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
.
13
, N
o.
4
,
D
e
c
e
m
be
r
2024
, pp.
4516
~
4525
I
S
S
N
:
2252
-
8938
,
D
O
I
:
10.11591/
ij
a
i.
v
13
.i
4
.pp
4516
-
4525
4516
Jou
r
n
al
h
om
e
page
:
ht
tp
:
//
ij
ai
.
ia
e
s
c
or
e
.c
om
A
c
c
u
r
ac
y b
ase
d
-
st
a
c
k
e
d
e
n
se
m
b
l
e
l
e
ar
n
i
n
g
m
o
d
e
l
f
or
t
h
e
p
r
e
d
i
c
t
i
on
of
c
or
on
ar
y
h
e
ar
t
d
i
se
ase
S
an
t
os
in
i
B
h
u
t
ia
,
B
ic
h
i
t
r
an
an
d
a
P
at
r
a,
M
it
r
ab
in
d
a
R
ay
D
e
pa
r
t
m
e
nt
of
C
om
put
e
r
S
c
i
e
nc
e
a
nd
E
ngi
ne
e
r
i
ng,
S
i
ks
ha
‘
O
’
A
nus
a
ndha
n
U
n
i
ve
r
s
i
t
y,
O
di
s
ha
,
I
ndi
a
A
r
t
ic
le
I
n
f
o
A
B
S
T
R
A
C
T
A
r
ti
c
le
h
is
to
r
y
:
R
e
c
e
iv
e
d
O
c
t
17
,
2023
R
e
vi
s
e
d
A
pr
1
,
2024
A
c
c
e
pt
e
d
A
pr
17
,
2024
Coronar
y
heart
disease
(CHD)
is
the
primary
cause
of
silen
t
and
noncommunicable
deaths.
Early
detection
is
essential
for
slowi
ng
the
progression
of
death
and
saving
lives.
Medical
researcher
s
use
m
achine
learning te
chniques to
predict CHD
. This artic
le propose
s an acc
uracy
based
-
stacked
ensemble
learning
(AB
-
SEL)
model
to
predict
CHD
while
minimizing
computatio
nal
time
(CT)
.
The
dataset
undergoes
the
l
ogistic
regression
recursive
feature
eliminati
on
(LR
-
RFE)
method
to
identify
the
important
feature
s.
The
three
strong
classifi
ers,
logistic
regressio
n
(LR),
random
forest
(RF),
and
AdaBoost,
are
chosen
to
build
ensemble
ma
chine
-
learning
models,
including
techniques
like
bagging,
majority
votin
g,
and
stacking,
for
the
Cleveland
dataset
accessible
from
Kaggle.
Data
scali
ng
was
done
using
the
normal
scalar
method,
and
hyperpar
ameter
optimizati
on
was
done
using
random
search
and
gri
d
search.
Effectiveness
is
measu
red
by
accuracy,
precision
,
recall,
F1
score,
and
CT
is
validated
through
5
-
fold
cross
-
validation.
The
suggested
approach
achieved
a
90.16%
accuracy
rate,
required
only 0.2 seconds of
CT
, and yielded an
area under t
he curve (
AUC
)
of 0.892.
K
e
y
w
o
r
d
s
:
H
e
a
r
t
di
s
e
a
s
e
M
a
c
hi
ne
l
e
a
r
ni
ng
E
ns
e
m
bl
e
l
e
a
r
ni
ng
L
ogi
s
ti
c
r
e
gr
e
s
s
io
n
-
r
e
c
ur
s
iv
e
f
e
a
tu
r
e
e
li
m
in
a
ti
on f
e
a
tu
r
e
s
e
le
c
ti
on
G
r
id
s
e
a
r
c
h
R
a
ndom s
e
a
r
c
h
5
-
f
ol
d c
r
os
s
-
va
li
da
ti
on
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
:
S
a
nt
os
in
i
B
hut
ia
D
e
pa
r
tm
e
nt
of
C
om
put
e
r
S
c
ie
nc
e
a
nd
E
ngi
ne
e
r
in
g,
S
ik
s
ha
‘
O
’
A
nus
a
ndha
n
U
ni
ve
r
s
it
y
J
-
15,
K
ha
nda
gi
r
i
M
a
r
g,
D
ha
r
a
m
V
ih
a
r
,
J
a
g
a
m
a
r
a
,
B
huba
ne
s
w
a
r
,
O
di
s
ha
751030,
I
ndi
a
E
m
a
il
:
s
a
nt
os
in
i.
bhut
ia
@
gm
a
il
.c
om
1.
I
N
T
R
O
D
U
C
T
I
O
N
C
or
ona
r
y
he
a
r
t
di
s
e
a
s
e
(
C
H
D
)
i
s
a
m
a
jo
r
c
hr
oni
c
di
s
e
a
s
e
.
I
t
in
c
l
ude
s
he
a
r
t
a
nd
bl
ood
ve
s
s
e
l
di
s
or
de
r
s
.
T
he
pr
im
a
r
y
f
unc
ti
on
of
th
e
he
a
r
t
is
to
pum
p
bl
ood
p
r
om
pt
ly
to
ke
e
p
hum
a
ns
a
li
ve
[
1]
.
I
f
th
e
he
a
r
t
f
a
il
s
,
e
s
s
e
nt
ia
l
or
ga
ns
s
uc
h a
s
th
e
br
a
in
m
a
y
c
e
a
s
e
to
f
unc
ti
on,
le
a
di
n
g
to
a
pot
e
nt
ia
l
s
hut
dow
n, a
nd
th
e
pe
r
s
on
m
a
y
s
uc
c
um
b.
P
oor
di
e
t,
he
a
vy a
lc
ohol
c
on
s
um
pt
io
n, pa
ti
e
nt
s
e
x, a
n
d a
ge
a
r
e
a
ll
r
is
k f
a
c
to
r
s
[
2]
. T
he
s
e
f
a
c
to
r
s
c
a
n
ne
ga
ti
ve
ly
im
pa
c
t
bl
ood
pr
e
s
s
ur
e
,
bl
ood
s
uga
r
le
ve
ls
,
a
nd
bl
ood
li
pi
d
pr
of
il
e
s
a
nd
c
a
us
e
obe
s
it
y.
T
he
s
e
a
s
pe
c
t
s
le
a
d t
o
a
r
is
e
i
n he
a
r
t
di
s
or
de
r
s
a
nd
ot
he
r
c
om
pl
e
xi
ti
e
s
[
3]
. T
he
W
or
ld
H
e
a
lt
h O
r
ga
ni
s
a
ti
on (
W
H
O
)
c
la
im
s
t
h
a
t
th
e
s
e
c
or
ona
r
y he
a
r
t
di
s
or
de
r
s
a
r
e
t
he
l
e
a
di
ng globa
l
c
a
us
e
of
de
a
th
. C
H
D
c
a
u
s
e
d 17.9 mi
ll
io
n de
a
th
s
i
n 2019,
38%
of
a
ll
f
a
ta
li
ti
e
s
.
S
in
c
e
th
e
s
e
num
be
r
s
a
r
e
e
xpe
c
t
e
d
to
a
ppr
oa
c
h
21
m
il
li
on
in
th
e
c
om
in
g
de
c
a
de
,
th
is
s
houl
d
be
a
ddr
e
s
s
e
d
s
e
r
io
us
ly
[
4]
.
E
le
c
tr
oni
c
de
vi
c
e
s
,
li
ke
e
le
c
tr
oc
a
r
di
ogr
a
m
s
a
nd
c
om
put
e
d
to
m
ogr
a
phy
s
c
a
ns
,
a
r
e
us
e
d
to
de
te
c
t
C
H
D
.
B
ut
th
e
s
e
de
vi
c
e
s
a
r
e
ve
r
y
e
xpe
n
s
iv
e
.
I
n
poor
e
r
c
ount
r
ie
s
,
th
e
r
e
m
a
y
not
be
e
nough
doc
to
r
s
to
di
a
gnos
e
C
H
D
pa
ti
e
nt
s
.
D
ia
gnos
ti
c
te
s
ts
c
a
n
be
e
r
r
one
ous
due
to
in
f
r
a
s
tr
uc
tu
r
a
l
is
s
ue
s
,
c
a
us
in
g
m
or
e
pr
obl
e
m
s
a
nd
pa
ti
e
nt
de
a
th
s
[
5]
.
I
t
a
ls
o
ta
ke
s
m
or
e
c
om
put
a
ti
o
na
l
ti
m
e
(
C
T
)
f
or
a
s
s
e
s
s
m
e
nt
s
.
T
he
r
e
f
or
e
,
it
is
c
r
uc
ia
l
to
ha
ve
a
r
e
li
a
bl
e
a
nd
pr
e
c
is
e
m
a
c
hi
ne
l
e
a
r
ni
ng
(
M
L
)
m
ode
l
f
or
pr
e
di
c
ti
ng
C
H
D
a
t
a
n
e
a
r
ly
s
ta
g
e
to
pr
e
ve
nt
de
a
th
.
H
ow
e
ve
r
,
le
ve
r
a
gi
ng
M
L
te
c
hni
que
s
to
pr
e
di
c
t
C
H
D
is
a
c
r
it
ic
a
l
c
ha
ll
e
nge
,
a
nd
s
e
le
c
ti
ng
th
e
opt
im
um
f
e
a
tu
r
e
s
f
r
om
t
he
da
ta
s
e
t
is
a
not
he
r
c
ha
ll
e
nge
i
n d
e
ve
l
opi
ng M
L
m
ode
ls
.
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
c
c
ur
ac
y
bas
e
d
-
s
ta
c
k
e
d
e
ns
e
m
bl
e
l
e
ar
ni
ng m
ode
l
fo
r
t
he
pr
e
di
c
ti
on of c
or
onar
y
…
(
Sant
os
in
i
B
hut
ia
)
4517
E
ns
e
m
bl
e
l
e
a
r
ni
ng me
th
ods
i
n M
L
a
r
e
be
in
g
us
e
d i
n s
e
ve
r
a
l
m
e
di
c
a
l
s
e
c
to
r
s
, a
nd f
e
a
tu
r
e
s
e
le
c
ti
on i
s
c
r
uc
ia
l
to
th
e
ir
pe
r
f
o
r
m
a
nc
e
[
6]
,
[
7
]
.
T
he
f
e
a
tu
r
e
s
e
le
c
ti
on
a
ppr
oa
c
h
e
li
m
in
a
te
s
c
e
r
ta
in
unne
c
e
s
s
a
r
y
f
e
a
tu
r
e
s
th
a
t
m
a
y
hi
nde
r
a
lg
or
it
hm
pe
r
f
or
m
a
nc
e
[
8]
,
[
9]
.
C
hoo
s
in
g
a
n
a
ppr
opr
ia
te
f
e
a
tu
r
e
s
ubs
e
t
a
nd
c
l
a
s
s
if
ic
a
ti
on
a
lg
or
it
hm
im
pr
ove
s
th
e
m
ode
l'
s
pe
r
f
or
m
a
nc
e
.
E
ns
e
m
bl
e
te
c
hni
que
s
li
ke
ba
ggi
ng,
boos
ti
ng,
m
a
jo
r
it
y
vot
in
g
,
a
nd
s
ta
c
ki
ng
a
r
e
r
e
por
te
d
by
L
a
th
a
a
nd
J
e
e
va
[
10]
us
in
g
th
e
C
le
ve
la
nd
he
a
r
t
di
s
e
a
s
e
da
ta
s
e
t.
T
he
s
e
e
ns
e
m
bl
e
m
ode
ls
w
e
r
e
us
e
d
to
e
nha
nc
e
th
e
c
la
s
s
if
ic
a
ti
on
a
c
c
ur
a
c
y
of
w
e
a
k
c
la
s
s
if
ie
r
s
.
T
he
s
tu
dy
f
ound
th
a
t
m
a
jo
r
it
y
vot
in
g
e
ns
e
m
bl
e
s
of
w
e
a
k
c
la
s
s
if
ie
r
s
im
pr
ove
d
a
c
c
ur
a
c
y
by
7.26%
,
w
he
r
e
a
s
f
e
a
tu
r
e
s
e
le
c
ti
on
a
ppr
oa
c
he
s
boos
te
d
pe
r
f
or
m
a
nc
e
e
v
e
n
m
or
e
.
U
ddi
n
a
nd
H
a
ld
e
r
[
11]
pr
o
pos
e
d
a
n
e
ns
e
m
bl
e
-
ba
s
e
d
m
ul
ti
la
ye
r
dyn
a
m
ic
s
ys
te
m
(
M
L
D
S
)
f
or
be
tt
e
r
c
a
r
di
ova
s
c
ul
a
r
di
s
e
a
s
e
pr
e
di
c
ti
on.
F
e
a
tu
r
e
s
e
le
c
ti
on
m
e
th
ods
li
ke
th
e
c
or
r
e
la
ti
on,
ga
in
r
a
ti
o,
in
f
or
m
a
ti
on
ga
in
,
la
s
s
o,
a
nd
e
xt
r
a
tr
e
e
s
c
la
s
s
if
ie
r
s
w
e
r
e
us
e
d
to
s
e
l
e
c
t
r
e
le
va
nt
f
e
a
tu
r
e
s
.
T
he
e
ns
e
m
bl
e
m
ode
l
w
a
s
s
ubs
e
que
nt
ly
c
r
e
a
te
d by u
s
in
g r
a
ndom f
or
e
s
t
(
R
F
)
,
n
a
ïv
e
B
a
ye
s
(
N
B
)
, a
nd
gr
a
di
e
nt
boo
s
ti
ng
. A
f
te
r
di
vi
di
ng 70,000 ins
ta
nc
e
s
i
nt
o 50:50,
60:
40, 70:30,
80:
20, a
nd
87.5:12.5, t
he
y got
be
tt
e
r
r
e
s
ul
ts
. F
u
r
th
e
r
m
or
e
,
th
e
a
r
e
a
und
e
r
th
e
c
ur
ve
(
AUC
)
c
ur
ve
m
e
a
s
ur
e
s
th
e
pr
oba
bi
l
it
y
of
a
c
c
ur
a
te
c
la
s
s
if
ic
a
ti
on.
