I
A
E
S
I
n
t
e
r
n
at
io
n
al
Jou
r
n
al
of
A
r
t
if
ic
ia
l
I
n
t
e
ll
ig
e
n
c
e
(
I
J
-
AI
)
V
ol
.
14
, N
o.
4
,
A
ugus
t
2025
, pp.
2909
~
2921
I
S
S
N
:
2252
-
8938
,
D
O
I
:
10.11591/
ij
a
i.
v
14
.i
4
.pp
2909
-
2921
2909
Jou
r
n
al
h
om
e
page
:
ht
tp
:
//
ij
ai
.
ia
e
s
c
or
e
.c
om
D
at
a
-
d
r
i
ve
n
su
p
p
or
t
ve
c
t
o
r
r
e
g
r
e
ss
i
o
n
-
ge
n
e
t
i
c
al
gor
i
t
h
m
m
od
e
l
f
or
p
r
e
d
i
c
t
i
n
g t
h
e
d
i
p
h
t
h
e
r
i
a d
i
st
r
i
b
u
t
i
on
Wi
w
ik
A
n
ggr
ae
n
i
1
, Y
e
ye
n
S
u
d
ia
r
t
i
1
, M
u
h
am
m
ad
I
lh
am
P
e
r
d
an
a
2
, E
d
w
in
R
ik
s
ak
om
ar
a
1
,
A
d
r
i
G
ab
r
ie
l
S
ooai
3
1
D
e
pa
r
t
m
e
nt
of
I
nf
or
m
a
t
i
on
S
ys
t
e
m
s
, F
a
c
ul
t
y of
I
nt
e
l
l
i
ge
nt
E
l
e
c
t
r
i
c
a
l
a
nd I
n
f
or
m
a
t
i
c
s
T
e
c
hnol
ogy, I
ns
t
i
t
ut
T
e
knol
ogi
S
e
pul
uh
N
ope
m
be
r
, S
ur
a
ba
ya
, I
ndone
s
i
a
2
D
e
pa
r
t
m
e
nt
of
I
nf
or
m
a
t
i
c
s
, F
a
c
ul
t
y of
E
ngi
ne
e
r
i
ng, U
ni
ve
r
s
i
t
a
s
M
uha
m
m
a
di
y
a
h M
a
l
a
ng, M
a
l
a
ng, I
ndo
ne
s
i
a
3
D
e
pa
r
t
m
e
n
t
of
C
om
put
e
r
S
c
i
e
nc
e
, F
a
c
ul
t
y of
E
ngi
ne
e
r
i
ng, U
ni
ve
r
s
i
t
a
s
K
a
t
ol
i
k W
i
dya
M
a
ndi
r
a
, K
upa
ng, I
ndone
s
i
a
A
r
t
ic
le
I
n
f
o
A
B
S
T
R
A
C
T
A
r
ti
c
le
h
is
to
r
y
:
R
e
c
e
iv
e
d
F
e
b
9
,
2024
R
e
vi
s
e
d
F
e
b
20
,
2025
A
c
c
e
pt
e
d
M
a
r
15
,
2025
Indonesia
is
one
of
the
countries
with
the
largest
number
of
dip
htheria
sufferers
in the world. Diphther
ia is
a case of
re
-
emerging
disease,
esp
ecially
in
Indonesia.
Diphtheria
c
an
be
prev
ented
by
immunization.
Dip
htheria
immunization
has
drastically
reduce
d
mortality
and
susceptibility
to
diphtheria,
but
it
is
still
a
significant
childhood
health
problem.
This
study
predicted
the
number
of
diphtheria
patients
in
several
regions
using
s
upport
vector
regression
(
SVR
)
combined
with
the
genetic
algorit
hm
(G
A)
for
parameter
optimization.
The
area
is
grouped
into
3
clusters
based
on
the
number
of
cases.
The
proposed
m
ethod
is
proven
to
overcome
ove
rfitting
and
avoid
local
optima
.
Model
robustness
tests
were
carried
out
in
several
other
regions
in
each
cluster.
Based
on
the
experiments
in
three
sc
enarios
and
12
areas,
the
hybrid
model
shows
good
forecasting
results
w
ith
an
average
mean
squared
error
(
MSE
)
of
0.036
and
a
sy
mmetric
mean
a
bsolute
percentage
error
(SMAPE)
of
41.2%
with
a
standard
deviation
of
0.0
75
an
d
0.442,
respectively.
Based on exp
eriments
in various
scenarios,
the S
VR
-
GA
model
shows
better
per
formance
than
others.
Compares
two
-
means
t
ests
on
MSE
and
SMAPE
were
given
to
prove
that
SVR
-
GA
models
have
better
performance. The
results of this forecasting can be use
d as a basis for policy
-
making to minimize
the sprea
d of diphther
ia cases
.
K
e
y
w
o
r
d
s
:
D
ip
ht
he
r
ia
D
is
e
a
s
e
F
or
e
c
a
s
ti
ng
G
e
ne
ti
c
a
lg
or
it
hm
S
uppor
t
ve
c
to
r
r
e
gr
e
s
s
io
n
This is an
open
acce
ss artic
le unde
r the
CC BY
-
SA
license.
C
or
r
e
s
pon
di
n
g A
u
th
or
:
W
iwi
k A
nggr
a
e
ni
D
e
pa
r
tm
e
nt
of
I
nf
or
m
a
ti
on
S
ys
te
m
s
, F
a
c
ul
ty
of
I
nt
e
ll
ig
e
nt
E
le
c
tr
ic
a
l
a
nd I
nf
or
m
a
ti
c
s
T
e
c
hnol
ogy
I
ns
ti
tu
t
T
e
knol
ogi
S
e
pul
uh N
ope
m
be
r
K
e
put
ih
S
ukol
il
o, S
ur
a
ba
ya
,
I
ndone
s
ia
E
m
a
il
:
w
iwi
k
@
is
.i
ts
.a
c
.i
d
1.
I
N
T
R
O
D
U
C
T
I
O
N
D
ip
ht
he
r
ia
is
a
da
nge
r
ous
in
f
e
c
ti
ous
di
s
e
a
s
e
c
a
us
e
d
by
th
e
ba
c
te
r
iu
m
C
or
y
ne
bac
t
e
r
iu
m
di
pht
he
r
ia
e
[
1]
.
D
ip
ht
he
r
ia
s
pr
e
a
ds
th
r
ough
a
n
in
f
e
c
te
d
pe
r
s
on'
s
s
a
li
va
or
c
ont
a
c
t
w
it
h
a
n
in
f
e
c
te
d
pe
r
s
on'
s
s
ki
n
or
pe
r
s
ona
l
e
f
f
e
c
ts
[
2]
.
T
he
di
s
e
a
s
e
c
a
n
le
a
d
to
d
e
a
th
,
due
to
a
bl
oc
ke
d
r
e
s
pi
r
a
to
r
y
tr
a
c
t
a
nd
he
a
r
t
f
a
il
ur
e
,
e
s
pe
c
ia
ll
y
in
c
hi
ld
r
e
n.
T
he
m
or
ta
li
ty
r
a
te
of
di
pht
he
r
ia
a
ve
r
a
ge
s
5
–
10%
in
c
hi
ld
r
e
n
unde
r
5
ye
a
r
s
of
a
ge
[
2]
, [
3
]
. A
lt
hough
di
pht
he
r
ia
i
s
r
a
r
e
i
n c
ount
r
ie
s
w
it
h hi
gh
va
c
c
in
a
ti
on c
ove
r
a
ge
, i
t
r
e
m
a
in
s
a
c
onc
e
r
n i
n
a
r
e
a
s
w
it
h i
na
de
qua
te
va
c
c
in
a
ti
on a
nd he
a
lt
hc
a
r
e
i
nf
r
a
s
tr
uc
tu
r
e
[
4]
.
D
ip
ht
he
r
ia
i
s
a
va
c
c
in
e
-
pr
e
ve
nt
a
bl
e
di
s
e
a
s
e
[
5]
.
I
m
m
uni
z
a
ti
on
a
ga
in
s
t
di
pht
he
r
ia
ha
s
dr
a
s
ti
c
a
ll
y
r
e
duc
e
d
m
or
ta
li
ty
a
nd
s
us
c
e
pt
ib
il
it
y
to
di
pht
he
r
ia
[
6]
.
G
lo
ba
l
da
ta
publ
is
he
d
by
W
H
O
a
nd
U
N
I
C
E
F
in
2022
s
how
th
a
t
th
e
r
e
ha
s
be
e
n
a
s
us
ta
in
e
d
de
c
li
ne
in
th
e
num
be
r
o
f
c
hi
ld
hood
va
c
c
in
a
ti
ons
in
a
bout
30
y
e
a
r
s
[
7]
.
I
n
2021
a
lo
ne
,
25
m
il
li
on
c
hi
ld
r
e
n
m
is
s
e
d
on
e
or
m
or
e
dos
e
s
of
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
, V
ol
.
14
, N
o.
4
,
A
ugus
t
20
25
:
2909
-
2921
2910
th
e
di
pht
he
r
ia
va
c
c
in
e
.
18
m
il
li
on
of
th
os
e
25
m
il
li
on
c
hi
ld
r
e
n
di
d
not
r
e
c
e
iv
e
a
s
in
gl
e
dos
e
of
di
pht
he
r
ia
,
te
ta
nus
,
a
nd
pe
r
tu
s
s
i
s
(
D
T
P
)
ye
a
r
-
r
ound,
a
nd
m
os
t
of
th
e
m
li
v
e
d
in
I
ndi
a
,
N
ig
e
r
ia
,
I
ndone
s
ia
,
E
th
io
pi
a
,
a
nd
P
hi
li
ppi
ne
s
[
7]
.
I
ndone
s
ia
ha
s
f
a
c
e
d
th
e
c
ha
ll
e
nge
of
di
pht
he
r
ia
out
br
e
a
ks
pe
r
io
di
c
a
ll
y
unt
il
2022,
e
s
pe
c
ia
ll
y
in
de
ns
e
ly
popula
te
d a
r
e
a
s
. I
ndone
s
ia
w
a
s
onc
e
t
he
c
ount
r
y w
it
h t
he
s
e
c
ond
-
la
r
ge
s
t
di
pht
he
r
ia
c
a
s
e
s
i
n t
he
w
or
ld
a
f
te
r
I
ndi
a
,
w
hi
c
h
w
a
s
3
,
203
c
a
s
e
s
[
8]
.
T
hi
s
is
e
ve
n
m
or
e
da
ng
e
r
ous
be
c
a
u
s
e
of
th
e
de
c
li
ne
in
im
m
uni
z
a
ti
on
c
ove
r
a
ge
in
I
ndone
s
ia
due
to
th
e
im
pa
c
t
of
C
O
V
I
D
-
19
.
W
H
O
I
ndone
s
ia
S
it
ua
ti
on
R
e
por
t
-
13
s
ta
te
s
th
a
t
th
e
de
c
r
e
a
s
e
oc
c
ur
r
e
d
by
10
-
40%
in
M
a
r
c
h
-
A
pr
il
2020
c
om
p
a
r
e
d
to
th
e
pr
e
vi
ou
s
ye
a
r
[
2]
.
B
a
s
e
d
on
th
e
s
e
c
ondi
ti
ons
,
th
e
I
ndone
s
ia
n
M
in
is
tr
y
of
H
e
a
lt
h
ne
e
ds
to
pr
e
pa
r
e
a
s
tr
a
te
gy
to
d
e
a
l
w
it
h
th
e
in
c
r
e
a
s
e
in
th
e
num
be
r
of
c
a
s
e
s
.
T
he
s
tr
a
te
gy
w
il
l
b
e
ti
m
e
ly
a
nd
ta
r
ge
te
d
i
f
it
is
ba
s
e
d
on
in
f
or
m
a
ti
on
on
th
e
po
s
s
ib
le
di
s
tr
ib
ut
io
n
of
th
e
num
be
r
of
c
a
s
e
s
a
nd
va
c
c
in
e
s
in
e
a
c
h
r
e
gi
on.
F
or
th
is
r
e
a
s
on,
th
is
s
tu
dy
is
c
a
r
r
ie
d
out
to
f
or
e
c
a
s
t
th
e
num
be
r
of
di
pht
he
r
ia
s
uf
f
e
r
e
r
s
in
s
e
v
e
r
a
l
r
e
gi
ons
in
I
ndone
s
ia
by
pa
yi
ng
a
tt
e
nt
io
n
to
v
a
r
io
us
va
c
c
in
e
s
a
s
s
oc
ia
te
d
w
it
h
di
pht
he
r
ia
.
T
he
r
e
s
ul
ts
of
f
or
e
c
a
s
ti
ng
th
e
num
be
r
of
di
pht
he
r
ia
c
a
s
e
s
r
e
s
ul
ti
ng
f
r
om
th
is
s
tu
dy
c
a
n
b
e
us
e
d
to
a
s
s
is
t
th
e
he
a
lt
h
of
f
ic
e
in
de
te
r
m
in
in
g
pol
ic
ie
s
th
a
t
w
il
l
be
c
a
r
r
ie
d
out
to
r
e
duc
e
th
e
num
be
r
of
di
ph
th
e
r
ia
s
uf
f
e
r
e
r
s
.
F
o
r
e
c
a
s
ti
ng
r
e
la
te
d
to
th
e
he
a
lt
h
dom
a
in
is
ne
e
de
d,
e
s
pe
c
ia
ll
y
if
th
e
f
or
e
c
a
s
ti
ng i
nvol
ve
s
m
ul
ti
va
r
ia
bl
e
[
9]
.
B
a
s
e
d
on
our
be
s
t
knowle
dge
,
c
ur
r
e
nt
ly
th
e
r
e
is
s
ti
ll
a
la
c
k
of
r
e
s
e
a
r
c
h
r
e
la
te
d
to
di
pht
he
r
ia
di
r
e
c
te
d
to
f
or
e
c
a
s
ti
ng.
