I
nte
rna
t
io
na
l J
o
urna
l o
f
E
lect
rica
l a
nd
Co
m
pu
t
er
E
ng
ineering
(
I
J
E
CE
)
Vo
l.
15
,
No
.
6
,
Decem
b
er
20
25
,
p
p
.
5
7
2
8
~
5
7
4
5
I
SS
N:
2088
-
8
7
0
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijece.
v
15
i
6
.
pp
5
7
2
8
-
5
7
4
5
5728
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ec
e.
ia
esco
r
e.
co
m
Securing
heal
thca
re data
and o
p
tim
izing
digita
l ma
rk
eting
throug
h ma
chine
lea
rning
:
th
e C
A
M
L
-
EHDS
f
ra
me
wo
rk
F
a
t
hi Abderr
a
hm
a
ne
1
,
M
o
uy
a
s
s
ir
K
a
wt
a
r
1
,
Ali Wa
qa
s
2
,
F
a
nd
i F
a
t
im
a
Z
a
hra
3
,
K
a
rt
it
Ali
1
1
N
a
t
i
o
n
a
l
S
c
h
o
o
l
o
f
A
p
p
l
i
e
d
S
c
i
e
n
c
e
s,
Th
e
I
n
f
o
r
ma
t
i
o
n
T
e
c
h
n
o
l
o
g
y
La
b
o
r
a
t
o
r
y
a
t
C
h
o
u
a
i
b
D
o
u
k
k
a
l
i
U
n
i
v
e
r
s
i
t
y
E
l
Ja
d
i
d
a
,
M
o
r
o
c
c
o
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
,
U
n
i
v
e
r
s
i
t
y
o
f
E
n
g
i
n
e
e
r
i
n
g
a
n
d
Te
c
h
n
o
l
o
g
y
,
La
h
o
r
e
,
P
a
k
i
s
t
a
n
3
LTI
M
,
H
a
ss
a
n
I
I
U
n
i
v
e
r
si
t
y
,
C
a
sa
b
l
a
n
c
a
,
M
o
r
o
c
c
o
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Feb
1
,
2
0
2
5
R
ev
is
ed
J
u
l 3
0
,
2
0
2
5
Acc
ep
ted
Sep
1
5
,
2
0
2
5
C
u
r
r
e
n
t
h
e
a
l
t
h
c
a
re
d
a
t
a
s
y
s
tem
s
f
a
c
e
m
a
j
o
r
c
h
a
l
le
n
g
e
s
i
n
p
r
e
v
e
n
t
i
n
g
u
n
a
u
t
h
o
r
i
z
e
d
a
c
c
e
s
s
,
e
n
s
u
r
i
n
g
c
o
m
p
l
i
a
n
c
e
w
i
t
h
d
a
ta
p
r
i
v
a
c
y
r
e
g
u
l
a
t
i
o
n
s
,
a
n
d
e
n
a
b
l
i
n
g
i
n
t
e
l
l
i
g
e
n
t
s
e
c
o
n
d
a
r
y
u
s
e
o
f
p
a
t
i
e
n
t
i
n
f
o
r
m
a
t
i
o
n
.
T
o
a
d
d
r
e
s
s
t
h
e
se
i
s
s
u
e
s
,
w
e
i
n
t
r
o
d
u
c
e
c
l
u
s
t
e
r
-
b
a
se
d
a
n
a
l
y
s
i
s
w
i
t
h
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
f
o
r
e
n
h
a
n
c
e
d
h
e
a
l
t
h
c
a
re
d
a
ta
se
c
u
r
i
t
y
(
C
A
M
L
-
E
H
D
S
)
,
a
u
n
i
f
i
e
d
fr
a
m
e
w
o
r
k
t
h
a
t
c
o
m
b
i
n
e
s
h
o
m
o
m
o
r
p
h
i
c
e
n
c
r
y
p
t
i
o
n
,
a
t
t
r
i
b
u
t
e
-
b
a
se
d
e
l
l
i
p
t
i
c
c
u
r
v
e
c
r
y
p
t
o
g
r
a
p
h
y
(
E
C
C
)
,
a
n
d
s
e
m
a
n
t
i
c
c
l
u
s
t
e
r
i
n
g
w
i
t
h
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
.
C
A
M
L
-
E
HD
S
i
m
p
ro
v
e
s
u
p
o
n
e
x
i
s
t
i
n
g
m
o
d
e
l
s
b
y
o
f
f
e
r
i
n
g
f
i
n
e
-
g
r
a
i
n
e
d
a
c
c
e
s
s
c
o
n
t
r
o
l
,
a
d
a
p
t
i
v
e
t
h
r
e
a
t
d
e
t
e
c
t
i
o
n
,
a
n
d
d
a
t
a
-
d
r
i
v
e
n
i
n
s
i
g
h
t
s
w
h
i
l
e
p
r
e
s
e
r
v
i
n
g
p
r
i
v
a
c
y
.
E
x
p
e
r
i
m
e
n
t
a
l
r
e
s
u
l
t
s
s
h
o
w
t
h
a
t
C
A
M
L
-
EH
D
S
a
c
h
i
e
v
e
s
u
p
t
o
9
8
%
c
l
a
s
s
i
f
ic
a
t
i
o
n
a
c
c
u
r
a
c
y
w
i
t
h
l
o
w
n
o
d
e
c
o
u
n
t
,
a
n
d
m
a
i
n
t
a
i
n
s
9
4
%
a
c
c
u
r
a
c
y
e
v
e
n
a
t
h
i
g
h
n
o
d
e
d
i
s
t
r
i
b
u
t
i
o
n
l
e
v
e
l
s
,
w
h
i
l
e
e
n
s
u
r
i
n
g
e
n
c
r
y
p
t
i
o
n
t
i
m
e
u
n
d
e
r
2
4
s
e
c
o
n
d
s
a
n
d
a
c
c
e
p
t
a
b
l
e
d
a
t
a
l
o
ss
b
e
l
o
w 2
9
%
.
M
o
r
e
o
v
e
r
,
i
n
c
o
m
p
a
ra
t
i
v
e
a
n
a
l
y
s
i
s w
i
t
h
s
ta
t
e
-
of
-
t
h
e
-
a
r
t
m
o
d
e
l
s
(
s
u
p
p
o
r
t
v
e
c
t
o
r
m
a
c
h
i
n
e
(
S
V
M
)
,
ra
n
d
o
m
f
o
r
e
s
t
(
R
F
)
,
a
n
d
d
e
c
i
s
i
o
n
t
r
e
e
(D
T
)
)
,
C
A
M
L
-
EHD
S
o
u
t
p
e
r
f
o
r
m
s
a
l
l
i
n
k
e
y
m
e
t
r
ic
s
w
i
t
h
a
n
a
c
c
u
ra
c
y
o
f
0
.
9
6
.
T
h
e
s
e
r
e
s
u
l
t
s
d
e
m
o
n
s
t
r
a
t
e
C
A
M
L
-
E
HD
S
’
s
p
o
t
e
n
t
ia
l
f
o
r
r
e
a
l
-
w
o
r
l
d
d
e
p
l
o
y
m
e
n
t
i
n
s
e
c
u
r
e
,
s
c
a
l
a
b
l
e
,
a
n
d
i
n
t
e
l
l
i
g
e
n
t
h
e
a
l
t
h
c
a
r
e
e
n
v
i
r
o
n
m
e
n
t
s
,
i
n
c
l
u
d
i
n
g
p
r
i
v
a
c
y
-
a
w
a
re
d
i
g
i
t
a
l
m
a
r
k
e
t
i
n
g
i
n
t
e
g
r
a
t
i
o
n
.
K
ey
w
o
r
d
s
:
C
AM
L
-
E
HDS
C
r
y
p
to
g
r
ap
h
y
Dig
ital m
ar
k
etin
g
Hea
lth
ca
r
e
d
ata
s
ec
u
r
ity
Ma
ch
in
e
lear
n
in
g
L
o
n
g
s
h
o
r
t
-
ter
m
m
e
m
o
r
y
n
etwo
r
k
s
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
Fath
i A
b
d
er
r
ah
m
a
n
e
Natio
n
al
Sch
o
o
l o
f
Ap
p
lied
Sc
ien
ce
s
,
T
h
e
I
n
f
o
r
m
atio
n
T
ec
h
n
o
lo
g
y
L
ab
o
r
ato
r
y
at
C
h
o
u
aib
Do
u
k
k
ali
Un
iv
er
s
ity
Natio
n
al
R
o
ad
No
.
1
,
P.O.
B
o
x
2
9
9
,
E
l Jad
id
a
2
4
0
0
0
,
Mo
r
o
cc
o
E
m
ail:
Ab
d
o
u
.
ft@
g
m
a
il
.
c
o
m
1.
I
NT
RO
D
UCT
I
O
N
I
n
th
is
p
ap
er
,
we
in
tr
o
d
u
ce
c
lu
s
ter
-
b
ased
an
aly
s
is
with
m
a
ch
in
e
lear
n
in
g
f
o
r
e
n
h
an
ce
d
h
ea
lth
ca
r
e
d
ata
s
ec
u
r
ity
(
C
AM
L
-
E
HDS)
,
a
r
o
b
u
s
t
f
r
a
m
ewo
r
k
d
esig
n
ed
to
ad
d
r
ess
th
e
escalatin
g
ch
all
en
g
es
o
f
p
r
o
tectin
g
s
en
s
itiv
e
h
ea
lth
ca
r
e
d
ata
in
an
in
cr
ea
s
in
g
ly
d
ig
ital
lan
d
s
ca
p
e.
T
h
e
d
ig
itizatio
n
o
f
m
ed
i
ca
l
r
ec
o
r
d
s
,
wh
ile
o
f
f
er
in
g
n
u
m
er
o
u
s
b
e
n
ef
its
,
h
as
also
h
eig
h
ten
ed
th
e
r
is
k
o
f
cy
b
er
th
r
ea
ts
,
n
ec
ess
itatin
g
ad
v
an
ce
d
a
n
aly
tical
tech
n
iq
u
es
to
s
af
eg
u
ar
d
p
at
ien
t
in
f
o
r
m
atio
n
[
1
]
.
W
ith
in
th
is
f
r
am
ewo
r
k
,
cr
y
p
to
g
r
ap
h
ic
m
eth
o
d
s
an
d
au
th
o
r
izatio
n
m
ec
h
a
n
is
m
s
ar
e
s
ea
m
less
ly
in
teg
r
ated
to
s
af
eg
u
ar
d
h
ea
lth
ca
r
e
d
ata.
L
ev
e
r
ag
i
n
g
tech
n
iq
u
es
s
u
ch
as
h
o
m
o
m
o
r
p
h
ic
e
n
cr
y
p
tio
n
a
n
d
attr
ib
u
te
-
b
ased
ellip
tic
cu
r
v
e
cr
y
p
t
o
g
r
a
p
h
y
(
E
C
C
)
s
ch
em
es
[
2
]
,
th
is
m
o
d
el
en
s
u
r
es
th
at
s
en
s
itiv
e
in
f
o
r
m
atio
n
r
em
ain
s
en
c
r
y
p
ted
a
n
d
ac
ce
s
s
ib
le
o
n
ly
to
a
u
th
o
r
ized
en
titi
es,
th
u
s
m
itig
atin
g
th
e
r
is
k
o
f
u
n
au
th
o
r
ized
ac
ce
s
s
o
r
d
ata
b
r
ea
ch
es.
T
h
is
ad
d
r
ess
es
a
cr
itical
g
ap
id
en
tifie
d
in
p
r
io
r
r
esear
ch
,
wh
er
e
r
o
b
u
s
t a
cc
ess
co
n
tr
o
l a
n
d
d
ata
p
r
o
tectio
n
m
e
ch
an
is
m
s
ar
e
o
f
ten
lac
k
in
g
.
Fu
r
th
er
m
o
r
e
,
C
AM
L
-
E
HDS
in
co
r
p
o
r
ates
clu
s
ter
-
b
ased
an
aly
s
is
with
s
ta
te
-
of
-
th
e
-
ar
t
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
s
to
u
n
co
v
e
r
h
id
d
e
n
p
atter
n
s
an
d
id
en
tif
y
p
o
ten
tial
s
ec
u
r
ity
th
r
ea
ts
with
in
h
ea
lth
ca
r
e
d
ata.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
S
ec
u
r
in
g
h
ea
lth
c
a
r
e
d
a
ta
a
n
d
o
p
timiz
in
g
d
ig
ita
l m
a
r
ke
tin
g
…
(
F
a
th
i A
b
d
err
a
h
ma
n
e
)
5729
B
y
em
p
lo
y
in
g
s
em
an
tic
clu
s
ter
in
g
,
r
a
n
k
in
g
clu
s
ter
s
,
an
d
c
o
m
p
u
tin
g
s
im
ilar
ity
in
d
ices,
th
e
m
o
d
el
p
r
o
v
id
es
in
v
alu
ab
le
i
n
s
ig
h
ts
in
to
t
h
e
u
n
d
er
ly
in
g
s
tr
u
ctu
r
e
o
f
h
ea
lth
c
ar
e
d
atasets
,
th
er
eb
y
in
c
r
ea
s
in
g
class
if
ier
s
ec
u
r
it
y
an
d
th
r
ea
t
d
etec
tio
n
ca
p
ab
ilit
ies.
T
h
r
o
u
g
h
th
e
u
tili
za
tio
n
o
f
lo
n
g
s
h
o
r
t
-
te
r
m
m
em
o
r
y
n
et
wo
r
k
s
(
L
STM
s
)
[
3
]
an
d
tr
a
n
s
f
er
lear
n
in
g
,
s
p
ec
if
ic
ally
u
s
in
g
p
r
e
-
tr
ai
n
ed
m
o
d
els
lik
e
b
i
d
ir
ec
tio
n
al
en
co
d
er
r
e
p
r
esen
tatio
n
s
f
r
o
m
tr
an
s
f
o
r
m
er
s
(
B
E
R
T
)
,
th
e
m
o
d
el
en
ab
les
h
ea
lth
ca
r
e
o
r
g
an
i
za
tio
n
s
to
class
if
y
s
ec
u
r
ity
th
r
ea
ts
with
p
r
ec
is
io
n
an
d
ef
f
icac
y
,
th
e
r
ef
o
r
e
estab
lis
h
in
g
a
m
o
r
e
s
ec
u
r
e
an
d
r
e
s
ilien
t
h
ea
lth
ca
r
e
d
ata
ec
o
s
y
s
tem
.
