T
E
L
K
O
M
NIKA
T
elec
o
mm
un
ica
t
io
n,
Co
m
pu
t
ing
,
E
lect
ro
nics
a
nd
Co
ntr
o
l
Vo
l.
18
,
No
.
4
,
A
u
g
u
s
t
2020
,
p
p
.
1
7
7
7
~
1
7
8
3
I
SS
N:
1
6
9
3
-
6
9
3
0
,
ac
cr
ed
ited
First Gr
ad
e
b
y
Ke
m
e
n
r
is
te
k
d
i
k
ti,
Dec
r
ee
No
: 2
1
/E/KPT
/2
0
1
8
DOI
: 1
0
.
1
2
9
2
8
/
T
E
L
KOM
NI
KA
.
v
1
8
i4
.
1
3
8
4
9
1777
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
Enha
nce
m
ent
of s
tudent
per
for
m
a
n
ce pre
di
ction
using
m
o
di
fied
K
-
nea
rest neighbo
r
Sa
j
a
T
a
ha
Ah
m
e
d
1
,
Ra
f
a
h Al
-
H
a
m
da
ni
2
,
M
ua
y
a
d Sa
dik
Cro
o
ck
3
1
,2
T
h
e
In
f
o
rm
a
ti
c
s In
stit
u
te f
o
r
P
o
stg
ra
d
u
a
te S
tu
d
ies
,
Ira
q
i
C
o
m
m
is
sio
n
f
o
r
Co
m
p
u
ters
&
I
n
f
o
rm
a
ti
c
s
(II
P
S
-
ICCI)
,
Ira
q
3
Co
m
p
u
ter E
n
g
in
e
e
rin
g
De
p
a
rtm
e
n
t,
Un
iv
e
rsity
o
f
T
e
c
h
n
o
lo
g
y
,
Ira
q
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Au
g
9
,
2019
R
ev
i
s
ed
Feb
2
5
,
2
0
2
0
A
cc
ep
ted
Ma
r
1
8
,
2
0
2
0
T
h
e
trad
it
io
n
a
l
K
-
n
e
a
re
st
n
e
ig
h
b
o
r
(KN
N)
a
lg
o
rit
h
m
u
se
s
a
n
e
x
h
a
u
stiv
e
se
a
rc
h
f
o
r
a
c
o
m
p
lete
train
in
g
se
t
to
p
re
d
ict
a
sin
g
le
tes
t
sa
m
p
le.
T
h
is
p
r
o
c
e
d
u
re
c
a
n
slo
w
d
o
w
n
th
e
sy
ste
m
to
c
o
n
su
m
e
m
o
re
ti
m
e
f
o
r
h
u
g
e
d
a
tas
e
ts.
T
h
e
se
lec
ti
o
n
o
f
c
las
se
s
f
o
r
a
n
e
w
sa
m
p
le
d
e
p
e
n
d
s
o
n
a
sim
p
le
m
a
jo
rit
y
v
o
ti
n
g
sy
ste
m
th
a
t
d
o
e
s
n
o
t
re
f
lec
t
th
e
v
a
rio
u
s
sig
n
if
ica
n
c
e
o
f
d
iff
e
re
n
t
sa
m
p
les
(i.
e
.
ig
n
o
ri
n
g
th
e
sim
il
a
rit
ies
a
m
o
n
g
sa
m
p
le
s).
It
a
lso
lea
d
s
to
a
m
isc
la
ss
if
ic
a
ti
o
n
p
ro
b
lem
d
u
e
t
o
t
h
e
o
c
c
u
rre
n
c
e
o
f
a
d
o
u
b
le
m
a
jo
rit
y
c
las
s.
In
r
e
fe
re
n
c
e
to
th
e
a
b
o
v
e
-
m
e
n
t
io
n
e
d
issu
e
s,
t
h
i
s
w
o
rk
a
d
o
p
ts
a
c
o
m
b
in
a
ti
o
n
o
f
m
o
m
e
n
t
d
e
sc
rip
to
r
a
n
d
KN
N
to
o
p
ti
m
ize
th
e
sa
m
p
le
s
e
lec
ti
o
n
.
T
h
is
is
d
o
n
e
b
a
se
d
o
n
th
e
f
a
c
t
th
a
t
c
las
sify
in
g
th
e
train
i
n
g
sa
m
p
les
b
e
f
o
re
th
e
se
a
rc
h
in
g
a
c
t
u
a
ll
y
tak
e
s
p
lac
e
c
a
n
sp
e
e
d
u
p
a
n
d
im
p
ro
v
e
th
e
p
re
d
icti
v
e
p
e
rf
o
r
m
a
n
c
e
o
f
t
h
e
n
e
a
re
st
n
e
ig
h
b
o
r
.
T
h
e
p
ro
p
o
se
d
m
e
th
o
d
c
a
n
b
e
c
a
ll
e
d
a
s
f
a
st
KN
N
(F
KN
N).
T
h
e
e
x
p
e
ri
m
e
n
tal
re
su
lt
s
sh
o
w
th
a
t
t
h
e
p
r
o
p
o
se
d
F
KN
N
m
e
th
o
d
d
e
c
re
a
se
s
o
rig
in
a
l
KN
N
c
o
n
s
u
m
in
g
ti
m
e
w
it
h
in
a
ra
n
g
e
o
f
(7
5
.
4
%
)
to
(9
0
.
2
5
%
),
a
n
d
im
p
ro
v
e
th
e
c
las
sif
i
c
a
ti
o
n
a
c
c
u
ra
c
y
p
e
rc
e
n
tag
e
in
th
e
ra
n
g
e
f
ro
m
(2
0
%
)
to
(3
6
.
3
%
)
u
ti
l
izin
g
th
re
e
ty
p
e
s
o
f
st
u
d
e
n
t
d
a
tas
e
ts
to
p
re
d
ict
w
h
e
th
e
r
th
e
stu
d
e
n
t
c
a
n
p
a
ss
o
r
f
a
il
th
e
e
x
a
m
a
u
to
m
a
t
ica
ll
y
.
K
ey
w
o
r
d
s
:
C
o
n
s
u
m
i
n
g
ti
m
e
E
d
u
ca
tio
n
al
d
ata
m
i
n
i
n
g
Mo
m
en
t
s
KNN
P
r
ed
ictio
n
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
:
Saj
a
T
ah
a
Ah
m
ed
,
T
h
e
I
n
f
o
r
m
atic
s
I
n
s
t
itu
te
f
o
r
P
o
s
tg
r
ad
u
ate
St
u
d
ies,
I
r
aq
i Co
m
m
i
s
s
io
n
f
o
r
C
o
m
p
u
t
er
s
&
I
n
f
o
r
m
atic
s
(
I
I
P
S
-
I
C
C
I
)
,
I
r
aq
.
E
m
ail:
s
aj
atah
a
@
y
m
ail.
co
m
1.
I
NT
RO
D
UCT
I
O
N
T
h
e
g
r
o
w
th
o
f
t
h
e
in
ter
n
et
an
d
co
m
m
u
n
icatio
n
tech
n
o
lo
g
ie
s
h
av
e
co
n
tr
ib
u
ted
to
th
e
d
is
s
e
m
i
n
atio
n
o
f
e
-
lear
n
i
n
g
to
s
u
p
p
o
r
t
ce
r
tain
co
u
n
tr
ie
s
co
n
f
r
o
n
ti
n
g
a
r
i
s
in
g
s
ca
r
cit
y
o
f
i
n
s
tr
u
cto
r
s
[
1
]
.
I
t
is
e
v
id
en
t
t
h
at
u
n
d
er
s
t
u
d
ies,
o
r
in
d
i
v
id
u
al
s
in
g
en
er
al,
w
h
o
ar
e
lo
o
k
i
n
g
f
o
r
in
f
o
r
m
atio
n
ca
n
ac
h
iev
e
t
h
i
s
e
f
f
ec
tiv
e
l
y
a
n
d
w
it
h
m
i
n
i
m
al
e
f
f
o
r
t
at
a
n
y
ti
m
e
a
n
d
an
y
w
h
er
e.
T
h
is
en
co
u
r
a
g
e
s
v
ar
io
u
s
co
lle
g
es
a
n
d
in
s
tr
u
ctiv
e
o
r
g
a
n
izatio
n
s
to
a
d
o
p
t
an
o
n
lin
e
lear
n
in
g
f
r
a
m
e
w
o
r
k
w
it
h
an
ex
te
n
s
io
n
o
f
th
e
s
tu
d
e
n
t
d
ata
v
o
lu
m
e.
Ho
w
e
v
er
,
e
-
lear
n
i
n
g
h
as
a
lo
t
o
f
i
m
p
ed
i
m
e
n
ts
a
n
d
ch
allen
g
es
an
d
d
r
o
p
-
o
u
t
r
ates
f
o
r
s
tu
d
en
ts
ar
e
m
o
r
e
co
m
m
o
n
th
a
n
co
n
v
en
t
io
n
al
lear
n
in
g
[
2
,
3
].
T
h
e
ed
u
ca
tio
n
al
d
ata
m
i
n
i
n
g
(
E
DM
)
is
u
t
ilized
to
d
ev
elo
p
a
m
o
d
el
t
h
at
ca
n
i
n
f
lu
e
n
ce
e
-
lear
n
i
n
g
s
y
s
te
m
b
ec
a
u
s
e
d
ata
g
at
h
er
ed
f
r
o
m
e
-
lear
n
in
g
s
y
s
te
m
o
f
ten
e
x
ce
ed
s
lar
g
e
n
u
m
b
er
s
o
f
s
t
u
d
en
t
s
[
4
]
.
T
h
e
in
v
o
lv
e
m
e
n
t
o
f
d
if
f
er
en
t
v
ar
iab
les
th
at
E
DM
ca
n
ex
p
lo
i
t
f
o
r
m
o
d
el
b
u
ild
i
n
g
h
elp
s
th
e
ed
u
ca
ti
o
n
al
s
y
s
te
m
to
p
er
f
o
r
m
b
etter
.
T
h
e
d
ev
elo
p
ed
m
o
d
el
ca
n
s
u
p
p
o
r
t
th
e
d
ec
is
io
n
m
ak
in
g
o
f
ed
u
ca
tio
n
al
in
s
tit
u
tio
n
s
an
d
u
n
i
v
er
s
itie
s
ab
o
u
t
f
u
tu
r
e
o
f
th
e
ir
s
t
u
d
en
t
s
,
f
o
r
ex
a
m
p
le
,
d
is
tin
g
u
is
h
i
n
g
e
f
f
ec
ti
v
e
s
tu
d
en
ts
f
r
o
m
a
g
i
v
e
n
co
u
r
s
e
an
d
p
er
ce
i
v
i
n
g
s
t
u
d
en
t
s
w
h
o
w
ill
d
r
o
p
o
u
t
o
r
f
ail
to
p
a
y
m
o
r
e
co
n
s
id
er
atio
n
d
u
r
i
n
g
co
u
r
s
e
p
r
o
g
r
ess
[
5
].
