I
nte
rna
t
io
na
l J
o
urna
l o
f
E
lect
rica
l a
nd
Co
m
p
ute
r
E
ng
in
ee
ring
(
I
J
E
CE
)
Vo
l.
7
,
No
.
1
,
Feb
r
u
ar
y
201
7
,
p
p
.
4
8
6
~
4
9
5
I
SS
N:
2
0
8
8
-
8708
,
DOI
: 1
0
.
1
1
5
9
1
/
i
j
ec
e
.
v7
i
1
.
pp
486
-
4
9
5
486
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ia
e
s
jo
u
r
n
a
l.c
o
m/o
n
lin
e/in
d
ex
.
p
h
p
/I
JE
C
E
Desig
n and
Analy
sis
Sys
te
m
of
K
N
N and ID3
Algo
ri
th
m
f
o
r
M
usic Cla
ss
ificati
o
n
b
a
sed
o
n Mo
o
d F
ea
tu
re Ex
trac
t
io
n
M
a
de
Su
da
r
m
a
1
,
I
G
ede
H
a
rse
m
a
di
2
1
De
p
a
rtme
n
t
o
f
El
e
c
tri
c
a
l
a
n
d
Co
m
p
u
ter E
n
g
in
e
e
rin
g
,
F
a
c
u
lt
y
o
f
En
g
in
e
e
rin
g
,
Ud
a
y
a
n
a
Un
iv
e
rsit
y
,
Ba
li
,
In
d
o
n
e
sia
2
De
p
a
rtme
n
t
o
f
In
f
o
rm
a
ti
o
n
S
y
ste
m
,
S
T
M
IK
-
S
T
IKO
M
B
a
li
,
In
d
o
n
e
sia
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Au
g
1
4
,
2
0
1
6
R
ev
i
s
ed
Oct
1
6
,
201
6
A
cc
ep
ted
No
v
1
,
2
0
1
6
Eac
h
o
f
m
u
sic
w
h
ich
h
a
s
b
e
e
n
c
re
a
ted
,
h
a
s
it
s
o
w
n
m
o
o
d
w
h
ich
is
e
m
it
ted
,
th
e
re
f
o
re
,
th
e
re
h
a
s
b
e
e
n
m
a
n
y
re
se
a
rc
h
e
s
in
M
u
sic
I
n
f
o
rm
a
ti
o
n
Re
tri
e
v
a
l
(M
IR)
f
ield
th
a
t
h
a
s
b
e
e
n
d
o
n
e
f
o
r
re
c
o
g
n
it
io
n
o
f
m
o
o
d
to
m
u
sic
.
T
h
is
re
se
a
rc
h
p
ro
d
u
c
e
d
so
f
tw
a
r
e
to
c
la
ss
ify
m
u
sic
to
th
e
m
o
o
d
b
y
u
sin
g
K
-
Ne
a
re
st
Ne
ig
h
b
o
r
a
n
d
ID3
a
lg
o
rit
h
m
.
In
th
is
re
se
a
rc
h
a
c
c
u
ra
c
y
p
e
rf
o
r
m
a
n
c
e
c
o
m
p
a
riso
n
a
n
d
m
e
a
su
re
m
e
n
t
o
f
a
v
e
ra
g
e
c
las
sif
ic
a
ti
o
n
ti
m
e
is
c
a
rried
o
u
t
w
h
ich
is
o
b
tai
n
e
d
b
a
se
d
o
n
th
e
v
a
lu
e
p
ro
d
u
c
e
d
f
ro
m
m
u
sic
f
e
a
tu
re
e
x
trac
ti
o
n
p
ro
c
e
ss
.
F
o
r
m
u
sic
f
e
a
tu
re
e
x
tr
a
c
ti
o
n
p
r
o
c
e
ss
it
u
se
s
9
t
y
p
e
s
o
f
sp
e
c
tral
a
n
a
ly
sis,
c
o
n
sists
o
f
4
0
0
p
ra
c
ti
c
in
g
d
a
ta
a
n
d
4
0
0
tes
ti
n
g
d
a
ta.
T
h
e
s
y
ste
m
p
ro
d
u
c
e
d
o
u
tco
m
e
a
s cla
ss
i
f
ica
ti
o
n
lab
e
l
o
f
m
o
o
d
ty
p
e
th
o
se
a
re
c
o
n
ten
tm
e
n
t,
e
x
u
b
e
ra
n
c
e
,
d
e
p
re
ss
io
n
a
n
d
a
n
x
io
u
s.
C
las
sif
ica
ti
o
n
b
y
u
sin
g
a
lg
o
rit
h
m
o
f
KN
N
is
g
o
o
d
e
n
o
u
g
h
t
h
a
t
is
8
6
.
5
5
%
a
t
k
v
a
lu
e
=
3
a
n
d
a
v
e
ra
g
e
p
ro
c
e
ss
in
g
ti
m
e
is
0
.
0
1
0
2
1
.
W
h
e
re
a
s
b
y
u
sin
g
ID3
it
re
su
lt
s
a
c
c
u
ra
c
y
o
f
5
9
.
3
3
%
a
n
d
a
v
e
ra
g
e
o
f
p
ro
c
e
ss
in
g
ti
m
e
is
0
.
0
5
0
9
1
se
c
o
n
d.
K
ey
w
o
r
d
:
C
lasi
f
icat
io
n
I
D3
KNN
Mo
o
d
Mu
s
ic
Co
p
y
rig
h
t
©
2
0
1
7
In
stit
u
te o
f
A
d
v
a
n
c
e
d
E
n
g
i
n
e
e
rin
g
a
n
d
S
c
ien
c
e
.
Al
l
rig
h
ts
re
se
rv
e
d
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
Ma
d
e
Su
d
ar
m
a,
Dep
ar
t
m
en
t o
f
E
lectr
ical
an
d
C
o
m
p
u
ter
E
n
g
in
ee
r
i
n
g
Facu
lt
y
o
f
E
n
g
i
n
ee
r
i
n
g
,
Ud
a
y
an
a
U
n
i
v
er
s
it
y
,
J
im
b
ar
an
,
B
ad
u
n
g
,
B
ali
,
I
n
d
o
n
esia.
E
m
ail:
i
m
as
u
d
ar
m
a
@
g
m
ail.
co
m
1.
I
NT
RO
D
UCT
I
O
N
T
h
er
e
is
d
ir
ec
t
r
elatio
n
s
h
ip
b
et
w
ee
n
m
u
s
ic
a
n
d
e
m
o
tio
n
[
1
]
.
Mu
s
ic
h
a
s
m
a
n
y
ad
v
a
n
ta
g
es
t
o
h
u
m
a
n
’
s
b
o
d
y
an
d
s
o
u
l,
m
u
s
ic
ca
n
ca
l
m
e
v
en
i
n
s
p
ir
i
n
g
th
e
p
er
s
o
n
w
h
o
lis
te
n
to
t
h
e
m
u
s
ic.
P
s
y
ch
o
lo
g
is
t
s
tated
t
h
at
th
er
e
is
p
ar
t
in
h
u
m
an
’
s
b
r
ain
t
h
at
f
ee
ls
m
u
s
ic
i
s
clo
s
e
to
th
e
b
r
ai
n
p
ar
t
th
a
t
r
elate
d
to
h
u
m
a
n
’
s
e
m
o
tio
n
al
ex
p
r
ess
io
n
.
R
esear
c
h
er
s
o
f
M
u
s
ic
I
n
f
o
r
m
atio
n
R
etr
iev
a
l
(
MI
R
)
f
ield
d
ev
elo
p
ed
alg
o
r
ith
m
to
d
ete
ct
m
o
o
d
to
m
u
s
ic
au
to
m
atica
l
l
y
.
T
h
e
y
t
r
ied
to
u
s
e
e
m
o
tio
n
i
n
m
u
s
i
c
to
g
et
co
n
v
en
tio
n
al
m
etad
ata
s
u
ch
as
s
t
y
le,
s
i
m
ilar
it
y
,
g
e
n
r
e,
an
d
m
o
o
d
.
I
n
r
esear
ch
p
er
f
o
r
m
ed
b
y
So
n
g
,
et
al.
[
2
]
s
tated
th
at
ea
ch
o
f
m
u
s
ic
cr
ea
ted
h
as
its
o
w
n
e
m
itt
ed
e
m
o
tio
n
e
n
er
g
y
.
T
h
a
y
er
[
3
]
p
r
o
p
o
s
ed
2
d
im
en
s
io
n
s
m
o
d
el
t
h
at
m
ap
p
in
g
m
o
o
d
i
n
m
u
s
ic,
th
i
s
2
d
im
e
n
s
io
n
s
ap
p
r
o
ac
h
d
esig
n
ate
th
eo
r
y
t
h
at
s
tated
th
at
m
o
o
d
is
ca
u
s
ed
b
y
t
w
o
f
ac
t
o
r
s
,
th
o
s
e
ar
e
s
tr
ess
(
h
ap
p
in
ess
a
n
d
an
x
io
u
s
)
an
d
en
er
g
y
(
r
elax
an
d
e
n
er
g
etic)
.
Fro
m
th
i
s
2
d
i
m
en
s
io
n
s
m
o
d
e
l
it
is
d
iv
id
ed
in
to
4
clu
s
ter
s
th
o
s
e
ar
e
co
n
te
n
t
m
e
n
t
,
ex
u
b
er
an
ce
,
d
ep
r
ess
io
n
,
an
d
an
x
io
u
s
.
I
n
v
ar
io
u
s
r
esear
ch
r
elate
d
to
MI
R
,
it
u
s
ed
v
ar
io
u
s
k
i
n
d
s
o
f
d
ata
m
i
n
i
n
g
m
et
h
o
d
f
o
r
g
r
o
u
p
in
g
in
cl
u
d
in
g
d
ata
cla
s
s
i
f
ica
tio
n
a
n
d
cl
u
s
ter
i
n
g
s
u
ch
as
C
4
.
5
[
4
]
,
d
ec
is
s
io
n
tr
ee
[
5
]
,
Su
p
p
o
r
t
Vec
to
r
Mo
d
el
[
1
]
,
A
r
ti
f
icial
Ne
u
r
al
Net
w
o
r
k
[
6
]
,
S
elf
Or
g
a
n
izatio
n
[
7
]
,
K
-
Me
an
s
[
8
]
,
[
9
]
,
etc.
C
la
s
s
i
f
icatio
n
p
r
o
ce
s
s
b
y
u
s
i
n
g
th
is
d
ata
m
i
n
i
n
g
al
g
o
r
ith
m
w
a
s
in
itiated
w
it
h
p
r
e
-
p
r
o
ce
s
s
in
g
s
tag
e.
I
n
th
is
s
ta
g
e
m
u
s
ic
p
ar
t
to
b
e
u
s
ed
is
th
e
r
ef
r
ain
,
w
h
ich
t
h
e
p
ar
t
w
i
th
f
r
eq
u
en
t
w
o
r
d
s
an
d
n
o
tes
r
ep
etit
io
n
an
d
th
i
s
in
t
h
e
p
ar
t
th
at
m
o
s
t
d
eter
m
in
e
m
o
o
d
t
y
p
e
in
c
lu
d
e
i
n
t
h
e
m
u
s
ic
[
1
0
]
.
