I
A
E
S
I
n
t
e
r
n
at
io
n
al
Jou
r
n
al
of
A
r
t
if
ic
ia
l
I
n
t
e
ll
ig
e
n
c
e
(
I
J
-
AI
)
V
ol
.
14
, N
o.
3
,
J
une
20
25
, pp.
1910
~
1918
I
S
S
N
:
2252
-
8938
,
D
O
I
:
10.11591/
ij
a
i.
v
14
.i
3
.pp
1910
-
1918
1910
Jou
r
n
al
h
om
e
page
:
ht
tp
:
//
ij
ai
.
ia
e
s
c
or
e
.c
om
A
ve
r
age
d
b
ar
s f
o
r
c
r
yp
t
o
c
u
r
r
e
n
c
y p
r
i
c
e
f
or
e
c
ast
i
n
g ac
r
oss
d
i
f
f
e
r
e
n
t
h
or
i
z
on
s
A
h
m
e
d
E
l
Y
ou
s
s
e
f
i
1
,
A
b
d
e
la
az
iz
H
e
s
s
an
e
1,
2
,
I
m
ad
Z
e
r
ou
al
1
, Y
ou
s
e
f
F
ar
h
aou
i
1
1
I
M
I
A
L
a
bor
a
t
or
y, T
-
I
D
M
S
, F
a
c
ul
t
y of
S
c
i
e
nc
e
s
a
nd T
e
c
hni
que
s
of
E
r
r
a
c
hi
di
a
,
M
oul
a
y I
s
m
a
i
l
U
ni
ve
r
s
i
t
y of
M
e
knè
s
,
M
e
kne
s
,
M
or
oc
c
o
2
D
e
pa
r
t
m
e
nt
of
C
om
put
e
r
S
c
i
e
nc
e
, F
a
c
ul
t
y of
S
c
i
e
nc
e
of
M
e
knè
s
,
M
oul
a
y I
s
m
a
i
l
U
ni
ve
r
s
i
t
y of
M
e
knè
s
,
M
e
kne
s
,
M
or
oc
c
o
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
A
pr
1
,
2024
R
e
vi
s
e
d
N
ov
18
,
2024
A
c
c
e
pt
e
d
N
ov
24
,
2024
Tec
hni
ca
l
an
al
ysi
s
us
es
p
as
t
pr
ice
mov
em
ent
s
an
d
pa
tt
er
ns
to
pr
edi
ct
f
utu
re
tre
nds
a
nd
h
elp
t
ra
der
s
ma
ke
in
fo
rme
d
de
cis
io
ns
ab
ou
t
t
he
ir
c
ryp
to
cu
rre
ncy
por
tfo
li
os.
T
his
s
tud
y
inv
es
tig
at
es
th
e
e
ff
ect
iv
en
ess
of
d
iff
er
en
t
f
or
e
cas
tin
g
alg
ori
th
ms
an
d
fe
atu
re
s
i
n
pr
edi
ct
ing
t
he
f
utu
re
lo
g
re
tur
n
of
c
ryp
to
cu
rre
ncy
clo
se
p
ri
ce
a
cro
ss
v
ar
io
us
h
ori
zo
ns.
Spe
ci
fi
cal
ly
,
we
c
om
pa
re
t
he
p
er
for
man
ce
of
Ad
aB
oos
t,
l
igh
t
gr
adi
en
t
bo
ost
in
g
mac
hi
ne
(
Lig
ht
GBM
)
,
r
and
om
fo
re
st
(RF
)
,
and
k
-
ne
ar
est
ne
ig
hbo
r
(K
NN
)
re
gr
ess
ors
u
sin
g
K
li
ne
ope
n,
hi
g
h,
l
ow
,
clo
se
(
OH
LC)
pr
ice
s
d
ata
an
d
av
er
age
d
b
ars
(H
eik
in
-
As
hi
)
fe
at
ure
s.
O
ur
ana
lys
is
c
ov
ers
ten
of
t
he
m
ost
cap
it
ali
ze
d
cr
ypt
oc
ur
ren
ci
es:
Ca
rda
no,
Ava
lan
ch
e,
B
ina
nc
e
Co
in
,
Bit
co
in
,
Do
gec
oi
n,
P
olk
ad
ot,
Eth
er
eum
,
Sol
ana
,
Tro
n,
a
nd
R
ip
ple
.
We
ha
ve
o
bse
rv
ed
n
uan
ce
d
pa
tt
ern
s
in
pr
e
dic
tiv
e
per
for
ma
nc
e
ac
ros
s
di
ff
ere
nt
cry
pt
ocu
rr
enc
ie
s,
f
ore
ca
st
ing
hor
iz
on
s
and
fea
tur
es
.
The
n
we
h
ave
f
ou
nd
th
at
A
daB
oo
st
an
d
RF
mo
de
ls
co
ns
i
ste
ntl
y
exh
ibi
t a co
mpe
ti
tiv
e pe
rf
orm
an
ce,
w
it
h Lig
ht
GBM
s
ho
win
g pr
omi
si
ng
res
ult
s
for
s
pe
ci
fic
cry
pt
ocu
rr
en
cie
s.
T
he
i
mp
ac
t
of
f
or
ec
ast
hor
iz
ons
on
f
or
e
cas
tin
g
per
for
ma
nc
e
un
de
rsc
or
es
the
ne
ed
fo
r
t
ail
or
ed
f
or
ec
ast
in
g
mo
d
els
.
I
n
sum
mar
y,
t
he
u
se
o
f
Kl
ine
OHL
C
da
ta
a
s
fe
at
ure
s
ou
tp
erf
or
ms
a
ver
ag
ed
ba
rs
in
fo
re
ca
sti
ng
t
he
f
ir
st
a
nd
s
ec
ond
ho
riz
on
s,
w
hil
e
a
ver
ag
ed
b
ars
ou
tp
erf
orm
Kli
ne
OH
LC
d
ata
f
or
mi
d
-
t
o
rel
at
ive
ly
l
ong
-
t
erm
h
or
izo
ns
(
sta
rt
ing
f
r
om
th
e
thi
rd
hor
iz
on
).
Our
fi
nd
ing
s
sug
ge
st
tha
t
ave
ra
ged
ba
rs
me
ri
t
m
ore
a
t
ten
tio
n
fro
m r
es
ea
rch
er
s i
ns
tea
d
of
re
lyi
ng
so
le
ly
on
Kl
in
e O
HL
C d
at
a.
K
e
y
w
o
r
d
s
:
A
ve
r
a
ge
d ba
r
s
B
oos
ti
ng a
lg
or
it
hm
s
C
ha
r
ti
ng t
e
c
hni
que
s
C
r
yp
to
c
ur
r
e
n
c
y
pr
i
c
e
f
or
e
c
a
s
ti
n
g
H
e
ik
in
-
a
s
hi
c
a
ndl
e
s
ti
c
k
s
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
:
A
hm
e
d E
l
Y
ous
s
e
f
i
I
M
I
A
L
a
bor
a
to
r
y, T
-
I
D
M
S
, F
a
c
ul
ty
of
S
c
ie
nc
e
s
a
nd T
e
c
hni
qu
e
s
of
E
r
r
a
c
hi
di
a
M
oul
a
y I
s
m
a
il
U
ni
ve
r
s
it
y of
M
e
knè
s
M
e
kne
s
,
M
or
oc
c
o
E
m
a
il
:
a
h.e
ly
ous
s
e
f
i@e
du.umi
.a
c
.m
a
1.
I
N
T
R
O
D
U
C
T
I
O
N
C
r
ypt
oc
ur
r
e
nc
ie
s
,
known
f
or
th
e
ir
hi
gh
vol
a
ti
li
ty
,
of
f
e
r
tr
a
de
r
s
lu
c
r
a
ti
ve
r
e
tu
r
ns
a
nd
hi
gh
r
is
ks
b
a
s
e
d
on
th
e
ir
de
c
is
io
n
s
’
a
c
c
ur
a
c
y
[
1]
.
T
r
a
d
e
r
s
c
om
m
onl
y
d
e
c
id
e
w
h
e
th
e
r
to
go
lo
ng
or
s
hor
t
on
a
c
r
ypt
oc
ur
r
e
nc
y
ba
s
e
d
on
th
e
ir
f
or
e
c
a
s
ti
ng
of
th
e
f
ut
ur
e
t
r
e
nd
of
th
e
c
r
ypt
oc
ur
r
e
nc
y
th
e
y
hol
d
or
to
w
hi
c
h
th
e
y
w
a
nt
to
ta
ke
a
pos
it
io
n.
F
or
e
c
a
s
ti
ng
c
r
ypt
oc
ur
r
e
nc
y
tr
e
nd
s
pos
e
s
a
s
ig
ni
f
ic
a
nt
c
ha
ll
e
nge
th
a
t
pr
om
pt
s
tr
a
de
r
s
to
e
m
pl
oy
va
r
io
us
te
c
hni
que
s
,
in
c
lu
di
ng
te
c
hni
c
a
l
a
na
ly
s
i
s
,
f
unda
m
e
nt
a
l
a
na
ly
s
is
,
s
ta
ti
s
ti
c
a
l
m
e
th
od
s
,
a
nd
a
r
ti
f
ic
ia
l
in
te
ll
ig
e
nc
e
-
ba
s
e
d a
ppr
oa
c
h
e
s
.
T
e
c
hni
c
a
l
a
n
a
ly
s
is
s
e
e
k
s
to
pr
e
di
c
t
f
ut
ur
e
m
a
r
ke
t
be
ha
vi
or
th
r
ough
pa
tt
e
r
n
r
e
c
ogni
ti
on
a
n
d
a
lg
or
it
hm
ic
tr
a
di
ng
s
tr
a
te
gi
e
s
.
T
hi
s
a
ppr
oa
c
h
le
ve
r
a
ge
s
te
c
h
ni
c
a
l
in
di
c
a
to
r
s
a
nd
c
ha
r
t
f
or
m
a
ti
ons
le
ve
ls
[
2]
,
[
3
]
,
de
r
iv
e
d
f
r
om
hi
s
to
r
ic
a
l
pr
ic
e
a
nd
vol
um
e
da
ta
(
e
.g.,
K
l
in
e
ope
n,
hi
gh,
lo
w
,
c
lo
s
e
(
O
H
L
C
)
va
lu
e
s
)
w
it
hi
n a
de
f
in
e
d t
im
e
f
r
a
m
e
. T
he
s
e
i
n
s
ig
ht
s
i
nf
or
m
c
r
ypt
oc
ur
r
e
nc
y por
tf
ol
io
m
a
na
ge
m
e
nt
de
c
is
io
ns
.
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
A
v
e
r
age
d bar
s
f
or
c
r
y
pt
oc
ur
r
e
nc
y
pr
ic
e
f
o
r
e
c
a
s
ti
ng ac
r
o
s
s
di
ff
e
r
e
nt
hor
iz
on
s
(
A
hm
e
d E
l
Y
ous
s
e
fi
)
1911
A
lt
e
r
na
ti
ve
ly
,
f
unda
m
e
nt
a
l
a
n
a
ly
s
is
c
ons
ti
tu
te
s
a
n
a
s
s
e
t
v
a
lu
a
ti
on
m
e
th
odol
ogy
[
4]
c
e
nt
e
r
e
d
on
de
te
r
m
in
in
g
th
e
in
tr
in
s
ic
va
lu
e
of
a
c
r
ypt
oc
ur
r
e
nc
y.
T
hi
s
a
ppr
oa
c
h
in
vol
ve
s
a
c
om
pr
e
he
ns
iv
e
e
v
a
lu
a
ti
on
of
th
e
f
ounda
ti
ona
l
c
om
pone
nt
s
dr
iv
in
g
a
pa
r
ti
c
ul
a
r
c
oi
n
pr
oj
e
c
t,
e
nc
om
pa
s
s
in
g
f
a
c
to
r
s
s
uc
h
a
s
it
s
und
e
r
ly
in
g
bl
oc
kc
ha
in
te
c
hnol
ogy
a
nd
upc
om
in
g
pr
oj
e
c
t
-
r
e
la
te
d
e
ve
nt
s
(
s
uc
h
a
s
p
a
r
tn
e
r
s
hi
ps
,
h
a
lv
in
g
e
ve
nt
s
,
or
in
tr
oduc
in
g
a
ne
w
c
ons
e
ns
us
a
lg
or
it
h
m
)
[
5]
.
A
ddi
ti
ona
ll
y,
f
un
da
m
e
nt
a
l
a
na
ly
s
is
c
on
s
id
e
r
s
th
e
a
dopt
io
n
r
a
te
of
th
e
pr
oj
e
c
t
’
s
s
e
r
vi
c
e
s
,
upda
te
s
f
r
om
th
e
pr
oj
e
c
t
te
a
m
,
a
nd
ot
he
r
pe
r
ti
ne
nt
pr
oj
e
c
t
-
r
e
la
te
d
in
f
o
r
m
a
ti
on.
U
s
in
g
th
e
s
e
in
s
ig
ht
s
,
f
unda
m
e
nt
a
li
s
ts
f
or
e
c
a
s
t
th
e
tr
e
nd
of
a
c
r
ypt
oc
ur
r
e
nc
y
’
s
pr
ic
e
.
