I
nte
rna
t
io
na
l J
o
urna
l o
f
E
lect
rica
l a
nd
Co
m
p
ute
r
E
ng
in
ee
ring
(
I
J
E
CE
)
Vo
l.
8
,
No
.
5
,
Octo
b
e
r
2
0
1
8
,
p
p
.
4
0
3
3
~
4
0
4
1
I
SS
N:
2088
-
8708
,
DOI
: 1
0
.
1
1
5
9
1
/
i
j
ec
e
.
v
8
i
5
.
pp
4
0
3
3
-
4041
4033
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ia
e
s
co
r
e
.
co
m/
jo
u
r
n
a
ls
/in
d
ex
.
p
h
p
/
I
JE
C
E
4
Da
ta M
ining
A
p
pro
a
ch of Acc
i
de
nt
O
ccur
rences I
dentif
i
ca
tion
w
ith
Effec
tive Me
thodo
lo
g
y
and I
m
ple
m
entatio
n
M
ee
nu
G
up
t
a
1
,
Vij
ender
K
u
m
a
r
So
la
n
ki
2
,
Vij
a
y
K
u
m
a
r
Sin
g
h
3
,
Vice
nte
G
a
rc
ía
-
Día
z
4
1
,3
A
n
sa
l
Un
iv
e
rsit
y
,
Ha
r
y
a
n
a
,
In
d
i
a
2
CM
RIT
,
In
d
ia
4
De
p
a
rtme
n
t
o
f
Co
m
p
u
ter S
c
ien
c
e
,
Un
iv
e
rsity
o
f
Ov
ied
o
,
Ov
ied
o
,
S
p
a
in
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Feb
20
,
2
0
1
8
R
ev
i
s
ed
Ma
y
2
8
,
2
0
1
8
A
cc
ep
ted
J
u
l
1
0
,
2
0
1
8
Da
ta
m
in
in
g
is
u
se
d
in
v
a
rio
u
s
d
o
m
a
in
s
o
f
re
se
a
rc
h
to
id
e
n
ti
fy
a
n
e
w
c
a
u
se
f
o
r
tan
e
ff
e
c
t
in
th
e
so
c
iet
y
o
v
e
r
th
e
g
lo
b
e
.
T
h
is
a
rti
c
le
in
c
lu
d
e
s
th
e
sa
m
e
re
a
so
n
f
o
r
u
sin
g
th
e
d
a
ta
m
in
in
g
to
i
d
e
n
ti
f
y
th
e
Ac
c
id
e
n
t
Oc
c
u
rre
n
c
e
s
in
d
if
fe
re
n
t
re
g
io
n
s
a
n
d
to
id
e
n
t
i
fy
th
e
m
o
st
v
a
li
d
re
a
so
n
f
o
r
h
a
p
p
e
n
in
g
a
c
c
id
e
n
ts
o
v
e
r
th
e
g
lo
b
e
.
Da
ta
M
in
i
n
g
a
n
d
A
d
v
a
n
c
e
d
M
a
c
h
in
e
L
e
a
rn
in
g
a
lg
o
rit
h
m
s
a
re
u
se
d
in
th
is
re
se
a
rc
h
a
p
p
r
o
a
c
h
a
n
d
th
is
a
rti
c
le
d
isc
u
ss
e
s
a
b
o
u
t
h
y
p
e
rli
n
e
,
c
las
si
f
ic
a
ti
o
n
s,
p
re
-
p
r
o
c
e
ss
in
g
o
f
th
e
d
a
ta,
train
in
g
th
e
m
a
c
h
in
e
w
it
h
th
e
sa
m
p
le
d
a
tas
e
ts
w
h
ich
a
re
c
o
ll
e
c
ted
f
ro
m
d
iff
e
re
n
t
re
g
io
n
s
in
w
h
ich
w
e
h
a
v
e
stru
c
tu
ra
l
a
n
d
se
m
i
-
st
ru
c
tu
ra
l
d
a
ta.
W
e
w
il
l
d
iv
e
in
t
o
d
e
e
p
o
f
m
a
c
h
in
e
lea
rn
in
g
a
n
d
d
a
ta
m
in
in
g
c
las
si
f
ic
a
ti
o
n
a
lg
o
rit
h
m
s
to
f
in
d
o
r
p
re
d
ict
so
m
e
th
in
g
n
o
v
e
l
a
b
o
u
t
t
h
e
a
c
c
id
e
n
t
o
c
c
u
rre
n
c
e
s
o
v
e
r
th
e
g
lo
b
e
.
W
e
m
a
jo
rl
y
c
o
n
c
e
n
trate
o
n
tw
o
c
las
si
f
ica
ti
o
n
a
lg
o
rit
h
m
s
to
m
in
i
fy
th
e
r
e
se
a
r
c
h
a
n
d
tas
k
a
n
d
th
e
y
a
re
v
e
r
y
b
a
sic
a
n
d
im
p
o
rtan
t
c
las
si
f
ica
ti
o
n
a
lg
o
rit
h
m
s.
S
V
M
(S
u
p
p
o
rt
v
e
c
to
r
m
a
c
h
in
e
),
CNB
Clas
sif
ier.
T
h
is
d
isc
u
ss
io
n
w
il
l
b
e
q
u
it
e
in
tere
stin
g
w
it
h
W
EK
A
to
o
l
f
o
r
CNB
c
las
si
f
ier,
B
a
g
o
f
W
o
rd
s
Id
e
n
ti
f
ica
ti
o
n
,
W
o
rd
Co
u
n
t
a
n
d
F
re
q
u
e
n
c
y
Ca
lc
u
latio
n
.
K
ey
w
o
r
d
:
A
cc
id
e
n
ts
B
ag
o
f
w
o
r
d
s
C
las
s
i
f
ier
C
NB
Data
m
i
n
i
n
g
Fre
q
u
en
c
y
c
alcu
latio
n
H
y
p
er
lin
e
Ma
ch
i
n
e
l
ea
r
n
i
n
g
SVN
W
E
KA
W
o
r
d
c
o
u
n
t
Co
p
y
rig
h
t
©
2
0
1
8
In
stit
u
te o
f
A
d
v
a
n
c
e
d
E
n
g
i
n
e
e
rin
g
a
n
d
S
c
ien
c
e
.
Al
l
rig
h
ts
re
se
rv
e
d
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
Vice
n
te
Gar
cía
-
Díaz
,
Dep
ar
t
m
en
t o
f
C
o
m
p
u
ter
Scie
n
ce
,
Un
i
v
er
s
it
y
o
f
O
v
ied
o
,
Ov
ied
o
,
Sp
ain
.
E
m
ail:
g
ar
ciav
ice
n
te
@
u
n
io
v
i.
es
1.
I
NT
RO
D
UCT
I
O
N
Data
m
i
n
i
n
g
is
t
h
e
p
r
o
m
i
n
en
t
tech
n
o
lo
g
y
to
p
r
ed
ict
o
r
d
o
s
o
m
e
an
al
y
tic
s
o
n
a
d
o
m
a
in
.
T
r
af
f
ic
m
an
a
g
e
m
e
n
t
a
n
d
ac
cid
en
t
o
c
cu
r
r
en
ce
s
i
n
d
i
f
f
er
e
n
t
p
lace
s
o
v
er
th
e
g
lo
b
e
an
d
t
h
e
r
ea
s
o
n
f
o
r
ac
cid
e
n
t
s
m
a
y
v
ar
y
.
B
u
t
w
e
n
ee
d
to
lo
o
k
af
ter
s
o
m
e
o
f
t
h
e
t
h
i
n
g
s
w
h
ic
h
a
r
e
r
elate
d
to
t
h
e
m
i
n
i
n
g
t
h
e
m
o
s
t
c
h
an
ce
s
o
f
ac
cid
en
t
o
cc
u
r
r
e
n
ce
s
.
L
et
’
s
ta
k
e
a
s
u
r
v
e
y
o
n
d
if
f
er
e
n
t
m
ac
h
in
e
lear
n
in
g
cla
s
s
i
f
ica
tio
n
a
l
g
o
r
ith
m
s
w
h
ich
ar
e
u
s
ed
o
n
d
i
f
f
er
e
n
t d
ata
s
e
ts
co
ll
ec
ted
f
r
o
m
d
if
f
er
e
n
t r
eg
io
n
an
d
w
e
ca
n
m
a
k
e
a
d
ec
is
io
n
o
n
w
h
ic
h
clas
s
i
f
icatio
n
r
u
le
o
r
as
s
o
ciatio
n
r
u
le
h
av
e
to
u
s
e
f
o
r
o
u
r
d
ata
s
e
t.
W
e
h
av
e
a
p
u
b
licl
y
a
v
ailab
le
d
at
a
s
et
o
n
w
h
ic
h
w
e
i
m
p
le
m
en
ted
SVM
clas
s
i
f
ier
an
d
C
NB
class
i
f
ier
w
ith
W
E
KA
to
o
l.
T
h
e
r
eq
u
ir
ed
r
esu
lt
is
to
id
en
tify
w
h
ic
h
class
i
f
icatio
n
alg
o
r
it
h
m
i
s
b
etter
f
o
r
t
h
e
m
in
i
n
g
th
e
ac
t
u
al
d
ata
an
d
p
r
e
d
ict
b
etter
w
it
h
t
h
e
r
esu
lt
s
.
T
h
e
m
ai
n
m
o
tto
b
eh
i
n
d
th
is
k
i
n
d
o
f
ar
tic
le
is
b
ec
au
s
e
o
f
m
o
r
e
ca
s
e
s
b
ein
g
r
ec
o
r
d
ed
b
y
t
h
e
r
eg
io
n
al
h
o
s
p
itals
as a
cc
id
e
n
t
ca
s
es.
T
h
e
in
j
u
r
ies,
d
a
m
ag
e
s
f
o
r
v
eh
icle
s
an
d
s
o
o
n
ca
n
b
e
c
o
n
s
id
er
ed
as
th
e
m
ai
n
r
ea
s
o
n
s
.
T
h
e
m
a
in
r
ea
s
o
n
s
f
o
r
th
e
d
ea
th
s
o
n
r
o
ad
is
tr
af
f
ic
ac
cid
en
ts
[
1
]
,
th
at
is
n
o
t
f
o
llo
w
i
n
g
th
e
tr
af
f
ic
r
u
le
s
,
o
v
e
r
tak
in
g
in
a
w
r
o
n
g
w
a
y
,
o
v
er
s
p
ee
d
,
n
o
t
f
o
llo
w
in
g
s
a
f
et
y
m
ea
s
u
r
es
o
f
r
o
ad
.
A
s
p
er
W
HO
(
W
o
r
ld
Hea
lth
Or
g
an
iza
tio
n
)
o
v
er
4
m
ill
io
n
ca
s
e
s
h
a
v
e
b
ee
n
r
ec
o
r
d
ed
ea
ch
y
ea
r
w
o
r
ld
w
id
e
b
ec
au
s
e
o
f
t
h
e
tr
a
f
f
ic
a
n
d
r
o
ad
ac
cid
en
ts
.
T
h
e
m
ai
n
r
ea
s
o
n
s
w
h
ic
h
W
HO
s
ta
tes
is
n
o
t
f
o
llo
w
i
n
g
tr
af
f
ic
r
u
le
s
,
n
o
t
f
o
llo
w
i
n
g
s
a
f
et
y
m
ea
s
u
r
es
lik
e
s
ea
t
b
elt,
h
el
l
m
ate,
o
v
er
s
p
ee
d
,
w
r
o
n
g
cr
o
s
s
in
g
,
m
in
o
r
d
r
iv
i
n
g
,
la
k
e
o
f
li
ter
ac
y
o
n
th
e
tr
a
f
f
ic
an
d
r
o
ad
s
af
et
y
r
u
le
s
,
d
r
u
n
k
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
8
,
No
.
5
,
Octo
b
er
2
0
1
8
:
4
0
3
3
–
4
0
4
1
4034
an
d
d
r
iv
e.
W
e
ca
n
p
r
o
v
id
e
th
e
m
ea
s
u
r
es to
av
o
id
th
i
s
k
in
d
o
f
th
i
n
g
s
w
it
h
s
m
all
m
ea
s
u
r
es
w
h
ic
h
ar
e
d
is
cu
s
s
ed
b
y
o
t
h
er
r
esear
ch
er
s
[
2
]
.
Data
m
i
n
in
g
i
s
m
ai
n
l
y
u
s
ed
to
id
en
t
if
y
th
e
s
e
v
er
it
y
o
f
ac
c
id
en
ts
o
n
r
o
ad
s
[
3
]
.
DM
DW
(
Data
Min
i
n
g
Data
W
ar
eh
o
u
s
i
n
g
)
[
4
]
h
a
v
e
all
t
h
e
tec
h
n
iq
u
es
to
b
e
u
s
ed
to
p
r
ed
ict
o
r
id
en
ti
f
y
t
h
e
s
e
v
er
it
y
o
f
ac
cid
e
n
ts
o
n
r
o
ad
s
.
