I
AE
S In
t
er
na
t
io
na
l J
o
urna
l o
f
Art
if
icia
l In
t
ellig
ence
(
I
J
-
AI
)
Vo
l.
14
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
5
,
p
p
.
769
~
779
I
SS
N:
2
2
5
2
-
8
9
3
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijai.v
14
.i
1
.
p
p
7
6
9
-
7
7
9
769
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
i
.
ia
esco
r
e.
co
m
Electroencep
ha
lo
g
ra
m deno
ising
using
discrete
wav
elet
trans
form a
nd ad
a
ptive no
ise ca
nce
lla
tion ba
sed o
n i
nform
a
tion
theory
H
a
s
hem
Abdo
la
hn
iy
a
1
,
Ali A
k
ba
r
K
ha
za
ei
1
,
M
a
hd
i A
za
rno
o
s
h
2
,
Sey
ed
E
h
s
a
n
Ra
za
v
i
1
1
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
,
M
a
s
h
h
a
d
B
r
a
n
c
h
,
I
sl
a
mi
c
A
z
a
d
U
n
i
v
e
r
si
t
y
,
M
a
s
h
h
a
d
,
I
r
a
n
2
D
e
p
a
r
t
me
n
t
o
f
B
i
o
m
e
d
i
c
a
l
E
n
g
i
n
e
e
r
i
n
g
,
M
a
s
h
h
a
d
B
r
a
n
c
h
,
I
sl
a
mi
c
A
z
a
d
U
n
i
v
e
r
s
i
t
y
,
M
a
sh
h
a
d
,
I
r
a
n
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
an
23
,
2
0
2
4
R
ev
is
ed
Sep
17
,
2
0
2
4
Acc
ep
ted
Sep
30
,
2
0
2
4
On
e
o
f
t
h
e
m
o
st
fre
q
u
e
n
tl
y
u
se
d
tec
h
n
iq
u
e
s
fo
r
re
m
o
v
in
g
b
a
c
k
g
ro
u
n
d
n
o
ise
fro
m
e
lec
tro
e
n
c
e
p
h
a
lo
g
ra
m
(EE
G)
d
a
ta
is
a
d
a
p
ti
v
e
n
o
ise
c
a
n
c
e
ll
a
ti
o
n
(AN
C).
No
n
e
th
e
les
s,
th
e
re
e
x
ist
tw
o
p
r
ima
ry
d
isa
d
v
a
n
tag
e
s
a
ss
o
c
iate
d
with
th
e
a
d
a
p
ti
v
e
n
o
ise
re
d
u
c
ti
o
n
o
f
EE
G
sig
n
a
ls:
t
h
e
a
d
a
p
ti
v
e
f
il
ter,
w
h
ich
is
su
p
p
o
se
d
to
b
e
a
n
a
p
p
r
o
x
ima
ti
o
n
o
f
c
o
n
tam
in
a
ted
n
o
ise
,
lac
k
s
t
h
e
re
fe
re
n
c
e
s
ig
n
a
l.
Th
e
m
e
a
n
sq
u
a
re
d
e
rro
r
(
M
S
E)
c
rit
e
r
io
n
is
fre
q
u
e
n
tl
y
e
m
p
lo
y
e
d
t
o
a
c
h
iev
e
th
is
g
o
a
l
i
n
a
d
a
p
ti
v
e
fil
ters
.
T
h
e
M
S
E
c
rit
e
rio
n
,
wh
ich
o
n
ly
c
o
n
si
d
e
rs
se
c
o
n
d
-
o
rd
e
r
e
rro
rs,
c
a
n
n
o
t
b
e
u
se
d
sin
c
e
n
e
it
h
e
r
th
e
EE
G
sig
n
a
l
n
o
r
t
h
e
EOG
a
rti
fa
c
t
a
re
G
a
u
ss
ian
.
In
th
is
wo
rk
,
we
e
m
p
lo
y
a
n
ANC
sy
ste
m
,
d
e
riv
i
n
g
a
n
e
stim
a
te
o
f
EOG
n
o
ise
with
a
d
isc
re
te
wa
v
e
le
t
tran
sfo
rm
(DWT
)
a
n
d
in
p
u
t
t
h
is
sig
n
a
l
in
to
th
e
re
fe
re
n
c
e
o
f
th
e
AN
C
s
y
ste
m
.
Th
e
e
n
tr
o
p
y
-
b
a
se
d
e
rro
r
m
e
tri
c
is
u
se
d
to
re
d
u
c
e
th
e
e
rro
r
sig
n
a
l
in
ste
a
d
o
f
t
h
e
MSE
.
Re
su
l
ts
fro
m
c
o
m
p
u
ter
sim
u
latio
n
s
d
e
m
o
n
stra
te
t
h
a
t
t
h
e
su
g
g
e
ste
d
sy
ste
m
o
u
t
p
e
rfo
rm
s
c
o
m
p
e
ti
n
g
m
e
th
o
d
s
wit
h
re
sp
e
c
t
to
ro
o
t
-
mean
-
sq
u
a
re
-
e
rr
o
r,
sig
n
a
l
-
to
-
n
o
ise
ra
ti
o
,
a
n
d
c
o
h
e
re
n
c
e
m
e
a
su
re
m
e
n
ts.
K
ey
w
o
r
d
s
:
Ad
ap
tiv
e
n
o
is
e
ca
n
ce
llatio
n
Dis
cr
ete
wav
elet
tr
an
s
f
o
r
m
E
r
r
o
r
e
n
tr
o
p
y
I
n
f
o
r
m
atio
n
th
e
o
r
y
Me
an
s
q
u
ar
ed
e
r
r
o
r
No
is
e
r
ed
u
ctio
n
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
:
Ali A
k
b
ar
Kh
az
ae
i
Dep
ar
tm
en
t o
f
E
lectr
ical
E
n
g
i
n
ee
r
in
g
,
Ma
s
h
h
ad
B
r
an
ch
,
I
s
lam
ic
Aza
d
Un
iv
er
s
ity
Ma
s
h
h
ad
,
I
r
a
n
E
m
ail:
k
h
az
ae
i@
m
s
h
d
iau
.
ac
.
ir
1.
I
NT
RO
D
UCT
I
O
N
E
lectr
o
en
ce
p
h
al
o
g
r
am
(
E
E
G)
s
ig
n
al
is
r
ef
er
r
ed
to
as
m
ea
s
u
r
em
en
ts
o
f
elec
tr
ical
ac
tiv
ity
o
f
th
e
b
r
ain
,
wh
ich
is
p
er
f
o
r
m
e
d
b
y
elec
tr
o
d
es
p
lace
d
o
v
er
th
e
s
ca
lp
.
B
lin
k
in
g
o
r
m
o
v
in
g
th
e
h
ea
d
wh
ile
r
ec
o
r
d
in
g
th
e
E
E
G
s
ig
n
al
ca
n
lead
to
co
n
tam
in
atio
n
o
f
th
e
s
ig
n
al
with
th
e
elec
tr
o
o
cu
lo
g
r
am
(
E
OG)
a
r
tifa
ct.
I
n
m
an
y
ap
p
licatio
n
s
u
s
ed
in
E
E
G,
lik
e
th
e
b
r
ain
-
co
m
p
u
ter
in
ter
f
ac
e
(
B
C
I
)
,
th
e
ex
is
ten
ce
o
f
ar
tifa
cts
with
i
n
th
e
E
E
G
s
ig
n
al
d
im
in
is
h
es
th
e
ef
f
icien
cy
o
f
th
e
s
y
s
tem
[
1
]
.
T
h
er
ef
o
r
e,
th
e
elim
in
atio
n
o
f
ar
tifa
cts
is
cr
u
cial
in
E
E
G
s
ig
n
al
an
aly
s
is
.
R
eg
r
ess
io
n
-
b
ased
tech
n
iq
u
es
f
in
d
ap
p
licatio
n
s
in
two
d
is
tin
ct
d
o
m
ain
s
,
tim
e
a
n
d
f
r
eq
u
en
c
y
[
2
]
,
[
3
]
,
f
o
r
o
m
itti
n
g
E
OG
ar
tifa
ct
o
f
E
E
G
s
ig
n
al,
wh
ich
is
r
ec
o
r
d
ed
f
o
r
lik
ely
p
r
o
v
o
k
e
d
s
tu
d
ies.
I
n
th
ese
m
eth
o
d
s
,
we
ca
n
u
s
e
E
E
G
d
ata
wh
ic
h
is
g
r
ad
ed
f
o
r
d
if
f
e
r
en
t
ca
u
s
es
f
o
r
o
b
tain
in
g
th
e
m
ea
n
co
ef
f
icie
n
ts
o
f
tr
a
n
s
m
is
s
io
n
with
in
ea
ch
E
E
G
ch
an
n
el
an
d
E
OG.
Ho
wev
e
r
,
with
in
th
e
p
r
o
ce
s
s
,
th
e
m
ea
n
tr
an
s
f
er
co
ef
f
i
cien
ts
f
o
r
r
em
o
v
i
n
g
E
OG
f
r
o
m
r
ea
l
-
tim
e
E
E
G
s
ig
n
als
is
in
s
u
f
f
icien
t
b
ec
au
s
e
E
OG
an
d
E
E
G
s
ig
n
als
ex
h
ib
it
d
y
n
am
ic
ch
ar
ac
ter
is
tics
.
Hen
ce
,
th
e
n
ec
ess
ity
f
o
r
ad
ap
t
iv
e
f
ilter
s
ar
is
es in
o
r
d
er
to
m
o
n
ito
r
th
e
d
y
n
am
ic
alter
atio
n
s
in
th
e
s
ig
n
al
[
4
]
.
T
h
e
u
tili
za
tio
n
o
f
a
d
ap
tiv
e
n
o
is
e
ca
n
ce
llatio
n
(
ANC)
in
b
io
m
ed
ical
s
ig
n
als
is
ex
ten
s
iv
e
[
5
]
–
[
8
]
.
Nev
er
th
eless
,
th
ese
m
eth
o
d
s
ty
p
ically
p
r
esu
p
p
o
s
e
th
e
av
ailab
ilit
y
o
f
a
r
ef
er
en
ce
d
s
i
g
n
al
f
o
r
ANC.
An
alter
n
ativ
e
ap
p
r
o
ac
h
em
p
lo
y
s
an
ad
ap
tiv
e
n
o
is
e
r
ed
u
ctio
n
m
eth
o
d
with
s
elf
-
co
n
tain
ed
co
m
p
o
n
en
t
an
aly
s
is
[
9
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
14
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
5
:
769
-
7
7
9
770
T
h
is
in
v
o
lv
es
th
e
ap
p
licatio
n
o
f
in
d
ep
e
n
d
en
t
co
m
p
o
n
e
n
t
an
aly
s
is
(
I
C
A)
o
n
E
E
G
s
ig
n
als
p
r
o
x
im
ate
to
th
e
ey
e
(
F_
p
1
o
r
F_
p
2
)
t
o
d
er
i
v
e
ac
ti
v
e
n
o
is
e
ca
n
ce
llatio
n
b
asis
s
ig
n
als.
I
t
s
h
o
u
ld
b
e
n
o
te
d
th
at
th
is
m
eth
o
d
ca
n
b
e
ap
p
lied
s
o
lely
to
m
u
ltip
le
-
elec
tr
o
d
e
E
E
G
s
ig
n
al
a
n
aly
s
is
s
o
,
as e
r
esu
lt m
ay
n
o
t su
itab
le
f
o
r
p
o
r
tab
le
d
ev
ices.
T
h
e
lo
ca
l
s
i
n
g
u
la
r
s
p
ec
t
r
u
m
a
n
aly
s
is
(
SS
A
)
m
et
h
o
d
[
1
0
]
,
[
1
1
]
h
as
f
o
u
n
d
a
p
p
li
ca
t
io
n
i
n
th
e
el
im
i
n
a
ti
o
n
o
f
h
i
g
h
-
l
ev
el
E
O
G
a
r
t
if
ac
ts
f
r
o
m
s
i
n
g
le
-
ch
an
n
el
E
E
G
s
i
g
n
a
ls
[
1
2
]
,
[
1
3
]
.
T
h
r
o
u
g
h
t
h
e
p
r
o
p
o
s
ed
m
et
h
o
d
,
v
e
ct
o
r
s
o
f
f
e
at
u
r
e
a
r
e
g
e
n
e
r
a
te
d
t
h
r
o
u
g
h
s
o
r
ti
n
g
o
f
d
el
ay
e
d
s
ig
n
als,
f
o
llo
we
d
b
y
cl
u
s
t
er
in
g
u
s
i
n
g
th
e
K
-
m
ea
n
s
a
lg
o
r
it
h
m
[
1
4
]
.
Ad
d
i
ti
o
n
all
y
,
i
n
t
h
e
p
r
o
c
e
s
s
o
f
s
i
n
g
u
la
r
v
al
u
e
d
ec
o
m
p
o
s
it
io
n
,
w
e
c
alc
u
l
at
e
th
e
e
ig
en
v
ec
t
o
r
s
a
n
d
ei
g
e
n
v
al
u
es
o
f
t
h
e
c
o
v
a
r
ia
n
ce
m
at
r
i
x
f
o
r
ea
ch
ca
te
g
o
r
y
.
