an
aly
s
is
u
s
in
g
d
if
f
e
r
en
t
t
y
p
es
o
f
g
r
ap
h
s
was
p
r
esen
ted
in
Fig
u
r
e
2
to
m
ea
s
u
r
e
th
e
s
ev
er
ity
o
f
ac
c
id
en
ts
.
T
h
e
g
r
a
p
h
s
s
h
o
w
th
at
f
o
r
th
e
ac
cid
e
n
t
ca
s
es
p
r
esen
ted
in
th
is
d
atab
ase
3
7
%
o
f
p
eo
p
le
ar
e
k
illed
i
n
t
h
e
ac
cid
en
ts
wh
ic
h
is
an
im
p
o
r
tan
t
n
u
m
b
er
in
two
y
ea
r
s
.
I
t
also
s
h
o
ws
th
at
f
o
r
non
-
f
atal
ac
ci
d
en
ts
,
alm
o
s
t 5
0
% o
f
p
eo
p
le
in
v
o
l
v
ed
in
ac
ci
d
en
ts
ar
e
in
ju
r
ed
with
1
2
% ser
i
o
u
s
ly
in
ju
r
e
d
.
Fig
u
r
e
2
.
Acc
id
e
n
t sev
er
ity
an
d
f
atality
p
er
ce
n
tag
e
f
o
r
2
0
1
5
-
2016
I
n
ad
d
itio
n
to
th
e
p
r
im
ar
y
d
a
ta
f
ield
s
,
th
e
d
atab
ase
co
n
tain
s
d
etailed
m
ap
p
in
g
s
f
o
r
a
m
u
ltit
u
d
e
o
f
co
d
ed
v
alu
es,
o
f
f
e
r
in
g
d
escr
i
p
tio
n
s
f
o
r
v
ar
io
u
s
co
n
d
itio
n
s
an
d
ca
teg
o
r
ies,
s
u
ch
as
d
r
iv
er
'
s
licen
s
e
ty
p
e,
v
eh
icle
ca
teg
o
r
ies,
o
cc
u
p
atio
n
s
,
an
d
in
j
u
r
y
t
y
p
es,
am
o
n
g
o
th
er
s
.
T
h
is
allo
ws
f
o
r
a
d
et
ailed
an
d
n
u
an
ce
d
an
aly
s
is
o
f
ac
cid
en
ts
.
T
h
e
a
d
d
itio
n
al
attr
ib
u
tes
ad
d
e
d
,
s
u
c
h
as
th
e
to
tal
n
u
m
b
er
o
f
f
atal
ities
an
d
in
ju
r
ies
as
well
as
th
e
esti
m
ated
s
ev
er
ity
o
f
in
ju
r
ies,
f
u
r
th
er
e
n
r
ich
t
h
e
d
ata
s
et.
T
h
ese
f
ield
s
allo
w
f
o
r
an
ac
cu
r
ate
ass
es
s
m
en
t
o
f
th
e
im
p
ac
t
o
f
ea
ch
ac
cid
en
t,
f
ac
ilit
atin
g
ad
v
an
ce
d
s
tatis
tical
an
aly
s
e
s
to
d
eter
m
in
e
th
e
m
o
s
t
s
ig
n
if
ican
t f
ac
to
r
s
co
n
t
r
ib
u
tin
g
to
r
o
a
d
in
cid
e
n
ts
.
Hea
tm
ap
s
ar
e
o
t
h
er
wa
y
s
to
ea
s
ily
v
is
u
alize
an
d
an
aly
ze
co
m
p
lex
d
ata.
T
h
e
h
ea
tm
ap
i
n
Fig
u
r
e
3
d
is
p
lay
s
th
e
av
er
ag
e
n
u
m
b
er
o
f
p
eo
p
le
in
v
o
lv
e
d
in
th
e
ac
c
id
en
t
f
o
r
d
if
f
er
en
t
h
o
u
r
s
o
f
th
e
d
ay
an
d
d
if
f
er
en
t
d
ay
s
o
f
th
e
wee
k
.
T
h
e
in
ten
s
i
ty
o
f
th
e
co
lo
r
r
e
p
r
esen
ts
th
e
m
ag
n
itu
d
e,
with
d
ar
k
er
co
lo
r
s
in
d
icatin
g
lik
ely
h
ig
h
er
a
v
er
ag
es.
T
h
e
h
ea
tm
ap
was
im
p
lem
en
ted
u
s
in
g
R
ap
id
Min
er
to
o
l
an
d
it
s
h
o
ws
th
at
th
er
e
is
a
r
elatio
n
s
h
ip
b
etwe
en
th
e
n
u
m
b
er
o
f
ac
cid
e
n
ts
th
at
h
ap
p
en
d
u
r
in
g
d
if
f
er
en
t
d
a
y
s
o
v
er
tim
e
.
B
y
o
b
s
er
v
in
g
h
o
w
ce
ll
co
lo
r
s
ch
an
g
e
ac
r
o
s
s
ea
ch
ax
is
,
we
ca
n
n
o
tice
th
at
m
o
s
t
ac
cid
en
ts
o
cc
u
r
d
u
r
in
g
n
ig
h
ttime
u
n
til
alm
o
s
t
9
a
m
.
T
h
is
ca
n
b
e
e
x
p
la
in
e
d
b
y
t
h
e
l
ess
ill
u
m
in
ati
o
n
at
n
i
g
h
t
an
d
t
h
e
t
r
a
f
f
ic
t
im
in
g
f
r
o
m
7
t
o
9
am
.
T
h
e
c
o
l
o
r
is
less
d
a
r
k
b
etw
ee
n
1
2
a
n
d
1
3
p
m
o
n
w
ee
k
d
a
y
s
.
T
h
is
ca
n
b
e
e
x
p
lai
n
ed
b
y
t
h
e
t
r
a
f
f
ic
d
u
r
i
n
g
l
u
n
c
h
ti
m
e
as
we
ll
.
Fig
u
r
e
2
.
Hea
tm
ap
o
f
d
aily
ac
cid
en
ts
b
y
h
o
u
r
s
d
is
tr
ib
u
tio
n
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2252
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
1
4
,
No
.
6
,
Dec
em
b
er
2
0
2
5
:
4
4
6
1
-
4
4
7
3
4466
T
h
e
co
r
r
elatio
n
m
atr
ix
o
f
v
ar
i
ab
les
was
al
s
o
im
p
lem
en
ted
as
s
h
o
wn
in
Fig
u
r
e
4
.
T
h
is
m
at
r
ix
s
h
o
ws
th
e
co
r
r
elatio
n
co
ef
f
icien
ts
b
etwe
en
p
air
s
o
f
v
ar
iab
les
in
o
u
r
d
ataset.
Dar
k
er
co
lo
r
s
(
r
e
d
o
r
v
io
let)
in
d
icate
s
tr
o
n
g
er
p
o
s
itiv
e
o
r
n
e
g
ativ
e
co
r
r
elatio
n
s
.