C
o
m
p
ar
ed
to
ex
is
tin
g
s
ch
em
es
s
u
ch
as
GH
Z
,
J
,
HZ
,
XZ
Y,
a
n
d
SC
H,
wh
ich
o
f
te
n
lack
p
air
i
n
g
-
f
r
ee
o
p
e
r
atio
n
s
,
E
C
C
-
b
ased
m
eth
o
d
s
,
o
r
k
ey
-
escr
o
w
m
ec
h
an
is
m
s
,
C
AM
L
-
E
HDS
o
f
f
er
s
a
m
o
r
e
co
m
p
r
e
h
en
s
iv
e
s
ec
u
r
it
y
s
o
lu
tio
n
.
I
n
ad
d
itio
n
t
o
its
co
r
e
s
ec
u
r
ity
f
ea
tu
r
es,
th
e
r
esear
ch
m
o
d
e
l
also
o
f
f
er
s
s
ea
m
less
in
teg
r
atio
n
with
d
ig
ital
m
ar
k
etin
g
s
tr
ateg
ies.
B
y
lev
er
ag
in
g
in
s
ig
h
ts
g
ain
e
d
f
r
o
m
u
n
if
ied
th
r
ea
t
d
etec
ti
o
n
an
d
clu
s
ter
-
b
ased
an
aly
s
is
,
C
AM
L
-
E
HD
S
en
ab
les
o
r
g
an
izatio
n
s
to
tailo
r
th
eir
d
ig
ital
m
ar
k
etin
g
s
tr
ateg
ies
ef
f
ec
tiv
ely
.
T
h
r
o
u
g
h
tar
g
eted
ad
v
er
tis
em
en
ts
an
d
p
er
s
o
n
alize
d
c
o
n
ten
t
f
o
r
h
ea
l
th
ca
r
e
p
r
o
d
u
cts
an
d
s
er
v
ices
b
ased
o
n
class
if
ied
d
ata
an
d
d
etec
ted
th
r
ea
ts
,
th
e
m
o
d
el
f
ac
ilit
ates
en
r
ich
ed
en
g
ag
em
en
t
an
d
c
u
s
to
m
er
s
atis
f
ac
tio
n
[
4
]
.
M
o
r
eo
v
er
,
th
e
p
r
o
p
o
s
ed
m
o
d
el
en
s
u
r
es
t
h
at
th
e
d
ata
u
s
ed
f
o
r
d
ig
ital
m
ar
k
etin
g
is
s
ec
u
r
e
an
d
co
m
p
lian
t
with
p
r
iv
ac
y
r
eg
u
latio
n
s
,
th
er
eb
y
p
r
o
v
id
in
g
o
r
g
an
izatio
n
s
with
p
ea
ce
o
f
m
in
d
wh
ile
m
ax
im
izin
g
th
e
ef
f
ec
tiv
en
ess
o
f
th
eir
m
ar
k
etin
g
ef
f
o
r
ts
[
5
]
.
T
h
is
in
t
eg
r
atio
n
,
d
em
o
n
s
tr
atin
g
th
e
p
r
ac
tical
ap
p
licatio
n
o
f
o
u
r
s
ec
u
r
ity
f
r
am
ewo
r
k
,
is
s
h
o
wn
th
r
o
u
g
h
co
m
p
ar
ativ
e
a
n
aly
s
es a
n
d
p
er
f
o
r
m
an
ce
ev
al
u
atio
n
s
in
th
e
r
esu
lts
s
ec
tio
n
,
h
ig
h
lig
h
tin
g
C
AM
L
-
E
HDS
'
s
s
u
p
er
io
r
p
er
f
o
r
m
an
ce
in
m
ain
tain
in
g
d
ata
s
ec
u
r
ity
w
h
ile
o
p
tim
izin
g
m
a
r
k
etin
g
s
tr
ateg
ies
.
T
h
e
m
eth
o
d
o
lo
g
ical
n
o
v
elty
o
f
th
is
s
tu
d
y
lies
in
th
e
in
teg
r
atio
n
o
f
h
o
m
o
m
o
r
p
h
ic
en
cr
y
p
tio
n
,
attr
ib
u
te
-
b
ased
E
C
C
)
an
d
s
em
an
tic
clu
s
ter
in
g
with
m
ac
h
in
e
lear
n
in
g
(
L
STM
an
d
tr
a
n
s
f
er
lear
n
in
g
v
ia
B
E
R
T
)
with
in
a
u
n
if
ie
d
f
r
am
ew
o
r
k
t
ailo
r
ed
f
o
r
h
ea
lth
ca
r
e
d
ata
s
e
cu
r
ity
.
C
AM
L
-
E
HDS
ad
d
r
ess
es
cr
itical
g
ap
s
in
ex
is
tin
g
f
r
am
ewo
r
k
s
,
n
o
ta
b
ly
th
e
lack
o
f
s
ec
u
r
e,
p
r
iv
ac
y
-
co
m
p
lian
t
m
o
d
els
ca
p
ab
le
o
f
r
ea
l
-
tim
e
th
r
ea
t
d
etec
tio
n
an
d
en
c
r
y
p
te
d
d
at
a
p
r
o
ce
s
s
in
g
.
Un
lik
e
c
o
n
v
e
n
tio
n
al
m
o
d
els,
C
AM
L
-
E
HDS
s
im
u
ltan
eo
u
s
ly
en
h
an
ce
s
d
ata
co
n
f
i
d
en
tiality
,
im
p
r
o
v
es
an
o
m
al
y
class
if
icat
io
n
ac
cu
r
ac
y
(
9
6
%),
an
d
s
u
p
p
o
r
ts
s
ec
u
r
e
d
ig
ital
m
ar
k
etin
g
s
tr
ateg
ies
alig
n
ed
with
GDPR
an
d
h
ea
lth
in
s
u
r
an
ce
p
o
r
ta
b
ilit
y
an
d
ac
c
o
u
n
ta
b
ilit
y
ac
t
(
HI
PAA)
.
T
h
e
co
m
p
a
r
ativ
e
r
esu
lts
p
r
e
s
en
ted
in
th
is
p
ap
er
h
ig
h
lig
h
t
its
s
u
p
er
io
r
en
cr
y
p
tio
n
e
f
f
icien
cy
,
r
e
d
u
ce
d
co
m
p
u
tatio
n
al
o
v
er
h
ea
d
,
a
n
d
in
cr
ea
s
ed
r
esil
ien
ce
to
cy
b
er
attac
k
s
.
T
h
ese
c
o
n
tr
ib
u
tio
n
s
o
f
f
er
p
r
o
m
is
in
g
im
p
licatio
n
s
f
o
r
f
u
tu
r
e
a
p
p
lic
atio
n
s
in
s
ec
u
r
e
h
ea
lth
ca
r
e
in
f
r
astru
ctu
r
es,
in
clu
d
in
g
clo
u
d
-
b
ased
s
y
s
tem
s
,
I
o
T
en
v
ir
o
n
m
en
ts
,
an
d
AI
-
d
r
i
v
en
m
ed
ical
d
ata
s
er
v
ices.
2.
L
I
T
E
R
AT
U
RE
R
E
VI
E
W
T
h
e
escalatin
g
d
ig
itizatio
n
o
f
m
ed
ical
r
ec
o
r
d
s
an
d
th
e
in
cr
e
asin
g
ly
s
o
p
h
is
ticated
lan
d
s
ca
p
e
o
f
cy
b
e
r
th
r
ea
ts
h
av
e
u
n
d
er
s
co
r
ed
t
h
e
cr
itical
n
ee
d
f
o
r
r
o
b
u
s
t
h
ea
lth
ca
r
e
d
ata
s
ec
u
r
ity
.
W
h
ile
p
r
io
r
r
esear
ch
h
as
ex
p
lo
r
ed
v
ar
io
u
s
f
ac
ets
o
f
t
h
is
ch
allen
g
e,
s
ig
n
if
ican
t
lim
itat
io
n
s
p
er
s
is
t,
wh
ich
C
AM
L
-
E
HDS
is
d
esig
n
ed
to
ad
d
r
ess
.
Acar
et
a
l.
[
6
]
d
em
o
n
s
tr
ated
th
e
p
r
o
m
is
e
o
f
h
o
m
o
m
o
r
p
h
ic
en
cr
y
p
tio
n
f
o
r
s
a
f
eg
u
ar
d
in
g
s
en
s
itiv
e
m
ed
ical
r
ec
o
r
d
s
.
Ho
wev
e
r
,
th
eir
ap
p
r
o
ac
h
lack
ed
th
e
f
i
n
e
-
g
r
ain
ed
ac
ce
s
s
co
n
tr
o
l
n
ec
ess
ar
y
in
co
llab
o
r
ativ
e
h
ea
lth
ca
r
e
en
v
ir
o
n
m
en
ts
,
leav
in
g
d
ata
v
u
ln
er
a
b
le
to
in
ter
n
al
b
r
ea
ch
es.
Similar
ly
,
I
m
am
et
a
l.
[
7
]
p
r
o
p
o
s
ed
an
attr
ib
u
te
-
b
ased
E
C
C
s
ch
em
e,
b
u
t
th
ey
d
id
n
o
t
ad
eq
u
ately
ad
d
r
ess
th
e
co
m
p
lex
ities
o
f
k
ey
m
an
ag
em
en
t
i
n
d
y
n
am
ic
h
ea
lth
ca
r
e
s
ettin
g
s
,
wh
ich
C
AM
L
-
E
HDS
tack
les
with
its
r
o
b
u
s
t
k
ey
-
escr
o
w
s
y
s
tem
.
C
lu
s
ter
-
b
ased
an
aly
s
is
h
as
also
b
ee
n
e
x
p
lo
r
ed
f
o
r
a
n
o
m
aly
d
etec
tio
n
,
as
ev
id
en
ce
d
b
y
Fes
tag
et
a
l.
[
8
]
,
wh
o
in
v
esti
g
ated
s
em
an
tic
clu
s
ter
in
g
alg
o
r
ith
m
s
.
Ho
wev
er
,
th
eir
wo
r
k
d
i
d
n
o
t
in
teg
r
ate
r
ea
l
-
tim
e
m
ac
h
in
e
lear
n
in
g
f
o
r
d
y
n
am
ic
th
r
ea
t
d
etec
tio
n
,
a
c
r
itical
co
m
p
o
n
en
t
o
f
C
AM
L
-
E
HDS.
Pra
s
ad
et
a
l.
[
9
]
ex
p
lo
r
ed
s
im
ilar
ity
-
b
ased
clu
s
ter
in
g
,
b
u
t
th
eir
a
p
p
r
o
ac
h
lack
ed
th
e
tem
p
o
r
al
a
n
aly
s
is
ca
p
ab
ilit
ies
p
r
o
v
id
e
d
b
y
C
AM
L
-
E
HDS’
s
L
STM
-
b
ased
co
m
p
o
n
en
t.
Mo
r
eo
v
er
,
B
alh
ar
eth
an
d
I
ly
as
[
1
0
]
u
tili
z
ed
C
NNs
f
o
r
s
ec
u
r
ity
b
r
ea
ch
d
etec
tio
n
in
m
ed
ical
im
ag
in
g
,
an
d
R
ajk
o
m
ar
et
a
l.
[
1
1
]
em
p
l
o
y
ed
L
STM
s
f
o
r
t
em
p
o
r
al
p
atter
n
r
ec
o
g
n
itio
n
in
elec
tr
o
n
ic
h
ea
lth
r
ec
o
r
d
s
;
h
o
we
v
er
,
th
ese
s
tu
d
ies
f
o
cu
s
ed
o
n
is
o
lated
a
s
p
ec
ts
o
f
d
ata
s
ec
u
r
ity
a
n
d
d
id
n
o
t
o
f
f
er
a
co
m
p
r
eh
e
n
s
iv
e
f
r
am
ew
o
r
k
th
a
t in
teg
r
ates m
u
ltip
le
s
ec
u
r
ity
l
ay
er
s
.
C
AM
L
-
E
HDS,
in
co
n
tr
ast,
co
m
b
in
es
cr
y
p
to
g
r
ap
h
ic
m
eth
o
d
s
,
ad
v
an
ce
d
clu
s
ter
in
g
,
an
d
s
o
p
h
is
ticated
m
ac
h
in
e
lear
n
in
g
,
in
clu
d
in
g
b
o
th
L
STM
an
d
B
E
R
T
,
t
o
p
r
o
v
id
e
a
m
u
lti
-
lay
er
ed
s
ec
u
r
ity
ap
p
r
o
ac
h
.
Fu
r
th
er
m
o
r
e
,
a
s
ig
n
i
f
ican
t
g
a
p
ex
is
ts
in
th
e
liter
atu
r
e
r
e
g
a
r
d
in
g
t
h
e
in
teg
r
atio
n
o
f
s
ec
u
r
ity
m
ea
s
u
r
es
with
d
ig
ital
m
ar
k
etin
g
s
tr
ateg
ies.
P
r
io
r
r
esear
ch
h
as
lar
g
ely
o
v
er
l
o
o
k
ed
th
is
in
ter
s
ec
tio
n
.
C
AM
L
-
E
HDS
ad
d
r
ess
es
th
is
g
ap
b
y
lev
er
ag
in
g
in
s
ig
h
ts
f
r
o
m
u
n
if
ied
th
r
ea
t
d
etec
ti
o
n
an
d
clu
s
ter
-
b
ased
an
aly
s
is
to
o
p
tim
ize
d
ig
ital
m
ar
k
etin
g
e
f
f
o
r
ts
wh
ile
en
s
u
r
in
g
d
ata
s
ec
u
r
ity
co
m
p
lia
n
ce
.
B
y
tailo
r
in
g
tar
g
eted
ad
v
er
tis
em
en
ts
an
d
p
er
s
o
n
alize
d
co
n
ten
t
b
ased
o
n
class
if
ied
d
ata
an
d
d
ete
cted
th
r
ea
ts
,
C
AM
L
-
E
HDS
im
p
r
o
v
es
cu
s
to
m
e
r
en
g
ag
em
e
n
t
wh
ile
ad
h
er
in
g
t
o
s
tr
in
g
en
t
p
r
iv
ac
y
r
eg
u
latio
n
s
.
T
h
is
in
teg
r
atio
n
o
f
s
ec
u
r
ity
an
d
m
ar
k
etin
g
,
co
u
p
led
with
its
r
o
b
u
s
t
cr
y
p
to
g
r
ap
h
ic
an
d
m
ac
h
i
n
e
lear
n
in
g
co
m
p
o
n
en
ts
,
d
is
tin
g
u
is
h
es
C
AM
L
-
E
HDS
a
s
a
co
m
p
r
eh
e
n
s
iv
e
an
d
in
n
o
v
ativ
e
s
o
lu
tio
n
,
s
u
r
p
ass
in
g
th
e
lim
itatio
n
s
o
f
p
r
e
v
io
u
s
m
o
d
els.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
15
,
No
.
6
,
Decem
b
e
r
20
25
:
5
7
2
8
-
5
7
4
5
5730
3.
