I
nte
rna
t
io
na
l J
o
urna
l o
f
E
lect
rica
l a
nd
Co
m
p
ute
r
E
ng
in
ee
ring
(
I
J
E
CE
)
Vo
l.
8
,
No
.
1
,
Feb
r
u
ar
y
201
8
,
p
p
.
4
8
3
~4
9
6
I
SS
N:
2
0
8
8
-
8708
,
DOI
: 1
0
.
1
1
5
9
1
/
i
j
ec
e
.
v8
i
1
.
p
p
4
8
3
-
4
9
6
483
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ia
e
s
co
r
e
.
co
m/
jo
u
r
n
a
ls
/in
d
ex
.
p
h
p
/
I
JE
C
E
O
pti
m
i
zed
K
er
ne
l Ex
tre
m
e Lea
r
ni
ng
Ma
chine f
o
r
M
y
o
elect
ric
Patter
n
Recog
niti
o
n
K
ha
irul Ana
m
1
,
Adel
A
l
-
J
um
a
ily
2
1
De
p
a
rtme
n
t
o
f
El
e
c
tri
c
a
l
En
g
in
e
e
rin
g
,
Un
iv
e
rsity
o
f
J
e
m
b
e
r
,
In
d
o
n
e
sia
2
De
p
a
rtme
n
t
o
f
El
e
c
tri
c
a
l
En
g
in
e
e
rin
g
,
Un
iv
e
rsity
o
f
T
e
c
h
n
o
lo
g
y
,
S
y
d
n
e
y
,
A
u
stra
li
a
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Oct
22
,
2
0
1
7
R
ev
i
s
ed
J
an
5
,
2
0
1
8
A
cc
ep
ted
J
an
20
,
2
0
1
8
M
y
o
e
lec
tri
c
p
a
tt
e
rn
re
c
o
g
n
it
io
n
(
M
P
R)
is
u
se
d
to
d
e
tec
t
u
se
r’s
in
ten
ti
o
n
to
a
c
h
iev
e
a
s
m
o
o
th
in
tera
c
ti
o
n
b
e
t
w
e
e
n
h
u
m
a
n
a
n
d
m
a
c
h
in
e
.
T
h
e
p
e
rf
o
r
m
a
n
c
e
o
f
M
P
R
is
in
f
lu
e
n
c
e
d
b
y
th
e
f
e
a
tu
re
s ex
tra
c
ted
a
n
d
t
h
e
c
las
sif
ier e
m
p
lo
y
e
d
.
A
k
e
rn
e
l
e
x
tre
m
e
lea
rn
in
g
m
a
c
h
in
e
e
sp
e
c
iall
y
ra
d
ial
b
a
sis
f
u
n
c
ti
o
n
e
x
tre
m
e
lea
rn
in
g
m
a
c
h
in
e
(RBF
-
EL
M
)
h
a
s
e
m
e
r
g
e
d
a
s
o
n
e
o
f
th
e
p
o
ten
ti
a
l
c
las
si
f
iers
f
o
r
M
P
R.
Ho
w
e
v
e
r,
R
BF
-
EL
M
s
h
o
u
ld
b
e
o
p
ti
m
ize
d
to
w
o
rk
e
ff
ici
e
n
tl
y
.
T
h
is
p
a
p
e
r
p
ro
p
o
se
d
a
n
o
p
ti
m
iza
ti
o
n
o
f
RBF
-
E
L
M
p
a
ra
m
e
ters
u
sin
g
h
y
b
rid
iza
ti
o
n
o
f
p
a
rti
c
le
sw
a
r
m
o
p
ti
m
iz
a
ti
o
n
(
P
S
O)
a
n
d
a
w
a
v
e
let
f
u
n
c
ti
o
n
.
T
h
e
se
p
ro
p
o
se
d
sy
ste
m
s
a
r
e
e
m
p
lo
y
e
d
to
c
las
sify
f
in
g
e
r
m
o
v
e
m
e
n
ts
o
n
t
h
e
a
m
p
u
tee
s
a
n
d
a
b
le
-
b
o
d
ied
su
b
j
e
c
ts
u
sin
g
e
lec
tro
m
y
o
g
ra
p
h
y
si
g
n
a
ls.
T
h
e
e
x
p
e
ri
m
e
n
tal
re
su
lt
s
sh
o
w
th
a
t
t
h
e
a
c
c
u
ra
c
y
o
f
th
e
o
p
ti
m
ize
d
RBF
-
EL
M
is
9
5
.
7
1
%
a
n
d
9
4
.
2
7
%
in
th
e
h
e
a
lt
h
y
su
b
jec
ts
a
n
d
th
e
a
m
p
u
tee
s,
re
sp
e
c
ti
v
e
l
y
.
M
e
a
n
w
h
il
e
,
th
e
o
p
t
im
iza
ti
o
n
u
si
n
g
P
S
O
o
n
ly
a
tt
a
in
e
d
th
e
a
v
e
ra
g
e
a
c
c
u
ra
c
y
o
f
9
5
.
5
3
%
,
a
n
d
9
2
.
5
5
%
,
o
n
th
e
h
e
a
lt
h
y
su
b
jec
ts
a
n
d
th
e
a
m
p
u
tee
s,
re
sp
e
c
ti
v
e
l
y
.
T
h
e
e
x
p
e
ri
m
e
n
tal
re
su
lt
s
a
lso
sh
o
w
th
a
t
S
W
-
RBF
-
EL
M
a
c
h
iev
e
d
th
e
a
c
c
u
ra
c
y
th
a
t
is
b
e
tt
e
r
th
a
n
o
th
e
r
w
e
ll
-
k
n
o
w
n
c
las
si
f
i
e
rs
su
c
h
a
s
su
p
p
o
rt
v
e
c
to
r
m
a
c
h
in
e
(S
V
M
)
,
li
n
e
a
r
d
isc
rim
in
a
n
t
a
n
a
l
y
sis
(
L
D
A
)
a
n
d
k
-
n
e
a
re
st n
e
ig
h
b
o
r
(k
NN
).
K
ey
w
o
r
d
:
C
las
s
i
f
icatio
n
E
lectr
o
m
y
o
g
r
ap
h
y
E
x
tr
e
m
e
lear
n
i
n
g
m
ac
h
in
e
P
atter
n
r
ec
o
g
n
itio
n
W
av
elet
Co
p
y
rig
h
t
©
2
0
1
8
In
stit
u
te o
f
A
d
v
a
n
c
e
d
E
n
g
i
n
e
e
rin
g
a
n
d
S
c
ien
c
e
.
All r
ig
h
ts
re
se
rv
e
d
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
Kh
air
u
l
A
n
a
m
,
Dep
ar
t
m
en
t o
f
E
lectr
ical
E
n
g
i
n
ee
r
in
g
,
Un
i
v
er
s
it
y
o
f
J
e
m
b
er
,
J
l.
Kali
m
an
ta
n
3
7
J
em
b
er
6
8
1
2
1
I
n
d
o
n
esia
.
E
m
ail:
k
h
a
ir
u
l
@
u
n
ej
.
ac
.
id
1.
I
NT
RO
D
UCT
I
O
N
E
x
tr
e
m
e
lear
n
i
n
g
m
ac
h
in
e
(
E
L
M)
is
a
k
i
n
d
o
f
s
i
n
g
le
la
y
er
f
eed
-
f
o
r
w
ar
d
n
et
w
o
r
k
s
(
S
L
FN
s
)
th
at
h
as
f
ast
tr
ai
n
i
n
g
ti
m
e
[
1
]
.
E
L
M
is
a
g
r
ea
t
i
m
p
r
o
v
e
m
en
t
o
f
f
ee
d
-
f
o
r
w
ar
d
n
eu
r
al
n
e
t
w
o
r
k
s
(
F
FNNs
)
,
w
h
ic
h
v
er
y
co
n
s
id
er
ab
l
y
r
ed
u
ce
s
th
e
tr
ai
n
in
g
ti
m
e
o
f
FF
NNs
b
y
o
m
it
tin
g
th
e
iter
ati
v
e
lear
n
i
n
g
p
r
o
ce
s
s
.
I
n
E
L
M,
t
h
e
w
ei
g
h
ts
an
d
b
iases
o
f
t
h
e
h
i
d
d
en
n
o
d
e
ar
e
d
eter
m
i
n
ed
r
an
d
o
m
l
y
,
w
h
il
e
t
h
e
o
u
tp
u
t
w
e
ig
h
ts
ar
e
ca
lcu
lated
an
al
y
ticall
y
.
T
h
er
ef
o
r
e,
th
e
tr
ain
i
n
g
ti
m
e
i
s
v
er
y
s
h
o
r
t c
o
m
p
a
r
ed
to
th
e
tr
ad
itio
n
al
n
eu
r
al
n
e
t
w
o
r
k
s
.
T
h
e
d
ev
elo
p
m
e
n
t
o
f
E
L
M
i
s
v
er
y
f
a
s
t
a
n
d
en
co
m
p
as
s
i
n
g
m
a
n
y
ap
p
licatio
n
s
.
T
o
im
p
r
o
v
e
th
e
s
tab
ilit
y
o
f
E
L
M,
W
an
g
,
et
al.
[
2
]
p
r
o
p
o
s
ed
a
m
eth
o
d
to
f
in
d
a
h
i
g
h
q
u
alit
y
o
f
f
ea
t
u
r
e
m
a
p
p
in
g
in
t
h
e
f
ea
t
u
r
e
s
tag
e.
T
h
er
ef
o
r
e,
th
e
o
u
tp
u
t
w
ei
g
h
t
ca
lcu
lat
io
n
u
s
in
g
r
id
g
e
r
eg
r
ess
io
n
ca
n
b
e
o
p
tim
ized
.
On
th
e
o
th
er
h
an
d
,
th
e
id
ea
to
co
n
s
tr
u
ct
a
co
m
p
ac
t
E
L
M
w
as
p
r
o
p
o
s
ed
b
y
ad
d
in
g
a
n
e
w
ap
p
r
o
p
r
iate
h
id
d
en
n
e
u
r
o
n
[
3
]
.
L
i
k
e
w
is
e,
th
e
m
et
h
o
d
to
r
ed
u
c
e
o
f
th
e
s
ize
E
L
M
w
a
s
al
s
o
p
r
o
p
o
s
ed
in
[
4
]
,
[
5
]
.
Ma
n
y
o
t
h
er
d
ev
elo
p
m
e
n
ts
o
f
E
L
M
h
a
v
e
b
ee
n
p
r
o
p
o
s
ed
,
s
u
ch
as
E
L
M
o
n
o
n
li
n
e
s
eq
u
en
tial
d
ata
[
6
]
,
[
7
]
,
en
s
e
m
b
l
e
E
L
M
[
8
]
,
s
em
i
-
s
u
p
er
v
i
s
ed
an
d
u
n
s
u
p
er
v
i
s
e
d
E
L
M
[
9
]
,
[
1
0
]
,
E
L
M
f
o
r
i
m
b
al
an
ce
d
d
ata
[
1
1
]
,
an
d
in
cr
e
m
e
n
tal
E
L
M
[
1
2
]
.
E
L
M
m
et
h
o
d
h
a
s
b
ee
n
u
s
ed
f
o
r
a
w
id
e
r
a
n
g
e
o
f
ap
p
licati
o
n
[
13]
.