T
h
is
p
ar
t
w
it
h
3
0
s
ec
o
n
d
s
d
u
r
atio
n
[
1
]
,
[
6
]
w
it
h
f
o
r
m
at
o
f
*
.
w
av
m
o
n
o
ch
an
n
el
i
s
f
u
r
th
er
m
o
r
e
p
r
o
ce
s
s
ed
b
y
u
s
i
n
g
s
i
g
n
al
p
r
o
ce
s
s
i
n
g
o
f
Fas
t
Fo
u
r
ier
T
r
an
s
f
o
r
m
(
FF
T
)
an
d
n
in
e
t
y
p
es
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
E
C
E
I
SS
N:
2
0
8
8
-
8708
Desig
n
a
n
d
A
n
a
lysi
s
S
ystem
o
f KN
N
a
n
d
I
D3
A
l
g
o
r
ith
m
fo
r
Mu
s
ic
C
la
s
s
ifica
tio
n
B
a
s
ed
…
(
Ma
d
e
S
u
d
a
r
ma
)
487
o
f
Sp
ec
tr
al
An
al
y
s
i
s
Featu
r
e
s
,
w
it
h
t
h
e
p
u
r
p
o
s
e
to
ch
a
n
g
e
s
ig
n
a
l
f
r
o
m
ti
m
e
d
o
m
ai
n
to
f
r
e
q
u
en
c
y
d
o
m
ai
n
to
m
ak
e
ea
s
y
f
ea
t
u
r
e
e
x
tr
ac
tio
n
p
r
o
ce
s
s
to
g
et
a
s
et
o
f
f
ea
tu
r
e
v
alu
e
t
h
at
w
ill
b
e
p
r
o
ce
s
s
ed
as
p
r
ac
ticin
g
d
ata
a
n
d
test
i
n
g
d
ata
i
n
clas
s
i
f
icatio
n
p
r
o
ce
s
s
u
s
in
g
K
-
Nea
r
es
t
Nei
g
h
b
o
r
(
KNN)
a
n
d
I
ter
ativ
e
D
ich
o
to
m
i
s
er
3
(
I
D3
)
alg
o
r
ith
m
.
KNN
i
s
k
n
o
w
n
a
s
s
tr
o
n
g
al
g
o
r
ith
m
i
n
clas
s
i
f
icat
io
n
p
r
o
ce
s
s
o
f
test
i
n
g
d
ata
to
p
r
ac
ticin
g
d
ata
t
h
at
h
as
m
u
c
h
n
o
is
e,
s
o
it
is
e
f
f
ec
tiv
e
if
it
u
s
es
b
ig
p
r
ac
ticin
g
d
ata.
W
h
er
ea
s
,
I
D3
al
g
o
r
ith
m
u
s
es
co
n
ce
p
t
f
r
o
m
in
f
o
r
m
atio
n
en
tr
o
p
y
,
w
h
ic
h
h
as
s
u
p
er
io
r
it
y
i
n
p
er
f
o
r
m
i
n
g
s
ea
r
c
h
i
n
g
th
r
o
u
g
h
i
n
all
e
x
is
t
in
g
cla
s
s
i
f
icat
io
n
p
o
s
s
ib
ilit
ies i
n
d
ec
is
io
n
tr
ee
.
T
h
is
r
esear
ch
h
as
p
u
r
p
o
s
e
to
b
e
a
b
le
to
b
u
ild
a
s
y
s
te
m
t
h
at
ca
n
clas
s
i
f
y
m
u
s
ic
in
to
f
o
u
r
m
o
o
d
p
ar
am
eter
s
an
d
to
f
i
n
d
o
u
t
ac
cu
r
ac
y
le
v
el
a
n
d
p
r
o
ce
s
s
in
g
ti
m
e
p
er
f
o
r
m
ed
b
y
u
s
in
g
K
NN
an
d
I
D3
alg
o
r
ith
m
.
T
h
e
s
ec
o
n
d
p
ar
t
o
f
th
is
r
esear
ch
p
ap
er
w
i
ll
d
is
cu
s
s
ab
o
u
t
liter
atu
r
e
s
t
u
d
y
,
m
e
th
o
d
o
lo
g
y
s
y
s
te
m
w
ill
b
e
d
is
cu
s
s
ed
in
t
h
e
th
ir
d
p
ar
t,
r
esear
ch
r
esu
lt
is
p
r
esen
ted
o
n
th
e
f
o
u
r
t
h
p
ar
t,
an
d
o
n
th
e
f
if
th
p
ar
t
is
th
e
co
n
clu
s
io
n
w
il
l a
n
d
s
y
s
te
m
d
e
v
elo
p
m
e
n
t
in
t
h
e
f
u
tu
r
e.
2.
CO
M
P
RE
H
E
NSI
V
E
T
H
E
O
RE
T
I
CA
L
B
ASE
T
h
e
f
o
llo
w
i
n
g
w
il
l
ex
p
lai
n
ab
o
u
t
s
u
p
p
o
r
tin
g
t
h
eo
r
ies
t
h
at
w
ill
b
ec
o
m
e
t
h
e
t
h
eo
r
y
b
ase
o
f
t
h
is
r
esear
ch
,
a
m
o
n
g
o
th
er
s
:
2
.
1
.
T
ha
y
er
’
s
2
Dim
e
ns
io
n
s
M
o
del
Mo
o
d
is
th
e
co
n
d
itio
n
o
f
e
m
o
tio
n
th
a
t
r
elati
v
el
y
last
lo
n
g
.
Mo
o
d
is
d
if
f
er
en
t
w
it
h
s
i
m
p
l
e
e
m
o
tio
n
w
h
er
e
e
m
o
tio
n
i
s
less
s
p
ec
i
f
ic
,
less
in
te
n
s
e,
an
d
les
s
lik
e
l
y
t
o
b
e
tr
ig
g
er
ed
b
y
s
ti
m
u
lu
s
o
r
c
er
tain
ev
e
n
t
[
3
]
.
I
n
1
9
8
9
R
o
b
er
t
T
h
ay
er
s
u
g
g
es
t
e
d
a
m
o
o
d
m
o
d
el
in
t
w
o
d
i
m
e
n
s
io
n
s
,
it
o
f
f
er
ed
a
s
i
m
p
le
m
e
t
h
o
d
b
u
t
ef
f
ec
t
iv
e
to
r
ep
r
esen
t th
e
m
o
o
d
.
T
h
ay
er
ad
o
p
ted
a
d
if
f
er
e
n
t
a
p
p
r
o
ac
h
f
r
o
m
He
v
n
er
’
s
m
o
d
e
l.
T
h
a
y
er
’
s
m
o
d
el
s
u
g
g
es
ted
th
at
m
o
o
d
d
ep
en
d
s
o
n
t
w
o
f
ac
to
r
s
,
th
o
s
e
ar
e:
s
tr
ess
(
h
ap
p
in
ess
an
d
an
x
iet
y
)
a
n
d
en
er
g
y
(
ca
l
m
a
n
d
en
er
g
y
)
co
m
b
i
n
ed
in
t
w
o
d
i
m
en
s
io
n
s
a
x
is
t
h
at
f
o
r
m
ed
f
o
u
r
d
if
f
er
en
t
q
u
ad
r
an
t
s
,
th
o
s
e
ar
e:
co
n
ten
t
m
e
n
t,
r
ep
r
esen
ti
n
g
m
u
s
ic
m
o
o
d
t
y
p
e
o
f
ca
l
m
a
n
d
h
ap
p
y
;
d
ep
r
ess
io
n
,
r
ep
r
esen
t
in
g
m
u
s
ic
m
o
o
d
ty
p
e
o
f
an
x
iet
y
an
d
d
ep
r
ess
io
n
;
ex
u
b
er
an
ce
,
r
ep
r
esen
tin
g
m
u
s
ic
m
o
o
d
t
y
p
e
th
at
r
ef
er
s
to
h
ap
p
in
es
s
a
n
d
e
n
er
g
etic
;
a
n
d
an
x
iet
y
,
r
ep
r
ese
n
ti
n
g
m
u
s
ic
t
y
p
e
o
f
p
an
ic,
an
x
iet
y
a
n
d
co
n
f
u
s
ed
.
Fo
r
m
o
r
e
d
etail
ca
n
b
e
s
ee
n
in
Fig
u
r
e
1.
Fig
u
r
e
1
.
T
h
ay
er
’
s
e
m
o
t
io
n
d
i
m
en
s
io
n
m
o
d
el
[
3
]
On
e
o
f
w
ea
k
n
es
s
es
i
n
th
i
s
m
o
d
el
is
th
e
lo
w
o
f
g
r
an
u
lar
it
y
w
h
ic
h
is
les
s
ab
le
to
d
ef
in
e
d
if
f
er
en
t
e
m
o
tio
n
.
On
o
th
er
s
id
e,
th
e
s
u
p
er
io
r
it
y
o
f
th
is
m
o
d
el
i
s
t
h
e
s
i
m
p
licit
y
i
n
ca
te
g
o
r
izin
g
m
o
o
d
th
at
ca
n
lo
w
er
m
ea
n
in
g
a
m
b
i
g
u
it
y
.
2.
2
.
F
a
s
t
F
o
urier
T
ra
ns
f
o
r
m
T
h
e
F
a
s
t
F
o
u
r
ier
Tr
a
n
s
fo
r
m
is
an
alg
o
r
ith
m
to
co
u
n
t
tr
an
s
f
o
r
m
at
io
n
o
f
Fo
u
r
ier
d
is
cr
ete
q
u
ick
l
y
a
n
d
ef
f
icien
tl
y
.
D
u
e
to
co
n
tin
u
o
u
s
s
ig
n
al
s
in
co
m
m
u
n
icatio
n
s
y
s
te
m
,
f
o
r
co
n
ti
n
u
o
u
s
s
ig
n
al
s
ca
s
e
s
u
c
h
as
au
d
io
s
ig
n
al
ca
n
u
s
e
Fo
u
r
ier
tr
an
s
f
o
r
m
atio
n
.
T
h
e
Fo
u
r
ier
tr
an
s
f
o
r
m
atio
n
is
d
ef
i
n
ed
b
y
E
q
u
atio
n
1
:
(
)
∫
(
)
(
1
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
J
E
C
E
Vo
l.
7
,
No
.
1
,
Feb
r
u
ar
y
2
0
1
7
:
4
8
6
–
495
488
w
h
er
e
s
(
f
)
i
s
s
i
g
n
al
i
n
f
r
eq
u
en
c
y
d
o
m
ain
,
s
(
t)
is
s
i
g
n
a
l
in
ti
m
e
d
o
m
ai
n
,
an
d
is
co
n
s
ta
n
t
o
f
a
s
ig
n
al
’
s
v
alu
e,
f
is
f
r
eq
u
e
n
c
y
a
n
d
t is ti
m
e.