H
ow
e
ve
r
,
a
pp
ly
i
ng
f
und
a
m
e
n
ta
l
a
n
a
ly
s
i
s
to
c
r
ypt
oc
ur
r
e
n
c
y
pr
i
c
e
f
or
e
c
a
s
t
in
g
r
e
m
a
in
s
a
c
o
m
pl
e
x
t
a
s
k
du
e
t
o t
he
s
ig
ni
f
i
c
a
nt
i
nf
lu
e
n
c
e
of
s
p
e
c
ul
a
ti
ve
b
e
h
a
v
io
r
,
m
a
r
k
e
t
s
e
n
ti
m
e
n
t
[
6]
,
a
n
d
ot
h
e
r
f
a
c
t
or
s
s
uc
h
a
s
t
h
e
hi
gh
m
a
r
ke
t
v
ol
a
ti
li
t
y [
7]
.
M
a
c
hi
ne
le
a
r
ni
ng
a
lg
or
it
hm
s
ha
ve
a
ls
o
ga
r
ne
r
e
d
s
ig
ni
f
ic
a
nt
a
tt
e
nt
io
n
a
s
vi
a
bl
e
a
ppr
oa
c
he
s
f
or
de
ve
lo
pi
ng
c
r
ypt
oc
ur
r
e
nc
y
f
or
e
c
a
s
ti
ng
m
ode
ls
.
B
y
l
e
ve
r
a
gi
ng
th
e
s
e
a
lg
or
it
hm
s
,
r
e
s
e
a
r
c
h
e
r
s
a
nd
tr
a
de
r
s
a
im
to
c
a
pi
ta
li
z
e
on
th
e
ir
a
bi
li
ty
to
pr
oc
e
s
s
la
r
ge
da
ta
s
e
ts
a
nd
u
nc
ove
r
tr
e
nds
a
nd
pa
tt
e
r
ns
th
a
t
c
oul
d
s
ig
na
l
f
or
th
c
om
in
g
pr
ic
e
f
lu
c
tu
a
ti
ons
.
T
ypi
c
a
ll
y,
th
e
s
e
a
lg
or
it
hm
s
ut
il
iz
e
hi
s
to
r
ic
a
l
K
li
ne
O
H
L
C
da
ta
,
s
our
c
e
d
di
r
e
c
tl
y
f
r
om
e
xc
ha
nge
a
r
c
hi
ve
s
or
a
ggr
e
ga
te
d
f
r
om
di
ve
r
s
e
pl
a
tf
or
m
s
,
to
tr
a
in
m
ode
ls
f
or
p
r
e
di
c
ti
ng
f
ut
ur
e
c
r
ypt
oc
ur
r
e
nc
y
pr
ic
e
s
ove
r
s
in
gl
e
or
m
ul
ti
pl
e
ti
m
e
s
te
ps
(
h
or
iz
ons
)
.
T
he
r
e
a
r
e
two
m
a
in
a
ppr
oa
c
he
s
to
c
r
ypt
oc
ur
r
e
nc
y
f
or
e
c
a
s
ti
ng:
c
la
s
s
if
ic
a
ti
on
a
nd
r
e
gr
e
s
s
io
n)
.
C
la
s
s
if
ic
a
ti
on
a
ppr
oa
c
he
s
pr
e
di
c
t
pr
ic
e
tr
e
nds
by
le
a
r
ni
ng
f
r
om
la
be
le
d
da
ta
(
e
.g.,
tr
e
nd
d
ir
e
c
ti
on
ba
s
e
d
on
K
l
in
e
O
H
L
C
)
a
nd
in
c
or
po
r
a
ti
ng
a
ddi
ti
ona
l
f
e
a
tu
r
e
s
[
8]
.
M
e
a
nw
hi
le
,
r
e
gr
e
s
s
io
n
m
ode
ls
ut
il
iz
e
hi
s
to
r
ic
a
l
pr
ic
e
da
ta
to
di
r
e
c
tl
y
f
or
e
c
a
s
t
f
ut
ur
e
c
r
ypt
oc
ur
r
e
nc
y
va
lu
e
s
w
it
hi
n a
s
p
e
c
if
ie
d t
im
e
f
r
a
m
e
.
S
ta
ti
s
ti
c
a
l
a
ppr
oa
c
he
s
a
r
e
a
ls
o
e
m
pl
oye
d
to
f
or
e
c
a
s
t
c
r
ypt
oc
ur
r
e
nc
ie
s
.
T
he
s
e
m
e
th
ods
in
c
lu
de
a
ut
or
e
gr
e
s
s
iv
e
in
te
gr
a
te
d
m
ovi
ng
a
ve
r
a
ge
(
A
R
I
M
A
)
m
ode
ls
f
or
ti
m
e
s
e
r
ie
s
a
na
ly
s
i
s
[
9]
a
nd
ot
he
r
s
ta
ti
s
ti
c
a
l
te
c
hni
que
s
th
a
t
c
a
pt
ur
e
pa
tt
e
r
ns
a
nd
tr
e
nds
in
hi
s
to
r
ic
a
l
p
r
ic
e
da
ta
.
S
ta
ti
s
ti
c
a
l
m
ode
ls
pr
ovi
de
a
qua
nt
it
a
ti
ve
f
r
a
m
e
w
or
k
f
or
unde
r
s
ta
ndi
ng
a
nd
pr
e
di
c
ti
ng
c
r
ypt
oc
ur
r
e
nc
y
pr
ic
e
m
ove
m
e
nt
s
.
H
ow
e
v
e
r
,
th
e
y
a
ls
o
n
e
e
d
h
e
lp
w
it
h
c
ha
ll
e
nge
s
,
in
c
lu
di
ng
a
s
s
um
pt
io
ns
a
bout
s
ta
ti
ona
r
it
y
a
nd
th
e
di
f
f
ic
ul
ty
of
c
a
pt
ur
in
g
th
e
c
om
pl
e
x
dyna
m
ic
s
of
th
e
c
r
ypt
oc
ur
r
e
nc
y
m
a
r
ke
t.
T
r
a
de
r
s
a
nd
a
na
ly
s
ts
of
te
n
c
om
bi
ne
m
ul
ti
pl
e
m
e
th
ods
,
dr
a
w
in
g
on
e
a
c
h ot
he
r
’
s
s
tr
e
ngt
hs
t
o e
nh
a
nc
e
t
he
r
obus
tn
e
s
s
of
t
he
ir
f
or
e
c
a
s
ts
[
10]
.
O
ur
s
tu
dy
a
im
s
to
c
om
pa
r
e
th
e
e
f
f
e
c
ti
ve
ne
s
s
of
us
in
g
a
ve
r
a
ge
d
ba
r
s
a
nd
K
li
ne
O
H
L
C
da
ta
a
s
f
e
a
tu
r
e
s
f
or
m
a
c
hi
ne
le
a
r
ni
ng
a
lg
or
it
hm
s
to
f
or
e
c
a
s
t
c
lo
s
e
pr
i
c
e
lo
g
r
e
tu
r
ns
a
c
r
os
s
va
r
io
us
c
r
ypt
oc
ur
r
e
nc
ie
s
a
nd
hor
iz
ons
.
W
hi
le
pr
e
vi
ous
li
te
r
a
tu
r
e
pr
im
a
r
il
y
f
oc
us
e
s
on
one
-
s
te
p
f
ut
ur
e
f
or
e
c
a
s
ti
ng
w
it
h
f
e
w
e
r
s
tu
di
e
s
ut
il
iz
in
g
a
ve
r
a
ge
d
ba
r
s
,
our
r
e
s
e
a
r
c
h
e
xt
e
nds
be
yond
th
e
s
e
li
m
it
a
ti
ons
.
B
y
ta
r
ge
ti
ng
di
f
f
e
r
e
nt
hor
iz
ons
a
nd
c
lo
s
e
pr
ic
e
s
of
te
n
c
r
ypt
oc
ur
r
e
nc
ie
s
,
w
e
s
e
e
k
to
a
ddr
e
s
s
s
e
ve
r
a
l
ke
y
que
s
ti
ons
th
a
t
ha
ve
not
be
e
n
th
or
oughly
e
xpl
or
e
d:
‒
W
hi
c
h
c
ha
r
ti
ng
te
c
hni
que
e
xhi
bi
ts
gr
e
a
te
r
e
f
f
ic
a
c
y
w
he
n
f
or
e
c
a
s
ti
ng
de
e
pe
r
hor
iz
ons
in
to
th
e
f
ut
ur
e
a
nd
us
in
g w
hi
c
h
m
a
c
hi
ne
l
e
a
r
ni
ng
a
lg
or
it
hm
?
‒
I
s
th
e
r
e
c
ons
is
te
nt
be
ha
vi
or
a
m
ong
th
e
di
f
f
e
r
e
nt
a
lg
o
r
it
hm
s
c
o
ns
id
e
r
e
d
in
th
is
s
tu
dy
w
he
n
a
ppl
ie
d
a
c
r
os
s
di
f
f
e
r
e
nt
c
r
ypt
oc
ur
r
e
nc
ie
s
?
T
he
nove
lt
y
of
our
r
e
s
e
a
r
c
h
li
e
s
in
it
s
c
om
pr
e
he
ns
iv
e
a
ppr
oa
c
h,
a
s
w
e
e
xt
e
nd
b
e
yond
pr
io
r
s
tu
di
e
s
by
e
xa
m
in
in
g
m
ul
ti
pl
e
c
r
ypt
oc
ur
r
e
nc
ie
s
a
nd
e
xpl
or
in
g
a
m
or
e
e
xt
e
ns
iv
e
r
a
nge
of
f
or
e
c
a
s
ti
ng
hor
iz
ons
.
A
ddi
ti
ona
ll
y,
unl
ik
e
our
pr
e
vi
ous
r
e
s
e
a
r
c
h
[
5]
w
hi
c
h
f
oc
us
e
d
p
r
im
a
r
il
y
on
B
it
c
oi
n
,
one
hor
iz
on,
a
nd
di
f
f
e
r
e
nt
ti
m
e
s
a
m
pl
in
g
w
in
dow
s
us
in
g
K
li
ne
O
H
L
C
da
ta
a
nd
H
e
ik
in
-
A
s
hi
,
th
is
s
tu
dy
br
oa
de
ns
th
e
s
c
ope
to
in
c
lu
de
a
ddi
ti
ona
l
c
r
ypt
oc
ur
r
e
nc
ie
s
a
nd
m
or
e
hor
iz
ons
.
T
hr
ough
th
is
c
om
pa
r
a
ti
ve
a
na
ly
s
is
,
w
e
a
im
to
pr
ovi
de
in
s
ig
ht
s
in
to
th
e
pr
e
di
c
ti
ve
c
a
pa
bi
li
ti
e
s
of
th
e
s
e
c
ha
r
ti
ng
te
c
hni
que
s
,
ul
ti
m
a
te
ly
c
ont
r
ib
ut
in
g
to
a
de
e
pe
r
unde
r
s
ta
ndi
ng of
c
r
ypt
oc
ur
r
e
nc
y m
a
r
ke
t
dyna
m
ic
s
.
T
he
s
ub
s
e
que
nt
s
e
c
ti
ons
of
th
is
pa
pe
r
a
r
e
or
ga
ni
z
e
d
a
s
f
ol
lo
w
s
:
T
he
s
e
c
ond
s
e
c
ti
on
w
il
l
in
tr
oduc
e
c
or
e
f
or
m
ul
a
s
f
or
c
a
lc
ul
a
ti
ng
a
ve
r
a
ge
d
b
a
r
s
a
nd
pr
ovi
de
a
n
ov
e
r
vi
e
w
of
r
e
c
e
nt
s
tu
di
e
s
a
bout
th
e
a
ppl
ic
a
ti
on
of
H
e
ik
in
-
A
s
hi
c
a
ndl
e
s
ti
c
ks
in
f
or
e
c
a
s
ti
ng
s
to
c
k
a
nd
c
r
ypt
oc
ur
r
e
nc
y
pr
ic
e
s
.
F
ol
lo
w
in
g
th
a
t,
th
e
th
ir
d
s
e
c
ti
on
w
il
l
de
s
c
r
ib
e
th
e
pr
oc
e
dur
e
s
f
or
da
ta
c
ol
le
c
ti
on,
pr
e
pr
oc
e
s
s
i
ng,
a
nd
th
e
m
e
th
odol
ogi
e
s
e
m
pl
oye
d.
I
n
th
e
f
our
th
s
e
c
ti
on,
w
e
w
il
l
d
e
lv
e
in
to
th
e
pr
e
s
e
nt
a
ti
on
a
nd
di
s
c
us
s
i
on
of
our
r
e
s
ul
ts
.
T
he
c
onc
lu
di
ng
f
if
th
s
e
c
ti
on
w
il
l
s
um
m
a
r
iz
e
t
he
pa
pe
r
a
nd outl
in
e
pot
e
nt
ia
l
di
r
e
c
ti
ons
f
or
f
ut
ur
e
r
e
s
e
a
r
c
h e
nde
a
vor
s
.
2.
B
A
C
K
G
R
O
U
N
D
A
N
D
R
E
L
A
T
E
D
WORKS
K
li
ne
c
a
ndl
e
s
ti
c
k
s
a
r
e
e
xt
e
ns
iv
e
ly
us
e
d
in
r
e
s
e
a
r
c
h
r
e
la
te
d
t
o
s
to
c
ks
a
nd
c
r
ypt
oc
ur
r
e
nc
y
pr
ic
e
s
[
11]
−
[
14]
.
H
ow
e
ve
r
,
th
e
a
ve
r
a
ge
d
b
a
r
s
known
a
s
H
e
ik
in
-
A
s
hi
c
a
ndl
e
s
ti
c
k
s
ne
e
d
m
or
e
a
tt
e
nt
io
n
f
r
om
th
e
s
c
ie
nt
if
ic
c
om
m
uni
ty
.
