I
n
t
e
r
n
at
io
n
al
Jou
r
n
al
of
A
d
van
c
e
s
i
n
A
p
p
li
e
d
S
c
ie
n
c
e
s
(
I
JA
A
S
)
V
ol
.
14
, N
o.
4
,
D
e
c
e
m
be
r
20
25
, pp.
1028
~
1039
I
S
S
N
:
2252
-
8814
,
D
O
I
:
10.11591/
ij
a
a
s
.
v14.
i
4
.
pp1028
-
1039
1028
Jou
r
n
al
h
om
e
page
:
ht
tp
:
//
ij
aas
.i
ae
s
c
or
e
.c
om
Im
p
l
e
m
e
n
t
at
i
on
of
X
G
B
oost
f
or
d
i
ab
e
t
e
s m
e
l
l
i
t
u
s r
i
sk
p
r
e
d
i
c
t
i
on
b
ase
d
on
h
e
al
t
h
h
i
st
or
y
A
n
d
i
R
ia
n
s
yah
1
,
G
h
u
f
r
on
1
, L
ai
la
t
u
l
F
it
r
iy
ah
1
, S
u
yan
t
o
2
1
D
e
pa
r
t
m
e
nt
o
f
I
nf
or
m
a
t
i
c
s
E
ngi
ne
e
r
i
ng
,
F
a
c
ul
t
y
o
f
I
ndus
t
r
i
a
l
T
e
c
hnol
ogy, U
ni
ve
r
s
i
t
a
s
I
s
l
a
m
S
ul
t
a
n A
gung, S
e
m
a
r
a
ng, I
ndone
s
i
a
2
F
a
c
ul
t
y of
N
ur
s
i
ng,
U
ni
ve
r
s
i
t
a
s
I
s
l
a
m
S
ul
t
a
n A
gung, S
e
m
a
r
a
ng
,
I
ndone
s
i
a
A
r
t
ic
le
I
n
f
o
A
B
S
T
R
A
C
T
A
r
ti
c
le
h
is
to
r
y
:
R
e
c
e
iv
e
d
S
e
p 12, 2024
R
e
vi
s
e
d
O
c
t
30, 2025
A
c
c
e
pt
e
d
O
c
t
20, 2025
Diabetes
mellitu
s
(DM)
is
a
chronic
disease
with
a
growing
global
burden
and
specific
challenges
for
early
management,
particularly
in
region
s
with
limited
access
to
h
ealthcar
e.
T
his
study
develops
a
web
-
based
syst
em
to
classify
diabetes
risk
from
medical
history
using
extreme
gradient
b
oostin
g
(
XGBoost
)
,
an
ensemble
model
of
decision
trees.
The
dataset
compris
ed
520
respondents
(320
DM,
200
non
-
DM)
and
underwent
la
beling,
standardization,
and
an
80:20
train
–
test
split,
followed
by
hyperparameter
selection
via
grid
search
and
5
-
fold
cross
-
validation
(CV).
On
the
t
est
set,
the
model
achieved
an
accuracy
of
0.9888,
precision
of
1.0000,
re
call
of
0.9718,
and
an
F1
-
score
of
0.9857;
discriminative
pe
rformance
wa
s
also
strong
with
an
area
under
the
receiver
operating
characteristic
curve
(AUC
-
ROC)
of
0.
839.
These
findings
confirm
that
XGBoost
effectively
h
andles
complex
or
imbalanced
medical
data
while
providi
ng
probabil
istic
output
s
that
are
clinically
meaningful.
Deployed
as
a
web
application,
the
system
can
support
early
screening,
triage,
and
clinical
decision
-
making,
t
hereby
expediti
ng
referrals
and
personali
zing
intervent
ions
in
primary
care
and
hospital settings, especially in resource
-
constrain
ed enviro
nments.
Thi
s work
lays
the
groundwork
for
further
development,
including
the
integrat
ion
of
expl
ainable
artificial
intell
igence
(XAI)
techniqu
es
to
enhance
c
linical
transpar
ency.
K
e
y
w
o
r
d
s
:
D
e
c
is
io
n t
r
e
e
D
ia
be
te
s
m
e
ll
it
us
M
a
c
hi
ne
l
e
a
r
ni
ng
P
r
e
di
c
ti
on
X
G
B
oos
t
This is an
open
acce
ss artic
le unde
r the
CC BY
-
SA
license.
C
or
r
e
s
pon
di
n
g A
u
th
or
:
A
ndi
R
ia
ns
ya
h
D
e
pa
r
tm
e
nt
of
I
nf
or
m
a
ti
c
s
E
ngi
ne
e
r
in
g, F
a
c
ul
ty
of
I
ndus
tr
ia
l
T
e
c
hnol
ogy
,
U
ni
ve
r
s
it
a
s
I
s
la
m
S
ul
ta
n A
gung
K
a
li
ga
w
e
S
tr
e
e
t
K
m
. 4, S
e
m
a
r
a
ng 50112, I
ndone
s
ia
E
m
a
il
:
a
ndi
@
uni
s
s
ul
a
.a
c
.i
d.
1.
I
N
T
R
O
D
U
C
T
I
O
N
I
n ge
ne
r
a
l,
di
a
be
te
s
r
e
f
e
r
s
t
o a
gr
oup of
c
ondi
ti
ons
t
ha
t
a
f
f
e
c
t
pr
oc
e
s
s
e
s
i
n t
he
body, c
ha
r
a
c
t
e
r
iz
e
d by
a
s
us
ta
in
e
d
in
c
r
e
a
s
e
in
bl
ood
s
uga
r
le
ve
l
s
due
to
da
m
a
ge
to
th
e
f
unc
ti
on
of
or
ga
ns
s
uc
h
a
s
ne
r
ve
s
,
ki
dne
y
s
,
he
a
r
t,
a
nd
bl
ood
ve
s
s
e
ls
.
T
hi
s
di
s
e
a
s
e
c
a
n
le
a
d
to
va
r
io
us
s
e
r
i
ous
c
om
pl
ic
a
ti
ons
th
a
t
im
pa
c
t
m
a
ny
or
ga
n
s
in
th
e
body,
w
it
h
a
r
a
nge
of
s
ym
pt
om
s
[
1]
.
I
n
I
ndone
s
ia
,
th
e
pr
e
va
le
nc
e
of
di
a
gnos
e
d
di
a
be
te
s
c
ont
in
ue
s
to
r
is
e
,
a
s
r
e
f
le
c
te
d
in
th
e
R
is
ke
s
da
s
da
ta
s
in
c
e
2018,
w
hi
c
h
pr
ovi
de
s
a
de
ta
il
e
d
pi
c
tu
r
e
of
th
e
di
s
tr
ib
ut
io
n
of
di
a
be
te
s
a
c
r
os
s
th
e
a
r
c
hi
pe
la
go.
T
hi
s
c
ondi
ti
on
r
e
qui
r
e
s
a
d
e
e
p
unde
r
s
t
a
ndi
ng
to
pl
a
n
e
f
f
e
c
ti
ve
in
te
r
ve
nt
io
ns
f
or
th
e
pr
e
ve
nt
io
n
a
nd
m
a
na
ge
m
e
nt
of
di
a
be
te
s
in
th
e
c
om
m
uni
ty
[
2]
,
[
3]
.
D
ia
be
te
s
ha
s
e
vol
ve
d
in
to
a
gl
oba
l
he
a
lt
h
is
s
ue
w
it
h
in
c
r
e
a
s
in
g
num
be
r
s
e
v
e
r
y
ye
a
r
.
T
hi
s
in
di
c
a
te
s
th
a
t
di
a
be
te
s
is
not
ju
s
t
a
lo
c
a
l
he
a
lt
h
pr
obl
e
m
but
ha
s
be
c
om
e
a
gl
oba
l
c
h
a
ll
e
nge
th
a
t
r
e
qui
r
e
s
a
tt
e
nt
io
n
a
nd
c
o
ll
e
c
ti
ve
a
c
ti
on
f
r
om
va
r
io
us
pa
r
ti
e
s
w
ho
a
r
e
a
w
a
r
e
[
4]
,
[
5]
.
I
n
r
e
c
e
nt
ye
a
r
s
,
th
e
us
e
of
m
a
c
hi
ne
le
a
r
ni
ng
te
c
hni
que
s
ha
s
be
c
om
e
ve
r
y
popula
r
f
or
di
s
e
a
s
e
pr
e
di
c
ti
on.
H
ow
e
ve
r
,
th
e
a
c
c
ur
a
c
y
of
di
s
e
a
s
e
pr
e
di
c
ti
ons
r
e
m
a
i
ns
a
s
ig
ni
f
ic
a
nt
c
ha
ll
e
nge
in
th
e
m
e
di
c
a
l
f
ie
ld
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
dv A
ppl
S
c
i
I
S
S
N
:
2252
-
8814
I
m
pl
e
m
e
nt
at
io
n of
X
G
B
oos
t
fo
r
di
abe
te
s
m
e
ll
it
us
r
i
s
k
pr
e
di
c
ti
on bas
e
d on he
al
th
hi
s
to
r
y
(
A
ndi
R
ia
ns
y
ah)
1029
W
it
h t
he
pr
e
s
e
nc
e
of
a
r
ti
f
ic
ia
l
in
te
ll
ig
e
nc
e
, i
t
is
hope
d t
ha
t
it
c
a
n pr
e
di
c
t
di
s
e
a
s
e
s
, e
s
pe
c
ia
ll
y di
a
be
te
s
m
e
ll
it
us
(
D
M
)
, s
o t
ha
t
pa
ti
e
nt
s
c
a
n r
e
c
e
iv
e
pr
om
pt
a
nd a
c
c
ur
a
te
c
a
r
e
[
6]
,
[
7]
.
T
he
e
xt
r
e
m
e
gr
a
di
e
nt
boos
ti
ng
(
X
G
B
oos
t
)
pr
oc
e
s
s
c
a
n
e
nha
n
c
e
pr
e
di
c
ti
on
a
c
c
ur
a
c
y
th
r
ough
m
ul
ti
-
th
r
e
a
di
ng,
w
hi
c
h
e
f
f
ic
ie
nt
ly
ut
il
iz
e
s
C
P
U
c
or
e
s
to
s
ig
ni
f
ic
a
nt
ly
a
c
c
e
le
r
a
te
th
e
a
lg
or
it
hm
'
s
pe
r
f
or
m
a
nc
e
.
T
hi
s
a
lg
or
it
hm
a
ppl
ie
s
e
ns
e
m
bl
e
t
e
c
hni
qu
e
s
a
nd f
oc
u
s
e
s
on t
he
gr
a
di
e
nt
of
t
he
opt
im
iz
e
d l
os
s
f
unc
ti
on, r
e
s
ul
ti
ng i
n
a
m
ode
l
th
a
t
is
ve
r
y
r
obus
t
a
nd
a
c
c
ur
a
te
[
8]
–
[
10
]
.
W
it
h
X
G
B
o
os
t,
th
is
r
e
s
e
a
r
c
h
a
im
s
to
pr
oduc
e
a
m
ode
l
th
a
t
c
a
n
pr
e
di
c
t
th
e
r
is
k
of
D
M
ba
s
e
d
on
a
n
in
di
vi
dua
l'
s
he
a
lt
h
hi
s
t
or
y
to
e
nha
nc
e
e
a
r
ly
di
a
gnos
i
s
[
11]
–
[
13]
.
T
he
r
e
s
e
a
r
c
he
r
s
hop
e
to
c
ont
r
ib
ut
e
to
th
e
d
e
ve
lo
pm
e
nt
of
m
or
e
a
c
c
ur
a
te
pr
e
di
c
ti
ve
to
ol
s
th
a
t
c
a
n
a
s
s
i
s
t
in
e
a
r
ly
pr
e
ve
nt
io
n
a
nd
m
a
na
ge
m
e
nt
f
or
th
os
e
a
t
r
is
k,
a
s
w
e
ll
a
s
pr
ovi
de
a
ddi
ti
ona
l
in
f
or
m
a
ti
on
a
bout
r
is
k
f
a
c
to
r
s
a
s
s
oc
ia
t
e
d w
it
h
DM
.
