I
nte
rna
t
io
na
l J
o
urna
l o
f
E
lect
rica
l a
nd
Co
m
pu
t
er
E
ng
ineering
(
I
J
E
CE
)
Vo
l.
1
6
,
No
.
1
,
Feb
r
u
ar
y
20
2
6
,
p
p
.
545
~
554
I
SS
N:
2088
-
8
7
0
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijece.
v
1
6
i
1
.
pp
5
4
5
-
5
5
4
545
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ec
e.
ia
esco
r
e.
co
m
A sy
stema
tic
revi
ew of sof
tware
fault pre
diction
tec
hniques
:
mo
dels, clas
sifiers
,
a
nd da
ta p
ro
cess
ing
appro
a
ches
R.
K
a
nes
a
ra
j
Ra
m
a
s
a
my
1
,
V
enus
h
ini
Ra
j
endra
n
1
,
P
a
ra
m
eswa
ra
n Sub
ra
m
a
nia
n
2
1
F
a
c
u
l
t
y
o
f
C
o
mp
u
t
i
n
g
a
n
d
I
n
f
o
r
mat
i
c
s,
M
u
l
t
i
me
d
i
a
U
n
i
v
e
r
si
t
y
,
C
y
b
e
r
j
a
y
a
,
M
a
l
a
y
si
a
2
S
c
h
o
o
l
o
f
B
u
s
i
n
e
ss
a
n
d
M
a
n
a
g
e
me
n
t
,
C
H
R
I
S
T
U
n
i
v
e
r
si
t
y
,
P
u
n
e
,
I
n
d
i
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
u
n
1
2
,
2
0
2
5
R
ev
is
ed
Sep
2
3
,
2
0
2
5
Acc
ep
ted
No
v
2
3
,
2
0
2
5
S
o
ftwa
re
fa
u
lt
p
re
d
icti
o
n
(S
F
P
)
p
lay
s
a
c
rit
ica
l
r
o
le
i
n
imp
ro
v
in
g
so
f
twa
re
re
li
a
b
il
it
y
b
y
e
n
a
b
li
n
g
e
a
rly
d
e
te
c
ti
o
n
a
n
d
c
o
rre
c
ti
o
n
o
f
d
e
fe
c
ts.
Th
is
p
a
p
e
r
p
re
se
n
ts
a
c
o
m
p
re
h
e
n
siv
e
re
v
iew
o
f
2
5
re
c
e
n
t
a
n
d
si
g
n
ifi
c
a
n
t
st
u
d
i
e
s
o
n
S
F
P
tec
h
n
iq
u
e
s,
f
o
c
u
sin
g
o
n
d
a
ta
p
re
p
r
o
c
e
ss
in
g
stra
teg
ies
,
c
la
ss
ifi
c
a
ti
o
n
a
lg
o
rit
h
m
s,
a
n
d
th
e
ir
e
ffe
c
ti
v
e
n
e
ss
a
c
ro
ss
v
a
rio
u
s
d
a
tas
e
ts.
Th
e
re
v
iew
c
a
teg
o
rize
s
th
e
a
p
p
r
o
a
c
h
e
s
in
t
o
tra
d
it
i
o
n
a
l
sta
ti
stica
l
m
o
d
e
ls,
m
a
c
h
in
e
lea
rn
in
g
m
e
th
o
d
s,
d
e
e
p
lea
rn
i
n
g
a
rc
h
it
e
c
tu
re
s,
a
n
d
h
y
b
ri
d
t
e
c
h
n
iq
u
e
s
.
No
tab
ly
,
wra
p
p
e
r
-
b
a
se
d
fe
a
tu
re
se
lec
ti
o
n
,
n
e
u
ra
l
n
e
two
r
k
c
las
sifiers
,
a
n
d
su
p
p
o
rt
v
e
c
to
r
m
a
c
h
in
e
s
(S
VM
)
a
re
id
e
n
ti
fied
a
s
t
h
e
m
o
st
e
f
fe
c
ti
v
e
in
a
c
h
iev
in
g
h
ig
h
a
c
c
u
ra
c
y
,
p
a
rti
c
u
l
a
rly
wh
e
n
d
e
a
li
n
g
wi
th
imb
a
lan
c
e
d
o
r
n
o
isy
d
a
tas
e
ts.
Th
e
p
a
p
e
r
a
lso
h
ig
h
li
g
h
ts
a
d
v
a
n
c
e
d
a
p
p
ro
a
c
h
e
s
su
c
h
a
s
v
a
riatio
n
a
l
a
u
to
e
n
c
o
d
e
rs
(VA
E),
Ba
y
e
sia
n
c
las
sifiers
,
a
n
d
fu
z
z
y
c
lu
ste
rin
g
fo
r
fa
u
lt
p
re
d
ictio
n
.
C
o
m
p
a
ra
ti
v
e
a
n
a
l
y
sis
is
p
ro
v
i
d
e
d
t
o
a
ss
e
ss
p
e
rfo
rm
a
n
c
e
m
e
tri
c
s
su
c
h
a
s
a
c
c
u
ra
c
y
,
F
-
m
e
a
su
re
,
a
n
d
a
re
a
u
n
d
e
r
th
e
c
u
rv
e
(AU
C)
.
Th
e
f
in
d
in
g
s
su
g
g
e
st
th
a
t
n
o
sin
g
le
m
e
th
o
d
fit
s
a
ll
sc
e
n
a
rio
s,
b
u
t
a
c
o
m
b
in
a
ti
o
n
o
f
a
p
p
ro
p
riate
p
re
p
r
o
c
e
ss
in
g
a
n
d
r
o
b
u
st
c
las
sifica
ti
o
n
y
ield
s
o
p
ti
m
a
l
re
su
lt
s.
Th
is
re
v
iew
p
ro
v
id
e
s
v
a
lu
a
b
le
in
sig
h
ts
fo
r
re
se
a
rc
h
e
rs
a
n
d
p
r
a
c
ti
ti
o
n
e
rs
a
imin
g
t
o
e
n
h
a
n
c
e
so
ftwa
re
q
u
a
li
t
y
th
r
o
u
g
h
p
re
d
ict
iv
e
a
n
a
l
y
ti
c
s.
F
u
tu
re
wo
rk
sh
o
u
l
d
e
x
p
l
o
re
e
n
se
m
b
le
lea
rn
in
g
a
n
d
re
a
l
-
ti
m
e
S
F
P
sy
ste
m
s
fo
r
b
ro
a
d
e
r
a
p
p
li
c
a
b
il
i
ty
.
K
ey
w
o
r
d
s
:
C
las
s
if
icatio
n
alg
o
r
ith
m
s
Data
p
r
ep
r
o
ce
s
s
in
g
m
et
h
o
d
s
Ma
ch
in
e
lear
n
in
g
tech
n
iq
u
es
So
f
twar
e
f
au
lt p
r
e
d
ictio
n
So
f
twar
e
r
eliab
ilit
y
en
g
in
ee
r
in
g
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
:
R
.
Kan
esar
aj
R
am
asam
y
Facu
lty
o
f
C
o
m
p
u
tin
g
an
d
I
n
f
o
r
m
atics,
Mu
ltime
d
ia
Un
iv
er
s
ity
Per
s
iar
an
Mu
ltime
d
ia,
6
3
1
0
0
C
y
b
er
jay
a,
Selan
g
o
r
,
Ma
lay
s
ia
E
m
ail: r
.
k
an
esar
aj@
m
m
u
.
ed
u
.
m
y
1.
I
NT
RO
D
UCT
I
O
N
Dea
lin
g
with
s
o
f
twar
e
f
a
u
lts
an
d
f
ailu
r
es
is
an
im
p
o
r
tan
t
ta
s
k
in
th
e
s
o
f
twar
e
q
u
ality
a
n
d
r
eliab
ilit
y
to
p
ic.
R
aju
et
a
l.
[
1
]
d
ef
i
n
ed
s
o
f
twar
e
f
au
lt
as
a
d
ef
ec
t
in
s
o
f
twar
e
co
d
e
r
esu
ltin
g
i
n
s
o
f
twar
e
f
ailu
r
e
d
u
r
i
n
g
ex
ec
u
tio
n
.
W
h
er
ea
s
,
Gu
p
ta
et
a
l.
[
2
]
m
e
n
tio
n
ed
th
at
s
y
s
te
m
f
ailu
r
e
is
wh
en
th
e
s
y
s
tem
is
n
o
t
b
eh
av
in
g
th
e
way
it
s
h
o
u
ld
b
e
.
Failu
r
es
in
clo
u
d
s
y
s
tem
s
co
u
ld
ca
u
s
e
s
er
io
u
s
p
r
o
b
le
m
s
s
u
ch
as
d
ata
d
a
m
ag
e,
ac
co
r
d
in
g
to
L
o
u
et
a
l.
[
3
]
.
So
f
twar
e
f
ailu
r
es
in
s
o
f
twar
e
s
y
s
tem
s
ca
n
lead
to
p
er
f
o
r
m
an
ce
d
e
g
r
ad
atio
n
an
d
p
o
o
r
s
er
v
ice,
wh
ich
n
ee
d
t
o
b
e
r
e
d
u
ce
d
to
av
o
id
cu
s
to
m
e
r
d
is
s
atis
f
ac
tio
n
an
d
in
cr
ea
s
in
g
co
s
ts
by
Pit
ak
r
at
et
a
l.
[
4
]
.
I
t
is
d
if
f
icu
lt
f
o
r
ex
p
e
r
ien
ce
d
p
r
o
g
r
am
m
er
s
to
av
o
id
s
o
f
twar
e
f
a
u
lts
in
s
o
f
twar
e
p
r
o
jects.
C
o
m
m
o
n
ly
,
C
h
atter
jee
an
d
Ma
ji
[
5
]
s
tated
t
h
at
th
e
d
a
ta
f
ailu
r
e
is
av
ailab
le
d
u
r
in
g
th
e
test
in
g
o
r
d
e
p
lo
y
m
e
n
t
p
h
ase.
W
h
en
a
p
r
o
b
lem
o
cc
u
r
r
e
d
,
th
e
s
y
s
tem
lo
g
wo
u
l
d
b
e
t
h
e
f
ir
s
t
p
lace
p
eo
p
le
wo
u
ld
g
o
to
d
etec
t
th
e
ca
u
s
e
o
f
f
ailu
r
e
in
J
au
k
et
a
l.
[
6
]
.
Ho
we
v
er
,
Su
n
et
a
l.
[
7
]
s
tated
th
at
th
e
am
o
u
n
t
o
f
f
ail
u
r
e
d
ata
is
p
r
ac
tically
m
u
ch
le
s
s
th
an
n
o
n
-
f
ailu
r
e
d
ata
in
s
o
f
twar
e
an
d
r
esu
ltin
g
in
an
im
b
alan
ce
d
d
ata
d
is
t
r
ib
u
tio
n
,
wh
ich
m
a
y
ca
u
s
e
p
o
o
r
p
er
f
o
r
m
a
n
ce
o
f
s
o
f
twar
e
p
r
ed
ictio
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
1
6
,
No
.
1
,
Feb
r
u
ar
y
20
2
6
:
5
4
5
-
554
546
2.
L
I
T
E
R
AT
U
RE
R
E
VI
E
W
2
.
1
.
At
t
ribute
s
elec
t
io
n,
s
a
mp
lin
g
t
ec
hn
o
lo
g
ies,
a
nd
ens
e
m
ble a
lg
o
rit
hm
(
ASRA)
m
o
del
D
i
n
g
e
t
a
l
.
[
8
]
p
r
o
p
o
s
e
a
n
A
SR
A
m
o
d
e
l
t
h
at
h
a
s
a
c
o
m
b
i
n
a
t
i
o
n
o
f
a
t
t
r
i
b
u
t
e
s
e
le
c
t
io
n
,
s
a
m
p
l
i
n
g
t
e
c
h
n
o
l
o
g
i
e
s
a
n
d
e
n
s
e
m
b
l
e
a
l
g
o
r
i
t
h
m
s
t
o
s
o
l
v
e
t
h
e
i
m
b
a
l
a
n
c
ed
d
a
t
a
s
et
s
p
r
o
b
l
e
m
s
.
T
h
e
ASR
A
m
o
d
e
l
b
o
r
r
o
w
e
d
t
h
e
c
h
i
-
s
q
u
a
r
e
t
es
t
o
f
at
t
r
i
b
u
t
e
s
el
e
c
t
i
o
n
.
T
h
e
s
e
l
ec
t
i
o
n
al
g
o
r
i
t
h
m
w
o
r
k
s
li
k
e
t
h
is
:
f
i
r
s
t
,
t
h
e
c
a
n
d
i
d
a
t
e
’
s
f
e
at
u
r
e
s
u
b
s
e
t
i
s
g
e
n
e
r
a
t
e
d
,
t
h
e
n
t
h
e
s
u
b
s
e
ts
a
r
e
e
v
al
u
a
t
e
d
a
n
d
f
i
lt
e
r
e
d
u
s
i
n
g
a
f
il
t
e
r
m
o
d
e
l
.
T
h
e
n
,
t
h
e
s
y
n
t
h
e
t
ic
m
i
n
o
r
i
t
y
o
v
e
r
-
s
a
m
p
l
i
n
g
t
e
c
h
n
i
q
u
e
(
SM
O
T
E
)
a
l
g
o
r
i
t
h
m
o
v
e
r
-
s
a
m
p
l
i
n
g
t
e
c
h
n
i
q
u
e
a
n
d
t
h
e
r
e
s
a
m
p
l
e
al
g
o
r
i
t
h
m
u
n
d
e
r
-
s
a
m
p
l
i
n
g
t
e
c
h
n
i
q
u
e
a
r
e
p
e
r
f
o
r
m
e
d
.
L
a
s
t
l
y
,
t
h
e
a
d
a
p
ti
v
e
b
o
o
s
t
i
n
g
(
A
d
aB
o
o
s
t
)
a
l
g
o
r
i
t
h
m
is
u
s
e
d
.
T
h
e
e
n
s
e
m
b
l
e
a
l
g
o
r
i
t
h
m
f
u
n
c
ti
o
n
is
t
o
t
u
r
n
a
we
a
k
c
l
ass
i
f
i
e
r
i
n
to
a
s
t
r
o
n
g
o
n
e
.
F
r
o
m
t
h
e
e
x
p
e
r
i
m
e
n
t
,
t
h
e
A
SR
A
a
l
g
o
r
i
t
h
m
h
a
s
a
h
i
g
h
v
a
l
u
e
o
f
F
-
m
e
a
s
u
r
es
a
n
d
a
r
e
a
u
n
d
e
r
t
h
e
c
u
r
v
e
(
A
UC
)
,
w
h
i
c
h
a
r
e
m
o
r
e
t
h
a
n
0
.
7
a
n
d
0
.
8
,
r
e
s
p
e
c
t
i
v
el
y
,
o
n
a
l
l
t
h
e
d
at
a
s
e
ts
c
o
m
p
a
r
e
d
t
o
t
h
e
o
r
i
g
i
n
a
l
a
n
d
SR
A
d
es
i
g
n
.
2
.
2
.
Sp
ira
l lif
e
cy
cle
m
o
del
-
ba
s
ed
B
a
y
esia
n c
la
s
s
if
ica
t
io
n (
SL
M
B
C)
Dh
an
ajay
an
an
d
Pil
lai
[
9
]
p
r
o
p
o
s
e
a
s
p
ir
al
life
cy
cle
m
o
d
el
-
b
ased
B
ay
esian
class
if
icatio
n
tech
n
iq
u
e
wh
er
e
th
e
s
p
ir
al
life
cy
cle
is
in
teg
r
ated
with
th
e
B
ay
esian
class
if
icatio
n
with
th
e
h
elp
o
f
th
e
r
o
b
u
s
t
s
im
ilar
ity
clu
s
ter
in
g
tech
n
iq
u
e
(
R
SC
)
.
T
h
is
m
eth
o
d
h
as
4
p
h
ases
.
T
h
e
f
ir
s
t
p
h
ase
is
to
p
in
p
o
in
t
th
e
o
b
jectiv
e,
f
u
n
ctio
n
ality
,
alter
n
ativ
es
an
d
co
n
s
tr
ain
ts
o
f
th
e
s
o
f
twar
e
p
r
o
d
u
ct.
T
h
en
,
th
e
alter
n
ativ
es
will
b
e
ev
alu
ated
.
T
h
e
th
ir
d
p
h
ase
is
th
e
d
ev
elo
p
m
en
t
an
d
test
in
g
p
h
ase
an
d
l
astl
y
is
th
e
p
lan
n
in
g
p
h
ase
f
o
r
th
e
n
ex
t
iter
atio
n
.
T
h
en
,
th
e
s
o
f
twar
e
r
eliab
ilit
y
m
o
d
el
is
p
er
f
o
r
m
ed
,
f
o
llo
we
d
b
y
th
e
B
ay
esian
clas
s
if
icatio
n
m
o
d
el.
T
o
p
r
ed
ict
th
e
f
ailu
r
e,
we
will
lo
o
k
at
th
e
b
ig
g
est
p
o
s
ter
io
r
p
r
o
b
a
b
ilit
y
.
Af
ter
class
if
y
in
g
th
e
m
o
d
u
le,
r
o
b
u
s
t
s
im
ilar
clu
s
ter
in
g
(
R
SC
)
is
ca
r
r
ied
o
u
t.
Usi
n
g
th
e
m
in
im
u
m
d
is
tan
ce
m
ea
s
u
r
e
o
f
th
e
s
im
ilar
f
ea
tu
r
es,
a
f
ew
clu
s
ter
s
ar
e
g
en
er
ate
d
.
Fro
m
t
h
e
ex
p
e
r
im
en
t,
th
e
SLM
B
C
ac
h
iev
ed
0
.
5
2
p
er
ce
n
t
i
n
th
e
d
etec
tio
n
o
f
f
au
lt
y
m
o
d
u
les
wh
ich
ar
e
n
o
t
f
au
lty
,
0
.
0
0
5
in
f
alse
p
o
s
itiv
e
r
ate
(
FP
R
)
an
d
0
.
0
2
i
n
th
e
o
v
er
all
er
r
o
r
r
ate.
T
h
e
f
alse
n
eg
ativ
e
r
ate
(
FNR
)
,
FP
R
an
d
o
v
er
all
er
r
o
r
r
ate
ar
e
lo
w
u
s
in
g
SLM
B
C
co
m
p
ar
ed
to
o
th
er
s
.
2
.
3
.
M
et
ric
ba
s
ed
o
n neura
l net
wo
rk
cla
s
s
if
ier
J
ay
an
th
i
an
d
Flo
r
en
ce
[
1
0
]
p
r
o
p
o
s
e
an
in
teg
r
atio
n
o
f
p
r
in
ci
p
al
co
m
p
o
n
en
t
an
aly
s
is
(
PC
A
)
,
a
s
ch
em
e
o
f
f
ea
tu
r
e
r
ed
u
ctio
n
with
th
e
ap
p
licatio
n
o
f
a
n
eu
r
al
n
etwo
r
k
-
b
ased
class
if
icatio
n
tech
n
i
q
u
e.
PC
A
wo
r
k
s
in
s
u
ch
a
way
b
y
a
d
d
in
g
s
ec
o
n
d
-
o
r
d
er
m
o
m
e
n
t
co
m
p
u
tatio
n
t
o
an
y
r
an
d
o
m
v
ec
to
r
’
s
c
h
ar
ac
ter
is
tics
.
T
h
e
PC
A
d
ata
r
ec
o
n
s
tr
u
ctio
n
m
ig
h
t
h
a
v
e
er
r
o
r
s
;
th
e
r
ef
o
r
e
,
th
e
PC
A
is
im
p
r
o
v
ed
b
y
in
teg
r
atin
g
m
a
x
im
u
m
-
lik
elih
o
o
d
esti
m
atio
n
.
T
h
en
,
n
e
u
r
al
n
etw
o
r
k
s
ar
e
im
p
lem
en
te
d
.
T
h
e
n
eu
r
al
n
etwo
r
k
h
as
th
r
ee
lay
er
s
an
d
th
e
s
o
f
twar
e
d
ata
will
b
e
p
r
o
ce
s
s
ed
b
y
r
ef
e
r
r
in
g
t
o
its
weig
h
ts
.
At
ea
ch
l
ay
er
,
an
in
p
u
t
o
f
n
eu
r
o
n
s
will
b
e
g
i
v
en
af
te
r
th
e
weig
h
ts
h
av
e
b
ee
n
a
d
ju
s
ted
f
o
llo
win
g
th
e
r
eq
u
ir
e
m
en
t.