FF
T
(
F
a
s
t
F
o
u
r
ier
Tr
a
n
s
fo
r
m
)
is
o
n
e
o
f
m
et
h
o
d
s
f
o
r
tr
an
s
f
o
r
m
at
io
n
o
f
au
d
io
s
i
g
n
al
in
ti
m
e
d
o
m
ai
n
in
to
s
i
g
n
al
i
n
f
r
eq
u
e
n
c
y
d
o
m
a
in
,
w
h
ic
h
m
ea
n
s
t
h
at
t
h
e
au
d
i
o
r
ec
o
r
d
in
g
p
r
o
ce
s
s
is
k
ep
t
in
d
ig
ital
f
o
r
m
i
n
th
e
f
o
r
m
o
f
a
u
d
io
s
p
ec
tr
u
m
w
av
e
w
it
h
f
r
eq
u
en
c
y
b
ase
s
o
i
t
i
s
ea
s
ier
in
an
al
y
zi
n
g
r
ec
o
r
d
ed
a
u
d
io
f
r
eq
u
e
n
c
y
s
p
ec
tr
u
m
.
O
n
t
h
e
o
th
er
h
an
d
,
th
is
F
FT
i
m
p
le
m
e
n
tatio
n
h
elp
s
in
t
h
e
p
r
o
ce
s
s
in
g
o
f
f
ilt
er
in
g
i
n
p
u
t
s
i
g
n
a
l
p
r
o
p
er
ly
to
b
e
f
r
eq
u
en
c
y
s
i
g
n
a
l
[
1
1
]
,
[
1
2
]
.
2.
3
.
F
e
a
t
ure
E
x
t
ra
ct
io
n
T
h
e
Featu
r
e
E
x
tr
ac
tio
n
is
a
p
r
o
ce
s
s
ca
r
r
ied
o
u
t
to
ta
k
e
c
h
ar
ac
ter
is
tics
o
f
a
n
in
p
u
t
d
ata.
F
u
r
th
er
m
o
r
e,
th
is
d
ata
w
ill
b
ec
o
m
e
b
ase
to
d
o
a
ce
r
tain
j
o
b
in
a
d
if
f
er
e
n
t
s
tag
e.
T
h
e
f
ea
tu
r
e
s
p
r
o
d
u
ce
d
w
il
l
d
eter
m
i
n
e
cla
s
s
o
f
in
co
m
in
g
i
n
p
u
t
s
ig
n
al.
Fea
tu
r
e
ex
tr
ac
tio
n
in
v
o
lv
e
s
i
n
p
u
t
an
al
y
s
is
f
r
o
m
a
u
d
io
s
ig
n
al.
I
n
Mu
s
ic
I
n
fo
r
ma
ti
on
R
etri
ev
a
l,
s
o
m
e
r
esear
ch
er
s
ag
r
ee
d
th
at
f
ea
t
u
r
e
e
x
tr
ac
tio
n
p
la
y
m
o
r
e
i
m
p
o
r
tan
t
r
o
le
f
r
o
m
o
t
h
er
p
h
a
s
es
w
h
et
h
er
f
o
r
m
u
s
ic
class
i
f
icat
i
o
n
p
u
r
p
o
s
e
o
f
f
o
r
m
u
s
ic
in
tr
o
d
u
ctio
n
p
u
r
p
o
s
e.
So
m
e
f
ea
t
u
r
e
ex
tr
ac
tio
n
m
et
h
o
d
u
s
ed
in
t
h
i
s
r
esear
ch
a
m
o
n
g
o
th
er
s
[
1
3
]
:
First,
b
ased
o
n
s
ta
tis
t
ical
p
r
o
p
er
t
y
f
r
o
m
a
u
d
io
s
i
g
n
al,
w
h
er
e
au
d
io
f
ea
tu
r
e
is
an
al
y
ze
d
b
as
ed
o
n
au
d
io
s
i
g
n
al
b
lo
ck
’
s
le
n
g
t
h
an
d
to
n
e
le
v
e
l
r
esu
lted
f
r
o
m
t
h
e
ex
tr
ac
tio
n
p
r
o
ce
s
s
.
I
n
t
h
is
ca
s
e
a
u
d
io
f
ea
t
u
r
e
v
al
u
e
w
as
o
b
tain
ed
b
y
u
s
i
n
g
s
p
ec
tr
a
l S
ke
w
n
ess
a
n
d
ku
r
to
s
is
a
n
al
y
s
i
s
.
1.
S
p
ec
tr
a
l
S
ke
w
n
es
is
asy
m
m
etr
ic
p
r
o
b
ab
ilit
y
d
is
tr
ib
u
tio
n
m
ea
s
u
r
e
m
en
t
f
r
o
m
r
ea
l
v
alu
e
r
an
d
o
m
v
ar
iab
le
in
th
is
ca
s
e
is
t
h
e
a
u
d
io
s
i
g
n
al
s
p
ec
tr
u
m
.
T
h
is
f
ea
tu
r
e
s
h
o
w
s
w
h
et
h
er
t
h
er
e
is
o
r
t
h
er
e
i
s
n
o
t
s
p
ec
tr
u
m
t
h
at
s
k
e
w
ed
to
w
ar
d
s
a
v
er
ag
e
r
an
g
e
o
f
ce
r
tai
n
ar
it
h
m
etic
v
al
u
e.
Fo
r
s
a
m
p
le
o
f
N
v
al
u
e
it
f
o
r
m
s
E
q
u
atio
n
2
,
w
h
er
e
s
k
e
w
n
e
s
s
v
al
u
e
is
:
⁄
∑
(
(
)
̅
)
(
∑
(
(
)
̅
)
)
⁄
(
2
)
I
n
ab
o
v
e
E
q
u
at
io
n
̅
s
h
o
w
s
t
h
e
av
er
ag
e
o
f
m
a
g
n
itu
d
e,
is
s
a
m
p
le
at
th
ir
d
s
p
ec
tr
u
m
ce
n
ter
,
an
d
is
v
ar
ian
t
s
a
m
p
le.
2.
S
p
ec
tr
a
l
K
u
r
to
s
is
s
h
o
w
s
ce
n
tr
aliza
tio
n
o
f
au
d
io
s
p
ec
tr
u
m
t
h
at
ca
n
b
e
u
s
ed
to
s
h
o
w
to
n
e
le
v
el
co
u
n
ted
at
f
o
u
r
t
h
iter
atio
n
.
T
o
co
u
n
t
s
p
ec
tr
a
l k
u
r
to
s
is
it
u
s
e
s
E
q
u
atio
n
3
an
d
E
q
u
atio
n
4
th
o
s
e
ar
e:
∫
(
)
(
)
(
3
)
An
d
k
u
r
to
s
is
v
al
u
e
is
:
(
4
)
w
h
er
e
,
=
m
ea
n
a
n
d
=
d
ev
iati
o
n
s
tan
d
ar
d
.
Seco
n
d
,
au
d
io
f
ea
t
u
r
e
is
o
b
tai
n
ed
b
ased
o
n
s
p
ec
tr
al
s
h
ap
e,
in
t
h
is
ca
s
e
it
ca
n
b
e
f
o
u
n
d
o
u
t
b
ased
o
n
th
e
ti
m
b
r
e
(
s
o
u
n
d
/a
u
d
io
co
lo
r
)
,
p
itch
(
h
ig
h
-
lo
w
o
f
to
n
e)
,
an
d
lo
u
d
n
e
s
s
(
p
o
w
er
f
u
l
-
w
ea
k
o
f
s
o
u
n
d
)
.
T
o
g
et
au
d
io
f
ea
t
u
r
e
v
al
u
e
i
n
t
h
is
s
p
e
ctr
al
s
h
ap
e
it
i
s
o
b
tain
ed
b
y
s
p
ec
tr
a
l
ce
n
tr
o
id
,
r
o
llo
ff,
s
lo
p
e,
s
p
r
ea
d
,
d
ec
r
ea
s
e,
an
d
flu
x
.
1.
S
p
ec
tr
a
l
C
en
tr
o
id
is
t
h
e
e
x
is
tin
g
s
p
ec
tr
u
m
b
r
i
g
h
t
n
e
s
s
i
n
d
i
ca
to
r
,
an
d
it
s
h
o
w
s
g
r
av
it
y
c
en
tr
e
s
p
ec
tr
al.
Sp
ec
tr
al
ce
n
tr
o
id
s
h
o
w
s
s
o
u
n
d
’
s
clar
it
y
le
v
el.
Sp
ec
tr
al
ce
n
tr
o
id
ca
n
b
e
co
u
n
ted
b
y
u
s
i
n
g
E
q
u
atio
n
5
,
6
,
7
as th
e
f
o
llo
w
i
n
g
:
∫
(
)
(
5
)
W
h
er
e
to
co
u
n
t
p
(
f
)
:
(
)
(
)
∑
(
)
(
6
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
E
C
E
I
SS
N:
2
0
8
8
-
8708
Desig
n
a
n
d
A
n
a
lysi
s
S
ystem
o
f KN
N
a
n
d
I
D3
A
l
g
o
r
ith
m
fo
r
Mu
s
ic
C
la
s
s
ifica
tio
n
B
a
s
ed
…
(
Ma
d
e
S
u
d
a
r
ma
)
489
An
d
to
co
u
n
t
A
(
f
)
:
(
)
|
[
(
)
]
|
(
7
)
T
h
e
x
v
ar
iab
le
is
f
r
eq
u
e
n
c
y
f
r
o
m
co
llected
s
a
m
p
le.
W
h
er
ea
s
v
ar
iab
le
p
(
f)
is
p
r
o
b
ab
ilit
y
to
o
b
s
er
v
e
f
.
2.
S
p
ec
tr
a
l
R
o
llo
ff
is
b
an
d
w
id
th
m
ea
s
u
r
e
f
r
o
m
n
b
lo
ck
w
h
ic
h
is
an
a
l
y
ze
d
f
r
o
m
au
d
io
s
a
m
p
le.
Sp
ec
tr
al
R
o
llo
f
i
s
d
e
f
in
ed
a
s
f
r
eq
u
e
n
c
y
b
elo
w
ac
c
u
m
u
lat
io
n
o
f
ST
F
T
(
s
h
o
r
t
time
F
o
u
r
ier
t
r
a
n
s
fo
r
m
)
p
o
w
er
t
h
at
r
ea
ch
p
er
ce
n
tag
e
o
f
ce
r
tai
n
v
alu
e
b
et
w
ee
n
0
.
8
5
(
8
5
%)
o
r
0
.
9
5
(
9
5
%).
E
q
u
atio
n
8
f
o
r
s
p
ec
tr
al
r
o
llo
f
f
is
as th
e
f
o
llo
w
in
g
:
∑
|
(
)
|
∑
|
(
)
|
⁄
(
8
)
3.
S
p
ec
tr
a
l
S
lo
p
e
r
e
p
r
esen
ts
a
m
o
u
n
t
o
f
s
lo
p
e
f
r
o
m
e
n
er
g
y
s
p
ec
tr
al
as
f
r
eq
u
e
n
c
y
f
u
n
ctio
n
,
w
it
h
as
s
u
m
p
tio
n
th
at
a
m
p
l
itu
d
e
s
p
ec
tr
u
m
f
o
llo
w
s
th
e
li
n
ea
r
m
o
d
el,
w
h
ich
i
s
:
(
)
(
9
)
S
lo
p
e
m
is
co
u
n
ted
b
y
u
s
in
g
li
n
ea
r
r
eg
r
ess
io
n
o
f
E
q
u
atio
n
10
∑
(
)
∑
∑
(
)
∑
(
∑
)
(
1
0
)
I
n
th
e
E
q
u
atio
n
10
K
is
to
tal
am
o
u
n
t
o
f
f
r
eq
u
en
c
y
v
al
u
e,
A
(
k)
is
s
p
ec
tr
al
m
a
g
n
i
tu
d
e
w
it
h
f
r
eq
u
en
c
y
in
d
ex
k.