T
hi
s
r
e
s
e
a
r
c
h
li
e
s
in
th
e
im
pl
e
m
e
nt
a
ti
on
of
th
e
X
G
B
oos
t
a
lg
or
i
th
m
f
or
p
r
e
di
c
ti
ng
D
M
r
is
k
ba
s
e
d
on
pa
ti
e
nt
s
’
he
a
lt
h
hi
s
to
r
y
a
nd
it
s
in
te
gr
a
ti
on
in
to
a
w
e
b
-
ba
s
e
d
s
ys
te
m
[
1]
.
U
nl
ik
e
m
a
ny
pr
e
vi
ous
s
tu
di
e
s
th
a
t
pr
im
a
r
il
y
r
e
li
e
d
on
c
onve
nt
io
na
l
m
ode
ls
s
uc
h
a
s
lo
gi
s
ti
c
r
e
gr
e
s
s
io
n
(
L
R
)
,
s
uppor
t
ve
c
to
r
m
a
c
hi
ne
(
S
V
M
)
,
or
r
a
ndom
f
or
e
s
t
(
R
F
)
,
th
is
s
tu
dy
de
m
ons
tr
a
te
s
th
e
s
upe
r
io
r
a
b
il
it
y
of
X
G
B
oos
t
to
ha
ndl
e
i
m
ba
la
nc
e
d
a
nd
c
om
pl
e
x
m
e
di
c
a
l
da
ta
s
e
ts
,
f
ur
th
e
r
opt
im
iz
e
d
th
r
ough
hype
r
p
a
r
a
m
e
te
r
tu
ni
ng
w
it
h
gr
id
s
e
a
r
c
h
a
nd
c
r
os
s
-
va
li
da
ti
on [
2]
. T
he
m
a
in
c
ont
r
ib
ut
io
ns
o
f
t
hi
s
w
or
k i
nc
lu
de
t
he
de
ve
lo
pm
e
nt
of
a
hi
gh
-
pe
r
f
o
r
m
a
nc
e
pr
e
di
c
ti
ve
m
ode
l
w
it
h
th
e
c
r
e
a
ti
on
of
a
us
e
r
-
f
r
ie
ndl
y
w
e
b
a
ppl
ic
a
ti
on
to
f
a
c
il
it
a
te
e
a
r
ly
de
te
c
ti
on
of
di
a
be
te
s
r
is
k,
a
nd
th
e
pr
ovi
s
io
n
of
pr
oba
bi
li
s
ti
c
out
put
s
th
a
t
of
f
e
r
m
or
e
nua
nc
e
d
de
c
is
io
n
s
uppor
t
f
or
he
a
lt
hc
a
r
e
pr
of
e
s
s
io
na
ls
.
I
n
a
ddi
ti
on,
th
is
r
e
s
e
a
r
c
h
c
ont
r
ib
ut
e
s
to
th
e
s
c
ie
nt
if
ic
c
om
m
uni
ty
by
hi
ghl
ig
ht
in
g
th
e
pr
a
c
ti
c
a
l
a
dva
nt
a
ge
s
of
X
G
B
oos
t
in
m
e
di
c
a
l
in
f
or
m
a
ti
c
s
a
nd
pr
ovi
di
ng
a
f
ounda
ti
on
f
or
f
ut
u
r
e
e
xpl
or
a
ti
on
of
hyb
r
id
or
e
ns
e
m
bl
e
m
ode
ls
f
or
di
s
e
a
s
e
pr
e
di
c
ti
on
[
3]
.
2.
R
E
V
I
E
W OF
L
I
T
E
R
A
T
U
R
E
T
h
is
s
e
c
ti
on
r
e
v
ie
w
s
pr
i
or
s
t
udi
e
s
a
pp
ly
i
ng
t
he
X
G
B
oo
s
t
a
l
g
or
it
h
m
i
n
m
e
di
c
a
l
a
nd
r
e
l
a
t
e
d
f
i
e
ld
s
.
R
e
s
e
a
r
c
h
[
14]
hi
g
hl
i
ght
s
t
he
im
p
or
t
a
n
c
e
of
f
e
a
t
ur
e
a
n
a
l
ys
is
in
X
G
B
o
os
t
f
or
c
a
r
di
o
va
s
c
u
la
r
di
s
e
a
s
e
pr
e
di
c
ti
on
,
s
h
ow
i
ng
s
tr
on
g
a
c
c
u
r
a
c
y
c
o
m
p
a
r
e
d
t
o
o
th
e
r
m
od
e
l
s
(
95%
f
or
RF
,
90%
f
or
S
V
M
,
a
nd
86%
f
or
LR
)
.
A
n
ot
h
e
r
s
t
udy
a
ppl
ie
d
X
G
B
oo
s
t
w
it
h
f
e
a
t
ur
e
s
e
l
e
c
ti
on a
n
d
h
yp
e
r
p
a
r
a
m
e
t
e
r
o
pt
i
m
i
z
a
ti
o
n
i
n s
p
a
m
de
te
c
ti
on
, a
c
hi
e
vi
ng a
n
a
c
c
ur
a
c
y
of
92
.67%
a
nd
o
ut
p
e
r
f
or
m
in
g
C
h
i
a
n
d
pr
in
c
ip
a
l
c
om
po
ne
n
t
a
n
a
l
ys
is
(
P
C
A
)
-
b
a
s
e
d
m
e
th
od
s
[
15]
,
[
16]
.
I
n
th
is
s
tu
dy,
hyp
e
r
pa
r
a
m
e
te
r
s
s
uc
h
a
s
m
a
x_d
e
pt
h,
le
a
r
ni
ng_r
a
te
,
a
nd
n_e
s
ti
m
a
to
r
s
w
e
r
e
s
ys
te
m
a
ti
c
a
ll
y
opt
im
iz
e
d
u
s
in
g
gr
id
s
e
a
r
c
h
w
it
h
5
-
f
ol
d
c
r
os
s
-
va
li
da
ti
on
to
b
a
la
nc
e
m
ode
l
c
om
pl
e
xi
ty
,
le
a
r
ni
ng
e
f
f
ic
ie
nc
y,
a
nd
ove
r
f
it
ti
ng
r
is
ks
[
4]
.
E
va
lu
a
ti
on
m
e
t
r
ic
s
in
c
lu
de
d
a
c
c
ur
a
c
y,
pr
e
c
is
io
n,
r
e
c
a
ll
,
a
nd
F1
-
s
c
or
e
to
e
ns
ur
e
c
li
ni
c
a
l
r
e
le
va
nc
e
in
di
a
be
te
s
pr
e
di
c
ti
o
n.
P
r
e
c
is
io
n
r
e
duc
e
s
f
a
ls
e
di
a
gno
s
e
s
,
r
e
c
a
ll
m
in
im
iz
e
s
unde
te
c
te
d c
a
s
e
s
, a
nd F
1
-
s
c
or
e
ba
la
n
c
e
s
bot
h, e
s
pe
c
i
a
ll
y us
e
f
ul
f
or
i
m
ba
la
nc
e
d da
ta
s
e
t
s
[
3]
.
W
hi
le
X
G
B
oos
t
de
m
ons
tr
a
te
s
s
tr
ong
pe
r
f
or
m
a
nc
e
f
or
di
a
be
te
s
r
is
k
pr
e
di
c
ti
on,
c
om
pa
r
is
ons
w
it
h
ot
he
r
m
ode
ls
a
r
e
e
s
s
e
nt
ia
l.
RF
of
f
e
r
s
r
obus
tn
e
s
s
but
hi
ghe
r
c
om
put
a
ti
on
c
os
ts
[
17]
;
LR
i
s
s
im
pl
e
r
y
e
t
le
s
s
e
f
f
e
c
ti
ve
w
it
h
non
-
li
ne
a
r
da
ta
[
18]
;
S
V
M
pe
r
f
or
m
s
w
e
ll
w
i
th
hi
gh
-
di
m
e
ns
io
na
l
da
ta
but
s
tr
uggl
e
s
w
it
h
s
c
a
la
bi
li
ty
;
a
nd
A
N
N
s
c
a
pt
ur
e
c
om
pl
e
x
p
a
tt
e
r
ns
but
r
e
qui
r
e
la
r
ge
da
ta
s
e
ts
a
nd
c
a
r
e
f
ul
tu
ni
ng.
B
e
n
c
hm
a
r
ki
ng
a
ga
in
s
t
th
e
s
e
m
ode
l
s
,
R
F
,
S
V
M
,
L
R
,
a
nd
a
r
ti
f
ic
ia
l
ne
ur
a
l
n
e
twor
k
(
A
N
N
)
is
ne
e
d
e
d
to
c
onf
ir
m
X
G
B
oos
t’
s
s
upe
r
io
r
it
y
in
c
om
pl
e
x
a
nd
im
ba
la
nc
e
d
m
e
di
c
a
l
d
a
ta
s
e
t
s
[
19]
.
F
ut
ur
e
r
e
s
e
a
r
c
h
s
h
oul
d
e
x
pa
nd
t
o
la
r
g
e
r
a
n
d
m
or
e
di
ve
r
s
e
d
a
t
a
s
e
t
s
,
i
nc
lu
di
n
g
r
e
a
l
-
ti
m
e
e
l
e
c
tr
o
ni
c
h
e
a
lt
h
r
e
c
o
r
ds
(
E
H
R
s
)
.
H
y
br
id
e
ns
e
m
bl
e
m
e
th
od
s
s
u
c
h
a
s
s
t
a
c
ki
ng
a
nd
ba
gg
in
g
m
a
y
e
nh
a
n
c
e
r
o
bu
s
tn
e
s
s
,
w
hi
le
in
t
e
g
r
a
ti
ng
e
x
pl
a
in
a
b
le
A
I
(
X
A
I
)
i
s
c
r
uc
i
a
l
f
or
in
t
e
r
pr
e
t
a
bi
li
t
y
a
n
d
c
li
ni
c
a
l
a
do
pt
i
on.
2.1. Diab
e
t
e
s
m
e
ll
it
u
s
DM
is
a
c
om
m
on
di
s
e
a
s
e
c
ha
r
a
c
te
r
iz
e
d
by
hi
gh
le
v
e
ls
of
s
uga
r
in
ur
in
e
a
nd
bl
ood.
T
he
te
r
m
“
di
a
be
te
s
”
c
om
e
s
f
r
om
th
e
G
r
e
e
k
di
a
ba
in
e
n,
m
e
a
ni
ng
“
to
f
lo
w
c
ont
in
uous
ly
,”
w
hi
le
“
m
e
ll
it
us
”
c
om
e
s
f
r
om
th
e
L
a
ti
n
m
e
ll
it
us
,
m
e
a
ni
ng
“
hone
y.”
T
hi
s
c
ondi
ti
on
oc
c
ur
s
w
he
n
th
e
pa
nc
r
e
a
s
doe
s
not
pr
oduc
e
s
uf
f
ic
ie
nt
in
s
ul
in
[
5]
.
I
ns
ul
in
pl
a
ys
a
c
r
uc
ia
l
r
ol
e
in
tr
a
ns
por
ti
ng
bl
ood
s
uga
r
,
de
r
iv
e
d
f
r
om
c
a
r
bohyd
r
a
te
br
e
a
kdown,
in
to
c
e
ll
s
f
or
e
ne
r
gy.
A
la
c
k
of
in
s
ul
in
le
a
d
s
to
e
l
e
va
te
d
bl
oo
d
s
uga
r
le
ve
ls
,
w
hi
c
h
,
ove
r
ti
m
e
m
a
y
da
m
a
ge
or
ga
ns
a
nd
body
ti
s
s
ue
s
.
C
om
m
on
s
ym
pt
om
s
in
c
lu
de
f
r
e
que
nt
ur
in
a
ti
on,
e
xc
e
s
s
iv
e
th
ir
s
t,
in
c
r
e
a
s
e
d
a
ppe
ti
te
,
s
udde
n w
e
ig
ht
l
os
s
, f
a
ti
gue
, obe
s
it
y, s
lo
w
-
he
a
li
ng w
ounds
, vi
s
i
on pr
obl
e
m
s
, a
nd i
tc
hi
ng
[
17]
,
[
19]
.