R
a
th
e
r
,
pr
e
vi
ous
s
tu
di
e
s
te
nd
to
be
m
or
e
in
te
r
e
s
te
d
in
th
e
f
a
c
to
r
a
na
ly
s
e
s
in
f
lu
e
nt
ia
l
to
th
e
r
is
in
g
of
di
pht
he
r
ia
c
a
s
e
s
[
3]
,
[
4]
,
a
na
ly
s
is
of
th
e
e
f
f
e
c
ti
vi
ty
of
va
c
c
in
a
ti
on
to
th
e
p
e
opl
e
’
s
he
a
lt
h
s
pe
c
if
ie
d
by
a
ge
[
10]
‒
[
12]
,
r
e
gi
ona
l
de
m
ogr
a
phy
[
13]
,
[
14]
,
c
ount
r
ie
s
’
in
c
o
m
e
[
1
5
]
,
a
nd
popula
ti
on
[
16]
.
T
hi
s
s
tu
dy
tr
ie
s
to
f
or
e
c
a
s
t
th
e
s
pr
e
a
d
of
th
e
di
pht
he
r
ia
c
a
s
e
num
be
r
in
vol
vi
ng othe
r
va
r
ia
bl
e
s
r
e
la
te
d
to
va
c
c
in
a
ti
on
th
a
t
ha
ve
be
e
n
gi
ve
n
pr
e
vi
ous
ly
.
T
he
num
be
r
of
va
c
c
in
e
s
gi
ve
n
is
in
vol
ve
d
a
s
a
r
e
gr
e
s
s
or
be
c
a
us
e
va
c
c
in
a
ti
on
is
ve
r
y
c
r
uc
ia
l
to
th
e
di
pht
he
r
ia
c
a
s
e
num
be
r
f
lu
c
tu
a
ti
on
[
10]
,
[
11]
.
A
di
pht
he
r
ia
c
a
s
e
f
or
e
c
a
s
ti
ng
s
tu
dy
ha
s
be
e
n
done
by
[
16]
.
H
ow
e
ve
r
,
A
nggr
a
e
ni
e
t
al
.
[
16]
onl
y
doe
s
s
o
by
us
in
g
ju
s
t
a
s
in
gl
e
va
r
ia
bl
e
,
w
hi
c
h
is
th
e
di
pht
he
r
ia
to
ta
l
c
a
s
e
s
.
I
n
a
ddi
ti
on,
A
nggr
a
e
ni
e
t
al
.
[
17]
ha
s
a
ls
o
done
a
f
or
e
c
a
s
ti
ng
s
tu
dy
in
vol
vi
ng
m
or
e
in
f
lu
e
nt
ia
l
va
r
ia
bl
e
s
, but
t
he
i
m
pl
e
m
e
nt
e
d m
ode
l
s
ti
ll
ha
s
s
e
r
io
us
l
im
it
a
ti
ons
.
O
n
ly
ve
r
y
f
e
w
s
t
u
di
e
s
ha
ve
be
e
n
f
o
un
d
t
o
ha
ve
d
o
ne
a
f
o
r
e
c
a
s
t
i
ng
a
na
ly
s
is
on
th
e
c
a
s
e
s
tu
d
y
of
d
i
p
ht
he
r
ia
.
T
h
is
m
a
ke
s
t
he
s
t
a
t
e
-
of
-
th
e
-
a
r
t
s
t
u
dy
a
b
ou
t
t
hi
s
m
e
th
o
d
v
e
r
y
li
m
i
te
d.
A
p
r
e
v
i
ous
s
t
ud
y
m
e
nt
i
on
e
d
e
a
r
li
e
r
d
oe
s
a
f
or
e
c
a
s
t
i
ng
of
th
e
n
u
m
b
e
r
o
f
d
ip
ht
h
e
r
i
a
c
a
s
e
s
us
in
g
th
e
r
a
di
a
l
b
a
s
i
s
f
u
nc
t
io
n
(
R
B
F
)
ne
t
w
o
r
k
a
p
pr
oa
c
h
[
1
6
]
.
H
ow
e
ve
r
,
th
e
s
t
ud
y
in
vo
l
ve
s
o
n
l
y
t
he
n
u
m
b
e
r
o
f
c
a
s
e
s
in
t
he
pa
s
t
w
i
th
ou
t
c
o
ns
i
de
r
i
ng
o
th
e
r
in
f
lu
e
n
t
ia
l
f
a
c
t
o
r
s
.
I
n
a
dd
i
ti
o
n,
t
he
pe
r
f
o
r
m
a
nc
e
o
f
t
he
m
od
e
l
i
s
a
b
it
l
a
c
ki
n
g.
T
he
s
tu
dy
w
o
ul
d
be
de
ve
l
op
e
d
f
u
r
th
e
r
b
y
in
v
ol
v
in
g
a
n
ot
he
r
i
n
f
l
ue
n
ti
a
l
v
a
r
i
a
b
le
,
w
h
ic
h
is
t
he
nu
m
be
r
o
f
v
a
c
c
i
na
t
io
ns
[
1
7
]
,
by
i
m
pl
e
m
e
nt
i
ng
t
he
f
uz
z
y
a
p
p
r
oa
c
h.
T
he
da
t
a
in
vol
ve
d
in
th
i
s
r
e
s
e
a
r
c
h
is
not
s
uf
f
ic
ie
nt
in
qu
a
nt
it
y, s
o
th
e
c
h
a
nc
e
of
ove
r
f
it
ti
ng
oc
c
ur
r
in
g
is
hi
gh
be
c
a
u
s
e
th
e
m
ode
l
c
a
nnot
le
a
r
n
f
r
om
e
nough
da
ta
.
F
or
e
c
a
s
ti
ng
w
it
h
a
s
m
a
ll
da
ta
s
e
t
in
tr
a
in
in
g
is
c
ha
ll
e
ngi
ng
[
9]
.
K
now
in
g
th
is
c
ondi
ti
on,
th
is
s
tu
dy
im
pl
e
m
e
nt
s
th
e
s
uppor
t
ve
c
to
r
r
e
gr
e
s
s
io
n
(
S
V
R
)
m
e
th
od
to
do a
f
or
e
c
a
s
ti
ng on the
numbe
r
of
di
pht
he
r
ia
c
a
s
e
s
. T
he
s
upp
or
t
ve
c
to
r
m
a
c
hi
ne
(
S
V
M
)
ha
s
be
e
n pr
ove
n t
o
s
ol
ve
th
e
pr
obl
e
m
of
ove
r
f
it
ti
ng
,
s
o
it
is
s
ui
ta
bl
e
to
be
im
pl
e
m
e
nt
e
d
to
pr
oduc
e
lo
ng
-
te
r
m
f
or
e
c
a
s
ti
ng
w
it
h
a
r
e
la
ti
ve
ly
s
m
a
ll
d
a
ta
s
e
t
[
18]
.
T
he
S
V
R
is
a
ls
o
m
or
e
c
ons
i
s
te
nt
in
f
or
e
c
a
s
ti
ng
a
c
a
s
e
th
a
n
ot
he
r
r
e
gr
e
s
s
io
n
a
ppr
oa
c
he
s
[
19]
a
nd
i
s
hybr
id
iz
e
d
w
it
h
ot
he
r
s
[
20]
.
I
n
a
ddi
ti
on,
th
e
S
V
R
ha
s
a
be
tt
e
r
g
e
ne
r
a
li
z
in
g
a
bi
li
ty
,
w
hi
c
h
is
s
ui
ta
bl
e
f
or
s
m
a
ll
da
ta
s
e
t
a
na
ly
s
e
s
w
it
h
non
-
li
ne
a
r
da
ta
[
21]
,
[
22
]
.
I
n
s
pi
te
of
it
s
s
upe
r
io
r
f
e
a
tu
r
e
s
,
th
e
pa
r
a
m
e
te
r
va
lu
e
of
th
e
S
V
R
m
us
t
be
de
f
in
e
d
w
it
h
pr
e
c
is
e
a
c
c
ur
a
c
y
to
c
ons
tr
uc
t
a
n
opt
im
iz
e
d
S
V
R
m
od
e
l
[
22]
.
U
ns
ui
ta
bl
e
pa
r
a
m
e
te
r
va
lu
e
in
th
e
S
V
R
a
na
ly
s
is
w
oul
d,
ne
e
dl
e
s
s
to
s
a
y,
in
f
lu
e
nc
e
it
s
f
or
e
c
a
s
ti
ng
pe
r
f
or
m
a
nc
e
[
18]
.
P
a
r
a
m
e
te
r
de
f
in
it
io
n
ba
s
e
d
on
tr
ia
l
a
nd e
r
r
or
c
oul
d
c
a
us
e
s
e
r
io
us
pr
obl
e
m
s
a
nd
ne
e
ds
m
or
e
ti
m
e
t
o a
na
ly
z
e
[
18]
.
T
h
is
r
e
s
e
a
r
c
h
hy
br
i
di
z
e
s
t
h
e
S
V
R
m
e
th
o
d w
it
h t
h
e
ge
ne
ti
c
a
lg
or
i
th
m
(
G
A
)
.
T
h
e
G
A
i
s
i
m
p
le
m
e
nt
e
d
to
de
f
i
ne
th
e
pa
r
a
m
e
t
e
r
va
lu
e
of
t
he
S
V
R
.
T
h
e
f
i
tn
e
s
s
v
a
lu
e
of
t
he
G
A
i
s
c
a
pa
bl
e
o
f
a
vo
id
i
ng
th
e
l
oc
a
l
op
ti
m
um
a
nd
d
e
f
in
in
g
th
e
g
lo
b
a
l
opt
im
u
m
in
a
s
h
or
t
ti
m
e
[
2
3]
.
T
he
GA
,
a
s
on
e
o
f
th
e
m
a
n
y
ki
nd
s
of
m
e
t
a
h
e
ur
i
s
ti
c
a
lg
or
it
hm
s
,
h
a
s
b
e
e
n
pr
o
ve
n
c
a
p
a
b
le
of
opt
im
i
z
in
g
th
e
p
e
r
f
or
m
a
nc
e
of
a
h
e
u
r
i
s
ti
c
op
ti
m
iz
a
ti
on
a
lg
or
it
hm
[
2
0]
.
T
h
e
S
V
R
-
G
A
m
e
t
hod
h
a
s
b
e
e
n
a
pp
li
e
d t
o
f
o
r
e
c
a
s
t
t
our
i
s
m
d
e
m
a
nd i
n
C
hi
na
[
22]
.
T
h
e
s
t
ud
y f
i
nd
s
t
ha
t
t
he
S
V
R
-
G
A
o
ut
p
e
r
f
or
m
s
b
ot
h
t
he
ba
c
k
pr
o
pa
ga
ti
o
n
n
e
ur
a
l
n
e
t
w
or
k
(
B
P
N
N
)
a
nd
a
u
to
r
e
gr
e
s
s
iv
e
in
te
gr
a
te
d
m
o
vi
n
g
a
v
e
r
a
ge
(
A
R
I
M
A
)
m
od
e
l
s
a
c
c
o
r
di
ng
t
o
n
or
m
a
li
z
e
d
m
e
a
n
s
qu
a
r
e
e
r
r
or
(
M
S
E
)
a
n
d
m
e
a
n
a
bs
ol
ut
e
p
e
r
c
e
n
ta
ge
e
r
r
or
(
M
A
P
E
)
[
22]
.
T
hi
s
S
V
R
-
G
A
m
e
t
hod
h
a
s
a
ls
o
b
e
e
n
a
p
p
li
e
d
to
m
a
k
e
a
f
or
e
c
a
s
t
on
th
e
q
ua
li
t
y
of
th
e
a
qu
a
c
ul
tu
r
e
w
a
t
e
r
[
24]
.
T
h
e
r
e
s
e
a
r
c
h
f
i
nd
s
th
a
t
t
he
S
V
R
-
G
A
o
ut
p
e
r
f
or
m
s
th
e
tr
a
di
ti
on
a
l
S
V
R
a
nd
th
e
B
P
NN
m
od
e
l
[
24]
. T
h
e
hy
br
id
iz
a
t
io
n
of
th
e
G
A
a
nd
S
V
R
i
n
th
e
o
pt
i
m
i
z
a
ti
o
n
o
f
p
a
r
a
m
e
t
e
r
v
a
l
ue
i
s
a
n
e
f
f
e
c
ti
ve
w
a
y
to
in
c
r
e
a
s
e
th
e
a
c
c
ur
a
c
y
o
f
t
he
pr
e
d
ic
ti
o
n
a
n
d
t
he
a
bi
l
it
y
t
o
g
e
ne
r
a
li
z
e
a
m
od
e
l
[
20
]
.
I
n
th
i
s
s
tu
dy,
t
he
S
V
R
-
GA
m
e
t
hod
i
s
u
s
e
d
to
m
od
e
l
a
n
d
f
o
r
e
c
a
s
t
th
e
num
be
r
of
in
f
e
c
ti
o
n
s
by
d
ip
h
th
e
r
i
a
w
it
h
s
om
e
i
nd
e
p
e
n
de
nt
v
a
r
i
a
bl
e
s
,
na
m
e
l
y
t
he
im
m
un
iz
a
ti
on
r
a
ng
e
a
nd
po
pu
la
t
io
n
d
e
n
s
i
ty
.
W
ha
t
th
i
s
r
e
s
e
a
r
c
h
w
a
nt
s
to
c
ont
r
ib
u
te
is
l
i
s
te
d
a
s
f
ol
lo
w
s
:
i)
t
hi
s
s
tu
dy
pr
opos
e
s
a
hybr
id
m
ode
l
c
om
bi
ni
ng
th
e
S
V
R
a
nd
G
A
m
e
th
ods
to
pr
oduc
e
a
m
or
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
D
at
a
-
dr
iv
e
n s
uppor
t
v
e
c
to
r
r
e
gr
e
s
s
io
n
-
ge
ne
ti
c
al
gor
it
hm
m
ode
l
f
or
pr
e
di
c
ti
ng
…
(
W
iw
ik
A
nggr
ae
ni
)
2911
opt
im
a
l
f
or
e
c
a
s
ti
ng
m
ode
l
w
it
hout
a
tr
ia
l
-
a
nd
-
e
r
r
or
e
xpe
r
im
e
n
t
in
f
in
di
ng
a
pa
r
a
m
e
te
r
th
a
t
w
oul
d
pr
oduc
e
a
f
or
e
c
a
s
ti
ng
m
ode
l
w
it
h
f
e
w
e
r
e
r
r
or
s
.
