T
E
L
K
O
M
N
I
K
A
T
elec
o
m
m
un
ica
t
io
n,
Co
m
pu
t
ing
,
E
lect
ro
nics
a
nd
Co
ntr
o
l
Vo
l.
18
,
No
.
5
,
Octo
b
er
2
0
2
0
,
p
p
.
2
3
7
8
~
2
3
8
4
I
SS
N:
1
6
9
3
-
6
9
3
0
,
ac
cr
ed
ited
First Gr
ad
e
b
y
Kem
en
r
is
tek
d
i
k
ti,
Dec
r
ee
No
: 2
1
/E/KPT
/2
0
1
8
DOI
: 1
0
.
1
2
9
2
8
/TE
L
KOM
NI
K
A.
v
1
8
i5
.
1
3
3
9
8
2378
J
o
ur
na
l
ho
m
ep
a
g
e
:
h
ttp
:
//jo
u
r
n
a
l.u
a
d
.
a
c.
id
/in
d
ex
.
p
h
p
/TELK
OM
N
I
K
A
Viterbi
optimiza
ti
o
n f
o
r cr
ime
de
te
ction a
nd id
enti
fi
ca
tion
Ree
m
Ra
zz
a
q Abdu
l H
us
s
e
i
n
1
,
Sa
lih
M
a
h
di Al
-
Q
a
ra
a
wi
2
,
M
ua
y
a
d Sa
dik
Cro
o
c
k
3
1
Co
ll
e
g
e
o
f
Bu
si
n
e
ss
In
fo
rm
a
ti
c
s,
Un
iv
e
rsit
y
o
f
In
f
o
rm
a
ti
o
n
Tec
h
n
o
lo
g
y
a
n
d
C
o
m
m
u
n
ica
ti
o
n
s,
Ira
q
2,
3
Co
m
p
u
ter E
n
g
i
n
e
e
rin
g
De
p
a
rt
m
e
n
t,
Un
i
v
e
rsity
o
f
Tec
h
n
o
l
o
g
y
,
Ira
q
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
u
n
2
4
,
2
0
1
9
R
ev
is
ed
Ma
r
3
,
2
0
2
0
Acc
ep
ted
Ma
r
1
3
,
2
0
2
0
In
t
h
is
p
a
p
e
r,
we
i
n
tro
d
u
c
e
two
t
y
p
e
s
o
f
h
y
b
ri
d
iza
ti
o
n
.
Th
e
first
c
o
n
tri
b
u
t
io
n
is
th
e
h
y
b
ri
d
iza
ti
o
n
b
e
twe
e
n
t
h
e
Viterb
i
a
lg
o
rit
h
m
a
n
d
Ba
u
m
Welc
h
in
o
rd
e
r
t
o
p
re
d
ict
c
rime
lo
c
a
ti
o
n
s
.
Wh
il
e
t
h
e
se
c
o
n
d
c
o
n
tri
b
u
ti
o
n
c
o
n
sid
e
rs
th
e
o
p
t
imiz
a
ti
o
n
b
a
se
d
o
n
d
e
c
isi
o
n
tree
(DT)
in
c
o
m
b
in
a
ti
o
n
wit
h
t
h
e
Viterb
i
a
lg
o
rit
h
m
fo
r
c
rimin
a
l
i
d
e
n
ti
f
ica
ti
o
n
u
sin
g
Ira
q
a
n
d
In
d
ia
c
rime
d
a
tas
e
t.
Th
is
wo
rk
is
b
a
se
d
o
n
o
u
r
p
re
v
i
o
u
s
wo
r
k
[1
].
T
h
e
m
a
in
g
o
a
l
is
t
o
e
n
h
a
n
c
e
th
e
re
su
lt
s
o
f
th
e
m
o
d
e
l
in
b
o
th
c
o
n
su
m
i
n
g
ti
m
e
s
a
n
d
t
o
g
e
t
a
m
o
re
a
c
c
u
ra
te
m
o
d
e
l.
Th
e
o
b
tain
e
d
re
su
lt
s
p
ro
v
e
d
t
h
e
a
c
h
iev
e
m
e
n
t
o
f
b
o
th
g
o
a
ls
i
n
a
n
e
fficie
n
t
wa
y
.
K
ey
w
o
r
d
s
:
B
au
m
W
elch
alg
o
r
ith
m
Dec
is
io
n
tr
ee
Hy
b
r
id
izatio
n
Viter
b
i a
lg
o
r
ith
m
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
:
R
ee
m
R
az
za
q
Ab
d
u
l H
u
s
s
ein
,
C
o
lleg
e
o
f
B
u
s
in
ess
I
n
f
o
r
m
ati
cs,
Un
iv
er
s
ity
o
f
I
n
f
o
r
m
atio
n
T
ec
h
n
o
lo
g
y
an
d
C
o
m
m
u
n
icatio
n
s
,
B
ag
h
d
ad
,
I
r
aq
.
E
m
ail:
r
ee
m
r
az
za
k
@
y
ah
o
o
.
co
m
1.
I
NT
RO
D
UCT
I
O
N
T
h
e
d
ata
m
in
in
g
tec
h
n
iq
u
es
a
r
e
wid
ely
u
s
ed
in
e
-
g
o
v
er
n
m
en
t,
esp
ec
ially
in
th
e
cr
im
in
o
lo
g
y
f
ield
,
wh
er
ein
th
ey
h
elp
p
o
lice
d
ep
ar
tm
en
ts
p
r
ed
icate
cr
im
in
als
an
d
d
etec
t
in
f
o
r
m
atio
n
ab
o
u
t
cr
im
e
lo
ca
tio
n
s
.
T
h
e
s
tu
d
y
o
f
cr
im
e
is
ex
p
ec
te
d
n
o
t
o
n
ly
to
c
o
n
tr
o
l
p
r
es
en
t
cr
im
e
b
u
t
also
to
a
n
aly
ze
th
e
cr
im
in
al
ac
tiv
ities
so
th
at
f
u
tu
r
e
o
cc
u
r
r
en
ce
s
o
f
th
e
s
am
e
in
ci
d
en
ts
ca
n
b
e
o
v
e
r
co
m
e.
Go
v
e
r
n
m
en
t
is
t
r
y
in
g
to
im
p
r
o
v
e
th
e
ef
f
ec
tiv
en
ess
o
f
p
r
ev
e
n
tio
n
o
f
cr
im
es
th
at
ca
n
b
e
h
a
p
p
en
ed
in
clu
s
ter
in
g
cr
im
i
n
al
f
ea
tu
r
es
d
ep
en
d
i
n
g
o
n
s
p
atial
an
d
tem
p
o
r
al
cr
it
er
ia
wh
ich
ar
e
d
if
f
er
en
t
in
ea
ch
co
u
n
tr
y
[1
-
1
1
]
.
I
t
also
tr
ies
to
d
is
co
v
er
th
e
co
r
r
elatio
n
b
etwe
en
d
if
f
e
r
e
n
t
attr
ib
u
tes,
s
u
ch
as
cr
im
e
an
d
d
em
o
g
r
ap
h
ics
.
An
aly
zin
g
h
is
to
r
ical
d
ata
is
u
s
ef
u
l
to
d
is
co
v
er
th
e
cr
im
e
p
atter
n
[
1
2
]
.
Pas
t
r
esear
ch
e
r
s
in
cr
im
e
d
etec
tio
n
d
e
p
en
d
e
d
o
n
d
ata
m
in
in
g
an
d
m
ac
h
in
e
lear
n
in
g
to
d
ev
elo
p
s
y
s
tem
s
th
at
an
aly
ze
p
atter
n
s
to
h
av
e
th
e
h
ig
h
est
p
r
o
b
a
b
ilit
y
o
f
o
cc
u
r
r
in
g
a
n
d
p
r
e
d
ict
th
e
r
eg
io
n
s
wh
er
e
s
u
ch
p
atter
n
s
ar
e
lik
ely
t
o
h
ap
p
e
n
.
So
m
e
r
e
s
ea
r
ch
er
s
u
tili
ze
d
th
e
alg
o
r
ith
m
o
f
B
au
m
-
W
elch
in
p
ast
s
tu
d
ies,
wh
ich
co
llecte
d
GPS
d
ata
f
o
r
p
r
ed
ictin
g
cr
im
in
al
m
o
v
em
en
ts
.
T
h
e
wo
r
k
o
f
[
1
3
]
u
s
ed
attr
ib
u
tes
ac
co
r
d
in
g
to
th
eir
c
h
ar
ac
ter
is
tics
an
d
tr
ain
ed
a
n
HM
M
f
o
r
ea
ch
clu
s
ter
,
o
b
tain
in
g
an
ac
c
u
r
a
cy
r
esu
l
t
o
f
1
3
.
8
5
%.
