I
AE
S In
t
er
na
t
io
na
l J
o
urna
l o
f
Art
if
icia
l In
t
ellig
ence
(
I
J
-
AI
)
Vo
l.
13
,
No
.
4
,
Dec
em
b
er
2
0
2
4
,
p
p
.
4
4
5
3
~
4
4
7
1
I
SS
N:
2
2
5
2
-
8
9
3
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijai.v
13
.i
4
.
p
p
4
4
5
3
-
4
4
7
1
4453
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
i
.
ia
esco
r
e.
co
m
Inno
v
a
tive ma
chine learning
appro
a
ches for pr
edicti
o
n of
hy
po
g
ly
cemia
in
pa
tient
s
with
t
y
pe
2
d
ia
betes
Sa
chin Ra
m
na
t
h G
a
ik
wa
d
1
,
Seet
a
Dev
i
2
,
Sa
m
ee
r
Sh
ek
ha
r
3
,
Dipa
li Dum
bre
4
1
D
e
p
a
r
t
me
n
t
o
f
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
M
a
c
h
i
n
e
L
e
a
r
n
i
n
g
,
S
y
m
b
i
o
si
s
I
n
st
i
t
u
t
e
o
f
T
e
c
h
n
o
l
o
g
y
,
S
y
m
b
i
o
si
s
I
n
t
e
r
n
a
t
i
o
n
a
l
D
e
e
m
e
d
U
n
i
v
e
r
si
t
y
,
P
u
n
e
,
I
n
d
i
a
2
D
e
p
a
r
t
me
n
t
o
f
O
b
s
t
e
t
r
i
c
s
a
n
d
G
y
n
a
e
c
o
l
o
g
i
c
a
l
N
u
r
s
i
n
g
,
S
y
mb
i
o
s
i
s
C
o
l
l
e
g
e
o
f
N
u
r
si
n
g
,
S
y
mb
i
o
si
s I
n
t
e
r
n
a
t
i
o
n
a
l
D
e
e
me
d
U
n
i
v
e
r
s
i
t
y
,
P
u
n
e
,
I
n
d
i
a
3
D
e
p
a
r
t
me
n
t
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
a
n
d
E
n
g
i
n
e
e
r
i
n
g
,
S
y
mb
i
o
s
i
s I
n
st
i
t
u
t
e
o
f
Te
c
h
n
o
l
o
g
y
,
S
y
m
b
i
o
si
s
I
n
t
e
r
n
a
t
i
o
n
a
l
D
e
e
me
d
U
n
i
v
e
r
s
i
t
y
,
P
u
n
e
,
I
n
d
i
a
4
D
e
p
a
r
t
me
n
t
o
f
M
e
d
i
c
a
l
a
n
d
S
u
r
g
i
c
a
l
N
u
r
si
n
g
,
S
y
mb
i
o
s
i
s
C
o
l
l
e
g
e
o
f
N
u
r
si
n
g
,
S
y
m
b
i
o
si
s
I
n
t
e
r
n
a
t
i
o
n
a
l
D
e
e
m
e
d
U
n
i
v
e
r
s
i
t
y
,
P
u
n
e
,
I
n
d
i
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
No
v
18
,
2
0
2
3
R
ev
is
ed
Mar
23
,
2
0
2
4
Acc
ep
ted
J
u
n
1
,
2
0
2
4
M
e
d
ica
l
d
a
ta
sc
ien
c
e
a
d
v
a
n
c
e
s
u
sin
g
m
a
c
h
i
n
e
lea
rn
i
n
g
,
wh
ic
h
p
re
d
icts
g
lu
c
o
se
lev
e
ls.
A
s
u
p
e
rv
ise
d
m
a
c
h
in
e
lea
rn
in
g
tec
h
n
iq
u
e
is
e
m
p
lo
y
e
d
i
n
wh
ich
re
g
re
ss
io
n
a
n
d
c
las
sifica
ti
o
n
m
e
th
o
d
s
a
re
u
se
d
t
o
c
h
e
c
k
t
h
e
p
re
d
ictio
n
p
e
rfo
rm
a
n
c
e
.
Th
e
u
n
s
u
p
e
rv
ise
d
m
a
c
h
in
e
lea
rn
in
g
tec
h
n
iq
u
e
m
a
k
e
s
c
lu
ste
rs
b
a
se
d
o
n
v
a
riab
les
'
sim
il
a
rit
ies
.
F
u
r
th
e
rm
o
re
,
th
e
p
re
d
icti
o
n
a
c
c
u
ra
c
y
o
f
c
o
n
v
e
n
ti
o
n
a
l
m
a
c
h
in
e
lea
rn
i
n
g
tec
h
n
i
q
u
e
s
is
imp
r
o
v
e
d
b
y
p
r
o
p
o
sin
g
a
tran
sfe
r
lea
rn
in
g
tec
h
n
iq
u
e
.
Ba
se
d
o
n
a
m
e
d
ian
v
a
lu
e
o
f
6
7
m
g
/
d
L,
th
e
d
a
ta
se
t
is
d
iv
i
d
e
d
i
n
t
o
two
g
r
o
u
p
s:
g
r
o
u
p
1
(
BS
L
5
7
m
g
/
d
L
t
o
6
7
m
g
/d
L)
h
a
s
5
0
.
6
7
%
o
f
th
e
sa
m
p
les
,
a
n
d
g
ro
u
p
2
(with
B
S
L
6
8
m
g
/
d
L
to
7
9
m
g
/
d
L)
h
a
s
4
9
.
3
3
%
o
f
th
e
sa
m
p
les
.
In
re
g
re
ss
io
n
a
n
a
ly
si
s,
5
-
fo
l
d
c
ro
ss
-
v
a
li
d
a
ti
o
n
is p
e
rf
o
r
m
e
d
.
Th
e
d
e
c
isio
n
tree
(
DT
)
a
n
d
g
ra
d
ie
n
t
b
o
o
sti
n
g
(
GB
)
in
d
i
v
id
u
a
ll
y
p
ro
v
id
e
a
p
re
d
ictio
n
a
c
c
u
ra
c
y
o
f
1
8
.
2
%
.
R
e
g
a
rd
in
g
c
las
sifica
ti
o
n
a
n
a
ly
sis
,
a
1
0
-
fo
l
d
c
ro
ss
-
v
a
li
d
a
ti
o
n
c
o
n
fi
g
u
ra
ti
o
n
is
u
se
d
fo
r
train
in
g
a
n
d
tes
ti
n
g
t
h
e
m
o
d
e
l
.
Ad
a
Bo
o
st,
G
B,
ra
n
d
o
m
fo
re
st,
a
n
d
n
e
u
ra
l
n
e
two
r
k
a
c
h
iev
e
a
n
a
c
c
u
ra
c
y
ra
te
o
f
6
6
.
3
%
a
n
d
a
n
a
re
a
u
n
d
e
r
c
u
r
v
e
(
AUC
)
sc
o
re
o
f
0
.
7
3
1
.
I
n
u
n
su
p
e
rv
ise
d
lea
rn
in
g
,
th
e
d
a
tas
e
ts
a
re
d
iv
id
e
d
in
to
th
re
e
c
lu
ste
rs.
T
h
e
c
lu
ste
rin
g
re
su
lt
i
s
u
se
d
i
n
re
g
re
ss
io
n
a
n
d
c
las
sifica
ti
o
n
m
o
d
e
ls
u
si
n
g
tran
sfe
r
lea
r
n
in
g
.
Th
e
a
c
c
u
ra
c
y
a
n
d
p
re
c
isio
n
o
f
t
h
e
Ad
a
Bo
o
st
a
n
d
G
B
a
re
a
s
fo
ll
o
ws
:
6
9
.
6
%
,
0
.
6
9
6
with
f1
0
.
6
6
1
a
n
d
6
9
.
6
%
,
0
.
7
0
8
w
it
h
f
1
0
.
7
0
8
,
re
sp
e
c
ti
v
e
ly
.
K
ey
w
o
r
d
s
:
Hig
h
r
is
k
f
ac
to
r
s
Hy
p
o
g
ly
ca
e
m
ia
I
n
n
o
v
ativ
e
m
ac
h
in
e
lear
n
in
g
ap
p
r
o
ac
h
es
Pre
d
icitio
n
T
y
p
e
2
d
iab
etes
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
:
Seeta
Dev
i
Sy
m
b
io
s
is
C
o
lleg
e
o
f
Nu
r
s
in
g
,
Sy
m
b
io
s
is
I
n
ter
n
atio
n
al
Dee
m
ed
Un
iv
er
s
ity
Pu
n
e
4
1
2
1
1
5
,
I
n
d
ia
E
m
ail: d
r
s
ee
tad
ev
i1
9
8
1
@
g
m
a
il.c
o
m
1.
I
NT
RO
D
UCT
I
O
N
Diab
etes
m
ellitu
s
(
DM
)
is
a
lo
n
g
-
ter
m
m
ed
ical
d
is
o
r
d
er
th
at
in
v
o
lv
es h
ig
h
b
lo
o
d
g
lu
c
o
s
e
(
B
G)
lev
els
ca
u
s
ed
b
y
an
in
ab
ilit
y
to
m
ak
e
o
r
u
s
e
in
s
u
lin
p
r
o
p
er
ly
.
Kee
p
in
g
n
o
r
m
al
b
lo
o
d
s
u
g
ar
lev
el
s
ca
n
h
elp
p
r
ev
en
t
d
iab
etic
co
m
p
licatio
n
s
af
f
ec
ti
n
g
b
o
th
m
in
o
r
a
n
d
lar
g
e
b
lo
o
d
v
ess
els
[
1
]
–
[
4
]
.
T
ig
h
t
m
an
a
g
e
m
en
t
o
f
b
l
o
o
d
s
u
g
ar
lev
els
elev
ates
th
e
ch
a
n
ce
s
o
f
e
x
p
er
ien
cin
g
h
y
p
o
g
ly
ce
m
ic
ep
i
s
o
d
es
b
u
t
c
o
n
cu
r
r
en
tly
d
im
in
u
tio
n
s
th
e
lik
elih
o
o
d
o
f
e
n
d
u
r
i
n
g
ch
r
o
n
ic
c
o
m
p
lica
tio
n
s
[
5
]
.
Hy
p
o
g
ly
ca
em
ia
is
c
o
m
m
o
n
l
y
d
etec
te
d
as
a
b
l
o
o
d
s
u
g
ar
lev
el
o
f
less
th
an
7
0
m
g
/d
L
,
wh
ic
h
ca
u
s
es
two
p
h
y
s
io
lo
g
ical
r
esp
o
n
s
es:
a
co
u
n
ter
r
eg
u
lato
r
y
r
ea
c
tio
n
an
d
d
ec
r
ea
s
ed
co
g
n
itiv
e
a
b
ilit
y
[
6
]
–
[
8
]
.
Hy
p
o
g
ly
ce
m
ia
ca
n
ca
u
s
e
s
ig
n
if
ica
n
t
ca
r
d
i
o
v
ascu
lar
ev
en
ts
a
n
d
r
esu
lts
in
f
ea
r
o
f
lo
w
b
lo
o
d
s
u
g
ar
,
m
elan
c
h
o
ly
,
an
d
ab
n
o
r
m
al
m
u
s
cu
lar
m
o
v
em
en
ts
,
in
a
d
d
itio
n
to
d
ep
r
ess
io
n
an
d
d
e
v
iatio
n
i
n
p
h
y
s
ical
well
-
b
ein
g
[
9
]
–
[
1
3
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
13
,
No
.
