I
nte
rna
t
io
na
l J
o
urna
l o
f
E
lect
rica
l a
nd
Co
m
pu
t
er
E
ng
ineering
(
I
J
E
CE
)
Vo
l.
15
,
No
.
6
,
Decem
b
er
20
25
,
p
p
.
5
3
4
7
~
5
3
5
9
I
SS
N:
2088
-
8
7
0
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijece.
v
15
i
6
.
pp
5
3
4
7
-
5
3
5
9
5347
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ec
e.
ia
esco
r
e.
co
m
An ensembl
e ma
c
hine learning
ba
sed mo
del
for pred
iction a
nd
dia
g
no
sis
of diabetes melli
tus
M
o
a
t
a
z
M
o
ha
m
ed
E
l Sherbi
ny
1
,
Asm
a
a
H
a
m
dy
Ra
bie
2
,
M
o
ha
m
ed
G
a
m
a
l A
bd
el
F
a
t
t
a
h
1
,
Ali El
s
herbi
ny
T
a
k
i El
din
3
,
H
o
s
s
a
m
E
l
-
Din M
o
s
t
a
f
a
1
1
D
e
p
a
r
t
me
n
t
o
f
El
e
c
t
r
o
n
i
c
s a
n
d
C
o
m
mu
n
i
c
a
t
i
o
n
En
g
i
n
e
e
r
i
n
g
,
F
a
c
u
l
t
y
o
f
E
n
g
i
n
e
e
r
i
n
g
,
M
a
n
s
o
u
r
a
U
n
i
v
e
r
s
i
t
y
,
M
a
n
so
u
r
a
,
E
g
y
p
t
2
D
e
p
a
r
t
me
n
t
o
f
C
o
m
p
u
t
e
r
a
n
d
C
o
n
t
r
o
l
S
y
s
t
e
m
s E
n
g
i
n
e
e
r
i
n
g
S
c
i
e
n
c
e
,
F
a
c
u
l
t
y
o
f
En
g
i
n
e
e
r
i
n
g
,
M
a
n
s
o
u
r
a
U
n
i
v
e
r
si
t
y
,
M
a
n
so
u
r
a
,
E
g
y
p
t
3
D
e
p
a
r
t
me
n
t
o
f
C
y
b
e
r
S
e
c
u
r
i
t
y
,
F
a
c
u
l
t
y
o
f
A
r
t
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
,
D
e
l
t
a
U
n
i
v
e
r
si
t
y
,
G
a
masa
,
E
g
y
p
t
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Feb
2
2
,
2
0
2
5
R
ev
is
ed
J
u
l 1
7
,
2
0
2
5
Acc
ep
ted
Sep
1
4
,
2
0
2
5
Dia
b
e
tes
m
e
ll
it
u
s
(DM)
is
a
c
h
ro
n
ic
m
e
tab
o
li
c
d
iso
r
d
e
r
t
h
a
t
p
o
se
s sig
n
ifi
c
a
n
t
h
e
a
lt
h
risk
s
a
n
d
g
lo
b
a
l
e
c
o
n
o
m
i
c
b
u
rd
e
n
s.
Early
p
re
d
icti
o
n
a
n
d
a
c
c
u
ra
te
d
iag
n
o
sis
a
re
c
ru
c
ial
fo
r
e
ffe
c
ti
v
e
m
a
n
a
g
e
m
e
n
t
a
n
d
trea
tme
n
t.
Th
is
stu
d
y
p
re
se
n
ts
a
n
e
n
se
m
b
le
m
a
c
h
in
e
le
a
rn
in
g
-
b
a
se
d
m
o
d
e
l
d
e
sig
n
e
d
to
p
re
d
ict
a
n
d
d
iag
n
o
se
Dia
b
e
tes
M
e
ll
it
u
s
u
si
n
g
c
li
n
ica
l
a
n
d
d
e
m
o
g
ra
p
h
ic
d
a
ta.
Th
e
p
ro
p
o
se
d
a
p
p
r
o
a
c
h
in
te
g
ra
tes
m
u
lt
i
p
le
m
a
c
h
i
n
e
lea
rn
i
n
g
a
lg
o
rit
h
m
s,
in
c
lu
d
in
g
ra
n
d
o
m
fo
re
st
(RF
),
e
x
t
re
m
e
g
ra
d
ien
t
b
o
o
stin
g
(XG
B),
a
n
d
l
o
g
isti
c
re
g
re
ss
io
n
(LR),
to
le
v
e
ra
g
e
t
h
e
ir
in
d
i
v
id
u
a
l
stre
n
g
t
h
s
a
n
d
e
n
h
a
n
c
e
th
e
e
n
ti
re
p
e
rfo
rm
a
n
c
e
.
Th
e
e
n
se
m
b
le
m
o
d
e
l
wa
s
train
e
d
a
n
d
v
a
li
d
a
ted
o
n
m
u
lt
ip
le
c
o
m
p
re
h
e
n
siv
e
d
a
tas
e
ts.
P
e
rfo
rm
a
n
c
e
m
e
a
su
re
s
d
e
m
o
n
stra
te
th
e
ro
b
u
st
n
e
ss
o
f
p
r
o
p
o
se
d
m
o
d
e
l
a
n
d
it
s
re
li
a
b
il
it
y
in
d
isti
n
g
u
ish
in
g
d
iab
e
ti
c
c
a
se
s
fro
m
non
-
d
iab
e
ti
c
c
a
se
s
a
fter
a
p
p
ly
i
n
g
se
v
e
ra
l
p
re
p
r
o
c
e
ss
in
g
ste
p
s.
Th
is
w
o
rk
e
n
su
re
s
th
e
c
a
p
a
b
il
it
y
o
f
m
a
c
h
in
e
lea
rn
i
n
g
in
a
d
v
a
n
c
i
n
g
h
e
a
l
th
c
a
re
b
y
p
ro
v
id
i
n
g
e
fficie
n
t,
d
a
ta
-
d
r
iv
e
n
to
o
ls
f
o
r
d
iab
e
tes
m
a
n
a
g
e
m
e
n
t,
a
i
d
in
g
c
li
n
icia
n
s
i
n
e
a
rl
y
d
ia
g
n
o
sis,
a
n
d
c
o
n
tr
ib
u
ti
n
g
to
p
e
rso
n
a
li
z
e
d
trea
tme
n
t
stra
teg
ies
.
Co
m
p
a
ra
ti
v
e
a
n
a
l
y
sis
a
g
a
in
st
sta
n
d
a
lo
n
e
m
o
d
e
ls
h
i
g
h
li
g
h
ts
th
e
su
p
e
rio
r
p
re
d
ictiv
e
c
a
p
a
b
il
i
ti
e
s o
f
th
e
e
n
se
m
b
le ap
p
ro
a
c
h
.
Re
su
lt
s h
a
d
sh
o
wn
th
a
t
e
n
se
m
b
le
m
o
d
e
l
a
c
h
iev
e
d
a
n
a
c
c
u
ra
c
y
o
f
9
6
.
8
8
%
a
n
d
p
r
e
c
isio
n
o
f
8
9
.
8
5
%
o
u
t
p
e
rfo
rm
in
g
i
n
d
i
v
id
u
a
l
c
las
sifiers
.
K
ey
w
o
r
d
s
:
C
las
s
if
icatio
n
Diab
etes m
ellitu
s
E
n
s
em
b
le
Ma
ch
in
e
lear
n
in
g
Per
f
o
r
m
an
ce
m
ea
s
u
r
es
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
:
Mo
ataz
Mo
h
am
ed
E
l Sh
er
b
in
y
Dep
ar
tm
en
t o
f
E
lectr
o
n
ics an
d
C
o
m
m
u
n
icatio
n
E
n
g
in
ee
r
i
n
g
,
Facu
lty
o
f
E
n
g
in
ee
r
in
g
,
Ma
n
s
o
u
r
a
Un
iv
e
r
s
ity
Ma
n
s
o
u
r
a,
DK
3
5
5
1
6
,
E
g
y
p
t
E
m
ail: m
o
ataz
elsh
er
b
in
y
@
m
a
n
s
.
ed
u
.
eg
1.
I
NT
RO
D
UCT
I
O
N
Dia
b
et
es
m
e
llit
u
s
is
o
n
e
o
f
th
e
m
o
s
t
p
r
e
v
ale
n
t
c
h
r
o
n
i
c
d
is
ea
s
es
a
n
d
h
as
b
ec
o
m
e
a
m
aj
o
r
p
u
b
li
c
h
e
alt
h
ch
all
en
g
e
[
1
]
.
Di
ab
et
es
m
e
llit
u
s
(
D
M)
is
a
c
h
r
o
n
i
c
m
e
ta
b
o
li
c
d
is
ea
s
e
ch
a
r
ac
te
r
iz
e
d
b
y
el
e
v
at
e
d
b
l
o
o
d
g
l
u
co
s
e
lev
els
,
r
es
u
l
ti
n
g
f
r
o
m
e
it
h
e
r
i
n
s
u
f
f
i
cie
n
t
i
n
s
u
li
n
p
r
o
d
u
c
ti
o
n
o
r
i
n
ef
f
e
cti
v
e
i
n
s
u
li
n
u
t
ili
za
t
io
n
[
2
]
.
T
h
e
p
r
e
v
a
le
n
c
e
o
f
d
ia
b
etes
h
as
r
ea
ch
e
d
ala
r
m
i
n
g
l
ev
els
wi
th
p
r
o
j
ec
t
io
n
s
in
d
ic
ati
n
g
f
u
r
t
h
e
r
g
r
o
wt
h
in
t
h
e
c
o
m
i
n
g
d
e
ca
d
es
.
Acc
o
r
d
in
g
t
o
t
h
e
I
n
te
r
n
a
ti
o
n
a
l
D
ia
b
e
tes
Fe
d
er
ati
o
n
(
I
DF
)
,
5
3
7
m
il
li
o
n
ad
u
lts
a
g
e
d
2
0
–
7
9
y
ea
r
s
w
er
e
li
v
i
n
g
wit
h
d
ia
b
etes
g
lo
b
all
y
i
n
2
0
2
1
[
3
]
.
T
h
is
n
u
m
b
er
is
p
r
o
je
cte
d
to
r
is
e
t
o
6
4
3
m
il
li
o
n
b
y
2
0
3
0
an
d
7
8
3
m
ill
io
n
b
y
2
0
4
5
,
r
e
f
le
cti
n
g
a
s
i
g
n
i
f
ic
an
t
u
p
wa
r
d
t
r
e
n
d
.
I
n
2
0
2
1
,
d
i
a
b
et
es
was
est
im
ate
d
to
b
e
th
e
ca
u
s
e
o
f
6
.
7
m
illi
o
n
f
at
ali
ties
,
i
n
d
ic
ati
n
g
t
h
a
t
o
n
e
p
er
s
o
n
d
ies
ev
e
r
y
f
i
v
e
s
e
co
n
d
s
d
u
e
t
o
d
i
ab
et
es
-
r
e
lat
ed
c
o
m
p
l
i
ca
ti
o
n
s
.
Di
ab
etes
is
co
n
s
i
d
e
r
e
d
a
s
i
le
n
t
k
i
lle
r
w
h
e
r
e
th
e
n
u
m
b
e
r
o
f
u
n
d
i
a
g
n
o
s
e
d
ca
s
es
g
l
o
b
all
y
is
n
ea
r
l
y
2
4
0
m
ill
io
n
in
d
i
v
i
d
u
a
ls
ac
c
o
u
n
ti
n
g
f
o
r
1
i
n
2
a
d
u
lts
wit
h
d
ia
b
e
tes
[
4
]
.
E
g
y
p
t
r
a
n
k
ed
n
i
n
t
h
g
l
o
b
al
ly
i
n
t
h
e
n
u
m
b
er
o
f
d
ia
b
e
tes
ca
s
es
,
wit
h
1
0
.
9
m
i
lli
o
n
ad
u
l
ts
l
iv
i
n
g
wit
h
t
h
e
d
is
ea
s
e
.
Ar
o
u
n
d
5
0
%
o
f
d
ia
b
etes
c
ases
i
n
E
g
y
p
t
r
e
m
ain
u
n
d
ia
g
n
o
s
ed
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
15
,
No
.