DM
is
u
s
ed
to
ex
tr
ac
t th
e
s
e
m
an
tic
th
in
g
s
o
v
er
th
e
d
ata
s
et
t
h
at
is
a
m
ea
n
in
g
f
u
l
ex
tr
ac
t
f
r
o
m
t
h
e
d
at
a
av
ailab
le
[
5
]
.
T
h
e
cla
s
s
i
f
icatio
n
tech
n
iq
u
e
s
lik
e
cl
u
s
ter
i
n
g
,
a
n
o
m
al
y
d
etec
tio
n
,
clu
s
ter
in
g
a
n
d
class
i
f
icatio
n
r
u
le
s
[
6
]
ar
e
u
s
ed
f
o
r
m
o
s
t o
f
th
e
DM
o
p
er
atio
n
s
o
n
th
e
r
o
ad
ac
cid
en
ts
.
I
n
th
i
s
ar
ticle
w
e
w
o
u
ld
li
k
e
to
s
h
ar
e
s
o
m
e
liter
at
u
r
e
s
u
r
v
e
y
o
n
d
if
f
er
en
t p
r
ev
io
u
s
o
p
er
atio
n
s
d
o
n
e
o
n
d
if
f
er
e
n
t
d
ata
s
ets
an
d
al
s
o
th
e
c
u
r
r
en
t
r
esear
ch
w
e
w
o
u
ld
to
d
o
o
n
t
h
e
d
if
f
er
en
t
d
ata
s
et
r
elate
d
to
th
e
r
o
ad
ac
cid
en
ts
an
d
s
ev
er
it
y
.
T
h
e
n
ex
t
s
ec
t
io
n
w
ill
d
is
c
u
s
s
s
h
o
r
t
liter
at
u
r
e
s
u
r
v
e
y
,
later
cu
r
r
en
t
w
o
r
k
wh
at
th
is
ar
tic
le
w
il
l
s
p
ea
k
,
ex
p
er
i
m
e
n
tal
r
es
u
lt
s
,
r
eso
u
r
ce
s
an
d
f
i
n
all
y
co
n
clu
d
e.
2.
L
I
T
E
R
AT
U
RE
SU
RVE
Y
As
w
e
n
ee
d
to
co
n
s
id
er
b
asic
s
o
f
S
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
e
s
an
d
C
NB
clas
s
i
f
ier
s
to
u
n
d
e
r
s
tan
d
t
h
e
liter
atu
r
e
r
ev
ie
w
,
let
’
s
m
a
k
e
a
s
a
m
p
le
co
llec
tio
n
o
f
k
n
o
w
l
ed
g
e
o
n
S
VM
as
it
i
s
i
m
p
o
r
tan
t
i
n
t
h
is
r
esear
c
h
s
co
p
e.
I
n
m
ac
h
i
n
e
lear
n
in
g
,
S
VM
s
ar
e
co
n
tr
o
lled
lear
n
in
g
m
o
d
el
s
w
it
h
r
elate
d
lear
n
i
n
g
co
u
n
t
s
th
at
s
ep
ar
ate
d
ata
u
s
ed
f
o
r
co
u
r
s
e
o
f
ac
tio
n
an
d
b
ac
k
s
lid
e
ex
a
m
i
n
atio
n
.
Giv
e
n
a
co
u
r
s
e
o
f
ac
tio
n
o
f
p
r
ep
ar
in
g
ca
s
e
s
,
ea
ch
s
et
ap
a
r
t
as
h
a
v
in
g
a
p
lace
w
it
h
b
o
th
o
f
t
w
o
g
r
o
u
p
i
n
g
s
,
a
SVM
ar
r
an
g
i
n
g
ch
ec
k
s
et
ti
n
g
u
p
a
f
o
r
m
a
t)
.
A
Su
p
p
o
r
t
Vec
to
r
Ma
ch
in
e
p
o
in
ts
a
d
elin
ea
tio
n
o
f
t
h
e
m
eth
o
d
as
in
d
icate
s
in
a
p
lo
t,
p
o
in
ted
o
r
co
n
n
ec
ted
w
i
th
th
e
tar
g
et
t
h
at
t
h
e
e
x
a
m
p
le
s
o
f
th
e
i
n
s
tan
ce
o
f
cla
s
s
e
s
ar
e
d
i
s
en
g
ag
ed
b
y
a
s
e
n
s
ib
le
m
a
n
n
e
r
th
at
i
s
as
w
id
e
as
it
co
u
ld
b
e
s
e
n
s
ib
le.
Ne
w
i
n
s
ta
n
ce
s
ar
e
th
e
n
i
n
d
en
t
if
ied
an
d
co
n
n
ec
ted
in
to
th
at
s
a
m
e
h
y
p
o
th
e
s
is
an
d
an
ticip
ated
to
h
a
v
e
a
p
lace
w
i
th
a
cla
s
s
i
n
co
n
te
x
t
o
f
w
h
ic
h
s
id
e
o
f
t
h
e
i
n
s
ta
n
ce
t
h
e
y
f
al
l.
No
t
w
ith
s
ta
n
d
in
g
p
lay
in
g
o
u
t
th
e
p
r
o
m
p
t d
e
m
an
d
,
Su
p
p
o
r
t V
ec
to
r
Ma
ch
in
es c
an
b
en
e
f
iciall
y
ac
t b
e
y
o
n
d
th
e
b
o
u
n
d
ar
y
as a
n
o
n
-
s
tr
aig
h
t
d
ep
ictio
n
u
s
in
g
th
e
t
h
in
g
w
h
at
is
ac
t
u
all
y
id
e
n
ti
f
ied
as
th
e
p
ar
t
-
tr
ap
,
ch
ec
k
in
g
an
d
co
n
n
ec
tin
g
th
eir
d
u
ties
r
e
g
ar
d
in
g
h
ig
h
-
i
n
s
ta
n
c
e
p
o
r
tio
n
s
p
ac
es.
R
i
g
h
t
w
h
e
n
th
e
d
ata
i
s
n
't
s
ta
m
p
ed
,
s
tr
ai
g
h
t
f
o
r
w
ar
d
th
in
g
s
r
elate
d
to
lear
n
in
g
is
n
'
t
ac
ce
p
tab
le,
an
d
an
u
n
-
s
u
p
er
v
is
ed
lear
n
in
g
m
et
h
o
d
o
lo
g
y
is
m
an
d
ato
r
y
,
w
h
ic
h
i
s
lead
in
g
to
id
en
t
if
y
tr
ad
e
m
ar
k
g
ath
er
i
n
g
o
f
t
h
e
i
n
f
o
r
m
atio
n
t
o
g
et
-
to
g
et
h
er
s
,
a
n
d
a
f
ter
t
h
at
g
u
id
e
r
ele
v
a
n
t
d
ata
to
th
ese
s
u
r
r
o
u
n
d
ed
s
o
cial
g
r
o
u
p
s
.
T
h
e
g
r
o
u
p
in
g
id
e
n
ti
f
ies
w
h
ic
h
lead
s
to
a
c
h
a
n
ce
o
f
m
o
d
i
f
icat
io
n
to
t
h
e
SVM’
s
is
ca
lled
s
u
p
p
o
r
t
v
ec
to
r
ass
e
m
b
li
n
g
a
n
d
it
is
o
n
ce
in
a
w
h
ile
u
s
ed
as
a
b
it
o
f
m
ec
h
a
n
ical
m
e
th
o
d
o
lo
g
y
e
ith
er
w
h
e
n
t
h
e
d
ata
is
n
'
t
c
h
ec
k
ed
o
r
w
h
e
n
j
u
s
t
t
w
o
o
r
th
r
ee
d
ata
ar
e
n
a
m
ed
as
a
p
r
e
-
p
r
o
ce
s
s
i
n
g
f
o
r
a
d
ep
ictio
n
m
et
h
o
d
.
Ask
i
n
g
f
o
r
d
ata
is
a
g
en
er
al
u
n
d
er
ta
k
in
g
i
n
ML
.
E
x
p
ec
t
s
o
m
e
s
h
o
w
n
d
ata
s
h
o
w
s
ev
er
y
p
o
in
t
as
a
p
lace
eith
er
o
f
th
e
a
v
ailab
le
cl
ass
es
a
n
d
th
e
p
u
r
p
o
s
e
is
to
p
ic
k
ex
ac
t
cla
s
s
alter
n
ati
v
e
Data
p
o
in
t
w
ill
b
e
u
s
i
n
g
.
B
y
id
ea
ls
o
f
SVM
’
s
,
a
d
ata
p
o
in
t
is
id
en
ti
f
ied
as
a
p
d
i
m
e
n
s
io
n
al
v
ec
to
r
(
a
q
u
ick
o
v
er
v
iew
o
f
p
id
en
ti
f
ier
s
)
,
an
d
th
e
th
i
n
g
w
e
h
av
e
to
id
en
ti
f
y
i
s
th
at
p
o
s
s
ib
le
th
at
we
ca
n
is
o
late
s
u
c
h
p
o
in
ter
s
w
it
h
a
(
p
-
1)
-
m
u
lti
-
d
i
m
en
s
io
n
al
h
y
p
er
p
la
n
e.
T
h
is
ca
n
b
e
id
en
t
if
ied
as
d
ir
ec
ted
class
i
f
ier
.
T
h
er
e
ar
e
d
i
f
f
er
e
n
t
h
y
p
er
li
n
es
th
at
m
a
y
to
tal
d
ata
r
eg
ar
d
in
g
th
e
p
o
in
ts
.
T
h
e
o
n
e
s
en
s
it
iv
e
o
p
in
io
n
as
th
e
b
etter
h
y
p
er
-
p
la
n
e
is
th
e
o
n
e
th
at
te
n
d
s
to
t
h
e
b
est
p
ar
titi
o
n
,
o
r
p
o
in
t,
b
et
w
e
en
t
h
e
d
if
f
er
e
n
t
clas
s
es.
So
we
s
elec
t
t
h
e
h
y
p
er
-
li
n
e
s
o
t
h
e
is
o
latio
n
f
r
o
m
it
to
th
e
clo
s
est
d
ata
-
p
o
in
t
o
n
o
th
e
r
s
id
e
is
i
m
p
r
o
v
ed
.
I
n
s
u
c
h
d
ata
-
p
o
in
t
t
h
at
h
y
p
er
-
li
n
e
id
en
t
i
f
ies,
it
is
k
n
o
w
n
as
th
e
b
est
f
it
ted
h
y
p
er
-
li
n
e
a
n
d
th
e
q
u
ic
k
id
e
n
ti
f
ier
it
p
o
r
tr
ay
s
i
s
m
en
tio
n
ed
as
a
m
o
s
t
o
v
er
th
e
to
p
d
ata
class
i
f
ier
; o
r
p
r
o
p
o
r
tio
n
atel
y
,
t
h
e
p
er
ce
p
tr
o
n
o
f
f
la
w
le
s
s
s
ec
u
r
it
y
A
ll
th
e
m
o
r
e
g
e
n
er
all
y
,
a
SV
M
d
ev
elo
p
s
a
h
y
p
er
-
li
n
e
o
r
s
e
t
o
f
h
y
p
er
-
li
n
es
in
a
h
i
g
h
-
o
r
tr
e
m
en
d
o
u
s
d
i
m
en
s
io
n
al
p
lan
e,
w
h
ic
h
w
a
s
u
s
ed
f
o
r
d
ep
ictio
n
,
f
al
l
a
wa
y
f
r
o
m
t
h
e
f
ait
h
,
o
r
v
ar
io
u
s
u
n
d
er
tak
i
n
g
s
li
k
e
ir
r
eg
u
lar
itie
s
af
f
ir
m
atio
n
.
R
e
g
u
lar
l
y
,
a
m
in
d
b
lo
w
i
n
g
p
ac
k
a
g
e
is
r
ef
i
n
ed
b
y
th
e
h
y
p
er
-
l
in
e
th
at
h
as
t
h
e
b
est
d
iv
is
io
n
to
th
e
clo
s
est
p
r
ep
ar
in
g
i
n
f
o
r
m
atio
n
p
u
r
p
o
s
e
b
eh
in
d
an
y
clas
s
(
attested
ac
co
m
m
o
d
atin
g
ed
g
e)
,
s
i
n
ce
all
ar
o
u
n
d
th
e
m
o
r
e
p
r
o
m
i
n
e
n
t
th
e
ed
g
e
th
e
lo
w
er
t
h
e
h
y
p
o
th
esis
s
p
o
il o
f
t
h
e
clas
s
i
f
ier
Fig
u
r
e
1
.