So
m
e
v
ar
iab
les,
s
u
ch
as
k
illed
_
in
s
tan
tly
,
n
o
n
_
k
illed
,
s
er
io
u
s
ly
_
in
ju
r
ed
,
an
d
lig
h
tly
_
in
ju
r
ed
,
s
h
o
w
h
ig
h
e
r
co
r
r
elatio
n
s
am
o
n
g
th
em
s
elv
es a
s
ex
p
e
cted
.
Fig
u
r
e
4
.
Ma
tr
ix
o
f
c
o
r
r
elatio
n
o
f
v
ar
ia
b
les
2
.
3
.
Su
perv
is
ed
m
a
chine le
a
rning
a
lg
o
rit
hm
s
I
n
p
r
ev
io
u
s
wo
r
k
,
u
n
s
u
p
er
v
is
ed
m
ac
h
i
n
e
lear
n
i
n
g
m
eth
o
d
s
wer
e
ap
p
lied
o
n
a
d
if
f
er
en
t
r
ea
l
r
o
a
d
tr
af
f
ic
ac
cid
en
ts
d
ataset
to
g
ain
v
alu
ab
le
in
s
ig
h
ts
in
to
ac
cid
en
t
p
atter
n
s
a
n
d
t
r
en
d
s
.
R
esu
lts
s
h
o
wed
th
at
ac
cid
en
ts
co
u
ld
b
e
ca
teg
o
r
ize
d
b
y
d
ay
an
d
n
ig
h
t
b
ased
o
n
f
o
u
r
attr
ib
u
tes:
ty
p
e
o
f
co
llis
io
n
,
in
itial
s
h
o
ck
,
an
d
th
e
m
o
v
em
e
n
t
in
th
e
ac
cid
e
n
t.
Valu
ab
le
s
u
g
g
esti
o
n
s
wer
e
th
en
s
en
t
to
th
e
Min
is
tr
y
o
f
T
r
an
s
p
o
r
t
to
h
elp
r
ed
u
ce
r
o
a
d
ac
cid
en
ts
[
2
3
]
an
d
in
jecte
d
in
to
a
f
u
zz
y
lo
g
ic
co
n
tr
o
ller
to
tr
ain
a
s
em
i
-
au
to
n
o
m
o
u
s
ca
r
to
tak
e
th
e
r
ig
h
t
d
ec
is
io
n
wh
e
n
th
e
d
r
iv
er
d
o
esn
’
t
r
ea
ct
tim
ely
a
n
d
p
r
o
p
e
r
ly
[
2
4
]
.
Ho
wev
er
,
s
in
ce
th
e
co
u
n
tr
y
o
f
Mo
r
o
cc
o
h
ad
a
h
ig
h
co
s
t
o
f
r
o
ad
cr
ash
es
in
2
0
2
2
esti
m
ated
at
E
UR
1
.
6
b
illi
o
n
with
o
u
t
c
o
u
n
tin
g
th
e
co
s
t
o
f
s
lig
h
t
in
ju
r
ies
an
d
p
r
o
p
er
ty
d
am
ag
e
[
7
]
as
s
h
o
wn
in
T
ab
le
2
,
m
o
r
e
d
ata
a
n
aly
tics
u
s
in
g
s
u
p
er
v
is
ed
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
s
was
co
n
d
u
cted
in
th
is
r
esear
ch
to
p
e
r
f
o
r
m
a
n
ea
r
ly
p
r
ed
ict
o
f
ac
c
id
en
ts
s
ev
er
ity
an
d
th
er
ef
o
r
e
h
elp
r
e
d
u
ce
th
is
co
s
t
in
th
e
co
u
n
tr
y
.
T
ab
le
2
.
C
o
s
t o
f
r
o
ad
c
r
ash
es in
Mo
r
o
cc
o
,
2
0
2
2
F
a
t
a
l
i
t
i
e
s
U
n
i
t
c
o
st
(
EU
R
)
N
u
mb
e
r
To
t
a
l
c
o
st
(
EU
R
)
2
5
6
1
3
6
3
4
9
9
0
.
9
b
i
l
l
i
o
n
S
e
r
i
o
u
sl
y
i
n
j
u
r
e
d
6
4
0
3
3
1
0
9
2
9
0
.
7
b
i
l
l
i
o
n
To
t
a
l
1
.
6
b
i
l
l
i
o
n
T
h
e
d
ataset
u
s
ed
i
n
th
is
r
esear
ch
co
n
tain
s
v
ar
io
u
s
in
f
o
r
m
atio
n
s
u
ch
as
ac
cid
en
t
tim
e
,
lo
ca
tio
n
,
ev
e
n
t
ty
p
e,
an
d
b
r
ig
h
tn
ess
wh
ich
p
r
o
v
id
e
cr
u
cial
in
s
ig
h
ts
in
to
ac
c
id
en
t
cir
cu
m
s
tan
ce
s
f
o
r
th
e
m
o
d
el
d
ev
elo
p
m
en
t.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
P
r
ed
ictin
g
th
e
s
ev
erit
y
o
f ro
a
d
tr
a
ffic a
cc
id
en
ts
Mo
r
o
cc
o
:
a
s
u
p
ervis
ed
ma
ch
in
e
…
(
Ha
lim
a
Dri
s
s
i To
u
z
a
n
i)
4467
T
h
e
tar
g
et
v
ar
iab
le,
f
atalities
,
o
r
in
ju
r
ies
was
s
elec
ted
to
m
ea
s
u
r
e
th
e
s
ev
e
r
ity
o
f
th
e
a
cc
id
en
t.
I
n
ca
s
e
th
e
ac
cid
en
t
is
n
o
t
f
atal,
an
o
th
er
tar
g
et
v
ar
iab
le
is
ch
o
s
en
to
r
ep
r
esen
t
a
ca
teg
o
r
izatio
n
o
f
in
j
u
r
y
s
ev
er
ity
in
th
e
ac
cid
en
t.
Dif
f
er
en
t
m
ac
h
in
e
le
ar
n
in
g
m
o
d
els
to
p
r
e
d
ict
th
e
f
atality
o
r
th
e
in
ju
r
y
s
ev
er
ity
o
f
th
e
ac
cid
en
ts
wer
e
ap
p
lied
an
d
ev
alu
ate
d
in
th
is
r
esear
ch
.
T
h
e
KNN
alg
o
r
ith
m
was
u
s
ed
to
p
r
ed
ict
th
e
s
ev
er
it
y
o
f
r
o
ad
ac
ci
d
en
ts
b
ased
o
n
th
e
ch
ar
ac
ter
is
tics
o
f
ea
ch
ac
cid
en
t.