6
,
Decem
b
e
r
20
25
:
5
3
4
7
-
5
3
5
9
5348
T
h
e
d
is
ea
s
e
is
a
s
s
o
ciate
d
wit
h
s
ev
er
e
co
m
p
licatio
n
s
[
5
]
,
in
clu
d
in
g
ca
r
d
io
v
ascu
lar
d
is
o
r
d
er
s
,
r
en
al
f
ailu
r
e,
n
eu
r
o
p
ath
y
,
an
d
r
etin
o
p
ath
y
,
wh
ich
s
ig
n
if
ican
tly
i
m
p
ac
t
th
e
q
u
ality
o
f
life
a
n
d
in
cr
ea
s
e
m
o
r
tality
r
ates.
E
ar
ly
d
iag
n
o
s
is
an
d
ef
f
ec
tiv
e
m
an
ag
em
e
n
t
ar
e,
t
h
er
ef
o
r
e,
cr
itical
in
m
itig
atin
g
t
h
e
d
ev
elo
p
m
en
t
o
f
d
iab
etes
an
d
its
ac
co
m
p
an
y
i
n
g
co
m
p
licatio
n
s
.
T
r
a
d
itio
n
al
d
iag
n
o
s
tic
m
eth
o
d
s
f
o
r
d
iag
n
o
s
is
o
f
d
iab
etes
[
6
]
,
s
u
ch
as
f
asti
n
g
p
lasma
g
lu
c
o
s
e
(
FP
G)
,
o
r
al
g
lu
c
o
s
e
to
ler
an
ce
test
s
(
OGT
T
)
,
an
d
h
em
o
g
l
o
b
in
A1
c
(
Hb
A1
c)
lev
els,
ar
e
r
eliab
le
b
u
t
m
ay
b
e
lim
ited
b
y
co
s
t,
ac
ce
s
s
ib
ilit
y
,
an
d
th
e
n
ee
d
f
o
r
lab
o
r
a
to
r
y
in
f
r
astru
ct
u
r
e.
Mo
r
eo
v
er
,
th
ese
m
eth
o
d
s
o
f
te
n
f
ail
to
p
r
ed
ict
th
e
o
n
s
et
o
f
d
iab
etes
in
p
r
ed
iab
etic
in
d
iv
i
d
u
als,
em
p
h
asizin
g
th
e
n
ee
d
f
o
r
in
n
o
v
ativ
e
ap
p
r
o
ac
h
es
to
en
h
an
ce
ea
r
ly
d
etec
tio
n
.
I
n
r
ec
en
t
y
ea
r
s
,
ad
v
an
ce
s
i
n
m
ac
h
in
e
lear
n
i
n
g
(
ML
)
h
av
e
d
e
m
o
n
s
tr
ated
s
i
g
n
if
ican
t
p
o
te
n
tial
in
h
ea
lth
ca
r
e,
o
f
f
e
r
in
g
d
ata
-
d
r
iv
e
n
s
o
lu
tio
n
s
f
o
r
d
is
ea
s
e
p
r
ed
ictio
n
,
d
iag
n
o
s
is
,
an
d
p
er
s
o
n
alize
d
tr
ea
tm
en
t.
T
elem
ed
icin
e
h
as
b
ec
o
m
e
a
g
am
e
-
ch
a
n
g
in
g
s
o
lu
tio
n
[
7
]
,
wh
er
e
it
im
p
r
o
v
es
h
ea
lth
ca
r
e
ac
ce
s
s
ib
ilit
y
b
y
elim
in
atin
g
t
h
e
n
ee
d
f
o
r
i
n
-
p
er
s
o
n
h
o
s
p
ital
v
is
its
th
r
o
u
g
h
th
e
u
tili
za
tio
n
o
f
d
ig
ital c
o
m
m
u
n
icatio
n
tech
n
o
l
o
g
y
t
h
at
en
ab
les
d
is
tan
t c
o
n
s
u
ltatio
n
s
.
Mo
s
t
o
f
p
r
ev
io
u
s
s
tu
d
ies
em
p
lo
y
ed
th
e
Pima
I
n
d
ian
s
Diab
etes
Data
s
et
(
PID
D)
.
I
t
is
co
n
s
id
er
ed
as
o
n
e
o
f
th
e
m
o
s
t
well
-
k
n
o
wn
d
atasets
in
b
in
ar
y
class
if
icatio
n
o
f
d
iab
etes
u
s
in
g
m
ac
h
in
e
lear
n
in
g
.
Feb
r
ian
et
a
l.
[
8
]
ap
p
lied
two
s
u
p
er
v
is
ed
m
ac
h
i
n
e
lear
n
in
g
a
lg
o
r
ith
m
s
o
n
th
e
PID
D.
T
r
ai
n
an
d
test
s
p
lit
wer
e
p
er
f
o
r
m
ed
with
o
u
t
c
r
o
s
s
v
ali
d
atio
n
.
T
h
e
r
esu
lts
o
f
K
-
n
ea
r
est
n
eig
h
b
o
r
(
KNN)
wer
e
o
u
t
p
er
f
o
r
m
ed
b
y
n
aïv
e
B
ay
es
(
NB
)
in
b
o
th
ex
p
er
im
e
n
ts
.
Au
th
o
r
s
co
m
p
ar
e
d
r
esu
lts
in
ter
m
s
o
f
ac
cu
r
ac
y
,
r
ec
all
as
well
as
p
r
ec
is
io
n
.
NB
ac
h
iev
ed
th
e
h
ig
h
est
ac
cu
r
ac
y
o
f
7
8
.
5
2
%.
Kan
g
r
a
an
d
Sin
g
h
[
9
]
s
p
lit
d
ata
in
to
tr
ain
i
n
g
an
d
test
in
g
u
s
in
g
10
-
f
o
l
d
cr
o
s
s
-
v
alid
atio
n
f
o
r
p
r
ep
r
o
ce
s
s
in
g
s
tag
e.
Au
t
h
o
r
s
co
m
p
ar
e
d
s
ix
s
u
p
e
r
v
is
ed
m
ac
h
in
e
lea
r
n
in
g
alg
o
r
ith
m
s
u
s
in
g
th
r
ee
ev
alu
at
io
n
m
etr
ics
wh
ich
ar
e
ac
cu
r
ac
y
,
p
r
ec
is
io
n
an
d
r
ec
all.
T
h
e
y
NB
,
KNN,
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
e
(
SVM)
,
d
ec
is
io
n
tr
ee
(
DT
)
,
r
an
d
o
m
f
o
r
est
(
R
F)
an
d
lo
g
is
tic
r
eg
r
ess
io
n
(
L
R
)
o
n
t
h
e
PID
D
in
d
icatin
g
th
at
SVM
ac
h
iev
ed
h
ig
h
est
ac
c
u
r
ac
y
s
co
r
e
o
f
7
4
.
3
%
f
o
llo
wed
b
y
L
R
wh
ich
ac
h
iev
ed
7
4
%.
C
h
an
g
et
a
l.
[
1
0
]
co
n
d
u
cted
th
r
ee
ex
p
er
im
en
ts
o
n
th
e
PID
D.
T
h
e
f
ir
s
t e
x
p
er
im
en
t sh
o
wed
th
at
R
F
o
u
tp
er
f
o
r
m
ed
b
o
th
DT
an
d
NB
b
y
ac
h
iev
in
g
7
9
.
5
7
% a
n
d
8
9
.
4
% in
ter
m
s
o
f
ac
c
u
r
ac
y
an
d
p
r
ec
is
io
n
r
esp
ec
tiv
e
ly
.
Au
th
o
r
s
ap
p
lied
f
ea
tu
r
e
s
elec
tio
n
o
f
3
-
f
ac
t
o
r
o
f
th
e
en
tire
d
ataset
in
th
e
s
ec
o
n
d
ex
p
er
im
en
t.
NB
r
ea
ch
e
d
a
cc
u
r
ac
y
o
f
7
9
.
1
3
%,
an
d
F1
-
s
co
r
e
o
f
8
4
.
7
1
%.
I
n
t
h
eir
f
in
al
e
x
p
er
im
e
n
t,
au
t
h
o
r
s
u
tili
ze
d
f
ea
tu
r
e
s
elec
tio
n
o
f
5
-
f
ac
to
r
.
Ho
wev
er
,
ac
cu
r
ac
y
wen
t
d
o
w
n
to
7
7
.
8
3
%
b
y
NB
.
Mu
s
h
taq
et
a
l.
[
1
1
]
em
p
l
o
y
ed
a
two
-
s
ta
g
e
m
o
d
el
s
elec
tio
n
m
eth
o
d
o
l
o
g
y
.
L
R
,
SVM,
KNN,
GB
,
NB
an
d
R
F
ap
p
lied
to
d
eter
m
in
e
th
e
ef
f
icien
cy
o
f
p
r
ed
ictio
n
m
o
d
els.
R
F
was
f
o
u
n
d
to
b
e
th
e
b
est
with
ac
cu
r
ac
y
o
f
8
0
.
7
%
af
ter
ap
p
l
y
in
g
s
m
o
te.
T
h
e
e
n
s
em
b
le
o
f
th
e
b
est
3
m
o
d
els
y
ield
ed
ac
c
u
r
ac
y
o
f
8
2
%
o
n
o
r
ig
in
al
d
ataset
an
d
8
1
.
7
%
o
n
b
alan
ce
d
d
ataset.
R
awa
t
et
a
l.
[
1
2
]
ass
u
r
es
th
e
u
s
ef
u
ln
ess
o
f
d
ata
m
in
in
g
tech
n
iq
u
es
to
ev
alu
ate
th
e
u
n
k
n
o
wn
p
atter
n
s
o
n
th
e
PID
D.
Au
th
o
r
s
p
r
o
p
o
s
ed
m
u
ltip
le
tech
n
iq
u
es
s
u
ch
as
A
d
aBo
o
s
t
an
d
Naïv
e
B
ay
es
f
o
r
th
e
an
aly
s
is
an
d
p
r
ed
ictio
n
o
f
DM
p
atien
ts
.
T
h
e
r
esu
lts
co
m
p
u
ted
ar
e
f
o
u
n
d
to
b
e
7
9
.
6
9
%
class
if
icatio
n
ac
cu
r
ac
y
b
y
Ad
aBo
o
s
t
m
eth
o
d
.
B
ar
ik
et
a
l.
[
1
3
]
u
s
ed
two
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
s
o
n
PID
D.
I
n
th
e
ca
s
e
o
f
R
F,
th
e
p
r
ed
ictio
n
v
alu
e
was
7
1
.
9
%
b
u
t
XGBo
o
s
t
y
ield
ed
h
ig
h
e
r
ac
cu
r
ac
y
o
f
7
4
.
1
%.
Palim
k
ar
et
a
l.
[
1
4
]
u
tili
ze
d
m
u
ltip
le
m
ac
h
in
e
lear
n
in
g
m
o
d
els
o
n
a
q
u
esti
o
n
n
air
e
d
ataset
s
u
ch
a
s
L
R
,
SVM,
n
aïv
e
B
ay
es
a
n
d
ad
a
p
tiv
e
b
o
o
s
tin
g
(
Ad
aB
o
o
s
t)
.
R
esu
lts
wer
e
co
m
p
ar
ed
u
s
in
g
7
0
%
-
3
0
%
tr
ain
in
g
an
d
test
in
g
ac
cu
r
ac
y
r
esp
ec
tiv
ely
in
ad
d
itio
n
to
m
ea
n
s
q
u
ar
e
er
r
o
r
(
MSE
)
.
L
R
ac
h
iev
ed
9
3
.
5
9
%,
SVM
y
ield
ed
9
4
.
2
3
%,
Gau
s
s
ian
NB
9
1
.
0
2
%
a
n
d
Ad
aBo
o
s
t
9
4
.
8
7
in
ter
m
s
o
f
test
in
g
ac
cu
r
ac
y
.
Vo
ca
l
b
io
m
ar
k
er
p
r
ed
ictio
n
o
f
d
is
ea
s
e
h
as
b
ee
n
em
p
lo
y
e
d
in
a
v
ar
iety
o
f
d
is
ea
s
es,
in
clu
d
in
g
C
OVI
D
-
1
9
d
etec
tio
n
,
Par
k
in
s
o
n
'
s
d
is
ea
s
e,
p
u
lm
o
n
ar
y
f
u
n
ctio
n
,
an
d
co
r
o
n
ar
y
ar
ter
y
d
is
ea
s
e.
Fag
h
er
az
zi
et
a
l.
