I
A
E
S
I
n
t
e
r
n
at
io
n
al
Jou
r
n
al
of
A
r
t
if
ic
ia
l
I
n
t
e
ll
ig
e
n
c
e
(
I
J
-
AI
)
V
ol
. 14, No. 6, D
e
c
e
m
be
r
2025
, pp.
5231
~
5239
I
S
S
N
:
2252
-
8938
,
D
O
I
:
10.11591/
ij
a
i.
v
16
.i
6
.pp
5231
-
5239
5231
Jou
r
n
al
h
om
e
page
:
ht
tp
:
//
ij
ai
.
ia
e
s
c
or
e
.c
om
E
xp
l
or
i
n
g t
h
e
i
n
f
l
u
e
n
c
e
of
sof
t
i
n
f
or
m
at
i
on
f
r
om
e
c
on
om
i
c
n
e
w
s on
e
xc
h
an
ge
r
at
e
an
d
gol
d
p
r
i
c
e
m
ove
m
e
n
t
s
R
ah
ar
d
it
o D
io
P
r
as
t
ow
o, I
n
d
r
a B
u
d
i,
A
m
an
ah
R
am
ad
ia
h
, A
r
is
B
u
d
i
S
an
t
os
o, P
r
ab
u
K
r
e
s
n
a P
u
t
r
a
F
a
c
ul
t
y of
C
om
put
e
r
S
c
i
e
nc
e
, U
ni
ve
r
s
i
t
a
s
I
ndone
s
i
a
, D
e
pok, I
ndone
s
i
a
A
r
t
ic
le
I
n
f
o
A
B
S
T
R
A
C
T
A
r
ti
c
le
h
is
to
r
y
:
R
e
c
e
iv
e
d
J
ul
16, 2024
R
e
vi
s
e
d
O
c
t
5, 2025
A
c
c
e
pt
e
d
O
c
t
18, 2025
Information
on
business
condition
s
is
an
important
conc
ern
for
market
players
and
regulators.
Hard
information
relates
to
easily
va
lidated
characterist
ics
such
as
producti
on
levels
and
employm
ent
conditi
o
ns.
In
contrast,
soft
informat
ion
such
as
consumer
and
public
perceptio
ns
—
is
subjective
and
difficult
to
verify.
Although
previous
studies
on
hard
a
nd
soft
information
mainly
focus
on
microeco
nomics
and
banking,
current
developments
in
big
data
and
machine
learning
enable
broader
applications
in
financial
m
arket
analysis.
This
study
combined
VADER
sen
timent
analysis
and
support
vector
machine
(
SVM
)
classifi
cation
(accuracy
=
85%)
to
analyze
economic
news,
followed
by
Granger
causality
and
m
ultiple
linear
regression
to
examine
causal
effects
and
predictive
relationship
s.
The
findings
reveal
that
negative
news
sentiment
and
the
Indonesian
Rupiah
(
IDR
)
exchange
rate
influence
each
other,
while
positi
ve
sentimen
t
has
no
causal
impact
on
the
exchange
rate.
Both
negative
and
positive
sent
iments
affect
gold
prices,
whereas
gold
price
movements
do
not
inf
luence
sentiment.
Regression
analysis
shows
that
negative
sentiment
has
a
stronger
effect
in
decreasing
the
IDR
exchange
rate
than
positive
sentiment,
w
ith
the
model
explaining
approximately
20%
of
the
variance.
Integrating
sen
timent
and
exchange
rate
data
enhances
the
predictive
model
for
gold
price
for
ecasting
and
highli
ghts
the
asymmetr
ic
roles
of
positi
ve
and
n
egative
news in financial dynamics.
K
e
y
w
o
r
d
s
:
E
c
onomi
c
ne
w
s
s
e
nt
im
e
nt
E
xc
ha
nge
r
a
te
G
ol
d pr
ic
e
m
ove
m
e
nt
s
G
r
a
nge
r
c
a
us
a
li
ty
M
ul
ti
pl
e
l
in
e
a
r
r
e
gr
e
s
s
io
n
This is an
open
acce
ss artic
le unde
r the
CC BY
-
SA
license.
C
or
r
e
s
pon
di
n
g A
u
th
or
:
R
a
ha
r
di
to
D
io
P
r
a
to
w
o
F
a
c
ul
ty
of
C
om
put
e
r
S
c
ie
nc
e
, U
ni
ve
r
s
it
a
s
I
ndone
s
ia
M
a
r
gonda
R
a
ya
,
D
e
pok, W
e
s
t
J
a
v
a
16424
, I
ndone
s
ia
E
m
a
il
:
r
a
ha
r
di
to
.di
o@
ui
.a
c
.i
d
1.
I
N
T
R
O
D
U
C
T
I
O
N
B
ot
h
now
a
nd
in
th
e
f
ut
ur
e
,
bus
in
e
s
s
c
ondi
ti
ons
w
il
l
a
lwa
y
s
be
a
n
im
por
ta
nt
c
onc
e
r
n
f
or
m
a
r
ke
t
pl
a
ye
r
s
a
nd
r
e
gul
a
to
r
s
a
s
a
ba
s
is
f
or
de
te
r
m
in
in
g
pol
ic
y
[
1]
.
T
w
o
ty
pe
s
of
in
f
or
m
a
ti
on
c
a
n
be
u
s
e
d
in
f
or
e
c
a
s
ti
ng,
"
ha
r
d"
in
f
or
m
a
ti
on
in
th
e
f
or
m
o
f
r
e
a
l
da
ta
th
a
t
c
a
n
be
obt
a
in
e
d
di
r
e
c
tl
y,
s
uc
h
a
s
pr
oduc
ti
on
le
ve
ls
a
nd
e
m
pl
oym
e
nt
c
ondi
ti
ons
.
I
n
c
ont
r
a
s
t,
"
s
of
t"
in
f
o
r
m
a
ti
on
is
da
ta
us
ua
ll
y
obt
a
in
e
d
f
r
om
th
e
pe
r
c
e
pt
io
ns
of
c
ons
um
e
r
s
a
nd t
he
ge
ne
r
a
l
publ
ic
[
2]
.
E
ve
n t
hou
gh dis
c
us
s
io
ns
r
e
la
te
d t
o ha
r
d i
nf
or
m
a
ti
on a
nd
s
of
t
in
f
or
m
a
ti
on
ha
ve
pr
e
vi
ous
ly
be
e
n
w
id
e
ly
di
s
c
u
s
s
e
d
in
th
e
c
ont
e
xt
of
m
ic
r
oe
c
onomi
c
s
,
e
s
pe
c
ia
ll
y
ba
nki
ng,
te
c
hnol
ogi
c
a
l
de
ve
lo
pm
e
nt
s
ha
v
e
e
nc
our
a
ge
d
th
e
u
s
e
of
th
is
c
onc
e
pt
in
f
in
a
nc
ia
l
m
a
r
ke
ts
a
nd
in
s
ti
tu
ti
ons
out
s
id
e
ba
nki
ng
[
3]
.
H
a
r
d
in
f
o
r
m
a
t
io
n
ha
s
p
r
o
pe
r
t
ie
s
t
ha
t
c
a
n
be
di
r
e
c
tl
y
va
l
id
a
te
d
a
nd
ge
ne
r
a
l
ly
a
gr
e
e
d
u
po
n
by
va
r
io
us
pa
r
t
ie
s
,
s
uc
h
a
s
de
m
o
g
r
a
ph
ic
da
ta
[
4]
.
A
no
th
e
r
e
x
a
m
p
le
is
f
in
a
nc
ia
l
r
e
p
o
r
t
da
ta
,
w
hi
c
h
ba
nks
c
om
m
on
ly
us
e
d
to
p
r
ov
id
e
l
oa
ns
t
o
s
m
a
l
l
bus
i
ne
s
s
e
s
[
5]
.
A
no
th
e
r
c
ha
r
a
c
t
e
r
is
t
ic
o
f
h
a
r
d
in
f
o
r
m
a
t
io
n
is
t
ha
t
it
is
ve
r
i
f
ia
bl
e
a
n
d
ha
s
a
s
s
e
s
s
m
e
n
t
s
ta
n
da
r
ds
us
e
d
by
pe
op
le
[
6]
,
r
e
s
u
lt
in
g
i
n
s
pe
c
i
f
ic
in
f
o
r
m
a
t
io
n
a
n
d
c
le
a
r
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
,
V
ol
. 14, No. 6, D
e
c
e
m
be
r
2025
:
5231
-
5239
5232
a
ns
w
e
r
[
7
]
.
M
e
a
nw
h
il
e
,
s
o
f
t
i
nf
or
m
a
ti
o
n
is
da
t
a
t
ha
t
c
a
n
b
e
in
te
r
p
r
e
te
d
d
if
f
e
r
e
nt
ly
a
nd
r
e
q
ui
r
e
s
m
o
r
e
e
f
f
o
r
t
to
ve
r
i
f
y
i
ts
v
a
lu
e
[
4
]
.
A
dva
nc
e
s
in
bi
g
da
ta
te
c
hnol
ogy
in
r
e
c
e
nt
y
e
a
r
s
h
a
ve
e
n
a
bl
e
d
th
e
tr
a
ns
f
or
m
a
ti
on
of
qua
li
ta
ti
ve
da
ta
in
to
qua
nt
it
a
ti
ve
da
ta
[
8]
,
e
s
p
e
c
ia
ll
y
in
f
in
a
nc
i
a
l
te
c
hnol
ogy,
w
hi
c
h
ha
s
op
e
ne
d
up
th
e
pos
s
ib
il
it
y
of
ut
il
iz
in
g
s
of
t
in
f
or
m
a
ti
on
to
be
c
onve
r
te
d
in
to
ha
r
d
in
f
or
m
a
ti
on
to
be
c
om
e
m
or
e
e
f
f
ic
ie
nt
[
9]
.
F
il
om
e
ni
e
t
al
.
[
10]
e
xa
m
in
e
d
how
to
in
te
gr
a
te
s
of
t
in
f
or
m
a
ti
on
c
ons
is
ti
ng
of
pr
e
-
de
f
in
e
d
que
s
ti
onna
ir
e
s
a
nd
s
e
nt
im
e
nt
a
na
ly
s
is
f
r
om
lo
a
n
a
ppl
ic
a
ti
on
te
xt
s
in
to
s
ta
ti
s
ti
c
a
l
da
ta
to
pr
e
di
c
t
c
or
po
r
a
te
de
f
a
ul
t,
w
hi
c
h
s
how
s
a
s
ig
ni
f
ic
a
nt
im
pa
c
t
in
i
nc
r
e
a
s
in
g pr
e
di
c
ti
ve
a
bi
li
ty
w
he
n i
nt
e
gr
a
ti
ng t
he
s
e
da
ta
. I
n a
br
oa
de
r
c
ont
e
xt
, ne
w
s
he
a
dl
in
e
s
e
nt
im
e
nt
ha
s
be
gun
to
be
w
id
e
ly
a
ppl
ie
d
a
nd
is
c
ons
id
e
r
e
d
m
or
e
e
f
f
ic
ie
nt
in
m
e
a
s
ur
in
g
in
ve
s
to
r
s
e
nt
im
e
nt
.
I
nve
s
to
r
s
e
nt
im
e
nt
is
a
pa
r
t
th
a
t
c
a
nnot
b
e
le
f
t
out
w
he
n
c
a
r
r
yi
ng
out
pr
ic
e
pr
e
di
c
ti
on
a
na
ly
s
is
f
or
a
c
om
m
odi
ty
in
a
ddi
ti
on
to
in
di
c
a
to
r
s
s
uc
h
a
s
pr
ic
e
in
de
x,
e
xc
ha
ng
e
r
a
te
,
a
nd
p
r
oduc
ti
on
le
ve
ls
[
11]
.