T
o
g
et
th
e
f
in
al
r
esu
lt,
all
th
e
in
p
u
t
s
will
b
e
m
u
ltip
lied
b
y
th
eir
r
esp
ec
tiv
e
weig
h
ts
.
T
h
e
ex
p
er
im
e
n
t
u
s
ed
f
o
u
r
d
a
tasets
,
wh
ich
ar
e
KC
I
,
J
MI
,
PC
3
an
d
PC
4
.
T
h
e
ac
cu
r
ac
y
o
f
t
h
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
is
as
f
o
llo
ws:
KC
I
wi
th
8
6
.
9
1
%,
J
M1
with
8
3
.
0
3
%,
PC
3
with
8
9
%
an
d
last
ly
PC
4
with
9
3
.
6
4
%.
2
.
4
.
G
re
y
s
y
s
t
em
t
heo
ry
-
ba
s
ed
m
et
ho
d
Ma
o
[
1
1
]
p
r
o
p
o
s
ed
a
p
r
ed
icti
o
n
f
r
am
ewo
r
k
b
ased
o
n
a
g
r
e
y
m
o
d
el.
T
h
e
g
r
ey
m
o
d
el
is
co
m
b
in
ed
to
g
eth
er
with
in
ter
v
al
p
r
ed
icti
o
n
o
f
s
o
f
twar
e
f
au
lts
an
d
p
r
e
d
ictio
n
o
f
f
au
lt
n
u
m
b
er
b
ased
o
n
r
elate
d
f
ac
to
r
s
.
T
h
e
g
r
ey
th
e
o
r
y
is
u
s
ed
to
g
et
th
e
p
o
ten
tial
law
o
f
a
d
ata
s
et
th
r
o
u
g
h
m
in
i
n
g
.
T
h
e
p
r
o
c
ess
is
ca
lled
g
r
ey
s
eq
u
en
ce
g
en
e
r
atio
n
.
T
h
en
,
t
h
r
o
u
g
h
s
o
m
e
o
p
er
atio
n
,
th
e
r
a
n
d
o
m
n
ess
o
f
th
e
g
r
ey
s
eq
u
en
ce
ca
n
b
e
r
e
d
u
ce
d
.
A
n
ew
d
ata
s
eq
u
en
ce
ca
n
b
e
o
b
tain
ed
b
y
ap
p
ly
in
g
a
t
r
an
s
f
o
r
m
atio
n
o
p
er
atio
n
(
g
r
e
y
s
eq
u
e
n
c
e
g
en
er
atio
n
)
t
o
th
e
g
r
ey
s
eq
u
en
ce
.
T
h
e
n
ew
d
ata
is
ca
lled
a
tr
an
s
f
o
r
m
d
ata
s
eq
u
en
ce
.
A
p
r
ed
ictio
n
ca
n
b
e
em
p
lo
y
ed
wh
e
n
a
r
elativ
e
lev
el
is
r
ea
ch
ed
th
r
o
u
g
h
th
e
s
m
o
o
th
n
ess
o
f
th
e
s
eq
u
en
ce
.
I
n
g
r
ey
th
e
o
r
y
,
g
r
ey
m
o
d
ellin
g
is
u
s
ed
to
ex
p
r
ess
s
eq
u
en
ce
s
u
s
in
g
a
p
p
r
o
x
im
ate
d
if
f
er
en
tial
e
q
u
atio
n
s
.
GM
(
1
,
1
)
is
u
s
ed
i
n
th
is
m
et
h
o
d
.
T
h
en
,
b
y
u
s
in
g
GM
,
th
e
f
au
lt
n
u
m
b
er
ca
n
b
e
p
r
ed
icted
.
T
o
p
r
ed
ict
th
e
in
t
er
v
al
o
f
a
f
au
lt
n
u
m
b
er
,
p
r
o
p
o
r
tio
n
al
b
an
d
-
b
ased
an
d
d
ev
elo
p
m
en
t
-
b
an
d
m
et
h
o
d
s
ar
e
u
s
ed
.
T
h
e
ap
p
r
o
ac
h
is
b
ased
o
n
th
e
m
in
im
u
m
an
d
m
ax
im
u
m
f
r
o
m
th
e
s
eq
u
en
ce
,
in
a
d
d
itio
n
to
a
f
e
w
s
tep
s
th
at
n
ee
d
to
b
e
d
o
n
e.
T
h
e
m
eth
o
d
is
p
r
o
v
en
t
o
r
ed
u
ce
th
e
c
o
s
t
o
f
m
ain
ten
an
ce
an
d
allo
w
th
e
o
r
g
an
izatio
n
to
g
et
b
etter
id
ea
s
o
n
h
o
w
to
h
a
n
d
le
f
ailu
r
es.
2
.
5
.
F
uzzy
rules a
nd
da
t
a
a
na
ly
s
is
-
ba
s
ed
m
et
ho
d
Din
g
et
a
l.
[
1
2
]
p
r
o
p
o
s
ed
n
ew
o
n
lin
e
p
r
ed
ictio
n
tech
n
i
q
u
es
u
s
in
g
f
u
zz
y
r
u
les
an
d
d
ata
an
a
ly
s
is
.
T
h
is
m
eth
o
d
h
as
th
r
ee
p
h
ases
:
th
e
f
ir
s
t
o
n
e
is
th
e
o
n
lin
e
tr
ain
in
g
,
th
e
r
eq
u
ir
em
en
t
d
o
c
u
m
en
tati
o
n
a
n
d
th
e
r
u
n
n
in
g
o
f
th
e
s
y
s
tem
.
T
h
is
is
to
g
et
th
e
lo
g
f
ile
co
r
r
esp
o
n
d
in
g
t
o
th
e
s
am
p
lin
g
tim
e.
T
h
e
s
e
co
n
d
p
h
ase
is
th
e
p
r
ed
ictio
n
m
o
d
el
b
u
ild
in
g
,
wh
er
e
a
f
u
zz
y
r
u
le
is
ap
p
lied
to
g
et
th
e
v
ar
iab
les' r
elatio
n
s
h
ip
an
d
th
e
ev
o
lu
tio
n
ar
y
tr
en
d
u
s
in
g
th
e
au
to
r
e
g
r
ess
iv
e
in
teg
r
ated
m
o
v
in
g
av
er
ag
e
(
AR
I
MA
)
m
o
d
el.
T
h
e
last
p
h
ase
is
th
e
f
ailu
r
e
p
r
ed
ictio
n
,
wh
e
r
e
b
o
th
th
e
v
alu
es
f
r
o
m
th
e
AR
I
MA
m
o
d
el
ar
e
co
m
p
ar
e
d
with
th
e
f
u
zz
y
r
u
le.
I
f
th
e
d
if
f
er
e
n
ce
b
etwe
en
th
e
v
alu
es
e
x
ce
ed
s
,
th
en
th
er
e
wo
u
ld
b
e
an
e
r
r
o
r
.
Fro
m
th
e
ex
p
er
im
en
t
th
at
was
co
n
d
u
cted
o
n
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
A
s
ystema
tic
r
ev
ie
w
o
f so
ftw
a
r
e
fa
u
lt p
r
ed
ictio
n
…
(
R
.
K
a
n
e
s
a
r
a
j R
a
ma
s
a
my
)
547
m
u
ltip
le
m
o
n
ito
r
e
d
v
a
r
iab
les,
it
s
h
o
ws
th
at
th
e
p
r
o
p
o
s
ed
m
e
th
o
d
was
ab
le
to
g
et
2
6
T
P
f
r
o
m
2
8
n
u
m
b
er
s
o
f
f
ailu
r
es in
jecte
d
,
wh
ich
is
o
n
ly
2
FP
p
r
ed
icted
wr
o
n
g
ly
.
2
.
6
.
E
nerg
y
-
ba
s
ed
a
no
m
a
ly
det
ec
t
io
n
Mo
n
n
i
an
d
Pezz
e
[
1
3
]
p
r
esen
ted
a
n
ew
ap
p
r
o
ac
h
o
f
en
er
g
y
-
b
ased
m
o
d
els
to
p
r
ed
ict
f
ailu
r
es
b
ased
o
n
th
e
o
b
s
er
v
atio
n
o
f
an
al
o
g
ies
am
o
n
g
co
m
p
lex
s
o
f
twar
e
s
y
s
tem
s
,
p
h
y
s
ical
s
y
s
tem
s
,
an
d
n
etwo
r
k
s
.
T
h
e
f
ea
s
ib
ilit
y
o
f
th
e
ap
p
r
o
ac
h
is
ev
alu
ated
to
r
ev
ea
l
s
o
m
e
p
r
elim
in
ar
y
r
esu
lts
b
y
m
ea
s
u
r
in
g
th
e
p
r
ec
is
io
n
o
f
th
e
r
estricte
d
B
o
ltzm
an
n
m
ac
h
in
e
(
R
B
M)
u
s
ed
in
u
s
in
g
m
o
d
el
s
to
r
ev
ea
l
f
ailu
r
e
-
p
r
o
n
e
an
o
m
alies.
T
h
e
r
esu
lts
s
u
g
g
ested
th
at
r
e
v
ea
lin
g
co
ll
ec
tiv
e
an
o
m
alies
o
f
k
e
y
p
e
r
f
o
r
m
an
ce
in
d
icato
r
s
(
KPI
s
)
v
alu
es
m
ay
p
r
ed
ict
f
ailu
r
es
in
co
m
p
lex
s
o
f
twar
e
s
y
s
tem
s
.
KPI
s
ar
e
th
e
an
o
m
aly
d
etec
to
r
s
f
o
u
n
d
in
m
an
y
d
i
f
f
er
en
t
p
ar
ts
o
f
th
e
s
o
f
twar
e
s
y
s
tem
s
th
at
ar
e
u
s
ed
to
co
llect
v
ar
io
u
s
m
etr
ics.
T
h
e
en
er
g
y
-
b
ased
a
p
p
r
o
ac
h
s
u
r
m
o
u
n
ts
lim
itatio
n
s
o
f
s
ev
er
al
d
if
f
er
e
n
t
ap
p
r
o
ac
h
es,
wh
ich
ar
e
s
ig
n
atu
r
e
-
b
ased
ap
p
r
o
ac
h
es
as
well
as
s
ee
d
ed
an
d
n
o
n
-
s
ee
d
ed
d
ata
-
d
r
iv
e
n
ap
p
r
o
ac
h
es.
2
.
7
.
D
e
e
p
l
e
a
rn
in
g
t
e
c
hn
iq
u
e
-
b
a
s
ed
mo
d
e
l
c
a
ll
e
d
V
AE
S
u
n
e
t
a
l
.
[
7
]
p
r
o
p
o
s
e
d
a
d
e
e
p
l
e
a
r
n
i
n
g
t
e
c
h
n
i
q
u
e
o
f
a
g
e
n
e
r
a
t
i
v
e
m
o
d
e
l
c
a
l
l
e
d
t
h
e
v
a
r
i
a
t
i
o
n
a
l
a
u
t
o
e
n
c
o
d
e
r
(
V
A
E
)
m
e
t
h
o
d
f
o
r
p
r
e
d
i
c
t
i
n
g
s
o
f
t
w
a
r
e
f
a
u
l
ts
.
T
h
e
p
r
o
p
o
s
e
d
m
o
d
e
l
i
s
u
s
e
d
t
o
g
en
e
r
a
t
e
n
e
w
s
a
m
p
l
e
s
o
f
f
a
i
l
u
r
e
d
at
a
t
o
o
v
e
r
c
o
m
e
t
h
e
i
m
b
a
l
a
n
c
e
d
d
at
a
o
f
s
o
f
t
wa
r
e
f
au
l
t
s
,
i
n
w
h
i
c
h
t
h
e
r
e
is
m
o
r
e
f
a
ilu
r
e
d
a
t
a
t
o
i
n
d
i
ca
t
e
t
h
e
f
a
i
l
u
r
e
m
o
d
u
l
e
(
m
i
n
o
r
i
t
y
)
t
h
a
n
n
o
n
-
f
a
i
l
u
r
e
d
a
t
a
(
m
a
j
o
r
i
t
y
)
t
o
i
n
d
i
c
a
t
e
t
h
e
n
o
n
-
f
a
i
l
u
r
e
m
o
d
u
l
e
.
T
h
u
s
,
t
h
e
V
A
E
i
s
d
es
i
g
n
e
d
t
o
b
al
a
n
c
e
t
h
e
d
a
ta
s
e
ts
a
n
d
t
o
i
m
p
r
o
v
e
t
h
e
a
c
c
u
r
a
c
y
o
f
t
h
e
c
l
ass
i
f
i
e
r
.
I
n
s
h
o
r
t
,
d
a
t
a
p
r
o
c
e
s
s
i
n
g
,
as
w
e
l
l
a
s
V
A
E
a
n
d
P
as
s
m
e
t
h
o
d
s
,
h
a
v
e
b
e
e
n
u
s
e
d
i
n
t
h
e
e
x
p
e
r
im
e
n
t
,
a
n
d
s
e
v
e
r
al
m
e
t
r
i
c
s
h
a
v
e
b
e
e
n
c
h
o
s
e
n
f
o
r
t
h
e
r
e
s
u
l
ts
e
v
a
l
u
a
t
i
o
n
.
I
n
c
o
n
c
l
u
s
i
o
n
,
t
h
e
r
e
s
u
l
t
s
r
e
p
o
r
t
e
d
t
h
a
t
t
h
e
u
t
i
l
i
z
at
i
o
n
o
f
t
h
e
V
A
E
m
e
t
h
o
d
i
m
p
r
o
v
e
s
t
h
e
a
b
i
l
i
t
y
t
o
p
r
e
d
i
c
t
f
a
i
l
u
r
e
d
at
a
w
h
il
e
t
h
e
p
r
e
d
i
c
t
i
o
n
o
f
n
o
n
-
f
a
i
l
u
r
e
d
a
t
a
is
b
e
i
n
g
p
e
r
f
o
r
m
e
d
.
2
.
8
.
B
a
y
esia
n
belief
net
wo
rk
-
ba
s
ed
m
o
del
A
s
tu
d
y
b
y
C
h
atter
jee
an
d
Ma
ji
[
5
]
ex
p
lain
ed
a
B
ay
esian
-
b
a
s
ed
m
o
d
el,
d
e
v
elo
p
e
d
to
p
r
ed
i
ct
th
e
n
et
n
u
m
b
er
o
f
f
a
u
lts
d
u
r
in
g
th
e
e
ar
ly
d
e
v
elo
p
m
e
n
t
p
h
ase
o
f
s
o
f
twar
e.
First
o
f
all,
th
e
p
r
o
p
o
s
ed
B
ay
esian
b
elief
n
etwo
r
k
is
a
d
ir
ec
ted
ac
y
clic
g
r
ap
h
co
n
s
is
tin
g
o
f
n
o
d
es,
ea
ch
is
ass
o
ciate
d
with
a
n
o
d
e
p
r
o
b
a
b
ilit
y
tab
le
o
r
n
o
d
e
p
r
o
b
ab
ilit
y
tab
le
(
NPT)
,
wh
ich
h
as
th
e
v
alu
es
o
f
co
n
d
itio
n
al
p
r
o
b
ab
ilit
y
an
d
th
e
e
x
p
ec
ted
f
au
lt
in
d
e
x
.
Seco
n
d
,
a
ty
p
e
o
f
f
u
zz
y
c
o
n
tr
o
l
s
y
s
tem
,
ca
lled
an
in
ter
v
al
ty
p
e
-
2
f
u
zz
y
lo
g
ic
s
y
s
tem
,
is
u
s
ed
to
ca
lcu
late
th
e
p
r
o
b
a
b
ilit
y
v
alu
es
o
f
th
e
m
o
d
el.
B
esid
es,
an
ar
tific
ial
n
eu
r
al
n
etwo
r
k
(
ANN)
is
ap
p
lied
to
id
en
tify
th
e
o
u
t
p
u
t
f
r
o
m
th
e
in
p
u
t
o
f
t
h
e
d
ata
f
r
o
m
s
im
ilar
o
r
ea
r
lier
p
r
o
jects.
T
h
ir
d
,
s
ix
m
etr
ics
ar
e
u
s
ed
to
im
p
lem
en
t
th
e
p
r
o
p
o
s
ed
m
eth
o
d
o
lo
g
y
.
So
f
t
war
e
m
etr
ics
h
av
e
q
u
alitativ
e
in
f
o
r
m
atio
n
ab
o
u
t
th
e
s
o
f
tw
ar
e
d
u
r
in
g
its
ea
r
ly
p
h
ase.
I
n
s
u
m
,
th
e
p
r
o
p
o
s
ed
m
o
d
el
ca
n
p
r
ed
ict
to
tal
f
a
u
lts
in
s
o
f
twar
e.
2
.
9
.
S
u
pp
o
r
t
v
e
c
t
o
r
ma
c
h
in
e
c
l
a
s
s
i
f
i
e
r
A
s
t
u
d
y
b
y
R
a
j
u
e
t
a
l
.
[
1
]
p
r
o
p
o
s
e
d
a
w
o
r
k
t
h
a
t
u
s
e
s
a
s
u
p
p
o
r
t
v
e
c
t
o
r
m
a
c
h
i
n
e
(
S
VM
)
c
l
a
s
s
i
f
ie
r
t
o
c
l
a
s
s
i
f
y
f
a
u
l
t
a
n
d
n
o
n
-
f
a
u
l
t
y
m
o
d
u
l
e
s
.
A
l
o
n
g
w
i
t
h
S
V
M
,
th
e
a
u
t
h
o
r
s
a
l
s
o
i
m
p
l
e
m
e
n
t
e
d
f
e
a
t
u
r
e
e
x
t
r
a
c
t
i
o
n
,
o
r
k
n
o
w
n
a
s
a
tt
r
i
b
u
t
e
s
e
le
c
t
i
o
n
,
t
o
f
i
n
d
a
p
p
r
o
p
r
i
a
t
e
f
e
at
u
r
e
s
f
o
r
t
h
e
c
l
as
s
i
f
i
c
at
i
o
n
m
o
d
e
l
.
I
n
t
e
g
r
a
t
i
n
g
a
f
e
at
u
r
e
e
x
t
r
a
c
t
i
o
n
m
et
h
o
d
w
i
t
h
a
c
la
s
s
i
f
i
e
r
(
S
V
M
)
i
s
a
l
s
o
c
a
l
led
t
h
e
w
r
a
p
p
e
r
a
p
p
r
o
a
c
h
.
O
v
e
r
a
l
l
,
t
h
e
p
r
o
p
o
s
e
d
m
e
t
h
o
d
o
l
o
g
y
f
o
l
l
o
w
e
d
s
i
m
p
l
e
s
t
e
p
s
t
o
p
r
e
d
i
c
t
t
h
e
s
o
f
tw
a
r
e
f
a
u
l
t
s
:
d
e
f
e
c
t
e
d
d
a
ta
s
et
‘
s
’
as
i
n
p
u
t
;
1
:
F
e
a
t
u
r
e
e
x
t
r
a
c
t
i
o
n
m
e
t
h
o
d
i
s
u
s
e
d
i
n
p
r
e
-
p
r
o
c
e
s
s
o
f
t
h
e
d
a
t
a
s
e
t
;
2
:
1
0
-
f
o
l
d
c
r
o
s
s
v
a
l
i
d
a
t
i
o
n
i
s
a
p
p
l
ie
d
b
y
d
i
v
i
d
i
n
g
t
h
e
d
a
t
a
s
e
t
i
n
t
o
t
r
a
i
n
i
n
g
m
o
d
e
l
a
n
d
t
e
s
t
i
n
g
m
o
d
e
l
;
3
:
S
V
M
cl
as
s
if
i
e
r
i
s
u
s
e
d
f
o
r
c
l
a
s
s
i
f
i
c
at
i
o
n
.
T
h
e
d
a
t
a
s
e
ts
u
s
e
d
as
t
h
e
i
n
p
u
t
t
o
t
h
e
p
r
o
p
o
s
e
d
m
o
d
e
l
a
r
e
C
M
1
,
J
M
1
,
KC
1
,
K
C
2
,
PC
1
,
a
n
d
D
A
T
A
T
R
I
E
V
E
.