4
,
Dec
em
b
er
2
0
2
4
:
4
4
5
3
-
4
4
7
1
4454
C
lin
ically
,
h
y
p
o
g
ly
ce
m
ia
is
ca
teg
o
r
ized
as
m
o
d
er
ate
(
MH
)
o
r
s
ev
er
e
(
SH)
b
ased
o
n
w
h
eth
er
th
e
p
atien
t
r
eq
u
ir
es
m
ed
ical
ass
is
t
an
ce
d
u
r
in
g
ep
is
o
d
es
o
f
lo
w
b
lo
o
d
s
u
g
ar
o
r
if
th
er
e
is
a
lo
s
s
o
f
co
n
s
cio
u
s
n
ess
[
1
4
]
.
I
t
h
as
b
ee
n
esti
m
ated
th
at
p
atien
ts
with
ty
p
e
1
d
iab
etes
(
T
1
D)
h
ad
ar
o
u
n
d
1
.
6
ep
is
o
d
es
o
f
m
ild
h
y
p
o
g
ly
ce
m
ia
an
d
0
.
0
2
9
ep
is
o
d
es
o
f
s
ev
er
e
h
y
p
o
g
l
y
ce
m
ia
e
v
er
y
wee
k
[
1
5
]
.
Fu
r
th
er
m
o
r
e,
am
o
n
g
p
eo
p
le
with
T
2
d
iab
etes
r
ec
eiv
in
g
tr
ea
tm
e
n
t,
th
e
f
r
eq
u
e
n
cy
o
f
s
ev
er
e
h
y
p
o
g
ly
ce
m
ia
was
ar
o
u
n
d
0
.
4
4
ep
is
o
d
es
p
er
p
er
s
o
n
p
er
y
ea
r
[
1
6
]
.
F
u
r
t
h
e
r
m
o
r
e
,
h
y
p
o
g
l
y
c
e
m
i
a
i
s
m
o
r
e
c
o
m
m
o
n
i
n
h
o
s
p
i
t
a
l
i
z
e
d
d
i
a
b
e
t
i
c
p
a
t
i
e
n
t
s
[
1
7
]
,
[
1
8
]
.
S
e
v
e
r
e
h
y
p
o
g
l
y
c
e
m
i
a
i
s
a
s
i
g
n
i
f
i
c
a
n
t
r
i
s
k
,
e
s
p
e
c
i
a
l
l
y
f
o
r
p
e
o
p
l
e
o
n
i
n
s
u
l
i
n
f
o
r
a
l
o
n
g
t
i
m
e
[
5
]
.
T
h
e
b
o
d
y
'
s
c
o
u
n
t
e
r
-
r
e
g
u
l
a
t
o
r
y
r
e
s
p
o
n
s
e
i
s
c
r
i
t
i
c
a
l
i
n
p
r
e
v
e
n
t
i
n
g
m
o
d
e
r
a
t
e
h
y
p
o
g
l
y
c
e
m
i
a
(
M
H
)
f
r
o
m
b
e
c
o
m
i
n
g
s
e
v
e
r
e
.
H
o
w
e
v
e
r
,
f
o
r
p
e
o
p
l
e
w
i
t
h
r
e
c
u
r
r
i
n
g
e
p
i
s
o
d
e
s
o
f
l
o
w
b
l
o
o
d
s
u
g
a
r
,
t
h
e
f
r
e
q
u
e
n
c
y
o
f
h
y
p
o
g
l
y
c
e
m
i
a
w
a
r
n
i
n
g
i
n
d
i
c
a
t
o
r
s
e
v
e
n
t
u
a
l
l
y
d
e
c
r
e
a
s
e
s
,
l
e
a
d
i
n
g
t
o
d
i
m
i
n
i
s
h
e
d
a
w
a
r
e
n
e
s
s
o
f
h
y
p
o
g
l
y
c
e
m
i
a
a
n
d
t
h
e
a
p
p
e
a
r
a
n
c
e
o
f
s
y
m
p
t
o
m
l
e
s
s
h
y
p
o
g
l
y
c
e
m
i
a
[
1
9
]
,
[
2
0
]
.
I
n
t
u
r
n
,
i
t
s
u
r
g
e
s
t
h
e
l
i
k
e
l
i
h
o
o
d
o
f
S
H
r
e
s
u
l
t
i
n
g
i
n
a
v
i
c
i
o
u
s
c
y
c
l
e
[
1
5
]
,
[
2
1
]
.
Hy
p
o
g
ly
ca
e
m
ia
d
u
r
in
g
a
h
o
s
p
ital
s
tay
h
as
b
ee
n
lin
k
ed
to
a
n
in
cr
ea
s
ed
r
is
k
o
f
m
o
r
tality
,
lo
n
g
e
r
h
o
s
p
ital
s
tay
s
,
m
o
r
e
p
r
o
b
lem
s
,
a
n
d
h
ig
h
er
co
s
ts
[
2
2
]
,
[
2
3
]
.
Alth
o
u
g
h
th
er
e
ar
e
h
o
s
p
ital
-
b
ased
g
ly
c
ae
m
ic
m
an
ag
em
e
n
t
s
o
f
twar
e
s
o
lu
tio
n
s
av
ailab
le,
s
u
ch
as
Glu
co
m
m
an
d
er
,
E
n
d
o
T
o
o
l,
an
d
Glu
co
tab
,
th
ey
ar
e
e
x
p
en
s
iv
e,
less
p
r
ec
is
e
f
o
r
th
e
p
r
e
d
ictio
n
o
f
h
y
p
o
g
ly
ce
m
ia,
lack
o
f
ass
ess
m
en
t
f
o
r
p
atien
ts
wh
o
a
r
e
cr
itically
il
l
,
ar
e
in
ca
p
ab
le
o
f
p
r
ed
ictin
g
g
lu
co
s
e
lev
els
f
o
r
s
ev
er
al
h
o
u
r
s
in
ad
v
an
ce
,
an
d
a
r
e
n
o
t
wid
ely
u
s
ed
[
2
4
]
–
[
2
6
]
.
I
n
clin
ical
s
ettin
g
s
,
p
r
ed
ictin
g
h
y
p
o
g
ly
ce
m
ia
is
cr
itical.
Sev
er
al
s
tu
d
ies
h
av
e
e
m
er
g
ed
in
th
e
last
ten
y
ea
r
s
,
ap
p
ly
in
g
tr
ad
itio
n
al
m
eth
o
d
o
l
o
g
ies
b
ased
o
n
p
h
y
s
i
o
lo
g
ical
a
n
d
clin
ical
p
a
r
am
eter
s
to
p
r
ed
ict
h
y
p
o
g
ly
ce
m
ia
o
cc
u
r
r
en
ce
s
.
T
y
p
ically
,
th
ese
tech
n
iq
u
es
we
r
e
tr
ain
e
d
u
s
in
g
h
is
to
r
ical
d
ata
f
r
o
m
el
ec
tr
o
n
ic
h
ea
lth
r
ec
o
r
d
s
,
in
clu
d
in
g
d
em
o
g
r
ap
h
ic
in
f
o
r
m
atio
n
,
lab
test
r
esu
lts
,
d
i
ab
etic
m
ed
icin
es,
an
d
o
th
er
p
e
r
tin
en
t in
d
icato
r
s
co
n
n
ec
ted
with
h
y
p
o
g
ly
ce
m
ia.
Fo
r
f
o
r
ec
asti
n
g
h
y
p
o
g
ly
ce
m
ia,
th
r
ee
m
eth
o
d
s
h
a
v
e
b
ee
n
p
r
o
p
o
s
ed
:
o
n
e
b
ased
o
n
p
h
y
s
io
lo
g
ical
p
r
ed
ictio
n
s
,
an
o
t
h
er
o
n
d
ata
an
aly
s
is
,
an
d
a
h
y
b
r
i
d
th
at
co
m
b
in
es
b
o
th
tech
n
iq
u
es
[
2
7
]
,
[
2
8
]
.
Nu
m
er
o
u
s
d
ata
-
ce
n
tr
ic
m
o
d
els
h
av
e
ar
is
en
with
in
th
ese
m
eth
o
d
o
l
o
g
ies,
ex
p
lo
itin
g
co
n
tin
u
o
u
s
g
lu
co
s
e
m
o
n
ito
r
in
g
(
C
GM
)
d
ata
in
co
m
b
in
atio
n
with
v
ar
i
ab
les
s
u
ch
as
d
ietar
y
in
tak
e,
p
h
y
s
ical
ac
tiv
ity
,
u
s
e
o
f
in
s
u
lin
,
an
d
o
th
er
f
ac
to
r
s
[
2
7
]
.
Nu
m
er
o
u
s
r
esear
ch
p
r
o
j
ec
ts
[
2
9
]
–
[
3
1
]
p
r
im
ar
ily
f
o
c
u
s
o
n
d
ev
elo
p
in
g
tim
e
s
er
ies
m
o
d
els
ca
p
a
b
le
o
f
f
o
r
ec
asti
n
g
BG
lev
els
.
Gallo
[
2
9
]
u
s
ed
a
f
ir
s
t
-
o
r
d
er
au
to
r
e
g
r
ess
iv
e
(
AR
)
m
o
d
el
an
d
a
f
ir
s
t
-
o
r
d
er
p
o
l
y
n
o
m
ial
m
o
d
el
f
o
r
p
r
e
d
ictin
g
BG
lev
e
ls
,
f
o
cu
s
in
g
o
n
p
r
e
d
ictio
n
in
te
r
v
als
o
f
3
0
o
r
4
5
m
i
n
u
tes.
A
n
ew
tech
n
iq
u
e
was
r
ec
en
tly
p
r
o
p
o
s
ed
th
at
co
m
b
i
n
es
an
au
to
r
eg
r
ess
iv
e
in
teg
r
a
ted
m
o
v
in
g
-
a
v
er
ag
e
(
AR
I
MA
)
tech
n
iq
u
e
with
a
p
r
o
g
r
a
m
m
ab
le
alg
o
r
ith
m
[
3
0
]
.
W
h
en
u
s
ed
to
f
o
r
ec
ast
h
y
p
o
g
ly
ce
m
ia
with
in
a
3
0
-
m
in
u
t
e,
th
is
n
o
v
el
m
o
d
el
d
is
p
lay
ed
f
aster
d
etec
tio
n
tim
es
an
d
f
ewe
r
f
alse
alar
m
s
th
a
n
co
n
v
e
n
tio
n
al
tech
n
iq
u
es
in
c
lu
d
in
g
an
ad
ap
tiv
e
u
n
iv
ar
iate
m
o
d
el
a
n
d
a
class
ic
AR
I
MA
tech
n
iq
u
e.
Nev
er
t
h
eless
,
it
is
ess
en
ti
al
to
n
o
te
th
at
th
is
ap
p
r
o
ac
h
n
ec
ess
itates
s
ig
n
if
ican
t
p
r
o
ce
s
s
in
g
r
eso
u
r
ce
s
an
d
p
r
esen
ts
is
s
u
es
f
o
r
r
ea
l
-
tim
e
im
p
lem
e
n
tatio
n
.