T
h
e
au
th
o
r
s
s
u
g
g
ested
d
ata
m
in
in
g
m
eth
o
d
s
f
o
r
cr
im
e
d
e
tectio
n
an
d
i
d
en
tific
atio
n
o
f
cr
im
in
al
f
o
r
I
n
d
ia
n
co
u
n
tr
y
d
u
r
in
g
th
e
p
er
i
o
d
o
f
2
0
0
0
to
2
0
1
2
.
T
h
e
y
d
ep
e
n
d
ed
o
n
clu
s
ter
in
g
,
u
s
ed
k
-
m
ea
n
s
,
ar
e
b
ased
o
n
cr
im
e
attr
ib
u
tes,
an
d
m
ad
e
v
is
u
aliza
tio
n
s
u
s
in
g
Go
o
g
le
Ma
p
s
.
T
h
e
y
also
u
s
ed
ANNs
an
d
f
o
r
est
d
ec
is
io
n
tr
ee
s
(
DT
s
)
f
o
r
class
if
icatio
n
.
T
h
e
o
u
tco
m
es we
r
e
co
m
p
u
ted
b
y
th
e
W
ek
a
to
o
l.
T
h
e
ac
cu
r
ac
y
o
f
th
e
A
NN
was 9
0
.
0
2
.
T
h
ey
ap
p
lied
1
0
-
f
o
l
d
cr
o
s
s
v
alid
ati
o
n
f
o
r
th
e
f
o
r
est
DT
s
,
th
e
ac
c
u
r
ac
y
in
R
MSE
was
0
.
0
7
5
1
,
a
n
d
th
e
tim
e
tak
e
n
to
b
u
ild
th
e
m
o
d
el
was
4
.
2
4
s
[
1
4
]
.
I
n
[
1
5
]
,
th
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
h
el
p
ed
ag
en
cies
to
em
p
h
asize
th
e
s
ec
u
r
ity
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
KA
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l Co
n
tr
o
l
V
iter
b
i o
p
timiz
a
tio
n
fo
r
crime
d
etec
tio
n
a
n
d
id
en
tifi
ca
tio
n
(
R
ee
m
R
a
z
z
a
q
A
b
d
u
l H
u
s
s
ein
)
2379
in
I
n
d
ia
n
,
th
eir
wo
r
k
n
ee
d
s
to
i
m
p
r
o
v
e
with
m
o
r
e
d
ata
m
in
in
g
m
eth
o
d
.
Oth
er
r
esear
ch
er
s
c
o
n
s
id
er
ed
t
h
e
u
s
e
o
f
th
e
Viter
b
i
al
g
o
r
ith
m
f
o
r
i
d
en
tify
in
g
lo
ca
tio
n
s
with
a
h
i
g
h
p
r
o
b
ab
ilit
y
o
f
h
av
in
g
a
cr
im
in
al
d
e
p
en
d
in
g
o
n
th
e
r
elatio
n
s
h
ip
b
etwe
en
t
h
e
lo
ca
tio
n
an
d
ty
p
es
o
f
cr
im
es,
s
u
ch
as
m
u
r
d
er
,
t
h
ief
,
a
n
d
ass
au
lt,
th
ey
h
an
d
le
id
lin
g
tim
e
r
ath
er
th
an
ac
cu
r
ac
y
p
r
o
b
lem
.
T
h
e
L
aten
t
h
id
d
e
n
Ma
r
k
o
v
m
o
d
els ar
e
ty
p
es o
f
al
g
o
r
ith
m
s
th
at
h
av
e
b
ee
n
d
esig
n
e
d
to
d
etec
t
cr
i
m
e
ac
tiv
ities
b
y
o
b
tain
in
g
a
s
eq
u
en
ce
o
f
o
b
s
er
v
atio
n
s
f
r
o
m
h
id
d
e
n
v
alu
es.
T
h
ey
d
e
v
elo
p
ed
an
alg
o
r
ith
m
th
at
f
its
r
eg
u
lar
v
i
n
e
co
p
u
la
to
g
en
er
ate
tr
ee
s
tr
u
ctu
r
es.
T
h
ese
tr
ee
s
h
ad
b
ee
n
em
p
lo
y
ed
t
o
p
r
o
d
u
ce
a
n
em
is
s
io
n
m
atr
ix
f
o
r
h
id
d
e
n
Ma
r
k
o
v
m
o
d
el
(
HM
M)
alg
o
r
ith
m
,
t
h
e
f
u
s
io
n
o
f
c
o
u
p
le
d
p
ar
am
eter
s
with
two
ty
p
es
o
f
HM
M
alg
o
r
ith
m
s
,
th
is
wo
r
k
n
ee
d
s
to
im
p
r
o
v
e
ac
cu
r
ac
y
a
n
d
r
ed
u
ce
ex
ec
u
tio
n
tim
e.
[
1
]
.
I
n
th
is
wo
r
k
,
we
attem
p
t to
o
v
e
r
co
m
e
t
h
e
p
r
o
b
lem
s
in
p
r
ev
i
o
u
s
s
tu
d
ies an
d
d
e
v
el
o
p
th
e
wo
r
k
o
f
p
ast
r
esear
ch
er
[
1
,
3
,
1
4
]
,
b
y
p
r
o
p
o
s
in
g
a
h
y
b
r
id
izatio
n
b
etwe
e
n
B
au
m
W
elch
alg
o
r
ith
m
an
d
Viter
b
i
alg
o
r
ith
m
s
in
a
s
id
e,
a
n
d
Viter
b
i
alg
o
r
ith
m
s
f
o
r
o
p
tim
izatio
n
o
f
d
ata
s
e
t
th
en
a
p
p
lied
DT
,
at
th
e
o
th
er
s
id
e.
T
h
e
o
b
jectiv
e
o
f
p
r
o
p
o
s
ed
wo
r
k
to
h
an
d
les
a
n
im
p
o
r
tan
t
to
p
ic
o
f
m
o
n
ito
r
in
g
cr
im
in
al
ac
tiv
ities
an
d
m
o
v
e
m
en
ts
an
d
in
d
icate
s
th
e
lev
el
o
f
d
an
g
e
r
in
lo
ca
tio
n
s
,
th
e
p
r
ep
o
s
d
h
y
b
r
id
izatio
n
m
eth
o
d
r
e
d
u
ce
s
th
e
c
o
n
s
u
m
ed
tim
e
an
d
in
c
r
ea
s
es
ac
cu
r
ac
y
,
r
esp
ec
tiv
ely
.
2.
H
M
M
T
h
e
HM
M
is
a
r
o
b
u
s
t
m
o
d
el
wh
ich
ca
n
b
e
u
s
ed
wh
e
n
s
tat
es
in
a
p
r
o
c
ess
ar
e
n
o
t
o
b
s
er
v
ab
le,
b
u
t
o
b
s
er
v
ed
d
ata
d
ep
e
n
d
s
o
n
th
ese
h
id
d
en
s
tates.
HM
M
d
ep
en
d
s
o
n
two
m
ain
p
r
o
p
e
r
ties
,
wh
ich
ar
e:
−
T
h
e
o
b
s
er
v
atio
n
at
tim
e
t
is
p
r
o
d
u
ce
d
b
y
a
p
r
o
ce
s
s
wh
o
s
e
s
tate
H
t
is
h
id
d
en
f
r
o
m
th
e
o
b
s
er
v
er
.
−
T
h
e
s
tate
o
f
th
e
h
i
d
d
en
p
r
o
ce
s
s
r
ep
r
esen
ts
th
e
Ma
r
k
o
v
ch
ain
[
1
6
-
2
3
]
.
At
th
e
o
t
h
er
h
an
d
,
th
e
Viter
b
i
is
a
d
y
n
am
ic
p
r
o
g
r
a
m
m
in
g
a
lg
o
r
ith
m
th
at
d
ep
en
d
s
o
n
tr
an
s
itio
n
an
d
em
is
s
io
n
m
atr
ices.
I
n
p
ar
ticu
l
ar
,
th
e
Viter
b
i
alg
o
r
ith
m
ca
n
o
b
tain
p
ath
(
s
tate
s
eq
u
en
ce
s
)
to
g
en
er
ate
o
u
tp
u
t
s
eq
u
en
ce
s
.
I
t
wo
r
k
s
b
y
f
in
d
in
g
th
e
m
a
x
im
u
m
o
v
e
r
all
p
o
s
s
ib
le
s
tate
s
eq
u
en
ce
by
c
o
n
s
id
er
in
g
a
f
o
r
war
d
in
g
alg
o
r
ith
m
.
T
h
e
B
au
m
(
1
9
7
0
)
p
r
o
p
o
s
ed
th
e
u
s
e
o
f
th
e
B
au
m
–
W
elch
alg
o
r
ith
m
b
ased
o
n
th
e
co
m
p
u
tatio
n
o
f
th
e
p
r
o
b
ab
ilis
tic
m
eth
o
d
s
o
f
th
e
Ma
r
k
o
v
m
o
d
el.
