I
nte
rna
t
io
na
l J
o
urna
l o
f
Adv
a
nces in Applie
d Science
s
(
I
J
AAS)
Vo
l.
1
5
,
No
.
2
,
J
u
n
e
20
2
6
,
p
p
.
790
~
8
0
3
I
SS
N:
2252
-
8
8
1
4
,
DOI
:
1
0
.
1
1
5
9
1
/ijaas
.
v
1
5
.
i
2
.
pp
790
-
8
0
3
790
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
a
s
.
ia
esco
r
e.
co
m
Using
ma
chine le
a
rning
t
o
unders
t
a
nd t
he
ro
o
t
ca
us
es o
f
type
2
dia
betes in
Saud
i
Ara
bia
M
o
ha
m
m
a
d Sa
ee
d Al
G
ha
md
i
1
,
Ala
a
K
ha
did
o
s
1,
2
,
Adil K
ha
did
o
s
3
1
D
e
p
a
r
t
me
n
t
o
f
I
n
f
o
r
mat
i
o
n
S
y
s
t
e
ms
,
F
a
c
u
l
t
y
o
f
C
o
m
p
u
t
i
n
g
a
n
d
I
n
f
o
r
m
a
t
i
o
n
T
e
c
h
n
o
l
o
g
y
,
K
i
n
g
A
b
d
u
l
a
z
i
z
U
n
i
v
e
r
si
t
y
,
J
e
d
d
a
h
,
S
a
u
d
i
A
r
a
b
i
a
2
C
e
n
t
e
r
o
f
R
e
sea
r
c
h
E
x
c
e
l
l
e
n
c
e
i
n
A
r
t
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
a
n
d
D
a
t
a
S
c
i
e
n
c
e
,
K
i
n
g
A
b
d
u
l
a
z
i
z
U
n
i
v
e
r
s
i
t
y
,
Je
d
d
a
h
,
S
a
u
d
i
A
r
a
b
i
a
3
D
e
p
a
r
t
me
n
t
o
f
I
n
f
o
r
mat
i
o
n
Te
c
h
n
o
l
o
g
y
,
F
a
c
u
l
t
y
o
f
C
o
m
p
u
t
i
n
g
a
n
d
I
n
f
o
r
m
a
t
i
o
n
T
e
c
h
n
o
l
o
g
y
,
K
i
n
g
A
b
d
u
l
a
z
i
z
U
n
i
v
e
r
s
i
t
y
,
Jed
d
a
h
,
S
a
u
d
i
A
r
a
b
i
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
u
l 7
,
2
0
2
5
R
ev
is
ed
Ma
y
3
,
2
0
2
6
Acc
ep
ted
Ma
y
2
3
,
2
0
2
6
Th
is
st
u
d
y
e
x
a
m
in
e
s
th
e
e
le
v
a
ted
p
re
v
a
le
n
c
e
o
f
ty
p
e
2
d
iab
e
te
s
m
e
ll
it
u
s
(T2
DM)
i
n
S
a
u
d
i
Ara
b
ia
b
y
in
t
e
g
ra
ti
n
g
a
larg
e
-
sc
a
le,
re
g
io
n
a
ll
y
sp
e
c
ifi
c
d
a
tas
e
t
fro
m
t
h
e
S
a
u
d
i
M
in
istry
o
f
He
a
lt
h
(
1
0
0
,
0
0
0
p
a
ti
e
n
t
re
c
o
rd
s).
Th
e
p
rima
ry
c
o
n
tri
b
u
ti
o
n
is
th
is
S
a
u
d
i
-
fo
c
u
se
d
d
a
ta
in
teg
ra
ti
o
n
—
a
d
d
re
ss
in
g
a
c
rit
ica
l
g
a
p
i
n
p
rio
r
stu
d
ies
th
a
t
re
ly
o
n
g
l
o
b
a
l
d
a
tas
e
ts
(e
.
g
.
,
P
IM
A,
UK
Bio
b
a
n
k
)
—
a
n
d
a
c
o
m
p
a
ra
ti
v
e
e
v
a
lu
a
ti
o
n
o
f
trad
it
io
n
a
l
m
a
c
h
in
e
lea
rn
in
g
(
ML
)
m
o
d
e
ls
(
ra
n
d
o
m
fo
re
st
(RF
)
,
a
d
a
p
ti
v
e
b
o
o
sti
n
g
(
Ad
a
Bo
o
st
)
)
a
n
d
fe
w
-
sh
o
t
larg
e
lan
g
u
a
g
e
m
o
d
e
ls
(LL
M
s)
fo
r
e
a
rly
risk
d
e
tec
ti
o
n
.
Ke
y
fe
a
tu
re
s
in
c
lu
d
e
a
g
e
,
b
o
d
y
m
a
ss
in
d
e
x
(
B
M
I
)
,
Hb
A1
c
,
b
lo
o
d
g
lu
c
o
se
,
h
y
p
e
rten
sio
n
,
a
n
d
sm
o
k
in
g
h
ist
o
ry
.
Clas
s
imb
a
lan
c
e
(9
%
d
iab
e
ti
c
)
wa
s
re
so
lv
e
d
u
si
n
g
sy
n
th
e
ti
c
m
i
n
o
rit
y
o
v
e
r
-
sa
m
p
li
n
g
tec
h
n
i
q
u
e
wit
h
e
d
it
e
d
n
e
a
re
st
n
e
ig
h
b
o
r
s
(
S
M
OTEE
NN
)
.
E
n
se
m
b
le
M
L
m
o
d
e
ls
a
c
h
iev
e
d
u
p
t
o
9
7
%
a
c
c
u
ra
c
y
,
wh
il
e
fe
w
-
sh
o
t
LL
M
re
a
c
h
e
d
9
8
%
a
c
ro
ss
a
c
c
u
ra
c
y
,
p
re
c
isio
n
,
re
c
a
ll
,
a
n
d
F
1
-
sc
o
r
e
(
a
re
a
u
n
d
e
r
t
h
e
c
u
r
v
e
(
AUC
)
0
.
9
9
).
F
e
a
tu
re
imp
o
r
tan
c
e
c
o
n
fi
rm
e
d
th
a
t
Hb
A1
c
a
n
d
g
l
u
c
o
se
we
re
th
e
t
o
p
p
re
d
ict
o
rs.
T
h
is
re
g
i
o
n
a
ll
y
tail
o
r
e
d
,
h
i
g
h
-
p
e
rfo
rm
in
g
fra
m
e
wo
rk
e
n
a
b
les
e
a
rly
d
e
tec
ti
o
n
o
f
T
2
DM
a
n
d
c
u
lt
u
ra
ll
y
se
n
siti
v
e
in
terv
e
n
ti
o
n
s
in
h
ig
h
-
p
re
v
a
len
c
e
se
tt
in
g
s s
u
c
h
a
s S
a
u
d
i
Ar
a
b
ia.
K
ey
w
o
r
d
s
:
E
n
s
em
b
le
lear
n
in
g
Featu
r
e
im
p
o
r
tan
ce
Hb
A1
c
L
ar
g
e
lan
g
u
ag
e
m
o
d
els
Ma
ch
in
e
lear
n
in
g
SMOT
E
E
NN
T
y
p
e
2
d
iab
etes m
ellitu
s
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
Alaa
Kh
ad
id
o
s
C
en
ter
o
f
R
esear
ch
E
x
ce
llen
c
e
in
Ar
tific
ial
I
n
tellig
en
ce
an
d
Data
Scien
ce
,
Kin
g
Ab
d
u
lazi
z
Un
iv
er
s
ity
J
ed
d
ah
,
Sau
d
i A
r
a
b
ia
E
m
ail:
ao
k
h
ad
id
o
s
@
k
au
.
ed
u
.
s
a
1.
I
NT
RO
D
UCT
I
O
N
I
n
Sau
d
i
Ar
ab
ia,
m
an
y
p
eo
p
le
h
av
e
ty
p
e
2
d
ia
b
etes
m
ellitu
s
(
T
2
DM
)
,
an
d
th
is
n
u
m
b
er
is
g
ettin
g
h
ig
h
er
e
v
er
y
y
ea
r
.
So
m
e
r
ep
o
r
ts
s
aid
th
at
m
o
r
e
th
an
1
8
.
7
%
o
f
ad
u
lts
in
Sau
d
i
Ar
a
b
ia
h
av
e
i
t
[
1
]
.
T
h
is
n
u
m
b
e
r
is
o
n
e
o
f
th
e
h
ig
h
est
in
th
e
wo
r
ld
,
an
d
it
ca
u
s
es
a
l
o
t
o
f
p
r
o
b
lem
s
,
s
u
ch
as
f
u
ll
h
o
s
p
itals
,
tire
d
m
ed
ical
team
s
,
an
d
m
o
r
e
d
ea
th
s
.
T
h
e
co
u
n
tr
y
allo
ca
tes
s
u
b
s
tan
tial
f
u
n
d
s
t
o
th
is
p
r
o
b
lem
.
Ma
n
y
r
ese
ar
c
h
er
s
ar
e
n
o
w
u
s
in
g
ar
tific
ial
in
tellig
en
ce
(
AI
)
an
d
m
ac
h
in
e
lear
n
in
g
(
ML
)
to
ad
d
r
ess
th
is
.
T
h
ese
n
ew
ap
p
r
o
ac
h
es
ca
n
h
elp
id
en
tify
wh
o
is
at
r
is
k
ea
r
ly
a
n
d
en
a
b
le
p
h
y
s
ician
s
to
tr
ea
t
th
em
m
o
r
e
ef
f
ec
tiv
el
y
[
2
]
.
H
o
wev
er
,
m
o
s
t
p
r
io
r
r
esear
ch
u
s
ed
d
ata
f
r
o
m
o
th
er
co
u
n
tr
ies.
No
t m
a
n
y
s
tu
d
ies u
s
e
r
ea
l d
ata
f
r
o
m
Sau
d
i p
e
o
p
le
.
T
h
is
p
ap
er
u
s
es
d
ata
f
r
o
m
1
0
0
,
0
0
0
p
eo
p
le
f
r
o
m
th
e
Sau
d
i
Min
is
tr
y
o
f
Hea
lth
[
3
]
.
T
h
i
s
is
m
o
r
e
ac
cu
r
ate
f
o
r
Sau
d
i
ca
s
es.
Ad
d
itio
n
ally
,
t
h
e
im
b
ala
n
ce
in
th
e
d
ata
was
ad
d
r
ess
ed
.
B
e
ca
u
s
e
o
n
ly
9
%
o
f
p
ar
ticip
an
ts
h
ad
d
iab
etes,
th
e
s
y
n
th
etic
m
in
o
r
ity
o
v
e
r
-
s
a
m
p
lin
g
tech
n
iq
u
e
with
ed
ite
d
n
ea
r
est
n
ei
g
h
b
o
r
s
(
SMOT
E
E
NN)
was
u
s
ed
to
ad
d
r
ess
th
is
im
b
alan
ce
[
4
]
.
Ma
n
y
m
o
d
els
wer
e
u
s
ed
.
So
m
e
o
f
th
em
ar
e
estab
lis
h
ed
,
s
u
ch
as
r
an
d
o
m
f
o
r
est
(
R
F)
an
d
ad
ap
tiv
e
b
o
o
s
tin
g
(
Ad
aBo
o
s
t)
.
On
e
o
f
th
em
g
av
e
9
7
%
ac
cu
r
ac
y
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ad
v
Ap
p
l Sci
I
SS
N:
2252
-
8
8
1
4
Usi
n
g
ma
ch
in
e
lea
r
n
in
g
to
u
n
d
ers
ta
n
d
th
e
r
o
o
t c
a
u
s
es o
f ty
p
e
2
d
i
a
b
etes …
(
Mo
h
a
mma
d
S
a
ee
d
A
l G
h
a
md
i)
791
Ad
d
itio
n
ally
,
a
n
ew
ty
p
e
o
f
m
o
d
el,
a
lar
g
e
lan
g
u
ag
e
m
o
d
el
(
L
L
M)
,
th
at
lea
r
n
s
f
r
o
m
a
s
m
all
n
u
m
b
er
o
f
ex
am
p
les,
was u
s
ed
.
Desp
ite
th
is
,
it a
ch
iev
ed
g
o
o
d
r
esu
lts
,
in
clu
d
in
g
9
8
% a
cc
u
r
ac
y
[
5
]
,
[
6
]
.
T
h
is
p
ap
er
ad
d
r
ess
es
th
r
ee
th
i
n
g
s
:
it
u
s
es
a
lar
g
e
Sa
u
d
i
d
ataset,
it
ad
d
r
ess
es
class
im
b
alan
ce
,
an
d
it
co
m
p
ar
es
s
tan
d
ar
d
ML
m
o
d
el
s
with
L
L
M
m
o
d
els.