A
ve
r
a
ge
d
ba
r
s
c
a
ndl
e
s
ti
c
ks
e
xhi
bi
t
s
m
oot
he
r
pa
tt
e
r
ns
c
om
pa
r
e
d
to
K
li
ne
-
ba
s
e
d
c
a
ndl
e
s
ti
c
ks
due
to
th
e
ir
ut
il
iz
a
ti
on
of
a
n
a
ve
r
a
ge
of
th
e
pr
e
c
e
di
ng
c
a
ndl
e
s
ti
c
ks
’
ope
n
a
nd
c
lo
s
e
pr
ic
e
s
to
de
te
r
m
in
e
th
e
ope
n
pr
ic
e
.
W
it
hi
n
th
is
pa
pe
r
,
a
ve
r
a
ge
d
ba
r
s
a
nd
H
e
ik
in
-
A
s
hi
w
il
l
be
us
e
d
in
te
r
c
ha
nge
a
bl
y.
H
e
ik
in
-
A
s
hi
c
a
ndl
e
s
ti
c
k
s
a
r
e
de
r
iv
e
d
di
r
e
c
tl
y
f
r
om
K
li
ne
O
H
L
C
da
ta
.
T
he
f
or
m
ul
a
s
[
15]
to
c
a
lc
ul
a
te
t
he
ir
O
H
L
C
(
H
A
_O
pe
n, H
A
_H
ig
h, H
A
_L
ow
, H
A
_
C
l
os
e
)
da
ta
a
r
e
a
s
s
how
n i
n (
1)
t
o (
7)
:
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
,
V
ol
.
14
, N
o.
3
,
J
une
20
25
:
1910
-
1918
1912
HA
_
O
p
e
n
=
p
r
e
v
io
u
s
_
HA
_
O
p
e
n
+
p
r
e
v
io
u
s
_
HA
_
C
l
o
s
e
2
(
1)
HA
_
H
ig
h
=
m
a
x
(
K
L
in
e
_
ℎ
,
HA
_
O
p
e
n
,
HA
_
C
l
o
s
e
)
(
2)
HA
_
L
o
w
=
m
in
(
_
,
HA
_
O
p
e
n
,
HA
_
C
l
o
s
e
)
(
3)
HA
_
=
_
+
_
ℎ
+
_
+
_
4
(
4)
E
a
c
h vi
s
ua
li
z
e
d
H
e
ik
in
-
A
s
hi
c
a
ndl
e
s
ti
c
k ha
s
a
body a
nd t
w
o t
hi
n s
ha
dow
s
(
u
ppe
r
a
nd l
ow
e
r
)
.
_
=
_
−
_
(
5)
_
ℎ
=
_
ℎ
−
(
_
,
_
)
(
6)
HA
_
L
o
we
r
S
ha
d
o
w
=
m
in
(
_
,
_
)
−
_
L
o
w
(
7)
A
ppl
yi
ng
H
e
ik
in
-
A
s
hi
c
a
ndl
e
s
ti
c
ks
in
a
na
ly
z
in
g
c
r
ypt
oc
ur
r
e
nc
y
a
nd
s
to
c
k
m
a
r
ke
ts
ha
s
r
e
c
e
iv
e
d
r
e
la
ti
ve
ly
li
m
it
e
d
a
tt
e
nt
io
n
in
th
e
e
xi
s
ti
ng
li
te
r
a
tu
r
e
.
S
ha
li
ni
e
t
al
.
[
16]
e
m
pl
oys
th
e
a
ve
r
a
ge
di
r
e
c
ti
ona
l
in
de
x
(
A
D
X
)
on
H
e
ik
in
-
A
s
hi
w
it
h
ot
he
r
te
c
hni
c
a
l
in
di
c
a
to
r
s
.
T
he
b
a
c
kt
e
s
ti
ng
pr
oc
e
s
s
f
oc
u
s
e
s
on
pr
ovi
di
ng
e
nt
r
y
a
nd
e
xi
t
poi
nt
s
f
or
s
to
c
k
m
a
r
ke
t
pa
r
ti
c
ip
a
nt
s
.
T
h
e
ir
a
na
ly
s
is
r
e
ve
a
ls
th
a
t
th
e
ge
ne
r
a
te
d
te
c
hni
c
a
l
in
di
c
a
to
r
us
in
g
H
e
ik
in
-
A
s
hi
w
a
s
one
of
t
he
e
f
f
e
c
ti
ve
i
ndi
c
a
to
r
s
f
or
m
os
t
s
tu
di
e
d s
to
c
ks
.
E
l
Y
ous
s
e
f
i
e
t
al
.
[
5]
c
om
pa
r
e
d
H
e
ik
in
-
A
s
hi
a
nd
J
a
pa
ne
s
e
c
a
ndl
e
s
ti
c
ks
,
s
in
gl
e
-
s
te
p
f
ut
ur
e
lo
g
r
e
tu
r
n
o
f
th
e
c
lo
s
e
pr
ic
e
of
B
it
c
oi
n
ove
r
di
f
f
e
r
e
nt
ti
m
e
w
in
dow
s
r
a
ngi
ng
f
r
o
m
1
da
y
to
5
m
in
ut
e
s
,
a
nd
e
m
pl
oyi
ng
va
r
io
us
r
e
gr
e
s
s
io
n
a
lg
or
it
hm
s
in
c
lu
di
ng
k
-
ne
a
r
e
s
t
ne
ig
hbor
(
K
N
N
)
r
e
gr
e
s
s
or
,
li
ne
a
r
r
e
gr
e
s
s
io
n,
li
ght
gr
a
di
e
nt
boos
ti
ng
m
a
c
hi
ne
(
L
ig
ht
G
B
M
)
,
th
e
H
ube
r
r
e
gr
e
s
s
or
,
a
nd
r
a
ndom
f
or
e
s
t
(
R
F
)
r
e
gr
e
s
s
or
.
T
h
e
ir
ke
y
f
in
di
ngs
s
ugg
e
s
t
th
a
t
us
in
g
O
H
L
C
c
a
ndl
e
s
ti
c
k
s
c
on
s
is
te
nt
ly
out
pe
r
f
or
m
s
H
e
ik
in
-
A
s
hi
c
a
ndl
e
s
ti
c
k
s
a
c
r
os
s
a
ll
c
on
s
id
e
r
e
d
pe
r
io
ds
.
P
ia
s
e
c
ki
a
nd
H
a
nć
kow
ia
k
[
17]
u
s
e
d
f
uz
z
y
num
be
r
s
to
r
e
pr
e
s
e
nt
th
e
H
e
ik
in
-
A
s
hi
tr
a
ns
f
or
m
a
ti
on,
a
c
c
ount
in
g
f
or
th
e
in
he
r
e
nt
unc
e
r
ta
in
ty
in
hi
s
to
r
ic
a
l
pr
ic
e
da
ta
.
T
h
e
y
r
e
por
te
d
th
a
t
de
s
pi
te
in
tr
oduc
in
g
a
ddi
ti
ona
l
im
pr
e
c
is
io
n
th
r
ough
th
e
ir
a
ve
r
a
gi
ng
a
ppr
oa
c
h,
H
e
ik
in
-
A
s
hi
c
a
ndl
e
s
ti
c
ks
e
f
f
ic
ie
nt
ly
de
te
c
t
tr
e
nds
in
vol
a
ti
le
pr
ic
e
da
ta
,
r
e
s
ul
ti
ng
in
not
a
bl
e
f
or
e
c
a
s
ti
ng
a
c
c
ur
a
c
y
.
M
a
dbou
ly
e
t
al
.
[
18]
i
nt
e
gr
a
te
d
c
lo
ud
m
ode
ls
,
f
uz
z
y
ti
m
e
s
e
r
ie
s
,
a
nd
H
e
ik
in
-
A
s
hi
c
a
ndl
e
s
ti
c
ks
to
pr
e
di
c
t
a
nd
c
onf
ir
m
s
to
c
k
tr
e
nds
t
o
a
ddr
e
s
s
nonl
in
e
a
r
it
y
a
nd
noi
s
e
i
n s
to
c
k m
a
r
ke
t
da
ta
, t
he
m
ode
l
le
ve
r
a
ge
d a
m
ode
l
to
c
ove
r
t
he
r
a
ndomne
s
s
ga
p i
n f
uz
z
y l
ogi
c
, br
id
gi
n
g
qua
li
ta
ti
ve
a
nd
qu
a
nt
it
a
ti
ve
c
onc
e
pt
s
.
T
he
m
ode
l
h
a
ndl
e
d
a
m
bi
gui
ty
a
nd
unc
e
r
ta
in
ty
in
J
a
pa
ne
s
e
c
a
ndl
e
s
ti
c
k
de
f
in
it
io
ns
a
nd
a
c
tu
a
l
s
to
c
k
pr
ic
e
s
,
c
on
s
tr
uc
ti
ng
dyna
m
ic
w
e
ig
ht
e
d
f
uz
z
y
lo
gi
c
a
l
r
e
la
ti
ons
hi
ps
f
or
O
H
L
C
pr
ic
e
s
f
or
e
c
a
s
ti
ng
.
E
l
Y
ou
s
s
e
f
i
e
t
al
.
[
19]
us
e
d
kM
e
n
s
c
lu
s
te
r
in
g
to
c
a
te
gor
iz
e
a
ve
r
a
ge
d
b
a
r
s
c
a
ndl
e
s
ti
c
k
s
a
nd
lo
ga
r
it
hm
ic
r
e
tu
r
ns
of
pr
ic
e
s
,
a
nd
to
de
te
r
m
in
opt
im
a
l
c
la
s
s
num
be
r
s
f
or
c
r
ypt
oc
ur
r
e
nc
y
lo
ga
r
it
hm
ic
r
e
tu
r
ns
.
T
he
s
tu
dy
e
s
ta
bl
is
he
s
th
e
im
por
ta
nc
e
of
c
lu
s
te
r
in
g
in
f
e
a
tu
r
e
pr
e
pr
oc
e
s
s
in
g
f
or
e
f
f
e
c
ti
ve
c
la
s
s
if
ic
a
ti
on
in
c
r
ypt
oc
ur
r
e
nc
y f
or
e
c
a
s
ti
ng.
3.
M
E
T
H
O
D
S
A
N
D
M
A
T
E
R
IAL
S
O
ur
s
tu
dy
be
ga
n
by
c
on
s
tr
uc
ti
ng
our
da
ta
s
e
t
us
in
g
hi
s
to
r
ic
a
l
da
ta
f
r
om
B
in
a
nc
e
,
f
oc
us
in
g
on
t
e
n
s
e
le
c
te
d
c
r
ypt
oc
ur
r
e
nc
ie
s
.
A
s
de
pi
c
te
d
in
F
ig
ur
e
1,
w
e
obt
a
in
e
d
K
li
ne
O
H
L
C
da
ta
f
r
om
B
in
a
n
c
e
.
N
e
xt
,
w
e
ge
ne
r
a
te
d
va
r
io
us
f
e
a
tu
r
e
s
in
c
lu
di
ng
K
li
ne
O
H
L
C
c
a
ndl
e
s
ti
c
k
s
a
nd
a
v
e
r
a
ge
d
b
a
r
s
c
a
ndl
e
s
ti
c
k
s
.
A
ddi
ti
ona
ll
y,
w
e
c
a
lc
ul
a
te
d
di
f
f
e
r
e
nt
hor
iz
on
ta
r
ge
ts
,
s
pe
c
if
ic
a
ll
y
th
e
lo
ga
r
it
hm
ic
r
e
tu
r
ns
f
r
om
one
to
te
n
hor
iz
ons
.
I
n
th
e
da
ta
pr
e
pr
oc
e
s
s
in
g s
ta
ge
, w
e
a
ddr
e
s
s
e
d m
is
s
in
g va
lu
e
s
us
in
g s
i
m
pl
e
m
e
a
n
-
ba
s
e
d i
m
put
a
ti
on. W
e
t
he
n a
ppl
ie
d
a
ti
m
e
s
e
r
ie
s
s
pl
it
s
tr
a
te
gy
w
it
h
te
n
f
ol
ds
to
di
vi
de
e
a
c
h
da
t
a
s
e
t
in
to
a
tr
a
in
in
g
s
pl
it
(
70%
of
th
e
da
ta
)
a
nd
a
te
s
t
s
pl
it
(
30%
of
th
e
da
ta
)
.
T
he
da
ta
w
e
r
e
nor
m
a
li
z
e
d
us
in
g
th
e
r
obus
t
s
c
a
le
r
.
F
in
a
ll
y,
w
e
a
ppl
ie
d
our
s
e
le
c
te
d
r
e
gr
e
s
s
or
s
to
th
e
r
e
s
ul
ti
ng
da
ta
s
e
ts
.
T
he
pe
r
f
or
m
a
nc
e
of
th
e
s
e
r
e
gr
e
s
s
or
s
w
a
s
e
va
lu
a
te
d
us
in
g
th
e
c
oe
f
f
ic
ie
nt
of
de
te
r
m
in
a
ti
on, R
-
s
qua
r
e
d
(
R
2
)
, t
o a
s
s
e
s
s
t
he
a
c
c
ur
a
c
y of
our
f
or
e
c
a
s
ts
.
3.1.