4.
Sp
ec
tr
al
Sp
r
ea
d
,
also
o
f
ten
c
alled
as
in
s
tan
t
o
r
a
m
o
m
e
n
t
b
an
d
w
id
t
h
s
p
ec
tr
al,
t
h
is
s
p
ec
tr
al
d
escr
ib
es
co
n
ce
n
tr
atio
n
o
f
s
p
ec
tr
u
m
p
o
w
er
ar
o
u
n
d
s
p
ec
tr
al
ce
n
tr
o
id
.
I
t
ca
n
b
e
m
ea
n
t
as
d
ev
iatio
n
s
tan
d
ar
d
v
alu
e
o
f
s
p
ec
tr
u
m
p
o
w
er
ar
o
u
n
d
s
p
e
ctr
al
ce
n
tr
o
id
.
E
q
u
atio
n
1
1
as th
e
f
o
llo
w
i
n
g
:
√
∑
(
(
)
)
|
(
)
|
∑
|
(
)
|
(
1
1
)
5.
Sp
ec
tr
al
Dec
r
ea
s
e
is
a
m
o
u
n
t
o
f
d
ec
r
ea
s
e
o
f
s
p
ec
tr
al
am
p
lit
u
d
e.
T
h
is
E
q
u
atio
n
ca
m
e
f
r
o
m
p
er
ce
p
tio
n
an
d
m
o
r
e
r
elate
d
to
f
r
eq
u
e
n
c
y
ac
c
o
r
d
in
g
to
h
u
m
a
n
’
s
p
er
ce
p
tio
n
.
T
h
e
f
o
r
m
u
la
is
:
∑
(
)
∑
(
)
(
)
(
1
2
)
I
n
th
i
s
E
q
u
atio
n
12
(
)
r
ep
r
esen
ts
f
r
eq
u
en
c
y
o
r
m
a
g
n
i
tu
d
e
q
u
ali
t
y
v
al
u
e
o
f
v
al
u
e.
6.
S
p
ec
tr
a
l
F
lu
x
is
u
s
ed
to
m
ea
s
u
r
e
ch
a
n
g
e
o
f
s
p
ec
tr
al
s
h
ap
e
,
w
h
ich
is
d
ef
i
n
ed
as
d
i
f
f
er
e
n
ce
o
f
av
er
ag
e
b
et
w
ee
n
f
r
a
m
es o
n
a
u
d
io
s
p
ec
tr
u
m
r
esp
ec
ti
v
el
y
.
√
∑
(
|
(
)
|
|
(
)
|
)
⁄
⁄
(
1
3
)
S
p
ec
tr
a
l flu
x
ca
n
b
e
m
ea
n
t a
s
b
asic a
p
p
r
o
ac
h
to
b
e
a
b
le
to
s
e
e
r
o
u
g
h
n
es
s
lev
e
l o
n
au
d
io
s
p
e
ctr
u
m
.
T
h
ir
d
,
au
d
io
f
ea
tu
r
e
is
o
b
tain
ed
b
ased
o
n
au
d
io
’
s
s
i
g
n
al
p
r
o
p
er
ties
,
w
h
er
e
a
u
d
io
f
ea
tu
r
e
to
b
e
an
al
y
ze
d
is
b
ased
o
n
to
n
e
al
o
n
g
th
e
a
u
d
io
s
ig
n
al,
w
h
ic
h
d
escr
ib
es
h
ar
m
o
n
ie
s
in
m
u
s
i
c.
T
o
o
b
tain
au
d
io
f
ea
t
u
r
e
v
al
u
e
b
ased
o
n
th
i
s
s
i
g
n
al
p
r
o
p
er
ties
it u
s
es
s
p
ec
tr
al
f
latn
es
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
J
E
C
E
Vo
l.
7
,
No
.
1
,
Feb
r
u
ar
y
2
0
1
7
:
4
8
6
–
495
490
7.
S
p
ec
tr
a
l
F
la
tn
ess
is
to
n
e
d
i
s
tr
i
b
u
tio
n
m
ea
s
u
r
e
a
n
d
s
p
ec
tr
al
p
o
w
er
i
n
a
u
d
io
s
p
ec
tr
u
m
.
S
p
ec
t
r
a
l
F
la
tn
ess
i
s
co
u
n
ted
b
y
d
iv
id
i
n
g
a
v
er
ag
e
g
eo
m
etr
ic
o
f
s
p
ec
tr
u
m
p
o
w
er
b
y
av
er
ag
e
ar
it
h
m
etic
f
r
o
m
th
e
s
p
ec
tr
u
m
p
o
w
er
.
S
p
ec
tr
a
l F
la
tn
ess
is
u
s
ed
in
m
ea
s
u
r
e
m
en
t o
f
all
s
p
ec
t
r
u
m
s
.
E
q
u
atio
n
1
4
is
as t
h
e
f
o
l
lo
w
i
n
g
:
√
∏
(
)
∑
(
)
(
1
4
)
I
n
th
i
s
E
q
u
atio
n
,
x
(
n
)
v
al
u
e
is
m
ag
n
it
u
d
e
o
f
n
v
al
u
e
o
f
s
p
ec
tr
u
m
p
o
w
er
w
it
h
len
g
t
h
N.
2.
4
.
K
-
Nea
re
s
t
Neig
hb
o
r
(
K
-
NN)
K
-
N
ea
r
es
t
N
eig
h
b
o
r
alg
o
r
ith
m
o
r
u
s
u
all
y
ca
l
led
K
-
NN
i
s
a
d
ata
class
if
icat
io
n
m
e
th
o
d
th
at
w
o
r
k
s
r
elativ
el
y
s
i
m
p
ler
co
m
p
ar
e
to
o
th
er
d
ata
class
if
icatio
n
m
eth
o
d
.
T
h
is
alg
o
r
it
h
m
tr
ies
t
o
class
if
y
n
e
w
d
ata
w
h
ic
h
th
e
clas
s
is
s
till
u
n
k
n
o
w
n
b
y
c
h
o
o
s
in
g
d
ata
o
f
s
o
m
e
k
lo
ca
ted
to
th
e
n
ea
r
est
o
f
n
e
w
d
ata
[
1
4
]
.
T
h
e
m
o
s
t
class
o
f
n
ea
r
es
t
d
ata
o
f
s
o
m
e
k
is
ch
o
s
e
n
a
s
clas
s
p
r
ed
icted
f
o
r
th
e
n
e
w
d
ata.
Ge
n
er
all
y
,
k
is
d
eter
m
i
n
ed
in
o
d
d
am
o
u
n
ts
to
av
o
id
t
h
e
e
m
e
r
g
in
g
o
f
s
a
m
e
d
is
ta
n
ce
a
m
o
u
n
t
s
in
clas
s
i
f
icatio
n
p
r
o
ce
s
s
[
1
5
]
.
K
-
NN
ta
k
es
d
ec
is
io
n
t
h
at
n
e
w
d
ata
d
in
clu
d
es
in
C
cla
s
s
b
ased
o
n
s
o
m
e
n
ea
r
es
t
n
ei
g
h
b
o
r
o
f
d
.
I
f
d
is
tan
ce
e
u
clid
en
is
u
s
ed
as
m
ea
s
u
r
e
m
e
n
t
o
f
clo
s
e
n
es
s
th
e
n
d
w
ill
b
ec
o
m
e
t
h
e
ce
n
ter
o
f
h
y
p
er
s
p
h
er
e
w
i
t
h
r
ad
iu
s
r
eq
u
al
to
th
e
eu
clid
en
d
is
tan
ce
.
W
h
at
h
as
to
b
e
d
o
n
e
is
to
in
cr
ea
s
e
r
s
o
h
y
p
er
s
p
h
er
e
lo
ad
s
k
d
ata.
C
las
s
f
o
r
d
ata
d
is
g
i
v
e
n
b
ase
d
o
n
th
e
a
m
o
u
n
t
o
f
t
h
e
m
o
s
t
c
lass
m
e
m
b
er
s
ap
p
ea
r
in
h
y
p
er
s
p
h
er
e
ce
n
ter
ed
o
n
th
e
d
.
T
h
e
n
ea
r
o
f
f
ar
o
f
n
ei
g
h
b
o
r
u
s
u
all
y
is
co
u
n
ted
b
ased
o
n
E
u
clid
en
d
is
tan
ce
w
i
th
E
q
u
a
tio
n
15
as
th
e
f
o
llo
w
in
g
:
√
(
)
(
)
(
)
√
∑
(
)
(
1
5
)
C
las
s
i
f
icatio
n
o
f
K
-
NN
is
ca
r
r
ied
o
u
t
b
y
s
ea
r
c
h
i
n
g
o
f
k
n
u
m
b
er
s
i
f
n
ea
r
es
t
n
ei
g
h
b
o
r
an
d
ch
o
o
s
i
n
g
class
w
it
h
t
h
e
m
o
s
t
k
i
in
cla
s
s
ω
i
2.
5
.
I
t
er
a
t
iv
e
Dicho
t
o
m
izer
T
re
e
(
I
D3
)
I
D3
is
class
i
f
icatio
n
alg
o
r
it
h
m
b
y
u
s
i
n
g
r
eg
u
latio
n
in
d
u
ctio
n
m
et
h
o
d
w
h
ic
h
is
u
s
ed
to
p
r
o
d
u
ce
m
o
d
el
o
f
a
d
ata
co
llectio
n
t
h
at
is
d
e
v
elo
p
ed
b
ased
o
n
s
u
p
er
v
i
s
ed
l
ea
r
n
in
g
s
y
s
te
m
b
y
J
.
R
o
s
e
Q
u
in
la
n
in
1
9
8
6
[
1
6
]
.
B
r
ief
l
y
,
I
D3
alg
o
r
ith
m
i
s
ex
p
l
ain
ed
as th
e
f
o
llo
w
i
n
g
:
1.
T
o
tak
e
o
u
t a
ll u
n
u
s
ed
attr
ib
u
t
es a
n
d
th
e
e
n
tr
o
p
y
is
ca
lc
u
late
d
w
h
ich
i
s
co
r
r
elate
d
to
d
ata
tr
ain
i
n
g
.
2.
C
h
o
o
s
e
attr
ib
u
te
w
h
er
e
it
h
as
m
i
n
i
m
u
m
e
n
tr
o
p
y
d
ata.
3.
Ma
k
e
n
o
d
e
w
it
h
th
e
co
n
ten
t o
f
th
o
s
e
attr
ib
u
te
s
.
B
asic
m
et
h
o
d
o
f
I
D3
is
to
c
h
o
o
s
e
attr
ib
u
tes
f
o
r
class
i
f
ica
ti
o
n
b
y
u
s
i
n
g
s
tati
s
tic
m
et
h
o
d
s
tar
ted
f
r
o
m
th
e
u
p
p
er
tr
ee
.
