I
nd
o
ne
s
ia
n
J
o
urna
l o
f
E
lect
rica
l En
g
ineering
a
nd
Co
m
pu
t
er
Science
Vo
l.
40
,
No
.
1
,
Octo
b
er
2
0
2
5
,
p
p
.
1
64
~
1
72
I
SS
N:
2
5
0
2
-
4
7
5
2
,
DOI
: 1
0
.
1
1
5
9
1
/ijeecs.v
40
.i
1
.
pp
1
64
-
1
72
164
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ee
cs.ia
esco
r
e.
co
m
Autoenco
der
-
ba
sed
G
a
uss
ia
n
mi
x
t
ure mo
del
for dia
g
no
sing
ea
rly
ons
et
o
f
dia
betic
r
etinopa
thy
P
riy
a
nk
a
S
re
eniv
a
s
,
K
a
v
it
a
V
.
H
o
ra
di
,
K
a
lpa
Ra
j
a
s
hek
a
r
D
e
p
a
r
t
me
n
t
o
f
C
o
mp
u
t
e
r
S
c
i
e
n
c
e
a
n
d
En
g
i
n
e
e
r
i
n
g
,
B
.
N
.
M
I
n
st
i
t
u
t
e
o
f
Te
c
h
n
o
l
o
g
y
,
B
e
n
g
a
l
u
r
u
,
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
Feb
7
,
2
0
2
5
R
ev
is
ed
Ap
r
8
,
2
0
2
5
Acc
ep
ted
J
u
l
4
,
2
0
2
5
Th
e
c
u
rre
n
t
st
u
d
y
p
re
se
n
ts
a
si
m
p
li
fied
y
e
t
in
n
o
v
a
ti
v
e
so
l
u
ti
o
n
to
wa
rd
s
e
ffe
c
ti
v
e
e
a
rly
d
iag
n
o
sis
o
f
d
iab
e
ti
c
re
ti
n
o
p
a
th
y
(DR)
th
a
t
lea
d
s
to
irrev
e
rsib
le
b
li
n
d
n
e
ss
.
A
re
v
iew
o
f
c
u
rre
n
t
l
it
e
ra
tu
re
sh
o
ws
a
c
o
n
sid
e
ra
b
le
n
u
m
b
e
r
o
f
m
a
c
h
in
e
lea
rn
i
n
g
a
n
d
d
e
e
p
lea
rn
in
g
a
p
p
r
o
a
c
h
e
s
h
a
v
e
b
e
e
n
p
re
se
n
ted
;
h
o
we
v
e
r,
th
e
re
a
re
sig
n
ifi
c
a
n
t
iss
u
e
s
with
th
e
e
a
rly
d
e
tec
ti
o
n
o
f
DR.
He
n
c
e
,
t
h
e
p
ro
p
o
se
d
stu
d
y
d
e
p
l
o
y
s
a
n
o
v
e
l
a
rc
h
it
e
c
tu
re
u
si
n
g
a
n
a
u
to
e
n
c
o
d
e
r
t
h
a
t
e
x
trac
ts
a
h
id
d
e
n
re
p
re
se
n
tati
o
n
o
f
re
ti
n
a
l
ima
g
e
s
wh
il
e
b
in
a
ry
c
la
ss
ifi
c
a
ti
o
n
is
c
a
rried
o
u
t
u
si
n
g
a
G
a
u
ss
ian
m
ix
t
u
re
m
o
d
e
l
.
T
h
e
p
rime
c
o
n
t
rib
u
ti
o
n
is
t
h
e
j
o
in
t
i
n
teg
ra
ti
o
n
o
f
d
e
e
p
lea
rn
in
g
wi
th
sta
ti
stica
l
m
o
d
e
ll
in
g
t
o
wa
rd
s
e
fficie
n
t
fe
a
tu
re
e
x
trac
ti
o
n
a
n
d
a
n
o
m
a
ly
d
e
tec
ti
o
n
,
su
p
p
o
rti
n
g
e
a
rly
d
e
term
in
a
ti
o
n
o
f
DR.
Th
e
stu
d
y
o
u
tco
m
e
sh
o
ws
a
p
ro
p
o
se
d
sy
ste
m
to
sig
n
ifi
c
a
n
tl
y
e
x
h
ib
i
t
9
6
.
5
%
a
c
c
u
ra
c
y
,
9
4
.
2
%
se
n
sit
iv
it
y
,
a
n
d
9
8
.
3
%
s
p
e
c
ifi
c
it
y
o
n
two
sta
n
d
a
rd
b
e
n
c
h
m
a
rk
e
d
d
a
tas
e
ts
in
c
o
m
p
a
riso
n
t
o
e
x
isti
n
g
m
o
d
e
ls f
re
q
u
e
n
t
ly
u
se
d
f
o
r
th
e
d
iag
n
o
sis
o
f
DR.
K
ey
w
o
r
d
s
:
C
las
s
if
icatio
n
Dee
p
l
ea
r
n
in
g
Dete
ctio
n
Diab
etic
r
etin
o
p
ath
y
Ma
ch
in
e
l
ea
r
n
in
g
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
:
Priy
an
k
a
Sre
en
iv
as
Dep
ar
tm
en
t o
f
C
o
m
p
u
ter
Scie
n
ce
an
d
E
n
g
in
ee
r
in
g
,
B
.
N.
M.
I
n
s
titu
te
o
f
T
ec
h
n
o
lo
g
y
B
an
ash
an
k
ar
i Stag
e
I
I
,
B
an
ash
an
k
ar
i
,
B
en
g
alu
r
u
,
I
n
d
ia
E
m
ail:
p
r
iy
an
k
as@
b
n
m
it.in
,
p
r
iy
an
k
as.v
a
h
in
@
g
m
ail.
co
m
1.
I
NT
RO
D
UCT
I
O
N
Diab
etic
r
etin
o
p
at
h
y
(
DR
)
is
co
n
s
id
er
ed
o
n
e
o
f
th
e
c
r
itical
m
ed
ical
co
m
p
licatio
n
s
o
f
d
ia
b
etes
th
at
lead
s
to
ir
r
e
v
er
s
ib
le
v
is
io
n
lo
s
s
[
1
]
.
T
h
e
p
r
o
g
r
ess
io
n
o
f
DR
is
ch
ar
ac
ter
ized
b
y
v
ar
io
u
s
s
tag
es
,
r
ig
h
t
f
r
o
m
m
il
d
s
co
r
e
to
v
er
y
h
ig
h
s
co
r
e
o
f
s
ev
er
ity
.
T
h
is
p
r
o
g
r
ess
io
n
c
an
b
e
s
lo
wed
d
o
wn
b
y
v
itre
cto
m
y
,
a
n
ti
-
VE
GF
in
jectio
n
,
laser
th
er
ap
y
,
etc.
Ho
wev
er
,
th
e
m
o
s
t
ch
allen
g
i
n
g
p
ar
t
o
f
th
is
m
ed
ical
co
n
d
itio
n
is
th
at
it
i
s
v
er
y
p
r
o
b
lem
atic
t
o
id
en
tif
y
in
its
ea
r
ly
s
tag
es
o
win
g
t
o
its
asy
m
p
to
m
atic
f
o
r
m
.
T
h
is
is
th
e
m
ajo
r
r
ea
s
o
n
th
at
th
e
m
ajo
r
ity
o
f
c
o
n
v
e
n
tio
n
al
s
cr
e
en
in
g
f
o
r
o
cu
lar
d
is
ea
s
es
f
ails
to
id
en
tif
y
DR
in
its
ea
r
l
y
s
tag
es.
T
h
er
e
ar
e
v
ar
io
u
s
s
cr
ee
n
in
g
m
eth
o
d
s
f
o
r
DR
in
cu
r
r
en
t
tim
es
in
m
o
d
er
n
clin
ical
s
etu
p
s
,
v
iz.
,
f
u
n
d
u
s
p
h
o
to
g
r
ap
h
y
,
ex
am
in
atio
n
o
f
d
ilated
f
u
n
d
u
s
,
o
p
ti
ca
l
co
h
er
e
n
ce
to
m
o
g
r
ap
h
y
(
OC
T
)
,
f
lu
o
r
escein
a
n
g
io
g
r
a
p
h
y
an
d
au
to
m
ate
d
r
etin
al
s
cr
ee
n
in
g
s
y
s
tem
[
2
]
.
All
th
ese
ex
is
tin
g
s
cr
ee
n
in
g
a
p
p
r
o
ac
h
es
ar
e
n
o
n
-
in
v
asiv
e
with
h
ig
h
-
r
eso
lu
tio
n
,
d
etailed
im
ag
es;
h
o
we
v
er
,
t
h
er
e
ar
e
c
h
allen
g
es
ass
o
ciate
d
with
alm
o
s
t
all
o
f
th
em
.
Fo
r
ex
am
p
le,
f
u
n
d
u
s
p
h
o
to
g
r
ap
h
y
a
n
d
d
ir
ec
t o
p
h
th
a
lm
o
s
co
p
y
a
r
e
n
o
t
s
u
itab
le
f
o
r
p
atien
t
wh
o
h
as
o
t
h
er
o
cu
lar
c
o
n
d
itio
n
s
,
a
n
d
th
ey
d
em
an
d
p
u
p
il
b
e
d
ilated
d
u
r
i
n
g
ex
am
in
atio
n
(
wh
ich
m
ay
n
o
t
b
e
s
u
itab
le
f
o
r
m
an
y
p
atien
ts
)
.
OC
T
d
em
an
d
s
h
ig
h
ly
s
p
ec
ialized
d
ev
ices
a
n
d
y
et
t
h
ey
h
av
e
lim
itatio
n
s
in
d
etec
tin
g
t
h
e
ea
r
l
y
o
n
s
et
o
f
DR
u
n
til
an
d
u
n
less
th
er
e
ar
e
n
o
p
o
ten
tial
ch
a
n
g
e
s
o
b
s
er
v
ed
in
th
e
m
ac
u
lar
o
r
r
etin
al
ar
ea
.
Flu
o
r
escein
an
g
io
g
r
ap
h
y
is
in
v
asiv
e
an
d
th
er
e
f
o
r
e
it
ca
n
n
o
t
b
e
ex
p
ec
ted
to
b
e
u
s
ed
r
o
u
tin
ely
f
o
r
p
atien
ts
.
Au
t
o
m
ated
r
etin
a
l
s
cr
ee
n
in
g
s
y
s
tem
s
u
f
f
er
s
f
r
o
m
u
n
a
v
o
id
ab
le
o
u
tlier
s
an
d
d
em
an
d
c
o
n
s
is
ten
t
u
p
d
ates
,
wh
ile
t
h
er
e
is
a
h
ig
h
e
r
co
s
t
o
f
eq
u
ip
m
en
t
f
o
r
r
etin
al
s
ca
n
n
in
g
tech
n
o
lo
g
y
.
As
a
p
ar
t
o
f
th
e
ev
o
lv
i
n
g
s
o
lu
tio
n
,
it
is
n
o
ted
th
at
ar
tific
ial
in
te
llig
en
ce
(
AI
)
h
as
b
ee
n
a
s
ig
n
if
ica
n
t
co
n
tr
i
b
u
to
r
to
war
d
s
m
ed
ical
im
ag
e
p
r
o
ce
s
s
in
g
u
s
in
g
its
m
ac
h
i
n
e
lear
n
in
g
(
ML
)
an
d
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2502
-
4
7
5
2
A
u
to
en
co
d
er
-
b
a
s
ed
G
a
u
s
s
ia
n
mixtu
r
e
mo
d
el
fo
r
d
ia
g
n
o
s
in
g
ea
r
ly
o
n
s
et
…
(
P
r
iya
n
ka
S
r
ee
n
iva
s
)
165
d
ee
p
lear
n
i
n
g
(
DL
)
alg
o
r
ith
m
s
[
3
]
.
T
h
e
ad
o
p
tio
n
o
f
AI
m
o
d
els
ca
n
b
e
u
s
ed
f
o
r
an
aly
zin
g
r
etin
al
im
ag
es
to
id
en
tify
v
ar
io
u
s
a
b
n
o
r
m
alities
an
d
h
e
n
ce
d
ia
g
n
o
s
tic
p
r
o
ce
s
s
is
ac
ce
ler
ated
wh
ile
AI
m
in
im
izes
its
d
ep
en
d
e
n
cy
o
n
h
u
m
an
e
x
p
er
t
s
.
Hea
lth
ca
r
e
p
r
o
v
i
d
er
s
ar
e
as
s
is
ted
with
v
ar
io
u
s
v
alu
ab
le
c
lin
ical
s
u
g
g
esti
o
n
s
b
ased
o
n
d
ata
an
aly
ze
d
f
o
r
b
e
tter
an
d
f
aster
d
ec
is
io
n
-
m
ak
i
n
g
.
Hen
ce
,
AI
o
f
f
er
s
in
cr
ea
s
ed
ef
f
icien
cy
,
r
e
d
u
ce
d
h
u
m
an
er
r
o
r
,
a
n
d
r
em
o
te
s
cr
ee
n
in
g
to
war
d
s
a
d
iag
n
o
s
is
o
f
DR
.
Fu
r
th
er
,
ML
is
a
s
u
b
s
et
o
f
AI
an
d
f
o
r
m
u
lates
d
ec
is
io
n
s
b
y
lear
n
in
g
p
atter
n
s
o
f
s
cr
ee
n
ed
d
ata.
T
h
e
r
etin
al
im
ag
es c
an
b
e
class
if
ied
b
y
M
L
b
ased
o
n
d
if
f
er
e
n
t
s
tag
es
o
f
DR
wh
ile
it
also
f
ac
i
litates
d
if
f
er
en
tiatin
g
DR
f
r
o
m
o
th
er
o
cu
lar
d
is
ea
s
es.
Hen
ce
,
ML
co
n
tr
ib
u
tes
to
en
h
an
ce
d
p
r
ed
ictio
n
o
f
r
is
k
with
im
p
r
o
v
e
d
d
iag
n
o
s
is
,
s
ca
lab
ilit
y
,
an
d
p
e
r
s
o
n
alize
d
ca
r
e.
Fin
ally
,
DL
is
an
ad
v
an
ce
d
s
u
b
s
et
o
f
th
e
ML
a
lg
o
r
ith
m
th
at
u
s
es
m
an
y
la
y
e
r
s
to
in
cr
ea
s
e
th
e
lear
n
in
g
ca
p
ab
ilit
y
f
o
r
g
ain
in
g
co
m
p
lex
in
s
ig
h
ts
.
