IA
E
S
In
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
Ar
t
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
(I
J
-
AI
)
Vo
l
.
8
, N
o
.
1
,
Ma
r
c
h
201
9
, p
p
.
5
4
~
6
2
IS
S
N
:
2252
-
8938
,
DOI
:
1
0
.
1
1
5
9
1
/
i
j
a
i
.
v
8
.i
1
.p
p
5
4
-
6
2
r
54
Jou
r
n
al
h
om
e
page
:
ht
t
p:
/
/
i
ae
s
co
r
e
.c
o
m
/o
n
lin
e
/in
d
e
x
.p
h
p
/I
J
A
I
Di
s
t
a
n
c
e
w
ei
g
h
t
ed
K
-
Me
a
n
s
a
l
g
o
r
i
t
h
m
f
o
r
c
e
n
t
e
r
s
e
l
e
c
t
i
o
n
i
n
tr
a
i
n
i
n
g
r
a
d
i
a
l
b
a
s
i
s
fu
n
c
ti
o
n
n
e
tw
o
r
k
s
Li
m
En
g
A
i
k
1
, T
a
n
W
e
i H
o
n
g
2
, A
h
m
a
d
K
a
d
r
i J
u
n
o
h
3
1
,3
In
s
t
i
t
u
t
M
a
t
e
m
a
t
i
k
K
e
j
u
ru
t
e
ra
a
n
, U
n
iv
e
r
s
iti M
a
la
y
s
ia
P
e
r
lis
, 0
2
6
0
0
A
r
a
u
, P
e
r
lis
, M
a
la
y
s
ia
.
2
Sc
h
o
o
l
o
f
M
e
c
h
a
t
r
o
n
i
c
En
g
i
n
e
e
r
i
n
g
,
U
n
i
v
e
r
s
i
t
i
M
a
l
a
y
s
i
a
Pe
r
l
i
s
,
0
2
6
0
0
A
r
a
u
,
Pe
r
l
i
s
,
M
a
l
a
y
s
i
a
.
Ar
t
i
c
l
e
I
n
f
o
AB
S
T
RACT
Ar
t
i
c
l
e
h
i
s
t
o
r
y
:
Re
c
e
i
v
e
d
N
o
v
2
7
, 2
0
1
8
Re
v
i
s
e
d
J
a
n
3
1
, 2
0
1
9
Ac
c
e
p
t
e
d
Fe
b
1
1
,
201
9
Th
e
a
c
c
u
r
a
c
i
e
s
r
a
t
e
s
o
f
t
h
e
n
e
u
r
a
l
n
e
t
w
o
r
k
s
m
a
i
n
l
y
d
e
p
e
n
d
o
n
t
h
e
s
e
l
e
c
t
i
o
n
of
t
he
c
or
r
e
c
t
da
t
a
c
e
nt
e
r
s
.
T
he
K
-
me
a
n
s
a
l
g
o
r
i
t
h
m
i
s
a
wi
d
e
l
y
u
s
e
d
cl
u
st
er
i
n
g
al
g
o
r
i
t
h
m
i
n
v
ar
i
o
u
s
d
i
sci
p
l
i
n
es
f
o
r
cen
t
er
s
sel
ect
i
o
n
.
H
o
w
ev
er
,
t
h
e
me
t
h
o
d
i
s
k
n
o
wn
f
o
r
i
t
s
sen
si
t
i
v
i
t
y
t
o
i
n
i
t
i
al
cen
t
er
s
sel
ect
i
o
n
.
I
t
su
f
f
er
s
n
o
t
onl
y
f
r
om
a
hi
gh
de
pe
nde
nc
y
on
t
he
a
l
gor
i
t
hm
'
s
i
ni
t
i
a
l
c
e
nt
e
r
s
s
e
l
e
c
t
i
on
but
,
al
so
f
r
o
m
d
at
a
p
o
i
n
t
s.
T
h
e
p
er
f
o
r
m
an
ce
o
f
K
-
me
a
n
s
h
a
s
b
e
e
n
e
n
h
a
n
c
e
d
f
r
o
m
di
f
f
e
r
e
nt
pe
r
s
pe
c
t
i
ve
s
,
i
nc
l
udi
ng
c
e
nt
r
oi
d
i
ni
t
i
al
i
zat
i
o
n
p
r
o
b
l
em
o
v
er
t
h
e
ye
a
r
s
.
U
nf
or
t
una
t
e
l
y,
t
he
s
ol
ut
i
on
doe
s
not
pr
ovi
de
a
good
t
r
a
de
-
of
f
be
t
w
e
e
n
qua
l
i
t
y
a
nd
e
f
f
i
c
i
e
nc
y
of
t
he
c
e
nt
e
r
s
pr
oduc
e
s
by
t
he
a
l
gor
i
t
hm
.
T
o
s
ol
ve
t
hi
s
pr
obl
e
m
,
a
ne
w
m
e
t
hod
t
o
f
i
nd
t
he
i
ni
t
i
a
l
c
e
nt
e
r
s
a
nd
i
m
pr
ove
t
he
s
e
ns
i
t
i
vi
t
y
to
th
e
in
itia
l
c
e
n
te
r
s
o
f
K
-
me
a
n
s
a
l
g
o
r
i
t
h
m
i
s
p
r
o
p
o
s
e
d
.
T
h
i
s
p
a
p
e
r
p
r
e
s
e
n
t
e
d
a
t
r
ai
n
i
n
g
al
g
o
r
i
t
h
m
f
o
r
t
h
e
r
ad
i
al
b
asi
s
f
u
n
ct
i
o
n
n
et
w
o
r
k
(
R
B
F
N
)
u
si
n
g
im
p
r
o
v
e
d
K
-
me
a
n
s
(
KM
)
a
l
g
o
r
i
t
h
m,
wh
i
c
h
i
s
t
h
e
mo
d
i
f
i
e
d
v
e
r
s
i
o
n
o
f
KM
al
g
o
r
i
t
h
m
b
ased
o
n
d
i
st
an
ce
-
we
i
g
ht
e
d
a
dj
us
t
m
e
nt
f
or
e
a
c
h
c
e
nt
e
r
s
,
know
n
a
s
di
s
t
a
nc
e
-
we
i
g
h
t
e
d
K
-
me
a
n
s
(
DW
KM
)
a
l
g
o
r
i
t
h
m.
T
h
e
p
r
o
p
o
s
e
d
t
r
a
i
n
i
n
g
al
g
o
r
i
t
h
m
,
w
h
i
ch
u
ses
D
W
K
M
al
g
o
r
i
t
h
m
sel
ect
cen
t
er
s
f
o
r
t
r
ai
n
i
n
g
R
B
F
N
obt
a
i
ne
d
be
t
t
e
r
a
c
c
ur
a
c
y
i
n
pr
e
di
c
t
i
ons
a
nd
r
e
duc
e
d
ne
t
w
or
k
a
r
c
hi
t
e
c
t
ur
e
co
m
p
ar
ed
t
o
t
h
e
st
an
d
ar
d
R
B
F
N
.
T
h
e
p
r
o
p
o
sed
t
r
ai
n
i
n
g
al
g
o
r
i
t
h
m
w
as
im
p
le
m
e
n
te
d
in
M
A
T
L
A
B
e
n
v
ir
o
n
m
e
n
t;
h
e
n
c
e
,
th
e
n
e
w
n
e
tw
o
r
k
w
a
s
unde
r
goi
ng
a
hybr
i
d
l
e
a
r
ni
ng
pr
oc
e
s
s
.
T
he
ne
t
w
or
k
c
a
l
l
e
d
D
W
K
M
-
RBF
N
wa
s
t
e
s
t
e
d
a
g
a
i
n
s
t
t
h
e
s
t
a
n
d
a
r
d
R
B
F
N
i
n
p
r
e
d
i
c
t
i
o
n
s
.
T
h
e
e
x
p
e
r
i
m
en
t
al
mo
d
e
l
s
we
r
e
t
e
s
t
e
d
o
n
f
o
u
r
l
i
t
e
r
a
t
u
r
e
s
n
o
n
l
i
n
e
a
r
f
u
n
c
t
i
o
n
a
n
d
f
o
u
r
r
e
a
l
-
wo
r
l
d
ap
p
l
i
cat
i
o
n
p
r
o
b
l
em
s,
p
ar
t
i
cu
l
ar
l
y
i
n
A
i
r
p
o
l
l
u
t
an
t
p
r
o
b
l
em
,
B
i
o
ch
em
i
cal
Ox
y
g
e
n
De
ma
n
d
(
B
OD)
p
r
o
b
l
e
m,
P
h
y
t
o
p
l
a
n
k
t
o
n
p
r
o
b
l
e
m,
a
n
d
f
o
r
e
x
p
a
i
r
EU
R
U
SD
.
Th
e
r
e
s
u
l
t
s
a
r
e
c
o
m
p
a
r
e
d
to
p
r
o
p
o
s
e
d
m
e
th
o
d
f
o
r
r
o
o
t
m
e
a
n
sq
u
ar
e
er
r
o
r
(
R
M
S
E
)
i
n
r
ad
i
al
b
asi
s
f
u
n
ct
i
o
n
n
et
w
o
r
k
(
R
B
F
N
)
.
T
h
e
p
r
o
p
o
sed
me
t
h
o
d
y
i
e
l
d
e
d
a
p
r
o
mi
s
i
n
g
r
e
s
u
l
t
wi
t
h
a
n
a
v
e
r
a
g
e
i
mp
r
o
v
e
me
n
t
p
e
r
c
e
n
t
a
g
e
mo
r
e
t
h
a
n
5
0
p
e
r
c
e
n
t
i
n
R
M
S
E
.
Ke
y
wo
r
d
s
:
C
en
t
er
s
el
ect
i
o
n
D
at
as
et
s
i
ze
r
ed
u
ct
i
o
n
K
-
me
a
n
s
a
l
g
o
r
i
t
h
m
N
eu
r
al
n
et
w
o
r
k
R
ad
i
al
b
as
i
s
f
u
n
ct
i
o
n
n
et
w
o
r
k
Co
p
y
r
i
g
h
t
©
2
0
1
9
In
s
t
i
t
u
t
e
o
f
A
d
v
a
n
c
e
d
E
n
g
i
n
e
e
r
i
n
g
a
n
d
S
c
i
e
n
c
e
.
Al
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
Co
r
r
e
s
p
o
n
d
i
n
g
Au
t
h
o
r
:
Li
m
En
g
A
i
k
,
In
s
t
i
t
u
t
M
a
t
e
m
a
t
i
k
K
e
j
u
ru
t
e
ra
a
n
,
Un
i
v
e
r
s
i
t
i
M
a
l
a
y
s
i
a
P
e
r
l
i
s
,
02600
A
r
a
u,
P
e
r
l
i
s
,
M
a
l
a
ys
i
a
.
Em
a
i
l
:
e.
a.
l
i
m
8
0
@
g
m
ai
l
1.
IN
T
R
O
D
U
C
T
IO
N
Ra
d
i
a
l
Ba
s
i
s
F
u
n
c
t
i
o
n
n
e
t
w
o
r
k
s
(
RBF
N
)
f
o
r
m
a
t
y
p
e
o
f
A
r
t
i
f
i
c
i
a
l
N
e
u
r
a
l
N
e
t
w
o
r
k
s
(
A
N
N
s
)
,
w
h
i
c
h
ha
s
c
e
r
t
a
i
n
a
dva
nt
a
ge
s
ove
r
ot
he
r
ki
nds
of
A
N
N
s
,
w
i
t
h
be
t
t
e
r
a
ppr
oxi
m
a
t
i
on
c
a
pa
bi
l
i
t
i
e
s
,
s
i
m
pl
e
r
ne
t
w
or
k
st
r
u
c
t
u
r
e
s
a
n
d
f
a
st
e
r
l
e
a
r
n
i
n
g
a
l
g
o
r
i
t
h
m
s.
A
s
a
r
e
su
l
t
,
R
BF
n
e
t
w
o
r
k
s
t
u
r
n
p
o
p
u
l
a
r
a
m
o
n
g
r
e
s
e
a
r
c
h
e
r
s
.
M
a
n
y
re
s
e
a
rc
h
e
rs
w
h
o
m
h
a
v
e
b
e
e
n
w
o
rk
i
n
g
t
o
p
ro
d
u
c
e
m
o
re
e
ffe
c
t
i
v
e
t
ra
i
n
i
n
g
a
l
g
o
ri
t
h
m
s
,
s
e
t
a
l
o
n
g
s
i
d
e
t
h
e
st
a
n
d
a
r
d
t
e
c
h
n
i
q
u
e
s
[1
-
7]
.
RBF
n
e
t
w
o
r
k
s
a
r
e
u
s
e
f
u
l
i
n
a
ppr
oxi
m
a
t
i
on
pr
obl
e
m
s
,
but
i
t
r
e
qui
r
e
s
a
l
ong
t
i
m
e
t
o
te
a
c
h
th
e
n
e
tw
o
r
k
s
a
s
it
p
e
r
ta
in
s
to
a
h
u
g
e
n
u
m
b
e
r
o
f
tr
a
in
in
g
d
a
ta
,
y
e
t
c
r
e
a
te
a
h
ig
h
e
r
r
o
r
b
e
c
a
u
s
e
o
f
p
o
s
s
ib
le
Evaluation Warning : The document was created with Spire.PDF for Python.
IJ
-
AI
IS
S
N
:
2252
-
8938
r
Di
s
t
a
n
c
e
w
e
i
g
h
t
e
d
K
-
Me
a
n
s
a
l
g
o
r
i
t
h
m
f
o
r
c
e
n
t
e
r
s
e
l
e
c
t
i
o
n
i
n
t
r
a
i
n
i
n
g
r
a
d
i
a
l
b
a
s
i
s
…
(
L
i
m
E
n
g
A
i
k
)
55
in
v
a
lid
d
a
ta
o
r
o
u
tlie
r
in
th
e
tr
a
in
in
g
d
a
ta
.
T
h
o
u
g
h
a
c
o
m
b
in
a
tio
n
o
f
c
lu
s
te
r
in
g
m
e
th
o
d
s
in
R
B
F
n
e
tw
o
r
k
s
h
a
s
be
e
n
pr
ove
d
by
S
a
r
i
m
ve
i
s
[1
]
to
b
e
f
a
s
te
r
in
tr
a
in
in
g
,
it
s
till
p
r
o
d
u
c
e
s
a
m
o
r
e
s
u
b
s
ta
n
tia
l
e
r
r
o
r
.
T
h
is
is
d
u
e
to
th
e
s
ta
n
d
a
r
d
c
lu
s
te
r
in
g
a
lg
o
r
ith
m
s
,
w
h
ic
h
s
till
la
c
k
th
e
a
b
ility
to
c
h
o
o
s
e
th
e
m
o
s
t
a
c
c
u
r
a
te
a
n
d
in
f
or
m
a
t
i
ve
cen
t
er
s
.
B
y
u
s
i
n
g
d
i
s
t
an
ce
-
we
i
g
h
t
e
d
K
-
me
a
n
s
(
D
W
K
M
)
a
l
g
o
r
i
t
h
m,
w
e
c
a
n
f
i
x
t
h
e
p
r
o
b
l
e
m
s
t
a
t
e
d
a
b
o
v
e
.
No
t
e
d
t
h
a
t
t
h
e
m
o
r
e
a
c
c
u
r
a
t
e
t
h
e
c
e
n
t
e
r
s
c
h
o
s
e
n
,
t
h
e
m
o
r
e
a
c
c
u
r
a
t
e
t
h
e
i
n
f
o
r
m
a
t
i
o
n
t
h
a
t
f
e
e
d
s
t
o
t
h
e
t
r
a
i
n
ne
t
w
or
k,
t
hi
s
l
e
a
ds
t
o
m
or
e
a
c
c
ur
a
t
e
r
e
s
u
lts
.
I
n
th
is
p
a
p
e
r
,
a
f
a
s
t
a
lg
o
r
ith
m
f
o
r
tr
a
in
in
g
R
B
F
n
e
tw
o
r
k
s
w
h
ic
h
yi
e
l
d
hi
gh
a
c
c
ur
a
c
i
e
s
i
s
pr
e
s
e
nt
e
d,
w
he
r
e
t
he
i
nput
c
e
nt
e
r
s
a
r
e
s
e
l
e
c
t
e
d
t
hr
ough
t
he
D
W
K
M
a
l
gor
i
t
hm
.
T
he
me
t
h
o
d
o
l
o
g
y
i
s
i
l
l
u
s
t
r
a
t
e
d
t
h
r
o
u
g
h
t
h
e
a
p
p
l
i
c
a
t
i
o
n
o
f
e
i
g
h
t
e
x
p
e
r
i
me
n
t
a
l
mo
d
e
l
s
,
w
ith
f
o
u
r
f
r
o
m
lite
r
a
tu
r
e
s
fu
n
c
t
i
o
n
,
a
n
d
fo
u
r
re
a
l
-
wo
r
l
d
p
r
o
b
l
e
m
s
d
a
t
a
o
b
t
a
i
n
e
d
f
r
o
m
L
i
m
[8
]
.
T
h
e
a
d
v
a
n
ta
g
e
s
o
f
th
e
p
r
e
s
e
n
te
d
le
a
r
n
in
g
st
r
a
t
e
g
y
,
D
W
K
M
-
RBF
N
,
a
r
e
i
d
e
n
t
i
f
i
e
d
a
n
d
t
h
e
r
e
s
u
l
t
s
a
r
e
c
o
m
p
a
r
e
d
w
i
t
h
s
t
a
n
d
a
r
d
RBF
N
.
2.
