I
n
d
on
e
s
ian
Jou
r
n
al
o
f
E
lec
t
r
ica
l
E
n
gin
e
e
r
in
g
a
n
d
Com
p
u
t
e
r
S
c
ience
Vo
l
.
25
,
N
o
.
2
,
F
e
b
r
ua
r
y
20
22
,
pp.
1115
~
112
2
I
S
S
N:
2502
-
4752,
DO
I
:
10
.
11591/i
j
e
e
c
s
.
v
25
.i
2
.
pp
1115
-
112
2
1115
Jou
r
n
al
h
o
m
e
page
:
ht
tp:
//
ij
e
e
c
s
.
iaes
c
or
e
.
c
om
Hy
b
r
id
d
y
n
a
m
ic
c
h
u
n
k
e
n
se
m
b
le
m
od
e
l
f
or
m
u
lti
-
c
la
ss d
a
t
a
st
r
e
am
s
Var
s
h
a
S
ac
h
in
K
h
an
d
e
k
a
r
,
P
r
avin
S
h
r
in
at
h
D
e
pa
r
tm
e
nt
of
C
o
mput
e
r
E
ngi
ne
e
r
in
g, M
uke
s
h P
a
te
l
S
c
h
ool
of
T
e
c
hn
o
l
o
g
y
M
a
na
ge
m
e
nt
a
nd
E
ngi
n
e
e
r
in
g, N
M
I
M
S
U
ni
v
e
r
s
it
y,
M
umba
i,
I
ndi
a
Ar
t
ic
l
e
I
n
f
o
AB
S
T
RA
CT
A
r
ti
c
le
h
is
tor
y
:
R
e
c
e
i
ve
d
A
u
g
19
,
2021
R
e
vi
s
e
d
No
v
18
,
2021
A
c
c
e
pt
e
d
De
c
1
,
2021
I
n
t
h
e
an
a
l
y
s
i
s
mo
r
e
s
p
ec
i
f
i
c
al
l
y
i
n
t
h
e
c
l
as
s
i
fi
c
at
i
o
n
o
f
co
n
t
i
n
u
o
u
s
d
at
a
s
t
re
a
m
u
s
i
n
g
m
a
c
h
i
n
e
l
e
arn
i
n
g
a
l
g
o
ri
t
h
m
s
j
o
i
n
t
o
ccu
rr
e
n
ce
o
f
c
o
n
ce
p
t
d
ri
ft
an
d
i
m
b
al
an
ce
d
i
s
s
u
e
b
ec
o
me
s
m
o
r
e
p
ro
v
o
c
at
i
v
e
.
A
l
s
o
,
i
m
b
al
an
ce
i
s
s
u
e
i
s
a
g
ai
n
mo
re
c
h
a
l
l
en
g
i
n
g
w
h
en
t
h
e
d
a
t
a
s
t
re
a
m
i
s
m
u
l
t
i
-
cl
as
s
w
i
t
h
mi
n
o
ri
t
y
c
l
as
s
an
d
t
h
at
i
s
t
o
o
w
i
t
h
d
at
a
-
d
i
ff
i
c
u
l
t
y
fa
c
t
o
rs
.
In
c
r
eme
n
t
al
l
e
a
rn
i
n
g
w
i
t
h
e
n
s
em
b
l
e
mo
d
e
l
s
f
o
u
n
d
m
o
r
e
p
ro
m
i
s
i
n
g
i
n
h
an
d
l
i
n
g
t
h
e
s
e
s
i
s
s
u
e
s
.
B
u
t
m
o
s
t
o
f
t
h
e
ap
p
ro
ac
h
e
s
ar
e
f
o
r
t
w
o
-
cl
as
s
d
at
a
s
t
re
am
s
w
h
i
c
h
c
an
’t
b
e
u
t
i
l
i
z
ed
f
o
r
mu
l
t
i
-
c
l
as
s
d
at
a
s
t
re
am
s
.
I
n
t
h
i
s
p
ap
e
r
w
e
h
av
e
d
e
s
i
g
n
ed
h
y
b
ri
d
d
y
n
a
m
i
c
ch
u
n
k
e
n
s
em
b
l
e
mo
d
el
(H
D
CE
M)
fo
r
t
h
e
cl
as
s
i
fi
c
at
i
o
n
o
f
mu
l
t
i
-
c
l
as
s
i
n
s
e
c
t
-
d
at
a
s
t
re
am
f
o
r
h
a
n
d
l
i
n
g
i
m
b
al
a
n
ce
an
d
c
o
n
ce
p
t
d
ri
ft
i
s
s
u
e
.
T
o
d
e
al
w
i
t
h
i
m
b
al
a
n
ce
i
s
s
u
e
w
e
h
av
e
p
r
o
p
o
s
e
d
e
ffec
t
i
v
e
s
p
l
i
t
b
ag
g
i
n
g
a
l
g
o
ri
t
h
m
w
h
i
c
h
h
as
a
c
h
i
e
v
ed
b
e
t
t
e
r
p
e
rfo
r
m
an
ce
o
n
m
i
n
o
r
i
t
y
cl
as
s
rec
a
l
l
an
d
F
-
me
as
u
r
e
o
n
arr
i
v
i
n
g
d
y
n
ami
c
c
h
u
n
k
s
o
f
d
at
a
fro
m
m
u
l
t
i
-
c
l
as
s
d
at
a
s
t
re
a
m.
H
D
CE
M
mo
d
el
c
a
n
ad
ap
t
t
o
ab
ru
p
t
an
d
g
rad
u
al
d
r
i
ft
b
e
c
a
u
s
e
i
t
h
as
c
o
m
b
i
n
ed
f
e
at
u
r
e
s
o
f
b
o
t
h
o
n
l
i
n
e
an
d
c
h
u
n
k
-
b
as
e
d
l
e
ar
n
i
n
g
t
o
g
e
t
h
e
r.
I
t
h
as
a
ch
i
e
v
ed
a
v
e
rag
e
7
8
%
mi
n
o
ri
t
y
c
l
as
s
rec
a
l
l
i
n
ab
ru
p
t
i
n
s
ec
t
d
at
a
s
t
re
a
m
an
d
7
1
%
i
n
g
rad
u
al
d
ri
ft
i
n
s
ec
t
s
t
re
a
m
.
K
e
y
w
o
r
d
s
:
C
o
n
c
e
pt
dr
i
f
t
Da
t
a
s
t
r
e
a
m
D
y
n
a
mi
c
c
h
u
n
k
E
n
s
e
m
b
l
e
l
e
a
r
ni
n
g
I
m
b
a
l
a
n
c
e
da
t
a
Th
i
s
i
s
a
n
o
p
en
a
c
ces
s
a
r
t
i
c
l
e
u
n
d
e
r
t
h
e
CC
B
Y
-
SA
l
i
cen
s
e.
C
or
r
e
s
pon
din
g
A
u
th
or
:
Va
r
s
ha
S
a
c
hi
n
Kh
a
n
de
ka
r
De
pa
r
t
m
e
n
t
o
f
C
o
m
put
e
r
E
n
g
i
n
e
e
r
i
n
g,
M
uke
s
h
P
a
te
l
S
c
h
o
o
l
o
f
T
e
c
hn
o
l
o
g
y
M
a
n
a
ge
m
e
n
t
a
n
d
E
n
g
i
ne
e
r
i
n
g
NM
I
M
S
Uni
ve
r
s
i
t
y
,
V.
L
,
P
h
e
r
o
z
e
s
h
a
h
M
e
h
t
a
R
d,
V
i
l
e
P
a
r
l
e
W
e
s
t
,
M
um
b
a
i
M
a
h
a
r
a
s
h
t
r
a
400056,
I
n
d
i
a
E
m
a
i
l
:
v
a
r
s
h
a
.
k
h
a
n
de
ka
r
@
g
m
a
il
.
c
o
m
1.
I
NT
RODU
C
T
I
ON
I
n
to
da
y
'
s
d
i
g
i
t
a
l
wo
r
l
d
due
t
o
e
v
o
l
ut
i
o
n
i
n
t
e
c
hn
o
l
o
gy
v
o
l
u
m
i
n
o
u
s
da
t
a
i
s
i
nc
r
e
a
s
i
n
g
a
t
r
a
p
i
d
s
pe
e
d.
T
h
e
da
t
a
ge
n
e
r
a
t
e
d
o
n
f
l
y
i
s
c
a
ll
e
d
a
s
a
s
t
r
e
a
m
i
ng
da
t
a
.
T
h
e
s
tr
e
a
m
i
ng
da
t
a
a
n
a
l
y
s
i
s
us
i
ng
m
a
c
hi
ne
l
e
a
r
ni
ng
t
e
c
h
ni
que
s
e
n
dur
e
m
a
ny
c
h
a
ll
e
n
ge
s
li
ke
hi
g
h
da
t
a
v
e
l
o
c
i
t
y
,
hi
g
h
v
o
l
u
m
e
o
f
da
t
a
,
c
h
a
n
g
e
i
n
t
he
u
n
de
r
lyi
ng
d
i
s
t
r
i
b
ut
i
o
n
o
v
e
r
t
h
e
t
i
m
e
.
F
o
r
e
x
a
m
p
l
e
,
t
h
e
pr
o
p
e
r
t
i
e
s
o
f
m
a
li
c
i
o
us
u
ni
f
o
r
m
r
e
s
o
ur
c
e
l
o
c
a
to
r
(
UR
L
s
)
a
nd
f
r
a
udu
l
e
n
t
t
r
a
n
s
a
c
t
i
o
n
s
a
s
we
ll
a
s
s
pa
m
t
we
e
t
s
p
o
s
t
e
d
by
s
pa
mm
e
r
s
a
r
e
c
h
a
n
g
i
ng
c
o
n
t
i
n
uo
us
l
y
[
1]
,
[
2]
.
I
n
t
h
e
a
n
a
ly
s
i
s
c
l
a
s
s
i
f
i
c
a
t
i
o
n
o
f
i
ns
t
a
n
c
e
s
f
r
o
m
da
t
a
s
t
r
e
a
m
,
m
o
de
l
h
a
ve
a
hy
pot
h
e
s
i
s
w
hi
c
h
f
i
nds
m
a
pp
i
ng
b
e
t
we
e
n
f
e
a
t
ur
e
v
a
r
i
a
bl
e
s
(
X)
a
n
d
t
a
r
ge
t
v
a
r
i
a
bl
e
s
(
Y)
whi
c
h
a
r
e
c
a
l
l
e
d
a
s
l
a
b
e
l
s
o
f
t
h
e
i
n
s
t
a
nc
e
s
.
T
h
e
r
e
i
s
a
n
e
e
d
o
f
a
da
pt
i
v
e
m
a
c
hi
ne
l
e
a
r
ni
ng
m
o
de
l
s
w
hi
c
h
a
r
e
a
bl
e
t
o
a
d
a
p
t
t
h
e
m
s
e
l
ve
s
to
n
e
w
un
de
r
l
yi
ng
d
i
s
t
r
i
b
ut
i
o
n
whi
c
h
i
s
kn
o
wn
a
s
a
c
o
n
c
e
pt
dr
i
f
t
.
T
h
e
c
o
n
c
e
pt
d
r
i
f
t
s
a
r
e
m
a
i
nly
f
ur
t
h
e
r
c
a
t
e
g
o
r
i
z
e
d
a
s
vi
r
t
ua
l
a
n
d
r
e
a
l
dr
i
f
t
.
V
i
r
t
ua
l
dr
i
f
t
:
T
h
e
r
e
i
s
a
c
h
a
n
ge
i
n
t
h
e
di
s
t
r
i
b
ut
i
o
n
o
f
f
e
a
t
ur
e
s
(
)
o
f
i
n
s
t
a
n
c
e
s
o
r
c
h
a
n
ge
i
n
t
h
e
d
i
s
t
r
i
b
ut
i
o
n
o
f
c
o
n
c
e
pt
s
o
r
t
a
r
ge
t
v
a
r
i
a
bl
e
s
(
)
.
R
e
a
l
dr
if
t
:
T
h
e
r
e
l
a
t
i
o
ns
hi
p
be
t
we
e
n
i
nput
v
a
r
i
a
bl
e
s
a
n
d
t
a
r
ge
t
c
o
n
c
e
pt
i
s
c
h
a
n
g
i
ng.
