I
AE
S In
t
er
na
t
io
na
l J
o
urna
l o
f
Art
if
icia
l In
t
ellig
ence
(
I
J
-
AI
)
Vo
l.
1
5
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
6
,
p
p
.
493
~
505
I
SS
N:
2
2
5
2
-
8
9
3
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijai.v
15
.i
1
.
p
p
4
9
3
-
5
0
5
493
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
i
.
ia
esco
r
e.
co
m
Predic
ting trapp
e
d victims
in
debri
s usin
g
sig
na
l ana
ly
sis
ensem
ble
cla
ss
ifi
c
a
tion
E
no
ch
Ada
m
a
J
iy
a
1
,
I
lesa
nm
i B.
O
luwa
f
em
i
2
,
O
la
y
ink
a
O
.
O
g
un
dil
e
3
,
O
luwa
s
ey
i P
.
B
a
ba
lo
la
4
1
D
e
p
a
r
t
me
n
t
o
f
El
e
c
t
r
i
c
a
l
a
n
d
I
n
f
o
r
mat
i
o
n
E
n
g
i
n
e
e
r
i
n
g
,
C
o
l
l
e
g
e
o
f
E
n
g
i
n
e
e
r
i
n
g
,
La
n
d
mar
k
U
n
i
v
e
r
si
t
y
,
O
m
u
-
A
r
a
n
,
N
i
g
e
r
i
a
2
D
e
p
a
r
t
me
n
t
o
f
El
e
c
t
r
i
c
a
l
a
n
d
El
e
c
t
r
o
n
i
c
s E
n
g
i
n
e
e
r
i
n
g
,
E
k
i
t
i
S
t
a
t
e
U
n
i
v
e
r
si
t
y
,
A
d
o
Ek
i
t
i
,
N
i
g
e
r
i
a
3
D
e
p
a
r
t
me
n
t
o
f
I
n
d
u
s
t
r
i
a
l
Te
c
h
n
i
c
a
l
E
d
u
c
a
t
i
o
n
,
T
a
i
S
o
l
a
r
i
n
U
n
i
v
e
r
s
i
t
y
o
f
E
d
u
c
a
t
i
o
n
,
I
j
a
g
u
n
,
N
i
g
e
r
i
a
4
D
e
p
a
r
t
me
n
t
o
f
El
e
c
t
r
i
c
a
l
,
E
l
e
c
t
r
o
n
i
c
s
,
a
n
d
C
o
m
p
u
t
e
r
En
g
i
n
e
e
r
i
n
g
,
F
r
e
n
c
h
S
o
u
t
h
A
f
r
i
c
a
n
I
n
st
i
t
u
t
e
o
f
T
e
c
h
n
o
l
o
g
y
,
C
a
p
e
P
e
n
i
n
s
u
l
a
U
n
i
v
e
r
si
t
y
o
f
Te
c
h
n
o
l
o
g
y
,
B
e
l
l
v
i
l
l
e
,
S
o
u
t
h
A
f
r
i
c
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Ma
r
1
8
,
2
0
2
4
R
ev
is
ed
Oct
9
,
2
0
2
4
Acc
ep
ted
Dec
1
5
,
2
0
2
5
On
e
m
a
jo
r
d
iffi
c
u
lt
y
in
p
e
rv
a
siv
e
c
o
m
p
u
ti
n
g
is
trap
p
e
d
h
u
m
a
n
d
e
tec
ti
o
n
in
se
a
rc
h
a
n
d
re
sc
u
e
(S
AR)
sc
e
n
a
rio
s.
Ac
c
u
ra
tely
id
e
n
ti
fy
in
g
trap
p
e
d
in
d
i
v
id
u
a
ls
is
c
h
a
ll
e
n
g
i
n
g
d
u
e
t
o
n
o
isy
d
a
ta
a
n
d
t
h
e
c
u
rse
o
f
d
ime
n
sio
n
a
li
ty
.
Wh
e
n
n
o
n
-
l
in
e
-
of
-
si
g
h
t
(NLOS)
c
o
n
d
i
ti
o
n
s
a
re
p
re
se
n
t
d
u
ri
n
g
c
a
t
a
stro
p
h
ic
o
c
c
u
rre
n
c
e
s,
th
e
c
u
rse
o
f
d
im
e
n
sio
n
a
li
t
y
c
a
n
re
su
lt
in
b
l
in
d
sp
o
ts
in
d
e
tec
ti
o
n
s b
e
c
a
u
se
o
f
n
o
ise
a
n
d
u
n
c
o
rre
late
d
d
a
ta.
Be
c
a
u
se
m
a
c
h
in
e
lea
rn
in
g
a
lg
o
rit
h
m
s
a
re
i
n
c
re
d
ib
l
y
a
c
c
u
ra
te,
th
is
wo
r
k
f
o
c
u
se
s
o
n
u
s
in
g
u
lt
ra
-
wid
e
b
a
n
d
(UWB)
ra
d
a
r
wa
v
e
s
to
d
e
tec
t
in
d
i
v
id
u
a
ls
i
n
NLOS
sc
e
n
a
rio
s
a
n
d
lev
e
ra
g
in
g
wire
les
s
c
o
m
m
u
n
ica
t
io
n
to
h
a
rm
o
n
ize
in
f
o
rm
a
ti
o
n
.
Th
e
p
a
p
e
r
u
se
s
e
n
se
m
b
le
m
e
th
o
d
s
to
e
x
tra
c
t
fe
a
tu
re
s
u
sin
g
i
n
d
e
p
e
n
d
e
n
t
c
o
m
p
o
n
e
n
t
a
n
a
ly
sis
(ICA)
a
n
d
e
v
a
l
u
a
te
c
las
sifica
ti
o
n
p
e
rf
o
rm
a
n
c
e
o
n
b
o
t
h
s
tatic
a
n
d
d
y
n
a
m
ic
d
a
tas
e
ts.
Th
e
tes
ti
n
g
re
su
lt
s
c
o
n
firm
th
e
e
ffe
c
ti
v
e
n
e
ss
o
f
th
e
p
ro
p
o
se
d
stra
teg
y
,
wit
h
c
las
sifica
ti
o
n
a
c
c
u
ra
c
ies
o
f
8
7
.
2
0
%
fo
r
d
y
n
a
m
ic
d
a
ta
a
n
d
8
8
.
0
0
%
fo
r
sta
ti
c
d
a
ta
.
Las
tl
y
,
d
u
rin
g
S
AR
o
p
e
ra
ti
o
n
s,
o
u
r
a
p
p
ro
a
c
h
c
a
n
a
ss
ist
e
n
g
in
e
e
rs an
d
sc
ien
ti
sts in
m
a
k
in
g
q
u
ic
k
d
e
c
isio
n
s.
K
ey
w
o
r
d
s
:
Dete
ctio
n
Dim
en
s
io
n
ality
Sear
ch
an
d
r
escu
e
T
r
ap
p
ed
Ultr
a
-
wid
eb
an
d
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
:
E
n
o
ch
Ad
a
m
a
J
iy
a
Dep
ar
tm
en
t o
f
E
lectr
ical
an
d
I
n
f
o
r
m
atio
n
E
n
g
in
ee
r
in
g
,
C
o
lleg
e
o
f
E
n
g
in
ee
r
i
n
g
,
L
a
n
d
m
ar
k
Un
iv
er
s
ity
Km
.
4
I
p
et
u
R
o
ad
,
PMB
1
0
0
1
,
Om
u
-
Ar
an
,
Kwa
r
a
State,
Nig
er
ia
E
m
ail:
jiy
a.
ad
am
a@
lm
u
.
ed
u
.
n
g
1.
I
NT
RO
D
UCT
I
O
N
Ar
o
u
n
d
th
e
g
l
o
b
e
,
n
at
u
r
al
d
is
aste
r
s
s
u
c
h
as
ea
r
t
h
q
u
ak
es,
i
n
f
e
r
n
o
,
h
u
r
r
i
ca
n
es
,
ts
u
n
a
m
is
,
a
n
d
s
e
is
m
i
c,
ac
ti
v
i
ty
o
c
cu
r
wit
h
al
ar
m
i
n
g
f
r
eq
u
e
n
cy
,
o
f
te
n
le
a
d
i
n
g
t
o
th
e
c
o
ll
ap
s
e
o
f
in
f
r
ast
r
u
ct
u
r
e
c
r
e
ate
d
b
y
h
u
m
a
n
s
,
s
u
c
h
as
b
u
il
d
i
n
g
s
an
d
b
r
id
g
es
[
1
]
.
I
n
s
u
c
h
s
ce
n
a
r
i
o
s
,
s
u
r
v
iv
o
r
s
c
an
b
e
co
m
e
t
r
a
p
p
e
d
wit
h
i
n
t
h
e
d
e
b
r
is
,
p
o
s
in
g
a
s
ig
n
i
f
ic
a
n
t
ch
all
e
n
g
e
t
o
s
ea
r
c
h
an
d
r
ec
o
v
e
r
y
te
am
s
,
wh
o
o
f
te
n
la
ck
p
r
ec
is
e
i
n
f
o
r
m
a
ti
o
n
ab
o
u
t
th
e
p
r
ese
n
ce
,
lo
c
ati
o
n
,
an
d
n
u
m
b
e
r
o
f
t
r
ap
p
e
d
i
n
d
iv
id
u
a
ls
.
Gi
v
en
t
h
a
t
t
h
e
li
f
e
e
x
p
e
ct
an
cy
o
f
p
a
tie
n
t
s
i
n
p
r
is
o
n
d
r
o
p
s
s
ig
n
i
f
ic
a
n
tl
y
a
f
t
er
t
h
r
ee
d
a
y
s
,
a
r
es
cu
e
o
p
e
r
ati
o
n
is
r
eq
u
i
r
e
d
in
th
is
s
it
u
at
io
n
[
2
]
,
[
3
]
.
O
v
e
r
t
h
e
p
ast
f
ew
d
ec
ad
es,
th
es
e
c
at
ast
r
o
p
h
ic
e
v
e
n
ts
h
av
e
cla
im
ed
m
i
lli
o
n
s
o
f
li
v
es
,
u
n
d
er
s
co
r
i
n
g
t
h
e
n
ec
ess
it
y
f
o
r
ef
f
ec
ti
v
e
p
o
s
t
-
d
is
aste
r
r
es
cu
e
o
p
e
r
a
ti
o
n
s
t
o
m
i
n
im
i
ze
f
ata
liti
es
[
4
]
.
D
ev
el
o
p
i
n
g
r
eli
a
b
le
t
ec
h
n
i
q
u
es
f
o
r
r
o
b
o
ti
c
n
av
ig
ati
o
n
a
n
d
m
o
n
i
to
r
i
n
g
in
u
n
c
h
a
r
te
d
e
n
v
i
r
o
n
m
en
ts
r
e
m
ai
n
s
a
c
r
it
ic
al
r
es
e
ar
ch
a
r
e
a
in
s
e
c
u
r
i
ty
a
n
d
p
r
o
te
cti
o
n
[
5
]
.
On
e
p
r
o
m
is
in
g
tech
n
iq
u
e
f
o
r
d
etec
tin
g
h
u
m
a
n
tar
g
ets
th
r
o
u
g
h
b
ar
r
ier
s
is
u
ltra
-
wid
e
b
a
n
d
(
UW
B
)
r
ad
ar
.
UW
B
r
ad
ar
o
f
f
er
s
n
u
m
er
o
u
s
ad
v
an
tag
es,
s
u
c
h
as
ex
c
ellen
t
p
en
etr
ab
ilit
y
,
p
r
ec
is
e
p
l
ac
em
en
t,
lo
w
p
o
wer
co
n
s
u
m
p
tio
n
,
h
ig
h
d
ef
in
itio
n
,
an
d
r
esil
ien
ce
to
ch
an
n
el
d
eg
r
ad
atio
n
[
6
]
,
[
7
]
.
T
h
ese
f
ea
tu
r
e
s
m
ak
e
it
id
ea
l
f
o
r
lo
ca
tin
g
h
i
d
d
en
tar
g
ets
b
en
ea
t
h
b
u
ild
in
g
m
ater
ials
,
m
ak
i
n
g
i
t
in
v
alu
ab
le
f
o
r
law
e
n
f
o
r
ce
m
en
t
an
d
em
er
g
en
cy
r
ec
o
v
er
y
ef
f
o
r
ts
.
UW
B
r
ad
ar
h
as
p
r
o
v
en
e
f
f
ec
tiv
e
in
ap
p
lic
atio
n
s
s
u
ch
as
tar
g
et
id
en
tific
atio
n
,
lo
ca
lizatio
n
,
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
1
5
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
6
:
493
-
5
0
5
494
h
ea
lth
m
o
n
ito
r
in
g
,
an
d
th
r
o
u
g
h
-
wall
h
u
m
an
d
etec
tio
n
[
8
]
.
