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
.
4
,
A
u
g
u
s
t
20
2
6
,
p
p
.
1
735
~
1
7
5
4
I
SS
N:
2088
-
8
7
0
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijece.
v
1
6
i
4
.
pp
1
7
3
5
-
1
7
5
4
1735
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ec
e.
ia
esco
r
e.
co
m
A new
diag
no
stic
metho
d bas
ed on
supp
o
rt
v
ector
m
a
chine f
o
r
sho
rt
circui
t
w
ind
ing
f
a
ults in
indu
ction mo
tors
H
icha
m
Z
a
im
en¹,
T
a
wf
ik
T
hela
idjia
²,
M
a
k
hlo
uf
Cho
uk
i³
,
Abdu
rr
a
hm
a
n Üns
a
l⁴
1
La
b
o
r
a
t
o
r
y
o
f
E
l
e
c
t
r
i
c
a
l
E
n
g
i
n
e
e
r
i
n
g
,
El
e
c
t
r
o
n
i
c
a
n
d
R
e
n
e
w
a
b
l
e
E
n
e
r
g
y
,
U
n
i
v
e
r
s
i
t
y
o
f
S
o
u
k
A
h
r
a
s,
S
o
u
k
A
h
r
a
s,
A
l
g
e
r
i
a
2
D
e
p
a
r
t
m
e
n
t
o
f
E
l
e
c
t
r
i
c
a
l
E
n
g
i
n
e
e
r
i
n
g
,
U
n
i
v
e
r
s
i
t
y
o
f
S
o
u
k
A
h
r
a
s,
S
o
u
k
A
h
r
a
s,
A
l
g
e
r
i
a
3
D
e
p
a
r
t
me
n
t
o
f
El
e
c
t
r
i
c
a
l
En
g
i
n
e
e
r
i
n
g
,
U
n
i
v
e
r
si
t
y
o
f
J
i
j
e
l
,
Ji
j
e
l
,
A
l
g
e
r
i
a
4
D
e
p
a
r
t
me
n
t
o
f
El
e
c
t
r
i
c
a
l
a
n
d
El
e
c
t
r
o
n
i
c
s E
n
g
i
n
e
e
r
i
n
g
,
K
ü
t
a
h
y
a
D
u
ml
u
p
i
n
a
r
U
n
i
v
e
r
si
t
y
,
K
ü
t
a
h
y
a
,
Tu
r
k
e
y
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Ma
r
5
,
2
0
2
6
R
ev
is
ed
Ma
y
1
5
,
2
0
2
6
Acc
ep
ted
J
u
l 2
2
,
2
0
2
6
In
ter
-
tu
r
n
sh
o
r
t
c
ircu
it
s
(IT
S
Cs
)
in
in
d
u
c
ti
o
n
m
o
to
r
(IM
)
wi
n
d
in
g
s
a
re
a
m
o
n
g
th
e
m
o
st
c
rit
ica
l
a
n
d
fre
q
u
e
n
t
fa
u
lt
s
in
i
n
d
u
strial
e
n
v
iro
n
m
e
n
ts,
a
s
th
e
y
c
a
n
ra
p
i
d
l
y
e
v
o
l
v
e
in
to
se
v
e
re
d
a
m
a
g
e
,
lea
d
in
g
t
o
u
n
p
lan
n
e
d
d
o
wn
t
ime
a
n
d
c
o
stly
m
a
in
te
n
a
n
c
e
.
To
e
n
h
a
n
c
e
th
e
re
li
a
b
il
it
y
o
f
IM
s,
t
h
is
p
a
p
e
r
p
ro
p
o
se
s
a
m
a
c
h
in
e
lea
rn
in
g
–
b
a
se
d
d
iag
n
o
sis
m
e
th
o
d
d
e
d
ica
ted
to
ITS
C
fa
il
u
re
s.
Th
e
d
e
v
e
lo
p
e
d
d
iag
n
o
sti
c
to
o
l
c
o
m
b
i
n
e
s
a
su
p
p
o
rt
v
e
c
to
r
m
a
c
h
in
e
(S
VM)
c
las
sifier
with
F
ish
e
r’s
ra
ti
o
(
F
R)
-
b
a
se
d
fe
a
tu
re
se
lec
ti
o
n
.
T
h
e
p
ro
p
o
se
d
fra
m
e
wo
rk
u
se
s
e
x
p
e
r
ime
n
tally
a
c
q
u
ired
c
u
rre
n
t
sig
n
a
ls
u
n
d
e
r
h
e
a
lt
h
y
c
o
n
d
it
io
n
s
a
n
d
fi
v
e
ITS
C
fa
u
lt
se
v
e
rit
y
lev
e
ls
(
1
%
–
5
%
),
e
v
a
lu
a
ted
a
c
ro
ss
fo
u
r
lo
a
d
c
o
n
d
it
i
o
n
s
(
2
5
%
,
5
0
%
,
7
5
%
,
a
n
d
1
0
0
%
).
Eac
h
sig
n
a
l
is
se
g
m
e
n
ted
in
t
o
2
0
0
n
o
n
-
o
v
e
rla
p
p
in
g
se
g
m
e
n
ts
(
5
0
0
sa
m
p
les
e
a
c
h
),
fro
m
wh
ich
n
in
e
ti
m
e
-
d
o
m
a
in
fe
a
tu
r
e
s
a
re
e
x
trac
ted
to
c
a
p
t
u
re
fa
u
lt
-
re
late
d
c
h
a
ra
c
teristics
.
Th
e
se
fe
a
tu
re
s
a
re
th
e
n
u
se
d
fo
r
train
i
n
g
a
n
d
tes
ti
n
g
a
S
V
M
c
las
sifier
c
a
p
a
b
le
o
f
d
isti
n
g
u
ish
in
g
b
e
twe
e
n
h
e
a
lt
h
y
sta
tes
a
n
d
lev
e
ls
o
f
se
v
e
rit
y
o
f
I
TS
C
fa
u
l
ts.
To
o
p
ti
m
ize
th
e
c
las
sifica
ti
o
n
p
r
o
c
e
ss
,
th
e
FR
a
lg
o
rit
h
m
is
e
m
p
lo
y
e
d
to
se
le
c
t
th
e
m
o
st
in
f
o
rm
a
ti
v
e
fe
a
tu
r
e
s
wh
il
e
d
isc
a
rd
in
g
t
h
o
se
wi
th
lo
w
re
lev
a
n
c
e
.
Ou
r
fi
n
d
in
g
s
u
n
v
e
i
led
t
h
a
t
th
e
p
r
o
p
o
se
d
h
y
b
rid
F
R
-
S
VM
-
b
a
se
d
d
iag
n
o
sis
a
c
h
iev
e
s
h
ig
h
d
iag
n
o
stic
a
c
c
u
ra
c
y
ra
n
g
i
n
g
fro
m
9
9
.
5
4
%
t
o
1
0
0
%
.
F
u
rth
e
rm
o
re
,
th
e
o
u
tco
m
e
s
p
ro
v
e
th
a
t
t
h
e
in
teg
ra
ted
tec
h
n
ica
l
fra
m
e
wo
rk
(t
ime
-
d
o
m
a
in
fe
a
tu
re
s
+
F
R
+
S
VM)
p
ro
v
id
e
s
z
e
ro
fa
lse
a
larm
s
a
n
d
a
b
a
lan
c
e
d
d
iag
n
o
stic
s
y
ste
m
th
a
t
c
o
m
b
in
e
s
c
o
m
p
u
tati
o
n
a
l
sp
e
e
d
with
h
i
g
h
p
re
c
isio
n
.
K
ey
w
o
r
d
s
:
C
o
n
d
itio
n
m
o
n
ito
r
in
g
Fau
lt c
lass
if
icat
io
n
Fis
h
er
’
s
r
atio
alg
o
r
ith
m
I
n
ter
-
tu
r
n
s
h
o
r
t c
ir
c
u
it f
au
lts
Ma
ch
in
e
lear
n
in
g
Su
p
p
o
r
t
v
ec
to
r
m
ac
h
in
e
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
:
Hich
am
Z
aim
en
L
ab
o
r
ato
r
y
o
f
E
lectr
ical
E
n
g
i
n
ee
r
in
g
,
E
lectr
o
n
ic
an
d
R
en
ewa
b
le
E
n
er
g
y
,
Un
iv
e
r
s
ity
o
f
So
u
k
Ah
r
as
So
u
k
Ah
r
as,
Alg
er
ia
E
m
ail:
h
ich
am
.
za
im
en
@
u
n
iv
-
s
o
u
k
ah
r
as.d
z
1.
I
NT
RO
D
UCT
I
O
N
W
ith
th
e
o
n
g
o
in
g
i
n
cr
ea
s
e
in
in
d
u
s
tr
ial
s
y
s
tem
s
o
p
h
is
ticatio
n
,
th
e
lik
elih
o
o
d
o
f
f
au
lts
co
r
r
esp
o
n
d
in
g
ly
r
is
es.
As
a
r
esu
lt,
th
e
to
p
ic
o
f
f
a
u
lt
d
iag
n
o
s
is
h
as
b
ec
o
m
e
m
o
r
e
im
p
o
r
t
an
t.
No
wad
ay
s
,
t
h
e
s
af
ety
an
d
r
eliab
ilit
y
o
f
elec
tr
ical
m
ac
h
in
es
h
av
e
b
ec
o
m
e
a
m
ajo
r
f
o
c
u
s
with
in
th
e
r
esear
ch
c
o
m
m
u
n
ity
[
1
]
.
Am
o
n
g
th
e
elec
tr
ical
m
ac
h
in
e
s
,
in
d
u
ctio
n
m
o
to
r
s
(
I
Ms)
p
lay
a
v
ital r
o
le
in
m
o
d
er
n
in
d
u
s
tr
ial
ap
p
licatio
n
s
d
u
e
to
th
eir
s
im
p
licity
,
r
o
b
u
s
tn
es
s
,
co
s
t
-
ef
f
ec
tiv
en
ess
,
an
d
ab
i
lity
to
o
p
er
ate
at
wid
e
s
p
ee
d
r
an
g
es.
