I
AE
S In
t
er
na
t
io
na
l J
o
urna
l o
f
Art
if
icia
l In
t
ellig
ence
(
I
J
-
AI
)
Vo
l.
14
,
No
.
6
,
Dec
em
b
er
2
0
2
5
,
p
p
.
4
4
2
7
~
4
4
4
1
I
SS
N:
2
2
5
2
-
8
9
3
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijai.v
14
.i
6
.
p
p
4
4
2
7
-
4
4
4
1
4427
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
i
.
ia
esco
r
e.
co
m
A review
on lo
ng
sho
rt
-
ter
m memo
ry
com
bina
tion d
ev
elo
pment
Ahm
a
d Riy
a
di
1,
2
,
Nur
Ro
k
h
m
a
n
1
,
L
uk
m
a
n H
er
y
a
wa
n
1
1
D
e
p
a
r
t
me
n
t
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
a
n
d
E
l
e
c
t
r
o
n
i
c
s,
F
a
c
u
l
t
y
o
f
M
a
t
h
e
ma
t
i
c
s
a
n
d
N
a
t
u
r
a
l
S
c
i
e
n
c
e
s
,
U
n
i
v
e
r
s
i
t
a
s Ga
d
j
a
h
M
a
d
a
,
Y
o
g
y
a
k
a
r
t
a
,
I
n
d
o
n
e
si
a
2
D
e
p
a
r
t
me
n
t
o
f
I
n
f
o
r
mat
i
c
s
,
F
a
c
u
l
t
y
o
f
S
c
i
e
n
c
e
a
n
d
T
e
c
h
n
o
l
o
g
y
,
U
n
i
v
e
r
si
t
a
s PG
R
I
Y
o
g
y
a
k
a
r
t
a
,
Y
o
g
y
a
k
a
r
t
a
,
I
n
d
o
n
e
s
i
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Oct
26
,
2
0
2
3
R
ev
is
ed
Sep
13
,
2
0
2
5
Acc
ep
ted
Oct
16
,
2
0
2
5
Lo
n
g
sh
o
rt
-
term
m
e
m
o
ry
(LS
TM
)
h
a
s
c
o
n
ti
n
u
e
d
t
o
d
e
v
e
l
o
p
si
n
c
e
it
wa
s
p
ro
p
o
se
d
i
n
1
9
9
7
.
LS
TM
h
a
s
o
p
t
imiz
e
d
so
lu
ti
o
n
s
t
o
v
a
ri
o
u
s
p
r
o
b
l
e
m
s.
Th
e
LS
TM
c
e
ll
,
a
rc
h
it
e
c
tu
re
,
a
n
d
m
e
m
o
ry
m
o
d
e
l
h
a
v
e
b
e
e
n
re
v
iew
e
d
.
A
re
v
iew
o
f
LS
TM
imp
lem
e
n
tatio
n
h
a
s
b
e
e
n
c
a
rried
o
u
t
in
v
a
ri
o
u
s
p
ro
b
lem
d
o
m
a
in
s
.
Th
e
re
a
re
c
o
m
b
in
a
ti
o
n
s
o
f
LS
T
M
with
o
t
h
e
r
m
e
th
o
d
s
t
o
o
p
ti
m
iz
e
so
lu
ti
o
n
s.
Ho
we
v
e
r,
th
e
re
is
n
o
re
v
iew
o
n
th
e
d
e
v
e
lo
p
m
e
n
t
o
f
L
S
TM
c
o
m
b
in
a
ti
o
n
(LC).
Th
is
re
se
a
rc
h
re
v
iew
s
th
e
d
e
v
e
lo
p
m
e
n
t
o
f
t
h
e
LC
m
o
d
e
l
o
n
n
i
n
e
re
se
a
rc
h
q
u
e
stio
n
s,
n
a
m
e
ly
:
d
e
v
e
lo
p
m
e
n
t
fra
m
e
wo
rk
,
d
a
ta,
p
re
p
ro
c
e
ss
in
g
,
lea
rn
in
g
p
ro
c
e
ss
,
tas
k
s,
o
p
ti
m
iz
a
ti
o
n
a
n
d
e
v
a
l
u
a
ti
o
n
,
d
o
m
a
in
p
ro
b
lem
s,
tren
d
s,
a
n
d
c
h
a
ll
e
n
g
e
s.
Th
e
re
su
l
ts
sh
o
w
th
a
t
th
e
LC
m
o
d
e
l
is
in
c
re
a
sin
g
l
y
wid
e
sp
re
a
d
i
n
s
o
lv
in
g
p
ro
b
lem
s.
Th
e
LC
m
o
d
e
l
h
a
s
c
o
m
p
lete
d
2
6
t
y
p
e
s
o
f
tas
k
s.
P
re
d
ictio
n
,
d
e
tec
ti
o
n
,
fo
re
c
a
stin
g
,
c
las
sifica
ti
o
n
,
a
n
d
re
c
o
g
n
it
io
n
a
re
th
e
m
o
st
fre
q
u
e
n
tl
y
p
e
rf
o
rm
e
d
t
a
sk
s.
LC
m
o
d
e
l
d
e
v
e
lo
p
m
e
n
t
tre
n
d
s
sh
o
w
th
a
t
LS
TM
is
i
n
c
re
a
sin
g
ly
c
o
ll
a
b
o
ra
ti
v
e
wi
th
o
t
h
e
r
m
e
th
o
d
s
o
n
a
wi
d
e
r
sc
o
p
e
.
Th
e
c
h
a
ll
e
n
g
e
s
id
e
n
t
ifi
e
d
in
c
lu
d
e
re
se
a
rc
h
a
re
a
s,
d
a
t
a
,
m
o
d
e
l
d
e
v
e
lo
p
m
e
n
ts,
t
h
e
a
re
a
o
f
imp
lem
e
n
tatio
n
,
p
e
rf
o
rm
a
n
c
e
,
a
n
d
e
fficie
n
c
y
.
K
ey
w
o
r
d
s
:
C
o
m
b
in
atio
n
Dee
p
lear
n
in
g
L
o
n
g
s
h
o
r
t
-
ter
m
m
e
m
o
r
y
Op
tim
izatio
n
Sy
s
tem
atic
liter
atu
r
e
r
ev
iew
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
Nu
r
R
o
k
h
m
an
Dep
ar
tm
en
t o
f
C
o
m
p
u
ter
Scie
n
ce
an
d
E
lectr
o
n
ics,
Facu
lty
o
f
Ma
th
em
atics a
n
d
Natu
r
al
Sci
en
ce
s
Un
iv
er
s
itas
Gad
jah
Ma
d
a
B
u
ild
in
g
C
,
4
th
Flo
o
r
,
Sek
ip
Utar
a,
B
u
lak
s
u
m
u
r
,
Yo
g
y
ak
a
r
ta
5
5
2
8
1
,
I
n
d
o
n
esia
E
m
ail: n
u
r
r
o
k
h
m
a
n
@
u
g
m
.
ac
.
i
d
1.
I
NT
RO
D
UCT
I
O
N
L
o
n
g
s
h
o
r
t
-
ter
m
m
em
o
r
y
(
L
STM
)
is
a
r
ec
u
r
r
en
t
co
n
n
ec
tio
n
n
etwo
r
k
a
r
ch
itectu
r
e
th
at
en
ab
les
u
p
d
atin
g
th
e
cu
r
r
en
t
s
tate
b
ased
o
n
p
ast
s
tates
an
d
cu
r
r
en
t
in
p
u
t
d
ata.
L
STM
is
a
n
ew
m
e
th
o
d
to
ad
d
r
ess
th
e
wea
k
n
ess
o
f
r
ec
u
r
r
en
t
n
e
u
r
al
n
etwo
r
k
(
R
NN)
.
L
STM
co
n
s
is
ts
o
f
m
em
o
r
y
,
in
p
u
t
g
ate,
f
o
r
g
et
g
ate,
an
d
o
u
tp
u
t
g
ate.
L
STM
s
ca
n
b
e
s
tack
e
d
to
cr
ea
te
d
ee
p
L
STM
n
etwo
r
k
s
th
at
ca
n
lear
n
m
o
r
e
co
m
p
l
ex
s
eq
u
en
tial
d
ata.
L
STM
is
ca
p
ab
le
o
f
lear
n
i
n
g
m
o
r
e
th
an
1
0
0
0
s
tep
s
in
ad
v
an
ce
[
1
]
.
T
wo
r
ev
iews
o
n
L
STM
ar
ch
it
ec
tu
r
e
f
o
cu
s
o
n
L
STM
ce
lls
an
d
L
STM
co
m
p
o
n
en
t.
T
h
e
L
STM
ce
ll
r
ev
iew
aim
s
to
ex
p
lo
r
e
th
e
le
ar
n
in
g
ca
p
ac
ity
in
d
ea
lin
g
wi
th
lo
n
g
-
ter
m
d
ep
e
n
d
en
c
y
p
r
o
b
lem
s
.
T
h
is
r
ev
iew
h
as
f
o
u
n
d
t
h
at
n
o
L
STM
v
ar
ia
n
t
o
u
tp
er
f
o
r
m
s
in
all
asp
ec
ts
[
2
]
.
T
h
e
L
STM
co
m
p
o
n
e
n
t
r
ev
iew
f
o
u
n
d
it
ca
n
b
e
ap
p
lied
to
in
ter
esti
n
g
task
s
,
in
clu
d
in
g
tex
t
r
ec
o
g
n
itio
n
,
tim
e
s
er
ies
f
o
r
ec
asti
n
g
,
n
atu
r
al
la
n
g
u
ag
e
p
r
o
ce
s
s
in
g
,
co
m
p
u
ter
v
is
io
n
,
tex
t,
im
ag
es,
an
d
v
id
e
o
.
T
h
is
r
e
v
iew
f
o
u
n
d
th
at
th
e
co
m
b
i
n
atio
n
o
f
L
STM
a
n
d
a
co
n
v
o
l
u
tio
n
al
n
e
u
r
al
n
etwo
r
k
(
C
NN
)
ca
n
im
p
r
o
v
e
s
y
s
tem
p
e
r
f
o
r
m
a
n
ce
o
p
tim
izatio
n
[
3
]
.
T
wo
r
ev
iews
o
n
L
STM
ap
p
li
ca
tio
n
f
o
cu
s
o
n
s
to
ck
m
ar
k
et
p
r
ed
ictio
n
a
n
d
an
o
m
aly
d
ete
ctio
n
.
T
h
e
L
STM
ap
p
licatio
n
in
s
to
ck
m
ar
k
et
p
r
ed
ictio
n
r
ev
iew
s
h
o
ws
th
at
L
STM
p
lay
s
an
im
p
o
r
tan
t
r
o
le
in
s
to
ck
m
ar
k
et
f
o
r
ec
asti
n
g
.
T
h
is
r
ev
ie
w
r
ec
o
m
m
en
d
s
th
at
L
STM
s
h
o
u
ld
b
e
co
m
b
in
ed
with
o
th
er
m
eth
o
d
s
to
im
p
r
o
v
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
14
,
No
.
6
,
Dec
em
b
er
20
25
:
4
4
2
7
-
4
4
4
1
4428
ac
cu
r
ac
y
b
y
co
n
s
id
er
in
g
ex
ter
n
al
f
ac
to
r
s
[
4
]
.
T
h
e
L
STM
ap
p
licatio
n
in
an
o
m
aly
d
etec
tio
n
r
ev
iew
s
h
o
ws
th
at
d
if
f
er
en
t a
r
ch
itectu
r
es a
r
e
ca
p
ab
le
o
f
d
etec
tin
g
v
ar
i
o
u
s
co
m
p
lex
an
o
m
alies c
o
n
tex
tu
ally
a
n
d
co
llectiv
ely
[
5
]
.
B
o
th
r
ev
iews
ab
o
v
e
o
n
ly
d
is
cu
s
s
th
e
L
STM
ce
ll
ar
ch
itectu
r
e
its
elf
an
d
h
a
v
e
n
o
t
d
is
cu
s
s
ed
th
e
co
m
b
in
atio
n
a
r
ch
itectu
r
e
o
f
L
STM
with
o
th
er
m
eth
o
d
s
.
T
h
e
y
o
n
ly
d
is
cu
s
s
th
e
im
p
lem
en
tatio
n
o
f
L
STM
in
a
lim
ited
s
co
p
e,
ev
en
th
o
u
g
h
L
STM
h
as
b
ee
n
im
p
lem
en
ted
i
n
a
wid
e
r
an
g
e
o
f
ar
ea
s
.
T
h
is
r
esear
ch
will
r
ev
iew
th
e
d
ev
elo
p
m
en
t
o
f
L
STM
c
o
m
b
in
atio
n
s
(
L
C
)
in
v
ar
i
o
u
s
ar
ea
s
an
d
ar
c
h
itectu
r
es.
I
t
will
b
e
co
n
d
u
cted
in
th
e
f
o
r
m
o
f
a
s
y
s
tem
atic
liter
atu
r
e
r
ev
iew
(
SLR)
t
o
en
s
u
r
e
a
m
o
r
e
d
etailed
a
n
d
f
o
c
u
s
ed
an
aly
s
is
.