Sig
n
if
ica
n
t
ad
v
an
ce
s
in
m
e
d
ical
d
ata
s
cien
ce
h
av
e
b
ee
n
m
a
d
e
v
ia
m
ac
h
in
e
lear
n
in
g
tec
h
n
iq
u
es,
i
n
clu
d
in
g
th
eir
ap
p
licatio
n
in
p
r
ed
ic
tin
g
g
lu
c
o
s
e
lev
els
an
d
h
y
p
o
g
ly
ce
m
ia
[
3
2
]
.
I
n
o
n
e
o
f
th
e
r
esear
ch
s
tu
d
ies,
an
ar
tifi
cial
n
eu
r
al
n
etwo
r
k
(
ANN)
was
u
s
ed
wh
ich
in
clu
d
ed
f
i
v
e
co
n
s
ec
u
tiv
e
C
GM
o
b
s
er
v
atio
n
s
as
well
as
th
e
m
ea
s
u
r
em
en
t
tim
e
[
3
3
]
.
W
h
ile
th
e
ANN
o
u
tp
e
r
f
o
r
m
ed
tim
e
s
er
ies
m
o
d
els
s
u
ch
as
th
e
AR
m
o
d
el
in
p
r
ed
ictin
g
ac
c
u
r
a
cy
[
2
9
]
,
it
s
tr
u
g
g
le
d
to
d
etec
t
ab
r
u
p
t
v
a
r
iatio
n
s
in
BG
lev
els
in
d
u
ce
d
b
y
ev
en
ts
s
u
ch
as
m
ea
ls
o
r
in
s
u
lin
i
n
jectio
n
s
.
Desp
ite
th
is
in
cr
ea
s
ed
co
m
p
lex
ity
,
th
is
is
s
u
e
m
ay
b
e
o
v
er
c
o
m
e
b
y
ad
d
in
g
ad
d
itio
n
al
d
ata,
in
clu
d
in
g
n
u
tr
itio
n
al
in
tak
e
[
3
4
]
–
[
3
6
]
an
d
in
s
u
lin
[
3
6
]
.
Pap
p
ad
a
et
a
l
.
[
3
7
]
p
r
esen
ted
an
alter
n
ativ
e
f
ee
d
-
f
o
r
war
d
n
eu
r
al
n
etwo
r
k
m
o
d
el;
h
o
wev
e
r
,
tr
ain
in
g
th
is
m
o
d
el
r
e
q
u
ir
ed
ex
ten
s
iv
e
in
p
u
ts
,
in
clu
d
in
g
in
s
u
lin
d
o
s
es,
d
ietar
y
d
etails,
life
s
ty
le
el
em
en
ts
,
em
o
tio
n
al
f
ac
to
r
s
,
an
d
co
n
s
ec
u
tiv
e
C
GM
d
ata
p
o
in
ts
.
Asi
d
e
f
r
o
m
th
e
n
eu
r
al
n
etwo
r
k
ap
p
r
o
ac
h
,
n
u
m
er
o
u
s
attem
p
ts
to
f
o
r
ec
ast
h
y
p
o
g
ly
ce
m
ia
[
2
4
]
h
a
v
e
b
ee
n
attem
p
te
d
,
in
clu
d
in
g
t
h
e
u
s
e
o
f
d
ec
is
io
n
-
tr
ee
tech
n
i
q
u
es
[
3
8
]
,
[
3
9
]
an
d
m
ac
h
in
e
lear
n
i
n
g
tec
h
n
iq
u
es
em
p
lo
y
in
g
s
elf
-
m
o
n
ito
r
in
g
BG
d
ata
[
4
0
]
.
No
n
eth
eless
,
m
a
n
y
o
f
th
ese
p
r
ev
io
u
s
ef
f
o
r
ts
ess
en
tial
v
ar
io
u
s
m
an
u
a
l
in
p
u
ts
to
im
p
r
o
v
e
th
e
p
r
ec
is
io
n
o
f
h
y
p
o
g
ly
ce
m
ia
p
r
ed
ictio
n
.
Alth
o
u
g
h
m
u
ltip
le
r
esear
ch
p
ap
er
s
h
av
e
s
h
o
wn
s
u
cc
ess
f
u
l
m
eth
o
d
s
f
o
r
f
o
r
ec
asti
n
g
o
v
er
n
ig
h
t
h
y
p
o
g
ly
ce
m
ia,
p
r
ed
ictin
g
p
o
s
t
-
m
ea
l
h
y
p
o
g
ly
ce
m
ia
r
em
ain
s
d
if
f
icu
lt
d
u
e
to
th
e
lar
g
e
g
lu
co
s
e
v
ar
iatio
n
s
d
u
r
in
g
m
ea
l
h
o
u
r
s
.
T
h
e
m
ain
o
b
jectiv
e
o
f
u
n
s
u
p
er
v
is
ed
lear
n
i
n
g
is
to
f
in
d
u
s
ef
u
l
an
d
h
id
d
en
p
atter
n
s
f
r
o
m
th
e
u
n
k
n
o
wn
d
ata.
I
t
h
as
m
u
ch
m
o
r
e
r
esem
b
lan
ce
to
ar
tific
ial
in
tell
ig
en
ce
(
AI
)
.
Als
o
,
t
h
is
ty
p
e
o
f
lear
n
in
g
u
s
es
r
ea
l
-
tim
e
d
ata
an
aly
s
is
an
d
h
el
p
s
p
r
e
-
p
r
o
ce
s
s
in
g
.
Su
p
er
v
is
ed
lear
n
in
g
is
u
s
ed
to
tr
ain
th
e
d
ata
a
n
d
p
r
ed
ict
o
u
tp
u
t,
wh
er
ea
s
lab
eled
d
ata
is
u
s
ed
to
tr
ain
th
ese
lear
n
in
g
alg
o
r
ith
m
s
.
Acc
u
r
ate
r
esu
lts
ar
e
p
r
o
d
u
ce
d
u
s
in
g
s
u
p
er
v
is
ed
lear
n
in
g
al
g
o
r
ith
m
s
.
Su
p
er
v
is
ed
lear
n
in
g
d
o
es
n
o
t
clo
s
ely
r
ese
m
b
le
AI
.
So
,
we
h
a
v
e
also
p
r
o
p
o
s
ed
a
n
u
n
s
u
p
er
v
is
ed
lear
n
i
n
g
m
eth
o
d
o
lo
g
y
wh
er
e
u
n
lab
eled
d
ata
is
u
s
ed
t
o
tr
ain
u
n
s
u
p
er
v
is
ed
lear
n
in
g
al
g
o
r
ith
m
s
.
T
h
e
m
ain
o
b
jectiv
e
o
f
u
n
s
u
p
er
v
is
ed
lear
n
in
g
is
to
f
in
d
u
s
ef
u
l
an
d
h
id
d
e
n
p
atter
n
s
f
r
o
m
th
e
u
n
k
n
o
wn
d
at
a.
I
t
h
as
m
u
c
h
m
o
r
e
r
esem
b
la
n
ce
to
AI
.
Als
o
,
t
h
is
ty
p
e
o
f
lear
n
in
g
u
s
es
r
ea
l
-
tim
e
d
ata
an
aly
s
is
an
d
h
elp
s
in
p
r
e
-
p
r
o
ce
s
s
in
g
.
T
r
an
s
f
e
r
lear
n
in
g
m
eth
o
d
is
a
m
ac
h
in
e
lear
n
in
g
tech
n
i
q
u
e
in
wh
ich
an
ex
is
tin
g
m
o
d
el
d
e
v
elo
p
e
d
f
o
r
o
n
e
task
is
r
ep
u
r
p
o
s
ed
as
th
e
f
o
u
n
d
atio
n
f
o
r
an
o
th
er
ac
tiv
i
ty
.
T
h
is
m
eth
o
d
is
co
m
m
o
n
l
y
u
s
ed
in
d
ee
p
lea
r
n
in
g
,
wh
er
e
p
r
e
-
b
u
ilt
m
o
d
els
ar
e
t
h
e
f
o
u
n
d
atio
n
f
o
r
co
m
p
u
ter
v
is
io
n
a
n
d
n
at
u
r
al
lan
g
u
ag
e
p
r
o
ce
s
s
in
g
task
s
.
T
h
is
is
esp
ec
ially
ad
v
an
tag
eo
u
s
g
iv
en
th
e
lar
g
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
I
n
n
o
v
a
tive
ma
ch
in
e
lea
r
n
i
n
g
a
p
p
r
o
a
ch
es fo
r
p
r
ed
ictio
n
o
f h
y
p
o
g
lyce
mia
…
(
S
a
ch
i
n
R
a
mn
a
th
Ga
ikw
a
d
)
4455
co
m
p
u
tatio
n
al
r
eso
u
r
ce
s
an
d
ef
f
o
r
t
r
eq
u
ir
ed
to
cr
ea
te
n
eu
r
al
n
etwo
r
k
m
o
d
els
f
o
r
s
u
ch
i
s
s
u
es
an
d
th
e
h
u
g
e
p
er
f
o
r
m
an
ce
im
p
r
o
v
e
m
en
ts
th
ey
p
r
o
v
id
e
in
r
elate
d
d
o
m
ain
s
.
T
h
e
au
th
o
r
s
ev
alu
ated
r
elev
a
n
t
r
esear
ch
p
a
p
er
s
f
r
o
m
s
ev
er
al
d
atab
ases
u
n
d
er
t
h
e
r
eq
u
ir
em
en
ts
p
r
o
v
id
e
d
in
th
e
liter
atu
r
e
s
ec
tio
n
cr
iter
ia
(
L
SC
)
.
T
h
e
p
r
esen
t
r
esear
ch
r
ev
iewe
d
d
if
f
e
r
en
t
d
a
tab
ases
,
co
m
p
r
is
in
g
o
f
SC
OPUS,
Pu
b
Me
d
,
an
d
I
E
E
E
Xp
lo
r
e.
A
r
ticles
o
f
p
r
ev
io
u
s
r
esear
ch
f
in
d
in
g
s
ass
o
ciate
d
with
th
e
c
u
r
r
e
n
t
s
tu
d
y
.
T
h
e
s
u
r
v
ey
p
er
io
d
u
n
d
e
r
co
n
s
id
er
atio
n
f
o
r
in
clu
s
io
n
e
x
ten
d
ed
f
r
o
m
2
0
1
5
to
Au
g
u
s
t
2
0
2
3
to
g
ain
a
m
o
r
e
in
-
d
ep
th
k
n
o
wled
g
e
o
f
p
ast
s
t
u
d
y
f
in
d
i
n
g
s
.
B
ak
ar
et
a
l.
[
4
1
]
r
ep
o
r
ted
th
at
h
y
p
o
g
ly
ce
m
ia
c
an
r
esu
lt
i
n
p
atien
ts
'
d
ea
th
s
an
d
n
eu
r
o
lo
g
ical
a
b
n
o
r
m
alities
.
Ar
o
u
n
d
6
6
.
7
%
o
f
d
iab
etes
p
atien
ts
wh
o
atte
n
d
ed
ca
s
u
alty
h
ad
h
y
p
o
g
ly
ce
m
ia.