ARCH
I
T
E
C
T
UR
E
O
F
CAML
-
E
H
DS MO
D
E
L
C
AM
L
-
E
HDS
ar
ch
itectu
r
e
r
ep
r
esen
ts
a
m
eticu
lo
u
s
ly
en
g
in
ee
r
ed
f
o
r
tr
ess
,
d
esig
n
ed
to
p
r
o
v
i
d
e
u
n
p
ar
alleled
h
ea
lth
ca
r
e
d
ata
s
ec
u
r
ity
wh
ile
s
im
u
ltan
eo
u
s
ly
o
p
tim
izin
g
d
ig
ital
m
ar
k
et
in
g
s
tr
ateg
ies.
T
h
e
p
r
o
ce
s
s
in
itiates
with
a
f
o
r
tifie
d
d
ata
co
llectio
n
an
d
p
r
ep
r
o
ce
s
s
in
g
p
h
ase,
en
s
u
r
in
g
th
e
s
ec
u
r
e
g
ath
er
i
n
g
an
d
m
eticu
lo
u
s
p
r
ep
a
r
atio
n
o
f
h
e
alth
ca
r
e
d
ata,
th
er
e
b
y
estab
lis
h
in
g
an
im
p
r
e
g
n
ab
le
f
o
u
n
d
a
tio
n
f
o
r
s
u
b
s
eq
u
en
t
an
aly
s
es
[
1
2
]
.
Fo
llo
win
g
th
is
,
a
d
u
al
-
lay
er
ed
cr
y
p
to
g
r
a
p
h
ic
s
h
ield
is
d
ep
lo
y
ed
,
in
co
r
p
o
r
a
tin
g
h
o
m
o
m
o
r
p
h
ic
en
cr
y
p
tio
n
an
d
E
C
C
-
b
ased
au
th
o
r
izatio
n
.
T
h
is
r
o
b
u
s
t
co
m
b
in
atio
n
g
u
ar
an
tees
im
p
e
n
etr
ab
le
d
ata
en
c
r
y
p
tio
n
an
d
e
n
f
o
r
ce
s
s
tr
in
g
en
t,
g
r
a
n
u
lar
ac
ce
s
s
p
o
licies,
ef
f
ec
tiv
el
y
th
war
tin
g
u
n
au
th
o
r
ized
ac
c
ess
an
d
m
itig
atin
g
p
o
ten
tial
b
r
ea
c
h
es.
Nex
t,
C
A
ML
-
E
HDS
em
p
lo
y
s
an
ad
v
a
n
ce
d
clu
s
ter
-
b
ased
an
aly
s
is
,
l
ev
er
ag
in
g
s
em
an
tic
clu
s
ter
in
g
,
r
an
k
in
g
clu
s
ter
s
with
p
r
ec
is
io
n
,
co
m
p
u
tin
g
s
im
ilar
ity
in
d
ices,
an
d
ex
ec
u
tin
g
d
o
m
ain
tr
an
s
f
o
r
m
atio
n
s
[
1
3
]
.
T
h
is
s
o
p
h
is
ticated
p
r
o
ce
s
s
u
n
co
v
er
s
h
id
d
en
p
atter
n
s
an
d
b
o
ls
ter
s
class
if
ier
s
ec
u
r
ity
,
tr
an
s
f
o
r
m
in
g
r
aw
d
ata
in
t
o
ac
t
io
n
ab
le
in
tellig
en
ce
.
T
h
e
m
o
d
el'
s
an
aly
tical
p
r
o
wess
is
f
u
r
th
er
am
p
lifie
d
b
y
a
p
o
wer
f
u
l
m
ac
h
i
n
e
lear
n
i
n
g
c
o
r
e.
I
n
d
iv
i
d
u
al
m
o
d
els,
in
clu
d
in
g
th
e
im
p
r
ess
iv
e
L
STM
s
an
d
tr
an
s
f
er
lear
n
in
g
m
o
d
els,
u
n
d
e
r
g
o
r
ig
o
r
o
u
s
tr
ain
in
g
to
class
if
y
s
ec
u
r
ity
th
r
ea
ts
with
u
n
m
atch
ed
ac
cu
r
ac
y
[
1
4
]
.
A
p
iv
o
tal
m
o
d
el
f
u
s
io
n
s
tag
e
th
en
in
teg
r
ates
th
e
o
u
t
p
u
ts
o
f
th
ese
m
o
d
els
th
r
o
u
g
h
weig
h
te
d
av
er
a
g
in
g
an
d
en
s
em
b
le
p
r
ed
ictio
n
,
g
e
n
er
atin
g
a
f
i
n
al,
e
x
ce
p
tio
n
ally
r
o
b
u
s
t
o
u
tp
u
t.
T
h
is
f
u
s
io
n
cr
ea
tes
a
s
y
n
er
g
is
tic
d
ef
en
s
e,
ex
ce
e
d
in
g
th
e
ca
p
a
b
ilit
ies
o
f
an
y
s
in
g
le
m
o
d
el
an
d
p
r
o
v
id
i
n
g
a
u
n
if
ied
,
im
p
e
n
et
r
ab
le
th
r
ea
t
d
etec
tio
n
s
y
s
tem
.
Fin
ally
,
C
AM
L
-
E
HD
S
s
ea
m
less
ly
in
teg
r
ate
s
d
ig
ital
m
ar
k
etin
g
s
tr
ateg
ies,
lev
er
ag
in
g
t
h
e
alar
m
in
g
in
s
ig
h
ts
d
er
iv
ed
f
r
o
m
u
n
if
ied
t
h
r
ea
t
d
etec
tio
n
an
d
clu
s
ter
-
b
ased
an
aly
s
is
.
T
h
is
in
teg
r
atio
n
en
ab
les
th
e
i
m
p
lem
en
tatio
n
o
f
tar
g
eted
ad
v
er
tis
em
en
ts
an
d
p
er
s
o
n
alize
d
co
n
ten
t,
en
s
u
r
i
n
g
b
o
th
m
ar
k
etin
g
ef
f
ec
tiv
en
es
s
an
d
u
n
wav
er
i
n
g
co
m
p
lian
ce
with
d
ata
s
ec
u
r
ity
r
eg
u
latio
n
s
[
1
5
]
.
T
h
is
co
m
p
r
e
h
en
s
iv
e
ar
ch
itectu
r
e
,
with
its
m
u
lti
-
lay
er
ed
d
ef
en
s
es
an
d
in
t
eg
r
ated
in
tellig
en
ce
,
estab
lis
h
es C
AM
L
-
E
HDS
as a
p
ar
ag
o
n
o
f
r
o
b
u
s
t a
n
d
s
ec
u
r
e
h
ea
lth
ca
r
e
d
ata
m
a
n
ag
em
e
n
t.
T
h
e
o
v
er
all
ar
ch
itectu
r
e
o
f
th
e
C
AM
L
-
E
HDS
m
o
d
el
is
v
i
s
u
ally
s
u
m
m
ar
ized
in
Fig
u
r
e
1
.
T
h
is
f
ig
u
r
e
illu
s
tr
ates
th
e
en
d
-
to
-
en
d
p
ip
elin
e
o
f
th
e
f
r
am
ewo
r
k
,
in
clu
d
in
g
d
ata
p
r
ep
r
o
ce
s
s
in
g
,
cr
y
p
to
g
r
ap
h
ic
m
eth
o
d
s
,
clu
s
ter
-
b
ased
an
aly
s
is
,
m
ac
h
in
e
lear
n
in
g
in
teg
r
atio
n
,
u
n
if
ied
th
r
ea
t
d
etec
tio
n
,
an
d
d
ig
ital
m
ar
k
etin
g
ap
p
licatio
n
s
.
T
h
e
d
iag
r
am
h
ig
h
lig
h
ts
h
o
w
ea
ch
co
m
p
o
n
en
t
i
n
ter
ac
ts
to
en
h
an
ce
h
ea
lt
h
ca
r
e
d
ata
s
ec
u
r
ity
wh
ile
s
u
p
p
o
r
tin
g
p
r
iv
ac
y
-
co
m
p
lian
t
m
ar
k
etin
g
s
tr
ateg
ies.
Fig
u
r
e
1
.
C
AM
L
-
E
HDS
m
o
d
e
l’
s
ar
ch
itectu
r
e
f
o
r
s
ec
u
r
in
g
h
e
alth
ca
r
e
d
ata
an
d
o
p
tim
izin
g
d
ig
ital m
ar
k
etin
g
4.
M
E
T
H
O
D
4
.
1
.
Da
t
a
c
o
llect
io
n a
nd
ex
perim
ent
a
l set
up
T
h
e
h
ea
lth
ca
r
e
d
ata,
en
c
o
m
p
a
s
s
in
g
p
atien
t
in
f
o
r
m
atio
n
o
n
v
ar
io
u
s
d
is
ea
s
es
s
o
u
r
ce
d
f
r
o
m
h
ea
lth
ca
r
e
web
s
ites
,
wa
s
m
eticu
lo
u
s
ly
g
ath
er
ed
.
Ho
we
v
er
,
a
cr
u
cia
l
b
ias
an
aly
s
is
r
ev
ea
led
p
o
ten
tial
d
em
o
g
r
a
p
h
ic
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
S
ec
u
r
in
g
h
ea
lth
c
a
r
e
d
a
ta
a
n
d
o
p
timiz
in
g
d
ig
ita
l m
a
r
ke
tin
g
…
(
F
a
th
i A
b
d
err
a
h
ma
n
e
)
5731
o
v
er
r
e
p
r
esen
tatio
n
with
in
th
e
d
ataset.
W
e
em
p
lo
y
e
d
d
ata
au
g
m
en
tatio
n
,
f
air
n
ess
-
awa
r
e
m
ac
h
in
e
lea
r
n
in
g
alg
o
r
ith
m
s
,
an
d
s
en
s
itiv
ity
an
aly
s
es,
th
o
u
g
h
we
ac
k
n
o
wled
g
e
th
at
in
h
er
en
t
b
iases
m
ay
p
er
s
is
t.
Fu
tu
r
e
wo
r
k
will
f
o
cu
s
o
n
e
x
p
an
d
in
g
d
a
taset
d
iv
er
s
ity
an
d
e
x
p
lo
r
i
n
g
ad
v
a
n
ce
d
b
ias
m
itig
atio
n
t
o
en
s
u
r
e
f
air
a
n
d
g
en
er
aliza
b
le
m
o
d
el
p
er
f
o
r
m
an
ce
.
C
AM
L
-
E
HDS
m
o
d
el
in
co
r
p
o
r
ates
s
tr
ateg
ies
to
m
ain
tain
co
m
p
u
tatio
n
a
l
ef
f
icien
cy
,
th
e
f
r
am
ewo
r
k
em
p
lo
y
s
o
p
tim
ized
p
r
e
p
r
o
ce
s
s
in
g
tech
n
iq
u
es
to
r
ed
u
ce
d
ata
d
im
en
s
io
n
ality
an
d
co
m
p
lex
ity
,
a
n
d
is
d
esig
n
e
d
to
lev
er
ag
e
p
ar
allel
p
r
o
ce
s
s
in
g
to
h
a
n
d
le
lar
g
e
d
atasets
with
o
u
t
s
ig
n
if
ican
t
p
er
f
o
r
m
an
ce
d
eg
r
ad
atio
n
.
Fo
r
th
e
im
p
lem
en
tatio
n
an
d
test
in
g
o
f
th
e
r
esear
ch
m
o
d
el,
Py
th
o
n
was
s
elec
ted
d
u
e
to
its
v
er
s
atility
an
d
th
e
ex
te
n
s
iv
e
r
an
g
e
o
f
lib
r
a
r
ies
[
1
6
]
,
in
clu
d
in
g
T
en
s
o
r
Flo
w,
s
cik
it
-
lear
n
,
Nu
m
Py
,
an
d
Pan
d
as.
T
h
ese
lib
r
ar
ies
ar
e
cr
u
cial
f
o
r
d
ev
el
o
p
in
g
an
d
test
in
g
m
ac
h
in
e
lear
n
in
g
m
o
d
els.
I
n
te
r
m
s
o
f
s
im
u
latio
n
p
ar
am
eter
s
,
s
y
m
m
etr
ic
k
ey
en
cr
y
p
tio
n
was
em
p
lo
y
ed
to
s
ec
u
r
e
d
ata
d
u
r
i
n
g
ex
p
er
im
en
ts
[
1
7
]
,
h
ig
h
lig
h
tin
g
th
e
cr
itical
r
o
le
o
f
ef
f
icien
t k
ey
m
an
a
g
em
en
t in
m
ain
tain
in
g
d
ata
s
ec
u
r
ity
.
T
h
e
d
ataset
co
n
s
is
t
s
o
f
m
ed
ical
d
ata
with
a
m
ea
n
v
al
u
e
o
f
5
0
7
k
an
d
a
s
tan
d
a
r
d
d
ev
iatio
n
o
f
1
2
.
5
k
.
Fo
r
h
ea
lth
ca
m
p
I
Ds,
th
e
m
ea
n
is
ca
lcu
lated
as
6
.
5
7
k
with
a
s
tan
d
ar
d
d
e
v
iatio
n
o
f
1
3
.
2
k
.
Similar
ly
,
f
o
r
p
atien
t
d
ata,
th
e
m
ea
n
is
3
8
7
k
with
a
s
tan
d
ar
d
d
e
v
iatio
n
o
f
3
9
.
6
k
.
T
h
e
o
v
er
all
d
escr
ip
tio
n
o
f
th
e
d
ataset
is
d
etailed
in
T
ab
le
1
.