T
h
e
q
u
ick
d
ev
elo
p
m
en
t
o
f
m
ac
h
in
e
lear
n
in
g
,
p
ar
ticu
lar
ly
d
ee
p
lear
n
in
g
,
h
as
r
ev
o
lu
tio
n
ize
d
th
e
a
b
ilit
y
to
m
ak
e
o
p
tim
al
d
ec
is
io
n
s
q
u
ick
ly
u
s
in
g
h
ig
h
-
p
er
f
o
r
m
an
ce
c
o
m
p
u
tin
g
s
y
s
te
m
s
th
at
ex
tr
ac
t
h
ig
h
-
lev
el
f
ea
tu
r
es
f
r
o
m
lar
g
e
,
m
u
lti
-
d
im
en
s
io
n
al
d
atasets
[
9
]
.
T
h
is
ad
v
an
ce
m
en
t
is
p
a
r
ticu
lar
ly
r
elev
an
t
in
d
is
aster
r
elief
an
d
h
u
m
an
itar
ian
aid
[
1
0
]
,
[
1
1
]
.
Pre
v
io
u
s
r
esear
c
h
p
r
o
p
o
s
ed
a
d
o
wn
s
tr
ea
m
ch
an
n
el
ass
es
s
m
en
t
n
etwo
r
k
u
tili
zin
g
co
n
v
o
lu
tio
n
al
n
eu
r
al
n
etwo
r
k
s
(
C
NNs)
f
o
r
f
ea
tu
r
e
ex
tr
ac
tio
n
(
FE)
a
n
d
r
e
cu
r
r
en
t
n
etwo
r
k
s
f
o
r
ch
an
n
el
esti
m
atio
n
[
1
2
]
.
H
o
wev
er
,
th
e
s
e
m
eth
o
d
s
d
id
n
o
t
s
p
ec
if
ically
ad
d
r
ess
n
o
n
-
lin
e
-
of
-
s
ig
h
t
(
NL
OS)
d
etec
tio
n
a
n
d
o
f
te
n
o
v
e
r
lo
o
k
ed
s
o
m
e
d
ata
in
th
e
ch
a
n
n
el
im
p
u
ls
e
r
esp
o
n
s
e
(
C
I
R
)
.
I
n
th
is
r
esear
ch
,
we
em
p
lo
y
an
in
d
ep
en
d
en
t
co
m
p
o
n
en
t
a
n
aly
s
is
(
I
C
A
)
ap
p
r
o
ac
h
f
o
r
FE
an
d
an
en
s
em
b
le
alg
o
r
ith
m
f
o
r
th
e
class
if
icatio
n
an
d
id
en
tific
atio
n
o
f
h
u
m
a
n
tar
g
ets
b
u
r
ie
d
b
en
ea
th
o
b
s
tacle
s
.
Ad
d
itio
n
ally
,
Stan
d
ar
d
S
ca
ler
is
in
tr
o
d
u
ce
d
,
a
d
ata
n
o
r
m
ali
za
tio
n
alg
o
r
ith
m
,
a
n
d
co
n
d
u
c
t
s
tr
in
g
en
t
tr
ain
in
g
an
d
ev
alu
atio
n
to
class
if
y
an
d
d
etec
t
h
u
m
an
tar
g
ets
in
v
ar
io
u
s
s
ce
n
ar
io
s
.
T
h
is
will
im
p
r
o
v
e
th
e
ab
ilit
y
to
id
en
tify
an
d
d
etec
t
b
o
th
m
o
v
i
n
g
an
d
s
tatio
n
ar
y
h
u
m
a
n
tar
g
ets
b
eh
in
d
walls.
T
h
is
will
b
e
ac
co
m
p
lis
h
ed
b
y
u
s
in
g
an
en
s
em
b
le
alg
o
r
ith
m
f
o
r
class
if
icatio
n
an
d
r
ec
o
g
n
izi
n
g
h
u
m
an
v
ictim
s
co
n
ce
aled
b
eh
in
d
o
b
s
tacle
s
,
as
well
as
I
C
A
f
o
r
(
f
ix
e
d
an
d
d
y
n
am
ic)
FE
.
T
o
ass
ess
th
e
ef
f
ic
ien
cy
o
f
th
e
two
d
atasets
(
s
tatic
an
d
d
y
n
am
ic)
,
a
co
m
p
ar
ativ
e
s
tu
d
y
was a
ls
o
co
n
d
u
cted
.
T
h
e
r
est
o
f
th
e
d
o
cu
m
en
t
is
ar
r
an
g
ed
as
f
o
llo
ws:
t
he
s
e
ctio
n
s
2
an
d
3
,
r
esp
ec
tiv
ely
,
f
o
cu
s
o
n
ca
tast
r
o
p
h
e
m
a
n
ag
em
en
t
in
s
e
ar
ch
an
d
r
escu
e
(
SAR
)
as
wel
l
as
th
e
g
u
id
in
g
p
r
i
n
cip
les
o
f
t
h
e
r
elate
d
ac
tiv
ity
.
T
h
e
s
ec
tio
n
4
d
is
cu
s
s
es
th
e
r
eso
u
r
ce
s
an
d
tech
n
ical
ap
p
r
o
a
ch
es,
as
well
as
th
e
d
ev
elo
p
m
en
t
an
d
tr
ain
in
g
o
f
th
e
d
ata.
T
h
e
s
ec
tio
n
s
5
,
6
,
an
d
7
,
r
esp
ec
tiv
ely
,
p
r
o
v
id
e
d
escr
ip
tio
n
s
o
f
th
e
p
er
f
o
r
m
an
ce
ev
alu
atio
n
,
ap
p
licatio
n
a
n
d
test
in
g
m
eth
o
d
o
lo
g
y
,
an
d
o
u
tco
m
es
a
n
aly
s
is
.
I
n
s
ec
t
io
n
8
,
th
e
co
m
p
ar
is
o
n
a
n
aly
s
is
is
p
r
esen
ted
.
Sectio
n
9
p
r
esen
ts
t
h
e
co
n
clu
s
io
n
.
2.
NL
O
S D
E
T
E
C
T
I
O
N
F
O
R
H
UM
AN
SA
R
O
P
E
RA
T
I
O
NS IN C
A
T
AC
L
YS
M
M
AN
AG
E
M
E
N
T
I
t
is
f
e
as
i
b
l
e
t
o
p
r
e
v
e
n
t
,
p
r
e
p
a
r
e
f
o
r
,
r
e
s
p
o
n
d
t
o
,
r
e
c
o
v
e
r
f
r
o
m
,
a
n
d
l
e
s
s
e
n
t
h
e
e
f
f
e
ct
s
o
f
d
i
s
as
t
e
r
s
t
h
r
o
u
g
h
d
i
s
a
s
t
e
r
m
a
n
a
g
e
m
e
n
t
.
P
r
e
v
e
n
t
i
n
g
d
i
s
a
s
t
e
r
s
b
e
f
o
r
e
t
h
ey
o
c
c
u
r
,
r
e
a
c
t
i
n
g
q
u
i
c
k
l
y
t
o
d
is
a
s
t
e
r
s
,
a
n
d
h
e
l
p
i
n
g
t
o
r
e
b
u
i
l
d
s
o
c
i
e
t
ie
s
a
f
t
e
r
a
d
is
a
s
t
e
r
a
r
e
a
ll
p
a
r
t
s
o
f
e
m
e
r
g
e
n
c
y
m
a
n
a
g
e
m
e
n
t
.
E
v
e
r
y
b
o
d
y
'
s
s
e
c
u
r
i
t
y
d
e
p
e
n
d
s
o
n
e
m
e
r
g
e
n
c
y
m
a
n
a
g
e
m
e
n
t
,
w
h
i
ch
s
h
o
u
l
d
b
e
c
o
n
s
i
d
e
r
e
d
i
n
a
l
l
d
a
i
l
y
d
e
c
i
s
i
o
n
s
r
a
t
h
e
r
t
h
a
n
j
u
s
t
i
n
t
h
e
c
as
e
o
f
a
t
r
a
g
e
d
y
,
w
h
i
c
h
i
s
b
e
c
o
m
i
n
g
a
ll
t
o
o
c
o
m
m
o
n
[
1
3
]
.
E
f
f
e
c
t
i
v
e
em
e
r
g
e
n
c
y
m
a
n
a
g
e
m
e
n
t
o
p
e
r
a
tio
n
s
a
f
t
e
r
a
n
a
t
u
r
al
o
r
m
a
n
-
m
a
d
e
d
i
s
a
s
t
e
r
d
e
p
e
n
d
o
n
s
t
r
o
n
g
c
o
m
m
u
n
i
c
a
t
i
o
n
n
e
t
w
o
r
k
s
.
U
n
f
o
r
t
u
n
a
t
e
l
y
,
l
a
r
g
e
-
s
c
a
l
e
d
i
s
a
s
t
e
r
s
h
a
v
e
t
h
e
p
o
t
e
n
t
i
a
l
t
o
d
is
r
u
p
t
SAR
o
p
e
r
a
ti
o
n
s
a
n
d
d
e
s
t
r
o
y
t
el
e
c
o
m
m
u
n
i
ca
t
i
o
n
s
n
e
tw
o
r
k
s
[
1
4
]
.
R
eso
lv
in
g
th
e
is
s
u
es
n
o
w
ex
is
tin
g
o
n
th
e
g
r
o
u
n
d
is
cr
u
ci
al
d
u
e
to
th
e
g
r
o
wth
o
f
h
u
m
an
itar
ian
ac
tiv
ities
an
d
n
ee
d
s
.
I
s
s
u
es
li
k
e
tr
af
f
ic,
d
ela
y
s
,
u
n
ac
co
u
n
tab
ilit
y
,
an
d
in
a
d
eq
u
ate
co
n
n
ec
t
iv
ity
co
u
l
d
s
er
v
e
as
test
s
ite
s
f
o
r
th
e
alleg
ed
ad
v
an
tag
es
o
f
n
ew
tech
n
o
lo
g
ic
al
ad
v
an
ce
m
en
ts
[
1
5
]
.
T
h
e
cu
r
r
en
t
r
elian
ce
o
n
ce
n
tr
alize
d
p
h
y
s
ical
in
f
r
astru
ctu
r
e
m
eth
o
d
s
p
o
s
es
a
s
ev
er
e
th
r
ea
t
to
th
e
m
an
ag
em
en
t
s
y
s
tem
as
a
wh
o
le.
Fu
r
th
er
m
o
r
e
,
cu
r
r
en
t
p
r
o
to
c
o
ls
f
o
r
co
m
m
u
n
icatio
n
d
u
r
in
g
s
y
s
tem
o
u
tag
es
m
o
s
tly
d
ep
en
d
o
n
u
s
in
g
tem
p
o
r
a
r
y
f
ac
ilit
ies,
in
clu
d
in
g
telec
o
m
to
wer
s
.
Han
d
lin
g
aid
r
e
q
u
ests
b
ased
o
n
c
u
r
r
e
n
t
d
is
aster
in
f
o
r
m
atio
n
an
d
ef
f
icien
tly
r
esp
o
n
d
i
n
g
to
s
u
ch
r
eq
u
ests
b
y
d
is
tr
ib
u
tin
g
th
e
m
an
ag
em
en
t'
s
av
ailab
le
lim
i
ted
r
eso
u
r
ce
s
is
th
e
m
ain
g
o
al
o
f
th
e
en
tire
p
r
o
ce
s
s
[
1
6
]
.
B
lo
ck
ch
ain
-
b
ased
tech
n
o
lo
g
ies
with
s
o
p
h
is
ticated
f
e
atu
r
es
ca
n
s
u
p
p
o
r
t
SAR
o
p
er
atio
n
s
b
y
lev
er
ag
i
n
g
ad
v
a
n
ce
m
en
ts
in
in
f
o
r
m
atio
n
an
d
co
m
m
u
n
icatio
n
s
tec
h
n
o
lo
g
y
(
I
C
T
)
,
d
r
o
n
es,
th
e
in
ter
n
et
o
f
th
in
g
s
(
I
o
T
)
,
clo
u
d
-
b
ased
co
m
p
u
tin
g
,
im
ag
e
a
n
aly
s
is
,
an
d
au
to
n
o
m
o
u
s
ae
r
ia
l sy
s
tem
s
[
1
7
]
.
Ho
wev
er
,
b
lo
c
k
ch
ain
,
I
o
T
,
a
n
d
cr
o
wd
s
o
u
r
cin
g
m
et
h
o
d
s
ca
n
o
f
f
er
v
al
u
ab
le
in
s
ig
h
ts
th
at
ca
n
f
ac
ilit
ate
th
e
p
r
o
v
is
io
n
in
g
p
r
o
ce
s
s
.