A
c
c
or
di
ng
to
th
e
A
m
e
r
ic
a
n
D
ia
be
te
s
A
s
s
oc
ia
ti
on
(
A
D
A
)
,
a
ne
w
pa
ti
e
nt
is
di
a
gnos
e
d
w
it
h
di
a
be
te
s
e
ve
r
y
21
s
e
c
onds
,
r
e
f
le
c
ti
ng
a
hi
gh
in
c
id
e
nc
e
r
a
te
[
2]
.
D
ia
be
te
s
is
a
le
a
di
ng
c
a
u
s
e
of
pr
e
m
a
tu
r
e
de
a
th
w
or
ld
w
id
e
a
nd c
ont
r
ib
ut
e
s
t
o s
e
ve
r
e
c
om
pl
ic
a
ti
ons
s
uc
h
a
s
bl
in
dne
s
s
, he
a
r
t
di
s
e
a
s
e
, a
nd kidne
y f
a
il
ur
e
[
18]
.
D
M
is
of
te
n
c
a
ll
e
d
a
“
s
il
e
nt
ki
ll
e
r
”
be
c
a
us
e
m
a
ny
p
a
ti
e
nt
s
r
e
m
a
in
una
w
a
r
e
of
th
e
ir
c
ondi
ti
on
unt
il
bl
ood
ve
s
s
e
l
da
m
a
ge
or
ot
he
r
c
om
pl
ic
a
ti
ons
oc
c
ur
.
R
is
k
f
a
c
to
r
s
a
r
e
not
li
m
it
e
d
to
ol
de
r
a
dul
ts
,
a
s
a
nyone
f
r
om
young
to
e
ld
e
r
ly
c
a
n
de
ve
lo
p
th
e
di
s
e
a
s
e
[
20]
.
L
if
e
s
ty
le
f
a
c
to
r
s
pl
a
y
a
s
ig
ni
f
ic
a
nt
r
ol
e
,
w
it
h
D
M
f
r
e
que
nt
ly
a
s
s
oc
ia
t
e
d w
it
h c
hr
oni
c
c
ondi
ti
ons
s
u
c
h a
s
hyp
e
r
te
ns
io
n a
nd hi
gh c
hol
e
s
te
r
ol
l
e
ve
ls
[
18]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8814
I
nt
J
A
dv A
ppl
S
c
i
,
V
ol
.
14
, N
o.
4
,
D
e
c
e
m
be
r
20
25
:
1028
-
1039
1030
2.2. P
r
e
d
ic
t
io
n
P
r
e
di
c
ti
on
in
th
e
c
ont
e
xt
of
m
a
c
hi
ne
le
a
r
ni
ng
is
th
e
pr
oc
e
s
s
of
e
s
ti
m
a
ti
ng
f
ut
ur
e
out
c
om
e
s
or
va
lu
e
s
ba
s
e
d
on
hi
s
to
r
ic
a
l
da
ta
[
21]
.
T
hi
s
hi
s
to
r
ic
a
l
in
f
or
m
a
ti
on
m
a
y
c
om
e
f
r
om
th
e
c
ur
r
e
nt
da
ta
s
e
t
or
f
r
om
s
pe
c
if
ic
in
put
da
ta
[
21]
.
T
he
goa
l
is
to
de
ve
lo
p
a
m
od
e
l
th
a
t
c
a
n
le
a
r
n
pa
tt
e
r
ns
f
r
om
e
xi
s
ti
ng
da
ta
a
nd
th
e
n
us
e
it
to
pr
e
di
c
t
out
c
om
e
s
on uns
e
e
n da
ta
, a
ll
ow
in
g pr
e
ve
nt
iv
e
m
e
a
s
ur
e
s
t
o be
i
m
pl
e
m
e
nt
e
d e
a
r
li
e
r
[
22]
.
2.3.
D
e
c
is
io
n
t
r
e
e
I
n
ge
ne
r
a
l,
a
de
c
is
io
n
tr
e
e
i
s
a
d
e
c
is
io
n
-
m
a
ki
ng
te
c
hni
qu
e
th
a
t
or
ga
ni
z
e
s
e
a
c
h
c
hoi
c
e
in
to
a
br
a
nc
hi
ng
s
tr
uc
tu
r
e
[
23]
.
A
c
c
or
di
ng
to
[
24]
,
de
c
is
io
n
tr
e
e
s
a
r
e
a
popula
r
a
nd
e
f
f
e
c
ti
ve
m
e
th
od
in
th
e
f
ie
ld
s
of
c
la
s
s
if
ic
a
ti
on a
nd pr
e
di
c
ti
on t
ha
t
tr
a
ns
f
or
m
da
ta
i
nt
o
a
n e
a
s
il
y unde
r
s
ta
nda
bl
e
de
c
is
io
n r
ul
e
t
r
e
e
. T
hi
s
r
ul
e
c
a
n
be
e
xpr
e
s
s
e
d
by s
e
a
r
c
hi
ng
f
or
or
f
il
te
r
in
g
da
ta
a
c
c
or
di
ng
to
th
e
c
r
it
e
r
ia
ge
ne
r
a
te
d
by
th
e
c
la
s
s
if
ic
a
ti
on
m
ode
l.
D
e
c
is
io
n t
r
e
e
s
r
e
duc
e
a
l
a
r
ge
da
ta
s
e
t
in
to
a
s
m
a
ll
e
r
da
ta
s
e
t.
W
h
e
r
e
e
a
c
h l
e
a
f
node
r
e
pr
e
s
e
nt
s
a
c
la
s
s
l
a
b
e
l.
T
he
r
oot
a
nd
in
te
r
na
l
node
s
c
ont
a
in
a
tt
r
ib
ut
e
te
s
t
c
ondi
ti
ons
,
u
s
ua
ll
y
in
th
e
s
ha
pe
of
a
n
ova
l,
w
hi
le
th
e
le
a
f
node
s
a
r
e
s
qua
r
e
-
s
ha
pe
d.
A
de
c
is
io
n t
r
e
e
c
la
s
s
if
ie
s
da
ta
by a
s
ki
ng que
s
ti
ons
r
e
la
te
d t
o t
h
e
c
ha
r
a
c
te
r
is
ti
c
s
of
e
a
c
h f
e
a
tu
r
e
i
n t
he
da
ta
.
E
a
c
h
que
s
ti
on
is
r
e
pr
e
s
e
nt
e
d
by
a
n
in
te
r
na
l
node
c
onne
c
te
d
to
c
hi
ld
node
s
th
a
t
r
e
pr
e
s
e
nt
th
e
a
ns
w
e
r
s
to
th
e
que
s
ti
on.
I
n
th
is
w
a
y,
th
e
qu
e
s
ti
ons
f
or
m
a
hi
e
r
a
r
c
hi
c
a
l
s
tr
uc
tu
r
e
in
th
e
s
ha
p
e
of
a
tr
e
e
.
T
he
de
c
is
io
n
tr
e
e
is
bui
lt
by
gr
a
dua
ll
y
a
ddi
ng
que
s
ti
on
node
s
,
gui
de
d
by
la
b
e
le
d
tr
a
in
in
g
da
ta
.
D
e
c
is
io
n
tr
e
e
s
na
tu
r
a
ll
y
di
vi
de
is
s
ue
s
in
to
s
m
a
ll
e
r
c
om
pon
e
nt
s
in
or
de
r
to
s
ol
ve
th
e
m
,
u
s
e
d
to
di
vi
de
da
ta
in
to
m
or
e
s
pe
c
if
ic
s
ubn
e
ts
by
c
ont
in
uous
ly
a
s
ki
ng
qu
e
s
ti
ons
unt
il
r
e
a
c
hi
ng
a
c
ondi
ti
on
w
h
e
r
e
f
ur
th
e
r
di
vi
s
io
n
is
no
lo
nge
r
ne
c
e
s
s
a
r
y
s
how
n
in
F
ig
ur
e
1
[
15]
.
F
ig
ur
e
1.
D
e
c
is
io
n t
r
e
e
s
tr
uc
tu
r
e
2.4. E
xt
r
e
m
e
g
r
ad
ie
n
t
b
oos
t
in
g
X
G
B
oos
t
w
a
s
in
tr
oduc
e
d
by
W
a
ng
e
t
al
.
[
10]
a
s
a
de
ve
lo
pm
e
nt
of
th
e
gr
a
di
e
nt
boos
ti
ng
de
c
is
io
n
tr
e
e
(
G
B
D
T
)
pr
e
vi
ous
ly
in
tr
oduc
e
d
by
[
25]
,
[
26]
.
T
hi
s
a
lg
or
it
hm
bui
ld
s
de
c
is
io
n
tr
e
e
s
it
e
r
a
ti
ve
ly
,
w
he
r
e
e
a
c
h
ne
w
tr
e
e
c
or
r
e
c
ts
th
e
e
r
r
or
s
of
th
e
pr
e
vi
ous
one
.
T
he
w
e
ig
ht
s
of
e
a
c
h
tr
e
e
a
r
e
a
l
s
o
a
dj
u
s
te
d,
r
e
s
ul
ti
ng
in
a
s
tr
onge
r
a
nd mor
e
r
e
li
a
bl
e
m
ode
l
[
27]
–
[
30]
.
I
n
th
is
s
tu
dy,
X
G
B
oos
t
w
a
s
c
om
pa
r
e
d
w
it
h
two
ba
s
ic
m
ode
l
s
c
om
m
onl
y
us
e
d
in
m
e
di
c
a
l
pr
e
di
c
ti
on:
LR
a
nd
de
c
i
s
io
n
tr
e
e
s
.
L
R
w
or
ks
w
e
ll
f
or
s
im
pl
e
r
c
a
s
e
s
,
but
l
a
c
ks
a
c
c
ur
a
c
y
in
ha
ndl
in
g
c
om
pl
e
x
non
-
li
ne
a
r
r
e
la
ti
ons
hi
ps
[
6]
.
D
e
c
i
s
io
n
tr
e
e
s
pr
ovi
de
b
e
tt
e
r
in
te
r
pr
e
ta
bi
li
ty
but
a
r
e
pr
one
to
ove
r
f
it
ti
ng
w
he
n
de
a
li
ng
w
it
h
da
ta
s
e
ts
c
ont
a
in
in
g
m
a
ny
f
e
a
tu
r
e
s
or
noi
s
e
.
I
n
c
ont
r
a
s
t,
X
G
B
oos
t
c
om
bi
ne
s
m
ul
ti
pl
e
de
c
is
io
n
tr
e
e
s
th
r
ough
e
ns
e
m
bl
e
le
a
r
ni
ng,
e
f
f
e
c
ti
ve
ly
ha
ndl
in
g
f
e
a
tu
r
e
in
te
r
a
c
ti
on
s
a
nd
d
e
li
ve
r
in
g
m
or
e
s
t
a
bl
e
a
nd
a
c
c
ur
a
t
e
pr
e
di
c
ti
ons
f
or
c
om
pl
e
x m
e
di
c
a
l
da
ta
s
e
t
s
.
T
he
m
a
in
a
dva
nt
a
ge
of
X
G
B
oos
t
li
e
s
i
n i
ts
a
bi
li
ty
t
o pr
ovi
de
h
i
ghl
y a
c
c
ur
a
te
pr
e
di
c
ti
ons
of
D
M
r
is
k,
a
c
hi
e
vi
ng
99.79%
tr
a
in
in
g
a
c
c
ur
a
c
y
a
nd
98.88%
te
s
ti
ng
a
c
c
ur
a
c
y,
s
ig
ni
f
ic
a
nt
ly
out
pe
r
f
or
m
in
g
LR
a
nd
de
c
is
io
n
tr
e
e
s
[
7]
.
M
or
e
ove
r
,
X
G
B
oos
t
is
m
or
e
e
f
f
e
c
ti
ve
a
t
ha
ndl
in
g
im
ba
la
nc
e
d
da
ta
,
w
hi
c
h
is
c
om
m
on
in
di
s
e
a
s
e
pr
e
di
c
ti
on
c
a
s
e
s
.