E
a
c
h
d
a
t
a
s
e
t
h
a
s
a
g
o
o
d
n
u
m
b
e
r
o
f
i
n
s
t
a
n
c
e
s
i
n
d
i
c
at
i
n
g
s
o
f
t
w
a
r
e
m
o
d
u
l
e
s
a
n
d
a
ls
o
h
a
s
s
e
v
e
r
al
s
o
f
t
w
a
r
e
m
e
t
r
i
cs
,
w
h
i
c
h
c
a
n
h
e
l
p
t
o
i
d
e
n
t
i
f
y
f
a
u
l
ts
i
n
t
h
e
e
x
i
s
t
i
n
g
s
o
f
t
w
a
r
e
m
o
d
u
l
e
s
.
F
o
r
t
h
e
e
x
p
er
i
m
e
n
t
a
t
i
o
n
p
r
o
c
es
s
,
t
h
e
C
K
-
m
e
t
r
i
c
s
a
r
e
a
p
p
l
i
e
d
t
o
t
h
e
d
a
t
a
s
et
s
KC
1
a
n
d
K
C
2
s
in
c
e
b
o
t
h
a
r
e
r
e
l
a
t
e
d
t
o
t
h
e
o
b
j
e
c
t
-
o
r
i
e
n
t
e
d
a
p
p
r
o
a
c
h
.
T
h
e
n
,
a
n
e
x
p
e
r
i
m
e
n
t
i
s
p
e
r
f
o
r
m
e
d
o
n
a
l
l
t
h
e
d
a
t
a
s
ets
t
o
e
v
a
l
u
a
te
t
h
e
p
a
r
a
m
e
te
r
s
,
w
h
i
c
h
a
r
e
a
c
c
u
r
a
c
y
,
s
e
n
s
i
t
i
v
ity
,
c
o
m
p
l
e
t
e
n
e
s
s
o
r
s
p
e
c
i
f
i
ci
t
y
,
c
u
t
-
o
f
f
o
r
p
r
e
c
i
s
i
o
n
,
a
n
d
F
-
m
e
a
s
u
r
e
.
I
n
c
o
n
c
l
u
s
i
o
n
,
t
h
e
e
x
p
e
r
i
m
e
n
t
a
l
r
es
u
l
ts
p
r
o
v
e
d
t
h
a
t
t
h
e
p
r
o
p
o
s
e
d
S
V
M
-
b
as
e
d
m
o
d
e
l
as
a
b
i
n
a
r
y
c
l
a
s
s
i
f
ie
r
p
e
r
f
o
r
m
e
d
b
e
s
t
i
n
te
r
m
s
o
f
c
l
as
s
i
f
i
c
a
ti
o
n
a
c
c
u
r
a
c
y
.
2
.
1
0
.
M
a
chine
lea
rning
a
lg
o
rit
hm
s
a
nd
t
ec
hn
iqu
e
s
C
am
p
o
s
et
a
l.
[
1
4
]
p
r
esen
ted
an
an
aly
s
is
o
f
s
ev
er
al
m
ac
h
in
e
lear
n
in
g
tec
h
n
iq
u
es
a
n
d
alg
o
r
ith
m
s
f
o
r
s
u
p
p
o
r
tin
g
o
n
lin
e
f
ailu
r
e
p
r
e
d
ictio
n
(
OFP).
I
n
th
e
e
x
p
er
i
m
en
ts
,
m
u
ltip
le
alg
o
r
ith
m
s
an
d
d
if
f
er
e
n
t
d
ata
p
r
o
ce
s
s
in
g
m
eth
o
d
s
ar
e
co
n
s
id
er
ed
.
A
c
o
m
p
ar
is
o
n
is
m
ad
e
with
SVM,
an
alg
o
r
ith
m
u
s
ed
in
OFP.
I
n
co
n
clu
s
io
n
,
th
e
r
esu
lts
s
h
o
w
th
at
SVMs
ca
n
p
r
ed
ict
a
s
in
g
le
f
ailu
r
e
m
o
d
e,
b
u
t
n
o
t
wh
en
co
n
s
id
er
in
g
m
u
lti
-
class
f
ailu
r
es.
Fo
r
s
in
g
le
an
d
m
u
ltip
le
f
ailu
r
e
m
o
d
es
,
d
ec
is
io
n
t
r
ee
(
DT
)
,
n
eu
r
al
n
etwo
r
k
(
NN)
,
a
n
d
B
ag
g
in
g
ca
n
p
r
ed
ict
th
em
well.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
1
6
,
No
.
1
,
Feb
r
u
ar
y
20
2
6
:
5
4
5
-
554
548
2
.
1
1
.
H
i
er
a
r
c
h
i
ca
l
o
n
l
i
n
e
f
a
i
l
u
r
e
p
r
e
d
i
c
t
io
n
a
p
p
ro
a
c
h
(
H
O
R
A
)
P
i
t
a
k
r
a
t
e
t
a
l
.
[
4
]
p
r
o
p
o
s
ed
an
a
p
p
r
o
a
ch
c
a
l
l
ed
H
O
R
A
,
u
s
e
d
f
o
r
O
F
P
,
wh
i
c
h
c
o
n
s
i
s
t
s
o
f
t
w
o
k
i
n
d
s
o
f
a
r
ch
i
t
e
c
tu
r
a
l
m
o
d
e
l
s
a
n
d
h
i
e
r
ar
c
h
ic
a
l
o
n
l
in
e
f
a
i
lu
r
e
p
r
e
d
ic
t
o
r
s
.
T
h
e
o
b
j
e
c
t
i
v
e
i
s
t
o
p
r
ed
i
c
t
th
e
p
r
o
p
ag
a
t
i
o
n
o
f
th
e
f
a
i
lu
r
e
s
i
f
t
h
e
f
a
i
lu
r
e
o
f
e
a
c
h
co
m
p
o
n
e
n
t
c
an
b
e
p
r
e
d
i
c
te
d
an
d
t
h
e
d
ep
e
n
d
e
n
c
i
e
s
am
o
n
g
t
h
e
m
a
r
e
a
ck
n
o
w
l
e
d
g
e
d
.
F
i
r
s
t
o
f
a
l
l
,
b
a
s
ed
o
n
B
a
y
e
s
i
an
n
et
w
o
r
k
s
,
t
h
e
f
a
i
l
u
r
e
p
r
o
p
ag
a
t
i
o
n
m
o
d
e
l
(
F
P
M
)
i
s
a
r
c
h
i
t
ec
t
e
d
,
wh
i
c
h
co
m
p
r
i
s
e
s
t
h
e
p
r
ed
i
c
t
i
o
n
o
u
tp
u
t
s
an
d
d
e
p
en
d
e
n
c
i
e
s
a
m
o
n
g
th
e
c
o
m
p
o
n
en
t
s
o
b
t
a
in
e
d
f
r
o
m
t
h
e
ar
c
h
i
t
ec
t
u
r
a
l
m
o
d
e
ls
.
S
e
c
o
n
d
,
t
h
e
a
r
ch
i
t
e
c
tu
r
a
l
d
e
p
en
d
e
n
cy
m
o
d
el
(
A
D
M)
is
a
m
i
d
d
l
e
m
o
d
e
l
r
e
p
r
e
s
e
n
t
in
g
t
h
e
d
e
p
en
d
e
n
c
ies
a
m
o
n
g
co
m
p
o
n
e
n
t
s
o
f
th
e
ar
c
h
i
t
e
c
tu
r
e
.
T
h
ir
d
,
an
a
u
to
m
a
ti
c
ex
t
r
a
c
t
io
n
o
f
t
h
e
A
D
M
r
e
q
u
i
r
e
s
m
o
n
i
to
r
in
g
d
a
t
a
t
h
a
t
c
o
m
p
r
i
s
e
s
s
y
s
t
e
m
-
l
e
v
e
l
r
e
s
o
u
r
c
e
u
s
ag
e
.
T
h
e
r
e
s
u
l
t
i
s
c
a
l
l
e
d
th
e
S
L
A
s
t
i
c
m
o
d
e
l
,
w
h
ic
h
in
c
l
u
d
e
s
t
h
e
r
e
l
a
t
i
o
n
s
h
ip
s
a
n
d
co
m
p
o
n
e
n
t
s
i
n
th
e
s
o
f
t
w
a
r
e,
i
n
f
o
r
m
a
t
io
n
a
b
o
u
t
d
e
p
lo
y
m
e
n
t,
an
d
t
h
e
n
u
m
b
er
o
f
ca
l
l
s
o
f
e
a
c
h
co
m
p
o
n
e
n
t.
A
D
M
i
s
c
r
ea
t
e
d
b
y
c
o
m
b
in
in
g
t
h
e
i
n
f
o
r
m
a
t
io
n
f
r
o
m
th
e
S
L
A
s
t
i
c
m
o
d
e
l
to
o
b
t
a
i
n
k
n
o
w
l
ed
g
e
ab
o
u
t
t
h
e
d
e
p
e
n
d
en
c
i
e
s
o
f
t
h
e
co
m
p
o
n
en
t
s
.
I
n
c
o
n
c
lu
s
i
o
n
,
t
h
e
p
r
o
p
o
s
e
d
a
p
p
r
o
a
ch
c
an
p
r
e
d
i
c
t
co
m
p
o
n
e
n
t
f
a
i
lu
r
e
s
i
n
t
h
e
s
y
s
t
e
m
an
d
t
h
e
p
r
o
b
ab
i
l
i
t
i
e
s
o
f
t
h
e
p
r
o
p
ag
a
t
i
o
n
o
f
th
o
s
e
f
a
i
lu
r
e
s
t
o
o
th
e
r
c
o
m
p
o
n
en
t
s
.
2
.
1
2
.
I
m
pro
v
ing
s
o
f
t
w
a
re
f
a
ult
predict
io
n wit
h t
hres
ho
ld
v
a
lue
Acc
o
r
d
in
g
t
o
Sh
atn
awi
[
1
5
]
,
th
r
esh
o
ld
r
ec
o
g
n
itio
n
an
d
d
ef
ec
t
p
r
ed
ictio
n
a
r
e
two
m
eth
o
d
s
th
at
ar
e
u
s
ed
f
o
r
a
n
aly
zin
g
s
tatis
tical
s
o
f
twar
e.
T
h
o
s
e
m
et
h
o
d
s
a
r
e
c
o
m
b
in
ed
in
th
eir
r
esear
ch
to
in
clu
d
e
a
n
ew
f
o
r
m
o
f
s
tatic
an
aly
s
is
.
Sh
atn
awi
[
1
5
]
p
r
o
p
o
s
ed
a
n
ew
d
ep
e
n
d
en
t
v
ar
iab
le
b
y
u
s
in
g
th
r
esh
o
ld
v
a
lu
es.
T
h
e
v
alu
es
o
f
th
e
th
r
esh
o
l
d
ar
e
u
s
ed
to
i
d
en
tify
t
h
e
s
o
f
twar
e
s
y
s
tem
s
th
at
n
ee
d
f
o
c
u
s
in
d
ev
elo
p
m
en
t,
test
in
g
an
d
m
ain
ten
an
ce
.
I
f
th
ey
h
a
v
e
n
o
f
au
lts
,
th
ese
m
o
d
u
les
a
r
e
lab
el
led
as
a
m
ed
iu
m
g
r
o
u
p
.
T
h
e
f
au
lty
m
o
d
u
les
h
av
e
b
ee
n
d
ef
in
e
d
as
s
tr
o
n
g
,
wh
er
ea
s
n
o
n
-
f
au
lty
s
y
s
tem
s
ar
e
d
ef
in
ed
as
n
o
th
in
g
at
all.
Af
ter
war
d
s
,
th
e
latest
id
en
tity
h
as
b
ee
n
u
s
ed
i
n
p
en
t
u
p
le
alg
o
r
ith
m
s
.
Fiv
e
class
if
ie
r
s
also
ev
alu
ate
th
e
o
r
ig
in
al
v
ar
iab
le,
as
well
as
th
e
o
u
tco
m
es
o
f
th
e
latter
ar
e
ev
alu
ated
th
r
o
u
g
h
test
in
g
o
f
s
tatis
tics
.
W
ilco
x
o
n
s
ig
n
ed
a
r
an
k
test
ab
le
to
f
in
d
th
r
ee
ty
p
es
o
f
class
if
ier
s
as
o
f
v
ar
iab
les
o
f
two
.
W
h
en
c
o
m
p
ar
ed
to
th
e
th
r
e
e
class
if
ier
s
,
th
ey
n
o
te
d
th
at
alth
o
u
g
h
J
R
ip
o
u
tp
u
t
d
im
in
is
h
es,
n
aiv
e
B
ay
es
(
NB
)
,
an
d
l
o
g
is
tic
r
eg
r
ess
io
n
(
L
R
)
ar
e
n
o
tab
ly
g
r
ea
ter
in
th
e
r
ec
o
m
m
en
d
ed
p
ar
am
eter
.
C
o
n
clu
s
iv
ely
,
th
eir
f
in
d
in
g
s
in
d
icate
th
at
in
ce
r
tain
class
if
ier
s
,
th
e
s
u
g
g
ested
p
ar
am
eter
eith
er
en
h
a
n
ce
s
th
e
p
er
f
o
r
m
an
ce
o
r
ju
s
t z
er
o
im
p
a
ct
o
n
th
e
p
er
f
o
r
m
an
ce
o
f
o
t
h
er
s
.
2
.
1
3
.
Dee
p
lea
rning
s
y
s
t
em
s
o
f
t
wa
re
f
a
ult
predict
io
n
Acc
o
r
d
in
g
to
Qiao
et
a
l.
[
1
6
]
,
th
ey
d
ev
elo
p
ed
th
e
c
o
n
ce
p
t
o
f
a
d
ee
p
lear
n
i
n
g
co
n
ce
p
t
f
o
r
p
r
ed
ictin
g
v
u
ln
er
ab
ilit
ies
in
a
p
p
licatio
n
s
y
s
tem
s
.
T
h
eir
p
r
e
f
er
r
ed
m
eth
o
d
d
ev
elo
p
s
a
co
m
p
r
eh
e
n
s
iv
e
lear
n
in
g
ap
p
r
o
ac
h
to
esti
m
ate
th
e
s
er
io
u
s
f
law.
T
h
e
im
p
r
o
v
em
e
n
t
o
f
th
e
s
u
g
g
ested
tech
n
iq
u
e
o
n
th
e
s
u
p
p
o
r
t
v
ec
to
r
r
eg
r
ess
io
n
(
SVR
)
,
f
u
zz
y
s
u
p
p
o
r
t
v
ec
to
r
r
eg
r
ess
io
n
(
FS
VR
)
,
an
d
d
ec
is
io
n
tr
ee
r
e
g
r
ess
io
n
(
DT
R
)
is
ess
en
tial
f
o
r
th
e
co
llected
d
ata.
Paled
in
c
o
m
p
ar
is
o
n
to
s
u
ch
s
tate
-
of
-
th
e
-
ar
t
m
eth
o
d
s
o
n
two
well
-
k
n
o
w
n
m
o
d
els,
th
e
m
eth
o
d
ac
h
iev
ed
a
s
u
b
s
tan
tial r
ed
u
cti
o
n
in
th
e
s
tan
d
ar
d
e
r
r
o
r
an
d
i
n
cr
ea
s
ed
th
e
co
m
p
o
u
n
d
ed
c
o
r
r
e
latio
n
co
ef
f
icien
t.
2
.
1
4
.
C
o
mp
r
e
h
e
ns
i
v
e
mo
d
el
f
o
r
s
o
f
t
w
a
r
e
f
a
u
l
t
p
r
ed
i
c
t
i
o
n
Acc
o
r
d
in
g
to
Sin
g
h
[
1
7
]
,
h
e
h
as
u
s
ed
th
e
f
iv
e
d
ev
el
o
p
m
e
n
ts
f
r
o
m
o
r
g
an
ic
f
ield
s
f
o
r
t
esti
n
g
an
d
tr
ain
in
g
.
I
t
is
s
ee
n
th
at
in
th
e
ca
s
e
o
f
a
p
r
o
ject,
C
4
.
5
ex
ec
u
t
ed
well
ar
o
u
n
d
a
m
ea
n
v
alu
e
o
f
r
ec
eiv
e
r
-
o
p
er
ato
r
ch
ar
ac
t
er
is
tic
(
R
OC
)
th
an
th
e
p
r
o
p
o
s
ed
eig
h
t
r
u
le
-
b
ased
class
if
icatio
n
with
th
e
m
ajo
r
ity
o
f
R
OC
.
T
h
e
ab
ilit
y
to
d
ea
l w
ith
th
e
p
r
o
g
r
am
d
ef
ec
t p
r
ed
ictio
n
p
r
o
b
lem
-
b
ased
lea
r
n
er
r
u
le
also
co
n
tr
ib
u
ted
to
im
p
r
o
v
e
d
ef
f
icien
cy
d
u
e
to
i
n
ad
eq
u
ate
tr
ain
in
g
d
at
a
f
o
r
a
g
iv
e
n
class
.
Dec
is
io
n
tab
le
-
Naiv
e
B
ay
es
h
y
b
r
id
class
if
ier
(
DT
NB
)
h
as
s
u
r
p
ass
ed
s
ev
er
al
r
u
le
-
b
ased
lear
n
er
s
in
c
r
o
s
s
-
p
r
o
jects
a
s
well
an
d
th
e
f
i
n
d
in
g
s
als
o
b
ec
o
m
e
clo
s
e
to
R
OC
6
9
p
er
ce
n
t
in
s
id
e
th
e
p
r
o
g
r
am
m
ed
.
T
h
er
e
f
o
r
e,
d
a
ta
f
r
o
m
v
a
r
io
u
s
ac
tiv
ities
r
e
latin
g
to
th
e
v
er
y
s
am
e
ar
ea
ca
n
b
e
u
s
ed
to
esti
m
ate
in
ter
-
p
r
o
ject
f
ailu
r
e
,
ju
s
t
in
ca
s
e
o
f
task
d
ata
s
et
is
ab
s
en
t
an
d
ca
n
wo
r
k
s
im
ilar
ly
ac
r
o
s
s
task
s
.
2
.
1
5
.
T
he
s
o
f
t
wa
re
def
ec
t
pr
edict
io
n c
o
ncept
re
lies
o
n t
h
e
Alt
a
Rica
la
ng
ua
g
e
Acc
o
r
d
in
g
to
So
n
g
et
a
l.
[
1
8
]
,
a
m
eth
o
d
u
s
in
g
th
e
AltaR
ica
alg
o
r
ith
m
is
p
r
o
p
o
s
ed
in
th
ei
r
r
esear
c
h
f
o
r
s
o
f
twar
e
f
a
u
lt
p
r
e
d
ictio
n
.
T
h
e
p
r
o
p
o
s
ed
m
eth
o
d
co
m
p
r
is
es
th
r
ee
co
m
p
o
n
e
n
ts
:
o
n
e
is
AltaR
ica
-
b
ased
s
o
f
twar
e
r
eq
u
ir
em
en
t
m
o
d
elin
g
,
th
e
two
is
lin
e
tem
p
o
r
al
lo
g
ic
(LTL)
-
b
ased
s
af
ety
lim
itatio
n
s
,
an
d
t
h
e
th
r
ee
is
a
m
o
d
el
-
b
ased
f
a
u
lt
p
r
e
d
ictio
n
alg
o
r
ith
m
.
E
v
e
n
tu
ally
,
th
ey
a
p
p
lied
th
is
m
o
d
el
to
th
e
tr
a
d
itio
n
al
s
tu
d
y
p
r
o
g
r
am
f
o
r
th
e
av
iatio
n
aid
s
o
f
twar
e
s
y
s
tem
.
T
h
e
test
r
esu
lts
s
h
o
w
t
h
at
th
is
latest
d
esig
n
will
b
o
o
s
t
th
e
ef
f
icac
y
an
d
v
alid
ity
o
f
th
e
d
ef
ec
t
p
r
ed
ictio
n
th
at
ca
n
r
eliab
ly
ch
ar
ac
ter
iz
e
th
e
o
p
er
atin
g
ch
a
r
ac
ter
is
tics
o
f
th
e
av
iatio
n
aid
s
o
f
twar
e
s
y
s
tem
,
an
d
ef
f
ec
tiv
ely
r
ec
o
g
n
ize
d
y
n
am
ic
d
ef
e
cts
s
u
ch
as
th
e
s
tate
tr
an
s
itio
n
d
is
p
u
te,
a
n
d
t
h
e
ir
r
eg
u
lar
f
ea
tu
r
e
s
er
ies.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
A
s
ystema
tic
r
ev
ie
w
o
f so
ftw
a
r
e
fa
u
lt p
r
ed
ictio
n
…
(
R
.