RE
L
AT
E
D
W
O
RK
S
On
e
o
f
t
h
e
b
e
s
t
f
e
a
t
u
r
e
s
o
f
n
e
u
r
a
l
n
e
t
wo
r
k
s
i
s
i
t
s
a
b
i
l
i
t
y
t
o
ge
ne
r
a
l
i
z
e
a
nd
a
ppr
oxi
m
a
t
e
a
s
a
m
pl
e
da
t
a
wi
t
h
o
u
t
t
h
e
n
e
e
d
o
f
s
p
e
c
i
f
y
e
q
u
a
t
i
o
n
a
n
d
c
o
e
f
f
i
c
i
e
n
t
s
,
p
a
r
t
i
c
u
l
a
r
l
y
wh
e
n
a
n
u
n
k
n
o
wn
m
o
d
e
l
d
e
s
c
r
i
b
i
n
g
a
n
unknow
n
c
om
pl
e
x
r
e
l
a
t
i
on
a
nd
t
r
a
i
ni
ng
da
t
a
a
bunda
nt
.
D
ue
t
o
t
he
i
r
a
bi
l
i
t
y
t
o
ge
ne
r
a
l
i
z
e
s
ubs
t
a
nt
i
a
l
l
y,
R
a
d
ia
l
Ba
s
i
s
F
u
n
c
t
i
o
n
n
e
t
w
o
r
k
s
(
RBF
N
)
a
r
e
u
s
u
a
l
l
y
s
e
l
e
c
t
e
d
f
o
r
t
h
i
s
p
u
r
p
o
s
e
[9
-
21
]
.
F
u
r
th
e
r
m
o
r
e
,
in
th
is
b
ig
d
a
ta
er
a,
m
an
y
d
o
m
ai
n
s
s
u
ch
as
i
m
ag
e
p
r
o
ces
s
i
n
g
,
t
ex
t
cat
eg
o
r
i
zat
i
o
n
,
b
i
o
m
et
r
i
c,
m
i
cr
o
ar
r
ay
,
et
c.
h
ad
t
h
e
s
i
ze
o
f
da
t
a
s
e
t
s
s
o
l
a
r
ge
,
t
ha
t
r
e
a
l
-
tim
e
s
y
s
te
m
r
e
q
u
ir
e
s
lo
n
g
tim
e
a
n
d
m
e
m
o
r
y
s
to
r
a
g
e
to
p
r
o
c
e
s
s
th
e
m
.
Un
d
e
r
s
u
c
h
co
n
d
i
t
i
o
n
s
,
ap
p
r
o
x
i
m
at
i
o
n
t
as
k
u
s
i
n
g
av
ai
l
ab
l
e
d
at
as
et
s
can
b
eco
m
e
a
ch
al
l
en
g
i
n
g
t
as
k
an
d
di
f
f
i
cul
t
.
T
h
is
pr
obl
e
m
i
s
m
or
e
ch
al
l
en
g
i
n
g
i
n
d
i
s
t
an
ce
b
as
ed
l
ear
n
i
n
g
al
g
o
r
i
t
h
m
s
s
u
ch
as
R
B
F
N
[1
4
,
22
,
23]
,
k
-
ne
a
r
e
s
t
ne
i
ghbor
[2
4
-
25]
,
c
lu
s
te
r
in
g
m
e
th
o
d
[2
1
,
26
-
29]
an
d
s
u
p
p
o
r
t
v
ect
o
r
m
ach
i
n
e
[3
0
-
32]
.
By
d
e
f
a
u
l
t
,
t
h
e
N
N
al
g
o
r
i
t
h
m
mu
s
t
s
e
a
r
c
h
t
h
r
o
u
g
h
a
l
l
a
v
a
i
l
a
b
l
e
t
r
a
i
n
i
n
g
s
a
mp
l
e
s
w
h
i
c
h
r
e
q
u
i
r
e
s
l
a
r
g
e
me
mo
r
y
,
a
n
d
p
e
r
f
o
r
ms
di
s
t
a
nc
e
t
o
c
e
nt
e
r
c
a
l
c
ul
a
t
i
on,
i
s
s
l
ow
dur
i
ng
t
r
a
i
ni
ng
of
N
N
f
or
a
ppr
oxi
m
a
t
i
on
pur
pos
e
s
.
A
ddi
t
i
ona
l
l
y,
due
t
o
NN
s
t
o
r
e
s
a
l
l
s
a
m
p
l
e
s
d
i
s
t
a
n
c
e
s
f
o
r
t
r
a
i
n
i
n
g
d
a
t
a
s
e
t
s,
t
h
u
s,
n
o
i
se
d
i
st
a
n
c
e
s
a
r
e
st
o
r
e
d
a
s
w
e
l
l
,
w
h
i
c
h
c
a
n
c
a
u
se
de
gr
a
de
i
n
a
ppr
oxi
m
a
t
i
on
a
c
c
ur
a
c
y.
R
e
c
e
nt
l
y,
M
i
r
j
a
l
i
l
i
[3
3
]
de
m
ons
t
r
a
t
e
d
t
ha
t
t
he
hybr
i
d
of
e
vol
ut
i
ona
r
y
al
g
o
r
i
t
h
m
s
u
ch
as
p
ar
t
i
cl
e
s
w
ar
m
o
p
t
i
m
i
zat
i
o
n
(
P
S
O
)
w
i
t
h
R
B
F
N
s
h
o
w
s
a
g
o
o
d
p
er
f
o
r
m
an
ce
i
n
cl
as
s
i
f
i
cat
i
o
n
ro
b
l
e
m
s
a
n
d
a
p
p
ro
x
i
m
a
t
i
o
n
p
ro
b
l
e
m
s
.
T
h
e
u
s
e
d
o
f
e
v
o
l
u
t
i
o
n
a
ry
a
l
g
o
ri
t
h
m
a
s
a
t
o
o
l
t
o
s
e
l
e
c
t
m
o
re
a
c
c
u
ra
t
e
cen
t
er
s
ar
e
al
s
o
r
ep
o
r
t
ed
i
n
m
an
y
r
ecen
t
l
i
t
er
at
u
r
es
[7
,
33
-
38]
fo
r
R
B
F
N
t
ra
i
n
i
n
g
i
n
d
e
e
d
i
s
a
g
o
o
d
m
e
t
h
o
d
i
f
th
e
n
e
tw
o
r
k
s
tr
a
in
in
g
s
p
e
e
d
a
n
d
c
o
m
p
u
ta
tio
n
c
o
s
t a
r
e
n
o
t m
a
in
c
o
n
c
e
r
n
s
.
Al
o
n
g
t
h
e
i
n
c
r
e
a
s
es
o
f
R
B
F
N
p
o
p
u
l
ar
i
t
y
,
t
h
e
ap
p
l
i
cat
i
o
n
o
f
R
B
F
N
i
n
ar
eas
s
u
ch
as
cl
as
s
i
f
i
cat
i
o
n
[1
0
-
14]
,
p
a
tte
r
n
r
e
c
o
g
n
itio
n
[4
,
16
,
39]
an
d
p
r
ed
i
ct
i
o
n
[1
5
,
17
-
20,
40
-
43]
in
c
r
e
a
s
e
s
,
th
u
s
p
r
o
o
f
th
e
w
id
e
u
s
e
s
a
n
d
r
e
lia
b
ility
o
f
R
B
F
N
in
m
a
n
y
f
ie
ld
s
.
H
o
w
e
v
e
r
,
R
B
F
N
a
c
c
u
r
a
c
y
m
a
in
ly
d
e
p
e
n
d
in
g
o
n
th
e
in
itia
l
c
e
n
te
r
s
s
e
le
c
te
d
f
r
o
m
d
a
ta
s
e
t
b
e
f
o
r
e
n
e
tw
o
r
k
tr
a
in
in
g
b
e
g
in
s
[1
5
,
3
9
,
4
1
,
4
4
,
45]
.
B
e
s
id
e
s
,
th
e
s
iz
e
o
f
tr
a
in
in
g
d
a
ta
s
e
ts
a
n
d
in
v
a
lid
d
a
ta
f
o
u
n
d
in
d
a
ta
s
e
ts
a
ls
o
p
la
y
a
n
im
p
o
r
ta
n
t
r
o
le
in
d
e
te
r
m
in
in
g
n
e
tw
o
r
k
s
tr
a
in
in
g
s
p
e
e
d
a
n
d
a
c
c
u
r
a
c
y
[4
6
-
50]
.
F
u
r
th
e
r
m
o
r
e
,
it
is
a
ls
o
r
e
p
o
r
te
d
th
a
t
th
e
le
a
r
n
in
g
a
lg
o
r
ith
m
f
o
r
n
e
tw
o
r
k
s
tr
a
in
in
g
m
a
y
p
e
r
f
o
r
m
w
o
r
s
e
w
ith
th
e
in
c
r
e
a
s
e
s
o
f
d
a
ta
s
e
t
[5
1
]
.
H
e
n
c
e
,
to
s
o
lv
e
s
th
e
s
e
m
e
n
tio
n
e
d
p
r
o
b
le
m
s
,
re
s
e
a
rc
h
e
rs
p
ro
p
o
s
e
d
t
h
e
u
s
e
s
o
f
c
l
u
s
t
e
ri
n
g
a
l
g
o
ri
t
h
m
s
i
n
c
e
n
t
e
rs
s
e
l
e
c
t
i
o
n
[1
,
2
8
,
2
9
,
4
4
,
5
2
-
58
]
fo
r
R
B
F
N
fo
r
o
b
t
a
i
n
i
n
g
b
e
t
t
e
r
a
c
c
u
ra
c
y
a
n
d
a
v
o
i
d
p
o
s
s
i
b
l
e
i
n
v
a
l
i
d
d
a
t
a
s
e
t
s
i
n
c
l
u
d
e
s
i
n
t
o
n
e
t
w
o
rk
s
t
ra
i
n
i
n
g
.
T
h
e
m
o
s
t
wi
d
e
l
y
u
s
e
c
l
u
s
t
e
r
i
n
g
a
l
g
o
r
i
t
h
m
i
n
c
e
n
t
e
r
s
s
e
l
e
c
t
i
o
n
i
s
K
-
me
a
n
s
a
l
g
o
r
i
t
h
m,
a
s
i
t
i
s
t
h
e
f
a
s
t
e
s
t
,
l
e
s
s
c
o
mp
l
e
x
al
g
o
r
i
t
h
m
,
an
d
y
i
el
d
g
o
o
d
a
c
c
ur
a
c
y
f
or
c
e
nt
e
r
s
s
e
l
e
c
t
i
on
i
n
c
om
pa
r
e
d
t
o
ot
he
r
a
va
i
l
a
bl
e
c
l
us
t
e
r
i
ng
a
l
gor
i
t
hm
.
Ho
we
v
e
r
,
a
t
p
r
e
s
e
n
t
a
v
a
i
l
a
b
l
e
c
o
n
v
e
n
t
i
o
n
a
l
c
l
u
s
t
e
r
i
n
g
a
l
g
o
r
i
t
h
m
a
r
e
n
o
t
f
l
a
wl
e
s
s
a
l
g
o
r
i
t
h
m
s
t
h
a
t
g
u
a
r
a
n
t
e
e
hi
gh
a
c
c
ur
a
c
y
i
n
c
hoos
i
ng
c
e
nt
e
r
s
.
I
t
i
s
r
e
por
t
e
d
t
ha
t
m
os
t
of
co
n
v
en
t
i
o
n
al
cl
u
s
t
er
i
n
g
al
g
o
r
i
t
h
m
s
ar
e
w
eak
i
n
ex
t
r
act
i
n
g
cen
t
er
s
f
r
o
m
t
i
m
e
-
se
r
i
e
s
d
a
t
a
se
t
s
[5
9
]
.
F
u
r
th
e
r
m
o
r
e
,
in
th
is
w
o
r
k
,
th
e
r
e
e
x
is
t
tw
o
r
e
a
l
-
wo
r
l
d
t
i
m
e
-
se
r
i
e
s
d
a
t
a
se
t
s,
t
h
e
a
i
r
-
pol
l
ut
a
nt
da
t
a
s
e
t
s
a
nd
t
he
f
or
e
x
E
U
R
U
S
D
pa
i
r
da
t
a
s
e
t
s
.
T
hus
,
r
e
s
e
a
r
c
he
r
s
f
ur
t
he
r
f
i
nd
wa
y
s
t
o
i
m
p
r
o
v
e
t
h
e
we
a
k
n
e
s
s
o
f
e
x
i
s
t
i
n
g
c
l
u
s
t
e
r
i
n
g
a
l
g
o
r
i
t
h
m
s
.
In
re
c
e
n
t
l
i
t
e
ra
t
u
re
s
,
Is
m
k
h
a
n
[5
2
]
pr
opos
e
d
a
n
i
m
pr
ove
d
K
-
me
a
n
s
a
l
g
o
r
i
t
h
m
w
i
t
h
i
t
e
r
a
t
i
v
e
mu
l
t
i
-
cl
u
s
t
er
i
n
g
ap
p
r
o
ach
es
f
o
r
s
el
ect
i
n
g
cen
t
er
s
w
h
i
l
e
m
i
n
i
m
i
zi
n
g
t
h
e
o
b
j
ect
i
v
e
f
u
n
ct
i
o
n
.
T
h
e
me
t
h
o
d
p
e
r
f
o
r
me
d
fi
rs
t
l
e
v
e
l
c
l
u
s
t
e
ri
n
g
a
n
d
s
t
o
re
d
t
h
e
fi
rs
t
-
ge
ne
r
a
t
i
on
c
e
nt
e
r
s
,
t
he
n
f
ur
t
he
r
r
un
t
he
s
e
c
ond
l
e
ve
l
c
l
us
t
e
r
i
ng
t
o
sh
o
r
t
e
n
t
h
e
c
e
n
t
e
r
s
d
i
st
a
n
c
e
s.
T
h
i
s
m
e
t
h
o
d
y
i
e
l
d
g
o
o
d
a
c
c
u
r
a
c
y
b
u
t
h
e
a
v
y
i
n
c
o
m
p
u
t
a
t
i
o
n
c
o
st
.
S
i
m
i
l
a
r
ap
p
r
o
ach
es
i
s
o
b
s
er
v
ed
i
n
Yu
e
t
a
l
.
[5
3
]
wo
r
k
s
.
Yu
t
e
a
m
u
s
e
s
a
t
r
i
-
le
v
e
l
K
-
me
a
n
s
f
o
r
d
e
t
e
r
mi
n
i
n
g
t
h
e
c
e
n
t
e
r
s
.
Th
is
a
lg
o
r
ith
m
c
a
n
o
v
e
r
c
o
m
e
n
o
is
e
a
n
d
f
ilte
r
in
v
a
lid
d
a
ta
,
b
u
t
c
o
s
tly
in
c
o
m
p
u
ta
tio
n
lo
a
d
.
Th
e
w
o
r
k
i
n
im
p
r
o
v
in
g
K
-
me
a
n
s
a
l
g
o
r
i
t
h
m
w
a
s
f
u
r
t
h
e
r
c
a
r
r
y
-
on
us
i
ng
de
ns
i
t
y
c
onc
e
pt
s
.
Z
ha
ng
e
t
a
l
.
[2
8
]
in
tr
o
d
u
c
e
d
th
e
us
e
s
of
de
ns
i
t
y
c
a
nopy
t
o
i
m
pr
ove
d
K
-
me
a
n
s
a
l
g
o
r
i
t
h
m
a
n
d
y
i
e
l
d
g
o
o
d
accu
r
acy
.
T
h
e
m
et
h
o
d
al
s
o
t
u
r
n
K
-
me
a
n
s
a
l
g
o
r
i
t
h
m
b
e
c
o
me
l
e
s
s
s
e
n
s
i
t
i
v
e
t
o
n
o
i
s
e
d
a
t
a
.
T
h
i
s
me
t
h
o
d
ma
i
n
l
y
f
o
c
u
s
e
s
o
n
c
a
l
c
u
l
a
t
i
n
g
d
e
n
s
i
t
y
o
f
each
o
b
j
ect
o
r
d
at
a
p
o
i
n
t
s
i
n
d
at
as
et
an
d
s
el
ect
ed
o
n
l
y
t
h
e
cen
t
er
s
w
i
t
h
h
i
g
h
es
t
d
en
s
i
t
y
p
o
i
n
t
s
n
ear
b
y
i
t
.
Si
m
i
l
a
r
w
o
r
k
s
w
as
f
o
u
n
d
i
n
l
i
t
er
at
u
r
es
f
r
o
m
ear
l
y
y
ear
s
[5
6
,
60
-
62]
,
u
s
in
g
d
e
n
s
ity
c
o
n
c
e
p
ts
th
a
t
f
o
c
u
s
o
n
co
m
p
u
t
i
n
g
d
en
s
i
t
y
o
f
d
at
a
p
o
i
n
t
s
ar
o
u
n
d
t
h
e
cen
t
er
s
.
In
l
i
t
e
ra
t
u
re
b
y
K
a
n
t
a
n
d
A
n
s
a
ri
[6
3
]
,
in
tr
o
d
u
c
e
s
a
n
in
itia
l
c
e
n
te
r
s
s
e
le
c
tio
n
m
e
th
o
d
f
o
r
K
-
me
a
n
s
al
g
o
r
i
t
h
m
u
s
i
n
g
A
t
k
i
n
s
o
n
i
n
d
ex
es
.
I
n
s
t
ead
o
f
r
an
d
o
m
l
y
se
l
e
c
t
i
n
g
t
h
e
i
n
i
t
i
a
l
c
e
n
t
e
r
s
f
o
r
p
r
o
c
e
e
d
i
n
g
K
-
me
a
n
s
Evaluation Warning : The document was created with Spire.PDF for Python.
r
IS
S
N
:
2252
-
8938
IJ
-
AI
Vo
l
.
8
, N
o
.
1
,
Ma
r
c
h
20
1
9
:
5
4
–
6
2
56
al
g
o
r
i
t
h
m
f
o
r
f
u
r
t
h
er
ad
j
u
s
t
m
en
t
o
f
cen
t
er
s
,
t
h
e
au
t
h
o
r
s
ch
o
o
s
e
t
o
u
s
es
t
h
i
s
ap
p
r
o
ach
es
r
ed
u
ce
t
h
e
p
o
s
s
i
b
l
e
er
r
o
r
o
f
i
n
v
al
i
d
d
at
a
f
r
o
m
r
an
d
o
m
i
ze
s
el
ect
i
o
n
ap
p
r
o
ach
es
.
