I
n
t
e
r
n
at
ion
al
Jou
r
n
a
l
of
I
n
f
o
r
m
at
ics
an
d
Com
m
u
n
icat
ion
T
e
c
h
n
ol
ogy
(
I
J
-
I
CT
)
Vo
l
.
1
5
,
N
o
.
4
,
D
e
c
e
m
b
e
r
20
2
6
,
pp.
1944~1953
I
S
S
N:
2252
-
8776
,
DO
I
:
10
.
11591/i
ji
c
t
.
v
1
5
i
4
.
pp
19
44
-
1953
1944
Jou
r
n
al
h
o
m
e
page
:
ht
tp:
//
ij
ict
.
iaes
c
or
e
.
c
om
A
d
v
e
r
sar
ia
l
-
r
e
si
li
e
n
t
i
ot
-
e
n
ab
le
d
f
in
an
c
ia
l
i
n
t
e
ll
ig
e
n
c
e
sy
st
e
m
u
si
n
g e
xp
la
in
a
b
le
d
e
e
p
l
e
ar
n
in
g f
or
se
n
t
i
m
e
n
t
-
d
r
iv
e
n
st
oc
k
t
r
ad
in
g
d
e
c
is
io
n
s
S
u
r
e
s
h
T
a
l
a
m
a
l
a
1
,
S
u
k
h
ad
a
Al
on
i
2
,
J.
Vij
ayal
ak
s
h
m
i
3
,
S
u
b
h
am
P
an
k
aj
S
am
an
t
ar
ay
4
,
K.
S
wan
t
h
an
a
5
,
Nis
ar
g
Gan
d
h
e
war
6
1
S
c
h
oo
l
of
B
us
in
e
s
s
, A
di
t
y
a
U
ni
ve
r
s
it
y
,
K
a
ki
na
da
, I
ndi
a
2
D
e
pa
r
tm
e
nt
of
C
o
mpu
te
r
E
ngi
n
e
e
r
in
g, A
. P
. S
ha
h
I
ns
ti
tu
te
of
T
e
c
hn
o
l
o
g
y
,
T
ha
n
e
(
W
e
s
t)
, M
a
ha
r
a
s
ht
r
a
, I
ndi
a
3
D
e
pa
r
tm
e
nt
of
A
r
ti
f
i
c
ia
l
I
nt
e
ll
ig
e
nc
e
a
nd D
a
ta
S
c
i
e
n
c
e
, S
t
J
o
s
e
ph
’
s
C
o
ll
e
ge
of
E
ngi
n
e
e
r
in
g, O
M
R
, C
he
nna
i,
I
ndi
a
4
D
e
pa
r
tm
e
nt
of
C
S
E
, N
ig
a
m I
ns
ti
tu
t
e
of
E
ngi
n
e
e
r
in
g a
nd
T
e
c
h
no
l
o
g
y
, N
I
E
T
, U
G
P
G
D
ip
l
o
ma
E
ngi
ne
e
r
in
g, O
di
s
ha
, I
ndi
a
5
D
e
pa
r
tm
e
nt
of
C
o
mpu
te
r
S
c
ie
n
c
e
a
nd
E
ngi
n
e
e
r
in
g,
K
o
n
e
r
u
L
a
ks
hma
ia
h E
duc
a
ti
o
n
F
o
unda
ti
o
n,
H
y
de
r
a
ba
d, I
ndi
a
6
D
e
pa
r
tm
e
nt
of
A
I
M
L
, R
a
mde
o
ba
ba
U
ni
v
e
r
s
it
y
, N
a
gpur
, I
ndi
a
Ar
t
ic
l
e
I
n
f
o
AB
S
T
RA
CT
A
r
ti
c
le
h
is
tor
y
:
R
e
c
e
i
ve
d
J
a
n
8,
2026
R
e
vi
s
e
d
A
ug
27,
2026
A
c
c
e
pt
e
d
S
e
p
3
,
2026
Fi
n
a
n
ci
al
s
t
o
c
k
m
ark
e
t
t
r
e
n
d
d
eci
s
i
o
n
s
y
s
t
em
s
s
u
p
p
o
rt
t
h
e
i
n
v
e
s
t
men
t
d
eci
s
i
o
n
s
b
y
p
r
e
d
i
c
t
i
n
g
fu
t
u
r
e
p
ri
ce
mo
v
eme
n
t
s
an
d
a
n
al
y
z
i
n
g
l
arg
e
-
s
c
al
e
m
ark
e
t
d
at
a.
D
e
s
p
i
t
e
,
I
o
T
e
n
ab
l
ed
t
rad
i
n
g
s
y
s
t
em
s
c
o
n
t
i
n
u
o
u
s
l
y
c
o
l
l
ec
t
s
an
d
e
x
ch
an
g
e
s
en
t
i
me
n
t
an
d
m
ark
e
t
i
n
f
o
r
m
at
i
o
n
ar
e
e
x
p
o
s
e
d
t
o
a
d
v
e
rs
ari
a
l
at
t
ac
k
s
,
w
h
i
ch
mi
s
l
e
ad
s
a
u
t
o
m
at
ed
t
rad
i
n
g
al
g
o
ri
t
h
m
s
a
n
d
r
e
s
u
l
t
s
i
n
i
n
a
cc
u
rat
e
i
n
v
e
s
t
me
n
t
d
ec
i
s
i
o
n
s
y
s
t
em
s
.
T
o
fa
ce
t
h
i
s
c
h
a
l
l
en
g
e,
t
h
i
s
s
t
u
d
y
p
ro
p
o
s
e
s
an
ad
v
e
rs
ari
al
-
r
e
s
i
l
i
e
n
t
e
x
p
l
ai
n
ab
l
e
art
i
f
i
ci
al
i
n
t
el
l
i
g
e
n
ce
fo
r
s
e
n
t
i
me
n
t
-
d
r
i
v
en
s
t
o
ck
t
rad
i
n
g
(A
R
X
A
I
-
SST
)
ar
c
h
i
t
ec
t
u
r
e
.
T
h
e
p
ro
p
o
s
e
d
framew
o
rk
c
o
m
b
i
n
e
s
g
rap
h
n
e
u
ra
l
n
e
t
w
o
r
k
s
(G
N
N
)
f
o
r
ad
v
e
rs
ari
al
at
t
ac
k
d
e
t
ec
t
i
o
n
,
F
i
n
B
E
RT
b
as
e
d
s
en
t
i
me
n
t
r
e
p
re
s
e
n
t
at
i
o
n
,
B
i
L
ST
M
w
i
t
h
at
t
e
n
t
i
o
n
(B
i
L
ST
M
-
A
t
t
)
fo
r
t
h
e
s
en
t
i
me
n
t
an
a
l
y
s
i
s
,
an
d
S
H
A
P
-
b
as
ed
e
x
p
l
ai
n
ab
l
e
art
i
fi
ci
al
i
n
t
e
l
l
i
g
e
n
ce
fo
r
en
h
a
n
c
i
n
g
m
o
d
e
l
i
n
t
e
rp
r
e
t
ab
i
l
i
t
y
.
T
h
e
e
x
p
e
ri
men
t
at
i
o
n
o
u
t
co
me
s
show
t
h
at
t
h
e
p
ro
p
o
s
e
d
fra
mew
o
rk
a
cc
o
m
p
l
i
s
h
e
s
h
i
g
h
e
r
a
ccu
ra
cy
an
d
ro
b
u
s
t
n
e
s
s
co
m
p
ared
w
i
t
h
t
h
e
e
x
i
s
t
i
n
g
ap
p
ro
a
c
h
e
s
w
h
i
l
e
p
ro
v
i
d
i
n
g
t
ran
s
p
are
n
t
i
n
s
i
g
h
t
s
i
n
t
o
t
h
e
t
rad
i
n
g
d
ec
i
s
i
o
n
s
.
T
h
e
p
ro
p
o
s
e
d
mo
d
e
l
o
ff
e
rs
an
e
ff
i
ci
e
n
t
an
d
r
el
i
ab
l
e
s
o
l
u
t
i
o
n
fo
r
s
ecu
re
s
e
n
t
i
men
t
d
r
i
v
en
s
t
o
c
k
t
rad
i
n
g
i
n
Io
T
en
ab
l
e
d
fi
n
an
ci
al
en
v
i
ro
n
me
n
t
s
.
K
e
y
w
o
r
d
s
:
A
d
v
e
r
s
a
r
i
a
l
a
t
t
a
c
ks
De
e
p
l
e
a
r
ni
ng
F
e
a
t
ur
e
s
e
l
e
c
t
i
o
n
F
i
na
n
c
i
a
l
i
n
du
s
t
r
y
H
y
pe
r
pa
r
a
m
e
t
e
r
t
uni
n
g
I
oT
s
S
e
n
t
i
m
e
n
t
a
n
a
ly
s
i
s
S
Ha
p
l
e
y
a
dd
i
t
i
ve
e
x
p
l
a
na
t
i
o
n
s
S
to
c
k
m
a
r
ke
t
Th
i
s
i
s
a
n
o
p
en
a
c
ces
s
a
r
t
i
c
l
e
u
n
d
e
r
t
h
e
CC
B
Y
-
SA
l
i
cen
s
e.
C
or
r
e
s
pon
din
g
A
u
th
or
:
S
ur
e
s
h
T
a
l
a
m
a
l
a
S
c
h
o
o
l
o
f
B
us
i
ne
s
s
,
A
d
i
t
y
a
U
ni
ve
r
s
i
t
y
S
ur
a
m
pa
l
e
m
,
K
a
k
i
na
da
,
An
d
h
r
a
P
r
a
de
s
h
-
533437,
I
n
d
i
a
E
m
a
i
l
:
dr
.
s
ur
e
s
h
s
o
ny
@g
m
a
il
.
c
o
m
1.
I
NT
RODU
C
T
I
ON
W
i
t
h
t
h
e
f
a
s
t
e
x
pa
n
s
i
o
n
o
f
t
h
e
I
oT
,
m
a
ny
d
e
vi
c
e
s
a
r
e
c
o
n
n
e
c
t
e
d
to
t
h
e
I
o
T
[
1]
.
T
h
e
s
e
I
o
T
de
vi
c
e
s
w
i
ll
pr
o
duc
e
a
m
a
s
s
i
ve
qua
n
t
i
t
y
o
f
da
t
a
t
h
a
t
c
a
n
c
o
n
s
u
m
e
a
hi
g
h
a
m
o
un
t
o
f
c
o
m
put
i
n
g
a
n
d
s
to
r
a
ge
r
e
s
o
ur
c
e
s
o
f
a
n
I
o
T
.
T
h
e
r
e
f
o
r
e
,
t
h
e
pur
po
s
e
i
s
to
r
e
qui
r
e
lo
we
r
de
l
a
y
,
w
i
d
e
r
c
o
n
n
e
c
t
i
o
n
,
a
n
d
b
e
t
t
e
r
c
o
n
t
r
o
l
o
f
c
l
o
ud
c
o
m
put
i
n
g
i
n
I
oT
s
,
whi
c
h
i
s
e
v
o
l
vi
ng
i
n
t
o
a
w
o
r
l
dw
i
de
d
i
s
t
r
i
b
ut
e
d
n
e
t
w
o
r
k
[
2
]
.
Due
to
t
h
e
d
i
f
f
i
c
u
l
t
i
e
s
o
f
e
xi
s
t
i
n
g
n
e
t
wo
r
k
s
e
c
ur
i
t
y
i
n
I
o
T
s
,
t
h
e
c
l
a
s
s
i
c
a
l
s
e
c
ur
i
t
y
t
e
c
hn
o
l
o
g
y
i
s
i
ne
f
f
e
c
t
i
ve
.
Al
s
o
,
t
h
e
e
r
u
pt
i
o
n
o
f
i
nc
r
e
a
s
i
n
g
n
e
t
wo
r
k
da
t
a
r
a
i
s
e
s
n
e
w
ne
e
ds
f
o
r
t
h
e
s
t
udy
o
f
n
e
t
wo
r
k
s
e
c
ur
i
t
y
.
M
a
l
i
c
i
o
us
a
tt
a
c
ks
o
n
l
a
n
gua
g
e
s
y
s
t
e
m
s
h
a
ve
de
v
e
l
o
pe
d
a
s
i
g
nif
i
c
a
n
t
s
t
udy
f
o
c
us
i
n
n
a
t
ur
a
l
l
a
n
gua
ge
pr
o
c
e
s
s
i
ng
(
NL
P
)
,
pa
r
t
i
c
u
l
a
r
ly
c
o
n
c
e
r
ni
ng
l
a
r
ge
l
a
n
gua
ge
m
o
de
l
s
(
L
L
M
s
)
[
3]
.
T
h
e
s
e
a
t
t
a
c
ks
e
x
p
l
o
i
t
we
a
kn
e
s
s
e
s
i
n
m
e
t
h
o
ds
to
de
c
e
i
v
e
pr
e
d
i
c
t
i
o
n
s
,
w
i
t
h
po
s
s
i
bly
s
e
ve
r
e
c
o
n
s
e
que
n
c
e
s
i
n
pe
r
c
e
pt
i
v
e
f
i
e
l
ds
li
ke
f
i
na
n
c
e
,
whe
r
e
a
tt
a
c
ks
c
a
n
l
e
a
d
to
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t
J
I
n
f
&
C
o
m
m
u
n
T
e
c
hn
o
l
I
S
S
N:
2252
-
8776
A
dv
e
r
s
ar
ial
-
r
e
s
il
ient
iot
-
e
nabled
f
inancial
int
e
ll
ige
nc
e
s
y
s
tem
us
ing
e
x
plai
nable
…
(
Sur
e
s
h
T
alamala
)
1945
d
i
s
r
upt
i
n
g
b
u
s
i
ne
s
s
pr
o
c
e
s
s
e
s
,
b
r
e
a
k
i
ng
da
t
a
pr
i
v
a
c
y
,
s
pr
e
a
d
i
ng
m
i
s
i
n
f
o
r
m
a
t
i
o
n
,
a
n
d
f
i
na
n
c
i
a
l
l
o
s
s
e
s
[
4]
.
C
o
n
s
e
que
n
t
l
y
,
o
n
e
n
e
e
d
s
to
un
de
r
s
t
a
n
d
t
h
e
f
l
a
ws
o
f
t
h
e
s
e
m
e
t
h
o
ds
to
r
e
c
o
gni
s
e
a
n
d
m
o
de
r
a
t
e
t
h
e
i
r
v
u
l
ne
r
a
bil
i
t
i
e
s
.
F
i
na
n
c
i
a
l
s
e
n
t
i
m
e
n
t
a
n
a
ly
s
i
s
t
e
c
hni
que
s
t
h
a
t
c
a
n
e
x
t
r
a
c
t
i
m
po
r
t
a
n
c
e
f
r
o
m
v
a
s
t
qua
n
t
i
t
i
e
s
o
f
r
a
w
da
t
a
a
c
t
a
s
a
vi
t
a
l
p
a
r
t
i
n
s
e
n
t
i
m
e
n
t
-
b
a
s
e
d
f
i
na
n
c
i
a
l
de
c
i
s
i
o
n
s
[
5]
.