Su
p
p
o
r
t V
ec
to
r
Ma
c
h
in
e
Sa
m
p
le
p
lo
tti
n
g
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8708
Da
ta
Min
in
g
A
p
p
r
o
a
ch
o
f A
cc
id
en
t O
cc
u
r
r
en
ce
s
I
d
en
tifi
ca
tio
n
w
ith
E
ffectiv
e
.
..
(
Meen
u
Gu
p
ta
)
4035
T
h
e
Fig
u
r
e
1
is
a
m
o
d
el
o
cc
u
r
r
en
ce
o
f
a
SVM
class
if
ier
,
i.e
.
,
a
SVM
class
if
ier
th
at
li
m
it
s
a
s
tr
ateg
y
o
f
t
h
i
n
g
s
i
n
to
t
h
eir
d
i
v
er
s
e
s
o
cial
e
v
e
n
ts
(
GR
E
E
N,
R
E
D)
in
d
icate
s
a
h
y
p
er
-
li
n
e.
Mo
s
t
a
s
s
e
m
b
li
n
g
u
n
d
er
ta
k
in
g
s
,
r
e
g
ar
d
less
,
ar
e
n
o
t
th
at
cr
u
cia
l,
an
d
r
eliab
l
y
m
o
r
e
f
an
ci
f
u
l
s
tr
u
ct
u
r
e
is
r
eq
u
i
r
ed
r
ec
o
llectin
g
t
h
e
g
en
u
i
n
e
g
o
al
to
m
ak
e
a
f
la
w
le
s
s
s
ep
ar
atio
n
,
i.e
.
,
d
ec
is
iv
e
l
y
p
o
r
tr
ay
n
e
w
d
i
f
f
icu
lties
(
te
s
t
i
n
s
tan
ce
s
)
in
li
g
h
t
o
f
th
e
d
ep
ictio
n
s
t
h
at
ar
e
o
p
er
ate
s
(
p
r
ep
ar
e
in
s
tan
ce
s
)
.
T
h
is
s
itu
atio
n
is
d
ep
icted
in
th
e
s
tr
u
c
tu
r
e
b
elo
w
.
E
m
er
g
ed
f
r
o
m
t
h
e
p
r
ev
io
u
s
s
e
m
a
n
tic,
u
n
m
i
s
ta
k
ab
l
y
a
co
m
p
le
te
s
ec
t
io
n
o
f
t
h
e
co
lo
r
s
Gr
ee
n
a
n
d
also
R
ed
in
d
icat
io
n
co
u
ld
b
e
r
eq
u
ir
e
a
w
i
n
d
(
w
h
i
ch
i
s
m
o
r
e
p
u
zz
l
in
g
th
a
n
a
h
y
p
er
-
li
n
e)
.
T
h
e
C
o
u
r
s
e
o
f
ac
ti
v
it
y
a
s
s
i
g
n
m
e
n
t
s
i
n
lig
h
t
o
f
attr
ac
t
in
g
h
y
p
er
li
n
e
s
to
s
ee
m
et
h
o
d
s
o
f
d
if
f
er
e
n
t
o
b
j
ec
ts
p
ar
ticip
atin
g
ar
e
d
e
f
i
n
ed
as
h
y
p
er
-
li
n
e
class
i
f
ier
s
as
s
h
o
w
n
i
n
Fi
g
u
r
e
2
.
Su
p
p
o
r
t
Vec
to
r
Ma
ch
in
es
ar
e
esp
ec
iall
y
s
u
ited
to
o
v
er
s
ee
th
at
k
i
n
d
o
f
er
r
an
d
s
.
T
h
e
Fig
u
r
e
3
b
elo
w
d
i
s
p
la
y
s
th
e
cr
itica
l
t
h
o
u
g
h
t
b
eh
i
n
d
SVM’
s
.
Her
e
w
e
ca
n
o
b
s
er
v
e
th
e
b
asi
c
d
if
f
er
e
n
ce
s
(
r
ed
p
ar
t
o
f
th
e
s
em
an
tic)
co
n
n
ec
ted
,
i.e
.
,
ad
j
u
s
ted
,
u
s
in
g
a
g
a
m
e
-
p
la
n
o
f
s
en
s
ib
le
cu
to
f
f
p
o
in
t
s
s
p
ec
if
ied
a
s
s
ec
tio
n
s
.
T
h
e
p
r
o
ce
s
s
o
f
m
o
d
if
y
i
n
g
t
h
e
ar
ticle
s
is
d
ef
in
ed
a
s
co
n
n
ec
t
in
g
.
Ma
k
e
a
n
o
te
t
h
at
i
n
th
i
s
n
e
w
o
p
er
ati
o
n
s
,
th
e
m
ap
p
ed
o
b
j
ec
ts
(
Gr
ee
n
p
ar
t
o
f
th
e
s
em
an
tic)
is
s
tr
ai
g
h
tl
y
u
n
m
is
ta
k
ab
le
an
d
,
in
li
k
e
m
an
n
er
,
in
s
tead
o
f
b
u
ild
in
g
th
e
co
n
f
u
s
in
g
tu
r
n
(
lef
t
s
e
m
a
n
tic)
,
w
e
s
h
o
u
ld
j
u
s
t
to
lo
ca
te
an
i
m
p
ec
ca
b
le
lin
e
th
at
ca
n
d
i
s
en
g
a
g
e
th
e
Gr
ee
n
an
d
also
th
e
R
ed
t
h
i
n
g
s
.
Fig
u
r
e
2
.
Dif
f
er
en
tiat
io
n
b
et
wee
n
p
lo
ts
Fig
u
r
e
3
.
I
n
p
u
t a
n
d
o
u
tp
u
t sp
a
ce
d
if
f
er
e
n
tiatio
n
SVM
is
p
er
h
ap
s
a
ch
a
m
p
io
n
a
m
o
n
g
t
h
e
m
o
s
t
w
ell
k
n
o
w
n
a
n
d
talk
ed
ab
o
u
t
m
ac
h
i
n
e
lear
n
in
g
esti
m
atio
n
s
.
T
h
e
y
w
er
e
i
n
cr
e
d
ib
ly
s
ta
n
d
ar
d
ar
o
u
n
d
t
h
e
ti
m
e
th
e
y
w
er
e
d
eli
v
er
ed
in
th
e
1
9
9
0
s
an
d
co
n
tin
u
e
b
ein
g
t
h
e
g
o
-
to
s
y
s
te
m
f
o
r
a
h
ig
h
-
p
er
f
o
r
m
i
n
g
co
u
n
t
w
it
h
a
litt
le
tu
n
i
n
g
.
I
n
th
i
s
p
o
s
t,
y
o
u
w
ill
d
is
co
v
er
th
e
SVM
m
ac
h
i
n
e
lear
n
i
n
g
f
i
g
u
r
i
n
g
.
I
n
t
h
e
w
a
k
e
o
f
e
x
a
m
i
n
i
n
g
t
h
is
p
o
s
t
y
o
u
w
ill
k
n
o
w
:
W
ell
o
r
d
er
ed
g
u
id
elin
es
to
d
i
s
en
ta
n
g
le
t
h
e
v
ar
io
u
s
n
a
m
es
u
s
ed
to
i
n
s
i
n
u
ate
h
elp
v
ec
to
r
m
ac
h
in
e
s
.
T
h
e
d
ep
ictio
n
u
s
ed
b
y
SV
M
w
h
e
n
t
h
e
m
o
d
el
i
s
tr
u
l
y
s
ec
u
r
ed
to
th
e
p
late.
Ho
w
an
in
f
o
r
m
ed
SVM
d
em
o
n
s
tr
ate
d
ep
ictio
n
ca
n
b
e
u
s
ed
to
m
ak
e
d
e
s
ir
es
f
o
r
n
e
w
d
ata.
W
ell
o
r
d
er
e
d
d
i
r
ec
tio
n
s
to
tak
e
i
n
an
SV
M
s
h
o
w
f
r
o
m
g
ett
in
g
r
ea
d
y
d
ata.
Gu
id
elin
e
s
to
b
est
s
et
u
p
y
o
u
r
d
ata
f
o
r
th
e
SVM
esti
m
atio
n
.
W
h
er
e
y
o
u
m
a
y
lik
e
to
g
et
m
o
r
e
i
n
f
o
r
m
atio
n
o
n
SVM.
SVM
i
s
a
s
ti
m
u
lati
n
g
esti
m
atio
n
a
n
d
th
e
th
o
u
g
h
ts
ar
e
b
y
a
n
d
lar
g
e
d
ir
ec
t.
T
h
is
p
o
s
t
w
as
cr
ea
ted
f
o
r
ar
ch
itects
w
it
h
b
asicall
y
n
o
estab
lis
h
m
e
n
t
in
est
i
m
a
tio
n
s
an
d
a
s
tr
aig
h
t
f
ac
to
r
b
ased
m
a
th
.
T
h
e
Ma
x
i
m
al
-
Ma
r
g
in
C
las
s
if
ier
is
a
th
eo
r
etica
l
class
i
f
ier
th
at
b
est
clea
r
s
u
p
h
o
w
S
VM
w
o
r
k
s
ev
en
t
u
all
y
.
T
h
e
n
u
m
er
ic
d
ata
f
ac
to
r
s
(
x
)
in
y
o
u
r
d
ata
(
th
e
s
ec
tio
n
s
)
o
u
t
lin
e
a
n
n
-
d
i
m
e
n
s
io
n
al
s
p
ac
e.
Fo
r
ex
a
m
p
le,
if
y
o
u
h
ad
t
w
o
in
f
o
r
m
atio
n
f
ac
to
r
s
,
t
h
i
s
w
o
u
ld
s
h
a
p
e
a
t
w
o
-
d
i
m
en
s
io
n
al
s
p
ac
e.
A
h
y
p
er
p
lan
e
i
s
a
lin
e
t
h
at
p
ar
ts
t
h
e
d
ata
v
ar
iab
le
s
p
ac
e.
I
n
SV
M,
a
h
y
p
er
p
lan
e
is
b
ested
is
o
late
th
e
c
o
n
ce
n
tr
atio
n
s
i
n
t
h
e
in
f
o
r
m
atio
n
v
ar
iab
le
s
p
ac
e
b
y
th
eir
class
,
eith
er
clas
s
0
o
r
c
lass
1
.
I
n
t
w
o
-
e
s
ti
m
atio
n
s
,
y
o
u
ca
n
p
ictu
r
e
th
i
s
as
a
lin
e
a
n
d
w
e
o
u
g
h
t
to
e
x
p
ec
t
th
at
t
h
e
lar
g
er
p
ar
t
o
f
o
u
r
d
ata
ce
n
ter
s
ca
n
b
e
co
m
p
letel
y
s
eg
r
eg
a
ted
b
y
th
is
lin
e.
Fo
r
ex
a
m
p
le:
B
0
+
(
B
1
*
X1
)
+
(
B
2
*
X2
)
=
0
W
h
er
e
th
e
co
ef
f
icie
n
ts
(
B
1
an
d
B
2
)
th
at
c
h
o
o
s
e
t
h
e
i
n
cli
n
atio
n
o
f
t
h
e
li
n
e
a
n
d
t
h
e
ca
tc
h
(
B
0
)
ar
e
f
o
u
n
d
b
y
th
e
lear
n
i
n
g
co
m
p
u
t
atio
n
,
an
d
X1
an
d
X2
ar
e
th
e
t
w
o
d
ata
f
ac
to
r
s
.
Yo
u
ca
n
ta
k
e
co
u
r
s
es
o
f
ac
tio
n
u
s
i
n
g
t
h
is
li
n
e.
B
y
in
ter
f
ac
i
n
g
w
it
h
e
n
ter
in
g
r
e
g
ar
d
s
i
n
to
t
h
e
lin
e
co
n
d
itio
n
,
y
o
u
ca
n
p
r
o
ce
s
s
w
h
et
h
er
a
n
o
th
er
p
o
in
t
is
ab
o
v
e
o
r
u
n
d
er
n
ea
th
t
h
e
lin
e.
O
v
er
th
e
lin
e,
t
h
e
co
n
d
itio
n
r
ee
s
tab
lis
h
es
a
r
eg
ar
d
m
o
r
e
n
o
ticea
b
le
th
an
0
an
d
th
e
p
o
in
t
h
as
a
p
lace
w
it
h
t
h
e
f
iv
e
s
tar
(
class
0
)
.
Un
d
er
n
ea
t
h
t
h
e
lin
e,
t
h
e
co
n
d
itio
n
r
ee
s
tab
lis
h
es
a
r
eg
ar
d
u
n
d
er
0
an
d
th
e
p
o
in
t
h
as
a
p
lace
w
it
h
t
h
e
b
en
ea
t
h
n
o
r
m
al
(
class
1
)
.
A
r
eg
ar
d
clo
s
e
to
th
e
li
n
e
r
ee
s
tab
lis
h
e
s
a
r
eg
ar
d
al
m
o
s
t
ze
r
o
an
d
th
e
p
o
in
t
m
a
y
b
e
d
if
f
ic
u
lt
to
m
as
ter
m
i
n
d
.