T
h
e
m
i
n
i
m
u
m
d
e
s
cr
i
p
t
io
n
le
n
g
t
h
(
MD
L
)
m
et
r
i
c
[
1
5
]
q
u
an
ti
f
ies
t
h
e
d
at
a
n
ec
ess
ar
y
f
o
r
esti
m
a
ti
n
g
E
OG
s
ig
n
al,
c
o
n
s
i
d
e
r
i
n
g
th
e
d
im
e
n
s
i
o
n
o
f
s
i
g
n
a
l
s
u
b
ca
te
g
o
r
y
o
r
t
h
e
n
u
m
b
e
r
o
f
eig
e
n
v
ec
t
o
r
s
.
I
t'
s
i
m
p
o
r
ta
n
t
t
o
em
p
h
asi
ze
t
h
at
ac
h
ie
v
in
g
a
n
a
cc
u
r
ate
esti
m
at
e
o
f
t
h
e
s
i
g
n
al
s
u
b
ca
te
g
o
r
y
r
e
q
u
ir
es
cr
ea
ti
n
g
s
u
f
f
ic
ie
n
t
s
e
p
a
r
a
ti
o
n
a
m
o
n
g
th
e
ei
g
e
n
v
al
u
es t
h
at
d
ef
i
n
e
E
E
G
a
n
d
E
OG
s
i
g
n
als
[
1
0
]
.
E
f
f
ec
tiv
en
ess
o
f
th
e
l
o
ca
l
SS
A
ap
p
r
o
ac
h
p
r
im
ar
ily
lies
in
i
ts
ca
p
ab
ilit
y
to
clea
n
f
r
o
n
tal
E
E
G
s
ig
n
als
f
r
o
m
E
OG
in
ter
f
er
en
ce
,
in
s
itu
atio
n
s
ch
ar
ac
ter
ized
b
y
a
s
u
b
s
tan
tial
r
is
e
in
E
OG
am
p
litu
d
e
with
in
th
e
E
E
G.
I
n
co
n
tr
ast,
E
E
G
s
ig
n
als
ar
e
ty
p
i
ca
lly
m
o
s
t
r
eli
ab
le
at
c
h
an
n
els
C
3
an
d
C
4
[
1
0
]
,
an
d
d
u
e
to
th
e
s
p
atial
s
ep
ar
atio
n
b
etwe
en
th
e
elec
tr
o
d
es
a
n
d
t
h
e
ey
e,
th
er
e
is
n
'
t
a
s
u
b
s
tan
tial
m
ag
n
itu
d
e
co
n
t
r
ast
am
o
n
g
th
e
eig
en
v
alu
es
p
r
esen
tin
g
E
E
G
an
d
E
OG
s
ig
n
als.
C
o
n
s
eq
u
en
tly
,
MD
L
cr
it
er
io
n
m
a
y
f
all
s
h
o
r
t
in
ac
c
u
r
a
tely
esti
m
atin
g
th
e
tr
u
e
d
im
en
s
io
n
f
o
r
th
e
s
ig
n
al
s
u
b
ca
teg
o
r
y
,
lead
in
g
to
r
esid
u
a
l E
OG
ar
tifa
cts in
th
e
d
en
o
is
e
d
E
E
G
s
ig
n
al.
An
ad
ap
tiv
e
lin
e
en
h
an
ce
r
(
AL
E
)
was
em
p
lo
y
ed
f
o
r
s
ep
a
r
atin
g
th
e
elec
tr
o
ca
r
d
io
g
r
am
(
E
C
G)
an
d
elec
tr
o
m
y
o
g
r
am
(
E
MG
)
b
ase
d
o
n
th
e
p
r
in
cip
les
o
f
SS
A
[
1
6
]
.
I
n
th
is
ap
p
r
o
ac
h
,
th
e
A
L
E
is
d
esig
n
ed
to
ex
clu
s
iv
ely
elim
in
ate
th
e
in
ter
f
er
en
ce
s
ig
n
al
b
y
u
tili
zin
g
th
e
c
y
clic
n
atu
r
e
o
f
t
h
e
ar
tifa
ct
an
d
i
ts
d
elay
ed
v
er
s
io
n
,
with
th
e
d
elay
d
eter
m
in
e
d
ac
c
o
r
d
in
g
t
o
th
e
p
er
i
o
d
icity
o
f
t
h
e
co
r
r
u
p
ted
s
ig
n
al.
No
n
et
h
eless
,
g
i
v
en
th
at
E
E
G
an
d
E
OG
s
ig
n
als
ar
e
in
h
e
r
en
t
ly
d
y
n
a
m
ic,
th
e
p
r
esen
ted
ap
p
r
o
ac
h
d
em
o
n
s
tr
ates
lim
ited
ef
f
ec
tiv
en
ess
in
th
e
elim
in
atio
n
o
f
E
OG
ar
tifa
cts f
r
o
m
E
E
G
s
ig
n
als.
C
o
m
b
in
atio
n
o
f
d
is
cr
ete
wav
e
let
tr
an
s
f
o
r
m
(
DW
T
)
,
a
n
d
ANC
h
as
b
ee
n
u
s
ed
to
c
u
t
o
u
t
E
OG
ar
tifa
ct
f
r
o
m
E
E
G
s
ig
n
als
[
1
7
]
.
T
h
r
o
u
g
h
th
e
m
eth
o
d
,
t
h
e
DW
T
is
ap
p
lied
to
th
e
E
E
G
s
ig
n
al
to
o
b
tain
th
e
E
OG
b
asis
s
ig
n
al,
th
en
th
e
E
OG
a
r
tifa
ct
is
d
y
n
am
ically
d
elete
d
f
r
o
m
t
h
e
EEG
s
ig
n
al.
I
t
is
q
u
ite
cle
ar
th
at
th
e
ty
p
e
o
f
wav
elet
f
u
n
ctio
n
an
d
f
ig
u
r
e
o
f
d
is
in
teg
r
atio
n
le
v
els
ar
e
im
p
o
r
tan
t
f
o
r
a
g
o
o
d
esti
m
ate
o
f
t
h
e
r
ef
er
en
ce
s
ig
n
al.
T
h
e
ef
f
icien
cy
o
f
t
h
is
m
eth
o
d
p
r
im
ar
ily
r
elies o
n
esti
m
atio
n
o
f
th
e
r
e
f
er
en
ce
s
ig
n
al.
T
h
e
m
ajo
r
ity
o
f
al
g
o
r
ith
m
s
e
m
p
lo
y
ed
f
o
r
er
r
o
r
r
e
d
u
ctio
n
in
ad
ap
tiv
e
f
ilter
s
r
ely
o
n
th
e
m
ea
n
s
q
u
ar
ed
er
r
o
r
(
MSE
)
m
ea
s
u
r
e.
Ho
wev
er
,
it'
s
im
p
o
r
tan
t
to
n
o
te
th
at
th
e
MSE
cr
iter
io
n
is
p
r
im
ar
ily
s
u
itab
le
f
o
r
Gau
s
s
ian
an
d
s
tatio
n
ar
y
s
ig
n
als
as
it
s
o
lely
tak
es
in
to
ac
co
u
n
t
t
h
e
m
ea
n
an
d
v
ar
ian
ce
o
f
th
e
e
r
r
o
r
d
is
tr
ib
u
tio
n
.
T
h
is
cr
iter
io
n
m
ay
n
o
t
y
ield
an
o
p
tim
al
r
esp
o
n
s
e
f
o
r
s
ig
n
als
th
at
ex
h
ib
it
n
o
n
-
Gau
s
s
ian
ch
ar
ac
ter
is
tics
o
r
co
n
tain
s
ig
n
if
ican
t in
f
o
r
m
atio
n
in
t
h
eir
h
ig
h
e
r
-
o
r
d
er
s
tatis
tics
[
3
]
.
I
f
th
e
p
r
o
b
a
b
ilit
y
d
en
s
ity
f
u
n
ctio
n
(
PDF)
o
f
th
e
er
r
o
r
is
n
o
n
-
Gau
s
s
ian
,
th
en
an
ap
p
r
o
p
r
iate
co
s
t
f
u
n
ctio
n
f
o
r
th
e
ad
a
p
tiv
e
f
ilter
m
u
s
t
b
e
u
s
ed
.
E
r
d
o
g
m
u
s
an
d
Prin
cip
e
[
1
8
]
h
av
e
u
s
ed
th
e
er
r
o
r
en
tr
o
p
y
cr
iter
io
n
in
a
n
o
n
-
lin
ea
r
ad
ap
tiv
e
s
y
s
tem
.
So
f
ar
,
m
an
y
alg
o
r
ith
m
s
lik
e
m
in
im
u
m
er
r
o
r
en
tr
o
p
y
with
s
to
c
h
asti
c
in
f
o
r
m
atio
n
g
r
ad
ie
n
t
(
MEE
-
SI
G
)
[
1
9
]
,
m
in
im
u
m
er
r
o
r
e
n
tr
o
p
y
with
s
elf
ad
ju
s
tin
g
s
tep
-
s
ize
(
ME
E
-
SAS)
[
2
0
]
,
an
d
n
o
r
m
alize
d
m
in
im
u
m
er
r
o
r
en
tr
o
p
y
(
NM
E
E
)
[
2
1
]
,
h
av
e
b
ee
n
p
r
esen
ted
b
ased
o
n
th
is
cr
iter
io
n
.
W
h
en
th
e
g
o
al
o
f
ad
ap
tatio
n
is
to
r
em
o
v
e
u
n
ce
r
tain
ty
f
r
o
m
th
e
er
r
o
r
s
ig
n
al
as
m
u
ch
as
f
ea
s
ib
le,
th
e
er
r
o
r
en
tr
o
p
y
cr
iter
io
n
is
p
r
ef
er
ab
le
to
t
h
e
MSE
cr
iter
io
n
[
3
]
.
Alg
o
r
ith
m
s
th
at
o
p
e
r
ate
o
n
th
e
p
r
in
cip
le
o
f
er
r
o
r
en
tr
o
p
y
ten
d
to
e
x
h
ib
it
h
ig
h
r
esil
ien
ce
ag
ain
s
t
o
u
tlier
s
,
n
o
n
-
Gau
s
s
ian
,
an
d
n
o
n
-
s
tatio
n
ar
y
n
o
is
e
s
o
u
r
ce
s
.
C
o
n
s
eq
u
en
tly
,
th
ese
alg
o
r
ith
m
s
p
r
o
v
e
to
b
e
p
ar
ticu
lar
ly
well
-
s
u
ited
f
o
r
e
f
f
ec
tiv
ely
elim
in
atin
g
E
OG
ar
t
if
ac
ts
f
r
o
m
E
E
G
s
ig
n
als,
as
b
o
th
th
e
E
E
G
s
ig
n
al
an
d
E
OG
ar
tifa
ct
ex
h
i
b
it c
h
ar
a
cter
is
tics
th
at
ar
e
n
eith
er
Ga
u
s
s
ian
n
o
r
s
tatio
n
ar
y
.
Als
o
,
th
e
e
n
tr
o
p
y
-
b
ased
e
r
r
o
r
cr
iter
io
n
co
n
s
id
er
s
th
e
h
ig
h
e
r
-
o
r
d
er
s
tatis
tical
co
n
d
u
ct
o
f
th
e
s
y
s
tem
s
an
d
s
ig
n
als
[
3
]
.
I
n
th
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
,
we
in
tr
o
d
u
ce
t
h
e
DW
T
-
ME
E
alg
o
r
ith
m
f
o
r
r
em
o
v
in
g
E
OG
ar
ti
f
ac
ts
f
r
o
m
E
E
G
s
ig
n
als.
C
o
n
v
en
tio
n
al
le
ast
m
ea
n
s
q
u
ar
es
(
L
MS)
alg
o
r
ith
m
m
ay
n
o
t
p
o
s
s
ess
th
e
o
p
tim
al
ca
p
ab
ilit
y
to
ac
cu
r
ately
tr
ac
k
E
OG
ar
tifa
cts
d
u
e
to
th
eir
n
o
n
-
Gau
s
s
ian
ch
ar
ac
ter
is
tics
.
Fu
r
th
er
m
o
r
e,
m
in
i
m
u
m
er
r
o
r
en
tr
o
p
y
(
ME
E
)
m
eth
o
d
r
ep
r
esen
ts
a
s
u
b
s
tan
tial
im
p
r
o
v
em
e
n
t
o
v
er
t
h
e
L
MS
alg
o
r
ith
m
,
p
r
im
ar
ily
b
ec
au
s
e
it
co
n
s
tr
ain
s
er
r
o
r
e
n
tr
o
p
y
r
ath
e
r
th
an
s
o
lel
y
f
o
cu
s
in
g
o
n
m
ea
n
a
n
d
v
a
r
ian
ce
co
n
s
id
er
atio
n
s
[
3
]
.
I
n
th
e
s
u
g
g
ested
ap
p
r
o
ac
h
,
an
esti
m
atio
n
o
f
th
e
E
OG
i
n
ter
f
er
en
ce
is
d
er
iv
ed
as
th
e
b
asis
s
ig
n
al
th
r
o
u
g
h
th
e
ap
p
licatio
n
o
f
DW
T
o
n
th
e
E
E
G
s
ig
n
al
co
n
tam
in
ated
b
y
ar
tifa
cts.
T
h
e
s
ig
n
al
is
g
iv
en
to
th
e
ANC
r
ef
er
en
ce
in
p
u
t.