S
to
c
k
pr
e
di
c
t
i
o
n
h
a
s
a
hi
g
h
s
i
g
nif
i
c
a
n
c
e
i
n
t
h
e
f
i
na
nc
i
a
l
do
m
a
i
n
.
Dur
i
n
g
t
h
e
s
t
o
c
k
f
l
uc
t
ua
t
i
o
n
,
whi
c
h
i
s
v
e
r
y
im
pu
l
s
i
ve
,
r
e
s
e
a
r
c
h
wo
r
ke
r
s
h
a
v
e
m
a
d
e
e
v
e
r
y
e
f
f
o
r
t
to
i
m
i
t
a
t
e
t
h
e
s
to
c
k
di
f
f
e
r
e
n
c
e
s
i
nc
e
a
c
o
m
pa
r
a
t
i
v
e
ly
s
e
n
s
i
b
l
e
pr
e
d
i
c
t
i
o
n
c
a
n
m
a
ke
h
u
ge
pr
o
f
i
t
s
a
n
d
h
e
l
p
de
c
r
e
a
s
e
r
i
s
k
s
[
6]
.
E
c
o
n
o
m
i
c
c
o
n
d
i
t
i
o
n
s
,
I
nv
e
s
to
r
s
’
s
e
n
t
i
m
e
n
t
s
,
a
n
d
publi
c
s
e
n
t
i
m
e
n
t
s
a
r
e
a
ll
e
s
s
e
n
t
i
a
l
va
r
i
a
bl
e
s
t
h
a
t
m
u
s
t
b
e
s
upp
l
e
m
e
n
t
e
d
i
n
t
o
a
s
to
c
k
pr
e
di
c
t
i
o
n
m
o
de
l
[
7]
.
E
s
pe
c
i
a
ll
y
,
s
e
n
t
i
m
e
n
t
a
n
a
ly
s
i
s
(
S
A
)
wa
s
r
e
c
o
gni
s
e
d
a
s
a
n
e
f
f
e
c
t
i
v
e
t
oo
l
i
n
t
h
e
l
a
s
t
f
e
w
y
e
a
r
s
.
Not
j
us
t
pr
i
c
e
a
n
d
pr
i
c
e
-
r
e
l
a
t
e
d
v
a
l
ue
s
,
r
e
s
e
a
r
c
h
wo
r
ke
r
s
h
a
v
e
pa
i
d
m
o
r
e
c
o
n
s
i
de
r
a
t
i
o
n
t
o
t
h
e
e
x
t
r
a
c
t
i
o
n
o
f
s
e
n
t
i
m
e
n
t
da
t
a
f
r
o
m
j
o
ur
n
a
l
s
,
n
e
w
s
pa
pe
r
s
,
a
n
d
a
l
s
o
s
o
c
i
a
l
m
e
d
i
a
[
8]
.
A
dve
r
s
a
r
i
a
l
a
t
t
a
c
ks
i
n
m
a
c
hi
ne
l
e
a
r
ni
ng
(
M
L
)
m
e
t
h
o
ds
a
r
e
a
pr
i
m
a
r
y
c
o
n
c
e
r
n
,
a
s
t
h
e
s
e
a
tt
a
c
ks
m
a
y
s
e
r
i
o
us
ly
a
f
f
e
c
t
t
h
e
pe
r
f
o
r
m
a
n
c
e
o
f
e
v
e
n
t
h
e
m
o
s
t
de
v
e
l
o
pe
d
s
y
s
t
e
m
s
[
9]
.
T
h
e
s
e
a
tt
a
c
ks
h
a
v
e
s
e
ve
r
a
l
m
e
t
h
o
ds
a
n
d
t
y
pe
s
,
s
uc
h
a
s
e
v
a
s
i
o
n
a
tt
a
c
ks
,
wh
e
r
e
a
n
a
tt
a
c
k
e
r
de
pl
o
y
s
t
h
e
i
n
put
da
t
a
to
p
o
i
s
o
n
a
tt
a
c
ks
o
r
a
v
o
i
d
de
t
e
c
t
i
o
n
[
10]
.
Kim
e
t
al.
[
11]
pr
e
s
e
n
t
e
d
a
s
to
c
k
t
r
a
di
n
g
s
ugge
s
t
i
o
n
m
e
t
h
o
d,
whi
c
h
c
o
m
bi
ne
s
r
e
l
a
t
i
v
e
s
t
r
e
n
gt
h
i
n
de
x
(
R
S
I
)
wi
t
h
t
h
e
n
e
w
s
S
A
f
o
r
pr
o
vi
d
i
ng
k
n
o
w
l
e
d
ge
a
bl
e
b
u
y
–
s
e
ll
de
c
i
s
i
o
ns
.
T
hi
s
m
e
t
h
o
d
e
m
p
l
o
y
e
d
a
r
u
l
e
-
dr
i
v
e
n
NL
P
t
e
c
h
ni
que
to
a
n
a
ly
s
e
t
h
e
c
ur
r
e
n
t
n
e
ws
pa
pe
r
r
e
por
t
s
a
n
d
i
n
t
e
gr
a
t
e
s
t
h
e
c
o
n
s
e
que
n
t
s
e
n
t
i
m
e
n
t
s
c
o
r
e
s
a
m
o
n
g
t
h
e
R
S
I
t
h
a
t
c
a
n
f
o
l
l
o
w
t
h
e
s
to
c
k
l
e
v
e
l
.
T
h
e
a
b
o
v
e
-
m
e
n
t
i
o
n
e
d
m
e
t
h
o
d
a
l
l
o
c
a
t
e
s
a
s
e
n
t
i
m
e
n
t
s
c
o
r
e
t
h
a
t
r
e
pl
i
c
a
t
e
s
m
a
r
ke
t
s
e
n
t
i
m
e
n
t
,
whi
l
e
t
h
e
R
S
I
de
t
e
c
t
s
o
v
e
r
s
o
l
d
o
r
o
v
e
r
b
o
ugh
t
s
t
a
t
e
s
.
T
hi
s
in
t
e
gr
a
t
e
d
m
e
t
h
o
d
e
n
a
bl
e
s
t
r
a
de
r
s
to
c
r
e
a
t
e
da
t
a
-
b
a
s
e
d
s
e
ll
,
b
u
y
,
o
r
h
o
l
d
de
c
i
s
i
o
ns
i
n
r
e
a
l
li
f
e
.
Na
r
a
y
a
n
a
e
t
al.
[
12]
pr
o
duc
e
d
a
hy
b
r
i
d
c
o
m
po
s
i
t
e
m
e
t
h
o
d,
whi
c
h
i
n
t
e
gr
a
t
e
s
t
r
a
di
t
i
o
n
a
l
s
t
o
c
k
m
a
r
ke
t
e
s
t
i
m
a
t
i
o
n
s
y
s
t
e
m
s
t
h
r
o
ugh
S
A
o
f
n
e
w
s
a
r
t
i
c
l
e
s
t
o
e
s
c
a
l
a
t
e
t
h
e
pr
e
di
c
t
i
o
n
a
c
c
ur
a
c
y
.
T
h
r
o
ugh
t
h
e
b
e
n
e
f
i
t
s
o
f
a
c
o
l
l
a
b
o
r
a
t
i
v
e
a
r
c
hi
t
e
c
t
ur
e
,
t
h
e
m
e
t
h
o
d
’
s
c
a
pa
bi
li
t
y
t
o
s
ui
t
t
h
e
v
a
r
y
i
ng
r
e
qu
i
r
e
m
e
n
t
s
o
f
i
nv
e
s
to
r
s
pr
o
vi
de
s
t
h
e
b
e
s
t
r
e
s
ul
t
s
i
t
h
a
s
a
tt
a
i
n
e
d
o
v
e
r
r
i
s
k
pr
o
f
i
l
e
s
.
P
a
n
d
i
r
i
e
t
al.
[
13]
de
v
e
l
o
pe
d
a
m
e
t
h
o
d
f
o
r
a
n
a
l
y
s
i
ng
t
h
e
pe
r
s
pe
c
t
i
v
e
o
f
r
e
c
ur
r
e
n
t
n
e
ur
a
l
n
e
t
wo
r
ks
(
R
NN
s
)
i
n
pe
r
f
o
r
m
i
n
g
S
A
o
f
f
i
na
n
c
ia
l
i
nf
o
r
m
a
t
i
o
n
f
o
r
i
m
pr
o
vi
ng
t
h
e
m
a
r
ke
t
pr
e
di
c
t
i
o
n
a
n
d
im
pr
o
vi
n
g
i
nve
s
t
m
e
n
t
m
e
t
h
o
ds
.
T
h
e
r
e
f
o
r
e
,
us
i
n
g
th
e
s
e
a
d
v
a
n
c
e
d
NL
P
t
e
c
hni
que
s
a
n
d
n
e
ur
a
l
n
e
t
wo
r
k
s
(
NN
)
.
Ya
n
g
e
t
al.
[
14]
pr
e
s
e
n
t
e
d
a
DL
-
b
a
s
e
d
s
t
o
c
k
m
a
r
ke
t
S
A
a
n
d
pr
e
d
i
c
t
i
o
n
m
e
t
h
o
d
t
h
a
t
de
pe
n
ds
o
n
t
h
e
c
o
m
bi
ne
d
t
e
c
hni
qu
e
o
f
L
S
T
M
a
n
d
C
NN
.
T
h
e
NL
P
t
e
c
hn
o
l
o
g
y
i
s
ut
i
li
s
e
d
f
o
r
e
x
t
r
a
c
t
i
n
g
t
h
e
s
e
n
t
i
m
e
n
t
c
h
a
r
a
c
t
e
r
i
s
t
i
c
s
o
f
e
c
o
n
o
m
i
c
n
e
w
s
a
n
d
s
o
c
i
a
l
n
e
t
wor
k
i
n
g
c
o
n
t
e
n
t,
a
n
d
a
hi
g
h
e
r
-
d
i
m
e
ns
i
o
n
f
a
c
t
o
r
s
pa
c
e
i
s
b
u
i
l
t
by
i
n
t
e
gr
a
t
i
n
g
t
h
e
m
a
r
ke
t
da
t
a
t
r
a
n
s
a
c
t
i
o
n
.
R
a
hm
a
n
e
t
al.
[
15]
pr
e
s
e
n
t
e
d
a
n
A
nn
o
t
a
to
r
s
’
I
n
s
t
r
uc
t
i
o
n
As
s
i
s
t
e
d
P
r
o
m
pt
,
a
n
i
nn
o
va
t
i
ve
a
s
s
e
s
s
m
e
n
t
pr
o
m
pt
s
i
n
t
e
n
de
d
t
o
r
e
e
v
a
l
ua
t
e
t
h
e
c
l
a
s
s
if
i
c
a
t
i
o
n
o
f
F
S
A
f
o
r
L
L
M
s
.
T
h
r
o
ugh
t
h
e
c
o
m
bi
na
t
i
o
n
o
f
c
o
m
p
l
e
t
e
t
a
s
ks
i
ni
t
i
a
l
ly
de
s
i
g
n
e
d
f
o
r
h
u
m
a
n
i
m
pr
e
s
s
i
o
n
s
w
i
t
h
t
h
e
L
L
M
s
’
pr
o
m
pt
a
r
c
hi
t
e
c
t
ur
e
,
A
I
A
P
wa
n
t
s
to
r
e
gul
a
t
e
t
h
e
c
o
m
p
r
e
h
e
n
s
i
o
n
o
f
s
e
n
t
i
m
e
n
t
o
v
e
r
e
i
t
h
e
r
m
a
c
hi
ne
o
r
h
u
m
a
n
c
l
a
r
i
f
i
c
a
t
i
o
n
s
,
g
i
vi
ng
a
b
e
t
t
e
r
a
n
d
c
o
n
t
e
x
t
-
r
e
l
a
t
e
d
f
o
un
d
a
t
i
o
n
f
o
r
S
A
.
S
m
a
t
o
v
e
t
al.
[
16]
p
r
e
s
e
n
t
e
d
a
n
i
nn
o
v
a
t
i
v
e
m
e
t
h
o
d
f
o
r
S
A
,
pa
r
t
i
c
u
l
a
r
ly
de
v
e
l
o
pe
d
f
o
r
f
o
r
e
c
a
s
t
i
n
g
s
t
o
c
k
m
a
r
ke
t
a
c
t
i
vi
t
i
e
s
.
T
h
e
a
uth
o
r
us
e
s
c
o
m
p
l
e
x
m
a
t
h
e
m
a
t
i
c
a
l
m
o
de
l
li
ng
m
e
t
h
o
ds
to
i
m
pr
o
v
e
b
o
t
h
t
h
e
i
n
t
e
r
pr
e
t
a
bi
li
t
y
a
n
d
pr
e
c
i
s
i
o
n
o
f
t
h
e
t
e
c
h
ni
que
s
.
E
xi
s
t
i
n
g
s
t
ud
i
e
s
by
Kim
e
t
al.
[
11]
,
N
a
r
a
y
a
n
a
e
t
al.
[
12]
,
P
a
n
d
i
r
i
e
t
al.
[
13]
,
a
n
d
Ya
n
g
e
t
al.
[
14]
m
a
i
n
ly
f
o
c
u
s
o
n
t
h
e
s
e
n
t
i
m
e
n
t
a
n
a
ly
s
i
s
-
b
a
s
e
d
s
t
oc
k
pr
e
di
c
t
i
o
n
by
c
o
m
bi
n
i
ng
C
NN
,
L
S
T
M
,
R
S
I
a
n
d
NL
P
m
o
de
l
s
f
o
r
e
nh
a
n
c
i
ng
t
r
a
di
n
g
de
c
i
s
i
o
n
s
.
R
a
hm
a
n
e
t
al.
[
15]
a
n
d
S
m
a
t
o
v
e
t
al.
[
16]
f
ur
t
h
e
r
i
m
pr
o
v
e
f
i
na
n
c
i
a
l
s
e
n
t
i
m
e
n
t
a
n
a
ly
s
i
s
by
a
d
v
a
n
c
e
d
pr
o
m
pt
-
b
a
s
e
d
l
e
a
r
ni
ng
a
n
d
m
a
t
h
e
m
a
t
i
c
a
l
m
o
de
ll
i
ng
m
e
t
h
o
ds
.
H
o
w
e
v
e
r
,
t
h
e
c
h
a
ll
e
n
g
e
s
a
s
s
o
c
i
a
t
e
d
w
i
t
h
s
e
c
ur
e
f
i
na
n
c
i
a
l
da
t
a
t
r
a
n
s
m
i
s
s
i
o
n
,
I
o
T
-
b
a
s
e
d
v
u
l
ne
r
a
bil
i
t
i
e
s
a
n
d
a
d
v
e
r
s
a
r
i
a
l
a
tt
a
c
ks
r
e
m
a
i
n
i
ns
u
f
f
i
c
i
e
n
t
l
y
a
ddr
e
s
s
e
d.
T
h
e
r
e
f
o
r
e
,
a
n
i
n
t
e
gr
a
t
e
d
I
oT
e
n
a
bl
e
d
a
d
v
e
r
s
a
r
i
a
l
d
e
f
e
ns
e
a
r
c
hi
t
e
c
t
ur
e
f
o
r
r
e
l
i
a
bl
e
a
n
d
s
e
c
ur
e
s
e
n
t
i
m
e
n
t
dr
i
v
e
n
s
t
o
c
k
m
a
r
ke
t
a
n
a
l
y
s
i
s
i
s
s
t
i
ll
n
e
e
de
d.
