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
.
6
,
Dec
em
b
er
2
0
2
0
,
p
p
.
3
3
3
1
~
3
3
3
8
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
i6
.
1
6
3
0
0
3331
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
New pr
ediction
m
ethod for
data
sp
read
ing
in
so
cia
l
networks
ba
sed o
n ma
chine
learning
alg
o
rith
m
M
a
y
t
ha
m
N.
M
eqda
d
1
,
Ra
wy
a
Al
-
Aka
m
2
,
Seif
edine K
a
d
ry
3
1
Al
-
M
u
sta
q
b
a
l
Un
i
v
e
rsity
C
o
ll
e
g
e
,
Ira
q
2
Ko
b
len
z
-
Lan
d
a
u
U
n
iv
e
rsit
y
,
G
e
rm
a
n
3
Be
iru
t
Ara
b
Un
iv
e
rsity
,
Leb
a
n
o
n
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Ap
r
9
,
2
0
2
0
R
ev
is
ed
J
u
n
7
,
2
0
2
0
Acc
ep
ted
J
u
l 1
6
,
2
0
2
0
In
fo
rm
a
ti
o
n
d
iffu
si
o
n
p
re
d
ictio
n
is
th
e
st
u
d
y
o
f
t
h
e
p
a
th
o
f
d
isse
m
in
a
ti
o
n
o
f
n
e
ws
,
in
fo
rm
a
ti
o
n
,
o
r
t
o
p
ics
i
n
a
stru
c
tu
re
d
d
a
ta
su
c
h
a
s
a
g
ra
p
h
.
R
e
se
a
rc
h
in
th
is
a
re
a
is
fo
c
u
se
d
o
n
tw
o
g
o
a
ls,
trac
in
g
t
h
e
in
fo
rm
a
ti
o
n
d
iffu
si
o
n
p
a
th
a
n
d
fin
d
i
n
g
t
h
e
m
e
m
b
e
rs
t
h
a
t
d
e
term
in
e
fu
t
u
re
t
h
e
n
e
x
t
p
a
th
.
T
h
e
m
a
jo
r
p
r
o
b
lem
o
f
trad
i
ti
o
n
a
l
a
p
p
ro
a
c
h
e
s
i
n
t
h
is
a
r
e
a
is
th
e
u
se
o
f
sim
p
le
p
r
o
b
a
b
il
isti
c
m
e
th
o
d
s
ra
th
e
r
th
a
n
in
tel
li
g
e
n
t
m
e
th
o
d
s.
R
e
c
e
n
t
y
e
a
rs
h
a
v
e
se
e
n
g
ro
win
g
i
n
t
e
re
st
in
th
e
u
se
o
f
m
a
c
h
in
e
lea
rn
i
n
g
a
lg
o
rit
h
m
s
in
th
is
field
.
Re
c
e
n
tl
y
,
d
e
e
p
lea
rn
i
n
g
,
wh
ich
is a b
ra
n
c
h
o
f
m
a
c
h
in
e
lea
rn
in
g
,
h
a
s b
e
e
n
i
n
c
re
a
sin
g
l
y
u
se
d
i
n
th
e
field
o
f
in
f
o
rm
a
ti
o
n
d
iff
u
sio
n
p
re
d
icti
o
n
.
T
h
is
p
a
p
e
r
p
re
se
n
ts
a
m
a
c
h
in
e
lea
rn
in
g
m
e
th
o
d
b
a
se
d
o
n
th
e
g
ra
p
h
n
e
u
ra
l
n
e
two
r
k
a
lg
o
rit
h
m
,
wh
ic
h
i
n
v
o
l
v
e
s
th
e
se
lec
ti
o
n
o
f
in
a
c
ti
v
e
v
e
rti
c
e
s
fo
r
a
c
ti
v
a
ti
o
n
b
a
se
d
o
n
th
e
n
e
ig
h
b
o
ri
n
g
v
e
rti
c
e
s
th
a
t
a
re
a
c
ti
v
e
in
a
g
i
v
e
n
sc
ien
t
ifi
c
to
p
ic.
Ba
sic
a
ll
y
,
i
n
th
is
m
e
th
o
d
,
i
n
fo
rm
a
ti
o
n
d
iffu
si
o
n
p
a
t
h
s
a
re
p
re
d
icte
d
t
h
r
o
u
g
h
th
e
a
c
ti
v
a
ti
o
n
o
f
in
a
c
ti
v
e
v
e
rti
c
e
s
b
y
a
c
ti
v
e
v
e
r
ti
c
e
s.
Th
e
m
e
th
o
d
is
tes
t
e
d
o
n
th
re
e
sc
ien
t
ifi
c
b
ib
li
o
g
ra
p
h
y
d
a
tas
e
ts:
Th
e
Dig
it
a
l
Bi
b
li
o
g
ra
p
h
y
a
n
d
Li
b
r
a
ry
P
ro
jec
t
(
DBLP
)
,
P
u
b
m
e
d
,
a
n
d
Co
ra
.
Th
e
m
e
th
o
d
a
tt
e
m
p
ts
to
a
n
sw
e
r
th
e
q
u
e
stio
n
th
a
t
w
h
o
will
b
e
th
e
p
u
b
li
sh
e
r
o
f
t
h
e
n
e
x
t
a
rti
c
le
i
n
a
s
p
e
c
ifi
c
field
o
f
sc
ien
c
e
.
Th
e
c
o
m
p
a
ris
o
n
o
f
th
e
p
ro
p
o
se
d
m
e
th
o
d
wit
h
o
t
h
e
r
m
e
th
o
d
s
sh
o
w
s
1
0
%
a
n
d
5
%
im
p
ro
v
e
d
p
re
c
isio
n
in
DBL
P
a
n
d
P
u
b
m
e
d
d
a
tas
e
ts,
re
sp
e
c
ti
v
e
l
y
.
K
ey
w
o
r
d
s
:
Data
s
p
r
ea
d
in
g
Ma
ch
in
e
lear
n
in
g
Pre
d
ictio
n
So
cial
n
etwo
r
k
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
:
Seif
ed
in
e
Kad
r
y
,
Dep
ar
tm
en
t o
f
Ma
th
em
atics a
n
d
C
o
m
p
u
te
r
Scien
ce
,
Facu
lty
o
f
Scien
ce
,
B
eir
u
t A
r
ab
Un
iv
er
s
ity
,
B
eir
u
t,
L
eb
an
o
n
.
E
m
ail:
s
.
k
ad
r
y
@
b
au
.
ed
u
.
lb
1.
I
NT
RO
D
UCT
I
O
N
I
n
f
o
r
m
atio
n
d
if
f
u
s
io
n
in
h
o
m
o
g
en
o
u
s
an
d
h
eter
o
g
en
e
o
u
s
n
etwo
r
k
s
is
a
d
y
n
am
ic
p
r
o
ce
s
s
o
f
k
ee
n
in
ter
est
to
r
esear
c
h
er
s
.
T
h
is
co
n
ce
p
t
r
e
f
er
s
to
h
o
w
in
f
o
r
m
atio
n
lik
e
n
ews
o
f
e
v
en
ts
o
u
tb
r
ea
k
s
etc.
s
p
r
ea
d
f
r
o
m
a
s
et
o
f
o
r
ig
i
n
n
o
d
es
to
o
th
e
r
n
o
d
es
ac
r
o
s
s
th
e
n
etwo
r
k
[
1
-
3
]
.
I
n
f
o
r
m
atio
n
d
if
f
u
s
io
n
h
as
b
ee
n
s
tu
d
ied
in
m
an
y
f
ield
s
r
an
g
in
g
f
r
o
m
h
ea
lth
ca
r
e
[
4
-
6
]
to
s
o
cial
n
etwo
r
k
s
[
7
,
8
]
.
On
e
o
f
th
e
m
o
s
t
im
p
o
r
tan
t
task
s
o
f
n
etwo
r
k
-
b
ased
s
y
s
tem
s
is
to
u
n
d
er
s
tan
d
,
m
o
d
el,
an
d
p
r
e
d
ict
r
ap
id
ly
d
ev
elo
p
i
n
g
ev
e
n
ts
with
in
th
e
n
etwo
r
k
.
A
f
ter
d
is
co
v
er
in
g
th
e
s
tr
u
ctu
r
e
o
f
a
n
etwo
r
k
,
it
is
p
o
s
s
ib
le
to
p
r
ed
i
ct
th
e
p
atter
n
s
o
f
e
v
en
ts
in
clu
d
in
g
th
eir
s
h
ap
e,
s
ize
an
d
d
e
v
elo
p
m
e
n
t,
wh
ich
ca
n
b
e
d
escr
ib
ed
as
in
f
o
r
m
atio
n
d
i
f
f
u
s
io
n
[
9
]
.
O
v
er
th
e
y
ea
r
s
,
r
e
s
ea
r
ch
er
s
h
av
e
tr
ied
s
ev
er
al
m
eth
o
d
s
to
m
o
d
el
in
f
o
r
m
atio
n
d
if
f
u
s
io
n
in
h
o
m
o
g
en
eo
u
s
an
d
h
eter
o
g
e
n
eo
u
s
n
etwo
r
k
s
[
1
0
-
1
3
]
.
Fo
r
m
ally
,
a
d
ata
n
etwo
r
k
is
r
e
p
r
esen
ted
b
y
a
g
r
a
p
h
G
=
(
V,
E
)
wh
er
e
V
is
th
e
s
et
o
f
v
er
tic
es
an
d
E
is
th
e
s
et
o
f
ed
g
es.
T
h
is
g
r
a
p
h
i
s
ca
lled
h
o
m
o
g
en
eo
u
s
if
th
e
v
e
r
tices
an
d
th
eir
ed
g
es
ar
e
o
f
t
h
e
s
am
e
ty
p
e,
an
d
is
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
.
6
,
Dec
em
b
e
r
2
0
2
0
:
3
3
3
1
-
333
8
3332
ca
lled
h
eter
o
g
en
e
o
u
s
o
th
er
wis
e.
Ho
m
o
g
en
e
o
u
s
n
etwo
r
k
s
h
av
e
b
ee
n
th
e
s
u
b
ject
o
f
m
an
y
s
t
u
d
ies
with
th
e
f
o
cu
s
b
ein
g
o
n
s
em
an
tic
an
aly
s
is
,
co
m
m
u
n
icab
le
d
is
ea
s
e
co
n
tr
o
l
[
1
4
-
1
7
]
,
an
d
lin
k
p
r
ed
ictio
n
[
1
8
-
2
0
]
.
R
ec
en
tly
,
m
o
r
e
atten
tio
n
h
as
b
ee
n
p
ai
d
to
h
et
er
o
g
en
e
o
u
s
n
etwo
r
k
s
,
as
th
ey
co
u
ld
p
r
o
v
id
e
a
m
o
r
e
r
ea
lis
tic
r
ep
r
esen
tatio
n
o
f
r
ea
l
-
wo
r
ld
p
h
e
n
o
m
en
a
[
2
1
,
2
2
]
.
I
n
a
s
tu
d
y
b
y
W
att
[
2
3
]
,
h
e
in
v
esti
g
ated
th
e
r
o
le
o
f
th
r
esh
o
ld
v
alu
es
an
d
n
etwo
r
k
s
tr
u
ct
u
r
e
in
in
f
o
r
m
ati
o
n
d
i
f
f
u
s
io
n
in
th
ese
n
etwo
r
k
s
.
I
n
[
1
0
]
,
i
n
f
o
r
m
atio
n
d
if
f
u
s
io
n
in
h
eter
o
g
en
eo
u
s
n
etwo
r
k
s
th
r
o
u
g
h
h
y
p
er
p
ath
s
was
s
tu
d
ied
.
T
h
is
s
tu
d
y
p
r
o
p
o
s
ed
a
m
eth
o
d
ca
lled
ML
T
M
-
R
f
o
r
an
aly
zin
g
in
f
o
r
m
atio
n
d
if
f
u
s
io
n
p
o
wer
i
n
d
if
f
e
r
en
t
h
y
p
e
r
p
at
h
s
.
I
n
t
h
is
m
eth
o
d
,
p
r
e
d
ictio
n
s
wer
e
m
ad
e
with
th
e
Path
Sim
alg
o
r
ith
m
u
s
ed
to
weig
h
t th
e
lin
k
s
b
etwe
en
ea
ch
two
n
o
d
es [
2
4
,
2
5
]
.
