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.
7
,
No
.
1
,
Feb
r
u
ar
y
201
7
,
p
p
.
5
05
~
5
1
2
I
SS
N:
2088
-
8708
,
DOI
: 1
0
.
1
1
5
9
1
/
i
j
ec
e
.
v
7
i
1
.
p
p
5
0
5
-
5
1
2
505
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ia
e
s
jo
u
r
n
a
l.c
o
m/o
n
lin
e/in
d
ex
.
p
h
p
/I
JE
C
E
O
pti
m
i
zed
Active
Lea
rning
for
Use
r’s Beha
v
io
r Mo
delling
b
a
sed o
n
No
n
-
In
t
rusiv
e S
m
a
rtpho
ne
I
k
a
K
us
u
m
a
nin
g
P
utr
i
1
,
Der
o
n L
ia
ng
2
,
Sh
o
leh H
a
di P
ra
m
o
no
3
,
Ra
h
m
a
dw
a
t
i
4
1
De
p
a
rtme
n
t
o
f
Co
m
p
u
ter S
c
ien
c
e
a
n
d
In
f
o
rm
a
ti
o
n
E
n
g
in
e
e
rin
g
,
N
a
ti
o
n
a
l
Ce
n
t
ra
l
Un
iv
e
rsity
,
Tai
w
a
n
1,
2,
3,
4
De
p
a
rt
m
e
n
t
o
f
El
e
c
tri
c
a
l
En
g
in
e
e
rin
g
,
Un
iv
e
rsity
o
f
Bra
w
ij
a
y
a
,
M
a
lan
g
,
In
d
o
n
e
sia
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Sep
9
,
2
0
1
6
R
ev
i
s
ed
No
v
15
,
2
0
1
6
A
cc
ep
ted
No
v
3
0
,
2
0
1
6
In
o
rd
e
r
to
p
ro
tec
t
th
e
d
a
ta
in
th
e
sm
a
rtp
h
o
n
e
,
t
h
e
re
is
so
m
e
p
ro
tec
ti
o
n
m
e
c
h
a
n
is
m
th
a
t
h
a
s
b
e
e
n
u
se
d
.
T
h
e
c
u
rre
n
t
a
u
th
e
n
ti
c
a
ti
o
n
u
se
s
P
IN,
p
a
ss
w
o
rd
,
a
n
d
b
i
o
m
e
tri
c
-
b
a
se
d
m
e
th
o
d
.
T
h
e
se
a
u
th
e
n
t
ica
ti
o
n
m
e
th
o
d
s
a
re
n
o
t
su
f
f
icie
n
t
d
u
e
t
o
c
o
n
v
e
n
i
e
n
c
e
a
n
d
se
c
u
rit
y
issu
e
.
No
n
-
In
tru
siv
e
a
u
th
e
n
ti
c
a
ti
o
n
is
m
o
re
c
o
m
f
o
rtab
le
b
e
c
a
u
se
it
ju
st
c
o
ll
e
c
ts
u
se
r’s
b
e
h
a
v
io
r
to
a
u
th
e
n
ti
c
a
te
th
e
u
se
r
t
o
th
e
sm
a
rt
p
h
o
n
e
.
S
e
v
e
ra
l
n
o
n
-
in
tr
u
siv
e
a
u
th
e
n
ti
c
a
ti
o
n
m
e
c
h
a
n
is
m
s
w
e
re
p
ro
p
o
se
d
b
u
t
t
h
e
y
d
o
n
o
t
c
a
re
a
b
o
u
t
t
h
e
train
i
n
g
sa
m
p
le
th
a
t
h
a
s
a
lo
n
g
d
a
ta
c
o
ll
e
c
ti
o
n
ti
m
e
.
T
h
is
p
a
p
e
r
p
ro
p
o
se
a
m
e
th
o
d
to
c
o
l
lec
t
d
a
ta
m
o
re
e
ff
icie
n
t
u
sin
g
Op
ti
m
ize
d
A
c
ti
v
e
L
e
a
rn
in
g
.
T
h
e
S
u
p
p
o
rt
V
e
c
to
r
M
a
c
h
in
e
(S
V
M
)
u
se
d
t
o
i
d
e
n
ti
f
y
th
e
e
ff
e
c
t
o
f
so
m
e
s
m
a
ll
a
m
o
u
n
t
o
f
train
in
g
d
a
ta.
T
h
is
p
ro
p
o
se
d
sy
ste
m
h
a
s
tw
o
m
a
in
f
u
n
c
ti
o
n
a
li
ti
e
s,
t
o
re
d
u
c
e
th
e
train
in
g
d
a
ta
u
sin
g
o
p
ti
m
ize
d
sto
p
r
u
le
a
n
d
m
a
in
tain
th
e
Err
o
r
Ra
te
u
sin
g
m
o
d
if
ied
m
o
d
e
l
a
n
a
l
y
sis
to
d
e
t
e
r
m
in
e
th
e
train
in
g
d
a
ta
th
a
t
f
it
f
o
r
e
a
c
h
u
se
r.
F
in
a
ll
y
,
a
f
ter
w
e
d
o
n
e
th
e
e
x
p
e
rime
n
t,
we
c
o
n
c
lu
d
e
th
a
t
o
u
r
p
ro
p
o
se
d
s
y
ste
m
is
b
e
tt
e
r
th
a
n
T
h
re
sh
o
ld
-
b
a
se
d
A
c
ti
v
e
Lea
rn
in
g
.
T
h
e
ti
m
e
r
e
q
u
ired
to
c
o
ll
e
c
t
th
e
d
a
ta
c
a
n
re
d
u
c
e
d
to
4
1
%
f
ro
m
1
7
to
1
0
m
in
u
tes
w
it
h
th
e
sa
m
e
Err
o
r
Ra
te.
K
ey
w
o
r
d
:
A
cti
v
e
lear
n
in
g
No
n
-
i
n
tr
u
s
i
v
e
au
th
e
n
ticat
io
n
Su
p
p
o
r
t v
ec
to
r
m
ac
h
i
n
e
User
au
t
h
en
t
icatio
n
Co
p
y
rig
h
t
©
2
0
1
6
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
:
I
k
a
Ku
s
u
m
a
n
i
n
g
P
u
tr
i
,
Dep
ar
t
m
en
t o
f
E
lectr
ical
E
n
g
i
n
ee
r
in
g
,
Un
i
v
er
s
it
y
o
f
B
r
a
w
ij
ay
a,
J
l.
Vete
r
an
,
Ma
lan
g
6
5
1
4
5
,
E
a
s
t J
av
a,
I
n
d
o
n
esia.
E
m
ail:
i
k
ae
m
an
g
u
k
a
@
g
m
ail.
c
o
m
1.
I
NT
RO
D
UCT
I
O
N
I
n
r
ec
en
t
y
ea
r
s
,
s
m
ar
tp
h
o
n
e
ar
e
u
s
ed
m
o
r
e
f
r
eq
u
en
t
l
y
f
o
r
v
ar
io
u
s
ap
p
licatio
n
s
o
th
er
th
a
n
telec
o
m
m
u
n
icatio
n
.
T
h
e
b
u
s
i
n
ess
ac
ti
v
it
y
li
k
e
o
n
li
n
e
m
o
n
e
y
tr
an
s
ac
tio
n
an
d
a
n
y
o
t
h
er
th
i
n
g
s
m
ak
e
s
u
s
s
to
r
e
o
u
r
p
er
s
o
n
al
d
ata
o
n
th
e
s
m
a
r
tp
h
o
n
e.
T
h
at
p
h
e
n
o
m
e
n
o
n
m
ak
es
u
s
m
o
r
e
co
n
ce
r
n
ed
ab
o
u
t
s
ec
u
r
i
t
y
is
s
u
e
s
i
n
o
u
r
s
m
ar
tp
h
o
n
es.
W
e
n
ee
d
a
p
r
o
tectio
n
f
o
r
ac
ce
s
s
i
n
g
o
u
r
p
er
s
o
n
al
s
e
n
s
i
tiv
e
d
ata
o
n
o
u
r
s
m
ar
tp
h
o
n
es.
T
h
e
cu
r
r
en
t
p
r
o
tectio
n
m
ec
h
a
n
is
m
s
o
f
m
o
b
ile
d
ev
ice
s
ar
e
u
s
u
all
y
b
ased
o
n
P
I
N
co
d
es
o
r
p
ass
w
o
r
d
s
,
an
d
b
io
m
etr
ic
b
ased
m
et
h
o
d
s
.
F
in
g
er
p
r
in
t
is
o
n
e
o
f
th
e
b
io
m
etr
ic
s
y
s
te
m
s
t
h
at
h
a
v
e
a
s
t
r
o
n
g
f
ea
t
u
r
es
w
i
t
h
h
ig
h
p
r
ec
is
io
n
.
I
t
ca
n
b
e
ap
p
lied
o
n
d
esk
to
p
a
u
th
e
n
ticat
io
n
[
1
]
an
d
m
o
b
ile
au
t
h
e
n
ticatio
n
[
2
]
b
y
u
s
i
n
g
th
e
f
i
n
g
er
p
r
in
t
d
e
v
ices
a
n
d
m
o
u
n
ted
f
i
n
g
er
p
r
in
t
i
n
m
o
b
ile
d
ev
ice.