KNN,
ch
o
s
en
f
o
r
its
s
im
p
licity
an
d
ad
ap
tab
ilit
y
to
m
ed
iu
m
-
s
ized
d
atasets
,
class
if
ies
p
o
in
ts
b
y
d
eter
m
in
in
g
th
e
m
ajo
r
it
y
class
am
o
n
g
th
eir
k
n
ea
r
est
n
eig
h
b
o
r
s
.
I
n
th
e
co
n
tex
t
o
f
ac
cid
e
n
t
s
ev
er
ity
p
r
ed
ictio
n
,
KNN
e
x
am
in
es
ac
ci
d
en
t
ch
a
r
ac
ter
is
tics
,
id
en
tifie
s
n
ea
r
est
n
eig
h
b
o
r
s
,
an
d
ass
ig
n
s
a
s
ev
er
ity
lev
el
b
a
s
ed
o
n
th
e
m
ajo
r
ity
class
am
o
n
g
th
ese
n
eig
h
b
o
r
s
.
T
h
e
L
R
an
d
SVM
alg
o
r
ith
m
s
wer
e
also
u
s
ed
.
T
h
ese
alg
o
r
it
h
m
s
ar
e
k
n
o
wn
b
y
th
eir
r
o
b
u
s
tess
,
an
d
th
ey
wer
e
o
f
ten
u
s
ed
f
o
r
class
if
icatio
n
an
d
p
r
ed
ictio
n
.
T
h
e
NB
class
if
ier
was
also
ch
o
s
en
to
b
e
u
s
ed
in
o
u
r
r
esear
c
h
f
o
r
its
s
p
ee
d
an
d
ef
f
icien
cy
in
m
ak
in
g
p
r
ed
ictio
n
s
with
lar
g
e
d
atasets
in
clu
d
in
g
in
d
ep
en
d
e
n
t
f
ea
tu
r
es.
T
h
en
th
e
ANN
wi
th
b
ac
k
p
r
o
p
ag
atio
n
was
also
em
p
lo
y
ed
in
th
is
r
esear
ch
to
p
r
e
d
ict
th
e
s
ev
er
ity
o
f
ac
cid
en
ts
s
in
ce
it
is
a
v
er
y
p
o
wer
f
u
l
alg
o
r
ith
m
to
p
r
o
v
id
e
s
o
lu
tio
n
s
to
co
m
p
lex
p
r
o
b
lem
s
.
I
n
o
u
r
ca
s
e,
t
h
e
A
NN
in
clu
d
ed
1
5
n
o
d
es
as
in
p
u
t
lay
er
s
,
2
h
id
d
en
lay
er
s
with
6
n
o
d
es,
an
d
o
u
tp
u
t
lay
er
s
with
s
in
g
le
n
o
d
es.
T
h
e
r
ec
tifie
d
lin
ea
r
u
n
it
(
R
eL
U
)
ac
tiv
atio
n
f
u
n
ctio
n
was
ap
p
lied
to
in
tr
o
d
u
ce
n
o
n
-
lin
ea
r
ity
in
to
th
e
m
o
d
el.
B
in
ar
y
c
r
o
s
s
-
en
tr
o
p
y
l
o
s
s
,
s
u
itab
le
f
o
r
b
in
a
r
y
class
if
icatio
n
p
r
o
b
lem
s
,
was
u
s
ed
f
o
r
f
atality
p
r
ed
ictio
n
.
Fo
r
in
ju
r
y
s
ev
er
ity
p
r
e
d
ictio
n
,
wh
ich
in
v
o
lv
es
m
u
lti
-
class
class
if
ica
tio
n
,
ca
teg
o
r
ical
cr
o
s
s
-
en
tr
o
p
y
lo
s
s
was
em
p
lo
y
ed
.
R
an
d
o
m
f
o
r
est
alg
o
r
ith
m
was
als
o
ap
p
lied
o
n
th
e
d
ataset
to
f
i
n
d
t
h
e
m
o
s
t im
p
o
r
tan
t f
ac
to
r
s
th
at
co
n
tr
ib
u
ted
to
in
cr
ea
s
e
th
e
ac
c
id
en
ts
s
ev
er
ity
.
3.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
T
h
e
d
ataset
was
d
iv
id
ed
in
to
tr
ain
in
g
(
8
0
%)
an
d
test
in
g
(
2
0
%)
s
ets.
T
h
is
d
iv
is
io
n
en
s
u
r
es
th
at
m
o
d
els
will
n
o
t
o
n
ly
m
e
m
o
r
iz
e
th
e
tr
ain
in
g
d
ata
u
s
ed
f
o
r
th
e
tr
ain
in
g
p
h
ase
b
u
t
ca
n
g
en
er
alize
its
p
r
ed
ictio
n
s
to
n
ew
o
b
s
er
v
atio
n
s
.
T
h
e
m
o
d
el'
s
p
r
ed
ictio
n
was
ass
ess
ed
o
n
th
e
test
s
et
f
o
r
all
t
h
e
alg
o
r
ith
m
s
as
s
h
o
wn
in
T
ab
le
3
,
with
ac
cu
r
ac
y
s
er
v
in
g
as
th
e
m
ain
p
er
f
o
r
m
an
ce
m
etr
ic.
A
h
ig
h
ac
cu
r
ac
y
im
p
lies
th
at
th
e
m
o
d
el
ca
n
ef
f
ec
tiv
ely
class
if
y
r
o
ad
ac
cid
en
t sev
er
ity
o
n
u
n
s
ee
n
d
ata.
Featu
r
es
wer
e
also
s
tan
d
ar
d
iz
ed
u
s
in
g
th
e
s
tan
d
ar
d
d
ev
iatio
n
m
eth
o
d
(
n
o
r
m
aliza
tio
n
m
et
h
o
d
f
o
r
th
e
ANN)
.
T
h
is
p
r
o
ce
s
s
is
cr
u
ci
al
to
en
s
u
r
e
th
at
ea
ch
f
ea
tu
r
e
co
n
tr
ib
u
tes
eq
u
itab
ly
to
t
h
e
p
r
ed
ictio
n
,
th
u
s
p
r
ev
en
tin
g
d
if
f
er
en
ce
s
in
s
ca
le
f
r
o
m
b
iasi
n
g
t
h
e
m
o
d
el
in
f
av
o
r
o
f
ce
r
tain
v
a
r
iab
les.
Fo
llo
win
g
th
e
tr
ain
in
g
p
h
ase,
th
e
m
eth
o
d
s
wer
e
em
p
l
o
y
ed
to
p
r
ed
ict
th
e
s
ev
er
ity
o
f
th
e
ac
cid
en
t o
n
th
e
test
in
g
d
at
a.
3
.
1
.