T
h
e
y
m
ig
h
t
also
b
e
a
m
ea
n
s
o
f
f
o
s
ter
in
g
a
d
y
n
am
ic
m
u
tu
al
t
r
u
s
t
b
etwe
en
p
eo
p
le
wh
o
p
r
o
v
id
e,
r
ec
eiv
e,
an
d
s
ee
k
h
elp
to
e
n
co
u
r
ag
e
th
is
im
p
r
o
v
em
en
t.
Fu
r
th
e
r
m
o
r
e
,
it c
an
b
e
u
s
ed
as a
b
asis
f
o
r
in
co
r
p
o
r
atin
g
tech
n
o
lo
g
ies
s
u
ch
as
b
lo
ck
ch
ai
n
,
th
r
ee
-
d
im
en
s
io
n
al
p
r
in
tin
g
,
an
d
ar
tif
icial
in
tellig
en
ce
to
im
p
r
o
v
e
t
h
e
f
lo
w
o
f
in
f
o
r
m
ati
o
n
,
p
r
o
d
u
cts,
an
d
f
u
n
d
s
in
h
u
m
an
itar
ian
s
u
p
p
l
y
ch
ain
s
[
1
5
]
.
3.
L
I
T
E
R
AT
U
RE
R
E
VI
E
W
T
h
e
u
s
e
o
f
m
ac
h
in
e
lear
n
in
g
tech
n
iq
u
es
to
id
en
tify
an
d
lo
ca
te
v
ictim
s
tr
ap
p
ed
b
eh
i
n
d
co
llap
s
ed
s
tr
u
ctu
r
es
h
as
g
ar
n
er
ed
a
lo
t
o
f
atten
tio
n
d
u
r
in
g
th
e
last
2
0
y
ea
r
s
.
Ho
wev
er
,
o
b
s
tacle
s
in
th
e
lin
e
o
f
s
ig
h
t
(
L
OS)
b
etwe
en
th
e
s
en
d
in
g
a
n
d
th
e
r
ec
ei
v
in
g
d
ev
ice
ad
v
er
s
ely
im
p
air
b
asic
m
ea
s
u
r
em
e
n
ts
o
f
r
ad
io
s
ig
n
als
.
T
h
ese
m
ea
s
u
r
em
en
ts
,
lik
e
r
ec
eiv
ed
s
ig
n
al
s
tr
en
g
th
in
d
icato
r
(
R
SS
I
)
v
alu
es,
r
en
d
er
in
g
th
e
m
in
ap
p
r
o
p
r
iate
f
o
r
th
e
co
n
d
itio
n
s
u
n
d
e
r
in
v
esti
g
a
tio
n
[
1
8
]
.
Yu
et
a
l.
[
1
9
]
p
r
o
p
o
s
ed
a
te
ch
n
iq
u
e
th
at
u
tili
ze
s
a
h
ig
h
er
-
lev
el
cy
clo
s
tatio
n
ar
ity
to
d
et
ec
t
h
u
m
an
r
esp
ir
atio
n
a
n
d
p
u
ls
e
to
g
et
b
ey
o
n
d
th
ese
r
estrictio
n
s
.
B
y
em
p
lo
y
in
g
th
e
t
h
ir
d
-
o
r
d
er
c
y
clic
cu
m
u
la
n
t,
th
is
tech
n
iq
u
e
e
f
f
ec
tiv
ely
r
ed
u
ce
s
h
ar
m
o
n
ic
in
ter
m
o
d
u
latio
n
,
r
a
n
d
o
m
b
o
d
y
m
o
tio
n
s
,
an
d
clu
t
ter
n
o
is
e,
allo
win
g
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
P
r
ed
ictin
g
tr
a
p
p
ed
victims in
d
eb
r
is
u
s
in
g
s
ig
n
a
l a
n
a
lysi
s
en
s
emb
le
cla
s
s
ifica
tio
n
(
E
n
o
ch
A
d
a
ma
Jiya
)
495
r
ad
ar
s
en
s
o
r
s
to
d
etec
t
wea
k
s
ig
n
als
with
lo
w
s
ig
n
al
-
to
-
n
o
is
e
r
atio
s
(
SNR
)
.
T
h
ese
d
ev
el
o
p
m
en
ts
ar
e
ess
en
tial
f
o
r
tar
g
et
m
o
n
ito
r
in
g
in
m
ilit
ar
y
an
d
em
er
g
e
n
cy
r
esp
o
n
s
e
o
p
er
atio
n
s
.
B
y
b
r
ea
k
in
g
d
o
w
n
r
esp
ir
ato
r
y
s
ig
n
als
in
to
d
is
tin
ct
s
u
b
-
s
ig
n
als,
v
ar
iatio
n
al
m
o
d
e
d
ec
o
m
p
o
s
itio
n
(
VM
D)
h
as
also
b
ee
n
u
s
ed
to
tr
ac
k
a
v
ar
iety
o
f
o
b
jects
b
eh
in
d
walls.
Fo
r
m
icr
o
wav
e
r
ad
ar
d
ev
ices,
th
is
ap
p
r
o
ac
h
—
wh
ich
u
s
es
th
e
Hilb
er
t
tr
an
s
f
o
r
m
,
d
is
tan
ce
b
in
s
,
an
d
tr
a
v
elin
g
r
esp
ir
ato
r
y
id
en
tific
atio
n
ca
lcu
latio
n
s
—
p
er
f
o
r
m
e
d
b
ett
er
th
an
co
n
v
en
tio
n
al
f
ast
Fo
u
r
ier
tr
an
s
f
o
r
m
(
FFT
)
m
eth
o
d
s
[
2
0
]
.
T
o
in
cr
ea
s
e
g
en
er
aliza
tio
n
a
b
ilit
y
u
n
d
er
a
v
ar
iety
o
f
NL
OS
cir
cu
m
s
t
an
ce
s
,
m
ac
h
in
e
lear
n
in
g
tech
n
iq
u
es
h
a
v
e
b
ee
n
cr
ea
ted
.
Fo
r
ex
am
p
le,
Kh
an
et
a
l.
[
2
1
]
u
s
ed
W
i
-
Fi
s
en
s
o
r
s
to
cr
ea
te
a
C
NN
th
at
clas
s
if
ied
an
d
ass
es
s
ed
h
u
m
an
r
esp
ir
ato
r
y
ac
tiv
ity
with
a
9
4
.
8
5
% a
cc
u
r
ac
y
r
ate.
T
h
e
m
o
s
t
r
ec
en
t
d
ev
elo
p
m
en
t
s
in
UW
B
th
r
o
u
g
h
-
wall
r
ad
ar
r
esear
ch
ar
e
co
v
er
e
d
in
d
etai
l,
with
an
em
p
h
asis
o
n
s
ig
n
al
-
p
r
o
ce
s
s
in
g
m
eth
o
d
s
f
o
r
m
o
n
ito
r
in
g
v
it
al
s
ig
n
s
an
d
id
e
n
tify
in
g
p
eo
p
le.
T
h
r
ee
p
r
im
ar
y
m
eth
o
d
s
f
o
r
h
u
m
an
d
etec
tio
n
with
NL
OS
s
en
s
o
r
s
wer
e
d
eter
m
in
ed
.
T
h
ese
ar
e:
i)
m
u
lti
p
ath
p
atter
n
s
o
f
th
e
r
ad
ar
r
e
f
lecte
d
s
ig
n
al;
ii)
n
u
m
er
ical
ch
ar
ac
ter
is
tics
o
f
th
e
r
e
ce
iv
ed
s
ig
n
al,
s
u
c
h
as
s
k
ewn
e
s
s
an
d
k
u
r
t
o
s
is
;
an
d
iii)
co
n
s
tan
t f
alse a
lar
m
r
ate
(
C
FA
R
)
,
wh
ich
d
eter
m
in
es e
n
e
r
g
y
lev
els f
o
r
tar
g
et
r
esp
o
n
s
es.
Par
k
et
a
l.
[
2
2
]
in
v
esti
g
ated
a
tr
an
s
f
er
lear
n
in
g
-
b
ased
UW
B
NL
OS
d
etec
tio
n
tech
n
iq
u
e
th
a
t
ac
h
iev
ed
p
r
ec
is
io
n
o
n
p
ar
with
d
ee
p
lea
r
n
in
g
m
eth
o
d
s
d
ev
elo
p
ed
with
s
p
ar
s
e
d
ata.
Oth
er
d
esig
n
s
an
d
h
u
m
an
d
etec
tio
n
task
s
wer
e
n
o
t
ex
am
in
ed
in
t
h
eir
s
tu
d
y
,
th
o
u
g
h
.
Similar
ly
,
two
r
ec
eiv
in
g
an
te
n
n
as
wer
e
u
tili
ze
d
to
m
ea
s
u
r
e
v
ital
s
ig
n
s
in
th
e
in
v
esti
g
atio
n
[
2
3
]
.
B
u
t
th
ey
o
n
ly
d
ec
id
ed
t
o
lo
o
k
at
th
e
s
tr
o
n
g
s
ig
n
al.
T
h
e
m
o
b
ile
f
ilter
was
s
u
b
s
eq
u
en
tly
f
itted
to
th
e
s
elec
ted
s
ig
n
al,
wh
ich
elim
in
ated
th
e
q
u
asi
-
s
tatic
n
o
is
e.
Du
al
-
f
r
eq
u
en
cy
h
ar
m
o
n
ic
co
n
tin
u
o
u
s
wav
e
(
CW
)
r
ad
ar
s
en
s
in
g
s
y
s
tem
s
h
av
e
b
ee
n
u
s
ed
in
o
th
er
s
tu
d
ies
to
in
cr
ea
s
e
SNR
an
d
d
ec
r
ea
s
e
f
lick
er
in
ter
f
er
e
n
ce
[
2
4
]
.
Desp
ite
th
ese
ad
v
an
ce
m
en
ts
,
ch
alle
n
g
es
s
u
ch
as
th
e
ze
r
o
-
p
o
in
t p
h
en
o
m
en
o
n
,
m
o
tio
n
ar
tifa
cts,
an
d
in
ter
f
er
e
n
ce
f
r
o
m
m
u
ltip
le
s
o
u
r
ce
s
s
till
af
f
ec
t
th
e
ef
f
icac
y
o
f
m
icr
o
wa
v
e
r
ad
ar
s
en
s
o
r
s
in
d
etec
tin
g
v
ital sig
n
s
[
2
5
]
,
[
2
6
]
.
T
h
is
r
esear
ch
aim
s
to
ad
d
r
ess
th
ese
ch
allen
g
es
b
y
ex
am
in
in
g
r
ad
ar
an
d
co
m
m
u
n
icatio
n
tech
n
o
lo
g
ies
f
o
r
S
A
R
o
p
er
atio
n
s
.
T
h
e
r
ev
ie
w
co
v
er
s
co
n
tin
u
o
u
s
C
W
UW
B
r
ad
ar
s
en
s
o
r
s
,
th
eir
o
p
e
r
atio
n
al
p
r
in
cip
les,
an
d
th
eir
p
h
y
s
ical
co
n
s
tr
u
ctio
n
.
T
h
e
co
n
clu
s
io
n
a
n
d
f
u
tu
r
e
o
u
tlo
o
k
ar
e
p
r
esen
ted
at
th
e
en
d
o
f
t
h
is
p
ap
er
.
4.
M
AT
E
R
I
AL
S AN
D
M
E
T
H
O
D
B
o
th
s
tatio
n
ar
y
an
d
d
y
n
am
ic
d
ata
wer
e
in
clu
d
ed
i
n
th
e
d
ata
s
ets
u
s
ed
f
o
r
th
is
in
v
esti
g
atio
n
.
Fig
u
r
e
1
p
r
o
v
id
es
a
s
u
m
m
ar
y
o
f
th
e
s
u
g
g
ested
f
r
am
ewo
r
k
.
T
h
e
p
h
as
e
o
f
o
b
tain
in
g
p
e
r
tin
en
t
in
f
o
r
m
atio
n
f
r
o
m
a
g
iv
e
n
d
ataset
ch
an
g
es
th
e
p
r
o
ce
d
u
r
e
wh
en
u
tili
zin
g
t
h
e
I
C
A
FE
a
p
p
r
o
ac
h
es
f
o
r
b
o
th
s
tatio
n
ar
y
an
d
f
ix
e
d
d
atasets
.
T
h
e
r
esu
lts
o
f
th
e
en
s
em
b
le
class
if
icatio
n
ap
p
r
o
ac
h
ar
e
co
n
tr
asted
with
cu
r
r
en
t
ap
p
r
o
ac
h
es
to
ev
alu
ate
th
e
NL
OS d
ataset
'
s
p
er
f
o
r
m
an
ce
.
Fig
u
r
e
1
.