T
h
e
E
L
M
h
as
b
ee
n
ap
p
lied
to
elec
tr
o
m
y
o
g
r
ap
h
y
(
E
MG
)
-
b
as
ed
p
atter
n
r
ec
o
g
n
itio
n
[
1
4
]
,
f
ac
e
r
ec
o
g
n
it
io
n
[
1
5
]
,
ch
ar
ac
te
r
r
ec
o
g
n
itio
n
[
1
6
]
,
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
8
,
No
.
1
,
Feb
r
u
ar
y
2
0
1
8
:
4
8
3
–
4
9
6
484
[
1
7
]
.
Mo
r
eo
v
er
,
it
h
as
b
ee
n
i
m
p
le
m
e
n
ted
i
n
p
r
o
tein
s
tr
u
ctu
r
e
p
r
ed
ictio
n
[
1
8
]
,
ca
n
ce
r
d
ete
ctio
n
[
1
9
]
,
elec
tr
ical
p
o
w
er
s
y
s
te
m
p
r
o
b
le
m
[
2
0
]
an
d
p
h
y
s
ical
p
ar
a
m
eter
es
ti
m
atio
n
[
2
1
]
.
Nev
er
th
e
less
,
t
h
e
h
id
d
en
n
o
d
e
p
ar
am
eter
s
,
t
h
e
i
n
p
u
t
w
e
ig
h
ts
,
a
n
d
b
iases
,
w
h
ich
ar
e
d
eter
m
i
n
ed
ar
b
itra
r
ily
,
r
es
u
lt
in
a
n
o
n
-
o
p
t
i
m
al
s
y
s
te
m
.
T
h
er
ef
o
r
e,
s
o
m
e
ef
f
o
r
ts
d
ea
lin
g
w
it
h
th
e
o
p
ti
m
izatio
n
p
r
o
b
lem
i
n
E
L
M
h
av
e
b
ee
n
m
ad
e.
Sel
f
-
a
d
ap
tiv
e
ev
o
l
u
tio
n
ar
y
E
L
M
(
S
A
E
-
E
L
M)
[
2
2
]
,
an
d
p
ar
ticle
s
w
ar
m
o
p
ti
m
izatio
n
E
L
M(
P
SO
-
E
L
M)
[
2
3
]
ar
e
s
o
m
e
m
e
th
o
d
s
d
ev
elo
p
ed
to
o
p
tim
ize
t
h
e
h
id
d
en
n
o
d
e
p
ar
am
et
er
s
.
E
L
M
is
n
o
t
m
er
el
y
w
o
r
k
in
g
o
n
t
h
e
n
o
d
e
s
t
y
le.
A
k
er
n
el
f
o
r
m
ca
n
b
e
i
n
co
r
p
o
r
ated
in
E
L
M
b
y
r
ep
lacin
g
t
h
e
n
o
d
e
p
r
o
ce
s
s
in
g
s
tr
u
ct
u
r
e
w
it
h
a
k
er
n
el
f
u
n
ctio
n
.
T
h
is
k
er
n
el
E
L
M
ca
n
b
e
co
n
s
id
e
r
ed
as
a
v
ar
ian
ce
o
f
lea
s
t
s
q
u
ar
e
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
i
n
e
(
L
S
-
SVM)
w
it
h
o
u
t
t
h
e
o
u
tp
u
t
b
ia
s
[
2
4
]
.
Si
m
i
lar
to
t
h
e
n
o
d
e
-
b
ased
E
L
M,
th
e
k
er
n
el
E
L
M
f
ac
es
th
e
o
p
ti
m
iza
tio
n
p
r
o
b
le
m
to
o
.
T
h
e
ef
f
icac
y
o
f
th
e
k
er
n
el
E
L
M
g
r
ea
tl
y
d
ep
en
d
s
o
n
th
e
o
p
ti
m
u
m
co
m
b
in
atio
n
o
f
t
h
e
k
er
n
el
p
ar
a
m
et
er
s
[
2
5
]
.
T
h
e
p
o
p
u
lar
g
r
id
s
ea
r
ch
alg
o
r
ith
m
t
h
at
i
s
s
i
m
p
le
w
a
s
u
s
ed
to
s
ea
r
ch
t
h
e
o
p
tim
a
l
k
er
n
el
[
1
4
]
.
Ho
w
e
v
er
,
th
e
ex
h
a
u
s
ti
v
e
g
r
id
s
ea
r
ch
o
n
a
lar
g
e
n
u
m
b
er
o
f
th
e
p
ar
a
m
eter
s
p
ac
es
m
a
y
r
es
u
lt in
a
v
er
y
ti
m
e
-
co
n
s
u
m
i
n
g
p
r
o
ce
s
s
.
A
p
o
p
u
lar
p
ar
ticle
s
w
ar
m
o
p
ti
m
izatio
n
(
P
SO)
alg
o
r
ith
m
ca
n
b
e
a
p
r
o
m
is
i
n
g
s
o
l
u
tio
n
f
o
r
o
p
ti
m
izi
n
g
th
e
k
er
n
el
p
ar
a
m
eter
s
in
t
h
e
k
er
n
el
E
L
M.
T
h
e
P
SO
h
as
b
ee
n
i
m
p
le
m
e
n
ted
in
m
a
n
y
ar
ea
s
s
u
c
h
as
m
ed
ical
[
2
6
]
,
p
o
w
er
s
y
s
te
m
[
2
7
]
,
an
d
cir
cu
it
d
esi
g
n
[
2
8
]
.
T
o
th
e
b
est
o
f
th
e
a
u
t
h
o
r
’
s
k
n
o
w
led
g
e,
n
o
o
n
e
e
m
p
lo
y
s
P
SO
to
o
p
ti
m
ize
t
h
e
k
er
n
el
E
L
M.
I
n
t
h
e
p
r
ac
tical
ap
p
licatio
n
,
L
i
n
g
,
e
t
al.
[
2
9
]
f
o
u
n
d
t
h
at
s
o
m
e
ti
m
e
s
,
P
SO
is
b
ein
g
tr
ap
p
ed
in
th
e
lo
ca
l
o
p
ti
m
a.
T
h
er
ef
o
r
e,
th
e
y
p
r
o
p
o
s
ed
P
SO
m
u
tated
b
y
w
a
v
elet.
T
h
e
ex
is
t
en
ce
o
f
th
e
w
a
v
elet
m
u
tatio
n
in
P
SO
d
ep
en
d
s
o
n
t
h
e
m
u
tatio
n
p
r
o
b
ab
ilit
y
.
T
h
e
h
ig
h
er
t
h
e
m
u
tatio
n
p
r
o
b
ab
ilit
y
i
s
,
t
h
e
g
r
ea
ter
th
e
ch
an
ce
o
f
th
e
w
a
v
elet
i
s
u
p
d
at
in
g
t
h
e
p
ar
ticles o
f
P
SO.
T
h
is
p
ap
er
in
tr
o
d
u
ce
s
a
s
w
ar
m
r
ad
i
al
b
asis
f
u
n
ctio
n
ex
tr
e
m
e
lear
n
i
n
g
m
ac
h
in
e
(
S
R
B
F
-
E
L
M)
,
th
e
r
ad
ial
b
asis
f
u
n
ct
io
n
k
er
n
el
E
L
M
o
p
ti
m
ized
b
y
P
SO
.
I
n
ad
d
itio
n
,
th
e
p
ap
er
p
r
o
p
o
s
es
a
s
w
ar
m
w
a
v
elet
r
ad
ial
b
asis
f
u
n
c
tio
n
e
x
tr
e
m
e
lear
n
i
n
g
m
ac
h
in
e
(
SW
-
R
B
F
-
E
L
M)
,
th
e
o
p
ti
m
izatio
n
o
f
r
ad
ial
b
a
s
is
f
u
n
c
tio
n
k
er
n
e
l
E
L
M
u
s
in
g
co
m
b
in
at
io
n
P
SO
an
d
w
a
v
elet
.
T
h
e
w
a
v
elet
d
if
f
er
s
S
R
B
F
-
L
E
M
a
n
d
SW
-
R
B
F
-
E
L
M.
T
h
e
w
a
v
elet
is
i
m
p
le
m
en
ted
u
s
in
g
a
m
u
ta
tio
n
p
r
o
b
ab
ilit
y
.
S
R
B
F
-
E
L
M
ca
n
b
e
co
n
s
id
er
ed
as
SW
-
RBF
-
E
L
M
w
it
h
ze
r
o
m
u
tatio
n
p
r
o
b
ab
ilit
y
.
I
n
th
i
s
p
ap
er
,
SR
B
F
-
E
L
M
an
d
S
W
-
R
B
F
-
E
L
M
ar
e
ap
p
lied
to
m
y
o
elec
tr
ic
p
atter
n
r
ec
o
g
n
itio
n
(
M
-
P
R
)
to
class
i
f
y
th
e
in
d
i
v
id
u
al
a
n
d
co
m
b
in
ed
f
in
g
er
m
o
v
e
m
e
n
ts
u
s
i
n
g
t
w
o
E
MG
ch
an
n
el
s
.
T
h
e
m
ai
n
co
n
tr
ib
u
tio
n
o
f
t
h
is
p
ap
er
is
o
n
th
e
o
p
ti
m
izat
io
n
o
f
k
er
n
el
e
x
tr
e
m
e
lear
n
in
g
m
ac
h
in
e
P
SO
an
d
w
av
e
let.
T
h
e
s
ec
o
n
d
co
n
t
r
ib
u
tio
n
i
s
t
h
e
i
m
p
le
m
e
n
tatio
n
o
f
t
h
e
p
r
o
p
o
s
ed
s
y
s
te
m
o
n
m
y
o
elec
tr
ic
p
atter
n
r
ec
o
g
n
itio
n
to
i
m
p
r
o
v
e
th
e
p
er
f
o
r
m
a
n
ce
o
f
MP
R
.
T
h
e
s
tr
u
ctu
r
e
o
f
t
h
is
p
ap
er
is
as
f
o
llo
w
s
.
T
h
e
s
ec
o
n
d
s
ec
ti
o
n
w
ill
d
is
c
u
s
s
t
h
e
b
asic
t
h
eo
r
y
o
f
P
SO
an
d
th
e
h
y
b
r
id
izati
o
n
o
f
w
a
v
elet
an
d
P
SO.
T
h
en
,
th
e
e
x
p
er
i
m
en
tal
s
et
u
p
is
p
r
ese
n
ted
i
n
t
h
e
th
ir
d
s
ec
t
io
n
.
Nex
t,
i
n
t
h
e
f
o
u
r
t
h
s
ec
tio
n
,
t
h
e
e
x
p
er
i
m
e
n
tal
r
es
u
lt
s
o
n
t
h
e
ab
le
-
b
o
d
ied
s
u
b
j
ec
ts
ar
e
d
i
s
cu
s
s
ed
.
A
d
d
itio
n
al
ex
p
er
i
m
e
n
t
o
n
th
e
a
m
p
u
tee
s
u
b
j
ec
ts
is
also
p
r
o
v
id
ed
.
Fin
all
y
,
t
h
e
f
i
f
t
h
s
ec
tio
n
e
n
d
s
th
e
p
ap
er
w
it
h
th
e
co
n
clu
s
io
n
.
2.
RE
S
E
ARCH
M
E
T
H
O
D
2
.
1
.
K
er
nel
E
x
t
re
m
e
L
ea
rning
M
a
chine
E
L
M
is
a
lear
n
i
n
g
alg
o
r
it
h
m
f
o
r
s
in
g
le
la
y
er
f
ee
d
f
o
r
w
ar
d
n
et
w
o
r
k
s
(
SLFNs).