M
a
chine le
a
rning
m
o
de
ls
ev
a
lua
t
io
n
T
h
e
m
o
d
el'
s
ef
f
ec
tiv
en
ess
was
ass
es
s
ed
u
s
in
g
th
e
d
if
f
er
en
t
m
etr
ics,
F1
-
s
co
r
e,
r
ec
all,
p
r
ec
is
io
n
,
an
d
ac
cu
r
ac
y
g
au
g
in
g
th
e
r
atio
o
f
co
r
r
ec
t
p
r
ed
ictio
n
s
to
th
e
to
t
al
p
r
ed
ictio
n
s
.
All
th
e
m
en
tio
n
ed
m
etr
ics
wer
e
ca
lcu
lated
f
r
o
m
th
e
co
n
f
u
s
io
n
m
atr
ix
g
en
e
r
ated
b
y
ea
ch
m
o
d
el
u
s
in
g
th
e
(
1
)
to
(
3
)
.
=
TP
+
(
1
)
=
TP
+
(
2
)
1
−
=
2
+
(
3
)
W
h
er
e
TP
is
tr
u
e
p
o
s
itiv
e
,
FP
is
f
alse
p
o
s
itiv
e
,
T
N
is
tr
u
e
n
eg
ativ
e
,
an
d
FN
is
f
alse
n
eg
ativ
e.
T
h
ese
ar
e
th
e
v
alu
es p
r
o
v
i
d
ed
b
y
th
e
c
o
ef
f
ic
ien
t m
atr
ix
f
o
r
ea
ch
m
o
d
el.
T
h
e
r
ea
l a
cc
u
r
ac
y
was c
alcu
lated
u
s
in
g
th
e
(
4
)
.
=
TP
+
FN
+
+
+
(
4
)
A
s
u
m
m
ar
y
o
f
r
esu
lts
f
o
r
all
m
ac
h
in
e
lear
n
in
g
m
o
d
els
u
s
ed
in
th
is
r
esear
ch
to
p
r
ed
ict
th
e
f
atality
o
r
th
e
in
ju
r
ies
s
ev
er
ity
in
r
o
ad
a
cc
id
en
ts
o
n
Mo
r
o
cc
o
is
s
h
o
w
n
in
T
ab
les
3
an
d
4
.
R
esu
lts
s
h
o
wed
th
at
th
e
m
o
s
t
ef
f
ec
tiv
e
m
o
d
els
th
at
g
av
e
th
e
b
est
ac
cu
r
ac
y
to
p
r
ed
ict
ac
c
id
en
ts
f
atality
o
r
ac
cid
en
ts
in
j
u
r
ies
s
ev
er
ity
wer
e
th
e
SVM
(
0
.
9
9
ac
cu
r
ac
y
f
o
r
ac
cid
en
t
f
atality
an
d
0
.
7
ac
cu
r
ac
y
f
o
r
ac
cid
en
ts
in
ju
r
ies
s
ev
er
ity
)
an
d
th
e
L
R
(
0
.
9
8
ac
c
u
r
ac
y
f
o
r
ac
cid
en
t
f
a
tality
an
d
0
.
7
ac
cu
r
ac
y
f
o
r
ac
cid
en
ts
in
ju
r
ies
s
ev
er
ity
)
.
ANN
g
av
e
also
a
h
ig
h
ac
cu
r
ac
y
f
o
r
th
e
ac
cid
en
t
’
s
f
atality
(
0
.
9
8
)
wh
ile
it
g
a
v
e
th
e
lo
west
ac
cu
r
ac
y
f
o
r
th
e
i
n
ju
r
y
’
s
s
ev
er
ity
(
0
.
5
7
)
th
is
m
ig
h
t
b
e
d
u
e
to
th
e
m
ajo
r
ity
o
f
ca
te
g
o
r
ical
f
ea
t
u
r
es
u
s
ed
in
th
e
d
ataset.
T
o
o
v
er
c
o
m
e
th
e
ch
allen
g
e
o
f
lim
ited
r
ea
l
-
tim
e
tr
af
f
ic
i
n
f
o
r
m
atio
n
,
th
is
p
r
e
d
ictiv
e
m
o
d
el
ca
n
lead
p
u
b
lic
s
ec
u
r
ity
f
o
r
ce
s
to
war
d
s
ar
ea
s
with
a
s
ig
n
if
ican
t
r
is
k
o
f
s
er
io
u
s
ac
cid
en
ts
,
f
ac
ilit
atin
g
p
r
o
ac
tiv
e
i
n
ter
v
en
tio
n
an
d
th
er
ef
o
r
e
lim
i
tin
g
f
atal
ac
cid
en
ts
f
r
o
m
h
ap
p
e
n
in
g
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2252
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
1
4
,
No
.
6
,
Dec
em
b
er
2
0
2
5
:
4
4
6
1
-
4
4
7
3
4468
T
ab
le
3
.
E
v
alu
atio
n
m
et
r
ics co
m
p
ar
is
o
n
f
o
r
th
e
p
r
ed
ictiv
e
m
eth
o
d
s
u
s
ed
f
o
r
ac
cid
e
n
ts
f
atality
M
a
c
h
i
n
e
l
e
a
r
n
i
n
g
m
o
d
e
l
P
r
e
c
i
s
i
o
n
R
e
c
a
l
l
F1
-
s
c
o
r
e
A
c
c
u
r
a
c
y
S
V
M
0
.
9
9
0
.
9
8
0
.
9
8
0
.
9
9
K
N
N
0
.
9
9
0
.
9
6
0
.
9
7
0
.
9
9
NB
0
.
9
9
0
.
9
7
0
.
9
7
0
.
9
7
LR
0
.
9
9
0
.
9
8
0
.
9
8
0
.
9
9
ANN
0
.
9
8
0
.
9
7
0
.
9
7
0
.
9
9
T
ab
le
4
.
E
v
alu
atio
n
m
et
r
ics co
m
p
ar
is
o
n
f
o
r
th
e
p
r
ed
ictiv
e
m
eth
o
d
s
u
s
ed
f
o
r
ac
cid
e
n
ts
in
ju
r
ies s
ev
er
ity
M
a
c
h
i
n
e
l
e
a
r
n
i
n
g
m
o
d
e
l
P
r
e
c
i
s
i
o
n
R
e
c
a
l
l
F1
-
s
c
o
r
e
A
c
c
u
r
a
c
y
S
V
M
0
.
7
0
.
9
9
0
.
8
2
0
.
7
K
N
N
0
.
7
1
0
.
9
2
0
.
8
0
.
6
8
NB
0
.
7
0
.
9
1
0
.
8
0
.
6
7
LR
0
.
7
0
.
9
9
0
.
8
2
0
.
7
ANN
0
.
6
6
0
.
5
7
0
.
6
2
0
.
5
7
Acc
id
en
ts
f
atality
o
r
in
ju
r
ies s
ev
er
ity
p
r
ed
ictio
n
g
en
er
ally
s
e
ar
ch
es th
e
r
elatio
n
s
h
ip
b
etwe
en
v
ictim
s
’
s
ev
er
ity
an
d
r
elev
an
t
f
ac
to
r
s
(
s
u
ch
as
h
u
m
an
,
v
eh
icle,
an
d
eq
u
ip
m
e
n
t
o
r
en
v
i
r
o
n
m
en
tal
f
ac
to
r
s
)
.