Sin
ce
th
is
E
OG
ar
tifa
ct
is
n
o
t
r
eg
en
er
ated
well
[
2
2
]
,
th
e
ex
p
licit
d
im
in
u
tio
n
o
f
th
is
p
r
ed
icted
s
ig
n
al
f
r
o
m
th
e
E
E
G
s
ig
n
al
d
o
es
n
o
t
r
esu
lt
in
r
em
o
v
al
o
f
all
o
f
t
h
e
E
OG
elem
en
ts
f
r
o
m
th
e
co
r
r
u
p
ted
E
E
G
s
ig
n
al.
T
o
elim
in
ate
an
y
r
e
m
ain
in
g
tr
ac
es
o
f
E
OG
ar
tifa
ct,
DW
T
is
u
n
ited
with
ME
E
.
Fin
ally
,
th
e
ME
E
-
b
ased
AN
C
ac
co
m
p
lis
h
es th
e
r
em
o
v
al
o
f
t
h
e
E
OG
ar
tifa
ct
th
r
o
u
g
h
co
n
ti
n
u
o
u
s
ad
a
p
tatio
n
o
f
f
ilter
co
e
f
f
icien
ts
.
T
h
e
s
tr
u
ctu
r
e
o
f
ANC
,
DW
T
,
an
d
th
e
en
tr
o
p
y
-
b
ased
er
r
o
r
m
etr
ic
ar
e
elab
o
r
ate
d
u
p
o
n
in
th
e
s
ec
o
n
d
p
ar
t
o
f
th
is
s
tu
d
y
.
Deta
iled
d
escr
ip
tio
n
ab
o
u
t
th
e
p
r
o
p
o
s
ed
m
eth
o
d
is
p
r
o
v
id
e
d
in
s
ec
tio
n
3
.
T
h
e
an
aly
s
is
o
f
s
im
u
latio
n
s
is
p
r
o
v
id
ed
i
n
s
ec
tio
n
4
.
Fin
ally
,
th
e
c
o
n
clu
s
io
n
i
s
in
s
ec
tio
n
5.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
E
lectro
en
ce
p
h
a
lo
g
r
a
m
d
en
o
is
in
g
u
s
in
g
d
is
crete
w
a
ve
let
tr
a
n
s
fo
r
m
a
n
d
a
d
a
p
tive
…
(
Ha
s
h
em
A
b
d
o
la
h
n
iya
)
771
2.
M
E
T
H
O
D
I
n
th
is
s
ec
tio
n
,
th
e
p
r
o
p
o
s
ed
m
eth
o
d
f
o
r
E
E
G
s
ig
n
al
n
o
is
e
ca
n
ce
latio
n
h
as
b
ee
n
in
tr
o
d
u
c
ed
.
I
n
th
is
s
ec
tio
n
,
th
e
s
tr
u
ctu
r
e
o
f
th
e
a
d
ap
tiv
e
n
o
is
e
ca
n
ce
latio
n
is
in
tr
o
d
u
ce
d
an
d
in
p
u
t
-
o
u
t
p
u
t
s
ig
n
als
ar
e
s
p
ec
if
ied
.
T
h
en
,
th
e
weig
h
t
u
p
d
atin
g
alg
o
r
ith
m
f
o
r
ad
ap
tiv
e
f
ilter
,
b
ased
o
n
th
e
er
r
o
r
en
t
r
o
p
y
m
in
im
i
za
tio
n
is
in
tr
o
d
u
ce
d
th
at
ca
lled
ME
E
alg
o
r
ith
m
.
Als
o
,
th
e
m
ath
em
atica
l
eq
u
atio
n
s
o
f
u
p
d
atin
g
th
e
weig
h
ts
ar
e
ex
p
r
ess
ed
.
Af
ter
th
at,
th
e
DW
T
tr
an
s
f
o
r
m
atio
n
an
d
it
s
m
ath
em
atica
l
r
elatio
n
s
h
ip
s
ar
e
ex
am
in
ed
.
Fin
ally
,
t
h
e
s
tr
u
ct
u
r
e
o
f
th
e
p
r
o
p
o
s
ed
m
eth
o
d
n
am
ed
DW
T
-
ANC
-
MEE an
d
h
o
w
it wo
r
k
s
will b
e
d
escr
ib
ed
.
2
.
1
.
Ada
ptiv
e
no
is
e
ca
ncella
t
io
n
T
h
e
p
r
im
ar
y
elem
en
ts
o
f
AN
C
co
n
s
is
t
o
f
b
lo
ck
s
r
esp
o
n
s
ib
le
f
o
r
weig
h
t
u
p
d
ates
an
d
f
ilter
in
g
,
as
illu
s
tr
ated
in
Fig
u
r
e
1
.
Var
io
u
s
alg
o
r
ith
m
s
,
s
u
ch
as
L
MS,
r
ec
u
r
s
iv
e
least
s
q
u
ar
es
(
R
L
S
)
,
o
r
ME
E
,
ca
n
b
e
em
p
lo
y
ed
to
u
p
d
ate
th
e
weig
h
t
s
.
I
n
o
u
r
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
,
w
e
o
p
t
f
o
r
ME
E
o
win
g
to
its
ex
ce
p
tio
n
al
ca
p
ab
ilit
y
to
ef
f
ec
tiv
ely
tr
ac
k
n
o
n
-
Gau
s
s
ian
an
d
n
o
n
-
s
tatio
n
ar
y
n
o
is
e,
wh
ile
also
lev
e
r
ag
in
g
h
ig
h
e
r
-
d
eg
r
ee
s
tatis
tical
an
aly
s
is
.
Usi
n
g
th
e
p
r
esen
ted
m
eth
o
d
,
th
e
f
ilter
co
ef
f
icien
ts
r
ea
ch
th
eir
o
p
tim
u
m
v
alu
es
with
a
r
e
d
u
ce
d
n
u
m
b
e
r
o
f
iter
atio
n
s
in
co
m
p
ar
is
o
n
t
o
alter
n
ativ
e
tech
n
iq
u
es
[
3
]
.
Fig
u
r
e
1
.
Ov
e
r
v
iew
o
f
ANC co
n
f
ig
u
r
atio
n
2
.
2
.
M
ini
m
um
er
ro
r
ent
ro
py
m
et
ho
d
T
h
e
m
ea
s
u
r
e
o
f
av
er
a
g
e
d
ata
co
n
ten
t
with
in
a
s
p
ec
if
ic
d
ata
ar
r
an
g
em
e
n
t
is
co
m
m
o
n
ly
r
e
f
er
r
ed
to
as
th
e
co
n
ce
p
t
o
f
en
tr
o
p
y
.
I
n
d
ee
d
,
MSE
r
ep
r
esen
ts
a
s
p
ec
if
ic
in
s
tan
ce
o
f
en
t
r
o
p
y
,
as
MSE
ex
c
lu
s
iv
ely
d
ea
ls
with
th
e
s
ec
o
n
d
-
o
r
d
er
s
tatis
tics
o
f
th
e
er
r
o
r
PDF
,
wh
er
ea
s
en
tr
o
p
y
en
co
m
p
ass
es th
e
en
tire
ty
o
f
th
e
PDF o
f
th
e
er
r
o
r
d
is
tr
ib
u
tio
n
.
I
n
f
ac
t,
wh
en
en
tr
o
p
y
is
m
i
n
im
ized
,
all
t
h
e
m
o
m
en
ts
o
f
e
r
r
o
r
d
is
tr
ib
u
tio
n
ar
e
m
in
im
ized
,
n
o
t
j
u
s
t
th
e
s
ec
o
n
d
m
o
m
e
n
t.
T
h
er
e
f
o
r
e
,
th
e
en
tr
o
p
y
cr
iter
io
n
ca
n
b
e
c
h
o
s
en
as
a
b
etter
alte
r
n
ativ
e
to
MSE
.
I
n
th
is
s
tu
d
y
,
we
em
p
lo
y
th
e
f
o
llo
win
g
ex
t
en
d
ed
d
e
f
in
itio
n
s
o
f
en
t
r
o
p
y
to
o
f
f
er
v
er
s
atility
with
in
a
p
ar
am
etr
ic
s
et,
with
Sh
an
n
o
n
'
s
co
n
ce
p
ts
s
er
v
in
g
as
th
e
s
p
ec
if
ic
s
itu
atio
n
wh
en
th
e
p
ar
am
eter
'
a'
eq
u
als
1
.
T
h
e
R
en
y
i
en
tr
o
p
y
o
f
'
e,
'
with
o
r
d
er
'
a,
'
is
d
ef
in
ed
as
(
1
)
[
1
8
]
:
H
α
(
e
)
=
1
1
−
α
l
og
∫
f
α
(
e
)
de
(
1
)
Her
e,
f
(
e
)
r
ep
r
esen
ts
er
r
o
r
r
an
d
o
m
v
ar
iab
le’
s
PDF.
I
n
th
e
p
r
esen
t
r
esear
ch
,
we
h
av
e
o
p
te
d
f
o
r
Seco
n
d
-
o
r
d
e
r
R
en
y
i’
s
en
tr
o
p
y
d
en
o
te
d
as
=
2
,
wh
ich
is
ex
p
r
ess
ed
as (
2
)
:
H
2
(
e
)
=
−
l
og
∫
f
2
(
e
)
de
(
2
)
As
ev
id
en
t
f
r
o
m
(
1
)
an
d
(
2
)
,
th
e
co
m
p
u
tatio
n
o
f
en
tr
o
p
y
n
ec
ess
itate
s
k
n
o
wled
g
e
o
f
th
e
PDF
o
f
a
r
an
d
o
m
v
ar
iab
le.
C
o
n
s
eq
u
en
tl
y
,
in
th
is
s
tu
d
y
,
th
e
Par
ze
n
win
d
o
win
g
m
eth
o
d
is
em
p
lo
y
ed
f
o
r
th
e
esti
m
atio
n
o
f
th
is
PDF.
T
h
e
Par
ze
n
win
d
o
w
in
g
tech
n
iq
u
e
is
r
ep
r
esen
ted
b
y
(
3
)
:
f
̂
(
e
)
=
1
N
∑
k
σ
N
i
=
1
(
e
−
e
(
i
)
)
(
3
)
Her
e,
k
(
e
)
r
ep
r
esen
ts
th
e
Ker
n
el
f
u
n
ctio
n
,
σ
d
en
o
tes
th
e
Ker
n
el
s
ize,
an
d
{
{
e
(
1
)
,
e
(
2
)
,
…
,
e
(
N
)
}
r
ep
r
esen
ts
th
e
er
r
o
r
d
ata
p
o
in
ts
.
W
h
ile
v
ar
io
u
s
k
er
n
el
f
u
n
ctio
n
s
ca
n
b
e
em
p
lo
y
ed
f
o
r
th
e
Par
ze
n
win
d
o
w,
in
th
is
s
tu
d
y
,
we
u
tili
ze
a
m
u
ltid
im
e
n
s
io
n
al
Gau
s
s
ian
d
is
tr
ib
u
tio
n
f
u
n
ctio
n
with
a
r
a
d
ially
s
y
m
m
etr
ic
al
v
ar
ian
ce
o
f
σ
2
.
C
o
n
s
eq
u
en
tly
,
th
e
esti
m
atio
n
o
f
R
en
y
i'
s
q
u
ad
r
atic
en
tr
o
p
y
f
o
r
er
r
o
r
s
am
p
les is
ca
lcu
lated
as f
o
llo
ws:
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
14
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
5
:
769
-
7
7
9
772
H
̂
2
(
e
)
=
−
l
og
∫
(
1
N
∑
G
σ
(
e
−
e
(
i
)
)
N
i
=
1
)
2
de
+
∞
−
∞
H
̂
2
(
e
)
=
−
l
og
1
N
2
∫
(
∑
∑
G
σ
(
e
−
e
(
j
)
)
G
σ
(
e
−
e
(
i
)
)
N
j
=
1
)
de
N
i
=
1
+
∞
−
∞
H
̂
2
(
e
)
=
−
l
og
1
N
2
(
∑
∑
∫
G
σ
(
e
−
e
(
j
)
)
G
σ
(
e
−
e
(
i
)
)
+
∞
−
∞
N
j
=
1
)
de
N
i
=
1
(
4
)
H
̂
2
(
e
)
=
−
l
og
(
1
N
2
∑
∑
G
σ
√
2
(
e
(
j
)
−
e
(
i
)
)
N
j
=
1
N
i
=
1
)
(
5
)
In
(
4
)
a
n
d
(
5
)
,
G
σ
(
.
)
r
ep
r
esen
ts
th
e
k
er
n
el
f
u
n
ctio
n
with
a
Gau
s
s
ian
co
r
e.
T
h
e
ter
m
en
clo
s
e
d
with
in
th
e
lo
g
ar
ith
m
ic
f
u
n
ctio
n
,
o
f
ten
r
ef
er
r
ed
to
as
in
f
o
r
m
atio
n
p
o
te
n
tial
(
I
P),
is
p
r
esen
ted
as
(
6
)
:
V
̂
2
(
e
)
=
1
N
2
∑
∑
G
σ
√
2
(
e
(
j
)
−
e
(
i
)
)
N
j
=
1
N
i
=
1
(
6
)
Hen
ce
,
th
e
en
tr
o
p
y
e
q
u
atio
n
f
o
r
th
e
er
r
o
r
r
an
d
o
m
v
a
r
iab
le
c
an
b
e
ex
p
r
ess
ed
as
(
7
)
:
H
̂
2
(
e
)
=
−
l
og
(
V
̂
2
(
e
)
)
(
7
)
As
th
e
lo
g
ar
ith
m
f
u
n
ctio
n
is
a
m
o
n
o
to
n
ic
tr
an
s
f
o
r
m
atio
n
,
it
m
ay
b
e
d
ed
u
ce
d
th
at
th
e
lo
wes
t
en
tr
o
p
y
co
r
r
esp
o
n
d
s
to
th
e
h
ig
h
est I
P.