T
he
f
o
r
e
c
a
s
ti
ng
is
de
duc
e
d
by
in
vol
vi
ng
s
om
e
va
r
ia
bl
e
s
in
f
lu
e
nt
ia
l
on
th
e
numbe
r
of
s
uf
f
e
r
e
r
s
.
ii
)
T
he
m
ode
l
th
a
t
w
e
pr
opos
e
i
s
s
ui
ta
b
le
f
or
s
pe
c
ul
a
ti
ng da
ta
w
it
h m
or
e
t
r
a
in
in
g da
ta
but
f
e
w
e
r
ir
r
e
gul
a
r
pa
tt
e
r
ns
. T
hi
s
m
ode
l
c
a
n s
uc
c
e
s
s
f
ul
ly
t
a
c
kl
e
a
n i
nc
r
e
a
s
e
i
n da
ta
l
oc
a
te
d outs
id
e
i
ts
i
nt
e
r
va
l
a
nd
hi
s
to
r
y.
ii
i)
T
hi
s
s
tu
dy
is
e
qui
ppe
d
w
it
h
a
de
e
pe
r
hypot
he
ti
c
a
l
te
s
ti
ng
to
pr
ove
th
a
t
th
e
im
p
le
m
e
nt
e
d
hybr
id
a
ppr
oa
c
h ha
s
a
l
ot
be
tt
e
r
pe
r
f
or
m
a
nc
e
t
ha
n t
he
ot
he
r
s
.
T
he
r
e
s
t
of
th
is
pa
pe
r
i
s
a
s
f
ol
lo
w
s
.
T
he
f
ir
s
t
s
e
c
ti
on
of
th
e
in
tr
oduc
ti
on
c
ont
a
in
s
th
e
c
ondi
ti
ons
a
nd
pr
obl
e
m
s
th
a
t
oc
c
ur
,
r
e
la
te
d
w
or
ks
th
a
t
h
a
ve
e
xi
s
t
e
d
be
f
or
e
,
ga
ps
,
a
nd
c
ont
r
ib
ut
io
ns
of
f
e
r
e
d.
I
n
th
e
ne
xt
s
e
c
ti
on,
th
e
r
e
is
a
m
e
th
od
c
ons
is
ti
ng
of
a
r
e
s
e
a
r
c
h
lo
c
a
ti
on,
a
da
ta
s
e
t,
a
nd
a
m
e
th
odol
ogy.
N
e
xt
is
th
e
r
e
s
ul
t
a
nd
di
s
c
us
s
io
n.
F
in
a
ll
y,
a
c
onc
lu
s
io
n
w
hi
c
h
i
s
a
ls
o
c
om
pl
e
m
e
nt
e
d
by
th
e
de
ve
lo
pm
e
nt
pl
a
n
of
th
is
pa
pe
r
pr
e
s
e
nt
e
d a
t
th
e
e
nd.
2.
M
E
T
H
O
D
2.1
.
R
e
s
e
ar
c
h
l
oc
at
io
n
an
d
d
at
as
e
t
T
he
dom
a
in
th
is
r
e
s
e
a
r
c
h
w
a
s
c
a
r
r
ie
d
out
i
s
s
om
e
r
e
gi
ons
or
c
it
ie
s
in
th
e
E
a
s
t
J
a
va
P
r
ovi
nc
e
,
I
ndone
s
ia
.
E
a
s
t
J
a
v
a
is
a
pr
ovi
nc
e
w
h
e
r
e
di
pht
he
r
ia
c
a
s
e
s
a
r
e
f
ound
to
be
m
os
t
f
r
e
que
nt
in
I
ndone
s
ia
,
w
it
h
a
pe
r
c
e
nt
a
ge
of
74%
[
7
]
.
T
hi
s
is
be
c
a
us
e
di
pht
he
r
ia
is
a
r
e
-
e
m
e
r
gi
ng
c
a
s
e
in
th
is
pr
ovi
nc
e
.
T
he
hi
ghe
s
t
r
e
c
or
de
d
num
be
r
of
c
a
s
e
s
ha
pp
e
ne
d
in
2012
,
w
it
h
a
to
ta
l
of
955
c
a
s
e
s
a
nd
37
de
a
th
s
.
B
ut
it
s
e
e
m
s
th
a
t
f
r
om
ti
m
e
to
ti
m
e
,
r
e
por
te
d
c
a
s
e
s
of
di
pht
he
r
ia
ke
e
p
in
c
r
e
a
s
in
g,
a
s
w
e
ll
a
s
th
os
e
w
ho
a
r
e
r
e
por
te
d
a
s
de
a
d
be
c
a
us
e
of
th
is
il
ln
e
s
s
.
E
a
s
t
J
a
va
ha
s
be
e
n
pr
onounc
e
d
m
ul
ti
pl
e
ti
m
e
s
a
s
a
pr
ovi
nc
e
w
it
h
a
n
out
br
e
a
k
of
di
pht
he
r
ia
c
a
s
e
s
.
M
e
a
nw
hi
le
,
th
e
da
ta
pr
oc
e
s
s
e
d
in
th
is
r
e
s
e
a
r
c
h
a
r
e
t
hos
e
c
onc
e
r
ni
ng
th
os
e
w
ho
a
r
e
in
f
e
c
te
d
by
di
pht
he
r
ia
in
s
om
e
r
e
gi
ons
/
c
it
ie
s
in
E
a
s
t
J
a
va
w
it
h
a
m
ont
hl
y
t
im
e
hor
iz
on.
I
n
a
ddi
ti
on,
th
e
r
e
a
r
e
s
om
e
d
a
ta
in
vol
ve
d
be
c
a
us
e
th
e
y
a
r
e
in
f
lu
e
nt
ia
l
on
th
e
num
be
r
o
f
di
pht
he
r
ia
c
a
s
e
s
,
na
m
e
ly
th
e
im
m
uni
z
a
ti
on
r
a
nge
(
D
P
T
1, D
P
T
2, D
P
T
3, a
nd D
P
T
4 va
c
c
in
e
s
)
a
s
w
e
ll
a
s
t
he
popul
a
ti
on de
ns
it
y.
2.2
.
M
e
t
h
od
ol
ogy
T
he
s
te
ps
unde
r
ta
ke
n
to
f
or
e
c
a
s
t
th
e
num
be
r
of
di
pht
he
r
ia
c
a
s
e
s
in
th
e
ne
xt
pe
r
io
ds
a
r
e
s
ys
te
m
a
ti
c
a
ll
y
il
lu
s
tr
a
te
d
in
F
ig
ur
e
1.
T
hi
s
f
or
e
c
a
s
ti
ng
f
r
a
m
e
w
or
k
in
te
gr
a
te
s
e
pi
d
e
m
io
lo
gi
c
a
l
da
ta
w
it
h
S
V
R
e
nha
nc
e
d
by
a
G
A
f
or
pa
r
a
m
e
te
r
opt
im
iz
a
ti
on.
T
he
GA
is
e
m
pl
oye
d
to
id
e
nt
if
y
th
e
opt
im
a
l
c
om
bi
na
ti
on
of
S
V
R
pa
r
a
m
e
te
r
s
,
th
us
im
pr
ovi
ng
th
e
m
ode
l’
s
pr
e
di
c
ti
ve
pe
r
f
or
m
a
nc
e
.
T
hi
s
hybr
id
a
ppr
oa
c
h
e
na
bl
e
s
m
or
e
a
c
c
ur
a
te
a
nd
r
e
li
a
bl
e
f
or
e
c
a
s
ti
ng,
w
hi
c
h
is
e
s
s
e
nt
ia
l
f
or
ti
m
e
ly
publ
ic
he
a
lt
h
in
te
r
ve
nt
io
ns
a
nd
e
f
f
ic
ie
nt
r
e
s
our
c
e
pl
a
nni
ng.
F
ig
ur
e
1. R
e
s
e
a
r
c
h m
e
th
odol
ogy
2.2.1.
D
at
a c
ol
le
c
t
io
n
T
he
f
ir
s
t
s
te
p
w
a
s
da
ta
c
ol
le
c
ti
on,
w
hi
c
h
is
hi
ghl
y
c
r
uc
ia
l
to
th
e
e
s
s
e
nc
e
a
nd
pe
r
f
or
m
a
nc
e
of
th
e
pr
e
s
e
nt
s
tu
dy.
D
a
ta
c
onc
e
r
ni
ng
th
e
num
be
r
of
in
f
e
c
te
d
in
di
vi
dua
ls
in
s
om
e
r
e
gi
on
s
/c
it
ie
s
in
E
a
s
t
J
a
va
w
e
r
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
, V
ol
.
14
, N
o.
4
,
A
ugus
t
20
25
:
2909
-
2921
2912
c
ol
le
c
te
d
f
r
om
th
e
E
a
s
t
J
a
va
pr
ovi
nc
ia
l
he
a
lt
h
de
pa
r
tm
e
nt
.
F
r
o
m
th
e
s
a
m
e
de
pa
r
tm
e
nt
,
w
e
a
ls
o
got
th
e
da
ta
c
onc
e
r
ni
ng
th
e
im
m
uni
z
a
ti
on
r
a
nge
f
r
om
th
e
D
P
T
1,
D
P
T
2,
D
P
T
3,
a
nd
D
P
T
4
va
c
c
in
e
s
.
H
ow
e
ve
r
,
da
ta
c
onc
e
r
ni
ng
th
e
popula
ti
on
de
n
s
it
y
w
e
r
e
c
ol
le
c
te
d
f
r
om
th
e
C
e
n
tr
a
l
B
ur
e
a
u
of
S
ta
ti
s
ti
c
s
in
E
a
s
t
J
a
va
P
r
ovi
nc
e
.
T
he
c
ol
le
c
te
d
s
t
a
ti
s
ti
c
a
l
de
s
c
r
ip
ti
ve
da
ta
a
r
e
s
how
n i
n T
a
bl
e
1. D
a
ta
i
n T
a
bl
e
1 s
how
s
t
ha
t
th
e
r
e
i
s
no
da
ta
t
h
a
t
is
di
s
tr
ib
u
te
d
nor
m
a
l
ly
.
T
hi
s
s
ho
w
s
t
ha
t
t
he
s
k
e
w
ne
s
s
va
lu
e
a
s
w
e
ll
a
s
t
h
e
kur
t
o
s
i
s
de
vi
a
te
f
r
o
m
z
e
r
o.
T
a
bl
e
1.
S
ta
ti
s
ti
c
a
l
de
s
c
r
ip
ti
ve
da
ta
V
a
r
i
a
bl
e
M
i
n
M
a
x
M
e
a
n
S
t
d.
d
e
vi
a
t
i
on
S
ke
w
ne
s
s
K
ur
t
os
i
s
P
opul
a
t
i
on D
e
ns
i
t
y
528
,
793
8
,
231
,
742
1
,
684
,
752
2
,
081
,
239
2
,
450
4
,
771
S
uf
f
e
r
e
r
N
um
be
r
0.000
22
,
000
1
,
032
1
,
802
3
,
763
23
,
636
D
P
T
-
1
0.000
98
,
330
8
,
169
4
,
394
16
,
747
336
,
146
D
P
T
-
2
0.000
97
,
370
8
,
072
4
,
337
16
,
550
331
,
411
D
P
T
-
3
0.000
96
,
170
8
,
031
4
,
333
16
,
284
323
,
839
D
P
T
-
4
0.000
127
,
020
5
,
123
7
,
835
7
,
769
94
,
756
2.2.2. P
r
e
p
r
oc
e
s
s
in
g
T
he
pr
e
pr
oc
e
s
s
in
g
da
ta
w
a
s
done
to
f
il
l
out
th
e
e
m
pt
y
da
ta
a
n
d
th
e
z
e
r
o
da
ta
,
w
hos
e
va
lu
e
s
a
r
e
not
s
uppos
e
d
to
be
z
e
r
o.
M
ul
ti
pl
e
im
put
a
ti
on
s
a
r
e
n
e
c
e
s
s
a
r
y
t
o
do
s
o.
I
n
a
ddi
ti
on,
da
ta
r
e
duc
ti
on
is
a
ls
o
ne
c
e
s
s
a
r
y,
e
s
pe
c
ia
ll
y
f
or
da
t
a
th
a
t
h
a
s
no
c
le
a
r
in
f
or
m
a
ti
on.
D
a
ta
nor
m
a
li
z
a
ti
on
w
a
s
c
a
r
r
ie
d
out
s
in
c
e
th
e
da
ta
s
how
n
in
T
a
bl
e
1
ha
s
a
ve
r
y
w
id
e
r
a
nge
.
A
f
te
r
th
e
da
ta
w
a
s
r
e
a
dy,
w
e
di
vi
de
d
it
in
to
two
s
e
c
ti
on
s
:
th
e
tr
a
in
in
g s
e
t
a
nd t
he
t
e
s
ti
ng s
e
t,
w
it
h a
r
a
ti
o of
75:
25.
2.2.3.
D
at
a c
lu
s
t
e
r
in
g
D
a
ta
in
va
r
io
us
r
e
gi
ons
a
t
th
e
n
e
xt
c
a
s
e
s
tu
dy
lo
c
a
ti
on
w
e
r
e
gr
oupe
d
in
to
3
c
lu
s
te
r
s
.
G
r
oupi
ng
w
a
s
ba
s
e
d
on
th
e
hi
ghe
s
t
num
be
r
of
s
uf
f
e
r
e
r
s
in
e
a
c
h
r
e
gi
on.