A
lt
hough
in
te
r
pr
e
ta
bi
li
ty
r
e
m
a
in
s
a
c
ha
ll
e
nge
,
X
G
B
oos
t’
s
a
bi
li
ty
to
m
a
na
ge
da
ta
c
om
pl
e
xi
ty
a
nd
de
li
ve
r
c
ons
is
te
nt
r
e
s
ul
ts
m
a
ke
s
it
a
s
tr
ong
c
a
ndi
da
te
f
or
m
e
di
c
a
l
a
ppl
ic
a
ti
ons
,
pa
r
ti
c
ul
a
r
ly
in
e
a
r
ly
di
a
be
te
s
r
is
k de
te
c
ti
on.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
dv A
ppl
S
c
i
I
S
S
N
:
2252
-
8814
I
m
pl
e
m
e
nt
at
io
n of
X
G
B
oos
t
fo
r
di
abe
te
s
m
e
ll
it
us
r
i
s
k
pr
e
di
c
ti
on bas
e
d on he
al
th
hi
s
to
r
y
(
A
ndi
R
ia
ns
y
ah)
1031
T
e
c
hni
c
a
ll
y,
X
G
B
oos
t
w
or
ks
by
bui
ld
in
g
r
e
gr
e
s
s
io
n
tr
e
e
s
th
a
t
m
a
p
da
ta
poi
nt
s
to
le
a
ve
s
w
it
h
c
ont
in
uous
s
c
or
e
s
.
I
t
m
in
im
iz
e
s
a
r
e
gul
a
r
iz
e
d
obj
e
c
ti
ve
f
unc
ti
on
(
L
1
a
nd
L
2)
a
nd
a
c
onve
x
lo
s
s
f
unc
ti
on,
us
in
g
gr
a
di
e
nt
de
s
c
e
nt
to
r
e
duc
e
e
r
r
or
.
E
a
c
h
ne
w
tr
e
e
is
tr
a
in
e
d
to
pr
e
di
c
t
th
e
r
e
s
id
ua
ls
of
th
e
pr
e
vi
ou
s
one
s
,
a
nd
th
e
e
ns
e
m
bl
e
of
tr
e
e
s
ul
ti
m
a
te
ly
pr
oduc
e
s
th
e
f
in
a
l
pr
e
di
c
ti
on
[
25]
.
F
ig
ur
e
2
il
lu
s
tr
a
te
s
how
X
G
B
oos
t
w
or
ks
,
w
he
r
e
\
pr
opt
o_i
is
th
e
r
e
gul
a
r
iz
a
ti
on
pa
r
a
m
e
te
r
a
nd
r
_i
i
s
th
e
r
e
s
id
ue
c
a
lc
ul
a
te
d
w
it
h
th
e
f
ir
s
t
tr
e
e
,
a
nd
h_i
is
a
f
unc
ti
on
th
a
t
ha
s
b
e
e
n
ta
ught
to
f
or
e
c
a
s
t
r
e
s
id
ua
l
s
,
r
_i
us
in
g
X
f
or
th
e
f
ir
s
t
tr
e
e
.
T
o
c
a
lc
ul
a
te
\
pr
opt
o_i
us
in
g t
he
r
e
s
id
ue
c
a
l
c
ul
a
te
d r
_i
i
t
c
a
n be
c
om
put
e
d u
s
in
g
(
1
).
arg
∝
=
∑
(
,
−
1
(
)
+
∝
ℎ
(
,
−
1
)
)
−
1
(
1)
(
,
(
)
)
is
a
di
f
f
e
r
e
nt
ia
te
d l
os
s
f
unc
ti
on
.
I
nf
or
m
a
ti
on
:
–
D
a
ta
c
ol
le
c
ti
on (
X,
Y
)
:
th
e
pr
oc
e
s
s
be
gi
ns
w
it
h a
da
t
a
s
e
t
c
on
s
is
ti
ng of
f
e
a
tu
r
e
X
a
nd t
a
r
ge
t
Y
.
–
T
r
e
e
1 (
F
1(
X
)
)
:
th
e
f
ir
s
t
m
ode
l
is
t
r
a
in
e
d us
in
g a
da
ta
s
e
t.
A
f
te
r
t
he
t
r
e
e
i
s
c
r
e
a
te
d, r
1 i
s
c
a
lc
ul
a
te
d, w
hi
c
h
is
t
he
di
f
f
e
r
e
nc
e
be
twe
e
n t
he
pr
e
di
c
ti
ons
of
t
r
e
e
1 a
nd t
h
e
a
c
tu
a
l
va
lu
e
Y
.
–
C
a
lc
ul
a
te
1
:
th
e
va
lu
e
of
α1
is
c
a
lc
ul
a
te
d,
w
hi
c
h
i
s
a
r
e
gul
a
r
iz
a
ti
on
pa
r
a
m
e
te
r
th
a
t
he
lp
s
r
e
duc
e
ove
r
f
it
ti
ng
.
–
T
r
e
e
2
(
F
2(
X
)
)
:
th
e
s
e
c
ond
m
ode
l,
tr
a
in
e
d
to
pr
e
di
c
t
th
e
r
e
s
id
u
a
l
r
1.
A
f
te
r
tr
e
e
2
is
tr
a
in
e
d,
th
e
r
e
s
id
ua
l
r
2 i
s
c
a
lc
ul
a
te
d, w
hi
c
h i
s
t
h
e
di
f
f
e
r
e
nc
e
be
twe
e
n t
he
pr
e
di
c
ti
on
s
of
tr
e
e
2 a
nd t
he
r
e
s
id
ua
l
r
1
–
C
a
lc
ul
a
te
:
th
e
va
lu
e
of
α2 is
c
a
lc
ul
a
te
d f
or
t
r
e
e
2.
–
I
te
r
a
ti
on
pr
oc
e
s
s
:
th
is
pr
oc
e
s
s
is
r
e
pe
a
te
d
f
or
m
a
ny
t
r
e
e
s
.
(
f
r
om
tr
e
e
3
to
tr
e
e
m
)
.
E
a
c
h
ne
w
t
r
e
e
is
tr
a
in
e
d t
o pr
e
di
c
t
t
he
r
e
s
id
ua
ls
f
r
om
t
he
pr
e
vi
ous
t
r
e
e
, a
nd ne
w
r
e
s
id
ua
ls
a
r
e
c
a
lc
ul
a
te
d a
t
e
a
c
h s
te
p. T
he
va
lu
e
of
α
i
is
a
ls
o c
a
lc
ul
a
te
d e
a
c
h t
im
e
f
or
e
a
c
h t
r
e
e
i
.
–
T
h
e
f
i
na
l
m
od
e
l
F
m
(
X
)
:
a
c
om
bi
n
a
t
io
n
of
e
a
c
h
t
r
e
e
t
h
a
t
h
a
s
b
e
e
n
tr
a
in
e
d.
T
hi
s
c
o
m
b
in
a
ti
on
i
s
w
r
it
te
n
a
s
(
2)
.
(
)
=
−
1
(
)
+
ℎ
(
,
−
1
)
(
2)
W
h
e
r
e
hi
i
s
a
f
unc
ti
on t
r
a
in
e
d t
o pr
e
di
c
t
th
e
r
e
s
id
ua
l
r
i
a
t
th
e
i
-
th
t
r
e
e
.
–
O
pt
i
m
i
z
in
g
α: t
o c
a
lc
ul
a
t
e
t
he
va
lu
e
of
α, the
n m
in
im
iz
e
t
he
di
f
f
e
r
e
nt
ia
bl
e
l
os
s
f
unc
ti
on L
, us
in
g
(
3)
.
∝
∑
(
,
−
1
(
)
+
∝
ℎ
(
,
−
1
)
)
=
1
(
3)
I
nf
or
m
a
ti
on
:
L
is
lo
s
s
f
unc
ti
on
,
is
th
e
a
c
tu
a
l
ta
r
ge
t
va
lu
e
(
0
or
1)
,
−
1
(
)
is
pr
e
di
c
ti
on
of
th
e
m
ode
l
in
th
e
pr
e
vi
ous
it
e
r
a
ti
on
,
∝
is
th
e
c
oe
f
f
ic
ie
nt
th
a
t
de
te
r
m
in
e
s
th
e
e
xt
e
nt
of
th
e
c
ont
r
ib
ut
io
n
o
f
th
e
ne
w
m
ode
l
h_i
,
ℎ
(
,
−
1
)
is
th
e
ne
w
m
ode
l
a
dde
d
in
th
e
i
it
e
r
a
ti
on
,
a
nd
−
1
is
r
e
s
id
ua
l
or
e
r
r
or
in
th
e
pr
e
vi
ous
i
te
r
a
ti
on.
F
ig
ur
e
2. H
ow
X
G
B
oos
t
w
or
ks
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8814
I
nt
J
A
dv A
ppl
S
c
i
,
V
ol
.
14
, N
o.
4
,
D
e
c
e
m
be
r
20
25
:
1028
-
1039
1032
T
he
us
e
of
te
c
hnol
ogy
,
s
uc
h
a
s
m
a
c
hi
ne
le
a
r
ni
ng
a
lg
or
it
hm
s
in
pr
e
di
c
ti
ng
di
a
be
te
s
,
is
a
l
s
o
r
a
pi
dl
y
a
dva
nc
in
g, m
a
ki
ng i
t
im
por
ta
nt
t
o r
e
c
ogni
z
e
w
he
th
e
r
s
om
e
one
f
a
ll
s
i
nt
o t
he
c
a
te
gor
y of
pa
ti
e
nt
s
w
it
h di
a
be
te
s
be
f
or
e
th
os
e
w
it
h
D
M
e
xp
e
r
ie
nc
e
s
e
r
io
us
c
om
pl
ic
a
ti
ons
,
s
o
th
a
t
th
e
y
c
a
n
r
e
c
e
iv
e
a
ppr
opr
ia
te
a
nd
ti
m
e
ly
tr
e
a
tm
e
nt
.
T
he
X
G
B
oos
t
a
lg
or
it
hm
ha
s
be
c
om
e
popula
r
f
or
pr
e
di
c
ti
ng
di
s
e
a
s
e
s
.
T
h
e
r
e
f
or
e
,
in
th
is
r
e
por
t,
w
e
w
il
l
ut
il
iz
e
th
e
X
G
B
oos
t
a
lg
or
it
hm
a
s
one
of
th
e
im
pl
e
m
e
nt
a
ti
ons
to
ge
ne
r
a
te
pr
e
di
c
ti
ons
f
or
th
e
r
is
k
of
DM
.
T
hi
s
s
ys
te
m
is
e
xpe
c
te
d
to
be
be
ne
f
ic
ia
l
f
or
he
a
lt
hc
a
r
e
w
or
k
e
r
s
a
nd
to
he
lp
im
p
r
ove
hos
pi
ta
l
s
e
r
vi
c
e
s
f
or
pa
ti
e
nt
s
, e
s
p
e
c
ia
ll
y i
n e
a
r
ly
de
te
c
ti
on, pa
r
ti
c
ul
a
r
ly
of
di
a
be
te
s
m
e
ll
it
us
, a
nd i
n pe
r
s
ona
li
z
in
g t
r
e
a
tm
e
nt
.
3.
M
E
T
H
O
D
A
n
a
ddi
ti
ona
l
e
xpl
a
na
ti
on
to
s
tr
e
ngt
he
n
th
e
a
ppl
ie
d
s
c
ie
nc
e
f
r
a
m
in
g
a
nd
it
s
r
e
le
va
nc
e
to
I
ndone
s
ia
’
s
he
a
lt
hc
a
r
e
c
ont
e
xt
:
X
G
B
oos
t
c
a
n be
ope
r
a
ti
ona
li
z
e
d a
s
a
n i
nt
e
ll
ig
e
nc
e
c
om
pone
nt
i
n
c
li
ni
c
a
l
de
c
is
io
n s
uppor
t
a
t
P
us
ke
s
m
a
s
a
nd
hos
pi
ta
l
s
f
or
e
a
r
ly
di
a
be
te
s
s
c
r
e
e
ni
ng
,
f
or
e
xa
m
pl
e
,
in
te
gr
a
te
d
w
it
h
e
le
c
tr
oni
c
m
e
di
c
a
l
r
e
c
or
ds
a
nd
r
out
in
e
a
na
m
ne
s
is
f
or
m
s
s
o
th
a
t
he
a
lt
h
w
or
ke
r
s
r
e
c
e
iv
e
a
r
e
a
l
-
ti
m
e
r
is
k
s
c
or
e
a
nd
it
s
c
ont
r
ib
ut
in
g
f
a
c
to
r
s
.