I
n
o
r
d
e
r
to
r
eg
u
late
th
e
f
r
eq
u
e
n
cy
o
f
h
y
p
o
g
ly
ce
m
ia
in
d
ia
b
etic
p
atien
ts
,
th
is
s
tu
d
y
ex
am
i
n
ed
th
e
ef
f
ec
ts
o
f
ch
a
r
ac
ter
is
tics
,
in
clu
d
in
g
o
cc
u
p
atio
n
,
ed
u
c
atio
n
,
k
n
o
wled
g
e,
g
e
n
d
er
,
an
d
ag
e.
T
h
e
o
u
tco
m
e
r
ev
ea
led
a
co
n
s
id
er
ab
le
im
p
ac
t
o
n
th
e
l
ik
elih
o
o
d
o
f
h
y
p
o
g
l
y
ce
m
ia.
Sex
,
jo
b
p
r
o
f
ile
,
ag
e
,
e
d
u
c
a
t
i
o
n
,
a
n
d
k
n
o
w
l
e
d
g
e
a
r
e
d
e
f
i
n
i
n
g
f
a
c
t
o
r
s
i
n
h
y
p
o
g
l
y
c
e
m
i
a
i
n
c
i
d
e
n
c
e
.
A
n
o
t
h
e
r
s
t
u
d
y
b
y
A
l
k
h
a
l
d
i
e
t
a
l
.
[
4
2
]
r
ep
o
r
ted
th
at
2
9
%
o
f
p
atien
ts
(
o
r
m
o
r
e
th
an
h
alf
)
r
e
p
o
r
ted
h
av
in
g
m
o
r
e
th
an
f
o
u
r
h
y
p
o
g
l
y
c
ae
m
ic
ep
is
o
d
es
am
o
n
g
ty
p
e
1
d
iab
etic
p
atien
ts
.
Nu
m
er
o
u
s
r
is
k
v
ar
iab
les
wer
e
f
o
u
n
d
,
in
clu
d
in
g
y
o
u
n
g
ag
e
,
ty
p
e
-
1
d
iab
etes,
lo
n
g
-
ter
m
d
iab
etes,
an
d
u
s
e
o
f
in
s
u
lin
.
I
n
ad
d
itio
n
,
Kalr
a
et
a
l.
[
4
3
]
r
ep
o
r
ted
th
at
th
e
r
is
k
o
f
h
y
p
o
g
ly
ce
m
ia
ty
p
ically
in
cr
ea
s
ed
af
ter
s
tr
o
n
g
g
ly
ca
em
ic
co
n
tr
o
l
attem
p
ts
.
D
iab
etes
m
o
r
tality
is
s
ix
tim
es
h
ig
h
er
i
n
p
e
o
p
le
with
s
ev
er
e
h
y
p
o
g
ly
ce
m
ia
th
a
n
th
o
s
e
with
o
u
t
it.
Fre
q
u
en
t h
y
p
o
g
ly
ce
m
ia
e
p
is
o
d
es
ca
n
im
p
ai
r
th
e
co
u
n
ter
-
r
e
g
u
lato
r
y
s
y
s
tem
.
I
t
was
co
n
clu
d
e
d
th
at
k
ey
s
tr
ateg
ies
f
o
r
m
ain
tain
in
g
g
o
o
d
s
u
g
ar
co
n
tr
o
l,
d
ec
r
ea
s
in
g
th
e
ev
en
t
o
f
h
y
p
o
g
ly
ce
m
ia,
an
d
p
r
ev
e
n
tin
g
lo
n
g
-
ter
m
co
m
p
licatio
n
s
in
cl
u
d
e
ea
r
ly
i
d
en
tific
atio
n
o
f
r
i
s
k
f
ac
to
r
s
an
d
r
eg
u
lar
c
h
ec
k
i
n
g
o
f
s
u
g
ar
lev
els.
Fo
n
s
ek
a
et
a
l.
[
4
4
]
.
s
tated
t
h
at
h
y
p
o
g
ly
ce
m
ia
b
ec
a
m
e
a
s
ig
n
i
f
ican
t iss
u
e
g
lo
b
ally
,
p
ar
ticu
la
r
ly
wh
ile
m
an
ag
i
n
g
d
iab
etic
p
atien
ts
’
an
tid
iab
etic
m
ed
icatio
n
s
.
Mo
s
t
p
atien
ts
ex
p
er
ien
ce
d
h
y
p
o
g
ly
ca
em
ic
ep
is
o
d
es
with
co
m
m
o
n
s
ig
n
s
o
f
u
n
co
n
s
cio
u
s
n
ess
,
wh
ich
ar
e
m
o
s
t
f
r
e
q
u
en
tly
o
b
s
er
v
e
d
in
p
atien
ts
r
ec
eiv
in
g
in
jectio
n
in
s
u
lin
.
T
h
e
s
tu
d
y
f
o
u
n
d
th
at
p
atien
ts
tak
in
g
in
s
u
lin
an
d
o
r
al
m
e
d
icatio
n
s
s
h
o
u
ld
b
e
in
f
o
r
m
ed
o
f
f
r
eq
u
e
n
t
h
y
p
o
g
ly
ca
em
ic
s
y
m
p
to
m
s
to
en
ab
le
ea
r
ly
d
et
ec
tio
n
an
d
tr
ea
tm
e
n
t
to
p
r
ev
e
n
t
n
eu
r
o
g
l
y
co
p
e
n
ia.
T
h
e
m
o
s
t
p
r
e
v
alen
t
ca
u
s
e
o
f
h
y
p
o
g
ly
ce
m
ia
was sk
ip
p
in
g
m
ea
ls
wh
ile
tak
in
g
in
s
u
lin
.
Yan
g
et
a
l.
[
4
5
]
m
en
tio
n
ed
th
at
h
y
p
o
g
ly
ce
m
ia
is
a
f
r
eq
u
en
t
ad
v
er
s
e
r
ea
ctio
n
to
d
iab
etes
m
ed
icatio
n
.
Au
th
o
r
s
u
s
ed
th
e
v
alid
ate
d
m
a
ch
in
e
lear
n
in
g
tech
n
iq
u
es
f
o
r
t
h
e
p
r
ed
ictio
n
o
f
h
y
p
o
g
ly
ce
m
i
a
am
o
n
g
th
e
t
y
p
e
2
d
iab
etic
p
atien
ts
.
T
h
e
em
b
ed
d
in
g
m
ac
h
in
e
lear
n
in
g
m
o
d
el
(
XGBo
o
s
t3
)
p
er
f
o
r
m
e
d
th
e
b
es
t
o
f
all
th
e
m
o
d
els
ex
am
in
ed
in
th
is
s
tu
d
y
.
Fo
r
XGBo
o
s
t,
th
e
ac
cu
r
ac
y
is
0
.
8
2
,
an
d
f
o
r
th
e
ar
ea
u
n
d
er
th
e
cu
r
v
e
(
AUC
)
,
th
e
ac
cu
r
ac
y
is
0
.
9
3
.
T
h
is
s
tu
d
y
r
ep
o
r
ted
th
at
XGBo
o
s
t
tech
n
iq
u
e
was
im
p
r
o
v
ed
b
y
v
ec
to
r
-
d
is
tr
ib
u
ted
m
em
o
r
y
tech
n
iq
u
es
th
at
co
u
ld
p
r
ec
is
ely
f
o
r
ec
ast
h
y
p
o
g
ly
ce
m
ia
am
o
n
g
d
ia
b
etic
p
atien
ts
.
Seo
et
a
l
.
[
3
2
]
attem
p
ted
to
p
r
ed
ict
h
y
p
o
g
ly
ce
m
ia
alar
m
b
y
th
ir
ty
m
in
u
te
p
r
e
d
ictio
n
win
d
o
w;
th
e
r
an
d
o
m
f
o
r
est
(
R
F)
tech
n
iq
u
es
d
ep
icted
th
e
b
etter
r
esu
lts
.
I
t
h
ad
an
A
UC
o
f
0
.
9
6
6
,
a
s
en
s
itiv
ity
o
f
0
.
8
9
6
,
a
n
d
a
s
p
ec
if
icity
o
f
0
.
9
1
3
%.
T
h
e
p
r
e
v
io
u
s
f
in
d
in
g
s
wer
e
r
ea
f
f
ir
m
ed
in
a
m
o
r
e
r
ec
en
t r
ev
iew.
Va
r
io
u
s
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
s
,
n
eu
r
al
n
etwo
r
k
s
,
an
d
s
u
p
er
v
is
ed
lear
n
in
g
a
p
p
r
o
ac
h
e
s
wer
e
s
h
o
wn
to
ac
cu
r
ately
p
r
e
d
ict
th
e
r
is
k
f
ac
to
r
s
ass
o
ciate
d
with
h
y
p
o
g
ly
ce
m
ia
ea
r
ly
,
ass
is
tin
g
in
av
o
id
in
g
co
m
p
licatio
n
s
in
d
iab
etic
p
atien
t
s
[
4
6
]
.
T
h
is
wo
r
k
u
s
es
a
m
ix
tu
r
e
o
f
f
ea
tu
r
e
m
is
ce
llan
y
s
tr
ateg
ies
to
s
o
lv
e
th
e
p
r
o
b
lem
o
f
d
ea
lin
g
with
u
n
b
alan
ce
d
d
atasets
.
I
t
ev
al
u
ates
th
e
ef
f
icac
y
o
f
v
ar
io
u
s
s
u
p
er
v
is
ed
a
n
d
u
n
s
u
p
e
r
v
is
ed
l
ea
r
n
in
g
m
o
d
els.
I
t
in
v
esti
g
ates
th
e
v
alu
e
o
f
tr
a
n
s
f
er
lear
n
in
g
ap
p
r
o
ac
h
es
s
u
ch
as
k
-
m
ea
n
s
r
eg
r
ess
io
n
an
d
class
if
icatio
n
in
p
r
ed
ictin
g
t
y
p
e
2
d
iab
etic
h
y
p
o
g
ly
ce
m
ia.
On
th
e
o
th
er
h
a
n
d
,
p
r
ev
io
u
s
r
esear
ch
s
tu
d
ies
f
o
cu
s
ed
m
o
s
tly
o
n
d
iv
id
in
g
b
alan
ce
d
d
atasets
,
an
d
s
o
m
e
s
tu
d
ies
f
o
cu
s
s
ed
o
n
d
ev
elo
p
in
g
ef
f
ec
tiv
e
class
if
i
ca
tio
n
o
r
r
eg
r
ess
io
n
m
o
d
els o
f
m
ac
h
in
e
lear
n
i
n
g
m
eth
o
d
s
f
o
r
p
r
ed
ictin
g
h
y
p
o
g
ly
ce
m
ia
.
2.
M
E
T
H
O
D
T
h
is
wo
r
k
p
r
o
p
o
s
ed
an
in
n
o
v
ativ
e
m
eth
o
d
o
l
o
g
y
f
o
r
h
y
p
o
g
ly
ce
m
ia
p
r
ed
ictio
n
in
t
y
p
e
2
d
iab
etes
d
ep
lo
y
in
g
m
ac
h
i
n
e
lear
n
i
n
g
t
ec
h
n
iq
u
es.
Firstl
y
,
r
esear
ch
e
r
s
u
s
ed
s
u
p
er
v
is
ed
m
ac
h
in
e
le
ar
n
in
g
m
eth
o
d
s
in
wh
ich
r
eg
r
ess
io
n
an
d
class
if
icatio
n
tech
n
iq
u
es
wer
e
u
s
ed
.