Fre
e
f
r
o
m
an
y
m
a
n
u
al
ex
t
r
ac
tio
n
o
f
f
ea
tu
r
es
o
r
p
r
ep
r
o
ce
s
s
in
g
,
r
eti
n
al
im
ag
es
ca
n
b
e
d
ir
ec
tly
s
u
b
jecte
d
to
DL
.
DL
is
ca
p
ab
le
o
f
au
to
n
o
m
o
u
s
ly
id
e
n
tify
in
g
th
e
ap
p
r
o
p
r
iate
f
ea
tu
r
es
e.
g
.
,
ha
em
o
r
r
h
ag
e,
e
x
u
d
ates,
ch
a
n
g
es
in
b
lo
o
d
v
ess
els,
etc.
th
at
co
u
ld
r
ef
lect
th
e
p
o
s
itiv
e
c
ase
o
f
DR
.
Var
io
u
s
ad
v
an
ce
d
lev
els
o
f
m
u
lti
-
ca
teg
o
r
y
class
if
icatio
n
ca
n
b
e
c
ar
r
ied
o
u
t
b
y
DL
o
f
f
er
in
g
m
u
ch
g
r
a
n
u
lar
ity
in
d
is
ea
s
e
p
r
o
g
r
ess
io
n
.
Hen
ce
,
t
h
e
DL
m
o
d
el
co
n
tr
ib
u
tes
to
h
i
g
h
er
ac
cu
r
ac
y
,
ea
r
ly
d
etec
tio
n
,
an
d
au
to
m
ati
o
n
o
f
wo
r
k
lo
ad
.
Ho
wev
er
,
t
h
er
e
a
r
e
c
h
allen
g
e
s
to
o
ass
o
ciate
d
with
AI
,
M
L
,
an
d
DL
wh
ich
ar
e
m
ain
ly
ass
o
ciate
d
with
d
ata
q
u
ality
,
in
ter
p
r
etab
ilit
y
,
an
d
o
v
er
f
itti
n
g
is
s
u
es.
T
h
er
e
ar
e
s
till
v
ar
io
u
s
c
o
n
tr
a
d
i
cto
r
y
t
h
eo
r
ies
to
claim
wh
eth
er
ML
o
r
DL
co
u
l
d
p
r
o
v
e
f
r
u
itfu
l to
war
d
s
th
e
ea
r
ly
d
etec
tio
n
o
f
DR
in
r
ea
l
-
tim
e
s
ce
n
ar
io
s
.
2.
RE
L
AT
E
D
WO
RK
V
a
r
i
o
u
s
r
e
l
at
e
d
w
o
r
k
s
h
a
v
e
b
e
e
n
s
t
u
d
i
e
d
t
o
u
n
d
e
r
s
t
a
n
d
d
i
f
f
e
r
e
n
t
v
a
r
i
a
n
t
s
o
f
m
e
t
h
o
d
s
u
s
e
d
f
o
r
d
i
a
g
n
o
s
i
n
g
D
R
[
4
]
-
[
1
2
]
.
T
h
e
f
i
r
s
t
f
r
e
q
u
e
n
t
l
y
u
s
e
d
l
e
a
r
n
i
n
g
m
o
d
e
l
i
s
t
h
e
c
o
n
v
o
l
u
t
i
o
n
a
l
n
eu
r
a
l
n
e
t
w
o
r
k
(
C
NN
)
w
h
i
c
h
i
s
p
r
o
v
e
n
t
o
h
a
v
e
h
i
g
h
er
a
d
o
p
t
i
o
n
i
n
a
d
d
r
e
s
s
i
n
g
c
la
s
s
if
i
c
a
t
i
o
n
t
as
k
s
r
e
l
a
t
e
d
t
o
m
e
d
i
ca
l
i
m
a
g
e
s
.
V
a
r
i
o
u
s
e
x
i
s
ti
n
g
r
e
s
e
a
r
c
h
w
o
r
k
s
u
s
i
n
g
C
N
N
h
a
v
e
c
l
as
s
i
f
i
e
d
r
e
t
i
n
al
i
m
a
g
e
s
f
o
r
s
c
r
e
e
n
i
n
g
DR
[
1
3
]
-
[
1
7
]
.
T
h
e
s
ec
o
n
d
wid
ely
-
u
s
ed
m
ac
h
in
e
lear
n
in
g
m
o
d
el
f
o
r
s
cr
ee
n
in
g
DR
is
th
e
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
e
(
SV
M)
to
f
in
d
its
h
ig
h
er
s
id
e
o
f
ad
o
p
tio
n
to
war
d
s
th
e
f
ea
tu
r
e
-
b
ased
class
if
icatio
n
o
f
r
etin
al
im
ag
es
[
1
8
]
-
[
2
1
]
;
h
o
w
ev
er
,
th
e
y
ar
e
f
o
u
n
d
to
b
e
less
ef
f
ec
tiv
e
w
h
en
s
u
b
j
ec
ted
to
co
m
p
le
x
p
atter
n
s
o
f
DR
im
ag
es.
T
h
e
n
ex
t
f
r
eq
u
en
t
ad
o
p
tio
n
is
n
o
ted
f
o
r
th
e
d
ee
p
b
elief
n
etwo
r
k
(
D
B
N)
wh
ich
is
a
ty
p
e
o
f
d
ee
p
le
ar
n
in
g
ap
p
r
o
ac
h
wh
er
e
th
e
R
e
s
tr
icted
B
o
ltzm
an
n
Ma
ch
in
e
is
u
s
ed
f
o
r
d
ev
elo
p
in
g
its
m
u
ltip
le
lay
er
s
s
o
th
at
h
i
er
ar
ch
ical
f
ea
tu
r
e
lea
r
n
in
g
ca
n
b
e
ca
r
r
ied
o
u
t.
At
p
r
esen
t,
th
e
r
e
ar
e
v
a
r
io
u
s
e
v
o
l
v
in
g
s
tu
d
ies
ad
o
p
tin
g
DB
N
to
war
d
s
tr
ain
in
g
ex
tr
ac
te
d
f
ea
t
u
r
es
f
o
r
th
e
d
etec
tio
n
o
f
DR
[
2
2
]
-
[
2
4
]
.
R
esNet
-
5
0
is
an
o
th
er
wid
el
y
u
s
ed
ap
p
r
o
ac
h
f
o
r
DR
s
cr
ee
n
in
g
wh
ich
is
a
ty
p
e
o
f
d
ee
p
r
esid
u
al
n
etwo
r
k
th
at
is
ca
p
ab
le
o
f
ad
d
r
ess
in
g
v
an
is
h
in
g
g
r
ad
ien
t
p
r
o
b
lem
s
.
C
u
r
r
en
t
s
tu
d
ies
h
av
e
b
ee
n
witn
ess
ed
to
ad
o
p
t
R
esNet
-
5
0
in
cr
ea
s
in
g
ly
f
o
r
its
ca
p
ab
il
ity
o
f
in
cr
ea
s
ed
ac
cu
r
ac
y
an
d
h
ig
h
er
s
p
ec
if
icity
p
er
f
o
r
m
an
ce
[
2
5
]
-
[
2
8
]
.
A
f
t
e
r
r
e
v
i
e
wi
n
g
e
x
i
s
t
i
n
g
s
t
u
d
ie
s
,
t
h
e
f
o
ll
o
w
i
n
g
r
e
s
e
a
r
c
h
p
r
o
b
l
e
m
s
h
a
v
e
b
e
e
n
i
d
e
n
ti
f
i
e
d
:
i
)
e
a
r
l
y
-
s
t
a
g
e
d
e
t
e
c
t
i
o
n
:
E
x
i
s
t
i
n
g
s
t
at
e
-
of
-
t
h
e
-
a
r
t
d
e
e
p
le
a
r
n
i
n
g
m
e
t
h
o
d
s
a
n
d
o
t
h
e
r
a
s
s
o
c
i
at
e
d
a
p
p
r
o
a
c
h
es
e
n
c
o
u
n
t
e
r
p
o
t
e
n
t
i
al
c
h
a
l
l
e
n
g
e
s
i
n
d
i
f
f
e
r
e
n
t
i
a
ti
n
g
e
a
r
l
y
s
i
g
n
s
o
f
DR
f
r
o
m
n
o
r
m
a
l
i
m
a
g
e
s
o
f
t
h
e
r
e
t
i
n
a
.
ii
)
i
n
a
d
eq
u
a
t
e
g
e
n
e
r
a
l
i
z
a
ti
o
n
o
v
e
r
d
i
v
e
r
s
e
d
a
t
a:
E
x
is
t
i
n
g
m
et
h
o
d
s
l
i
k
e
R
es
N
et
-
5
0
a
n
d
C
NN
,
c
l
a
i
m
e
d
t
o
b
e
p
o
t
e
n
ti
a
l
l
y
s
t
r
o
n
g
w
e
r
e
f
o
u
n
d
t
o
o
v
e
r
f
i
t
o
n
t
h
e
p
a
r
t
i
c
u
l
a
r
d
a
t
a
s
et
a
n
d
h
e
n
c
e
t
h
e
y
e
x
h
i
b
i
t
m
u
c
h
p
o
o
r
g
e
n
e
r
a
l
i
z
a
t
i
o
n
w
h
e
n
a
s
s
e
s
s
e
d
o
n
a
d
i
f
f
e
r
e
n
t
d
a
t
a
s
et
,
i
ii
)
C
o
m
p
l
e
x
i
t
y
o
f
l
e
ar
n
i
n
g
a
p
p
r
o
a
c
h
:
T
h
e
c
o
m
p
l
e
x
a
r
c
h
i
t
e
ct
u
r
e
s
l
i
k
e
R
es
N
et
-
5
0
d
e
m
a
n
d
s
i
g
n
i
f
i
ca
n
t
c
o
m
p
u
t
a
t
i
o
n
a
l
r
es
o
u
r
c
e
s
(
f
o
r
b
o
t
h
i
n
f
e
r
e
n
c
e
a
n
d
t
r
a
i
n
i
n
g
)
a
c
t
i
n
g
as
a
n
i
m
p
e
d
i
m
e
n
t
t
o
wa
r
d
s
d
e
p
l
o
y
m
e
n
t
in
p
r
e
s
e
n
c
e
o
f
e
n
v
i
r
o
n
m
e
n
t
w
it
h
r
e
s
o
u
r
c
e
-
c
o
n
s
t
r
a
i
n
ts
.
i
v
)
S
u
b
-
o
p
t
i
m
a
l
i
n
t
e
r
p
r
e
t
a
b
i
li
t
y
o
f
d
ee
p
l
e
a
r
n
i
n
g
m
o
d
e
l:
A
l
t
h
o
u
g
h
R
es
N
et
-
5
0
a
n
d
C
NN
a
r
e
k
n
o
w
n
f
o
r
h
i
g
h
e
r
a
c
c
u
r
a
c
y
,
t
h
e
i
r
o
u
t
c
o
m
e
is
l
es
s
e
a
s
i
l
y
i
n
t
e
r
p
r
e
t
a
b
l
e
f
o
r
a
d
o
p
t
i
n
g
a
b
l
a
c
k
b
o
x
m
o
d
e
l
i
n
t
h
e
i
r
u
n
d
e
r
l
y
i
n
g
a
r
c
h
i
t
e
ct
u
r
e
.
T
h
e
p
r
i
m
e
a
i
m
o
f
t
h
e
p
r
o
p
o
s
e
d
s
t
u
d
y
i
s
t
o
d
e
v
e
l
o
p
a
s
i
m
p
l
i
f
i
ed
y
e
t
i
n
n
o
v
a
t
i
v
e
d
i
a
g
n
o
s
is
f
r
am
e
w
o
r
k
f
o
r
t
h
e
d
e
t
e
c
ti
o
n
a
n
d
c
l
as
s
i
f
i
c
at
i
o
n
o
f
D
R
at
it
s
e
a
r
l
y
o
n
s
et
.
T
h
e
v
a
l
u
e
-
a
d
d
e
d
c
o
n
t
r
i
b
u
t
i
o
n
o
f
t
h
e
s
t
u
d
y
a
r
e
a
s
f
o
l
l
o
w
s
:
i
)
A
n
o
v
e
l
a
u
t
o
e
n
c
o
d
e
r
-
b
a
s
e
d
a
r
c
h
it
e
c
t
u
r
e
h
a
s
b
e
en
p
r
e
s
e
n
t
e
d
t
o
d
i
a
g
n
o
s
e
D
R
in
e
a
r
l
y
s
t
a
g
es
w
i
t
h
h
i
g
h
e
r
a
c
c
u
r
a
c
y
,
i
i
)
th
e
s
tu
d
y
m
o
d
els f
ac
ilit
ate
id
en
tific
atio
n
o
f
ab
n
o
r
m
alities
as DR in
d
ica
to
r
s
u
s
in
g
Gau
s
s
ian
m
ix
tu
r
e
m
o
d
els
(
GM
M)
an
d
r
ec
o
n
s
tr
u
ctio
n
er
r
o
r
s
,
iii)
an
ex
h
au
s
tiv
e
ass
ess
m
en
t
i
s
ca
r
r
ied
o
u
t
to
co
m
p
ar
e
p
r
o
p
o
s
ed
m
o
d
el
with
C
NN,
SVM,
an
d
R
esNet
-
5
0
o
n
two
w
id
ely
u
s
ed
s
tan
d
a
r
d
b
e
n
ch
m
a
r
k
ed
d
ataset,
an
d
iv
)
a
p
o
ten
tial
g
en
er
aliza
tio
n
is
d
em
o
n
s
tr
ated
b
y
t
h
e
m
o
d
el
o
v
e
r
d
iv
e
r
s
e
d
ataset
wh
ich
s
h
o
ws
its
ap
p
licab
ilit
y
o
n
n
ea
r
r
ea
l
-
wo
r
ld
s
cr
ee
n
in
g
o
f
D
R
.
3.
M
E
T
H
O
D
T
h
e
p
r
im
e
o
b
jectiv
e
o
f
th
e
p
r
o
p
o
s
ed
r
esear
ch
wo
r
k
is
to
war
d
s
ev
o
lv
in
g
a
n
o
v
el
d
ee
p
lear
n
i
n
g
m
o
d
el
th
at
ca
n
f
ac
ilit
ate
p
o
te
n
tial
lea
r
n
i
n
g
o
f
h
id
d
en
r
ep
r
esen
tatio
n
o
f
r
etin
al
im
ag
es
f
o
r
ea
r
ly
d
e
tectio
n
o
f
DR
.