−
h
ttp
s
:
//w
w
w
.
ka
g
g
le.
co
m/d
a
ta
s
ets/
meh
r
a
d
a
r
ia
/co
vid
1
9
-
lu
n
g
-
ct
-
s
ca
n
s
;
−
h
ttp
s
:
//w
w
w
.
ka
g
g
le.
co
m/d
a
ta
s
ets/
u
ciml/p
ima
-
in
d
ia
n
s
-
d
ia
b
et
es
-
d
a
ta
b
a
s
e
;
−
h
ttp
s
:
//w
w
w
.
ka
g
g
le.
co
m/d
a
ta
s
ets/
ma
th
ch
i/d
ia
b
etes
-
d
a
ta
-
s
et
;
−
h
ttp
s
:
//w
w
w
.
ka
g
g
le.
co
m/co
mp
etitio
n
s
/d
ia
b
etes
-
cla
s
s
ifica
tio
n
/d
a
ta
;
−
h
ttp
s
:
//w
w
w
.
ka
g
g
le.
co
m/d
a
ta
s
ets/
s
a
u
r
a
b
h
0
0
0
0
7
/d
ia
b
etescsv
;
−
h
ttp
s
:
//w
w
w
.
ka
g
g
le.
co
m/co
d
e/
p
a
u
ltimo
th
ymo
o
n
ey
/p
r
ed
ict
-
d
i
a
b
etes
-
fr
o
m
-
med
ica
l
-
r
ec
o
r
d
s
/d
a
ta
;
−
h
ttp
s
:
//w
w
w
.
ka
g
g
le.
co
m/d
a
ta
s
ets/
ku
ma
r
g
h
/p
ima
in
d
ia
n
s
d
ia
b
etescsv
;
−
h
ttp
s
:
//w
w
w
.
ka
g
g
le.
co
m/d
a
ta
s
ets/
r
i
s
ch
a
n
/d
ia
b
etes
-
d
a
ta
s
et
;
−
h
ttp
s
:
//w
w
w
.
ka
g
g
le.
co
m/d
a
ta
s
ets/
jill
a
n
is
o
fttech
/d
ia
b
etes
-
d
is
ea
s
e
-
u
p
d
a
te
d
-
d
a
ta
s
et
;
−
h
ttp
s
:
//w
w
w
.
ka
g
g
le.
co
m/co
d
e/
ma
th
ch
i/d
ia
g
n
o
s
tic
-
a
-
p
a
tien
t
-
has
-
d
ia
b
etes/d
a
ta
;
−
h
ttp
s
:
//w
w
w
.
ka
g
g
le.
co
m/d
a
ta
s
ets/
p
a
u
ltimo
th
ymo
o
n
ey
/b
lo
o
d
-
ce
lls
;
−
h
ttp
s
:
//w
w
w
.
ka
g
g
le.
co
m/d
a
ta
s
ets/
d
r
a
a
s
la
n
/b
lo
o
d
-
ce
ll
-
d
etec
ti
o
n
-
d
a
ta
s
et
;
−
h
ttp
s
:
//w
w
w
.
ka
g
g
le.
co
m/co
mp
etitio
n
s
/3
md
3
0
7
0
-
d
lmi/d
a
ta
;
T
ab
le
1
.
Data
s
et
d
is
tr
ib
u
tio
n
D
a
t
a
s
e
t
M
e
a
n
S
t
a
n
d
a
r
d
d
e
v
i
a
t
i
o
n
H
e
a
l
t
h
c
a
m
p
u
s
5
0
7
k
1
2
.
5
C
o
v
i
d
-
19
3
8
7
k
3
9
.
6
Lu
n
g
5
8
5
6
1
.
2
8
H
e
a
r
t
5
4
.
4
9
.
0
7
I
r
i
s
7
5
.
5
4
3
.
3
4
.
2
.
Da
t
a
clea
nin
g
a
nd
f
ilte
r
ing
Po
s
t
-
d
ata
co
llectio
n
,
a
co
m
p
r
eh
en
s
iv
e
clea
n
in
g
an
d
f
ilter
in
g
p
r
o
to
c
o
l
was
im
p
lem
en
ted
to
en
s
u
r
e
d
ata
in
teg
r
ity
a
n
d
c
o
n
s
is
ten
cy
.
T
h
is
p
r
o
to
co
l
e
n
co
m
p
ass
ed
th
e
r
em
o
v
al
o
f
ir
r
ele
v
an
t
in
f
o
r
m
atio
n
,
er
r
o
r
co
r
r
ec
tio
n
,
an
d
f
o
r
m
at
s
tan
d
a
r
d
izatio
n
.
Giv
e
n
th
e
cr
itical
n
atu
r
e
o
f
m
is
s
in
g
d
ata
in
h
ea
lth
ca
r
e
an
aly
tics
,
a
m
u
ltifa
ce
ted
im
p
u
tatio
n
s
tr
ateg
y
was
ad
o
p
ted
.
Fo
r
n
u
m
er
ica
l
v
ar
iab
les,
s
u
ch
as
p
atien
t
ag
e,
m
ea
n
im
p
u
tatio
n
was
u
tili
ze
d
to
p
r
o
v
id
e
s
tatis
ti
ca
lly
r
ep
r
esen
tativ
e
v
alu
es.
Fo
r
ca
teg
o
r
ical
v
ar
iab
les,
in
clu
d
in
g
p
atien
t
g
en
d
er
,
m
o
d
e
im
p
u
tatio
n
was
ap
p
lied
,
ass
ig
n
in
g
th
e
m
o
s
t
f
r
eq
u
en
t
ca
teg
o
r
y
.
I
n
in
s
tan
ce
s
wh
er
e
m
is
s
in
g
d
ata
was
d
ee
m
ed
a
n
aly
tically
s
ig
n
if
ic
an
t
o
r
wh
er
e
s
im
p
le
im
p
u
ta
tio
n
co
u
ld
in
t
r
o
d
u
ce
s
u
b
s
tan
tial
b
ias,
k
-
n
ea
r
est
n
eig
h
b
o
r
s
(
k
-
NN)
im
p
u
tatio
n
was
u
tili
ze
d
,
lev
er
a
g
in
g
s
im
ilar
d
ata
p
o
in
t
v
alu
es
to
esti
m
ate
m
is
s
in
g
v
alu
es.
T
h
is
ap
p
r
o
ac
h
was
s
elec
ted
to
m
in
im
ize
d
ata
lo
s
s
an
d
p
r
eser
v
e
d
ataset
in
te
g
r
ity
,
p
ar
ticu
la
r
ly
in
ca
s
es
wh
er
e
m
is
s
in
g
d
ata
p
atter
n
s
co
u
l
d
y
ield
v
alu
ab
le
in
s
ig
h
ts
.
T
o
ad
d
r
ess
th
e
is
s
u
e
o
f
im
b
alan
ce
d
d
ata,
wh
er
e
ce
r
tain
s
ec
u
r
ity
th
r
ea
t
ca
teg
o
r
ies
wer
e
less
f
r
eq
u
en
t
th
an
o
th
er
s
,
t
h
e
Sy
n
th
etic
Min
o
r
ity
Ov
er
-
s
am
p
lin
g
T
ec
h
n
iq
u
e
(
SMOT
E
)
was
s
u
b
s
eq
u
en
tly
ap
p
lied
.
SMOT
E
was
ch
o
s
en
to
g
en
e
r
ate
s
y
n
th
etic
i
n
s
tan
ce
s
o
f
th
e
m
i
n
o
r
it
y
class
es,
cr
ea
tin
g
a
m
o
r
e
b
alan
ce
d
d
ataset
f
o
r
m
o
d
el
tr
ain
i
n
g
.
T
h
is
tech
n
iq
u
e
h
elp
s
p
r
e
v
en
t
th
e
m
o
d
el
f
r
o
m
b
ein
g
b
iased
to
war
d
s
th
e
m
aj
o
r
ity
class
an
d
im
p
r
o
v
es
its
ab
ilit
y
to
ac
cu
r
ately
d
etec
t
r
ar
e
b
u
t
cr
itical
s
ec
u
r
ity
th
r
ea
ts
.
Fu
r
th
er
m
o
r
e,
tex
tu
al
d
ata
u
n
d
er
wen
t
to
k
e
n
izatio
n
,
lo
wer
ca
s
in
g
,
s
to
p
wo
r
d
r
e
m
o
v
al,
s
tem
m
in
g
,
an
d
lem
m
atiza
tio
n
[
1
8
]
,
p
r
ep
ar
in
g
it f
o
r
ef
f
ec
tiv
e
an
d
r
eliab
le
an
aly
s
is
[
1
9
]
.
4
.
3
.
Cry
pt
o
g
ra
ph
ic
m
e
t
ho
ds
a
nd
a
utho
riza
t
io
n
4
.
3
.
1
.
Cry
pto
g
r
a
ph
ic
pro
ce
s
s
es a
nd
k
ey
ma
na
g
ement
T
h
e
C
AM
L
-
E
HDS
m
o
d
el
im
p
lem
en
ts
a
r
o
b
u
s
t
cr
y
p
t
o
g
r
ap
h
ic
p
r
o
to
c
o
l
to
s
af
eg
u
ar
d
h
ea
lth
ca
r
e
d
ata,
f
ea
tu
r
in
g
k
e
y
elem
en
ts
s
u
ch
as
th
e
d
ata
o
wn
er
(
DO)
,
k
ey
g
en
er
atio
n
ce
n
ter
(
KGC),
clo
u
d
s
to
r
ag
e
(
C
S),
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
15
,
No
.
6
,
Decem
b
e
r
20
25
:
5
7
2
8
-
5
7
4
5
5732
d
ec
r
y
p
tio
n
s
er
v
er
(
DS)
,
an
d
d
ata
r
ec
eiv
er
(
DR
)
.
T
h
e
DO
o
v
er
s
ee
s
d
ata
ex
ch
an
g
e
an
d
e
n
cr
y
p
tio
n
f
o
r
cl
o
u
d
s
to
r
ag
e,
en
s
u
r
in
g
th
e
s
ec
u
r
e
tr
an
s
f
er
o
f
p
atie
n
t
d
ata.
T
h
e
KGC
co
o
r
d
in
ates
k
e
y
g
en
er
a
tio
n
an
d
in
teg
r
ates
p
r
iv
ate
k
e
y
s
b
ased
o
n
u
s
er
a
ttrib
u
tes,
f
ac
ilit
atin
g
en
cr
y
p
te
d
in
f
o
r
m
atio
n
e
x
ch
an
g
e
with
in
th
e
clo
u
d
[
2
0
]
.
Ser
v
in
g
as
a
s
em
i
-
tr
u
s
ted
en
t
ity
,
th
e
C
S
en
ab
les
d
ata
s
h
ar
in
g
an
d
s
to
r
ag
e
wh
ile
g
e
n
er
ati
n
g
s
ec
r
et
k
ey
s
f
o
r
u
s
er
s
[
2
1
]
.
T
h
e
DS
en
ab
les
d
e
cr
y
p
tio
n
o
f
tr
a
n
s
m
itted
in
f
o
r
m
atio
n
,
d
eter
m
in
in
g
d
ec
r
y
p
tio
n
ca
p
a
b
ilit
ies
at
th
e
r
ec
eiv
er
'
s
en
d
[
2
2
]
.
Me
an
w
h
ile,
th
e
DR
en
s
u
r
es
s
ec
u
r
e
d
ata
an
aly
s
is
b
y
in
teg
r
atin
g
at
tr
ib
u
te
s
ets
d
u
r
in
g
d
ec
r
y
p
tio
n
an
d
ac
co
m
m
o
d
atin
g
r
eso
u
r
ce
c
o
n
s
tr
ain
ts
f
o
r
li
g
h
t
weig
h
t m
o
b
ile
d
e
v
ices
[
2
3
]
.
4
.
3
.
2
.
H
o
m
o
m
o
rphic
e
ncry
ptio
n o
f
CAML
-
E
H
D
S m
o
del
a.
Setu
p
T
h
e
C
AM
L
-
E
HDS
m
o
d
el
u
ti
lizes
attr
ib
u
te
-
b
ased
E
C
C
to
s
af
eg
u
ar
d
h
ea
lth
ca
r
e
d
ata,
al
o
n
g
s
id
e
a
k
ey
-
b
ased
ap
p
r
o
ac
h
an
d
a
lig
h
tweig
h
t
m
o
d
el
tailo
r
ed
f
o
r
at
tr
ib
u
te
-
b
ased
E
C
C
p
r
o
ce
s
s
in
g
[
2
4
]
.
T
h
e
s
ec
u
r
ity
f
ea
tu
r
es
en
co
m
p
ass
d
o
m
ain
-
s
p
ec
if
ic
f
ea
tu
r
e
p
a
r
am
eter
s
,
le
v
er
ag
in
g
an
ellip
tical
cu
r
v
e
m
o
d
el
f
o
r
esti
m
atin
g
an
d
co
m
p
u
tin
g
p
u
b
lic
p
ar
am
eter
s
.
I
n
th
is
s
etu
p
,
th
e
C
S
is
in
teg
r
ated
with
th
e
KGC
[
2
5
]
,
wh
er
e
a
r
an
d
o
m
n
u
m
b
er
is
co
m
p
u
ted
as
∈
∗
with
th
e
au
th
o
r
izatio
n
o
f
∈
,
wh
er
e
ω
r
ep
r
esen
ted
as
th
e
attr
ib
u
te
s
s
et
f
o
r
th
e
au
th
o
r
izatio
n
o
f
=
{
1
,
2
,
3
,
…
…
.
,
}
,
=
1
to
an
d
∈
.
W
ith
th
e
s
et
u
p
o
f
KGC
th
e
s
ec
r
et
ley
f
o
r
t
h
e
m
aster
is
co
m
p
u
te
d
as k
∈
Z
*
q
with
E
C
C
f
o
r
th
e
co
m
p
u
tatio
n
o
f
th
e
p
u
b
lic
k
ey
s
tated
as
:
=
.
i.e
.
,
{
=
,
=
.
}
(
1
)
C
S Setu
p
: T
h
e
m
aster
s
ec
r
et
k
ey
is
elec
ted
b
ased
o
n
c
∈
Z
*
q
an
d
th
e
p
u
b
lic
k
ey
is
esti
m
ated
as
:
=
.
;
{
=
,
=
.
}
(
2
)
T
h
e
p
u
b
lic
p
ar
am
eter
o
u
tp
u
t is d
en
o
ted
as
=
{
,
,
}
.
b.
E
n
cr
y
p
tio
n
an
d
re
-
e
n
cr
y
p
tio
n
W
ith
in
DO
,
d
ata
i
s
u
p
lo
ad
ed
f
o
r
m
ess
ag
e
s
h
ar
in
g
an
d
ex
ec
u
tio
n
,
u
s
in
g
a
s
tr
u
ctu
r
e
wit
h
d
ef
in
ed
attr
ib
u
tes
f
o
r
au
th
o
r
izatio
n
,
d
en
o
ted
as
.
T
h
is
p
h
ase
in
clu
d
es
co
m
p
u
tin
g
a
n
d
esti
m
atin
g
th
e
en
cr
y
p
te
d
m
ess
ag
e
f
o
r
th
e
d
ata
in
p
u
t
.
T
h
e
ac
ce
s
s
tr
ee
is
r
ep
r
esen
ted
as
T
,
with
th
e
m
ess
a
g
e
en
cr
y
p
tio
n
o
f
u
s
in
g
a
r
an
d
o
m
n
u
m
b
er
est
im
ated
as
∈
∗
,
f
o
r
t
h
e
en
cr
y
p
tio
n
an
d
in
teg
r
ity
o
f
s
y
m
m
etr
ic
d
ata
co
m
p
u
tatio
n
[
2
6
]
.
T
h
e
C
S
ex
ec
u
tio
n
p
r
o
ce
s
s
in
v
o
lv
es
d
is
tr
ib
u
tin
g
an
d
s
to
r
in
g
cip
h
er
tex
t
d
ata
f
o
r
th
e
d
ata
g
en
er
ated
b
y
th
e
DO.
