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.
10
,
No
.
6
,
Dec
em
b
er
2
0
2
0
,
p
p
.
6
6
2
9
~
6
6
4
3
I
SS
N:
2
0
8
8
-
8708
,
DOI
: 1
0
.
1
1
5
9
1
/
i
j
ec
e
.
v
1
0
i
6
.
pp
6
6
2
9
-
6
6
4
3
6629
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ec
e.
ia
esco
r
e.
co
m/in
d
ex
.
p
h
p
/I
JE
C
E
A
s
y
ste
m
a
tic
r
ev
ie
w
of
t
ex
t
cla
ss
ifi
ca
tion
r
esea
rch
b
a
sed o
n
d
eep
l
ea
rning
m
o
dels
in
A
ra
bic
l
a
n
g
ua
g
e
Ahla
m
Wa
hd
a
n
1
,
Se
nd
ey
a
h
H
a
nto
o
bi
2
,
Sa
id A.
Sa
llo
u
m
3
,
K
ha
led Sha
a
la
n
4
1
,
2,
3,
4
F
a
c
u
lt
y
o
f
En
g
in
e
e
rin
g
a
n
d
I
T
,
T
h
e
Brit
ish
Un
iv
e
rsity
in
Du
b
a
i,
Un
it
e
d
A
ra
b
Em
irate
s
3
Re
se
a
rc
h
In
stit
u
te o
f
S
c
ien
c
e
s a
n
d
En
g
i
n
e
e
rin
g
,
Un
iv
e
rsity
o
f
S
h
a
rjah
,
Un
i
ted
A
ra
b
Em
irate
s
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
A
p
r
10
,
2
0
20
R
ev
i
s
ed
J
u
n
6
,
2020
A
cc
ep
ted
J
u
n
1
7
,
2
0
2
0
C
las
sify
in
g
o
r
c
a
teg
o
rizin
g
te
x
ts
is
th
e
p
ro
c
e
ss
b
y
w
h
ich
d
o
c
u
m
e
n
ts
a
r
e
c
las
si
f
ied
in
to
g
ro
u
p
s
b
y
su
b
jec
t,
ti
tl
e
,
a
u
t
h
o
r
,
e
tc.
T
h
is
p
a
p
e
r
u
n
d
e
rtak
e
s
a
s
y
st
e
m
a
ti
c
re
v
ie
w
o
f
th
e
late
st
re
se
a
rc
h
in
th
e
f
ield
o
f
th
e
c
l
a
ss
i
f
ica
ti
o
n
o
f
A
ra
b
ic
te
x
ts.
S
e
v
e
ra
l
m
a
c
h
in
e
lea
rn
in
g
tec
h
n
i
q
u
e
s
c
a
n
b
e
u
se
d
f
o
r
tex
t
c
las
si
f
ica
ti
o
n
,
b
u
t
w
e
h
a
v
e
f
o
c
u
se
d
o
n
ly
o
n
th
e
re
c
e
n
t
tren
d
o
f
n
e
u
ra
l
n
e
tw
o
rk
a
lg
o
rit
h
m
s.
In
th
is
p
a
p
e
r,
t
h
e
c
o
n
c
e
p
t
o
f
c
las
sify
in
g
te
x
ts
a
n
d
c
las
sif
i
c
a
ti
o
n
p
ro
c
e
ss
e
s
a
re
r
e
v
ie
we
d
.
De
e
p
lea
rn
in
g
tec
h
n
iq
u
e
s
in
c
las
sif
ica
ti
o
n
a
n
d
it
s
ty
p
e
a
re
d
isc
u
ss
e
d
in
th
is
p
a
p
e
r
a
s
w
e
ll
.
N
e
u
ra
l
n
e
tw
o
rk
s
o
f
v
a
rio
u
s
ty
p
e
s,
n
a
m
e
l
y
,
RNN
,
CNN
,
F
F
NN
,
a
n
d
L
S
T
M
,
a
re
id
e
n
ti
f
ied
a
s
t
h
e
su
b
jec
t
o
f
stu
d
y
.
T
h
ro
u
g
h
sy
ste
m
a
ti
c
stu
d
y
,
1
2
re
se
a
rc
h
p
a
p
e
rs
re
late
d
to
t
h
e
f
ield
o
f
th
e
c
las
sif
ica
ti
o
n
o
f
A
ra
b
ic
tex
ts
u
sin
g
n
e
u
ra
l
n
e
tw
o
rk
s
a
re
o
b
tai
n
e
d
:
f
o
r
e
a
c
h
p
a
p
e
r
th
e
m
e
th
o
d
o
lo
g
y
f
o
r
e
a
c
h
ty
p
e
o
f
n
e
u
ra
l
n
e
tw
o
rk
a
n
d
th
e
a
c
c
u
ra
c
y
ra
ti
o
n
f
o
r
e
a
c
h
ty
p
e
is
d
e
ter
m
in
e
d
.
T
h
e
e
v
a
lu
a
ti
o
n
c
rit
e
ria
u
se
d
in
th
e
a
lg
o
rit
h
m
s
o
f
d
iffere
n
t
n
e
u
ra
l
n
e
tw
o
rk
t
y
p
e
s
a
n
d
h
o
w
th
e
y
p
la
y
a
larg
e
ro
le
in
th
e
h
ig
h
ly
a
c
c
u
ra
te
c
l
a
ss
if
ic
a
ti
o
n
o
f
A
r
a
b
ic
tex
ts
a
re
d
isc
u
ss
e
d
.
Ou
r
re
su
lt
s
p
ro
v
id
e
so
m
e
f
in
d
in
g
s
re
g
a
rd
in
g
h
o
w
d
e
e
p
lea
rn
in
g
m
o
d
e
ls
can
b
e
u
se
d
t
o
im
p
ro
v
e
tex
t
c
las
sif
i
c
a
t
io
n
re
se
a
rc
h
in
A
ra
b
ic l
a
n
g
u
a
g
e
.
K
ey
w
o
r
d
s
:
L
ea
r
n
i
n
g
m
a
n
a
g
e
m
en
t
s
y
s
te
m
Neu
r
al
n
et
w
o
r
k
alg
o
r
it
h
m
s
S
y
s
te
m
a
tic
r
ev
ie
w
T
ex
t
class
if
icatio
n
V
ir
tu
al
lear
n
in
g
en
v
ir
o
n
m
e
n
t
Co
p
y
rig
h
t
©
2
0
2
0
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
:
Said
A
.
Sal
lo
u
m
,
R
esear
ch
I
n
s
ti
tu
te
o
f
Scie
n
ce
s
an
d
E
n
g
in
ee
r
i
n
g
,
Un
i
v
er
s
it
y
o
f
S
h
ar
j
ah
,
A
ca
d
e
m
ic
cit
y
R
o
ad
,
P
.
O.
B
o
x
2
7
2
7
2
Sh
ar
j
ah
,
Un
ited
A
r
ab
E
m
ir
ates
.
T
el:
+9
7
1
6
5
5
8
5
0
0
0
,
Fax
: +
9
7
1
6
5
5
8
5
0
9
9
E
m
ail: s
s
allo
u
m
@
s
h
ar
j
ah
.
ac
.
ae
1.
I
NT
RO
D
UCT
I
O
N
T
h
e
class
if
icat
io
n
o
f
te
x
ts
i
s
a
m
eth
o
d
o
f
s
ea
r
ch
in
g
f
o
r
d
ata
an
d
ex
p
lo
r
in
g
t
h
e
d
ata
a
m
o
n
g
lar
g
e
d
ata
an
d
clas
s
i
f
y
i
n
g
t
h
e
m
i
n
to
g
r
o
u
p
s
f
o
r
ea
s
y
r
e
f
er
en
ce
[1
-
5]
.
I
n
ter
n
et
p
ag
es,
b
o
o
k
s
[
6
]
,
m
a
g
az
in
e
s
,
a
n
d
s
o
cia
l
m
ed
ia
[7
-
2
0
]
,
h
a
v
e
b
ec
o
m
e
a
r
ich
s
o
u
r
ce
o
f
i
n
f
o
r
m
atio
n
t
h
a
t
n
ee
d
s
to
b
e
ca
teg
o
r
ized
an
d
o
r
g
an
ized
f
o
r
ea
s
y
r
ef
er
en
ce
[
2
1
-
25]
.
T
h
er
e
h
as
b
ee
n
a
lo
t
o
f
r
esear
ch
in
t
h
i
s
f
ield
,
b
u
t
m
o
s
t
o
f
it
i
n
v
o
lv
e
s
th
e
clas
s
i
f
icatio
n
o
f
E
n
g
l
is
h
an
d
Sp
an
is
h
tex
ts
an
d
tex
ts
i
n
o
th
er
lan
g
u
a
g
e
s
.
T
h
er
e
is
a
lack
o
f
r
esear
ch
in
A
r
ab
ic
tex
t
class
i
f
icatio
n
,
an
d
th
e
tech
n
iq
u
es
an
d
al
g
o
r
ith
m
s
u
s
ed
o
n
E
n
g
l
is
h
tex
t
s
d
o
n
o
t
f
it
th
e
A
r
a
b
ic
lan
g
u
ag
e
te
x
ts
.
T
h
is
is
b
ec
au
s
e
th
e
A
r
ab
ic
lan
g
u
ag
e
h
a
s
ce
r
tai
n
c
h
ar
ac
ter
is
tics
w
h
e
r
e
th
e
s
tr
u
ct
u
r
e
o
f
t
h
e
w
o
r
d
i
s
co
n
ce
r
n
e
d
[
2
6
]
.
T
h
e
m
ai
n
ai
m
o
f
th
is
s
tu
d
y
w
as
to
m
a
k
e
r
ev
ie
w
i
n
g
A
r
ab
ic
tex
t
clas
s
i
f
icati
o
n
b
ased
o
n
n
eu
r
a
l
n
et
w
o
r
k
s
.
So
,
th
is
r
esear
c
h
s
u
m
m
ar
ized
th
e
l
is
t
o
f
n
eu
r
al
n
et
w
o
r
k
tech
n
iq
u
es
t
h
at
w
e
r
e
u
s
ed
to
class
i
f
y
A
r
ab
ic
te
x
ts
a
n
d
d
eter
m
i
n
ed
w
h
ic
h
o
n
es
w
er
e
m
o
r
e
ef
f
ic
i
en
t
a
n
d
ac
cu
r
ate.
Fi
n
d
in
g
t
h
e
g
ap
in
t
h
e
c
u
r
r
en
t
liter
atu
r
e
w
a
s
a
n
o
th
er
i
m
p
o
r
t
an
t
ai
m
.
I
n
th
i
s
r
esear
c
h
,
w
e
f
o
c
u
s
ed
m
o
r
e
o
n
th
e
u
s
e
o
f
n
eu
r
al
n
et
w
o
r
k
s
in
th
e
cla
s
s
i
f
icatio
n
o
f
A
r
ab
ic
t
ex
ts
.