I
t
also
ca
lled
t
h
e
f
o
r
war
d
-
b
ac
k
war
d
m
eth
o
d
,
wh
er
ei
n
th
e
b
ac
k
war
d
p
ar
t r
ep
r
esen
ted
th
e
p
r
o
b
a
b
ilit
y
o
f
p
a
r
tial o
b
s
er
v
atio
n
s
eq
u
en
ce
s
f
r
o
m
th
e
tim
e
(
t+
1)
to
en
d
,
ca
n
b
e
co
m
p
u
ted
iter
ativ
ely
[
2
4
]
.
3.
O
VE
RVI
E
W
O
F
ADO
P
T
E
D
DATA
SE
T
S
I
n
th
is
wo
r
k
,
two
d
atasets
h
av
e
b
ee
n
u
s
ed
in
th
e
cr
im
in
al
f
iel
d
.
T
h
e
f
ir
s
t
o
n
e
is
th
e
I
r
aq
i
d
at
aset
,
wh
ile
th
e
s
ec
o
n
d
d
ataset
is
I
n
d
ia
d
at
aset [
2
5
]
.
T
h
e
I
r
aq
d
ataset
co
n
s
is
t
s
o
f
f
ea
tu
r
es
s
u
ch
as
as{
ag
e
,
g
en
d
e
r
,
I
D,
cr
im
e
ty
p
es,
lo
ca
tio
n
s
,
g
an
g
,
lo
n
g
itu
d
e,
latitu
d
e
}
t
h
e
ty
p
e
o
f
f
ea
t
u
r
es
is
ca
teg
o
r
izatio
n
w
h
er
ea
s
th
e
f
ea
tu
r
es
o
f
I
n
d
ian
d
ata
s
et
is
in
clu
d
ed
{states
,
m
u
r
d
er
,
attem
p
ted
t
o
m
u
r
d
er
,
…,
th
ief
},
m
ak
in
g
a
p
a
r
titi
o
n
in
g
b
ased
o
n
clu
s
ter
in
g
n
ee
d
s
m
o
r
e
in
v
esti
g
atio
n
an
d
a
n
aly
s
is
o
f
th
e
co
n
te
x
ts
in
a
d
is
c
r
ete
s
eq
u
en
ce
d
ataset.
Her
e,
th
e
n
ee
d
e
d
to
p
r
e
d
ict
th
e
lo
ca
tio
n
s
o
f
th
e
cr
im
es
,
an
d
th
e
HM
M
to
o
b
tain
th
e
h
id
d
en
p
atter
n
is
r
eq
u
ir
ed
.
As
m
en
tio
n
ed
ea
r
lier
,
th
e
d
ataset
is
co
llected
f
r
o
m
a
web
s
ite
an
d
s
o
cial
m
ed
ia,
wh
er
e
B
ag
h
d
ad
city
ca
n
b
e
d
iv
id
e
d
in
to
d
if
f
er
en
t
m
ain
lo
ca
tio
n
s
as sh
o
wn
in
p
r
o
p
o
s
e
d
T
ab
le
1
.
W
h
ile
th
e
r
ep
r
esen
tatio
n
o
f
th
e
s
ec
o
n
d
d
ataset
is
s
h
o
wn
in
p
r
o
p
s
ed
T
ab
le
2
.
T
h
e
p
r
o
p
o
s
ed
s
y
s
tem
is
also
ap
p
lied
t
o
I
n
d
ia
d
ataset.
As
m
en
tio
n
ed
ab
o
v
e,
th
e
HM
M
m
u
s
t
g
en
e
r
ate
ten
lo
ca
tio
n
s
:
L
o
c1
,
L
o
c2
,
….
,
L
oc
32
,
wh
er
e
th
ese
lo
ca
tio
n
s
r
e
p
r
esen
t
th
e
wo
r
k
f
lo
w
o
f
th
e
p
r
o
p
o
s
ed
s
y
s
tem
in
th
e
clu
s
ter
in
g
.
Her
e,
th
e
n
ee
d
e
d
to
p
r
ed
ict
th
e
lo
ca
tio
n
s
o
f
t
h
e
cr
im
es
,
an
d
th
e
HM
M
to
o
b
tain
th
e
h
id
d
en
p
atter
n
in
th
is
wo
r
k
as
well
as
th
e
t
wo
h
id
d
en
s
tates with
o
n
e
o
b
s
er
v
e
th
e
s
tate.
T
ab
le
1
.
L
o
ca
tio
n
o
f
I
r
a
q
r
ep
r
esen
tatio
n
T
ab
le
2
.
L
o
ca
tio
n
o
f
I
n
d
ian
s
tates r
ep
r
esen
tatio
n
S
e
q
Lo
c
a
t
i
o
n
R
e
p
r
e
se
n
t
a
t
i
o
n
1
A
d
a
m
y
a
Lo
c
a
t
i
o
n
1
(
L
o
c
_
1
)
2
A
l
b
a
y
a
a
Lo
c
a
t
i
o
n
2
(
L
o
c
_
2
)
3
H
a
y
a
l
a
m
e
l
Lo
c
a
t
i
o
n
3
(
L
o
c
_
3
)
4
k
a
d
mi
y
a
Lo
c
a
t
i
o
n
4
(
L
o
c
_
4
)
5
M
a
d
a
a
n
Lo
c
a
t
i
o
n
5
(
L
o
c
_
5
)
6
M
a
n
s
o
u
r
Lo
c
a
t
i
o
n
6
(
L
o
c
_
6
)
7
B
a
g
h
d
a
d
g
e
d
e
e
d
a
Lo
c
a
t
i
o
n
7
(
L
o
c
_
7
)
8
H
a
san
i
a
Lo
c
a
t
i
o
n
8
(
L
o
c
_
8
)
9
M
a
d
e
e
n
a
Lo
c
a
t
i
o
n
9
(
L
o
c
_
9
)
10
A
b
o
g
r
e
e
b
Lo
c
a
t
i
o
n
1
0
(
Lo
c
_
1
0
)
st
a
t
e
R
e
p
r
e
se
n
t
a
t
i
o
n
1
A
N
D
H
R
A
P
R
A
D
ESH
Lo
c
a
t
i
o
n
1
(
L
o
c
_
1
)
2
A
R
U
N
A
C
H
A
L
P
R
A
D
ESH
Lo
c
a
t
i
o
n
2
(
L
o
c
_
2
)
3
A
S
S
A
M
Lo
c
a
t
i
o
n
3
(
L
o
c
_
3
)
4
B
I
H
A
R
Lo
c
a
t
i
o
n
4
(
L
o
c
_
4
)
5
C
H
H
A
TTI
S
G
A
R
H
Lo
c
a
t
i
o
n
5
(
L
o
c
_
5
)
.
.
.
.
32
P
U
D
U
C
H
ER
R
Y
Lo
c
a
t
i
o
n
3
2
(
L
o
c
_
3
2
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6
9
3
0
T
E
L
KOM
NI
KA
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l Co
n
tr
o
l
,
Vo
l.
18
,
No
.
5
,
Octo
b
e
r
2
0
2
0
:
2
3
7
8
-
2384
2380
4.
P
RO
P
O
SE
D
M
E
T
H
O
D
T
h
e
Viter
b
i
alg
o
r
ith
m
as
an
o
p
tim
izer
alg
o
r
ith
m
in
two
s
tr
a
teg
ies
is
p
r
o
p
o
s
ed
.
T
h
e
f
ir
s
t
o
n
e
is
b
u
ilt
with
B
au
m
W
elch
,
wh
ile
t
h
e
s
ec
o
n
d
o
n
e
h
as
b
ee
n
cr
ea
ted
i
n
co
m
b
in
atio
n
with
th
e
DT
al
g
o
r
ith
m
.
Viter
b
i
is
a
s
u
p
er
v
is
ed
clu
s
ter
in
g
alg
o
r
ith
m
th
at
m
a
k
es
p
ar
t
itio
n
in
g
-
b
a
s
ed
clu
s
ter
in
g
f
o
r
lear
n
i
n
g
c
o
n
tex
ts
in
a
d
is
cr
ee
t
s
eq
u
en
ce
d
ataset.
An
d
it
wo
r
k
s
as
an
o
u
tlier
d
etec
to
r
th
at
r
em
o
v
es
lo
w
-
p
r
o
b
ab
ilit
y
d
ata
(
a
b
n
o
r
m
al
d
ata)
th
at
ex
p
lain
s
in
d
etail
th
e
f
o
llo
win
g
s
ec
tio
n
.
4
.
1
.