I
ts
e
f
f
ic
ie
nc
y
a
nd
s
c
a
la
bi
li
ty
e
na
bl
e
de
pl
oym
e
nt
in
o
f
f
li
ne
-
f
i
r
s
t
m
obi
le
a
ppl
ic
a
ti
ons
f
or
a
r
e
a
s
w
it
h
li
m
it
e
d c
onne
c
ti
vi
ty
, w
hi
le
i
ts
i
nt
e
r
pr
e
ta
bi
li
ty
he
lp
s
c
li
ni
c
ia
ns
unde
r
s
ta
nd t
he
i
nf
lu
e
nc
e
of
l
oc
a
l
va
r
ia
bl
e
s
s
uc
h
a
s
body
m
a
s
s
in
de
x
,
f
a
s
ti
ng
bl
ood
gl
uc
os
e
,
bl
ood
pr
e
s
s
ur
e
,
f
a
m
il
y
hi
s
to
r
y,
di
e
t,
a
nd
phys
ic
a
l
a
c
ti
vi
ty
,
m
a
ki
ng
f
ol
lo
w
-
up
r
e
c
om
m
e
nda
ti
ons
m
or
e
a
c
ti
ona
bl
e
.
I
n
a
ddi
ti
on,
th
is
a
ppr
oa
c
h
f
it
s
I
ndone
s
ia
’
s
ne
e
d
f
o
r
r
e
s
our
c
e
-
e
f
f
ic
ie
nt
s
ol
ut
io
ns
:
X
G
B
oo
s
t
ha
ndl
e
s
im
ba
la
nc
e
d
d
a
t
a
a
nd
m
a
ny
f
e
a
tu
r
e
s
w
i
th
o
ut
c
o
s
tl
y
in
f
r
a
s
tr
uc
tu
r
e
,
it
s
a
c
c
ur
a
c
y c
a
n b
e
m
oni
to
r
e
d a
c
r
o
s
s
r
e
g
io
n
s
, a
nd
it
c
a
n
b
e
a
li
gn
e
d
w
it
h
l
oc
a
l
po
li
c
ie
s
s
o
i
t
i
s
r
e
a
d
y
f
or
u
s
e
i
n
non
-
c
om
m
uni
c
a
bl
e
di
s
e
a
s
e
s
pr
ogr
a
m
m
a
na
g
e
m
e
nt
da
s
hbo
a
r
ds
a
nd e
ve
r
yda
y c
li
ni
c
a
l
w
or
kf
lo
w
s
.
3.1.
D
at
as
e
t
T
hi
s
da
ta
s
e
t
is
c
a
ll
e
d
e
a
r
ly
-
s
ta
ge
di
a
be
t
e
s
r
is
k pr
e
di
c
ti
on dia
be
te
s
a
nd w
a
s
dow
nl
oa
d
e
d f
r
om
K
a
ggl
e
.
T
he
da
ta
us
e
d
in
th
is
r
e
s
e
a
r
c
h
is
s
e
c
onda
r
y
da
ta
obt
a
in
e
d
f
r
om
que
s
ti
onna
ir
e
s
f
il
le
d
out
by
pa
ti
e
nt
s
a
t
S
yl
he
t
D
ia
be
ti
c
H
o
s
pi
ta
l
in
S
yl
he
t,
B
a
ngl
a
de
s
h, a
nd
ha
s
be
e
n
a
ppr
ove
d
by
a
doc
to
r
.
T
hi
s
r
e
s
e
a
r
c
h
in
vol
ve
s
a
to
ta
l
of
520 da
ta
poi
nt
s
, c
ons
i
s
ti
ng of
320 da
ta
poi
nt
s
f
or
t
he
pos
it
iv
e
D
M
c
a
te
gor
y a
nd 200 f
or
t
he
non
-
D
M
c
a
te
gor
y.
T
hi
s
da
ta
s
e
t
c
ove
r
s
a
v
a
r
ie
ty
of
s
ym
pt
om
s
a
s
s
o
c
ia
te
d
w
it
h
di
a
be
te
s
[
29]
.
T
he
da
ta
s
uppor
ti
ng
th
e
f
in
di
ngs
of
th
is
s
tu
dy
w
e
r
e
obt
a
in
e
d
f
r
om
va
r
io
us
s
c
ie
nt
if
ic
a
r
ti
c
le
s
a
nd
m
e
di
c
a
l
r
e
c
or
ds
r
e
la
te
d
to
di
a
be
t
e
s
r
is
k
pr
e
di
c
ti
on. T
he
da
ta
w
e
r
e
pr
oc
e
s
s
e
d a
nd a
nonymi
z
e
d pr
io
r
t
o u
s
e
i
n t
he
de
ve
lo
pm
e
nt
of
t
he
pr
e
di
c
ti
on mode
l
.
D
ue
to
pr
iv
a
c
y
a
nd
e
th
ic
a
l
c
ons
id
e
r
a
ti
ons
,
th
e
da
ta
a
r
e
not
pu
bl
ic
ly
a
va
il
a
bl
e
but
m
a
y
be
obt
a
in
e
d
f
r
om
th
e
c
or
r
e
s
ponding a
ut
hor
upon r
e
a
s
ona
bl
e
r
e
que
s
t.
3.2.
P
r
e
p
r
oc
e
s
s
in
g
T
hi
s
pr
oc
e
s
s
pr
e
pa
r
e
s
th
e
da
ta
f
or
pr
oc
e
s
s
in
g
in
m
ode
li
ng
w
it
h
th
e
X
G
B
oos
t
a
lg
or
it
hm
.
T
he
in
it
ia
l
s
te
p
in
vol
ve
s
la
be
li
ng
(
la
be
l
e
nc
odi
ng)
,
w
hi
c
h
is
th
e
pr
oc
e
s
s
of
c
onve
r
ti
ng
da
ta
f
e
a
tu
r
e
s
th
a
t
a
r
e
or
ig
in
a
ll
y
te
xt
in
to
num
e
r
ic
va
lu
e
s
[
29]
.
S
e
c
ond,
f
e
a
tu
r
e
s
a
nd
ta
r
ge
ts
a
r
e
s
e
pa
r
a
te
d
to
e
ns
ur
e
th
a
t
th
e
m
ode
l
is
tr
a
in
e
d
onl
y
on
r
e
le
va
nt
f
e
a
tu
r
e
s
a
nd
u
s
e
s
th
e
ta
r
ge
t
to
m
a
ke
pr
e
di
c
ti
ons
[
9]
.
N
e
xt
,
s
t
a
nda
r
di
z
e
th
e
da
ta
to
a
dj
us
t
it
s
o
th
a
t
a
ll
f
e
a
tu
r
e
s
a
r
e
w
it
hi
n
a
uni
f
or
m
r
a
nge
[
25]
. T
he
n,
th
e
da
ta
is
di
vi
de
d
in
to
two
w
it
h
a
di
s
tr
ib
ut
io
n
r
a
ti
o
of
80:
20,
r
e
s
ul
ti
ng
in
416
da
ta
poi
nt
s
f
or
tr
a
in
in
g
a
nd
104
da
ta
poi
nt
s
f
or
te
s
ti
ng.
D
a
ta
pa
r
ti
ti
oni
ng
is
c
a
r
r
ie
d
out
to
e
ns
ur
e
t
ha
t
th
e
m
ode
l
is
t
r
a
in
e
d us
in
g s
pe
c
if
ic
da
ta
a
nd i
ts
pe
r
f
or
m
a
nc
e
i
s
t
e
s
te
d us
in
g f
or
e
ig
n
da
ta
t
ha
t
ha
s
not
be
e
n s
e
e
n dur
in
g t
r
a
in
in
g
[
29]
.
3.3.
M
od
e
ll
in
g
A
f
te
r
th
e
da
ta
s
pl
it
ti
ng
pr
oc
e
s
s
,
th
e
f
or
m
a
t
is
c
onve
r
te
d
in
to
D
m
a
tr
ix
,
w
hi
c
h
is
a
s
pe
c
ia
l
f
or
m
a
t
f
or
X
G
B
oos
t
de
s
ig
ne
d
f
or
da
ta
s
to
r
a
ge
.
T
he
n
e
xt
pr
oc
e
s
s
is
to
de
t
e
r
m
in
e
th
e
hype
r
pa
r
a
m
e
te
r
s
us
in
g
gr
id
s
e
a
r
c
h
w
it
h 5
-
f
ol
d
c
r
os
s
-
va
li
da
ti
on
. T
he
hype
r
pa
r
a
m
e
te
r
s
us
e
d a
r
e
a
s
s
how
n
in
T
a
bl
e
1
.
T
a
bl
e
1. H
ype
r
pa
r
a
m
e
te
r
s
us
e
d
H
ype
r
pa
r
a
m
e
t
e
r
V
a
l
ue
I
nf
or
m
a
t
i
on
m
a
x_de
pt
h
[
3, 4, 5]
T
he
m
a
xi
m
um
de
pt
h of
e
a
c
h t
r
e
e
.
l
e
a
r
ni
ng_r
a
t
e
[
0.01, 0.1, 0.2]
C
ont
r
ol
l
i
ng t
he
s
t
e
p s
i
z
e
/
how
qui
c
kl
y t
he
l
e
a
r
ni
ng f
r
om
m
i
s
t
a
ke
s
m
ode
ope
r
a
t
e
s
i
n
e
a
c
h i
t
e
r
a
t
i
on.
n_e
s
t
i
m
a
t
or
s
[
550, 100, 200]
T
he
t
ot
a
l
num
be
r
of
t
r
e
e
s
t
ha
t
w
i
l
l
be
bui
l
t
by t
he
m
ode
l
i
n t
he
boos
t
i
ng pr
oc
e
s
s
.
T
he
be
s
t
hype
r
pa
r
a
m
e
te
r
c
om
bi
na
ti
on
ba
s
e
d
on
pe
r
f
or
m
a
nc
e
dur
in
g
5
it
e
r
a
ti
ons
(
c
r
os
s
-
va
li
da
ti
on)
w
il
l
be
s
a
v
e
d
a
s
th
e
m
ode
l
f
or
m
a
ki
ng
pr
e
di
c
ti
ons
on
th
e
te
s
ti
n
g
da
ta
. T
hi
s
pr
oc
e
s
s
i
s
c
a
r
r
ie
d
out
us
in
g
5
-
f
ol
d
c
r
os
s
-
va
li
da
ti
on,
w
he
r
e
th
e
m
ode
l
is
e
va
lu
a
te
d
f
or
e
a
c
h
hype
r
pa
r
a
m
e
te
r
c
om
bi
na
ti
on
to
obt
a
in
th
e
be
s
t
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
dv A
ppl
S
c
i
I
S
S
N
:
2252
-
8814
I
m
pl
e
m
e
nt
at
io
n of
X
G
B
oos
t
fo
r
di
abe
te
s
m
e
ll
it
us
r
i
s
k
pr
e
di
c
ti
on bas
e
d on he
al
th
hi
s
to
r
y
(
A
ndi
R
ia
ns
y
ah)
1033
pa
r
a
m
e
te
r
c
om
bi
na
ti
on
th
a
t
yi
e
ld
s
th
e
m
os
t
opt
im
a
l
m
ode
l
pe
r
f
or
m
a
nc
e
.
T
he
n,
to
m
a
xi
m
iz
e
th
e
m
ode
l'
s
pe
r
f
or
m
a
nc
e
,
th
e
m
ode
l
is
tr
a
in
e
d
w
it
h
tr
a
in
in
g
da
t
a
a
nd
e
va
lu
a
te
d
us
in
g
te
s
t
d
a
ta
[
16]
.