A
pa
r
t
f
r
om
th
a
t,
s
e
ve
r
a
l
s
tu
di
e
s
ha
ve
a
l
s
o
pr
ove
d
a
s
tr
ong
r
e
la
ti
ons
hi
p
be
twe
e
n
ne
w
s
a
nd
s
oc
ia
l
m
e
di
a
in
f
or
m
a
ti
on
on
in
ve
s
to
r
s
e
nt
im
e
nt
,
th
e
vol
um
e
of
a
c
om
m
odi
ty
on
th
e
m
a
r
ke
t,
a
n
d
a
s
s
e
t
pr
ic
e
s
by
ut
il
iz
in
g
te
xt
m
in
in
g
a
nd
c
la
s
s
if
ic
a
ti
on t
e
c
hni
que
s
[
12]
.
I
t
ha
s
be
e
n
pr
ove
n
th
a
t
m
a
c
r
oe
c
onomi
c
ne
w
s
s
e
nt
im
e
nt
c
a
n
be
us
e
d
to
a
na
ly
z
e
da
il
y
e
xc
ha
nge
r
a
te
m
ove
m
e
nt
s
,
one
of
w
hi
c
h
is
r
e
s
e
a
r
c
h
c
onduc
te
d
by
M
a
o
e
t
al
.
[
13]
,
w
he
r
e
th
e
r
e
s
e
a
r
c
h
in
te
gr
a
te
s
ne
w
s
s
e
nt
im
e
nt
w
it
h
th
e
pr
e
di
c
ti
on
m
ode
l
to
f
in
d
out
w
he
th
e
r
ne
w
s
s
e
nt
im
e
nt
c
a
n
im
pr
ove
th
e
p
e
r
f
or
m
a
nc
e
of
m
ode
ls
.
A
na
ly
s
is
of
th
e
r
e
la
ti
ons
hi
p
be
twe
e
n
ne
w
s
s
e
nt
im
e
nt
a
nd
m
ove
m
e
nt
s
in
th
e
va
lu
e
of
o
th
e
r
e
c
onomi
c
in
di
c
a
to
r
s
ha
s
a
ls
o be
e
n c
a
r
r
ie
d out, one
of
w
hi
c
h i
s
t
he
pr
ic
e
o
f
gol
d, w
he
r
e
t
hi
s
a
na
ly
s
is
s
how
s
t
ha
t
ne
ga
ti
ve
s
e
nt
im
e
nt
ha
s
m
or
e
in
f
lu
e
nc
e
on
gol
d
pr
ic
e
r
e
s
pon
s
e
s
[
14]
.
B
a
s
e
d
on
th
e
r
e
s
e
a
r
c
h,
th
e
r
e
is
a
ga
p
f
or
f
ur
th
e
r
e
xa
m
in
a
ti
on
th
a
t
f
oc
us
e
s
on
th
e
di
f
f
e
r
e
nc
e
s
in
th
e
in
f
lu
e
nc
e
of
pos
it
iv
e
a
nd
ne
ga
ti
ve
s
e
nt
im
e
nt
on
m
a
c
r
oe
c
onomi
c
ne
w
s
on
e
xc
ha
nge
r
a
te
s
a
nd
gol
d
pr
ic
e
s
.
A
pa
r
t
f
r
om
th
a
t,
ot
he
r
pot
e
nt
ia
l
r
e
s
e
a
r
c
h
is
r
e
la
te
d
to
pr
e
vi
ous
r
e
s
e
a
r
c
h,
w
hi
c
h
in
te
gr
a
te
d
s
of
t
in
f
or
m
a
ti
on
to
im
pr
ove
th
e
qu
a
li
ty
of
pr
e
di
c
ti
ve
a
na
ly
s
i
s
c
om
pa
r
e
d
to
pr
e
vi
ous
ly
onl
y
ut
il
iz
in
g
ha
r
d
in
f
o
r
m
a
ti
on,
r
e
ga
r
di
ng
th
e
im
pa
c
t
of
in
te
gr
a
ti
ng
ne
w
s
s
e
nt
im
e
nt
w
it
h t
he
e
xc
ha
nge
r
a
te
a
nd i
t
s
r
e
la
ti
ons
hi
p w
it
h gold pr
ic
e
m
ove
m
e
nt
s
.
2.
M
E
T
H
O
D
P
r
e
di
c
ti
ve
a
na
ly
s
is
of
a
n
e
xc
ha
nge
r
a
te
of
te
n
us
e
s
ti
m
e
s
e
r
ie
s
m
e
th
ods
to
de
te
r
m
in
e
th
e
pr
ic
e
tr
e
nd
of
a
c
ur
r
e
nc
y. M
a
c
hi
ne
l
e
a
r
ni
ng i
s
us
e
d t
o i
nc
r
e
a
s
e
s
pe
e
d a
nd e
f
f
ic
ie
nc
y, a
nd s
e
nt
im
e
nt
a
na
ly
s
is
i
s
i
nt
e
gr
a
te
d
to
in
c
r
e
a
s
e
pr
e
di
c
ti
ve
c
a
p
a
bi
li
ti
e
s
,
a
s
r
e
s
e
a
r
c
h
c
onduc
te
d
by
X
ue
li
ng
e
t
al
.
[
15]
T
he
s
tu
dy
us
e
d
th
e
C
N
N
m
e
th
od
to
e
xt
r
a
c
t
lo
c
a
l
f
e
a
tu
r
e
s
f
r
om
th
e
te
xt
,
c
om
bi
ne
d
w
i
th
L
S
T
M
to
c
a
r
r
y
out
tr
e
nd
e
xc
ha
nge
r
a
te
a
na
ly
s
is
.
M
a
o
e
t
al
.
[
13]
c
onduc
te
d
a
s
tu
dy
on
pr
e
di
c
ti
ng
c
o
m
pl
e
x
e
xc
ha
nge
r
a
te
m
ove
m
e
nt
s
us
in
g
C
N
N
-
L
S
T
M
a
nd
tr
a
ns
f
or
m
e
r
m
ode
ls
c
om
bi
ne
d
w
it
h
B
E
R
T
-
ba
s
e
d
ne
w
s
s
e
nt
im
e
nt
,
s
how
in
g
th
a
t
lo
ng
-
te
r
m
ne
w
s
e
f
f
e
c
ts
c
a
n
e
nha
n
c
e
pr
e
di
c
ti
on
a
c
c
ur
a
c
y
.
P
r
e
di
c
ti
ve
a
na
ly
s
is
o
f
gol
d
pr
ic
e
s
th
a
t
in
te
gr
a
te
s
ne
w
s
s
e
nt
im
e
nt
f
a
c
to
r
s
ha
s
a
ls
o
b
e
e
n
w
id
e
ly
c
a
r
r
ie
d
out
.
J
unj
ie
a
nd
M
e
ngoni
[
16]
us
e
d
P
e
a
r
s
on
c
or
r
e
la
ti
on
to
c
om
pa
r
e
th
e
r
e
la
ti
ons
hi
p be
twe
e
n 1
-
da
y a
nd 5
-
da
y ne
w
s
s
e
nt
im
e
nt
a
nd gold
pr
ic
e
m
ove
m
e
nt
s
.
I
n
th
is
s
tu
dy,
t
he
a
ut
hor
s
a
im
e
d
to
d
e
te
r
m
in
e
ne
w
s
s
e
nt
im
e
n
t'
s
i
m
pa
c
t
o
n
th
e
e
xc
h
a
ng
e
r
a
t
e
a
nd
g
ol
d
pr
ic
e
m
ove
m
e
nt
.
T
he
a
u
th
or
s
u
s
e
d
th
e
G
r
a
nge
r
c
a
u
s
a
li
ty
a
na
ly
s
is
to
f
in
d
out
w
he
t
he
r
a
ti
m
e
s
e
r
i
e
s
v
a
r
ia
bl
e
c
a
n
b
e
u
s
e
d
to
pr
e
di
c
t
ot
h
e
r
ti
m
e
s
e
r
ie
s
va
r
i
a
bl
e
,
s
uc
h
a
s
r
e
s
e
a
r
c
h
c
a
r
r
ie
d
o
ut
by
J
i
a
ng
e
t
al
.
[
17]
,
w
ho
u
s
e
d
G
r
a
nge
r
c
a
u
s
a
li
ty
t
o s
tu
dy t
he
e
f
f
e
c
t
of
va
r
i
a
bl
e
c
r
ud
e
oi
l
pr
ic
e
s
on t
he
e
x
c
ha
nge
r
a
t
e
, t
hi
s
r
e
s
e
a
r
c
h s
h
ow
s
t
ha
t
a
s
ig
ni
f
i
c
a
nt
c
a
u
s
a
l
im
p
a
c
t
b
e
tw
e
e
n
c
r
u
de
oi
l
pr
ic
e
s
a
nd
th
e
e
xc
ha
nge
r
a
te
w
il
l
onl
y oc
c
ur
w
he
n
th
e
va
lu
e
of
a
c
ur
r
e
nc
y
is
a
t
a
c
e
r
t
a
in
e
xt
r
e
m
e
c
on
di
ti
on
.
G
r
a
ng
e
r
c
a
us
a
li
ty
a
n
a
ly
s
i
s
h
a
s
a
l
s
o
be
e
n
c
a
r
r
ie
d
out
on
g
ol
d
pr
i
c
e
m
ove
m
e
nt
s
,
w
h
e
r
e
th
e
r
e
s
e
a
r
c
h
w
a
s
c
a
r
r
ie
d
out
to
s
tu
d
y
th
e
i
m
pa
c
t
of
th
e
s
pr
e
a
d
of
C
O
V
I
D
-
19 on
gol
d pr
ic
e
m
ove
m
e
nt
s
.
T
h
e
r
e
s
e
a
r
c
h
s
how
s
a
s
ig
ni
f
ic
a
nt
r
e
s
po
ns
e
to
th
e
gol
d
pr
ic
e
m
ov
e
m
e
nt
f
r
om
th
e
in
c
r
e
a
s
e
in
C
O
V
I
D
-
1
9 c
a
s
e
s
, but
no
c
a
u
s
a
l
r
e
la
ti
on
s
hi
p
w
a
s
f
oun
d w
h
e
n t
h
e
oppo
s
it
e
te
s
t
w
a
s
c
a
r
r
i
e
d ou
t
[
18]
.
F
ur
th
e
r
m
or
e
,
m
ul
ti
pl
e
li
ne
a
r
r
e
gr
e
s
s
io
n
a
na
ly
s
is
w
a
s
c
onduc
te
d
to
in
ve
s
ti
ga
te
th
e
in
f
lu
e
nc
e
of
I
ndone
s
ia
n
R
upi
a
h
(
I
D
R
)
e
xc
ha
nge
r
a
te
a
nd
e
c
onomi
c
ne
w
s
s
e
nt
im
e
nt
on
gol
d
p
r
ic
e
s
a
s
th
e
de
pe
nde
nt
va
r
ia
bl
e
.
T
hi
s
m
e
th
od
w
a
s
us
e
d
to
a
s
s
e
s
s
w
he
th
e
r
s
e
nt
im
e
nt
,
a
s
a
f
or
m
of
s
of
t
in
f
or
m
a
ti
on,
c
oul
d
e
nha
nc
e
th
e
pr
e
di
c
ti
ve
a
bi
li
ty
of
m
ode
ls
ty
pi
c
a
ll
y
ba
s
e
d
on
ha
r
d
in
f
or
m
a
ti
on.
I
n
a
ddi
ti
on,
th
is
s
tu
dy
e
xa
m
in
e
s
w
he
th
e
r
th
e
r
e
is
a
di
f
f
e
r
e
nc
e
be
twe
e
n
th
e
im
pa
c
t
of
ne
ga
ti
ve
a
nd
pos
it
iv
e
s
e
nt
im
e
nt
.
R
e
s
e
a
r
c
h
c
onduc
t
e
d
by
A
bdou
e
t
al
.