H
y
bridi
za
t
io
n Vit
er
bi
a
lg
o
rit
hm
wit
h
B
a
um
Welch
W
e
ex
p
lain
th
e
elem
en
ts
o
f
th
e
p
r
o
p
o
s
ed
wo
r
k
f
l
o
w
th
at
in
clu
d
e
th
e
h
y
b
r
id
izatio
n
o
f
th
e
Viter
b
i
alg
o
r
ith
m
with
B
au
m
W
elch
t
o
ap
p
ea
r
in
th
r
ee
s
tag
es
as
s
h
o
wn
in
Fig
u
r
e
1
.
T
h
e
Viter
b
i
p
lay
s
a
m
ajo
r
r
o
le
in
en
h
an
cin
g
th
e
clu
s
ter
in
g
m
o
d
el;
in
ad
d
itio
n
,
th
e
d
atas
et
f
o
r
m
u
lates
th
e
s
tr
u
ctu
r
e
o
f
th
e
d
y
n
am
ic
m
o
d
el
u
s
in
g
v
in
e
wh
ile
th
e
m
o
v
em
e
n
t
o
f
cr
im
in
als
is
d
eter
m
in
e
d
b
y
th
e
g
e
n
er
atio
n
o
f
d
y
n
am
i
c
tr
an
s
itio
n
m
atr
ix
f
r
o
m
d
ataset.
.
Fig
u
r
e
1
.
Hy
b
r
id
izatio
n
Viter
b
i a
lg
o
r
ith
m
with
B
au
m
W
elch
4
.
1
.
1
.
E
m
is
s
io
n
m
a
t
ri
x
g
ener
a
t
io
n sta
g
e
T
h
e
s
tr
u
ctu
r
e
of
m
atr
i
x
is
g
e
n
er
ated
u
s
in
g
a
s
im
p
le
v
in
e
c
o
u
p
le
to
p
r
o
d
u
ce
a
s
im
p
le
tr
e
e
s
tr
u
ctu
r
e
s
tr
ateg
y
f
o
r
s
elec
tin
g
a
tr
ee
m
o
d
el
b
ased
o
n
v
in
e
alg
o
r
ith
m
an
d
g
e
n
er
atin
g
a
co
m
p
ac
t
m
o
d
el.
C
o
n
d
itio
n
al
p
r
o
b
a
b
ilit
y
(
B
ay
es
th
eo
r
em
)
is
ap
p
lied
to
th
e
t
r
ee
s
tr
u
ctu
r
e
an
d
u
s
ed
f
o
r
ass
u
m
in
g
th
at
th
e
co
u
p
le
p
ar
am
ete
r
{m
o
n
th
,
lo
ca
tio
n
}
is
co
m
b
in
e
d
with
c
r
im
e
t
y
p
es
to
p
r
o
d
u
c
e
th
r
ee
d
im
en
s
io
n
s
r
ath
er
th
a
n
two
d
im
en
s
io
n
s
.
T
h
e
E
m
is
s
io
n
m
atr
ix
ca
n
b
e
em
b
ed
d
e
d
in
to
HM
M
(
B
au
m
-
W
elch
an
d
Viter
b
i
al
g
o
r
it
h
m
s
)
an
d
ev
alu
ated
u
s
in
g
i
n
(
1
)
,
E
m
i
=
[
(
1
/
,
)
⋯
(
N
/
,
)
⋮
⋱
⋮
(
1
/
ℎ
,
)
⋯
(
N
/
ℎ
,
)
]
(
1
)
4
.
1
.
2
.
T
ra
ns
it
io
n
m
a
t
ri
x
g
en
er
a
t
io
n sta
g
e
T
h
e
t
r
an
s
itio
n
m
atr
ix
is
u
s
ed
to
r
ep
r
esen
t
th
e
Ma
r
k
o
v
c
h
ain
,
an
d
it
is
d
ef
in
ed
as
a
s
et
o
f
(
lo
ca
tio
n
)
s
tates
(
S
=
{
s1
_
,
s
2
.
.
.
,
s
n
})
,
wh
er
e
L
o
ca
tio
n
s
ar
e
r
ep
r
esen
te
d
b
y
cr
im
in
al
m
o
v
em
en
ts
in
a
d
if
f
er
en
t
lo
ca
tio
n
.
T
h
e
tr
an
s
itio
n
is
r
ep
r
esen
ted
as
th
e
s
q
u
ar
e
ar
r
ay
T
N×
N
,
wh
er
e
N=
1
0
,
f
o
r
I
r
a
q
d
ataset,
an
d
N
=3
2
f
o
r
I
n
d
ia
d
ataset.
T
h
e
tr
an
s
itio
n
m
atr
i
x
ca
n
b
e
c
o
n
s
tr
u
ctio
n
as
th
e
f
o
llo
win
g
al
g
o
r
ith
m
,
wh
er
e
th
e
s
u
m
m
atio
n
o
f
t
h
e
p
r
o
b
a
b
ilit
y
o
f
ea
ch
r
o
w
in
th
e
m
atr
ix
m
u
s
t b
e
an
e
q
u
al
o
n
e.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
KA
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l Co
n
tr
o
l
V
iter
b
i o
p
timiz
a
tio
n
fo
r
crime
d
etec
tio
n
a
n
d
id
en
tifi
ca
tio
n
(
R
ee
m
R
a
z
z
a
q
A
b
d
u
l H
u
s
s
ein
)
2381
An algorithm of Transition Matrix
Function
Transition (x), where x is loc
ation of crime
Returns
Probability matrix of criminal’s movements
m = maximum value(x);
y = zeros(m,
1)
\
\
vertice
construct probability
matrix p
m×m
for k=1to n
-
1
y(x(k)) = y(x(k)) + 1;
p(x(k
), x
(k+1)) = p(x(k
), x
(k+1)) + 1;
w
here
k
is the number of movements
end loop
if y ==0 then
p=0
else
p
= (p
div y);
end if
4
.
1
.
3
.
Sequ
ence
g
ener
a
t
io
n sta
g
e
T
h
e
s
eq
u
en
ce
is
th
e
th
ir
d
p
ar
am
eter
o
f
HM
M,
in
th
is
s
tep
is
fr
i
s
t
co
n
tr
ib
u
tio
n
wh
ich
d
ef
in
es
th
e
s
eq
u
en
ce
o
f
c
r
im
e
th
at
h
a
p
p
en
s
,
to
o
b
tain
th
e
m
o
s
t
p
r
o
b
ab
le
lo
ca
tio
n
o
f
(
cr
im
es
)
s
tates.
I
n
tr
ad
itio
n
al
an
d
p
r
o
p
o
s
al
wo
r
k
,
th
e
s
witch
h
as b
ee
n
m
ad
e
b
etwe
en
s
tates a
n
d
s
eq
u
en
ce
s
(
cr
im
es)
in
two
im
p
o
r
tan
t p
h
ases
:
−
R
ep
lacin
g
th
e
s
eq
u
en
ce
v
alu
e
with
a
s
tate
v
alu
e
,
(
i.e
.
th
e
s
e
q
u
en
ce
as
co
u
p
led
p
ar
am
eter
s
{
cr
im
e
t
y
p
es
}
is
r
ep
lace
d
with
{lo
ca
tio
n
s
}.
−
I
n
s
er
tin
g
s
tates
(
lo
ca
tio
n
s
)
as
a
s
eq
u
en
ce
in
to
t
h
e
Viter
b
i
al
g
o
r
ith
m
to
g
en
er
ate
th
e
m
o
s
t
p
r
o
b
ab
le
s
eq
u
e
n
ce
o
f
cr
im
es a
s
an
o
u
tp
u
t.
4
.
2
.
DT
a
lg
o
rit
hm
s
wit
h Vit
er
bi
a
lg
o
rit
hm
s
Her
e,
th
e
s
ec
o
n
d
s
tr
ateg
y
o
f
i
m
p
r
o
v
i
n
g
d
ec
is
io
n
tr
ee
alg
o
r
i
th
m
is
ex
p
lain
ed
.
T
h
e
ad
o
p
ted
s
tep
s
o
f
th
e
p
r
o
p
o
s
ed
alg
o
r
ith
m
ar
e
illu
s
tr
ated
in
Fig
u
r
e
2
f
o
llo
win
g
th
e
s
tep
s
o
f
:
Step
1
:
Star
t.
Step
2
:
Gen
er
ate
in
itialize
th
e
v
alu
e
o
f
th
e
E
I
MS
m
atr
ix
as
m
en
ti
o
n
in
s
ec
tio
n
2
th
at
co
m
p
u
tes
th
e
p
r
o
b
ab
ilit
y
b
etwe
en
{c
r
im
es ty
p
e,
m
o
n
th
}
an
d
{lo
ca
tio
n
s
}.