I
f
th
e
s
p
an
o
f
th
e
r
eg
ar
d
is
g
en
er
o
u
s
,
th
e
m
o
d
el
m
a
y
h
a
v
e
m
o
r
e
tr
u
s
t
in
t
h
e
d
esire
.
T
h
e
d
iv
is
io
n
b
et
w
ee
n
t
h
e
lin
e
a
n
d
th
e
clo
s
est
d
at
a
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
8
,
No
.
5
,
Octo
b
er
2
0
1
8
:
4
0
3
3
–
4
0
4
1
4036
s
h
o
w
s
is
i
m
p
lied
a
s
t
h
e
ed
g
e.
T
h
e
b
est
o
r
p
er
f
ec
t
lin
e
th
at
c
an
s
ep
ar
ate
t
h
e
t
w
o
cla
s
s
e
s
is
th
e
li
n
e
t
h
at
a
s
t
h
e
g
r
ea
test
ed
g
e.
T
h
is
is
k
n
o
wn
as
th
e
Ma
x
i
m
al
-
Ma
r
g
i
n
h
y
p
er
p
la
n
e.
T
h
e
ed
g
e
is
f
ig
u
r
ed
as
th
e
co
n
tr
ar
y
d
etac
h
m
e
n
t
f
r
o
m
t
h
e
li
n
e
to
o
n
l
y
th
e
clo
s
e
s
t
co
r
e
in
ter
e
s
t
s
.
Si
m
p
l
y
t
h
ese
co
n
ce
n
tr
atio
n
s
ar
e
p
er
ti
n
e
n
t
i
n
p
o
r
tr
ay
in
g
t
h
e
lin
e
an
d
i
n
th
e
i
m
p
r
o
v
e
m
e
n
t
o
f
t
h
e
class
i
f
ier
.
T
h
ese
co
n
ce
n
tr
atio
n
s
ar
e
k
n
o
w
n
a
s
th
e
as
s
is
tan
ce
v
ec
to
r
s
.
T
h
e
y
s
u
p
p
o
r
t
o
r
d
escr
ib
e
th
e
h
y
p
er
p
lan
e.
T
h
e
h
y
p
er
p
lan
e
is
p
ic
k
ed
u
p
f
r
o
m
p
lan
n
i
n
g
d
ata
u
s
i
n
g
a
s
tr
ea
m
li
n
i
n
g
f
r
a
m
e
w
o
r
k
t
h
at
li
f
ts
t
h
e
ed
g
e.
W
h
en
all
i
s
s
aid
i
n
d
o
n
e,
au
t
h
en
t
ic
d
ata
is
d
is
o
r
g
a
n
ized
an
d
ca
n
'
t
b
e
s
ep
ar
ated
i
m
p
ec
ca
b
ly
w
it
h
a
h
y
p
er
p
lan
e.
T
h
e
b
asic
o
f
g
r
o
w
i
n
g
th
e
ed
g
e
o
f
th
e
li
n
e
th
at
s
eg
r
eg
ate
s
th
e
class
e
s
m
u
s
t
b
e
ea
s
y
g
o
i
n
g
.
T
h
is
is
r
o
u
tin
el
y
ca
l
led
th
e
f
r
ag
ile
ed
g
e
cla
s
s
i
f
ier
.
T
h
is
ch
a
n
g
e
allo
w
s
a
co
u
p
le
o
f
d
e
m
o
n
s
tr
ates
i
n
t
h
e
ar
r
an
g
e
m
e
n
t
d
ata
m
an
h
a
n
d
le
th
e
s
ec
l
u
d
in
g
lin
e.
An
ad
d
itio
n
al
g
a
m
e
p
la
n
o
f
co
e
f
f
icien
ts
ar
e
e
x
h
ib
ited
th
at
g
i
v
e
t
h
e
ed
g
e
s
q
u
ir
m
r
o
o
m
i
n
e
v
er
y
est
i
m
atio
n
.
T
h
ese
co
e
f
f
icien
ts
ar
e
r
ar
el
y
ca
lled
s
lac
k
v
ar
iab
les.
T
h
is
g
r
o
w
s
t
h
e
m
u
lti
f
ac
eted
id
ea
o
f
th
e
m
o
d
el
as
th
er
e
ar
e
m
o
r
e
p
ar
am
e
t
er
s
f
o
r
th
e
m
o
d
el
to
f
it
to
th
e
d
ata
to
g
iv
e
th
is
ca
p
r
icio
u
s
n
e
s
s
.
A
t
u
n
i
n
g
p
ar
a
m
eter
is
d
is
p
la
y
ed
ca
lled
b
asicall
y
C
t
h
at
p
o
r
tr
ay
s
th
e
s
p
an
o
f
t
h
e
s
q
u
ir
m
allo
w
ed
o
v
er
all
e
s
ti
m
atio
n
s
.
T
h
e
C
p
ar
a
m
eter
s
d
escr
ib
es
t
h
e
m
ea
s
u
r
e
o
f
e
n
cr
o
ac
h
m
e
n
t
o
f
th
e
ed
g
e
allo
w
ed
.
A
C
=0
is
n
o
e
n
cr
o
ac
h
m
en
t
a
n
d
w
e
ar
e
b
ac
k
to
t
h
e
u
n
b
en
d
a
b
le
Ma
x
i
m
al
-
Ma
r
g
in
C
la
s
s
i
f
ie
r
d
ep
icted
alr
ea
d
y
.
T
h
e
g
r
ea
ter
th
e
esti
m
atio
n
o
f
C
th
e
g
r
ea
ter
en
cr
o
ac
h
m
e
n
t
o
f
th
e
h
y
p
er
p
lan
e
ar
e
p
er
m
itted
.
T
h
e
tak
in
g
o
f
th
e
h
y
p
er
p
lan
e
f
r
o
m
d
ata,
all
r
ea
d
in
es
s
ca
s
es
th
a
t
ex
is
t
i
n
th
e
d
iv
is
io
n
o
f
th
e
ed
g
e
w
ill
i
m
p
ac
t
th
e
cir
cu
m
s
ta
n
ce
o
f
th
e
h
y
p
er
p
lan
e
a
n
d
ar
e
s
u
g
g
ested
as
h
elp
v
ec
to
r
s
.
L
i
k
e
w
i
s
e,
as
C
i
m
p
ac
ts
t
h
e
a
m
o
u
n
t
o
f
ev
e
n
ts
t
h
at
ar
e
allo
w
ed
to
f
all
i
n
s
id
e
th
e
ed
g
e,
C
i
m
p
ac
ts
t
h
e
a
m
o
u
n
t o
f
as
s
is
tan
ce
v
ec
to
r
s
u
s
ed
b
y
t
h
e
m
o
d
el.
I
n
th
is
s
h
o
r
t
liter
atu
r
e
s
u
r
v
e
y
w
e
w
o
u
ld
lik
e
to
d
is
cu
s
s
ab
o
u
t
d
if
f
er
e
n
t
ap
p
r
o
ac
h
es
w
o
r
k
ed
o
u
t
b
y
d
if
f
er
e
n
t
r
esear
c
h
er
s
o
v
er
th
e
g
lo
b
e.
Ma
ch
i
n
e
L
ea
r
n
i
n
g
i
s
t
h
e
b
ase
co
n
ce
p
t
b
eh
i
n
d
t
h
e
m
i
n
in
g
th
e
s
e
v
er
it
y
o
f
ac
cid
en
ts
.
A
s
w
e
d
is
cu
s
s
ed
p
r
ev
io
u
s
o
v
er
4
m
illi
o
n
ca
s
es
ar
e
b
ein
g
r
ec
o
r
d
ed
as
r
o
a
d
ac
cid
en
ts
ev
er
y
y
ea
r
.
So
m
e
o
f
t
h
e
m
ac
h
i
n
e
lear
n
i
n
g
alg
o
r
ith
m
s
li
k
e
cl
u
s
ter
i
n
g
is
u
s
ed
as
u
n
s
u
p
er
v
is
ed
lear
n
i
n
g
t
ec
h
n
iq
u
e.
W
e
n
ee
d
to
co
n
s
id
er
clu
s
ter
s
f
o
r
a
s
p
ec
if
ic
f
u
n
ctio
n
i
n
th
e
d
ata
s
et.
T
h
e
f
u
n
c
tio
n
m
a
y
b
e
a
r
ea
s
o
n
o
f
g
etti
n
g
ac
cid
en
t.
Fo
r
ex
a
m
p
le
o
v
er
s
p
ee
d
m
i
g
h
t
b
e
o
n
e
r
ea
s
o
n
s
o
w
ill b
e
co
n
s
i
d
er
in
g
t
h
at
as o
n
e
o
f
t
h
e
f
u
n
cti
o
n
.
A
N
N
(
A
r
ti
f
icial
Neu
r
al
Net
wo
r
k
s
)
[
7
]
w
i
ll b
e
h
elp
in
g
f
o
r
a
n
al
y
z
in
g
t
h
e
r
o
ad
ac
cid
en
t
s
w
i
th
d
i
f
f
er
e
n
t
p
ar
am
eter
s
.
T
r
ee
b
ased
an
al
y
z
in
g
is
o
n
e
o
th
er
co
n
ce
p
t
[
8
]
,
if
w
e
co
n
s
id
er
L
C
C
(
L
ate
n
t
C
la
s
s
C
lu
s
ter
i
n
g
)
i
t
is
f
aster
a
n
d
ac
cu
r
ate
th
a
n
k
-
NN
w
it
h
s
o
m
e
f
u
n
ctio
n
s
o
f
t
h
e
d
ata
s
et.
[
9
]
-
[
1
3
]
.
let’
s
tak
e
a
s
h
o
r
e
r
ev
ie
w
o
n
t
h
e
d
ata
m
i
n
i
n
g
tech
n
iq
u
es
w
h
ic
h
ar
e
b
ei
n
g
u
s
ed
in
d
if
f
er
en
t
d
o
m
ai
n
s
o
f
r
e
s
ea
r
ch
o
v
er
t
h
e
g
lo
b
e
b
y
d
i
f
f
er
en
t
r
esear
ch
er
s
.
T
h
e
r
ea
s
o
n
to
k
n
o
w
ab
o
u
t
th
e
o
th
er
r
esear
ch
d
o
m
a
in
s
r
eg
ar
d
in
g
th
e
d
ata
m
i
n
in
g
tec
h
n
iq
u
es
is
t
o
k
n
o
w
t
h
e
m
a
in
f
u
n
ctio
n
alit
y
o
f
ea
ch
an
d
ev
er
y
th
in
g
.
T
h
er
e
ar
e
f
e
w
f
u
n
d
a
m
e
n
tal
o
p
er
atio
n
s
in
t
h
e
d
ata
m
i
n
in
g
a
n
d
o
n
e
a
m
o
n
g
t
h
o
s
e
i
s
to
s
p
lit
th
e
d
ata
s
et
i
n
to
d
if
f
er
en
t
clu
s
ter
s
f
o
r
th
e
b
etter
clu
s
ter
i
n
g
o
p
er
atio
n
s
.
C
lu
s
te
r
i
n
g
is
u
n
s
u
p
er
v
i
s
ed
lea
r
n
in
g
in
w
h
ich
w
e
h
a
v
e
n
o
s
p
e
cif
ic
p
r
ed
icted
o
u
tp
u
t b
a
s
ed
o
n
t
h
e
a
v
ailab
le
d
ata
an
d
p
ast
d
ata
a
v
ailab
le
w
e
n
e
ed
to
p
er
f
o
r
m
t
h
e
o
p
er
atio
n
s
an
d
o
b
tain
t
h
e
p
r
ed
ictio
n
r
e
s
u
lts
[
1
4
]
,
[
1
5
]
.
I
f
w
e
co
n
s
id
er
th
e
cl
u
s
ter
in
g
w
e
n
ee
d
to
s
p
lit
t
h
e
d
ata
s
et
to
i
d
en
tify
t
h
e
co
m
m
o
n
a
n
d
s
a
m
e
ca
te
g
o
r
y
o
f
th
e
f
u
n
ctio
n
s
i
n
th
e
d
ata
s
et.
S
u
p
p
o
s
e
if
w
e
ar
e
co
n
s
id
er
in
g
th
e
ac
cid
en
t
s
e
v
er
it
y
i
n
o
u
r
ca
s
e
th
er
e
m
a
y
b
e
d
if
f
er
e
n
t
f
u
n
c
tio
n
s
to
b
e
co
n
s
id
er
ed
an
d
s
o
m
e
ca
s
e
s
w
e
n
e
ed
to
co
n
s
id
er
t
h
e
co
m
b
in
at
i
o
n
o
f
t
h
e
f
u
n
ctio
n
s
f
r
o
m
t
h
e
d
ataset.