T
o
ac
h
iev
e
th
e
m
ain
o
b
jectiv
es,
th
e
f
o
llo
win
g
n
in
e
r
esear
ch
q
u
esti
o
n
s
wer
e
d
ef
in
ed
:
i)
R
Q1
:
wh
at
is
th
e
L
C
m
o
d
el
d
ev
elo
p
m
e
n
t
f
r
am
ewo
r
k
?
;
ii)
R
Q2
:
h
o
w
is
th
e
d
ata
u
s
ed
in
L
C
m
o
d
el
d
ev
elo
p
m
en
t?
;
iii)
R
Q3
:
h
o
w
i
s
p
r
ep
r
o
ce
s
s
in
g
u
s
ed
in
L
C
m
o
d
el
d
ev
elo
p
m
en
t?
;
iv
)
R
Q4
:
wh
at
is
th
e
lear
n
i
n
g
p
r
o
ce
s
s
in
L
C
m
o
d
el
d
ev
elo
p
m
e
n
t?
;
v
)
R
Q5
:
h
o
w
t
o
o
p
tim
ize
an
d
ev
alu
ate
L
C
m
o
d
el
d
ev
elo
p
m
en
t?
;
v
i)
R
Q6
:
wh
at
task
s
d
o
es
L
C
m
o
d
el
d
ev
elo
p
m
en
t
p
e
r
f
o
r
m
?
;
v
ii)
R
Q7
:
wh
at
p
r
o
b
lem
s
d
o
es
th
e
d
ev
elo
p
m
e
n
t
o
f
th
e
L
C
m
o
d
el
s
o
lv
e?
;
v
iii)
R
Q8
:
wh
at
i
s
th
e
r
ec
en
t
tr
en
d
in
th
e
d
ev
elo
p
m
en
t
o
f
L
C
m
o
d
els?
;
an
d
ix
)
R
Q9
:
wh
at
ar
e
th
e
r
ec
en
t
ch
allen
g
es in
L
C
m
o
d
el
d
ev
el
o
p
m
en
t?
2.
M
E
T
H
O
D
T
h
is
r
esea
r
c
h
h
as
s
ev
e
n
s
te
p
s
,
n
a
m
el
y
f
o
r
m
u
la
ti
n
g
t
h
e
p
r
o
b
le
m
,
s
ea
r
c
h
i
n
g
t
h
e
li
te
r
at
u
r
e
,
s
c
r
ee
n
i
n
g
f
o
r
in
c
lu
s
io
n
,
ass
ess
in
g
q
u
ali
ty
,
e
x
t
r
a
cti
n
g
d
ata
,
a
n
a
ly
zi
n
g
,
a
n
d
s
y
n
th
esiz
in
g
d
a
ta
as
s
h
o
w
n
i
n
Fi
g
u
r
e
1
[
6
]
,
[
7
]
.
T
e
x
t
cl
ass
if
ic
ati
o
n
is
u
s
e
d
as
th
e
a
n
al
y
s
is
tec
h
n
i
q
u
e
[
8
]
.
Pa
p
e
r
s
we
r
e
o
b
t
ai
n
e
d
th
r
o
u
g
h
a
s
ea
r
c
h
p
r
o
c
ess
o
n
s
ev
er
al
o
n
l
in
e
ac
ad
em
ic
s
ea
r
c
h
e
n
g
in
es
,
s
u
c
h
as
S
c
o
p
u
s
,
S
ci
en
ce
D
ir
ec
t
,
I
E
E
E
X
p
l
o
r
e
,
Sp
r
i
n
g
e
r
L
i
n
k
,
E
m
er
al
d
,
an
d
Pr
o
Q
u
es
t,
u
s
i
n
g
t
h
e
k
e
y
w
o
r
d
s
"
L
ST
M"
wit
h
a
f
ilte
r
f
o
r
2
0
2
3
,
an
d
t
h
e
f
iel
d
o
f
c
o
m
p
u
t
e
r
s
cie
n
ce
.
T
h
e
in
clu
s
io
n
c
r
iter
ia
u
s
ed
in
th
is
r
esear
ch
in
clu
d
e
p
ap
er
s
d
i
s
cu
s
s
in
g
th
e
d
ev
elo
p
m
en
t
o
f
LC
,
p
ap
er
s
co
n
tain
in
g
a
f
r
am
ewo
r
k
f
o
r
d
ev
elo
p
in
g
LC
to
s
o
lv
e
r
esear
ch
p
r
o
b
lem
s
i
n
a
p
ar
ticu
la
r
d
o
m
ain
,
p
ap
er
s
ex
p
lain
in
g
th
e
task
s
ca
r
r
ied
o
u
t
b
y
LC
,
an
d
p
ap
er
s
ex
p
lain
in
g
d
ata
an
d
p
e
r
f
o
r
m
an
ce
.
Me
an
wh
ile,
th
e
ex
clu
s
io
n
cr
iter
ia
u
s
ed
in
th
is
r
esear
ch
ar
e
p
a
p
er
s
co
n
tain
in
g
th
e
w
o
r
d
s
"L
STM
"
o
r
"L
o
n
g
s
h
o
r
t
-
ter
m
m
em
o
r
y
"
in
th
e
titl
e.
T
h
e
p
ap
er
is
a
p
u
b
licatio
n
o
f
r
esear
ch
o
r
ex
p
er
im
en
tal
r
esu
lts
.
T
h
e
s
tag
es
o
f
s
elec
tin
g
p
ap
er
s
b
ased
o
n
in
clu
s
io
n
an
d
ex
clu
s
io
n
cr
iter
ia
ar
e
s
h
o
wn
in
Fig
u
r
e
2
.
Fig
u
r
e
1
.
R
esear
ch
m
eth
o
d
Fig
u
r
e
2
.
Stag
es o
f
s
elec
tin
g
p
ap
er
s
3.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
T
h
is
r
esear
ch
f
o
cu
s
es
o
n
r
ev
i
ewin
g
th
e
LC
in
SLR
f
o
r
m
i
n
a
m
o
r
e
d
etailed
an
d
f
o
cu
s
e
d
m
an
n
e
r
.
Pre
v
io
u
s
r
ev
iews
o
n
ly
c
o
v
er
e
d
L
STM
in
a
lim
ited
s
co
p
e
a
n
d
in
th
e
f
o
r
m
o
f
g
en
e
r
al
r
ev
iews
th
at
wer
e
less
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
A
r
ev
iew
o
n
lo
n
g
s
h
o
r
t
-
term me
mo
r
y
co
mb
in
a
tio
n
d
ev
elo
p
me
n
t (
A
h
ma
d
R
iya
d
i)
4429
f
o
cu
s
ed
an
d
d
etailed
.
T
h
is
r
esear
ch
d
escr
ib
es
th
e
p
ap
er
s
u
s
ed
,
th
e
r
esu
lts
o
f
th
e
L
C
r
e
v
iew,
co
m
p
ar
ativ
e
r
ev
iews,
an
d
f
u
r
t
h
er
r
esear
ch
.
T
h
e
p
ap
er
d
escr
ip
tio
n
is
u
s
e
d
to
i
n
d
icate
th
e
q
u
ality
an
d
q
u
an
tity
o
f
r
esear
ch
s
o
u
r
ce
s
.
T
h
e
r
esu
lts
o
f
th
e
r
e
v
iew
ar
e
u
s
ed
t
o
s
h
o
w
a
d
es
cr
ip
tio
n
o
f
L
C
m
o
d
el
d
e
v
elo
p
m
en
t
f
r
o
m
v
ar
io
u
s
p
r
ed
eter
m
in
e
d
p
o
i
n
ts
o
f
v
iew.
C
o
m
p
ar
ativ
e
r
ev
iews
wer
e
u
s
ed
to
d
em
o
n
s
tr
ate
th
e
r
ec
en
cy
an
d
s
u
p
er
io
r
ity
o
f
th
e
r
ev
iew
ca
r
r
ied
o
u
t
c
o
m
p
a
r
ed
to
p
r
ev
i
o
u
s
r
ev
iews.
Fu
r
th
er
r
esear
ch
is
n
ee
d
ed
to
r
e
f
in
e
th
e
lim
itatio
n
s
o
f
th
is
r
esear
ch
.
3
.
1
.
P
a
pers
des
cr
iptio
n
Pap
er
s
wer
e
co
llected
th
r
o
u
g
h
o
n
lin
e
ac
ad
em
ic
s
ea
r
ch
en
g
in
es,
in
clu
d
in
g
Sco
p
u
s
,
Scien
ce
Dir
ec
t,
I
E
E
E
Xp
lo
r
e,
Sp
r
in
g
er
L
in
k
,
E
m
er
ald
,
an
d
Pro
Qu
est.
Al
l
p
ap
er
s
wer
e
p
u
b
lis
h
ed
in
2
0
2
3
.
T
h
e
p
ap
e
r
s
s
y
n
th
esized
in
th
is
s
tu
d
y
h
a
v
e
g
o
n
e
th
r
o
u
g
h
9
s
tag
es
o
f
s
elec
tio
n
with
in
clu
s
io
n
an
d
ex
clu
s
io
n
cr
iter
ia
.
T
h
e
s
elec
ted
p
ap
e
r
s
wer
e
1
4
6
o
u
t
o
f
4
7
2
,
o
r
3
0
%.
Fig
u
r
e
3
s
h
o
ws
th
at
th
e
s
y
n
t
h
esized
p
ap
er
s
wer
e
p
u
b
lis
h
ed
in
5
5
p
o
p
u
lar
i
n
ter
n
atio
n
al
j
o
u
r
n
als.
Mo
s
t
o
f
th
e
p
a
p
er
s
ar
e
p
u
b
lis
h
ed
in
jo
u
r
n
als
in
th
e
f
ield
o
f
co
m
p
u
ter
s
cien
ce
,
with
o
t
h
er
s
in
th
e
f
ield
o
f
ap
p
lied
c
o
m
p
u
ter
s
cien
ce
.
3
3
% o
f
th
e
p
ap
er
s
wer
e
p
u
b
li
s
h
ed
b
y
jo
u
r
n
als
in
th
e
f
ield
o
f
co
m
p
u
ter
s
cien
c
e
an
d
its
ap
p
licatio
n
s
,
in
clu
d
in
g
I
E
E
E
Acc
ess
,
Sen
s
o
r
s
,
E
n
er
g
ies,
Ap
p
lied
Scien
ce
,
an
d
B
io
m
ed
ical
Sig
n
al
Pro
ce
s
s
in
g
an
d
C
o
n
tr
o
l.
T
h
e
q
u
ality
o
f
s
elec
ted
jo
u
r
n
als a
n
d
p
ap
e
r
p
u
b
lis
h
er
s
s
p
r
ea
d
ac
r
o
s
s
eig
h
t
co
u
n
tr
ies
is
s
h
o
wn
in
Fig
u
r
e
s
4
an
d
5
.
Mo
s
t
o
f
th
e
p
ap
er
s
wer
e
p
u
b
lis
h
ed
b
y
MD
PI,
I
E
E
E
,
E
ls
ev
ier
,
an
d
Sp
r
in
g
er
.
Fig
u
r
e
6
s
h
o
ws
th
at
all
jo
u
r
n
als
o
cc
u
p
y
q
u
a
r
tile
o
n
e
o
r
q
u
ar
tile
2
in
th
e
Scim
ag
o
J
o
u
r
n
al
a
n
d
C
o
u
n
tr
y
R
an
k
.
Fig
u
r
e
3
.
Pu
b
licatio
n
jo
u
r
n
al
o
f
L
C
Fig
u
r
e
4
.
Pu
b
lis
h
er
o
f
L
C
Fig
u
r
e
5
.
Pu
b
licatio
n
co
u
n
tr
y
o
f
L
C
Fig
u
r
e
6
.
J
o
u
r
n
al
q
u
ar
til
e
o
f
L
C
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
14
,
No
.
6
,
Dec
em
b
er
20
25
:
4
4
2
7
-
4
4
4
1
4430
3
.
2
.
Resea
r
ch
re
s
ult
T
h
is
r
esear
ch
aim
s
to
d
escr
ib
e
th
e
d
ev
elo
p
m
e
n
t
o
f
th
e
L
C
m
o
d
el
,
in
clu
d
i
n
g
th
e
f
r
am
e
wo
r
k
,
d
ata,
p
r
ep
r
o
ce
s
s
in
g
,
lear
n
in
g
p
r
o
ce
s
s
,
o
p
tim
izatio
n
,
ev
alu
atio
n
,
t
ask
s
,
p
r
o
b
lem
d
o
m
ain
s
,
tr
e
n
d
s
,
an
d
ch
allen
g
es,
wh
ich
wer
e
n
o
t
d
is
cu
s
s
ed
in
p
r
ev
io
u
s
r
ev
iews.
T
h
e
d
escr
ip
tio
n
o
f
r
esear
ch
r
esu
lts
is
ac
co
m
p
an
ied
b
y
a
d
is
cu
s
s
io
n
to
p
r
esen
t
th
e
r
esear
ch
r
esu
lts
m
o
r
e
co
m
p
r
eh
en
s
i
v
ely
.