A
co
m
p
r
eh
e
n
s
iv
e
an
aly
s
is
o
f
all
f
ac
to
r
s
co
n
tr
i
b
u
tin
g
t
o
th
e
s
e
v
er
ity
o
f
ac
cid
en
ts
is
n
ec
ess
ar
y
to
d
ef
i
n
e
th
e
r
ea
l
n
ee
d
s
o
f
p
e
o
p
le
in
r
o
ad
t
r
af
f
i
c
ac
cid
en
ts
[
2
5
]
,
[
2
6
]
.
T
h
is
a
n
aly
s
is
g
iv
es
cr
itical
in
f
o
r
m
at
io
n
to
em
er
g
en
cy
s
er
v
ices
an
d
tr
af
f
ic
m
an
a
g
er
s
to
im
p
lem
en
t
m
ea
s
u
r
es
to
r
ed
u
ce
th
e
s
id
e
ef
f
ec
ts
o
f
th
e
ac
c
id
en
t
lik
e
o
f
f
e
r
in
g
f
aster
m
ed
ical
ass
is
tan
ce
to
p
e
o
p
le
in
j
u
r
ed
in
th
e
ac
ci
d
en
ts
a
n
d
th
e
r
ef
o
r
e,
m
in
im
izin
g
d
ea
t
h
s
[
5
]
.
T
o
f
i
n
d
o
u
t
th
e
m
o
s
t
im
p
o
r
tan
t
f
ac
to
r
s
th
a
t
ca
n
in
cr
ea
s
e
r
o
ad
tr
af
f
ic
ac
c
id
en
ts
s
ev
er
ity
f
r
o
m
all
th
e
g
i
v
en
in
f
o
r
m
atio
n
in
o
u
r
d
ataset,
an
o
th
er
m
ac
h
i
n
e
l
ea
r
n
in
g
al
g
o
r
ith
m
,
r
a
n
d
o
m
f
o
r
est
,
was
ap
p
lied
o
n
th
e
s
am
e
d
ataset.
R
esu
lt
s
wil
l
b
e
d
is
cu
s
s
ed
in
th
e
n
e
x
t sectio
n
.
3
.
2
.
E
m
bedd
ed
f
ea
t
ure
s
elec
t
io
n in m
a
chine le
a
rning
R
an
d
o
m
f
o
r
est
m
eth
o
d
co
llect
s
d
if
f
er
en
t
r
esu
lts
f
r
o
m
d
ec
is
io
n
tr
ee
s
an
d
co
m
b
in
es
th
em
to
f
ig
u
r
e
o
u
t
wh
ich
attr
ib
u
tes
ar
e
th
e
m
o
s
t
im
p
o
r
tan
t
in
m
ak
in
g
a
d
ec
is
io
n
.
T
h
is
tech
n
iq
u
e
h
elp
s
u
n
d
er
s
t
an
d
wh
ich
f
ea
tu
r
es
im
p
ac
t
th
e
o
u
tco
m
e
m
o
s
t.
Fig
u
r
e
5
s
h
o
w
th
e
1
0
m
o
s
t
im
p
o
r
tan
t
f
ea
tu
r
es
th
at
lead
s
to
th
e
d
ec
is
io
n
s
elec
te
d
b
y
th
e
r
an
d
o
m
f
o
r
est
alg
o
r
ith
m
,
wh
er
e
Fig
u
r
e
5
(
a)
s
h
o
ws
th
e
f
atality
p
r
ed
ictio
n
an
d
Fig
u
r
e
5
(
b
)
s
h
o
ws
th
e
in
ju
r
y
'
s
s
ev
er
ity
p
r
ed
ictio
n
.
R
esu
lts
s
h
o
w
th
at
th
e
m
o
s
t
i
m
p
o
r
tan
t
f
ac
to
r
s
th
at
co
n
tr
ib
u
tes
th
e
f
atality
o
f
ac
cid
en
ts
o
r
i
n
ju
r
ies
s
ev
er
ity
ar
e
r
elate
d
to
t
h
e
h
u
m
an
f
a
cto
r
(
d
r
iv
er
’
s
a
g
e,
d
r
iv
in
g
lic
en
s
e
y
ea
r
,
d
r
iv
e
r
’
s
p
r
o
f
ess
io
n
,
an
d
m
is
tak
es
co
m
m
itted
b
y
th
e
d
r
iv
er
)
f
o
llo
wed
b
y
th
e
v
eh
icle
f
ac
to
r
(
ty
p
e
o
f
v
eh
icu
le
an
d
v
eh
icu
le
u
s
ag
e)
.
T
h
e
e
n
v
ir
o
n
m
en
t
f
ac
to
r
(
wea
th
e
r
,
r
o
ad
c
o
n
d
itio
n
s
,
a
n
d
v
is
ib
ilit
y
)
ex
h
ib
it
a
co
m
p
ar
ativ
ely
wea
k
p
o
s
itiv
e
co
r
r
elatio
n
with
th
e
s
ev
er
ity
o
f
in
ju
r
ies an
d
f
at
alities
in
r
o
ad
tr
af
f
ic
ac
cid
e
n
ts
.
(
a)
(
b
)
Fig
u
r
e
5
.
R
an
d
o
m
f
o
r
est
im
p
o
r
tan
ce
f
ea
tu
r
e
s
elec
tio
n
tec
h
n
i
q
u
e
ap
p
lied
to
th
e
d
ataset
f
o
r
(
a)
f
atality
p
r
e
d
ictio
n
an
d
(
b
)
i
n
ju
r
y
'
s
s
ev
er
ity
p
r
ed
ictio
n
On
e
o
f
t
h
e
m
o
s
t
im
p
o
r
tan
t
f
a
cto
r
f
r
o
m
th
e
h
u
m
an
f
ac
t
o
r
s
was
th
e
d
r
iv
er
’
s
ag
e.
T
o
f
i
n
d
o
u
t
ex
ac
tly
th
e
ag
e
r
an
g
e
o
f
th
e
d
r
iv
er
t
h
at
lead
s
to
in
cr
ea
s
e
th
e
s
ev
er
ity
o
f
th
e
ac
cid
en
ts
,
two
co
lu
m
n
s
’
ch
ar
ts
wer
e
d
ed
u
ce
d
f
r
o
m
t
h
e
d
ata
s
et.