Fi
n
g
er
p
r
in
ts
a
n
d
p
ass
w
o
r
d
en
tr
y
r
eq
u
ir
e
ex
p
licit
ac
tio
n
f
r
o
m
t
h
e
u
s
er
,
w
h
ic
h
is
n
o
t
co
n
v
e
n
ie
n
t
i
n
a
f
r
eq
u
en
t
u
s
e.
T
h
e
p
ass
w
o
r
d
-
b
ased
au
t
h
en
ticatio
n
m
et
h
o
d
s
u
s
ed
o
n
m
o
b
ile
d
ev
i
ce
s
h
av
e
a
s
u
r
f
in
g
is
s
u
e,
p
eo
p
le
ca
n
u
s
e
s
o
cial
en
g
in
ee
r
i
n
g
to
g
et
th
e
p
ass
w
o
r
d
.
So
m
e
cr
ac
k
er
s
al
s
o
ca
n
u
s
e
th
e
b
r
u
te
-
f
o
r
ce
m
et
h
o
d
to
g
et
o
u
r
P
I
N
co
d
e.
A
b
io
m
etr
ic
m
et
h
o
d
lik
e
f
ac
e
r
ec
o
g
n
itio
n
also
h
a
v
e
s
o
m
e
is
s
u
es
w
h
en
p
eo
p
le
in
p
u
t o
u
r
f
ac
e
o
n
l
y
t
h
r
o
u
g
h
o
u
r
p
h
o
to
.
A
cc
o
r
d
in
g
to
r
ec
e
n
t
s
u
r
v
e
y
s
[
3
-
5
]
,
6
0
% to
8
0
% o
f
u
s
er
s
c
h
o
o
s
e
to
tu
r
n
t
h
e
s
e
v
er
if
icat
io
n
f
ea
tu
r
es o
f
f
s
i
m
p
l
y
b
ec
a
u
s
e
o
f
it
s
i
n
co
n
v
e
n
ien
ce
.
I
n
o
r
d
er
to
i
m
p
r
o
v
e
t
h
e
co
n
v
e
n
ie
n
ce
a
n
d
s
ec
u
r
it
y
o
f
th
e
m
o
b
ile
d
ev
ice
s
,
n
o
n
-
i
n
tr
u
s
iv
e
a
u
t
h
en
ticatio
n
m
ec
h
a
n
i
s
m
s
ar
e
d
e
s
ir
ab
le
[
6
]
.
N
o
t
o
n
l
y
f
o
r
m
o
b
ile,
n
o
n
-
i
n
tr
u
s
i
v
e
au
th
e
n
ticat
io
n
also
u
s
ed
f
o
r
d
esk
to
p
s
ec
u
r
it
y
.
So
m
e
d
esk
to
p
p
r
o
tectio
n
r
esear
ch
[
7
]
,
[
8
]
ex
p
lo
its
m
o
u
s
e
d
y
n
a
m
ics
t
h
at
is
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
J
E
C
E
Vo
l.
7
,
No
.
1
,
Feb
r
u
ar
y
201
7
:
5
05
–
512
506
ef
f
ec
tiv
e
f
o
r
co
n
tin
u
o
u
s
v
er
if
i
ca
tio
n
b
ec
au
s
e
it d
o
es
n
o
t
n
ee
d
s
p
ec
ial
h
ar
d
w
ar
e
to
co
ll
ec
t b
io
m
etr
ic
d
ata.
No
n
-
in
tr
u
s
i
v
e
a
u
t
h
en
tica
tio
n
d
o
es
n
o
t
n
ee
d
an
in
ter
f
ac
e
b
ec
a
u
s
e
it
co
llects
u
s
er
's
b
e
h
av
io
r
i
n
t
h
e
b
ac
k
g
r
o
u
n
d
t
h
en
co
n
tin
u
e
au
t
h
e
n
ticati
n
g
to
p
r
o
tect
d
ata
o
n
th
e
s
m
ar
tp
h
o
n
e.
T
h
er
e
ar
e
s
ev
er
al
m
et
h
o
d
No
n
-
I
n
tr
u
s
iv
e
au
th
e
n
tica
tio
n
t
h
at
h
as
b
ee
n
p
r
o
p
o
s
ed
,
s
u
ch
as
u
s
i
n
g
th
e
o
r
ien
tatio
n
s
e
n
s
o
r
o
r
a
to
u
ch
s
cr
ee
n
an
d
u
s
i
n
g
B
atc
h
L
ea
r
n
i
n
g
[9
-
11]
an
d
A
ctiv
e
L
ea
r
n
in
g
[1
2
]
m
et
h
o
d
to
co
llect
th
e
tr
ain
i
n
g
d
ata
th
at
co
n
s
tr
u
c
t
th
e
No
n
-
in
tr
u
s
i
v
e
au
th
e
n
tica
tio
n
m
ec
h
a
n
i
s
m
.
T
h
e
y
u
s
e
s
m
ar
tp
h
o
n
e
s
as
s
e
n
s
o
r
s
to
id
en
tify
u
s
er
b
eh
av
io
r
w
h
ile
o
p
er
atin
g
t
h
e
s
m
ar
tp
h
o
n
e.
T
h
eir
tr
ain
i
n
g
d
a
ta
co
llected
f
r
o
m
t
h
e
w
a
y
o
f
u
s
er
u
s
e
t
h
eir
p
h
o
n
e.
T
h
ese
r
esear
ch
to
o
k
a
f
lic
k
to
u
ch
d
ata
f
r
o
m
th
e
u
s
er
as
t
h
e
tr
ain
in
g
d
ata.
T
h
e
d
ata
is
r
etr
iev
ed
u
s
i
n
g
a
d
is
p
la
y
li
s
t
t
h
e
n
an
al
y
ze
d
as
u
s
er
b
eh
av
io
r
.
T
h
e
B
atch
L
ea
r
n
i
n
g
an
d
T
h
r
esh
o
ld
-
b
ased
Acti
v
e
L
ea
r
n
i
n
g
p
r
o
v
id
e
n
o
n
-
in
tr
u
s
iv
e
a
u
t
h
e
n
t
icatio
n
to
th
e
s
m
ar
tp
h
o
n
e,
alt
h
o
u
g
h
ea
ch
o
f
t
h
e
m
h
a
s
it
s
w
ea
k
n
e
s
s
.
B
at
ch
p
r
o
ce
s
s
i
n
g
is
a
m
o
d
el
o
f
d
a
ta
p
r
o
ce
s
s
in
g
,
w
it
h
co
llectin
g
d
ata
f
ir
s
t,
an
d
o
r
g
a
n
ize
clu
s
ter
in
g
d
ata
in
to
g
r
o
u
p
s
ca
lled
b
atch
es
[
1
3
]
,
[
1
4
]
.
T
h
e
B
atch
L
ea
r
n
in
g
Au
t
h
en
t
icatio
n
n
ee
d
s
ab
o
u
t
2
4
0
tr
ain
in
g
d
ata
f
o
r
ea
ch
u
s
er
to
b
u
ild
th
e
m
o
d
el
th
at
w
il
l
b
e
u
s
ed
as
th
e
au
th
e
n
tica
tio
n
to
ac
h
ie
v
e
h
ig
h
r
ec
o
g
n
itio
n
r
ate.
Ho
w
ev
er
,
to
g
et
a
lo
t
o
f
d
ata
ta
k
es
a
lo
n
g
t
i
m
e
d
ata
co
llectio
n
an
d
it
is
n
o
t
ef
f
ic
ien
t
f
o
r
th
e
u
s
er
.
T
h
e
A
ctiv
e
L
ea
r
n
i
n
g
s
el
ec
t
th
e
d
ata
f
r
o
m
t
h
e
lear
n
i
n
g
p
r
o
ce
s
s
to
im
p
r
o
v
e
p
er
f
o
r
m
a
n
ce
w
it
h
le
s
s
tr
ain
i
n
g
d
ata
[
1
2
]
.
T
h
r
esh
o
ld
-
b
ased
A
cti
v
e
L
ea
r
n
i
n
g
u
s
es
a
d
i
f
f
er
e
n
t
w
a
y
o
f
co
llec
t
in
g
d
ata,
t
h
e
a
m
o
u
n
t
s
o
f
tr
ai
n
i
n
g
d
ata
f
o
r
ea
ch
u
s
er
,
v
ar
iet
y
o
f
p
o
s
itio
n
s
a
n
d
t
y
p
e
o
f
f
l
ick
i
s
d
i
f
f
er
en
t
ac
co
r
d
in
g
to
ea
ch
u
s
er
’
s
h
ab
it.
A
cti
v
e
L
ea
r
n
in
g
w
i
ll
co
llect
tr
ain
i
n
g
d
ata
in
s
o
m
e
s
ta
g
es
i
f
t
h
e
v
alid
ati
o
n
ac
cu
r
ac
y
i
n
t
h
e
p
r
ev
io
u
s
s
ta
g
e
d
o
es
n
o
t
ac
h
iev
e
a
g
o
o
d
r
esu
lt
y
et.