Step
3
:
Gen
er
ate
in
itialize
th
e
v
alu
e
o
f
th
e
T
R
NS
m
atr
ix
,
in
th
e
s
am
e
way
,
th
at
m
e
n
tio
n
in
s
ec
tio
n
,
to
d
eter
m
in
e
cr
im
in
al
’
s
m
o
v
em
e
n
ts
.
Step
4
:
A
s
eq
u
en
ce
o
f
c
o
u
p
l
e
p
ar
am
et
er
in
itializatio
n
.
Fig
u
r
e
2
.
DT
alg
o
r
ith
m
s
with
Viter
b
i
alg
o
r
ith
m
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6
9
3
0
T
E
L
KOM
NI
KA
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l Co
n
tr
o
l
,
Vo
l.
18
,
No
.
5
,
Octo
b
e
r
2
0
2
0
:
2
3
7
8
-
2384
2382
Step
5
:
ap
p
ly
t
h
e
Viter
b
i
alg
o
r
ith
m
wh
ich
ca
n
b
e
u
s
ed
as
an
o
p
ti
m
izatio
n
alg
o
r
ith
m
b
y
th
e
elim
in
atio
n
o
f
th
e
lo
ca
tio
n
with
lo
w
p
r
o
b
ab
ili
ty
as
an
o
u
tlier
to
m
ax
im
ize
t
o
in
cr
ea
s
e
th
e
e
f
f
icien
cy
o
f
tim
e
ex
ec
u
tio
n
,
Step
6
:
R
em
o
v
e
lo
ca
tio
n
s
with
lo
w
p
r
o
b
ab
ilit
y
f
r
o
m
th
e
d
ataset.
Step
7
:
Gen
er
ate
n
ew
d
ataset,
with
im
p
r
o
v
e
d
q
u
ality
o
f
th
e
d
ataset.
Step
8
:
Ap
p
ly
Dec
is
io
n
T
r
ee
alg
o
r
ith
m
with
h
y
p
er
p
ar
a
m
eter
,
o
b
tai
n
in
g
Nid
as
a
lab
el
f
o
r
f
ir
s
t
d
ataset,
an
d
k
id
n
ap
p
in
g
as a
lab
el
f
o
r
th
e
s
ec
o
n
d
d
ataset.
Step
9
:
E
n
d
5.
CO
NT
RI
B
U
T
I
O
N
S
T
h
e
Vin
e
C
o
p
u
la
B
au
m
-
W
elch
alg
o
r
ith
m
ca
n
ac
h
iev
e
g
o
o
d
r
esu
lts
with
th
e
h
y
b
r
id
Vin
e
C
o
p
u
la
Viter
b
i
alg
o
r
ith
m
f
o
r
d
eter
m
i
n
in
g
t
h
e
s
eq
u
e
n
tial
r
elatio
n
o
f
p
ast
cr
im
e
ty
p
es,
d
ate,
a
n
d
l
o
ca
tio
n
s
.
T
h
e
s
p
ee
d
o
f
ex
e
cu
tio
n
an
d
ac
cu
r
ac
y
o
f
th
e
B
au
m
-
W
elch
alg
o
r
ith
m
ar
e
en
h
a
n
ce
d
Nid
o
f
I
r
a
q
ca
n
h
elp
to
f
ac
ilit
ate
th
e
id
en
tific
atio
n
o
f
cr
im
in
als
b
y
th
e
p
o
lice
an
d
ch
ec
k
p
o
in
ts
.
T
h
e
co
n
tr
i
b
u
tio
n
o
f
th
is
wo
r
k
,
in
clu
d
in
g
a
d
ataset,
is
o
p
tim
ized
u
s
in
g
a
Viter
b
i
alg
o
r
ith
m
f
o
r
o
u
tlier
d
etec
tio
n
an
d
g
e
n
er
ates
a
n
ew
d
ataset.
wh
ich
is
u
s
ed
to
g
en
er
ate
DT
.
6.
RE
SU
L
T
AND
I
M
P
L
E
M
E
N
T
AT
I
O
N
T
h
e
im
p
lem
en
tatio
n
o
f
th
e
p
r
o
p
o
s
ed
alg
o
r
ith
m
in
b
o
t
h
s
tr
ateg
ies
ca
n
b
e
ex
p
lain
ed
in
n
u
m
er
o
u
s
s
tag
es
:
Stag
e
1
:
Gen
er
ate
th
e
E
m
is
s
io
n
m
atr
ix
,
E
ac
h
v
alu
e
o
f
th
e
p
r
o
b
ab
ilit
y
m
atr
ix
r
elate
d
to
cr
im
e
ty
p
e
in
a
s
p
ec
if
ic
lo
ca
tio
n
.
T
h
e
m
ain
lo
ca
tio
n
s
ca
n
th
en
b
e
d
eter
m
in
e
d
f
r
o
m
th
e
co
lu
m
n
o
f
c
r
im
e
ty
p
e
af
ter
tr
an
s
p
o
s
itio
n
,
as
m
en
tio
n
E
m
is
s
io
n
m
atr
ix
in
itially
I
t
was
b
u
ilt
b
y
th
e
im
p
lem
en
tatio
n
o
f
th
e
th
e
o
r
y
o
f
B
ay
es,
th
en
was
tr
ain
ed
u
s
in
g
laten
t
Ma
r
k
o
v
m
eth
o
d
,
t
h
e
I
m
p
lem
en
tatio
n
o
f
g
en
e
r
ated
m
atr
ix
o
f
th
e
f
ir
s
t
d
ataset
is
s
h
o
wn
in
T
ab
le
3
.
W
h
ile
th
e
I
m
p
lem
en
tatio
n
o
f
g
en
er
ated
E
m
is
s
io
n
m
atr
ix
o
f
th
e
s
ec
o
n
d
d
ataset
is
s
h
o
wn
in
T
ab
le
4
.
T
h
e
two
f
ea
tu
r
es
s
elec
ted
{
th
ief
,
o
th
er
I
PC
cr
im
es
},
th
e
p
o
r
o
b
ab
ilty
cr
im
e
o
cc
r
in
g
ap
p
er
i
n
g
i
n
(
3
2
)
s
tates
in
in
d
ain
co
u
n
tr
y
,
as
m
en
tio
n
E
m
is
s
io
n
m
atr
ix
in
itially
I
t
was
b
u
ilt
b
y
B
ay
es
th
o
r
y
,
th
en
was
tr
ain
ed
u
s
in
g
laten
t
Ma
r
k
o
v
m
eth
o
d
.
Fin
ally
,
T
h
e
E
m
is
s
io
n
m
atr
ix
h
ad
b
ee
n
c
o
n
s
tr
u
cted
with
th
r
ee
p
ar
am
eter
s
wh
ich
h
ad
th
e
m
ain
r
o
le
HM
M
alg
o
r
ith
m
s
.
T
ab
le
3
.
T
h
e
im
p
lem
en
tatio
n
em
is
s
io
n
m
atr
ix
o
f
th
e
f
r
is
t d
a
taset
Lo
c
1
Lo
c
2
Lo
c
3
Lo
c
1
0
Th
e
f
t
0
.
1
2
5
0
0
0
…
0
M
u
r
d
e
r
0
.
3
3
3
3
0
0
…
0
O
t
h
e
r
s
0
.
5
0
0
0
0
0
…
0
Th
e
f
t
0
.
4
2
8
6
0
.
2
8
5
7
0
…
0
M
u
r
d
e
r
0
.
0
2
1
3
0
.
4
8
9
4
0
.
1
2
7
7
…
0
.
2
3
4
0
O
t
h
e
r
s
0
1
.
0
0
0
0
0
…
0
T
ab
le
4
.
E
m
is
s
io
n
m
atr
ix
o
f
t
h
e
s
ec
o
n
d
d
ataset
l
o
c
1
l
o
c
2
l
o
c
3
…….
l
o
c
3
2
Th
i
e
f
0
.
0
5
3
8
0
.
0
0
1
0
0
.
0
1
3
7
……….
.
0
.
0
0
7
6
O
t
h
e
r
I
P
C
C
r
i
me
s
0
.
0
0
1
9
0
0
.
0
2
5
9
…………
.
0
.
0
0
1
9
Stag
e
2
:
Gen
e
r
atin
g
a
tr
a
n
s
itio
n
m
at
r
ix
f
o
r
b
o
t
h
d
atasets
.
T
h
e
lo
ca
tio
n
s
r
ep
r
esen
tatio
n
as
a
s
tate
d
ia
g
r
am
o
f
B
ag
h
d
ad
city
is
s
h
o
wn
in
Fig
u
r
e
3
.
I
t
is
im
p
o
r
tan
t
to
n
o
te
t
h
at
th
e
tr
a
n
s
itio
n
s
tag
es
ex
p
r
e
s
s
th
e
m
o
v
em
e
n
t
o
f
cr
im
in
als
am
o
n
g
lo
ca
tio
n
.