L
et
s
ta
k
e
a
n
ex
a
m
p
le
r
eg
ar
d
i
n
g
th
e
cl
u
s
ter
in
g
th
e
d
ata
s
et.
C
o
n
s
id
er
t
h
e
s
a
m
p
le
T
ab
le
2
b
elo
w
w
h
ich
i
s
h
a
v
i
n
g
s
o
m
e
c
o
m
m
o
n
th
i
n
g
s
i
n
t
h
e
d
ataset.
B
y
co
n
s
id
er
in
g
t
h
e
T
ab
le
1
w
e
ca
n
s
a
y
t
h
at
m
o
s
t
o
f
th
e
a
cc
id
en
ts
ar
e
h
ap
p
en
in
g
to
t
h
e
ca
r
r
id
er
s
,
r
ea
s
o
n
s
m
a
y
b
e
o
v
er
s
p
ee
d
,
d
r
u
n
k
a
n
d
d
r
iv
e
etc.
W
e
n
ee
d
to
f
o
r
m
th
e
cl
u
s
ter
s
b
ased
o
n
th
e
m
o
s
t
w
ei
g
h
t
r
ea
s
o
n
f
o
r
th
e
ac
cid
en
t.
T
ab
le
1
.
Sam
p
le
Data
f
r
o
m
Da
taset to
i
m
p
le
m
en
t sa
m
p
le
clu
s
ter
in
g
S
t
a
t
e
V
e
h
i
c
l
e
Ty
p
e
s
Est
i
m
a
t
e
d
A
c
c
i
d
e
n
t
R
e
a
so
n
Est
i
m
a
t
e
d
c
o
u
n
t
AP
C
a
r
s
O
v
e
r
sp
e
e
d
,
d
r
u
n
k
a
n
d
d
r
i
v
e
1
5
0
UP
C
a
r
s,
b
i
k
e
s
O
v
e
r
S
p
e
e
d
,
l
a
c
k
o
f
safe
t
y
me
a
su
r
e
s
1
2
0
,
5
0
MH
B
i
k
e
s
L
a
c
k
o
f
safe
t
y
M
e
a
su
r
e
s
2
0
0
K
e
r
a
l
a
C
a
r
s,
B
i
k
e
s,
b
u
s
O
v
e
r
sp
e
e
d
,
D
r
u
n
k
a
n
d
d
r
i
v
e
,
L
a
c
k
o
f
safe
t
y
me
a
su
r
e
s
5
0
,
2
5
,
1
5
K
a
r
n
a
t
a
k
a
C
a
r
s
O
v
e
r
S
p
e
e
d
,
V
i
o
l
a
t
i
n
g
t
r
a
f
f
i
c
r
u
l
e
s,
L
a
c
k
o
f
safe
t
y
me
a
su
r
e
s
1
5
0
TN
B
u
s
,
C
a
r
,
L
o
r
r
y
,
W
a
l
k
e
r
s
U
si
n
g
p
h
o
n
e
s
o
n
r
o
a
d
,
O
v
e
r
sp
e
e
d
,
r
o
a
d
i
ssu
e
s,
d
r
u
n
k
a
n
d
d
r
i
v
e
1
5
,
1
2
0
,
2
0
0
,
5
0
TS
C
a
r
s,
B
u
s
O
v
e
r
S
p
e
e
d
,
R
o
a
d
S
a
f
e
t
y
1
5
0
,
2
0
0
3.
P
RO
P
O
SE
D
AP
P
RO
ACH
W
e
h
av
e
s
ee
n
s
o
m
e
o
f
t
h
e
c
l
ass
i
f
icatio
n
al
g
o
r
ith
m
s
[
1
6
]
-
[
19
]
an
d
r
u
les
w
h
ich
ar
e
b
ase
d
o
n
lates
t
m
ac
h
in
e
lear
n
in
g
tec
h
n
iq
u
es.
C
lu
s
ter
in
g
is
b
ased
o
n
u
n
s
u
p
er
v
is
ed
lear
n
in
g
,
K
-
NN,
K
-
Me
an
s
[
20
]
is
also
u
n
d
er
u
n
s
u
p
er
v
is
ed
lear
n
in
g
tech
n
o
lo
g
y
.
L
et
u
s
ta
k
e
a
ti
m
e
a
n
d
ex
ec
u
te
t
h
e
s
a
m
e
d
ata
s
ets
w
h
ic
h
ar
e
av
ailab
le
i
n
s
u
p
er
v
i
s
ed
lear
n
i
n
g
.
SVM
(
S
u
p
p
o
r
t
Vec
to
r
Ma
ch
in
e
s
)
,
C
NB
C
la
s
s
i
f
ier
ar
e
t
h
e
t
w
o
clas
s
if
icatio
n
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8708
Da
ta
Min
in
g
A
p
p
r
o
a
ch
o
f A
cc
id
en
t O
cc
u
r
r
en
ce
s
I
d
en
tifi
ca
tio
n
w
ith
E
ffectiv
e
.
..
(
Meen
u
Gu
p
ta
)
4037
alg
o
r
ith
m
s
w
h
ic
h
w
e
ar
e
ex
p
l
ain
i
n
g
i
n
t
h
is
ar
ticle
.
B
ased
o
n
th
e
th
r
ee
ca
teg
o
r
ies
w
e
w
o
u
ld
lik
e
to
ex
p
lain
o
u
r
w
o
r
k
in
ac
cid
en
t
s
ev
er
i
t
y
.
B
O
W
(
B
ag
o
f
W
o
r
d
s
)
,
w
o
r
d
f
r
eq
u
en
c
y
a
n
d
w
o
r
d
r
ak
i
n
g
.
B
OW
is
co
n
s
is
tin
g
o
f
t
h
e
s
et
o
f
p
r
e
-
d
ef
i
n
ed
w
o
r
d
s
w
h
ic
h
ar
e
m
o
s
tl
y
u
s
ed
to
ex
p
lain
t
h
e
r
esear
ch
co
m
p
o
n
e
n
t
in
t
h
e
ap
p
licatio
n
.
Su
p
p
o
r
t
if
w
e
ar
e
h
av
i
n
g
d
ata
s
et
w
i
th
s
o
m
e
w
o
r
d
s
li
k
e
h
ell
m
ate,
s
ea
tb
elt,
s
p
ee
d
etc
t
h
o
s
e
t
h
i
n
g
s
w
ill
b
e
co
n
s
id
er
ed
as
b
ag
o
f
w
o
r
d
s
.
First
w
e
n
ee
d
to
p
er
f
o
r
m
th
e
p
r
e
-
p
r
o
ce
s
s
in
g
o
f
th
e
d
ata
s
et.
W
e
n
ee
d
to
i
d
en
tify
t
h
e
m
i
s
s
i
n
g
v
alu
e
s
in
t
h
e
d
ata
s
et
an
d
w
e
n
ee
d
to
s
u
b
s
tit
u
te
th
e
m
is
s
in
g
v
alu
e
s
w
it
h
th
e
r
elate
d
v
a
lu
e
s
,
w
h
e
th
er
it
m
a
y
b
e
co
n
s
id
er
in
g
t
h
e
m
ea
n
o
r
m
ed
i
an
o
f
t
h
e
v
al
u
es
o
f
th
a
t
f
u
n
cti
o
n
o
r
o
b
j
ec
t.
L
ets
tak
e
a
lo
o
k
o
f
t
h
e
s
a
m
p
le
tab
le
w
h
ic
h
w
ill co
n
s
is
t
in
g
o
f
t
h
e
s
a
m
p
le
d
ata
w
h
ic
h
m
ig
h
t b
e
av
a
ilab
le
w
it
h
t
h
e
d
ata
s
et.
T
h
is
s
a
m
p
le
d
ata
s
et
f
r
o
m
T
ab
le
2
w
il
l
b
e
u
s
ed
f
o
r
p
r
e
p
r
o
ce
s
s
in
g
in
m
ac
h
i
n
e
lear
n
i
n
g
tech
n
iq
u
e
m
a
y
b
e
u
s
i
n
g
p
y
t
h
o
n
o
r
R
p
r
o
g
r
a
m
m
i
n
g
.
I
n
t
h
is
p
r
o
ce
s
s
we
n
ee
d
to
eli
m
i
n
ate
o
r
h
an
d
le
th
e
m
is
s
i
n
g
v
al
u
e
s
.
W
h
ile
h
a
n
d
lin
g
th
e
m
i
s
s
i
n
g
v
alu
es
w
e
n
ee
d
to
id
en
ti
f
y
t
h
e
t
ex
t
v
al
u
es
a
n
d
n
ee
d
to
co
n
v
er
t
th
o
s
e
to
n
u
m
er
ical
f
o
r
m
at
to
ap
p
l
y
p
r
ed
ictio
n
o
r
d
ata
m
in
in
g
clas
s
i
f
icatio
n
alg
o
r
ith
m
.
A
l
g
o
r
ith
m
s
w
e
ar
e
u
s
i
n
g
ca
n
’
t
b
e
ab
le
to
h
an
d
le
t
h
e
s
tr
in
g
f
o
r
m
at
in
t
h
e
d
ata
s
et
al
w
a
y
s
.
T
h
er
e
is
a
s
eq
u
en
ce
to
f
o
llo
w
to
p
r
ed
ict
th
e
ac
cu
r
ac
y
o
r
to
p
r
ed
ict
th
e
m
ai
n
r
ea
s
o
n
b
eh
i
n
d
th
ese
ac
cid
en
t
s
.
L
et
s
tak
e
a
clea
r
lo
o
k
o
n
th
e
f
lo
w
w
i
th
F
i
g
u
r
e
4
.
T
ab
le
2
.
Sam
p
le
Data
s
et
w
i
th
s
o
m
e
m
is
s
in
g
v
a
lu
e
s
S
t
a
t
e
N
u
mb
e
r
o
f
a
c
c
i
d
e
n
t
s
D
e
a
d
C
a
se
s
I
n
j
u
r
e
d
C
a
se
s
R
e
a
so
n
I
d
e
n
t
i
f
i
c
a
t
i
o
n
s
A
n
d
h
r
a
P
r
a
d
e
sh
1
5
0
25
1
2
5
L
a
k
e
o
f
h
e
l
l
mat
e
,
o
v
e
r
sp
e
e
d
,
w
r
o
n
g
c
u
t
V
e
h
i
c
l
e
d
a
m
a
g
e
d
se
v
e
r
e
l
y
,
w
r
o
n
g
c
u
t
R
a
j
a
st
h
a
n
1
0
0
50
50
S
e
a
t
b
e
l
t
,
o
v
e
r
sp
e
e
d
W
r
o
n
g
c
u
t
M
a
h
a
r
a
s
t
r
a
1
0
0
25
V
e
h
i
c
l
e
d
a
m
a
g
e
d
Fig
u
r
e
4
.
Stru
ct
u
r
e
o
f
th
e
m
i
n
i
n
g
t
h
e
d
ata
s
et
First
w
e
n
ee
d
to
lo
ad
th
e
d
ata
s
et
w
h
ic
h
w
e
n
ee
d
to
p
r
o
ce
s
s
.
L
ater
d
o
s
o
m
e
p
r
e
-
p
r
o
ce
s
s
i
n
g
s
tep
s
li
k
e
eli
m
i
n
ati
n
g
th
e
m
i
s
s
i
n
g
v
a
lu
e
s
an
d
s
u
b
s
tit
u
ti
n
g
t
h
o
s
e
w
it
h
th
e
v
a
lid
in
f
o
r
m
at
io
n
li
k
e
m
ea
n
o
f
t
h
e
d
ata
o
f
m
ed
ian
.
T
h
en
s
elec
t
t
h
e
cla
s
s
if
icatio
n
al
g
o
r
ith
m
w
it
h
w
h
ic
h
w
e
n
ee
d
to
ap
p
ly
.
T
h
e
m
is
s
in
g
v
alu
e
s
clea
n
ed
d
ata
s
et
m
u
s
t
b
e
s
ep
ar
ated
as
t
r
ain
in
g
a
n
d
test
d
ata
s
et.
T
h
e
t
r
ain
in
g
d
atase
t
w
ill
b
e
u
s
ed
f
o
r
tr
ain
t
h
e
m
ac
h
i
n
e
o
r
class
i
f
icatio
n
a
lg
o
r
it
h
m
w
h
ich
w
e
ar
e
w
r
i
tin
g
;
test
d
ata
s
et
is
u
s
ed
to
co
r
r
elate
t
h
e
t
h
in
g
s
w
i
th
th
e
r
eq
u
ir
ed
r
esu
lt.
W
e
n
ee
d
to
test
th
e
v
a
l
u
es
o
f
t
h
e
d
ata
s
et
w
it
h
t
h
e
tr
ain
i
n
g
s
et
an
d
h
a
v
e
to
co
r
r
elate
w
it
h
t
h
e
p
r
ev
io
u
s
w
o
r
k
o
r
w
it
h
th
e
tr
ai
n
i
n
g
d
ata
s
et
[2
1
]
-
[2
3
].