D
at
a
c
ol
le
c
t
io
n
an
d
p
r
e
p
r
oc
e
s
s
in
g
B
a
s
e
d
on
m
a
r
k
e
t
c
a
pi
t
a
li
z
a
ti
o
n
a
va
il
a
bl
e
a
t
[
20]
,
w
e
h
a
ve
u
s
e
d
th
e
hi
s
to
r
i
c
a
l
da
t
a
pr
o
vi
de
d
by
B
in
a
n
c
e
of
t
h
e
t
e
n m
o
s
t
c
a
pi
t
a
li
z
e
d c
r
ypt
oc
ur
r
e
nc
i
e
s
a
s
of
J
a
nu
a
r
y 7, 2024, a
t
1
6:
23 G
M
T
. A
not
he
r
c
r
it
e
r
io
n i
s
th
a
t
th
e
c
r
ypt
o
c
ur
r
e
n
c
ie
s
s
h
oul
d
b
e
tr
a
da
bl
e
in
th
e
U
S
D
T
e
th
e
r
(
U
S
D
T
)
m
a
r
k
e
t.
T
he
c
ho
s
e
n
c
r
y
pt
oc
ur
r
e
nc
ie
s
a
lp
ha
be
ti
c
a
ll
y
or
d
e
r
e
d
a
r
e
C
a
r
da
no
(
A
D
A
)
,
A
v
a
la
n
c
he
(
A
V
A
X
)
,
B
in
a
nc
e
c
oi
n
(
B
N
B
)
,
B
it
c
oi
n
(
B
T
C
)
,
D
og
e
(
D
O
G
E
)
,
P
ol
ka
dot
(
D
O
T
)
,
E
th
e
r
e
um
(
E
T
H
)
,
S
o
la
n
a
(
S
O
L
)
,
T
r
o
n
(
T
R
X
)
,
a
nd
R
ip
pl
e
(
X
R
P
)
.
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
A
v
e
r
age
d bar
s
f
or
c
r
y
pt
oc
ur
r
e
nc
y
pr
ic
e
f
o
r
e
c
a
s
ti
ng ac
r
o
s
s
di
ff
e
r
e
nt
hor
iz
on
s
(
A
hm
e
d E
l
Y
ous
s
e
fi
)
1913
W
e
dow
nl
oa
de
d a
ll
t
he
K
li
ne
O
H
L
C
da
ta
f
r
om
t
he
ir
s
ta
r
t
da
te
o
f
t
r
a
di
ng on B
in
a
nc
e
unt
il
N
ove
m
be
r
30,
2023,
f
or
e
a
c
h
c
r
ypt
oc
ur
r
e
nc
y.
T
he
s
e
da
ta
in
c
lu
de
th
e
O
H
L
C
pr
ic
e
s
,
th
e
vol
um
e
,
a
nd
th
e
num
be
r
of
tr
a
de
s
.
H
e
ik
in
-
A
s
hi
O
H
L
C
a
nd
th
e
c
a
ndl
e
s
ti
c
ks
’
c
h
a
r
a
c
te
r
is
ti
c
s
(
body,
uppe
r
s
ha
dow
,
a
nd
lo
w
e
r
s
ha
dow
)
f
or
K
li
ne
a
nd
H
e
ik
in
-
A
s
hi
ha
ve
be
e
n
th
e
n
c
a
lc
ul
a
te
d
a
nd
a
ppe
nd
e
d
to
th
e
da
ta
.
T
h
e
n,
w
e
c
a
lc
ul
a
te
d
lo
ga
r
it
hm
ic
r
e
tu
r
ns
f
r
om
one
to
te
n
hor
iz
ons
f
or
e
a
c
h
d
a
ta
poi
nt
of
da
t
a
.
L
oga
r
it
hm
ic
r
e
tu
r
n
is
a
tr
a
ns
f
or
m
a
ti
on
w
id
e
ly
us
e
d
in
s
to
c
k
a
nd
c
r
ypt
oc
ur
r
e
nc
y
r
e
gr
e
s
s
io
n
-
ba
s
e
d
f
or
e
c
a
s
t
in
g
us
in
g
m
a
c
hi
ne
le
a
r
ni
ng
a
lg
or
it
hm
s
.
T
o
c
a
lc
ul
a
te
t
he
l
oga
r
it
hm
ic
r
e
tu
r
n f
or
t
he
n
th
hor
iz
on, us
in
g t
he
c
lo
s
e
pr
ic
e
of
a
c
r
ypt
oc
ur
r
e
nc
y,
th
e
(
8)
i
s
us
e
d:
L
og
_
r
e
tu
r
n
n
=
ln
(
_
(
)
)
−
ln
(
_
(
0
)
)
(
8)
W
he
r
e
_
(
0
)
is
th
e
c
ur
r
e
nt
c
a
ndl
e
s
ti
c
k
c
lo
s
e
pr
ic
e
a
nd
_
(
)
is
th
e
c
lo
s
e
pr
ic
e
of
th
e
n
th
f
ut
ur
e
K
li
ne
c
a
ndl
e
s
ti
c
k.
L
e
ve
r
a
gi
ng
th
e
P
yC
a
r
e
t
a
ut
om
a
te
d
m
a
c
hi
ne
l
e
a
r
ni
ng
li
br
a
r
y
[
21]
,
w
e
ha
ve
a
ppl
ie
d
a
ti
m
e
s
e
r
ie
s
c
r
os
s
-
va
li
da
ti
on
s
tr
a
te
gy
w
it
h
te
n
f
ol
ds
.
T
hi
s
a
ppr
oa
c
h
pr
io
r
it
iz
e
s
te
m
por
a
l
f
id
e
li
ty
by
e
xc
lu
s
iv
e
ly
tr
a
in
in
g
m
ode
ls
on
pa
s
t
da
ta
a
nd
ke
e
pi
ng
f
ut
ur
e
uns
e
e
n
da
ta
f
or
unbi
a
s
e
d
e
va
lu
a
ti
on.
T
hi
s
a
li
gns
w
it
h
th
e
in
he
r
e
nt
s
tr
uc
tu
r
e
of
c
r
ypt
oc
ur
r
e
nc
y
ti
m
e
s
e
r
ie
s
,
w
he
r
e
pr
e
di
c
ti
ons
b
a
s
e
d
on
una
va
il
a
bl
e
da
ta
a
r
e
m
or
e
a
ppl
ic
a
bl
e
.
A
ddi
ti
ona
ll
y, w
e
ha
ve
a
ddr
e
s
s
e
d m
is
s
in
g va
lu
e
s
t
hr
ough me
a
n i
m
put
a
ti
on. T
he
da
ta
s
e
t
w
a
s
s
ub
s
e
que
nt
ly
s
pl
it
in
to
a
70/
30
tr
a
in
in
g
-
te
s
ti
ng
r
a
ti
o
f
o
r
m
ode
l
tr
a
in
in
g
a
nd
pe
r
f
or
m
a
nc
e
a
s
s
e
s
s
m
e
nt
.
F
in
a
ll
y,
r
obus
t
s
c
a
li
ng
ha
s
be
e
n a
ppl
ie
d t
o m
it
ig
a
te
t
he
i
nf
lu
e
nc
e
of
out
li
e
r
s
on t
he
e
m
pl
oy
e
d a
lg
or
it
hm
s
.
F
ig
ur
e
1
. W
or
kf
lo
w
f
or
da
ta
c
ol
le
c
ti
on, f
e
a
tu
r
e
e
ngi
ne
e
r
in
g, da
ta
pr
e
pr
oc
e
s
s
in
g, a
nd mode
l
tr
a
in
in
g a
nd
e
va
lu
a
ti
on
3.2. L
e
ar
n
in
g r
e
gr
e
s
s
io
n
al
gor
it
h
m
s
W
e
e
m
p
lo
y
e
d
f
our
r
e
gr
e
s
s
io
n
m
od
e
l
s
—
A
d
a
B
o
os
t,
L
ig
ht
G
B
M
,
RF
,
a
nd
K
N
N
r
e
gr
e
s
s
or
s
—
to
c
om
pa
r
e
lo
g
r
e
tu
r
ns
a
c
r
o
s
s
di
f
f
e
r
e
nt
f
or
e
c
a
s
t
in
g
h
or
i
z
o
n
s
f
or
va
r
io
us
c
r
y
pt
o
c
ur
r
e
n
c
i
e
s
.
E
a
c
h
of
th
e
s
e
m
od
e
l
s
w
a
s
im
pl
e
m
e
n
te
d
u
s
i
ng
th
e
S
c
i
ki
t
-
le
a
r
n
li
br
a
r
y
[
22]
.
B
y
ut
il
iz
in
g
di
ve
r
s
e
e
n
s
e
m
bl
e
a
n
d
n
on
-
pa
r
a
m
e
tr
ic
t
e
c
hni
qu
e
s
,
w
e
a
im
e
d
to
c
a
pt
ur
e
d
if
f
e
r
e
nt
a
s
p
e
c
t
s
of
th
e
u
nd
e
r
ly
in
g
da
ta
p
a
tt
e
r
n
s
a
nd
a
s
s
e
s
s
th
e
ir
pr
e
d
ic
t
iv
e
c
a
p
a
b
il
it
ie
s
a
c
r
o
s
s
m
ul
ti
pl
e
t
im
e
h
or
i
z
o
ns
.
3.2.1.
A
d
aB
oos
t
r
e
gr
e
s
s
or
A
da
B
oos
t
e
m
pl
oy
s
a
boo
s
ti
ng
a
ppr
oa
c
h.
I
t
it
e
r
a
ti
ve
ly
tr
a
in
s
a
s
e
que
nc
e
of
w
e
a
k
l
e
a
r
ne
r
s
(
m
ode
ls
w
it
h
s
li
ght
ly
be
tt
e
r
a
c
c
ur
a
c
y
th
a
n
r
a
ndom
gue
s
s
in
g,
li
ke
s
im
pl
e
de
c
is
io
n
tr
e
e
s
)
.
T
he
s
e
w
e
a
k
le
a
r
ne
r
s
a
r
e
tr
a
in
e
d
on
m
odi
f
ie
d
v
e
r
s
io
ns
of
th
e
da
t
a
s
e
t.
T
he
ir
pr
e
di
c
ti
ons
a
r
e
th
e
n
c
om
bi
ne
d
us
in
g
a
w
e
ig
ht
e
d
vot
e
(
s
um
f
or
r
e
gr
e
s
s
io
n)
to
c
r
e
a
te
th
e
f
in
a
l
pr
e
di
c
ti
on.
A
t
e
a
c
h
boos
ti
ng
it
e
r
a
ti
on,
th
e
da
t
a
unde
r
goe
s
m
odi
f
ic
a
ti
ons
w
he
r
e
w
e
ig
ht
s
1
,
2
,
…
,
a
r
e
a
s
s
ig
ne
d
to
e
a
c
h
tr
a
in
in
g
s
a
m
pl
e
.
I
ni
ti
a
ll
y,
th
e
s
e
w
e
ig
ht
s
a
r
e
uni
f
or
m
ly
s
e
t
to
1
,
a
ll
ow
in
g
th
e
f
ir
s
t
s
t
e
p
to
tr
a
in
a
w
e
a
k
le
a
r
ne
r
on
th
e
or
ig
in
a
l
d
a
ta
.
A
d
a
B
oos
t
a
dj
us
t
s
s
a
m
pl
e
w
e
ig
ht
s
dyna
m
ic
a
ll
y.
S
a
m
pl
e
s
m
is
c
la
s
s
if
ie
d
by
th
e
p
r
e
vi
ous
boos
te
d
m
ode
l
ga
in
hi
ghe
r
w
e
ig
ht
s
,
w
hi
le
c
or
r
e
c
tl
y
c
la
s
s
if
ie
d
s
a
m
pl
e
s
ha
ve
th
e
ir
w
e
ig
ht
s
de
c
r
e
a
s
e
d.
T
hi
s
it
e
r
a
ti
ve
pr
oc
e
s
s
f
oc
u
s
e
s
th
e
le
a
r
ni
ng
a
lg
or
it
hm
on pr
e
vi
ous
ly
c
ha
ll
e
ngi
ng s
a
m
pl
e
s
i
n s
ub
s
e
que
nt
i
te
r
a
ti
ons
[
23]
.
3.2.2.
L
ig
h
t
gr
ad
ie
n
t
b
oos
t
in
g m
ac
h
in
e
L
ig
ht
G
B
M
is
a
gr
a
di
e
nt
boos
ti
ng
de
c
is
io
n
tr
e
e
a
lg
or
it
hm
w
hi
c
h
us
e
a
f
or
w
a
r
d
s
ta
ge
-
w
is
e
a
ppr
oa
c
h,
in
it
ia
ll
y
bui
ld
in
g
a
ba
s
e
m
ode
l
to
pr
e
di
c
t
th
e
ta
r
ge
t
va
r
ia
bl
e
’
s
m
e
a
n.
S
ubs
e
que
nt
ly
,
it
it
e
r
a
ti
ve
ly
r
e
f
in
e
s
th
is
m
ode
l
by
c
ons
tr
uc
ti
ng
de
c
is
io
n
tr
e
e
s
th
a
t
f
oc
us
on
th
e
r
e
s
id
u
a
ls
(
e
r
r
or
s
)
be
twe
e
n
th
e
a
c
tu
a
l
va
lu
e
s
a
nd
th
e
c
ur
r
e
nt
pr
e
di
c
ti
ons
.
T
h
e
s
e
r
e
s
id
ua
l
s
a
r
e
c
a
lc
ul
a
te
d
a
s
th
e
n
e
ga
ti
ve
gr
a
di
e
nt
of
th
e
c
hos
e
n
lo
s
s
f
unc
ti
on.
T
he
m
ode
l
is
th
e
n
upda
te
d
by
in
c
or
por
a
ti
ng
a
por
ti
on
of
e
a
c
h
n
e
w
tr
e
e
’
s
pr
e
di
c
ti
ons
.