[
19]
us
e
d
li
ne
a
r
r
e
gr
e
s
s
io
n a
na
ly
s
i
s
to
a
na
ly
z
e
th
e
in
f
lu
e
nc
e
of
T
w
it
te
r
s
e
nt
im
e
nt
by
c
om
p
a
r
in
g
th
r
e
e
r
e
gr
e
s
s
io
n
m
ode
ls
,
w
hi
c
h
s
how
s
th
a
t
th
e
r
e
is
a
w
e
a
k
c
or
r
e
la
ti
on
be
twe
e
n
T
w
it
te
r
s
e
nt
im
e
nt
a
nd
gol
d
pr
ic
e
s
.
A
not
he
r
r
e
s
e
a
r
c
h
w
a
s
c
onduc
t
e
d
by
J
ia
nyi
e
t
al
.
[
20]
,
w
hi
c
h
a
na
ly
z
e
d
th
e
c
ondi
ti
on
of
C
O
V
I
D
-
19
on
gol
d pr
ic
e
s
. T
he
r
e
s
ul
ts
s
how
e
d t
ha
t
th
e
C
O
V
I
D
-
19
c
a
s
e
c
oul
d
be
us
e
d t
o e
xpl
a
in
gol
d pr
ic
e
m
ove
m
e
nt
s
.
F
ig
ur
e
1
s
how
s
th
e
r
e
s
e
a
r
c
h
m
e
th
odol
ogy
us
e
d
by
th
e
a
ut
ho
r
s
to
a
ns
w
e
r
th
e
r
e
s
e
a
r
c
h
que
s
ti
ons
.
T
hr
e
e
ty
pe
s
of
d
a
ta
w
e
r
e
c
ol
le
c
t
e
d:
e
c
onomi
c
ne
w
s
s
e
nt
im
e
nt
da
ta
,
I
D
R
e
xc
ha
ng
e
r
a
te
,
a
nd
gol
d
pr
ic
e
m
ove
m
e
nt
s
.
B
e
f
or
e
th
e
c
a
u
s
a
li
ty
a
na
ly
s
is
,
s
e
nt
im
e
nt
a
na
ly
s
i
s
w
a
s
c
a
r
r
ie
d
out
on
ne
w
s
da
ta
to
c
la
s
s
if
y
th
e
pos
it
iv
e
a
nd
ne
ga
ti
ve
s
e
nt
im
e
nt
.
T
he
c
a
us
a
li
ty
a
n
a
ly
s
is
is
c
a
r
r
ie
d
out
to
a
s
s
e
s
s
th
e
im
pa
c
t
of
e
c
onomi
c
n
e
w
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
E
x
pl
or
in
g t
he
i
nf
lu
e
nc
e
of
s
of
t
in
fo
r
m
at
io
n f
r
om
e
c
onomic
ne
w
s
on
…
(
R
ahar
di
to
D
io
P
r
as
to
w
o
)
5233
s
e
nt
im
e
nt
on
th
e
I
D
R
e
xc
ha
nge
r
a
te
a
nd
gol
d
pr
ic
e
m
ove
m
e
nt
.
A
f
te
r
th
a
t,
a
r
e
gr
e
s
s
io
n
a
na
ly
s
is
is
c
a
r
r
ie
d
out
to
de
te
r
m
in
e
t
he
e
f
f
e
c
t
of
e
c
onomi
c
ne
w
s
s
e
nt
im
e
nt
on t
he
m
od
e
l
us
e
d t
o pr
e
di
c
t
gol
d pr
ic
e
s
.
2.1. Dat
a c
ol
le
c
t
io
n
T
he
a
ut
hor
s
us
e
D
e
ti
kF
in
a
nc
e
a
s
a
n
onl
in
e
m
e
di
a
s
our
c
e
to
obt
a
in
ne
w
s
da
ta
r
e
la
te
d
to
th
e
e
c
onomy
in
I
ndone
s
ia
be
c
a
us
e
D
e
ti
kF
in
a
nc
e
is
c
ons
id
e
r
e
d
a
n
onl
in
e
ne
w
s
m
e
di
a
pl
a
tf
or
m
w
it
h
m
a
ny
r
e
a
de
r
s
in
I
ndone
s
ia
.
T
he
a
ut
hor
s
s
c
r
a
pe
d
th
e
da
ta
us
in
g
th
e
b
e
a
ut
if
ul
s
ou
p
li
br
a
r
y
in
th
e
P
yt
hon
pr
og
r
a
m
m
in
g
la
ngua
ge
f
r
om
J
a
nua
r
y
1,
2020
to
A
pr
il
30,
2024.
B
e
a
ut
if
ul
S
oup
is
a
l
ib
r
a
r
y
th
a
t
c
a
n
e
xt
r
a
c
t
H
T
M
L
a
nd
X
M
L
da
ta
[
21]
.
T
he
ot
he
r
two
da
ta
w
e
r
e
obt
a
in
e
d
by
dow
nl
oa
di
ng
da
ta
s
e
ts
f
r
om
tr
us
te
d
s
our
c
e
s
.
S
pe
c
if
ic
a
ll
y,
th
e
a
ut
hor
s
dow
nl
oa
de
d
da
ta
f
r
om
in
ve
s
ti
ng.c
om
to
obt
a
in
th
e
I
D
R
e
xc
h
a
nge
r
a
te
a
ga
in
s
t
U
S
D
a
nd
gol
d
pr
ic
e
m
ove
m
e
nt
s
. T
he
s
e
da
ta
s
e
ts
a
r
e
vi
s
ua
li
z
e
d i
n F
ig
ur
e
2, w
hi
c
h pr
e
s
e
nt
s
t
he
t
r
e
nd
s
of
bot
h va
r
ia
bl
e
s
ove
r
t
im
e
.
F
ig
ur
e
1. R
e
s
e
a
r
c
h
m
e
th
odol
ogy
F
ig
ur
e
2. I
D
R
e
xc
ha
nge
r
a
te
a
nd
gol
d pr
ic
e
m
ove
m
e
nt
c
h
a
r
t
(
in
ve
s
ti
ng.c
om
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
,
V
ol
. 14, No. 6, D
e
c
e
m
be
r
2025
:
5231
-
5239
5234
2.2. S
e
n
t
im
e
n
t
an
al
ys
is
T
h
e
a
ut
h
or
s
us
e
d
a
s
e
nt
im
e
nt
a
n
a
ly
s
i
s
t
ool
c
a
l
le
d
V
A
D
E
R
,
w
hi
c
h
i
s
g
e
n
e
r
a
ll
y
u
s
e
d
to
c
a
r
r
y
ou
t
s
e
m
a
nt
i
c
s
c
or
i
ng
f
or
s
o
c
i
a
l
m
e
di
a
w
it
h
t
h
e
b
a
s
i
c
s
e
nt
im
e
nt
le
xi
c
o
n
[
22
]
.
T
h
e
s
e
nt
im
e
nt
r
e
s
ul
t
s
w
it
h v
a
l
ue
s
c
l
os
e
to
+
1
a
r
e
p
os
it
i
ve
,
va
lu
e
s
c
lo
s
e
t
o
-
1
a
r
e
ne
ga
ti
v
e
,
a
nd
v
a
lu
e
s
a
r
oun
d
-
0.5
t
o
0
.5
a
r
e
n
e
ut
r
a
l
[
2
3]
.
T
hi
s
s
tu
dy
c
om
bi
n
e
s
V
A
D
E
R
w
i
th
th
e
s
up
por
t
v
e
c
to
r
m
a
c
h
in
e
(
S
V
M
)
m
a
c
hi
n
e
l
e
a
r
ni
n
g
a
lg
or
it
hm
,
s
im
il
a
r
to
th
e
a
ppr
oa
c
h
us
e
d
in
[
22]
,
[
23]
S
V
M
i
s
u
s
e
d
to
c
la
s
s
i
f
y
pr
e
vi
ou
s
ly
ve
c
t
or
i
z
e
d
d
a
t
a
u
s
in
g
t
h
e
T
F
-
I
D
F
m
e
th
o
d,
w
hi
c
h
r
e
s
ul
t
s
f
r
om
c
ha
ngi
ng
t
e
xt
d
a
t
a
i
nt
o
a
n
um
e
r
i
c
a
l
r
e
pr
e
s
e
nt
a
ti
on
[
2
4]
.
B
a
s
e
d
on
th
e
r
e
s
ul
ts
of
c
la
s
s
if
ic
a
ti
on
m
ode
li
ng
c
a
r
r
ie
d
o
ut
w
it
h
S
V
M
,
T
a
bl
e
1
s
how
s
th
e
c
la
s
s
if
ic
a
ti
on
r
e
por
t
f
r
om
th
e
m
ode
l
w
it
h
th
e
s
c
or
e
s
of
pr
e
c
is
i
on,
r
e
c
a
ll
,
F
1
-
s
c
or
e
,
a
nd
a
c
c
ur
a
c
y
to
m
e
a
s
ur
e
th
e
c
la
s
s
if
ic
a
ti
on
pe
r
f
or
m
a
nc
e
[
25]
.
T
he
pr
e
c
is
io
n
s
c
or
e
s
how
s
th
a
t
c
or
r
e
c
t
pos
it
iv
e
pr
e
di
c
ti
ons
w
e
r
e
83
%
(
ne
ga
ti
ve
a
nd
ne
ut
r
a
l
la
be
ls
)
a
nd
87%
(
pos
it
iv
e
la
be
ls
)
of
th
e
to
ta
l
pos
it
iv
e
pr
e
di
c
ti
ons
.
T
he
r
e
c
a
ll
s
c
or
e
s
how
s
t
ha
t
th
e
c
or
r
e
c
t
pos
it
iv
e
pr
e
di
c
ti
ons
w
e
r
e
78%
(
ne
ga
ti
ve
l
a
be
l)
, 87%
(
ne
ut
r
a
l
la
be
l)
, a
nd 89%
(
pos
it
iv
e
la
be
l)
of
th
e
to
ta
l
a
c
tu
a
l
po
s
it
iv
e
s
. T
he
F
1
-
s
c
or
e
s
ho
w
s
th
e
ha
r
m
oni
z
e
d
va
lu
e
of
pr
e
c
is
io
n a
nd
r
e
c
a
ll
,
w
it
h
a
ll
va
lu
e
s
be
in
g
a
bove
85%
,
a
nd
th
e
ove
r
a
ll
a
c
c
ur
a
c
y
of
th
e
m
od
e
l
s
how
s
a
v
a
lu
e
of
85%
.
F
ig
ur
e
3
di
s
pl
a
ys
a
c
onf
us
io
n
m
a
tr
ix
gr
a
ph,
w
hi
c
h
in
di
c
a
te
s
th
a
t
th
e
num
be
r
s
s
ho
w
in
g
c
onf
or
m
it
y
a
r
e
s
ig
ni
f
ic
a
nt
ly
m
or
e
th
a
n
th
e
numbe
r
s
s
how
in
g nonc
onf
or
m
it
y.
T
a
bl
e
1
. S
V
M
c
l
a
s
s
if
ic
a
ti
on r
e
por
t
S
e
n
t
i
m
e
n
t
l
a
b
e
l
P
r
e
c
i
s
i
o
n
R
e
c
a
l
l
F1
-
s
c
o
r
e
N
e
ga
t
i
ve
0.8
3
0.7
4
0.7
8
N
e
ut
r
a
l
0.8
3
0.8
7
0.8
5
P
os
i
t
i
ve
0.8
7
0.8
9
0.8
8
A
c
c
u
r
a
c
y
0.8
5
F
ig
ur
e
3. L
a
gs
on
gr
a
nge
r
c
a
us
a
li
ty
t
e
s
ts
2.3. Gr
an
ge
r
c
au
s
al
it
y an
al
ys
is
U
s
in
g
th
e
G
r
a
nge
r
c
a
us
a
li
ty
m
e
th
od,
th
e
a
ut
hor
a
na
ly
z
e
s
how
ne
w
s
s
e
nt
im
e
nt
in
f
lu
e
nc
e
s
e
x
c
ha
nge
r
a
te
s
a
nd
gol
d
pr
ic
e
s
a
nd
vi
c
e
ve
r
s
a
.