T
h
e
ch
ar
t
p
r
esen
ted
in
Fig
u
r
e
6
c
o
n
f
ir
m
s
th
at
th
e
d
r
iv
e
r
’
s
ag
e
h
as
an
im
p
ac
t
o
n
th
e
s
ev
er
ity
o
f
t
h
e
ac
cid
en
ts
,
wh
er
e
Fig
u
r
e
6
(
a)
s
h
o
ws
th
e
f
at
al
ac
cid
en
ts
an
d
Fig
u
r
e
6
(
b
)
s
h
o
ws
th
e
ac
cid
en
ts
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
P
r
ed
ictin
g
th
e
s
ev
erit
y
o
f ro
a
d
tr
a
ffic a
cc
id
en
ts
Mo
r
o
cc
o
:
a
s
u
p
ervis
ed
ma
ch
in
e
…
(
Ha
lim
a
Dri
s
s
i To
u
z
a
n
i)
4469
with
at
least
o
n
e
p
er
s
o
n
s
u
f
f
er
in
g
s
er
io
u
s
in
ju
r
ies
.
T
h
e
co
m
p
ar
is
o
n
s
h
o
ws
th
at
m
o
s
t
s
e
v
er
e
ac
cid
en
ts
ar
e
ca
u
s
ed
b
y
y
o
u
n
g
d
r
i
v
er
s
(
ag
e
r
an
g
e
b
etwe
en
1
8
a
n
d
2
7
)
.
T
h
is
ca
n
b
e
e
x
p
lain
ed
b
y
th
e
f
ac
t
th
at
y
o
u
n
g
d
r
i
v
er
s
d
o
n
o
t
h
av
e
e
n
o
u
g
h
ex
p
e
r
ien
c
e
with
d
r
iv
in
g
,
r
o
a
d
d
an
g
e
r
s
,
d
is
tan
ce
s
,
an
d
esti
m
atin
g
s
p
ee
d
.
T
h
ey
o
v
er
v
alu
e
th
eir
d
r
iv
in
g
s
k
ills
th
at
ca
n
lead
to
s
p
ee
d
i
n
g
a
n
d
n
o
n
-
ad
ju
s
tm
en
t
o
f
d
r
iv
in
g
n
o
t
a
war
e
with
th
e
r
o
ad
d
if
f
icu
lties
an
d
s
u
r
f
ac
es
[
2
5
]
.
(
a)
(
b
)
Fig
u
r
e
6
.
T
o
tal
n
u
m
b
er
b
y
d
r
iv
er
'
s
ag
e
r
an
g
e
o
f
(
a)
f
atal
ac
cid
en
ts
an
d
(
b
)
ac
cid
en
ts
with
a
t le
ast o
n
e
p
er
s
o
n
s
u
f
f
er
in
g
s
er
io
u
s
in
ju
r
ies
B
ased
o
n
th
e
r
esu
lts
f
o
u
n
d
in
th
is
r
esear
ch
,
th
er
e
s
h
o
u
ld
b
e
an
ad
ju
s
tm
en
t
o
f
th
e
d
r
iv
in
g
e
d
u
ca
tio
n
al
p
r
o
g
r
a
m
s
an
d
m
an
a
g
em
en
t
m
eth
o
d
o
lo
g
ies
in
th
e
f
ield
o
f
r
o
ad
s
af
ety
.
Giv
in
g
m
o
r
e
im
p
o
r
tan
ce
to
t
h
e
h
u
m
an
f
ac
to
r
an
d
d
iv
id
i
n
g
it
in
to
m
o
r
e
s
p
ec
if
ic
attr
ib
u
tes
o
r
im
p
ac
t
f
ac
to
r
s
will
d
ef
in
itely
h
elp
th
e
tr
af
f
ic
au
th
o
r
ities
to
s
tim
u
late
ed
u
ca
tio
n
al
p
r
o
g
r
am
s
esp
ec
ially
in
th
e
b
eg
i
n
n
er
d
r
iv
er
ed
u
ca
tio
n
.
As
a
n
ex
ten
s
io
n
to
t
h
is
r
esear
ch
,
th
e
s
ev
er
ity
o
f
th
e
ac
cid
en
ts
ca
n
b
e
m
ea
s
u
r
ed
b
y
d
r
iv
er
’
s
a
g
e
a
n
d
th
e
tim
e
o
f
d
ay
(
d
ay
tim
e
o
r
n
ig
h
ttime
)
t
o
h
ig
h
lig
h
t
ex
ac
tly
th
e
p
er
io
d
o
f
t
h
e
d
ay
w
h
en
th
e
y
o
u
n
g
an
d
m
i
d
d
le
-
ag
e
d
d
r
iv
er
s
ar
e
m
o
r
e
lik
ely
to
s
u
s
tain
f
atal
o
r
s
ev
er
e
in
j
u
r
y
ac
cid
en
ts
.
4.
P
RO
P
O
SE
D
SO
L
UT
I
O
N
SVM
is
th
e
m
ac
h
in
e
lear
n
in
g
m
eth
o
d
th
at
c
o
u
ld
p
r
e
d
ict
m
o
s
t
ac
cu
r
ately
t
h
e
ac
cid
e
n
ts
f
atality
o
r
s
ev
er
ity
,
f
o
r
th
is
r
ea
s
o
n
,
t
h
e
al
g
o
r
ith
m
was
ap
p
lied
ag
ain
tak
in
g
in
to
ac
c
o
u
n
t
o
n
ly
th
e
m
o
s
t
im
p
o
r
tan
t
f
ac
to
r
s
lead
in
g
to
in
cr
ea
s
e
ac
cid
en
ts
s
ev
er
ity
.
T
h
e
ac
cu
r
ac
y
o
f
th
e
m
o
d
el
was
ag
ain
9
9
%
f
o
r
ac
cid
en
ts
f
atality
an
d
7
0
%
f
o
r
ac
cid
e
n
ts
s
ev
er
ity
.
No
wad
ay
s
,
ca
r
s
a
r
e
eq
u
ip
p
e
d
with
ADAS
th
at
ass
is
t
d
r
i
v
er
s
with
th
e
s
af
e
o
p
er
atio
n
o
f
a
v
e
h
icle
s
u
ch
as
au
to
m
atic
em
er
g
e
n
cy
b
r
a
k
in
g
(
AE
B
)
a
n
d
f
u
ll
au
t
o
b
r
ak
e
an
d
p
ed
estria
n
d
etec
tio
n
(
C
W
A
B
-
PD)
s
y
s
tem
s
as
s
tated
in
[
2
7
]
.
As
a
p
r
o
p
o
s
ed
s
o
lu
tio
n
to
th
is
s
er
io
u
s
r
ea
l
-
wo
r
ld
p
r
o
b
lem
,
a
ca
r
eq
u
ip
p
e
d
with
a
p
r
ed
ictio
n
m
o
d
el
s
u
ch
as
th
e
SVM
m
o
d
el,
a
ca
m
er
a,
s
en
s
o
r
s
as
wel
l
as
ADA
S
s
y
s
tem
s
ca
n
p
er
f
o
r
m
an
ea
r
ly
p
r
e
d
ictio
n
o
f
ac
cid
e
n
ts
s
ev
er
ity
.