T
hi
s
a
ppr
oa
c
h
e
nha
n
c
e
s
m
ode
l
pe
r
f
or
m
a
nc
e
w
hi
le
m
a
in
ta
in
in
g c
om
put
a
ti
ona
l
e
f
f
ic
ie
nc
y
[
24]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
,
V
ol
.
14
, N
o.
3
,
J
une
20
25
:
1910
-
1918
1914
3.2.3. Rand
om
f
or
e
s
t
r
e
gr
e
s
s
o
r
RF
is
a
m
a
c
hi
ne
le
a
r
ni
ng
a
lg
or
it
hm
th
a
t
ut
il
iz
e
s
a
m
ul
ti
tu
de
of
de
c
is
io
n
tr
e
e
s
f
or
pr
e
di
c
ti
on.
D
ur
in
g
tr
a
in
in
g,
it
c
ons
tr
uc
ts
a
f
or
e
s
t
of
unpr
une
d
de
c
is
io
n
tr
e
e
s
,
w
h
e
r
e
e
a
c
h
tr
e
e
is
bui
lt
on
a
r
a
ndom
s
ubs
e
t
of
f
e
a
tu
r
e
s
a
nd
a
r
a
ndom
s
ubs
e
t
of
tr
a
in
in
g
da
ta
.
P
r
e
di
c
ti
ons
a
r
e
m
a
de
by
a
ve
r
a
gi
ng
th
e
pr
e
di
c
ti
ons
f
r
om
a
ll
in
di
vi
dua
l
tr
e
e
s
w
it
hi
n
th
e
e
n
s
e
m
bl
e
.
T
hi
s
a
ppr
oa
c
h
e
nha
nc
e
s
m
ode
l
r
obus
tn
e
s
s
a
nd
r
e
duc
e
s
ove
r
f
it
ti
ng
c
om
pa
r
e
d t
o s
in
gl
e
de
c
is
io
n t
r
e
e
s
.
3.2.4.
K
-
n
e
ar
e
s
t
n
e
ig
h
b
or
r
e
gr
e
s
s
or
K
N
N
is
a
non
-
pa
r
a
m
e
tr
ic
,
in
s
ta
nc
e
-
ba
s
e
d
le
a
r
ni
ng
m
e
th
od
pr
e
va
le
nt
in
bot
h
s
ta
ti
s
ti
c
s
a
nd
m
a
c
hi
ne
le
a
r
ni
ng.
U
nl
ik
e
pa
r
a
m
e
tr
ic
a
ppr
oa
c
he
s
,
K
N
N
a
voi
ds
a
s
s
um
pt
io
ns
a
bout
da
ta
di
s
tr
ib
ut
io
n
a
nd
r
e
li
e
s
on
th
e
tr
a
in
in
g
da
ta
f
or
p
r
e
di
c
ti
ons
.
I
t
pr
e
di
c
ts
c
ont
in
uous
va
r
ia
bl
e
s
by
c
ons
id
e
r
in
g
th
e
‘
K
’
c
lo
s
e
s
t
ne
ig
hbor
s
in
th
e
tr
a
in
in
g
s
e
t.
T
h
e
a
lg
or
it
hm
id
e
nt
if
ie
s
th
e
s
e
n
e
ig
hbor
s
us
in
g
a
di
s
ta
n
c
e
m
e
tr
ic
,
e
.g.,
E
uc
li
de
a
n
di
s
ta
nc
e
.
T
he
f
in
a
l
pr
e
di
c
ti
on
is
th
e
n
c
a
lc
ul
a
te
d
a
s
th
e
a
v
e
r
a
ge
of
th
e
de
pe
nde
nt
va
r
ia
bl
e
va
lu
e
s
a
s
s
oc
ia
te
d
w
it
h
th
e
s
e
‘
K
’
ne
ig
hbor
s
[
5]
a
s
s
how
n i
n (
9)
.
y
̂
=
1
K
∑
y
i
K
i
=
1
(
9)
W
hi
le
th
e
K
N
N
r
e
gr
e
s
s
or
of
f
e
r
s
f
le
xi
bi
li
ty
a
nd
s
im
pl
ic
it
y
a
nd
c
a
n
e
f
f
e
c
ti
ve
ly
c
a
pt
ur
e
c
om
pl
e
x
va
r
ia
bl
e
r
e
la
ti
ons
hi
ps
in
s
p
e
c
if
ic
da
ta
s
e
ts
,
it
s
pe
r
f
or
m
a
nc
e
r
e
li
e
s
h
e
a
vi
l
y
on
f
a
c
to
r
s
s
uc
h
a
s
th
e
da
ta
’
s
di
m
e
ns
io
na
li
ty
a
nd s
c
a
li
ng, t
he
c
ho
s
e
n va
lu
e
of
‘
K
’
a
nd dis
ta
nc
e
m
e
tr
ic
.
3.3. E
val
u
at
io
n
m
e
t
r
ic
s
T
he
c
oe
f
f
ic
ie
nt
of
de
te
r
m
in
a
ti
on,
or
R
²,
is
a
s
ta
ti
s
ti
c
a
l
m
e
tr
ic
th
a
t
in
di
c
a
te
s
how
w
e
ll
a
r
e
gr
e
s
s
io
n
m
ode
l
f
it
s
th
e
da
ta
.
I
t
m
e
a
s
ur
e
s
th
e
pr
opor
ti
on
of
va
r
ia
ti
on
in
t
he
de
pe
nde
nt
va
r
ia
bl
e
(
Y
)
th
a
t
is
e
xpl
a
in
e
d
by
th
e
in
de
pe
nde
nt
va
r
ia
bl
e
s
(
X
)
in
c
lu
de
d
in
th
e
m
ode
l.
E
s
s
e
nt
ia
ll
y,
R
²
is
th
e
r
a
ti
o
of
th
e
e
xpl
a
in
e
d
s
um
of
s
qua
r
e
s
(
E
S
S
)
to
th
e
to
ta
l
s
um
of
s
qua
r
e
s
(
T
S
S
)
[
25]
.
E
S
S
r
e
f
le
c
ts
th
e
va
r
ia
ti
on
c
a
pt
ur
e
d
by
th
e
m
ode
l,
w
hi
le
T
S
S
r
e
pr
e
s
e
nt
s
t
he
t
ot
a
l
va
r
ia
ti
on i
n t
he
de
pe
nd
e
nt
va
r
ia
bl
e
. T
h
e
f
or
m
ul
a
f
or
c
a
lc
ul
a
ti
ng R
2
a
s
s
how
n i
n
(
10)
:
2
=
=
1
−
=
1
−
∑
(
−
)
2
=
1
∑
(
−
̅
)
2
=
1
(
10)
A
hi
ghe
r
R
2
va
lu
e
in
di
c
a
te
s
th
a
t
a
m
or
e
s
ig
ni
f
ic
a
nt
p
r
opor
ti
on
of
th
e
to
ta
l
va
r
ia
nc
e
in
th
e
de
pe
nde
nt
va
r
ia
bl
e
is
a
c
c
ount
e
d
f
or
by
th
e
in
de
pe
nde
nt
va
r
ia
bl
e
s
in
th
e
r
e
gr
e
s
s
io
n
m
od
e
l,
s
ugg
e
s
ti
ng
a
be
tt
e
r
f
it
of
th
e
m
ode
l
to
th
e
obs
e
r
ve
d
da
ta
.
T
hi
s
s
tu
dy
us
e
d
R
2
a
s
a
n
e
v
a
lu
a
ti
on
m
e
tr
ic
f
or
di
f
f
e
r
e
nt
c
r
ypt
oc
ur
r
e
nc
y
f
or
e
c
a
s
ti
ng
m
ode
ls
.
R
2
qua
nt
if
ie
s
th
e
pr
opor
ti
on
of
va
r
ia
nc
e
in
th
e
de
pe
nde
nt
va
r
ia
bl
e
e
xpl
a
in
e
d
by
th
e
m
ode
l,
th
e
r
e
by
of
f
e
r
in
g
a
s
tr
a
ig
ht
f
or
w
a
r
d
a
s
s
e
s
s
m
e
nt
of
m
ode
l
pe
r
f
or
m
a
nc
e
.
N
ot
a
bl
y,
R
2
is
s
c
a
le
-
in
de
pe
nde
nt
,
f
a
c
il
it
a
ti
ng
c
om
pa
r
is
ons
a
c
r
o
s
s
di
ve
r
s
e
d
a
ta
s
e
t
s
o
r
s
c
a
le
s
.
A
s
R
2
w
a
s
r
e
c
om
m
e
nde
d
in
[
26]
,
w
e
w
il
l
r
e
por
t
a
gr
a
phi
c
a
l
r
e
pr
e
s
e
nt
a
ti
on of
R
2
r
e
gr
e
s
s
io
n va
lu
e
s
on
ly
.
4.
R
E
S
U
L
T
S
A
N
D
D
I
S
C
U
S
S
I
O
N
T
hi
s
s
tu
dy
in
ve
s
ti
ga
te
d
th
e
pe
r
f
or
m
a
nc
e
of
f
our
f
o
r
e
c
a
s
ti
ng
a
lg
or
it
hm
s
:
A
da
B
oos
t,
L
ig
ht
G
B
M
,
R
F
,
a
nd
KNN
r
e
gr
e
s
s
or
s
,
us
in
g
K
li
ne
O
H
L
C
a
nd
H
e
ik
in
-
A
s
hi
(
a
ve
r
a
ge
d
ba
r
s
)
f
e
a
tu
r
e
s
to
f
or
e
c
a
s
t
th
e
f
ut
u
r
e
lo
g
r
e
tu
r
n
of
th
e
c
lo
s
e
pr
ic
e
ove
r
di
f
f
e
r
e
nt
ho
r
iz
ons
(
H
1
to
H
10)
.
W
hi
le
e
a
r
li
e
r
s
tu
di
e
s
ha
ve
e
xpl
or
e
d
va
r
io
us
f
or
e
c
a
s
ti
ng
m
e
th
ods
,
th
e
y
ha
ve
not
e
xpl
ic
it
ly
a
ddr
e
s
s
e
d
th
e
in
f
lu
e
nc
e
of
f
e
a
tu
r
e
ty
pe
s
a
nd
f
or
e
c
a
s
t
hor
iz
on
s
on
pr
e
di
c
ti
ve
pe
r
f
or
m
a
nc
e
f
o
r
c
r
ypt
oc
ur
r
e
nc
ie
s
.
F
ig
ur
e
s
2
to
11
pr
e
s
e
nt
th
e
R
2
va
lu
e
s
of
th
e
c
ons
id
e
r
e
d
a
lg
or
it
hm
s
a
c
r
os
s
va
r
io
us
hor
iz
ons
f
or
bot
h
f
e
a
tu
r
e
ty
pe
s
.
N
ot
a
bl
y,
a
ll
ne
ga
ti
ve
R
2
va
lu
e
s
w
e
r
e
e
xc
lu
de
d
to
f
oc
us
on me
a
ni
ngf
ul
pos
it
iv
e
da
ta
.
O
ur
ke
y
f
in
di
ngs
r
e
ve
a
l
th
a
t
th
e
us
e
of
a
ve
r
a
ge
d
b
a
r
s
g
e
ne
r
a
ll
y
le
a
ds
to
be
tt
e
r
R
2
v
a
lu
e
s
,
out
pe
r
f
or
m
in
g
K
li
ne
O
H
L
C
f
e
a
tu
r
e
s
s
ta
r
ti
ng
f
r
om
th
e
3
rd
hor
iz
on.
T
hi
s
im
pr
ove
m
e
nt
i
s
e
vi
de
nt
in
c
r
ypt
oc
ur
r
e
nc
ie
s
s
uc
h
a
s
E
th
e
r
e
um
,
B
it
c
oi
n,
a
nd
C
a
r
da
n
o.
S
pe
c
if
ic
a
ll
y,
A
da
B
oo
s
t
a
nd
R
F
m
ode
ls
c
ons
is
te
nt
ly
de
m
on
s
tr
a
te
hi
gh
pe
r
f
or
m
a
nc
e
w
it
h
a
ve
r
a
ge
d
b
a
r
s
f
e
a
tu
r
e
s
.
C
onv
e
r
s
e
ly
,
L
ig
ht
G
B
M
s
how
s
pr
om
is
e
f
or
s
pe
c
if
ic
c
r
ypt
oc
ur
r
e
nc
ie
s
,
not
a
bl
y
B
it
c
oi
n
a
nd
E
th
e
r
e
um
,
w
hi
le
K
N
N
m
od
e
ls
e
xhi
bi
t
le
s
s
c
ons
is
te
nt
pe
r
f
or
m
a
nc
e
,
w
it
h
lo
w
e
r
R
2
va
lu
e
s
a
c
r
os
s
m
os
t
c
r
ypt
oc
ur
r
e
nc
ie
s
a
nd
hor
iz
ons
e
xc
e
pt
f
or
A
va
la
nc
he
a
nd
B
in
a
nc
e
c
oi
n.
O
ur
s
tu
dy
s
ugge
s
ts
th
a
t
a
ve
r
a
ge
d
ba
r
s
of
f
e
r
be
tt
e
r
r
e
s
ul
ts
f
or
lo
nge
r
hor
iz
ons
in
f
or
e
c
a
s
ti
ng
c
r
ypt
oc
ur
r
e
nc
y
pr
ic
e
s
,
w
he
r
e
a
s
K
li
ne
O
H
L
C
d
a
t
a
a
r
e
m
or
e
e
f
f
e
c
ti
ve
f
or
s
hor
t
-
te
r
m
hor
iz
ons
(1
st
to
3
rd
hor
iz
ons
)
, de
s
pi
te
t
he
ir
ge
ne
r
a
ll
y l
ow
e
r
R
2
va
lu
e
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
A
v
e
r
age
d bar
s
f
or
c
r
y
pt
oc
ur
r
e
nc
y
pr
ic
e
f
o
r
e
c
a
s
ti
ng ac
r
o
s
s
di
ff
e
r
e
nt
hor
iz
on
s
(
A
hm
e
d E
l
Y
ous
s
e
fi
)
1915
F
ig
ur
e
2
.