I
n
clas
s
ic
al
SL
F
Ns,
n
et
w
o
r
k
p
ar
a
m
e
ter
s
ar
e
tu
n
ed
iter
ativ
el
y
w
h
ile
i
n
E
L
M;
m
o
s
t o
f
t
h
ese
p
ar
a
m
eter
s
ar
e
d
eter
m
i
n
ed
an
al
y
ticall
y
.
Hid
d
en
p
ar
am
eter
s
ca
n
b
e
in
d
ep
en
d
en
tl
y
ca
lc
u
lated
f
r
o
m
th
e
tr
ain
i
n
g
d
ata,
an
d
o
u
tp
u
t
p
ar
am
eter
s
ca
n
b
e
d
eter
m
in
ed
b
y
t
h
e
p
s
e
u
d
o
-
in
v
er
s
e
m
et
h
o
d
.
A
s
a
r
es
u
lt,
t
h
e
l
ea
r
n
in
g
o
f
E
L
M
ca
n
b
e
ca
r
r
ied
o
u
t
f
ast
co
m
p
ar
ed
to
th
e
o
th
er
lear
n
i
n
g
al
g
o
r
ith
m
s
[
2
5
]
.
As d
escr
ib
ed
in
[
2
5
]
,
th
e
o
u
tp
u
t o
f
E
L
M
is
d
e
f
in
ed
b
y
:
(
)
(
)
(
)
(
1
)
w
h
er
e
g
(
x
)
is
th
e
f
ea
t
u
r
e
m
ap
p
in
g
i
n
t
h
e
h
id
d
en
la
y
er
,
T
i
s
th
e
tar
g
et
an
d
C
i
s
th
e
r
eg
u
lat
io
n
p
ar
am
eter
o
f
E
L
M.
T
h
e
f
ea
t
u
r
e
m
ap
p
in
g
in
th
e
h
id
d
en
la
y
er
o
f
E
L
M
ca
n
b
e
r
ep
lace
d
b
y
a
k
er
n
el
f
u
n
ctio
n
.
T
h
er
ef
o
r
e,
th
e
f
o
r
m
u
latio
n
o
f
th
e
k
er
n
el
b
ase
d
E
L
M
is
d
ef
i
n
ed
b
y
:
(
)
[
(
)
(
)
]
(
)
(
2
)
w
h
er
e
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:
2
0
8
8
-
8708
Op
timiz
ed
K
ern
el
E
xt
r
eme
Le
a
r
n
in
g
Ma
ch
in
e
f
o
r
Myo
elec
tr
ic
P
a
tter
n
R
ec
o
g
n
itio
n
(
K
h
a
ir
u
l A
n
a
m
)
485
(
)
(
)
(
)
(
3
)
an
d
K
is
a
k
er
n
el
f
u
n
ctio
n
as
s
h
o
w
n
i
n
Eq
u
at
io
n
(
4
)
to
E
q
u
at
io
n
(
6
)
.
R
a
dia
l
b
a
s
is fun
c
tio
n
:
(
)
(
‖
‖
)
(
4
)
L
in
e
a
r
:
(
)
(
5
)
Pol
ynom
ia
l
:
(
)
(
)
(
6
)
2
.
2
.
P
a
rt
icle
Sw
a
rm
O
pt
i
m
i
za
t
io
n
(
P
SO
)
P
ar
ticle
s
w
ar
m
o
p
ti
m
izatio
n
(
P
SO)
is
a
p
o
p
u
latio
n
-
b
ased
s
t
o
ch
asti
c
o
p
ti
m
izatio
n
al
g
o
r
ith
m
t
h
at
h
a
s
b
ee
n
ap
p
lied
w
id
el
y
in
m
an
y
o
p
tim
izatio
n
ar
ea
s
[
2
9
-
31]
.
PS
O
is
in
s
p
ir
ed
b
y
th
e
s
o
cial
b
eh
av
io
r
s
o
f
a
n
i
m
als
lik
e
f
i
s
h
s
c
h
o
o
lin
g
a
n
d
b
ir
d
f
lo
ck
i
n
g
[
2
9
]
.
T
h
e
p
ar
ticle
s
w
ar
m
d
o
es
n
o
t
u
s
e
s
elec
tio
n
.
I
t
m
ea
n
s
t
h
at
all
p
o
p
u
latio
n
m
e
m
b
er
s
s
u
r
v
iv
e
f
r
o
m
t
h
e
b
eg
i
n
n
i
n
g
u
n
til
t
h
e
en
d
[
3
2
]
.
I
n
th
e
P
SO,
a
s
war
m
o
f
i
n
ter
ac
ti
n
g
p
ar
ticles
m
o
v
es
i
n
a
n
n
-
d
i
m
e
n
s
io
n
al
s
ea
r
ch
s
p
ac
e
o
f
t
h
e
p
r
o
b
le
m
’
s
p
o
s
s
ib
le
s
o
lu
t
io
n
.
Fo
u
r
ele
m
e
n
ts
th
a
t
ar
e
a
p
o
s
itio
n
⃗
,
a
v
elo
cit
y
⃗
,
th
e
b
est
p
r
ev
io
u
s
(
lo
ca
l)
p
o
s
itio
n
⃗
an
d
th
e
b
est
g
lo
b
al
p
o
s
itio
n
⃗
r
ep
r
esen
t
a
p
ar
ticle
in
th
e
s
w
ar
m
.
So
m
e
g
en
er
atio
n
s
ar
e
g
en
er
ated
to
u
p
d
ate
th
e
p
ar
ticle’
s
p
o
s
iti
ons
an
d
v
elo
cities.
T
h
e
pa
r
ticles
ex
p
lo
r
e
th
e
p
r
o
m
i
s
i
n
g
d
o
m
ai
n
to
f
in
d
t
h
e
b
est
s
o
l
u
tio
n
s
,
w
h
ich
s
p
r
ea
d
th
r
o
u
g
h
o
u
t
th
e
s
w
ar
m
.
T
h
e
p
ar
am
eter
ad
ap
tatio
n
s
ar
e
g
iv
e
n
b
y
:
(
)
(
)
(
)
(
7
)
(
)
(
)
(
(
)
(
)
)
(
(
)
(
)
)
(
8
)
W
h
er
e
[
]
[
]
I
n
t
h
e
ab
o
v
e
eq
u
atio
n
s
,
q
i
p
d
en
o
tes
th
e
b
est
p
r
ev
io
u
s
(
lo
ca
l)
p
o
s
itio
n
a
n
d
q
i
g
d
en
o
tes
t
h
e
b
est
g
lo
b
al
p
o
s
itio
n
.
Mo
r
eo
v
er
,
t
r
ep
r
esen
ts
t
h
e
g
en
er
at
io
n
,
k
d
e
n
o
tes
th
e
n
u
m
b
er
o
f
t
h
e
p
ar
ticle
s
i
n
t
h
e
s
w
ar
m
,
d
d
en
o
te
s
th
e
n
u
m
b
er
o
f
d
i
m
en
s
io
n
s
,
i
s
in
er
tia
w
ei
g
h
t,
an
d
c
1
an
d
c
2
ar
e
ac
ce
ler
atio
n
co
n
s
tan
t
s
w
h
i
ch
ar
e
w
ei
g
h
ted
b
y
r
1
an
d
r
2
.
2
.
3
.
P
SO
w
it
h
Wa
v
elet
M
u
t
a
t
io
n
P
SO
t
y
p
icall
y
co
n
v
er
g
es
i
n
t
h
e
ea
r
ly
s
ta
g
e
o
f
t
h
e
s
ea
r
ch
in
g
p
r
o
ce
s
s
.
T
h
is
in
d
icate
s
t
h
at
P
S
O
ten
d
s
to
b
e
tr
ap
p
e
d
in
th
e
lo
ca
l
o
p
ti
m
a.
T
h
is
s
h
o
r
tco
m
i
n
g
m
a
y
in
f
l
u
en
ce
t
h
e
p
er
f
o
r
m
a
n
ce
o
f
t
h
e
m
y
o
elec
tr
ic
f
i
n
g
er
class
i
f
icatio
n
.
O
n
e
o
f
t
h
e
s
o
lu
tio
n
s
o
f
t
h
e
lo
ca
l
o
p
ti
m
a
is
b
y
in
j
ec
tin
g
a
w
a
v
elet
f
u
n
ctio
n
i
n
s
id
e
t
h
e
P
SO.
T
h
e
w
a
v
elet
m
u
tates
t
h
e
s
w
ar
m
p
ar
ticle
in
s
m
all
p
r
o
b
ab
ilit
y
to
cr
ea
te
a
p
o
s
s
ib
ilit
y
f
o
r
th
e
s
war
m
p
ar
ticle
to
g
et
o
u
t f
r
o
m
t
h
e
lo
ca
l o
p
ti
m
a.
T
h
e
w
a
v
elet
m
u
tatio
n
i
n
P
SO
w
a
s
p
r
o
p
o
s
ed
b
y
L
i
n
g
e
t
al
.
[
2
9
]
.
A
m
u
tatio
n
c
h
a
n
ce
is
d
r
iv
en
b
y
a
m
u
tatio
n
p
r
o
b
ab
ilit
y
p
m
[
0
1
]
.
I
f
x
i
(
t)
is
s
elec
ted
to
b
e
m
u
ta
ted
th
e
n
a
n
e
w
p
o
s
i
tio
n
is
g
i
v
e
n
b
y
:
(
)
{
(
)
(
(
)
)
(
)
(
(
)
)
(
9
)
w
h
er
e
par
max
an
d
par
min
ar
e
th
e
m
ax
i
m
u
m
an
d
m
in
i
m
u
m
p
o
s
itio
n
,
r
e
s
p
ec
tiv
el
y
.
A
s
f
o
r
,
it is
th
e
Mo
r
let
w
a
v
elet
f
u
n
ctio
n
d
ef
i
n
ed
b
y
:
√
(
)
⁄
(
(
)
)
(
1
0
)
T
h
e
v
ar
iab
le
“
a”
in
t
h
e
Mo
r
l
et
w
a
v
elet
is
d
eter
m
in
ed
b
y
eq
u
atio
n
:
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
8
,
No
.
1
,
Feb
r
u
ar
y
2
0
1
8
:
4
8
3
–
4
9
6
486
(
)
(
)
(
)
(
1
1
)
T
h
e
o
b
j
ec
tiv
e
o
f
th
e
o
p
ti
m
izat
io
n
u
s
in
g
t
h
e
w
av
ele
t
-
P
SO i
s
t
o
f
in
d
th
e
o
p
ti
m
u
m
p
ar
a
m
eter
s
o
f
th
e
k
er
n
el
E
L
M
t
h
at
m
in
i
m
ize
t
h
e
class
i
f
icatio
n
er
r
o
r
o
f
th
e
f
i
n
g
er
m
o
tio
n
r
ec
o
g
n
i
tio
n
.
A
3
-
f
o
l
d
cr
o
s
s
v
alid
atio
n
w
a
s
e
m
p
lo
y
ed
to
m
ea
s
u
r
e
t
h
e
er
r
o
r
.
M
o
r
eo
v
er
,
th
e
f
itn
e
s
s
f
u
n
ctio
n
o
f
p
ar
ticle
x
is
d
ef
i
n
ed
b
y
(
)
∑
(
)
(
1
2
)
w
h
er
e
N
v
is
t
h
e
n
u
m
b
er
o
f
cr
o
s
s
v
a
lid
atio
n
s
,
E
n
is
t
h
e
er
r
o
r
in
ea
ch
v
alid
atio
n
p
r
o
ce
s
s
.