G
a
o
e
t
al
.
[
12]
im
pr
ove
d
he
a
r
t
di
s
e
a
s
e
pr
e
di
c
ti
on
w
it
h
e
n
s
e
m
bl
e
le
a
r
ni
ng
m
e
t
hods
,
in
c
lu
di
ng
ba
ggi
ng
a
nd
boo
s
ti
ng.
T
o
f
in
d
im
por
ta
nt
f
e
a
tu
r
e
s
,
th
e
y
u
s
e
d
pr
in
c
ip
a
l
c
om
pone
nt
a
na
ly
s
i
s
(
P
C
A
)
a
nd
li
ne
a
r
di
s
c
r
im
in
a
nt
a
na
ly
s
is
(
L
D
A
)
.
T
h
e
a
ut
hor
s
a
ls
o
c
om
pa
r
e
d
th
e
s
e
m
e
th
ods
w
it
h
f
iv
e
tr
a
di
ti
ona
l
M
L
m
e
th
ods
us
in
g
va
r
io
us
pe
r
f
or
m
a
nc
e
m
e
tr
ic
s
.
T
he
ba
ggi
ng
e
ns
e
m
bl
e
le
a
r
ni
ng
a
ppr
oa
c
h,
u
s
in
g
th
e
de
c
i
s
io
n
t
r
e
e
(
D
T
)
c
la
s
s
if
ie
r
a
nd
P
C
A
f
e
a
tu
r
e
s
e
le
c
ti
on,
ha
s
98.6%
a
c
c
ur
a
c
y.
R
a
hi
m
e
t
al
.
[
13]
in
tr
oduc
e
d
th
e
m
a
c
hi
ne
le
a
r
ni
ng
-
ba
s
e
d
c
a
r
di
ova
s
c
ul
a
r
di
s
e
a
s
e
di
a
gnos
is
(
M
a
L
C
a
D
D
)
s
ys
te
m
.
T
he
y
s
ta
r
te
d
by
r
e
pl
a
c
in
g
m
is
s
in
g
da
t
a
w
it
h
th
e
m
e
a
n
r
e
pl
a
c
e
m
e
nt
te
c
hni
que
a
nd
im
ba
la
nc
e
d
da
ta
w
it
h
th
e
s
ynt
he
ti
c
m
in
or
it
y
ove
r
-
s
a
m
pl
in
g
te
c
h
ni
que
.
T
he
y
s
e
le
c
te
d
th
e
m
os
t
r
e
le
v
a
nt
f
e
a
tu
r
e
s
us
in
g S
e
le
c
tKB
e
s
t
a
nd
bui
lt
a
n e
n
s
e
m
bl
e
m
ode
l
u
s
in
g l
ogi
s
ti
c
r
e
gr
e
s
s
io
n
(
L
R
)
a
nd k
-
ne
a
r
e
s
t
ne
ig
hbor
(
K
N
N
)
.
O
n
C
le
ve
la
nd,
F
r
a
m
in
gha
m
,
a
nd
he
a
r
t
di
s
e
a
s
e
,
th
e
s
ys
t
e
m
a
c
hi
e
ve
d
95.5%
,
99.1%
,
a
nd
98.0%
a
c
c
ur
a
c
y.
M
ie
nye
e
t
al
.
[
14]
r
a
ndoml
y
pa
r
ti
ti
one
d
th
e
da
ta
s
e
t
a
nd
u
s
e
d
c
la
s
s
if
ic
a
ti
on
a
nd
r
e
gr
e
s
s
io
n
tr
e
e
(
C
A
R
T
)
to
m
ode
l
e
a
c
h
s
ubs
e
t.
S
ubs
e
qu
e
nt
ly
,
th
e
y
de
ve
lo
pe
d
a
hom
oge
ne
ous
e
ns
e
m
bl
e
m
ode
l
f
r
om
th
e
s
e
m
ul
ti
pl
e
C
A
R
T
m
ode
ls
by
a
ppl
yi
ng
a
n
a
c
c
ur
a
c
y
-
ba
s
e
d
w
e
ig
ht
e
d
a
gi
ng
(
A
B
-
W
A
E
)
c
la
s
s
i
f
ie
r
.
T
hi
s
m
ode
l
a
c
hi
e
ve
d
93%
a
nd
91%
c
la
s
s
if
ic
a
ti
on a
c
c
ur
a
c
y on the
C
le
v
e
la
nd a
nd F
r
a
m
in
gha
m
da
ta
s
e
ts
, r
e
s
pe
c
ti
ve
ly
.
A
c
c
or
di
ng
to
th
e
li
te
r
a
tu
r
e
,
m
a
ny
r
e
s
e
a
r
c
he
r
s
us
e
e
n
s
e
m
bl
e
a
ppr
oa
c
he
s
to
im
pr
ove
c
la
s
s
if
ic
a
ti
on
a
c
c
ur
a
c
y
ove
r
s
in
gl
e
c
la
s
s
if
ie
r
s
.
A
ddi
ti
ona
ll
y,
f
e
a
tu
r
e
s
e
le
c
ti
on
be
f
or
e
c
la
s
s
if
ic
a
ti
on
a
ls
o
im
pr
ove
s
m
od
e
l
e
f
f
ic
ie
nc
y.
H
ow
e
ve
r
,
th
e
y
di
d
not
pr
ovi
de
th
e
C
T
f
or
m
ode
l
f
it
ti
ng.
T
he
li
te
r
a
tu
r
e
r
e
vi
e
w
doe
s
not
e
xpl
a
in
hype
r
pa
r
a
m
e
te
r
tu
ni
ng’
s
e
f
f
e
c
ts
on
m
ode
l
s
.
B
a
s
e
d
on
m
ode
l
p
e
r
f
or
m
a
nc
e
a
nd
f
e
a
tu
r
e
in
te
r
a
c
ti
ons
,
L
R
w
it
h
r
e
c
ur
s
iv
e
f
e
a
tu
r
e
e
li
m
in
a
ti
on
(
R
F
E
)
r
e
m
ove
s
f
e
a
tu
r
e
s
r
e
pe
a
t
e
dl
y.
L
R
-
R
F
E
is
m
ode
l
-
dr
iv
e
n
a
nd
id
e
nt
if
ie
s
LR
-
r
e
le
va
nt
f
e
a
tu
r
e
s
, m
a
ki
ng
f
e
a
tu
r
e
s
e
le
c
ti
on mor
e
pe
r
s
ona
li
z
e
d a
nd opti
m
iz
e
d t
ha
n S
e
le
c
tKB
e
s
t
or
P
C
A
.
I
n
th
is
s
tu
dy,
th
e
L
R
-
R
F
E
f
e
a
tu
r
e
s
e
le
c
ti
on
te
c
hni
que
is
us
e
d
to
s
e
le
c
t
th
e
s
ig
ni
f
ic
a
nt
f
e
a
tu
r
e
s
f
or
pr
e
di
c
ti
ng
C
H
D
.
T
he
r
e
a
f
te
r
,
num
e
r
ous
M
L
a
lg
or
it
hm
s
a
r
e
e
xpl
or
e
d
to
s
e
le
c
t
th
e
th
r
e
e
hi
gh
-
a
c
c
ur
a
c
y
le
a
r
ne
r
s
to
bui
ld
e
ns
e
m
bl
e
le
a
r
ni
ng
m
ode
ls
s
uc
h
a
s
m
a
jo
r
it
y
vot
in
g,
ba
gg
in
g,
a
nd
s
ta
c
ki
ng.
C
om
pa
r
in
g
th
e
pe
r
f
or
m
a
nc
e
of
th
e
s
e
e
ns
e
m
bl
e
m
ode
ls
,
a
s
ta
c
ke
d
e
ns
e
m
bl
e
m
ode
l
na
m
e
d
t
he
a
c
c
ur
a
c
y
ba
s
e
d
-
s
ta
c
ke
d
e
n
s
e
m
bl
e
le
a
r
ni
ng
(
A
B
-
S
E
L
)
m
ode
l
is
pr
opos
e
d, w
hi
c
h pe
r
f
or
m
s
be
tt
e
r
f
or
pr
e
di
c
t
in
g C
H
D
w
it
h t
he
l
e
a
s
t
C
T
.
T
he
f
ol
lo
w
in
g i
s
a
s
um
m
a
r
y of
t
he
c
ont
r
ib
ut
io
ns
t
ha
t
th
is
a
r
ti
c
le
m
a
ke
s
:
−
T
o
e
xpl
or
e
s
ix
tr
a
di
ti
ona
l
M
L
m
e
th
ods
:
L
R
,
K
N
N
,
R
F
,
D
T
,
N
B
,
a
nd
s
uppor
t
ve
c
to
r
m
a
c
hi
ne
(
S
V
M
)
,
a
nd
f
our
boos
ti
ng
m
e
th
od
s
:
A
d
a
B
oos
t,
G
B
oos
t,
X
G
B
oos
t,
a
nd
L
ig
ht
G
B
M
.
T
o
im
pr
ove
m
od
e
l
a
c
c
ur
a
c
y
,
th
e
LR
-
R
F
E
f
e
a
tu
r
e
s
e
le
c
ti
on
te
c
hni
que
is
u
s
e
d
to
s
e
le
c
t
a
s
ig
ni
f
ic
a
nt
f
e
a
tu
r
e
s
ub
s
e
t.
T
he
hype
r
pa
r
a
m
e
te
r
s
a
r
e
opt
im
iz
e
d us
in
g
bot
h
gr
id
s
e
a
r
c
h a
nd r
a
ndom s
e
a
r
c
h a
ppr
oa
c
he
s
.
−
T
hr
e
e
hi
gh
-
a
c
c
ur
a
c
y
s
tr
ong
le
a
r
ne
r
s
-
L
R
,
R
F
,
a
nd
A
da
B
oo
s
t
-
a
r
e
c
hos
e
n
to
c
ons
tr
uc
t
ba
ggi
ng,
m
a
jo
r
it
y
vot
in
g,
a
nd
s
ta
c
ki
ng
e
n
s
e
m
bl
e
m
ode
ls
.
I
n
c
li
ni
c
a
l
a
ppl
ic
a
ti
on
s
,
e
ns
e
m
bl
e
m
od
e
l
s
a
r
e
e
v
a
lu
a
te
d
f
or
e
f
f
ic
ie
nc
y
us
in
g
m
e
a
s
ur
e
s
in
c
lu
di
ng
a
c
c
ur
a
c
y,
pr
e
c
is
io
n,
r
e
c
a
ll
,
F
1
s
c
or
e
,
C
T
,
a
nd
A
U
C
to
a
s
c
e
r
ta
in
th
e
m
os
t
s
ui
ta
bl
e
one
.
−
A
ll
m
ode
ls
a
r
e
c
ons
tr
uc
te
d
us
in
g
th
e
publ
ic
ly
a
c
c
e
s
s
ib
le
C
le
v
e
l
a
nd
da
ta
s
e
t,
w
hi
c
h
i
s
a
va
il
a
bl
e
on
K
a
ggl
e
.
P
yt
hon
-
ba
s
e
d J
upyt
e
r
N
ot
e
book is
e
m
pl
oye
d f
or
a
ll
da
ta
pr
oc
e
s
s
in
g a
nd c
om
put
a
ti
ona
l
ta
s
k
s
.
T
he
r
e
m
a
in
de
r
of
th
e
pa
pe
r
is
s
tr
uc
tu
r
e
d
a
s
f
ol
lo
w
s
:
s
e
c
ti
on
2
pr
ovi
de
s
a
de
ta
il
e
d
de
s
c
r
ip
ti
on
of
th
e
m
e
th
odol
ogy, highl
ig
ht
in
g t
he
pr
opos
e
d
a
ppr
oa
c
h f
or
pr
e
di
c
ti
n
g C
H
D
. I
t
a
ls
o
c
ove
r
s
t
he
c
om
pr
e
he
n
s
iv
e
s
tu
dy
pr
oc
e
dur
e
e
m
pl
oye
d t
o e
ns
ur
e
t
he
r
e
li
a
bi
li
ty
a
nd a
c
c
ur
a
c
y of
t
he
pr
e
di
c
ti
ons
. S
e
c
ti
on 3 pr
e
s
e
nt
s
t
he
f
in
di
ng
of
th
e
s
tu
dy,
di
s
c
us
s
io
n
a
nd
how
th
e
y
c
om
pa
r
e
w
it
h
e
xi
s
ti
ng
li
te
r
a
t
ur
e
.
F
in
a
ll
y,
s
e
c
ti
on
4
c
onc
lu
de
s
th
e
pa
pe
r
w
it
h
a
s
um
m
a
r
y, of
f
e
r
in
g f
in
a
l
r
e
m
a
r
ks
a
nd s
ugge
s
ti
ng pote
nt
ia
l
di
r
e
c
ti
ons
f
or
f
ut
u
r
e
r
e
s
e
a
r
c
h i
n t
hi
s
f
ie
ld
.
2.
M
E
T
H
O
D
O
L
O
G
Y
L
e
t
us
c
on
s
id
e
r
th
e
da
ta
s
e
t
as
=
{
(
,
)
,
=
1
,
2
,
3
,
…
}
,
w
he
r
e
is
th
e
in
de
pe
nde
nt
va
r
ia
bl
e
a
nd
de
pe
nde
nt
va
r
ia
bl
e
us
e
d
f
or
th
e
pr
e
di
c
ti
on
of
C
H
D
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
,
V
ol
.
