I
AE
S In
t
er
na
t
io
na
l J
o
urna
l o
f
Art
if
icia
l In
t
ellig
ence
(
I
J
-
AI
)
Vo
l.
1
5
,
No
.
4
,
A
u
g
u
s
t 2
0
2
6
,
p
p
.
3
5
2
8
~
3
5
3
6
I
SS
N:
2
2
5
2
-
8
9
3
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijai.v
15
.i
4
.
p
p
3
5
2
8
-
3
5
3
6
3528
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
i
.
ia
esco
r
e.
co
m
A ma
chine l
ea
rni
ng
f
ra
mewo
rk
for
predic
ting a
nd o
ptimizing
return
on inv
estm
ent
a
cros
s ma
rke
t
ing
channels
Cha
nd
ra
Cha
t
hu
ra
,
K
ee
rt
ha
n Sa
y
a
,
Sa
t
his
hk
um
a
r
M
a
ni
D
e
p
a
r
t
me
n
t
o
f
C
o
mp
u
t
e
r
S
c
i
e
n
c
e
a
n
d
En
g
i
n
e
e
r
i
n
g
,
G
I
TA
M
S
c
h
o
o
l
o
f
T
e
c
h
n
o
l
o
g
y
,
G
I
TA
M
U
n
i
v
e
r
s
i
t
y
,
B
e
n
g
a
l
u
r
u
,
I
n
d
i
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Ma
y
2
,
2
0
2
5
R
ev
is
ed
J
u
n
1
7
,
2
0
2
6
Acc
ep
ted
J
u
l 9
,
2
0
2
6
In
th
e
c
u
rre
n
t
fa
st
-
p
a
c
e
d
c
o
m
p
e
t
it
iv
e
m
a
rk
e
ti
n
g
e
n
v
iro
n
m
e
n
t,
fir
m
s
re
q
u
ire
d
a
ta
-
c
e
n
tri
c
m
e
th
o
d
s
to
m
a
x
imiz
e
th
e
ir
i
n
v
e
stm
e
n
ts
i
n
se
v
e
ra
l
a
v
e
n
u
e
s.
T
h
i
s
re
se
a
rc
h
wo
rk
a
p
p
li
e
s
m
a
c
h
in
e
le
a
rn
in
g
tec
h
n
iq
u
e
s
to
e
stim
a
te
th
e
re
tu
rn
o
n
in
v
e
stm
e
n
t
(ROI)
fo
r
m
a
rk
e
ti
n
g
c
o
sts,
w
h
ich
h
e
lp
s
o
r
g
a
n
i
z
a
ti
o
n
s
i
n
b
u
d
g
e
ti
n
g
m
o
re
e
ffe
c
ti
v
e
ly
.
F
o
u
r
m
o
d
e
ls
i
n
c
lu
d
in
g
ra
n
d
o
m
f
o
re
st
,
e
x
trem
e
g
ra
d
ien
t
b
o
o
sti
n
g
(
XG
Bo
o
st
)
,
g
r
a
d
ien
t
b
o
o
stin
g
,
a
n
d
li
n
e
a
r
re
g
re
ss
io
n
we
re
u
ti
li
z
e
d
f
o
r
th
e
ir
a
c
c
u
ra
c
y
in
m
a
k
in
g
p
re
d
ict
io
n
s
.
Re
su
lt
s
sh
o
we
d
th
a
t
t
h
e
h
ig
h
e
st ac
c
u
ra
c
y
wa
s
a
c
h
iev
e
d
b
y
li
n
e
a
r
re
g
re
ss
io
n
a
t
9
9
.
3
9
%
,
ra
n
d
o
m
fo
re
st
a
t
9
9
.
0
7
%
,
g
ra
d
ien
t
b
o
o
sti
n
g
a
t
9
9
.
0
1
%
,
a
n
d
XG
Bo
o
st
a
t
9
8
.
8
1
%
.
I
t
wa
s
fu
rth
e
r
n
o
ted
t
h
a
t
d
i
g
it
a
l
m
a
rk
e
ti
n
g
a
v
e
n
u
e
s
su
c
h
a
s
so
c
ial
m
e
d
ia
a
n
d
o
n
li
n
e
sto
re
s
g
a
v
e
t
h
e
h
ig
h
e
st
ROI,
in
d
ica
ti
n
g
t
h
a
t
c
o
m
p
a
n
ies
s
h
o
u
ld
p
ri
o
rit
iz
e
d
ig
it
a
l
m
a
rk
e
ti
n
g
m
o
re
t
h
a
n
trad
i
ti
o
n
a
l
m
a
rk
e
ti
n
g
.
O
n
a
p
ra
c
ti
c
a
l
lev
e
l,
th
i
s
a
p
p
ro
a
c
h
h
e
lp
s
m
a
rk
e
ti
n
g
tea
m
f
o
r
c
h
o
o
sin
g
h
ig
h
p
e
rfo
rm
in
g
c
h
a
n
n
e
ls
sin
c
e
it
e
stim
a
tes
e
x
p
e
c
ted
re
t
u
rn
s
f
ro
m
e
a
c
h
m
a
rk
e
ti
n
g
c
h
a
n
n
e
ls
a
n
d
m
a
k
e
sm
a
rter
b
u
d
g
e
t
a
ll
o
c
a
ti
o
n
.
Y
e
t,
t
h
e
stu
d
y
is
d
o
n
e
b
y
u
sin
g
Ka
g
g
le
d
a
tas
e
t
.
I
n
o
rd
e
r
t
o
imp
ro
v
e
it
s
g
e
n
e
ra
li
t
y
,
fu
tu
re
re
se
a
rc
h
m
a
y
u
se
larg
e
r
r
e
a
l
-
wo
rld
d
a
tas
e
ts an
d
e
x
ten
si
v
e
v
is
u
a
li
z
a
ti
o
n
tec
h
n
iq
u
e
s.
K
ey
w
o
r
d
s
:
B
u
s
in
ess
in
tell
ig
en
ce
Data
-
d
r
iv
en
d
ec
is
io
n
m
ak
in
g
Ma
ch
in
e
lear
n
in
g
Ma
r
k
etin
g
ch
a
n
n
el
o
p
tim
izatio
n
R
etu
r
n
o
n
in
v
estme
n
t
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
Sath
is
h
k
u
m
ar
Ma
n
i
Dep
ar
tm
en
t o
f
C
o
m
p
u
ter
Scie
n
ce
an
d
E
n
g
in
ee
r
in
g
,
GI
T
AM
Sch
o
o
l o
f
T
ec
h
n
o
lo
g
y
,
GI
T
A
M
Un
iv
er
s
ity
B
en
g
alu
r
u
,
I
n
d
ia
E
m
ail:
s
ath
is
h
k
u
m
ar
m
an
i1
7
@
g
m
ail.
co
m
1.
I
NT
RO
D
UCT
I
O
N
Ma
r
k
etin
g
h
as
u
n
d
er
g
o
n
e
h
u
g
e
ch
a
n
g
es,
m
o
v
in
g
f
r
o
m
tr
a
d
itio
n
al
m
ar
k
etin
g
lik
e
telev
i
s
io
n
(
T
V)
,
b
illb
o
ar
d
s
,
a
n
d
n
ewsp
ap
er
s
to
m
o
d
er
n
m
a
r
k
etin
g
lik
e
s
o
cia
l
m
ed
ia,
o
n
lin
e
s
to
r
es
,
an
d
m
o
b
ile
ap
p
licatio
n
s
.
T
ec
h
n
o
lo
g
y
d
ev
elo
p
m
en
t
is
o
n
e
o
f
th
e
m
ain
r
ea
s
o
n
s
b
eh
in
d
th
is
an
d
also
ch
a
n
g
e
i
n
cu
s
to
m
er
b
e
h
av
io
r
.
C
o
m
p
an
ies
d
esire
to
r
ea
ch
co
n
s
u
m
er
s
ef
f
icien
tly
an
d
g
r
o
w
th
eir
b
u
s
in
ess
wh
ile
ad
ap
ti
n
g
to
tech
n
o
l
o
g
ical
ch
an
g
es
also
ad
d
s
to
th
is
.
Ma
r
k
etin
g
is
o
n
e
o
f
th
e
m
ain
asp
ec
ts
wh
ich
b
r
in
g
s
co
n
s
u
m
er
s
clo
s
e
to
co
m
p
a
n
ies
p
r
o
d
u
cts
an
d
d
r
iv
es
s
ales
wh
i
ch
r
esu
lts
in
p
r
o
f
its
.
T
h
e
b
ig
g
est
ch
allen
g
e
f
o
r
c
o
m
p
a
n
ies
is
to
id
e
n
tify
t
h
e
r
i
g
h
t
m
ar
k
etin
g
ch
an
n
el
to
ac
h
ie
v
e
h
ig
h
r
et
u
r
n
o
n
in
v
estme
n
t
(
R
OI
)
,
a
n
d
it
is
im
p
o
r
tan
t
to
u
n
d
er
s
tan
d
wh
at
tr
ad
itio
n
al
m
ar
k
etin
g
an
d
m
o
d
er
n
m
ar
k
etin
g
ar
e
b
r
in
g
in
g
t
o
th
e
tab
le
f
o
r
co
m
p
a
n
ies.
I
n
o
r
d
er
to
co
v
er
m
ass
au
d
ien
ce
s
in
th
e
m
ar
k
et
c
o
m
p
an
ies
u
s
u
ally
g
o
f
o
r
t
r
ad
itio
n
a
l
m
ar
k
etin
g
lik
e
b
illb
o
a
r
d
s
,
T
V,
an
d
n
ewsp
ap
er
s
.
T
h
ese
ch
an
n
els
r
esu
lt
in
co
v
er
in
g
a
lar
g
e
v
o
lu
m
e
o
f
au
d
ien
c
e,
h
elp
s
in
in
cr
ea
s
in
g
b
r
an
d
r
e
co
g
n
itio
n
b
u
t
th
ese
ch
an
n
els f
ail
in
d
ir
ec
tly
im
p
ac
tin
g
tar
g
eted
co
n
s
u
m
er
s
.
Fu
r
th
er
s
h
ap
in
g
m
ar
k
etin
g
s
tr
ateg
ie
s
in
r
ea
l
-
tim
e
to
b
e
p
r
ec
is
e
p
er
s
o
n
alize
d
a
d
v
er
tis
in
g
,
s
in
ce
co
m
p
an
ies
ar
e
m
o
r
e
f
o
cu
s
ed
o
n
tailo
r
ed
ca
m
p
a
ig
n
s
ac
co
r
d
in
g
to
tar
g
eted
au
d
ien
ce
s
.
I
n
s
tead
o
f
u
s
in
g
d
ata
-
d
r
iv
e
n
in
s
ig
h
ts
m
o
s
t
o
f
th
e
b
u
s
in
ess
es
r
ely
o
n
p
ast
ex
p
er
ien
ce
s
an
d
in
d
u
s
tr
y
t
r
en
d
s
w
h
ile
allo
ca
tin
g
m
ar
k
etin
g
b
u
d
g
ets.
T
h
is
m
a
y
r
esu
lt
i
n
wastin
g
r
eso
u
r
ce
s
a
n
d
p
o
o
r
in
v
estme
n
t
ch
o
ices
wh
ich
ag
ain
r
esu
lt
in
d
ec
r
ea
s
in
g
p
r
o
f
itab
ilit
y
.
B
y
d
e
ter
m
in
in
g
th
e
m
o
s
t
ef
f
icien
t
c
h
an
n
els
to
in
v
est
in
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
A
ma
ch
in
e
lea
r
n
in
g
fr
a
mewo
r
k
fo
r
p
r
ed
ictin
g
a
n
d
o
p
timiz
in
g
r
etu
r
n
o
n
in
ve
s
tmen
t
…
(
C
h
a
n
d
r
a
C
h
a
th
u
r
a
)
3529
an
d
a
d
ju
s
tin
g
th
e
s
p
en
d
i
n
g
h
elp
s
b
u
s
in
ess
to
o
v
er
co
m
e
th
ese
ch
allen
g
es.
T
o
m
a
k
e
th
is
p
o
s
s
ib
le,
m
ar
k
etin
g
ca
m
p
aig
n
s
r
eq
u
ir
e
c
o
n
tin
u
o
u
s
ev
alu
atio
n
t
o
m
ain
tain
lo
n
g
-
t
er
m
g
r
o
wth
,
f
u
r
th
er
it
s
h
o
u
ld
h
av
e
th
e
ca
p
ac
ity
to
f
o
r
ec
ast
ac
cu
r
ately
an
d
f
in
d
R
OI
f
r
o
m
v
ar
io
u
s
m
a
r
k
etin
g
ch
an
n
els
b
y
u
s
in
g
m
ac
h
i
n
e
lear
n
in
g
.
Ma
ch
in
e
lear
n
in
g
a
n
d
p
r
ed
ictiv
e
an
aly
tics
p
r
esen
t
an
o
p
p
o
r
tu
n
ity
f
o
r
r
ev
o
lu
tio
n
izin
g
m
ar
k
etin
g
d
ec
is
io
n
-
m
ak
in
g
b
y
p
r
ed
ictin
g
p
o
ten
tial r
etu
r
n
s
u
s
in
g
p
ast d
ata.
Yah
ia
an
d
E
lB
o
lo
k
[
1
]
o
f
f
er
e
d
a
s
to
ch
asti
c
n
o
n
lin
ea
r
p
r
o
g
r
am
m
in
g
ap
p
r
o
ac
h
f
o
r
o
p
tim
iz
in
g
b
u
d
g
et
allo
ca
tio
n
f
o
r
d
ig
ital
m
ar
k
eti
n
g
ca
m
p
aig
n
s
;
h
o
wev
er
,
th
e
p
ap
er
d
i
d
n
o
t
p
r
o
v
i
d
e
a
co
m
p
ar
ativ
e
s
tu
d
y
f
o
r
d
if
f
er
en
t
lear
n
in
g
m
o
d
els
in
ter
m
s
o
f
m
ar
k
etin
g
c
h
an
n
els
b
u
t
f
o
cu
s
ed
m
o
r
e
o
n
m
ath
em
atica
l
o
p
tim
izatio
n
r
ath
er
th
an
p
r
ed
ictin
g
R
OI
u
s
in
g
m
ac
h
in
e
lear
n
in
g
m
o
d
els.
Nay
y
ar
et
a
l.