S
ec
o
n
d
ly
,
a
n
u
n
s
u
p
er
v
is
ed
lear
n
in
g
m
eth
o
d
u
s
in
g
k
-
m
ea
n
clu
s
ter
in
g
was
u
s
ed
f
o
r
d
ata
an
aly
s
is
.
Fig
u
r
e
1
d
ep
icts
t
h
e
s
tep
s
th
at
wer
e
u
s
ed
in
th
e
p
r
o
p
o
s
ed
m
eth
o
d
o
l
o
g
y
f
o
r
p
r
ed
ictin
g
h
y
p
o
g
ly
ce
m
ia
i
n
ty
p
e
2
d
iab
etes.
2
.
1
.
Da
t
a
c
o
llect
io
n a
nd
da
t
a
pre
-
pro
ce
s
s
ing
T
h
e
d
ata
u
s
ed
in
th
e
p
r
o
p
o
s
ed
m
o
d
el
is
r
ea
l
-
tim
e
d
ata
co
llected
f
r
o
m
p
a
r
ticip
an
ts
b
y
p
er
s
o
n
al
in
ter
v
iews
with
th
e
p
atien
ts
a
d
m
itted
to
th
e
h
o
s
p
ital.
Ma
in
l
y
,
th
ese
q
u
esti
o
n
n
air
es
i
n
clu
d
ed
r
is
k
f
ac
to
r
s
th
at
d
ir
ec
tly
im
p
ac
t
ca
u
s
in
g
h
y
p
o
g
ly
ce
m
ia.
T
h
ese
d
ata
o
f
r
is
k
f
ac
to
r
s
f
o
r
th
e
p
r
e
d
ictio
n
o
f
h
y
p
o
g
ly
ce
m
ia.
T
h
e
d
ata
was
n
o
r
m
alize
d
u
s
in
g
a
lab
el
en
co
d
in
g
tech
n
iq
u
e,
u
n
d
e
r
w
h
ich
d
ata
was
co
n
v
er
ted
in
to
n
u
m
b
er
s
,
an
d
ea
ch
f
ac
to
r
was
s
co
r
ed
with
0
o
r
1
f
o
r
ev
er
y
tr
u
e
an
d
f
alse
v
alu
e.
Du
e
to
th
is
d
ata
n
o
r
m
aliza
tio
n
,
th
e
cu
r
r
en
t
m
o
d
el
co
u
ld
p
e
r
f
o
r
m
ac
cu
r
ate
p
r
ed
ic
tio
n
s
with
o
u
t o
v
e
r
f
ee
d
in
g
an
d
o
v
er
f
itti
n
g
d
ata.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
13
,
No
.
4
,
Dec
em
b
er
2
0
2
4
:
4
4
5
3
-
4
4
7
1
4456
No
n
e
o
f
th
e
c
o
lu
m
n
s
h
a
d
a
lar
g
e
in
f
lu
en
ce
d
u
r
in
g
p
r
ed
ictio
n
s
.
T
h
e
r
ea
l
-
d
ata
s
et
co
n
tain
e
d
3
0
3
s
am
p
les
an
d
1
8
co
lu
m
n
s
with
n
o
m
is
s
in
g
v
alu
es,
o
f
wh
ich
1
6
h
ad
ca
t
eg
o
r
ical
v
alu
es.
T
h
is
d
ata
ea
s
ily
f
it th
e
r
eg
r
ess
io
n
m
o
d
el,
wh
ile
th
e
class
if
icatio
n
m
o
d
el
n
ee
d
e
d
a
ca
teg
o
r
ical
tar
g
et
v
ar
iab
le.
T
h
e
d
ata
was
g
r
o
u
p
ed
in
to
two
ca
teg
o
r
ies
b
ased
o
n
th
e
m
ed
ia
n
o
f
th
e
b
elo
w
s
u
g
ar
lev
el
(
B
SL
)
v
alu
e.
T
h
e
ca
teg
o
r
ies
o
f
g
r
o
u
p
s
wer
e
0
an
d
1
,
m
ak
in
g
th
ese
g
r
o
u
p
s
a
ca
teg
o
r
ical
tar
g
et
v
ar
iab
le
an
d
s
k
ip
p
in
g
th
e
B
SL
v
alu
e.
Af
ter
t
h
is
,
d
ata
was
f
it
f
o
r
th
e
class
if
icatio
n
m
o
d
el
p
r
ed
ictio
n
.
Fig
u
r
e
1
.
Me
th
o
d
o
lo
g
y
u
s
ed
i
n
cu
r
r
e
n
t stu
d
y
f
o
r
p
r
ed
ictio
n
o
f
h
y
p
o
g
l
y
ce
m
ia
in
ty
p
e
-
2
d
ia
b
etes
2
.
2
.
Su
perv
is
ed
a
nd
u
ns
up
e
rv
is
ed
m
a
chine le
a
rning
t
ec
hn
iq
ues
f
o
r
predict
io
n o
f
hy
po
g
ly
ce
m
ia
R
eg
r
ess
io
n
is
a
s
u
p
er
v
is
ed
m
ac
h
in
e
lear
n
in
g
tech
n
iq
u
e
f
o
r
p
r
e
d
ictio
n
o
f
a
co
n
tin
u
o
u
s
n
u
m
er
i
ca
l
o
u
tp
u
t
o
r
tar
g
et
v
ar
ia
b
le
b
ased
o
n
o
n
e
o
r
m
o
r
e
in
p
u
ts
.
I
n
th
e
cu
r
r
en
t
s
tu
d
y
,
th
e
tar
g
et
v
ar
ia
b
le
was
th
e
n
u
m
er
ical
v
al
u
e
o
f
B
SL.
T
h
u
s
,
t
h
e
p
r
ed
ictio
n
tech
n
iq
u
e
was
co
m
p
atib
le
wit
h
m
o
d
el
d
ev
elo
p
m
en
t,
tr
ain
in
g
,
an
d
p
r
ed
ictio
n
s
.
C
las
s
if
icatio
n
is
also
a
s
u
p
er
v
is
ed
m
ac
h
in
e
lear
n
in
g
tech
n
iq
u
e
u
s
ed
t
o
p
r
ed
ict
ca
teg
o
r
ical
d
ata.
T
h
e
d
ata
was
o
r
g
an
ized
a
n
d
p
r
e
-
p
r
o
ce
s
s
ed
in
to
ca
teg
o
r
ical
tar
g
ets
an
d
t
h
en
u
s
ed
f
o
r
f
u
r
th
e
r
tr
ain
in
g
a
n
d
p
r
e
d
ictio
n
s
.
T
h
e
k
-
m
ea
n
clu
s
ter
in
g
u
n
s
u
p
er
v
is
ed
m
ac
h
i
n
e
lear
n
in
g
m
eth
o
d
was
u
s
ed
f
o
r
d
ata
an
aly
s
is
an
d
to
f
in
d
th
e
p
atter
n
s
in
d
ata
.
2
.
2
.
1
.
Su
perv
is
ed
lea
rning
u
s
ing
re
g
re
s
s
io
n a
nd
cla
s
s
if
ica
t
io
n
In
s
u
p
er
v
is
ed
lear
n
i
n
g
u
s
in
g
r
eg
r
ess
io
n
m
o
d
el,
s
ev
en
s
tatis
tical
alg
o
r
ith
m
s
f
o
r
p
r
ed
ictio
n
a
n
d
r
esu
lts
h
av
e
b
ee
n
u
s
ed
.
Fig
u
r
e
2
s
h
o
ws
th
e
p
r
o
p
o
s
ed
wo
r
k
f
lo
w
f
o
r
p
r
e
d
ictin
g
h
y
p
o
g
ly
ce
m
ia
t
y
p
e
-
2
d
iab
etes
u
s
in
g
Or
an
g
e
d
ata
m
i
n
in
g
to
o
l.
I
n
S
u
p
er
v
is
ed
lear
n
in
g
u
s
in
g
class
if
icatio
n
m
o
d
el,
t
h
e
d
ata
was
c
ateg
o
r
ized
in
to
two
g
r
o
u
p
s
as
th
is
r
ea
l
-
tim
e
d
ataset
h
ad
th
e
p
r
o
b
lem
o
f
u
n
b
alan
ce
d
d
ata.
I
n
tr
ials
,
th
e
f
ir
s
t
d
if
f
icu
lty
was
ca
teg
o
r
izin
g
th
e
d
ata
in
a
b
al
an
ce
d
m
a
n
n
er
.
T
h
e
class
im
b
alan
ce
was
th
e
m
ajo
r
co
n
ce
r
n
d
u
r
in
g
th
e
m
o
d
el
d
ev
elo
p
m
e
n
t.
I
t
h
as b
ee
n
ad
d
r
ess
ed
b
y
ex
er
cisi
n
g
v
ar
io
u
s
tr
ials
,
in
clu
d
in
g
f
ea
tu
r
es,
clu
s
ter
i
n
g
,
an
d
d
escr
ip
tiv
e
s
tatis
t
ics.
Af
ter
m
an
y
tr
ials
an
d
er
r
o
r
s
,
in
v
esti
g
ato
r
s
f
o
u
n
d
t
h
e
s
o
lu
tio
n
o
f
ca
teg
o
r
izi
n
g
th
e
d
ata
b
ased
o
n
th
e
m
ed
ian
o
f
th
e
B
SL
v
a
lu
es,
w
h
ich
was
6
7
m
g
/
d
l.
T
h
e
d
ata
was
d
is
tr
ib
u
ted
in
two
b
alan
c
ed
class
es:
g
r
o
u
p
1
co
n
s
is
ted
o
f
5
0
.
6
7
%
s
am
p
les,
an
d
g
r
o
u
p
2
h
ad
4
9
.
3
3
%.
W
ith
th
is
g
r
o
u
p
in
g
,
in
v
esti
g
ato
r
s
h
ad
ap
p
lied
d
if
f
e
r
en
t
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
I
n
n
o
v
a
tive
ma
ch
in
e
lea
r
n
i
n
g
a
p
p
r
o
a
ch
es fo
r
p
r
ed
ictio
n
o
f h
y
p
o
g
lyce
mia
…
(
S
a
ch
i
n
R
a
mn
a
th
Ga
ikw
a
d
)
4457
s
tatis
t
ical
alg
o
r
ith
m
s
,
wh
ich
was
n
o
t
p
o
s
s
ib
le
in
o
u
r
m
ain
d
ata
d
u
e
to
th
e
n
u
m
er
ic
tar
g
et
v
ar
iab
le.
T
h
u
s
,
g
r
o
u
p
1
h
ad
a
B
SL
r
an
g
e
o
f
5
7
m
g
/d
l to
6
7
m
g
/d
l,
a
n
d
g
r
o
u
p
2
h
a
d
a
6
8
m
g
/
d
l r
an
g
e
o
f
7
9
m
g
/d
l.
W
e
r
em
o
v
ed
th
e
B
SL
v
alu
e
v
a
r
iab
le
f
r
o
m
o
u
r
d
ata
s
et
an
d
ad
d
ed
an
o
th
er
v
ar
iab
le
n
a
m
ed
g
r
o
u
p
,
w
h
ich
is
s
elec
ted
as
th
e
tar
g
et
v
ar
iab
le.