T
h
e
r
a
p
i
d
e
x
pa
n
s
i
o
n
o
f
i
nf
o
r
m
a
t
i
o
n
a
n
d
c
o
m
m
u
nica
t
i
o
n
t
e
c
hn
o
l
o
g
y
(
I
C
T
)
h
a
s
t
r
a
n
s
f
o
r
m
e
d
f
i
na
n
c
i
a
l
i
n
t
e
ll
i
ge
n
c
e
s
y
s
t
e
m
s
by
f
a
c
il
i
t
a
t
i
n
g
de
c
i
s
i
o
n
-
m
a
k
i
n
g,
pr
o
c
e
s
s
i
n
g
a
n
d
r
e
a
l
-
t
i
m
e
da
t
a
a
c
qui
s
i
t
i
o
n
,
t
h
r
o
ugh
i
n
t
e
r
c
o
nn
e
c
t
e
d
t
e
c
h
ni
que
s
.
P
a
r
t
i
c
u
l
a
r
ly
,
I
o
T
-
e
n
a
bled
f
i
na
nc
i
a
l
p
l
a
t
f
o
r
m
s
pr
o
vi
d
e
c
o
n
t
i
n
uo
us
a
c
c
e
s
s
to
v
a
r
i
o
us
m
a
r
ke
t
i
nf
o
r
m
a
t
i
o
n
,
i
n
c
l
ud
i
ng
s
e
n
t
i
m
e
n
t
-
r
e
l
a
t
e
d
da
t
a
,
i
nve
s
t
o
r
i
n
t
e
r
a
c
t
i
o
n
s
a
n
d
t
r
a
d
i
n
g
a
c
t
i
vi
t
i
e
s
.
De
s
p
i
t
e
,
t
h
e
i
nc
r
e
a
s
i
n
g
d
e
pe
n
de
n
c
e
o
n
A
I
dr
i
v
e
n
t
r
a
d
i
n
g
s
y
s
t
e
m
s
i
n
t
r
o
duc
e
s
s
e
c
ur
i
t
y
c
h
a
ll
e
n
ge
s
,
s
pe
c
i
f
i
c
a
ll
y
a
d
v
e
r
s
a
r
i
a
l
a
tt
a
c
ks
whi
c
h
m
a
ni
pu
l
a
t
e
s
i
n
put
da
t
a
a
n
d
de
gr
a
de
t
h
e
r
e
l
i
a
bil
i
t
y
o
f
t
h
e
a
uto
m
a
t
e
d
i
nv
e
s
t
m
e
n
t
d
e
c
i
s
i
o
n
s
.
He
n
c
e
,
de
v
e
l
o
p
i
ng
a
d
v
e
r
s
a
r
i
a
l
r
e
s
il
i
e
n
t
t
e
c
h
ni
que
s
i
s
e
s
s
e
n
t
i
a
l
f
o
r
e
ns
ur
i
n
g
t
r
us
t
w
o
r
t
hy
a
n
d
s
e
c
ur
e
f
i
na
n
c
i
a
l
i
n
t
e
ll
i
ge
n
c
e
.
T
h
e
r
e
f
o
r
e
,
t
hi
s
s
t
ud
y
pr
o
po
s
e
s
a
n
a
dv
e
r
s
a
r
i
a
l
-
r
e
s
il
i
e
n
t
e
x
p
l
a
i
na
bl
e
a
r
t
i
f
i
c
i
a
l
i
n
t
e
ll
i
g
e
n
c
e
f
o
r
s
e
n
t
i
m
e
n
t
-
dr
i
v
e
n
s
t
o
c
k
t
r
a
di
n
g
(
AR
X
A
I
-
S
S
T
)
a
r
c
h
i
t
e
c
t
ur
e
.
T
h
e
ke
y
c
o
n
t
r
i
b
ut
i
o
ns
a
r
e
g
i
v
e
n
a
s
f
o
l
l
o
w
s
.
De
s
i
g
n
a
n
a
d
v
e
r
s
a
r
i
a
l
a
t
t
a
c
k
de
t
e
c
t
i
o
n
m
o
du
l
e
b
a
s
e
d
o
n
gr
a
ph
n
e
ur
a
l
n
e
t
wo
r
ks
(
GN
Ns
)
.
De
v
e
l
o
p
a
s
e
n
t
i
m
e
n
t
a
n
a
ly
s
i
s
m
o
du
l
e
us
i
ng
t
h
e
B
i
L
S
T
M
-
A
t
t
m
o
de
l
.
E
m
p
l
o
y
t
h
e
A
da
m
o
pt
i
mi
s
e
r
f
o
r
hy
pe
r
pa
r
a
m
e
t
e
r
t
u
ni
ng.
I
n
t
e
gr
a
t
e
S
HA
P
-
b
a
s
e
d
X
A
I
to
p
r
o
vi
de
c
l
e
a
r
i
ns
i
g
ht
s
i
n
t
o
t
h
e
m
o
de
l
’
s
pr
e
d
i
c
t
i
o
n
s
.
C
o
n
duc
t
e
x
t
e
n
s
i
v
e
e
x
pe
r
im
e
n
t
a
l
e
v
a
l
ua
t
i
o
ns
o
n
b
e
n
c
hm
a
r
k
da
t
a
s
e
t
s
de
m
o
n
s
t
r
a
t
i
n
g
s
upe
r
i
o
r
pe
r
f
o
r
m
a
n
c
e
a
c
r
o
s
s
m
u
l
t
i
p
l
e
e
v
a
l
ua
t
i
o
n
m
e
t
r
i
c
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8776
I
n
t
J
I
n
f
&
C
o
m
m
u
n
T
e
c
hn
o
l
,
Vo
l
.
1
5
,
N
o.
4
,
D
e
c
e
m
be
r
20
2
6
:
1944
-
1953
1946
2.
T
HE
P
ROP
OS
E
D
M
E
T
HO
D
I
n
t
hi
s
s
t
ud
y
,
t
h
e
A
R
X
A
I
-
S
S
T
a
r
c
hi
t
e
c
t
ur
e
f
o
l
l
o
ws
a
s
t
r
uc
t
u
r
e
d
w
o
r
kf
l
o
w
f
o
r
e
x
p
l
a
i
na
bl
e
s
e
n
t
i
m
e
n
t
-
dr
i
v
e
n
,
a
c
c
ur
a
t
e
a
n
d
s
e
c
ur
e
s
t
o
c
k
tr
a
di
n
g,
a
s
g
i
ve
n
i
n
F
i
g
ur
e
1.
I
ni
t
i
a
ll
y
,
t
h
e
a
d
v
e
r
s
a
r
ia
l
a
t
t
a
c
k
de
t
e
c
t
i
o
n
m
o
du
l
e
pr
e
pr
o
c
e
s
s
e
s
t
h
e
i
nput
da
t
a
by
f
e
a
t
ur
e
s
c
a
l
i
ng
a
n
d
R
F
E
f
o
r
s
e
l
e
c
t
i
n
g
t
h
e
m
o
s
t
r
e
l
e
v
a
n
t
f
e
a
t
ur
e
s
,
whi
c
h
a
r
e
t
h
e
n
c
a
t
e
gor
i
z
e
d
us
i
n
g
GN
N
to
de
t
e
c
t
t
h
e
a
dv
e
r
s
a
r
i
a
l
a
c
t
i
vi
t
i
e
s
[
17]
.
Ne
x
t
,
t
h
e
s
e
n
t
i
m
e
n
t
a
n
a
ly
s
i
s
m
o
du
l
e
pr
e
pr
o
c
e
s
s
e
s
f
i
na
nc
i
a
l
t
e
x
t
,
ge
ts
c
o
n
t
e
x
t
ua
l
e
m
be
dd
i
n
g
s
by
F
i
nB
E
R
T
,
a
n
d
c
a
t
e
go
r
i
z
e
s
m
a
r
ke
t
s
e
n
t
i
m
e
n
t
by
a
B
i
L
S
T
M
-
A
t
t
m
o
de
l
o
p
t
i
mi
z
e
d
t
h
r
o
ugh
A
da
m
o
pt
i
mi
z
e
r
.
F
i
na
ll
y
,
S
HA
P
is
us
e
d
t
o
i
n
t
e
r
pr
e
t
t
h
e
m
o
de
l
pr
e
d
i
c
t
i
o
n
s
by
f
i
nd
i
ng
t
h
e
ke
y
s
e
n
t
i
m
e
n
t
f
e
a
t
ur
e
s
whi
c
h
i
nf
l
ue
n
c
e
s
t
h
e
tr
a
d
i
n
g
de
c
i
s
i
o
ns
.
F
i
gur
e
1.
Vi
s
ua
l
o
v
e
r
vi
e
w
o
f
t
h
e
AR
X
A
I
-
S
S
T
s
t
r
u
c
t
ur
e
I
n
t
h
e
pr
o
p
o
s
e
d
a
r
c
hi
t
e
c
t
ur
e
,
c
o
n
n
e
c
t
e
d
d
i
g
i
t
a
l
pl
a
t
f
o
r
m
s
a
n
d
I
o
T
e
n
a
bl
e
d
f
i
na
n
c
i
a
l
de
vi
c
e
s
c
o
n
t
i
n
uo
us
l
y
ge
n
e
r
a
t
e
s
c
o
m
m
u
ni
c
a
t
i
o
n
a
n
d
t
r
a
n
s
a
c
t
i
o
n
da
t
a
,
t
h
a
t
wa
s
t
r
a
n
s
m
i
t
t
e
d
by
t
h
e
I
C
T
i
nf
r
a
s
t
r
uc
t
ur
e
to
t
h
e
pr
o
p
o
s
e
d
m
o
de
l
.
T
h
e
c
o
l
l
e
c
t
e
d
da
t
a
g
o
t
h
r
o
ug
h
pr
e
pr
o
c
e
s
s
i
n
g,
a
dv
e
r
s
a
r
i
a
l
de
t
e
c
t
i
o
n
,
s
e
n
t
i
m
e
n
t
a
n
a
ly
s
i
s
,
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t
J
I
n
f
&
C
o
m
m
u
n
T
e
c
hn
o
l
I
S
S
N:
2252
-
8776
A
dv
e
r
s
ar
ial
-
r
e
s
il
ient
iot
-
e
nabled
f
inancial
int
e
ll
ige
nc
e
s
y
s
tem
us
ing
e
x
plai
nable
…
(
Sur
e
s
h
T
alamala
)
1947
a
n
d
pr
e
d
i
c
t
i
o
n
f
o
r
s
uppo
r
t
i
n
g
s
e
c
ur
e
f
i
na
n
c
i
a
l
de
c
is
i
o
n
m
a
k
i
ng.
F
ur
t
h
e
r
m
o
r
e
,
t
h
e
S
HA
P
ba
s
e
d
e
x
p
l
a
in
a
bil
i
t
y
m
o
du
l
e
s
h
o
ws
t
r
a
n
s
pa
r
e
n
t
i
n
t
e
r
pr
e
t
a
t
i
o
n
s
w
i
t
hi
n
f
in
a
n
c
i
a
l
I
C
T
i
nf
r
a
s
t
r
uc
t
ur
e
s
.
GN
N,
F
i
n
B
E
R
T
,
B
i
L
S
T
M
-
A
t
t,
a
n
d
S
H
A
P
i
s
s
e
l
e
c
t
e
d
f
o
r
e
x
p
l
a
i
na
bl
e
a
n
d
s
e
c
ur
e
f
i
na
n
c
i
a
l
i
n
t
e
ll
i
ge
n
c
e
.
F
o
r
a
dv
e
r
s
a
r
i
a
l
a
t
t
a
c
k
de
t
e
c
t
i
o
n
GN
N
i
s
u
s
e
d,
due
to
i
t
s
c
a
pa
bil
i
t
y
o
f
c
a
pt
ur
i
n
g
c
o
m
p
l
e
x
r
e
l
a
t
i
o
n
s
hi
ps
o
ve
r
i
n
t
e
r
c
o
n
n
e
c
t
e
d
I
o
T
b
a
s
e
d
t
r
a
di
n
g
da
t
a
.
F
i
nB
E
R
T
i
s
us
e
d
f
o
r
s
e
n
t
i
m
e
n
t
f
e
a
t
ur
e
e
x
t
r
a
c
t
i
o
n
a
s
i
t
pr
o
vi
de
s
do
m
a
i
n
s
pe
c
if
i
c
c
o
n
t
e
x
t
ua
l
r
e
pr
e
s
e
n
t
a
t
i
o
n
s
.
B
i
L
S
T
M
-
A
t
t
i
s
u
s
e
d
t
o
c
a
pt
u
r
e
s
e
que
n
t
i
a
l
s
e
n
t
i
m
e
n
t
de
pe
n
de
n
c
i
e
s
.
F
i
na
ll
y
,
S
HA
P
i
s
i
n
t
e
gr
a
t
e
d
f
o
r
e
nh
a
n
c
i
n
g
m
o
de
l
t
r
a
ns
pa
r
e
n
c
y
.
T
h
e
AR
X
A
I
-
S
S
T
a
r
c
hi
t
e
c
t
u
r
e
i
s
de
s
i
g
ne
d
f
o
r
I
oT
e
n
a
bl
e
d
f
i
na
nc
i
a
l
e
nvi
r
o
nm
e
n
t
s
,
wh
e
r
e
m
a
r
ke
t
i
n
f
o
r
m
a
t
i
o
n
,
r
e
l
a
t
e
d
c
o
m
m
u
ni
c
a
t
i
o
n
s
a
n
d
f
i
na
n
c
i
a
l
t
r
a
n
s
a
c
t
i
o
n
da
t
a
a
r
e
c
o
n
t
i
n
uo
us
ly
c
r
e
a
t
e
d.
T
h
e
s
e
h
e
t
e
r
o
ge
n
e
o
us
da
t
a
s
t
r
e
a
m
s
a
r
e
s
e
n
t
t
h
r
o
ugh
t
h
e
I
C
T
f
r
a
m
e
t
o
t
h
e
pr
o
p
o
s
e
d
m
o
de
l
f
o
r
pr
e
pr
o
c
e
s
s
i
n
g,
s
to
c
k
t
r
e
n
d
pr
e
di
c
t
i
o
n
,
a
dv
e
r
s
a
r
i
a
l
de
t
e
c
t
i
o
n
,
a
n
d
s
e
n
t
i
m
e
n
t
a
n
a
ly
s
i
s
.
2.
1.
D
at
a
p
r
e
p
r
o
c
e
s
s
in
g
an
d
f
e
at
u
r
e
s
c
a
l
in
g
T
h
e
c
o
l
l
e
c
t
e
d
da
t
a
s
e
t
s
we
r
e
pr
e
pr
o
c
e
s
s
e
d
f
o
r
r
e
m
o
vi
ng
i
nc
o
n
s
i
s
t
e
n
c
i
e
s
,
e
l
im
i
na
t
i
n
g
i
r
r
e
l
e
va
n
t
i
n
f
o
r
m
a
t
i
o
n
,
a
n
d
to
c
o
n
v
e
r
t
t
h
e
da
t
a
i
n
t
o
t
h
e
we
l
l
-
m
a
t
c
h
e
d
f
o
r
m
a
t
s
f
o
r
t
h
e
m
o
de
l
tr
a
i
ni
ng.