T
h
e
p
r
o
p
o
s
ed
s
tu
d
y
h
as
co
u
p
led
a
n
o
v
el
a
u
to
en
co
d
er
with
s
im
p
lifie
d
m
ath
em
atica
l
m
o
d
ellin
g
to
id
en
tify
as
well
as
class
if
y
th
e
ab
n
o
r
m
alities
ass
o
ciate
d
with
th
e
r
etin
a
d
u
r
in
g
DR
.
An
e
x
p
licit
r
esear
ch
m
eth
o
d
with
a
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
5
2
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
40
,
No
.
1
,
Octo
b
er
20
25
:
1
64
-
1
72
166
s
eq
u
en
tial
an
d
p
r
o
g
r
ess
iv
e
o
p
er
atio
n
al
s
tep
h
as
b
ee
n
p
er
f
o
r
m
ed
to
ac
co
m
p
lis
h
th
is
s
tu
d
y
g
o
al.
Fig
u
r
e
1
h
ig
h
lig
h
ts
th
e
ar
c
h
itectu
r
e
o
f
t
h
e
p
r
o
p
o
s
ed
s
y
s
tem
.
Fig
u
r
e
1
.
T
h
e
ar
ch
itectu
r
e
o
f
t
h
e
p
r
o
p
o
s
ed
s
y
s
tem
Acc
o
r
d
in
g
t
o
Fig
u
r
e
1
,
th
e
r
e
a
r
e
v
ar
io
u
s
s
ets
o
f
o
p
er
atio
n
s
b
ein
g
ca
r
r
ied
o
u
t
b
y
th
e
p
r
o
p
o
s
ed
s
y
s
tem
wh
ich
s
tar
t
f
r
o
m
th
e
m
an
ag
e
m
en
t
o
f
th
e
d
ataset
to
o
b
tain
in
g
th
e
o
u
tco
m
e
o
f
th
e
s
tu
d
y
.
T
h
e
m
eth
o
d
u
s
es
r
ec
o
n
s
tr
u
ctio
n
er
r
o
r
s
to
f
in
d
an
o
m
alies
s
u
g
g
esti
v
e
o
f
ea
r
l
y
-
s
tag
e
d
iab
etic
r
etin
o
p
ath
y
(
DR
)
b
y
c
o
m
b
in
i
n
g
au
to
en
co
d
er
s
with
G
au
s
s
ian
m
ix
tu
r
e
m
o
d
els
in
a
n
o
v
el
way
.
T
o
th
e
b
est
o
f
o
u
r
k
n
o
wled
g
e,
th
e
in
teg
r
atio
n
o
f
au
to
en
co
d
er
s
an
d
GM
M
in
to
a
s
in
g
le
f
r
am
ewo
r
k
s
p
ec
if
ically
f
o
r
ea
r
ly
DR
d
etec
ti
o
n
h
as
n
o
t
b
ee
n
in
v
esti
g
ated
,
ev
e
n
th
o
u
g
h
b
o
t
h
co
m
p
o
n
e
n
ts
h
av
e
b
ee
n
em
p
lo
y
ed
i
n
d
ep
e
n
d
en
tly
in
t
h
e
lit
er
atu
r
e
to
d
ate.
T
h
is
h
y
b
r
id
tec
h
n
iq
u
e
i
n
cr
ea
s
es
th
e
s
en
s
itiv
ity
o
f
ea
r
ly
-
s
tag
e
d
etec
tio
n
b
y
en
a
b
lin
g
b
o
th
f
ea
tu
r
e
ex
tr
ac
tio
n
(
u
s
in
g
au
to
en
co
d
er
s
)
an
d
class
if
icati
o
n
(
u
s
in
g
G
MM
)
.
Fo
llo
win
g
is
th
e
elab
o
r
ated
d
is
cu
s
s
io
n
o
f
th
e
o
p
e
r
atio
n
a
l
p
r
o
ce
s
s
b
ein
g
u
n
d
er
ta
k
en
in
th
e
p
r
o
p
o
s
ed
s
y
s
tem
:
3
.
1
.
Da
t
a
m
a
na
g
e
m
ent
T
h
e
in
p
u
t
im
ag
e
is
s
u
b
jecte
d
to
s
tan
d
ar
d
izatio
n
wh
er
e
r
escalin
g
is
p
er
f
o
r
m
ed
o
n
all
r
etin
al
im
ag
es
to
f
ix
ed
W
x
H
p
ix
els,
w
h
ich
is
a
n
ess
en
tial
s
tep
to
war
d
s
e
n
s
u
r
i
n
g
p
r
o
p
er
c
o
m
p
atib
ilit
y
o
f
th
e
in
p
u
t
im
ag
e
to
th
e
ar
ch
itectu
r
e
o
f
th
e
n
eu
r
al
n
et
wo
r
k
.
C
o
n
s
id
er
th
at,
th
e
r
eti
n
al
im
ag
e
is
r
ep
r
esen
ted
as
X
wh
er
e
X
∈
R
WxHxC
,
wh
ile
th
e
v
ar
iab
le
C
r
ep
r
esen
ts
th
e
ca
r
d
in
ality
o
f
c
o
lo
u
r
c
h
an
n
els.
A
s
im
p
le
way
to
u
n
d
er
s
tan
d
th
is
is
th
at
C
=1
will
r
ep
r
esen
t
a
g
r
a
y
s
ca
le
im
ag
e
wh
ile
C
=3
will
r
ep
r
esen
t
a
n
R
GB
im
ag
e.
T
h
e
m
ath
e
m
atica
l
r
ep
r
esen
tatio
n
o
f
n
o
r
m
aliza
tio
n
is
as,
=
1
(
1
)
I
n
(
1
)
,
th
e
c
o
m
p
u
tatio
n
o
f
n
o
r
m
alize
d
im
ag
e
X
norm
is
ca
r
r
ied
o
u
t
b
y
d
iv
id
in
g
v
ar
iab
le
A
1
b
y
s
tan
d
ar
d
d
ev
iatio
n
σ
,
wh
er
e
A
1
r
e
p
r
esen
ts
a
d
if
f
e
r
en
ce
o
f
m
ea
n
μ
f
r
o
m
m
ain
p
ix
el
X
i.e
.
A
1
=
X
-
μ
.
Af
ter
o
b
tain
in
g
a
n
o
r
m
alize
d
im
ag
e,
d
ata
a
u
g
m
en
tatio
n
is
p
e
r
f
o
r
m
ed
to
m
a
x
i
m
ize
th
e
v
ar
iab
ilit
y
a
n
d
s
ize
o
f
th
e
d
ataset.
T
h
e
m
ath
em
atic
r
ep
r
esen
tatio
n
o
f
d
ata
au
g
m
e
n
tatio
n
is
as:
X
’=
R
θ
(
X
)
(
2
)
I
n
(
2
)
,
X’
r
ep
r
esen
ts
au
g
m
en
ted
d
ata
wh
ile
R
θ
r
ep
r
esen
ts
r
o
tatio
n
tr
an
s
f
o
r
m
atio
n
co
n
s
id
er
in
g
θ
as
r
an
d
o
m
r
o
tatio
n
.
Ap
ar
t
f
r
o
m
th
is
,
im
ag
es
co
u
ld
b
e
tr
an
s
f
o
r
m
ed
in
to
g
r
ay
s
ca
le
im
ag
es
if
th
e
m
o
d
el
is
an
ticip
ated
to
f
o
c
u
s
m
o
r
e
o
n
t
ex
tu
r
e
a
n
d
s
h
ar
e
d
esp
ite
co
lo
u
r
attr
ib
u
tes.
T
h
e
s
tu
d
y
will
p
er
f
o
r
m
s
eg
m
en
tatio
n
u
s
in
g
r
eg
io
n
-
of
-
in
ter
est
(
R
o
I
)
to
co
n
f
in
e
its
atten
tio
n
t
o
th
e
r
etin
al
ar
ea
a
n
d
elim
in
at
e
an
y
u
n
n
ec
ess
ar
y
r
eg
io
n
s
in
th
e
b
ac
k
g
r
o
u
n
d
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2502
-
4
7
5
2
A
u
to
en
co
d
er
-
b
a
s
ed
G
a
u
s
s
ia
n
mixtu
r
e
mo
d
el
fo
r
d
ia
g
n
o
s
in
g
ea
r
ly
o
n
s
et
…
(
P
r
iya
n
ka
S
r
ee
n
iva
s
)
167
3
.
2
.
Aut
o
enco
der
des
ig
n
T
h
e
p
r
o
p
o
s
ed
s
tu
d
y
will
u
s
e
a
co
n
v
o
lu
tio
n
au
to
en
c
o
d
er
(
C
AE
)
wh
ich
is
d
esig
n
ed
u
s
in
g
th
r
ee
co
r
e
co
m
p
o
n
en
ts
i.e
.
,
en
c
o
d
er
,
late
n
t
s
p
ac
e
r
ep
r
esen
tatio
n
,
an
d
d
ec
o
d
er
.
T
h
e
en
co
d
er
e
x
tr
ac
ts
f
ea
tu
r
e
m
ap
s
f
r
o
m
th
e
in
p
u
t
im
ag
e
u
s
in
g
th
e
a
ctiv
atio
n
f
u
n
ctio
n
,
p
o
o
lin
g
lay
er
s
,
an
d
co
n
v
o
lu
tio
n
lay
e
r
s
.
T
h
e
laten
t
s
p
ac
e
r
ep
r
esen
tatio
n
is
an
o
u
tco
m
e
f
r
o
m
a
p
r
io
r
en
co
d
er
th
at
ca
p
tu
r
es
all
p
o
ten
tial
f
ea
t
u
r
es
f
r
o
m
th
e
r
etin
a
e.
g
.
,
m
icr
o
an
eu
r
y
s
m
s
,
ex
u
d
ates,
an
d
b
lo
o
d
v
ess
els.
T
h
e
d
ec
o
d
er
p
er
f
o
r
m
s
r
ec
o
n
s
tr
u
ctio
n
o
f
t
h
e
f
in
al
im
a
g
e
u
s
in
g
tr
an
s
p
o
s
ed
lay
er
s
o
f
co
n
v
o
l
u
tio
n
f
r
o
m
th
e
p
r
i
o
r
m
o
d
u
le
i.e
.
laten
t
s
p
ac
e.
T
h
e
m
ath
em
atica
l
r
ep
r
esen
tatio
n
o
f
th
e
en
co
d
e
r
o
p
e
r
atio
n
is
as f
o
l
lo
ws:
Z
=
E
(
X
)
(
3
)
I
n
(
3
)
,
th
e
v
ar
iab
le
Z
r
e
p
r
ese
n
ts
th
e
laten
t
r
ep
r
esen
tatio
n
o
f
th
e
in
p
u
t
im
ag
e
wh
er
e
E
r
e
p
r
esen
ts
an
en
co
d
er
f
u
n
ctio
n
.
T
h
e
e
x
p
r
ess
io
n
ca
n
b
e
f
u
r
th
e
r
m
o
d
if
ied
a
s
E
(
X
)=
f
θ1
(X
1
)
wh
er
e
E
is
th
e
en
co
d
in
g
f
u
n
ctio
n
f
θ1
r
ep
r
esen
ts
th
e
r
ec
tifie
d
lin
ea
r
u
n
it
(
R
eL
U)
f
u
n
ctio
n
wh
ile
X
1
r
ep
r
esen
ts
th
e
f
ir
s
t c
o
n
v
o
lu
tio
n
o
p
er
atio
n
.
T
h
e
co
n
v
o
l
u
tio
n
is
a
p
p
lied
in
ea
c
h
s
u
b
s
eq
u
en
t
lay
er
f
o
llo
wed
b
y
ac
tiv
atio
n
f
u
n
ctio
n
a
n
d
p
o
o
li
n
g
.
T
h
e
in
p
u
t
i
m
ag
e
X
is
tr
an
s
f
o
r
m
ed
to
Z
,
a
laten
t
r
ep
r
esen
tatio
n
,
a
f
ter
p
ass
in
g
v
ia
d
if
f
er
e
n
t
lay
er
s
.
I
t
s
h
o
u
ld
b
e
n
o
ted
th
at
Z
,
a
laten
t
v
ec
to
r
,
is
a
h
ig
h
ly
co
m
p
r
ess
ed
f
o
r
m
o
f
a
n
im
ag
e
with
lo
w
d
im
en
s
io
n
wh
ile
th
e
f
ea
tu
r
e
ex
tr
ac
tio
n
ca
p
ab
ilit
y
o
f
th
e
au
to
en
c
o
d
er
is
d
ec
id
ed
b
y
d
i.e
.
,
th
e
d
im
e
n
s
io
n
ality
o
f
Z
laten
t
v
ec
to
r
.
I
n
th
e
f
in
al
s
tep
,
th
e
d
ec
o
d
in
g
o
p
e
r
a
tio
n
is
ap
p
lied
wh
ich
is
m
ath
em
atica
lly
r
ep
r
e
s
en
ted
as,
̂
=
(
1
)
(
4
)
I
n
(
4
)
,
th
e
v
ar
iab
le
̂
R
ep
r
esen
ts
a
r
ec
o
n
s
tr
u
cted
im
a
g
e
w
h
ile
D
r
e
p
r
esen
ts
th
e
d
ec
o
d
in
g
f
u
n
ctio
n
u
s
in
g
th
e
s
am
e
ac
tiv
atio
n
an
d
Z
1
r
ep
r
esen
ts
tr
an
s
p
o
s
ed
co
n
v
o
lu
tio
n
o
p
er
ati
o
n
.
I
t
s
h
o
u
l
d
b
e
n
o
te
d
th
at
th
e
p
r
im
e
p
u
r
p
o
s
e
o
f
th
is
C
AE
is
to
war
d
s
lear
n
in
g
a
r
ep
r
esen
ta
tio
n
o
f
b
o
th
n
o
r
m
al
an
d
DR
im
ag
es
f
o
llo
wed
b
y
in
p
u
t
im
a
g
e
r
ec
o
n
s
tr
u
ctio
n
u
s
in
g
lo
w
-
d
im
en
s
io
n
al
f
ea
tu
r
es.