Un
d
er
th
is
class
if
icatio
n
p
r
e
d
ictio
n
,
ten
s
tatis
tical
alg
o
r
ith
m
s
h
av
e
b
ee
n
u
s
ed
.
Fig
u
r
e
3
s
h
o
ws
t
h
e
alg
o
r
ith
m
s
u
s
ed
:
A
d
aBo
o
s
t,
g
r
a
d
ien
t b
o
o
s
tin
g
(
GB
)
,
lo
g
i
s
tic
r
eg
r
ess
io
n
(
L
R
)
,
n
aïv
e
B
ay
es
(
NB
)
,
k
-
n
ea
r
est
n
eig
h
b
o
u
r
s
(
k
NN)
,
RF
,
n
eu
r
al
n
etwo
r
k
,
C
N2
r
u
le
in
d
u
ctio
n
,
d
ec
is
io
n
t
r
ee
(
DT
)
,
an
d
s
ch
o
h
asti
c
g
r
ad
ien
t
d
escen
t
(
SGD)
.
W
e
tr
ain
ed
th
e
d
ata
a
n
d
ca
m
e
u
p
wi
th
a
p
r
ed
ictio
n
m
o
d
el.
Fig
u
r
e
2
.
W
o
r
k
f
lo
w
d
iag
r
am
f
o
r
r
eg
r
ess
io
n
p
r
ed
ictio
n
o
f
h
y
p
o
g
ly
ce
m
ia
i
n
ty
p
e
-
2
d
iab
etes
Fig
u
r
e
3
.
W
o
r
k
f
lo
w
d
iag
r
am
f
o
r
class
if
icatio
n
p
r
ed
ictio
n
o
f
h
y
p
o
g
ly
ce
m
ia
in
t
y
p
e
-
2
d
iab
et
es
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
13
,
No
.
4
,
Dec
em
b
er
2
0
2
4
:
4
4
5
3
-
4
4
7
1
4458
2
.
2
.
2
.
Uns
up
er
v
is
e
d lea
rning
us
ing
k
-
m
ea
ns
K
-
m
ea
n
s
clu
ter
in
g
m
o
d
el
is
co
m
m
o
n
ly
u
s
ed
wh
en
th
e
d
ata
t
o
b
e
g
r
o
u
p
e
d
with
s
im
ilar
ch
ar
a
cter
is
tics
.
T
h
is
m
o
d
el
clu
s
ter
th
e
d
ataset
in
to
K
d
is
tin
ct,
n
o
n
-
o
v
e
r
lap
p
i
n
g
s
u
b
g
r
o
u
p
s
o
r
clu
s
ter
s
in
a
s
im
p
le
an
d
ef
f
icien
t
m
an
n
er
.
T
h
e
ce
n
tr
o
i
d
o
f
ea
c
h
clu
s
ter
is
d
er
iv
ed
as
th
e
av
er
ag
e
o
f
all
th
e
d
ata
p
o
in
ts
with
in
th
at
clu
s
ter
.
No
n
eth
eless
,
K
-
m
ea
n
s
h
as
s
o
m
e
lim
itatio
n
s
,
in
clu
d
in
g
as
its
s
en
s
itiv
ity
to
o
u
tlier
s
an
d
i
ts
ass
u
m
p
tio
n
th
at
clu
s
ter
s
ar
e
ess
en
tially
s
p
h
er
ical
an
d
o
f
e
q
u
al
s
izes,
wh
ich
m
ay
n
o
t n
ec
ess
ar
ily
co
r
r
e
s
p
o
n
d
to
th
e
co
m
p
le
x
ities
o
f
r
ea
l
-
wo
r
l
d
d
ata.
B
elo
w
is
th
e
d
iag
r
am
s
h
o
win
g
th
e
clu
s
t
er
s
f
o
r
m
ed
o
n
th
e
d
ata
u
s
ed
.
K
-
m
ea
n
clu
s
ter
in
g
ca
teg
o
r
izes
th
e
d
ata
in
to
3
clu
s
ter
s
as
s
h
o
wn
in
F
ig
u
r
e
4
.
T
h
e
d
is
tr
ib
u
tio
n
o
f
s
am
p
les
in
C
1
,
C
2
,
an
d
C
3
a
r
e
7
3
,
1
0
3
,
an
d
1
2
7
,
r
esp
ec
tiv
ely
.
A
ll
th
e
clu
s
ter
s
ar
e
d
is
tr
ib
u
ted
ac
co
r
d
in
g
to
s
im
ilar
p
atter
n
s
i
n
th
e
d
ata
f
o
r
ea
s
y
an
aly
s
is
.
T
h
e
clu
s
ter
s
u
s
ed
in
t
h
is
s
tu
d
y
wer
e
th
e
s
am
e
s
ize
b
e
ca
u
s
e
th
ey
u
s
ed
s
o
m
e
s
im
ila
r
ities
an
d
co
r
r
elatio
n
s
in
th
e
d
ata
s
et.
I
n
v
esti
g
ato
r
s
tr
ied
to
f
o
r
m
m
o
r
e
clu
s
ter
s
.
Ho
wev
er
,
th
r
ee
cl
u
s
ter
s
o
n
ly
f
o
r
m
ed
as
it
was
g
iv
in
g
b
etter
r
esu
lts
.
W
h
en
th
ese
cl
u
s
ter
s
ar
e
p
r
o
v
id
ed
to
th
e
p
r
ed
ictio
n
m
o
d
el,
g
o
o
d
ac
cu
r
a
cy
was
o
b
tain
ed
as
r
esear
ch
er
s
r
ed
u
ce
d
th
e
m
ac
h
i
n
e'
s
ef
f
o
r
t
an
d
m
a
d
e
th
e
d
ata
p
r
e
-
r
ea
d
y
f
o
r
t
h
e
p
r
ed
ictio
n
to
o
l
s
o
th
at
th
e
m
ac
h
in
e
f
o
cu
s
es
m
o
r
e
o
n
g
ettin
g
r
esu
lts
r
ath
er
th
an
s
o
r
tin
g
an
d
th
e
n
p
r
ed
ictin
g
.
T
h
is
clu
s
ter
in
g
m
eth
o
d
also
h
el
p
ed
an
aly
ze
th
e
d
ata
b
ec
au
s
e
it
co
u
ld
m
ak
e
b
u
n
c
h
es
o
f
d
ata
with
s
o
m
e
s
im
ilar
ities
an
d
co
r
r
elati
o
n
s
,
m
ak
in
g
it
ea
s
y
to
in
ter
p
r
et.
Du
e
to
t
h
is
clu
s
ter
in
g
,
we
co
u
ld
ea
s
ily
id
e
n
tify
th
e
s
ig
n
if
ican
t
co
lu
m
n
s
an
d
f
in
d
s
o
m
e
co
r
r
elatio
n
s
,
wh
ich
was v
er
y
h
el
p
f
u
l in
u
n
d
er
s
tan
d
in
g
th
e
d
ata
.
Fig
u
r
e
4
.
Data
u
n
d
er
s
tan
d
i
n
g
u
s
in
g
k
-
m
ea
n
s
clu
s
ter
in
g
2
.
3
.
P
r
o
po
s
ed
t
ra
ns
f
er
lea
rning
m
et
ho
d f
o
r
i
m
pro
v
ing
pr
edict
io
n a
cc
ura
cy
I
n
th
is
wo
r
k
,
r
esear
ch
e
r
s
f
o
u
n
d
th
e
p
r
ed
ictio
n
ac
cu
r
ac
y
is
s
u
es
an
d
im
p
r
o
v
em
e
n
t
o
n
r
ea
l
-
wo
r
ld
d
ata.
T
h
ese
is
s
u
es
wer
e
r
eso
lv
ed
b
y
ap
p
ly
in
g
th
e
tr
an
s
f
er
lear
n
in
g
m
eth
o
d
,
as
s
h
o
wn
in
Fig
u
r
e
5
.
T
r
an
s
f
er
lear
n
in
g
was
b
en
ef
icial
f
o
r
a
s
m
all
d
a
taset,
s
o
it
wa
s
s
u
itab
le
f
o
r
p
r
ed
ictio
n
m
o
d
els
an
d
d
ata
ev
alu
atio
n
.
B
o
th
th
e
p
r
ed
ictio
n
m
o
d
els
-
r
eg
r
ess
io
n
an
d
class
if
icatio
n
wer
e
u
s
ed
in
th
is
tech
n
iq
u
e.
T
h
e
d
ata
was
tr
an
s
f
er
r
ed
to
p
r
ed
icto
r
s
with
in
ea
ch
cl
u
s
ter
in
th
e
r
e
g
r
ess
io
n
m
o
d
el.
W
h
er
ea
s
,
in
th
e
class
if
icatio
n
tech
n
iq
u
e,
th
e
d
ata
was
class
if
ied
with
in
ea
ch
clu
s
ter
as st
ated
in
th
e
class
if
icatio
n
tech
n
iq
u
e.
2
.
3
.
1
.
Reg
re
s
s
io
n us
ing
t
ra
ns
f
er
lea
rning
m
et
ho
d
T
r
an
s
f
er
lear
n
in
g
is
an
m
ac
h
in
e
lear
n
in
g
m
o
d
el
in
wh
ich
a
m
o
d
el
is
tr
ain
ed
o
n
o
n
e
task
an
d
f
in
e
-
tu
n
e
d
f
o
r
a
r
elate
d
task
.
I
n
s
tead
o
f
t
r
a
in
in
g
f
r
o
m
s
cr
atch
,
th
e
r
esear
ch
er
s
p
r
e
-
tr
ain
ed
th
e
m
o
d
el
th
at
h
as
alr
ea
d
y
lear
n
e
d
u
s
ef
u
l
f
ea
tu
r
es,
an
d
th
e
h
y
p
e
r
p
ar
am
eter
s
o
f
all
th
e
alg
o
r
it
h
m
s
u
s
ed
ar
e
also
f
in
ely
tu
n
ed
.
T
h
e
p
r
o
p
o
s
ed
r
eg
r
ess
io
n
tech
n
iq
u
e
was
u
s
e
d
an
d
p
ass
ed
o
n
to
all
th
r
ee
cl
u
s
ter
s
(
C
1
,
C
2
,
an
d
C
3
)
s
ep
a
r
ately
an
d
ev
alu
ated
p
r
ed
ictio
n
r
esu
lts
.
All
th
e
alg
o
r
ith
m
s
wer
e
tr
ain
e
d
,
a
n
d
t
h
e
n
th
e
b
elo
w
wo
r
k
f
lo
w
was
u
s
ed
f
o
r
th
e
p
r
ed
ictio
n
m
o
d
el,
as sh
o
wn
in
Fig
u
r
e
5
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
I
n
n
o
v
a
tive
ma
ch
in
e
lea
r
n
i
n
g
a
p
p
r
o
a
ch
es fo
r
p
r
ed
ictio
n
o
f h
y
p
o
g
lyce
mia
…
(
S
a
ch
i
n
R
a
mn
a
th
Ga
ikw
a
d
)
4459
Fig
u
r
e
5
.
W
o
r
k
f
lo
w
f
o
r
r
eg
r
ess
io
n
u
s
in
g
tr
an
s
f
er
lea
r
n
in
g
2
.