T
h
en
,
t
h
e
co
s
t f
u
n
ctio
n
J
(
e)
f
o
r
th
e
ME
E
cr
iter
io
n
ca
n
b
e
d
ef
in
e
d
as
(
8
)
:
J
ME
E
(
e
)
=
V
(
e
)
w
m
ax
(
8
)
I
n
o
n
lin
e
tr
ain
in
g
s
ce
n
ar
i
o
s
,
th
e
s
to
ch
asti
c
in
f
o
r
m
atio
n
g
r
a
d
ien
t
(
SIG
)
ca
n
b
e
em
p
lo
y
ed
to
esti
m
ate
th
e
IP
,
as
d
ep
icted
in
(
9
)
.
As
a
co
m
m
o
n
r
esu
lt,
th
e
s
to
ch
asti
c
g
r
ad
ien
t
o
f
in
f
o
r
m
atio
n
is
d
er
iv
ed
,
co
m
p
u
ted
as
th
e
s
u
m
m
atio
n
o
f
L
n
ew
s
am
p
les at
tim
e
n
.
V
̂
2
(
e
(
n
)
)
≈
1
L
∑
G
σ
√
2
(
e
(
n
)
−
e
(
i
)
)
n
−
1
i
=
n
−
L
(
9
)
B
y
im
p
lem
en
tin
g
th
e
ME
E
-
SI
G
m
eth
o
d
f
o
r
c
o
ef
f
icien
t
u
p
d
a
tes,
as
d
escr
ib
ed
i
n
[
2
1
]
,
th
e
a
im
is
to
r
e
d
u
ce
th
e
s
ig
n
al
e(
n
)
en
t
r
o
p
y
with
in
th
e
ad
ap
tiv
e
f
ilter
illu
s
tr
ated
in
Fi
g
u
r
e
1
.
(
+
1
)
=
(
)
+
.
∇
(
(
)
)
(
1
0
)
I
n
th
is
co
n
tex
t,
th
e
g
r
a
d
ien
t is
ca
lcu
lated
as
(
1
1
)
:
∇
(
(
)
)
=
1
2
2
∑
√
2
(
(
)
−
(
)
)
−
1
=
−
{
(
)
−
(
)
}
{
(
)
−
(
)
}
(
1
1
)
2
.
3
.
Dis
cr
et
e
wa
v
ele
t
t
ra
ns
f
o
rm
T
h
e
wav
elet
an
aly
s
is
is
am
o
n
g
th
e
m
o
s
t
p
o
wer
f
u
l
to
o
ls
f
o
r
ca
p
tu
r
in
g
b
o
th
tem
p
o
r
al
a
n
d
s
p
ec
tr
al
in
f
o
r
m
atio
n
with
in
a
s
ig
n
al.
I
t
ex
ce
ls
in
p
r
o
v
id
in
g
en
h
a
n
ce
d
tem
p
o
r
al
r
eso
lu
tio
n
f
o
r
s
eg
m
en
ts
with
h
ig
h
er
f
r
eq
u
e
n
cies
an
d
im
p
r
o
v
ed
f
r
eq
u
en
cy
r
eso
lu
tio
n
f
o
r
co
m
p
o
n
e
n
ts
with
lo
wer
f
r
eq
u
e
n
cies.
T
h
e
s
ig
n
al,
d
en
o
ted
as
y
(
t)
,
co
u
ld
b
e
ex
p
r
ess
ed
th
r
o
u
g
h
a
wav
elet
an
al
y
s
is
as o
u
tlin
ed
in
[
2
3
]
.
(
)
=
∑
ϕ
+
∑
∑
(
)
−
1
(
1
2
)
W
h
er
e
is
th
e
p
r
o
x
im
ate
an
d
ar
e
th
e
co
e
f
f
icien
ts
d
etail.
T
o
r
ec
o
n
s
tr
u
ct
t
h
e
p
r
im
ar
y
s
ig
n
al
y
(
t)
a
t
p
r
esen
ted
d
ec
o
m
p
o
s
itio
n
lev
el
(
e.
g
.
,
M
)
,
th
e
p
r
o
ce
s
s
is
as
(
1
3
)
:
(
)
=
(
)
+
∑
(
)
=
1
(
1
3
)
2
.
4
.
P
r
o
po
s
ed
d
is
c
re
t
e
w
a
v
el
et
t
r
a
ns
f
o
r
m
-
a
d
a
pt
iv
e
n
o
is
e
ca
n
ce
l
la
t
i
o
n
-
m
ini
m
u
m e
rr
o
r
ent
ro
p
y
t
ec
hn
iqu
e
Fig
u
r
e
2
d
ep
icts
th
e
p
r
o
ce
s
s
o
f
E
OG
ar
tifa
ct
ca
n
ce
llatio
n
u
s
in
g
t
h
e
DW
T
-
ME
E
tech
n
iq
u
e.
T
h
e
co
r
r
u
p
te
d
E
E
G
s
ig
n
al
(
)
,
o
b
tain
ed
as
th
e
s
u
m
o
f
th
e
E
E
G
s
ig
n
al
(
)
an
d
a
c
o
n
s
tan
t
r
atio
(
)
o
f
th
e
E
OG
s
ig
n
al
1
(
)
,
is
th
e
in
p
u
t.
T
o
g
en
er
a
te
th
e
r
e
f
er
en
ce
s
ig
n
al
r
eq
u
ir
e
d
f
o
r
th
e
ANC,
th
e
DW
T
is
em
p
lo
y
ed
o
n
th
e
s
ig
n
al
v
ec
to
r
=
[
(
1
)
,
(
2
)
,
…
,
(
)
]
,
wh
ich
is
h
eld
in
a
b
u
f
f
er
o
f
s
ize
n
.
T
h
e
s
u
b
s
eq
u
en
t
s
u
b
s
ec
tio
n
s
will
d
etail
th
e
p
r
im
ar
y
s
tep
s
in
v
o
lv
ed
in
cr
ea
tin
g
th
e
s
o
u
r
ce
s
ig
n
a
l f
o
r
ANC.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
E
lectro
en
ce
p
h
a
lo
g
r
a
m
d
en
o
is
in
g
u
s
in
g
d
is
crete
w
a
ve
let
tr
a
n
s
fo
r
m
a
n
d
a
d
a
p
tive
…
(
Ha
s
h
em
A
b
d
o
la
h
n
iya
)
773
Fig
u
r
e
2
.
DW
T
-
ME
E
m
eth
o
d
b
lo
ck
d
ia
g
r
am
2
.
4
.
1
.
E
x
t
ra
ct
io
n t
he
s
o
urce
s
ig
na
l f
o
r
a
da
ptiv
e
no
is
e
ca
n
ce
lla
t
io
n
us
ing
dis
cr
et
e
wa
v
e
let
t
ra
ns
f
o
rm
I
n
th
is
s
tu
d
y
,
we
h
a
v
e
co
n
d
u
cted
a
m
u
lti
-
lev
el
wav
elet
d
e
co
m
p
o
s
itio
n
u
s
in
g
DW
T
u
p
t
o
lev
el
8
.
T
h
is
p
r
o
ce
s
s
en
ab
les
th
e
p
r
e
cise
id
en
tific
atio
n
o
f
wav
elet
co
ef
f
icien
ts
ass
o
ciate
d
with
th
e
E
OG
ar
tifa
ct,
tak
in
g
in
to
c
o
n
s
id
er
atio
n
t
h
e
ef
f
ec
tiv
e
f
ac
to
r
s
s
u
ch
as
ey
e
m
o
v
em
e
n
t
(
0
-
7
Hz)
an
d
b
lin
k
in
g
(
8
-
1
3
Hz)
(
lev
el
8
:
0
.
2
5
-
0
.
5
Hz,
lev
el
7
:
0
.
5
-
1
Hz,
lev
el
6
:
1
-
2
Hz,
lev
e
l
5
:
2
-
4
Hz,
lev
el
4
:
4
-
8
Hz,
le
v
el
3
:
8
-
1
6
Hz,
lev
el
2
:
1
6
-
3
2
Hz
,
an
d
lev
el
1
:
3
2
-
6
4
Hz)
.
T
h
r
esh
o
ld
in
g
h
as
b
ee
n
ap
p
lied
to
th
e
d
etail
co
ef
f
icien
ts
f
r
o
m
lev
el
8
d
o
wn
to
lev
el
3
f
o
r
E
OG
d
en
o
is
in
g
.
Dec
o
m
p
o
s
in
g
th
e
E
E
G
s
ig
n
al
to
lev
el
3
p
r
o
v
id
es
th
e
ess
en
tial
o
cu
lar
-
r
elate
d
wav
elet
co
ef
f
icien
ts
n
ec
ess
ar
y
f
o
r
n
o
is
e
r
ed
u
ctio
n
,
as
in
d
icate
d
in
[
2
4
]
.
T
o
o
b
tain
t
h
e
r
ef
e
r
en
ce
s
ig
n
al,
we
ap
p
l
y
th
e
DW
T
to
th
e
n
o
is
y
s
ig
n
al,
r
esu
ltin
g
in
th
e
E
OG
s
ig
n
al
v
ec
to
r
,
d
en
o
ted
as
2
,
wh
ich
s
er
v
es
as
th
e
p
r
im
ar
y
in
p
u
t f
o
r
th
e
ANC s
y
s
tem
.
2
.
4
.
2
.
D
is
cr
et
e
wa
v
elet
t
r
a
ns
f
o
rm
-
m
ini
m
um
er
ro
r
ent
ro
py
I
n
th
e
DW
T
-
ME
E
m
eth
o
d
p
r
o
p
o
s
ed
,
t
h
e
ANC
s
y
s
tem
ta
k
es
as
its
in
itia
l
an
d
s
o
u
r
ce
in
p
u
ts
th
e
co
n
tam
in
ated
E
E
G
s
ig
n
al
v
ec
to
r
(
)
an
d
th
e
elicited
E
OG
s
ig
n
al
v
ec
to
r
(
2
,
)
,
r
esp
ec
tiv
ely
.
B
y
r
en
ewin
g
th
e
co
ef
f
icien
ts
o
f
th
e
f
ilter
u
s
in
g
th
e
ME
E
m
eth
o
d
,
a
n
esti
m
ate
o
f
th
e
s
ig
n
al
̂
1
(
)
is
o
b
tain
ed
f
r
o
m
th
e
s
ig
n
al
2
.
T
o
o
b
tain
a
clea
n
E
E
G
s
ig
n
al,
in
ea
ch
b
lo
ck
o
f
d
ata,
th
e
esti
m
ated
s
ig
n
al
̂
1
(
)
is
d
ed
u
cte
d
f
r
o
m
th
e
c
o
r
r
u
p
ted
E
E
G
s
ig
n
al
y
(
n
)
.
T
h
e
tim
e
r
eq
u
ir
ed
to
d
er
iv
e
th
e
c
o
r
r
ec
ted
E
E
G
s
ig
n
al
f
o
r
ea
ch
b
lo
ck
m
a
tch
es
th
e
co
m
b
in
ed
co
m
p
u
tin
g
d
u
r
atio
n
s
o
f
th
e
s
er
ial
-
to
-
p
ar
allel
co
n
v
e
r
ter
,
th
e
DW
T
p
r
o
ce
s
s
,
p
ar
allel
-
to
-
s
er
i
al
co
n
v
er
ter
,
an
d
th
e
ANC
o
p
er
atio
n
.
T
h
is
m
eth
o
d
is
p
ar
ticu
lar
ly
ad
v
an
tag
eo
u
s
as
th
e
co
m
p
u
tatio
n
al
tim
e
to
o
b
tain
th
e
ac
cu
r
at
e
E
E
G
s
ig
n
al
is
s
h
o
r
ter
th
an
th
e
s
am
p
lin
g
in
ter
v
al
o
f
th
e
E
E
G
s
ig
n
al.
3.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
Dele
tin
g
th
e
E
OG
ar
tifa
ct
f
r
o
m
liv
e
E
E
G
d
ata
allo
ws
r
esear
ch
er
s
to
test
h
o
w
well
th
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
wo
r
k
s
.
E
v
en
m
o
r
e
s
o
at
f
r
o
n
tal
s
ites
,
th
e
E
E
G
s
ig
n
a
l
m
ig
h
t
b
e
co
n
tam
in
ate
d
b
y
E
OG
ar
tifa
ct.
C
ar
ef
u
l
in
s
tr
u
ctio
n
s
ca
n
r
ed
u
ce
th
e
s
e
v
er
ity
o
f
th
e
p
r
o
b
lem
,
b
u
t
th
ey
r
ar
ely
elim
in
ate
it
e
n
tire
ly
.
Fig
u
r
e
3
s
h
o
ws
a
s
am
p
le
o
f
ex
p
e
r
im
en
tal
s
ig
n
al
s
o
b
tain
ed
f
r
o
m
B
C
I
co
m
p
etitio
n
d
ataset
th
at
u
s
ed
in
th
e
ex
p
r
im
en
ts
.