T
h
e
d
is
cu
s
s
io
n
s
co
p
e
o
f
th
e
r
esear
ch
r
esu
lts
was
b
ased
o
n
n
in
e
p
r
ed
ete
r
m
i
n
ed
r
esear
ch
q
u
esti
o
n
s
.
3
.
2
.
1
.
RQ
1
:
wha
t
is
t
he
L
C
m
o
del dev
elo
pm
ent
f
r
a
m
ew
o
rk
?
W
e
f
o
u
n
d
t
h
at
L
STM
d
ev
e
lo
p
m
en
t
ca
n
b
e
class
if
ied
i
n
to
two
d
e
v
elo
p
m
e
n
t
m
o
d
els,
n
am
ely
d
ev
elo
p
m
e
n
t
in
L
STM
ce
lls
a
n
d
L
C
m
o
d
el
d
ev
elo
p
m
en
t,
a
s
s
h
o
wn
in
Fig
u
r
e
7
.
L
STM
c
ell
d
ev
elo
p
m
e
n
t
is
ca
r
r
ied
o
u
t
b
y
m
o
d
if
y
i
n
g
th
e
ar
ch
itectu
r
e,
weig
h
ts
,
o
r
f
u
n
ctio
n
s
o
f
th
e
L
STM
ce
ll.
L
C
m
o
d
el
d
ev
elo
p
m
en
t
is
ca
r
r
ied
o
u
t
b
y
co
m
b
in
in
g
L
STM
with
its
elf
o
r
co
m
b
in
in
g
L
STM
with
o
th
er
m
et
h
o
d
s
.
T
h
e
co
m
b
i
n
atio
n
o
f
L
STM
with
o
th
er
m
eth
o
d
s
ca
n
b
e
s
er
ial,
p
ar
allel,
o
r
m
ix
ed
.
Fig
u
r
e
7
.
C
lass
if
icatio
n
o
f
L
C
m
o
d
el
d
e
v
elo
p
m
e
n
t
T
h
e
d
e
v
elo
p
m
e
n
t
o
f
L
STM
c
ell
is
less
th
an
L
C
m
o
d
el.
T
h
e
d
e
v
elo
p
m
e
n
t
o
f
L
STM
ce
lls
aim
s
to
o
p
tim
ize
n
etwo
r
k
m
o
d
els
[
9
]
–
[
1
2
]
a
n
d
h
id
d
e
n
lay
er
o
p
tim
i
za
tio
n
with
ce
r
tain
alg
o
r
ith
m
s
,
in
clu
d
in
g
p
ar
ticle
s
war
m
o
p
tim
izatio
n
(
PSO
)
a
n
d
g
en
etic
alg
o
r
ith
m
[
1
3
]
–
[
1
7
]
o
r
m
o
d
if
y
in
g
n
etwo
r
k
wei
g
h
ts
[
1
8
]
,
[
1
9
]
.
Fo
r
ex
am
p
le,
two
s
ig
m
o
id
f
u
n
ctio
n
s
an
d
th
e
tan
h
f
u
n
ctio
n
in
th
e
L
STM
ce
ll
ar
e
r
ep
lace
d
with
a
s
in
u
s
o
id
al
f
u
n
ctio
n
to
in
cr
ea
s
e
ac
c
u
r
ac
y
in
p
r
o
b
lem
s
wh
o
s
e
o
u
tp
u
t
i
s
p
er
io
d
ic,
a
n
d
th
e
u
s
e
o
f
R
ad
ix
-
r
o
f
f
s
et
b
in
a
r
y
co
d
in
g
(
OB
C
)
as
a
r
ec
u
r
r
en
t
co
n
n
ec
tio
n
weig
h
t
at
ea
ch
g
a
te
o
f
th
e
L
STM
ce
ll
to
in
cr
ea
s
e
th
e
ex
p
o
n
en
tial
g
r
o
wth
o
f
th
e
m
o
d
el
s
ize.
L
C
m
o
d
el
d
e
v
elo
p
m
e
n
t
ca
n
b
e
cl
ass
if
ied
in
to
two
m
o
d
els,
n
am
ely
,
a
co
m
b
in
atio
n
o
f
L
STM
with
its
elf
[
2
0
]
o
r
a
co
m
b
in
atio
n
o
f
L
STM
with
o
t
h
er
m
eth
o
d
s
.
Mo
s
t
L
C
m
o
d
el
d
ev
elo
p
m
e
n
ts
ar
e
a
co
m
b
in
atio
n
o
f
L
STM
an
d
o
n
e
o
r
m
o
r
e
o
th
er
m
et
h
o
d
s
[
2
1
]
–
[
2
3
]
.
T
h
e
L
C
m
o
d
el
f
o
r
m
at
is
s
er
ial
[
2
4
]
–
[
2
7
]
,
p
ar
allel
[
2
8
]
–
[
3
4
]
,
o
r
a
m
i
x
tu
r
e
[
3
5
]
.
T
h
e
L
C
m
o
d
el
d
ev
elo
p
m
en
t
aim
s
to
g
et
b
etter
m
o
d
el
p
er
f
o
r
m
a
n
ce
,
b
u
t
th
er
e
ar
e
an
o
m
alies in
ce
r
tain
ca
s
es
[
3
6
]
.
W
e
f
o
u
n
d
th
e
g
en
er
al
L
C
d
ev
elo
p
m
en
t a
r
c
h
itectu
r
e
as sh
o
wn
in
Fig
u
r
e
8
.
T
h
e
ar
c
h
itectu
r
e
is
d
iv
id
ed
in
to
th
r
ee
s
tag
es:
d
ata
p
r
ep
a
r
atio
n
,
tr
ain
in
g
m
o
d
el,
a
n
d
ev
alu
atio
n
m
o
d
el.
T
h
ese
th
r
ee
s
tag
es
ar
e
r
elate
d
s
er
ially
.
Data
p
r
ep
ar
atio
n
co
n
s
is
ts
o
f
d
ata
co
llectio
n
an
d
d
ata
r
ed
u
ctio
n
.
T
h
e
r
ed
u
ce
d
d
ata
is
s
av
ed
as
a
d
ataset
an
d
r
ea
d
y
to
b
e
u
s
ed
f
o
r
th
e
n
ex
t
p
r
o
ce
s
s
.
C
o
llectin
g
d
ata
ca
n
b
e
ch
allen
g
i
n
g
d
u
e
to
d
if
f
icu
lt
ies
in
f
in
d
in
g
s
o
u
r
ce
s
an
d
ex
t
r
ac
tin
g
d
ata,
as
well
as
d
ea
lin
g
with
d
iv
er
s
e
f
o
r
m
s
an
d
ty
p
es
o
f
d
ata.
T
h
e
tr
ain
in
g
m
o
d
el
c
o
n
s
is
ts
o
f
in
p
u
t
f
ea
tu
r
es,
d
iv
id
in
g
tr
ain
in
g
d
ata
a
n
d
v
alid
atio
n
d
ata
,
an
d
lear
n
in
g
p
r
o
ce
s
s
es.
So
m
e
L
C
m
o
d
el
d
ev
elo
p
m
en
ts
ca
r
r
y
o
u
t
p
r
ep
r
o
ce
s
s
in
g
[
3
7
]
,
[
3
8
]
,
f
ea
tu
r
e
ex
t
r
ac
tio
n
[
3
9
]
–
[
4
1
]
o
r
b
o
th
.
Featu
r
e
en
g
in
ee
r
in
g
is
th
e
p
r
o
ce
s
s
o
f
tr
an
s
f
o
r
m
in
g
r
aw
d
ata
in
to
a
f
o
r
m
at
s
u
itab
le
f
o
r
an
aly
s
is
.
T
h
is
in
v
o
lv
es
ex
tr
a
ctin
g
n
ec
ess
ar
y
f
ea
tu
r
es
u
s
in
g
s
p
ec
if
ic
m
eth
o
d
s
.
T
h
e
d
e
v
elo
p
m
en
t
o
f
th
e
L
C
m
o
d
el
is
f
u
r
th
er
d
em
o
n
s
tr
ated
b
y
th
e
lear
n
in
g
p
r
o
ce
s
s
ar
ch
i
tectu
r
e.
Op
tim
al
m
o
d
el
p
er
f
o
r
m
an
ce
is
o
b
tain
ed
f
r
o
m
r
e
p
ea
ted
tr
ain
in
g
an
d
v
alid
atio
n
,
as
well
as
m
o
d
if
y
in
g
th
e
p
r
o
p
o
r
tio
n
o
f
tr
ain
in
g
d
ata
an
d
v
alid
atio
n
d
ata.
T
r
ain
in
g
a
m
o
d
el
in
v
o
l
v
es
f
in
d
in
g
th
e
b
est
h
y
p
e
r
p
ar
a
m
eter
s
f
o
r
o
p
tim
al
p
er
f
o
r
m
an
ce
.
T
h
e
e
v
alu
atio
n
m
o
d
el
aim
s
to
m
ea
s
u
r
e
t
h
e
p
er
f
o
r
m
a
n
ce
o
f
th
e
L
C
m
o
d
el.
T
h
e
e
v
alu
atio
n
m
eth
o
d
f
o
r
a
m
o
d
el
s
h
o
u
ld
b
e
ap
p
r
o
p
r
iate
f
o
r
its
in
p
u
t,
o
u
tp
u
t,
an
d
task
s
to
u
n
d
e
r
s
tan
d
its
p
er
f
o
r
m
an
ce
.
C
h
o
o
s
in
g
a
b
aselin
e
m
eth
o
d
f
o
r
co
m
p
ar
is
o
n
is
a
k
e
y
ch
allen
g
e
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
A
r
ev
iew
o
n
lo
n
g
s
h
o
r
t
-
term me
mo
r
y
co
mb
in
a
tio
n
d
ev
elo
p
me
n
t (
A
h
ma
d
R
iya
d
i)
4431
Fig
u
r
e
8
.
L
C
d
e
v
elo
p
m
e
n
t f
r
a
m
ewo
r
k
3
.
2
.
2
.
RQ
2
:
ho
w
is
da
t
a
us
ed
in L
C
m
o
del dev
elo
pm
ent
?
W
e
f
o
u
n
d
th
at
th
e
d
ata
u
s
ed
f
o
r
d
ev
elo
p
in
g
th
e
L
C
m
o
d
el
ca
n
b
e
class
if
ied
in
to
th
r
ee
ca
teg
o
r
ies,
n
am
ely
p
u
b
lic
d
atasets
,
o
f
f
icial
d
ata,
an
d
ex
p
er
im
en
tal
d
ata
.
Fig
u
r
e
9
s
h
o
ws
th
e
p
r
o
p
o
r
tio
n
s
o
f
th
e
th
r
ee
d
ata
ca
teg
o
r
ies.
E
x
p
er
im
e
n
tal
d
ata
o
cc
u
p
ies
th
e
h
ig
h
est
p
r
o
p
o
r
t
io
n
.
T
h
is
p
r
o
p
o
r
tio
n
s
h
o
ws
th
at
m
o
s
t
o
f
th
e
L
C
m
o
d
el
d
ev
elo
p
m
en
t
is
b
ased
o
n
r
ea
l
-
wo
r
ld
p
r
o
b
lem
s
an
d
lab
o
r
ato
r
y
test
s
[
4
2
]
–
[
4
5
]
.
Pu
b
li
c
d
ata
u
s
ed
f
o
r
L
C
m
o
d
el
d
ev
elo
p
m
en
t
i
n
clu
d
e
I
E
E
E
b
ea
r
in
g
d
ataset
[
1
1
]
,
th
e
Dee
p
MI
MO
d
ataset
[
4
6
]
,
C
-
MA
PS
S
d
ataset
[
4
7
]
,
B
SL
d
ataset
[
4
8
]
,
a
p
u
b
licly
a
v
ailab
le
d
ataset
f
r
o
m
New
So
u
th
W
ales
[
4
9
]
,
wea
th
e
r
d
atas
et
in
Qu
in
lan
d
[
5
0
]
,
PHM2
0
1
0
to
o
l
-
wea
r
d
ataset
[
5
1
]
,
n
ew
p
lan
t
d
is
ea
s
e
d
ataset
[
5
2
]
,
an
d
I
MD
b
[
5
3
]
.
T
h
e
h
ig
h
p
r
o
p
o
r
tio
n
o
f
o
f
f
icial
d
ata
u
s
ag
e
in
d
icate
s
t
h
at
th
e
d
ev
elo
p
m
en
t
o
f
th
e
L
C
m
o
d
el
s
o
lv
es
in
s
titu
tio
n
al
b
u
s
in
ess
p
r
o
b
lem
s
.