3
.
2
.
Cla
s
s
if
ica
t
io
n us
ing
t
ra
ns
f
er
lea
rning
C
las
s
if
icatio
n
o
f
d
ata
was
p
er
f
o
r
m
ed
i
n
ea
ch
o
f
th
e
clu
s
ter
s
f
o
r
m
ed
u
s
in
g
k
-
m
ea
n
s
.
T
h
e
s
am
p
les
in
th
e
d
ata
ar
e
g
r
o
u
p
e
d
in
t
o
1
an
d
2
an
d
th
e
n
t
r
an
s
f
er
r
e
d
to
m
a
ch
in
e
lear
n
in
g
alg
o
r
ith
m
.
Fig
u
r
e
6
illu
s
tr
ates
th
e
wo
r
k
f
lo
w
f
o
r
class
if
icatio
n
u
s
in
g
tr
an
s
f
er
lear
n
in
g
.
C
lu
s
ter
1
,
clu
s
ter
2
,
an
d
clu
s
ter
3
ar
e
p
r
o
v
id
e
d
s
ep
ar
ately
g
en
er
ated
u
s
in
g
k
-
m
ea
n
.
E
ac
h
clu
s
ter
was
th
en
tr
an
s
f
er
r
e
d
to
th
e
class
if
icatio
n
p
r
ed
ict
io
n
m
o
d
el
a
n
d
th
en
tr
ain
ed
,
with
p
r
ed
ictio
n
r
esu
lts
ev
alu
ated
.
Fig
u
r
e
6
.
W
o
r
k
f
lo
w
f
o
r
class
if
icatio
n
u
s
in
g
tr
an
s
f
er
lear
n
in
g
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
13
,
No
.
4
,
Dec
em
b
er
2
0
2
4
:
4
4
5
3
-
4
4
7
1
4460
3.
RE
SU
L
T
S
T
h
e
B
SL
v
alu
e
in
m
g
/
d
l
in
n
u
m
er
ic
d
ata
was
co
n
s
id
er
e
d
t
h
e
tar
g
et
v
ar
ia
b
le.
All
3
0
0
s
am
p
les
wer
e
u
s
ed
to
t
r
ain
th
e
m
ac
h
in
e
lear
n
in
g
m
o
d
el,
a
n
d
s
tatis
tical
alg
o
r
ith
m
s
wer
e
f
in
el
y
tu
n
ed
a
n
d
u
s
ed
f
o
r
th
is
p
u
r
p
o
s
e.
T
h
e
p
r
e
d
ictio
n
m
o
d
el
was
ev
alu
ated
u
s
in
g
5
-
f
o
ld
cr
o
s
s
-
v
alid
atio
n
f
o
r
p
r
ec
is
e
m
o
d
el
t
r
ain
in
g
.
W
e
o
b
tain
m
ax
im
u
m
p
er
f
o
r
m
an
ce
o
f
o
u
r
m
o
d
el
in
5
f
o
ld
s
o
f
test
in
g
,
an
d
th
en
th
is
m
o
d
el
was u
s
ed
f
o
r
p
r
ed
ictio
n
s
.
3
.
1
.
I
np
ut
da
t
a
T
h
er
e
wer
e
a
to
tal
o
f
1
8
f
ea
tu
r
es
in
clu
d
ed
i
n
th
e
d
ataset.
T
h
er
e
ar
e
th
r
ee
f
ea
tu
r
es
b
ased
o
n
th
e
r
o
le:
in
p
u
t,
o
u
tp
u
t,
a
n
d
s
k
ip
p
ed
.
T
h
e
s
am
p
le
f
ea
t
u
r
e
was
s
k
ip
p
ed
;
B
SL
is
th
e
tar
g
et,
a
n
d
th
e
r
e
m
ain
in
g
f
ea
tu
r
es
a
r
e
th
e
in
p
u
t,
as
p
r
esen
ted
in
T
ab
l
e
1
.
Fig
u
r
e
7
s
h
o
ws
th
e
d
ata
im
b
alan
ce
.
T
h
e
m
a
x
im
u
m
f
r
e
q
u
en
cy
o
f
p
ar
ticip
an
ts
was in
th
e
r
an
g
e
o
f
B
SL
6
8
m
g
/d
l to
7
0
m
g
/d
l.
T
h
is
im
b
alan
ce
d
d
ata
d
is
tr
ib
u
tio
n
af
f
ec
ts
th
e
p
r
ed
ictio
n
r
esu
lts
as
m
o
r
e
n
u
m
b
e
r
s
ar
e
b
etwe
en
6
8
an
d
7
0
.
As
a
r
esu
lt
o
f
t
h
is
,
th
e
m
ax
im
u
m
ac
cu
r
ac
y
o
b
tain
ed
is
1
8
%.
A
class
if
icatio
n
tech
n
iq
u
e
was
p
r
o
p
o
s
ed
t
o
r
eso
lv
e
th
is
is
s
u
e
a
n
d
o
b
tain
b
etter
p
r
ed
ictio
n
r
esu
lts
.
Fig
u
r
e
8
d
i
p
ict
th
e
d
ata
d
is
tr
ib
u
tio
n
f
o
r
B
SL
in
a
m
o
r
e
d
is
cr
ete
a
n
d
i
n
f
o
r
m
ativ
e
f
o
r
m
.
Mo
s
t
p
a
r
ticip
an
ts
ar
e
6
6
≤
B
SL
≤
7
2
,
6
2
.
6
7
%,
a
n
d
th
e
r
em
ai
n
in
g
is
3
7
.
3
0
%.
Fu
r
t
h
er
,
we
tr
y
to
u
n
d
er
s
tan
d
th
e
ca
u
s
es
b
eh
in
d
h
y
p
o
g
ly
ce
m
ia
ty
p
e
2
d
iab
etes.
T
h
e
m
ajo
r
ca
u
s
e
in
cl
u
d
es
u
s
e
o
f
in
s
u
lin
(
6
6
%),
p
o
o
r
d
iet
in
tak
e
(
6
6
%),
g
astric
p
r
o
b
lem
s
(
5
8
%),
a
n
d
in
cr
ea
s
ed
ex
er
cise
(
6
4
%).
Oth
er
ca
u
s
es
ar
e
ad
v
an
cin
g
a
g
e
(
5
0
%),
s
k
ip
p
in
g
m
ea
ls
(
4
4
%),
u
n
u
s
u
al
weig
h
t
lo
s
s
(
3
6
%),
d
iab
etes
o
v
er
a
lo
n
g
ti
m
e
(
3
6
%),
an
tid
iab
etic
m
ed
ici
n
es
(
3
4
%),
co
n
s
u
m
p
ti
o
n
o
f
al
co
h
o
l
(
3
4
%),
u
s
e
o
f
s
am
e
in
jectio
n
s
ite
f
r
eq
u
en
tly
(
3
6
%),
s
o
m
e
o
f
th
em
h
ad
im
p
air
ed
liv
er
/k
id
n
ey
(
1
8
%)
an
d
p
r
eg
n
an
t
(
1
7
%).
T
ab
le
1
.
L
is
t o
f
v
ar
iab
les u
s
ed
in
d
ata
S
r
.
N
o
.
Li
st
o
f
V
a
r
i
a
b
l
e
s
Ty
p
e
o
f
d
a
t
a
F
e
a
t
u
r
e
s
1
U
si
n
g
i
n
su
l
i
n
,
u
s
i
n
g
a
n
t
i
d
i
a
b
e
t
i
c
me
d
i
c
i
n
e
,
i
m
p
a
i
r
e
d
l
i
v
e
r
/
k
i
d
n
e
y
,
d
i
a
b
e
t
e
s
l
o
n
g
t
i
me
,
i
n
c
r
e
a
s
e
d
e
x
e
r
c
i
se
,
l
a
c
k
o
f
k
n
o
w
l
e
d
g
e
a
b
o
u
t
h
y
p
o
g
l
y
c
e
mi
a
,
c
o
n
s
u
m
p
t
i
o
n
o
f
a
l
c
o
h
o
l
,
p
o
o
r
d
i
e
t
a
r
y
i
n
t
a
k
e
,
s
k
i
p
p
i
n
g
me
a
l
s
,
a
d
v
a
n
c
i
n
g
a
g
e
,
u
n
u
s
u
a
l
w
e
i
g
h
t
l
o
ss,
t
a
k
i
n
g
b
e
t
a
b
l
o
c
k
e
r
s
u
sa
g
e
,
u
se
o
f
s
a
me
i
n
j
e
c
t
i
o
n
si
t
e
f
r
e
q
u
e
n
t
l
y
,
t
h
y
r
o
i
d
p
r
o
b
l
e
m
s,
g
a
st
r
i
c
p
r
o
b
l
e
ms
a
n
d
p
r
e
g
n
a
n
c
y
C
a
t
e
g
o
r
i
c
a
l
I
n
p
u
t
2
S
a
mp
l
e
N
u
meri
c
S
k
i
p
p
e
d
3
B
S
L
v
a
l
u
e
(
mg
/
d
l
)
N
u
meri
c
Ta
r
g
e
t
Fig
u
r
e
7
.
T
h
e
d
is
tr
ib
u
tio
n
g
r
a
p
h
f
o
r
B
SL
v
alu
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
I
n
n
o
v
a
tive
ma
ch
in
e
lea
r
n
i
n
g
a
p
p
r
o
a
ch
es fo
r
p
r
ed
ictio
n
o
f h
y
p
o
g
lyce
mia
…
(
S
a
ch
i
n
R
a
mn
a
th
Ga
ikw
a
d
)
4461
Fig
u
r
e
8
.
Scatter
p
l
o
t f
o
r
th
e
B
SL
v
alu
e
v
s
f
r
eq
u
en
c
y
o
f
s
am
p
les
3
.
2
.
Reg
re
s
s
io
n mo
dels
T
h
e
r
eg
r
ess
io
n
m
o
d
el
f
o
r
p
r
ed
ictio
n
was
d
ev
elo
p
ed
u
s
i
n
g
v
ar
io
u
s
s
tatis
tical
a
lg
o
r
ith
m
s
u
n
d
er
s
ec
tio
n
2
.
2
.
1
.
C
r
o
s
s
v
alid
atio
n
m
eth
o
d
was
d
ep
lo
y
e
d
to
tr
a
in
an
d
test
th
e
p
e
r
f
o
r
m
an
ce
o
f
p
r
ed
ictio
n
m
o
d
el
.
Dif
f
er
en
t
K
-
f
o
ld
s
wer
e
tr
ied
an
d
test
ed
u
n
d
er
cr
o
s
s
v
alid
at
io
n
.
B
etter
p
r
e
d
ictio
n
r
esu
lts
wer
e
o
b
tain
ed
with
5
-
f
o
ld
s
o
f
tr
ain
in
g
an
d
test
in
g
f
o
r
th
e
in
p
u
t
d
ata.
Fig
u
r
e
9
s
h
o
ws
th
at
th
e
p
r
ed
ictio
n
m
o
d
el
is
ev
alu
ated
in
5
-
f
o
ld
cr
o
s
s
-
v
alid
atio
n
,
ea
c
h
f
o
ld
co
n
tain
in
g
2
0
%
s
am
p
les.