An
y
f
o
r
m
o
f
an
o
m
alies
p
r
ese
n
ted
in
th
e
f
o
r
m
o
f
er
r
o
r
s
d
u
r
in
g
th
e
r
ec
o
n
s
tr
u
ctio
n
p
r
o
ce
s
s
ca
n
o
f
f
e
r
in
d
icat
o
r
s
o
f
ea
r
ly
DR
.
T
h
e
n
o
v
elty
o
f
th
is
C
AE
m
o
d
u
le
d
if
f
er
en
t
f
r
o
m
a
n
y
e
x
is
tin
g
s
y
s
tem
is
i)
T
h
is
s
ch
em
e
u
s
e
s
an
atten
tio
n
m
ec
h
an
is
m
to
e
m
p
h
asize
R
o
I
(
e.
g
.
ex
u
d
ates,
m
icr
o
a
n
eu
r
y
s
m
s
)
th
at
ar
e
clin
ically
p
r
o
v
e
n
to
b
e
cr
itical
f
o
r
th
e
d
iag
n
o
s
is
o
f
DR
an
d
ii)
T
h
e
s
tu
d
y
also
u
s
es
d
ilated
co
n
v
o
lu
tio
n
s
(
also
k
n
o
wn
as
v
ar
iatio
n
al
au
to
en
co
d
er
s
(
VAE
)
)
to
m
a
n
ag
e
f
ea
tu
r
es
o
f
r
etin
al
im
ag
es with
h
ig
h
er
c
o
m
p
lex
iti
es a
s
well
as f
lu
ctu
atio
n
in
s
tag
es o
f
DR
.
3
.
3
.
L
o
s
s
f
un
ct
io
n
T
h
e
p
r
im
e
n
o
tio
n
o
f
th
e
au
to
en
co
d
er
is
to
war
d
s
r
ed
u
ctio
n
o
f
all
p
o
s
s
ib
le
s
co
r
es
o
f
r
ec
o
n
s
tr
u
ctio
n
er
r
o
r
b
etwe
en
th
e
X
an
d
̂
,
i.e
.
,
in
p
u
t
an
d
r
ec
o
n
s
tr
u
ct
ed
im
a
g
e.
T
h
e
s
tu
d
y
u
s
es
m
ea
n
s
q
u
a
r
ed
er
r
o
r
(
MSE
)
as
a
lo
s
s
f
u
n
ctio
n
in
t
h
e
in
itial r
o
u
n
d
s
o
f
o
p
er
atio
n
.
I
t is em
p
i
r
ically
ex
p
r
ess
ed
as,
=
1
∑
|
|
−
̂
|
|
2
=
1
(
5
)
I
n
(
5
)
,
th
e
co
m
p
u
tatio
n
o
f
lo
s
s
d
u
e
to
r
ec
o
n
s
tr
u
ctio
n
L
recon
is
ca
r
r
ied
o
u
t
co
n
s
id
er
i
n
g
N
as
t
h
e
n
u
m
b
er
o
f
im
ag
es
wh
ile
th
e
o
r
ig
in
al
an
d
r
ec
o
n
s
tr
u
cted
im
ag
e
is
r
e
p
r
esen
ted
as
X
i
an
d
̂
R
esp
ec
ti
v
ely
.
Ap
ar
t
f
r
o
m
th
is
,
th
e
p
r
o
p
o
s
ed
s
ch
em
e
also
u
s
es
L
2
r
eg
u
lar
izatio
n
to
c
o
n
tr
o
l
t
h
e
o
v
er
f
itti
n
g
p
r
o
b
lem
.
T
h
e
m
ath
em
atica
l
r
ep
r
esen
tatio
n
o
f
lo
s
s
d
u
e
to
r
eg
u
lar
izatio
n
is
s
h
o
wn
as,
=
∑
|
|
|
|
2
(
6
)
I
n
(
6
)
,
th
e
lo
s
s
o
f
r
eg
u
lar
izatio
n
L
regu
is
co
m
p
u
ted
co
n
s
id
er
in
g
n
etwo
r
k
weig
h
ts
as
θ
j
an
d
th
e
co
n
s
tan
t
o
f
r
e
g
u
lar
izatio
n
as
λ
.
I
t
s
h
o
u
ld
b
e
n
o
te
d
th
at
th
e
id
ea
o
f
u
s
in
g
L2
r
eg
u
lar
izatio
n
is
t
o
in
d
u
ce
a
p
en
alt
y
to
war
d
s
lar
g
e
weig
h
ts
p
r
esen
t
with
in
th
e
n
etwo
r
k
.
Hen
ce
,
t
h
e
to
tal
lo
s
s
f
u
n
ctio
n
L
total
is
c
o
m
p
u
ted
b
y
ad
d
i
n
g
(
5
)
an
d
(
6
)
i.e
.
L
total
=
L
recon
+
L
reg
u
.
3
.
4
.
Ano
m
a
ly
det
ec
t
io
n us
in
g
re
co
ns
t
ruct
io
n e
rr
o
r
Af
ter
th
e
co
m
p
u
ta
tio
n
o
f
th
e
l
o
s
s
f
u
n
ctio
n
,
th
e
p
r
o
p
o
s
ed
s
y
s
tem
u
s
es
b
ac
k
p
r
o
p
ag
atio
n
an
d
g
r
ad
ien
t
d
escen
t
f
o
r
tr
ain
in
g
th
e
au
to
e
n
co
d
er
.
Ass
u
m
in
g
t
h
e
n
etwo
r
k
p
ar
a
m
eter
b
e
d
e
n
o
ted
b
y
θ=
{
θ
1
,
θ
2
},
wh
er
e
θ
1
an
d
θ
2
r
e
p
r
esen
t b
o
t
h
en
co
d
er
an
d
d
ec
o
d
er
r
esp
ec
tiv
ely
.
Hen
ce
,
th
e
u
p
d
a
te
r
u
le
ca
n
b
e
d
ef
i
n
ed
as:
θ
θ
-
ηΔ
θ
.L
total
(
7
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
5
2
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
40
,
No
.
1
,
Octo
b
er
20
25
:
1
64
-
1
72
168
I
n
(
7
)
,
th
e
lear
n
i
n
g
r
ate
is
s
ig
n
if
ied
as
η
wh
ile
Δ
θ
.L
total
r
ep
r
esen
ts
th
e
g
r
ad
ien
t
ass
o
ciate
d
with
th
e
f
u
n
ctio
n
o
f
a
t
o
tal
lo
s
s
co
n
s
id
er
in
g
n
etwo
r
k
p
ar
a
m
eter
s
.
Af
t
er
th
e
tr
ain
in
g
o
f
th
e
en
co
d
er
i
s
ac
co
m
p
lis
h
ed
,
th
e
p
o
s
s
ib
le
ab
n
o
r
m
alities
ar
e
id
en
tifie
d
co
n
ce
r
n
in
g
t
h
e
r
ec
o
n
s
tr
u
ctio
n
er
r
o
r
co
n
s
id
er
in
g
r
e
tin
al
im
ag
e
s
wh
ich
h
av
e
h
ig
h
er
p
o
s
s
ib
ilit
ies
o
f
D
R
.
T
h
e
em
p
ir
ical
f
o
r
m
u
latio
n
o
f
a
r
ec
o
n
s
tr
u
ctio
n
er
r
o
r
f
o
r
X
as
an
in
p
u
t
im
ag
e
is
r
ep
r
esen
ted
as f
o
llo
ws:
E
(
X
)
=|
|
X
-
̂
||
(
8
)
Acc
o
r
d
in
g
to
(
8
)
,
if
th
e
e
r
r
o
r
o
f
r
ec
o
n
s
tr
u
ctio
n
E
(
X)
is
f
o
u
n
d
to
b
e
h
ig
h
er
t
h
an
a
ce
r
tain
c
ut
-
o
f
f
v
al
u
e
T
th
en
th
e
im
ag
e
u
n
d
er
o
b
s
er
v
atio
n
is
co
n
clu
d
ed
to
b
e
p
o
s
s
ess
in
g
ab
n
o
r
m
alities
.
T
h
is
co
u
ld
f
u
r
th
er
h
av
e
p
o
s
s
ib
ilit
ies
o
f
p
o
s
itiv
e
DR
.
T
h
e
d
eter
m
in
atio
n
o
f
cu
t
-
o
f
f
T
is
ca
r
r
ied
o
u
t
o
n
v
ar
i
o
u
s
ex
p
er
im
en
ts
o
n
th
e
v
alid
atio
n
s
et.
T
h
e
m
o
d
ellin
g
o
f
r
ec
o
n
s
tr
u
ctio
n
e
r
r
o
r
d
is
t
r
ib
u
tio
n
is
ca
r
r
ied
o
u
t
b
y
t
h
e
G
au
s
s
ian
m
ix
tu
r
e
m
o
d
el
wh
ich
is
u
tili
ze
d
f
o
r
c
o
m
p
u
tin
g
th
e
p
r
o
b
a
b
ilit
ies
o
f
all
er
r
o
r
s
d
u
r
in
g
t
h
e
r
ec
o
n
s
tr
u
ctio
n
p
r
o
ce
s
s
i.e
.
E
(
X
)
.
T
h
e
id
ea
is
to
war
d
s
cla
s
s
if
y
in
g
th
e
n
o
r
m
al
im
ag
e
f
r
o
m
DR
.
T
h
e
m
ath
em
atica
l
r
e
p
r
esen
tatio
n
o
f
th
e
p
r
o
b
a
b
ilit
y
o
f
er
r
o
r
i
n
im
ag
e
r
ec
o
n
s
tr
u
ctio
n
is
as f
o
llo
ws:
=
∑
.
(
)
=
1
(
9
)
I
n
(
9
)
,
th
e
co
m
p
u
tatio
n
o
f
P
r
o
b
(
p
r
o
b
a
b
ilit
y
o
f
r
ec
o
n
s
tr
u
ctio
n
er
r
o
r
)
is
ca
r
r
ie
d
o
u
t
c
o
n
s
id
er
in
g
α
as
a
m
ix
in
g
co
e
f
f
icien
t
an
d
gd
as
G
au
s
s
ian
Dis
tr
ib
u
tio
n
co
n
ce
r
n
in
g
E
as
er
r
o
r
,
m
ea
n
a
n
d
s
tan
d
ar
d
d
ev
iatio
n
.
T
h
e
im
ag
es
ar
e
f
u
r
t
h
er
s
u
b
jecte
d
to
class
if
icatio
n
b
ased
o
n
th
e
p
r
o
b
a
b
ilit
ies
o
f
th
eir
r
ec
o
n
s
tr
u
ctio
n
er
r
o
r
t
h
at
b
elo
n
g
s
to
a
class
o
f
DR
as f
o
l
lo
ws,
C
pred
=
a
r
g
max
P
r
o
b
(
E
|
C
)
(
1
0
)
I
n
(
1
0
)
,
th
e
co
m
p
u
tatio
n
o
f
cla
s
s
o
f
p
r
e
d
ictio
n
C
pred
is
ca
r
r
ied
o
u
t
co
n
s
id
er
i
n
g
a
m
a
x
im
u
m
n
u
m
b
er
o
f
ar
g
u
m
en
ts
ass
o
ciate
d
with
co
m
p
u
ted
p
r
o
b
a
b
ilit
y
P
r
o
b
in
th
e
p
r
ev
io
u
s
s
tep
co
n
ce
r
n
in
g
e
r
r
o
r
E
an
d
class
C
.
B
y
p
er
f
o
r
m
in
g
th
is
last
s
tep
,
th
e
o
b
jectiv
e
to
war
d
s
th
e
d
etec
tio
n
an
d
class
if
icatio
n
o
f
DR
is
ac
co
m
p
lis
h
ed
.
T
h
e
n
ex
t sectio
n
o
u
tlin
es th
e
s
tu
d
y
o
u
tco
m
es.
4.
RE
SU
L
T
T
h
e
im
p
lem
en
tatio
n
o
f
th
e
p
r
o
p
o
s
ed
s
tu
d
y
is
ca
r
r
ied
o
u
t
co
n
s
id
er
in
g
s
tan
d
ar
d
r
etin
al
im
a
g
e
d
atasets
.
T
h
e
s
tu
d
y
u
s
es
DI
AR
E
T
DB
1
,
wh
ich
co
n
s
is
ts
o
f
8
9
im
ag
e
s
[
2
9
]
,
a
n
d
E
y
ePAC
S,
wh
ich
co
m
p
r
is
es
3
5
,
0
0
0
im
ag
es
[
3
0
]
.
T
ab
le
1
s
h
o
wca
s
es
th
e
h
y
p
er
p
ar
a
m
eter
s
u
s
ed
f
o
r
th
e
an
aly
s
is
co
n
s
id
er
in
g
s
tan
d
ar
d
p
e
r
f
o
r
m
an
c
e
m
etr
ics
o
f
ac
cu
r
ac
y
,
s
en
s
itiv
ity
,
s
p
ec
if
icity
,
an
d
F1
-
Sc
o
r
e.
Fu
r
th
er
,
th
e
ass
ess
m
en
t
is
ca
r
r
ied
o
u
t
b
y
co
m
p
ar
in
g
th
e
p
r
o
p
o
s
ed
m
o
d
el
P
r
o
p
,
with
C
NN,
SVM,
DB
N,
an
d
R
e
s
Net
-
5
0
,
wh
ich
ar
e
f
o
u
n
d
to
b
e
f
r
eq
u
e
n
tly
ad
o
p
ted
to
war
d
s
in
v
esti
g
atin
g
DR
.
T
ab
le
1
.
Hy
p
er
p
ar
a
m
eter
s
an
d
th
eir
ad
o
p
ted
v
alu
es
H
y
p
e
r
p
a
r
a
me
t
e
r
V
a
l
u
e
/
S
e
t
t
i
n
g
A
u
t
o
e
n
c
o
d
e
r
a
r
c
h
i
t
e
c
t
u
r
e
C
o
n
v
o
l
u
t
i
o
n
a
l
l
a
y
e
r
s wi
t
h
3
2
,
6
4
,
1
2
8
,
2
5
6
f
i
l
t
e
r
s,
l
a
t
e
n
t
s
p
a
c
e
d
i
me
n
s
i
o
n
o
f
5
1
2
A
c
t
i
v
a
t
i
o
n
f
u
n
c
t
i
o
n
R
e
LU
(
i
n
t
e
r
m
e
d
i
a
t
e
)
,
S
i
g
mo
i
d
(
o
u
t
p
u
t
l
a
y
e
r
)
La
t
e
n
t
s
p
a
c
e
d
i
me
n
si
o
n
5
1
2
O
p
t
i
mi
z
e
r
A
d
a
m,
Le
a
r
n
i
n
g
r
a
t
e
=
0
.