[
2
]
ex
p
lo
r
ed
d
ata
-
d
r
iv
en
m
ac
h
in
e
lear
n
in
g
tech
n
iq
u
es
to
o
p
tim
ize
ad
v
er
tis
in
g
ca
m
p
ai
g
n
s
in
m
o
d
er
n
m
ed
ia
a
n
d
f
o
cu
s
ed
o
n
ly
o
n
ca
m
p
aig
n
o
p
tim
izatio
n
s
tr
ateg
ies
a
n
d
d
i
d
n
o
t
p
r
o
v
id
e
an
an
aly
s
is
o
f
m
ac
h
in
e
lear
n
in
g
m
o
d
els
f
o
r
R
OI
p
r
ed
ictio
n
s
with
v
ar
io
u
s
m
ar
k
etin
g
ch
an
n
els.
Ma
an
d
Su
n
[
3
]
d
is
cu
s
s
ed
th
e
im
p
o
r
tan
ce
o
f
m
ac
h
in
e
lear
n
in
g
to
im
p
r
o
v
e
h
u
m
an
d
ec
is
io
n
-
m
ak
in
g
in
m
ar
k
etin
g
.
T
h
is
wo
r
k
em
p
h
asizes
m
atch
in
g
co
m
p
u
tatio
n
al
m
o
d
e
ls
with
co
n
s
u
m
er
p
s
y
ch
o
lo
g
y
an
d
ap
p
r
o
ac
h
es.
I
t
d
o
esn
'
t
in
clu
d
e
f
o
r
ec
asti
n
g
o
f
f
in
a
n
cial
p
er
f
o
r
m
a
n
ce
s
lik
e
R
OI
p
r
ed
ictio
n
o
r
in
d
icatin
g
q
u
a
n
titativ
e
m
ar
k
et
in
g
in
v
estme
n
t
o
p
tim
izatio
n
o
r
ca
m
p
aig
n
b
u
d
g
etin
g
.
Mik
lo
s
ik
an
d
E
v
a
n
s
[
4
]
p
r
o
v
id
e
an
o
v
er
v
iew
o
f
h
o
w
b
ig
d
ata
an
d
m
ac
h
i
n
e
lear
n
in
g
tech
n
iq
u
es
ar
e
ch
an
g
in
g
m
a
r
k
et
s
ec
to
r
.
B
r
ei
[
5
]
p
r
o
v
id
es
a
n
o
v
er
v
iew
o
f
m
ac
h
in
e
lear
n
in
g
s
tr
ateg
ies
in
m
a
r
k
etin
g
.
T
h
is
p
ap
e
r
d
o
es
n
o
t
f
o
cu
s
o
n
a
p
a
r
ticu
lar
ap
p
licatio
n
m
o
d
el
n
o
r
q
u
a
n
titativ
e
ass
es
s
m
en
t
o
r
alg
o
r
ith
m
co
m
p
ar
is
o
n
o
f
R
OI
.
Her
h
au
s
en
et
a
l.
[
6
]
d
is
cu
s
s
th
e
ch
an
g
in
g
lan
d
s
ca
p
e
o
f
m
a
ch
in
e
lear
n
in
g
f
o
r
m
a
r
k
etin
g
ap
p
licatio
n
s
.
T
h
is
wo
r
k
r
eq
u
i
r
es
m
o
r
e
em
p
ir
ical
s
tu
d
ies with
co
m
p
ar
is
o
n
s
o
f
m
ac
h
in
e
lear
n
in
g
m
o
d
els.
C
h
en
[
7
]
d
is
cu
s
s
es
m
ac
h
in
e
l
ea
r
n
in
g
f
o
r
e
-
co
m
m
er
ce
s
tr
at
eg
y
,
b
u
t
it
d
o
esn
'
t
co
n
ce
n
tr
at
e
o
n
R
OI
f
o
r
ec
asts
o
r
m
u
lti
-
ch
a
n
n
el
m
ar
k
etin
g
ef
f
ec
tiv
en
ess
,
wh
ich
r
estricts
its
u
s
ag
e
in
s
tr
ateg
ic
b
u
d
g
et
p
lan
n
i
n
g
.
No
r
d
in
an
d
R
av
ald
[
8
]
e
x
am
in
e
d
ec
is
io
n
-
m
ak
i
n
g
in
ch
an
g
in
g
m
ar
k
etin
g
e
n
v
ir
o
n
m
en
ts
.
Ng
ai
an
d
W
u
[
9
]
d
em
o
n
s
tr
ated
a
f
r
am
ewo
r
k
f
o
r
u
n
d
er
s
tan
d
in
g
m
ac
h
in
e
lear
n
in
g
th
at
ca
n
au
g
m
en
t
m
a
r
k
et
in
g
ef
f
icie
n
cy
.
T
h
is
m
o
d
el
o
u
tlin
es
m
an
y
m
ac
h
in
e
lear
n
in
g
tech
n
iq
u
es
f
o
r
tar
g
etin
g
an
d
p
e
r
s
o
n
aliza
tio
n
.
Mik
lo
s
ik
et
a
l.
[
1
0
]
ex
p
lo
r
ed
th
e
u
s
ag
e
o
f
m
ac
h
in
e
lear
n
in
g
to
o
ls
.
T
h
e
R
OI
p
r
e
d
ictiv
e
m
o
d
el
is
n
o
t
in
clu
d
ed
in
th
is
p
ap
er
,
an
d
to
o
ls
ar
e
n
o
t
test
ed
a
g
ain
s
t
q
u
an
titativ
e
p
er
f
o
r
m
an
ce
m
ea
s
u
r
es.
Vash
is
h
th
et
a
l.
[
1
1
]
s
h
ar
e
tr
en
d
s
in
a
r
tific
ial
in
tellig
en
ce
ad
o
p
tio
n
.
An
d
ay
a
n
i
et
a
l.
[
1
2
]
ex
am
in
e
a
p
p
licatio
n
o
f
b
ig
d
ata
b
u
t
d
id
n
'
t
d
is
cu
s
s
ch
an
n
el
s
p
ec
if
ic
R
OI
f
o
r
ec
asti
n
g
.
Sab
a
et
a
l.
[
1
3
]
s
u
r
v
e
y
s
m
ac
h
in
e
lear
n
in
g
ap
p
licatio
n
s
in
m
ar
k
etin
g
tr
an
s
f
o
r
m
atio
n
.
T
h
is
p
ap
er
lack
s
in
elab
o
r
atin
g
m
o
d
el
m
eth
o
d
s
,
ev
alu
atio
n
m
etr
ics
an
d
p
r
ac
tical
im
p
lem
en
tatio
n
f
o
r
R
OI
esti
m
atio
n
.
Ku
m
ar
[
1
4
]
d
escr
ib
ed
a
m
ar
k
etin
g
s
u
p
p
o
r
t
h
y
b
r
id
m
o
d
el,
wh
ich
co
n
ce
n
tr
at
es
o
n
d
ata
m
in
in
g
r
ath
er
th
an
f
in
an
cial
r
etu
r
n
p
r
ed
ictio
n
an
d
lack
s
in
m
ar
k
et
in
g
ch
an
n
el
b
r
ea
k
d
o
wn
s
o
r
i
n
v
estme
n
t
an
aly
s
is
.
Z
ag
h
lo
u
l
et
a
l.
[
1
5
]
r
esear
ch
i
n
v
o
lv
es
a
co
m
p
a
r
is
o
n
o
f
tr
ad
i
tio
n
al
an
d
d
ee
p
m
et
h
o
d
s
in
p
r
ed
ictin
g
co
n
s
u
m
er
s
atis
f
ac
tio
n
in
an
e
-
co
m
m
er
c
e
co
n
tex
t,
alth
o
u
g
h
th
ese
r
esear
ch
er
s
s
p
ec
if
ically
ex
am
in
e
r
esu
lts
r
eg
ar
d
in
g
cu
s
to
m
er
ex
p
er
ie
n
ce
an
d
n
o
t
R
OI
p
r
ed
ictio
n
o
r
b
u
d
g
et
d
is
tr
ib
u
tio
n
in
a
v
a
r
iety
o
f
o
n
lin
e
m
ar
k
etin
g
ch
a
n
n
els.
Sad
r
n
ia
[
1
6
]
f
o
cu
s
es
o
n
p
r
e
d
ictiv
e
m
o
d
ellin
g
f
o
r
ca
m
p
a
ig
n
s
tr
ateg
y
.
G
o
o
ljar
et
a
l.
[
1
7
]
s
y
s
tem
atica
lly
r
ev
iewe
d
th
e
u
s
e
o
f
p
r
e
d
ictiv
e
m
o
d
els
b
ased
o
n
s
en
tim
en
t
s
f
o
r
p
u
r
ch
ases
co
n
d
u
cted
o
n
lin
e,
in
r
elatio
n
to
“M
ar
k
etin
g
5
.
0
”
s
p
ec
if
ically
,
alth
o
u
g
h
th
ei
r
r
esear
ch
is
ce
n
ter
ed
o
n
t
h
e
an
aly
s
is
o
f
c
o
n
s
u
m
er
s
en
tim
en
ts
r
ath
er
th
an
d
ir
ec
tly
r
eg
a
r
d
in
g
th
e
p
r
ed
ictio
n
in
ter
m
s
o
f
R
OI
r
etu
r
n
o
r
b
u
d
g
et
d
is
tr
ib
u
tio
n
i
n
v
ar
io
u
s
av
ailab
le
o
n
lin
e
m
a
r
k
etin
g
ch
a
n
n
els.
Sa
n
g
s
awa
n
g
[
1
8
]
d
is
cu
s
s
ed
wo
r
k
o
n
click
-
th
r
o
u
g
h
r
ates
(
C
T
R
)
p
r
ed
ictio
n
r
ath
e
r
th
an
m
u
lti
-
ch
a
n
n
el
in
v
estme
n
t a
n
aly
s
is
.
2.
M
E
T
H
O
D
T
h
e
m
eth
o
d
o
l
o
g
y
p
lay
s
a
s
ig
n
if
ican
t
r
o
le
in
p
r
o
m
o
tin
g
u
s
e
r
ex
p
er
ien
ce
an
d
d
ec
is
io
n
m
a
k
in
g
.
T
h
is
p
r
ed
ictiv
e
m
o
d
el
s
tar
ts
with
th
e
ch
an
n
els
th
at
th
e
u
s
er
ca
n
in
v
est
in
,
an
d
u
s
er
s
s
elec
t
o
n
e
o
r
m
o
r
e
ch
a
n
n
els.
T
h
en
th
e
u
s
er
in
p
u
ts
th
e
am
o
u
n
t to
in
v
est in
th
e
m
ar
k
etin
g
c
h
an
n
els.
T
h
e
m
o
d
el
th
en
p
r
e
d
i
cts th
e
R
OI
an
d
th
e
s
y
s
tem
p
r
o
d
u
ce
s
a
p
ie
c
h
ar
t
w
h
ich
r
ep
r
esen
ts
th
e
co
n
tr
ib
u
tio
n
o
f
ea
ch
m
ar
k
etin
g
ch
a
n
n
el
t
o
o
v
e
r
all
R
OI
.
T
h
is
en
ab
les
b
u
s
in
ess
es
to
m
ak
e
in
f
o
r
m
ed
d
ec
is
io
n
s
o
n
th
e
ch
a
n
n
els
th
at
th
ey
ca
n
in
v
est
in
an
d
av
o
id
th
e
c
h
an
n
el
wh
er
e
th
e
R
OI
will
b
e
lo
w,
h
en
ce
s
av
in
g
o
n
c
o
s
ts
.
F
ig
u
r
e
1
s
h
o
ws
th
e
s
y
s
tem
atic
wo
r
k
f
lo
w
o
f
th
e
p
r
ed
ictiv
e
m
ac
h
in
e
lear
n
in
g
m
o
d
el
d
e
v
elo
p
ed
f
o
r
m
ar
k
etin
g
R
OI
o
p
tim
izatio
n
.
T
h
e
p
r
o
ce
s
s
b
eg
in
s
with
th
e
d
ataset
lo
ad
in
g
.
I
t
co
n
tain
s
h
is
to
r
ical
m
ar
k
etin
g
p
er
f
o
r
m
a
n
ce
d
at
a,
s
u
ch
as
ex
p
en
s
es
o
n
ca
m
p
aig
n
s
,
m
etr
ics
o
f
cu
s
to
m
er
en
g
ag
em
en
t,
an
d
r
e
v
en
u
e
g
en
er
ate
d
.
T
h
is
in
f
o
r
m
atio
n
f
o
r
m
s
th
e
f
o
u
n
d
atio
n
f
o
r
tr
ain
in
g
an
d
test
in
g
m
ac
h
in
e
lear
n
in
g
m
o
d
els.
T
h
en
th
e
d
ataset
is
p
r
ep
r
o
ce
s
s
ed
,
wh
ich
in
cl
u
d
es
clea
n
in
g
an
d
tr
an
s
f
o
r
m
atio
n
o
f
th
e
r
aw
d
ataset.
Mis
s
in
g
v
alu
es
ar
e
h
an
d
led
,
n
u
m
e
r
ical
attr
ib
u
tes
ar
e
n
o
r
m
alize
d
,
ca
teg
o
r
ical
v
ar
iab
les
ar
e
en
co
d
ed
,
an
d
k
e
y
p
e
r
f
o
r
m
an
ce
in
d
icato
r
s
ar
e
d
er
i
v
ed
.
T
h
en
f
ea
t
u
r
e
e
x
tr
ac
tio
n
is
d
o
n
e,
b
ased
o
n
p
r
o
f
it
p
er
f
o
r
m
an
ce
a
n
d
ca
m
p
aig
n
an
aly
s
is
.
T
h
e
d
ataset
is
d
i
v
id
ed
in
to
test
in
g
d
ata
an
d
t
r
ain
in
g
d
ata
af
ter
p
r
ep
r
o
ce
s
s
in
g
.
T
h
e
tr
ain
i
n
g
s
a
m
p
le
is
u
s
ed
to
b
u
ild
m
ac
h
in
e
lear
n
in
g
m
o
d
els,
an
d
th
e
test
s
am
p
le
is
k
ep
t
f
o
r
later
to
m
ea
s
u
r
e
th
e
ac
c
u
r
ac
y
an
d
g
e
n
er
aliza
tio
n
o
f
th
e
m
o
d
els.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
1
5
,
No
.