T
h
ey
ar
e
b
r
o
ad
ly
u
s
ed
in
ap
p
licatio
n
s
r
an
g
in
g
f
r
o
m
m
an
u
f
ac
tu
r
in
g
to
HVAC
s
y
s
tem
s
an
d
tr
an
s
p
o
r
tatio
n
.
Desp
ite
th
ese
b
en
ef
its
,
th
e
co
m
p
o
n
e
n
ts
o
f
in
d
u
ctio
n
m
o
to
r
s
ar
e
s
u
s
ce
p
tib
le
to
n
u
m
er
o
u
s
ty
p
es
o
f
f
au
lts
,
wh
ic
h
a
r
e
ca
teg
o
r
ized
as
m
ec
h
an
ical
an
d
elec
tr
ical
[
2
]
.
E
lectr
ical
f
au
lts
in
clu
d
e
s
h
o
r
t
-
cir
cu
it,
b
r
o
k
en
-
b
a
r
,
an
d
o
p
e
n
-
p
h
ase
f
au
lts
,
wh
ile
m
ec
h
an
ic
al
o
n
es
in
clu
d
e
b
ea
r
in
g
,
ec
ce
n
tr
icity
,
an
d
m
is
alig
n
m
e
n
t
f
a
u
lts
[
3
]
.
T
h
e
I
T
SC
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
.
4
,
Au
g
u
s
t
20
2
6
:
1
7
3
5
-
1
754
1736
f
au
lts
in
th
e
s
tato
r
win
d
in
g
s
ar
e
am
o
n
g
th
e
m
o
s
t
cr
itical.
T
h
ese
f
au
lts
h
ap
p
e
n
wh
en
t
h
e
in
s
u
latio
n
b
etwe
en
tu
r
n
s
in
th
e
s
am
e
win
d
in
g
p
h
ase
b
r
ea
k
s
d
o
wn
,
ca
u
s
in
g
a
lo
ca
lized
s
h
o
r
t
cir
cu
it.
I
f
n
o
t
d
et
ec
te
d
a
t
a
n
ea
r
l
y
s
tag
e,
I
T
SC
f
au
lts
c
an
p
r
o
g
r
es
s
r
a
p
i
d
l
y
,
le
ad
in
g
t
o
s
e
v
er
e
m
o
to
r
d
am
a
g
e
,
u
n
p
l
an
n
e
d
s
h
u
t
d
o
wn
s
,
a
n
d
e
x
p
e
n
s
i
v
e
r
e
p
a
ir
s
.
Acc
o
r
d
i
n
g
t
o
[
4
]
,
[
5
]
,
am
o
n
g
a
ll
f
a
u
lts
o
f
i
n
d
u
ct
i
o
n
m
o
to
r
s
,
I
T
SC
is
t
h
e
n
u
m
b
e
r
o
n
e
o
cc
u
r
r
e
n
c
e.
Stu
d
i
es i
n
d
i
ca
t
e
t
h
at
a
p
p
r
o
x
im
atel
y
3
0
–
4
0
%
o
f
i
n
d
u
cti
o
n
m
o
to
r
f
a
il
u
r
es
ar
e
att
r
i
b
u
te
d
t
o
th
i
s
t
y
p
e
o
f
f
a
u
l
t.
T
h
e
d
ia
g
n
o
s
is
o
f
an
I
T
SC
f
a
u
lt
in
its
ea
r
ly
s
tag
e
is
cr
u
ci
al
f
o
r
r
ed
u
cin
g
d
o
wn
tim
e
a
n
d
k
ee
p
in
g
s
y
s
tem
r
eliab
ilit
y
.
Ov
er
r
ec
en
t
d
ec
ad
es,
r
esear
ch
er
s
in
th
e
d
iag
n
o
s
is
an
d
class
if
icatio
n
to
p
i
cs
h
av
e
d
ev
elo
p
ed
n
u
m
er
o
u
s
s
tr
ateg
ies
to
d
iag
n
o
s
e
an
d
d
etec
t
f
au
lts
in
elec
tr
ical
m
ac
h
in
es,
aim
in
g
to
en
h
a
n
ce
th
eir
s
af
ety
an
d
s
er
v
ice
life
wh
ile
r
ed
u
cin
g
s
h
u
td
o
wn
lo
s
s
es.
B
o
n
n
ett
et
a
l.
[
6
]
p
r
o
v
id
ed
a
n
in
-
d
ep
th
an
a
ly
s
is
o
f
s
tato
r
an
d
r
o
to
r
f
ailu
r
e
ca
u
s
es,
em
p
h
asi
zin
g
th
e
im
p
o
r
tan
ce
o
f
ea
r
ly
d
etec
tio
n
o
f
I
T
SC
f
au
lts
to
av
o
id
ca
tast
r
o
p
h
ic
co
n
s
eq
u
en
ce
s
.
T
h
e
d
iag
n
o
s
is
ap
p
r
o
ac
h
es
ar
e
b
r
o
a
d
ly
d
iv
i
d
ed
in
to
s
ig
n
al
p
r
o
ce
s
s
in
g
-
b
ased
,
m
o
d
el
-
b
ased
,
ar
tific
ial
in
tellig
en
ce
-
b
ased
,
a
n
d
h
y
b
r
id
-
b
ased
tech
n
i
q
u
es.
C
las
s
ical
f
au
lt
d
iag
n
o
s
is
m
eth
o
d
s
,
s
u
ch
as
th
er
m
al
m
o
n
ito
r
in
g
[
7
]
,
[
8
]
v
ib
r
atio
n
an
al
y
s
is
[
9
]
,
[
1
0
]
an
d
ac
o
u
s
tic
em
is
s
io
n
,
o
f
te
n
r
eq
u
ir
e
ad
d
itio
n
al
s
en
s
o
r
s
an
d
ar
e
s
en
s
itiv
e
to
o
p
er
ati
n
g
co
n
d
itio
n
s
a
n
d
en
v
ir
o
n
m
en
tal
n
o
is
e.
I
n
co
n
tr
ast,
m
o
to
r
cu
r
r
e
n
t
s
ig
n
atu
r
e
a
n
aly
s
is
(
MCS
A)
h
as
em
er
g
ed
as
a
co
s
t
-
ef
f
ec
tiv
e,
non
-
in
v
asiv
e
tech
n
i
q
u
e
t
h
at
u
tili
ze
s
ex
is
tin
g
cu
r
r
en
t
s
en
s
o
r
s
f
o
r
r
ea
l
-
tim
e
m
o
n
ito
r
in
g
[
1
1
]
,
[
1
2
]
.
W
h
ile
f
ast
Fo
u
r
ier
tr
an
s
f
o
r
m
(
FF
T
)
[
1
3
]
,
[
1
4
]
is
wid
ely
u
s
ed
in
M
C
SA
to
ex
tr
ac
t
f
au
lt
s
ig
n
atu
r
es
f
r
o
m
f
r
eq
u
e
n
cy
s
p
ec
tr
a,
its
lim
itatio
n
s
u
n
d
er
n
o
n
-
s
tatio
n
ar
y
o
p
er
atin
g
c
o
n
d
itio
n
s
h
av
e
p
r
o
m
p
te
d
th
e
u
s
e
o
f
tim
e
-
f
r
eq
u
e
n
c
y
tech
n
iq
u
es
s
u
ch
as
W
ig
n
er
-
Ville
d
is
tr
ib
u
tio
n
s
[
1
5
]
,
[
1
6
]
an
d
wav
elet
tr
a
n
s
f
o
r
m
s
[
1
7
]
.
T
im
e
–
f
r
eq
u
en
cy
tech
n
iq
u
es
ar
e
ex
ten
s
iv
ely
u
s
ed
f
o
r
th
e
an
aly
s
is
o
f
n
o
n
-
s
tatio
n
ar
y
s
ig
n
als
in
f
au
lt
d
ia
g
n
o
s
is
.
Ho
wev
er
,
th
ese
m
eth
o
d
s
s
h
ar
e
co
m
m
o
n
d
r
awb
ac
k
s
,
in
clu
d
in
g
h
ig
h
c
o
m
p
u
tatio
n
al
d
em
a
n
d
s
f
o
r
r
ea
l
-
tim
e
ap
p
licatio
n
s
an
d
in
ter
p
r
etatio
n
c
o
m
p
lex
ity
t
h
at
n
ee
d
s
s
ig
n
if
ican
t e
x
p
e
r
tis
e.
T
h
e
ap
p
licatio
n
o
f
ar
tific
ial
in
tellig
en
ce
(
AI
)
alg
o
r
ith
m
s
in
th
e
elec
tr
ical
m
ac
h
in
e
f
au
lt
d
iag
n
o
s
is
to
p
ic
h
as
em
er
g
e
d
as
a
p
r
o
m
is
in
g
r
esear
ch
d
ir
ec
tio
n
i
n
r
ec
en
t
y
ea
r
s
.
T
h
r
o
u
g
h
le
v
er
ag
in
g
ad
v
a
n
ce
d
tech
n
o
lo
g
ies
s
u
ch
as
d
ee
p
le
ar
n
in
g
(
DL
)
,
p
atter
n
r
ec
o
g
n
itio
n
(
PR
)
,
an
d
m
ac
h
i
n
e
lear
n
i
n
g
(
ML
)
,
AI
-
b
ased
ap
p
r
o
ac
h
es
ar
e
ca
p
ab
le
o
f
p
er
f
o
r
m
in
g
r
ea
l
-
tim
e
f
au
lt
d
etec
t
io
n
an
d
d
eliv
er
i
n
g
ac
cu
r
ate
p
r
ed
ictio
n
s
o
f
m
o
to
r
f
ailu
r
es.
T
h
is
is
ac
h
iev
ed
th
r
o
u
g
h
th
e
au
to
m
atic
lear
n
in
g
an
d
r
ec
o
g
n
itio
n
o
f
d
is
cr
im
in
ativ
e
f
ea
tu
r
es
an
d
p
atter
n
s
f
r
o
m
b
ig
v
o
lu
m
es o
f
s
am
p
led
d
ata.