T
h
e
ci
p
h
er
tex
t
d
ata
p
a
r
am
eter
s
ar
e
ca
lcu
lated
b
ased
o
n
th
e
in
p
u
t
d
ata
an
d
th
e
C
S
m
aster
k
ey
cip
h
e
r
tex
t
[
2
7
]
,
wi
th
th
e
m
aster
k
ey
g
en
er
ate
d
as
(
3
)
:
=
(
,
,
,
,
=
(
,
)
)
(
3
)
Her
e
.
=
(
,
)
.
c.
Key
g
en
er
atio
n
,
k
e
y
u
p
d
ate
an
d
d
ec
r
y
p
tio
n
T
h
e
k
e
y
g
e
n
er
atio
n
p
h
ase
ce
n
ter
s
o
n
p
r
o
d
u
cin
g
th
e
KGC
k
ey
K,
ass
o
ciate
d
with
t
h
e
attr
i
b
u
te
s
et
S
f
o
r
th
e
r
ec
eiv
er
d
ata.
T
h
e
p
r
iv
ate
k
ey
f
o
r
th
e
KGC
[
2
8
]
,
d
er
i
v
ed
u
s
in
g
a
r
an
d
o
m
n
u
m
b
er
r
∈
Z
*
q
,
is
ex
p
r
ess
ed
as
(
4
)
:
=
.
,
∀
∈
(
4
)
I
n
th
is
eq
u
atio
n
,
th
e
r
an
d
o
m
n
u
m
b
er
g
en
er
ate
d
f
o
r
th
e
ai
t
ak
es
in
to
ac
co
u
n
t
t
h
e
s
etu
p
p
h
ases
.
T
h
e
C
AM
L
-
E
HDS
m
o
d
el
co
m
p
r
is
es
o
f
th
r
ee
p
h
ases
s
u
ch
as
K
GC
,
C
S
an
d
DR
f
o
r
th
e
esti
m
atio
n
o
f
C
S
an
d
KG
C
.
T
h
e
k
ey
g
e
n
er
atio
n
o
f
th
e
c
o
m
p
o
n
en
ts
co
m
p
r
is
es o
f
th
e
f
o
llo
win
g
s
tep
s
th
at
ar
e
s
tated
as b
elo
w:
−
I
n
i
t
i
a
ll
y
,
t
h
e
s
ec
r
e
t
k
e
y
is
g
e
n
e
r
a
t
e
d
a
s
a
n
d
w
i
t
h
t
h
e
g
e
n
e
r
at
i
o
n
o
f
t
h
e
s
e
c
r
e
t
k
e
y
as
C
S
r
e
p
r
ese
n
t
e
d
a
s
.
−
B
ased
o
n
th
e
esti
m
ated
v
alu
e
s
o
f
,
an
d
th
e
co
m
p
u
tatio
n
p
r
o
ce
s
s
is
p
er
f
o
r
m
e
d
with
th
e
i
n
f
o
r
m
atio
n
tr
an
s
f
er
r
ed
th
r
o
u
g
h
th
e
C
S.
−
W
ith
th
e
v
alu
e
o
f
r
ec
ep
tio
n
with
th
e
C
S th
e
r
an
d
o
m
n
u
m
b
er
is
g
en
er
ated
∈
∗
co
m
p
u
tatio
n
o
f
(
)
.
f
o
r
th
e
KGC v
alu
es.
−
T
h
e
KGC
v
alu
es
ar
e
co
m
p
u
te
d
with
th
e
esti
m
atio
n
o
f
v
alu
e
=
.
2
is
co
n
v
er
s
io
n
o
f
v
al
u
e
B
with
in
th
e
C
S.
−
T
h
e
esti
m
ated
C
S v
alu
e
f
o
r
th
e
co
m
p
o
n
en
ts
is
d
en
o
ted
as:
′
=
.
=
.
2
.
=
(
)
.
.
2
.
=
(
+
)
∗
1
2
.
=
(
+
)
.
(
5
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
S
ec
u
r
in
g
h
ea
lth
c
a
r
e
d
a
ta
a
n
d
o
p
timiz
in
g
d
ig
ita
l m
a
r
ke
tin
g
…
(
F
a
th
i A
b
d
err
a
h
ma
n
e
)
5733
T
h
e
d
ec
r
y
p
tio
n
p
r
o
ce
s
s
is
ev
alu
ated
th
r
o
u
g
h
th
e
in
te
g
r
atio
n
o
f
DS
an
d
DR
,
em
p
h
asizin
g
li
g
h
tweig
h
t
o
p
er
atio
n
s
.
T
h
e
k
e
y
co
m
p
o
n
e
n
ts
in
v
o
lv
ed
in
th
is
p
r
o
ce
s
s
ar
e
d
en
o
ted
as
,
,
co
r
r
esp
o
n
d
in
g
to
ea
ch
k
ey
elem
e
n
t r
eq
u
ir
ed
f
o
r
d
ec
r
y
p
tin
g
th
e
DR
p
r
o
ce
s
s
.
4
.
3
.
3
.
At
t
ribute
-
b
a
s
ed
E
CC
f
o
r
a
utho
riza
t
io
n
E
llip
tic
cu
r
v
e
cr
y
p
t
o
g
r
ap
h
y
(
E
C
C
)
u
tili
ze
s
s
p
ec
if
ic
p
ar
am
eter
s
to
d
ef
in
e
th
e
ellip
tic
cu
r
v
e
an
d
th
e
cr
y
p
to
g
r
ap
h
ic
o
p
e
r
atio
n
s
e
x
ec
u
ted
o
n
it.
T
h
e
m
o
s
t
co
m
m
o
n
ly
em
p
lo
y
ed
E
C
C
p
ar
am
eter
s
in
clu
d
e
th
e
cu
r
v
e
eq
u
atio
n
,
th
e
p
r
i
m
e
m
o
d
u
l
u
s
,
th
e
b
ase
p
o
in
t
[
2
9
]
,
an
d
th
e
o
r
d
er
o
f
th
e
b
ase
p
o
in
t.
T
h
e
E
C
C
p
ar
am
eter
s
f
o
r
th
e
co
m
m
o
n
l
y
u
s
ed
NI
ST
P
-
2
5
6
cu
r
v
e
a
r
e
p
r
o
v
i
d
ed
:
T
h
e
cu
r
v
e
p
ar
am
eter
s
elec
ted
f
o
r
th
e
a
n
aly
s
is
is
s
h
o
wn
in
eq
u
atio
n
:
2
=
3
−
3
+
(
6
)
T
h
e
Prim
e
m
o
d
u
lu
s
(
f
ield
s
ize
)
with
th
e
C
AM
L
-
E
HDS
m
o
d
el
is
p
r
esen
ted
as
=
2
256
−
2
224
+
2
1
92
+
2
96
−
1
(
7
)
T
h
r
o
u
g
h
th
e
eq
u
atio
n
th
e
co
o
r
d
in
ates a
n
d
v
alu
e
b
is
co
m
p
u
t
ed
as
=
410583637
251521
421293
261297
8004726
840911
444101
599372
555483
525631
403946
740129
1
with
b
ased
g
en
er
at
o
r
o
f
G
=
(
x
,
y
)
w
h
er
e:
x
=
484395612
939064
517590
525852
527979
142027
629495
260417
479958
440807
1708240
463528
6
y
=
36134250
956749
795798
585127
919587
881956
611106
672985
015071
877198
253568
4144051
0
9
T
h
e
o
r
d
er
o
f
p
air
is
co
m
p
u
ted
as:
=
11579208921
035624
876269
744694
940757
352999
695522
413576
034242
2259061
068512
04436
9
T
h
ese
p
ar
am
eter
s
d
ef
in
e
th
e
ellip
tic
cu
r
v
e
an
d
ar
e
u
s
ed
in
E
C
C
o
p
er
atio
n
s
lik
e
k
e
y
g
en
er
atio
n
,
p
o
in
t
m
u
ltip
licatio
n
,
an
d
d
ig
ital sig
n
atu
r
es.
4
.
4
.
Clus
t
er
-
ba
s
ed
a
na
ly
s
is
T
h
e
co
n
s
tr
u
ctio
n
o
f
th
e
clu
s
ter
is
ass
es
s
ed
b
y
co
n
s
id
er
in
g
th
e
o
b
s
er
v
ed
s
em
an
tic
d
o
m
ain
s
.
B
y
co
m
p
u
tin
g
th
e
C
AM
L
-
E
HDS
m
o
d
el,
clu
s
ter
s
ar
e
r
an
k
e
d
b
as
ed
o
n
th
e
esti
m
atio
n
o
f
th
e
m
e
an
v
alu
e
with
in
th
e
clu
s
ter
g
r
o
u
p
.
T
h
e
th
clu
s
ter
r
elatio
n
s
h
ip
is
ev
alu
ated
b
ased
o
n
th
e
len
g
th
o
f
th
e
clu
s
ter
m
o
d
el
in
th
e
d
o
m
ain
as
(
−
1
ℎ
an
d
+
1
ℎ
)
.
W
ith
co
m
p
u
tatio
n
o
f
th
e
s
im
ilar
ity
in
d
ex
in
th
e
th
clu
s
ter
is
d
esig
n
ed
with
,
.
W
ith
in
th
e
d
o
m
ain
o
f
th
clu
s
ter
with
d
o
m
ain
p
an
d
q
v
alu
es
is
m
ea
s
u
r
ed
as
1
[
3
0
]
.
Similar
ly
,
f
o
r
th
e
d
o
m
ai
n
p
an
d
q
th
e
ass
ig
n
ed
v
alu
es
is
s
tated
as
0
.
5
o
th
er
it
is
ass
i
g
n
ed
as
th
e
0
.
T
h
e
tr
an
s
f
o
r
m
atio
n
o
f
th
e
s
o
u
r
ce
d
o
m
ain
is
ev
alu
ated
b
y
m
ap
p
in
g
th
e
tar
g
et
f
u
n
ctio
n
with
th
e
laten
t
s
p
ac
e
d
o
f
attac
k
s
[
3
1
]
.
T
h
r
o
u
g
h
th
e
co
n
v
er
s
io
n
o
f
th
e
attac
k
er
'
s
d
o
m
ain
,
th
e
tr
an
s
f
o
r
m
atio
n
o
f
th
e
laten
t
s
p
ac
e
is
a
s
s
e
s
s
ed
u
s
in
g
ab
u
n
d
a
n
t
lab
el
in
s
tan
ce
s
to
clas
s
if
y
th
e
h
ea
lt
h
ca
r
e
tar
g
et
d
o
m
ain
f
o
r
s
ec
u
r
ity
.
T
o
en
h
an
ce
th
e
s
ec
u
r
ity
o
f
h
ea
lth
ca
r
e
d
ata,
lab
ellin
g
is
ap
p
lie
d
to
th
e
tar
g
et
in
s
tan
ce
s
with
th
e
tr
ain
in
g
o
f
th
e
class
if
ier
[
3
2
]
.
W
ith
th
e
p
r
o
p
o
s
ed
m
o
d
el
d
ee
p
lear
n
in
g
f
o
cu
s
ed
o
n
th
e
ass
ig
n
m
en
t
o
f
th
e
s
co
r
e
to
th
e
clu
s
ter
g
r
o
u
p
f
o
r
th
e
attac
k
p
r
ev
en
tio
n
.
I
n
itially
,
ea
ch
clu
s
ter
s
o
u
r
ce
is
ass
ig
n
ed
as
th
e
“n
o
r
m
al”
o
r
“a
ttack
er
”
with
th
e
ass
ig
n
ed
lab
els
to
t
h
e
clu
s
ter
.
T
h
e
d
o
m
ain
s
o
u
r
ce
co
m
p
r
is
es
o
f
th
e
tar
g
et
d
o
m
ain
d
en
o
ted
as
1
an
d
2
with
th
e
E
u
clid
ea
n
d
is
tan
ce
.
T
h
e
lab
el
f
o
r
th
e
s
o
u
r
ce
in
th
e
th
clu
s
ter
is
r
an
k
ed
as
th
e
,
+
1
,
−
1
in
th
is
m
o
d
el
th
o
s
e
ar
e
la
b
elled
as
th
e
f
o
llo
ws:
−
Step
1
: I
n
itially
,
s
et
th
e
v
alu
e
as z
er
o
f
o
r
t
h
e
lab
el
−
Step
2
:
Up
o
n
th
e
r
an
k
in
g
o
f
t
h
e
s
o
u
r
ce
clu
s
ter
an
d
attac
k
er
is
d
en
o
ted
as
α
with
th
e
eli
m
in
atio
n
o
f
th
e
clu
s
ter
v
alu
e.
−
Step
3
:
W
ith
th
e
s
o
u
r
ce
n
o
d
e
clu
s
ter
is
r
an
k
ed
as
+
1
with
th
e
attac
k
d
en
o
ted
as
2
will
b
e
in
cl
u
d
ed
in
th
e
clu
s
ter
else it
will b
e
elim
in
ated
.
−
Step
4
:
W
ith
th
e
s
o
u
r
ce
n
o
d
es
th
e
r
an
k
o
f
clu
s
ter
is
s
tated
a
s
−
1
an
d
attac
k
er
is
d
ef
in
ed
as
2
in
clu
d
e
d
with
in
th
e
s
y
s
tem
else n
o
d
e
is
elim
in
ated
f
r
o
m
th
e
clu
s
ter
g
r
o
u
p
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
15
,
No
.
6
,
Decem
b
e
r
20
25
:
5
7
2
8
-
5
7
4
5
5734
T
h
r
o
u
g
h
th
e
esti
m
atio
n
o
f
th
e
s
co
r
e
tar
g
et
as
0
an
d
1
th
e
cl
u
s
ter
in
s
tan
ce
s
ar
e
n
o
r
m
alize
d
with
th
e
n
o
r
m
al
o
r
attac
k
.
W
ith
th
e
ass
ig
n
ed
s
o
f
t
lab
els
th
e
in
s
tan
ce
s
f
o
r
th
e
th
r
esh
o
ld
is
1
is
co
n
s
id
er
ed
as
th
e
attac
k
else
th
e
th
r
esh
o
ld
2
is
co
n
s
id
er
ed
as
th
e
t
h
r
esh
o
ld
d
ef
in
e
d
as
th
e
n
o
r
m
al”.
T
h
e
i
n
s
tan
ce
s
f
o
r
th
e
tar
g
et
a
r
e
d
ef
in
ed
as:
1
=
\
\
Set
as
attac
k
la
b
el,
an
d
2
=
1
−
\
\
Set
as
a
n
o
r
m
al
la
b
el
in
th
is
lab
el,
th
e
ass
ig
n
m
e
n
t
s
ch
em
e
with
lab
elled
in
s
tan
ce
s
attac
k
s
i
s
clas
s
if
ied
an
d
elim
in
ate
th
e
in
co
r
p
o
r
atio
n
o
f
th
e
attac
k
s
in
th
e
n
etwo
r
k
b
y
s
o
f
t
lab
ellin
g
.