Of
f
icial
d
ata
u
s
ed
f
o
r
L
C
m
o
d
el
d
ev
elo
p
m
en
t
in
clu
d
e
t
h
e
Natio
n
al
Ma
r
in
e
Data
C
en
ter
[
5
4
]
,
e
x
p
er
im
e
n
tal
d
ata
f
r
o
m
th
e
Un
iv
er
s
ity
o
f
C
in
cin
n
ati’
s
I
n
tellig
en
t
Ma
in
ten
an
ce
Sy
s
tem
s
C
en
ter
[
5
5
]
,
air
p
o
llu
tio
n
d
ata,
a
n
d
m
eteo
r
o
lo
g
y
d
ata
f
r
o
m
3
5
m
o
n
ito
r
in
g
s
tatio
n
s
in
B
eijin
g
[
5
6
]
,
p
r
ice
o
f
C
h
in
a
R
ea
l
E
s
tate
I
n
d
ex
[
5
7
]
,
E
p
ilep
s
y
R
esear
ch
C
en
ter
at
B
o
n
n
Un
iv
er
s
ity
in
Ger
m
an
y
[
5
8
]
,
an
d
ac
tu
al
o
p
e
r
atio
n
d
ata
o
f
Den
m
a
r
k
’
s
DK1
r
eg
io
n
in
th
e
No
r
d
ic
elec
tr
icity
m
ar
k
et
[
5
9
]
.
W
e
al
s
o
f
o
u
n
d
t
h
a
t
t
h
e
d
a
ta
u
s
e
d
i
n
L
C
m
o
d
e
l
d
e
v
el
o
p
m
e
n
t
c
a
n
b
e
c
la
s
s
i
f
i
e
d
b
as
e
d
o
n
i
t
s
t
y
p
e
as
s
h
o
w
n
i
n
F
i
g
u
r
e
1
0
.
T
h
e
n
u
m
er
i
c
a
l
d
a
t
a
t
y
p
e
i
s
t
h
e
m
o
s
t
w
i
d
e
l
y
u
s
e
d
,
r
e
a
c
h
i
n
g
1
4
%
[
6
0
]
–
[
6
2
]
.
M
e
a
n
w
h
i
l
e
,
t
h
e
s
i
g
n
a
l
d
at
a
t
y
p
e
[
6
3
]
is
2
0
%
,
a
n
d
t
h
e
i
m
a
g
e
d
a
t
a
t
y
p
e
[
6
4
]
is
1
4
%
.
T
e
x
t
a
n
d
m
u
l
t
i
v
a
r
i
a
t
e
d
a
t
a
t
y
p
es
[
6
5
]
–
[
6
7
]
e
a
c
h
a
c
c
o
u
n
t
f
o
r
1
0
%
.
T
h
e
r
e
a
r
e
o
n
l
y
a
f
e
w
o
t
h
e
r
d
a
t
a
t
y
p
e
s
,
i
n
cl
u
d
i
n
g
v
i
d
e
o
[
2
6
]
,
s
y
m
b
o
l
s
[
6
8
]
,
[
6
9
]
,
a
n
d
s
p
e
e
c
h
[
7
0
]
.
Fig
u
r
e
9
.
T
h
e
d
ata
s
o
u
r
ce
o
f
L
C
Fig
u
r
e
1
0
.
T
h
e
d
ata
ty
p
e
o
f
L
C
3
.
2
.
3
.
RQ
3
:
ho
w
is
prepro
ce
s
s
ing
in
L
C
m
o
del dev
elo
pm
ent
?
M
o
s
t
o
f
t
h
e
d
a
t
a
ca
n
n
o
t
b
e
a
n
a
l
y
z
e
d
d
i
r
e
c
t
l
y
i
n
t
h
e
l
e
a
r
n
i
n
g
p
r
o
c
e
s
s
o
f
L
C
m
o
d
el
d
e
v
e
l
o
p
m
e
n
t
.
W
e
d
is
c
o
v
e
r
e
d
v
a
r
i
o
u
s
m
e
t
h
o
d
s
o
f
p
r
e
p
r
o
c
e
s
s
i
n
g
d
a
t
a
,
as
s
h
o
w
n
i
n
F
i
g
u
r
e
1
1
.
T
h
es
e
m
e
t
h
o
d
s
i
n
c
l
u
d
e
Pe
a
r
s
o
n
c
o
r
r
e
l
a
t
i
o
n
[
4
1
]
,
n
o
r
m
a
l
i
za
t
i
o
n
[
7
1
]
,
i
n
t
e
r
p
o
l
a
t
i
o
n
[
7
2
]
,
l
ab
e
l
i
n
g
[
7
3
]
,
w
a
v
e
l
et
[
7
4
]
,
wo
r
d
2
v
e
c
[
7
5
]
,
a
n
d
c
o
n
v
e
r
t
i
n
g
,
G
l
o
V
e
[
7
6
]
.
T
h
e
n
e
e
d
f
o
r
p
r
e
p
r
o
c
e
s
s
i
n
g
m
e
t
h
o
d
s
i
s
d
e
t
e
r
m
i
n
e
d
b
a
s
e
d
o
n
th
e
d
a
t
a
t
y
p
e
,
d
a
t
a
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
14
,
No
.
6
,
Dec
em
b
er
20
25
:
4
4
2
7
-
4
4
4
1
4432
c
l
e
a
n
i
n
g
,
a
n
d
a
n
a
l
y
s
is
m
e
t
h
o
d
t
o
b
e
u
s
e
d
[
2
7
]
,
[
7
7
]
.
S
o
m
e
L
C
m
o
d
e
l
d
e
v
e
l
o
p
m
e
n
t
s
u
s
e
m
o
r
e
t
h
a
n
o
n
e
p
r
e
p
r
o
c
e
s
s
i
n
g
m
e
t
h
o
d
[
7
8
]
.
N
o
r
m
a
l
i
z
a
ti
o
n
i
s
t
h
e
m
o
s
t
wi
d
e
l
y
u
s
e
d
m
e
t
h
o
d
[
7
9
]
.
S
o
m
e
L
C
m
o
d
e
l
d
e
v
e
l
o
p
m
e
n
t
s
u
s
e
a
c
o
m
b
i
n
a
t
i
o
n
o
f
n
o
r
m
a
l
i
z
a
t
i
o
n
a
n
d
o
t
h
e
r
m
e
t
h
o
d
s
[
8
0
]
,
[
8
1
]
.
P
r
e
p
r
o
c
e
s
s
i
n
g
f
o
r
t
e
x
t
d
a
t
a
u
s
e
s
G
l
o
V
e
[
2
9
]
,
[
7
6
]
,
o
r
w
o
r
d
2
v
e
c
[
8
2
]
,
w
h
i
l
e
p
r
e
p
r
o
c
es
s
i
n
g
f
o
r
s
i
g
n
a
l
d
a
t
a
u
s
es
w
a
v
e
l
et
s
[
3
8
]
,
[
7
4
]
,
o
r
i
n
t
e
r
p
o
l
a
t
i
o
n
[
8
3
]
,
[
8
4
]
.
P
r
e
p
r
o
c
e
s
s
i
n
g
i
n
t
h
e
L
C
-
s
u
p
e
r
v
i
s
e
d
m
o
d
e
l
d
e
v
e
l
o
p
m
e
n
t
m
o
d
e
l
u
s
e
s
d
at
a
l
a
b
el
i
n
g
[
8
5
]
,
[
8
6
]
.
3
.
2
.
4
.
RQ
4
:
wha
t
is
t
he
lea
r
nin
g
pro
ce
s
s
i
n L
C
dev
elo
p
m
ent
?
W
e
f
o
u
n
d
th
at
L
C
m
o
d
el
d
ev
e
lo
p
m
en
t is ca
r
r
ied
o
u
t b
y
co
m
b
in
in
g
L
STM
with
m
o
d
u
les,
alg
o
r
ith
m
s
,
o
r
o
t
h
er
m
eth
o
d
s
to
im
p
r
o
v
e
p
er
f
o
r
m
an
ce
in
p
r
o
b
lem
-
s
o
l
v
in
g
.
Mo
d
u
les
t
h
at
ar
e
o
f
te
n
u
s
ed
i
n
L
C
m
o
d
el
d
ev
elo
p
m
e
n
t
in
clu
d
e
atten
tio
n
[
2
1
]
,
[
2
9
]
,
[
3
5
]
,
R
e
L
u
[
2
2
]
,
[
2
3
]
,
p
o
o
lin
g
[
2
4
]
,
[
3
0
]
,
S
o
f
tMa
x
[
2
6
]
,
[
3
9
]
,
co
n
d
itio
n
al
r
a
n
d
o
m
f
iel
d
(
C
R
F
)
[
2
9
]
,
d
r
o
p
o
u
t
[
3
0
]
,
[
8
0
]
,
f
l
atten
[
3
1
]
,
[
5
2
]
,
f
u
lly
co
n
n
ec
t
ed
[
4
7
]
,
[
5
9
]
,
d
en
s
e
[
5
0
]
,
[
7
8
]
.
Me
th
o
d
s
th
at
ar
e
o
f
ten
co
m
b
in
ed
with
L
STM
in
clu
d
e
C
NN
[
2
1
]
,
[
2
4
]
,
[
2
8
]
,
R
NN
[
6
9
]
,
t
r
an
s
f
o
r
m
er
[
4
2
]
,
[
6
4
]
,
g
ate
d
r
ec
u
r
r
en
t
u
n
it
(
GR
U
)
,
an
d
L
STM
its
elf
[
2
0
]
,
[
3
2
]
,
[
5
6
]
.
Bi
L
STM
an
d
Stack
ed
B
iLST
M
ar
e
ex
am
p
le
s
o
f
LC
m
eth
o
d
s
th
em
s
elv
es
[
8
7
]
.
M
u
lti
-
lay
er
s
eq
u
en
tial
L
S
T
M
an
d
m
u
lti
-
h
ea
d
L
STM
ar
e
co
m
b
in
atio
n
s
o
f
L
STM
th
em
s
elv
es
in
s
eq
u
en
tial
[
8
8
]
.
C
NN
is
th
e
m
o
s
t
o
f
ten
u
s
ed
m
eth
o
d
i
n
L
C
m
o
d
els,
as sh
o
wn
in
Fig
u
r
e
1
2
.
Fig
u
r
e
1
1
.
T
h
e
m
et
h
o
d
s
u
s
ed
i
n
L
C
p
r
ep
r
o
ce
s
s
in
g
Fig
u
r
e
1
2
.
T
h
e
m
et
h
o
d
c
o
m
b
i
n
ed
with
L
STM
in
L
C
lear
n
in
g
T
h
r
ee
ty
p
es
o
f
LC
ar
e
s
er
ial
co
m
b
in
atio
n
,
p
ar
allel
co
m
b
in
at
io
n
,
an
d
m
ix
ed
c
o
m
b
in
atio
n
.
MC
-
L
STM
[
2
7
]
,
E
R
-
L
STM
[
8
0
]
,
C
NN
-
L
STM
[
8
8
]
,
a
n
d
B
iLST
M
-
So
f
tm
ax
[
8
9
]
ar
e
ex
am
p
les
o
f
th
e
L
C
s
er
ial
co
m
b
in
atio
n
s
.
C
NN
-
B
iLST
M
[
2
9
]
,
AC
N
-
L
STM
[
3
0
]
,
h
y
b
r
i
d
1
DC
NN
-
L
STM
[
3
1
]
,
an
d
C
NN
-
L
STM
[
8
2
]
a
r
e
ex
am
p
les
o
f
th
e
L
C
p
ar
allel
c
o
m
b
in
atio
n
s
.
R
eu
s
ab
le
L
STM
n
etwo
r
k
(
R
L
N)
[
3
2
]
,
Hy
b
r
id
C
NN
-
L
STM
with
m
u
lti
-
lev
el
atten
tio
n
f
u
s
io
n
[
7
4
]
,
d
is
tr
ib
u
ted
e
n
s
em
b
le
L
STM
[
7
5
]
,
an
d
E
PKSL
[
9
0
]
ar
e
ex
am
p
les
o
f
th
e
L
C
m
ix
ed
co
m
b
in
atio
n
s
.
3
.
2
.
5
.
RQ
5
:
ho
w
t
o
o
ptim
ize
a
nd
ev
a
lua
t
e
L
C
m
o
del dev
elo
pm
ent
?
So
m
e
L
C
m
o
d
el
d
e
v
elo
p
m
e
n
ts
u
s
e
o
p
tim
izatio
n
m
eth
o
d
s
to
im
p
r
o
v
e
m
o
d
el
p
er
f
o
r
m
an
ce
.
Fig
u
r
e
1
3
s
h
o
ws
th
e
p
r
o
p
o
r
tio
n
o
f
alg
o
r
ith
m
s
d
is
co
v
er
ed
f
o
r
o
p
tim
izi
n
g
L
C
m
o
d
el
d
ev
elo
p
m
e
n
t.
T
h
e
PS
O
an
d
Ad
am
alg
o
r
ith
m
s
ar
e
th
e
m
o
s
t
o
f
ten
u
s
ed
f
o
r
t
h
is
p
u
r
p
o
s
e.
An
L
C
m
o
d
el
u
s
in
g
PS
O
ca
n
p
r
o
d
u
ce
g
o
o
d
p
er
f
o
r
m
a
n
ce
,
as it a
ch
i
e
v
ed
a
h
ig
h
F1
-
s
co
r
e
[
9
1
]
.
Dev
el
o
p
in
g
a
n
L
C
m
o
d
e
l u
s
in
g
th
e
Ad
am
alg
o
r
ith
m
ca
n
also
lead
to
g
o
o
d
p
er
f
o
r
m
an
ce
,
with
h
ig
h
F1
-
s
co
r
e,
ac
cu
r
ac
y
,
a
n
d
AUC
[
7
0
]
,
[
8
9
]
,
[
9
2
]
.