[
1
5
]
im
p
lem
en
ted
th
eir
s
tu
d
y
o
n
C
o
liv
e
s
tu
d
y
v
o
ice
d
ataset.
au
th
o
r
s
u
tili
ze
d
th
r
ee
class
if
ier
alg
o
r
ith
m
s
s
u
ch
as
lo
g
is
tic
r
eg
r
ess
io
n
,
s
u
p
p
o
r
t v
ec
to
r
m
ac
h
in
e
a
n
d
m
u
lti
-
lay
e
r
p
er
ce
p
tr
o
n
class
if
ier
s
(
ML
P).
R
esu
lts
in
d
icate
d
th
at
ML
P
y
ield
e
d
th
e
h
i
g
h
est
ac
cu
r
ac
y
o
f
6
7
%
o
n
f
em
ale
g
r
o
u
p
with
6
6
%,6
7
%
s
p
ec
if
ic
ity
an
d
s
en
s
itiv
ity
r
esp
ec
tiv
ely
.
Fu
r
th
e
r
m
o
r
e
,
ML
P
ac
h
iev
ed
7
1
%,
7
0
%
an
d
7
3
%
in
ter
m
s
o
f
ac
cu
r
ac
y
,
s
p
ec
if
icity
a
n
d
s
en
s
itiv
ity
r
esp
ec
tiv
ely
.
Kau
f
m
an
et
a
l.
[
1
6
]
in
v
esti
g
ated
th
e
p
r
o
s
p
ec
t
o
f
s
p
ee
ch
an
aly
s
is
as
a
p
r
escr
ee
n
in
g
o
r
tr
ac
k
in
g
to
o
l
f
o
r
ty
p
e
2
d
ia
b
etes
m
ellitu
s
(
T
2
DM
)
th
r
o
u
g
h
co
n
tr
asti
n
g
t
h
e
v
o
ice
r
ec
o
r
d
in
g
s
b
etwe
en
n
o
n
-
d
iab
etic
an
d
T
2
DM
in
d
iv
i
d
u
a
ls
.
T
o
tal
2
6
7
p
a
r
ticip
an
ts
wer
e
d
iag
n
o
s
ed
as
n
o
n
-
d
iab
etic
o
r
d
iab
etic
b
ased
o
n
Am
er
ican
Diab
etes
Ass
o
ciatio
n
(
ADA)
g
u
id
elin
es.
Sam
p
les r
ec
r
u
ited
i
n
I
n
d
ia
u
s
in
g
a
s
m
ar
tp
h
o
n
e
ap
p
licatio
n
r
ec
o
r
d
in
g
a
fix
e
d
p
h
r
ase
in
a
d
d
itio
n
to
d
em
o
g
r
a
p
h
ic
f
ea
t
u
r
es
s
u
ch
as
ag
e
an
d
b
o
d
y
m
ass
in
d
ex
.
Au
th
o
r
s
im
p
lem
en
ted
two
s
u
p
er
v
is
ed
m
ac
h
in
e
lear
n
in
g
m
o
d
els
wh
i
ch
ar
e
l
o
g
is
tic
r
eg
r
ess
io
n
a
n
d
n
aïv
e
B
ay
es.
L
R
ac
h
iev
ed
test
in
g
ac
cu
r
ac
y
o
f
7
0
%
o
n
wo
m
en
v
o
ice
d
ataset.
Acc
u
r
ac
y
wen
t
u
p
to
8
2
%
wh
en
all
f
ea
tu
r
es
wer
e
im
p
lem
en
ted
.
On
Me
n
v
o
ice
d
ataset,
L
R
s
co
r
ed
a
test
in
g
ac
c
u
r
ac
y
o
f
6
9
%.
Mo
r
e
o
v
er
,
wh
e
n
all
f
ea
tu
r
es
wer
e
co
n
s
id
er
ed
,
ac
c
u
r
ac
y
we
n
t u
p
to
8
6
%.
T
h
is
s
tu
d
y
aim
s
to
d
esig
n
a
n
ac
cu
r
ate
m
ac
h
in
e
lear
n
in
g
m
o
d
el
th
at
tr
ain
s
ea
ch
class
in
d
e
p
en
d
en
t
o
f
d
ataset
o
r
ig
in
al
d
is
tr
ib
u
tio
n
a
n
d
s
ize
th
r
o
u
g
h
th
e
em
p
lo
y
m
en
t
o
f
a
p
r
o
p
er
p
r
e
p
r
o
ce
s
s
in
g
ap
p
r
o
ac
h
to
h
an
d
le
th
e
n
o
n
-
ex
is
tin
g
v
alu
es,
d
ata
r
escalin
g
an
d
class
im
b
alan
ce
.
Fu
r
th
er
m
o
r
e,
I
m
p
r
o
v
in
g
t
h
e
p
er
f
o
r
m
an
ce
o
f
p
r
o
p
o
s
ed
m
o
d
el
th
r
o
u
g
h
b
alan
cin
g
th
e
d
ataset
an
d
e
n
s
em
b
le
tech
n
iq
u
es.
T
h
e
p
r
o
p
o
s
ed
m
o
d
el
was
ap
p
lied
o
n
d
iv
er
s
e
d
atasets
to
en
s
u
r
e
th
e
g
en
er
aliza
b
ilit
y
o
f
th
e
r
esu
lts
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
A
n
en
s
emb
le
ma
ch
in
e
lea
r
n
in
g
b
a
s
ed
mo
d
el
fo
r
p
r
ed
ictio
n
a
n
d
…
(
Mo
a
ta
z
Mo
h
a
med
E
l
S
h
erb
in
y
)
5349
T
h
e
r
e
m
ain
d
er
o
f
th
is
m
an
u
s
cr
ip
t
is
o
r
g
an
ized
as
f
o
llo
ws:
Sectio
n
2
r
ev
iews
m
eth
o
d
s
u
s
ed
in
th
e
ap
p
licatio
n
o
f
m
ac
h
in
e
lear
n
i
n
g
f
o
r
d
iab
etes
d
iag
n
o
s
is
,
f
o
llo
win
g
th
e
d
escr
ip
tio
n
o
f
th
e
d
atasets
em
p
lo
y
ed
in
th
is
s
tu
d
y
.
Sectio
n
3
p
r
esen
ts
ex
p
er
im
e
n
tal
r
esu
lts
an
d
d
is
cu
s
s
io
n
.
Fin
ally
,
a
c
o
n
clu
s
io
n
an
d
p
o
s
s
ib
le
f
u
tu
r
e
wo
r
k
in
s
ec
tio
n
4
.
2.
M
E
T
H
O
D
T
h
is
s
ec
tio
n
im
p
lem
en
ts
th
e
s
tated
ar
ch
itectu
r
e
o
f
th
is
p
ap
e
r
.
I
t
d
elv
es
in
to
f
iv
e
m
ain
p
ar
ts
.
First
ly
,
d
ataset
d
escr
ip
tio
n
,
th
en
d
ata
p
r
ep
r
o
ce
s
s
in
g
,
tr
ain
-
test
s
p
lit
an
d
c
r
o
s
s
v
alid
atio
n
,
ML
alg
o
r
ith
m
s
an
d
f
in
ally
p
er
f
o
r
m
an
ce
ev
al
u
atio
n
m
etr
ic
s
as sh
o
wn
in
Fig
u
r
e
1
.
Fig
u
r
e
1
.
Pro
p
o
s
ed
f
r
am
ewo
r
k
b
lo
ck
d
iag
r
am
2
.
1
.
Da
t
a
s
et
L
ab
eled
d
ata
is
a
v
ital
in
p
u
t
to
s
u
p
er
v
is
ed
m
ac
h
in
e
lear
n
in
g
an
d
d
ee
p
lear
n
in
g
class
if
icatio
n
p
r
o
b
lem
s
[
1
7
]
.
A
r
elev
an
t
co
llectio
n
o
f
d
ata
ai
d
s
to
b
etter
m
ac
h
in
e
lear
n
in
g
class
if
icatio
n
.
T
h
er
e
ar
e
f
o
u
r
d
atasets
im
p
lem
en
ted
in
th
is
s
tu
d
y
th
at
d
if
f
er
in
n
u
m
b
er
o
f
s
am
p
les
as
well
a
s
th
e
n
u
m
b
er
an
d
th
e
ty
p
e
o
f
th
eir
attr
ib
u
tes.
T
h
ey
wer
e
g
a
th
er
ed
f
r
o
m
p
u
b
lic
h
o
s
ts
an
d
b
y
ag
r
ee
m
en
ts
with
m
e
d
ical
c
en
ter
s
an
d
d
o
cto
r
s
.
T
h
ey
ar
e
p
u
b
licly
av
ailab
le
o
n
lin
e
h
o
s
ted
b
y
UC
I
Ma
ch
i
n
e
L
ea
r
n
i
n
g
.
Deta
iled
d
escr
i
p
tio
n
o
f
f
ea
tu
r
es
in
en
tire
d
atasets
is
illu
s
tr
ated
th
r
o
u
g
h
T
ab
les
1
to
4
.
T
ab
le
1
.
Descr
ip
tiv
e
f
ea
t
u
r
es
o
f
PID
D
A
t
t
r
i
b
u
t
e
D
e
scri
p
t
i
o
n
N
u
l
l
v
a
l
u
e
s
c
o
u
n
t
R
a
n
g
e
P
r
e
g
n
a
n
c
i
e
s
N
u
mb
e
r
o
f
t
i
mes
a
p
a
t
i
e
n
t
h
a
s
b
e
e
n
p
r
e
g
n
a
n
t
-
0
-
17
G
l
u
c
o
s
e
C
o
n
c
e
n
t
r
a
t
i
o
n
o
f
p
l
a
sm
a
g
l
u
c
o
se
a
t
t
w
o
h
o
u
r
s
i
n
a
n
o
r
a
l
g
l
u
c
o
se
t
o
l
e
r
a
n
c
e
t
e
st
(
G
TI
T)
1
8
0
0
-
1
9
9
BP
D
i
a
st
o
l
i
c
b
l
o
o
d
p
r
e
ss
u
r
e
(
mm
H
g
)
2
2
1
0
-
1
2
2
ST
S
k
i
n
f
o
l
d
t
h
i
c
k
n
e
ss
i
n
Tr
i
c
e
p
s (m
m)
2
9
2
0
-
99
I
n
su
l
i
n
S
e
r
u
m I
n
su
l
i
n
f
o
r
t
w
o
h
o
u
r
s (
/
ml
)
4
9
8
0
-
8
4
6
B
M
I
B
o
d
y
mas
s i
n
d
e
x
(
k
g
/
m)
80
0
-
6
7
.
1
D
P
F
D
i
a
b
e
t
e
s
p
e
d
i
g
r
e
e
f
u
n
c
t
i
o
n
-
0
.
0
7
8
–
2
.
4
2
A
g
e
A
g
e
i
n
y
e
a
r
s
-
2
1
-
81
O
u
t
c
o
m
e
B
i
n
a
r
y
t
a
r
g
e
t
i
n
d
i
c
a
t
i
n
g
d
i
a
b
e
t
i
c
o
r
n
o
t
-
0
-
1
T
h
e
f
ir
s
t
d
ataset
n
am
ely
Pima
I
n
d
ian
Diab
etes
Data
s
et
(
PID
D)
.
T
h
e
PID
D
is
a
wid
ely
u
s
e
d
m
ed
ical
d
ata
r
ec
o
r
d
s
in
m
ac
h
in
e
lear
n
in
g
.
I
t
was
g
at
h
er
ed
b
y
th
e
Natio
n
al
I
n
s
titu
te
o
f
Diab
etes
an
d
Dig
esti
v
e
an
d
Kid
n
ey
Dis
ea
s
es
(
NI
DDK)
.
I
t
f
o
cu
s
es
o
n
p
r
e
d
ictin
g
w
h
eth
er
a
p
atien
t
h
as
d
iab
etes
b
a
s
ed
o
n
d
iag
n
o
s
tic
m
ea
s
u
r
em
en
ts
an
d
p
e
r
s
o
n
al
d
ata.
T
h
e
d
ataset
is
p
ar
t
o
f
t
h
e
UC
I
Ma
ch
in
e
L
ea
r
n
in
g
R
ep
o
s
ito
r
y
av
ailab
le
o
n
lin
e
o
n
Kag
g
le
[
1
8
]
.
I
t
co
n
s
is
ts
o
f
7
6
8
in
s
tan
ce
s
with
9
attr
ib
u
tes
in
clu
d
in
g
th
e
tar
g
et
v
a
r
iab
le.
All
f
ea
tu
r
es
ar
e
n
u
m
er
ic
an
d
d
escr
ib
e
d
in
T
ab
le
2
.