A
t
k
i
n
s
o
n
i
n
d
ex
es
u
s
es
b
o
u
n
d
ar
y
an
d
in
e
q
u
a
lity
to
de
t
e
r
m
i
ne
t
he
r
a
nge
of
c
e
nt
e
r
s
s
e
l
e
c
t
i
on
f
or
K
-
me
a
n
s
a
l
g
o
r
i
t
h
m.
I
n
e
a
r
l
y
w
o
r
k
s
b
y
D
i
n
g
e
t
a
l
.
[6
4
]
,
u
s
e
s
si
m
i
l
a
r
c
o
n
c
e
p
t
s
i
n
t
h
e
i
r
p
r
o
p
o
se
d
Y
i
n
y
a
n
g
K
-
me
a
n
s
a
l
g
o
r
i
t
h
m.
T
h
i
s
a
l
g
o
r
i
t
h
m
u
s
e
s
u
p
p
e
r
a
n
d
l
o
w
e
r
b
o
u
n
ds
be
t
w
e
e
n
a
poi
nt
a
nd
r
a
ndom
i
ni
t
i
a
l
c
e
nt
e
r
s
f
or
s
e
l
e
c
t
i
ng
s
ui
t
a
bl
e
c
e
nt
e
r
s
.
T
hi
s
a
l
gor
i
t
hm
a
ppl
i
e
d
a
co
m
p
l
i
cat
ed
p
r
o
ces
s
t
o
p
er
f
o
r
m
t
h
e
co
m
p
u
t
at
i
o
n
;
i
t
i
s
n
o
t
eas
i
l
y
u
n
d
er
s
t
an
d
ab
l
e
f
o
r
r
eg
u
l
ar
r
es
ear
ch
er
.
In
ad
d
i
t
i
o
n
,
t
h
er
e
w
er
e
al
s
o
r
es
ear
ch
er
s
ap
p
l
i
ed
ke
r
ne
l
f
unc
t
i
ons
f
or
di
s
t
a
nc
e
c
a
l
c
ul
a
t
i
on
i
n
K
-
me
a
n
s
a
l
g
o
r
i
t
h
m
to
o
b
ta
in
e
d
b
e
tte
r
c
la
s
s
if
ic
a
tio
n
r
e
s
u
lts
[5
8
]
.
B
y
u
s
in
g
k
e
r
n
e
l
f
u
n
c
tio
n
,
th
e
c
o
m
p
lic
a
te
d
d
a
ta
s
e
t
a
r
e
m
a
p
p
e
d
in
to
h
ig
h
e
r
d
im
e
n
s
io
n
w
h
ic
h
tu
r
n
in
g
th
e
d
a
ta
s
e
t
in
to
m
o
r
e
e
a
s
ily
s
e
p
a
r
a
b
le
a
n
d
c
la
s
s
if
y
d
a
ta
s
e
ts
.
T
z
o
r
tz
is
a
n
d
Li
k
a
s
[6
5
]
in
tr
o
d
u
c
e
d
a
m
e
th
o
d
to
a
s
s
ig
n
e
d
w
e
ig
h
ts
to
c
lu
s
te
r
s
r
e
la
tiv
e
to
th
e
ir
v
a
r
ia
n
c
e
.
T
h
is
w
o
r
k
s
s
h
e
d
a
ne
w
i
ns
i
ght
on
us
i
ng
s
t
a
t
i
s
t
i
c
a
l
a
ppr
oa
c
he
s
i
n
m
a
c
hi
ne
l
e
a
r
ni
ng
a
l
gor
i
t
hm
,
how
e
ve
r
i
t
i
s
t
oo
c
om
pl
i
c
a
t
e
d
a
nd
co
s
t
l
y
i
n
co
m
p
u
t
at
i
o
n
.
M
el
n
y
k
o
v
an
d
M
el
n
y
k
o
v
[6
6
]
pr
opos
e
d
t
he
us
e
s
of
M
a
ha
l
a
nobi
s
di
s
t
a
nc
e
t
o
r
e
pl
a
c
e
th
e
Eu
c
l
i
d
e
a
n
d
i
s
t
a
n
c
e
i
n
K
-
me
a
n
s
a
l
g
o
r
i
t
h
m.
T
h
e
a
p
p
r
o
a
c
h
e
s
y
i
e
l
d
n
o
t
s
i
g
n
i
f
i
c
a
n
t
im
p
r
o
v
e
m
e
n
t in
a
c
c
u
r
a
c
ie
s
.
Ab
o
v
e
l
i
t
e
r
a
t
u
r
e
s
n
o
t
f
o
c
u
s
o
n
d
i
s
t
a
n
c
e
-
ba
s
e
d
r
a
t
i
o
or
w
e
i
ght
a
ge
f
or
s
e
l
e
c
t
i
ng
a
nd
de
t
e
r
m
i
ne
t
he
accu
r
at
e
p
o
s
i
t
i
o
n
cen
t
er
s
.
H
o
w
ev
er
,
s
o
m
e
l
i
t
er
at
u
r
es
t
h
at
f
o
cu
s
o
n
d
en
s
i
t
y
o
f
each
p
o
i
n
t
p
r
o
v
i
d
ed
i
n
s
i
g
h
t
f
o
r
our
w
or
k
t
o
pr
ogr
e
s
s
i
n
us
i
ng
di
s
t
a
nc
e
-
ba
s
e
d
w
e
i
g
ht
f
or
s
e
l
e
c
t
i
ng
be
t
t
e
r
c
e
nt
e
r
s
f
or
R
B
F
N
t
r
a
i
ni
ng.
U
s
i
ng
j
us
t
di
s
t
a
nc
e
-
ra
t
i
o
a
s
w
e
i
g
h
t
,
t
h
e
c
e
n
t
e
rs
s
e
l
e
c
t
i
o
n
a
l
g
o
ri
t
h
m
i
n
K
-
me
a
n
s
c
a
n
p
r
e
c
e
d
e
f
a
s
t
e
r
c
o
mp
u
t
a
t
i
o
n
w
i
t
h
o
u
t
ha
ve
t
o
gone
t
hr
ough
c
om
pl
i
c
a
t
e
d
a
l
gor
i
t
hm
a
nd
c
os
t
l
y
c
om
put
a
t
i
on.
T
hi
s
pa
pe
r
i
s
or
ga
n
iz
e
d
w
ith
th
e
fo
l
l
o
w
i
n
g
s
e
c
t
i
o
n
d
e
s
c
ri
b
i
n
g
t
h
e
s
t
a
n
d
a
rd
K
-
me
a
n
s
a
l
g
o
r
i
t
h
m,
f
o
l
l
o
w
s
b
y
i
mp
r
o
v
e
me
n
t
d
o
n
e
f
o
r
K
-
me
a
n
s
al
g
o
r
i
t
h
m
,
an
d
f
i
n
al
l
y
t
h
e
p
r
o
p
o
s
ed
t
r
ai
n
i
n
g
m
et
h
o
d
u
s
ed
f
o
r
s
i
m
u
l
at
i
n
g
an
d
p
r
ed
i
ct
i
n
g
t
h
e
ei
g
h
t
ex
p
er
i
m
en
t
al
mo
d
e
l
s
.
T
h
e
n
,
i
n
s
e
c
t
i
o
n
4
d
i
s
c
us
s
e
d
t
he
r
e
s
ul
t
s
of
e
a
c
h
m
ode
l
s
a
nd
c
om
pa
r
e
d
w
i
t
h
s
t
a
nda
r
d
R
B
F
N
f
or
accu
r
aci
es
.
F
i
n
al
l
y
,
s
ect
i
o
n
5
co
n
cl
u
d
es
t
h
e
f
i
n
d
i
n
g
s
an
d
d
i
s
cu
s
s
ed
s
o
m
e
f
u
t
u
r
e
w
o
r
k
t
h
at
w
o
u
l
d
h
el
p
in
im
p
r
o
v
in
g
th
e
p
r
o
p
o
s
e
d
tr
a
in
in
g
m
e
th
o
d
.
3.
ME
T
H
O
D
O
L
O
G
Y
3.
1.
S
ta
n
d
a
r
d
K
-
me
a
n
s
al
gor
i
t
h
m
K
-
me
a
n
s
(
K
M
)
a
l
g
o
r
i
t
h
m
[6
2
]
is
o
n
e
o
f
th
e
s
im
p
le
s
t
u
n
s
u
p
e
r
v
is
e
d
le
a
r
n
in
g
a
lg
o
r
ith
m
s
th
a
t
s
o
lv
e
th
e
we
l
l
-
know
n
c
l
us
t
e
r
i
ng
pr
obl
e
m
.
I
t
i
s
a
n
a
l
gor
i
t
hm
ba
s
e
d
on
f
i
ndi
ng
da
t
a
c
l
us
t
e
r
s
i
n
a
da
t
a
s
e
t
s
uc
h
t
ha
t
a
c
os
t
fu
n
ct
i
o
n
o
f
d
i
s
s
i
m
i
l
ar
i
t
y
(
d
i
s
t
an
ce)
m
eas
u
r
e
i
s
m
i
n
i
m
i
zed
.
T
h
e
p
r
o
ced
u
r
e
f
o
l
l
o
w
s
a
s
i
m
p
l
e
an
d
eas
y
t
o
cl
as
s
i
f
y
w
i
t
h
a
g
i
v
en
d
at
a
s
et
t
h
r
o
u
g
h
s
o
m
e
cl
u
s
t
er
s
f
i
x
ed
a
p
r
i
o
r
i
.
T
h
e
m
ai
n
i
d
ea
i
s
t
o
d
ef
i
n
e
K
cen
t
r
es
,
o
n
e
fo
r
e
a
c
h
c
l
u
s
t
e
r.
In
o
t
h
e
r
w
o
rd
;
K
-
me
a
n
s
a
l
g
o
r
ith
m
is
a
n
a
lg
o
r
ith
m
to
c
la
s
s
if
y
o
r
g
r
o
u
p
o
b
je
c
ts
b
a
s
e
d
o
n
at
t
r
i
b
u
t
es
o
r
f
eat
u
r
es
i
n
t
o
K
n
u
m
b
er
o
f
g
r
o
u
p
s
.
K
i
s
a
p
o
s
i
t
i
v
e
i
n
t
eg
er
n
u
m
b
er
.
T
h
e
g
r
o
u
p
i
n
g
i
s
d
o
n
e
b
y
mi
n
i
mi
z
i
n
g
t
h
e
s
u
m
s
q
u
a
r
e
s
o
f
d
i
s
t
a
n
c
e
s
b
e
t
w
e
e
n
d
a
t
a
a
n
d
t
h
e
c
o
r
r
e
s
p
o
n
d
i
n
g
c
l
u
s
t
e
r
c
e
n
t
r
e
[2
9
,
56]
.
T
h
u
s
,
th
e
p
u
r
p
o
s
e
o
f
K
-
me
a
n
s
a
l
g
o
r
i
t
h
m
i
s
t
o
c
l
a
s
s
i
f
y
t
h
e
d
a
t
a
.
Th
e
b
a
s
i
c
s
t
e
p
o
f
K
-
m
ean
s
al
g
o
r
i
t
h
m
i
s
s
i
m
p
l
e:
It
e
ra
t
e
u
n
t
i
l
s
t
a
b
l
e
(n
o
o
b
j
e
c
t
m
o
v
e
g
ro
u
p
)
a)
De
t
e
r
m
i
n
e
t
h
e
c
e
n
t
r
e
c
o
o
r
d
i
n
a
t
e
.
b)
De
t
e
r
m
i
n
e
t
h
e
d
i
s
t
a
n
c
e
o
f
e
a
c
h
o
b
j
e
c
t
t
o
t
h
e
c
e
n
t
r
e
.
c)
Gr
o
u
p
t
h
e
o
b
j
e
c
t
b
a
s
e
d
o
n
m
i
n
i
m
u
m
d
i
s
t
a
n
c
e
.
To
d
e
s
c
r
i
b
e
t
h
e
a
l
g
o
r
i
t
h
m
,
w
e
n
e
e
d
s
o
m
e
n
o
t
a
t
i
o
n
s
.
A
s
e
t
o
f
n
ve
c
t
or
s
x
j
X
R
∈
,
j
=
1
,
…
,
n
,
a
r
e
to
b
e
pa
r
t
i
t
i
one
d
i
nt
o
c
gr
oups
G
i
,
i
=
1
,
…
,
c
.
T
h
e
c
o
s
t
f
u
n
c
tio
n
,
b
a
s
e
d
o
n
th
e
E
u
c
lid
e
a
n
d
is
ta
n
c
e
b
e
tw
e
e
n
a
v
e
c
to
r
x
k
in
g
r
o
u
p
j
an
d
t
h
e
co
r
r
es
p
o
n
d
i
n
g
cl
u
s
t
er
cen
t
r
e
c
i
,
is
d
e
f
in
e
d
b
y
:
2
11
,
ki
CC
ik
i
ii
k
x
G
JJ
x
c
==
∈
==
−
⎛⎞
⎜⎟
⎝⎠
∑∑
∑
(1
)
wh
e
r
e
2
,
ki
ik
i
kx
G
Jx
c
∈
=
−
∑
is
th
e
c
o
s
t f
u
n
c
tio
n
w
ith
in
g
r
o
u
p
i
.
Th
e
p
a
r
t
i
t
i
o
n
e
d
g
r
o
u
p
s
a
r
e
d
e
f
i
n
e
d
b
y
a
c
x
n
bi
na
r
y
m
e
m
be
r
s
hi
p
m
a
t
r
i
x
U
,
w
h
e
r
e
th
e
e
le
m
e
n
t
u
ij
is
1,
i
f
t
he
j
th
d
a
ta
p
o
in
t
x
j
be
l
ongs
t
o
gr
oups
i
an
d
0
o
t
h
er
w
i
s
e.
O
n
ce
t
h
e
cl
u
s
t
er
cen
t
r
es
c
i
,
a
r
e
f
ix
e
d
,
th
e
mi
n
i
mi
z
i
n
g
u
ij
fo
r
(1
)
i
s
d
e
ri
v
e
d
a
s
fo
l
l
o
w
s
:
Evaluation Warning : The document was created with Spire.PDF for Python.
IJ
-
AI
IS
S
N
:
2252
-
8938
r
Di
s
t
a
n
c
e
w
e
i
g
h
t
e
d
K
-
Me
a
n
s
a
l
g
o
r
i
t
h
m
f
o
r
c
e
n
t
e
r
s
e
l
e
c
t
i
o
n
i
n
t
r
a
i
n
i
n
g
r
a
d
i
a
l
b
a
s
i
s
…
(
L
i
m
E
n
g
A
i
k
)
57
22
1
,
,
(
_
ta
n
)
0,
ji
j
k
ij
if
x
c
x
c
k
i
S
h
o
s
te
s
t
d
is
t
U
e
ls
e
−≤
−
≠
=
⎧
⎨
⎩
(2
)
wh
i
c
h
m
e
a
n
s
t
h
a
t
x
j
be
l
ongs
t
o
gr
oup
i
if
c
i
is
th
e
c
lo
s
e
s
t c
e
n
tr
e
a
m
o
n
g
a
ll c
e
n
tr
e
s
.
On
t
h
e
o
t
h
e
r
h
a
n
d
,
i
f
t
h
e
m
e
m
b
e
r
s
h
i
p
m
a
t
r
i
x
i
s
f
i
x
e
d
,
i
.
e
.
i
f
u
ij
is
f
ix
e
d
,
th
e
n
th
e
o
p
tim
a
l
c
e
n
tr
e
c
i
th
a
t
mi
n
i
mi
z
e
s
(1
)
i
s
t
h
e
m
e
a
n
o
f
a
l
l
v
e
c
t
o
rs
i
n
g
ro
u
p
i
:
,
1
ki
ik
kx
G
i
cx
G
∈
=
∑
(3
)
wh
e
r
e
i
G
is
th
e
s
iz
e
o
f
G
i
, o
r
1
n
ii
j
j
Gu
=
=
∑
.
Th
e
a
l
g
o
r
i
t
h
m
i
s
p
r
e
s
e
n
t
e
d
b
e
l
o
w
:
a)
In
i
t
i
a
l
t
h
e
c
l
u
s
t
e
r
c
e
n
t
re
c
i
,
i
=
1
,
…
,
c
. T
h
is
is
d
o
n
e
b
y
r
a
n
d
o
m
ly
s
e
le
c
tin
g
c
poi
nt
a
m
ong
a
l
l
da
t
a
poi
nt
s
.
b)
De
t
e
r
m
i
n
e
m
e
m
b
e
r
s
h
i
p
m
a
t
r
i
x
U
by
(
2)
.
c)
Co
m
p
u
t
e
t
h
e
co
s
t
fu
n
c
t
i
o
n
a
c
c
o
rd
i
n
g
t
o
(1
).
S
t
o
p
i
f
e
i
t
h
e
r
i
t
i
s
b
e
l
o
w
a
c
e
rt
a
i
n
t
o
l
e
ra
n
c
e
v
a
l
u
e
o
r
i
t
s
im
p
r
o
v
e
m
e
n
t o
v
e
r
p
r
e
v
io
u
s
ite
r
a
tio
n
is
b
e
lo
w
a
c
e
r
ta
in
th
r
e
s
h
o
ld
. U
p
d
a
te
c
lu
s
t
er
cen
t
r
e
acco
r
d
i
n
g
t
o
(3
).
d)
Re
p
e
a
t
s
t
e
p
2
a
n
d
3
u
n
t
i
l
c
o
n
v
e
r
g
e
n
c
e
i
s
a
c
h
i
e
v
e
d
,
u
n
t
i
l
t
he
r
e
ar
e
n
o
m
o
r
e
o
b
j
ect
m
o
v
e
g
r
o
u
p
s
,
th
e
n
s
to
p
.