As
ca
n
b
e
s
ee
n
th
e
o
r
ig
i
n
al
E
E
G
d
ata
s
h
o
ws
in
Fig
u
r
e
3
(
a)
.
Als
o
Fig
u
r
e
3
(
b
)
d
is
p
lay
s
E
OG
ar
tif
ac
t
an
d
Fig
u
r
e
3
(
c
)
r
ep
r
esen
t
s
ig
n
al
co
n
tam
in
ated
with
n
o
is
e.
A
clea
n
E
E
G
s
ig
n
al,
E
OG
ar
tifa
ct,
an
d
an
E
E
G
s
ig
n
al
co
r
r
u
p
ted
with
E
OG
ar
tifa
ct
ca
n
b
e
s
ee
n
i
n
t
h
is
f
ig
u
r
e
.
Als
o
,
i
n
o
r
d
er
to
s
ee
th
e
f
r
eq
u
en
cy
s
p
ec
tr
u
m
o
f
t
h
e
clea
n
s
ig
n
al,
t
h
e
E
OG
ar
tifa
ct
an
d
th
e
c
o
r
r
u
p
ted
E
E
G
s
ig
n
al
with
th
e
E
OG
ar
tif
ac
t,
th
eir
p
o
wer
s
p
ec
tr
al
d
en
s
it
y
(
PS
D)
ar
e
s
h
o
wn
in
Fig
u
r
e
s
4
(
a)
to
4
(
c)
r
esp
ec
ti
v
ely
.
I
n
t
h
is
f
ig
u
r
e,
t
h
e
h
o
r
izo
n
tal
ax
is
r
e
p
r
esen
ts
f
r
e
q
u
en
c
y
an
d
th
e
v
e
r
tical
ax
is
r
ep
r
esen
ts
PS
D.
Mu
l
ti
p
le
E
E
G
s
i
g
n
al
r
ec
o
r
d
i
n
g
s
c
o
v
e
r
i
n
g
a
w
id
e
v
ar
i
et
y
o
f
w
av
e
s
h
a
p
es
h
a
v
e
b
e
e
n
u
s
e
d
t
o
v
ali
d
at
e
th
e
p
r
o
p
o
s
e
d
t
ec
h
n
i
q
u
e
i
n
t
h
e
2
0
0
8
B
C
I
co
m
p
et
iti
o
n
d
at
aset
[
2
5
]
.
W
e
h
a
v
e
e
m
p
l
o
y
e
d
r
ea
l
-
wo
r
l
d
E
OG
s
i
g
n
al
ar
t
if
ac
ts
f
r
o
m
a
d
ata
b
ase
u
tili
z
ed
i
n
t
h
e
B
C
I
c
o
m
p
etit
io
n
i
n
o
r
d
e
r
t
o
ass
ess
t
h
e
e
n
h
a
n
c
em
e
n
t
p
e
r
f
o
r
m
a
n
c
e
w
it
h
i
n
a
d
y
n
am
ic
e
n
v
ir
o
n
m
e
n
t
.
B
C
I
d
atas
ets
ty
p
i
ca
l
l
y
co
n
s
is
t
o
f
t
h
r
e
e
E
OG
c
h
a
n
n
els
a
n
d
twe
n
t
y
-
t
wo
E
E
G
m
o
n
o
p
o
la
r
ch
an
n
els.
A
ll
s
i
g
n
als
w
er
e
b
a
n
d
p
ass
f
ilt
er
ed
b
etwe
e
n
0
.
5
a
n
d
1
0
0
H
z
a
n
d
ca
p
t
u
r
e
d
at
2
5
0
Hz.
A
d
d
it
io
n
all
y
,
a
n
o
tc
h
f
i
lte
r
wit
h
a
f
r
e
q
u
e
n
c
y
o
f
5
0
H
z
was
u
s
ed
o
n
t
h
is
d
at
a
s
et.
Ni
n
e
h
e
alt
h
y
v
o
l
u
n
tee
r
s
w
er
e
u
s
e
d
to
c
o
l
le
ct
E
E
G
d
ata
o
v
e
r
th
e
c
o
u
r
s
e
o
f
t
wo
s
es
s
i
o
n
s
.
E
ac
h
s
ess
i
o
n
b
e
g
i
n
s
wi
th
t
h
r
ee
tas
k
s
:
o
n
e
wit
h
e
y
es
cl
o
s
e
d
,
o
n
e
wi
th
ey
es
o
p
e
n
,
a
n
d
o
n
e
i
n
cl
u
d
i
n
g
ey
e
m
o
v
e
m
e
n
ts
[
2
5
]
.
C
o
m
p
ar
ed
to
C
r
o
f
t'
s
i
n
te
n
d
e
d
s
t
r
ate
g
y
[
2
6
]
,
t
h
e
el
ec
t
r
o
d
e
p
la
ce
m
e
n
t
f
o
r
t
h
e
E
OG
s
ig
n
als
in
t
h
is
d
ata
s
et
is
u
n
i
q
u
e
.
A
ll
E
E
G
s
i
g
n
als
we
r
e
als
o
f
ilt
er
e
d
w
ith
a
l
o
w
-
p
ass
f
il
te
r
s
et
a
t
4
5
Hz
f
o
r
t
h
is
s
t
u
d
y
.
T
o
av
o
i
d
co
n
f
o
u
n
d
i
n
g
t
h
e
E
E
G
r
e
ad
i
n
g
s
,
a
lo
w
-
p
ass
f
ilt
er
s
et
at
2
0
Hz
is
u
s
e
d
t
o
th
e
E
OG
r
ea
d
i
n
g
s
i
n
t
h
is
m
et
h
o
d
[
2
6
]
.
Fi
g
u
r
e
3
(
b
)
d
e
p
i
cts
a
n
E
O
G
ar
t
if
ac
t,
w
h
i
ch
we
u
s
e
d
as
t
h
e
m
o
d
el'
s
r
e
f
er
en
ce
s
ig
n
al
x
(
n
)
.
T
h
e
in
p
u
t
t
o
an
ad
ap
t
iv
e
f
ilt
er
is
a
n
E
E
G
s
i
g
n
al
t
h
at
h
as
b
ee
n
c
o
n
ta
m
i
n
a
te
d
b
y
E
OG
a
r
t
if
ac
ts
.
I
n
th
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
14
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
5
:
769
-
7
7
9
774
s
u
g
g
este
d
te
ch
n
i
q
u
e
,
t
h
e
PS
D
,
c
o
h
er
e
n
ce
e
x
a
m
i
n
a
ti
o
n
a
n
d
s
i
g
n
al
-
to
-
n
o
is
e
r
at
io
(
S
NR
)
we
r
e
u
s
e
d
t
o
ass
ess
t
h
e
q
u
ali
ty
o
f
n
o
is
y
E
E
G
s
i
g
n
als
.
T
h
e
eq
u
ati
o
n
f
o
r
e
n
h
a
n
c
in
g
t
h
e
q
u
ali
ty
o
f
a
n
E
E
G
s
i
g
n
al
is
a
s
(
1
4
)
:
=
10
l
og
10
(
(
)
)
(
(
)
−
̂
(
)
)
(
1
4
)
T
h
e
v
ar
ian
ce
o
p
er
ato
r
is
d
en
o
t
ed
b
y
v
a
r
,
an
d
th
e
r
ea
l
(
)
s
ig
n
al,
an
d
th
e
tr
u
e
E
E
G
s
ig
n
al
̂
(
)
wh
i
ch
is
esti
m
ated
u
s
in
g
th
e
ad
ap
tiv
e
f
ilter
,
r
esp
ec
tiv
ely
.
s
am
p
les
(
a)
s
am
p
les
(
b
)
(
c)
Fig
u
r
e
3
.
A
s
am
p
le
o
f
ex
p
er
im
en
tal
s
ig
n
als f
r
o
m
B
C
I
co
m
p
et
itio
n
d
ataset
(
a)
E
E
G
with
o
u
t
n
o
is
e,
(
b
)
E
OG
ar
tifa
ct,
a
n
d
(
c
)
s
ig
n
a
l c
o
n
tam
in
ated
with
n
o
is
e
F
r
e
q
u
e
n
c
y
(
Hz)
(
a)
F
r
e
q
u
e
n
c
y
(
Hz)
(
b
)
(
c)
Fig
u
r
e
4
.
PSD
o
f
e
x
p
er
im
e
n
tal
s
ig
n
als (
a)
E
E
G
s
ig
n
al,
(
b
)
E
OG,
an
d
(
c)
E
E
G
co
n
tam
i
n
ated
b
y
E
OG
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
E
lectro
en
ce
p
h
a
lo
g
r
a
m
d
en
o
is
in
g
u
s
in
g
d
is
crete
w
a
ve
let
tr
a
n
s
fo
r
m
a
n
d
a
d
a
p
tive
…
(
Ha
s
h
em
A
b
d
o
la
h
n
iya
)
775
Sin
ce
th
e
E
E
G
s
ig
n
al
is
n
o
t
co
n
s
tan
t,
th
is
ar
ticle
em
p
lo
y
s
W
elch
's
ap
p
r
o
ac
h
to
esti
m
ate
th
e
p
o
wer
co
m
p
o
n
en
t
u
s
in
g
a
Ham
m
in
g
-
win
d
o
w
o
f
1
0
2
4
d
ata
p
o
in
ts
o
f
1
s
ec
o
n
d
d
u
r
atio
n
with
5
0
%
o
v
er
lap
.
W
e
ass
e
s
s
ed
s
p
ec
tr
al
co
h
er
e
n
ce
b
etwe
en
d
en
o
is
ed
an
d
clea
n
E
E
G
s
ig
n
a
ls
as
a
n
u
m
e
r
ical
m
ea
s
u
r
em
e
n
t
f
o
r
th
e
p
r
o
p
o
s
e
d
ef
f
icien
t
d
e
n
o
is
in
g
ap
p
r
o
ac
h
at
th
e
tar
g
et
f
r
eq
u
e
n
cy
.
T
h
e
f
o
llo
win
g
is
t
h
e
c
o
n
s
is
ten
cy
e
q
u
atio
n
b
etwe
en
th
e
̂
(
)
s
ig
n
al
an
d
th
e
(
)
s
ig
n
al.
ℎ
(
)
=
|
̂
(
)
|
2
(
)
̂
(
)
(
1
5
)
T
h
e
cr
o
s
s
-
s
p
ec
tr
al
d
e
n
s
ity
b
etwe
en
s
ig
n
als
(
)
an
d
̂
(
)
is
r
ep
r
esen
ted
b
y
̂
(
)
,
wh
er
e
(
)
an
d
̂
(
)
ar
e
th
e
co
r
r
esp
o
n
d
in
g
a
u
to
s
p
ec
tr
a.
T
ab
le
1
d
is
p
lay
s
th
e
r
esu
lts
o
f
a
co
m
p
ar
is
o
n
b
etwe
en
L
MS
-
b
ased
an
d
ME
E
-
b
ased
alg
o
r
ith
m
s
ap
p
lied
to
f
iv
e
s
ep
ar
ate
r
ec
o
r
d
in
g
s
o
f
elec
tr
o
en
ce
p
h
al
o
g
r
ap
h
ic
(
E
E
G)
ac
tiv
ity
.
As ca
n
b
e
s
ee
n
in
th
is
tab
le
th
e
SNR
o
f
th
e
d
e
-
n
o
is
ed
E
E
G
s
ig
n
al
av
e
r
ag
es
at
3
.
3
2
d
B
wh
en
ap
p
ly
i
n
g
th
e
L
MS
m
eth
o
d
a
n
d
r
is
es
to
4
.
7
2
d
B
wh
en
u
tili
zin
g
m
ee
m
eth
o
d
.
T
h
ese
r
esu
lts
s
h
o
w
th
e
s
ig
n
if
ican
t
s
u
p
er
io
r
ity
o
f
th
e
ME
E
alg
o
r
it
h
m
o
v
er
th
e
L
MS
alg
o
r
ith
m
in
te
r
m
s
o
f
im
p
r
o
v
i
n
g
th
e
SNR
r
ate
.
T
ab
le
1
.
C
o
m
p
a
r
in
g
th
e
SNR
b
etwe
en
th
e
L
MS
an
d
th
e
ME
E
m
eth
o
d
s
S
N
R
o
u
t
(
d
B
)
R
M
S
E
D
W
T_
L
M
S
D
W
T_
M
EE
D
W
T_
L
M
S
D
W
T_
M
EE
S
u
b
j
e
c
t
1
0
.
4
2
1
.
6
9
0
.
1
0
0
.
0
9
S
u
b
j
e
c
t
2
2
.
6
0
2
.
9
8
0
.
1
5
0
.
1
4
S
u
b
j
e
c
t
3
5
.
8
7
6
.
5
2
0
.
1
1
0
.
1
0
S
u
b
j
e
c
t
4
2
.
2
7
5
.
2
5
0
.
3
0
0
.
2
1
S
u
b
j
e
c
t
5
5
.
4
4
7
.
1
6
0
.
1
5
0
.
1
2
A
v
e
r
a
g
e
3
.
3
2
4
.
7
2
0
.
1
6
0
.