T
h
e
lo
ca
tio
n
s
ar
e
r
ep
r
esen
ted
in
d
ig
r
a
p
h
o
f
t
h
e
Ma
r
k
o
v
ch
ain
.
W
h
er
e
th
e
cir
cle
s
h
ap
e
r
ep
r
esen
ts
th
e
n
o
d
es
(
cities)
in
wh
ich
th
e
cr
im
es
o
cc
u
r
r
ed
an
d
th
e
ed
g
es
co
n
n
ec
tin
g
th
e
n
o
d
es
b
etwe
en
th
em
,
to
r
e
p
r
esen
t th
e
p
o
s
s
ib
ilit
y
o
f
tr
an
s
itio
n
b
etwe
en
n
o
d
e
s
.
Stag
e
3
: E
n
ter
in
g
th
e
s
eq
u
en
c
e
o
f
cr
im
es f
o
r
b
o
th
d
ata
s
ets.
Stag
e
4
:
T
ab
le
5
to
T
ab
le
6
illu
s
tr
ates
th
e
co
m
p
ar
is
o
n
o
f
th
e
p
r
o
p
o
s
ed
alg
o
r
ith
m
with
th
e
tr
ad
itio
n
al
m
eth
o
d
s
in
ter
m
s
o
f
ac
cu
r
ac
y
,
r
elativ
e
m
ea
n
s
q
u
ar
e
e
r
r
o
r
(
R
MSE
)
an
d
co
n
s
u
m
e
d
tim
e.
T
h
e
R
MSE
is
co
n
s
id
er
ed
f
o
r
ac
cu
r
ac
y
ex
p
r
ess
io
n
f
o
r
H
MM
m
eth
o
d
s
th
at
in
clu
d
e
s
B
au
m
W
elch
an
d
th
e
p
r
o
p
o
s
ed
B
au
m
W
elch
+V
etr
ib
i
as
s
h
o
wn
in
T
ab
le
5
.
I
t
is
well
s
h
o
wn
th
at
th
e
p
r
o
p
o
s
ed
alg
o
r
ith
m
f
o
r
b
o
th
d
atasets
d
ec
r
ea
s
es
th
e
R
MSE
,
clea
r
ly
b
y
(
0
.
1
3
)
.
At
th
e
o
th
er
h
an
d
,
T
a
b
le
6
e
x
p
lain
s
th
e
co
m
p
ar
is
o
n
b
etwe
en
th
e
p
r
o
p
o
s
ed
(
DT
with
Viter
b
i)
an
d
tr
ad
itio
n
al
DT
.
I
n
th
is
tab
le,
th
e
ac
cu
r
ac
y
o
f
a
co
n
f
u
s
io
n
m
atr
ix
is
co
n
s
id
er
ed
as
th
e
m
ain
in
d
icato
r
.
T
h
e
DT
b
ased
o
n
Ma
tlab
lib
r
a
r
y
f
o
r
im
p
lem
en
tatio
n
.
T
h
e
ac
cu
r
ac
y
o
f
DT
b
ef
o
r
e
th
e
o
p
tim
izatio
n
is
ap
p
r
o
x
im
atel
y
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
KA
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l Co
n
tr
o
l
V
iter
b
i o
p
timiz
a
tio
n
fo
r
crime
d
etec
tio
n
a
n
d
id
en
tifi
ca
tio
n
(
R
ee
m
R
a
z
z
a
q
A
b
d
u
l H
u
s
s
ein
)
2383
(
9
8
%)
f
o
r
th
e
f
ir
s
t
d
ataset,
w
h
er
ea
s
it
is
n
ea
r
ly
)
9
3
.
8
%
(
f
o
r
th
e
s
ec
o
n
d
d
ataset.
T
h
e
ac
c
u
r
ac
y
o
f
DT
af
ter
o
p
tim
izatio
n
is
im
p
r
o
v
e
d
b
y
a
p
p
r
o
x
im
ately
)
9
8
.
8
%
(
f
o
r
th
e
f
ir
s
t
d
ataset.
Fo
r
th
e
s
ec
o
n
d
d
a
taset,
th
e
ac
cu
r
ac
y
is
also
im
p
r
o
v
ed
b
y
ar
o
u
n
d
)
9
5
.
9
%
(
af
ter
o
p
tim
izatio
n
.
T
h
e
o
b
tain
ed
r
esu
lts
s
h
o
w
th
at
th
e
p
r
o
p
o
s
ed
DT
+V
iter
b
i
in
cr
ea
s
es th
e
ac
cu
r
ac
y
f
o
r
b
o
t
h
d
atasets
.
Fig
u
r
e
3
.
State
tr
an
s
itio
n
b
e
f
o
r
e
o
p
tim
izatio
n
o
f
I
r
a
q
d
ataset
T
ab
le
5
.
C
o
m
p
a
r
is
o
n
o
f
r
esu
lts
f
o
r
R
MSE
B
a
u
m W
e
l
c
h
[
2
]
B
a
u
m W
e
l
c
h
+
V
i
t
e
r
b
i
(
p
r
e
p
o
s
e
d
w
o
r
k
)
D
a
t
a
s
e
t
1
0
.
1
7
0
.
1
2
D
a
t
a
s
e
t
2
0
.
0
4
0
.
0
3
T
ab
le
6
.
C
o
m
p
a
r
is
o
n
o
f
r
esu
lts
f
o
r
ac
cu
r
ac
y
in
th
e
co
n
f
u
s
io
n
m
atr
ix
A
c
c
u
r
a
c
y
i
n
t
h
e
c
o
n
f
u
si
o
n
ma
t
r
i
x
D
T
[
3
]
P
r
e
p
o
d
e
d
D
T+
v
i
t
e
r
b
i
(
p
r
e
p
o
se
d
w
o
r
k
)
D
a
t
a
s
e
t
1
~
%9
8
~
%9
8
.
8
D
a
t
a
s
e
t
2
~
%9
3
.
8
~
%9
5
.
9
T
ab
le
7
.
Sh
o
ws
th
e
tim
e
-
co
n
s
u
m
in
g
co
m
p
ar
is
o
n
o
f
th
e
p
r
o
p
o
s
ed
m
eth
o
d
s
(
B
au
m
W
elch
+
Viter
b
i)
with
th
e
tr
ad
itio
n
al
(
B
au
m
W
elch
)
alg
o
r
ith
m
s
.
A
p
p
lied
i
n
two
d
atasets
,
b
o
th
im
p
r
o
v
e
d
to
0
.
0
5
an
d
0
.
0
8
r
esp
ec
tiv
ely
.
I
t
is
p
r
o
v
e
d
th
at
th
e
p
r
o
p
o
s
ed
m
eth
o
d
s
in
clu
d
in
g
B
au
n
W
elch
o
u
t
p
er
f
o
r
m
th
e
tr
ad
itio
n
al
ap
p
r
o
ac
h
es
wh
ich
h
ad
b
ee
n
i
m
p
r
o
v
e
d
b
y
(
~0
.
1
5
s
ec
)
wh
ile
Dee
p
ik
a
K.
K
an
d
at
[
1
4
]
,
a
ch
iev
es
4
.
2
4
s
with
0
.
0
7
5
1
as R
MSE
.
T
ab
le
7
.
C
o
m
p
a
r
is
o
n
o
f
r
esu
lts
f
o
r
th
e
co
n
s
u
m
ed
tim
e
C
o
n
s
u
me
d
T
i
me
i
n
(
se
c
o
n
d
)
B
a
u
m
W
e
l
c
h
[
1]
B
a
u
m W
e
l
c
h
+
V
i
t
e
r
b
i
(
p
r
e
p
o
s
e
d
w
o
r
k
)
D
a
t
a
s
e
t
1
0
.
2
0
.
0
5
D
a
t
a
s
e
t
2
0
.
1
0
.
0
8
7.
CO
NCLU
SI
O
N
I
n
th
is
p
a
p
er
,
a
p
r
o
p
o
s
ed
HM
M
h
as
p
r
o
d
u
ce
d
to
p
r
ed
icate
th
e
lev
el
o
f
d
a
n
g
er
in
a
s
p
ec
if
ic
r
eg
io
n
.
T
h
e
co
m
b
in
atio
n
was
d
o
n
e
u
s
in
g
B
au
m
W
elch
with
Viter
b
i
f
r
o
m
a
s
id
e
a
n
d
DT
with
Viter
b
i
at
th
e
o
th
e
r
s
id
e
to
g
et
a
m
o
r
e
ac
cu
r
ate
m
o
d
el.
T
h
e
p
r
o
p
o
s
ed
m
eth
o
d
s
c
an
ac
h
iev
e
p
r
o
m
is
in
g
r
esu
lts
f
o
r
c
o
n
s
id
er
in
g
th
e
Viter
b
i
alg
o
r
ith
m
t
o
d
eter
m
in
e
th
e
s
eq
u
en
tial
r
elatio
n
o
f
p
ast
cr
im
e
ty
p
es,
with
two
laten
t
p
ar
a
m
eter
s
{m
o
n
th
an
d
lo
ca
tio
n
s
}.