T
h
e
s
ec
o
n
d
d
ataset
is
s
u
b
m
itted
u
s
in
g
a
q
u
esti
o
n
n
air
e
f
o
r
d
iab
etes
p
r
ed
ictio
n
ca
s
e
s
tu
d
y
[
1
9
]
.
I
t
co
n
tain
s
5
2
0
s
am
p
les
a
n
d
1
7
p
r
ed
ictiv
e
f
ea
tu
r
es.
T
h
e
th
ir
d
d
ataset
[
2
0
]
co
n
s
is
ts
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
15
,
No
.
6
,
Decem
b
e
r
20
25
:
5
3
4
7
-
5
3
5
9
5350
o
f
1
0
0
,
0
0
0
s
am
p
les
with
7
f
ea
tu
r
es.
I
t
is
co
n
s
id
er
ed
th
e
lar
g
est
d
ataset
in
th
is
s
tu
d
y
am
o
n
g
th
e
f
o
u
r
im
p
lem
en
ted
o
n
es.
T
h
e
f
o
u
r
t
h
an
d
last
d
ataset
n
am
ed
Vo
i
ce
-
an
d
-
d
iab
etes
-
VOCADI
AB
[
2
1
]
is
av
ailab
le
o
n
GitHu
b
r
ep
o
s
ito
r
y
.
I
t
is
a
p
a
r
t
o
f
t
h
e
C
o
liv
e
Vo
ice
s
tu
d
y
,
t
h
at
f
o
c
u
s
es
o
n
u
s
in
g
v
o
ice
an
aly
s
is
to
s
cr
ee
n
f
o
r
ty
p
e
2
d
iab
etes
(
T
2
DM
)
in
th
e
ad
u
lt
p
o
p
u
latio
n
o
f
th
e
Un
ited
States
.
T
h
e
g
o
al
o
f
th
e
s
tu
d
y
is
to
an
aly
ze
ac
o
u
s
tic
r
ec
o
r
d
in
g
s
wh
ich
ar
e
in
th
e
f
o
r
m
o
f
v
o
ice
em
b
e
d
d
in
g
s
.
Par
ticip
an
ts
lik
ely
p
r
o
v
id
ed
s
tan
d
a
r
d
ized
v
o
ice
r
ec
o
r
d
in
g
s
,
s
u
ch
as:
Su
s
tain
ed
v
o
wels
(
/a
a
/
o
r
/o
o
/
)
.
Ad
d
itio
n
ally
,
it
in
v
o
l
v
es
th
e
ass
o
ciate
d
p
ar
ticip
an
t
m
eta
d
ata
to
d
e
v
elo
p
a
m
ac
h
in
e
lear
n
in
g
-
b
ased
s
cr
ee
n
in
g
to
o
l f
o
r
ty
p
e
2
d
iab
etes.
T
ab
le
2
.
Descr
ip
tiv
e
f
ea
t
u
r
es
o
f
q
u
esti
o
n
n
air
e
d
ataset
A
t
t
r
i
b
u
t
e
D
e
scri
p
t
i
o
n
R
a
n
g
e
(
D
i
s
t
r
i
b
u
t
i
o
n
)
A
g
e
A
g
e
o
f
p
e
r
s
o
n
i
n
y
e
a
r
s
16
-
90
G
e
n
d
e
r
S
e
x
o
f
p
a
t
i
e
n
t
M
a
l
e
(
6
3
%)
o
r
F
e
m
a
l
e
(
3
7
%)
P
o
l
y
u
r
i
a
Ex
c
e
ss
u
r
i
n
a
t
i
o
n
Y
e
s (5
0
%
o
r
N
o
(
5
0
%
)
P
o
l
y
d
i
p
si
a
Ex
c
e
ss
t
h
i
r
st
Tr
u
e
(
4
5
%
)
o
r
F
a
l
se
(
5
5
%
)
S
u
d
d
e
n
w
e
i
g
h
t
l
o
ss
U
n
i
n
t
e
n
t
i
o
n
a
l
a
n
d
r
a
p
i
d
w
e
i
g
h
t
l
o
ss
Tr
u
e
(
4
2
%
)
o
r
F
a
l
se
(
5
8
%
)
W
e
a
k
n
e
ss
R
e
d
u
c
e
d
e
n
e
r
g
y
Tr
u
e
(
5
9
%
)
o
r
F
a
l
se
(
4
1
%
)
P
o
l
y
p
h
a
g
i
a
Ex
c
e
ssi
v
e
h
u
n
g
e
r
o
r
i
n
c
r
e
a
se
d
a
p
p
e
t
i
t
e
Tr
u
e
(
4
6
%
)
o
r
F
a
l
se
(
5
4
%
)
G
e
n
i
t
a
l
t
h
r
u
sh
F
u
n
g
a
l
i
n
f
e
c
t
i
o
n
Tr
u
e
(
2
2
%
)
o
r
F
a
l
se
(
7
8
%
)
V
i
su
a
l
b
l
u
r
r
i
n
g
D
i
f
f
i
c
u
l
t
y
i
n
se
e
i
n
g
c
l
e
a
r
l
y
Tr
u
e
(
4
5
%
)
o
r
F
a
l
se
(
5
5
%
)
I
t
c
h
i
n
g
P
e
r
si
st
e
n
t
s
k
i
n
p
r
u
r
i
t
u
s
Tr
u
e
(
4
9
%
)
o
r
F
a
l
se
(
5
1
%
)
I
r
r
i
t
a
b
i
l
i
t
y
Emo
t
i
o
n
a
l
se
n
si
t
i
v
i
t
y
o
r
mo
o
d
sw
i
n
g
s
Tr
u
e
(
2
4
%
)
o
r
F
a
l
se
(
7
6
%
)
D
e
l
a
y
e
d
h
e
a
l
i
n
g
S
l
o
w
w
o
u
n
d
h
e
a
l
i
n
g
Tr
u
e
(
4
6
%
)
o
r
F
a
l
se
(
5
4
%
)
P
a
r
t
i
a
l
p
a
r
e
si
s
W
e
a
k
n
e
ss
o
r
p
a
r
a
l
y
s
i
s
o
f
m
u
sc
l
e
g
r
o
u
p
Tr
u
e
(
4
3
%
)
o
r
F
a
l
se
(
5
7
%
)
M
u
s
c
l
e
s
t
i
f
f
n
e
ss
R
e
d
u
c
e
d
f
l
e
x
i
b
i
l
i
t
y
i
n
m
u
sc
l
e
s
Tr
u
e
(
3
8
%
)
o
r
F
a
l
se
(
6
2
%
)
A
l
o
p
e
c
i
a
H
a
i
r
l
o
ss
a
n
d
h
o
r
mo
n
e
i
m
b
a
l
a
n
c
e
Tr
u
e
(
3
4
%
)
o
r
F
a
l
se
(
6
6
%
)
O
b
e
si
t
y
Ex
c
e
ss
b
o
d
y
f
a
t
Tr
u
e
(
1
7
%
)
o
r
F
a
l
se
(
8
3
%
)
C
l
a
s
s
B
i
n
a
r
y
t
a
r
g
e
t
i
n
d
i
c
a
t
i
n
g
d
i
a
b
e
t
i
c
o
r
n
o
t
P
o
si
t
i
v
e
(
6
2
%)
o
r
N
e
g
a
t
i
v
e
(
3
8
%)
T
ab
le
3
.
Descr
ip
tiv
e
f
ea
t
u
r
es
o
f
th
ir
d
d
ataset
A
t
t
r
i
b
u
t
e
D
e
scri
p
t
i
o
n
R
a
n
g
e
G
e
n
d
e
r
S
e
x
o
f
p
a
t
i
e
n
t
M
a
l
e
(
1
)
–
F
e
m
a
l
e
(
0
)
A
g
e
A
g
e
i
n
y
e
a
r
s
0
-
80
Ht
H
y
p
e
r
t
e
n
si
o
n
Y
e
s (1
)
–
N
o
(
0
)
Hd
H
e
a
r
t
D
i
sea
se
Y
e
s (1
)
–
N
o
(
0
)
S
mo
k
i
n
g
P
a
r
t
i
c
i
p
a
n
t
h
i
st
o
r
y
o
f
s
mo
k
i
n
g
C
u
r
r
e
n
t
,
n
e
v
e
r
,
f
o
r
mer
,
n
o
_
i
n
f
o
B
M
I
B
o
d
y
mas
s i
n
d
e
x
(
k
g
/
m)
1
0
.
0
1
–
9
5
.
6
9
H
b
A
1
c
G
l
y
c
a
t
e
d
H
e
m
o
g
l
o
b
i
n
:
a
b
l
o
o
d
t
e
s
t
t
h
a
t
mea
s
u
r
e
s
t
h
e
a
v
e
r
a
g
e
b
l
o
o
d
su
g
a
r
(
g
l
u
c
o
se)
l
e
v
e
l
s
o
v
e
r
t
h
e
p
a
st
2
-
3
m
o
n
t
h
s
3
.
5
-
9
B
g
l
B
l
o
o
d
g
l
u
c
o
s
e
l
e
v
e
l
8
0
-
300
d
i
a
b
e
t
e
s
B
i
n
a
r
y
t
a
r
g
e
t
i
n
d
i
c
a
t
i
n
g
d
i
a
b
e
t
i
c
o
r
n
o
t
D
i
a
b
e
t
i
c
(
1
)
–
non
-
D
i
a
b
e
t
i
c
(
0
)
T
ab
le
4
.
Descr
ip
tiv
e
f
ea
t
u
r
es
o
f
VOCADI
AB
A
t
t
r
i
b
u
t
e
D
e
scri
p
t
i
o
n
R
a
n
g
e
B
y
o
l
s
_
e
mb
e
d
d
i
n
g
s
N
u
meri
c
a
l
r
e
p
r
e
se
n
t
a
t
i
o
n
s
o
f
k
e
y
a
c
o
u
st
i
c
a
n
d
sp
e
e
c
h
c
h
a
r
a
c
t
e
r
i
s
t
i
c
s e
x
t
r
a
c
t
e
d
f
r
o
m
p
a
r
t
i
c
i
p
a
n
t
s’
v
o
i
c
e
r
e
c
o
r
d
i
n
g
s
-
G
e
n
d
e
r
S
e
x
o
f
p
a
r
t
i
c
i
p
a
n
t
M
a
l
e
(
1
)
–
F
e
m
a
l
e
(
0
)
A
g
e
A
g
e
i
n
y
e
a
r
s
1
8
-
81
B
M
I
B
o
d
y
mas
s i
n
d
e
x
(
k
g
/
m)
1
5
.
8
2
t
o
6
6
.
9
3
Et
h
n
i
c
i
t
y
R
a
c
e
o
f
p
a
r
t
i
c
i
p
a
n
t
La
t
i
n
o
–
w
h
i
t
e
–
b
l
a
c
k
–
mi
x
e
d
–
o
t
h
e
r
–
A
s
i
a
n
-
u
n
k
n
o
w
n
A
D
A
_
sco
r
e
D
i
a
b
e
t
e
s
-
r
e
l
a
t
e
d
sc
o
r
e
b
a
s
e
d
o
n
t
h
e
A
meric
a
n
D
i
a
b
e
t
e
s
A
sso
c
i
a
t
i
o
n
's
c
l
a
ssi
f
i
c
a
t
i
o
n
.
I
n
t
e
g
e
r
f
r
o
m
0
t
o
7
D
i
a
b
e
t
e
s
B
i
n
a
r
y
t
a
r
g
e
t
i
n
d
i
c
a
t
i
n
g
d
i
a
b
e
t
i
c
o
r
n
o
t
D
i
a
b
e
t
i
c
(
1
)
–
non
-
D
i
a
b
e
t
i
c
(
0
)
2
.
2
.
Da
t
a
prepro
ce
s
s
ing
Pre
p
r
o
ce
s
s
in
g
i
s
an
im
p
o
r
tan
t
b
u
ild
in
g
b
lo
ck
i
n
th
e
p
r
o
ce
s
s
o
f
d
ev
el
o
p
m
e
n
t
o
f
t
h
e
p
r
o
p
o
s
e
d
m
o
d
el
.
W
h
er
e
th
e
ef
f
icien
cy
o
f
th
e
p
r
ed
ictio
n
m
o
d
e
l
i
s
a
l
t
e
r
e
d
b
y
t
h
e
i
n
c
o
n
s
i
s
t
e
n
t
d
a
t
a
.