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6930
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
,
Vo
l.
18
,
No
.
4
,
A
u
g
u
s
t 2
0
2
0
:
1
7
7
7
-
1
7
8
3
1778
T
h
e
K
-
t
h
e
n
ea
r
est
n
ei
g
h
b
o
r
is
o
n
e
o
f
th
e
s
i
m
p
lest
E
DM
al
g
o
r
ith
m
s
[
6
]
.
I
t
is
co
m
p
u
tatio
n
a
ll
y
s
i
m
p
le
b
ased
o
n
s
i
m
ilar
i
t
y
m
ea
s
u
r
es
s
u
ch
a
s
a
E
u
clid
ea
n
d
i
s
tan
ce
m
etr
ic
w
it
h
m
aj
o
r
ity
v
o
tin
g
o
f
t
h
e
K
clo
s
est
tr
ain
i
n
g
s
a
m
p
le
clas
s
ass
i
g
n
ed
to
th
e
test
s
a
m
p
le
[
7
]
.
KNN
is
an
in
s
ta
n
ce
-
b
ased
lear
n
er
,
s
o
m
e
ti
m
es
ca
lled
a
lazy
lear
n
er
,
as
it
d
ef
er
s
t
h
e
tr
ain
in
g
u
n
til
a
n
e
w
s
tu
d
e
n
t
(
test
s
a
m
p
le)
s
h
o
u
ld
b
e
class
i
f
ied
(
i.e
.
th
er
e
i
s
n
o
tr
ain
i
n
g
p
h
a
s
e)
w
it
h
m
o
s
t t
h
e
p
o
w
er
r
elies o
n
m
atc
h
in
g
s
c
h
e
m
e
[
8
]
.
KNN
h
a
s
s
o
m
e
co
n
s
th
a
t c
an
b
e
lis
ted
as [
9
,
10
]:
-
C
o
m
p
u
tatio
n
al
o
v
er
h
ea
d
i
s
e
x
tr
e
m
e
l
y
h
i
g
h
as
ea
ch
n
e
w
s
tu
d
en
t
n
ee
d
s
to
ca
lcu
late
t
h
e
d
is
tan
ce
to
all
tr
ain
i
n
g
s
a
m
p
le
s
.
-
T
h
e
ca
p
ac
ity
r
eq
u
ir
e
m
en
t i
s
h
u
g
e
i
n
p
r
o
p
o
r
tio
n
to
th
e
tr
ain
i
n
g
s
ize
s
et.
-
KNN
w
it
h
m
u
ltid
i
m
e
n
s
io
n
al
d
ata
s
ets h
a
s
a
m
in
i
m
u
m
ac
cu
r
a
c
y
r
ate.
R
esear
ch
er
s
o
f
f
er
ed
a
v
ar
iet
y
o
f
tech
n
iq
u
es
f
o
r
d
ea
lin
g
w
it
h
th
e
is
s
u
es
o
f
tr
ad
itio
n
al
KNN
alg
o
r
ith
m
an
d
i
m
p
r
o
v
i
n
g
it
s
p
er
f
o
r
m
a
n
c
e.
T
h
e
au
th
o
r
s
o
f
[
1
1
]
p
r
o
p
o
s
ed
th
at
t
h
e
g
en
et
ic
al
g
o
r
ith
m
(
GA
)
a
n
d
KNN
w
er
e
co
m
b
i
n
ed
to
i
m
p
r
o
v
e
t
h
e
c
lass
i
f
icatio
n
p
er
f
o
r
m
a
n
ce
.
G
A
w
as
u
s
ed
to
in
s
ta
n
tl
y
p
i
ck
u
p
k
-
n
eig
h
b
o
r
s
an
d
ca
lcu
late
t
h
e
d
is
tan
ce
to
cl
ass
i
f
y
t
h
e
tes
t
s
a
m
p
les.
T
h
e
p
r
o
p
o
s
ed
m
et
h
o
d
w
as
co
m
p
ar
ed
w
it
h
th
e
tr
ad
i
tio
n
al
KNN,
C
AR
T
an
d
SVM
cla
s
s
i
f
ier
s
.
T
h
e
r
es
u
lts
s
h
o
w
ed
th
at
t
h
e
p
r
o
p
o
s
ed
m
eth
o
d
r
ed
u
ce
d
co
m
p
le
x
it
y
an
d
i
m
p
r
o
v
e
ac
cu
r
ac
y
.
T
h
e
au
th
o
r
s
in
[
1
2
]
s
o
lv
ed
th
e
lar
g
e
s
a
m
p
le
co
m
p
u
tat
io
n
p
r
o
b
lem
u
s
in
g
a
c
u
r
e
cl
u
s
ter
i
n
g
al
g
o
r
ith
m
w
it
h
KNN
to
o
b
tain
r
ep
r
esen
tativ
e
s
a
m
p
les
o
f
th
e
o
r
ig
i
n
al
d
ataset
f
o
r
tex
t
ca
teg
o
r
izatio
n
.
T
h
e
p
r
o
p
o
s
ed
m
et
h
o
d
class
i
f
ied
6
5
0
0
n
e
w
s
e
s
s
a
y
s
f
r
o
m
8
ca
te
g
o
r
ies
o
f
Si
n
a
w
eb
s
i
tes
w
it
h
i
m
p
r
o
v
ed
co
m
p
u
tatio
n
s
p
ee
d
co
m
p
ar
ed
to
tr
ad
itio
n
al
KNN
b
u
t
d
i
d
n
o
t
en
h
an
ce
th
e
ac
cu
r
ac
y
o
f
KNN,
w
h
ich
is
co
n
s
i
d
er
ed
a
s
a
m
aj
o
r
lim
ita
tio
n
o
f
th
e
p
r
o
p
o
s
ed
m
et
h
o
d
.
T
h
e
au
th
o
r
in
[
1
3
]
f
o
cu
s
ed
o
n
im
p
r
o
v
i
n
g
th
e
p
er
f
o
r
m
an
ce
o
f
KNN
b
y
co
m
b
i
n
i
n
g
lo
ca
l
m
ea
n
b
ased
KNN
w
ith
d
is
ta
n
ce
w
ei
g
h
t
KNN.
T
h
e
p
r
o
p
o
s
ed
m
eth
o
d
w
as
ap
p
lied
to
f
o
u
r
d
atasets
f
r
o
m
UC
I
,
k
a
g
g
le,
an
d
k
ee
l,
in
ad
d
itio
n
to
a
r
ea
l
d
ataset
f
r
o
m
p
u
b
lic
s
e
n
io
r
h
ig
h
s
c
h
o
o
l.
T
h
e
o
b
tain
ed
r
esu
lt
s
ap
p
ea
r
ed
th
at
th
e
clas
s
i
f
icatio
n
ac
c
u
r
ac
y
o
f
t
h
e
p
r
o
p
o
s
ed
m
et
h
o
d
co
m
p
ar
ed
to
KNN
w
a
s
i
n
cr
ea
s
ed
,
b
u
t
t
h
is
r
esear
ch
i
g
n
o
r
ed
th
e
co
m
p
le
x
it
y
o
f
ex
ec
u
tio
n
ti
m
e
r
esu
lted
f
r
o
m
t
h
e
m
ix
in
g
o
f
p
r
o
p
o
s
ed
m
et
h
o
d
s
.
T
h
e
KNN
co
m
p
u
tatio
n
al
co
m
p
lex
it
y
f
o
r
class
if
y
i
n
g
a
s
in
g
le
n
e
w
i
n
s
ta
n
ce
is
O(
n
)
,
w
h
er
e
n
is
a
n
u
m
b
er
o
f
tr
ain
i
n
g
s
a
m
p
le
s
[
1
4
]
.
T
h
er
ef
o
r
e,
in
t
h
i
s
s
t
u
d
y
,
th
e
p
r
o
to
t
y
p
e
s
to
r
ag
e,
co
m
p
u
tat
io
n
ti
m
e
an
d
ac
cu
r
ac
y
h
av
e
a
g
r
ea
t
d
ea
l
o
f
an
a
ly
s
is
.
T
h
is
p
ap
er
p
r
o
p
o
s
ed
a
s
o
lu
tio
n
b
y
in
tr
o
d
u
ci
n
g
an
ac
ce
ler
atio
n
s
c
h
e
m
e
to
o
v
er
co
m
e
KNN
d
r
a
w
b
ac
k
s
v
ia
a
co
m
b
in
atio
n
o
f
m
o
m
e
n
t
d
escr
ip
to
r
s
w
ith
tr
ad
itio
n
al
KNN.
T
h
e
m
o
m
en
t
d
escr
ip
to
r
s
h
av
e
b
ee
n
u
tili
ze
d
w
ell
i
n
m
u
lti
m
e
d
ia
r
esear
ch
f
o
r
v
ar
io
u
s
ap
p
li
ca
tio
n
s
,
s
u
ch
a
s
m
u
s
ica
l
s
i
m
il
ar
it
y
an
d
s
o
n
g
y
ea
r
p
r
ed
ictio
n
[
15
]
,
s
p
ee
d
u
p
c
o
lo
r
im
a
g
e
f
r
ac
tal
co
m
p
r
es
s
io
n
[
16
]
an
d
en
h
an
ce
f
r
ac
tal
au
d
io
co
m
p
r
ess
io
n
[
17
].
T
h
e
tr
ain
in
g
s
et
w
ill
b
e
ar
r
an
g
ed
i
n
to
s
u
b
s
et
s
;
s
a
m
p
les
b
elo
n
g
to
th
e
s
a
m
e
s
u
b
s
et
h
av
e
s
i
m
ilar
d
escr
ip
to
r
n
u
m
b
er
.
T
h
e
p
r
o
p
o
s
ed
FKNN
d
o
es
n
o
t
h
a
v
e
to
test
e
ac
h
n
e
w
s
a
m
p
le
(
i.e
.
,
co
m
p
u
te
it
s
d
is
tan
ce
)
w
it
h
a
ll
tr
ai
n
in
g
s
a
m
p
les.
b
u
t
,
ea
ch
te
s
t
s
a
m
p
le
(
n
e
w
s
t
u
d
en
t)
w
h
en
th
e
p
r
o
p
o
s
ed
FKNN
co
m
p
u
tes
it
s
d
escr
ip
to
r
v
alu
e
is
m
atc
h
ed
o
n
l
y
w
it
h
a
p
r
ed
e
ter
m
i
n
ed
s
u
b
s
et
o
f
tr
ain
in
g
s
a
m
p
le
s
w
h
ic
h
h
a
s
s
i
m
ila
r
d
escr
ip
to
r
v
alu
e.