0
0
1
,
β
1
=
0
.
9
,
β2
=
0
.
9
9
9
,
ϵ=
1
0
−7
Lo
ss f
u
n
c
t
i
o
n
M
e
a
n
S
q
u
a
r
e
d
Er
r
o
r
(
M
S
E)
,
L
2
R
e
g
u
l
a
r
i
z
a
t
i
o
n
B
a
t
c
h
si
z
e
32
Ep
o
c
h
s
5
0
(
w
i
t
h
e
a
r
l
y
s
t
o
p
p
i
n
g
)
D
r
o
p
o
u
t
r
a
t
e
0
.
3
(
a
f
t
e
r
c
o
n
v
o
l
u
t
i
o
n
a
l
l
a
y
e
r
s)
W
e
i
g
h
t
i
n
i
t
i
a
l
i
z
a
t
i
o
n
X
a
v
i
e
r
I
n
i
t
i
a
l
i
z
a
t
i
o
n
G
M
M
c
o
mp
o
n
e
n
t
s
2
(
H
e
a
l
t
h
y
,
D
R
)
G
M
M
c
o
v
a
r
i
a
n
c
e
t
y
p
e
F
u
l
l
c
o
v
a
r
i
a
n
c
e
ma
t
r
i
x
G
M
M
ma
x
i
t
e
r
a
t
i
o
n
s
1
0
0
0
Tr
a
i
n
-
Te
s
t
S
p
l
i
t
7
0
%
-
1
5
%
-
1
5
%
T
ab
le
2
an
d
Fig
u
r
es
2
-
5
s
h
o
wca
s
e
th
e
ac
tu
al
s
tu
d
y
o
u
tc
o
m
e
to
s
h
o
w
th
at
Pro
p
e
x
ce
ls
s
ig
n
if
ican
tly
b
etter
p
er
f
o
r
m
an
ce
in
c
o
n
tr
ast
to
f
r
eq
u
e
n
tly
ex
is
tin
g
lear
n
in
g
m
o
d
els
to
war
d
s
d
etec
tio
n
a
n
d
class
if
icatio
n
o
f
DR
.
An
in
ter
esti
n
g
f
ac
t
to
o
b
s
er
v
e
in
th
is
ev
alu
atio
n
s
co
r
e
e
x
h
ib
ited
in
T
ab
le
2
is
t
h
at
it
o
f
f
er
s
n
ea
r
l
y
s
im
ilar
p
er
f
o
r
m
an
ce
co
n
ce
r
n
in
g
b
o
th
th
e
d
atasets
an
d
h
en
ce
its
co
n
s
is
ten
cy
ca
n
b
e
ev
en
tu
all
y
ascer
tain
ed
o
v
er
ad
o
p
ted
p
er
f
o
r
m
an
ce
m
etr
ics.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2502
-
4
7
5
2
A
u
to
en
co
d
er
-
b
a
s
ed
G
a
u
s
s
ia
n
mixtu
r
e
mo
d
el
fo
r
d
ia
g
n
o
s
in
g
ea
r
ly
o
n
s
et
…
(
P
r
iya
n
ka
S
r
ee
n
iva
s
)
169
T
ab
le
2
.
C
o
m
p
r
eh
en
s
iv
e
s
tu
d
y
o
u
tco
m
es
M
o
d
e
l
D
a
t
a
s
e
t
A
c
c
u
r
a
c
y
(
%)
S
e
n
s
i
t
i
v
i
t
y
(
%)
S
p
e
c
i
f
i
c
i
t
y
(
%)
F1
-
S
c
o
r
e
P
r
o
p
D
I
A
R
ETD
B
1
9
6
.
5
9
4
.
2
9
8
.
3
0
.
9
6
P
r
o
p
Ey
e
P
A
C
S
9
5
.
1
9
2
.
8
9
7
.
5
0
.
9
4
C
N
N
D
I
A
R
ETD
B
1
9
2
.
1
8
8
.
3
9
5
.
2
0
.
9
0
C
N
N
Ey
e
P
A
C
S
9
3
.
6
8
9
.
5
9
6
.
3
0
.
9
1
S
V
M
D
I
A
R
ETD
B
1
8
5
.
4
8
3
.
2
8
9
.
6
0
.
8
4
S
V
M
Ey
e
P
A
C
S
8
7
.
3
8
5
.
1
9
0
.
8
0
.
8
6
D
B
N
D
I
A
R
ETD
B
1
9
0
.
2
8
6
.
5
9
3
.
5
0
.
8
8
D
B
N
Ey
e
P
A
C
S
9
1
.
3
8
7
.
9
9
4
.
8
0
.
8
9
R
e
sN
e
t
-
50
D
I
A
R
ETD
B
1
9
3
.
4
8
9
.
8
9
5
.
8
0
.
9
1
R
e
sN
e
t
-
50
Ey
e
P
A
C
S
9
4
.
3
9
0
.
2
9
6
.
1
0
.
8
9
Fig
u
r
e
2
.
Ass
ess
m
en
t o
f
ac
cu
r
ac
y
Fig
u
r
e
3
.
Ass
ess
m
en
t o
f
s
p
ec
if
icity
Fig
u
r
e
4
.
Ass
ess
m
en
t o
f
s
en
s
itiv
it
y
Fig
u
r
e
5
.
Ass
ess
m
en
t o
f
F1
-
s
co
r
e
4
.
1
.
Abla
t
io
n
s
t
ud
y
T
ab
le
3
p
r
esen
ts
th
e
r
esu
lts
o
f
th
e
ab
latio
n
s
tu
d
y
,
wh
er
e
v
ar
io
u
s
co
m
p
o
n
en
ts
o
f
th
e
m
o
d
el
wer
e
alter
ed
to
d
eter
m
in
e
th
eir
i
n
f
lu
en
ce
o
n
th
e
o
v
er
all
p
e
r
f
o
r
m
an
ce
.
E
ac
h
e
x
p
er
im
en
t
test
ed
th
e
m
o
d
el'
s
p
er
f
o
r
m
an
ce
i
n
ter
m
s
o
f
ac
c
u
r
ac
y
,
s
en
s
itiv
ity
,
s
p
ec
if
icity
,
a
n
d
F1
-
Sco
r
e
u
s
in
g
th
e
DI
AR
E
T
DB
1
d
ataset.
T
ab
le
3
s
h
o
wca
s
es
th
e
o
u
tco
m
e
o
f
th
e
ab
latio
n
s
tu
d
y
.
W
ith
9
6
.
5
%
ac
cu
r
ac
y
,
9
4
.
2
%
s
en
s
itiv
ity
,
an
d
9
8
.
3
%
s
p
ec
if
icity
,
th
e
f
u
ll
m
o
d
el
(
Pro
p
)
p
er
f
o
r
m
s
b
est
ac
r
o
s
s
all
m
ea
s
u
r
es,
d
em
o
n
s
tr
atin
g
t
h
e
v
alu
e
o
f
co
m
b
in
in
g
au
to
e
n
co
d
er
s
,
GM
M,
a
n
d
o
th
er
m
eth
o
d
s
.
Acc
u
r
ac
y
an
d
s
en
s
itiv
ity
s
ig
n
if
ican
tly
d
ec
r
ea
s
e
wh
en
th
e
au
to
en
co
d
er
is
r
em
o
v
ed
,
s
u
g
g
esti
n
g
th
at
au
to
en
co
d
er
s
ar
e
es
s
en
tial f
o
r
f
ea
tu
r
e
e
x
tr
ac
tio
n
i
n
DR
d
etec
tio
n
.
T
ab
le
3
.
Ab
latio
n
s
tu
d
y
r
esu
lt
s
o
n
DI
AR
E
T
DB
1
d
at
aset
Ex
p
e
r
i
m
e
n
t
A
c
c
u
r
a
c
y
(
%)
S
e
n
s
i
t
i
v
i
t
y
(
%)
S
p
e
c
i
f
i
c
i
t
y
(
%)
F1
-
S
c
o
r
e
F
u
l
l
m
o
d
e
l
(
P
r
o
p
)
9
6
.
5
9
4
.
2
9
8
.
3
0
.
9
6
W
i
t
h
o
u
t
a
u
t
o
e
n
c
o
d
e
r
9
2
.
3
8
9
.
1
9
4
.
7
0
.
9
0
W
i
t
h
o
u
t
G
M
M
9
3
.
8
9
1
.
4
9
6
.
1
0
.
9
2
W
i
t
h
o
u
t
d
a
t
a
a
u
g
me
n
t
a
t
i
o
n
9
4
.
1
9
1
.
7
9
6
.
8
0
.
9
3
W
i
t
h
o
u
t
R
o
I
seg
m
e
n
t
a
t
i
o
n
9
5
.
2
9
2
.
8
9
7
.
9
0
.
9
4
W
i
t
h
o
u
t
re
g
u
l
a
r
i
z
a
t
i
o
n
(
L
2
)
9
3
.
7
9
0
.
3
9
6
.
4
0
.
9
1
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
5
2
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
40
,
No
.
1
,
Octo
b
er
20
25
:
1
64
-
1
72
170
Per
f
o
r
m
an
ce
is
m
o
d
er
ately
r
ed
u
ce
d
wh
e
n
GM
M
is
d
is
ab
led
,
d
em
o
n
s
tr
atin
g
th
e
s
ig
n
if
ican
ce
o
f
r
ec
o
n
s
tr
u
ctio
n
er
r
o
r
m
o
d
e
l
lin
g
f
o
r
d
if
f
er
en
tiatin
g
b
etwe
en
n
o
r
m
al
an
d
DR
p
ictu
r
e
s
.
R
em
o
v
in
g
R
OI
s
eg
m
en
tatio
n
o
r
d
ata
au
g
m
e
n
tatio
n
wea
k
en
s
th
e
m
o
d
el,
d
em
o
n
s
tr
atin
g
th
at
t
h
ese
m
eth
o
d
s
im
p
r
o
v
e
th
e
m
o
d
el'
s
ca
p
ac
ity
f
o
r
g
en
er
aliz
atio
n
.
Desp
ite
h
av
in
g
a
n
eg
li
g
ib
le
im
p
ac
t,
L
2
r
eg
u
lar
izati
o
n
h
elp
s
to
im
p
r
o
v
e
m
o
d
el
g
e
n
er
aliza
tio
n
an
d
less
en
o
v
er
f
itti
n
g
.
4
.
2
.
Dis
cus
s
io
n
T
h
e
k
ey
f
in
d
in
g
s
ar
e
s
u
m
m
ar
i
ze
d
as
f
o
llo
ws
:
we
d
e
v
elo
p
ed
a
n
ew
d
ee
p
-
lear
n
in
g
m
o
d
el
f
o
r
th
e
ea
r
l
y
id
en
tific
atio
n
o
f
DR
b
y
co
m
b
i
n
in
g
Gau
s
s
ian
m
ix
tu
r
e
m
o
d
els
f
o
r
an
o
m
aly
d
etec
tio
n
with
C
AE
.
Acr
o
s
s
s
ev
er
al
d
atasets
,
in
clu
d
in
g
DI
AR
E
T
DB
1
an
d
E
y
ePAC
S,
th
e
s
u
g
g
ested
m
o
d
el,
k
n
o
w
n
as
Pro
p
,
o
u
tp
er
f
o
r
m
ed
m
o
r
e
estab
lis
h
ed
m
o
d
els
,
in
clu
d
in
g
C
NN,
SVM,
D
B
N,
an
d
R
es
Net
-
5
0
.
On
th
e
DI
AR
E
T
DB
1
d
ataset,
th
e
Pro
p
m
o
d
el
d
em
o
n
s
tr
ated
a
r
em
ar
k
ab
le
9
6
.
5
%
ac
cu
r
ac
y
an
d
9
4
.
2
%
s
en
s
itiv
ity
,
d
em
o
n
s
tr
atin
g
its
ef
f
icac
y
in
id
en
tify
in
g
DR
at
an
ea
r
ly
s
tag
e.
T
h
ese
f
in
d
in
g
s
d
em
o
n
s
tr
ate
th
e
m
o
d
el'
s
r
esil
ien
ce
a
n
d
its
ca
p
ac
ity
to
ac
h
iev
e
h
ig
h
s
p
ec
if
icity
(
9
8
.
3
%)
wh
ile
r
ed
u
cin
g
f
alse p
o
s
itiv
es.
T
h
e
Pro
p
m
o
d
el
c
o
n
tin
u
o
u
s
ly
s
u
r
p
ass
ed
ea
r
lier
r
esear
ch
,
in
clu
d
in
g
th
o
s
e
co
n
d
u
cte
d
with
C
NN
o
r
R
esNet
-
5
0
,
esp
ec
ially
in
ter
m
s
o
f
s
en
s
itiv
ity
,
wh
ich
is
ess
en
tial
f
o
r
id
e
n
tify
in
g
DR
in
its
ea
r
ly
,
m
o
s
t
cu
r
a
b
le
p
h
ases
.
T
h
is
was f
u
r
th
er
co
r
r
o
b
o
r
ated
b
y
th
e
ab
latio
n
in
v
esti
g
atio
n
,
wh
ich
d
e
m
o
n
s
tr
ated
th
at
ev
er
y
elem
en
t
—
in
clu
d
in
g
d
ata
au
g
m
en
tatio
n
,
R
OI
s
eg
m
en
tatio
n
,
an
d
GM
M
in
teg
r
atio
n
—
m
ad
e
a
s
u
b
s
tan
t
ial
co
n
tr
ib
u
tio
n
to
th
e
m
o
d
el'
s
p
er
f
o
r
m
an
ce
.
Hig
h
class
if
icatio
n
ac
cu
r
ac
y
an
d
r
eliab
le
g
en
er
aliza
tio
n
ac
r
o
s
s
a
v
ar
iety
o
f
d
atasets
wer
e
m
a
d
e
p
o
s
s
ib
le
b
y
th
e
co
m
b
in
atio
n
o
f
au
to
e
n
co
d
e
r
s
f
o
r
f
ea
tu
r
e
e
x
tr
ac
tio
n
an
d
GM
M
f
o
r
an
o
m
aly
d
etec
tio
n
.