4
,
Au
g
u
s
t 2
0
2
6
:
3
5
2
8
-
3
5
3
6
3530
T
o
p
r
ed
ict
R
OI
f
r
o
m
m
ar
k
e
tin
g
ex
p
en
s
es,
th
e
s
tu
d
y
r
elies
o
n
alg
o
r
ith
m
s
lik
e
r
an
d
o
m
f
o
r
est
r
eg
r
ess
o
r
[
1
9
]
,
[
2
0
]
,
g
r
ad
ien
t
b
o
o
s
tin
g
[
2
1
]
,
e
x
tr
em
e
g
r
ad
ien
t
b
o
o
s
tin
g
(
XGBo
o
s
t)
[
2
2
]
,
[
2
3
]
,
an
d
lin
ea
r
r
eg
r
ess
io
n
[
2
4
]
,
[
2
5
]
.
T
h
ese
m
o
d
els
p
ick
u
p
o
n
p
atter
n
s
an
d
r
elatio
n
s
h
ip
s
b
etwe
en
m
ar
k
etin
g
co
s
ts
an
d
p
r
o
f
its
,
h
elp
in
g
th
e
s
y
s
tem
m
ak
e
k
n
o
wled
g
ea
b
le
d
ec
is
io
n
s
b
ased
o
n
t
h
e
n
u
m
b
er
s
.
Up
o
n
co
m
p
letio
n
o
f
th
e
tr
ain
in
g
p
h
ase,
th
e
n
e
x
t
s
tep
is
th
e
g
en
er
atio
n
o
f
th
e
R
OI
p
r
ed
ictio
n
.
I
n
th
is
p
h
ase,
r
etu
r
n
esti
m
ates
f
o
r
ea
ch
m
ar
k
etin
g
ch
an
n
el
ar
e
cr
ea
ted
b
ased
o
n
th
e
u
s
er
’
s
in
p
u
t
r
eg
ar
d
in
g
th
e
am
o
u
n
t
to
in
v
est
in
ea
ch
o
f
th
em
.
T
h
is
p
h
ase
aim
s
to
p
r
o
v
id
e
b
u
s
in
es
s
es
with
in
s
ig
h
ts
in
to
p
r
o
s
p
ec
t
iv
e
r
etu
r
n
s
b
ef
o
r
e
co
m
m
itti
n
g
r
eso
u
r
ce
s
.
Fu
r
th
e
r
,
th
e
ac
cu
r
ac
y
e
v
alu
atio
n
p
h
ase
is
in
co
r
p
o
r
ated
to
en
s
u
r
e
th
e
r
eliab
ilit
y
o
f
th
e
s
y
s
tem
.
Fig
u
r
e
1
.
Ma
ch
i
n
e
lear
n
in
g
f
r
a
m
ewo
r
k
f
o
r
m
ar
k
etin
g
R
OI
p
r
ed
ictio
n
2
.
1
.
E
x
perim
ent
a
l set
up
T
h
e
m
o
d
er
n
to
o
ls
an
d
lan
g
u
a
g
es
lik
e
Py
th
o
n
,
Pan
d
as
,
an
d
Ma
tp
lo
tlib
ar
e
ad
o
p
ted
f
o
r
th
e
p
r
o
p
o
s
ed
wo
r
k
f
o
r
d
ata
m
an
ip
u
latio
n
a
n
d
d
ata
v
is
u
aliza
tio
n
.
Fu
r
t
h
er
s
cik
it
-
lear
n
f
o
r
m
ac
h
in
e
lear
n
in
g
m
o
d
el
wh
ic
h
in
clu
d
es
R
-
s
q
u
ar
ed
(
R
2
)
s
co
r
e
,
m
ea
n
s
q
u
ar
e
d
er
r
o
r
(
MSE
)
f
o
r
m
o
d
el
ev
alu
atio
n
.
T
h
ese
lib
r
ar
ies
p
r
o
v
id
e
th
e
f
o
u
n
d
atio
n
al
to
o
ls
n
ee
d
ed
f
o
r
d
ata
p
r
o
ce
s
s
in
g
,
v
is
u
aliza
tio
n
,
an
d
m
ac
h
in
e
lear
n
in
g
m
o
d
el
d
ev
elo
p
m
e
n
t.
Data
p
r
ep
r
o
ce
s
s
in
g
is
ca
r
r
ied
o
u
t
b
y
en
s
u
r
in
g
th
er
e
ar
e
n
o
m
is
s
in
g
v
alu
es.
C
o
lu
m
n
m
ea
n
s
ar
e
u
s
ed
to
r
ep
lace
m
is
s
in
g
n
u
m
er
ical
v
alu
es
to
a
v
o
id
lo
s
in
g
th
e
d
ata
an
d
i
n
f
u
r
th
er
it
en
s
u
r
es
co
n
s
is
ten
cy
.
T
h
e
p
r
o
f
it
m
ar
g
in
%
is
co
m
p
u
ted
,
wh
ich
is
o
n
e
o
f
th
e
m
o
s
t
im
p
o
r
ta
n
t
f
in
an
cial
m
etr
ics
f
o
r
ass
ess
in
g
p
r
o
f
ita
b
ilit
y
f
o
r
d
if
f
er
en
t
m
ar
k
etin
g
ca
m
p
aig
n
s
.
Selecti
o
n
o
f
th
e
f
ea
t
u
r
e
s
et
p
lay
s
a
n
im
p
o
r
ta
n
t
r
o
le
i
n
th
is
wo
r
k
,
s
o
th
at
v
a
r
iab
les
wh
ich
ar
e
r
elev
an
t
an
d
s
ig
n
i
f
ican
t
in
p
r
ed
ictin
g
R
OI
ar
e
id
en
tifie
d
.
An
o
p
en
-
s
o
u
r
ce
d
is
tr
ib
u
tio
n
ca
lled
an
ac
o
n
d
a
is
u
s
ed
in
th
is
wo
r
k
to
m
an
ag
e
lib
r
ar
ies
r
eq
u
i
r
ed
f
o
r
m
ac
h
in
e
lear
n
in
g
m
o
d
els
an
d
p
y
t
h
o
n
is
wr
itten
an
d
ex
ec
u
ted
u
s
in
g
th
e
o
p
en
-
s
o
u
r
ce
en
v
ir
o
n
m
e
n
t
ca
lled
J
u
p
y
ter
No
teb
o
o
k
.
T
h
is
wo
r
k
w
as
s
im
u
lated
u
s
in
g
16
GB
R
AM
,
I
n
tel
C
o
r
e
i5
p
r
o
ce
s
s
o
r
,
an
d
1
2
0
GB
o
f
ha
r
d
d
is
k
s
p
ac
e
in
W
in
d
o
ws 1
0
o
p
er
atin
g
en
v
ir
o
n
m
en
t.
2
.
2
.
Da
t
a
s
o
urce
T
h
e
d
ataset
f
o
r
th
e
an
aly
s
is
o
f
m
ar
k
etin
g
c
h
an
n
els
an
d
p
r
e
d
ictin
g
R
OI
is
tak
en
f
r
o
m
Ka
g
g
le
.
T
h
e
d
ataset
co
n
tain
s
s
o
m
e
s
ig
n
if
ican
t
p
ar
am
eter
s
lik
e
m
o
n
e
y
in
v
ested
,
p
r
o
d
u
ct
ty
p
e
,
a
n
d
t
ar
g
eted
p
eo
p
le
f
o
r
s
p
ec
if
ic
m
ar
k
etin
g
ch
a
n
n
els.
On
ly
th
e
p
ar
am
eter
v
ar
iab
les
an
d
o
u
tp
u
t
v
ar
iab
les
ar
e
f
e
d
in
to
th
e
s
y
s
tem
to
av
o
id
lo
g
is
tical
an
d
p
r
iv
ac
y
co
n
ce
r
n
s
.
T
h
e
d
ataset
h
as
6
0
,
0
0
0
r
o
ws
an
d
1
2
c
o
lu
m
n
s
f
o
r
6
d
if
f
er
en
t
m
ar
k
etin
g
ch
an
n
els
an
d
th
e
m
a
r
k
etin
g
c
h
an
n
els
ar
e
s
o
cial
m
e
d
ia,
T
V,
m
o
b
ile
a
p
p
licatio
n
s
,
b
illb
o
ar
d
,
n
ewsp
ap
er
s
,
an
d
o
n
lin
e
s
to
r
e.
2
.
3
.
Alg
o
rit
hm
s
a
nd
perf
o
r
m
a
nce
ev
a
lua
t
i
o
n m
et
rics
Ma
ch
in
e
lear
n
i
n
g
a
n
d
p
r
ed
ic
tiv
e
an
aly
tics
p
r
esen
t
an
o
p
p
o
r
tu
n
ity
f
o
r
r
e
v
o
lu
tio
n
izin
g
m
ar
k
etin
g
d
ec
is
io
n
-
m
ak
in
g
b
y
p
r
ed
ictin
g
p
o
te
n
tial
r
etu
r
n
s
u
s
in
g
p
ast
d
ata.
T
h
is
wo
r
k
aim
s
at
cr
ea
tin
g
a
p
r
ed
ictiv
e
m
ac
h
in
e
lear
n
i
n
g
m
o
d
el
th
at
p
r
ed
icts
m
ar
k
etin
g
R
OI
an
d
h
elp
s
co
m
p
a
n
ies
to
m
ax
im
i
ze
th
eir
m
ar
k
etin
g
ef
f
o
r
ts
.
T
h
is
m
o
d
el
u
s
es
m
ac
h
in
e
lear
n
in
g
m
eth
o
d
s
,
f
o
r
in
s
tan
ce
,
r
an
d
o
m
f
o
r
est
r
eg
r
ess
o
r
,
g
r
ad
ien
t
b
o
o
s
tin
g
,
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
A
ma
ch
in
e
lea
r
n
in
g
fr
a
mewo
r
k
fo
r
p
r
ed
ictin
g
a
n
d
o
p
timiz
in
g
r
etu
r
n
o
n
in
ve
s
tmen
t
…
(
C
h
a
n
d
r
a
C
h
a
th
u
r
a
)
3531
XGBo
o
s
t,
an
d
lin
ea
r
r
eg
r
e
s
s
io
n
to
p
r
ed
ict
R
OI
ac
cu
r
ately
an
d
co
n
s
is
ten
tly
.
T
h
is
r
esear
ch
wo
r
k
is
in
ten
d
ed
to
cr
ea
te
a
b
r
id
g
e
b
etwe
en
tr
ad
it
io
n
al
m
ar
k
etin
g
an
d
d
ig
ital
m
ar
k
etin
g
b
y
p
r
o
v
id
in
g
a
d
ee
p
an
aly
s
is
o
f
ea
c
h
o
f
th
em
.
Dig
ital
m
ar
k
etin
g
allo
ws
f
o
r
ad
v
an
ce
d
tar
g
etin
g
an
d
tr
ac
k
in
g
,
b
u
t
tr
ad
itio
n
al
m
ar
k
eti
n
g
is
s
till
g
o
o
d
f
o
r
r
ea
ch
in
g
b
r
o
a
d
d
em
o
g
r
a
p
h
ics.
C
o
m
p
an
ies
th
at
ap
p
l
y
b
o
t
h
ap
p
r
o
ac
h
es
r
ec
eiv
e
th
e
b
est
r
esu
lts
b
ec
au
s
e
th
ey
cr
ea
te
th
e
m
o
s
t
b
alan
ce
d
m
ar
k
etin
g
m
ix
an
d
g
et
th
e
h
ig
h
est
R
OI
.
C
o
m
p
an
ies
th
at
u
s
e
m
ac
h
in
e
lear
n
in
g
f
o
r
m
ar
k
etin
g
an
aly
tics
h
a
v
e
a
co
m
p
etitiv
e
ad
v
a
n
tag
e
b
ec
au
s
e
t
h
ey
ca
n
m
a
k
e
th
eir
m
ar
k
etin
g
in
v
estme
n
t
m
u
ch
m
o
r
e
ef
f
icien
t a
n
d
r
ec
eiv
e
th
e
h
ig
h
est p
o
s
s
ib
le
r
etu
r
n
.
T
h
e
ac
cu
r
ac
y
is
ev
alu
ated
b
y
co
m
p
ar
in
g
th
e
f
o
r
ec
asted
R
OI
to
th
e
ac
tu
al
test
d
ata.
T
h
e
p
e
r
f
o
r
m
a
n
ce
in
d
icato
r
s
u
s
ed
f
o
r
ev
alu
atio
n
o
f
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
s
ac
cu
r
ac
y
ar
e
MSE
as
in
(
1
)
,
R
2
as
in
(
2
)
,
an
d
m
ea
n
ab
s
o
lu
te
er
r
o
r
(
MA
E
)
a
s
in
(
3
)
.
T
h
e
m
ath
em
atica
l
m
o
d
els
to
co
m
p
u
te
th
e
p
er
f
o
r
m
an
ce
in
d
icato
r
s
ar
e
as f
o
llo
w
s.
=
1
∑
(
−
ˆ
)
2
=
1
(
1
)
2
=
1
−
∑
(
−
ˆ
)
2
∑
(
−
̅
)
2
(
2
)
=
1
×
∑
|
−
ˆ
|
(
3
)
W
h
er
e
is
n
u
m
b
er
o
f
d
ata
p
o
i
n
ts
,
is
ac
tu
al
v
alu
e
,
ˆ
is
p
r
ed
icted
v
alu
e
,
(
−
ˆ
)
2
is
s
q
u
ar
ed
er
r
o
r
f
o
r
th
e
i
-
th
d
ata
p
o
i
n
t
,
̅
is
m
ea
n
o
f
ac
t
u
al
v
alu
es
,
an
d
Σ
is
s
u
m
m
atio
n
n
o
tatio
n
.
2
.
4
.
Da
t
a
prepro
ce
s
s
ing
Data
p
r
ep
r
o
ce
s
s
in
g
is
ess
en
tial
f
o
r
d
ata
a
n
aly
s
is
an
d
m
ac
h
in
e
lear
n
i
n
g
,
in
v
o
l
v
in
g
s
tep
s
lik
e
d
ata
co
llectio
n
,
clea
n
i
n
g
,
in
teg
r
at
io
n
,
tr
a
n
s
f
o
r
m
atio
n
,
a
n
d
r
ed
u
ctio
n
,
u
ltima
tely
lead
in
g
to
im
p
r
o
v
ed
m
o
d
el
p
er
f
o
r
m
an
ce
th
r
o
u
g
h
ca
r
ef
u
l
d
ataset
p
r
ep
ar
atio
n
.