T
h
is
s
ig
n
i
f
ica
n
tl
y
r
ed
u
ce
s
t
h
e
ex
ec
u
tio
n
ti
m
e
(
co
m
p
ar
is
o
n
d
is
ta
n
ce
ti
m
e)
an
d
m
e
m
o
r
y
r
eq
u
ir
e
m
en
t
s
.
I
n
ad
d
itio
n
,
ea
ch
tr
ai
n
i
n
g
s
u
b
s
et
is
f
o
r
m
ed
o
n
t
h
e
b
asi
s
o
f
a
w
ei
g
h
ted
m
o
m
e
n
t
d
escr
ip
to
r
th
at
ca
p
tu
r
es
th
e
i
m
p
o
r
tan
ce
o
f
s
elec
ted
attr
ib
u
tes
f
o
r
d
if
f
er
en
t
s
a
m
p
le
s
,
th
is
en
ab
les
ea
c
h
tr
ain
i
n
g
s
u
b
s
e
t
to
co
n
tain
th
e
m
o
s
t
s
i
m
ilar
s
a
m
p
le
s
.
I
t,
in
t
u
r
n
,
i
n
c
r
ea
s
es th
e
ac
c
u
r
ac
y
o
f
a
class
i
f
icatio
n
an
d
av
o
id
s
d
o
u
b
le
m
a
j
o
r
ity
class
if
icatio
n
(
i.e
.
m
i
s
clas
s
i
f
icatio
n
)
.
2.
RE
S
E
ARCH
M
E
T
H
O
D
T
h
is
s
tu
d
y
w
ill i
n
cl
u
d
e
t
w
o
p
h
ases
as a
p
ar
t o
f
th
e
m
et
h
o
d
o
lo
g
y
,
as
f
o
llo
w
:
2
.
1
.
Da
t
a
s
et
co
llect
i
o
n a
nd
prepa
ra
t
io
n
T
h
e
co
llectio
n
,
p
r
ep
r
o
ce
s
s
in
g
an
d
f
ea
t
u
r
e
s
elec
tio
n
o
f
d
ata
s
ets
ar
e
d
o
n
e
b
ased
o
n
o
u
r
r
esear
ch
w
o
r
k
o
f
[
1
8
]
.
T
h
is
s
tu
d
y
u
s
ed
t
h
r
ee
d
atasets
,
th
e
f
ir
s
t
b
ei
n
g
t
h
e
I
r
aq
i
s
tu
d
en
t
p
er
f
o
r
m
an
ce
p
r
ed
ictio
n
d
ataset,
w
h
ic
h
is
co
llected
th
r
o
u
g
h
ap
p
ly
in
g
(
o
r
s
u
b
m
itti
n
g
)
q
u
esti
o
n
n
air
e
i
n
th
r
ee
I
r
aq
i
s
ec
o
n
d
ar
y
s
ch
o
o
ls
f
o
r
b
o
th
ap
p
licab
le
an
d
b
io
lo
g
y
b
r
an
ch
es
o
f
th
e
f
in
a
l
s
ta
g
e
d
u
r
in
g
th
e
s
ec
o
n
d
s
e
m
e
s
ter
o
f
th
e
2
0
1
8
y
ea
r
a
n
d
u
p
lo
ad
ed
to
[
19
]
w
ith
f
u
ll
d
escr
ip
tio
n
.
W
h
ile
th
e
s
ec
o
n
d
a
n
d
th
ir
d
d
atasets
(
s
tu
d
e
n
t
alco
h
o
l
co
n
s
u
m
p
t
io
n
d
ataset
)
,
ar
e
o
b
tain
ed
f
r
o
m
UC
I
P
o
r
tu
g
al
[
20
]
,
w
h
ic
h
in
co
r
p
o
r
ates
tw
o
d
atasets
:
s
tu
d
e
nt
-
m
a
t.c
s
v
an
d
s
tu
d
e
n
t
-
p
o
r
.
cs
v
.
Data
s
et
p
r
ep
r
o
ce
s
s
in
g
in
cl
u
d
e
s
th
e
f
o
llo
w
in
g
s
tep
s
:
-
Data
s
et
en
co
d
in
g
:
th
e
d
ataset
co
n
tain
s
attr
ib
u
te
s
o
f
v
ar
io
u
s
d
ata
ty
p
es,
f
o
r
in
s
ta
n
ce
:
b
in
ar
y
,
in
ter
v
al
,
n
u
m
er
ic
an
d
ca
teg
o
r
ical
(
n
o
m
in
al
,
o
r
d
in
al
)
.
T
h
e
KNN
r
eq
u
ir
es
d
ata
to
b
e
in
th
e
n
u
m
er
ical
f
o
r
m
u
latio
n
.
T
h
is
is
d
u
e
to
t
h
at
t
h
er
e
ar
e
m
an
y
f
ea
t
u
r
e
e
n
co
d
in
g
m
eth
o
d
s
f
o
r
tr
an
s
f
o
r
m
in
g
ca
te
g
o
r
ical
d
ata
to
n
u
m
er
ic
o
n
es,
s
u
ch
as
lab
el
e
n
co
d
in
g
o
r
in
te
g
er
en
co
d
in
g
,
o
n
e
-
h
o
t
en
c
o
d
in
g
,
b
in
ar
ized
a
n
d
h
a
s
h
i
n
g
.
I
n
t
h
is
r
e
s
ea
r
ch
,
th
e
d
atasets
ar
e
en
co
d
ed
u
s
in
g
L
ab
el
E
n
co
d
er
,
w
h
ic
h
is
t
h
e
m
o
s
t
co
m
m
o
n
m
et
h
o
d
to
tr
an
s
f
o
r
m
ca
teg
o
r
ical
f
ea
t
u
r
es i
n
to
n
u
m
er
ical
lab
els.
Nu
m
er
ical
lab
els ar
e
al
w
a
y
s
b
ein
g
b
et
w
ee
n
0
an
d
(
#
attr
ib
u
te_
v
alu
e
-
1
)
.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
E
n
h
a
n
ce
me
n
t o
f stu
d
en
t
p
erfo
r
ma
n
ce
p
r
ed
ictio
n
u
s
in
g
m
o
d
i
fied
K
-
n
ea
r
est n
eig
h
b
o
r
(
S
a
ja
Ta
h
a
A
h
med
)
1779
-
Data
s
et
n
o
r
m
aliza
tio
n
:
in
th
e
m
ac
h
in
e
lear
n
i
n
g
al
g
o
r
ith
m
wh
er
e
t
h
e
d
is
ta
n
ce
p
la
y
s
a
v
ital
r
o
le
lik
e
K
NN,
th
e
d
atasets
m
u
s
t
b
e
n
o
r
m
a
liz
ed
f
o
r
a
b
etter
p
r
ed
icto
r
(
i.e
.
a
v
o
id
m
is
cla
s
s
i
f
icatio
n
)
an
d
to
ef
f
icien
tl
y
tr
ai
n
th
e
al
g
o
r
ith
m
.
T
h
e
n
o
r
m
aliza
ti
o
n
is
th
e
p
r
o
ce
s
s
o
f
s
ca
li
n
g
att
r
ib
u
te
v
a
lu
e
s
w
it
h
i
n
a
s
p
ec
i
f
ic
r
an
g
e
(
s
u
c
h
a
s
0
to
1
)
,
in
a
m
an
n
er
t
h
at
all
at
tr
ib
u
tes
h
a
v
e
ap
p
r
o
x
i
m
atel
y
s
i
m
ilar
m
a
g
n
i
tu
d
es.
T
h
i
s
r
esear
ch
n
o
r
m
alize
s
th
e
attr
ib
u
te
v
al
u
es
u
s
i
n
g
Min
-
Ma
x
n
o
r
m
aliza
tio
n
at
t
h
e
r
an
g
e
[
-
1
,
1
]
.
-
Featu
r
e
s
elec
tio
n
:
th
e
r
es
u
lts
o
f
th
e
p
r
o
p
o
s
ed
f
ea
tu
r
e
s
elec
t
io
n
in
[
1
8
]
o
b
v
io
u
s
l
y
s
h
o
w
t
h
at
th
e
h
ig
h
est
p
er
f
o
r
m
a
n
ce
ac
cu
r
ac
y
i
s
ac
h
ie
v
ed
b
y
s
o
cial
f
ac
to
r
s
in
co
m
b
i
n
atio
n
w
it
h
m
ar
k
s
.
T
h
is
r
esear
ch
s
elec
ts
to
p
eig
h
t
f
ea
tu
r
e
s
s
u
b
s
et
b
ased
o
n
P
ea
r
s
o
n
co
r
r
elatio
n
r
an
k
in
g
cr
iter
ia.
Featu
r
e
s
u
b
s
et
o
f
I
r
aq
i
d
ataset
i
n
cl
u
d
es
th
e
f
o
llo
w
i
n
g
q
u
es
tio
n
s
:
“
Q3
7
W
o
r
r
y
E
f
f
ec
t”,
“
Q2
0
Fa
m
il
y
E
co
n
o
m
ic
L
e
v
el”,
“
Q2
5
R
e
aso
n
o
f
s
tu
d
y
”,
“
Q2
7
Fail
u
r
e
Yea
r
”,
“Q8
Fath
er
A
li
v
e”
,
“
Q1
7
Seco
n
d
ar
y
J
o
b
”,
“
Q3
3
Stu
d
y
Ho
u
r
”,
“
Q2
3
Sp
ec
ializatio
n
”
,
w
h
ile
U
C
I
.
s
t
u
d
en
t
-
p
o
r
.
csv
f
e
atu
r
e
s
u
b
s
et
i
n
cl
u
d
es:
“
Q1
0
r
ea
s
o
n
”
,
“
Q8
Mj
o
b
”,
“
Q2
1
in
ter
n
e
t”,
“Q3
ad
d
r
ess
”,
“
Q7
Fed
u
”,
“
Q6
Me
d
u
”,
“
Q1
3
s
tu
d
y
ti
m
e”
,
“
Q2
0
h
ig
h
er
”,
an
d
UC
I
.
s
t
u
d
en
t
-
m
at.
cs
v
f
ea
t
u
r
e
s
u
b
s
et
h
a
s
”Q1
7
p
aid
”,
“
Q8
Mj
o
b
”
,
“
Q1
s
e
x
”,
“
Q3
ad
d
r
ess
”,
“Q1
0
r
ea
s
o
n
”,
“
Q7
Fed
u
”,
“
Q2
0
h
ig
h
er
”
,
“
Q6
Me
d
u
f
ea
t
u
r
es”.
2
.
2
.
T
he
pro
po
s
ed
m
et
ho
d
I
n
th
i
s
s
tu
d
y
,
t
h
e
p
r
o
p
o
s
ed
FKNN
u
t
ilizes
t
h
e
co
n
ce
p
t
o
f
m
o
m
e
n
t
d
escr
ip
to
r
w
h
ich
is
a
s
et
o
f
p
ar
a
m
eter
s
t
h
at
d
escr
ib
e
t
h
e
d
is
tr
ib
u
tio
n
o
f
m
ater
ial
[
21
]
.