T
h
e
r
e
ar
e
s
o
m
e
s
er
io
u
s
o
b
s
er
v
atio
n
s
in
th
ese
d
atasets
s
u
ch
as
n
o
n
-
e
x
is
tin
g
v
alu
es
o
r
ze
r
o
v
alu
es,
n
o
m
in
al
f
ea
tu
r
es
an
d
tar
g
et
class
u
n
eq
u
al
d
is
tr
ib
u
tio
n
s
.
T
h
e
p
r
ep
r
o
ce
s
s
in
g
o
f
d
ata
is
im
p
le
m
en
ted
i
n
th
r
ee
d
i
f
f
e
r
e
n
t
s
tag
es
w
h
i
c
h
ar
e
class
im
b
alan
ce
,
h
a
n
d
lin
g
m
is
s
in
g
v
alu
es,
an
d
e
n
c
o
d
i
n
g
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
A
n
en
s
emb
le
ma
ch
in
e
lea
r
n
in
g
b
a
s
ed
mo
d
el
fo
r
p
r
ed
ictio
n
a
n
d
…
(
Mo
a
ta
z
Mo
h
a
med
E
l
S
h
erb
in
y
)
5351
2
.
2
.
1
.
Cla
s
s
im
ba
la
nce
C
las
s
im
b
alan
ce
o
cc
u
r
s
wh
en
th
er
e
ar
e
c
o
m
p
ar
ativ
el
y
m
o
r
e
s
am
p
les
in
o
n
e
class
o
f
a
d
ata
s
et
th
an
in
o
th
er
class
.
Ma
ch
in
e
lear
n
in
g
m
o
d
els
m
i
g
h
t
h
av
e
d
if
f
icu
lti
es
in
p
r
e
d
ictio
n
s
tag
e
b
ec
au
s
e
o
f
th
is
im
b
alan
ce
.
As
th
ey
ten
d
to
f
av
o
r
th
e
m
aj
o
r
ity
class
r
esu
ltin
g
in
a
b
iased
m
o
d
el
p
r
ed
ictio
n
a
n
d
m
is
le
ad
in
g
p
er
f
o
r
m
an
ce
,
p
ar
ticu
lar
ly
f
o
r
th
e
m
in
o
r
ity
class
[
1
6
]
.
C
las
s
im
b
alan
ce
i
s
s
u
es
ca
n
b
e
ad
d
r
ess
ed
in
s
e
v
er
al
way
s
s
u
ch
as
u
n
d
er
s
am
p
lin
g
th
e
d
o
m
in
a
n
t
class
o
r
o
v
er
s
am
p
lin
g
th
e
m
in
o
r
ity
class
.
Ad
d
itio
n
ally
,
d
ata
au
g
m
en
tatio
n
,
wh
ich
is
co
m
m
o
n
l
y
em
p
lo
y
ed
in
im
ag
e
d
atasets
.
All
d
ataset
s
im
p
lem
en
ted
in
th
is
s
tu
d
y
s
u
f
f
er
f
r
o
m
class
im
b
alan
ce
as
s
h
o
wn
in
Fig
u
r
e
2
ex
ce
p
t
f
o
r
VOCADI
A
B
wh
er
e
th
e
tar
g
et
class
d
is
tr
ib
u
tio
n
is
eq
u
al.
PID
D
in
clu
d
es
5
0
0
d
ia
b
etic
p
atien
ts
wh
ile
th
e
n
u
m
b
er
o
f
n
o
n
-
d
ia
b
etic
in
d
iv
id
u
als
is
2
6
8
.
T
h
e
th
ir
d
d
atas
et
co
n
tain
s
9
1
,
5
0
0
n
o
n
-
d
iab
et
ic
in
d
iv
id
u
als
an
d
o
n
ly
8
,
5
0
0
d
iab
etic
p
atien
ts
.
T
h
e
q
u
esti
o
n
n
ai
r
e
d
ataset
co
n
ta
in
s
3
2
0
d
ia
b
etics a
n
d
2
0
0
n
o
n
-
d
iab
etics.
Fig
u
r
e
2
.
Data
s
et
tar
g
et
class
d
is
tr
ib
u
tio
n
2
.
2
.
2
.
M
is
s
i
ng
v
a
lues
T
h
er
e
ar
e
s
o
m
e
s
er
io
u
s
o
b
s
er
v
atio
n
s
in
th
ese
d
atasets
s
u
ch
as
n
u
ll
o
r
ze
r
o
v
alu
es.
So
m
e
f
ea
tu
r
es,
s
u
ch
as Glu
co
s
e,
B
lo
o
d
Pre
s
s
u
r
e,
Sk
in
T
h
ick
n
ess
,
I
n
s
u
lin
,
an
d
B
MI
,
h
av
e
ze
r
o
v
alu
es,
wh
ich
ar
e
u
n
lik
ely
in
a
r
ea
l
-
wo
r
ld
m
ed
ical
s
ettin
g
.
F
o
r
e
x
am
p
le,
a
ze
r
o
B
MI
o
r
g
l
u
co
s
e
is
b
io
l
o
g
ically
im
p
o
s
s
ib
le.
Fo
r
a
v
ar
iety
o
f
r
ea
s
o
n
s
,
p
atien
ts
o
f
ten
n
eg
lec
t
m
u
ltip
le
r
eq
u
i
r
ed
test
s
.
T
h
e
r
e
f
o
r
e
,
n
o
n
-
e
x
is
tin
g
v
alu
es
w
ill
ap
p
ea
r
in
d
ata,
r
eq
u
ir
in
g
th
e
em
p
lo
y
m
e
n
t o
f
s
u
itab
le
im
p
u
tatio
n
tec
h
n
iq
u
es.
T
h
er
e
ar
e
m
u
ltip
le
ap
p
r
o
ac
h
es
to
h
an
d
le
n
o
n
-
e
x
is
tin
g
attr
ib
u
tes
lik
e
ex
ch
an
g
in
g
th
e
m
with
a
co
n
s
tan
t,
m
ea
n
,
m
ed
ian
an
d
m
o
s
t
f
r
eq
u
e
n
t.
Dea
lin
g
with
in
co
m
p
lete
m
e
d
ical
r
ec
o
r
d
s
ca
n
b
e
p
er
f
o
r
m
ed
th
r
o
u
g
h
d
if
f
er
en
t
m
et
h
o
d
s
[
2
2
]
.
Swap
p
in
g
o
u
t
m
is
s
in
g
f
ea
tu
r
es
with
a
co
n
s
tan
t
“z
er
o
”
h
as
n
o
ef
f
ec
t
o
n
th
e
p
r
ed
ictio
n
b
iasi
n
g
o
f
m
o
d
el.
On
th
e
o
t
h
er
h
an
d
,
th
is
ass
u
m
p
tio
n
is
b
io
lo
g
ically
im
p
o
s
s
ib
le.
Neg
lectin
g
in
co
m
p
lete
r
ec
o
r
d
b
y
s
im
p
l
y
r
em
o
v
in
g
th
em
ca
n
af
f
e
ct
s
m
al
l
-
s
ca
le
d
atasets
.
Oth
er
m
ath
e
m
atica
l
ap
p
r
o
ac
h
es
s
u
ch
as r
ep
lacin
g
n
o
n
-
ex
is
tin
g
v
alu
es with
a
co
n
s
tan
t,
m
ea
n
,
m
ed
ian
o
r
m
o
s
t f
r
e
q
u
en
t.
2
.
2
.
3
N
o
m
ina
l f
e
a
t
ures
T
h
e
m
ac
h
i
n
e
lear
n
in
g
al
g
o
r
it
h
m
n
ee
d
s
to
tr
a
n
s
f
o
r
m
n
o
m
in
al
v
alu
es
in
to
n
u
m
er
ical
v
alu
es
s
o
th
at
it
ca
n
co
m
p
r
eh
en
d
th
e
d
ata
it
r
ec
eiv
es
to
en
ab
le
f
u
r
t
h
er
p
r
o
c
ess
in
g
.
C
ateg
o
r
ical
v
a
r
i
a
b
l
e
s
wer
e
en
co
d
e
d
u
s
i
n
g
o
n
e
h
o
t
en
c
o
d
er
.
I
t
tr
an
s
f
o
r
m
s
ea
ch
u
n
iq
u
e
v
al
u
e
in
th
e
n
o
m
in
al
attr
ib
u
te
in
to
a
b
in
ar
y
v
ec
to
r
.
E
v
er
y
u
n
iq
u
e
v
alu
e
is
r
ep
r
esen
ted
b
y
a
v
ec
t
o
r
with
a
s
in
g
le
“1
”
in
d
icatin
g
th
e
p
r
esen
ce
o
f
th
at
ca
teg
o
r
y
wh
ile
th
e
r
em
ain
in
g
ca
teg
o
r
ies
ar
e
r
ep
r
esen
ted
b
y
“0
”.
E
n
co
d
in
g
is
cr
u
cial
b
ec
au
s
e
m
ac
h
in
e
lear
n
in
g
m
o
d
els
wo
r
k
s
wit
h
n
u
m
er
ical
d
ata,
n
o
t c
ateg
o
r
ica
l la
b
els.
2
.
3
.
K
-
F
o
ld cr
o
s
s
v
a
lid
a
t
io
n
C
r
o
s
s
v
alid
atio
n
an
d
tr
ain
-
test
s
p
lit
ar
e
tech
n
iq
u
es
u
s
ed
in
m
ac
h
in
e
lear
n
in
g
to
ev
al
u
ate
m
o
d
el
p
er
f
o
r
m
an
ce
.
Sin
ce
th
e
y
esti
m
ate
h
o
w
well
a
m
o
d
el
will
g
en
er
alize
to
u
n
s
ee
n
d
ata.
I
n
th
e
tr
ain
-
test
s
p
lit
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
15
,
No
.
6
,
Decem
b
e
r
20
25
:
5
3
4
7
-
5
3
5
9
5352
m
eth
o
d
,
th
e
d
ataset
is
d
iv
id
ed
in
to
two
p
ar
ts
.
T
h
e
tr
ain
s
et
wh
ich
is
u
s
ed
to
tr
ai
n
th
e
m
o
d
el,
an
d
th
e
test
Set
th
at
is
u
s
ed
t
o
ev
alu
ate
th
e
p
e
r
f
o
r
m
a
n
ce
o
n
u
n
k
n
o
wn
d
ata.
I
t
is
co
n
s
id
er
ed
th
e
s
im
p
lest
ap
p
r
o
ac
h
wh
er
e
d
ata
is
d
iv
id
ed
in
to
a
7
0
%
-
8
0
%
f
o
r
tr
ain
in
g
p
h
ase
an
d
a
2
0
%
-
3
0
%
f
o
r
test
in
g
p
h
ase.
I
t
is
s
im
p
le,
q
u
ick
,
ea
s
y
to
im
p
lem
en
t
an
d
c
o
m
p
u
tatio
n
ally
ef
f
icien
t.
I
t
wo
r
k
s
well
f
o
r
lar
g
e
d
atasets
.
On
t
h
e
o
t
h
er
h
an
d
,
t
h
er
e
is
a
h
ig
h
v
ar
ian
ce
wh
er
e
p
er
f
o
r
m
an
ce
d
ep
en
d
s
o
n
h
o
w
th
e
d
ata
is
s
p
lit.
T
h
e
r
esu
lts
m
ig
h
t
v
ar
y
with
d
if
f
er
en
t
r
a
n
d
o
m
s
ee
d
s
.
I
t is less
r
eliab
le
f
o
r
s
m
all
d
atasets
wh
er
e
th
e
s
in
g
le
s
p
lit m
ig
h
t n
o
t c
a
p
tu
r
e
t
h
e
v
ar
i
ab
ilit
y
in
th
e
d
ata.
T
h
e
d
ataset
is
d
iv
id
ed
in
to
a
n
u
m
b
er
“K”
o
f
s
u
b
s
ets
th
at
ar
e
ap
p
r
o
x
im
ately
eq
u
al
s
ize
in
cr
o
s
s
v
alid
atio
n
.
T
h
e
m
o
d
el
is
tr
ain
ed
an
d
test
ed
K
tim
es,
with
ea
ch
f
o
ld
u
s
ed
as
th
e
test
s
et
e
x
ac
tly
o
n
ce
an
d
th
e
r
em
ain
in
g
f
o
ld
s
as
th
e
tr
ain
in
g
s
et
as
s
h
o
wn
in
Fig
u
r
e
3
.