T
o
o
v
e
r
co
m
e
th
e
lim
itatio
n
s
an
d
d
r
aw
b
ac
k
s
o
f
class
ical
f
au
lt
d
iag
n
o
s
is
ap
p
r
o
ac
h
es,
r
ec
en
t
r
esear
ch
h
as
s
h
if
ted
to
war
d
DL
a
n
d
M
L
-
b
ased
m
eth
o
d
s
.
T
h
ese
in
clu
d
e
th
e
a
r
tific
ial
n
eu
r
al
n
etwo
r
k
s
(
ANN)
[
5
]
,
[
1
8
]
–
[
2
1
]
,
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
e
(
SVM)
[
2
2
]
–
[
2
5
]
,
Dec
is
io
n
T
r
ee
s
(
DT
)
[
2
6
]
,
r
an
d
o
m
f
o
r
ests
(
R
F)
[
2
3
]
–
[
3
1
]
,
k
-
n
ea
r
est
n
eig
h
b
o
r
s
(
k
-
NN)
[
2
3
]
–
[
2
6
]
,
[
3
2
]
,
[
3
3
]
,
co
n
v
o
l
u
tio
n
al
n
eu
r
al
n
etwo
r
k
s
(
C
NN)
[
2
5
]
,
[
3
4
]
,
[
3
5
]
,
Naïv
e
B
ay
es
(
NB
)
[
2
2
]
,
[
2
6
]
,
ANFI
S
[
2
6
]
an
d
d
ee
p
lear
n
in
g
ar
c
h
itectu
r
es
[
2
9
]
,
[
3
6
]
,
[
3
7
]
,
wh
ich
o
f
f
er
p
r
o
m
is
in
g
r
esu
lts
b
y
a
u
to
m
atica
lly
lear
n
in
g
f
ea
tu
r
es
f
r
o
m
r
aw
s
ig
n
als
o
r
e
n
g
in
ee
r
e
d
in
d
icato
r
s
.
Fau
lt
d
iag
n
o
s
is
m
eth
o
d
s
b
ased
o
n
ML
/DL
alg
o
r
ith
m
s
o
f
ten
u
s
e
s
p
ec
if
ic
s
ig
n
als
s
u
ch
as
s
o
u
n
d
s
,
v
ib
r
atio
n
s
,
cu
r
r
en
ts
,
an
d
th
er
m
al
im
a
g
es
[
3
8
]
.
J
ae
n
-
C
u
ellar
et
a
l.
[
1
9
]
s
u
g
g
ested
a
d
iag
n
o
s
is
tech
n
iq
u
e
f
o
r
th
e
d
etec
tio
n
o
f
I
T
SC
b
ased
o
n
t
h
e
p
r
o
ce
s
s
in
g
o
f
m
ag
n
etic
s
tr
ay
-
f
lu
x
,
v
ib
r
ati
o
n
,
an
d
s
tato
r
cu
r
r
en
t
s
ig
n
als.
Stati
s
tical
tim
e
-
d
o
m
ain
f
ea
tu
r
es
ar
e
r
ed
u
ce
d
t
h
r
o
u
g
h
th
e
li
n
ea
r
d
is
cr
im
in
an
t
an
aly
s
is
(
L
DA)
an
d
th
en
class
if
ied
b
y
a
n
eu
r
al
n
etwo
r
k
(
NN)
,
ac
h
iev
in
g
a
n
ac
cu
r
ac
y
o
f
9
9
.
4
%.
Ho
wev
er
,
t
h
e
s
tu
d
y
is
lim
ited
to
o
n
ly
th
r
ee
I
T
SC
s
ev
er
ity
lev
els
(
1
.
4
2
%,
2
.
8
5
%,
an
d
4
.
2
8
%)
an
d
co
n
s
id
er
s
o
n
ly
two
lo
ad
c
o
n
d
itio
n
s
(
1
5
%
an
d
2
5
%
o
f
th
e
n
o
m
in
al
lo
ad
)
,
wh
ich
m
ay
r
estrict
its
ap
p
licab
ilit
y
u
n
d
er
b
r
o
a
d
er
o
p
e
r
atin
g
c
o
n
d
itio
n
s
.
C
h
an
d
et
a
l.
[
2
1
]
d
ev
el
o
p
e
d
a
n
ar
r
o
w
n
e
u
r
al
n
etwo
r
k
(
NNN)
-
b
ased
class
if
icatio
n
m
eth
o
d
u
s
in
g
p
r
in
cip
al
co
m
p
o
n
en
t
an
aly
s
is
(
PC
A)
an
d
tim
e
d
o
m
ain
f
ea
tu
r
es.
Desp
ite
ad
d
r
ess
in
g
f
o
u
r
f
a
u
lt
s
ev
er
ity
lev
els
(
5
.
7
7
%,
6
.
8
5
%,
8
.
4
2
%,
an
d
1
0
.
9
2
%)
u
n
d
er
t
h
r
ee
lo
a
d
co
n
d
itio
n
s
(
n
o
-
lo
ad
,
2
5
%
lo
a
d
,
an
d
4
0
%
lo
a
d
)
,
th
e
ac
cu
r
ac
y
is
r
elativ
ely
lim
ited
,
r
ea
ch
i
n
g
9
8
.
2
3
%.
I
n
[
2
5
]
,
th
e
AI
-
b
ased
to
o
ls
ar
e
u
s
ed
f
o
r
I
T
SC
f
au
lt
lev
el
class
if
icat
io
n
u
s
in
g
SVM
an
d
C
NN
m
o
d
els.
R
esu
lt
s
s
h
o
w
th
at
b
o
th
m
o
d
els
ca
n
r
ea
lize
an
ac
cu
r
ac
y
o
f
9
9
%.
Ho
wev
e
r
,
th
e
s
tu
d
y
is
lim
ited
to
o
n
ly
t
h
r
ee
I
T
SC
s
ev
er
ity
lev
els
(
5
%,
1
0
%,
an
d
1
5
%)
an
d
co
n
s
id
er
s
o
n
ly
two
lo
ad
c
o
n
d
itio
n
s
(
n
o
-
lo
ad
a
n
d
a
n
o
m
in
al
to
r
q
u
e
o
f
0
.
0
5
Nm
)
.
I
n
[
2
6
]
th
e
au
th
o
r
s
p
r
o
p
o
s
ed
a
m
eth
o
d
b
ased
o
n
ANFI
S a
n
d
ANN
m
o
d
els to
d
iag
n
o
s
e
I
T
SC
in
p
u
m
p
in
g
s
y
s
tem
s
.
W
h
ile
th
ese
m
o
d
els
ac
h
iev
e
s
atis
f
ac
to
r
y
ac
cu
r
ac
y
lev
els,
9
9
.
6
%
f
o
r
ANN
an
d
9
4
.
6
%
f
o
r
ANFI
S,
th
ey
lack
test
in
g
o
n
r
ea
l
in
s
tallatio
n
s
to
d
em
o
n
s
tr
ate
r
o
b
u
s
tn
ess
u
n
d
er
o
p
er
atio
n
al
v
ar
iab
ilit
ies.
I
n
ad
d
itio
n
,
th
is
r
esear
ch
d
id
n
o
t
s
tu
d
y
th
e
m
o
d
el’
s
ab
ilit
y
to
p
r
ed
ict
v
er
y
in
cip
ien
t
f
au
lts
.
An
I
T
SC
f
au
lt
d
iag
n
o
s
is
tech
n
iq
u
e
f
o
r
a
p
er
m
a
n
en
t
m
ag
n
et
s
y
n
ch
r
o
n
o
u
s
m
o
to
r
(
PMSM)
was
d
ev
elo
p
ed
in
[
2
7
]
.
T
h
is
tech
n
iq
u
e
em
p
lo
y
s
a
s
p
ar
s
e
r
ep
r
esen
tati
o
n
f
o
r
ex
t
r
ac
tin
g
f
e
atu
r
es
f
r
o
m
v
ib
r
atio
n
an
d
cu
r
r
en
t
s
ig
n
a
ls
.
T
h
e
f
ea
tu
r
es
ar
e
th
en
u
tili
ze
d
as
in
p
u
ts
to
an
SVM
f
o
r
f
au
lt
d
iag
n
o
s
tics
.
Alth
o
u
g
h
th
is
tech
n
iq
u
e
ex
h
ib
its
h
ig
h
ac
cu
r
ac
y
,
it
o
v
er
lo
o
k
s
m
o
to
r
lo
ad
v
a
r
ia
tio
n
s
an
d
th
e
d
etec
tio
n
o
f
d
if
f
er
en
t
s
h
o
r
t
-
cir
c
u
it
f
au
lt
s
ev
er
ities
(
b
in
ar
y
class
if
icatio
n
:
h
ea
lth
y
an
d
f
a
u
lty
)
.
Das
et
a
l.
[
2
8
]
ap
p
lied
c
o
n
tin
u
o
u
s
wav
elet
tr
a
n
s
f
o
r
m
(
C
W
T
)
f
o
r
f
ea
t
u
r
e
ex
tr
ac
tio
n
an
d
a
SVM
f
o
r
cl
ass
if
icatio
n
.
Alth
o
u
g
h
th
is
s
tu
d
y
ad
d
r
ess
es
n
in
e
m
in
o
r
I
T
SC
f
au
lt
class
e
s
,
it
o
v
er
lo
o
k
s
m
o
to
r
lo
ad
c
o
n
d
itio
n
s
,
an
d
th
e
ac
h
iev
ed
ac
cu
r
ac
y
is
clo
s
e
to
9
0
%.
I
n
[
3
0
]
,
a
R
F
m
eth
o
d
is
u
s
ed
to
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
n
ew d
ia
g
n
o
s
tic
meth
o
d
b
a
s
ed
o
n
s
u
p
p
o
r
t v
ec
to
r
ma
ch
in
e
fo
r
…
(
Hich
a
m
Za
imen
)
1737
d
iag
n
o
s
e
th
e
s
h
o
r
te
d
in
ter
-
tu
r
n
f
ailu
r
es
i
n
I
PMSMs.