Data
clea
n
i
n
g
in
v
o
lv
es
id
en
tif
y
in
g
an
d
r
e
ctif
y
in
g
er
r
o
r
s
an
d
in
co
n
s
is
ten
cies
in
d
ata
s
u
ch
a
s
f
illi
n
g
in
m
is
s
in
g
v
alu
es,
co
r
r
ec
tin
g
in
ac
cu
r
ac
ies,
an
d
ad
d
r
ess
in
g
o
u
tlier
s
to
en
h
an
ce
th
e
d
ataset'
s
q
u
ality
an
d
r
ea
d
in
ess
f
o
r
an
aly
s
is
o
r
m
o
d
ellin
g
.
Data
i
n
teg
r
atio
n
m
er
g
es
d
ata
f
r
o
m
d
iv
er
s
e
s
o
u
r
ce
s
in
to
a
c
o
h
e
r
en
t
d
ataset,
wh
ile
d
ata
tr
an
s
f
o
r
m
atio
n
p
r
ep
ar
es
th
at
d
at
a
f
o
r
an
al
y
s
is
b
y
co
n
v
er
tin
g
it
in
to
an
ap
p
r
o
p
r
iate
f
o
r
m
at
an
d
en
s
u
r
in
g
f
ea
tu
r
e
co
m
p
ar
ab
ilit
y
.
Data
r
ed
u
ctio
n
in
v
o
l
v
es
m
in
im
izin
g
d
ata
v
o
lu
m
e
wh
il
e
m
ain
tain
in
g
k
ey
i
n
f
o
r
m
atio
n
,
en
h
an
cin
g
c
o
m
p
u
tatio
n
al
e
f
f
icien
cy
an
d
m
o
d
el
s
im
p
licity
an
d
f
ea
t
u
r
e
s
elec
tio
n
,
wh
ich
h
elp
p
r
ev
e
n
t
o
v
er
f
itti
n
g
an
d
im
p
r
o
v
e
p
r
o
ce
s
s
in
g
s
p
ee
d
with
o
u
t g
r
ea
tly
co
m
p
r
o
m
is
in
g
p
e
r
f
o
r
m
an
ce
.
Data
s
p
litt
in
g
in
v
o
lv
es
d
iv
i
d
in
g
a
d
ataset
in
to
d
is
tin
ct
s
u
b
s
ets
f
o
r
m
o
d
el
d
ev
elo
p
m
e
n
t
ty
p
ically
a
tr
ain
i
n
g
s
et
f
o
r
f
itti
n
g
t
h
e
m
o
d
el,
a
v
alid
atio
n
s
et
f
o
r
p
ar
a
m
eter
tu
n
in
g
,
a
n
d
a
test
s
et
f
o
r
f
in
al
p
er
f
o
r
m
an
c
e
ev
alu
ati
o
n
.
T
h
is
r
esear
c
h
wo
r
k
ad
o
p
ts
8
0
%
tr
ain
in
g
an
d
2
0
%
test
in
g
r
atio
to
en
s
u
r
e
th
e
m
o
d
el
g
e
n
er
alize
s
ef
f
ec
tiv
ely
to
u
n
s
ee
n
d
ata.
3.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
T
h
e
v
is
u
aliza
tio
n
aid
s
b
u
s
in
ess
es
in
id
en
tify
in
g
th
e
m
o
s
t
ef
f
icien
t
ch
an
n
els
a
n
d
allo
ca
tin
g
r
eso
u
r
ce
s
ac
co
r
d
in
g
l
y
.
T
h
e
s
ig
n
if
ica
n
ce
o
f
s
elec
ted
m
ac
h
in
e
lear
n
i
n
g
alg
o
r
ith
m
s
is
ju
s
tifie
d
in
th
e
f
o
llo
win
g
s
ec
tio
n
.
R
an
d
o
m
f
o
r
est
r
eg
r
ess
o
r
is
a
m
o
d
el
wh
ich
r
ep
r
esen
ts
m
a
n
y
d
ec
is
io
n
tr
ee
s
.
E
ac
h
t
r
ee
o
n
th
eir
o
wn
lear
n
s
f
r
o
m
a
s
u
b
s
et
o
f
d
ataset
an
d
g
i
v
es
its
o
wn
p
r
ed
ictio
n
.
T
h
e
f
in
al
an
s
wer
is
th
e
av
er
ag
e
o
f
all
o
f
th
e
p
r
e
d
ictio
n
s
.
I
t
allo
ws
to
d
ec
r
ea
s
e
o
v
er
f
itti
n
g
,
to
m
ak
e
th
e
m
o
d
el
m
o
r
e
r
o
b
u
s
t
[
2
0
]
.
I
n
th
e
c
o
n
tex
t
o
f
p
r
ed
ictin
g
m
ar
k
etin
g
R
OI
,
th
is
m
o
d
el
d
em
o
n
s
tr
ates
h
o
w
co
m
p
licated
th
e
r
elatio
n
s
h
ip
is
b
etwe
en
th
e
in
v
ested
m
o
n
ey
an
d
r
ev
en
u
e.
XGBo
o
s
t
r
eg
r
ess
o
r
i
s
an
ad
v
an
ce
d
b
o
o
s
tin
g
alg
o
r
ith
m
.
I
t
im
p
r
o
v
es
d
ec
is
io
n
tr
ee
s
v
ia
an
iter
ativ
e
lear
n
in
g
p
r
o
ce
d
u
r
e.
XGBo
o
s
t c
o
n
s
tr
u
ct
s
tr
ee
s
in
s
er
ial
m
an
n
er
an
d
ev
er
y
n
ex
t t
r
ee
co
r
r
ec
ts
th
e
er
r
o
r
s
o
f
p
r
ev
i
o
u
s
tr
ee
s
.
T
h
e
f
in
al
p
r
ed
ictio
n
is
th
e
ag
g
r
eg
ated
v
alu
e
th
at
t
h
e
a
lg
o
r
ith
m
p
r
o
v
i
d
es
af
ter
th
e
n
ec
ess
ar
y
o
p
tim
izatio
n
[
2
2
]
,
[
2
3
]
.
T
h
is
ap
p
r
o
ac
h
is
esp
ec
ially
u
s
ef
u
l
in
m
ar
k
etin
g
p
r
ed
ictio
n
s
.
Sin
ce
it
is
ca
p
a
b
le
o
f
h
an
d
lin
g
lar
g
e
d
ata
an
d
r
ed
u
c
in
g
er
r
o
r
r
ates.
L
in
ea
r
r
e
g
r
ess
io
n
p
r
o
v
id
es
a
s
tr
aig
h
t
f
o
r
war
d
an
d
i
n
ter
p
r
eta
b
le
b
aselin
e
f
o
r
p
r
ed
ictin
g
R
OI
,
r
ea
s
o
n
ab
ly
ef
f
ec
tiv
e
at
m
o
d
elin
g
s
u
b
tle
r
elatio
n
s
h
ip
s
[
2
4
]
,
[
2
5
]
.
Fig
u
r
e
2
is
th
e
s
ca
tter
p
lo
t
o
f
p
r
ed
icted
R
OI
v
alu
es
v
er
s
u
s
ac
t
u
al
R
OI
v
alu
es
wh
en
e
m
p
lo
y
i
n
g
th
e
lin
ea
r
r
eg
r
ess
io
n
m
o
d
el.
T
h
e
id
ea
l
f
it
y
=x
r
ep
r
esen
ted
b
y
a
d
ash
ed
r
ed
lin
e
s
ig
n
if
i
es
p
er
f
ec
t
p
r
ed
ictio
n
wh
er
e
p
r
ed
icted
an
d
ac
tu
al
v
alu
es
ar
e
eq
u
iv
alen
t.
B
lu
e
p
o
in
ts
s
ig
n
if
y
in
d
iv
id
u
al
p
r
e
d
ictio
n
s
,
an
d
t
h
e
d
o
t'
s
p
o
s
iti
o
n
o
n
th
e
r
e
d
lin
e
in
d
icate
s
th
e
ac
cu
r
ac
y
o
f
th
e
m
o
d
el'
s
f
it.
Gr
ad
ien
t
b
o
o
s
tin
g
r
eg
r
ess
o
r
b
u
ild
s
s
ev
er
al
d
ec
i
s
io
n
tr
ee
s
o
n
e
af
ter
an
o
th
er
th
e
d
ataset
(
D)
is
d
iv
id
ed
in
to
s
u
b
s
ets
(
D1
,
D2
,
..
.
,
Dn
)
.
E
ac
h
s
u
b
s
et
is
u
s
ed
f
o
r
t
r
ain
in
g
th
e
d
ec
is
io
n
tr
ee
an
d
ea
ch
tr
ee
lear
n
s
f
r
o
m
th
e
m
is
tak
es
o
f
th
e
p
r
ev
io
u
s
t
r
ee
.
T
h
e
f
in
al
r
esu
lt
is
o
b
tain
e
d
b
y
av
er
a
g
in
g
th
e
p
r
ed
ictio
n
s
o
f
all
tr
ee
s
,
wh
ic
h
ac
h
iev
es
h
ig
h
e
r
ac
c
u
r
ac
y
[
2
1
]
.
T
h
is
ap
p
r
o
ac
h
is
h
i
g
h
ly
r
ec
o
m
m
en
d
e
d
f
o
r
m
ar
k
etin
g
a
n
aly
tics
b
ec
au
s
e
it h
elp
s
to
p
r
e
d
ict
th
e
R
OI
with
h
ig
h
ac
cu
r
ac
y
an
d
also
d
ec
r
ea
s
es th
e
er
r
o
r
r
ate.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
1
5
,
No
.
4
,
Au
g
u
s
t 2
0
2
6
:
3
5
2
8
-
3
5
3
6
3532
T
o
co
m
p
ar
e
th
e
ef
f
ec
tiv
en
es
s
o
f
t
h
e
in
v
estme
n
t,
p
r
o
d
u
ct
p
er
f
o
r
m
a
n
ce
a
n
aly
s
is
(
Fig
u
r
e
3
)
an
d
m
ar
k
etin
g
ca
m
p
aig
n
a
n
aly
s
is
(
Fig
u
r
e
4
)
ar
e
ca
r
r
ied
o
u
t
u
p
o
n
p
r
e
p
r
o
ce
s
s
in
g
o
f
t
h
e
d
ataset.
I
d
en
tific
atio
n
o
f
th
e
m
o
s
t
an
d
least
p
r
o
f
itab
le
p
r
o
d
u
cts
ca
n
b
e
d
o
n
e
u
s
in
g
a
b
ar
ch
ar
t
th
at
co
m
p
ar
es
th
e
to
tal
p
r
o
f
it
an
d
to
tal
co
s
t
f
o
r
ea
ch
p
r
o
d
u
ct.
Fig
u
r
e
4
r
ep
r
esen
tin
g
th
e
p
er
ce
n
tag
e
o
f
to
tal
r
ev
en
u
es
v
er
s
u
s
to
tal
ex
p
en
s
es,
f
u
r
th
e
r
it
p
r
o
v
id
es
an
o
v
er
all
p
er
s
p
ec
ti
v
e
o
n
f
ir
m
p
r
o
f
itab
ilit
y
.
Ad
v
er
tis
in
g
s
tr
ateg
y
ca
n
b
e
o
p
ti
m
ized
b
y
ca
m
p
aig
n
an
aly
s
is
o
f
m
ar
k
etin
g
ca
m
p
ai
g
n
s
r
ep
r
esen
tin
g
th
e
o
v
er
all
c
o
s
t
in
cu
r
r
e
d
f
o
r
ea
c
h
ca
m
p
aig
n
an
d
th
e
n
u
m
b
er
o
f
ch
an
n
els
u
tili
ze
d
f
o
r
m
ar
k
eti
n
g
.
T
h
e
R
OI
co
n
tr
ib
u
tio
n
o
f
ea
ch
m
ar
k
etin
g
ch
an
n
el
an
d
th
e
b
r
ea
k
d
o
wn
o
f
m
ar
k
etin
g
co
s
ts
o
n
a
ch
a
n
n
el
-
wis
e
b
asis
ar
e
r
ep
r
esen
ted
u
s
in
g
a
p
ie
c
h
ar
t;
it
h
elp
s
th
e
m
o
d
el
to
id
en
tify
th
e
m
o
s
t
p
r
o
f
itab
le
m
ar
k
etin
g
ch
an
n
el
f
o
r
in
v
estme
n
ts
.
T
o
en
s
u
r
e
b
etter
d
ec
is
io
n
s
in
b
u
d
g
et
allo
ca
tio
n
,
th
is
m
o
d
el
f
o
r
ec
asts
th
e
R
OI
f
r
o
m
v
ar
io
u
s
m
ar
k
etin
g
ch
an
n
els.
R
an
d
o
m
f
o
r
est
r
eg
r
ess
o
r
al
g
o
r
ith
m
av
er
a
g
es
th
e
o
u
tco
m
es a
n
d
m
in
im
izes v
ar
ia
n
ce
u
s
in
g
d
ec
is
io
n
tr
ee
s
.
T
h
e
XGBo
o
s
t
r
eg
r
ess
o
r
is
an
ad
v
an
ce
d
g
r
ad
ien
t
b
o
o
s
tin
g
m
o
d
el,
an
d
it
ca
n
m
an
a
g
e
th
e
m
is
s
in
g
d
ata
ef
f
icien
tly
an
d
c
o
n
tin
u
ally
e
n
h
an
ce
th
e
p
r
ec
is
io
n
.
T
h
e
lin
ea
r
r
eg
r
ess
io
n
is
s
im
p
le
an
d
u
n
d
er
s
tan
d
ab
le
m
o
d
el
to
p
r
ed
ict
R
OI
,
an
d
th
e
g
r
ad
i
en
t
b
o
o
s
tin
g
r
e
g
r
ess
o
r
en
h
a
n
c
es
th
e
p
r
ed
ictio
n
s
b
y
c
o
r
r
ec
ti
n
g
th
e
er
r
o
r
s
o
f
th
e
p
r
ec
ed
in
g
m
o
d
els.
