I
AE
S In
t
er
na
t
io
na
l J
o
urna
l o
f
Ro
bo
t
ics a
nd
Aut
o
m
a
t
io
n
(
I
J
RA)
Vo
l.
14
,
No
.
4
,
Dec
em
b
er
2
0
2
5
,
p
p
.
5
7
0
~
5
8
0
I
SS
N:
2722
-
2
5
8
6
,
DOI
:
1
0
.
1
1
5
9
1
/i
jr
a
.
v
14
i
4
.
pp
5
7
0
-
580
570
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
r
a
.
ia
esco
r
e.
co
m
Enha
ncing
health
status pr
ediction
a
nd da
ta securi
ty
using
trans
former
-
ba
sed deep
learning
a
rchitec
tures
Su
bra
m
a
niy
a
n Sent
ha
m
a
ra
i
1
,
Ra
j
a
M
a
nick
a
m
M
a
la
2
,
Ve
lla
iy
a
n P
a
la
nis
a
m
y
1
1
D
e
p
a
r
t
me
n
t
o
f
C
o
m
p
u
t
e
r
A
p
p
l
i
c
a
t
i
o
n
s,
A
l
a
g
a
p
p
a
U
n
i
v
e
r
si
t
y
,
K
a
r
a
i
k
u
d
i
,
T
a
mi
l
n
a
d
u
,
I
n
d
i
a
.
2
D
e
p
a
r
t
me
n
t
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
G
o
v
e
r
n
m
e
n
t
A
r
t
s
a
n
d
S
c
i
e
n
c
e
C
o
l
l
e
g
e
f
o
r
W
o
me
n
,
P
a
r
a
ma
k
u
d
i
,
Ta
m
i
l
n
a
d
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
Au
g
17
,
2
0
2
4
R
ev
is
ed
J
u
l
26
,
2
0
2
5
Acc
ep
ted
Au
g
26
,
2
0
2
5
Th
is
p
a
p
e
r
p
r
o
p
o
se
s
a
p
riv
a
c
y
-
p
re
se
rv
in
g
tran
sfo
rm
e
r
-
b
a
se
d
fe
d
e
ra
ted
lea
rn
in
g
(P
P
TF
L)
fra
m
e
wo
rk
d
e
sig
n
e
d
to
e
n
h
a
n
c
e
p
r
iv
a
c
y
,
a
c
c
u
ra
c
y
,
a
n
d
c
o
m
p
u
tati
o
n
a
l
e
fficie
n
c
y
i
n
h
e
a
lt
h
c
a
re
d
a
ta
a
n
a
ly
sis.
F
e
d
e
ra
ted
lea
rn
in
g
(F
L)
h
a
s
e
m
e
rg
e
d
a
s
a
p
ro
m
isin
g
so
l
u
ti
o
n
f
o
r
d
istri
b
u
te
d
m
a
c
h
in
e
lea
rn
in
g
wh
il
e
p
re
se
rv
in
g
d
a
ta
p
ri
v
a
c
y
,
e
sp
e
c
ially
i
n
se
n
si
ti
v
e
se
c
to
rs
li
k
e
h
e
a
lt
h
c
a
re
.
Ho
we
v
e
r,
c
h
a
ll
e
n
g
e
s
su
c
h
a
s
m
a
in
tain
in
g
h
i
g
h
a
c
c
u
ra
c
y
a
n
d
m
a
n
a
g
in
g
c
o
m
m
u
n
ica
ti
o
n
o
v
e
rh
e
a
d
re
m
a
in
.
T
h
e
p
ro
p
o
se
d
P
P
T
F
L
f
ra
m
e
wo
rk
lev
e
ra
g
e
s
th
e
p
o
we
r
o
f
tran
sfo
r
m
e
r
m
o
d
e
ls
t
o
im
p
ro
v
e
t
h
e
p
e
rf
o
rm
a
n
c
e
o
f
fe
d
e
ra
ted
lea
rn
i
n
g
wh
il
e
i
n
teg
r
a
ti
n
g
p
ri
v
a
c
y
-
p
re
se
rv
i
n
g
tec
h
n
i
q
u
e
s.
Th
e
m
o
d
e
l
d
e
m
o
n
stra
tes
su
p
e
rio
r
p
e
rfo
rm
a
n
c
e
with
a
n
a
c
c
u
ra
c
y
o
f
9
2
.
8
7
%
,
a
n
F
1
sc
o
re
o
f
9
2
.
3
7
%
,
a
n
d
a
p
ri
v
a
c
y
b
u
d
g
e
t
(ϵ
)
o
f
1
.
6
,
o
u
tp
e
rf
o
rm
in
g
e
x
ist
in
g
a
p
p
ro
a
c
h
e
s
in
term
s
o
f
b
o
t
h
p
ri
v
a
c
y
a
n
d
a
c
c
u
ra
c
y
.
Th
e
m
o
d
e
l
a
lso
e
x
h
i
b
it
s
c
o
m
p
u
tati
o
n
a
l
e
fficie
n
c
y
,
wi
th
l
o
we
r
c
o
m
m
u
n
ica
ti
o
n
c
o
st
a
n
d
r
e
a
so
n
a
b
le
train
in
g
ti
m
e
.
Co
m
p
a
ra
ti
v
e
e
v
a
lu
a
ti
o
n
s
wit
h
fo
u
r
re
lev
a
n
t
li
tera
tu
re
m
o
d
e
ls
fu
rth
e
r
v
a
li
d
a
te
th
e
e
ffe
c
ti
v
e
n
e
ss
o
f
th
e
p
r
o
p
o
se
d
P
P
TF
L
fra
m
e
wo
rk
.
T
h
is
wo
rk
h
i
g
h
li
g
h
ts
t
h
e
p
o
te
n
ti
a
l
o
f
P
P
TF
L
to
re
v
o
lu
ti
o
n
ize
h
e
a
lt
h
c
a
re
in
fo
rm
a
ti
c
s
b
y
p
ro
v
id
in
g
se
c
u
r
e
,
a
c
c
u
ra
te,
a
n
d
e
fficie
n
t
so
l
u
ti
o
n
s
fo
r
fe
d
e
ra
ted
lea
rn
in
g
a
p
p
li
c
a
ti
o
n
s.
K
ey
w
o
r
d
s
:
Atten
tio
n
m
ec
h
an
is
m
C
las
s
if
icatio
n
Dee
p
lear
n
in
g
Fed
er
ated
lear
n
in
g
Pre
d
ictio
n
T
r
an
s
f
o
r
m
e
r
b
ased
lear
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
:
Su
b
r
am
an
iy
a
n
Sen
th
am
ar
ai
Dep
ar
tm
en
t o
f
C
o
m
p
u
ter
Scie
n
ce
,
Alag
ap
p
a
Un
iv
er
s
ity
Kar
aik
u
d
i,
I
n
d
ia
E
m
ail
-
s
ak
th
is
iv
a2
1
2
5
@
g
m
ail.
co
m
1.
I
NT
RO
D
UCT
I
O
N
T
h
e
r
a
p
id
d
i
g
itizatio
n
o
f
h
e
alth
ca
r
e
an
d
th
e
wid
esp
r
ea
d
ad
o
p
ti
o
n
o
f
elec
tr
o
n
ic
h
ea
l
th
r
ec
o
r
d
s
(
E
HR
s
)
,
wea
r
ab
le
d
e
v
ices,
an
d
co
n
n
ec
ted
m
ed
ical
s
en
s
o
r
s
h
av
e
s
ig
n
if
ica
n
tly
tr
a
n
s
f
o
r
m
e
d
m
o
d
er
n
m
ed
ici
n
e,
m
ak
in
g
d
ata
a
cr
itical
ass
et
i
n
clin
ical
d
ec
is
io
n
-
m
ak
in
g
an
d
h
ea
lth
ca
r
e
d
eliv
er
y
.
W
ith
th
e
p
r
o
life
r
atio
n
o
f
m
ed
ical
b
ig
d
ata,
th
e
r
e
is
a
g
r
o
win
g
o
p
p
o
r
tu
n
ity
to
u
tili
ze
m
ac
h
in
e
lear
n
in
g
an
d
d
ee
p
lear
n
in
g
m
o
d
els
to
ex
tr
ac
t
ac
tio
n
a
b
le
in
s
ig
h
ts
,
p
r
ed
ict
d
is
ea
s
e
o
u
tco
m
es,
p
e
r
s
o
n
alize
tr
ea
tm
en
ts
,
a
n
d
u
ltima
t
ely
im
p
r
o
v
e
p
atien
t
ca
r
e
[
1
]
–
[
3
]
.
Ho
wev
er
,
t
h
e
in
c
r
ea
s
in
g
s
en
s
itiv
ity
o
f
m
ed
ical
d
ata
an
d
s
tr
in
g
e
n
t
r
eg
u
lato
r
y
f
r
am
ewo
r
k
s
s
u
ch
as
th
e
g
e
n
e
r
a
l d
at
a
p
r
o
tec
ti
o
n
r
e
g
u
la
ti
o
n
(
GDPR
)
p
o
s
e
m
aj
o
r
ch
alle
n
g
es
f
o
r
d
a
ta
s
h
ar
i
n
g
an
d
c
en
t
r
aliz
e
d
l
ea
r
n
i
n
g
.
I
n
th
is
co
n
tex
t,
f
e
d
er
ated
lea
r
n
in
g
(
FL)
h
as
em
er
g
e
d
as
a
p
r
o
m
is
in
g
p
a
r
ad
ig
m
t
h
at
en
ab
les
co
llab
o
r
ativ
e
m
o
d
el
tr
ain
in
g
ac
r
o
s
s
m
u
ltip
l
e
d
ec
en
tr
alize
d
h
ea
lth
ca
r
e
in
s
titu
tio
n
s
with
o
u
t
ex
p
o
s
in
g
s
en
s
itiv
e
p
atien
t
d
ata.
Desp
ite
its
th
eo
r
etica
l
ad
v
an
t
ag
es,
co
n
v
en
tio
n
al
f
ed
er
ate
d
l
ea
r
n
in
g
ap
p
r
o
ac
h
es
f
ac
e
s
ev
e
r
al
lim
itatio
n
s
s
u
ch
as
v
u
ln
er
ab
ilit
y
to
d
ata
h
eter
o
g
en
eity
,
lack
o
f
r
o
b
u
s
tn
ess
ag
ain
s
t
ad
v
er
s
ar
ial
attac
k
s
,
h
ig
h
co
m
m
u
n
icatio
n
co
s
ts
,
an
d
in
ad
eq
u
ate
s
u
p
p
o
r
t
f
o
r
d
y
n
am
ic
m
o
d
el
u
p
d
atin
g
in
r
eso
u
r
ce
-
c
o
n
s
tr
ain
ed
e
d
g
e
en
v
ir
o
n
m
en
ts
[
4
]
,
[
5
]
.
Mo
r
eo
v
er
,
tr
ad
itio
n
al
r
ec
u
r
r
en
t
n
eu
r
al
n
etwo
r
k
(
R
NN)
-
b
ased
ar
c
h
itectu
r
es
o
f
te
n
f
all
s
h
o
r
t
i
n
h
a
n
d
lin
g
co
m
p
lex
tem
p
o
r
al
d
e
p
en
d
en
ci
es a
n
d
lo
n
g
-
r
an
g
e
in
ter
ac
tio
n
s
p
r
esen
t in
m
ed
ical
tim
e
-
s
er
ies
d
ata.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
AE
S
I
n
t
J
R
o
b
&
A
u
to
m
I
SS
N:
2722
-
2
5
8
6
E
n
h
a
n
cin
g
h
ea
lth
s
ta
tu
s
p
r
ed
ictio
n
a
n
d
d
a
ta
s
ec
u
r
ity
…
(
S
u
b
r
a
ma
n
iya
n
S
en
th
a
ma
r
a
i
)
571
T
o
ad
d
r
ess
th
ese
ch
allen
g
es,
t
h
e
in
tr
o
d
u
ctio
n
o
f
tr
an
s
f
o
r
m
er
-
b
ased
ar
ch
itectu
r
es,
o
r
ig
in
all
y
d
esig
n
ed
f
o
r
n
at
u
r
al
lan
g
u
ag
e
p
r
o
ce
s
s
in
g
,
o
f
f
er
s
a
p
o
wer
f
u
l
alter
n
ativ
e
d
u
e
to
th
eir
s
elf
-
atte
n
tio
n
m
ec
h
a
n
is
m
s
,
s
ca
lab
ilit
y
,
an
d
p
ar
allel
p
r
o
ce
s
s
in
g
ca
p
ab
ilit
ies.