Desp
ite
ac
h
iev
in
g
s
atis
f
ac
to
r
y
clas
s
if
icatio
n
ac
cu
r
ac
y
an
d
in
co
r
p
o
r
atin
g
v
ar
y
in
g
m
o
to
r
lo
a
d
co
n
d
itio
n
s
,
th
is
ap
p
r
o
ac
h
is
co
n
s
tr
ain
ed
b
y
its
lim
ited
f
au
lt
r
ep
r
esen
tatio
n
,
as
it
co
n
s
id
er
s
o
n
ly
th
r
ee
I
T
SC
s
ev
er
ity
lev
els.
I
n
ad
d
itio
n
,
th
e
n
eg
lect
o
f
lo
w
-
s
ev
er
ity
f
au
lt
co
n
d
itio
n
s
p
r
ev
en
ts
th
e
m
o
d
el
f
r
o
m
ca
p
tu
r
in
g
i
n
cip
ien
t
f
a
u
lts
at
an
ea
r
ly
s
tag
e.
I
n
[
3
2
]
,
a
m
u
lti
-
alg
o
r
ith
m
AI
-
b
ased
ap
p
r
o
ac
h
is
p
r
o
p
o
s
ed
f
o
r
th
e
d
etec
tio
n
o
f
I
T
SC
f
au
lts
.
T
h
is
ap
p
r
o
ac
h
r
elies
o
n
MA
T
L
AB
-
b
ased
s
im
u
latio
n
s
with
o
u
t
ex
p
er
im
en
tal
v
alid
atio
n
,
lim
itin
g
its
p
r
ac
tical
ap
p
licab
ilit
y
.
A
2
D
C
NN
-
b
ased
d
ee
p
lear
n
in
g
to
o
l
is
d
ev
elo
p
e
d
in
[
3
4
]
to
d
ia
g
n
o
s
e
th
e
I
T
SC
f
au
lts
an
d
f
in
d
th
eir
s
ev
er
ity
.
T
h
e
r
esu
lts
s
h
o
w
th
at
th
is
d
iag
n
o
s
tic
to
o
l
is
ca
p
ab
le
o
f
d
etec
tin
g
th
e
f
au
lt
u
n
d
er
d
if
f
er
en
t
lo
a
d
in
g
c
o
n
d
itio
n
s
with
an
ac
cu
r
ac
y
o
f
9
7
.
7
5
%
f
o
r
all
d
ef
in
ed
f
a
u
lt
lev
els.
I
n
[
3
9
]
,
ML
alg
o
r
ith
m
s
ar
e
u
s
ed
f
o
r
I
T
SC
f
au
lt
d
etec
tio
n
in
PMSMs.
Ho
wev
er
,
t
h
e
r
esear
c
h
r
elies
o
n
a
s
im
u
latio
n
d
ataset,
co
n
s
id
er
s
o
n
ly
h
ea
lth
y
a
n
d
5
%
f
a
u
lt
co
n
d
itio
n
s
,
a
n
d
u
s
es
a
lim
ited
f
ea
tu
r
e
s
et
(
R
MS
an
d
n
eg
ativ
e
s
eq
u
en
ce
c
u
r
r
en
t)
,
wh
ic
h
m
ay
r
estrict
th
e
r
o
b
u
s
tn
ess
an
d
g
en
er
aliza
tio
n
o
f
th
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
.
As
h
ig
h
lig
h
ted
ea
r
lier
,
n
u
m
e
r
o
u
s
s
tu
d
ies
h
av
e
ap
p
lied
AI
-
b
ased
tech
n
iq
u
es
f
o
r
d
iag
n
o
s
in
g
f
au
lts
in
elec
tr
ical
m
ac
h
in
es.
W
h
ile
s
o
m
e
o
f
th
ese
ap
p
r
o
ac
h
es
r
ely
o
n
m
an
u
al
f
ea
tu
r
e
ex
tr
ac
tio
n
,
o
th
er
s
ad
o
p
t
d
ee
p
lear
n
in
g
–
b
ased
au
to
m
ated
f
e
atu
r
e
lear
n
in
g
.
Ho
we
v
er
,
r
elativ
ely
f
ew
s
tu
d
ies
h
a
v
e
co
n
s
id
er
ed
I
T
SC
f
au
lt
d
iag
n
o
s
is
u
n
d
er
v
ar
y
in
g
lo
a
d
co
n
d
itio
n
s
,
w
h
ich
r
e
p
r
esen
t
a
cr
u
cial
f
ac
to
r
in
r
ea
lis
tic
o
p
er
atin
g
en
v
i
r
o
n
m
e
n
ts
.
Fu
r
th
er
m
o
r
e
,
m
u
ch
o
f
th
e
p
r
io
r
r
esear
ch
h
as
ad
d
r
ess
ed
a
lim
ited
n
u
m
b
er
o
f
f
au
lt
s
ev
er
ity
lev
els,
wh
ich
co
n
s
tr
ain
s
th
eir
ab
ilit
y
to
r
eli
ab
ly
d
is
cr
im
in
ate
am
o
n
g
d
if
f
er
en
t
lev
els
o
f
I
T
SC
.
Ad
d
itio
n
ally
,
th
e
b
o
d
y
o
f
r
esear
ch
d
ed
icate
d
to
th
e
d
iag
n
o
s
is
o
f
s
h
o
r
t
-
cir
cu
it
f
au
lts
is
r
elativ
ely
lim
ited
wh
en
co
m
p
ar
ed
to
th
e
s
tu
d
ies
ad
d
r
ess
in
g
b
ea
r
in
g
,
o
p
en
cir
cu
it,
b
r
o
k
en
b
a
r
,
an
d
m
u
ltip
l
e
f
au
lts
.
Mo
r
eo
v
e
r
,
it
s
h
o
u
l
d
b
e
n
o
ted
th
at
th
e
p
r
o
p
o
s
ed
m
eth
o
d
f
o
cu
s
es
o
n
lo
w
-
s
ev
er
ity
I
T
SC
f
au
lts
(
1
%
–
5
%)
in
co
n
tr
ast
to
m
o
s
t
p
r
io
r
wo
r
k
s
th
at
m
ain
ly
co
n
s
id
er
m
o
d
er
ate
-
o
r
s
er
io
u
s
-
s
ev
er
ity
I
T
SC
f
au
lts
(
e.
g
.
,
g
r
ea
ter
th
an
1
0
%),
wh
ich
ar
e
r
elativ
ely
ea
s
y
t
o
id
en
tify
d
u
e
t
o
th
eir
p
r
o
n
o
u
n
ce
d
s
ig
n
atu
r
es.
T
o
b
r
id
g
e
s
o
m
e
o
f
th
ese
g
ap
s
,
th
is
p
ap
er
p
r
o
p
o
s
es
a
m
ac
h
in
e
lear
n
in
g
-
b
ased
d
iag
n
o
s
tic
m
eth
o
d
th
at
co
m
b
i
n
es
a
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
e
with
Fis
h
er
’
s
r
atio
-
b
ased
f
ea
tu
r
e
s
elec
tio
n
f
o
r
I
T
SC
f
au
lt
d
etec
tio
n
an
d
s
ev
er
it
y
class
if
icatio
n
.
Hen
ce
,
th
e
s
cien
tific
c
o
n
tr
i
b
u
tio
n
o
f
th
is
p
ap
e
r
r
esid
es
in
th
e
f
o
ll
o
win
g
p
o
in
t
s
:
T
h
e
ML
-
b
ased
d
iag
n
o
s
tic
m
eth
o
d
a
d
d
r
ess
ed
f
i
v
e
m
in
o
r
-
s
ev
er
ity
I
T
SC
f
au
lts
(
1
%
–
5
%)
u
n
d
er
f
o
u
r
lo
ad
-
lev
e
l c
o
n
d
itio
n
s
o
f
in
d
u
ctio
n
m
o
to
r
s
.
−
A
s
et
o
f
f
ea
t
u
r
es
is
co
m
p
u
te
d
u
s
in
g
th
e
tim
e
-
d
o
m
ain
f
ea
tu
r
e
ex
tr
ac
tio
n
f
r
o
m
th
e
s
tat
o
r
cu
r
r
en
ts
an
d
ap
p
ly
in
g
Fis
h
er
’
s
r
atio
(
FR
)
m
eth
o
d
to
f
in
d
th
e
m
o
s
t d
is
cr
im
in
ativ
e
f
ea
tu
r
es f
o
r
SVM
class
if
icatio
n
task
s
.
−
T
h
e
d
ev
el
o
p
ed
m
eth
o
d
is
p
e
r
f
o
r
m
e
d
o
f
f
lin
e,
wh
e
r
ein
th
e
cu
r
r
en
t
s
ig
n
als
f
r
o
m
th
e
I
M
u
n
d
e
r
test
ar
e
co
llected
ex
p
er
im
en
tally
.
Ou
t
co
m
es
d
em
o
n
s
tr
ate
th
at
th
e
p
r
o
p
o
s
ed
h
y
b
r
id
FR
-
SVM
-
b
ased
m
eth
o
d
is
ab
le
to
d
etec
t th
e
lo
w
-
s
ev
er
ity
I
T
SC
f
au
lts
with
h
ig
h
ac
cu
r
ac
y
(
r
an
g
in
g
f
r
o
m
9
9
.
5
4
% to
1
0
0
%)
.
T
h
e
s
u
b
s
eq
u
en
t
s
ec
tio
n
s
o
f
th
is
p
ap
er
ar
e
ar
r
an
g
ed
as
f
o
llo
ws:
s
ec
tio
n
2
p
r
esen
ts
th
e
ex
p
er
im
en
tal
s
etu
p
.
Sectio
n
3
is
d
iv
id
ed
in
to
two
p
ar
ts
;
th
e
f
ir
s
t
in
tr
o
d
u
ce
s
th
e
f
ea
tu
r
e
ex
tr
ac
tio
n
m
eth
o
d
b
ased
o
n
tim
e
-
d
o
m
ain
in
d
icato
r
s
,
an
d
th
e
s
e
co
n
d
f
o
cu
s
es
o
n
th
e
p
r
o
p
o
s
ed
d
iag
n
o
s
is
m
eth
o
d
b
ased
o
n
SVM
an
d
FR
.