Si
n
c
e
w
e
a
r
e
n
o
t
t
a
l
k
i
n
g
a
b
o
u
t
t
h
e
l
o
c
a
t
i
o
n
o
f
t
h
e
c
e
n
t
r
e
,
w
e
n
e
e
d
t
o
a
d
j
u
s
t
t
h
e
c
e
n
t
r
e
l
o
c
a
t
i
o
n
b
a
s
e
d
on
t
he
c
ur
r
e
nt
upda
t
e
d
da
t
a
.
T
he
n
w
e
a
s
s
i
gn
a
l
l
t
he
da
t
a
t
o
t
hi
s
ne
w
c
e
nt
r
e
.
T
hi
s
pr
oc
e
s
s
i
s
r
e
pe
a
t
e
d
unt
il
th
e
r
e
is
n
o
m
o
r
e
d
a
ta
m
o
v
in
g
to
a
n
o
th
e
r
c
lu
s
te
r
.
3.
2
.
Di
s
t
a
n
c
e
W
e
i
g
h
t
e
d
K
-
M
ea
n
s
(D
W
K
M
)
Al
g
o
r
i
t
h
m
Ty
p
i
c
a
l
l
y
,
K
-
me
a
n
s
al
g
o
r
i
t
h
m
ch
o
s
e
i
t
s
cen
t
er
s
u
n
i
f
o
r
m
l
y
an
d
r
an
d
o
m
l
y
f
r
o
m
v
ect
o
r
X
R
.
To
r
e
d
u
c
e
K
-
me
a
n
s
al
g
o
r
i
t
h
m
we
a
k
n
e
s
s
d
u
r
i
n
g
c
e
n
t
e
r
s
s
e
l
e
c
t
i
o
n
,
a
n
i
m
p
r
o
v
e
m
e
n
t
i
s
ma
d
e
a
t
S
t
e
p
1
i
n
S
e
c
t
i
o
n
2
.
1
.
We
p
r
o
p
o
s
e
d
a
n
i
m
p
r
o
v
e
d
c
e
n
t
e
r
s
s
e
l
e
c
t
i
o
n
m
e
t
h
o
d
f
o
r
K
-
me
a
n
s
a
l
g
o
r
i
t
hm
us
i
ng
di
s
t
a
nc
e
a
s
w
e
i
ght
a
ge
.
Th
e
i
m
p
r
o
v
e
d
K
-
me
a
n
s
a
l
g
o
r
i
t
h
m
i
s
p
r
e
s
e
n
t
e
d
a
s
f
o
l
l
o
w
s
:
1a
.
S
e
l
e
c
t
a
cen
t
er
1
c
, u
n
if
o
r
m
ly
c
h
o
s
e
n
b
y
r
a
n
d
o
m
f
r
o
m
v
e
c
to
r
X
R
.
1b.
A
s
s
i
gn
a
ne
w
c
e
nt
e
r
i
c
, c
h
o
s
e
n
f
r
o
m
v
e
c
to
r
X
R
wi
t
h
p
r
o
b
a
b
i
l
i
t
y
o
f
2
2
1
m
a
x
,
1
,
2,
3
,
...
,
ii
N
ki
k
xc
iC
xc
=
−
=
−
⎛⎞
⎜⎟
⎜⎟
⎜⎟
⎝⎠
∑
an
d
N
is
th
e
to
ta
l n
u
m
b
e
r
o
f
d
a
ta
s
e
t in
X
R
.
1c
.
R
e
pe
a
t
S
t
e
p
1b.
unt
i
l
a
l
l
t
he
k
cen
t
er
s
i
s
o
b
t
ai
n
ed
.
Th
e
n
t
h
e
r
e
m
a
i
n
i
n
g
s
t
e
p
s
o
f
K
-
me
a
n
s
,
f
r
o
m
S
t
e
p
2
t
o
S
t
e
p
4
,
p
r
o
c
e
e
d
a
s
t
h
e
s
t
a
n
d
a
r
d
K
-
me
a
n
s
a
l
g
o
r
i
t
h
m.
3.
3
.
Di
s
t
a
n
c
e
W
e
i
g
h
t
e
d
K
-
M
ea
n
s
(D
W
K
M
)
Al
g
o
r
i
t
h
m
Th
e
R
B
F
N
w
a
s
c
o
n
s
i
d
e
r
e
d
a
t
h
r
e
e
l
a
y
e
r
n
e
t
w
o
r
k
.
Th
e
i
n
p
u
t
n
o
d
e
s
p
a
s
s
t
h
e
i
n
p
u
t
v
a
l
u
e
s
t
o
t
h
e
co
n
n
ect
i
o
n
ar
cs
.
T
h
e
i
n
t
er
n
al
u
n
i
t
s
f
o
r
m
a
s
i
n
g
l
e
l
ay
er
o
f
L
-
RBF
n
o
d
e
s
,
w
h
e
r
e
t
h
e
G
a
u
s
s
i
a
n
f
u
n
c
t
i
o
n
w
a
s
us
e
d
i
n
t
hi
s
l
a
ye
r
t
ha
t
l
oc
a
l
i
z
e
d
r
e
s
pons
e
f
u
n
ct
i
o
n
s
i
n
t
h
e
i
n
p
u
t
s
p
ace.
T
h
e
h
i
d
d
en
n
o
d
e
r
es
p
o
n
s
es
ar
e
we
i
g
h
t
e
d
a
n
d
t
h
e
o
u
t
p
u
t
n
o
d
e
s
a
r
e
s
i
m
p
l
e
s
u
m
m
a
t
i
o
n
s
o
f
t
h
e
we
i
g
h
t
e
d
r
e
s
p
o
n
s
e
s
.
T
h
e
f
o
r
m
u
l
a
t
i
o
n
o
f
t
h
e
tr
a
in
in
g
a
lg
o
r
ith
m
in
v
o
lv
e
s
a
s
e
t
o
f
in
p
u
t
-
out
put
pa
i
r
s
[
x(
i
)
,
y(
i
)
],
i
=
1
,…
,K
,
w
h
e
r
e
x(
i
)
is
th
e
N
-
di
m
e
ns
i
ona
l
in
p
u
t
v
e
c
to
r
,
y(
i
)
is
th
e
c
o
r
r
e
s
p
o
n
d
in
g
ta
r
g
e
t
o
r
d
e
s
ir
e
d
M
-
di
m
e
ns
i
ona
l
out
put
ve
c
t
or
a
nd
K
is
th
e
n
u
m
b
e
r
o
f
tr
a
in
in
g
e
x
a
m
p
le
s
. T
h
e
s
e
t o
f
in
p
u
t
-
out
put
e
xa
m
pl
e
s
i
s
t
he
i
nf
or
m
a
t
i
on
ba
s
e
us
e
t
o
de
t
e
r
m
i
ne
t
he
va
l
ue
s
of
t
he
unknow
n
pa
r
a
m
e
t
er
s
,
i
.
e.
t
h
e
h
i
d
d
en
n
o
d
e
cen
t
er
s
an
d
r
ad
i
i
an
d
t
h
e
co
n
n
ect
i
o
n
w
ei
g
h
t
s
b
et
w
een
t
h
e
h
i
d
d
en
an
d
t
h
e
o
u
t
p
u
t
l
ay
er
.
T
h
e
r
em
ai
n
i
n
g
o
f
t
h
e
R
B
F
N
p
r
o
ced
u
r
e
ar
e
cal
cu
l
at
ed
u
s
i
n
g
s
t
an
d
ar
d
m
et
h
o
d
s
.
4.
RE
S
UL
T
S
A
ND
DI
S
CUS
S
I
O
N
Th
e
p
r
o
p
o
s
e
d
D
W
K
M
-
RBF
N
w
a
s
t
e
s
t
e
d
u
s
i
n
g
4
n
o
n
l
i
n
e
a
r
f
u
n
c
t
i
o
n
f
r
o
m
l
i
t
e
r
a
t
u
r
e
s
,
w
h
i
c
h
a
r
e
Sa
n
t
n
e
r
e
t
a
l
.
[6
7
]
fu
n
c
t
i
o
n
g
i
v
e
n
i
n
(4
),
L
i
m
e
t
a
l
.
[6
8
]
fu
n
c
t
i
o
n
i
n
(
5)
,
D
e
t
t
e
a
nd
P
e
pe
l
ys
he
v
[6
9
]
fu
n
c
t
i
o
n
i
n
(6
),
a
n
d
F
ri
e
d
m
a
n
[7
0
]
fu
n
c
t
i
o
n
i
n
(7
).
F
o
r
a
l
l
t
h
e
s
e
4
fu
n
c
t
i
o
n
s
,
t
h
e
t
ra
i
n
i
n
g
s
e
t
fo
r
R
B
F
N
Evaluation Warning : The document was created with Spire.PDF for Python.
r
IS
S
N
:
2252
-
8938
IJ
-
AI
Vo
l
.
8
, N
o
.
1
,
Ma
r
c
h
20
1
9
:
5
4
–
6
2
58
co
n
s
i
s
t
s
o
f
4
0
0
s
et
s
o
f
r
an
d
o
m
g
en
er
at
ed
d
at
a
p
o
i
n
t
s
an
d
t
es
t
s
et
co
m
p
r
i
s
es
4
0
0
s
et
s
o
f
r
an
d
o
m
g
en
er
at
ed
d
at
a
poi
nt
s
,
bot
h
i
n
r
a
nge
of
[
0,
1]
.
[
]
(
)
e
x
p
(
1
.
4
)
c
o
s
(3
.
5
)
,
0
,
1
fx
x
x
x
π
=
−
∈
(4
)
(
)
(
)
(
)
(
)
(
)
[
]
[
]
11
2
1
3
0
5
s
in
5
4
e
x
p
5
1
0
0
,
0
,
1
,
1
,
2
.
6
i
fx
x
x
x
x
i
=+
+
−
−
∈
∀
=
(5
)
(
)
(
)
(
)
(
)
[
]
2
22
2
12
2
2
3
3
42
8
8
3
4
1
6
1
2
1
,
0
,
1
,
1
,
2
,
3
.
i
fx
x
x
x
x
x
x
x
i
=
−
+
−
+
−
+
+
−
∈
∀
=
(6
)
(
)
(
)
(
)
[
]
2
12
3
4
5
10
s
i
n
20
0.
5
10
5
,
0,
1
,
1
,
2,
3
,
4,
5.
i
fx
x
x
x
x
x
x
i
π
=+
−
+
+
∈
∀
=
(7
)
DW
KM
-
RBF
N
w
a
s
a
l
s
o
t
e
s
t
e
d
o
n
4
r
e
a
l
-
wo
r
l
d
d
a
t
a
s
e
t
s
f
o
r
i
t
s
p
e
r
f
o
r
m
a
n
c
e
.
T
h
e
B
i
o
c
h
e
m
i
c
a
l
Ox
y
g
e
n
De
m
a
n
d
(
B
OD)
c
o
n
c
e
n
t
r
a
t
i
o
n
d
a
t
a
s
e
t
,
p
h
y
t
o
p
l
a
n
k
t
o
n
g
r
o
wt
h
r
a
t
e
a
n
d
d
e
a
t
h
r
a
t
e
d
a
t
a
s
e
t
a
n
d
T
e
x
a
s
ai
r
p
o
l
l
u
t
an
t
d
at
as
et
w
er
e
o
b
t
ai
n
ed
f
r
o
m
A
i
k
an
d
Z
ai
n
u
d
d
i
n
[7
1
]
.
A
n
o
th
e
r
is
th
e
d
a
ta
s
e
t
f
r
o
m
f
o
r
e
x
f
o
r
EU
R
U
S
D
p
a
i
r
s
i
s
c
o
l
l
e
c
t
e
d
f
r
o
m
X
M
M
e
t
a
t
ra
d
e
r
4
d
a
t
a
b
a
s
e
[7
2
]
.
T
h
e
B
O
D
d
a
ta
s
e
t
a
n
d
p
h
y
to
p
la
n
k
to
n
da
t
a
s
e
t
c
ons
i
s
t
s
of
100
s
e
t
s
of
da
t
a
a
nd
t
he
t
e
s
t
s
e
t
c
om
pr
i
s
e
s
100
s
e
t
s
of
da
t
a
.
M
e
a
nw
hi
l
e
,
f
or
a
i
r
pol
l
ut
a
nt
da
t
a
s
e
t
,
t
he
t
r
a
i
ni
ng
s
e
t
ha
s
480
s
e
t
s
of
da
t
a
a
nd
t
he
t
e
s
t
s
e
t
ha
s
72
se
t
s
o
f
d
a
t
a
,
w
h
i
c
h
b
o
t
h
w
e
r
e
t
a
k
e
n
f
r
o
m
hour
l
y
a
i
r
da
t
a
.
F
or
E
U
R
U
S
D
pa
i
r
s
,
t
he
t
r
a
i
ni
ng
s
e
t
c
ons
i
s
t
s
of
519
s
e
t
s
of
da
t
a
t
a
ke
n
f
r
om
ye
a
r
2016
t
o
e
nd
of
ye
a
r
2017.
T
he
t
e
s
t
s
e
t
c
ons
i
s
t
s
of
155
s
e
t
s
of
da
t
a
t
a
ke
n
f
r
om
e
a
r
l
y
ye
a
r
2018
t
o
A
ugus
t
2018.
Th
e
ex
p
er
i
m
en
t
w
as
i
m
p
l
em
en
t
ed
b
y
u
s
i
n
g
t
h
e
n
ew
r
b
f
u
n
ct
i
o
n
b
ecau
s
e
i
t
r
ep
r
es
en
t
s
t
h
e
g
en
er
al
f
o
r
m
o
f
an
R
B
F
ne
t
w
or
k.
F
ur
t
he
r
m
or
e
,
t
he
pr
opos
e
d
c
l
us
t
e
r
i
ng
m
e
t
hod
ha
s
be
e
n
i
m
pl
e
m
e
nt
e
d
by
us
i
ng
M
A
T
L
A
B
’
s
f
unc
t
i
on.
Ga
u
s
s
i
a
n
b
a
s
i
s
f
u
n
c
t
i
o
n
h
a
s
b
e
e
n
u
s
e
d
f
o
r
b
o
t
h
n
e
t
wo
r
ks
w
i
t
h
ot
he
r
pa
r
a
m
e
t
e
r
s
s
uc
h
a
s
s
pr
e
a
d
w
a
s
s
e
t
t
o
de
f
a
ul
t
va
l
ue
,
s
o
t
he
pe
r
f
or
m
a
nc
e
of
t
he
pr
opos
e
d
ne
t
w
or
k
w
a
s
e
va
l
ua
t
e
d
e
f
f
e
c
t
i
ve
l
y
[
8]
.
P
e
r
f
or
m
a
nc
e
of
st
a
n
d
a
r
d
R
B
F
N
a
n
d
D
W
K
M
-
RBF
N
i
n
t
h
i
s
e
x
p
e
r
i
m
e
n
t
h
a
s
b
e
e
n
m
e
a
s
u
r
e
d
b
y
c
o
m
p
a
r
i
n
g
t
h
e
c
o
m
p
u
t
a
t
i
o
n
tim
e
ta
ke
n
f
or
t
r
a
i
ni
ng
w
i
t
h
num
be
r
of
i
t
e
r
a
t
i
on
t
a
ke
n
f
or
c
onve
r
ge
nc
e
a
nd
t
he
R
oot
M
e
a
n
S
qua
r
e
d
E
r
r
or
(R
M
S
E
)
t
o
m
e
a
s
u
re
h
o
w
w
e
l
l
b
o
t
h
n
e
t
w
o
rk
s
a
p
p
ro
x
i
m
a
t
e
s
t
h
e
c
h
o
s
e
n
fu
n
c
t
i
o
n
s
a
n
d
i
t
i
s
g
i
v
e
n
b
y
(
)
2
1
1
n
ii
i
RM
S
E
O
P
n
=
=
−
∑
wh
e
r
e
n
is
th
e
n
u
m
b
e
r
o
f
p
r
e
d
ic
te
d
r
e
sp
o
n
d
e
r
;
i
O
is
th
e
ta
r
g
e
t
v
a
lu
e
f
o
r
tim
e
-
st
e
p
i
,
a
n
d
i
P
is
th
e
p
r
e
d
ic
te
d
va
l
ue
of
t
he
m
ode
l
a
t
t
i
m
e
-
st
e
p
i
.
Th
e
n
u
m
b
e
r
o
f
c
e
n
t
e
r
s
i
s
f
i
x
e
d
t
o
1
0
c
e
n
t
e
r
s
f
o
r
a
l
l
8
d
a
t
a
s
e
t
s
t
r
a
i
n
i
n
g
.
F
o
r
a
i
r
p
o
l
l
u
t
a
n
t
p
r
o
b
l
e
m
,
t
h
e
pol
l
ut
a
nt
m
oni
t
or
e
d
i
nc
l
ude
s
c
a
r
bon
m
onoxi
de
,
ni
t
r
i
c
oxi
de
,
ni
t
r
oge
n
di
oxi
de
,
oz
one
,
a
nd
oxi
de
s
of
ni
t
r
oge
n.
Fo
r
e
x
p
e
r
i
m
e
n
t
a
l
p
u
r
p
o
s
e
s
,
h
o
u
r
l
y
u
p
d
a
t
e
d
ai
r
q
u
al
i
t
y
d
at
a
o
b
t
ai
n
ed
f
r
o
m
A
i
k
an
d
Z
ai
n
u
d
d
i
n
[7
1
]
ha
s
be
e
n
us
e
d
t
o
pr
e
di
c
t
t
he
t
r
e
nd
of
i
nt
e
r
e
s
t
e
d
pol
l
ut
a
nt
s
f
or
N
i
t
r
i
c
O
xi
de
,
N
i
t
r
oge
n
D
i
oxi
de
a
nd
O
xi
de
s
of
N
i
t
r
oge
n.
Wh
i
l
e
f
o
r
P
h
y
t
o
p
l
a
n
k
t
o
n
p
r
o
b
l
e
m
,
g
r
o
w
t
h
r
a
t
e
a
n
d
d
e
a
t
h
r
a
t
e
h
a
v
e
b
e
e
n
u
s
e
d
a
s
t
h
e
i
n
t
e
r
e
s
t
e
d
v
a
l
u
e
s
.