1
3
T
h
e
f
in
d
in
g
s
illu
s
tr
ate
th
at
th
e
ME
E
ap
p
r
o
ac
h
s
u
r
p
ass
es
th
e
L
MS
alg
o
r
ith
m
i
n
ef
f
ec
tiv
el
y
r
em
o
v
in
g
th
e
E
OG
ar
tifa
ct.
Ou
r
ca
lc
u
latio
n
s
in
th
is
p
ap
e
r
ar
e
b
ased
o
n
a
d
ataset
o
f
2
0
0
0
E
E
G
s
am
p
le
s
,
em
p
lo
y
in
g
a
s
tep
s
ize
co
ef
f
icien
t
(
μ
)
o
f
0
.
0
1
i
n
L
MS
an
d
a
v
alu
e
o
f
1
f
o
r
ME
E
.
Ad
d
itio
n
ally
s
h
o
wn
a
r
e
th
e
o
u
tco
m
es
o
f
a
s
im
u
lated
E
E
G
with
a
s
ig
n
al
-
to
-
n
o
is
e
r
atio
o
f
-
2
d
B
d
u
e
to
c
o
n
tam
in
atio
n
.
Fig
u
r
e
5
d
is
p
la
y
s
th
e
ef
f
ec
tiv
en
ess
o
f
th
e
ME
E
an
d
L
MS
alg
o
r
it
h
m
s
in
r
em
o
v
in
g
n
o
is
e.
Fig
u
r
e
5
(
a)
s
h
o
ws
d
en
o
is
ed
s
ig
n
al
u
s
in
g
DW
T
-
L
MS
alg
o
r
ith
m
an
d
Fig
u
r
e
5
(
b
)
s
h
o
ws
d
en
o
is
ed
s
ig
n
al
u
s
in
g
DW
T
-
ME
E
m
eth
o
d
.
I
n
ad
d
itio
n
,
Fig
u
r
e
6
s
h
o
wca
s
es
th
e
PS
D
b
o
th
p
r
io
r
to
an
d
f
o
llo
win
g
f
iltra
tio
n
u
s
in
g
th
e
L
M
S
an
d
ME
E
alg
o
r
ith
m
s
.
Fig
u
r
e
6
(
a)
s
h
o
ws
PS
D
o
f
co
n
tam
in
ated
E
E
G.
Als
o
Fig
u
r
e
6
(
b
)
r
e
p
r
esen
t
f
ilter
ed
PS
D
with
L
MS,
an
d
Fig
u
r
e
6
(
c)
d
is
p
lay
f
ilter
ed
PS
D
u
s
in
g
ME
E
m
eth
o
d
.
S
a
mpl
e
s
(
a)
(
b
)
Fig
u
r
e
5
.
Den
o
is
in
g
r
esu
lts
: (
a)
d
en
o
is
ed
u
s
in
g
DW
T
-
L
MS
alg
o
r
ith
m
,
a
n
d
(
b
)
d
e
n
o
is
ed
u
s
i
n
g
DW
T
-
ME
E
m
eth
o
d
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
-2
-1
0
1
2
a
A
m
p
l
i
t
u
d
e
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
-2
-1
0
1
2
b
A
m
p
l
i
t
u
d
e
S
am
p
l
e
s
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
-2
-1
0
1
2
a
A
m
p
l
i
t
u
d
e
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
-2
-1
0
1
2
b
A
m
p
l
i
t
u
d
e
S
am
p
l
e
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
14
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
5
:
769
-
7
7
9
776
F
r
e
q
u
e
n
c
y
(
Hz)
(
a)
F
r
e
q
u
e
n
c
y
(
Hz)
(
b
)
(
c)
Fig
u
r
e
6
.
PSD
b
ef
o
r
e
an
d
af
ter
f
iltra
tio
n
(
a)
PS
D
o
f
c
o
n
tam
in
ated
E
E
G,
(
b
)
f
ilter
ed
PS
D
with
L
MS,
an
d
(
c)
f
ilter
ed
PS
D
u
s
in
g
ME
E
m
eth
o
d
F
i
g
u
r
e
7
r
e
p
r
e
s
e
n
t
s
t
h
e
p
r
o
g
r
es
s
i
o
n
o
f
M
E
E
a
n
d
L
M
S
t
e
c
h
n
i
q
u
e
s
a
s
w
el
l
as
t
h
e
d
i
s
p
a
r
it
i
es
i
n
t
h
e
t
h
e
r
e
c
o
n
s
t
r
u
c
t
e
d
a
n
d
c
l
ea
n
s
i
g
n
a
l
.
F
i
g
u
r
e
7
(
a
)
d
is
p
l
a
y
s
t
h
e
o
u
t
c
o
m
e
s
o
f
E
O
G
r
e
m
o
v
a
l
w
i
t
h
t
h
e
L
M
S
al
g
o
r
i
t
h
m
,
a
n
d
F
i
g
u
r
e
7
(
b
)
s
h
o
w
s
t
h
e
o
u
t
c
o
m
es
o
f
E
O
G
r
e
m
o
v
al
w
it
h
t
h
e
M
E
E
m
e
t
h
o
d
.
T
h
e
r
e
s
u
l
ts
s
h
o
w
t
h
at
t
h
e
M
E
E
al
g
o
r
i
t
h
m
o
u
t
p
e
r
f
o
r
m
s
t
h
e
L
M
S
a
l
g
o
r
i
t
h
m
a
t
r
e
d
u
c
i
n
g
s
t
e
a
d
y
-
s
t
a
t
e
e
r
r
o
r
.
A
ls
o
c
o
h
e
r
e
n
c
e
o
f
t
h
e
o
r
i
g
i
n
a
l
a
n
d
r
ec
o
n
s
t
r
u
c
t
e
d
s
i
g
n
a
ls
e
m
p
l
o
y
i
n
g
A
NC
f
i
l
t
e
r
w
i
t
h
L
M
S
,
a
n
d
M
E
E
al
g
o
r
i
t
h
m
a
r
e
d
i
s
p
l
a
y
e
d
i
n
F
i
g
u
r
e
s
8
(
a
)
a
n
d
8
(
b
)
,
r
esp
ec
tiv
ely
.
Su
b
s
titu
tin
g
th
e
er
r
o
r
en
tr
o
p
y
cr
iter
io
n
f
o
r
th
e
MSE
cr
iter
io
n
d
em
o
n
s
tr
ates
a
g
en
er
al
im
p
r
o
v
em
en
t
in
co
h
e
r
en
ce
v
alu
es,
p
ar
ticu
lar
ly
i
n
th
e
lo
w
er
f
r
eq
u
en
cy
r
an
g
e
.
I
n
co
n
clu
s
io
n
,
we
u
tili
ze
th
e
r
e
lativ
e
r
o
o
t
m
ea
n
s
q
u
ar
e
e
r
r
o
r
(
R
R
M
SE)
cr
iter
io
n
t
o
ass
ess
an
d
co
m
p
ar
e
th
e
ef
f
ec
tiv
en
ess
o
f
th
is
ap
p
r
o
ac
h
ag
ain
s
t o
th
er
r
ec
en
tly
i
n
tr
o
d
u
ce
d
m
eth
o
d
s
.
R
R
MSE
i
s
ca
lcu
lated
as (
1
6
)
:
=
(
−
̂
)
(
)
(
1
6
)
I
n
th
e
(
1
6
)
,
th
e
r
o
o
t
m
ea
n
s
q
u
ar
e
o
f
th
e
o
r
i
g
in
al
s
ig
n
al
d
en
o
ted
as
R
MS(
S),
wh
ile
R
MS(
̂
)
d
en
o
tes
th
e
r
o
o
t
m
ea
n
s
q
u
ar
e
o
f
clea
n
e
d
E
E
G
s
ig
n
als.
Fig
u
r
e
9
illu
s
tr
ates
an
ef
f
ec
ti
v
en
ess
co
m
p
a
r
is
o
n
o
f
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
u
s
in
g
SS
A
-
ANC
an
d
t
h
e
m
eth
o
d
f
r
o
m
[
2
7
]
,
DW
T
-
ANC,
with
r
eg
ar
d
to
th
e
R
R
MSE
m
etr
ic.
I
n
th
is
f
ig
u
r
e,
th
e
h
o
r
iz
o
n
tal
ax
is
s
h
o
ws
th
e
SNR
an
d
th
e
v
er
tical
a
x
is
s
h
o
ws th
e
R
R
MSE
in
p
er
ce
n
t.
As ca
n
b
e
s
ee
n
in
th
e
f
i
g
u
r
e,
th
e
m
eth
o
d
p
r
esen
ted
i
n
th
is
wo
r
k
in
all
SNR
s
h
as
a
lo
wer
R
R
MSE
th
an
o
th
er
m
eth
o
d
s
,
wh
ich
i
n
d
icate
s
th
e
b
etter
p
er
f
o
r
m
an
ce
o
f
t
h
e
m
eth
o
d
p
r
esen
ted
in
all
SNR
s
.
Fo
r
ex
am
p
le,
in
SNR
eq
u
al
to
2
,
th
e
R
R
MSE
cr
iter
io
n
f
o
r
th
e
p
r
o
p
o
s
ed
DW
T
-
ANS
-
C
ME
E
m
eth
o
d
is
eq
u
al
to
4
6
,
wh
ich
is
4
1
%
s
u
p
e
r
io
r
to
th
e
SS
A
-
ANC
-
R
L
S
alg
o
r
ith
m
.
T
h
is
s
h
o
ws
th
e
ef
f
ec
tiv
en
ess
o
f
th
e
p
r
o
p
o
s
ed
m
eth
o
d
in
co
n
d
itio
n
s
with
h
ig
h
n
o
is
e
p
o
wer
(
l
o
w
SNR
s
)
.
sa
mp
l
e
s
(
a)
(
b
)
Fig
u
r
e
7
.
T
h
e
er
r
o
r
s
ig
n
als af
t
er
d
en
o
is
in
g
:
(
a)
th
e
o
u
tco
m
es
o
f
E
OG
r
em
o
v
al
with
th
e
L
M
S a
lg
o
r
ith
m
,
an
d
(
b
)
th
e
o
u
tco
m
es o
f
E
OG
r
em
o
v
al
with
th
e
ME
E
m
et
h
o
d
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
E
lectro
en
ce
p
h
a
lo
g
r
a
m
d
en
o
is
in
g
u
s
in
g
d
is
crete
w
a
ve
let
tr
a
n
s
fo
r
m
a
n
d
a
d
a
p
tive
…
(
Ha
s
h
em
A
b
d
o
la
h
n
iya
)
777
Fr
e
q
u
e
n
c
y
(
Hz)
(
a)
(
b
)
Fig
u
r
e
8
.
T
h
e
co
n
s
is
ten
cy
b
etwe
en
th
e
p
r
im
a
r
y
an
d
r
ec
r
ea
te
d
s
ig
n
als:
(
a)
o
u
tc
o
m
es o
f
E
O
G
r
em
o
v
al
u
s
in
g
th
e
L
MS
m
eth
o
d
a
n
d
(
b
)
o
u
tc
o
m
es o
f
E
OG
r
em
o
v
al
u
s
in
g
th
e
ME
E
m
eth
o
d
Fig
u
r
e
9
.
E
v
alu
atio
n
o
f
th
e
SS
A
-
ANC an
d
DW
T
-
ANC m
eth
o
d
s
in
r
elatio
n
t
o
R
R
MSE
I
t
is
wo
r
th
m
en
tio
n
in
g
th
at
th
is
r
esear
ch
h
as
in
v
esti
g
ated
s
i
m
u
ltan
eo
u
s
ly
all
th
e
cr
iter
ia
o
f
R
R
M
SE
cr
iter
ia,
o
u
tp
u
t
SNR
af
ter
f
il
ter
in
g
,
er
r
o
r
am
p
litu
d
e
cu
r
v
e
af
ter
d
en
o
is
in
g
o
p
er
atio
n
,
PS
D
an
d
co
h
er
en
ce
an
aly
s
es.
W
h
ile
p
r
ev
io
u
s
s
tu
d
i
es
h
av
e
n
o
t
in
v
esti
g
a
ted
th
e
s
im
u
ltan
eo
u
s
ef
f
ec
t
o
f
all
th
ese
c
r
iter
ia,
an
d
p
r
ev
i
o
u
s
s
tu
d
ies
h
av
e
in
v
esti
g
ated
s
o
m
e
o
f
th
ese
cr
iter
ia.
we
f
o
u
n
d
t
h
at
th
e
o
u
tp
u
t
SNR
m
ea
s
u
r
e
wer
e
co
r
r
elate
d
with
th
e
R
MSE
m
ea
s
u
r
e.
T
h
e
p
r
o
p
o
s
ed
m
eth
o
d
in
th
is
s
tu
d
y
ten
d
s
to
h
av
e
an
ex
tr
em
ely
l
o
w
r
atio
o
f
R
MSE
co
m
p
ar
ed
to
o
th
e
r
p
r
ev
io
u
s
s
tu
d
ies
wh
ich
co
n
f
i
r
m
s
th
e
r
ed
u
ctio
n
o
f
f
iltra
tio
n
e
r
r
o
r
in
th
e
p
r
o
p
o
s
ed
m
eth
o
d
.
Ou
r
s
tu
d
y
s
u
g
g
ests
th
at
lo
wer
R
MSE
is
n
o
t
ass
o
ciate
d
wi
th
p
o
o
r
p
er
f
o
r
m
a
n
ce
in
o
u
tp
u
t
SNR
.