T
h
e
pr
o
c
e
s
s
e
d
da
t
a
wa
s
t
h
e
n
s
p
l
i
t
t
e
d
to
t
r
a
i
ni
ng
a
n
d
t
e
s
t
i
n
g
ph
a
s
e
s
f
o
r
t
h
e
pe
r
f
o
r
m
a
n
c
e
a
s
s
e
s
s
m
e
n
t
.
F
e
a
t
u
r
e
s
c
a
l
i
ng
wa
s
u
t
i
li
z
e
d
to
e
n
s
ur
e
t
h
a
t
a
l
l
t
he
i
n
put
f
e
a
t
ur
e
s
we
r
e
c
h
a
n
ge
d
i
n
t
o
a
c
o
n
s
i
s
t
e
n
t
r
a
n
ge
,
whi
c
h
pr
e
v
e
n
t
s
t
h
e
f
e
a
t
ur
e
s
w
i
t
h
l
a
r
ge
r
v
a
l
ue
s
f
r
o
m
c
o
n
t
r
o
l
l
i
ng
t
h
e
l
e
a
r
ni
ng
pr
o
c
e
s
s
.
2.
2.
A
d
ve
r
s
a
r
ia
l
n
ois
e
in
f
in
an
c
ial
s
e
n
t
im
e
n
t
A
d
v
e
r
s
a
r
i
a
l
n
o
i
s
e
r
e
f
e
r
s
t
o
s
m
a
ll
v
a
r
i
a
t
i
o
ns
i
n
s
o
c
i
a
l
m
e
d
i
a
c
o
n
t
e
n
t
,
r
e
p
o
r
t
s
o
r
f
i
na
n
c
i
a
l
n
e
ws
t
h
a
t
c
a
n
m
i
s
l
e
a
d
s
e
n
t
i
m
e
n
t
a
l
c
l
a
s
s
if
i
c
a
t
i
o
n
m
o
de
l
s
w
i
t
h
o
u
t
s
ub
s
t
a
n
t
i
a
l
ly
a
l
t
e
r
i
n
g
t
h
e
or
i
g
i
na
l
m
e
a
ni
ng.
T
h
e
pr
o
p
o
s
e
d
m
o
dul
e
i
de
n
t
i
f
i
e
s
s
uc
h
v
a
r
i
e
d
i
n
put
s
b
e
f
o
r
e
s
e
n
t
i
m
e
n
t
a
n
a
ly
s
i
s
,
t
h
us
i
m
pr
o
vi
ng
t
h
e
r
e
l
i
a
bil
i
t
y
a
n
d
r
o
b
us
t
n
e
s
s
o
f
t
h
e
f
r
a
m
e
wo
r
k.
2.
3.
A
d
ve
r
s
a
r
ia
l
at
t
ac
k
d
e
t
e
c
t
ion
m
od
u
l
e
T
h
e
a
dve
r
s
a
r
i
a
l
a
t
t
a
c
k
de
t
e
c
t
i
o
n
m
o
du
l
e
i
s
de
s
i
g
n
e
d
to
i
de
n
t
i
f
y
m
a
li
c
i
o
us
a
c
t
i
vi
t
i
e
s
t
h
a
t
m
a
y
c
o
m
pr
o
m
i
s
e
t
h
e
r
e
l
i
a
bil
i
t
y
o
f
s
e
n
t
i
m
e
n
t
-
dr
i
v
e
n
s
to
c
k
tr
a
di
n
g
de
c
i
s
i
o
n
s
.
I
ni
t
i
a
ll
y
,
t
h
e
i
nput
da
t
a
a
r
e
pr
e
-
pr
o
c
e
s
s
e
d
to
e
n
s
ur
e
c
o
n
s
i
s
t
e
nc
y
a
n
d
im
pr
o
v
e
m
o
de
l
t
r
a
i
ni
ng.
R
e
c
ur
s
i
v
e
f
e
a
t
ur
e
e
l
im
i
na
t
i
o
n
(
R
F
E
)
i
s
t
h
e
n
ut
i
li
z
e
d
to
c
h
o
o
s
e
t
h
e
i
nf
o
r
m
a
t
i
v
e
a
tt
r
i
b
ut
e
s
by
i
t
e
r
a
t
i
v
e
ly
r
e
m
o
vi
ng
l
e
s
s
r
e
l
e
v
a
n
t
a
tt
r
i
b
ut
e
s
,
t
h
e
r
e
by
r
e
duc
i
n
g
r
e
dun
da
n
c
y
a
n
d
im
pr
o
vi
ng
c
l
a
s
s
i
f
i
c
a
t
i
o
n
p
e
r
f
o
r
m
a
n
c
e
[
18]
.
T
h
e
c
h
o
s
e
n
f
e
a
t
ur
e
s
c
a
n
b
e
s
u
b
s
e
que
n
t
l
y
pr
o
vi
de
d
to
a
GN
N
f
o
r
a
dv
e
r
s
a
r
i
a
l
a
t
t
a
c
k
de
t
e
c
t
i
o
n.
GN
N
i
s
a
do
pt
e
d
b
e
c
a
us
e
i
t
e
f
f
e
c
t
i
v
e
ly
c
a
pt
ur
e
s
s
tr
uc
t
u
r
a
l
r
e
l
a
t
i
o
n
s
hi
p
s
a
m
o
n
g
i
n
t
e
r
c
o
nn
e
c
t
e
d
f
i
na
n
c
i
a
l
da
t
a
,
e
n
a
bli
ng
r
o
b
us
t
i
d
e
n
t
i
f
i
c
a
t
i
o
n
o
f
a
d
v
e
r
s
a
r
i
a
l
p
a
tt
e
r
n
s
i
n
gr
a
ph
-
ba
s
e
d
f
i
na
nc
i
a
l
n
e
t
wo
r
ks
[
19
]
.
2.
4.
S
e
n
t
im
e
n
t
an
al
ys
is
m
od
u
le
T
h
e
S
A
m
o
du
l
e
e
v
a
l
ua
t
e
s
m
a
r
ke
t
s
e
n
t
i
m
e
n
t
f
r
o
m
t
h
e
t
e
x
t
ua
l
f
i
na
nc
i
a
l
da
t
a
.
I
ni
t
i
a
ll
y
,
t
h
e
r
a
w
t
e
x
t
i
s
pr
e
pr
o
c
e
s
s
e
d
t
o
i
m
pr
o
v
e
t
h
e
da
t
a
qua
l
i
t
y
.
F
i
nB
E
R
T
i
s
t
h
e
n
us
e
d
to
ge
n
e
r
a
t
e
c
o
n
t
e
x
t
w
o
r
d
e
m
b
e
dd
i
n
gs
,
whi
c
h
a
r
e
gi
ve
n
a
s
i
n
put
to
a
B
i
L
S
T
M
w
i
t
h
a
n
a
t
t
e
n
t
i
o
n
m
e
c
h
a
ni
s
m
f
o
r
t
h
e
s
e
n
t
i
m
e
n
t
c
l
a
s
s
if
i
c
a
t
i
o
n
i
n
t
o
p
os
i
t
i
v
e
o
r
n
e
ga
t
i
v
e
c
l
a
s
s
e
s
[
20]
.
T
h
e
a
tt
e
n
t
i
o
n
m
e
c
ha
ni
s
m
hi
g
hli
g
h
t
s
t
h
e
m
o
s
t
i
nf
o
r
m
a
t
i
v
e
c
o
n
t
e
x
t
f
e
a
t
ur
e
s
,
t
h
e
r
e
by
e
nh
a
nc
i
n
g
c
l
a
s
s
if
i
c
a
t
i
o
n
pe
r
f
o
r
m
a
n
c
e
.
F
i
na
ll
y
,
t
h
e
m
o
de
l
pa
r
a
m
e
t
e
r
s
a
r
e
o
p
t
i
mi
z
e
d
by
t
h
e
A
da
m
o
p
t
i
mi
z
e
r
to
a
c
hi
e
ve
hi
g
h
c
o
nve
r
ge
n
c
e
a
n
d
im
pr
o
v
e
d
pr
e
d
i
c
t
i
o
n
a
c
c
ur
a
c
y
[
21]
.
S
pe
c
i
f
i
c
a
ll
y
,
Al
go
r
i
t
hm
1
de
t
e
c
t
s
a
dv
e
r
s
a
r
i
a
l
a
t
t
a
c
ks
by
pr
o
c
e
s
s
i
ng
f
i
na
n
c
i
a
l
da
t
a
s
e
t
t
h
r
o
ugh
R
F
E
f
e
a
t
ur
e
s
e
l
e
c
t
i
o
n
,
gr
a
ph
c
o
ns
t
r
uc
t
i
o
n
GN
N,
a
n
d
c
r
o
ss
-
e
n
t
r
o
py
l
o
s
s
o
pt
i
mi
z
a
t
i
o
n
to
o
u
t
pu
t
a
n
a
t
t
a
c
k
l
a
b
e
l
,
w
h
e
r
e
a
s
Al
go
r
i
t
hm
2
pe
r
f
o
r
m
s
s
e
n
t
i
m
e
n
t
a
n
a
ly
s
i
s
o
n
f
i
na
n
c
i
a
l
t
e
x
t
by
e
x
t
r
a
c
t
i
n
g
c
o
n
t
e
x
t
ua
l
f
e
a
t
ur
e
s
w
i
t
h
F
i
nB
E
R
T
a
n
d
A
tt
e
n
t
i
o
n
-
b
a
s
e
d
B
i
L
S
T
M
o
p
t
i
mi
z
e
d
by
Ada
m
t
o
o
u
t
pu
t
a
p
o
s
i
t
i
v
e
o
r
n
e
ga
ti
ve
s
e
n
t
i
m
e
n
t
l
a
b
e
l
.
T
h
e
a
l
go
r
i
t
hm
s
f
o
r
a
dv
e
r
s
a
r
i
a
l
a
tt
a
c
k
de
t
e
c
t
i
o
n
m
o
du
l
e
a
n
d
s
e
n
t
i
m
e
n
t
a
n
a
ly
s
i
s
m
o
du
l
e
a
r
e
pr
e
s
e
n
t
e
d
a
s
f
o
l
l
o
w
.
Al
go
r
i
t
hm
1.
A
d
v
e
r
s
a
r
i
a
l
a
t
t
a
c
k
de
t
e
c
t
i
o
n
Input: Financial dataset D
Output: Adversarial attack label
1. Loading dataset D.
2. Prepro
cessing D.
3. Selecting informative features using RFE.
4. Constructing graph G (V, E).
5. Training GNN with selected features.
6. Updating model parameters by cross
-
entropy loss.
7. Predicting attack class.
8. Returning the attack label.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8776
I
n
t
J
I
n
f
&
C
o
m
m
u
n
T
e
c
hn
o
l
,
Vo
l
.
1
5
,
N
o.
4
,
D
e
c
e
m
be
r
20
2
6
:
1944
-
1953
1948
Al
go
r
i
t
hm
2.
S
e
n
t
i
m
e
n
t
a
n
a
l
y
s
i
s
Input: Financial text T
Output: Sentiment label
1. Loading text dataset T.
2. Preprocessing text.
3. Generating embeddings using FinBERT.
4. Extracting contextual features by BiLSTM with Attention.
5. Optimizing model using Adam.
6.
Predicting Positive or Negative sentiment.
7. Returning sentiment label.
2.
5.
M
od
e
l
in
t
e
r
p
r
e
t
ab
il
it
y
via
S
HAP
Due
to
t
h
e
i
r
r
e
s
t
r
i
c
t
e
d
i
n
t
e
r
pr
e
t
a
bi
li
t
y
,
D
L
m
e
t
h
o
ds
a
r
e
ge
n
e
r
a
l
ly
de
s
c
r
i
be
d
a
s
bl
a
c
k
b
o
x
e
s
.
T
o
e
n
h
a
n
c
e
m
o
de
l
t
r
a
n
s
pa
r
e
n
c
y
,
S
HA
P
i
s
e
m
p
l
o
ye
d
to
e
v
a
l
a
ut
e
t
h
e
i
m
pa
c
t
o
f
e
v
e
r
y
f
e
a
t
ur
e
f
o
r
t
h
e
p
r
e
d
i
c
t
i
o
n
[
22]
.
S
HA
P
a
s
s
i
g
ns
a
s
i
g
ni
f
i
c
a
n
c
e
v
a
l
ue
t
o
e
v
e
r
y
f
e
a
t
ur
e
by
e
va
l
u
a
t
i
n
g
i
t
s
i
nf
l
ue
n
c
e
o
n
t
h
e
m
o
d
e
l
r
e
s
u
l
t
,
e
n
a
bli
ng
t
h
e
i
de
n
t
i
f
i
c
a
t
i
o
n
o
f
im
po
r
t
a
n
t
f
a
c
to
r
s
t
h
a
t
d
r
i
ve
f
i
na
n
c
i
a
l
s
e
n
t
i
m
e
n
t
a
l
pr
e
d
i
c
t
i
o
ns
a
n
d
e
nh
a
nc
i
n
g
t
h
e
i
n
t
e
r
pr
e
t
a
bi
li
t
y
o
f
t
h
e
pr
o
p
o
s
e
d
f
r
a
m
e
wo
r
k.
3.
RE
S
UL
T
S
AN
D
DI
S
CU
S
S
I
ON
I
n
t
hi
s
s
e
c
t
i
o
n
,
t
h
e
e
m
p
i
r
i
c
a
l
v
a
li
da
t
i
o
n
o
f
t
h
e
AR
X
A
I
-
S
S
T
t
e
c
h
ni
que
i
s
i
ns
pe
c
t
e
d
ut
i
l
i
s
i
ng
t
h
r
e
e
da
t
a
s
e
t
s
,
n
a
m
e
ly
a
dv
e
r
s
a
r
i
a
l
M
L
d
a
t
a
s
e
t
[
23
]
,
s
tock
-
m
a
r
ke
t
s
e
n
t
i
m
e
n
t
da
t
a
s
e
t
[
24]
,
a
n
d
f
i
na
n
c
i
a
l
s
e
n
t
i
m
e
n
t
a
n
a
ly
s
i
s
da
t
a
s
e
t
[
25]
.
T
h
e
t
e
c
hni
que
i
s
e
x
e
c
ut
e
d
by
us
i
ng
P
y
t
h
o
n
3.
6.
5
wi
t
h
a
n
i
5
-
8600k
C
P
U,
4
G
B
GPU,
16
GB
R
A
M
,
250
GB
S
S
D,
a
n
d
1
T
B
HD
D.
T
a
bl
e
1
s
h
o
ws
t
h
e
pa
r
a
m
e
t
e
r
s
e
tt
i
n
gs
o
f
t
h
e
pr
o
p
o
s
e
d
m
e
t
h
o
d
.
T
a
bl
e
1.