T
h
e
o
p
tim
izatio
n
o
f
u
s
in
g
DT
m
eth
o
d
h
el
p
ed
th
e
p
r
o
p
o
s
ed
alg
o
r
ith
m
in
m
in
i
n
g
t
h
e
o
p
tim
al
s
o
lu
tio
n
.
I
n
ad
d
itio
n
,
th
e
c
o
n
s
u
m
ed
tim
e
is
r
ed
u
ce
d
in
ef
f
icien
t p
e
r
f
o
r
m
an
ce
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6
9
3
0
T
E
L
KOM
NI
KA
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l Co
n
tr
o
l
,
Vo
l.
18
,
No
.
5
,
Octo
b
e
r
2
0
2
0
:
2
3
7
8
-
2384
2384
RE
F
E
R
E
NC
E
S
[1
]
Re
e
m
R
.
A
.
,
M
u
a
y
a
d
S
.
C
.
,
S
a
li
h
M
.
A
.
,
“
De
v
e
lo
p
e
d
Crime
Lo
c
a
ti
o
n
P
re
d
ica
ti
o
n
Us
in
g
Latten
M
a
rk
o
v
M
o
d
e
l
,
”
J
o
u
rn
a
l
o
f
T
h
e
o
re
ti
c
a
l
a
n
d
A
p
p
l
ied
In
f
o
rm
a
ti
o
n
T
e
c
h
n
o
lo
g
y
,
v
o
l
.
96
,
n
o
.
1
,
p
p
.
2
9
0
-
3
0
1
,
2
0
1
9
.
[2
]
Re
e
m
R
.
A
.
,
S
a
li
h
M
.
A
.
,
M
u
a
y
a
d
S
.
C
.
,
“
Ro
le
o
f
Da
ta
M
i
n
in
g
in
E
-
G
o
v
e
rn
m
e
n
t
F
ra
m
e
wo
rk
,
”
Ira
q
i
J
o
u
r
n
a
l
fo
r
Co
mp
u
ter
s
a
n
d
In
f
o
rm
a
ti
c
s
,
v
o
l.
4
4
,
n
o
.
1
,
p
p
.
1
4
,
2
0
1
8
.
[3
]
Re
e
m
R
.
A
.,
M
u
a
y
a
d
S
.
C
.
,
S
a
li
h
M
.
A.
,
“
Im
p
ro
v
e
m
e
n
t
o
f
Cr
imin
a
l
Id
e
n
t
ifi
c
a
ti
o
n
b
y
S
m
a
rt
Op
ti
m
iza
ti
o
n
M
e
t
h
o
d
,
"
M
AT
EC
W
e
b
o
f
C
o
n
fer
e
n
c
e
s
ED
P
S
c
ien
c
e
s,
2
0
1
9
.
[4
]
Ch
e
n
H.,
“
S
p
e
c
ial
Iss
u
e
Dig
it
a
l
Go
v
e
rn
m
e
n
t:
tec
h
n
o
l
o
g
ies
a
n
d
p
ra
c
ti
c
e
s
,
”
De
c
isio
n
S
u
p
p
o
rt
S
y
ste
ms
,
vol
.
34
,
n
o
.
3
,
pp
.
2
2
3
-
2
2
7
,
2
0
0
3
.
[5
]
To
n
g
we
i
Y.
,
P
e
n
g
C
.
,
“
Da
ta
M
in
in
g
Ap
p
li
c
a
ti
o
n
s
in
g
o
v
e
rn
m
e
n
t
In
fo
rm
a
ti
o
n
S
e
c
u
ri
ty
,
”
Pr
o
c
e
d
i
a
En
g
in
e
e
rin
g
,
vol
.
29
,
pp
.
2
3
5
-
2
4
0
,
2
0
1
2
.
[6
]
Qa
y
y
u
m
,
S
h
a
m
a
il
a
,
Ha
fsa
S
.
D
.
,
Teh
m
in
a
A.
,
“
A
S
u
r
v
e
y
o
f
Da
ta
M
in
in
g
Tec
h
n
i
q
u
e
s
fo
r
Cri
m
e
De
te
c
ti
o
n
,”
Un
iv
e
rs
it
y
o
f
S
i
n
d
h
J
o
u
rn
a
l
o
f
In
f
o
rm
a
ti
o
n
a
n
d
Co
mm
u
n
ica
ti
o
n
T
e
c
h
n
o
l
o
g
y
,
v
o
l.
2
,
n
o
.
1
,
p
p
.
1
-
6
,
2
0
1
8
.
[7
]
Ke
y
v
a
n
p
o
u
r
,
M
o
h
a
m
m
a
d
R
.
,
M
o
s
tafa
J
.
,
a
n
d
M
o
h
a
m
m
a
d
R
.
E.
,
"
De
tec
ti
n
g
a
n
d
in
v
e
stig
a
ti
n
g
c
rime
b
y
m
e
a
n
s
o
f
d
a
ta
m
in
in
g
:
a
g
e
n
e
ra
l
c
rime
m
a
tch
in
g
fra
m
e
wo
rk
,
"
Pr
o
c
e
d
ia
C
o
mp
u
ter
S
c
ien
c
e
,
v
o
l.
3
,
p
p
.
8
7
2
-
8
8
0
,
2
0
1
1
.
[8
]
V.
G
u
p
ta
a
n
d
G
.
S
.
Leh
a
l,
“
A
s
u
rv
e
y
o
f
tex
t
m
in
i
n
g
tec
h
n
i
q
u
e
s
a
n
d
a
p
p
li
c
a
ti
o
n
s,
”
J
o
u
rn
a
l
o
f
Eme
rg
in
g
T
e
c
h
n
o
lo
g
ies
in
W
e
b
In
telli
g
e
n
c
e
,
v
o
l.
1
,
n
o
.
1
,
p
p
.
6
0
-
7
6
,
2
0
0
9
.
[9
]
Ka
d
h
im
B
.
A
.
,
S
wa
d
i
,
“
A p
ro
p
o
s
e
d
fr
a
m
e
wo
rk
f
o
r
a
n
a
ly
z
i
n
g
c
rime
d
a
ta
se
t
u
sin
g
d
e
c
isio
n
tree
a
n
d
sim
p
le
k
-
m
e
a
n
s
m
in
in
g
a
lg
o
rit
h
m
s
m”
J
o
u
r
n
a
l
o
f
Ku
fa
fo
r
M
a
t
h
e
ma
ti
c
s a
n
d
Co
mp
u
ter
,
v
o
l.
1
,
n
o
.
3
,
p
p
.
8
-
24
,
2
0
1
1
.
[1
0
]
J.
Ho
ss
e
in
k
h
a
n
i,
S
.
I
b
ra
h
im,
S
.
C
h
u
p
ra
t,
a
n
d
J.
H.
Na
n
iz,
“
Web
Crime
M
in
in
g
b
y
M
e
a
n
s
o
f
Da
ta
M
in
i
n
g
Tec
h
n
i
q
u
e
s,
”
Res
e
a
rc
h
J
o
u
rn
a
l
o
f
A
p
p
li
e
d
S
c
ie
n
c
e
s,
En
g
in
e
e
rin
g
a
n
d
T
e
c
h
n
o
lo
g
y
,
v
o
l
.
7
,
p
p
.
2
0
2
7
-
2
0
3
2
,
2
0
1
4
.
[1
1
]
Li
Xin
g
a
n
,
“
Ap
p
li
c
a
ti
o
n
o
f
d
a
ta
m
in
in
g
m
e
th
o
d
s
in
t
h
e
stu
d
y
o
f
c
ri
m
e
b
a
se
d
o
n
in
tern
a
ti
o
n
a
l
d
a
ta
so
u
rc
e
s
,”
T
a
mp
e
re
Un
ive
rs
it
y
Pre
ss
,
2
0
1
4
.
[1
2
]
Re
e
m
R
.
A
.
,
M
u
a
y
a
d
S
.
C
.
,
S
a
li
h
M
a
h
d
i
Al
-
Qa
ra
a
wi
,
"
M
u
lt
ista
g
e
Tree
M
o
d
e
l
fo
r
Crime
Da
tas
e
t
in
Ira
q
,
"
Ir
a
q
i
J
o
u
rn
a
l
o
f
Co
m
p
u
ter
s
,
C
o
mm
u
n
ic
a
ti
o
n
,
a
n
d
C
o
n
tro
l
&
S
y
ste
ms
En
g
in
e
e
rin
g
,
v
o
l.
1
9
,
n
o
.
2
,
p
p
.
1
-
8
,
2
0
1
9
.