W
e
d
is
c
u
s
s
ed
t
h
e
co
n
c
ep
t
o
f
t
h
e
clas
s
i
f
i
ca
tio
n
o
f
t
ex
ts
f
ir
s
t
alo
n
g
s
id
e
th
e
t
y
p
e
s
o
f
cla
s
s
i
f
icatio
n
t
h
e
n
w
e
p
r
esen
ted
t
h
e
to
p
ic
o
f
n
eu
r
al
n
et
w
o
r
k
s
,
t
h
eir
t
y
p
es,
a
n
d
th
e
cla
s
s
i
f
icatio
n
p
r
o
ce
s
s
es
f
o
r
th
e
d
if
f
er
en
t
t
y
p
es.
T
h
r
o
u
g
h
s
y
s
te
m
atic
r
e
v
ie
w
,
th
e
r
esear
ch
q
u
e
s
tio
n
s
w
er
e
in
i
tiall
y
id
e
n
ti
f
ied
,
th
en
t
h
e
r
esear
ch
s
tr
ate
g
y
w
a
s
d
ev
elo
p
ed
to
o
b
tain
th
e
to
p
ics
m
o
r
e
ac
c
u
r
atel
y
.
S
u
b
s
eq
u
e
n
tl
y
,
q
u
alit
y
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.
10
,
No
.
6
,
Dec
em
b
er
2
0
2
0
:
6
6
2
9
-
6
6
4
3
6630
ass
es
s
m
en
t
cr
iter
ia
w
er
e
d
ev
elo
p
ed
to
ass
ess
th
e
q
u
ali
t
y
o
f
th
e
r
ese
ar
c
h
p
ap
er
s
i
n
o
r
d
er
to
ex
tr
ac
t
th
e
i
m
p
o
r
ta
n
t o
n
es.
I
n
t
h
e
d
is
c
u
s
s
io
n
s
ec
t
io
n
,
all
r
esear
ch
q
u
esti
o
n
s
w
er
e
an
s
w
er
ed
in
d
etai
l.
I
n
o
r
d
er
to
ac
h
iev
e
t
h
e
m
ai
n
a
i
m
o
f
t
h
is
s
t
u
d
y
,
w
e
d
ef
i
n
ed
o
u
r
r
esear
ch
q
u
es
tio
n
s
as f
o
llo
w
s
:
-
R
Q1
:
W
h
ich
co
r
p
o
r
a
w
er
e
m
ain
l
y
u
s
ed
f
o
r
A
r
ab
ic
tex
t
cla
s
s
i
f
icatio
n
?
W
er
e
th
e
y
o
p
en
s
o
u
r
ce
o
r
cr
ea
ted
b
y
t
h
e
au
th
o
r
?
-
R
Q2
:
W
h
ich
co
u
n
tr
ie
s
f
o
c
u
s
e
d
o
n
p
u
b
lis
h
i
n
g
r
esear
ch
o
n
AT
C
?
-
R
Q3
:
W
h
ic
h
d
atab
ases
h
ad
m
o
r
e
p
u
b
licatio
n
s
?
-
R
Q4
:
Du
r
i
n
g
w
h
ic
h
y
ea
r
s
w
as
th
er
e
m
o
r
e
f
o
cu
s
o
n
A
T
C
?
-
R
Q5
:
W
h
at
w
er
e
th
e
t
y
p
es
o
f
n
eu
r
al
n
et
w
o
r
k
alg
o
r
it
h
m
s
u
s
ed
to
class
if
y
A
r
ab
ic
tex
ts
?
An
d
w
h
ic
h
o
n
e
w
a
s
m
o
s
t u
s
ed
?
-
R
Q6
:
W
h
at
w
er
e
th
e
e
f
f
icie
n
c
y
m
ea
s
u
r
e
s
u
s
ed
?
-
R
Q7
:
Did
th
e
au
t
h
o
r
co
m
p
ar
e
NN
w
it
h
o
th
er
tech
n
iq
u
e
s
?
An
d
w
h
ic
h
o
n
es
w
er
e
th
e
m
o
s
t
f
r
eq
u
en
t
l
y
co
m
p
ar
ed
w
it
h
NN?
-
R
Q8
:
W
h
ic
h
t
y
p
e
o
f
n
eu
r
al
n
e
t
w
o
r
k
al
g
o
r
ith
m
p
r
o
v
ed
m
o
s
t
ef
f
icien
t?
2.
L
I
T
E
R
AT
U
RE
R
E
VI
E
W
2
.
1
.
Def
ini
t
io
n o
f
T
C
T
h
e
tex
t
cla
s
s
i
f
icat
io
n
p
r
o
ce
s
s
is
t
h
e
p
r
o
ce
s
s
o
f
a
u
to
m
ati
n
g
a
s
et
o
f
d
o
cu
m
en
ts
in
to
s
p
ec
if
ic
g
r
o
u
p
s
b
ased
o
n
th
e
co
n
ten
t
o
f
t
h
e
tex
t
its
el
f
t
h
r
o
u
g
h
th
e
u
s
e
o
f
ce
r
tain
tech
n
o
lo
g
ie
s
an
d
alg
o
r
ith
m
s
[
2
4
,
2
7
-
3
0
]
.
E
lh
as
s
an
a
n
d
Ah
m
ed
[
3
0
]
al
s
o
d
ef
in
ed
th
e
cla
s
s
i
f
icatio
n
o
f
tex
t
s
as
a
m
et
h
o
d
o
f
s
ea
r
ch
in
g
f
o
r
d
ata
an
d
ex
p
lo
r
in
g
t
h
e
m
a
m
o
n
g
lar
g
e
d
ata
an
d
cla
s
s
i
f
y
in
g
th
e
m
i
n
t
o
g
r
o
u
p
s
f
o
r
ea
s
y
r
ef
er
e
n
ce
.
A
cc
o
r
d
in
g
to
[
3
1
]
,
th
e
clas
s
i
f
icatio
n
o
f
te
x
ts
w
a
s
d
ef
i
n
ed
as
t
h
e
p
r
o
ce
s
s
o
f
o
r
g
an
izi
n
g
d
o
c
u
m
en
t
s
ac
co
r
d
in
g
to
a
k
n
o
w
n
a
n
d
p
r
e
-
ex
is
tin
g
s
tr
u
ct
u
r
e
o
f
s
p
ec
if
ic
ca
te
g
o
r
ies
th
a
t
s
u
ited
th
i
s
t
y
p
e
o
f
te
x
t.
Ot
h
er
s
m
e
n
tio
n
ed
th
at
t
h
er
e
w
as
a
s
li
g
h
t
d
i
f
f
er
e
n
ce
b
et
w
ee
n
te
x
t
ca
te
g
o
r
izatio
n
an
d
cla
s
s
i
f
ic
atio
n
.
W
h
er
ea
s
th
e
ca
te
g
o
r
izatio
n
o
f
tex
ts
e
n
taile
d
so
r
tin
g
th
e
m
ac
co
r
d
in
g
to
th
e
ir
co
n
ten
t,
t
h
eir
clas
s
i
f
icatio
n
p
lace
d
th
e
m
i
n
ce
r
tai
n
g
r
o
u
p
s
th
at
s
u
ited
t
h
ei
r
co
n
ten
t
ac
co
r
d
in
g
to
a
u
t
h
o
r
,
s
u
b
j
ec
t,
titl
e,
lan
g
u
ag
e,
a
n
d
o
th
er
clas
s
i
f
icatio
n
s
[
3
2
]
.
Sp
ec
ialized
r
esear
ch
o
n
th
e
cla
s
s
i
f
icatio
n
o
f
tex
ts
h
as
in
cr
e
ased
s
i
g
n
i
f
ican
tl
y
d
u
e
to
th
e
e
n
o
r
m
o
u
s
d
ata
a
v
ailab
le
f
r
o
m
m
a
n
y
s
o
u
r
ce
s
,
in
cl
u
d
in
g
I
n
ter
n
e
t
p
ag
es,
e
-
m
ail
m
es
s
ag
e
s
,
n
e
w
s
p
ag
e
s
,
tex
ts
cir
cu
lated
th
r
o
u
g
h
s
o
cial
m
ed
ia,
r
ep
o
r
ts
,
an
d
j
o
u
r
n
al
ar
ticles.
T
h
er
ef
o
r
e,
th
is
r
esear
ch
f
o
cu
s
ed
atten
tio
n
i
n
o
r
d
er
to
m
a
k
e
th
e
b
est
p
o
s
s
ib
le
u
s
e
o
f
all
th
e
s
e
d
ata
an
d
class
i
f
y
t
h
e
m
[
3
3
]
.
2
.
2
.
Ara
bic
t
ex
t
cla
s
s
if
ica
t
io
n
A
r
ab
ic
is
o
n
e
o
f
t
h
e
s
i
x
p
r
im
ar
y
lan
g
u
ag
e
s
w
o
r
ld
w
id
e.
I
t
s
u
s
e
i
s
s
p
r
ea
d
ac
r
o
s
s
m
a
n
y
c
o
u
n
tr
ies
i
n
th
e
w
o
r
ld
,
an
d
it
is
t
h
e
b
asi
s
o
f
th
e
A
r
ab
w
o
r
ld
,
b
u
t
s
o
m
e
I
s
la
m
ic
co
u
n
tr
ies
a
ls
o
u
s
e
t
h
e
A
r
ab
ic
la
n
g
u
a
g
e
a
s
th
eir
p
r
i
m
ar
y
la
n
g
u
a
g
e
b
ec
au
s
e
it
i
s
t
h
e
la
n
g
u
a
g
e
o
f
t
h
e
Q
u
r
'
a
n
.
T
h
e
A
r
ab
ic
la
n
g
u
a
g
e
c
o
n
s
is
ts
o
f
2
8
letter
s
,
in
cl
u
d
in
g
h
a
m
za
s
a
n
d
el
m
o
d
o
d
.
E
ac
h
letter
h
as
ce
r
tain
s
tr
u
ctu
r
es
t
h
at
d
ep
en
d
o
n
th
e
p
o
s
i
tio
n
o
f
th
e
le
tter
in
th
e
w
h
o
le,
t
h
at
is
,
w
h
et
h
er
i
t
is
a
t
t
h
e
b
e
g
in
n
i
n
g
o
f
t
h
e
wo
r
d
,
in
th
e
m
id
d
le
o
f
th
e
w
o
r
d
,
o
r
at
th
e
en
d
o
f
th
e
w
o
r
d
[
3
4
]
.
T
h
er
e
ar
e
n
o
u
p
p
er
-
an
d
lo
w
er
-
ca
s
e
let
ter
s
i
n
A
r
ab
ic
u
n
li
k
e
E
n
g
li
s
h
o
r
o
t
h
er
lan
g
u
ag
e
s
[
3
1
]
.