T
h
e
p
s
eu
d
o
co
d
e
o
f
th
e
w
a
v
el
et
m
u
tatio
n
f
o
r
o
p
ti
m
izin
g
t
h
e
p
ar
am
eter
s
o
f
t
h
e
k
er
n
el
b
ase
d
E
L
M
is
p
r
esen
t
ed
in
Fi
g
u
r
e
1
.
Begin
Load emg_features, classes
t
1
// iteration number
Initialize x(t)
// x(t) : position, a particle swarm
Evaluate f(x)
// f(x): fitness function Eq.(12)
Initialize v
// v : velocity
̃
//
̃
: personal best position
̂
̃
//
̂
: global best position
While (condition satisfied) do
i
i+1
update position of particle x(i)
//
Eq
.(7)
update velocity v(i)
//
Eq.(8)
if v(i)> vmax, v(i)=vmax end
if v(
i)<
-
vmax, v(i)=
-
vmax end
update
̃
if new
̃
better th
an previuos
̃
update
̂
if new
̂
better than previuos
̂
perform wavelet mutation operation with
p
m
Updating x(i
)
//
Eq
.
(9)
Evaluate f(x(i)) // f(x): fitness functi
on
Eq.(12)
end
end
Fig
u
r
e
1
.
T
h
e
p
s
eu
d
o
co
d
e
f
o
r
P
SO
w
it
h
w
av
e
let
m
u
tatio
n
f
o
r
o
p
tim
izi
n
g
t
h
e
p
ar
a
m
eter
s
o
f
E
L
M
2
.
4
.
T
he
E
x
per
i
m
ent
a
l Set
up
Fig
u
r
e
2
s
h
o
w
s
th
e
d
iag
r
a
m
b
lo
ck
o
f
th
e
ex
p
er
i
m
en
t
co
n
d
u
cted
in
t
h
i
s
s
ec
t
io
n
.
T
h
e
E
M
G
d
ata
w
as
co
llected
f
r
o
m
eig
h
t
ab
le
-
b
o
d
ied
s
u
b
j
ec
ts
,
t
w
o
f
e
m
ales
a
n
d
s
i
x
m
ale
s
a
g
ed
2
4
-
6
0
y
e
ar
s
o
ld
.
T
w
o
E
M
G
M
y
o
Scan
™
T
9
5
0
3
M
s
en
s
o
r
s
o
r
elec
tr
o
d
es
w
er
e
p
lace
d
o
n
th
e
f
o
r
ea
r
m
o
f
t
h
e
s
u
b
j
ec
t
to
co
llect
m
y
o
elec
tr
ic
s
ig
n
a
l
f
r
o
m
f
lex
o
r
p
o
licis
lo
n
g
u
s
(
FP
L
)
an
d
f
le
x
o
r
d
ig
ito
r
iu
m
s
u
p
er
f
iciali
s
(
FDS)
m
u
s
cles,
as
s
h
o
w
n
in
Fig
u
r
e
3
.
T
h
e
Fle
x
C
o
m
p
I
n
f
in
iti™
S
y
s
te
m
f
r
o
m
T
h
o
u
g
h
t
T
ec
h
n
o
lo
g
y
ac
q
u
ir
ed
t
h
e
E
MG
s
i
g
n
als
w
it
h
a
s
a
m
p
li
n
g
f
r
eq
u
e
n
c
y
o
f
2
0
0
0
H
z
an
d
th
en
a
m
p
li
f
ied
th
e
s
ig
n
a
ls
w
it
h
a
to
tal
g
ai
n
o
f
1
0
0
0
.
Fig
u
r
e
2
.
T
h
e
ex
p
er
im
e
n
tal
s
e
t
u
p
o
f
th
e
P
SO
-
w
a
v
elet
m
u
tat
i
o
n
f
o
r
E
L
M
p
ar
a
m
eter
s
o
p
ti
m
i
za
tio
n
W
i
n
d
o
w
i
n
g
F
e
a
t
u
r
e
E
x
t
r
a
c
t
i
o
n
D
i
m
e
n
s
i
o
n
a
l
i
t
y
r
e
d
u
c
t
i
o
n
:
S
R
D
A
T
w
o
E
M
G
C
h
a
n
n
e
l
s
P
o
s
t
-
P
r
o
c
e
s
s
i
n
g
M
A
V
,
M
A
V
S
,
Z
C
,
S
S
C
,
S
S
,
W
L
,
R
M
S
,
A
R
6
,
H
T
D
C
l
a
s
s
i
f
i
c
a
t
i
o
n
:
S
W
-
R
B
F
-
E
L
M
K
e
r
n
e
l
E
L
M
p
a
r
a
m
e
t
e
r
s
:
C
,
g
P
S
O
-
W
a
v
e
l
e
t
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:
2
0
8
8
-
8708
Op
timiz
ed
K
ern
el
E
xt
r
eme
Le
a
r
n
in
g
Ma
ch
in
e
f
o
r
Myo
elec
tr
ic
P
a
tter
n
R
ec
o
g
n
itio
n
(
K
h
a
ir
u
l A
n
a
m
)
487
Fig
u
r
e
3
.
T
h
e
p
lace
m
en
t o
f
t
h
e
elec
tr
o
d
es
T
h
e
d
ata
co
llectio
n
an
d
th
e
m
y
o
elec
tr
ic
p
atter
n
r
ec
o
g
n
it
io
n
p
r
o
ce
s
s
w
er
e
co
n
d
u
c
ted
u
s
i
n
g
t
h
e
Ma
tlab
2
0
1
2
b
in
s
talled
i
n
th
e
I
n
tel
C
o
r
e
i5
3
.
1
GHz
d
esk
to
p
co
m
p
u
ter
w
it
h
4
GB
R
A
M
r
u
n
n
i
n
g
o
n
W
i
n
d
o
w
s
7
o
p
er
atin
g
s
y
s
te
m
.
Di
g
ital
l
y
,
th
e
E
MG
d
ata
i
s
f
ilter
ed
u
s
i
n
g
a
b
an
d
-
p
ass
f
i
lter
w
h
ic
h
f
ilter
s
t
h
e
s
ig
n
al
s
i
n
t
h
e
f
r
eq
u
en
c
y
r
an
g
e
b
et
w
ee
n
2
0
an
d
5
0
0
Hz
an
d
a
n
o
tch
f
i
lt
er
w
as
u
s
ed
as
w
e
ll
to
r
e
m
o
v
e
th
e
5
0
Hz
li
n
e
in
ter
f
er
e
n
ce
.
T
h
e
co
llected
d
at
a
w
as d
o
w
n
-
s
a
m
p
led
to
1
0
0
0
Hz.
I
n
th
i
s
p
ap
er
,
th
e
e
x
p
er
i
m
e
n
t
co
n
s
id
er
ed
ten
clas
s
es
o
f
t
h
e
i
n
d
iv
id
u
al
an
d
co
m
b
i
n
ed
f
i
n
g
er
m
o
v
e
m
e
n
ts
.
T
h
e
in
d
iv
id
u
al
f
i
n
g
er
s
co
n
s
is
t
o
f
th
e
f
le
x
io
n
o
f
th
u
m
b
(
T
)
,
in
d
ex
(
I
)
,
m
id
d
le
(
M)
,
r
in
g
(
R
)
,
an
d
litt
le
(
L
)
f
in
g
er
s
,
w
h
ile
t
h
e
co
m
b
in
ed
f
i
n
g
er
co
n
s
is
t
s
o
f
t
h
e
p
in
ch
i
n
g
o
f
t
h
u
m
b
a
n
d
i
n
d
ex
f
in
g
er
s
(
T
–
I
)
,
th
u
m
b
an
d
m
id
d
le
f
in
g
er
s
(
T
–
M
)
,
th
u
m
b
a
n
d
r
in
g
f
in
g
er
s
(
T
–
R
)
,
th
u
m
b
a
n
d
litt
le
f
i
n
g
er
s
(
T
–
L
)
,
an
d
clo
s
in
g
th
e
h
an
d
(
HC
)
.
Du
r
i
n
g
th
e
d
ata
c
o
llectio
n
,
th
e
s
u
b
j
ec
ts
w
er
e
as
k
ed
to
p
er
f
o
r
m
o
n
e
f
i
n
g
er
m
o
v
e
m
e
n
t
f
o
r
5
s
an
d
th
en
ta
k
e
a
r
est
f
o
r
5
s
.
T
h
e
s
u
b
j
ec
t
r
ep
ea
ted
ea
ch
m
o
v
e
m
e
n
t
s
i
x
ti
m
e
s
.
T
h
e
d
ata
co
llecte
d
w
er
e
d
iv
id
ed
in
to
tr
ain
i
n
g
d
ata
an
d
tes
tin
g
d
ata
u
s
i
n
g
3
-
f
o
ld
cr
o
s
s
v
alid
atio
n
.
I
n
th
e
ex
p
er
i
m
e
n
t
s
,
th
e
m
y
o
el
ec
tr
ic
p
atter
n
r
ec
o
g
n
itio
n
(
M
-
P
R
)
ex
tr
ac
ts
f
ea
t
u
r
es
o
f
w
a
v
e
f
o
r
m
le
n
g
t
h
(
W
L
)
,
s
lo
p
e
s
i
g
n
c
h
an
g
es
(
S
SC
)
,
n
u
m
b
er
o
f
ze
r
o
cr
o
s
s
i
n
g
s
(
Z
C
C
)
,
s
a
m
p
le
s
k
e
w
n
es
s
(
SS
)
,
m
ea
n
ab
s
o
lu
te
v
alu
e
(
M
A
V)
,
m
ea
n
ab
s
o
lu
te
v
alu
e
s
lo
p
e
(
MA
VS)
,
r
o
o
t
m
e
an
s
q
u
ar
e
(
R
MS)
,
s
o
m
e
p
ar
a
m
eter
s
f
r
o
m
Hj
o
r
th
ti
m
e
d
o
m
ain
p
ar
a
m
eter
s
(
HT
D)
an
d
6
-
o
r
d
er
au
to
r
eg
r
es
s
i
v
e
(
AR
6
)
m
o
d
el
p
ar
a
m
eter
s
ar
e
in
cl
u
d
ed
.
Mo
r
eo
v
er
,
SR
D
A
w
i
l
l
p
r
o
j
ec
t
an
d
r
e
d
u
ce
th
e
d
i
m
e
n
s
io
n
o
f
t
h
e
f
ea
t
u
r
e
ex
tr
ac
ted
.
T
h
e
ex
p
er
im
e
n
t
i
n
v
o
l
v
ed
th
e
s
tead
y
s
tate
s
ig
n
al
o
n
l
y
a
n
d
r
e
m
o
v
e
d
th
e
tr
a
n
s
ie
n
t
s
tate
o
f
th
e
m
y
o
elec
tr
ic
s
i
g
n
al.
T
h
e
m
aj
o
r
it
y
v
o
te
w
it
h
f
o
u
r
p
r
ev
io
u
s
s
ta
tes
m
a
y
b
e
u
s
ed
to
r
ef
in
e
t
h
e
clas
s
i
f
icatio
n
p
er
f
o
r
m
an
ce
.
As
d
ep
icted
in
Fi
g
u
r
e
2
,
P
SO
m
u
tated
b
y
w
a
v
elet
i
s
u
s
ed
to
o
p
tim
ize
th
e
p
ar
a
m
eter
s
o
f
r
a
d
ial
b
asis
f
u
n
ctio
n
ex
tr
e
m
e
lear
n
i
n
g
m
ac
h
in
e
(
R
B
F
-
E
L
M)
.