T
h
e
m
a
in
id
ea
i
s
t
h
e
s
i
m
ilar
it
y
b
et
w
ee
n
a
ttrib
u
te
s
v
alu
e
o
f
n
e
w
s
t
u
d
en
t (
test
s
a
m
p
le)
an
d
p
r
ev
io
u
s
l
y
r
e
g
is
ter
ed
s
tu
d
e
n
ts
(
tr
ai
n
ed
ex
a
m
p
le
s
)
,
s
i
n
ce
if
t
w
o
s
a
m
p
le
s
h
av
e
s
a
m
e
d
escr
ip
to
r
s
th
e
y
ar
e
g
o
in
g
to
h
av
e
ap
p
r
o
x
i
m
a
tel
y
s
i
m
ilar
p
er
f
o
r
m
a
n
ce
.
Fro
m
th
is
p
o
in
t
o
f
v
ie
w
,
th
is
r
e
s
ea
r
ch
co
m
e
s
o
u
t
w
i
th
t
h
e
co
n
tr
ib
u
tio
n
o
f
e
n
h
an
c
in
g
th
e
p
er
f
o
r
m
a
n
ce
o
f
K
NN
b
y
e
m
p
lo
y
i
n
g
m
o
m
en
t
d
escr
ip
to
r
to
p
r
e
-
class
i
f
y
s
t
u
d
en
ts
.
T
h
is
s
tr
ateg
y
u
s
e
s
th
e
d
escr
ip
to
r
s
as
a
r
ef
er
en
ce
i
n
d
ic
ato
r
to
p
r
e
-
class
if
y
th
e
tr
ain
in
g
s
a
m
p
les
i
n
to
g
r
o
u
p
s
w
it
h
a
s
p
ec
i
f
ic
d
escr
ip
to
r
v
alu
e
b
ased
o
n
s
o
cial
a
n
d
ac
ad
em
ic
f
ac
to
r
s
.
T
h
e
r
ea
s
o
n
f
o
r
ad
o
p
tin
g
th
is
c
lass
i
f
icatio
n
co
n
ce
p
t
is
t
h
at
th
e
d
escr
ip
to
r
o
f
ea
ch
s
t
u
d
en
t
r
e
p
r
esen
ts
a
s
i
g
n
atu
r
e
to
d
if
f
er
en
tiate
t
h
e
s
t
u
d
en
t b
eh
av
io
r
.
T
h
er
ef
o
r
e,
in
s
tead
o
f
m
ak
in
g
t
h
e
f
u
ll
s
ea
r
ch
d
u
r
i
n
g
d
is
tan
ce
co
m
p
u
tatio
n
w
it
h
t
h
e
w
h
o
le
tr
ain
i
n
g
s
et,
o
n
l
y
a
s
u
b
s
e
t
o
f
th
ese
s
a
m
p
les
is
co
m
p
u
ted
.
T
h
e
m
o
m
e
n
ts
ar
e
d
eter
m
in
ed
b
y
e
x
p
lo
itin
g
t
w
o
f
ir
s
t
-
o
r
d
er
m
o
m
e
n
t
s
D1
an
d
D2
,
as sh
o
wn
m
at
h
e
m
atica
l
l
y
i
n
(
1
)
:
1
,
2
=
∑
[
−
1
=
0
]
∗
[
,
]
(
1
)
w
h
er
e
v
is
t
h
e
len
g
th
o
f
t
h
e
f
ea
tu
r
e
s
u
b
s
e
t.
S
r
ep
r
esen
ts
attr
ib
u
te
j
o
f
s
am
p
le
i
in
a
d
ataset.
W
is
a
m
at
h
e
m
atica
l
r
ep
r
esen
tatio
n
c
h
o
s
en
f
o
r
a
b
etter
s
ep
ar
atio
n
co
n
tr
o
l.
T
h
e
ad
o
p
ted
w
eig
h
t
s
in
t
h
is
r
esear
c
h
a
r
e:
1
[
]
=
[
2j
−
1
]
(
2
)
2
[
]
=
{
2
−
1
≤
−
1
2
−
[
2
(
−
1
)
−
]
−
1
ℎ
(
3
)
w
h
er
e
j
=0
…
v
-
1
,
T
h
e
d
escr
ip
to
r
o
f
b
o
th
tr
ain
i
n
g
a
n
d
te
s
tin
g
s
a
m
p
le
i
s
d
eter
m
i
n
ed
u
s
in
g
t
h
e
f
o
llo
w
i
n
g
m
at
h
e
m
a
tical
(
4
)
:
=
1
2
−
2
2
1
2
+
2
2
(
4
)
T
h
e
p
r
o
p
o
s
ed
FKNN
n
ee
d
s
t
o
d
eter
m
in
e
t
h
e
in
d
ex
v
a
lu
e
f
o
r
ea
ch
s
a
m
p
le
(
i.e
.
s
tu
d
e
n
t)
,
th
e
d
eter
m
i
n
ed
d
escr
ip
to
r
(
Des)
is
co
n
v
er
ted
in
to
in
te
g
er
v
a
lu
e
w
i
th
i
n
r
an
g
e
[
0
,
No
_
s
u
b
]
,
w
h
er
e
No
_
s
u
b
is
th
e
n
u
m
b
er
o
f
tr
ain
i
n
g
s
u
b
s
et
s
,
th
e
d
escr
ip
to
r
in
d
ex
v
a
lu
e
f
o
r
ea
ch
s
a
m
p
le
(
Des_
I
n
d
ex
)
is
co
m
p
u
ted
u
s
in
g
th
e
(
5
)
:
Des_
I
n
d
ex
=
R
o
u
n
d
(
A
b
s
o
l
u
te(
Des)
*
No
_
s
u
b
)
(
5
)
I
n
ad
d
itio
n
,
th
e
p
r
o
p
o
s
ed
FKNN
n
ee
d
s
to
co
n
s
tr
u
ct
a
d
ata
s
tr
u
ctu
r
e
(
DS)
to
w
ar
r
an
t
y
f
ast
er
ac
ce
s
s
to
th
e
s
a
m
p
le
s
.
T
h
is
d
ata
s
tr
u
ctu
r
e
co
n
tain
s
th
e
id
en
t
if
ica
tio
n
n
u
m
b
er
an
d
d
escr
ip
to
r
in
d
ex
f
o
r
all
tr
ain
in
g
s
a
m
p
le
s
.
T
h
e
s
am
p
les
o
f
a
d
ata
s
tr
u
ctu
r
e
ar
e
ar
r
an
g
ed
in
ascen
d
i
n
g
o
r
d
er
ac
co
r
d
in
g
to
s
a
m
p
les
’
d
escr
ip
to
r
in
d
ex
.
T
h
er
ef
o
r
e,
all
s
a
m
p
les
t
h
at
h
a
v
e
th
e
s
a
m
e
d
escr
ip
to
r
w
ill
f
o
r
m
a
class
(
i.e
.
tr
ai
n
in
g
s
a
m
p
le
s
u
b
s
et)
in
co
n
t
ig
u
o
u
s
lo
ca
tio
n
s
.
T
h
e
p
r
e
-
class
if
icati
o
n
o
f
a
tr
ain
in
g
s
et
in
to
s
u
b
s
et
s
is
clar
if
ied
b
y
th
e
f
o
llo
w
i
n
g
p
s
eu
d
o
-
co
d
e
w
r
itte
n
as Alg
o
r
it
h
m
1
.
T
h
e
n
ex
t
s
tep
i
s
t
o
ca
lcu
late
t
h
e
f
r
eq
u
en
c
y
f
o
r
ea
ch
d
escr
ip
to
r
in
d
ex
i
n
th
e
s
o
r
ted
d
ata
s
tr
u
ctu
r
e
(
DS)
an
d
s
et
a
n
ar
r
a
y
o
f
p
o
in
ter
to
i
n
d
icate
t
h
e
s
tar
t
a
n
d
en
d
f
o
r
ea
ch
tr
ain
in
g
s
u
b
s
et.
I
n
s
u
c
h
a
w
a
y
,
t
h
e
li
m
itatio
n
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6930
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
,
Vo
l.
18
,
No
.
4
,
A
u
g
u
s
t 2
0
2
0
:
1
7
7
7
-
1
7
8
3
1780
o
f
ea
ch
s
u
b
s
et
ar
e
i
n
d
icate
d
b
y
p
o
in
ter
s
th
at
ac
t
as
le
ad
in
g
s
ig
n
s
to
r
ea
ch
t
h
e
i
n
te
n
tio
n
al
class
i
m
m
ed
iatel
y
.
T
h
e
s
tep
s
f
o
r
b
u
ild
i
n
g
an
a
r
r
ay
o
f
p
o
in
ter
s
ar
e
i
llu
s
tr
at
ed
in
t
h
e
f
o
llo
w
i
n
g
p
s
e
u
d
o
-
co
d
e,
p
r
esen
ted
as
A
l
g
o
r
ith
m
2
.
Algorithm 1. Training Set Pre
-
Classification
Input: Training sample as a matrix
[# students, # attributes]
Output: Sorted Data Structure contains samples classifying according to their
descriptor index field value.
Define DS as a data structure which is an array of records contains two elements
student descriptor index and his positi
oning (or, identifier) in training set.
Set the number of descriptor classes to No_sub.
For each i
ndex j of feature vector length
// Calculate weights in the range of
feature vector length.
Begin
Compute w1[j] using equation 2
Compute w2[j] using equation 3
End
For each student i in the Training set
Begin
For each attribute j in the feature vector
Compute D1 and D2 based on equation 1.
Compute Descriptor of student i (Des) by using equation 4
Compute Samp
le Descriptor Index (Des_Index) using equation 5
Set DS[i]. Index=Des_Index
Set DS[i]. Identifier=i
End
Sort elements of data structure (DS) according to descriptor field.
Return DS
Algorithm 2. Pointers
Input: Sorted data structure (DS) of samples’
descriptors and identifiers
Output: an array of pointers Pointer[#No_sub]
Define Freq [#No_sub] as an array of integer hold the occurrences of descriptor
index (Des_Index) in DS.
For each student i in training dataset
Begin
Set X=DS[i]. Index
Increment Freq[X] by one
End
Set Pointer [0] =0
For each value n in No_sub
Set Pointer[n]=Pointer[n
-
1] + Freq[n
-
1]
Return Pointer
Af
ter
co
m
p
leti
n
g
th
e
tas
k
o
f
s
o
r
tin
g
tr
ain
i
n
g
s
a
m
p
le
s
,
th
e
m
atch
i
n
g
p
r
o
ce
s
s
tak
es
p
lace
b
y
ap
p
ly
in
g
KNN.