T
r
an
s
f
o
r
m
er
s
h
av
e
s
h
o
wn
ex
ce
p
tio
n
al
p
er
f
o
r
m
an
ce
in
m
o
d
elin
g
s
eq
u
en
tial
d
ata
b
y
ca
p
tu
r
in
g
g
lo
b
al
d
ep
en
d
en
cie
s
m
o
r
e
ef
f
ec
tiv
ely
th
an
R
NNs
o
r
C
NNs,
m
ak
in
g
th
em
p
ar
ticu
lar
ly
s
u
itab
le
f
o
r
h
ea
lth
ca
r
e
ap
p
licatio
n
s
s
u
ch
a
s
p
atien
t tr
ajec
to
r
y
p
r
ed
ictio
n
,
d
is
ea
s
e
p
r
o
g
r
ess
io
n
m
o
d
elin
g
,
an
d
clin
ical
r
is
k
s
tr
atif
icatio
n
[
6
]
–
[
8
]
.
Ho
wev
er
,
th
eir
in
teg
r
atio
n
with
in
p
r
iv
ac
y
-
p
r
eser
v
in
g
f
ed
er
ated
en
v
ir
o
n
m
en
ts
r
em
ai
n
s
r
elativ
ely
u
n
ex
p
l
o
r
ed
an
d
d
em
an
d
s
n
o
v
el
d
esig
n
c
o
n
s
id
er
atio
n
s
to
b
alan
ce
co
m
p
u
tatio
n
al
e
f
f
icien
cy
,
p
r
iv
ac
y
,
in
ter
p
r
etab
ilit
y
,
an
d
p
r
e
d
ictiv
e
p
er
f
o
r
m
an
ce
[
9
]
,
[
1
0
]
.
T
h
e
m
o
ti
v
a
ti
o
n
b
e
h
i
n
d
t
h
is
r
esea
r
c
h
s
t
em
s
f
r
o
m
th
e
p
r
ess
i
n
g
n
e
e
d
to
d
e
v
e
lo
p
i
n
t
elli
g
en
t,
p
r
i
v
a
c
y
-
awa
r
e
s
y
s
t
em
s
t
h
at
ca
n
le
v
e
r
a
g
e
d
is
tr
ib
u
t
e
d
m
ed
ic
al
d
a
ta
f
o
r
a
cc
u
r
at
e
p
r
ed
ict
io
n
wh
ile
e
n
s
u
r
in
g
c
o
m
p
li
a
n
ce
wit
h
d
at
a
p
r
o
t
ec
ti
o
n
la
ws
a
n
d
m
i
n
i
m
iz
in
g
c
o
m
p
u
ta
ti
o
n
al
o
v
e
r
h
ea
d
.
T
h
is
is
es
p
ec
ial
ly
c
r
u
ci
al
in
e
d
g
e
-
b
ase
d
h
e
alt
h
c
ar
e
s
c
en
ar
io
s
,
w
h
e
r
e
h
o
s
p
ita
ls
,
cli
n
i
cs,
a
n
d
w
ea
r
a
b
l
e
h
e
alt
h
m
o
n
it
o
r
i
n
g
d
ev
ices
o
p
er
ate
u
n
d
er
d
i
v
er
s
e
in
f
r
ast
r
u
ct
u
r
e
c
o
n
s
tr
ai
n
ts
a
n
d
o
f
te
n
lac
k
t
h
e
b
a
n
d
wi
d
t
h
o
r
p
r
o
c
ess
i
n
g
p
o
we
r
r
e
q
u
i
r
e
d
f
o
r
t
r
ad
i
ti
o
n
al
c
en
tr
ali
ze
d
d
e
ep
le
ar
n
i
n
g
m
o
d
els
.
F
u
r
t
h
er
m
o
r
e,
th
e
im
p
o
r
ta
n
c
e
o
f
r
o
b
u
s
t
a
n
d
s
ec
u
r
e
l
ea
r
n
i
n
g
f
r
am
e
wo
r
k
s
h
as
b
ee
n
u
n
d
er
s
co
r
ed
b
y
th
e
in
c
r
ea
s
in
g
v
o
l
u
m
e
o
f
cy
b
er
t
h
r
ea
ts
a
n
d
d
at
a
b
r
e
ac
h
es
i
n
t
h
e
h
e
alt
h
c
a
r
e
s
ec
t
o
r
.
I
n
p
a
r
a
lle
l,
th
e
h
ea
lt
h
ca
r
e
i
n
d
u
s
tr
y
is
e
x
p
e
r
i
en
ci
n
g
a
g
r
o
wi
n
g
d
e
m
a
n
d
f
o
r
AI
-
d
r
iv
en
m
o
d
e
ls
t
h
at
ar
e
i
n
t
er
p
r
e
ta
b
le
,
cli
n
ic
all
y
r
ele
v
an
t,
a
n
d
c
a
p
a
b
l
e
o
f
d
el
iv
er
in
g
r
ea
l
-
t
im
e
i
n
s
i
g
h
ts
in
c
r
i
tic
al
ca
r
e
s
e
tti
n
g
s
.
E
x
is
t
in
g
s
tu
d
ies
h
av
e
p
r
o
p
o
s
e
d
v
ar
i
o
u
s
te
ch
n
i
q
u
es
i
n
cl
u
d
i
n
g
f
e
d
e
r
ate
d
a
v
e
r
a
g
i
n
g
,
s
ec
u
r
e
m
u
lti
p
a
r
t
y
c
o
m
p
u
t
ati
o
n
,
an
d
h
o
m
o
m
o
r
p
h
i
c
en
cr
y
p
ti
o
n
t
o
e
n
h
a
n
ce
p
r
i
v
ac
y
i
n
f
e
d
e
r
a
te
d
s
y
s
te
m
s
,
y
et
th
es
e
m
et
h
o
d
s
o
f
t
en
c
o
m
p
r
o
m
is
e
eit
h
er
m
o
d
el
ac
c
u
r
a
cy
o
r
s
y
s
t
em
s
c
ala
b
i
lit
y
.
A
d
d
iti
o
n
a
ll
y
,
w
h
il
e
s
e
v
er
al
r
ec
e
n
t
f
r
a
m
e
wo
r
k
s
i
n
co
r
p
o
r
ate
f
e
d
er
ate
d
le
ar
n
i
n
g
wit
h
c
o
n
v
o
l
u
ti
o
n
al
a
n
d
r
e
cu
r
r
e
n
t
n
et
wo
r
k
s
,
t
h
ey
o
f
te
n
f
ail
to
ac
c
o
m
m
o
d
at
e
r
a
p
i
d
l
y
e
v
o
lv
in
g
c
li
n
ic
al
c
o
n
d
iti
o
n
s
an
d
co
m
p
le
x
p
ati
en
t
-
le
v
el
v
a
r
iati
o
n
s
.
A
s
i
g
n
i
f
i
ca
n
t
g
a
p
p
er
s
is
ts
i
n
t
h
e
i
n
t
eg
r
a
ti
o
n
o
f
tr
a
n
s
f
o
r
m
e
r
-
b
ase
d
d
e
ep
lea
r
n
i
n
g
m
o
d
els
wit
h
f
e
d
e
r
a
te
d
s
y
s
te
m
s
t
ail
o
r
e
d
s
p
e
cif
ic
all
y
f
o
r
h
ea
lt
h
ca
r
e
a
n
al
y
ti
cs.
T
h
e
p
r
o
p
o
s
e
d
r
es
ea
r
c
h
aim
s
t
o
d
ev
el
o
p
a
p
r
iv
ac
y
-
p
r
es
er
v
i
n
g
,
t
r
a
n
s
f
o
r
m
e
r
-
en
h
an
c
ed
f
e
d
e
r
at
e
d
l
ea
r
n
i
n
g
f
r
am
e
wo
r
k
t
h
a
t
n
o
t
o
n
l
y
ad
d
r
ess
es
t
h
e
s
c
ala
b
i
lit
y
a
n
d
h
ete
r
o
g
en
eit
y
is
s
u
es
i
n
ex
is
ti
n
g
s
y
s
te
m
s
b
u
t
als
o
im
p
r
o
v
es
p
r
ed
i
cti
o
n
o
u
t
co
m
es
,
s
u
p
p
o
r
ts
c
r
o
s
s
-
d
e
v
i
ce
s
y
n
c
h
r
o
n
iz
ati
o
n
,
a
n
d
e
n
s
u
r
es
r
e
al
-
ti
m
e
a
d
a
p
t
ab
ilit
y
i
n
cli
n
i
ca
l
e
n
v
i
r
o
n
m
e
n
ts
.
2.
L
I
T
E
R
AT
U
RE
R
E
VI
E
W
T
h
e
f
ield
o
f
p
r
e
cisi
o
n
ag
r
icu
lt
u
r
e
h
as
s
ig
n
if
ican
tly
ad
v
a
n
ce
d
with
th
e
ad
o
p
tio
n
o
f
m
ac
h
i
n
e
lear
n
in
g
tech
n
iq
u
es
in
r
ec
e
n
t
y
ea
r
s
,
t
h
e
in
ter
s
ec
tio
n
o
f
h
ea
lth
ca
r
e
an
d
m
ac
h
in
e
lear
n
i
n
g
h
as
s
ee
n
tr
an
s
f
o
r
m
ativ
e
g
r
o
wth
,
p
a
r
ticu
lar
ly
with
th
e
em
er
g
en
ce
o
f
p
r
iv
ac
y
-
p
r
eser
v
in
g
tech
n
iq
u
es
f
o
r
s
en
s
itiv
e
d
ata.
Ullah
et
a
l.
[
1
1
]
in
tr
o
d
u
ce
d
a
p
r
i
v
ac
y
-
p
r
eser
v
in
g
f
e
d
er
ated
lear
n
in
g
a
p
p
r
o
ac
h
s
p
ec
if
ically
d
esig
n
ed
f
o
r
ed
g
e
h
ea
lth
ca
r
e
in
f
o
r
m
atics.
T
h
eir
m
eth
o
d
,
wh
ich
em
p
h
asizes
d
ec
en
tr
alize
d
m
o
d
el
tr
ain
in
g
ac
r
o
s
s
m
u
ltip
le
ed
g
e
n
o
d
es,
en
s
u
r
es
th
at
r
aw
p
atien
t
d
ata
r
em
ain
s
lo
ca
lized
,
th
er
eb
y
m
iti
g
atin
g
p
o
ten
tial b
r
ea
ch
es.
Ho
wev
er
,
th
e
s
y
s
tem
is
n
o
t
with
o
u
t
ch
allen
g
es.
T
h
e
r
elian
ce
o
n
ed
g
e
d
ev
ices
in
tr
o
d
u
ce
s
co
m
p
u
tatio
n
al
lim
itatio
n
s
,
an
d
in
co
n
s
is
ten
t
d
ata
d
is
tr
ib
u
tio
n
s
ac
r
o
s
s
n
o
d
es
ca
n
im
p
ac
t
c
o
n
v
e
r
g
en
ce
.
Als
o
,
s
y
n
ch
r
o
n
izatio
n
d
elay
s
m
ay
o
cc
u
r
wh
e
n
m
u
ltip
le
clien
ts
p
ar
ticip
ate
in
tr
ain
in
g
.
B
ar
ati
et
a
l.
[
1
2
]
p
r
o
p
o
s
ed
a
n
o
v
el
cl
o
u
d
au
d
itin
g
ap
p
r
o
ac
h
aim
ed
at
en
s
u
r
in
g
GDPR
co
m
p
lian
ce
with
in
o
n
lin
e
h
ea
lth
ca
r
e
p
latf
o
r
m
s
.
T
h
ei
r
p
r
iv
ac
y
-
awa
r
e
a
u
d
itin
g
m
ec
h
an
is
m
en
ab
les
co
n
tin
u
o
u
s
m
o
n
ito
r
i
n
g
o
f
clo
u
d
-
b
ased
s
er
v
ices
to
v
alid
ate
ad
h
er
en
ce
to
r
e
g
u
lato
r
y
f
r
am
ewo
r
k
s
.
Utilizin
g
an
in
tellig
en
t a
u
d
it e
n
g
in
e
an
d
s
ec
u
r
e
lo
g
m
an
a
g
e
m
en
t sy
s
tem
,
th
e
m
eth
o
d
en
h
a
n
ce
s
tr
an
s
p
ar
en
cy
in
h
o
w
p
atien
t
d
ata
is
ac
ce
s
s
ed
an
d
p
r
o
ce
s
s
ed
.
A
s
ig
n
if
ican
t
ad
v
an
tag
e
o
f
th
is
ap
p
r
o
ac
h
is
its
ab
ilit
y
to
b
r
i
d
g
e
th
e
tr
u
s
t
g
ap
b
etwe
en
h
ea
lth
ca
r
e
p
r
o
v
id
er
s
an
d
r
eg
u
lato
r
y
b
o
d
ies.
I
t
also
s
u
p
p
o
r
ts
au
to
m
ate
d
aler
ts
in
ca
s
es
o
f
p
o
licy
v
io
latio
n
s
,
en
ab
lin
g
tim
ely
in
ter
v
en
tio
n
.
I
n
a
r
elate
d
v
ein
,
Z
h
o
u
et
a
l.
[
1
3
]
d
ev
elo
p
ed
a
f
ed
e
r
ated
lear
n
in
g
f
r
am
ewo
r
k
f
o
r
e
d
g
e
co
m
p
u
tin
g
t
h
at
in
tr
o
d
u
ce
s
en
h
a
n
ce
m
en
ts
to
p
r
iv
ac
y
,
s
ca
lab
ilit
y
,
an
d
r
o
b
u
s
tn
ess
in
h
ea
lth
ca
r
e
ap
p
licatio
n
s
.