T
hi
s
c
h
a
n
ge
i
s
c
h
a
n
ge
i
n
t
h
e
l
i
ke
li
h
o
o
d
(
|
)
t
h
us
(
|
)
≠
+
1
(
|
)
or
c
h
a
n
g
e
i
n
po
s
t
e
r
i
o
r
pr
o
b
a
bil
i
t
y
d
i
s
t
r
i
b
ut
i
o
n
(
|
)
t
h
us
(
|
)
≠
+
1
(
|
)
.
T
h
e
r
e
a
l
dr
i
f
t
a
f
f
e
c
t
s
o
n
t
h
e
de
c
i
s
i
o
n
b
o
un
da
r
y
.
I
n
m
a
ny
r
e
a
l
-
wo
r
l
d
pr
o
bl
e
m
s
,
m
o
s
t
o
f
th
e
t
i
m
e
t
h
e
y
o
c
c
ur
a
t
s
a
m
e
t
i
m
e
.
De
a
l
i
ng
w
i
t
h
c
o
n
c
e
pt
dr
i
f
t
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2502
-
4752
I
n
do
n
e
s
i
a
n
J
E
l
e
c
E
n
g
&
C
o
m
p
S
c
i
,
Vo
l
.
25
,
N
o
.
2
,
F
e
b
r
ua
r
y
20
22
:
1115
-
112
2
1116
i
s
s
ue
i
n
t
h
e
da
t
a
s
t
r
e
a
m
c
l
a
s
s
i
f
i
c
a
t
i
o
n
t
a
s
k
h
a
s
be
c
o
m
e
c
h
a
ll
e
n
g
i
ng
a
n
d
h
a
s
b
e
c
o
m
e
a
t
t
e
n
t
i
v
e
i
n
r
e
s
e
a
r
c
h
c
o
m
m
u
ni
t
y
.
E
a
r
l
i
e
r
a
ppr
o
a
c
h
e
s
ha
v
e
us
e
d
s
o
m
e
s
t
a
t
i
s
t
i
c
a
l
c
h
a
nge
de
t
e
c
t
i
o
n
t
e
s
t
s
to
m
o
ni
t
o
r
c
o
n
c
e
pt
d
r
i
f
t
.
I
n
s
t
e
a
d
o
f
us
i
ng
a
c
c
ur
a
c
y
,
dr
i
f
t
de
t
e
c
t
i
o
n
m
e
t
h
o
d
f
o
r
o
nl
i
n
e
c
l
a
s
s
im
ba
l
a
n
c
e
(
DD
M
-
OC
I
)
[
3]
m
o
ni
t
o
r
s
t
h
e
c
l
a
s
s
r
e
c
a
l
l
to
de
a
l
w
i
t
h
im
ba
l
a
n
c
e
i
s
s
ue
.
F
o
r
de
t
e
c
t
i
n
g
dr
i
f
t
o
n
po
s
i
t
i
v
e
a
n
d
ne
ga
t
i
v
e
c
l
a
s
s
,
l
i
ne
a
r
f
o
ur
r
a
t
e
s
[
4]
h
a
v
e
u
s
e
d
t
r
ue
p
o
s
i
t
i
ve
r
a
t
e
,
tr
ue
n
e
ga
t
i
v
e
r
a
t
e
,
p
o
s
i
t
i
v
e
pr
e
d
i
c
t
e
d
v
a
l
ue
a
n
d
n
e
ga
t
i
v
e
pr
e
d
i
c
t
e
d
v
a
l
u
e
.
P
a
ge
–
hi
nk
l
e
y
(
P
A
UC
-
P
H)
[
5]
us
e
s
P
H
-
t
e
s
t
[
6]
to
de
t
e
c
t
dr
i
f
t
.
An
o
t
h
e
r
i
m
po
r
t
a
n
t
i
s
s
ue
w
hi
c
h
a
f
f
e
c
t
s
t
h
e
c
l
a
s
s
if
i
c
a
t
i
o
n
m
o
de
l
’
s
pe
r
f
o
r
m
a
n
c
e
i
s
c
l
a
s
s
im
ba
l
a
n
c
e
w
h
e
r
e
n
u
m
be
r
o
f
i
ns
t
a
n
c
e
s
o
f
o
n
e
o
f
t
h
e
c
l
a
s
s
i
s
do
m
i
na
n
t
o
v
e
r
t
h
e
ot
h
e
r
.
W
h
e
n
im
ba
l
a
n
c
e
a
n
d
c
o
n
c
e
pt
dr
i
f
t
b
o
t
h
t
h
e
s
e
pr
o
bl
e
m
s
o
c
c
ur
a
t
s
a
m
e
t
i
m
e
i
n
da
t
a
s
t
r
e
a
m
,
t
h
e
y
w
i
l
l
t
e
nd
to
e
x
a
s
pe
r
a
t
e
e
a
c
h
o
t
h
e
r
.
W
he
n
c
l
a
s
s
i
m
ba
l
a
n
c
e
o
c
c
ur
s
,
i
t
b
e
c
o
m
e
s
d
i
f
f
i
c
u
l
t
to
de
t
e
c
t
t
h
e
c
o
n
c
e
pt
dr
i
f
t
a
n
d
c
o
nf
o
r
m
t
h
e
m
o
de
l
to
n
e
w
di
s
t
r
i
b
ut
i
o
n
.
A
c
t
i
v
e
a
n
d
P
a
s
s
i
ve
a
ppr
o
a
c
h
e
s
h
a
v
e
be
e
n
us
e
d
f
o
r
h
a
n
d
li
ng
t
h
e
c
o
n
c
e
pt
dr
i
f
t
s
.
A
c
t
i
ve
a
ppr
o
a
c
h
i
nv
o
l
v
e
s
e
x
p
li
c
i
t
de
t
e
c
t
i
o
n
t
e
c
h
ni
qu
e
whil
e
pa
s
s
i
ve
a
ppr
o
a
c
h
i
s
ba
s
e
d
o
n
a
da
pt
i
o
n
o
f
m
o
de
l
.
P
a
s
s
i
ve
a
ppr
o
a
c
h
i
s
m
o
r
e
s
uc
c
e
s
s
f
u
l
a
s
c
o
m
pa
r
e
d
to
a
c
t
i
v
e
w
hi
c
h
o
v
e
r
c
o
m
e
s
t
h
e
l
im
i
t
a
t
i
o
n
s
i
n
a
n
a
c
t
i
v
e
a
ppr
o
a
c
h
.
C
l
a
s
s
i
m
ba
l
a
nc
e
i
n
s
t
a
t
i
o
n
a
r
y
o
r
i
n
s
t
a
t
i
c
e
nvi
r
o
nm
e
n
t
i
s
m
o
s
t
a
ddr
e
s
s
e
d
pr
o
bl
e
m
us
i
ng
v
a
r
i
o
us
t
e
c
h
ni
que
s
.
B
ut
t
h
e
r
e
a
r
e
o
nl
y
f
e
w
m
o
de
l
s
ha
v
e
b
e
e
n
f
o
un
d
whi
c
h
a
r
e
de
a
li
ng
w
i
t
h
b
o
t
h
c
o
n
c
e
pt
dr
i
f
t
a
n
d
c
l
a
s
s
im
ba
l
a
n
c
e
s
i
m
u
l
t
a
ne
o
us
l
y
.
T
h
e
s
e
m
o
de
l
s
a
r
e
c
a
t
e
go
r
i
z
e
d
a
s
o
nli
ne
a
n
d
c
h
u
n
k
-
b
a
s
e
d
m
o
de
l
s
.
C
h
u
nk
b
a
s
e
d
m
o
de
l
s
m
o
s
t
l
y
h
a
v
e
u
s
e
d
e
n
s
e
m
b
l
e
l
e
a
r
ni
ng
a
ppr
o
a
c
h
.
Onl
i
ne
l
e
a
r
ni
ng
m
o
de
l
s
[
6]
,
[
7]
a
da
p
t
t
h
e
m
s
e
l
v
e
s
f
o
r
e
v
e
r
y
i
ns
t
a
n
c
e
a
r
r
i
vi
ng
i
n
t
h
e
s
t
r
e
a
m
.
T
h
e
s
e
a
r
e
m
o
r
e
e
f
f
e
c
t
i
v
e
i
n
h
a
n
d
li
ng
a
b
r
upt
ki
n
d
o
f
dr
i
f
t
.
I
n
c
h
u
n
k
-
b
a
s
e
d
l
e
a
r
ni
ng,
m
o
de
l
i
s
n
o
t
a
da
p
t
i
n
g
i
t
s
e
l
f
u
n
t
i
l
c
e
r
t
a
i
n
nu
m
b
e
r
o
f
i
ns
t
a
n
c
e
s
a
r
e
n
o
t
c
o
l
l
e
c
t
e
d
i
n
a
b
u
f
f
e
r
,
who
s
e
s
i
z
e
i
s
m
o
s
t
l
y
pr
e
-
de
c
i
de
d
a
n
d
i
t
i
s
f
i
xe
d
t
h
r
o
ugh
o
u
t
t
h
e
a
n
a
ly
s
i
s
o
f
da
t
a
s
tr
e
a
m
.
S
o
m
e
c
h
u
n
k
-
b
a
s
e
d
m
e
t
ho
ds
h
a
v
e
us
e
d
a
s
s
i
g
nm
e
n
t
o
f
dy
na
m
i
c
we
i
g
h
t
s
to
c
o
m
po
ne
n
t
c
l
a
s
s
if
i
e
r
s
i
n
e
ns
e
m
b
l
e
m
o
de
l
b
a
s
e
d
o
n
t
h
e
a
c
c
ur
a
c
y
m
e
a
s
ur
e
[
8]
.
T
h
e
r
e
we
r
e
s
o
m
e
f
i
xe
d
s
i
z
e
c
h
u
n
k
b
a
s
e
d
m
e
t
h
o
ds
pr
o
p
o
s
e
d
whi
c
h
we
r
e
us
e
d
f
o
r
c
l
a
s
s
i
f
i
c
a
t
i
o
n
im
ba
l
a
n
c
e
d
n
o
n
-
s
t
a
t
i
o
n
a
r
y
da
t
a
s
tr
e
a
m
s
[
9]
,
[
10]
.
I
n
un
c
o
r
r
e
l
a
t
e
d
b
a
gg
i
n
g
[
11]
c
ur
r
e
n
t
c
h
un
k
i
s
b
a
lan
c
e
d
b
y
pr
e
s
e
r
vi
ng
t
h
e
m
i
n
o
r
i
t
y
c
l
a
s
s
e
x
a
m
p
l
e
s
f
r
o
m
pr
e
vio
us
c
h
un
k
s
.
B
ut
h
e
r
e
t
h
e
l
im
i
t
a
t
i
o
n
i
s
us
a
ge
o
f
m
e
m
o
r
y
f
o
r
s
to
r
i
n
g
pa
s
t
da
t
a
i
n
s
t
a
n
c
e
s
a
n
d
a
l
s
o
t
hi
s
c
a
n’
t
a
da
pt
to
n
e
w
c
o
n
c
e
pt
r
a
pi
d
ly
.
I
m
pr
o
v
e
m
e
n
t
i
n
t
hi
s
t
e
c
hni
que
i
s
o
b
s
e
r
v
e
d
i
n
s
e
l
e
c
t
i
ve
l
y
r
e
c
ur
s
i
ve
(
S
E
R
A
)
[
12
]
a
nd
i
n
r
e
c
ur
s
i
v
e
e
n
s
e
m
b
l
e
a
ppr
o
a
c
h
(
R
E
A
)
[
13
]
by
s
e
l
e
c
t
i
n
g
o
nl
y
m
o
s
t
s
i
mi
l
a
r
pa
s
t
m
i
n
o
r
i
t
y
i
ns
t
a
nc
e
s
.
D
i
t
z
l
e
r
a
n
d
P
o
l
i
ka
r
[
14]
pr
o
p
o
s
e
d
t
w
o
c
h
un
k
-
b
a
s
e
d
e
n
s
e
m
b
les
c
a
l
l
e
d
l
e
a
r
n
++.