13
, N
o.
4
,
D
e
c
e
m
be
r
2024
:
4516
-
4525
4518
=
[
1
,
2
,
3
,
…
,
]
a
nd
∈
{
0
,
1
}
T
he
c
on
c
e
pt
ua
l
f
r
a
m
e
w
or
k
f
or
th
e
pr
e
di
c
ti
on
of
C
H
D
is
s
how
n
in
F
ig
ur
e
1.
T
hi
s
m
e
th
odol
ogy
in
c
lu
de
s
f
e
a
tu
r
e
s
e
le
c
ti
on,
c
la
s
s
if
ie
r
m
ode
li
ng
w
it
h
hype
r
pa
r
a
m
e
t
e
r
opt
im
iz
a
ti
on,
va
li
da
ti
on,
a
nd
pe
r
f
or
m
a
nc
e
a
na
ly
s
is
.
T
he
f
ir
s
t
s
te
p
id
e
nt
if
ie
s
a
s
e
t
of
f
e
a
tu
r
e
s
th
a
t
a
r
e
m
o
s
t
r
e
le
va
nt
to
de
te
c
ti
ng
C
H
D
.
T
he
LR
-
R
F
E
m
e
th
od
is
us
e
d
to
r
e
duc
e
th
e
w
e
a
k
e
s
t
f
e
a
tu
r
e
s
unt
il
th
e
de
s
ir
e
d
num
be
r
is
r
e
a
c
h
e
d.
T
he
n
e
xt
pha
s
e
bui
ld
s
th
e
m
ode
l.
B
a
s
e
d
on
th
e
ir
le
a
r
ni
ng
ty
pe
,
s
ix
c
om
m
on
s
ta
nd
a
r
d
ML
m
e
th
ods
-
L
R
,
K
N
N
,
R
F
,
D
T
,
N
B
,
a
nd
S
V
M
-
a
nd
f
our
boos
ti
ng
a
lg
or
it
hm
s
-
A
da
B
oos
t,
G
B
oos
t,
X
G
B
oos
t,
a
nd
L
ig
ht
G
B
M
-
a
r
e
c
hos
e
n
to
e
xpl
or
e
.
T
he
a
f
or
e
m
e
nt
io
ne
d
m
ode
ls
a
r
e
opt
im
iz
e
d
us
in
g
gr
id
s
e
a
r
c
h
a
nd
r
a
ndom
s
e
a
r
c
h
w
it
h
5
-
f
ol
d
c
r
os
s
-
va
li
da
ti
on.
T
he
c
la
s
s
if
ic
a
ti
on
a
lg
or
it
hm
s
m
e
nt
io
ne
d
a
bove
f
or
m
th
e
ba
s
is
f
or
th
e
c
la
s
s
if
ic
a
ti
on
a
na
ly
s
is
a
nd
e
f
f
e
c
ti
ve
ne
s
s
c
om
pa
r
is
on
s
out
li
ne
d
in
s
e
c
ti
on
3.
T
he
t
hr
e
e
hi
gh
-
a
c
c
ur
a
c
y
s
tr
ong
le
a
r
ne
r
s
-
L
R
,
R
F
,
a
nd
A
da
B
oos
t
-
a
r
e
c
ho
s
e
n
to
bui
ld
ba
ggi
ng,
m
a
jo
r
it
y
vot
in
g,
a
nd
s
t
a
c
ki
ng
e
n
s
e
m
bl
e
m
ode
ls
.
F
in
a
ll
y
,
an
a
s
s
e
s
s
m
e
nt
is
c
a
r
r
ie
d
out
to
m
e
a
s
ur
e
th
e
a
c
c
ur
a
c
y,
pr
e
c
i
s
io
n,
r
e
c
a
ll
,
F1
s
c
or
e
,
CT
,
a
nd
AUC
of
th
e
s
e
e
ns
e
m
bl
e
m
ode
ls
.
F
ig
ur
e
1.
F
r
a
m
e
w
or
k
f
or
p
r
e
di
c
ti
on
of
C
H
D
2.1. Dat
as
e
t
T
hi
s
s
tu
dy
us
e
s
K
a
ggl
e
’
s
C
le
v
e
la
nd
da
ta
s
e
t
to
a
na
ly
z
e
C
H
D
.
T
a
bl
e
1
s
how
s
303
in
s
ta
n
c
e
s
a
nd
14
f
e
a
tu
r
e
s
or
a
tt
r
ib
ut
e
s
.
It
ha
s
165
C
H
D
pa
ti
e
nt
s
,
a
nd
th
e
r
e
m
a
in
de
r
a
r
e
he
a
lt
hy.
T
he
d
a
ta
s
e
t
ha
s
be
e
n
pr
e
pr
oc
e
s
s
e
d
us
in
g
th
e
s
ta
nd
a
r
d
s
c
a
le
r
m
e
th
od
(
1)
f
o
r
ML
a
lg
or
it
hm
s
.
T
a
bl
e
2
li
s
ts
a
ge
,
ge
nde
r
,
c
hol
e
s
te
r
ol
,
bl
ood
pr
e
s
s
ur
e
,
a
lc
ohol
us
e
,
di
a
be
te
s
,
a
nd
ot
he
r
he
a
lt
h
f
a
c
to
r
s
in
th
e
da
ta
s
e
t.
T
he
di
s
c
u
s
s
io
n
s
ugg
e
s
ts
th
a
t
5
-
f
ol
d
c
r
os
s
-
va
li
da
ti
on
ha
s
pr
oduc
e
d
th
e
m
os
t
a
c
c
ur
a
te
a
nd
in
s
i
ght
f
ul
out
c
om
e
s
f
or
th
e
s
a
m
e
da
ta
s
e
t.
S
ta
nda
r
di
z
a
ti
on:
′
=
−
̅
(
1)
w
he
r
e
̅
(
m
e
a
n)
=
(
1
)
∑
=
1
a
nd
σ
(
s
ta
nda
r
d
de
vi
a
ti
on)
=
√
1
∑
(
−
̅
)
2
=
1
T
a
bl
e
1
.
O
ve
r
vi
e
w
of
th
e
d
a
ta
s
e
t
D
a
t
a
s
e
t
I
ns
t
a
nc
e
s
A
t
t
r
i
but
e
s
C
l
e
ve
l
a
nd
303
14
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
c
c
ur
ac
y
bas
e
d
-
s
ta
c
k
e
d
e
ns
e
m
bl
e
l
e
ar
ni
ng m
ode
l
fo
r
t
he
pr
e
di
c
ti
on of c
or
onar
y
…
(
Sant
os
in
i
B
hut
ia
)
4519
T
a
bl
e
2
.
D
e
ta
il
e
d
a
tt
r
ib
ut
e
de
s
c
r
ip
ti
on
of
th
e
d
a
ta
s
e
t
A
t
t
r
i
but
e
c
ode
A
t
t
r
i
but
e
na
m
e
A
t
t
r
i
but
e
de
s
c
r
i
pt
i
on
a
ge
A
ge
A
ge
i
n ye
a
r
s
s
e
x
S
e
x
0
-
F
e
m
a
l
e
1
-
M
a
l
e
cp
C
he
s
t
pa
i
n
t
ype
s
1
-
T
ypi
c
a
l
A
ngi
na
2
-
A
t
ypi
c
a
l
A
ngi
na
3
-
N
on
-
A
ngi
na
P
a
i
n
4
-
A
s
ym
pt
om
a
t
i
c
P
a
i
n
t
r
e
s
t
bps
R
e
s
t
i
ng bl
ood pr
e
s
s
ur
e
R
e
s
t
i
ng bl
ood pr
e
s
s
ur
e
(
i
n m
m
H
g)
c
hol
C
hol
e
s
t
e
r
ol
S
e
r
um
c
hol
e
s
t
e
r
ol
(
i
n m
m
/
dl
)
f
bs
F
a
s
t
i
ng bl
ood s
uga
r
F
a
s
t
i
ng bl
ood s
uga
r
>120 mg/
dl
0
-
F
a
l
s
e
1
-
T
r
ue
r
e
s
t
e
c
g
R
e
s
t
i
ng e
l
e
c
t
r
oc
a
r
di
ogr
a
phi
c
R
e
s
t
i
ng
e
l
e
c
t
r
oc
a
r
di
ogr
a
phi
c
r
e
s
ul
t
s
V
a
l
ue
0
-
N
or
m
a
l
V
a
l
ue
1
-
H
a
vi
ng S
T
-
T
w
a
ve
a
bnor
m
a
l
i
t
y
V
a
l
ue
2
-
S
how
i
ng pr
oba
bl
e
or
de
f
i
ne
d l
e
f
t
ve
nt
r
i
c
ul
a
r
hype
r
t
r
ophy
t
ha
l
a
c
h
M
a
xi
m
um
he
a
r
t
r
a
t
e
M
a
xi
m
um
he
a
r
t
r
a
t
e
a
t
t
a
i
ne
d a
t
t
he
t
i
m
e
of
t
ha
l
l
i
um
t
e
s
t
e
xa
ng
E
xe
r
c
i
s
e
-
i
nduc
e
d a
ngi
na
0
-
No
1
-
Y
e
s
ol
dpe
a
k
S
T
de
pr
e
s
s
i
on
S
T
de
pr
e
s
s
i
on i
nduc
e
d by e
xe
r
c
i
s
e
r
e
l
a
t
i
ve
t
o r
e
s
t
s
l
ope
S
T
s
l
ope
T
he
s
l
ope
of
t
he
pe
a
k e
xe
r
c
i
s
e
S
T
s
e
gm
e
nt
1
-
U
p s
l
opi
ng
2
-
F
l
a
t
3
-
D
ow
n s
l
opi
ng
ca
N
um
be
r
of
m
a
j
or
ve
s
s
e
l
s
N
um
be
r
of
m
a
j
or
ve
s
s
e
l
s
(
0
-
3)
c
ol
or
e
d by
f
l
uor
os
c
opy
t
ha
l
T
ha
l
l
i
um
he
a
r
t
t
e
s
t
T
ha
l
a
s
s
e
m
i
a
va
l
ue
1
-
N
or
m
a
l
2
-
F
i
xe
d de
f
e
c
t
3
-
R
e
ve
r
s
i
bl
e
de
f
e
c
t
t
a
r
ge
t
H
e
a
r
t
di
s
e
a
s
e
0
-
P
a
t
i
e
nt
not
s
uf
f
e
r
i
ng f
r
om
he
a
r
t
di
s
e
a
s
e
1
-
P
a
t
i
e
nt
s
uf
f
e
r
i
ng f
r
om
he
a
r
t
di
s
e
a
s
e
2.2. F
e
at
u
r
e
s
e
le
c
t
io
n
F
e
a
tu
r
e
s
e
le
c
ti
on
r
e
duc
e
s
noi
s
y,
in
s
ig
ni
f
ic
a
nt
,
a
nd
r
e
dunda
nt
f
e
a
tu
r
e
s
,
im
pr
ovi
ng
pr
e
di
c
ti
on
m
ode
l
e
f
f
ic
ie
nc
y
[
15]
.
R
F
E
is
a
w
r
a
ppe
r
f
e
a
tu
r
e
s
e
le
c
ti
on
m
e
th
od
[
16]
.
T
he
w
r
a
ppe
r
m
e
th
od
ge
ne
r
a
te
s
a
f
e
a
tu
r
e
s
ubs
e
t
w
it
h
th
e
hi
ghe
s
t
c
la
s
s
if
ie
r
a
c
c
ur
a
c
y.
W
he
n
f
e
w
e
r
f
e
a
tu
r
e
s
a
r
e
e
m
pl
oye
d,
w
r
a
ppe
r
m
e
th
ods
pr
ovi
de
im
pr
ove
d
a
c
c
ur
a
c
y.
H
a
or
a
n
W
u
s
ugge
s
te
d
a
n
L
R
-
R
F
E
m
e
th
od
f
or
e
li
m
in
a
ti
ng
f
e
a
tu
r
e
s
a
nd
e
nha
nc
in
g
th
e
e
f
f
ic
ie
nc
y
of
th
e
m
ode
l
in
a
not
he
r
w
or
k
[
17]
.
F
or
th
is
r
e
a
s
on,
w
e
ha
ve
c
hos
e
n
L
R
-
R
F
E
in
our
s
tu
dy.
L
R
w
il
l
be
u
s
e
d
to
s
c
or
e
a
nd
r
a
nk
th
e
f
e
a
tu
r
e
s
.
T
he
L
R
-
R
F
E
a
lg
or
it
hm
r
e
duc
e
s
th
e
f
e
a
tu
r
e
s
r
e
c
ur
s
iv
e
ly
unt
il
th
e
ta
r
ge
t
num
be
r
i
s
r
e
a
c
he
d. T
hi
s
a
ppr
oa
c
h
i
nc
lu
de
s
t
he
f
ol
lo
w
in
g s
te
p
s
:
i)
L
R
i
s
u
s
e
d t
o f
it
t
he
m
ode
l
;
ii
)
r
a
nk
th
e
m
os
t
im
por
ta
nt
f
e
a
tu
r
e
s
;
ii
i)
e
li
m
in
a
te
th
e
le
a
s
t
im
por
ta
nt
f
e
a
tu
r
e
s
;
a
nd
iv
)
r
e
f
i
t
t
he
m
ode
l
unt
il
th
e
r
e
qui
r
e
d f
e
a
tu
r
e
s
a
r
e
f
ound.
2.3. B
as
e
m
od
e
ls
f
or
e
n
s
e
m
b
le
l
e
a
r
n
in
g
2.3.1.