K
a
n
e
s
a
r
a
j R
a
ma
s
a
my
)
549
2
.
1
6
.
Resea
rc
h o
n
s
o
f
t
wa
re
m
et
ric
s
elec
t
io
n f
o
r
s
o
f
t
wa
re
def
ec
t
predict
io
n
Acc
o
r
d
in
g
to
W
an
g
a
n
d
Kh
o
s
h
g
o
f
taar
[
1
9
]
,
th
e
y
h
av
e
m
a
d
e
a
co
m
p
ar
is
o
n
with
th
r
ee
class
es
o
f
s
elec
tio
n
wh
ich
is
f
ilter
-
b
ased
s
u
b
s
et
ev
al
u
ato
r
s
,
wr
a
p
p
er
-
b
ased
s
u
b
s
et
s
elec
to
r
s
a
n
d
f
ilter
-
b
ased
f
ea
tu
r
e
r
an
k
in
g
.
T
h
e
y
th
en
b
u
ilt
th
e
m
o
d
el
ac
co
r
d
in
g
ly
to
r
eso
lv
e
s
o
f
twar
e
f
ailu
r
e
p
r
ed
ictio
n
s
.
T
h
e
n
,
th
e
r
eliab
ilit
y
o
f
its
id
en
tific
atio
n
is
ev
alu
ated
a
cc
o
r
d
in
g
to
t
h
e
ar
ea
u
n
d
er
t
h
e
r
ec
eiv
er
o
p
er
atin
g
ch
ar
ac
te
r
is
tic
AUC
ef
f
icien
cy
m
etr
ic.
T
h
ey
h
av
e
u
s
ed
d
ata
o
b
tain
ed
f
r
o
m
th
e
f
o
u
r
r
elea
s
es
o
f
th
e
m
ass
iv
e
n
etwo
r
k
s
y
s
tem
,
s
p
ec
if
ically
telec
o
m
m
u
n
icatio
n
s
.
T
h
e
p
r
ed
ictiv
e
m
o
d
els ar
e
d
e
v
elo
p
e
d
u
s
in
g
f
iv
e
d
if
f
er
e
n
t c
lass
if
ier
s
,
w
h
ich
ar
e
K
-
n
e
ar
est
n
eig
h
b
o
r
s
,
B
ay
es,
m
u
ltil
ay
er
p
er
ce
p
tr
o
n
,
lo
g
is
tic
r
eg
r
ess
io
n
an
d
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
es.
T
h
e
test
r
esu
lts
r
ev
ea
l
th
at,
ac
co
r
d
i
n
g
to
th
e
AUC
ef
f
icien
cy
m
ea
s
u
r
e,
th
e
wr
ap
p
er
-
b
ased
a
p
p
r
o
ac
h
to
s
elec
tio
n
o
f
a
s
u
b
s
et
p
er
f
o
r
m
ed
b
etter
th
an
t
h
e
r
est.
T
h
e
ap
p
s
'
r
an
k
in
g
d
id
wo
r
s
e.
Fu
r
th
er
m
o
r
e
,
wh
en
co
m
p
ar
ed
to
th
e
f
iv
e
lear
n
er
s
in
o
u
r
s
am
p
le
,
lo
g
is
tic
r
eg
r
ess
io
n
(
L
R
)
r
e
p
r
esen
ts
th
e
h
ig
h
est o
u
tp
u
t,
th
en
,
m
u
ltil
ay
er
p
e
r
ce
p
tr
o
n
(
ML
P)
.
2
.
1
7
.
L
o
a
d
-
ca
pa
cit
y
m
o
del
T
h
e
tr
ad
itio
n
al
lo
ad
ca
p
ac
ity
m
o
d
el
h
as
two
s
tates,
wh
ich
ar
e
th
e
n
o
r
m
al
s
tate
an
d
th
e
f
au
lty
s
tate.
Ji
et
a
l.
[
2
0
]
p
r
o
p
o
s
ed
a
th
ir
d
s
tate
th
at
is
ca
lled
s
tate
-
co
n
g
esti
o
n
to
m
ee
t th
e
c
h
ar
ac
ter
is
tics
o
f
co
m
m
u
n
icatio
n
n
etwo
r
k
s
.
T
h
ey
im
p
r
o
v
ed
t
h
e
tr
ad
itio
n
al
m
o
d
el
an
d
m
a
d
e
it
ab
le
to
b
e
m
o
r
e
r
elia
b
le
in
te
r
m
s
o
f
ac
c
u
r
ac
y
an
d
p
r
ec
is
io
n
o
f
f
a
u
lt
p
r
ed
ictio
n
an
d
p
r
ed
ictio
n
ef
f
ec
t.
Af
ter
c
o
m
p
ar
is
o
n
with
th
e
m
o
d
els,
t
h
ey
f
o
u
n
d
th
at
th
e
av
er
ag
e
ac
cu
r
ac
y
o
f
t
h
e
p
r
o
p
o
s
ed
f
au
lt p
r
ed
ictio
n
m
o
d
el
is
a
r
o
u
n
d
6
6
.
6
1
%
an
d
im
p
r
o
v
ed
b
y
9
.
3
5
%
c
o
m
p
ar
e
d
with
n
o
n
-
lin
ea
r
lo
a
d
ca
p
ac
ity
m
o
d
el
in
W
an
g
et
a
l.
[
2
1
]
.
T
h
e
p
r
o
p
o
s
ed
m
o
d
el
h
as
a
h
ig
h
ac
cu
r
ac
y
r
ate
b
u
t
it
i
s
s
en
s
itiv
e
to
th
e
p
r
o
p
ag
atio
n
p
r
ed
ictio
n
in
th
e
n
etwo
r
k
’
s
k
e
y
n
o
d
e
f
au
lts
.
2
.
1
8
.
Art
if
icia
l
neura
l net
wo
rk
a
nd
qu
euing
t
heo
ry
Usi
n
g
ar
tific
ial
n
eu
r
al
n
etwo
r
k
s
(
ANN)
an
d
q
u
eu
in
g
th
e
o
r
y
,
T
r
ip
ath
i
an
d
Sar
aswat
[
2
2
]
w
er
e
ab
le
to
in
cr
ea
s
e
th
e
r
eliab
ilit
y
o
f
s
o
f
t
war
e
p
r
o
d
u
cts
to
esti
m
ate
f
ailu
r
e
r
ate,
d
etec
tio
n
,
co
r
r
ec
tio
n
,
an
d
allo
ca
tio
n
.
T
h
e
q
u
eu
in
g
th
eo
r
y
m
o
d
el
is
m
a
in
ly
ab
o
u
t
s
im
u
latin
g
th
e
cu
s
to
m
er
ar
r
i
v
al
p
atter
n
in
a
q
u
eu
e
th
at
h
as
b
ee
n
co
r
r
elate
d
with
th
e
er
r
o
r
d
etec
tio
n
p
atter
n
in
th
e
p
r
o
ce
s
s
o
f
s
o
f
twar
e
d
ev
elo
p
m
en
t.
T
h
e
p
a
p
er
r
ev
ea
ls
th
at
less
n
u
m
b
er
o
f
test
er
s
ar
e
r
eq
u
ir
ed
in
th
e
ea
r
ly
p
h
ases
o
f
s
o
f
twar
e
d
ev
elo
p
m
en
t
life
c
y
cle
(
SDL
C
)
,
b
u
t
t
h
e
n
u
m
b
e
r
in
cr
ea
s
es
as
th
e
s
o
f
twar
e
d
ev
elo
p
m
en
t
p
h
ases
in
cr
ea
s
e.
T
ea
m
lead
er
s
o
f
s
o
f
twar
e
co
m
p
an
ies
ca
n
u
s
e
th
is
p
ap
er
’
s
r
esu
lts
to
p
r
ed
ict
th
e
n
u
m
b
er
o
f
test
er
s
th
ey
s
h
o
u
ld
h
ir
e
f
r
o
m
th
e
in
itial
s
tag
e
to
th
e
m
id
d
le
s
tag
e
a
n
d
at
th
e
en
d
s
tag
e
o
f
s
o
f
twar
e
d
ev
elo
p
m
en
t.
I
n
th
e
m
i
d
d
le
s
ta
g
e,
th
e
m
ax
im
u
m
n
u
m
b
er
o
f
s
o
f
twar
e
test
er
s
i
s
n
ee
d
ed
,
b
u
t it
d
ec
r
ea
s
es in
th
e
in
itial p
h
ase
an
d
th
e
last
p
h
ase
o
f
th
e
s
o
f
twar
e
d
ev
elo
p
m
en
t l
if
e
cy
cle
(
SDLC)
.
2
.
1
9
.
I
nh
er
it
a
nce
m
et
rics
a
nd
a
rt
if
icia
l neura
l net
wo
rk
Aziz
et
a
l.
[
2
3
]
s
elec
ted
C
h
id
am
b
er
an
d
Kem
er
er
m
etr
ics
(
C
K)
to
ev
alu
ate
in
h
er
itan
ce
ef
f
ec
ts
o
n
So
f
twar
e
f
au
lt
p
r
ed
ictio
n
(
SF
P)
.
T
h
ey
s
p
lit
th
e
d
atasets
in
to
two
s
ets,
th
e
f
ir
s
t
o
n
e
is
C
K
with
in
h
er
itan
ce
a
n
d
th
e
o
th
er
d
ataset
is
C
K
with
o
u
t
in
h
er
itan
c
e
f
o
r
c
o
m
p
ar
is
o
n
.
T
o
b
u
ild
th
is
m
o
d
el,
th
e
y
u
s
ed
an
ANN.
T
h
e
r
esu
lts
h
av
e
s
h
o
wn
th
at
in
h
er
itan
ce
s
h
o
ws
an
ac
ce
p
tab
le
co
n
tr
ib
u
tio
n
in
SF
P
an
d
it
i
s
s
af
e,
b
u
t
h
ig
h
in
h
er
itan
ce
is
not
b
ec
au
s
e
it
ca
n
lead
to
s
o
f
twar
e
f
au
lts
.
T
h
ey
s
u
g
g
ested
k
ee
p
in
g
th
e
in
h
er
itan
ce
m
etr
ics
m
in
im
u
m
;
th
e
test
in
g
c
o
m
m
u
n
ity
ca
n
s
af
ely
u
s
e
in
h
er
itan
c
e
m
etr
ics
to
p
r
ed
ict
f
au
lts
,
b
u
t
h
ig
h
in
h
er
itan
ce
is
not
r
ec
o
m
m
en
d
e
d
b
ec
au
s
e
it c
an
ca
u
s
e
f
au
lts
.
2
.
2
0
.
Resid
ua
l
e
rr
o
rs:
J
-
M
a
nd
G
-
M
T
h
e
h
y
b
r
i
d
m
o
d
e
l
J
a
b
e
e
n
e
t
a
l
.
[
2
4
]
p
r
o
p
o
s
e
s
a
c
o
m
b
i
n
a
t
i
o
n
o
f
J
-
M
a
n
d
G
-
M
b
a
s
e
d
o
n
r
e
s
i
d
u
a
l
e
r
r
o
r
s
;
t
h
e
m
o
d
e
l
s
el
e
c
ts
t
h
e
m
o
s
t
s
u
i
ta
b
l
e
p
r
e
d
i
c
t
e
d
v
al
u
e
f
r
o
m
b
o
t
h
m
o
d
e
l
s
.
G
-
M
g
i
v
e
s
h
i
g
h
p
r
e
d
i
c
t
i
o
n
a
c
c
u
r
a
c
y
wi
th
s
m
a
l
l
s
a
m
p
l
es
o
f
d
a
t
a
,
b
u
t
wi
th
l
o
w
p
r
e
d
i
c
t
i
n
g
a
c
c
u
r
a
c
y
wi
t
h
r
a
n
d
o
m
f
l
u
c
t
u
a
t
i
n
g
d
a
t
a
.
J
-
M
i
s
m
o
r
e
s
u
it
a
b
l
e
f
o
r
f
o
r
e
c
a
s
t
i
n
g
v
i
b
r
a
ti
n
g
a
n
d
c
h
a
n
g
i
n
g
s
t
o
c
h
a
s
t
i
c
d
a
ta
s
e
q
u
e
n
c
es
b
u
t
i
t
n
e
e
d
s
a
l
a
r
g
e
a
m
o
u
n
t
o
f
d
a
t
a
.
T
h
e
h
y
b
r
i
d
m
o
d
e
l
c
o
m
b
i
n
e
s
t
h
e
b
es
t
-
p
r
e
d
i
c
t
e
d
v
a
l
u
es
o
f
t
h
e
m
o
d
e
ls
a
n
d
g
i
v
e
s
b
et
t
e
r
f
o
r
e
ca
s
t
i
n
g
r
es
u
l
t
s
f
o
r
f
a
i
l
u
r
e
d
a
ta
s
e
q
u
e
n
c
e
s
.
T
h
e
m
o
d
e
l
h
as
g
o
o
d
p
e
r
f
o
r
m
a
n
c
e
a
c
c
u
r
a
c
y
a
n
d
a
p
p
l
i
c
a
b
i
li
t
y
i
n
S
o
f
tw
a
r
e
f
a
u
lt
p
r
ed
i
c
t
i
o
n
.
2
.
2
1
.
Wra
pp
er
-
ba
s
ed
s
elec
t
i
o
n m
et
ho
d by
pa
rt
icle
s
wa
r
m
o
ptim
iz
a
t
io
n
-
m
ulti
-
G
a
us
s
ia
n a
pp
ro
a
ch
Usi
n
g
d
atasets
th
at
co
n
s
is
t
o
f
n
o
is
y
an
d
ir
r
elev
an
t
r
ec
o
r
d
s
m
ay
r
esu
lt
in
u
n
n
ec
ess
ar
y
waste
o
f
r
eso
u
r
ce
s
an
d
co
u
l
d
lead
to
f
a
ilu
r
es
if
u
s
in
g
class
if
icatio
n
with
o
u
t
s
elec
tio
n
m
eth
o
d
s
.
T
h
e
r
e
ar
e
m
ain
ly
th
r
ee
ty
p
es
f
o
r
class
if
icatio
n
:
f
ilte
r
-
b
ased
f
ea
tu
r
e
s
elec
tio
n
,
e
m
b
ed
d
e
d
f
ea
tu
r
e
s
elec
tio
n
a
n
d
wr
ap
p
er
f
ea
t
u
r
e
s
elec
tio
n
.
B
an
g
a
an
d
B
an
s
al
[
2
5
]
ch
o
s
e
th
e
wr
ap
p
er
-
b
ased
s
elec
tio
n
m
eth
o
d
b
ec
au
s
e
it
h
as
th
e
b
est
ac
cu
r
ac
y
am
o
n
g
th
ese
ap
p
r
o
ac
h
es.
T
h
e
m
eth
o
d
is
a
h
y
b
r
i
d
o
f
alg
o
r
ith
m
s
u
s
ed
to
im
p
r
o
v
e
th
e
ac
cu
r
a
cy
th
e
r
eliab
ilit
y
o
f
th
e
s
o
f
twar
e
esti
m
atio
n
u
s
in
g
f
ea
t
u
r
e
s
elec
tio
n
b
y
p
ar
t
icle
s
war
m
o
p
tim
izatio
n
(
PS
O)
-
m
u
lti
-
Gau
s
s
ian
ap
p
r
o
ac
h
(
MG
A)
to
r
ed
u
ce
n
o
is
e
an
d
s
elec
t
r
elev
an
t
attr
ib
u
tes.
Me
an
weig
h
ted
least
s
q
u
ar
es
twi
n
SV
M
(
MW
-
L
STSVM
)
r
esu
lt
s
s
h
o
w
th
at
th
e
clas
s
if
icatio
n
alg
o
r
ith
m
h
as
r
ea
ch
ed
th
e
h
ig
h
est
ac
cu
r
ac
y
am
o
n
g
all
o
th
er
class
if
ier
s
wi
th
9
1
.
3
%,
ef
f
icien
cy
in
cr
ea
s
ed
b
y
9
.
8
%
th
an
th
e
o
th
er
ex
is
tin
g
m
eth
o
d
s
in
class
if
y
in
g
d
ef
ec
tiv
e
an
d
n
o
n
-
d
ef
ec
ti
v
e
a
p
p
r
o
ac
h
es.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
1
6
,
No
.
1
,
Feb
r
u
ar
y
20
2
6
:
5
4
5
-
554
550
2
.
2
2
.
M
a
chine
lea
rning
im
pl
em
ent
a
t
io
n
Fro
m
th
e
r
esear
ch
,
W
ó
jcick
i
an
d
Dąb
r
o
wsk
i
[
2
6
]
f
o
u
n
d
th
at
th
ey
ca
n
au
to
m
atica
lly
p
r
ed
ict
th
e
p
o
s
s
ib
ly
f
au
lty
f
r
ag
m
e
n
t
in
th
e
s
o
u
r
ce
co
d
e
th
ey
c
o
d
ed
a
n
d
wh
ich
allo
ws
d
ev
elo
p
er
s
to
f
o
cu
s
m
o
r
e
o
n
d
ev
elo
p
m
e
n
t,
b
u
t
n
o
t c
o
n
s
u
m
e
lo
ts
o
f
tim
e
in
f
ailu
r
e
f
in
d
i
n
g
.
T
h
e
tech
n
i
q
u
es th
at
th
ey
u
s
e
im
p
r
o
v
e
th
e
f
ailu
r
e
p
r
ed
ictio
n
b
y
s
tr
u
ctu
r
al
p
r
o
g
r
am
m
in
g
to
r
estrict
th
e
co
n
t
r
o
l
f
lo
w
to
a
h
ier
ar
ch
ical
s
tr
u
ctu
r
e.
T
h
e
n
ex
t
im
p
r
o
v
em
e
n
t
is
th
at
th
e
o
b
ject
-
o
r
ien
ted
p
ar
a
d
ig
m
will
b
e
th
e
m
ain
p
r
o
g
r
am
m
in
g
c
o
n
ce
p
t.
As
a
r
esu
lt,
th
e
r
esear
ch
p
r
o
v
es
th
at
f
au
lt
-
p
r
ed
ictio
n
m
eth
o
d
s
ca
n
b
e
ap
p
lied
s
u
cc
ess
f
u
lly
to
Py
th
o
n
,
J
av
a,
an
d
C
/
C
++
p
r
o
g
r
a
m
s
,
an
d
g
et
th
e
r
esu
lt
o
f
0
.
6
4
r
ec
all
r
ate
an
d
0
.
2
3
f
alse
p
o
s
itiv
e
r
ate.
T
h
is
r
e
s
ea
r
ch
also
s
u
p
p
o
r
ts
ap
p
ly
in
g
th
e
s
p
ec
ial
lo
g
ar
ith
m
ic
f
ilter
,
s
in
ce
th
e
f
au
lt p
r
ed
ict
o
r
im
p
lem
en
ted
with
o
u
t u
s
in
g
th
is
f
ilter
ac
h
iev
ed
a
0
.
3
2
8
o
f
r
ec
all
r
ate
with
a
0
.
1
0
8
o
f
f
alse
p
o
s
itiv
e
r
ate.
I
n
co
n
clu
s
io
n
,
m
ac
h
in
e
lear
n
i
n
g
ca
n
b
e
u
s
ed
f
o
r
cr
ea
tin
g
a
f
a
u
lt p
r
ed
icto
r
f
o
r
Py
th
o
n
,
J
av
a
an
d
C
/C
++
p
r
o
ject
s
.
2
.
2
3
.
B
R
t
ec
hn
iqu
e
I
n
th
is
r
esear
ch
,
Ma
h
ajan
et
a
l.
[
2
7
]
f
o
u
n
d
B
ay
esian
r
e
g
u
lar
izatio
n
(
B
R
)
tech
n
iq
u
e
h
elp
s
with
s
o
f
twar
e
f
au
lts
p
r
ed
icted
.
T
h
e
m
ain
f
u
n
ctio
n
o
f
th
e
B
R
te
ch
n
iq
u
e
is
to
elim
in
ate
th
e
m
ath
er
r
o
r
s
an
d
also
d
eter
m
in
e
th
e
s
u
itab
le
co
m
b
in
atio
n
to
p
r
o
d
u
ce
a
n
ef
f
icien
t
n
etwo
r
k
.
T
h
e
ac
c
u
r
ac
y
o
f
t
h
e
B
R
alg
o
r
ith
m
is
co
m
p
ar
ed
with
th
e
L
ev
e
n
b
er
g
-
Ma
r
q
u
a
r
d
t
(
LM
)
alg
o
r
ith
m
an
d
th
e
b
ac
k
p
r
o
p
ag
atio
n
al
g
o
r
ith
m
(
B
PA)
f
o
r
s
ea
r
ch
in
g
s
o
f
twar
e
f
a
u
lts
.