T
h
e
r
esu
lts
an
d
th
eir
an
aly
s
es a
r
e
p
r
o
v
id
ed
in
s
ec
tio
n
5
.
T
h
e
c
o
n
c
lu
s
io
n
an
d
f
u
tu
r
e
t
o
p
ic
ar
e
g
iv
en
in
s
ec
tio
n
6
.
2.
T
H
E
E
XP
E
R
I
M
E
N
T
A
L
L
A
B
O
RATOR
Y
SE
T
UP
2
.
1
.
E
x
perim
ent
a
l set
up
des
cr
ipt
io
n
T
h
e
ex
p
er
im
e
n
tal
s
etu
p
p
latf
o
r
m
s
ch
em
atic
is
s
h
o
wn
in
Fig
u
r
e
1
.
I
t
c
o
n
s
is
ts
o
f
a
s
tar
-
co
n
n
ec
ted
3
k
W
s
q
u
ir
r
el
ca
g
e
in
d
u
ctio
n
m
o
to
r
,
p
o
wer
ed
d
ir
ec
tly
f
r
o
m
a
t
h
r
ee
-
p
h
ase
s
u
p
p
ly
an
d
co
u
p
le
d
to
a
v
ar
ia
b
le
lo
ad
.
T
h
e
m
o
to
r
d
r
iv
es
a
5
k
VA
s
elf
-
ex
cited
s
y
n
c
h
r
o
n
o
u
s
g
en
er
at
o
r
,
wh
ich
s
u
p
p
lies
a
th
r
ee
-
p
h
a
s
e
r
esis
tiv
e
lo
ad
o
f
1
0
×
0
.
5
k
W
.
Data
ac
q
u
is
itio
n
is
p
er
f
o
r
m
e
d
u
s
in
g
Flu
k
e
i2
0
0
cu
r
r
e
n
t
clam
p
s
co
n
n
ec
ted
to
a
Natio
n
al
I
n
s
tr
u
m
en
ts
d
ata
ac
q
u
is
itio
n
s
y
s
tem
,
NI
c
-
DAQ
9
1
7
4
,
with
NI
9
2
2
5
m
o
d
u
les.
Flu
k
e
cu
r
r
en
t
p
r
o
b
es
with
a
b
an
d
wid
th
o
f
4
0
k
Hz
wer
e
u
s
ed
to
r
ec
o
r
d
th
e
c
u
r
r
en
t
s
ig
n
als.
T
h
e
cu
r
r
e
n
t
p
r
o
b
es
wer
e
d
ir
ec
tly
tr
an
s
f
er
r
ed
t
o
th
e
co
m
p
u
ter
v
ia
th
e
Natio
n
al
I
n
s
tr
u
m
en
t N
I
9
2
2
5
m
o
d
u
le
[
4
0
]
.
Fig
u
r
e
1
.
Sch
em
atic
o
f
ex
p
er
i
m
en
tal
s
etu
p
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
.
4
,
Au
g
u
s
t
20
2
6
:
1
7
3
5
-
1
754
1738
Fau
lt
d
iag
n
o
s
is
was
p
er
f
o
r
m
e
d
with
th
e
th
r
ee
-
p
h
ase
cu
r
r
en
t
s
,
wh
ich
wer
e
ac
q
u
ir
ed
an
d
r
ec
o
r
d
ed
at
2
5
0
0
0
s
am
p
les
p
er
s
ec
o
n
d
.
T
h
e
in
ter
-
tu
r
n
s
h
o
r
t
-
ci
r
cu
it
(
S
C
)
is
cr
ea
ted
b
y
r
ewin
d
in
g
t
h
e
s
tato
r
win
d
i
n
g
as
s
h
o
wn
in
Fig
u
r
e
2
(
a)
.
T
h
e
th
r
ee
AB
C
p
h
ases
o
f
th
e
s
ta
to
r
wer
e
r
ewo
u
n
d
in
o
r
d
er
to
b
e
a
b
le
to
o
p
er
ate
u
n
d
e
r
f
au
lt
co
n
d
itio
n
s
.
T
h
e
r
ewo
u
n
d
s
tato
r
is
also
s
h
o
wn
i
n
Fig
u
r
e
2
(
b
)
.
I
n
a
d
d
itio
n
,
th
e
s
ev
er
ity
o
f
th
e
SC
f
au
lts
(
μ
cc
)
is
in
d
icate
d
as
a
p
er
ce
n
t
ag
e.
T
h
e
p
e
r
ce
n
tag
e
o
f
SC
f
ailu
r
e
is
co
n
s
id
er
ed
to
b
e
th
e
r
atio
b
etwe
en
th
e
n
u
m
b
er
o
f
s
h
o
r
t
-
cir
cu
ited
tu
r
n
s
(
n
cc
)
an
d
t
h
e
to
tal
n
u
m
b
er
o
f
tu
r
n
s
in
ea
ch
win
d
in
g
(
n
s
):
=
.
100
(
1
)
T
h
e
s
ev
er
ity
o
f
th
e
I
T
SC
was
d
ef
in
ed
as
%1
,
%2
,
%3
,
%4
,
an
d
5
%.
E
ac
h
p
h
ase
o
f
th
e
m
o
to
r
h
as
a
to
tal
o
f
1
0
0
tu
r
n
s
.
T
h
e
m
o
to
r
with
a
h
ea
lth
y
a
n
d
s
h
o
r
ted
-
t
u
r
n
s
tato
r
win
d
in
g
was
test
ed
with
lo
ad
s
o
f
2
5
%,
5
0
%,
7
5
%,
an
d
1
0
0
%
o
f
th
e
f
u
ll
lo
ad
.
T
h
e
I
T
SC
s
wer
e
in
tr
o
d
u
ce
d
in
p
h
ase
A,
wh
ile
th
e
o
th
er
p
h
ases
r
em
ain
ed
in
a
h
ea
lth
y
co
n
d
itio
n
.
(
a)
(
b
)
Fig
u
r
e
2
.
Stato
r
win
d
in
g
s
h
o
r
t
-
cir
cu
its
f
au
lts
(
a)
I
T
SC
f
au
lt l
ev
els an
d
(
b
)
d
ef
ec
tiv
e
s
tato
r
win
d
in
g
s
2
.
2
.
E
f
f
ec
t
o
f
I
T
SC f
a
ults o
n sta
t
o
r
t
hree
-
ph
a
s
e
curr
ent
s
Fig
u
r
e
3
s
h
o
ws
th
e
ex
p
er
im
e
n
tally
r
ec
o
r
d
e
d
th
r
ee
-
p
h
ase
s
tato
r
cu
r
r
en
ts
f
o
r
th
e
h
ea
lth
y
ca
s
e
an
d
d
if
f
er
en
t
I
T
SC
f
au
lt
s
ev
er
ities
th
at
o
cc
u
r
r
ed
in
p
h
ase
A
.
Fro
m
Fig
u
r
es
3
(
a)
-
3
(
f
)
,
it
is
ev
id
en
t
th
at
th
e
im
b
alan
ce
o
f
wav
ef
o
r
m
s
is
m
o
r
e
m
a
r
k
ed
with
th
e
m
o
r
e
s
ev
er
e
I
T
SC
f
au
lt.
Fu
r
th
er
m
o
r
e,
th
e
am
p
litu
d
e
o
f
p
h
ase
-
A
cu
r
r
e
n
t
in
cr
ea
s
es
with
th
e
in
c
r
ea
s
e
in
I
T
SC
f
au
lt
s
e
v
er
ity
.
T
h
er
e
is
a
s
ig
n
if
ican
t
im
p
ac
t
o
n
th
e
p
h
ase
-
A
cu
r
r
en
t,
with
m
in
o
r
e
f
f
ec
ts
o
n
o
th
e
r
p
h
ases
.
Fig
u
r
e
3
.
Me
asu
r
e
d
s
tato
r
cu
r
r
en
t w
av
ef
o
r
m
s
(
ca
s
e
o
f
2
5
% o
f
f
u
ll lo
ad
an
d
2
5
0
0
s
am
p
les)
(
a)
Hea
lth
y
s
tate
,
(
b
)
1
% SC
,
(
c)
2
% SC
,
(
d
)
3
% SC
,
(
e)
4
% SC
,
an
d
(
f
)
5
% SC
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
n
ew d
ia
g
n
o
s
tic
meth
o
d
b
a
s
ed
o
n
s
u
p
p
o
r
t v
ec
to
r
ma
ch
in
e
fo
r
…
(
Hich
a
m
Za
imen
)
1739
3.
F
E
AT
U
RE
S E
X
T
RAC
T
I
O
N
Featu
r
e
ex
tr
ac
tio
n
in
th
e
ML
to
p
ic
is
th
e
p
r
o
ce
s
s
o
f
s
witch
in
g
r
aw
d
ata
in
to
a
s
et
o
f
f
ea
tu
r
es
o
r
in
p
u
ts
th
at
ML
alg
o
r
ith
m
s
m
ay
u
s
e.
T
h
is
s
tep
is
v
ital
to
im
p
r
o
v
e
b
o
th
t
h
e
ef
f
icien
c
y
an
d
th
e
p
r
ed
ictiv
e
ac
cu
r
ac
y
o
f
m
ac
h
i
n
e
lear
n
in
g
m
o
d
els.
Featu
r
e
ex
tr
ac
tio
n
will
r
ed
u
ce
th
e
m
ass
iv
e
tim
e
s
er
ies
d
ata
p
o
in
ts
in
to
a
m
an
ag
ea
b
le
s
y
n
o
p
tic
d
ata
s
tr
u
ctu
r
e
wh
ile
co
n
s
er
v
in
g
m
o
s
t
o
f
th
e
ch
a
r
ac
ter
is
tics
o
f
th
e
tim
e
s
er
ies
[
2
2
]
.