T
h
is
h
y
b
r
id
izatio
n
is
ca
lled
s
w
ar
m
-
w
a
v
elet
b
ased
R
B
F
-
E
L
M
o
r
SW
-
R
B
F
-
E
L
M.
So
m
e
p
ar
a
m
eter
s
s
h
o
u
ld
b
e
d
et
er
m
in
ed
at
th
e
b
e
g
i
n
n
i
n
g
o
f
t
h
e
ex
p
er
i
m
en
t.
T
w
o
p
ar
am
eter
s
o
f
R
B
F
-
E
L
M
ar
e
C
an
d
g
(
s
ee
Eq
u
atio
n
(4
))
.
T
h
e
y
ar
e
i
n
t
h
e
r
an
g
e
o
f
[
2
-
7
, 2
10
]
,
an
d
[
2
-
7
, 2
10
]
f
o
r
C
an
d
g
,
r
esp
ec
ti
v
el
y
.
T
h
en
,
t
h
e
p
ar
a
m
eter
s
o
f
P
SO
(
s
ee
E
q
u
atio
n
(7
)
an
d
E
q
u
a
tio
n
(8
)
)
ar
e
s
et
a
s
f
o
llo
w
s
.
P
ar
am
eter
c
1
an
d
c
2
ar
e
s
et
a
t
2
.
0
5
,
an
d
is
0
.
9
.
P
ar
am
et
er
s
r
1
an
d
r
2
ar
e
r
an
d
o
m
f
u
n
c
tio
n
s
i
n
t
h
e
r
an
g
e
o
f
[
0
-
1
]
.
I
n
ad
d
itio
n
,
th
e
o
p
tim
izatio
n
w
as
d
o
n
e
u
n
ti
l
1
5
0
g
en
er
atio
n
s
w
er
e
co
m
p
leted
w
it
h
3
0
p
ar
ticles
in
ea
ch
g
en
er
atio
n
.
As
f
o
r
t
h
e
p
ar
am
eter
o
f
th
e
w
av
e
let,
t
h
e
w
o
r
k
i
n
th
i
s
s
ec
tio
n
w
ill
v
a
r
y
t
h
e
v
al
u
e
o
f
t
h
e
w
a
v
elet
p
ar
a
m
eter
s
,
a
s
s
ee
n
in
E
q
u
a
tio
n
(9
)
a
n
d
E
q
u
atio
n
(
1
0
)
)
ex
ce
p
t
f
o
r
α
;
it
is
d
et
er
m
in
ed
r
a
n
d
o
m
l
y
,
ac
co
r
d
in
g
to
[
3
3
]
.
T
o
test
th
e
ef
f
icac
y
o
f
th
e
p
r
o
p
o
s
ed
s
y
s
te
m
,
s
o
m
e
ex
p
er
i
m
e
n
t
s
w
ill b
e
co
n
d
u
cted
.
T
h
e
y
ar
e:
a.
T
h
e
ex
p
er
im
e
n
t o
n
t
h
e
i
n
f
lu
e
n
ce
o
f
th
e
m
u
ta
tio
n
p
r
o
b
ab
ilit
y
p
m
b.
T
h
e
ex
p
er
im
e
n
t o
n
t
h
e
s
h
ap
e
p
ar
am
eter
(
E
q
ua
tio
n
(
11)
)
c.
T
h
e
ex
p
er
im
e
n
t o
n
t
h
e
p
ar
a
m
e
ter
g
(Eq
u
atio
n
(
1
1
))
d.
T
h
e
ex
p
er
im
e
n
t o
n
t
h
e
p
atter
n
r
ec
o
g
n
itio
n
p
er
f
o
r
m
an
ce
3.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
3
.
1
.
E
x
peri
m
e
nt
o
n t
he
Able
-
bo
d
ied S
ub
j
ec
t
s
3
.
1
.
1
.
M
ut
a
t
io
n P
ro
ba
bil
it
y
p
m
T
h
is
s
ec
tio
n
test
ed
t
h
e
in
f
l
u
e
n
ce
o
f
th
e
m
u
tatio
n
p
r
o
b
ab
ilit
y
p
m
to
th
e
SW
-
R
B
F
-
E
L
M
p
er
f
o
r
m
a
n
ce
.
T
h
e
p
m
v
alu
e
is
v
ar
ied
f
r
o
m
0
to
0
.
6
.
T
h
e
p
ar
am
eter
p
m
=
0
m
ea
n
s
n
o
w
a
v
elet
m
u
tatio
n
i
n
th
e
P
SO.
B
esid
e
s
,
i
s
eq
u
al
to
0
.
2
an
d
g
is
eq
u
al
to
1
0
0
0
0
.
T
h
e
ex
p
er
im
e
n
tal
r
esu
lt
s
ar
e
p
r
esen
ted
in
F
ig
u
r
e
4
.
Fig
u
r
e
4
in
d
icate
s
t
h
at
o
n
th
e
p
a
r
am
eter
p
m
=
0
,
t
h
e
f
it
n
ess
v
alu
e
o
f
t
h
e
P
SO
i
s
lar
g
er
t
h
a
n
t
h
at
w
it
h
p
m
m
o
r
e
t
h
an
0
,
e
v
en
w
h
en
it
is
th
e
lar
g
e
s
t
v
al
u
e.
T
h
e
lo
w
er
th
e
f
i
tn
e
s
s
v
al
u
e,
th
e
b
etter
th
e
s
y
s
te
m
,
s
o
t
h
e
P
SO
w
it
h
w
a
v
elet
m
u
tatio
n
i
s
b
etter
th
a
n
w
it
h
o
u
t
w
a
v
elet
m
u
tatio
n
.
T
h
er
ef
o
r
e,
t
h
e
w
a
v
elet
m
u
tatio
n
ca
n
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
8
,
No
.
1
,
Feb
r
u
ar
y
2
0
1
8
:
4
8
3
–
4
9
6
488
en
h
a
n
ce
th
e
o
p
ti
m
izatio
n
p
r
o
ce
s
s
.
Mo
r
eo
v
er
,
in
g
en
er
al,
t
h
e
f
i
g
u
r
e
also
s
h
o
w
s
t
h
at
t
h
e
m
o
r
e
th
e
m
u
tatio
n
p
r
o
b
a
b
ilit
y
,
th
e
le
s
s
t
h
e
f
i
tn
e
s
s
v
alu
e.
Ho
w
e
v
er
,
th
e
p
m
=
0
.
5
is
th
e
o
p
ti
m
u
m
v
alu
e
a
m
o
n
g
t
h
e
test
ed
v
a
lu
e
s
.
Fig
u
r
e
4
.
T
h
e
f
itn
es
s
v
al
u
es
f
o
r
v
ar
iab
le
p
m
w
h
e
n
=0
.
2
an
d
g
=1
0
0
0
0
o
v
er
eig
h
t su
b
j
ec
ts
T
ab
le
1
g
iv
es
m
o
r
e
i
n
f
o
r
m
at
io
n
r
eg
ar
d
in
g
th
e
m
u
tatio
n
p
r
o
b
ab
ilit
y
p
m
ac
r
o
s
s
d
if
f
er
en
t
s
u
b
j
ec
ts
.
In
T
ab
le
1
,
th
e
u
n
d
er
li
n
ed
v
al
u
e
in
d
icate
s
th
e
m
in
i
m
u
m
v
alu
e
f
o
r
ea
ch
s
u
b
j
ec
t.
T
h
is
tab
le
e
m
p
h
a
s
izes
th
e
f
ac
t
in
Fig
u
r
e
4
t
h
at
p
m
=
0
.
5
is
th
e
m
o
s
t
ac
cu
r
ate
P
SO
ac
r
o
s
s
s
ev
en
s
u
b
j
ec
ts
,
o
u
t
o
f
ei
g
h
t.
A
l
th
o
u
g
h
t
h
e
ac
c
u
r
ac
y
o
f
t
h
e
p
ar
a
m
eter
p
m
=0
.
6
is
t
h
e
h
i
g
h
est,
it
o
cc
u
r
r
ed
i
n
f
i
v
e
s
u
b
j
ec
ts
o
n
l
y
.
An
o
t
h
er
in
ter
esti
n
g
f
ac
t
i
s
al
s
o
f
o
u
n
d
in
t
h
e
T
ab
le.
T
h
e
m
u
tatio
n
w
a
v
elet
d
o
es
n
o
t
p
r
o
v
id
e
a
b
en
ef
it
to
th
e
o
p
ti
m
izat
io
n
p
r
o
ce
s
s
o
n
t
w
o
s
u
b
j
ec
ts
,
S5
an
d
S8
b
ec
au
s
e
t
h
e
ac
cu
r
ac
y
o
f
t
h
e
s
y
s
te
m
w
it
h
w
a
v
elet
m
u
tatio
n
an
d
w
i
th
o
u
t
i
s
v
er
y
s
i
m
ilar
.
T
h
is
f
ac
t
s
h
o
w
s
t
h
at
t
h
e
w
a
v
elet
m
u
ta
tio
n
i
n
t
h
e
P
SO
d
o
es
n
o
t
f
u
l
l
y
e
n
s
u
r
e
t
h
e
i
m
p
r
o
v
e
m
en
t
i
n
t
h
e
cla
s
s
i
f
icatio
n
p
er
f
o
r
m
a
n
ce
.
Ho
w
e
v
er
,
th
er
e
i
s
a
h
i
g
h
p
r
o
b
ab
ili
ty
th
at
th
e
o
p
ti
m
izat
io
n
p
r
o
ce
s
s
w
il
l b
e
i
m
p
r
o
v
ed
.
Fin
all
y
,
th
e
p
ar
am
eter
p
m
=
0
.
5
is
s
elec
ted
f
o
r
th
e
r
est o
f
t
h
e
ex
p
er
i
m
en
t.
T
ab
le
1
.
T
h
e
ac
cu
r
ac
y
o
f
SW
-
R
B
F
-
E
L
M
w
h
e
n
=0
.
2
an
d
g
=1
0
0
0
0
u
s
in
g
3
-
f
o
ld
cr
o
s
s
v
ali
d
atio
n
S
u
b
j
e
c
t
M
u
t
a
t
i
o
n
p
a
r
a
me
t
e
r
(
A
c
c
u
r
a
c
y
i
n
%)
0
0
.
1
0
.
2
0
.
3
0
.
4
0
.
5
0
.
6
S1
9
2
.
2
7
8
9
2
.
4
1
7
9
2
.
4
1
7
9
2
.
8
6
9
9
2
.
8
6
9
9
2
.
8
6
9
9
2
.
8
6
9
S2
9
8
.
0
9
8
9
8
.
0
9
8
9
8
.
0
2
8
9
8
.
0
2
8
9
8
.
0
2
8
9
8
.
1
2
9
9
8
.
0
9
8
S3
9
5
.
0
7
0
9
5
.
0
7
0
9
5
.
0
7
0
9
5
.
1
3
9
9
5
.
1
3
9
9
5
.
4
4
0
9
5
.
5
4
6
S4
9
3
.
2
4
0
9
3
.
2
3
8
9
3
.
2
3
8
9
3
.
3
4
4
9
3
.
3
4
4
9
3
.
3
4
4
9
3
.
3
1
0
S5
9
6
.
7
3
1
9
6
.
6
6
0
9
6
.