T
hi
s
pr
oc
e
s
s
in
vol
ve
s
r
e
pe
a
te
d
it
e
r
a
ti
ons
ove
r
5
ti
m
e
s
to
m
in
im
iz
e
pr
e
di
c
ti
on e
r
r
or
s
a
nd
r
e
duc
e
ove
r
f
it
ti
ng,
m
a
ki
ng
th
e
m
ode
l
m
or
e
r
e
li
a
bl
e
in
m
a
ki
ng
pr
e
di
c
ti
ons
on
d
a
ta
th
a
t
ha
s
not
be
e
n
s
e
e
n
a
t
a
ll
dur
in
g
th
e
m
ode
l
tr
a
in
in
g
pr
oc
e
s
s
.
A
f
te
r
th
e
tr
a
in
in
g
a
nd
opt
im
iz
a
ti
on
pr
oc
e
s
s
,
th
e
m
ode
l
i
s
s
a
ve
d
f
or
f
ut
ur
e
us
e
a
nd
c
a
n
be
a
ppl
ie
d
in
r
e
a
l
-
w
or
ld
a
ppl
ic
a
ti
ons
f
or
a
ut
om
a
ti
c
di
a
be
te
s
r
is
k
pr
e
di
c
ti
on.
W
it
h
th
is
a
ppr
oa
c
h,
a
n
a
c
c
ur
a
te
a
nd
r
e
li
a
bl
e
di
a
be
te
s
r
is
k
pr
e
di
c
ti
on
m
ode
l
c
a
n
be
bui
lt
u
s
in
g
th
e
X
G
B
oos
t
a
lg
or
it
hm
.
T
he
f
ol
lo
w
in
g
is
th
e
s
y
s
te
m
m
ode
li
ng
d
e
s
ig
n
in
th
e
f
or
m
of
a
f
lo
w
c
ha
r
t
,
a
s
s
how
n i
n F
ig
ur
e
3.
F
ig
ur
e
3.
S
ys
te
m
m
ode
li
ng f
lo
w
c
ha
r
t
3.4.
M
od
e
l
e
val
u
at
io
n
T
he
m
ode
l
th
a
t
w
il
l
be
e
va
lu
a
te
d
in
th
is
r
e
s
e
a
r
c
h
us
e
s
a
c
onf
us
io
n
m
a
tr
ix
,
w
hi
c
h
is
one
w
a
y
to
m
e
a
s
ur
e
a
m
ode
l'
s
pe
r
f
or
m
a
nc
e
in
c
la
s
s
if
ic
a
ti
on
pr
obl
e
m
s
[
24]
,
w
he
r
e
th
e
c
la
s
s
if
ic
a
ti
on
r
e
s
ul
ts
c
a
n
be
lo
ng
to
two
or
m
o
r
e
c
la
s
s
e
s
.
E
va
lu
a
ti
on
us
in
g
th
e
c
onf
us
io
n
m
a
tr
ix
h
e
lp
s
in
id
e
nt
if
yi
ng
how
th
e
m
ode
l
a
c
c
ur
a
te
ly
pr
e
di
c
ts
di
a
be
te
s
c
ondi
ti
ons
(
pos
it
iv
e
a
nd
n
e
ga
ti
ve
)
[
9]
.
T
he
a
c
c
ur
a
c
y
va
lu
e
i
s
th
e
c
om
pa
r
is
on
b
e
twe
e
n
th
e
num
be
r
of
da
ta
th
a
t
ha
s
be
e
n
c
or
r
e
c
tl
y
id
e
nt
i
f
ie
d
a
nd
th
e
to
ta
l
da
ta
th
a
t
ha
s
be
e
n
te
s
te
d
[
29]
.
T
he
pr
e
c
is
io
n
va
lu
e
is
th
e
c
om
pa
r
is
on
be
tw
e
e
n
th
e
qua
nt
it
y
of
a
c
c
ur
a
te
f
or
e
c
a
s
ts
a
s
w
e
ll
a
s
th
e
ov
e
r
a
ll
num
be
r
of
f
or
e
c
a
s
ts
[
31]
.
T
he
r
e
c
a
ll
va
lu
e
is
a
c
om
pa
r
is
on
of
how
m
a
n
y
pr
e
di
c
ti
ons
a
r
e
c
or
r
e
c
t
c
om
pa
r
e
d
to
a
ll
th
e
a
c
tu
a
l
da
ta
[
32]
.
T
he
F
1
-
s
c
or
e
va
lu
e
c
om
bi
ne
s
pr
e
c
is
io
n
a
nd
r
e
c
a
ll
in
to
a
s
in
gl
e
m
e
tr
ic
,
w
hi
c
h
is
ve
r
y
us
e
f
ul
f
or
de
a
li
ng w
it
h i
m
ba
la
nc
e
d da
ta
[
33]
–
[
35]
.
=
+
(
+
+
+
)
(4
)
=
+
(5
)
=
+
(6
)
1
−
=
2
×
×
+
(7
)
4.
E
X
P
E
R
I
M
E
N
T
A
N
A
L
Y
S
I
S
M
ode
li
ng
us
in
g
th
e
X
G
B
oos
t
a
lg
or
i
th
m
ha
s
be
e
n
s
uc
c
e
s
s
f
ul
ly
a
ppl
ie
d
to
pr
e
di
c
t
th
e
r
is
k
of
di
a
be
te
s
ba
s
e
d
on
he
a
lt
h
hi
s
to
r
y.
T
hi
s
m
ode
l
w
a
s
c
hos
e
n
f
or
it
s
a
dva
nt
a
ge
s
in
ha
ndl
in
g
c
om
pl
e
x
da
ta
a
nd
i
m
ba
la
nc
e
d
da
ta
di
s
tr
ib
ut
io
ns
. T
he
r
e
s
ul
ts
of
t
he
m
ode
li
ng s
how
good pe
r
f
o
r
m
a
nc
e
.
O
ne
of
th
e
li
m
it
a
ti
ons
of
us
in
g
X
G
B
oos
t
in
th
i
s
s
tu
dy
i
s
it
s
r
e
la
ti
ve
ly
lo
w
in
te
r
pr
e
ta
bi
li
ty
,
w
hi
c
h
is
a
c
r
uc
ia
l
a
s
pe
c
t
in
m
e
di
c
a
l
a
ppl
ic
a
ti
ons
.
A
lt
hough
X
G
B
oos
t
ha
s
de
m
ons
tr
a
te
d
hi
gh
pr
e
di
c
ti
ve
a
c
c
ur
a
c
y,
it
s
e
ns
e
m
bl
e
-
ba
s
e
d
a
nd
c
om
pl
e
x
na
tu
r
e
m
a
ke
s
th
e
de
c
is
io
n
-
m
a
ki
ng
pr
oc
e
s
s
di
f
f
ic
ul
t
to
unde
r
s
ta
nd
di
r
e
c
tl
y
b
y
he
a
lt
hc
a
r
e
pr
of
e
s
s
io
na
ls
.
I
n
c
li
ni
c
a
l
pr
a
c
ti
c
e
,
phys
ic
ia
n
s
r
e
qui
r
e
not
onl
y
th
e
pr
e
di
c
ti
on
r
e
s
ul
ts
but
a
ls
o
c
le
a
r
e
xpl
a
na
ti
ons
of
th
e
c
ont
r
ib
ut
in
g
f
a
c
to
r
s
be
hi
nd
th
os
e
r
e
s
ul
ts
i
n
or
de
r
to
e
ns
ur
e
e
th
ic
a
l
us
e
a
nd
bui
ld
tr
us
t.
T
he
r
e
f
or
e
,
f
ut
ur
e
r
e
s
e
a
r
c
h
s
houl
d
c
ons
id
e
r
in
te
gr
a
ti
ng
X
A
I
a
ppr
oa
c
he
s
,
s
uc
h
a
s
S
ha
pl
e
y
a
ddi
ti
ve
e
xpl
a
na
ti
ons
(
S
H
A
P
)
or
lo
c
a
l
in
te
r
pr
e
ta
bl
e
m
ode
l
a
gnos
ti
c
e
xpl
a
na
ti
on
(
L
I
M
E
)
,
to
pr
ovi
de
in
s
ig
ht
s
in
to
th
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8814
I
nt
J
A
dv A
ppl
S
c
i
,
V
ol
.
14
, N
o.
4
,
D
e
c
e
m
be
r
20
25
:
1028
-
1039
1034
dom
in
a
nt
f
e
a
tu
r
e
s
in
f
lu
e
nc
in
g
e
a
c
h
pr
e
di
c
ti
on.
B
y
doi
ng
s
o,
th
e
m
ode
l
w
il
l
not
onl
y
be
a
c
c
ur
a
te
but
a
l
s
o
tr
a
ns
pa
r
e
nt
, t
r
us
twor
th
y, a
nd be
tt
e
r
a
li
gne
d w
it
h t
he
ne
e
ds
of
m
e
di
c
a
l
de
c
is
io
n
s
uppor
t.
4.1. P
r
e
p
r
oc
e
s
s
in
g
T
hi
s
s
e
c
ti
on
out
li
ne
s
th
e
pr
e
pr
oc
e
s
s
in
g
s
te
p
s
,
in
c
lu
di
ng
l
a
be
l
e
nc
odi
ng,
f
e
a
tu
r
e
-
ta
r
ge
t
s
e
pa
r
a
ti
on,
a
nd
da
ta
s
ta
nd
a
r
di
z
a
ti
on
to
pr
e
pa
r
e
th
e
da
ta
s
e
t
f
or
X
G
B
oos
t
m
ode
li
ng.
F
ig
ur
e
4
s
how
s
th
e
e
nc
odi
ng
pr
oc
e
s
s
c
onve
r
ts
a
ll
c
a
te
gor
ic
a
l
f
e
a
tu
r
e
s
in
to
bi
na
r
y
va
lu
e
s
(
0
a
nd
1)
.
F
ig
ur
e
5
s
how
s
th
a
t
th
e
p
r
e
di
c
to
r
va
r
ia
bl
e
s
(
x
)
c
ons
is
t
of
pa
ti
e
nt
s
ym
pt
om
s
a
nd
c
ha
r
a
c
te
r
is
ti
c
s
,
w
hi
le
th
e
ta
r
ge
t
va
r
ia
bl
e
(
y)
is
th
e
c
la
s
s
c
ol
um
n
r
e
pr
e
s
e
nt
in
g
th
e
di
a
gnos
is
l
a
be
l.
4.2.
D
e
t
e
r
m
in
e
h
yp
e
r
p
ar
a
m
e
t
e
r
s
T
hi
s
s
e
c
ti
on
out
li
ne
s
th
e
d
e
te
r
m
in
a
ti
on
of
hype
r
pa
r
a
m
e
te
r
s
f
o
r
X
G
B
oos
t
m
ode
li
ng.
F
ig
ur
e
6
s
how
th
a
t
th
e
be
s
t
X
G
B
oos
t
m
ode
l
us
e
s
e
v
a
l_
m
e
tr
ic
=
lo
gl
os
s
,
l
e
a
r
ni
ng_r
a
te
=
0.1,
m
a
x_de
pt
h=
4,
n_e
s
ti
m
a
to
r
s
=
100,
a
nd obje
c
ti
ve
=
bi
na
r
y:
lo
gi
s
ti
c
.
F
ig
ur
e
4
.
O
ut
put
a
f
te
r
e
nc
odi
ng da
ta
F
ig
ur
e
5
.
O
ut
put
of
s
e
pa
r
a
ti
ng f
e
a
tu
r
e
s
a
nd t
a
r
ge
ts
F
ig
ur
e
6
.
O
ut
put
de
f
in
e
s
t
he
hype
r
pa
r
a
m
e
te
r
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
dv A
ppl
S
c
i
I
S
S
N
:
2252
-
8814
I
m
pl
e
m
e
nt
at
io
n of
X
G
B
oos
t
fo
r
di
abe
te
s
m
e
ll
it
us
r
i
s
k
pr
e
di
c
ti
on bas
e
d on he
al
th
hi
s
to
r
y
(
A
ndi
R
ia
ns
y
ah)
1035
4.3.