W
h
en
s
a
m
p
les
o
f
tr
ai
n
in
g
s
u
b
s
et
ar
e
ar
r
an
g
ed
at
co
n
tig
u
o
u
s
lo
ca
tio
n
s
s
in
ce
t
h
e
y
s
h
ar
ed
a
s
i
m
ilar
d
escr
ip
to
r
in
d
ex
,
as
a
r
esu
lt,
ea
ch
test
s
a
m
p
le
is
o
n
l
y
co
m
p
a
r
ed
w
i
th
t
h
e
s
p
ec
i
f
ic
tr
ain
i
n
g
s
u
b
s
et
b
ased
o
n
it
s
d
escr
ip
to
r
in
d
ex
.
A
b
s
o
lu
tel
y
,
th
is
tr
ain
i
n
g
s
u
b
s
et
h
as
f
e
w
er
s
a
m
p
les
th
a
n
t
h
o
s
e
f
o
u
n
d
w
it
h
in
th
e
f
u
ll
tr
ai
n
i
n
g
d
ataset.
I
n
ad
d
itio
n
,
t
h
e
b
est
s
i
m
ilar
s
a
m
p
le
s
(
in
ter
m
s
o
f
t
h
eir
attr
ib
u
tes)
ar
e
m
o
s
t
p
r
o
b
ab
l
y
a
v
ailab
le
i
n
t
h
is
tr
ain
i
n
g
s
u
b
s
e
t
th
a
t
h
a
s
s
i
m
ila
r
d
escr
ip
to
r
in
d
ex
.
T
h
is
led
to
a
s
u
b
s
ta
n
tia
l
r
ed
u
ctio
n
i
n
r
u
n
n
in
g
t
i
m
e
o
f
KN
N
an
d
i
m
p
r
o
v
es
t
h
e
ac
cu
r
ac
y
o
f
class
i
f
icatio
n
.
T
h
e
s
i
m
ilar
it
y
m
ea
s
u
r
e
m
e
n
t
i
s
b
ased
o
n
th
e
E
u
clid
ea
n
d
is
ta
n
ce
b
et
w
ee
n
t
h
e
te
s
t
s
a
m
p
le
an
d
s
a
m
p
les
o
f
th
e
tr
ain
in
g
s
u
b
s
e
t.
T
h
e
ca
lcu
lated
d
i
s
tan
ce
s
ar
e
s
to
r
ed
in
a
s
o
r
ted
ascen
d
i
n
g
o
r
d
er
ar
r
ay
.
I
f
t
h
e
d
is
tan
ce
h
as
ze
r
o
v
al
u
e,
th
e
la
b
el
o
f
th
e
co
r
r
esp
o
n
d
in
g
s
a
m
p
le
is
co
n
s
id
er
ed
as
tar
g
et
clas
s
d
ir
ec
tl
y
,
o
th
er
w
i
s
e,
th
e
k
tr
ai
n
in
g
s
a
m
p
le
i
s
p
ick
ed
o
u
t
an
d
t
h
e
tar
g
et
cla
s
s
o
f
th
e
n
e
w
s
a
m
p
le
i
s
d
eter
m
in
ed
b
y
th
e
u
s
e
o
f
th
e
m
aj
o
r
it
y
v
o
ti
n
g
co
n
ce
p
t.
T
h
e
f
o
llo
w
i
n
g
p
s
eu
d
o
-
co
d
e
(
alg
o
r
ith
m
3
)
ex
p
lain
s
th
e
s
tep
s
i
n
v
o
lv
ed
in
ap
p
l
y
in
g
th
e
p
r
o
p
o
s
ed
(
FKNN)
f
o
r
test
s
a
m
p
les
:
Algorithm 3. The Proposed FKNN
Input: test set as matrix [#students, # attributes]
Output: target class for a test samples
For each student
t in the test set
Begin
//Define DS2 to contain distance and training sample identifier.
Calculate descriptor of student t using equation 4
Calculate descriptor index of student t to get Des
-
test using equation 5
//Determine the start and
end index for training subset which has the same
descriptor index as student t
using an array of the pointer.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
E
n
h
a
n
ce
me
n
t o
f stu
d
en
t
p
erfo
r
ma
n
ce
p
r
ed
ictio
n
u
s
in
g
m
o
d
i
fied
K
-
n
ea
r
est n
eig
h
b
o
r
(
S
a
ja
Ta
h
a
A
h
med
)
1781
Set Start=Pointer[Des
-
test]
Set End=Freq[Des
-
test] +Pointer[Des
-
test]
Set x=0
For each training sample i in the range from Star
t to End
Begin
//Calculate Distance between new student and train students which have the
same
descriptor in the sorted data structure DS
Set ID=DS[i]. Identifier
For each attribute j in feature vector //calculate Eucl
idean
distance
[
]
=
√
∑
[
,
]
−
[
,
]
−
=
Set DS2[x]. Distance=Distance[x].
Set DS2[x]. ID=ID.
Increment x by one
End
Sort DS2 according to the distance in ascendi
ng order
If DS2[0]. Distance is equal to zero,
then the target class is the label of the sample that has DS2[0].
ID.
Else Begin
//Matching via Majority Vote
Pick the first K entries from Distance
Get the labels of the selected K entries
End
End
Return the target class of the majority K labels
3.
RE
SU
L
T
S
AND
AN
AL
Y
SI
S
T
h
e
ex
p
er
i
m
en
ts
a
n
d
th
e
ap
p
licatio
n
s
y
s
te
m
ar
e
p
er
f
o
r
m
ed
b
ased
o
n
v
is
u
al
s
tu
d
io
.
n
e
t
C
#
2
0
1
5
.
T
h
e
ev
alu
atio
n
o
f
t
h
e
p
r
o
p
o
s
e
d
m
et
h
o
d
is
p
er
f
o
r
m
ed
u
s
in
g
h
o
ld
o
u
t
v
alid
atio
n
,
w
h
ic
h
s
p
li
ts
d
atasets
in
to
t
w
o
s
ets:
7
0
% tr
ain
in
g
an
d
3
0
% te
s
t.
A
cc
u
r
ac
y
(
AC
C
)
is
co
n
s
id
er
ed
to
m
ea
s
u
r
es t
h
e
d
eg
r
ee
to
w
h
ic
h
th
e
i
n
s
tan
ce
s
co
r
r
ec
tly
clas
s
i
f
ied
b
y
t
h
e
m
ac
h
in
e
lear
n
in
g
alg
o
r
it
h
m
i
n
p
r
o
p
o
r
tio
n
to
th
e
en
tire
test
ed
in
s
tan
ce
s
[
2
2
]
.
As
m
en
t
io
n
ed
ea
r
lier
,
th
e
m
ai
n
ai
m
o
f
th
i
s
w
o
r
k
is
t
h
e
p
r
ed
ictio
n
o
f
s
t
u
d
en
t
p
er
f
o
r
m
an
ce
.
Fo
r
th
is
p
u
r
p
o
s
e,
th
e
tar
g
et
c
lass
lab
el
i
s
f
o
r
m
u
lated
f
o
r
ea
ch
d
atase
t,
w
h
ic
h
ca
n
b
e
eit
h
er
“
P
as
s
”
o
r
“
Fa
il”.
T
h
er
e
ar
e
th
r
ee
av
er
ag
es
o
f
G1
,
G2
,
an
d
G3
in
t
h
e
U
C
I
d
ataset
w
it
h
v
al
u
es
f
r
o
m
0
to
2
0
.
T
h
er
ef
o
r
e,
if
t
h
e
s
t
u
d
en
t
h
as
a
g
r
ad
e
e
q
u
al
to
o
r
g
r
ea
ter
th
an
1
0
,
it
s
h
o
u
ld
b
e
class
if
ied
u
n
d
er
th
e
“
P
ass
”
lab
el,
o
th
er
w
i
s
e,
it
s
h
o
u
ld
b
e
class
i
f
ied
as
a
“
Fa
il”
lab
el.
I
n
I
r
aq
i
d
ataset,
g
r
ad
e
v
al
u
es
ar
e
w
it
h
i
n
r
a
n
g
e
o
f
(
0
-
1
0
0
)
.
I
f
th
e
s
tu
d
e
n
t h
as
a
g
r
ad
e
eq
u
al
o
r
h
ig
h
er
th
a
n
5
0
,
it sh
o
u
ld
b
e
d
ef
in
ed
w
it
h
i
n
th
e
“
P
as
s
”
lab
el,
o
th
er
w
i
s
e
is
cla
s
s
i
f
ie
d
as “
Fail”
s
tu
d
e
n
t.
T
h
e
s
tu
d
en
t
s
’
p
er
f
o
r
m
an
ce
o
f
th
e
UC
I
d
ataset
s
is
p
r
ed
icted
b
as
ed
o
n
f
in
al
s
e
m
e
s
ter
g
r
ad
es
(
G3
)
as
th
e
o
b
j
ec
tiv
e
class
.
T
h
e
I
r
a
q
i
d
ataset
p
r
ed
ictio
n
o
f
th
e
tar
g
et
class
is
d
o
n
e
u
s
in
g
th
e
s
ec
o
n
d
-
s
e
m
ester
av
er
ag
e
(
Av
g
2
)
.
I
n
th
i
s
w
o
r
k
;
f
o
r
th
e
p
u
r
p
o
s
e
o
f
co
m
p
ar
in
g
r
esu
lts
a
m
o
n
g
d
ataset
s
,
ce
r
tain
p
ar
a
m
eter
s
m
u
s
t
b
e
estab
lis
h
ed
s
u
c
h
as
t
h
e
n
u
m
b
er
o
f
d
escr
ip
to
r
class
es
(
i.e
.
a
n
u
m
b
er
o
f
b
in
s
)
w
h
ic
h
s
et
to
a
v
alu
e
o
f
f
i
v
e
an
d
th
e
v
a
lu
e
o
f
K
co
n
s
id
er
ed
to
b
e
th
r
ee
n
ea
r
est n
ei
g
h
b
o
r
s
.
I
n
th
e
p
er
s
p
ec
ti
v
e
o
f
tr
ad
itio
n
a
l
KNN
is
s
u
es,
t
h
e
p
r
o
p
o
s
ed
FKNN
h
a
s
p
r
o
v
ed
th
at
it
r
u
n
s
f
a
s
te
r
f
o
r
all
test
s
a
m
p
les
t
h
an
tr
ad
itio
n
al
K
NN
s
in
ce
FKNN
r
eq
u
ir
es
a
s
m
aller
n
u
m
b
er
o
f
co
m
p
ar
is
o
n
s
b
ased
o
n
th
e
d
is
ta
n
ce
ca
lcu
latio
n
o
f
ea
c
h
n
e
w
s
a
m
p
l
e
in
f
o
r
m
atio
n
f
r
o
m
a
s
u
b
s
et
o
f
tr
ain
i
n
g
d
ata
co
n
ta
in
i
n
g
t
h
e
s
a
m
e
d
escr
ip
to
r
in
d
ex
as
th
e
n
e
w
s
a
m
p
le.