T
h
ese
th
in
g
s
ca
n
h
elp
Sau
d
i
h
ea
lth
ca
r
e
⎯
an
d
m
ay
b
e
o
th
e
r
co
u
n
tr
ies
as we
ll
.
T
h
is
s
tu
d
y
m
ak
es th
e
f
o
llo
win
g
co
n
tr
i
b
u
t
io
n
s
:
i)
Used
a
lar
g
e,
n
atio
n
ally
r
ep
r
e
s
en
tativ
e
S
au
d
i
d
ataset
o
f
1
0
0
,
0
0
0
r
ec
o
r
d
s
f
r
o
m
th
e
Min
is
tr
y
o
f
Hea
lth
,
ad
d
r
ess
in
g
th
e
lack
o
f
r
e
g
io
n
-
s
p
ec
if
ic
an
d
lar
g
e
-
s
ca
le
co
h
o
r
ts
in
p
r
io
r
T
2
DM
p
r
ed
ictio
n
s
tu
d
ies.
ii)
Ap
p
lied
SMOT
E
E
NN
to
co
r
r
ec
t
s
ev
er
e
class
im
b
alan
ce
an
d
in
teg
r
ated
a
clin
ical
-
g
r
a
d
e
ev
alu
atio
n
p
ip
elin
e
,
in
clu
d
i
n
g
r
ec
eiv
e
r
o
p
er
atin
g
c
h
ar
ac
ter
is
tic
ar
ea
u
n
d
er
th
e
c
u
r
v
e
(
R
OC
-
AUC)
,
B
r
ier
s
co
r
e,
co
n
f
id
en
ce
in
ter
v
als
(
C
I
)
,
a
n
d
ca
lib
r
atio
n
an
aly
s
is
to
en
s
u
r
e
r
o
b
u
s
t a
n
d
r
eliab
le
p
er
f
o
r
m
a
n
c
e.
iii)
B
en
ch
m
ar
k
ed
h
ig
h
-
p
er
f
o
r
m
in
g
en
s
em
b
le
m
o
d
els
(
Ad
aBo
o
st
an
d
RF
)
ag
ain
s
t
a
f
ew
-
s
h
o
t
LLM
,
o
f
f
er
i
n
g
th
e
f
ir
s
t d
ir
ec
t c
o
m
p
ar
is
o
n
b
et
wee
n
th
ese
ap
p
r
o
ac
h
es f
o
r
T
2
DM
p
r
ed
ictio
n
in
th
e
Sau
d
i
p
o
p
u
latio
n
.
iv
)
E
n
s
u
r
ed
clin
ical
in
ter
p
r
etab
ili
ty
b
y
a
n
aly
zin
g
lo
g
is
tic
-
r
eg
r
e
s
s
io
n
co
ef
f
icien
ts
an
d
m
o
d
el
-
d
r
iv
en
f
ea
tu
r
e
im
p
o
r
tan
c
e,
co
n
f
i
r
m
in
g
t
h
at
Hb
A1
c,
g
lu
c
o
s
e,
b
o
d
y
m
ass
in
d
ex
(
B
MI
)
,
ag
e,
a
n
d
h
y
p
er
t
en
s
io
n
ar
e
th
e
d
o
m
in
an
t
p
r
ed
icto
r
s
.
T
2
DM
is
a
m
ajo
r
g
lo
b
al
p
u
b
lic
h
ea
lth
ch
allen
g
e
with
r
ap
id
ly
in
cr
ea
s
in
g
p
r
e
v
alen
c
e
in
m
an
y
co
u
n
tr
ies.
I
n
Sau
d
i
Ar
ab
ia,
m
o
r
e
th
an
1
8
.
7
%
o
f
ad
u
lts
ar
e
esti
m
ated
to
b
e
af
f
ec
ted
,
p
lacin
g
th
e
co
u
n
tr
y
am
o
n
g
t
h
o
s
e
with
th
e
h
ig
h
e
s
t
d
iab
etes
p
r
ev
alen
ce
wo
r
l
d
wid
e
[
1
]
.
T
h
is
r
is
in
g
b
u
r
d
en
p
lace
s
s
ig
n
if
ican
t
p
r
ess
u
r
e
o
n
h
ea
lth
ca
r
e
s
y
s
tem
s
an
d
h
ig
h
lig
h
ts
th
e
n
ee
d
f
o
r
e
f
f
ec
tiv
e
ea
r
ly
d
etec
tio
n
a
n
d
p
r
ev
en
tio
n
s
tr
ateg
ies.
AI
an
d
ML
h
av
e
r
ec
en
tly
e
m
er
g
ed
as
p
o
we
r
f
u
l
to
o
ls
f
o
r
p
r
ed
ictin
g
c
h
r
o
n
ic
d
is
ea
s
es,
in
clu
d
in
g
d
iab
etes.
T
h
ese
ap
p
r
o
ac
h
es
ca
n
an
aly
z
e
lar
g
e
clin
ical
d
atasets
to
id
en
tify
co
m
p
lex
p
atter
n
s
ass
o
ciate
d
with
d
is
ea
s
e
r
is
k
[
2
]
.
Su
ch
p
r
ed
ictiv
e
s
y
s
tem
s
ca
n
ass
i
s
t
h
ea
lth
ca
r
e
p
r
o
f
ess
io
n
als
in
id
en
tify
in
g
h
ig
h
-
r
i
s
k
in
d
iv
id
u
als
an
d
im
p
lem
en
tin
g
p
r
ev
e
n
tiv
e
in
te
r
v
en
tio
n
s
ea
r
lier
.
Ho
wev
e
r
,
m
an
y
e
x
is
tin
g
d
iab
etes
p
r
ed
i
ctio
n
s
tu
d
ies
r
ely
o
n
in
ter
n
atio
n
al
d
atasets
r
ath
er
th
an
r
eg
io
n
-
s
p
ec
if
ic
clin
ical
d
ata.
Mo
d
els
d
ev
elo
p
ed
u
s
in
g
f
o
r
eig
n
d
atasets
m
ay
n
o
t
f
u
lly
r
e
f
lect
th
e
d
em
o
g
r
a
p
h
ic
a
n
d
life
s
ty
le
ch
ar
ac
ter
is
tics
o
f
th
e
Sau
d
i
p
o
p
u
latio
n
.
T
h
er
ef
o
r
e
,
lo
ca
lly
r
elev
an
t d
atasets
ar
e
ess
en
tial f
o
r
d
e
v
elo
p
in
g
ac
cu
r
ate
a
n
d
c
o
n
tex
t
-
s
p
ec
if
ic
p
r
ed
ictiv
e
m
o
d
els.
T
h
is
s
tu
d
y
ad
d
r
ess
es
th
is
lim
itatio
n
b
y
an
al
y
zin
g
a
la
r
g
e
d
ataset
o
f
1
0
0
,
0
0
0
an
o
n
y
m
iz
ed
p
atien
t
r
ec
o
r
d
s
o
b
tain
ed
f
r
o
m
th
e
Sau
d
i
Min
is
tr
y
o
f
Hea
lth
[
3
]
.
U
s
in
g
a
lar
g
e
n
atio
n
al
d
ataset
i
m
p
r
o
v
es
m
o
d
ellin
g
r
eliab
ilit
y
an
d
en
h
a
n
ce
s
th
e
clin
ical
r
elev
an
ce
o
f
th
e
r
esu
lts
.
I
n
ad
d
itio
n
,
a
d
v
an
ce
d
p
r
e
p
r
o
ce
s
s
in
g
tech
n
iq
u
es
ar
e
ap
p
lied
to
en
s
u
r
e
d
ata
q
u
ality
an
d
m
o
d
ellin
g
s
tab
ilit
y
.
An
o
th
er
ch
allen
g
e
in
m
e
d
ic
al
d
atasets
i
s
cla
s
s
im
b
alan
ce
,
wh
er
e
n
o
n
-
d
iab
eti
c
ca
s
es
g
r
ea
tly
o
u
tn
u
m
b
er
d
i
ab
etic
ca
s
es.
Su
ch
im
b
alan
ce
ca
n
b
ias
p
r
e
d
ictiv
e
m
o
d
els
to
war
d
th
e
m
ajo
r
ity
class
an
d
r
ed
u
ce
d
etec
tio
n
o
f
tr
u
e
d
iab
etic
ca
s
es.
T
o
m
itig
ate
th
is
is
s
u
e,
th
e
SMOT
E
E
NN
h
y
b
r
id
r
esam
p
lin
g
m
eth
o
d
is
ap
p
lied
to
b
alan
ce
th
e
d
ataset
an
d
im
p
r
o
v
e
class
if
icatio
n
p
er
f
o
r
m
an
ce
[
4
]
.
Var
io
u
s
ML
alg
o
r
ith
m
s
h
a
v
e
b
ee
n
ex
p
lo
r
e
d
f
o
r
d
iab
etes
p
r
e
d
ictio
n
.
T
r
ad
itio
n
al
m
o
d
els
s
u
c
h
as
lo
g
is
tic
r
eg
r
ess
io
n
(
L
R
)
,
d
ec
is
io
n
tr
ee
s
(
DT
s
)
,
an
d
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
es
(
SVMs)
h
av
e
b
ee
n
wid
ely
u
s
ed
to
an
aly
ze
clin
ical
r
is
k
f
ac
to
r
s
.
Ho
wev
er
,
en
s
em
b
le
l
ea
r
n
in
g
m
eth
o
d
s
o
f
ten
p
r
o
v
id
e
h
ig
h
er
p
r
ed
ictiv
e
ac
cu
r
ac
y
an
d
r
o
b
u
s
tn
ess
co
m
p
ar
ed
with
in
d
iv
i
d
u
al
m
o
d
els
[
6
]
,
[
7
]
.
R
esear
ch
er
s
h
av
e
also
ex
p
lo
r
e
d
th
e
u
s
e
o
f
g
en
etic
an
d
g
en
o
m
ic
in
f
o
r
m
atio
n
f
o
r
d
iab
etes
p
r
ed
ictio
n
.
T
ec
h
n
iq
u
es
s
u
ch
as
g
e
n
o
m
e
-
wid
e
ass
o
ciatio
n
s
tu
d
ies
(
GW
AS)
co
m
b
in
ed
with
ML
h
av
e
b
ee
n
u
s
ed
to
id
en
tify
g
en
etic
m
ar
k
er
s
lin
k
ed
t
o
d
ia
b
etes
r
is
k
[
8
]
–
[
1
0
]
.
T
h
ese
ap
p
r
o
ac
h
es
d
em
o
n
s
tr
ate
th
e
p
o
t
en
tial
o
f
in
teg
r
atin
g
b
io
m
ed
ical
d
ata
with
co
m
p
u
t
atio
n
al
m
o
d
els
to
b
etter
u
n
d
e
r
s
tan
d
d
is
ea
s
e
m
ec
h
a
n
is
m
s
.
C
lin
ical
an
d
life
s
ty
le
f
ac
to
r
s
r
em
ain
th
e
m
o
s
t
wid
el
y
u
s
ed
p
r
e
d
icto
r
s
in
d
iab
etes
r
is
k
m
o
d
ellin
g
.
Var
iab
l
es
s
u
ch
as
ag
e,
B
MI
,
b
lo
o
d
g
lu
co
s
e
lev
el,
an
d
f
am
ily
h
is
to
r
y
h
a
v
e
co
n
s
is
ten
tly
b
ee
n
id
en
tifie
d
as
s
tr
o
n
g
in
d
icato
r
s
o
f
d
iab
etes
r
is
k
.
ML
m
o
d
els tr
ain
ed
o
n
th
ese
f
ea
tu
r
es h
av
e
s
h
o
wn
s
tr
o
n
g
p
r
ed
ictiv
e
p
er
f
o
r
m
an
ce
in
h
ea
lth
ca
r
e
d
atasets
[
6
]
,
[
7
]
.
R
ec
en
t
s
tu
d
ies
h
av
e
also
in
co
r
p
o
r
ated
ex
p
lain
ab
le
a
r
tific
ial
in
tellig
en
ce
(
XAI
)
tec
h
n
iq
u
es
to
im
p
r
o
v
e
m
o
d
el
tr
an
s
p
a
r
en
cy
.
Me
th
o
d
s
s
u
c
h
as
Sh
ap
le
y
a
d
d
itiv
e
ex
p
lan
atio
n
s
(
SHAP
)
an
d
least
ab
s
o
lu
te
s
h
r
in
k
ag
e
a
n
d
s
elec
tio
n
o
p
e
r
ato
r
(
L
ASSO
)
h
elp
in
ter
p
r
et
f
ea
tu
r
e
co
n
tr
ib
u
tio
n
s
a
n
d
ex
p
lain
m
o
d
el
p
r
ed
ictio
n
s
to
clin
ician
s
[
6
]
,
[
1
1
]
,
[
1
2
]
.