T
h
ey
also
f
in
d
ac
cu
r
ac
y
with
NN
class
if
ier
s
.
Fo
r
th
e
r
esu
lt,
th
e
B
R
alg
o
r
ith
m
p
r
o
v
id
es
9
2
.
4
4
%
ac
c
u
r
ac
y
,
wh
ich
is
th
e
h
ig
h
est
ac
cu
r
ac
y
co
m
p
ar
ed
to
an
o
t
h
er
a
lg
o
r
ith
m
th
at
was
test
ed
in
th
is
ex
p
er
im
en
t.
I
n
c
o
n
clu
s
io
n
,
th
e
B
R
alg
o
r
ith
m
is
th
e
m
o
s
t
r
eliab
le
alg
o
r
ith
m
with
th
e
co
m
p
ar
is
o
n
to
o
th
er
alg
o
r
ith
m
s
.
2
.
2
4
.
I
t
er
a
t
ed
f
ea
t
ure
s
elec
t
io
n a
lg
o
rit
hm
a
nd
la
y
er
-
re
cu
rr
ent
neura
l net
wo
rk
Fro
m
th
e
r
esear
ch
,
T
u
r
ab
ieh
e
t
a
l.
[
2
8
]
u
s
ed
b
in
a
r
y
g
en
etic
alg
o
r
ith
m
(
B
GA
)
,
b
in
a
r
y
p
ar
ti
cle
s
war
m
o
p
tim
izatio
n
(
B
PS
O
)
,
an
d
b
in
ar
y
an
t
c
o
lo
n
y
o
p
tim
izatio
n
(
B
AC
O
)
as
wr
ap
p
er
FS
alg
o
r
i
th
m
s
.
T
h
e
r
esu
lt
is
co
m
p
ar
ed
with
n
aïv
e
B
ay
es
(
NB
)
,
ANN
,
lo
g
is
tic
r
eg
r
ess
io
n
(
LR
),
k
-
n
ea
r
est
n
eig
h
b
o
r
s
(
k
-
NN
)
a
n
d
C
4
.
5
d
ec
is
io
n
tr
ee
s
.
T
h
e
p
r
o
p
o
s
ed
alg
o
r
ith
m
g
en
e
r
ates
th
e
test
in
g
d
ataset.
T
h
ey
test
th
e
p
er
f
o
r
m
a
n
ce
o
f
th
e
alg
o
r
ith
m
s
in
am
p
lify
in
g
th
e
s
o
f
twar
e
f
au
lt
an
d
f
ailu
r
e
p
r
ed
ictio
n
s
y
s
tem
.
T
h
ey
a
p
p
lied
th
e
p
r
o
p
o
s
ed
alg
o
r
ith
m
s
th
at
th
ey
p
r
o
p
o
s
ed
with
th
e
lay
er
-
r
ec
u
r
r
en
t n
eu
r
al
n
etwo
r
k
(
L
-
R
NN
)
class
if
ier
.
Fo
r
th
e
r
esu
lt,
th
ey
f
o
u
n
d
th
at
th
e
p
er
f
o
r
m
an
ce
o
f
L
-
R
NN
d
ep
en
d
s
o
n
th
e
in
p
u
t d
ata
ch
ar
ac
ter
is
tics
.
Fin
d
in
g
th
e
im
p
o
r
tan
t
m
etr
ic
will
en
h
an
ce
th
e
tr
ain
in
g
p
r
o
ce
s
s
.
I
n
co
n
clu
s
io
n
,
th
e
p
r
o
p
o
s
ed
alg
o
r
ith
m
ca
n
ch
o
o
s
e
th
e
ess
en
tial
s
o
f
twar
e
m
etr
ic
b
y
u
s
in
g
d
is
tin
ct
f
ea
tu
r
e
s
elec
tio
n
(
FS
)
alg
o
r
ith
m
s
.
2
.
2
5
.
Sem
i
-
s
up
er
v
is
ed
deep
f
uzzy
c
-
m
e
a
n c
lus
t
er
ing
m
et
ho
d
Fro
m
th
is
r
ev
iew
p
ap
er
,
Ar
s
h
ad
et
a
l.
[
2
9
]
d
ea
l
with
th
e
s
u
p
er
v
is
ed
an
d
u
n
s
u
p
er
v
is
ed
d
ata
to
u
tili
ze
th
e
f
u
zz
y
in
f
o
r
m
atio
n
f
r
o
m
l
ab
eled
to
u
n
lab
eled
d
ata
to
s
u
s
tain
th
e
b
u
ild
i
n
g
o
f
th
e
c
u
r
r
en
t
m
o
d
el
p
atter
n
.
T
h
ey
u
tili
ze
d
ee
p
f
u
zz
y
c
-
m
e
an
clu
s
ter
in
g
(
DFC
M)
clu
s
ter
in
g
to
r
ep
lace
h
u
m
an
l
o
g
ic.
T
o
p
r
o
v
e
th
e
m
et
h
o
d
,
th
ey
p
r
esen
t
a
m
o
d
el
f
o
r
s
o
f
twar
e
f
au
lt
p
r
ed
ictio
n
.
T
h
e
y
s
h
o
w
th
e
ca
p
ab
ilit
y
o
f
th
ei
r
m
eth
o
d
with
o
th
e
r
m
eth
o
d
s
an
d
all
r
esu
lts
ar
e
p
er
f
o
r
m
e
d
b
y
av
e
r
ag
in
g
1
0
0
r
u
n
s
.
T
h
ey
o
b
s
er
v
e
th
at
th
e
p
er
f
o
r
m
a
n
ce
o
f
class
m
ass
n
o
r
m
aliza
tio
n
(
C
MN
)
is
wo
r
s
e
th
an
non
-
n
eg
ativ
e
s
p
a
r
s
e
g
r
ap
h
b
ased
lab
eled
p
r
o
p
a
g
atio
n
,
NT
C
(
NB
)
,
an
d
DFC
M.
T
h
ey
also
f
o
u
n
d
th
at
th
e
p
e
r
f
o
r
m
an
ce
o
f
FTF
is
p
o
o
r
b
ec
au
s
e
FTF
ap
p
lies
s
u
p
er
v
is
ed
d
ata
.
I
t
co
n
clu
d
es
th
at
s
em
i
-
s
u
p
er
v
is
e
d
d
ata
f
o
r
tr
ai
n
in
g
m
o
d
els
im
p
r
o
v
es
ca
p
ab
ilit
y
.
I
n
s
u
m
m
ar
y
,
th
e
m
eth
o
d
ca
n
b
u
ild
a
f
in
e
p
r
ed
ictio
n
s
y
s
tem
b
y
g
en
e
r
atin
g
g
o
o
d
f
ea
t
u
r
es
an
d
r
e
m
o
v
in
g
e
x
ce
s
s
iv
e
f
ea
tu
r
es
to
d
ec
r
ea
s
e
th
e
u
n
u
s
ed
d
ata
f
o
r
class
if
icatio
n
.
3.
ANALY
SI
S
T
h
is
s
ec
tio
n
will
an
aly
ze
all
t
h
e
av
ailab
le
s
o
f
twar
e
p
r
ed
icti
o
n
m
eth
o
d
s
th
at
h
av
e
b
ee
n
r
e
s
ea
r
ch
ed
.
T
h
is
in
clu
d
es
th
e
an
aly
s
is
o
f
d
ata
p
r
o
ce
s
s
in
g
tech
n
i
q
u
es,
class
if
icatio
n
tech
n
iq
u
es,
m
etr
ics
u
s
ed
,
as
well
as
h
o
w
th
is
will
im
p
r
o
v
e
th
e
ac
cu
r
ac
y
o
f
a
s
o
f
twar
e
p
r
e
d
ictio
n
.
T
a
b
le
1
s
h
o
ws
a
co
m
p
ar
i
s
o
n
tab
le
b
etwe
en
d
if
f
er
en
t
p
r
ed
ictio
n
m
eth
o
d
s
with
th
e
co
r
r
esp
o
n
d
in
g
ad
v
a
n
t
ag
es a
n
d
d
is
ad
v
a
n
tag
es o
f
ea
c
h
m
eth
o
d
.
3
.
1
.
Da
t
a
s
et
s
Data
s
ets
p
lay
a
cr
u
cial
r
o
le
in
th
e
ac
cu
r
ac
y
o
f
t
h
e
class
if
ier
.
I
f
th
e
d
ataset
is
n
o
t
ch
o
s
en
ca
r
ef
u
lly
,
th
e
class
if
ier
ca
n
b
ec
o
m
e
b
iased
to
war
d
th
e
n
o
n
-
f
au
lt
-
p
r
o
n
e
m
o
d
u
le,
w
h
ich
lead
s
to
a
d
ec
r
e
ase
in
th
e
s
o
f
twar
e
p
r
ed
ictio
n
’
s
class
if
icatio
n
ac
c
u
r
ac
y
.
T
h
er
e
ar
e
a
lo
t
o
f
is
s
u
e
s
r
eg
ar
d
in
g
th
e
d
ataset,
s
u
ch
a
s
th
e
im
b
alan
ce
o
f
d
atasets
o
r
th
e
s
elec
tio
n
o
f
f
ea
tu
r
es.
B
elo
w
is
th
e
an
aly
s
is
f
o
r
th
is
p
ar
ticu
lar
p
r
o
b
lem
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
A
s
ystema
tic
r
ev
ie
w
o
f so
ftw
a
r
e
fa
u
lt p
r
ed
ictio
n
…
(
R
.
K
a
n
e
s
a
r
a
j R
a
ma
s
a
my
)
551
T
ab
le
1
.
C
o
m
p
a
r
is
o
n
o
f
s
o
f
twar
e
p
r
ed
ictio
n
m
eth
o
d
s
No
Pred
i
c
t
i
o
n
T
ec
h
n
i
q
u
e
A
d
v
an
t
a
g
e
D
i
s
a
d
v
a
n
t
a
g
e
1.
L
o
a
d
-
ca
p
a
ci
t
y
m
o
d
e
l
˗
E
n
h
an
ci
n
g
t
h
e
l
o
ad
ca
p
a
ci
t
y
m
o
d
e
l
.
˗
Rep
res
en
t
a
t
i
v
e
m
o
d
e
l
.
˗
Imp
ro
v
e
me
n
t
i
n
t
h
e
t
ra
d
i
t
i
o
n
a
l
m
o
d
e
l
.
˗
H
i
g
h
acc
u
ra
cy
ra
t
e.
˗
Fo
c
u
s
e
d
o
n
l
ar
g
e
-
s
ca
l
e
a
p
p
s
.
˗
Co
m
m
u
n
i
ca
t
i
o
n
n
et
w
o
r
k
s
b
a
s
e
d
.
2.
In
h
er
i
t
an
ce
me
t
ri
cs
a
n
d
A
N
N
˗
U
s
es
C
K
me
t
ri
cs
.
˗
U
s
es
A
N
N
.
N
/
A
3.
Res
i
d
u
al
err
o
r
s
:
J
M
an
d
GM
˗
Co
m
b
i
n
at
i
o
n
o
f
G
M
a
n
d
J
M
m
o
d
el
s
.
˗
S
e
l
e
c
t
s
t
h
e
s
u
i
t
a
b
l
e
p
r
e
d
i
c
t
e
d
v
a
l
u
e
f
r
o
m
b
o
t
h
m
o
d
e
l
s
.
N
/
A
4.
A
N
N
a
n
d
q
u
e
u
i
n
g
t
h
eo
ry
˗
E
s
t
i
m
a
t
e
f
a
i
l
u
r
e
r
a
t
e
,
d
e
t
e
c
t
i
o
n
,
c
o
r
r
e
c
t
i
o
n
a
n
d
a
l
l
o
c
a
t
i
o
n
.
˗
Si
m
u
l
a
t
e
a
q
u
eu
e.
N
/
A
5.
Imb
al
a
n
c
ed
d
at
a
p
r
o
ce
s
s
i
n
g
mo
d
e
l
˗
L
o
w
er
l
i
m
i
t
o
f
l
ear
n
i
n
g
acc
u
ra
cy
i
s
n
o
t
n
eed
ed
˗
Rem
o
v
e
r
ed
u
n
d
a
n
t
at
t
r
i
b
u
t
e
˗
Crea
t
e
a
b
a
l
a
n
ce
d
d
a
t
a
s
e
t
.
˗
O
v
er
s
am
p
l
i
n
g
w
i
l
l
i
n
cre
as
e
t
ra
i
n
i
n
g
t
i
me.
˗
U
n
d
er
-
s
am
p
l
i
n
g
c
an
ca
u
s
e
t
h
e
l
o
s
s
o
f
i
m
p
o
r
t
a
n
t
i
n
f
o
r
mat
i
o
n
.
6.
Sp
i
ra
l
l
i
fe
c
y
c
l
e
m
o
d
el
-
b
as
ed
Bay
es
i
a
n
Cl
a
s
s
i
fi
cat
i
o
n
˗
Can
h
an
d
l
e
an
u
n
cer
t
a
i
n
d
at
as
et
.
˗
H
i
g
h
acc
u
ra
cy
,
s
t
ab
i
l
i
t
y
a
n
d
c
o
n
s
i
s
t
e
n
c
y
.
˗
Co
m
p
u
t
at
i
o
n
t
i
me
r
ed
u
c
ed
.
˗
T
h
e
f
ai
l
u
re
ra
t
i
o
i
s
l
o
w
.
˗
Can
o
n
l
y
d
e
t
ec
t
8
1
%
o
f
f
au
l
t
y
mo
d
u
l
es
u
s
i
n
g
t
h
e
cl
as
s
i
fi
ca
t
i
o
n
a
n
d
cl
u
s
t
er
i
n
g
-
b
as
ed
a
l
g
o
r
i
t
h
m.
7.
Met
ri
c
b
as
ed
o
n
a
n
e
u
ra
l
n
e
t
w
o
rk
c
l
a
s
s
i
f
i
er
˗
T
i
me
an
d
s
t
o
ra
g
e
s
p
ac
e
c
an
b
e
re
d
u
c
ed
.
˗
Rem
o
v
al
o
f
m
u
l
t
i
c
o
l
l
i
n
ear
i
t
y
N
/
A
8.
G
rey
s
y
s
t
e
m
t
h
e
o
ry
-
b
as
ed
p
re
d
i
ct
i
o
n
˗
H
e
l
p
s
t
o
r
ev
ea
l
t
h
e
f
l
u
ct
u
a
t
i
n
g
ra
n
g
e
o
f
t
h
e
fa
u
l
t
.
˗
H
a
n
d
l
e
s
ma
l
l
a
n
d
u
n
cer
t
a
i
n
d
at
as
et
s
.
˗
L
i
mi
t
e
d
t
o
i
t
s
i
n
t
ri
n
s
i
c
l
i
mi
t
a
t
i
o
n
s
.
9.
O
n
l
i
n
e
f
ai
l
u
re
p
r
ed
i
c
t
i
o
n
b
a
s
e
d
o
n
f
u
zz
y
ru
l
e
a
n
d
d
at
a
an
a
l
y
s
i
s
˗
N
o
fa
i
l
u
re
p
a
t
t
er
n
o
r
ex
p
e
ct
ed
v
al
u
e
i
s
n
eed
ed
.
˗
Can
d
ea
l
w
i
t
h
a
v
ar
i
a
b
l
e
t
h
at
i
s
d
i
s
cre
t
e,
co
n
t
i
n
u
o
u
s
a
n
d
l
i
n
g
u
i
s
t
i
c
.
˗
A
v
o
i
d
f
ak
e
re
g
r
es
s
i
o
n
˗
A
f
u
z
zy
r
u
l
e
i
s
n
o
t
a
l
w
a
y
s
a
cc
u
ra
t
e.
10
E
n
er
g
y
-
b
as
ed
a
n
o
ma
l
y
d
e
t
ec
t
i
o
n
˗
O
v
erc
o
m
e
l
i
mi
t
a
t
i
o
n
s
o
f
s
i
g
n
a
t
u
re
-
b
a
s
e
d
as
w
el
l
as
s
ee
d
e
d
an
d
n
o
n
-
s
ee
d
e
d
d
a
t
a
-
d
r
i
v
e
n
a
p
p
r
o
ac
h
e
s
.
˗
T
h
e
e
x
p
er
i
m
en
t
a
l
r
es
u
l
t
s
are
far
fro
m
b
e
i
n
g
co
n
c
l
u
s
i
v
e.
11
D
ee
p
l
e
ar
n
i
n
g
t
ec
h
n
i
q
u
e
-
b
a
s
e
d
mo
d
e
l
c
al
l
e
d
V
A
E
˗
Imp
ro
v
e
s
t
h
e
ab
i
l
i
t
y
t
o
p
re
d
i
ct
fa
i
l
u
re
d
a
t
a
w
h
i
l
e
t
h
e
p
re
d
i
ct
i
o
n
o
f
n
o
n
-
fa
i
l
u
r
e
d
a
t
a
i
s
b
e
i
n
g
p
e
rfo
r
me
d
.
N
/
A
12
Bay
es
i
a
n
b
e
l
i
ef
n
e
t
w
o
r
k
-
b
a
s
e
d
m
o
d
e
l
˗
A
s
s
i
s
t
s
d
e
v
e
l
o
p
ers
i
n
ac
h
i
ev
i
n
g
a
s
t
a
t
e
o
f
t
o
t
al
s
o
f
t
w
ar
e
fau
l
t
s
t
h
a
t
h
a
s
b
ee
n
t
a
rg
et
ed
.
N
/
A
13
SV
M
c
l
a
s
s
i
f
i
er
˗
Imp
ro
v
e
s
cl
a
s
s
i
fi
cat
i
o
n
ac
cu
rac
y
.
N
/
A
14
Mac
h
i
n
e
l
ear
n
i
n
g
al
g
o
ri
t
h
m
s
an
d
t
ec
h
n
i
q
u
e
s
˗
D
T
,
N
N
,
a
n
d
B
ag
g
i
n
g
ca
n
p
re
d
i
c
t
s
i
n
g
l
e
an
d
m
u
l
t
i
p
l
e
fai
l
u
re
cl
a
s
s
e
s
.
˗
SV
Ms
ca
n
n
o
t
p
r
ed
i
c
t
mu
l
t
i
p
l
e
fai
l
u
re
s
a
n
d
t
h
ei
r
p
erf
o
r
ma
n
ce
d
r
o
p
s
w
h
e
n
p
re
d
i
ct
i
n
g
i
n
d
i
v
i
d
u
al
fa
i
l
u
res
.
15
H
O
R
A
˗
H
a
s
h
i
g
h
e
r
m
o
d
u
l
a
r
i
t
y
b
y
r
e
u
s
i
n
g
d
i
f
f
e
r
e
n
t
t
e
c
h
n
i
q
u
e
s
t
o
p
r
e
d
i
c
t
f
a
i
l
u
r
e
a
m
o
n
g
t
h
e
c
o
m
p
o
n
e
n
t
s
o
f
t
h
e
s
y
s
t
e
m
.
˗
Can
b
e
u
s
ed
i
f
m
o
n
i
t
o
r
i
n
g
o
f
d
a
t
a
i
s
co
n
s
t
a
n
t
l
y
co
l
l
ec
t
e
d
.
1
6
.
Imp
ro
v
i
n
g
s
o
f
t
w
are
fa
u
l
t
p
re
d
i
ct
i
o
n
w
i
t
h
t
h
re
s
h
o
l
d
v
a
l
u
es
˗
T
h
e
r
es
u
l
t
w
i
l
l
h
el
p
k
n
o
w
t
h
e
e
ffec
t
o
f
t
h
e
t
h
re
s
h
o
l
d
f
o
r
t
h
e
t
ec
h
n
i
q
u
e
o
f
p
re
-
p
r
o
ce
s
s
i
n
g
t
o
e
n
h
a
n
ce
t
h
e
s
o
ft
w
ar
e
d
efe
ct
p
re
d
i
ct
i
o
n
˗
Fo
c
u
s
o
n
l
y
o
n
C
K
me
t
r
i
c
s
˗
T
h
e
r
e
a
r
e
s
e
v
e
r
a
l
m
e
t
h
o
d
s
f
o
r
t
e
s
t
i
n
g
s
o
f
t
w
a
r
e
a
n
d
c
e
r
t
a
i
n
m
e
t
r
i
c
s
m
a
y
h
a
v
e
d
i
f
f
e
r
e
n
t
m
e
a
n
i
n
g
s
o
f
v
a
r
i
a
n
t
s
.