T
h
e
d
ata
co
llectio
n
o
f
A
cti
v
e
L
e
ar
n
in
g
d
ep
en
d
s
o
n
ac
cu
r
ac
y
i
n
t
h
e
v
alid
atio
n
p
r
o
ce
s
s
.
T
h
r
esh
o
ld
-
b
a
s
ed
A
ctiv
e
L
ea
r
n
in
g
d
ata
co
llectio
n
ti
m
e
f
aster
t
h
a
n
B
atc
h
L
ea
r
n
i
n
g
b
ec
au
s
e
it
j
u
s
t
ta
k
es
th
e
i
m
p
o
r
ta
n
t
d
ata
ac
co
r
d
in
g
to
u
s
er
h
ab
it
s
.
Ho
w
e
v
er
,
th
e
e
r
r
o
r
r
ate
is
g
r
ea
ter
th
an
B
atc
h
L
ea
r
n
in
g
w
h
ich
m
ea
n
s
it
h
a
s
lo
w
er
r
ec
o
g
n
itio
n
r
ate.
B
o
th
ex
is
ti
n
g
t
y
p
es
o
f
r
esear
ch
h
av
e
ea
c
h
o
th
er
's
w
ea
k
n
es
s
es,
B
atch
L
ea
r
n
in
g
h
a
s
a
lo
n
g
d
ata
co
llectio
n
ti
m
e
a
n
d
T
h
r
esh
o
ld
-
b
ased
Activ
e
L
ea
r
n
i
n
g
h
a
s
a
b
ad
r
ec
o
g
n
itio
n
r
ate.
T
h
e
v
alu
e
o
f
E
q
u
al
E
r
r
o
r
R
ate
(
E
E
R
)
an
d
co
llectio
n
d
ata
ti
m
e
o
f
t
h
e
p
r
ev
i
o
u
s
r
esear
ch
s
h
o
w
n
in
Fig
u
r
e
1
.
B
atch
L
ea
r
n
i
n
g
h
a
v
e
b
est
E
E
R
t
h
an
t
h
e
o
t
h
er
p
o
in
ts
b
u
t
t
h
e
co
llectio
n
ti
m
e
ab
o
u
t
2
0
m
i
n
u
tes
i
s
n
o
t
ef
f
icien
t
to
u
s
er
w
h
ile
co
llec
tin
g
th
e
d
ata.
T
h
e
r
ec
tan
g
le
an
d
tr
ian
g
le
p
o
in
t
i
s
th
eT
h
r
e
s
h
o
ld
-
b
ased
A
ct
iv
e
L
ea
r
n
i
n
g
th
a
t
u
s
in
g
9
0
%
an
d
9
5
%
as
th
e
th
r
es
h
o
ld
to
s
to
p
th
e
co
llectin
g
d
ata.
T
h
e
T
h
r
e
s
h
o
ld
-
b
ased
A
cti
v
e
L
ea
r
n
i
n
g
(
T
B
A
L
)
h
a
v
e
f
aster
-
co
llecti
n
g
d
ata
t
h
an
th
e
B
atch
L
ea
r
n
i
n
g
.
I
t
i
s
p
o
s
s
ib
le
t
o
co
llect
d
ata
f
r
o
m
a
u
s
er
i
n
th
e
r
e
al
i
m
p
le
m
e
n
tatio
n
.
Fig
u
r
e
1
.
E
E
R
an
d
Data
C
o
lle
ctin
g
T
im
e
o
f
B
atch
L
ea
r
n
i
n
g
an
d
T
h
r
esh
o
ld
B
ased
A
cti
v
e
L
ea
r
n
in
g
Fig
u
r
e
2
s
h
o
w
t
h
at
T
B
AL
n
o
t
ef
f
ec
ti
v
e
to
r
ed
u
ce
th
e
tr
ain
i
n
g
d
ata.
Yello
w
lin
e
i
n
th
a
t
f
i
g
u
r
e
is
11
s
tag
e
s
t
h
at
m
ea
n
s
it
co
llect
t
h
e
s
a
m
e
tr
ai
n
i
n
g
d
ata
n
u
m
b
er
w
it
h
t
h
e
B
atch
L
ea
r
n
i
n
g
.
T
h
er
e
ar
e
1
9
o
f
4
5
u
s
er
s
th
at
ca
n
n
o
t
ef
f
ec
ti
v
el
y
r
ea
c
h
th
e
s
to
p
r
u
le.
I
t
s
till
h
a
s
m
o
r
e
tr
ain
i
n
g
d
ata
t
h
an
th
e
b
atch
lear
n
i
n
g
.
T
h
e
m
ai
n
p
u
r
p
o
s
e
o
f
th
is
r
esear
ch
ar
e
in
cr
ea
s
in
g
th
e
r
ec
o
g
n
itio
n
r
ate
a
n
d
r
ed
u
ce
th
e
a
m
o
u
n
t
o
f
tr
ai
n
i
n
g
d
ata
s
o
w
e
j
u
s
t
n
ee
d
less
t
i
m
e
to
co
llect
d
ata.
T
h
e
ex
p
ec
ted
er
r
o
r
r
ate
s
h
o
u
ld
b
e
less
th
a
n
o
r
eq
u
al
an
d
th
e
ti
m
e
ta
k
e
n
ex
p
ec
ted
to
b
e
les
s
t
h
a
n
o
r
eq
u
al
to
T
B
AL
9
5
%.
T
o
d
em
o
n
s
tr
ate
th
e
f
ea
s
ib
ilit
y
o
f
t
h
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
,
an
ap
p
h
as
b
ee
n
i
m
p
le
m
en
ted
o
n
t
h
e
An
d
r
o
id
4
.
3
o
p
er
atin
g
s
y
s
te
m
to
co
llect
t
h
e
b
io
m
e
tr
ics
f
r
o
m
th
e
to
u
c
h
s
en
s
o
r
o
f
4
5
u
s
er
s
w
h
e
n
t
h
e
y
o
p
er
ate
th
e
s
m
ar
tp
h
o
n
es i
n
th
e
ir
h
an
d
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
E
C
E
I
SS
N:
2088
-
8708
Op
timiz
ed
A
ctive
Lea
r
n
in
g
fo
r
User
’
s
B
eh
a
vio
r
Mo
d
ellin
g
B
a
s
ed
o
n
…
(
I
ka
K
u
s
u
ma
n
i
n
g
P
u
tr
i
)
507
Fig
u
r
e
2
.
T
h
e
E
f
f
ec
ti
v
en
e
s
s
o
f
T
h
r
esh
o
ld
-
b
ased
A
ct
iv
e
L
ea
r
n
in
g
2.
RE
S
E
ARCH
M
E
T
H
O
D
Fig
u
r
e
3
s
h
o
w
s
th
e
d
esig
n
s
c
en
ar
io
o
f
th
is
p
r
o
p
o
s
ed
r
esear
ch
.
I
n
th
is
s
ec
tio
n
,
t
h
e
d
ata
co
llectio
n
s
tag
e
s
,
s
to
p
p
in
g
cr
iter
ia
to
co
llect
u
s
er
’
s
tr
ai
n
i
n
g
d
ata
a
n
d
m
o
d
el
an
a
l
y
s
is
ar
e
p
r
esen
t
ed
as
f
o
llo
w
s
.
T
h
is
s
y
s
te
m
co
llects
6
0
tr
ain
i
n
g
d
ata
in
th
e
f
ir
s
t
s
tag
e
t
h
en
d
o
th
e
tr
ai
n
i
n
g
u
s
i
n
g
SVM.
I
f
t
h
e
SV
M
v
alid
atio
n
ac
cu
r
ac
y
m
e
t
t
h
e
cr
iter
ia
to
s
to
p
,
th
en
t
h
e
s
y
s
te
m
w
i
ll
s
to
p
co
llectin
g
d
ata
an
d
s
tar
t
cr
ea
tin
g
Mo
d
el
u
s
i
n
g
SVM.
B
u
t,
i
f
t
h
e
S
VM
v
alid
atio
n
ac
cu
r
ac
y
d
id
n
’
t
m
ee
t
th
e
s
to
p
p
in
g
cr
iter
ia,
t
h
e
s
y
s
te
m
w
il
l
an
a
l
y
ze
th
e
tr
ain
i
n
g
d
ata
in
th
at
s
ta
g
e
to
k
n
o
w
w
h
ic
h
b
e
h
av
io
r
t
h
at
f
it
u
s
er
.
1
8
tr
ain
i
n
g
d
ata
w
ill
b
e
co
llected
b
ased
o
n
b
eh
av
io
r
r
atio
th
at
alr
ea
d
y
an
a
l
y
ze
d
an
d
s
tar
t tr
ai
n
i
n
g
a
g
ain
u
s
i
n
g
all
s
ta
g
e
tr
ain
in
g
d
ata.
Fig
u
r
e
3
.
R
esea
r
ch
De
s
ig
n
2
.
1
.