Su
ch
in
ter
p
r
eta
b
ilit
y
is
p
a
r
ticu
lar
ly
im
p
o
r
tan
t
in
h
ea
lt
h
ca
r
e
a
p
p
licatio
n
s
wh
er
e
tr
an
s
p
ar
en
cy
an
d
tr
u
s
t
ar
e
c
r
itical.
Sev
er
al
s
tu
d
ies
h
av
e
ev
alu
ated
en
s
em
b
le
le
ar
n
in
g
m
o
d
els
f
o
r
d
iab
etes
p
r
ed
i
ctio
n
.
Deb
er
n
e
h
an
d
Kim
[
1
3
]
ap
p
lied
ex
tr
e
m
e
g
r
ad
ien
t
b
o
o
s
tin
g
(
XGBo
o
s
t
)
-
b
ased
en
s
em
b
le
lear
n
in
g
to
p
r
e
d
ict
T
2
DM
u
s
in
g
a
Ko
r
ea
n
clin
ical
d
ataset.
T
h
eir
m
o
d
el
ac
h
iev
e
d
a
p
p
r
o
x
im
ately
9
9
%
class
if
icatio
n
ac
cu
r
ac
y
u
s
in
g
k
ey
p
r
ed
icto
r
s
s
u
ch
as
Hb
A1
c
lev
els
an
d
B
MI
.
T
h
ese
r
esu
lts
d
em
o
n
s
tr
ate
th
e
s
tr
o
n
g
p
r
ed
ictiv
e
ca
p
ab
ilit
y
o
f
en
s
em
b
le
lea
r
n
in
g
m
eth
o
d
s
an
d
h
ig
h
lig
h
t
th
eir
s
u
p
er
io
r
i
ty
o
v
e
r
tr
ad
itio
n
al
s
in
g
le
-
class
if
ier
m
o
d
els f
o
r
cli
n
ical
r
is
k
p
r
ed
ictio
n
.
Similar
ly
,
Yo
u
s
ef
et
a
l.
[
1
4
]
a
n
aly
ze
d
clin
ical
d
ata
f
r
o
m
2
1
,
4
3
1
p
atien
ts
in
th
e
Sau
d
i
Natio
n
al
Gu
ar
d
p
o
p
u
latio
n
.
T
h
eir
s
tu
d
y
r
ep
o
r
t
ed
p
r
ed
ictio
n
ac
cu
r
ac
ies
b
etw
ee
n
8
5
%
an
d
9
0
%
u
s
in
g
d
ata
m
in
in
g
tech
n
i
q
u
es.
Ho
wev
er
,
m
is
s
in
g
d
ata
a
n
d
d
ataset
lim
itatio
n
s
r
em
ain
e
d
s
ig
n
if
ican
t
ch
alle
n
g
es.
El
-
So
f
an
y
et
a
l.
[
1
5
]
d
ev
elo
p
e
d
a
s
em
i
-
s
u
p
er
v
is
ed
ML
m
o
d
el
in
te
g
r
ated
in
t
o
a
m
o
b
ile
h
ea
lth
a
p
p
licatio
n
f
o
r
d
iab
etes
p
r
ed
ictio
n
.
T
h
e
f
r
am
ewo
r
k
u
s
ed
a
d
ataset
o
f
3
0
0
s
am
p
les
co
llected
f
r
o
m
Sau
d
i
an
d
E
g
y
p
tian
p
o
p
u
la
tio
n
s
.
B
y
ap
p
ly
in
g
s
y
n
th
etic
m
in
o
r
ity
o
v
e
r
-
s
am
p
lin
g
tech
n
iq
u
e
(
SMOT
E
)
to
ad
d
r
ess
class
im
b
alan
ce
,
t
h
e
p
r
o
p
o
s
ed
m
o
d
el
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
8
1
4
I
n
t J Ad
v
Ap
p
l Sci
,
Vo
l.
1
5
,
No
.
2
,
J
u
n
e
2
0
2
6
:
7
9
0
-
803
792
ac
h
iev
ed
ap
p
r
o
x
i
m
ately
9
2
%
p
r
ed
ictio
n
ac
c
u
r
ac
y
a
n
d
d
e
m
o
n
s
tr
ated
th
e
f
ea
s
ib
ilit
y
o
f
m
o
b
ile
-
h
ea
lth
–
b
ased
d
is
ea
s
e
p
r
ed
ictio
n
s
y
s
tem
s
.
Hasan
et
a
l.
[
1
6
]
p
r
o
p
o
s
ed
a
n
au
to
m
ated
m
ac
h
in
e
lear
n
in
g
(
Au
to
ML
)
f
r
am
ew
o
r
k
c
o
m
b
in
ed
with
XAI
tech
n
iq
u
es
f
o
r
d
iab
ete
s
p
r
ed
ictio
n
.
T
h
eir
ap
p
r
o
ac
h
i
n
teg
r
ates
Au
to
ML
with
SHAP
an
d
lo
ca
l
in
ter
p
r
etab
le
m
o
d
el
-
a
g
n
o
s
tic
ex
p
lan
atio
n
s
(
L
I
ME
)
to
i
m
p
r
o
v
e
tr
an
s
p
a
r
en
cy
in
m
o
d
el
p
r
e
d
ictio
n
s
.
T
h
e
f
r
am
ewo
r
k
ac
h
iev
ed
a
p
p
r
o
x
i
m
ately
9
8
%
ac
cu
r
ac
y
w
h
ile
p
r
o
v
id
i
n
g
in
ter
p
r
etab
le
ex
p
lan
atio
n
s
th
at
h
elp
clin
ician
s
u
n
d
er
s
tan
d
th
e
co
n
tr
ib
u
tio
n
o
f
k
e
y
r
is
k
f
ac
to
r
s
.
Z
h
ao
et
a
l.
[
1
7
]
ap
p
lied
XAI
tech
n
iq
u
es to
v
is
u
alize
d
iab
etes
r
is
k
p
r
ed
ictio
n
s
u
s
in
g
SHAP
an
d
L
I
ME
.
T
h
eir
f
r
a
m
ewo
r
k
g
en
er
ate
d
in
ter
p
r
etab
le
ex
p
lan
atio
n
s
th
at
h
elp
clin
ician
s
a
n
d
p
atien
ts
u
n
d
er
s
tan
d
th
e
k
e
y
f
ac
to
r
s
c
o
n
tr
i
b
u
tin
g
to
d
iab
etes
r
is
k
.
Su
ch
v
is
u
aliza
tio
n
-
b
ased
ex
p
lan
atio
n
m
et
h
o
d
s
im
p
r
o
v
e
m
o
d
el
tr
an
s
p
ar
e
n
cy
a
n
d
s
u
p
p
o
r
t
b
etter
p
atien
t
awa
r
e
n
ess
an
d
d
ec
is
io
n
-
m
a
k
in
g
in
d
iab
etes
p
r
ev
e
n
t
io
n
an
d
m
a
n
ag
em
en
t.
Neta
y
awijit
et
a
l.
[
1
8
]
p
r
o
p
o
s
ed
an
in
ter
p
r
etab
le
ML
f
r
am
ewo
r
k
f
o
r
d
iab
etes
p
r
ed
ictio
n
th
at
i
n
teg
r
ates
r
esam
p
lin
g
s
tr
ateg
ies
with
XAI
tech
n
iq
u
es.
T
h
eir
m
eth
o
d
c
o
m
b
in
e
s
SMOT
E
-
b
ased
class
b
alan
cin
g
with
SHAP
ex
p
lan
atio
n
s
to
s
im
u
ltan
eo
u
s
ly
ad
d
r
ess
class
im
b
alan
ce
an
d
m
o
d
el
in
te
r
p
r
etab
ilit
y
.
T
h
e
s
tu
d
y
s
h
o
ws
th
at
in
te
g
r
atin
g
d
ata
b
alan
cin
g
with
ex
p
lain
a
b
le
m
o
d
els
ca
n
im
p
r
o
v
e
b
o
th
p
r
ed
ictiv
e
p
e
r
f
o
r
m
an
ce
a
n
d
tr
an
s
p
ar
e
n
cy
i
n
clin
ical
d
ec
is
io
n
-
s
u
p
p
o
r
t sy
s
tem
s
.
Desp
ite
r
ec
en
t a
d
v
an
ce
s
,
m
an
y
d
iab
etes p
r
ed
ictio
n
s
tu
d
ies r
ely
o
n
r
elativ
ely
s
m
all
d
atasets
o
r
lim
ited
f
ea
tu
r
e
s
ets.
I
n
ad
d
itio
n
,
f
ew
s
tu
d
ies
d
ir
ec
tly
co
m
p
ar
e
co
n
v
en
tio
n
al
ML
m
o
d
els
with
em
er
g
in
g
a
p
p
r
o
ac
h
es
s
u
ch
as
LLM
s
.
T
h
ese
lim
itatio
n
s
h
ig
h
lig
h
t
th
e
n
ee
d
f
o
r
m
o
r
e
co
m
p
r
eh
e
n
s
iv
e
f
r
am
ewo
r
k
s
th
at
in
te
g
r
ate
la
r
g
e
d
atasets
an
d
ad
v
an
ce
d
AI
t
ec
h
n
iq
u
es
.
T
h
is
s
tu
d
y
d
e
v
el
o
p
s
a
d
iab
etes
p
r
ed
ictio
n
f
r
am
ewo
r
k
u
s
in
g
a
lar
g
e
-
s
ca
le
Sau
d
i
h
ea
lth
ca
r
e
d
ataset.
T
h
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
co
m
p
ar
es
co
n
v
en
tio
n
al
ML
m
o
d
els
with
a
f
ew
-
s
h
o
t
LLM
to
ev
alu
ate
p
r
ed
ictiv
e
p
er
f
o
r
m
an
ce
.
B
y
in
teg
r
atin
g
class
b
alan
cin
g
,
e
n
s
em
b
le
lear
n
in
g
,
an
d
in
ter
p
r
etab
ilit
y
an
aly
s
is
,
th
e
f
r
am
ewo
r
k
aim
s
to
p
r
o
v
id
e
a
r
o
b
u
s
t
an
d
clin
ically
r
elev
an
t
to
o
l
f
o
r
d
iab
etes
r
is
k
p
r
e
d
ictio
n
.
A
lis
t o
f
t
h
e
wo
r
k
s
an
d
m
o
d
els is
s
h
o
wn
in
T
ab
le
1
.
T
ab
le
1
.