R
ec
en
t
y
ea
r
s
h
av
e
s
ee
n
a
g
r
o
win
g
in
ter
est
in
th
e
u
s
e
o
f
d
ee
p
lear
n
in
g
i
n
h
eter
o
g
en
eo
u
s
n
etwo
r
k
s
[
2
6
,
2
7
]
.
I
n
[
2
8
]
,
a
co
r
e
d
ee
p
l
ea
r
n
i
n
g
(
C
DL
)
f
r
am
ewo
r
k
was
u
s
ed
to
s
o
l
v
e
th
e
p
r
o
b
lem
o
f
h
eter
o
g
en
e
o
u
s
v
is
u
al
v
er
s
u
s
n
ea
r
-
in
f
r
ar
ed
(
VI
S
-
NI
R
)
im
a
g
e
m
atch
in
g
th
r
o
u
g
h
to
p
ic
d
i
f
f
u
s
io
n
in
n
etwo
r
k
s
.
T
an
g
et
al.
[
2
9
]
p
r
o
p
o
s
ed
a
L
I
NE
alg
o
r
ith
m
f
o
r
e
m
b
ed
d
in
g
l
ea
r
n
in
g
th
at
tr
av
er
s
es
all
ed
g
e
ty
p
es
an
d
s
am
p
les
o
n
e
e
d
g
e
at
a
tim
e
f
o
r
ea
c
h
e
d
g
e
ty
p
e
.
C
h
an
g
et
al
[
3
0
]
d
ev
el
o
p
ed
a
d
ee
p
a
r
ch
itectu
r
e
f
o
r
in
f
o
r
m
atio
n
d
if
f
u
s
io
n
p
r
ed
ictio
n
th
r
o
u
g
h
in
f
o
r
m
ati
o
n
en
co
d
in
g
in
h
eter
o
g
en
e
o
u
s
n
etwo
r
k
s
.
I
n
[
3
1
]
,
a
n
e
w
alg
o
r
ith
m
ca
lled
Me
tap
ath
2
v
ec
was
p
r
esen
ted
f
o
r
i
n
f
o
r
m
atio
n
en
c
o
d
in
g
i
n
h
et
er
o
g
en
e
o
u
s
n
etwo
r
k
s
,
wh
er
e
c
o
n
ce
p
ts
an
d
p
atter
n
s
ar
e
m
ap
p
ed
b
y
th
e
u
s
e
o
f
h
y
p
e
r
p
ath
s
.
T
h
e
r
ev
iew
o
f
p
r
ev
i
o
u
s
wo
r
k
s
r
ev
ea
ls
s
o
m
e
s
tr
en
g
th
s
an
d
wea
k
n
ess
es
in
th
e
cu
r
r
e
n
t
ap
p
r
o
ac
h
to
in
f
o
r
m
atio
n
d
if
f
u
s
io
n
in
h
eter
o
g
en
eo
u
s
n
etwo
r
k
s
.
T
h
e
u
s
e
o
f
d
ee
p
l
ea
r
n
in
g
i
n
th
e
s
tu
d
y
o
f
in
f
o
r
m
atio
n
d
if
f
u
s
io
n
p
r
o
ce
s
s
es
s
u
ch
as
to
p
ic
d
if
f
u
s
io
n
an
d
in
f
o
r
m
atio
n
ca
s
ca
d
es
ca
n
h
elp
av
o
id
t
h
e
p
r
o
b
lem
s
o
f
m
o
r
e
tr
a
d
itio
n
al
m
eth
o
d
s
.
T
h
e
m
aj
o
r
d
is
ad
v
an
tag
e
o
f
th
e
p
r
ev
io
u
s
wo
r
k
s
is
th
at
m
o
s
t
to
p
ic
d
if
f
u
s
io
n
m
eth
o
d
s
u
s
e
lo
ca
l
s
im
ilar
ity
an
d
en
co
d
in
g
b
ased
o
n
n
eig
h
b
o
r
in
g
n
o
d
es.
Fo
r
l
ar
g
e
h
eter
o
g
en
eo
u
s
n
etwo
r
k
s
,
it
is
tim
e
-
co
n
s
u
m
in
g
an
d
d
if
f
icu
lt
to
p
e
r
f
o
r
m
lo
ca
l
s
im
ilar
ity
ca
lcu
latio
n
s
f
o
r
ea
ch
two
co
r
r
esp
o
n
d
in
g
n
o
d
es.
As
a
r
esu
lt,
th
er
e
is
a
n
ee
d
f
o
r
a
m
o
r
e
co
m
p
r
eh
en
s
iv
e
y
et
less
co
m
p
le
x
a
u
to
m
atic
m
e
th
o
d
f
o
r
m
ea
s
u
r
i
n
g
th
e
s
im
ilar
ity
o
f
n
o
d
es a
n
d
f
in
d
in
g
d
if
f
u
s
io
n
p
ath
s
in
h
eter
o
g
en
eo
u
s
n
etwo
r
k
s
.
I
n
th
is
p
ap
er
,
th
e
p
r
o
b
lem
o
f
p
r
ed
ictin
g
th
e
p
ath
o
f
in
f
o
r
m
a
tio
n
d
if
f
u
s
io
n
in
a
n
etwo
r
k
is
m
ap
p
ed
to
a
d
ee
p
lear
n
i
n
g
p
r
o
b
lem
.
Sin
ce
p
r
ed
ictin
g
t
h
e
n
ew
u
s
er
s
wh
o
will
b
e
in
th
e
p
at
h
o
f
i
n
f
o
r
m
atio
n
f
lo
w
is
a
r
ec
o
g
n
itio
n
p
r
o
ce
s
s
,
th
is
p
r
o
b
lem
ca
n
b
e
s
o
lv
ed
b
y
m
ac
h
in
e
lear
n
i
n
g
al
g
o
r
ith
m
s
.
As
n
o
ted
in
s
ec
tio
n
X,
r
ec
en
tly
,
d
ee
p
m
ac
h
in
e
lea
r
n
in
g
alg
o
r
ith
m
s
h
a
v
e
b
ee
n
wi
d
ely
u
s
ed
in
th
is
f
ield
.
Als
o
,
r
esear
ch
er
s
h
av
e
d
ev
elo
p
e
d
d
ee
p
m
ac
h
in
e
lea
r
n
i
n
g
alg
o
r
ith
m
s
th
at
ca
n
u
s
e
g
r
ap
h
d
ata
in
th
e
lear
n
in
g
p
r
o
ce
s
s
.
T
h
is
p
ap
er
p
r
esen
ts
a
m
ac
h
in
e
lear
n
in
g
m
eth
o
d
b
a
s
ed
o
n
g
r
ap
h
n
eu
r
al
n
etwo
r
k
s
,
wh
ich
in
v
o
lv
es
s
elec
tin
g
th
e
in
ac
tiv
e
n
o
d
e
to
b
e
ac
tiv
ated
b
ased
o
n
its
n
ei
g
h
b
o
r
in
g
ac
tiv
e
n
o
d
es
in
ea
ch
s
cien
tific
to
p
ic.
I
n
o
th
er
wo
r
d
s
,
in
th
is
m
eth
o
d
,
in
f
o
r
m
atio
n
d
if
f
u
s
io
n
p
ath
s
ar
e
p
r
ed
icted
th
r
o
u
g
h
th
e
ac
tiv
at
io
n
o
f
in
ac
tiv
e
n
o
d
es
b
y
ac
tiv
e
n
o
d
es.
T
o
e
v
alu
ate
th
e
p
r
o
p
o
s
ed
m
eth
o
d
,
it
is
test
ed
o
n
th
r
ee
h
eter
o
g
e
n
eo
u
s
s
c
ien
tific
d
atab
ases
:
T
h
e
Dig
ital
B
ib
lio
g
r
ap
h
y
an
d
L
ib
r
ar
y
Pro
ject
(
DB
L
P
)
,
Pu
b
m
ed
,
a
n
d
C
o
r
a.
T
h
e
m
eth
o
d
s
ee
k
s
to
a
n
s
wer
th
e
q
u
esti
o
n
t
h
at
wh
o
will
b
e
th
e
p
u
b
lis
h
er
o
f
th
e
n
ex
t
ar
ticle
in
a
p
ar
ticu
lar
f
ield
o
f
s
cien
ce
.
T
h
e
co
m
p
a
r
is
o
n
o
f
th
e
p
r
o
p
o
s
ed
m
eth
o
d
with
o
th
er
m
eth
o
d
s
s
h
o
ws 1
0
% a
n
d
5
% i
m
p
r
o
v
e
m
en
t in
p
r
ec
is
io
n
in
D
B
L
P a
n
d
Pu
b
m
ed
d
atasets
,
r
esp
ec
tiv
ely
.
I
n
s
u
m
m
ar
y
,
th
e
m
o
s
t im
p
o
r
ta
n
t in
n
o
v
atio
n
s
o
f
t
h
e
p
r
esen
t
wo
r
k
ar
e
as f
o
llo
ws:
−
Pre
s
en
tin
g
a
d
ee
p
lear
n
i
n
g
m
o
d
el
wh
er
e
th
e
i
n
f
o
r
m
atio
n
o
f
a
h
eter
o
g
en
e
o
u
s
n
etwo
r
k
is
en
co
d
e
d
in
th
e
f
o
r
m
o
f
a
d
ee
p
lear
n
in
g
g
r
ap
h
,
wh
ich
ca
n
m
o
d
el
th
e
in
f
o
r
m
atio
n
d
if
f
u
s
io
n
p
ath
.
−
Pro
v
id
in
g
a
f
ea
tu
r
e
ex
tr
ac
tio
n
m
ec
h
an
is
m
to
f
in
d
t
h
e
d
eg
r
ee
o
f
c
o
r
r
elatio
n
o
f
n
ei
g
h
b
o
r
in
g
v
er
tices
in
d
if
f
er
en
t
g
r
ap
h
h
y
p
er
p
ath
s
.
−
T
esti
n
g
th
e
m
eth
o
d
o
n
th
e
h
eter
o
g
en
eo
u
s
d
atas
ets
DB
L
P,
Pu
b
m
ed
,
a
n
d
C
o
r
a,
wh
ic
h
h
av
e
r
ea
l
-
wo
r
ld
ap
p
licatio
n
s
,
in
o
r
d
er
to
d
em
o
n
s
tr
ate
th
e
ap
p
licab
ilit
y
o
f
th
e
p
r
o
p
o
s
ed
m
eth
o
d
.
T
h
e
r
em
ain
i
n
g
s
ec
tio
n
s
o
f
th
e
ar
ticle
ar
e
o
r
g
an
ized
as
f
o
llo
w
s
.
Sectio
n
2
d
escr
ib
es
th
e
d
if
f
e
r
en
t
co
m
p
o
n
e
n
ts
o
f
th
e
p
r
o
p
o
s
ed
m
eth
o
d
.
Sectio
n
3
d
escr
ib
es
th
e
test
in
g
p
r
o
ce
d
u
r
e
an
d
a
n
aly
ze
s
th
e
r
esu
lts
.
An
d
Sectio
n
4
p
r
esen
ts
th
e
co
n
clu
s
io
n
s
.
2.
P
RO
P
O
SE
D
M
E
T
H
O
D
I
n
f
o
r
m
atio
n
d
if
f
u
s
io
n
is
a
wid
ely
d
is
cu
s
s
ed
d
y
n
am
ic
n
etwo
r
k
p
r
o
ce
s
s
with
p
o
ten
tial
a
p
p
li
ca
tio
n
s
in
v
ar
io
u
s
f
ield
s
o
f
s
cien
ce
.
T
h
is
ter
m
r
ef
er
s
to
th
e
s
p
r
ea
d
in
g
o
f
in
f
o
r
m
atio
n
o
r
s
im
ilar
co
n
ce
p
ts
s
u
ch
as
n
ews,
in
n
o
v
atio
n
,
v
ir
u
s
o
r
m
alwa
r
e
a
s
et
o
f
v
er
tices
to
o
th
er
v
er
ti
ce
s
ac
r
o
s
s
th
e
n
etwo
r
k
.
T
h
er
e
is
a
r
ich
b
o
d
y
o
f
liter
atu
r
e
o
n
in
f
o
r
m
atio
n
d
if
f
u
s
io
n
in
co
m
p
le
x
n
etwo
r
k
s
,
wh
er
e
d
if
f
er
e
n
t
m
o
d
els
an
d
t
h
eir
in
ter
ac
tio
n
s
with
n
etwo
r
k
to
p
o
lo
g
y
h
av
e
b
ee
n
a
n
aly
ze
d
[
1
]
.