A
s
f
o
r
th
e
B
O
D
p
r
o
b
le
m
,
th
e
B
O
D
c
o
n
c
e
n
tr
a
tio
n
h
a
s
b
e
e
n
ta
k
e
n
a
s
th
e
in
te
r
e
s
te
d
v
a
lu
e
.
F
in
a
lly
,
th
e
f
o
r
e
x
E
U
R
U
S
D
da
t
a
s
e
t
c
ons
i
s
t
s
of
t
hr
e
e
va
r
i
a
bl
e
s
i
s
t
a
ke
n
c
ons
i
de
r
e
d
f
or
t
r
a
i
ni
ng
a
r
e
t
he
da
i
l
y
hi
ghe
s
t
pr
i
c
e
,
da
i
l
y
l
ow
e
s
t
pr
i
c
e
a
nd
ope
n
pr
i
c
e
,
w
hi
l
e
t
he
c
l
os
e
pr
i
c
e
i
s
us
e
d
f
or
pr
e
di
c
t
i
on.
Ta
b
l
e
1
.
Pe
r
f
o
r
m
a
n
c
e
o
f
D
W
K
M
-
RBF
N
a
n
d
S
t
a
n
d
a
r
d
RBF
N
p
r
e
d
i
c
t
i
o
n
r
e
s
u
l
t
s
f
o
r
d
a
t
a
s
e
t
s
.
Da
t
a
s
e
t
St
a
n
d
a
r
d
R
B
F
N
DW
KM
-
RBF
N
Av
e
r
a
g
e
o
f
R
M
S
E
St
a
n
d
a
r
d
D
e
v
i
a
t
i
o
n
o
f
R
M
SE
Av
e
r
a
g
e
o
f
R
M
S
E
St
a
n
d
a
r
d
D
e
v
i
a
t
i
o
n
o
f
R
M
SE
Sa
n
t
n
e
r
0.
160980
0.
132510
0.
032115
0.
015732
Li
m
0.
151190
0.
053404
0.
081227
0.
006912
De
t
t
e
1.
938260
1.
318204
0.
707767
0.
157561
Fr
i
e
d
m
a
n
1.
333590
0.
700492
0.
066972
0.
011266
BO
D
0.
000252
3.
87e
-
07
0.
000245
4.
01e
-
06
Ph
y
t
o
p
l
a
n
k
t
o
n
0.
004980
0.
000933
0.
003884
0.
000858
Ai
r
P
o
l
l
u
t
a
n
t
4.
203630
2.
860265
0.
137055
0.
004483
EU
R
U
SD
0.
031649
0.
004812
0.
026758
0.
005649
Evaluation Warning : The document was created with Spire.PDF for Python.
IJ
-
AI
IS
S
N
:
2252
-
8938
r
Di
s
t
a
n
c
e
w
e
i
g
h
t
e
d
K
-
Me
a
n
s
a
l
g
o
r
i
t
h
m
f
o
r
c
e
n
t
e
r
s
e
l
e
c
t
i
o
n
i
n
t
r
a
i
n
i
n
g
r
a
d
i
a
l
b
a
s
i
s
…
(
L
i
m
E
n
g
A
i
k
)
59
Ta
b
l
e
2
.
Pe
r
c
e
n
t
a
g
e
o
f
I
m
p
r
o
v
e
m
e
n
t
f
o
r
D
W
K
M
-
RBF
N
o
v
e
r
S
t
a
n
d
a
r
d
RBF
N
b
y
RM
S
E
.
Da
t
a
s
e
t
Pe
r
c
e
n
t
a
g
e
o
f
I
m
p
r
o
v
e
m
e
n
t
(
%
)
Sa
n
t
n
e
r
80.
05
Li
m
46.
27
De
t
t
e
63.
48
Fr
i
e
d
m
a
n
94.
98
BO
D
2.
72
Ph
y
t
o
p
l
a
n
k
t
o
n
22.
01
Ai
r
P
o
l
l
u
t
a
n
t
96.
74
EU
R
U
SD
15.
45
Re
s
u
l
t
s
f
r
o
m
T
a
b
l
e
1
sh
o
w
s
t
h
a
t
D
W
K
M
-
RBF
N
n
e
t
w
o
r
k
s
o
u
t
p
e
r
f
o
r
m
s
t
a
n
d
a
r
d
RBF
N
i
n
a
v
e
r
a
g
e
RM
S
E
a
n
d
s
t
a
n
d
a
r
d
d
e
v
i
a
t
i
o
n
o
f
RM
S
E
.
A
l
l
r
e
s
u
l
t
s
o
f
a
v
e
r
a
g
e
RM
S
E
i
n
T
a
b
l
e
1
a
r
e
c
a
l
c
u
l
a
t
e
d
u
s
i
n
g
t
e
n
tim
e
s
r
u
n
o
f
e
a
c
h
n
e
tw
o
r
k
s
.
F
r
o
m
T
a
b
le
1
,
D
W
K
M
-
RBF
N
n
e
t
w
o
r
k
s
u
r
p
a
s
s
e
s
t
h
e
s
t
a
n
d
a
r
d
RBF
i
n
accu
r
acy
an
d
n
et
w
o
r
k
ar
ch
i
t
ect
u
r
e
b
y
u
s
i
n
g
t
r
ai
n
i
n
g
s
et
w
h
i
ch
co
n
s
i
s
t
s
o
n
l
y
8
1
.
8
%
,
7
1
.
8
%
,
6
8
.
5
%
,
an
d
6
5
.
8
%
o
f
t
o
t
al
da
t
a
s
e
t
s
i
z
e
f
or
S
a
nt
ne
r
da
t
a
s
e
t
,
L
i
m
da
t
a
s
e
t
,
D
e
t
t
e
da
t
a
s
e
t
,
a
nd
F
r
i
e
dm
a
n
da
t
a
s
e
t
,
r
e
s
pe
c
t
i
ve
l
y.
W
hi
l
e
DW
KM
-
RBF
N
n
e
t
w
o
r
k
t
r
a
i
n
i
n
g
f
o
r
r
e
a
l
-
wo
r
l
d
d
a
t
a
s
e
t
i
n
v
o
l
v
e
s
t
h
e
B
OD
d
a
t
a
s
e
t
,
P
h
y
t
o
p
l
a
n
k
t
o
n
d
a
t
a
s
e
t
,
Ai
r
pol
l
ut
a
nt
da
t
a
s
e
t
a
nd
f
or
e
x
E
U
R
U
S
D
da
t
a
s
e
t
us
e
d
onl
y
81.
8%
,
71.
8%
,
68.
5%
a
nd
66.
8%
,
r
e
s
pe
c
t
i
ve
l
y.
T
hi
s
me
a
n
s
t
h
a
t
,
i
t
i
s
p
o
s
s
i
b
l
e
t
o
s
u
i
t
a
b
l
e
n
u
mb
e
r
o
f
d
a
t
a
s
e
t
s
u
c
h
t
h
a
t
,
i
t
w
i
l
l
p
r
o
v
i
d
e
a
ne
t
w
or
k
w
i
t
h
r
e
duc
e
d
co
m
p
l
ex
i
t
y
,
f
as
t
er
t
r
ai
n
i
n
g
t
i
m
e
an
d
i
m
p
r
o
v
ed
accu
r
acy
.
Ta
b
l
e
2
s
h
o
w
s
t
h
e
r
e
s
u
l
t
s
o
f
t
h
e
p
e
r
c
e
n
t
a
g
e
o
f
i
m
p
r
o
v
e
m
e
n
t
o
f
R
M
S
E
f
o
r
D
W
K
M
-
RBF
N
n
e
t
w
o
r
k
in
c
o
m
p
a
r
e
d
to
s
ta
n
d
a
r
d
R
B
F
N
.
R
e
s
u
lts
s
h
o
w
e
d
th
a
t
D
W
K
M
-
RBF
N
n
e
t
w
o
r
k
o
u
t
p
e
r
f
o
r
m
s
t
a
n
d
a
r
d
R
BF
N
ne
t
w
or
k
i
n
t
e
r
m
of
a
c
c
ur
a
c
y
m
or
e
t
ha
n
80%
f
or
S
a
nt
ne
r
da
t
a
s
e
t
,
F
r
i
e
dm
a
n
da
t
a
s
e
t
a
nd
a
i
r
pol
l
ut
a
nt
da
t
a
s
e
t
.
Me
a
n
w
h
i
l
e
,
f
o
r
L
i
m
d
a
t
a
s
e
t
,
D
e
t
t
e
d
a
t
a
s
e
t
,
P
h
y
t
o
p
l
a
n
k
t
o
n
d
a
t
a
s
e
t
a
n
d
f
o
r
e
x
E
U
R
U
S
D
d
a
t
a
s
e
t
,
e
a
c
h
o
b
t
a
i
n
e
d
im
p
r
o
v
e
m
e
n
t
in
r
a
n
g
e
o
f
1
5
%
to
6
4
%
.
H
o
w
e
ve
r
,
B
O
D
da
t
a
s
e
t
s
how
s
no
s
i
gni
f
i
c
a
nt
i
m
pr
ove
m
e
nt
f
or
DW
KM
-
RBF
N
n
e
t
w
o
r
k
o
v
e
r
s
t
a
n
d
a
r
d
RBF
N
w
i
t
h
p
e
r
c
e
n
t
a
g
e
l
e
s
s
t
h
a
n
3
%
.
T
h
i
s
i
s
d
u
e
t
o
t
h
e
h
i
g
h
nonl
i
ne
a
r
i
t
y
of
B
O
D
da
t
a
,
be
s
i
de
s
,
t
he
l
a
c
k
of
a
ddi
t
i
ona
l
i
nput
va
r
i
a
bl
e
t
ha
t
c
ont
r
ol
t
he
c
ha
nge
s
i
n
B
O
D
c
a
n
le
a
d
to
w
e
a
k
pr
e
di
c
t
i
on
r
e
s
ul
t
s
.
Th
e
r
e
s
u
l
t
s
f
r
o
m
r
e
a
l
-
wo
r
l
d
d
a
t
a
s
e
t
s
f
o
r
P
h
y
t
o
p
l
a
n
k
t
o
n
d
a
t
a
s
e
t
a
n
d
f
o
r
e
x
E
UR
US
D
d
a
t
a
s
e
t
s
h
o
we
d
im
p
r
o
v
e
m
e
n
t
o
f
2
2
.0
1
%
a
n
d
1
5
.4
5
%
,
r
e
s
p
e
c
tiv
e
ly
.
B
o
th
o
f
th
is
d
a
ta
s
e
ts
a
r
e
h
ig
h
ly
n
o
n
lin
e
a
r
d
a
ta
.
P
o
s
s
ib
le
ex
i
s
t
e
nc
e
of
ot
he
r
e
nvi
r
onm
e
nt
a
l
f
a
c
t
or
not
i
nc
l
ude
d
i
n
da
t
a
s
e
t
i
s
t
he
r
e
a
s
on
be
hi
nd
t
he
l
ow
i
m
pr
ove
m
e
nt
pe
r
c
e
nt
a
ge
oc
c
ur
s
f
or
P
hyt
opl
a
nkt
on
da
t
a
s
e
t
.
A
ddi
t
i
ona
l
l
y,
f
or
f
or
e
x
E
U
R
U
S
D
da
t
a
s
e
t
,
t
he
l
ow
i
m
pr
ove
m
e
nt
pe
r
c
e
nt
a
ge
i
s
due
t
o
m
a
ny
pos
s
i
bl
e
f
a
c
t
or
s
not
i
nc
l
u
de
d
i
n
t
he
da
t
a
s
e
t
t
ha
t
a
l
s
o
dr
i
ve
s
t
he
m
ove
m
e
nt
of
t
hi
s
cu
r
r
en
cy
v
al
u
es
.
F
act
o
r
t
h
at
can
n
o
t
b
e
q
u
an
t
i
f
i
ed
,
s
u
ch
as
p
o
l
i
t
i
cal
i
n
f
l
u
en
ces
o
r
n
at
u
r
al
d
i
s
as
t
er
,
can
af
f
ect
th
e
c
u
r
r
e
n
c
y
f
lu
c
tu
a
tio
n
.
H
e
n
c
e
,
if
s
u
c
h
f
a
c
to
r
s
c
a
n
q
u
a
n
tif
y
a
n
d
in
c
lu
d
e
s
in
th
e
d
a
ta
s
et
,
i
m
p
r
o
v
em
en
t
i
n
pe
r
c
e
nt
a
ge
i
s
e
xpe
c
t
e
d.
F
r
om
T
a
bl
e
1
a
nd
T
a
bl
e
2,
w
e
obs
e
r
ve
d
t
ha
t
D
W
K
M
-
RBF
N
n
e
t
w
o
r
k
p
r
o
v
i
d
e
s
u
c
h
co
n
s
i
s
t
en
t
r
es
u
l
t
s
ev
en
t
h
at
i
t
u
s
es
l
es
s
d
at
as
et
an
d
s
t
i
l
l
ab
l
e
t
o
p
er
f
o
r
m
s
u
ch
s
at
i
s
f
y
i
n
g
r
es
u
l
t
s
.
T
h
e
r
es
u
l
t
s
o
f
pr
e
di
c
t
i
on
f
r
om
D
W
K
M
-
RB
FN
n
e
t
w
o
r
k
a
r
e
c
o
n
s
i
s
t
e
n
t
a
s
w
e
o
b
s
e
r
v
e
d
f
r
o
m
T
a
b
l
e
1
,
t
h
e
s
t
a
n
d
a
r
d
d
e
v
i
a
t
i
o
n
of
R
M
S
E
w
a
s
m
uc
h
l
ow
e
r
t
ha
n
s
t
a
nda
r
d
R
B
F
N
.
F
i
gur
e
1
s
how
s
t
he
e
r
r
or
ba
r
pl
ot
t
ha
t
di
s
pl
a
ye
d
t
he
co
m
p
ar
i
s
o
n
o
f
er
r
o
r
s
f
o
r
b
o
t
h
t
h
e
n
et
w
o
r
k
s
.
C
l
ear
l
y
,
t
h
e
s
t
an
d
ar
d
R
B
F
N
h
as
l
ar
g
er
er
r
o
r
i
n
co
m
p
ar
ed
t
o
DW
KM
-
RBF
N
n
e
t
w
o
r
k
f
o
r
a
l
l
d
a
t
a
s
e
t
s
.
Th
e
D
W
K
M
-
RBF
N
n
e
t
w
o
r
k
a
n
d
S
t
a
n
d
a
r
d
RBF
N
p
e
r
f
o
r
m
e
d
w
e
l
l
i
n
t
h
e
e
x
p
e
r
i
m
e
n
t
s
f
o
r
d
a
t
a
s
e
t
wi
t
h
s
m
a
l
l
e
r
r
a
n
g
e
t
h
a
t
l
i
e
s
i
n
[
0
,
1
]
.
Ho
we
v
e
r
,
f
o
r
l
a
r
g
e
r
r
a
n
g
e
o
f
d
a
t
a
s
e
t
s
u
c
h
a
s
a
i
r
p
o
l
l
u
t
a
n
t
d
a
t
a
s
e
t
,
t
h
e
s
ta
n
d
a
r
d
R
B
F
N
p
e
r
f
o
r
m
e
d
p
o
o
r
ly
in
a
c
c
u
r
a
c
y
.
B
e
s
id
e
s
,
th
e
D
W
K
M
-
RBF
N
n
e
t
w
o
r
k
i
s
s
u
p
e
r
i
o
r
i
n
a
c
c
u
r
a
c
i
e
s
an
d
t
h
e
ar
ch
i
t
ect
u
r
e
o
f
t
h
e
n
et
w
o
r
k
b
u
t
a
p
r
o
p
er
ad
j
u
s
t
m
en
t
i
n
t
h
e
n
u
m
b
er
o
f
cen
t
er
s
f
o
r
each
d
at
as
et
can
en
h
an
ce
t
h
e
n
et
w
o
r
k
s
accu
r
acy
t
o
m
u
ch
h
i
g
h
er
l
ev
el
.
Th
e
r
e
s
u
l
t
s
i
n
Ta
b
l
e
1
s
h
o
w
s
t
h
a
t
e
v
e
n
w
i
t
h
l
e
s
s
d
a
t
a
s
e
t
us
e
d
f
or
t
r
a
i
ni
ng,
i
f
t
he
r
i
ght
c
e
nt
e
r
s
a
r
e
c
hos
e
n,
c
a
n
i
m
pa
c
t
i
n
pr
e
di
c
t
i
on
a
c
c
ur
a
c
y.
T
he
us
e
s
of
huge
num
be
r
of
da
t
a
s
e
t
doe
s
not
a
l
w
a
ys
gua
r
a
nt
e
e
t
he
de
s
i
r
a
bl
e
pr
e
di
c
t
i
on
a
c
c
ur
a
c
i
e
s
,
but
s
uf
f
i
c
i
e
nt
num
be
r
of
da
t
a
s
e
t
t
ha
t
can
r
ep
r
es
en
t
t
h
e
s
h
ap
e
o
f
t
h
e
d
i
s
t
r
i
b
u
t
i
o
n
f
o
r
t
h
e
d
at
as
et
i
s
en
o
u
g
h
f
o
r
p
r
o
v
i
d
i
n
g
g
o
o
d
p
r
ed
i
ct
i
o
n
r
es
u
l
t
s
.
Fu
r
t
h
e
r
m
o
r
e
,
m
a
n
y
o
f
t
r
a
i
n
i
n
g
d
a
t
a
s
e
t
m
i
g
h
t
c
o
n
t
a
i
n
i
n
v
a
l
i
d
d
a
t
a
t
h
a
t
c
o
u
l
d
j
e
o
p
a
r
d
i
z
e
t
h
e
d
e
s
i
r
e
d
a
c
c
u
r
a
c
y
,
not
m
e
nt
i
oni
ng
t
he
s
i
z
e
of
ne
t
w
or
k
i
t
w
oul
d
c
r
e
a
t
e
a
nd
t
he
t
i
m
e
t
a
ke
n
f
or
t
r
a
i
ni
ng.