Su
m
m
a
r
y
o
f
ML
ap
p
r
o
ac
h
es u
s
ed
f
o
r
T
2
DM
p
r
ed
ict
io
n
A
p
p
r
o
a
c
h
M
e
t
h
o
d
s
K
e
y
f
i
n
d
i
n
g
s
G
e
n
e
t
i
c
a
n
d
g
e
n
o
mi
c
G
W
A
S
,
RF
,
a
n
d
m
u
l
t
i
-
o
m
i
c
s
i
n
t
e
g
r
a
t
i
o
n
I
d
e
n
t
i
f
i
e
d
n
o
v
e
l
g
e
n
e
s
a
n
d
si
n
g
l
e
-
n
u
c
l
e
o
t
i
d
e
p
o
l
y
mo
r
p
h
i
sms
(
S
N
P
s)
a
sso
c
i
a
t
e
d
w
i
t
h
T
2
D
M
[
8
]
–
[
1
0
]
C
l
i
n
i
c
a
l
a
n
d
d
e
m
o
g
r
a
p
h
i
c
LR
,
DT
s,
a
n
d
g
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
(
G
B
)
A
g
e
,
B
M
I
,
a
n
d
b
l
o
o
d
g
l
u
c
o
se
a
r
e
k
e
y
p
r
e
d
i
c
t
o
r
s
[
6
]
,
[
7
]
M
e
t
a
b
o
l
o
mi
c
s
Tr
e
e
-
b
a
se
d
b
o
o
st
i
n
g
a
n
d
S
H
A
P
e
x
p
l
a
n
a
t
i
o
n
s
P
h
e
n
y
l
a
c
e
t
a
t
e
a
n
d
t
a
u
r
i
n
e
a
r
e
s
t
r
o
n
g
p
r
e
d
i
c
t
o
r
s
[
6
]
C
a
u
sa
l
i
n
f
e
r
e
n
c
e
C
a
u
sa
l
d
i
sc
o
v
e
r
y
t
e
c
h
n
i
q
u
e
s
a
n
d
l
i
n
e
a
r
n
o
n
-
G
a
u
ss
i
a
n
a
c
y
c
l
i
c
m
o
d
e
l
(
Li
N
G
A
M
)
P
h
y
s
i
c
a
l
i
n
a
c
t
i
v
i
t
y
a
n
d
a
l
c
o
h
o
l
u
se
w
e
r
e
i
d
e
n
t
i
f
i
e
d
a
s
d
i
r
e
c
t
c
a
u
s
e
s
[
1
9
]
,
[
2
0
]
Li
f
e
st
y
l
e
a
n
d
e
n
v
i
r
o
n
m
e
n
t
a
l
X
G
B
o
o
st
a
n
d
S
H
A
P
e
x
p
l
a
n
a
t
i
o
n
s
H
y
p
e
r
t
e
n
si
o
n
a
n
d
a
g
e
a
r
e
si
g
n
i
f
i
c
a
n
t
r
i
sk
f
a
c
t
o
r
s;
s
o
c
i
a
l
a
t
t
r
i
b
u
t
e
s
a
r
e
p
r
o
t
e
c
t
i
v
e
[
2
1
]
XAI
S
H
A
P
a
n
d
LA
S
S
O
f
e
a
t
u
r
e
s
e
l
e
c
t
i
o
n
I
mp
r
o
v
e
d
m
o
d
e
l
i
n
t
e
r
p
r
e
t
a
b
i
l
i
t
y
a
n
d
f
e
a
t
u
r
e
t
r
a
n
s
p
a
r
e
n
c
y
[
6
]
,
[
1
1
]
,
[
1
2
]
P
o
p
u
l
a
t
i
o
n
-
l
e
v
e
l
r
i
s
k
C
l
a
i
ms
d
a
t
a
a
n
a
l
y
si
s
a
n
d
p
o
l
y
s
o
c
i
a
l
r
i
sk
s
c
o
r
e
Ea
r
l
y
a
n
d
l
a
t
e
-
st
a
g
e
r
i
s
k
f
a
c
t
o
r
s i
d
e
n
t
i
f
i
e
d
f
r
o
m
r
e
a
l
-
w
o
r
l
d
d
a
t
a
[
2
2
]
,
[
2
3
]
G
e
n
d
e
r
-
sp
e
c
i
f
i
c
a
n
a
l
y
si
s
S
t
r
a
t
i
f
i
e
d
a
n
a
l
y
si
s
b
y
g
e
n
d
e
r
F
a
st
i
n
g
g
l
u
c
o
se
a
n
d
B
M
I
a
r
e
c
r
i
t
i
c
a
l
f
o
r
w
o
me
n
;
su
g
a
r
c
o
n
su
mp
t
i
o
n
f
o
r
me
n
[
7
]
,
[
2
4
]
A
l
g
o
r
i
t
h
m
c
o
mp
a
r
i
s
o
n
Li
g
h
t
g
r
a
d
i
e
n
t
b
o
o
st
i
n
g
,
X
G
B
o
o
st
,
a
n
d
RF
En
se
mb
l
e
me
t
h
o
d
s
o
u
t
p
e
r
f
o
r
m
t
r
a
d
i
t
i
o
n
a
l
M
L
i
n
b
o
t
h
a
c
c
u
r
a
c
y
a
n
d
st
a
b
i
l
i
t
y
[
6
]
,
[
1
2
]
2.
M
E
T
H
O
D
T
h
is
s
tu
d
y
f
o
llo
wed
a
s
tr
u
ct
u
r
ed
ML
p
ip
elin
e
co
n
s
is
tin
g
o
f
f
iv
e
s
tag
es:
d
ataset
ac
q
u
is
itio
n
,
d
ata
p
r
ep
r
o
ce
s
s
in
g
,
class
im
b
alan
ce
co
r
r
ec
tio
n
,
m
o
d
el
d
ev
el
o
p
m
en
t,
an
d
ev
alu
atio
n
.
First,
a
lar
g
e
-
s
ca
le
d
ataset
co
n
tain
in
g
1
0
0
,
000
a
n
o
n
y
m
i
ze
d
p
atien
t
r
ec
o
r
d
s
was
o
b
tain
ed
f
r
o
m
t
h
e
Sau
d
i
Min
is
tr
y
o
f
Hea
lth
.
Nex
t,
p
r
ep
r
o
ce
s
s
in
g
tech
n
iq
u
es
,
in
cl
u
d
in
g
o
u
tlier
d
etec
tio
n
,
ca
teg
o
r
ical
en
co
d
i
n
g
,
n
o
r
m
aliza
tio
n
,
an
d
m
is
s
in
g
d
ata
h
an
d
lin
g
,
wer
e
ap
p
lied
to
en
s
u
r
e
d
ata
q
u
ality
.
B
ec
au
s
e
th
e
d
ataset
co
n
tain
ed
o
n
ly
9
%
d
iab
etic
ca
s
es,
th
e
SMOT
E
E
NN
h
y
b
r
id
r
esam
p
l
in
g
m
eth
o
d
was
em
p
lo
y
ed
t
o
b
alan
ce
th
e
class
es.
Su
b
s
eq
u
en
tly
,
s
ev
er
al
ML
m
o
d
els
—
in
clu
d
in
g
LR
,
SVM
,
k
-
n
ea
r
est
n
eig
h
b
o
r
s
(
KNN)
,
DT
,
RF
,
an
d
Ad
aBo
o
s
t
—
wer
e
tr
ain
ed
an
d
ev
alu
ated
.
I
n
ad
d
itio
n
,
a
f
e
w
-
s
h
o
t
L
L
M
was
test
ed
to
ass
es
s
its
ab
ilit
y
to
p
er
f
o
r
m
s
tr
u
ctu
r
ed
cli
n
ical
class
if
icatio
n
ta
s
k
s
.
Fin
ally
,
t
h
e
m
o
d
els
wer
e
ev
alu
ated
u
s
in
g
s
tan
d
ar
d
m
etr
ics
s
u
ch
as
ac
cu
r
ac
y
,
p
r
ec
is
io
n
,
r
ec
all,
F1
-
s
co
r
e,
R
OC
-
AUC,
a
n
d
B
r
ier
s
co
r
e
.
2
.
1
.
Da
t
a
s
o
urce
a
nd
des
cr
iptio
n
T
h
e
d
ataset
u
s
ed
in
th
is
s
tu
d
y
was
o
b
tain
ed
f
r
o
m
th
e
Sau
d
i
Min
is
tr
y
o
f
Hea
lth
[
3
]
v
ia
th
e
Sau
d
i
Op
en
Data
Po
r
tal.
I
t
co
m
p
r
is
es
1
0
0
,
0
0
0
an
o
n
y
m
ized
p
atie
n
t
r
ec
o
r
d
s
,
m
a
k
in
g
it
o
n
e
o
f
th
e
m
o
s
t
ex
ten
s
iv
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ad
v
Ap
p
l Sci
I
SS
N:
2252
-
8
8
1
4
Usi
n
g
ma
ch
in
e
lea
r
n
in
g
to
u
n
d
ers
ta
n
d
th
e
r
o
o
t c
a
u
s
es o
f ty
p
e
2
d
i
a
b
etes …
(
Mo
h
a
mma
d
S
a
ee
d
A
l G
h
a
md
i)
793
d
atasets
av
ailab
le
f
o
r
th
e
Sau
d
i
p
o
p
u
latio
n
with
T
2
DM
.
E
ac
h
r
ec
o
r
d
in
clu
d
es
d
e
m
o
g
r
ap
h
i
c
v
ar
iab
les
,
s
u
ch
as
ag
e
an
d
g
en
d
e
r
,
an
th
r
o
p
o
m
etr
ic
in
d
icato
r
s
,
s
u
ch
as
B
MI
,
cl
in
ical
m
ea
s
u
r
em
en
ts
,
s
u
ch
as
Hb
A1
c
an
d
b
lo
o
d
gl
u
co
s
e,
an
d
life
s
ty
le
attr
ib
u
tes
,
s
u
ch
as
s
m
o
k
i
n
g
h
is
to
r
y
.
T
h
e
d
em
o
g
r
ap
h
ic
attr
ib
u
tes
in
cl
u
d
e
ag
e
an
d
g
en
d
er
,
wh
ile
an
th
r
o
p
o
m
etr
ic
m
ea
s
u
r
e
m
en
ts
in
clu
d
e
B
MI
.
C
lin
ical
i
n
d
icato
r
s
co
n
s
is
t
o
f
lab
o
r
a
t
o
r
y
an
d
h
ea
lth
-
r
elate
d
m
ea
s
u
r
em
en
ts
s
u
ch
as Hb
A1
c
lev
els,
b
lo
o
d
g
lu
co
s
e
lev
els,
h
y
p
er
ten
s
io
n
s
tatu
s
,
an
d
h
ea
r
t
d
is
ea
s
e
s
ta
tu
s
.
I
n
ad
d
itio
n
,
life
s
ty
le
in
f
o
r
m
a
tio
n
,
s
u
ch
as
s
m
o
k
in
g
h
is
to
r
y
(
n
e
v
er
,
f
o
r
m
e
r
,
cu
r
r
en
t,
an
d
ev
er
)
is
in
clu
d
ed
in
t
h
e
d
ataset.
T
h
e
t
ar
g
et
v
ar
iab
le
r
ep
r
esen
ts
th
e
d
iab
etes
s
tatu
s
o
f
ea
c
h
p
atien
t,
in
d
icatin
g
wh
eth
er
th
e
in
d
iv
id
u
al
is
d
iag
n
o
s
ed
with
d
iab
etes
o
r
n
o
t.
G
en
etic
m
ar
k
e
r
s
wer
e
in
ten
tio
n
ally
ex
clu
d
ed
f
r
o
m
th
is
s
tu
d
y
to
en
s
u
r
e
th
at
th
e
p
r
o
p
o
s
ed
f
r
am
ewo
r
k
r
em
ai
n
s
p
r
ac
tical
f
o
r
d
ep
l
o
y
m
en
t
in
r
o
u
tin
e
p
r
im
ar
y
h
ea
lth
ca
r
e
s
ettin
g
s
.
2
.
2
.
1
.
O
utlier
r
emo
v
a
l
Ou
tlier
s
in
th
e
co
n
tin
u
o
u
s
v
a
r
iab
les,
in
clu
d
in
g
B
MI
,
b
lo
o
d
g
lu
co
s
e,
a
n
d
Hb
A1
c
,
wer
e
id
en
tifie
d
u
s
in
g
th
e
Z
-
s
co
r
e
m
eth
o
d
.
R
ec
o
r
d
s
with
ab
s
o
lu
te
Z
-
s
co
r
e
v
alu
es
g
r
ea
ter
th
an
3
wer
e
co
n
s
id
er
ed
ex
tr
em
e
o
b
s
er
v
atio
n
s
an
d
r
em
o
v
ed
f
r
o
m
th
e
d
ataset.
T
h
is
s
tep
h
elp
e
d
r
ed
u
ce
n
o
is
e
an
d
i
m
p
r
o
v
e
d
a
ta
co
n
s
is
ten
cy
.
As a
r
esu
lt,
th
e
m
o
d
els we
r
e
tr
ain
e
d
o
n
clea
n
er
an
d
m
o
r
e
r
eliab
le
clin
ical
d
ata.
2
.
2
.
2
.
E
nco
din
g
C
ateg
o
r
ical
v
ar
iab
les
wer
e
tr
a
n
s
f
o
r
m
ed
in
t
o
n
u
m
e
r
ical
f
o
r
m
b
ef
o
r
e
m
o
d
el
tr
ai
n
in
g
.
B
in
ar
y
v
ar
iab
les
s
u
ch
as
g
e
n
d
er
,
h
y
p
er
ten
s
io
n
,
an
d
h
ea
r
t
d
is
ea
s
e
wer
e
en
co
d
ed
u
s
in
g
lab
el
en
c
o
d
in
g
.
M
u
lti
-
class
v
ar
iab
les,
p
ar
ticu
lar
ly
s
m
o
k
in
g
h
is
to
r
y
,
wer
e
c
o
n
v
er
te
d
th
r
o
u
g
h
o
n
e
-
h
o
t
e
n
co
d
i
n
g
t
o
a
v
o
id
in
tr
o
d
u
cin
g
f
a
ls
e
o
r
d
in
al
r
elatio
n
s
h
ip
s
.
T
h
is
en
s
u
r
ed
t
h
a
t c
ateg
o
r
ical
in
f
o
r
m
atio
n
was r
ep
r
esen
ted
ap
p
r
o
p
r
iately
f
o
r
ML
m
o
d
els.
2
.
2
.
3
.