T
h
is
ca
n
also
r
ed
u
ce
m
e
m
o
r
y
r
eq
u
ir
e
m
e
n
ts
s
i
g
n
i
f
ican
tl
y
.
I
n
co
n
tr
ast
t
o
tr
ad
itio
n
al
KNN,
it is
b
ei
n
g
s
lo
w
b
ec
a
u
s
e
o
f
t
h
e
d
ep
en
d
en
c
y
o
n
t
h
e
e
x
h
a
u
s
ti
v
e
s
ea
r
ch
o
f
ea
c
h
n
e
w
s
a
m
p
le
w
it
h
a
ll tr
ai
n
i
n
g
d
ata
an
d
r
eq
u
ir
es
m
o
r
e
m
e
m
o
r
y
c
ap
ac
it
y
to
s
to
r
e
d
is
ta
n
ce
s
o
f
w
h
o
le
tr
a
i
n
in
g
s
a
m
p
le
s
.
Fi
g
u
r
e
1
in
d
icate
s
t
h
at
th
e
r
u
n
n
in
g
ti
m
e
o
f
t
h
e
p
r
o
p
o
s
ed
m
et
h
o
d
is
i
m
p
r
o
v
ed
co
m
p
a
r
ed
to
th
e
tr
ad
itio
n
al
KNN
ti
m
e.
C
o
m
p
ar
is
o
n
a
co
m
m
o
n
w
a
y
t
o
m
ea
s
u
r
e
t
h
e
p
r
o
ce
s
s
in
g
ef
f
e
ct
is
to
co
m
p
ar
e
th
e
o
u
tco
m
e
o
f
in
ter
est
b
ef
o
r
e
p
r
o
ce
s
s
in
g
w
it
h
th
at
a
f
t
er
p
r
o
ce
s
s
in
g
.
T
h
e
p
er
ce
n
tag
e
ch
an
g
e
m
ea
s
u
r
es
an
ite
m
’
s
ch
a
n
g
e
i
n
v
al
u
e
r
elati
v
e
to
its
o
r
ig
in
al
v
al
u
e.
Su
p
p
o
s
e
x
is
th
e
b
aselin
e
v
al
u
e,
y
is
th
e
p
o
s
t
-
p
r
o
ce
s
s
i
n
g
v
alu
e.
T
h
e
P
er
c
en
tag
e
c
h
an
g
e
ca
n
b
e
ca
lcu
lated
u
s
i
n
g
(
6
)
[
2
3
]
:
P
C
=(
(
X
-
Y)
/X)
*
1
0
0
(
6
)
T
ab
le
1
s
u
m
m
ar
ize
s
t
h
e
p
er
ce
n
tag
e
ch
a
n
g
e
o
f
r
u
n
n
i
n
g
ti
m
e
b
ased
o
n
th
e
r
es
u
lt
s
s
h
o
w
n
i
n
Fig
u
r
e
1
.
I
t
ca
n
b
e
s
ee
n
th
a
t
th
e
p
r
o
p
o
s
ed
FKNN
r
ed
u
ce
s
th
e
t
i
m
e
c
o
m
p
le
x
it
y
o
f
t
h
e
tr
ad
itio
n
al
KNN
b
y
(
9
0
.
2
5
%),
(
8
7
.
5
3
%),
an
d
(
7
5
.
4
%)
f
o
r
Po
r
,
Ma
th
,
an
d
I
r
aq
,
r
esp
ec
tiv
el
y
.
T
h
e
p
er
f
o
r
m
a
n
ce
o
f
t
h
e
p
r
o
p
o
s
ed
FKNN
ac
h
ie
v
es
b
etter
clas
s
if
ica
tio
n
ac
c
u
r
ac
y
t
h
a
n
tr
ad
iti
o
n
al
KNN
f
o
r
all
d
atase
ts
.
T
h
is
is
d
u
e
to
th
at
th
e
p
r
o
p
o
s
ed
FKNN
r
eli
es
o
n
th
e
w
ei
g
h
ted
m
o
m
en
t
d
escr
ip
to
r
s
a
m
p
les
to
co
n
s
tr
u
ct
tr
ain
in
g
s
u
b
s
e
ts
t
h
at
h
a
v
e
h
ig
h
er
cla
s
s
d
i
s
cr
i
m
i
n
at
o
r
y
in
f
o
r
m
atio
n
.
T
h
is
ca
n
lea
d
to
g
etti
n
g
t
h
e
b
es
t
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6930
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
,
Vo
l.
18
,
No
.
4
,
A
u
g
u
s
t 2
0
2
0
:
1
7
7
7
-
1
7
8
3
1782
m
atc
h
in
g
d
i
s
tan
ce
a
m
o
n
g
n
e
w
an
d
tr
ain
in
g
s
a
m
p
le
s
.
Via
t
h
i
s
s
elec
ti
v
e
s
ch
e
m
e,
t
h
e
m
i
s
cla
s
s
i
f
icatio
n
p
r
o
b
le
m
o
f
tr
ad
itio
n
al
KNN
ca
n
b
e
s
ig
n
if
ica
n
tl
y
o
v
er
co
m
e.
R
e
f
er
r
in
g
t
o
T
ab
le
2
th
e
p
r
o
p
o
s
ed
m
et
h
o
d
o
b
tain
s
th
e
h
ig
h
e
s
t
ac
cu
r
ac
y
o
f
1
0
0
%
f
o
r
I
r
aq
i
s
t
u
d
en
t
p
er
f
o
r
m
a
n
ce
d
ataset.
I
n
ad
d
itio
n
,
it
ca
n
b
e
s
ee
n
th
at
th
e
p
r
o
p
o
s
ed
m
et
h
o
d
is
ab
le
to
en
h
an
ce
cla
s
s
i
f
icat
io
n
ac
cu
r
ac
y
f
o
r
f
i
n
al
s
e
m
ester
g
r
ad
e
p
r
ed
ictio
n
(
G3
)
b
y
o
b
tain
in
g
t
h
e
p
er
ce
n
tag
e
ch
an
g
e
i
n
a
cc
u
r
ac
y
o
f
(
3
6
.
3
%)
,
(
2
3
.
7
%),
an
d
(
2
0
%),
f
o
r
Po
r
,
Ma
th
an
d
I
r
aq
d
atasets
,
r
esp
ec
tiv
el
y
.
T
ab
le
3
s
h
o
w
s
a
co
m
p
ar
is
o
n
o
f
th
e
p
r
o
p
o
s
ed
FKNN
w
it
h
th
e
r
esear
ch
w
o
r
k
o
f
[
2
4
]
.
T
h
is
r
esear
ch
u
s
e
s
P
o
r
d
ataset
f
r
o
m
UC
I
to
p
r
e
d
ict
s
tu
d
en
t
p
er
f
o
r
m
an
ce
b
ased
o
n
eig
h
t
f
e
atu
r
e
s
G2
,
G1
,
f
ailu
r
es,
h
i
g
h
er
,
Me
d
u
,
s
ch
o
o
l,
s
t
u
d
y
t
i
m
e,
Fe
d
u
.
I
n
ad
d
itio
n
,
a
co
m
p
ar
is
o
n
o
f
th
e
p
r
o
p
o
s
ed
FKNN
w
it
h
th
e
r
esear
ch
w
o
r
k
o
f
[
2
5
]
.
T
h
is
r
esear
ch
u
s
e
s
Ma
th
d
ataset
f
r
o
m
UC
I
to
p
r
ed
ict
s
tu
d
en
t
p
er
f
o
r
m
a
n
ce
b
ased
o
n
1
9
f
ea
tu
r
es
in
cl
u
d
in
g
t
h
e
clas
s
attr
ib
u
te:
s
ex
,
f
a
m
s
ize,
ad
d
r
ess
,
p
s
tat
u
s
,
m
ed
u
,
f
ed
u
,
m
j
o
b
,
f
j
o
b
,
tr
a
v
elti
m
e,
s
t
u
d
y
ti
m
e,
s
ch
o
o
ls
u
p
,
h
ig
h
er
,
in
ter
n
et,
r
o
m
an
tic,
f
r
ee
ti
m
e,
Dalc
,
W
alc,
h
ea
lt
h
,
s
u
cc
e
s
s
.
I
t
is
clea
r
th
at
th
e
p
r
o
p
o
s
ed
DD
T
s
u
r
p
ass
al
l
m
et
h
o
d
s
u
t
ilized
in
th
ese
r
esear
ch
e
s
f
o
r
t
w
o
UC
I
(
P
o
r
an
d
Ma
th
)
d
atasets
.
Fig
u
r
e
1
.
T
h
e
co
m
p
ar
is
o
n
o
f
t
h
e
p
r
o
p
o
s
ed
FKNN
an
d
tr
ad
itio
n
al
KNN
r
u
n
n
in
g
ti
m
e
f
o
r
f
i
n
al
g
r
ad
e
p
r
ed
ictio
n
T
ab
le
1
.
P
er
ce
n
tag
e
ch
a
n
g
e
i
n
r
u
n
n
i
n
g
ti
m
e
o
f
p
r
o
p
o
s
ed
FKNN
D
a
t
a
se
t
P
C
T
i
me
P
o
r
9
0
.
2
5
%
M
a
t
h
8
7
.
5
3
%
I
r
a
q
7
5
.
4
%
T
ab
le
2
.
T
h
e
ac
cu
r
ac
y
i
m
p
r
o
v
e
m
en
t
o
f
t
h
e
p
r
o
p
o
s
ed
m
eth
o
d
v
s
.
tr
ad
itio
n
al
KN
N
f
o
r
f
i
n
al
g
r
ad
e
p
r
ed
ictio
n
D
a
t
a
se
t
s
T
r
a
d
i
t
i
o
n
a
l
K
N
N
P
r
o
p
o
se
d
F
K
N
N
A
c
c
u
r
a
c
y
D
i
f
f
e
r
e
n
c
e
s
P
C
i
n
A
c
c
u
r
a
c
y
P
o
r
6
9
.
2
9
4
.
3
2
5
.
1
3
6
.
3
%
Ma
th
7
8
.
1
9
6
.
6
1
8
.
5
2
3
.
7
%
I
r
aq
8
3
.
3
1
0
0
1
6
.
7
2
0
%
T
ab
le
3
.