T
hi
s
te
c
hni
que
is
c
om
m
onl
y
a
ppl
ie
d
f
or
ti
m
e
s
e
r
ie
s
a
na
ly
s
is
in
va
r
io
us
s
c
ie
nt
if
ic
di
s
c
ip
li
ne
s
[
26]
.
G
r
a
nge
r
c
a
us
a
li
ty
te
s
ts
w
he
th
e
r
one
ti
m
e
s
e
r
ie
s
c
a
n
b
e
u
s
e
d
to
pr
e
di
c
t
a
not
he
r
ti
m
e
s
e
r
ie
s
[
27]
.
T
h
e
r
e
f
or
e
,
th
is
s
tu
dy
u
s
e
s
la
gs
1,
2,
a
nd
3 a
s
la
g
pe
r
io
ds
to
a
s
s
e
s
s
how
ne
w
s
s
e
nt
im
e
nt
a
f
f
e
c
ts
th
e
e
xc
ha
nge
r
a
te
a
nd
gol
d
pr
ic
e
.
L
a
gs
a
r
e
th
e
de
la
y
pe
r
io
d
in
th
e
obs
e
r
va
ti
on
th
a
t
is
in
f
lu
e
nc
e
d,
but
th
e
di
s
ta
nc
e
of
th
e
s
e
l
a
gs
c
a
nnot
b
e
th
e
s
a
m
e
f
or
di
f
f
e
r
e
nt
ti
m
e
s
e
r
ie
s
d
a
ta
[
26]
.
F
ig
ur
e
4
f
ur
th
e
r
il
lu
s
tr
a
te
s
s
om
e
of
th
e
la
gs
c
ons
id
e
r
e
d i
n t
hi
s
s
tu
dy.
F
ig
ur
e
4. L
a
gs
on
gr
a
nge
r
c
a
us
a
li
ty
t
e
s
ts
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
E
x
pl
or
in
g t
he
i
nf
lu
e
nc
e
of
s
of
t
in
fo
r
m
at
io
n f
r
om
e
c
onomic
ne
w
s
on
…
(
R
ahar
di
to
D
io
P
r
as
to
w
o
)
5235
2.4. Re
gr
e
s
s
io
n
an
al
ys
is
T
he
a
ut
hor
c
a
r
r
ie
d
out
f
ur
th
e
r
a
na
ly
s
is
r
e
ga
r
di
ng
th
e
a
bi
li
ty
of
ne
w
s
s
e
nt
im
e
nt
to
m
a
k
e
pr
e
di
c
ti
ons
th
r
ough
in
te
gr
a
te
d
m
ode
li
ng
w
it
h
e
xc
ha
nge
r
a
te
da
ta
to
pr
e
di
c
t
gol
d
pr
ic
e
s
,
a
s
pr
e
vi
ous
r
e
s
e
a
r
c
h
s
ta
te
d
th
a
t
s
of
t
in
f
or
m
a
ti
on
c
oul
d
im
p
r
ove
th
e
qua
li
ty
of
pr
e
di
c
ti
ons
f
r
o
m
m
ode
ls
c
r
e
a
te
d
us
in
g
ha
r
d
in
f
or
m
a
ti
on
[
10]
.
L
in
e
a
r
r
e
gr
e
s
s
io
n
a
na
ly
s
is
is
a
s
ta
ti
s
ti
c
a
l
a
na
ly
s
is
th
a
t
is
ge
ne
r
a
ll
y
us
e
d
f
or
pr
e
di
c
ti
ve
a
na
ly
s
is
a
nd
to
a
na
ly
z
e
th
e
r
e
la
ti
ons
hi
p
be
twe
e
n
th
e
de
pe
nde
nt
va
r
ia
bl
e
a
nd
one
or
m
or
e
in
de
pe
nde
nt
va
r
ia
bl
e
s
[
28]
.
B
y
in
t
e
gr
a
ti
ng
ne
w
s
s
e
nt
im
e
nt
w
it
h
e
xc
ha
nge
r
a
te
s
,
th
e
a
ut
hor
us
e
s
a
m
ul
ti
pl
e
li
ne
a
r
r
e
gr
e
s
s
io
n
m
ode
l
w
he
r
e
th
e
r
e
a
r
e
two
or
m
or
e
i
nde
pe
nde
nt
va
r
ia
bl
e
s
[
29]
.
3.
R
E
S
U
L
T
S
A
N
D
D
I
S
C
U
S
S
I
O
N
T
he
a
ut
hor
s
be
ga
n t
he
da
ta
a
na
ly
s
i
s
pr
oc
e
s
s
by obs
e
r
vi
ng t
r
e
nds
i
n t
he
ne
w
s
s
e
nt
im
e
nt
da
ta
t
ha
t
ha
d
be
e
n
ge
ne
r
a
te
d.
T
he
d
a
ta
is
di
s
pl
a
ye
d
a
s
a
li
ne
gr
a
ph
in
F
ig
ur
e
5.
T
hi
s
f
ig
ur
e
doe
s
not
in
di
c
a
t
e
a
dow
nw
a
r
d
or
upw
a
r
d
tr
e
nd
f
o
r
ne
ga
ti
ve
a
nd
pos
it
iv
e
s
e
nt
im
e
nt
th
r
oughout
th
e
pe
r
io
d
us
e
d.
H
ow
e
ve
r
,
th
e
r
e
w
a
s
a
n
in
c
r
e
a
s
e
im
m
e
di
a
te
ly
f
ol
lo
w
e
d
by
a
de
c
r
e
a
s
e
in
th
e
ne
w
s
w
it
h
pos
it
iv
e
a
nd
ne
ga
ti
ve
s
e
nt
im
e
nt
in
m
id
-
2020
a
nd e
a
r
ly
2021.
F
ig
ur
e
5. D
a
il
y
pos
it
iv
e
a
nd ne
ga
ti
ve
ne
w
s
s
e
nt
im
e
n
t
3.1. Cau
s
al
it
y
b
e
t
w
e
e
n
n
e
w
s
s
e
n
t
im
e
n
t
an
d
I
D
R
e
xc
h
an
ge
r
at
e
B
a
s
e
d
on
th
e
G
r
a
nge
r
c
a
us
a
li
ty
te
s
t
th
a
t
ha
s
be
e
n
done
,
th
e
p
-
va
lu
e
s
how
s
a
s
ta
ti
s
ti
c
a
ll
y
s
ig
ni
f
ic
a
nt
"
c
a
us
a
l"
r
e
la
ti
ons
hi
p
if
th
e
p
-
va
lu
e
<
0.05,
w
hi
c
h
m
e
a
n
s
ne
w
s
s
e
nt
im
e
nt
c
a
u
s
e
s
c
ha
nge
s
in
I
D
R
e
x
c
ha
nge
r
a
te
.
I
n
c
ont
r
a
s
t,
th
e
p
-
va
lu
e
in
di
c
a
te
s
a
"
non
-
c
a
us
a
l"
r
e
la
ti
ons
hi
p
if
th
e
p
-
va
lu
e
is
>
0.05,
w
hi
c
h
m
e
a
ns
th
e
ne
w
s
s
e
nt
im
e
nt
doe
s
not
c
a
us
e
c
ha
ng
e
s
in
I
D
R
e
xc
ha
nge
r
a
te
.
T
a
bl
e
2
s
how
s
th
e
r
e
s
ul
ts
of
th
e
G
r
a
nge
r
c
a
us
a
li
ty
te
s
t
be
twe
e
n
n
e
w
s
s
e
nt
im
e
nt
a
nd
I
D
R
e
xc
ha
nge
r
a
t
e
us
in
g
la
gs
of
1,
2,
a
nd
3
da
ys
.
A
p
a
r
t
f
r
om
th
a
t,
th
e
di
r
e
c
ti
on of
t
he
a
r
r
ow
i
s
us
e
d t
o s
how
t
he
di
r
e
c
ti
on of
c
a
us
a
li
ty
of
t
he
t
e
s
t
c
a
r
r
ie
d out.
T
he
"
V
a
r
ia
bl
e
s
"
c
ol
um
n
c
ons
is
ts
of
th
e
di
r
e
c
ti
on
of
th
e
c
a
us
a
li
ty
te
s
t
be
twe
e
n
ne
w
s
s
e
nt
im
e
nt
a
nd
I
D
R
e
xc
ha
nge
r
a
te
,
a
nd
th
e
"
L
a
g"
c
ol
um
n
c
ons
is
ts
of
3
va
lu
e
s
of
la
g
th
a
t
ha
ve
be
e
n
us
e
d
in
th
is
s
tu
dy.
T
he
"
p
-
va
lu
e
"
c
ol
um
n
c
ons
is
ts
of
c
a
us
a
li
ty
te
s
t
r
e
s
ul
ts
,
w
he
r
e
ba
s
e
d
on
th
e
r
e
s
ul
ts
in
th
e
T
a
bl
e
2
,
it
c
a
n
be
s
a
id
th
a
t
5
r
e
la
ti
ons
hi
ps
s
how
p
-
va
lu
e
<
0.05
(
H
2,
H
5,
H
6,
H
9,
H
10)
.
T
hi
s
in
di
c
a
te
s
th
a
t
ne
ga
ti
ve
ne
w
s
s
e
nt
im
e
nt
s
how
s
a
c
a
u
s
a
l
r
e
la
ti
ons
hi
p
to
th
e
I
D
R
e
xc
h
a
nge
r
a
te
a
nd
vi
c
e
v
e
r
s
a
,
e
ve
n
th
ough
th
e
r
e
w
a
s
a
non
-
c
a
us
a
l
r
e
la
ti
ons
hi
p
a
t
la
g
1
(
H
1)
.
I
n
c
ont
r
a
s
t,
th
e
pos
it
iv
e
n
e
w
s
e
nt
im
e
nt
s
how
s
a
non
-
c
a
u
s
a
l
r
e
la
ti
ons
hi
p
to
t
he
I
D
R
e
xc
ha
nge
r
a
te
a
nd vi
c
e
ve
r
s
a
w
it
h p
-
va
lu
e
>
0.05 (
H
3, H
4, H
7, H
8, H
11, H12)
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
,
V
ol
. 14, No. 6, D
e
c
e
m
be
r
2025
:
5231
-
5239
5236
T
a
bl
e
2
. R
e
s
ul
ts
f
or
gr
a
nge
r
c
a
us
a
li
ty
t
e
s
t
be
twe
e
n n
e
w
s
s
e
nt
im
e
nt
w
it
h
I
D
R
e
xc
ha
nge
r
a
te
C
ode
V
a
r
i
a
b
l
e
s
L
a
g
p
-
va
l
ue
C
a
us
a
l
i
t
y
H1
(
-
)
S
e
nt
i
m
e
nt
–
I
D
R
R
a
t
e
1
0.0
595
non
-
c
a
us
a
l
H2
I
D
R
R
a
t
e
–
(
-
)
S
e
n
t
i
m
e
n
t
1
1.5
9e
-
05
c
a
us
a
l
H3
(
+)
S
e
nt
i
m
e
nt
–
I
D
R
R
a
t
e
1
0.1
538
non
-
c
a
us
a
l
H4
I
D
R
R
a
t
e
–
(
+)
S
e
nt
i
m
e
nt
1
0.0
912
non
-
c
a
us
a
l
H5
(
-
)
S
e
nt
i
m
e
nt
–
I
D
R
R
a
t
e
2
0.0
015
c
a
us
a
l
H6
I
D
R
R
a
t
e
–
(
-
)
S
e
n
t
i
m
e
n
t
2
0.0
014
c
a
us
a
l
H7
(
+)
S
e
nt
i
m
e
nt
–
I
D
R
R
a
t
e
2
0.0
511
non
-
c
a
us
a
l
H8
I
D
R
R
a
t
e
–
(
+)
S
e
nt
i
m
e
nt
2
0.3
688
non
-
c
a
us
a
l
H9
(
-
)
S
e
nt
i
m
e
nt
–
I
D
R
R
a
t
e
3
0.0
041
c
a
us
a
l
H
10
I
D
R
R
a
t
e
–
(
-
)
S
e
n
t
i
m
e
n
t
3
0.0
142
c
a
us
a
l
H
11
(
+)
S
e
nt
i
m
e
nt
–
I
D
R
R
a
t
e
3
0.0
836
non
-
c
a
us
a
l
H
12
I
D
R
R
a
t
e
–
(
+)
S
e
nt
i
m
e
nt
3
0.5
518
non
-
c
a
us
a
l
3.2. Cau
s
al
it
y b
e
t
w
e
e
n
n
e
w
s
s
e
n
t
im
e
n
t
an
d
gol
d
p
r
ic
e
s
T
a
bl
e
3
s
how
s
th
e
r
e
s
ul
t
s
of
th
e
G
r
a
nge
r
c
a
us
a
li
ty
te
s
t
c
a
r
r
ie
d
out
by
th
e
a
ut
hor
s
in
a
na
ly
z
in
g
th
e
c
a
us
a
li
ty
r
e
la
ti
ons
hi
p
be
twe
e
n
ne
w
s
s
e
nt
im
e
nt
a
nd
gol
d
pr
ic
e
.