No
rma
liza
t
io
n
C
o
n
tin
u
o
u
s
v
ar
iab
les
wer
e
s
tan
d
ar
d
ized
u
s
in
g
z
-
s
co
r
e
n
o
r
m
aliza
tio
n
to
b
r
in
g
th
em
o
n
to
a
co
m
m
o
n
s
ca
le.
T
h
is
p
r
o
ce
s
s
ad
ju
s
ted
ea
ch
f
ea
tu
r
e
b
y
s
u
b
t
r
ac
tin
g
t
h
e
m
ea
n
an
d
d
i
v
id
in
g
b
y
th
e
s
tan
d
ar
d
d
ev
iatio
n
.
No
r
m
aliza
tio
n
h
elp
e
d
p
r
ev
en
t
f
ea
tu
r
es
with
lar
g
e
r
n
u
m
er
ic
r
an
g
es
f
r
o
m
d
o
m
in
atin
g
th
e
l
ea
r
n
in
g
p
r
o
ce
s
s
.
I
t
also
im
p
r
o
v
e
d
tr
ain
in
g
s
tab
ilit
y
an
d
c
o
n
v
e
r
g
en
ce
f
o
r
t
h
e
ap
p
lied
m
o
d
e
ls
.
2
.
2
.
4
.
M
is
s
i
ng
da
t
a
ha
nd
lin
g
Miss
in
g
v
alu
es
wer
e
m
an
a
g
ed
u
s
in
g
a
two
-
s
tep
ap
p
r
o
ac
h
.
R
ec
o
r
d
s
lack
in
g
cr
itical
clin
ical
f
ea
tu
r
es,
s
u
ch
as
Hb
A1
c
o
r
g
lu
co
s
e
m
ea
s
u
r
em
en
ts
,
wer
e
e
x
clu
d
ed
t
o
av
o
i
d
co
m
p
r
o
m
is
in
g
d
ia
g
n
o
s
tic
ac
cu
r
ac
y
.
L
ess
ess
en
tial
attr
ib
u
tes,
in
clu
d
in
g
life
s
ty
le
v
ar
iab
les,
wer
e
im
p
u
ted
u
s
in
g
KNN
to
p
r
eser
v
e
lo
ca
l
s
im
ilar
ity
p
atter
n
s
.
T
h
e
clea
n
e
d
d
ataset
was th
en
ass
ess
ed
f
o
r
co
m
p
let
en
ess
an
d
co
n
s
is
ten
cy
b
ef
o
r
e
m
o
d
el
tr
ain
in
g
.
2
.
3
.
Cla
s
s
b
a
la
ncing
Mo
s
t
in
d
iv
id
u
als
in
th
e
d
atase
t
d
i
d
n
o
t
h
av
e
d
iab
etes.
On
ly
9
%
h
ad
d
iab
etes.
T
h
is
m
ak
es
it
h
ar
d
f
o
r
th
e
m
o
d
els
to
lear
n
.
So
,
th
e
SMOT
E
E
NN
m
eth
o
d
is
u
s
ed
[
2
5
]
.
T
h
is
to
o
l
p
er
f
o
r
m
s
two
task
s
:
i
t
g
en
er
ates
ad
d
itio
n
al
s
y
n
th
etic
s
am
p
les
f
o
r
th
e
s
m
all
g
r
o
u
p
(
d
iab
etics)
an
d
r
em
o
v
es
o
u
tlier
r
ec
o
r
d
s
f
r
o
m
th
e
lar
g
e
g
r
o
u
p
(
n
o
n
-
d
iab
etics).
Su
b
s
eq
u
e
n
tly
,
th
e
d
ata
wer
e
m
o
r
e
b
alan
ce
d
,
with
ap
p
r
o
x
im
ately
5
2
.
5
%
o
f
p
ar
ticip
an
ts
h
av
in
g
d
iab
etes.
T
h
is
h
elp
ed
th
e
m
o
d
el
f
o
cu
s
m
o
r
e
ef
f
ec
tiv
e
ly
an
d
a
v
o
id
s
elec
tin
g
th
e
m
aj
o
r
ity
class
.
2
.
4
.
M
o
del
d
ev
elo
pm
ent
T
h
e
d
em
o
g
r
ap
h
ic
attr
ib
u
tes
in
clu
d
e
ag
e
an
d
g
en
d
e
r
,
wh
ile
an
th
r
o
p
o
m
etr
ic
m
ea
s
u
r
em
e
n
ts
in
clu
d
e
B
MI
.
C
lin
ical
in
d
icato
r
s
in
clu
d
e
lab
o
r
ato
r
y
an
d
h
ea
lth
-
r
elate
d
m
ea
s
u
r
em
en
ts
s
u
ch
as
Hb
A1
c,
b
lo
o
d
g
lu
co
s
e,
h
y
p
er
ten
s
io
n
s
tatu
s
,
an
d
h
ea
r
t
d
is
ea
s
e
s
tatu
s
.
I
n
ad
d
itio
n
,
life
s
ty
le
in
f
o
r
m
atio
n
s
u
ch
as
s
m
o
k
in
g
h
is
to
r
y
(
n
ev
er
,
f
o
r
m
er
,
cu
r
r
en
t,
an
d
ev
er
)
is
in
clu
d
e
d
in
th
e
d
ata
s
et.
T
h
e
tar
g
et
v
ar
iab
le
in
d
ic
ates
ea
ch
p
atien
t'
s
d
iab
etes
s
tatu
s
,
wh
eth
er
d
iag
n
o
s
ed
o
r
n
o
t.
Gen
etic
m
ar
k
er
s
wer
e
in
ten
tio
n
ally
ex
c
lu
d
ed
f
r
o
m
th
is
s
tu
d
y
to
en
s
u
r
e
th
at
th
e
p
r
o
p
o
s
ed
f
r
a
m
ewo
r
k
r
em
ai
n
s
p
r
ac
tical
f
o
r
d
e
p
lo
y
m
en
t i
n
r
o
u
tin
e
p
r
im
a
r
y
h
ea
lth
ca
r
e
s
ettin
g
s
.
2
.
4
.
1
.
M
o
del
a
rc
hite
ct
ures a
nd
hy
perpa
ra
m
et
er
t
un
ing
Sev
er
al
ML
m
o
d
els
wer
e
i
m
p
lem
en
ted
an
d
tu
n
ed
to
id
en
tify
th
e
m
o
s
t
ef
f
ec
tiv
e
ap
p
r
o
ac
h
f
o
r
d
iab
etes
p
r
ed
ictio
n
.
Hy
p
e
r
p
ar
am
eter
o
p
tim
izatio
n
was
ca
r
r
ied
o
u
t
to
en
s
u
r
e
f
air
co
m
p
ar
is
o
n
an
d
im
p
r
o
v
ed
m
o
d
el
p
er
f
o
r
m
a
n
ce
.
T
h
e
s
elec
ted
co
n
f
ig
u
r
atio
n
s
ar
e
d
escr
ib
ed
.
i)
L
R
:
lin
ea
r
m
o
d
el
with
L
2
r
eg
u
lar
izatio
n
(
R
id
g
e)
;
ar
ch
itectu
r
e:
s
o
f
tm
ax
o
u
tp
u
t
f
o
r
b
i
n
ar
y
class
if
icatio
n
.
T
u
n
ed
v
ia
g
r
id
s
ea
r
ch
cr
o
s
s
-
v
alid
atio
n
(
Gr
id
Sear
ch
C
V
)
:
C
∈
{0
.
0
1
,
0
.
1
,
1
,
1
0
,
1
0
0
}
,
s
o
lv
er
='
lb
f
g
s
',
m
ax
_
iter
=2
0
0
.
Op
tim
al:
C
=1
.
0
.
ii)
SVM:
k
er
n
el
-
b
ased
class
if
ier
;
r
ad
ial
b
asis
f
u
n
ctio
n
(
R
B
F)
k
er
n
el.
T
u
n
e
d
:
C
∈
{
0
.
1
,
1
,
1
0
},
g
a
m
m
a
∈
{'
s
ca
le
'
,
'
au
to
'
,
0
.
0
0
1
}.
O
p
tim
al:
C
=1
,
g
am
m
a=
'
s
ca
le
'
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
8
1
4
I
n
t J Ad
v
Ap
p
l Sci
,
Vo
l.
1
5
,
No
.
2
,
J
u
n
e
2
0
2
6
:
7
9
0
-
803
794
iii)
K
N
N
:
in
s
ta
n
c
e
-
b
a
s
e
d
;
E
u
c
l
i
d
ea
n
d
i
s
t
a
n
c
e
m
et
r
i
c
.
T
u
n
e
d
:
n
_
n
e
i
g
h
b
o
r
s
∈
{
3
,
5,
7,
9
}
,
weig
h
ts
='
d
is
tan
ce
'
.
Op
tim
al:
n
_
n
eig
h
b
o
r
s
=5
.
iv
)
DT
:
si
n
g
le
tr
ee
with
Gin
i
im
p
u
r
ity
;
m
ax
_
d
ep
th
=1
0
to
cu
r
b
o
v
er
f
itti
n
g
.
T
u
n
ed
:
m
ax
_
d
e
p
th
∈
{5
,
1
0
,
1
5
},
m
in
_
s
am
p
les_
s
p
lit
=1
0
.
(
u
s
ed
as R
F/Ad
aBo
o
s
t b
ase.
)
v)
R
F:
e
n
s
em
b
le
o
f
1
0
0
DT
s
;
b
ag
g
in
g
with
r
e
p
lace
m
en
t.
T
u
n
ed
v
ia
Gr
id
Sear
ch
C
V:
n
_
esti
m
ato
r
s
∈
{5
0
,
100,
2
0
0
},
m
a
x
_
d
e
p
th
∈
{1
0
,
2
0
,
No
n
e}
,
m
i
n
_
s
am
p
les_
s
p
lit
∈
{2
,
5
}.
Op
tim
al:
n
_
esti
m
ato
r
s
=1
0
0
,
m
ax
_
d
e
p
th
=
No
n
e.
2
.
5
.
E
v
a
lua
t
i
o
n
m
et
rics
Sev
er
al
s
tan
d
ar
d
e
v
alu
atio
n
m
etr
ics
wer
e
u
s
ed
to
ass
ess
th
e
p
er
f
o
r
m
a
n
ce
o
f
th
e
p
r
ed
ict
io
n
m
o
d
els.
Acc
u
r
ac
y
was
u
s
ed
to
m
ea
s
u
r
e
th
e
p
r
o
p
o
r
tio
n
o
f
co
r
r
ec
tly
class
if
ied
in
s
tan
ce
s
am
o
n
g
all
p
r
ed
ictio
n
s
.
Pre
cisi
o
n
was
ca
lcu
lated
to
d
eter
m
in
e
th
e
p
r
o
p
o
r
tio
n
o
f
p
r
ed
icted
p
o
s
itiv
e
ca
s
es
th
at
wer
e
ac
tu
ally
tr
u
e
p
o
s
itiv
es.
R
ec
all,
also
r
ef
er
r
e
d
to
as
s
en
s
itiv
ity
,
m
ea
s
u
r
es
th
e
ab
ilit
y
o
f
th
e
m
o
d
el
to
co
r
r
e
ctly
id
en
tify
ac
tu
al
d
iab
etic
ca
s
es.
I
n
ad
d
itio
n
to
th
ese
m
etr
ics,
th
e
F1
-
s
co
r
e
was
co
m
p
u
ted
as
th
e
h
ar
m
o
n
ic
m
ea
n
o
f
p
r
ec
is
io
n
an
d
r
ec
all,
p
r
o
v
i
d
in
g
a
b
alan
c
ed
m
ea
s
u
r
e
o
f
m
o
d
el
p
er
f
o
r
m
a
n
ce
wh
en
class
d
is
tr
ib
u
tio
n
s
ar
e
u
n
ev
e
n
.
T
h
e
R
OC
-
AUC
wa
s
u
s
ed
to
ev
alu
ate
th
e
m
o
d
el’
s
ab
ilit
y
to
d
is
tin
g
u
is
h
b
etwe
en
d
ia
b
etic
an
d
non
-
d
iab
etic
in
d
iv
id
u
als
ac
r
o
s
s
d
if
f
er
en
t
class
if
icatio
n
th
r
esh
o
ld
s
.
Fu
r
th
er
m
o
r
e,
th
e
B
r
ier
s
co
r
e
was
ca
lcu
lated
to
ass
ess
th
e
ca
lib
r
atio
n
q
u
ality
o
f
p
r
e
d
icted
p
r
o
b
a
b
ilit
ies,
in
d
icatin
g
h
o
w
well
th
e
p
r
ed
icted
p
r
o
b
a
b
ilit
ies
co
r
r
esp
o
n
d
to
ac
tu
al
o
u
tco
m
es.