T
h
e
m
eth
o
d
to
ch
o
o
s
e
attr
ib
u
te
is
b
y
u
s
i
n
g
s
t
atis
tical
p
r
o
p
er
t
y
ca
lled
i
n
f
o
r
m
atio
n
g
ai
n
,
w
h
ic
h
is
d
ef
i
n
ed
to
d
eter
m
i
n
e
th
e
m
ea
s
u
r
e
o
f
v
a
lu
e
o
f
a
n
attr
ib
u
t
e
.
P
r
ev
io
u
s
l
y
,
en
tr
o
p
y
v
alu
e
o
f
an
o
b
j
ec
t
class
if
ied
in
t
h
e
tr
ee
s
h
o
u
ld
b
e
test
ed
.
E
n
tr
o
p
y
is
m
ea
s
u
r
e
o
f
i
n
f
o
r
m
atio
n
t
h
eo
r
y
t
h
at
ca
n
f
in
d
o
u
t
ch
ar
ac
ter
i
s
tic
o
f
i
m
p
u
r
it
y
a
n
d
h
o
m
o
g
e
n
eit
y
o
f
d
ata
co
llectio
n
.
Fro
m
t
h
e
en
t
r
o
p
y
v
a
lu
e,
t
h
e
n
in
f
o
r
m
at
io
n
g
ain
(
I
G)
v
alu
e
o
f
ea
ch
attr
ib
u
te
s
is
co
u
n
ted
.
E
n
t
r
o
p
y
v
al
u
e
is
d
e
f
i
n
ed
as th
e
f
o
llo
w
i
n
g
:
(
)
∑
(
)
(
1
6
)
w
h
er
e
P
i
is
r
atio
o
f
C
i
class
i
n
s
et
o
f
s
a
m
p
le
d
ata
S =
{x
1
,x
2
,... x
k
}
∑
(
1
7
)
Fro
m
en
tr
o
p
y
f
o
r
m
u
la
ab
o
v
e
it
ca
n
b
e
co
n
c
lu
d
ed
t
h
at
d
ef
i
n
itio
n
o
f
e
n
tr
o
p
y
(
S)
i
s
n
u
m
b
er
o
f
b
y
te
esti
m
ated
th
at
is
n
ee
d
ed
to
b
e
ab
le
to
ex
tr
ac
t
a
class
(
+
o
r
-
)
f
r
o
m
a
n
u
m
b
er
o
f
r
an
d
o
m
d
at
a
in
a
s
a
m
p
le
r
o
o
m
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
E
C
E
I
SS
N:
2
0
8
8
-
8708
Desig
n
a
n
d
A
n
a
lysi
s
S
ystem
o
f KN
N
a
n
d
I
D3
A
l
g
o
r
ith
m
fo
r
Mu
s
ic
C
la
s
s
ifica
tio
n
B
a
s
ed
…
(
Ma
d
e
S
u
d
a
r
ma
)
491
S.
E
n
tr
o
p
y
ca
n
b
e
s
aid
a
s
n
ec
ess
it
y
o
f
b
y
te
to
s
ta
te
a
clas
s
.
T
h
e
s
m
aller
e
n
tr
o
p
y
v
al
u
e
t
h
e
b
etter
it
i
s
to
b
e
u
s
ed
in
e
x
tr
ac
ti
n
g
a
cla
s
s
.
Af
ter
e
n
tr
o
p
y
v
al
u
e
i
s
o
b
tain
e
d
f
o
r
a
d
ata
co
llectio
n
,
w
e
ca
n
m
ea
s
u
r
e
e
f
f
ec
tiv
e
n
es
s
o
f
an
a
ttrib
u
te
i
n
class
i
f
y
in
g
d
ata.
T
h
is
e
f
f
ec
ti
v
en
es
s
m
ea
s
u
r
e
i
s
ca
l
led
i
n
f
o
r
m
a
tio
n
g
ain
.
Ma
t
h
e
m
a
ticall
y
,
in
f
o
r
m
atio
n
g
ain
o
f
an
attr
ib
u
t
e
A
i
s
s
tated
as t
h
e
f
o
llo
w
i
n
g
:
(
)
(
)
∑
|
|
|
|
(
)
(
)
(
1
8
)
w
h
er
e:
a.
Qu
alit
y
|
|
|
|
is
r
atio
o
f
d
ata
w
it
h
attr
ib
u
te
v
in
s
a
m
p
le
s
et
b.
A
: a
ttrib
u
te;
V:
a
p
o
s
s
ib
le
v
al
u
e
f
o
r
attr
ib
u
te
A
c.
Valu
es (
A
)
: p
o
s
s
ib
le
s
e
t f
o
r
att
r
ib
u
te
A
d.
|S
v
|
: a
m
o
u
n
t o
f
s
a
m
p
le
f
o
r
v
al
u
e
v
e.
|
S
|
: to
tal
a
m
o
u
n
t o
f
d
ata
s
a
m
p
l
e
f.
E(S
v
)
:
en
tr
o
p
y
f
o
r
s
a
m
p
le
s
th
at
h
av
e
v
al
u
e
v
3.
RE
S
E
ARCH
M
E
T
H
O
DO
L
O
G
Y
Mu
s
ic
m
o
o
d
class
i
f
icat
io
n
s
y
s
te
m
f
lo
w
g
en
er
all
y
ca
n
b
e
s
ee
n
i
n
F
ig
u
r
e
2
.
Fro
m
a
n
u
m
b
er
o
f
o
b
tain
ed
m
o
o
d
ca
teg
o
r
y
it
ca
m
e
f
r
o
m
s
o
cial
an
n
o
tat
io
n
/ta
g
f
r
o
m
s
ite
.
Fu
r
t
h
er
m
o
r
e
it
is
c
ar
r
ied
o
u
t
m
u
s
ic
f
ile
p
r
e
-
p
r
o
ce
s
s
in
g
,
b
y
u
s
i
n
g
t
h
e
r
ef
r
ain
p
ar
t
f
r
o
m
th
e
m
u
s
ic.
M
u
s
ic
cl
ip
r
e
f
r
ain
d
u
r
atio
n
is
d
e
ter
m
i
n
ed
to
b
e
o
n
l
y
3
0
s
ec
o
n
d
s
an
d
f
u
r
t
h
er
m
o
r
e
i
t
is
k
ep
t
w
it
h
f
o
r
m
at
o
f
.
w
a
v
w
ith
m
o
n
o
au
d
io
ch
a
n
n
el.
E
x
tr
ac
tio
n
p
r
o
ce
s
s
is
s
tar
ted
b
y
ch
a
n
g
in
g
i
n
p
u
t
m
u
s
ic
f
ile
s
i
g
n
al
to
b
e
f
r
eq
u
e
n
c
y
d
o
m
ai
n
b
y
u
s
in
g
F
a
s
t
F
o
u
r
ie
r
Tr
a
n
s
fo
r
m
(
FFT
)
m
et
h
o
d
.
FF
T
is
u
s
ed
f
o
r
au
d
i
o
s
ig
n
al
tr
an
s
f
o
r
m
atio
n
in
ti
m
e
d
o
m
ain
to
b
e
s
ig
n
al
in
f
r
eq
u
en
c
y
d
o
m
ai
n
.
T
h
i
s
m
u
s
ic
f
i
le
p
r
o
d
u
ce
d
b
y
FF
T
f
u
r
th
er
m
o
r
e
en
ter
in
g
f
ea
t
u
r
e
ex
tr
ac
ti
n
g
s
ta
g
e
b
y
u
s
in
g
s
p
ec
t
r
al
an
al
y
s
i
s
.
R
esu
lt
f
r
o
m
s
p
ec
t
r
a
l a
n
a
lysi
s
(
s
p
ec
tr
a
l c
en
tr
o
i
d
,
s
p
ec
tr
a
l skewn
es
s
,
s
p
ec
tr
a
l ro
llo
ff,
s
p
ec
tr
a
l
s
lo
p
e,
s
p
ec
tr
a
l k
u
r
to
s
is
,
s
p
ec
tr
a
l
s
p
r
ea
d
,
s
p
ec
tr
a
l
d
ec
r
ea
s
e,
s
p
ec
tr
a
l
flu
x,
an
d
s
p
ec
tr
a
l
fla
tn
ess
)
i
s
a
s
et
o
f
f
ea
tu
r
e
s
et
v
al
u
e
co
n
s
is
t
s
o
f
9
attr
ib
u
te
v
al
u
e
s
f
o
r
ea
c
h
o
f
m
u
s
ic
f
ile.
T
h
ese
9
attr
i
b
u
te
v
al
u
e
s
ar
e
th
e
v
al
u
es
t
h
at
b
ec
o
m
e
s
p
ec
ial
ch
ar
ac
ter
is
tic
o
f
m
u
s
ic
f
i
le
th
at
is
u
s
ed
as
tr
ain
i
n
g
d
ata
an
d
test
in
g
d
ata
f
o
r
class
if
ica
tio
n
p
r
o
ce
s
s
u
s
i
n
g
t
h
e
KNN
an
d
I
D3
alg
o
r
ith
m
.
Fig
u
r
e
2
.
Mu
s
ic
C
las
s
if
icatio
n
S
y
s
te
m
F
lo
w
to
Mo
o
d
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
J
E
C
E
Vo
l.
7
,
No
.
1
,
Feb
r
u
ar
y
2
0
1
7
:
4
8
6
–
495
492
S
y
s
te
m
tes
tin
g
is
p
er
f
o
r
m
ed
to
m
ea
s
u
r
e
p
er
ce
n
ta
g
e
o
f
s
y
s
te
m
’
s
c
lass
if
icatio
n
ac
cu
r
ac
y
b
y
u
s
in
g
b
o
th
o
f
th
e
s
e
clas
s
i
f
icatio
n
a
lg
o
r
it
h
m
s
.
T
esti
n
g
i
s
ca
r
r
ied
o
u
t
w
it
h
b
en
c
h
m
ar
k
i
n
g
th
at
r
e
f
er
s
to
r
ef
er
en
ce
s
tan
d
ar
d
to
o
b
j
ec
t
th
at
w
ill
b
e
co
m
p
ar
e
d
.
I
n
th
is
ca
s
e
co
m
p
ili
n
g
o
f
tr
ain
i
n
g
d
ata
an
d
test
i
n
g
d
ata
h
av
e
b
ee
n
class
i
f
ied
b
ased
o
n
m
o
o
d
t
y
p
e
i
n
s
ite,
t
h
o
s
e
ar
e
co
n
ten
tmen
t
,
ex
u
b
era
n
ce
,
d
ep
r
ess
io
n
an
d
a
n
xio
u
s
.
W
ith
to
tal
o
f
tr
ai
n
i
n
g
d
ata
is
4
0
0
f
iles
f
o
r
th
e
f
o
u
r
t
h
m
o
o
d
t
y
p
e
a
n
d
test
i
n
g
d
ata
is
4
0
0
f
iles
.