B
a
s
e
d
on
th
e
r
e
s
ul
ts
,
it
c
a
n
be
s
a
id
th
a
t
th
e
r
e
w
e
r
e
5
r
e
la
ti
ons
hi
ps
th
a
t
s
how
p
-
va
lu
e
<
0.05
(
H
13,
H
15,
H
17,
H
19,
H
21)
.
T
hi
s
in
di
c
a
te
s
th
a
t
ne
w
s
s
e
nt
im
e
nt
s
how
s
a
c
a
us
a
l
r
e
la
ti
ons
hi
p
to
th
e
gol
d
pr
ic
e
,
bot
h
ne
ga
ti
ve
a
nd
pos
it
iv
e
s
e
nt
im
e
nt
,
e
ve
n
th
ough
th
e
r
e
w
a
s
a
non
-
c
a
u
s
a
l
r
e
la
ti
ons
hi
p
a
t
L
a
g
3
(
H
23)
,
w
hi
le
th
e
gol
d
pr
ic
e
s
how
s
non
-
c
a
us
a
l
r
e
la
ti
ons
hi
p
to
th
e
ne
w
s
s
e
nt
im
e
nt
w
it
h p
-
va
lu
e
>
0.05 (
H
14, H16, H
18, H20, H
22
, H
24)
, bot
h ne
ga
ti
ve
a
nd pos
it
iv
e
s
e
nt
im
e
nt
.
3.3. Re
gr
e
s
s
io
n
an
al
ys
is
b
e
t
w
e
e
n
n
e
w
s
s
e
n
t
im
e
n
t
an
d
I
D
R
e
xc
h
an
ge
r
at
e
s
w
it
h
t
h
e
gol
d
p
r
ic
e
I
n
th
e
ne
xt
a
na
ly
s
i
s
pr
oc
e
s
s
,
th
e
a
ut
hor
s
c
a
r
r
ie
d
out
s
ta
ti
s
ti
c
a
l
a
na
ly
s
is
us
in
g
m
ul
ti
pl
e
li
ne
a
r
r
e
gr
e
s
s
io
n
to
te
s
t
th
e
r
e
la
ti
ons
hi
p
be
twe
e
n
one
de
pe
nde
nt
va
r
ia
bl
e
a
nd
s
e
ve
r
a
l
in
de
pe
nde
nt
va
r
ia
bl
e
s
.
I
n
th
is
s
tu
dy, t
he
de
pe
nde
nt
va
r
ia
bl
e
i
s
t
he
gol
d pr
ic
e
, w
hi
le
t
he
i
nde
pe
nde
nt
va
r
ia
bl
e
s
c
ons
is
t
of
ne
w
s
s
e
nt
im
e
nt
a
nd
th
e
I
D
R
e
xc
ha
nge
r
a
te
.
T
a
bl
e
4
s
how
s
t
he
r
e
s
ul
ts
of
t
e
s
ti
ng w
it
h m
ul
ti
pl
e
l
in
e
a
r
r
e
gr
e
s
s
io
n.
T
a
bl
e
3
. R
e
s
ul
ts
f
or
gr
a
nge
r
c
a
us
a
li
ty
t
e
s
t
be
twe
e
n n
e
w
s
s
e
nt
im
e
nt
w
it
h
gol
d
pr
ic
e
C
ode
V
a
r
i
a
b
l
e
s
L
a
g
p
-
va
l
ue
C
a
us
a
l
i
t
y
H
13
(
-
)
S
e
nt
i
m
e
nt
–
G
ol
d
P
r
i
c
e
1
1.2
3e
-
04
c
a
us
a
l
H
14
G
ol
d
P
r
i
c
e
–
(
-
)
S
e
n
t
i
m
e
nt
1
0.5
013
non
-
c
a
us
a
l
H
15
(
+)
S
e
nt
i
m
e
nt
–
G
ol
d
P
r
i
c
e
1
2.2
5e
-
03
c
a
us
a
l
H
16
G
ol
d
P
r
i
c
e
–
(
+)
S
e
nt
i
m
e
n
t
1
0.5
649
non
-
c
a
us
a
l
H
17
(
-
)
S
e
nt
i
m
e
nt
–
G
ol
d
P
r
i
c
e
2
0.0
003
c
a
us
a
l
H
18
G
ol
d
P
r
i
c
e
–
(
-
)
S
e
n
t
i
m
e
nt
2
0.7
318
non
-
c
a
us
a
l
H
19
(
+)
S
e
nt
i
m
e
nt
–
G
ol
d
P
r
i
c
e
2
0.0
160
c
a
us
a
l
H
20
G
ol
d
P
r
i
c
e
–
(
+)
S
e
nt
i
m
e
n
t
2
0.7
222
non
-
c
a
us
a
l
H
21
(
-
)
S
e
nt
i
m
e
nt
–
G
ol
d
P
r
i
c
e
3
0.0
020
c
a
us
a
l
H
22
G
ol
d
P
r
i
c
e
–
(
-
)
S
e
n
t
i
m
e
nt
3
0.6
296
non
-
c
a
us
a
l
H
23
(
+)
S
e
nt
i
m
e
nt
–
G
ol
d
P
r
i
c
e
3
0.0
961
non
-
c
a
us
a
l
H
24
G
ol
d
P
r
i
c
e
–
(
+)
S
e
nt
i
m
e
n
t
3
0.7
219
non
-
c
a
us
a
l
T
a
b
l
e
4
.
R
e
s
ul
t
s
f
o
r
r
e
gr
e
s
s
io
n
a
n
a
l
y
s
i
s
b
e
tw
e
e
n
n
e
w
s
s
e
nt
im
e
nt
a
n
d
I
D
R
e
x
c
h
a
ng
e
r
a
t
e
o
n
g
ol
d
pr
i
c
e
m
o
v
e
m
e
nt
s
E
s
t
i
m
a
t
e
S
t
d.
e
r
r
o
r
t
va
l
ue
P
r
(
>|
t
|
)
N
e
ga
t
i
ve
s
e
nt
i
m
e
nt
&
I
D
R
e
xc
ha
nge
r
a
t
e
(
I
nt
e
r
c
e
pt
)
365.788690
97.810362
3.740
0.000194
***
s
e
nt
_ne
ga
t
i
ve
-
21.215032
4.321684
-
4.909
1.06e
-
06
***
I
D
R
0.101696
0.006585
15.444
<
2e
-
16
***
A
dj
.
R
-
s
qua
r
e
d
0.1977
p
-
va
l
ue
<
2.2e
-
16
P
os
i
t
i
ve
s
e
nt
i
m
e
nt
&
I
D
R
e
x
c
ha
nge
r
a
t
e
(
I
nt
e
r
c
e
pt
)
358.554030
98.309433
3.647
0.000278
***
s
e
nt
_ne
ga
t
i
ve
-
15.053169
4.024521
-
3.740
0.000194
***
I
D
R
0.102149
0.006616
15.441
<
2e
-
16
***
A
dj
.
R
-
s
qua
r
e
d
0.1977
p
-
va
l
ue
<
2.2e
-
16
T
he
f
ir
s
t
pa
r
t
of
T
a
bl
e
4
di
s
pl
a
ys
t
he
r
e
gr
e
s
s
io
n a
na
ly
s
is
r
e
s
ul
ts
w
it
h ne
ga
ti
ve
s
e
nt
im
e
nt
a
nd t
he
I
D
R
e
xc
ha
nge
r
a
te
a
s
in
de
pe
nde
nt
va
r
ia
bl
e
s
a
nd
th
e
gol
d
pr
ic
e
a
s
th
e
de
pe
nde
nt
va
r
ia
bl
e
w
it
h
p
-
va
lu
e
<
0.05,
in
di
c
a
ti
ng
th
a
t
th
e
ne
ga
ti
ve
s
e
nt
im
e
nt
a
nd
th
e
I
D
R
e
xc
ha
nge
r
a
te
s
ig
ni
f
ic
a
nt
ly
a
f
f
e
c
t
th
e
m
ode
l.
T
h
e
s
e
c
ond
pa
r
t
of
T
a
bl
e
4
di
s
pl
a
ys
th
e
r
e
gr
e
s
s
io
n
a
na
ly
s
i
s
r
e
s
ul
t
s
w
it
h
po
s
it
iv
e
s
e
nt
im
e
nt
a
nd
th
e
I
D
R
e
xc
ha
nge
r
a
te
a
s
in
de
pe
nde
nt
va
r
ia
bl
e
s
a
nd
th
e
gol
d
pr
ic
e
a
s
th
e
d
e
pe
nde
nt
va
r
i
a
bl
e
w
it
h
a
p
-
va
lu
e
<
0.05,
a
ls
o
in
di
c
a
ti
ng
th
a
t
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
E
x
pl
or
in
g t
he
i
nf
lu
e
nc
e
of
s
of
t
in
fo
r
m
at
io
n f
r
om
e
c
onomic
ne
w
s
on
…
(
R
ahar
di
to
D
io
P
r
as
to
w
o
)
5237
pos
it
iv
e
s
e
nt
im
e
nt
a
nd
th
e
I
D
R
e
xc
ha
ng
e
r
a
te
s
ig
ni
f
ic
a
nt
ly
a
f
f
e
c
t
th
e
m
od
e
l.
M
e
a
nw
hi
le
,
th
e
a
dj
u
s
te
d
R
-
s
qua
r
e
d va
lu
e
of
0.1977 s
how
s
t
ha
t
th
e
m
ode
l
c
a
n
e
xpl
a
in
19
%
of
t
he
va
r
ia
ti
on i
n t
he
de
pe
nde
nt
va
r
ia
bl
e
.
4.
C
O
N
C
L
U
S
I
O
N
T
hi
s
s
tu
dy
a
im
e
d
to
de
te
r
m
in
e
th
e
e
f
f
e
c
t
of
s
of
t
in
f
o
r
m
a
ti
on
f
r
om
e
c
onomi
c
ne
w
s
s
e
nt
im
e
nt
on
th
e
I
D
R
e
xc
ha
nge
r
a
te
a
nd
gol
d
pr
ic
e
f
r
om
J
a
nua
r
y
1,
2020
to
A
pr
il
30,
2024.
F
ir
s
t,
th
e
a
ut
hor
s
c
a
r
r
ie
d
out
s
e
nt
im
e
nt
a
na
ly
s
is
u
s
in
g t
he
V
A
D
E
R
l
e
xi
c
on me
th
od a
nd va
li
d
a
te
d i
t
us
in
g t
he
S
V
M
m
ode
l.
I
n a
ddi
ti
on, da
ta
r
e
la
te
d
to
th
e
I
D
R
e
xc
ha
nge
r
a
te
a
nd
gol
d
pr
ic
e
w
e
r
e
obt
a
in
e
d
f
r
om
I
nve
s
ti
ng.c
om
.