Pro
p
o
s
ed
f
r
am
ewo
r
k
T
h
e
NL
OS
d
atasets
u
s
ed
in
th
is
wo
r
k
in
clu
d
ed
1
7
,
4
0
8
(
s
tatio
n
er
y
)
a
n
d
2
3
,
5
5
2
(
d
y
n
am
ic)
ca
s
es,
with
2
5
6
s
am
p
les
p
er
win
d
o
w.
Stan
d
ar
d
s
ca
lar
p
r
ep
r
o
ce
s
s
in
g
o
r
SC
is
ty
p
ically
u
s
ed
f
o
r
d
ata
c
lean
in
g
to
m
ak
e
th
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
1
5
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
6
:
493
-
5
0
5
496
im
p
u
ted
p
r
im
ar
y
d
ata
u
n
if
o
r
m
f
o
r
p
r
ec
is
e
p
r
ed
ictio
n
.
T
o
id
en
tify
th
e
cr
u
cial
d
is
cr
e
p
an
cy
with
f
ewe
r
in
d
ep
en
d
en
t
co
m
p
o
n
en
ts
an
d
o
f
f
er
s
u
b
s
tan
tial
ev
id
en
ce
o
f
m
o
r
e
r
esear
ch
,
th
e
I
C
A
m
eth
o
d
was
u
tili
ze
d
to
r
em
o
v
e
in
ac
tiv
e
elem
e
n
ts
f
r
o
m
th
e
UW
B
NL
OS
d
ata.
T
h
e
I
C
A
FE
p
r
o
ce
s
s
is
o
lates
an
d
r
em
o
v
es
co
n
f
lictin
g
v
ar
iab
les.
T
h
e
d
ata
c
o
n
tain
in
g
r
ec
o
v
e
r
ed
I
C
A
laten
t
s
ig
n
if
ican
t
f
ea
tu
r
es
is
class
if
ied
u
s
in
g
th
e
en
s
em
b
le
Ad
aBo
o
s
t
ca
teg
o
r
izatio
n
tech
n
iq
u
e.
E
n
s
em
b
le
class
if
ier
s
ar
e
u
tili
ze
d
to
ev
alu
ate
th
e
ef
f
ic
ac
y
o
f
class
if
icatio
n
ap
p
licatio
n
s
,
em
p
lo
y
in
g
s
ev
e
n
ty
p
er
ce
n
t
o
f
th
e
d
ata
f
o
r
tr
ain
i
n
g
an
d
30%
f
o
r
ev
al
u
atin
g
p
r
e
d
ictio
n
ac
cu
r
ac
y
.
T
o
ass
ess
th
e
d
ata
an
al
y
s
is
le
a
r
n
in
g
e
n
a
ct
m
e
n
t
m
et
h
o
d
o
l
o
g
i
e
s
,
th
e
c
o
ll
ec
t
ed
d
a
ta
was
t
ak
e
n
f
r
o
m
t
h
e
m
ea
s
u
r
e
m
e
n
t
d
at
a
th
at
is
p
u
b
lic
ly
av
ail
ab
le
.
T
h
e
r
e
p
o
s
i
to
r
y
is
l
o
c
at
ed
a
t
‘
h
tt
p
s
:/
/g
it
h
u
b
.
c
o
m
/
d
is
i
u
n
i
b
o
-
n
l
u
/
u
w
b
-
n
l
o
s
-
h
u
m
a
n
-
d
et
ec
t
io
n
”.
Ma
n
y
m
a
ch
i
n
e
le
ar
n
i
n
g
t
ec
h
n
iq
u
es
h
a
v
e
b
e
en
r
e
p
li
ca
t
ed
t
o
an
al
y
z
e
an
d
p
r
e
d
ic
t
N
L
OS
h
u
m
a
n
l
o
c
ali
za
tio
n
d
at
a
o
f
s
e
v
e
r
al
c
o
m
p
le
x
m
at
er
ials
.
A
c
o
n
v
en
ti
o
n
al
s
c
a
ler
w
as
a
p
p
lie
d
t
o
h
u
m
an
-
d
et
ec
t
a
b
le
f
ea
t
u
r
es
,
o
b
s
tacl
es,
o
b
j
ec
ts
wit
h
d
i
f
f
e
r
e
n
t
an
g
l
es,
a
n
d
o
b
j
ec
ts
at
d
i
f
f
e
r
e
n
t
d
is
ta
n
c
es
f
r
o
m
t
h
e
co
l
le
cte
d
d
atas
et
[
2
7
]
.
Fo
r
b
ett
er
p
r
ed
icti
o
n
,
t
h
e
o
u
t
p
u
t s
h
o
ws
a
m
o
r
e
c
o
n
d
en
s
e
d
an
d
f
ilt
er
ed
c
o
ll
ec
t
io
n
o
f
d
at
a.
E
x
p
er
im
en
ts
wer
e
co
n
d
u
cte
d
in
a
v
ar
iety
o
f
in
d
o
o
r
l
o
ca
tio
n
s
o
n
th
e
f
ir
s
t
f
lo
o
r
o
f
th
e
U
n
iv
er
s
ity
o
f
B
o
lo
g
n
a'
s
Sch
o
o
l
o
f
E
n
g
in
ee
r
in
g
'
s
C
esen
a
C
am
p
u
s
.
First,
c
o
n
s
id
er
atio
n
was
g
iv
en
to
th
e
f
ix
ed
m
ea
s
u
r
em
en
t,
wh
ich
en
tails
m
ain
tain
i
n
g
th
e
r
ad
ar
m
o
tio
n
less
o
n
a
r
o
llin
g
wag
o
n
at
a
d
is
tan
ce
o
f
ab
o
u
t
1
3
0
cm
f
r
o
m
th
e
ea
r
th
.
T
h
e
tar
g
eted
o
b
ject
(
d
)
,
wh
ich
was
p
lace
d
2
0
cm
af
ter
th
e
o
b
s
tr
u
ctio
n
s
,
was
3
0
,
6
0
,
a
n
d
9
0
c
m
f
r
o
m
th
e
r
ad
ar
(
r
)
.
A
n
o
th
er
p
o
ten
tial o
b
s
tacle
was
th
e
th
ick
n
ess
o
f
th
e
m
ater
ial.
Data
co
llectio
n
ex
p
l
icitly
co
n
s
id
er
s
th
e
f
o
llo
win
g
m
ater
ials
:
a
cr
y
s
tal
f
r
am
e
s
et
at
2
c
m
,
a
b
r
ick
wa
ll
s
et
at
1
5
cm
,
w
o
o
d
e
n
g
ates
s
et
at
3
an
d
5
c
m
,
an
d
d
o
u
b
le
-
g
lazin
g
s
et
at
1
0
cm
.
A
m
o
r
e
r
ea
lis
tic
s
ce
n
ar
io
is
p
r
esen
ted
in
th
e
s
ec
o
n
d
ex
am
p
le,
wh
er
e
th
e
r
ad
ar
was
h
an
d
led
at
d
if
f
er
e
n
t
h
eig
h
ts
,
ca
u
s
in
g
th
e
ac
q
u
is
itio
n
to
b
e
d
y
n
am
ic
v
ia
s
m
all
m
o
tio
n
s
[
2
7
]
.
C
o
n
s
eq
u
en
tly
,
th
e
f
o
llo
win
g
is
a
s
u
m
m
ar
y
o
f
th
e
r
esear
ch
s
tr
ateg
y
:
Ma
k
e
a
d
ataset
o
f
UW
B
NL
OS
u
s
in
g
th
e
m
ater
ials
o
f
d
if
f
e
r
en
t
b
ar
r
ie
r
s
an
d
th
e
b
o
d
y
o
r
ie
n
tatio
n
s
o
f
d
if
f
er
en
t
v
ictim
s
.
i)
Descr
ib
e
h
o
w
d
ata
n
o
r
m
aliza
tio
n
im
p
ac
ts
th
e
d
ataset'
s
ac
cu
r
ac
y
.
ii)
Use I
C
A
FE
tech
n
iq
u
es to
id
en
tify
th
e
laten
t c
o
m
p
o
n
en
t in
t
h
e
d
ataset.
i
i
i
)
D
e
s
c
r
i
b
e
a
c
l
a
s
s
i
f
i
e
r
t
r
a
i
n
i
n
g
a
p
p
r
o
a
c
h
t
h
a
t
i
n
c
r
e
a
s
e
s
p
r
e
d
i
c
t
i
o
n
a
c
c
u
r
a
c
y
b
y
l
e
v
e
r
a
g
i
n
g
e
n
s
e
m
b
l
e
c
l
a
s
s
i
f
i
c
a
t
i
o
n
.
iv
)
Ma
k
in
g
ca
teg
o
r
izatio
n
s
u
g
g
esti
o
n
s
u
s
in
g
th
e
en
s
em
b
le
class
i
f
ier
.
v)
Usi
n
g
p
er
tin
en
t liter
atu
r
e
t
o
co
m
p
ar
e
th
e
o
u
tco
m
es to
alter
n
a
tiv
e
ap
p
r
o
ac
h
es.
4
.
1
.
M
et
ho
ds
MA
T
L
AB
was
u
tili
ze
d
to
e
x
a
m
in
e
th
e
in
f
o
r
m
atio
n
g
ath
e
r
ed
f
r
o
m
[
2
7
]
ex
p
er
im
en
tally
,
an
d
I
C
A
was
u
s
ed
to
ex
tr
ac
t
f
ea
tu
r
es.
R
etr
iev
ab
le
ch
ar
ac
ter
is
tics
wer
e
u
s
ed
f
o
r
class
if
icatio
n
u
tili
zin
g
an
e
n
s
em
b
le
alg
o
r
ith
m
ic
tec
h
n
iq
u
e
.
T
o
h
o
m
o
g
en
ize
co
llected
d
ata,
elim
i
n
ate
n
o
is
y
v
alu
es
a
n
d
o
u
tlier
s
,
tr
an
s
f
o
r
m
th
e
d
ata,
an
d
r
eg
u
lar
ize
it,
th
e
s
tu
d
y
e
m
p
lo
y
s
SC
.
T
h
e
g
o
al
o
f
SC
is
to
tr
an
s
f
o
r
m
d
if
f
er
e
n
t
eig
en
v
alu
es
in
to
a
p
r
eset
r
an
g
e
o
f
ze
r
o
s
a
n
d
o
n
es.
B
y
u
s
in
g
item
s
ca
lin
g
,
th
is
m
eth
o
d
m
ak
es
s
u
r
e
th
at
f
ea
tu
r
es
ar
e
a
b
o
u
t
th
e
s
am
e
s
ize,
wh
ich
m
ak
es
th
em
m
o
r
e
m
a
n
ag
ea
b
le
f
o
r
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
s
.
T
h
e
SC
n
o
r
m
a
lizes
elem
en
ts
,
b
y
r
em
o
v
in
g
th
e
m
e
d
ian
an
d
n
o
r
m
alizin
g
th
e
v
ar
ian
ce
to
o
n
e,
ac
h
iev
in
g
u
n
it
v
ar
ian
ce
b
y
d
iv
id
i
n
g
ev
e
r
y
p
ar
am
eter
b
y
th
e
s
tan
d
a
r
d
d
ev
iatio
n
.
I
n
th
e
ca
s
e
o
f
Gau
s
s
ian
-
d
is
tr
ib
u
ted
d
ata,
t
h
is
s
tan
d
ar
d
izatio
n
—
also
k
n
o
wn
as
n
o
r
m
aliza
tio
n
—
is
esp
ec
ially
h
elp
f
u
l
s
in
ce
it
m
ak
es
m
ac
h
in
e
lear
n
in
g
tec
h
n
iq
u
es
ea
s
ier
to
h
an
d
le.
Stan
d
ar
d
izatio
n
m
a
y
h
a
v
e
a
n
im
p
ac
t
o
n
d
ata
o
u
tlier
s
s
in
ce
it
d
o
esn
'
t
h
av
e
a
b
o
u
n
d
ar
y
r
a
n
g
e,
u
n
lik
e
n
o
r
m
aliza
tio
n
.
N
o
r
m
alizin
g
d
ata,
h
o
wev
er
,
m
ig
h
t
n
o
t a
lway
s
b
e
n
ec
ess
ar
y
[
2
8
]
.
W
h
i
l
e
l
e
a
r
n
i
n
g
t
h
e
f
e
a
t
u
r
e
r
e
p
r
e
s
e
n
t
a
t
i
o
n
,
t
h
e
e
n
s
e
m
b
l
e
i
s
u
t
i
l
i
z
e
d
t
o
a
n
i
m
a
t
e
d
l
y
o
p
t
i
m
i
z
e
b
o
t
h
t
h
e
f
e
a
t
u
r
e
s
a
n
d
t
h
e
c
l
a
s
s
i
f
i
e
r
s
.