W
e
c
alcu
late
t
h
e
r
atio
o
f
th
e
l
ev
el
o
f
ac
cu
r
ac
y
o
f
cla
s
s
i
f
ica
t
io
n
b
y
ca
te
g
o
r
y
m
o
o
d
u
s
i
n
g
K
-
NN
a
n
d
I
D3
alg
o
r
ith
m
d
e
f
i
n
ed
as th
e
f
o
llo
w
i
n
g
:
(
4
.
1
.
)
Fo
r
class
i
f
icat
io
n
ti
m
e
o
f
b
o
th
al
g
o
r
ith
m
s
ar
e
co
u
n
ted
w
h
e
n
u
s
er
p
u
s
h
es
t
h
e
“
cla
s
s
i
f
icati
o
n
”
b
u
tto
n
o
n
th
e
s
y
s
te
m
,
s
o
it
w
i
ll r
esu
lt
lab
el
o
f
m
u
s
ic
m
o
o
d
clas
s
i
f
ica
tio
n
.
4.
RE
SU
L
T
S AN
D
AN
AL
Y
SI
S
4
.
1
.
Cha
ra
c
t
er
is
t
ics o
f
T
ra
ini
ng
Da
t
a
F
ea
t
ure
I
n
th
e
s
e
e
x
p
er
i
m
e
n
ts
w
e
d
id
a
tr
ain
i
n
g
p
h
ase
u
s
in
g
4
0
0
au
d
io
f
iles
h
a
v
e
a
s
o
u
r
ce
o
n
a
n
n
o
tatio
n
/ta
g
o
f
m
u
s
ic
e
x
p
er
tis
e
i
n
s
ite
www
.
a
u
d
io
n
et
w
o
r
k
.
co
m
w
h
ich
ar
e
u
s
ed
i
n
th
i
s
r
esear
ch
a
s
tr
ai
n
in
g
d
ata,
w
h
ich
is
s
p
r
ea
d
to
b
e
1
0
0
au
d
io
f
iles
o
f
ca
teg
o
r
y
1
(
C
o
n
te
n
t
m
e
n
t)
,
1
0
0
au
d
io
f
iles
o
f
ca
te
g
o
r
y
2
(
E
x
u
b
er
an
ce
)
,
1
0
0
au
d
io
f
ile
s
o
f
ca
te
g
o
r
y
3
(
De
p
r
ess
io
n
)
,
an
d
1
0
0
au
d
io
f
iles
o
f
ca
te
g
o
r
y
4
(
An
x
io
u
s
)
.
E
ac
h
a
u
d
io
f
ile
u
s
ed
i
n
th
e
tr
ain
i
n
g
p
h
ase
a
n
d
class
i
f
i
ca
tio
n
h
a
s
a
d
u
r
atio
n
o
f
3
0
s
ec
o
n
d
clip
s
.
I
n
th
e
p
r
e
-
p
r
o
ce
s
s
i
n
g
s
tag
e
i
s
an
a
u
d
io
f
ile
is
co
n
v
er
ted
i
n
to
W
A
V
f
o
r
m
at
w
it
h
m
o
n
o
t
y
p
e,
s
a
m
p
lin
g
r
ate
o
f
4
4
.
1
k
Hz.
T
h
e
s
y
s
te
m
i
n
ter
f
ac
e
o
f
d
ata
tr
ain
i
n
g
d
ev
e
lo
p
ed
.
Gen
er
all
y
,
in
an
a
u
d
io
f
ile
th
at
i
s
u
s
ed
as
tr
ain
in
g
d
ata
it
h
as
1
0
attr
ib
u
t
es
v
al
u
e
s
u
ch
a
s
9
attr
ib
u
tes
f
r
o
m
s
p
ec
tr
al
a
n
al
y
s
is
a
n
d
1
attr
ib
u
te
as
co
d
e
o
f
m
u
s
ic
m
o
o
d
ca
teg
o
r
y
.
T
h
is
w
il
l
b
e
k
ep
t
in
a
tr
ain
i
n
g
d
ataset
f
ile.
4
0
0
r
ec
o
r
d
s
in
th
is
tr
ai
n
in
g
d
ata
h
as
co
m
p
ar
ati
v
e
f
ea
t
u
r
e
c
h
ar
ac
ter
is
tic
v
alu
e
s
o
f
t
h
e
t
y
p
e
o
f
m
o
o
d
.
Fig
u
r
e
3
s
h
o
w
s
th
at
in
a
tr
ai
n
in
g
d
ata
f
o
r
m
ed
it
h
a
v
e
h
ig
h
e
s
t
f
ea
tu
r
e
v
al
u
es
r
e
s
p
ec
tiv
el
y
in
s
p
ec
tr
al
r
o
llo
f
f
,
s
p
ec
tr
al
ce
n
tr
o
id
,
s
p
ec
tr
al
s
p
r
ea
d
an
d
s
p
ec
tr
al
k
u
r
to
s
i
s
.
T
h
i
s
h
ig
h
est
v
a
lu
e
in
ea
ch
o
f
m
o
o
d
ca
n
b
e
s
ee
n
in
m
o
o
d
o
f
co
n
ten
t
m
e
n
t,
ex
u
b
er
an
ce
,
an
d
an
x
io
u
s
,
w
h
er
e
as
t
h
e
lo
w
est
v
al
u
e
is
in
t
h
e
m
o
o
d
o
f
d
ep
r
ess
io
n
.
T
h
e
ex
ce
p
tio
n
i
s
i
n
s
p
ec
tr
al
k
u
r
to
s
is
w
h
er
e
t
h
e
h
ig
h
e
s
t
s
p
ec
tr
al
f
ea
t
u
r
e
v
al
u
e
s
ar
e
r
esp
ec
tiv
e
l
y
i
n
t
h
e
m
o
o
d
o
f
co
n
ten
t
m
e
n
t,
d
ep
r
ess
io
n
,
an
d
an
x
io
u
s
,
a
n
d
th
e
lo
w
est i
s
in
t
h
e
m
o
o
d
o
f
ex
u
b
er
an
ce
.
Ov
er
all
tr
a
n
i
n
i
n
g
d
ata
th
e
n
s
to
r
ed
in
a
f
ile
ca
lled
tr
ain
i
n
g
d
ataset
f
o
r
th
e
n
ex
t
p
h
ases
o
f
cla
s
s
i
f
icatio
n
p
r
o
ce
s
s
u
s
i
n
g
K
-
N
N
an
d
I
D3
.
Fig
u
r
e
3
.
C
o
m
p
ar
is
o
n
o
f
9
s
p
e
ctr
al
f
ea
tu
r
e
v
alu
e
s
to
4
m
o
o
d
ca
teg
o
r
ies
4
.
2
T
esting
Resul
t
o
f
K
-
Nea
r
est
Neig
hb
o
r
Cla
s
s
if
ica
t
io
n
C
las
s
i
f
icatio
n
in
ter
f
ac
e
h
as
f
u
n
ctio
n
a
s
i
n
ter
f
ac
e
f
o
r
u
s
er
to
p
er
f
o
r
m
clas
s
i
f
icatio
n
p
r
o
ce
s
s
b
y
u
s
i
n
g
KNN
o
r
I
D3
m
et
h
o
d
b
y
u
s
in
g
tr
ai
n
i
n
g
d
ata
s
et
w
h
ic
h
h
as
b
ee
n
in
p
u
t
p
r
ev
io
u
s
l
y
.
T
h
e
d
if
f
er
en
ce
i
s
t
h
at
in
4
1
3
8
.
5
0
9
8
2
7
.
2
8
4
6
1
2
6
1
1
.
0
0
4
-
2
.
4
4
6
8
E
-
0
7
8
4
0
.
2
3
9
2
3
8
4
0
.
0
1
7
4
0
.
0
8
7
3
1
9
3
7
0
.
0
0
0
1
3
6
7
8
2
0
.
2
6
1
1
0
7
7
9
3
9
7
7
.
0
0
4
2
4
.
0
1
8
6
6
1
1
8
8
8
.
9
3
3
-
8
.
3
4
3
2
2
E
-
0
7
6
9
3
.
4
0
0
8
3
7
5
0
.
0
7
3
0
.
0
7
8
1
2
9
0
2
0
.
0
0
0
1
4
1
0
5
5
0
.
1
7
6
4
7
3
6
8
1
7
3
5
.
8
5
2
1
2
6
.
8
3
6
0
7
6
2
7
0
.
0
6
-
3
.
0
6
2
E
-
0
7
8
1
6
.
6
2
7
3
1
5
9
5
.
5
1
6
6
0
.
0
7
4
5
2
7
2
0
.
0
0
0
1
1
3
9
1
7
0
.
0
4
8
0
9
6
1
3
5
3
2
8
9
.
3
3
0
5
2
3
.
9
1
7
3
4
1
1
2
1
9
.
3
5
5
-
9
.
5
4
1
5
5
E
-
0
7
6
8
5
.
8
6
3
3
2
7
.
8
5
7
0
.
0
7
5
4
7
6
0
1
0
.
0
0
0
1
3
2
1
4
3
0
.
1
8
4
8
2
3
9
1
C
E
N
T
R
O
I
D
S
K
E
W
N
E
S
S
R
O
L
L
O
F
F
S
L
O
P
E
K
U
R
T
O
S
I
S
S
P
R
E
A
D
D
E
C
R
E
A
S
E
F
L
U
X
F
L
A
T
N
E
S
S
C
O
M
P
A
R
I
S
O
N
O
F
9
S
P
E
C
T
R
A
L
F
E
A
T
U
R
E
V
A
L
U
E
S
T
O
4
M
O
O
D
C
A
T
E
G
O
R
I
E
S
0
1
C
o
n
t
e
n
m
e
n
t
0
2
E
x
u
b
e
r
a
n
c
e
0
3
D
e
p
r
e
s
s
i
o
n
0
4
A
n
x
i
o
u
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
E
C
E
I
SS
N:
2
0
8
8
-
8708
Desig
n
a
n
d
A
n
a
lysi
s
S
ystem
o
f KN
N
a
n
d
I
D3
A
l
g
o
r
ith
m
fo
r
Mu
s
ic
C
la
s
s
ifica
tio
n
B
a
s
ed
…
(
Ma
d
e
S
u
d
a
r
ma
)
493
KNN
in
ter
f
ac
e
t
h
er
e
is
a
tex
tb
o
x
th
at
i
s
u
s
ed
to
in
p
u
t
t
h
e
K
NN
n
ei
g
h
b
o
r
in
g
k
v
al
u
e.
Data
in
p
u
t
i
n
t
h
is
s
y
s
te
m
is
th
e
a
u
d
io
d
ata
w
it
h
f
o
r
m
at
m
o
n
o
*
.
w
a
v
.
T
h
e
tr
ain
i
n
g
d
ata
th
at
is
u
s
ed
i
n
b
o
th
o
f
t
h
ese
i
n
ter
f
ac
es a
r
e
a
s
et
o
f
f
ea
t
u
r
e
s
et
v
al
u
e
r
es
u
lt
f
r
o
m
s
p
ec
tr
al
an
al
y
s
is
t
h
at
h
as
s
o
u
r
ce
f
r
o
m
a
u
d
io
f
ile
w
it
h
f
o
u
r
t
y
p
es
o
f
m
o
o
d
ca
teg
o
r
y
.