R
e
s
u
lt
t
r
ai
n
an
d
t
e
s
t
T
hi
s
s
e
c
ti
on
di
s
c
us
s
e
s
th
e
m
ode
l'
s
pe
r
f
or
m
a
nc
e
,
il
lu
s
tr
a
ti
ng
th
e
tr
a
in
in
g
a
nd
te
s
ti
ng
pe
r
f
or
m
a
nc
e
in
pr
e
di
c
ti
ng
da
ta
.
F
ig
ur
e
7
s
ho
w
s
th
e
r
e
s
ul
t
of
s
a
vi
ng
a
m
ode
l
t
ha
t
ha
s
gon
e
th
r
ough
th
e
gr
id
s
e
a
r
c
h
s
ta
ge
,
a
m
ode
l
w
it
h
th
e
be
s
t
hype
r
pa
r
a
m
e
te
r
c
om
bi
na
ti
on
ba
s
e
d
on
pe
r
f
or
m
a
nc
e
ove
r
5
it
e
r
a
ti
ons
.
(
c
r
os
s
-
va
li
da
ti
on)
.
T
hi
s
m
ode
l
w
il
l
b
e
us
e
d
to
m
a
ke
pr
e
di
c
ti
ons
on
th
e
te
s
ti
ng
da
ta
.
T
he
m
e
tr
ic
r
e
s
ul
ts
ge
ne
r
a
te
d
dur
in
g
th
e
m
ode
l
tr
a
in
in
g
a
r
e
di
s
pl
a
ye
d
in
a
m
o
r
e
r
e
a
da
bl
e
f
or
m
a
t
a
n
d
di
s
pl
a
yi
ng
f
our
de
c
im
a
l
pl
a
c
e
s
.
P
r
ove
n
by
di
s
pl
a
yi
ng/
c
a
lc
ul
a
ti
ng
e
va
lu
a
ti
on
m
e
tr
ic
s
on
th
e
tr
a
in
in
g
da
ta
.
A
c
onf
us
io
n
m
a
tr
ix
is
us
e
d
to
s
e
e
how
m
a
ny
c
or
r
e
c
t
a
nd i
nc
or
r
e
c
t
pr
e
di
c
ti
ons
t
he
m
ode
l
m
a
de
on t
he
t
r
a
in
in
g da
ta
.
T
he
a
c
c
ur
a
c
y
pr
oduc
e
d
f
r
om
th
e
im
a
ge
a
bove
is
to
c
a
l
c
ul
a
te
th
e
m
ode
l'
s
a
c
c
ur
a
c
y
in
pr
e
di
c
ti
ng
c
or
r
e
c
tl
y,
w
it
h
a
n
a
c
c
ur
a
c
y
v
a
lu
e
of
0.9979.
T
he
r
a
ti
o
of
r
e
a
l
pos
it
iv
e
f
or
e
c
a
s
t
s
to
a
ll
pos
it
iv
e
pr
e
di
c
ti
on
s
is
known
a
s
pr
e
c
i
s
io
n,
r
e
s
ul
ti
ng
in
0.9960.
T
he
pe
r
c
e
nt
a
ge
of
r
e
a
l
pos
it
iv
e
in
s
ta
nc
e
s
th
a
t
th
e
m
ode
l
a
c
c
ur
a
t
e
ly
de
te
c
ts
is
kno
w
n
a
s
r
e
c
a
ll
,
yi
e
ld
in
g
a
r
e
s
ul
t
of
1.000.
T
he
F
1
-
s
c
or
e
is
th
e
h
a
r
m
oni
z
a
ti
on
be
twe
e
n
r
e
c
a
ll
a
nd
pr
e
c
is
io
n, r
e
s
ul
ti
ng i
n 0.9980.
F
ig
ur
e
8
s
how
s
th
a
t
m
ode
l
e
va
lu
a
ti
on
is
th
e
pr
oc
e
s
s
of
a
s
s
e
s
s
in
g
th
e
pe
r
f
or
m
a
nc
e
of
a
m
ode
l
th
a
t
ha
s
be
e
n
tr
a
in
e
d
a
nd
te
s
te
d
us
in
g
f
or
e
ig
n
da
ta
.
T
he
obj
e
c
ti
ve
is
to
e
v
a
lu
a
te
th
e
m
ode
l'
s
pe
r
f
or
m
a
nc
e
w
he
n
a
ppl
ie
d t
o unf
a
m
il
ia
r
da
ta
. T
he
f
ol
lo
w
in
g i
s
t
he
r
e
s
ul
t
of
e
a
c
h
s
c
o
r
e
i
n t
a
bl
e
f
or
m
.
E
a
c
h
s
c
or
e
di
s
pl
a
ye
d
in
T
a
bl
e
s
2
to
4
is
th
e
r
e
s
ul
t
of
th
e
m
ode
l'
s
pe
r
f
or
m
a
nc
e
e
va
lu
a
ti
on
a
ga
in
s
t
th
e
da
ta
us
e
d.
T
o
m
e
a
s
ur
e
th
e
m
ode
l
in
m
a
ki
ng
a
c
c
ur
a
te
pr
e
di
c
ti
ons
(
a
c
c
ur
a
c
y)
,
to
a
s
s
e
s
s
th
e
tr
ue
pos
it
iv
e
pr
e
di
c
ti
ons
c
om
pa
r
e
d
to
th
e
pos
it
iv
e
pr
e
di
c
ti
ons
m
a
de
by
th
e
m
ode
l
f
or
e
a
c
h
c
a
te
gor
y
(
pr
e
c
is
io
n)
,
th
e
n
to
e
va
lu
a
te
how
m
a
ny
a
c
tu
a
l
pos
it
iv
e
c
a
s
e
s
a
r
e
c
or
r
e
c
tl
y
de
te
c
te
d
by
th
e
m
ode
l
f
r
om
th
e
to
ta
l
pos
it
iv
e
c
a
s
e
s
(
r
e
c
a
ll
)
,
a
nd
to
c
a
lc
ul
a
te
th
e
pr
e
c
is
io
n
a
nd
r
e
c
a
ll
ha
r
m
oni
c
m
e
a
n,
w
hi
c
h
pr
ovi
de
s
a
n
ove
r
a
ll
pi
c
tu
r
e
of
th
e
ba
la
nc
e
be
tw
e
e
n
th
e
two
(
F
1
-
s
c
or
e
)
.
F
ig
ur
e
9
s
how
s
th
a
t
th
e
X
G
B
oos
t
m
ode
l
a
c
hi
e
ve
d
th
e
be
s
t
pe
r
f
or
m
a
nc
e
w
it
h a
n A
U
C
of
0.839 c
om
pa
r
e
d t
o
L
R
(
0.823)
, l
in
e
a
r
S
V
M
(
0.815)
, a
nd
RF
(
0.812)
.
F
ig
ur
e
7. M
ode
l
tr
a
in
F
ig
ur
e
8. M
ode
l
e
va
lu
a
ti
on
T
a
bl
e
2.
C
onf
us
io
n m
a
tr
ix
va
lu
e
s
T
r
ue
pos
i
t
i
ve
(
T
P
)
F
a
l
s
e
pos
i
t
i
ve
(
F
P
)
F
a
l
s
e
ne
ga
t
i
ve
(
F
N
)
T
r
ue
ne
ga
t
i
ve
(
T
N
)
33
0
2
69
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8814
I
nt
J
A
dv A
ppl
S
c
i
,
V
ol
.
14
, N
o.
4
,
D
e
c
e
m
be
r
20
25
:
1028
-
1039
1036
T
a
bl
e
3
.
M
od
e
l
pe
r
f
or
m
a
nc
e
r
e
s
ul
ts
D
a
t
a
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
T
r
a
i
ni
ng
99%
99%
100%
99%
T
e
s
t
i
ng
98%
100%
97%
98%
T
a
bl
e
4
.
C
a
te
gor
y of
e
a
c
h c
la
s
s
C
a
t
e
gor
y
P
r
e
c
i
s
i
on
R
e
c
a
l
l
F1
-
s
c
or
e
D
i
a
be
t
e
s
(
1)
94%
100%
97%
N
on
-
d
i
a
be
t
e
s
(
0)
100%
97%
99%
F
ig
ur
e
9
.
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
s
4.4.
We
b
s
it
e
a
p
p
e
ar
an
c
e
T
he
de
ve
lo
p
e
d
m
ode
l
ha
s
th
e
n
be
e
n
im
pl
e
m
e
nt
e
d
in
th
e
f
or
m
of
a
w
e
bs
it
e
to
r
e
a
c
h
m
a
ny
pe
opl
e
out
th
e
r
e
w
ho ne
e
d pr
e
di
c
ti
ons
of
D
M
r
is
k ba
s
e
d on the
ir
he
a
lt
h hi
s
to
r
y.
F
ig
ur
e
10
s
how
s
t
he
r
e
s
ul
ts
of
a
s
e
r
ie
s
of
pr
oc
e
s
s
e
s
th
a
t
o
c
c
ur
a
f
te
r
th
e
u
s
e
r
in
put
s
th
e
ir
he
a
lt
h
hi
s
to
r
y
d
a
ta
.
F
ir
s
t,
it
in
vol
ve
s
a
s
t
a
nda
r
di
z
a
ti
on
pr
oc
e
s
s
f
or
th
e
da
ta
th
a
t
ha
s
be
e
n
e
nt
e
r
e
d,
s
o
th
a
t
th
e
ne
w
ly
in
put
va
lu
e
s
a
r
e
on
a
di
ve
r
s
e
s
c
a
le
ba
s
e
d
on
th
e
pa
r
a
m
e
te
r
s
th
a
t
ha
v
e
be
e
n
pr
oc
e
s
s
e
d
by
th
e
pr
e
vi
ous
ly
tr
a
in
e
d m
ode
l,
a
nd
th
e
r
e
s
ul
ts
w
il
l
b
e
di
s
pl
a
y
e
d
on
th
e
m
a
in
pa
ge
.
T
he
n
th
is
m
od
e
l
w
il
l
ge
n
e
r
a
te
pr
e
di
c
ti
ons
f
or
di
a
be
te
s
r
is
k
a
nd
th
e
pr
oba
bi
li
ty
of
di
a
b
e
te
s
r
is
k
ba
s
e
d
on
th
e
in
put
da
ta
pr
ov
id
e
d.
T
he
m
ode
l
w
il
l
de
te
r
m
in
e
w
he
th
e
r
th
e
r
is
k
of
di
a
be
te
s
f
r
om
th
e
e
nt
e
r
e
d
da
ta
i
s
pos
it
iv
e
(
in
di
c
a
ti
ng dia
be
te
s
)
or
ne
ga
ti
ve
(
doe
s
not
i
ndi
c
a
te
di
a
be
te
s
).
T
he
pr
oba
bi
li
ty
va
lu
e
of
di
a
be
te
s
r
is
k
i
s
a
r
e
s
ul
t
w
it
h
a
v
a
lu
e
b
e
twe
e
n
0
a
nd
1
to
in
di
c
a
te
ho
w
hi
gh
th
e
m
ode
l'
s
c
onf
id
e
nc
e
le
ve
l
is
in
pr
e
di
c
ti
ng
w
he
th
e
r
s
om
e
one
ha
s
a
r
is
k
of
di
a
be
te
s
ba
s
e
d
on
th
e
gi
ve
n
da
ta
.
A
va
lu
e
of
0.5
(
50%
)
is
u
s
e
d
a
s
th
e
th
r
e
s
hol
d
d.
I
f
th
e
pr
ob
a
bi
li
ty
va
lu
e
is
a
bove
0.5
(
a
ppr
oa
c
hi
ng
1)
,
it
m
e
a
ns
th
e
m
ode
l
is
ve
r
y
c
onf
id
e
nt
th
a
t
th
e
r
e
i
s
a
r
is
k
of
di
a
be
te
s
,
a
nd
th
e
r
e
s
ul
t
w
il
l
be
c
la
s
s
if
ie
d
in
th
e
pos
it
iv
e
c
a
te
gor
y
(
di
a
be
t
e
s
)
.
C
o
nv
e
r
s
e
ly
,
if
th
e
pr
o
ba
bi
l
it
y
v
a
l
ue
i
s
u
nd
e
r
0.