6
6
0
9
6
.
7
3
1
9
6
.
6
6
0
9
6
.
7
3
1
9
6
.
7
3
1
S6
9
7
.
0
8
8
9
7
.
2
1
5
9
7
.
2
1
5
9
7
.
2
1
5
9
7
.
2
1
5
9
7
.
2
5
0
9
7
.
2
5
0
S7
9
3
.
8
9
8
9
4
.
1
0
6
9
4
.
1
0
6
9
3
.
9
6
7
9
4
.
0
0
5
9
4
.
0
3
8
9
4
.
0
0
4
S8
9
7
.
8
8
0
9
7
.
8
8
0
9
7
.
8
8
0
9
7
.
8
8
0
9
7
.
8
8
0
9
7
.
8
8
0
9
7
.
8
8
0
A
v
e
r
a
g
e
9
5
.
5
3
5
9
5
.
5
8
5
9
5
.
5
7
7
9
5
.
6
4
7
9
5
.
6
4
3
9
5
.
7
1
0
9
5
.
7
1
1
*
T
h
e
u
n
d
e
r
l
i
n
e
d
v
a
l
u
e
i
s
t
h
e
h
i
g
h
e
st
o
n
e
T
h
e
h
i
g
h
er
v
al
u
e
o
f
t
h
e
p
ar
am
eter
p
m
in
cr
ea
s
e
s
th
e
s
ea
r
ch
in
g
s
p
ac
e
o
f
th
e
o
p
ti
m
izat
io
n
in
P
SO.
I
f
th
e
n
u
m
b
er
o
f
ele
m
e
n
ts
in
a
p
ar
ticle
is
s
m
all,
i
t
is
p
r
ef
er
ab
l
e
to
in
cr
ea
s
e
t
h
e
v
al
u
e
o
f
th
e
p
ar
am
eter
.
Fig
u
r
e
4
i
m
p
lies
t
h
at
t
h
e
h
i
g
h
er
v
alu
e
o
f
p
m
te
n
d
s
to
g
iv
e
g
o
o
d
o
p
tim
izatio
n
p
er
f
o
r
m
a
n
ce
.
T
h
i
s
p
h
en
o
m
en
a
m
atc
h
es
w
it
h
th
e
f
ac
t
s
u
g
g
ested
b
y
L
i
n
g
et
al
.
[
2
9
]
.
T
h
e
y
r
ec
o
m
m
en
d
ed
a
h
i
g
h
er
v
al
u
e
o
f
p
m
in
b
et
w
ee
n
0
.
5
–
0
.
8
f
o
r
a
s
m
al
l n
u
m
b
er
o
f
ele
m
e
n
ts
i
n
a
p
ar
ticle.
I
n
th
is
r
esear
ch
,
t
h
e
n
u
m
b
er
o
f
ele
m
en
t
s
is
t
w
o
.
T
o
ex
am
i
n
e
t
h
e
b
en
e
f
it
o
f
w
a
v
elet
m
u
tatio
n
s
tati
s
ticall
y
,
an
an
al
y
s
is
o
f
v
ar
ian
ce
(
A
NO
V
A
)
te
s
t
w
as
co
n
d
u
ct
ed
o
n
t
h
e
f
itn
e
s
s
v
al
u
e
o
f
th
e
P
SO
w
i
th
o
u
t
w
a
v
elet
m
u
tatio
n
a
n
d
w
it
h
w
a
v
elet
m
u
tatio
n
p
m
=
0
.
5
.
T
h
e
co
n
f
id
e
n
ce
le
v
el
p
i
s
s
et
at
0
.
0
5
.
A
NOV
A
test
p
r
o
d
u
ce
d
p
=
3
.
6
9
x
1
0
-
7
.
T
h
is
r
es
u
lt
co
n
cl
u
d
es
t
h
at
t
h
e
en
h
a
n
ce
m
en
t p
r
o
d
u
ce
d
b
y
w
a
v
elet
m
u
ta
tio
n
i
s
s
tati
s
tic
all
y
s
ig
n
if
ican
t.
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:
2
0
8
8
-
8708
Op
timiz
ed
K
ern
el
E
xt
r
eme
Le
a
r
n
in
g
Ma
ch
in
e
f
o
r
Myo
elec
tr
ic
P
a
tter
n
R
ec
o
g
n
itio
n
(
K
h
a
ir
u
l A
n
a
m
)
489
3
.
1
.
2
.
Sh
a
pe
-
pa
ra
m
et
er
T
h
is
s
ec
tio
n
v
ar
ied
t
h
e
v
al
u
e
o
f
s
h
ap
e
p
ar
a
m
eter
in
E
q
u
at
io
n
(
1
1
)
.
T
h
e
s
h
ap
e
p
ar
am
eter
is
v
ar
ied
a
m
o
n
g
0
.
1
,
0
.
2
,
0
.
3
,
0
.
5
,
2
a
n
d
5
.
T
h
e
v
a
l
u
e
o
f
th
e
p
ar
am
eter
p
m
is
0
.
5
f
o
llo
w
in
g
th
e
r
esu
l
t
i
n
s
ec
tio
n
0
.
Fu
r
t
h
er
m
o
r
e,
g
is
eq
u
al
to
1
0
0
0
0
.
T
h
e
ex
p
er
im
e
n
tal
r
es
u
lt i
s
p
r
esen
ted
in
Fi
g
u
r
e
5
.
Fig
u
r
e
5
in
d
icate
s
t
h
at
=
2
c
o
n
v
er
g
ed
ea
r
lier
t
h
a
n
th
e
o
th
e
r
s
d
id
.
T
h
e
f
in
al
f
it
n
es
s
v
alu
e
o
f
it
i
s
th
e
s
ec
o
n
d
w
o
r
s
t
af
ter
=
5
.
On
th
e
o
th
er
h
an
d
,
t
h
e
s
m
all
v
alu
e
o
f
g
a
v
e
a
g
o
o
d
o
p
ti
m
izati
o
n
p
r
o
ce
s
s
.
T
h
ese
f
ac
ts
i
m
p
l
y
t
h
at
t
h
e
h
i
g
h
v
al
u
e
o
f
is
n
o
t
a
g
o
o
d
o
p
tio
n
f
o
r
o
p
ti
m
izatio
n
o
f
SW
-
RBF
-
E
L
M.
T
h
e
b
est
o
p
tim
izatio
n
p
r
o
ce
s
s
is
s
h
o
w
n
w
h
en
=
0
.
2
.
Fig
u
r
e
5
.
T
h
e
f
itn
es
s
v
al
u
es
f
o
r
v
ar
iab
le
w
h
e
n
p
m
=
0
.
5
an
d
g
=
1
0
0
0
0
o
v
er
eig
h
t s
u
b
j
ec
t
s
T
ab
le
2
d
r
aw
s
d
if
f
er
e
n
t
f
in
d
i
n
g
f
r
o
m
Fi
g
u
r
e
5
.
T
h
e
tab
le
s
h
o
w
s
th
a
t
SW
-
R
B
F
-
E
L
M
w
i
th
=
0
.
1
ac
h
iev
ed
th
e
h
ig
h
est
av
er
a
g
e
ac
cu
r
ac
y
,
n
o
t
=
0
.
2
.
B
esi
d
es,
it
attai
n
s
t
h
e
h
i
g
h
e
s
t
ac
cu
r
ac
y
ac
r
o
s
s
f
o
u
r
s
u
b
j
ec
ts
,
w
h
ic
h
is
s
i
m
ilar
to
=
0
.
2
.
B
y
co
n
s
id
er
in
g
th
e
f
i
tn
es
s
v
al
u
e
an
d
t
h
e
av
er
a
g
e
ac
cu
r
ac
y
p
er
f
o
r
m
ed
,
=
0
.
2
is
s
elec
ted
as th
e
o
p
tim
al
s
h
ap
e
p
ar
a
m
eter
.
T
ab
le
2
.
T
h
e
ac
cu
r
ac
y
o
f
SW
-
R
B
F
-
E
L
M
w
h
e
n
p
m
=0
.
5
an
d
g
=1
0
0
0
0
u
s
in
g
3
-
f
o
ld
cr
o
s
s
v
ali
d
atio
n
S
u
b
j
e
c
t
(
A
c
c
u
r
a
c
y
i
n
%)
0
.
1
0
.
2
0
.
3
0
.
5
2
5
S1
9
2
.
8
6
9
9
2
.
8
6
9
9
2
.
8
6
9
9
2
.
8
6
9
9
2
.
8
6
9
9
2
.
8
6
9
S2
9
8
.
0
2
8
9
8
.
1
2
9
9
8
.
0
2
8
9
8
.
0
9
8
9
8
.
0
2
8
9
8
.
0
2
8
S3
9
5
.
8
9
3
9
5
.
4
4
0
9
5
.
8
9
3
9
5
.
1
3
9
9
5
.
0
7
0
9
5
.
1
3
9
S4
9
3
.
3
1
0
9
3
.
3
4
4
9
3
.
3
1
0
9
3
.
3
4
4
9
3
.
2
4
0
9
3
.
3
0
9
S5
9
6
.
7
3
1
9
6
.
7
3
1
9
6
.
6
6
0
9
6
.
7
3
1
9
6
.
6
6
0
9
6
.
7
3
1
S6
9
7
.
3
2
1
9
7
.
2
5
0
9
7
.
2
5
0
9
7
.
2
5
0
9
7
.
2
1
5
9
7
.
1
2
3
S7
9
4
.
1
0
6
9
4
.
0
3
8
9
4
.
0
0
2
9
4
.
0
0
4
9
3
.
8
9
8
9
3
.
8
9
8
S8
9
7
.
8
4
5
9
7
.
8
8
0
9
7
.
8
8
0
9
7
.
8
4
5
9
7
.
8
4
5
9
7
.
8
4
5
A
v
e
r
a
g
e
9
5
.
7
6
3
9
5
.
7
1
0
9
5
.
7
3
7
9
5
.
6
6
0
9
5
.
6
0
3
9
5
.
6
1
8
*
T
h
e
u
n
d
e
r
l
i
n
e
d
v
a
l
u
e
i
s
t
h
e
h
i
g
h
e
st
o
n
e
3
.
1
.
3
.
P
a
ra
m
et
er
g
T
h
e
p
r
ev
io
u
s
t
w
o
e
x
p
er
i
m
e
n
t
s
h
a
v
e
s
elec
ted
t
w
o
o
p
ti
m
u
m
p
ar
am
eter
s
,
p
m
=
0
.
5
an
d
=
0
.
2
.
T
h
is
s
ec
tio
n
tr
ies
to
g
et
th
e
o
p
ti
m
u
m
g
p
ar
a
m
eter
.
T
h
e
p
ar
am
e
ter
g
(Eq
u
atio
n
(
1
1
)
)
is
v
ar
ied
f
r
o
m
1
0
0
,
1
0
0
0
,
1
0
0
0
0
an
d
1
0
0
0
0
0
.
T
h
e
ex
p
er
im
en
tal
r
esu
lts
ar
e
p
r
esen
ted
i
n
Fig
u
r
e
6
.
Fig
u
r
e
6
d
ep
icts
th
e
f
it
n
ess
v
alu
e
s
o
f
f
o
u
r
d
if
f
er
en
t
g
v
a
lu
es.
T
h
is
f
ig
u
r
e
in
d
icate
s
th
at
th
e
b
ig
n
u
m
b
er
o
f
g
v
al
u
e
g
iv
e
b
etter
ac
cu
r
ac
y
t
h
a
n
t
h
e
s
m
al
l o
n
e.