All
L
C
m
o
d
el
d
ev
elo
p
m
en
t
g
o
es
th
r
o
u
g
h
a
m
o
d
el
e
v
alu
atio
n
cy
cle.
Fig
u
r
e
1
4
s
h
o
ws
th
e
p
r
o
p
o
r
tio
n
o
f
e
v
alu
atio
n
m
eth
o
d
s
d
is
co
v
er
ed
in
L
C
m
o
d
el
d
ev
el
o
p
m
en
t.
T
h
e
m
o
s
t
o
f
ten
ev
al
u
atio
n
m
eth
o
d
s
ar
e
ac
cu
r
ac
y
,
r
o
o
t
m
ea
n
s
q
u
ar
e
d
er
r
o
r
(
R
MSE
)
,
an
d
F1
-
s
co
r
e
.
Sev
er
al
L
C
m
o
d
el
d
ev
elo
p
m
e
n
ts
h
av
e
ac
h
iev
ed
h
ig
h
ac
c
u
r
ac
y
,
ap
p
r
o
ac
h
in
g
9
9
%
[
9
3
]
–
[
9
5
]
,
b
u
t
th
e
r
e
ar
e
s
till
m
o
d
els
with
ac
cu
r
ac
y
u
n
d
er
7
2
%
[
9
6
]
.
So
m
e
L
C
m
o
d
el
d
ev
elo
p
m
en
ts
h
av
e
ac
h
iev
ed
r
elativ
ely
s
m
all
R
MSE
[
9
7
]
,
[
9
8
]
,
b
u
t
th
er
e
ar
e
s
ti
ll
m
o
d
els
with
h
ig
h
R
MSE
[
9
9
]
.
Sev
er
al
L
C
d
ev
elo
p
m
en
ts
h
av
e
ac
h
iev
e
d
h
ig
h
F1
-
s
co
r
es
[
1
0
0
]
–
[
1
0
2
]
,
b
u
t
th
er
e
ar
e
s
till
m
o
d
els
with
lo
w
F1
-
s
co
r
es
[
1
0
3
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
A
r
ev
iew
o
n
lo
n
g
s
h
o
r
t
-
term me
mo
r
y
co
mb
in
a
tio
n
d
ev
elo
p
me
n
t (
A
h
ma
d
R
iya
d
i)
4433
Fig
u
r
e
1
3
.
Op
tim
izatio
n
alg
o
r
i
th
m
s
in
L
C
Fig
u
r
e
1
4
.
E
v
alu
atio
n
m
eth
o
d
in
L
C
3
.
2
.
6
.
RQ
6
:
wha
t
t
a
s
k
s
do
es L
C
m
o
del dev
elo
pm
ent
perf
o
rm
?
T
h
e
d
ev
elo
p
m
en
t
o
f
th
e
L
C
m
o
d
el
was
ca
r
r
ied
o
u
t
to
co
m
p
lete
ce
r
tain
task
s
in
s
o
lv
in
g
p
r
o
b
lem
s
in
th
e
r
ea
l
w
o
r
ld
.
T
h
is
m
o
d
el
h
as
co
m
p
leted
v
ar
io
u
s
task
s
.
Fiv
e
m
ain
task
s
th
at
ar
e
o
f
te
n
ca
r
r
ie
d
o
u
t
in
L
C
m
o
d
el
d
e
v
elo
p
m
e
n
t
in
clu
d
e
r
ec
o
g
n
itio
n
[
1
2
]
,
[
2
6
]
,
[
8
0
]
,
[
8
5
]
,
[
8
6
]
,
[
1
0
0
]
,
p
r
e
d
ictio
n
[
1
0
4
]
–
[
1
0
6
]
,
d
etec
tio
n
[
1
0
7
]
,
[
1
0
8
]
,
f
o
r
ec
asti
n
g
[
1
0
9
]
–
[
1
1
4
]
,
an
d
class
if
icatio
n
[
1
1
5
]
–
[
1
1
7
]
as
s
h
o
wn
in
T
a
b
le
1
.
Oth
e
r
task
s
p
er
f
o
r
m
ed
b
y
L
C
m
o
d
el
d
e
v
elo
p
m
en
t
i
n
a
lim
ited
s
co
p
e
in
clu
d
e
an
aly
s
is
,
m
o
d
elin
g
[
1
1
8
]
,
o
p
tim
izatio
n
[
1
1
9
]
,
s
en
s
in
g
,
an
d
d
iag
n
o
s
is
.
Pre
d
ictio
n
is
th
e
task
m
o
s
t L
C
m
o
d
el
d
e
v
elo
p
m
e
n
t p
er
f
o
r
m
s
.
T
ab
le
1
.
T
h
e
p
r
o
p
o
r
tio
n
o
f
L
C
d
ev
elo
p
m
en
t ta
s
k
s
No
Ta
sk
P
r
o
p
o
r
t
i
o
n
(
%)
1
P
r
e
d
i
c
t
i
o
n
23
2
D
e
t
e
c
t
i
o
n
14
3
F
o
r
e
c
a
st
i
n
g
14
4
C
l
a
s
si
f
i
c
a
t
i
o
n
12
5
R
e
c
o
g
n
i
t
i
o
n
8
6
V
a
r
i
o
u
s
o
t
h
e
r
s
t
a
sk
29
3
.
2
.
7
.
RQ
7
:
wha
t
pro
blem
s
do
es t
he
dev
elo
pm
ent
o
f
t
he
L
C
m
o
del so
lv
e?
Pro
b
lem
s
in
th
e
r
ea
l
wo
r
ld
ar
e
n
u
m
er
o
u
s
an
d
r
a
p
id
ly
d
ev
el
o
p
in
g
.
Var
io
u
s
d
ev
elo
p
m
e
n
ts
h
av
e
b
ee
n
ca
r
r
ied
o
u
t
to
s
o
l
v
e
th
ese
p
r
o
b
lem
s
.
W
e
f
o
u
n
d
th
at
th
e
d
ev
elo
p
m
en
t
o
f
th
e
L
C
m
o
d
el
h
a
s
p
en
etr
ated
v
ar
io
u
s
r
ea
l
p
r
o
b
lem
ar
ea
s
in
th
e
wo
r
ld
,
as
s
h
o
wn
in
Fig
u
r
e
1
5
.
T
h
ese
p
r
o
b
lem
d
o
m
ai
n
s
in
clu
d
e
th
e
en
v
ir
o
n
m
e
n
t,
m
ec
h
an
ical,
elec
tr
ical,
h
ea
lth
,
an
d
f
in
an
cial.
E
n
v
ir
o
n
m
en
tal
p
r
o
b
lem
s
in
clu
d
e
u
r
b
an
wate
r
[
3
8
]
,
u
r
b
a
n
f
lo
o
d
in
g
[
4
5
]
,
wate
r
p
o
llu
tio
n
[
1
0
9
]
,
d
o
m
esti
c
waste
g
en
er
atio
n
[
1
1
4
]
,
an
d
tr
an
s
p
o
r
tatio
n
[
1
2
0
]
.
Me
ch
an
ical
p
r
o
b
lem
s
in
clu
d
e
s
er
v
o
s
y
s
tem
s
[
6
6
]
,
m
is
s
ile
m
an
eu
v
e
r
tr
a
jecto
r
ies
[
6
7
]
,
air
cr
a
f
t
en
g
in
es
[
7
1
]
,
a
n
d
m
ac
h
in
e
cu
tter
h
ea
d
s
[
1
2
1
]
.
E
n
e
r
g
y
p
r
o
b
lem
s
in
clu
d
e
s
o
lar
p
o
wer
[
1
0
5
]
an
d
win
d
p
o
wer
[
1
2
2
]
.
I
n
d
u
s
t
r
i
a
l
p
r
o
b
l
e
m
s
i
n
c
l
u
d
e
c
h
e
m
i
c
a
l
p
r
o
c
es
s
es
[
1
8
]
,
g
a
s
a
n
a
l
y
s
is
[
9
4
]
,
a
n
d
c
o
m
p
l
e
x
p
r
o
d
u
c
t
d
e
s
i
g
n
[
1
1
8
]
.
H
e
a
l
t
h
p
r
o
b
l
e
m
s
i
n
c
l
u
d
e
d
r
u
g
r
e
a
c
t
i
o
n
s
[
1
9
]
,
c
a
r
d
i
o
v
a
s
cu
l
a
r
i
s
s
u
e
s
[
8
8
]
,
b
r
a
i
n
t
u
m
o
r
s
[
9
5
]
,
f
o
o
d
s
a
f
e
t
y
[
1
0
3
]
,
a
n
d
i
n
f
l
u
e
n
z
a
[
1
2
3
]
.
Hu
m
an
-
s
ty
le
p
r
o
b
lem
s
in
clu
d
e
f
ac
ial
em
o
tio
n
s
[
2
6
]
,
s
leep
s
tag
in
g
[
3
3
]
,
g
ait
p
h
ases
[
8
6
]
,
a
n
d
d
r
iv
i
n
g
s
ty
le
[
1
1
5
]
.
Fin
an
cial
p
r
o
b
lem
s
in
clu
d
e
C
h
in
a’
s
r
ea
l
estate
s
to
ck
tr
en
d
[
5
7
]
,
s
to
ck
p
r
ice
[
9
8
]
,
an
d
c
r
ed
it
ca
r
d
f
r
au
d
[
1
2
4
]
.
So
cial
m
ed
ia
p
r
o
b
lem
s
in
clu
d
e
tex
t
[
5
3
]
,
s
o
cial
n
etwo
r
k
s
[
7
8
]
,
an
d
C
h
in
ese
n
ews
[
8
2
]
.
Ag
r
icu
ltu
r
al
p
r
o
b
lem
s
in
clu
d
e
p
lan
t d
is
ea
s
e
[
5
2
]
,
t
o
m
ato
s
ee
d
cu
ltiv
ar
s
[
1
2
5
]
,
a
n
d
ag
r
icu
lt
u
r
al
p
r
o
d
u
cts
[
1
2
6
]
.
E
lectr
ical
p
r
o
b
lem
s
in
clu
d
e
elec
tr
ical
co
n
s
u
m
p
tio
n
f
o
r
s
h
ip
s
[
2
1
]
a
n
d
elec
tr
ical
lo
a
d
[
4
9
]
.
Geo
g
r
ap
h
y
p
r
o
b
lem
s
i
n
clu
d
e
lith
o
lo
g
y
[
1
6
]
,
f
i
b
er
o
p
tic
ca
b
le
[
9
3
]
,
g
r
o
u
n
d
m
o
tio
n
[
1
2
7
]
,
ac
u
te
m
o
u
n
tain
s
ick
n
ess
[
1
2
8
]
,
an
d
f
a
u
lt
lo
ca
tio
n
[
1
2
9
]
.
Net
wo
r
k
p
r
o
b
lem
s
in
clu
d
e
I
o
T
en
v
ir
o
n
m
en
t
[
1
3
0
]
a
n
d
clo
u
d
co
m
p
u
tin
g
[
1
3
1
]
.
Ma
r
itime
p
r
o
b
lem
s
in
clu
d
e
s
h
ip
m
o
tio
n
[
1
7
]
an
d
wav
e
h
e
ig
h
t
[
5
4
]
.
I
m
ag
e
p
r
o
b
lem
s
in
clu
d
e
h
y
p
e
r
s
p
ec
tr
al
im
ag
e
[
6
4
]
,
im
ag
e
ca
p
tio
n
g
en
er
atio
n
[
1
3
2
]
,
a
n
d
f
u
s
ed
m
u
ltimo
d
ality
m
e
d
ical
im
ag
e
[
1
3
3
]
.
L
a
n
g
u
a
g
e
p
r
o
b
lem
s
in
clu
d
e
b
ab
y
s
ig
n
l
an
g
u
ag
e
[
4
8
]
an
d
m
u
ltil
in
g
u
a
l
h
u
m
o
r
a
n
d
ir
o
n
y
[
7
6
]
.
E
lectr
o
m
ed
ical
p
r
o
b
lem
s
in
clu
d
e
e
p
ilep
tic
E
E
G
s
ig
n
a
ls
[
5
8
]
,
an
d
E
E
G
[
1
3
4
]
.
O
th
er
p
r
o
b
lem
s
i
n
clu
d
e
p
ar
ti
al
d
is
ch
ar
g
e
[
2
5
]
,
m
u
lti
-
d
o
m
ain
[
7
5
]
,
an
d
p
er
m
a
f
r
o
s
t d
eg
r
a
d
atio
n
[
1
3
5
]
.
3
.
2
.
8
.
RQ
8
:
wha
t
is
t
he
re
ce
nt
t
re
nd
in t
he
dev
elo
pm
ent
o
f
L
C
m
o
dels
?
W
e
f
o
u
n
d
th
at
th
e
tr
en
d
o
f
d
ata
u
s
ed
i
n
L
C
m
o
d
el
d
e
v
e
lo
p
m
en
t
is
g
ettin
g
clo
s
er
to
r
ea
l
-
wo
r
ld
p
r
o
b
lem
s
.