Fo
r
m
o
d
el
ev
a
lu
atio
n
,
th
e
d
ataset
was
d
iv
i
d
ed
in
to
tr
ain
in
g
an
d
test
in
g
s
u
b
s
ets
u
s
in
g
a
7
0
%
–
3
0
%
s
p
lit.
T
o
en
s
u
r
e
r
o
b
u
s
tn
ess
an
d
r
ed
u
ce
r
an
d
o
m
n
ess
in
t
h
e
r
esu
lts
,
th
e
ex
p
er
im
en
ts
wer
e
r
ep
ea
ted
t
h
r
ee
tim
es
u
s
in
g
d
if
f
er
en
t
r
an
d
o
m
s
ee
d
s
,
an
d
th
e
f
in
al
p
er
f
o
r
m
an
ce
m
etr
ics
wer
e
r
ep
o
r
ted
as th
e
av
er
a
g
e
ac
r
o
s
s
th
ese
r
u
n
s
.
3.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
3
.
1
.
Da
t
a
s
et
c
ha
ra
c
t
er
is
t
ics
T
h
e
d
ataset
h
as
1
0
0
,
0
0
0
r
ec
o
r
d
s
.
All
p
atien
t
n
am
es
wer
e
r
em
o
v
ed
,
s
o
th
e
d
ata
is
an
o
n
y
m
o
u
s
.
T
h
e
g
en
d
er
p
ar
t
s
h
o
ws
th
at
4
1
.
4
%
ar
e
m
en
(
n
=4
1
,
4
3
9
)
,
a
n
d
5
8
.
6
%
ar
e
wo
m
en
(
n
=5
8
,
5
6
1
)
.
T
h
e
s
m
o
k
in
g
h
is
to
r
y
was
wr
itten
in
f
iv
e
k
in
d
s
:
ev
e
r
cu
r
r
en
t,
n
o
t
c
u
r
r
en
t,
f
o
r
m
er
,
n
ev
er
,
an
d
n
o
in
f
o
r
m
atio
n
.
A
t
f
ir
s
t,
o
n
ly
9
%
o
f
p
eo
p
le
h
ad
d
iab
etes.
T
h
is
in
d
icate
s
th
at
th
e
d
ata
ar
e
n
o
t
b
alan
ce
d
.
SMOT
E
E
NN
wa
s
u
s
ed
to
f
ix
th
is
p
r
o
b
lem
[
4
]
.
Su
b
s
eq
u
en
tl
y
,
t
h
e
n
u
m
b
er
o
f
d
iab
etic
r
e
co
r
d
s
in
cr
ea
s
ed
to
ap
p
r
o
x
im
ately
5
2
.
5
%.
Als
o
,
th
is
m
eth
o
d
r
e
m
o
v
e
d
n
o
is
y
r
ec
o
r
d
s
f
r
o
m
th
e
b
ig
class
.
T
h
e
d
ata
is
n
o
w
m
o
r
e
b
ala
n
ce
d
an
d
r
o
b
u
s
t
f
o
r
t
h
e
p
r
ed
ictio
n
m
o
d
els.
−
=
ℎ
2
.
(
2
+
2
)
(
1
)
All sy
m
b
o
ls
th
at
h
av
e
b
ee
n
u
s
ed
in
th
e
e
q
u
atio
n
s
s
h
o
u
ld
b
e
d
ef
in
ed
in
th
e
f
o
llo
win
g
tex
t.
3
.
2
.
M
o
del
perf
o
rma
nce
o
v
e
rv
iew
3
.
2
.
1
.
Acc
ura
cy
Acc
u
r
ac
y
is
th
e
p
r
o
p
o
r
tio
n
o
f
tim
es
th
e
m
o
d
el
co
r
r
ec
tly
p
r
ed
icts
.
T
h
e
b
est
-
p
er
f
o
r
m
in
g
m
o
d
el
was
Ad
aBo
o
s
t,
ac
h
iev
in
g
9
7
.
2
6
%,
f
o
llo
wed
b
y
RF
at
9
7
.
0
6
%.
T
h
e
lo
west
was
KNN
at
8
5
.
5
%,
f
o
llo
wed
b
y
SVM
at
8
6
.
7
3
%.
T
h
ese
two
m
o
d
els
wer
e
test
ed
af
ter
u
s
in
g
u
n
d
er
s
am
p
lin
g
.
T
h
e
ac
cu
r
ac
y
r
esu
lts
f
o
r
all
m
o
d
els
ar
e
s
u
m
m
ar
ized
in
Fig
u
r
e
1
.
3
.
2
.
2
.
Rec
eiv
er
o
pera
t
ing
ch
a
ra
ct
er
is
t
ic
a
nd
a
re
a
un
der
t
he
curv
e
R
OC
AU
C
s
h
o
ws
h
o
w
well
th
e
m
o
d
el
d
is
tin
g
u
is
h
es
b
et
wee
n
d
iab
etic
an
d
n
o
n
-
d
ia
b
e
tic
p
eo
p
le.
A
s
co
r
e
o
f
1
i
n
d
icate
s
p
er
f
ec
t
p
er
f
o
r
m
an
ce
.
I
n
t
h
is
s
tu
d
y
,
Ad
aBo
o
s
t
g
o
t
0
.
9
8
an
d
R
F
g
o
t
0
.
9
7
.
T
h
e
two
m
o
d
els
p
er
f
o
r
m
ed
well
in
th
is
test
.
KNN
an
d
SVM
h
ad
lo
wer
s
co
r
es
.
R
OC
cu
r
v
es
f
o
r
all
m
o
d
els
ar
e
p
r
esen
ted
in
Fig
u
r
e
2
.
3
.
2
.
3
.
P
re
cisi
o
n
Pre
cisi
o
n
in
d
icate
s
th
e
p
r
o
p
o
r
t
io
n
o
f
p
r
ed
icted
d
iab
etic
ca
s
es
th
at
ar
e
ac
tu
ally
tr
u
e.
T
h
is
is
im
p
o
r
tan
t
b
ec
au
s
e
if
th
e
m
o
d
el
in
d
icate
s
th
at
s
o
m
e
o
n
e
h
as
d
ia
b
etes,
co
n
f
id
en
ce
in
t
h
e
d
i
ag
n
o
s
is
is
n
e
ed
ed
.
I
n
th
is
s
tu
d
y
,
Ad
aBo
o
s
t
ac
h
iev
ed
th
e
h
ig
h
e
s
t
p
r
ec
is
io
n
o
f
9
8
%
o
n
im
b
al
an
ce
d
d
ata.
T
h
e
lo
west
-
p
er
f
o
r
m
in
g
m
eth
o
d
was
KNN,
with
6
6
%.
Fig
u
r
e
3
p
r
e
s
en
ts
a
co
m
p
ar
is
o
n
o
f
p
r
ec
is
io
n
ac
r
o
s
s
all
m
o
d
els.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ad
v
Ap
p
l Sci
I
SS
N:
2252
-
8
8
1
4
Usi
n
g
ma
ch
in
e
lea
r
n
in
g
to
u
n
d
ers
ta
n
d
th
e
r
o
o
t c
a
u
s
es o
f ty
p
e
2
d
i
a
b
etes …
(
Mo
h
a
mma
d
S
a
ee
d
A
l G
h
a
md
i)
795
Fig
u
r
e
1
.
Acc
u
r
ac
y
o
f
s
ev
er
al
ML
m
eth
o
d
s
Fig
u
r
e
2
.
R
OC
cu
r
v
es f
o
r
all
e
v
alu
ated
m
o
d
els
Fig
u
r
e
3
.
T
h
e
p
r
ec
is
io
n
co
m
p
a
r
is
o
n
f
o
r
all
m
o
d
els
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
8
1
4
I
n
t J Ad
v
Ap
p
l Sci
,
Vo
l.
1
5
,
No
.
2
,
J
u
n
e
2
0
2
6
:
7
9
0
-
803
796
3
.
2
.
4
.
Rec
a
ll
T
h
e
p
r
o
p
o
r
tio
n
o
f
ac
tu
al
d
iab
etics
th
at
th
e
m
o
d
el
id
en
tifie
s
.
I
t
is
e
s
s
en
tial
in
h
ea
lth
b
ec
au
s
e
m
is
s
in
g
a
r
ea
l
ca
s
e
i
s
d
an
g
er
o
u
s
.
Ad
aBo
o
s
t
with
u
n
d
er
-
s
am
p
lin
g
ac
h
iev
ed
th
e
b
est
r
ec
all
o
f
9
1
%.
KNN
o
n
im
b
alan
ce
d
d
ata
y
ield
ed
th
e
lo
west sco
r
e
o
f
7
7
%.
Fig
u
r
e
4
p
r
esen
ts
th
e
r
ec
all
s
co
r
es f
o
r
all
m
o
d
els.
Fig
u
r
e
4
.
R
ec
all
co
m
p
a
r
is
o
n
f
o
r
all
m
o
d
els
3
.
2
.
5
.
F1
-
s
co
re
F1
-
s
co
r
e
is
th
e
av
e
r
ag
e
o
f
p
r
ec
is
io
n
an
d
r
ec
all.
I
n
th
e
m
ac
r
o
-
av
er
a
g
ed
s
ettin
g
,
Ad
aBo
o
s
t
ac
h
iev
ed
9
0
%,
wh
er
ea
s
KNN
ac
h
iev
ed
7
0
%.
T
h
is
im
p
lies
th
at
tr
ee
m
o
d
els,
s
u
ch
as
Ad
aBo
o
s
t,
ar
e
m
o
r
e
s
tab
le.
Fo
r
th
e
weig
h
ted
av
e
r
ag
e,
Ad
aBo
o
s
t
an
d
RF
g
o
t
9
7
%.
LR
h
ad
t
h
e
lo
west
with
8
0
%.
Fig
u
r
e
5
p
r
esen
ts
th
e
F1
-
s
co
r
es
f
o
r
all
m
o
d
els.
Fig
u
r
e
5
.
F1
-
s
co
r
e
m
ac
r
o
av
g
co
m
p
ar
is
o
n
f
o
r
all
m
o
d
els
T
h
e
m
ac
r
o
av
e
r
ag
e
tr
ea
ts
b
o
th
class
es
eq
u
ally
,
ev
e
n
w
h
en
th
e
n
u
m
b
er
o
f
d
iab
etics
is
u
s
u
ally
im
b
alan
ce
d
in
th
e
r
ea
l
wo
r
ld
.
T
h
is
is
im
p
o
r
tan
t
i
n
m
ed
ical
d
ata
b
ec
au
s
e
it
is
s
en
s
itiv
e,
a
n
d
it
is
n
ec
ess
ar
y
t
o
know
h
o
w
well
th
e
m
o
d
el
id
e
n
tifie
s
th
e
ac
tu
al
d
iab
et
ic
ca
s
es,
ev
en
wh
en
th
ey
ar
e
r
ar
e.
T
h
e
weig
h
ted
av
er
a
g
e
ass
ig
n
s
g
r
ea
ter
weig
h
t
to
t
h
e
l
ar
g
er
class
.
T
h
is
is
also
u
s
ef
u
l
f
o
r
u
n
d
er
s
tan
d
in
g
o
v
er
all
p
er
f
o
r
m
an
ce
.
Fig
u
r
e
6
p
r
esen
ts
th
e
F1
weig
h
ted
s
co
r
es f
o
r
all
m
o
d
els.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ad
v
Ap
p
l Sci
I
SS
N:
2252
-
8
8
1
4
Usi
n
g
ma
ch
in
e
lea
r
n
in
g
to
u
n
d
ers
ta
n
d
th
e
r
o
o
t c
a
u
s
es o
f ty
p
e
2
d
i
a
b
etes …
(
Mo
h
a
mma
d
S
a
ee
d
A
l G
h
a
md
i)
797
Fig
u
r
e
6
.
F1
-
s
co
r
e
weig
h
ted
a
v
g
co
m
p
a
r
is
o
n
f
o
r
all
m
o
d
els
I
n
th
e
r
esu
lts
,
Ad
aBo
o
s
t
ac
h
iev
ed
a
m
ac
r
o
-
a
v
er
ag
e
o
f
9
0
%,
wh
er
ea
s
KNN
ac
h
iev
ed
7
0
%.
T
h
is
in
d
icate
s
th
at
tr
ee
m
o
d
els,
s
u
ch
as
Ad
aBo
o
s
t,
ar
e
m
o
r
e
s
tab
le.