C
DS
t
h
a
t
i
s
c
o
nc
e
pt
dr
i
f
t
w
i
t
h
s
m
o
t
e
a
n
d
l
e
a
r
n
++.
NI
E
whi
c
h
i
s
n
o
n
-
s
t
a
t
i
o
n
a
r
y
im
b
a
l
a
n
c
e
d
e
nvi
r
o
nm
e
n
t
.
B
ot
h
a
r
e
i
ns
p
i
r
e
d
f
r
o
m
l
e
a
r
n
+
+
.
NS
E
to
h
a
n
d
l
e
i
m
ba
l
a
nc
e
d
da
t
a
s
t
r
e
a
m
s
w
i
t
h
c
o
n
c
e
pt
dr
i
f
t
[
15]
wh
e
r
e
l
e
a
r
n
+
+
.
NSE
de
a
l
s
w
i
t
h
c
o
n
c
e
pt
dr
i
f
t
us
i
n
g
a
d
y
n
a
mi
c
we
i
g
h
t
i
n
g
s
t
r
a
t
e
gy
a
n
d
S
M
OT
E
f
o
r
ba
l
a
n
c
i
ng
t
h
e
m
i
n
o
r
i
t
y
c
l
a
s
s
i
ns
t
a
n
c
e
s
.
An
e
n
s
e
m
b
l
e
o
f
s
u
bs
e
t
o
f
o
nli
ne
s
e
que
n
t
i
a
l
e
x
t
r
e
m
e
l
e
a
r
ni
ng
m
a
c
hi
ne
(
E
S
OS
-
E
L
M
)
[
16]
h
a
ve
c
o
n
s
t
r
uc
t
e
d
a
n
d
s
tor
e
d
we
i
g
h
t
m
a
t
r
i
c
e
s
f
o
r
e
v
e
r
y
c
h
u
n
k.
Gr
a
dua
l
r
e
s
a
m
p
l
i
ng
e
n
s
e
m
ble
(
GR
E
)
[
17]
us
e
d
c
l
u
s
t
e
r
i
n
g
t
e
c
h
ni
que
f
o
r
s
e
l
e
c
t
i
n
g
th
e
m
i
n
o
r
i
t
y
c
l
a
s
s
s
a
m
p
l
e
s
f
r
o
m
pr
e
vi
o
us
c
h
u
n
k.
T
o
ge
n
e
r
a
t
e
t
r
a
i
ni
ng
da
t
a
s
e
t,
t
h
e
y
h
a
v
e
us
e
d
de
ns
i
t
y
b
a
s
e
d
s
pa
t
i
a
l
o
f
a
pp
l
i
c
a
t
i
o
ns
w
i
t
h
n
o
i
s
e
(
DB
S
C
A
N)
c
l
u
s
t
e
r
i
n
g
w
i
t
h
mi
n
o
r
i
t
y
c
l
a
s
s
a
n
d
t
r
i
e
d
to
m
i
nim
i
z
e
o
v
e
r
l
a
pp
i
ng
wi
t
h
m
a
j
o
r
i
t
y
c
l
a
s
s
.
Al
s
o
,
i
n
f
e
w
c
h
u
n
k
-
b
a
s
e
d
m
e
t
h
o
ds
pr
e
s
e
r
v
e
t
h
e
mi
n
o
r
i
t
y
s
a
m
p
l
e
s
f
r
o
m
pr
e
vi
o
us
c
h
u
n
k
w
hi
c
h
a
r
e
m
e
r
ge
d
w
i
t
h
t
h
e
m
i
n
o
r
i
t
y
s
a
m
p
l
e
s
i
n
t
h
e
s
uc
c
e
e
d
i
n
g
c
h
u
n
k
to
ge
t
e
n
o
ugh
n
u
m
b
e
r
o
f
m
i
n
o
r
i
t
y
s
a
m
p
l
e
s
,
h
o
we
v
e
r
,
t
hi
s
a
s
s
u
m
pt
i
o
n
m
a
y
f
a
il
a
s
i
m
ba
l
a
n
c
e
r
a
t
i
o
m
a
y
n
o
t
b
e
f
i
xe
d
a
n
d
m
a
y
be
c
ha
n
g
i
n
g
o
v
e
r
th
e
t
i
m
e
.
R
e
vi
e
w
s
h
o
ws
t
h
a
t
ba
gg
i
n
g
b
a
s
e
d
e
ns
e
m
b
l
e
s
a
r
e
us
e
f
u
l
f
o
r
i
m
pr
o
vi
ng
t
h
e
pe
r
f
o
r
m
a
n
c
e
o
f
c
l
a
s
s
i
f
i
e
r
f
o
r
de
a
l
i
n
g
w
i
t
h
i
m
ba
l
a
n
c
e
i
s
s
ue
.
P
r
o
p
o
s
e
d
m
e
t
h
o
d
i
s
b
a
s
e
d
o
n
b
a
gg
i
ng
a
ppr
o
a
c
h
a
n
d
c
o
m
pa
r
e
d
w
i
t
h
f
o
l
l
o
w
i
ng
s
t
a
t
e
-
of
-
t
h
e
-
a
r
t
b
a
gg
i
n
g
m
e
t
h
o
ds
.
i
)
o
v
e
r
b
a
gg
i
ng
[
18]
,
t
hi
s
t
e
c
h
ni
que
r
e
l
i
e
s
o
n
a
r
a
n
do
m
o
v
e
r
s
a
m
p
li
ng
o
f
mi
n
o
r
i
t
y
c
l
a
s
s
t
o
a
c
qui
r
e
e
a
c
h
s
u
b
s
e
t
o
f
da
t
a
s
e
t
.
He
r
e
e
v
e
r
y
s
ubs
e
t
w
i
l
l
i
nc
l
ude
a
l
l
t
h
e
o
r
i
g
i
na
l
e
x
a
m
p
l
e
s
a
n
d
dup
l
i
c
a
t
e
s
a
m
p
l
e
s
o
f
r
a
n
do
m
l
y
s
e
l
e
c
t
e
d
i
ns
t
a
n
c
e
s
o
f
t
h
e
m
i
n
o
r
i
t
y
c
l
a
s
s
;
ii
)
s
y
n
t
he
t
i
c
m
i
n
o
r
i
t
y
o
v
e
r
s
a
m
p
li
ng
tec
hni
qu
e
(
S
M
OT
E
)
b
a
gg
i
n
g
[
18]
,
t
h
i
s
a
ppr
o
a
c
h
m
a
ke
s
us
e
o
f
S
M
OT
E
a
l
go
r
i
t
hm
f
o
r
c
r
e
a
t
i
n
g
n
e
w
i
ns
t
a
n
c
e
s
f
r
o
m
mi
n
o
r
i
t
y
c
l
a
s
s
.
T
o
i
n
c
r
e
a
s
e
d
i
ve
r
s
i
t
y
i
n
s
u
b
s
e
t
m
a
j
o
r
i
t
y
c
l
a
s
s
i
ns
t
a
n
c
e
s
a
r
e
s
e
l
e
c
t
e
d
r
a
n
do
m
ly
;
iii
)
u
n
de
r
b
a
gg
i
ng
[
19]
,
h
e
r
e
i
n
s
t
e
a
d
o
f
us
i
n
g
un
de
r
s
a
m
p
l
i
ng,
i
t
us
e
s
ov
e
r
s
a
m
p
li
ng
t
e
c
hni
qu
e
f
o
r
ge
n
e
r
a
t
i
n
g
s
u
b
s
e
t
s
f
r
o
m
o
r
i
g
i
na
l
da
t
a
s
e
t.
B
e
c
a
us
e
o
f
un
de
r
s
a
m
p
l
i
ng
s
i
z
e
o
f
s
u
b
s
e
t
ge
t
s
r
e
duc
e
d
;
a
n
d
iv
)
u
n
de
r
o
v
e
r
b
a
gg
i
n
g
[
19]
,
t
hi
s
a
ppr
o
a
c
h
us
e
s
b
o
t
h
un
de
r
s
a
m
p
li
ng
a
n
d
o
v
e
r
s
a
m
p
li
ng
a
l
o
n
g
-
w
i
t
h
S
M
OT
E
b
a
gg
i
n
g
.
M
o
s
t
of
t
h
e
m
e
t
h
o
ds
r
e
vi
e
we
d
de
s
i
g
n
e
d
f
o
r
t
w
o
c
l
a
s
s
c
l
a
s
s
if
i
c
a
t
i
o
n
i
n
da
t
a
s
tr
e
a
m
,
s
o
t
h
e
r
e
wa
s
o
n
e
mi
n
o
r
i
t
y
w
hil
e
a
n
o
t
h
e
r
o
n
e
wa
s
m
a
j
o
r
i
t
y
c
l
a
s
s
.
B
ut
t
h
e
s
e
m
e
t
h
o
ds
we
r
e
f
a
i
l
e
d
to
h
a
n
d
l
e
m
u
l
t
i
p
l
e
mi
n
o
r
i
t
y
c
l
a
s
s
e
s
a
n
d
t
h
e
dy
n
a
mi
c
i
m
ba
l
a
nc
e
r
a
t
i
o
i
n
m
u
l
t
i
-
c
l
a
s
s
da
t
a
s
tr
e
a
m
s
.
Al
s
o
,
t
h
e
s
i
z
e
o
f
c
h
u
n
k
c
o
n
s
i
de
r
e
d
a
f
f
e
c
t
t
h
e
pe
r
f
o
r
m
a
nc
e
o
f
a
m
o
de
l
,
e
x
p
l
i
c
i
t
ly
w
he
n
t
h
e
d
a
t
a
s
tr
e
a
m
i
s
i
m
ba
l
a
n
c
e
d.
I
n
t
hi
s
pa
pe
r
,
we
h
a
ve
pr
o
p
o
s
e
d
a
hy
br
i
d
d
y
n
a
mi
c
c
h
u
n
k
e
n
s
e
m
b
l
e
m
o
de
l
(
HD
C
E
M
)
f
o
r
c
l
a
s
s
if
i
c
a
t
i
o
n
o
f
m
u
l
t
i
-
c
l
a
s
s
i
ns
e
c
t
i
m
ba
l
a
n
c
e
d
da
t
a
s
t
r
e
a
m
s
.
I
n
pr
o
p
o
s
e
d
e
n
s
e
m
b
l
e
m
o
de
l
de
c
i
s
i
o
n
tr
e
e
i
s
us
e
d
a
s
a
c
a
n
d
i
da
t
e
c
l
a
s
s
i
f
i
e
r
.
F
o
r
t
e
s
t
da
t
a
c
l
a
s
s
if
i
c
a
t
i
o
n
d
y
n
a
mi
c
e
ns
e
m
bl
e
s
e
l
e
c
t
i
o
n
i
s
us
e
d.
T
h
e
pr
o
p
o
s
e
d
m
o
de
l
h
a
s
f
o
l
l
o
w
i
n
g
a
d
va
n
t
a
ge
s
:
i
)
I
t
i
s
a
bl
e
to
pe
r
f
o
r
m
t
h
e
m
u
l
t
i
-
c
l
a
s
s
c
l
a
s
s
if
i
c
a
t
i
o
n
i
n
n
o
n
-
s
t
a
t
i
o
n
a
r
y
d
a
t
a
s
tr
e
a
m
s
,
ii
)
I
m
ba
l
a
n
c
e
i
s
s
ue
i
s
r
e
s
o
l
v
e
d
us
i
ng
n
o
ve
l
s
p
l
i
t
b
a
s
e
d
r
e
s
a
m
p
li
ng
e
ns
e
m
b
l
e
a
l
go
r
i
t
hm
,
i
ii
)
I
t
c
a
n
h
a
n
d
l
e
a
b
r
upt
a
n
d
g
r
a
dua
l
c
o
n
c
e
pt
dr
i
f
t
s
,
a
s
f
e
a
t
ur
e
s
o
f
b
o
th
o
nl
i
ne
a
n
d
c
h
u
n
k
-
b
a
s
e
d
l
e
a
r
ni
ng
a
r
e
c
o
m
bi
ne
d,
a
n
d
iv
)
F
o
r
t
e
s
t
da
t
a
dy
n
a
mi
c
e
n
s
e
m
b
l
e
s
e
l
e
c
t
i
o
n
i
s
a
pp
li
e
d
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
do
n
e
s
i
a
n
J
E
l
e
c
E
n
g
&
C
o
m
p
S
c
i
I
S
S
N:
2502
-
4752
Hy
br
id
dy
namic
c
hunk
e
ns
e
mble
mode
l
f
or
multi
-
c
las
s
data
s
tr
e
ams
(
V
a
r
s
ha
Sac
hin
K
hande
k
ar
)
1117
2.