T
h
is
tr
ain
i
n
g
d
ata
is
u
s
ed
as
r
ef
er
e
n
ce
d
ata
to
f
i
n
d
o
u
t
lev
el
o
f
class
if
ica
tio
n
ac
cu
r
ac
y
an
d
ti
m
e
n
ee
d
ed
in
class
if
icatio
n
p
r
o
ce
s
s
i
n
m
u
s
ic
f
ile
i
n
p
u
t
to
m
o
o
d
.
T
h
e
f
o
llo
w
i
n
g
s
ar
e
ap
p
ea
r
an
ce
f
o
r
in
ter
f
ac
e
o
f
KNN
an
d
I
D3
class
i
f
icatio
n
s
y
s
te
m
(
Fi
g
u
r
e
4
,
5
).
B
o
th
o
f
th
e
s
e
in
ter
f
ac
es
h
a
v
e
4
p
an
els co
n
s
i
s
t o
f
,
f
ir
s
t p
a
n
el,
th
e
Mu
s
ic
Mo
o
d
Data
P
r
o
ce
s
s
in
g
w
h
er
e
in
cl
u
d
in
g
o
f
s
ev
er
al
b
u
tto
n
s
t
h
at
ar
e
u
s
ed
as
tr
ig
g
er
o
f
clas
s
if
i
ca
tio
n
s
ta
g
es.
Seco
n
d
,
th
e
g
r
ap
h
ic
p
an
el
r
esu
l
t
o
f
s
i
g
n
al
s
a
m
p
li
n
g
,
co
n
s
i
s
t
o
f
o
r
ig
in
al
a
u
d
io
s
ig
n
al
s
a
m
p
li
n
g
an
d
s
ig
n
al
s
a
m
p
li
n
g
o
f
au
d
io
r
esu
lt
o
f
FF
T
.
T
h
e
th
ir
d
is
th
e
g
r
ap
h
ic
p
an
el
o
f
o
r
ig
in
al
au
d
io
s
i
g
n
a
l
s
p
ec
tr
o
g
r
a
m
an
d
au
d
io
s
ig
n
al
s
p
ec
tr
o
g
r
a
m
r
es
u
l
t
o
f
FF
T
.
T
h
e
f
o
u
r
t
h
i
s
p
an
el
o
f
clas
s
if
icatio
n
r
es
u
lt
th
a
t
co
n
s
is
t
s
o
f
clas
s
i
f
icatio
n
r
es
u
lt
lab
el
an
d
cla
s
s
i
f
icatio
n
p
r
o
ce
s
s
in
g
t
i
m
e.
Fo
r
class
if
icatio
n
r
es
u
lt
u
s
i
n
g
KNN
alg
o
r
ith
m
at
4
0
-
4
0
0
th
is
test
in
g
d
ata
is
ca
r
r
ied
o
u
t
w
it
h
k
v
al
u
e
th
at
i
s
1
to
1
6
,
it
is
o
b
tain
ed
p
er
ce
n
tag
e
o
f
a
v
er
ag
e
cla
s
s
i
f
ic
atio
n
ac
cu
r
ac
y
f
o
r
h
i
g
h
e
s
t
v
al
u
e
is
in
k
v
al
u
e
=
3
f
o
r
8
6
.
5
5
%,
an
d
av
er
ag
e
p
r
o
ce
s
s
i
n
g
ti
m
e
p
er
m
u
s
ic
f
ile
i
s
0
.
0
1
0
2
1
s
ec
o
n
d
s
.
T
h
is
r
es
u
lt
ca
n
b
e
s
ee
n
i
n
Fig
u
r
e
6
a
n
d
T
ab
le
1
.
W
h
er
ea
s
f
o
r
cla
s
s
i
f
icat
io
n
r
es
u
lt
u
s
i
n
g
I
D3
al
g
o
r
ith
m
u
s
e
s
4
0
-
4
0
0
t
esti
n
g
d
ata
a
n
d
it
is
o
b
tain
ed
p
er
ce
n
tag
e
o
f
av
er
a
g
e
class
i
f
ica
tio
n
ac
cu
r
ac
y
f
o
r
5
9
.
3
3
%,
an
d
av
er
ag
e
p
r
o
ce
s
s
in
g
ti
m
e
p
er
m
u
s
ic
f
ile
is
0
.
0
5
0
9
1
s
ec
o
n
d
s
.
Fig
u
r
e
4
.
I
n
ter
f
ac
e
o
f
KNN
cl
ass
i
f
icatio
n
s
y
s
te
m
Fig
u
r
e
5
.
I
n
ter
f
ac
e
o
f
I
D3
clas
s
if
ica
tio
n
s
y
s
te
m
Fig
u
r
e
6
.
R
esu
lt o
f
clas
s
i
f
icati
o
n
ac
cu
r
ac
y
u
s
e
s
K
-
NN
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8708
I
J
E
C
E
Vo
l.
7
,
No
.
1
,
Feb
r
u
ar
y
2
0
1
7
:
4
8
6
–
495
494
T
ab
le
1
.
A
cc
u
r
ac
y
a
n
d
Av
er
ag
e
C
lass
if
ica
tio
n
P
r
o
ce
s
s
in
g
T
im
e
o
f
KNN
No
A
mo
u
n
t
o
f
t
e
st
i
n
g
d
a
t
a
K
N
N
I
D
3
A
c
c
u
r
a
c
y
o
f
k
=
3
A
v
e
r
a
g
e
T
i
me
(
se
c
o
n
d
s)
A
c
c
u
r
a
c
y
A
v
e
r
a
g
e
T
i
me
(
se
c
o
n
d
s)
1
40
8
7
,
5
%
0
.
0
1
0
5
9
6
7
.
5
%
0
.
0
6
9
1
5
2
80
8
5
%
0
.
0
1
0
8
4
6
5
%
0
.
0
4
7
4
6
3
1
2
0
8
9
,
2
%
0
.
0
1
0
9
1
6
0
.
8
%
0
.
0
5
4
4
3
4
1
6
0
8
9
,
3
%
0
.
0
1
0
4
7
6
0
.
6
%
0
.
0
4
7
5
2
5
2
0
0
8
7
,
5
%
0
.
0
1
0
0
5
5
9
%
0
.
0
4
3
3
6
6
2
4
0
8
7
,
5
%
0
.
0
1
0
3
8
5
7
.
1
%
0
.
0
5
4
7
3
7
2
8
0
8
8
,
7
%
0
.
0
0
9
9
1
5
5
%
0
.
0
5
1
3
3
8
3
2
0
8
5
%
0
.
0
0
9
7
4
5
5
.
6
%
0
.
0
4
8
8
8
9
3
6
0
8
4
,
7
%
0
.
0
0
9
6
4
5
6
.
4
%
0
.
0
4
6
9
1
10
4
0
0
8
4
%
0
.
0
0
9
5
6
5
6
.
2
%
0
.
0
4
5
3
0
A
c
c
u
r
a
c
y
a
n
d
a
v
e
r
a
g
e
c
l
a
ssi
f
i
c
a
t
i
o
n
8
6
,
5
5
%
0
.
0
1
0
2
1
se
c
o
n
d
s
5
9
.
3
3
%
0
.
0
5
0
9
1
se
c
o
n
d
s
5.
CO
NCLU
SI
O
N
Gen
er
all
y
,
r
es
u
lt
o
f
s
y
s
te
m
cl
ass
i
f
icatio
n
b
u
i
lt
b
y
u
s
i
n
g
K
NN
al
g
o
r
ith
m
h
as
b
ee
n
ab
le
to
p
er
f
o
r
m
m
u
s
ic
cla
s
s
i
f
icatio
n
to
m
o
o
d
b
y
u
s
in
g
T
h
a
y
er
’
s
m
o
d
el
w
i
th
p
er
ce
n
tag
e
o
f
ac
c
u
r
ac
y
i
s
g
o
o
d
en
o
u
g
h
w
h
ich
i
s
8
6
.
5
5
%,
an
d
av
er
ag
e
p
r
o
ce
s
s
in
g
ti
m
e
is
0
.
0
1
0
2
1
s
ec
o
n
d
s
.
W
h
er
ea
s
s
y
s
te
m
w
it
h
I
D3
alg
o
r
ith
m
p
r
o
d
u
ce
d
class
i
f
icatio
n
ac
cu
r
ac
y
f
o
r
5
9
.
3
3
%
an
d
av
er
ag
e
p
r
o
ce
s
s
in
g
ti
m
e
is
0
.
0
5
0
9
1
1
2
7
5
s
ec
o
n
d
s
.
Su
g
g
e
s
tio
n
f
o
r
f
u
r
t
h
er
s
y
s
te
m
d
e
v
elo
p
m
e
n
t,
i
n
m
u
s
ic
f
ile
p
r
e
-
p
r
o
ce
s
s
i
n
g
s
ta
g
e
it
ca
n
b
e
tr
ied
to
u
s
e
f
ile
w
i
th
f
o
r
m
at
o
f
s
te
r
eo
*
.
w
av
w
it
h
m
o
r
e
o
p
ti
m
al
r
e
f
r
ain
d
u
r
atio
n
.
Mo
r
eo
v
er
,
t
h
e
K
-
N
N
al
g
o
r
ith
m
h
as
w
ea
k
n
ess
i
n
d
eter
m
in
in
g
ac
cu
r
ate
k
v
al
u
e
to
p
r
o
d
u
ce
th
e
b
est
s
y
s
te
m
ac
c
u
r
ac
y
.
T
h
er
ef
o
r
e,
it
is
ex
p
ec
ted
th
a
t
in
f
u
r
th
er
s
y
s
te
m
d
ev
elo
p
m
en
t o
p
ti
m
izi
n
g
i
s
ca
r
r
ied
o
u
t in
ch
o
o
s
i
n
g
t
h
e
k
v
al
u
e
p
ar
am
eter
au
to
m
atica
ll
y
.
ACK
NO
WL
E
D
G
E
M
E
NT
S
T
h
e
ap
p
r
ec
iatio
n
is
d
eliv
er
ed
to
th
e
to
t
h
e
Dep
ar
t
m
e
n
t
o
f
E
lectr
ical
an
d
C
o
m
p
u
ter
E
n
g
in
ee
r
in
g
–
Ud
a
y
an
a
U
n
i
v
er
s
it
y
a
n
d
ST
MI
K
-
ST
I
KOM
B
ali
th
at
h
as
g
i
v
en
o
p
p
o
r
tu
n
it
y
to
p
er
f
o
r
m
t
h
e
r
esear
ch
,
an
d
also
fo
r
all
p
ar
ties
th
at
al
w
a
y
s
g
i
v
e
s
u
p
p
o
r
t
d
ir
ec
tly
an
d
i
n
d
ir
ec
tl
y
i
n
th
e
f
i
n
is
h
i
n
g
o
f
th
i
s
s
cien
ti
f
ic
j
o
u
r
n
al
esp
ec
iall
y
w
e
w
o
u
ld
lik
e
to
t
h
an
k
y
o
u
a
g
ain
.
RE
F
E
R
E
NC
E
S
[1
]
S
.