Fu
r
th
er
,
t
h
e
f
o
ld
-
wis
e
d
is
tr
ib
u
t
io
n
o
f
th
e
s
am
p
les
was a
n
aly
ze
d
.
I
t
is
o
b
s
er
v
ed
th
at
th
e
m
ajo
r
ity
o
f
th
e
p
a
r
ticip
a
n
ts
b
elo
n
g
to
6
8
≤
B
SL
≤
7
0
ex
ce
p
t f
o
ld
-
3
.
Fo
l
d
-
1
co
n
s
is
ts
o
f
2
3
.
3
3
%
o
f
p
ar
ticip
a
n
ts
,
f
o
ld
-
2
co
n
s
is
ts
o
f
3
1
.
6
7
%
o
f
p
ar
ticip
a
n
ts
,
f
o
ld
-
4
co
n
s
is
ts
o
f
2
1
.
6
7
%
s
am
p
les,
an
d
f
o
ld
-
5
co
n
s
is
ts
o
f
3
5
%.
W
h
er
ea
s
f
o
ld
th
r
ee
co
n
s
is
ts
o
f
2
3
.
3
3
%
in
th
e
6
6
m
g
/d
t
o
6
8
m
g
/d
l
r
an
g
e.
All
f
in
e
-
tu
n
e
d
h
y
p
er
p
ar
am
eter
s
f
o
r
th
e
r
eg
r
ess
io
n
m
o
d
el
ar
e
p
r
esen
ted
in
T
ab
le
2
.
DT
an
d
GB
s
h
o
wed
b
etter
p
r
ed
ictio
n
o
u
tp
u
t,
1
8
.
2
%,
th
an
th
e
r
em
ain
in
g
p
r
e
d
icto
r
s
.
Fig
u
r
e
9
.
Scatter
p
l
o
t f
o
r
th
e
5
-
f
o
ld
s
cr
o
s
s
v
alid
atio
n
f
o
r
r
eg
r
ess
io
n
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
13
,
No
.
4
,
Dec
em
b
er
2
0
2
4
:
4
4
5
3
-
4
4
7
1
4462
T
ab
le
2
.
L
is
t m
o
d
els an
d
f
in
e
-
tu
n
ed
h
y
p
er
p
ar
am
eter
s
u
s
ed
in
r
eg
r
ess
io
n
A
l
g
o
r
i
t
h
m
H
y
p
e
r
p
a
r
a
me
t
e
r
V
a
l
u
e
DT
M
i
n
i
m
u
m
n
u
m
b
e
r
o
f
d
a
t
a
p
o
i
n
t
s
w
i
t
h
i
n
l
e
a
f
n
o
d
e
s
1
D
o
n
o
t
d
i
v
i
d
e
su
b
s
e
t
s
t
h
a
t
a
r
e
l
e
ss t
h
a
n
a
5
Li
mi
t
t
h
e
m
a
x
i
m
u
m
d
e
p
t
h
o
f
t
h
e
t
r
e
e
t
o
a
sp
e
c
i
f
i
c
v
a
l
u
e
.
1
0
0
S
t
o
p
w
h
e
n
t
h
e
si
g
n
i
f
i
c
a
n
t
t
h
r
e
s
h
o
l
d
i
s
a
c
h
i
e
v
e
d
[
%]
95
Tr
e
e
s
n
u
mb
e
r
38
RF
Th
e
c
o
u
n
t
o
f
f
e
a
t
u
r
e
s
d
e
l
i
b
e
r
a
t
e
d
a
t
e
v
e
r
y
s
p
l
i
t
5
A
v
o
i
d
sp
l
i
t
t
i
n
g
s
u
b
se
t
s
i
f
sma
l
l
e
r
t
h
a
n
5
Te
c
h
n
i
q
u
e
Ex
t
r
e
me
G
B
Tr
e
e
s
n
u
mb
e
r
75
R
a
t
e
o
f
l
e
a
r
n
i
n
g
0
.
3
R
e
p
l
i
c
a
b
l
e
t
r
a
i
n
i
n
g
y
e
s
GB
r
e
g
u
l
a
r
i
z
a
t
i
o
n
La
mb
d
a
:
0
.
0
0
0
1
C
o
m
p
e
l
t
h
e
d
e
p
t
h
o
f
i
n
d
i
v
i
d
u
a
l
t
r
e
e
s
7
Q
u
a
n
t
i
t
y
o
f
t
r
a
i
n
i
n
g
e
x
a
m
p
l
e
s
1
Q
u
a
n
t
i
t
y
o
f
a
t
t
r
i
b
u
t
e
s
u
t
i
l
i
z
e
d
i
n
e
a
c
h
t
r
e
e
1
Q
u
a
n
t
i
t
y
o
f
a
t
t
r
i
b
u
t
e
s
i
n
c
l
u
d
e
d
a
t
e
a
c
h
l
e
v
e
l
.
1
Q
u
a
n
t
i
t
y
o
f
a
t
t
r
i
b
u
t
e
s
u
se
d
a
t
e
a
c
h
d
i
v
i
si
o
n
1
B
a
se
e
s
t
i
m
a
t
o
r
Tr
e
e
Est
i
m
a
t
o
r
q
u
a
n
t
i
t
y
33
A
d
a
B
o
o
st
R
a
t
e
o
f
l
e
a
r
n
i
n
g
0
.
1
F
i
x
e
d
see
d
-
r
a
n
d
o
m
g
e
n
e
r
a
t
o
r
50
R
e
g
r
e
ssi
o
n
l
o
ss
f
u
n
c
t
i
o
n
Li
n
e
a
r
C
o
s
t
(
C
)
3
R
e
g
r
e
ssi
o
n
c
o
s
t
e
p
s
i
l
o
n
(
ε)
0
.
5
S
V
M
K
e
r
n
e
l
f
u
n
c
t
i
o
n
RBF
N
u
meri
c
a
l
t
o
l
e
r
a
n
c
e
0
.
2
I
t
e
r
a
t
i
o
n
l
i
mi
t
1
5
0
C
l
a
s
si
f
i
c
a
t
i
o
n
S
q
u
a
r
e
d
ε
i
n
s
e
n
si
t
i
v
e
ε
0
.
1
4
R
e
g
r
e
ssi
o
n
ε se
n
si
t
i
v
e
ε
0
.
8
R
e
g
u
l
a
r
i
z
a
t
i
o
n
La
sso
(
L
1
)
S
t
r
e
n
g
t
h
(
α)
0
.
0
0
0
1
1
S
G
D
Le
a
r
n
i
n
g
r
a
t
e
C
o
n
st
a
n
t
I
n
i
t
i
a
l
l
e
a
r
n
i
n
g
r
a
t
e
0
.
0
0
1
2
N
u
mb
e
r
o
f
i
t
e
r
a
t
i
o
n
s
5
0
0
0
To
l
e
r
a
n
c
e
0
.
0
0
1
S
h
u
f
f
l
e
d
a
t
a
a
f
t
e
r
e
a
c
h
i
t
e
r
a
t
i
o
n
Y
e
s
F
i
x
e
d
see
d
f
o
r
r
a
n
d
o
m
sh
u
f
f
l
i
n
g
15
M
LP
N
e
u
r
o
n
s w
i
t
h
i
n
c
o
n
c
e
a
l
e
d
l
a
y
e
r
s
1
0
0
;
5
0
;
2
5
;
1
0
I
n
i
t
i
a
t
i
o
n
R
e
l
.
u
S
o
l
u
t
i
o
n
A
d
a
m
R
e
g
u
l
a
r
i
z
a
t
i
o
n
(
α)
0
Th
e
ma
x
i
mu
m
a
l
l
o
w
a
b
l
e
c
o
u
n
t
o
f
i
t
e
r
a
t
i
o
n
s
5
0
0
Tr
a
i
n
i
n
g
t
h
a
t
c
a
n
b
e
r
e
p
l
i
c
a
b
l
e
Y
e
s
T
h
e
ev
alu
atio
n
p
ar
a
m
eter
s
m
e
an
s
q
u
ar
e
e
r
r
o
r
(
MSE
)
,
r
o
o
t
m
ea
n
s
q
u
ar
e
e
r
r
o
r
(
R
MSE
)
,
m
ea
n
ab
s
o
lu
te
er
r
o
r
(
MA
E
)
,
an
d
c
o
ef
f
icien
t
o
f
d
eter
m
in
atio
n
(
R
2
)
ar
e
co
n
s
id
er
ed
f
o
r
p
r
ed
ictio
n
as
s
h
o
wn
in
T
ab
le
3
.
DT
an
d
GB
h
av
e
b
etter
ac
cu
r
ac
y
th
a
n
o
th
er
alg
o
r
ith
m
s
.
T
h
e
DT
an
d
GB
p
r
ed
icted
v
alu
es
r
em
ain
ed
b
etwe
en
6
6
≤
a
n
d
DT
GB
<
6
8
,
4
6
%.
T
h
e
m
ajo
r
in
f
lu
en
cin
g
f
ea
tu
r
es
u
n
d
er
t
h
is
r
a
n
g
e
ar
e
g
astric
p
r
o
b
lem
s
6
5
.
2
%,
u
s
in
g
in
s
u
lin
s
7
8
.
2
%,
p
o
o
r
d
ietar
y
in
ta
k
e
6
5
.
2
%,
an
d
lack
o
f
k
n
o
wled
g
e
a
b
o
u
t
h
y
p
o
g
ly
ce
m
ia
6
0
.
8
%.
T
h
e
s
e
f
ea
tu
r
es
p
lay
e
d
an
im
p
o
r
ta
n
t
r
o
le
i
n
m
ak
in
g
t
h
e
p
r
ed
ictio
n
.
T
h
e
m
ax
im
u
m
er
r
o
r
f
o
r
th
e
p
r
e
d
icted
v
alu
es
u
n
d
er
DT
an
d
GB
is
±
1
0
m
g
/d
l.
Fro
m
t
h
ese
s
am
p
les,
DT
an
d
GB
p
r
ed
icted
a
b
o
u
t
4
4
.
9
3
%
ac
cu
r
ately
.
T
h
e
r
e
m
ain
in
g
s
am
p
les
h
ad
s
o
m
e
m
in
o
r
p
r
ed
ictio
n
er
r
o
r
s
.
T
ab
le
3
.
R
eg
r
ess
io
n
m
o
d
el
ev
alau
tio
n
M
o
d
e
l
M
S
E
R
M
S
E
M
A
E
R
2
DT
1
2
.
7
7
3
.
5
7
4
2
.
8
7
7
0
.
1
8
2
RF
1
2
.
8
6
2
3
.
5
8
6
2
.
8
7
6
0
.
1
7
6
GB
1
2
.
7
7
0
3
.
5
7
4
2
.
8
7
7
0
.
1
8
2
A
d
a
B
o
o
st
1
2
.
9
1
9
3
.
5
9
4
2
.
9
4
0
0
.
1
7
3
S
V
M
1
4
.
3
3
1
3
.
7
8
6
3
.
1
7
1
0
.
0
8
2
S
G
D
1
4
.
9
7
7
3
.
8
7
0
2
.
9
7
9
0
.
0
4
N
e
u
r
a
l
N
e
t
w
o
r
k
1
2
.
8
4
0
3
.
5
8
3
2
.
9
0
0
0
.
1
7
8
Evaluation Warning : The document was created with Spire.PDF for Python.