Fo
r
co
n
d
itio
n
m
o
n
ito
r
i
n
g
an
d
class
if
icatio
n
o
f
th
e
I
T
SC
f
au
lts
,
a
s
et
o
f
tim
e
-
d
o
m
ain
f
ea
tu
r
es
(
T
DFs
)
ar
e
u
s
ed
in
th
e
p
r
o
p
o
s
ed
ML
alg
o
r
ith
m
.
T
h
ese
f
ea
tu
r
es
allo
w
ca
p
tu
r
in
g
th
e
s
h
a
p
e,
s
p
r
ea
d
,
an
d
d
is
tr
ib
u
tio
n
o
f
th
e
s
ig
n
al’
s
v
alu
es.
T
h
e
tim
e
-
b
ased
f
ea
tu
r
es
ex
tr
ac
ted
a
r
e
th
e
r
o
o
t
m
ea
n
s
q
u
ar
e
(
R
MS
)
,
m
ea
n
,
v
ar
ian
ce
(
V
a
r
)
,
p
ea
k
-
to
-
p
ea
k
,
k
u
r
to
s
is
,
s
k
ewn
ess
,
s
tan
d
ar
d
d
ev
iatio
n
(
Std
)
,
m
ax
im
u
m
,
a
n
d
m
i
n
im
u
m
.
Fo
r
a
d
is
cr
ete
s
ig
n
al
(
)
,
∈
[
1
,
]
,
th
e
f
o
llo
win
g
9
f
ea
tu
r
es
ar
e
u
s
ed
in
th
e
SVM
-
b
ased
m
ac
h
in
e
lear
n
in
g
m
o
d
el:
(
)
=
√
1
∑
(
)
2
=
1
(
2
)
(
)
=
1
∑
(
)
=
1
(
3
)
(
)
=
1
∑
(
(
)
−
̄
)
=
1
2
(
4
)
(
)
=
1
∑
(
(
)
−
̄
)
=
1
3
[
1
∑
(
(
)
−
̄
)
=
1
2
]
3
/
2
(
5
)
(
)
=
1
∑
(
(
)
−
̄
)
=
1
4
4
(
6
)
2
(
)
=
[
(
)
]
−
[
(
)
]
(
7
)
(
)
=
√
1
∑
[
(
)
−
̄
]
2
=
1
(
8
)
(
)
=
[
(
)
]
,
(
)
=
[
(
)
]
(
9
)
T
h
e
T
DFs
ar
e
s
ep
ar
ately
ap
p
lied
to
th
e
s
tato
r
p
h
ase
cu
r
r
en
ts
.
E
ac
h
ac
q
u
ir
e
d
cu
r
r
en
t
s
ig
n
al
i
x
(
x
=
a
,
b
,
c
)
co
m
p
o
s
ed
o
f
s
am
p
les is
s
eg
m
en
ted
in
to
eq
u
al
p
a
r
ts
(
N
s
eg
m
en
ts
)
o
f
a
len
g
t
h
l
,
as (
1
0
)
s
h
o
ws:
=
[
1
:
⏟
1
,
+
1
:
2
⏟
2
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1
+
2
−
.
:
⏟
]
(
1
0
)
Fig
u
r
e
4
s
h
o
ws th
e
p
r
o
ce
s
s
o
f
r
aw
s
ig
n
al
s
eg
m
en
tatio
n
f
o
r
f
ea
tu
r
e
ex
tr
ac
tio
n
.
E
ac
h
s
ig
n
al
was seg
m
en
ted
in
to
500
-
s
am
p
le
win
d
o
ws.
Fig
u
r
e
4
.
Featu
r
e
ex
tr
ac
tio
n
f
r
o
m
th
r
ee
-
p
h
ase
r
aw
s
ig
n
al
s
eg
m
en
tatio
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
.
4
,
Au
g
u
s
t
20
2
6
:
1
7
3
5
-
1
754
1740
4.
DIAG
NO
SI
S M
E
T
H
O
D
4
.
1
.
M
et
ho
d ste
ps
T
h
e
d
iag
r
am
o
f
th
e
p
r
o
p
o
s
ed
f
au
lt
d
iag
n
o
s
tic
m
eth
o
d
f
o
r
I
T
SC
f
au
lts
is
d
ep
icted
in
Fig
u
r
e
5
.
All
s
tep
s
ar
e
illu
s
tr
ated
as f
o
llo
ws:
Step
1
:
Data
p
r
o
ce
s
s
in
g
:
T
h
e
h
ea
lth
y
s
tate
an
d
I
T
SC
f
au
lt
m
o
d
es
ar
e
d
ef
in
e
d
.
T
h
e
f
a
u
lt
m
o
n
ito
r
in
g
o
f
th
r
ee
p
h
ases
’
cu
r
r
en
ts
u
n
d
er
h
ea
lth
y
co
n
d
itio
n
s
an
d
d
if
f
er
en
t I
T
SC
lev
els
is
ac
q
u
ir
ed
an
d
s
am
p
le
d
.
Step
2
:
T
h
e
f
ea
tu
r
e
v
ec
to
r
is
e
x
tr
ac
ted
f
r
o
m
th
e
m
o
n
ito
r
ed
c
u
r
r
en
t
s
ig
n
als
b
ased
o
n
T
DFs
.
T
h
ese
f
ea
tu
r
es
a
r
e
th
en
n
o
r
m
alize
d
.
Step
3
:
T
h
e
Fis
h
er
’
s
r
atio
a
lg
o
r
ith
m
is
u
s
ed
to
s
elec
t
th
e
1
0
b
etter
an
d
m
o
r
e
in
f
lu
en
tial
f
e
atu
r
es.
T
h
is
p
r
o
ce
s
s
allo
ws
d
ec
r
ea
s
in
g
th
e
ca
lcu
latio
n
co
s
ts
an
d
im
p
r
o
v
i
n
g
t
h
e
ef
f
icien
cy
o
f
th
e
m
ac
h
i
n
e
lear
n
in
g
class
if
ier
.
Step
4
: T
h
e
s
u
p
p
o
r
t v
ec
t
o
r
m
a
ch
in
e
alg
o
r
ith
m
is
u
s
ed
as a
cl
ass
if
ier
to
ac
h
iev
e
th
e
I
T
SC
f
a
u
lt c
lass
if
icatio
n
.
Fig
u
r
e
5
.
Diag
r
a
m
o
f
th
e
p
r
o
p
o
s
ed
f
au
lt d
iag
n
o
s
is
m
eth
o
d
u
s
in
g
th
e
FR
–
SVM
alg
o
r
ith
m
T
o
m
itig
ate
th
e
is
s
u
e
o
f
th
e
s
ig
n
if
ican
t
v
a
r
iatio
n
in
th
e
s
ca
le
o
f
s
am
p
le
f
ea
tu
r
e
v
alu
es,
we
n
ee
d
t
o
n
o
r
m
alize
th
ese
f
ea
tu
r
es.
T
h
e
n
o
r
m
aliza
tio
n
p
r
o
ce
s
s
m
ak
es
f
ea
tu
r
es
co
m
p
a
r
ab
le,
h
el
p
s
alg
o
r
ith
m
s
co
n
v
er
g
e
f
aster
,
an
d
o
f
ten
im
p
r
o
v
es
M
L
m
o
d
el
ac
cu
r
ac
y
.
I
n
th
is
r
esear
ch
,
th
e
f
ea
tu
r
es
ar
e
n
o
r
m
al
ized
u
s
in
g
th
e
m
in
-
m
ax
s
ca
lin
g
m
eth
o
d
,
wh
ich
is
d
ef
in
ed
as
=
−
(
1
1
)
W
h
er
e:
is
th
e
f
ea
tu
r
e
n
u
m
b
e
r
i
(
=
1
:
27
)
,
is
th
e
n
o
r
m
alize
d
f
ea
tu
r
e,
an
d
th
e
v
alu
e
r
a
n
g
e
is
b
etwe
en
0
an
d
1
.
an
d
ar
e
th
e
m
in
im
u
m
an
d
m
ax
im
u
m
v
al
u
es in
th
e
s
am
p
les,
r
esp
ec
tiv
ely
.
T
h
e
lab
els ass
o
ciate
d
with
s
tat
o
r
s
h
o
r
t c
ir
c
u
it f
au
lt c
lass
es a
r
e
lis
ted
in
T
ab
le
1
.
T
ab
le
1
.
L
a
b
els o
f
s
tato
r
s
h
o
r
t
cir
cu
it f
au
lt c
lass
es
S
t
a
t
e
D
e
scri
p
t
i
o
n
La
b
e
l
C
l
a
s
s
H
e
a
l
t
h
y
N
o
f
a
u
l
t
i
n
st
a
t
o
r
w
i
n
d
i
n
g
H
e
a
l
t
h
y
1
I
TSC
1
%
0
1
t
u
r
n
sh
o
r
t
e
d
i
n
p
h
a
se
a
S
C
1
%
2
I
TSC
2
%
0
2
t
u
r
n
sh
o
r
t
e
d
i
n
p
h
a
se
a
S
C
2
%
3
I
TSC
3
%
0
3
t
u
r
n
sh
o
r
t
e
d
i
n
p
h
a
se
a
S
C
3
%
4
I
TSC
4
%
0
4
t
u
r
n
sh
o
r
t
e
d
i
n
p
h
a
se
a
S
C
4
%
5
I
TSC
5
%
0
5
t
u
r
n
sh
o
r
t
e
d
i
n
p
h
a
se
a
S
C
5
%
6
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
n
ew d
ia
g
n
o
s
tic
meth
o
d
b
a
s
ed
o
n
s
u
p
p
o
r
t v
ec
to
r
ma
ch
in
e
fo
r
…
(
Hich
a
m
Za
imen
)
1741
4
.
2
.
Su
pp
o
rt
v
ec
t
o
r
ma
chine
princip
le
I
n
f
au
lt
class
if
icatio
n
r
ec
o
g
n
it
io
n
,
th
e
class
if
ier
s
’
r
o
le
is
to
d
eter
m
in
e
th
e
ty
p
e
o
f
f
a
u
lt
to
wh
ich
th
e
test
s
am
p
le
b
elo
n
g
s
b
ased
o
n
well
-
lab
eled
tr
ai
n
in
g
d
ata
wi
th
d
if
f
e
r
en
t
f
au
lt
ty
p
es
[
3
0
]
.