P
a
r
a
m
e
t
e
r
s
e
tt
i
n
gs
o
f
t
h
e
A
R
X
A
I
-
S
S
T
m
e
t
h
o
d
M
o
dul
e
H
y
p
e
r
pa
r
a
m
e
t
e
r
O
pt
im
a
l
v
a
lu
e
F
in
B
E
R
T
e
mb
e
ddi
ng
M
a
x
im
um
s
e
qu
e
nc
e
l
e
ngt
h
256
E
mbe
ddi
ng
d
im
e
ns
i
o
n
768
D
r
o
p
o
ut
r
a
te
0.2
G
N
N
N
umbe
r
of
g
r
a
ph l
a
y
e
r
s
3
H
id
d
e
n
di
m
e
ns
io
n
128
A
tt
e
nt
i
o
n
h
e
a
ds
4
B
i
L
S
T
M
w
i
th
a
tt
e
nt
i
o
n
B
i
L
S
T
M
l
a
y
e
r
s
2
H
id
d
e
n
u
ni
ts
128
A
tt
e
nt
i
o
n
h
e
a
ds
4
D
r
o
p
o
ut
r
a
te
0.2
T
r
a
in
in
g
c
o
n
f
ig
ur
a
ti
o
n
O
pt
im
iz
e
r
A
da
m
L
e
a
r
ni
ng
r
a
te
0.001
B
a
tc
h
s
i
z
e
32
E
poc
hs
50
W
e
ig
ht
d
e
c
a
y
1×
10⁻
⁴
E
a
r
l
y
s
to
ppi
ng
p
a
ti
e
n
c
e
10
3.
1.
R
e
s
u
l
t
an
al
ys
is
of
ad
ve
r
s
a
r
ial
m
l
d
at
as
e
t
T
h
e
e
x
pe
r
i
m
e
n
t
a
t
i
o
n
a
n
a
ly
s
i
s
o
f
t
h
e
A
R
X
A
I
-
S
S
T
m
e
t
h
o
d
i
s
i
ns
pe
c
t
e
d
un
de
r
t
h
e
a
dv
e
r
s
a
r
i
a
l
M
L
d
a
t
a
s
e
t
[
26
]
-
[
29]
.
DN
S
t
un
n
e
ll
i
ng
de
a
l
s
w
i
t
h
ne
t
wo
r
k
da
t
a
ga
i
n
e
d
dur
i
n
g
a
DN
S
t
un
ne
l
li
ng
a
t
t
a
c
k.
T
h
e
tot
a
l
c
o
un
t
o
f
f
e
a
t
ur
e
s
i
s
12,
a
n
d
a
l
l
12
f
e
a
t
ur
e
s
we
r
e
c
h
o
s
e
n
.
T
h
e
da
t
a
s
e
t
e
n
c
o
m
pa
s
s
e
s
5999
i
n
s
t
a
n
c
e
s
un
de
r
t
w
o
c
l
a
s
s
e
.
T
a
bl
e
2
de
p
i
c
t
s
t
h
e
c
o
m
pa
r
i
s
o
n
a
s
s
e
s
e
m
e
n
t
o
f
t
h
e
A
R
X
A
I
-
S
S
T
f
r
a
m
e
wo
r
k
wi
t
h
e
xi
s
t
i
n
g
s
y
s
t
e
m
s
.
T
h
e
pe
r
f
o
r
m
a
nc
e
o
u
t
c
o
m
e
s
h
o
we
d
t
h
a
t
t
h
e
A
R
X
A
I
-
S
S
T
m
o
de
l
o
b
t
a
i
ns
c
o
n
s
i
de
r
a
bl
e
o
u
t
c
o
m
e
s
w
i
t
h
o
f
97.
22%
,
o
f
95.
95%
,
o
f
97.
22%
,
a
n
d
1
o
f
96.
56%
,
c
o
r
r
e
s
p
o
n
d
i
n
g
ly
.
A
d
d
i
t
i
o
n
a
ll
y
,
t
h
e
i
s
o
l
a
t
i
o
n
f
o
r
e
s
t
(
I
F
)
a
n
d
J
48
m
o
de
l
s
ge
t
s
om
e
w
h
a
t
l
e
s
s
e
r
pe
r
f
o
r
m
a
n
c
e
w
i
t
h
o
f
96.
33%
a
n
d
96.
11%
,
o
f
91.
13%
a
n
d
90.
71%
,
o
f
92.
15%
a
n
d
88.
97%
,
a
n
d
1
o
f
89.
89%
a
n
d
95.
32%
,
r
e
s
pe
c
t
i
ve
ly
.
F
i
gur
e
2
e
x
p
l
a
i
ns
b
o
t
h
gl
o
ba
l
a
n
d
l
o
c
a
l
i
n
t
e
r
pr
e
t
a
t
i
o
ns
o
f
t
h
e
m
o
de
l
e
m
p
l
o
y
i
ng
S
H
A
P
v
a
l
ue
s
.
T
h
e
g
l
o
b
a
l
f
e
a
t
ur
e
s
i
g
ni
f
i
c
a
n
c
e
p
l
o
t
d
i
s
p
l
a
y
s
t
h
a
t
mA
i
s
t
h
e
m
o
s
t
p
o
we
r
f
u
l
f
e
a
t
ur
e
,
c
on
t
r
i
b
ut
i
n
g
s
i
g
nif
i
c
a
n
t
l
y
m
o
r
e
to
t
h
e
m
o
de
l
’
s
pr
e
d
i
c
t
i
o
ns
t
h
a
n
t
h
e
ot
h
e
r
s
.
F
e
a
t
ur
e
s
l
i
ke
m
Q,
vA
,
a
n
d
s
A
a
r
e
pr
e
s
e
n
t
e
d
a
s
o
f
r
e
a
s
o
n
a
bl
e
i
m
po
r
t
a
n
c
e
,
wh
e
r
e
a
s
t
h
e
r
e
s
i
du
a
l
f
e
a
t
ur
e
s
c
a
n
h
a
ve
c
o
m
pa
r
a
t
i
v
e
ly
s
m
a
ll
e
r
i
m
pa
c
t
s
.
T
h
e
l
o
c
a
l
e
x
p
l
a
n
a
t
i
o
n
s
u
mm
a
r
i
s
e
s
h
o
w
i
n
d
i
v
i
d
ua
l
f
e
a
t
ur
e
v
a
l
ue
s
a
f
f
e
c
t
t
h
e
o
u
t
pu
t
v
a
r
i
a
bl
e
g
a
c
r
o
s
s
v
a
r
i
o
us
i
ns
t
a
nc
e
s
.
P
o
s
i
t
i
ve
S
HA
P
v
a
l
ue
s
s
pe
c
if
y
a
n
ups
ur
ge
i
n
t
he
pr
e
d
i
c
t
i
o
n
,
whil
e
n
e
ga
t
i
v
e
v
a
l
ue
s
s
h
o
w
a
r
e
duc
t
i
o
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t
J
I
n
f
&
C
o
m
m
u
n
T
e
c
hn
o
l
I
S
S
N:
2252
-
8776
A
dv
e
r
s
ar
ial
-
r
e
s
il
ient
iot
-
e
nabled
f
inancial
int
e
ll
ige
nc
e
s
y
s
tem
us
ing
e
x
plai
nable
…
(
Sur
e
s
h
T
alamala
)
1949
T
h
e
c
o
l
o
ur
g
r
a
d
i
e
n
t
(
bl
u
e
to
r
e
d)
de
n
o
t
e
s
l
o
we
r
to
hi
g
h
e
r
f
e
a
t
ur
e
v
a
l
ue
s
,
un
de
r
s
c
o
r
i
n
g
h
o
w
d
i
s
s
im
il
a
r
i
t
i
e
s
i
n
f
e
a
t
ur
e
m
a
g
ni
t
ude
i
m
pa
c
t
t
h
e
m
o
de
l
’
s
b
e
h
a
vi
o
ur
.
T
h
e
S
HA
P
a
na
l
y
s
i
s
f
ur
t
h
e
r
m
o
r
e
v
a
li
da
t
e
s
t
h
a
t
t
h
e
m
o
de
l
pr
e
d
i
c
t
i
o
ns
a
r
e
b
a
s
i
c
a
ll
y
dr
i
ve
n
by
m
e
a
ni
ng
f
u
l
e
c
o
n
o
m
i
c
i
nd
i
c
a
t
or
s
i
n
s
t
e
a
d
o
f
a
dv
e
r
s
a
r
i
a
l
t
r
i
gge
r
pa
tt
e
r
n
s
.
T
h
e
c
o
n
tr
o
l
o
f
e
c
o
n
o
m
i
c
a
ll
y
r
e
lev
a
n
t
f
e
a
t
ur
e
s
a
n
d
t
h
e
l
im
i
t
e
d
c
o
n
t
r
i
b
ut
i
o
ns
o
f
un
r
e
l
a
t
e
d
f
e
a
t
ur
e
s
i
n
d
i
c
a
t
e
t
h
a
t
t
h
e
m
o
de
l
c
a
pt
ur
e
s
m
a
r
ke
t
r
e
l
a
t
e
d
s
i
g
na
l
s
a
n
d
l
o
we
r
s
de
p
e
n
de
n
c
e
o
n
f
e
a
s
i
b
ly
m
a
ni
pu
l
a
t
e
d
ke
y
wo
r
ds
.
T
a
bl
e
2.
C
o
m
pa
r
i
s
o
n
o
f
o
u
t
c
o
m
e
s
o
f
t
h
e
AR
X
A
I
-
S
S
T
m
e
t
h
o
d
wi
t
h
e
xi
s
t
i
n
g
m
o
de
l
s
o
n
t
h
e
a
dv
e
r
s
a
r
i
a
l
M
L
da
t
a
s
e
t
M
o
de
ls
1
IF
96.33
91.13
92.15
89.89
S
V
M
89.33
84.25
94.59
89.90
J
48
m
o
de
l
96.11
90.71
88.97
95.32
N
a
ïv
e
B
a
y
e
s
92.40
86.20
89.56
92.66
K
-
m
e
a
ns
c
lu
s
t
e
r
in
g
95.67
84.73
87.06
84.16
N
e
ur
a
l
N
e
t
81.10
88.55
94.09
85.75
R
a
ndo
m
f
o
r
e
s
t
82.00
84.20
95.01
91.34
A
R
X
A
I
-
S
S
T
97.22
95.95
97.22
96.56
F
i
gur
e
2
.
S
HA
P
-
b
a
s
e
d
g
l
o
b
a
l
a
n
d
l
o
c
a
l
i
n
t
e
r
pr
e
t
a
b
il
i
t
y
a
n
a
ly
s
i
s
o
f
a
d
v
e
r
s
a
r
i
a
l
M
L
da
t
a
s
e
t
3.
2.
R
e
s
u
l
t
an
al
ys
is
of
s
t
oc
k
-
m
a
r
k
e
t
s
e
n
t
im
e
n
t
d
at
as
e
t
T
h
e
e
m
p
i
r
i
c
a
l
a
s
s
e
s
s
m
e
n
t
o
f
t
h
e
A
R
X
A
I
-
S
S
T
m
e
t
h
o
d
i
s
a
c
hi
e
v
e
d
un
de
r
t
h
e
s
to
c
k
-
m
a
r
ke
t
s
e
n
t
i
m
e
n
t
da
t
a
s
e
t.
Ga
t
h
e
r
e
d
s
to
c
k
n
e
ws
f
r
o
m
m
u
l
t
i
p
l
e
T
w
i
tt
e
r
h
a
n
d
l
e
s
r
e
ga
r
de
d
e
c
o
n
o
m
i
c
n
e
w
s
,
s
e
pa
r
a
t
i
n
g
i
n
t
o
t
w
o
pa
r
t
s
:
n
e
ga
t
i
ve
(
-
1)
a
n
d
p
o
s
i
t
i
v
e
(
1)
.
T
h
e
da
t
a
s
e
t
e
n
c
o
m
pa
s
s
e
s
a
tot
a
l
o
f
5791
c
o
un
t
s
un
de
r
n
e
ga
t
i
v
e
h
o
l
d
s
,
2,
106,
a
n
d
p
o
s
i
t
i
ve
h
o
l
ds
,
3,
685.
T
a
bl
e
3
i
ll
us
t
r
a
t
e
t
h
e
c
o
m
pa
r
i
s
o
n
pe
r
f
o
r
m
a
n
c
e
o
f
th
e
pr
o
p
o
s
e
d
f
r
a
m
e
wo
r
k
wi
t
h
e
xi
s
t
i
n
g
m
e
t
h
o
ds
o
n
t
h
e
s
to
c
k
-
m
a
r
ke
t
da
t
a
s
e
t
.
T
h
e
S
VM
,
L
R
,
a
n
d
R
F
m
o
de
l
s
a
c
hi
e
v
e
d
l
o
we
r
pe
r
f
o
r
m
a
n
c
e
,
w
hil
e
B
E
R
T
,
F
i
nB
E
R
T
,
GPT
-
4
o,
a
n
d
GPT
-
4o
-
m
i
n
i
s
h
o
we
d
i
mpr
o
v
e
d
r
e
s
u
l
t
s
.
Unl
i
k
e
t
h
e
pr
o
p
o
s
e
d
A
R
X
A
I
-
S
S
T
m
e
t
h
o
d
a
c
c
o
m
p
li
s
he
d
hi
g
h
pe
r
f
o
r
m
a
nc
e
w
i
t
h
97.
74%
a
c
c
ur
a
c
y
,
97.
90%
pr
e
c
i
s
i
o
n
,
97.
74%
r
e
c
a
l
l
,
a
n
d
97.
82%
F1
-
s
c
o
r
e
.
T
h
e
s
e
r
e
s
u
l
t
s
de
m
o
ns
t
r
a
t
e
t
h
e
e
f
f
i
c
i
e
n
c
y
o
f
pr
o
p
o
s
e
d
m
o
de
l
i
n
e
nha
n
c
i
ng
s
e
n
t
i
m
e
n
t
dr
i
ve
n
s
to
c
k
t
r
a
di
n
g
de
c
i
s
i
o
n
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8776
I
n
t
J
I
n
f
&
C
o
m
m
u
n
T
e
c
hn
o
l
,
Vo
l
.
1
5
,
N
o.
4
,
D
e
c
e
m
be
r
20
2
6
:
1944
-
1953
1950
T
a
bl
e
3.
C
o
m
pa
r
a
t
i
ve
o
u
t
c
o
m
e
o
f
A
R
X
A
I
-
S
S
T
m
e
t
h
o
d
wi
t
h
e
xi
s
t
i
n
g
m
o
de
l
s
o
n
s
to
c
k
-
m
a
r
ke
t
da
t
a
s
e
t
M
o
de
l
1
G
P
T
-
4o
87.79
87.86
87.79
87.69
G
P
T
-
4o
-
mi
n
i
87.79
87.74
87.79
87.73
B
E
R
T
81.20
81.50
81.20
81.16
F
in
B
E
R
T
80.04
80.04
80.04
80.04
S
V
M
64.53
65.68
64.53
64.38
R
F
m
o
d
e
l
65.31
65.55
65.31
65.31
L
o
g
is
ti
c
-
r
e
g
r
e
s
s
io
n
64.92
65.89
64.92
64.74
A
R
X
A
I
-
S
S
T
97.74
97.90
97.74
97.82
T
h
e
a
bl
a
t
i
o
n
r
e
s
u
l
t
s
us
i
n
g
a
c
c
ur
a
c
y
,
pr
e
c
i
s
i
o
n
,
r
e
c
a
l
l
,
a
n
d
F
1
-
s
c
o
r
e
de
m
o
ns
t
r
a
t
e
t
h
e
c
o
n
t
r
i
b
ut
i
o
n
o
f
e
a
c
h
c
o
m
po
n
e
n
t
i
n
AR
X
A
I
-
S
S
T
.