He
n
c
e
,
t
h
e
r
e
i
s
n
o
d
e
n
i
a
l
o
n
t
h
e
a
b
i
l
i
t
y
o
f
t
h
e
DW
KM
-
RBF
N
n
e
t
w
o
r
k
a
n
d
s
t
a
n
d
a
r
d
RBF
N
i
n
pr
e
di
c
t
i
on,
but
i
t
c
om
e
s
w
i
t
h
a
he
f
t
y
c
om
pe
ns
a
t
i
on
f
or
t
he
a
c
c
ur
a
c
y
i
f
t
he
pr
ope
r
va
l
ue
of
num
be
r
of
c
e
nt
e
r
s
is
n
o
t
s
e
le
c
te
d
.
As
t
h
e
n
u
m
b
e
r
o
f
d
a
t
a
s
e
t
s
f
o
r
t
h
e
n
e
t
wo
r
k
b
e
c
o
m
e
s
l
e
s
s
e
r
a
n
d
i
t
r
e
s
u
l
t
s
m
u
c
h
s
i
m
p
l
e
r
n
e
t
wo
r
k
ar
ch
i
t
ect
u
r
e
an
d
p
o
s
s
i
b
l
y
f
r
ee
o
f
i
n
v
al
i
d
d
at
as
et
.
A
l
t
h
o
u
g
h
b
o
t
h
m
o
d
el
s
p
r
o
v
i
d
e
s
at
i
s
f
y
i
n
g
r
es
u
l
t
s
,
t
h
e
n
et
w
o
r
k
st
r
u
c
t
u
r
e
a
n
d
a
c
c
u
r
a
c
y
of
t
he
D
W
K
M
-
RBF
N
n
e
t
w
o
r
k
i
s
s
u
p
e
r
i
o
r
c
o
m
p
a
r
e
d
t
o
t
h
e
s
t
a
n
d
a
r
d
RBF
N
n
e
t
w
o
r
k
.
Evaluation Warning : The document was created with Spire.PDF for Python.
r
IS
S
N
:
2252
-
8938
IJ
-
AI
Vo
l
.
8
, N
o
.
1
,
Ma
r
c
h
20
1
9
:
5
4
–
6
2
60
Fi
g
u
r
e
1
.
Er
r
o
r
b
ar
p
lo
t f
o
r
d
a
ta
s
e
ts
u
s
in
g
s
ta
n
d
a
r
d
R
B
F
N
a
n
d
D
W
K
M
-
RBF
N
.
5.
CO
NCL
US
I
O
N
Fo
u
r
l
i
t
e
r
a
t
u
r
e
s
n
o
n
l
i
n
e
a
r
f
u
n
c
t
i
o
n
s
a
n
d
f
o
u
r
r
e
a
l
-
wo
r
l
d
p
r
o
b
l
e
m
s
h
a
v
e
b
e
e
n
s
i
m
u
l
a
t
e
d
i
n
t
hi
s
pa
pe
r
,
wh
e
r
e
we
a
p
p
l
i
e
d
f
o
r
r
e
a
l
-
wo
r
l
d
p
r
o
b
l
e
m
s
o
n
p
r
e
d
i
c
t
i
o
n
o
f
B
OD
p
r
o
b
l
e
m
,
p
h
y
t
o
p
l
a
n
k
t
o
n
p
r
o
b
l
e
m
,
a
i
r
pol
l
ut
i
on
pr
obl
e
m
a
nd
f
or
e
x
E
U
R
U
S
D
pr
i
c
e
pr
e
di
c
t
i
on
pr
obl
e
m
.
T
he
pe
r
f
or
m
a
nc
e
of
bot
h
ne
t
w
or
ks
ha
s
be
e
n
co
m
p
ar
ed
t
o
t
h
e
cas
e
u
s
i
n
g
t
h
e
R
o
o
t
M
ean
Sq
u
a
r
e
d
E
r
r
o
r
(
R
M
SE
)
a
n
d
s
t
a
n
d
a
r
d
d
e
v
i
a
t
i
o
n
a
s
t
h
e
c
r
i
t
e
r
i
a
f
o
r
pe
r
f
or
m
a
nc
e
m
e
a
s
ur
e
m
e
nt
a
nd
ne
t
w
or
k
pr
e
di
c
t
i
on
c
ons
i
s
t
e
nc
y.
Re
s
u
l
t
s
f
r
o
m
a
l
l
e
i
g
h
t
s
t
u
d
i
e
s
s
h
o
w
t
h
a
t
t
h
e
DW
KM
-
RBF
N
n
e
t
w
o
r
k
i
s
b
e
t
t
e
r
t
h
a
n
t
h
e
s
t
a
n
d
a
r
d
RBF
N
i
n
p
r
e
d
i
c
t
i
o
n
a
c
c
u
r
a
c
y
a
n
d
n
e
t
w
o
r
k
a
rc
h
i
t
e
c
t
u
re
.
Th
u
s
,
i
t
i
s
p
o
s
s
i
b
l
e
t
o
i
m
p
r
o
v
e
t
h
e
a
c
c
u
r
a
c
y
o
f
t
h
e
p
r
o
p
o
s
e
d
n
e
t
w
o
r
k
b
y
u
s
i
n
g
s
t
a
t
i
s
t
i
c
a
l
m
e
t
h
o
d
s
t
o
c
h
o
o
s
e
th
e
b
e
s
t
v
a
lu
e
o
f
n
u
m
b
e
r
o
f
c
e
n
te
r
to
b
e
u
s
e
d
f
o
r
d
if
f
e
r
e
n
t
ty
p
e
o
f
d
a
ta
s
e
t.
A
s
c
o
n
c
lu
s
io
n
,
th
e
p
r
o
p
o
s
e
d
ne
t
w
or
k
i
s
f
a
r
s
upe
r
i
or
t
o
t
he
st
a
n
d
a
r
d
R
B
F
N
n
e
t
w
o
r
k
a
s
f
o
r
n
e
t
w
o
r
k
a
r
c
h
i
t
e
c
t
u
r
e
a
n
d
a
c
c
u
r
a
c
y
.
Si
n
c
e
s
e
l
f
-
or
ga
ni
z
e
d
s
e
l
e
c
t
i
on
of
c
e
nt
e
r
s
c
a
n
pe
r
f
or
m
e
d
by
c
l
us
t
e
r
i
ng
a
l
gor
i
t
hm
s
f
or
s
e
l
e
c
t
i
ng
s
i
gni
f
i
c
a
nt
c
e
nt
e
r
s
,
su
f
f
i
c
i
e
n
t
t
o
r
e
p
r
e
se
n
t
t
h
e
d
i
st
r
i
b
u
t
i
o
n
o
f
d
a
t
a
se
t
f
o
r
t
h
e
h
i
d
d
e
n
n
o
d
e
s
h
a
s
b
e
e
n
u
se
d
,
i
t
w
o
u
l
d
b
e
i
n
t
e
r
e
st
i
n
g
i
f
th
e
n
e
tw
o
r
k
s
to
b
e
te
s
te
d
w
ith
h
ig
h
n
o
is
e
tr
a
in
in
g
d
a
ta
to
v
e
r
if
y
th
e
e
f
f
ic
ie
n
c
y
o
f
th
e
c
h
o
s
e
n
c
lu
s
te
r
in
g
al
g
o
r
i
t
h
m
i
n
cen
t
er
s
el
ect
i
o
n
.
RE
F
E
RE
NCE
S
[1
]
H.
S
a
r
i
mv
e
i
s
,
A.
Al
e
x
a
n
d
r
i
d
i
s
,
a
n
d
G.
B
a
f
a
s
,
“
A
f
a
s
t
t
r
a
i
n
i
n
g
a
l
g
o
r
i
t
h
m
f
o
r
R
B
F
n
e
t
wo
r
k
s
b
a
s
e
d
o
n
s
u
b
t
r
a
c
t
i
v
e
cl
u
st
er
i
n
g
,
”
Ne
u
r
o
c
o
m
p
u
t
i
n
g
, v
o
l. 5
1
, p
p
. 5
0
1
-
505,
2003.
[2
]
A.
Al
e
x
a
n
d
r
i
d
i
s
,
E
.
C
h
o
n
d
r
o
d
i
ma
,
N.
Gi
a
n
n
o
p
o
u
l
o
s
,
a
n
d
H.
S
a
r
i
mv
e
i
s
,
“
A
F
a
s
t
a
n
d
E
f
f
i
c
ie
n
t
M
e
th
o
d
f
o
r
T
r
a
in
in
g
Ca
t
e
g
o
r
i
c
a
l
Ra
d
i
a
l
Ba
s
i
s
F
u
n
c
t
i
o
n
N
e
t
w
o
r
k
s
,
”
IE
E
E
T
r
a
n
s
.
N
e
u
r
a
l
N
e
t
w
o
r
k
s
L
e
a
r
n
.
S
y
s
t
.
, p
p
. 1
-
6,
2016.
[3
]
A.
Al
e
x
a
n
d
r
i
d
i
s
,
E
.
C
h
o
n
d
r
o
d
i
ma
,
N.
Gi
a
n
n
o
p
o
u
l
o
s
,
a
n
d
H.
S
a
r
i
mv
e
i
s
,
“
A
f
a
s
t
a
n
d
e
f
f
i
c
i
e
n
t
me
t
h
o
d
f
o
r
t
r
a
i
n
i
n
g
cat
eg
o
r
i
cal
r
ad
ia
l
b
a
s
is
f
u
n
c
tio
n
n
e
tw
o
r
k
s
,”
IE
E
E
T
r
a
n
s
.
N
e
u
r
a
l
N
e
t
w
o
r
k
s
L
e
a
r
n
.
S
y
s
t
.
,
v
o
l.
2
8
,
n
o
.
1
1
,
p
p
.
2
8
3
1
-
2836,
2017.
[4
]
Y.
Hu
,
J
.
J
.
Yo
u
,
J
.
N.
K.
L
i
u
,
a
n
d
T
.
He
,
“
An
e
i
g
e
n
v
e
c
t
o
r
b
a
s
e
d
c
e
n
t
e
r
s
e
l
e
c
t
i
o
n
f
o
r
f
a
s
t
t
r
a
i
n
i
n
g
s
c
h
e
me
o
f
RBF
N
N
,
”
In
f
.
S
c
i
.
(N
y
).
, v
o
l. 4
2
8,
pp.
62
-
75,
2018.
[5
]
E.
M
.
d
a
Si
l
v
a
,
R
.
D
.
M
a
i
a
,
a
n
d
C
.
D
.
C
a
b
a
c
i
n
h
a
,
“
B
e
e
-
in
s
p
ir
e
d
R
B
F
n
e
tw
o
r
k
f
o
r
v
o
lu
m
e
e
s
tim
a
tio
n
o
f
in
d
iv
id
u
a
l
tr
e
e
s
,”
Co
m
p
u
t
.
E
l
e
c
t
r
o
n
.
A
g
r
i
c
.
, v
o
l. 1
5
2
, p
p
. 4
0
1
-
408,
2018.
[6
]
D.
S
h
a
n
a
n
d
X.
Xu
,
“
M
u
l
t
i
-
la
b
e
l
L
e
a
r
n
in
g
M
o
d
e
l
B
a
s
e
d
o
n
M
ul
t
i
-
la
b
e
l
R
a
d
ia
l
B
a
s
is
F
u
n
c
tio
n
N
e
u
r
a
l
N
e
tw
o
r
k
a
n
d
Re
g
u
l
a
r
i
z
e
d
E
x
t
r
e
m
e
L
e
a
r
n
i
n
g
M
a
c
h
i
n
e
,
”
Mo
s
h
i
S
h
i
b
i
e
y
u
R
e
n
g
o
n
g
Z
h
i
n
e
n
g
/
P
a
t
t
e
r
n
R
e
c
o
g
n
i
t
.
A
r
t
i
f
.
I
n
t
e
l
l
.
,
v
o
l.
30,
no.
9,
2017.
[7
]
Y.
S
u
n
et
a
l
.
,
“
T
h
e
a
p
p
lic
a
tio
n
o
f
R
B
F
n
e
u
r
a
l
n
e
tw
o
r
k
b
a
s
e
d
o
n
a
n
t
c
o
l
ony
c
l
us
t
e
r
i
ng
a
l
gor
i
t
hm
t
o
pr
e
s
s
ur
e
s
e
ns
or
,
”
Ch
i
n
e
s
e
J
.
S
e
n
s
o
r
s
A
c
t
u
a
t
o
r
s
, v
o
l. 2
6
, n
o
. 6
, 2
0
1
3
.
[8
]
L.
E.
A
i
k
,
“
M
o
d
i
f
i
e
d
C
l
u
s
t
e
r
i
n
g
A
l
g
o
r
i
t
h
m
s
f
o
r
R
a
d
i
a
l
B
a
s
i
s
Fu
n
c
t
i
o
n
N
e
t
w
o
r
k
s
,
”
U
n
i
v
e
r
s
i
t
i
Sa
i
n
s
M
a
l
a
y
s
i
a
,
2
0
0
6
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IJ
-
AI
IS
S
N
:
2252
-
8938
r
Di
s
t
a
n
c
e
w
e
i
g
h
t
e
d
K
-
Me
a
n
s
a
l
g
o
r
i
t
h
m
f
o
r
c
e
n
t
e
r
s
e
l
e
c
t
i
o
n
i
n
t
r
a
i
n
i
n
g
r
a
d
i
a
l
b
a
s
i
s
…
(
L
i
m
E
n
g
A
i
k
)
6
1
[9
]
S.
M
i
r
j
a
l
i
l
i
,
“
Ev
o
l
u
t
i
o
n
a
r
y
R
a
d
i
a
l
B
a
si
s
F
u
n
ct
i
o
n
N
et
w
o
r
k
s,
”
i
n
St
udi
e
s
i
n
C
om
put
at
i
onal
I
nt
e
l
l
i
ge
nc
e
,
2
0
1
9
,
p
p
.
105
-
139.
[1
0
]
A.
Os
ma
n
o
v
i
ć
,
S
.
Ha
l
i
l
o
v
i
ć
,
L
.
A.
I
l
a
h
,
A.
F
o
j
n
i
c
a
,
a
n
d
Z
.
Gr
o
mi
l
i
ć
,
“
M
a
c
h
i
n
e
l
e
a
r
n
i
n
g
t
e
c
h
n
i
q
u
e
s
f
o
r
cl
assi
f
i
cat
i
o
n
o
f
b
r
east
can
cer
,
”
i
n
IF
M
B
E
P
r
o
c
e
e
d
i
n
g
s
, 2
0
1
9
.
[1
1]
M.
Mo
h
a
m
m
a
d
i
,
A
.
K
r
i
s
h
n
a
,
S
.
N
a
l
e
s
h
,
a
n
d
S
.
K
.
N
a
n
d
y
,
“
A
H
a
r
d
w
a
r
e
A
r
c
h
i
t
e
c
t
u
r
e
f
o
r
R
a
d
i
a
l
B
a
s
i
s
F
u
n
c
t
i
o
n
Ne
u
r
a
l
Ne
t
wo
r
k
C
l
a
s
s
i
f
i
e
r
,
”
IE
E
E
T
r
a
n
s
.
P
a
r
a
l
l
e
l
D
i
s
t
r
i
b
.
S
y
s
t
.
, 2
0
1
8
.
[1
2
]
M.
W.
L
.
Mo
r
e
i
r
a
,
J
.
J
.
P
.
C
.
R
o
d
r
i
g
u
e
s
,
N
.
K
u
m
a
r
,
J
.
A
l
-
Mu
h
t
a
d
i
,
a
n
d
V
.
Ko
r
o
t
a
e
v
,
“
E
v
o
l
u
t
i
o
n
a
r
y
r
a
d
i
a
l
b
a
s
i
s
fu
n
c
t
i
o
n
n
e
t
w
o
rk
fo
r
g
e
s
t
a
t
i
o
n
a
l
d
i
a
b
e
t
e
s
d
a
t
a
a
n
a
l
y
t
i
c
s
,
”
J.
C
o
m
p
u
t
.
S
ci
.
, 2
0
1
8
.
[1
3
]
K.
Ka
v
a
k
l
i
o
g
l
u
,
M
.
F
.
Ko
s
e
o
g
l
u
,
a
n
d
O.
C
a
l
i
s
k
a
n
,
“
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
He
a
t
a
n
d
M
a
s
s
T
r
a
n
s
f
e
r
E
x
p
e
r
i
me
n
t
a
l
in
v
e
s
tig
a
tio
n
a
n
d
r
a
d
ia
l
b
a
s
is
f
u
n
c
tio
n
n
e
tw
o
r
k
m
o
d
e
lin
g
o
f
d
ir
e
c
t
e
v
a
p
o
r
a
tiv
e
c
o
o
lin
g
s
y
s
te
m
s
,”
In
t
.
J
.
H
e
a
t
M
a
s
s
Tr
a
n
s
f
.
, v
o
l. 1
2
6
, p
p
. 1
3
9
-
150
, 2
0
1
8
.
[1
4
]
M.
S
m
o
l
i
k
,
V
.
S
k
a
l
a
,
a
n
d
Z
.
Ma
j
d
i
s
o
v
a
,
“
A
d
v
a
n
c
e
s
i
n
E
n
g
i
n
e
e
r
i
n
g
S
o
f
t
w
a
r
e
V
e
c
t
o
r
f
i
e
l
d
r
a
d
i
a
l
b
a
s
i
s
f
u
n
c
t
i
o
n
ap
p
r
o
x
i
m
at
i
o
n
☆
,”
Ad
v
.
En
g
.
S
o
f
t
w
.
, v
o
l. 1
2
3
, n
o
. 1
7
, p
p
. 1
1
7
-
129,
2018.
[1
5
]
Y.
L
i
,
X.
W
a
n
g
,
S
.
S
u
n
,
X.