[1
3
]
M
a
th
e
w,
Wes
ley
,
R
u
b
e
n
R
.
,
a
n
d
Bru
n
o
M
.
,
"
P
re
d
ictin
g
fu
t
u
re
lo
c
a
ti
o
n
s
wit
h
h
id
d
e
n
M
a
r
k
o
v
m
o
d
e
ls
,
"
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
2
0
1
2
ACM
c
o
n
fer
e
n
c
e
o
n
u
b
iq
u
it
o
u
s c
o
m
p
u
ti
n
g
,
2
0
1
2
.
[1
4
]
K
.
De
e
p
ik
a
,
Vin
o
d
,
S
m
it
h
a
,
“
Cr
ime
a
n
a
ly
sis
in
In
d
ia
u
si
n
g
d
a
ta
m
in
in
g
tec
h
n
i
q
u
e
s
,”
I
n
ter
n
a
t
io
n
a
l
J
o
u
r
n
a
l
o
f
En
g
i
n
e
e
rin
g
&
T
e
c
h
n
o
lo
g
y
, v
ol
.
7
,
n
o
.
2
.
6
,
p
p
.
2
5
3
-
2
5
8
,
2
0
1
8
[1
5
]
M
e
k
a
th
o
ti
R
.
B
.,
Ko
n
d
a
p
a
ll
i
,
“O
p
t
imiz
in
g
c
rime
h
o
ts
p
o
ts
a
n
d
c
o
ld
s
p
o
ts
u
sin
g
Hid
d
e
n
M
a
rk
o
v
M
o
d
e
l
,
"
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
Res
e
a
rc
h
,
v
ol
.
4
,
n
o
.
1
7
,
p
p
.
2
3
4
8
-
6
8
4
8
,
2
0
1
7
.
[1
6
]
Jo
sh
i
J.
C.
,
Tan
k
e
sh
wa
r
Ku
m
a
r,
S
u
n
i
ta
S
.
,
a
n
d
Di
v
y
a
S
a
c
h
d
e
v
a
,
“
Op
ti
m
isa
ti
o
n
o
f
Hid
d
e
n
M
a
r
k
o
v
M
o
d
e
l
u
si
n
g
Ba
u
m
–
Welc
h
a
lg
o
rit
h
m
fo
r
p
re
d
i
c
ti
o
n
o
f
m
a
x
im
u
m
a
n
d
m
in
imu
m
tem
p
e
ra
tu
re
o
v
e
r
In
d
ian
Him
a
lay
a
,”
J
o
u
r
n
a
l
o
f
Ea
rth
S
y
ste
m S
c
ien
c
e
,
v
o
l,
1
2
6
,
n
o
.
3,
2
0
1
7
.
[1
7
]
G
ö
rn
it
z
,
Nic
o
,
M
i
k
io
B
.
,
a
n
d
M
a
ri
u
s
K.
,
“
Hid
d
e
n
M
a
rk
o
v
a
n
o
m
a
ly
d
e
tec
ti
o
n
,”
I
n
ter
n
a
t
io
n
a
l
C
o
n
fer
e
n
c
e
o
n
M
a
c
h
i
n
e
L
e
a
rn
in
g
,
p
p
.
1
8
3
3
-
1
8
4
2
,
2
0
1
5
.
[1
8
]
Leh
é
ricy
Lu
c
,
“
S
tate
-
by
-
sta
te
m
in
ima
x
a
d
a
p
ti
v
e
e
stim
a
ti
o
n
f
o
r
n
o
n
p
a
ra
m
e
tri
c
h
id
d
e
n
M
a
r
k
o
v
m
o
d
e
l
s
,”
T
h
e
J
o
u
rn
a
l
o
f
M
a
c
h
i
n
e
L
e
a
r
n
in
g
Res
e
a
rc
h
,
v
o
l.
1
9
,
n
o
.
1
,
p
p
.
1
4
3
2
-
1
4
7
7
,
2
0
1
8
.
[1
9
]
Zh
e
n
g
,
Yu
h
u
i
B
.
J
.
,
Le
S
.
,
Jia
n
we
i
Z
.
,
a
n
d
Hu
i
Z
,
"
S
tu
d
e
n
t’s
t
-
h
id
d
e
n
M
a
rk
o
v
m
o
d
e
l
fo
r
u
n
s
u
p
e
rv
ise
d
lea
rn
in
g
u
sin
g
lo
c
a
li
z
e
d
fe
a
tu
re
se
lec
ti
o
n
,
"
IEE
E
T
ra
n
s
a
c
ti
o
n
s
o
n
Circ
u
i
ts
a
n
d
S
y
ste
ms
fo
r
Vi
d
e
o
T
e
c
h
n
o
l
o
g
y
,
v
o
l.
2
8
,
n
o
.
1
0
,
pp.
2
5
8
6
-
2
5
9
8
,
2
0
1
8
.
[2
0
]
Wo
n
,
Ky
o
u
n
g
-
Ja
e
,
A
.
P
ru
g
e
l
-
Be
n
n
e
tt
,
a
n
d
A
n
d
e
rs
K
.,
“
E
v
o
l
v
i
n
g
th
e
stru
c
t
u
re
of
h
id
d
e
n
M
a
rk
o
v
m
o
d
e
ls
,”
IEE
E
T
ra
n
sa
c
ti
o
n
s
on
Evo
l
u
ti
o
n
a
ry
Co
mp
u
ta
ti
o
n
,
v
o
l
.
10
,
n
o
.
1
,
p
p
.
39
-
49
,
2
0
0
6
.
[
2
1
]
G
h
o
s
h
,
S
o
u
m
y
a
K
.
,
a
n
d
S
h
r
e
y
a
G.
,
“
M
o
d
e
l
i
n
g
I
n
d
i
v
i
d
u
a
l
'
s
M
o
v
e
m
e
n
t
P
a
t
t
e
r
n
s
to
I
n
f
e
r
N
e
x
t
L
o
c
a
t
i
o
n
f
r
o
m
S
p
a
r
s
e
T
r
a
j
e
c
t
o
r
y
T
r
a
c
e
s
,”
2018
I
E
E
E
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
on
S
y
s
t
e
m
s
,
Man,
a
n
d
C
y
b
e
r
n
e
t
i
c
s
(
S
M
C
)
,
p
p
.
6
9
3
-
6
9
8
,
2
0
1
8
.
[2
2
]
Emd
a
d
i,
Ak
ra
m
,
F
a
tem
e
h
A
.
M
.
,
F
a
tem
e
h
Y
.
M
.
,
a
n
d
Ch
a
n
g
iz
E.
,
“
A
n
o
v
e
l
a
lg
o
rit
h
m
f
o
r
p
a
ra
m
e
ter
e
stim
a
ti
o
n
o
f
Hid
d
e
n
M
a
rk
o
v
M
o
d
e
l
in
s
p
ired
b
y
An
t
C
o
l
o
n
y
Op
t
imiz
a
ti
o
n
,”
He
li
y
o
n
,
v
o
l
.
5
,
n
o
.
3
,
2
0
1
9
.
[2
3
]
Ro
b
i
n
so
n
,
Wi
ll
iam
N.
,
a
n
d
An
d
r
e
a
A.
,
“
S
e
q
u
e
n
ti
a
l
fra
u
d
d
e
tec
ti
o
n
f
o
r
p
re
p
a
i
d
c
a
rd
s
u
si
n
g
h
i
d
d
e
n
M
a
rk
o
v
m
o
d
e
l
d
iv
e
rg
e
n
c
e
,”
Exp
e
rt S
y
ste
ms
wit
h
Ap
p
li
c
a
ti
o
n
s
,
v
o
l
.
91
,
p
p
.
2
3
5
-
2
5
1
,
2
0
1
8
.
[2
4
]
Zh
a
n
g
,
Ya
n
x
u
e
,
Do
n
g
m
e
i
Z
.
,
a
n
d
Jin
x
i
n
g
Li
u
,
“
T
h
e
a
p
p
li
c
a
ti
o
n
of
Ba
u
m
-
Welc
h
a
lg
o
rit
h
m
in
t
h
e
m
u
lt
istep
a
tt
a
c
k
,”
T
h
e
S
c
ien
t
if
ic
W
o
rld
J
o
u
r
n
a
l
,
v
o
l.
2
0
1
4
,
p
p
.
1
-
7,
2
0
1
4
.
[2
5
]
Ka
g
g
le,
“
Da
ra
se
t:
Crime
in
In
d
ia,
”
2
0
1
7
.
[O
n
li
n
e
].
Av
a
il
a
b
le
h
tt
p
s:
//
ww
w.k
a
g
g
le.c
o
m
/raja
n
a
n
d
/crim
e
-
in
-
in
d
ia/
.
Evaluation Warning : The document was created with Spire.PDF for Python.