T
a
bl
e
4
de
m
o
n
s
t
r
a
t
e
s
t
h
e
a
bl
a
t
i
o
n
a
n
a
ly
s
i
s
o
f
t
h
e
pr
o
po
s
e
d
m
e
t
h
o
d.
F
o
r
t
h
e
a
dv
e
r
s
a
r
i
a
l
M
L
da
t
a
s
e
t
,
A
R
X
A
I
-
S
S
T
(
S
c
a
l
i
ng
+
R
F
E
+
GN
N)
a
c
hi
e
v
e
s
97.
22%
a
c
c
ur
a
c
y
a
n
d
96.
56%
F1
-
s
c
o
r
e
.
F
or
t
h
e
s
to
c
k
m
a
r
ke
t
da
t
a
s
e
t
,
t
h
e
c
o
m
p
l
e
t
e
m
o
de
l
(
F
i
nB
E
R
T
+
B
i
L
S
T
M
-
A
t
t
+
A
da
m
)
o
b
t
a
i
n
s
97.
74%
a
c
c
ur
a
c
y
a
n
d
97.
82%
F
1
-
s
c
o
r
e
,
whi
l
e
t
h
e
f
i
na
n
c
i
a
l
s
e
n
t
i
m
e
n
t
da
t
a
s
e
t
a
c
hi
e
v
e
s
97.
75%
a
c
c
ur
a
c
y
a
n
d
95.
26%
F
1
-
s
c
o
r
e
.
T
h
e
s
e
r
e
s
u
l
t
s
c
o
nf
i
r
m
t
h
a
t
t
h
e
c
o
m
bi
na
t
i
o
n
o
f
f
e
a
t
ur
e
s
e
l
e
c
t
i
o
n
,
GN
N,
F
i
n
B
E
R
T
,
a
tt
e
n
t
i
o
n
m
e
c
h
a
ni
s
m
,
a
n
d
A
da
m
o
pt
i
mi
z
e
r
i
m
pr
o
v
e
s
t
h
e
o
v
e
r
a
l
l
r
o
b
us
t
n
e
s
s
a
n
d
pr
e
d
i
c
t
i
o
n
pe
r
f
o
r
m
a
n
c
e
.
T
a
bl
e
4.
A
bl
a
t
i
o
n
s
t
ud
y
o
f
t
h
e
AR
X
A
I
-
S
S
T
m
e
t
h
o
d
M
o
de
ls
1
A
dve
r
s
a
r
ia
l
M
L
da
ta
s
e
t
W
it
ho
ut
t
h
e
R
F
E
f
e
a
tu
r
e
s
e
l
e
c
t
i
o
n
95.91
94.65
96.00
95.27
G
N
N
o
nl
y
96.71
95.15
96.62
95.91
A
R
X
A
I
-
S
S
T
(
S
c
a
li
ng +
R
F
E
+
G
N
N
)
97.22
95.95
97.22
96.56
S
to
c
k
-
m
a
r
ke
t
da
ta
s
e
t
B
i
L
S
T
M
o
n
l
y
(
w
it
h
o
ut
a
tt
e
nt
i
o
n)
95.93
95.96
95.63
95.88
B
i
L
S
T
M
-
A
tt
o
nl
y
96.68
96.49
96.19
96.53
F
in
B
E
R
T
+
B
i
L
S
T
M
-
A
tt
(
w
it
h
o
u
t
A
da
m o
pt
im
i
z
e
r
)
97.20
97.20
96.99
97.03
A
R
X
A
I
-
S
S
T
(
F
in
B
E
R
T
+
B
i
L
S
T
M
-
A
tt
+
A
da
m)
97.74
97.9
97.74
97.82
F
in
a
nc
ia
l
s
e
nt
i
me
nt
a
na
l
y
s
is
da
ta
s
e
t
B
i
L
S
T
M
o
n
l
y
(
w
it
h
o
ut
a
tt
e
nt
i
o
n)
95.54
94.52
92.51
93.51
B
i
L
S
T
M
-
A
tt
o
nl
y
96.29
95.08
93.11
94.09
F
in
B
E
R
T
+
B
i
L
S
T
M
-
A
tt
(
w
it
h
o
u
t
A
da
m o
pt
im
i
z
e
r
)
96.96
95.75
93.69
94.72
A
R
X
A
I
-
S
S
T
(
F
in
B
E
R
T
+
B
i
L
S
T
M
-
A
tt
+
A
da
m)
97.75
96.32
94.40
95.26
4.
CONC
L
USI
ON
I
n
t
hi
s
pa
pe
r
,
t
h
e
A
R
X
A
I
-
S
S
T
m
e
t
h
o
d
i
s
pr
e
s
e
nt
e
d
a
s
a
n
e
x
p
l
a
i
na
bl
e
a
n
d
s
e
c
ur
e
f
r
a
m
e
wo
r
k
f
o
r
s
e
n
t
i
m
e
n
t
dr
i
ve
n
s
to
c
k
tr
a
di
n
g
de
c
i
s
i
o
ns
i
n
I
o
T
e
n
a
bl
e
d
f
i
na
n
c
i
a
l
e
nvi
r
o
nm
e
n
t
s
.
T
h
e
pr
o
p
o
s
e
d
m
e
t
h
o
d
e
n
a
bl
e
s
us
e
r
s
to
un
de
r
s
t
a
n
d
t
h
e
j
u
s
t
i
f
i
c
a
t
i
o
n
be
hi
nd
m
a
r
ke
t
tr
e
n
d
pr
e
d
i
c
t
i
o
n
s
a
n
d
to
v
e
r
i
f
y
w
h
e
t
h
e
r
de
c
i
s
i
o
ns
a
r
e
b
a
s
e
d
o
n
t
h
e
ge
n
u
i
n
e
m
a
r
ke
t
b
e
h
a
vi
o
ur
.
T
h
e
pr
o
p
o
s
e
d
m
e
t
h
o
d
e
n
s
ur
e
s
t
r
a
n
s
pa
r
e
n
t
,
a
c
c
ur
a
t
e
a
n
d
r
o
b
us
t
t
r
a
di
n
g
de
c
i
s
i
o
ns
.
T
h
e
e
x
pe
r
i
m
e
n
t
a
l
a
s
s
e
s
s
m
e
n
t
on
b
e
n
c
hm
a
r
k
da
t
a
s
e
t
s
s
h
o
ws
t
h
a
t
t
h
e
pr
o
p
o
s
e
d
f
r
a
m
e
wo
r
k
a
tt
a
i
ns
gr
e
a
t
e
r
pe
r
f
o
r
m
a
n
c
e
t
h
a
n
t
h
e
e
xi
s
t
i
n
g
s
ys
t
e
m
s
.
B
e
y
o
n
d
pe
r
f
o
r
m
a
n
c
e
i
m
pr
o
v
e
m
e
n
t
s
,
t
h
e
pr
o
p
o
s
e
d
a
r
c
hi
t
e
c
t
ur
e
h
a
s
s
ubs
t
a
n
t
i
a
l
po
t
e
n
t
i
a
l
f
o
r
r
e
a
l
wo
r
l
d
de
p
l
o
y
m
e
n
t
i
n
f
i
n
t
e
c
h
p
l
a
t
f
o
r
m
s
.
I
t
s
e
x
p
l
a
i
na
bil
i
t
y
c
a
pa
bi
li
t
y
c
a
n
a
l
s
o
s
uppo
r
t
t
h
e
r
e
gul
a
t
o
r
y
c
o
m
p
li
a
n
c
e
by
pr
o
vi
d
i
ng
t
h
e
t
r
a
n
s
pa
r
e
n
t
i
ns
i
g
h
t
s
i
n
t
o
A
I
dr
i
v
e
n
t
r
a
di
n
g
de
c
i
s
i
o
n
s
.
Al
t
h
o
ugh
t
h
e
pr
o
p
o
s
e
d
f
r
a
m
e
w
o
r
k
de
m
o
ns
t
r
a
t
e
s
hi
g
h
pe
r
f
o
r
m
a
n
c
e
,
t
h
e
e
x
pe
r
im
e
n
t
s
we
r
e
c
o
n
duc
t
e
d
by
b
e
n
c
hm
a
r
k
da
t
a
s
e
t
s
,
whi
c
h
m
a
y
n
o
t
c
o
m
p
l
e
t
e
l
y
c
a
pt
ur
e
r
e
a
l
wo
r
l
d
f
i
na
n
c
i
a
l
m
a
r
ke
t
d
y
na
mi
c
s
.
I
n
a
dd
i
t
i
o
n
,
l
i
ve
t
r
a
d
i
n
g
a
s
s
e
s
s
m
e
n
t
s
a
n
d
f
i
na
n
c
ia
l
b
a
c
k
t
e
s
t
i
n
g
we
r
e
o
v
e
r
t
h
e
s
c
o
pe
o
f
t
hi
s
s
t
udy
.
T
h
e
s
e
li
mi
t
a
t
i
o
ns
m
a
y
im
pa
c
t
t
h
e
e
v
a
l
ua
t
i
o
n
o
f
pr
a
c
t
i
c
a
l
de
p
l
o
y
m
e
n
t
,
a
n
d
t
h
e
f
ut
ur
e
w
o
r
k
wi
ll
hi
g
hli
g
h
t
o
n
v
a
li
da
t
i
o
n
o
f
t
h
e
a
r
c
hi
t
e
c
t
ur
e
i
n
r
e
a
l
t
i
m
e
I
o
T
e
n
a
bled
f
i
na
nc
i
a
l
e
nvi
r
o
nm
e
n
t
s
.
F
UN
DI
NG
I
N
F
ORM
AT
I
ON
T
h
e
a
ut
h
o
r
s
r
e
c
e
i
v
e
d
n
o
f
u
n
d
i
ng
f
o
r
t
hi
s
s
t
ud
y
.
AU
T
HO
R
CONT
RI
B
UT
I
ONS
S
T
AT
E
M
E
NT
T
hi
s
j
o
ur
n
a
l
u
s
e
s
t
h
e
C
o
n
t
r
i
b
ut
or
R
o
l
e
s
T
a
x
o
n
o
m
y
(
C
R
e
d
i
T
)
to
r
e
c
o
gni
z
e
i
n
d
i
v
i
dua
l
a
ut
h
o
r
c
o
n
t
r
i
b
ut
i
o
ns
,
r
e
duc
e
a
ut
h
or
s
hi
p
d
i
s
put
e
s
,
a
n
d
f
a
c
il
i
t
a
t
e
c
o
l
l
a
b
o
r
a
t
i
o
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t
J
I
n
f
&
C
o
m
m
u
n
T
e
c
hn
o
l
I
S
S
N:
2252
-
8776
A
dv
e
r
s
ar
ial
-
r
e
s
il
ient
iot
-
e
nabled
f
inancial
int
e
ll
ige
nc
e
s
y
s
tem
us
ing
e
x
plai
nable
…
(
Sur
e
s
h
T
alamala
)
1951
Nam
e
of
Aut
h
or
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
S
ur
e
s
h
T
a
l
a
m
a
l
a
S
ukh
a
da
Al
o
ni
J
.
Vi
j
a
y
a
l
a
ks
hmi
S
ubh
a
m
P
a
n
k
a
j
S
a
m
a
n
t
a
r
a
y
K
.
S
wa
n
t
h
a
n
a
N
i
s
a
r
g
Ga
n
d
h
e
w
a
r
C
:
C
o
n
c
e
pt
ua
li
z
a
ti
o
n
M
:
M
e
th
o
d
o
l
o
g
y
So
:
So
f
twa
r
e
Va
:
Va
li
da
ti
o
n
Fo
:
Fo
r
ma
l
a
na
l
y
s
is
I
:
I
nve
s
ti
ga
ti
o
n
R
:
R
e
s
o
u
r
c
e
s
D
:
D
a
ta
C
ur
a
ti
o
n
O
:
W
r
it
in
g
-
O
r
ig
in
a
l
D
r
a
f
t
E
:
W
r
it
in
g
-
R
e
v
i
e
w
&
E
di
ti
ng
Vi
:
Vi
s
ua
li
z
a
ti
o
n
Su
:
Su
pe
r
v
is
io
n
P
:
P
r
o
j
e
c
t
a
dmi
ni
s
tr
a
ti
o
n
Fu
:
Fu
ndi
ng a
c
qui
s
it
i
o
n
CONF
L
I
CT
OF
I
NT
E
RE
S
T
S
T
AT
E
M
E
NT
A
ut
h
o
r
s
s
t
a
t
e
n
o
c
o
nf
li
c
t
o
f
i
n
t
e
r
e
s
t
.
DA
T
A
AV
AI
L
AB
I
L
I
T
Y
S
T
A
T
E
M
E
NT
T
h
e
da
t
a
t
h
a
t
s
uppo
r
t
t
h
e
f
i
nd
i
ngs
o
f
t
hi
s
s
t
ud
y
a
r
e
o
pe
nl
y
a
v
a
il
a
bl
e
i
n
t
h
e
K
a
gg
l
e
r
e
po
s
i
to
r
y
a
t
h
t
t
ps
:/
/www
.
ka
ggl
e
.
c
o
m
/dat
a
s
e
t
s
/cn
r
i
e
i
i
t
/ad
ve
r
s
a
r
i
a
l
-
m
a
c
hi
ne
-
le
a
r
ni
ng
-
da
t
a
s
e
t
,
h
t
t
ps
:/
/www
.
ka
ggl
e
.
c
o
m
/dat
a
s
e
t
s
/
y
a
s
h
612/
s
t
o
c
km
a
r
ke
t
-
s
e
n
t
i
m
e
n
t
-
da
t
a
s
e
t
,
a
n
d
h
t
t
ps
:/
/www
.
ka
ggl
e
.
c
o
m
/dat
a
s
e
t
s
/s
bh
a
t
t
i
/
f
i
na
nc
i
a
l
-
s
e
n
t
i
m
e
n
t
-
a
n
a
ly
s
i
s
,
r
e
f
e
r
e
n
c
e
n
u
m
be
r
s
[
23
]
-
[
25]
.
RE
F
E
R
E
NC
E
S
[
1]
E
.
C
ic
e
k
y
ur
t
a
nd
G
.
B
a
ka
l,
“
E
nha
n
c
in
g
s
e
nt
i
me
nt
a
na
l
y
s
is
in
s
to
c
k
ma
r
ke
t
twe
e
ts
th
r
o
ugh
B
E
R
T
-
ba
s
e
d
kn
o
w
le
dg
e
t
r
a
ns
f
e
r
,”
C
om
put
at
io
nal
E
c
onomic
s
, vo
l.
67, n
o
. 2, pp. 855
–
877, F
e
b. 20
26, do
i:
10.1007/s
10614
-
025
-
10901
-
8.
[
2]
E
.
M
ut
hukuma
r
,
H
.
P
.