T
he
c
ha
r
a
c
te
r
is
ti
c
s
of
e
a
c
h c
lu
s
te
r
,
na
m
e
ly
c
lu
s
te
r
1,
a
r
e
da
ta
f
or
c
it
ie
s
/r
e
gi
ons
th
a
t
ha
ve
th
e
hi
ghe
s
t
num
be
r
of
s
uf
f
e
r
e
r
s
w
it
h
a
s
c
or
e
m
or
e
th
a
n
10,
c
lu
s
te
r
2
is
in
th
e
r
a
nge
f
r
om
6
to
10,
a
nd
c
lu
s
te
r
3
is
f
or
le
s
s
th
a
n
or
e
qua
l
t
o
5.
F
or
e
a
c
h
c
lu
s
te
r
,
on
e
da
ta
s
e
t
w
a
s
s
e
le
c
te
d
a
s
a
r
e
pr
e
s
e
nt
a
ti
v
e
us
e
d
in
c
r
e
a
ti
ng
th
e
S
V
R
-
G
A
m
ode
l.
T
hi
s
s
e
le
c
ti
on
w
a
s
m
a
de
b
a
s
e
d
on
a
r
e
a
s
th
a
t
ha
ve
m
or
e
c
om
pl
e
x
da
ta
pa
tt
e
r
ns
.
T
he
da
ta
us
e
d
in
m
a
ki
ng
th
e
m
ode
l
in
c
lu
s
te
r
1
is
S
R
C
ity
,
c
lu
s
te
r
2
is
M
L
R
e
ge
nc
y
, a
nd
c
lu
s
te
r
3 i
s
S
M
R
e
ge
nc
y
.
2.2.4.
C
or
r
e
la
t
io
n
an
al
ys
is
T
he
c
or
r
e
la
ti
on
te
s
t
w
a
s
c
a
r
r
ie
d
out
to
de
te
r
m
in
e
th
e
r
e
la
ti
ons
hi
p
be
twe
e
n
th
e
in
de
pe
nd
e
nt
va
r
ia
bl
e
a
nd t
he
de
pe
nde
nt
va
r
ia
bl
e
. T
he
c
or
r
e
la
ti
on t
e
s
t
us
e
d i
s
t
he
P
e
a
r
s
on c
or
r
e
la
ti
on t
e
s
t
a
s
c
a
r
r
ie
d out
in
[
2
5
]
. T
he
r
e
gr
e
s
s
or
s
w
hos
e
c
or
r
e
la
ti
on
w
a
s
a
na
ly
z
e
d
w
e
r
e
im
m
uni
z
a
ti
o
n
c
ove
r
a
ge
a
nd
popula
ti
on
de
n
s
it
y,
w
hi
le
th
e
de
pe
nde
nt
va
r
ia
bl
e
w
a
s
th
e
num
be
r
of
di
pht
he
r
ia
s
uf
f
e
r
e
r
s
.
T
he
im
m
uni
z
a
ti
on
c
ove
r
a
ge
va
r
ia
bl
e
in
c
lu
de
s
D
P
T
1, D
P
T
2, D
P
T
3,
a
nd D
P
T
4. A
ll
of
t
he
s
e
t
ype
s
of
i
m
m
uni
z
a
ti
on a
r
e
i
nc
lu
de
d i
n t
he
f
or
e
c
a
s
ti
ng mode
l.
2.2.5.
T
h
e
S
V
R
m
od
e
l
c
on
s
t
r
u
c
t
io
n
T
he
r
e
a
r
e
3 m
ode
l
s
t
ha
t
w
e
r
e
f
or
m
e
d, n
a
m
e
l
y
th
e
c
lu
s
t
e
r
1 m
od
e
l,
w
hi
c
h u
s
e
s
t
he
S
R
C
it
y d
a
ta
s
e
t
, t
he
c
lu
s
t
e
r
2
m
od
e
l,
w
hi
c
h
u
s
e
s
t
he
M
L
R
e
g
e
n
c
y
d
a
ta
s
e
t
,
a
nd
t
he
c
lu
s
t
e
r
3
m
od
e
l,
w
hi
c
h
u
s
e
s
t
he
S
M
R
e
ge
n
c
y
da
ta
s
e
t
.
T
h
e
ke
r
ne
l
th
a
t
w
a
s
u
s
e
d
in
f
or
m
in
g
t
he
m
od
e
l
s
is
t
he
R
B
F
k
e
r
ne
l.
T
hi
s
S
V
R
ha
s
th
r
e
e
p
a
r
a
m
e
t
e
r
s
us
e
d
,
na
m
e
ly
p
a
r
a
m
e
t
e
r
s
C
(
c
on
s
t
a
nt
)
,
γ
(
g
a
m
m
a
)
,
a
nd
ε
(
e
ps
il
on)
.
I
n
th
e
f
ir
s
t
s
c
e
na
r
i
o,
th
e
t
hr
e
e
p
a
r
a
m
e
te
r
s
us
e
d
a
n
in
it
ia
l
v
a
lu
e
,
na
m
e
ly
0.1
,
a
nd
th
e
n
th
e
M
S
E
a
nd
s
ym
m
e
tr
ic
m
e
a
n
a
b
s
ol
ut
e
pe
r
c
e
nt
a
ge
e
r
r
or
(
S
M
A
P
E
)
va
lu
e
s
f
r
om
th
i
s
s
c
e
na
r
i
o
w
oul
d
be
c
a
lc
ul
a
te
d
.
S
V
M
i
s
a
m
a
c
hi
ne
l
e
a
r
ni
n
g
m
e
th
od
th
a
t
a
ppl
ie
s
t
he
pr
in
c
i
pl
e
of
in
duc
ti
ve
r
i
s
k m
in
im
iz
a
ti
o
n
to
obt
a
i
n
go
od
g
e
ne
r
a
li
z
a
ti
on
a
c
r
os
s
a
n
um
be
r
of
le
a
r
ni
ng
pa
t
te
r
n
s
[
26
]
.
T
he
S
V
M
im
pl
e
m
e
nt
s
le
a
r
ni
ng
a
lg
or
it
hm
s
th
a
t
a
r
e
us
e
f
ul
f
or
r
e
c
ogni
z
in
g pa
tt
e
r
n
s
i
n c
om
pl
e
x
d
a
ta
s
e
t
s
.
S
V
R
i
s
a
f
or
m
of
S
V
M
but
f
or
th
e
r
e
gr
e
s
s
io
n
c
a
s
e
[
26
]
.
T
he
go
a
l
of
th
e
S
V
R
i
s
to
f
in
d
a
f
un
c
ti
on
f
(
x)
a
s
a
h
ype
r
pl
a
n
e
in
th
e
f
or
m
of
a
r
e
gr
e
s
s
i
on
f
unc
ti
on
th
a
t
f
it
s
a
l
l
of
th
e
in
p
ut
da
t
a
by
m
a
xi
m
iz
i
ng
th
e
m
a
r
gi
n
b
e
tw
e
e
n
tw
o
c
la
s
s
e
s
a
nd
m
in
im
iz
in
g
th
e
e
r
r
or
a
s
li
tt
l
e
a
s
po
s
s
i
bl
e
[
27
]
.
S
up
po
s
e
t
he
f
u
n
c
ti
on
i
n
(
1)
i
s
a
r
e
gr
e
s
s
io
n
li
n
e
a
s
a
n
opt
im
a
l
hype
r
pl
a
ne
[
18]
, [
19]
.
(
)
=
(
)
+
(
1)
M
a
xi
m
iz
in
g
th
e
m
a
r
gi
n
in
c
r
e
a
s
e
s
th
e
pr
oba
bi
li
ty
of
th
e
da
ta
f
a
ll
in
g
w
it
hi
n
th
e
±ε
r
a
di
us
.
T
he
r
e
f
or
e
,
to
m
a
xi
m
iz
e
m
a
r
gi
n,
a
m
in
im
um
‖
w
‖
is
r
e
qui
r
e
d
[
27
]
.
I
t
is
a
s
s
um
e
d
th
a
t
a
ll
poi
nt
s
a
r
e
in
th
e
r
a
ng
e
f
(
x)
±ε
(
f
e
a
s
ib
le
)
,
w
he
r
e
th
e
r
e
a
r
e
s
e
v
e
r
a
l
poi
nt
s
th
a
t
m
a
y
be
out
of
th
e
r
a
nge
,
th
e
n
s
la
c
k
v
a
r
ia
bl
e
s
ξ
a
nd
ξ
^
*
a
r
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
D
at
a
-
dr
iv
e
n s
uppor
t
v
e
c
to
r
r
e
gr
e
s
s
io
n
-
ge
ne
ti
c
al
gor
it
hm
m
ode
l
f
or
pr
e
di
c
ti
ng
…
(
W
iw
ik
A
nggr
ae
ni
)
2913
a
dde
d
to
ove
r
c
om
e
th
e
pr
obl
e
m
of
in
f
e
a
s
ib
le
c
ons
tr
a
in
ts
in
th
e
opt
im
iz
a
ti
on
p
r
obl
e
m
[
27
]
.
N
e
xt
,
th
e
opt
im
iz
a
ti
on pr
obl
e
m
c
a
n be
f
or
m
ul
a
te
d a
s
(
2)
to
(
5)
[
21
]
, [
27
]
.
1
2
‖
‖
2
+
∑
(
+
∗
)
=
1
(
2)
W
it
h
c
ondi
ti
ons
:
−
(
)
−
−
≤
unt
uk
=
1
,
…
,
(
3)
(
)
−
+
−
∗
≤
unt
uk
=
1
,
…
,
(
4)
,
∗
≥
0
(
5)
w
he
r
e
x
i
is
t
he
in
put
da
ta
,
y
i
is
t
he
out
put
,
ω
,
ξ
,
ε
a
r
e
t
he
p
a
r
a
m
e
te
r
s
w
hos
e
va
lu
e
s
w
il
l
be
s
e
a
r
c
he
d f
or
.
I
n
th
e
S
V
R
,
th
e
r
e
is
a
ke
r
ne
l
f
unc
ti
on. T
he
k
e
r
ne
l
f
unc
ti
on
is
us
e
d
to
m
a
ke
non
-
li
ne
a
r
da
ta
s
e
pa
r
a
bl
e
by
m
ovi
ng
it
to
a
hi
ghe
r
-
d
im
e
ns
io
na
l
s
pa
c
e
.
T
he
ke
r
ne
l
th
a
t
w
il
l
be
us
e
d
in
th
e
S
V
R
is
th
e
R
B
F
ke
r
ne
l
be
c
a
us
e
it
ha
s
be
e
n
pr
ove
n
to
pr
oduc
e
be
tt
e
r
p
r
e
di
c
ti
on
pe
r
f
o
r
m
a
nc
e
[
28
]
.
T
he
R
B
F
ke
r
ne
l
f
or
m
ul
a
ti
on
a
s
s
how
n
in
(
6)
[
20]
,
[
29
]
.
I
n
th
e
S
V
R
,
s
e
le
c
ti
ng
hype
r
-
opt
i
m
a
l
pa
r
a
m
e
te
r
s
is
a
n
im
por
ta
nt
s
te
p.
T
he
s
e
pa
r
a
m
e
te
r
s
in
c
lu
de
a
r
e
pa
r
a
m
e
te
r
C
th
a
t
de
te
r
m
in
e
s
th
e
tr
a
d
e
-
of
f
c
os
ts
a
s
a
pe
na
lt
y
due
to
a
n
in
f
e
a
s
ib
le
pr
obl
e
m
,
pa
r
a
m
e
te
r
γ
th
a
t
s
how
s
th
e
ba
ndw
id
th
of
th
e
ke
r
ne
l
f
unc
ti
on
w
hi
c
h
r
e
pr
e
s
e
nt
s
th
e
va
r
ia
nc
e
of
th
e
R
B
F
ke
r
ne
l
f
unc
ti
on,
a
nd
ε
-
in
s
e
ns
it
iv
e
lo
s
s
f
unc
ti
on
w
hi
c
h
i
s
th
e
di
s
ta
nc
e
be
tw
e
e
n
th
e
hype
r
pl
a
ne
a
nd
2
bounda
r
y l
in
e
s
e
qua
l
to
a
c
c
ur
a
c
y of
t
he
e
s
ti
m
a
te
pl
a
c
e
d on the
t
r
a
in
in
g da
ta
poi
nt
s
[
18]
, [
19]
, [
27
].
(
,
)
=
e
x
p
(
−
1
2
2
‖
−
‖
2
)
(
6)
2.2.6.
T
h
e
S
V
R
-
G
A
m
od
e
l
c
on
s
t
r
u
c
t
io
n
P
a
r
a
m
e
te
r
tu
ni
ng
a
im
s
to
f
in
d
th
e
opt
im
a
l
S
V
R
hype
r
pa
r
a
m
e
te
r
va
lu
e
s
.
I
n
th
i
s
s
tu
dy,
opt
im
a
l
pa
r
a
m
e
te
r
de
f
in
it
io
n
w
a
s
c
a
r
r
ie
d
out
us
in
g
th
e
G
A
m
e
th
od.
G
A
is
a
m
e
ta
he
ur
is
ti
c
m
e
th
od
in
s
pi
r
e
d
by
th
e
na
tu
r
a
l
s
e
le
c
ti
on
pr
oc
e
s
s
[
3
0
]
.
G
A
us
e
s
c
r
os
s
ove
r
a
nd
m
ut
a
ti
o
n
ope
r
a
to
r
s
,
w
hi
c
h
m
a
ke
th
e
popula
ti
on
m
or
e
di
ve
r
s
e
a
nd
th
u
s
m
or
e
im
m
une
to
be
in
g
tr
a
ppe
d
in
a
lo
c
a
l
opt
im
um
.
I
n
th
e
or
y,
di
ve
r
s
it
y
a
ls
o
h
e
lp
s
th
e
a
lg
or
it
hm
be
f
a
s
te
r
in
r
e
a
c
hi
ng
th
e
gl
oba
l
opt
im
um
be
c
a
us
e
it
w
il
l
a
ll
ow
th
e
a
lg
or
it
hm
to
e
xpl
or
e
th
e
s
ol
ut
io
n
s
pa
c
e
m
or
e
qui
c
kl
y [
23]
.