RE
S
E
AR
CH
M
E
T
HO
D
2.
1.
Dat
a
s
e
t
P
r
o
p
o
s
e
d
m
o
de
l
h
a
s
b
e
e
n
e
v
a
l
ua
t
e
d
us
i
n
g
m
o
s
qu
it
o
i
ns
e
c
t
s
tr
e
a
m
da
t
a
s
e
t
r
e
l
e
a
s
e
d
by
a
ut
h
o
r
s
[
20]
.
T
hi
s
da
t
a
s
e
t
h
a
s
33
f
e
a
t
ur
e
s
a
n
d
s
i
x
c
l
a
s
s
l
a
b
e
l
s
.
T
h
e
c
l
a
s
s
l
a
b
e
l
s
a
r
e
t
h
e
s
pe
c
i
e
s
o
f
t
h
r
e
e
t
y
pe
s
o
f
m
o
s
qu
i
t
o
e
s
f
r
o
m
b
o
t
h
s
e
x
e
s
.
De
t
a
i
l
s
o
f
t
h
e
s
e
m
o
s
qu
i
t
o
s
pe
c
i
e
s
a
r
e
gi
ve
n
a
s
s
h
o
wn
.
A
e
de
s
a
e
g
y
pt
i
.
T
hi
s
m
o
s
qu
i
t
o
s
pe
c
i
e
s
i
s
c
o
m
m
o
nly
k
n
o
w
n
a
s
y
e
ll
o
w
f
e
v
e
r
m
o
s
qu
i
t
o
.
I
t
i
s
i
nv
o
l
v
e
d
in
s
pr
e
a
d
i
n
g
d
e
n
gue
f
e
v
e
r
,
z
i
ka
f
e
v
e
r
,
c
hi
ku
n
gu
ny
a
,
m
a
y
a
r
o
,
y
e
l
l
o
w
f
e
v
e
r
vi
r
us
e
s
,
a
n
d
o
t
h
e
r
d
i
s
e
a
s
e
a
g
e
n
t
s
[
21]
.
A
e
de
s
a
l
b
o
p
i
c
t
us
.
T
hi
s
m
o
s
qu
i
t
o
s
pe
c
i
e
s
c
a
ll
e
d
a
s
As
i
a
n
t
i
ge
r
m
o
s
qu
i
t
o
o
r
f
o
r
e
s
t
m
o
s
qu
i
t
o
.
I
t
c
a
n
s
pr
e
a
d
d
i
s
e
a
s
e
s
i
n
c
l
ud
i
ng
y
e
ll
o
w
f
e
v
e
r
,
de
n
gue
f
e
v
e
r
,
a
n
d
c
hi
ku
n
gu
ny
a
f
e
v
e
r
[
21]
.
C
u
l
e
x
qu
i
nque
f
a
s
c
i
a
t
us
.
T
hi
s
i
s
kn
o
wn
a
s
t
h
e
S
o
u
t
h
e
r
n
h
o
us
e
m
o
s
qu
i
t
o.
I
t
i
s
a
m
e
d
i
u
m
-
s
i
z
e
d
m
o
s
qu
i
t
o
f
o
u
n
d
i
n
t
r
o
pi
c
a
l
a
n
d
s
u
b
t
r
o
pi
c
a
l
r
e
g
i
o
ns
o
f
t
h
e
wo
r
l
d.
I
t
i
s
i
m
po
r
t
a
n
t
i
n
t
r
a
n
s
mi
s
s
i
o
n
o
f
wuc
he
r
e
r
i
a
b
a
n
c
r
o
f
t
i
,
a
vi
a
n
m
a
l
a
r
i
a
,
a
n
d
a
r
b
o
vi
r
u
s
e
s
i
nc
l
ud
i
ng
v
St
,
a
n
d
l
o
u
i
s
e
n
c
e
p
h
a
li
t
i
s
vi
r
u
s
[
22]
.
F
e
a
t
ur
e
s
of
t
hi
s
da
t
a
s
e
t
a
r
e
e
x
t
r
a
c
t
e
d
by
pr
o
c
e
s
s
i
ng
o
p
t
i
c
a
l
s
i
g
na
l
by
us
i
ng
s
i
g
n
a
l
pr
o
c
e
s
s
i
ng
t
e
c
h
ni
que
s
.
T
h
e
s
e
f
e
a
t
ur
e
s
i
n
c
l
ude
w
i
ng
be
a
t
f
r
e
que
nc
y
,
v
a
r
i
o
us
s
t
a
t
i
s
t
i
c
s
f
r
o
m
t
e
m
po
r
a
l
r
e
pr
e
s
e
n
t
a
t
i
o
n
,
c
o
m
p
l
e
xi
t
y
m
e
a
s
ur
e
s
o
f
s
i
g
n
a
l
s
pe
c
t
r
um
a
n
d
s
o
on
.
T
h
e
r
e
a
r
e
t
h
r
e
e
v
a
r
i
a
t
i
o
n
s
o
f
t
h
e
s
e
da
t
a
s
e
t
s
wi
t
h
a
b
r
upt
,
gr
a
dua
l
a
n
d
r
e
c
ur
r
i
n
g
c
o
n
c
e
pt
d
r
i
f
t
s
.
A
ut
h
o
r
s
h
a
ve
ge
n
e
r
a
t
e
d
t
h
i
s
i
n
s
e
c
t
s
tr
e
a
m
da
t
a
s
e
t
s
wi
t
h
c
o
n
c
e
pt
dr
i
f
t
s
us
i
n
g
o
pt
i
c
a
l
s
e
ns
o
r
b
a
s
e
d
s
m
a
r
t
tr
a
p
f
o
r
c
a
t
c
hi
ng
t
h
e
i
ns
e
c
t
s
.
T
h
e
da
t
a
s
e
t
wi
t
h
d
i
f
f
e
r
e
n
t
c
o
n
c
e
pt
dr
i
f
t
s
i
s
ge
n
e
r
a
t
e
d
by
do
i
n
g
t
h
e
v
a
r
i
a
t
i
o
n
s
i
n
t
h
e
t
e
m
pe
r
a
t
ur
e
whi
c
h
m
a
y
a
f
f
e
c
t
o
n
t
h
e
di
s
t
r
i
b
ut
i
o
n
o
f
f
e
a
t
ur
e
s
o
f
t
h
e
i
ns
e
c
t
s
.
2.
2
.
De
s
c
r
ip
t
ion
of
p
r
op
os
e
d
m
od
e
l
I
n
t
hi
s
s
e
c
t
i
o
n
we
de
s
c
r
i
be
o
ur
p
r
o
p
o
s
e
d
m
o
de
l
w
it
h
ps
e
udo
-
c
o
de
.
T
h
e
pr
o
p
o
s
e
d
m
o
de
l
i
s
s
h
o
w
n
i
n
F
i
gur
e
1.
C
o
n
s
i
de
r
a
da
t
a
s
t
r
e
a
m
=
{
,
}
,
wh
e
r
e
is
a
n
i
np
ut
f
e
a
t
ur
e
v
e
c
to
r
a
n
d
∈
{
1
,
2
,
.
.
.
,
}
i
s
o
u
t
pu
t
v
a
r
i
a
bl
e
o
r
c
l
a
s
s
l
a
be
l
o
f
.
Ov
e
r
a
l
l
pr
o
po
s
e
d
m
o
de
l
i
s
de
s
c
r
i
be
d
i
n
ps
e
udo
-
c
o
de
un
de
r
F
i
gur
e
2
.
I
n
t
h
e
f
i
r
s
t
ph
a
s
e
,
t
h
e
m
o
de
l
i
s
t
r
a
i
n
e
d
o
n
d
y
n
a
mi
c
s
i
z
e
d
c
h
u
n
ks
.
I
n
s
t
e
a
d
o
f
f
i
x
i
ng
t
h
e
s
i
z
e
o
f
c
h
u
n
k
i
ni
t
i
a
ll
y
,
t
h
e
c
h
u
n
k
s
o
f
dy
na
m
i
c
s
i
z
e
a
r
e
f
o
r
m
e
d
by
c
o
n
s
i
de
r
i
ng
e
n
o
ugh
n
u
m
b
e
r
o
f
i
n
s
t
a
nc
e
s
f
r
o
m
a
l
l
c
l
a
s
s
e
s
o
f
t
h
e
da
t
a
s
e
t
(
l
i
ne
s
5
-
6)
.
T
h
e
s
e
c
h
u
n
k
s
a
r
e
us
e
d
f
o
r
t
r
a
i
ni
ng
t
h
e
e
n
s
e
m
bl
e
m
o
de
l
.
S
t
a
bi
li
t
y
o
f
m
o
de
l
i
s
a
c
hiev
e
d
by
m
o
ni
t
o
r
i
n
g
t
h
e
e
r
r
o
r
r
a
t
e
a
n
d
a
ppl
yi
ng
s
t
a
t
i
s
t
i
c
a
l
t
e
s
t
o
n
t
h
e
v
a
r
i
a
nc
e
s
i
n
pr
e
d
i
c
t
i
o
n
e
r
r
or
[
23]
(
l
i
ne
s
7
-
14)
.
He
r
e
,
hy
br
i
d
e
ns
e
m
b
l
e
m
o
de
l
i
s
t
r
a
i
n
e
d
us
i
ng
t
h
e
w
h
o
l
e
c
h
u
n
k,
a
t
t
h
e
s
a
m
e
t
i
m
e
o
n
e
s
pe
c
i
a
l
c
o
m
pe
t
e
n
t
c
l
a
s
s
i
f
i
e
r
is
t
r
a
i
n
e
d
us
i
ng
e
v
e
r
y
i
ns
t
a
n
c
e
f
r
o
m
t
h
e
c
h
u
n
k,
s
o
a
b
r
upt
o
r
gr
a
dua
l
dr
i
f
t
,
i
f
e
xi
s
t
s
,
c
a
n
b
e
ha
n
d
l
e
d
e
f
f
e
c
t
i
v
e
ly
.
F
o
r
h
a
n
d
li
ng
i
m
ba
l
a
n
c
e
d
i
s
s
ue
i
n
c
h
u
n
k
s
,
s
p
l
i
t
ba
s
e
d
r
e
s
a
m
p
li
ng
a
l
go
r
i
t
hm
i
s
us
e
d
whi
c
h
i
s
de
s
c
r
i
b
e
d
i
n
ps
e
udo
-
c
o
de
un
de
r
F
i
gur
e
3
.
F
o
r
t
e
s
t
da
t
a
k
-
n
e
a
r
e
s
t
n
e
i
g
hb
o
r
s
(
K
NN
)
b
a
s
e
d
d
y
na
m
i
c
e
ns
e
m
b
l
e
s
e
lec
t
i
o
n
i
s
us
e
d
(
l
i
ne
s
14
-
21)
.
F
i
gur
e
1.
F
r
a
m
e
wo
r
k
f
o
r
p
r
o
p
o
s
e
d
m
o
de
l
T
o
de
a
l
w
i
t
h
i
m
ba
l
a
n
c
e
i
s
s
ue
i
n
da
t
a
s
t
r
e
a
m
we
ha
v
e
im
p
l
e
m
e
n
t
e
d
s
p
l
i
t
b
a
gg
i
ng
t
e
c
hni
que
w
hi
c
h
i
s
pr
e
s
e
n
t
e
d
i
n
p
s
e
udo
-
c
o
de
un
de
r
F
i
gur
e
3
.