R
2
va
lu
e
s
f
or
A
D
A
/US
D
T
p
a
ir
c
lo
s
e
pr
ic
e
lo
g r
e
tu
r
n f
or
e
c
a
s
ti
ng
F
ig
ur
e
3
.
R
2
va
lu
e
s
f
or
A
V
A
X
/US
D
T
p
a
ir
c
lo
s
e
pr
ic
e
lo
g r
e
tu
r
n f
or
e
c
a
s
ti
ng
F
ig
ur
e
4
.
R
2
va
lu
e
s
f
or
B
N
B
/US
D
T
pa
ir
c
lo
s
e
pr
ic
e
lo
g r
e
tu
r
n f
or
e
c
a
s
ti
ng
F
ig
ur
e
5
.
R
2
va
lu
e
s
f
or
B
T
C
/US
D
T
pa
ir
c
lo
s
e
pr
ic
e
lo
g r
e
tu
r
n f
or
e
c
a
s
ti
ng
F
ig
ur
e
6
.
R
2
va
lu
e
s
f
or
D
O
G
E
/US
D
T
pa
ir
c
lo
s
e
pr
ic
e
lo
g r
e
tu
r
n f
or
e
c
a
s
ti
ng
F
ig
ur
e
7
.
R
2
va
lu
e
s
f
or
D
O
T
/US
D
T
pa
ir
c
lo
s
e
pr
ic
e
lo
g r
e
tu
r
n f
or
e
c
a
s
ti
ng
F
ig
ur
e
8
.
R
2
va
lu
e
s
f
or
B
T
C
/US
D
T
pa
ir
c
lo
s
e
pr
ic
e
lo
g r
e
tu
r
n
f
or
e
c
a
s
ti
ng
F
ig
ur
e
9
.
R
2
va
lu
e
s
f
or
S
O
L
/US
D
T
pa
ir
c
lo
s
e
pr
ic
e
lo
g r
e
tu
r
n f
or
e
c
a
s
ti
ng
0
0
.5
H1
H2
H3
H4
H5
H6
H7
H8
H9
H1
0
R
2
H
o
r
i
zo
n
s
Ca
r
d
a
n
o
a
d
a
l
i
g
h
tg
b
m
rf
k
n
n
a
d
a
_a
b
l
i
g
h
tg
b
m
_a
b
0
0
.2
H1
H2
H3
H4
H5
H6
H7
H8
H9
H1
0
R
2
H
o
r
i
zo
n
s
A
v
a
l
a
n
c
h
e
a
d
a
l
i
g
h
tg
b
m
rf
k
n
n
a
d
a
_a
b
l
i
g
h
tg
b
m
_a
b
r
f
_a
b
k
n
n
_a
b
0
0
.5
H1
H2
H3
H4
H5
H6
H7
H8
H9
H1
0
R
2
H
o
r
i
zo
n
s
B
i
n
a
n
c
e
Co
i
n
a
d
a
l
i
g
h
tg
b
m
rf
k
n
n
a
d
a
_a
b
l
i
g
h
tg
b
m
_a
b
0
0
.5
H1
H2
H3
H4
H5
H6
H7
H8
H9
H1
0
R
2
H
o
r
i
zo
n
s
B
i
t
c
o
i
n
a
d
a
l
i
g
h
tg
b
m
rf
k
n
n
a
d
a
_a
b
l
i
g
h
tg
b
m
_a
b
0
0
.5
H1
H2
H3
H4
H5
H6
H7
H8
H9
H1
0
R
2
H
o
r
i
zo
n
s
D
o
g
e
Co
i
n
a
d
a
l
i
g
h
t
g
b
m
rf
k
n
n
a
d
a
_a
b
l
i
g
h
tg
b
m
_a
b
0
0
.5
H1
H2
H3
H4
H5
H6
H7
H8
H9
H1
0
R
2
H
o
r
i
zo
n
s
P
o
l
k
a
d
o
t
a
d
a
l
i
g
h
tg
b
m
rf
k
n
n
a
d
a
_
a
b
l
i
g
h
tg
b
m
_
a
b
0
0
.2
0
.4
0
.6
H1
H2
H3
H4
H5
H6
H7
H8
H9
H
1
0
R
2
H
o
r
i
zo
n
s
E
t
h
er
eu
m
a
d
a
l
i
g
h
tg
b
m
rf
k
n
n
a
d
a
_a
b
l
i
g
h
tg
b
m
_a
b
0
0
.2
0
.4
H1
H2
H3
H4
H5
H6
H7
H8
H9
H1
0
R
2
H
o
r
i
zo
n
s
S
o
l
a
n
a
a
d
a
l
i
g
h
tg
b
m
rf
k
n
n
a
d
a
_a
b
l
i
g
h
tg
b
m
_a
b
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
,
V
ol
.
14
, N
o.
3
,
J
une
20
25
:
1910
-
1918
1916
F
ig
ur
e
10
.
R
2
va
lu
e
s
f
or
T
R
X
/US
D
T
p
a
ir
c
lo
s
e
pr
ic
e
lo
g r
e
tu
r
n f
or
e
c
a
s
ti
ng
F
ig
ur
e
11
.
R
2
va
lu
e
s
f
or
X
R
P
/US
D
T
pa
ir
c
lo
s
e
pr
ic
e
lo
g r
e
tu
r
n f
or
e
c
a
s
ti
ng
5.
C
O
N
C
L
U
S
I
O
N
T
hi
s
s
tu
dy e
xpl
or
e
d
th
e
pe
r
f
or
m
a
nc
e
of
f
our
f
or
e
c
a
s
ti
ng
a
lg
or
it
hm
s
:
A
da
B
oos
t,
L
ig
ht
G
B
M
,
RF
,
a
nd
K
N
N
r
e
gr
e
s
s
or
s
.
W
e
us
e
d
K
li
ne
O
H
L
C
da
ta
f
or
one
s
e
t
of
f
e
a
tu
r
e
s
a
nd
H
e
ik
in
-
a
s
hi
f
e
a
tu
r
e
s
f
or
a
not
he
r
.
T
he
s
e
r
e
gr
e
s
s
io
n
a
lg
or
it
hm
s
w
e
r
e
a
ppl
ie
d
to
f
or
e
c
a
s
t
th
e
f
ut
ur
e
da
il
y
lo
g
r
e
tu
r
n
of
th
e
c
lo
s
in
g
pr
ic
e
ove
r
va
r
io
us
hor
iz
ons
(
H
1
to
H
10)
.
K
e
y
in
s
ig
ht
s
e
m
e
r
ge
d
f
r
om
e
va
lu
a
ti
ng
m
ul
ti
pl
e
c
r
ypt
oc
ur
r
e
nc
ie
s
a
nd
f
or
e
c
a
s
ti
ng
hor
iz
ons
.
F
ir
s
tl
y,
f
e
a
tu
r
e
s
e
le
c
ti
on
is
c
r
uc
ia
l
f
o
r
pr
e
di
c
ti
ve
a
c
c
ur
a
c
y.
W
hi
le
H
e
ik
in
-
A
s
hi
f
e
a
tu
r
e
s
yi
e
ld
be
tt
e
r
r
e
s
ul
ts
f
or
lo
nge
r
hor
iz
ons
in
f
or
e
c
a
s
ti
ng
c
r
ypt
oc
ur
r
e
nc
y
pr
ic
e
s
,
th
e
ir
s
upe
r
io
r
it
y
is
not
uni
ve
r
s
a
l
a
c
r
os
s
a
ll
c
r
ypt
oc
ur
r
e
nc
ie
s
a
nd
f
or
e
c
a
s
t
hor
iz
on
s
.
T
he
y
p
e
r
f
or
m
be
s
t
a
t
th
e
th
ir
d
hor
iz
on
a
nd
th
e
n
va
r
y
f
or
lo
nge
r
hor
iz
ons
,
in
di
c
a
ti
ng
th
a
t
f
e
a
tu
r
e
e
f
f
e
c
ti
ve
ne
s
s
de
p
e
nds
on
th
e
s
pe
c
if
ic
c
h
a
r
a
c
te
r
is
ti
c
s
of
e
a
c
h
c
r
ypt
oc
ur
r
e
nc
y
’
s
pr
ic
e
be
h
a
vi
or
a
nd
th
e
f
or
e
c
a
s
t
hor
iz
on.
S
e
c
ondl
y,
th
e
pe
r
f
or
m
a
nc
e
of
f
or
e
c
a
s
ti
ng
a
lg
or
it
hm
s
va
r
ie
s
s
ig
ni
f
ic
a
nt
ly
.
A
da
B
oos
t
a
nd
RF
m
ode
ls
s
h
ow
c
om
pe
ti
ti
ve
pe
r
f
or
m
a
nc
e
a
c
r
os
s
m
ul
ti
pl
e
c
r
ypt
oc
ur
r
e
nc
ie
s
a
nd hor
iz
ons
, w
hi
le
L
ig
ht
G
B
M
pr
oduc
e
s
pr
o
m
is
in
g r
e
s
ul
ts
f
or
s
pe
c
if
ic
c
r
ypt
oc
ur
r
e
nc
ie
s
a
nd
hor
iz
ons
, hi
ghl
ig
ht
in
g i
ts
e
f
f
ic
a
c
y i
n m
ode
li
ng nonli
ne
a
r
r
e
la
ti
o
ns
hi
ps
. L
a
s
tl
y, t
he
i
m
pa
c
t
of
f
o
r
e
c
a
s
t
hor
iz
ons
on
pr
e
di
c
ti
ve
pe
r
f
or
m
a
nc
e
e
m
pha
s
iz
e
s
th
e
im
por
ta
nc
e
of
ta
il
or
in
g
f
or
e
c
a
s
ti
ng
m
ode
ls
to
s
pe
c
if
ic
ti
m
e
f
r
a
m
e
s
a
nd
m
a
r
ke
t
c
ondi
ti
ons
.
S
hor
t
-
te
r
m
hor
iz
ons
m
a
y
f
a
vor
a
lg
or
it
hm
s
th
a
t
a
da
pt
to
r
a
pi
d
p
r
ic
e
f
lu
c
tu
a
ti
ons
,
w
he
r
e
a
s
lo
nge
r
-
te
r
m
hor
iz
ons
m
a
y
r
e
qui
r
e
m
ode
ls
th
a
t
id
e
nt
if
y
s
us
ta
in
e
d
tr
e
nds
in
c
r
ypt
oc
ur
r
e
nc
y
pr
ic
e
s
.
O
ur
f
in
di
ngs
e
m
pha
s
iz
e
th
e
im
por
ta
nc
e
of
in
c
or
por
a
ti
ng
a
ve
r
a
ge
d
ba
r
s
(
H
e
ik
in
-
A
s
hi
)
a
nd
c
on
s
id
e
r
in
g
f
or
e
c
a
s
t
hor
iz
ons
in
c
r
ypt
oc
ur
r
e
nc
y
f
or
e
c
a
s
ti
ng
r
e
s
e
a
r
c
h.
F
ur
th
e
r
r
e
s
e
a
r
c
h
is
r
e
qui
r
e
d
to
r
e
f
in
e
th
e
s
e
m
e
th
odol
ogi
e
s
a
nd
de
ve
lo
p
r
obus
t
m
ode
ls
th
a
t
c
a
n
c
a
pt
ur
e
th
e
c
om
pl
e
xi
ti
e
s
of
c
r
ypt
oc
ur
r
e
nc
y
p
r
ic
e
dyna
m
ic
s
.
I
t
is
im
por
ta
nt
to
not
e
th
e
li
m
it
a
ti
ons
of
our
s
tu
dy,
in
c
lu
di
ng
th
e
e
xc
lu
s
iv
e
u
s
e
of
hi
s
to
r
ic
a
l
da
ta
f
r
om
B
in
a
nc
e
a
nd
th
e
f
oc
us
on
a
s
pe
c
if
ic
s
e
t
of
c
r
ypt
oc
ur
r
e
n
c
ie
s
a
nd
f
or
e
c
a
s
t
hor
iz
ons
.
T
o
va
li
da
te
th
e
s
e
r
e
s
ul
ts
,
f
ut
ur
e
r
e
s
e
a
r
c
h
s
houl
d
e
xt
e
nd
to
a
br
oa
de
r
r
a
nge
of
c
r
ypt
oc
ur
r
e
nc
ie
s
a
nd
e
xpl
or
e
di
f
f
e
r
e
nt
ti
m
e
s
a
m
pl
in
gs
be
yond
th
e
1
-
d
a
y
c
a
ndl
e
s
ti
c
k
s
us
e
d
in
th
is
s
tu
dy.
A
ddi
ti
ona
ll
y,
f
ut
ur
e
s
tu
di
e
s
m
ig
ht
in
ve
s
ti
ga
te
m
or
e
s
ophi
s
ti
c
a
te
d
f
e
a
tu
r
e
e
ngi
ne
e
r
in
g
m
e
th
ods
a
nd
th
e
a
ppl
i
c
a
ti
on
of
a
lt
e
r
na
ti
ve
r
e
gr
e
s
s
io
n
a
lg
or
it
hm
s
to
f
ur
th
e
r
e
nha
nc
e
pr
e
di
c
ti
ve
a
c
c
ur
a
c
y.