I
n
th
e
weig
h
ted
a
v
er
ag
e,
Ad
aBo
o
s
t
an
d
RF
b
o
th
ac
h
iev
ed
9
7
%,
w
h
ile
LR
h
ad
th
e
lo
we
s
t
at
8
0
%.
T
h
e
e
n
s
em
b
le
m
o
d
els
p
er
f
o
r
m
ed
b
e
tter
ev
en
wh
en
t
h
e
d
ata
wer
e
im
b
alan
ce
d
.
T
h
e
S
MO
T
E
E
NN
m
eth
o
d
h
elp
e
d
b
a
lan
ce
th
e
d
ata,
a
n
d
th
e
m
o
d
els
b
ec
am
e
s
tr
o
n
g
er
.
3
.
3
.
F
ew
-
s
ho
t
l
a
rg
e
la
ng
ua
g
e
m
o
del
e
v
a
lua
t
i
o
n
L
L
M,
tr
ain
ed
u
s
in
g
a
f
ew
-
s
h
o
t
lear
n
in
g
ap
p
r
o
ac
h
,
y
ield
ed
h
ig
h
class
if
icatio
n
m
etr
ic
s
:
a
cc
u
r
ac
y
(
0
.
9
8
)
,
p
r
ec
is
io
n
(
0
.
9
8
)
,
r
ec
al
l
(
0
.
9
8
)
,
a
n
d
F1
-
s
co
r
e
(
0
.
9
8
)
.
T
h
e
AUC
f
o
r
d
ia
b
etes
class
if
icatio
n
was
0
.
9
9
,
in
d
icatin
g
s
tr
o
n
g
d
is
cr
im
in
ati
v
e
p
o
wer
.
T
h
ese
r
esu
lts
af
f
ir
m
th
e
p
r
o
m
is
e
o
f
L
L
Ms
in
s
tr
u
ctu
r
ed
clin
ical
p
r
ed
ictio
n
task
s
.
T
h
e
p
r
ed
icti
v
e
p
er
f
o
r
m
a
n
ce
o
f
th
e
m
o
d
el
was
f
u
r
th
er
ass
ess
ed
u
s
in
g
R
OC
cu
r
v
es
ac
r
o
s
s
v
ar
io
u
s
clin
ical
co
n
d
itio
n
s
.
T
h
e
m
o
d
el
d
em
o
n
s
tr
ated
h
ig
h
d
is
cr
im
in
atio
n
ac
r
o
s
s
all
ev
alu
ated
co
n
d
itio
n
s.
Sp
ec
if
ically
,
th
e
AUC
f
o
r
d
iab
etes
was
0
.
9
9
,
in
d
icatin
g
h
ig
h
ac
cu
r
ac
y
in
d
is
tin
g
u
is
h
in
g
d
iab
etic
f
r
o
m
non
-
d
iab
etic
s
u
b
jects
an
d
d
e
m
o
n
s
tr
atin
g
ef
f
ec
tiv
e
ca
p
a
b
ilit
y
to
id
en
tif
y
in
d
iv
i
d
u
als
with
d
iab
etes.
T
h
e
h
i
g
h
AUC
v
alu
es
co
n
f
ir
m
th
e
m
o
d
el’
s
ef
f
ec
tiv
e
class
if
icatio
n
o
f
in
d
iv
id
u
als
in
to
th
eir
c
o
r
r
ec
t
d
iag
n
o
s
tic
g
r
o
u
p
s
,
u
n
d
er
s
co
r
i
n
g
its
s
tr
o
n
g
p
o
ten
t
ial
to
aid
ea
r
ly
,
p
r
ec
is
e
d
iag
n
o
s
is
o
f
d
iab
etes.
Fig
u
r
e
7
p
r
e
s
en
ts
a
co
m
p
ar
ativ
e
p
er
f
o
r
m
an
ce
an
aly
s
is
o
f
Ad
a
B
o
o
s
t,
RF
,
an
d
L
L
M
m
o
d
els,
ev
alu
ated
ac
r
o
s
s
f
o
u
r
k
ey
m
etr
ics:
ac
cu
r
ac
y
,
p
r
ec
is
io
n
,
r
ec
all
,
an
d
F1
-
s
co
r
e.
L
L
M
c
o
n
s
is
ten
tly
o
u
tp
e
r
f
o
r
m
s
Ad
aBo
o
s
t
an
d
RF
,
ac
h
ie
v
in
g
9
8
%
ac
r
o
s
s
all
m
etr
ics,
wh
ile
Ad
aBo
o
s
t
an
d
RF
s
h
o
w
h
ig
h
b
u
t
s
lig
h
tly
v
ar
ied
r
esu
lts
.
No
tab
ly
,
A
d
a
B
o
o
s
t
ac
h
iev
es
th
e
h
ig
h
est
p
r
ec
is
io
n
am
o
n
g
th
e
M
L
m
o
d
els,
at
9
8
%,
alth
o
u
g
h
it
h
as
a
co
m
p
a
r
ativ
ely
lo
wer
r
ec
all
o
f
9
1
%.
T
h
is
f
ig
u
r
e
u
n
d
e
r
s
co
r
es
th
e
p
r
ed
ictiv
e
p
o
wer
o
f
f
ew
-
s
h
o
t
-
tr
ain
ed
L
L
Ms
an
d
h
i
g
h
lig
h
ts
th
e
co
m
p
etitiv
e
p
er
f
o
r
m
an
ce
o
f
en
s
em
b
le
ML
m
eth
o
d
s
.
Fig
u
r
e
7
.
C
o
m
p
a
r
is
o
n
b
etwe
e
n
th
e
h
ig
h
est p
e
r
f
o
r
m
an
ce
m
eth
o
d
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
8
1
4
I
n
t J Ad
v
Ap
p
l Sci
,
Vo
l.
1
5
,
No
.
2
,
J
u
n
e
2
0
2
6
:
7
9
0
-
803
798
3
.
4
.
Co
nfidence
inte
rv
a
ls
a
n
d m
o
del r
o
bu
s
t
nes
s
I
n
th
is
p
ar
t,
th
e
s
tab
ilit
y
an
d
r
o
b
u
s
tn
ess
o
f
th
e
Ad
aBo
o
s
t m
o
d
el
wer
e
ev
alu
ated
to
d
eter
m
i
n
e
wh
eth
er
its
p
er
f
o
r
m
an
ce
r
e
m
ain
ed
co
n
s
is
ten
t
ac
r
o
s
s
d
if
f
er
en
t
d
ata
s
am
p
les
r
ath
er
th
an
b
ein
g
s
tr
o
n
g
o
n
ly
i
n
a
s
in
g
le
test
s
et
tin
g
.
T
o
d
o
th
is
,
th
e
m
o
d
el
was
tr
ain
ed
m
u
ltip
le
tim
es
an
d
test
ed
o
n
u
n
s
ee
n
d
ata
in
ea
ch
r
u
n
.
W
h
en
tr
ain
ed
o
n
8
,
0
0
0
r
e
co
r
d
s
an
d
e
v
alu
ated
o
n
2
,
0
0
0
r
e
co
r
d
s
,
th
e
m
o
d
el
ac
h
iev
e
d
an
ac
cu
r
ac
y
o
f
9
5
.
8
%
with
a
9
5
%
C
I
o
f
9
4
.
9
9
%
–
9
6
.
4
6
%,
a
p
r
ec
is
io
n
o
f
9
1
.
6
%
(
9
5
%
C
I
:
9
0
.
3
6
%
–
9
2
.
6
9
%),
a
r
ec
all
o
f
8
7
.
5
%
(
9
5
%
C
I
:
8
6
.
0
5
%
–
8
8
.
8
3
%),
a
n
d
an
F1
-
s
co
r
e
o
f
8
9
.
5
%
(
9
5
%
C
I
:
8
8
.
1
6
%
–
9
0
.
7
2
%).
I
n
a
d
d
itio
n
,
th
e
m
o
d
el
p
r
o
d
u
ce
d
a
R
OC
-
AU
C
o
f
0
.
9
8
8
,
a
p
r
ec
is
io
n
-
r
ec
all
-
AUC
o
f
0
.
9
6
,
an
d
a
B
r
ier
s
co
r
e
o
f
0
.
0
2
9
.
T
h
ese
r
esu
lts
in
d
icate
th
at
th
e
Ad
aBo
o
s
t
m
o
d
el
m
ain
tain
ed
h
ig
h
p
r
ed
ictiv
e
p
er
f
o
r
m
an
ce
a
n
d
r
eliab
le
p
r
o
b
ab
ilit
y
esti
m
atio
n
ac
r
o
s
s
r
ep
ea
ted
ev
alu
atio
n
s
.
3
.
5
.
St
a
t
is
t
ica
l
s
ig
nifica
nce
t
esting
T
-
test
s
,
as
s
h
o
wn
in
T
ab
le
2
,
r
ev
ea
l
s
tatis
tically
s
ig
n
if
ican
t
d
if
f
er
en
ce
s
in
all
co
n
tin
u
o
u
s
f
ea
tu
r
es
(
p
<0
.
0
0
1
)
,
in
clu
d
in
g
ag
e,
B
M
I
,
Hb
A1
c,
an
d
b
lo
o
d
g
lu
c
o
s
e.
C
h
i
-
s
q
u
ar
e
test
s
,
as
s
h
o
wn
in
T
ab
le
3
,
co
n
f
ir
m
ed
th
at
all
ca
teg
o
r
ical
v
ar
iab
les
(
g
en
d
e
r
,
h
y
p
e
r
ten
s
io
n
,
h
ea
r
t
d
is
ea
s
e,
an
d
s
m
o
k
in
g
h
is
to
r
y
)
wer
e
s
ig
n
if
ican
tly
ass
o
ciate
d
with
d
iab
etes
s
tatu
s
(
p
<0
.
0
0
1
)
.
Similar
ly
,
th
e
c
h
i
-
s
q
u
ar
e
test
f
o
r
ca
teg
o
r
ical
v
ar
iab
les
(
T
ab
le
3
)
in
d
icate
s
th
at
all
ca
teg
o
r
ical
p
r
ed
icto
r
s
ar
e
s
ig
n
if
ican
tly
ass
o
ciate
d
with
d
iab
etes statu
s
.
T
ab
le
2
.
T
-
test
s
(
co
n
tin
u
o
u
s
v
ar
iab
les)
F
e
a
t
u
r
e
t
-
st
a
t
i
s
t
i
c
p
-
v
a
l
u
e
A
g
e
S
i
g
n
i
f
i
c
a
n
t
p
<
0
.
0
0
1
B
M
I
S
i
g
n
i
f
i
c
a
n
t
p
<
0
.
0
0
1
H
b
A
1
c
l
e
v
e
l
S
i
g
n
i
f
i
c
a
n
t
p
<
0
.
0
0
1
B
l
o
o
d
g
l
u
c
o
s
e
l
e
v
e
l
S
i
g
n
i
f
i
c
a
n
t
p
<
0
.
0
0
1
T
ab
le
3
.
C
h
i
-
s
q
u
ar
e
test
s
(
ca
teg
o
r
ical
v
ar
ia
b
les)
F
e
a
t
u
r
e
χ2
-
st
a
t
i
st
i
c
p
-
v
a
l
u
e
G
e
n
d
e
r
S
i
g
n
i
f
i
c
a
n
t
p
<
0
.
0
0
1
H
y
p
e
r
t
e
n
si
o
n
S
i
g
n
i
f
i
c
a
n
t
p
<
0
.
0
0
1
H
e
a
r
t
d
i
s
e
a
s
e
S
i
g
n
i
f
i
c
a
n
t
p
<
0
.
0
0
1
S
mo
k
i
n
g
h
i
st
o
r
y
S
i
g
n
i
f
i
c
a
n
t
p
<
0
.
0
0
1
3
.
6
.
F
e
a
t
ure
i
m
po
rt
a
nce
A
s
im
p
le
m
o
d
el
ca
lled
LR
was
u
s
ed
.
I
t
g
av
e
s
co
r
es
to
ea
ch
h
ea
lth
f
ea
tu
r
e.
Hig
h
er
s
co
r
es
i
n
d
icate
a
g
r
ea
ter
lik
elih
o
o
d
o
f
a
d
iab
ete
s
d
iag
n
o
s
is
.
T
h
e
to
p
s
ig
n
s
wer
e
Hb
A1
c
an
d
b
lo
o
d
s
u
g
ar
.
T
h
en
ca
m
e
B
MI
,
ag
e,
an
d
b
lo
o
d
p
r
ess
u
r
e.
T
h
ese
r
e
s
u
lts
ar
e
co
n
s
is
ten
t
wit
h
wh
at
d
o
cto
r
s
ty
p
ically
ch
ec
k
.