T
h
is
is
s
h
o
wn
b
y
th
e
p
r
o
p
o
r
tio
n
o
f
d
ata,
n
am
ely
6
4
%
co
llected
f
r
o
m
ex
p
e
r
im
en
tal
d
ata
o
r
o
f
f
icial
d
ata,
as
s
h
o
wn
in
Fig
u
r
e
9
.
T
h
e
d
ev
elo
p
m
en
t
o
f
th
e
L
C
m
o
d
el
u
s
in
g
o
f
f
icial
d
ata
s
h
o
ws
th
at
L
C
m
o
d
el
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
14
,
No
.
6
,
Dec
em
b
er
20
25
:
4
4
2
7
-
4
4
4
1
4434
d
ev
elo
p
m
e
n
t
is
n
ee
d
ed
to
p
r
o
v
id
e
s
o
lu
tio
n
s
to
b
u
s
in
ess
p
r
o
b
lem
s
.
L
C
m
o
d
el
d
ev
elo
p
m
en
t
u
s
in
g
p
u
b
lic
d
atasets
aim
s
to
f
in
d
th
e
b
est
p
er
f
o
r
m
i
n
g
.
T
h
e
tr
e
n
d
o
f
L
C
m
o
d
el
d
ev
elo
p
m
en
t
b
ased
o
n
r
esear
ch
p
r
o
b
lem
d
o
m
ain
s
s
h
o
ws
th
at
t
h
e
s
co
p
e
o
f
L
C
m
o
d
el
r
esear
ch
is
in
cr
e
asin
g
ly
b
r
o
ad
,
as
s
h
o
wn
in
Fig
u
r
e
1
6
.
T
h
is
s
h
o
ws
th
at
all
ar
ea
s
o
f
r
esear
ch
ar
e
p
o
s
s
ib
le
to
ca
r
r
y
o
u
t
th
r
o
u
g
h
th
e
d
ev
elo
p
m
en
t
o
f
L
C
m
o
d
e
ls
.
Me
an
wh
ile,
th
e
tr
en
d
o
f
L
C
m
o
d
el
d
ev
elo
p
m
en
t
b
ased
o
n
th
e
task
s
s
h
o
ws
th
at
L
C
d
ev
elo
p
m
e
n
t
is
u
s
ed
to
s
o
lv
e
p
r
ed
ictio
n
,
d
etec
tio
n
,
f
o
r
ec
asti
n
g
,
class
if
icatio
n
,
an
d
r
ec
o
g
n
itio
n
task
s
.
W
e
f
o
u
n
d
th
at
t
h
e
tr
e
n
d
o
f
p
r
ep
r
o
ce
s
s
in
g
in
L
C
m
o
d
el
d
e
v
elo
p
m
e
n
t is n
o
r
m
ali
za
tio
n
,
as sh
o
wn
in
Fig
u
r
e
1
1
.
C
NN
is
th
e
m
o
s
t tr
en
d
in
g
m
et
h
o
d
f
o
r
co
m
b
in
in
g
with
L
STM
in
th
e
lear
n
in
g
p
r
o
ce
s
s
,
as
s
h
o
wn
in
Fig
u
r
e
1
2
.
Me
an
wh
ile,
th
e
tr
en
d
o
f
L
C
m
o
d
el
d
ev
el
o
p
m
en
t
o
n
test
in
g
r
esu
lts
s
h
o
ws th
at
L
C
m
o
d
el
p
er
f
o
r
m
an
ce
is
s
till
v
ar
ied
.
Fig
u
r
e
1
5
.
T
h
e
p
r
o
b
lem
d
o
m
ai
n
s
s
o
lv
ed
b
y
L
C
Fig
u
r
e
1
6
.
L
C
d
ev
elo
p
m
en
t c
h
allen
g
es
3
.
2
.
9
.
RQ
9
:
wha
t
a
re
t
he
re
ce
nt
cha
lleng
es in L
C
m
o
del dev
elo
pm
ent
?
W
e
f
o
u
n
d
th
at
th
e
ch
allen
g
e
s
o
f
d
ev
elo
p
in
g
an
L
C
m
o
d
e
l
r
elate
d
t
o
d
ata
p
r
o
b
lem
s
in
clu
d
e
d
ata
g
r
o
wth
,
d
ata
e
x
p
an
s
io
n
,
d
ata
e
x
p
lo
r
atio
n
,
d
ata
au
g
m
en
tatio
n
,
d
ata
m
o
d
elin
g
,
an
d
d
ata
ex
tr
a
ctio
n
,
as
s
h
o
wn
in
Fig
u
r
e
1
6
.
Fu
r
t
h
er
d
ata
g
r
o
wth
s
tu
d
ies
wer
e
ca
r
r
ied
o
u
t
to
in
cr
ea
s
e
th
e
am
o
u
n
t
o
f
d
ata
[
1
3
6
]
,
[
1
3
7
]
.
T
h
e
g
r
o
wth
o
f
d
ata
r
eq
u
ir
es
f
u
r
th
er
s
tu
d
ies
o
n
au
g
m
en
tatio
n
[
8
1
]
o
r
ex
p
lo
r
atio
n
[
1
0
2
]
.
F
u
r
t
h
e
r
s
t
u
d
i
es
o
n
d
at
a
e
x
p
a
n
s
i
o
n
w
e
r
e
c
a
r
r
i
e
d
o
u
t
b
y
e
x
p
a
n
d
i
n
g
t
h
e
s
c
o
p
e
a
n
d
r
a
n
g
e
o
f
d
a
t
a
[
1
3
8
]
–
[
1
4
0
]
.
F
u
r
t
h
e
r
s
t
u
d
i
e
s
o
n
d
a
t
a
m
o
d
e
l
i
n
g
a
n
d
d
a
t
a
e
x
t
r
a
c
ti
o
n
a
r
e
n
e
e
d
e
d
t
o
s
i
m
p
li
f
y
t
h
e
p
r
o
c
es
s
i
n
g
a
n
d
r
e
p
r
e
s
e
n
t
a
ti
o
n
o
f
d
a
t
a
[
7
6
]
,
[
1
1
7
]
,
[
1
2
5
]
.
C
h
allen
g
es
in
L
C
m
o
d
el
d
ev
elo
p
m
en
t
r
elate
d
to
im
p
r
o
v
in
g
m
o
d
el
p
er
f
o
r
m
a
n
ce
in
cl
u
d
e
in
cr
ea
s
in
g
ac
cu
r
ac
y
,
en
h
an
cin
g
g
en
er
a
lizatio
n
ab
ilit
y
,
co
n
d
u
ctin
g
m
eth
o
d
co
m
b
i
n
atio
n
tr
ial
s
,
an
d
co
m
p
ar
in
g
co
m
b
in
atio
n
m
eth
o
d
s
.
Fu
r
t
h
e
r
r
esear
ch
r
elate
d
to
ac
c
u
r
ac
y
in
clu
d
es
in
v
o
l
v
in
g
u
n
ce
r
tain
ty
th
eo
r
y
[
1
4
1
]
,
in
cr
ea
s
in
g
d
ata
f
r
o
m
m
u
ltip
le
s
o
u
r
ce
s
[
5
6
]
,
[
1
2
8
]
,
co
m
b
in
i
n
g
v
ar
io
u
s
m
eth
o
d
s
[
3
9
]
,
[
8
3
]
,
[
8
4
]
,
[
9
9
]
,
[
1
2
4
]
,
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
A
r
ev
iew
o
n
lo
n
g
s
h
o
r
t
-
term me
mo
r
y
co
mb
in
a
tio
n
d
ev
elo
p
me
n
t (
A
h
ma
d
R
iya
d
i)
4435
an
d
co
m
p
ar
in
g
th
em
[
1
1
1
]
.
Ad
d
itio
n
al
s
tu
d
ies
r
elate
d
to
g
en
er
aliza
tio
n
in
v
o
lv
e
u
s
in
g
m
o
r
e
co
m
p
lex
d
ata
[
1
8
]
,
lar
g
er
d
atasets
[
9
5
]
,
ad
d
itio
n
al
eq
u
i
p
m
en
t
[
1
4
2
]
,
a
n
d
in
teg
r
atin
g
id
ea
s
a
n
d
p
ar
a
m
eter
s
in
to
m
o
d
els
[
1
2
1
]
.
A
s
ig
n
if
ican
t
is
s
u
e
in
L
C
m
o
d
el
d
ev
elo
p
m
en
t
i
s
ef
f
icien
cy
.
Fu
r
t
h
er
r
esear
c
h
r
elate
d
to
m
o
d
el
ef
f
icien
cy
in
clu
d
es
co
m
p
u
tat
io
n
,
r
eso
u
r
ce
s
,
to
o
ls
,
an
d
o
p
tim
izatio
n
.
C
o
m
p
u
tatio
n
al
ch
allen
g
es
in
clu
d
e
alg
o
r
ith
m
s
,
ca
lcu
latio
n
s
,
a
n
d
t
im
e
co
m
p
lex
ity
[
1
9
]
,
[
8
8
]
,
[
1
0
9
]
.
L
ac
k
o
f
r
eso
u
r
ce
s
an
d
p
o
o
r
to
o
l
q
u
ality
ar
e
o
n
g
o
in
g
is
s
u
e
in
L
C
m
o
d
el
d
ev
elo
p
m
en
t
[
8
5
]
,
[
1
0
0
]
,
[
1
3
3
]
,
[
1
4
3
]
,
[
1
4
4
]
.
Ho
t
to
p
ics
r
elate
d
to
L
C
m
o
d
el
d
ev
elo
p
m
e
n
t
o
p
tim
izatio
n
in
clu
d
e
o
p
tim
izatio
n
tech
n
iq
u
e
s
,
s
tr
u
ctu
r
e
o
p
tim
izatio
n
,
al
g
o
r
ith
m
s
,
s
ec
u
r
ity
,
o
v
er
f
itti
n
g
,
an
d
er
r
o
r
s
[
2
3
]
,
[
6
8
]
,
[
6
9
]
,
[
9
8
]
,
[
1
1
2
]
,
[
1
1
5
]
,
[
1
1
8
]
,
[
1
1
9
]
.
Fu
r
th
er
r
esear
ch
o
n
L
C
m
o
d
el
d
ev
elo
p
m
e
n
t
ca
n
b
e
co
n
d
u
cte
d
at
th
e
im
p
lem
e
n
tatio
n
lev
el.
Stu
d
y
in
g
m
o
d
el
im
p
lem
en
tat
io
n
in
v
ar
io
u
s
f
ield
is
a
f
ascin
atin
g
ar
ea
to
e
x
p
l
o
r
e
[
1
1
]
,
[
4
1
]
,
[
8
7
]
,
[
9
1
]
,
[
9
7
]
,
[
1
3
1
]
,
[
1
4
5
]
,
[
1
4
6
]
.
E
x
p
a
n
d
in
g
t
h
e
s
co
p
e
o
f
im
p
lem
en
tatio
n
is
also
a
ch
all
en
g
in
g
s
tu
d
y
[
5
8
]
,
[
9
2
]
,
[
1
2
9
]
.
T
h
e
d
ev
el
o
p
m
en
t
o
f
t
h
e
r
esu
ltin
g
L
C
m
o
d
el
is
a
n
ev
er
-
e
n
d
in
g
c
h
allen
g
e.
C
h
allen
g
es
r
elate
d
to
m
o
d
el
d
e
v
elo
p
m
e
n
t
in
clu
d
e
an
aly
s
is
ex
p
an
s
io
n
,
m
o
d
e
l
in
teg
r
atio
n
,
e
m
b
ed
d
ed
m
o
d
els,
an
d
m
o
d
e
l
ex
p
lo
r
atio
n
.
An
aly
s
is
ex
p
a
n
s
io
n
is
ca
r
r
ie
d
o
u
t
to
in
cr
ea
s
e
t
h
e
m
o
d
el'
s
s
o
lu
tio
n
ca
p
ab
ilit
ies
f
o
r
th
e
p
r
o
b
lem
s
b
ein
g
s
o
lv
ed
[
3
7
]
,
[
7
5
]
,
[
1
0
1
]
,
[
1
1
3
]
,
[
1
3
4
]
,
[
1
4
7
]
–
[
1
4
9
]
.
M
o
d
el
in
teg
r
atio
n
an
d
em
b
ed
d
e
d
m
o
d
els ar
e
ca
r
r
ied
o
u
t
r
eg
a
r
d
in
g
th
e
u
s
e
o
f
r
eso
u
r
ce
s
o
r
to
o
ls
to
m
ak
e
t
h
e
m
o
d
el
ea
s
ier
[
2
9
]
,
[
3
4
]
,
[
5
3
]
,
[
7
4
]
,
[
1
5
0
]
.
Fu
r
t
h
er
s
tu
d
ies
o
n
m
o
d
el
ex
p
lo
r
atio
n
in
clu
d
e
d
ata
ex
p
lo
r
atio
n
,
tech
n
iq
u
es,
r
eso
u
r
ce
s
,
an
d
alg
o
r
ith
m
s
s
o
th
at
th
e
m
o
d
el
p
r
o
v
i
d
es
m
o
r
e
ad
d
e
d
v
alu
e
[
3
1
]
,
[
7
8
]
,
[
7
9
]
,
[
9
7
]
,
[
1
0
4
]
,
[
1
0
5
]
,
[
1
0
8
]
,
[
1
2
7
]
,
[
1
5
1
]
.