T
h
e
s
u
g
g
e
s
t
e
d
s
y
s
t
e
m
c
o
n
s
i
s
t
s
o
f
m
o
d
u
l
e
s
f
o
r
F
E
,
c
l
a
s
s
i
f
i
c
a
t
i
o
n
,
r
e
s
u
l
t
e
x
t
r
a
c
t
i
o
n
,
a
n
d
d
a
t
a
s
e
t
l
o
a
d
i
n
g
.
A
f
t
e
r
n
o
r
m
a
l
i
z
i
n
g
t
h
e
d
a
t
a
s
e
t
i
n
t
h
e
p
e
r
s
o
n
d
e
t
e
c
t
i
o
n
m
o
d
u
l
e
,
t
h
e
F
E
u
n
i
t
i
m
p
o
r
t
s
i
t
a
n
d
r
u
n
s
t
h
e
I
C
A
a
l
g
o
r
i
t
h
m
o
n
i
t
.
T
h
e
F
E
m
o
d
u
l
e
r
e
c
e
i
v
e
s
t
h
e
o
u
t
p
u
t
m
o
d
u
l
e
a
n
d
a
p
p
l
i
e
s
F
E
o
n
i
t
s
o
w
n
.
T
h
e
r
e
s
p
o
n
s
e
i
s
s
h
o
w
n
i
n
t
h
e
o
u
t
c
o
m
e
m
o
d
u
l
e
o
n
c
e
t
h
e
c
l
a
s
s
i
f
i
c
a
t
i
o
n
p
r
o
c
e
s
s
u
s
i
n
g
E
n
s
e
m
b
l
e
h
a
s
c
o
r
r
e
c
t
l
y
c
l
a
s
s
i
f
i
e
d
i
t
.
S
C
f
e
a
t
u
r
e
s
e
l
e
c
t
i
o
n
i
n
N
L
O
S
d
a
t
a
a
n
a
l
y
s
i
s
i
s
c
o
m
b
i
n
e
d
w
i
t
h
t
h
e
F
E
t
e
c
h
n
i
q
u
e
s
"
I
C
A
"
a
n
d
"
e
n
s
e
m
b
l
e
.
"
T
h
ese
tech
n
iq
u
es
will
b
e
co
m
b
in
ed
an
d
u
s
ed
to
d
e
v
elo
p
a
m
o
d
el
o
f
ev
alu
atio
n
m
ea
s
u
r
es.
T
h
e
ar
ticle's
m
eth
o
d
o
lo
g
y
is
as
f
o
llo
ws
.
W
e
as
s
es
s
ed
o
u
r
r
esu
lt
s
in
ter
m
s
o
f
ac
cu
r
ac
y
,
p
r
ec
is
io
n
,
F1
-
s
co
r
e,
a
n
d
s
en
s
itiv
ity
am
o
n
g
o
t
h
er
s
b
y
u
tili
zin
g
en
s
em
b
le
class
if
icatio
n
,
SC
d
ata
s
tan
d
ar
d
izati
o
n
,
an
d
I
C
A
FE
to
en
h
an
ce
th
e
ef
f
ec
tiv
en
ess
o
f
cl
ass
if
icatio
n
.
4
.
2
.
M
a
t
er
ia
ls
T
o
en
h
a
n
ce
th
e
ef
f
ec
tiv
e
n
ess
o
f
th
e
h
u
m
a
n
p
r
esen
ce
d
e
tectio
n
d
ataset,
th
is
s
tu
d
y
ex
p
lo
r
es
th
e
ap
p
licatio
n
o
f
en
s
em
b
le
class
if
icatio
n
an
d
I
C
A
f
o
r
d
im
e
n
s
io
n
ality
r
ed
u
ctio
n
o
f
m
u
ltifa
ce
te
d
NL
OS
d
ata.
T
h
e
tr
ain
in
g
p
r
o
ce
s
s
es
u
s
ed
,
s
u
p
er
v
is
ed
lear
n
in
g
tech
n
i
q
u
es
th
at
ad
h
er
e
to
th
e
tr
ain
in
g
d
at
aset
'
s
o
r
g
an
izatio
n
,
wh
ich
n
o
r
m
ally
tak
e
d
esig
n
ate
d
in
s
tan
ce
s
as in
p
u
t.
C
o
n
s
eq
u
e
n
tly
,
f
o
r
ev
er
y
in
s
tan
ce
,
th
e
cl
ass
o
f
co
n
ce
r
n
will
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
P
r
ed
ictin
g
tr
a
p
p
ed
victims in
d
eb
r
is
u
s
in
g
s
ig
n
a
l a
n
a
lysi
s
en
s
emb
le
cla
s
s
ifica
tio
n
(
E
n
o
ch
A
d
a
ma
Jiya
)
497
b
e
d
eter
m
in
ed
.
th
is
is
d
o
n
e
b
y
lab
elin
g
wav
ef
o
r
m
in
s
tan
ce
s
as
eith
er
"
p
er
s
o
n
p
r
esen
t"
o
r
"p
er
s
o
n
a
b
s
en
t."
T
h
is
allo
ws
th
e
last
m
o
d
el
to
ca
teg
o
r
ize
f
u
r
t
h
er
p
r
ep
r
o
ce
s
s
ed
h
ar
m
o
n
ics
in
to
a
p
ar
ticu
la
r
o
f
th
e
f
o
llo
win
g
s
ce
n
ar
io
s
.
T
o
d
r
asti
ca
lly
lo
wer
th
e
d
im
en
s
io
n
ality
o
f
h
u
m
an
d
etec
tio
n
d
ata,
th
is
s
tu
d
y
co
n
s
id
er
s
I
C
A
an
d
en
s
em
b
le
class
if
icatio
n
alg
o
r
it
h
m
s
.
An
I
-
J
m
atr
i
x
M
is
u
s
ed
to
d
is
p
lay
two
d
atasets
.
T
h
er
e
is
a
to
tal
o
f
2
3
,
5
5
2
in
s
tan
ce
s
in
th
e
s
tatic
s
itu
ati
o
n
an
d
1
7
,
4
0
8
in
th
e
d
y
n
a
m
ic
ca
s
e.
B
o
th
s
itu
atio
n
s
ap
p
ly
:
J
=2
5
6
(
K)
.
T
o
d
eter
m
in
e
th
e
o
v
er
all
n
u
m
b
e
r
o
f
r
o
ws
g
e
n
er
ated
b
y
co
m
b
in
in
g
ea
c
h
p
u
ls
e
s
et
f
r
o
m
t
h
e
d
ata
c
o
llectio
n
in
q
u
ir
y
,
a
q
u
alitativ
e
an
aly
s
i
s
was
u
tili
ze
d
to
r
em
o
v
e
th
e
in
ac
cu
r
ate
p
ar
ts
(
i.e
.
,
th
o
s
e
with
s
ig
n
if
ican
t
d
is
p
lace
m
en
t
an
d
lo
w
d
ata
co
n
ten
t)
[
2
7
]
,
[
2
9
]
.
Data
f
r
o
m
t
h
e
C
esen
a
C
am
p
u
s
Sch
o
o
l
o
f
E
n
g
in
ee
r
in
g
at
th
e
Un
iv
er
s
ity
o
f
B
o
lo
g
n
a.
T
ab
le
1
g
iv
e
a
th
o
r
o
u
g
h
o
v
er
v
iew
o
f
th
e
d
ataset,
in
clu
d
i
n
g
f
ea
tu
r
e
s
am
p
le
o
cc
u
r
r
e
n
ce
s
an
d
s
am
p
lin
g
ch
a
r
ac
ter
is
tics
.
T
h
e
s
p
ec
if
ic
ch
ar
ac
ter
is
tics
o
f
th
e
d
ataset
u
tili
ze
d
in
th
is
r
esear
ch
ar
e
s
h
o
wn
in
T
a
b
le
1
.
T
ab
le
1
.
Featu
r
es o
f
th
e
d
atase
t
D
a
t
a
s
e
t
f
e
a
t
u
r
es
D
e
scri
p
t
i
o
n
M
o
v
i
n
g
r
a
d
a
r
2
3
,
5
5
2
S
t
a
t
i
o
n
e
r
y
r
a
d
a
r
1
7
,
4
0
8
O
b
serv
a
t
i
o
n
2
5
6
S
o
u
r
c
e
U
n
i
v
e
r
si
t
y
o
f
B
o
l
o
g
n
a
's
C
e
s
e
n
a
C
a
m
p
u
s
C
h
a
r
a
c
t
e
r
i
s
t
i
c
s
H
u
ma
n
b
o
d
y
a
l
i
g
n
me
n
t
,
c
o
n
st
r
u
c
t
i
o
n
mat
e
r
i
a
l
s,
r
u
b
b
l
e
s,
a
n
d
se
n
s
o
r
d
i
st
a
n
c
e
s
A
c
c
e
ss
i
b
i
l
i
t
y
P
u
b
l
i
c
l
y
a
v
a
i
l
a
b
l
e
d
a
t
a
se
t
[
2
7
]
4
.
2
.
1
.
Dim
ens
io
na
lity
re
du
ct
io
n
R
e
g
a
r
d
i
n
g
c
o
m
p
u
t
a
t
i
o
n
a
l
i
n
tr
i
c
a
c
y
,
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
,
a
n
d
s
t
a
t
is
t
i
ca
l
e
v
a
l
u
a
ti
o
n
,
d
i
m
e
n
s
i
o
n
a
l
it
y
r
e
d
u
c
t
i
o
n
t
e
c
h
n
i
q
u
es
a
r
e
es
s
e
n
ti
a
l
m
et
h
o
d
s
f
o
r
o
v
e
r
c
o
m
i
n
g
t
h
e
d
i
f
f
i
c
u
l
ti
e
s
p
r
e
s
e
n
t
e
d
b
y
la
r
g
e
-
s
c
a
l
e
d
a
t
a
f
i
l
es
.
T
o
l
e
s
s
e
n
t
h
e
c
u
r
s
e
o
f
d
i
m
e
n
s
i
o
n
a
li
t
y
,
t
h
es
e
m
e
t
h
o
d
s
s
e
e
k
t
o
d
e
c
r
e
a
s
e
t
h
e
n
u
m
b
e
r
o
f
k
e
y
v
a
r
i
a
b
le
s
b
ei
n
g
e
x
a
m
i
n
e
d
.
D
i
m
e
n
s
i
o
n
r
e
d
u
c
t
i
o
n
is
f
r
e
q
u
e
n
t
l
y
u
s
e
d
as
a
p
r
e
p
r
o
c
e
s
s
i
n
g
s
t
e
p
b
e
f
o
r
e
u
s
i
n
g
u
n
s
u
p
e
r
v
is
e
d
t
e
c
h
n
i
q
u
e
s
s
u
c
h
a
s
c
l
u
s
t
e
r
i
n
g
a
l
g
o
r
i
t
h
m
s
[
3
0
]
.
B
y
r
e
m
o
v
i
n
g
m
u
l
t
i
c
o
l
l
i
n
e
a
r
i
t
i
es
,
d
i
m
e
n
s
i
o
n
a
li
t
y
r
e
d
u
c
ti
o
n
m
a
k
e
s
i
t
e
a
s
i
e
r
to
u
n
d
e
r
s
t
a
n
d
h
o
w
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
a
l
g
o
r
i
t
h
m
s
r
e
a
d
i
n
p
u
t
i
n
f
o
r
m
a
t
i
o
n
,
w
h
i
c
h
l
o
w
e
r
s
t
h
e
c
o
m
p
u
t
i
n
g
t
i
m
e
a
n
d
s
p
a
c
e
n
e
e
d
e
d
,
a
n
d
p
e
r
m
i
ts
d
is
p
l
a
y
o
f
d
a
t
a
i
n
l
o
w
e
r
-
d
i
m
e
n
s
i
o
n
a
l
e
n
v
i
r
o
n
m
e
n
t
s
,
li
k
e
2
D
o
r
3
D.
FE
an
d
f
ea
tu
r
e
s
elec
tio
n
ar
e
th
e
two
p
r
im
ar
y
s
tep
s
in
d
im
en
s
io
n
ality
r
ed
u
ctio
n
in
m
ac
h
in
e
lear
n
in
g
[
3
1
]
.
T
o
p
r
o
d
u
ce
a
s
m
aller
co
llectio
n
o
f
m
ea
n
in
g
f
u
l
f
ea
tu
r
es,
FE
en
tails
lo
ca
tin
g
an
d
r
em
o
v
in
g
r
elate
d
ch
ar
ac
ter
is
tics
f
r
o
m
h
ig
h
-
d
im
en
s
io
n
al
d
ata.