F
o
r
e
x
a
m
p
l
e
,
t
h
e
W
M
C
m
e
t
r
i
c
h
a
s
t
w
o
m
a
i
n
c
o
n
c
e
p
t
s
,
a
n
d
t
h
e
r
e
a
r
e
s
e
v
e
r
a
l
v
a
r
i
a
n
t
s
t
o
t
h
e
L
C
O
M
m
e
t
r
i
c
.
B
u
t
i
n
t
h
i
s
s
t
u
d
y
,
t
h
e
m
e
t
r
i
c
s
d
a
t
a
a
r
e
b
a
s
e
d
o
n
o
r
i
g
i
n
a
l
c
o
n
c
e
p
t
s
f
o
r
p
o
s
s
i
b
l
e
c
o
m
p
a
r
i
s
o
n
s
.
17
D
ee
p
l
e
ar
n
i
n
g
-
b
a
s
e
d
s
o
ft
w
a
re
d
ef
ec
t
p
re
d
i
ct
i
o
n
˗
T
h
e
c
o
u
n
t
o
f
fa
u
l
t
p
re
d
i
ct
i
o
n
i
s
v
ery
eff
e
ct
i
v
e
˗
V
ar
i
o
u
s
d
a
t
a
s
e
t
s
t
h
at
u
s
e
d
d
i
ffe
re
n
t
met
r
i
c
s
co
u
l
d
af
fec
t
t
h
e
p
erf
o
r
ma
n
ce
o
f
fa
u
l
t
p
re
d
i
ct
i
o
n
˗
T
h
e
a
p
p
r
o
ac
h
w
as
o
n
l
y
e
v
a
l
u
at
ed
b
a
s
e
d
o
n
t
w
o
o
p
e
n
d
a
t
a
s
e
t
s
,
s
o
t
h
e
o
u
t
c
o
me
ma
y
n
o
t
b
e
t
h
e
s
ame
f
o
r
o
t
h
e
r
c
o
m
merc
i
a
l
l
y
a
v
a
i
l
a
b
l
e
s
o
ft
w
a
re.
18
T
h
e
s
o
ft
w
are
fa
u
l
t
p
re
d
i
c
t
i
o
n
mo
d
e
l
b
a
s
e
d
o
n
t
h
e
A
l
t
a
Ri
ca
l
a
n
g
u
a
g
e
˗
A
p
t
f
o
r
s
t
at
e
t
ra
n
s
i
t
i
o
n
,
d
at
a
i
n
t
er
act
i
o
n
,
ab
l
e
t
o
p
r
o
ce
s
s
co
m
p
l
e
x
s
y
s
t
e
ms
p
rec
i
s
el
y
an
d
u
s
e
t
r
av
e
rs
a
l
s
ea
rc
h
met
h
o
d
t
o
d
et
erm
i
n
e
t
h
e
s
o
ft
w
a
re
v
i
o
l
a
t
i
o
n
NA
19
A
s
t
u
d
y
o
n
s
o
ft
w
a
re
met
ri
c
s
e
l
ec
t
i
o
n
f
o
r
s
o
f
t
w
ar
e
fa
u
l
t
p
re
d
i
ct
i
o
n
˗
L
R
p
erf
o
rme
d
t
h
e
b
es
t
NA
20
Co
m
p
r
eh
en
s
i
v
e
m
o
d
e
l
f
o
r
s
o
ft
w
a
re
f
au
l
t
p
re
d
i
ct
i
o
n
˗
Man
y
ru
l
e
-
b
as
e
d
l
ear
n
er
s
h
a
v
e
b
ee
n
u
s
e
d
f
o
r
c
o
m
p
ar
i
s
o
n
t
o
g
et
a
n
a
p
p
r
o
p
ri
at
e
re
s
u
l
t
N
/
A
21
W
ra
p
p
er
-
b
as
ed
s
el
e
ct
i
o
n
met
h
o
d
b
y
PS
O
-
MG
A
˗
H
i
g
h
e
s
t
ac
cu
rac
y
r
at
e
a
mo
n
g
o
t
h
er
t
y
p
es
o
f
cl
a
s
s
i
f
i
ca
t
i
o
n
me
t
h
o
d
s
.
˗
Sh
o
u
l
d
r
ed
u
c
e
n
o
i
s
e
a
n
d
rem
o
v
e
i
rre
l
e
v
a
n
t
d
a
t
a
f
i
r
s
t
.
˗
Co
u
l
d
b
e
i
m
p
r
o
v
ed
b
y
u
s
i
n
g
A
I.
22
ML
me
t
h
o
d
fo
r
s
o
ft
w
a
re
fau
l
t
p
re
d
i
ct
i
o
n
˗
Pro
v
i
d
e
a
s
a
t
i
s
fac
t
o
ry
re
s
u
l
t
i
n
p
re
d
i
c
t
i
n
g
f
au
l
t
s
˗
T
h
e
met
h
o
d
i
s
s
t
i
l
l
i
n
t
h
e
ex
p
er
i
me
n
t
a
l
s
t
ag
e
a
n
d
f
u
t
u
re
res
e
arc
h
i
s
n
ee
d
e
d
.
23
Bay
es
i
a
n
r
eg
u
l
ar
i
za
t
i
o
n
(BR
)
t
ec
h
n
i
q
u
e
˗
Fau
l
t
p
re
d
i
ct
i
o
n
d
u
ri
n
g
t
h
e
d
e
s
i
g
n
p
h
a
s
e
˗
Bes
t
al
g
o
ri
t
h
m
t
o
ap
p
l
y
c
o
m
p
ar
ed
t
o
o
t
h
ers
N
/
A
24
It
er
at
e
d
fea
t
u
re
s
e
l
e
ct
i
o
n
al
g
o
ri
t
h
m
an
d
L
-
R
N
N
˗
A
b
l
e
t
o
s
el
e
ct
t
h
e
m
o
s
t
i
mp
o
r
t
a
n
t
s
o
f
t
w
a
re
me
t
r
i
c
s
˗
A
b
l
e
t
o
rec
ei
v
e
a
g
o
o
d
cl
as
s
i
f
i
ca
t
i
o
n
r
at
e
N
/
A
25
Sem
i
-
s
u
p
er
v
i
s
ed
D
FC
M
met
h
o
d
˗
Mu
l
t
i
p
l
e
c
l
u
s
t
er
s
ca
n
b
e
ama
l
g
am
at
ed
˗
In
c
o
r
p
o
ra
t
e
l
a
b
e
l
e
d
d
a
t
a
a
n
d
u
n
l
a
b
e
l
e
d
d
at
a
˗
Rem
o
v
i
n
g
e
x
ce
s
s
i
v
e
fe
at
u
re
s
N
/
A
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
1
6
,
No
.
1
,
Feb
r
u
ar
y
20
2
6
:
5
4
5
-
554
552
I
n
th
e
ASR
A
m
o
d
el,
th
e
c
h
i
-
s
q
u
ar
e
m
eth
o
d
is
u
s
ed
t
o
d
is
ca
r
d
o
r
i
g
n
o
r
e
ir
r
ele
v
an
t
a
n
d
r
ed
u
n
d
an
t
f
ea
tu
r
es.
Af
ter
f
ilter
in
g
th
e
s
u
b
s
et,
b
o
th
o
v
er
-
s
am
p
lin
g
a
n
d
u
n
d
er
-
s
am
p
lin
g
ar
e
u
s
ed
,
b
u
t
th
e
d
is
ad
v
a
n
tag
e
o
f
th
is
tech
n
iq
u
e
is
th
at
o
v
er
-
s
am
p
lin
g
co
u
ld
in
c
r
ea
s
e
th
e
tr
ain
in
g
tim
e
an
d
u
n
d
e
r
-
s
am
p
lin
g
co
u
ld
lead
to
lo
s
s
o
f
cr
itical
in
f
o
r
m
atio
n
b
y
Z
h
o
u
e
t
a
l.
[
8
]
.
E
v
en
t
h
o
u
g
h
u
n
d
er
-
s
am
p
lin
g
ca
n
p
r
o
d
u
ce
f
aster
tr
ain
in
g
p
r
o
ce
s
s
es,
it
ca
n
ca
u
s
e
lo
wer
p
e
r
f
o
r
m
an
ce
i
n
C
am
p
o
s
et
a
l.
[
1
4
]
.
An
o
th
er
m
et
h
o
d
is
th
e
wr
a
p
p
er
-
b
ased
s
elec
tio
n
m
eth
o
d
,
wh
er
e
th
e
o
p
tim
al
p
a
r
am
et
er
ca
n
b
e
o
b
tain
ed
b
y
lear
n
in
g
f
r
o
m
th
e
tr
ain
in
g
o
f
th
e
d
atasets
.
T
h
e
ad
v
an
tag
es
a
r
e
th
at
it
is
m
o
r
e
ac
cu
r
ate
co
m
p
ar
ed
to
o
th
e
r
f
ea
tu
r
e
s
elec
tio
n
m
et
h
o
d
s
,
s
u
c
h
as
em
b
ed
d
ed
f
ea
t
u
r
e
s
elec
tio
n
an
d
f
ilter
-
b
ased
s
elec
tio
n
b
y
B
an
g
a
an
d
B
an
s
al
[
2
5
]
.
I
n
t
h
e
p
ap
er
b
y
T
u
r
ab
ieh
et
a
l.
[
2
8
]
,
wr
ap
p
er
-
b
ased
s
elec
tio
n
is
f
u
r
th
er
im
p
r
o
v
e
d
b
y
p
er
f
o
r
m
in
g
iter
atio
n
.
T
h
e
ad
v
an
tag
es
o
f
th
is
m
eth
o
d
a
r
e
th
a
t
it
s
to
p
s
iter
atin
g
if
th
e
o
p
tim
al
v
alu
e
is
r
ea
ch
ed
an
d
allo
ws y
o
u
to
f
in
d
th
e
m
o
s
t v
alu
ab
le
m
etr
ics.
Oth
er
th
an
t
h
at,
in
th
e
m
etr
ic
-
b
ased
n
eu
r
al
n
etwo
r
k
class
if
ier
p
r
o
p
o
s
ed
b
y
J
ay
an
t
h
i
an
d
Flo
r
en
ce
[
1
0
]
,
p
r
in
cip
al
co
m
p
o
n
e
n
t
an
a
ly
s
is
(
P
C
A)
is
u
s
ed
,
wh
er
e
th
ey
lo
wer
th
e
p
r
o
p
o
r
tio
n
o
f
th
e
d
ataset
as
well
a
s
th
e
d
im
en
s
io
n
ality
wh
ile
r
e
d
u
cin
g
t
h
e
in
f
o
r
m
atio
n
lo
s
s
.
T
h
e
PC
A
is
th
en
f
u
r
th
er
im
p
r
o
v
e
d
b
y
a
d
d
in
g
m
ax
im
u
m
-
lik
elih
o
o
d
esti
m
atio
n
to
r
ed
u
ce
th
e
r
ec
o
n
s
tr
u
ctio
n
er
r
o
r
.
T
h
e
ad
v
an
tag
es
o
f
th
i
s
m
eth
o
d
ar
e
th
at
it
allo
ws
th
e
d
ataset
to
b
e
d
i
s
tr
ib
u
ted
eq
u
ally
with
o
u
t
af
f
ec
tin
g
o
th
e
r
u
n
d
is
tr
ib
u
ted
d
ata.
B
ased
o
n
th
e
ex
p
er
im
en
t,
all
th
e
d
if
f
e
r
en
t
d
atasets
h
av
e
an
ac
cu
r
ac
y
o
f
m
o
r
e
th
an
8
0
%,
wh
ich
s
h
o
ws
th
at
PC
A
h
as
an
im
p
o
r
tan
t
r
o
le
in
th
e
p
r
ed
ictio
n
ac
c
u
r
ac
y
.
L
astl
y
,
i
n
th
e
p
a
p
er
f
r
o
m
Ma
o
[
1
1
]
,
g
r
ey
s
eq
u
en
ce
g
en
er
atio
n
is
u
s
ed
to
h
an
d
le
d
atasets
,
b
u
t
th
e
d
r
awb
ac
k
is
th
at
it
ca
n
o
n
ly
h
a
n
d
le
s
m
all
d
ata
s
izes
co
m
p
ar
e
d
to
o
th
e
r
m
eth
o
d
s
th
at
h
a
n
d
le
b
ig
d
atasets
.
3
.
2
.
Cla
s
s
if
ica
t
io
n
T
h
e
s
ec
o
n
d
asp
ec
t
is
th
e
cla
s
s
if
icatio
n
p
r
o
ce
s
s
in
p
r
ed
ict
in
g
s
o
f
twar
e
f
au
lts
.
T
h
e
r
e
ar
e
s
ev
er
al
class
if
ier
s
u
s
ed
f
o
r
f
au
lty
an
d
n
o
n
-
f
a
u
lty
s
o
f
twar
e
co
m
p
o
n
e
n
ts
class
if
icatio
n
.
On
e
o
f
th
e
class
if
ier
s
is
SVM.
I
n
th
e
s
tu
d
y
b
y
R
aju
et
a
l.
[
1
]
,
th
is
cla
s
s
if
ier
i
s
u
s
ed
a
s
a
b
in
ar
y
class
if
ier
f
o
r
class
if
y
in
g
s
ev
er
al
d
atasets
in
p
u
t
an
d
e
x
p
er
im
en
tal
r
esu
lts
s
h
o
w
ed
th
at
it
h
as
th
e
ac
cu
r
ac
y
a
b
o
v
e
9
7
%
f
o
r
C
M1
d
ata
s
et,
9
9
.
5
4
%
f
o
r
J
M1
d
ata
s
et,
9
6
.
5
8
%
f
o
r
KC
1
d
ata
s
et,
9
9
.
9
5
%
f
o
r
KC
2
d
ata
s
et,
9
9
.
5
7
f
o
r
PC
1
d
ata
s
et,
an
d
9
3
.
8
9
f
o
r
th
e
Data
T
r
iev
e
d
ata
s
et.
B
e
s
id
es,
ac
co
r
d
in
g
to
B
an
g
a
an
d
B
an
s
al
[
2
5
]
,
a
v
ar
iatio
n
o
f
th
e
SVM
clas
s
if
icatio
n
alg
o
r
ith
m
k
n
o
wn
as
least
s
q
u
ar
es
twin
S
VM
(
MW
-
L
STSVM
)
also
h
as
th
e
h
ig
h
est
ac
cu
r
ac
y
,
with
9
1
.
3
%,
th
an
o
th
er
class
if
ier
s
lik
e
k
-
NN
an
d
L
STSVM
,
an
d
th
e
ef
f
icien
cy
o
f
th
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
(
PS
O
-
MG
A)
is
9
.
8
%
h
ig
h
er
th
an
o
t
h
er
tech
n
iq
u
es in
f
a
u
lt a
n
d
n
o
n
-
f
a
u
lty
m
o
d
u
les cla
s
s
if
icatio
n
.
Alth
o
u
g
h
SVMs
ca
n
p
r
ed
ict
o
n
e
f
ailu
r
e
class
,
h
o
wev
er
,
it
ca
n
n
o
t
b
e
u
s
ed
wh
en
co
n
s
id
er
in
g
m
u
ltip
le
f
ailu
r
e
m
o
d
es
b
y
C
am
p
o
s
et
a
l.
[
1
4
]
.
T
h
e
au
th
o
r
s
s
tated
th
at
wh
en
co
n
s
id
er
i
n
g
b
o
th
o
n
e
a
n
d
m
u
ltip
le
f
ailu
r
es,
DT
,
NN,
an
d
B
ag
g
in
g
alg
o
r
i
th
m
s
ca
n
p
r
ed
ict
th
em
well.
Nex
t,
o
th
er
ty
p
ical
class
if
ier
s
b
y
Su
n
et
a
l.
[
7
]
,
wh
er
e
f
iv
e
ty
p
ical
class
if
ier
s
a
r
e
u
s
ed
to
p
r
ed
ict
s
o
f
twar
e
f
ail
u
r
e,
in
clu
d
in
g
SVM,
wh
ich
ar
e
R
F,
DT
,
N
B
,
an
d
L
R
.
I
n
Dh
an
ajay
an
an
d
Pil
lai
[
9
]
,
B
ay
esian
class
if
icatio
n
i
s
u
s
ed
in
th
e
p
r
o
p
o
s
ed
tech
n
i
q
u
e
ca
lled
SLM
B
C
to
g
r
o
u
p
o
r
class
if
y
f
au
lty
an
d
n
o
n
-
f
au
lt
y
m
o
d
u
les
u
s
in
g
a
p
r
o
b
ab
ilit
y
d
is
tr
ib
u
tio
n
.
As
a
r
esu
lt,
FNR
,
F
PR
,
an
d
o
v
er
all
er
r
o
r
r
ate
ca
n
b
e
g
r
ea
tly
r
ed
u
ce
d
u
s
in
g
SLM
B
C
th
an
o
th
er
tech
n
iq
u
es.
Af
ter
th
at,
in
J
ay
an
th
i
an
d
Flo
r
en
ce
[
1
0
]
an
d
Ma
h
aja
n
et
a
l.
[
2
7
]
,
NN
class
if
ier
i
s
u
s
ed
wh
ile
T
u
r
a
b
ieh
et
a
l
.
[
2
8
]
p
r
o
p
o
s
ed
a
class
if
icatio
n
tech
n
iq
u
e
ca
lle
d
L
-
R
NN
u
s
ed
in
th
e
p
r
o
p
o
s
ed
alg
o
r
ith
m
ca
lled
iter
ated
f
ea
tu
r
e
s
elec
tio
n
alg
o
r
ith
m
an
d
as
a
co
n
s
eq
u
en
ce
,
it
ca
n
g
ain
well
r
ate
o
f
cla
s
s
if
icatio
n
b
ased
o
n
AUC
r
esu
lts
,
o
u
tp
er
f
o
r
m
in
g
NB
,
ANN,
k
-
NN,
an
d
C
4
.
5
.
4.
RE
CO
M
M
E
NDA
T
I
O
N
Dea
lin
g
with
s
o
f
twar
e
f
au
lts
is
a
v
er
y
im
p
o
r
tan
t
task
wh
e
n
d
ea
lin
g
with
s
o
f
twar
e
r
elia
b
ilit
y
an
d
q
u
ality
ass
u
r
an
ce
.
I
n
o
u
r
r
e
v
iew
p
ap
er
,
we
r
ev
iewe
d
r
esea
r
ch
p
ap
er
s
a
n
d
m
o
d
els
th
at
f
o
cu
s
o
n
Fau
lt
an
d
Failu
r
e
Pre
d
ictio
n
T
ec
h
n
iq
u
es
.
o
u
r
an
aly
s
is
is
b
ased
o
n
d
ata
p
r
o
ce
s
s
in
g
tech
n
iq
u
es,
class
if
icatio
n
tech
n
iq
u
es
an
d
m
etr
ics
to
co
m
e
u
p
with
a
r
ec
o
m
m
en
d
atio
n
o
n
wh
ich
SF
P
tech
n
iq
u
es
ar
e
m
o
r
e
co
n
v
en
ien
t
f
o
r
s
o
m
e
s
ce
n
ar
io
s
.
So
m
e
o
f
th
e
tr
ad
iti
o
n
al
m
o
d
els
in
SF
P
ca
n
b
e
im
p
r
o
v
e
d
an
d
ap
p
lied
t
o
s
o
f
twar
e
s
y
s
tem
s
to
en
s
u
r
e
a
b
etter
q
u
ality
i
n
p
r
e
d
ictin
g
f
au
lts
an
d
f
ailu
r
es
.
All
o
f
th
e
r
ev
iewe
d
SF
P
tech
n
iq
u
es
h
av
e
th
eir
o
wn
a
d
v
an
tag
es
an
d
lim
itatio
n
s
u
n
d
e
r
d
if
f
er
e
n
t
s
ce
n
ar
io
s
.
T
h
u
s
,
th
e
team
h
as
a
f
ew
r
ec
o
m
m
en
d
atio
n
s
o
n
h
o
w
to
im
p
r
o
v
e
th
e
r
eliab
ilit
y
o
f
s
o
f
twar
e
p
r
o
d
u
cts
u
s
in
g
s
o
f
twar
e
f
au
lt
a
n
d
f
ailu
r
e
p
r
ed
ictio
n
tech
n
iq
u
es
.