T
h
eir
m
o
d
el
u
tili
ze
s
en
cr
y
p
ted
g
r
ad
ien
t
d
e
s
ce
n
t,
s
ec
u
r
e
m
u
ltip
ar
ty
co
m
p
u
tatio
n
,
an
d
d
if
f
er
en
tial
p
r
iv
ac
y
to
m
ain
tain
th
e
co
n
f
id
en
tiality
o
f
p
atien
t
r
ec
o
r
d
s
d
u
r
in
g
m
o
d
el
tr
ain
in
g
.
B
y
lev
er
ag
in
g
h
ier
a
r
ch
ical
ag
g
r
e
g
atio
n
an
d
ad
a
p
tiv
e
lear
n
in
g
r
ates,
th
eir
f
r
am
ewo
r
k
is
ab
le
to
o
v
e
r
co
m
e
co
m
m
o
n
is
s
u
es
s
u
ch
as
s
lo
w
co
n
v
er
g
e
n
ce
an
d
d
ata
im
b
alan
ce
.
On
e
o
f
th
e
m
ain
a
d
v
an
tag
es
is
its
m
o
d
u
lar
ar
ch
i
tectu
r
e,
allo
win
g
it
to
b
e
a
d
ap
ted
ac
r
o
s
s
v
ar
io
u
s
clin
ical
s
ettin
g
s
with
ea
s
e.
Ho
wev
er
,
en
c
r
y
p
tio
n
an
d
p
r
iv
ac
y
-
p
r
eser
v
in
g
m
ec
h
a
n
is
m
s
ca
n
s
o
m
etim
es
r
esu
lt
in
s
lo
wer
tr
ain
in
g
tim
es
an
d
in
c
r
ea
s
ed
s
y
s
tem
co
m
p
lex
ity
.
T
h
e
tr
ad
e
-
o
f
f
b
etwe
en
p
r
iv
ac
y
an
d
c
o
m
p
u
tatio
n
al
ef
f
icien
cy
r
e
m
ain
s
a
k
ey
co
n
s
i
d
er
atio
n
in
its
im
p
lem
en
tatio
n
.
W
an
g
et
a
l.
[
1
4
]
ex
p
l
o
r
ed
a
s
im
ilar
d
ir
ec
tio
n
b
y
ap
p
ly
in
g
p
r
iv
ac
y
-
p
r
eser
v
in
g
f
ed
er
ated
lear
n
in
g
to
th
e
in
ter
n
et
o
f
m
ed
ical
th
in
g
s
(
I
o
MT
)
u
n
d
e
r
ed
g
e
co
m
p
u
tin
g
en
v
ir
o
n
m
en
ts
.
T
h
eir
m
eth
o
d
in
teg
r
ates
b
lo
c
k
c
h
ain
f
o
r
tam
p
e
r
-
p
r
o
o
f
m
o
d
e
l
v
er
if
icatio
n
an
d
em
p
lo
y
s
s
ec
u
r
e
h
ash
-
b
ased
co
m
m
u
n
icatio
n
p
r
o
to
co
ls
b
etwe
en
d
ev
ices.
A
n
o
tab
le
s
tr
en
g
th
o
f
th
is
f
r
am
ewo
r
k
is
it
s
f
o
cu
s
o
n
h
eter
o
g
en
e
o
u
s
d
ev
ice
co
m
m
u
n
icatio
n
in
an
I
o
MT
s
ettin
g
,
wh
ich
r
ef
lects
r
ea
l
-
wo
r
ld
s
ce
n
ar
io
s
.
I
t
en
h
an
ce
s
b
o
th
d
ata
in
teg
r
it
y
an
d
in
ter
o
p
e
r
ab
ilit
y
am
o
n
g
d
ev
ices.
Ho
wev
er
,
th
e
u
s
e
o
f
b
lo
ck
ch
ain
,
wh
ile
b
en
ef
icial
f
o
r
s
ec
u
r
ity
,
ad
d
s
s
u
b
s
tan
tial la
ten
cy
an
d
s
to
r
a
g
e
d
em
an
d
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
7
2
2
-
2
5
8
6
I
AE
S I
n
t J Ro
b
&
Au
to
m
,
Vo
l
.
1
4
,
No
.
4
,
Dec
em
b
er
2
0
2
5
:
5
7
0
-
5
8
0
572
Ad
d
itio
n
ally
,
en
e
r
g
y
co
n
s
u
m
p
tio
n
ass
o
ciate
d
with
co
n
t
in
u
o
u
s
b
l
o
ck
ch
ain
o
p
e
r
atio
n
s
ca
n
p
o
s
e
s
u
s
tain
ab
ilit
y
co
n
ce
r
n
s
in
lo
w
-
p
o
wer
ed
g
e
d
e
v
ices.
Me
an
wh
ile,
Pan
et
a
l.
[
1
5
]
p
r
o
p
o
s
ed
an
a
d
ap
tiv
e
f
ed
er
ated
lear
n
i
n
g
f
r
am
ewo
r
k
f
o
r
clin
ical
r
is
k
p
r
ed
ictio
n
th
at
lev
er
ag
es
elec
tr
o
n
ic
h
e
alth
r
ec
o
r
d
s
f
r
o
m
m
u
ltip
le
h
o
s
p
itals
.
T
h
eir
m
o
d
el
s
tan
d
s
o
u
t
b
y
in
tr
o
d
u
cin
g
d
y
n
am
ic
clien
t
s
elec
tio
n
an
d
m
o
d
el
p
er
s
o
n
aliza
tio
n
tech
n
iq
u
es.
T
h
ese
s
tr
ateg
ies
a
llo
w
th
e
s
y
s
tem
to
ad
ap
t
to
v
ar
y
in
g
lo
ca
l
d
ata
ch
a
r
ac
ter
is
tics
,
im
p
r
o
v
in
g
th
e
m
o
d
el’
s
g
en
er
aliza
tio
n
a
n
d
r
e
lev
an
ce
.
Mo
r
eo
v
er
,
th
e
f
r
am
e
wo
r
k
ad
d
r
ess
es
d
ata
d
r
if
t
an
d
h
eter
o
g
en
eity
b
y
u
p
d
atin
g
m
o
d
el
p
ar
a
m
eter
s
in
a
co
n
tex
t
-
awa
r
e
m
a
n
n
er
.
On
e
k
ey
b
en
e
f
it
o
f
th
is
m
eth
o
d
is
its
ab
ilit
y
to
en
h
an
ce
p
r
ed
ictio
n
ac
c
u
r
ac
y
with
o
u
t
co
m
p
r
o
m
is
in
g
p
atien
t
p
r
iv
ac
y
.
H
o
wev
er
,
t
h
e
co
m
p
lex
ity
o
f
co
o
r
d
in
atin
g
m
u
ltip
le
clien
ts
an
d
e
n
s
u
r
in
g
co
n
s
is
ten
t
m
o
d
e
l
u
p
d
ates
ca
n
lead
t
o
s
y
n
ch
r
o
n
izatio
n
o
v
er
h
ea
d
.
Mo
r
eo
v
er
,
d
i
f
f
er
en
ce
s
i
n
el
ec
tr
o
n
ic
h
e
alth
r
ec
o
r
d
(
E
H
R
)
s
tr
u
ctu
r
es
ac
r
o
s
s
in
s
titu
tio
n
s
m
ig
h
t
r
eq
u
ir
e
ad
d
itio
n
al
p
r
e
p
r
o
ce
s
s
in
g
a
n
d
s
tan
d
ar
d
izatio
n
e
f
f
o
r
ts
.
B
ad
awy
et
a
l.
[
1
6
]
p
r
esen
te
d
an
e
x
ten
s
iv
e
s
u
r
v
ey
c
o
v
er
i
n
g
m
ac
h
in
e
lear
n
in
g
a
n
d
d
ee
p
lear
n
in
g
tech
n
iq
u
es
f
o
r
h
ea
lth
ca
r
e
p
r
e
d
ictiv
e
an
aly
tics
.
B
ad
awy
an
d
co
lleag
u
es
em
p
h
asized
th
e
n
ee
d
f
o
r
p
r
iv
ac
y
-
p
r
eser
v
in
g
ap
p
r
o
ac
h
es,
esp
ec
ially
wh
en
wo
r
k
in
g
with
r
ea
l
-
tim
e
p
atien
t
d
ata,
th
er
eb
y
ec
h
o
in
g
th
e
d
ir
ec
tio
n
tak
en
b
y
o
th
er
r
esear
ch
er
s
in
th
is
d
o
m
ain
.
L
at
h
a
et
a
l.
[
1
7
]
in
v
esti
g
ated
th
e
u
s
e
o
f
s
tack
i
n
g
alg
o
r
ith
m
s
o
v
e
r
b
ig
d
ata
in
h
ea
lth
ca
r
e
c
o
m
m
u
n
ities
f
o
r
d
is
ea
s
e
p
r
ed
ictio
n
.
T
h
eir
wo
r
k
in
v
o
lv
es
c
o
m
b
in
i
n
g
m
u
ltip
le
m
ac
h
i
n
e
lear
n
in
g
m
o
d
els
s
u
ch
as
r
an
d
o
m
f
o
r
ests
,
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
es,
an
d
g
r
ad
ien
t
b
o
o
s
tin
g
class
if
ier
s
in
a
lay
er
ed
en
s
em
b
le
f
o
r
m
at.
T
h
e
s
tack
in
g
ap
p
r
o
a
ch
im
p
r
o
v
e
s
class
if
icatio
n
ac
cu
r
ac
y
b
y
lear
n
in
g
f
r
o
m
th
e
p
r
ed
ictio
n
er
r
o
r
s
o
f
b
ase
m
o
d
els.
On
e
clea
r
ad
v
a
n
tag
e
o
f
th
eir
m
et
h
o
d
is
its
r
esil
ien
ce
ag
ain
s
t
n
o
is
y
o
r
in
co
m
p
lete
d
ata,
wh
ich
is
co
m
m
o
n
in
h
ea
lth
ca
r
e
d
atab
ases
.
Mo
r
eo
v
er
,
en
s
em
b
le
m
o
d
els
g
en
er
ally
r
e
d
u
ce
th
e
r
is
k
o
f
m
o
d
el
b
ias
an
d
v
ar
ia
n
ce
.
Ho
wev
er
,
th
e
co
m
p
u
tati
o
n
al
co
m
p
lex
ity
a
n
d
tim
e
r
e
q
u
ir
ed
f
o
r
tr
ain
in
g
m
u
ltip
le
b
ase
lear
n
er
s
ca
n
b
e
s
u
b
s
tan
tial.
Ad
d
itio
n
ally
,
i
n
ter
p
r
etab
ilit
y
b
ec
o
m
es
m
o
r
e
ch
allen
g
in
g
as
th
e
en
s
em
b
le
m
o
d
el
g
r
o
ws in
d
e
p
t
h
an
d
s
tr
u
ct
u
r
e.
Z
h
ao
et
a
l.
[
1
8
]
c
o
n
tr
ib
u
ted
a
p
r
iv
ac
y
-
p
r
eser
v
in
g
f
e
d
er
ated
l
ea
r
n
in
g
f
r
am
ew
o
r
k
d
esig
n
ed
s
p
ec
if
ically
f
o
r
p
r
o
g
n
o
s
is
p
r
ed
ictio
n
u
s
in
g
m
u
lti
-
s
o
u
r
ce
E
HR
s
.
T
h
eir
a
p
p
r
o
ac
h
em
p
l
o
y
s
atten
tio
n
-
b
a
s
ed
m
ec
h
an
is
m
s
to
id
en
tify
an
d
weig
h
t
h
e
m
o
s
t
r
elev
an
t
f
ea
tu
r
es
ac
r
o
s
s
d
at
a
s
o
u
r
ce
s
.
T
h
is
atten
tio
n
-
e
n
h
an
ce
d
ar
ch
itectu
r
e
en
ab
les
m
o
r
e
ac
cu
r
ate
p
r
e
d
ictio
n
s
b
y
f
o
c
u
s
in
g
o
n
clin
ically
s
ig
n
if
ican
t
v
ar
iab
les
w
h
ile
m
a
in
tain
in
g
lo
ca
l
d
ata
p
r
iv
ac
y
.
M
o
d
el
[
1
9
]
s
u
p
p
o
r
ts
s
ec
u
r
e
co
m
m
u
n
icatio
n
th
r
o
u
g
h
h
o
m
o
m
o
r
p
h
ic
en
c
r
y
p
t
io
n
,
e
n
s
u
r
in
g
th
at
in
d
iv
id
u
al
h
ea
lth
r
ec
o
r
d
s
ar
e
n
ev
er
ex
p
o
s
ed
.
T
h
e
a
d
v
an
ta
g
e
s
o
f
t
h
is
m
eth
o
d
i
n
clu
d
e
its
f
l
ex
ib
ilit
y
in
ad
ap
tin
g
to
d
if
f
er
e
n
t
E
HR
f
o
r
m
ats
an
d
its
r
o
b
u
s
tn
ess
in
h
an
d
lin
g
d
at
a
s
p
ar
s
ity
.