W
ith
in
th
e
clu
s
ter
g
r
o
u
p
,
lab
els
ar
e
ass
ig
n
ed
to
ea
ch
clu
s
ter
,
in
co
r
p
o
r
atin
g
th
r
ee
co
m
p
o
n
en
ts
o
f
h
ea
lth
ca
r
e
d
ata
s
ec
u
r
ity
an
d
class
if
icatio
n
.
T
h
e
n
o
d
e
clu
s
ter
s
s
h
o
u
ld
in
cl
u
d
e
v
ar
io
u
s
f
ac
to
r
s
s
u
ch
as
p
r
io
r
k
n
o
wled
g
e
,
p
r
o
b
ab
ilit
y
o
f
ed
g
es,
an
d
c
o
n
d
iti
o
n
al
p
r
o
b
ab
ilit
y
tab
le
(
C
PT)
[
3
3
]
.
O
u
r
p
r
o
ce
s
s
f
o
cu
s
es
o
n
esti
m
atin
g
n
etwo
r
k
attac
k
s
b
y
co
m
p
u
tin
g
ca
u
s
ality
an
d
in
teg
r
atin
g
it
with
th
e
ML
-
b
ased
tr
an
s
f
er
lear
n
in
g
p
r
o
ce
s
s
.
T
h
e
C
AM
L
-
E
HDS
p
r
o
ce
s
s
,
co
m
b
in
ed
with
th
e
tr
an
s
f
er
lear
n
in
g
p
r
o
ce
s
s
f
o
r
ass
ig
n
in
g
lab
els
an
d
d
etec
tin
g
attac
k
s
,
is
d
escr
ib
ed
in
(
8
)
.
T
h
r
o
u
g
h
th
e
ass
ig
n
ed
la
b
el
in
s
tan
ce
1
an
d
2
u
n
k
n
o
wn
attac
k
s
ar
e
co
m
p
u
ted
an
d
esti
m
ated
with
co
n
s
id
er
atio
n
o
f
C
PT
attac
k
s
=
(
=
|
=
)
.
T
h
e
p
r
o
ce
s
s
f
lo
w
o
f
o
u
r
m
o
d
el
f
o
r
attac
k
d
etec
tio
n
an
d
p
r
ev
en
tio
n
is
ev
alu
ated
with
th
e
ML
m
o
d
el
f
o
r
th
e
tr
ain
in
g
an
d
co
m
p
u
tatio
n
o
f
th
e
tr
u
s
t v
alu
e
s
in
th
e
d
atab
ase.
=
{
+
(
1
−
)
−
1
,
(
|
)
=
1
(
|
)
=
1
(
1
−
)
−
1
,
(
|
)
=
1
(
|
)
=
0
ij
t
-
1
o
th
er
wis
e
(
8
)
4
.
5
.
CAML
-
E
H
DS
a
lg
o
rit
hm
f
o
r
k
ey
ma
na
g
em
ent
s
t
ra
t
eg
ies
W
ith
th
e
ass
ig
n
ed
lab
el
in
s
tan
ce
s
o
f
th
e
attac
k
d
ata
elim
in
ated
th
at
was
id
en
tifie
d
as
(
=
{
1
,
2
,
3
.
.
.
.
.
.
}
)
f
o
r
th
e
attac
k
d
ata
esti
m
atio
n
d
en
o
ted
as
.
T
h
e
C
AM
L
-
E
HD
S a
ttack
s
ce
n
ar
io
is
es
tim
ated
as
=
(
1
,
2
,
.
.
.
.
.
)
with
th
e
ass
ig
n
ed
lab
el
o
f
ML
b
ased
d
ee
p
lear
n
in
g
m
o
d
el
f
o
r
th
e
esti
m
atio
n
o
f
th
e
attac
k
s
.
T
h
e
m
o
d
el
attac
k
s
f
o
r
th
e
esti
m
atio
n
o
f
th
e
v
ar
iab
les
ar
e
co
m
p
u
ted
f
o
r
o
u
r
m
o
d
el
is
p
r
esen
ted
in
Alg
o
r
ith
m
1
.
Alg
o
r
ith
m
1
.
Par
am
eter
e
s
tim
atio
n
Input: Network Attack =
{
(
1
,
2
.
.
.
.
.
)
(
3
,
4
.
.
.
.
.
)
.
.
.
.
}
Output:
+
1
=
(
+
1
,
+
1
,
+
1
)
// Start
For n = 0 estimate
0
For
0
=
0
set
(
)
0
Compute the attacks those are unknown as n =0,1,2
.
.
.
do
Compute using
(7)
Compute using
(8)
End for
End for
Set the values for estimation
Set values for the comparison
If
(
=
1
|
=
1
)
>
then
Calculate the
1
and
2
based estimated values
End if
for
(
)
set as the attack value
If
(
)
>
then
Calculate the set
End if
End for
End for
T
h
e
ML
f
r
am
ewo
r
k
f
o
cu
s
es
o
n
g
en
er
atin
g
s
o
u
r
ce
m
ap
p
in
g
s
an
d
co
n
s
tr
u
ctin
g
th
e
tar
g
e
t
d
o
m
ain
with
in
th
e
laten
t
s
p
ac
e.
Up
o
n
co
n
v
er
tin
g
th
e
laten
t
s
p
ac
e,
th
e
s
o
u
r
ce
d
o
m
ain
co
n
s
is
ts
o
f
p
r
o
b
a
b
le
in
s
tan
ce
lab
els
f
o
r
attac
k
class
if
icatio
n
.
T
h
e
ac
cu
r
ac
y
o
f
th
e
tr
ain
i
n
g
lab
el
class
if
ier
f
o
r
t
h
e
tar
g
eted
in
s
tan
ce
s
is
ev
alu
ated
u
s
in
g
th
e
ass
ig
n
ed
s
o
f
t
lab
els.
T
h
e
C
AM
L
-
E
HDS
s
o
lu
tio
n
in
v
o
lv
es
g
en
e
r
atin
g
,
d
is
tr
ib
u
tin
g
,
an
d
u
p
d
atin
g
e
n
cr
y
p
tio
n
k
e
y
s
f
o
r
v
ar
io
u
s
en
titi
es,
in
clu
d
in
g
t
h
e
d
ata
o
wn
er
,
t
h
e
clo
u
d
s
e
r
v
er
,
an
d
th
e
d
ata
r
ec
ip
ien
t.
Her
e
ar
e
s
o
m
e
co
n
s
i
d
er
atio
n
s
f
o
r
k
ey
m
an
ag
e
m
en
t
in
th
is
en
v
ir
o
n
m
en
t:
−
Key
g
en
er
atio
n
:
E
n
cr
y
p
tio
n
k
ey
s
ar
e
p
ar
am
o
u
n
t
f
o
r
s
af
eg
u
ar
d
in
g
h
ea
lth
ca
r
e
d
ata
[
3
4
]
.
Secu
r
e
m
eth
o
d
s
s
u
ch
as
r
eliab
le
r
a
n
d
o
m
n
u
m
b
er
g
e
n
er
ato
r
s
o
r
tr
u
s
ted
k
e
y
m
an
ag
em
e
n
t
s
y
s
tem
s
ar
e
es
s
en
tial
f
o
r
th
eir
g
en
er
atio
n
[
3
5
]
.
Fu
r
th
e
r
m
o
r
e
,
th
ese
k
ey
s
m
u
s
t
p
o
s
s
e
s
s
ad
eq
u
ate
s
tr
en
g
th
to
with
s
tan
d
b
r
u
te
-
f
o
r
ce
attac
k
s
an
d
ad
h
er
e
to
r
ec
o
m
m
e
n
d
ed
k
ey
s
ize
g
u
id
elin
es sp
ec
if
ied
f
o
r
th
e
en
cr
y
p
tio
n
alg
o
r
ith
m
in
u
s
e
[
3
6
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
S
ec
u
r
in
g
h
ea
lth
c
a
r
e
d
a
ta
a
n
d
o
p
timiz
in
g
d
ig
ita
l m
a
r
ke
tin
g
…
(
F
a
th
i A
b
d
err
a
h
ma
n
e
)
5735
−
Key
d
is
tr
ib
u
tio
n
: O
n
ce
e
n
cr
y
p
tio
n
k
ey
s
ar
e
g
en
er
ated
,
s
ec
u
r
e
an
d
r
eliab
le
m
et
h
o
d
s
ar
e
im
p
er
ativ
e
f
o
r
th
eir
d
is
tr
ib
u
tio
n
to
d
esig
n
ated
en
tit
ies
[
3
7
]
.
T
h
is
p
r
o
ce
s
s
o
f
ten
in
v
o
lv
es
th
e
u
tili
za
tio
n
o
f
s
ec
u
r
e
ch
an
n
els
s
u
ch
as
en
cr
y
p
ted
e
m
ail
o
r
s
ec
u
r
e
f
ile
tr
an
s
f
er
p
r
o
t
o
co
ls
[
3
8
]
.
E
n
s
u
r
in
g
th
e
s
ec
u
r
e
t
r
an
s
m
is
s
io
n
an
d
p
r
o
tectio
n
o
f
k
ey
s
d
u
r
in
g
d
is
tr
ib
u
tio
n
is
cr
itical
to
p
r
ev
en
t
u
n
au
t
h
o
r
ized
ac
ce
s
s
[
3
9
]
.
−
Key
u
p
d
ates:
I
n
th
e
d
y
n
am
i
c
h
ea
lth
ca
r
e
lan
d
s
ca
p
e,
r
eg
u
lar
u
p
d
ates
to
en
cr
y
p
tio
n
k
ey
s
m
ay
b
ec
o
m
e
n
ec
ess
ar
y
d
u
e
to
v
ar
io
u
s
f
a
cto
r
s
[
4
0
]
.
T
h
ese
f
ac
to
r
s
in
c
lu
d
e
k
ey
ex
p
ir
atio
n
,
c
o
m
p
r
o
m
is
ed
k
ey
s
,
o
r
ch
an
g
es
in
u
s
er
ac
ce
s
s
p
er
m
is
s
io
n
s
.
A
well
-
d
ef
in
ed
p
r
o
ce
s
s
m
u
s
t
m
an
ag
e
k
ey
u
p
d
ates
ef
f
ec
tiv
ely
,
in
co
r
p
o
r
atin
g
m
ec
h
an
is
m
s
f
o
r
r
ev
o
k
in
g
an
d
r
ep
lacin
g
k
e
y
s
as r
eq
u
ir
ed
[
4
1
]
.
−
Acc
ess
c
o
n
tr
o
l:
Pro
p
er
ac
ce
s
s
co
n
tr
o
l
m
ec
h
an
is
m
s
ar
e
v
ital
to
r
estrict
ac
ce
s
s
to
en
cr
y
p
tio
n
k
ey
s
to
o
n
ly
au
th
o
r
ized
e
n
titi
es
[
4
2
]
.
T
h
i
s
m
ay
en
tail
im
p
lem
en
tin
g
r
o
le
-
b
ased
ac
ce
s
s
co
n
tr
o
l,
cr
y
p
to
g
r
ap
h
ic
k
ey
m
an
ag
em
en
t
s
y
s
tem
s
,
o
r
o
th
er
ac
ce
s
s
co
n
tr
o
l
p
o
licies
to
s
af
eg
u
ar
d
s
en
s
itiv
e
in
f
o
r
m
atio
n
f
r
o
m
u
n
au
th
o
r
ized
ac
ce
s
s
[
4
3
]
.
−
Key
s
to
r
ag
e:
Secu
r
e
s
to
r
ag
e
o
f
en
cr
y
p
tio
n
k
ey
s
is
cr
u
cial
to
p
r
ev
en
t
u
n
a
u
th
o
r
ize
d
ac
ce
s
s
an
d
p
o
ten
tial
b
r
ea
ch
es
[
4
4
]
u
tili
zin
g
h
ar
d
w
ar
e
s
ec
u
r
ity
m
o
d
u
les
(
HSMs)
o
r
o
th
er
s
ec
u
r
e
s
to
r
ag
e
s
o
lu
tio
n
s
ca
n
h
elp
s
af
eg
u
ar
d
k
ey
s
f
r
o
m
b
o
t
h
p
h
y
s
ical
an
d
lo
g
ical
attac
k
s
,
th
er
e
b
y
en
h
an
cin
g
o
v
er
all
s
ec
u
r
ity
[
4
5
]
.
−
Key
b
ac
k
u
p
an
d
r
ec
o
v
e
r
y
:
R
eg
u
lar
b
ac
k
u
p
s
o
f
e
n
cr
y
p
tio
n
k
ey
s
ar
e
n
ec
ess
ar
y
to
m
itig
ate
th
e
r
is
k
o
f
d
ata
lo
s
s
in
th
e
ev
en
t
o
f
k
e
y
c
o
m
p
r
o
m
is
e
o
r
s
y
s
tem
f
ailu
r
es
[
4
6
]
.
E
s
tab
lis
h
in
g
a
r
o
b
u
s
t
k
ey
r
ec
o
v
er
y
p
r
o
c
ess
is
v
ital to
r
esto
r
e
ac
ce
s
s
to
en
cr
y
p
ted
d
ata
p
r
o
m
p
tly
if
k
e
y
s
ar
e
lo
s
t o
r
b
ec
o
m
e
in
ac
ce
s
s
ib
le
[
4
7
]
.
−
C
o
m
p
lian
ce
an
d
a
u
d
itin
g
:
Key
m
an
ag
em
e
n
t
p
r
o
ce
s
s
es
m
u
s
t
co
m
p
ly
with
r
ele
v
an
t
r
eg
u
lato
r
y
r
eq
u
ir
em
e
n
ts
,
s
u
ch
as
h
ea
lth
in
s
u
r
an
ce
p
o
r
tab
ilit
y
an
d
ac
c
o
u
n
tab
ilit
y
ac
t
(
H
I
PAA
)
f
o
r
h
ea
lth
ca
r
e
d
ata
[
4
8
]
.
R
eg
u
lar
au
d
its
an
d
co
n
t
in
u
o
u
s
m
o
n
ito
r
in
g
s
h
o
u
ld
b
e
co
n
d
u
cte
d
to
en
s
u
r
e
co
m
p
lian
ce
an
d
id
en
tif
y
an
y
p
o
te
n
tial v
u
ln
er
a
b
ilit
ies in
th
e
k
ey
m
an
ag
em
e
n
t sy
s
tem
[
4
9
]
.