T
h
e
m
o
r
p
h
o
lo
g
y
o
f
th
e
A
r
ab
i
c
lan
g
u
a
g
e
is
n
o
t
ea
s
y
;
it
m
a
y
b
e
co
m
p
licated
as
i
t
h
a
s
r
ad
ical
w
o
r
d
s
,
p
r
ef
i
x
es,
an
d
s
u
f
f
ix
e
s
;
m
o
r
eo
v
er
,
u
n
li
k
e
th
o
s
e
in
o
th
er
lan
g
u
ag
e
s
,
its
w
o
r
d
s
d
o
n
o
t
co
n
s
is
t
o
f
s
eq
u
en
t
ial
f
o
r
m
s
as
its
w
o
r
d
s
tr
u
c
tu
r
es
ar
e
co
m
p
let
el
y
d
i
f
f
er
e
n
t
a
n
d
d
ep
en
d
o
n
p
o
s
itio
n
in
th
e
s
e
n
ten
ce
an
d
m
ea
n
i
n
g
[
3
1
]
.
As
m
e
n
tio
n
ed
ab
o
v
e,
th
e
cla
s
s
if
ica
tio
n
o
f
tex
t
s
h
as
b
ec
o
m
e
w
id
esp
r
ea
d
,
an
d
m
o
s
t
o
f
t
h
e
r
esear
ch
in
c
lu
d
es
th
e
class
if
icatio
n
o
f
o
th
er
la
n
g
u
a
g
es,
s
u
c
h
as
E
n
g
lis
h
.
L
ittl
e
o
f
th
e
r
esear
ch
o
n
th
e
clas
s
i
f
icatio
n
o
f
tex
t
s
h
as
f
o
cu
s
ed
o
n
A
r
ab
ic.
Ma
n
y
tec
h
n
iq
u
es
an
d
alg
o
r
it
h
m
s
h
a
v
e
b
ee
n
cr
ea
ted
to
class
if
y
E
n
g
l
is
h
te
x
ts
,
an
d
th
e
y
g
iv
e
e
x
ce
lle
n
t
r
es
u
lt
s
a
n
d
h
i
g
h
ac
c
u
r
ac
y
d
u
e
to
t
h
e
n
at
u
r
e
o
f
t
h
e
la
n
g
u
a
g
e,
letter
s
,
a
n
d
w
o
r
d
s
.
B
u
t
w
h
en
it
co
m
e
s
to
t
h
e
A
r
ab
ic
la
n
g
u
a
g
e,
th
e
ap
p
licati
o
n
o
f
th
e
s
e
al
g
o
r
i
th
m
s
d
o
es
n
o
t
g
iv
e
s
i
m
ilar
ac
c
u
r
ac
y
an
d
v
alid
it
y
in
cla
s
s
i
f
icat
io
n
.
T
h
er
ef
o
r
e,
ap
p
r
o
p
r
iate
alg
o
r
ith
m
s
h
av
e
b
ee
n
cr
ea
ted
to
s
y
n
t
h
esize
wo
r
d
s
in
t
h
e
A
r
ab
ic
lan
g
u
a
g
e,
b
u
t t
h
e
y
h
av
e
n
o
t
y
e
t d
em
o
n
s
tr
ated
f
u
l
l
y
r
eliab
le
m
er
it f
o
r
t
h
e
class
if
ica
tio
n
p
r
o
ce
s
s
[
3
4
]
.
2
.
3
.
Cla
s
s
if
ica
t
io
n t
ec
hn
iqu
e
T
h
er
e
ar
e
t
w
o
b
asic
m
et
h
o
d
s
o
f
cla
s
s
i
f
y
in
g
tex
t
s
.
O
n
e
i
s
t
h
e
g
r
a
m
m
ar
o
r
li
n
g
u
i
s
tic
b
a
s
e
d
m
et
h
o
d
,
w
h
ic
h
d
ep
en
d
s
o
n
p
r
ep
ar
in
g
ce
r
tain
r
u
les
an
d
o
p
er
atin
g
th
e
m
i
n
s
o
m
e
e
x
p
er
t
s
y
s
te
m
s
i
n
o
r
d
er
to
clas
s
i
f
y
th
e
te
x
t
s
.
T
h
e
s
ec
o
n
d
m
et
h
o
d
is
t
h
e
p
r
o
ce
s
s
o
f
f
ee
d
in
g
t
h
e
t
ex
t
e
x
tr
ap
o
latio
n
p
r
o
ce
s
s
w
it
h
a
s
et
o
f
d
o
cu
m
e
n
ts
ca
lled
tr
ain
in
g
d
o
cu
m
en
ts
t
h
at
h
a
v
e
p
r
ev
io
u
s
l
y
b
ee
n
c
lass
i
f
ied
ac
co
r
d
in
g
to
s
p
ec
i
f
ic
ca
teg
o
r
ies
[
3
0
]
.
B
elo
w
ar
e
th
e
ex
p
la
n
atio
n
s
o
f
th
e
t
w
o
m
et
h
o
d
s
.
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:
2
0
8
8
-
8708
A
s
ystema
tic
r
ev
ie
w
o
f te
xt
cla
s
s
ifica
tio
n
r
esea
r
ch
b
a
s
ed
o
n
.
.
.
(
A
h
la
m
W
a
h
d
a
n
)
6631
2
.
3
.
1.
M
a
nu
a
l a
nd
s
t
a
t
is
t
ica
l
t
ec
hn
iqu
e
s
Ma
n
u
al
tex
t
c
lass
if
ica
tio
n
(
T
C
)
is
d
o
n
e
b
y
w
r
i
tin
g
s
p
ec
i
f
ic
q
u
er
ie
s
m
a
n
u
all
y
f
o
r
ea
ch
ca
teg
o
r
y
,
f
o
r
ex
a
m
p
le
s
p
o
r
ts
,
n
u
tr
itio
n
,
clo
th
in
g
,
an
d
h
ea
lt
h
.
T
h
en
th
e
tex
t
en
ter
ed
w
it
h
i
n
a
s
p
ec
if
ic
s
ea
r
ch
en
g
in
e
is
ca
teg
o
r
ized
b
ased
o
n
th
e
p
r
ed
ef
i
n
ed
q
u
er
ies.
Ho
w
ev
er
,
t
h
i
s
m
et
h
o
d
w
o
r
k
s
o
n
s
m
all
te
x
t
s
,
n
o
t o
n
lar
g
e
o
n
es
o
r
o
n
m
a
n
y
d
o
cu
m
e
n
t
s
.
T
h
e
ac
cu
r
ac
y
o
f
th
i
s
ap
p
r
o
ac
h
to
class
i
f
icatio
n
d
ep
en
d
s
o
n
t
h
e
v
a
l
id
it
y
o
f
q
u
ir
es
a
n
d
th
e
s
k
il
l
b
eh
i
n
d
t
h
e
q
u
er
y
d
esig
n
.
E
x
p
er
t
s
y
s
te
m
s
h
a
v
e
b
ee
n
cr
ea
ted
th
at
p
r
ep
ar
e
r
u
les
an
d
in
q
u
ir
ies.
An
ex
a
m
p
le
is
C
o
n
s
tr
u
e
-
T
I
S,
th
e
o
n
e
cr
ea
ted
f
o
r
R
eu
ter
s
Ne
w
s
,
w
h
ic
h
w
a
s
ab
le
to
ap
p
ly
6
7
4
ca
teg
o
r
ies
an
d
to
g
et
to
k
n
o
w
m
o
r
e
t
h
a
n
1
7
0
,
0
0
0
co
m
p
an
ie
s
[
3
2
]
.
T
h
is
cla
s
s
if
ica
tio
n
w
as
n
o
t
b
ased
o
n
in
d
iv
id
u
al
w
o
r
d
s
b
u
t
o
n
co
n
ce
p
ts
g
r
ad
ed
f
r
o
m
th
e
ac
tu
al
te
x
t
a
s
s
h
o
w
n
i
n
Fi
g
u
r
e
1
.
T
ec
h
n
iq
u
es
w
h
ic
h
r
el
y
o
n
m
at
h
e
m
a
tical
r
u
les
an
d
p
r
in
cip
les
ar
e
also
ca
lled
s
tatis
t
ical
tex
t
cla
s
s
i
f
icat
io
n
.
T
h
ese
tech
n
iq
u
es
ar
e
s
u
itab
le
f
o
r
s
m
all
d
ata,
s
u
ch
as “
Fre
q
u
en
t
is
t
p
r
o
ce
d
u
r
es,
B
ay
esia
n
p
r
o
ce
d
u
r
es,
an
d
th
e
B
i
n
ar
y
a
n
d
m
u
lticla
s
s
p
r
o
ce
d
u
r
es”
[
3
4
]
.
Fig
u
r
e
1
.
T
h
e
r
u
le
o
f
th
e
ca
teg
o
r
y
Au
s
tr
alia
n
Do
llar
2
.
3
.
2.
M
a
chine le
a
rning
t
ec
hn
iq
ues
Du
e
to
th
e
lar
g
e
a
m
o
u
n
ts
o
f
in
f
o
r
m
at
io
n
d
u
r
in
g
th
e
p
ast
t
w
o
d
ec
ad
es,
th
er
e
em
er
g
ed
a
n
ee
d
f
o
r
tech
n
ical
m
et
h
o
d
s
o
f
clas
s
i
f
y
i
n
g
h
u
g
e
te
x
ts
as
s
tatis
tical
an
d
m
a
n
u
al
m
et
h
o
d
s
h
ad
n
e
v
er
b
ec
o
m
e
u
s
ele
s
s
i
n
th
is
f
ie
ld
[
3
5
-
48]
.
T
h
er
ef
o
r
e,
m
ac
h
in
e
lear
n
i
n
g
c
lass
if
i
ca
t
io
n
tech
n
iq
u
es
ap
p
ea
r
ed
w
h
i
ch
ai
m
to
clas
s
if
y
u
n
s
tr
u
ct
u
r
ed
tex
t
s
an
d
d
o
cu
m
en
ts
b
ased
o
n
ce
r
tain
al
g
o
r
ith
m
s
d
esi
g
n
ed
f
o
r
th
is
p
u
r
p
o
s
e.
Ma
ch
in
e
lear
n
i
n
g
tech
n
iq
u
es
ca
n
b
e
d
i
v
id
ed
in
to
g
r
o
u
p
s
as
s
h
o
w
n
in
Fi
g
u
r
e
2
[
3
4
]
.
I
n
th
e
Su
p
er
v
is
ed
L
ea
r
n
i
n
g
tec
h
n
iq
u
e,
d
if
f
er
e
n
t
d
ata
ar
e
r
ec
o
g
n
ized
,
w
h
et
h
er
v
i
s
u
al,
a
u
d
io
,
o
r
tex
t.
T
h
e
d
ata
ar
e
co
m
p
ar
ed
w
it
h
th
e
ex
p
ec
ted
id
ea
l
d
ata
o
r
r
esu
lt
s
t
h
r
o
u
g
h
a
p
r
o
ce
s
s
ca
l
led
b
ac
k
p
r
o
p
ag
atio
n
in
w
h
ic
h
t
h
e
d
ata
ar
e
d
ir
ec
ted
f
r
o
m
t
h
e
o
u
tp
u
t
la
y
er
s
to
th
e
in
p
u
t
la
y
er
s
an
d
er
r
o
r
s
ar
e
co
r
r
ec
ted
an
d
m
i
n
i
m
iz
ed
to
ac
h
iev
e
b
etter
ac
cu
r
ac
y
.