Fin
ally
,
Ab
id
i
et
a
l.
[
2
0
]
ex
p
l
o
r
ed
t
h
e
in
teg
r
atio
n
o
f
cr
o
s
s
o
v
er
-
b
ase
d
m
u
ltil
ay
er
p
e
r
ce
p
tr
o
n
s
(
M
L
Ps
)
in
s
m
ar
t
h
ea
lth
ca
r
e
s
y
s
tem
s
f
o
r
d
is
ea
s
e
d
etec
tio
n
.
T
h
eir
wo
r
k
f
o
cu
s
e
s
o
n
c
o
m
b
in
in
g
g
en
etic
alg
o
r
ith
m
s
with
n
e
u
r
al
n
etwo
r
k
s
to
im
p
r
o
v
e
f
ea
tu
r
e
s
elec
tio
n
an
d
class
if
icatio
n
ac
cu
r
ac
y
.
B
y
u
s
in
g
cr
o
s
s
o
v
er
o
p
er
atio
n
s
in
th
e
m
o
d
el’
s
ar
ch
i
tectu
r
e,
th
ey
en
s
u
r
e
b
etter
g
en
er
aliza
tio
n
a
n
d
r
o
b
u
s
tn
ess
to
o
v
er
f
itti
n
g
.
T
h
e
s
y
s
t
em
d
em
o
n
s
tr
ates
h
ig
h
p
r
ec
is
io
n
in
d
is
tin
g
u
is
h
in
g
b
etwe
en
s
im
ilar
d
is
ea
s
e
class
e
s
,
m
ak
in
g
it su
itab
le
f
o
r
co
m
p
l
ex
d
iag
n
o
s
tic
task
s
.
Ad
d
itio
n
ally
,
Gu
p
ta
[
2
1
]
p
r
o
v
id
ed
a
co
m
p
ar
ativ
e
s
tu
d
y
o
n
s
u
p
er
v
is
ed
m
ac
h
in
e
lear
n
in
g
al
g
o
r
ith
m
s
,
o
f
f
er
in
g
in
s
ig
h
ts
in
to
m
o
d
el
s
elec
tio
n
an
d
p
er
f
o
r
m
a
n
ce
ev
a
lu
atio
n
,
wh
ich
is
f
o
u
n
d
atio
n
al
f
o
r
h
ea
lth
ca
r
e
M
L
ap
p
licatio
n
s
.
B
atis
ta
an
d
E
v
s
u
k
o
f
f
[
2
2
]
co
n
d
u
cted
a
s
y
s
tem
atic
r
ev
iew
o
n
th
e
u
s
e
o
f
tr
an
s
f
o
r
m
er
-
b
ased
m
eth
o
d
s
in
an
aly
zin
g
elec
tr
o
n
ic
m
ed
ical
r
ec
o
r
d
s
,
h
ig
h
lig
h
tin
g
th
eir
s
u
p
er
io
r
ity
i
n
h
an
d
lin
g
s
eq
u
en
tial
an
d
co
n
tex
tu
al
h
ea
lth
ca
r
e
d
ata.
K
u
m
ar
et
a
l.
[
2
3
]
d
is
cu
s
s
ed
th
e
co
m
b
i
n
ed
a
p
p
licatio
n
s
o
f
m
a
ch
in
e
lear
n
i
n
g
a
n
d
n
atu
r
al
lan
g
u
ag
e
p
r
o
ce
s
s
in
g
(
NL
P)
in
h
ea
lth
ca
r
e,
d
em
o
n
s
tr
atin
g
h
o
w
d
ata
-
d
r
iv
en
tech
n
i
q
u
es
ca
n
s
tr
ea
m
lin
e
clin
ical
wo
r
k
f
lo
ws.
B
er
r
o
s
et
a
l.
[
2
4
]
e
m
p
h
asized
t
h
e
r
o
le
o
f
b
i
g
d
ata
an
aly
tics
in
e
n
h
an
cin
g
d
ig
ital
h
ea
lth
s
er
v
ices,
s
h
o
wca
s
in
g
im
p
r
o
v
e
d
d
ec
is
io
n
-
m
ak
in
g
th
r
o
u
g
h
r
e
al
-
tim
e
d
ata
p
r
o
ce
s
s
in
g
.
Z
h
o
u
[
2
5
]
ex
p
lo
r
e
d
th
e
d
ev
elo
p
m
e
n
t
o
f
NL
P
to
o
ls
to
ex
tr
ac
t
in
teg
r
ativ
e
h
ea
lth
in
f
o
r
m
atio
n
f
r
o
m
E
HR
s
,
wh
ich
is
cr
u
cial
f
o
r
ex
p
an
d
i
n
g
th
e
s
co
p
e
o
f
p
er
s
o
n
alize
d
an
d
p
r
ev
en
tiv
e
ca
r
e.
3.
P
RO
P
O
SE
D
WO
RK
I
n
r
ec
e
n
t
y
ea
r
s
,
t
h
e
h
ea
lth
ca
r
e
s
ec
to
r
h
as
ex
p
er
ien
ce
d
an
ex
p
o
n
en
tial
g
r
o
wth
in
d
ata,
esp
e
cially
f
r
o
m
elec
tr
o
n
ic
h
ea
lth
r
ec
o
r
d
s
(
E
HR
s
)
,
m
ed
ical
im
ag
in
g
,
s
en
s
o
r
-
b
ased
m
o
n
ito
r
in
g
d
e
v
ices,
an
d
wea
r
ab
le
tech
n
o
lo
g
ies.
T
h
e
in
te
g
r
atio
n
o
f
th
is
d
ata
is
ess
en
tial
f
o
r
b
u
ild
in
g
p
r
ed
ictiv
e
m
o
d
els
th
at
c
an
ass
is
t
in
clin
ical
d
ec
is
io
n
-
m
ak
in
g
,
d
is
ea
s
e
d
etec
tio
n
,
an
d
p
er
s
o
n
alize
d
tr
ea
t
m
en
t
p
lan
s
.
Ho
wev
er
,
th
e
ce
n
tr
aliza
tio
n
o
f
s
u
ch
s
en
s
itiv
e
h
ea
lth
d
ata
r
aises
s
i
g
n
if
ican
t
p
r
i
v
ac
y
co
n
ce
r
n
s
an
d
r
eg
u
lato
r
y
ch
allen
g
es.
I
t
h
a
s
b
ee
n
estab
lis
h
ed
th
at
p
r
iv
ac
y
v
io
latio
n
s
,
s
u
c
h
as
d
ata
b
r
ea
ch
es
an
d
u
n
au
th
o
r
ized
ac
ce
s
s
,
ca
n
s
ev
er
ely
u
n
d
er
m
in
e
th
e
tr
u
s
t
in
h
ea
lth
ca
r
e
s
y
s
tem
s
.
Mo
r
eo
v
er
,
d
ata
p
r
iv
ac
y
r
eg
u
latio
n
s
lik
e
th
e
GDP
R
an
d
h
ea
lth
in
s
u
r
an
ce
p
o
r
tab
ilit
y
an
d
ac
co
u
n
tab
ilit
y
ac
t
(
HI
PAA)
i
m
p
o
s
e
s
tr
ict
co
n
s
tr
ain
ts
o
n
h
o
w
p
atien
t
d
ata
ca
n
b
e
s
h
ar
e
d
an
d
ac
c
ess
ed
.
T
h
is
h
as
led
to
a
cr
itical
n
ee
d
f
o
r
f
r
am
ewo
r
k
s
t
h
at
en
s
u
r
e
p
r
iv
a
cy
wh
ile
s
till
en
ab
lin
g
co
llab
o
r
ativ
e
d
ata
-
d
r
iv
e
n
lear
n
in
g
.
T
h
e
p
r
o
p
o
s
ed
f
r
am
e
wo
r
k
s
ee
k
s
to
o
v
er
co
m
e
th
es
e
ch
allen
g
es
b
y
co
m
b
i
n
in
g
f
e
d
er
ated
lear
n
in
g
to
en
s
u
r
e
d
ata
p
r
iv
ac
y
with
th
e
tr
an
s
f
o
r
m
er
a
r
ch
itectu
r
e
t
o
en
h
a
n
ce
m
o
d
el
p
er
f
o
r
m
a
n
ce
,
p
a
r
ticu
lar
ly
in
h
an
d
lin
g
Evaluation Warning : The document was created with Spire.PDF for Python.
I
AE
S
I
n
t
J
R
o
b
&
A
u
to
m
I
SS
N:
2722
-
2
5
8
6
E
n
h
a
n
cin
g
h
ea
lth
s
ta
tu
s
p
r
ed
ictio
n
a
n
d
d
a
ta
s
ec
u
r
ity
…
(
S
u
b
r
a
ma
n
iya
n
S
en
th
a
ma
r
a
i
)
573
co
m
p
lex
,
m
u
ltimo
d
al
d
atasets
.
T
h
is
co
m
b
in
atio
n
is
ex
p
ec
ted
to
im
p
r
o
v
e
th
e
ac
c
u
r
ac
y
an
d
ef
f
icien
c
y
o
f
p
r
ed
ictiv
e
m
o
d
els
in
h
ea
lth
ca
r
e
ap
p
licatio
n
s
with
o
u
t
co
m
p
r
o
m
is
in
g
p
atien
t
co
n
f
id
en
tiali
ty
.
Fig
u
r
e
1
s
h
o
ws
th
e
o
v
er
all
ar
c
h
itectu
r
e
o
f
th
e
p
r
o
p
o
s
ed
wo
r
k
.
Fig
u
r
e
1
.
Ov
e
r
v
iew
o
f
th
e
p
r
o
p
o
s
ed
m
o
d
el
3
.
1
.
Da
t
a
s
et
det
a
ils
T
h
e
PTFL
-
HPA
f
r
am
ewo
r
k
h
as
b
ee
n
im
p
lem
en
ted
u
s
in
g
p
u
b
licly
av
ailab
le
h
ea
lth
ca
r
e
d
at
asets
,
s
u
ch
as
th
e
Me
d
ical
I
n
f
o
r
m
atio
n
M
ar
t
f
o
r
I
n
ten
s
iv
e
C
ar
e
(
MI
MI
C
-
I
I
I
)
.
h
is
d
ataset
co
m
p
r
is
es
d
e
-
id
en
tifie
d
h
ea
lth
-
r
elate
d
d
ata
o
f
o
v
er
6
0
,
0
0
0
I
C
U
ad
m
is
s
io
n
s
at
th
e
B
et
h
I
s
r
ae
l
Dea
co
n
ess
Me
d
ical
C
en
ter
in
B
o
s
to
n
,
Ma
s
s
ac
h
u
s
etts
,
co
llected
b
etw
ee
n
2
0
0
1
an
d
2
0
1
2
.
T
h
e
PTFL
-
HPA
f
r
am
ewo
r
k
is
d
esig
n
ed
to
en
ab
le
d
ec
en
tr
a
lized
an
d
p
r
iv
ac
y
-
p
r
eser
v
i
n
g
h
ea
lth
ca
r
e
an
aly
tics
ac
r
o
s
s
m
u
ltip
le
in
s
titu
tio
n
s
.
T
h
is
ap
p
r
o
a
ch
en
s
u
r
es
th
at
s
en
s
itiv
e
d
ata,
s
u
c
h
as
p
atien
t
h
ea
lth
r
ec
o
r
d
s
,
la
b
r
esu
lts
,
a
n
d
clin
ical
n
o
tes,
r
e
m
ain
s
with
in
t
h
e
lo
ca
l
in
s
titu
tio
n
,
an
d
o
n
ly
m
o
d
el
u
p
d
ates
ar
e
s
h
ar
e
d
with
a
ce
n
tr
al
f
e
d
er
ated
s
er
v
er
.
B
y
u
s
in
g
tr
a
n
s
f
o
r
m
e
r
m
o
d
els,
th
e
p
r
o
p
o
s
ed
f
r
am
ewo
r
k
aim
s
to
ex
tr
ac
t
m
ea
n
in
g
f
u
l
in
s
ig
h
ts
f
r
o
m
b
o
t
h
tim
e
-
s
er
ies
d
ata
(
e.
g
.
,
p
atie
n
t
v
itals
,
lab
r
esu
lts
)
an
d
u
n
s
tr
u
ctu
r
ed
tex
t
d
ata
(
e.
g
.
,
clin
ical
n
o
tes),
wh
ich
a
r
e
co
m
m
o
n
in
h
ea
lth
ca
r
e
s
ettin
g
s
.
T
h
e
tr
a
n
s
f
o
r
m
er
’
s
s
elf
-
atten
tio
n
m
ec
h
a
n
is
m
allo
ws
th
e
m
o
d
el
to
ca
p
tu
r
e
l
o
n
g
-
ter
m
d
e
p
en
d
e
n
cies
in
s
e
q
u
en
tial
d
ata
an
d
lear
n
co
m
p
lex
r
elatio
n
s
h
ip
s
i
n
m
u
ltimo
d
al
in
p
u
ts
.