I
n
m
u
l
t
i
-
c
l
a
s
s
i
m
ba
l
a
nc
e
da
t
a
s
e
t
,
a
l
o
n
g
w
i
t
h
d
y
na
mi
c
im
ba
l
a
n
c
e
r
a
t
i
o
o
f
c
l
a
s
s
e
s
,
t
h
e
r
e
a
r
e
ot
h
e
r
da
t
a
di
f
f
i
c
u
l
t
y
f
a
c
to
r
s
l
i
ke
n
o
i
s
y
mi
n
o
r
i
t
y
i
ns
t
a
n
c
e
s
a
n
d
o
v
e
r
l
a
pp
i
n
g
c
l
a
s
s
e
s
whi
c
h
de
t
e
r
i
o
r
a
t
e
s
t
h
e
pe
r
f
o
r
m
a
n
c
e
o
f
c
l
a
s
s
i
f
i
e
r
s
,
a
l
t
h
o
ugh
we
ge
n
e
r
a
t
e
d
t
h
e
b
a
l
a
nc
e
d
da
t
a
s
e
t.
I
n
a
d
d
i
t
i
o
n
to
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2502
-
4752
I
n
do
n
e
s
i
a
n
J
E
l
e
c
E
n
g
&
C
o
m
p
S
c
i
,
Vo
l
.
25
,
N
o
.
2
,
F
e
b
r
ua
r
y
20
22
:
1115
-
112
2
1118
t
hi
s
d
i
ve
r
s
i
t
y
o
f
t
r
a
i
ni
ng
s
u
bs
e
t
s
i
s
a
l
s
o
c
r
uc
i
a
l
f
o
r
i
m
pr
o
vi
n
g
t
h
e
pe
r
f
o
r
m
a
nc
e
o
f
c
l
a
s
s
if
i
e
r
.
P
r
o
p
o
s
e
d
a
l
go
r
i
t
hm
h
a
s
s
y
s
t
e
m
a
t
i
c
a
ll
y
c
r
e
a
t
e
d
n
u
m
be
r
o
f
s
u
bs
e
t
s
by
s
p
l
i
t
t
i
n
g
t
h
e
c
l
a
s
s
e
s
i
n
t
o
pa
r
t
i
t
i
o
n
s
b
a
s
e
d
upo
n
t
he
s
i
z
e
o
f
mi
n
o
r
i
t
y
c
l
a
s
s
a
n
d
e
ve
r
y
pa
r
t
i
t
i
o
n
i
s
f
il
l
e
d
w
i
t
h
t
he
e
qua
l
n
u
m
be
r
o
f
i
ns
t
a
n
c
e
s
f
r
o
m
m
a
j
o
r
i
t
y
c
l
a
s
s
whi
c
h
a
r
e
s
e
l
e
c
t
e
d
us
i
n
g
ne
ga
t
i
v
e
bi
n
o
mi
a
l
d
i
s
t
r
i
b
ut
i
o
n
[
24]
us
i
n
g
a
s
(
1)
(
l
i
ne
s
1
-
13)
a
n
d
s
y
n
t
h
e
t
i
c
mi
n
o
r
i
t
y
e
x
a
m
p
l
e
s
.
(
|
)
=
(
+
−
1
)
(
1)
W
h
e
r
e
,
m
=
n
u
m
be
r
f
a
i
l
ur
e
s
f
o
r
gi
ve
n
n
=
n
u
m
be
r
o
f
s
uc
c
e
s
s
e
s
a
n
d
p
=
q
=
0
.
5
.
W
hil
e
ge
n
e
r
a
t
i
n
g
s
y
n
t
he
t
i
c
mi
n
o
r
i
t
y
i
ns
t
a
n
c
e
s
i
ns
t
e
a
d
o
f
ge
ne
r
a
t
i
n
g
r
a
n
do
m
ly
,
s
y
n
t
h
e
t
i
c
e
x
a
m
p
l
e
s
a
r
e
ge
ne
r
a
t
e
d
o
nl
y
f
r
o
m
t
h
e
s
a
f
e
mi
n
o
r
i
t
y
s
a
m
p
l
e
s
a
n
d
us
i
ng
t
h
e
n
e
a
r
e
s
t
n
e
i
g
hb
o
r
s
f
o
r
e
v
e
r
y
c
l
a
s
s
.
S
o
,
t
h
e
s
y
n
t
h
e
t
i
c
e
xa
m
p
l
e
s
w
hi
c
h
w
i
ll
be
ge
n
e
r
a
t
e
d
w
i
l
l
n
o
t
c
o
n
t
a
i
n
a
ny
n
o
i
s
y
o
r
h
a
r
d
to
l
e
a
r
n
i
n
s
t
a
nc
e
s
(
l
i
ne
s
14
-
15)
.
T
r
a
i
n
t
h
e
c
l
a
s
s
if
i
e
r
s
i
n
e
n
s
e
m
ble
m
o
de
l
u
s
i
ng
ba
l
a
n
c
e
d
s
u
b
s
e
t
s
(
l
i
ne
s
16
-
18)
.
F
i
gur
e
2
.
P
s
e
udo
c
o
de
f
o
r
HD
C
E
M
F
i
gur
e
3
.
P
s
e
udo
c
o
de
f
o
r
s
pl
i
t
ba
gg
i
n
g
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
do
n
e
s
i
a
n
J
E
l
e
c
E
n
g
&
C
o
m
p
S
c
i
I
S
S
N:
2502
-
4752
Hy
br
id
dy
namic
c
hunk
e
ns
e
mble
mode
l
f
or
multi
-
c
las
s
data
s
tr
e
ams
(
V
a
r
s
ha
Sac
hin
K
hande
k
ar
)
1119
P
r
o
p
o
s
e
d
m
o
de
l
us
e
s
d
y
na
mi
c
e
n
s
e
m
bl
e
s
e
l
e
c
t
i
o
n
f
o
r
c
l
a
s
s
if
i
c
a
t
i
o
n
o
f
t
e
s
t
da
t
a
.
P
s
e
udo
-
c
o
de
f
o
r
K
NN
-
d
y
na
mi
c
e
n
s
e
m
b
l
e
s
e
l
e
c
t
i
o
n
(
K
NN
-
DE
S
)
i
s
s
h
o
w
n
i
n
F
i
gur
e
4
.
He
r
e
f
o
r
e
v
e
r
y
i
ns
t
a
n
c
e
f
r
o
m
tes
t
da
t
a
,
r
e
g
i
o
n
o
f
c
o
m
p
e
t
e
n
c
e
i
s
c
o
m
put
e
d
b
a
s
e
d
o
n
k
-
n
e
a
r
e
s
t
n
e
i
g
h
b
o
r
s
f
r
o
m
t
r
a
i
ni
ng
d
a
t
a
s
e
t
.
Onl
y
t
h
o
s
e
k
c
l
a
s
s
if
i
e
r
s
o
f
s
e
l
e
c
t
e
d
k
-
n
e
a
r
e
s
t
n
e
i
g
hb
o
r
s
a
r
e
us
e
d
f
o
r
de
c
i
d
in
g
t
h
e
l
a
b
e
l
o
f
t
e
s
t
i
n
s
t
a
n
c
e
by
m
a
xim
i
z
i
ng
t
h
e
pr
o
b
a
bil
i
t
y
o
f
pr
e
d
i
c
t
i
o
n
o
f
t
h
a
t
r
e
s
pe
c
t
i
ve
c
l
a
s
s
l
a
b
e
l
.
F
i
gur
e
4.
P
s
e
udo
c
o
de
f
o
r
K
NN
-
DE
S
3.
RE
S
UL
T
S
A
ND
D
IS
CU
S
S
I
ON
I
n
t
hi
s
s
e
c
t
i
o
n
,
we
h
a
v
e
do
n
e
c
o
m
pa
r
a
t
i
v
e
a
n
a
ly
s
is
o
f
pr
o
p
o
s
e
d
m
o
de
l
HD
C
E
M
-
s
p
li
t
b
a
gg
i
ng
w
i
t
h
HD
C
E
M
-
S
M
OT
E
b
a
gg
i
n
g
,
HD
C
E
M
-
o
v
e
r
b
a
gg
i
n
g
a
n
d
HD
C
E
M
-
un
de
r
o
v
e
r
b
a
gg
i
ng
.
W
e
i
n
t
e
n
t
to
v
e
r
i
f
y
t
h
e
pot
e
n
c
y
o
f
t
h
e
pr
o
p
o
s
e
d
s
p
l
i
t
b
a
gg
i
ng
i
n
m
u
l
t
i
-
c
las
s
i
m
ba
l
a
nc
e
d
da
t
a
s
t
r
e
a
m
s
.
W
h
e
n
da
t
a
i
s
i
m
ba
lan
c
e
d
f
o
r
a
s
s
e
s
s
i
ng
t
h
e
pe
r
f
o
r
m
a
n
c
e
o
f
m
o
de
l
r
a
t
h
e
r
t
h
a
n
us
i
n
g
a
c
c
ur
a
c
y
F
-
m
e
a
ur
e
,
pr
e
c
i
s
i
o
n
m
i
n
o
r
i
t
y
c
l
a
s
s
r
e
c
a
l
l
pe
r
f
o
r
m
a
n
c
e
m
e
a
s
ur
e
s
a
r
e
us
e
d.
T
h
e
pre
c
i
s
i
o
n
,
a
l
s
o
c
a
l
l
e
d
a
s
T
rue
po
s
i
t
i
ve
ra
t
e
o
r
s
pe
c
i
fi
c
i
t
y
,
i
s
t
he
r
a
t
i
o
o
f
c
o
r
r
e
c
t
l
y
pr
e
d
i
c
t
e
d
po
s
i
t
i
ve
i
ns
t
a
n
c
e
s
to
tot
a
l
pr
e
d
ict
e
d
p
o
s
i
t
i
v
e
i
ns
t
a
n
c
e
s
.
t
h
e
r
e
c
a
l
l
,
a
l
s
o
c
a
ll
e
d
a
s
s
e
n
s
i
t
i
v
i
t
y
,
i
s
t
h
e
r
a
t
i
o
o
f
tr
ue
p
o
s
i
t
i
v
e
i
ns
t
a
n
c
e
s
t
o
a
c
t
ua
l
po
s
i
t
i
v
e
i
ns
t
a
n
c
e
s
.
F
1
s
c
o
r
e
o
r
F
-
m
e
a
s
ur
e
,
i
s
we
i
g
h
t
e
d
a
v
e
r
a
ge
o
f
pr
e
c
i
s
i
o
n
a
n
d
r
e
c
a
ll
.
T
h
e
s
e
pe
r
f
o
r
m
a
n
c
e
m
e
a
s
ur
e
s
a
r
e
de
f
i
ne
d
a
s
(
2)
,
(
3)
,
a
n
d
(
4)
.
P
r
e
cis
io
n
=
TP
TP
+
FP
(
2
)
Re
c
al
l
=
TP
TP
+
FN
(
3
)
F
−
M
e
a
s
u
r
e
=
2
∗
(
P
r
e
c
i
s
i
on
∗
R
e
c
a
ll
)
(
P
r
e
c
i
s
i
on
+
R
e
c
a
ll
)
(
4
)
3.
1
.
E
x
p
e
r
im
e
n
t
a
l
r
e
s
u
l
t
s
T
h
e
F
-
m
e
a
s
ur
e
a
n
d
mi
n
o
r
i
t
y
c
l
a
s
s
r
e
c
a
ll
f
o
r
i
m
b
a
l
a
n
c
e
d
i
ns
e
c
t
da
t
a
s
tr
e
a
m
s
w
i
t
h
a
b
r
upt
,
gr
a
dua
l
c
o
n
c
e
pt
dr
i
f
t
s
a
t
di
f
f
e
r
e
n
t
c
h
un
k
s
i
z
e
a
r
e
s
h
o
wn
us
i
ng
f
o
l
l
o
w
i
ng
gr
a
phs
.
F
r
o
m
F
i
gur
e
s
5
(
a
)
a
n
d
(
b
)
i
t
i
s
c
l
e
a
r
t
h
a
t
t
h
e
pr
o
p
o
s
e
d
HD
C
E
M
w
i
t
h
s
p
li
t
b
a
gg
i
ng
m
o
d
e
l
h
a
s
g
i
ve
n
s
t
a
bl
e
p
e
r
f
o
r
m
a
n
c
e
f
o
r
b
o
t
h
a
b
r
upt
a
nd
gr
a
dua
l
dr
i
f
t
i
n
s
e
c
t
da
t
a
s
t
r
e
a
m
s
a
n
d
o
ut
pe
r
f
o
r
m
e
d
o
v
e
r
S
M
OT
E
b
a
gg
i
ng
,
o
v
e
r
b
a
gg
i
ng
a
n
d
u
n
de
r
o
v
e
r
b
a
g
g
i
n
g
f
o
r
d
i
f
f
e
r
e
n
t
c
h
u
n
k
s
i
z
e
s
.