A
cc
u
r
ac
y
co
m
p
ar
is
o
n
o
f
o
u
r
p
r
o
p
o
s
ed
DDT
an
d
o
th
er
m
e
th
o
d
s
f
o
r
UC
I
d
atasets
D
a
t
a
se
t
R
e
se
a
r
c
h
W
o
r
k
M
e
t
h
o
d
A
c
c
u
r
a
c
y
P
o
r
[
2
4
]
(
2
0
1
9
)
N
a
ï
v
e
B
a
y
e
s
D
e
c
i
si
o
n
T
r
e
e
R
a
n
d
o
mT
r
e
e
R
EPT
r
e
e
JR
i
p
O
n
e
R
S
i
mp
l
e
L
o
g
i
st
i
c
Z
e
r
o
R
7
3
.
1
8
%
7
6
.
2
7
%
6
7
.
9
5
%
7
6
.
7
3
%
7
4
.
1
1
%
7
6
.
7
3
%
7
3
.
6
5
%
3
0
.
9
7
%
O
u
r
P
r
o
p
o
se
d
M
o
d
e
l
T
h
e
P
r
o
p
o
se
d
F
K
N
N
9
4
.
3
%
M
a
t
h
[
2
5
]
(
2
0
1
6
)
P
C
F
w
i
t
h
k
-
me
d
o
i
d
s
a
l
g
o
r
i
t
h
m
P
C
F
w
i
t
h
k
-
me
a
n
s
a
l
g
o
r
i
t
h
m
6
5
.
8
2
%
6
3
.
5
0
%
O
u
r
P
r
o
p
o
se
d
M
o
d
e
l
T
h
e
P
r
o
p
o
se
d
F
K
N
N
9
6
.
6
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
E
n
h
a
n
ce
me
n
t o
f stu
d
en
t
p
erfo
r
ma
n
ce
p
r
ed
ictio
n
u
s
in
g
m
o
d
i
fied
K
-
n
ea
r
est n
eig
h
b
o
r
(
S
a
ja
Ta
h
a
A
h
med
)
1783
4.
CO
NCLU
SI
O
NS
T
h
is
s
tu
d
y
p
r
ese
n
ted
th
e
FK
NN
alg
o
r
ith
m
w
h
ic
h
co
m
b
in
es
th
e
s
a
m
p
le
in
d
e
x
in
g
m
ec
h
an
is
m
w
it
h
KNN
to
d
ea
l
w
ith
t
h
e
m
aj
o
r
p
r
o
b
lem
s
o
f
th
e
tr
ad
itio
n
al
KNN.
T
h
e
co
m
p
u
tatio
n
al
o
v
er
h
ea
d
,
m
e
m
o
r
y
r
eq
u
ir
e
m
en
t,
m
u
l
tid
i
m
e
n
s
io
n
a
lit
y
(
th
e
n
u
m
b
er
o
f
s
a
m
p
le
s
)
an
d
m
is
cla
s
s
i
f
icatio
n
p
r
o
b
lem
s
w
er
e
s
u
b
s
ta
n
tiall
y
r
ed
u
ce
d
d
u
e
to
th
e
p
r
e
-
class
i
f
icatio
n
o
f
tr
ain
i
n
g
d
ata
b
ased
o
n
th
e
d
escr
ip
to
r
s
u
s
ed
b
y
t
h
e
s
elec
ti
v
e
s
ea
r
ch
s
tr
ateg
y
.
T
h
e
class
i
f
icatio
n
ac
c
u
r
ac
y
w
as
e
n
h
an
ce
d
u
s
i
n
g
t
h
e
p
r
o
p
o
s
ed
FKNN
m
et
h
o
d
s
in
ce
th
e
tr
ain
i
n
g
s
a
m
p
le
g
r
o
u
p
ed
ac
co
r
d
in
g
to
th
eir
s
i
m
ilar
it
y
.
T
h
e
r
esu
lts
s
h
o
w
ed
a
s
i
g
n
i
f
ica
n
t
e
n
h
an
ce
m
e
n
t
in
ac
cu
r
ac
y
w
i
th
th
e
h
i
g
h
est
in
cr
ea
s
e
r
ea
ch
ed
to
(
3
6
.
3
%)
an
d
im
p
r
o
v
ed
th
e
co
m
p
u
tatio
n
ti
m
e
o
f
KNN
w
i
th
th
e
h
ig
h
est
ti
m
e
r
ed
u
ctio
n
r
ea
ch
to
9
0
.
2
5
%
f
o
r
UC
I
.
s
t
u
d
en
t
-
p
r
o
.
csv
d
ata
s
et.
T
h
e
ad
o
p
ted
ex
p
er
im
e
n
ts
co
n
f
ir
m
ed
th
a
t
th
e
p
r
o
p
o
s
ed
FKNN
o
u
tp
er
f
o
r
m
ed
th
e
tr
ad
itio
n
al
KNN
f
o
r
all
ed
u
ca
tio
n
al
d
ata
s
ets.
T
h
er
ef
o
r
e,
th
e
p
r
o
p
o
s
ed
alg
o
r
ith
m
w
as
v
er
y
u
s
ef
u
l
f
o
r
th
e
r
ea
l
-
ti
m
e
s
y
s
te
m
s
u
c
h
a
s
e
lear
n
in
g
en
v
ir
o
n
m
en
t
an
d
co
u
ld
b
e
u
s
ed
f
o
r
lar
g
er
d
atasets
.
RE
F
E
R
E
NC
E
S
[1
]
A
n
d
e
rss
o
n
A
.
,
“
S
e
v
e
n
m
a
jo
r
c
h
a
ll
e
n
g
e
s
f
o
r
e
-
lea
rn
in
g
in
d
e
v
e
lo
p
i
n
g
c
o
u
n
t
ries
:
Ca
se
stu
d
y
e
B
IT
,
S
ri
L
a
n
k
a
,”
In
ter
n
a
t
io
n
a
l
J
o
u
rn
a
l
o
f
E
d
u
c
a
ti
o
n
a
n
d
De
v
e
lo
p
me
n
t
u
sin
g
ICT
,
v
o
l.
4
,
n
o
.
3
,
p
p
.
45
-
62
,
2
0
0
8
.
[2
]
Ha
sib
u
a
n
Z
.
A.
,
“
S
tep
-
F
u
n
c
ti
o
n
A
p
p
ro
a
c
h
f
o
r
E
-
L
e
a
rn
in
g
P
e
rs
o
n
a
li
z
a
ti
o
n
,”
T
EL
KOM
NIKA
T
e
l
e
c
o
mm
u
n
ica
ti
o
n
,
Co
mp
u
t
in
g
,
El
e
c
tro
n
ics
a
n
d
Co
n
t
ro
l,
v
o
l.
1
5
,
n
o
.
3
,
p
p
.
1
3
6
2
-
1
3
6
7
,
2
0
1
7
.
[3
]
Do
g
ru
e
r
N
.
,
Ey
y
a
m
R
.
,
M
e
n
e
v
is
I.
,
“
T
h
e
u
se
o
f
t
h
e
i
n
tern
e
t
f
o
r
e
d
u
c
a
ti
o
n
a
l
p
u
rp
o
se
s
,”
Pro
c
e
d
i
a
-
S
o
c
ia
l
a
n
d
Beh
a
v
io
r
a
l
S
c
ien
c
e
s
,
v
o
l.
2
8
,
p
p
.
6
0
6
-
6
1
1
,
2
0
1
1
.
[4
]
Ak
ib
u
M
.
A
.
,
M
o
k
h
a
iri
M
.
,
S
u
h
a
il
a
n
S
.
,
“
T
h
e
p
a
tt
e
rn
s
o
f
a
c
c
e
ss
in
g
lea
rn
in
g
m
a
n
a
g
e
m
e
n
t
s
y
st
e
m
a
m
o
n
g
stu
d
e
n
ts
,
”
In
d
o
n
e
sia
n
J
o
u
rn
a
l
o
f
El
e
c
trica
l
En
g
i
n
e
e
rin
g
a
n
d
C
o
mp
u
ter
S
c
ien
c
e
,
v
o
l.
13
,
n
o
.
1
,
p
p
.
15
-
21
,
2
0
1
9
.
[5
]
M
in
i
n
g
T
.
E.
,
“
En
h
a
n
c
in
g
tea
c
h
in
g
a
n
d
lea
rn
i
n
g
th
r
o
u
g
h
e
d
u
c
a
ti
o
n
a
l
d
a
ta
m
in
in
g
a
n
d
lea
rn
in
g
a
n
a
ly
ti
c
s:
A
n
issu
e
b
rief
,
”
Pro
c
e
e
d
in
g
s o
f
c
o
n
fer
e
n
c
e
o
n
a
d
v
a
n
c
e
d
tec
h
n
o
lo
g
y
fo
r e
d
u
c
a
ti
o
n
,
2
0
1
2
.
[6
]
Ag
ra
w
a
l
R.
,
“
K
-
n
e
a
re
st
n
e
ig
h
b
o
r
f
o
r
u
n
c
e
rtain
d
a
ta
,”
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
C
o
mp
u
ter
Ap
p
li
c
a
ti
o
n
s
,
v
o
l
.
1
0
5
,
n
o
.
1
1
,
p
p
.
1
3
3
-
1
6
,
2
0
1
4
.
[7
]
W
isit
L
.
,
S
a
k
o
l
U.,
“
Im
a
g
e
c
las
si
f
i
c
a
ti
o
n
o
f
m
a
laria
u
sin
g
h
y
b
rid
a
lg
o
rit
h
m
s:
c
o
n
v
o
l
u
ti
o
n
a
l
n
e
u
ra
l
n
e
t
w
o
rk
a
n
d
m
e
th
o
d
to
f
in
d
a
p
p
ro
p
riate
K
f
o
r
K
-
Ne
a
re
st
n
e
ig
h
b
o
r
,
”
In
d
o
n
e
sia
n
J
o
u
rn
a
l
o
f
E
lec
trica
l
E
n
g
in
e
e
rin
g
a
n
d
C
o
mp
u
ter
S
c
ien
c
e
,
v
o
l.
16
,
n
o
.
1
,
p
p
.
3
8
2
-
3
8
8
,
2
0
1
9
.
[8
]
G
a
r
c
ia E
.
K
.
,
F
e
l
d
m
a
n
S
.
,
G
u
p
ta
M
.
R
.
,
S
riv
a
sta
v
a
S
.
,
“
Co
m
p
lete
ly
laz
y
lea
rn
in
g
,”
IEE
E
T
ra
n
sa
c
ti
o
n
s o
n
K
n
o
wled
g
e
a
n
d
Da
t
a
E
n
g
i
n
e
e
rin
g
,
v
o
l.
22
,
n
o
.
9
,
p
p
.
1
2
7
4
-
85
,
2
0
0
9
.
[9
]
Ha
ll
M
.
A.
,
“
Co
rre
latio
n
-
b
a
se
d
f
e
a
tu
re
se
lec
ti
o
n
f
o
r
m
a
c
h
in
e
lea
rn
in
g
,
”
P
h
D T
h
e
sis,
1
9
9
9
.