I
n
U
n
s
u
p
er
v
is
ed
L
ea
r
n
i
n
g
,
t
h
e
lear
n
i
n
g
p
r
o
ce
s
s
tak
e
s
p
lace
d
u
r
i
n
g
t
h
e
p
r
o
ce
s
s
in
g
.
T
h
er
e
ar
e
n
o
p
r
ev
io
u
s
d
ata
f
o
r
co
m
p
ar
i
s
o
n
as
th
e
n
et
w
o
r
k
an
a
l
y
ze
s
t
h
e
d
ata
an
d
p
r
o
ce
s
s
es
t
h
e
m
;
t
h
e
n
,
b
u
ild
s
a
f
u
n
ctio
n
to
d
eter
m
i
n
e
th
e
er
r
o
r
r
ate
an
d
r
ed
u
ce
it
to
o
b
tain
h
i
g
h
ac
c
u
r
ac
y
.
T
h
e
Se
m
i
-
Su
p
er
v
is
ed
lea
r
n
in
g
tec
h
n
iq
u
e
i
s
a
co
m
b
i
n
at
io
n
o
f
t
h
e
p
r
ev
io
u
s
t
w
o
tec
h
n
iq
u
e
s
.
R
e
s
ea
r
ch
er
s
s
a
w
t
h
at
t
h
er
e
w
as
litt
le
u
n
n
a
m
ed
d
ata
th
at
co
u
ld
co
n
tr
i
b
u
te
to
i
m
p
r
o
v
in
g
th
e
ac
cu
r
ac
y
o
f
t
h
e
lear
n
i
n
g
p
r
o
ce
s
s
th
at
to
o
k
p
lace
d
u
r
in
g
t
h
e
d
ata
cla
s
s
i
f
icatio
n
s
tag
e.
So
,
a
s
m
al
l
p
er
ce
n
tag
e
o
f
t
h
e
d
ata
w
a
s
u
s
e
d
in
Se
m
i
-
Su
p
er
v
i
s
ed
lear
n
i
n
g
.
Fig
u
r
e
2
.
Ma
ch
in
e
l
ea
r
n
i
n
g
t
e
ch
n
iq
u
es
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.
10
,
No
.
6
,
Dec
em
b
er
2
0
2
0
:
6
6
2
9
-
6
6
4
3
6632
2
.
4
.
Neura
l n
et
w
o
rk
A
l
g
o
r
ith
m
s
o
f
n
e
u
r
al
n
et
w
o
r
k
s
g
iv
e
v
er
y
ac
c
u
r
ate
r
esu
lt
s
in
th
e
ar
ea
o
f
N
L
P
an
d
s
u
p
p
o
r
t
d
ee
p
lear
n
in
g
v
er
y
s
i
g
n
if
ican
tl
y
as
th
e
y
s
o
lv
e
s
o
m
e
er
r
o
r
s
an
d
p
r
o
b
lem
s
r
elate
d
to
th
e
v
ar
ian
c
e
o
f
d
ata
r
esu
ltin
g
f
r
o
m
t
h
e
p
r
o
ce
s
s
o
f
d
ee
p
lear
n
in
g
[
3
6
-
39]
.
Neu
r
al
n
et
w
o
r
k
s
ar
e
u
s
ed
i
n
t
h
e
clas
s
i
f
icatio
n
o
f
te
x
t
s
to
ad
d
r
ess
lin
ea
r
an
d
n
o
n
-
li
n
ea
r
p
r
o
b
lem
s
.
T
h
e
b
ac
k
p
r
o
p
ag
atio
n
m
o
d
el
class
if
ies
tex
t
s
u
s
in
g
n
e
u
r
al
n
et
w
o
r
k
s
b
y
m
ea
n
s
o
f
a
g
r
o
u
p
o
f
n
o
d
es
t
h
at
f
o
r
m
an
d
r
ep
r
esen
t
a
m
ath
e
m
atica
l
m
o
d
el
o
f
b
io
lo
g
ical
n
er
v
e
s
.
No
in
te
r
f
er
en
ce
a
n
d
s
elf
-
lear
n
in
g
a
f
f
ec
t
th
e
s
e
n
eu
r
al
n
et
w
o
r
k
s
a
s
t
h
e
y
ar
e
u
s
ed
t
o
id
en
ti
f
y
s
y
s
te
m
s
an
d
p
atter
n
s
an
d
to
class
if
y
te
x
t
an
d
i
m
a
g
e
p
r
o
ce
s
s
in
g
[
4
0
]
.
Ma
n
y
co
n
s
id
er
n
eu
r
al
n
e
t
w
o
r
k
s
to
b
e
u
n
its
th
at
ar
e
co
n
n
ec
ted
to
ea
ch
o
th
er
an
d
in
ter
t
w
i
n
ed
s
i
m
ilar
l
y
t
o
th
e
d
y
n
as
ticit
y
o
f
th
e
n
e
u
r
o
n
.
I
n
ad
d
itio
n
,
th
r
o
u
g
h
t
h
is
p
r
o
ce
s
s
o
f
en
ta
n
g
le
m
en
t,
th
e
i
n
p
u
ts
ar
e
m
ad
e
a
n
d
t
h
e
o
th
er
n
eu
r
o
n
allo
w
s
t
h
e
e
x
it
o
f
th
e
tr
an
s
m
itted
o
u
tp
u
ts
.
T
h
e
d
esig
n
o
f
A
NN
is
s
i
m
ilar
in
s
tr
u
ctu
r
e
to
th
at
o
f
th
e
b
r
ain
an
d
its
n
e
u
r
al
n
et
wo
r
k
.
A
ll
t
y
p
e
s
o
f
s
ab
o
tag
e
o
f
th
e
n
e
u
r
al
n
et
w
o
r
k
h
av
e
a
b
asic
p
r
i
n
cip
le,
w
h
ich
is
t
h
at
e
v
er
y
n
e
u
r
o
n
i
n
t
h
e
n
et
w
o
r
k
r
ec
eiv
es
i
n
p
u
ts
,
p
r
o
ce
s
s
es
th
e
m
,
a
n
d
s
e
n
d
s
o
u
tp
u
ts
.
E
ac
h
n
e
u
r
o
n
is
lin
k
ed
to
at
lea
s
t o
n
e
n
eu
r
o
n
,
an
d
ea
ch
co
n
n
ec
tio
n
o
r
as
s
o
ciatio
n
h
as a
s
p
ec
if
ic
d
ig
i
tal
w
ei
g
h
t
ca
lled
t
h
e
w
e
ig
h
t
f
ac
to
r
,
w
h
ich
w
o
r
k
s
to
r
ef
lect
t
h
e
i
m
p
o
r
tan
ce
o
f
t
h
e
co
m
m
u
n
ica
tio
n
b
et
w
ee
n
n
eu
r
o
n
s
i
n
th
e
n
et
w
o
r
k
[
4
1
]
.
So
m
e
p
o
p
u
lar
p
r
i
m
ar
y
t
y
p
es o
f
n
eu
r
al
n
et
w
o
r
k
s
f
o
llo
w
.
2
.
4
.
1.
F
ee
dfo
r
w
a
rd
neura
l net
w
o
rk
(
F
NN)
I
t
is
a
n
eu
r
al
n
et
w
o
r
k
w
h
o
s
e
ce
lls
ar
e
lin
ea
r
l
y
co
n
n
ec
ted
to
ea
ch
o
th
er
an
d
d
o
n
o
t
r
e
p
r
esen
t
a
c
y
cle
lik
e
an
y
o
th
er
n
e
u
r
al
n
e
t
w
o
r
k
.
T
h
e
s
in
g
le
la
y
er
p
er
ce
p
tr
o
n
(
SL
P
)
is
th
e
s
i
m
p
les
t
f
o
r
m
o
f
t
h
is
t
y
p
e
o
f
n
eu
r
al
n
et
w
o
r
k
s
in
ce
t
h
e
in
p
u
ts
ar
e
d
ir
ec
tly
r
elate
d
to
th
e
o
u
tp
u
t.
I
n
t
h
e
S
L
P
,
th
e
i
n
p
u
ts
g
o
th
r
o
u
g
h
s
ev
er
al
la
y
er
s
o
f
tr
a
n
s
f
o
r
m
atio
n
,
s
o
t
h
e
y
ar
e
co
n
s
id
er
ed
v
er
y
s
u
ita
b
le
f
o
r
ca
teg
o
r
izin
g
tex
t
s
[
4
2
]
.
T
h
e
s
tr
u
ct
u
r
e
o
f
th
e
f
ee
d
f
o
r
w
ar
d
n
e
u
r
al
n
e
t
w
o
r
k
is
s
h
o
w
n
in
F
ig
u
r
e
3
.
Fig
u
r
e
3
.
FNN
2
.
4
.
2.
Co
nv
o
lutio
na
l
neura
l
net
w
o
rk
(
CNN)
C
o
n
v
o
l
u
tio
n
al
n
eu
r
al
n
et
w
o
r
k
s
ar
e
in
s
p
ir
ed
b
y
t
h
e
m
u
lti
-
l
a
y
er
p
er
ce
p
tr
o
n
(
ML
P
)
an
d
ar
e
d
esig
n
ed
to
ex
tr
ac
t
t
h
e
s
p
atial
s
tr
u
ct
u
r
e
in
i
m
a
g
e
d
ata
an
d
t
h
e
p
o
s
itio
n
s
o
f
o
b
j
ec
ts
th
at
ca
n
b
e
u
s
ed
in
t
h
e
i
m
a
g
e
[
4
3
]
.
T
h
e
s
a
m
e
p
r
i
n
cip
le
a
n
d
id
ea
w
er
e
u
s
ed
to
clas
s
i
f
y
te
x
ts
an
d
o
n
e
-
d
i
m
e
n
s
io
n
al
w
o
r
d
s
,
an
d
t
h
is
w
as
d
o
n
e
th
r
o
u
g
h
in
ter
ac
tio
n
w
i
th
n
e
ig
h
b
o
r
in
g
n
eu
r
o
n
s
.
I
ts
n
a
m
e
is
“
c
o
n
v
o
l
u
tio
n
al”
d
u
e
to
f
o
ld
i
n
g
p
r
o
ce
s
s
th
a
t
o
cc
u
r
s
b
et
w
ee
n
n
e
u
r
o
n
s
d
u
r
i
n
g
t
h
e
class
i
f
icatio
n
p
r
o
ce
s
s
.
T
h
e
s
tr
u
ctu
r
e
o
f
t
h
e
co
n
v
o
l
u
tio
n
a
l
n
eu
r
al
n
et
w
o
r
k
is
s
h
o
w
n
in
F
ig
u
r
e
4
.
Fig
u
r
e
4
.
C
NN
2
.
4
.
3.