M
a
,
a
n
d
G.
L
u
,
“
F
o
r
e
c
a
s
t
i
n
g
s
hor
t
-
te
r
m
s
u
b
w
a
y
p
a
s
s
e
n
g
e
r
f
lo
w
u
n
d
e
r
s
p
e
c
ia
l
e
v
e
n
ts
scen
ar
i
o
s
u
si
n
g
m
u
l
t
i
scal
e
r
ad
i
al
b
asi
s
f
u
n
ct
i
o
n
n
et
w
o
r
k
s,
”
Tr
a
n
s
p
.
Re
s
.
Pa
r
t
C
Em
e
r
g
.
Te
c
h
n
o
l
.
, 2
0
1
7
.
[1
6
]
G.
W
.
C
h
a
n
g
,
H.
J
.
L
u
,
Y.
R
.
C
h
a
n
g
,
a
n
d
Y.
D.
L
e
e
,
“
An
i
mp
r
o
v
e
d
n
e
u
r
a
l
n
e
t
wo
r
k
-
ba
s
e
d
a
ppr
oa
c
h
f
or
s
hor
t
-
te
r
m
wi
n
d
s
p
e
e
d
a
n
d
p
o
we
r
f
o
r
e
c
a
s
t
,
”
Re
n
e
w
.
En
e
r
g
y
, 2
0
1
7
.
[1
7
]
Q.
P
.
Ha
,
H.
W
a
h
i
d
,
H.
Du
c
,
a
n
d
M
.
Az
z
i
,
“
E
n
h
a
n
c
e
d
r
a
d
i
a
l
b
a
s
i
s
f
u
n
c
t
i
o
n
n
e
u
r
a
l
n
e
t
wo
r
k
s
f
o
r
o
z
o
n
e
l
e
v
e
l
est
i
m
at
i
o
n
,
”
Ne
u
r
o
c
o
m
p
u
t
i
n
g
, 2
0
1
5
.
[1
8
]
Z.
M
a
j
d
i
s
o
v
a
a
n
d
V
.
Sk
a
l
a
,
“
R
a
d
i
a
l
b
a
s
i
s
f
u
n
ct
i
o
n
ap
p
r
o
x
i
m
at
i
o
n
s
:
co
m
p
ar
i
so
n
an
d
ap
p
l
i
cat
i
o
n
s,
”
Ap
p
l
.
M
a
t
h
.
Mo
d
e
l
.
, v
o
l. 5
1
, p
p
. 7
2
8
-
743,
2017.
[1
9
]
Y.
L
e
i
,
L
.
Di
n
g
,
a
n
d
W
.
Z
h
a
n
g
,
“
Ge
n
e
r
a
l
i
z
a
t
i
o
n
P
e
r
f
o
r
ma
n
c
e
o
f
R
a
d
i
a
l
B
a
s
i
s
F
u
n
c
t
i
o
n
Ne
t
wo
r
k
s
,
”
Ne
u
r
a
l
Ne
t
wo
r
k
s
L
e
a
r
n
.
S
y
s
t
.
I
E
E
E
T
r
a
n
s
.
, v
o
l. 2
6
,
no.
3,
pp.
551
-
564,
2015.
[2
0
]
C.
A
r
t
e
a
g
a
a
n
d
I
.
M
a
r
r
e
r
o
,
“
U
n
i
v
e
r
s
a
l
a
p
p
r
o
x
i
m
a
t
i
o
n
b
y
r
a
d
i
a
l
b
a
s
i
s
f
u
n
c
t
i
o
n
n
e
t
w
o
r
k
s
o
f
D
e
l
s
a
r
t
e
t
r
a
n
s
l
a
t
e
s
,
”
Ne
u
r
a
l
Ne
t
wo
r
k
s
, v
o
l. 4
6
, p
p
. 2
9
9
-
305,
2013.
[2
1
]
S.
Li
n
,
“
Li
n
e
a
r
a
n
d
n
o
n
l
i
n
e
a
r
a
p
p
r
o
x
i
m
a
t
i
o
n
o
f
s
p
h
e
r
i
c
a
l
r
a
d
i
a
l
ba
s
i
s
f
unc
t
i
on
ne
t
w
or
ks
✩
,”
v
o
l.
3
5
,
p
p
.
8
6
-
101,
2016.
[2
2
]
M.
L
.
A
l
b
a
l
a
t
e
,
“
D
a
t
a
r
e
d
u
c
t
i
o
n
t
e
c
h
n
i
q
u
e
s
i
n
c
l
a
s
s
i
f
i
c
a
t
i
o
n
p
r
o
c
e
s
s
e
s
,
”
2
0
0
7
.
[2
3
]
S.
H
a
y
k
i
n
,
“
N
e
u
r
a
l
n
e
t
w
o
r
k
s
-
A
c
o
mp
r
e
h
e
n
s
i
v
e
f
o
u
n
d
a
t
i
o
n
,
”
Ne
w
Y
o
r
k
:
I
E
E
E
P
r
e
s
s
.
H
e
r
r
m
a
n
n
,
M
.
,
B
a
u
e
r
,
H
.
-
U.
,
&
De
r,
R
, v
o
l. p
s
y
c
h
o
lo
g
y
. p
p
. p
p
1
0
7
-
116,
1994.
[2
4
]
B.
V
.
D
a
s
a
r
a
t
h
y
,
“
A
c
o
m
p
u
t
a
t
i
o
n
a
l
d
e
m
a
n
d
o
p
t
i
m
i
z
a
t
i
o
n
a
i
d
e
f
o
r
n
e
a
r
e
s
t
-
ne
i
ghbor
-
ba
s
e
d
de
c
i
s
i
on
s
ys
t
e
m
s
,
”
i
n
Co
n
f
e
r
e
n
c
e
P
r
o
c
e
e
d
i
n
g
s
1
9
9
1
I
E
E
E
I
n
t
e
r
n
a
t
i
o
n
a
l
Co
n
f
e
r
e
n
c
e
o
n
S
y
s
t
e
m
s
,
M
a
n
,
a
n
d
Cy
b
e
r
n
e
t
i
c
s
,
1
9
9
1
,
p
p.
1777
-
1782.
[2
5
]
B.
D
a
s
a
r
a
t
h
y
,
“
D
a
t
a
m
i
n
i
n
g
t
a
s
k
s
a
n
d
m
e
t
h
o
d
s
:
Cl
a
s
s
i
f
i
c
a
t
i
o
n
:
n
e
a
r
e
s
t
-
ne
i
ghbor
a
ppr
oa
c
he
s
,
”
Ha
n
d
b
.
d
a
t
a
M
i
n
.
Kn
o
wl
.
D
i
s
c
o
v
.
, 2
0
0
2
.
[2
6
]
J.
J
ȩ
d
r
zej
o
w
i
cz
an
d
P
.
J
ȩ
d
r
zej
o
w
i
cz,
“D
i
st
an
ce
-
ba
s
e
d
onl
i
ne
c
l
a
s
s
i
f
i
e
r
s
,
”
Ex
p
e
r
t
S
y
s
t
.
Ap
p
l
.
,
v
o
l.
60,
pp.
249
-
257,
2016.
[2
7
]
M.
K
i
r
s
t
e
n
a
n
d
S
.
Wr
o
b
e
l
,
“
R
e
l
a
t
i
o
n
a
l
d
i
s
t
a
n
c
e
-
ba
s
e
d
c
l
us
t
e
r
i
ng,
”
i
n
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
8
t
h
i
n
t
e
r
n
a
t
i
o
n
a
l
co
n
f
eren
ce
o
n
I
n
d
u
ct
i
ve
l
o
g
i
c
p
ro
g
ra
m
m
i
n
g
,
I
L
P
-
98,
J
ul
y
22
-
24
, 1
9
9
8
, n
o
. 2
0
2
3
7
, p
p
. 2
6
1
-
270.
[2
8
]
G.
Z
h
a
n
g
,
C
.
Z
h
a
n
g
,
a
n
d
H.
Z
h
a
n
g
,
“
I
mp
r
o
v
e
d
K
-
me
a
n
s
a
l
g
o
r
i
t
h
m
b
a
s
e
d
o
n
d
e
n
s
i
t
y
C
a
n
o
p
y
,
”
Kn
o
wl
e
d
g
e
-
Ba
s
e
d
Sy
s
t
.
, 2
0
1
8
.
[2
9
]
W.
L
.
Z
h
a
o
,
C
.
H
.
D
e
n
g
,
a
n
d
C
.
W.
N
g
o
,
“
k
-
me
a
n
s
:
A
r
e
v
i
s
i
t
,
”
Ne
u
r
o
c
o
m
p
u
t
i
n
g
, 2
0
1
8
.
[3
0
]
K.
Da
h
i
y
a
,
“
R
e
d
u
c
i
n
g
Ne
u
r
a
l
Ne
t
wo
r
k
T
r
a
i
n
i
n
g
D
at
a
u
si
n
g
S
u
p
p
o
r
t
V
ect
o
r
s,
”
v
o
l
.
0
,
n
o
.
2
,
p
p
.
6
-
8,
2014.
[3
1
]
M.
A
.
S
e
b
t
o
s
h
e
i
k
h
a
n
d
A
.
S
a
l
e
h
i
,
“
L
i
t
h
o
l
o
g
y
p
r
e
d
i
c
t
i
o
n
b
y
s
u
p
p
o
r
t
v
e
c
t
o
r
c
l
a
s
s
i
f
i
e
r
s
u
s
i
n
g
i
n
v
e
r
t
e
d
s
e
i
s
m
i
c
at
t
r
i
b
u
t
es
d
at
a
an
d
p
et
r
o
p
h
y
si
cal
l
o
g
s
as
a
n
ew
ap
p
r
o
ach
an
d
i
n
v
est
i
g
at
i
o
n
o
f
t
r
ai
n
i
n
g
da
t
a
s
e
t
s
i
z
e
e
f
f
e
c
t
on
i
t
s
pe
r
f
or
m
a
nc
e
i
n
a
he
t
e
r
oge
ne
ous
c
a
r
bona
t
e
r
e
s
e
r
voi
r
,
”
J.
P
et
.
S
ci
.
E
n
g
.
, v
o
l. 1
3
4
, p
p
. 1
4
3
-
149,
2015.
[3
2
]
S.
W
a
n
g
,
Z.
Li
,
C
.
Li
u
,
X
.
Zh
a
n
g
,
a
n
d
H
.
Zh
a
n
g
,
“
Tr
a
i
n
i
n
g
d
a
t
a
r
e
d
u
c
t
i
o
n
t
o
s
p
e
e
d
u
p
SV
M
t
r
a
i
n
i
n
g
,
”
Ap
p
l
.
I
n
t
e
l
l
.
,
vol
.
41,
no.
2,
pp.
405
–
420,
2014.
[3
3
]
S.
M
i
r
j
a
l
i
l
i
,
“
Ev
o
l
u
t
i
o
n
a
r
y
R
a
d
i
a
l
B
a
s
i
s
Fu
n
c
t
i
o
n
N
e
t
w
o
r
k
s
,
”
i
n
St
udi
e
s
i
n
C
om
put
at
i
onal
I
nt
e
l
l
i
ge
nc
e
,
2
0
1
9
,
p
p
.
105
-
139.
[3
4
]
C.
C.
L
i
a
o
,
“
G
e
n
e
t
i
c
k
-
me
a
n
s
a
l
g
o
r
i
t
h
m
b
a
s
e
d
R
B
F
n
e
t
wo
r
k
f
o
r
p
h
o
t
o
v
o
l
t
a
i
c
M
P
P
p
r
e
d
i
c
t
i
o
n
,
”
En
e
rg
y
, 2
0
1
0
.
[3
5
]
M.
A
.
R
a
h
m
a
n
a
n
d
M.
Z
.
I
s
l
a
m
,
“
A
h
y
b
r
i
d
c
l
u
s
t
e
r
i
n
g
t
e
c
h
n
i
q
u
e
c
o
m
b
i
n
i
n
g
a
n
o
v
e
l
g
e
n
e
t
i
c
a
l
g
o
r
i
t
h
m
w
i
t
h
K
-
Me
a
n
s
,
”
Kn
o
wl
e
d
g
e
-
Ba
s
e
d
S
y
s
t
.
, 2
0
1
4
.
[3
6
]
J.
W
u
,
J.
L
o
n
g
,
an
d
M
.
L
i
u
,
“E
v
o
l
v
i
n
g
R
B
F
n
eu
r
al
n
et
w
o
r
k
s
f
o
r
r
ai
n
f
al
l
p
r
ed
i
ct
i
o
n
u
si
n
g
h
y
b
r
i
d
pa
r
t
i
c
l
e
s
w
a
r
m
opt
i
m
i
z
a
t
i
on
a
nd
ge
ne
t
i
c
a
l
gor
i
t
hm
,
”
Ne
u
r
o
c
o
m
p
u
t
i
n
g
, 2
0
1
5
.
[3
7
]
L.
Ta
n
,
“
A
c
l
u
s
t
e
r
i
n
g
K
-
me
a
n
s
a
l
g
o
r
i
t
h
m
b
a
s
e
d
o
n
i
mp
r
o
v
e
d
P
S
O
a
l
g
o
r
i
t
h
m,
”
i
n
Pr
o
c
e
e
d
i
n
g
s
-
2015
5t
h
In
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
C
o
m
m
u
n
i
c
a
t
i
o
n
S
y
s
t
e
m
s
a
n
d
N
e
t
w
o
r
k
T
e
c
h
n
o
l
o
g
i
e
s
,
CS
NT
2
0
1
5
, 2
0
1
5
.
[3
8
]
E.
M
.
d
a
Si
l
v
a
,
R
.
D
.
M
a
i
a
,
a
n
d
C
.
D
.
C
a
b
a
c
i
n
h
a
,
“
B
e
e
-
in
s
p
ir
e
d
R
B
F
n
e
tw
o
r
k
f
o
r
v
o
lu
m
e
e
s
tim
a
tio
n
o
f
in
d
iv
id
u
a
l
tr
e
e
s
,”
Co
m
p
u
t
.
E
l
e
c
t
r
o
n
.
A
g
r
i
c
.
, v
o
l. 1
5
2
, 2
0
1
8
.
[3
9
]
M.
D
i
e
z
,
S
.
V
o
l
p
i
,
A
.
S
e
r
a
n
i
,
F
.
S
t
e
r
n
,
a
n
d
E
.
F
.
C
a
m
p
a
n
a
,
“
S
i
m
u
l
a
t
i
on
-
Ba
s
e
d
D
e
s
i
g
n
O
p
t
i
m
i
z
a
t
i
o
n
b
y
S
e
q
u
e
n
t
i
a
l
Mu
l
t
i
-
cr
i
t
er
i
o
n
A
d
ap
t
i
v
e
S
am
p
l
i
n
g
an
d
D
y
n
am
i
c
R
ad
i
al
B
asi
s
F
u
n
ct
i
o
n
s,
”
i
n
Co
m
p
u
t
a
t
i
o
n
a
l
M
e
t
h
o
d
s
i
n
A
p
p
l
i
e
d
Sc
i
e
nc
e
s
, 2
0
1
9
.
[4
0
]
H.
d
e
L
e
o
n
-
De
l
g
a
d
o
,
R
.
J
.
P
r
a
g
a
-
Al
e
j
o
,
D.
S
.
Go
n
z
a
l
e
z
-
Go
n
z
a
l
e
z
,
a
n
d
M
.
C
a
n
t
ú
-
Si
f
u
e
n
t
e
s,
“M
u
l
t
i
v
ar
i
at
e
st
at
i
st
i
cal
in
f
e
r
e
n
c
e
in
a
r
a
d
ia
l b
a
s
is
f
u
n
c
tio
n
n
e
u
r
a
l n
e
tw
o
r
k
,”
Ex
p
e
r
t
S
y
s
t
.
Ap
p
l
.
, 2
0
1
8
.
[4
1
]
Y.
C
u
i
,
J
.
S
h
i
,
a
n
d
Z
.
W
a
n
g
,
“
L
a
z
y
Qu
a
n
t
u
m
c
l
u
s
t
e
r
i
n
g
i
n
d
u
c
e
d
r
a
d
i
a
l
b
a
s
i
s
f
u
n
c
t
i
o
n
n
e
t
wo
r
k
s
(
L
QC
-
RBF
N
)
w
i
t
h
ef
f
ect
i
v
e
cen
t
er
s
sel
ect
i
o
n
an
d
r
ad
i
i
d
et
er
m
i
n
at
i
o
n
,
”
Ne
u
r
o
c
o
m
p
u
t
i
n
g
, 2
0
1
6
.
Evaluation Warning : The document was created with Spire.PDF for Python.
r
IS
S
N
:
2252
-
8938
IJ
-
AI
Vo
l
.
8
, N
o
.
1
,
Ma
r
c
h
20
1
9
:
5
4
–
6
2
62
[4
2
]
G.
A.
M
o
n
t
a
z
e
r
a
n
d
D.
Gi
v
e
k
i
,
“
An
i
mp
r
o
v
e
d
r
a
d
i
a
l
b
a
s
i
s
f
u
n
c
t
i
o
n
n
e
u
r
a
l
n
e
t
wo
r
k
f
o
r
o
b
j
e
c
t
i
ma
g
e
r
e
t
r
i
e
v
a
l
,
”
Ne
u
r
o
c
o
m
p
u
t
i
n
g
, 2
0
1
5
.
[4
3
]
Z.
R
a
m
e
d
a
n
i
,
M
.
O
m
i
d
,
A
.
K
e
y
h
a
n
i
,
S.
Sh
a
m
s
h
i
r
b
a
n
d
,
a
n
d
B
.
K
h
o
s
h
n
e
v
i
s
a
n
,
“
Po
t
e
n
t
i
a
l
o
f
r
a
d
i
a
l
b
a
s
i
s
f
u
n
c
t
i
o
n
ba
s
e
d
s
uppor
t
ve
c
t
or
r
e
gr
e
s
s
i
on
f
or
gl
oba
l
s
ol
a
r
r
a
di
a
t
i
on
pr
e
di
c
t
i
on,
”
Re
n
e
w
.