P
o
u
y
a
n
f
a
r
a
n
d
H.
S
a
m
e
ti
,
“
M
u
sic
e
m
o
ti
o
n
re
c
o
g
n
it
i
o
n
u
s
in
g
tw
o
lev
e
l
c
las
si
f
ica
ti
o
n
”
,
i
n
2
0
1
4
Ira
n
i
a
n
Co
n
fer
e
n
c
e
o
n
In
tell
ig
e
n
t
S
y
ste
ms
(
ICIS
)
,
2
0
1
4
,
p
p
.
1
–
6.
[2
]
Y.
S
o
n
g
,
S
.
Dix
o
n
,
a
n
d
M
.
P
e
a
rc
e
,
“
Ev
a
lu
a
ti
o
n
o
f
M
u
sic
a
l
F
e
a
tu
re
s
f
o
r
E
m
o
ti
o
n
Clas
sif
ica
ti
o
n
”
,
in
IS
M
IR
,
2
0
1
2
,
p
p
.
5
2
3
–
5
2
8
.
[3
]
T
h
a
y
e
r
R.
E.
,
T
h
e
Bi
o
p
sy
c
h
o
l
o
g
y
o
f
M
o
o
d
a
n
d
Aro
u
sa
l
.
Ox
f
o
rd
Un
i
v
e
rsit
y
P
re
ss
,
1
9
9
0
.
[4
]
V
.
Ha
m
p
ih
o
li
,
“
A
m
e
th
o
d
f
o
r
m
u
sic
c
las
si
f
ica
ti
o
n
b
a
se
d
o
n
p
e
rc
e
iv
e
d
m
o
o
d
d
e
tec
ti
o
n
f
o
r
In
d
ian
b
o
ll
y
w
o
o
d
m
u
sic
”
,
in
Pro
c
e
e
d
i
n
g
s
o
f
W
o
rl
d
Aca
d
e
my
o
f
S
c
ien
c
e
,
En
g
in
e
e
rin
g
a
n
d
T
e
c
h
n
o
l
o
g
y
,
2
0
1
2
,
p
.
5
0
7
.
[5
]
B.
G
.
P
a
tra,
D.
Da
s,
a
n
d
S
.
Ba
n
d
y
o
p
a
d
h
y
a
y
,
“
A
u
to
m
a
ti
c
m
u
sic
m
o
o
d
c
las
si
f
ica
ti
o
n
o
f
Hin
d
i
so
n
g
s”
,
in
S
ixth
In
ter
n
a
t
io
n
a
l
J
o
in
t
Co
n
fer
e
n
c
e
o
n
Na
tu
r
a
l
L
a
n
g
u
a
g
e
Pro
c
e
ss
in
g
,
2
0
1
3
,
p
.
2
4
.
[6
]
M
.
B.
M
o
k
h
si
n
,
N.B.
Ro
sli,
S
.
Zam
b
ri,
N.D.
A
h
m
a
d
,
a
n
d
S
.
R
a
h
a
h
,
“
A
u
to
m
a
ti
c
M
u
sic
E
m
o
ti
o
n
Clas
sif
ic
a
ti
o
n
Us
in
g
A
rt
if
i
c
ial
Ne
u
ra
l
N
e
t
w
o
rk
Ba
se
d
o
n
V
o
c
a
l
a
n
d
In
str
u
m
e
n
tal
S
o
u
n
d
T
im
b
re
s
”
,
J
.
Co
mp
u
t.
S
c
i.
,
v
o
l.
1
0
,
n
o
.
1
2
,
p
p
.
2
5
8
4
–
2
5
9
2
,
De
c
.
2
0
1
4
.
[7
]
K.C.
De
w
i
a
n
d
L
.
A
.
A
.
R.
P
u
tr
i,
“
M
u
sic
M
o
o
d
P
lay
e
r
I
m
p
le
m
e
n
tat
io
n
A
p
p
li
e
d
In
Da
y
c
a
re
Us
in
g
S
e
lf
Org
a
n
izin
g
M
a
p
M
e
th
o
d
”
,
J
.
B
u
a
n
a
I
n
fo
rm
.
,
v
o
l.
2
,
n
o
.
2
,
2
0
1
1
.
[8
]
A
.
S
e
ti
a
w
a
n
,
“
A
n
a
li
sis
Kla
si
f
ik
a
si
S
u
a
ra
Be
rd
a
sa
rk
a
n
G
e
n
d
e
r
d
e
n
g
a
n
F
o
rm
a
t
W
a
v
M
e
n
g
g
u
n
a
k
a
n
A
lg
o
rit
m
a
K
-
M
e
a
n
s”
,
S
a
i
n
s Da
n
T
e
k
n
o
l.
,
v
o
l.
2
,
n
o
.
2
,
2
0
0
9
.
[9
]
G
.
V
y
a
s
a
n
d
M
.
K.
Du
tt
a
,
“
A
u
to
m
a
ti
c
m
o
o
d
d
e
tec
ti
o
n
o
f
in
d
ian
m
u
sic
u
sin
g
M
F
CCs
a
n
d
K
-
m
e
a
n
s
a
lg
o
rit
h
m
”
,
in
S
e
v
e
n
th
I
n
ter
n
a
t
io
n
a
l
C
o
n
fer
e
n
c
e
o
n
C
o
n
tem
p
o
ra
ry
C
o
mp
u
ti
n
g
(
IC3
)
,
2
0
1
4
,
p
p
.
1
1
7
–
1
2
2
.
[1
0
]
S
.
Oh
,
M
.
Ha
h
n
,
a
n
d
J.
Kim
,
“
M
u
sic
m
o
o
d
c
las
si
f
ica
ti
o
n
u
sin
g
in
tro
a
n
d
re
f
ra
in
p
a
rts
o
f
l
y
rics
”
,
in
In
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
In
f
o
rm
a
ti
o
n
S
c
ien
c
e
a
n
d
Ap
p
li
c
a
ti
o
n
s (
ICIS
A)
,
2
0
1
3
,
p
p
.
1
–
3.
[1
1
]
R.
Y.
S
ip
a
su
lt
a
,
A
.
S
.
L
u
m
e
n
ta,
a
n
d
S
.
R.
S
o
m
p
ie,
“
S
im
u
las
i
S
iste
m
P
e
n
g
a
c
a
k
S
in
y
a
l
De
n
g
a
n
M
e
t
o
d
e
F
F
T
(F
a
st
F
o
u
rier T
ra
n
sf
o
rm
)”
,
J.
Tek
.
El
e
k
tro
Da
n
K
o
m
p
u
t.
UN
S
RA
T
,
v
o
l.
3
,
n
o
.
2
,
p
p
.
1
–
9
,
2
0
1
4
.
[1
2
]
S
.
G
o
p
in
a
t
h
a
n
,
R
.
Ko
k
il
a
,
a
n
d
P
.
T
h
a
n
g
a
v
e
l,
“
W
a
v
e
let
a
n
d
F
F
T
Ba
se
d
Im
a
g
e
De
n
o
isin
g
Us
in
g
No
n
-
li
n
e
a
r
F
il
ters
”
,
IJ
ECE
,
v
o
l.
5
,
n
o
.
5
,
2
0
1
5
.
[1
3
]
A
.
Lerc
h
,
An
i
n
tro
d
u
c
ti
o
n
to
a
u
d
io
c
o
n
te
n
t
a
n
a
lys
is:
a
p
p
li
c
a
ti
o
n
s
in
sig
n
a
l
p
ro
c
e
ss
in
g
a
n
d
mu
sic
in
fo
rm
a
ti
c
s
.
Ho
b
o
k
e
n
,
NJ
:
W
il
e
y
,
2
0
1
2
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
E
C
E
I
SS
N:
2
0
8
8
-
8708
Desig
n
a
n
d
A
n
a
lysi
s
S
ystem
o
f KN
N
a
n
d
I
D3
A
l
g
o
r
ith
m
fo
r
Mu
s
ic
C
la
s
s
ifica
tio
n
B
a
s
ed
…
(
Ma
d
e
S
u
d
a
r
ma
)
495
[1
4
]
M
.
A
b
d
a
r,
S
.
R.
N.
Ka
lh
o
r
i,
T
.
S
u
ti
k
n
o
,
I.
M
.
I.
S
u
b
ro
t
o
,
a
n
d
G
.
A
rji
,
“
Co
m
p
a
rin
g
P
e
rf
o
rm
a
n
c
e
o
f
Da
ta
M
in
in
g
A
l
g
o
rit
h
m
s in
P
re
d
icti
o
n
He
a
rt
D
ise
a
se
s
”
,
IJ
ECE
,
v
o
l.
5
,
n
o
.
6
,
p
p
.
1
5
6
9
–
1
5
7
6
,
2
0
1
5
.
[1
5
]
Ek
o
P
ra
se
ty
o
,
Da
ta M
in
in
g
:
Ko
n
se
p
d
a
n
A
p
li
k
a
si M
e
n
g
g
u
n
a
k
a
n
M
a
tl
a
b
.
Yo
g
y
a
k
a
rta:
A
ND
I,
2
0
1
2
.
[1
6
]
V
a
su
d
e
v
a
n
,
“
Itera
ti
v
e
Dic
h
o
to
m
ise
r
-
3
A
lg
o
rit
h
m
in
Da
ta
M
in
in
g
A
p
p
li
e
d
to
Dia
b
e
tes
Da
tab
a
se
”
,
J
.
Co
m
p
u
t.
S
c
i
.
,
v
o
l.
1
0
,
n
o
.
7
,
p
p
.
1
1
5
1
–
1
1
5
5
,
Ju
l
.
2
0
1
4
.
B
I
O
G
RAP
H
I
E
S
O
F
AUTH
O
RS
Dr
.
Ir
.
M
a
d
e
S
u
d
a
r
m
a
,
M
.
A.S
c
.
De
p
a
rtme
n
t
o
f
El
e
c
tri
c
a
l
a
n
d
Co
m
p
u
ter E
n
g
in
e
e
rin
g
F
a
c
u
lt
y
o
f
En
g
in
e
e
rin
g
,
Ud
a
y
a
n
a
Un
iv
e
rsit
y
Ud
a
y
a
n
a
Ca
m
p
u
s Rd
,
Jim
b
a
ra
n
,
Ba
d
u
n
g
Ba
li
-
IND
ON
ES
IA
Em
a
il
:
i
m
a
su
d
a
r
m
a
@
g
m
a
il
.
c
o
m
I G
e
d
e
H
a
r
se
m
a
d
i,
S
.
K
o
m
.
,
M
.
T.
De
p
a
rtme
n
t
o
f
In
f
o
rm
a
ti
o
n
S
y
ste
m
S
T
M
IK S
T
IKO
M
B
A
L
I
Jl.
Ra
y
a
P
u
p
u
tan
No
.
8
6
,
De
n
p
a
sa
r
Ba
li
8
0
2
3
4
,
I
n
d
o
n
e
sia
.
T
e
l.
/F
a
x
:
+
6
2
3
6
1
2
4
4
4
4
5
Em
a
il
:
g
ed
e
h
ar
s
e
m
ad
i@
y
a
h
o
o
.
co
m
Evaluation Warning : The document was created with Spire.PDF for Python.