Ad
d
itio
n
ally
,
th
e
f
r
am
ew
o
r
k
in
co
r
p
o
r
ates
p
r
iv
ac
y
-
p
r
eser
v
in
g
tech
n
iq
u
es
s
u
ch
as
h
o
m
o
m
o
r
p
h
ic
en
cr
y
p
tio
n
an
d
s
ec
u
r
e
a
g
g
r
e
g
atio
n
to
e
n
s
u
r
e
th
at
m
o
d
el
t
r
ain
in
g
a
n
d
u
p
d
ates
r
em
ai
n
p
r
i
v
a
te.
T
h
e
s
y
s
tem
will
allo
w
m
u
ltip
le
h
ea
lth
ca
r
e
in
s
titu
tio
n
s
to
co
llab
o
r
ate
o
n
p
r
e
d
ictiv
e
an
aly
tics
task
s
s
u
ch
as
d
is
ea
s
e
p
r
ed
ictio
n
,
p
atien
t
r
is
k
ass
es
s
m
en
t,
an
d
clin
ical
d
ec
is
io
n
-
m
ak
in
g
with
o
u
t
ex
p
o
s
in
g
s
en
s
itiv
e
p
atien
t
d
ata.
T
h
is
co
llab
o
r
ativ
e
m
o
d
el
tr
ain
i
n
g
,
with
o
u
t
d
ir
ec
t
d
ata
s
h
a
r
in
g
,
a
lig
n
s
with
th
e
p
r
iv
ac
y
r
eq
u
ir
e
m
en
ts
s
et
f
o
r
th
b
y
d
ata
p
r
o
tectio
n
r
eg
u
latio
n
s
lik
e
GDPR
an
d
HI
PAA,
en
s
u
r
in
g
co
m
p
lian
ce
with
p
r
iv
ac
y
s
tan
d
ar
d
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
7
2
2
-
2
5
8
6
I
AE
S I
n
t J Ro
b
&
Au
to
m
,
Vo
l
.
1
4
,
No
.
4
,
Dec
em
b
er
2
0
2
5
:
5
7
0
-
5
8
0
574
3
.
1
.
1
.
L
o
ca
l
da
t
a
re
po
s
it
o
ries
a
nd
inp
ut
f
ea
t
ures
Hea
lth
ca
r
e
d
ata
at
ea
ch
p
ar
ticip
atin
g
in
s
titu
tio
n
h
av
e
b
ee
n
r
etain
ed
in
s
ec
u
r
e
lo
ca
l
en
v
i
r
o
n
m
en
ts
.
T
h
e
in
p
u
t
d
ata
co
n
s
is
t o
f
th
r
ee
m
ain
ca
teg
o
r
ies:
−
Stru
ctu
r
ed
n
u
m
er
ical
d
ata:
Vit
al
s
ig
n
s
,
lab
m
ea
s
u
r
em
en
ts
−
C
ateg
o
r
ical
an
d
tex
tu
al
d
ata:
C
lin
ical
n
o
tes,
d
iag
n
o
s
es,
m
ed
icatio
n
s
−
T
im
e
-
s
er
ies d
ata:
L
o
n
g
itu
d
i
n
a
l E
HR
s
eq
u
en
ce
s
L
et
d
en
o
te
th
e
d
ataset
s
to
r
ed
lo
ca
lly
at
s
ite
i,
s
u
ch
th
at
=
{
(
1
,
1
)
,
(
2
,
2
)
,
…
,
(
,
)
}
,
wh
er
e
is
a
f
ea
tu
r
e
v
ec
to
r
an
d
is
th
e
co
r
r
esp
o
n
d
in
g
la
b
el
o
r
tar
g
et
v
a
r
iab
le.
3
.
1
.
2
.
Da
t
a
prepro
ce
s
s
ing
a
n
d f
ea
t
ure
t
ra
ns
f
o
rma
t
io
n
T
o
s
tan
d
ar
d
ize
n
u
m
er
ical
d
ata,
m
in
-
m
ax
s
ca
lin
g
o
r
z
-
s
co
r
e
n
o
r
m
aliza
tio
n
h
as b
ee
n
ap
p
lied
.
−
Min
-
m
ax
n
o
r
m
aliza
tio
n
:
′
=
−
−
−
Z
-
s
co
r
e
n
o
r
m
aliza
tio
n
:
′
=
−
Fo
r
tex
tu
al
d
ata,
em
b
e
d
d
in
g
s
∈
×
h
av
e
b
ee
n
g
en
e
r
ated
u
s
in
g
p
r
etr
ain
e
d
tr
an
s
f
o
r
m
er
s
lik
e
B
E
R
T
,
wh
er
e
n
is
th
e
n
u
m
b
er
o
f
to
k
e
n
s
an
d
d
is
th
e
em
b
ed
d
i
n
g
d
im
e
n
s
io
n
.
−
Po
s
itio
n
al
en
co
d
in
g
PE
h
as b
e
en
ad
d
e
d
to
r
etain
s
eq
u
en
tial i
n
f
o
r
m
atio
n
:
,
2
=
s
in
(
10000
2
⁄
)
,
,
2
+
1
=
c
os
(
10000
2
⁄
)
3
.
1
.
3
.
T
ra
ns
f
o
rm
er
-
ba
s
ed
lo
ca
l le
a
rning
m
o
del
E
ac
h
lo
ca
l
n
o
d
e
h
as
u
s
ed
a
t
r
an
s
f
o
r
m
er
en
co
d
er
f
o
r
m
o
d
e
lin
g
th
e
h
ea
lth
ca
r
e
s
eq
u
e
n
ce
s
.
T
h
e
s
elf
-
atten
tio
n
m
ec
h
an
is
m
is
d
ef
i
n
e
d
as:
−
Scaled
Do
t
-
p
r
o
d
u
ct
atten
tio
n
(
,
,
)
=
(
√
)
wh
er
e
Q,
K,
an
d
V
ar
e
th
e
q
u
er
y
,
k
ey
,
an
d
v
alu
e
m
atr
ices
r
esp
ec
tiv
ely
,
an
d
is
th
e
d
im
en
s
io
n
ality
o
f
th
e
k
ey
s
.
−
Mu
lti
-
h
ea
d
atten
tio
n
(
,
,
)
=
(
ℎ
1
,
…
,
ℎ
ℎ
)
ℎ
ℎ
=
(
,
,
)
T
h
ese
lay
er
s
h
av
e
b
ee
n
f
o
llo
wed
b
y
la
y
er
n
o
r
m
aliza
tio
n
,
d
r
o
p
o
u
t,
a
n
d
f
ee
d
f
o
r
war
d
lay
er
s
:
(
)
=
(
0
,
1
+
1
)
2
+
2
3
.
1
.
4
.
F
eder
a
t
ed
lea
rning
co
m
po
nent
L
o
ca
l
m
o
d
els
h
av
e
b
ee
n
tr
ain
ed
in
d
ep
en
d
en
tly
at
ea
ch
n
o
d
e
u
s
in
g
lo
ca
l
d
atasets
.
T
h
e
m
o
d
el
u
p
d
ates
f
r
o
m
n
o
d
e
iii
h
av
e
b
ee
n
en
cr
y
p
ted
an
d
s
h
ar
ed
with
th
e
ce
n
tr
al
ag
g
r
eg
ato
r
.
−
Fed
er
ated
a
v
er
a
g
in
g
(
Fed
A
v
g
)
:
=
∑
=
1
wh
er
e:
N
is
th
e
n
u
m
b
er
o
f
p
ar
t
icip
atin
g
n
o
d
es
;
is
th
e
n
u
m
b
e
r
o
f
s
am
p
les at
n
o
d
e
i
;
=
∑
=
1
.
E
n
cr
y
p
te
d
u
p
d
ates
(
)
h
av
e
b
ee
n
ag
g
r
eg
ated
u
s
in
g
h
o
m
o
m
o
r
p
h
i
c
en
cr
y
p
tio
n
o
p
er
atio
n
s
,
wh
e
r
e:
(
)
=
∑
(
)
T
h
is
allo
ws co
m
p
u
tatio
n
with
o
u
t d
ec
r
y
p
tin
g
t
h
e
d
ata.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
AE
S
I
n
t
J
R
o
b
&
A
u
to
m
I
SS
N:
2722
-
2
5
8
6
E
n
h
a
n
cin
g
h
ea
lth
s
ta
tu
s
p
r
ed
ictio
n
a
n
d
d
a
ta
s
ec
u
r
ity
…
(
S
u
b
r
a
ma
n
iya
n
S
en
th
a
ma
r
a
i
)
575
3
.
1
.
5
.
P
riv
a
c
y
-
preserv
ing
m
e
cha
nis
m
s
Dif
f
er
en
tial
p
r
iv
ac
y
:
T
o
e
n
s
u
r
e
d
if
f
er
e
n
tial
p
r
iv
ac
y
,
ca
lib
r
ated
n
o
is
e
(
0
,
2
)
h
as
b
ee
n
a
d
d
ed
t
o
g
r
ad
ien
ts
:
̇
=
+
(
0
,
2
)
Priv
ac
y
B
u
d
g
et:
T
h
e
p
r
iv
ac
y
g
u
ar
an
tee
(
ϵ
,
δ)
is
co
n
tr
o
lled
b
y
s
elec
tin
g
ap
p
r
o
p
r
iate
n
o
is
e
s
ca
le
σ
,
b
atch
s
ize,
an
d
n
u
m
b
er
o
f
iter
atio
n
s
.
T
h
e
to
tal
p
r
iv
ac
y
lo
s
s
ϵ
o
v
e
r
T
iter
atio
n
s
h
as
b
ee
n
b
o
u
n
d
e
d
u
s
in
g
th
e
m
o
m
e
n
ts
ac
co
u
n
tan
t te
ch
n
iq
u
e.
3
.
1
.
6
.
M
ultim
o
da
l
f
ea
t
ure
f
u
s
io
n la
y
er
R
ep
r
esen
tatio
n
s
f
r
o
m
s
tr
u
ctu
r
ed
d
ata
ℎ
,
tex
tu
al
em
b
e
d
d
in
g
s
ℎ
,
an
d
tim
e
-
s
er
ies
f
ea
tu
r
es
ℎ
h
a
v
e
b
ee
n
f
u
s
ed
u
s
in
g
co
n
ca
ten
atio
n
f
o
llo
wed
b
y
a
d
en
s
e
p
r
o
jectio
n
:
ℎ
=
(
[
ℎ
;
ℎ
;
ℎ
]
+
)
T
h
is
r
ep
r
esen
tatio
n
h
as
b
ee
n
p
ass
ed
in
to
a
f
in
al
class
if
icatio
n
o
r
r
eg
r
ess
io
n
lay
er
d
ep
en
d
in
g
o
n
th
e
u
s
e
ca
s
e.
3
.
1
.
7
.
L
o
s
s
f
un
ct
io
n a
nd
m
o
del o
ptim
iza
t
io
n
T
o
ef
f
ec
tiv
el
y
tr
ain
d
ee
p
lear
n
in
g
m
o
d
els,
th
e
ch
o
ice
o
f
l
o
s
s
f
u
n
ctio
n
s
an
d
o
p
tim
izatio
n
alg
o
r
ith
m
s
p
lay
s
a
cr
itical
r
o
le
in
en
s
u
r
in
g
co
n
v
er
g
e
n
ce
an
d
ac
h
iev
i
n
g
h
ig
h
p
r
e
d
ictiv
e
p
er
f
o
r
m
an
ce
.
L
o
s
s
f
u
n
ctio
n
s
p
r
o
v
id
e
a
q
u
a
n
titativ
e
m
ea
s
u
r
e
o
f
th
e
d
is
cr
ep
an
c
y
b
etwe
en
p
r
ed
icted
o
u
tp
u
ts
an
d
ac
tu
al
ta
r
g
et
v
al
u
es,
g
u
id
in
g
th
e
m
o
d
el
to
m
in
im
ize
e
r
r
o
r
s
d
u
r
in
g
t
r
ain
in
g
.
De
p
en
d
in
g
o
n
th
e
ty
p
e
o
f
task
—
class
if
icat
io
n
o
r
r
e
g
r
ess
io
n
—
d
if
f
er
en
t
l
o
s
s
f
u
n
ctio
n
s
a
r
e
em
p
lo
y
ed
to
ca
p
tu
r
e
t
h
e
le
ar
n
in
g
o
b
jectiv
e.
O
n
ce
th
e
lo
s
s
is
co
m
p
u
ted
,
o
p
tim
izatio
n
alg
o
r
ith
m
s
s
u
ch
as
Ad
am
ar
e
u
s
ed
to
iter
at
iv
ely
u
p
d
ate
m
o
d
el
p
ar
a
m
eter
s
,
en
ab
lin
g
ef
f
icien
t
co
n
v
er
g
en
ce
to
war
d
th
e
o
p
tim
al
s
o
lu
tio
n
.