I
n
s
u
m
m
a
r
y
,
o
u
r
s
tu
d
y
h
as
s
h
o
wn
t
h
at
a
h
y
b
r
i
d
au
to
e
n
co
d
e
r
-
GM
M
tech
n
i
q
u
e,
wh
ich
o
f
f
er
s
a
b
alan
ce
o
f
h
ig
h
ac
cu
r
ac
y
,
s
e
n
s
itiv
ity
,
an
d
co
m
p
u
tatio
n
al
ec
o
n
o
m
y
,
i
s
a
p
o
ten
tial
to
o
l
f
o
r
ea
r
ly
-
s
tag
e
DR
id
en
tific
atio
n
.
T
h
e
s
u
g
g
ested
s
y
s
tem
o
f
f
er
s
a
s
tr
o
n
g
b
asis
f
o
r
u
p
co
m
i
n
g
d
ev
elo
p
m
en
ts
in
au
to
m
ated
r
etin
a
l
s
cr
ee
n
in
g
an
d
,
ev
e
n
tu
ally
,
im
p
r
o
v
in
g
p
atien
t o
u
tco
m
es b
y
id
en
tify
in
g
d
iab
etic
p
r
o
b
lem
s
ea
r
ly
.
5.
CO
NCLU
SI
O
N
T
h
e
cu
r
r
en
t
p
ap
e
r
h
as
p
r
esen
t
ed
a
n
o
v
el
d
ee
p
-
lear
n
in
g
m
o
d
el
u
s
in
g
a
n
a
u
to
en
co
d
er
to
d
et
ec
t
DR
in
its
ea
r
ly
s
tag
es.
B
y
cr
ea
tin
g
a
u
n
iq
u
e
a
u
to
en
c
o
d
er
-
b
ased
m
o
d
el
an
d
i
n
teg
r
atin
g
it
with
Gau
s
s
ian
m
ix
tu
r
e
m
o
d
els
,
th
is
s
tu
d
y
ef
f
ec
tiv
ely
tack
les
th
e
m
ain
o
b
s
tacl
es
in
t
h
e
ea
r
ly
d
iag
n
o
s
is
o
f
DR
.
T
h
i
s
g
r
ea
tly
im
p
r
o
v
es
th
e
s
en
s
itiv
ity
an
d
s
p
ec
if
icity
o
f
DR
d
etec
tio
n
in
its
ea
r
l
y
s
tag
es.
W
e
s
h
o
w
th
at
o
u
r
m
eth
o
d
n
o
t
o
n
l
y
ac
h
ie
v
es
h
ig
h
class
if
icatio
n
ac
cu
r
ac
y
b
u
t
also
g
u
a
r
an
tees
s
tr
o
n
g
g
en
er
aliza
tio
n
b
y
v
alid
atin
g
th
e
p
r
o
p
o
s
ed
m
o
d
el
o
n
s
ev
er
al
b
en
ch
m
ar
k
d
atasets
,
ac
h
iev
in
g
th
e
g
o
al
o
f
o
f
f
er
in
g
a
p
r
ac
tical,
s
ca
lab
le,
an
d
in
ter
p
r
etab
le
s
o
lu
tio
n
f
o
r
DR
d
iag
n
o
s
is
in
r
ea
l
-
wo
r
ld
clin
ical
s
ettin
g
s
.
T
h
e
p
r
im
ar
y
n
o
v
elty
an
d
co
n
tr
ib
u
tio
n
o
f
th
e
s
tu
d
y
ar
e
as
f
o
llo
ws:
i)
th
e
is
s
u
es
o
f
ea
r
ly
-
s
tag
e
d
etec
tio
n
id
en
tifie
d
in
r
e
lated
wo
r
k
h
av
e
b
ee
n
ad
d
r
ess
ed
b
y
t
h
e
p
r
o
p
o
s
e
d
h
y
b
r
id
izatio
n
o
f
GM
M
an
d
Au
to
en
co
d
er
b
y
u
s
in
g
r
ec
o
n
s
tr
u
ctio
n
er
r
o
r
s
f
o
r
d
eter
m
in
in
g
g
r
an
u
lar
an
o
m
alies
ass
o
ciate
d
with
r
etin
al
im
a
g
es.
T
h
is
d
ir
ec
tly
im
p
r
o
v
es
th
e
s
e
n
s
it
iv
ity
o
f
th
e
s
y
s
tem
t
o
war
d
s
ea
r
ly
d
etec
tio
n
o
f
DR
.
ii)
T
h
e
is
s
u
e
o
f
in
a
d
eq
u
ate
g
en
er
aliza
tio
n
o
v
e
r
d
iv
er
s
e
d
ata
is
ad
d
r
ess
ed
in
t
h
e
cu
r
r
en
t
wo
r
k
b
y
co
n
s
id
er
in
g
two
d
if
f
e
r
en
t
d
at
asets
o
f
E
y
ePA
C
S
an
d
DI
AR
E
T
DB
1
f
o
r
en
h
an
ce
d
ap
p
lica
b
ilit
y
an
d
im
p
r
o
v
ed
g
en
er
ali
za
tio
n
to
n
ea
r
-
p
r
ac
tic
al
clin
ical
s
ettin
g
s
.
iii)
th
e
is
s
u
es
o
f
th
e
C
o
m
p
lex
ity
o
f
th
e
L
ea
r
n
in
g
Ap
p
r
o
ac
h
ar
e
ad
d
r
ess
ed
b
y
u
s
in
g
an
au
t
o
en
co
d
e
r
f
o
r
an
o
m
aly
d
etec
ti
o
n
with
r
ed
u
ce
d
c
o
m
p
u
tatio
n
al
b
u
r
d
en
an
d
m
o
r
e
ef
f
icien
t
th
an
R
esNet
-
5
0
,
a
d
e
ep
er
n
etwo
r
k
.
i
v
)
t
h
e
is
s
u
es
o
f
Su
b
-
o
p
tim
al
in
ter
p
r
etab
ilit
y
o
f
th
e
d
ee
p
lear
n
in
g
m
o
d
el
ar
e
a
d
d
r
ess
ed
b
y
jo
in
t i
n
teg
r
atio
n
o
f
GM
M
with
r
ec
o
n
s
tr
u
ctio
n
er
r
o
r
f
o
r
b
etter
in
ter
p
r
etab
ilit
y
.
Fu
tu
r
e
wo
r
k
will
b
e
ca
r
r
ied
o
u
t
to
war
d
s
b
e
n
ch
m
ar
k
in
g
th
e
p
r
o
p
o
s
ed
s
y
s
tem
with
o
t
h
er
ass
o
ci
ated
o
cu
lar
d
is
ea
s
es
lik
e
g
lau
co
m
a
.
A
n
o
v
el
a
p
p
r
o
ac
h
t
h
at
ca
n
d
is
tin
g
u
is
h
DR
f
r
o
m
g
lau
co
m
ic
p
atien
ts
in
its
ea
r
ly
s
tag
e
will b
e
d
ev
elo
p
ed
h
ar
n
ess
in
g
a
m
o
r
e
i
n
n
o
v
ativ
e
AI
-
b
as
ed
ap
p
r
o
ac
h
.
F
UNDING
I
NF
O
R
M
A
T
I
O
N
Au
th
o
r
s
s
tate
n
o
f
u
n
d
in
g
in
v
o
lv
ed
.
AUTHO
R
CO
NT
RI
B
UT
I
O
NS ST
A
T
E
M
E
N
T
T
h
is
jo
u
r
n
al
u
s
es
th
e
C
o
n
tr
ib
u
to
r
R
o
les
T
ax
o
n
o
m
y
(
C
R
ed
iT)
to
r
ec
o
g
n
ize
in
d
iv
id
u
al
au
th
o
r
co
n
tr
ib
u
tio
n
s
,
r
ed
u
ce
au
th
o
r
s
h
ip
d
is
p
u
tes,
an
d
f
ac
ilit
ate
co
llab
o
r
atio
n
.
Na
m
e
o
f
Aut
ho
r
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
P
riy
a
n
k
a
S
re
e
n
i
v
a
s
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
Ka
v
it
a
V.
Ho
ra
d
i
✓
✓
✓
✓
✓
✓
✓
Ka
lp
a
Ra
jas
h
e
k
a
r
✓
✓
✓
✓
✓
✓
✓
✓
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2502
-
4
7
5
2
A
u
to
en
co
d
er
-
b
a
s
ed
G
a
u
s
s
ia
n
mixtu
r
e
mo
d
el
fo
r
d
ia
g
n
o
s
in
g
ea
r
ly
o
n
s
et
…
(
P
r
iya
n
ka
S
r
ee
n
iva
s
)
171
C
:
C
o
n
c
e
p
t
u
a
l
i
z
a
t
i
o
n
M
:
M
e
t
h
o
d
o
l
o
g
y
So
:
So
f
t
w
a
r
e
Va
:
Va
l
i
d
a
t
i
o
n
Fo
:
Fo
r
mal
a
n
a
l
y
s
i
s
I
:
I
n
v
e
s
t
i
g
a
t
i
o
n
R
:
R
e
so
u
r
c
e
s
D
:
D
a
t
a
C
u
r
a
t
i
o
n
O
:
W
r
i
t
i
n
g
-
O
r
i
g
i
n
a
l
D
r
a
f
t
E
:
W
r
i
t
i
n
g
-
R
e
v
i
e
w
&
E
d
i
t
i
n
g
Vi
:
Vi
su
a
l
i
z
a
t
i
o
n
Su
:
Su
p
e
r
v
i
s
i
o
n
P
:
P
r
o
j
e
c
t
a
d
mi
n
i
st
r
a
t
i
o
n
Fu
:
Fu
n
d
i
n
g
a
c
q
u
i
si
t
i
o
n
CO
NF
L
I
C
T
O
F
I
N
T
E
R
E
S
T
ST
A
T
E
M
E
NT
Au
th
o
r
s
s
tate
n
o
co
n
f
lict o
f
in
t
er
est.
DATA AV
AI
L
AB
I
L
I
T
Y
T
h
e
d
ata
th
at
s
u
p
p
o
r
t
th
e
f
in
d
in
g
s
o
f
th
is
s
tu
d
y
ar
e
av
aila
b
le
f
r
o
m
th
e
c
o
r
r
esp
o
n
d
in
g
au
th
o
r
u
p
o
n
r
ea
s
o
n
ab
le
r
eq
u
est.
RE
F
E
R
E
NC
E
S
[
1
]
S
.
A
j
i
t
h
K
u
mar
a
n
d
J
.
S
a
t
h
e
e
sh
K
u
mar,
“
R
e
t
i
n
a
l
l
e
s
i
o
n
s
c
l
a
ss
i
f
i
c
a
t
i
o
n
f
o
r
d
i
a
b
e
t
i
c
r
e
t
i
n
o
p
a
t
h
y
u
si
n
g
c
u
s
t
o
m
R
e
sN
e
t
-
b
a
se
d
c
l
a
ss
i
f
i
e
r
,
”
I
n
d
o
n
e
s
i
a
n
J
o
u
r
n
a
l
o
f
E
l
e
c
t
ri
c
a
l
E
n
g
i
n
e
e
ri
n
g
a
n
d
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
(
I
J
EE
C
S
)
,
v
o
l
.
3
3
,
n
o
.
1
,
p
.
4
0
5
,
2
0
2
4
,
d
o
i
:
1
0
.
1
1
5
9
1
/
i
j
e
e
c
s
.
v
3
3
.
i
1
.
p
p
4
0
5
-
415
.
[
2
]
M
.
K
ą
p
a
,
I
.
K
o
r
y
c
i
a
r
z
,
N
.
K
u
st
o
si
k
,
P
.
Ju
r
o
w
sk
i
,
a
n
d
Z
.
P
n
i
a
k
o
w
s
k
a
,
“
M
o
d
e
r
n
a
p
p
r
o
a
c
h
t
o
d
i
a
b
e
t
i
c
r
e
t
i
n
o
p
a
t
h
y
d
i
a
g
n
o
s
t
i
c
s,
”
D
i
a
g
n
o
s
t
i
c
s
(
B
a
sel
)
,
v
o
l
.
1
4
,
n
o
.
1
7
,
p
.
1
8
4
6
,
2
0
2
4
,
d
o
i
:
1
0
.
3
3
9
0
/
d
i
a
g
n
o
s
t
i
c
s
1
4
1
7
1
8
4
6
.
[
3
]
M
.
A
v
a
n
z
o
,
J.
S
t
a
n
c
a
n
e
l
l
o
,
G
.
P
i
r
r
o
n
e
,
A
.
D
r
i
g
o
,
a
n
d
A
.
R
e
t
i
c
o
,
“
Th
e
e
v
o
l
u
t
i
o
n
o
f
a
r
t
i
f
i
c
i
a
l
i
n
t
e
l
l
i
g
e
n
c
e
i
n
me
d
i
c
a
l
i
m
a
g
i
n
g
:
F
r
o
m
c
o
m
p
u
t
e
r
sci
e
n
c
e
t
o
m
a
c
h
i
n
e
a
n
d
d
e
e
p
l
e
a
r
n
i
n
g
,
”
C
a
n
c
e
rs
,
v
o
l
.
1
6
,
n
o
.
2
1
,
p
.
3
7
0
2
,
2
0
2
4
,
d
o
i
:
1
0
.
3
3
9
0
/
c
a
n
c
e
r
s1
6
2
1
3
7
0
2
.
[
4
]
D
.
M
u
t
h
u
s
a
my
a
n
d
P
.
P
a
l
a
n
i
,
"
D
e
e
p
l
e
a
r
n
i
n
g
m
o
d
e
l
u
si
n
g
c
l
a
ssi
f
i
c
a
t
i
o
n
f
o
r
d
i
a
b
e
t
i
c
r
e
t
i
n
o
p
a
t
h
y
d
e
t
e
c
t
i
o
n
:
a
n
o
v
e
r
v
i
e
w
,
v
o
l
.
5
7
,
n
o
.
1
8
5
,
2
0
2
4
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
0
4
6
2
-
0
2
4
-
1
0
8
0
6
-
2.
[
5
]
D
.
M
u
t
h
u
s
a
m
y
a
n
d
P
.
P
a
l
a
n
i
,
"
D
e
e
p
n
e
u
r
a
l
n
e
t
w
o
r
k
m
o
d
e
l
f
o
r
d
i
a
g
n
o
si
n
g
d
i
a
b
e
t
i
c
r
e
t
i
n
o
p
a
t
h
y
d
e
t
e
c
t
i
o
n
:
A
n
e
f
f
i
c
i
e
n
t
me
c
h
a
n
i
sm
f
o
r
d
i
a
b
e
t
i
c
m
a
n
a
g
e
me
n
t
,
"
B
i
o
m
e
d
i
c
a
l
S
i
g
n
a
l
P
ro
c
e
ss
i
n
g
a
n
d
C
o
n
t
r
o
l
,
v
o
l
.