[1
0
]
A
li
z
a
d
e
h
H
.
,
M
in
a
e
i
-
Bi
d
g
o
li
B
.
,
Am
ir
g
h
o
li
p
o
u
r
S
.
K.
,
“
A
n
e
w
m
e
th
o
d
f
o
r
im
p
ro
v
in
g
th
e
p
e
rf
o
rm
a
n
c
e
o
f
k
n
e
a
re
st
n
e
ig
h
b
o
r
u
si
n
g
c
lu
ste
rin
g
tec
h
n
i
q
u
e
,”
J
o
u
rn
a
l
o
f
C
o
n
v
e
rg
e
n
c
e
In
f
o
rm
a
ti
o
n
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
4
,
n
o
.
2
,
p
p
.
84
-
92
,
2
0
0
9
.
[1
1
]
S
u
g
u
n
a
N
.
,
T
h
a
n
u
sh
k
o
d
i
K.
,
“
A
n
im
p
ro
v
e
d
k
-
n
e
a
re
st
n
e
ig
h
b
o
r
c
las
sif
ic
a
ti
o
n
u
sin
g
g
e
n
e
ti
c
a
lg
o
rit
h
m
,”
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
Co
m
p
u
ter
S
c
ien
c
e
Iss
u
e
s
,
v
o
l.
7
,
n
o
.
2
,
p
p
.
18
-
21
,
2
0
0
9
.
[1
2
]
Ch
e
n
S
.
,
“
K
-
n
e
a
re
st
n
e
ig
h
b
o
r
a
l
g
o
rit
h
m
o
p
ti
m
iza
ti
o
n
in
tex
t
c
a
teg
o
riza
ti
o
n
,”
IOP
Co
n
fer
e
n
c
e
S
e
rie
s:
Ea
rth
a
n
d
En
v
iro
n
me
n
ta
l
S
c
ien
c
e
,
2
0
1
8
.
[1
3
]
S
y
a
li
m
a
n
K
.
U
.
,
Na
b
a
b
a
n
E
.
B
.
,
S
it
o
m
p
u
l
O
.
S.
,
“
Im
p
ro
v
in
g
th
e
a
c
c
u
ra
c
y
o
f
k
-
n
e
a
re
st
n
e
ig
h
b
o
r
u
sin
g
lo
c
a
l
m
e
a
n
b
a
se
d
a
n
d
d
istan
c
e
w
e
ig
h
t
,
”
J
o
u
r
n
a
l
o
f
P
h
y
sic
s: Co
n
fer
e
n
c
e
S
e
rie
s
,
v
o
l.
9
7
8
,
n
o
.
1
,
p
p
.
28
-
3
0
,
2
0
1
8
.
[1
4
]
Ha
ss
a
n
a
t
A
.
B
.
,
A
b
b
a
d
i
M
.
A
.
,
Altara
w
n
e
h
G
.
A
.
,
A
lh
a
sa
n
a
t
A
.
A.
,
“
S
o
lv
in
g
th
e
p
ro
b
lem
o
f
th
e
K
p
a
ra
m
e
ter
in
th
e
KN
N
c
las
sif
ier
u
sin
g
a
n
e
n
se
m
b
le
lea
rn
in
g
a
p
p
ro
a
c
h
,”
In
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
Co
mp
u
ter
S
c
ien
c
e
a
n
d
I
n
fo
rm
a
ti
o
n
S
e
c
u
rity
,
v
o
l.
1
2
,
n
o
.
8
,
2
0
1
4
.
[1
5
]
F
o
ste
r
P
.
,
M
a
u
c
h
M
.
,
Dix
o
n
S
.
,
“
S
e
q
u
e
n
ti
a
l
c
o
m
p
lex
it
y
a
s
a
d
e
sc
rip
t
o
r
f
o
r
m
u
sic
a
l
sim
il
a
ri
ty
,”
IEE
E/
ACM
T
ra
n
sa
c
ti
o
n
s
o
n
Au
d
i
o
,
S
p
e
e
c
h
a
n
d
L
a
n
g
u
a
g
e
Pro
c
e
ss
in
g
(
T
AS
L
P
)
,
v
o
l.
22
,
n
o
.
12
,
p
p
.
1
9
6
5
-
77
,
2
0
1
4
.
[1
6
]
G
e
o
rg
e
L
.
E
.
,
A
l
-
Hilo
E
.
A.
,
“
S
p
e
e
d
in
g
-
u
p
F
ra
c
tal
Co
l
o
r
Im
a
g
e
Co
m
p
re
ss
io
n
Us
in
g
M
o
m
e
n
ts
F
e
a
tu
re
s
Ba
se
d
o
n
S
y
m
m
e
tr
y
P
re
d
icto
r
,”
2
0
1
1
E
ig
h
t
h
In
ter
n
a
ti
o
n
a
l
C
o
n
fer
e
n
c
e
o
n
In
f
o
rm
a
ti
o
n
T
e
c
h
n
o
lo
g
y
:
Ne
w
Ge
n
e
ra
ti
o
n
s
,
2
0
1
1
.
[1
7
]
Be
d
a
n
A
.
K
.
,
G
e
o
rg
e
L
.
E.
,
“
S
p
e
e
d
in
g
-
u
p
f
ra
c
tal
a
u
d
i
o
c
o
m
p
re
ss
i
o
n
u
si
n
g
m
o
m
e
n
t
d
e
sc
rip
t
o
rs
,”
L
a
mb
e
rt
Ac
a
d
e
mic
Pu
b
li
s
h
in
g
(
L
AP
)
,
2
0
1
3
.
[1
8
]
S
a
ja
T
.
A
.
,
Ra
f
a
h
S
.
H.,
M
u
a
y
a
d
S
.
C.
,
“
EDM
P
re
p
ro
c
e
ss
in
g
a
n
d
Hy
b
rid
F
e
a
tu
re
S
e
lec
ti
o
n
f
o
r
Im
p
ro
v
in
g
Clas
si
f
ica
ti
o
n
A
c
c
u
ra
c
y
,
”
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
I
n
fo
rm
a
ti
o
n
T
e
c
h
n
o
lo
g
y
,
vol
.
96
,
n
o
1
,
n
o
.
1
9
9
2
-
8
6
4
5
,
2
0
1
9
.
[1
9
]
S
a
ja
T
a
h
a
,
“
Ira
q
i
S
tu
d
e
n
t
P
e
rf
o
r
m
a
n
c
e
P
re
d
ictio
n
,
”
M
e
n
d
e
ley
Da
ta
,
2
0
1
8
.
[
On
li
n
e
]
.
A
v
a
il
a
b
le:
h
tt
p
:
//
d
x
.
d
o
i.
o
rg
/
1
0
.
1
7
6
3
2
/sm
g
x
6
s5
p
w
r.
1
,
DO
I:
1
0
.
1
7
6
3
2
/sm
g
x
6
s5
p
w
r.
1
.
2
0
1
8
[2
0
]
Co
rtez
P
.
,
S
il
v
a
A
.
M.
,
“
Us
in
g
d
a
ta m
in
in
g
to
p
re
d
ict
se
c
o
n
d
a
ry
sc
h
o
o
l
stu
d
e
n
t
p
e
rf
o
rm
a
n
c
e
,
”
EUROS
IS
,
2
0
0
8
.
[2
1
]
G
h
iri
n
g
h
e
ll
i
L
.
M
.
,
V
y
b
iral
J
.
,
A
h
m
e
tcik
E
.
,
Ou
y
a
n
g
R
.
,
L
e
v
c
h
e
n
k
o
S
.
V
.
,
Dra
x
l
C
.
,
S
c
h
e
f
f
l
e
r
M
.
,
“
L
e
a
rn
in
g
p
h
y
sic
a
l
d
e
sc
rip
to
rs f
o
r
m
a
teria
ls sc
ien
c
e
b
y
c
o
m
p
re
ss
e
d
se
n
sin
g
,”
Ne
w
J
o
u
rn
a
l
o
f
P
h
y
sic
s
,
v
o
l.
19
,
n
o
.
2
,
2
0
1
7
.
[2
2
]
M
.
Z
.
H
.
J
.
,
Ho
ss
e
n
A
.
,
Ho
ss
e
n
J
.
,
Ra
ja
J
.
E
.
,
T
h
a
n
g
a
v
e
l
B
.
,
S
a
y
e
e
d
S
.
,
“
A
U
T
O
-
CDD
:
a
u
to
m
a
ti
c
c
lea
n
in
g
d
irt
y
d
a
ta
u
sin
g
m
a
c
h
in
e
lea
rn
in
g
tec
h
n
iq
u
e
s
,”
T
EL
KOM
NIKA
T
e
lec
o
m
mu
n
ica
t
io
n
,
C
o
mp
u
ti
n
g
,
El
e
c
tro
n
ics
a
n
d
Co
n
tro
l
,
v
o
l.
1
7
,
n
o
.
4
,
p
p
.
2
0
7
6
-
2
0
8
6
,
2
0
1
9
.
[
2
3
]
T
u
Y
.
K.
,
“
T
e
s
t
i
n
g
t
h
e
r
e
l
a
t
i
o
n
b
e
t
w
e
e
n
p
e
r
c
e
n
t
a
g
e
c
h
a
n
g
e
a
n
d
b
a
s
e
l
i
n
e
v
a
l
u
e
,”
S
c
i
e
n
t
i
f
i
c
R
e
p
o
r
t
s
,
v
o
l
.
6
,
p
p
.
1
-
8
,
2
0
1
6
.
[2
4
]
S
a
lal
Y
.
K
.
,
A
b
d
u
ll
a
e
v
S
.
M
.
,
Ku
m
a
r
M
.
,
“
Ed
u
c
a
ti
o
n
a
l
Da
ta
M
i
n
in
g
:
S
tu
d
e
n
t
P
e
rf
o
rm
a
n
c
e
P
re
d
ictio
n
in
A
c
a
d
e
m
i
c
,
”
v
o
l.
8
,
n
o
.
4
C,
p
p
.
54
-
5
9
,
2
0
1
9
.
[2
5
]
S
a
ti
N
.
U.
,
“
P
re
d
icti
o
n
o
f
S
tu
d
e
n
ts'
su
c
c
e
ss
in
M
a
th
e
m
a
ti
c
s
b
y
a
Clas
sif
i
c
a
ti
o
n
T
e
c
h
n
iq
u
e
V
ia
P
o
ly
h
e
d
ra
l
Co
n
ic
F
u
n
c
ti
o
n
s
,”
T
h
e
E
u
ra
si
a
Pro
c
e
e
d
in
g
s
o
f
E
d
u
c
a
ti
o
n
a
l
&
S
o
c
ia
l
S
c
ien
c
e
s
,
2
0
1
6
.
Evaluation Warning : The document was created with Spire.PDF for Python.