T
h
e
q
u
ali
ty
o
f
ea
ch
m
o
d
el
is
m
ea
s
u
r
ed
b
y
th
e
R
2
co
ef
f
icien
t,
w
h
ich
in
d
icate
s
th
e
p
r
o
p
o
r
tio
n
o
f
v
ar
ian
ce
in
R
OI
.
T
h
e
MSE
,
in
d
icate
s
th
e
er
r
o
r
in
p
r
ed
ictio
n
.
T
h
e
s
y
s
tem
co
m
p
ar
es
all
m
o
d
els
an
d
c
h
o
o
s
es
th
e
b
est
p
r
ed
ictio
n
.
T
h
is
is
ac
h
iev
e
d
th
r
o
u
g
h
a
p
ie
ch
ar
t,
wh
ich
s
h
o
ws
th
e
p
er
ce
n
ta
g
e
o
f
t
h
e
co
n
tr
ib
u
tio
n
o
f
ea
c
h
m
a
r
k
etin
g
ch
a
n
n
el
to
th
e
to
tal
R
OI
.
T
h
e
ac
tu
al
R
OI
an
d
p
r
e
d
icted
R
OI
ar
e
ca
lcu
lated
u
s
in
g
(
4
)
an
d
(
5
)
.
(
%
)
=
−
×
100
(
4
)
(
%
)
=
(
ma
r
k
e
ti
n
g
c
ha
n
n
e
l
s
pe
n
d
,
c
ost
of
pr
od
uc
t
,
pr
o
fit
ma
r
gin
(
%
)
)
(
5
)
Fig
u
r
e
2
.
Pre
d
icte
d
v
s
ac
tu
al
R
OI
u
s
in
g
lin
ea
r
r
eg
r
ess
io
n
Fig
u
r
e
3
.
Pro
d
u
ct
-
wis
e
p
r
o
f
it
an
d
co
s
t c
o
m
p
ar
is
o
n
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
A
ma
ch
in
e
lea
r
n
in
g
fr
a
mewo
r
k
fo
r
p
r
ed
ictin
g
a
n
d
o
p
timiz
in
g
r
etu
r
n
o
n
in
ve
s
tmen
t
…
(
C
h
a
n
d
r
a
C
h
a
th
u
r
a
)
3533
Fig
u
r
e
4
.
R
OI
d
is
tr
ib
u
tio
n
ac
r
o
s
s
m
ar
k
etin
g
ch
a
n
n
els
T
h
e
m
o
d
el
ac
cu
r
ac
y
is
ch
ec
k
e
d
b
y
f
o
r
ec
asti
n
g
th
e
R
OI
an
d
co
m
p
ar
in
g
it
with
th
e
en
ter
ed
u
s
er
v
alu
e
f
o
r
ac
tu
al
R
OI
.
Fo
r
ea
c
h
m
ac
h
in
e
lear
n
in
g
m
o
d
el,
a
b
s
o
lu
te
p
er
ce
n
tag
e
er
r
o
r
(
APE)
is
co
m
p
u
ted
b
y
t
h
e
s
y
s
tem
to
id
en
tify
th
e
p
r
e
d
ictio
n
ac
c
u
r
ac
y
.
T
ab
le
1
h
as
th
e
ac
tu
al
R
OI
v
alu
e,
th
e
p
r
ed
icted
R
OI
v
alu
e,
an
d
th
e
p
er
ce
n
tag
e
o
f
ac
cu
r
ac
y
.
T
h
is
f
in
al
s
tep
aim
s
at
h
elp
in
g
c
o
m
p
an
ies
s
elec
t
th
e
m
o
s
t
ac
cu
r
ate
m
ac
h
in
e
lea
r
n
in
g
alg
o
r
ith
m
to
in
cr
ea
s
e
th
eir
m
ar
k
etin
g
r
etu
r
n
.
Gen
er
ally
,
th
is
d
ep
lo
y
m
e
n
t
m
er
g
es
d
at
a
v
is
u
aliza
tio
n
an
d
m
ac
h
in
e
lear
n
in
g
-
b
ased
R
OI
p
r
ed
ictio
n
,
en
a
b
lin
g
c
o
m
p
a
n
ies
to
m
ak
e
in
v
estme
n
t
d
ec
is
io
n
s
b
ased
o
n
d
ata
,
r
ed
u
ce
in
ef
f
icien
cies,
an
d
in
c
r
ea
s
e
p
r
o
f
itab
ilit
y
.
L
in
ea
r
r
eg
r
ess
io
n
g
iv
es
h
ig
h
ac
cu
r
ac
y
i
n
p
r
ed
ictin
g
R
OI
,
f
u
r
th
er
it
is
ev
id
en
t
t
h
at
th
is
alg
o
r
ith
m
ca
n
b
e
u
s
ed
t
o
m
o
d
el
an
d
d
e
m
o
n
s
tr
ate
th
e
in
ter
ac
tio
n
b
etwe
en
m
ar
k
etin
g
e
x
p
en
d
itu
r
e
an
d
R
OI
ef
f
ec
tiv
ely
.
Per
f
o
r
m
an
ce
o
f
m
o
d
els
is
ev
alu
ated
b
ased
o
n
MSE
an
d
R
2
wh
ich
ex
p
lain
s
h
o
w
e
f
f
ec
tiv
ely
ea
ch
m
o
d
el
r
ed
u
ce
s
p
r
ed
ictio
n
er
r
o
r
an
d
ex
p
lain
s
v
ar
iat
io
n
o
f
R
OI
.
T
h
e
p
er
f
o
r
m
an
ce
o
f
th
e
r
an
d
o
m
f
o
r
est
r
eg
r
ess
o
r
lies
b
etwe
en
X
GB
o
o
s
t,
wh
ich
h
a
d
t
h
e
lo
west
ac
cu
r
ac
y
,
a
n
d
lin
ea
r
r
eg
r
ess
io
n
,
wh
ich
s
h
o
wed
t
h
e
h
ig
h
est.
I
t
is
ab
le
to
p
r
ed
ict
R
OI
o
u
tco
m
es
with
g
r
ea
ter
f
id
e
lity
th
an
XGBo
o
s
t
an
d
g
r
a
d
ien
t b
o
o
s
tin
g
,
alth
o
u
g
h
it is
s
lig
h
tly
less
p
r
ec
is
e
co
m
p
ar
ed
to
lin
ea
r
r
eg
r
ess
io
n
.
T
h
e
r
an
d
o
m
f
o
r
est
r
eg
r
ess
o
r
ex
h
ib
its
g
o
o
d
o
v
er
all
p
er
f
o
r
m
an
ce
.
I
n
g
e
n
er
al,
it
p
er
f
o
r
m
s
well
wi
th
b
o
th
f
itti
n
g
a
n
d
g
en
er
aliza
tio
n
m
o
d
els to
th
e
d
ata.
XGBo
o
s
t h
as a
ls
o
s
h
o
wn
s
tr
o
n
g
g
en
er
ali
za
tio
n
ca
p
ab
ilit
ies.
I
t
is
v
er
y
s
tr
o
n
g
at
f
in
d
in
g
g
en
er
al
p
atter
n
s
in
th
e
d
ataset.
Ho
wev
er
,
it
wasn
'
t
am
o
n
g
t
h
e
b
est
o
n
es
in
th
e
co
m
p
etitio
n
.
I
t
h
ad
lar
g
er
er
r
o
r
r
ates
co
m
p
a
r
ed
to
o
th
er
m
o
d
els.
T
h
is
m
ea
n
s
th
at
it
m
ig
h
t
n
o
t
b
e
th
e
b
est
f
it
f
o
r
th
e
d
ataset.
I
t
is
ex
ce
p
tio
n
al
at
h
a
n
d
lin
g
co
m
p
lex
f
ea
tu
r
es.
T
h
e
g
r
a
d
ien
t
b
o
o
s
tin
g
r
eg
r
ess
o
r
h
as
s
h
o
w
n
b
etter
p
er
f
o
r
m
a
n
ce
t
h
an
XG
B
o
o
s
t.
I
t
h
as
a
lo
t
lo
wer
p
r
e
d
ictio
n
er
r
o
r
d
em
o
n
s
tr
ated
,
th
e
ef
f
ec
tiv
e
n
ess
in
lear
n
in
g
f
r
o
m
d
ata
p
atter
n
s
,
a
n
d
r
ed
u
ci
n
g
in
ac
cu
r
ac
ies.
I
ts
ab
ilit
y
to
ca
p
tu
r
e
co
m
p
lex
r
elat
io
n
s
h
ip
s
with
in
th
e
d
ata
m
ad
e
it o
n
e
o
f
t
h
e
m
o
s
t a
cc
u
r
ate
an
d
c
o
n
s
is
ten
t m
o
d
els
in
th
e
s
tu
d
y
.
L
in
ea
r
r
e
g
r
ess
io
n
m
o
d
el
d
eliv
er
ed
a
s
u
r
p
r
is
e,
as
it
was
r
ate
d
b
est
in
ter
m
s
o
f
p
r
ed
ictio
n
ac
cu
r
ac
y
.
T
h
e
m
o
d
el
h
as
s
h
o
wn
th
e
s
m
allest
er
r
o
r
,
as
it
h
as
alm
o
s
t
p
er
f
ec
tly
em
b
r
ac
e
d
th
e
d
ata.
T
h
is
im
p
lies
th
at
th
e
r
elatio
n
s
h
ip
b
etwe
en
in
p
u
t
f
ea
tu
r
es
an
d
R
OI
is
h
ig
h
ly
lin
ea
r
,
allo
win
g
th
e
lin
ea
r
r
e
g
r
ess
io
n
m
o
d
el
to
p
r
o
v
id
e
an
alm
o
s
t
ac
cu
r
ate
p
r
e
d
ictio
n
.
As
p
er
th
e
p
er
f
o
r
m
an
ce
,
lin
e
ar
r
eg
r
ess
io
n
m
o
d
el
ap
p
ea
r
s
to
b
e
th
e
b
est
m
o
d
el
to
p
r
ed
ict
th
e
R
OI
in
s
u
c
h
c
ir
cu
m
s
tan
ce
s
.
Alth
o
u
g
h
th
e
v
ar
iatio
n
in
ac
c
u
r
ac
ies
am
o
n
g
m
o
d
els
ar
e
v
er
y
m
in
im
al,
it
is
ev
id
en
t
th
at
all
m
o
d
els
wer
e
ef
f
ec
tiv
e
in
R
OI
p
r
ed
ictio
n
f
o
r
m
ar
k
etin
g
ex
p
en
d
itu
r
e,
f
u
r
th
er
s
u
p
p
o
r
tin
g
t
h
e
ef
f
icac
y
o
f
m
ac
h
in
e
lear
n
in
g
i
n
m
ar
k
etin
g
b
u
d
g
et
o
p
tim
izatio
n
.
T
a
b
le
1
p
r
e
s
en
ts
a
co
m
p
ar
is
o
n
o
f
ac
tu
al
a
n
d
p
r
ed
icted
R
OI
of
d
i
f
f
er
en
t
m
ar
k
etin
g
ch
a
n
n
els
f
o
r
th
e
f
o
u
r
m
ac
h
in
e
le
ar
n
in
g
m
o
d
els
also
p
r
o
v
id
es
m
o
r
e
in
s
ig
h
ts
.
T
h
e
o
v
er
all
ac
cu
r
ac
y
as
in
Fig
u
r
e
5
o
f
ea
ch
m
o
d
el
is
th
e
av
er
ag
e
o
f
th
e
ac
cu
r
ac
ies
o
f
all
in
d
iv
id
u
al
m
a
r
k
etin
g
c
h
an
n
els.
T
ab
le
1
.
Actu
al
an
d
p
r
ed
icted
R
OI
v
alu
es a
cr
o
s
s
m
ar
k
etin
g
c
h
an
n
els
M
a
r
k
e
t
i
n
g
c
h
a
n
n
e
l
R
a
n
d
o
m f
o
r
e
s
t
r
e
g
r
e
ss
o
r
X
G
B
o
o
st
G
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
r
e
g
r
e
ss
o
r
Li
n
e
a
r
r
e
g
r
e
s
si
o
n
A
c
t
u
a
l
R
O
I
(
%)
P
r
e
d
i
c
t
e
d
R
O
I
(
%)
A
c
t
u
a
l
R
O
I
(
%)
P
r
e
d
i
c
t
e
d
R
O
I
(
%)
A
c
t
u
a
l
R
O
I
(
%)
P
r
e
d
i
c
t
e
d
R
O
I
(
%)
A
c
t
u
a
l
R
O
I
(
%)
P
r
e
d
i
c
t
e
d
R
O
I
(
%)
B
i
l
l
b
o
a
r
d
1
1
.
2
2
1
1
.
1
1
1
.
2
2
1
1
.
1
1
1
.
2
2
1
1
.
2
1
1
.
2
2
1
1
.
1
N
e
w
sp
a
p
e
r
1
1
.
2
2
1
1
.
1
1
1
.
2
2
1
1
.
1
1
1
.
2
2
1
1
.
1
1
1
.
2
2
1
1
.
2
S
o
c
i
a
l
me
d
i
a
3
3
.
2
9
3
3
.
5
3
3
.
2
9
3
3
.
4
3
3
.
2
9
3
3
.
3
3
3
.
2
9
3
3
.
2
TV
1
1
.
2
1
1
1
.
1
1
1
.
2
1
11
1
1
.
2
1
11
1
1
.
2
1
1
1
.
1
O
n
l
i
n
e
st
o
r
e
s
2
1
.
8
7
2
2
.
1
2
1
.
8
7
2
2
.
3
2
1
.
8
7
2
2
.
3
2
1
.
8
7
2
2
.
1
M
o
b
i
l
e
a
p
p
l
i
c
a
t
i
o
n
s
1
1
.
1
9
1
1
.
1
1
1
.
1
9
1
1
.
1
1
1
.
1
9
1
1
.
1
1
1
.
1
9
1
1
.
2
O
v
e
r
a
l
l
a
c
c
u
r
a
c
y
9
9
.
0
7
9
8
.
8
1
9
9
.
0
1
9
9
.
3
9
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
1
5
,
No
.
4
,
Au
g
u
s
t 2
0
2
6
:
3
5
2
8
-
3
5
3
6
3534
Fig
u
r
e
5
.