T
h
e
p
r
ev
io
u
s
s
tu
d
ies
h
av
e
b
ee
n
m
o
s
tly
f
o
cu
s
e
d
o
n
h
eter
o
g
en
e
o
u
s
n
etwo
r
k
s
.
An
in
f
o
r
m
atio
n
n
et
wo
r
k
lik
e
G
=
(
V,
E
)
wh
er
e
V
is
th
e
s
et
o
f
v
er
tices
an
d
E
is
th
e
s
et
o
f
ed
g
es
is
h
o
m
o
g
en
eo
u
s
if
th
e
ed
g
es
a
n
d
v
er
tices
ar
e
o
f
th
e
s
am
e
ty
p
e
.
C
o
n
v
er
s
ely
,
th
e
n
etwo
r
k
s
with
m
o
r
e
th
an
o
n
e
ty
p
e
o
f
n
o
d
e
o
r
ed
g
e
ar
e
ca
lled
h
eter
o
g
en
eo
u
s
[
8
-
1
0
]
.
Fo
r
ex
a
m
p
le,
in
DB
L
P,
wh
ich
is
an
im
p
o
r
tan
t
co
m
p
u
ter
s
cien
ce
b
ib
lio
g
r
ap
h
y
d
atab
ase
,
th
e
v
er
tices
co
u
ld
b
e
au
th
o
r
s
,
ar
ticles,
an
d
v
en
u
es
(
jo
u
r
n
al
s
/co
n
f
er
en
ce
s
)
an
d
ed
g
es
co
u
ld
b
e
t
h
e
au
th
o
r
-
au
t
h
o
r
r
elatio
n
s
h
ip
in
th
e
s
en
s
e
th
a
t
th
ey
h
a
v
e
wo
r
k
ed
in
th
e
s
am
e
ar
ea
,
an
d
atten
d
e
d
th
e
s
am
e
co
n
f
er
e
n
ce
s
.
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
N
ew p
r
ed
ictio
n
meth
o
d
fo
r
d
a
t
a
s
p
r
ea
d
in
g
in
s
o
cia
l n
etw
o
r
ks b
a
s
ed
o
n
… (
Ma
yth
a
m
N
.
Me
q
d
a
d
)
3333
Her
e,
we
m
o
d
el
in
f
o
r
m
atio
n
d
if
f
u
s
io
n
an
d
s
p
ec
if
ically
to
p
ic
d
if
f
u
s
io
n
in
h
eter
o
g
e
n
eo
u
s
in
f
o
r
m
atio
n
n
etwo
r
k
s
.
T
o
th
is
en
d
,
we
u
s
e
a
co
n
ce
p
t
ca
lled
m
eta
-
p
ath
.
T
h
e
m
eta
-
p
ath
p
o
n
th
e
g
r
id
T
G
=
(
A,
R
)
,
wh
er
e
A
an
d
R
r
ep
r
esen
t
v
er
tices a
n
d
r
elatio
n
s
h
ip
s
,
is
d
ef
in
ed
as f
o
ll
o
ws:
1
1
→
2
2
→
…
→
+
1
(
1
)
Her
e,
l
is
an
in
d
ex
o
f
th
e
m
e
ta
-
p
ath
.
T
h
e
s
u
m
m
atio
n
r
elati
o
n
s
h
ip
b
etwe
en
d
if
f
er
e
n
t
ty
p
e
s
o
f
v
er
tices
(
A1
-
Al+1
)
is
g
iv
en
b
y
:
=
1
∘
2
∘
…
(
2
)
wh
er
e
o
is
th
e
c
o
m
b
in
atio
n
o
p
er
ato
r
.
I
n
DB
L
P,
f
o
r
ex
am
p
l
e,
ea
ch
au
th
o
r
-
au
th
o
r
o
r
au
th
o
r
-
co
n
f
e
r
en
ce
-
a
u
th
o
r
r
elatio
n
s
h
ip
is
co
n
s
id
er
ed
a
s
in
g
le
m
eta
-
p
ath
.
Fig
u
r
e
1
s
h
o
ws
an
ex
am
p
le
o
f
th
e
d
if
f
u
s
io
n
o
f
th
e
to
p
ic
o
f
“d
ata
m
in
in
g
”
in
DB
L
P,
wh
er
e
a
u
th
o
r
s
ca
n
b
e
lin
k
ed
th
r
o
u
g
h
d
if
f
er
en
t
m
e
ta
-
p
ath
s
.
T
h
is
p
a
p
er
p
r
o
v
id
es
a
m
ac
h
in
e
lear
n
in
g
m
eth
o
d
b
a
s
ed
o
n
g
r
ap
h
n
e
u
r
al
n
etwo
r
k
s
in
wh
ic
h
an
in
ac
tiv
e
n
o
d
e
is
ac
tiv
ated
b
y
its
ac
tiv
e
n
eig
h
b
o
r
s
in
a
p
ar
ticu
lar
s
cien
tific
to
p
ic.
Giv
e
n
th
at
th
e
p
r
ed
ictio
n
o
f
n
ew
u
s
er
s
wh
o
will
b
e
in
t
h
e
p
ath
o
f
in
f
o
r
m
atio
n
f
lo
w
is
a
r
ec
o
g
n
itio
n
p
r
o
ce
s
s
,
th
is
p
r
o
b
lem
ca
n
b
e
s
o
l
v
ed
b
y
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
s
.
T
h
e
g
en
er
al
f
r
am
ewo
r
k
o
f
t
h
e
m
eth
o
d
co
n
s
is
ts
o
f
two
m
ain
p
h
ases
:
1
)
d
esig
n
in
g
a
m
ac
h
in
e
lear
n
in
g
s
ch
em
e
(
lear
n
in
g
m
ac
h
in
e
)
f
o
r
th
e
p
r
e
d
ictio
n
p
r
o
ce
s
s
,
an
d
2
)
ev
alu
atin
g
th
e
ac
cu
r
ac
y
o
f
th
e
d
esig
n
ed
s
ch
em
e
(
m
ac
h
in
e)
in
p
r
ed
ictin
g
th
e
f
lo
w
o
f
in
f
o
r
m
atio
n
in
th
e
d
at
aset
o
f
in
ter
est.
T
h
e
f
ir
s
t
s
te
p
in
v
o
lv
es
tr
ain
in
g
a
lear
n
in
g
m
ac
h
in
e,
wh
e
r
e
th
e
in
p
u
t
is
t
h
e
d
ata
co
llected
f
r
o
m
th
e
in
f
o
r
m
atio
n
n
etwo
r
k
g
r
a
p
h
an
d
th
e
o
u
t
p
u
t
is
th
e
tag
“Ye
s
”
o
r
“No
”,
s
h
o
win
g
wh
eth
er
o
r
n
o
t
th
e
n
o
d
e
s
p
ec
if
ied
in
th
e
in
p
u
t
will
b
e
s
elec
t
ed
as
th
e
n
ex
t
p
ath
o
f
in
f
o
r
m
atio
n
d
i
f
f
u
s
io
n
.
T
h
e
p
u
r
p
o
s
e
o
f
th
is
m
ac
h
i
n
e
is
to
cr
ea
te
a
r
e
g
r
ess
io
n
f
u
n
ctio
n
f
o
r
o
p
tim
al
m
ap
p
i
n
g
b
etwe
en
in
p
u
t d
ata
an
d
o
u
tp
u
t
tag
s
.
I
n
th
e
s
ec
o
n
d
p
h
ase,
a
test
d
ataset,
wh
ich
i
s
tak
en
f
r
o
m
th
e
co
llected
d
ata,
is
u
s
ed
to
test
th
e
d
esig
n
ed
m
ac
h
in
e.
I
n
th
e
test
in
g
an
d
ac
cu
r
ac
y
ev
alu
atio
n
p
h
ase,
th
e
class
if
icatio
n
p
r
o
c
ess
is
d
o
n
e
o
n
ce
r
a
n
d
o
m
ly
an
d
an
o
th
er
tim
e
with
th
e
d
esig
n
ed
m
ac
h
in
e.
I
n
th
e
en
d
,
th
e
q
u
ality
o
f
th
e
v
er
tices
o
b
tain
ed
f
r
o
m
th
ese
two
m
et
h
o
d
s
is
co
m
p
ar
ed
.
Fig
u
r
e
1
.
An
ex
am
p
le
o
f
a
h
et
er
o
g
en
e
o
u
s
n
etwo
r
k
[
1
5
]
2
.
1
.
G
ra
ph
co
nv
o
lutio
n net
wo
rk
a
lg
o
rit
hm
T
h
is
s
ec
tio
n
d
escr
ib
es
th
e
d
eta
ils
o
f
g
r
ap
h
c
o
n
v
o
lu
tio
n
al
n
etwo
r
k
(
GC
N
)
an
d
th
e
n
ex
t
s
ec
ti
o
n
ex
p
lain
s
h
o
w
it is
u
s
ed
in
th
e
lear
n
in
g
f
r
o
m
th
e
g
r
ap
h
d
ata
o
f
th
is
s
tu
d
y
.
Gr
ap
h
d
ata
ca
n
b
e
b
r
o
k
en
d
o
wn
in
to
two
m
ain
elem
en
ts
: v
er
tices v
ij a
n
d
ed
g
es a
ij.
A
g
r
ap
h
ca
n
b
e
d
escr
ib
e
d
b
y
th
e
f
o
llo
win
g
3
‑
t
u
p
le.
(
3
)
=
(
,
)
wh
er
e
ℝ
×
is
th
e
v
er
te
x
s
i
g
n
al
m
atr
i
x
d
escr
ib
i
n
g
N
v
e
r
tices
ea
ch
with
f
f
ea
tu
r
es,
ℝ
×
is
th
e
a
d
jace
n
cy
m
atr
ix
wh
ich
en
co
d
es
th
e
ed
g
e
s
in
f
o
r
m
atio
n
as
d
escr
ib
ed
in
s
ec
tio
n
2
,
an
d
ea
c
h
elem
en
t
A
is
d
ef
in
ed
as
f
o
llo
ws
;
a
ij
=
{
wi
j
,
if
t
h
e
r
e
is
an
e
dge
b
e
tw
e
e
n
i
a
n
d
j
0
,
ot
h
e
r
w
ise
(
4
)
An
ex
am
p
le
g
r
ap
h
an
d
its
v
er
t
ex
m
atr
ix
V
an
d
ad
jace
n
c
y
m
a
tr
ix
A
ar
e
s
h
o
wn
i
n
Fig
u
r
e
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
.
6
,
Dec
em
b
e
r
2
0
2
0
:
3
3
3
1
-
333
8
3334
Fig
u
r
e
2
.
An
ex
am
p
le
o
f
a
g
r
a
p
h
an
d
its
ad
jace
n
cy
m
atr
i
x
2
.
2
.
G
ra
ph
co
nv
o
lutio
n
Gr
ap
h
d
ata
ca
n
p
r
o
v
id
e
a
b
r
ie
f
r
ep
r
esen
tatio
n
o
f
in
f
o
r
m
atio
n
in
v
er
tices
an
d
ed
g
es.
T
o
p
r
o
ce
s
s
an
d
lear
n
th
is
in
f
o
r
m
atio
n
,
o
n
e
h
as
to
u
s
e
a
co
n
v
o
lu
tio
n
f
ilter
in
g
m
eth
o
d
to
f
ilter
b
o
th
v
er
te
x
in
f
o
r
m
atio
n
an
d
ed
g
e
in
f
o
r
m
atio
n
.
T
h
is
is
a
s
p
atia
l
ap
p
r
o
ac
h
r
ela
ted
to
t
h
e
g
r
ap
h
c
o
n
v
o
lu
tio
n
m
eth
o
d
,
wh
ich
u
s
es
th
e
lo
ca
l
n
eig
h
b
o
r
h
o
o
d
g
r
a
p
h
f
ilter
in
g
s
tr
ateg
y
.
T
h
e
g
r
a
p
h
co
n
v
o
l
u
tio
n
o
p
er
atio
n
is
b
ased
o
n
th
e
p
o
l
y
n
o
m
ials
o
f
th
e
ad
jace
n
cy
m
atr
ix
o
f
th
e
g
r
ap
h
.