An
o
t
h
e
r
i
n
te
r
es
ti
n
g
ch
all
en
g
e
i
n
L
C
d
e
v
el
o
p
m
e
n
t
is
p
ar
am
ete
r
ex
p
an
s
io
n
.
Pa
r
a
m
et
er
e
x
p
a
n
s
i
o
n
ai
m
s
t
o
i
m
p
r
o
v
e
t
h
e
m
o
d
el
b
y
co
n
s
i
d
e
r
i
n
g
o
t
h
e
r
i
n
f
lu
e
n
ti
al
f
ac
t
o
r
s
t
h
at
h
av
e
n
o
t
b
ee
n
c
o
n
s
id
e
r
ed
in
p
r
e
v
i
o
u
s
m
o
d
e
ls
[
4
0
]
,
[
5
7
]
,
[
7
1
]
,
[
7
7
]
,
[
9
0
]
,
[
1
0
6
]
,
[
1
0
7
]
,
[
1
1
6
]
,
[
1
2
2
]
,
[
1
2
3
]
,
[
1
2
6
]
,
[
1
3
0
]
,
[
1
4
9
]
,
[
1
5
1
]
–
[
1
5
3
]
.
T
h
e
d
e
v
el
o
p
m
e
n
t
o
f
t
h
e
L
C
m
o
d
el
als
o
d
e
p
e
n
d
s
o
n
th
e
s
p
ec
if
ie
d
r
ese
a
r
c
h
t
o
p
ic
o
r
f
iel
d
.
R
ese
ar
c
h
f
i
e
ld
s
t
h
a
t
r
e
q
u
ir
e
f
u
r
t
h
e
r
s
t
u
d
y
i
n
cl
u
d
e
m
u
lti
li
n
g
u
al
[
8
2
]
,
g
ai
t
r
e
c
o
g
n
i
ti
o
n
[
8
6
]
,
m
u
lti
-
em
o
t
io
n
[
9
6
]
,
s
e
m
a
n
t
ics
[
1
3
2
]
,
a
n
d
f
e
at
u
r
e
ex
tr
ac
t
i
o
n
[
1
5
4
]
.
3
.
3
.
Dis
cus
s
io
n
T
h
is
r
esear
ch
s
p
ec
if
ically
co
n
d
u
cts
a
r
ev
iew
p
ap
er
th
at
ex
p
l
ain
s
th
e
d
ev
elo
p
m
en
t
o
f
th
e
L
C
m
eth
o
d
.
T
h
e
p
ap
er
s
s
elec
ted
f
o
r
th
i
s
r
esear
ch
wer
e
f
ilter
ed
r
ig
o
r
o
u
s
ly
to
e
n
s
u
r
e
h
ig
h
q
u
a
lity
.
I
t
p
r
o
v
id
es
a
co
m
p
r
eh
e
n
s
iv
e
d
escr
ip
tio
n
o
f
th
e
r
ev
iewe
d
p
ap
er
s
,
d
em
o
n
s
tr
atin
g
th
e
lev
el
o
f
q
u
ality
o
f
th
e
p
a
p
er
s
u
s
ed
,
wh
ich
h
as
n
o
t
b
ee
n
d
o
n
e
in
p
r
ev
io
u
s
r
ev
iews.
T
h
e
r
ev
iew
ap
p
lied
a
n
SLR
m
eth
o
d
,
wh
ich
h
as
n
o
t
b
ee
n
u
s
ed
in
an
y
L
STM
r
ev
iew.
T
h
e
S
L
R
m
eth
o
d
g
av
e
a
clea
r
er
a
n
d
m
o
r
e
f
o
c
u
s
ed
r
ev
iew.
T
ab
l
e
2
s
h
o
ws
th
at
th
is
s
tu
d
y
ex
p
licitly
f
in
d
s
a
co
m
p
l
ete
ag
g
r
eg
atio
n
o
f
L
C
d
ev
elo
p
m
en
t
an
d
o
u
tlin
es
ch
allen
g
es
f
o
r
f
u
r
th
er
r
esear
c
h
o
n
L
C
.
Ho
wev
er
,
th
is
r
esea
r
ch
h
as
lim
itatio
n
s
.
T
h
e
r
ev
iew
s
co
p
e
is
lim
ited
to
th
e
L
C
m
eth
o
d
o
n
ly
.
T
h
e
r
ev
iew
in
clu
d
es
p
ap
er
s
p
u
b
lis
h
ed
in
2
0
2
3
with
titl
es
co
n
tain
in
g
"L
STM
"
o
r
"L
o
n
g
s
h
o
r
t
-
ter
m
m
em
o
r
y
".
T
h
e
r
e
v
iew
is
lim
ited
to
n
in
e
r
esear
ch
q
u
esti
o
n
s
.
T
o
en
h
an
ce
th
e
r
ev
iew,
ad
d
itio
n
al
r
esear
c
h
q
u
esti
o
n
s
ca
n
b
e
in
clu
d
ed
,
a
n
d
m
o
r
e
r
ec
e
n
t p
ap
er
s
ca
n
b
e
co
n
s
id
er
ed
d
u
e
to
t
h
e
r
ap
id
p
ac
e
o
f
r
esear
ch
d
ev
e
lo
p
m
en
ts
.
T
ab
le
2
.
Ag
g
r
eg
atio
n
o
f
L
C
d
ev
elo
p
m
en
t
R
e
se
a
r
c
h
q
u
e
st
i
o
n
A
g
g
r
e
g
a
t
i
o
n
M
o
d
e
l
d
e
v
e
l
o
p
m
e
n
t
f
r
a
m
e
w
o
r
k
D
a
t
a
p
r
e
p
a
r
a
t
i
o
n
,
t
r
a
i
n
i
n
g
mo
d
e
l
,
a
n
d
e
v
a
l
u
a
t
i
o
n
mo
d
e
l
D
a
t
a
so
u
r
c
e
E
x
p
e
r
i
m
e
n
t
a
l
d
a
t
a
a
n
d
p
u
b
l
i
c
d
a
t
a
s
e
t
D
a
t
a
t
y
p
e
N
u
meri
c
,
s
i
g
n
a
l
,
i
m
a
g
e
,
a
n
d
t
e
x
t
M
e
t
h
o
d
s
u
se
d
i
n
p
r
e
p
r
o
c
e
ss
i
n
g
N
o
r
mal
i
z
a
t
i
o
n
M
e
t
h
o
d
c
o
m
b
i
n
e
d
w
i
t
h
LS
TM
i
n
l
e
a
r
n
i
n
g
C
N
N
,
d
e
n
se,
a
t
t
e
n
t
i
o
n
,
S
o
f
t
M
a
x
,
a
n
d
R
e
L
u
O
p
t
i
mi
z
a
t
i
o
n
a
l
g
o
r
i
t
h
ms
P
S
O
a
n
d
A
d
a
m
Ev
a
l
u
a
t
i
o
n
me
t
h
o
d
A
c
c
u
r
a
c
y
,
R
M
S
E,
a
n
d
F
1
-
sc
o
r
e
Ta
sk
P
r
e
d
i
c
t
i
o
n
,
d
e
t
e
c
t
i
n
g
,
f
o
r
e
c
a
s
t
i
n
g
,
a
n
d
c
l
a
ssi
f
i
c
a
t
i
o
n
P
r
o
b
l
e
m
d
o
ma
i
n
s
s
o
l
v
e
d
A
l
l
d
o
mai
n
R
e
c
e
n
t
t
r
e
n
d
i
n
t
h
e
d
e
v
e
l
o
p
me
n
t
R
e
a
l
-
w
o
r
l
d
p
r
o
b
l
e
ms
R
e
c
e
n
t
c
h
a
l
l
e
n
g
e
s
i
n
t
h
e
d
e
v
e
l
o
p
me
n
t
R
e
se
a
r
c
h
f
i
e
l
d
,
d
a
t
a
,
m
o
d
e
l
,
p
e
r
f
o
r
m
a
n
c
e
,
a
n
d
i
mp
l
e
me
n
t
a
t
i
o
n
4.
CO
NCLU
SI
O
N
T
h
is
SLR
is
b
a
s
ed
o
n
1
4
6
p
a
p
er
s
s
elec
ted
f
r
o
m
Q1
o
r
Q2
-
in
d
ex
ed
jo
u
r
n
als
in
co
m
p
u
ter
s
cien
ce
in
2
0
2
3
.
All
p
ap
er
s
co
n
tain
"L
STM
"
o
r
"L
o
n
g
s
h
o
r
t
-
te
r
m
m
e
m
o
r
y
"
in
t
h
eir
titl
e.
T
h
e
L
C
m
o
d
el
d
ev
el
o
p
m
en
t
f
r
am
ewo
r
k
in
clu
d
es
d
ata
p
r
e
p
ar
atio
n
,
p
r
e
p
r
o
ce
s
s
in
g
,
lear
n
in
g
,
v
alid
atio
n
,
an
d
test
in
g
,
with
d
ata
m
ain
ly
s
o
u
r
ce
d
f
r
o
m
ex
p
er
im
en
tal
o
r
o
f
f
icial
s
o
u
r
c
es.
L
C
m
o
d
els
ai
m
to
ad
d
r
ess
r
ea
l
-
wo
r
ld
p
r
o
b
l
em
s
ac
r
o
s
s
v
ar
i
o
u
s
d
o
m
ain
s
s
u
ch
as
en
v
ir
o
n
m
en
t
,
m
ec
h
an
ical,
en
e
r
g
y
,
n
etwo
r
k
,
in
d
u
s
tr
y
,
g
eo
g
r
ap
h
y
,
elec
tr
ical,
h
ea
lth
,
h
u
m
a
n
s
ty
le,
ag
r
icu
ltu
r
e,
an
d
s
o
cial
m
ed
ia.
C
NN
is
a
p
o
p
u
la
r
m
et
h
o
d
in
L
C
m
o
d
el
d
ev
elo
p
m
en
t,
an
d
PS
O
is
wid
ely
u
s
ed
f
o
r
o
p
tim
izatio
n
.
R
esear
ch
o
n
L
STM
ce
ll
d
e
v
elo
p
m
e
n
t
is
lim
ited
.
Pre
d
ictio
n
,
d
ete
ctio
n
,
f
o
r
ec
asti
n
g
,
class
if
icatio
n
,
an
d
r
ec
o
g
n
itio
n
ar
e
co
m
m
o
n
task
s
in
L
C
m
o
d
el
d
ev
elo
p
m
en
t.
C
o
llab
o
r
atio
n
an
d
in
teg
r
atio
n
o
f
L
STM
with
o
th
er
m
eth
o
d
s
ar
e
tr
en
d
s
in
L
C
m
o
d
el
d
ev
elo
p
m
en
t,
o
f
f
er
in
g
s
o
l
u
tio
n
s
to
r
ea
l
-
wo
r
ld
ch
allen
g
es.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
14
,
No
.
6
,
Dec
em
b
er
20
25
:
4
4
2
7
-
4
4
4
1
4436
Data
ex
p
an
s
io
n
,
p
e
r
f
o
r
m
an
ce
en
h
an
ce
m
e
n
t,
m
o
d
el
c
o
m
p
a
r
is
o
n
,
r
esear
ch
f
ield
e
x
p
an
s
io
n
,
a
n
d
im
p
lem
en
tatio
n
ar
e
c
h
al
l
en
g
e
s
in
L
C
m
o
d
el
d
ev
elo
p
m
en
t
with
d
iv
er
s
e
p
ar
a
m
eter
s
an
d
d
o
m
ain
s
.
T
h
is
r
esear
ch
ad
d
r
ess
es
n
in
e
r
esear
ch
q
u
esti
o
n
s
.
T
h
e
r
ev
ie
w
h
ig
h
lig
h
ts
th
e
r
a
p
id
g
r
o
wt
h
o
f
L
C
m
o
d
els,
d
em
o
n
s
tr
atin
g
th
e
v
er
s
atility
o
f
L
STM
in
s
o
lv
in
g
co
m
p
lex
r
ea
l
-
wo
r
ld
p
r
o
b
lem
s
ac
r
o
s
s
v
ar
io
u
s
f
ield
s
.
Alth
o
u
g
h
L
C
d
ev
elo
p
m
e
n
t
is
in
cr
ea
s
in
g
ly
c
o
m
p
lex
,
it
p
r
o
v
id
es
p
r
o
m
is
in
g
o
p
p
o
r
tu
n
itie
s
to
ad
d
r
ess
ch
allen
g
es
in
s
o
lv
in
g
r
ea
l
-
w
o
r
ld
p
r
o
b
lem
s
.
F
UNDING
I
NF
O
R
M
A
T
I
O
N
T
h
is
r
esear
ch
was
f
u
n
d
ed
b
y
th
e
Fin
al
Pro
ject
R
ec
o
g
n
itio
n
Pro
g
r
am
o
f
Gad
jah
Ma
d
a
Un
iv
er
s
ity
B
atch
I
,
Fis
ca
l Y
ea
r
2
0
2
2
,
as s
tated
in
th
e
Ass
ig
n
m
en
t L
etter
Nu
m
b
er
: 3
5
5
0
/UN1
.