Acc
u
r
ac
y
c
o
m
p
ar
is
o
n
o
f
m
ac
h
in
e
lear
n
in
g
m
o
d
els
T
h
e
r
esu
lt
s
h
o
ws
th
at
lin
ea
r
r
eg
r
ess
io
n
h
as
th
e
h
ig
h
est
o
v
er
all
ac
cu
r
ac
y
at
9
9
.
3
9
%,
th
en
r
an
d
o
m
f
o
r
est
at
9
9
.
0
7
%,
f
o
llo
wed
b
y
g
r
ad
ien
t
b
o
o
s
tin
g
at
9
9
.
0
1
%,
an
d
last
ly
XGBo
o
s
t
at
9
8
.
8
1
%
.
T
h
is
is
p
r
o
v
en
b
y
th
e
MSE
an
d
R
2
v
alu
es
th
at
also
p
r
o
v
e
th
at
lin
ea
r
r
eg
r
ess
io
n
was
th
e
m
o
s
t
ef
f
ec
tiv
e
in
R
OI
p
r
ed
ictio
n
.
T
h
e
an
aly
s
is
is
ev
id
en
t
th
at
m
ac
h
in
e
lear
n
in
g
is
also
ef
f
ec
tiv
e
in
o
p
tim
izatio
n
o
f
m
a
r
k
etin
g
b
u
d
g
et.
Ma
c
h
in
e
lear
n
in
g
alg
o
r
ith
m
s
p
er
f
o
r
m
e
d
well
in
an
aly
zin
g
m
ar
k
etin
g
ch
a
n
n
els
in
d
iv
id
u
ally
an
d
p
r
o
v
i
d
ed
ac
c
u
r
ate
f
o
r
ec
asts
,
f
u
r
th
er
th
e
v
ar
iatio
n
in
f
o
r
ec
ast
is
v
er
y
m
in
im
al
am
o
n
g
alg
o
r
ith
m
s
.
C
h
an
n
els
o
f
s
o
cial
m
ed
ia
an
d
o
n
lin
e
s
to
r
es
alwa
y
s
h
ad
th
e
h
ig
h
est
R
OI
,
wh
ich
co
r
r
esp
o
n
d
s
to
th
e
tr
en
d
o
f
h
ig
h
er
p
r
o
f
itab
ilit
y
o
f
d
ig
ital
m
ar
k
etin
g
s
tr
ateg
ies o
v
e
r
th
e
t
r
ad
itio
n
al
o
n
es,
s
u
c
h
as b
illb
o
ar
d
s
an
d
n
ewsp
a
p
er
s
.
T
h
e
an
aly
s
is
g
iv
es
an
in
s
ig
h
t
th
at
in
v
esti
n
g
in
d
ig
ital
p
latf
o
r
m
s
is
m
o
r
e
p
r
o
f
itab
le
th
a
n
in
v
esti
n
g
in
tr
ad
itio
n
al
ad
v
e
r
tis
in
g
.
T
h
e
p
r
ed
icted
R
OI
o
f
t
h
e
ch
a
n
n
els
o
f
TV
,
b
illb
o
a
r
d
s
,
an
d
n
ewsp
a
p
er
s
co
in
cid
e
d
with
th
e
ac
tu
al
R
OI
m
o
s
t
o
f
all
in
all
th
e
m
o
d
els,
wh
ich
s
h
o
ws
th
at
th
e
tr
ad
itio
n
al
ch
a
n
n
els
o
f
ad
v
e
r
tis
in
g
h
av
e
r
elativ
ely
s
tab
le
R
OI
,
an
d
less
p
r
o
f
itab
le
th
an
d
ig
ital
ch
a
n
n
els,
b
u
t
th
e
R
OI
o
f
th
ese
ch
an
n
els
is
m
o
r
e
p
r
ed
ictab
le
an
d
less
d
y
n
a
m
ic.
Am
o
n
g
th
e
en
s
em
b
le
m
o
d
els
,
r
an
d
o
m
f
o
r
est
an
d
g
r
ad
ien
t
b
o
o
s
tin
g
p
er
f
o
r
m
e
d
s
lig
h
tly
b
etter
th
an
XGBo
o
s
t
.
Ov
er
all
r
esu
lts
ar
e
also
v
alid
at
in
g
th
at
m
ac
h
i
n
e
lear
n
in
g
alg
o
r
ith
m
s
d
o
h
a
v
e
th
e
ab
ilit
y
to
g
en
er
ate
h
ig
h
ac
cu
r
ate
R
OI
esti
m
ates,
en
ab
lin
g
o
r
g
an
izatio
n
s
to
m
a
k
e
s
o
lid
in
v
estme
n
t
d
ec
is
io
n
s
b
ac
k
ed
b
y
d
ata
an
d
h
ig
h
co
n
f
id
en
ce
lev
els.
T
h
e
d
etailed
a
n
aly
s
is
in
d
icate
s
th
at
d
ig
ital
m
ar
k
etin
g
c
h
an
n
els
d
o
m
in
ated
t
h
e
R
OI
p
r
ed
ictio
n
s
an
d
s
u
g
g
ests
th
at
b
u
s
in
ess
es
s
h
o
u
ld
p
r
i
o
r
itize
in
v
estme
n
ts
in
s
o
cial
m
ed
ia
an
d
o
n
lin
e.
T
h
e
s
tu
d
y
s
h
o
ws
th
at
m
ac
h
in
e
lear
n
in
g
-
b
ased
R
OI
p
r
ed
ictio
n
s
ar
e
u
s
ef
u
l
f
o
r
d
ec
is
io
n
m
ak
in
g
in
m
ar
k
etin
g
,
allo
win
g
b
u
s
in
ess
es
to
d
is
tr
ib
u
te
th
eir
b
u
d
g
et
p
r
o
p
er
l
y
,
m
in
im
ize
f
i
n
an
cial
r
is
k
s
an
d
,
th
er
ef
o
r
e,
m
ax
im
ize
th
e
R
OI
.
T
h
o
u
g
h
r
esu
lts
h
av
e
b
ee
n
s
h
o
wn
with
in
th
e
p
r
esen
t
ex
p
er
im
en
t,
it
is
p
o
s
s
ib
le
to
d
iv
e
d
ee
p
er
in
to
th
e
r
esu
lts
b
y
co
n
d
u
ctin
g
ex
p
er
im
e
n
ts
o
n
b
ig
g
e
r
d
atasets
.
Ad
d
itio
n
al
r
esear
ch
wo
r
k
co
u
ld
b
e
d
o
n
e
to
im
p
r
o
v
e
in
ter
p
r
etab
ilit
y
m
e
th
o
d
s
to
f
u
r
th
er
im
p
r
o
v
e
th
e
a
p
p
licatio
n
o
f
th
e
ap
p
r
o
ac
h
.
4.
CO
NCLU
SI
O
N
T
h
is
r
esear
ch
wo
r
k
was
co
n
d
u
cted
to
in
co
r
p
o
r
ate
m
ac
h
in
e
le
ar
n
in
g
m
et
h
o
d
s
f
o
r
th
e
p
r
e
d
ict
io
n
o
f
th
e
R
OI
o
f
m
ar
k
etin
g
ca
m
p
aig
n
s
.
T
h
e
r
esear
ch
p
r
o
v
id
es
b
u
s
in
e
s
s
es
ev
id
en
ce
-
b
ased
r
ec
o
m
m
e
n
d
atio
n
s
o
n
h
o
w
to
m
o
d
if
y
th
eir
b
u
d
g
et
d
is
tr
ib
u
ti
o
n
.
Fu
r
th
er
th
is
r
esear
ch
wo
r
k
h
as
s
h
o
w
n
th
at
m
ac
h
i
n
e
lear
n
in
g
tec
h
n
iq
u
es
ca
n
ef
f
icien
tly
p
r
ed
ict
m
ar
k
etin
g
p
er
f
o
r
m
a
n
ce
with
a
h
ig
h
lev
el
o
f
ac
cu
r
ac
y
.
T
h
e
r
esu
lt
s
h
o
ws
th
at
th
e
b
est
m
o
d
el
f
o
r
t
h
is
ty
p
e
o
f
an
al
y
s
is
is
lin
ea
r
r
eg
r
ess
io
n
,
wh
ic
h
h
as
ac
h
iev
ed
th
e
lo
west
MSE
a
n
d
th
e
h
ig
h
est
R
2
s
co
r
e,
f
u
r
th
er
it
in
d
icate
s
th
at
th
e
d
a
taset
f
o
llo
ws
a
h
ig
h
ly
lin
ea
r
r
elatio
n
s
h
ip
b
etwe
en
in
p
u
t
f
ea
t
u
r
es
an
d
R
OI
.
T
h
e
h
ig
h
est
R
OI
was
r
eg
u
lar
ly
d
e
r
iv
ed
f
r
o
m
d
i
g
ital
m
ar
k
etin
g
ch
an
n
els,
f
o
r
ex
am
p
le,
s
o
cial
m
ed
ia
an
d
o
n
lin
e
s
to
r
es.
T
h
is
r
esear
ch
wo
r
k
co
n
f
ir
m
s
d
ig
ital
m
ar
k
etin
g
in
v
estme
n
ts
ar
e
a
h
ea
d
o
f
tr
ad
itio
n
al
m
eth
o
d
s
.
Pre
d
ictiv
e
an
aly
tics
h
as
p
r
o
v
en
to
b
e
r
ea
lly
h
elp
f
u
l
f
o
r
m
ak
in
g
m
ar
k
etin
g
d
ec
is
io
n
s
.
T
h
e
b
est
ac
cu
r
ac
y
is
ac
h
iev
ed
b
y
lin
ea
r
r
e
g
r
ess
io
n
at
9
9
.
3
9
%,
f
o
llo
wed
clo
s
ely
b
y
r
an
d
o
m
f
o
r
est
at
9
9
.
0
7
%,
g
r
ad
ien
t
b
o
o
s
tin
g
at
9
9
.
0
1
%,
a
n
d
XGBo
o
s
t
at
9
8
.
8
1
%.
T
h
e
im
p
lem
en
te
d
m
o
d
el
d
o
es
n
o
t
ca
p
t
u
r
e
s
u
d
d
e
n
b
e
h
a
v
io
r
al
ch
an
g
es
s
in
ce
th
is
r
esear
ch
is
b
ased
o
n
h
is
to
r
ical
d
ata.
Su
b
s
eq
u
en
t
r
esear
c
h
ca
n
b
e
d
o
n
e
b
y
u
s
in
g
ex
ten
s
iv
e
d
atasets
wh
ich
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
A
ma
ch
in
e
lea
r
n
in
g
fr
a
mewo
r
k
fo
r
p
r
ed
ictin
g
a
n
d
o
p
timiz
in
g
r
etu
r
n
o
n
in
ve
s
tmen
t
…
(
C
h
a
n
d
r
a
C
h
a
th
u
r
a
)
3535
in
clu
d
e
d
ata
co
llected
o
v
e
r
tim
e,
th
e
b
eh
av
io
r
o
f
co
n
s
u
m
er
s
an
d
th
e
tactics
o
f
th
e
p
r
ice
d
y
n
am
ics
to
en
h
an
ce
th
e
ac
cu
r
ac
y
a
n
d
p
er
s
o
n
alize
d
p
r
ed
ictio
n
s
.
F
UNDING
I
NF
O
R
M
A
T
I
O
N
N
o
f
u
n
d
in
g
in
v
o
lv
e
d
.
AUTHO
R
CO
NT
RI
B
UT
I
O
NS ST
A
T
E
M
E
N
T
T
h
is
jo
u
r
n
al
u
s
es
th
e
C
o
n
tr
ib
u
to
r
R
o
les
T
ax
o
n
o
m
y
(
C
R
ed
iT)
to
r
ec
o
g
n
ize
in
d
iv
id
u
al
au
th
o
r
co
n
tr
ib
u
tio
n
s
,
r
ed
u
ce
au
th
o
r
s
h
ip
d
is
p
u
tes,
an
d
f
ac
ilit
ate
co
llab
o
r
atio
n
.
Na
m
e
o
f
Aut
ho
r
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
C
h
an
d
r
a
C
h
ath
u
r
a
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
Kee
r
th
an
Say
a
✓
✓
✓
✓
✓
✓
✓
✓
✓
Sath
is
h
k
u
m
ar
Ma
n
i
✓
✓
✓
✓
✓
✓
✓
C
:
C
o
n
c
e
p
t
u
a
l
i
z
a
t
i
o
n
M
:
M
e
t
h
o
d
o
l
o
g
y
So
:
So
f
t
w
a
r
e
Va
:
Va
l
i
d
a
t
i
o
n
Fo
:
Fo
r
mal
a
n
a
l
y
s
i
s
I
:
I
n
v
e
s
t
i
g
a
t
i
o
n
R
:
R
e
so
u
r
c
e
s
D
:
D
a
t
a
C
u
r
a
t
i
o
n
O
:
W
r
i
t
i
n
g
-
O
r
i
g
i
n
a
l
D
r
a
f
t
E
:
W
r
i
t
i
n
g
-
R
e
v
i
e
w
&
E
d
i
t
i
n
g
Vi
:
Vi
su
a
l
i
z
a
t
i
o
n
Su
:
Su
p
e
r
v
i
s
i
o
n
P
:
P
r
o
j
e
c
t
a
d
mi
n
i
st
r
a
t
i
o
n
Fu
:
Fu
n
d
i
n
g
a
c
q
u
i
si
t
i
o
n
CO
NF
L
I
C
T
O
F
I
N
T
E
R
E
S
T
ST
A
T
E
M
E
NT
N
o
co
n
f
lict o
f
in
ter
est.
DATA AV
AI
L
AB
I
L
I
T
Y
T
h
e
d
ata
th
at
s
u
p
p
o
r
t
th
e
f
i
n
d
in
g
s
o
f
th
is
s
tu
d
y
ar
e
a
v
ailab
le
f
r
o
m
th
e
c
o
r
r
esp
o
n
d
in
g
au
t
h
o
r
,
[
SM]
,
u
p
o
n
r
ea
s
o
n
ab
le
r
eq
u
est.
RE
F
E
R
E
NC
E
S
[
1
]
Z.
Y
a
h
i
a
a
n
d
M
.