L
ogi
s
t
ic
r
e
gr
e
s
s
io
n
T
he
m
os
t
popula
r
s
im
pl
e
m
e
th
od
f
or
c
la
s
s
if
ic
a
ti
on
in
s
upe
r
vi
s
e
d
ML
is
LR
[
18]
.
W
he
n
us
e
d
on
a
c
a
te
gor
ic
a
l
de
pe
nde
nt
v
a
r
ia
bl
e
, t
he
r
e
s
ul
t
c
a
n b
e
di
s
c
r
e
te
or
bi
n
a
r
y. T
he
LR
m
ode
l
us
e
s
a
s
ig
m
oi
d f
unc
ti
on (
2)
a
s
a
c
os
t
f
unc
ti
on
in
s
te
a
d
of
a
li
ne
a
r
f
unc
ti
on.
T
he
s
ig
m
oi
d
f
u
nc
ti
on
c
onve
r
ts
a
pr
e
di
c
te
d
a
c
tu
a
l
va
lu
e
in
to
a
pr
oba
bi
li
s
ti
c
va
lu
e
be
twe
e
n '
0'
a
nd '
1'
.
S
ig
m
oi
d f
unc
ti
on:
(
)
=
1
1
+
−
(
2)
H
e
r
e
,
σ
(
x)
is
th
e
pr
oba
bi
li
ty
e
va
lu
a
ti
on
f
unc
ti
on
w
hos
e
va
lu
e
li
e
s
be
twe
e
n
0
a
nd
1,
x
i
s
th
e
in
put
of
th
e
pr
oba
bi
li
ty
f
unc
ti
on,
a
nd
e
is
E
ul
e
r
’
s
nu
m
be
r
a
m
a
th
e
m
a
ti
c
a
l
c
ons
ta
nt
.
T
he
pr
e
s
e
nt
e
d
da
ta
s
e
t
s
how
s
th
a
t
0
im
pl
ie
s
no he
a
r
t
r
is
ks
a
nd 1 im
pl
ie
s
a
hi
gh
r
is
k of
C
H
D
.
2.3.2.
R
an
d
om
f
or
e
s
t
T
he
R
F
i
s
a
c
ol
le
c
ti
on
of
DT
th
a
t
ta
k
e
s
i
nt
o
a
c
c
ount
m
ul
ti
pl
e
DT
be
f
or
e
ge
n
e
r
a
ti
ng a
r
e
s
ul
t
[
19]
. T
hi
s
s
tr
a
te
gy
is
ba
s
e
d
on
th
e
a
s
s
um
pt
io
n
th
a
t
m
or
e
tr
e
e
s
w
il
l
c
o
nve
r
ge
on
th
e
c
or
r
e
c
t
c
on
c
lu
s
io
n.
I
n
bi
na
r
y
c
la
s
s
if
ic
a
ti
on
s
it
ua
ti
ons
,
th
e
R
F
pe
r
f
or
m
s
r
e
m
a
r
ka
bl
y
w
e
ll
.
T
he
c
om
pl
e
te
da
ta
s
e
t
ha
s
be
e
n
s
pl
it
in
to
s
ub
-
da
ta
s
e
ts
,
w
it
h
e
a
c
h
s
ub
-
da
ta
s
e
t
tr
a
in
e
d
in
n
DT
.
E
a
c
h
DT
i
s
tr
a
in
e
d
s
e
pa
r
a
te
ly
a
nd
m
a
k
e
s
pr
e
di
c
ti
ons
f
or
it
s
r
e
s
pe
c
ti
ve
s
ub
-
da
ta
s
e
t.
T
he
f
in
a
l
pr
e
di
c
ti
on
r
e
li
e
s
he
a
vi
ly
on t
he
pr
e
di
c
ti
ve
out
c
om
e
s
of
th
e
s
e
s
ubgr
oups
.
In
(
3)
out
li
ne
s
t
he
pr
oba
bi
li
ty
f
or
e
a
c
h s
ubs
e
t
be
lo
ngi
ng t
o a
pr
e
di
c
ti
ve
c
la
s
s
.
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
.
13
, N
o.
4
,
D
e
c
e
m
be
r
2024
:
4516
-
4525
4520
(
)
=
1
+
2
+
⋯
+
∑
(
(
)
)
=
1
(
3)
w
he
r
e
,
C
is
c
la
s
s
,
F
is
f
e
a
tu
r
e
s
,
n
is
num
be
r
of
s
ub
-
da
ta
s
e
ts
,
a
nd
1
…
is
pr
oba
bi
li
ty
of
e
a
c
h
f
e
a
tu
r
e
a
nd
c
l
a
s
s
.
2.3.3.
A
d
aB
oos
t
A
da
B
oo
s
t
is
a
t
ype
of
e
s
ti
m
a
t
or
th
a
t
pr
odu
c
e
s
a
s
e
r
ie
s
of
w
e
a
k
c
l
a
s
s
if
ie
r
s
,
w
it
h
th
e
a
l
gor
it
hm
de
te
r
m
in
in
g
th
e
be
s
t
c
l
a
s
s
if
ie
r
a
t
e
a
c
h
it
e
r
a
ti
on
ba
s
e
d
o
n
th
e
m
o
s
t
r
e
c
e
nt
s
a
m
pl
e
w
e
i
ght
s
.
I
ni
ti
a
ll
y
,
a
ll
t
he
da
t
a
poi
nt
s
a
r
e
a
s
s
ig
ne
d
e
qu
a
l
w
e
ig
ht
s
.
I
n
th
e
(
k
+
1)
th
it
e
r
a
ti
on,
th
e
s
a
m
pl
e
s
th
a
t
w
e
r
e
pr
e
vi
ou
s
ly
m
i
s
c
l
a
s
s
if
i
e
d
r
e
c
e
iv
e
in
c
r
e
a
s
e
d
im
por
t
a
nc
e
,
w
h
e
r
e
a
s
t
he
s
a
m
pl
e
s
th
a
t
w
e
r
e
c
or
r
e
c
tl
y
c
la
s
s
if
i
e
d
a
r
e
a
s
s
ig
ne
d
r
e
du
c
e
d
i
m
por
ta
nc
e
.
S
ubs
e
que
n
tl
y,
a
ll
da
ta
poi
nt
s
w
it
h
hi
gh
e
r
w
e
ig
ht
s
a
r
e
g
iv
e
n
pr
io
r
it
y.
I
n
e
a
c
h
it
e
r
a
ti
o
n,
a
s
t
a
ge
w
e
ig
ht
is
de
te
r
m
in
e
d
ba
s
e
d
on
th
e
e
r
r
or
r
a
t
e
a
t
th
a
t
p
a
r
ti
c
ul
a
r
it
e
r
a
t
io
n.
T
he
e
n
s
e
m
b
le
f
or
m
e
d
by
th
e
s
e
w
e
ig
ht
e
d
c
la
s
s
if
i
e
r
s
a
s
a
w
h
ol
e
ha
s
a
hi
gh
e
r
pr
ob
a
bi
li
ty
of
b
e
in
g
c
or
r
e
c
tl
y
c
la
s
s
if
i
e
d t
ha
n a
ny of
t
he
i
ndi
vi
d
ua
l
c
la
s
s
if
ie
r
s
[
20]
.
2.4. E
n
s
e
m
b
le
m
e
t
h
od
s
T
he
e
ns
e
m
bl
e
a
ppr
oa
c
h
is
u
s
e
d
to
e
nha
nc
e
c
la
s
s
if
ie
r
a
c
c
ur
a
c
y
by
in
te
gr
a
ti
ng
w
e
a
k
a
nd
s
tr
ong
le
a
r
ne
r
s
.
I
nt
e
gr
a
ti
ng
num
e
r
ous
c
la
s
s
if
ie
r
s
a
im
s
to
im
pr
ove
pe
r
f
or
m
a
nc
e
,
r
e
s
ul
ti
ng
in
be
tt
e
r
out
c
om
e
s
th
a
n
us
in
g
s
in
gl
e
c
la
s
s
if
ie
r
s
.
T
hi
s
is
a
r
e
m
a
r
ka
bl
e
e
ns
e
m
bl
e
le
a
r
ni
ng
te
c
hni
que
f
or
e
nha
nc
in
g
th
e
a
c
c
ur
a
c
y
of
m
ul
ti
pl
e
a
lg
or
it
hm
s
in
pr
e
di
c
ti
ng
C
H
D
.
2.4.1.
B
aggi
n
g
A
m
ode
l
is
tr
a
in
e
d
f
or
each
s
ubs
e
t
of
a
r
e
pl
a
c
e
m
e
nt
tr
a
in
in
g
s
e
t
w
it
h
m
ul
ti
pl
e
s
ubs
e
ts
us
in
g
b
a
ggi
ng
c
la
s
s
if
ie
r
s
[
21]
.
T
he
f
in
a
l
c
la
s
s
if
ic
a
ti
on
out
c
om
e
is
de
t
e
r
m
in
e
d
by
s
e
le
c
ti
ng
th
e
m
a
jo
r
it
y
ba
s
e
d
on
th
e
a
v
e
r
a
ge
of
th
e
pr
e
di
c
te
d
out
c
om
e
s
of
th
e
s
ub
-
m
ode
ls
.
In
th
is
s
tu
dy,
b
a
ggi
ng
e
ns
e
m
bl
e
m
ode
ls
a
r
e
c
on
s
tr
uc
te
d
us
in
g
ba
s
e
c
la
s
s
if
ie
r
s
a
lo
ng
w
it
h
gr
id
s
e
a
r
c
h
a
nd
r
a
ndom
s
e
a
r
c
h
to
de
t
e
r
m
in
e
th
e
opt
im
um
hype
r
pa
r
a
m
e
te
r
,
a
nd
th
e
i
r
pe
r
f
or
m
a
nc
e
is
va
li
da
te
d
u
s
in
g
5
-
f
ol
d
c
r
os
s
-
va
li
da
ti
on.
T
he
s
e
m
ode
ls
a
r
e
m
or
e
a
c
c
ur
a
te
,
but
th
e
ir
CT
is
lo
ng
e
r
.
2.4.2.
V
ot
in
g
T
h
e
m
a
j
or
it
y
v
ot
i
ng
c
l
a
s
s
if
i
e
r
i
s
a
t
yp
e
of
m
e
ta
-
c
l
a
s
s
if
i
e
r
t
ha
t
c
om
bi
ne
s
pr
e
di
c
ti
on
s
f
r
o
m
m
ul
ti
pl
e
in
di
vi
d
u
a
l
c
l
a
s
s
if
i
e
r
s
to
m
a
k
e
a
f
i
na
l
d
e
c
i
s
io
n
b
a
s
e
d
o
n
t
he
m
a
jo
r
it
y
vot
e
of
t
he
m
o
de
ls
[
2
2]
.
I
n
t
hi
s
s
t
udy
,
L
R
,
R
F
,
a
n
d
A
d
a
B
oo
s
t
a
r
e
th
e
c
ho
s
e
n
c
la
s
s
if
ie
r
s
t
o
f
or
m
t
he
e
n
s
e
m
bl
e
m
od
e
l
.
T
he
e
n
s
e
m
b
le
m
od
e
l
e
m
pl
oy
s
‘
h
a
r
d
vot
i
ng’
,
w
h
e
r
e
e
a
c
h
c
l
a
s
s
if
i
e
r
c
o
nt
r
i
but
e
s
a
s
in
gl
e
v
ot
e
to
th
e
f
in
a
l
pr
e
di
c
t
io
n
,
a
n
d
hy
pe
r
pa
r
a
m
e
t
e
r
op
ti
m
i
z
a
ti
on
te
c
hn
iq
u
e
s
a
r
e
a
p
pl
i
e
d
t
o
f
in
e
-
tu
n
e
t
he
p
e
r
f
or
m
a
n
c
e
of
e
a
c
h
c
la
s
s
if
ie
r
,
e
n
s
ur
in
g
t
he
m
os
t
a
c
c
ur
a
te
e
n
s
e
m
bl
e
ou
tp
u
t.
2.4.3.
S
t
ac
k
in
g
S
ta
c
ki
ng
is
a
ty
pe
of
e
ns
e
m
bl
e
le
a
r
ni
ng
in
w
hi
c
h
f
in
a
l
pr
e
di
c
ti
ons
a
r
e
m
a
de
us
in
g
m
ul
ti
pl
e
la
ye
r
s
of
m
ode
ls
[
23]
.
T
he
s
ta
c
ki
ng
c
la
s
s
if
ie
r
w
a
s
bui
lt
in
th
is
w
or
k
us
in
g
RF
a
nd
A
da
B
oos
t
as
th
e
ba
s
is
c
la
s
s
if
ie
r
s
a
nd
LR
as
th
e
m
e
ta
c
la
s
s
if
ie
r
.
T
h
e
m
ode
l
w
a
s
tr
a
in
e
d
u
s
in
g
5
-
f
ol
d
c
r
os
s
-
va
li
da
ti
on
a
lo
ng
w
it
h
gr
id
s
e
a
r
c
h
a
nd
r
a
ndom
s
e
a
r
c
h
to
id
e
nt
if
y
th
e
opt
im
um
hype
r
pa
r
a
m
e
te
r
s
.
T
he
s
ta
c
ki
ng
c
la
s
s
if
ie
r
w
a
s
tr
a
in
e
d
a
ga
in
a
f
te
r
s
e
le
c
ti
ng
th
e
opt
im
um
hype
r
pa
r
a
m
e
te
r
s
,
a
nd
it
s
pe
r
f
or
m
a
nc
e
w
a
s
e
va
lu
a
te
d.
T
he
a
bove
-
m
e
nt
io
ne
d
e
ns
e
m
bl
e
m
ode
ls
a
r
e
im
pl
e
m
e
nt
e
d
f
or
th
e
pr
e
di
c
ti
on
of
C
H
D
.