T
he
G
r
a
nge
r
c
a
us
a
li
ty
te
s
t
r
e
ve
a
le
d
th
a
t
ne
g
a
ti
ve
ne
w
s
s
e
nt
im
e
nt
s
ig
ni
f
ic
a
nt
ly
a
f
f
e
c
ts
th
e
I
D
R
e
x
c
ha
nge
r
a
te
,
w
hi
le
po
s
it
iv
e
s
e
nt
im
e
nt
s
how
s
no
s
ig
ni
f
ic
a
nt
in
f
lu
e
nc
e
.
I
n
c
ont
r
a
s
t,
bot
h
pos
it
iv
e
a
nd
ne
ga
ti
ve
s
e
nt
im
e
nt
s
im
pa
c
t
gol
d
pr
ic
e
s
,
but
th
e
gol
d
pr
ic
e
do
e
s
not
s
ig
ni
f
ic
a
nt
ly
a
f
f
e
c
t
ne
w
s
s
e
nt
im
e
nt
.
T
he
s
e
f
in
di
ngs
in
di
c
a
te
th
a
t
ne
ga
ti
ve
s
e
nt
im
e
nt
pl
a
ys
a
s
tr
onge
r
r
ol
e
in
c
ur
r
e
nc
y
f
lu
c
tu
a
ti
ons
,
w
h
il
e
gol
d
pr
ic
e
s
r
e
s
pond
to
ove
r
a
ll
s
e
nt
im
e
nt
tr
e
nds
.
A
ddi
ti
ona
ll
y,
m
ul
ti
pl
e
li
ne
a
r
r
e
gr
e
s
s
io
n
a
na
ly
s
is
c
onf
ir
m
e
d
th
a
t
e
c
onomi
c
ne
w
s
s
e
nt
im
e
nt
a
nd
th
e
I
D
R
e
xc
ha
nge
r
a
te
s
ig
ni
f
ic
a
nt
ly
in
f
lu
e
nc
e
gol
d
pr
ic
e
s
.
T
hi
s
s
u
gge
s
ts
th
a
t
bot
h
f
a
c
to
r
s
c
a
n
be
in
te
gr
a
te
d
in
to
pr
e
di
c
ti
ve
m
ode
ls
f
or
go
ld
pr
ic
e
f
or
e
c
a
s
ti
ng.
F
ur
th
e
r
r
e
s
e
a
r
c
h
c
a
n
e
xpl
or
e
th
e
f
in
d
in
gs
f
r
om
th
is
s
tu
dy
,
e
s
pe
c
ia
ll
y
r
e
g
a
r
di
ng
th
e
unde
r
ly
in
g
r
e
a
s
ons
f
or
th
e
di
f
f
e
r
e
nc
e
be
twe
e
n
ne
ga
ti
ve
a
nd
po
s
it
iv
e
s
e
nt
im
e
nt
in
f
lu
e
nc
e
on
th
e
e
xc
ha
nge
r
a
te
or
a
not
he
r
e
c
onomi
c
in
di
c
a
to
r
.
O
th
e
r
f
in
di
ngs
r
e
la
te
d
to
th
e
m
ul
t
ip
le
li
ne
a
r
r
e
gr
e
s
s
io
n
m
ode
l
f
or
p
r
e
di
c
ti
ng
gol
d
p
r
ic
e
s
ha
ve
th
e
pot
e
nt
ia
l
f
or
f
u
r
th
e
r
r
e
s
e
a
r
c
h
r
e
ga
r
di
ng
ot
he
r
e
c
onomi
c
in
di
c
a
to
r
s
t
ha
t
c
a
n be
us
e
d t
o i
m
pr
ove
t
he
i
nde
pe
nde
nt
va
r
ia
bl
e
s
'
a
bi
li
ty
t
o e
xpl
a
in
t
he
de
pe
nde
nt
va
r
ia
bl
e
.
A
C
K
N
O
WL
E
D
G
M
E
N
T
S
T
hi
s
s
tu
dy
w
a
s
c
onduc
te
d
unde
r
th
e
a
us
pi
c
e
s
of
th
e
U
ni
ve
r
s
it
y
of
I
ndone
s
ia
,
a
nd
w
e
a
r
e
e
s
p
e
c
ia
ll
y
th
a
nkf
ul
f
or
t
he
a
c
a
de
m
ic
s
uppor
t
a
nd f
a
c
il
it
ie
s
pr
ovi
de
d.
F
U
N
D
I
N
G
I
N
F
O
R
M
A
T
I
O
N
T
hi
s
s
tu
dy
w
a
s
c
a
r
r
ie
d
out
in
d
e
pe
nde
nt
ly
w
it
hout
a
ny
f
in
a
nc
ia
l
s
uppor
t
f
r
om
e
xt
e
r
na
l
in
s
ti
tu
ti
on
or
f
undi
ng body.
A
U
T
H
O
R
C
O
N
T
R
I
B
U
T
I
O
N
S
S
T
A
T
E
M
E
N
T
T
o
pr
om
ot
e
tr
a
ns
pa
r
e
nc
y
a
nd
pr
ope
r
a
tt
r
ib
ut
io
n
o
f
w
or
k,
th
is
s
tu
dy
us
e
s
th
e
C
ont
r
ib
ut
or
R
ol
e
s
T
a
xonomy
(
C
R
e
di
T
)
t
o s
pe
c
if
y t
he
c
ont
r
ib
ut
io
n of
e
a
c
h a
ut
hor
.
N
am
e
o
f
A
u
t
h
or
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
R
a
ha
r
di
to
D
io
P
r
a
s
to
w
o
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
I
ndr
a
B
udi
✓
✓
✓
✓
✓
✓
✓
A
m
a
na
h R
a
m
a
di
a
h
✓
✓
✓
✓
✓
✓
✓
A
r
is
B
udi
S
a
nt
os
o
✓
✓
✓
✓
✓
✓
P
r
a
bu K
r
e
s
na
P
ut
r
a
✓
✓
✓
✓
✓
✓
C
:
C
onc
e
pt
ua
l
i
z
a
t
i
on
M
:
M
e
t
hodol
ogy
So
:
So
f
t
w
a
r
e
Va
:
Va
l
i
da
t
i
on
Fo
:
Fo
r
m
a
l
a
na
l
ys
i
s
I
:
I
nve
s
t
i
ga
t
i
on
R
:
R
e
s
our
c
e
s
D
:
D
a
t
a
C
ur
a
t
i
on
O
:
W
r
i
t
i
ng
-
O
r
i
gi
na
l
D
r
a
f
t
E
:
W
r
i
t
i
ng
-
R
e
vi
e
w
&
E
di
t
i
ng
Vi
:
Vi
s
ua
l
i
z
a
t
i
on
Su
:
Su
pe
r
vi
s
i
on
P
:
P
r
oj
e
c
t
a
dm
i
ni
s
t
r
a
t
i
on
Fu
:
Fu
ndi
ng a
c
qui
s
i
t
i
on
C
O
N
F
L
I
C
T
O
F
I
N
T
E
R
E
S
T
S
T
A
T
E
M
E
N
T
T
he
a
ut
hor
s
de
c
la
r
e
th
a
t
th
is
s
tu
dy
w
a
s
c
a
r
r
ie
d
out
w
it
hout
a
ny c
om
m
e
r
c
ia
l
or
f
in
a
nc
ia
l
in
vol
ve
m
e
nt
th
a
t
c
oul
d be
i
nt
e
r
pr
e
te
d a
s
a
pot
e
nt
ia
l
c
onf
li
c
t
of
i
nt
e
r
e
s
t.
I
N
F
O
R
M
E
D
C
O
N
S
E
N
T
T
hi
s
s
tu
d
y
di
d
n
ot
i
nvo
lv
e
d
ir
e
c
t
i
nt
e
r
a
c
ti
on
w
i
th
hu
m
a
n
pa
r
ti
c
ip
a
nt
s
o
r
d
a
ta
ob
ta
in
e
d
f
r
om
a
n
y
s
pe
c
i
f
i
c
i
ns
t
it
u
ti
on
.
O
n
ly
pu
bl
ic
l
y
a
va
il
a
b
le
s
e
c
o
nda
r
y
da
ta
w
e
r
e
us
e
d
.
T
he
r
e
f
o
r
e
,
in
f
o
r
m
e
d
c
ons
e
nt
w
a
s
not
r
e
qu
i
r
e
d
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
,
V
ol
. 14, No. 6, D
e
c
e
m
be
r
2025
:
5231
-
5239
5238
E
T
H
I
C
A
L
A
P
P
R
O
V
A
L
A
s
th
is
s
tu
dy
di
d
not
in
vol
ve
hum
a
n
s
ubj
e
c
ts
,
in
s
ti
tu
ti
ona
l
da
ta
,
or
a
ny
c
onf
id
e
nt
ia
l
or
pr
opr
ie
ta
r
y
in
f
or
m
a
ti
on, e
th
ic
a
l
a
ppr
ova
l
w
a
s
not
a
ppl
ic
a
bl
e
.
D
A
T
A
A
V
A
I
L
A
B
I
L
I
T
Y
T
he
d
a
ta
th
a
t
s
uppor
t
th
e
f
in
di
ngs
of
th
is
s
tu
dy
a
r
e
op
e
nl
y
a
va
il
a
bl
e
in
D
e
ti
kF
in
a
nc
e
a
t
ht
tp
s
:/
/f
in
a
nc
e
.de
ti
k.c
om
a
nd I
nve
s
ti
ng.c
om
a
t
ht
tp
s
:/
/ww
w
.i
nv
e
s
ti
ng.c
om
.
R
E
F
E
R
E
N
C
E
S
[
1]
A
. H
. S
ha
pi
r
o, M
.
S
udhof
, a
nd D
. J
. W
i
l
s
on, “
M
e
a
s
ur
i
ng n
e
w
s
s
e
nt
i
m
e
nt
,”
J
ou
r
nal
of
E
c
onom
e
t
r
i
c
s
, vol
. 228,
no. 2, pp.
221
–
243,
J
un. 2022, doi
:
10.1016/
j
.j
e
c
onom
.2020.07.053.
[
2]
M
.
F
.
H
s
u,
T
.
M
.
C
ha
ng,
a
nd
S
.
J
.
L
i
n,
“
N
e
w
s
-
ba
s
e
d
s
of
t
i
nf
or
m
a
t
i
on
a
s
a
c
or
por
a
t
e
c
om
pe
t
i
t
i
ve
a
dva
nt
a
ge
,”
T
e
c
hnol
ogi
c
al
and
E
c
onom
i
c
D
e
v
e
l
opm
e
nt
of
E
c
onom
y
, vol
. 26, no. 1, pp. 48
–
70, J
a
n. 2020, doi
:
10.3846/
t
e
de
.2019.11328.
[
3]
J
.
M
.
L
i
be
r
t
i
a
nd
M
.
A
.
P
e
t
e
r
s
e
n,
“
I
nf
or
m
a
t
i
on:
H
a
r
d
a
nd
s
of
t
,”
R
e
v
i
e
w
of
C
or
por
at
e
F
i
nanc
e
St
udi
e
s
,
vol
.
8,
no.
1,
pp.
1
–
41,
M
a
r
. 2019, doi
:
10.1093/
r
c
f
s
/
c
f
y009.
[
4]
S
.
E
s
t
r
i
n,
S
.
K
h
a
vu
l
,
a
nd
M
.
W
r
i
gh
t
,
“
S
o
f
t
a
nd
h
a
r
d
i
nf
or
m
a
t
i
on
i
n
e
qu
i
t
y
c
r
o
w
df
un
di
ng:
ne
t
w
o
r
k
e
f
f
e
c
t
s
i
n
t
he
d
i
g
i
t
a
l
i
z
a
t
i
o
n
o
f
e
nt
r
e
p
r
e
ne
u
r
i
a
l
f
i
n
a
nc
e
,”
Sm
a
l
l
B
us
i
ne
s
s
E
c
on
om
i
c
s
,
vo
l
.
58
,
no.
4
, p
p.
17
61
–
17
8
1, A
p
r
.