I
n
s
e
a
r
c
hi
ng
f
or
S
V
R
pa
r
a
m
e
te
r
s
w
it
h
th
e
G
A
,
th
e
r
e
a
r
e
s
ta
ge
s
of
in
it
ia
l
s
ol
ut
io
n,
f
it
ne
s
s
f
unc
ti
on,
s
e
le
c
ti
on,
c
r
os
s
ov
e
r
,
a
nd
m
ut
a
ti
on.
T
he
le
ngt
h
of
th
e
c
hr
om
os
om
e
is
3,
e
a
c
h
r
e
pr
e
s
e
nt
in
g
th
e
S
V
R
pa
r
a
m
e
te
r
w
hos
e
opt
im
a
l
va
lu
e
w
il
l
be
s
ought,
na
m
e
ly
c
on
s
ta
nt
,
ga
m
m
a
,
a
nd
e
p
s
il
on.
A
t
th
e
in
it
ia
l
s
ol
ut
io
n
s
ta
ge
,
a
r
a
ndom
m
e
th
od
w
a
s
us
e
d
to
obt
a
in
10
c
hr
om
os
om
e
s
.
T
he
n
th
e
10
c
hr
om
os
om
e
s
w
il
l
unde
r
go
s
e
le
c
ti
on,
c
r
os
s
ove
r
,
a
nd
m
ut
a
ti
on
to
r
e
a
c
h
100
c
hr
om
os
om
e
s
.
E
a
c
h
c
hr
o
m
os
om
e
'
s
va
lu
e
w
il
l
be
e
nt
e
r
e
d
in
to
th
e
S
V
R
m
ode
l
s
o
th
a
t
th
e
M
S
E
a
nd
S
M
A
P
E
c
a
lc
ul
a
ti
ons
c
a
n
be
c
a
r
r
ie
d
out
.
T
he
f
it
ne
s
s
f
unc
ti
on
us
e
d
is
th
e
s
m
a
ll
e
s
t
M
S
E
a
nd S
M
A
P
E
va
lu
e
s
. T
he
pa
r
a
m
e
te
r
t
uni
ng i
s
c
a
r
r
ie
d out on e
a
c
h m
ode
l
th
a
t
ha
s
b
e
e
n f
or
m
e
d.
2.2.7.
S
e
le
c
t
io
n
of
t
h
e
b
e
s
t
c
om
b
in
at
io
n
of
var
ia
b
le
s
A
f
te
r
obt
a
in
in
g
th
e
be
s
t
m
ode
l
a
nd
pa
r
a
m
e
te
r
s
,
th
e
ne
xt
s
te
p
is
to
c
r
e
a
te
a
c
om
bi
na
ti
on
of
va
r
ia
bl
e
s
.
T
he
s
e
c
om
bi
na
ti
ons
a
r
e
be
twe
e
n
de
pe
nd
e
nt
va
r
ia
bl
e
na
m
e
ly
n
um
be
r
of
s
uf
f
e
r
e
r
a
nd
i
n
de
p
e
nde
pe
nt
va
r
ia
bl
e
s
li
ke
im
m
uni
z
a
ti
on
r
a
nge
a
nd
popula
ti
on
de
n
s
it
y,
im
m
uni
z
a
ti
on
r
a
nge
,
a
nd
popula
ti
on
d
e
ns
it
y
.
T
he
c
om
bi
na
ti
on
of
va
r
ia
bl
e
s
w
it
h
th
e
s
m
a
ll
e
s
t
M
S
E
a
nd
S
M
A
P
E
va
lu
e
s
w
il
l
be
s
e
le
c
te
d
a
s
th
e
m
ode
l
f
o
r
c
a
r
r
yi
ng out t
he
ne
xt
f
or
e
c
a
s
ti
ng pr
oc
e
s
s
.
2.2.8.
P
e
r
f
or
m
an
c
e
e
val
u
at
io
n
T
o
m
e
a
s
ur
e
th
e
pe
r
f
or
m
a
nc
e
of
f
or
e
c
a
s
ti
ng
r
e
s
ul
ts
,
th
is
s
tu
dy
us
e
s
M
S
E
a
nd
S
M
A
P
E
[2
5
]
.
M
S
E
is
th
e
s
um
of
th
e
di
f
f
e
r
e
nc
e
s
be
twe
e
n
f
or
e
c
a
s
t
da
ta
a
nd
a
c
tu
a
l
da
ta
.
B
e
s
id
e
s
,
S
M
A
P
E
is
a
n
a
lt
e
r
na
ti
ve
to
c
a
lc
ul
a
ti
ng M
A
P
E
w
he
n t
he
a
c
tu
a
l
da
ta
i
s
0 or
c
lo
s
e
t
o 0.
2.2.9.
F
or
e
c
as
t
in
g t
h
e
n
e
xt
p
e
r
io
d
F
or
e
c
a
s
ti
ng
is
pe
r
f
or
m
e
d us
in
g a
pr
e
di
c
ti
ve
m
ode
l
c
onf
ig
ur
e
d
w
it
h t
he
opt
im
a
l
s
e
t
of
pa
r
a
m
e
te
r
s
a
nd
th
e
m
os
t
r
e
le
va
nt
c
om
bi
na
ti
on
of
in
put
va
r
ia
bl
e
s
.
T
hi
s
a
ppr
oa
c
h
is
de
s
ig
ne
d
to
e
nha
nc
e
th
e
a
c
c
ur
a
c
y
of
th
e
f
or
e
c
a
s
ti
ng r
e
s
ul
ts
.
T
he
m
ode
l
is
t
he
n a
ppl
ie
d t
o e
s
ti
m
a
te
t
he
nu
m
be
r
of
di
pht
he
r
ia
c
a
s
e
s
th
a
t
c
a
n ha
ppe
n
ove
r
th
e
ne
xt
24
pe
r
io
ds
in
S
R
C
it
y,
M
L
R
e
ge
nc
y,
a
nd
S
M
R
e
ge
n
c
y
,
th
e
r
e
by
pr
ovi
di
ng
va
lu
a
bl
e
in
s
ig
ht
s
f
or
lo
c
a
l
publ
ic
he
a
lt
h pl
a
nni
ng a
nd i
nt
e
r
ve
nt
io
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
, V
ol
.
14
, N
o.
4
,
A
ugus
t
20
25
:
2909
-
2921
2914
2.2.10.
R
ob
u
s
t
n
e
s
s
t
e
s
t
m
od
e
l
T
he
m
ode
l
w
it
h
th
e
be
s
t
pa
r
a
m
e
te
r
s
a
nd
c
om
bi
na
ti
on
of
va
r
ia
bl
e
s
w
il
l
be
te
s
te
d
on
ot
he
r
da
ta
s
e
ts
th
a
t
a
r
e
s
ti
ll
in
th
e
s
a
m
e
c
lu
s
te
r
a
s
th
e
da
ta
s
e
t
u
s
e
d
to
c
r
e
a
te
t
he
m
ode
l.
T
hi
s
i
s
done
to
f
in
d
out
how
r
obus
t
th
e
S
V
R
-
G
A
m
ode
l
th
a
t
ha
s
be
e
n
c
r
e
a
te
d
is
.
T
he
da
ta
s
e
t
w
a
s
s
e
le
c
te
d
us
in
g
a
s
im
pl
e
r
a
ndom
s
a
m
pl
in
g
m
e
th
od
.
C
lu
s
te
r
1
,
w
it
h
th
e
hi
ghe
s
t
num
be
r
of
r
e
gi
ons
,
w
il
l
be
te
s
te
d
f
or
r
obus
tn
e
s
s
in
4
ot
he
r
r
e
gi
ons
.
A
s
f
or
c
lu
s
te
r
s
2 a
nd 3, the
y w
e
r
e
t
e
s
te
d i
n 3 a
nd 2 othe
r
r
e
gi
ons
,
r
e
s
pe
c
ti
ve
ly
.
3.
R
E
S
U
L
T
S
A
N
D
D
I
S
C
U
S
S
I
O
N
T
he
r
e
s
ul
t
s
of
th
e
c
or
r
e
la
ti
on
te
s
t
u
s
e
d
to
de
te
r
m
in
e
th
e
in
f
l
ue
nc
e
be
tw
e
e
n
va
r
ia
bl
e
s
in
v
a
r
io
us
s
a
m
pl
e
a
r
e
a
s
a
r
e
s
how
n
in
T
a
bl
e
2
.
T
a
bl
e
2
s
how
s
th
a
t,
f
or
th
e
c
it
y
of
SB
,
th
e
D
P
T
1
va
r
ia
bl
e
ha
s
a
di
r
e
c
tl
y
pr
opor
ti
ona
l
r
e
la
ti
ons
hi
p
w
it
h
th
e
num
be
r
of
s
uf
f
e
r
e
r
s
,
w
h
il
e
th
e
va
r
ia
bl
e
s
D
P
T
2,
D
P
T
3,
D
P
T
4,
a
nd
popula
ti
on
de
ns
it
y
ha
ve
a
n
in
ve
r
s
e
ly
pr
opor
ti
ona
l
r
e
la
ti
ons
hi
p. M
e
a
nw
hi
le
, a
ll
p
-
va
lu
e
s
a
r
e
gr
e
a
te
r
th
a
n
0.05,
s
o
th
e
r
e
is
no s
ig
ni
f
ic
a
nt
r
e
la
ti
ons
hi
p
be
tw
e
e
n
va
r
ia
bl
e
s
.
H
ow
e
ve
r
,
th
e
P
-
va
lu
e
f
or
th
e
popula
ti
on
de
ns
it
y
a
nd
num
be
r
of
s
uf
f
e
r
e
r
s
va
r
ia
bl
e
s
is
s
m
a
ll
e
r
th
a
n
th
e
ot
he
r
va
r
ia
b
le
s
,
s
o
e
v
e
n
th
ough
th
e
va
lu
e
is
gr
e
a
t
e
r
th
a
n
0.05,
th
e
popula
ti
on
de
n
s
it
y
va
r
ia
bl
e
ha
s
a
gr
e
a
te
r
in
f
lu
e
nc
e
on
th
e
num
b
e
r
of
s
uf
f
e
r
e
r
s
th
a
n
th
e
ot
he
r
va
r
ia
bl
e
s
.
T
a
bl
e
2. C
or
r
e
la
ti
on
te
s
t
r
e
s
ul
t
be
twe
e
n r
e
gr
e
s
s
or
va
r
ia
bl
e
s
a
nd
th
e
num
be
r
of
s
uf
f
e
r
e
r
s
V
a
r
i
a
bl
e
S
B
C
i
t
y
M
L
R
e
ge
nc
y
S
M
R
e
ge
nc
y
P
e
a
r
s
on
c
or
r
e
l
a
t
i
on
c
oe
f
f
i
c
i
e
nt
P
-
va
l
ue
P
e
a
r
s
on
c
or
r
e
l
a
t
i
on
c
oe
f
f
i
c
i
e
nt
P
-
va
l
ue
P
e
a
r
s
on
c
or
r
e
l
a
t
i
on
c
oe
f
f
i
c
i
e
nt
P
-
va
l
ue
D
P
T
-
1
–
s
uf
f
e
r
e
r
nu
m
be
r
0.054
0.655
-
0.011
0.931
0.061
0.619
D
P
T
-
2
–
s
uf
f
e
r
e
r
nu
m
be
r
-
0.007
0.953
-
0.008
0.948
0.102
0.395
D
P
T
-
3
–
s
uf
f
e
r
e
r
nu
m
be
r
-
0.034
0.780
-
0.010
0.936
0.112
0.348
D
P
T
-
4
–
s
uf
f
e
r
e
r
nu
m
be
r
-
0.083
0.488
-
0.131
0.273
-
0.027
0.824
P
opul
a
t
i
on
de
ns
i
t
y
–
s
uf
f
e
r
e
r
num
be
r
-
0.187
0.117
0.035
0.769
0.052
0.665
M
e
a
nw
hi
le
,
f
or
M
L
R
e
ge
nc
y
,
th
e
va
r
ia
bl
e
s
D
P
T
1,
D
P
T
2,
D
P
T
3
,
a
nd
D
P
T
4
ha
ve
a
n
in
ve
r
s
e
r
e
la
ti
ons
hi
p
w
it
h
th
e
num
be
r
of
s
uf
f
e
r
e
r
s
.
M
e
a
nw
hi
le
,
th
e
popula
ti
on
de
ns
it
y
va
r
ia
bl
e
ha
s
a
di
r
e
c
tl
y
pr
opor
ti
ona
l
r
e
la
ti
ons
hi
p
w
it
h
th
e
num
be
r
of
s
uf
f
e
r
e
r
s
.
A
ll
p
-
va
lu
e
s
a
r
e
gr
e
a
te
r
th
a
n
th
e
pr
e
de
te
r
m
in
e
d
th
r
e
s
hol
d
va
lu
e
,
s
o
th
a
t
th
e
r
e
is
no
s
ig
ni
f
ic
a
nt
r
e
la
ti
ons
hi
p
be
twe
e
n
va
r
ia
bl
e
s
.
H
ow
e
ve
r
,
th
e
p
-
va
lu
e
of
th
e
D
P
T
4 va
r
ia
bl
e
i
s
s
m
a
ll
e
r
t
ha
n t
he
ot
he
r
va
r
ia
bl
e
s
, s
o e
ve
n t
hough the
va
lu
e
i
s
gr
e
a
te
r
t
ha
n t
he
r
e
a
l
le
ve
l
va
lu
e
(
0.05)
, t
he
D
P
T
4 va
r
ia
bl
e
ha
s
a
gr
e
a
te
r
i
nf
lu
e
nc
e
on t
he
numbe
r
of
s
uf
f
e
r
e
r
s
t
ha
n t
he
ot
he
r
va
r
ia
bl
e
s
.