Rec
urre
nt
neura
l net
wo
rk
(
RNN)
T
h
e
s
tr
u
ctu
r
e
i
n
th
is
t
y
p
e
o
f
n
et
w
o
r
k
i
s
d
o
n
e
co
n
ti
n
u
o
u
s
l
y
an
d
s
eq
u
en
t
iall
y
s
i
n
ce
t
h
e
o
u
tp
u
ts
o
f
th
e
p
r
ev
io
u
s
s
ta
g
e
ar
e
th
e
i
n
p
u
ts
o
f
th
e
n
e
x
t
s
ta
g
e.
B
y
co
n
t
r
ast,
w
i
th
tr
ad
itio
n
al
n
e
u
r
al
n
et
w
o
r
k
s
,
th
e
in
p
u
ts
an
d
o
u
tp
u
ts
ar
e
s
ep
ar
ate
u
n
i
t
s
f
r
o
m
ea
c
h
o
t
h
er
.
So
m
e
ti
m
es
t
h
er
e
is
a
n
e
ed
to
p
r
ed
ict
th
e
d
esire
d
w
o
r
d
,
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:
2
0
8
8
-
8708
A
s
ystema
tic
r
ev
ie
w
o
f te
xt
cla
s
s
ifica
tio
n
r
esea
r
ch
b
a
s
ed
o
n
.
.
.
(
A
h
la
m
W
a
h
d
a
n
)
6633
an
d
th
er
e
f
o
r
e,
th
e
n
ee
d
d
ev
e
lo
p
s
to
r
ef
er
to
th
e
w
o
r
d
s
p
r
io
r
to
th
e
r
ev
ie
w
,
a
n
d
th
is
is
w
h
a
t
r
ec
u
r
r
en
t
n
eu
r
al
n
et
w
o
r
k
s
d
o
.
A
h
id
d
en
la
y
er
h
a
s
b
ee
n
u
s
ed
to
h
elp
w
it
h
th
i
s
t
y
p
e
o
f
n
et
w
o
r
k
an
d
to
co
m
p
lete
th
e
p
r
o
ce
s
s
[
4
4
]
.
A
n
ex
a
m
p
le
o
f
a
f
r
eq
u
e
n
t n
e
u
r
al
n
et
w
o
r
k
i
s
s
h
o
w
n
i
n
Fi
g
u
r
e
5
.
Fig
u
r
e
5
.
R
NN
2
.
4
.
4.
L
o
ng
s
ho
rt
-
t
er
m
m
e
mo
ry
(
L
ST
M
)
L
ST
M
ar
e
lo
n
g
s
h
o
r
t
-
ter
m
m
e
m
o
r
y
n
et
w
o
r
k
s
as
th
e
y
a
r
e
a
s
p
ec
if
ic
r
ec
u
r
r
en
t
n
eu
r
a
l
n
et
w
o
r
k
s
tr
u
ct
u
r
e
(
R
NN)
d
esi
g
n
ed
f
o
r
u
s
e
i
n
m
o
d
elin
g
lo
n
g
-
r
a
n
g
e
ti
m
e
s
er
ie
s
t
h
at
ar
e
o
b
s
er
v
ed
to
o
p
er
ate
m
o
r
e
p
r
ec
is
el
y
t
h
a
n
R
NN.
T
h
e
n
eu
r
o
n
s
o
f
t
h
is
n
et
w
o
r
k
a
r
e
m
ad
e
u
p
o
f
u
n
its
t
h
at
ar
e
g
ates
; i
n
t
h
is
w
a
y
,
t
h
e
n
et
w
o
r
k
ca
n
co
n
tr
o
l
t
h
e
f
lo
w
o
f
t
h
e
i
n
p
u
ts
t
h
at
lead
to
t
h
e
f
i
n
al
o
u
t
p
u
t.
T
h
u
s
,
o
n
l
y
a
f
e
w
in
p
u
ts
m
a
y
p
ar
ticip
ate
i
n
th
e
o
u
tp
u
t,
s
o
th
e
er
r
o
r
r
ate
is
r
ed
u
ce
d
[
4
5
]
.
L
ST
M
co
n
tai
n
s
u
n
i
ts
ca
lled
m
e
m
o
r
y
b
lo
c
k
s
,
a
n
d
th
e
y
ar
e
lo
ca
ted
in
th
e
h
id
d
en
la
y
er
.
T
h
ese
b
lo
ck
s
h
a
v
e
s
el
f
-
co
n
n
ec
tio
n
s
i
n
m
e
m
o
r
y
ce
lls
t
h
at
r
ec
o
r
d
th
e
ti
m
e
s
tat
u
s
o
f
th
e
n
e
t
w
o
r
k
d
u
r
i
n
g
t
h
e
w
o
r
k
.
T
h
ey
al
s
o
co
n
tai
n
u
n
it
s
ca
lle
d
g
ates
t
h
at
co
n
tr
ib
u
te
to
co
n
tr
o
llin
g
t
h
e
f
lo
w
o
f
in
p
u
t
s
an
d
o
u
tp
u
t
s
.
T
h
eir
in
p
u
t
g
ates
co
n
tr
o
l
th
e
ac
tiv
a
tio
n
p
r
o
ce
s
s
f
o
r
en
ter
in
g
in
f
o
r
m
atio
n
in
to
th
e
m
e
m
o
r
y
.
As
f
o
r
t
h
e
o
u
tp
u
t
p
o
r
t,
it
co
n
t
r
o
ls
th
e
o
u
tp
u
t
a
f
ter
th
e
ac
t
iv
a
tio
n
p
r
o
ce
s
s
,
w
h
ic
h
ta
k
es
p
lac
e
at
th
e
e
n
tr
y
g
ate.
A
p
o
r
tal
ca
lled
t
h
e
f
o
r
g
et
g
at
e,
w
h
ic
h
ad
d
r
ess
e
s
t
h
e
w
ea
k
n
es
s
o
f
L
ST
M
in
d
eter
m
i
n
i
n
g
f
lo
w
s
f
o
r
s
p
ec
if
ic
u
n
i
ts
,
h
a
s
also
b
ee
n
ad
d
ed
[
4
6
]
.
Fig
u
r
e
6
s
h
o
w
s
th
e
s
tr
u
ct
u
r
e
o
f
L
ST
M
R
NN.
Fig
u
r
e
6
.
L
ST
M
2
.
5
.
Ara
bic Tex
t
cla
s
s
if
ica
t
io
n pro
ce
s
s
T
h
er
e
ar
e
th
r
ee
m
a
in
s
ta
g
e
s
in
th
e
clas
s
i
f
icatio
n
o
f
tex
t
s
,
wh
ich
b
ased
o
n
t
h
e
t
y
p
e
o
f
n
e
u
r
al
n
et
w
o
r
k
u
s
ed
as s
h
o
w
n
i
n
Fi
g
u
r
e
7
.
Fig
u
r
e
7
.
A
r
ab
ic
t
ex
t
c
la
s
s
i
f
ic
atio
n
p
r
o
ce
s
s
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.
10
,
No
.
6
,
Dec
em
b
er
2
0
2
0
:
6
6
2
9
-
6
6
4
3
6634
2
.
5
.
1.
Da
t
a
pre
-
pro
ce
s
s
ing
A
t
th
is
s
ta
g
e,
t
h
e
w
o
r
d
i
s
f
il
ter
ed
an
d
r
etu
r
n
ed
to
i
ts
o
r
ig
in
al
r
o
o
t
b
y
r
e
m
o
v
i
n
g
all
th
e
ch
a
n
g
in
g
p
o
lices
r
elate
d
to
th
e
w
h
o
le,
f
o
r
in
s
tan
ce
,
h
a
m
za
h
,
T
a
Ma
r
b
o
u
ta
(
“
ة
”
)
,
tash
k
ee
l
f
r
o
m
all
t
h
e
w
o
r
d
s
,
n
u
m
b
er
s
,
an
d
p
u
n
c
tu
at
io
n
m
a
r
k
s
.
Als
o
,
th
e
w
o
r
d
s
t
h
at
l
i
n
k
w
o
r
d
s
,
s
u
c
h
as
(
“
كل
ذ
ل
،
ل
ة
ب
س
ن
ل
ا
ب
”)
,
ar
e
r
em
o
v
ed
.
P
r
ef
ix
es
a
n
d
s
u
f
f
i
x
es
ar
e
also
r
e
m
o
v
ed
,
an
d
all
w
o
r
d
s
o
f
th
e
s
a
m
e
r
o
o
t
ar
e
g
r
o
u
p
ed
.
T
h
is
s
tag
e
ai
m
s
to
g
i
v
e
p
r
ec
is
io
n
to
th
e
class
i
f
icatio
n
p
r
o
ce
s
s
an
d
s
av
e
ti
m
e
[
3
4
]
.
A
f
ter
r
e
m
o
v
i
n
g
all
o
f
th
e
ap
p
en
d
ag
es,
th
e
w
o
r
d
is
r
etu
r
n
e
d
to
th
e
r
o
o
t
b
y
t
h
r
e
e
m
et
h
o
d
s
:
“
th
e
r
o
o
t
-
b
a
s
ed
s
te
m
m
er
,
t
h
e
li
g
h
t
s
te
m
m
er
,
an
d
th
e
s
tati
s
tical
s
te
m
m
er
.
”
T
h
en
t
h
e
d
o
cu
m
e
n
t
is
class
i
f
ied
b
ased
o
n
th
e
v
ec
t
o
r
s
f
o
r
ea
ch
d
o
cu
m
e
n
t
[
3
0
]
.
2
.
5
.
2.
T
ex
t
cla
s
s
if
ica
t
io
n
T
h
e
tr
ain
in
g
p
h
a
s
e
ta
k
es
p
la
ce
h
er
e;
a
s
p
ec
if
ic
alg
o
r
it
h
m
is
ap
p
lied
to
th
e
w
o
r
d
s
o
b
tain
ed
f
r
o
m
th
e
p
r
ev
io
u
s
s
tag
e
in
o
r
d
er
to
co
m
p
lete
th
e
clas
s
i
f
icatio
n
p
r
o
ce
s
s
.
Se
v
er
al
co
m
p
ar
is
o
n
s
ca
n
also
b
e
m
ad
e
h
er
e
b
y
u
s
in
g
a
d
i
f
f
er
en
t
al
g
o
r
it
h
m
to
clas
s
i
f
y
th
e
tex
t
an
d
t
h
e
n
ch
o
o
s
i
n
g
t
h
e
b
est
i
n
ter
m
s
o
f
p
er
f
o
r
m
a
n
ce
an
d
th
e
r
esu
lt o
f
ac
c
u
r
ac
y
i
n
class
i
f
icatio
n
[
3
0
]
.
2
.
5
.
3.
E
v
a
lua
t
io
n
Af
ter
w
ar
d
s
,
an
d
at
th
is
s
ta
g
e,
th
e
ef
f
ec
tiv
e
n
es
s
o
f
class
if
ica
tio
n
is
ev
al
u
ated
;
s
ev
er
al
tec
h
n
iq
u
e
s
ar
e
u
s
ed
,
b
u
t
th
e
m
o
s
t
f
a
m
o
u
s
an
d
f
r
eq
u
e
n
tl
y
u
s
ed
o
n
e
i
n
t
h
e
class
if
icatio
n
o
f
te
x
t
s
is
“
F1
,
p
r
ec
is
io
n
a
n
d
r
ec
all”
[
3
0
]
.