=
ℎ
0
+
ℎ
1
1
+
ℎ
2
2
+
ℎ
3
3
+
⋯
+
ℎ
(
5
)
T
h
is
f
ilter
is
d
ef
in
e
d
as
th
e
k
th
d
eg
r
e
e
p
o
ly
n
o
m
ial
o
f
th
e
ad
jace
n
cy
m
at
r
ix
.
T
h
e
ex
p
o
n
en
t
o
f
th
is
p
o
ly
n
o
m
ial
en
co
d
es
th
e
n
u
m
b
er
o
f
s
tep
s
f
r
o
m
th
e
v
er
tex
o
f
i
n
ter
est,
wh
ich
is
m
u
ltip
lied
b
y
th
e
ass
u
m
ed
f
ilter
f
ac
to
r
s
.
T
h
e
s
ca
lar
f
ac
to
r
h
i
d
eter
m
in
es
h
o
w
m
u
c
h
ea
ch
n
ei
g
h
b
o
r
o
f
a
v
er
tex
c
o
n
tr
ib
u
tes
to
th
e
co
n
v
o
lu
tio
n
o
p
er
atio
n
.
T
h
er
ef
o
r
e,
th
e
f
ilte
r
m
atr
ix
is
o
b
tain
ed
as
∈
ℝ
×
.
T
h
e
co
n
v
o
lu
tio
n
o
f
t
h
e
v
er
tices
with
th
e
f
ilter
is
d
ef
in
ed
as th
e
f
o
llo
win
g
m
atr
ix
m
u
ltip
licatio
n
,
wh
er
e
,
∈
ℝ
.
=
(
6
)
T
h
is
m
o
d
el
ca
n
b
e
ad
ju
s
ted
in
th
r
ee
way
s
.
T
h
e
f
ir
s
t
way
is
to
av
o
id
A
b
ec
o
m
in
g
ex
p
o
n
e
n
tiated
an
d
s
im
p
lif
y
th
e
ad
jace
n
c
y
p
o
ly
n
o
m
ial
in
(
2
)
in
t
o
th
e
lin
ea
r
f
o
r
m
g
iv
en
i
n
(
6
)
.
T
h
e
r
ea
s
o
n
b
eh
in
d
t
h
is
ap
p
r
o
ac
h
is
th
at,
as
s
h
o
wn
b
y
VGGN
et,
a
ca
s
ca
d
e
o
f
f
ilter
s
ca
n
ef
f
ec
tiv
ely
esti
m
ate
th
e
r
ec
ep
tiv
e
f
ield
o
f
a
lar
g
e
f
ilter
.
≈
ℎ
0
+
ℎ
1
(
7)
T
h
e
n
e
x
t
s
t
e
p
i
s
t
o
c
r
e
a
t
e
t
h
e
a
d
j
a
c
e
n
c
y
t
e
n
s
o
r
.
T
h
i
s
t
e
n
s
o
r
c
o
n
s
i
s
t
s
o
f
m
u
l
t
i
p
l
e
a
d
j
a
c
e
n
c
y
m
a
t
r
i
c
e
s
A
,
w
h
i
c
h
a
r
e
t
h
e
s
l
i
c
e
s
o
f
t
h
i
s
t
e
n
s
o
r
,
e
a
c
h
e
n
c
o
d
i
n
g
a
s
p
e
c
i
f
i
c
e
d
g
e
f
e
a
t
u
r
e
.
T
h
e
r
e
f
o
r
e
,
t
h
e
l
i
n
e
a
r
f
i
l
t
e
r
m
a
t
r
i
x
i
n
(
6
)
i
s
d
e
f
i
n
e
d
a
s
a
c
o
n
v
e
x
c
o
m
b
i
n
a
t
i
o
n
o
f
a
d
j
a
c
e
n
c
y
m
a
t
r
i
c
e
s
a
s
g
i
v
e
n
i
n
(
7
)
.
T
h
i
s
e
q
u
a
t
i
o
n
c
a
n
b
e
s
i
m
p
l
i
f
i
e
d
i
n
t
o
(
8
)
.
=
ℎ
0
+
ℎ
1
1
+
ℎ
2
2
+
⋯
+
ℎ
−
1
(
8
)
≈
∑
h
A
e
=
0
(
9
)
Mu
ltip
le
ed
g
e
f
ea
t
u
r
es
ar
e
en
co
d
ed
b
y
m
u
ltip
le
ad
jace
n
cy
m
atr
ices,
ea
ch
o
f
w
h
ich
en
c
o
d
es
a
s
in
g
le
f
ea
tu
r
e.
Als
o
,
as
s
h
o
wn
in
Fig
u
r
e
3
,
th
e
ed
g
es
ar
e
s
u
b
d
iv
id
e
d
in
to
m
u
ltip
le
m
atr
ices.
Fig
u
r
e
3
s
h
o
ws
th
e
d
ef
au
lt
lin
ea
r
GC
N
f
ilter
in
an
im
ag
e
ap
p
licatio
n
.
A
f
ilter
f
ac
to
r
is
is
o
tr
o
p
ically
ap
p
lied
to
all
v
er
tices
at
a
g
iv
e
n
d
is
tan
ce
.
I
n
th
is
ca
s
e,
h
0
is
ap
p
lied
to
th
e
v
er
tex
o
f
th
e
0
th
s
t
ep
an
d
h
1
is
ap
p
lied
to
all
ad
ja
ce
n
t
v
er
tices.
I
f
t
h
is
f
ig
u
r
e
is
en
clo
s
ed
i
n
an
o
t
h
er
s
et
o
f
p
ix
els,
ea
ch
p
ix
el
in
t
h
at
s
et
will b
e
m
u
ltip
lied
b
y
th
e
f
ilter
f
ac
to
r
h
2
.
As
s
h
o
wn
in
Fig
u
r
e
3
,
to
cr
ea
t
e
th
e
a
d
j
ac
en
cy
ten
s
o
r
,
th
e
a
d
jace
n
cy
m
atr
i
x
ca
n
b
e
s
u
b
d
i
v
id
ed
in
t
o
9
ad
jace
n
cy
m
atr
ices.
E
ac
h
o
f
t
h
ese
ad
jace
n
cy
m
atr
ices
s
h
o
w
s
a
d
if
f
er
e
n
t
r
elativ
e
lin
k
(
ed
g
e
f
ea
tu
r
e)
to
a
g
i
v
en
v
er
tex
.
T
h
e
n
ex
t
s
tep
is
to
ap
p
ly
a
u
n
iq
u
e
f
ilter
to
ea
c
h
ad
jac
en
cy
m
atr
ix
to
p
e
r
f
o
r
m
co
n
v
o
lu
tio
n
f
o
llo
wed
b
y
ag
g
r
eg
atio
n
.
T
h
is
g
iv
es
a
d
ir
ec
tio
n
to
th
e
GC
N
f
ilter
.
T
h
is
is
eq
u
iv
alen
t
to
a
3
×3
FIR
f
ilter
i
n
tr
ad
itio
n
al
GC
Ns
.
All
th
e
ab
o
v
e
-
d
escr
ib
ed
f
ilter
s
ar
e
f
o
r
a
s
in
g
le
v
er
tex
f
ea
tu
r
e.
Fo
r
th
e
ex
ten
s
io
n
to
m
u
ltip
le
v
er
tex
f
ea
tu
r
es,
ea
c
h
h
l
m
u
s
t
b
e
i
n
ℝ
s
o
th
at
H
h
as
a
d
im
en
s
io
n
o
f
ℝ
×
×
.
T
h
er
ef
o
r
e,
ea
c
h
v
er
te
x
f
ea
tu
r
e
h
as
a
f
ilter
m
atr
ix
H
o
f
s
ize
N×
N.
T
h
e
r
ef
o
r
e,
i
n
(
9
)
ca
n
b
e
r
ewr
itten
as
i
n
(
1
0
)
,
wh
er
e
H(
c)
is
an
N×
N
s
lice
o
f
H
an
d
h
(
c)
is
a
s
ca
lar
r
elate
d
to
an
in
p
u
t f
ea
tu
r
e
an
d
a
s
lice
o
f
A
l
.
≈
∑
h
A
e
=
0
(
1
0
)
B
ased
o
n
(
c)
,
th
e
v
er
tex
s
ig
n
al
is
d
escr
ib
ed
b
y
i
n
(
1
1
)
,
wh
er
e
V
in
(
c)
is
a
co
lu
m
n
o
f
V
in
wh
er
e
c
is
th
e
o
n
ly
v
er
tex
f
ea
tu
r
e.
A
b
ias in
th
e
f
o
r
m
o
f
b
∈
ℝ
is
also
a
d
d
ed
to
th
e
f
o
r
m
u
latio
n
,
r
esu
lt
s
in
∈
ℝ
.
S
R
Q
P
0
1
1
0
P
0
0
1
1
Q
1
0
0
1
R
0
1
0
0
S
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
N
ew p
r
ed
ictio
n
meth
o
d
fo
r
d
a
t
a
s
p
r
ea
d
in
g
in
s
o
cia
l n
etw
o
r
ks b
a
s
ed
o
n
… (
Ma
yth
a
m
N
.
Me
q
d
a
d
)
3335
=
∑
V
c
=
1
+
(
11)
Fig
u
r
e
3
.
T
y
p
es o
f
f
ilter
s
th
at
ca
n
b
e
d
e
f
in
ed
in
th
e
GC
N
alg
o
r
ith
m
2
.
3
.
Arc
hite
ct
ure
us
ed
f
o
r
G
CN
T
h
is
ar
ch
itectu
r
e
o
f
C
NN
co
n
s
is
t
s
o
f
th
r
ee
g
e
n
er
al
s
tep
s
.
T
h
e
f
ir
s
t
s
tep
is
th
e
p
r
e
p
ar
atio
n
o
f
th
e
i
n
p
u
t
g
r
ap
h
,
w
h
ich
co
u
l
d
b
e
DB
L
P
f
o
r
ex
am
p
le,
s
o
th
at
th
e
ac
tiv
e
an
d
in
ac
tiv
e
v
e
r
tices
ar
e
eq
u
alize
d
an
d
lev
eled
.
Ho
wev
er
,
th
is
s
tep
also
h
as a
s
ep
ar
ate
in
p
u
t in
t
h
e
f
o
r
m
o
f
a
t
ag
.
T
h
e
o
u
t
p
u
t o
f
th
is
s
tep
is
t
h
e
g
r
a
p
h
0
,
wh
ich
will b
e
f
ed
to
th
e
n
ex
t step
.
T
h
e
s
ec
o
n
d
s
tep
is
to
p
er
f
o
r
m
th
e
co
n
v
o
lu
tio
n
o
p
e
r
atio
n
o
n
th
is
g
r
ap
h
in
s
ev
er
al
lay
er
s
.
I
n
th
is
s
tep
,
a
f
ilter
is
ap
p
lied
to
ea
ch
v
er
te
x
b
ased
o
n
its
ad
jace
n
t e
d
g
e.
T
h
e
in
p
u
t o
f
th
e
f
ir
s
t c
o
n
v
o
l
u
tio
n
lay
er
1
is
f
ed
to
th
e
s
ec
o
n
d
co
n
v
o
l
u
tio
n
lay
e
r
,
wh
ich
p
r
o
d
u
ce
s
th
e
o
u
tp
u
t
2
.
I
n
th
e
th
ir
d
s
tep
,
th
e
cr
ea
ted
f
e
atu
r
es
ar
e
lin
ea
r
ized
an
d
g
i
v
en
to
a
So
f
tm
ax
lay
er
,
wh
ich
d
ec
id
es
wh
ich
n
o
d
e
s
h
o
u
ld
b
e
ac
tiv
at
e
d
n
e
x
t.
Fo
r
t
h
is
p
u
r
p
o
s
e,
m
u
ltip
le
lin
ea
r
v
ec
to
r
s
ar
e
cr
ea
ted
b
y
co
n
ca
ten
atin
g
all
th
e
g
r
ap
h
s
cr
ea
ted
in
th
e
p
r
ev
i
o
u
s
s
tep
s
an
d
th
eir
d
if
f
er
en
t
p
e
r
m
u
tatio
n
s
f
r
o
m
0
to
2
.
E
ac
h
o
f
t
h
ese
v
ec
to
r
s
is
ca
lled
a
f
ea
tu
r
e
v
ec
to
r
.