T
h
e
g
=
1
0
0
0
0
ex
h
ib
its
th
e
b
est
p
er
f
o
r
m
an
ce
.
T
h
is
f
ac
t is
s
u
p
p
o
r
ted
b
y
t
h
e
ac
cu
r
a
c
y
o
f
SW
-
R
B
F
-
E
L
M
in
T
ab
le
3
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
8
,
No
.
1
,
Feb
r
u
ar
y
2
0
1
8
:
4
8
3
–
4
9
6
490
Fig
u
r
e
6
.
T
h
e
f
itn
es
s
v
al
u
es
f
o
r
v
ar
iatio
n
o
f
t
h
e
p
ar
a
m
eter
g
w
h
e
n
p
m
=0
.
5
an
d
=0
.
2
o
v
er
e
ig
h
t s
u
b
j
ec
ts
A
lt
h
o
u
g
h
th
e
ac
c
u
r
ac
y
o
f
t
h
e
p
ar
am
eter
g
=
1
0
0
0
0
is
t
h
e
lo
w
e
s
t
o
n
e
o
n
av
er
a
g
e
ac
r
o
s
s
eig
h
t
s
u
b
j
ec
ts
,
it
i
s
t
h
e
h
i
g
h
e
s
t
i
n
t
h
e
o
v
er
h
al
f
o
f
t
h
e
s
u
b
j
ec
ts
,
w
h
ich
i
s
f
i
v
e
o
u
t
o
f
ei
g
h
t.
T
h
ese
r
esu
lt
s
co
n
f
ir
m
t
h
e
r
ec
o
m
m
e
n
d
atio
n
o
f
L
i
n
g
et
al
.
[
2
9
]
.
T
h
e
y
f
o
u
n
d
t
h
at
b
y
s
et
tin
g
th
e
p
ar
a
m
eter
g
in
t
h
e
h
i
g
h
v
al
u
e,
th
e
o
t
h
er
p
ar
am
eter
co
u
ld
b
e
ch
o
s
e
n
b
y
tr
ial
an
d
er
r
o
r
.
T
ab
le
3
.
T
h
e
ac
cu
r
ac
y
o
f
SW
-
R
B
F
-
E
L
M
w
h
e
n
p
m
=0
.
5
an
d
=
0
.
2
u
s
in
g
3
-
f
o
ld
cr
o
s
s
v
a
lid
atio
n
S
u
b
j
e
c
t
P
a
r
a
me
t
e
r
g
(
A
c
c
u
r
a
c
y
i
n
%)
1
0
0
1
0
0
0
1
0
0
0
0
1
0
0
0
0
0
S1
9
2
.
8
6
9
9
2
.
8
6
9
9
2
.
8
3
4
9
2
.
8
0
0
S2
9
8
.
0
2
8
9
8
.
0
9
8
9
8
.
1
2
9
9
8
.
1
2
9
S3
9
5
.
7
3
2
9
5
.
8
9
3
9
5
.
4
4
0
9
5
.
5
4
6
S4
9
3
.
3
4
7
9
3
.
3
1
0
9
3
.
3
4
4
9
3
.
2
3
8
S5
9
6
.
7
3
1
9
6
.
6
6
0
9
6
.
7
3
1
9
6
.
7
3
1
S6
9
7
.
2
5
0
9
7
.
2
1
5
9
7
.
2
5
0
9
7
.
2
1
5
S7
9
4
.
0
3
8
9
4
.
0
0
4
9
4
.
0
3
8
9
4
.
1
0
6
S8
9
7
.
8
4
5
9
7
.
8
4
5
9
7
.
8
8
0
9
7
.
8
8
0
A
v
e
r
a
g
e
9
5
.
7
3
0
9
5
.
7
3
7
9
5
.
7
0
6
9
5
.
7
0
6
*
T
h
e
u
n
d
e
r
l
i
n
e
d
v
a
l
u
e
i
s
t
h
e
h
i
g
h
e
st
o
n
e
3
.
1
.
4
.
P
a
t
t
er
n Re
co
g
nitio
n P
er
f
o
r
m
a
nce
a
cr
o
s
s
S
ub
j
ec
t
s
T
h
e
p
r
ev
io
u
s
s
ec
tio
n
s
co
n
d
u
cted
s
o
m
e
ex
p
er
i
m
en
ts
to
d
eter
m
i
n
e
t
h
e
o
p
ti
m
u
m
p
ar
a
m
e
ter
s
o
f
t
h
e
w
a
v
elet.
T
h
e
y
ar
e
p
m
=0
.
5
,
=
0
.
2
an
d
g
=
1
0
0
0
0
.
T
h
is
s
ec
tio
n
ap
p
lied
th
o
s
e
v
al
u
es
to
SW
-
R
B
F
-
E
L
M
a
n
d
d
id
an
al
y
s
is
o
n
t
h
e
r
es
u
lts
e
s
p
ec
iall
y
o
n
th
e
co
m
p
ar
is
o
n
b
et
w
ee
n
P
SO
w
it
h
w
a
v
elet
m
u
tatio
n
an
d
w
i
th
o
u
t
m
u
tatio
n
.
T
h
e
r
esu
l
t is s
h
o
w
n
in
Fi
g
u
r
e
7
.
Fig
u
r
e
7
.
T
h
e
ac
cu
r
ac
y
o
f
R
B
F
-
E
L
M
w
it
h
m
u
tatio
n
a
n
d
w
i
t
h
o
u
t
m
u
tatio
n
u
s
i
n
g
3
-
f
o
ld
cr
o
s
s
v
a
lid
atio
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:
2
0
8
8
-
8708
Op
timiz
ed
K
ern
el
E
xt
r
eme
Le
a
r
n
in
g
Ma
ch
in
e
f
o
r
Myo
elec
tr
ic
P
a
tter
n
R
ec
o
g
n
itio
n
(
K
h
a
ir
u
l A
n
a
m
)
491
Fig
u
r
e
7
d
ep
icts
th
e
a
v
er
a
g
e
a
cc
u
r
ac
y
o
f
r
ad
ial
b
asis
f
u
n
ct
io
n
E
L
M
(
R
B
F
-
E
L
M)
th
a
t is o
p
ti
m
ized
b
y
P
SO
w
it
h
w
a
v
elet
m
u
tatio
n
(
SW
-
R
B
F
-
E
L
M)
an
d
w
ith
o
u
t
m
u
tatio
n
(
S
R
B
F
-
E
L
M)
.
T
h
e
f
ig
u
r
e
i
n
d
icate
s
th
a
t
SW
-
R
B
F
-
E
L
M
ac
h
ie
v
es
b
ette
r
ac
cu
r
ac
y
t
h
a
n
S
R
B
F
-
E
L
M
a
cr
o
s
s
s
e
v
en
s
u
b
j
ec
ts
.
SR
B
F
-
E
L
M
i
s
as
ac
cu
r
ate
as
SW
-
R
B
F
-
E
L
M
i
n
o
n
e
s
u
b
j
ec
t
o
n
l
y
,
w
h
ic
h
is
s
u
b
j
ec
t
S8
.
T
h
er
ef
o
r
e,
th
e
p
r
o
b
a
b
ili
t
y
o
f
t
h
e
i
m
p
r
o
v
e
m
e
n
t
o
f
th
e
p
er
f
o
r
m
an
ce
u
s
i
n
g
w
a
v
el
et
m
u
tat
io
n
i
s
7
/8
x
1
0
0
%
=
8
7
.
5
%.
On
av
er
a
g
e,
SW
-
RBF
-
E
L
M
attai
n
ed
a
n
ac
cu
r
ac
y
o
f
9
5
.
7
1
%
w
h
ile
S
R
B
F
-
E
L
M
ac
h
ie
v
ed
th
e
ac
c
u
r
ac
y
o
f
9
5
.
5
4
%.
3
.
1
.
5
.
P
a
t
t
er
n Re
co
g
nitio
n P
er
f
o
r
m
a
nce
o
n t
he
M
o
v
em
e
nt
T
h
is
s
ec
tio
n
i
n
v
e
s
ti
g
ates
th
e
p
er
f
o
r
m
an
ce
o
f
b
o
th
s
y
s
te
m
s
,
SR
B
F
-
E
L
M
a
n
d
SW
-
RBF
-
E
L
M,
i
n
class
i
f
y
in
g
f
in
g
er
m
o
v
e
m
e
n
ts
.
T
h
e
m
y
o
elec
tr
ic
p
atter
n
r
ec
o
g
n
itio
n
class
if
ie
s
ten
f
in
g
er
m
o
v
e
m
e
n
ts
.
T
h
e
y
in
cl
u
d
e
th
u
m
b
(
T
)
,
in
d
ex
(
I
)
,
m
id
d
le
(
M)
,
r
in
g
(
R
)
,
an
d
litt
l
e
(
L
)
f
in
g
er
m
o
v
e
m
e
n
t
s
.
T
h
e
o
th
er
m
o
v
e
m
e
n
ts
ar
e
th
u
m
b
–
in
d
e
x
(
T
I
)
,
th
u
m
b
–
m
i
d
d
le
(
T
M)
,
th
u
m
b
–
r
i
n
g
(
T
R
)
,
th
u
m
b
–
litt
le
(
T
L
)
,
an
d
t
h
e
h
a
n
d
clo
s
e
(
H
C
)
m
o
v
e
m
e
n
ts
.
Fi
g
u
r
e
8
p
r
esen
ts
th
e
clas
s
i
f
icatio
n
r
es
u
lt
s
o
f
S
R
B
F
-
E
L
M
(
w
it
h
o
u
t
w
a
v
ele
t
m
u
tatio
n
)
a
n
d
SW
-
R
B
F
-
E
L
M
(
w
it
h
w
a
v
elet
m
u
t
atio
n
)
.
Fig
u
r
e
8
s
h
o
w
s
th
at
SW
-
R
B
F
-
E
L
M
is
b
etter
th
an
S
R
B
F
-
E
L
M
in
cla
s
s
i
f
y
in
g
t
w
o
in
d
iv
i
d
u
al
f
i
n
g
er
m
o
v
e
m
e
n
ts
(
T
,
an
d
M)
,
an
d
f
o
u
r
co
m
b
i
n
ed
m
o
v
e
m
e
n
ts
(
T
I
,
T
M,
T
R
,
an
d
T
L
)
.
On
th
e
o
th
er
h
an
d
,
SR
B
F
-
E
L
M
is
b
etter
t
h
a
n
SW
-
R
B
F
-
E
L
M
i
n
t
w
o
m
o
v
e
m
en
t
s
o
n
l
y
:
L
an
d
H
C
.
A
s
f
o
r
f
i
n
g
er
m
o
v
e
m
en
t
I
an
d
R
,
b
o
t
h
s
y
s
te
m
s
e
x
h
ib
ited
a
s
i
m
ilar
p
er
f
o
r
m
an
ce
.
Ov
er
all,
th
e
SW
-
R
B
F
-
E
L
M
i
s
b
etter
t
h
a
n
S
R
B
F
-
E
L
M.
I
n
o
th
er
w
o
r
d
s
,
t
h
e
w
a
v
elet
m
u
ta
tio
n
in
P
SO
e
n
h
a
n
ce
s
t
h
e
cla
s
s
i
f
icatio
n
p
er
f
o
r
m
a
n
ce
o
f
t
h
e
p
atter
n
r
ec
o
g
n
itio
n
s
y
s
te
m
.