S
u
s
t
a
i
n
.
En
e
r
g
y
Re
v
.
, 2
0
1
4
.
[4
4
]
T.
W
a
n
g
c
h
a
m
h
a
n
,
S.
C
h
i
e
w
c
h
a
n
w
a
t
t
a
n
a
,
an
d
K
.
S
u
n
at
,
“E
f
f
i
ci
en
t
al
g
o
r
i
t
h
m
s
b
ased
o
n
t
h
e
k
-
me
a
n
s
a
n
d
C
h
a
o
t
i
c
Le
a
g
u
e
C
h
a
m
p
i
o
n
s
h
i
p
A
l
g
o
r
i
t
h
m
f
o
r
n
u
m
e
r
i
c
,
c
a
t
e
g
o
r
i
c
a
l
,
a
n
d
m
i
x
e
d
-
ty
p
e
d
a
ta
c
lu
s
te
r
in
g
,”
Ex
p
e
r
t
S
y
s
t
.
Ap
p
l
.
,
2017.
[4
5
]
Q.
Qu
e
a
n
d
M
.
B
e
l
k
i
n
,
“
B
a
c
k
t
o
t
h
e
F
u
t
u
r
e
:
R
a
d
i
a
l
B
a
s
i
s
F
u
n
c
t
i
o
n
N
et
w
o
r
k
s
R
ev
i
si
t
ed
,
”
i
n
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
1
9
t
h
In
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
r
e
n
c
e
o
n
A
r
t
i
f
i
c
i
a
l
In
t
e
l
l
i
g
e
n
c
e
a
n
d
S
t
a
t
i
s
t
i
c
s
, 2
0
1
6
.
[4
6
]
V.
K.
C
h
a
u
h
a
n
,
A.
S
h
a
r
ma
,
a
n
d
K.
Da
h
i
y
a
,
“
F
a
s
t
e
r
l
e
a
r
n
i
n
g
b
y
r
e
d
u
c
t
i
o
n
o
f
d
a
t
a
a
c
c
e
s
s
t
i
me
,
”
Ap
p
l
.
I
n
t
e
l
l
.
, 2
0
1
8
.
[4
7
]
G.
Af
e
n
d
r
a
s
an
d
M
.
M
ar
k
at
o
u
,
“O
p
t
i
m
al
i
t
y
o
f
t
r
ai
n
i
n
g
/
t
est
si
ze
an
d
r
esam
p
l
i
n
g
ef
f
ect
i
v
en
ess
i
n
cr
o
ss
-
va
l
i
da
t
i
on,
”
J.
S
t
a
t
.
P
l
a
n
.
I
n
f
eren
ce
, v
o
l. 1
6
, n
o
. x
x
x
x
, p
p
. 1
-
16,
2018.
[4
8
]
S.
O
u
g
i
a
r
o
g
l
o
u
,
K
.
I
.
D
i
a
m
a
n
t
a
r
a
s
,
a
n
d
G
.
Ev
a
n
g
e
l
i
d
i
s
,
“
Ex
p
l
o
r
i
n
g
t
h
e
e
f
f
e
c
t
o
f
d
a
t
a
r
e
duc
t
i
on
on
N
e
ur
a
l
N
e
t
w
or
k
an
d
S
u
p
p
o
r
t
V
ect
o
r
M
ach
i
n
e
cl
assi
f
i
cat
i
o
n
,
”
Ne
u
r
o
c
o
m
p
u
t
i
n
g
, v
o
l. 2
8
0
, p
p
. 1
0
1
-
110,
2017.
[4
9
]
M.
B
a
t
a
i
n
e
h
a
n
d
T
.
Ma
r
l
e
r
,
“
N
e
u
r
a
l
n
e
t
w
o
r
k
f
o
r
r
e
g
r
e
s
s
i
o
n
p
r
o
b
l
e
m
s
w
i
t
h
r
e
d
u
c
e
d
t
r
a
i
n
i
n
g
s
e
t
s
,
”
Ne
u
r
a
l
Ne
t
wo
r
k
s
,
vol
.
95,
pp.
1
–
9,
2017.
[5
0
]
V.
C
h
o
u
v
a
t
u
t
,
W
.
J
i
n
d
a
l
u
a
n
g
,
a
n
d
E
.
B
o
o
n
c
h
i
e
n
g
,
“
T
r
a
i
n
i
n
g
s
e
t
s
i
z
e
r
e
d
u
c
t
i
o
n
i
n
l
a
r
g
e
d
a
t
a
s
e
t
p
r
o
b
l
e
ms
,
”
2015
I
nt
.
Co
m
p
u
t
.
S
c
i
.
E
n
g
.
Co
n
f
.
, p
p
. 1
-
5,
2015.
[5
1
]
W.
A
.
Y
o
u
s
e
f
a
n
d
S
.
K
u
n
d
u
,
“
L
e
a
r
n
i
n
g
a
l
g
o
r
i
t
h
m
s
m
a
y
p
e
r
f
o
r
m
w
o
r
s
e
w
i
t
h
i
n
c
r
e
a
s
i
n
g
t
r
a
i
n
i
n
g
set
si
ze:
A
l
g
o
r
i
t
h
m
-
da
t
a
i
nc
om
pa
t
i
bi
l
i
t
y,
”
Co
m
p
u
t
.
S
t
a
t
.
D
a
t
a
A
n
a
l
.
, v
o
l. 7
4
, p
p
. 1
8
1
-
197,
2014.
[5
2
]
H.
I
s
mk
h
a
n
,
“
I
-
k
-
me
a
n
s
−
+
:
An
i
t
e
r
a
t
i
v
e
c
l
u
s
t
e
r
i
n
g
a
l
g
o
r
i
t
h
m
b
a
s
e
d
o
n
a
n
e
n
h
a
n
c
e
d
v
e
r
s
i
o
n
o
f
t
h
e
k
-
me
a
n
s
,
”
Pa
t
t
e
r
n
Re
c
o
g
n
i
t
.
, 2
0
1
8
.
[5
3
]
S.
S.
Y
u
,
S.
W
.
C
hu,
C
.
M
.
W
a
ng,
Y
.
K
.
C
ha
n,
a
nd
T
.
C
.
C
ha
ng,
“
T
w
o
i
m
pr
ove
d
k
-
me
a
n
s
a
l
g
o
r
i
t
h
ms
,
”
Ap
p
l
.
S
o
f
t
Co
m
p
u
t
.
J
.
, 2
0
1
8
.
[5
4
]
Z.
K
a
k
u
s
h
a
d
z
e
a
n
d
W
.
Y
u
,
“
*
K
-
me
a
n
s
a
n
d
c
l
u
s
t
e
r
mo
d
e
l
s
f
o
r
c
a
n
c
e
r
s
i
g
n
a
t
u
r
e
s
,
”
Bi
o
m
o
l
.
D
e
t
e
c
t
.
Q
u
a
n
t
i
f
.
, 2
0
1
7
.
[5
5
]
M.
C
a
p
ó
,
A
.
P
é
r
e
z
,
a
n
d
J
.
A
.
L
o
z
a
n
o
,
“
A
n
e
f
f
i
c
i
e
n
t
a
p
p
r
o
x
i
m
a
t
i
o
n
t
o
t
h
e
K
-
me
a
n
s
c
l
u
s
t
e
r
i
n
g
f
o
r
ma
s
s
i
v
e
d
a
t
a
,
”
Kn
o
wl
e
d
g
e
-
Ba
s
e
d
S
y
s
t
.
, 2
0
1
7
.
[5
6
]
C.
X
i
o
n
g
,
Z
.
H
u
a
,
K
.
L
v
,
a
n
d
X
.
L
i
,
“
A
n
i
m
p
r
o
v
e
d
K
-
me
a
n
s
t
e
x
t
c
l
u
s
t
e
r
i
n
g
a
l
g
o
r
i
t
h
m
b
y
o
p
t
i
mi
z
i
n
g
i
n
i
t
i
a
l
c
l
u
s
t
e
r
cen
t
er
s,
”
i
n
Pr
o
c
e
e
d
i
n
g
s
-
2016
7t
h
I
nt
e
r
nat
i
onal
C
onf
e
r
e
nc
e
on
C
l
oud
C
om
put
i
ng
and
B
i
g
D
at
a,
C
C
B
D
2016
,
2017.
[5
7
]
R.
J
o
t
h
i
,
S
.
K
.
M
o
h
a
n
t
y
,
a
n
d
A
.
O
j
h
a
,
“
D
K
-
me
a
n
s
:
a
d
e
t
e
r
mi
n
i
s
t
i
c
K
-
me
a
n
s
c
l
u
s
t
e
r
i
n
g
a
l
g
o
r
i
t
h
m
f
o
r
g
e
n
e
ex
p
r
essi
o
n
an
al
y
si
s,
”
Pa
t
t
e
r
n
An
a
l
y
s
is
a
n
d
A
p
p
lic
a
tio
n
s
, 2
0
1
7
.
[5
8
]
S.
M
a
l
d
o
n
a
d
o
,
E.
C
a
r
r
i
z
o
s
a
,
a
n
d
R
.
W
e
b
e
r
,
“
K
e
r
n
e
l
Pe
n
a
l
i
z
e
d
K
-
me
a
n
s
:
A
f
e
a
t
u
r
e
s
e
l
e
c
t
i
o
n
me
t
h
o
d
b
a
s
e
d
o
n
Ke
r
n
e
l
K
-
me
a
n
s
,
”
In
f
.
S
c
i
.
(N
y
).
, 2
0
1
5
.
[5
9
]
X.
Hu
a
n
g
,
Y.
Ye
,
L
.
Xi
o
n
g
,
R
.
Y.
K.
L
a
u
,
N.
J
i
a
n
g
,
a
n
d
S
.
W
a
n
g
,
“
T
i
me
s
e
r
i
es
k
-
me
a
n
s
:
A
n
e
w
k
-
me
a
n
s
t
y
p
e
sm
o
o
t
h
su
b
sp
ace
cl
u
st
er
i
n
g
f
o
r
t
i
m
e
ser
i
es
d
at
a,
”
In
f
.
S
c
i
.
(N
y
).
, 2
0
1
6
.
[6
0
]
K.
M
.
Ku
ma
r
a
n
d
A.
R
.
M
.
R
e
d
d
y
,
“
An
e
f
f
i
c
i
e
n
t
k
-
me
a
n
s
c
l
u
s
t
e
r
i
n
g
f
i
l
t
e
r
i
n
g
a
l
g
o
r
i
t
h
m
u
s
i
n
g
d
e
n
s
i
t
y
b
a
s
e
d
i
n
i
t
i
a
l
cl
u
st
er
cen
t
er
s,
”
In
f
.
S
c
i
.
(N
y
).
,
2017.
[6
1
]
Y.
Ha
n
mi
n
,
L
.
Ha
o
,
a
n
d
S
.
Qi
a
n
t
i
n
g
,
“
An
i
mp
r
o
v
e
d
s
e
mi
-
su
p
er
v
i
sed
K
-
me
a
n
s
c
l
u
s
t
e
r
i
n
g
a
l
g
o
r
i
t
h
m,
”
i
n
Pr
o
c
e
e
d
i
n
g
s
o
f
2
0
1
6
I
EEE
I
n
f
o
r
m
a
t
i
o
n
Te
c
h
n
o
l
o
g
y
,
N
e
t
w
o
r
k
i
n
g
,
El
e
c
t
r
o
n
i
c
a
n
d
Au
t
o
m
a
t
i
o
n
C
o
n
t
r
o
l
C
o
n
f
e
r
e
n
c
e
,
IT
N
E
C
2
0
1
6
, 2
0
1
6
.
[6
2
]
Y.
L
i
u
,
H.
P
.
Yi
n
,
a
n
d
Y.
C
h
a
i
,
“
An
i
mp
r
o
v
e
d
k
e
r
n
e
l
k
-
me
a
n
s
c
l
u
s
t
e
r
i
n
g
a
l
g
o
r
i
t
h
m,
”
i
n
Le
c
t
u
r
e
N
o
t
e
s
i
n
El
e
c
t
r
i
c
a
l
En
g
i
n
e
e
r
i
n
g
, 2
0
1
6
.
[6
3
]
S.
K
a
n
t
a
n
d
I
.
A
.
A
n
s
a
r
i
,
“
A
n
i
m
p
r
o
v
e
d
K
m
e
a
n
s
c
l
u
s
t
e
r
i
n
g
w
i
t
h
A
t
k
i
n
s
o
n
i
n
d
e
x
t
o
c
l
a
s
s
i
f
y
l
i
v
e
r
p
a
t
i
e
n
t
d
a
t
a
s
e
t
,
”
In
t
.
J.
S
yst
.
A
ssu
r.
E
n
g
.
M
a
n
a
g
.
, 2
0
1
6
.
[6
4
]
Y.
Di
n
g
,
Y.
Z
h
a
o
,
X.
S
h
e
n
,
M
.
M
u
s
u
v
a
t
h
i
,
a
n
d
T
.
M
y
t
k
o
wi
c
z
,
“
Yi
n
y
a
n
g
K
-
me
a
n
s
:
A
d
r
o
p
-
in
r
e
p
la
c
e
m
e
n
t
o
f
th
e
cl
assi
c
K
-
me
a
n
s
wi
t
h
c
o
n
s
i
s
t
e
n
t
s
p
e
e
d
u
p
,
”
Ic
m
l
-
2015
, 2
0
1
5
.
[6
5
]
G.
T
z
o
r
t
z
i
s
a
n
d
A.
L
i
k
a
s
,
“
T
h
e
M
i
n
M
a
x
k
-
Me
a
n
s
cl
u
st
er
i
n
g
al
g
o
r
i
t
h
m
,
”
i
n
Pa
t
t
e
r
n
Re
c
o
g
n
i
t
i
o
n
, 2
0
1
4
.
[6
6
]
I.
M
e
l
n
y
k
o
v
a
n
d
V
.
M
e
l
n
y
k
o
v
,
“
O
n
K
-
me
a
n
s
a
l
g
o
r
i
t
h
m
wi
t
h
t
h
e
u
s
e
o
f
ma
h
a
l
a
n
o
b
i
s
d
i
s
t
a
n
c
e
s
,
”
St
at
.
P
r
obab.
L
e
t
t
.
,
2014.
[6
7
]
T.
J
.
Sa
n
t
n
e
r
,
B
.
J
.
W
i
l
l
i
a
m
s
,
a
n
d
W
.
I
.
N
o
t
z
,
Th
e
D
e
s
i
g
n
a
n
d
An
a
l
y
s
i
s
o
f
Co
m
p
u
t
e
r
E
x
p
e
r
i
m
e
n
t
s
,
1
s
t
e
d
.
S
p
r
in
g
e
r
-
Ve
r
l
a
g
Ne
w
Yo
r
k
,
2
0
0
3
.
[6
8
]
Y.
B
.
L
i
m,
J
.
S
a
c
k
s
,
W
.
J
.
S
t
u
d
d
e
n
,
a
n
d
W
.
J
.
W
e
l
c
h
,
“
De
s
i
g
n
a
n
d
a
n
a
l
y
s
i
s
o
f
c
o
mp
u
t
e
r
e
x
p
e
r
i
me
n
t
s
wh
e
n
t
h
e
out
put
i
s
hi
ghl
y
c
or
r
e
l
a
t
e
d
ove
r
t
he
i
nput
s
pa
c
e
,
”
Ca
n
.
J
.
S
t
a
t
.
, v
o
l. 3
0
, n
o
. 1
,
pp.
109
-
126,
2002.
[6
9
]
H.
De
t
t
e
a
n
d
A.
P
e
p
e
l
y
s
h
e
v
,
“
Ge
n
e
r
a
l
i
z
e
d
l
a
t
i
n
h
y
p
e
r
c
u
b
e
d
e
s
i
g
n
f
o
r
c
o
mp
u
t
e
r
e
x
p
e
r
i
me
n
t
s
,
”
Te
c
h
n
o
m
e
t
r
i
c
s
,
v
o
l.
52,
no.
4,
pp.
421
-
429,
2010.
[7
0
]
J.
H
.
F
r
i
ed
m
an
,
M
.
A
d
ap
t
i
v
e,
an
d
R
.
S
p
l
i
n
es,
“M
u
l
t
i
v
ar
i
at
e
A
d
ap
t
i
v
e
R
eg
r
essi
o
n
S
p
l
i
n
es
,”
An
n
.
S
t
a
t
.
,
v
o
l.
1
9
,
n
o
.
1
,
pp.
1
-
67,
1991.
[7
1
]
L.
E.
A
i
k
a
n
d
Z.
Za
i
n
u
d
d
i
n
,
“
A
n
I
m
p
r
o
v
e
d
Fa
s
t
Tr
a
i
n
i
n
g
A
l
g
o
r
i
t
h
m
f
o
r
R
B
F
N
e
t
w
o
r
k
s
U
s
i
n
g
Sy
m
m
e
t
r
y
-
Ba
s
e
d
Fu
z
z
y
C
-
Me
a
n
s
C
l
u
s
t
e
r
i
n
g
O
v
e
r
v
i
e
w
o
f
F
u
z
z
y
C
-
Me
a
n
s
C
l
u
s
t
e
r
i
n
g
Me
t
h
o
d
,
”
MA
T
E
MA
T
I
K
A
,
v
o
l.
2
4
,
n
o
.
2
,
p
p
.
141
-
148,
2008.
[7
2
]
L.
X
M
G
l
o
b
a
l
,
“
X
M
M
e
t
a
t
r
a
d
e
r
4
,
”
2
0
1
8
.
[
O
n
l
i
n
e
]
.
A
v
a
i
l
a
b
l
e
:
h
t
t
p
s
:
/
/
w
w
w
.
x
m
.
c
o
m
/
m
t
4
.
[
A
c
c
e
s
s
e
d
:
2
0
-
Ju
l
-
2018]
.
Evaluation Warning : The document was created with Spire.PDF for Python.