T
h
ey
also
ag
r
ee
with
r
esu
lts
f
r
o
m
s
tr
o
n
g
er
m
o
d
els,
s
u
ch
as
Ad
aBo
o
s
t.
No
ad
v
an
c
ed
m
ath
was
u
s
ed
.
T
h
e
m
o
d
el
ass
ig
n
ed
ea
ch
item
a
n
u
m
b
er
(
a
co
ef
f
icie
n
t)
,
a
n
d
th
e
s
ig
n
if
ican
ce
o
f
th
at
n
u
m
b
er
was
e
x
am
in
e
d
.
Fig
u
r
e
8
s
h
o
ws
h
o
w
s
o
m
e
f
ea
tu
r
es st
o
o
d
o
u
t m
o
r
e
th
an
o
th
er
s
.
T
h
is
in
d
icate
s
th
at
th
e
m
o
d
el
o
b
s
er
v
es th
e
s
am
e
p
atter
n
s
as r
ea
l d
o
cto
r
s
.
Fig
u
r
e
8
.
Featu
r
e
im
p
o
r
ta
n
ce
b
ased
o
n
LR
co
e
f
f
icien
ts
co
n
f
i
r
m
s
th
e
s
tr
o
n
g
ass
o
ciatio
n
o
f
Hb
A1
c
an
d
g
lu
co
s
e
m
etr
ics with
d
iab
etes c
lass
if
ic
a
tio
n
3
.
7
.
M
o
del
c
a
lib
ra
t
io
n
A
ca
lib
r
atio
n
c
u
r
v
e
was
d
r
aw
n
u
s
in
g
th
e
Ad
aBo
o
s
t
m
o
d
el
(
tr
ain
ed
o
n
3
,
0
0
0
r
ec
o
r
d
s
)
.
T
h
e
r
esu
lts
f
o
llo
wed
th
e
4
5
-
d
e
g
r
ee
r
e
f
er
e
n
ce
lin
e
f
air
ly
well,
esp
ec
iall
y
in
th
e
m
id
d
le
-
an
d
h
ig
h
-
p
r
o
b
ab
ilit
y
ar
ea
s
.
T
h
is
in
d
icate
s
th
at
th
e
m
o
d
el
p
r
o
d
u
ce
d
r
eliab
le
p
r
o
b
ab
ilit
ies
an
d
co
u
ld
in
f
o
r
m
clin
ical
d
ec
is
io
n
-
m
ak
in
g
.
Fig
u
r
e
9
illu
s
tr
ates
th
e
ca
lib
r
atio
n
cu
r
v
e
co
m
p
ar
i
n
g
p
r
ed
icted
p
r
o
b
ab
ilit
ies
with
o
b
s
er
v
ed
o
u
t
co
m
es.
T
h
e
d
ash
ed
d
iag
o
n
al
lin
e
r
ep
r
esen
ts
p
er
f
ec
tly
ca
lib
r
ated
p
r
ed
ictio
n
s
,
wh
er
e
p
r
ed
icted
p
r
o
b
ab
ilit
i
es
ex
ac
tly
m
atch
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ad
v
Ap
p
l Sci
I
SS
N:
2252
-
8
8
1
4
Usi
n
g
ma
ch
in
e
lea
r
n
in
g
to
u
n
d
ers
ta
n
d
th
e
r
o
o
t c
a
u
s
es o
f ty
p
e
2
d
i
a
b
etes …
(
Mo
h
a
mma
d
S
a
ee
d
A
l G
h
a
md
i)
799
o
b
s
er
v
ed
ev
e
n
t
f
r
eq
u
e
n
cies.
T
h
e
Ad
aBo
o
s
t
m
o
d
el
clo
s
ely
f
o
llo
ws
th
is
r
ef
er
en
ce
lin
e,
in
d
icatin
g
r
eliab
le
p
r
o
b
a
b
ilit
y
esti
m
atio
n
an
d
s
tr
o
n
g
clin
ical
ap
p
licab
ilit
y
.
Fig
u
r
e
9
.
C
alib
r
atio
n
c
u
r
v
e
s
h
o
win
g
p
r
e
d
icted
v
s
.
o
b
s
er
v
ed
p
r
o
b
a
b
ilit
ies f
o
r
th
e
Ad
aBo
o
s
t
m
o
d
el
3
.
8
.
Cro
s
s
-
v
a
lid
a
t
io
n str
a
t
eg
y
B
ey
o
n
d
th
e
in
itial
7
0
/3
0
tr
ain
-
test
s
p
lit
(
s
tr
atif
ied
)
,
a
5
-
f
o
ld
s
tr
atif
ied
k
-
f
o
ld
cr
o
s
s
-
v
alid
atio
n
(
CV
)
was
em
p
lo
y
ed
d
u
r
in
g
tu
n
in
g
(
k
=5
,
s
h
u
f
f
le
=
t
r
u
e)
to
esti
m
at
e
g
en
er
aliza
tio
n
.
Fo
r
th
e
f
in
al
ev
alu
atio
n
,
m
o
d
els
wer
e
r
etr
ain
ed
o
n
th
e
f
u
ll
tr
ai
n
in
g
f
o
ld
s
a
n
d
test
ed
o
n
an
u
n
s
ee
n
3
0
%
(
n
≈
2
9
,
6
0
0
)
.
Var
i
ab
ilit
y
was
ass
ess
e
d
v
ia
th
r
ee
r
ep
ea
ted
r
u
n
s
(
d
if
f
er
en
t
s
ee
d
s
:
4
2
,
1
2
3
,
4
5
6
)
,
an
d
m
ea
n
s
±
SD
wer
e
r
ep
o
r
ted
.
T
h
is
y
i
eld
ed
tig
h
t
C
I
s
(
e.
g
.
,
Ad
aBo
o
s
t a
cc
u
r
ac
y
: 9
7
.
2
6
%±
0
.
4
2
%),
co
n
f
ir
m
in
g
s
tab
ilit
y
.
3
.
9
.
I
nte
rpre
t
a
bil
it
y
a
na
ly
s
i
s
T
o
im
p
r
o
v
e
th
e
in
ter
p
r
etab
ilit
y
o
f
th
e
p
r
ed
ictio
n
m
o
d
els,
SHAP
v
0
.
4
2
.
0
was
em
p
lo
y
ed
t
o
co
m
p
u
te
k
er
n
el
SHAP
ap
p
r
o
x
im
atio
n
s
f
o
r
th
e
test
s
et
u
s
in
g
1
0
0
b
ac
k
g
r
o
u
n
d
s
am
p
les.
Glo
b
al
f
ea
t
u
r
e
im
p
o
r
tan
ce
was
ass
es
s
ed
u
s
in
g
th
e
m
ea
n
ab
s
o
lu
te
SHAP
v
alu
es
co
m
p
u
ted
ac
r
o
s
s
1
,
0
0
0
test
s
am
p
les.
T
h
e
an
aly
s
is
s
h
o
wed
th
at
Hb
A1
c
was
th
e
m
o
s
t
in
f
l
u
en
tial
p
r
e
d
icto
r
with
a
SHAP
v
alu
e
o
f
0
.
2
8
,
an
d
v
al
u
es
ab
o
v
e
6
.
5
%
ac
co
u
n
ted
f
o
r
6
2
%
o
f
d
iab
etic
attr
ib
u
tio
n
s
.
B
lo
o
d
g
lu
co
s
e
was
th
e
s
ec
o
n
d
m
o
s
t
im
p
o
r
ta
n
t
f
ea
tu
r
e
wit
h
a
SHAP
v
alu
e
o
f
0
.
2
2
,
s
h
o
win
g
a
s
tr
o
n
g
in
ter
ac
tio
n
with
Hb
A1
c,
wh
ile
f
asti
n
g
g
lu
co
s
e
lev
els
ab
o
v
e
1
2
6
m
g
/d
L
in
cr
ea
s
ed
th
e
r
is
k
o
d
d
s
b
y
3
.
2
tim
es.
B
MI
also
p
lay
e
d
a
s
ig
n
if
ican
t
r
o
le
with
a
SHAP
v
alu
e
o
f
0
.
1
5
,
wh
er
e
o
b
esit
y
lev
els
ab
o
v
e
3
0
e
x
p
lain
e
d
1
8
%
o
f
th
e
v
ar
ian
ce
,
f
o
llo
we
d
b
y
ag
e
w
ith
a
SHAP
v
alu
e
o
f
0
.
1
2
,
in
d
i
ca
tin
g
a
h
ig
h
er
r
is
k
am
o
n
g
i
n
d
iv
id
u
als
o
ld
er
t
h
an
4
5
y
ea
r
s
.
Hy
p
er
ten
s
io
n
co
n
t
r
ib
u
ted
a
SHAP
v
alu
e
o
f
0
.
0
9
,
wh
er
e
its
p
r
esen
ce
in
cr
ea
s
ed
th
e
lo
g
-
o
d
d
s
b
y
0
.
1
4
,
wh
ile
g
en
d
e
r
a
n
d
s
m
o
k
in
g
h
is
to
r
y
h
a
d
r
elativ
el
y
s
m
aller
co
n
tr
ib
u
tio
n
s
with
SHAP
v
alu
es
o
f
0
.
0
3
an
d
0
.
0
2
,
r
esp
ec
tiv
ely
,
alth
o
u
g
h
cu
r
r
en
t
s
m
o
k
in
g
s
h
o
wed
th
e
s
tr
o
n
g
e
s
t
ef
f
ec
t
with
in
th
e
s
m
o
k
in
g
ca
teg
o
r
ies.
3
.
1
0
.
Clini
ca
l
re
lev
a
nce
o
f
t
o
p pre
dict
o
rs
T
o
p
f
ea
tu
r
es
m
a
p
d
i
r
ec
tly
to
Am
er
ican
Diab
etes
Ass
o
ciatio
n
(
ADA)
/W
HO
g
u
id
eli
n
es
[
2
5
]
:
Hb
A1
c/g
lu
co
s
e
as
d
iag
n
o
s
tic
g
o
ld
s
tan
d
ar
d
s
(
s
en
s
itiv
ity
9
7
%
wh
en
c
o
m
b
in
e
d
)
;
B
MI
/ag
e/h
y
p
er
te
n
s
io
n
as
m
o
d
if
iab
le
r
is
k
s
in
Sau
d
i
co
n
tex
ts
(
e.
g
.
,
s
ed
en
tar
is
m
ex
ac
er
b
ates
o
b
esit
y
;
p
r
ev
alen
ce
3
5
%
[
1
]
)
.
SHAP
co
n
f
ir
m
s
ca
u
s
al
p
lau
s
ib
ilit
y
—
e.
g
.
,
+1
%
Hb
A1
c
s
h
if
ts
p
r
ed
ictio
n
p
r
o
b
ab
ilit
y
f
r
o
m
0
.
2
to
0
.
6
5
—
en
ab
lin
g
p
er
s
o
n
alize
d
in
ter
v
e
n
tio
n
s
(
e.
g
.
,
m
etf
o
r
m
in
f
o
r
h
ig
h
-
H
b
A1
c)
.
I
n
h
ig
h
-
p
r
ev
ale
n
ce
Sau
d
i
s
ettin
g
s
,
th
is
p
r
io
r
itizes
co
s
t
-
ef
f
ec
tiv
e
s
cr
ee
n
in
g
(
e.
g
.
,
Hb
A1
c
test
s
at
SA
R
5
0
)
,
p
o
te
n
tially
r
ed
u
ci
n
g
in
cid
en
ce
b
y
2
0
%
v
ia
ea
r
ly
f
lag
g
i
n
g
[
2
6
]
.
L
im
itatio
n
s
:
SHAP
as
s
u
m
es
f
ea
tu
r
e
in
d
ep
en
d
en
ce
;
f
u
t
u
r
e
ca
u
s
al
ML
(
e.
g
.
,
Do
W
h
y
)
co
u
ld
d
is
en
tan
g
le
co
n
f
o
u
n
d
er
s
lik
e
d
iet.
3
.
1
1
.
Co
m
pa
riso
n wit
h
ex
is
t
ing
m
o
dels
T
ab
le
4
p
r
esen
ts
a
co
m
p
ar
ativ
e
o
v
er
v
iew
o
f
th
e
p
r
o
p
o
s
ed
f
r
am
ewo
r
k
ag
ain
s
t
r
ep
r
esen
tativ
e
d
iab
etes
p
r
ed
ictio
n
ap
p
r
o
ac
h
es
r
ep
o
r
t
ed
in
p
r
io
r
s
tu
d
ies.
Mo
s
t
ex
is
tin
g
wo
r
k
s
r
ely
o
n
r
elativ
ely
s
m
all
d
atasets
,
Evaluation Warning : The document was created with Spire.PDF for Python.