1
0
0
,
p
.
1
0
7
0
3
5
,
2
0
2
5
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
b
s
p
c
.
2
0
2
4
.
1
0
7
0
3
5
.
[
6
]
A
.
B
i
sw
a
s
a
n
d
R
.
B
a
n
i
k
,
"
D
e
t
e
c
t
i
o
n
a
n
d
c
l
a
ss
i
f
i
c
a
t
i
o
n
o
f
d
i
a
b
e
t
i
c
r
e
t
i
n
o
p
a
t
h
y
b
a
s
e
d
o
n
e
n
s
e
m
b
l
e
l
e
a
r
n
i
n
g
,
"
Ad
v
a
n
c
e
s
i
n
C
o
m
p
u
t
a
t
i
o
n
a
l
I
n
t
e
l
l
i
g
e
n
c
e
,
v
o
l
.
4
,
n
o
.
9
,
2
0
2
4
,
d
o
i
:
1
0
.
1
0
0
7
/
s4
3
6
7
4
-
0
2
4
-
0
0
0
7
6
-
4.
[
7
]
G
.
N
a
d
a
r
a
n
d
K
.
S
o
n
a
w
a
n
e
,
"
D
e
t
e
c
t
i
o
n
a
n
d
c
l
a
ssi
f
i
c
a
t
i
o
n
o
f
d
i
a
b
e
t
i
c
r
e
t
i
n
o
p
a
t
h
y
u
s
i
n
g
i
ma
g
e
p
r
o
c
e
ssi
n
g
a
n
d
mac
h
i
n
e
l
e
a
r
n
i
n
g
t
e
c
h
n
i
q
u
e
s,"
C
o
n
g
ress
o
n
S
m
a
rt
C
o
m
p
u
t
i
n
g
T
e
c
h
n
o
l
o
g
i
e
s
,
p
p
.
3
3
-
4
6
,
2
0
2
3
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
981
-
99
-
2
4
6
8
-
4
_
4
.
[
8
]
S
.
R
.
S
y
e
d
a
n
d
S
.
D
u
r
a
i
M
A
,
"
A
d
i
a
g
n
o
s
i
s
m
o
d
e
l
f
o
r
d
e
t
e
c
t
i
o
n
a
n
d
c
l
a
s
si
f
i
c
a
t
i
o
n
o
f
d
i
a
b
e
t
i
c
r
e
t
i
n
o
p
a
t
h
y
u
si
n
g
d
e
e
p
l
e
a
r
n
i
n
g
,
"
N
e
t
w
o
rk
M
o
d
e
l
i
n
g
A
n
a
l
y
si
s
i
n
H
e
a
l
t
h
I
n
f
o
rm
a
t
i
c
s
a
n
d
Bi
o
i
n
f
o
rm
a
t
i
c
s
,
v
o
l
.
1
2
,
n
o
.
3
7
,
2
0
2
3
,
d
o
i
:
1
0
.
1
0
0
7
/
s1
3
7
2
1
-
0
2
3
-
0
0
4
3
2
-
3.
[
9
]
S
.
D
a
s
a
n
d
S
.
K
.
S
a
h
a
,
"
D
i
a
b
e
t
i
c
r
e
t
i
n
o
p
a
t
h
y
d
e
t
e
c
t
i
o
n
a
n
d
c
l
a
ss
i
f
i
c
a
t
i
o
n
u
s
i
n
g
C
N
N
t
u
n
e
d
b
y
g
e
n
e
t
i
c
a
l
g
o
r
i
t
h
m,"
Mu
l
t
i
m
e
d
i
a
T
o
o
l
s
a
n
d
Ap
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
8
1
,
p
p
.
8
0
0
7
–
8
0
2
0
,
2
0
2
2
,
d
o
i
:
1
0
.
1
0
0
7
/
s1
1
0
4
2
-
0
2
1
-
1
1
8
2
4
-
w.
[
1
0
]
U
.
B
h
i
m
a
v
a
r
a
p
u
,
N
.
C
h
i
n
t
a
l
a
p
u
d
i
a
n
d
G
.
B
a
t
t
i
n
e
n
i
,
"
A
u
t
o
m
a
t
i
c
d
e
t
e
c
t
i
o
n
a
n
d
c
l
a
ss
i
f
i
c
a
t
i
o
n
o
f
d
i
a
b
e
t
i
c
r
e
t
i
n
o
p
a
t
h
y
u
s
i
n
g
t
h
e
i
mp
r
o
v
e
d
p
o
o
l
i
n
g
f
u
n
c
t
i
o
n
i
n
t
h
e
c
o
n
v
o
l
u
t
i
o
n
n
e
u
r
a
l
n
e
t
w
o
r
k
,
"
D
i
a
g
n
o
s
t
i
c
s
,
v
o
l
.
1
3
,
n
o
.
1
5
,
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
d
i
a
g
n
o
st
i
c
s
1
3
1
5
2
6
0
6
.
[
1
1
]
B
.
L
a
l
i
t
h
a
d
e
v
i
a
n
d
S
.
K
r
i
s
h
n
a
v
e
n
i
,
"
D
i
a
b
e
t
i
c
r
e
t
i
n
o
p
a
t
h
y
d
e
t
e
c
t
i
o
n
a
n
d
s
e
v
e
r
i
t
y
c
l
a
s
s
i
f
i
c
a
t
i
o
n
u
s
i
n
g
o
p
t
i
m
i
z
e
d
d
e
e
p
l
e
a
r
n
i
n
g
w
i
t
h
e
x
p
l
a
i
n
a
b
l
e
A
I
t
e
c
h
n
i
q
u
e
,
"
M
u
l
t
i
m
e
d
i
a
T
o
o
l
s
a
n
d
A
p
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
8
3
,
p
p
.
8
9
9
4
9
–
9
0
0
1
3
,
2
0
2
4
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
1
0
4
2
-
024
-
1
8
8
6
3
-
z.
[
1
2
]
C
.
M
o
h
a
n
t
y
e
t
a
l
.
,
"
U
si
n
g
D
e
e
p
Le
a
r
n
i
n
g
A
r
c
h
i
t
e
c
t
u
r
e
s
f
o
r
D
e
t
e
c
t
i
o
n
a
n
d
C
l
a
ssi
f
i
c
a
t
i
o
n
o
f
D
i
a
b
e
t
i
c
R
e
t
i
n
o
p
a
t
h
y
,
"
S
e
n
s
o
rs
,
v
o
l
.
2
3
,
n
o
.
1
2
,
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
s2
3
1
2
5
7
2
6
.
[
1
3
]
C
.
S
u
e
d
u
mr
o
n
g
,
S
.
P
h
o
n
g
mo
o
,
T
.
A
k
a
r
a
j
a
k
a
a
n
d
K
.
Le
k
s
a
k
u
l
,
"
D
i
a
b
e
t
i
c
r
e
t
i
n
o
p
a
t
h
y
d
e
t
e
c
t
i
o
n
u
s
i
n
g
c
o
n
v
o
l
u
t
i
o
n
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
s
w
i
t
h
b
a
c
k
g
r
o
u
n
d
r
e
mo
v
a
l
,
a
n
d
d
a
t
a
a
u
g
me
n
t
a
t
i
o
n
,
"
Ap
p
l
i
e
d
S
c
i
e
n
c
e
s
,
v
o
l
.
1
4
,
n
o
.
1
9
,
p
.
8
8
2
3
,
2
0
2
4
,
d
o
i
:
1
0
.
3
3
9
0
/
a
p
p
1
4
1
9
8
8
2
3
.
[
1
4
]
Y
a
sash
v
i
n
i
,
V
.
R
a
j
a
S
a
r
o
b
i
n
M
.
,
R
.
P
a
n
j
a
n
a
t
h
a
n
,
G
.
Jasm
i
n
e
,
a
n
d
J.
A
n
b
a
r
a
si
,
“
D
i
a
b
e
t
i
c
r
e
t
i
n
o
p
a
t
h
y
c
l
a
ss
i
f
i
c
a
t
i
o
n
u
si
n
g
C
N
N
a
nd
h
y
b
r
i
d
d
e
e
p
c
o
n
v
o
l
u
t
i
o
n
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
s,
”
S
y
m
m
e
t
ry
,
v
o
l
.
1
4
,
n
o
.
9
,
p
.
1
9
3
2
,
2
0
2
2
,
d
o
i
:
1
0
.
3
3
9
0
/
s
y
m
1
4
0
9
1
9
32
.
[
1
5
]
S
.
S
u
n
d
a
r
,
S
.
S
u
b
r
a
ma
n
i
a
n
,
a
n
d
M
.
M
a
h
m
u
d
,
“
C
l
a
s
si
f
i
c
a
t
i
o
n
o
f
d
i
a
b
e
t
i
c
r
e
t
i
n
o
p
a
t
h
y
d
i
sea
s
e
l
e
v
e
l
s
b
y
e
x
t
r
a
c
t
i
n
g
s
p
e
c
t
r
a
l
f
e
a
t
u
r
e
s
u
si
n
g
W
a
v
e
l
e
t
C
N
N
,
”
D
i
a
g
n
o
s
t
i
c
s,
v
o
l
.
1
4
,
n
o
.
1
1
,
p
.
1
0
9
3
,
2
0
2
4
,
d
o
i
:
1
0
.
3
3
9
0
/
d
i
a
g
n
o
s
t
i
c
s1
4
1
1
1
0
9
3
.
[
1
6
]
M
.
Y
o
u
l
d
a
s
h
e
t
a
l
.
,
“
Ea
r
l
y
d
e
t
e
c
t
i
o
n
a
n
d
c
l
a
ss
i
f
i
c
a
t
i
o
n
o
f
d
i
a
b
e
t
i
c
r
e
t
i
n
o
p
a
t
h
y
:
A
d
e
e
p
l
e
a
r
n
i
n
g
a
p
p
r
o
a
c
h
,
”
AI
,
v
o
l
.
5
,
n
o
.
4
,
p
p
.
2
5
8
6
–
2
6
1
7
,
2
0
2
4
,
d
o
i
:
1
0
.
3
3
9
0
/
a
i
5
0
4
0
1
2
5
.
[
1
7
]
I
n
a
mu
l
l
a
h
,
S
.
H
a
ss
a
n
,
N
.
A
.
A
l
r
a
j
e
h
,
E.
A
.
M
o
h
a
mm
e
d
,
a
n
d
S
.
K
h
a
n
,
“
D
a
t
a
d
i
v
e
r
s
i
t
y
i
n
c
o
n
v
o
l
u
t
i
o
n
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
-
b
a
s
e
d
e
n
s
e
mb
l
e
mo
d
e
l
f
o
r
d
i
a
b
e
t
i
c
r
e
t
i
n
o
p
a
t
h
y
,
”
B
i
o
m
i
me
t
i
c
s,
v
o
l
.
8
,
n
o
.
2
,
p
.
1
8
7
,
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
b
i
o
m
i
me
t
i
c
s8
0
2
0
1
8
7
.
[
1
8
]
M
.
H
a
r
d
a
s,
S
.
M
a
t
h
u
r
,
A
.
B
h
a
s
k
a
r
,
a
n
d
M
.
K
a
l
l
a
,
“
R
e
t
i
n
a
l
f
u
n
d
u
s
i
m
a
g
e
c
l
a
ssi
f
i
c
a
t
i
o
n
f
o
r
d
i
a
b
e
t
i
c
r
e
t
i
n
o
p
a
t
h
y
u
s
i
n
g
S
V
M
p
r
e
d
i
c
t
i
o
n
s,”
P
h
y
s
i
c
a
l
a
n
d
e
n
g
i
n
e
e
ri
n
g
sc
i
e
n
c
e
s
i
n
m
e
d
i
c
i
n
e
,
v
o
l
.
4
5
,
n
o
.
3
,
p
p
.
7
8
1
–
7
9
1
,
2
0
2
2
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
3
2
4
6
-
0
2
2
-
0
1
1
4
3
-
1
.
[
1
9
]
M
.
K
.
B
e
h
e
r
a
a
n
d
S
.
C
h
a
k
r
a
v
a
r
t
y
,
"
D
i
a
b
e
t
i
c
r
e
t
i
n
o
p
a
t
h
y
i
m
a
g
e
c
l
a
ssi
f
i
c
a
t
i
o
n
u
si
n
g
s
u
p
p
o
r
t
v
e
c
t
o
r
ma
c
h
i
n
e
,
"
2
0
2
0
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
E
n
g
i
n
e
e
ri
n
g
a
n
d
Ap
p
l
i
c
a
t
i
o
n
s
(
I
C
C
S
E
A)
,
G
u
n
u
p
u
r
,
I
n
d
i
a
,
2
0
2
0
,
p
p
.
1
-
4
,
d
o
i
:
1
0
.
1
1
0
9
/
I
C
C
S
EA
4
9
1
4
3
.
2
0
2
0
.
9
1
3
2
8
7
5
.
[
2
0
]
S
.
S
u
b
r
a
ma
n
i
a
n
,
S
.
M
i
s
h
r
a
,
S
.
P
a
t
i
l
,
K
.
S
h
a
w
,
a
n
d
E.
A
g
h
a
j
a
r
i
,
“
M
a
c
h
i
n
e
l
e
a
r
n
i
n
g
st
y
l
e
s
f
o
r
d
i
a
b
e
t
i
c
r
e
t
i
n
o
p
a
t
h
y
d
e
t
e
c
t
i
o
n
:
A
r
e
v
i
e
w
a
n
d
b
i
b
l
i
o
met
r
i
c
a
n
a
l
y
s
i
s,”
Bi
g
D
a
t
a
a
n
d
C
o
g
n
i
t
i
v
e
C
o
m
p
u
t
i
n
g
,
v
o
l
.
6
,
n
o
.
4
,
p
.
1
5
4
,
2
0
2
2
,
d
o
i
:
1
0
.
3
3
9
0
/
b
d
c
c
6
0
4
0
1
5
4
.
[
2
1
]
T.
P
.
H
.
N
i
ss
e
n
e
t
a
l
.