T
h
is
p
r
o
ce
d
u
r
e
elim
in
ates
n
o
is
e
an
d
r
ed
u
n
d
an
c
y
wh
ile
im
p
r
o
v
in
g
th
e
d
ata'
s
in
ter
p
r
etab
ilit
y
.
I
n
co
n
tr
ast,
f
ea
tu
r
e
s
elec
tio
n
en
tails
elim
in
ati
n
g
s
u
p
er
f
lu
o
u
s
o
r
r
ed
u
n
d
an
t f
ea
tu
r
es wh
ile
s
elec
tin
g
a
s
u
b
s
et
o
f
f
ea
tu
r
es th
at
b
est ca
p
tu
r
e
th
e
d
ata
[
3
2
]
.
4
.
2
.
2
.
F
ea
t
ure
e
x
t
ra
ct
i
o
n
Hu
g
e
v
o
l
u
m
es
o
f
u
n
p
r
o
ce
s
s
ed
d
ata
ar
e
s
ep
a
r
ated
in
t
o
ca
teg
o
r
ies
th
at
ar
e
ea
s
ier
to
m
an
ag
e
u
tili
zin
g
a
d
im
en
s
io
n
ality
r
e
d
u
ctio
n
tech
n
iq
u
e
ter
m
ed
FE.
T
h
e
f
ac
t
t
h
at
th
ese
m
ass
iv
e
d
ata
s
ets
h
a
v
e
n
u
m
er
o
u
s
p
ar
ts
th
at
r
eq
u
ir
e
a
s
ig
n
if
ican
t
am
o
u
n
t
o
f
c
o
m
p
u
tin
g
p
o
wer
m
a
k
es
th
em
co
m
p
ar
ab
le
.
T
h
e
p
h
r
a
s
e
"FE
"
d
escr
ib
es
m
eth
o
d
s
th
at
c
h
o
o
s
e
p
er
tin
e
n
t
v
ar
iab
les
an
d
/o
r
co
m
b
i
n
e
th
em
to
p
r
o
d
u
ce
f
ea
tu
r
es,
w
h
ich
m
in
im
izes
th
e
v
o
lu
m
e
o
f
d
ata
th
at
n
ee
d
s
to
b
e
p
r
o
ce
s
s
ed
wh
ile
ac
cu
r
ately
an
d
f
u
lly
ch
ar
ac
ter
izin
g
th
e
o
r
ig
in
al
d
ataset
[
3
3
]
.
FE
is
a
cr
ea
tiv
e
s
u
b
s
titu
te
f
o
r
f
ea
tu
r
e
s
elec
tio
n
w
h
en
d
ea
lin
g
with
d
im
i
n
is
h
in
g
s
izes
o
f
la
r
g
e
-
s
ca
le
d
ata.
I
n
a
lo
wer
-
d
im
en
s
io
n
al
d
o
m
ain
,
it
is
r
ef
er
r
ed
to
as
"f
ea
tu
r
e
tr
an
s
latio
n
o
r
cr
ea
tio
n
.
"
T
h
e
F
E
m
eth
o
d
p
o
r
tr
ay
s
p
r
o
b
lem
s
in
a
m
o
r
e
u
s
ab
le
an
d
d
is
cr
im
in
atin
g
s
p
ac
e
b
y
ch
a
n
g
in
g
th
e
s
tar
tin
g
v
ar
ia
b
le
in
a
s
p
ac
e
with
f
ewe
r
d
im
en
s
io
n
s
,
wh
ich
in
cr
ea
s
es
th
e
ef
f
icien
cy
o
f
f
u
r
th
er
an
aly
s
is
.
L
in
ea
r
an
d
n
o
n
-
lin
ea
r
ap
p
r
o
ac
h
es
ar
e
th
e
two
m
ain
ca
teg
o
r
ies
o
f
FE
alg
o
r
ith
m
s
.
L
in
ea
r
p
r
o
ce
s
s
es
ar
e
g
en
er
ally
f
aster
,
m
o
r
e
d
ep
e
n
d
a
b
le,
an
d
s
im
p
ler
to
co
m
p
r
eh
e
n
d
th
a
n
n
o
n
-
lin
ea
r
p
r
o
ce
d
u
r
es.
C
o
m
p
lex
d
ata
s
tr
u
ctu
r
es,
o
r
em
b
ed
m
en
ts
,
th
a
t
lin
ea
r
alg
o
r
ith
m
s
ca
n
n
o
t r
ec
o
g
n
ize
a
r
e
d
etec
ted
b
y
n
o
n
-
lin
ea
r
ap
p
r
o
ac
h
es
[
3
4
]
.
FE
is
th
e
p
r
o
ce
s
s
o
f
co
n
v
er
tin
g
a
d
ataset
in
to
a
m
o
r
e
b
asic
f
o
r
m
o
f
c
h
ar
ac
ter
is
tics
s
o
th
at
m
o
r
e
laten
t
id
ea
l
co
m
p
o
n
en
t
f
ea
tu
r
es
ca
n
b
e
in
f
er
r
e
d
f
r
o
m
it.
I
t
o
f
f
e
r
s
an
o
p
en
d
ata
r
e
p
r
esen
tatio
n
o
f
th
e
ass
o
ciate
d
v
ar
iab
le
f
o
r
co
m
b
in
in
g
lin
ea
r
v
ar
iab
les
in
to
f
ea
tu
r
e
s
u
b
s
ets.
Mo
r
eo
v
er
,
FE
is
a
f
lex
ib
le
m
eth
o
d
th
at
m
ay
b
e
u
s
ed
in
v
a
r
io
u
s
co
n
tex
ts
[
3
5
]
.
T
h
is
s
tu
d
y
u
tili
ze
s
I
C
A
to
m
at
ch
co
n
n
ec
ted
p
ar
am
eter
s
in
t
h
e
s
y
s
tem
b
ec
au
s
e
it
n
ec
ess
itates an
o
r
th
o
g
o
n
al
tr
a
n
s
f
o
r
m
atio
n
with
r
ep
r
esen
tatio
n
s
o
f
u
n
i
n
ter
r
u
p
ted
ly
in
d
is
tin
g
u
is
h
ab
le
f
ea
tu
r
es.
4
.
2
.
3
.
I
nd
epen
dent
co
m
po
nent
a
na
ly
s
is
W
h
en
I
C
A
was
f
ir
s
t
in
tr
o
d
u
ce
d
in
t
h
e
1
9
8
0
s
,
it
s
u
g
g
ested
a
r
ein
f
o
r
ce
d
in
s
tan
tan
eo
u
s
m
eth
o
d
.
T
h
er
e
was
n
o
th
e
o
r
etica
l
ex
p
lan
atio
n
in
clu
d
ed
i
n
th
at
b
o
o
k
,
an
d
th
e
p
r
o
p
o
s
ed
m
et
h
o
d
was
ir
r
elev
an
t
in
s
ev
er
al
cir
cu
m
s
tan
ce
s
.
B
u
t
u
n
til
1
9
9
4
,
wh
en
th
e
wo
r
d
"
I
C
A"
f
ir
s
t
ap
p
ea
r
ed
an
d
was
m
ar
k
eted
as
a
n
ew
co
n
ce
p
t,
th
e
I
C
A
alg
o
r
ith
m
was ty
p
ically
u
n
k
n
o
wn
[
3
6
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
1
5
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
6
:
493
-
5
0
5
498
T
h
e
g
o
al
o
f
I
C
A
is
to
ex
tr
ac
t
p
er
tin
en
t
in
f
o
r
m
atio
n
o
r
f
u
n
d
a
m
en
tal
s
ig
n
als
—
a
co
l
lectio
n
o
f
m
ea
s
u
r
ed
m
ix
tu
r
e
s
ig
n
als
—
f
r
o
m
th
e
d
ata.
So
u
r
ce
s
ig
n
als
w
er
e
r
ec
o
v
e
r
ed
u
s
in
g
I
C
A.
W
h
en
I
C
A
ca
n
p
r
eser
v
e
o
r
r
em
o
v
e
a
s
p
ec
if
ic
s
o
u
r
c
e,
it
is
s
o
m
etim
es
s
ee
n
as
a
d
im
en
s
io
n
ality
r
ed
u
ctio
n
p
r
o
ce
d
u
r
e.
C
er
tain
in
f
o
r
m
atio
n
ca
n
b
e
elim
in
ated
o
r
f
ilter
ed
u
s
in
g
th
is
p
r
o
ce
s
s
,
wh
ich
is
also
k
n
o
wn
as a
f
ilter
in
g
o
p
e
r
atio
n
.
I
C
A
ca
n
id
en
tify
in
d
e
p
en
d
e
n
t
co
m
p
o
n
en
ts
an
d
im
p
r
o
v
e
h
ig
h
er
-
o
r
d
er
m
etr
ics
lik
e
k
u
r
to
s
is
.
Sev
er
al
I
C
A
alg
o
r
ith
m
s
ex
is
t,
s
u
ch
as
a
Fas
tI
C
A
p
r
o
jectio
n
p
u
r
s
u
it
an
d
I
n
f
o
m
a
x
[
3
7
]
.
T
h
e
m
ain
g
o
als
o
f
u
s
in
g
th
ese
tech
n
iq
u
es
to
id
en
tify
d
is
tin
ct
co
m
p
o
n
e
n
ts
ar
e
to
ap
p
ly
th
e
m
ax
im
u
m
lik
elih
o
o
d
(
ML
)
esti
m
ate
ap
p
r
o
ac
h
,
m
ax
im
ize
n
o
n
-
Gau
s
s
ian
ity
,
o
r
m
in
im
ize
m
u
tu
al
in
f
o
r
m
atio
n
[
3
8
]
.
T
h
e
s
tep
s
r
eq
u
ir
ed
to
r
u
n
th
e
I
C
A
p
r
o
g
r
am
f
r
o
m
b
eg
in
n
i
n
g
to
c
o
n
clu
s
io
n
ar
e
d
is
p
lay
ed
in
Alg
o
r
ith
m
1
.
Alg
o
r
ith
m
1
.
I
C
A
s
tar
ts
1
: Fir
s
tly
,
K
is
s
et
to
ze
r
o
.
2
:
Ascer
tain
in
g
th
e
d
is
tan
ce
a
m
o
n
g
t
h
e
tr
ain
in
g
in
s
tan
ce
s
an
d
th
e
in
p
u
t e
x
am
p
le.
3
: I
n
th
e
t
h
ir
d
s
tag
e,
s
o
r
t t
h
e
d
i
v
is
io
n
.
4
: I
n
s
tep
f
o
u
r
,
c
h
o
o
s
e
th
e
h
ig
h
est
-
r
an
k
in
g
k
-
n
ea
r
est n
eig
h
b
o
r
s
(
KNN)
.
5
: I
n
s
tep
f
iv
e
,
u
s
e
th
e
s
im
p
le
m
ajo
r
ity
.
6
: U
s
e
ad
d
itio
n
al
n
eig
h
b
o
r
s
f
o
r
id
en
tify
in
g
th
e
lo
a
d
ed
s
am
p
l
e'
s
s
u
b
class
lab
el.
Sto
p
4
.
2
.
4
.
Cla
s
s
if
ica
t
io
n
C
o
n
tem
p
o
r
ar
y
ad
v
an
ce
s
in
d
ata
in
v
esti
g
atio
n
e
m
p
h
asize
t
h
e
s
tate
-
of
-
th
e
-
ar
t
u
tili
za
tio
n
o
f
r
a
n
k
in
g
p
r
o
b
a
b
ilis
tic
m
o
d
els
d
er
iv
e
d
f
r
o
m
b
o
th
L
OS
a
n
d
NL
OS
d
ata
to
class
if
icatio
n
is
s
u
es.
Acc
o
r
d
i
n
g
t
o
R
ay
av
ar
ap
u
an
d
Ma
h
ap
atr
o
[
3
9
]
,
th
is
m
eth
o
d
en
tails
im
p
r
o
v
in
g
th
ese
r
e
p
r
esen
tatio
n
s
th
r
o
u
g
h
th
e
u
s
e
o
f
a
n
in
teg
r
ated
f
r
am
ewo
r
k
,
w
h
ich
m
ak
es
it
p
o
s
s
ib
le
to
d
e
v
elo
p
ef
f
icien
t
class
if
icatio
n
alg
o
r
it
h
m
s
.
Me
an
wh
ile,
as
d
escr
ib
ed
b
y
Mo
r
o
et
a
l.
[
2
7
]
d
ev
elo
p
m
en
ts
in
m
ac
h
in
e
l
ea
r
n
in
g
h
av
e
s
p
a
r
k
ed
t
h
e
cr
e
atio
n
o
f
en
s
em
b
le
d
ec
is
io
n
tr
ee
class
if
icatio
n
tech
n
iq
u
es,
in
clu
d
in
g
b
o
o
s
tin
g
,
b
a
g
g
in
g
,
a
n
d
r
an
d
o
m
f
o
r
ests
.