T
he
e
f
f
e
c
ti
ve
ne
s
s
of
th
e
s
e
e
ns
e
m
bl
e
m
ode
ls
is
e
v
a
lu
a
te
d
w
it
h
r
e
ga
r
d
to
a
c
c
ur
a
c
y,
pr
e
c
is
io
n,
r
e
c
a
ll
,
F1
s
c
or
e
,
a
nd
CT
,
a
nd
th
is
e
va
lu
a
ti
on
is
de
ta
il
e
d
in
s
e
c
ti
on
4.
T
he
f
ol
lo
w
in
g
is
th
e
ps
e
udoc
ode
f
or
th
e
AB
-
S
E
L
m
ode
l.
P
r
oc
e
dur
e
f
or
A
B
-
S
E
L
m
ode
l:
I
nput
:
C
or
ona
r
y he
a
r
t
di
s
e
a
s
e
da
ta
s
e
t
=
{
(
,
)
,
=
1
,
2
,
3
,
…
}
in
C
S
V
f
or
m
a
t
O
ut
put
:
P
e
r
f
or
m
a
nc
e
of
t
he
A
B
-
S
E
L
m
ode
l
S
ta
r
t
S
te
p 1:
I
m
por
t
li
b
r
a
r
ie
s
S
te
p 2:
I
nput
t
he
C
H
D
da
ta
s
e
t
S
te
p 3:
P
r
e
pr
oc
e
s
s
in
g i
s
c
le
a
ni
ng a
nd t
a
ki
ng c
a
r
e
of
m
is
s
in
g va
l
ue
s
S
te
p 4:
F
or
a
ll
f
e
a
tu
r
e
s
1 t
o n
F
it
t
he
f
e
a
tu
r
e
s
us
in
g t
he
L
R
-
R
F
E
a
ppr
oa
c
h
S
te
p 5:
S
pl
it
t
he
da
ta
s
e
t
in
a
n 80:20 r
a
ti
o be
twe
e
n t
he
tr
a
in
in
g a
nd t
e
s
t
s
e
ts
S
te
p 6:
S
c
a
le
t
he
f
e
a
tu
r
e
us
in
g s
ta
nda
r
d
s
c
a
la
r
a
s
(
1
)
S
te
p 7:
T
r
a
in
va
r
io
us
c
la
s
s
if
ic
a
ti
on mode
ls
on t
he
t
r
a
in
in
g s
e
t
a
l
ong with hyper
pa
r
a
m
e
te
r
opt
im
iz
a
ti
on a
nd
s
e
le
c
t
th
e
t
op t
hr
e
e
c
la
s
s
if
ie
r
s
ba
s
e
d on the
a
c
c
ur
a
c
y s
c
or
e
s
S
te
p 8:
A
ppl
y t
he
s
ta
c
ki
ng a
lg
or
it
hm
t
o i
m
pl
e
m
e
nt
A
B
-
S
E
L
m
o
de
l
S
te
p 9:
S
to
p
E
nd
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
c
c
ur
ac
y
bas
e
d
-
s
ta
c
k
e
d
e
ns
e
m
bl
e
l
e
ar
ni
ng m
ode
l
fo
r
t
he
pr
e
di
c
ti
on of c
or
onar
y
…
(
Sant
os
in
i
B
hut
ia
)
4521
3.
R
E
S
U
L
T
S
AND
D
I
S
C
U
S
S
I
O
N
T
hi
s
s
tu
dy
in
ve
s
ti
ga
t
e
d
th
e
e
f
f
e
c
ti
ve
ne
s
s
of
tr
a
di
ti
ona
l
M
L
,
boos
ti
ng,
a
nd
e
ns
e
m
bl
e
m
ode
ls
f
or
pr
e
di
c
ti
ng
C
H
D
us
in
g
th
e
C
le
ve
la
nd
da
ta
s
e
t
publ
ic
ly
a
va
il
a
bl
e
on
th
e
K
a
ggl
e
da
ta
ba
s
e
.
T
he
pe
r
f
or
m
a
nc
e
of
a
ll
th
e
m
ode
ls
is
m
e
a
s
ur
e
d
in
te
r
m
s
of
a
c
c
ur
a
c
y
a
nd
C
T
,
w
he
th
e
r
th
e
r
e
f
e
r
e
nc
e
li
te
r
a
tu
r
e
ha
s
not
a
ddr
e
s
s
e
d
th
e
C
T
.
F
ir
s
tl
y,
th
e
a
c
c
ur
a
c
y
of
th
e
m
ode
ls
is
a
s
s
e
s
s
e
d
by
c
on
s
id
e
r
in
g
a
ll
f
e
a
tu
r
e
s
in
th
e
da
ta
s
e
t.
S
ubs
e
que
nt
ly
,
th
e
LR
-
R
F
E
a
ppr
oa
c
h
is
a
ppl
ie
d
to
e
li
m
in
a
te
th
e
ir
r
e
le
va
nt
f
e
a
tu
r
e
s
.
U
s
in
g
th
e
s
e
f
e
a
tu
r
e
s
ub
s
e
ts
,
a
ll
m
ode
ls
a
r
e
opt
im
iz
e
d
th
r
ough
gr
id
s
e
a
r
c
h
a
nd
r
a
ndom
s
e
a
r
c
h
a
ppr
oa
c
he
s
.
We
f
ound
th
a
t
by
us
in
g
th
e
s
e
r
e
le
va
nt
f
e
a
tu
r
e
s
,
th
e
pe
r
f
or
m
a
nc
e
of
th
e
m
ode
ls
im
pr
ove
d,
as
pr
ovi
de
d
in
T
a
bl
e
s
3
a
nd
4
.
It
is
a
ls
o
ob
s
e
r
ve
d
th
a
t
gr
id
s
e
a
r
c
h
ge
ne
r
a
ll
y
ta
ke
s
lo
nge
r
th
a
n
r
a
ndom
s
e
a
r
c
h
a
c
r
o
s
s
a
ll
a
lg
or
it
hm
s
.
T
h
e
s
ta
nd
a
r
d
s
c
a
le
r
s
c
a
li
ng
a
ppr
oa
c
h
is
im
pl
e
m
e
nt
e
d
to
pr
e
ve
nt
a
lg
or
it
hm
s
f
r
o
m
e
xhi
bi
ti
ng
bi
a
s
to
w
a
r
d
hi
ghe
r
va
lu
e
s
.
T
he
hype
r
pa
r
a
m
e
te
r
s
e
m
pl
oye
d
to
a
c
hi
e
ve
opt
im
a
l
r
e
s
ul
ts
a
r
e
d
e
ta
il
e
d
in
T
a
bl
e
5.
T
he
5
-
f
ol
d
c
r
o
s
s
-
va
li
da
ti
on
te
c
hni
que
is
e
m
pl
oye
d
to
pr
e
ve
nt
pr
obl
e
m
s
r
e
la
te
d
to
ove
r
f
it
ti
ng.
F
ig
ur
e
s
2(
a
)
to
2(
d)
s
how
th
e
a
c
c
ur
a
c
ie
s
a
nd
CT
s
of
a
ll
m
ode
ls
b
a
s
e
d
on
th
e
r
e
le
va
nt
f
e
a
tu
r
e
s
s
e
le
c
te
d
by
th
e
LR
-
R
F
E
f
e
a
tu
r
e
s
e
le
c
ti
on
te
c
hni
que
.
U
pon
obs
e
r
va
ti
on,
we
id
e
nt
if
ie
d
LR
(
a
c
c
ur
a
c
y=
88.52%
,
C
T
=
0.1
s
)
,
RF
(
a
c
c
ur
a
c
y=
88.52%
,
C
T
=
6.4s
)
,
a
nd
A
da
B
oos
t
(
a
c
c
ur
a
c
y=
88.52%
,
C
T
=
2.7
s
)
as
th
e
to
p
th
r
e
e
hi
gh
-
a
c
c
ur
a
c
y
s
tr
ong
le
a
r
ne
r
s
.
T
a
bl
e
3.
E
f
f
e
c
ti
ve
ne
s
s
of
t
r
a
di
ti
ona
l
ML
m
ode
ls
T
r
a
di
t
i
ona
l
ML
a
l
gor
i
t
hm
s
A
c
c
ur
a
c
y w
i
t
h
f
ul
l
f
e
a
t
ur
e
s
e
t
LR
-
R
F
E
w
i
t
h
r
a
ndom
s
e
a
r
c
h
LR
-
R
F
E
w
i
t
h
gr
i
d s
e
a
r
c
h
A
c
c
ur
a
c
y
(%)
A
c
c
ur
a
c
y
(%)
CT
(
s
)
A
c
c
ur
a
c
y
(%)
CT
(
s
)
LR
88.52
88.52
0.1
86.89
0.3
KNN
63.93
83.60
0.1
83.60
0.6
RF
75.41
88.52
6.4
88.52
27.6
DT
77.05
80.32
0.1
85.24
0.7
NB
85.24
86.88
0.1
86.88
2.2
S
V
M
55.73
81.96
0.1
85.25
5.5
T
a
bl
e
4.
E
f
f
e
c
ti
ve
ne
s
s
of
boo
s
ti
ng
m
ode
ls
E
ns
e
m
bl
e
l
e
a
r
ni
ng
a
l
gor
i
t
hm
s
A
c
c
ur
a
c
y w
i
t
h
f
ul
l
f
e
a
t
ur
e
s
e
t
LR
-
R
F
E
w
i
t
h
r
a
ndom
s
e
a
r
c
h
LR
-
R
F
E
w
i
t
h
gr
i
d s
e
a
r
c
h
A
c
c
ur
a
c
y
(%)
A
c
c
ur
a
c
y
(%)
CT
(
s
)
A
c
c
ur
a
c
y
(%)
CT
(
m
)
A
da
B
oos
t
86.88
88.52
2.7
85.24
5.2
G
B
oos
t
i
ng
81.96
83.60
3.1
81.96
5.8
X
G
B
oos
t
83.60
85.24
3.8
85.24
6.3
L
i
ght
G
B
M
85.24
81.97
1.2
83.60
1.7
T
a
bl
e
5.
T
he
hype
r
pa
r
a
m
e
te
r
s
u
s
e
d
to
a
c
hi
e
v
e
opt
im
um
r
e
s
ul
ts
A
l
gor
i
t
hm
s
P
a
r
a
m
e
t
e
r
s
of
r
a
ndom
s
e
a
r
c
h
P
a
r
a
m
e
t
e
r
s
of
gr
i
d s
e
a
r
c
h
LR
C
:
0.23
P
e
na
l
t
y:
l
2
S
ol
ve
r
:
‘
s
a
ga
’
C
:
0.1
P
e
na
l
t
y:
l
2
S
ol
ve
r
:
‘
s
a
ga
’
KNN
N
o of
ne
i
ghbor
:
9
N
o of
ne
i
ghbor
:
9
RF
n_e
s
t
i
m
a
t
or
s
:
300
m
a
x_de
pt
h:
3
c
r
i
t
e
r
i
on:
gi
ni
n_e
s
t
i
m
a
t
or
s
:
300
m
a
x_de
pt
h:
3
c
r
i
t
e
r
i
on:
gi
ni
DT
M
i
n s
a
m
pl
e
s
l
e
a
f
:
7
m
a
x_de
pt
h:
N
one
c
r
i
t
e
r
i
on:
gi
ni
M
i
n
s
a
m
pl
e
s
l
e
a
f
:
7
m
a
x_de
pt
h:
N
one
c
r
i
t
e
r
i
on:
gi
ni
NB
va
r
_s
m
oot
hi
ng:
0.001
va
r
_s
m
oot
hi
ng:
0.001
S
V
M
ke
r
ne
l
:
‘
r
bf
’
ga
m
m
a
:
0.01
C
:
1.0
ke
r
ne
l
:
‘
s
i
gm
oi
d’
ga
m
m
a
:
0.01
C
:
1.0
A
da
B
oos
t
n_e
s
t
i
m
a
t
or
s
:
40
l
e
a
r
ni
ng_r
a
t
e
:
1.0
n_e
s
t
i
m
a
t
or
s
:
20
l
e
a
r
ni
ng_r
a
t
e
:
0.4
G
B
oos
t
i
ng
n_e
s
t
i
m
a
t
or
s
:
80
m
a
x_de
pt
h:
6
l
e
a
r
ni
ng_r
a
t
e
:
0.275
n_e
s
t
i
m
a
t
or
s
:
80
m
a
x_de
pt
h:
3
l
e
a
r
i
ng_r
a
t
e
:
0.1
X
G
B
oos
t
n_e
s
t
i
m
a
t
or
s
:
200
m
a
x_de
pt
h:
5
l
e
a
r
ni
ng_r
a
t
e
:
0.125
n_e
s
t
i
m
a
t
or
s
:
80
m
a
x_de
pt
h:
3
l
e
a
r
ni
ng_r
a
t
e
:
0.1
L
i
ght
G
B
M
n_e
s
t
i
m
a
t
or
s
:
80
m
a
x_de
pt
h:
6
l
e
a
r
ni
ng_r
a
t
e
:
0.25
n_e
s
t
i
m
a
t
or
s
:
80
m
a
x_de
pt
h:
3
l
e
a
r
ni
ng_r
a
t
e
:
0.1
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
.
13
, N
o.
4
,
D
e
c
e
m
be
r
2024
:
4516
-
4525
4522
(
a
)
(
b)
(
c
)
(
d)
F
ig
ur
e
2
.