,
“
P
e
r
f
o
r
m
a
n
c
e
o
f
a
s
u
p
p
o
r
t
v
e
c
t
o
r
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
t
o
o
l
f
o
r
d
i
a
g
n
o
s
i
n
g
d
i
a
b
e
t
i
c
r
e
t
i
n
o
p
a
t
h
y
i
n
c
l
i
n
i
c
a
l
p
r
a
c
t
i
c
e
,
”
J
o
u
r
n
a
l
o
f
Pe
rs
o
n
a
l
i
z
e
d
M
e
d
i
c
i
n
e
,
v
o
l
.
1
3
,
n
o
.
7
,
p
.
1
1
2
8
,
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
j
p
m
1
3
0
7
1
1
2
8
.
[
2
2
]
A
.
A
b
i
r
a
mi
a
n
d
R
.
K
a
v
i
t
h
a
,
“
A
n
e
f
f
i
c
i
e
n
t
e
a
r
l
y
d
e
t
e
c
t
i
o
n
o
f
d
i
a
b
e
t
i
c
r
e
t
i
n
o
p
a
t
h
y
u
s
i
n
g
d
w
a
r
f
m
o
n
g
o
o
se
o
p
t
i
mi
z
a
t
i
o
n
b
a
s
e
d
d
e
e
p
b
e
l
i
e
f
n
e
t
w
o
r
k
,
”
C
o
n
c
u
rr
e
n
c
y
a
n
d
C
o
m
p
u
t
a
t
i
o
n
.
,
v
o
l
.
3
4
,
n
o
.
2
8
,
2
0
2
2
,
d
o
i
:
1
0
.
1
0
0
2
/
c
p
e
.
7
3
6
4
.
[
2
3
]
A
.
S
i
n
g
h
,
R
.
K
u
mar,
a
n
d
A
.
H
.
G
a
n
d
o
mi
,
“
A
d
a
p
t
i
v
e
i
so
ma
p
f
e
a
t
u
r
e
e
x
t
r
a
c
t
i
v
e
g
r
a
d
i
e
n
t
d
e
e
p
b
e
l
i
e
f
n
e
t
w
o
r
k
c
l
a
s
si
f
i
e
r
f
o
r
d
i
a
b
e
t
i
c
r
e
t
i
n
o
p
a
t
h
y
i
d
e
n
t
i
f
i
c
a
t
i
o
n
,
”
Mu
l
t
i
m
e
d
i
a
T
o
o
l
s
a
n
d
Ap
p
l
i
c
a
t
i
o
n
s
,
2
0
2
4
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
1
0
4
2
-
0
2
4
-
1
9
2
1
6
-
6
.
[
2
4
]
S
.
K
r
i
sh
n
a
m
o
o
r
t
h
y
,
Y
.
W
e
i
f
e
n
g
,
J.
L
u
o
,
a
n
d
S
.
K
a
d
r
y
,
“
G
O
-
D
B
N
:
G
a
n
n
e
t
o
p
t
i
mi
z
e
d
d
e
e
p
b
e
l
i
e
f
n
e
t
w
o
r
k
b
a
s
e
d
w
a
v
e
l
e
t
k
e
r
n
e
l
ELM
f
o
r
d
e
t
e
c
t
i
o
n
o
f
d
i
a
b
e
t
i
c
r
e
t
i
n
o
p
a
t
h
y
,
”
Ex
p
e
r
t
S
y
st
e
m
s
w
i
t
h
Ap
p
l
i
c
a
t
i
o
n
s:
A
n
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
rn
a
l
,
v
o
l
.
2
2
9
,
n
o
.
1
2
0
4
0
8
,
p
.
1
2
0
4
0
8
,
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
e
sw
a
.
2
0
2
3
.
1
2
0
4
0
8
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
5
2
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
40
,
No
.
1
,
Octo
b
er
20
25
:
1
64
-
1
72
172
[
2
5
]
C.
-
L.
Li
n
a
n
d
K
.
-
C
.
W
u
,
“
D
e
v
e
l
o
p
m
e
n
t
o
f
r
e
v
i
se
d
R
e
sN
e
t
-
5
0
f
o
r
d
i
a
b
e
t
i
c
r
e
t
i
n
o
p
a
t
h
y
d
e
t
e
c
t
i
o
n
,
”
BM
C
Bi
o
i
n
f
o
rm
a
t
i
c
s
,
v
o
l
.
2
4
,
n
o
.
1
,
2
0
2
3
,
d
o
i
:
1
0
.
1
1
8
6
/
s
1
2
8
5
9
-
0
2
3
-
0
5
2
9
3
-
1.
[
2
6
]
Y
.
H
u
a
n
g
,
L
.
Li
n
,
P
.
C
h
e
n
g
,
J.
L
y
u
,
R
.
Ta
m
,
a
n
d
X
.
Ta
n
g
,
“
I
d
e
n
t
i
f
y
i
n
g
t
h
e
k
e
y
c
o
mp
o
n
e
n
t
s
i
n
R
e
sN
e
t
-
5
0
f
o
r
d
i
a
b
e
t
i
c
r
e
t
i
n
o
p
a
t
h
y
g
r
a
d
i
n
g
f
r
o
m
f
u
n
d
u
s
i
ma
g
e
s
:
A
sy
st
e
ma
t
i
c
i
n
v
e
st
i
g
a
t
i
o
n
,
”
D
i
a
g
n
o
s
t
i
c
s
,
v
o
l
.
1
3
,
n
o
.
1
0
,
p
.
1
6
6
4
,
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
d
i
a
g
n
o
s
t
i
c
s1
3
1
0
1
6
6
4
.
[
2
7
]
M
.
K
.
Y
a
q
o
o
b
,
S
.
F
.
A
l
i
,
M
.
B
i
l
a
l
,
M
.
S
.
H
a
n
i
f
,
a
n
d
U
.
M
.
A
l
-
S
a
g
g
a
f
,
“
R
e
s
N
e
t
b
a
s
e
d
d
e
e
p
f
e
a
t
u
r
e
s
a
n
d
r
a
n
d
o
m
f
o
r
e
st
c
l
a
ss
i
f
i
e
r
f
o
r
d
i
a
b
e
t
i
c
r
e
t
i
n
o
p
a
t
h
y
d
e
t
e
c
t
i
o
n
,
”
S
e
n
so
rs
,
v
o
l
.
2
1
,
n
o
.
1
1
,
p
.
3
8
8
3
,
2
0
2
1
,
d
o
i
:
1
0
.
3
3
9
0
/
s
2
1
1
1
3
8
8
3
.
[
2
8
]
U
.
I
sh
t
i
a
q
,
E.
R
.
M
.
F
.
A
b
d
u
l
l
a
h
,
a
n
d
Z.
I
sh
t
i
a
q
u
e
,
“
A
h
y
b
r
i
d
t
e
c
h
n
i
q
u
e
f
o
r
d
i
a
b
e
t
i
c
r
e
t
i
n
o
p
a
t
h
y
d
e
t
e
c
t
i
o
n
b
a
sed
o
n
e
n
sem
b
l
e
-
o
p
t
i
m
i
z
e
d
C
N
N
a
n
d
t
e
x
t
u
r
e
f
e
a
t
u
r
e
s,”
D
i
a
g
n
o
s
t
i
c
s
,
v
o
l
.
1
3
,
n
o
.
1
0
,
p
.
1
8
1
6
,
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
d
i
a
g
n
o
s
t
i
c
s1
3
1
0
1
8
1
6
.
[
2
9
]
M
.
Y
i
n
g
h
u
a
,
Y
.
H
e
n
g
,
R
.
A
mar
n
a
t
h
a
n
d
Z.
H
u
i
,
"
H
a
r
d
e
x
u
d
a
t
e
s
se
g
me
n
t
a
t
i
o
n
i
n
d
i
a
b
e
t
i
c
r
e
t
i
n
o
p
a
t
h
y
u
s
i
n
g
D
i
a
R
e
t
D
B
1
,
"
i
n
I
EE
E
Ac
c
e
ss
,
v
o
l
.
1
2
,
p
p
.
1
2
6
4
8
6
-
1
2
6
5
0
2
,
2
0
2
4
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
4
.
3
4
5
5
4
3
3
.
[
3
0
]
A
.
R
.
W
.
S
a
i
t
,
“
A
l
i
g
h
t
w
e
i
g
h
t
d
i
a
b
e
t
i
c
r
e
t
i
n
o
p
a
t
h
y
d
e
t
e
c
t
i
o
n
mo
d
e
l
u
s
i
n
g
a
d
e
e
p
-
l
e
a
r
n
i
n
g
t
e
c
h
n
i
q
u
e
,
”
D
i
a
g
n
o
s
t
i
c
s
,
v
o
l
.
1
3
,
n
o
.
1
9
,
p
.
3
1
2
0
,
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
d
i
a
g
n
o
s
t
i
c
s
1
3
1
9
3
1
2
0
.
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
Priy
a
n
k
a
S
r
e
e
n
iv
a
s
wit
h
te
a
c
h
in
g
e
x
p
e
rien
c
e
o
f
8
y
e
a
rs,
is
wo
rk
in
g
a
s
a
n
As
sista
n
t
p
ro
fe
ss
o
r
in
th
e
De
p
a
r
tme
n
t
o
f
C
o
m
p
u
ter
S
c
ien
c
e
a
n
d
En
g
in
e
e
rin
g
a
t
B
.
N.
M
In
stit
u
te
o
f
Tec
h
n
o
lo
g
y
,
Be
n
g
a
l
u
ru
,
I
n
d
ia.
S
h
e
is
a
re
se
a
rc
h
sc
h
o
lar
a
t
Da
y
a
n
a
n
d
S
a
g
a
r
Un
iv
e
rsity
,
Be
n
g
a
lu
r
u
.
S
h
e
h
o
l
d
s
a
n
M
.
Tec
h
d
e
g
re
e
i
n
C
o
m
p
u
ter
S
c
ien
c
e
a
n
d
E
n
g
i
n
e
e
rin
g
.
He
r
re
se
a
rc
h
a
re
a
s
a
re
D
e
e
p
Lea
rn
in
g
,
m
e
d
ica
l
ima
g
e
a
n
a
ly
sis
,
a
n
d
p
a
tt
e
r
n
re
c
o
g
n
it
io
n
.
S
h
e
is
c
e
rti
fied
a
s
a
c
o
a
c
h
fo
r
th
e
AI
Yo
u
th
P
r
o
g
ra
m
m
e
fro
m
De
ll
-
In
te
l.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
p
riy
a
n
k
a
s.v
a
h
i
n
@g
m
a
il
.
c
o
m
.
K
a
v
ita
V.
H
o
r
a
d
i
re
c
e
iv
e
d
h
e
r
B.
E.
,
M
.
Tec
h
a
n
d
P
h
.
D.
i
n
C
o
m
p
u
ter
S
c
ien
c
e
a
n
d
En
g
in
e
e
ri
n
g
fr
o
m
Visv
e
sv
a
ra
y
a
Tec
h
n
o
l
o
g
ica
l
Un
i
v
e
rsity
,
B
e
lg
a
u
m
.
Wi
th
1
5
y
e
a
rs
o
f
rich
e
x
p
e
rie
n
c
e
in
tea
c
h
in
g
in
h
ig
h
e
r
e
d
u
c
a
ti
o
n
sh
e
is
c
u
rre
n
t
ly
wo
rk
i
n
g
a
s
a
n
As
so
c
iate
P
ro
fe
ss
o
r
a
t
De
p
t.
o
f
C
o
m
p
u
ter
S
c
ien
c
e
a
n
d
E
n
g
i
n
e
e
rin
g
,
B.
N
.
M
.
In
st
it
u
te
o
f
Tec
h
n
o
l
o
g
y
,
Ba
n
g
a
lo
re
.
He
r
re
se
a
rc
h
in
tere
sts
a
re
d
o
c
u
m
e
n
t
ima
g
e
p
r
o
c
e
ss
in
g
,
c
o
m
p
u
ter
v
isio
n
,
m
a
c
h
i
n
e
lea
rn
in
g
,
a
n
d
d
e
e
p
lea
rn
in
g
.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
a
t
k
a
v
it
a
.
b
n
m
it
3
@g
m
a
il
.
c
o
m
.
K
a
lp
a
Ra
ja
s
h
e
k
a
r
is
a
r
e
se
a
rc
h
sc
h
o
lar
a
t
Visv
e
sv
a
ra
y
a
Tec
h
n
o
l
o
g
ica
l
Un
iv
e
rsity
.
S
h
e
h
a
s
h
e
r
M
-
Tec
h
in
C
o
m
p
u
ter
S
c
ien
c
e
fro
m
Visv
e
sv
a
ra
y
a
Tec
h
n
o
l
o
g
ica
l
Un
iv
e
rsity
,
Ka
rn
a
tak
a
.
He
r
M
a
ste
r’s
is
in
Co
m
p
u
ter
S
c
ie
n
c
e
fro
m
BNM
In
sti
tu
te
o
f
Tec
h
n
o
l
o
g
y
,
a
n
d
h
e
r
Ba
c
h
e
lo
r'
s
i
n
In
f
o
rm
a
ti
o
n
S
c
ien
c
e
fr
o
m
Am
rit
a
In
stit
u
te
o
f
Tec
h
n
o
lo
g
y
.
He
r
a
re
a
o
f
in
tere
st
is
in
Co
m
p
u
ter
Ne
two
rk
s
a
n
d
c
ry
p
to
g
ra
p
h
y
.
S
h
e
is
a
F
a
c
u
l
ty
o
f
th
e
Co
m
p
u
ter
S
c
ien
c
e
d
e
p
a
rtme
n
t
a
t
BNM
In
stit
u
te
o
f
Tec
h
n
o
l
o
g
y
.
Be
fo
re
se
rv
in
g
a
t
BNM
In
stit
u
te
o
f
Tec
h
n
o
lo
g
y
,
sh
e
h
a
d
a
q
u
ick
st
in
t
a
t
Da
y
a
n
a
n
d
a
S
a
g
a
r
I
n
stit
u
t
io
n
s.
S
h
e
sta
rted
h
e
r
c
a
re
e
r
a
t
th
e
In
d
ian
S
p
a
c
e
Re
se
a
rc
h
Org
a
n
iza
ti
o
n
(IS
RO).
S
h
e
c
a
n
b
e
c
o
n
tac
ted
a
t
k
a
lp
a
.
ra
jas
h
e
k
a
r@g
m
a
il
.
c
o
m
.
Evaluation Warning : The document was created with Spire.PDF for Python.