Ho
w
ev
er
,
t
h
e
a
n
al
y
s
i
s
o
f
v
ar
ia
n
ce
te
s
t (
A
NO
V
A
)
s
et
p
=
0
.
0
5
y
ield
s
p
is
eq
u
al
to
0
.
9
6
.
T
h
er
ef
o
r
e,
th
e
i
m
p
r
o
v
e
m
en
t
is
s
tat
is
tical
l
y
n
o
t
s
ig
n
i
f
ica
n
t.
T
h
is
r
esu
lt
co
n
f
r
o
n
ts
t
h
e
A
NO
V
A
te
s
t
r
es
u
lt
in
Sectio
n
0
t
h
a
t
p
r
o
v
ed
th
e
s
ig
n
if
ica
n
ce
o
f
t
h
e
ex
is
te
n
ce
o
f
th
e
w
a
v
elet
in
P
SO.
T
h
ese
t
w
o
r
esu
lt
s
ca
n
b
e
ac
co
m
m
o
d
ated
b
y
s
a
y
i
n
g
th
at
t
h
e
en
h
an
ce
m
e
n
t
o
f
w
av
ele
t
m
u
tatio
n
i
n
th
e
o
p
ti
m
izatio
n
p
r
o
ce
s
s
is
s
tatis
ticall
y
s
i
g
n
i
f
ica
n
t,
b
u
t
it
is
n
o
t si
g
n
i
f
ican
t in
t
h
e
clas
s
i
f
i
ca
tio
n
p
er
f
o
r
m
a
n
ce
.
Fig
u
r
e
8
.
T
h
e
ac
cu
r
ac
y
o
f
t
h
e
f
i
n
g
er
m
o
v
e
m
en
t c
lass
if
icatio
n
ac
r
o
s
s
eig
h
t
s
u
b
j
ec
ts
u
s
in
g
3
-
f
o
ld
cr
o
s
s
v
alid
atio
n
T
ab
le
4
.
T
h
e
co
n
f
u
s
io
n
m
atr
ix
o
f
th
e
clas
s
i
f
icatio
n
r
es
u
lt o
f
SW
-
R
B
F
-
E
L
M
C
l
a
ssi
f
i
e
d
I
n
t
e
n
d
e
d
T
I
M
R
L
TI
TM
TR
TL
HC
T
9
7
.
2
7
0
.
0
0
0
.
0
4
0
.
0
0
0
.
5
2
0
.
6
5
0
.
0
0
0
.
6
1
0
.
0
0
0
.
9
1
I
0
.
0
4
9
9
.
3
5
0
.
0
0
0
.
0
0
0
.
0
0
0
.
4
8
0
.
1
3
0
.
0
0
0
.
0
0
0
.
0
0
M
0
.
0
0
0
.
0
0
9
9
.
6
6
0
.
0
0
0
.
0
0
0
.
0
0
0
.
3
4
0
.
0
0
0
.
0
0
0
.
0
0
R
0
.
0
0
0
.
0
0
0
.
0
9
9
9
.
1
3
0
.
1
7
0
.
0
0
0
.
0
0
0
.
6
1
0
.
0
0
0
.
0
0
L
0
.
0
0
1
.
1
9
0
.
0
0
1
.
8
4
9
1
.
4
0
2
.
1
9
0
.
7
9
0
.
5
3
1
.
8
9
0
.
1
8
TI
0
.
5
6
1
.
9
1
0
.
0
4
0
.
0
0
1
.
4
8
9
4
.
5
7
1
.
0
0
0
.
1
3
0
.
3
0
0
.
0
0
TM
0
.
0
0
0
.
7
4
0
.
2
2
0
.
2
6
1
.
4
8
4
.
3
0
9
1
.
5
8
1
.
2
2
0
.
2
2
0
.
0
0
TR
1
.
2
7
0
.
1
7
0
.
0
0
0
.
1
7
0
.
3
8
0
.
5
5
1
.
4
4
9
4
.
8
1
0
.
5
5
0
.
6
8
TL
0
.
3
0
0
.
0
0
0
.
2
2
0
.
1
3
2
.
6
5
1
.
1
3
0
.
0
0
1
.
0
4
9
4
.
5
3
0
.
0
0
HC
0
.
7
5
0
.
3
5
0
.
0
0
0
.
0
0
1
.
2
7
0
.
1
7
1
.
9
6
0
.
7
0
0
.
0
0
9
4
.
8
0
An
o
th
er
f
ac
t
f
o
u
n
d
i
n
Fi
g
u
r
e
8
is
th
at
SR
B
F
-
E
L
M
an
d
SW
-
R
B
F
-
E
L
M
ex
h
ib
it
r
elat
iv
el
y
b
ad
p
er
f
o
r
m
a
n
ce
i
n
clas
s
if
y
i
n
g
al
l
co
m
b
i
n
ed
m
o
v
e
m
e
n
t
s
an
d
l
ittl
e
f
i
n
g
er
m
o
v
e
m
e
n
t.
T
h
e
p
h
en
o
m
en
a
ca
n
b
e
in
v
e
s
ti
g
ated
th
r
o
u
g
h
th
e
co
n
f
u
s
io
n
m
atr
i
x
in
T
ab
le
4
an
d
Fig
u
r
e
9
.
T
ab
le
4
s
h
o
w
s
th
at
th
e
SW
-
R
B
F
-
E
L
M
m
o
s
tl
y
m
i
s
clas
s
i
f
ied
t
h
e
l
ittl
e
f
i
n
g
er
m
o
v
e
m
e
n
t
(
L
)
to
t
h
u
m
b
-
in
d
e
x
m
o
tio
n
(
T
I
)
w
ith
t
h
e
ac
cu
r
ac
y
o
f
2
.
1
9
%.
B
esid
es,
th
e
s
y
s
te
m
also
m
i
s
c
lass
i
f
ies
L
to
m
o
v
e
m
en
t
R
a
n
d
T
L
.
As
f
o
r
t
h
e
co
m
b
in
ed
m
o
v
e
m
e
n
t,
SW
-
R
B
F
-
E
L
M
g
e
n
er
all
y
m
is
cla
s
s
i
f
ied
th
e
m
to
th
e
i
n
d
iv
id
u
al
m
o
v
e
m
en
t
t
h
e
y
b
elo
n
g
to
.
Fo
r
i
n
s
t
an
ce
,
t
h
e
m
o
v
e
m
e
n
t
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
n
t J
E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
8
,
No
.
1
,
Feb
r
u
ar
y
2
0
1
8
:
4
8
3
–
4
9
6
492
T
L
is
m
o
s
tl
y
m
is
cla
s
s
i
f
ied
to
th
e
m
o
v
e
m
e
n
t
L
b
y
ac
c
u
r
ac
y
2
.
6
5
%.
Nev
er
t
h
eles
s
,
it
d
i
d
n
o
t
o
cc
u
r
in
a
ll
co
m
b
i
n
ed
m
o
v
e
m
en
ts
.
I
n
ad
d
itio
n
to
T
ab
le
4
,
Fig
u
r
e
9
h
elp
s
t
h
e
r
ea
d
er
th
e
g
et
a
v
is
u
al
g
r
ap
h
o
f
t
h
e
co
n
f
u
s
io
n
m
atr
i
x
.
Fig
u
r
e
9
.
T
h
e
co
n
f
u
s
io
n
m
a
tr
i
x
p
lo
t o
f
th
e
clas
s
i
f
icatio
n
r
es
u
lt o
f
SW
-
R
B
F
-
E
L
M
3
.
1
.
6
.
S
W
-
RB
F
-
E
L
M
a
nd
o
t
her
well
-
kn
o
w
n C
la
s
s
if
iers
I
n
th
is
e
x
p
er
i
m
e
n
t,
th
e
p
er
f
o
r
m
an
ce
o
f
SW
-
R
B
F
-
E
L
M
is
c
o
m
p
ar
ed
to
o
th
er
w
e
ll
-
k
n
o
w
n
class
if
ier
s
s
u
c
h
as
o
r
ig
i
n
al
E
L
M
u
s
i
n
g
s
i
g
m
o
id
ac
ti
v
atio
n
f
u
n
ctio
n
(
Sig
-
E
L
M)
,
SR
B
F
-
E
L
M,
SV
M,
L
DA
,
a
n
d
k
NN.
T
h
e
ex
p
er
i
m
e
n
tal
r
es
u
lts
ar
e
d
ep
icted
in
Fig
u
r
e
1
0
.
Fig
u
r
e
10
.
T
h
e
ac
cu
r
ac
y
o
f
S
W
-
R
B
F
-
E
L
M
an
d
o
th
er
w
ell
-
k
n
o
w
n
clas
s
i
f
ier
s
f
o
r
f
in
g
er
m
o
v
e
m
e
n
t r
ec
o
g
n
itio
n
u
s
i
n
g
3
-
f
o
ld
cr
o
s
s
v
alid
atio
n
Fig
u
r
e
1
0
s
h
o
w
s
t
h
at
SW
-
R
B
F
-
E
L
M
is
th
e
m
o
s
t a
cc
u
r
ate
c
l
ass
i
f
ier
a
m
o
n
g
s
e
v
en
d
if
f
er
en
t
class
if
ier
s
in
r
ec
o
g
n
iz
in
g
ten
f
i
n
g
er
m
o
v
e
m
en
ts
u
s
i
n
g
E
MG
c
h
a
n
n
els
ac
r
o
s
s
eig
h
t
ab
le
-
b
o
d
ied
s
u
b
j
e
cts.
T
h
is
f
in
d
i
n
g
is
s
u
p
p
o
r
ted
b
y
T
ab
le
5
th
at
p
r
es
en
ts
th
e
a
v
er
ag
e
ac
c
u
r
ac
y
ac
h
i
ev
ed
b
y
ea
c
h
clas
s
if
ier
.
SW
-
RBF
-
E
L
M
ac
h
iev
ed
th
e
ac
cu
r
ac
y
o
f
9
5
.
7
1
%.
Fu
r
th
er
m
o
r
e,
SW
-
R
B
F
-
E
L
M
ac
h
i
ev
ed
th
e
h
ig
h
es
t
ac
cu
r
ac
y
o
n
f
o
u
r
s
u
b
j
ec
ts
,
w
h
ile
it a
ttain
ed
t
h
e
s
ec
o
n
d
lo
w
est a
cc
u
r
ac
y
o
n
t
h
e
s
u
b
j
ec
t S3
an
d
S4
.
T
ab
le
5
.
T
h
e
ac
cu
r
ac
y
o
f
v
ar
io
u
s
cla
s
s
i
f
ier
s
f
o
r
th
e
f
in
g
er
m
o
v
e
m
e
n
t r
ec
o
g
n
itio
n
u
s
in
g
3
-
f
o
ld
cr
o
s
s
v
alid
atio
n
C
l
a
ssi
f
i
e
r
A
c
c
u
r
a
c
y
M
e
a
n
(
%)
S
T
D
S
i
g
-
E
L
M
9
5
.
1
0
2
.
2
5
RBF
-
E
L
M
9
5
.
0
6
2
.
2
1
S
R
B
F
-
E
L
M
9
5
.
5
4
2
.
2
3
SW
-
RBF
-
E
L
M
9
5
.
7
1
2
.
0
9
S
V
M
9
5
.
3
9
1
.
8
6
L
D
A
9
4
.
3
7
2
.
3
8
k
N
N
9
5
.
0
6
2
.
3
7
Evaluation Warning : The document was created with Spire.PDF for Python.