E
f
f
e
c
t
i
ve
n
e
s
s
of
a
l
l
m
o
d
e
l
s
pe
r
f
or
m
a
n
c
e
of
(
a
)
M
L
a
lg
or
i
t
hm
s
i
n
t
e
r
m
s
of
a
c
c
ur
a
c
y
,
(
b
)
M
L
a
l
g
or
i
th
m
s
i
n
t
e
r
m
s
of
C
T
, (
c
)
bo
o
s
ti
ng
a
l
go
r
i
th
m
s
in
t
e
r
m
s
o
f
a
c
c
u
r
a
c
y,
a
nd
(
d)
b
o
o
s
t
in
g
a
lg
or
it
hm
s
i
n
t
e
r
m
s
of
C
T
T
he
s
tr
ong
le
a
r
ne
r
s
id
e
nt
if
ie
d
-
L
R
,
R
F
,
a
nd
A
da
B
oos
ta
r
e
e
m
pl
oye
d
to
im
pl
e
m
e
nt
th
r
e
e
e
ns
e
m
bl
e
m
ode
ls
,
na
m
e
ly
b
a
ggi
ng,
m
a
jo
r
it
y
vot
in
g,
a
nd
s
ta
c
ki
ng
.
T
h
e
e
f
f
e
c
ti
ve
ne
s
s
of
th
e
s
e
e
n
s
e
m
bl
e
m
ode
ls
is
e
va
lu
a
te
d
in
te
r
m
s
of
a
c
c
ur
a
c
y,
pr
e
c
is
io
n,
r
e
c
a
ll
,
F1
s
c
or
e
,
a
nd
CT
us
in
g
bot
h
r
a
ndom
s
e
a
r
c
h
a
nd
gr
id
s
e
a
r
c
h
a
ppr
oa
c
he
s
,
as
de
pi
c
te
d
in
T
a
bl
e
s
6
a
nd
7,
r
e
s
pe
c
ti
ve
ly
.
F
ig
ur
e
s
3
a
nd
4
pr
e
s
e
nt
th
e
c
om
pa
r
a
ti
ve
a
na
ly
s
i
s
of
th
e
s
e
e
ns
e
m
bl
e
m
ode
ls
,
a
nd
it
is
obs
e
r
ve
d
th
a
t
our
pr
opos
e
d
s
ta
c
ke
d
e
n
s
e
m
bl
e
m
ode
l,
na
m
e
d
th
e
AB
-
S
E
L
m
ode
l
w
it
h
r
a
ndom
s
e
a
r
c
h
hype
r
pa
r
a
m
e
te
r
opt
im
iz
a
ti
on,
pe
r
f
or
m
s
be
tt
e
r
th
a
n
m
a
jo
r
it
y
vot
in
g
a
nd
ba
ggi
ng
e
ns
e
m
bl
e
m
ode
l
s
.
T
hi
s
m
ode
l
a
c
hi
e
ve
s
an
a
c
c
ur
a
c
y
r
a
te
of
90.
16%
,
pr
e
c
is
io
n
of
92%
,
r
e
c
a
ll
of
85.19%
,
a
nd
an
F1
-
s
c
or
e
of
88.46%
.
T
he
CT
t
a
ke
n
by
th
e
s
e
e
n
s
e
m
bl
e
m
od
e
ls
us
in
g
bot
h
r
a
ndom
s
e
a
r
c
h
a
nd
gr
id
s
e
a
r
c
h
hype
r
pa
r
a
m
e
te
r
opt
im
iz
a
ti
on
is
c
om
pa
r
e
d
a
nd
s
how
n
in
F
ig
ur
e
5.
T
he
CT
ta
ke
n
by
th
e
AB
-
S
E
L
m
ode
l
is
0.2
s
e
c
onds
,
w
it
h
a
0.892
A
U
C
.
A
hi
ghe
r
AUC
va
lu
e
in
di
c
a
te
s
be
tt
e
r
pe
r
f
or
m
a
nc
e
[
24
]
.
F
ig
u
r
e
6
de
pi
c
ts
th
e
r
e
c
e
iv
e
r
ope
r
a
ti
ng
c
ha
r
a
c
te
r
is
ti
c
(
R
O
C
)
c
ur
ve
f
or
th
e
AB
-
S
E
L
m
ode
l.
T
a
bl
e
6.
P
e
r
f
or
m
a
nc
e
of
e
ns
e
m
bl
e
m
ode
l
s
us
in
g
r
a
ndom
s
e
a
r
c
h
E
ns
e
m
bl
e
of
R
F
, L
R
, A
da
B
oos
t
A
c
c
ur
a
c
y
(%)
P
r
e
c
i
s
i
on
(%)
R
e
c
a
l
l
(%)
F1
-
s
c
or
e
(%)
CT
S
t
a
c
ki
ng
90.16
92
85.19
88.46
0.2s
M
a
j
or
i
t
y V
ot
i
ng
88.52
88.57
91.18
89.85
1.6s
B
a
ggi
ng
88.52
80
88.89
84.21
4.3s
T
a
bl
e
7.
P
e
r
f
or
m
a
nc
e
of
e
ns
e
m
bl
e
m
ode
l
s
us
in
g
gr
id
s
e
a
r
c
h
E
ns
e
m
bl
e
of
R
F
, L
R
, A
da
B
oos
t
A
c
c
ur
a
c
y
(%)
P
r
e
c
i
s
i
on
(%)
R
e
c
a
l
l
(%)
F1
-
s
c
or
e
(%)
CT
S
t
a
c
ki
ng
90.16
92
85.19
88.46
22.9s
M
a
j
or
i
t
y V
ot
i
ng
88.52
88.57
91.18
89.85
3.5s
B
a
ggi
ng
88.52
80
88.89
84.21
1.8m
75
80
85
90
LR
KNN
RF
DT
NB
S
V
M
A
c
c
ur
a
c
y
i
n %
M
L
A
l
g
or
i
t
hm
s
C
om
par
i
s
on
of
M
L
A
l
gor
i
t
h
m
s
us
i
n
g
h
yper
par
am
et
er
opt
i
m
i
za
t
i
on
A
c
c
ur
a
c
y(
i
n %
)
us
i
ng R
a
ndom
S
e
a
r
c
h
A
c
c
ur
a
c
y(
i
n %
)
us
i
ng G
r
i
d S
e
a
r
c
h
0.1
0.1
6.4
0.1
0.1
0.1
0.3
0.6
27.6
0.7
2.2
5.5
0
5
10
15
20
25
30
LR
KNN
RF
DT
NB
S
V
M
C
T
(
i
n s
e
c
)
M
L
A
l
g
or
i
t
hm
s
C
T
(
i
n
s
ec
)
t
ake
n
b
y
M
L
A
l
gor
i
t
h
m
s
us
i
n
g
h
yper
par
am
et
er
opt
i
m
i
za
t
i
on
R
a
ndom
S
e
a
r
c
h
G
r
i
d S
e
a
r
c
h
2.7
3.1
3.8
1.2
312
348
378
102
0
100
200
300
400
A
da
B
oos
t
G
B
oos
t
i
ng
X
G
B
oos
t
L
i
ght
G
B
M
C
T
(
i
n s
e
c
)
B
oo
s
t
i
ng
A
l
g
or
i
t
hm
s
C
T
(
i
n
s
ec
)
t
ake
n
b
y
B
oos
t
i
n
g
A
l
gor
i
t
h
m
s
us
i
n
g
h
yper
par
am
et
er
opt
i
m
i
za
t
i
on
R
a
nd
om
S
e
a
r
c
h
G
r
i
d S
e
a
r
c
h
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
c
c
ur
ac
y
bas
e
d
-
s
ta
c
k
e
d
e
ns
e
m
bl
e
l
e
ar
ni
ng m
ode
l
fo
r
t
he
pr
e
di
c
ti
on of c
or
onar
y
…
(
Sant
os
in
i
B
hut
ia
)
4523
F
ig
ur
e
3. C
om
pa
r
a
ti
ve
r
e
s
ul
ts
gr
a
ph r
e
pr
e
s
e
nt
a
ti
on f
or
e
ns
e
m
bl
e
m
ode
ls
us
in
g R
a
ndom S
e
a
r
c
h
F
ig
ur
e
4. C
om
pa
r
a
ti
ve
r
e
s
ul
ts
gr
a
ph r
e
pr
e
s
e
nt
a
ti
on f
or
e
ns
e
m
bl
e
m
ode
ls
us
in
g G
r
id
S
e
a
r
c
h
F
ig
ur
e
5. C
om
pa
r
is
on of
C
T
t
a
ke
n by the
e
n
s
e
m
bl
e
m
ode
ls
F
ig
ur
e
6. R
O
C
c
ur
ve
of
AB
-
S
E
L
m
ode
l
O
ur
s
tu
dy
c
onduc
te
d
a
c
om
pr
e
he
n
s
iv
e
a
na
ly
s
is
of
e
xi
s
ti
ng
r
e
s
e
a
r
c
h
in
th
is
dom
a
in
,
a
nd
th
e
f
in
di
ngs
w
e
r
e
s
ys
te
m
a
ti
c
a
ll
y
or
ga
ni
z
e
d
in
T
a
bl
e
8.
I
t
s
how
s
a
n
a
c
c
ur
a
c
y
of
85.48%
by
L
a
th
a
a
nd
J
e
e
va
[
10]
,
88.88%
by
R
a
z
a
[
22]
,
87%
by
K
ha
nna
e
t
al
.
[
25]
,
80.14%
by
M
ia
o
e
t
al
.
[
26]
,
85.86%
by
T
a
m
a
e
t
al
.
[
27]
,
a
nd
87.37
%
by
M
e
ha
novi
ć
e
t
al
.
[
28]
us
in
g
di
f
f
e
r
e
nt
e
ns
e
m
bl
e
le
a
r
ni
ng
te
c
hni
que
s
.
T
he
c
om
pa
r
a
ti
ve
a
na
ly
s
is
r
e
ve
a
le
d
th
a
t
90.16
88.52
88.52
92
88.57
80
85.19
91.18
88.89
88.46
89.85
84.21
0.2
1.6
4.3
S
T
A
C
K
I
N
G
M
A
J
O
R
I
T
Y
V
O
T
I
N
G
B
A
G
G
I
N
G
PE
RF
OR
MA
NC
E
ME
A
SU
RE
S
ENSEMBLE MODELS
P
E
R
F
O
R
M
A
N
C
E
O
F
E
N
S
E
M
B
L
E
M
O
D
E
L
S
U
S
I
N
G
R
A
N
D
O
M
S
E
A
R
C
H
A
c
c
ur
a
c
y
P
r
e
c
i
s
i
on
R
e
c
a
l
l
F
-
s
c
or
e
C
T
(
i
n s
e
c
)
L
i
ne
a
r
(
C
T
(
i
n s
e
c
)
)
90
.16
88.52
88.52
92
88.57
80
85.19
91.18
88.89
88.46
89.85
84.21
22.9
3.5
108
S
T
A
C
K
I
N
G
M
A
J
O
R
I
T
Y
V
O
T
I
N
G
B
A
G
G
I
N
G
P
E
R
F
O
R
M
A
N
C
E
M
E
A
S
U
R
E
S
E
N
S
E
M
B
L
E
M
O
D
E
L
S
P
E
R
F
O
R
M
A
N
C
E
O
F
E
N
S
E
M
B
L
E
M
O
D
E
L
S
U
S
I
N
G
G
R
I
D
S
E
A
R
C
H
A
c
c
ur
a
c
y
P
r
e
c
i
s
i
on
R
e
c
a
l
l
F
-
s
c
or
e
C
T
(
i
n s
e
c
)
L
i
ne
a
r
(
C
T
(
i
n s
e
c
)
)
0.2
1.6
4.3
22.9
3.5
108
0
20
40
60
80
100
120
S
t
a
c
ki
ng
M
a
j
or
i
t
y V
ot
i
ng
B
a
ggi
ng
C
T
(
i
n s
e
c
)
E
ns
e
m
bl
e
M
ode
l
s
C
T
(
i
n
s
ec
)
t
ake
n
b
y
E
n
s
em
b
l
e
M
odel
s
R
a
n
d
o
m
S
e
a
r
c
h
G
r
i
d S
e
a
r
c
h
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
.
13
, N
o.
4
,
D
e
c
e
m
be
r
2024
:
4516
-
4525
4524
our
pr
opos
e
d
A
B
-
S
E
L
m
ode
l
a
c
hi
e
ve
d
a
s
ig
ni
f
ic
a
nt
ly
e
nh
a
nc
e
d
a
c
c
ur
a
c
y
of
90.16%
,
e
xc
e
pt
f
or
th
e
ba
ggi
ng
-
qua
nt
um
s
uppor
t
ve
c
to
r
c
la
s
s
if
ie
r
(
Q
S
V
C
)
m
ode
l
by
A
bdul
s
a
la
m
e
t
al
.
[
29]
.
T
he
B
a
ggi
ng
-
Q
S
V
C
m
ode
l
di
d
not
s
uppl
y
th
e
c
r
uc
ia
l
C
T
f
or
m
ode
l
f
it
ti
ng.
N
ot
a
bl
y,
our
pr
opos
e
d
m
ode
l
de
m
ons
tr
a
te
d
e
f
f
ic
ie
nt
f
it
ti
ng,
c
om
pl
e
ti
ng
th
e
pr
oc
e
s
s
in
a
m
e
r
e
0.2
s.
T
a
bl
e
8.
C
om
pa
r
is
on
of
th
e
p
r
opos
e
d
m
ode
ls
w
it
h
e
xi
s
ti
ng
m
od
e
ls
A
ut
hor
s
M
e
t
hods
F
e
a
t
ur
e
s
e
l
e
c
t
i
on
A
c
c
ur
a
c
y
(%)
P
r
opos
e
d A
B
-
S
E
L
M
ode
l
S
t
a
c
ke
d w
i
t
h R
F
, L
R
, A
d
a
B
oos
t
LR
-
R
F
E
90.16
L
a
t
ha
a
nd
J
e
e
va
[
10]
M
a
j
or
i
t
y vot
e
w
i
t
h N
B
, B
N
, M
P
,
a
nd
R
F
B
r
ut
e
F
or
c
e
85.48
R
a
z
a
[
22]
M
a
j
or
i
t
y
v
ot
i
ng
88.88
K
ha
nna
e
t
al
.