T
h
e
o
u
tp
u
t
o
f
th
ese
o
p
er
atio
n
s
is
a
lin
ea
r
v
ec
to
r
th
at
is
f
e
d
to
th
e
f
u
ll
y
co
n
n
ec
ted
n
eu
r
al
n
etwo
r
k
p
lace
d
in
th
e
last
lay
er
.
Fin
ally
,
th
is
n
etwo
r
k
,
wh
ich
is
k
n
o
wn
as
So
f
tm
ax
,
is
u
s
ed
to
d
ec
id
e
o
n
th
e
n
ex
t a
ctiv
e
n
o
d
e.
2
.
4
.
I
m
ple
m
ent
a
t
io
n o
f
cla
s
s
if
ica
t
io
n a
lg
o
rit
h
m
I
n
th
e
p
r
o
p
o
s
ed
alg
o
r
ith
m
,
a
lear
n
in
g
m
ac
h
in
e
f
o
r
im
a
g
e
c
lass
if
icatio
n
in
th
e
d
ataset
p
r
o
ce
s
s
ed
in
th
e
class
if
icatio
n
p
h
ase
is
u
s
ed
to
co
n
s
tr
u
c
t
a
m
o
d
el
th
at
p
r
o
d
u
ce
s
v
alu
es
th
at
ar
e
as
clo
s
e
as
p
o
s
s
ib
le
to
th
e
ex
p
ec
ted
v
alu
es.
Sev
er
al
d
if
f
er
en
t
ap
p
r
o
ac
h
es
h
av
e
b
ee
n
d
ev
elo
p
e
d
f
o
r
th
is
p
u
r
p
o
s
e.
I
n
th
is
p
ap
er
,
class
if
icatio
n
alg
o
r
ith
m
s
ar
e
u
s
ed
t
o
class
if
y
th
e
u
s
er
i
n
to
d
i
f
f
er
en
t
class
es
in
th
e
d
ataset.
T
h
e
im
a
g
e
class
if
icatio
n
is
d
o
n
e
b
ased
o
n
th
e
f
ea
tu
r
e
v
ec
to
r
e
x
tr
ac
ted
in
th
e
p
r
ev
i
o
u
s
p
h
ase.
T
h
e
p
r
o
p
o
s
e
d
alg
o
r
ith
m
u
s
es
a
m
o
d
if
ied
v
er
s
io
n
o
f
th
e
s
tan
d
ar
d
alg
o
r
ith
m
d
escr
ib
ed
ab
o
v
e
f
o
r
class
if
icatio
n
.
Sin
ce
th
e
o
u
tp
u
t
o
f
th
e
p
r
o
p
o
s
ed
m
eth
o
d
is
tag
g
ed
in
two
class
es,
th
e
m
u
lti
-
class
v
er
s
io
n
o
f
th
es
e
alg
o
r
ith
m
s
is
u
s
ed
.
T
h
e
f
ea
tu
r
es
in
clu
d
ed
in
th
e
f
ea
tu
r
e
m
a
t
r
ix
o
f
ea
ch
im
ag
e
f
o
r
m
an
n
-
d
im
en
s
io
n
al
v
ec
to
r
,
wh
ich
b
elo
n
g
s
to
o
n
e
o
f
two
c
lass
es.
=
(
1
,
2
,
⋯
,
)
(
1
2
)
=
(
1
,
2
)
(
1
3
)
=
(
1
,
2
,
⋯
,
)
,
L
=
N
umb
e
r
of
Sa
mpl
e
(
1
4
)
T
h
e
alg
o
r
ith
m
p
r
esen
ted
in
Fig
u
r
e
4
s
h
o
ws
h
o
w
th
e
le
ar
n
i
n
g
m
ac
h
in
e
is
b
u
ilt
an
d
test
ed
.
I
n
th
e
p
r
o
p
o
s
ed
alg
o
r
ith
m
,
ea
ch
m
ac
h
in
e
is
f
ir
s
t
tr
ain
ed
u
s
in
g
th
e
ex
tr
ac
ted
d
ata.
As
ca
n
b
e
s
ee
n
,
ea
ch
lear
n
in
g
m
ac
h
in
e
is
tr
ain
e
d
s
ep
a
r
ately
f
o
r
ea
ch
d
ataset
ex
t
r
ac
ted
i
n
th
e
alg
o
r
ith
m
.
T
h
is
is
d
o
n
e
u
s
in
g
th
e
“g
en
e
r
ateM
L
”
f
u
n
ctio
n
in
lin
e
5
.
T
h
en
,
th
e
f
o
ld
in
g
alg
o
r
ith
m
is
u
s
ed
to
test
ea
ch
m
ac
h
in
e
.
T
h
e
r
ef
o
r
e,
ea
ch
m
ac
h
in
e
is
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
.
6
,
Dec
em
b
e
r
2
0
2
0
:
3
3
3
1
-
333
8
3336
r
ep
ea
ted
ly
s
u
b
d
iv
id
ed
an
d
t
r
ain
ed
an
d
test
ed
b
y
d
if
f
er
e
n
t
d
ata
s
ets.
Mo
r
e
d
etails
o
n
th
e
m
ec
h
an
is
m
o
f
th
e
f
o
ld
in
g
a
lg
o
r
ith
m
an
d
m
ea
s
u
r
em
e
n
t o
f
th
e
ac
cu
r
ac
y
o
f
t
h
e
lear
n
in
g
m
ac
h
in
e
ar
e
p
r
o
v
id
ed
in
th
e
n
ex
t sectio
n
.
F
ig
u
r
e
4
.
P
s
eu
d
o
-
co
d
e
o
f
th
e
c
lass
if
icatio
n
s
tag
e
3.
T
E
ST
S
3
.
1
.
T
est
prepa
ra
t
io
n
T
h
e
p
r
o
p
o
s
ed
m
eth
o
d
was
test
ed
o
n
th
r
ee
r
ea
l
d
atasets
,
n
am
ely
DB
L
P,
Pu
b
m
ed
,
an
d
C
o
r
a,
wh
ich
h
av
e
b
ee
n
u
s
ed
in
n
u
m
er
o
u
s
em
p
ir
i
ca
l
s
tu
d
ies.
DB
L
P
:
th
is
is
a
co
m
p
u
ter
s
cien
ce
b
ib
lio
g
r
a
p
h
y
d
atab
ase
co
n
tain
in
g
th
e
n
am
e
o
f
m
ajo
r
au
th
o
r
s
,
c
o
n
f
er
e
n
ce
s
,
an
d
p
u
b
licatio
n
s
.
I
n
th
e
n
etwo
r
k
u
s
ed
f
o
r
DB
L
P,
o
b
jects
r
ep
r
esen
t
au
th
o
r
s
.
T
h
e
m
eta
-
p
ath
s
co
n
s
id
er
ed
in
t
h
is
n
etwo
r
k
ar
e
au
th
o
r
-
p
ap
er
-
au
th
o
r
(
APA
)
,
a
u
th
o
r
-
p
ap
er
-
au
th
o
r
-
p
ap
er
-
a
u
th
o
r
(
APAPA
)
,
au
th
o
r
-
co
n
f
e
r
en
ce
-
a
u
th
o
r
(
AC
A
)
,
an
d
au
th
o
r
-
co
n
f
er
en
ce
-
au
th
o
r
-
co
n
f
er
e
n
ce
-
au
th
o
r
(
AC
AC
A
)
.
T
h
is
d
ataset
is
ty
p
ically
u
s
ed
to
ex
tr
ac
t
d
if
f
er
e
n
t
to
p
ics
an
d
ex
am
i
n
e
th
e
d
if
f
u
s
io
n
o
f
in
f
o
r
m
atio
n
ab
o
u
t a
s
p
ec
if
ic
to
p
ic.
I
n
f
o
r
m
a
tio
n
co
n
tain
e
d
in
th
is
d
ataset
p
er
tain
s
to
th
e
p
er
io
d
b
etwe
en
1
9
5
4
a
n
d
2
0
1
6
.
Pu
b
m
ed
:
th
is
is
a
b
ib
lio
g
r
ap
h
y
d
ataset
f
o
r
th
e
f
ield
o
f
m
ed
ical
s
cien
ce
s
,
wh
ich
in
clu
d
es
au
th
o
r
s
,
co
n
f
er
e
n
ce
s
,
an
d
p
u
b
licatio
n
s
.
I
n
th
is
n
etwo
r
k
u
s
ed
f
o
r
Pu
b
m
ed
,
a
u
th
o
r
s
ar
e
r
ep
r
esen
t
ed
b
y
o
b
jects
an
d
th
e
co
n
s
id
er
ed
m
eta
-
p
at
h
s
ar
e
APAPA
an
d
APA.
I
n
f
o
r
m
atio
n
o
f
th
is
d
a
taset
is
f
o
r
th
e
p
er
io
d
b
etwe
en
1
9
9
4
an
d
2
0
0
3
.
C
o
r
a
:
t
h
is
is
an
o
th
e
r
co
m
p
u
ter
s
cien
ce
b
ib
lio
g
r
ap
h
y
d
ata
b
ase.
T
h
e
m
eta
-
p
at
h
s
u
s
ed
f
o
r
th
is
d
ataset
ar
e
APAPA a
n
d
APA.
T
h
is
d
ataset
co
n
tain
s
in
f
o
r
m
atio
n
f
r
o
m
1
9
9
0
to
2
0
1
2
.
I
n
t
h
e
e
v
a
l
u
at
i
o
n
p
r
o
c
e
s
s
,
t
h
e
d
i
f
f
u
s
i
o
n
p
r
o
c
es
s
w
as
m
o
d
e
le
d
f
o
r
s
e
v
e
r
al
t
o
p
i
cs
c
o
n
t
a
i
n
ed
i
n
t
h
es
e
d
a
t
a
s
et
s
,
w
h
ic
h
i
n
c
l
u
d
e
d
a
t
a
m
in
i
n
g
,
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
,
s
o
c
i
a
l
n
e
t
w
o
r
k
s
,
h
e
a
l
t
h
c
a
r
e
,
D
N
A
a
n
d
i
n
f
e
c
t
i
o
u
s
d
i
s
e
a
s
e
.
T
h
e
s
e
p
a
r
t
ic
u
l
a
r
t
o
p
i
cs
w
e
r
e
s
el
e
c
t
e
d
b
e
c
a
u
s
e
o
f
t
h
e
i
r
h
i
g
h
f
r
eq
u
e
n
c
y
i
n
t
h
e
d
a
t
a
s
et
a
n
d
t
h
e
co
n
s
i
d
e
r
a
b
l
e
a
m
o
u
n
t
o
f
d
a
t
a
a
v
a
i
l
a
b
l
e
f
o
r
c
o
m
p
a
r
is
o
n
a
n
d
c
o
n
c
l
u
s
i
o
n
.
T
r
a
i
n
i
n
g
a
n
d
t
es
t
i
n
g
o
p
e
r
a
ti
o
n
s
w
e
r
e
p
e
r
f
o
r
m
e
d
b
y
t
h
e
u
s
e
o
f
th
e
K
-
F
o
l
d
m
et
h
o
d
a
s
d
es
c
r
i
b
e
d
e
a
r
l
i
e
r
i
n
t
h
e
p
a
p
e
r
.
I
n
t
h
i
s
m
e
t
h
o
d
,
d
a
t
a
i
s
p
a
r
t
it
i
o
n
e
d
i
n
t
o
K
s
u
b
s
et
s
.
E
a
c
h
t
i
m
e
,
o
n
e
o
f
t
h
e
s
e
K
s
u
b
s
e
ts
is
u
s
e
d
f
o
r
t
es
t
i
n
g
a
n
d
t
h
e
o
t
h
e
r
K
-
1
a
r
e
u
s
e
d
f
o
r
t
r
a
i
n
i
n
g
.
T
h
is
p
r
o
c
e
d
u
r
e
is
r
e
p
ea
t
e
d
k
t
i
m
e
s
s
o
t
h
a
t
e
ac
h
d
a
t
a
is
u
s
e
d
e
x
a
c
tl
y
o
n
c
e
f
o
r
t
r
a
i
n
i
n
g
a
n
d
o
n
c
e
f
o
r
t
e
s
t
i
n
g
.
I
n
t
h
e
e
n
d
,
t
h
e
a
v
e
r
a
g
e
r
e
s
u
l
t
o
f
t
h
e
s
e
K
t
e
s
t
s
is
r
e
p
o
r
t
e
d
a
s
a
f
i
n
a
l
e
s
t
i
m
at
e
.