T
he
n,
f
or
SM
R
e
ge
nc
y,
th
e
va
r
ia
bl
e
s
D
P
T
1,
D
P
T
2,
D
P
T
3,
a
n
d
popula
ti
on
de
ns
it
y
ha
ve
a
di
r
e
c
tl
y
pr
opor
ti
ona
l
r
e
la
ti
ons
hi
p
w
i
th
th
e
num
be
r
of
s
uf
f
e
r
e
r
s
,
w
hi
le
t
he
D
P
T
4
va
r
ia
bl
e
ha
s
a
n
in
ve
r
s
e
r
e
la
ti
ons
hi
p
w
it
h
th
e
num
be
r
of
s
uf
f
e
r
e
r
s
.
A
ll
P
-
va
lu
e
va
lu
e
s
a
r
e
gr
e
a
te
r
th
a
n
th
e
r
e
a
l
le
ve
l
va
lu
e
s
th
a
t
ha
ve
be
e
n
de
te
r
m
in
e
d,
s
o
th
e
r
e
is
no
s
ig
ni
f
ic
a
nt
r
e
la
ti
ons
hi
p
be
twe
e
n
va
r
ia
bl
e
s
.
H
ow
e
ve
r
,
th
e
P
-
va
lu
e
of
th
e
va
r
ia
bl
e
s
D
P
T
2
a
nd
D
P
T
3
ha
s
a
s
m
a
ll
e
r
va
lu
e
th
a
n
ot
he
r
va
r
ia
bl
e
s
,
s
o
e
ve
n
th
ough
th
e
va
lu
e
is
gr
e
a
te
r
th
a
n
th
e
r
e
a
l
le
ve
l
va
lu
e
(
0.05)
, t
hi
s
va
r
ia
bl
e
ha
s
a
gr
e
a
t
e
r
i
nf
lu
e
nc
e
on t
he
n
um
be
r
of
s
uf
f
e
r
e
r
s
t
ha
n ot
he
r
va
r
ia
bl
e
s
.
T
he
c
or
r
e
la
ti
on t
e
s
t
r
e
s
ul
ts
i
n T
a
bl
e
2 s
how
t
ha
t
a
ll
r
e
gr
e
s
s
or
va
r
ia
bl
e
s
a
r
e
s
a
id
t
o ha
ve
no s
ig
ni
f
ic
a
nt
e
f
f
e
c
t
on t
he
numbe
r
of
s
uf
f
e
r
e
r
s
. T
he
ge
ne
r
a
l
c
onc
lu
s
io
n i
s
, of
c
our
s
e
, t
he
oppos
it
e
of
w
ha
t
w
a
s
c
onve
ye
d by
[
3]
,
[
4]
,
w
ho
a
r
gue
th
a
t
m
a
ny
f
a
c
to
r
s
c
a
n
in
f
lu
e
nc
e
th
e
num
be
r
of
s
uf
f
e
r
e
r
s
of
di
pht
he
r
ia
.
A
s
w
e
ll
a
s
f
or
[
11]
,
[
24]
w
ho
s
ta
te
th
a
t
v
a
c
c
in
a
ti
on
ha
s
a
bi
g
im
pa
c
t
on
r
e
du
c
in
g
th
e
num
be
r
of
s
uf
f
e
r
e
r
s
.
H
ow
e
ve
r
,
e
ve
n
th
ough
it
doe
s
not
ha
ve
a
s
ig
ni
f
ic
a
nt
e
f
f
e
c
t,
th
e
c
or
r
e
la
ti
on
c
o
e
f
f
ic
ie
nt
s
ti
ll
ha
s
a
non
-
nul
l
va
lu
e
, s
o
it
c
a
n s
ti
ll
be
c
ons
id
e
r
e
d
to
ha
ve
a
n
in
f
lu
e
nc
e
,
e
ve
n
th
ough
it
is
not
s
ig
ni
f
ic
a
nt
.
S
o,
in
de
te
r
m
in
in
g
th
e
be
s
t
m
ode
l
,
s
e
ve
r
a
l
s
c
e
na
r
io
s
a
r
e
u
s
e
d t
ha
t
in
vol
ve
th
e
s
e
r
e
gr
e
s
s
or
va
r
ia
bl
e
s
.
A
f
te
r
th
e
c
or
r
e
la
ti
on
te
s
t
is
c
a
r
r
ie
d
out
,
th
e
ne
xt
s
te
p
is
to
f
in
d
th
e
be
s
t
pa
r
a
m
e
t
e
r
va
lu
e
s
.
P
a
r
a
m
e
t
e
r
tu
ni
ng
is
c
a
r
r
ie
d
out
us
in
g
tr
a
in
in
g
d
a
ta
,
w
hi
c
h
is
in
f
lu
e
nc
e
d
by
a
ll
v
a
r
ia
bl
e
s
.
T
uni
ng
of
S
V
R
pa
r
a
m
e
te
r
s
(
C
,
ε
,
a
nd
γ
)
w
a
s
c
a
r
r
ie
d
out
us
in
g
th
e
G
A
m
e
th
od.
T
he
be
s
t
pa
r
a
m
e
te
r
va
lu
e
s
in
e
a
c
h
c
lu
s
te
r
a
r
e
s
ho
w
n
in
T
a
bl
e
3
.
I
n
e
a
c
h
c
lu
s
te
r
,
th
e
gr
e
a
te
r
th
e
C
va
lu
e
,
th
e
be
tt
e
r
th
e
r
e
s
ul
ti
ng
M
S
E
a
nd
S
M
A
P
E
va
lu
e
s
,
w
hi
le
th
e
s
m
a
ll
e
r
t
he
ga
m
m
a
a
nd e
ps
il
on va
lu
e
s
, t
he
be
tt
e
r
t
he
r
e
s
ul
ti
ng
M
S
E
a
nd S
M
A
P
E
va
lu
e
s
. H
ow
e
ve
r
, t
he
va
lu
e
s
of
t
he
s
e
t
hr
e
e
pa
r
a
m
e
te
r
s
ha
ve
di
f
f
e
r
e
nt
r
a
nge
s
i
n e
a
c
h c
lu
s
te
r
.
T
a
bl
e
3
s
how
s
th
a
t
th
e
di
f
f
e
r
e
nc
e
s
i
n t
he
M
S
E
a
nd
S
M
A
P
E
va
lu
e
s
of
th
e
in
it
ia
ti
on
pa
r
a
m
e
te
r
s
a
nd
th
e
be
s
t
pa
r
a
m
e
te
r
s
s
how
qui
te
s
ig
ni
f
ic
a
nt
r
e
s
ul
ts
,
s
o
it
c
a
n
be
c
onf
ir
m
e
d
th
a
t
th
e
r
e
s
ul
ts
of
pa
r
a
m
e
te
r
tu
ni
ng
a
r
e
th
e
be
s
t
pa
r
a
m
e
te
r
va
lu
e
s
f
r
om
a
ll
th
e
e
xpe
r
im
e
nt
s
th
a
t
ha
ve
be
e
n
c
a
r
r
ie
d
out
.
A
f
te
r
obt
a
in
in
g
th
e
be
s
t
pa
r
a
m
e
te
r
va
lu
e
s
,
th
e
ne
xt
e
xpe
r
im
e
nt
w
a
s
c
a
r
r
ie
d
ou
t
us
in
g
s
e
ve
r
a
l
s
c
e
na
r
io
c
om
bi
na
ti
on
s
of
r
e
gr
e
s
s
or
va
r
ia
bl
e
s
t
o
f
or
e
c
a
s
t
th
e
ne
xt
pe
r
io
d.
T
a
bl
e
4
s
how
s
a
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
D
at
a
-
dr
iv
e
n s
uppor
t
v
e
c
to
r
r
e
gr
e
s
s
io
n
-
ge
ne
ti
c
al
gor
it
hm
m
ode
l
f
or
pr
e
di
c
ti
ng
…
(
W
iw
ik
A
nggr
ae
ni
)
2915
c
om
pa
r
is
on
of
M
S
E
a
nd
S
M
A
P
E
va
lu
e
s
r
e
s
ul
ti
ng
f
r
om
va
r
io
u
s
c
om
bi
na
ti
ons
of
va
r
ia
bl
e
s
f
or
th
e
c
lu
s
te
r
1,
c
lu
s
te
r
2
,
a
nd c
lu
s
te
r
3 s
a
m
pl
e
a
r
e
a
s
.
T
a
bl
e
3. P
a
r
a
m
e
te
r
tu
ni
ng r
e
s
ul
t
us
in
g ge
n
e
ti
c
a
lg
or
it
hm
C
l
us
t
e
r
P
a
r
a
m
e
t
e
r
M
S
E
S
M
A
P
E
(
%
)
µ
σ
e
µ
Σ
I
ni
a
t
i
on
B
e
s
t
pa
r
a
m
e
t
e
r
va
l
ue
1
10
0.0022
0.0065
6,571
0.0001371
36.88
11.53
2
40
0.0001
0.0001
2,039
5.203E
-
07
40.90
22.65
3
8
0.0001
0.0001
0.675
6.951E
-
08
89.38
52.83
T
a
bl
e
4. P
a
r
a
m
e
te
r
tu
ni
ng r
e
s
ul
t
ba
s
e
d on dif
f
e
r
e
nt
r
e
gr
e
s
s
or
va
r
ia
bl
e
c
om
bi
na
ti
ons
f
or
e
a
c
h c
lu
s
te
r
I
nde
pe
nde
nt
va
r
i
a
bl
e
D
e
pe
nde
nt
va
r
i
a
bl
e
s
C
l
us
t
e
r
-
1
-
S
B
C
i
t
y
C
l
us
t
e
r
-
2
–
M
L
R
e
ge
nc
y
C
l
us
t
e
r
-
3
-
S
M
R
e
ge
nc
y
M
S
E
S
M
A
P
E
(
%
)
M
S
E
S
M
A
P
E
(
%
)
M
S
E
S
M
A
P
E
(
%
)
T
r
a
i
ni
ng
T
e
s
t
i
ng
T
r
a
i
ni
ng
T
e
s
t
i
ng
T
r
a
i
ni
ng
T
e
s
t
i
ng
T
r
a
i
ni
ng
T
e
s
t
i
ng
T
r
a
i
ni
ng
T
e
s
t
i
ng
T
r
a
i
ni
ng
T
e
s
t
i
ng
A
l
l
i
nde
pe
nde
nt
va
r
i
a
bl
e
s
S
uf
f
e
r
e
r
num
be
r
0.0001
0.218
11.52
6.42
0.000
0.000
26.42
23.53
0.000
0.000
67.92
52.94
I
m
m
uni
z
a
t
i
on
(
D
P
T
1
–
D
P
T
4
)
S
uf
f
e
r
e
r
num
be
r
0.4720
0.123
13.18
6.95
0.134
0.000
28.07
23.53
0.026
0.002
68.55
53.08
P
opul
a
t
i
on
de
ns
i
t
y
S
uf
f
e
r
e
r
num
be
r
0.0001
0.216
11.52
6.42
0.073
0.007
29.70
23.81
0.167
0.000
76.34
63.57
T
h
e
s
m
a
l
le
r
th
e
M
S
E
a
nd
S
M
A
P
E
va
lu
e
s
,
t
he
b
e
tt
e
r
t
he
m
o
de
l
pe
r
f
or
m
a
n
c
e
.
B
a
s
e
d
on
T
a
b
le
4
,
th
e
s
c
e
na
r
io
w
i
th
a
c
o
m
bi
na
ti
o
n
of
a
l
l
in
de
pe
nd
e
nt
v
a
r
ia
bl
e
s
i
n
c
l
u
s
t
e
r
1
,
r
e
pr
e
s
e
nt
e
d
b
y
S
B
C
it
y,
s
ho
w
s
th
e
b
e
s
t
pe
r
f
or
m
a
n
c
e
.
L
ik
e
w
is
e
,
f
or
c
l
u
s
t
e
r
2,
w
hi
c
h
i
s
r
e
pr
e
s
e
n
te
d
by
M
L
R
e
g
e
n
c
y
,
a
nd
c
lu
s
t
e
r
3
b
y
S
M
R
e
ge
nc
y.
I
t
c
a
n
b
e
s
e
e
n
t
ha
t
t
he
b
e
s
t
m
od
e
l
i
s
a
m
od
e
l
i
nv
ol
v
in
g
a
ll
r
e
gr
e
s
s
or
va
r
i
a
bl
e
s
,
a
n
d
t
he
p
a
r
a
m
e
te
r
s
u
s
e
d
a
r
e
a
s
s
h
ow
n
in
T
a
bl
e
4
.
C
om
pa
r
is
on of
a
c
tu
a
l
da
ta
a
n
d
f
o
r
e
c
a
s
t
r
e
s
ul
t
s
f
or
tr
a
i
ni
n
g a
nd
te
s
ti
n
g
d
a
t
a
i
s
s
ho
w
n
f
or
e
a
c
h
c
lu
s
t
e
r
,
a
s
s
ho
w
n
in
F
ig
ur
e
2
.
A
f
te
r
th
e
m
od
e
l
i
s
pr
ov
e
n
to
h
a
v
e
g
oo
d
p
e
r
f
or
m
a
nc
e
,
th
e
m
o
de
l
i
s
t
he
n
us
e
d
to
pr
e
di
c
t
t
he
num
be
r
of
di
p
ht
h
e
r
i
a
s
u
f
f
e
r
e
r
s
i
n
th
e
ne
xt
2
4
p
e
r
i
od
s
.
T
h
e
r
e
s
ul
t
s
of
f
ut
ur
e
f
or
e
c
a
s
t
in
g
f
or
SB
C
i
ty
,
M
L
R
e
g
e
n
c
y
,
a
nd
S
M
R
e
g
e
n
c
y
a
r
e
a
l
s
o
s
ho
w
n
i
n
F
i
gur
e
2
f
r
o
m
th
e
7
3
rd
p
e
r
i
od
on
w
a
r
d
s
.