C
las
s
if
icatio
n
alg
o
r
ith
m
s
ar
e
ess
en
tial
f
o
r
a
n
ticip
atin
g
o
b
s
tacle
s
b
ased
o
n
p
atter
n
s
o
f
b
o
d
y
p
o
s
itio
n
in
t
h
e
f
ield
o
f
"v
ictim
d
etec
tio
n
"
s
tu
d
ies
[
4
0
]
.
A
c
c
o
r
d
i
n
g
t
o
A
y
y
a
d
e
t
a
l
.
[
4
1
]
,
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
i
s
a
s
ci
e
n
t
i
f
i
c
a
p
p
r
o
a
c
h
t
h
a
t
s
e
e
k
s
to
i
m
p
r
o
v
e
c
o
m
p
u
t
e
r
l
e
a
r
n
i
n
g
t
h
r
o
u
g
h
ex
p
e
r
i
e
n
c
e
.
I
n
t
h
is
f
r
a
m
e
w
o
r
k
,
c
l
a
s
s
i
f
i
c
a
ti
o
n
e
n
t
a
i
ls
c
r
ea
t
i
n
g
d
e
c
i
s
i
o
n
c
r
i
t
e
r
ia
a
c
c
o
r
d
i
n
g
t
o
b
o
d
y
o
r
i
e
n
t
a
ti
o
n
an
d
e
n
v
i
r
o
n
m
e
n
t
a
l
f
e
a
t
u
r
es
,
w
h
ic
h
a
r
e
es
s
e
n
t
i
a
l
f
o
r
j
o
b
s
s
u
c
h
as
r
es
c
u
e
o
p
e
r
a
ti
o
n
s
.
T
h
i
s
f
i
el
d
f
r
e
q
u
e
n
t
l
y
e
m
p
l
o
y
s
a
v
a
r
i
e
t
y
o
f
c
l
as
s
i
f
i
e
r
s
,
s
u
c
h
a
s
d
e
ci
s
i
o
n
t
r
e
es
,
n
e
u
r
a
l
n
etw
o
r
k
s
,
a
r
ti
f
i
c
i
al
b
e
e
c
o
l
o
n
i
e
s
,
b
a
t
a
l
g
o
r
i
t
h
m
s
,
p
a
r
t
i
cl
e
s
wa
r
m
o
p
t
i
m
i
z
at
i
o
n
,
s
u
p
p
o
r
t
v
e
c
t
o
r
m
a
c
h
i
n
e
s
(
S
V
M
)
,
a
n
d
K
N
N
.
4
.
2
.
5
.
E
ns
em
ble
T
o
p
r
o
d
u
ce
in
c
r
ed
ib
ly
ac
cu
r
a
te
r
esu
lts
,
en
s
em
b
le
class
if
ier
s
—
lik
e
m
o
d
els
o
f
r
an
d
o
m
s
u
b
s
p
ac
es
—
co
m
b
in
e
d
is
p
ar
ate
s
ec
tio
n
s
o
f
tr
ain
in
g
d
ata
o
r
d
if
f
e
r
en
t
cla
s
s
if
ier
v
ar
iab
les.
T
h
ese
class
if
ier
s
ar
e
f
r
eq
u
e
n
tly
u
s
ed
in
m
ac
h
in
e
lear
n
i
n
g
,
esp
ec
ially
in
s
itu
atio
n
s
w
h
er
e
h
u
m
an
s
m
u
s
t
id
e
n
tify
th
em
s
elv
e
s
b
eh
in
d
o
b
s
tacle
s
in
b
o
th
L
OS
an
d
NL
OS
d
o
m
ain
s
.
E
n
s
em
b
le
clas
s
if
ier
s
m
ak
e
class
if
icatio
n
ju
d
g
m
en
ts
ef
f
icien
tly
b
y
co
m
b
in
i
n
g
th
e
o
u
tp
u
t f
r
o
m
v
ar
io
u
s
class
if
ier
s
[
4
1
]
.
B
y
m
er
g
in
g
t
h
e
r
esu
lts
o
f
s
ev
er
al
class
if
ier
s
,
en
s
em
b
le
tech
n
iq
u
es
—
s
o
m
etim
es
r
ef
e
r
r
ed
to
as
en
s
em
b
le
m
eth
o
d
o
l
o
g
ies
—
im
p
r
o
v
e
class
if
icatio
n
p
er
f
o
r
m
an
ce
.
I
n
en
s
em
b
le
class
if
icatio
n
,
b
o
o
s
tin
g
an
d
b
o
o
ts
tr
ap
a
g
g
r
eg
atin
g
(
b
ag
g
i
n
g
)
ar
e
co
m
m
o
n
tech
n
iq
u
es.
W
h
ile
b
o
o
s
tin
g
m
o
d
if
ies
th
e
weig
h
ts
o
f
tr
ain
in
g
in
s
tan
ce
s
ac
co
r
d
in
g
to
th
ei
r
im
p
ac
t
o
n
class
if
ier
p
er
f
o
r
m
an
ce
,
b
ag
g
in
g
en
tails
r
an
d
o
m
ly
ch
an
g
in
g
th
e
tr
ain
in
g
d
ata
to
p
r
o
d
u
ce
s
u
b
s
titu
te
tr
ain
in
g
cy
cles.
T
h
e
f
in
al
class
if
ier
is
d
er
iv
ed
f
r
o
m
th
e
weig
h
ted
d
ec
is
io
n
s
o
f
s
ep
ar
ate
class
if
ier
s
.
Ad
aBo
o
s
t
was
d
ev
elo
p
e
d
as
a
r
esu
lt
o
f
th
e
tech
n
iq
u
e
s
h
o
w
n
in
[
4
2
]
,
wh
ich
d
em
o
n
s
tr
ates
th
e
u
s
e
o
f
en
s
em
b
le
lear
n
in
g
to
ap
p
ly
b
o
o
s
tin
g
alg
o
r
ith
m
s
i
n
d
atasets
.
T
h
is
iter
ativ
e
m
e
th
o
d
d
em
o
n
s
tr
ates
im
p
r
o
v
em
e
n
ts
in
e
n
s
em
b
le
te
ch
n
iq
u
es
b
y
a
d
ju
s
tin
g
weig
h
ts
to
en
h
an
ce
class
if
ier
p
er
f
o
r
m
an
ce
.
Ass
u
m
e
th
at,
g
iv
en
,
an
y
th
in
g
ca
n
b
e
m
ad
e
s
o
th
at:
1
(
)
=
1
: g
iv
en
an
d
:
(
1
)
+
(
)
=
(
)
×
{
−
=
(
)
≠
(
)
(
2
)
+
(
)
=
(
)
(
−
(
)
)
(
3
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
P
r
ed
ictin
g
tr
a
p
p
ed
victims in
d
eb
r
is
u
s
in
g
s
ig
n
a
l a
n
a
lysi
s
en
s
emb
le
cla
s
s
ifica
tio
n
(
E
n
o
ch
A
d
a
ma
Jiya
)
499
i
s
th
e
n
o
r
m
aliza
tio
n
v
ar
iab
le
i
n
th
is
ca
s
e
,
wh
er
e
:
=
1
2
(
1
−
)
(
4
)
5.
P
E
RF
O
RM
A
NCE
E
VA
L
U
AT
I
O
N
T
o
ass
ess
th
e
m
ac
h
in
e
lear
n
in
g
m
o
d
el'
s
ef
f
icac
y
,
a
f
e
w
v
alid
atio
n
p
r
o
ce
d
u
r
es
ar
e
r
eq
u
ir
e
d
.
C
o
n
f
u
s
io
n
m
atr
ices
ar
e
co
m
m
o
n
ly
u
s
ed
i
n
ca
teg
o
r
izatio
n
m
o
d
els
to
ex
am
in
e
t
h
e
f
o
u
r
s
tan
d
ar
d
s
tr
u
e
p
o
s
itiv
e
(
T
P),
tr
u
e
n
eg
ativ
e
(
T
N)
,
f
alse
p
o
s
itiv
e
(
FP
)
,
a
n
d
f
alse
n
eg
at
iv
e
(
FN)
.
T
h
e
m
o
d
el
d
ataset
s
u
p
p
lied
t
o
ev
alu
ate
th
e
m
o
d
el
id
en
tifie
s
th
e
im
a
g
es
th
at
wer
e
co
r
r
ec
tly
an
d
i
n
co
r
r
ec
tly
id
e
n
tifie
d
.
Per
f
o
r
m
an
ce
m
ea
s
u
r
es
an
d
th
eir
ca
lcu
latio
n
m
eth
o
d
s
ar
e
d
escr
ib
ed
as
f
o
llo
ws
,
wh
er
ea
s
Fig
u
r
e
2
s
h
o
ws
th
e
d
ata
s
am
p
le
co
n
tain
in
g
th
e
f
ea
tu
r
es o
f
th
e
lo
ad
ed
r
aw
d
at
aset.
Fig
u
r
e
2
.
Data
s
am
p
le
s
h
o
win
g
co
n
ten
ts
o
f
th
e
o
r
ig
in
al
u
n
p
r
o
ce
s
s
ed
d
ataset
T
h
e
f
o
u
r
m
etr
ics
th
at
d
eter
m
i
n
e
th
e
co
n
d
itio
n
in
wh
ich
a
m
o
d
el
is
p
r
esen
t
ar
e
u
s
ed
to
ev
alu
ate
its
ac
cu
r
ac
y
.
B
ec
au
s
e
o
f
th
is
,
FP
f
in
d
s
th
e
s
tate
wh
en
it
is
n
'
t
th
er
e.
T
h
e
T
N
p
r
o
d
u
ct
ca
n
n
o
t
b
e
f
o
u
n
d
wh
en
e
v
er
th
e
s
tate
is
n
o
t p
r
esen
t.
T
h
e
F
N
p
r
o
d
u
ct
d
o
es n
o
t d
etec
t th
e
s
tate
if
it e
x
is
ts
.
=
+
+
+
+
(
5
)
T
h
e
p
er
ce
n
tag
e
o
f
ac
cu
r
ate
ly
id
en
tifie
d
in
s
tan
ce
s
with
p
o
s
itiv
e
p
o
s
itiv
es
i
s
k
n
o
wn
as
s
en
s
i
tiv
ity
.
T
h
e
d
ef
in
itio
n
o
f
s
en
s
itiv
ity
is
as f
o
llo
ws
.
(
)
=
+
(
6
)
T
h
e
q
u
a
n
tity
o
f
p
r
o
p
er
ly
id
e
n
t
if
ied
ev
en
ts
with
n
eg
ativ
es is
d
eter
m
in
ed
b
y
s
p
ec
if
icity
.
=
+
(
7
)
=
+
(
8
)
1
−
=
2
×
×
+
(
9
)
6.
AP
P
L
I
CA
T
I
O
N
Hu
m
an
u
n
ea
r
th
in
g
an
d
f
i
n
d
in
g
v
ictim
s
h
id
d
en
b
en
ea
th
th
e
wr
ec
k
ag
e
is
m
ad
e
ea
s
ier
with
NL
OS
d
ata
p
r
o
ce
s
s
in
g
.
T
h
e
n
ee
d
t
o
lo
ca
te
r
elev
an
t
d
ata
is
b
en
ef
icial
to
t
h
e
d
e
v
elo
p
m
e
n
t
o
f
s
ev
er
al
tec
h
n
o
lo
g
ies,
s
u
ch
as
th
e
ar
m
ed
f
o
r
ce
s
,
in
t
r
u
s
io
n
d
etec
tio
n
,
an
d
au
to
m
ated
d
et
ec
tio
n
,
to
n
am
e
a
f
ew.
C
o
n
t
r
ad
icto
r
y
f
ac
ts
an
d
d
esig
n
s
ca
n
b
e
f
o
u
n
d
b
y
u
s
in
g
m
ac
h
in
e
lear
n
in
g
tech
n
o
lo
g
ies.
I
t h
as g
o
o
d
alg
o
r
ith
m
s
,
wh
i
ch
ar
e
to
o
ls
u
tili
ze
d
in
m
an
y
d
if
f
er
e
n
t
in
d
u
s
tr
ies.
B
ec
au
s
e
o
f
its
u
s
er
-
f
r
ien
d
lin
e
s
s
an
d
ad
v
an
tag
eo
u
s
p
r
o
g
r
am
m
in
g
p
latf
o
r
m
f
o
r
en
g
in
ee
r
s
,
ar
ch
itects,
s
ch
o
lar
s
,
an
d
ac
ad
em
ia,
am
o
n
g
o
th
er
s
,
MA
T
L
AB
is
u
s
ed
f
o
r
ex
p
er
im
en
tatio
n
.