T
h
e
p
r
o
p
o
s
ed
m
eth
o
d
m
a
y
b
en
e
f
it
f
r
o
m
R
MSE
wit
h
o
u
t
ad
v
er
s
ely
o
n
o
u
tp
u
t
SNR
o
f
d
en
o
is
ed
s
ig
n
al.
in
th
is
p
ap
er
a
co
m
p
r
eh
e
n
s
iv
e
s
tu
d
y
in
E
E
G
s
ig
n
al
n
o
is
e
r
em
o
v
al
with
DW
T
-
ANC
-
ME
E
alg
o
r
ith
m
ex
p
lo
r
ed
.
Ho
wev
er
,
f
u
r
th
er
an
d
in
-
d
e
p
th
s
tu
d
ies
m
ay
b
e
n
ee
d
ed
to
co
n
f
ir
m
its
ef
f
ec
tiv
en
ess
o
n
m
o
r
e
s
am
p
les
o
f
E
E
G
s
i
g
n
als.
I
n
0
5
10
15
20
25
30
35
40
0
0
.
5
1
a
C
oh
e
r
e
n
c
e
0
5
10
15
20
25
30
35
40
0
0
.
5
1
b
C
oh
e
r
e
n
c
e
F
r
e
q
u
e
n
c
y
(
H
z
)
0
5
10
15
20
25
30
35
40
0
0
.
5
1
a
C
oh
e
r
e
n
c
e
0
5
10
15
20
25
30
35
40
0
0
.
5
1
b
C
oh
e
r
e
n
c
e
F
r
e
q
u
e
n
c
y
(
H
z
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
14
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
5
:
769
-
7
7
9
778
th
is
s
tu
d
y
,
it
was
s
h
o
wn
th
at
d
en
o
is
in
g
alg
o
r
ith
m
s
b
ased
o
n
er
r
o
r
en
tr
o
p
y
ar
e
m
o
r
e
ef
f
ic
ien
t
th
an
tr
ad
itio
n
al
m
eth
o
d
s
f
o
r
d
en
o
is
in
g
th
e
E
E
G
s
ig
n
al.
Fu
tu
r
e
s
tu
d
ies
m
a
y
i
n
v
esti
g
ate
m
o
r
e
u
p
-
to
-
d
ate
a
d
ap
tiv
e
alg
o
r
ith
m
s
f
o
r
d
en
o
is
in
g
th
e
E
E
G
s
ig
n
al.
R
e
ce
n
t
o
b
s
e
r
v
atio
n
s
in
E
E
G
s
ig
n
al
d
en
o
is
in
g
s
tu
d
ies
s
h
o
w
th
at
ad
ap
tiv
e
f
ilter
in
g
m
eth
o
d
s
ca
n
p
r
o
v
id
e
a
p
r
o
p
er
iate
p
er
f
o
r
m
a
n
ce
in
E
E
G
s
ig
n
al
d
en
o
is
in
g
with
lo
w
er
r
o
r
r
ates.
Ou
r
f
in
d
in
g
s
p
r
o
v
id
e
co
n
clu
s
iv
e
e
v
id
en
ce
t
h
at
p
r
o
p
er
u
p
d
atin
g
o
f
th
e
a
d
a
p
tiv
e
f
ilter
co
ef
f
icien
t
s
u
s
in
g
th
e
ME
E
alg
o
r
ith
m
ca
n
lead
to
b
etter
p
er
f
o
r
m
a
n
ce
o
f
th
e
ANC s
y
s
tem
.
T
h
e
r
esu
l
ts
o
b
tain
ed
in
th
is
wo
r
k
co
n
f
ir
m
th
is
is
s
u
e.
4.
CO
NCLU
SI
O
N
T
h
is
p
ap
er
in
tr
o
d
u
ce
s
a
h
y
b
r
i
d
ap
p
r
o
ac
h
em
p
lo
y
in
g
b
o
th
DW
T
an
d
ANC
f
o
r
th
e
ex
tr
ac
ti
o
n
o
f
E
OG
ar
tifa
cts
f
r
o
m
E
E
G
s
ig
n
als.
T
h
e
m
eth
o
d
o
lo
g
y
lev
er
a
g
es
DW
T
to
d
er
iv
e
an
a
p
p
r
o
x
im
atio
n
o
f
th
e
E
OG
s
ig
n
al,
wh
ich
s
er
v
es a
s
th
e
b
asis
in
p
u
t
f
o
r
th
e
ad
a
p
tiv
e
f
ilter
.
Fu
r
t
h
er
m
o
r
e,
u
s
e
o
f
t
h
e
ME
E
alg
o
r
ith
m
is
ad
v
o
ca
ted
f
o
r
n
o
is
e
r
ed
u
ctio
n
with
in
th
e
a
d
ap
tiv
e
s
y
s
tem
,
p
a
r
ticu
lar
ly
f
o
r
n
o
n
-
Gau
s
s
ian
s
ig
n
als,
wh
er
e
it
d
em
o
n
s
tr
ates
s
u
p
er
io
r
s
u
itab
ilit
y
co
m
p
ar
ed
to
MSE
-
b
ased
alg
o
r
ith
m
s
.
A
p
er
f
o
r
m
an
ce
ev
al
u
atio
n
o
f
th
e
p
r
o
p
o
s
ed
s
y
s
tem
ag
ain
s
t
MSE
-
b
ased
alg
o
r
ith
m
s
r
ev
ea
ls
its
s
u
p
er
io
r
p
er
f
o
r
m
an
ce
in
m
etr
ics
s
u
ch
as
R
R
M
SE,
SNR
o
f
o
u
tp
u
t
,
an
d
co
h
er
en
ce
a
n
aly
s
es.
RE
F
E
R
E
NC
E
S
[
1
]
D
.
G
o
r
j
a
n
,
K
.
G
r
a
ma
n
n
,
K
.
D
.
P
a
u
w
,
a
n
d
U
.
M
a
r
u
s
i
c
,
“
R
e
mo
v
a
l
o
f
m
o
v
e
m
e
n
t
-
i
n
d
u
c
e
d
E
EG
a
r
t
i
f
a
c
t
s:
C
u
r
r
e
n
t
st
a
t
e
o
f
t
h
e
a
r
t
a
n
d
g
u
i
d
e
l
i
n
e
s,
”
J
o
u
rn
a
l
o
f
N
e
u
r
a
l
En
g
i
n
e
e
ri
n
g
,
v
o
l
.
1
9
,
n
o
.
1
,
2
0
2
2
,
d
o
i
:
1
0
.
1
0
8
8
/
1
7
4
1
-
2
5
5
2
/
a
c
5
4
2
c
.
[
2
]
J.
Y
i
n
,
A
.
Li
u
,
C
.
Li
,
R
.
Q
i
a
n
,
a
n
d
X
.
C
h
e
n
,
“
F
r
e
q
u
e
n
c
y
i
n
f
o
r
ma
t
i
o
n
e
n
h
a
n
c
e
d
d
e
e
p
EEG
d
e
n
o
i
si
n
g
n
e
t
w
o
r
k
f
o
r
o
c
u
l
a
r
a
r
t
i
f
a
c
t
r
e
mo
v
a
l
,
”
I
EEE
S
e
n
s
o
rs
J
o
u
r
n
a
l
,
v
o
l
.
2
2
,
n
o
.
2
2
,
p
p
.
2
1
8
5
5
–
2
1
8
6
5
,
2
0
2
2
,
d
o
i
:
1
0
.
1
1
0
9
/
JS
EN
.
2
0
2
2
.
3
2
0
9
8
0
5
.
[
3
]
A
.
D
a
r
r
o
u
d
i
,
J.
P
a
r
c
h
a
m
i
,
K
.
R
a
z
a
v
i
,
a
n
d
G
.
S
a
r
b
i
s
h
e
i
e
,
“
E
EG
a
d
a
p
t
i
v
e
n
o
i
se
c
a
n
c
e
l
l
a
t
i
o
n
u
si
n
g
i
n
f
o
r
mat
i
o
n
t
h
e
o
r
e
t
i
c
a
p
p
r
o
a
c
h
,”
Bi
o
-
Me
d
i
c
a
l
Ma
t
e
r
i
a
l
s
a
n
d
En
g
i
n
e
e
r
i
n
g
,
v
o
l
.
2
8
,
n
o
.
4
,
p
p
.
3
2
5
–
3
3
8
,
2
0
1
7
,
d
o
i
:
1
0
.
3
2
3
3
/
B
M
E
-
1
7
1
6
8
0
.
[
4
]
V
.
S
r
i
v
a
st
a
v
a
,
“
A
n
o
p
t
i
mi
z
a
t
i
o
n
f
o
r
a
d
a
p
t
i
v
e
m
u
l
t
i
-
f
i
l
t
e
r
e
s
t
i
m
a
t
i
o
n
i
n
me
d
i
c
a
l
i
ma
g
e
s
a
n
d
EEG
b
a
se
d
s
i
g
n
a
l
d
e
n
o
i
si
n
g
,
”
B
i
o
m
e
d
i
c
a
l
S
i
g
n
a
l
Pr
o
c
e
ssi
n
g
a
n
d
C
o
n
t
ro
l
,
v
o
l
.
8
2
,
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
b
sp
c
.
2
0
2
2
.
1
0
4
5
1
3
.
[
5
]
V
.
D
a
l
a
l
a
n
d
S
.
B
h
a
i
r
a
n
n
a
w
a
r
,
“
Ef
f
i
c
i
e
n
t
d
e
-
n
o
i
si
n
g
t
e
c
h
n
i
q
u
e
f
o
r
e
l
e
c
t
r
o
e
n
c
e
p
h
a
l
o
g
r
a
m
s
i
g
n
a
l
p
r
o
c
e
ss
i
n
g
,
”
I
A
ES
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
f
Art
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
,
v
o
l
.
1
1
,
n
o
.
2
,
p
p
.
6
0
3
–
6
1
2
,
2
0
2
2
,
d
o
i
:
1
0
.
1
1
5
9
1
/
i
j
a
i
.
v
1
1
.
i
2
.
p
p
6
0
3
-
6
1
2
.
[
6
]
A
.
D
a
r
r
o
u
d
i
,
J.
P
a
r
c
h
a
mi
,
G
.
S
a
r
b
i
s
h
a
e
i
,
a
n
d
S
.
R
a
j
a
n
,
“
R
e
m
o
v
i
n
g
EC
G
n
o
i
se
f
r
o
m
su
r
f
a
c
e
EM
G
b
a
se
d
o
n
i
n
f
o
r
mat
i
o
n
t
h
e
o
r
y
,
”
i
n
El
e
c
t
r
i
c
a
l
En
g
i
n
e
e
r
i
n
g
(
I
C
EE),
I
r
a
n
i
a
n
C
o
n
f
e
re
n
c
e
o
n
,
2
0
1
8
,
p
p
.
1
4
0
3
–
1408
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
EE.
2
0
1
8
.
8
4
7
2
6
1
3
.
[
7
]
C
.
S
.
L.
P
r
a
sa
n
n
a
a
n
d
M
.
Z.
U
.
R
a
h
m
a
n
,
“
A
r
e
f
e
r
e
n
c
e
f
r
e
e
n
o
n
-
n
e
g
a
t
i
v
e
a
d
a
p
t
i
v
e
l
e
a
r
n
i
n
g
s
y
st
e
m
f
o
r
h
e
a
l
t
h
c
a
r
e
mo
n
i
t
o
r
i
n
g
a
n
d
a
d
a
p
t
i
v
e
p
h
y
s
i
o
l
o
g
i
c
a
l
a
r
t
i
f
a
c
t
e
l
i
mi
n
a
t
i
o
n
i
n
b
r
a
i
n
w
a
v
e
s,
”
H
e
a
l
t
h
c
a
re
An
a
l
y
t
i
c
s
,
v
o
l
.
4
,
p
.
1
0
0
2
2
5
,
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
h
e
a
l
t
h
.
2
0
2
3
.
1
0
0
2
2
5
.
[
8
]
N
.
A
n
s
a
r
i
,
Y
.
U
.
K
h
a
n
,
A
.
T
.
K
h
a
n
,
a
n
d
O
.
F
a
r
o
o
q
,
“
A
u
t
o
ma
t
i
c
r
e
m
o
v
a
l
o
f
o
c
u
l
a
r
a
n
d
c
a
r
d
i
a
c
a
r
t
i
f
a
c
t
s
a
n
d
r
e
t
e
n
t
i
o
n
o
f
e
m
b
e
d
d
e
d
EEG
i
n
f
o
r
ma
t
i
o
n
v
i
a
a
d
a
p
t
i
v
e
n
o
i
se
c
a
n
c
e
l
l
a
t
i
o
n
,
”
i
n
2
0
2
3
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
Re
c
e
n
t
Ad
v
a
n
c
e
s
i
n
E
l
e
c
t
ri
c
a
l
,
El
e
c
t
r
o
n
i
c
s
& D
i
g
i
t
a
l
H
e
a
l
t
h
c
a
re
T
e
c
h
n
o
l
o
g
i
e
s (
REEDCO
N
)
,
2
0
2
3
,
p
p
.
1
3
3
–
1
3
8
,
d
o
i
:
1
0
.
1
1
0
9
/
R
EED
C
O
N
5
7
5
4
4
.
2
0
2
3
.
1
0
1
5
1
1
0
1
.
[
9
]
S
.
S
h
a
r
m
a
a
n
d
U
.