P.I
I
I
/Dit
-
L
it/P
T
.
0
1
.
0
5
/2
0
2
2
d
ated
Ma
y
2
0
,
2
0
2
2
.
AUTHO
R
CO
NT
RI
B
UT
I
O
NS ST
A
T
E
M
E
N
T
T
h
is
jo
u
r
n
al
u
s
es
th
e
C
o
n
tr
ib
u
to
r
R
o
les
T
ax
o
n
o
m
y
(
C
R
ed
iT)
to
r
ec
o
g
n
ize
in
d
iv
id
u
al
au
th
o
r
co
n
tr
ib
u
tio
n
s
,
r
ed
u
ce
au
th
o
r
s
h
ip
d
is
p
u
tes,
an
d
f
ac
ilit
ate
co
llab
o
r
atio
n
.
Na
m
e
o
f
Aut
ho
r
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
Ah
m
ad
R
iy
ad
i
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
Nu
r
R
o
k
h
m
an
✓
✓
✓
✓
✓
✓
✓
✓
L
u
k
m
an
He
r
y
awa
n
✓
✓
✓
✓
✓
✓
✓
C
:
C
o
n
c
e
p
t
u
a
l
i
z
a
t
i
o
n
M
:
M
e
t
h
o
d
o
l
o
g
y
So
:
So
f
t
w
a
r
e
Va
:
Va
l
i
d
a
t
i
o
n
Fo
:
Fo
r
mal
a
n
a
l
y
s
i
s
I
:
I
n
v
e
s
t
i
g
a
t
i
o
n
R
:
R
e
so
u
r
c
e
s
D
:
D
a
t
a
C
u
r
a
t
i
o
n
O
:
W
r
i
t
i
n
g
-
O
r
i
g
i
n
a
l
D
r
a
f
t
E
:
W
r
i
t
i
n
g
-
R
e
v
i
e
w
&
E
d
i
t
i
n
g
Vi
:
Vi
su
a
l
i
z
a
t
i
o
n
Su
:
Su
p
e
r
v
i
s
i
o
n
P
:
P
r
o
j
e
c
t
a
d
mi
n
i
st
r
a
t
i
o
n
Fu
:
Fu
n
d
i
n
g
a
c
q
u
i
si
t
i
o
n
CO
NF
L
I
C
T
O
F
I
N
T
E
R
E
S
T
ST
A
T
E
M
E
NT
T
h
e
au
th
o
r
s
co
n
f
ir
m
th
at
th
er
e
is
n
o
co
n
f
lict
o
f
in
ter
est ass
o
ciate
d
with
th
is
p
u
b
licatio
n
.
DATA AV
AI
L
AB
I
L
I
T
Y
T
h
e
au
th
o
r
s
c
o
n
f
i
r
m
t
h
at
th
e
d
ata
s
u
p
p
o
r
ti
n
g
t
h
e
f
i
n
d
in
g
s
o
f
t
h
is
s
t
u
d
y
a
r
e
a
v
a
ila
b
l
e
wit
h
i
n
t
h
e
a
r
ti
cle
.
RE
F
E
R
E
NC
E
S
[
1
]
S
.
H
o
c
h
r
e
i
t
e
r
a
n
d
J.
S
c
h
m
i
d
h
u
b
e
r
,
“
Lo
n
g
s
h
o
r
t
-
t
e
r
m
m
e
m
o
r
y
,
”
N
e
u
r
a
l
C
o
m
p
u
t
a
t
i
o
n
,
v
o
l
.
9
,
n
o
.
8
,
p
p
.
1
7
3
5
–
1
7
8
0
,
1
9
9
7
,
d
o
i
:
1
0
.
1
1
6
2
/
n
e
c
o
.
1
9
9
7
.
9
.
8
.
1
7
3
5
.
[
2
]
Y
.
Y
u
,
X
.
S
i
,
C
.
H
u
,
a
n
d
J.
Z
h
a
n
g
,
“
A
r
e
v
i
e
w
o
f
r
e
c
u
r
r
e
n
t
n
e
u
r
a
l
n
e
t
w
o
r
k
s
:
LST
M
c
e
l
l
s
a
n
d
n
e
t
w
o
r
k
a
r
c
h
i
t
e
c
t
u
r
e
s
,
”
N
e
u
r
a
l
C
o
m
p
u
t
a
t
i
o
n
,
v
o
l
.
3
1
,
n
o
.
7
,
p
p
.
1
2
3
5
–
1
2
7
0
,
2
0
1
9
,
d
o
i
:
1
0
.
1
1
6
2
/
n
e
c
o
_
a
_
0
1
1
9
9
.
[
3
]
G
.
V
.
H
o
u
d
t
,
C
.
M
o
s
q
u
e
r
a
,
a
n
d
G
.
N
á
p
o
l
e
s,
“
A
r
e
v
i
e
w
o
n
t
h
e
l
o
n
g
sh
o
r
t
-
t
e
r
m
m
e
m
o
r
y
mo
d
e
l
,
”
Ar
t
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
Re
v
i
e
w
,
v
o
l
.
5
3
,
n
o
.
8
,
p
p
.
5
9
2
9
–
5
9
5
5
,
2
0
2
0
,
d
o
i
:
1
0
.
1
0
0
7
/
s1
0
4
6
2
-
0
2
0
-
0
9
8
3
8
-
1.
[
4
]
S
.
Ti
w
a
r
i
a
n
d
A
.
K
.
C
h
a
t
u
r
v
e
d
i
,
“
A
s
u
r
v
e
y
o
n
LS
TM
-
b
a
s
e
d
st
o
c
k
m
a
r
k
e
t
p
r
e
d
i
c
t
i
o
n
,
”
I
l
k
o
g
re
t
i
m
O
n
l
i
n
e
-
E
l
e
m
e
n
t
a
r
y
E
d
u
c
a
t
i
o
n
O
n
l
i
n
e
,
v
o
l
.
2
0
,
n
o
.
5
,
p
p
.
1
6
7
1
–
1
6
7
7
,
2
0
2
1
,
d
o
i
:
1
0
.
1
7
0
5
1
/
i
l
k
o
n
l
i
n
e
.
2
0
2
1
.
0
5
.
1
8
2
.
[
5
]
B
.
L
i
n
d
e
ma
n
n
,
B
.
M
a
sc
h
l
e
r
,
N
.
S
a
h
l
a
b
,
a
n
d
M
.
W
e
y
r
i
c
h
,
“
A
s
u
r
v
e
y
o
n
a
n
o
ma
l
y
d
e
t
e
c
t
i
o
n
f
o
r
t
e
c
h
n
i
c
a
l
s
y
st
e
ms
u
s
i
n
g
LST
M
n
e
t
w
o
r
k
s,”
C
o
m
p
u
t
e
rs
i
n
I
n
d
u
st
r
y
,
v
o
l
.
1
3
1
,
2
0
2
1
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
c
o
mp
i
n
d
.
2
0
2
1
.
1
0
3
4
9
8
.
[
6
]
M
.
T
e
m
p
l
i
e
r
a
n
d
G
.
P
a
r
é
,
“
A
f
r
a
me
w
o
r
k
f
o
r
g
u
i
d
i
n
g
a
n
d
e
v
a
l
u
a
t
i
n
g
l
i
t
e
r
a
t
u
r
e
r
e
v
i
e
w
s,”
C
o
m
m
u
n
i
c
a
t
i
o
n
s
o
f
t
h
e
Ass
o
c
i
a
t
i
o
n
f
o
r
I
n
f
o
rm
a
t
i
o
n
S
y
st
e
m
s
,
v
o
l
.
3
7
,
2
0
1
5
,
d
o
i
:
1
0
.
1
7
7
0
5
/
1
C
A
I
S
.
0
3
7
0
6
.
[
7
]
S
.
G
.
P
a
u
l
e
t
a
l
.
,
“
C
o
m
b
a
t
i
n
g
C
o
v
i
d
-
1
9
u
s
i
n
g
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
a
n
d
d
e
e
p
l
e
a
r
n
i
n
g
:
a
p
p
l
i
c
a
t
i
o
n
s
,
c
h
a
l
l
e
n
g
e
s,
a
n
d
f
u
t
u
r
e
p
e
r
s
p
e
c
t
i
v
e
s
,
”
Arr
a
y
,
v
o
l
.
1
7
,
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
a
r
r
a
y
.
2
0
2
2
.
1
0
0
2
7
1
.
[
8
]
R
.
J
i
n
d
a
l
,
R
.
M
a
l
h
o
t
r
a
,
a
n
d
A
.
J
a
i
n
,
“
Te
c
h
n
i
q
u
e
s
f
o
r
t
e
x
t
c
l
a
ssi
f
i
c
a
t
i
o
n
:
l
i
t
e
r
a
t
u
r
e
r
e
v
i
e
w
a
n
d
c
u
r
r
e
n
t
t
r
e
n
d
s,
”
W
e
b
o
l
o
g
y
,
v
o
l
.
1
2
,
n
o
.
2
,
2
0
1
5
.
[
9
]
L.
F
e
n
g
,
“
P
r
e
d
i
c
t
i
n
g
o
u
t
p
u
t
r
e
s
p
o
n
se
s
o
f
n
o
n
l
i
n
e
a
r
d
y
n
a
m
i
c
a
l
s
y
st
e
ms
w
i
t
h
p
a
r
a
me
t
r
i
z
e
d
i
n
p
u
t
s
u
si
n
g
LST
M
,
”
I
EEE
J
o
u
r
n
a
l
o
n
Mu
l
t
i
sc
a
l
e
a
n
d
Mu
l
t
i
p
h
y
si
c
s C
o
m
p
u
t
a
t
i
o
n
a
l
T
e
c
h
n
i
q
u
e
s
,
v
o
l
.
8
,
p
p
.
9
7
–
1
0
7
,
2
0
2
3
,
d
o
i
:
1
0
.
1
1
0
9
/
J
M
M
C
T
.
2
0
2
3
.
3
2
4
2
0
4
4
.
[
1
0
]
M
.
T.
K
h
a
n
,
H
.
E.
Y
a
n
t
i
r
,
K
.
N
.
S
a
l
a
ma,
a
n
d
A
.
M
.
E
l
t
a
w
i
l
,
“
A
r
c
h
i
t
e
c
t
u
r
a
l
t
r
a
d
e
-
o
f
f
a
n
a
l
y
s
i
s
f
o
r
a
c
c
e
l
e
r
a
t
i
n
g
LST
M
n
e
t
w
o
r
k
u
si
n
g
R
a
d
i
x
-
r
O
B
C
sc
h
e
m
e
,
”
I
E
EE
T
ra
n
s
a
c
t
i
o
n
s
o
n
C
i
rc
u
i
t
s
a
n
d
S
y
s
t
e
m
s
I
:
Re
g
u
l
a
r
P
a
p
e
rs
,
v
o
l
.
7
0
,
n
o
.
1
,
p
p
.
2
6
6
–
2
7
9
,
2
0
2
3
,
d
o
i
:
1
0
.
1
1
0
9
/
TC
S
I
.
2
0
2
2
.
3
2
1
7
0
9
1
.
[
1
1
]
S
.
X
i
a
n
g
,
J.
Z
h
o
u
,
J.
L
u
o
,
F
.
Li
u
,
a
n
d
Y
.
Q
i
n
,
“
C
o
c
k
t
a
i
l
LST
M
a
n
d
i
t
s
a
p
p
l
i
c
a
t
i
o
n
i
n
t
o
m
a
c
h
i
n
e
r
e
m
a
i
n
i
n
g
u
sef
u
l
l
i
f
e
p
r
e
d
i
c
t
i
o
n
,
”
I
EEE/
A
S
M
E
T
r
a
n
s
a
c
t
i
o
n
s
o
n
M
e
c
h
a
t
ro
n
i
c
s
,
v
o
l
.
2
8
,
n
o
.
5
,
p
p
.
2
4
2
5
–
2
4
3
6
,
2
0
2
3
,
d
o
i
:
1
0
.
1
1
0
9
/
T
M
E
C
H
.
2
0
2
3
.
3
2
4
4
2
8
2
.
[
1
2
]
H
.
W
u
e
t
a
l
.
,
“
A
n
o
v
e
l
p
e
d
a
l
mu
scu
l
o
s
k
e
l
e
t
a
l
r
e
s
p
o
n
s
e
b
a
s
e
d
o
n
d
i
f
f
e
r
e
n
t
i
a
l
s
p
a
t
i
o
-
t
e
mp
o
r
a
l
LS
TM
f
o
r
h
u
ma
n
a
c
t
i
v
i
t
y
r
e
c
o
g
n
i
t
i
o
n
,
”
K
n
o
w
l
e
d
g
e
-
Ba
s
e
d
S
y
st
e
m
s
,
v
o
l
.
2
6
1
,
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
k
n
o
sy
s.
2
0
2
2
.
1
1
0
1
8
7
.
Evaluation Warning : The document was created with Spire.PDF for Python.