Da
t
a
Co
llect
io
n St
a
g
es
T
h
e
u
s
er
’
s
o
p
er
atio
n
al
b
eh
a
v
i
o
r
ca
n
b
e
d
iv
id
ed
in
to
s
ix
ca
teg
o
r
ies
ac
co
r
d
in
g
to
t
h
eir
p
o
s
tu
r
e
a
n
d
f
lic
k
t
y
p
e
as
s
h
o
w
n
in
Fi
g
u
r
e
4
.
E
v
er
y
u
s
er
h
as
a
d
if
f
er
e
n
t
r
atio
f
o
r
ea
ch
o
p
e
r
atio
n
al
b
eh
av
io
r
b
ased
o
n
th
eir
h
ab
its
.
T
h
e
T
B
A
L
u
s
e
d
if
f
er
en
t
a
m
o
u
n
t
o
f
d
ata
r
eq
u
ir
ed
b
et
w
ee
n
th
e
o
p
er
atio
n
al
b
eh
a
v
io
r
s
.
W
e
d
ec
id
ed
s
itti
n
g
a
n
d
s
tan
d
i
n
g
as
th
e
p
o
s
tu
r
e
t
y
p
e
b
ec
au
s
e
t
h
ese
p
o
s
t
u
r
es
co
m
m
o
n
u
s
ed
b
y
e
v
er
y
o
n
e
w
h
e
n
o
p
er
atin
g
t
h
e
s
m
ar
tp
h
o
n
e
th
a
n
l
y
in
g
,
w
alk
i
n
g
o
r
r
u
n
n
i
n
g
p
o
s
tu
r
e.
W
e
ar
e
tr
y
i
n
g
to
b
u
ild
as
m
u
c
h
as
p
o
s
s
ib
le
f
o
r
b
eh
av
io
r
o
p
er
atio
n
to
g
et
th
e
s
p
ec
i
f
ic
u
s
er
’
s
b
eh
a
v
io
r
v
ar
iat
io
n
.
T
h
er
ef
o
r
e
w
e
d
ec
id
ed
to
u
s
e
s
h
o
r
t,
m
ed
i
u
m
a
n
d
lo
n
g
f
lic
k
t
y
p
e
f
o
r
u
s
er
’
s
b
eh
a
v
io
r
o
p
er
atio
n
in
s
tead
o
f
u
s
i
n
g
o
n
e
f
lic
k
t
y
p
e
o
r
j
u
s
t lo
n
g
a
n
d
s
h
o
r
t a
s
a
f
lic
k
t
y
p
e.
I
n
t
h
e
f
ir
s
t
p
h
ase
,
t
h
e
s
y
s
te
m
to
o
k
th
e
s
a
m
e
a
m
o
u
n
t
o
f
d
ata
f
r
o
m
th
e
6
t
y
p
e
s
o
f
b
eh
a
v
io
r
,
it
u
s
ed
to
d
etec
t
u
s
er
’
s
b
eh
a
v
io
r
w
h
ile
o
p
er
atin
g
t
h
e
s
m
ar
tp
h
o
n
e.
T
h
e
s
ec
o
n
d
p
h
ase
w
i
ll
ta
k
e
t
h
e
d
at
a
ac
co
r
d
in
g
to
u
s
er
b
eh
av
io
r
at
th
e
p
r
ev
io
u
s
s
ta
g
e.
T
h
ed
ata
co
llectio
n
am
o
u
n
t
f
o
r
ea
ch
d
if
f
er
e
n
t
b
eh
av
io
r
d
ep
en
d
s
o
n
th
e
r
atio
o
b
tain
ed
f
r
o
m
t
h
e
a
n
al
y
s
i
s
o
f
b
eh
av
io
r
S
VM
m
o
d
el
at
th
e
p
r
ev
io
u
s
s
ta
g
e.
T
h
e
s
y
s
te
m
u
s
e
s
ea
c
h
u
s
er
p
r
o
f
i
le
to
in
d
icate
w
h
ic
h
o
p
er
atio
n
w
il
l
b
e
ca
r
r
ied
o
u
t
to
cr
ea
te
a
m
o
d
el.
As
s
h
o
w
n
i
n
Fi
g
u
r
e
5
,
th
e
u
s
er
a
s
k
ed
to
s
tan
d
u
p
a
n
d
s
cr
o
ll
t
h
e
s
cr
ee
n
to
t
h
e
n
u
m
b
er
s
1
5
an
d
clic
k
o
n
it.
T
h
e
as
k
ed
n
u
m
b
er
s
r
ep
r
esen
t
t
h
e
t
y
p
e
o
f
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
J
E
C
E
Vo
l.
7
,
No
.
1
,
Feb
r
u
ar
y
201
7
:
5
05
–
512
508
f
lic
k
r
eq
u
ir
ed
.
I
n
th
e
u
s
er
p
r
o
f
ile,
w
e
m
u
s
t
d
eter
m
in
e
t
w
o
th
in
g
s
,
f
ir
s
t
h
o
w
m
u
ch
d
ata
s
h
o
u
ld
b
e
o
p
e
r
ated
b
y
th
e
u
s
er
an
d
th
e
s
ec
o
n
d
p
o
s
tu
r
e
an
d
n
u
m
b
er
in
d
icati
n
g
f
l
ick
t
y
p
e
o
f
th
e
u
s
er
’
s
o
p
er
atio
n
al
b
eh
av
io
r
.
Fig
u
r
e
4
.
B
eh
av
io
r
Op
er
atio
n
Fig
u
r
e
5
.
Dis
p
la
y
A
p
p
licatio
n
2
.
2
.
Sto
pp
ing
Crit
er
ia
So
m
e
p
eo
p
le
h
a
v
e
a
d
i
f
f
er
e
n
t
r
ate
to
r
ea
ch
t
h
eir
s
tab
ilit
y
w
h
ile
u
s
in
g
t
h
e
s
m
ar
tp
h
o
n
e.
T
h
ey
ca
n
s
tab
le
f
r
o
m
th
e
f
ir
s
t
s
e
s
s
io
n
o
r
th
e
y
w
i
ll
s
tab
le
a
f
ter
s
o
m
e
s
tag
e
s
t
h
at
i
n
d
icate
d
b
y
t
h
e
SVM
v
a
lid
atio
n
ac
cu
r
ac
y
.
I
f
th
e
y
r
ea
c
h
t
h
e
h
ig
h
e
s
t
v
alid
atio
n
ac
c
u
r
ac
y
,
th
e
s
y
s
te
m
d
o
es
n
o
t
n
ee
d
to
co
llect
o
th
er
d
ata
an
y
m
o
r
e.
B
u
t,
if
th
e
v
alid
atio
n
ac
c
u
r
ac
y
is
s
t
ill
i
m
p
r
o
v
i
n
g
,
w
e
n
ee
d
to
lo
o
k
a
n
o
th
er
s
ta
g
e
u
n
t
il
r
ea
ch
e
s
t
h
e
h
ig
h
e
s
t
ac
c
u
r
ac
y
.
Mo
s
t
u
s
er
s
h
av
e
th
e
s
a
m
e
p
atter
n
o
n
t
h
e
SVM
v
al
id
atio
n
ac
c
u
r
ac
y
.
I
f
t
h
e
ir
v
a
lid
atio
n
ac
cu
r
ac
y
h
i
g
h
a
n
d
w
ill
n
o
t
r
is
e
ag
ai
n
s
i
g
n
if
ican
t
l
y
,
t
h
e
m
o
d
el
o
f
t
h
e
h
i
g
h
est
ac
c
u
r
ac
y
th
at
h
as
b
ee
n
f
o
r
m
ed
i
s
th
e
b
est
m
o
d
el,
s
o
w
e
d
o
n
o
t n
ee
d
to
ad
d
m
o
r
e
d
ata
an
d
it
w
i
ll r
ed
u
ce
th
e
co
llecti
n
g
d
ata
ti
m
e.
Fi
g
u
r
e
6
s
h
o
w
s
s
o
m
e
u
s
er
th
a
t
h
av
e
SVM
v
a
lid
atio
n
ac
c
u
r
ac
y
m
o
r
e
t
h
a
n
o
r
eq
u
al
to
9
5
%
w
i
ll
s
tab
le
a
n
d
n
o
t
d
ec
r
ea
s
e
o
r
in
cr
ea
s
e
s
ig
n
i
f
ica
n
tl
y
.
I
n
t
h
i
s
s
tab
le
co
n
d
itio
n
,
w
e
ca
n
m
a
k
e
th
e
d
ata
co
llectio
n
s
to
p
ea
r
lie
r
b
ec
au
s
e
th
e
ea
r
l
y
m
o
d
el
alr
ea
d
y
h
as a
g
o
o
d
m
o
d
el
an
d
d
o
n
o
t n
ee
d
to
ad
d
m
o
r
e
d
ata.
Fig
u
r
e
6
.
Stab
le
SVM
Valid
ati
o
n
A
cc
u
r
ac
y
W
e
s
et
th
e
s
to
p
p
in
g
cr
iter
ia
f
o
r
co
llectin
g
d
ata
to
r
ed
u
ce
t
h
e
n
u
m
b
er
o
f
tr
ain
in
g
d
ata.