F
U
N
D
I
N
G
I
N
F
O
R
M
A
T
I
O
N
T
hi
s
r
e
s
e
a
r
c
h w
a
s
n’
t
f
unde
d by a
ny gr
a
nt
.
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
hi
s
jo
ur
na
l
us
e
s
th
e
C
ont
r
ib
ut
or
R
ol
e
s
T
a
xonomy
(
C
R
e
d
iT
)
to
r
e
c
ogni
z
e
in
di
vi
dua
l
a
ut
hor
c
ont
r
ib
ut
io
ns
, r
e
duc
e
a
ut
hor
s
hi
p di
s
put
e
s
,
a
nd f
a
c
il
it
a
te
c
ol
la
bo
r
a
ti
on
N
am
e
o
f
A
u
t
h
or
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
A
hm
e
d E
L
Y
ous
s
e
f
i
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
A
bde
la
a
z
iz
H
e
s
s
a
ne
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
I
m
a
d Z
e
r
oua
l
✓
✓
✓
✓
✓
Y
ous
e
f
F
a
r
ha
oui
✓
✓
✓
✓
✓
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
0
0
.5
H1
H2
H3
H4
H5
H6
H7
H8
H9
H
1
0
R
2
H
o
r
i
zo
n
s
T
r
o
n
a
d
a
l
i
g
h
tg
b
m
rf
k
n
n
a
d
a
_
a
b
l
i
g
h
tg
b
m
_a
b
0
0.5
H1
H2
H3
H4
H5
H6
H7
H8
H9
H
1
0
R
2
H
o
r
i
zo
n
s
R
i
p
p
l
e
a
d
a
l
i
g
h
tg
b
m
rf
k
n
n
a
d
a
_a
b
l
i
g
h
tg
b
m
_a
b
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
A
v
e
r
age
d bar
s
f
or
c
r
y
pt
oc
ur
r
e
nc
y
pr
ic
e
f
o
r
e
c
a
s
ti
ng ac
r
o
s
s
di
ff
e
r
e
nt
hor
iz
on
s
(
A
hm
e
d E
l
Y
ous
s
e
fi
)
1917
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
A
ut
hor
s
s
ta
te
no c
onf
li
c
t
of
i
nt
e
r
e
s
t.
D
A
T
A
A
V
A
I
L
A
B
I
L
I
T
Y
D
a
t
a
a
va
i
la
bi
l
it
y
i
s
not
a
pp
li
c
a
b
le
t
o t
hi
s
pa
pe
r
a
s
no
n
e
w
d
a
t
a
w
e
r
e
c
r
e
a
t
e
d
or
a
na
ly
z
e
d
in
th
i
s
s
tu
d
y.
R
E
F
E
R
E
N
C
E
S
[
1]
B
.
C
he
n
a
nd
Y
.
S
un,
“
R
i
s
k
c
ha
r
a
c
t
e
r
i
s
t
i
c
s
a
nd
c
onne
c
t
e
dne
s
s
i
n
c
r
ypt
oc
ur
r
e
nc
y
m
a
r
ke
t
s
:
n
e
w
e
vi
de
nc
e
f
r
om
a
non
-
l
i
ne
a
r
f
r
a
m
e
w
or
k,”
T
he
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
. 69, J
a
n. 2024, doi
:
10.1016/
j
.na
j
e
f
.2023.102036.
[
2]
S
. C
or
be
t
, V
. E
r
a
s
l
a
n, B
.
L
uc
e
y, a
nd A
.
S
e
ns
oy, “
T
he
e
f
f
e
c
t
i
ve
ne
s
s
of
t
e
c
hni
c
a
l
t
r
a
di
ng r
ul
e
s
i
n c
r
ypt
oc
ur
r
e
nc
y m
a
r
ke
t
s
,”
F
i
nanc
e
R
e
s
e
ar
c
h L
e
t
t
e
r
s
, vol
. 31, pp. 32
–
37, D
e
c
. 2019, doi
:
10.1016/
j
.f
r
l
.2019.04.027.
[
3]
I
.
L
yuke
vi
c
h,
I
.
G
or
ba
t
e
nko,
a
nd
E
.
B
e
s
s
onova
,
“
C
r
ypt
oc
ur
r
e
nc
y
m
a
r
ke
t
:
c
h
oi
c
e
of
t
e
c
hni
c
a
l
i
ndi
c
a
t
or
s
i
n
t
r
a
di
ng
s
t
r
a
t
e
gi
e
s
of
i
ndi
vi
dua
l
i
nve
s
t
or
s
,”
i
n
3
rd
I
nt
e
r
nat
i
onal
Sc
i
e
nt
i
f
i
c
C
onf
e
r
e
nc
e
on
I
nnov
at
i
o
ns
i
n
D
i
gi
t
al
E
c
onom
y
,
S
a
i
nt
-
P
e
t
e
r
s
bur
g
R
us
s
i
a
n
F
e
de
r
a
t
i
on:
A
C
M
, O
c
t
. 2021, pp. 408
–
416
, doi
:
10.1145/
3527049.3527089.
[
4]
A
.
T
ha
k
ka
r
a
nd
K
.
C
ha
udh
a
r
i
,
“
A
c
om
pr
e
he
ns
i
v
e
s
ur
ve
y
on
p
or
t
f
ol
i
o
o
pt
i
m
i
z
a
t
i
o
n,
s
t
oc
k
pr
i
c
e
a
nd
t
r
e
n
d
p
r
e
d
i
c
t
i
o
n
us
i
ng
p
a
r
t
i
c
l
e
s
w
a
r
m
o
pt
i
m
i
z
a
t
i
on
,”
A
r
c
h C
o
m
pu
t
a
t
M
e
t
h
ods
E
ngi
ne
e
r
i
n
g
,
vo
l
.
28
, n
o.
4,
p
p.
21
33
–
216
4,
20
21
, d
oi
:
10.
10
07
/
s
11
831
-
020
-
094
48
-
8.
[
5]
A
.
E
l
Y
ous
s
e
f
i
,
A
.
H
e
s
s
a
ne
,
I
.
Z
e
r
oua
l
,
a
nd
Y
.
F
a
r
ha
oui
,
“
U
t
i
l
i
z
i
ng
m
a
c
hi
ne
l
e
a
r
ni
ng
a
nd
d
e
e
p
l
e
a
r
ni
ng
f
o
r
p
r
e
di
c
t
i
ng
c
r
ypt
o
-
c
ur
r
e
nc
y
t
r
e
nds
,”
Sal
ud, C
i
e
nc
i
a y
T
e
c
nol
ogí
a
-
Se
r
i
e
d
e
C
onf
e
r
e
nc
i
as
, vol
. 3,
M
a
r
. 2024, doi
:
10.56294/
s
c
t
c
onf
2024638.
[
6]
G
. D
ude
k, P
. F
i
s
z
e
de
r
, P
. K
ubu
s
, a
nd
W
. O
r
z
e
s
z
ko, “
F
or
e
c
a
s
t
i
ng c
r
ypt
oc
ur
r
e
nc
i
e
s
vol
a
t
i
l
i
t
y us
i
ng
s
t
a
t
i
s
t
i
c
a
l
a
nd m
a
c
hi
n
e
l
e
a
r
ni
ng
m
e
t
hods
:
a
c
om
pa
r
a
t
i
ve
s
t
udy
,”
SSR
N
, 2023
, doi
:
10.2139/
s
s
r
n.4409549.
[
7]
B
.
Y
.
A
l
m
a
ns
our
,
M
.
M
.
A
l
s
ha
t
e
r
,
a
nd
A
.
Y
.
A
l
m
a
ns
our
,
“
P
e
r
f
or
m
a
nc
e
of
A
R
C
H
a
nd
G
A
R
C
H
m
ode
l
s
i
n
f
or
e
c
a
s
t
i
n
g
c
r
ypt
oc
ur
r
e
nc
y
m
a
r
ke
t
vol
a
t
i
l
i
t
y
,”
I
ndus
t
r
i
al
E
ngi
ne
e
r
i
ng
&
M
anage
m
e
nt
Sy
s
t
e
m
s
,
vol
.
20,
no.
2,
pp.
130
–
139,
J
un.
2021,
doi
:
10.7232/
i
e
m
s
.2021.20.2.130.
[
8]
E
.
A
kyi
l
di
r
i
m
,
A
.
G
onc
u,
a
nd
A
.
S
e
ns
oy,
“
P
r
e
di
c
t
i
on
of
c
r
ypt
oc
u
r
r
e
nc
y
r
e
t
ur
ns
us
i
ng
m
a
c
hi
ne
l
e
a
r
ni
ng,”
A
nnal
s
of
O
pe
r
at
i
on
s
R
e
s
e
ar
c
h
, vol
. 297, no. 1
–
2, pp. 3
–
36, F
e
b. 2021, doi
:
10.1007/
s
10479
-
020
-
03575
-
y.
[
9]
Z
.
W
a
ng,
Z
.
Y
a
ng,
Z
.
Z
he
ng,
a
nd
Y
.
Z
hu,
“
B
i
t
c
oi
n
pr
i
c
e
f
o
r
e
c
a
s
t
i
ng
ba
s
e
d
on
a
r
i
m
a
m
ode
l
a
nd
m
ul
t
i
f
a
c
t
or
i
a
l
l
i
ne
a
r
r
e
gr
e
s
s
i
on
,”
i
n
2023
I
nt
e
r
nat
i
onal
C
onf
e
r
e
nc
e
on
N
e
t
w
or
k
i
ng,
I
nf
or
m
at
i
c
s
and
C
om
put
i
ng
(
I
C
N
E
T
I
C
)
,
P
a
l
e
r
m
o,
I
t
a
l
y:
I
E
E
E
,
M
a
y
2023,
pp. 15
–
20
, doi
:
10.1109/
I
C
N
E
T
I
C
59568.2023.00009.
[
10]
A
.
E
l
Y
ous
s
e
f
i
,
A
.
H
e
s
s
a
ne
,
Y
.
F
a
r
ha
oui
,
a
nd
I
.
Z
e
r
oua
l
,
“
C
r
ypt
oc
ur
r
e
nc
y
r
e
t
ur
ns
c
l
us
t
e
r
i
ng
us
i
ng
j
a
pa
ne
s
e
c
a
ndl
e
s
t
i
c
ks
:
t
ow
a
r
d
s
a
pr
ogr
a
m
m
a
t
i
c
t
r
a
di
ng
s
ys
t
e
m
,”
i
n
A
dv
anc
e
d
T
e
c
hnol
ogy
f
or
Sm
ar
t
E
nv
i
r
onm
e
nt
and
E
ne
r
gy
,
C
ha
m
:
S
pr
i
nge
r
I
nt
e
r
na
t
i
ona
l
P
ubl
i
s
hi
ng, 2023, pp. 93
–
103
, doi
:
10.1007/
978
-
3
-
031
-
25662
-
2_8.
[
11]
R
.
K
.
A
l
khodha
i
r
i
,
S
.
R
.
A
l
j
a
l
ha
m
i
,
N
.
K
.
R
us
a
yni
,
J
.
F
.
A
l
s
hoba
i
l
i
,
A
.
A
.
A
l
-
S
ha
r
ga
bi
,
a
nd
A
.
A
l
a
bdul
a
t
i
f
,
“
B
i
t
c
oi
n
c
a
ndl
e
s
t
i
c
k
pr
e
di
c
t
i
on
w
i
t
h
de
e
p
n
e
ur
a
l
ne
t
w
or
ks
ba
s
e
d
on
r
e
a
l
t
i
m
e
da
t
a
,”
C
om
p
ut
e
r
s
,
M
at
e
r
i
al
s
&
C
ont
i
nua
,
vol
.
68,
no.
3,
pp. 3215
–
3233, 2021, doi
:
10.32604/
c
m
c
.2021.016881.
[
12]
M
.
S
he
r
a
z
,
S
.
D
e
du,
a
nd
V
.
P
r
e
da
,
“
V
ol
a
t
i
l
i
t
y
dyna
m
i
c
s
of
n
on
-
l
i
ne
a
r
vol
a
t
i
l
e
t
i
m
e
s
e
r
i
e
s
a
nd
a
na
l
ys
i
s
of
i
nf
or
m
a
t
i
on
f
l
ow
:
e
vi
de
nc
e
f
r
om
c
r
ypt
oc
ur
r
e
nc
y da
t
a
,”
E
nt
r
opy
, vol
. 24, no. 10, O
c
t
. 2022, doi
:
10.3390/
e
24101410.
[
13]
S
.
S
i
m
t
ha
r
a
ka
o
a
nd
D
.
S
ut
i
vong,
“
E
xpl
or
i
ng
nor
m
a
l
i
z
a
t
i
on
t
e
c
hni
que
s
i
n
ne
ur
a
l
ne
t
w
or
ks
f
or
bi
t
c
oi
n
c
a
ndl
e
s
t
i
c
k
pr
i
c
e
pr
e
di
c
t
i
on
,”
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
i
n
I
nf
or
m
at
i
on
and
C
om
m
uni
c
at
i
on
(
I
C
A
I
I
C
)
,
B
a
l
i
,
I
ndone
s
i
a
:
I
E
E
E
,
F
e
b. 2023, pp. 483
–
488
, doi
:
10.1109/
I
C
A
I
I
C
57133.2023.100
67086.
[
14]
S
.
S
i
va
pr
a
kka
s
h
a
nd
S
.