J
o
s
y
ul
a
,
S
.
K
.
G
a
ta
la
,
M
.
K
.
V
a
nda
na
pu,
V
.
M
is
tr
y
,
a
nd
N
.
S
in
gh,
“
A
I
-
dr
i
ve
n
p
r
e
d
ic
t
i
v
e
a
na
l
y
ti
c
s
f
or
f
in
a
n
c
ia
l
ma
r
k
e
t
f
o
r
e
c
a
s
ti
ng,”
in
P
r
oc
e
e
di
ngs
o
f
th
e
2025
I
nt
e
r
nat
io
nal
C
on
f
e
r
e
nc
e
on
T
e
c
hnol
ogy
E
nabl
e
d
E
c
onomic
C
han
ge
s
,
I
nT
e
c
h 2025
, I
E
E
E
, F
e
b. 2025, pp. 1389
–
1394
,
d
o
i:
10.1109/I
n
T
e
c
h64186.2025.11198418.
[
3]
H
.
R
.
B
.
S
e
s
ha
ka
ga
r
i,
A
.
U
ma
s
ha
nka
r
,
T
.
H
a
r
ik
a
la
,
L
.
J
a
y
a
s
r
e
e
,
a
nd
J
.
S
e
ve
r
a
n
c
e
,
“
D
y
na
mi
c
f
in
a
nc
ia
l
s
e
nt
im
e
nt
a
na
l
y
s
is
a
nd
ma
r
ke
t
f
o
r
e
c
a
s
ti
ng t
hr
o
ugh
l
a
r
g
e
l
a
ngua
ge
m
o
d
e
ls
,”
I
nt
e
r
nat
io
nal
J
our
nal
of
H
um
an C
om
put
at
io
ns
& i
nt
e
ll
ig
e
nc
e
, 2025,
[
o
nl
in
e
]
.
a
v
a
il
a
bl
e
:
h
tt
p:
//
c
r
e
a
ti
ve
c
o
m
m
o
ns
.o
r
g
/l
i
c
e
ns
e
s
/b
y
/4
.0
/3
97
[
4]
F
.
M
.
P
.
F
oz
a
p,
“
H
y
br
id
ma
c
hi
n
e
le
a
r
n
in
g
mo
d
e
ls
f
o
r
l
o
ng
-
te
r
m
s
t
oc
k
ma
r
ke
t
f
or
e
c
a
s
ti
ng:
in
t
e
gr
a
ti
ng
t
e
c
hni
c
a
l
in
di
c
a
t
or
s
,”
J
our
nal
of
R
is
k
and F
in
anc
ia
l
M
anage
m
e
nt
,
vo
l.
18, n
o
. 4, p. 2
01, Apr
. 2025, do
i:
10.3390
/j
r
f
m18040201.
[
5]
T
.
K
uma
r
,
“
D
e
e
p
l
e
a
r
ni
ng
-
ba
s
e
d
s
to
c
k
p
r
i
c
e
f
o
r
e
c
a
s
ti
ng
f
or
bu
s
in
e
s
s
ma
na
ge
me
nt
:
a
h
y
br
id
a
ppr
o
a
c
h
us
in
g
L
S
T
M
a
nd
s
e
nt
i
me
nt
a
na
l
y
s
is
,”
E
K
SP
L
O
R
I
U
M
-
B
U
L
E
T
I
N
P
U
S
A
T
T
E
K
N
O
L
O
G
I
B
A
H
A
N
G
A
L
I
A
N
N
U
K
L
I
R
,
vo
l.
46, n
o
. 1, pp. 1598
–
1613, 2025.
[
6]
K
.
A
la
m,
M
.
H
.
B
hui
y
a
n,
I
.
U
l
H
a
qu
e
,
M
.
F
.
M
o
ni
r
,
a
nd
T
.
A
hme
d,
“
E
nha
nc
in
g
s
t
oc
k
ma
r
ke
t
p
r
e
di
c
ti
o
n
:
a
r
o
bus
t
L
S
T
M
-
DNN
mo
d
e
l
a
na
l
y
s
is
o
n 26
r
e
a
l
-
li
f
e
da
ta
s
e
ts
,”
I
E
E
E
A
c
c
e
s
s
, v
o
l.
12,
pp. 122757
–
122768, 2024, do
i:
10.1109/AC
C
E
S
S
.2024.34345
24.
[
7]
D
.
G
.
T
a
ka
le
,
“
E
nha
nc
in
g
f
in
a
nc
ia
l
s
e
nt
im
e
nt
a
na
l
y
s
is
:
a
de
e
p
di
ve
in
t
o
na
tu
r
a
l
la
ngua
ge
pr
oc
e
s
s
in
g
f
o
r
ma
r
ke
t
pr
e
di
c
ti
o
n
in
dus
tr
ie
s
,”
J
our
nal
of
C
o
m
put
e
r
N
e
tw
or
k
s
and V
ir
tu
al
iz
at
io
n
,
vo
l.
2, n
o
. 2, pp. 1
–
5, 2024, d
o
i:
10.48001/
j
o
c
n
v
.2024.221
-
5.
[
8]
S
.
B
a
not
h
a
nd
S
.
Y
a
da
la
,
“
A
hy
br
id
de
e
p
le
a
r
n
in
g
f
r
a
m
e
w
or
k
f
o
r
s
t
o
c
k
p
r
ic
e
f
or
e
c
a
s
ti
ng
w
it
h
s
e
nt
im
e
nt
an
a
l
y
s
is
,”
in
P
r
oc
e
e
di
ngs
of
th
e
2024
13t
h
I
nt
e
r
nat
io
nal
C
on
f
e
r
e
nc
e
on
Sy
s
te
m
M
ode
li
n
g
and
A
dv
anc
e
m
e
nt
in
R
e
s
e
a
r
c
h
T
r
e
nds
,
SM
A
R
T
2024
,
I
E
E
E
,
D
e
c
.
2024, pp. 467
–
471
,
do
i:
10.1109/
S
M
A
R
T
63812.2024.1088221
5.
[9
]
V
.
K
ha
nde
lwa
l,
H
.
V
a
r
s
hne
y
,
a
nd
G
.
M
unj
a
l,
“
S
e
nt
im
e
nt
a
na
l
y
s
is
ba
s
e
d
s
to
c
k
pr
ic
e
pr
e
di
c
ti
o
n
us
in
g
ma
c
hi
n
e
l
e
a
r
ni
ng,
”
in
P
r
oc
e
e
di
ngs
-
2nd
I
nt
e
r
nat
io
nal
C
onf
e
r
e
nc
e
on
A
dv
anc
e
m
e
n
t
in
C
om
put
at
io
n
and
C
om
put
e
r
T
e
c
hnol
ogi
e
s
,
I
nC
A
C
C
T
20
24
,
I
E
E
E
, M
a
y
2024, pp.
182
–
187
,
d
o
i:
10.1109/
I
nC
A
C
C
T
61598.2
024.10551038.
[
10]
S
.
P
r
a
ba
ka
r
,
A
.
K
uma
r
,
R
.
J
a
y
a
ka
r
th
ik
,
D
.
V
e
nka
t
e
s
h,
R
.
S
.
P
r
a
th
e
e
ba
,
a
nd
B
.
K
ha
n,
“
E
mpi
r
ic
a
l
e
v
a
lu
a
ti
o
n
of
s
t
o
c
k
ma
r
ke
t
pr
e
di
c
ti
o
n
s
y
s
t
e
m
us
in
g
in
t
e
ll
ig
e
nt
l
e
a
r
ni
ng
s
c
h
e
m
e
w
it
h
d
a
ta
pr
oc
e
s
s
in
g
l
o
gi
c
,”
in
P
r
oc
e
e
di
ngs
-
2024
5t
h
I
nt
e
r
nat
i
onal
C
onf
e
r
e
nc
e
on
I
nt
e
ll
ig
e
nt
C
om
m
uni
c
at
io
n
T
e
c
hnol
ogi
e
s
and
V
ir
tu
al
M
obi
le
N
e
tw
or
k
s
,
I
C
I
C
V
2024
,
I
E
E
E
,
M
a
r
.
2
024,
pp. 290
–
297
,
do
i:
10.1109/
I
C
I
C
V
62344.2024.00050.
[
11]
Y
.
K
im
,
S
.
Y
oo
,
a
nd
S
.
P
a
r
k,
“
A
r
ul
e
-
ba
s
e
d
s
t
oc
k
tr
a
di
ng
r
e
c
o
mm
e
nda
ti
o
n
s
y
s
t
e
m
us
in
g
s
e
nt
im
e
nt
a
na
l
y
s
is
a
nd
te
c
h
ni
c
a
l
in
di
c
a
t
o
r
s
,”
E
le
c
tr
oni
c
s
(
Sw
it
z
e
r
la
nd)
, vo
l.
14, n
o
. 4, p. 773,
F
e
b. 2025, do
i:
10.3390/
e
l
e
c
tr
o
ni
c
s
14040773.
[
12]
M
.
L
.
N
a
r
a
y
a
na
,
A
.
J
.
K
a
r
th
a
,
A
.
K
.
M
a
nda
l,
R
.
P
,
A
.
S
ur
e
s
h,
a
nd
A
.
C
.
J
o
s
e
,
“
E
ns
e
mbl
e
ti
me
s
e
r
i
e
s
m
o
d
e
ls
f
o
r
s
to
c
k
p
r
ic
e
pr
e
di
c
ti
o
n
a
nd
p
o
r
t
f
ol
i
o
o
pt
i
mi
z
a
ti
o
n
w
it
h
s
e
nt
im
e
nt
a
na
l
y
s
is
,”
J
our
nal
of
I
nt
e
ll
ig
e
nt
I
nf
or
m
at
io
n
Sy
s
te
m
s
,
vo
l.
63,
n
o.
4,
pp. 1079
–
1103, Aug. 2025, do
i:
10.1007/s
10844
-
025
-
00928
-
6.
[
13]
R
.
P
a
ndi
r
i,
K
.
T
hi
la
ga
v
a
th
i,
A
.
M
it
ta
l,
M
.
M
.
A
lm
us
a
w
i,
P
.
N
a
ga
v
e
ni
,
a
nd
T
.
R
.
K
uma
r
,
“
U
ti
l
iz
in
g
r
e
c
ur
r
e
nt
ne
ur
a
l
ne
tw
o
r
ks
f
o
r
s
e
nt
im
e
nt
a
na
l
y
s
is
of
f
in
a
n
c
ia
l
ne
w
s
to
e
nha
nc
e
ma
r
ke
t
pr
e
di
c
ti
o
n
a
nd
in
ve
s
tm
e
nt
s
tr
a
te
g
i
e
s
,”
in
2025
I
nt
e
r
nat
io
nal
C
onf
e
r
e
nc
e
on
A
ut
om
at
io
n
and
C
om
put
at
io
n,
A
U
T
O
C
O
M
2025
,
I
E
E
E
,
M
a
r
.
2025,
pp.
285
–
289
,
do
i:
10.1109/AUT
O
C
O
M
64127.2025.10957345.
[
14]
W
.
Y
a
ng,
Y
.
L
in
,
H
.
X
u
e
,
a
nd
J
.
W
a
ng,
“
R
e
s
e
a
r
c
h
o
n
s
to
c
k
ma
r
k
e
t
s
e
nt
im
e
n
t
a
na
l
y
s
is
a
nd
pr
e
di
c
ti
o
n
m
e
th
o
d
ba
s
e
d
o
n
c
o
n
vo
lu
ti
o
na
l
n
e
ur
a
l
n
e
tw
o
r
k,”
in
P
r
oc
e
e
di
ngs
of
th
e
2025
I
nt
e
r
nat
io
nal
C
onf
e
r
e
nc
e
on
M
ac
hi
ne
L
e
ar
ni
ng
and
N
e
u
r
al
N
e
tw
or
k
s
,
M
L
N
N
2025
, N
e
w
Y
or
k, N
Y
, U
S
A
:
A
C
M
, A
pr
. 2025, pp. 91
–
96
,
do
i:
10.1145/3747227.3
747241.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8776
I
n
t
J
I
n
f
&
C
o
m
m
u
n
T
e
c
hn
o
l
,
Vo
l
.
1
5
,
N
o.
4
,
D
e
c
e
m
be
r
20
2
6
:
1944
-
1953
1952
[
15]
A
.
M
.
M
.
R
a
hma
n,
A
.
U
ddi
n,
a
nd
G
.
“
G
r
a
c
e
”
W
a
ng,
“
E
v
a
lu
a
ti
ng
f
in
a
n
c
ia
l
s
e
nt
i
me
n
t
a
na
l
y
s
is
w
it
h
a
nno
ta
t
o
r
s
in
s
tr
u
c
ti
o
n
a
s
s
is
te
d
pr
o
mp
ti
ng:
e
nha
nc
in
g
c
o
nt
e
x
tu
a
l
in
te
r
pr
e
ta
ti
o
n
a
nd
s
t
o
c
k
pr
e
di
c
ti
o
n
a
c
c
u
r
a
c
y
,”
ar
X
iv
pr
e
pr
in
t
ar
X
iv
:
2505.07871
,
2025,
[
O
nl
i
ne
]
.
A
v
a
il
a
bl
e
:
ht
tp
:/
/a
r
x
i
v
.
or
g/
a
bs
/2
505.07871
[
16]
N
.
S
ma
to
v
,
R
.
K
a
la
s
hni
kov
,
a
nd
A
.
K
a
r
tb
a
y
e
v,
“
D
e
ve
l
o
pm
e
nt
of
c
o
nt
e
x
t
-
ba
s
e
d
s
e
nt
i
me
nt
c
la
s
s
if
i
c
a
ti
o
n
f
or
in
te
l
li
ge
n
t
s
to
c
k
ma
r
ke
t
pr
e
di
c
ti
o
n,”
B
ig
D
at
a and C
ogni
ti
v
e
C
om
put
in
g
, vo
l.
8,
no
. 6, p. 51, M
a
y
2024, d
o
i
:
10.3390/bdcc
8060051.
[
17]
R
.
H
a
s
a
n
e
t
al
.
,
“
E
nha
nc
in
g
ma
lwa
r
e
de
t
e
c
ti
o
n
w
it
h
f
e
a
tu
r
e
s
e
l
e
c
t
i
o
n
a
nd
s
c
a
li
ng
te
c
hni
qu
e
s
us
in
g
ma
c
hi
ne
l
e
a
r
ni
ng
m
o
d
e
ls
,”
Sc
ie
nt
if
ic
R
e
por
t
s
, v
o
l.
15, n
o
. 1, p. 9122, M
a
r
. 2025, d
o
i:
10.10
38/
s
41598
-
025
-
93447
-
x.
[
18]
S
.
A
lk
hl
iwi
,
“
M
o
d
e
ll
in
g
of
h
y
b
r
id
d
e
e
p
l
e
a
r
ni
ng
f
r
a
m
e
w
or
k
w
it
h
r
e
c
u
r
s
iv
e
f
e
a
tu
r
e
e
l
im
in
a
ti
o
n
f
or
di
s
tr
ib
ut
e
d
d
e
ni
a
l
of
s
e
r
v
i
c
e
a
tt
a
c
k de
t
e
c
ti
o
n s
y
s
t
e
ms
,”
Sc
ie
nt
if
ic
R
e
por
ts
, v
o
l.