I
f
th
e
s
ta
g
e
r
ea
ch
es
t
h
e
cr
iter
ia,
th
e
co
llec
tin
g
d
ata
w
i
ll
b
e
s
to
p
p
ed
.
I
n
t
h
is
r
esear
ch
,
w
e
s
et
t
h
e
s
to
p
p
in
g
cr
iter
ia
u
n
t
il
t
h
e
SVM
v
alid
atio
n
ac
cu
r
ac
y
o
f
th
e
p
r
ev
io
u
s
s
tag
e
d
o
es
n
o
t
im
p
r
o
v
e
an
y
m
o
r
e
o
r
s
to
p
u
n
ti
l
SVM
v
alid
atio
n
ac
cu
r
ac
y
r
ea
ch
m
o
r
e
t
h
an
o
r
eq
u
al
to
9
5
%
an
d
s
et
1
1
s
ta
g
e
s
as
a
m
ax
i
m
u
m
n
u
m
b
er
o
f
t
h
e
s
tag
e.
W
e
ch
o
o
s
e
11
s
tag
es
as
t
h
e
m
a
x
i
m
u
m
n
u
m
b
er
o
f
s
ta
g
es
ac
co
r
d
in
g
to
t
h
e
n
u
m
b
er
o
f
tr
ai
n
i
n
g
d
ata
i
n
th
e
b
atch
lear
n
i
n
g
b
ec
au
s
e
b
atch
lear
n
i
n
g
g
i
v
es
t
h
e
b
est
E
E
R
w
i
th
th
e
s
a
m
e
n
u
m
b
e
r
o
f
d
ata
i
n
th
e
1
1
th
s
tag
e.
T
ab
le
1
s
h
o
w
s
t
h
e
s
to
p
r
u
le
p
s
eu
d
o
co
d
e
b
ef
o
r
e
c
o
llectin
g
th
e
n
ex
t d
ata
u
s
i
n
g
3
w
i
n
d
o
w
s
ta
g
e
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
E
C
E
I
SS
N:
2088
-
8708
Op
timiz
ed
A
ctive
Lea
r
n
in
g
fo
r
User
’
s
B
eh
a
vio
r
Mo
d
ellin
g
B
a
s
ed
o
n
…
(
I
ka
K
u
s
u
ma
n
i
n
g
P
u
tr
i
)
509
T
ab
le
1
.
P
s
eu
d
o
co
d
e
o
f
Data
C
o
llectio
n
S
to
p
R
u
le
if
sta
g
e
n
-
2
≥ sta
g
e
n
-
1
th
e
n
b
e
st
S
V
M
A
c
c
u
ra
c
y
=
sta
g
e
n
-
2
e
lse
th
e
n
b
e
st
S
V
M
A
c
c
u
ra
c
y
=
sta
g
e
n
-
1
e
n
d
if
th
re
sh
o
l
d
=
b
e
st
S
VM
A
c
c
u
ra
c
y
*
1
%
if
(sta
g
e
n
-
b
e
st
S
V
M
A
c
c
u
ra
c
y
)
>
th
re
sh
o
l
d
A
ND
S
V
M
A
c
c
u
ra
c
y
<
9
5
th
e
n
g
o
to
n
e
x
t
sta
g
e
e
lse
th
e
n
st
o
p
e
n
d
if
2
.
3
.
M
o
del A
na
ly
s
is
T
h
e
T
B
A
L
u
s
e
s
th
e
m
o
d
el
an
al
y
s
i
s
to
d
eter
m
i
n
e
t
h
e
n
e
x
t
s
t
ag
e
tr
ain
i
n
g
d
ata.
I
t
u
s
e
s
SV
M
class
i
f
ier
in
t
h
e
p
r
ev
io
u
s
s
tag
e
to
d
eter
m
i
n
e
w
h
at
b
eh
a
v
io
r
is
in
cl
u
d
ed
in
SVM
s
u
p
p
o
r
t
v
ec
to
r
.
T
h
en
th
e
r
atio
o
f
ea
ch
b
eh
av
io
r
o
n
SVM
s
u
p
p
o
r
t
v
ec
to
r
is
ca
lcu
lated
to
o
b
tain
th
e
a
m
o
u
n
t
o
f
d
ata
o
n
i
n
d
i
v
id
u
al
b
e
h
av
io
r
as
ad
d
itio
n
al
d
ata
in
th
e
s
ta
g
e.
I
n
th
is
r
esear
ch
,
w
e
u
s
e
t
h
e
SV
M
class
i
f
ier
u
s
in
g
s
o
f
t
m
ar
g
in
to
s
ep
ar
ate
th
e
n
o
n
-
lin
ea
r
s
ep
ar
ab
le
class
es.
T
o
d
e
ter
m
i
n
e
w
h
ic
h
b
eh
a
v
io
r
th
at
w
il
l
b
e
u
s
ed
f
o
r
n
ex
t
s
tag
e,
we
an
al
y
ze
t
h
e
clo
s
e
s
t
p
o
in
t
to
th
e
h
y
p
er
p
la
n
e
b
ec
au
s
e
SV
M
s
u
p
p
o
r
t
v
ec
to
r
is
n
o
t
th
e
m
o
s
t
s
en
s
iti
v
e
m
e
m
b
er
s
o
f
th
e
cla
s
s
s
i
n
ce
th
er
e
w
er
e
s
til
l
a
n
y
m
e
m
b
er
s
th
at
c
lo
s
er
to
t
h
e
h
y
p
er
p
lan
e.
Sin
ce
w
e
u
s
e
t
h
e
clo
s
e
s
t
p
o
in
t
s
to
f
i
n
d
t
h
e
f
it
test
u
s
er
’
s
b
eh
av
io
r
,
w
e
j
u
s
t
f
i
n
d
1
8
clo
s
est p
o
i
n
ts
to
t
h
e
h
y
p
er
p
l
an
e.
T
h
en
w
e
a
n
al
y
ze
al
l p
o
in
t
to
g
e
t
n
ex
t s
tag
e
’
s
b
eh
av
io
r
tr
ain
i
n
g
d
ata.
A
lt
h
o
u
g
h
t
h
e
o
p
ti
m
al
h
y
p
er
p
lan
e
co
n
s
tr
u
cted
b
y
s
u
p
p
o
r
t
v
e
cto
r
,
th
er
e
ar
e
s
o
m
e
p
o
in
t
s
t
h
a
t
clo
s
er
to
th
e
h
y
p
er
p
lan
e
th
a
n
t
h
e
s
u
p
p
o
r
t
v
ec
to
r
.
B
ec
au
s
e
o
f
t
h
eir
p
o
s
itio
n
clo
s
er
to
th
e
h
y
p
er
p
l
an
e,
th
e
y
ar
e
m
o
r
e
s
en
s
iti
v
e
t
h
an
o
t
h
er
an
d
p
o
ten
tiall
y
m
i
s
clas
s
i
f
ied
.
Fi
g
u
r
e
7
s
h
o
w
s
t
h
e
s
u
p
p
o
r
t
v
ec
to
r
an
d
t
h
e
p
o
in
t
s
b
et
w
ee
n
th
e
m
ar
g
i
n
a
n
d
h
y
p
er
p
lan
e
i
n
t
h
e
SV
M
s
o
f
t
m
ar
g
i
n
t
h
at
h
av
e
m
o
r
e
s
e
n
s
iti
v
e
e
f
f
ec
t
to
th
e
cla
s
s
i
f
icat
io
n
.
T
h
er
ef
o
r
e,
w
e
d
ec
id
ed
to
ch
o
o
s
e
th
e
m
to
g
et
m
o
r
e
ac
cu
r
ate
class
i
f
icatio
n
.
Fig
u
r
e
7
.
Su
p
p
o
r
t V
ec
to
r
an
d
Mo
r
e
Sen
s
iti
v
e
P
o
in
t in
SVM
So
f
t M
ar
g
i
n
3.
RE
SU
L
T
S
A
ND
D
I
SCU
SS
I
O
N
T
h
e
n
e
w
m
o
d
el
a
n
al
y
s
is
ap
p
lied
to
all
u
s
er
s
an
d
w
e
d
id
n
o
t
li
m
it
t
h
e
n
u
m
b
er
o
f
s
ta
g
es
to
k
n
o
w
t
h
e
i
m
p
ac
t
o
f
th
e
n
e
w
m
o
d
el
a
n
a
l
y
s
i
s
to
all
s
ta
g
es.
W
e
u
s
e
1
7
s
tag
e
s
t
h
at
s
a
m
e
w
it
h
t
h
e
m
ax
i
m
u
m
n
u
m
b
er
o
f
s
tag
e
s
in
T
h
r
esh
o
ld
-
b
ased
A
ct
iv
e
L
ea
r
n
i
n
g
to
an
al
y
ze
th
e
d
if
f
er
en
ce
o
f
SVM
v
al
id
atio
n
ac
cu
r
ac
y
w
h
e
n
u
s
i
n
g
th
e
d
if
f
er
en
t
m
o
d
el
an
al
y
s
i
s
.