3.
RE
S
E
ARCH
M
E
T
H
O
D
Fig
u
r
e
8
d
em
o
n
s
tr
ate
s
th
e
k
e
y
p
h
ase
s
w
e
f
o
llo
w
ed
.
Gen
er
all
y
,
r
esear
ch
q
u
e
s
tio
n
s
w
er
e
id
en
tif
ied
th
en
w
e
id
en
ti
f
ied
t
h
e
s
ea
r
ch
s
tr
ateg
y
a
n
d
th
e
k
e
y
w
o
r
d
s
w
e
u
s
ed
in
o
u
r
r
esear
ch
.
S
u
b
s
eq
u
en
tl
y
,
w
e
id
en
ti
f
ied
th
e
q
u
ali
t
y
a
s
s
es
s
m
en
t
q
u
esti
o
n
s
.
Af
ter
th
a
t,
ex
tr
ac
tio
n
o
f
t
h
e
p
ap
er
s
o
cc
u
r
r
ed
.
I
n
th
e
last
s
tep
,
w
e
d
id
cr
itical
an
al
y
s
is
o
f
th
e
c
h
o
s
e
n
p
ap
er
s
.
Fig
u
r
e
8
.
S
y
s
te
m
atic
r
e
v
ie
w
m
et
h
o
d
o
d
lo
y
3
.
1
.
Da
t
a
s
o
urce
T
h
is
s
tu
d
y
h
as
b
eg
u
n
i
n
J
an
u
ar
y
2
0
2
0
,
an
d
w
e
in
c
lu
d
ed
m
o
s
t
r
esear
ch
es
f
r
o
m
2
0
1
6
to
2
0
1
9
an
d
s
o
m
e
r
esear
ch
e
s
b
ef
o
r
e
t
h
at
p
er
io
d
.
T
h
e
d
atab
ases
w
e
u
s
ed
f
o
llo
w
:
I
E
E
E
,
Scien
ce
Dir
ec
t,
Sp
r
in
g
er
,
AC
M.
T
h
e
Fig
u
r
e
9
s
h
o
w
s
th
e
d
ata
s
o
u
r
ce
an
d
th
e
n
u
m
b
er
o
f
ar
ticl
es th
at
w
er
e
u
s
ed
.
Fig
u
r
e
9
.
Data
s
o
u
r
ce
3
.
2
.
Sea
rc
h str
a
t
eg
y
T
h
e
k
e
y
w
o
r
d
s
w
e
u
s
ed
to
co
ll
ec
t a
ll th
e
p
r
ev
io
u
s
s
t
u
d
ies
f
o
ll
o
w
:
-
“A
r
ab
ic
tex
t”
an
d
“
clas
s
if
icati
o
n
”
an
d
“n
e
u
r
al
n
et
w
o
r
k
s
”
-
“A
r
ab
ic
tex
t”
an
d
“
clas
s
if
icati
o
n
”
an
d
“
d
ee
p
lear
n
i
n
g
”
-
“A
r
ab
ic
s
cr
ip
t”
an
d
“
clas
s
i
f
ica
tio
n
”
an
d
“
d
ee
p
lear
n
in
g
”
-
“A
r
ab
ic
s
cr
ip
t”
an
d
“
ta
g
g
i
n
g
”
an
d
“
d
ee
p
lear
n
i
ng”
-
“A
r
ab
ic
s
cr
ip
t”
an
d
“
ca
teg
o
r
iz
atio
n
”
an
d
“
d
ee
p
lear
n
in
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:
2
0
8
8
-
8708
A
s
ystema
tic
r
ev
ie
w
o
f te
xt
cla
s
s
ifica
tio
n
r
esea
r
ch
b
a
s
ed
o
n
.
.
.
(
A
h
la
m
W
a
h
d
a
n
)
6635
3
.
3
.
I
nclus
io
n c
rit
er
ia
I
n
th
e
s
elec
tio
n
cr
iter
ia
s
tep
,
in
cl
u
s
io
n
an
d
e
x
cl
u
s
io
n
cr
iter
ia
w
er
e
s
et
to
en
s
u
r
e
th
at
th
e
r
esear
ch
in
cl
u
d
ed
in
th
i
s
s
t
u
d
y
w
a
s
v
al
u
ab
le
an
d
r
elev
a
n
t a
n
d
w
o
u
ld
lead
u
s
t
o
o
u
r
m
ai
n
ai
m
.
3
.
3
.
1.
E
x
clu
s
io
n
cr
it
er
ia
-
Mu
s
t b
e
a
tex
t c
la
s
s
i
f
ica
tio
n
s
t
u
d
y
-
Mu
s
t b
e
f
o
r
A
r
ab
ic
tex
t c
la
s
s
i
f
icatio
n
o
n
l
y
-
Mu
s
t
u
s
e
n
eu
r
al
n
et
w
o
r
k
f
o
r
tex
t c
las
s
i
f
icatio
n
-
Mu
s
t r
ep
o
r
t th
e
m
o
d
el
an
d
it
s
p
er
f
o
r
m
a
n
ce
m
ea
s
u
r
es,
f
o
r
in
s
tan
ce
,
ac
cu
r
ac
y
o
r
an
o
t
h
er
m
et
r
ic
-
Mu
s
t a
d
d
r
ess
s
u
m
m
ar
y
f
o
r
th
e
co
r
p
u
s
u
s
ed
.
3
.
3
.
2.
Select
io
n c
rit
er
ia
-
P
ap
er
s
n
o
t f
o
r
A
r
ab
ic
tex
t c
las
s
if
ica
tio
n
-
P
ap
er
d
id
n
’
t a
d
d
r
ess
th
e
ac
c
u
r
ac
y
-
P
ap
er
h
ad
n
o
t b
ee
n
p
u
b
lis
h
ed
in
j
o
u
r
n
al
o
r
co
n
f
er
en
ce
-
P
ap
er
d
id
n
o
t u
s
e
NN
3
.
4
.
Q
ua
lity
a
s
s
ess
m
ent
I
n
th
i
s
p
ar
t,
w
e
d
esig
n
ed
q
u
alit
y
as
s
es
s
m
e
n
t
q
u
e
s
tio
n
s
to
m
ak
e
a
ch
ec
k
li
s
t
f
o
r
th
e
r
e
s
ea
r
ch
an
d
en
s
u
r
e
th
at
it
w
o
u
ld
s
ati
s
f
y
t
h
e
ai
m
o
f
t
h
is
s
y
s
te
m
atic
r
ev
ie
w
.
B
elo
w
ar
e
t
h
e
q
u
est
io
n
s
.
Q1
: W
as th
e
co
r
p
u
s
id
en
ti
f
ied
an
d
d
escr
ib
ed
w
ell?
Q2
: W
as th
e
co
r
p
u
s
id
en
ti
f
ied
w
ell
r
e
g
ar
d
in
g
th
e
e
x
te
n
t o
f
tr
ain
i
n
g
a
n
d
test
i
n
g
?
Q3
: W
er
e
th
e
tex
t c
las
s
if
icatio
n
(
m
o
d
el
/ f
r
a
m
e
w
o
r
k
)
s
tep
s
d
escr
ib
ed
clea
r
ly
?
Q4
: D
id
th
e
au
t
h
o
r
m
a
k
e
co
m
p
ar
is
o
n
s
w
it
h
tech
n
iq
u
es o
th
er
th
an
t
h
e
NN
h
e
u
s
ed
?
Q5
: W
as th
e
p
er
f
o
r
m
an
ce
o
f
t
h
e
m
o
d
el
id
en
ti
f
ied
clea
r
l
y
?
Scale:
A
th
r
ee
-
p
o
in
t
s
ca
le
w
as
u
s
ed
in
t
h
is
ass
e
s
s
m
e
n
t.
I
f
th
e
p
ap
er
ad
d
r
ess
ed
th
e
q
u
e
s
tio
n
e
x
ac
tl
y
,
it
w
o
u
ld
b
e
g
r
ad
ed
1
,
an
d
if
it
d
id
n
o
t
ad
d
r
ess
it,
it
w
o
u
ld
b
e
g
r
ad
ed
0
,
b
u
t
if
it
an
s
w
er
ed
th
e
q
u
e
s
tio
n
p
ar
tiall
y
,
it
w
o
u
l
d
b
e
g
r
ad
ed
0
.
5
.
T
ab
le
1
s
h
o
w
s
t
h
e
r
es
u
l
t
o
f
ev
al
u
ati
n
g
t
h
e
r
esear
ch
u
s
in
g
t
h
e
d
esi
g
n
ed
ass
es
s
m
en
t q
u
esti
o
n
s
.
T
ab
le
1
.
Qu
alit
y
as
s
ess
m
e
n
t e
v
alu
a
tio
n
S
t
u
d
y
Q1
Q2
Q3
Q4
Q5
T
o
t
a
l
P
e
r
c
e
n
t
a
g
e
S1
1
1
1
1
0
4
8
0
%
S2
1
1
1
1
0
.
4
4
.
5
9
0
%
S3
1
1
1
1
1
5
1
0
0
%
S4
1
1
1
1
1
5
1
0
0
%
S5
1
1
1
1
1
5
1
0
0
%
S6
1
1
0
.
5
1
1
4
.
5
9
0
%
S7
1
1
1
0
1
4
8
0
%
S8
1
1
1
1
1
5
1
0
0
%
S9
1
1
1
1
1
5
1
0
0
%
S
1
0
1
0
.
5
1
0
1
3
.
5
7
0
%
S
1
1
1
0
.
5
1
1
0
3
.
5
7
0
%
S
1
2
1
1
0
.
5
1
1
4
.
5
9
0
%
4.
RE
SU
L
T
S
A
ND
D
I
SCU
SS
I
O
N
a.
R
Q1
:
W
h
ic
h
co
r
p
o
r
a
w
er
e
m
a
in
l
y
u
s
ed
f
o
r
A
r
ab
ic
tex
t
class
if
icatio
n
?
W
er
e
th
ey
o
p
en
s
o
u
r
ce
o
r
cr
e
ated
b
y
t
h
e
au
th
o
r
?
Fro
m
th
e
g
r
ap
h
s
h
o
w
n
i
n
Fi
g
u
r
e
01
,
it
is
clea
r
th
at
th
e
d
o
cu
m
e
n
t
s
th
at
w
er
e
u
s
ed
in
t
h
e
p
r
o
ce
s
s
o
f
class
i
f
y
in
g
t
h
e
A
r
ab
ic
tex
t
s
wer
e
m
o
s
t
l
y
o
p
en
s
o
u
r
ce
an
d
n
e
w
s
s
ites
w
i
th
4
s
o
u
r
ce
s
in
ea
ch
ca
s
e.
So
m
e
also
u
s
ed
b
o
o
k
s
:
t
w
o
r
esear
ch
p
ap
er
s
ap
p
lied
th
e
clas
s
i
f
icati
o
n
p
r
o
ce
s
s
to
u
s
ed
b
o
o
k
s
.