M
i
n
o
r
i
t
y
c
l
a
s
s
r
e
c
a
l
l
us
i
ng
p
r
o
p
o
s
e
d
m
o
de
l
i
s
b
e
t
t
e
r
t
h
a
n
ot
h
e
r
t
e
c
hni
qu
e
s
.
Av
e
r
a
ge
mi
n
o
r
i
t
y
c
l
a
s
s
r
e
c
a
ll
i
n
a
b
r
upt
dr
i
f
t
i
ns
e
c
t
da
t
a
s
t
r
e
a
m
i
s
78%
a
n
d
71%
i
n
gr
a
dua
l
dr
i
f
t
i
ns
e
c
t
da
t
a
s
t
r
e
a
m
.
I
n
a
b
r
upt
dr
i
f
t
,
t
h
e
a
c
c
ur
a
c
y
i
s
m
o
r
e
b
e
c
a
us
e
t
h
e
r
e
i
s
o
n
e
de
d
i
c
a
t
e
d
c
o
m
po
n
e
n
t
c
l
a
s
s
if
i
e
r
us
e
d
whi
c
h
l
e
a
r
n
s
e
v
e
r
y
i
ns
t
a
n
c
e
f
r
o
m
c
ur
r
e
n
t
c
h
u
n
k.
W
hil
e
f
o
r
gr
a
dua
l
dr
i
f
t
i
ns
e
c
t
da
t
a
s
t
r
e
a
m
,
a
l
t
h
o
ugh
c
hun
k
s
i
z
e
i
s
m
o
r
e
a
s
c
o
m
p
a
r
e
d
to
a
b
r
upt
dr
i
f
t
da
t
a
s
t
r
e
a
m
c
hun
k
s
i
z
e
,
m
i
n
o
r
i
t
y
c
l
a
s
s
r
e
c
a
ll
h
a
s
n
o
t
i
m
pr
o
v
e
d.
R
e
a
s
o
n
b
e
hi
nd
t
hi
s
i
s
t
h
a
t
t
h
e
r
e
m
i
g
h
t
b
e
i
n
c
r
e
a
s
e
d
n
u
m
be
r
o
f
h
a
r
d
t
o
l
e
a
r
n
i
ns
t
a
n
c
e
s
o
r
i
nc
r
e
a
s
e
d
im
ba
l
a
n
c
e
r
a
t
i
o
.
A
ve
r
a
ge
o
v
e
r
a
l
l
a
c
c
ur
a
c
y
a
c
hi
e
v
e
d
i
s
91%
,
b
ut
due
to
l
im
i
t
e
d
s
pa
c
e
gr
a
phi
c
a
l
a
n
a
lys
i
s
i
s
n
o
t
s
h
o
w
n
h
e
r
e
.
T
o
s
upp
o
r
t
t
hi
s
r
e
s
u
l
t
,
we
h
a
v
e
a
pp
l
ied
n
o
n
pa
r
a
m
e
t
r
i
c
M
a
nn
-
W
hi
t
ne
y
’
s
U
s
t
a
t
i
s
t
i
c
a
l
t
e
s
t
[
25
]
.
R
e
s
u
l
t
s
o
f
t
hi
s
s
t
a
t
i
s
t
i
c
a
l
t
e
s
t
i
n
t
h
e
f
o
r
m
o
f
a
v
e
r
a
g
e
r
a
n
k
v
a
l
ue
s
(
R
+
)
a
n
d
(
R
-
)
b
e
t
we
e
n
pr
o
p
o
s
e
d
m
o
de
l
w
i
t
h
s
p
l
i
t
ba
gg
i
ng
a
n
d
w
i
t
h
o
t
h
e
r
b
a
gg
i
n
g
t
e
c
hni
qu
e
s
a
r
e
s
h
o
wn
i
n
T
a
bl
e
1.
T
a
bl
e
1
de
pi
c
t
s
t
h
a
t
t
h
e
pr
o
p
o
s
e
d
m
o
de
l
h
a
s
o
u
t
f
o
r
m
e
d.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2502
-
4752
I
n
do
n
e
s
i
a
n
J
E
l
e
c
E
n
g
&
C
o
m
p
S
c
i
,
Vo
l
.
25
,
N
o
.
2
,
F
e
b
r
ua
r
y
20
22
:
1115
-
112
2
1120
(
a
)
(
b
)
F
i
gur
e
5.
C
o
m
pa
r
a
t
i
v
e
a
n
a
ly
s
i
s
u
s
i
ng
F
-
m
e
a
s
ur
e
a
n
d
m
i
n
o
r
i
t
y
c
l
a
s
s
r
e
c
a
ll
f
o
r
(
a
)
i
ns
e
c
t
da
t
a
s
t
r
e
a
m
w
i
t
h
a
b
r
upt
dr
i
f
t
a
n
d
(
b
)
i
ns
e
c
t
da
t
a
s
tr
e
a
m
w
i
t
h
gr
a
dua
l
dr
i
f
t
T
a
bl
e
1.
M
a
nn
-
W
hi
t
n
e
y
’
s
U
s
t
a
t
i
s
t
i
c
a
l
t
e
s
t
r
e
s
ul
t
s
C
o
mpa
r
is
o
n
F
-
M
e
a
s
ur
e
M
in
o
r
it
y
c
la
s
s
r
e
c
a
ll
P
-
v
a
lu
e
(
R
+
)
(R
-
)
P
-
v
a
lu
e
(
R
+
)
(R
-
)
H
D
C
E
M
-
s
pl
it
ba
ggi
ng
V
s
H
D
C
E
M
-
S
M
O
T
E
ba
ggi
ng
<
0.01*
166
87
<
0.01*
161
92
H
D
C
E
M
-
s
pl
it
ba
ggi
ng
V
s
H
D
C
E
M
-
ove
r
ba
ggi
ng
<
0.01*
181
72
<
0.01*
183
70
H
D
C
E
M
-
s
pl
it
ba
ggi
ng
V
s
H
D
C
E
M
-
und
e
r
-
ove
r
ba
ggi
ng
<
0.01*
143
110
<
0.01*
142
111
3.
2
.
Com
p
u
t
at
io
n
al
t
im
e
c
om
p
l
e
x
it
y
F
i
gur
e
6
s
h
o
ws
t
h
e
t
i
m
e
r
e
qu
i
r
e
d
f
o
r
c
h
u
n
k
pr
oc
e
s
s
i
ng.
A
s
t
h
e
i
ns
e
c
t
da
t
a
s
t
r
e
a
m
i
s
m
u
l
t
i
-
c
l
a
s
s
,
c
o
m
put
a
t
i
o
n
a
l
c
o
s
t
o
f
pr
o
p
o
s
e
d
m
o
de
l
i
s
hi
g
h
due
to
b
a
l
a
n
c
i
ng
t
e
c
hni
que
us
e
d
f
o
r
ge
ne
r
a
t
i
n
g
d
i
ve
r
s
e
b
a
l
a
n
c
e
d
s
ubs
e
t
s
w
i
t
h
ha
n
d
l
i
ng
o
f
s
o
m
e
o
f
t
h
e
da
t
a
di
f
f
i
c
u
l
t
y
f
a
c
to
r
s
.
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
F
-
m
e
as
u
r
e
Ch
u
n
k
s
i
z
e
H
D
C
E
M
-
S
pl
it
B
a
ggi
ng
H
D
C
E
M
-
S
mot
e
B
a
ggi
ng
H
D
C
E
M
-
O
ve
r
B
a
ggi
ng
H
D
C
E
M
-
U
nde
r
O
ve
r
B
a
ggi
ng
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
M
in
o
r
it
y
c
la
s
s
r
e
c
al
l
C
h
u
n
k
s
iz
e
H
D
C
E
M
-
S
pl
it
B
a
ggi
ng
H
D
C
E
M
-
S
mot
e
B
a
ggi
ng
H
D
C
E
M
-
O
ve
r
B
a
ggi
ng
H
D
C
E
M
-
U
nde
r
O
ve
r
B
a
ggi
ng
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
F
-
m
e
as
u
r
e
C
h
u
n
k
s
iz
e
H
D
C
E
M
-
S
pl
it
B
a
ggi
ng
H
D
C
E
M
-
S
M
O
T
E
B
a
ggi
ng
H
D
C
E
M
-
O
ve
r
B
a
ggi
ng
H
D
C
E
M
-
U
nde
r
O
ve
r
B
a
ggi
ng
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
M
i
n
o
r
i
t
y
c
l
as
s
r
e
c
al
l
C
h
u
n
k
s
iz
e
H
D
C
E
M
-
S
pl
it
B
a
ggi
ng
H
D
C
E
M
-
S
M
O
T
E
B
a
ggi
ng
H
D
C
E
M
-
O
ve
r
B
a
ggi
ng
H
D
C
E
M
-
U
nde
r
O
ve
r
B
a
ggi
ng
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
do
n
e
s
i
a
n
J
E
l
e
c
E
n
g
&
C
o
m
p
S
c
i
I
S
S
N:
2502
-
4752
Hy
br
id
dy
namic
c
hunk
e
ns
e
mble
mode
l
f
or
multi
-
c
las
s
data
s
tr
e
ams
(
V
a
r
s
ha
Sac
hin
K
hande
k
ar
)
1121
F
i
gur
e
6
.
C
o
m
put
a
t
i
o
n
a
l
t
i
m
e
f
o
r
c
h
un
k
pr
o
c
e
s
s
i
n
g
4.
CONC
L
USI
ON
F
o
r
t
h
e
c
l
a
s
s
if
i
c
a
t
i
o
n
o
f
da
t
a
s
t
r
e
a
m
s
,
c
o
n
c
e
pt
dr
i
f
t
a
n
d
i
m
ba
l
a
n
c
e
i
s
s
u
e
s
a
r
e
t
h
e
m
a
j
o
r
pr
o
bl
e
m
s
.
I
n
t
hi
s
pa
pe
r
we
h
a
v
e
pr
o
po
s
e
d
HD
C
E
M
f
o
r
m
u
l
t
i
-
c
l
a
s
s
da
t
a
s
t
r
e
a
m
t
o
de
a
l
w
i
t
h
t
h
e
s
e
i
s
s
ue
s
.
HD
C
E
M
ge
n
e
r
a
t
e
s
a
n
e
ns
e
m
b
l
e
m
o
de
l
whi
c
h
i
s
tr
a
i
ne
d
o
n
da
t
a
c
h
un
k
s
wh
o
s
e
s
i
z
e
i
s
de
c
i
de
d
d
y
n
a
mi
c
a
ll
y
r
a
t
h
e
r
t
h
a
n
f
ixi
ng
i
t
a
pr
i
o
r
.
Al
s
o
,
f
o
r
h
a
n
d
li
ng
d
y
n
a
mi
c
i
m
ba
l
a
n
c
e
i
s
s
ue
we
h
a
v
e
pr
o
p
o
s
e
d
S
pl
i
t
b
a
s
e
d
B
a
gg
i
n
g
a
l
go
r
i
t
hm
w
hi
c
h
c
a
n
h
a
n
d
l
e
n
o
i
s
y
,
ha
r
d
to
l
e
a
r
n
mi
n
o
r
i
t
y
a
n
d
m
a
j
o
r
i
t
y
in
s
t
a
n
c
e
s
pr
e
s
e
n
t
i
n
t
h
e
da
t
a
s
e
t
.
I
n
a
dd
i
t
i
o
n
to
t
hi
s
,
i
n
s
t
e
a
d
o
f
a
pp
l
yi
ng
d
i
r
e
c
t
m
a
j
o
r
i
t
y
v
o
t
i
n
g
e
n
s
e
m
b
l
e
a
l
go
r
i
t
hm
f
o
r
t
e
s
t
da
t
a
pr
e
di
c
t
i
o
n
,
k
-
n
e
a
r
e
s
t
n
e
i
g
hb
o
r
b
a
s
e
d
d
y
n
a
mi
c
e
ns
e
m
b
l
e
s
e
l
e
c
t
i
o
n
i
s
us
e
d.