Fig
u
r
e
8
s
h
o
w
s
th
at
t
h
e
av
er
ag
e
o
f
SVM
v
al
id
atio
n
ac
cu
r
ac
y
i
n
cr
ea
s
i
n
g
t
o
al
m
o
s
t
all
u
s
er
s
w
h
e
n
u
s
i
n
g
t
h
e
n
e
w
m
o
d
el
an
al
y
s
i
s
w
h
ich
u
s
es
t
h
e
clo
s
e
s
t
p
o
in
t
to
d
eter
m
in
e
t
h
e
r
atio
o
f
ea
ch
b
eh
av
io
r
o
p
er
atio
n
.
W
h
en
w
e
ap
p
l
y
t
h
e
t
h
r
es
h
o
ld
b
ased
s
to
p
r
u
le
to
t
h
is
n
e
w
m
o
d
el
an
al
y
s
is
,
t
h
e
co
llecti
n
g
d
ata
w
ill
b
e
s
to
p
p
ed
ea
r
lier
th
an
b
ef
o
r
e
b
ec
au
s
e
th
e
v
al
u
e
o
f
SVM
v
alid
atio
n
ac
c
u
r
ac
y
h
as
i
m
p
r
o
v
ed
.
A
s
s
h
o
w
n
i
n
Fig
u
r
e
9
,
ab
o
u
t
2
0
o
f
4
5
u
s
er
s
ca
n
s
to
p
ea
r
lier
th
an
u
s
i
n
g
th
e
s
u
p
p
o
r
t
v
ec
to
r
m
o
d
el
an
al
y
s
i
s
.
T
h
e
av
er
ag
e
n
u
m
b
er
o
f
t
h
e
s
tag
e
th
at
n
ee
d
ed
f
o
r
all
u
s
er
w
h
e
n
u
s
i
n
g
s
u
p
p
o
r
t
v
ec
to
r
m
o
d
el
a
n
al
y
s
i
s
is
a
b
o
u
t
9
.
5
s
tag
es
a
n
d
it
d
ec
r
ea
s
es
to
6
.
5
s
tag
es
w
h
i
le
u
s
i
n
g
t
h
e
clo
s
est
p
o
in
t
m
o
d
el
an
al
y
s
is
.
T
h
e
av
er
ag
e
o
f
co
llectin
g
d
ata
ti
m
e
ca
n
r
ed
u
ce
ab
o
u
t
2
3
%
f
r
o
m
1
7
m
i
n
u
tes
to
1
3
m
i
n
u
te
s
.
T
h
e
r
esu
lt
s
r
ev
ea
l
th
at
r
ed
u
ci
n
g
t
h
e
ti
m
e
o
f
th
e
m
o
d
el
an
al
y
s
is
b
ased
o
n
clo
s
e
s
t p
o
in
t
s
to
th
e
h
y
p
er
p
lan
e
is
s
ig
n
i
f
ica
n
t
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
J
E
C
E
Vo
l.
7
,
No
.
1
,
Feb
r
u
ar
y
201
7
:
5
05
–
512
510
Fig
u
r
e
8
.
SVM
Valid
atio
n
A
c
cu
r
ac
y
C
o
m
p
ar
is
o
n
Fig
u
r
e
9
.
Stag
es
R
ed
u
ctio
n
Af
ter
k
n
o
w
in
g
t
h
e
m
o
d
el
a
n
al
y
s
i
s
e
f
f
ec
ti
v
en
e
s
s
,
w
e
ap
p
lied
th
e
o
p
ti
m
ized
s
to
p
r
u
le
to
th
e
s
y
s
te
m
to
li
m
it
t
h
e
n
u
m
b
er
o
f
s
tag
e
s
.
T
o
d
eter
m
i
n
e
w
h
et
h
er
th
e
r
es
u
lts
o
f
th
e
ex
p
er
i
m
en
t
is
s
u
cc
ess
f
u
l,
w
e
d
id
a
co
m
p
ar
is
o
n
o
f
th
e
ti
m
e
d
ata
co
llectio
n
b
et
w
ee
n
th
e
e
x
p
er
i
m
en
tal
r
esu
lts
w
it
h
T
B
A
L
w
h
e
n
it
r
ea
ch
es
th
e
s
a
m
e
E
E
R
.
Fro
m
th
i
s
co
m
p
ar
is
o
n
,
as
s
h
o
w
n
in
Fi
g
u
r
e
1
0
,
it
is
k
n
o
w
n
t
h
at
th
er
e
is
a
r
ed
u
ce
d
d
ata
co
llectio
n
ti
m
e
f
r
o
m
T
h
r
es
h
o
ld
B
ased
A
cti
v
e
L
ea
r
n
i
n
g
to
Op
t
i
m
ized
A
cti
v
e
L
ea
r
n
i
n
g
as
m
u
c
h
as
4
1
%.
Ap
p
ly
i
n
g
n
e
w
m
o
d
el
an
al
y
s
is
an
d
o
p
ti
m
ized
s
to
p
r
u
le
to
t
h
e
s
y
s
te
m
g
iv
e
s
f
aster
au
th
e
n
tica
tio
n
t
i
m
e
th
a
n
T
B
AL
.
T
h
e
r
es
u
lts
s
h
o
w
th
at
t
h
i
s
ap
p
r
o
ac
h
is
m
o
r
e
f
e
asib
le
f
o
r
a
u
th
e
n
ticat
io
n
p
u
r
p
o
s
e
th
a
n
t
h
e
T
B
A
L
b
ec
au
s
e
f
aste
r
au
t
h
e
n
ticatio
n
ti
m
e
g
iv
e
s
m
o
r
e
co
m
f
o
r
tab
le
to
th
e
u
s
er
s
.
Fig
u
r
e
10
.
E
E
R
an
d
Data
C
o
ll
ec
tio
n
T
im
e
C
o
m
p
ar
is
o
n
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
E
C
E
I
SS
N:
2088
-
8708
Op
timiz
ed
A
ctive
Lea
r
n
in
g
fo
r
User
’
s
B
eh
a
vio
r
Mo
d
ellin
g
B
a
s
ed
o
n
…
(
I
ka
K
u
s
u
ma
n
i
n
g
P
u
tr
i
)
511
4.
CO
NCLU
SI
O
N
I
n
t
h
is
r
esear
ch
,
w
e
i
m
p
r
o
v
e
th
e
co
n
ce
p
t
o
f
A
cti
v
e
L
ea
r
n
i
n
g
f
o
r
co
llectin
g
th
e
tr
ain
in
g
d
ata
w
it
h
an
al
y
z
in
g
ea
c
h
u
s
er
b
eh
a
v
io
r
w
h
e
n
th
e
y
o
p
er
ate
th
e
s
m
ar
tp
h
o
n
e
to
s
h
o
r
ten
th
e
d
ata
tr
ai
n
in
g
an
d
co
llecti
n
g
d
ata
ti
m
e
u
n
t
il it c
o
u
ld
b
e
ac
ce
p
tab
le
to
th
e
u
s
er
.
I
n
o
r
d
er
to
ac
h
iev
e
th
i
s
g
o
al,
w
e
d
e
v
elo
p
th
e
A
cti
v
e
L
ea
r
n
i
n
g
b
y
c
h
an
g
i
n
g
t
h
e
s
ev
er
al
p
h
as
es
in
cl
u
d
in
g
th
e
m
o
d
el
an
al
y
s
is
,
s
to
p
p
in
g
cr
iter
ia
an
d
co
m
b
in
i
n
g
it
w
i
th
t
h
e
T
h
r
esh
o
ld
-
b
ased
A
ct
iv
e
L
ea
r
n
in
g
’
s
s
to
p
r
u
le.
I
n
th
e
f
ir
s
t
p
h
ase,
w
e
c
h
an
g
i
n
g
t
h
e
s
to
p
p
in
g
cr
iter
ia
o
f
th
e
co
llectin
g
d
ata
w
it
h
c
h
ec
k
i
n
g
if
th
e
ac
c
u
r
ac
y
w
il
l
n
o
t
i
m
p
r
o
v
e
s
ig
n
i
f
ica
n
tl
y
is
t
h
e
b
etter
s
to
p
r
u
le
to
g
et
th
e
o
p
tim
a
l
r
es
u
lt.T
h
e
s
ec
o
n
d
p
h
ase,
w
e
c
h
an
g
e
th
e
m
o
d
el
a
n
al
y
s
i
s
t
h
at
u
s
ed
to
d
eter
m
i
n
e
th
e
r
atio
o
f
ea
c
h
b
eh
av
io
r
w
h
e
n
s
e
lecti
n
g
t
h
e
t
r
ain
in
g
d
ata.
We
ch
o
o
s
e
th
e
c
lo
s
est
p
o
in
t
to
t
h
e
SVM
h
y
p
e
r
p
lan
e
to
ca
lcu
lat
e
th
e
b
eh
a
v
io
r
r
atio
.
T
h
is
s
t
u
d
y
s
h
o
w
s
t
h
at
4
3
o
f
4
5
p
er
s
o
n
's
tr
ain
in
g
d
ata
ca
n
r
ed
u
ce
u
s
in
g
t
h
e
n
e
w
A
cti
v
e
L
ea
r
n
i
n
g
m
et
h
o
d
.