[
25]
S
V
M
87
M
i
a
o
e
t
al
.
[
26]
A
da
pt
i
ve
b
oos
t
i
ng
80.14
A
bdul
s
a
l
a
m
e
t
al
.
[
29]
B
a
ggi
ng
-
Q
S
V
C
R
F
E
90.16
T
a
m
a
e
t
al
.
[
27]
S
t
a
c
ki
ng of
R
F
, G
B
M
,
a
nd X
G
B
oo
s
t
C
F
S
-
PSO
85.86
M
e
ha
novi
ć
e
t
al
.
[
28]
M
a
j
or
i
t
y
v
ot
i
ng
87.37
4.
C
O
N
C
L
U
S
I
O
N
In
t
hi
s
s
tu
dy
,
we
in
te
g
r
a
te
d
L
R
,
R
F
,
a
nd
A
da
B
oos
t
to
a
na
ly
z
e
th
e
pe
r
f
or
m
a
nc
e
of
ba
ggi
ng,
m
a
jo
r
it
y
vot
in
g,
a
nd
s
ta
c
ki
ng
e
n
s
e
m
bl
e
m
ode
l
s
f
or
th
e
C
le
ve
la
nd
da
ta
s
e
t
a
va
il
a
bl
e
on
K
a
ggl
e
.
T
he
s
ta
nda
r
d
s
c
a
l
a
r
m
e
th
od
is
u
s
e
d
f
or
th
e
pr
e
pr
oc
e
s
s
in
g
of
th
e
da
ta
s
e
t.
LR
-
R
F
E
is
us
e
d
to
s
e
le
c
t
a
r
e
le
v
a
nt
f
e
a
tu
r
e
s
ubs
e
t,
w
hi
c
h
im
pr
ove
s
th
e
pr
e
di
c
ti
on
pe
r
f
or
m
a
nc
e
a
nd
r
e
duc
e
s
th
e
ove
r
a
ll
CT
.
G
r
id
s
e
a
r
c
h
a
nd
r
a
ndom
s
e
a
r
c
h
a
r
e
two
s
e
a
r
c
hi
ng
a
ppr
oa
c
he
s
us
e
d
to
s
e
le
c
t
opt
im
a
l
hype
r
pa
r
a
m
e
te
r
s
,
w
it
h
r
a
ndom
s
e
a
r
c
h
ta
ki
ng
le
s
s
CT
th
a
n
gr
id
s
e
a
r
c
h
.
A
c
c
or
di
ng
to
th
e
f
in
di
ngs
of
th
is
r
e
s
e
a
r
c
h,
th
e
AB
-
S
E
L
m
ode
l
s
ugge
s
te
d
in
th
e
s
tu
dy
a
tt
a
in
s
an
im
pr
ove
d
a
c
c
ur
a
c
y
of
90.16%
,
r
e
qui
r
e
s
onl
y
0.2
s
e
c
onds
of
C
T
,
a
nd
a
c
hi
e
ve
s
an
AUC
of
0.892.
T
h
e
s
ugge
s
te
d
in
te
ll
ig
e
nt
m
ode
l
s
how
s
gr
e
a
t
pot
e
nt
ia
l,
but
it
ha
s
c
e
r
ta
in
r
e
s
tr
ic
ti
ons
be
c
a
us
e
of
th
e
s
ta
c
ki
ng
e
ns
e
m
bl
e
m
od
e
l,
w
hi
c
h
a
ll
ow
s
a
ny
m
ode
l
s
to
be
ut
il
iz
e
d
as
ba
s
e
m
ode
ls
a
nd
m
e
ta
-
m
ode
ls
.
F
or
th
is
r
e
a
s
on,
in
th
e
f
ut
ur
e
,
w
it
h
th
e
he
lp
of
th
e
a
ddi
ti
ona
l
d
a
ta
s
e
ts
,
we
w
il
l
be
a
bl
e
to
dr
a
w
m
or
e
tr
us
twor
th
y
c
onc
lu
s
io
ns
.
We
m
a
y
a
ls
o
u
s
e
m
e
ta
he
ur
is
ti
c
m
e
th
ods
a
nd
na
tu
r
e
-
in
s
pi
r
e
d
a
lg
or
it
hm
s
to
opt
im
i
z
e
th
e
c
la
s
s
if
ie
r
pa
r
a
m
e
te
r
s
a
nd
de
t
e
r
m
in
e
th
e
a
c
c
ur
a
c
y
of
th
e
a
lg
or
it
hm
s
.
S
uc
h
a
ty
pe
of
in
te
ll
ig
e
nt
m
ode
l
c
a
n
w
or
k
w
it
h
m
e
di
c
a
l
pr
of
e
s
s
io
na
ls
to
of
f
e
r
a
di
f
f
e
r
e
nt
pe
r
s
pe
c
ti
ve
a
nd
a
c
ti
ve
ly
a
s
s
is
t
pe
opl
e
in
de
te
c
ti
ng
C
H
D
.
R
E
F
E
R
E
N
C
E
S
[
1]
A
. K
a
da
m
, S
. P
a
t
i
l
, P
. P
e
t
hka
r
, R
. S
hi
ka
r
e
, a
nd
S
. S
a
r
na
ya
k,
“
A
c
a
r
di
ova
s
c
ul
a
r
di
s
e
a
s
e
pr
e
di
c
t
i
on s
y
s
t
e
m
us
i
ng m
a
c
hi
n
e
l
e
a
r
ni
ng,”
J
our
nal
of
P
har
m
ac
e
ut
i
c
al
N
e
gat
i
v
e
R
e
s
ul
t
s
, pp. 7216
–
7225, 2023
,
doi
:
10.477
50/
pnr
.2022.13.S
09.849
.
[
2]
R
. A
t
a
t
, L
.
L
i
u, J
. W
u, G
. L
i
, C
. Y
e
, a
nd
Y
. Y
a
ng, “
B
i
g
d
a
t
a
m
e
e
t
c
ybe
r
-
phys
i
c
a
l
s
ys
t
e
m
s
:
a
pa
nor
a
m
i
c
s
ur
ve
y
,”
I
E
E
E
A
c
c
e
s
s
, vol
.
6, pp. 73603
–
73636, 2018, doi
:
10.1109/
A
C
C
E
S
S
.2018.2878681.
[
3]
D
.
G
hos
hR
oy,
P
.
A
.
A
l
vi
,
a
nd
J
.
M
.
R
.
S
.
T
a
va
r
e
s
,
“
D
e
t
e
c
t
i
on
of
c
a
r
di
ova
s
c
ul
a
r
di
s
e
a
s
e
us
i
ng
e
ns
e
m
bl
e
f
e
a
t
ur
e
e
ngi
ne
e
r
i
ng
w
i
t
h
de
c
i
s
i
on t
r
e
e
,”
I
nt
e
r
nat
i
onal
J
our
nal
of
A
m
bi
e
nt
C
om
put
i
ng and I
nt
e
l
l
i
ge
nc
e
, v
ol
. 13, no. 1, 2022, doi
:
10.4018/
I
J
A
C
I
.300795.
[
4]
A
.
S
.
K
um
a
r
a
nd
N
.
S
i
nha
,
“
C
a
r
di
ova
s
c
ul
a
r
di
s
e
a
s
e
i
n
I
ndi
a
:
A
360
de
gr
e
e
ove
r
vi
e
w
,”
M
e
di
c
al
J
our
nal
A
r
m
e
d
F
or
c
e
s
I
ndi
a
,
vol
.
76, no. 1, pp. 1
–
3, 2020, doi
:
10.1016/
j
.m
j
a
f
i
.2019.12.005.
[
5]
V
.
S
hor
e
w
a
l
a
,
“
E
a
r
l
y
de
t
e
c
t
i
on
of
c
o
r
ona
r
y
he
a
r
t
di
s
e
a
s
e
us
i
ng
e
ns
e
m
bl
e
t
e
c
hn
i
que
s
,”
I
nf
or
m
at
i
c
s
i
n
M
e
di
c
i
ne
U
nl
oc
k
e
d
,
vol
.
26,
2021, doi
:
10.1016/
j
.i
m
u.2021.100655.
[
6]
P
.
G
hos
h
e
t
al
.
,
“
E
f
f
i
c
i
e
nt
pr
e
di
c
t
i
on
of
c
a
r
di
ova
s
c
ul
a
r
di
s
e
a
s
e
us
i
ng
m
a
c
hi
n
e
l
e
a
r
ni
ng
a
l
gor
i
t
hm
s
w
i
t
h
r
e
l
i
e
f
a
nd
l
a
s
s
o
f
e
a
t
ur
e
s
e
l
e
c
t
i
on t
e
c
hni
que
s
,”
I
E
E
E
A
c
c
e
s
s
, vol
. 9, pp. 19304
–
19326, 2021, doi
:
10.1109/
A
C
C
E
S
S
.2021.3053759.
[
7]
S
.
B
hut
i
a
,
B
.
P
a
t
r
a
,
a
nd
M
.
R
a
y,
“
C
O
V
I
D
-
19
e
pi
de
m
i
c
:
a
n
a
l
ys
i
s
a
nd
pr
e
di
c
t
i
on,”
I
A
E
S
I
nt
e
r
nat
i
onal
J
our
nal
of
A
r
t
i
f
i
c
i
al
I
nt
e
l
l
i
ge
nc
e
, vol
. 11, no. 2, pp. 736
–
745, 2022, doi
:
10.11591/
i
j
a
i
.v11.i
2.pp736
-
745.
[
8]
B
.
P
a
t
r
a
,
S
.
B
hut
i
a
,
a
nd
N
.
P
a
nda
,
“
M
a
c
hi
ne
l
e
a
r
ni
ng
t
e
c
hni
que
s
f
or
c
a
nc
e
r
r
i
s
k
pr
e
di
c
t
i
on
,”
T
E
ST
E
ngi
ne
e
r
i
ng
&
M
anage
m
e
nt
,
vol
. 83, 2020, pp. 7414
–
7420, 2020.
[
9]
S
.
B
hut
i
a
,
B
.
P
a
t
r
a
,
a
nd
M
. R
a
y, “
A
hybr
i
d
a
ppr
oa
c
h
f
or
c
a
nc
e
r
c
l
a
s
s
i
f
i
c
a
t
i
on
ba
s
e
d
on s
qui
r
r
e
l
s
e
a
r
c
h,”
J
ou
r
nal
of
I
nf
or
m
at
i
on
and
O
pt
i
m
i
z
at
i
on Sc
i
e
nc
e
s
, vol
. 43, no. 5, pp. 905
–
914, 2022, doi
:
10.1080/
025226
67.2022.2091095.
[
10]
C
.
B
.
C
.
L
a
t
ha
a
nd
S
.
C
.
J
e
e
va
,
“
I
m
pr
ovi
ng
t
he
a
c
c
ur
a
c
y
of
pr
e
di
c
t
i
on
of
h
e
a
r
t
di
s
e
a
s
e
r
i
s
k
b
a
s
e
d
on
e
n
s
e
m
bl
e
c
l
a
s
s
i
f
i
c
a
t
i
o
n
t
e
c
hni
que
s
,”
I
nf
or
m
at
i
c
s
i
n M
e
di
c
i
ne
U
nl
oc
k
e
d
, vol
. 16, 2019, doi
:
10.1016/
j
.i
m
u.2019.100203.
[
11]
M
.
N
.
U
ddi
n
a
nd
R
.
K
.
H
a
l
de
r
,
“
A
n
e
n
s
e
m
bl
e
m
e
t
hod
ba
s
e
d
m
ul
t
i
l
a
ye
r
dyna
m
i
c
s
ys
t
e
m
t
o
pr
e
di
c
t
c
a
r
di
ova
s
c
ul
a
r
di
s
e
a
s
e
u
s
i
ng
m
a
c
hi
ne
l
e
a
r
ni
ng a
ppr
oa
c
h,”
I
nf
or
m
at
i
c
s
i
n M
e
di
c
i
ne
U
nl
oc
k
e
d
, vol
. 24, 2021,
doi
:
10.1016/
j
.i
m
u.2021.100584.
[
12]
X
. Y
. G
a
o, A
. A
m
i
n
A
l
i
, H
.
S
.
H
a
s
s
a
n, a
nd E
. M
. A
nw
a
r
,
“
I
m
pr
ovi
ng
t
he
a
c
c
ur
a
c
y f
or
a
na
l
yz
i
ng he
a
r
t
di
s
e
a
s
e
s
pr
e
di
c
t
i
on ba
s
e
d on
t
he
e
ns
e
m
bl
e
m
e
t
hod,”
C
o
m
pl
e
x
i
t
y
, vol
. 2021, 2021, doi
:
10.1155/
2021/
6663455.
[
1
3]
A
.
R
a
hi
m
,
Y
.
R
a
s
h
e
e
d,
F
.
A
z
a
m
,
M
.
W
.
A
n
w
a
r
,
M
.
A
.
R
a
hi
m
,
a
nd
A
.
W
.
M
uz
a
f
f
a
r
,
“
A
n
i
n
t
e
g
r
a
t
e
d
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
f
r
a
m
e
w
o
r
k
f
or
e
f
f
e
c
t
i
ve
p
r
e
di
c
t
i
on
of
c
a
r
d
i
ov
a
s
c
u
l
a
r
d
i
s
e
a
s
e
s
,”
I
E
E
E
A
c
c
e
s
s
,
v
ol
.
9
,
p
p.
1
0
65
75
–
10
6
58
8
,
20
2
1,
d
o
i
:
10
.1
1
09
/
A
C
C
E
S
S
.