15, n
o
. 1, p. 3
9261, Nov
. 2025, d
o
i:
10.1038/s
41598
-
025
-
22936
-
w.
[
19]
Y
.
W
a
ng
a
nd
X
.
L
ia
ng,
“
A
ppl
ic
a
ti
o
n
of
r
e
in
f
or
c
e
m
e
nt
l
e
a
r
ni
ng
me
th
o
ds
c
o
mb
in
in
g
gr
a
ph
ne
ur
a
l
ne
tw
o
r
ks
a
nd
s
e
lf
-
a
tt
e
n
t
io
n
me
c
ha
ni
s
ms
i
n s
uppl
y
c
ha
in
r
o
u
te
o
pt
im
i
z
a
ti
o
n,”
Se
ns
or
s
, v
o
l.
25, no
. 3, p. 955, F
e
b. 2025, d
o
i:
10.3390/s
25030955.
[
20]
Z
.
T
ia
n,
D
.
C
he
n,
a
nd
L
.
Z
ha
o
,
“
S
h
o
r
t
-
t
e
r
m
e
n
e
r
g
y
c
o
ns
umpt
i
o
n
pr
e
di
c
ti
o
n
of
la
r
g
e
publ
i
c
bui
ld
in
gs
c
o
mbi
n
e
d
w
it
h
da
ta
f
e
a
tu
r
e
e
ngi
n
e
e
r
in
g a
nd bil
s
tm
-
a
tt
e
nt
i
o
n,”
A
ppl
ie
d Sc
ie
nc
e
s
(
Sw
it
z
e
r
la
nd)
, vo
l.
14, n
o
. 5, p. 2137, M
a
r
. 2024, d
o
i:
10.3390/app140521
3
7.
[
21]
H
.
S
un
e
t
al
.
,
“
A
ga
s
tr
o
in
te
s
ti
na
l
im
a
ge
c
la
s
s
i
f
i
c
a
ti
o
n
me
th
o
d
ba
s
e
d
o
n
im
pr
ov
e
d
a
da
m
a
lg
o
r
i
th
m,”
M
at
he
m
a
ti
c
s
,
v
o
l.
12,
no.
16,
p. 2452, Aug. 2024,
do
i:
10.3390/m
a
th
12162452.
[
22]
M
.
H
.
A
bdul
la
h
a
nd
Y
.
T
.
M
us
ta
f
a
,
“
M
a
c
hi
n
e
L
e
a
r
n
in
g
a
n
d
S
H
A
P
-
ba
s
e
d
a
na
l
y
s
is
of
d
e
f
or
e
s
ta
ti
o
n
a
nd
f
o
r
e
s
t
d
e
gr
a
da
t
io
n
d
y
na
mi
c
s
a
lo
ng
th
e
I
r
a
q
–
T
ur
k
e
y
b
o
r
d
e
r
,”
E
ar
th
(
Sw
it
z
e
r
la
nd)
,
vo
l.
6, n
o
. 2, p. 49, J
un. 2025, d
o
i:
10.3390/
e
a
r
th
6020049.
[
23]
K
a
ggl
e
,
“
A
dv
e
r
s
a
r
ia
l
ma
c
hi
ne
l
e
a
r
ni
ng
d
a
ta
s
e
t,
”
K
a
ggl
e
.
ht
tp
s
:/
/ww
w
.ka
ggl
e
.c
o
m
/d
a
ta
s
e
ts
/c
nr
ie
i
it
/a
dve
r
s
a
r
ia
l
-
ma
c
hi
n
e
-
l
e
a
r
ni
ng
-
da
ta
s
e
t
[
24]
Y
.
C
ha
udha
r
y
,
“
S
t
oc
k
-
ma
r
k
e
t
s
e
nt
im
e
nt
da
ta
s
e
t,
”
ka
ggl
e
.
ht
tp
s
:/
/ww
w
.ka
ggl
e
.c
om/
da
ta
s
e
ts
/
y
a
s
h612/s
t
o
c
kma
r
k
e
t
-
s
e
nt
i
me
nt
-
da
ta
s
e
t
[
25]
S
bha
tt
i,
“
F
in
a
nc
ia
l
s
e
nt
i
me
nt
a
na
l
y
s
is
,”
K
a
ggl
e
.
ht
tp
s
:/
/ww
w
.ka
ggl
e
.
c
o
m/
da
ta
s
e
ts
/s
bha
tt
i/
f
in
a
nc
ia
l
-
s
e
nt
im
e
nt
-
a
na
l
y
s
is
[
26]
G
.
D
e
B
e
r
na
r
di
,
G
.
B
.
G
a
gg
e
r
o
,
F
.
P
a
tr
o
n
e
,
S
.
Z
a
ppa
to
r
e
,
M
.
M
a
r
c
he
s
e
,
a
nd
M
.
M
o
nge
l
li
,
“
R
ul
e
-
ba
s
e
d
e
X
pl
a
in
a
bl
e
a
ut
oe
nc
o
d
e
r
f
or
D
N
S
tu
nn
e
li
ng d
e
te
c
ti
o
n,”
C
om
put
e
r
s
, vo
l.
14, n
o
. 9, p. 375
, S
e
p. 2025, d
o
i:
10.3390/
c
o
mput
e
r
s
14090375.
[
27]
D
.
R
.
I
ns
ua
,
R
.
N
a
v
e
ir
o
,
a
nd
V
.
G
a
ll
e
g
o
,
“
P
e
r
s
p
e
c
t
i
v
e
s
o
n
a
d
ve
r
s
a
r
ia
l
c
la
s
s
if
i
c
a
ti
o
n,”
M
at
he
m
at
ic
s
,
vo
l.
8,
n
o
.
11,
p.
1957,
N
ov
.
2020, do
i:
10.3390/m
a
th
8111957.
[
28]
D
.
K
.
N
a
s
io
po
ul
o
s
,
K
.
I
.
R
o
um
e
li
o
ti
s
,
D
.
P
.
S
a
ka
s
,
K
.
T
o
uda
s
,
a
nd
P
.
R
e
kl
it
is
,
“
F
in
a
nc
ia
l
s
e
nt
im
e
nt
a
na
l
y
s
is
a
nd
c
la
s
s
if
ic
a
ti
on:
a
c
o
mpa
r
a
ti
ve
s
tu
d
y
of
f
in
e
-
tu
ne
d
d
e
e
p
le
a
r
ni
ng
m
o
de
ls
,”
I
nt
e
r
nat
io
nal
J
our
nal
o
f
F
in
anc
ia
l
St
udi
e
s
,
vo
l.
13,
n
o
.
2,
p.
75,
M
a
y
2025, do
i:
10.3390/i
j
f
s
13020075.
[
29]
R
.
A
la
w
a
ji
a
nd
A
.
A
l
o
r
a
in
i,
“
S
e
nt
im
e
nt
a
na
l
y
s
is
of
di
gi
ta
l
ba
n
ki
ng
r
e
v
ie
w
s
us
in
g
ma
c
hi
n
e
l
e
a
r
ni
ng
a
nd
la
r
g
e
la
ngua
g
e
m
o
d
e
l
s
,”
E
le
c
tr
oni
c
s
(
Sw
it
z
e
r
la
nd)
, vo
l.
14, n
o
. 11, p. 2125, M
a
y
2025, d
o
i:
10.3390/
e
l
e
c
tr
o
ni
c
s
14112125.
B
I
OG
RA
P
HI
E
S
OF
AU
T
HO
RS
Su
res
h
Ta
l
a
m
a
l
a
i
s
an
as
s
i
s
t
an
t
p
ro
fe
s
s
o
r
i
n
t
h
e
S
ch
o
o
l
o
f
Bu
s
i
n
e
s
s
at
A
d
i
t
y
a
U
n
i
v
e
rs
i
t
y
,
Su
ra
m
p
al
em
,
K
ak
i
n
a
d
a,
A
n
d
h
ra
Prad
e
s
h
,
In
d
i
a.
H
i
s
a
c
ad
emi
c
i
n
t
e
r
e
s
t
s
i
n
c
l
u
d
e
b
u
s
i
n
e
s
s
a
n
al
y
t
i
c
s
,
art
i
f
i
c
i
a
l
i
n
t
e
l
l
i
g
e
n
ce
,
f
i
n
a
n
c
i
a
l
t
ec
h
n
o
l
o
g
i
e
s
,
a
n
d
d
at
a
-
d
ri
v
e
n
d
eci
s
i
o
n
-
m
ak
i
n
g
.
H
e
c
a
n
b
e
c
o
n
t
ac
t
e
d
at
em
ai
l
:
d
r.
s
u
r
e
s
h
s
o
n
y
@
g
m
ai
l
.
co
m
.
Sukh
a
da
A
l
o
n
i
i
s
an
as
s
i
s
t
an
t
p
ro
fe
s
s
o
r
i
n
t
h
e
D
e
p
art
men
t
o
f
Co
m
p
u
t
er
E
n
g
i
n
ee
r
i
n
g
at
A
.
P.
Sh
a
h
In
s
t
i
t
u
t
e
o
f
T
ec
h
n
o
l
o
g
y
,
T
h
a
n
e
(
W
e
s
t
),
Ma
h
aras
h
t
ra,
In
d
i
a.
H
e
r
re
s
e
ar
ch
i
n
t
e
r
e
s
t
s
i
n
c
l
u
d
e
co
m
p
u
t
e
r
e
n
g
i
n
ee
ri
n
g
,
art
i
fi
c
i
a
l
i
n
t
el
l
i
g
en
ce
,
m
a
c
h
i
n
e
l
e
arn
i
n
g
,
an
d
i
n
t
e
l
l
i
g
e
n
t
c
o
m
p
u
t
i
n
g
s
y
s
t
em
s
.
Sh
e
c
a
n
b
e
c
o
n
t
ac
t
e
d
at
em
ai
l
:
s
s
al
o
n
i
@
ap
s
i
t
.
ed
u
.
i
n
.
J
.
V
i
ja
y
a
l
a
k
s
hm
i
i
s
an
as
s
o
ci
at
e
p
ro
f
e
s
s
o
r
i
n
t
h
e
D
e
p
art
me
n
t
o
f
A
rt
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
ce
an
d
D
at
a
S
c
i
en
ce
at
St
.
J
o
s
e
p
h
’
s
Co
l
l
eg
e
o
f
E
n
g
i
n
ee
r
i
n
g
,
Ch
en
n
a
i
,
T
a
m
i
l
N
ad
u
,
I
n
d
i
a
.
H
e
r
r
e
s
e
ar
c
h
fo
cu
s
e
s
o
n
art
i
fi
ci
al
i
n
t
el
l
i
g
e
n
ce,
d
at
a
s
c
i
e
n
ce,
m
a
ch
i
n
e
l
e
arn
i
n
g
,
a
n
d
i
n
t
el
l
i
g
en
t
d
at
a
an
a
l
y
t
i
c
s
.
S
h
e
c
a
n
b
e
co
n
t
a
c
t
ed
at
em
ai
l
:
j
r.
v
i
j
i
o
f
f
i
c
i
a
l
@
g
m
ai
l
.
c
o
m
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t
J
I
n
f
&
C
o
m
m
u
n
T
e
c
hn
o
l
I
S
S
N:
2252
-
8776
A
dv
e
r
s
ar
ial
-
r
e
s
il
ient
iot
-
e
nabled
f
inancial
int
e
ll
ige
nc
e
s
y
s
tem
us
ing
e
x
plai
nable
…
(
Sur
e
s
h
T
alamala
)
1953
Su
b
h
a
m
P
a
nk
a
j
Sa
m
a
n
ta
ra
y
i
s
a
n
as
s
i
s
t
an
t
p
ro
fes
s
o
r
i
n
t
h
e
D
e
p
art
men
t
o
f
Co
m
p
u
t
e
r
S
c
i
en
ce
a
n
d
E
n
g
i
n
ee
r
i
n
g
at
N
i
g
a
m
In
s
t
i
t
u
t
e
o
f
E
n
g
i
n
ee
r
i
n
g
a
n
d
T
ech
n
o
l
o
g
y
,
O
d
i
s
h
a,
I
n
d
i
a.
H
i
s
ar
e
as
o
f
i
n
t
e
r
e
s
t
i
n
c
l
u
d
e
c
o
m
p
u
t
er
s
c
i
e
n
ce,
m
a
ch
i
n
e
l
e
ar
n
i
n
g
,
d
at
a
an
al
y
t
i
c
s
,
a
n
d
i
n
t
el
l
i
g
en
t
co
m
p
u
t
i
n
g
.
H
e
c
a
n
b
e
co
n
t
a
c
t
ed
at
em
a
i
l
:
j
p
p
4
2
0
0
3
@
y
a
h
o
o
.
c
o
.
i
n
.
K.
Sw
a
n
th
a
n
a
i
s
an
as
s
i
s
t
an
t
p
ro
fe
s
s
o
r
i
n
t
h
e
D
e
p
art
m
en
t
o
f
Co
m
p
u
t
e
r
Sc
i
e
n
c
e
an
d
E
n
g
i
n
ee
ri
n
g
at
K
o
n
e
ru
L
ak
s
h
m
ai
ah
E
d
u
c
at
i
o
n
Fo
u
n
d
at
i
o
n
,
H
y
d
e
rab
ad
,
T
e
l
an
g
an
a,
I
n
d
i
a.
H
e
r
re
s
e
arc
h
i
n
t
e
re
s
t
s
i
n
c
l
u
d
e
art
i
fi
c
i
al
i
n
t
e
l
l
i
g
e
n
c
e
,
mac
h
i
n
e
l
e
arn
i
n
g
,
d
at
a
an
al
y
t
i
c
s
,
an
d
s
o
f
t
w
are
e
n
g
i
n
ee
ri
n
g
.
Sh
e
c
an
b
e
c
o
n
t
ac
t
e
d
at
em
ai
l
:
k
s
w
an
t
h
an
a@
g
m
ai
l
.
co
m
.
N
i
s
a
rg
Ga
n
dh
ew
a
r
i
s
an
as
s
i
s
t
an
t
p
ro
fe
s
s
o
r
i
n
t
h
e
D
e
p
art
me
n
t
o
f
A
rt
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
ce
an
d
Ma
c
h
i
n
e
L
e
arn
i
n
g
at
Ra
m
d
eo
b
ab
a
U
n
i
v
e
rs
i
t
y
,
N
a
g
p
u
r,
Ma
h
aras
h
t
ra,
I
n
d
i
a.
H
i
s
r
e
s
e
ar
ch
i
n
t
e
r
e
s
t
s
i
n
c
l
u
d
e
art
i
f
i
c
i
a
l
i
n
t
e
l
l
i
g
e
n
ce
,
mac
h
i
n
e
l
e
ar
n
i
n
g
,
d
ee
p
l
e
arn
i
n
g
,
a
n
d
i
n
t
el
l
i
g
en
t
s
y
s
t
em
s
.
H
e
c
a
n
b
e
c
o
n
t
ac
t
e
d
at
em
a
i
l
:
n
i
s
arg
.
g
an
d
h
ew
ar
@
g
m
a
i
l
.
c
o
m
.
Evaluation Warning : The document was created with Spire.PDF for Python.