On
e
r
esear
c
h
p
ap
er
class
i
f
ied
g
u
es
t
r
ev
ie
w
s
o
f
a
h
o
tel,
w
h
ich
co
u
ld
b
e
co
n
s
id
er
ed
to
n
o
t
b
e
an
o
p
en
s
o
u
r
ce
.
T
h
at
is
,
th
e
n
u
m
b
er
o
f
o
p
en
s
o
u
r
ce
r
eso
u
r
ce
s
i
n
t
h
e
r
esear
ch
p
ap
er
s
u
n
d
er
s
tu
d
y
w
a
s
1
1
,
an
d
t
h
e
n
u
m
b
er
o
f
n
o
n
-
o
p
en
s
o
u
r
ce
o
n
es
w
a
s
o
n
l
y
o
n
e
as s
h
o
w
n
i
n
T
ab
le
2
.
b.
R
Q2
:
W
h
ich
co
u
n
tr
ie
s
f
o
c
u
s
i
n
g
o
n
p
u
b
lis
h
i
n
g
r
esear
c
h
es i
n
A
T
C
?
Ma
n
y
r
esear
c
h
er
s
f
r
o
m
A
r
ab
co
u
n
tr
ies
w
er
e
in
ter
e
s
ted
in
s
tu
d
y
i
n
g
an
d
p
u
b
lis
h
in
g
r
esear
ch
r
elate
d
to
th
e
cla
s
s
i
f
icatio
n
o
f
A
r
ab
ic
tex
t
s
.
J
o
r
d
an
w
as
at
t
h
e
f
o
r
ef
r
o
n
t
o
f
t
h
e
co
u
n
tr
ie
s
th
at
w
er
e
cla
s
s
i
f
ied
.
Af
ter
th
e
liq
u
id
atio
n
,
w
e
o
b
tain
ed
th
r
ee
r
esear
ch
p
ap
er
s
o
f
r
esear
ch
er
s
f
r
o
m
J
o
r
d
an
,
th
en
s
o
m
e
f
r
o
m
Mo
r
o
cc
o
,
th
e
Un
ited
A
r
ab
E
m
ir
ates,
a
n
d
th
e
U
n
ited
S
tates
o
f
Am
er
ica
at
a
r
ate
o
f
t
w
o
r
esear
ch
p
ap
er
s
f
r
o
m
ea
ch
co
u
n
tr
y
.
As
f
o
r
th
e
le
ast
p
u
b
lis
h
ed
co
u
n
tr
ies
in
t
h
e
p
er
io
d
id
en
tif
ied
i
n
t
h
i
s
r
esear
ch
,
A
l
g
er
ia,
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.
10
,
No
.
6
,
Dec
em
b
er
2
0
2
0
:
6
6
2
9
-
6
6
4
3
6636
Sau
d
i
A
r
ab
ia,
an
d
T
u
n
is
ia
q
u
alif
ied
,
w
ith
o
n
e
r
esear
c
h
p
ap
er
f
r
o
m
ea
ch
co
u
n
tr
y
.
I
t
is
clea
r
th
at
th
e
A
r
ab
co
u
n
tr
ies
h
av
e
s
tar
ted
to
s
t
u
d
y
a
n
d
p
u
b
li
s
h
o
n
th
e
s
u
b
j
ec
t
o
f
t
h
e
c
lass
if
ica
tio
n
o
f
A
r
ab
ic
t
ex
ts
in
o
r
d
er
to
u
s
e
th
e
la
n
g
u
a
g
e
i
n
al
m
o
s
t
all
f
iel
d
s
.
A
l
s
o
,
th
e
s
e
co
u
n
tr
ies
h
av
e
s
tar
ted
to
u
s
e
A
r
ti
f
icia
l
I
n
tel
li
g
en
ce
tech
n
iq
u
e
s
i
n
al
m
o
s
t
all
f
ield
s
.
T
h
er
ef
o
r
e,
th
e
A
r
ab
ic
lan
g
u
a
g
e
s
h
o
u
ld
m
ak
e
a
co
n
tr
ib
u
tio
n
in
t
h
e
f
ield
s
o
f
A
r
ti
f
icial
I
n
telli
g
en
ce
a
n
d
Ma
ch
i
n
e
L
ea
r
n
in
g
.
Fi
g
u
r
e
1
0
illu
s
tr
ates t
h
e
n
u
m
b
er
o
f
s
t
u
d
ied
ar
ticles
in
ea
ch
co
u
n
tr
y
.
Fig
u
r
e
10
.
co
r
p
o
r
a
w
er
e
m
ai
n
l
y
u
s
ed
in
th
e
s
elec
ted
p
ap
er
s
T
ab
le
2
.
Op
en
s
o
u
r
ce
s
co
r
p
o
r
a
u
s
ed
an
d
p
r
iv
ate
s
o
u
r
ce
s
S
t
u
d
y
N
u
mb
e
r
o
f
d
o
c
u
me
n
t
s
u
se
d
S
o
u
r
c
e
S1
3
1
9
,
2
5
4
,
1
2
4
w
o
r
d
s a
n
d
1
1
1
,
7
2
8
d
o
c
u
me
n
t
s1
T
h
r
e
e
N
e
w
sp
a
p
e
r
O
n
l
i
n
e
A
r
t
i
c
l
e
S2
2
0
0
K
a
r
t
i
c
l
e
s
S
A
N
A
D
(
A
l
a
r
a
b
i
y
a
.
n
e
t
,
a
l
k
h
a
l
e
e
j
.
a
e
a
n
d
a
k
h
b
a
r
o
n
a
.
c
o
m
S3
6
9
3
k
b
o
o
k
s
B
R
A
D
(
A
r
a
b
i
c
C
o
r
p
u
s
i
n
G
i
t
h
u
b
)
S4
5
0
7
0
d
o
c
u
me
n
t
s
C
N
N
A
r
a
b
i
c
n
e
w
s
S5
1
6
K
d
o
c
u
me
n
t
s
A
l
j
a
z
e
e
r
a
h
.
n
e
t
a
n
d
A
l
h
a
r
b
i
S
a
u
d
i
p
r
e
s
s a
g
e
n
c
y
S6
1
4
0
0
d
o
c
u
me
n
t
s
A
l
j
a
z
e
e
r
a
n
e
w
s
w
e
b
si
t
e
,
S
a
u
d
i
P
r
e
ss A
g
e
n
c
y
.
S7
1
5
K
t
w
e
e
t
s
A
r
a
S
e
n
T
i
S8
4
5
3
H
a
d
i
t
h
C
o
l
l
e
c
t
i
o
n
o
f
P
r
o
p
h
e
t
b
e
u
p
o
n
h
i
m (
ﷺ
)
h
a
d
i
t
h
f
r
o
m t
h
e
N
i
n
e
b
o
o
k
s
e
n
c
y
c
l
o
p
e
d
i
a
S9
1
0
6
5
H
a
d
i
t
h
C
o
l
l
e
c
t
i
o
n
o
f
P
r
o
p
h
e
t
b
e
u
p
o
n
h
i
m
h
a
d
i
t
h
(
ﷺ
)
S
1
0
2
2
,
4
2
9
d
o
c
u
me
n
t
s
O
p
e
n
S
o
u
r
c
e
A
r
a
b
i
c
C
o
r
p
o
r
a
(
O
S
A
C
)
S
1
1
1
1
b
o
o
k
s
1
0
T
a
sh
k
i
l
a
b
o
o
k
s
a
s
a
c
o
l
l
e
c
t
i
o
n
f
r
o
m
I
sl
a
mi
c
h
e
r
i
t
a
g
e
b
o
o
k
s
A
n
d
H
o
l
y
Q
u
r
a
n
S
1
2
2
2
9
1
r
e
v
i
e
w
s t
e
x
t
s
6
0
2
9
a
n
n
o
t
a
t
e
d
se
n
t
e
n
c
e
s
A
r
a
b
i
c
H
o
t
e
l
s’
r
e
v
i
e
w
s
Fig
u
r
e
1
1
.
C
o
u
n
tr
ie
s
p
u
p
lis
i
h
i
n
g
r
ea
s
r
ac
h
in
A
r
ab
ic
t
e
x
t
c
las
s
if
ica
tio
n
c.
R
Q3
:
W
h
ic
h
d
atab
ase
h
av
e
m
o
r
e
p
u
b
licatio
n
s
?
A
cc
o
r
d
in
g
to
th
e
c
h
ar
t
b
elo
w
in
Fi
g
u
r
e
1
2
,
a
n
u
m
b
er
o
f
w
ell
-
k
n
o
w
n
s
cien
tific
j
o
u
r
n
al
s
p
u
b
lis
h
ed
r
esear
ch
r
elate
d
to
t
h
e
clas
s
i
f
icatio
n
o
f
A
r
ab
ic
tex
ts
.
I
E
E
E
w
a
s
at
th
e
f
o
r
ef
r
o
n
t.
6
0
r
esear
ch
p
ap
er
s
w
er
e
o
b
tain
ed
f
r
o
m
I
E
E
E
,
4
0
w
er
e
o
b
tain
ed
f
r
o
m
Sp
r
in
g
er
,
an
d
1
7
w
er
e
o
b
tain
ed
f
r
o
m
A
C
M.
As
f
o
r
th
e
s
c
ien
ti
f
i
c
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:
2
0
8
8
-
8708
A
s
ystema
tic
r
ev
ie
w
o
f te
xt
cla
s
s
ifica
tio
n
r
esea
r
ch
b
a
s
ed
o
n
.
.
.
(
A
h
la
m
W
a
h
d
a
n
)
6637
j
o
u
r
n
als
in
w
h
ic
h
t
h
e
n
u
m
b
er
o
f
p
u
b
lis
h
ed
p
ap
er
s
o
n
th
e
t
o
p
ic
w
as
s
m
a
ll,
th
e
y
w
er
e
E
m
er
ald
an
d
E
ls
e
v
ier
w
it
h
8
an
d
4
r
esear
ch
p
ap
er
s
o
n
av
e
r
ag
e,
r
esp
ec
ti
v
el
y
.
T
h
e
r
ea
s
o
n
f
o
r
I
E
E
E
’
s
p
o
s
itio
n
at
th
e
to
p
w
a
s
th
e
h
o
ld
in
g
o
f
a
n
u
m
b
er
o
f
co
n
f
er
e
n
ce
s
w
i
th
th
i
s
f
o
cu
s
,
w
it
h
m
an
y
r
esear
ch
er
s
p
ar
ticip
ati
n
g
in
t
h
i
s
f
ield
w
it
h
th
eir
r
esear
ch
p
ap
er
s
.
I
n
all
c
ases
,
th
e
s
u
b
j
ec
t
o
f
th
e
c
lass
if
icatio
n
o
f
A
r
ab
ic
te
x
ts
i
s
s
til
l
u
n
d
er
d
ev
elo
p
m
e
n
t
w
h
e
n
co
m
p
ar
ed
to
th
at
o
f
o
th
e
r
lan
g
u
ag
e
s
,
s
u
ch
a
s
E
n
g
lis
h
a
n
d
Sp
an
is
h
.