V
e
ve
k,
“
P
r
i
c
e
vol
a
t
i
l
i
t
y
i
n
c
r
ypt
oc
ur
r
e
nc
i
e
s
:
a
m
ode
l
l
i
ng
a
ppr
oa
c
h
,”
i
n
A
dv
anc
e
s
i
n
F
i
nanc
e
,
A
c
c
ount
i
ng
,
and E
c
onom
i
c
s
,
I
G
I
G
l
oba
l
, 2023, pp. 29
–
43
, doi
:
10.4018/
978
-
1
-
6684
-
5691
-
0.
c
h002.
[
15]
O
.
A
.
H
a
s
s
e
n,
S
.
M
.
D
a
r
w
i
s
h,
N
.
A
.
A
bu,
a
nd
Z
.
Z
.
A
bi
di
n,
“
A
ppl
i
c
a
t
i
on
o
f
c
l
oud
m
ode
l
i
n
qua
l
i
t
a
t
i
ve
f
or
e
c
a
s
t
i
ng
f
or
s
t
o
c
k
m
a
r
ke
t
t
r
e
nds
,”
E
nt
r
opy
,
vol
. 22, no. 9, S
e
p. 2020, doi
:
10.3390/
e
22090991.
[
16]
T
.
S
ha
l
i
ni
,
S
.
P
r
a
na
v,
a
nd
S
.
U
t
ka
r
s
h,
“
P
i
c
ki
ng
buy
-
s
e
l
l
s
i
gna
l
s
:
a
pr
a
c
t
i
t
i
one
r
’
s
pe
r
s
pe
c
t
i
ve
on
k
e
y
t
e
c
hni
c
a
l
i
ndi
c
a
t
or
s
f
or
s
e
l
e
c
t
e
d i
ndi
a
n f
i
r
m
s
,
”
St
udi
e
s
i
n B
us
i
ne
s
s
and E
c
onom
i
c
s
, vol
. 14, no. 3, pp. 205
–
219, D
e
c
. 2019, doi
:
10.2478/
s
be
-
2019
-
0054.
[
17]
K
.
P
i
a
s
e
c
ki
a
nd
A
.
Ł
.
-
H
a
nć
kow
i
a
k,
“
H
e
i
ki
n
-
a
s
hi
t
e
c
hni
que
w
i
t
h
us
e
of
or
i
e
nt
e
d
f
uz
z
y
num
be
r
s
,”
U
nc
e
r
t
ai
nt
y
and
I
m
pr
e
c
i
s
i
on
i
n
D
e
c
i
s
i
on
M
ak
i
ng
and
D
e
c
i
s
i
on
Suppor
t
:
N
e
w
A
dv
anc
e
s
,
C
hal
l
e
nge
s
,
and
P
e
r
s
pe
c
t
i
v
e
s
,
C
ha
m
:
S
pr
i
nge
r
I
nt
e
r
na
t
i
ona
l
P
ubl
i
s
hi
ng,
2022, pp. 60
–
71
, doi
:
10.1007/
978
-
3
-
030
-
95929
-
6_5.
[
18]
M
.
M
.
M
a
dboul
y,
M
.
E
l
khol
y,
Y
.
M
.
G
ha
r
i
b,
a
nd
S
.
M
.
D
a
r
w
i
s
h,
“
P
r
e
di
c
t
i
ng
s
t
oc
k
m
a
r
ke
t
t
r
e
nds
f
or
j
a
pa
ne
s
e
c
a
ndl
e
s
t
i
c
k
us
i
n
g
c
l
oud
m
ode
l
,”
i
n
P
r
oc
e
e
di
ng
s
of
t
he
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
C
om
put
e
r
V
i
s
i
on
(
A
I
C
V
2020
)
,
C
ha
m
:
S
pr
i
nge
r
I
nt
e
r
na
t
i
ona
l
P
ubl
i
s
hi
ng, 2020, pp. 628
–
645
, doi
:
10.1007/
978
-
3
-
030
-
44289
-
7_59.
[
19]
A
.
E
l
Y
ous
s
e
f
i
,
A
.
H
e
s
s
a
ne
,
A
.
E
l
A
l
l
a
oui
,
I
.
Z
e
r
oua
l
,
a
nd
Y
.
F
a
r
ha
oui
,
“
H
e
i
ki
n
a
s
hi
c
a
ndl
e
s
t
i
c
ks
f
or
c
r
ypt
oc
ur
r
e
nc
y
r
e
t
ur
ns
c
l
us
t
e
r
i
ng
,”
A
r
t
i
f
i
c
i
al
I
nt
e
l
l
i
ge
nc
e
and
S
m
ar
t
E
nv
i
r
on
m
e
nt
,
C
h
a
m
:
S
pr
i
nge
r
I
nt
e
r
na
t
i
ona
l
P
ubl
i
s
hi
ng,
2023,
pp.
481
–
485
,
doi
:
10.1007/
978
-
3
-
031
-
26254
-
8_69.
[
20]
“
Y
a
hoo
F
i
na
nc
e
-
s
t
oc
k
m
a
r
ke
t
l
i
ve
,
quot
e
s
,
bus
i
ne
s
s
a
nd
f
i
na
nc
e
ne
w
s
,
”
Y
a
hoo
F
i
nanc
e
.
A
c
c
e
s
s
e
d:
M
a
r
.
22,
2024.
[
O
nl
i
ne
]
.
A
va
i
l
a
bl
e
:
ht
t
ps
:
/
/
f
i
na
nc
e
.ya
hoo.c
om
/
[
21]
“
P
yC
a
r
e
t
:
A
n
ope
n
s
our
c
e
,
l
ow
-
c
ode
m
a
c
hi
ne
l
e
a
r
ni
ng
l
i
br
a
r
y
i
n
P
yt
hon
,”
P
y
C
ar
e
t
,
2020.
A
c
c
e
s
s
e
d:
N
ov.
23,
2023.
[
O
nl
i
ne
]
.
A
va
i
l
a
bl
e
:
ht
t
ps
:
/
/
pyc
a
r
e
t
.or
g/
[
22]
F
.
P
e
dr
e
gos
a
e
t
al
.,
“
S
c
i
ki
t
-
l
e
a
r
n:
m
a
c
hi
ne
l
e
a
r
ni
ng
i
n
pyt
hon
,”
J
our
nal
of
M
a
c
hi
ne
L
e
ar
ni
ng
R
e
s
e
ar
c
h
,
vol
.
12,
pp.
2825
–
2830,
2011.
[
23]
“
1.11.
E
ns
e
m
bl
e
s
:
G
r
a
di
e
nt
boos
t
i
ng,
r
a
ndom
f
or
e
s
t
s
,
ba
ggi
ng,
vot
i
ng,
s
t
a
c
ki
ng
-
s
c
i
ki
t
-
l
e
a
r
n
1.4.0
doc
um
e
nt
a
t
i
on
,
”
Sc
i
k
i
t
L
e
ar
n
,
A
c
c
e
s
s
e
d:
J
a
n. 25, 2024. [
O
nl
i
ne
]
. A
va
i
l
a
bl
e
:
ht
t
ps
:
/
/
s
c
i
ki
t
-
l
e
a
r
n.or
g/
s
t
a
bl
e
/
m
odul
e
s
/
e
ns
e
m
bl
e
.ht
m
l
#a
da
boo
s
t
[
24]
G
.
K
e
e
t
al
.,
“
L
i
ght
G
B
M
:
a
hi
ghl
y
e
f
f
i
c
i
e
nt
gr
a
di
e
nt
boos
t
i
ng
de
c
i
s
i
on
t
r
e
e
,
”
i
n
A
dv
anc
e
s
i
n
N
e
ur
al
I
nf
or
m
at
i
on
P
r
o
c
e
s
s
i
ng
Sy
s
t
e
m
s
,
C
ur
r
a
n A
s
s
oc
i
a
t
e
s
I
nc
.,
pp. 1
-
9,
2017
.
[
25]
Y
.
J
i
a
ng,
H
.
L
i
,
G
.
Y
a
ng,
C
.
Z
ha
ng,
a
nd
K
.
Z
ha
o,
“
M
a
c
hi
ne
l
e
a
r
ni
ng
-
dr
i
ve
n
ont
ol
ogi
c
a
l
know
l
e
dge
ba
s
e
f
or
br
i
dge
c
or
r
os
i
on
e
va
l
ua
t
i
on
,”
I
E
E
E
A
c
c
e
s
s
, vol
. 11, pp. 144735
–
144746, 2023, doi
:
10.1109/
A
C
C
E
S
S
.2023.3344320.
[
26]
D
.
C
hi
c
c
o,
M
.
J
.
W
a
r
r
e
ns
,
a
nd
G
.
J
ur
m
a
n,
“
T
he
c
oe
f
f
i
c
i
e
nt
of
de
t
e
r
m
i
na
t
i
on
R
-
s
qua
r
e
d
i
s
m
or
e
i
nf
or
m
a
t
i
ve
t
ha
n
S
M
A
P
E
,
M
A
E
,
M
A
P
E
, M
S
E
a
nd R
M
S
E
i
n r
e
gr
e
s
s
i
on a
na
l
ys
i
s
e
va
l
ua
t
i
on,”
P
e
e
r
J
C
om
put
e
r
S
c
i
e
nc
e
, vol
. 7, J
ul
. 2021, doi
:
10.7717/
pe
e
r
j
-
c
s
.623.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
,
V
ol
.
14
, N
o.
3
,
J
une
20
25
:
1910
-
1918
1918
B
I
O
G
R
A
P
H
I
E
S
O
F
A
U
T
H
O
R
S
Ahmed
El
Youssefi
e
arned
his
Ph
.
D
.
in
software
engineering
and
artificial
intelligence
in
2025.
He
is
a
computer
science
inspector
working
at
th
e
Ministry
of
Education,
Morocco,
since
2020.
His
researc
h
focuses
on
technical
analysis
,
optimized
labeling
of
financial dat
a and artifici
al intell
igence applied
to cryptocu
rrency price forecasti
ng. He activel
y
contribu
tes to
the scien
tific com
munity
as a revi
ewer and serv
es on t
he scient
ific comm
ittees o
f
various international confere
nces. He can be c
ontacted at email: ah.elyoussefi@edu.umi.ac.ma.
Abdelaaziz
Hessane
is
an
Assistant
Professor
at
the
Faculty
of
Sci
ences,
Moulay
Ismail
University
of
Meknès,
Morocco.
He
earned
his
Ph.D.
in
science
and
techniques
in
2024
from
the
Faculty
of
Sciences
and
Techniques
of
Errachidia,
specializ
ing
in
computer
science,
software
engineering,
and
artificial
intelli
gence.
He
also
holds
an
M
.
S
.
in
business
intelligence
and
image
processin
g
from
the
same
institution
(2020)
and
has
exper
ience
teaching
computer
science
at
the
high
school
level.
His
researc
h
focuses
on
artificial
int
elligence
applications
in
precision
agriculture.
He
actively
contr
ibutes
to
the
scientific
com
munity
as
a
reviewer
for
several
esteemed
journals
and
serves
on
the
scientific
committees
of
various
international
conferences. H
e can be con
tacted at
email:
a.hessane@
edu.umi
.ac.ma.
Imad
Zeroual
is
currently
an
Associate
Professor
in
the
Departme
nt
of
Computer
Scienc
e, Fac
ulty of
Scienc
es an
d Tec
hnics,
Moulay
Ismail
Unive
rsity.
He re
ceive
d his Ph.
D. in
computer
science
from
Mohamed
First
University
in
2018.
He
is
al
so
a
member
of
severa
l
internationa
l
and
national
scientific
communities
such
as
the
Inter
national
Association
for
Educators
and
Researchers
(IAER/171101),
London,
UK,
the
Inte
rnational
Association
of
Engineers
(IAENG/206013),
and
the
Arabic
Language
Engineering
S
ociety
(ALESM),
Rabat,
Morocco.
His
areas
of
researc
h
are
artificial
intell
igence
and
data
scie
nce.
He
primarily
works
on
natural
language
processing,
machine
learning,
information
retrieval/extraction,
and
language
teaching a
nd learnin
g. He ca
n be conta
cted at e
mail: mr.imadine
@
gmail.com.
Yousef
Farhaoui
is
a
Professor
at
Moulay
Ismail
University
of
M
eknès,
Faculty
of
Sciences
and
Techniques,
Morocco.
Local
publishing
and
research
coordinator,
Cambridge
International
Academics
in
United
Kingdom.
He
obtained
his
Ph.
D.
degree
in
computer
security
from
Ibn
Zohr
University,
Faculty
of
Science.
His
researc
h
i
nterests
include
learning,
e
-
learning,
computer
security,
big
data
analytics,
and
business
intelligence
.
He
has
three
books
in
computer
science.
He
is
a
coordinator
and
member
of
the
orga
nizing
committee
and
a
member
of
the
scientific
committee
of
several
international
congres
ses
and
is
a
m
ember
of
various
international
associations.
He
has
authored
10
books
and
m
any
book
chapters
with
reputed
publishers
such
as
Springer
and
IGI.
He
served
as
a
reviewer
for
IEEE,
IET,
Springer,
Inderscience
,
and
Elsevier
journals.
He
is
also
the
guest
editor
of
m
any
journals
with
Wiley,
Springe
r, an
d Inde
rscie
nce.
He
has b
een th
e
gene
ral c
hair,
sessio
n cha
i
r, an
d pane
list in
seve
ral
conferences.
He
is
senior
member
of
IEEE,
IET,
ACM
,
and
EAI
Res
earch
Group.
He
can
be
contacted
at email
:
y.farhaoui@fste.umi.ac.ma.
Evaluation Warning : The document was created with Spire.PDF for Python.