I
n
t
er
m
s
o
f
th
e
u
s
ed
d
ataset,
s
o
m
e
m
eth
o
d
s
ca
n
not
d
is
ca
r
d
o
r
ig
n
o
r
e
ir
r
elev
an
t
a
n
d
r
ed
u
n
d
a
n
t
d
ata,
an
d
f
r
o
m
o
u
r
an
al
y
s
is
,
we
s
u
g
g
est
u
s
in
g
th
e
wr
ap
p
e
r
-
b
ased
s
elec
tio
n
m
eth
o
d
b
ec
au
s
e
it
is
th
e
m
o
s
t
ac
cu
r
ate
s
elec
tio
n
m
eth
o
d
a
m
o
n
g
th
e
o
th
er
m
ajo
r
s
elec
tio
n
m
eth
o
d
s
.
Als
o
,
it c
an
b
e
im
p
r
o
v
ed
b
y
u
s
in
g
iter
atio
n
s
.
I
n
ter
m
s
o
f
th
e
class
if
icatio
n
asp
ec
t,
we
s
u
g
g
est
u
s
in
g
t
h
e
n
eu
r
al
n
etwo
r
k
class
if
ier
p
r
o
p
o
s
ed
b
y
J
ay
an
th
i a
n
d
Flo
r
en
ce
[
1
0
]
.
T
h
e
ad
v
an
tag
e
o
f
th
is
m
eth
o
d
is
th
at
it a
llo
w
s
th
e
d
ataset
to
b
e
d
is
tr
ib
u
ted
eq
u
ally
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
A
s
ystema
tic
r
ev
ie
w
o
f so
ftw
a
r
e
fa
u
lt p
r
ed
ictio
n
…
(
R
.
K
a
n
e
s
a
r
a
j R
a
ma
s
a
my
)
553
with
o
u
t
af
f
ec
tin
g
o
th
e
r
u
n
d
i
s
tr
ib
u
ted
d
ata.
B
ased
o
n
th
e
ex
p
er
im
e
n
t,
all
th
e
d
if
f
er
e
n
t
d
atasets
h
av
e
an
ac
cu
r
ac
y
o
f
m
o
r
e
t
h
an
8
0
%.
I
n
B
an
g
a
an
d
B
an
s
al
[
2
5
]
,
a
v
ar
iatio
n
o
f
t
h
e
SVM
class
if
icatio
n
alg
o
r
ith
m
k
n
o
wn
as
MW
-
L
STSVM
also
h
as
th
e
h
ig
h
est
ac
cu
r
ac
y
w
ith
9
1
.
3
%,
th
an
o
th
er
class
if
i
er
s
lik
e
k
-
NN
an
d
L
STSVM
,
an
d
th
e
ef
f
icien
c
y
o
f
th
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
(
PS
O
-
MG
A)
is
in
cr
ea
s
ed
b
y
9
.
8
%
th
an
o
t
h
er
tech
n
iq
u
es.
SVMs
ca
n
p
r
ed
i
ct
o
n
e
f
ailu
r
e
m
o
d
e,
wh
ile
f
o
r
s
in
g
le
an
d
m
u
ltip
le
f
ailu
r
e
m
o
d
es,
DT
an
d
B
ag
g
in
g
ca
n
p
r
ed
ict
th
em
well.
5.
CO
NCLU
SI
O
N
I
n
co
n
clu
s
io
n
,
we
f
o
u
n
d
t
h
at
it
h
as
p
len
ty
o
f
way
s
an
d
te
ch
n
iq
u
es
to
p
er
f
o
r
m
s
o
f
twar
e
f
au
lt
an
d
f
ailu
r
e
p
r
e
d
ictio
n
.
B
ased
o
n
t
h
e
to
tal
o
f
2
5
r
ev
iews
in
th
is
r
esear
ch
p
a
p
er
,
we
ce
n
tr
alize
d
o
u
r
an
aly
s
is
an
d
ca
teg
o
r
ized
o
u
r
an
al
y
s
is
in
to
d
ata
p
r
o
ce
s
s
in
g
tech
n
iq
u
es
an
d
class
if
icatio
n
tech
n
iq
u
es.
Data
p
r
o
ce
s
s
in
g
tech
n
iq
u
es
m
ain
ly
co
n
tr
o
l
th
e
ac
cu
r
ac
y
o
f
t
h
e
class
if
ier
,
an
d
th
e
class
if
icatio
n
tech
n
iq
u
e
is
th
e
way
o
f
ca
teg
o
r
izin
g
in
p
u
t
d
atasets
.
T
h
r
o
u
g
h
th
e
a
n
aly
s
is
,
we
h
a
v
e
s
h
o
wn
s
ev
er
al
c
o
m
p
ar
is
o
n
s
o
n
d
ata
p
r
o
ce
s
s
in
g
tech
n
iq
u
es
as
well
as
class
i
f
icatio
n
tech
n
iq
u
es.
Fo
r
t
h
e
an
aly
s
is
r
esu
lt,
we
f
in
alize
d
o
u
r
an
aly
s
is
an
d
r
ec
o
m
m
en
d
ed
s
ev
er
al
tech
n
iq
u
es,
wh
ich
ar
e
m
eth
o
d
s
an
d
NN
c
lass
if
ier
f
o
r
d
ata
p
r
o
ce
s
s
in
g
tech
n
iq
u
e,
a
n
d
SVM
f
o
r
class
if
icatio
n
tech
n
iq
u
e
wr
ap
p
e
r
-
b
ased
s
elec
tio
n
.
T
h
ese
tech
n
iq
u
es
h
av
e
th
e
b
es
t
ac
cu
r
ac
y
an
d
also
ef
f
icien
cy
f
o
r
f
au
lt
an
d
f
ailu
r
e
p
r
e
d
ictio
n
b
ased
o
n
o
u
r
r
esea
r
ch
.
Fo
r
f
u
tu
r
e
wo
r
k
,
we
m
ay
lo
o
k
at
th
e
r
esear
ch
an
d
d
is
co
v
er
m
o
r
e
tech
n
iq
u
es
as
well
as
wr
ite
a
r
esear
ch
p
a
p
er
th
at
is
m
o
r
e
in
-
d
ep
th
in
to
th
e
p
r
o
ce
s
s
o
f
f
au
lt
an
d
f
ailu
r
e
p
r
ed
ictio
n
.
RE
F
E
R
E
NC
E
S
[
1
]
K
.
S
.
R
a
j
u
,
M
.
R
.
M
u
r
t
y
,
M
.
V
.
R
a
o
,
a
n
d
S
.
C
.
S
a
t
a
p
a
t
h
y
,
“
S
u
p
p
o
r
t
v
e
c
t
o
r
mac
h
i
n
e
w
i
t
h
k
-
f
o
l
d
c
r
o
ss
v
a
l
i
d
a
t
i
o
n
m
o
d
e
l
f
o
r
so
f
t
w
a
r
e
f
a
u
l
t
p
r
e
d
i
c
t
i
o
n
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
P
u
re
a
n
d
A
p
p
l
i
e
d
M
a
t
h
e
m
a
t
i
c
s
,
v
o
l
.
1
1
8
,
n
o
.
2
0
,
p
p
.
3
2
1
–
3
3
4
,
2
0
1
8
.
[
2
]
S
.
G
u
p
t
a
,
A
.
M
i
s
h
r
a
,
a
n
d
M
.
C
h
a
w
l
a
,
“
A
n
a
l
y
s
i
s
a
n
d
r
e
c
o
mm
e
n
d
a
t
i
o
n
o
f
c
o
mm
o
n
f
a
u
l
t
a
n
d
f
a
i
l
u
r
e
i
n
s
o
f
t
w
a
r
e
d
e
v
e
l
o
p
m
e
n
t
sy
st
e
ms,
”
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
S
i
g
n
a
l
Pr
o
c
e
ssi
n
g
,
C
o
m
m
u
n
i
c
a
t
i
o
n
,
P
o
w
e
r
a
n
d
Em
b
e
d
d
e
d
S
y
st
e
m
,
S
C
O
P
ES
2
0
1
6
-
Pro
c
e
e
d
i
n
g
s
,
p
p
.
1
7
3
0
–
1
7
3
4
,
2
0
1
7
,
d
o
i
:
1
0
.
1
1
0
9
/
S
C
O
P
ES.
2
0
1
6
.
7
9
5
5
7
3
9
.
[
3
]
C
.
L
o
u
,
P
.
H
u
a
n
g
,
a
n
d
S
.
S
mi
t
h
,
“
U
n
d
e
r
st
a
n
d
i
n
g
,
d
e
t
e
c
t
i
n
g
a
n
d
l
o
c
a
l
i
z
i
n
g
p
a
r
t
i
a
l
f
a
i
l
u
r
e
s
i
n
l
a
r
g
e
sy
s
t
e
m
so
f
t
w
a
r
e
,
”
i
n
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
1
7
th
U
S
E
N
I
X
S
y
m
p
o
si
u
m
o
n
N
e
t
w
o
r
k
e
d
S
y
s
t
e
m
s D
e
si
g
n
a
n
d
I
m
p
l
e
m
e
n
t
a
t
i
o
n
,
N
S
D
I
2
0
2
0
,
2
0
2
0
,
p
p
.
5
5
9
–
5
7
4
.
[
4
]
T.
P
i
t
a
k
r
a
t
,
D
.
O
k
a
n
o
v
i
ć
,
A
.
v
a
n
H
o
o
r
n
,
a
n
d
L.
G
r
u
n
s
k
e
,
“
H
O
R
A
:
A
r
c
h
i
t
e
c
t
u
r
e
-
a
w
a
r
e
o
n
l
i
n
e
f
a
i
l
u
r
e
p
r
e
d
i
c
t
i
o
n
,
”
J
o
u
rn
a
l
o
f
S
y
s
t
e
m
s
a
n
d
S
o
f
t
w
a
r
e
,
v
o
l
.
1
3
7
,
p
p
.
6
6
9
–
6
8
5
,
2
0
1
8
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
j
ss.
2
0
1
7
.
0
2
.
0
4
1
.
[
5
]
S
.
C
h
a
t
t
e
r
j
e
e
a
n
d
B
.
M
a
j
i
,
“
A
B
a
y
e
si
a
n
b
e
l
i
e
f
n
e
t
w
o
r
k
b
a
se
d
mo
d
e
l
f
o
r
p
r
e
d
i
c
t
i
n
g
s
o
f
t
w
a
r
e
f
a
u
l
t
s
i
n
e
a
r
l
y
p
h
a
se
o
f
s
o
f
t
w
a
r
e
d
e
v
e
l
o
p
me
n
t
p
r
o
c
e
ss,
”
A
p
p
l
i
e
d
I
n
t
e
l
l
i
g
e
n
c
e
,
v
o
l
.
4
8
,
n
o
.
8
,
p
p
.
2
2
1
4
–
2
2
2
8
,
2
0
1
8
,
d
o
i
:
1
0
.
1
0
0
7
/
s1
0
4
8
9
-
0
1
7
-
1
0
7
8
-
x.
[
6
]
D
.
Jau
k
,
D
.
Y
a
n
g
,
a
n
d
M
.
S
c
h
u
l
z
,
“
P
r
e
d
i
c
t
i
n
g
f
a
u
l
t
s
i
n
h
i
g
h
p
e
r
f
o
r
m
a
n
c
e
c
o
m
p
u
t
i
n
g
s
y
st
e
ms:
A
n
i
n
-
d
e
p
t
h
s
u
r
v
e
y
o
f
t
h
e
st
a
t
e
-
of
-
t
h
e
-
p
r
a
c
t
i
c
e
,
”
S
C
1
9
:
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
f
o
r
H
i
g
h
Pe
r
f
o
rm
a
n
c
e
C
o
m
p
u
t
i
n
g
,
N
e
t
w
o
rki
n
g
,
S
t
o
r
a
g
e
a
n
d
A
n
a
l
y
s
i
s
,
D
e
n
v
e
r
,
C
O
,
U
S
A
,
2
0
1
9
,
p
p
.
1
–
1
3
,
d
o
i
:
1
0
.
1
1
4
5
/
3
2
9
5
5
0
0
.
3
3
5
6
1
8
5
.
[
7
]
Y
.
S
u
n
,
L.
X
u
,
Y
.
L
i
,
L.
G
u
o
,
Z
.
M
a
,
a
n
d
Y
.
W
a
n
g
,
“
U
t
i
l
i
z
i
n
g
d
e
e
p
a
r
c
h
i
t
e
c
t
u
r
e
n
e
t
w
o
r
k
s
o
f
V
A
E
i
n
s
o
f
t
w
a
r
e
f
a
u
l
t
p
r
e
d
i
c
t
i
o
n
,
”
i
n
Pr
o
c
e
e
d
i
n
g
s
-
16
th
I
E
EE
I
n
t
e
r
n
a
t
i
o
n
a
l
S
y
m
p
o
si
u
m
o
n
P
a
r
a
l
l
e
l
a
n
d
D
i
st
ri
b
u
t
e
d
P
ro
c
e
ss
i
n
g
w
i
t
h
A
p
p
l
i
c
a
t
i
o
n
s
,
1
7
th
I
EEE
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
U
b
i
q
u
i
t
o
u
s
C
o
m
p
u
t
i
n
g
a
n
d
C
o
m
m
u
n
i
c
a
t
i
o
n
s
,
8
th
I
EEE
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
Bi
g
D
a
t
a
a
n
d
C
l
o
u
d
C
o
m
p
u
t
i
n
g
,
1
1
t
h
,
2
0
1
8
,
p
p
.
8
7
0
–
877
,
d
o
i
:
1
0
.
1
1
0
9
/
B
D
C
l
o
u
d
.
2
0
1
8
.
0
0
1
2
9
.
[
8
]
L.
Z
h
o
u
,
R
.
Li
,
S
.
Z
h
a
n
g
,
a
n
d
H
.
W
a
n
g
,
“
I
mb
a
l
a
n
c
e
d
d
a
t
a
p
r
o
c
e
ss
i
n
g
m
o
d
e
l
f
o
r
s
o
f
t
w
a
r
e
d
e
f
e
c
t
p
r
e
d
i
c
t
i
o
n
,
”
Wi
re
l
e
ss
Pe
rs
o
n
a
l
C
o
m
m
u
n
i
c
a
t
i
o
n
s
,
v
o
l
.
1
0
2
,
n
o
.
2
,
p
p
.
9
3
7
–
9
5
0
,
2
0
1
8
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
1
2
7
7
-
0
1
7
-
5
1
1
7
-
z.
[
9
]
C
.
G
.
R
.
D
h
a
n
a
j
a
y
a
n
a
n
d
S
.
A
.
P
i
l
l
a
i
,
“
S
L
M
B
C
:
s
p
i
r
a
l
l
i
f
e
c
y
c
l
e
mo
d
e
l
-
b
a
se
d
B
a
y
e
s
i
a
n
c
l
a
ssi
f
i
c
a
t
i
o
n
t
e
c
h
n
i
q
u
e
f
o
r
e
f
f
i
c
i
e
n
t
so
f
t
w
a
r
e
f
a
u
l
t
p
r
e
d
i
c
t
i
o
n
a
n
d
c
l
a
ss
i
f
i
c
a
t
i
o
n
,
”
S
o
f
t
C
o
m
p
u
t
i
n
g
,
v
o
l
.
2
1
,
n
o
.
2
,
p
p
.
4
0
3
–
4
1
5
,
2
0
1
7
,
d
o
i
:
1
0
.
1
0
0
7
/
s
0
0
5
0
0
-
016
-
2
3
1
6
-
6.
[
1
0
]
R
.
Ja
y
a
n
t
h
i
a
n
d
L.
F
l
o
r
e
n
c
e
,
“
S
o
f
t
w
a
r
e
d
e
f
e
c
t
p
r
e
d
i
c
t
i
o
n
t
e
c
h
n
i
q
u
e
s
u
si
n
g
me
t
r
i
c
s
b
a
se
d
o
n
n
e
u
r
a
l
n
e
t
w
o
r
k
c
l
a
ssi
f
i
e
r
,
”
C
l
u
st
e
r
C
o
m
p
u
t
i
n
g
,
v
o
l
.
2
2
,
n
o
.
s1
,
p
p
.
7
7
–
8
8
,
2
0
1
9
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
0
5
8
6
-
018
-
1
7
3
0
-
1.
[
1
1
]
C
.
M
a
o
,
“
S
o
f
t
w
a
r
e
f
a
u
l
t
s
p
r
e
d
i
c
t
i
o
n
b
a
s
e
d
o
n
g
r
e
y
sy
s
t
e
m
t
h
e
o
r
y
,
”
AC
M
S
I
G
S
O
FT
S
o
f
t
w
a
r
e
E
n
g
i
n
e
e
r
i
n
g
N
o
t
e
s
,
v
o
l
.
3
4
,
n
o
.
2
,
p
p
.
1
–
6
,
2
0
0
9
,
d
o
i
:
1
0
.
1
1
4
5
/
1
5
0
7
1
9
5
.
1
5
0
7
2
0
6
.
[
1
2
]
Z.
D
i
n
g
,
Y
.
Z
h
o
u
,
G
.
P
u
,
a
n
d
M
.
Z
h
o
u
,
“
O
n
l
i
n
e
f
a
i
l
u
r
e
p
r
e
d
i
c
t
i
o
n
f
o
r
r
a
i
l
w
a
y
t
r
a
n
sp
o
r
t
a
t
i
o
n
s
y
st
e
ms
b
a
se
d
o
n
f
u
z
z
y
r
u
l
e
s a
n
d
d
a
t
a
a
n
a
l
y
si
s
,
”
I
E
EE
T
r
a
n
sa
c
t
i
o
n
s
o
n
R
e
l
i
a
b
i
l
i
t
y
,
v
o
l
.
6
7
,
n
o
.
3
,
p
p
.
1
1
4
3
–
1
1
5
8
,
2
0
1
8
,
d
o
i
:
1
0
.
1
1
0
9
/
T
R
.
2
0
1
8
.
2
8
2
8
1
1
3
.
[
1
3
]
C
.
M
o
n
n
i
a
n
d
M
.
P
e
z
z
e
,
“
En
e
r
g
y
-
b
a
s
e
d
a
n
o
ma
l
y
d
e
t
e
c
t
i
o
n
a
n
e
w
p
e
r
sp
e
c
t
i
v
e
f
o
r
p
r
e
d
i
c
t
i
n
g
s
o
f
t
w
a
r
e
f
a
i
l
u
r
e
s,
”
i
n
Pr
o
c
e
e
d
i
n
g
s
-
2
0
1
9
I
E
EE/
AC
M
4
1
st
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
S
o
f
t
w
a
re
E
n
g
i
n
e
e
r
i
n
g
:
N
e
w
I
d
e
a
s
a
n
d
Em
e
rg
i
n
g
Re
s
u
l
t
s,
I
C
S
E
-
N
I
ER
2
0
1
9
,
2
0
1
9
,
p
p
.
6
9
–
72
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
S
E
-
N
I
ER
.
2
0
1
9
.
0
0
0
2
6
.
[
1
4
]
J.
R
.
C
a
m
p
o
s
,
M
.
V
i
e
i
r
a
,
a
n
d
E.
C
o
s
t
a
,
“
E
x
p
l
o
r
a
t
o
r
y
s
t
u
d
y
o
f
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
t
e
c
h
n
i
q
u
e
s
f
o
r
s
u
p
p
o
r
t
i
n
g
f
a
i
l
u
r
e
p
r
e
d
i
c
t
i
o
n
,
”
i
n
Pro
c
e
e
d
i
n
g
s
-
2
0
1
8
1
4
th
Eu
r
o
p
e
a
n
D
e
p
e
n
d
a
b
l
e
C
o
m
p
u
t
i
n
g
C
o
n
f
e
r
e
n
c
e
,
ED
C
C
2
0
1
8
,
2
0
1
8
,
p
p
.
9
–
16
,
d
o
i
:
1
0
.
1
1
0
9
/
ED
C
C
.
2
0
1
8
.
0
0
0
1
4
.
[
1
5
]
R
.
S
h
a
t
n
a
w
i
,
“
I
mp
r
o
v
i
n
g
s
o
f
t
w
a
r
e
f
a
u
l
t
p
r
e
d
i
c
t
i
o
n
w
i
t
h
t
h
r
e
sh
o
l
d
v
a
l
u
e
s
,
”
i
n
2
0
1
8
2
6
th
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
S
o
f
t
w
a
re
,
T
e
l
e
c
o
m
m
u
n
i
c
a
t
i
o
n
s
a
n
d
C
o
m
p
u
t
e
r
N
e
t
w
o
r
k
s,
S
o
f
t
C
O
M
2
0
1
8
,
2
0
1
8
,
p
p
.