I
n
t
h
e
K
-
F
o
l
d
m
et
h
o
d
,
t
h
e
r
a
t
i
o
o
f
d
a
t
a
o
f
c
l
a
s
s
es
i
n
e
a
c
h
s
u
b
s
et
s
h
o
u
l
d
m
a
t
c
h
t
h
i
s
r
at
i
o
i
n
t
h
e
m
a
i
n
s
e
t
.
Fin
ally
,
th
e
p
er
f
o
r
m
a
n
ce
o
f
th
e
m
eth
o
d
in
p
r
e
d
ictin
g
to
p
i
c
d
if
f
u
s
io
n
was
ev
alu
ated
in
ter
m
s
o
f
th
e
cr
iter
io
n
k
n
o
wn
as Pr
ec
is
i
o
n
.
T
h
is
cr
iter
io
n
was c
alcu
lated
u
s
in
g
th
e
f
o
llo
win
g
d
ef
in
iti
o
n
s
:
−
T
P: if
an
ac
tiv
e
n
o
d
e
is
co
r
r
ec
t
ly
lab
eled
as a
ctiv
e
−
T
N:
if
an
in
ac
tiv
e
n
o
d
e
is
co
r
r
ec
tly
lab
eled
as in
ac
tiv
e
−
FP
: if
an
ac
tiv
e
n
o
d
e
is
in
c
o
r
r
e
ctly
lab
eled
as in
ac
tiv
e
−
FN: if
an
in
ac
tiv
e
n
o
d
e
is
in
co
r
r
ec
tly
lab
eled
as a
ctiv
e
T
ab
le
1
p
r
esen
ts
th
e
p
a
r
am
ete
r
s
o
f
th
e
GC
N
alg
o
r
ith
m
illu
s
t
r
ated
in
Fig
u
r
e
4
.
All
test
s
o
f
th
is
s
tu
d
y
wer
e
p
er
f
o
r
m
ed
with
th
ese
p
a
r
am
eter
s
ettin
g
s
.
I
n
th
is
tab
le,
h
id
d
en
1
an
d
h
i
d
d
en
2
ar
e
th
e
n
u
m
b
er
o
f
n
o
d
es
in
th
e
two
co
n
v
o
lu
tio
n
lay
e
r
s
.
Als
o
,
ea
r
ly
_
s
to
p
p
in
g
r
ef
er
s
to
t
h
e
ea
r
ly
ter
m
in
atio
n
co
n
d
itio
n
o
f
th
e
alg
o
r
ith
m
,
wh
ich
is
co
n
v
er
g
en
ce
in
less
th
an
1
0
iter
atio
n
s
.
3
.
2
.
Co
m
pa
riso
n wit
h o
t
her
wo
rk
s
Fo
r
f
u
r
th
e
r
ev
alu
atio
n
o
f
th
e
p
r
o
p
o
s
ed
m
eth
o
d
,
it
was
co
m
p
ar
ed
with
o
t
h
er
m
eth
o
d
s
in
t
h
e
f
ield
o
f
in
f
o
r
m
atio
n
d
if
f
u
s
io
n
.
T
h
is
co
m
p
ar
is
o
n
was
m
ad
e
with
two
m
eth
o
d
s
,
h
eter
o
g
en
eo
u
s
p
r
o
b
ab
ilit
y
m
o
d
el
-
i
n
d
ep
e
n
d
en
t
ca
s
ca
d
e
(
HPM
-
IC
)
,
h
eter
o
g
e
n
eo
u
s
p
r
o
b
ab
ilit
y
m
o
d
el
-
l
in
ea
r
th
r
esh
o
ld
(
HPM
-
LT
)
[
1
5
]
an
d
m
u
lti
-
r
elatio
n
al
lin
ea
r
t
r
esh
o
ld
Mo
d
el
-
r
elatio
n
lev
el
ag
g
r
eg
atio
n
(
MLTM
-
R
)
[
1
6
]
,
w
h
ich
ar
e
b
ased
o
n
p
r
o
b
a
b
ilis
tic
f
u
n
ctio
n
s
.
T
h
ese
m
eth
o
d
s
wer
e
im
p
lem
en
ted
o
n
th
e
test
d
ataset
s
u
s
in
g
th
e
s
ett
in
g
s
r
ec
o
m
m
en
d
ed
in
th
e
r
esp
ec
tiv
e
r
e
f
er
en
ce
s
.
T
h
e
o
u
t
p
u
t
o
f
t
h
ese
m
eth
o
d
s
w
as
also
th
e
in
f
o
r
m
atio
n
d
if
f
u
s
io
n
p
at
h
f
o
r
s
ev
e
r
al
-
:
1
_
ℎ
=
{'
GCN
'}
2
=
{'
DBLP
',
P
ubmet
','C
ora
'}
3
for
e
ac
h
Do
4
for
e
ac
h
Do
5
for
e
ac
h
Do
6 (
Xtra
in,
Ytra
in, Xtest, Ytest)
m
ak
e
fold
(
)
7 ML
ge
n
e
r
ate
M
L
(
_
ℎ
,Xtra
in,
Ytra
in)
8 a
c
c
ura
c
y(
)
c
al
c
u
late
Ac
c
u
r
ac
y
(ML
,Yte
st,)
9
e
n
d
10
e
n
d
11
e
n
d
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
N
ew p
r
ed
ictio
n
meth
o
d
fo
r
d
a
t
a
s
p
r
ea
d
in
g
in
s
o
cia
l n
etw
o
r
ks b
a
s
ed
o
n
… (
Ma
yth
a
m
N
.
Me
q
d
a
d
)
3337
s
cien
tific
to
p
ics.
I
n
th
is
s
ec
tio
n
,
th
e
test
r
esu
lts
f
o
r
s
ev
er
al
t
o
p
ics
o
n
DB
L
P
an
d
Pu
b
m
e
d
d
atasets
ar
e
r
ep
o
r
ted
b
ased
o
n
p
r
ec
is
io
n
an
d
r
ec
all
cr
iter
ia.
T
ab
le
2
p
r
esen
t
s
th
e
r
esu
lts
o
f
th
is
ev
alu
atio
n
in
ter
m
s
o
f
p
r
ec
is
io
n
.
T
ab
le
1
.
I
n
itial p
ar
am
eter
s
o
f
GC
N
alg
o
r
ith
m
P
a
r
a
me
t
e
r
V
a
l
u
e
l
e
a
r
n
i
n
g
_
r
a
t
e
0
.
0
1
e
p
o
c
h
s
2
0
0
h
i
d
d
e
n
1
16
h
i
d
d
e
n
1
16
d
r
o
p
o
u
t
0
.
5
w
e
i
g
h
t
_
d
e
c
a
y
1
0
0
0
0
e
a
r
l
y
_
s
t
o
p
p
i
n
g
10
T
ab
le
2
.
R
esu
lts
o
f
co
m
p
a
r
is
o
n
o
f
th
e
p
r
o
p
o
s
ed
m
et
h
o
d
with
o
th
er
m
eth
o
d
s
f
o
r
d
if
f
er
en
t to
p
ics o
n
DB
L
P
u
s
in
g
th
e
p
r
ec
is
io
n
cr
iter
io
n
S
u
b
j
e
c
t
A
c
c
u
r
a
c
y
(
%)
G
C
N
H
P
M
-
IC
H
P
M
-
LT
M
TL
M
-
R
D
a
t
a
M
i
n
i
n
g
75
%
60
%
55
%
56
%
M
a
c
h
i
n
e
L
e
a
r
n
i
n
g
50
%
48
%
32
%
37
%
S
o
c
i
a
l
N
e
t
w
o
r
k
50
%
40
%
38
%
39
%
M
e
d
i
c
a
l
C
a
r
e
75
%
62
%
55
%
56
%
DNA
15
%
14
%
12
%
11
%
i
n
f
e
c
t
i
o
u
s
d
i
sea
s
e
25
%
21
%
20
%
22
%
S
o
f
t
w
a
r
e
E
n
g
i
n
e
e
r
i
n
g
30
%
25
%
20
%
10
%
B
i
g
D
a
t
a
25
%
22
%
21
%
14
%
N
e
t
w
o
r
k
25
%
21
%
19
%
16
%
G
e
n
e
t
i
c
75
%
50
%
40
%
33
%
B
i
o
l
o
g
y
50
%
37
%
11
%
10
%
N
e
u
r
a
l
e
t
w
o
r
k
25
%
22
%
21
%
20
%
A
s
t
h
es
e
r
es
u
l
ts
i
n
d
i
c
at
e
,
t
h
e
p
r
o
p
o
s
e
d
m
e
t
h
o
d
a
c
h
i
e
v
e
d
a
s
i
g
n
i
f
i
c
a
n
t
i
m
p
r
o
v
e
m
e
n
t
r
a
n
g
i
n
g
b
e
t
w
e
e
n
1
0
%
a
n
d
2
0
%
i
n
a
l
l
c
o
m
p
a
r
i
s
o
n
s
.
I
n
t
h
e
DB
L
P
d
a
t
as
e
t
,
f
o
r
e
x
a
m
p
l
e
,
t
h
e
p
r
o
p
o
s
e
d
m
et
h
o
d
h
a
s
a
1
0
%
h
i
g
h
e
r
p
r
e
c
i
s
i
o
n
t
h
a
n
o
t
h
e
r
m
e
t
h
o
d
s
.
T
h
e
r
e
a
s
o
n
f
o
r
t
h
is
i
m
p
r
o
v
e
m
en
t
c
o
u
l
d
b
e
t
h
a
t
o
t
h
e
r
m
e
t
h
o
d
s
a
r
e
p
u
r
e
l
y
b
a
s
e
d
o
n
p
r
o
b
a
b
i
l
i
t
y
f
u
n
c
t
i
o
n
s
a
n
d
c
al
c
u
l
a
t
i
o
n
o
f
p
r
o
b
a
b
i
l
i
t
y
b
et
w
e
e
n
n
e
i
g
h
b
o
r
i
n
g
v
e
r
t
i
c
e
s
.
T
h
i
s
m
e
a
n
s
t
h
a
t
t
h
e
s
e
m
e
t
h
o
d
s
h
a
v
e
n
o
s
u
c
h
t
h
i
n
g
a
s
f
e
a
t
u
r
e
l
e
a
r
n
i
n
g
o
r
i
n
t
e
l
l
i
g
e
n
t
p
r
o
c
ess
e
s
a
n
d
o
n
l
y
r
e
p
e
a
t
a
c
o
n
s
t
a
n
t
s
e
t
o
f
c
a
l
c
u
l
at
i
o
n
s
.
I
n
c
o
n
t
r
a
s
t
,
as
e
x
p
l
a
i
n
e
d
i
n
th
e
d
e
s
c
r
i
p
t
i
o
n
o
f
t
h
e
a
r
c
h
i
t
e
c
tu
r
e
,
t
h
e
p
r
o
p
o
s
e
d
m
e
t
h
o
d
u
s
es
d
i
f
f
e
r
e
n
t
l
ea
r
n
i
n
g
o
p
e
r
a
t
i
o
n
s
f
o
r
e
a
c
h
s
e
g
m
e
n
t
.
Fo
r
e
x
a
m
p
l
e
,
t
h
e
c
o
n
v
o
l
u
t
i
o
n
f
u
n
c
t
i
o
n
i
s
d
es
i
g
n
e
d
t
o
d
et
e
r
m
i
n
e
t
h
e
r
e
l
a
ti
o
n
o
f
e
a
c
h
n
o
d
e
t
o
i
t
s
n
e
i
g
h
b
o
r
s
t
h
r
o
u
g
h
a
l
e
a
r
n
i
n
g
p
r
o
c
e
s
s
.
T
h
e
s
e
o
p
e
r
a
ti
o
n
s
a
r
e
l
e
a
r
n
e
d
i
n
t
e
l
li
g
e
n
t
l
y
d
u
r
i
n
g
t
h
e
e
v
o
l
u
t
i
o
n
o
f
t
h
e
G
C
N
a
l
g
o
r
i
t
h
m
.