F
or
e
c
a
s
ti
n
g
r
e
s
u
lt
s
f
or
t
he
n
e
x
t
24
p
e
r
i
od
s
in
t
he
S
B
c
i
ty
,
a
s
d
is
pl
a
y
e
d
in
F
i
gur
e
2(
a
)
,
s
h
ow
p
a
tt
e
r
n
va
l
ue
s
th
a
t
a
r
e
s
im
i
la
r
to
tr
a
in
i
ng
a
n
d
t
e
s
ti
ng
da
ta
f
or
p
r
e
vi
o
u
s
p
e
r
io
d
s
c
om
pa
r
e
d
to
f
or
e
c
a
s
ti
ng
r
e
s
ul
t
s
in
ot
he
r
r
e
gi
on
s
.
T
hi
s
m
e
a
n
s
th
a
t
th
e
m
o
d
e
l
i
n
t
he
c
i
ty
of
S
B
ha
s
th
e
be
s
t
pe
r
f
or
m
a
n
c
e
c
om
pa
r
e
d
to
ot
he
r
d
a
t
a
s
e
t
s
.
L
ik
e
w
i
s
e
,
f
o
r
e
c
a
s
ti
n
g
f
or
t
he
ne
xt
24
p
e
r
i
od
s
i
n
M
L
R
e
g
e
n
c
y
i
n
F
ig
ur
e
2(
b)
s
h
ow
s
t
h
a
t
th
e
f
or
e
c
a
s
t
li
ne
c
a
n
f
ol
lo
w
t
h
e
a
c
t
ua
l
d
a
t
a
pa
t
te
r
n
e
v
e
n
th
o
ug
h
t
he
r
e
a
r
e
s
e
ve
r
a
l
p
e
r
io
d
s
t
ha
t
a
r
e
s
li
g
ht
l
y
d
if
f
e
r
e
nt
f
r
om
th
e
a
c
t
ua
l
da
t
a
.
T
he
f
or
e
c
a
s
ti
ng
r
e
s
u
lt
s
f
or
S
B
C
i
ty
a
n
d
M
L
R
e
g
e
n
c
y
s
h
ow
t
ha
t
th
e
S
V
R
-
G
A
a
lg
o
r
it
hm
i
s
s
u
it
a
bl
e
f
o
r
a
pp
li
c
a
ti
o
n t
o
da
ta
th
a
t
ha
s
a
no
n
-
l
in
e
a
r
pa
tt
e
r
n
but
te
nd
s
t
o f
lu
c
t
ua
t
e
a
r
o
und
th
e
a
ve
r
a
ge
, w
i
th
a
l
a
r
g
e
d
a
t
a
r
a
ng
e
a
nd
a
l
ot
of
s
m
a
l
l
d
a
t
a
.
H
ow
e
ve
r
,
th
is
i
s
di
f
f
e
r
e
nt
f
r
om
t
he
f
or
e
c
a
s
ti
ng
r
e
s
ul
ts
f
or
t
he
ne
xt
2
4
p
e
r
i
od
s
in
S
M
R
e
ge
nc
y
a
s
s
how
n
i
n
F
ig
ur
e
2(
c
)
.
T
h
i
s
s
up
por
t
s
w
h
a
t
w
a
s
s
a
i
d
by
[
18]
,
w
h
o
s
t
a
te
s
th
a
t
t
he
S
V
R
m
o
de
l
i
s
s
u
it
a
bl
e
f
or
l
on
g
-
t
e
r
m
f
or
e
c
a
s
ti
ng
w
i
th
a
lo
t
of
s
m
a
ll
d
a
ta
.
T
hi
s
f
in
di
ng
i
s
a
l
s
o
in
li
ne
w
it
h
th
e
f
in
di
n
g
s
of
[
2
1]
.
T
h
e
S
V
R
-
G
A
m
o
de
l
i
s
pr
ov
e
n t
o
b
e
a
bl
e
to
pr
e
d
ic
t
o
th
e
r
da
ta
w
e
ll
,
w
hi
c
h
s
ho
w
s
t
h
a
t
it
h
a
s
g
ood
g
e
n
e
r
a
li
z
a
ti
on
a
bi
li
ti
e
s
e
v
e
n
t
hou
gh
a
l
o
t
of
h
is
to
r
i
c
a
l
da
ta
i
s
u
s
e
d
s
p
a
r
i
ngl
y
a
n
d
is
s
om
e
ti
m
e
s
no
n
-
li
ne
a
r
[
21]
,
[
22]
.
C
om
pa
r
is
on
of
a
c
tu
a
l
a
nd
f
or
e
c
a
s
ti
ng
da
ta
in
ot
he
r
a
r
e
a
s
i
s
s
ho
w
n
in
F
ig
ur
e
3.
C
lu
s
te
r
1
is
s
how
n
in
F
ig
ur
e
3(
a
)
,
c
lu
s
te
r
2
in
F
ig
ur
e
3(
b
)
,
a
nd
F
ig
ur
e
3(
c
)
r
e
pr
e
s
e
nt
s
th
e
r
e
s
ul
ts
in
c
lu
s
te
r
3
a
r
e
a
.
F
ig
ur
e
3
s
how
s
th
a
t
th
e
r
e
s
ul
ti
ng
gr
a
ph
is
not
as
good
a
s
th
e
c
om
pa
r
is
on
s
h
ow
n
in
F
ig
ur
e
2.
T
hi
s
c
a
n
be
s
e
e
n
f
r
om
th
e
f
or
e
c
a
s
t
li
ne
,
w
hi
c
h i
s
not
a
bl
e
t
o f
ol
lo
w
t
he
a
c
tu
a
l
da
ta
pa
tt
e
r
n w
e
ll
. T
hi
s
c
ondi
ti
on c
a
n be
c
a
u
s
e
d by the
da
ta
pa
tt
e
r
n
he
r
e
be
in
g
di
f
f
e
r
e
nt
f
r
om
th
e
da
ta
s
e
t
in
th
e
c
lu
s
te
r
a
r
e
a
s
in
S
B
C
it
y
a
nd
M
L
R
e
ge
nc
y.
S
M
R
e
g
e
nc
y
ha
s
a
l
ot
of
z
e
r
o da
ta
w
he
n c
om
pa
r
e
d t
o
S
B
C
it
y
a
nd M
L
R
e
ge
nc
y. T
he
da
ta
r
a
nge
i
s
a
ls
o ve
r
y s
m
a
ll
, na
m
e
ly
onl
y
a
r
ound
0
–
5.
F
r
om
th
e
f
or
e
c
a
s
ti
ng
r
e
s
ul
ts
f
or
S
M
R
e
ge
nc
y
,
it
c
a
n
be
s
a
id
th
a
t
th
e
S
V
R
-
G
A
a
lg
or
it
hm
is
not
s
ui
ta
bl
e
w
h
e
n
a
ppl
ie
d
to
d
a
ta
th
a
t
is
c
ha
r
a
c
te
r
iz
e
d
by
ha
vi
ng
m
a
ny
z
e
r
o
va
lu
e
s
a
nd
a
s
m
a
ll
da
t
a
r
a
nge
.
T
hi
s
m
e
a
n
s
t
ha
t
f
or
e
c
a
s
ti
ng f
or
S
M
R
e
ge
nc
y c
a
nnot
be
us
e
d a
s
a
r
e
f
e
r
e
nc
e
f
or
de
c
is
io
n
-
m
a
ki
ng
in
t
he
ne
xt
24
pe
r
io
ds
.
T
he
s
e
f
in
di
ngs
s
uppor
t
w
ha
t
w
a
s
s
ta
t
e
d
by
[
2
5
]
,
w
ho
s
ta
te
s
th
a
t
f
or
e
c
a
s
ti
ng
w
it
h
da
ta
th
a
t
c
ont
a
in
s
a
lo
t
of
z
e
r
os
a
nd
s
m
a
ll
va
lu
e
s
is
di
f
f
ic
ul
t.
A
pa
r
t
f
r
om
th
a
t,
f
o
r
e
c
a
s
ti
ng
w
it
h
m
a
ny
z
e
r
o
va
lu
e
s
is
a
ls
o
di
f
f
ic
ul
t
to
ge
t
good pe
r
f
or
m
a
nc
e
[
3
1
].
A
f
te
r
obt
a
in
in
g
a
m
ode
l
w
i
th
th
e
be
s
t
pa
r
a
m
e
te
r
va
lu
e
s
a
nd
c
om
bi
na
ti
on
of
va
r
ia
bl
e
s
,
e
xpe
r
im
e
nt
s
w
e
r
e
c
a
r
r
ie
d
out
us
in
g
ot
he
r
r
e
gi
ona
l
da
ta
s
e
ts
.
F
or
e
c
a
s
ti
ng
us
in
g
ot
he
r
da
ta
s
e
ts
w
a
s
us
e
d
to
s
e
e
how
r
obus
t
th
e
m
ode
l
th
a
t
ha
d be
e
n c
r
e
a
te
d w
a
s
. T
h
e
t
e
s
t
r
e
s
ul
ts
c
a
n be
s
e
e
n i
n T
a
bl
e
5
. F
r
om
T
a
bl
e
5
,
it
c
a
n be
s
e
e
n t
ha
t
s
e
ve
r
a
l
S
M
A
P
E
va
lu
e
s
a
r
e
a
bove
50%
.
H
ow
e
v
e
r
,
th
e
S
M
A
P
E
va
lu
e
c
a
nnot
b
e
c
om
pl
e
te
ly
us
e
d
a
s
a
be
nc
hm
a
r
k
in
m
e
a
s
ur
in
g
pe
r
f
or
m
a
nc
e
.
I
t
m
us
t
a
ls
o
be
s
e
e
n
in
te
r
m
s
of
gr
a
phs
a
nd
th
e
M
S
E
va
lu
e
a
s
a
c
ons
id
e
r
a
ti
on.
A
ll
te
s
ts
pr
oduc
e
ve
r
y
good
M
S
E
va
lu
e
s
a
nd
a
ls
o
s
how
e
xc
e
ll
e
nt
gr
a
phi
c
s
.
E
xa
m
pl
e
s
of
c
om
pa
r
is
on
r
e
s
ul
ts
of
a
c
tu
a
l
a
nd
f
or
e
c
a
s
ti
ng
da
ta
f
or
ot
he
r
r
e
gi
ons
in
e
a
c
h
c
lu
s
te
r
a
r
e
s
how
n
in
F
ig
ur
e
3
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
, V
ol
.
14
, N
o.
4
,
A
ugus
t
20
25
:
2909
-
2921
2916
F
ig
ur
e
3
s
how
s
th
a
t
th
e
be
s
t
m
ode
l
f
or
e
a
c
h
c
lu
s
te
r
is
r
obus
t.
T
hi
s
is
s
how
n
in
th
e
f
or
e
c
a
s
ti
ng
r
e
s
ul
t
s
gr
a
ph
,
w
hi
c
h
s
how
s
a
p
a
tt
e
r
n
s
im
il
a
r
to
th
e
a
c
tu
a
l
da
ta
,
in
c
lu
di
ng
c
l
us
te
r
3
s
how
n
in
F
ig
ur
e
s
3
(
c
)
.
T
hi
s
c
ondi
ti
on
in
c
r
e
a
s
in
gl
y
s
how
s
th
a
t
th
e
S
V
R
-
G
A
m
ode
l
is
c
a
pa
bl
e
of
w
or
ki
ng
w
it
h
ot
he
r
da
ta
,
w
hi
c
h
s
how
s
th
a
t
it
ha
s
good ge
ne
r
a
li
z
a
ti
on a
bi
li
ti
e
s
, e
ve
n t
hough a
l
ot
of
hi
s
to
r
ic
a
l
da
t
a
i
s
us
e
d
. T
hi
s
s
uppor
ts
t
he
s
ta
te
m
e
nt
m
a
de
by
[
9]
,
[
21]
,
[
22]
.
T
o
de
m
ons
tr
a
te
w
he
th
e
r
th
e
pr
opos
e
d
m
ode
l
is
c
a
pa
bl
e
of
ge
ne
r
a
ti
ng
s
upe
r
io
r
f
or
e
c
a
s
ti
ng
c
om
pa
r
e
d
to
ot
he
r
s
,
S
V
R
-
G
A
is
c
om
pa
r
e
d
w
it
h
r
e
gr
e
s
s
io
n
m
e
th
ods
pr
e
vi
ous
ly
e
m
pl
oye
d.
T
h
e
be
nc
hm
a
r
k
m
e
th
ods
us
e
d
f
ol
lo
w
th
os
e
ut
il
iz
e
d
in
pr
io
r
r
e
s
e
a
r
c
h
,
na
m
e
ly
S
V
M
[
9]
,
[
31]
,
a
r
ti
f
ic
ia
l
ne
ur
a
l
ne
twor
k
(
ANN
)
[
31]
,
li
ne
a
r
r
e
gr
e
s
s
io
n
(
L
R
)
[
9]
,
s
to
c
ha
s
ti
c
gr
a
di
e
nt
d
e
s
c
e
nt
(
S
G
D
)
,
gr
a
di
e
nt
boos
ti
ng
(
G
B
)
,
a
nd
A
da
B
oos
t
.
T
he
r
e
s
ul
ts
of
t
he
pe
r
f
or
m
a
nc
e
c
om
pa
r
is
on
a
r
e
pr
e
s
e
nt
e
d i
n
T
a
b
le
5.
(
a
)
(
b)
(
c
)
F
ig
ur
e
2.