T
h
e
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
e
is
a
s
u
cc
ess
f
u
l
s
u
p
e
r
v
is
ed
m
ac
h
in
e
-
lear
n
in
g
alg
o
r
ith
m
th
at
co
n
v
er
ts
th
e
o
r
i
g
in
al
d
ata
with
a
lo
w
d
im
en
s
io
n
in
to
a
h
ig
h
-
d
im
en
s
i
o
n
al
f
ea
tu
r
e
s
p
ac
e.
SVM
aim
s
to
f
in
d
an
o
p
tim
al
h
y
p
er
p
la
n
e
th
at
ef
f
ec
tiv
ely
s
ep
ar
ates
d
if
f
er
en
t
class
es
o
f
d
ata
wh
ile
m
ax
im
izin
g
th
e
m
ar
g
in
,
wh
ic
h
is
d
ef
in
e
d
as
th
e
d
is
tan
ce
b
etwe
en
th
e
h
y
p
er
p
la
n
e
an
d
th
e
clo
s
est
d
ata
p
o
in
ts
f
r
o
m
ea
c
h
class
in
Fig
u
r
e
6
.
Fig
u
r
e
6
.
C
lass
if
icatio
n
u
s
in
g
SVM
(
2
class
es),
h
y
p
er
p
lan
e
i
n
2
D
s
p
ac
e
Giv
en
a
d
ataset
with
n
ex
am
p
les (
x
j
,
y
j
)
,
wh
er
e
ea
ch
x
j
is
an
o
b
s
er
v
atio
n
an
d
y
j
its
as
s
o
ciat
ed
d
ec
is
io
n
b
elo
n
g
s
to
[
-
1
,
1
]
.
B
y
u
s
in
g
a
n
o
n
lin
ea
r
m
a
p
p
in
g
Ф
(
x)
,
th
e
o
r
ig
in
al
d
ata
is
m
ap
p
ed
in
to
a
n
ew
f
ea
tu
r
e
s
p
ac
e
in
wh
ich
th
e
d
ata
ar
e
s
p
ar
s
e
an
d
p
o
s
s
ib
ly
m
o
r
e
s
ep
ar
ab
le,
a
n
d
th
e
m
ax
im
u
m
m
ar
g
in
s
ep
ar
atin
g
h
y
p
e
r
p
lan
e
ω
.Ф(
x)
+b
0
will
b
e
b
u
ilt.
Her
e
b
0
is
an
o
f
f
s
et
ter
m
[
4
1
]
.
I
n
o
r
d
er
to
f
i
n
d
th
e
o
p
tim
al
h
y
p
er
p
lan
e,
we
n
ee
d
to
m
in
im
ize
th
e
f
o
llo
win
g
f
u
n
ctio
n
al
F
cn
[
4
2
]
.
,
0
,
{
(
)
=
0
.
5
.
‖
‖
2
+
∑
=
1
}
(
1
2
)
Su
b
ject
to
:
(
.
+
0
)
≥
1
−
,
≥
0
,
∀
=
1
,
2
,
.
.
.
.
(
1
3
)
wh
er
e
w
is
th
e
weig
h
t
v
ec
to
r
o
f
th
e
h
y
p
er
p
lan
e
an
d
C
>
0
is
a
r
eg
u
lar
izatio
n
p
ar
am
et
er
th
at
co
n
tr
o
ls
th
e
tr
ad
e
-
o
f
f
b
etwe
en
th
e
m
ar
g
in
m
ax
im
izatio
n
an
d
th
e
class
if
icatio
n
er
r
o
r
.
ξ
j
ar
e
th
e
s
lack
v
ar
iab
les
th
at
allo
w
ce
r
tain
d
ata
p
o
i
n
ts
to
v
io
late
t
h
e
m
ar
g
in
co
n
s
tr
ain
ts
.
B
y
in
tr
o
d
u
cin
g
L
ag
r
an
g
e
m
u
lt
ip
lier
s
α
j
,
th
e
d
u
al
o
p
tim
izatio
n
p
r
o
b
lem
b
ec
o
m
es
{
2
(
)
=
∑
−
0
.
5
=
1
∑
∑
=
1
=
1
(
,
)
}
(
1
4
)
Su
b
ject
to
:
∑
=
1
=
0
,
0
≤
≤
(
1
5
)
I
n
th
is
r
esear
ch
,
th
e
p
o
p
u
lar
r
ad
ial
b
asis
Ker
n
el
f
u
n
ctio
n
(
R
B
F)
i
s
ad
o
p
ted
an
d
its
m
a
th
em
atica
l
f
o
r
m
u
la
is
g
iv
e
n
b
y
:
(
1
,
2
)
=
[
−
‖
1
−
2
‖
2
]
,
>
0
(
1
6
)
T
h
e
d
ec
is
io
n
f
u
n
ctio
n
is
g
i
v
en
as
(
)
=
[
∑
(
−
‖
−
‖
2
)
+
0
∈
]
(
1
7
)
Her
e,
SV
d
en
o
tes th
e
s
et
o
f
s
u
p
p
o
r
t
v
ec
to
r
s
.
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
.
4
,
Au
g
u
s
t
20
2
6
:
1
7
3
5
-
1
754
1742
4
.
3
.
F
is
her’
s
ra
t
io
f
ea
t
ure
s
e
lect
io
n m
et
ho
d
I
n
th
is
r
esear
ch
,
Fis
h
er
’
s
r
atio
(
FR
)
,
d
en
o
ted
as
Fr
,
i
s
u
s
ed
as
a
s
tatis
t
ical
cr
iter
io
n
to
ev
alu
ate
th
e
d
is
cr
im
in
ativ
e
p
o
wer
o
f
ea
c
h
ex
tr
ac
ted
f
ea
tu
r
e.
B
y
r
a
n
k
in
g
th
e
f
ea
tu
r
es
ac
co
r
d
i
n
g
to
th
eir
Fr
v
alu
es,
th
e
m
o
s
t
in
f
o
r
m
ativ
e
o
n
es
ar
e
s
elec
ted
an
d
th
en
u
s
ed
as
in
p
u
ts
to
th
e
SVM
alg
o
r
ith
m
.
T
h
is
m
eth
o
d
g
u
ar
an
tees
th
at
th
e
SVM
is
tr
ain
ed
with
f
ea
tu
r
es
th
at
m
ax
im
ize
class
s
ep
ar
ab
ilit
y
,
th
er
eb
y
e
n
h
an
ci
n
g
class
if
icatio
n
ac
cu
r
ac
y
wh
ile
r
ed
u
cin
g
co
m
p
u
tatio
n
al
co
m
p
lex
ity
.
T
h
e
in
teg
r
atio
n
o
f
FR
with
SVM
o
f
f
er
s
a
n
ef
f
ec
tiv
e
an
d
r
o
b
u
s
t
o
u
tlin
e
f
o
r
d
is
tin
g
u
is
h
in
g
b
et
wee
n
h
ea
lth
y
a
n
d
f
a
u
lty
o
p
e
r
a
tin
g
co
n
d
itio
n
s
in
th
e
s
tu
d
ied
s
y
s
tem
.
T
h
e
FR
ca
lcu
lates
th
e
r
atio
o
f
s
q
u
ar
ed
in
te
r
-
class
d
iv
er
g
en
c
e
to
in
tr
a
-
class
s
p
r
ea
d
o
f
a
f
e
atu
r
e
by
(
18
)
[
2
2
]
:
(
)
|
=
1
:
27
=
[
(
1
)
−
(
2
)
]
2
[
(
1
)
]
2
+
[
(
2
)
]
2
(
1
8
)
wh
er
e
(
)
an
d
(
)
ar
e
th
e
in
s
tan
ce
m
ea
n
an
d
v
a
r
ian
ce
o
f
f
ea
tu
r
e
r
esp
ec
tiv
ely
,
an
d
cla
s
s
=
1
,
2
r
ep
r
esen
t th
e
two
class
es,
an
d
i
is
th
e
i
th
f
ea
tu
r
e.
A
h
ig
h
Fr
v
alu
e
r
e
f
lects
a
s
t
r
o
n
g
d
is
cr
im
in
ativ
e
ca
p
ab
ilit
y
o
f
a
f
ea
tu
r
e.
As
in
d
icate
d
in
(
1
8
)
,
FR
ass
es
s
es
th
e
s
ig
n
if
ican
ce
o
f
f
e
atu
r
es
s
ep
ar
ately
,
wh
ich
n
o
t
o
n
ly
s
im
p
lifie
s
th
e
co
m
p
u
tatio
n
al
p
r
o
ce
s
s
b
u
t
also
en
ab
les
d
ir
ec
t
co
m
p
ar
is
o
n
o
f
th
e
r
elativ
e
im
p
o
r
tan
ce
b
et
wee
n
an
y
two
f
ea
tu
r
es.
Fig
u
r
e
7
h
ig
h
lig
h
ts
th
e
d
escr
ip
tio
n
o
f
th
e
Fis
h
er
’
s
r
ati
o
alg
o
r
ith
m
f
o
r
f
ea
tu
r
e
s
elec
ti
o
n
.
Fig
u
r
e
7
.
Flo
wch
ar
t
o
f
th
e
Fis
h
er
’
s
r
atio
alg
o
r
ith
m
4
.
4
.
E
v
a
lua
t
i
o
n m
et
rics
T
o
ass
ess
th
e
ef
f
ec
tiv
en
ess
o
f
th
e
d
ev
elo
p
e
d
d
iag
n
o
s
tic
to
o
l,
its
p
er
f
o
r
m
an
ce
was
q
u
an
tifie
d
u
s
in
g
a
s
et
o
f
ev
alu
atio
n
m
etr
ics.