−
B
in
ar
y
cr
o
s
s
-
en
tr
o
p
y
lo
s
s
f
o
r
class
if
icatio
n
task
s
:
=
1
∑
[
(
̂
)
+
(
1
−
)
(
1
−
̂
)
]
=
1
−
Me
an
s
q
u
ar
ed
e
r
r
o
r
(
MSE
)
f
o
r
r
eg
r
ess
io
n
task
s
:
=
1
∑
(
=
1
−
̂
)
2
−
Op
tim
izatio
n
h
as b
ee
n
ca
r
r
ied
o
u
t u
s
in
g
th
e
Ad
am
o
p
tim
izer
:
=
1
−
1
+
(
1
−
1
)
,
=
2
−
1
+
(
1
−
2
)
2
̂
=
−
1
−
√
+
wh
er
e
is
th
e
g
r
a
d
ien
t,
η
is
th
e
lear
n
in
g
r
ate,
an
d
1
,
2
ar
e
m
o
m
en
t
u
m
co
ef
f
icien
ts
.
T
h
e
p
r
o
ce
s
s
b
eg
in
s
with
th
e
i
n
itializatio
n
o
f
g
lo
b
al
m
o
d
el
p
ar
am
eter
s
0
,
ty
p
ically
t
h
r
o
u
g
h
r
an
d
o
m
weig
h
t
ass
ig
n
m
en
ts
.
T
h
e
tr
ai
n
in
g
p
r
o
ce
d
u
r
e
u
n
f
o
ld
s
o
v
er
T
co
m
m
u
n
icatio
n
r
o
u
n
d
s
,
d
u
r
in
g
wh
ich
m
u
ltip
le
clien
ts
(
h
o
s
p
itals
,
clin
ics,
o
r
ed
g
e
d
ev
ices)
,
ea
c
h
with
th
ei
r
o
wn
d
ataset
,
co
llab
o
r
ativ
el
y
tr
ain
a
s
h
a
r
ed
tr
an
s
f
o
r
m
er
m
o
d
el.
T
h
e
clien
t
s
o
p
er
ate
in
p
a
r
allel,
en
s
u
r
in
g
co
m
p
u
tatio
n
al
e
f
f
icien
cy
a
n
d
s
ca
lab
ilit
y
.
Fo
r
ea
ch
co
m
m
u
n
icatio
n
r
o
u
n
d
t,
ev
er
y
clien
t
k
r
ec
eiv
es
a
co
p
y
o
f
th
e
g
lo
b
al
m
o
d
el
−
1
an
d
p
er
f
o
r
m
s
lo
ca
l
tr
ai
n
in
g
u
s
in
g
its
p
r
iv
ate
d
ataset
.
T
h
e
d
at
aset
is
d
iv
id
ed
in
to
b
atch
es,
an
d
th
e
m
o
d
el
is
tr
ain
ed
lo
ca
lly
f
o
r
E
ep
o
c
h
s
u
s
in
g
s
tan
d
ar
d
f
o
r
war
d
a
n
d
b
ac
k
war
d
p
r
o
p
ag
atio
n
s
tep
s
.
Pre
d
ictio
n
s
̂
ar
e
co
m
p
u
ted
u
s
in
g
th
e
tr
a
n
s
f
o
r
m
er
m
o
d
el,
lo
s
s
es
ar
e
ca
lcu
late
d
v
ia
a
p
r
ed
ef
in
ed
lo
s
s
f
u
n
cti
o
n
L
,
an
d
g
r
a
d
ien
ts
ar
e
co
m
p
u
ted
to
u
p
d
ate
t
h
e
lo
ca
l w
eig
h
ts
th
r
o
u
g
h
g
r
a
d
ien
t d
escen
t w
ith
a
lear
n
in
g
r
ate
η
.
4.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
T
o
v
a
l
i
d
a
t
e
t
h
e
p
e
r
f
o
r
m
a
n
c
e
a
n
d
e
f
f
i
c
i
e
n
c
y
o
f
t
h
e
p
r
o
p
o
s
e
d
p
r
i
v
a
c
y
-
p
r
e
s
e
r
v
i
n
g
t
r
a
n
s
f
o
r
m
e
r
-
b
a
s
ed
f
e
d
e
r
a
t
e
d
l
e
a
r
n
i
n
g
(
PP
T
F
L
)
m
o
d
e
l
,
a
c
o
m
p
r
e
h
e
n
s
i
v
e
e
v
al
u
a
ti
o
n
w
a
s
c
o
n
d
u
c
t
e
d
u
s
i
n
g
r
e
al
-
w
o
r
l
d
d
a
t
as
e
ts
.
T
h
e
e
v
a
l
u
a
t
i
o
n
f
o
c
u
s
e
d
o
n
m
e
a
s
u
r
i
n
g
c
l
a
s
s
i
f
i
c
a
ti
o
n
a
c
c
u
r
a
c
y
,
p
r
e
c
i
s
i
o
n
,
r
e
c
a
ll
,
F
1
s
c
o
r
e
,
c
o
m
p
u
t
a
t
i
o
n
a
l
t
i
m
e
,
a
n
d
p
r
i
v
a
c
y
o
v
e
r
h
e
a
d
.
T
h
e
p
e
r
f
o
r
m
a
n
c
e
o
f
t
h
e
p
r
o
p
o
s
e
d
m
o
d
e
l
w
a
s
b
e
n
c
h
m
a
r
k
e
d
a
g
a
i
n
s
t
c
o
n
v
e
n
t
i
o
n
a
l
f
e
d
e
r
a
t
e
d
l
e
a
r
n
i
n
g
f
r
a
m
e
w
o
r
k
s
a
n
d
s
e
l
e
cte
d
m
e
t
h
o
d
s
f
r
o
m
e
x
i
s
t
i
n
g
l
it
e
r
at
u
r
e
.
B
o
t
h
c
e
n
t
r
a
l
i
ze
d
a
n
d
d
e
c
en
t
r
a
l
i
z
e
d
s
c
e
n
a
r
i
o
s
w
e
r
e
t
es
t
e
d
.
T
h
e
f
o
l
l
o
w
i
n
g
t
o
o
ls
a
n
d
s
o
f
t
wa
r
e
e
n
v
i
r
o
n
m
e
n
t
s
we
r
e
u
s
e
d
f
o
r
i
m
p
l
e
m
e
n
t
at
i
o
n
a
n
d
e
v
a
l
u
a
t
i
o
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
7
2
2
-
2
5
8
6
I
AE
S I
n
t J Ro
b
&
Au
to
m
,
Vo
l
.
1
4
,
No
.
4
,
Dec
em
b
er
2
0
2
5
:
5
7
0
-
5
8
0
576
−
Pro
g
r
am
m
in
g
l
an
g
u
ag
e:
Py
th
o
n
3
.
1
0
−
L
ib
r
ar
ies:
Py
T
o
r
ch
(
v
2
.
1
)
,
Sci
k
it
-
lear
n
,
Nu
m
Py
,
Pan
d
as,
Ma
tp
lo
tlib
,
T
r
an
s
f
o
r
m
er
s
(
Hu
g
g
i
n
g
Face
)
−
Fed
er
ated
lear
n
in
g
f
r
am
ew
o
r
k
: Flo
wer
(
FLwr
)
−
Priv
ac
y
l
ib
r
ar
y
: O
p
ac
u
s
(
f
o
r
d
i
f
f
er
en
tial p
r
i
v
ac
y
ev
al
u
atio
n
)
−
Har
d
war
e:
NVI
DI
A
R
T
X
3
0
9
0
GPU,
6
4
GB
R
AM
,
I
n
tel
i9
p
r
o
ce
s
s
o
r
−
Op
er
atin
g
s
y
s
tem
: U
b
u
n
tu
2
2
.
0
4
L
T
S
R
ec
en
t
ad
v
an
ce
m
en
ts
in
f
ed
er
ated
lear
n
in
g
(
FL)
h
a
v
e
d
em
o
n
s
tr
ated
th
e
ab
ilit
y
to
tr
ain
m
o
d
els
co
llab
o
r
ativ
ely
ac
r
o
s
s
d
is
tr
ib
u
ted
h
ea
lth
ca
r
e
d
ata
with
o
u
t
co
m
p
r
o
m
is
in
g
p
atien
t
p
r
iv
ac
y
Ullah
et
a
l.
[
1
1
]
,
[
1
3
]
.
H
o
wev
er
,
c
h
allen
g
es
p
er
s
is
t
in
ac
h
iev
in
g
h
ig
h
ac
c
u
r
ac
y
wh
ile
m
in
im
izin
g
c
o
m
m
u
n
icatio
n
co
s
t
an
d
m
ain
tain
in
g
s
tr
o
n
g
p
r
iv
ac
y
g
u
ar
an
tees.
T
o
ad
d
r
ess
th
ese,
a
P
PTFL
m
o
d
el
was p
r
o
p
o
s
ed
an
d
ev
alu
ated
ag
ain
s
t
co
n
v
en
tio
n
al
ce
n
tr
alize
d
a
n
d
FL
-
b
ased
m
o
d
els as in
T
ab
le
1
.
T
h
e
p
r
o
p
o
s
ed
PP
T
FL
m
o
d
el
o
u
tp
er
f
o
r
m
ed
all
b
aselin
e
a
p
p
r
o
ac
h
es
ac
r
o
s
s
k
e
y
p
e
r
f
o
r
m
an
ce
m
etr
ics.
I
t
ac
h
iev
ed
th
e
h
i
g
h
est
ac
cu
r
ac
y
(
9
2
.
8
7
%),
p
r
ec
is
io
n
(
9
1
.
4
2
%),
r
ec
all
(
9
3
.
3
5
%),
a
n
d
F1
s
co
r
e
(
9
2
.
3
7
%),
d
em
o
n
s
tr
atin
g
its
s
u
p
er
io
r
cla
s
s
if
icatio
n
ca
p
ab
ilit
y
in
h
ea
lth
ca
r
e
co
n
tex
ts
.
As
s
h
o
wn
in
T
a
b
le
2
,
th
e
ac
c
u
r
ac
y
an
d
F1
s
co
r
e
s
tead
ily
im
p
r
o
v
ed
with
an
in
cr
ea
s
e
in
th
e
n
u
m
b
er
o
f
clien
ts
,
r
ea
ch
in
g
a
p
ea
k
ac
c
u
r
ac
y
o
f
9
2
.
8
7
%
a
n
d
F1
s
co
r
e
o
f
9
2
.
3
7
%
with
2
0
clien
ts
.
T
h
is
in
d
i
ca
tes
th
at
th
e
p
r
o
p
o
s
ed
PP
T
FL
m
o
d
el
s
ca
les
well
with
th
e
ad
d
itio
n
o
f
m
o
r
e
clien
ts
,
lik
ely
to
b
e
n
ef
it
f
r
o
m
t
h
e
in
cr
ea
s
ed
d
ata
d
iv
er
s
ity
p
r
o
v
id
ed
b
y
m
u
ltip
le
clien
t m
o
d
els.
T
ab
le
3
s
h
o
ws th
e
co
m
p
u
tatio
n
ef
f
icien
cy
o
v
e
r
ep
o
ch
s
.
As
s
ee
n
in
Fig
u
r
e
2
,
th
e
ac
cu
r
ac
y
o
f
th
e
m
o
d
el
im
p
r
o
v
es
s
tead
ily
as
th
e
n
u
m
b
er
o
f
ep
o
ch
s
in
cr
ea
s
es.
Star
tin
g
with
8
6
.
2
1
%
ac
c
u
r
ac
y
at
ep
o
ch
1
0
,
th
e
ac
c
u
r
ac
y
r
is
es
to
9
2
.
8
7
%
b
y
ep
o
ch
4
0
,
in
d
icatin
g
th
e
m
o
d
el'
s
ab
ilit
y
to
en
h
an
ce
its
p
r
ed
icti
v
e
p
er
f
o
r
m
a
n
ce
with
m
o
r
e
t
r
ain
in
g
iter
atio
n
s
.
T
h
is
is
co
n
s
is
ten
t
with
ty
p
ical
d
ee
p
lear
n
in
g
m
o
d
els,
wh
er
e
m
o
r
e
tr
ain
in
g
ep
o
ch
s
g
en
e
r
ally
lead
to
b
etter
r
esu
lts
.
T
ab
le
1
.
Mo
d
el
p
er
f
o
r
m
a
n
ce
o
n
MI
MI
C
-
III
d
ataset
M
e
t
r
i
c
C
e
n
t
r
a
l
i
z
e
d
t
r
a
n
sf
o
r
mer
F
L
+
C
N
N
F
L
+
LSTM
P
r
o
p
o
se
d
P
P
TFL
A
c
c
u
r
a
c
y
(
%)
9
0
.
1
3
8
8
.
2
5
8
9
.
0
6
9
2
.
8
7
P
r
e
c
i
s
i
o
n
(
%)
8
9
.
0
1
8
7
.
1
2
8
8
.
5
4
9
1
.
4
2
R
e
c
a
l
l
(
%)
9
0
.
6
0
8
5
.
6
7
8
7
.
9
1
9
3
.