20
22,
d
oi
:
1
0.
100
7/
s
11
187
-
0
21
-
0
04
73
-
w.
[
5]
D
.
T
s
ur
ut
a
,
“
C
a
n
ba
nks
m
oni
t
or
s
m
a
l
l
bus
i
ne
s
s
bor
r
ow
e
r
s
e
f
f
e
c
t
i
ve
l
y
us
i
ng
ha
r
d
i
nf
or
m
a
t
i
on
?
,”
A
c
c
ount
i
ng
and
F
i
nanc
e
,
vol
.
60,
no. 4, pp. 4291
–
4330, D
e
c
. 2020, doi
:
10.1111/
a
c
f
i
.12544.
[
6]
S
.
N
.
A
l
i
,
N
.
H
a
ghpa
na
h,
X
.
L
i
n,
a
nd
R
.
S
i
e
ge
l
,
“
H
ow
t
o
s
e
l
l
ha
r
d
i
nf
or
m
a
t
i
o
n,”
T
he
Q
uar
t
e
r
l
y
J
ou
r
nal
of
E
c
onom
i
c
s
,
vol
.
137,
no. 1,
pp. 619
–
678, 2022, doi
:
10.1093/
qj
e
/
qj
a
b024.
[
7]
H
.
L
i
,
“
E
m
be
dde
d
m
i
c
r
opr
oc
e
s
s
or
w
i
r
e
l
e
s
s
c
om
m
uni
c
a
t
i
on
da
t
a
c
ol
l
e
c
t
i
on
a
i
ds
i
n
e
a
r
l
y
w
a
r
ni
ng
of
de
f
a
ul
t
r
i
s
k
f
or
i
nt
e
r
ne
t
f
i
na
nc
e
ba
nk c
us
t
om
e
r
s
,”
J
our
nal
of
Se
ns
o
r
s
,
vol
. 2021, pp. 1
-
10,
2021, doi
:
10
.1155/
2021/
1679907.
[
8]
M
.
C
a
r
on
a
nd
O
.
M
ul
l
e
r
,
“
H
a
r
de
ni
ng
S
o
f
t
I
nf
or
m
a
t
i
on:
A
T
r
a
ns
f
or
m
e
r
-
B
a
s
e
d
A
ppr
oa
c
h
t
o
F
or
e
c
a
s
t
i
ng
S
t
oc
k
R
e
t
ur
n
V
ol
a
t
i
l
i
t
y,”
in
2020 I
E
E
E
I
nt
e
r
nat
i
onal
C
onf
e
r
e
nc
e
on B
i
g D
at
a
, D
e
c
. 2020, pp. 4383
–
439
1
, doi
:
10.1109/
B
i
gD
a
t
a
50022.2020.9378134.
[
9]
T
.
S
he
ng,
“
T
he
e
f
f
e
c
t
of
f
i
nt
e
c
h
on
ba
nks
’
c
r
e
di
t
pr
ovi
s
i
on
t
o
S
M
E
s
:
E
vi
de
nc
e
f
r
om
C
hi
na
,”
F
i
nanc
e
R
e
s
e
ar
c
h
L
e
t
t
e
r
s
,
vol
.
39
,
M
a
r
. 2021, doi
:
10.1016/
j
.f
r
l
.2020.101558.
[
10]
S
.
F
i
l
o
m
e
n
i
, U
. B
os
e
, A
. M
e
ga
r
i
t
i
s
, a
nd A
.
T
r
i
a
n
t
a
f
y
l
l
ou
, “
C
a
n
m
a
r
k
e
t
i
n
f
o
r
m
a
t
i
o
n ou
t
pe
r
f
o
r
m
ha
r
d
a
nd s
of
t
i
n
f
o
r
m
a
t
i
on
i
n
pr
e
d
i
c
t
i
n
g
c
or
po
r
a
t
e
de
f
a
u
l
t
s
?
,”
I
n
t
e
r
n
at
i
o
na
l
J
our
n
al
o
f
F
i
na
nc
e
an
d
E
c
o
nom
i
c
s
, v
ol
. 2
9,
no
. 3
, p
p.
35
67
–
35
92
, 2
02
4,
do
i
:
10
.1
002
/
i
j
f
e
.
28
40.
[
11]
Y
.
L
i
,
S
.
J
i
a
ng,
X
.
L
i
,
a
nd
S
.
W
a
ng,
“
T
he
r
ol
e
o
f
ne
w
s
s
e
nt
i
m
e
nt
i
n
oi
l
f
ut
ur
e
s
r
e
t
ur
ns
a
nd
vo
l
a
t
i
l
i
t
y
f
or
e
c
a
s
t
i
ng:
D
a
t
a
-
de
c
om
pos
i
t
i
on ba
s
e
d de
e
p l
e
a
r
ni
ng a
ppr
oa
c
h,”
E
ne
r
gy
E
c
onom
i
c
s
, vol
. 95, M
a
r
. 2021, doi
:
10.1016/
j
.e
ne
c
o.2021.105140.
[
12]
A
.
T
a
dpha
l
e
,
H
.
S
a
r
a
s
w
a
t
,
O
.
S
ona
w
a
ne
,
a
nd
P
.
R
.
D
e
s
hm
ukh,
“
I
m
pa
c
t
of
ne
w
s
s
e
nt
i
m
e
nt
on
f
or
e
i
gn
e
xc
ha
nge
r
a
t
e
pr
e
di
c
t
i
on
,”
i
n
2023
3r
d
I
nt
e
r
nat
i
onal
C
onf
e
r
e
nc
e
on
I
nt
e
l
l
i
ge
nt
T
e
c
hnol
ogi
e
s
(
C
O
N
I
T
)
,
H
ubl
i
,
I
ndi
a
,
2023,
pp.
1
-
8,
doi
:
10.1109/
C
O
N
I
T
59222.2023.10205534.
[
13]
Y
.
M
a
o,
Z
.
C
he
n,
S
.
L
i
u,
a
nd
Y
.
L
i
,
“
U
nve
i
l
i
ng
t
he
pot
e
nt
i
a
l
:
E
xpl
or
i
ng
t
he
pr
e
di
c
t
a
bi
l
i
t
y
of
c
om
pl
e
x
e
xc
ha
nge
r
a
t
e
t
r
e
nds
,
”
E
ngi
ne
e
r
i
ng A
ppl
i
c
at
i
ons
of
A
r
t
i
f
i
c
i
al
I
nt
e
l
l
i
ge
nc
e
, vol
. 133, M
a
r
. 2024, doi
:
10
.1016/
j
.e
nga
ppa
i
.2024.108112.
[
14]
M
.
L
uka
u
s
ka
s
,
V
.
P
i
l
i
nki
e
ne
,
J
.
B
r
une
c
ki
e
ne
,
A
.
S
t
undz
i
e
ne
,
A
.
G
r
yba
us
k
a
s
,
a
nd
T
.
R
uz
ga
s
,
“
E
c
onom
i
c
A
c
t
i
vi
t
y
f
or
e
c
a
s
t
i
ng
ba
s
e
d on t
he
s
e
nt
i
m
e
nt
a
na
l
ys
i
s
of
n
e
w
s
,”
M
at
he
m
at
i
c
s
, vol
. 10, no. 3461, S
e
p.
2022, doi
:
10.3390/
m
a
t
h10193461.
[
15]
L
.
X
ue
l
i
ng,
X
.
X
i
ong,
a
nd
S
.
Y
uc
ong,
“
E
xc
ha
nge
r
a
t
e
m
a
r
ke
t
t
r
e
nd
p
r
e
di
c
t
i
on
ba
s
e
d
on
s
e
nt
i
m
e
nt
a
na
l
ys
i
s
,”
C
om
put
e
r
s
and
E
l
e
c
t
r
i
c
al
E
ngi
ne
e
r
i
ng
, vol
. 111, O
c
t
. 2023, doi
:
10.1016/
j
.c
om
pe
l
e
c
e
ng.2023.
108901.
[
16]
Z
.
J
unj
i
e
a
nd
P
.
M
e
ngoni
,
“
S
pot
gol
d
pr
i
c
e
pr
e
di
c
t
i
on
us
i
ng
f
i
na
nc
i
a
l
ne
w
s
s
e
nt
i
m
e
nt
a
na
l
ys
i
s
,
”
i
n
2020
I
E
E
E
/
W
I
C
/
A
C
M
I
nt
e
r
nat
i
onal
J
oi
nt
C
onf
e
r
e
nc
e
on
W
e
b
I
nt
e
l
l
i
ge
nc
e
and
I
nt
e
l
l
i
ge
nt
A
ge
nt
T
e
c
hnol
ogy
,
D
e
c
.
2020,
pp.
758
–
763
,
doi
:
10.1109/
W
I
I
A
T
50758.2020.00117.
[
17]
Y
.
J
i
a
ng,
Y
.
S
.
R
e
n,
S
.
N
a
r
a
ya
n,
C
.
Q
.
M
a
,
a
nd
X
.
G
.
Y
a
ng,
“
H
e
t
e
r
oge
ne
i
t
y
de
pe
nde
nc
e
be
t
w
e
e
n
oi
l
pr
i
c
e
s
a
nd
e
xc
ha
nge
r
a
t
e
:
E
vi
de
nc
e
f
r
om
a
pa
r
a
m
e
t
r
i
c
t
e
s
t
of
G
r
a
nge
r
c
a
us
a
l
i
t
y
i
n
qua
nt
i
l
e
s
,”
N
or
t
h
A
m
e
r
i
c
an
J
our
nal
of
E
c
onom
i
c
s
and
F
i
nanc
e
,
vol
.
62,
N
ov. 2022, doi
:
10.1016/
j
.na
j
e
f
.2022.101711.
[
18]
R
.
G
a
ut
a
m
,
Y
.
K
i
m
,
E
.
T
opa
l
,
a
nd
M
.
H
i
t
c
h,
“
C
or
r
e
l
a
t
i
on
be
t
w
e
e
n
C
O
V
I
D
-
19
c
a
s
e
s
a
nd
gol
d
pr
i
c
e
f
l
uc
t
ua
t
i
on,”
I
nt
e
r
nat
i
onal
J
our
nal
of
M
i
ni
ng, R
e
c
l
am
at
i
on and E
nv
i
r
onm
e
nt
, vol
. 36, no. 8, pp. 574
–
586, 2022, doi
:
10.1080/
17480930.2022.2077542.
[
19]
M
.
A
bdou,
M
.
S
ha
l
t
out
,
A
.
G
oda
h,
K
.
S
obh,
Y
.
E
i
d,
a
nd
W
.
M
e
dha
t
,
“
G
ol
d
pr
i
c
e
pr
e
di
c
t
i
on
us
i
ng
s
e
nt
i
m
e
nt
a
na
l
ys
i
s
,”
in
20t
h
C
onf
e
r
e
nc
e
on L
anguage
E
ngi
ne
e
r
i
ng, E
SO
L
E
C
2022
, 2022, pp. 41
–
44
, doi
:
10.1109/
E
S
O
L
E
C
54569.2022.10009529.
[
20]
Y
.
J
i
a
nyi
,
W
.
C
he
nya
ng,
H
.
Y
upe
ng,
a
nd
L
.
Z
i
c
he
ng,
“
R
e
s
e
a
r
c
h
on
t
he
r
e
l
a
t
i
ons
hi
p
be
t
w
e
e
n
c
ovi
d
-
19
e
pi
de
m
i
c
a
nd
gol
d
pr
i
c
e
t
r
e
nd
ba
s
e
d
on
l
i
ne
a
r
r
e
gr
e
s
s
i
on
m
ode
l
,”
i
n
I
E
E
E
9t
h
J
oi
nt
I
nt
e
r
nat
i
onal
I
nf
or
m
at
i
on
T
e
c
hnol
ogy
and
A
r
t
i
f
i
c
i
al
I
nt
e
l
l
i
ge
nc
e
C
onf
e
r
e
nc
e
2020
, 2020, pp. 1796
–
1798
, doi
:
10.1109/
I
T
A
I
C
49862.2020.93388
28.