T
h
ese
in
clu
d
e
ac
cu
r
ac
y
,
p
r
ec
is
io
n
,
r
ec
all,
an
d
th
e
F1
-
s
co
r
e,
wh
ich
ar
e
m
ath
em
atica
lly
f
o
r
m
u
lated
as:
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
n
ew d
ia
g
n
o
s
tic
meth
o
d
b
a
s
ed
o
n
s
u
p
p
o
r
t v
ec
to
r
ma
ch
in
e
fo
r
…
(
Hich
a
m
Za
imen
)
1743
{
(
%
)
=
+
+
+
+
×
100
(
%
)
=
+
×
100
(
%
)
=
+
×
100
1
−
(
%
)
=
2
.
×
+
×
100
(
1
9
)
w
h
er
e,
Fn
:
f
alse n
e
g
ativ
e
,
Tn
:
tr
u
e
n
eg
ativ
e
,
Fp
:
f
alse p
o
s
itiv
e
,
an
d
Tp
:
tr
u
e
p
o
s
itiv
e
.
5.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
T
h
is
s
ec
tio
n
s
ee
k
s
to
h
ig
h
lig
h
t
th
e
ca
p
a
b
ilit
y
o
f
th
e
in
t
r
o
d
u
ce
d
ML
alg
o
r
ith
m
f
o
r
I
T
SC
f
au
lt
d
iag
n
o
s
is
an
d
class
if
icatio
n
.
T
h
e
co
m
p
u
ter
p
r
o
g
r
am
s
wer
e
ca
r
r
ied
o
u
t
in
MA
T
L
AB
u
s
in
g
th
e
SVM
to
o
lb
o
x
lib
r
ar
y
.
T
h
e
d
atab
ase
is
co
m
p
o
s
ed
o
f
twen
ty
-
s
ev
en
(
9
×
3
)
f
ea
tu
r
es
(
0
9
T
DFs
ap
p
lied
to
e
ac
h
p
h
ase
cu
r
r
e
n
t)
;
s
ix
(
0
6
)
class
es,
wh
ich
ar
e
a
h
ea
lth
y
s
tato
r
win
d
in
g
an
d
f
i
v
e
I
T
SC
f
au
lts
o
f
s
ev
er
ity
(
1
%
,
2
%,
3
%,
4
%,
an
d
5
%);
an
d
f
o
u
r
lo
ad
lev
els (
2
5
%,
5
0
%,
7
5
%,
an
d
1
0
0
% o
f
f
u
ll lo
ad
)
.
T
h
e
d
is
tr
ib
u
tio
n
o
f
th
e
e
x
p
er
i
m
en
tal
d
ataset
u
s
ed
f
o
r
tr
ai
n
in
g
a
n
d
v
alid
atin
g
th
e
p
r
o
p
o
s
e
d
FR
-
SVM
f
r
am
ewo
r
k
is
s
u
m
m
ar
ize
d
in
T
ab
le
2
.
T
o
en
s
u
r
e
a
r
o
b
u
s
t
an
d
b
alan
ce
d
class
if
icatio
n
,
th
e
d
ataset
was
p
ar
titi
o
n
ed
ac
r
o
s
s
v
ar
io
u
s
s
ce
n
ar
io
s
,
in
clu
d
i
n
g
s
in
g
le
l
o
ad
l
ev
els
an
d
m
u
ltip
le
d
ata
f
u
s
io
n
co
m
b
in
atio
n
s
.
E
ac
h
o
f
th
e
s
ix
class
es
was
r
ep
r
esen
ted
b
y
2
0
0
s
eg
m
en
ts
p
e
r
lo
a
d
lev
el.
Fo
r
t
h
e
s
in
g
le
-
lo
a
d
ca
s
es,
a
to
tal
o
f
1
,
2
0
0
s
eg
m
en
ts
wer
e
u
tili
ze
d
,
co
r
r
es
p
o
n
d
in
g
to
6
0
0
,
0
0
0
r
aw
cu
r
r
e
n
t
s
am
p
les.
I
n
th
e
m
o
s
t
co
m
p
l
ex
s
ce
n
ar
io
,
wh
e
r
e
all
lo
ad
co
n
d
itio
n
s
wer
e
co
m
b
in
ed
,
th
e
d
ataset
ex
p
an
d
e
d
to
4
,
8
0
0
to
tal
s
eg
m
en
ts
(
2
.
4
1
0
6
r
aw
s
am
p
les).
As
s
h
o
wn
in
Fig
u
r
e
8
,
a
s
tr
atif
ied
1
0
-
f
o
ld
c
r
o
s
s
-
v
alid
atio
n
s
c
h
em
e
was
ad
o
p
te
d
to
e
n
s
u
r
e
r
o
b
u
s
t
an
d
u
n
b
iased
p
e
r
f
o
r
m
an
ce
e
v
alu
a
tio
n
.
I
n
ea
ch
iter
atio
n
,
9
0
%
o
f
th
e
d
ataset
was
u
s
ed
f
o
r
tr
ain
in
g
an
d
th
e
r
em
ain
in
g
1
0
%
f
o
r
test
in
g
.
T
h
is
p
r
o
ce
d
u
r
e
was
r
ep
ea
ted
ac
r
o
s
s
all
f
o
l
d
s
,
an
d
th
e
f
in
al
p
er
f
o
r
m
an
ce
was
r
ep
o
r
ted
as
th
e
av
er
a
g
e
o
v
er
th
e
1
0
r
u
n
s
,
th
e
r
eb
y
r
ed
u
cin
g
v
ar
ian
ce
an
d
m
itig
atin
g
o
v
er
f
itt
in
g
.
T
h
e
s
ettin
g
s
o
f
th
e
SVM
class
if
ier
ar
e
s
h
o
wn
in
T
ab
le
3
.
T
ab
le
2
.
Su
m
m
a
r
y
o
f
d
ata
s
am
p
le
s
ce
n
ar
io
s
D
a
t
a
s
e
t
s
c
e
n
a
r
i
o
s
(
Lo
a
d
=
%
o
f
f
u
l
l
l
o
a
d
)
N
o
.
o
f
c
l
a
ss
e
s
N
o
.
o
f
se
g
m
e
n
t
s
p
e
r
c
l
a
ss (N
)
N
o
.
o
f
t
o
t
a
l
seg
m
e
n
t
s
To
t
a
l
r
a
w
sa
mp
l
e
s
p
e
r
o
n
e
p
h
a
se
c
u
r
r
e
n
t
S
i
n
g
l
e
l
o
a
d
2
5
%
6
2
0
0
1
2
0
0
6
0
0
,
0
0
0
5
0
%
6
2
0
0
1
2
0
0
6
0
0
,
0
0
0
1
0
0
%
6
2
0
0
1
2
0
0
6
0
0
,
0
0
0
D
o
u
b
l
e
-
l
o
a
d
f
u
si
o
n
2
5
%
+
7
5
%
6
2
0
0
2
4
0
0
1
2
0
0
,
0
0
0
5
0
%
+
1
0
0
%
6
2
0
0
2
4
0
0
1
2
0
0
,
0
0
0
2
5
%
+
1
0
0
%
6
2
0
0
2
4
0
0
1
2
0
0
,
0
0
0
Tr
i
p
l
e
-
l
o
a
d
f
u
si
o
n
2
5
%
+
5
0
%
+
7
5
%
6
2
0
0
3
6
0
0
1
8
0
0
,
0
0
0
2
5
%
+
5
0
%
+
1
0
0
%
6
2
0
0
3
6
0
0
1
8
0
0
,
0
0
0
5
0
%
+
7
5
%
+
1
0
0
%
6
2
0
0
3
6
0
0
1
8
0
0
,
0
0
0
A
l
l
l
o
a
d
f
u
s
i
o
n
2
5
%
+
5
0
%
+
7
5
%
+
1
0
0
%
6
2
0
0
4
8
0
0
2
4
0
0
,
0
0
0
Fig
u
r
e
8
.
Sch
em
atic
d
iag
r
am
o
f
th
e
1
0
-
f
o
ld
cr
o
s
s
-
v
alid
atio
n
s
ch
em
e
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
.
4
,
Au
g
u
s
t
20
2
6
:
1
7
3
5
-
1
754
1744
T
ab
le
3
.
Settin
g
s
o
f
th
e
class
if
ier
S
e
t
t
i
n
g
s
S
p
e
c
i
f
i
c
a
t
i
o
n
K
e
r
n
e
l
r
a
d
i
a
l
b
a
si
s
f
u
n
c
t
i
o
n
k
e
r
n
e
l
(
R
B
F
)
K
e
r
n
e
l
c
o
e
f
f
i
c
i
e
n
t
γ
=
0
.
1
R
e
g
u
l
a
r
i
z
a
t
i
o
n
p
a
r
a
m
e
t
e
r
C =
1
0
0
5
.
1
.
SVM
cla
s
s
if
ica
t
io
n o
utput
s
wit
ho
ut
F
is
her’
s
ra
t
io
a
lg
o
rit
hm
T
h
e
in
itial
d
iag
n
o
s
tic
p
h
ase
ev
alu
ates
th
e
SVM
class
if
ier
’
s
p
er
f
o
r
m
an
ce
u
s
in
g
th
e
f
u
ll
2
7
-
elem
en
t
f
ea
tu
r
e
v
ec
to
r
,
wh
ic
h
co
n
s
is
ts
o
f
9
T
DI
s
ex
tr
ac
ted
f
r
o
m
ea
c
h
o
f
th
e
s
tato
r
cu
r
r
en
ts
.
Fig
u
r
es
9
an
d
1
0
p
r
o
v
id
e
3
D
v
is
u
aliza
tio
n
s
o
f
th
ese
f
ea
t
u
r
es (
test
in
g
d
ata
s
et)
f
o
r
b
o
t
h
s
in
g
le
-
lo
ad
an
d
m
u
lti
-
lo
ad
s
ce
n
ar
io
s
.
Fig
u
r
e
9
.
3
D
v
is
u
aliza
tio
n
o
f
th
e
2
7
f
ea
tu
r
es f
o
r
a
lo
ad
o
f
2
5
%
Fig
u
r
e
10
.
3
D
v
is
u
aliza
tio
n
o
f
th
e
2
7
f
ea
t
u
r
es f
o
r
all
m
o
to
r
lo
ad
s
Evaluation Warning : The document was created with Spire.PDF for Python.