3
5
F
1
s
c
o
r
e
(
%)
8
9
.
8
0
8
6
.
3
8
8
8
.
2
2
9
2
.
3
7
ϵ (P
r
i
v
a
c
y
)
∞
(
n
o
p
r
i
v
a
c
y
)
3
.
5
2
.
8
1
.
6
Tr
a
i
n
i
n
g
T
i
me
(
s)
4
1
0
3
8
5
3
7
2
4
4
2
C
o
mm
.
C
o
s
t
(
M
B
)
N
/
A
2
4
0
2
7
0
2
1
8
T
ab
le
2
.
I
m
p
ac
t o
f
clien
t c
o
u
n
t o
n
ac
cu
r
ac
y
N
o
.
o
f
C
l
i
e
n
t
s
A
c
c
u
r
a
c
y
(
%)
F
1
sc
o
r
e
(
%)
C
o
mm
.
O
v
e
r
h
e
a
d
(
M
B
)
5
8
9
.
7
4
8
8
.
3
0
1
9
0
10
9
1
.
4
2
9
0
.
8
6
2
0
5
15
9
2
.
0
1
9
1
.
6
2
2
1
5
20
9
2
.
8
7
9
2
.
3
7
2
1
8
T
ab
le
3
.
C
o
m
p
u
tatio
n
al
ef
f
icie
n
cy
o
v
er
ep
o
ch
s
Ep
o
c
h
s
A
c
c
u
r
a
c
y
(
%)
Tr
a
i
n
i
n
g
t
i
me
(
s)
P
r
i
v
a
c
y
b
u
d
g
e
t
(
ϵ)
10
8
6
.
2
1
98
2
.
5
20
8
8
.
7
9
1
9
0
2
.
2
30
9
1
.
3
4
2
9
5
1
.
9
40
9
2
.
8
7
4
4
2
1
.
6
Fig
u
r
e
2
.
C
o
m
p
a
r
is
o
n
o
f
m
etr
i
cs v
s
E
p
o
ch
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
AE
S
I
n
t
J
R
o
b
&
A
u
to
m
I
SS
N:
2722
-
2
5
8
6
E
n
h
a
n
cin
g
h
ea
lth
s
ta
tu
s
p
r
ed
ictio
n
a
n
d
d
a
ta
s
ec
u
r
ity
…
(
S
u
b
r
a
ma
n
iya
n
S
en
th
a
ma
r
a
i
)
577
T
ab
le
4
co
m
p
ar
es
th
e
p
er
f
o
r
m
an
ce
o
f
th
e
p
r
o
p
o
s
ed
PP
T
FL
m
o
d
el
with
f
o
u
r
r
elev
a
n
t
m
eth
o
d
s
f
r
o
m
th
e
liter
atu
r
e.
T
h
is
co
m
p
ar
is
o
n
is
b
ased
o
n
k
ey
e
v
alu
atio
n
m
etr
ics,
in
clu
d
in
g
ac
cu
r
ac
y
,
F1
s
co
r
e,
p
r
iv
ac
y
b
u
d
g
et
(
ϵ
)
,
a
n
d
c
o
m
m
u
n
icatio
n
co
s
t.
Fro
m
th
e
c
o
m
p
ar
ativ
e
ev
alu
atio
n
in
Fig
u
r
e
3
to
Fig
u
r
e
6
,
it
is
ev
id
e
n
t
th
at
th
e
p
r
o
p
o
s
ed
PP
T
FL
m
o
d
el
o
u
tp
er
f
o
r
m
s
all
th
e
o
th
er
m
eth
o
d
s
ac
r
o
s
s
th
e
lis
ted
m
etr
ics.
T
h
e
PP
T
FL
m
o
d
el
ac
h
iev
es
th
e
h
ig
h
est
ac
cu
r
ac
y
(
9
2
.
8
7
%)
an
d
F1
s
co
r
e
(
9
2
.
3
7
%)
wh
en
co
m
p
ar
ed
to
o
th
er
m
o
d
els
s
u
c
h
as P2
FL
(
9
0
.
0
1
%),
PF
L
F (
9
1
.
1
2
%),
an
d
m
u
lti
-
s
o
u
r
ce
FL
(
9
1
.
4
7
%).
T
ab
le
4
.
C
o
m
p
a
r
is
o
n
with
liter
atu
r
e
m
eth
o
d
s
M
o
d
e
l
/
M
e
t
h
o
d
A
c
c
u
r
a
c
y
(
%)
F
1
sc
o
r
e
(
%)
P
r
i
v
a
c
y
(
ϵ)
C
o
mm
.
C
o
s
t
(
M
B
)
P
2
F
L
[
1
1
]
9
0
.
0
1
8
9
.
3
3
2
.
2
2
4
0
P
F
LF
[
1
3
]
9
1
.
1
2
9
0
.
2
1
2
.
0
2
4
5
M
u
l
t
i
-
so
u
r
c
e
F
L
[
1
8
]
9
1
.
4
7
9
1
.
1
2
1
.
9
2
5
2
M
LP w
i
t
h
C
r
o
ss
o
v
e
r
[
2
0
]
8
9
.
8
8
8
8
.
7
2
—
2
5
5
P
P
TFL (Tr
a
n
sf
o
r
mer)
(
P
r
o
p
o
se
d
w
o
r
k
)
9
2
.
8
7
9
2
.
3
7
1
.
6
2
1
8
Fig
u
r
e
3
.
Acc
u
r
ac
y
c
o
m
p
ar
is
o
n
with
v
ar
io
u
s
m
o
d
els
Fig
u
r
e
4
.
F1
s
co
r
e
co
m
p
ar
is
o
n
with
v
ar
io
u
s
m
o
d
els
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
7
2
2
-
2
5
8
6
I
AE
S I
n
t J Ro
b
&
Au
to
m
,
Vo
l
.
1
4
,
No
.
4
,
Dec
em
b
er
2
0
2
5
:
5
7
0
-
5
8
0
578
Fig
u
r
e
5
.
Priv
ac
y
co
m
p
ar
is
o
n
with
v
ar
io
u
s
m
o
d
els
Fig
u
r
e
6
.
C
o
m
m
u
n
icatio
n
c
o
s
t a
cr
o
s
s
v
ar
io
u
s
m
o
d
els
5.
CO
NCLU
SI
O
N
I
n
th
is
s
tu
d
y
,
we
p
r
o
p
o
s
ed
a
PP
T
FL
f
r
am
ewo
r
k
d
esig
n
ed
to
en
h
an
ce
th
e
p
r
iv
a
cy
,
ac
c
u
r
ac
y
,
a
n
d
ef
f
icien
cy
o
f
h
ea
lth
ca
r
e
d
ata
p
r
o
ce
s
s
in
g
in
f
ed
er
ate
d
lear
n
in
g
en
v
ir
o
n
m
en
ts
.
T
h
e
m
o
d
el
w
as
ev
alu
ated
ac
r
o
s
s
s
ev
er
al
k
ey
m
etr
ics,
in
clu
d
in
g
ac
cu
r
ac
y
,
F1
s
co
r
e
,
p
r
iv
ac
y
b
u
d
g
et,
tr
ai
n
in
g
tim
e,
an
d
c
o
m
m
u
n
icatio
n
co
s
t,
u
s
in
g
a
s
im
u
lated
I
o
MT
d
at
aset.
T
h
e
r
esu
lts
d
em
o
n
s
tr
ate
th
at
th
e
PP
T
FL
m
o
d
el
o
u
tp
er
f
o
r
m
s
ex
is
tin
g
m
eth
o
d
s
in
ter
m
s
o
f
ac
c
u
r
ac
y
,
with
an
im
p
r
ess
iv
e
ac
cu
r
ac
y
o
f
9
2
.
8
7
%
an
d
a
n
F1
s
co
r
e
o
f
9
2
.
3
7
%.
Mo
r
eo
v
e
r
,
th
e
p
r
i
v
ac
y
b
u
d
g
et
(
ϵ
)
o
f
1
.
6
f
u
r
th
er
u
n
d
er
s
co
r
es
th
e
m
o
d
el'
s
r
o
b
u
s
t
p
r
iv
ac
y
-
p
r
eser
v
in
g
ca
p
ab
ilit
ies,
wh
ich
is
cr
itical
in
th
e
h
ea
lth
ca
r
e
d
o
m
ain
wh
er
e
d
ata
co
n
f
id
en
tiality
is
p
ar
am
o
u
n
t.
Fu
tu
r
e
wo
r
k
c
an
ex
p
l
o
r
e
f
u
r
th
er
o
p
tim
izatio
n
o
f
th
e
m
o
d
el'
s
ef
f
icien
cy
an
d
s
ca
lab
ilit
y
,
as we
ll a
s
i
ts
r
ea
l
-
wo
r
ld
d
ep
lo
y
m
en
t
f
o
r
m
o
r
e
co
m
p
lex
,
h
eter
o
g
en
e
o
u
s
h
ea
lth
ca
r
e
d
a
tasets
.
T
h
e
o
n
g
o
in
g
d
ev
elo
p
m
en
t
o
f
s
u
ch
f
r
am
ewo
r
k
s
will
co
n
tr
ib
u
te
to
ad
v
an
cin
g
s
ec
u
r
e
an
d
ef
f
icien
t
AI
s
o
lu
tio
n
s
f
o
r
th
e
h
ea
lth
ca
r
e
s
ec
to
r
,
e
n
s
u
r
in
g
b
o
th
th
e
p
r
o
tectio
n
o
f
s
en
s
itiv
e
p
atien
t d
ata
an
d
th
e
d
eliv
e
r
y
o
f
ac
cu
r
ate
clin
ical
p
r
ed
ictio
n
s
.
F
UNDING
I
NF
O
R
M
A
T
I
O
N
T
h
is
s
ec
tio
n
s
h
o
u
ld
d
escr
ib
e
s
o
u
r
ce
s
o
f
f
u
n
d
in
g
a
g
en
cy
th
at
h
av
e
s
u
p
p
o
r
ted
th
e
wo
r
k
.
Au
th
o
r
s
s
h
o
u
ld
s
tate
h
o
w
th
e
r
esear
ch
d
escr
ib
ed
in
th
eir
ar
ticle
wa
s
f
u
n
d
ed
,
in
clu
d
in
g
g
r
an
t
n
u
m
b
er
s
if
ap
p
licab
le.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
AE
S
I
n
t
J
R
o
b
&
A
u
to
m
I
SS
N:
2722
-
2
5
8
6
E
n
h
a
n
cin
g
h
ea
lth
s
ta
tu
s
p
r
ed
ictio
n
a
n
d
d
a
ta
s
ec
u
r
ity
…
(
S
u
b
r
a
ma
n
iya
n
S
en
th
a
ma
r
a
i
)
579
I
n
clu
d
e
th
e
f
o
llo
win
g
(
o
r
s
im
ilar
)
s
tatem
en
t
if
th
er
e
is
n
o
f
u
n
d
in
g
in
v
o
lv
e
d
:
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
Su
b
r
am
an
iy
a
n
Sen
th
am
ar
ai
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
R
aja
Ma
n
ick
am
Ma
la
✓
✓
✓
✓
✓
✓
✓
Vellaiy
an
Palan
is
am
y
✓
✓
✓
✓
✓
✓
✓
✓
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
Data
av
ailab
ilit
y
is
n
o
t
ap
p
li
ca
b
le
to
th
is
p
ap
er
as
n
o
n
e
w
d
ata
wer
e
cr
ea
ted
o
r
an
aly
ze
d
in
th
is
s
tu
d
y
.
RE
F
E
R
E
NC
E
S
[
1
]
M
.
K
u
l
i
h
a
a
n
d
S
.
V
e
r
m
a
,
“
S
e
c
u
r
e
i
n
t
e
r
n
e
t
o
f
me
d
i
c
a
l
t
h
i
n
g
s
b
a
s
e
d
e
l
e
c
t
r
o
n
i
c
h
e
a
l
t
h
r
e
c
o
r
d
s
sc
h
e
m
e
i
n
t
r
u
st
d
e
c
e
n
t
r
a
l
i
z
e
d
l
o
o
p
f
e
d
e
r
a
t
e
d
l
e
a
r
n
i
n
g
c
o
n
se
n
su
s
b
l
o
c
k
c
h
a
i
n
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
f
I
n
t
e
l
l
i
g
e
n
t
N
e
t
w
o
rks
,
v
o
l
.
5
,
p
p
.
1
6
1
–
1
7
4
,
2
0
2
4
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
i
j
i
n
.
2
0
2
4
.
0
3
.
0
0
1
.
[
2
]
A
.
K
.
A
l
H
w
a
i
t
a
t
e
t
a
l
.
,
“
A
n
e
w
b
l
o
c
k
c
h
a
i
n
-
b
a
s
e
d
a
u
t
h
e
n
t
i
c
a
t
i
o
n
f
r
a
mew
o
r
k
f
o
r
se
c
u
r
e
I
o
T
n
e
t
w
o
r
k
s,
”
E
l
e
c
t
r
o
n
i
c
s
,
v
o
l
.
1
2
,
n
o
.