20
2
1.
30
9
86
88
.
[
14]
I
.
D
.
M
i
e
nye
,
Y
.
S
un, a
nd
Z
.
W
a
ng,
“
A
n
i
m
pr
ove
d
e
ns
e
m
bl
e
l
e
a
r
ni
ng
a
ppr
oa
c
h f
or
t
he
pr
e
di
c
t
i
on
of
he
a
r
t
di
s
e
a
s
e
r
i
s
k,
”
I
nf
or
m
at
i
c
s
i
n M
e
di
c
i
ne
U
nl
oc
k
e
d
, vol
. 20, 2020, doi
:
10.1016/
j
.i
m
u.2020.100402.
[
15]
B
. P
a
t
r
a
, L
. J
e
na
, S
. B
hut
i
a
, a
nd S
.
N
a
ya
k, “
E
vol
ut
i
ona
r
y
hybr
i
d
f
e
a
t
ur
e
s
e
l
e
c
t
i
on f
or
c
a
nc
e
r
di
a
gnos
i
s
,”
Sm
ar
t
I
nnov
at
i
on, Sy
s
t
e
m
s
and T
e
c
hnol
ogi
e
s
, vol
. 153, pp. 279
–
287, 2021, doi
:
10.1007/
978
-
981
-
15
-
6202
-
0_28.
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
c
c
ur
ac
y
bas
e
d
-
s
ta
c
k
e
d
e
ns
e
m
bl
e
l
e
ar
ni
ng m
ode
l
fo
r
t
he
pr
e
di
c
ti
on of c
or
onar
y
…
(
Sant
os
in
i
B
hut
ia
)
4525
[
16]
B
.
P
a
t
r
a
,
S
.
B
hut
i
a
,
T
.
P
a
nde
y,
a
nd
L
.
J
e
na
,
“
A
n
i
nnova
t
i
ve
I
oT
-
ba
s
e
d
br
e
a
s
t
c
a
nc
e
r
m
oni
t
or
i
ng
s
ys
t
e
m
w
i
t
h
t
he
a
i
d
of
m
a
c
hi
ne
l
e
a
r
ni
ng
a
ppr
oa
c
h,”
T
he
I
nt
e
r
ne
t
of
M
e
di
c
al
T
hi
ng
s
:
E
nabl
i
ng
T
e
c
hnol
ogi
e
s
a
nd
E
m
e
r
gi
ng
A
ppl
i
c
at
i
ons
,
pp.
155
–
180,
2022,
doi
:
10.1049/
pbhe
034e
_c
h9.
[
17]
H
.
W
u,
“
A
de
e
p
l
e
a
r
ni
ng
-
ba
s
e
d
hybr
i
d
f
e
a
t
ur
e
s
e
l
e
c
t
i
on
a
ppr
oa
c
h
f
or
c
a
nc
e
r
di
a
gnos
i
s
,”
J
ou
r
nal
of
P
hy
s
i
c
s
:
C
onf
e
r
e
nc
e
Se
r
i
e
s
,
vol
.
1848, no. 1, 2021, doi
:
10.1088/
1742
-
6596/
1848/
1/
012019.
[
18]
I
. K
a
m
ka
r
, S
. K
.
G
upt
a
, D
. P
hung, a
nd
S
. V
e
nka
t
e
s
h, “
S
t
a
bl
e
f
e
a
t
ur
e
s
e
l
e
c
t
i
on f
or
c
l
i
ni
c
a
l
pr
e
di
c
t
i
on:
E
xpl
oi
t
i
ng
I
C
D
t
r
e
e
s
t
r
uc
t
ur
e
us
i
ng T
r
e
e
-
L
a
s
s
o,”
J
our
nal
of
B
i
om
e
di
c
al
I
nf
or
m
at
i
c
s
, vol
. 53, pp. 277
–
290, 20
15, doi
:
10.1016/
j
.j
bi
.2014.11.013.
[
19]
Y
.
L
uo
e
t
al
.
,
“
P
r
e
di
c
t
i
ng
c
onge
ni
t
a
l
he
a
r
t
de
f
e
c
t
s
:
A
c
om
pa
r
i
s
on
of
t
hr
e
e
da
t
a
m
i
ni
ng
m
e
t
hods
,”
P
L
oS
O
N
E
,
vol
.
12,
no.
5,
2017,
doi
:
10.1371/
j
our
na
l
.pone
.0177811.
[
20]
T
.
R
.
M
a
he
s
h
e
t
al
.
,
“
A
d
a
B
oos
t
e
n
s
e
m
bl
e
m
e
t
hod
s
us
i
ng
k
-
f
ol
d
c
r
os
s
va
l
i
da
t
i
o
n
f
or
s
ur
vi
va
bi
l
i
t
y
w
i
t
h
t
he
e
a
r
l
y
de
t
e
c
t
i
on
of
he
a
r
t
di
s
e
a
s
e
,
”
C
om
put
at
i
onal
I
nt
e
l
l
i
ge
nc
e
and N
e
u
r
os
c
i
e
n
c
e
, vol
. 2022, 2022, doi
:
10.1155/
2022/
9005278.
[
21]
L
. B
r
e
i
m
a
n, “
B
a
ggi
ng pr
e
di
c
t
or
s
,
”
M
ac
hi
ne
L
e
ar
ni
ng
, vol
. 24, no. 2, pp. 123
–
1
40, 1996, doi
:
10.1023/
A
:
1018054314350.
[
22]
K
.
R
a
z
a
,
“
I
m
pr
ovi
ng
t
he
pr
e
di
c
t
i
on
a
c
c
ur
a
c
y
of
he
a
r
t
di
s
e
a
s
e
w
i
t
h
e
ns
e
m
bl
e
l
e
a
r
ni
ng
a
nd
m
a
j
or
i
t
y
vot
i
ng
r
ul
e
,”
U
-
H
e
al
t
hc
ar
e
M
oni
t
or
i
ng Sy
s
t
e
m
s
:
V
ol
um
e
1:
D
e
s
i
gn and A
ppl
i
c
at
i
ons
, pp. 179
–
196, 2018, d
oi
:
10.1016/
B
978
-
0
-
12
-
815370
-
3.00008
-
6.
[
23]
J
.
W
a
ng
e
t
al
.
,
“
A
s
t
a
c
ki
ng
-
ba
s
e
d
m
ode
l
f
or
non
-
i
nva
s
i
ve
de
t
e
c
t
i
on
of
c
or
ona
r
y
he
a
r
t
di
s
e
a
s
e
,”
I
E
E
E
A
c
c
e
s
s
,
vol
.
8,
pp.
37124
–
37133, 2020, doi
:
10.1109/
A
C
C
E
S
S
.2020.2975377.
[
2
4]
F
.
T
e
n
g,
Z
.
M
a
,
J
.
C
he
n
,
M
.
X
i
a
o
,
a
n
d
L
.
H
u
a
n
g
,
“
A
u
t
o
m
a
t
i
c
m
e
di
c
a
l
c
od
e
a
s
s
i
g
n
m
e
nt
v
i
a
d
e
e
p
l
e
a
r
ni
ng
a
p
p
r
o
a
c
h
f
o
r
i
n
t
e
l
l
i
ge
n
t
h
e
a
l
t
h
c
a
r
e
,
”
I
E
E
E
J
our
n
al
o
f
B
i
o
m
e
di
c
a
l
an
d
H
e
a
l
t
h
I
n
f
o
r
m
at
i
c
s
,
vo
l
.
2
4
,
n
o
.
9
,
p
p
.
2
5
06
–
25
15
,
20
2
0,
do
i
:
1
0
.1
10
9
/
J
B
H
I
.
20
2
0.
29
9
69
37
.
[
25]
D
.
K
ha
nna
,
R
.
S
a
hu,
V
.
B
a
t
hs
,
a
nd
B
.
D
e
s
hpa
nde
,
“
C
om
pa
r
a
t
i
ve
s
t
udy
of
c
l
a
s
s
i
f
i
c
a
t
i
on
t
e
c
hni
que
s
(
S
V
M
,
l
ogi
s
t
i
c
r
e
gr
e
s
s
i
on
a
n
d
ne
ur
a
l
ne
t
w
or
ks
)
t
o
pr
e
di
c
t
t
h
e
pr
e
va
l
e
nc
e
of
he
a
r
t
di
s
e
a
s
e
,”
I
nt
e
r
nat
i
onal
J
ou
r
n
al
of
M
ac
hi
ne
L
e
a
r
ni
ng
and C
om
put
i
ng
,
vol
.
5,
no.
5, pp. 414
–
419, 2015, doi
:
10.7763/
i
j
m
l
c
.2015.v5.544.
[
26]
K
.
H
.
M
i
a
o
,
J
.
H
.
M
i
a
o
,
a
nd
G
.
J
.
M
i
a
o
,
“
D
i
a
gnos
i
ng
c
or
ona
r
y
he
a
r
t
di
s
e
a
s
e
us
i
ng
e
ns
e
m
bl
e
m
a
c
hi
ne
l
e
a
r
ni
ng
,”
I
nt
e
r
nat
i
onal
J
our
nal
of
A
dv
anc
e
d C
om
put
e
r
Sc
i
e
nc
e
and A
ppl
i
c
at
i
ons
, vol
. 7, no. 10, 2016,
doi
:
10.14569/
i
j
a
c
s
a
.2016.071004.
[
27]
B
.
A
.
T
a
m
a
,
S
.
I
m
,
a
nd
S
.
L
e
e
,
“
I
m
pr
ovi
ng
a
n
i
nt
e
l
l
i
ge
nt
de
t
e
c
t
i
on
s
ys
t
e
m
f
o
r
c
or
ona
r
y
he
a
r
t
di
s
e
a
s
e
us
i
ng
a
t
w
o
-
t
i
e
r
c
l
a
s
s
i
f
i
e
r
e
ns
e
m
bl
e
,”
B
i
oM
e
d R
e
s
e
ar
c
h I
nt
e
r
nat
i
onal
, vol
. 2020, 2020, doi
:
10.1155/
2020/
9816142.
[
28]
D
.
M
e
ha
novi
ć
,
Z
.
M
a
š
e
t
i
ć
,
a
nd
D
.
K
e
č
o,
“
P
r
e
di
c
t
i
on
of
he
a
r
t
di
s
e
a
s
e
s
u
s
i
ng
m
a
j
or
i
t
y
vot
i
ng
e
n
s
e
m
bl
e
m
e
t
hod,
”
I
F
M
B
E
P
r
oc
e
e
di
ngs
, vol
. 73, pp. 491
–
498, 2020, doi
:
10.1007/
978
-
3
-
030
-
17971
-
7_73.
[
29]
G
.
A
bdul
s
a
l
a
m
,
S
.
M
e
s
houl
,
a
nd
H
.
S
ha
i
ba
,
“
E
xpl
a
i
na
bl
e
he
a
r
t
di
s
e
a
s
e
pr
e
d
i
c
t
i
on
us
i
ng
e
ns
e
m
bl
e
-
qua
nt
um
m
a
c
hi
ne
l
e
a
r
ni
ng
a
ppr
o
a
c
h,”
I
nt
e
l
l
i
ge
nt
A
ut
om
at
i
on and Sof
t
C
om
put
i
ng
, vol
. 36, no. 1, pp. 761
–
779, 2023, doi
:
10.32604/
i
a
s
c
.2023.032262.
B
I
O
G
R
A
P
H
I
E
S
O
F
A
U
T
H
O
R
S
Santosini
Bhutia
is
currently
working
as
a
Research
Scholar
at
Siksha
‘O’
Anusandhan
(Deemed
to
be
University),
Bhubaneswar,
Odisha,
I
ndia.
She
received
her
Bachelo
r
of
Technol
ogy
(B.Tech
.)
in
Informa
tion
Technol
ogy,
Master
of
Technol
ogy
(M.Tech.) in C
omputer S
cience and Engi
neering, and p
ursuing h
er Ph.D. i
n Comput
er Science
and Engi
neering. S
he can be co
ntacted at
email:
santosini.bhutia@gmail.com
.
Bichitrananda
Patra
is
currently
working
as
a
Professor
in
the
Department
of
Computer
Applicat
ion
at
the
Institute
of
Technic
al
Educatio
n
an
d
Resear
ch,
Siksha
'
O'
Anusandhan
(Deemed
to
be
University).
Bhubaneswar,
Odisha,
India. He
received
his
M.Tech.
in
Computer
Science
from
Utkal
University,
Bhub
aneswar,
and
Ph.D.
from
Berhampur
University,
Orissa,
India.
He
has
published
more
research
papers
in
in
ternational
and
national
journals,
confere
nces,
and
book
chapters
in
differe
nt
books
and
also
has
membership
in
differe
nt
professional
bodies
like
ISTE
and
CSI.
H
e
can
be
contacted
at
email
:
bichitranandapatra@soa.ac.in
.
Mitrabin
da
Ray
is
currently
working
as
an
Associate
Prof.
in
the
Department
of
Computer
Science
and
Enginee
ring
at
Siksha
‘O’
Anusandh
an
(De
emed
to
be
Universi
ty),
Bhubanes
war,
Odisha,
India.
She
holds
her
Master
of
Technol
ogy
(
M.
Tech.)
and
Ph.D.
in
Computer
Science
and
Engineering
from
the
National
Institute
of
Technology,
Rourkela,
Odisha,
India.
She
has
more
than
15
years
of
teaching
and
resear
ch
experience.
She
has
published
more
than
25
international
journals
.
Her
research
areas
of
interest
include
software
testing
and
software
reliability
analysis.
Sh
e
c
an
be
contacted
at
email:
mitrabindar
ay@
soa.ac.in
.
Evaluation Warning : The document was created with Spire.PDF for Python.