C
om
p
a
r
i
s
on
of
a
c
tu
a
l
a
nd
f
or
e
s
ti
n
g
da
ta
f
or
t
h
e
ne
xt
24
pe
r
i
od
s
,
(
a
)
S
B
C
i
ty
,
(
b)
M
L
r
e
g
e
n
c
y
a
nd
(
c
)
S
M
r
e
ge
nc
y
T
a
bl
e
5. P
e
r
f
or
m
a
nc
e
c
om
pa
r
is
on of
t
he
pr
opos
e
d m
ode
l
w
it
h o
th
e
r
s
C
l
us
t
e
r
D
a
t
a
s
e
t
S
V
R
-
GA
S
V
M
S
G
D
ANN
LR
GB
A
da
B
oos
t
M
S
E
S
M
A
P
E
(%)
M
S
E
S
M
A
P
E
(%)
M
S
E
S
M
A
P
E
(%)
M
S
E
S
M
A
P
E
(%)
M
S
E
S
M
A
P
E
(%)
M
S
E
S
M
A
P
E
(%)
M
S
E
S
M
A
P
E
(%)
1
SB
0.218
6.420
18.392
99.655
22.828
99.744
22.828
99.708
17.053
99.736
14.143
90.987
16.167
42.421
BT
0.202
21.851
10.805
82.188
7.706
99.701
12.206
97.955
9.184
99.627
8.194
92.303
8.333
51.349
TL
0.001
29.791
1.657
85.925
2.297
88.975
2.698
94.600
2.160
90.216
2.588
95.387
2.222
59.775
BK
0.000
29.582
12.996
77.443
9.775
86.039
8.893
87.700
9.591
88.937
2.872
86.870
3.556
62.379
2
ML
0.000
23.530
5.222
99.524
5.260
88.674
5.175
86.688
5.295
99.780
8.632
89.278
8.167
63.889
BU
0.000
63.383
0.659
87.768
0.908
71.275
0.883
69.473
0.970
71.000
3.407
80.414
1.167
51.852
SD
0.000
14.090
4.169
99.754
5.543
99.438
6.538
99.833
5.714
99.831
6.260
99.800
6.444
40.444
TB
0.000
56.342
1.977
80.319
1.785
88.643
1.683
78.685
1.792
88.628
2.021
82.664
2.778
46.296
3
SM
0.000
52.940
1.012
79.291
1.693
80.249
1.042
84.514
1.705
81.089
0.890
52.382
1.222
57.407
BJ
0.008
64.229
1.177
55.060
0.982
77.795
1.236
77.047
1.007
64.945
0.218
77.514
0.500
46.296
MG
0.003
67.890
0.379
99.924
0.356
85.870
0.341
95.610
0.382
86.261
0.447
70.346
0.556
38.889
TG
0.000
61.975
0.621
74.826
0.539
81.821
1.013
85.245
0.551
81.879
1.162
57.557
0.944
43.519
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
D
at
a
-
dr
iv
e
n s
uppor
t
v
e
c
to
r
r
e
gr
e
s
s
io
n
-
ge
ne
ti
c
al
gor
it
hm
m
ode
l
f
or
pr
e
di
c
ti
ng
…
(
W
iw
ik
A
nggr
ae
ni
)
2917
(
a
)
(
b)
(
c
)
F
ig
ur
e
3. C
om
pa
r
is
on of
a
c
tu
a
l
a
nd f
or
e
c
a
s
ti
ng da
ta
i
n ot
he
r
a
r
e
a
s
i
n e
a
c
h
c
lu
s
te
r
(
a
)
c
lu
s
te
r
1
,
(
b
)
c
lu
s
te
r
2
,
a
nd
(
c
)
c
lu
s
te
r
3
T
a
bl
e
5
s
how
s
th
a
t
S
V
R
-
G
A
ha
s
be
tt
e
r
pe
r
f
or
m
a
nc
e
c
om
pa
r
e
d
to
ot
he
r
a
lg
or
it
hm
s
.
I
n
te
r
m
s
o
f
M
S
E
,
S
V
R
-
G
A
ha
s
a
s
m
a
ll
e
r
M
S
E
th
a
n
th
e
ot
he
r
s
in
a
ll
r
e
gi
o
ns
.
H
ow
e
ve
r
,
if
w
e
lo
ok
a
t
it
f
r
om
an
M
A
P
E
pe
r
s
pe
c
ti
ve
,
S
V
R
-
G
A
is
s
ti
ll
be
tt
e
r
th
a
n
th
e
ot
he
r
s
,
f
or
s
e
v
e
r
a
l
r
e
gi
ons
e
xc
e
pt
f
or
r
e
gi
ons
in
c
lu
s
te
r
3.
H
ow
e
ve
r
,
ove
r
a
ll
,
S
V
R
-
G
A
c
a
n
s
ti
ll
be
s
a
id
to
ha
ve
be
tt
e
r
pe
r
f
or
m
a
nc
e
th
a
n
th
e
ot
he
r
s
.
T
hi
s
is
s
how
n
in
th
e
r
e
s
ul
ts
of
th
e
two
-
m
e
a
ns
c
om
pa
r
is
on
te
s
t
s
how
n
in
T
a
bl
e
s
6
a
nd
7
.
I
n
-
de
pt
h
hypothe
s
is
te
s
ti
ng
i
s
ne
e
de
d
to
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
, V
ol
.
14
, N
o.
4
,
A
ugus
t
20
25
:
2909
-
2921
2918
pr
ove
th
a
t
th
e
m
ode
l
pe
r
f
or
m
a
nc
e
di
s
pl
a
ye
d
in
num
e
r
ic
a
l
f
or
m
a
c
tu
a
ll
y
ha
s
be
tt
e
r
pe
r
f
or
m
a
nc
e
or
vi
c
e
ve
r
s
a
[2
5
]
.
T
hi
s
2
m
e
a
ns
c
om
pa
r
is
on
te
s
t
w
a
s
c
a
r
r
ie
d
out
on
M
S
E
a
nd
M
A
P
E
f
or
S
V
R
-
G
A
w
it
h
th
e
c
om
pa
r
is
on
m
e
th
od
w
it
h
th
e
be
s
t
pe
r
f
or
m
a
nc
e
.
T
he
nul
l
hypothe
s
is
s
how
s
t
ha
t
S
V
R
-
G
A
is
not
be
tt
e
r
th
a
n
ot
he
r
m
e
th
ods
(H
0
:
M
S
E
S
V
R
-
GA
≥
M
S
E
be
s
t
ot
he
r
m
e
t
hod
)
a
nd
th
e
c
om
pe
ti
ng
hypothe
s
is
s
how
s
th
a
t
S
V
R
-
G
A
is
be
tt
e
r
th
a
n
ot
he
r
m
e
th
ods
(
H
1
:
M
S
E
S
V
R
-
GA
<
M
S
E
be
s
t
ot
he
r
m
e
t
hod
)
.
T
h
e
c
onf
id
e
nc
e
l
e
ve
l
us
e
d
i
s
95%
.
I
f
m
ode
le
d
m
a
th
e
m
a
ti
c
a
ll
y
,
it
be
c
om
e
s
a
s
s
how
n i
n
(
9
)
a
nd
(
10
)
f
or
M
S
E
a
nd (
11)
a
nd (
12
)
f
or
S
M
A
P
E
.
0
:
−
≥
ℎ
ℎ
(
9)
1
:
−
<
ℎ
ℎ
(
10)
0
:
−
≥
ℎ
ℎ
(
11)
1
:
−
<
ℎ
ℎ
(
12)
T
a
bl
e
6.
T
w
o
m
e
a
ns
c
om
pa
r
is
on t
e
s
t
r
e
s
ul
t
ba
s
e
d on
M
S
E
C
l
us
t
e
r
R
e
gi
on
M
S
E
S
t
d. D
e
v. M
S
E
t
-
va
l
ue
C
r
i
t
i
c
a
l
va
l
ue
S
t
a
t
us
S
V
R
-
GA
B
e
s
t
c
om
pa
r
i
s
on
m
e
t
h
od
S
V
R
-
GA
B
e
s
t
c
om
pa
r
i
s
on
m
e
t
h
od
1
SB
0.218
14.143
0.900
21.473
-
2.749
-
1.734
H
o R
e
j
e
c
t
e
d
BT
0.202
7.706
0.000
10.918
-
2.916
-
1.734
H
o R
e
j
e
c
t
e
d
TL
0.001
1.657
0.000
2.280
-
3.083
-
1.734
H
o R
e
j
e
c
t
e
d
BK
0.000
2.872
0.000
4.334
-
2.811
-
1.734
H
o R
e
j
e
c
t
e
d
2
ML
0.000
5.175
0.000
10.640
-
2.063
-
1.740
H
o R
e
j
e
c
t
e
d
BU
0.000
0.659
0.000
2.246
-
1.245
-
1.740
H
o R
e
j
e
c
t
e
d
SD
0.000
4.169
0.000
9.489
-
1.864
-
1.740
H
o R
e
j
e
c
t
e
d
TB
0.000
1.683
0.000
4.417
-
1.616
-
1.740
H
o
R
e
j
e
c
t
e
d
3
SM
0.000
1.012
0.000
2.209
-
1.943
-
1.740
H
o R
e
j
e
c
t
e
d
BJ
0.008
0.218
0.000
2.215
-
0.402
-
1.734
H
o R
e
j
e
c
t
e
d
MG
0.003
0.341
0.000
0.983
-
1.460
-
1.734
H
o R
e
j
e
c
t
e
d
TG
0.000
0.551
0.000
1.274
-
1.835
-
1.740
H
o R
e
j
e
c
t
e
d
T
a
bl
e
7.
T
w
o
m
e
a
ns
c
om
pa
r
is
on t
e
s
t
r
e
s
ul
t
ba
s
e
d
-
on S
M
A
P
E
C
l
us
t
e
r
R
e
gi
on
S
M
A
P
E
S
t
d. D
e
v. S
M
A
P
E
t
-
va
l
ue
C
r
i
t
i
c
a
l
va
l
ue
S
t
a
t
us
S
V
R
-
GA
B
e
s
t
c
om
pa
r
i
s
on
m
e
t
h
od
S
V
R
-
GA
B
e
s
t
c
om
pa
r
i
s
on
m
e
t
hod
1
SB
6.420
42.421
0.242
24.667
-
6.192
-
1.734
H
o R
e
j
e
c
t
e
d
BT
21.851
51.349
0.382
36.189
-
3.458
-
1.714
H
o R
e
j
e
c
t
e
d
TL
29.791
59.775
0.500
43.139
-
2.949
-
1.708
H
o R
e
j
e
c
t
e
d
BK
29.582
62.379
0.461
41.697
-
3.337
-
1.708
H
o R
e
j
e
c
t
e
d
2
ML
23.530
63.889
0.437
41.290
-
4.147
-
1.717
H
o R
e
j
e
c
t
e
d
BU
63.383
51.852
0.485
50.127
0.976
-
1.684
H
o N
ot
R
e
j
e
c
t
e
d
SD
14.090
40.444
0.323
36.141
-
3.094
-
1.717
H
o R
e
j
e
c
t
e
d
TB
56.342
46.296
0.485
50.018
0.852
-
1.684
H
o N
ot
R
e
j
e
c
t
e
d
3
SM
52.940
57.407
0.514
49.581
-
0.382
-
1.684
H
o R
e
j
e
c
t
e
d
BJ
64.229
46.296
0.514
50.018
1.521
-
1.684
H
o N
ot
R
e
j
e
c
t
e
d
MG
67.890
38.889
0.502
50.163
2.453
-
1.701
H
o N
ot
R
e
j
e
c
t
e
d
TG
61.975
43.519
0.461
48.216
1.624
-
1.684
H
o N
ot
R
e
j
e
c
t
e
d
T
a
bl
e
s
6
a
nd
7
s
how
th
a
t
H
0
is
r
e
je
c
te
d
in
a
ll
r
e
gi
ons
.
T
hi
s
s
how
s
th
a
t
th
e
pe
r
f
or
m
a
nc
e
of
S
V
R
-
G
A
is
be
tt
e
r
th
a
n
ot
he
r
m
e
th
ods
.
H
ow
e
ve
r
,
th
is
is
di
f
f
e
r
e
nt
f
r
om
w
ha
t
is
s
how
n
in
T
a
bl
e
7
,
w
he
r
e
pe
r
f
or
m
a
nc
e
is
s
e
e
n
f
r
om
S
M
A
P
E
.
F
or
s
e
ve
r
a
l
a
r
e
a
s
,
s
uc
h
a
s
B
T
in
c
lu
s
te
r
2,
th
e
n
B
J
,
M
G
,
a
nd
T
G
in
c
lu
s
te
r
3,
th
e
pe
r
f
or
m
a
nc
e
of
S
V
R
-
G
A
is
no
be
tt
e
r
th
a
n
ot
he
r
m
e
th
ods
.
T
hi
s
is
in
di
c
a
te
d
by
th
e
s
ta
tu
s
H
0
,
w
hi
c
h
is
not
r
e
je
c
te
d,
w
hi
c
h
m
e
a
ns
th
a
t
ot
he
r
m
e
th
ods
ha
ve
th
e
s
a
m
e
or
be
tt
e
r
pe
r
f
or
m
a
nc
e
th
a
n
S
V
R
-
G
A
.
T
he
be
tt
e
r
c
ondi
ti
on
of
S
V
R
-
G
A
c
om
pa
r
e
d
to
ot
he
r
s
,
e
s
pe
c
ia
ll
y
tr
a
di
t
io
na
l
S
V
M
a
nd
N
N
,
s
uppor
ts
th
e
f
in
di
ngs
pr
e
s
e
nt
e
d by
[
22]
,
[
24]
.
4.
C
O
N
C
L
U
S
I
O
N
F
or
e
c
a
s
ti
ng
th
e
num
be
r
of
di
pht
he
r
ia
s
uf
f
e
r
e
r
s
is
n
e
e
de
d
in
a
n
e
f
f
or
t
to
pr
e
pa
r
e
r
e
la
te
d
c
ont
r
ol
m
e
a
s
ur
e
s
.
T
h
e
da
ta
us
e
d
in
th
is
s
tu
dy
is
r
e
la
ti
ve
ly
s
m
a
ll
,
s
o
it
r
e
qui
r
e
s
a
n
a
ppr
oa
c
h
th
a
t
is
a
bl
e
to
ha
ndl
e
th
is
Evaluation Warning : The document was created with Spire.PDF for Python.