T
h
e
Op
ti
m
ized
A
c
tiv
e
L
ea
r
n
i
n
g
h
a
s
less
co
llecti
n
g
d
at
a
ti
m
e
t
h
a
n
T
h
r
esh
o
ld
-
b
ased
L
ea
r
n
i
n
g
an
d
B
atch
L
ea
r
n
i
n
g
.
W
e
ca
n
r
ed
u
ce
th
e
co
llectin
g
d
ata
ti
m
e
u
n
til
4
1
%
f
r
o
m
t
h
e
T
h
r
esh
o
ld
-
b
ased
A
cti
v
e
L
ea
r
n
i
n
g
.
ACK
NO
WL
E
D
G
E
M
E
NT
S
T
h
e
au
th
o
r
s
w
o
u
ld
lik
e
to
t
h
an
k
Sall
y
P
an
an
d
al
l
m
e
m
b
er
s
o
f
I
n
te
lli
g
en
t
So
f
t
w
ar
e
S
y
s
te
m
s
L
ab
o
r
ato
r
y
o
f
Natio
n
a
l
C
en
tr
al
Un
iv
er
s
it
y
T
ai
w
a
n
f
o
r
i
n
s
i
g
h
t
f
u
l
d
is
cu
s
s
io
n
s
d
u
r
i
n
g
t
h
is
r
esear
ch
a
n
d
al
s
o
r
ev
ie
w
er
s
f
o
r
th
eir
v
al
u
ab
le
co
m
m
en
t
s
.
RE
F
E
R
E
NC
E
S
[1
]
R
.
Ch
o
u
h
a
n
,
A
.
M
ish
ra
a
n
d
P
.
Kh
a
n
n
a
,
“
F
in
g
e
rp
ri
n
t
A
u
th
e
n
ti
c
a
ti
o
n
b
y
Wav
e
l
e
t
-
b
a
se
d
Dig
it
a
l
W
a
ter
m
a
rk
in
g
”
,
In
ter
n
a
t
io
n
a
l
J
o
u
rn
a
l
o
f
E
lec
trica
l
a
n
d
C
o
mp
u
ter
En
g
in
e
e
rin
g
,
V
o
l
.
2
,
N
o
.
4
,
p
p
.
5
2
3
~
5
2
8
,
2
0
1
2
.
[2
]
J.
Ha
n
,
“
F
in
g
e
rp
ri
n
t
A
u
th
e
n
ti
c
a
ti
o
n
S
c
h
e
m
e
sf
o
r
M
o
b
il
e
De
v
ice
s
”
,
In
ter
n
a
ti
o
n
a
l
J
o
u
r
n
a
l
o
f
El
e
c
trica
l
a
n
d
Co
mp
u
ter
E
n
g
in
e
e
rin
g
,
V
o
l
.
5
,
N
o
.
3
,
p
p
.
5
7
9
~
5
8
5
,
2
0
1
5
.
[3
]
O.
M
a
z
h
e
li
s,
J.
M
a
rk
u
u
la,
a
n
d
J.
V
e
ij
a
lain
e
n
,
“
A
n
in
teg
ra
ted
id
e
n
ti
ty
v
e
ri
f
ica
ti
o
n
s
y
ste
m
f
o
r
m
o
b
il
e
t
e
rm
in
a
ls”
,
In
fo
rm
a
t
io
n
M
a
n
a
g
e
me
n
t
&
Co
mp
u
ter
S
e
c
u
rity
,
v
o
l.
1
3
,
n
o
.
5
,
p
p
.
3
6
7
-
3
7
8
,
2
0
0
5
.
[4
]
S
m
a
rt
Cr
e
d
it
.
“
Co
n
su
m
e
r
Re
p
o
rts
su
rv
e
y
o
n
m
o
b
il
e
p
h
o
n
e
s
a
n
d
se
c
u
rit
y
”
,
2
0
1
1
P
re
ss
.
a
v
a
il
a
b
le
f
ro
m
:
h
tt
p
:
//
ww
w
.
s
m
a
rtcr
e
d
it
.
c
o
m
/b
lo
g
/2
0
1
1
/
0
9
/
0
2
/c
o
n
su
m
e
r
-
re
p
o
rts
-
su
r
v
e
y
o
n
-
m
o
b
il
e
-
p
h
o
n
e
s
-
a
n
d
-
se
c
u
rit
y
/
(2
0
1
1
/
1
1
/
1
5
)
.
[5
]
C.
T
h
e
riau
lt
,
“
S
u
rv
e
y
sa
y
s
7
0
%
d
o
n
'
t
p
a
ss
w
o
rd
-
p
ro
tec
t
m
o
b
il
e
s
”
,
2
0
1
1
P
re
ss
.
a
v
a
il
a
b
le
f
ro
m
:
ht
tp
:
//
n
a
k
e
d
se
c
u
rit
y
.
so
p
h
o
s.c
o
m
/
2
0
1
1
/
0
8
/
0
9
/f
re
e
so
p
h
o
s
-
m
o
b
il
e
-
se
c
u
rit
y
-
to
o
lk
it
/
(2
0
1
1
/1
1
/1
1
).
[6
]
N.
Clark
e
,
S
.
Ka
ra
tzo
u
n
i,
a
n
d
S
.
F
u
rn
e
ll
,
“
F
lex
ib
le
a
n
d
tran
sp
a
re
n
t
u
se
r
a
u
th
e
n
ti
c
a
ti
o
n
f
o
r
m
o
b
il
e
d
e
v
ice
s
”
,
IFI
P
Ad
v
a
n
c
e
s i
n
In
f
o
rm
a
ti
o
n
a
n
d
C
o
mm
u
n
ica
ti
o
n
T
e
c
h
n
o
lo
g
y
,
2
9
7
/2
0
0
9
,
1
-
1
2
,
2
0
0
9
.
[7
]
C.
C.
L
in
,
C.
C
.
C
h
a
n
g
,
a
n
d
D.
L
ian
g
,
“
A
Ne
w
No
n
-
in
tru
siv
e
A
u
th
e
n
ti
c
a
ti
o
n
A
p
p
ro
a
c
h
f
o
r
Da
ta P
ro
t
e
c
ti
o
n
Ba
se
d
o
n
M
o
u
se
Dy
n
a
m
ics
”
,
In
ter
n
a
ti
o
n
a
l
S
y
mp
o
si
u
m
o
n
Bi
o
me
trics
a
n
d
S
e
c
u
rity
T
e
c
h
n
o
lo
g
ies
,
T
a
ip
e
i,
T
a
iwa
n
,
M
a
rc
h
2
6
-
1
9
,
p
p
.
9
-
1
4
,
2
0
1
2
.
[8
]
A
.
A
.
E.
A
h
m
e
d
,
a
n
d
I.
T
ra
o
re
,
“
A
Ne
w
Bio
m
e
tri
c
T
e
c
h
n
o
lo
g
y
Ba
se
d
o
n
M
o
u
se
D
y
n
a
m
ics
”
,
I
EE
E
T
ra
n
s.
o
n
De
p
e
n
d
a
b
le a
n
d
S
e
c
u
re
Co
m
p
u
ti
n
g
,
v
o
l.
4
,
n
o
.
3
,
p
p
.
1
6
5
-
1
7
9
,
2
0
0
7
.
[9
]
H.
Ga
m
b
o
a
,
a
n
d
A
.
F
re
d
,
“
A
Us
e
r
A
u
th
e
n
ti
c
a
ti
o
n
T
e
c
h
n
ic
Us
in
g
a
Web
In
tera
c
ti
o
n
M
o
n
it
o
ri
n
g
S
y
ste
m
”
,
L
e
c
tu
re
No
tes
in
C
o
mp
u
ter
S
c
ien
c
e
(
P
a
tt
e
rn
Re
c
o
g
n
it
i
o
n
a
n
d
Im
a
g
e
A
n
a
l
y
s
is),
v
o
l.
2
6
5
2
,
p
p
.
2
4
6
-
2
5
4
,
2
0
0
3
.
[1
0
]
C.
C.
L
in
,
C.
C.
Ch
a
n
g
,
D.
R.
L
i
a
n
g
,
a
n
d
C.
H.
Ya
n
g
,
"
A
P
re
li
m
i
n
a
ry
S
tu
d
y
o
n
No
n
-
In
tru
siv
e
Us
e
r
A
u
th
e
n
ti
c
a
ti
o
n
M
e
th
o
d
Us
in
g
S
m
a
rtp
h
o
n
e
S
e
n
s
o
rs
"
,
Ap
p
li
e
d
M
e
c
h
a
n
ics
a
n
d
M
a
te
ria
ls
,
v
o
l.
2
8
4
,
p
p
.
3
2
7
0
-
3
2
7
4
,
2
0
1
3
.
[1
1
]
C.
C.
L
in
,
C.
C.
Ch
a
n
g
,
a
n
d
D.