El
B
o
l
o
k
,
“
A
st
o
c
h
a
s
t
i
c
n
o
n
l
i
n
e
a
r
p
r
o
g
r
a
mm
i
n
g
m
o
d
e
l
f
o
r
b
u
d
g
e
t
m
i
x
o
p
t
i
m
i
z
a
t
i
o
n
o
f
d
i
g
i
t
a
l
m
a
r
k
e
t
i
n
g
c
a
m
p
a
i
g
n
s
u
n
d
e
r
u
n
c
e
r
t
a
i
n
t
y
,
”
F
u
t
u
re
Bu
s
i
n
e
ss
J
o
u
r
n
a
l
,
v
o
l
.
1
1
,
n
o
.
1
,
O
c
t
.
2
0
2
5
,
d
o
i
:
1
0
.
1
1
8
6
/
s4
3
0
9
3
-
0
2
5
-
0
0
6
6
4
-
x.
[
2
]
G
.
N
a
y
y
a
r
,
S
.
S
u
m
a
n
,
a
n
d
T.
R
.
K
.
La
k
sh
mi
,
“
O
p
t
i
mi
z
i
n
g
a
d
c
a
m
p
a
i
g
n
s
w
i
t
h
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
:
d
a
t
a
-
d
r
i
v
e
n
a
p
p
r
o
a
c
h
e
s
i
n
mo
d
e
r
n
me
d
i
a
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
M
u
l
t
i
p
h
y
s
i
c
s
,
v
o
l
.
1
8
,
n
o
.
3
,
p
p
.
1
7
4
6
–
1
7
5
4
,
2
0
2
4
,
d
o
i
:
1
0
.
5
2
7
8
3
/
i
j
m.
v
1
8
.
1
4
8
9
.
[
3
]
L.
M
a
a
n
d
B
.
S
u
n
,
“
M
a
c
h
i
n
e
l
e
a
r
n
i
n
g
a
n
d
A
I
i
n
mar
k
e
t
i
n
g
–
c
o
n
n
e
c
t
i
n
g
c
o
mp
u
t
i
n
g
p
o
w
e
r
t
o
h
u
m
a
n
i
n
s
i
g
h
t
s,”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
f
Re
s
e
a
r
c
h
i
n
Ma
r
k
e
t
i
n
g
,
v
o
l
.
3
7
,
n
o
.
3
,
p
p
.
4
8
1
–
5
0
4
,
S
e
p
.
2
0
2
0
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
i
j
r
e
smar
.
2
0
2
0
.
0
4
.
0
0
5
.
[
4
]
A
.
M
i
k
l
o
s
i
k
a
n
d
N
.
E
v
a
n
s,
“
I
mp
a
c
t
o
f
b
i
g
d
a
t
a
a
n
d
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
o
n
d
i
g
i
t
a
l
t
r
a
n
sf
o
r
m
a
t
i
o
n
i
n
mar
k
e
t
i
n
g
:
a
l
i
t
e
r
a
t
u
r
e
r
e
v
i
e
w
,
”
I
EEE
A
c
c
e
ss
,
v
o
l
.
8
,
p
p
.
1
0
1
2
8
4
–
1
0
1
2
9
2
,
2
0
2
0
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
0
.
2
9
9
8
7
5
4
.
[
5
]
V
.
A
.
B
r
e
i
,
“
M
a
c
h
i
n
e
l
e
a
r
n
i
n
g
i
n
ma
r
k
e
t
i
n
g
:
o
v
e
r
v
i
e
w
,
l
e
a
r
n
i
n
g
st
r
a
t
e
g
i
e
s
,
a
p
p
l
i
c
a
t
i
o
n
s,
a
n
d
f
u
t
u
r
e
d
e
v
e
l
o
p
me
n
t
s,”
F
o
u
n
d
a
t
i
o
n
s
a
n
d
T
r
e
n
d
s®
i
n
Ma
r
k
e
t
i
n
g
,
v
o
l
.
1
4
,
n
o
.
3
,
p
p
.
1
7
3
–
2
3
6
,
A
u
g
.
2
0
2
0
,
d
o
i
:
1
0
.
1
5
6
1
/
1
7
0
0
0
0
0
0
6
5
.
[
6
]
D
.
H
e
r
h
a
u
se
n
,
S
.
F
.
B
e
r
n
r
i
t
t
e
r
,
E.
W
.
T.
N
g
a
i
,
A
.
K
u
mar,
a
n
d
D
.
D
e
l
e
n
,
“
M
a
c
h
i
n
e
l
e
a
r
n
i
n
g
i
n
mark
e
t
i
n
g
:
r
e
c
e
n
t
p
r
o
g
r
e
ss
a
n
d
f
u
t
u
r
e
r
e
s
e
a
r
c
h
d
i
r
e
c
t
i
o
n
s,”
J
o
u
r
n
a
l
o
f
Bu
s
i
n
e
ss
Re
se
a
r
c
h
,
v
o
l
.
1
7
0
,
Ja
n
.
2
0
2
4
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
j
b
u
sr
e
s.
2
0
2
3
.
1
1
4
2
5
4
.
[
7
]
N
.
C
h
e
n
,
“
R
e
se
a
r
c
h
o
n
e
-
c
o
mm
e
r
c
e
d
a
t
a
b
a
se
mark
e
t
i
n
g
b
a
s
e
d
o
n
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
a
l
g
o
r
i
t
h
m,”
C
o
m
p
u
t
a
t
i
o
n
a
l
I
n
t
e
l
l
i
g
e
n
c
e
a
n
d
N
e
u
ro
s
c
i
e
n
c
e
,
v
o
l
.
2
0
2
2
,
n
o
.
1
,
p
p
.
1
–
1
3
,
J
u
n
.
2
0
2
2
,
d
o
i
:
1
0
.
1
1
5
5
/
2
0
2
2
/
7
9
7
3
4
4
6
.
[
8
]
F
.
N
o
r
d
i
n
a
n
d
A
.
R
a
v
a
l
d
,
“
Th
e
m
a
k
i
n
g
o
f
mar
k
e
t
i
n
g
d
e
c
i
si
o
n
s
i
n
m
o
d
e
r
n
mark
e
t
i
n
g
e
n
v
i
r
o
n
me
n
t
s
,
”
J
o
u
rn
a
l
o
f
B
u
s
i
n
e
s
s
Re
se
a
rc
h
,
v
o
l
.
1
6
2
,
Ju
l
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
j
b
u
sr
e
s.
2
0
2
3
.
1
1
3
8
7
2
.
[
9
]
E.
W
.
T.
N
g
a
i
a
n
d
Y
.
W
u
,
“
M
a
c
h
i
n
e
l
e
a
r
n
i
n
g
i
n
mar
k
e
t
i
n
g
:
a
l
i
t
e
r
a
t
u
r
e
r
e
v
i
e
w
,
c
o
n
c
e
p
t
u
a
l
f
r
a
mew
o
r
k
,
a
n
d
r
e
sea
r
c
h
a
g
e
n
d
a
,
”
J
o
u
rn
a
l
o
f
Bu
s
i
n
e
ss
Re
s
e
a
r
c
h
,
v
o
l
.
1
4
5
,
p
p
.
3
5
–
4
8
,
J
u
n
.
2
0
2
2
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
j
b
u
sr
e
s.
2
0
2
2
.
0
2
.
0
4
9
.
[
1
0
]
A
.
M
i
k
l
o
s
i
k
,
M
.
K
u
c
h
t
a
,
N
.
E
v
a
n
s,
a
n
d
S
.
Za
k
,
“
T
o
w
a
r
d
s
t
h
e
a
d
o
p
t
i
o
n
o
f
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
-
b
a
se
d
a
n
a
l
y
t
i
c
a
l
t
o
o
l
s
i
n
d
i
g
i
t
a
l
mark
e
t
i
n
g
,
”
I
EEE
Ac
c
e
ss
,
v
o
l
.
7
,
p
p
.
8
5
7
0
5
–
8
5
7
1
8
,
2
0
1
9
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
1
9
.
2
9
2
4
4
2
5
.
[
1
1
]
T.
K
.
V
a
sh
i
sh
t
h
,
V
.
S
h
a
r
ma,
K
.
K
.
S
h
a
r
ma,
B
.
K
u
m
a
r
,
S
.
C
h
a
u
d
h
a
r
y
,
a
n
d
R
.
P
a
n
w
a
r
,
“
Em
b
r
a
c
i
n
g
A
I
a
n
d
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
f
o
r
t
h
e
f
u
t
u
r
e
o
f
d
i
g
i
t
a
l
mar
k
e
t
i
n
g
,
”
i
n
AI
,
Bl
o
c
k
c
h
a
i
n
,
a
n
d
M
e
t
a
v
e
rse
i
n
H
o
s
p
i
t
a
l
i
t
y
a
n
d
T
o
u
r
i
sm
I
n
d
u
s
t
ry
4
.
0
,
N
e
w
Y
o
r
k
:
C
h
a
p
ma
n
a
n
d
H
a
l
l
/
C
R
C
,
2
0
2
4
,
p
p
.
9
0
–
1
1
7
,
d
o
i
:
1
0
.
1
2
0
1
/
9
7
8
1
0
3
2
7
0
6
4
7
4
-
7.
[
1
2
]
D
.
A
n
d
a
y
a
n
i
,
M
.
M
a
d
a
n
i
,
H
.
A
g
u
st
i
a
n
,
N
.
S
e
p
t
i
a
n
i
,
a
n
d
L.
W
.
M
i
n
g
,
“
O
p
t
i
m
i
z
i
n
g
d
i
g
i
t
a
l
mar
k
e
t
i
n
g
s
t
r
a
t
e
g
i
e
s
t
h
r
o
u
g
h
b
i
g
d
a
t
a
a
n
d
mac
h
i
n
e
l
e
a
r
n
i
n
g
:
i
n
s
i
g
h
t
s
a
n
d
a
p
p
l
i
c
a
t
i
o
n
s,
”
J
o
u
r
n
a
l
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
a
n
d
T
e
c
h
n
o
l
o
g
y
A
p
p
l
i
c
a
t
i
o
n
,
v
o
l
.
1
,
n
o
.
2
,
p
p
.
1
0
4
–
1
1
0
,
A
u
g
.
2
0
2
4
,
d
o
i
:
1
0
.
3
3
0
5
0
/
c
o
r
i
si
n
t
a
.
v
1
i
2
.
2
9
.
[
1
3
]
N
.
S
.
S
a
b
a
,
R
.
G
a
n
d
h
i
,
S
.
R
.
R
a
j
e
n
d
r
a
n
,
a
n
d
N
.
D
.
A
b
r
a
h
a
m
,
“
R
e
v
o
l
u
t
i
o
n
i
z
i
n
g
d
i
g
i
t
a
l
mar
k
e
t
i
n
g
u
s
i
n
g
mac
h
i
n
e
l
e
a
r
n
i
n
g
,
”
i
n
C
o
n
t
e
m
p
o
r
a
ry
Ap
p
r
o
a
c
h
e
s
o
f
D
i
g
i
t
a
l
M
a
rk
e
t
i
n
g
a
n
d
t
h
e
R
o
l
e
o
f
M
a
c
h
i
n
e
I
n
t
e
l
l
i
g
e
n
c
e
,
N
e
w
Y
o
r
k
,
U
n
i
t
e
d
S
t
a
t
e
s
:
I
G
I
G
l
o
b
a
l
S
c
i
e
n
t
i
f
i
c
P
u
b
l
i
s
h
i
n
g
,
2
0
2
3
,
p
p
.
1
–
22
,
d
o
i
:
1
0
.
4
0
1
8
/
9
7
8
-
1
-
6
6
8
4
-
7
7
3
5
-
9
.
c
h
0
0
1
.
[
1
4
]
T.
S
.
K
u
mar,
“
D
a
t
a
m
i
n
i
n
g
-
b
a
s
e
d
mark
e
t
i
n
g
d
e
c
i
si
o
n
s
u
p
p
o
r
t
s
y
st
e
m
u
si
n
g
h
y
b
r
i
d
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
a
l
g
o
r
i
t
h
m,”
J
o
u
r
n
a
l
o
f
Art
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
a
n
d
C
a
p
s
u
l
e
N
e
t
w
o
rks
,
v
o
l
.
2
,
n
o
.
3
,
p
p
.
1
8
5
–
1
9
3
,
A
u
g
.
2
0
2
0
,
d
o
i
:
1
0
.
3
6
5
4
8
/
j
a
i
c
n
.
2
0
2
0
.
3
.
0
0
6
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
1
5
,
No
.
4
,
Au
g
u
s
t 2
0
2
6
:
3
5
2
8
-
3
5
3
6
3536
[
1
5
]
M
.
Za
g
h
l
o
u
l
,
S
.
B
a
r
a
k
a
t
,
a
n
d
A
.
R
e
z
k
,
“
P
r
e
d
i
c
t
i
n
g
E
-
c
o
mm
e
r
c
e
c
u
st
o
m
e
r
sa
t
i
sf
a
c
t
i
o
n
:
t
r
a
d
i
t
i
o
n
a
l
mac
h
i
n
e
l
e
a
r
n
i
n
g
v
s
.
d
e
e
p
l
e
a
r
n
i
n
g
a
p
p
r
o
a
c
h
e
s,
”
J
o
u
r
n
a
l
o
f
Re
t
a
i
l
i
n
g
a
n
d
C
o
n
s
u
m
e
r
S
e
rv
i
c
e
s
,
v
o
l
.
7
9
,
Ju
l
.
2
0
2
4
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
j
r
e
t
c
o
n
s
e
r
.
2
0
2
4
.
1
0
3
8
6
5
.
[
1
6
]
L.
S
a
d
r
n
i
a
,
“
T
h
e
f
u
t
u
r
e
o
f
mar
k
e
t
i
n
g
:
h
o
w
p
r
e
d
i
c
t
i
v
e
mo
d
e
l
i
n
g
o
p
t
i
m
i
z
e
s
c
a
m
p
a
i
g
n
st
r
a
t
e
g
i
e
s,”
i
B
u
si
n
e
ss
,
v
o
l
.
1
5
,
n
o
.
4
,
p
p
.
2
4
9
–
2
6
2
,
2
0
2
3
,
d
o
i
:
1
0
.
4
2
3
6
/
i
b
.
2
0
2
3
.
1
5
4
0
1
8
.
[
1
7
]
V
.
G
o
o
l
j
a
r
,
T.
I
ssa,
S
.
H
.
-
R
a
m
a
n
a
n
,
a
n
d
B
.
A
.