5
(
a
ppr
o
a
c
hi
ng
0
)
,
it
m
e
a
n
s
th
e
m
od
e
l
i
s
v
e
r
y
c
onf
id
e
nt
t
ha
t
t
he
r
e
is
n
o r
i
s
k o
f
d
ia
be
te
s
,
a
nd
t
h
e
r
e
s
ul
t
w
il
l
b
e
c
l
a
s
s
if
i
e
d
in
t
h
e
ne
ga
ti
v
e
c
a
t
e
gor
y.
F
ig
ur
e
10. M
a
in
di
s
pl
a
y
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
dv A
ppl
S
c
i
I
S
S
N
:
2252
-
8814
I
m
pl
e
m
e
nt
at
io
n of
X
G
B
oos
t
fo
r
di
abe
te
s
m
e
ll
it
us
r
i
s
k
pr
e
di
c
ti
on bas
e
d on he
al
th
hi
s
to
r
y
(
A
ndi
R
ia
ns
y
ah)
1037
5.
C
O
N
C
L
U
S
I
O
N
T
hi
s
r
e
s
e
a
r
c
h
de
m
ons
tr
a
te
s
th
a
t
th
e
XG
boos
t
a
lg
o
r
it
hm
s
how
s
opt
im
a
l
a
nd
a
c
c
ur
a
te
pe
r
f
or
m
a
nc
e
i
n
c
la
s
s
if
yi
ng
th
e
r
is
k
of
D
M
ba
s
e
d
on
he
a
lt
h
hi
s
to
r
y
to
e
nha
n
c
e
e
a
r
ly
di
a
gno
s
is
,
a
c
hi
e
vi
ng
a
n
a
c
c
ur
a
c
y
of
0.9808,
a
pr
e
c
i
s
io
n
of
1.000,
a
r
e
c
a
ll
of
0.9718,
a
nd a
n
F
1
-
s
c
or
e
of
0.9857. W
it
h
a
hi
gh
le
ve
l
of
a
c
c
ur
a
c
y,
th
e
m
ode
l
s
uc
c
e
s
s
f
ul
ly
pr
e
di
c
ts
w
he
th
e
r
a
n i
ndi
vi
dua
l
is
a
t
r
is
k of
h
a
vi
ng dia
be
te
s
ba
s
e
d on the
ir
m
e
di
c
a
l
hi
s
to
r
y,
in
or
de
r
t
o i
m
pr
ove
e
a
r
ly
di
a
gnos
is
.
A
C
K
N
O
WL
E
D
G
M
E
N
T
S
T
he
a
ut
hor
s
gr
a
te
f
ul
ly
a
c
knowle
dge
th
e
f
in
a
nc
ia
l
s
uppor
t
pr
ovi
de
d
by
th
e
U
ni
ve
r
s
it
a
s
I
s
la
m
S
ul
ta
n
A
gung,
S
e
m
a
r
a
ng,
I
ndone
s
ia
in
c
ondu
c
ti
ng
a
nd
c
om
pl
e
ti
ng
th
is
s
tu
dy.
W
e
w
oul
d
li
ke
to
e
xpr
e
s
s
our
a
ppr
e
c
ia
ti
on
to
th
e
ps
yc
hol
ogi
s
ts
w
ho
v
e
r
if
ie
d
th
e
da
ta
s
e
t
w
e
c
ol
le
c
te
d;
w
it
hout
th
e
ir
c
ont
r
ib
ut
io
ns
,
th
is
r
e
s
e
a
r
c
h
w
oul
d
not
ha
ve
be
e
n
pos
s
ib
le
.
I
n
a
ddi
ti
on,
w
e
a
r
e
ve
r
y
gr
a
te
f
ul
to
our
a
c
a
de
m
ic
a
dvi
s
or
s
f
or
th
e
ir
s
uppor
t
a
nd
va
lu
a
bl
e
f
e
e
dba
c
k
th
r
oughout
th
is
r
e
s
e
a
r
c
h
pr
oc
e
s
s
.
T
hi
s
r
e
s
e
a
r
c
h
is
de
di
c
a
te
d
to
a
ll
pa
r
ti
e
s
s
tr
iv
in
g t
o i
m
pr
ove
t
he
qua
li
ty
of
m
e
nt
a
l
he
a
lt
h c
a
r
e
f
or
a
dol
e
s
c
e
nt
s
i
n I
ndone
s
ia
.
F
U
N
D
I
N
G
I
N
F
O
R
M
A
T
I
O
N
T
hi
s
r
e
s
e
a
r
c
h w
a
s
f
unde
d by Univ
e
r
s
it
a
s
I
s
la
m
S
ul
ta
n A
gung, S
e
m
a
r
a
ng, I
ndone
s
ia
.
A
U
T
H
O
R
C
O
N
T
R
I
B
U
T
I
O
N
S
S
T
A
T
E
M
E
N
T
T
hi
s
jo
ur
na
l
us
e
s
th
e
C
ont
r
ib
ut
or
R
ol
e
s
T
a
xonomy
(
C
R
e
di
T
)
to
r
e
c
ogni
z
e
in
di
vi
dua
l
a
ut
hor
c
ont
r
ib
ut
io
ns
, r
e
duc
e
a
ut
hor
s
hi
p di
s
put
e
s
,
a
nd f
a
c
il
it
a
te
c
ol
la
bo
r
a
ti
on.
N
am
e
o
f
A
u
t
h
or
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
A
ndi
R
ia
ns
ya
h
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
G
huf
r
on
✓
✓
✓
✓
✓
✓
✓
✓
L
a
il
a
tu
l
F
it
r
iy
a
h
✓
✓
✓
✓
✓
✓
✓
✓
✓
S
uya
nt
o
✓
✓
✓
✓
C
:
C
onc
e
pt
ua
l
i
z
a
t
i
on
M
:
M
e
t
hodol
ogy
So
:
So
f
t
w
a
r
e
Va
:
Va
l
i
da
t
i
on
Fo
:
Fo
r
m
a
l
a
na
l
ys
i
s
I
:
I
nve
s
t
i
ga
t
i
on
R
:
R
e
s
our
c
e
s
D
:
D
a
t
a
C
ur
a
t
i
on
O
:
W
r
i
t
i
ng
-
O
r
i
gi
na
l
D
r
a
f
t
E
:
W
r
i
t
i
ng
-
R
e
vi
e
w
&
E
di
t
i
ng
Vi
:
Vi
s
ua
l
i
z
a
t
i
on
Su
:
Su
pe
r
vi
s
i
on
P
:
P
r
oj
e
c
t
a
dm
i
ni
s
t
r
a
t
i
on
Fu
:
Fu
ndi
ng a
c
qui
s
i
t
i
on
C
O
N
F
L
I
C
T
O
F
I
N
T
E
R
E
S
T
S
T
A
T
E
M
E
N
T
T
he
a
ut
ho
r
s
de
c
la
r
e
t
ha
t
th
e
y
ha
ve
n
o
kn
ow
n
c
on
f
l
ic
ts
o
f
i
nt
e
r
e
s
t,
w
he
th
e
r
f
in
a
nc
ia
l
or
non
-
f
i
na
nc
ia
l
,
th
a
t
c
ou
ld
ha
ve
in
f
l
ue
nc
e
d
th
e
c
o
nd
uc
t
o
r
ou
tc
om
e
s
o
f
th
is
r
e
s
e
a
r
c
h
.
T
h
e
s
t
ud
y
w
a
s
c
a
r
r
ie
d
out
i
nde
pe
n
de
n
tl
y
,
a
nd
a
l
l
a
na
l
ys
e
s
,
i
nt
e
r
p
r
e
ta
t
io
ns
,
a
nd
c
o
nc
lu
s
i
ons
a
r
e
ba
s
e
d
s
ol
e
ly
on
t
he
da
ta
o
bt
a
in
e
d
a
nd
th
e
m
e
t
ho
do
lo
g
ie
s
a
pp
l
ie
d
by
th
e
a
u
th
or
s
.
T
he
a
ut
ho
r
s
a
l
one
a
r
e
r
e
s
pons
ib
le
f
or
th
e
c
o
nt
e
nt
a
n
d
w
r
it
in
g
o
f
th
is
p
a
pe
r
.
D
A
T
A
A
V
A
I
L
A
B
I
L
I
T
Y
T
he
da
ta
s
e
t
us
e
d
in
th
is
s
tu
dy
to
de
ve
lo
p
a
nd
e
va
lu
a
te
th
e
X
G
B
oos
t
-
ba
s
e
d
di
a
be
te
s
r
is
k
pr
e
di
c
ti
on
m
ode
l
c
ons
is
t
s
of
a
nonymi
z
e
d
p
a
ti
e
nt
he
a
lt
h
r
e
c
or
ds
de
r
iv
e
d
f
r
om
hi
s
to
r
ic
a
l
he
a
lt
h
da
ta
.
A
ll
p
e
r
s
ona
l
id
e
nt
if
ie
r
s
w
e
r
e
r
e
m
ove
d
pr
io
r
to
a
na
ly
s
i
s
to
e
n
s
ur
e
c
om
pl
ia
n
c
e
w
it
h
e
th
ic
a
l
a
nd
pr
iv
a
c
y
s
ta
nda
r
ds
.
D
ue
to
in
s
ti
tu
ti
ona
l
a
nd
pr
iv
a
c
y
r
e
s
tr
ic
ti
ons
,
th
e
r
a
w
da
ta
a
r
e
not
publ
ic
ly
a
va
il
a
bl
e
,
but
th
e
y
m
a
y
be
obt
a
in
e
d
f
r
o
m
th
e
c
or
r
e
s
ponding a
ut
hor
[
A
R
]
upon r
e
a
s
ona
bl
e
r
e
que
s
t
a
nd w
it
h pe
r
m
is
s
io
n f
r
om
t
he
r
e
le
va
nt
da
ta
ow
ne
r
.
R
E
F
E
R
E
N
C
E
S
[
1]
Y
.
L
i
e
t
al
.
,
“
A
s
s
o
c
i
a
t
i
on
of
a
ge
a
t
di
a
gno
s
i
s
of
t
ype
2
di
a
b
e
t
e
s
m
e
l
l
i
t
us
w
i
t
h
t
he
r
i
s
ks
of
t
he
m
or
bi
di
t
y
of
c
a
r
di
ova
s
c
ul
a
r
di
s
e
a
s
e
,
c
a
nc
e
r
a
nd
a
l
l
-
c
a
u
s
e
m
or
t
a
l
i
t
y:
e
vi
de
nc
e
f
r
om
a
r
e
a
l
-
w
or
l
d
s
t
udy
w
i
t
h
a
l
a
r
ge
popul
a
t
i
on
-
ba
s
e
d
c
ohor
t
s
t
udy,”
D
i
abe
t
e
s
R
e
s
e
ar
c
h
and C
l
i
ni
c
al
P
r
ac
t
i
c
e
, vol
. 217, N
ov. 2024, doi
:
10.1016/
j
.di
a
br
e
s
.2024.111870.
[
2]
J
.
H
.
W
e
r
s
ä
l
l
,
P
.
A
dol
f
s
s
on,
J
.
E
ke
l
und,
G
.
F
or
s
a
nde
r
,
K
.
Å
ke
s
s
on,
a
nd
R
.
H
a
na
s
,
“
I
s
di
a
be
t
i
c
ke
t
oa
c
i
dos
i
s
a
t
di
a
gnos
i
s
of
t
y
pe
1
di
a
be
t
e
s
a
s
s
oc
i
a
t
e
d
w
i
t
h
s
e
c
onda
r
y
e
pi
s
ode
s
of
ke
t
oa
c
i
dos
i
s
?
a
na
t
i
onw
i
de
l
ongi
t
udi
na
l
s
t
udy
of
S
w
e
di
s
h
c
hi
l
dr
e
n
f
r
om
2012
t
o
2019,”
D
i
abe
t
e
s
R
e
s
e
a
r
c
h and C
l
i
ni
c
al
P
r
ac
t
i
c
e
, vol
. 225, J
ul
. 2025, doi
:
10.1016/
j
.di
a
br
e
s
.2025.112282.
Evaluation Warning : The document was created with Spire.PDF for Python.