[
21]
H
.
B
h
o
i
r
a
n
d
K
.
J
a
y
a
m
a
l
i
n
i
, “
W
e
b
c
r
a
w
l
i
n
g
o
n
ne
w
s
w
e
b p
a
ge
us
i
ng
d
i
f
f
e
r
e
n
t
f
r
a
m
e
w
o
r
ks
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
ur
n
a
l
o
f
S
c
i
e
n
t
i
f
i
c
R
e
s
e
a
r
c
h
i
n
C
om
p
u
t
e
r
Sc
i
e
n
c
e
,
E
n
g
i
ne
e
r
i
n
g
a
n
d
I
n
f
o
r
m
a
t
i
o
n
T
e
c
hn
o
l
o
gy
,
vo
l
.
7
,
n
o
.
4
,
p
p.
5
13
–
5
1
9
,
A
u
g
.
2
0
2
1,
d
o
i
:
1
0
.
3
26
2
8
/
c
s
e
i
t
2
1
74
1
2
0
.
[
22]
A
.
B
or
g
a
nd
M
.
B
ol
dt
,
“
U
s
i
ng
V
A
D
E
R
s
e
nt
i
m
e
nt
a
nd
S
V
M
f
or
pr
e
di
c
t
i
ng
c
us
t
om
e
r
r
e
s
pons
e
s
e
nt
i
m
e
nt
,”
E
x
pe
r
t
Sy
s
t
e
m
s
w
i
t
h
A
ppl
i
c
at
i
ons
, vol
. 162, D
e
c
. 2020, doi
:
10.1016/
j
.e
s
w
a
.2020.113746.
[
23]
D
.
M
a
r
ut
ho,
M
ul
j
ono,
S
.
R
u
s
t
a
d,
a
nd
P
ur
w
a
nt
o,
“
S
e
nt
i
m
e
nt
a
na
l
ys
i
s
opt
i
m
i
z
a
t
i
on
us
i
ng
v
a
de
r
l
e
xi
c
on
on
m
a
c
hi
ne
l
e
a
r
ni
ng
a
ppr
oa
c
h,
”
i
n
2022
I
nt
e
r
nat
i
onal
Se
m
i
nar
on
I
nt
e
l
l
i
ge
nt
T
e
c
hnol
ogy
and
I
t
s
A
ppl
i
c
at
i
ons
:
A
dv
anc
e
d
I
nnov
at
i
ons
of
E
l
e
c
t
r
i
c
al
Sy
s
t
e
m
s
f
o
r
H
um
ani
t
y
, 2022, pp. 98
–
103
, doi
:
10.1109/
I
S
I
T
I
A
56226.2022.985
5341.
[
24]
S
. J
a
i
s
w
a
l
, S
. S
r
i
va
s
t
a
va
, S
. G
a
r
g,
a
nd P
. S
i
ngh, “
E
f
f
e
c
t
of
ne
w
s
h
e
a
dl
i
ne
s
on
g
ol
d pr
i
c
e
pr
e
di
c
t
i
on us
i
ng
N
L
P
a
nd
de
e
p
l
e
a
r
ni
ng,”
i
n
2023
I
nt
e
r
nat
i
onal
C
onf
e
r
e
nc
e
on
A
r
t
i
f
i
c
i
al
I
nt
e
l
l
i
ge
nc
e
and
A
ppl
i
c
at
i
ons
(
I
C
A
I
A
)
A
l
l
i
anc
e
T
e
c
hnol
ogy
C
onf
e
r
e
nc
e
(
A
T
C
O
N
-
1)
,
B
a
nga
l
or
e
, I
ndi
a
, 2023, pp. 1
-
6,
doi
:
10.1109/
I
C
A
I
A
57370.2023.10169488.
[
25]
R
.
C
h
e
r
uk
u,
K
.
H
us
s
a
i
n
,
I
.
K
a
va
t
i
,
A
.
M
.
R
e
dd
y,
a
n
d
K
.
S
.
R
e
dd
y,
“
S
e
nt
i
m
e
n
t
c
l
a
s
s
i
f
i
c
a
t
i
o
n
w
i
t
h
m
o
di
f
i
e
d
R
oB
E
R
T
a
a
n
d
r
e
c
ur
r
e
n
t
ne
u
r
a
l
ne
t
w
or
ks
,”
M
ul
t
i
m
e
d
i
a
T
oo
l
s
a
nd
A
pp
l
i
c
at
i
o
ns
,
no
. 8
3,
pp
.
293
99
–
29
41
7,
S
e
p
. 2
02
3,
do
i
:
10
.1
007
/
s
1
104
2
-
02
3
-
16
83
3
-
5.
[
26]
A
.
S
hoj
a
i
e
a
nd
E
.
B
.
F
ox,
“
G
r
a
nge
r
c
a
us
a
l
i
t
y:
a
r
e
vi
e
w
a
nd
r
e
c
e
nt
a
dv
a
nc
e
s
,
”
A
nnual
R
e
v
i
e
w
of
St
at
i
s
t
i
c
s
and
I
t
s
A
ppl
i
c
at
i
on
,
vol
. 9, pp. 289
–
319, N
ov. 2021, doi
:
10.1146/
a
nnur
e
v
-
s
t
a
t
i
s
t
i
c
s
-
040120.
[
27]
C
.
Y
a
ng,
K
.
X
i
a
o,
Y
.
A
o,
Q
.
C
ui
,
X
.
J
i
ng,
a
nd
Y
.
W
a
ng,
“
T
he
t
ha
l
a
m
us
i
s
t
he
c
a
us
a
l
hub
of
i
nt
e
r
ve
nt
i
on
i
n
pa
t
i
e
nt
s
w
i
t
h
m
a
j
o
r
de
pr
e
s
s
i
ve
di
s
or
de
r
:
E
vi
de
nc
e
f
r
om
t
he
G
r
a
nge
r
c
a
us
a
l
i
t
y
a
na
l
ys
i
s
,”
N
e
ur
oI
m
age
:
C
l
i
ni
c
al
,
vol
.
37,
J
a
n.
2023,
do
i
:
10.1016/
j
.ni
c
l
.2022.103295.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
E
x
pl
or
in
g t
he
i
nf
lu
e
nc
e
of
s
of
t
in
fo
r
m
at
io
n f
r
om
e
c
onomic
ne
w
s
on
…
(
R
ahar
di
to
D
io
P
r
as
to
w
o
)
5239
[
28]
J
.
H
.
J
a
n
a
nd
A
.
K
.
G
opa
l
a
s
w
a
m
y,
“
I
de
nt
i
f
yi
ng
f
a
c
t
or
s
i
n
c
ur
r
e
nc
y
e
xc
ha
nge
r
a
t
e
e
s
t
i
m
a
t
i
on:
a
s
t
udy
on
A
U
D
a
ga
i
ns
t
U
S
D
,
”
J
our
nal
of
A
dv
anc
e
s
i
n M
anage
m
e
nt
R
e
s
e
a
r
c
h
, vol
. 16, no. 4, pp. 436
–
452, O
c
t
. 2019, doi
:
10.1108/
J
A
M
R
-
09
-
2018
-
0084.
[
29]
M
.
F
l
or
e
s
-
S
os
a
,
E
.
L
e
ón
-
C
a
s
t
r
o,
J
.
M
.
M
e
r
i
gó,
a
nd
R
.
R
.
Y
a
ge
r
,
“
F
o
r
e
c
a
s
t
i
ng
t
he
e
xc
ha
nge
r
a
t
e
w
i
t
h
m
ul
t
i
pl
e
l
i
ne
a
r
r
e
g
r
e
s
s
i
on
a
nd
he
a
vy
or
de
r
e
d
w
e
i
ght
e
d
a
ve
r
a
ge
ope
r
a
t
or
s
,”
K
now
l
e
dge
-
B
as
e
d
Sy
s
t
e
m
s
,
vol
.
248,
J
ul
.
2022,
doi
:
10.1016/
j
.knos
ys
.2022.108863.
B
I
O
G
R
A
P
H
I
E
S
O
F
A
U
T
H
O
R
S
Rahardito
Dio
Prastowo
is
a
student
in
the
Master
of
Informati
on
Technology
program
at
the
University
of
Indonesia.
He
is
also
an
associate
re
searcher
at
the
National
Resear
ch
and
Innovati
on
Agency
of
the
Republic
of
Indones
ia.
His
research
focuses
on
social
listening,
social
impact
analysis,
and
social
media
analytics,
with
t
he
aim
of
exploring
the
applicati
on
of
informat
ion
technolo
gy
to
analyze
social
behavior
and
support
the
formulat
ion
of public policy and corporate strate
gy
.
He can be contacted
at email:
rahardito.dio@
ui.ac.id or
raha012@
brin.go.i
d.
Indra
Budi
is
a
distinguished
researcher
and
academic
whose
pion
eering
work
in
computer
science
and
artificial
intell
igence
spans
diverse
domains
s
uch
as
natural
language
processing,
machine
learning,
and
data
mining.
With
a
keen
focu
s
on
enhancing
human
-
computer
interacti
on,
his
ResearchGate
profile
highli
ghts
a
prolifi
c
a
rray
of
publicat
ions
and
collaborat
ions
that
underscore
his
dedicatio
n
to
innovat
ive
approache
s.
He
is
also
a
respected
faculty
member
at
UI’s
Department
of
Computer
Science,
where
hi
s
insight
s
and
expertise
significantly
contribute
to
shaping
the
evolving
landscape
of
AI
rese
arch
and
application
.
He
can be cont
acted at em
ail: i
ndra@
cs.ui.ac.i
d.
Amanah
Ramadiah
is
a
proficient
Data
Science
Lead
at
FNA,
driving
Suptech
solutions
for
Asia
-
Pacifi
c
centr
al
banks
and
co
-
developing
Central
Bank
Digital
Currency
simulations.
Also,
as
an
adjunct
faculty
member
at
Universitas
Indon
esia,
her
interdisciplina
ry
expertise
encompass
es
computer
science,
financ
e,
and
economics
,
f
ocusing
on
systemi
c
risk
and
digital
currencies.
A
visiting
scholar
at
the
Bank
of
England
and
IMF,
she
holds
a
Ph
.
D
.
and
M
.
Sc
.
from
University
College
London
and
a
B
.
Sc
.
from
Univer
sitas
Indonesia
.
She
can
be contacted at email:
amanah.ramad
iah06@
ui.ac.id
.
Aris
Budi
Santoso
is
a
skilled
software
engineer
and
data
analytic
s
trainer
with
a
Master'
s
in
Information
Technology
from
the
University
of
Indones
ia.
He
works
as
an
App
Developer
at
the
Ministry
of
Finance,
Indonesia,
and
excels
in
full
-
stack
development,
web
technologie
s,
and
software
architec
ture.
With
a
meticulous
problem
-
solving
approach,
he
delivers
high
-
quality
solutions
and
actively
contributes
to
innovativ
e
projects.
Proficient
in
Java and Python,
he
is a proac
tive collabo
rator who
remains c
ommitted to stayin
g updated w
ith
industry tre
nds
.
He can be contacted at email:
aris.bud
i
@
ui.ac.id.
Prabu
Kresna
Putra
is
an
accomplished
researcher
in
comput
er
science
and
technology.
He
works
at
the
National
Research
and
Innovation
Agenc
y'
s
Center
for
Geospatial
Resear
ch.
With
a
m
aster'
s
degree
in
Information
Technology
from
the
University
of
Indonesia
.
H
is
specialized
research
spans
data
mining,
social
media
analytic
s,
and
spatial
analytics
.
Profic
ient
in
softwa
re
develo
pment
and
artifi
cial
intellige
nce,
his
impactf
ul
advan
cemen
ts
resonate
across
professional
platforms,
solidify
ing
his
dedication
to
excellence
as
a
prominent
researcher.
He can be contacted at email:
prabu.kresna@ui.ac.id.
Evaluation Warning : The document was created with Spire.PDF for Python.