-
S
a
l
i
h
,
“
S
e
n
t
i
m
e
n
t
-
b
a
se
d
p
r
e
d
i
c
t
i
v
e
m
o
d
e
l
s
f
o
r
o
n
l
i
n
e
p
u
r
c
h
a
s
e
s
i
n
t
h
e
e
r
a
o
f
mark
e
t
i
n
g
5
.
0
:
a
s
y
st
e
ma
t
i
c
r
e
v
i
e
w
,
”
J
o
u
rn
a
l
o
f
Bi
g
D
a
t
a
,
v
o
l
.
1
1
,
n
o
.
1
,
A
u
g
.
2
0
2
4
,
d
o
i
:
1
0
.
1
1
8
6
/
s4
0
5
3
7
-
0
2
4
-
0
0
9
4
7
-
0.
[
1
8
]
T.
S
a
n
g
sawa
n
g
,
“
P
r
e
d
i
c
t
i
n
g
a
d
c
l
i
c
k
-
t
h
r
o
u
g
h
r
a
t
e
s
i
n
d
i
g
i
t
a
l
m
a
r
k
e
t
i
n
g
u
si
n
g
s
u
p
p
o
r
t
v
e
c
t
o
r
m
a
c
h
i
n
e
s
,
”
J
o
u
r
n
a
l
o
f
D
i
g
i
t
a
l
M
a
rk
e
t
a
n
d
D
i
g
i
t
a
l
C
u
rre
n
c
y
,
v
o
l
.
1
,
n
o
.
3
,
p
p
.
2
2
5
–
2
4
6
,
D
e
c
.
2
0
2
4
,
d
o
i
:
1
0
.
4
7
7
3
8
/
j
d
md
c
.
v
1
i
3
.
2
0
.
[
1
9
]
B
.
H
.
H
a
y
a
d
i
a
n
d
I
.
M
.
M
.
E
l
Em
a
r
y
,
“
P
r
e
d
i
c
t
i
n
g
c
a
mp
a
i
g
n
R
O
I
u
si
n
g
d
e
c
i
si
o
n
t
r
e
e
s
a
n
d
r
a
n
d
o
m
f
o
r
e
st
s
i
n
d
i
g
i
t
a
l
mar
k
e
t
i
n
g
,
”
J
o
u
rn
a
l
o
f
D
i
g
i
t
a
l
M
a
rk
e
t
a
n
d
D
i
g
i
t
a
l
C
u
rre
n
c
y
,
v
o
l
.
1
,
n
o
.
1
,
p
p
.
1
–
2
0
,
Ju
n
.
2
0
2
4
,
d
o
i
:
1
0
.
4
7
7
3
8
/
j
d
m
d
c
.
v
1
i
1
.
5
.
[
2
0
]
K
.
Ji
n
,
Z.
Z
.
Z
h
o
n
g
,
E.
Y
.
Z
h
a
o
,
“
S
u
st
a
i
n
a
b
l
e
d
i
g
i
t
a
l
mark
e
t
i
n
g
u
n
d
e
r
b
i
g
d
a
t
a
:
a
n
A
I
r
a
n
d
o
m
f
o
r
e
st
mo
d
e
l
a
p
p
r
o
a
c
h
,
”
I
EE
E
T
ra
n
s
a
c
t
i
o
n
s
o
n
E
n
g
i
n
e
e
r
i
n
g
M
a
n
a
g
e
m
e
n
t
,
v
o
l
.
7
1
,
p
p
.
3
5
6
6
–
3
5
7
9
,
2
0
2
4
,
d
o
i
:
1
0
.
1
1
0
9
/
TE
M
.
2
0
2
3
.
3
3
4
8
9
9
1
.
[
2
1
]
J.
H
.
F
r
i
e
d
ma
n
,
“
G
r
e
e
d
y
f
u
n
c
t
i
o
n
a
p
p
r
o
x
i
ma
t
i
o
n
:
a
g
r
a
d
i
e
n
t
b
o
o
st
i
n
g
m
a
c
h
i
n
e
,
”
T
h
e
A
n
n
a
l
s o
f
S
t
a
t
i
s
t
i
c
s
,
v
o
l
.
2
9
,
n
o
.
5
,
O
c
t
.
2
0
0
1
,
d
o
i
:
1
0
.
1
2
1
4
/
a
o
s/
1
0
1
3
2
0
3
4
5
1
.
[
2
2
]
C
.
Ş
a
h
i
n
,
“
P
r
e
d
i
c
t
i
n
g
b
a
se
s
t
a
t
i
o
n
r
e
t
u
r
n
o
n
i
n
v
e
st
m
e
n
t
i
n
t
h
e
t
e
l
e
c
o
m
mu
n
i
c
a
t
i
o
n
s
i
n
d
u
st
r
y
:
ma
c
h
i
n
e
‐
l
e
a
r
n
i
n
g
a
p
p
r
o
a
c
h
e
s,
”
I
n
t
e
l
l
i
g
e
n
t
S
y
s
t
e
m
s
i
n
A
c
c
o
u
n
t
i
n
g
,
F
i
n
a
n
c
e
a
n
d
M
a
n
a
g
e
m
e
n
t
,
v
o
l
.
3
0
,
n
o
.
1
,
p
p
.
2
9
–
4
0
,
J
a
n
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
0
2
/
i
s
a
f
.
1
5
3
0
.
[
2
3
]
S
.
La
k
s
h
mi
n
a
r
a
y
a
n
a
n
,
R
.
S
e
r
a
n
m
a
d
e
v
i
,
N
.
B
.
M
u
d
d
a
n
g
a
l
a
,
S
.
S
u
s
e
n
d
i
r
a
n
,
S
.
S
.
S
u
n
d
a
r
,
a
n
d
B
.
N
a
m
a
si
v
a
y
a
m,
“
A
d
v
a
n
c
e
d
d
a
t
a
sci
e
n
c
e
t
e
c
h
n
i
q
u
e
s
f
o
r
d
y
n
a
m
i
c
o
p
t
i
mi
z
a
t
i
o
n
o
f
d
i
g
i
t
a
l
m
a
r
k
e
t
i
n
g
c
a
m
p
a
i
g
n
s:
l
e
v
e
r
a
g
i
n
g
r
e
i
n
f
o
r
c
e
m
e
n
t
l
e
a
r
n
i
n
g
,
N
LP,
a
n
d
g
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
,
”
i
n
3
r
d
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
O
p
t
i
m
i
z
a
t
i
o
n
T
e
c
h
n
i
q
u
e
s
i
n
t
h
e
Fi
e
l
d
o
f
E
n
g
i
n
e
e
ri
n
g
(
I
C
O
FE
-
2
0
2
4
)
,
2
0
2
5
,
p
p
.
1
–
11
,
d
o
i
:
1
0
.
2
1
3
9
/
ssr
n
.
5
0
8
6
6
6
6
.
[
2
4
]
K
.
C
h
e
n
,
L
.
D
o
n
g
,
a
n
d
L.
W
a
n
g
,
“
L
i
n
e
a
r
r
e
g
r
e
ssi
o
n
mo
d
e
l
f
o
r
b
u
si
n
e
ss
st
r
a
t
e
g
y
:
a
c
a
s
e
st
u
d
y
o
f
S
M
A
R
TFO
O
D
c
o
m
p
a
n
y
,
”
i
n
2
0
2
2
7
t
h
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
Fi
n
a
n
c
i
a
l
I
n
n
o
v
a
t
i
o
n
a
n
d
Ec
o
n
o
m
i
c
D
e
v
e
l
o
p
m
e
n
t
(
I
C
FI
ED
2
0
2
2
)
,
2
0
2
2
,
p
p
.
1
6
7
0
–
1
6
7
7
,
d
o
i
:
1
0
.
2
9
9
1
/
a
e
b
mr
.
k
.
2
2
0
3
0
7
.
2
7
3
.
[
2
5
]
Y
.
S
h
i
,
“
A
p
p
l
i
c
a
t
i
o
n
o
f
i
m
p
r
o
v
e
d
l
i
n
e
a
r
r
e
g
r
e
ss
i
o
n
a
l
g
o
r
i
t
h
m
i
n
b
u
si
n
e
ss
b
e
h
a
v
i
o
r
a
n
a
l
y
si
s
,
”
Pro
c
e
d
i
a
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
v
o
l
.
2
2
8
,
p
p
.
1
1
0
1
–
1
1
0
9
,
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
p
r
o
c
s.
2
0
2
3
.
1
1
.
1
4
4
.
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
Cha
n
d
r
a
Ch
a
th
u
r
a
is
a
sc
h
o
lar
o
f
G
ITAM
Un
iv
e
rsity
Be
n
g
a
lu
ru
,
Ka
rn
a
tak
a
,
In
d
ia
a
n
d
c
u
rre
n
tl
y
p
u
rsu
i
n
g
t
h
e
Ba
c
h
e
lo
r
o
f
Tec
h
n
o
l
o
g
y
d
e
g
re
e
in
Co
m
p
u
ter
S
c
ien
c
e
a
n
d
Bu
sin
e
ss
S
y
ste
m
s.
He
r
a
c
a
d
e
m
ic
a
n
d
p
r
o
jec
t
wo
r
k
re
flec
t
a
stro
n
g
in
tere
st
in
d
a
ta
sc
ien
c
e
a
n
d
m
a
c
h
in
e
lea
rn
in
g
,
wit
h
a
p
a
rti
c
u
lar
fo
c
u
s
o
n
t
h
e
ir
a
p
p
li
c
a
ti
o
n
s
in
b
u
si
n
e
ss
.
He
r
re
se
a
rc
h
in
tere
sts
li
e
a
t
th
e
in
ters
e
c
ti
o
n
o
f
d
a
ta
sc
ien
c
e
,
b
u
sin
e
ss
i
n
telli
g
e
n
c
e
,
m
a
rk
e
ti
n
g
a
n
a
ly
ti
c
s,
a
n
d
a
p
p
li
e
d
m
a
c
h
in
e
lea
rn
i
n
g
.
S
h
e
is
is
p
a
ss
io
n
a
te
a
b
o
u
t
c
o
m
b
i
n
i
n
g
tec
h
n
ica
l
e
x
p
e
rti
se
with
b
u
sin
e
ss
i
n
sig
h
ts.
S
h
e
c
a
n
b
e
c
o
n
t
a
c
ted
a
t
e
m
a
il
:
wo
rk
with
c
h
a
th
u
ra
c
h
a
n
d
ra
@g
m
a
il
.
c
o
m
.
K
e
e
r
th
a
n
S
a
y
a
is
a
sc
h
o
lar
o
f
G
ITAM
U
n
iv
e
rsit
y
Be
n
g
a
lu
r
u
,
Ka
rn
a
tak
a
,
In
d
ia
a
n
d
c
u
rre
n
tl
y
p
u
rsu
i
n
g
th
e
B.
Tec
h
.
d
e
g
re
e
i
n
Co
m
p
u
ter
S
c
ien
c
e
a
n
d
Bu
si
n
e
ss
S
y
ste
m
s.
He
h
a
s
p
a
rti
c
ip
a
ted
i
n
p
r
o
jec
ts
e
x
p
lo
r
in
g
th
e
in
ters
e
c
ti
o
n
o
f
b
u
sin
e
ss
a
n
a
ly
ti
c
s
a
n
d
fin
a
n
c
ial
m
o
d
e
li
n
g
,
a
imin
g
t
o
d
e
riv
e
a
c
ti
o
n
a
b
le
in
s
ig
h
ts
fo
r
stra
teg
ic
p
lan
n
in
g
.
Hi
s
in
tere
sts
li
e
i
n
f
in
a
n
c
ia
l
a
n
a
ly
ti
c
s,
b
u
si
n
e
ss
in
telli
g
e
n
c
e
,
F
in
Tec
h
,
a
n
d
th
e
a
p
p
l
ica
ti
o
n
o
f
d
a
t
a
-
d
riv
e
n
m
e
th
o
d
s in
m
o
d
e
rn
b
u
sin
e
ss
e
n
v
iro
n
m
e
n
ts.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
k
e
e
rth
a
n
7
7
5
@g
m
a
il
.
c
o
m
.
S
a
th
ish
k
u
m
a
r
M
a
n
i
h
a
s
o
b
tain
e
d
h
is
B.
E.
d
e
g
re
e
in
C
o
m
p
u
ter
S
c
ien
c
e
a
n
d
En
g
i
n
e
e
rin
g
fro
m
B
h
a
ra
th
iar
U
n
i
v
e
rsity
,
C
o
imb
a
t
o
re
,
In
d
ia
a
n
d
M
.
Tec
h
.
d
e
g
re
e
i
n
I
n
fo
rm
a
ti
o
n
Tec
h
n
o
l
o
g
y
fro
m
P
u
n
jab
i
Un
i
v
e
rsity
,
P
a
ti
a
la,
In
d
ia.
He
e
a
rn
e
d
h
is
P
h
.
D.
i
n
Co
m
p
u
ter
S
c
ien
c
e
a
n
d
E
n
g
i
n
e
e
rin
g
fro
m
S
a
v
e
e
th
a
Un
iv
e
rsit
y
,
C
h
e
n
n
a
i,
In
d
ia.
H
e
is
c
u
rre
n
tl
y
a
n
a
ss
o
c
iate
p
ro
fe
ss
o
r
in
t
h
e
De
p
a
rtme
n
t
o
f
Co
m
p
u
ter
S
c
ien
c
e
a
n
d
E
n
g
i
n
e
e
rin
g
a
t
G
ITAM
Un
iv
e
rsity
in
Be
n
g
a
lu
r
u
,
I
n
d
ia
.
He
h
a
s
o
v
e
r
2
5
y
e
a
rs
o
f
e
x
p
e
rien
c
e
in
m
u
l
ti
p
le
d
o
m
a
in
s
li
k
e
tea
c
h
in
g
,
re
se
a
rc
h
a
n
d
so
ftwa
re
d
e
v
e
lo
p
m
e
n
t.
His
re
se
a
rc
h
a
re
a
is
n
e
two
rk
se
c
u
rit
y
,
m
a
c
h
in
e
l
e
a
rn
in
g
,
a
n
d
i
n
tern
e
t
o
f
t
h
in
g
s
.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
sa
th
ish
k
u
m
a
r
m
a
n
i1
7
@g
m
a
il
.
c
o
m
.
Evaluation Warning : The document was created with Spire.PDF for Python.