T
h
e
ca
r
ca
n
also
b
e
e
q
u
ip
p
ed
with
3
s
ig
n
als,
o
n
e
in
d
icatin
g
th
at
th
e
ac
cid
e
n
t
will
b
e
f
atal,
th
e
o
th
er
two
lig
h
ts
will
in
d
icate
th
e
in
ju
r
ies
s
e
v
er
ity
lev
el
o
f
th
e
ac
cid
en
t th
at
is
ab
o
u
t t
o
h
ap
p
e
n
.
T
h
e
SVM
m
o
d
el
d
ep
lo
y
e
d
in
t
o
a
ca
r
will
co
llect
th
e
in
p
u
t
f
r
o
m
b
o
t
h
th
e
d
r
iv
er
an
d
th
e
en
v
ir
o
n
m
e
n
t.
Mo
s
t
o
f
th
e
v
ar
iab
les
a
r
e
s
tatic
an
d
ca
n
b
e
en
ter
ed
in
to
t
h
e
s
y
s
tem
b
ef
o
r
e
d
r
iv
in
g
s
u
c
h
as:
d
er
iv
e
r
’
s
ag
e,
p
r
o
f
ess
io
n
(
jo
b
)
,
g
e
n
d
er
,
d
r
iv
in
g
licen
s
e
ca
teg
o
r
y
(
D
L
_
ca
teg
o
r
y
)
,
d
r
i
v
in
g
licen
s
e
ex
p
ir
atio
n
d
ate
(
DL
_
E
x
p
ir
atio
n
)
,
v
eh
icle
u
s
ag
e
,
an
d
v
e
h
icle
ty
p
e.
Fo
r
th
e
o
t
h
er
f
ac
t
o
r
s
,
a
ca
m
er
a
will
b
e
a
s
s
o
ciate
d
with
th
e
s
y
s
tem
to
s
ca
n
th
e
p
atien
t’
s
p
h
ase.
Fatig
u
e
d
etec
tio
n
was
also
p
er
f
o
r
m
ed
u
s
in
g
d
ee
p
lear
n
in
g
alg
o
r
ith
m
s
[
2
8
]
.
I
n
o
u
r
r
esear
ch
,
we
p
r
o
p
o
s
ed
th
at
th
e
in
f
o
r
m
atio
n
is
s
en
t
to
a
f
u
zz
y
lo
g
ic
alg
o
r
ith
m
th
at
c
an
p
r
o
v
id
e
d
r
iv
er
’
s
p
h
y
s
ical
co
n
d
itio
n
s
as
a
n
o
u
tp
u
t
(
d
is
ab
ilit
y
,
d
r
u
n
k
,
n
ar
c
o
tics
+d
r
u
g
s
asleep
,
tire
d
,
s
u
d
d
e
n
s
ick
n
ess
o
r
o
th
er
s
)
.
C
ar
ca
m
er
a
alo
n
g
with
s
en
s
o
r
s
ca
n
also
s
ca
n
th
e
s
u
r
r
o
u
n
d
in
g
en
v
ir
o
n
m
en
t
a
n
d
c
o
llect
th
e
d
ata
r
elate
d
to
t
h
e
m
is
tak
es
co
m
m
itted
b
y
th
e
d
r
iv
er
(
f
ailu
r
e
to
o
b
e
y
r
ed
lig
h
t
s
o
r
s
to
p
s
ig
n
s
,
f
ailu
r
e
to
co
m
p
ly
with
p
r
io
r
ities
,
cr
o
s
s
ed
m
ix
ed
co
n
tin
u
o
u
s
lin
e,
u
n
r
ep
o
r
ted
/d
e
f
ec
tiv
e
ir
r
eg
u
lar
o
p
er
atio
n
,
was
d
r
iv
in
g
wi
th
o
u
t
p
r
ec
a
u
tio
n
s
,
s
p
ee
d
in
g
,
was
d
r
iv
in
g
in
a
p
r
o
h
ib
ited
ar
ea
,
d
ef
ec
tiv
e
s
to
p
o
r
p
ar
k
in
g
o
r
o
th
er
s
)
an
d
m
is
tak
es
n
o
t
r
elate
d
t
o
th
e
d
r
iv
er
(
an
im
als
o
n
t
h
e
r
o
ad
,
o
b
s
tacle
ab
a
n
d
o
n
ed
o
n
th
e
r
o
ad
way
,
tire
e
n
clo
s
u
r
e,
d
an
g
er
o
u
s
h
o
le
in
th
e
r
o
ad
way
,
a
cc
id
en
tal
b
r
ea
k
ag
e
o
f
th
e
win
d
s
h
ield
,
wo
r
k
s
ite
n
o
t
r
ep
o
r
ted
,
o
r
o
t
h
er
s
)
.
T
h
e
d
a
ta
th
en
ca
n
b
e
s
en
t
to
th
e
m
o
d
el
an
d
a
n
ea
r
ly
p
r
ed
ictio
n
o
f
th
e
ac
cid
e
n
t’
s
s
ev
er
ity
will
b
e
d
eter
m
in
ed
.
T
h
e
co
r
e
o
f
o
u
r
s
etu
p
in
v
o
lv
ed
a
s
im
u
lated
v
eh
icle
en
v
ir
o
n
m
en
t
eq
u
ip
p
ed
with
a
3
v
is
u
al
s
ig
n
alin
g
s
y
s
tem
,
o
n
e
in
d
icatin
g
th
at
th
e
ac
cid
en
t
will
b
e
f
atal,
th
e
o
th
e
r
two
lig
h
ts
will
in
d
icate
th
e
i
n
ju
r
ies
s
ev
er
ity
lev
el
o
f
th
e
ac
cid
en
t
th
at
is
ab
o
u
t
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2252
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
1
4
,
No
.
6
,
Dec
em
b
er
2
0
2
5
:
4
4
6
1
-
4
4
7
3
4470
to
h
ap
p
en
.
T
h
e
d
esig
n
o
f
th
i
s
s
etu
p
p
r
io
r
itizes
r
ep
r
o
d
u
cib
ilit
y
,
allo
win
g
o
th
er
r
esear
ch
er
s
to
r
e
p
licate
o
u
r
f
in
d
in
g
s
an
d
e
x
ten
d
th
is
wo
r
k
.
T
h
e
p
r
o
p
o
s
ed
m
o
d
el
is
s
u
m
m
ar
ized
in
th
e
f
lo
wc
h
ar
t
s
h
o
wn
in
Fig
u
r
e
7
.
I
t
im
p
lem
en
ts
a
SVM
to
p
r
ed
ict
ac
cid
en
t sev
er
ity
an
d
o
u
tlin
es
th
e
d
ec
is
io
n
-
m
ak
i
n
g
p
r
o
ce
s
s
at
ea
ch
s
tag
e.