I
m
p
lem
en
tin
g
a
co
m
p
r
eh
e
n
s
iv
e
k
ey
m
a
n
ag
em
e
n
t
s
tr
ateg
y
is
ess
en
tial
f
o
r
m
ain
tain
in
g
t
h
e
s
ec
u
r
ity
an
d
co
n
f
id
e
n
tiality
o
f
h
ea
lth
c
ar
e
d
ata
in
a
d
y
n
am
ic
en
v
ir
o
n
m
en
t
[
5
0
]
.
C
o
n
s
u
ltin
g
with
s
ec
u
r
ity
ex
p
er
ts
an
d
ad
h
er
in
g
to
in
d
u
s
tr
y
b
est
p
r
ac
tices
is
r
ec
o
m
m
en
d
e
d
to
d
esi
g
n
a
n
d
im
p
lem
en
t
a
n
ef
f
ec
tiv
e
k
ey
m
an
a
g
em
en
t
s
y
s
tem
th
at
m
ee
ts
th
e
s
p
ec
if
ic
s
ec
u
r
ity
n
ee
d
s
o
f
h
ea
lth
ca
r
e
o
r
g
an
izatio
n
s
[
5
1
]
.
4
.
6
.
M
a
chine
lea
rning
m
o
de
l
4
.
6
.
1
.
L
o
ng
s
ho
rt
-
t
er
m m
e
mo
ry
net
wo
r
k
s
(
L
ST
M
s
)
L
STM
s
ar
e
em
p
l
o
y
ed
to
an
aly
ze
th
e
in
h
er
e
n
t
tem
p
o
r
al
d
ep
e
n
d
en
cies
with
in
p
atien
t
r
ec
o
r
d
s
,
ess
en
tial
f
o
r
d
etec
tin
g
ev
o
l
v
in
g
s
ec
u
r
ity
th
r
ea
ts
th
at
m
an
if
est
o
v
er
tim
e
[
5
2
]
.
T
o
a
n
a
ly
ze
th
e
tem
p
o
r
al
d
ep
en
d
e
n
cies
in
h
er
en
t
in
-
p
ati
en
t
r
ec
o
r
d
s
,
cr
u
cial
f
o
r
d
ete
ctin
g
ev
o
lv
in
g
s
ec
u
r
ity
th
r
ea
ts
,
lo
n
g
s
h
o
r
t
-
ter
m
m
em
o
r
y
n
etwo
r
k
s
(
L
STM
s
)
wer
e
s
elec
ted
f
o
r
t
h
eir
o
p
tim
i
ze
d
ab
ilit
y
to
p
r
o
ce
s
s
s
eq
u
e
n
tial
tim
e
-
s
er
ies
d
ata,
a
co
m
m
o
n
f
o
r
m
at
in
elec
tr
o
n
i
c
h
ea
lth
r
ec
o
r
d
s
.
Un
lik
e
t
r
a
n
s
f
o
r
m
er
s
,
wh
ich
ex
ce
l
at
c
ap
tu
r
in
g
lo
n
g
-
r
an
g
e
d
ep
en
d
e
n
cies
ac
r
o
s
s
en
tire
s
e
q
u
en
ce
s
b
u
t
ar
e
co
m
p
u
tatio
n
ally
in
ten
s
iv
e,
o
r
co
n
v
o
lu
tio
n
al
n
eu
r
al
n
etwo
r
k
s
(
C
NNs),
wh
ich
ar
e
ef
f
ec
tiv
e
f
o
r
s
p
atial
d
ata
b
u
t
less
s
u
ited
f
o
r
tem
p
o
r
al
p
atter
n
s
,
L
STM
s
o
f
f
er
a
b
alan
ce
o
f
ef
f
icien
cy
a
n
d
e
f
f
ec
tiv
en
ess
in
id
en
tify
i
n
g
s
u
b
tle
an
o
m
al
ies
an
d
p
atter
n
s
t
h
at
em
er
g
e
o
v
er
tim
e.
T
h
eir
r
ec
u
r
r
en
t
ar
ch
itectu
r
e
allo
ws
th
em
to
m
ain
tain
m
em
o
r
y
ac
r
o
s
s
s
eq
u
en
ce
s
,
en
a
b
lin
g
t
h
e
d
etec
tio
n
o
f
th
r
ea
ts
th
at
m
an
if
est
as
ch
a
n
g
es
in
p
atien
t
d
ata
o
v
er
e
x
ten
d
ed
p
e
r
io
d
s
,
m
ak
i
n
g
th
e
m
a
m
o
r
e
p
r
a
ctica
l
an
d
ef
f
icien
t
ch
o
ice
f
o
r
th
is
s
p
ec
if
ic
ap
p
lica
tio
n
.
Fo
r
o
u
r
L
STM
im
p
lem
e
n
tatio
n
,
h
y
p
er
p
ar
a
m
eter
s
wer
e
m
eticu
lo
u
s
ly
s
elec
ted
th
r
o
u
g
h
a
co
m
b
in
atio
n
o
f
g
r
i
d
s
ea
r
ch
a
n
d
v
alid
atio
n
s
et
p
er
f
o
r
m
an
ce
e
v
alu
atio
n
[
5
3
]
.
W
e
u
tili
ze
d
a
m
u
lti
-
lay
e
r
ed
L
STM
ar
ch
itectu
r
e
with
1
2
8
h
i
d
d
en
u
n
its
p
er
lay
er
,
d
eter
m
in
e
d
to
b
ala
n
ce
m
o
d
el
co
m
p
lex
ity
an
d
c
o
m
p
u
tatio
n
al
ef
f
icien
c
y
.
T
h
e
s
eq
u
en
ce
len
g
th
was
s
et
to
5
0
,
ex
h
ib
ited
ef
f
icien
t
m
em
o
r
y
u
s
ag
e,
r
e
q
u
ir
in
g
ap
p
r
o
x
im
a
tely
4
GB
o
f
GPU
m
em
o
r
y
d
u
r
in
g
tr
ain
in
g
.
T
h
is
co
n
f
i
g
u
r
atio
n
r
esu
lted
in
a
n
av
er
ag
e
tr
ai
n
in
g
tim
e
o
f
3
h
o
u
r
s
o
n
o
u
r
d
ataset.
T
h
e
Ad
am
o
p
tim
izer
was
ch
o
s
en
with
a
lear
n
i
n
g
r
ate
o
f
0
.
0
0
1
,
an
d
b
atch
s
ize
wa
s
s
et
to
3
2
,
v
alu
es
d
eter
m
in
ed
th
r
o
u
g
h
g
r
id
s
ea
r
c
h
to
o
p
tim
ize
c
o
n
v
e
r
g
en
ce
an
d
p
r
ev
en
t
o
v
er
f
itti
n
g
.
4
.
6
.
2
.
T
ra
ns
f
er
lea
rning
wit
h
pre
-
t
ra
ined m
o
dels
C
o
m
p
lem
en
tin
g
L
STM
s
,
we
l
ev
er
ag
e
tr
a
n
s
f
er
lear
n
in
g
with
B
E
R
T
to
im
p
r
o
v
e
o
u
r
m
o
d
el'
s
ab
ilit
y
to
u
n
d
er
s
tan
d
t
h
e
s
em
an
tic
co
n
t
ex
t
o
f
h
ea
lth
ca
r
e
d
ata.
B
E
R
T
,
p
r
e
-
tr
ain
e
d
o
n
v
ast
am
o
u
n
ts
o
f
tex
t,
ex
ce
ls
in
ca
p
tu
r
in
g
co
m
p
lex
r
elatio
n
s
h
ip
s
b
etwe
en
wo
r
d
s
an
d
p
h
r
ases
[
5
4
]
.
T
h
is
allo
ws
f
o
r
th
e
d
etec
tio
n
o
f
s
u
b
tle
s
em
an
tic
an
o
m
alies
th
at
m
ay
in
d
icate
u
n
au
t
h
o
r
ized
ac
ce
s
s
o
r
d
ata
m
an
ip
u
latio
n
.
W
h
ile
L
STM
s
ar
e
o
p
tim
ized
f
o
r
tem
p
o
r
al
an
aly
s
is
,
B
E
R
T
p
r
o
v
id
es
a
d
ee
p
s
em
an
tic
u
n
d
er
s
tan
d
in
g
,
all
o
win
g
u
s
to
ca
p
tu
r
e
d
if
f
er
en
t
t
h
r
ea
t
v
ec
to
r
s
[
5
5
]
.
B
y
co
m
b
in
i
n
g
L
STM
s
f
o
r
tem
p
o
r
al
p
atter
n
r
ec
o
g
n
itio
n
an
d
B
E
R
T
f
o
r
s
em
a
n
tic
u
n
d
er
s
tan
d
in
g
,
o
u
r
m
o
d
el
ac
h
iev
es
a
co
m
p
r
e
h
en
s
iv
e
an
aly
s
is
o
f
h
ea
lth
ca
r
e
d
ata,
ad
d
r
ess
in
g
b
o
th
th
e
s
eq
u
en
tial
n
atu
r
e
an
d
th
e
co
m
p
lex
s
em
an
tic
co
n
ten
t
o
f
th
e
in
f
o
r
m
atio
n
[
5
6
]
.
T
h
i
s
h
y
b
r
id
ap
p
r
o
ac
h
o
p
tim
izes
th
r
ea
t
d
etec
tio
n
b
y
lev
er
ag
in
g
th
e
s
tr
en
g
th
s
o
f
b
o
th
r
ec
u
r
r
e
n
t
an
d
tr
a
n
s
f
o
r
m
er
-
b
ased
ar
ch
itectu
r
es,
in
cr
ea
s
in
g
th
e
o
v
er
all
s
ec
u
r
ity
o
f
h
ea
lth
ca
r
e
d
ata
ec
o
s
y
s
tem
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
7
0
8
I
n
t J E
lec
&
C
o
m
p
E
n
g
,
Vo
l.
15
,
No
.
6
,
Decem
b
e
r
20
25
:
5
7
2
8
-
5
7
4
5
5736
4
.
6
.
3
.
M
o
del
f
us
io
n:
a
m
pli
f
y
ing
det
ec
t
io
n c
a
pa
bil
it
ies
As
th
e
in
d
iv
id
u
al
m
o
d
els
em
e
r
g
e
f
r
o
m
th
e
co
n
tain
e
r
o
f
tr
ai
n
in
g
,
th
e
y
co
n
v
er
g
e
h
ar
m
o
n
io
u
s
ly
in
th
e
f
u
s
io
n
p
h
ase,
f
o
r
g
in
g
an
allia
n
ce
th
at
tr
an
s
ce
n
d
s
th
e
ca
p
a
b
ilit
ies
o
f
an
y
s
in
g
le
m
o
d
el.
H
er
e,
th
e
c
o
llectiv
e
in
tellig
en
ce
o
f
L
STM
s
an
d
tr
an
s
f
er
lear
n
in
g
m
o
d
els
co
m
b
i
n
e,
b
ir
th
in
g
a
h
y
b
r
id
f
u
s
io
n
a
p
p
r
o
ac
h
p
r
ep
ar
ed
to
r
ed
ef
in
e
h
ea
lth
ca
r
e
s
ec
u
r
ity
t
h
r
ea
t
d
etec
tio
n
.
E
m
p
lo
y
in
g
s
o
p
h
is
ticated
f
u
s
io
n
tech
n
iq
u
e
s
,
p
r
ed
ictio
n
s
f
r
o
m
in
d
iv
id
u
al
m
o
d
els
ar
e
co
m
b
in
ed
to
f
o
r
m
a
r
o
b
u
s
t
en
s
em
b
le.
Usi
n
g
th
e
weig
h
ted
av
er
a
g
in
g
m
eth
o
d
,
th
e
f
u
s
io
n
ap
p
r
o
ac
h
cr
ea
tes
a
f
in
al
en
s
e
m
b
le
p
r
ed
ictio
n
en
r
ich
e
d
with
co
llectiv
e
wis
d
o
m
.
I
n
th
is
p
r
o
ce
s
s
,
ea
ch
m
o
d
el'
s
p
r
ed
ictio
n
is
ass
ig
n
ed
a
weig
h
t
b
ased
o
n
its
p
er
f
o
r
m
an
ce
a
n
d
r
eliab
ilit
y
[
5
7
]
.
T
h
ese
weig
h
ted
s
co
r
es
ar
e
th
en
av
er
ag
ed
to
p
r
o
d
u
ce
a
u
n
if
ied
p
r
ed
ictio
n
.
T
h
is
s
y
n
t
h
esis
tr
an
s
ce
n
d
s
th
e
lim
itatio
n
s
o
f
in
d
iv
id
u
al
m
o
d
els
b
y
lev
er
ag
in
g
th
eir
d
iv
er
s
e
s
tr
en
g
th
s
.
B
y
ca
r
ef
u
lly
ass
ig
n
in
g
weig
h
ts
,
th
e
f
u
s
io
n
m
eth
o
d
e
n
s
u
r
es
th
at
th
e
m
o
s
t
ac
cu
r
ate
an
d
r
eliab
le
m
o
d
els h
av
e
a
g
r
ea
ter
in
f
lu
en
ce
o
n
t
h
e
f
in
al
p
r
ed
ictio
n
[
5
8
]
.
T
h
is
ap
p
r
o
ac
h
im
p
r
o
v
es th
e
o
v
er
all
r
esil
ien
ce
an
d
e
f
f
icac
y
o
f
th
e
h
ea
lth
ca
r
e
s
ec
u
r
ity
s
y
s
tem
,
p
r
o
v
id
in
g
a
c
o
m
p
r
e
h
en
s
i
v
e
d
ef
en
s
e
ag
ai
n
s
t
p
o
ten
tial th
r
ea
ts
.
4
.
7
.
Dig
it
a
l
ma
rk
e
t
ing
inte
g
ra
t
io
n
C
AM
L
-
E
HDS
u
n
iq
u
ely
in
teg
r
ates
d
ig
ital
m
ar
k
etin
g
s
tr
ateg
ies
with
r
o
b
u
s
t
d
ata
s
ec
u
r
ity
,
lev
er
ag
in
g
in
s
ig
h
ts
f
r
o
m
u
n
if
ied
th
r
ea
t
d
etec
tio
n
an
d
clu
s
ter
-
b
ased
a
n
aly
s
is
.
T
h
is
en
ab
les
h
ea
lth
c
ar
e
o
r
g
a
n
izatio
n
s
to
d
ev
elo
p
tar
g
eted
ad
v
er
tis
em
e
n
ts
an
d
p
e
r
s
o
n
alize
d
c
o
n
ten
t
b
ased
o
n
class
if
ied
d
ata
a
n
d
d
etec
ted
t
h
r
ea
ts
,
im
p
r
o
v
in
g
cu
s
to
m
er
en
g
a
g
e
m
en
t
an
d
m
ar
k
etin
g
ef
f
ec
ti
v
en
ess
.