Af
ter
s
elec
ti
n
g
th
e
cla
s
s
i
f
icati
o
n
alg
o
r
ith
m
,
i
f
w
e
s
elec
t
th
e
SVM
alg
o
r
it
h
m
,
w
e
n
ee
d
to
s
elec
t
h
o
w
m
ai
n
co
l
u
m
n
s
o
r
r
o
w
s
w
e
n
ee
d
to
u
s
e
f
o
r
t
h
e
te
s
t
s
et
to
co
r
r
elate
,
th
e
n
s
u
b
m
it
t
h
e
v
al
u
es.
T
h
e
r
esu
lt
w
ill
b
e
i
n
th
r
ee
t
y
p
e
s
.
I
t
w
i
ll d
o
B
OW
co
llectio
n
,
w
o
r
d
co
u
n
t
an
d
w
o
r
d
f
r
eq
u
en
c
y
.
B
ased
o
n
t
h
e
w
o
r
d
f
r
eq
u
en
c
y
w
e
ca
n
esti
m
ate
t
h
at
w
h
ic
h
i
s
t
h
e
m
ai
n
r
ea
s
o
n
b
eh
i
n
d
t
h
e
s
e
v
er
r
o
ad
ac
cid
en
ts
.
T
h
e
s
a
m
e
f
o
llo
w
s
w
it
h
C
NB
class
i
f
ier
,
b
u
t
t
h
e
th
i
n
g
w
ill
c
h
an
g
e
h
er
e
i
s
w
e
n
ee
d
to
g
i
v
e
s
a
m
p
le
co
u
n
t
o
f
co
lu
m
n
s
a
n
d
r
o
w
s
to
p
r
o
ce
s
s
,
it
w
il
l ta
k
e
e
n
tire
d
ataset
w
it
h
o
u
t
m
is
s
i
n
g
v
al
u
es a
n
d
i
m
p
l
y
W
E
KA
to
o
l o
n
it a
n
d
p
r
o
d
u
ce
th
e
esti
m
a
ted
r
esu
lt.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
8
,
No
.
5
,
Octo
b
er
2
0
1
8
:
4
0
3
3
–
4
0
4
1
4038
I
n
th
e
later
p
ar
t
o
f
th
e
s
ec
tio
n
w
e
w
ill
d
is
c
u
s
s
th
e
e
x
p
er
im
en
tal
r
esu
lts
w
it
h
r
elate
d
to
th
e
s
a
m
p
l
e
d
ata
s
et
w
e
ar
e
u
s
i
n
g
f
o
r
th
e
p
r
o
ce
s
s
in
g
o
f
t
h
e
d
ata.
T
o
b
e
p
r
ec
is
e
th
er
e
ar
e
th
r
ee
ty
p
es
o
f
r
esu
l
ts
w
e
ac
q
u
ir
e
an
d
w
e
h
a
v
e
alr
ea
d
y
d
is
cu
s
s
ed
th
e
t
y
p
es o
f
r
esu
l
ts
w
e
ar
e
g
o
i
n
g
to
g
et
w
it
h
th
i
s
ex
p
er
i
m
en
t.
As
w
e
d
is
cu
s
s
ed
t
h
e
p
r
o
p
o
s
e
d
ap
p
r
o
ac
h
to
id
en
tify
th
e
ac
cid
en
t
s
e
v
er
it
y
u
s
i
n
g
t
w
o
cla
s
s
i
f
icatio
n
alg
o
r
ith
m
s
it
w
o
r
t
h
to
k
n
o
w
a
b
o
u
t
th
e
w
h
et
h
er
t
h
ese
t
w
o
w
i
ll
co
m
p
letel
y
s
a
tis
f
y
o
u
r
r
eq
u
i
r
e
m
en
t
o
r
an
y
th
i
n
g
n
ee
d
to
b
e
in
c
lu
d
ed
.
C
o
m
i
n
g
to
p
r
o
s
o
f
th
e
s
e
t
w
o
ap
p
r
o
ac
h
es
is
w
e
n
ee
d
n
o
t
i
n
cl
u
d
e
ev
e
r
y
f
u
n
ctio
n
in
to
th
e
alg
o
r
ith
m
o
r
t
h
e
m
o
d
el
w
h
i
c
h
w
e
ar
e
u
s
i
n
g
.
T
h
e
e
n
tire
th
i
n
g
w
e
n
ee
d
is
li
m
ited
m
o
d
el
d
at
a
o
r
f
u
n
ct
io
n
s
to
b
e
i
m
p
le
m
en
ted
in
th
e
alg
o
r
it
h
m
.
T
h
ese
t
w
o
w
ill
g
i
v
e
q
u
ic
k
r
es
u
lt
s
t
h
an
o
th
er
al
g
o
r
ith
m
s
.
As
t
h
e
s
e
t
w
o
ar
e
o
ld
est
alg
o
r
ith
m
s
an
d
clas
s
i
f
i
ca
tio
n
m
o
d
els
th
e
e
x
p
ec
ted
r
esu
lt
s
m
a
y
b
e
v
ar
y
as
w
e
p
r
ed
icted
.
A
s
w
e
u
s
e
li
m
ited
n
u
m
b
er
o
f
f
u
n
ctio
n
s
we
ca
n
n
o
t g
e
t th
e
co
m
p
le
te
an
al
y
s
i
s
o
f
t
h
e
p
r
ed
icted
th
in
g
s
r
e
q
u
ir
ed
.
T
h
e
b
etter
w
a
y
to
s
o
lv
e
th
e
p
r
o
b
lem
r
e
g
ar
d
in
g
t
h
e
ac
ci
d
en
ts
s
ev
er
it
y
w
e
ca
n
m
ak
e
u
s
e
o
f
th
e
clu
s
ter
i
n
g
alg
o
r
it
h
m
s
,
K
-
Me
a
n
s
,
A
NN
etc.
So
t
h
at
w
e
ca
n
g
et
th
e
ap
t r
esu
lt
s
w
e
r
eq
u
ir
ed
p
r
ed
icted
r
esu
lts
.
4.
E
XP
E
R
I
M
E
NT
A
L
RE
SUL
T
S
T
h
e
r
esu
lts
w
e
ac
q
u
ir
e
h
er
e
h
av
e
th
r
ee
t
y
p
e
s
an
d
th
e
f
ir
s
t
t
h
in
g
i
s
b
ag
o
f
w
o
r
d
s
co
llectio
n
(
B
OW
)
.
B
ased
o
n
t
h
e
n
u
m
b
er
o
f
v
al
u
es
w
e
a
s
s
i
g
n
ed
w
e
ca
n
ca
l
cu
late
th
e
ac
cu
r
ac
y
o
f
t
h
e
a
lg
o
r
ith
m
.
Fig
u
r
e
5
Descr
ib
es
t
h
e
g
r
ap
h
o
f
p
r
ed
i
cted
r
esu
lt
s
w
h
ic
h
d
e
s
cr
ib
es
t
h
e
m
ai
n
r
ea
s
o
n
f
o
r
t
h
e
ac
cid
en
ts
i
n
t
h
o
s
e
ar
ea
s
.
A
cc
u
r
ac
y
i
s
b
ased
o
n
th
e
ti
m
e
tak
e
n
an
d
t
h
e
n
u
m
b
er
o
f
r
o
w
s
o
r
co
lu
m
n
s
p
r
o
ce
s
s
ed
w
it
h
th
e
g
iv
e
n
class
i
f
icatio
n
alg
o
r
it
h
m
u
s
in
g
Data
Min
i
n
g
o
r
Ma
ch
in
e
L
ea
r
n
in
g
[2
4
]
-
[2
6
]
.
Fig
u
r
e
5
.
Gr
ap
h
o
f
p
r
ed
icted
r
esu
lt
B
y
t
h
i
s
g
r
ap
h
w
e
ca
n
p
r
ed
ict
th
e
m
ain
r
ea
s
o
n
f
o
r
t
h
e
s
e
v
er
it
y
o
f
ac
cid
en
ts
i
n
d
if
f
er
e
n
t
lo
ca
tio
n
s
.
C
las
s
i
f
icatio
n
p
r
o
b
le
m
s
ar
e
m
o
r
e
r
elate
d
to
th
e
Ma
ch
i
n
e
L
e
ar
n
in
g
tec
h
n
iq
u
e
w
i
th
w
h
ic
h
w
e
n
ee
d
to
tr
ain
t
h
e
m
ac
h
in
e
w
it
h
a
n
al
g
o
r
ith
m
[
2
7
]
.
Usi
n
g
M
L
t
h
e
r
es
u
lt
w
e
g
o
t
h
er
e
i
s
clas
s
i
f
ied
in
to
s
o
m
e
o
f
t
h
e
f
u
n
ctio
n
s
.
L
e
t
th
e
F
u
n
ct
io
n
b
e
R
ea
s
o
n
t
y
p
e
b
eh
in
d
th
e
ac
cid
e
n
t.
L
e
t
th
e
C
it
y
1
m
a
y
h
a
v
e
2
0
0
ca
s
es
an
d
o
u
t
o
f
th
at
1
0
0
ar
e
d
r
u
n
k
a
n
d
d
r
iv
e
r
e
m
ai
n
i
n
g
ar
e
o
v
er
s
p
ee
d
,
an
d
f
o
r
cit
y
2
th
e
to
tal
ca
s
es
m
a
y
b
e
3
0
0
an
d
d
r
u
n
k
a
n
d
d
r
iv
e
ca
s
es
ar
e
1
5
0
an
d
r
em
ain
in
g
ar
e
o
v
er
s
p
ee
d
,
n
o
tr
a
f
f
ic
r
u
les
ar
e
f
o
llo
w
ed
etc.
[
28
]
,
[2
9
]
.
T
h
er
ef
o
r
e
w
e
ca
n
g
et
th
e
r
esu
lt t
h
at
d
r
u
n
k
a
n
d
d
r
iv
e
is
t
h
e
m
aj
o
r
f
u
n
ctio
n
w
h
ic
h
is
co
m
m
o
n
i
n
all
th
e
a
s
p
ec
ts
.
W
e
n
ee
d
to
u
s
e
Dec
i
s
io
n
T
r
e
es
[
29
]
,
A
NN
f
r
o
m
t
h
e
m
ac
h
in
e
lear
n
in
g
co
m
m
u
n
it
y
[
30
]
f
o
r
b
etter
p
r
ed
ictio
n
m
o
d
els
f
o
r
th
e
d
o
m
ai
n
o
f
r
e
s
ea
r
ch
.
A
NN
h
er
e
m
a
y
b
e
u
s
ed
to
p
r
ed
ict
th
e
f
u
tu
r
e
ca
u
s
e
o
f
t
h
e
ac
cid
en
ts
a
n
d
to
id
en
ti
f
y
t
h
e
r
atio
o
f
h
ap
p
en
i
n
g
o
f
t
h
e
ac
cid
en
t
to
t
h
e
s
p
ec
i
f
ic
r
ea
s
o
n
.
T
h
a
t
m
ea
n
s
w
e
n
ee
d
to
p
r
ed
ict
th
e
r
ea
s
o
n
w
h
ic
h
m
a
y
ca
u
s
e
a
n
d
ef
f
ec
t
in
f
u
t
u
r
e
an
d
h
o
w
m
u
c
h
r
atio
t
h
e
ca
u
s
e
m
a
y
tak
e
p
ar
t
in
t
h
e
h
ap
p
en
ed
ef
f
ec
t li
k
e
ac
cid
en
t
i
n
a
s
p
ec
if
ic
r
e
g
io
n
.
I
n
T
h
is
r
esear
ch
w
e
ar
e
p
lan
n
i
n
g
to
i
m
p
le
m
e
n
t so
m
e
o
f
th
e
a
d
v
an
ce
d
al
g
o
r
ith
m
s
l
ik
e
ANN
,
Dec
is
io
n
tr
ee
s
,
R
eg
r
es
s
io
n
al
g
o
r
ith
m
s
li
k
e
SV
R
(
Su
p
p
o
r
t V
ec
to
r
R
eg
r
ess
io
n
)
to
d
esi
g
n
b
etter
p
r
ed
ictio
n
alg
o
r
it
h
m
w
i
th
th
e
av
ailab
le
d
ata
s
ets.
W
e
co
llected
th
e
p
u
b
lic
d
ata
s
et
a
v
ailab
le
f
r
o
m
t
h
e
g
o
v
er
n
m
en
t
r
esear
ch
w
eb
s
it
e
w
h
ic
h
w
ill
g
i
v
e
th
e
b
r
ie
f
in
f
o
r
m
atio
n
ab
o
u
t
th
e
d
i
f
f
er
e
n
t
r
ea
s
o
n
s
b
eh
in
d
t
h
e
ac
cid
en
t
s
an
d
h
o
w
m
a
n
y
n
u
m
b
er
o
f
ca
s
e
s
ar
e
r
ec
o
r
d
ed
r
eg
io
n
w
i
s
e
i
n
th
e
s
p
an
o
f
y
ea
r
s
.