S
a
t
i
j
a
,
“
A
u
t
o
mat
e
d
o
c
u
l
a
r
a
r
t
i
f
a
c
t
s
r
e
m
o
v
a
l
f
r
a
m
e
w
o
r
k
b
a
s
e
d
o
n
a
d
a
p
t
i
v
e
c
h
i
r
p
m
o
d
e
d
e
c
o
mp
o
si
t
i
o
n
,
”
I
EE
E
S
e
n
so
rs J
o
u
rn
a
l
,
v
o
l
.
2
2
,
n
o
.
6
,
p
p
.
5
8
0
6
–
5
8
1
4
,
2
0
2
2
,
d
o
i
:
1
0
.
1
1
0
9
/
JS
EN
.
2
0
2
2
.
3
1
4
7
0
1
0
.
[
1
0
]
J.
Y
e
d
u
k
o
n
d
a
l
u
a
n
d
L.
D
.
S
h
a
r
ma,
“
C
i
r
c
u
l
a
n
t
s
i
n
g
u
l
a
r
s
p
e
c
t
r
u
m
a
n
a
l
y
s
i
s
a
n
d
d
i
scre
t
e
w
a
v
e
l
e
t
t
r
a
n
sf
o
r
m
f
o
r
a
u
t
o
mat
e
d
r
e
m
o
v
a
l
o
f
EO
G
a
r
t
i
f
a
c
t
s fr
o
m
EEG
s
i
g
n
a
l
s
,
”
S
e
n
so
rs
,
v
o
l
.
2
3
,
n
o
.
3
,
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
s
2
3
0
3
1
2
3
5
.
[
1
1
]
A
.
B
h
a
t
t
a
c
h
a
r
y
y
a
,
A
.
V
e
r
m
a
,
R
.
R
a
n
t
a
,
a
n
d
R
.
B
.
P
a
c
h
o
r
i
,
“
O
c
u
l
a
r
a
r
t
i
f
a
c
t
s
e
l
i
m
i
n
a
t
i
o
n
f
r
o
m
m
u
l
t
i
v
a
r
i
a
t
e
EEG
si
g
n
a
l
u
si
n
g
f
r
e
q
u
e
n
c
y
-
s
p
a
t
i
a
l
f
i
l
t
e
r
i
n
g
,
”
I
E
EE
T
r
a
n
sa
c
t
i
o
n
s
o
n
C
o
g
n
i
t
i
v
e
a
n
d
D
e
v
e
l
o
p
m
e
n
t
a
l
S
y
st
e
m
s
,
v
o
l
.
1
5
,
n
o
.
3
,
p
p
.
1
5
4
7
–
1
5
5
9
,
2
0
2
3
,
d
o
i
:
1
0
.
1
1
0
9
/
T
C
D
S
.
2
0
2
2
.
3
2
2
6
7
7
5
.
[
1
2
]
A
.
K
.
M
a
d
d
i
r
a
l
a
a
n
d
K
.
C
.
V
e
l
u
v
o
l
u
,
“
S
S
A
w
i
t
h
C
W
T
a
n
d
k
-
mea
n
s
f
o
r
e
y
e
-
b
l
i
n
k
a
r
t
i
f
a
c
t
r
e
mo
v
a
l
f
r
o
m
s
i
n
g
l
e
-
c
h
a
n
n
e
l
E
EG
si
g
n
a
l
s,
”
S
e
n
so
rs
,
v
o
l
.
2
2
,
n
o
.
3
,
2
0
2
2
,
d
o
i
:
1
0
.
3
3
9
0
/
s
2
2
0
3
0
9
3
1
.
[
1
3
]
H
.
H
u
,
Z.
P
u
,
a
n
d
P
.
W
a
n
g
,
“
A
f
l
e
x
i
b
l
e
a
n
d
a
c
c
u
r
a
t
e
me
t
h
o
d
f
o
r
e
l
e
c
t
r
o
e
n
c
e
p
h
a
l
o
g
r
a
p
h
y
r
h
y
t
h
ms
e
x
t
r
a
c
t
i
o
n
b
a
se
d
o
n
c
i
r
c
u
l
a
n
t
si
n
g
u
l
a
r
sp
e
c
t
r
u
m
a
n
a
l
y
s
i
s,”
Pe
e
rJ
,
v
o
l
.
1
0
,
2
0
2
2
,
d
o
i
:
1
0
.
7
7
1
7
/
p
e
e
r
j
.
1
3
0
9
6
.
[
1
4
]
C
.
M
.
B
i
sh
o
p
,
N
e
u
r
a
l
n
e
t
w
o
r
k
s f
o
r
p
a
t
t
e
r
n
r
e
c
o
g
n
i
t
i
o
n
.
O
x
f
o
r
d
,
U
n
i
t
e
d
K
i
n
g
d
o
m
:
C
l
a
r
e
n
d
o
n
P
r
e
ss,
1
9
9
5
.
[
1
5
]
M
.
T.
A
k
h
t
a
r
,
W
.
M
i
t
s
u
h
a
s
h
i
,
a
n
d
C
.
J
.
Jam
e
s,
“
Em
p
l
o
y
i
n
g
sp
a
t
i
a
l
l
y
c
o
n
st
r
a
i
n
e
d
I
C
A
a
n
d
w
a
v
e
l
e
t
d
e
n
o
i
s
i
n
g
,
f
o
r
a
u
t
o
m
a
t
i
c
r
e
mo
v
a
l
o
f
a
r
t
i
f
a
c
t
s
f
r
o
m
m
u
l
t
i
c
h
a
n
n
e
l
EEG
d
a
t
a
,
”
S
i
g
n
a
l
Pr
o
c
e
ss
i
n
g
,
v
o
l
.
9
2
,
n
o
.
2
,
p
p
.
4
0
1
–
4
1
6
,
2
0
1
2
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
s
i
g
p
r
o
.
2
0
1
1
.
0
8
.
0
0
5
.
[
1
6
]
S
.
S
a
n
e
i
,
T.
K
.
M
.
Le
e
,
a
n
d
V
.
A
b
o
l
g
h
a
sem
i
,
“
A
n
e
w
a
d
a
p
t
i
v
e
l
i
n
e
e
n
h
a
n
c
e
r
b
a
se
d
o
n
s
i
n
g
u
l
a
r
s
p
e
c
t
r
u
m
a
n
a
l
y
si
s
,
”
I
EE
E
T
ra
n
s
a
c
t
i
o
n
s
o
n
B
i
o
m
e
d
i
c
a
l
En
g
i
n
e
e
r
i
n
g
,
v
o
l
.
5
9
,
n
o
.
2
,
p
p
.
4
2
8
–
4
3
4
,
2
0
1
2
,
d
o
i
:
1
0
.
1
1
0
9
/
T
B
M
E.
2
0
1
1
.
2
1
7
3
9
3
6
.
[
1
7
]
H
.
A
b
d
o
l
l
a
h
n
i
y
a
,
A
.
A
.
K
h
a
z
a
e
i
,
a
n
d
M
.
A
z
a
r
n
o
o
s
h
,
“
A
D
W
T
-
A
N
C
e
r
r
o
r
e
n
t
r
o
p
y
c
r
i
t
e
r
i
o
n
b
a
s
e
d
si
n
g
l
e
-
c
h
a
n
n
e
l
EEG
si
g
n
a
l
EO
G
n
o
i
se
r
e
d
u
c
t
i
o
n
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
f
N
o
n
l
i
n
e
a
r
A
n
a
l
y
s
i
s
a
n
d
Ap
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
1
4
,
p
p
.
6
6
7
–
6
7
7
,
2
0
2
3
,
d
o
i
:
1
0
.
2
2
0
7
5
/
i
j
n
a
a
.
2
0
2
2
.
2
6
4
7
4
.
3
3
2
6
.
[
1
8
]
D
.
Er
d
o
g
m
u
s a
n
d
J
.
C
.
P
r
i
n
c
i
p
e
,
“
A
n
e
r
r
o
r
-
e
n
t
r
o
p
y
m
i
n
i
mi
z
a
t
i
o
n
a
l
g
o
r
i
t
h
m
f
o
r
su
p
e
r
v
i
se
d
t
r
a
i
n
i
n
g
o
f
n
o
n
l
i
n
e
a
r
a
d
a
p
t
i
v
e
sy
s
t
e
ms
,
”
I
EEE
T
r
a
n
s
a
c
t
i
o
n
s
o
n
S
i
g
n
a
l
Pr
o
c
e
ss
i
n
g
,
v
o
l
.
5
0
,
n
o
.
7
,
p
p
.
1
7
8
0
–
1
7
8
6
,
2
0
0
2
,
d
o
i
:
1
0
.
1
1
0
9
/
TSP
.
2
0
0
2
.
1
0
1
1
2
1
7
.
[
1
9
]
M
.
Li
,
Z.
J
i
n
g
,
a
n
d
H
.
L
e
u
n
g
,
“
R
o
b
u
st
mi
n
i
mu
m
e
r
r
o
r
e
n
t
r
o
p
y
b
a
se
d
c
u
b
a
t
u
r
e
i
n
f
o
r
ma
t
i
o
n
f
i
l
t
e
r
w
i
t
h
n
o
n
-
g
a
u
ss
i
a
n
me
a
su
r
e
me
n
t
n
o
i
se
,
”
I
E
EE
S
i
g
n
a
l
Pro
c
e
ss
i
n
g
L
e
t
t
e
rs
,
v
o
l
.
2
8
,
p
p
.
3
4
9
–
3
5
3
,
2
0
2
1
,
d
o
i
:
1
0
.
1
1
0
9
/
LSP
.
2
0
2
1
.
3
0
5
5
7
4
8
.
[
2
0
]
S
.
H
a
n
,
S
.
R
a
o
,
D
.
Er
d
o
g
m
u
s,
a
n
d
J.
P
r
i
n
c
i
p
e
,
“
A
n
i
m
p
r
o
v
e
d
m
i
n
i
m
u
m
e
r
r
o
r
e
n
t
r
o
p
y
c
r
i
t
e
r
i
o
n
w
i
t
h
sel
f
a
d
j
u
s
t
i
n
g
st
e
p
-
si
z
e
,
”
i
n
2
0
0
5
I
EEE
W
o
r
k
sh
o
p
o
n
M
a
c
h
i
n
e
L
e
a
rn
i
n
g
f
o
r
S
i
g
n
a
l
Pro
c
e
ss
i
n
g
,
2
0
0
5
,
p
p
.
3
1
7
–
3
2
2
,
d
o
i
:
1
0
.
1
1
0
9
/
M
LSP
.
2
0
0
5
.
1
5
3
2
9
2
1
.
[
2
1
]
C
.
S
a
f
a
r
i
a
n
a
n
d
T.
O
g
u
n
f
u
n
m
i
,
“
Q
u
a
t
e
r
n
i
o
n
k
e
r
n
e
l
n
o
r
m
a
l
i
z
e
d
mi
n
i
mu
m
e
r
r
o
r
e
n
t
r
o
p
y
a
d
a
p
t
i
v
e
a
l
g
o
r
i
t
h
ms,
”
i
n
2
0
1
8
Asi
a
-
P
a
c
i
f
i
c
S
i
g
n
a
l
a
n
d
I
n
f
o
rm
a
t
i
o
n
Pro
c
e
ss
i
n
g
Asso
c
i
a
t
i
o
n
A
n
n
u
a
l
S
u
m
m
i
t
a
n
d
C
o
n
f
e
re
n
c
e
(
AP
S
I
PA
A
S
C
)
,
2
0
1
8
,
p
p
.
5
7
2
–
5
7
8
,
d
o
i
:
1
0
.
2
3
9
1
9
/
A
P
S
I
P
A
.
2
0
1
8
.
8
6
5
9
7
8
7
.
[
2
2
]
M
.
K
.
Y
a
d
a
v
a
l
l
i
a
n
d
V
.
K
.
P
a
mu
l
a
,
“
A
n
e
f
f
i
c
i
e
n
t
f
r
a
mew
o
r
k
t
o
a
u
t
o
m
a
t
i
c
e
x
t
r
a
c
t
EO
G
a
r
t
i
f
a
c
t
s
f
r
o
m
s
i
n
g
l
e
c
h
a
n
n
e
l
EEG
r
e
c
o
r
d
i
n
g
s,”
i
n
2
0
2
2
I
EE
E
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
S
i
g
n
a
l
Pro
c
e
ss
i
n
g
a
n
d
C
o
m
m
u
n
i
c
a
t
i
o
n
s
(
S
PC
O
M)
,
2
0
2
2
,
p
p
.
1
–
5
,
d
o
i
:
1
0
.
1
1
0
9
/
S
P
C
O
M
5
5
3
1
6
.
2
0
2
2
.
9
8
4
0
8
4
9
.
[
2
3
]
B
.
A
z
z
e
r
b
o
n
i
,
M
.
C
a
r
p
e
n
t
i
e
r
i
,
F
.
L
.
F
o
r
e
st
a
,
a
n
d
F
.
C
.
M
o
r
a
b
i
t
o
,
“
N
e
u
r
a
l
-
I
C
A
a
n
d
w
a
v
e
l
e
t
t
r
a
n
sf
o
r
m
f
o
r
a
r
t
i
f
a
c
t
s
r
e
mo
v
a
l
i
n
s
u
r
f
a
c
e
Evaluation Warning : The document was created with Spire.PDF for Python.