2
8
6
–
2
9
0
,
d
o
i
:
1
0
.
2
3
9
1
9
/
S
O
F
T
C
O
M
.
2
0
1
8
.
8
5
5
5
8
1
8
.
[
1
6
]
L.
Q
i
a
o
,
X
.
L
i
,
Q
.
U
m
e
r
,
a
n
d
P
.
G
u
o
,
“
D
e
e
p
l
e
a
r
n
i
n
g
b
a
se
d
s
o
f
t
w
a
r
e
d
e
f
e
c
t
p
r
e
d
i
c
t
i
o
n
,
”
N
e
u
ro
c
o
m
p
u
t
i
n
g
,
v
o
l
.
3
8
5
,
p
p
.
1
0
0
–
1
1
0
,
2
0
2
0
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
n
e
u
c
o
m
.
2
0
1
9
.
1
1
.
0
6
7
.
[
1
7
]
P
.
S
i
n
g
h
,
“
C
o
m
p
r
e
h
e
n
si
v
e
m
o
d
e
l
f
o
r
so
f
t
w
a
r
e
f
a
u
l
t
p
r
e
d
i
c
t
i
o
n
,
”
i
n
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
I
n
v
e
n
t
i
v
e
C
o
m
p
u
t
i
n
g
a
n
d
I
n
f
o
rm
a
t
i
c
s,
I
C
I
C
I
2
0
1
7
,
2
0
1
8
,
p
p
.
1
1
0
3
–
1
1
0
8
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
I
C
I
.
2
0
1
7
.
8
3
6
5
3
1
1
.
[
1
8
]
J.
S
o
n
g
,
B
.
C
h
e
n
,
X
.
L
i
,
Y
.
Y
a
n
g
,
C
.
Li
u
,
a
n
d
H
.
L
i
,
“
T
h
e
so
f
t
w
a
r
e
f
a
u
l
t
p
r
e
d
i
c
t
i
o
n
m
o
d
e
l
b
a
se
d
o
n
t
h
e
A
l
t
a
R
i
c
a
l
a
n
g
u
a
g
e
,
”
i
n
Pro
c
e
e
d
i
n
g
s
o
f
2
0
1
9
I
EEE
3
rd
I
n
f
o
r
m
a
t
i
o
n
T
e
c
h
n
o
l
o
g
y
,
N
e
t
w
o
r
k
i
n
g
,
El
e
c
t
r
o
n
i
c
a
n
d
Au
t
o
m
a
t
i
o
n
C
o
n
t
ro
l
C
o
n
f
e
re
n
c
e
,
I
T
N
E
C
2
0
1
9
,
2
0
1
9
,
p
p
.
2
5
4
9
–
2
5
5
2
,
d
o
i
:
1
0
.
1
1
0
9
/
I
TN
EC
.
2
0
1
9
.
8
7
2
9
2
3
5
.
[
1
9
]
H
.
W
a
n
g
a
n
d
T
.
M
.
K
h
o
s
h
g
o
f
t
a
a
r
,
“
A
st
u
d
y
o
n
s
o
f
t
w
a
r
e
me
t
r
i
c
s
e
l
e
c
t
i
o
n
f
o
r
so
f
t
w
a
r
e
f
a
u
l
t
p
r
e
d
i
c
t
i
o
n
,
”
i
n
Pr
o
c
e
e
d
i
n
g
s
-
18
th
I
EEE
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
Ma
c
h
i
n
e
L
e
a
rn
i
n
g
a
n
d
A
p
p
l
i
c
a
t
i
o
n
s
,
I
C
MLA
2
0
1
9
,
2
0
1
9
,
p
p
.
1
0
4
5
–
1
0
5
0
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
M
LA
.
2
0
1
9
.
0
0
1
7
6
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
1
6
,
No
.
1
,
Feb
r
u
ar
y
20
2
6
:
5
4
5
-
554
554
[
2
0
]
X
.
Ji
,
X
.
L
i
a
n
g
,
Y
.
Z
h
o
u
,
Y
.
H
u
o
,
X
.
S
h
i
,
a
n
d
Y
.
Y
a
n
g
,
“
A
f
a
u
l
t
p
r
e
d
i
c
t
i
o
n
m
e
t
h
o
d
b
a
se
d
o
n
l
o
a
d
-
c
a
p
a
c
i
t
y
mo
d
e
l
i
n
t
h
e
c
o
mm
u
n
i
c
a
t
i
o
n
n
e
t
w
o
r
k
,
”
i
n
2
0
1
9
2
0
t
h
Asi
a
-
P
a
c
i
f
i
c
N
e
t
w
o
rk
O
p
e
ra
t
i
o
n
s
a
n
d
Ma
n
a
g
e
m
e
n
t
S
y
m
p
o
si
u
m
:
Ma
n
a
g
e
m
e
n
t
i
n
a
C
y
b
e
r
-
P
h
y
s
i
c
a
l
W
o
rl
d
,
AP
N
O
M
S
2
0
1
9
,
2
0
1
9
,
p
p
.
1
8
–
21
,
d
o
i
:
1
0
.
2
3
9
1
9
/
A
P
N
O
M
S
.
2
0
1
9
.
8
8
9
3
1
0
4
.
[
2
1
]
L.
W
a
n
g
,
Y
.
F
u
,
M
.
Z.
Q
.
C
h
e
n
,
a
n
d
X
.
Y
a
n
g
,
“
C
o
n
t
r
o
l
l
a
b
i
l
i
t
y
r
o
b
u
st
n
e
ss
f
o
r
sca
l
e
-
f
r
e
e
n
e
t
w
o
r
k
s
b
a
s
e
d
o
n
n
o
n
l
i
n
e
a
r
l
o
a
d
-
c
a
p
a
c
i
t
y
,
”
N
e
u
ro
c
o
m
p
u
t
i
n
g
,
v
o
l
.
2
5
1
,
p
p
.
9
9
–
1
0
5
,
2
0
1
7
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
n
e
u
c
o
m
.
2
0
1
7
.
0
4
.
0
1
1
.
[
2
2
]
R
.
Tr
i
p
a
t
h
i
a
n
d
M
.
S
a
r
a
sw
a
t
,
“
A
sse
s
smen
t
&
p
r
e
d
i
c
t
i
o
n
o
f
so
f
t
w
a
r
e
r
e
l
i
a
b
i
l
i
t
y
:
A
N
N
a
p
p
r
o
a
c
h
,
”
i
n
S
S
R
N
E
l
e
c
t
r
o
n
i
c
J
o
u
r
n
a
l
,
2
0
2
0
,
p
p
.
1
–
4
,
d
o
i
:
1
0
.
2
1
3
9
/
ssr
n
.
3
5
6
5
8
7
3
.
[
2
3
]
S
.
R
.
A
z
i
z
,
T.
K
h
a
n
,
a
n
d
A
.
N
a
d
e
e
m,
“
Ex
p
e
r
i
me
n
t
a
l
v
a
l
i
d
a
t
i
o
n
o
f
i
n
h
e
r
i
t
a
n
c
e
me
t
r
i
c
s’
i
m
p
a
c
t
o
n
so
f
t
w
a
r
e
f
a
u
l
t
p
r
e
d
i
c
t
i
o
n
,
”
I
EEE
Ac
c
e
ss
,
v
o
l
.
7
,
p
p
.
8
5
2
6
2
–
8
5
2
7
5
,
2
0
1
9
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
1
9
.
2
9
2
4
0
4
0
.
[
2
4
]
G
.
Ja
b
e
e
n
,
X
.
Y
a
n
g
,
L
.
P
i
n
g
,
S
.
R
a
h
i
m,
G
.
S
a
h
a
r
,
a
n
d
A
.
A
.
S
h
a
h
,
“
H
y
b
r
i
d
so
f
t
w
a
r
e
r
e
l
i
a
b
i
l
i
t
y
p
r
e
d
i
c
t
i
o
n
m
o
d
e
l
b
a
s
e
d
o
n
r
e
si
d
u
a
l
e
r
r
o
r
s,”
i
n
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
I
EEE
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
S
o
f
t
w
a
r
e
En
g
i
n
e
e
ri
n
g
a
n
d
S
e
rvi
c
e
S
c
i
e
n
c
e
s
,
I
C
S
E
S
S
,
2
0
1
7
,
p
p
.
4
7
9
–
482
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
S
ESS
.
2
0
1
7
.
8
3
4
2
9
5
9
.
[
2
5
]
M
.
B
a
n
g
a
a
n
d
A
.
B
a
n
sa
l
,
“
P
r
o
p
o
se
d
so
f
t
w
a
r
e
f
a
u
l
t
s
d
e
t
e
c
t
i
o
n
u
si
n
g
h
y
b
r
i
d
a
p
p
r
o
a
c
h
,
”
S
e
c
u
ri
t
y
a
n
d
Pr
i
v
a
c
y
,
v
o
l
.
6
,
n
o
.
4
,
p
p
.
1
–
1
4
,
D
e
c
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
0
2
/
s
p
y
2
.
1
0
3
.
[
2
6
]
B
.
W
ó
j
c
i
c
k
i
a
n
d
R
.
D
ą
b
r
o
w
s
k
i
,
“
A
p
p
l
y
i
n
g
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
t
o
so
f
t
w
a
r
e
f
a
u
l
t
p
r
e
d
i
c
t
i
o
n
,
”
E
-
I
n
f
o
rm
a
t
i
c
a
S
o
f
t
w
a
r
e
E
n
g
i
n
e
e
r
i
n
g
J
o
u
rn
a
l
,
v
o
l
.
1
2
,
n
o
.
1
,
p
p
.
1
9
9
–
2
1
6
,
2
0
1
8
,
d
o
i
:
1
0
.
5
2
7
7
/
e
-
I
n
f
1
8
0
1
0
8
.
[
2
7
]
R
.
M
a
h
a
j
a
n
,
S
.
K
.
G
u
p
t
a
,
a
n
d
R
.
K
.
B
e
d
i
,
“
D
e
s
i
g
n
o
f
so
f
t
w
a
r
e
f
a
u
l
t
p
r
e
d
i
c
t
i
o
n
m
o
d
e
l
u
si
n
g
B
R
t
e
c
h
n
i
q
u
e
,
”
Pr
o
c
e
d
i
a
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
v
o
l
.
4
6
,
p
p
.
8
4
9
–
8
5
8
,
2
0
1
5
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
p
r
o
c
s.
2
0
1
5
.
0
2
.
1
5
4
.
[
2
8
]
H
.
T
u
r
a
b
i
e
h
,
M
.
M
a
f
a
r
j
a
,
a
n
d
X
.
Li
,
“
I
t
e
r
a
t
e
d
f
e
a
t
u
r
e
se
l
e
c
t
i
o
n
a
l
g
o
r
i
t
h
ms
w
i
t
h
l
a
y
e
r
e
d
r
e
c
u
r
r
e
n
t
n
e
u
r
a
l
n
e
t
w
o
r
k
f
o
r
so
f
t
w
a
r
e
f
a
u
l
t
p
r
e
d
i
c
t
i
o
n
,
”
Ex
p
e
rt
S
y
st
e
m
s
w
i
t
h
A
p
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
1
2
2
,
p
p
.
2
7
–
4
2
,
2
0
1
9
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
e
sw
a
.
2
0
1
8
.
1
2
.
0
3
3
.
[
2
9
]
A
.
A
r
sh
a
d
,
S
.
R
i
a
z
,
L.
J
i
a
o
,
a
n
d
A
.
M
u
r
t
h
y
,
“
S
e
mi
-
s
u
p
e
r
v
i
se
d
d
e
e
p
f
u
z
z
y
c
-
mea
n
c
l
u
st
e
r
i
n
g
f
o
r
s
o
f
t
w
a
r
e
f
a
u
l
t
p
r
e
d
i
c
t
i
o
n
,
”
I
EEE
A
c
c
e
ss
,
v
o
l
.
6
,
p
p
.
2
5
6
7
5
–
2
5
6
8
5
,
2
0
1
8
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
1
8
.
2
8
3
5
3
0
4
.
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
R.
K
a
n
e
sa
r
a
j
R
a
m
a
sa
m
y
is
a
n
a
ss
o
c
iate
p
ro
fe
ss
o
r
a
t
th
e
F
a
c
u
lt
y
o
f
C
o
m
p
u
ti
n
g
a
n
d
In
f
o
rm
a
ti
c
s
a
n
d
De
p
u
ty
Di
re
c
to
r
o
f
t
h
e
To
u
c
h
p
o
in
t
M
a
n
a
g
e
m
e
n
t
Offic
e
,
M
u
lt
ime
d
i
a
Un
iv
e
rsity
.
W
it
h
o
v
e
r
1
5
y
e
a
rs
o
f
e
x
p
e
rien
c
e
a
c
ro
ss
a
c
a
d
e
m
ia,
in
d
u
str
y
,
a
n
d
n
a
ti
o
n
a
l
in
it
iati
v
e
s,
h
e
sp
e
c
ialize
s in
AI,
I
o
T,
a
n
d
p
re
d
icti
v
e
a
n
a
ly
ti
c
s,
b
rid
g
i
n
g
re
se
a
rc
h
with
re
a
l
-
wo
rl
d
a
p
p
li
c
a
ti
o
n
s.
A
P
h
.
D
.
h
o
ld
e
r
in
IT,
p
r
o
fe
ss
io
n
a
l
tec
h
n
o
l
o
g
ist
(
M
BOT),
a
n
d
S
e
n
i
o
r
I
EE
E
M
e
m
b
e
r,
h
e
h
a
s
led
re
se
a
rc
h
p
r
o
jec
ts
e
x
c
e
e
d
in
g
R
M
1
.
0
5
m
il
li
o
n
a
n
d
CS
R
in
i
ti
a
ti
v
e
s
w
o
rth
RM
1
3
7
,
0
0
0
,
p
ro
d
u
c
in
g
a
wa
rd
-
w
in
n
i
n
g
in
n
o
v
a
ti
o
n
s
su
c
h
a
s
P
u
lse
Wi
se
a
n
d
S
m
a
rt
To
il
e
t
.
He
h
a
s
p
u
b
li
sh
e
d
wi
d
e
ly
i
n
S
COPU
S
Q1
/Q2
j
o
u
r
n
a
ls,
h
o
ld
s
fi
v
e
c
o
p
y
ri
g
h
ts
a
n
d
o
n
e
p
a
ten
t,
a
n
d
h
a
s
a
d
v
a
n
c
e
d
M
M
U’s
p
re
d
icti
v
e
a
n
a
ly
ti
c
s
tNP
S
sy
ste
m
a
c
r
o
ss
3
8
d
e
p
a
rtme
n
ts.
His
c
o
n
tri
b
u
ti
o
n
s
e
x
ten
d
t
o
M
a
lay
sia
’s
Na
ti
o
n
a
l
I
o
T
P
o
li
c
y
,
wh
il
e
h
e
a
c
ti
v
e
l
y
su
p
e
rv
ise
s
p
o
stg
ra
d
u
a
te
re
se
a
rc
h
,
d
e
li
v
e
rs
p
r
o
fe
ss
io
n
a
l
train
in
g
to
o
rg
a
n
iza
ti
o
n
s
su
c
h
a
s
TM
,
M
C
M
C,
a
n
d
P
OS
M
a
lay
sia
,
a
n
d
c
h
a
m
p
i
o
n
s
ES
G
-
a
n
d
CS
R
-
d
riv
e
n
i
n
n
o
v
a
ti
o
n
s
in
h
e
a
lt
h
c
a
re
a
n
d
su
sta
in
a
b
il
it
y
.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
r.
k
a
n
e
sa
ra
j@m
m
u
.
e
d
u
.
m
y
.
Ve
n
u
shi
n
i
R
a
je
n
d
r
a
n
is
a
lec
tu
re
r
a
t
th
e
F
a
c
u
lt
y
o
f
Co
m
p
u
ti
n
g
a
n
d
In
f
o
rm
a
ti
c
s,
M
u
lt
ime
d
ia
Un
i
v
e
rsity
,
with
a
m
a
ste
r
o
f
sc
ien
c
e
sp
e
c
ializin
g
in
we
b
se
rv
ice
s
a
n
d
Io
T.
F
o
rm
e
rly
a
re
se
a
rc
h
sc
h
o
lar
with
Tele
k
o
m
M
a
lay
sia
R&
D,
sh
e
c
o
n
tri
b
u
te
d
to
p
r
o
jec
ts
in
we
b
,
m
o
b
il
e
,
a
n
d
I
o
T,
e
a
rn
i
n
g
re
c
o
g
n
it
io
n
a
s
a
M
a
lay
sia
n
Yo
u
n
g
Dig
i
tal
He
ro
.
P
a
ss
io
n
a
te
a
b
o
u
t
t
h
e
in
ters
e
c
ti
o
n
o
f
tec
h
n
o
lo
g
y
a
n
d
h
u
m
a
n
b
e
h
a
v
i
o
r,
sh
e
a
lso
h
o
ld
s
m
u
lt
i
p
le
M
icr
o
so
f
t
c
e
rti
fica
ti
o
n
s,
s
h
o
wc
a
sin
g
h
e
r
e
x
p
e
rti
se
i
n
m
o
d
e
rn
d
i
g
it
a
l
t
o
o
ls
a
n
d
p
o
siti
o
n
i
n
g
h
e
r
a
s
a
p
ro
m
isin
g
tale
n
t
i
n
d
i
g
it
a
l
tec
h
n
o
lo
g
y
a
n
d
i
n
n
o
v
a
ti
o
n
.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
v
e
n
u
s
h
in
i@m
m
u
.
e
d
u
.
m
y
.
Pa
r
a
m
e
sw
a
r
a
n
S
u
b
r
a
m
a
n
i
a
n
is
a
n
e
d
u
c
a
ti
o
n
a
l
a
d
m
in
istrato
r
with
o
v
e
r
2
2
y
e
a
r
s
o
f
e
x
p
e
rien
c
e
in
c
o
ll
e
g
iate
a
n
d
u
n
iv
e
rsit
y
e
d
u
c
a
ti
o
n
.
He
h
o
l
d
s
a
d
o
c
to
ra
te
a
n
d
s
p
e
c
ialize
s
in
tax
a
ti
o
n
,
s
u
p
p
o
rted
b
y
a
m
a
ste
r
d
e
g
re
e
i
n
a
c
c
o
u
n
ti
n
g
,
fi
n
a
n
c
e
,
a
n
d
h
u
m
a
n
re
so
u
rc
e
m
a
n
a
g
e
m
e
n
t.
He
h
a
s
e
x
ten
siv
e
e
x
p
e
rien
c
e
h
a
n
d
li
n
g
a
c
c
re
d
it
a
ti
o
n
p
ro
c
e
ss
e
s
fo
r
th
e
M
a
lay
sia
n
Qu
a
li
fica
ti
o
n
A
g
e
n
c
y
(
M
QA
),
UK
u
n
i
v
e
rsiti
e
s
(QA
A),
a
n
d
S
in
g
a
p
o
re
(S
QC)
.
Dr.
P
a
ra
m
e
sw
a
ra
n
h
a
s
m
a
n
a
g
e
d
g
o
v
e
rn
m
e
n
t
-
fu
n
d
e
d
p
ro
jec
ts
i
n
tax
a
ti
o
n
a
n
d
b
lo
c
k
c
h
a
i
n
so
lu
ti
o
n
s
f
o
r
b
u
sin
e
ss
.
A
c
e
rti
fie
d
train
e
r
a
n
d
a
d
v
o
c
a
te
fo
r
o
u
tco
m
e
-
b
a
se
d
e
d
u
c
a
ti
o
n
(OBE
)
,
Dr.
P
a
ra
m
e
sw
a
ra
n
h
a
s
b
e
e
n
wi
d
e
ly
re
c
o
g
n
ize
d
in
h
is
f
ield
.
He
h
o
ld
s
se
v
e
n
p
a
ten
ts
a
n
d
h
a
s
p
u
b
li
sh
e
d
2
7
re
se
a
rc
h
p
a
p
e
rs
in
S
c
o
p
u
s
,
IS
I,
a
n
d
p
e
e
r
-
re
v
iew
e
d
j
o
u
r
n
a
ls.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
p
a
ra
m
e
sw
a
ra
n
.
s@
c
h
ristu
n
i
v
e
rsity
.
i
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.