T
h
e
r
ea
s
o
n
s
w
i
ll
b
e
clea
r
w
i
th
a
p
ictu
r
e
t
h
at
th
e
m
ai
n
r
ea
s
o
n
m
a
y
b
e
n
o
t
f
o
llo
w
i
n
g
t
h
e
tr
a
f
f
ic
r
u
les
an
d
o
v
er
s
p
ee
d
ar
e
t
h
e
m
ai
n
r
ea
s
o
n
s
f
o
r
t
h
e
a
cc
id
en
ts
s
e
v
er
it
y
i
n
ev
er
y
r
eg
io
n
.
T
h
e
f
o
llo
w
in
g
i
m
ag
e
Fi
g
u
r
e
6
w
ill
ex
p
lai
n
th
e
s
a
m
p
le
ab
o
u
t
t
h
e
co
ef
f
icie
n
t
an
d
s
ta
n
d
ar
d
d
ev
iatio
n
le
v
els i
n
o
u
r
al
g
o
r
ith
m
r
elate
d
to
th
e
d
o
m
ain
o
f
r
es
ea
r
ch
.
Fo
r
b
etter
u
n
d
er
s
ta
n
d
in
g
o
f
t
h
e
d
ec
is
io
n
tr
ee
s
an
d
d
ec
is
io
n
alg
o
r
ith
m
s
,
a
n
d
d
ata
m
i
n
in
g
t
ec
h
n
iq
u
e
s
w
e
ca
n
ta
k
e
an
y
h
ea
lt
h
ca
r
e
ex
a
m
p
le
li
k
e
ca
n
ce
r
[
3
1
]
.
W
e
a
p
p
ly
s
o
m
e
o
f
th
e
d
ata
m
i
n
in
g
k
n
o
w
led
g
e
o
n
th
at
to
p
r
ed
ict
th
e
ca
n
ce
r
p
er
ce
n
ta
g
e
a
n
d
th
e
f
u
n
ctio
n
al
li
f
e
ti
m
e
o
f
t
h
at
p
atie
n
t
a
n
d
t
h
e
s
e
v
e
r
it
y
o
f
th
e
d
is
ea
s
e.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8708
Da
ta
Min
in
g
A
p
p
r
o
a
ch
o
f A
cc
id
en
t O
cc
u
r
r
en
ce
s
I
d
en
tifi
ca
tio
n
w
ith
E
ffectiv
e
.
..
(
Meen
u
Gu
p
ta
)
4039
[3
2
]
-
[
3
4
]
.
Data
m
in
in
g
a
n
d
Ma
ch
in
e
L
ea
r
n
i
n
g
ar
e
t
h
e
t
w
o
ar
ea
s
w
h
ic
h
ar
e
u
s
ed
f
o
r
th
e
f
u
r
t
h
e
r
r
esear
ch
o
f
th
e
d
o
m
ai
n
s
li
k
e
p
r
ed
ictin
g
t
h
e
ac
cid
en
t
p
r
o
n
e
ar
ea
s
an
d
t
y
p
es
o
f
r
ea
s
o
n
s
b
ased
o
n
t
h
e
lo
ca
lit
y
in
th
e
f
u
t
u
r
e.
T
h
e
f
u
tu
r
e
o
f
d
ata
m
in
in
g
is
m
ac
h
i
n
e
lear
n
i
n
g
.
Fig
u
r
e
6
.
C
o
ef
f
icien
ts
a
n
d
th
e
Stan
d
ar
d
E
r
r
o
r
ex
p
lan
atio
n
Fig
u
r
e
7
ex
p
lai
n
s
th
e
co
u
n
t
o
f
ac
cid
en
ts
to
tal
l
y
in
o
n
e
lo
ca
tio
n
.
L
et
it
b
e
o
n
e
cit
y
o
r
s
ta
t
e.
So
th
at
th
ese
ar
e
th
e
to
tal
n
u
m
b
er
o
f
ac
cid
en
ts
d
o
n
e
i
n
o
n
e
m
o
n
t
h
an
d
w
e
ca
n
m
a
k
e
a
co
n
clu
s
i
o
n
th
a
t
b
ec
au
s
e
o
f
L
o
r
r
ies
m
o
r
e
ac
cid
en
ts
ar
e
h
a
p
p
en
in
g
.
W
h
eth
er
i
t
m
a
y
b
e
b
ec
au
s
e
o
f
th
e
o
v
er
s
p
ee
d
o
r
d
r
u
n
k
a
n
d
d
r
iv
e.
W
e
ca
n
s
ee
t
h
e
co
m
b
in
a
tio
n
o
f
t
h
o
s
e
in
Fig
u
r
e
5
.
I
n
Fi
g
u
r
e
5
w
e
w
ill
g
et
t
h
e
co
m
b
in
a
tio
n
o
f
t
h
e
r
ea
s
o
n
o
f
ac
cid
en
ts
i
n
o
n
e
s
tate
f
o
r
o
n
e
m
o
n
t
h
.
Fig
u
r
e
7
.
P
lo
ttin
g
ac
cid
en
t
s
s
e
v
er
it
y
B
ased
o
n
F
ig
u
r
e
8
th
e
m
aj
o
r
r
ea
s
o
n
o
f
ac
cid
en
ts
i
n
o
n
e
s
ta
te
in
o
n
e
m
o
n
th
i
s
Dr
u
n
k
an
d
Dr
iv
e
an
d
No
t
Fo
llo
w
in
g
t
h
e
T
r
af
f
ic
R
u
l
es.
L
ik
e
th
i
s
w
e
ca
n
co
n
s
id
er
f
e
w
m
a
n
y
co
n
d
itio
n
s
b
ased
o
n
th
e
r
eq
u
ir
e
m
e
n
t
o
f
th
e
p
r
ed
ictio
n
m
o
d
el
an
d
its
ar
ch
itect
u
r
e
Fig
u
r
e
8
.
P
r
ed
ictin
g
m
aj
o
r
it
y
o
f
th
e
r
ea
s
o
n
f
o
r
ac
cid
en
ts
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
8
,
No
.
5
,
Octo
b
er
2
0
1
8
:
4
0
3
3
–
4
0
4
1
4040
5.
CO
NCLU
SI
O
N
T
h
e
d
ata
m
i
n
i
n
g
a
n
d
m
ac
h
i
n
e
lear
n
in
g
ar
e
t
h
e
th
i
n
g
s
w
e
n
ee
d
to
b
e
co
n
s
id
er
ed
to
id
en
ti
f
y
a
n
y
u
n
p
r
o
ce
s
s
ed
t
h
i
n
g
u
s
i
n
g
d
ata
s
ets.
I
n
t
h
i
s
ar
ticle
w
e
tr
ied
to
i
m
p
le
m
e
n
t
SVM
a
n
d
C
NB
class
i
f
ier
s
w
i
th
w
h
ich
w
e
ar
e
p
r
ed
ictin
g
t
h
e
m
a
in
r
ea
s
o
n
f
o
r
th
e
s
e
v
er
it
y
o
f
ac
cid
en
ts
an
d
w
e
also
p
r
ed
ict
th
e
m
ai
n
r
ea
s
o
n
o
n
o
v
er
all
r
esu
lt
s
.
Fo
r
ex
a
m
p
le
w
e
ca
n
c
o
n
s
id
er
ea
ch
s
tate
in
in
d
ia
an
d
w
e
ca
n
p
r
ed
ict
b
o
th
th
e
th
i
n
g
s
li
k
e
m
ai
n
r
ea
s
o
n
f
o
r
th
e
ac
cid
en
t
s
i
n
i
n
d
iv
id
u
a
l
s
tate
a
n
d
also
m
ai
n
r
ea
s
o
n
i
n
o
v
er
all
co
u
n
tr
y
.
Fo
r
s
o
m
e
ca
s
es
SVM
is
s
h
o
w
in
g
m
o
r
e
ac
c
u
r
ac
y
o
f
9
7
%
an
d
s
o
m
e
ca
s
es
C
NB
is
s
h
o
w
i
n
g
ac
c
u
r
ac
y
o
f
9
8
%.
W
ith
t
h
e
o
b
tain
ed
r
esu
lt
s
b
o
th
th
e
alg
o
r
ith
m
s
ar
e
w
o
r
k
i
n
g
w
ell
with
all
t
h
e
co
n
d
itio
n
s
co
n
s
id
er
ed
.
RE
F
E
R
E
NC
ES
[1
]
S
.
Ku
m
a
r
a
n
d
D.
T
o
sh
n
iw
a
l,
“
A
n
o
v
e
l
f
ra
m
e
w
o
rk
to
a
n
a
l
y
z
e
ro
a
d
a
c
c
id
e
n
t
ti
m
e
se
rie
s
d
a
ta
,
”
J
o
u
rn
a
l
o
f
Bi
g
Da
t
a
,
v
o
l
/i
ss
u
e
:
3
(
8
)
,
p
p
.
1
-
1
1
,
2
0
1
6
.
[2
]
M
.
Ka
rlaf
ti
s
a
n
d
A
.
T
a
rk
o
,
“
He
t
e
ro
g
e
n
e
it
y
c
o
n
sid
e
ra
ti
o
n
s
in
a
c
c
id
e
n
t
m
o
d
e
li
n
g
,
”
Acc
id
.
An
a
l.
Pre
v
.
,
v
o
l
.
30
,
n
o
.
4
,
p
p
.
4
2
5
-
4
3
3
,
1
9
9
8
.
[3
]
S
.
Ku
m
a
r
a
n
d
D.
T
o
sh
n
iw
a
l
,
“
A
n
a
l
y
si
s
o
f
Ho
u
rly
ro
a
d
Ac
c
id
e
n
t
Co
u
n
ts
u
sin
g
Hie
ra
rc
h
ica
l
Clu
ste
rin
g
a
n
d
Co
p
h
e
n
e
ti
c
Co
rre
lati
o
n
C
o
e
f
f
i
c
ie
n
t
(c
p
c
c
)”
,
J
o
u
r
n
a
l
o
f
Bi
g
Da
t
a
,
v
o
l
.
3
,
n
o
.
1
3
,
p
p
.
1
-
1
1
,
2
0
1
6
.
[4
]
P
.
N.
T
a
n
,
e
t
a
l.
,
“
In
tro
d
u
c
ti
o
n
to
Da
ta M
in
in
g
”
,
Bo
st
o
n
,
P
e
a
rso
n
A
d
d
iso
n
-
W
e
sle
y
,
p
.
7
6
9
,
2
0
0
6
.
[5
]
S
.
Ku
m
a
r
a
n
d
D.
T
o
sh
n
iw
a
l,
“
A
n
a
l
y
sin
g
ro
a
d
Ac
c
id
e
n
t
D
a
ta
u
sin
g
A
ss
o
c
iatio
n
ru
le
M
in
i
n
g
”
,
In
ter
n
a
t
io
n
a
l
Co
n
fer
e
n
c
e
o
n
Co
m
p
u
ti
n
g
Co
mm
u
n
ica
ti
o
n
a
n
d
S
e
c
u
rity (
ICCCS
-
2
0
1
5
)
,
Ka
n
y
a
k
u
m
a
ri
,
I
n
d
i
a
,
2
0
1
5
.
[6
]
J
.
Ha
n
a
n
d
M
.
Ka
m
b
e
r,
“
Da
ta
M
in
i
n
g
:
Co
n
c
e
p
ts
a
n
d
T
e
c
h
n
iq
u
e
s
”
,
Un
it
e
d
S
tate
s
,
M
o
rg
a
n
Ka
u
fm
a
n
n
P
u
b
l
ish
e
rs,
2
0
0
1
.
[7
]
L
.
M
u
ss
o
n
e
,
e
t
a
l
.
,
“
A
n
A
n
a
l
y
sis
o
f
u
rb
a
n
C
o
ll
isi
o
n
s u
sin
g
a
n
A
rti
f
icia
l
In
telli
g
e
n
c
e
M
o
d
e
l”
,
Acc
id
e
n
t
A
n
a
lys
is
a
n
d
Pre
v
e
n
ti
o
n
,
v
o
l.
3
1
,
p
p
.
7
0
5
-
7
1
8
,
1
9
9
9
.
[8
]
L
.
Ch
a
n
g
a
n
d
W
.
Ch
e
n
,
“
Da
ta
M
in
in
g
o
f
T
re
e
b
a
se
d
M
o
d
e
ls
to
A
n
a
ly
z
e
F
re
e
wa
y
A
c
c
id
e
n
t
F
re
q
u
e
n
c
y
”
,
J
o
u
rn
a
l
o
f
S
a
fety
Res
e
a
rc
h
,
v
o
l.