E
x
pe
r
i
m
e
n
t
a
l
r
e
s
u
l
t
s
s
h
o
we
d
t
h
a
t
pr
o
p
o
s
e
d
m
o
de
l
h
a
s
o
u
t
pe
r
f
o
r
m
e
d,
b
ut
i
t
i
s
c
o
m
put
a
t
i
o
n
a
ll
y
e
x
pe
ns
i
ve
.
T
h
e
t
i
m
e
r
e
qu
i
r
e
m
e
n
t
f
o
r
p
r
o
c
e
s
s
i
n
g
m
u
l
t
i
p
l
e
c
l
a
s
s
e
s
i
n
da
t
a
s
t
r
e
a
m
i
s
m
o
r
e
s
o
f
o
r
f
ut
ur
e
wor
k
we
c
a
n
i
m
p
l
e
m
e
n
t
p
r
o
p
o
s
e
d
m
o
de
l
us
in
g
d
i
s
t
r
i
b
ut
e
d
e
n
vi
r
o
nm
e
n
t
pl
a
t
f
o
r
m
s
l
i
ke
Ha
do
o
p
o
r
S
pa
r
k
.
AC
K
NOWL
E
DGE
M
E
NT
S
A
ut
h
o
r
s
t
h
a
n
ks
t
o
De
pa
r
t
m
e
n
t
o
f
C
o
m
put
e
r
E
n
g
i
n
e
e
r
i
ng,
M
uke
s
h
P
a
t
e
l
S
c
h
o
o
l
o
f
T
e
c
hn
o
l
o
g
y
M
a
n
a
ge
m
e
n
t
a
n
d
E
n
g
i
n
e
e
r
i
n
g
M
u
m
ba
i
,
I
n
d
i
a
,
a
nd
S
m
t
.
K
a
s
hi
ba
i
Na
v
a
l
e
C
o
l
l
e
ge
o
f
E
n
g
i
ne
e
r
i
ng,
P
un
e
f
o
r
pr
o
vi
d
i
n
g
a
n
i
nf
r
a
s
t
r
uc
t
ur
e
a
n
d
s
uppo
r
t
to
c
a
r
r
y
r
e
s
e
a
r
c
h
o
n
a
b
o
ve
m
e
n
t
i
o
n
e
d
t
o
pi
c
.
RE
F
E
R
E
NC
E
S
[
1]
H
. M
.
G
o
m
e
s
, J
. P
. B
a
r
dda
l,
A
.
F
. E
n
e
mbr
e
c
k, a
nd A
.
B
i
f
e
t,
“
A
s
ur
ve
y
o
n
e
ns
e
mb
le
le
a
r
ni
ng
f
o
r
da
ta
s
tr
e
a
m
c
la
s
s
if
i
c
a
ti
o
n,”
A
C
M
C
om
put
in
g
Sur
v
e
y
s
, v
o
l.
50, n
o
. 2, pp. 1
–
36, M
a
r
. 2017, d
o
i:
10
.1145/3054925.
[
2]
B
.
K
r
a
w
c
z
y
k,
L
.
L
.
M
in
ku,
J
.
G
a
ma
,
J
.
S
t
e
f
a
n
o
w
s
ki
,
a
nd
M
.
W
oź
ni
a
k,
“
E
ns
e
mb
le
l
e
a
r
ni
ng
f
or
da
ta
s
tr
e
a
m
a
na
l
y
s
is
:
A
s
ur
ve
y
,”
I
nf
or
m
at
io
n F
us
io
n
, vo
l.
37, pp. 132
–
156,
S
e
p. 2017, d
o
i
:
10.
1
016/
j.
in
f
f
us
.2017.02.004.
[
3]
S
.
W
a
ng,
L
.
L
.
M
in
ku,
a
nd
X
.
Y
a
o
,
“
A
S
y
s
te
ma
ti
c
S
tu
d
y
of
O
nl
in
e
C
la
s
s
I
mba
la
nc
e
L
e
a
r
n
in
g
w
it
h
C
o
nc
e
pt
D
r
i
f
t,
”
I
E
E
E
T
r
ans
ac
ti
ons
on
N
e
u
r
al
N
e
tw
or
k
s
and
L
e
ar
ni
ng
Sy
s
t
e
m
s
,
v
ol
.
29,
n
o
.
10,
pp.
4802
–
4821,
O
c
t.
2018,
do
i:
10.1109/
T
N
N
L
S
.2017.2771290.
[
4]
H
.
W
a
ng
a
nd
Z
.
A
br
a
h
a
m,
“
C
o
nc
e
pt
dr
i
f
t
d
e
t
e
c
t
i
o
n
f
or
s
tr
e
a
mi
ng
da
ta
,”
in
P
r
oc
e
e
di
ngs
o
f
th
e
I
nt
e
r
nat
io
nal
J
oi
nt
C
on
f
e
r
e
nc
e
on
N
e
ur
al
N
e
tw
or
k
s
, J
ul
. 2015, v
o
l.
2015
-
S
e
pt
e
mb
e
r
, d
oi
:
10.1109
/I
J
C
N
N
.2015.7280398.
[
5]
D
.
B
r
z
e
z
in
s
ki
a
nd
J
.
S
te
f
a
n
o
w
s
ki
,
“
P
r
e
qu
e
nt
ia
l
A
U
C
f
or
c
la
s
s
if
i
e
r
e
v
a
lu
a
ti
o
n
a
nd
dr
i
f
t
d
e
t
e
c
ti
o
n
in
e
vo
l
v
in
g
da
ta
s
tr
e
a
ms
,”
in
L
e
c
tu
r
e
N
ot
e
s
in
A
r
ti
f
ic
ia
l
I
nt
e
ll
ig
e
nc
e
(
Subs
e
r
i
e
s
o
f
L
e
c
tu
r
e
N
ot
e
s
in
C
om
put
e
r
Sc
ie
nc
e
)
,
v
o
l.
8983,
S
pr
in
ge
r
I
nt
e
r
na
ti
ona
l
P
ubl
is
hi
ng, 2015, pp. 87
–
101.
[
6]
S
.
W
a
ng,
L
.
L
.
M
in
ku,
a
nd
X
.
Y
a
o
,
“
R
e
s
a
mpl
in
g
-
ba
s
e
d
e
ns
e
mbl
e
me
th
o
ds
f
o
r
o
nl
in
e
c
la
s
s
im
ba
la
nc
e
le
a
r
ni
ng,”
I
E
E
E
T
r
ans
ac
ti
ons
on K
now
le
dge
and Data E
ngi
ne
e
r
in
g
, vo
l.
27, n
o
. 5, pp. 1356
–
1368, M
a
y
2015, d
o
i
:
10.1109/
T
K
D
E
.2014.2345380.
[
7]
S
.
W
a
ng,
L
.
L
.
M
in
ku,
D
.
G
h
e
z
z
i,
D
.
C
a
lt
a
bi
a
no
,
P
.
T
in
o
,
a
nd
X
.
Y
a
o
,
“
C
o
n
c
e
pt
d
r
i
f
t
d
e
t
e
c
ti
o
n
f
or
o
nl
in
e
c
la
s
s
im
ba
la
nc
e
le
a
r
n
in
g,”
A
ug. 2013, do
i:
10.1109/
I
J
C
N
N
.2013.6706768.
[
8]
K
.
W
u,
A
.
E
dw
a
r
ds
,
W
.
F
a
n,
J
.
G
a
o
,
a
nd
K
.
Z
ha
ng,
“
C
la
s
s
if
y
i
ng
im
ba
la
nc
e
d
da
ta
s
tr
e
a
ms
v
ia
d
y
na
mi
c
f
e
a
tu
r
e
gr
o
up
w
e
ig
ht
in
g
w
it
h
im
po
r
ta
n
c
e
s
a
mpl
in
g,”
in
SI
A
M
I
nt
e
r
nat
io
nal
C
on
f
e
r
e
nc
e
on
D
at
a
M
in
in
g
2014,
SD
M
2014
,
A
pr
.
2014,
v
o
l.
2,
pp.
722
–
7
30,
do
i:
10.1137/1.9781611
973440.83.
[
9]
Y
.
L
u,
Y
.
M
.
C
he
ung,
a
nd
Y
.
Y
a
n
T
a
ng,
“
A
da
pt
i
ve
c
hunk
-
ba
s
e
d
d
y
na
mi
c
w
e
ig
h
te
d
ma
j
o
r
it
y
f
or
i
mba
la
nc
e
d
da
ta
s
tr
e
a
ms
w
it
h
c
o
n
c
e
pt
dr
i
f
t,
”
I
E
E
E
T
r
ans
ac
ti
ons
on
N
e
ur
al
N
e
tw
or
k
s
and
L
e
ar
ni
ng
Sy
s
te
m
s
,
v
o
l.
31,
n
o
.
8,
pp.
2764
–
2778,
A
ug.
2020,
do
i:
10.1109/
T
N
N
L
S
.2019.2951814.
[
10]
G
.
D
it
z
l
e
r
,
M
.
R
ove
r
i,
C
.
A
li
ppi
,
a
nd
R
.
P
o
li
ka
r
,
“
L
e
a
r
ni
ng
i
n
N
o
ns
ta
ti
o
na
r
y
E
n
v
ir
o
n
me
nt
s
:
A
S
ur
ve
y
,”
I
E
E
E
C
om
put
at
io
nal
I
nt
e
ll
ig
e
nc
e
M
agaz
in
e
, vo
l.
10, n
o
. 4, pp. 12
–
25, N
ov
. 2015, d
oi
:
10.1109/M
C
I
.2015.2471196.
[
11]
J
.
G
a
o
,
W
.
F
a
n,
J
.
H
a
n,
a
nd
P
.
S
.
Y
u,
“
A
ge
ne
r
a
l
f
r
a
m
e
w
or
k
f
or
mi
ni
ng
c
o
n
c
e
pt
-
d
r
i
f
ti
ng
da
ta
s
tr
e
a
ms
w
it
h
s
ke
w
e
d
di
s
tr
ib
ut
i
ons
,”
in
P
r
oc
e
e
di
ngs
of
t
he
7t
h
SI
A
M
I
nt
e
r
nat
io
nal
C
onf
e
r
e
nc
e
on D
at
a M
in
in
g
, A
pr
. 2007, pp. 3
–
14, do
i:
10.1137/1.9781611
972771.1.
[
12]
S
.
C
he
n
a
nd
H
.
H
e
,
“
S
E
R
A
:
S
e
l
e
c
ti
ve
l
y
r
e
c
ur
s
i
ve
a
ppr
o
a
c
h
t
o
w
a
r
ds
no
ns
ta
ti
o
na
r
y
im
ba
la
nc
e
d
s
tr
e
a
m
da
ta
mi
ni
ng,”
in
P
r
oc
e
e
di
ngs
of
t
he
I
nt
e
r
nat
io
nal
J
oi
nt
C
on
f
e
r
e
nc
e
on N
e
u
r
al
N
e
tw
or
k
s
,
J
un.
2009, pp. 522
–
529, do
i:
10.1109/
I
J
C
N
N
.2009.5178874.
[
13]
H
.
H
e
a
nd
S
.
C
he
n,
“
T
o
w
a
r
ds
in
c
r
e
m
e
nt
a
l
le
a
r
ni
ng
of
n
o
ns
ta
ti
o
na
r
y
im
ba
la
n
c
e
d
da
ta
s
tr
e
a
m:
A
mul
ti
pl
e
s
e
le
c
ti
ve
l
y
r
e
c
ur
s
iv
e
a
ppr
o
a
c
h,”
E
v
ol
v
in
g Sy
s
te
m
s
, v
o
l.
2, n
o
. 1, pp. 35
–
50, N
ov
. 201
1, do
i:
10.1007/s
12530
-
010
-
9021
-
y.
0
200
400
600
800
1000
1200
1400
1600
1800
T
im
e
in
(
s
)
C
h
u
n
k
S
iz
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2502
-
4752
I
n
do
n
e
s
i
a
n
J
E
l
e
c
E
n
g
&
C
o
m
p
S
c
i
,
Vo
l
.