Fig
u
r
e
7
.
Flo
wch
ar
t
o
f
th
e
p
r
o
p
o
s
ed
SVM
-
b
ased
alg
o
r
ith
m
f
o
r
ac
cid
en
t sev
er
ity
p
r
e
d
ictio
n
An
o
th
er
ad
v
ice
to
th
e
tr
a
f
f
i
c
au
th
o
r
ities
,
th
er
e
s
h
o
u
ld
b
e
also
an
ad
ju
s
tm
en
t
o
f
t
h
e
d
r
iv
in
g
ed
u
ca
tio
n
al
p
r
o
g
r
am
s
a
n
d
m
a
n
ag
em
en
t
m
eth
o
d
o
lo
g
ies
in
t
h
e
f
ield
o
f
r
o
a
d
s
af
ety
.
Giv
in
g
m
o
r
e
im
p
o
r
ta
n
ce
to
th
e
h
u
m
a
n
f
ac
to
r
an
d
d
iv
id
i
n
g
it
in
to
m
o
r
e
s
p
ec
if
ic
attr
ib
u
te
s
o
r
im
p
ac
t
f
ac
t
o
r
s
will
d
ef
in
it
ely
h
elp
th
e
tr
af
f
ic
au
th
o
r
ities
to
s
tim
u
late
ed
u
ca
tio
n
al
p
r
o
g
r
am
s
esp
ec
ially
in
th
e
b
eg
in
n
er
d
r
iv
e
r
ed
u
ca
tio
n
.
As
an
ex
ten
s
io
n
to
th
is
r
esear
ch
,
th
e
s
ev
er
ity
o
f
th
e
ac
cid
en
ts
ca
n
b
e
m
ea
s
u
r
ed
b
y
d
r
iv
er
’
s
ag
e
an
d
th
e
tim
e
o
f
d
ay
(
d
a
y
tim
e
o
r
n
ig
h
ttime
)
t
o
h
ig
h
lig
h
t
ex
ac
tly
th
e
p
er
io
d
o
f
t
h
e
d
ay
w
h
en
th
e
y
o
u
n
g
an
d
m
i
d
d
le
-
ag
e
d
d
r
iv
er
s
ar
e
m
o
r
e
lik
ely
to
s
u
s
tain
f
atal
o
r
s
ev
er
e
in
ju
r
y
ac
cid
en
ts
.
T
h
e
p
o
p
u
latio
n
g
r
o
wth
in
Mo
r
o
cc
o
ca
n
also
b
e
an
im
p
o
r
tan
t
f
ac
t
o
r
to
co
n
s
id
er
in
d
ata
an
aly
tics
as m
en
tio
n
ed
[
1
2
]
.
Desp
ite
its
p
o
ten
tial,
th
e
ap
p
licatio
n
o
f
m
ac
h
in
e
lea
r
n
in
g
f
o
r
p
r
ed
ictin
g
r
o
ad
ac
cid
en
ts
f
atality
o
r
in
ju
r
ies
s
ev
er
ity
r
em
ain
s
an
a
r
ea
with
co
n
s
id
er
ab
le
s
co
p
e
f
o
r
f
u
r
th
e
r
d
ev
elo
p
m
en
t.
Sin
ce
th
e
d
ata
v
o
lu
m
es
in
cr
ea
s
e
an
d
co
m
p
u
tin
g
ca
p
ac
ity
b
ec
o
m
es
m
o
r
e
p
o
wer
f
u
l
an
d
af
f
o
r
d
ab
le,
r
o
a
d
ac
cid
en
ts
s
ev
er
ity
p
r
ed
ictio
n
ca
n
b
e
p
er
f
o
r
m
ed
with
co
m
p
lex
m
ac
h
in
e
lear
n
in
g
m
o
d
els
s
u
ch
as
d
ee
p
lear
n
in
g
f
r
am
e
wo
r
k
s
(
d
ee
p
b
elief
n
etwo
r
k
o
r
r
ec
u
r
r
e
n
t
n
eu
r
al
n
e
two
r
k
)
as
u
s
ed
[
2
9
]
.
Dee
p
lear
n
in
g
s
o
f
twar
e
ca
n
an
al
y
ze
b
ig
an
d
co
m
p
le
x
d
ata
s
ets
to
p
r
ed
ict
th
e
s
ev
er
ity
i
n
r
o
ad
ac
ci
d
en
ts
f
aster
an
d
m
o
r
e
ac
cu
r
ately
th
an
h
u
m
an
s
.
T
h
is
o
n
g
o
in
g
ev
o
lu
tio
n
in
co
m
p
u
tatio
n
al
ca
p
ab
ilit
ies
u
n
d
er
s
co
r
es
th
e
im
p
o
r
tan
ce
o
f
r
ep
r
o
d
u
cib
le
m
eth
o
d
o
l
o
g
ies,
allo
win
g
o
th
er
r
esear
ch
er
s
to
r
ep
licate
an
d
f
u
r
th
er
ex
ten
d
th
ese
p
r
e
d
ictiv
e
m
o
d
els to
en
h
a
n
ce
r
o
ad
s
af
ety
o
u
tco
m
es g
lo
b
ally
.
5.
CO
NCLU
SI
O
N
T
h
is
p
ap
er
u
s
es
m
ac
h
in
e
lea
r
n
in
g
alg
o
r
ith
m
s
to
p
r
ed
ict
th
e
s
ev
er
ity
o
f
ac
ci
d
en
ts
u
s
in
g
a
r
e
al
d
ataset
o
n
ac
cid
e
n
ts
in
th
e
c
o
u
n
tr
y
o
f
Mo
r
o
cc
o
.
T
h
e
ab
ilit
y
t
o
p
r
e
d
ict
an
d
u
n
d
er
s
tan
d
cr
ash
es
h
as
th
e
p
o
ten
tial
to
r
ad
ically
tr
a
n
s
f
o
r
m
cr
ash
p
r
ev
en
tio
n
a
n
d
r
ed
u
ctio
n
e
f
f
o
r
ts
,
s
av
in
g
liv
es
an
d
r
e
d
u
cin
g
th
e
e
co
n
o
m
ic
an
d
s
o
cial
co
s
ts
ass
o
ciate
d
with
th
ese
tr
ag
ed
ies.
T
h
e
p
ap
e
r
u
s
es
a
co
m
b
in
ati
o
n
o
f
k
n
o
wled
g
e
b
etwe
en
a
r
tific
ial
in
tellig
en
ce
,
s
tatis
tics
,
an
d
g
e
o
g
r
ap
h
ic
in
f
o
r
m
atio
n
s
y
s
tem
s
.
I
t
u
s
es
r
ea
l
d
ata
p
r
o
v
id
e
d
b
y
th
e
NARS
A
f
o
r
Evaluation Warning : The document was created with Spire.PDF for Python.