3
6
,
p
p
.
3
6
5
-
3
7
5
,
2
0
0
5
.
[9
]
J.
D.
Oñ
a
,
e
t
a
l.
,
“
A
n
a
l
y
sis
o
f
Traff
ic
A
c
c
id
e
n
ts
o
n
Ru
ra
l
Hig
h
wa
y
s
u
sin
g
L
a
ten
t
Clas
s
C
l
u
ste
rin
g
a
n
d
Ba
y
e
si
a
n
N
e
tw
o
rk
s”
,
Acc
id
An
a
l
Pre
v
,
v
o
l.
5
1
,
p
p
.
1
-
1
0
,
2
0
1
3
.
[1
0
]
S
.
Ku
m
a
r
a
n
d
D.
T
o
sh
n
iw
a
l,
“
A
Da
ta M
in
in
g
F
ra
m
e
w
o
rk
to
a
n
a
ly
z
e
ro
a
d
A
c
c
id
e
n
t
D
a
ta”
,
J
o
u
rn
a
l
o
f
Bi
g
Da
t
a
,
v
o
l
.
2
,
n
o
.
1
,
p
p
.
1
-
1
8
,
2
0
1
5
.
[1
1
]
V
.
K
.
S
o
lan
k
i
a
n
d
V
.
K
.
S
in
g
h
,
“
A
No
v
e
l
F
ra
m
e
w
o
rk
to
Us
e
A
ss
o
c
iatio
n
Ru
le
M
i
n
in
g
f
o
r
Clas
sif
i
c
a
ti
o
n
o
f
T
ra
ff
ic
A
c
c
id
e
n
t
S
e
v
e
rit
y
”
.
[1
2
]
M
.
G
u
p
ta
,
“
A
n
a
l
y
sis o
f
Da
ta
m
in
in
g
T
e
c
h
n
iq
u
e
f
o
r
T
ra
ff
ic
Ac
c
i
d
e
n
t
S
e
v
e
rit
y
P
r
o
b
lem
:
A
Re
v
ie
w
”
.
[1
3
]
M
.
G
u
p
ta
,
“
P
e
rf
o
rm
a
n
c
e
E
v
a
lu
a
ti
o
n
o
f
Clas
sif
ic
a
ti
o
n
Al
g
o
rit
h
m
s o
n
Dif
fe
re
n
t
Da
ta S
e
ts
”
.
[1
4
]
Z
.
Hu
a
n
g
,
“
A
F
a
st Cl
u
ste
rin
g
A
lg
o
rit
h
m
to
Cl
u
ste
r
V
e
ry
L
a
r
g
e
Ca
t
e
g
o
rica
l
Da
ta S
e
ts
in
Da
ta M
in
i
n
g
”
.
[1
5
]
Z
.
Hu
a
n
g
,
“
Ex
ten
sio
n
s t
o
t
h
e
k
-
M
e
a
n
s A
l
g
o
rit
h
m
f
o
r
Clu
ste
rin
g
L
a
rg
e
Da
ta S
e
ts
w
it
h
Ca
teg
o
rica
l
V
a
lu
e
s
”
.
[1
6
]
N
.
Do
g
a
n
a
n
d
Z
.
T
a
n
rik
u
l
u
,
“
A
Co
m
p
a
ra
ti
v
e
A
n
a
l
y
sis
o
f
Clas
si
f
i
c
a
ti
o
n
A
lg
o
rit
h
m
s
in
Da
ta
M
in
i
n
g
f
o
r
A
c
c
u
ra
c
y
,
S
p
e
e
d
a
n
d
R
o
b
u
st
n
e
ss
”
.
[1
7
]
M
a
im
o
n
O
.
a
n
d
R
o
k
a
c
h
L
.
,
“
T
h
e
Da
ta M
in
in
g
a
n
d
Kn
o
w
led
g
e
Dis
c
o
v
e
r
y
H
a
n
d
b
o
o
k
”
,
S
p
rin
g
e
r,
Be
r
li
n
,
2
0
1
0
.
[1
8
]
Ha
n
J
.
a
n
d
Ka
m
b
e
r
M
.,
“
Da
ta M
in
in
g
Co
n
c
e
p
ts
a
n
d
T
e
c
h
n
iq
u
e
s
”
,
2
n
d
e
d
n
.
M
o
rg
a
n
Ka
u
fm
a
n
n
,
US
A
,
2006
.
[1
9
]
Du
n
h
a
m
M
.
H
.,
“
Da
ta M
i
n
in
g
:
I
n
tro
d
u
c
to
ry
a
n
d
A
d
v
a
n
c
e
d
T
o
p
ics
”
,
P
re
n
ti
c
e
Ha
ll
,
Ne
w
Je
rse
y
,
2002
.
[2
0
]
T
.
N
.
P
h
y
u
,
“
S
u
rv
e
y
o
f
Clas
si
f
ica
t
io
n
T
e
c
h
n
iq
u
e
s i
n
Da
ta M
i
n
in
g
”
.
[2
1
]
P
u
tt
e
n
P
.,
e
t
a
l
.
,
“
P
ro
f
il
in
g
N
o
v
e
l
Clas
si
f
ic
a
ti
o
n
A
lg
o
r
it
h
m
s:
A
rti
f
icia
l
I
m
m
u
n
e
S
y
ste
m
”
,
Pro
c
e
e
d
in
g
s
o
f
t
h
e
7
th
IE
EE
In
ter
n
a
ti
o
n
a
l
C
o
n
fer
e
n
c
e
o
n
Cy
b
e
rn
e
ti
c
I
n
telli
g
e
n
t
S
y
ste
ms
(
CIS
2
0
0
8
),
L
o
n
d
o
n
,
UK
,
p
p
.
1
-
6
,
2
0
0
8
.
[2
2
]
He
rg
e
rt
F
.
,
e
t
a
l.
,
“
I
m
p
ro
v
in
g
M
o
d
e
l
S
e
lec
ti
o
n
b
y
D
y
n
a
m
i
c
Re
g
u
lariz
a
ti
o
n
M
e
th
o
d
s
”
,
in
P
e
tsc
h
e
T
.,
e
t
a
l
.,
“
Co
m
p
u
tatio
n
a
l
lea
rn
in
g
th
e
o
ry
a
n
d
n
a
t
u
ra
l
lea
rn
in
g
sy
st
e
m
s:
s
e
le
c
ti
n
g
g
o
o
d
m
o
d
e
ls
,”
M
IT
P
re
ss
,
Ca
m
b
rid
g
e
,
p
p
.
323
-
3
4
3
,
1
9
9
5
.
[2
3
]
Ka
e
lb
li
n
g
L
.
P
.,
“
A
ss
o
c
iati
v
e
m
e
th
o
d
s
in
re
i
n
f
o
rc
e
m
e
n
t
lea
rn
in
g
:
a
n
e
m
p
rica
l
stu
d
y
,
”
in
H
a
n
so
n
S
.
J
.
,
e
t
a
l.
,
Co
m
p
u
tatio
n
a
l
L
e
a
rn
in
g
T
h
e
o
ry
a
n
d
Na
tu
ra
l
L
e
a
rn
in
g
S
y
ste
m
s:
I
n
ters
e
c
ti
o
n
b
e
tw
e
e
n
T
h
e
o
ry
a
n
d
E
x
p
e
rime
n
t
,
M
IT
P
re
ss
,
Ca
m
b
rid
g
e
,
p
p
.
1
3
3
-
1
5
3
,
1
9
9
4
.
[2
4
]
G
e
E
.,
e
t
a
l.
,
“
Da
ta
M
in
in
g
f
o
r
L
if
e
ti
m
e
P
re
d
ic
ti
o
n
o
f
M
e
talli
c
C
o
m
p
o
n
e
n
ts
”
,
Pro
c
e
e
d
in
g
s
o
f
th
e
5
th
Au
stra
l
a
sia
n
Da
ta
M
in
in
g
C
o
n
fer
e
n
c
e
(
Au
sD
M
2
0
0
6
),
S
y
d
n
e
y
,
Au
str
a
li
a
,
p
p
.
75
-
81
,
2
0
0
6
.
[2
5
]
Ch
iarin
i
T
.
M
.
,
e
t
a
l
.
,
“
Id
e
n
t
ify
in
g
f
a
ll
-
re
late
d
In
ju
ries
:
T
e
x
t
M
in
in
g
th
e
El
e
c
tro
n
ic
M
e
d
ica
l
R
e
c
o
r
d
”
,
In
f
T
e
c
h
n
o
l
M
a
n
a
g
e
,
v
o
l
.
1
0
,
n
o
.
4
,
p
p
.
2
5
3
-
2
6
5
,
2
0
0
9
.
[2
6
]
Bre
ima
n
L
.
,
e
t
a
l.
,
“
Clas
si
f
ica
ti
o
n
a
n
d
R
e
g
re
ss
io
n
tree
”
,
W
a
d
s
w
o
r
th
&
Bro
o
k
s/Co
le
A
d
v
a
n
c
e
d
Bo
o
k
s
&
S
o
f
t
w
a
r
e
,
P
a
c
if
ic G
ro
v
e
,
1984
.
[2
7
]
R.
A
g
r
a
w
a
l,
e
t
a
l.
,
“
Da
tab
a
se
M
i
n
i
n
g
:
A
P
e
rf
o
rm
a
n
c
e
P
e
rsp
e
c
ti
v
e
”
,
IEE
E
T
ra
n
s
.
Kn
o
wle
d
g
e
a
n
d
Da
t
a
En
g
i
n
e
e
rin
g
,
v
o
l
.
5
,
n
o
.
6
,
p
p
.
9
1
4
-
9
2
5
,
1
9
9
3
.
[2
8
]
J.
R.
Qu
i
n
lan
,
“
C4
.
5
:
P
r
o
g
ra
m
s f
o
r
M
a
c
h
in
e
L
e
a
rn
in
g
”
,
M
o
rg
a
n
Ka
u
fm
a
n
n
,
1
9
9
3
.
[2
9
]
Y.
Be
n
g
io
,
e
t
a
l
.
,
“
In
tr
o
d
u
c
ti
o
n
to
t
h
e
S
p
e
c
ial
I
ss
u
e
o
n
Ne
u
ra
l
Ne
tw
o
rk
s
f
o
r
Da
ta
M
in
in
g
a
n
d
K
n
o
w
led
g
e
d
isc
o
v
e
r
y
,
”
IEE
E
T
ra
n
s.
Ne
u
ra
l
Ne
two
rk
s
,
v
o
l.
1
1
,
p
p
.
5
4
5
-
5
4
9
,
2
0
0
0
.
[3
0
]
D.
M
ich
ie,
e
t
a
l.
,
“
M
a
c
h
in
e
L
e
a
rn
in
g
,
Ne
u
ra
l
a
n
d
S
tatisti
c
a
l
Clas
sif
ic
a
ti
o
n
,
”
El
li
s
Ho
rw
o
o
d
S
e
ri
e
s
in
A
rti
f
icia
l
In
telli
g
e
n
c
e
,
1
9
9
4
.
[3
1
]
“
Co
m
p
a
ra
ti
v
e
A
n
a
l
y
sis o
f
Clas
si
fica
ti
o
n
A
lg
o
rit
h
m
s f
o
r
th
e
P
re
d
ict
io
n
o
f
L
e
u
k
e
m
i
a
Ca
n
c
e
r
.”
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8708
Da
ta
Min
in
g
A
p
p
r
o
a
ch
o
f A
cc
id
en
t O
cc
u
r
r
en
ce
s
I
d
en
tifi
ca
tio
n
w
ith
E
ffectiv
e
.
..
(
Meen
u
Gu
p
ta
)
4041
[3
2
]
S
.
V
ij
a
y
a
ra
n
i
, “
Co
m
p
a
ra
ti
v
e
A
n
a
l
y
sis o
f
Ba
y
e
s an
d
L
a
z
y
Cl
a
ss
i
f
ica
ti
o
n
A
lg
o
rit
h
m
s
.”
[3
3
]
“
A
No
v
e
l
D
e
sig
n
S
p
e
c
if
ic
a
ti
o
n
Dista
n
c
e
(D
SD
)
Ba
se
d
K
-
M
e
a
n
Cl
u
ste
rin
g
P
e
rf
o
rm
a
c
e
Ev
a
lu
a
ti
o
n
o
n
E
n
g
in
e
e
rin
g
M
a
teria
ls'
Da
tab
a
se
.”
[3
4
]
“
A
S
u
rv
e
y
o
n
De
c
isio
n
T
re
e
Ba
s
e
d
A
p
p
ro
a
c
h
e
s in
Da
ta M
in
i
n
g
.”
Evaluation Warning : The document was created with Spire.PDF for Python.