Al
l
d
ata
p
o
i
n
ts
ar
e
u
s
ed
f
o
r
b
o
t
h
tr
ain
in
g
an
d
s
o
it
is
co
n
s
id
er
ed
c
o
m
p
u
tatio
n
ally
e
x
p
en
s
iv
e
an
d
ca
n
b
e
tim
e
-
co
n
s
u
m
in
g
,
esp
ec
ially
f
o
r
la
r
g
e
d
atasets
o
r
co
m
p
lex
m
o
d
els.
Fig
u
r
e
3
.
Fiv
e
-
f
o
ld
cr
o
s
s
v
alid
atio
n
[
2
3
]
2
.
4
.
M
a
chine le
a
rning
Ma
ch
in
e
lear
n
in
g
(
ML
)
m
ak
e
u
s
e
o
f
m
ath
em
atica
l
an
d
s
tatis
tical
alg
o
r
ith
m
s
in
o
r
d
e
r
t
o
id
en
tify
p
atter
n
s
in
d
ata
s
o
th
at
it
ca
n
p
er
f
o
r
m
an
ac
cu
r
ate
an
d
p
r
ec
is
e
p
r
ed
ictio
n
s
[
2
4
]
.
ML
en
h
an
ce
th
eir
p
er
f
o
r
m
an
ce
o
v
er
tim
e
th
r
o
u
g
h
b
ein
g
ex
p
o
s
ed
to
m
o
r
e
d
ata
.
Su
p
er
v
is
ed
lear
n
in
g
tr
ai
n
s
o
n
lab
eled
d
ata
u
s
ed
in
class
if
icatio
n
as
in
o
u
r
ca
s
e.
T
h
is
s
tu
d
y
in
v
o
lv
es
an
e
n
s
em
b
le
o
f
m
ac
h
in
e
lear
n
i
n
g
class
if
ier
s
,
s
u
ch
as
r
an
d
o
m
f
o
r
est
(
R
F),
ex
tr
em
e
g
r
ad
ien
t
b
o
o
s
tin
g
(
XGB)
an
d
lo
g
is
tic
r
eg
r
ess
io
n
f
o
r
t
h
e
p
u
r
p
o
s
e
o
f
p
r
e
d
ictin
g
d
iab
etes
m
ellitu
s
.
2
.
4
.
1
.
L
o
g
is
t
ic
re
g
re
s
s
io
n
lo
g
is
tic
r
eg
r
ess
io
n
(
L
R
)
is
o
n
e
o
f
t
h
e
p
o
p
u
lar
s
u
p
er
v
is
ed
le
ar
n
in
g
alg
o
r
ith
m
s
i
n
h
ea
lth
ca
r
e
s
y
s
tem
s
.
I
t
is
k
n
o
wn
f
o
r
its
s
im
p
licity
an
d
ea
s
e
o
f
im
p
lem
e
n
tatio
n
,
m
ak
in
g
it
o
n
e
o
f
t
h
e
m
o
s
t
s
u
i
tab
le
alg
o
r
ith
m
s
f
o
r
b
in
ar
y
class
if
icatio
n
p
r
o
b
lem
s
.
T
h
e
L
R
u
s
es
a
co
llectio
n
o
f
in
d
ep
en
d
en
t
f
ea
tu
r
es
to
p
r
e
d
ict
th
e
lik
elih
o
o
d
o
f
th
e
class
o
u
tp
u
t
[
2
5
]
.
T
h
e
th
r
e
s
h
o
ld
u
s
ed
to
id
en
tify
wh
ich
d
ata
b
elo
n
g
s
to
a
p
a
r
ticu
lar
cl
ass
is
k
n
o
wn
as
th
e
d
ec
is
io
n
b
o
u
n
d
ar
y
[
2
6
]
.
T
h
e
lo
g
is
tic
s
ig
m
o
id
f
u
n
ctio
n
is
u
s
ed
to
g
et
th
is
ca
teg
o
r
izatio
n
p
r
o
b
ab
ilit
y
.
T
h
e
co
ef
f
icien
ts
o
f
L
R
p
r
o
v
id
e
cle
ar
in
s
ig
h
ts
in
to
th
e
r
elatio
n
s
h
ip
b
etwe
en
ea
ch
f
ea
tu
r
e
a
n
d
th
e
o
u
tco
m
e
class
.
2
.
4
.
2
.
Ra
nd
o
m
f
o
re
s
t
R
an
d
o
m
f
o
r
est
(
R
F)
cr
ea
tes
a
n
u
m
b
er
o
f
d
ec
is
io
n
tr
ee
s
an
d
g
iv
es
th
e
o
u
tp
u
t
class
o
f
ea
ch
tr
ee
in
th
e
tr
ain
in
g
p
h
ase
[
2
7
]
.
R
F
ca
n
h
an
d
le
a
lar
g
e
n
u
m
b
er
o
f
f
ea
tu
r
es
ev
en
if
th
ey
in
clu
d
e
m
is
s
i
n
g
d
ata,
m
ak
in
g
it
s
u
itab
le
f
o
r
r
e
al
-
wo
r
ld
d
atasets
.
Mo
r
eo
v
er
,
it
p
r
o
v
id
es
in
s
i
g
h
ts
to
f
ea
tu
r
e
im
p
o
r
tan
ce
t
h
at
d
eter
m
in
e
wh
ic
h
v
ar
iab
les
co
n
tr
ib
u
te
th
e
m
o
s
t
t
o
th
e
p
r
e
d
ictio
n
.
T
h
is
m
o
d
el
o
f
f
er
s
a
s
tr
aig
h
tf
o
r
wa
r
d
m
o
d
if
i
ca
tio
n
th
at
u
tili
ze
s
a
co
r
r
elate
d
tr
ee
in
th
e
b
ag
g
i
n
g
p
r
o
ce
s
s
,
th
is
.
A
ce
r
tain
am
o
u
n
t
o
f
attr
ib
u
tes
ar
e
ig
n
o
r
e
d
ac
r
o
s
s
all
co
lu
m
n
s
d
u
r
in
g
b
o
o
ts
tr
ap
p
in
g
[
2
8
]
.
T
h
is
tech
n
iq
u
e
aid
s
in
th
e
p
r
o
ce
s
s
o
f
r
ed
u
cin
g
v
ar
ia
n
ce
.
O
n
th
e
o
th
er
h
a
n
d
,
it
r
aises
th
e
p
r
o
b
ab
ilit
y
o
f
b
iasi
n
g
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
A
n
en
s
emb
le
ma
ch
in
e
lea
r
n
in
g
b
a
s
ed
mo
d
el
fo
r
p
r
ed
ictio
n
a
n
d
…
(
Mo
a
ta
z
Mo
h
a
med
E
l
S
h
erb
in
y
)
5353
2
.
4
.
3
.
E
x
t
re
m
e
g
ra
dient
bo
o
s
t
ing
An
ex
tr
em
e
g
r
ad
ien
t
b
o
o
s
tin
g
(
XGB
)
i
s
a
tr
ee
-
b
ased
s
eq
u
en
tial
DT
alg
o
r
ith
m
ap
p
lied
to
r
elativ
ely
s
m
all
o
r
m
ed
iu
m
s
ize
tab
u
lar
d
atasets
[
2
9
]
.
I
t
is
co
n
s
id
er
e
d
to
b
e
am
o
n
g
th
e
m
o
s
t
ef
f
ec
tiv
e
tech
n
iq
u
es
f
o
r
class
if
icatio
n
an
d
p
r
ed
ictio
n
.
I
t
is
k
n
o
wn
f
o
r
its
s
p
ee
d
an
d
p
er
f
o
r
m
an
ce
d
u
e
to
o
p
tim
ized
g
r
ad
ien
t
b
o
o
s
tin
g
alg
o
r
ith
m
s
.
B
y
co
m
b
in
in
g
co
m
p
ar
ativ
ely
wea
k
er
a
n
d
s
im
p
ler
m
o
d
els.
Scalab
ilit
y
is
co
n
s
id
er
ed
th
e
m
o
s
t
im
p
o
r
tan
t
f
ea
tu
r
e
in
XGB
[
3
0
]
,
wh
er
e
it
im
p
lem
en
t
lear
n
in
g
th
r
o
u
g
h
d
is
tr
ib
u
te
d
co
m
p
u
tin
g
an
d
m
em
o
r
y
u
s
ag
e
is
well
s
tr
u
ctu
r
ed
.
T
h
e
u
s
e
o
f
L
ass
o
an
d
R
id
g
e
r
eg
u
lar
izatio
n
aid
s
in
p
r
ev
en
tin
g
o
v
er
f
itti
n
g
.
XGB
ca
n
wo
r
k
with
d
if
f
er
en
t ty
p
es o
f
d
ata
m
ak
in
g
it v
e
r
s
atile
f
o
r
m
an
y
m
ed
ical
ap
p
licatio
n
s
.
2
.
4
.
4
.
E
ns
em
ble m
o
delin
g
Sin
ce
d
if
f
er
en
t
m
o
d
els
h
av
e
d
if
f
er
en
t
s
tr
en
g
th
s
an
d
wea
k
n
ess
es.
E
n
s
em
b
le
m
eth
o
d
s
u
tili
ze
th
e
co
llectiv
e
d
ec
is
io
n
o
f
m
u
ltip
l
e
b
ase
m
o
d
els
wh
ich
ar
e
m
o
r
e
r
o
b
u
s
t
an
d
ac
c
u
r
ate
t
h
an
a
n
y
in
d
i
v
id
u
al
m
o
d
el.
E
r
r
o
r
s
d
u
e
t
o
b
iasi
n
g
,
v
a
r
ian
ce
,
o
r
ev
en
n
o
is
e
in
t
h
e
d
ata
ca
n
b
e
m
i
n
im
ized
t
h
r
o
u
g
h
c
o
m
b
in
in
g
m
u
ltip
le
m
o
d
els
[
3
1
]
.
Pre
d
ictio
n
s
f
r
o
m
m
u
ltip
le
m
o
d
els
ar
e
av
er
a
g
ed
f
o
r
r
eg
r
ess
io
n
o
r
c
o
m
b
in
ed
d
u
e
to
m
ajo
r
ity
v
o
tin
g
f
o
r
class
if
icatio
n
as
in
o
u
r
ca
s
e.
E
n
s
em
b
le
m
eth
o
d
s
ca
n
b
etter
id
en
tify
a
n
d
u
tili
ze
im
p
o
r
tan
t
f
ea
tu
r
es.
Alth
o
u
g
h
in
d
iv
id
u
al
m
o
d
els
m
ay
b
e
co
m
p
u
tatio
n
ally
ef
f
i
cien
t,
en
s
em
b
les
ca
n
s
till
b
e
ef
f
icien
t
th
r
o
u
g
h
p
ar
allel
p
r
o
ce
s
s
in
g
o
r
o
p
tim
iz
ed
alg
o
r
ith
m
s
m
a
k
in
g
it c
o
m
p
ar
ed
to
tr
ain
i
n
g
a
s
in
g
le
c
o
m
p
l
ex
m
o
d
el.
2
.
5
.
E
v
a
lua
t
i
o
n m
ea
s
ures
Key
p
er
f
o
r
m
an
ce
m
etr
ics
in
class
if
icatio
n
task
s
in
clu
d
e
ac
cu
r
ac
y
,
p
r
ec
is
io
n
,
r
ec
all,
an
d
F1
-
s
co
r
e.
T
o
ca
lcu
late
th
ese
m
etr
ics,
r
ely
in
g
o
n
f
o
u
r
k
e
y
co
m
p
o
n
en
ts
:
tr
u
e
p
o
s
itiv
es
(
T
P),
f
alse
p
o
s
itiv
es
(
FP
)
,
tr
u
e
n
eg
ativ
es
(
T
N)
,
an
d
f
alse
n
eg
ativ
es
(
FN)
.
T
h
ese
co
m
p
o
n
en
t
s
ar
e
ty
p
ically
r
ep
r
esen
ted
in
a
co
n
f
u
s
io
n
m
at
r
ix
,
as illu
s
tr
ated
in
Fig
u
r
e
4
.
Fig
u
r
e
4
.
B
in
ar
y
class
co
n
f
u
s
io
n
m
atr
ix
Acc
u
r
ac
y
m
ea
s
u
r
es th
e
p
r
o
p
o
r
tio
n
o
f
c
o
r
r
ec
tly
class
if
ied
in
s
tan
ce
s
o
u
t o
f
th
e
to
tal
ca
s
es.
=
+
+
+
+
(
1
)
Pre
cisi
o
n
in
d
icate
s
th
e
r
atio
o
f
co
r
r
ec
tly
id
e
n
tifie
d
p
o
s
itiv
e
in
s
tan
ce
s
to
th
e
to
tal
p
r
ed
icte
d
p
o
s
itiv
es.