I
t
s
h
o
u
l
d
a
l
s
o
b
e
n
o
t
e
d
t
h
a
t
i
n
l
e
a
r
n
i
n
g
a
l
g
o
r
i
t
h
m
s
,
t
h
e
e
n
t
i
r
e
p
r
o
b
l
e
m
s
p
a
c
e
c
a
n
b
e
e
a
s
il
y
e
x
p
l
o
r
e
d
,
w
h
e
r
e
a
s
,
i
n
p
r
o
b
a
b
i
lit
y
f
u
n
c
t
i
o
n
-
b
a
s
e
d
m
e
t
h
o
d
s
,
o
n
ly
a
p
a
r
t
o
f
t
h
e
p
r
o
b
l
e
m
s
p
a
c
e
ca
n
b
e
s
e
a
r
c
h
e
d
.
4.
CO
NCLU
SI
O
N
T
h
is
p
ap
er
p
r
esen
te
d
a
m
ac
h
in
e
lear
n
in
g
m
eth
o
d
b
ased
o
n
th
e
g
r
ap
h
n
eu
r
al
n
etwo
r
k
alg
o
r
it
h
m
,
wh
ic
h
in
v
o
lv
es
th
e
s
elec
tio
n
o
f
in
ac
t
iv
e
v
er
tices
b
ased
o
n
th
eir
n
ei
g
h
b
o
r
in
g
ac
tiv
e
v
er
tices
in
ea
ch
s
cien
tific
to
p
ic.
B
asically
,
in
th
i
s
m
eth
o
d
,
in
f
o
r
m
atio
n
d
if
f
u
s
io
n
p
ath
s
ar
e
p
r
ed
icted
th
r
o
u
g
h
th
e
ac
tiv
atio
n
o
f
in
ac
tiv
e
v
er
tices
b
y
ac
tiv
e
v
e
r
tices.
Sin
ce
p
r
ed
i
ctin
g
th
e
n
ew
u
s
er
s
wh
o
will
b
e
in
th
e
p
ath
o
f
in
f
o
r
m
atio
n
f
lo
w
is
a
r
ec
o
g
n
itio
n
p
r
o
ce
s
s
,
th
is
p
r
o
b
lem
ca
n
b
e
s
o
lv
ed
b
y
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
s
.
T
h
e
p
r
o
p
o
s
ed
m
eth
o
d
was
tes
ted
o
n
th
r
ee
r
ea
l
d
atasets
,
DB
L
P,
Pu
b
m
ed
,
an
d
C
o
r
a,
wh
ic
h
ar
e
ex
te
n
s
iv
ely
u
s
ed
in
th
e
em
p
ir
ical
s
tu
d
i
es.
T
h
e
ev
alu
atio
n
p
r
o
ce
s
s
in
v
o
lv
ed
m
o
d
elin
g
th
e
d
if
f
u
s
io
n
p
r
o
ce
s
s
f
o
r
s
ev
er
al
to
p
ics
co
n
tain
ed
in
th
ese
d
ata
s
ets,
in
clu
d
in
g
d
ata
m
in
in
g
,
m
ac
h
in
e
lear
n
in
g
,
s
o
c
ial
n
etwo
r
k
s
,
h
ea
lth
ca
r
e
,
DNA
an
d
in
f
ec
tio
u
s
d
is
ea
s
e.
T
est
r
esu
lts
s
h
o
wed
th
at
t
h
e
p
r
o
p
o
s
ed
m
eth
o
d
o
u
tp
er
f
o
r
m
s
o
th
er
m
et
h
o
d
s
i
n
th
is
a
r
ea
.
As
a
p
o
ten
tial
id
ea
f
o
r
f
u
tu
r
e
s
tu
d
ies,
th
e
p
r
o
p
o
s
ed
s
y
s
tem
ca
n
b
e
im
p
lem
e
n
ted
in
a
p
ar
allel
p
latf
o
r
m
o
r
with
th
e
e
x
tr
ac
tio
n
an
d
co
m
b
in
atio
n
o
f
o
th
e
r
f
ea
tu
r
es
to
r
ea
ch
a
s
tr
o
n
g
er
s
y
s
tem
.
T
h
e
u
s
e
o
f
m
o
r
e
r
o
b
u
s
t
m
ac
h
in
e
lear
n
in
g
co
n
ce
p
ts
m
ay
also
en
h
a
n
ce
th
e
q
u
ality
o
f
th
e
m
et
h
o
d
.
T
h
e
m
et
h
o
d
s
with
p
o
s
s
ib
le
b
en
ef
its
in
th
is
ar
ea
in
clu
d
e
f
ea
tu
r
e
r
ed
u
ctio
n
an
d
f
ea
tu
r
e
lear
n
in
g
.
T
h
e
f
e
atu
r
e
r
ed
u
ctio
n
m
eth
o
d
is
p
ar
ticu
lar
ly
u
s
ef
u
l
f
o
r
r
ed
u
cin
g
th
e
o
v
e
r
all
co
m
p
lex
ity
o
f
th
e
r
ec
o
g
n
itio
n
m
eth
o
d
.
Featu
r
e
lear
n
in
g
is
a
p
r
o
ce
s
s
in
v
o
lv
in
g
th
e
tr
a
n
s
f
er
o
f
th
e
d
ata
p
r
o
ce
s
s
in
g
f
r
o
m
th
e
o
r
ig
in
al
f
ea
tu
r
e
s
p
ac
e
to
a
n
ew
s
p
ac
e
with
h
i
g
h
er
f
e
atu
r
e
r
eso
lu
tio
n
.
RE
F
E
R
E
NC
E
S
[
1
]
E.
Ba
k
s
h
y
,
I.
Ro
se
n
n
,
C.
M
a
rlo
w,
L.
Ad
a
m
ic,
“
T
h
e
ro
le
o
f
so
c
ial
n
e
two
rk
s i
n
in
f
o
rm
a
ti
o
n
d
iu
si
o
n
,
”
in
:
Pro
c
e
e
d
in
g
s
o
f
t
h
e
2
1
st i
n
ter
n
a
ti
o
n
a
l
c
o
n
fer
e
n
c
e
o
n
W
o
rld
W
id
e
W
e
b
,
AC
M
,
p
p
.
5
1
9
,
2
0
1
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
.
6
,
Dec
em
b
e
r
2
0
2
0
:
3
3
3
1
-
333
8
3338
[
2
]
M
.
S
.
G
ra
n
o
v
e
tt
e
r,
“
Th
e
stre
n
g
t
h
o
f
we
a
k
ti
e
s,”
In
S
o
c
ial
n
e
two
r
k
s,
Ac
a
d
e
m
ic P
re
ss
.
p
p
.
3
4
7
-
3
6
7
,
1
9
7
7
.
[
3
]
Y.
Hu
,
R.
J.
S
o
n
g
,
M
.
Ch
e
n
,
“
M
o
d
e
li
n
g
fo
r
I
n
fo
rm
a
ti
o
n
Di
u
si
o
n
in
On
li
n
e
S
o
c
ial
Ne
two
rk
s
v
ia
Hy
d
r
o
d
y
n
a
m
ics
,
”
IEE
E
Acc
e
ss
,
v
o
l
.
5
,
2
0
1
7
.
[
4
]
K.
Ik
e
d
a
,
e
t
a
l.
,
“
M
u
l
ti
-
a
g
e
n
t
i
n
fo
rm
a
ti
o
n
d
iff
u
sio
n
m
o
d
e
l
fo
r
twit
t
e
r
,”
In
Pro
c
e
e
d
in
g
s
o
f
t
h
e
2
0
1
4
I
EE
E/
W
IC/A
C
M
In
ter
n
a
t
io
n
a
l
J
o
in
t
Co
n
fer
e
n
c
e
s
o
n
W
e
b
In
telli
g
e
n
c
e
(W
I)
a
n
d
I
n
tell
ig
e
n
t
A
g
e
n
t
T
e
c
h
n
o
lo
g
ies
(IA
T
)
-
I
EE
E
Co
m
p
u
te
r
S
o
c
iety
v
ol
.
1
,
p
p
.
2
1
-
2
6
,
2
0
1
4
.
[
5
]
T.
Kip
f
,
N.
Th
o
m
a
s,
a
n
d
M
.
Well
in
g
,
“
S
e
m
i
-
S
u
p
e
rv
ise
d
Clas
sifica
ti
o
n
with
G
ra
p
h
Co
n
v
o
lu
ti
o
n
a
l
Ne
two
rk
s,
”
a
rXiv
p
re
p
rin
t
a
rXiv
:
1
6
0
9
.
0
2
9
0
7
,
2
0
1
6
.
[
6
]
Y.
M
o
re
n
o
,
R
.
P
a
sto
r
-
S
a
to
rra
s,
A.
Ve
sp
ig
n
a
n
i,
“
Ep
i
d
e
m
ic
o
u
tb
re
a
k
s
i
n
c
o
m
p
lex
h
e
ter
o
g
e
n
e
o
u
s
n
e
two
r
k
s,
”
T
h
e
Eu
r
o
p
e
a
n
P
h
y
sic
a
l
J
o
u
rn
a
l
B
,
v
o
l.
26
,
n
o
.
4
,
2
0
0
2
.
[
7
]
R.
Ya
n
g
,
B
.
-
H.
Wan
g
,
J.
Re
n
,
W.
J.
Ba
i,
Z
.
W
.
S
h
i
,
W
.
X.
W
a
n
g
,
T.
Z
h
o
u
,
“
Ep
i
d
e
m
ic
sp
re
a
d
i
n
g
o
n
h
e
tero
g
e
n
e
o
u
s
n
e
two
rk
s wit
h
i
d
e
n
ti
c
a
l
i
n
fe
c
ti
v
it
y
,
”
Ph
y
sic
s L
e
tt
e
rs
A
,
v
o
l.
3
6
4
,
p
p
.
3
-
4
,
2
0
0
7
.
[
8
]
M
.
S
a
leh
i,
R.
S
h
a
rm
a
,
M
.
M
a
rz
o
ll
a
,
M
.
M
a
g
n
a
n
i,
P
.
S
iy
a
ri,
D.
M
o
n
tes
i,
“
S
p
re
a
d
in
g
p
ro
c
e
ss
e
s
in
m
u
lt
il
a
y
e
r
n
e
two
rk
s,
”
IE
EE
T
r
a
n
sa
c
ti
o
n
s
o
n
Ne
two
rk
S
c
ien
c
e
a
n
d
E
n
g
in
e
e
rin
g
,
v
o
l.
2
,
n
o
.
2
,
2
0
1
5
.
[
9
]
L.
Wan
g
,
G
.
Z.
Da
i,
“
G
lo
b
a
l
sta
b
il
it
y
o
f
v
iru
s
sp
re
a
d
i
n
g
in
c
o
m
p
lex
h
e
tero
g
e
n
e
o
u
s
n
e
two
rk
s,
”
S
ia
m
J
o
u
rn
a
l
o
n
Ap
p
li
e
d
M
a
t
h
e
ma
ti
c
s
,
v
o
l.
68
,
n
o
.
5
,
2
0
0
8
.
[
1
0
]
M
.
Na
d
in
i
,
K.
S
u
n
,
E.
Ub
a
l
d
i,
M
.
S
ta
rn
i
n
i,
A.
Rizz
o
,
N.
P
e
rra
,
“
Ep
id
e
m
ic
sp
re
a
d
in
g
i
n
m
o
d
u
l
a
r
ti
m
e
-
v
a
ry
in
g
n
e
two
rk
s,
”
a
rXiv
p
re
p
r
in
t
a
rXiv
:1
7
1
0
.
0
1
3
5
5
,
2
0
1
7
.
[
1
1
]
G
.
D
e
m
i
r
e
l
,
E
.
B
a
r
t
e
r
,
T
.
G
r
o
s
s
,
“
D
y
n
a
m
i
c
s
o
f
e
p
i
d
e
m
i
c
d
i
s
e
a
s
e
s
o
n
a
g
r
o
w
i
n
g
a
d
a
p
t
i
v
e
n
e
t
w
o
r
k
,
”
S
c
i
e
n
t
i
c
r
e
p
o
r
t
s
,
v
o
l
.
7
,
2
0
1
7
.
[
1
2
]
P
.
S
e
rm
p
e
z
is,
T.
S
p
y
r
o
p
o
u
l
o
s,
“
In
fo
rm
a
ti
o
n
d
iffu
si
o
n
in
h
e
tero
g
e
n
e
o
u
s
n
e
tw
o
rk
s:
T
h
e
c
o
n
g
u
ra
ti
o
n
m
o
d
e
l
a
p
p
ro
a
c
h
,
”
in
:
Pro
c
e
e
d
i
n
g
s
-
IEE
E
INFOCO
M
,
p
p
.