Fig
u
r
e
1
2
.
Nu
m
b
er
o
f
A
r
ticles
b
ased
o
n
s
cien
ce
j
o
u
r
n
al
s
d.
R
Q4
:
Du
r
i
n
g
w
h
ic
h
y
ea
r
s
w
as
th
er
e
m
o
r
e
f
o
cu
s
o
n
A
T
C
?
B
ased
o
n
th
e
g
r
ap
h
ill
u
s
tr
ate
d
in
Fig
u
r
e
1
3
,
th
e
p
er
io
d
f
r
o
m
2
0
1
7
to
2
0
1
9
w
it
n
e
s
s
ed
a
s
ig
n
i
f
ica
n
t
d
ev
elo
p
m
en
t
in
th
e
f
ield
o
f
A
r
ab
ic
tex
t
clas
s
i
f
icatio
n
r
e
s
ea
r
ch
.
Si
x
r
esear
ch
p
ap
er
s
w
er
e
d
is
tr
ib
u
ted
b
et
w
ee
n
2
0
1
8
an
d
2
0
1
9
,
th
at
is
,
th
r
ee
r
esear
ch
p
ap
er
s
f
o
r
ea
ch
y
ea
r
.
I
n
th
e
p
er
io
d
b
et
w
ee
n
2
0
1
3
a
n
d
2
0
1
6
,
o
n
ly
f
o
u
r
r
esear
ch
p
ap
er
s
w
er
e
s
t
u
d
ied
;
th
e
y
w
er
e
d
is
tr
ib
u
ted
as
f
o
llo
w
s
:
o
n
e
ar
ticle
i
n
2
0
1
5
a
n
d
th
r
ee
in
2
0
1
6
.
On
l
y
t
w
o
r
esear
c
h
p
ap
er
s
w
e
r
e
o
b
tain
ed
b
etw
ee
n
2
0
0
9
an
d
2
0
1
2
.
Fr
o
m
t
h
e
d
ata
w
e
h
a
v
e,
it
is
clea
r
th
at
r
esear
ch
in
ter
est
i
n
t
h
e
f
ie
ld
o
f
t
h
e
clas
s
if
icatio
n
o
f
A
r
ab
ic
t
ex
ts
i
s
s
tar
ti
n
g
to
in
cr
ea
s
e
d
u
e
to
th
e
i
m
p
o
r
ta
n
ce
o
f
th
e
A
r
ab
ic
lan
g
u
ag
e,
it
s
u
s
e
in
m
o
r
e
th
a
n
o
n
e
f
ield
,
an
d
its
u
s
e
i
n
m
o
r
e
A
r
ab
I
s
la
m
ic
co
u
n
tr
ies.
Fig
u
r
e
1
3
.
Nu
m
b
er
o
f
a
r
ticles
s
elec
ted
p
er
y
ea
r
e.
R
Q5
:
W
h
at
w
er
e
t
h
e
t
y
p
e
s
o
f
n
eu
r
al
n
et
w
o
r
k
al
g
o
r
ith
m
s
u
s
ed
to
class
if
y
A
r
ab
ic
te
x
ts
?
An
d
w
h
ic
h
o
n
e
w
a
s
m
o
s
t u
s
ed
?
Ma
n
y
t
y
p
es
o
f
n
e
u
r
al
n
et
w
o
r
k
s
w
er
e
u
s
ed
i
n
p
r
ev
io
u
s
r
esear
ch
.
T
h
e
y
in
cl
u
d
ed
co
n
v
o
l
u
tio
n
al
n
e
u
r
al
n
et
w
o
r
k
s
(
C
NN)
,
r
ec
u
r
r
en
t
n
eu
r
al
n
et
w
o
r
k
s
(
R
NN)
,
B
ir
NN
,
L
ST
M,
th
e
d
ee
p
b
elie
f
n
et
w
o
r
k
(
DB
N)
,
b
ac
k
p
r
o
p
ag
atio
n
au
to
e
n
co
d
er
(
B
P
NN)
,
th
e
m
u
ltil
a
y
er
p
er
ce
p
tr
o
n
(
ML
P
)
,
an
d
f
ee
d
f
o
r
w
ar
d
n
eu
r
al
n
et
w
o
r
k
s
(
FF
NN)
.
L
ST
M
an
d
C
NN
w
e
r
e
th
e
m
o
s
t
u
s
ed
n
e
u
r
al
n
et
w
o
r
k
s
.
B
o
th
w
er
e
u
s
ed
in
f
o
u
r
s
t
u
d
ies.
L
ST
M
w
er
e
u
s
ed
i
n
[
4
7
-
50]
.
C
o
n
v
o
lu
tio
n
al
n
e
u
r
al
n
et
w
o
r
k
s
(
C
NN)
w
er
e
u
s
ed
to
clas
s
if
y
A
r
a
b
ic
tex
t
i
n
[
5
0
-
53]
.
R
ec
u
r
r
en
t
n
e
u
r
al
n
et
w
o
r
k
s
(
R
NN)
w
er
e
u
s
ed
i
n
o
n
l
y
o
n
e
s
tu
d
y
[
5
2
]
.
I
n
th
is
s
tu
d
y
,
t
h
e
a
u
th
o
r
al
s
o
tr
ied
to
b
u
ild
en
s
e
m
b
le
m
et
h
o
d
s
,
w
h
ic
h
co
m
b
i
n
ed
R
N
N
w
it
h
C
NN,
i
n
o
r
d
er
to
i
m
p
r
o
v
e
ef
f
icie
n
c
y
.
I
n
th
e
s
a
m
e
s
tu
d
y
,
an
o
th
er
m
o
d
el
w
a
s
b
u
il
t
u
s
in
g
a
s
p
ec
ial
t
y
p
e
o
f
R
NN:
B
i
R
N
N.
T
h
r
ee
-
la
y
er
f
ee
d
-
f
o
r
w
ar
d
NN
h
a
v
e
b
ee
n
u
s
ed
t
w
ice
i
n
[
4
0
]
an
d
in
[
5
4
]
.
B
u
t,
in
[
5
5
]
,
th
e
m
u
l
tila
y
er
p
e
r
ce
p
tr
o
n
(
ML
P
)
w
as
u
s
ed
as
a
p
ar
t
o
f
3
-
la
y
er
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.
10
,
No
.
6
,
Dec
em
b
er
2
0
2
0
:
6
6
2
9
-
6
6
4
3
6638
f
ee
d
-
f
o
r
w
ar
d
NN.
T
h
e
d
ee
p
b
elief
n
et
w
o
r
k
(
DB
N)
w
as
o
n
l
y
u
s
ed
i
n
o
n
e
s
tu
d
y
[
5
6
]
.
B
ac
k
p
r
o
p
ag
atio
n
au
to
en
co
d
er
w
as
u
s
ed
j
u
s
t
o
n
ce
in
[
5
7
]
.
Fig
u
r
e
1
4
s
h
o
w
s
t
h
e
t
y
p
es
o
f
NN
an
d
h
o
w
m
an
y
ti
m
e
s
it
h
as
b
ee
n
u
s
ed
in
all
r
ev
ie
w
ed
p
ap
er
s
.
T
ab
le
3
s
h
o
w
s
t
h
e
r
ef
er
en
ce
s
f
o
r
ea
ch
t
y
p
e
o
f
n
e
u
r
al
n
et
w
o
r
k
f
o
u
n
d
ed
in
ea
c
h
p
ap
er
.
T
h
e
p
ap
er
s
d
id
n
o
t
m
e
n
tio
n
th
e
r
ea
s
o
n
b
e
h
in
d
c
h
o
o
s
in
g
t
h
ese
t
y
p
es
o
f
NN.
W
e
b
el
iev
e
t
h
at
t
h
e
r
ea
s
o
n
s
h
o
u
ld
b
e
s
tated
.
Fig
u
r
e
1
4
.
T
y
p
es o
f
N
N
T
ab
le
3
.
T
y
p
es o
f
NN
Ty
p
e
s
o
f
N
N
R
e
f
e
r
e
n
c
e
s
C
N
N
[
4
7
-
5
0
]
L
S
T
M
[
5
0
-
5
3
]
R
N
N
[
5
2
]
C
N
N
&
R
N
N
[
5
2
]
.
F
e
e
d
f
o
r
w
a
r
d
N
N
(
F
F
N
N
)
[
4
0
,
5
4
]
M
L
P
[
5
5
]
D
B
N
[
5
6
]
B
a
c
k
p
r
o
p
a
g
a
t
i
o
n
A
u
t
o
_
e
n
c
o
d
e
r
3
l
a
y
e
r
s (BPN
N
)
[
5
7
]
f.
R
Q6
: W
h
at
w
er
e
th
e
e
f
f
icie
n
c
y
m
ea
s
u
r
e
s
u
s
ed
?
T
h
e
ef
f
icie
n
c
y
m
ea
s
u
r
e
m
ai
n
l
y
u
s
ed
i
n
th
e
s
e
s
t
u
d
ies
was
p
r
i
m
ar
il
y
ac
c
u
r
ac
y
;
t
h
e
o
th
er
s
u
s
ed
in
cl
u
d
ed
R
ec
all,
F
-
m
ea
s
u
r
e,
an
d
P
r
ec
is
io
n
a
s
s
h
o
w
n
i
n
Fi
g
u
r
e
1
5
.
A
cc
u
r
ac
y
w
as
u
s
ed
in
8
s
tu
d
ie
s
i
n
o
u
r
s
y
s
te
m
a
tic
r
ev
ie
w
as
s
h
o
w
n
in
T
a
b
le
4
,
b
u
t
n
o
t
all
o
f
th
ese
s
t
u
d
ies
id
en
ti
f
ied
it
w
ell.
A
c
cu
r
ac
y
w
as
id
en
ti
f
ied
w
ell
a
n
d
th
e
r
esear
c
h
er
s
clea
r
e
d
th
e
d
ef
in
itio
n
o
r
th
e
eq
u
at
io
n
in
o
n
l
y
f
o
u
r
s
t
u
d
ies
:
[
4
8
-
5
0
,
5
7
]
.
T
h
e
R
ec
all
ef
f
icie
n
c
y
m
ea
s
u
r
e
ca
m
e
i
n
s
ec
o
n
d
p
o
s
itio
n
:
it
w
as
u
s
ed
in
f
i
v
e
s
t
u
d
ies.
P
r
ec
is
io
n
an
d
F
-
m
ea
s
u
r
e
w
er
e
th
e
least
u
s
e
d
;
th
e
y
w
er
e
u
s
ed
in
o
n
l
y
4
s
t
u
d
ies.
T
ab
le
4
s
h
o
w
s
t
h
e
e
f
f
ic
ien
c
y
m
ea
s
u
r
e
t
h
at
u
s
ed
in
ea
c
h
p
ap
er
.
Fig
u
r
e
1
5
.
E
f
f
icie
n
c
y
m
ea
s
u
r
e
s
Evaluation Warning : The document was created with Spire.PDF for Python.