1
7
,
p
.
3
6
1
8
,
A
u
g
.
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
e
l
e
c
t
r
o
n
i
c
s
1
2
1
7
3
6
1
8
.
[
3
]
M
.
A
b
a
o
u
d
,
M
.
A
.
A
l
m
u
q
r
i
n
,
a
n
d
M
.
F
.
K
h
a
n
,
“
A
d
v
a
n
c
i
n
g
f
e
d
e
r
a
t
e
d
l
e
a
r
n
i
n
g
t
h
r
o
u
g
h
n
o
v
e
l
mec
h
a
n
i
sms
f
o
r
p
r
i
v
a
c
y
p
r
e
ser
v
a
t
i
o
n
i
n
h
e
a
l
t
h
c
a
r
e
a
p
p
l
i
c
a
t
i
o
n
s,
”
I
EEE
Ac
c
e
ss
,
v
o
l
.
1
1
,
p
p
.
8
3
5
6
2
–
8
3
5
7
9
,
2
0
2
3
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
ESS
.
2
0
2
3
.
3
3
0
1
1
6
2
.
[
4
]
A
.
A
.
La
g
h
a
r
i
,
A
.
A
.
K
h
a
n
,
R
.
A
l
k
a
n
h
e
l
,
H
.
El
m
a
n
n
a
i
,
a
n
d
S
.
B
o
u
r
o
u
i
s,
“
Li
g
h
t
w
e
i
g
h
t
-
B
I
o
V
:
B
l
o
c
k
c
h
a
i
n
d
i
st
r
i
b
u
t
e
d
l
e
d
g
e
r
t
e
c
h
n
o
l
o
g
y
(
B
D
LT)
f
o
r
i
n
t
e
r
n
e
t
o
f
v
e
h
i
c
l
e
s
(
I
o
V
s)
,
”
El
e
c
t
r
o
n
i
c
s
,
v
o
l
.
1
2
,
n
o
.
3
,
p
.
6
7
7
,
Jan
.
2
0
2
3
,
d
o
i
:
1
0
.
3
3
9
0
/
e
l
e
c
t
r
o
n
i
c
s
1
2
0
3
0
6
7
7
.
[
5
]
B
.
B
e
r
me
j
o
a
n
d
C
.
J
u
i
z
,
“
I
mp
r
o
v
i
n
g
c
l
o
u
d
/
e
d
g
e
su
s
t
a
i
n
a
b
i
l
i
t
y
t
h
r
o
u
g
h
a
r
t
i
f
i
c
i
a
l
i
n
t
e
l
l
i
g
e
n
c
e
:
A
sy
s
t
e
m
a
t
i
c
r
e
v
i
e
w
,
”
J
o
u
r
n
a
l
o
f
Pa
r
a
l
l
e
l
a
n
d
D
i
st
r
i
b
u
t
e
d
C
o
m
p
u
t
i
n
g
,
v
o
l
.
1
7
6
,
p
p
.
4
1
–
5
4
,
J
u
n
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
j
p
d
c
.
2
0
2
3
.
0
2
.
0
0
6
.
[
6
]
I
.
K
h
a
n
e
t
a
l
.
,
“
S
e
c
u
r
i
n
g
B
l
o
c
k
c
h
a
i
n
-
b
a
s
e
d
s
u
p
p
l
y
c
h
a
i
n
m
a
n
a
g
e
m
e
n
t
:
T
e
x
t
u
a
l
d
a
t
a
e
n
c
r
y
p
t
i
o
n
a
n
d
a
c
c
e
ss
c
o
n
t
r
o
l
,
”
T
e
c
h
n
o
l
o
g
i
e
s
,
v
o
l
.
1
2
,
n
o
.
7
,
2
0
2
4
,
d
o
i
:
1
0
.
3
3
9
0
/
t
e
c
h
n
o
l
o
g
i
e
s
1
2
0
7
0
1
1
0
.
[
7
]
T.
S
a
b
a
,
K
.
H
a
see
b
,
A
.
R
e
h
m
a
n
,
a
n
d
G
.
Jeo
n
,
“
B
l
o
c
k
c
h
a
i
n
-
e
n
a
b
l
e
d
i
n
t
e
l
l
i
g
e
n
t
I
o
T
p
r
o
t
o
c
o
l
f
o
r
h
i
g
h
-
p
e
r
f
o
r
ma
n
c
e
a
n
d
se
c
u
r
e
d
b
i
g
f
i
n
a
n
c
i
a
l
d
a
t
a
t
r
a
n
sa
c
t
i
o
n
,
”
I
EEE
T
r
a
n
s
a
c
t
i
o
n
s
o
n
C
o
m
p
u
t
a
t
i
o
n
a
l
S
o
c
i
a
l
S
y
s
t
e
m
s
,
v
o
l
.
1
1
,
n
o
.
2
,
p
p
.
1
6
6
7
–
1
6
7
4
,
2
0
2
4
,
d
o
i
:
1
0
.
1
1
0
9
/
T
C
S
S
.
2
0
2
3
.
3
2
6
8
5
9
2
.
[
8
]
O
.
C
h
e
i
k
h
r
o
u
h
o
u
,
K
.
M
e
r
sh
a
d
,
F
.
Ja
mi
l
,
R
.
M
a
h
m
u
d
,
A
.
K
o
u
b
a
a
,
a
n
d
S
.
R
.
M
o
o
s
a
v
i
,
“
A
l
i
g
h
t
w
e
i
g
h
t
b
l
o
c
k
c
h
a
i
n
a
n
d
f
o
g
-
e
n
a
b
l
e
d
sec
u
r
e
r
e
m
o
t
e
p
a
t
i
e
n
t
m
o
n
i
t
o
r
i
n
g
s
y
st
e
m,”
I
n
t
e
r
n
e
t
o
f
T
h
i
n
g
s (N
e
t
h
e
r
l
a
n
d
s)
,
v
o
l
.
2
2
,
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
i
o
t
.
2
0
2
3
.
1
0
0
6
9
1
.
[
9
]
P
.
S
h
a
r
m
a
,
S
.
N
a
mas
u
d
r
a
,
R
.
G
o
n
z
a
l
e
z
C
r
e
sp
o
,
J
.
P
a
r
r
a
-
F
u
e
n
t
e
,
a
n
d
M
.
C
h
a
n
d
r
a
Tr
i
v
e
d
i
,
“
EH
D
H
E:
E
n
h
a
n
c
i
n
g
s
e
c
u
r
i
t
y
o
f
h
e
a
l
t
h
c
a
r
e
d
o
c
u
m
e
n
t
s
i
n
I
o
T
-
e
n
a
b
l
e
d
d
i
g
i
t
a
l
h
e
a
l
t
h
c
a
r
e
e
c
o
s
y
st
e
ms
u
s
i
n
g
b
l
o
c
k
c
h
a
i
n
,
”
I
n
f
o
rm
a
t
i
o
n
S
c
i
e
n
c
e
s
,
v
o
l
.
6
2
9
,
p
p
.
7
0
3
–
7
1
8
,
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
i
n
s
.
2
0
2
3
.
0
1
.
1
4
8
.
[
1
0
]
S
.
G
u
l
a
t
i
,
K
.
G
u
l
e
r
i
a
,
a
n
d
N
.
G
o
y
a
l
,
“
P
r
i
v
a
c
y
-
p
r
e
s
e
r
v
i
n
g
a
n
d
c
o
l
l
a
b
o
r
a
t
i
v
e
f
e
d
e
r
a
t
e
d
l
e
a
r
n
i
n
g
m
o
d
e
l
f
o
r
t
h
e
d
e
t
e
c
t
i
o
n
o
f
o
c
u
l
a
r
d
i
s
e
a
se
s,”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
M
a
t
h
e
m
a
t
i
c
a
l
,
E
n
g
i
n
e
e
ri
n
g
a
n
d
M
a
n
a
g
e
m
e
n
t
S
c
i
e
n
c
e
s
,
v
o
l
.
1
0
,
n
o
.
1
,
p
p
.
2
1
8
–
2
4
8
,
F
e
b
.
2
0
2
5
,
d
o
i
:
1
0
.
3
3
8
8
9
/
I
JM
E
M
S
.
2
0
2
5
.
1
0
.
1
.
0
1
3
.
[
1
1
]
F
.
U
l
l
a
h
e
t
a
l
.
,
“
P
2
F
L
:
P
r
i
v
a
c
y
-
p
r
e
se
r
v
i
n
g
f
e
d
e
r
a
t
e
d
l
e
a
r
n
i
n
g
a
p
p
r
o
a
c
h
f
o
r
h
e
a
l
t
h
c
a
r
e
i
n
f
o
r
ma
t
i
c
s
a
t
t
h
e
e
d
g
e
,
”
L
e
c
t
u
re
N
o
t
e
s
o
n
D
a
t
a
E
n
g
i
n
e
e
ri
n
g
a
n
d
C
o
m
m
u
n
i
c
a
t
i
o
n
s Te
c
h
n
o
l
o
g
i
e
s
,
v
o
l
.
2
3
2
,
p
p
.
4
7
–
5
8
,
2
0
2
5
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
3
-
0
3
1
-
7
6
4
6
2
-
2
_
5
.
[
1
2
]
M
.
B
a
r
a
t
i
e
t
a
l
.
,
“
P
r
i
v
a
c
y
-
a
w
a
r
e
c
l
o
u
d
a
u
d
i
t
i
n
g
f
o
r
G
D
P
R
c
o
m
p
l
i
a
n
c
e
v
e
r
i
f
i
c
a
t
i
o
n
i
n
o
n
l
i
n
e
h
e
a
l
t
h
c
a
r
e
,
”
I
EEE
T
ra
n
s
a
c
t
i
o
n
s
o
n
I
n
d
u
st
r
i
a
l
I
n
f
o
rm
a
t
i
c
s
,
v
o
l
.
1
8
,
n
o
.
7
,
p
p
.
4
8
0
8
–
4
8
1
9
,
2
0
2
2
,
d
o
i
:
1
0
.
1
1
0
9
/
TI
I
.
2
0
2
1
.
3
1
0
0
1
5
2
.
[
1
3
]
H
.
Zh
o
u
,
G
.
Y
a
n
g
,
H
.
D
a
i
,
a
n
d
G
.
Li
u
,
“
P
F
LF
:
P
r
i
v
a
c
y
-
p
r
e
ser
v
i
n
g
f
e
d
e
r
a
t
e
d
l
e
a
r
n
i
n
g
f
r
a
m
e
w
o
r
k
f
o
r
e
d
g
e
c
o
mp
u
t
i
n
g
,
”
I
EE
E
T
ra
n
s
a
c
t
i
o
n
s
o
n
I
n
f
o
rm
a
t
i
o
n
Fo
re
n
si
c
s a
n
d
S
e
c
u
ri
t
y
,
v
o
l
.
1
7
,
p
p
.
1
9
0
5
–
1
9
1
8
,
2
0
2
2
.
[
1
4
]
R
.
W
a
n
g
,
J.
La
i
,
Z
.
Z
h
a
n
g
,
X
.
L
i
,
P
.
V
i
j
a
y
a
k
u
m
a
r
,
a
n
d
M
.
K
a
r
u
p
p
i
a
h
,
“
P
r
i
v
a
c
y
-
p
r
e
ser
v
i
n
g
f
e
d
e
r
a
t
e
d
l
e
a
r
n
i
n
g
f
o
r
i
n
t
e
r
n
e
t
o
f
med
i
c
a
l
t
h
i
n
g
s
u
n
d
e
r
e
d
g
e
c
o
mp
u
t
i
n
g
,
”
I
E
EE
J
o
u
r
n
a
l
o
f
B
i
o
m
e
d
i
c
a
l
a
n
d
H
e
a
l
t
h
I
n
f
o
rm
a
t
i
c
s
,
v
o
l
.
2
7
,
n
o
.
2
,
p
p
.
8
5
4
–
8
6
5
,
2
0
2
3
,
d
o
i
:
1
0
.
1
1
0
9
/
J
B
H
I
.
2
0
2
2
.
3
1
5
7
7
2
5
.
[
1
5
]
W
.
P
a
n
,
Z.
X
u
,
S
.
R
a
j
e
n
d
r
a
n
,
a
n
d
F
.
W
a
n
g
,
“
A
n
a
d
a
p
t
i
v
e
f
e
d
e
r
a
t
e
d
l
e
a
r
n
i
n
g
f
r
a
mew
o
r
k
f
o
r
c
l
i
n
i
c
a
l
r
i
sk
p
r
e
d
i
c
t
i
o
n
w
i
t
h
e
l
e
c
t
r
o
n
i
c
h
e
a
l
t
h
r
e
c
o
r
d
s
f
r
o
m m
u
l
t
i
p
l
e
h
o
sp
i
t
a
l
s
,
”
P
a
t
t
e
r
n
s
,
v
o
l
.
5
,
n
o
.
1
,
2
0
2
4
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
p
a
t
t
e
r
.
2
0
2
3
.
1
0
0
8
9
8
.
Evaluation Warning : The document was created with Spire.PDF for Python.