3
2
6
1
,
2
0
1
3
.
[
1
3
]
Y.
Zh
o
u
,
L
.
Li
u
,
“
S
o
c
ial
in
u
e
n
c
e
b
a
se
d
c
lu
ste
rin
g
o
f
h
e
ter
o
g
e
n
e
o
u
s
in
f
o
rm
a
ti
o
n
n
e
two
r
k
s,
”
in
:
Pr
o
c
e
e
d
in
g
s
o
f
th
e
1
9
th
ACM
S
IGKD
D i
n
ter
n
a
ti
o
n
a
l
c
o
n
fer
e
n
c
e
o
n
K
n
o
wle
d
g
e
d
i
sc
o
v
e
ry
a
n
d
d
a
t
a
min
i
n
g
,
AC
M
,
p
p
.
3
3
8
,
2
0
1
3
.
[
1
4
]
S
.
M
o
lae
i,
S
.
Ba
b
a
e
i,
M
.
S
a
leh
i,
M
.
Ja
li
li
,
“
In
fo
rm
a
ti
o
n
s
p
re
a
d
a
n
d
t
o
p
ic
d
iff
u
sio
n
i
n
h
e
tero
g
e
n
e
o
u
s
i
n
fo
rm
a
ti
o
n
n
e
two
rk
s,
”
S
c
ien
ti
c
Rep
o
rts
,
v
o
l
.
8
,
n
o
.
1
,
2
0
1
8
.
[
1
5
]
S
.
M
o
l
a
e
i
,
e
t
a
l
.
,
“
I
n
f
o
r
m
a
t
i
o
n
S
p
r
e
a
d
a
n
d
T
o
p
i
c
D
i
f
f
u
s
i
o
n
i
n
H
e
t
e
r
o
g
e
n
e
o
u
s
I
n
f
o
r
m
a
t
i
o
n
N
e
t
w
o
r
k
s
,”
S
c
i
.
R
e
p
.,
v
o
l
.
8
,
2
0
1
8
.
[
1
6
]
H.
G
u
i,
Y.
S
u
n
,
J.
Ha
n
,
G
.
Bro
v
a
,
“
M
o
d
e
li
n
g
T
o
p
ic
Diff
u
sio
n
in
M
u
lt
i
-
Re
lati
o
n
a
l
Bi
b
li
o
g
ra
p
h
ic
In
fo
rm
a
ti
o
n
Ne
two
rk
s
,”
In
Pr
o
c
e
e
d
in
g
s
o
f
t
h
e
2
3
rd
ACM
I
n
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
Co
n
fer
e
n
c
e
o
n
I
n
f
o
rm
a
ti
o
n
a
n
d
Kn
o
wl
e
d
g
e
M
a
n
a
g
e
me
n
t
-
CIKM
‘1
4
,
p
p
.
6
4
9
-
6
5
8
,
2
0
1
4
.
[
1
7
]
T.
M
i
k
o
l
o
v
,
K.
C
h
e
n
,
G
.
Co
rra
d
o
,
J.
De
a
n
,
“
Ecie
n
t
e
stim
a
ti
o
n
o
f
wo
r
d
re
p
re
se
n
tatio
n
s
i
n
v
e
c
to
r
sp
a
c
e
,
”
a
rXiv
p
re
p
rin
t
a
rXiv
:1
3
0
1
.
3
7
8
1
,
2
0
1
3
.
[
1
8
]
W.
Ch
e
n
g
,
C.
G
re
a
v
e
s,
M
.
War
re
n
,
“
F
ro
m
n
-
g
ra
m
to
sk
i
p
g
ra
m
to
c
o
n
c
g
ra
m
,
”
In
ter
n
a
ti
o
n
a
l
jo
u
rn
a
l
o
f
c
o
rp
u
s
li
n
g
u
isti
c
s
,
v
o
l
.
11
,
n
o
.
4
,
2
0
0
6
.
[
1
9
]
B.
P
e
ro
z
z
i,
R.
Al
-
Rfo
u
,
S
.
S
k
ien
a
,
“
De
e
p
wa
lk
:
On
li
n
e
lea
rn
in
g
o
f
so
c
ial
re
p
re
se
n
tatio
n
s,
”
in
:
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
2
0
t
h
ACM
S
IGKD
D
I
n
ter
n
a
ti
o
n
a
l
C
o
n
fer
e
n
c
e
o
n
Kn
o
wled
g
e
D
isc
o
v
e
ry
a
n
d
D
a
t
a
M
in
i
n
g
,
ACM
,
p
p
.
7
0
,
2
0
1
4
.
[
2
0
]
S
.
Ho
c
h
re
it
e
r,
J.
S
c
h
m
i
d
h
u
b
e
r,
“
L
o
n
g
sh
o
rt
-
term
m
e
m
o
ry
,
”
Ne
u
ra
l
c
o
mp
u
t
a
ti
o
n
,
v
o
l.
9
,
n
o
.
8
,
1
9
9
7
.
[
2
1
]
Q.
Ca
o
,
H.
S
h
e
n
,
K.
Ce
n
,
W.
Ou
y
a
n
g
,
X
.
Ch
e
n
g
,
“
De
e
p
h
a
wk
e
s:
Brid
g
in
g
t
h
e
g
a
p
b
e
twe
e
n
p
re
d
icti
o
n
a
n
d
u
n
d
e
rsta
n
d
i
n
g
o
f
i
n
fo
rm
a
ti
o
n
c
a
sc
a
d
e
s,
”
in
:
Pro
c
e
e
d
in
g
s
o
f
th
e
2
0
1
7
AC
M
o
n
C
o
n
fer
e
n
c
e
o
n
I
n
fo
rm
a
t
io
n
a
n
d
Kn
o
wled
g
e
M
a
n
a
g
e
me
n
t,
ACM
,
2
0
1
7
.
[
2
2
]
Y.
LeCu
n
,
Y.
Be
n
g
io
,
e
t
a
l.
,
“
Co
n
v
o
lu
ti
o
n
a
l
n
e
two
rk
s
f
o
r
ima
g
e
s,
sp
e
e
c
h
,
a
n
d
ti
m
e
se
ries
,
”
Th
e
h
a
n
d
b
o
o
k
o
f
b
ra
i
n
th
e
o
ry
a
n
d
n
e
u
ra
l
n
e
two
r
k
s
,
v
o
l.
3
3
6
1
,
n
o
.
10
,
1
9
9
5
.
[
2
3
]
F
.
J.
Ord
o
n
e
z
,
D.
Ro
g
g
e
n
,
“
De
e
p
c
o
n
v
o
lu
ti
o
n
a
l
a
n
d
lstm
re
c
u
rre
n
t
n
e
u
ra
l
n
e
two
r
k
s
fo
r
m
u
l
ti
m
o
d
a
l
w
e
a
ra
b
le
a
c
ti
v
it
y
re
c
o
g
n
it
i
o
n
,
”
S
e
n
s
o
rs
,
v
o
l.
16
,
n
o
.
1
,
2
0
1
6
.
[
2
4
]
Cit
a
ti
o
n
Ne
two
rk
Da
tas
e
t,
Av
a
il
a
b
le [o
n
li
n
e
],
URL
h
tt
p
:
//
k
o
n
e
c
t.
u
n
i
-
k
o
b
le
n
z
.
d
e
/n
e
tw
o
rk
s/su
b
e
lj
_
c
o
ra
[
2
5
]
P
.
Ku
m
a
ra
n
,
S
.
Ch
it
ra
k
a
la,
“
Co
m
m
u
n
it
y
f
o
rm
a
ti
o
n
b
a
se
d
in
fl
u
e
n
c
e
n
o
d
e
se
lec
ti
o
n
f
o
r
in
f
o
rm
a
ti
o
n
d
if
fu
sio
n
i
n
o
n
li
n
e
so
c
ial
n
e
two
r
k
,”
In
2
0
1
6
I
n
ter
n
a
ti
o
n
a
l
C
o
n
fer
e
n
c
e
o
n
C
o
mp
u
ti
n
g
T
e
c
h
n
o
lo
g
ies
a
n
d
In
telli
g
e
n
t
Da
ta
E
n
g
i
n
e
e
rin
g
(ICCT
IDE,
2
0
1
6
)
,
p
p
.
1
-
6
,
2
0
1
6
.
[
2
6
]
M
.
Lah
ir
i
a
n
d
M
.
Ce
b
ri
n
,
“
T
h
e
G
e
n
e
ti
c
Alg
o
ri
th
m
a
s
a
G
e
n
e
ra
l
Diffu
sio
n
M
o
d
e
l
f
o
r
S
o
c
ial
Ne
two
r
k
s,
”
Pro
c
.
o
f
t
h
e
2
4
t
h
AA
AI
Co
n
fer
e
n
c
e
o
n
Arti
fi
c
i
a
l
In
telli
g
e
n
c
e
,
2
0
1
0
.
[
2
7
]
L.
Li
,
S
.
Li
,
X.
C
h
e
n
,
“
A
n
e
w
g
e
n
e
ti
c
s
-
b
a
se
d
d
iff
u
si
o
n
m
o
d
e
l
fo
r
s
o
c
ial
n
e
two
rk
s,”
I
n
2
0
1
1
In
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
C
o
mp
u
ta
t
io
n
a
l
Asp
e
c
ts o
f
S
o
c
i
a
l
Ne
two
rk
s (CA
S
o
N
,
2
0
1
1
)
,
p
p
.
7
6
-
8
1
,
2
0
1
1
.
[
2
8
]
H.
Zh
u
,
C.
Hu
a
n
g
,
H.
Li
,
“
In
fo
rm
a
ti
o
n
d
iff
u
sio
n
m
o
d
e
l
b
a
se
d
o
n
p
r
iv
a
c
y
se
tt
in
g
in
o
n
li
n
e
s
o
c
ial
n
e
two
r
k
in
g
se
rv
ice
s,”
T
h
e
Co
m
p
u
ter
J
o
u
rn
a
l
,
v
o
l
.
58
,
n
o
.
4,
p
p
.
5
3
6
-
5
4
8
,
2
0
1
4
.
[
2
9
]
L.
Li
u
,
e
t
a
l
.
,
“
M
o
d
e
ll
in
g
o
f
in
f
o
rm
a
ti
o
n
d
iff
u
sio
n
o
n
so
c
ial
n
e
t
wo
rk
s
wit
h
a
p
p
li
c
a
ti
o
n
s
to
WeCh
a
t,
”
Ph
y
sic
a
A
:
S
ta
ti
st
ica
l
M
e
c
h
a
n
ics
a
n
d
i
ts A
p
p
li
c
a
ti
o
n
s,
v
o
l
.
4
9
6
,
p
p
.
3
1
8
-
3
2
9
,
2
0
1
8
.
[
3
0
]
Ay
m
a
n
M
a
d
i,
O
u
ss
a
m
a
Ka
ss
e
m
Zein
,
S
e
ifed
i
n
e
Ka
d
ry
,
"
On
th
e
i
m
p
ro
v
e
m
e
n
t
o
f
c
y
c
lo
m
a
ti
c
c
o
m
p
lex
it
y
m
e
tri
c
,
”
In
ter
n
a
t
io
n
a
l
J
o
u
rn
a
l
o
f
S
o
f
twa
re
En
g
in
e
e
rin
g
a
n
d
I
ts A
p
p
li
c
a
ti
o
n
s
,
v
o
l
.
7
,
no.
2,
p
p
.
67
-
8
2
,
2
0
1
3
.
[
3
1
]
S
e
ifed
in
e
Ka
d
ry
,
Ra
fic
Yo
u
n
è
s.
"
Et
u
d
e
P
ro
b
a
b
il
iste
d
’
u
n
S
y
ste
m
e
M
e
c
a
n
iq
u
e
a
P
a
ra
m
e
tres
In
c
e
rtain
s
p
a
r
u
n
e
Tec
h
n
iq
u
e
Ba
se
e
su
r
la M
e
th
o
d
e
d
e
tran
sfo
rm
a
ti
o
n
,
"
Pro
c
e
e
d
i
n
g
o
f
Ca
n
C
a
m.
Ca
n
a
d
a
,
2
0
0
5
.
Evaluation Warning : The document was created with Spire.PDF for Python.