B
y
em
p
lo
y
in
g
ad
v
a
n
ce
d
cr
y
p
t
o
g
r
ap
h
ic
tech
n
iq
u
es
lik
e
h
o
m
o
m
o
r
p
h
i
c
en
cr
y
p
tio
n
an
d
E
C
C
,
C
A
ML
-
E
HDS
en
s
u
r
es
th
at
m
ar
k
etin
g
d
ata
r
em
ain
s
s
ec
u
r
e
an
d
co
m
p
lian
t w
ith
p
r
i
v
ac
y
r
eg
u
latio
n
s
.
Ho
wev
er
,
th
is
in
teg
r
atio
n
n
ec
ess
itate
s
ca
r
e
f
u
l c
o
n
s
id
er
atio
n
o
f
eth
ical
co
n
ce
r
n
s
.
Sp
ec
if
ically
,
th
e
u
s
e
o
f
s
en
s
itiv
e
h
ea
lth
ca
r
e
d
ata
f
o
r
m
ar
k
etin
g
p
u
r
p
o
s
es
r
aises
q
u
esti
o
n
s
ab
o
u
t
in
f
o
r
m
ed
co
n
s
en
t,
d
ata
an
o
n
y
m
izatio
n
,
an
d
th
e
p
o
te
n
tial
f
o
r
d
is
cr
im
in
ato
r
y
tar
g
etin
g
.
T
o
m
itig
ate
th
ese
r
is
k
s
,
C
AM
L
-
E
HDS
in
co
r
p
o
r
ates
m
ec
h
an
is
m
s
f
o
r
t
r
an
s
p
ar
en
t
d
ata
u
s
ag
e,
r
o
b
u
s
t
an
o
n
y
m
izatio
n
tech
n
iq
u
es,
an
d
s
tr
ict
ad
h
er
e
n
ce
to
p
r
iv
ac
y
r
eg
u
latio
n
s
.
5.
RE
SU
L
T
S
5
.
1
.
E
v
a
lua
t
io
n o
f
hea
lt
hca
re
da
t
a
s
ec
urit
y
us
ing
CAM
L
-
E
H
DS m
o
del
Usi
n
g
th
e
p
r
o
p
o
s
ed
C
AM
L
-
E
HDS
tech
n
iq
u
es,
h
ea
lth
ca
r
e
d
ata
s
ec
u
r
ity
f
ea
tu
r
es
ar
e
ass
ess
ed
with
m
ac
h
in
e
lear
n
i
n
g
,
f
o
cu
s
in
g
o
n
th
r
ee
d
if
f
er
en
t
m
et
r
ics:
au
th
e
n
ticatio
n
,
en
cr
y
p
tio
n
,
an
d
m
ac
h
in
e
lear
n
in
g
.
T
h
e
m
o
d
el
in
cl
u
d
es
a
c
r
y
p
t
o
g
r
a
p
h
ic
p
r
o
ce
s
s
th
at
is
ex
a
m
in
ed
c
o
n
s
id
er
in
g
v
ar
i
o
u
s
f
ea
tu
r
es
f
o
r
s
ec
u
r
ity
,
co
m
m
u
n
icatio
n
o
v
e
r
h
ea
d
,
an
d
co
m
p
u
tatio
n
p
r
o
ce
s
s
.
T
h
e
h
o
m
o
m
o
r
p
h
ic
en
c
r
y
p
tio
n
s
ch
em
e
is
ev
alu
ated
u
s
in
g
th
e
attr
ib
u
te
-
b
ased
escr
o
w
m
o
d
el
f
o
r
a
n
aly
s
is
.
I
n
th
e
p
r
o
p
o
s
ed
m
o
d
el,
h
o
m
o
m
o
r
p
h
ic
en
cr
y
p
tio
n
is
u
s
ed
to
s
to
r
e
elec
tr
o
n
ic
m
e
d
ical
r
ec
o
r
d
s
o
n
t
h
e
escr
o
w
s
er
v
er
.
T
h
is
en
cr
y
p
tio
n
m
eth
o
d
is
ap
p
lied
to
th
e
m
e
d
ical
h
ea
lth
ca
r
e
r
ec
o
r
d
s
.
T
h
e
ex
am
in
ed
r
esu
lts
f
o
r
th
e
co
n
s
tr
u
cted
m
o
d
el
ar
e
p
r
e
s
en
ted
in
Fig
u
r
e
2
.
I
n
Fig
u
r
e
3
,
E
C
C
-
b
ased
au
th
o
r
izatio
n
is
co
n
d
u
cted
f
o
r
th
e
ev
alu
atio
n
an
d
c
o
m
p
u
tatio
n
o
f
m
ed
ical
d
ata.
T
h
e
ex
a
m
in
atio
n
in
v
o
l
v
es
au
th
o
r
izin
g
u
s
er
s
o
f
m
ed
ical
h
ea
lth
ca
r
e
d
ata.
C
o
m
p
u
ted
a
u
th
o
r
izatio
n
u
s
in
g
E
C
C
is
im
p
lem
en
ted
in
th
e
clo
u
d
to
en
h
an
c
e
s
ec
u
r
ity
.
5
.
2
.
Securit
y
f
e
a
t
ures
T
h
e
C
AM
L
-
E
HDS
m
o
d
el
is
d
esig
n
ed
to
b
o
o
s
t
th
e
s
ec
u
r
ity
o
f
h
ea
lth
ca
r
e
d
ata
b
y
in
c
o
r
p
o
r
ati
n
g
v
ar
io
u
s
ad
v
a
n
ce
d
s
ec
u
r
ity
f
ea
tu
r
es.
T
h
ese
f
ea
tu
r
es
ar
e
ev
al
u
ated
an
d
c
o
m
p
ar
e
d
ag
ain
s
t
ex
is
tin
g
m
o
d
els
to
h
ig
h
lig
h
t
th
e
ef
f
ec
tiv
en
ess
o
f
o
u
r
m
o
d
el
in
s
af
eg
u
ar
d
in
g
s
en
s
itiv
e
in
f
o
r
m
atio
n
.
T
h
e
k
e
y
s
ec
u
r
ity
f
ea
tu
r
es
ass
es
s
ed
in
clu
d
e
p
air
in
g
-
f
r
e
e
o
p
er
atio
n
s
,
E
C
C
b
ased
m
eth
o
d
s
,
k
ey
-
escr
o
w
m
ec
h
a
n
i
s
m
s
,
r
esis
tan
ce
to
co
llu
s
io
n
attac
k
s
,
p
r
o
v
ab
le
s
e
cu
r
ity
,
a
n
d
k
ey
au
th
o
r
ity
m
an
ag
em
en
t.
T
ab
le
2
p
r
esen
ts
a
c
o
m
p
ar
ativ
e
an
aly
s
is
o
f
th
ese
s
ec
u
r
ity
f
ea
tu
r
es a
cr
o
s
s
d
if
f
er
en
t sch
em
es.
T
h
e
p
er
f
o
r
m
a
n
ce
o
f
t
h
e
C
AM
L
-
E
HDS
m
o
d
el
is
co
m
p
ar
e
d
with
ex
is
tin
g
s
ch
em
es
s
u
ch
as
GHZ
,
J
,
HZ
,
XZ
Y,
a
n
d
SC
H.
T
h
e
c
o
m
p
ar
ativ
e
a
n
aly
s
is
f
o
cu
s
es
o
n
h
o
w
ea
ch
s
ch
em
e
h
an
d
les
t
h
e
s
ec
u
r
ity
f
ea
tu
r
es.
C
AM
L
-
E
HDS
m
o
d
el
ex
ce
ls
in
all
ca
teg
o
r
ies,
d
em
o
n
s
tr
atin
g
its
s
u
p
er
io
r
ity
in
p
r
o
v
i
d
in
g
co
m
p
r
eh
en
s
iv
e
s
ec
u
r
ity
f
o
r
h
ea
lth
ca
r
e
d
ata
.
−
GHZ
[
1
4
]
an
d
J
[
1
5
]
s
ch
em
es
lack
p
air
in
g
-
f
r
ee
o
p
er
atio
n
s
an
d
E
C
C
-
b
ased
m
eth
o
d
s
,
wh
ich
ar
e
cr
u
cial
f
o
r
ef
f
icien
t a
n
d
s
ec
u
r
e
d
ata
p
r
o
ce
s
s
in
g
in
r
eso
u
r
ce
-
co
n
s
tr
ain
ed
en
v
ir
o
n
m
en
ts
.
−
HZ
[
3
4
]
p
r
o
v
i
d
es
p
air
in
g
-
f
r
ee
o
p
er
atio
n
s
b
u
t
d
o
es
n
o
t
in
clu
d
e
E
C
C
-
b
ased
m
eth
o
d
s
o
r
k
ey
-
escr
o
w
m
ec
h
an
is
m
s
,
lim
itin
g
its
f
lex
ib
ilit
y
an
d
s
ec
u
r
ity
.
−
XZ
Y
[
7
]
an
d
SC
H
[
8
]
in
co
r
p
o
r
ate
b
o
t
h
p
air
in
g
-
f
r
ee
an
d
E
C
C
-
b
ased
m
eth
o
d
s
b
u
t
l
ac
k
k
ey
-
escr
o
w
f
ea
tu
r
es,
r
ed
u
cin
g
th
eir
ef
f
ec
ti
v
en
ess
in
k
ey
m
a
n
ag
em
e
n
t a
n
d
r
ec
o
v
e
r
y
s
ce
n
ar
i
o
s
.
T
h
e
p
r
o
p
o
s
ed
m
o
d
el
n
o
t
o
n
ly
ad
d
r
ess
es
th
ese
s
h
o
r
tco
m
in
g
s
b
u
t
also
in
tr
o
d
u
ce
s
a
lig
h
tw
eig
h
t
k
ey
-
escr
o
w
s
ch
em
e
an
d
r
o
b
u
s
t
k
ey
au
t
h
o
r
ity
m
an
ag
e
m
en
t,
m
ak
in
g
it
a
well
-
r
o
u
n
d
e
d
s
o
lu
tio
n
f
o
r
s
ec
u
r
in
g
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J E
lec
&
C
o
m
p
E
n
g
I
SS
N:
2088
-
8
7
0
8
S
ec
u
r
in
g
h
ea
lth
c
a
r
e
d
a
ta
a
n
d
o
p
timiz
in
g
d
ig
ita
l m
a
r
ke
tin
g
…
(
F
a
th
i A
b
d
err
a
h
ma
n
e
)
5737
h
ea
lth
ca
r
e
d
ata.
I
n
co
r
p
o
r
atin
g
ad
v
an
ce
d
cr
y
p
to
g
r
a
p
h
ic
tech
n
iq
u
es
an
d
r
o
b
u
s
t
k
ey
m
a
n
ag
e
m
en
t
s
tr
ateg
ies,
o
u
r
m
o
d
el
is
t
h
e
b
est
p
er
f
o
r
m
in
g
m
o
d
el
f
o
r
p
r
o
tectin
g
s
en
s
itiv
e
h
ea
lth
ca
r
e
in
f
o
r
m
atio
n
.
T
h
e
ev
al
u
atio
n
an
d
co
m
p
ar
ativ
e
an
aly
s
is
d
em
o
n
s
tr
ate
its
ef
f
ec
tiv
en
ess
in
m
itig
atin
g
v
ar
io
u
s
s
ec
u
r
ity
t
h
r
ea
ts
,
en
s
u
r
in
g
t
h
e
co
n
f
id
en
tiality
,
i
n
teg
r
ity
,
a
n
d
av
ailab
ilit
y
o
f
h
ea
lth
ca
r
e
d
ata.
Fig
u
r
e
2
.
Me
d
ical
r
ec
o
r
d
s
an
d
en
cr
y
p
tio
n
with
h
o
m
o
m
o
r
p
h
ic
p
r
o
ce
s
s
Fig
u
r
e
3
.
Patien
t d
em
o
g
r
a
p
h
ic
attr
ib
u
tes f
o
r
a
u
th
o
r
izatio
n
T
ab
le
2
.
C
o
m
p
a
r
is
o
n
o
f
s
ec
u
r
i
ty
f
ea
tu
r
es
in
C
AM
L
-
E
HDS
S
c
h
e
me
P
a
i
r
i
n
g
–
f
r
e
e
EC
C
b
a
se
d
K
e
y
–
e
scr
o
w
C
o
l
l
u
si
o
n
a
t
t
a
c
k
P
r
o
v
a
b
l
e
s
e
c
u
r
e
d
K
e
y
a
u
t
h
o
r
i
t
y
G
H
Z
[
1
4
]
No
No
Y
e
s
Y
e
s
Y
e
s
Y
e
s
J
[
1
5
]
No
No
Y
e
s
Y
e
s
No
Y
e
s
H
Z
[
3
4
]
Y
e
s
No
No
No
Y
e
s
No
X
ZY
[
7
]
Y
e
s
Y
e
s
No
No
Y
e
s
No
S
C
H
[
8
]
Y
e
s
Y
e
s
No
Y
e
s
Y
e
s
No
P
r
o
p
o
se
d
C
A
M
L
-
EH
D
S
Y
e
s
Y
e
s
Y
e
s
Y
e
s
Y
e
s
Y
e
s
5
.
3
.
P
er
f
o
r
m
a
nce
a
na
ly
s
is
o
f
no
de
co
nfig
ura
t
io
n in e
ncry
pte
d sy
s
t
em
s
T
ab
le
3
p
r
o
v
i
d
es
a
co
m
p
ar
ativ
e
an
aly
s
is
o
f
en
c
r
y
p
tio
n
tim
e,
lo
s
s
p
er
ce
n
tag
e
,
an
d
ac
cu
r
ac
y
p
er
ce
n
tag
e
f
o
r
d
if
f
er
en
t
n
u
m
b
er
s
o
f
n
o
d
es
in
a
s
y
s
tem
.
T
h
e
C
AM
L
-
E
HDS
m
o
d
el
d
em
o
n
s
tr
ates
s
ig
n
if
ican
t
s
tr
en
g
th
s
in
h
a
n
d
lin
g
en
cr
y
p
tio
n
an
d
m
ain
tain
i
n
g
h
i
g
h
ac
c
u
r
ac
y
in
h
ea
lth
ca
r
e
d
ata
s
ec
u
r
it
y
.
T
h
e
r
esu
lts
s
h
o
w
th
at
with
a
lo
w
n
u
m
b
er
o
f
n
o
d
es,
o
u
r
m
o
d
el
ac
h
iev
es
ex
ce
p
tio
n
ally
h
ig
h
ac
cu
r
ac
y
,
with
9
8
%
at
2
n
o
d
es,
an
d
a
m
in
im
al
lo
s
s
p
er
ce
n
tag
e
o
f
1
3
%.
Evaluation Warning : The document was created with Spire.PDF for Python.