25
,
N
o
.
2
,
F
e
b
r
ua
r
y
20
22
:
1115
-
112
2
1122
[
14]
G
.
D
it
z
le
r
a
nd
R
.
P
o
li
ka
r
,
“
I
nc
r
e
m
e
nt
a
l
le
a
r
ni
ng
of
c
o
n
c
e
pt
dr
if
t
f
r
o
m
s
tr
e
a
mi
ng
im
ba
la
n
c
e
d
da
ta
,”
I
E
E
E
T
r
ans
ac
ti
ons
on
K
now
le
dge
and Data E
ngi
ne
e
r
in
g
, vo
l.
25, n
o
. 10, pp. 2283
–
23
01, Oc
t.
2013, d
o
i:
10.1109/
T
K
D
E
.2012.136.
[
15]
R
.
E
lwe
ll
a
nd
R
.
P
o
li
ka
r
,
“
I
nc
r
e
me
nt
a
l
l
e
a
r
ni
ng
of
c
o
n
c
e
p
t
dr
if
t
in
no
ns
ta
ti
o
na
r
y
e
n
v
ir
o
nm
e
nt
s
,”
I
E
E
E
T
r
ans
ac
ti
ons
on
N
e
ur
al
N
e
tw
or
k
s
, v
o
l.
22, n
o
. 10, pp. 1517
–
1531, O
c
t.
2011, d
o
i:
10.11
09/
T
N
N
.2011.2160459.
[
16]
B
.
M
ir
z
a
,
Z
.
L
in
,
a
nd
N
.
L
iu
,
“
E
ns
e
mb
le
of
s
ubs
e
t
o
nl
in
e
s
e
q
ue
nt
ia
l
e
xt
r
e
me
l
e
a
r
ni
ng
ma
c
hi
n
e
f
o
r
c
la
s
s
im
ba
la
nc
e
a
nd
c
o
n
c
e
pt
dr
i
f
t,
”
N
e
ur
oc
om
put
in
g
, v
ol
. 149, n
o
. P
a
r
t
A
, pp. 316
–
329, F
e
b
.
2015, do
i:
10.1016/j
.n
e
u
c
o
m.2014.03.075.
[
17]
S
.
R
e
n,
B
.
L
ia
o
,
W
.
Z
hu,
Z
.
L
i,
W
.
L
iu
,
a
nd
K
.
L
i,
“
T
h
e
G
r
a
du
a
l
R
e
s
a
mpl
in
g
E
ns
e
mbl
e
f
or
mi
ni
ng
im
ba
la
nc
e
d
da
ta
s
tr
e
a
ms
w
it
h
c
o
n
c
e
pt
d
r
i
f
t,
”
N
e
ur
oc
om
put
in
g
, vo
l.
286, pp. 150
–
166, Apr
. 20
18, do
i:
10.1016/j
.n
e
u
c
o
m.2018.01.063.
[
18]
S
.
W
a
ng
a
nd
X
.
Y
a
o
,
“
D
iv
e
r
s
it
y
a
na
l
y
s
is
o
n
im
ba
la
nc
e
d
da
ta
s
e
ts
by
us
in
g
e
ns
e
mbl
e
m
o
de
ls
,”
in
2009
I
E
E
E
Sy
m
pos
iu
m
on
C
om
put
at
io
nal
I
nt
e
ll
ig
e
nc
e
and
D
at
a
M
in
in
g,
C
I
D
M
2009
-
P
r
oc
e
e
di
ngs
,
M
a
r
.
2009,
pp.
324
–
331,
do
i:
10.110
9/
C
I
D
M
.2009.4938667.
[
19]
R
.
B
a
r
a
nde
la
,
J
.
S
.
S
á
n
c
he
z
,
a
nd
R
.
M
.
V
a
ld
ov
in
o
s
,
“
N
e
w
A
ppl
ic
a
ti
o
ns
of
E
ns
e
mbl
e
s
of
C
la
s
s
if
i
e
r
s
,”
P
at
te
r
n
A
nal
y
s
is
and
A
ppl
ic
at
io
ns
, vo
l.
6, n
o
. 3, pp. 245
–
256, D
e
c
. 2003, d
o
i:
10.100
7/
s
10044
-
003
-
0192
-
z.
[
20]
V
.
M
.
A
.
S
o
u
z
a
,
D
.
M
.
d
o
s
R
e
is
,
A
.
G
.
M
a
le
t
z
k
e
,
a
nd
G
.
E
.
A
.
P
.
A
.
B
a
ti
s
ta
,
“
C
ha
ll
e
nge
s
in
b
e
n
c
hma
r
ki
ng
s
tr
e
a
m
le
a
r
ni
ng
a
lg
o
r
it
h
ms
w
it
h
r
e
a
l
-
w
o
r
ld
da
ta
,”
D
at
a
M
in
in
g
and
K
now
le
dge
D
is
c
o
v
e
r
y
,
v
o
l.
34,
n
o
.
6,
pp.
1805
–
1858,
J
ul
.
2020,
do
i:
10.1007/s
10618
-
020
-
00698
-
5.
[
21]
N
.
G
.
G
r
a
tz
,
“
C
r
it
i
c
a
l
r
e
v
ie
w
of
th
e
ve
c
t
o
r
s
ta
tu
s
of
A
e
d
e
s
a
lb
o
pi
c
tu
s
,”
M
e
di
c
al
and
V
e
te
r
in
ar
y
E
nt
om
ol
ogy
,
v
o
l.
18,
no
.
3,
pp.
215
–
227, S
e
p. 2004, d
o
i:
10.1111/j
.0269
-
283X.2004.00513.
x
.
[
22]
L
.
C
.
B
a
r
th
o
l
o
ma
y
e
t
al
.
,
“
P
a
th
o
g
e
n
o
mi
c
s
of
C
ul
e
x
qui
nq
ue
f
a
s
c
ia
tu
s
a
nd
me
ta
-
a
na
l
y
s
is
of
in
f
e
c
ti
o
n
r
e
s
p
o
ns
e
s
to
di
ve
r
s
e
pa
th
o
g
e
ns
,”
Sc
ie
nc
e
, v
o
l.
330, n
o
. 6000, pp. 88
–
90, O
c
t.
2010,
do
i:
10.1126/s
c
i
e
nc
e
.1193162.
[
23]
G
. S
n
e
d
e
c
or
a
nd W
. C
oc
hr
a
n, “
S
ta
ti
s
ti
c
a
l
m
e
th
o
ds
, 8t
hE
dn,”
St
at
is
ti
c
al
M
e
th
ods
, p. 503,
1989.
[
24]
S
.
H
id
o
a
nd
H
.
K
a
s
hi
ma
,
“
R
o
ughl
y
ba
la
nc
e
d
ba
ggi
ng
f
o
r
i
mba
la
nc
e
d
da
ta
,”
in
Soc
ie
ty
f
or
I
ndus
tr
ia
l
and
A
ppl
ie
d
M
at
he
m
at
ic
s
-
8t
h
SI
A
M
I
nt
e
r
nat
io
nal
C
onf
e
r
e
nc
e
on
D
at
a
M
in
in
g
2008,
P
r
o
c
e
e
di
ngs
in
A
ppl
ie
d
M
at
he
m
at
ic
s
130
,
A
pr
.
2008,
vo
l.
1,
pp.
1
43
–
152, do
i:
10.1137/1.9781611
972788.13.
[
25]
H
.
B
.
M
a
nn
a
nd
D
.
R
.
W
hi
tn
e
y
,
“
O
n
a
T
e
s
t
of
W
he
th
e
r
o
ne
of
T
w
o
R
a
nd
o
m
V
a
r
ia
bl
e
s
is
S
t
oc
ha
s
ti
c
a
ll
y
L
a
r
g
e
r
th
a
n
th
e
O
th
e
r
,”
T
he
A
nnal
s
of
M
at
he
m
at
ic
al
St
at
is
ti
c
s
, v
o
l.
18, n
o
. 1, pp. 50
–
60
, M
a
r
. 1947, do
i:
10.1214/
a
o
ms
/1
177730491.
B
I
OG
RA
P
HI
E
S
OF
AU
T
HO
RS
V
a
r
s
h
a
Kh
a
n
de
k
a
r
rece
i
v
e
d
t
h
e
M
.
E
.
d
e
g
r
ee
i
n
C
o
m
p
u
t
e
r
E
n
g
i
n
ee
r
i
n
g
fro
m
t
h
e
Sa
v
i
t
ri
b
ai
P
h
u
l
e
Pu
n
e
U
n
i
v
e
rs
i
t
y
,
Mah
aras
h
t
ra
Pu
n
e
,
In
d
i
a
i
n
2
0
0
8
.
Sh
e
i
s
w
o
rk
i
n
g
t
o
w
ard
t
h
e
Ph
.
D
.
d
e
g
r
ee
w
i
t
h
t
h
e
D
e
p
art
me
n
t
o
f
Co
m
p
u
t
e
r
E
n
g
i
n
ee
ri
n
g
,
N
MIMS's
Mu
k
e
s
h
Pat
e
S
c
h
o
o
l
o
f
T
ech
n
o
l
o
g
y
Man
a
g
e
me
n
t
a
n
d
E
n
g
i
n
ee
ri
n
g
,
M
u
m
b
ai
.
Sh
e
i
s
c
u
rr
en
t
l
y
a
n
A
s
s
i
s
t
an
t
Pro
fe
s
s
o
r
i
n
t
h
e
D
e
p
art
men
t
o
f
I
n
f
o
r
m
at
i
o
n
T
ech
n
o
l
o
g
y
o
f
Sm
t
.
K
as
h
i
b
ai
N
av
a
l
e
C
o
l
l
e
g
e
o
f
E
n
g
i
n
ee
r
i
n
g
Pu
n
e
,
I
n
d
i
a.
H
e
r
r
e
s
e
ar
ch
i
n
t
e
r
e
s
t
s
i
n
cl
u
d
e
D
at
a
-
M
i
n
i
n
g
,
E
n
s
em
b
l
e
L
e
arn
i
n
g
,
Ma
c
h
i
n
e
L
e
arn
i
n
g
,
D
at
a
Sci
e
n
ce
a
n
d
B
i
g
D
at
a
A
n
al
y
t
i
c
s
.
Sh
e
c
an
b
e
co
n
t
a
c
t
ed
at
em
a
i
l
:
v
ars
h
a.
k
h
an
d
ek
ar@
g
m
a
i
l
.
c
o
m
.
P
ra
v
i
n
S
h
ri
n
a
t
h
rece
i
v
e
d
t
h
e
M.
T
ec
h
d
eg
r
ee
i
n
Co
m
p
u
t
e
r
E
n
g
i
n
ee
ri
n
g
i
n
2
0
0
8
an
d
Ph
D
d
e
g
r
ee
i
n
2
0
1
6
fro
m
N
M
I
MS’s
Mu
k
es
h
Pat
e
l
S
c
h
o
o
l
o
f
T
ech
n
o
l
o
g
y
Man
ag
eme
n
t
a
n
d
E
n
g
i
n
ee
ri
n
g
,
Mu
m
b
ai
.
H
e
i
s
cu
rr
e
n
t
l
y
an
A
s
s
o
c
i
at
e
Pro
f
e
s
s
o
r
&
H
e
ad
o
f
D
e
p
art
me
n
t
Co
m
p
u
t
e
r
E
n
g
i
n
ee
ri
n
g
at
N
M
I
MS’s
MPST
ME
,
Mu
m
b
ai
.
H
i
s
re
s
e
ar
ch
i
n
t
e
r
e
s
t
s
i
n
cl
u
d
e
Im
ag
e
Pro
ce
s
s
i
n
g
,
A
rt
i
fi
ci
al
In
t
el
l
i
g
en
ce
,
Ma
ch
i
n
e
L
e
arn
i
n
g
.
H
e
c
an
b
e
c
o
n
t
ac
t
e
d
at
em
a
i
l
:
p
rav
i
n
.
s
ri
n
at
h
@
n
m
i
m
s
.
e
d
u
.
Evaluation Warning : The document was created with Spire.PDF for Python.