=
+
(
2
)
R
ec
all
r
ef
lects th
e
p
r
o
p
o
r
tio
n
o
f
ac
tu
al
p
o
s
itiv
e
ca
s
es th
at
wer
e
co
r
r
ec
tly
p
r
ed
icted
.
=
+
(
3
)
F1
-
Sco
r
e
is
th
e
weig
h
ted
a
v
er
ag
e
o
f
p
r
ec
is
io
n
an
d
r
ec
all.
1
−
=
2
∗
∗
+
(
4
)
2
.
6
.
E
x
perim
ent
a
l set
up
T
h
e
m
o
d
el
was
co
n
d
u
cte
d
o
n
th
e
Kag
g
le
p
latf
o
r
m
o
n
an
in
t
el
i7
-
10
th
g
en
er
atio
n
p
r
o
ce
s
s
o
r
.
T
h
e
co
d
e
was w
r
itten
in
Py
th
o
n
p
r
o
g
r
a
m
m
in
g
lan
g
u
ag
e.
T
h
e
s
cr
ip
t in
clu
d
es th
e
f
o
llo
win
g
k
ey
elem
en
ts
:
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
15
,
No
.
6
,
Decem
b
e
r
20
25
:
5
3
4
7
-
5
3
5
9
5354
a.
E
n
v
ir
o
n
m
en
tal
s
etu
p
:
I
m
p
o
r
ti
n
g
to
o
ls
an
d
lib
r
ar
ies
wh
ich
wer
e
im
p
lem
en
ted
in
t
h
is
s
tu
d
y
as
s
h
o
wn
in
Fig
u
r
e
5
an
d
illu
s
tr
ated
in
T
a
b
le
5
.
b.
Kag
g
le
c
o
n
f
i
g
u
r
atio
n
:
t
h
e
u
ti
lized
m
em
o
r
y
f
o
r
m
o
d
el
tr
ai
n
in
g
was
1
.
2
GiB
wh
ile
th
e
d
is
k
s
p
ac
e
was
2
.
3
GiB
.
Fu
r
th
er
m
o
r
e,
th
e
r
u
n
t
im
e
o
f
en
tire
c
o
d
e
was 1
,
2
7
5
s
ec
o
n
d
s
with
o
u
t a
cc
eler
at
o
r
.
c.
Hy
p
er
p
ar
a
m
eter
tu
n
in
g
f
o
r
i
m
p
lem
en
ted
m
ac
h
i
n
e
lear
n
in
g
m
o
d
els
as
s
h
o
wn
in
Fig
u
r
e
6
.
Valu
es
ar
e
d
is
cu
s
s
ed
in
T
ab
le
6
.
Fig
u
r
e
5
.
Scr
ee
n
s
h
o
t o
f
t
h
e
in
p
u
t lib
r
ar
ies an
d
to
o
ls
T
ab
le
5
.
I
m
p
o
r
te
d
lib
r
ar
ies
Li
b
r
a
r
y
I
mp
l
e
me
n
t
a
t
i
o
n
p
u
r
p
o
se
P
a
n
d
a
s
D
a
t
a
m
a
n
i
p
u
l
a
t
i
o
n
a
n
d
a
n
a
l
y
s
i
s
o
f
d
a
t
a
f
r
a
mes
,
h
a
n
d
l
i
n
g
s
t
r
u
c
t
e
d
d
a
t
a
,
c
l
e
a
n
i
n
g
d
a
t
a
a
n
d
t
r
a
n
sf
o
r
ma
t
i
o
n
.
N
u
mP
y
N
u
meri
c
a
l
c
o
m
p
u
t
i
n
g
f
o
r
m
u
l
t
i
d
i
m
e
n
si
o
n
a
l
a
r
r
a
y
s
a
n
d
ma
t
h
e
ma
t
i
c
a
l
o
p
e
r
a
t
o
r
s.
S
c
i
k
i
t
-
l
e
a
r
n
M
a
c
h
i
n
e
l
e
a
r
n
i
n
g
a
l
g
o
r
i
t
h
m f
o
r
c
l
a
ss
i
f
i
c
a
t
i
o
n
a
n
d
m
o
d
e
l
e
v
a
l
u
a
t
i
o
n
.
P
l
o
t
l
y
G
r
a
p
h
i
n
g
l
i
b
r
a
r
i
e
s f
o
r
q
u
a
l
i
t
y
v
i
s
u
a
l
i
z
a
t
i
o
n
.
Ti
me
M
e
a
su
r
i
n
g
a
n
d
m
a
n
a
g
i
n
g
t
i
me
-
r
e
l
a
t
e
d
t
a
s
k
s
i
n
p
r
o
g
r
a
ms.
Te
n
s
o
r
F
l
o
w
O
p
e
n
-
so
u
r
c
e
l
i
b
r
a
r
y
f
o
r
b
u
i
l
d
i
n
g
a
n
d
t
r
a
i
n
i
n
g
b
o
t
h
ma
c
h
i
n
e
a
n
d
d
e
e
p
l
e
a
r
n
i
n
g
m
o
d
e
l
s
.
Fig
u
r
e
6
.
Scr
ee
n
s
h
o
t
o
f
t
h
e
Ka
g
g
le
ML
m
o
d
els an
d
p
a
r
am
ete
r
tu
n
in
g
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
A
n
en
s
emb
le
ma
ch
in
e
lea
r
n
in
g
b
a
s
ed
mo
d
el
fo
r
p
r
ed
ictio
n
a
n
d
…
(
Mo
a
ta
z
Mo
h
a
med
E
l
S
h
erb
in
y
)
5355
T
ab
le
6
.
E
x
p
er
im
en
t
p
ar
am
ete
r
s
M
L
m
o
d
e
l
P
a
r
a
me
t
e
r
V
a
l
u
e
D
e
scri
p
t
i
o
n
LR
so
l
v
e
r
“
l
i
b
l
i
n
e
a
r
”
O
p
t
i
mi
z
a
t
i
o
n
a
l
g
o
r
i
t
h
m
.
RF
n
_
e
st
i
ma
t
o
r
s
1
0
0
N
u
mb
e
r
o
f
t
r
e
e
s
i
n
f
o
r
e
s
t
.
max
_
d
e
p
t
h
N
o
n
e
N
o
d
e
s s
p
r
e
a
d
t
i
l
l
e
v
e
r
y
l
e
a
f
i
s
p
u
r
e
t
o
e
n
s
u
r
e
e
a
c
h
l
e
a
f
n
o
d
e
r
e
p
r
e
se
n
t
s
a
d
i
s
t
i
n
c
t
c
l
a
ss
w
i
t
h
o
u
t
a
n
y
a
m
b
i
g
u
i
t
y
.
XGB
n
_
e
st
i
ma
t
o
r
s
1
0
0
N
u
mb
e
r
o
f
b
o
o
st
i
n
g
i
t
e
r
a
t
i
o
n
s
.
max
_
d
e
p
t
h
N
o
n
e
M
a
x
d
e
p
t
h
o
f
t
r
e
e
.
l
e
a
r
n
i
n
g
_
r
a
t
e
0
.
1
S
h
r
i
n
k
a
g
e
p
a
r
a
m
e
t
e
r
t
h
a
t
c
o
n
t
r
o
l
s
t
h
e
c
o
n
t
r
i
b
u
t
i
o
n
o
f
e
a
c
h
t
r
e
e
t
o
t
h
e
f
i
n
a
l
m
o
d
e
l
d
e
c
i
si
o
n
.
En
se
mb
l
e
e
st
i
mat
o
r
s
LR
,
R
F
,
X
G
B
Li
st
o
f
t
u
p
l
e
s w
h
e
r
e
e
a
c
h
e
st
i
ma
t
o
r
i
s
a
c
l
a
ss
i
f
i
e
r
.
v
o
t
i
n
g
“
H
a
r
d
”
M
a
j
o
r
i
t
y
v
o
t
i
n
g
c
l
a
ss
.
n
_
j
o
b
s
-
1
R
u
n
n
i
n
g
a
n
u
m
b
e
r
o
f
j
o
b
s f
o
r
f
i
t
t
i
n
g
a
n
d
p
r
e
d
i
c
t
i
o
n
i
n
p
a
r
a
l
l
e
l
w
h
e
r
e
-
1
mea
n
s a
l
l
p
r
o
c
e
sso
r
s
a
r
e
b
e
i
n
g
u
s
e
d
.
3.
RE
SU
L
T
S
AND
D
I
SCU
SS
I
O
N
A
co
n
s
id
er
ab
le
p
r
ep
r
o
ce
s
s
in
g
is
tak
en
in
to
ac
co
u
n
t
in
th
r
ee
m
ain
s
tep
s
af
ter
im
p
o
r
tin
g
lib
r
ar
ies
an
d
d
atasets
.
First
s
tep
,
d
ea
lin
g
with
class
im
b
alan
ce
is
s
u
es
t
h
r
o
u
g
h
a
p
p
ly
in
g
s
y
n
th
etic
m
i
n
o
r
ity
o
v
er
s
am
p
lin
g
tech
n
iq
u
e
(
SMOT
)
.
Seco
n
d
s
tep
is
r
e
p
lacin
g
n
o
n
-
ex
is
tin
g
v
alu
e
with
m
ea
n
v
alu
e
in
s
tead
o
f
s
im
p
ly
r
em
o
v
in
g
th
e
s
am
p
le
r
o
w
in
o
r
d
er
t
o
p
r
e
s
er
v
e
d
ataset
s
ize.
L
astl
y
,
c
o
n
v
er
tin
g
n
o
n
-
n
u
m
er
ic
f
ea
t
u
r
es
i
n
to
n
u
m
er
ic
o
n
e
b
y
ap
p
ly
in
g
o
n
e
-
h
o
t
e
n
co
d
e
r
.
Data
s
ets
a
r
e
d
iv
id
ed
in
to
tr
ain
in
g
an
d
test
in
g
p
a
r
titi
o
n
s
with
7
0
%
-
3
0
%
o
r
80%
-
2
0
%
th
en
5
-
f
o
ld
s
cr
o
s
s
v
alid
atio
n
is
ap
p
lied
.
A
n
en
s
em
b
le
m
o
d
el
f
o
r
th
r
ee
b
ase
c
lass
if
ier
s
wh
ich
ar
e
lo
g
is
tic
r
eg
r
ess
io
n
,
r
a
n
d
o
m
f
o
r
est
an
d
e
x
tr
em
e
g
r
ad
ien
t
b
o
o
s
tin
g
.
Ma
jo
r
ity
v
o
tin
g
was
ch
o
s
en
in
th
e
p
r
o
p
o
s
ed
p
r
ed
ictio
n
m
o
d
el.
R
esu
lts
ar
e
co
m
p
ar
ed
u
s
in
g
ac
cu
r
ac
y
,
p
r
e
cisi
o
n
,
r
ec
all
an
d
F1
-
s
co
r
e
as
s
h
o
wn
in
T
ab
le
7
.
Ad
d
itio
n
ally
,
co
n
f
u
s
io
n
m
atr
i
ce
s
ar
e
illu
s
tr
ated
in
Fig
u
r
es
7
to
1
0
,
(
a)
7
0
%
tr
ain
an
d
3
0
%
test
an
d
(
b
)
8
0
%
tr
ain
an
d
3
0
%
test
.
C
o
m
p
ar
is
o
n
b
etwe
en
th
e
p
r
o
p
o
s
ed
m
o
d
el
an
d
p
r
e
v
io
u
s
r
elate
d
wo
r
k
is
s
u
m
m
ar
ized
in
T
ab
le
8
.
T
h
e
en
s
em
b
le
tech
n
iq
u
e
co
m
b
in
es
th
e
p
r
ed
ictio
n
s
o
f
m
u
ltip
le
in
d
iv
id
u
al
m
o
d
els
to
cr
ea
te
a
m
o
r
e
r
o
b
u
s
t
an
d
ac
c
u
r
ate
p
r
e
d
ictio
n
m
eth
o
d
.
T
h
is
im
p
r
o
v
em
en
t
ar
is
es
b
ec
au
s
e
en
s
em
b
le
m
eth
o
d
s
lev
er
ag
e
th
e
s
tr
en
g
th
s
o
f
in
teg
r
ated
m
o
d
els
wh
ile
m
in
im
izin
g
th
eir
wea
k
p
o
in
ts
.