L
ian
g
,
"
A
No
v
e
l
No
n
-
in
tru
siv
e
Us
e
r
A
u
th
e
n
ti
c
a
ti
o
n
M
e
th
o
d
Ba
se
d
o
n
T
o
u
c
h
sc
re
e
n
o
f
S
m
a
rtp
h
o
n
e
s"
,
to
a
p
p
e
a
r
in
J
o
u
rn
a
l
o
f
I
n
ter
n
e
t
T
e
c
h
n
o
l
o
g
y
,
v
o
l.
1
6
,
p
.
1
-
1
0
,
2
0
1
5
.
[1
2
]
E
.
Ch
e
n
.
"
Us
in
g
Ac
ti
v
e
L
e
a
rn
in
g
to
Co
ll
e
c
t
Us
e
r’s
Be
h
a
v
io
r
f
o
r
T
ra
in
in
g
M
o
d
e
l.
Ba
se
o
n
No
n
-
in
tr
u
siv
e
S
m
a
rtp
h
o
n
e
A
u
th
e
n
ti
c
a
ti
o
n
"
,
M
a
ste
r
T
h
e
sis,
Na
ti
o
n
a
l
Ce
n
tral
Un
i
v
e
rsit
y
,
2
0
1
5
.
[1
3
]
J.
L
iu
a
n
d
J.
H
u
,
“
Dy
n
a
m
ic
b
a
tch
p
ro
c
e
ss
in
g
in
w
o
rk
f
lo
w
s:
M
o
d
e
l
a
n
d
im
p
lem
e
n
tatio
n
”
,
F
u
t
u
re
G
e
n
e
ra
ti
o
n
Co
m
p
u
ter S
y
ste
m
s,
v
o
l.
2
3
,
n
o
.
3
,
p
p
.
3
3
8
–
3
4
7
,
2
0
0
7
.
[1
4
]
L
.
P
u
f
a
h
l
a
n
d
M
.
W
e
sk
e
,
“
Ba
tch
A
c
ti
v
it
ies
in
P
ro
c
e
ss
M
o
d
e
li
n
g
a
n
d
Ex
e
c
u
ti
o
n
”
,
in
S
e
rv
ice
-
Orie
n
ted
Co
m
p
u
ti
n
g
.
S
p
rin
g
e
r,
2
0
1
3
,
p
p
.
2
8
3
–
2
9
7
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
J
E
C
E
Vo
l.
7
,
No
.
1
,
Feb
r
u
ar
y
201
7
:
5
05
–
512
512
B
I
O
G
RAP
H
I
E
S
O
F
AUTH
O
RS
I
k
a
K
u
s
u
m
a
n
in
g
Pu
t
r
i
re
c
e
iv
e
d
th
e
M
S
c
d
e
g
re
e
in
De
p
a
rtm
e
n
t
o
f
Co
m
p
u
ter
S
c
ien
c
e
a
n
d
In
f
o
rm
a
ti
o
n
En
g
in
e
e
rin
g
,
Na
ti
o
n
a
l
Ce
n
tral
Un
iv
e
rsit
y
,
T
a
i
wa
n
in
2
0
1
6
a
s
a
n
In
tern
a
ti
o
n
a
l
Du
a
l
De
g
re
e
M
a
ste
r
stu
d
e
n
t
b
e
tw
e
e
n
Un
iv
e
rsit
y
o
f
Bra
w
ij
a
y
a
,
In
d
o
n
e
sia
a
n
d
Na
ti
o
n
a
l
Ce
n
tral
Un
iv
e
rsit
y
,
Tai
w
a
n
.
S
h
e
c
o
m
p
lete
d
h
e
r
Ba
c
h
e
lo
r
d
e
g
re
e
in
De
p
a
rtme
n
t
o
f
In
f
o
r
m
a
ti
c
s
En
g
in
e
e
rin
g
,
Un
iv
e
rsity
o
f
Br
a
w
i
ja
y
a
,
In
d
o
n
e
sia
in
2
0
1
4
.
H
er
re
se
a
rc
h
in
tere
st
a
re
a
in
th
e
a
re
a
s
o
f
so
f
t
w
a
r
e
e
n
g
in
e
e
rin
g
a
n
d
in
f
o
r
m
a
ti
o
n
tec
h
n
o
lo
g
y
,
m
o
b
il
e
d
e
v
e
lo
p
m
e
n
t
,
a
n
d
i
n
telli
g
e
n
c
e
s
y
ste
m
s
.
De
r
o
n
Lia
n
g
is
a
P
ro
f
e
ss
o
r
in
th
e
De
p
a
rtm
e
n
t
o
f
Co
m
p
u
t
e
r
S
c
ien
c
e
a
n
d
In
f
o
rm
a
ti
o
n
En
g
in
e
e
rin
g
,
Na
ti
o
n
a
l
Ce
n
tral
Un
iv
e
rsit
y
,
T
a
i
w
a
n
.
He
re
c
e
iv
e
d
a
BS
d
e
g
r
e
e
in
e
l
e
c
tri
c
a
l
e
n
g
in
e
e
rin
g
f
ro
m
Na
ti
o
n
a
l
T
a
i
wa
n
Un
iv
e
rsity
in
1
9
8
3
,
a
n
M
S
a
n
d
a
P
h
D
in
c
o
m
p
u
ter
sc
ien
c
e
f
ro
m
th
e
Un
iv
e
rsity
o
f
M
a
r
y
lan
d
a
t
C
o
ll
e
g
e
P
a
rk
,
USA
in
1
9
9
1
a
n
d
1
9
9
2
re
sp
e
c
ti
v
e
l
y
.
He
a
lso
h
o
l
d
s
jo
i
n
t
a
p
p
o
in
tm
e
n
t
w
it
h
th
e
In
stit
u
te
o
f
In
f
o
rm
a
ti
o
n
S
c
ien
c
e
(IIS
),
A
c
a
d
e
m
i
a
S
in
ica
,
T
a
ip
e
i,
T
a
i
wa
n
,
Re
p
u
b
l
ic
o
f
Ch
i
n
a
.
He
w
a
s
w
it
h
IIS
f
ro
m
1
9
9
3
ti
ll
2
0
0
1
.
Dr.
L
ian
g
’s
c
u
rre
n
t
re
se
a
rc
h
in
tere
sts
a
r
e
in
th
e
a
re
a
s
o
f
so
f
t
wa
re
f
a
u
lt
-
to
lera
n
c
e
,
s
y
ste
m
se
c
u
rit
y
,
d
istri
b
u
ted
s
y
ste
m
s,
o
b
jec
to
r
ien
ted
a
n
d
sy
st
e
m
re
li
a
b
il
it
y
a
n
a
l
y
sis.
Dr.
L
ian
g
is
a
m
e
m
b
e
r
o
f
A
CM
a
n
d
IEE
E.
S
h
o
le
h
H
a
d
i
Pra
m
o
n
o
is
a
s
e
n
io
r
lec
tu
re
r
in
De
p
a
rtm
e
n
t
o
f
El
e
c
tri
c
a
l
En
g
in
e
e
rin
g
,
Un
iv
e
rsity
o
f
Bra
w
ij
a
y
a
,
In
d
o
n
e
sia
.
He
g
o
t
Do
c
to
r
d
e
g
re
e
f
ro
m
Un
iv
e
r
sity
o
f
In
d
o
n
e
sia
,
In
d
o
n
e
sia
.
He
p
re
se
n
tl
y
w
o
rk
in
T
e
l
e
c
o
m
m
u
n
ica
ti
o
n
L
a
b
o
ra
to
ry
,
Un
iv
e
rsit
y
o
f
Bra
w
ij
a
y
a
,
In
d
o
n
e
sia
a
s
a
n
o
p
t
ica
l
tele
c
o
m
m
u
n
ica
ti
o
n
sp
e
c
ialist.
His
re
se
a
rc
h
in
tere
st
a
re
a
in
th
e
a
re
a
s
o
f
o
p
t
ica
l
tele
c
o
m
m
u
n
ica
ti
o
n
,
tec
h
n
o
lo
g
y
o
f
a
n
ten
n
a
,
d
istr
ib
u
ted
sy
st
e
m
s,
a
n
d
tele
c
o
m
m
u
n
ica
ti
o
n
a
rc
h
it
e
c
tu
re
.
Ra
h
m
a
d
w
a
ti
is
a
lec
tu
re
r
in
De
p
a
rtme
n
t
o
f
El
e
c
tri
c
a
l
En
g
in
e
e
rin
g
Un
iv
e
rsity
o
f
Bra
w
ij
a
y
a
,
In
d
o
n
e
sia
.
S
h
e
g
o
t
P
h
D
d
e
g
re
e
f
r
o
m
Un
iv
e
rsit
y
o
f
W
o
ll
o
n
g
o
n
g
,
A
u
stra
li
a
.
He
r
re
se
a
rc
h
in
tere
st
a
re
a
in
th
e
a
re
a
s o
f
Im
a
g
e
P
r
o
c
e
ss
in
g
a
n
d
I
n
telli
g
e
n
c
e
S
y
ste
m
s.
Evaluation Warning : The document was created with Spire.PDF for Python.