C
o
m
b
in
in
g
m
o
d
els
with
co
m
p
lem
en
tar
y
ca
p
ab
ilit
ies
allo
ws
ca
p
tu
r
in
g
b
o
th
lin
ea
r
an
d
n
o
n
-
lin
ea
r
r
elatio
n
s
h
ip
s
an
d
p
atter
n
s
,
wh
er
e
L
R
ca
p
tu
r
es
s
im
p
le
p
atter
n
s
wh
ile
R
F
h
an
d
les
n
o
n
-
lin
ea
r
in
ter
ac
tio
n
s
.
A
d
d
i
tio
n
ally
,
XGB
f
o
cu
s
es
o
n
m
is
class
if
ied
d
ata.
Stan
d
alo
n
e
R
F
h
as
s
o
m
e
b
ias
es
d
ep
en
d
in
g
o
n
th
e
d
e
p
th
o
f
tr
ee
s
.
L
in
ea
r
L
R
m
o
d
els
h
a
v
e
h
ig
h
b
ias
wh
en
p
r
o
b
lem
s
ar
e
n
o
n
-
lin
ea
r
s
u
ch
as in
o
u
r
ca
s
e.
Ho
wev
e
r
,
o
v
er
f
itti
n
g
m
ig
h
t b
e
a
s
er
io
u
s
is
s
u
e.
T
ab
le
7
.
E
n
s
em
b
le
m
o
d
el
r
esu
lts
o
n
all
d
atasets
Tr
a
i
n
-
Te
s
t
S
p
l
i
t
D
a
t
a
s
e
t
A
c
c
u
r
a
c
y
P
r
e
c
i
s
i
o
n
R
e
c
a
l
l
F1
-
sc
o
r
e
8
0
-
2
0
%
P
I
D
D
8
1
%
8
0
%
8
3
%
8
1
%
Q
u
e
st
i
o
n
n
a
i
r
e
9
5
%
9
7
%
9
4
%
9
5
%
Th
i
r
d
d
a
t
a
s
e
t
9
6
.
8
8
%
8
9
.
8
5
%
7
9
.
6
6
%
7
1
.
5
5
%
V
O
C
A
D
I
B
9
0
.
9
8
%
8
9
.
4
7
%
9
0
.
2
7
9
1
.
0
7
%
7
0
–
3
0
%
P
I
D
D
8
2
%
8
1
%
8
4
%
8
2
%
Q
u
e
st
i
o
n
n
a
i
r
e
9
6
%
9
7
%
9
5
%
9
6
%
Th
i
r
d
d
a
t
a
s
e
t
9
6
.
8
3
%
8
9
.
6
7
%
7
9
.
1
7
%
7
0
.
8
7
%
V
O
C
A
D
I
B
9
2
.
3
5
%
9
5
.
1
8
%
9
1
.
8
6
%
8
8
.
7
6
%
(
a)
(
b
)
(
a)
(
b
)
Fig
u
r
e
7
.
C
o
n
f
u
s
io
n
m
atr
i
x
f
o
r
th
e
en
s
em
b
le
m
o
d
el
o
n
f
ir
s
t d
ataset
(
a)
7
0
% tr
ain
a
n
d
3
0
% test an
d
(
b
)
8
0
% tr
ain
an
d
3
0
% test
Fig
u
r
e
8
.
C
o
n
f
u
s
io
n
m
atr
i
x
f
o
r
th
e
en
s
em
b
le
m
o
d
el
o
n
s
ec
o
n
d
d
ataset
(
a)
7
0
% tr
ain
an
d
3
0
% test an
d
(
b
)
8
0
% tr
ain
an
d
3
0
% test
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
15
,
No
.
6
,
Decem
b
e
r
20
25
:
5
3
4
7
-
5
3
5
9
5356
(
a)
(
b
)
(
a)
(
b
)
Fig
u
r
e
9
.
C
o
n
f
u
s
io
n
m
atr
i
x
f
o
r
th
e
en
s
em
b
le
m
o
d
el
o
n
th
ir
d
d
ataset
(
a)
7
0
% tr
ain
an
d
3
0
% test an
d
(
b
)
8
0
% tr
ain
an
d
3
0
% test
Fig
u
r
e
1
0
.
C
o
n
f
u
s
io
n
m
atr
ix
f
o
r
th
e
en
s
em
b
le
m
o
d
el
o
n
f
o
u
r
th
d
ataset
(
a)
7
0
% tr
ain
an
d
3
0
% test
an
d
(
b
)
8
0
% tr
ai
n
an
d
3
0
% test
T
ab
le
8
.
Pro
p
o
s
ed
m
o
d
el
r
esu
l
ts
v
er
s
u
s
r
elate
d
wo
r
k
in
liter
a
tu
r
e
r
ev
iew
A
u
t
h
o
r
s
D
a
t
a
s
e
t
Te
c
h
n
i
q
u
e
A
c
c
u
r
a
c
y
F
e
b
r
i
a
n
e
t
a
l
.
[
8
]
P
i
ma
I
n
d
i
a
n
D
i
a
b
e
t
e
s
d
a
t
a
se
t
(PIDD)
K
N
N
NB
7
7
.
9
2
%
7
8
.
5
2
%
K
a
n
g
r
a
a
n
d
S
i
n
g
h
[
9
]
NB
K
N
N
S
V
M
DT
RF
LR
7
2
.
6
%
6
6
.
1
%
7
4
.
3
%
7
1
.
8
%
6
4
.
9
%
7
4
%
C
h
a
n
g
e
t
a
l
.
[
1
0
]
RF
N
B
a
n
d
f
e
a
t
u
r
e
s
e
l
e
c
t
i
o
n
(
3
-
F
a
c
t
o
r
)
N
B
a
n
d
f
e
a
t
u
r
e
s
e
l
e
c
t
i
o
n
(
5
-
F
a
c
t
o
r
)
7
9
.
5
7
%
7
9
.
1
3
%
7
7
.
8
3
%
M
u
s
h
t
a
q
e
t
a
l
.
[
1
1
]
S
t
a
n
d
a
l
o
n
e
R
F
En
se
mb
l
e
(
b
a
l
a
n
c
e
d
d
a
t
a
se
t
)
8
0
.
7
%
8
1
.
7
%
R
a
w
a
t
e
t
a
l
.
[
1
2
]
A
d
a
B
o
o
st
7
9
.
6
9
%
B
a
r
i
k
e
t
a
l
.
[
1
3
]
RF
XGB
7
1
.
9
%
7
4
.
1
%
P
a
l
i
m
k
a
r
e
t
a
l
.
[
1
4
]
C
a
se
st
u
d
y
d
a
t
a
set
LR
S
V
M
NB
A
d
a
B
o
o
st
9
3
.
5
9
%
9
4
.
2
3
%
9
1
.
0
2
%
9
4
.
8
7
%
F
a
g
h
e
r
a
z
z
i
e
t
a
l
.
[
1
5
]
V
O
C
A
D
I
A
B
F
e
mal
e
g
r
o
u
p
-
LR
F
e
mal
e
g
r
o
u
p
-
M
LP
F
e
mal
e
g
r
o
u
p
-
S
V
M
M
a
l
e
g
r
o
u
p
-
LR
M
a
l
e
g
r
o
u
p
-
S
V
M
M
a
l
e
g
r
o
u
p
-
M
LP
6
7
%
6
3
%
5
7
%
6
9
%
7
0
%
7
1
%
K
a
u
f
ma
n
e
t
a
l
.
[
1
6
]
V
o
i
c
e
r
e
c
o
r
d
s
d
a
t
a
s
e
t
LR
(
w
o
me
n
–
v
o
i
c
e
f
e
a
t
u
r
e
s)
LR
(
w
o
me
n
–
a
l
l
f
e
a
t
u
r
e
s)
N
B
(
me
n
–
v
o
i
c
e
)
N
B
(
me
n
–
a
l
l
f
e
a
t
u
r
e
s)
7
0
%
8
2
%
6
9
%
8
6
%
P
r
o
p
o
se
d
m
o
d
e
l
P
I
D
D
En
se
mb
l
e
o
f
L
R
,
R
F
a
n
d
X
G
B
8
2
%
C
a
se
st
u
d
y
d
a
t
a
set
9
6
%
Th
i
r
d
d
a
t
a
s
e
t
9
6
.
8
3
%
V
O
C
A
D
I
A
B
9
2
.
3
5
%
4.
CO
NCLU
SI
O
N
T
h
e
p
r
ed
ictio
n
o
f
d
ia
b
etes
m
ellitu
s
is
co
n
s
id
er
ed
a
ch
allen
g
in
g
m
ed
ical
r
esear
ch
to
p
ic.
T
h
i
s
r
esear
ch
in
v
o
lv
ed
t
h
e
d
ev
el
o
p
m
en
t
o
f
a
m
ac
h
in
e
lear
n
i
n
g
-
b
ased
p
ip
el
in
e
f
o
r
th
e
p
r
o
ce
s
s
o
f
p
r
ed
ictin
g
d
iab
etes m
ellitu
s
d
ep
en
d
i
n
g
o
n
f
o
u
r
d
i
f
f
er
en
t
d
atasets
.
T
h
ese
d
ataset
s
h
av
e
s
er
io
u
s
o
b
s
er
v
atio
n
s
s
u
ch
as
clas
s
im
b
alan
ce
,
m
is
s
in
g
v
alu
es
in
ad
d
itio
n
to
ca
teg
o
r
ical
f
ea
tu
r
es.
T
r
ai
n
in
g
an
d
test
in
g
wer
e
p
e
r
f
o
r
m
ed
b
y
ap
p
l
y
in
g
5
-
f
o
ld
cr
o
s
s
v
alid
atio
n
.
C
o
n
s
eq
u
e
n
tly
,
o
u
r
g
o
al
was
m
et
b
y
ap
p
ly
i
n
g
L
R
,
R
F
an
d
XGB
in
a
n
e
n
s
em
b
le
m
o
d
el.
T
h
e
p
r
o
p
o
s
ed
m
o
d
el
y
ield
e
d
r
esu
lt
s
wh
ich
ar
e
s
u
p
er
io
r
to
th
o
s
e
o
f
o
th
e
r
s
tu
d
ies
in
liter
atu
r
e
r
e
v
iew
r
ea
ch
in
g
8
2
%
8
1
%,
8
4
%,8
2
%
in
ter
m
s
o
f
a
cc
u
r
ac
y
,
p
r
ec
is
io
n
,
r
ec
all
an
d
F1
-
s
co
r
e
r
esp
ec
tiv
ely
o
n
th
e
PID
D.
T
h
e
r
esu
lts
wer
e
9
2
.
3
5
%,
9
5
.
1
8
%,
9
1
.
8
6
%
an
d
8
8
.
7
6
f
o
r
ac
cu
r
ac
y
,
p
r
ec
is
io
n
,
r
ec
all
an
d
F1
-
s
co
r
e
r
esp
ec
tiv
ely
wh
en
ap
p
ly
in
g
p
er
f
o
r
m
a
n
ce
m
etr
ics
v
o
ca
l
d
ataset.
T
h
e
h
ig
h
est
r
e
s
u
lts
ar
e
9
6
.
8
8
%,
8
9
.
8
5
%,
7
9
.
6
6
%,
an
d
7
1
.
5
5
%
o
n
th
e
th
ir
d
d
ataset.
R
esu
lts
wer
e
9
6
%,
9
7
%,
9
5
%
an
d
9
6
%
f
o
r
ac
cu
r
ac
y
,
p
r
ec
is
io
n
,
r
ec
all
an
d
F1
-
s
co
r
e
r
esp
ec
tiv
ely
o
n
q
u
esti
o
n
n
air
e
d
ataset.
I
n
f
u
tu
r
e
wo
r
k
,
it
is
s
u
g
g
ested
to
ap
p
ly
d
if
f
er
en
t
n
o
n
-
ex
is
tin
g
v
al
u
e
im
p
u
tatio
n
tech
n
iq
u
es
clo
s
e
to
r
ea
l
-
life
s
itu
atio
n
s
in
a
d
d
itio
n
to
v
ar
i
o
u
s
class
im
b
alan
ce
tech
n
i
q
u
es.
Fu
r
th
er
m
o
r
e
,
m
o
r
e
m
ac
h
in
e
lear
n
in
g
a
n
d
d
ee
p
lear
n
i
n
g
tech
n
iq
u
es will b
e
ap
p
lied
o
n
h
y
b
r
id
d
atasets
.
Evaluation Warning : The document was created with Spire.PDF for Python.