MA
T
L
AB
i
s
a
co
m
p
u
ter
lan
g
u
ag
e
an
d
m
u
lti
-
wo
r
ld
v
iew
m
a
th
em
atica
l
co
m
p
u
tatio
n
al
en
v
i
r
o
n
m
en
t
d
e
v
elo
p
e
d
b
y
Ma
th
W
o
r
k
s
.
I
t
en
ab
les
th
e
d
ev
elo
p
m
en
t
o
f
u
s
er
in
ter
f
ac
es
with
in
s
ev
er
al
lan
g
u
ag
es,
in
clu
d
in
g
C
,
C
++
,
C
#
,
J
av
a,
Fo
r
tr
an
,
a
n
d
Py
t
h
o
n
,
as
well
a
s
th
e
im
p
lem
en
tatio
n
o
f
alg
o
r
i
th
m
s
,
f
u
n
ctio
n
a
n
d
d
ata
c
h
ar
ti
n
g
,
a
n
d
co
n
ce
p
tu
al
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
1
5
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
6
:
493
-
5
0
5
500
m
o
d
el
co
n
tr
o
ls
.
T
h
e
m
ain
f
o
c
u
s
o
f
th
is
s
tu
d
y
is
th
e
u
s
e
o
f
co
llap
s
e
s
tr
u
ctu
r
al
tech
n
o
lo
g
y
,
wh
ich
m
ak
es
u
tili
ze
o
f
th
e
MA
T
L
AB
p
r
o
g
r
am
an
d
th
e
NL
OS
d
atab
ase,
to
an
tic
ip
ate
th
e
n
u
m
b
er
o
f
p
eo
p
le
en
tr
ap
p
ed
u
n
d
er
n
ea
th
d
eb
r
is
.
An
iC
o
r
e3
C
PU,
8
G
B
o
f
R
A
M,
a
6
4
-
b
it
o
p
er
atin
g
s
y
s
tem
,
an
d
MA
T
L
AB
2
0
1
5
b
as
its
ex
ec
u
tio
n
to
o
ls
m
ak
e
u
p
th
e
co
m
p
u
ter
c
o
n
f
ig
u
r
atio
n
u
tili
ze
d
f
o
r
t
h
is
s
tu
d
y
.
7.
RE
SU
L
T
S
AND
D
I
SCU
SS
I
O
N
B
o
th
th
e
f
in
d
in
g
s
an
d
a
co
m
m
en
tar
y
o
n
th
e
wo
r
k
ar
e
in
clu
d
e
d
in
th
is
s
ec
t
io
n
.
T
h
e
r
esu
lts
o
f
th
e
test
s
an
d
th
eir
im
p
licatio
n
s
f
o
r
f
u
r
t
h
er
r
esear
ch
will b
e
d
is
cu
s
s
ed
in
th
is
s
ec
t
io
n
.
T
h
e
f
in
d
in
g
s
n
o
t o
n
ly
d
em
o
n
s
tr
ate
th
e
im
p
o
r
tan
ce
o
f
o
u
r
s
tr
ateg
y
b
u
t
also
p
a
v
e
th
e
way
f
o
r
ad
d
itio
n
al
r
esear
ch
in
th
e
ar
ea
.
T
h
e
co
n
f
u
s
io
n
m
atr
ix
,
p
er
f
o
r
m
a
n
ce
m
etr
ic
ta
b
le,
an
d
b
ar
ch
a
r
t r
ep
r
esen
tatio
n
ar
e
th
e
o
u
tco
m
es.
7
.
1
.
Resul
t
s
T
h
e
ex
p
er
im
e
n
t
is
tr
ain
ed
,
ass
es
s
ed
,
an
d
ev
alu
ated
u
s
in
g
an
en
s
em
b
le
lear
n
in
g
te
ch
n
iq
u
e
to
ca
teg
o
r
ize
d
ata
a
n
d
v
er
if
y
its
ef
f
ec
tiv
en
ess
in
r
ed
u
cin
g
e
r
r
o
r
s
in
th
e
d
ataset.
T
h
e
o
u
tco
m
es
ar
e
ass
ess
ed
u
s
in
g
p
er
f
o
r
m
an
ce
m
etr
ics
an
d
c
o
m
p
u
tatio
n
al
o
u
tco
m
es.
T
h
e
m
o
d
els
ar
e
class
if
ied
u
s
in
g
a
n
A
d
aBo
o
s
t
en
s
em
b
le
class
if
ier
,
wh
ich
ac
h
iev
es
a
p
er
f
o
r
m
an
ce
ac
cu
r
ac
y
o
f
8
8
.
0
0
%
f
o
r
s
tatic
m
o
d
els
an
d
8
7
.
2
0
%
f
o
r
d
y
n
am
ic
m
o
d
els.
T
h
e
o
u
tco
m
es
a
n
d
p
r
o
ce
s
s
es
o
f
th
e
r
esu
lts
o
b
tain
e
d
f
o
r
t
h
e
two
d
atasets
ar
e
p
r
e
s
en
ted
in
Fig
u
r
e
3
,
wh
ile
Fig
u
r
e
4
co
n
tai
n
s
th
eir
g
r
ap
h
ical
p
e
r
f
o
r
m
an
ce
co
m
p
ar
is
o
n
.
T
h
e
I
C
A
FE
alg
o
r
ith
m
is
em
p
lo
y
ed
to
r
etr
iev
e
th
e
co
n
ce
aled
ch
a
r
ac
t
er
is
tics
f
r
o
m
th
e
NL
OS
d
ata
in
T
ab
le
1
.
T
h
e
o
u
tco
m
es
ar
e
s
h
o
wn
in
T
ab
le
2
,
wh
ich
cr
ea
tes
an
o
u
tp
u
t
f
o
r
th
e
p
er
f
o
r
m
an
ce
in
d
icato
r
s
i
n
F
ig
u
r
e
3
(
a)
f
o
r
s
tatic
an
d
Fig
u
r
e
3
(
b
)
f
o
r
d
y
n
am
ic
u
s
in
g
th
e
co
n
f
u
s
io
n
m
atr
ix
.
T
h
e
g
ath
er
e
d
attr
ib
u
tes ar
e
ca
te
g
o
r
ized
u
s
in
g
th
e
e
n
s
em
b
le
m
eth
o
d
.
(
a)
(
b
)
Fig
u
r
e
3
.
T
h
e
r
esu
ltan
t c
o
n
f
u
s
io
n
m
atr
ix
f
o
r
th
e
(
a)
s
tatic
an
d
(
b
)
d
y
n
a
m
ic
d
ataset
en
s
em
b
l
e
class
if
icatio
n
s
Fig
u
r
e
4
.
Per
f
o
r
m
an
c
e
in
d
icat
o
r
g
r
a
p
h
0
10
20
30
40
50
60
70
80
90
1
0
0
S
e
n
si
t
i
v
i
t
y
S
p
e
c
i
f
i
c
i
t
y
P
r
e
c
i
si
o
n
N
e
g
a
t
i
v
e
p
r
e
d
i
c
t
i
v
e
v
a
l
u
e
A
c
c
u
r
a
c
y
F
1
-
sc
o
r
e
M
a
t
t
h
e
w
’
s
c
o
r
r
e
l
a
t
i
o
n
c
o
e
f
f
i
c
i
e
n
t
P
e
r
c
e
n
t
a
g
e
(
%
)
C
o
n
f
u
si
o
n
mat
r
i
x
me
t
r
i
c
s
En
se
mb
l
e
C
l
a
ssi
f
i
c
a
t
i
o
n
S
t
a
t
i
c
d
a
t
a
D
y
n
a
mi
c
d
a
t
a
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
P
r
ed
ictin
g
tr
a
p
p
ed
victims in
d
eb
r
is
u
s
in
g
s
ig
n
a
l a
n
a
lysi
s
en
s
emb
le
cla
s
s
ifica
tio
n
(
E
n
o
ch
A
d
a
ma
Jiya
)
501
Fig
u
r
e
4
s
h
o
ws
th
e
p
er
f
o
r
m
an
ce
m
etr
ics
f
o
r
c
o
r
r
ec
tly
a
n
d
er
r
o
n
eo
u
s
ly
class
if
ied
v
alu
es,
wh
ile
Fig
u
r
e
3
s
h
o
w
th
e
d
atasets
th
at
wer
e
s
u
cc
es
s
f
u
lly
class
if
ie
d
.
Nu
m
b
er
s
an
d
g
r
a
p
h
s
th
at
ac
t
as
m
ea
s
u
r
es
f
o
r
p
ar
ticu
lar
p
o
in
ts
o
f
d
ata
ar
e
u
s
ed
to
r
ep
r
esen
t
th
e
v
ar
iab
le
s
’
v
alu
es.
T
h
e
r
elatio
n
s
h
ip
s
a
m
o
n
g
th
e
v
ar
iab
les
th
at
h
av
e
b
ee
n
ca
teg
o
r
ized
ar
e
d
is
p
lay
ed
in
th
is
g
r
a
p
h
.
Fig
u
r
e
3
d
is
p
lay
t
h
e
co
n
f
u
s
io
n
m
at
r
ix
f
o
r
t
h
e
b
o
o
s
ted
en
s
em
b
le
class
if
ier
p
r
ed
ictio
n
s
em
p
lo
y
ed
in
th
e
s
tu
d
y
.
T
h
e
co
n
f
u
s
io
n
m
atr
i
x
tab
le,
wh
ic
h
s
h
o
ws
th
e
tr
u
e,
f
alse,
tr
u
e
n
eg
ativ
e,
a
n
d
f
alse
n
eg
ativ
e
f
ig
u
r
es,
is
th
en
u
tili
ze
d
to
ex
p
lain
h
o
w
well
th
e
p
r
ed
ictio
n
ap
p
r
o
ac
h
p
er
f
o
r
m
s
g
iv
en
s
ets o
f
e
x
p
er
i
m
en
tal
d
ata
with
v
er
if
ie
d
tr
u
e
v
alu
es.
E
n
s
em
b
le
class
if
icatio
n
i
s
u
til
ized
to
p
r
ed
ict
p
er
f
o
r
m
an
ce
.
T
h
e
r
esu
lts
estab
li
s
h
ed
th
e
ef
f
ec
tiv
en
ess
o
f
d
ata
p
r
o
ce
s
s
in
g
wh
e
n
m
ea
s
u
r
ed
d
ata
is
u
s
ed
.
T
h
e
p
e
r
f
o
r
m
an
ce
o
u
tco
m
es
f
o
r
th
e
r
ec
o
m
m
en
d
ed
ap
p
r
o
ac
h
ar
e
s
h
o
wn
a
n
d
r
elate
d
in
T
ab
le
2
.
C
o
n
s
eq
u
e
n
tly
,
th
e
ac
cu
r
ac
y
an
d
o
th
e
r
p
er
f
o
r
m
a
n
ce
m
etr
ics
o
f
en
s
em
b
l
e
ca
teg
o
r
izatio
n
ar
e
s
ig
n
if
ican
tl
y
im
p
r
o
v
ed
.
Fig
u
r
e
4
illu
s
tr
ates
h
o
w
I
C
A
FE
tech
n
iq
u
es
in
c
o
n
ju
n
ctio
n
with
an
SC
n
o
is
e
r
ed
u
ctio
n
m
o
d
el
ca
n
en
h
an
ce
en
s
em
b
le
class
if
icatio
n
r
esu
lts
.
Fig
u
r
e
4
d
ep
icts
th
e
p
er
f
o
r
m
an
ce
cu
r
v
e
f
o
r
co
n
tr
asted
o
u
tp
u
t
p
r
o
d
u
cts.
As
p
r
ev
io
u
s
ly
s
aid
,
Fig
u
r
e
3
an
d
T
ab
le
2
p
r
esen
t
ed
th
e
in
v
esti
g
atio
n
an
d
p
e
r
f
o
r
m
an
ce
ass
ess
m
en
t
o
f
th
e
r
esear
ch
r
esu
lts
.
R
esear
ch
er
s
h
av
e
s
u
g
g
ested
s
ev
e
r
al
im
p
r
o
v
em
en
ts
u
s
in
g
th
e
p
e
r
f
o
r
m
an
ce
m
etr
ics
in
Fig
u
r
e
4
,
a
n
d
th
e
o
u
tco
m
es
h
a
v
e
d
em
o
n
s
tr
ated
th
at
u
s
in
g
I
C
A
FE
tech
n
iq
u
es
an
d
a
ty
p
ical
SC
d
im
en
s
io
n
ality
r
ed
u
ctio
n
s
tr
ateg
y
ca
n
in
c
r
ea
s
e
en
s
em
b
le
class
if
ier
ac
cu
r
ac
y
.
T
ab
le
2
.
Per
f
o
r
m
an
ce
m
etr
ics o
f
th
e
co
n
f
u
s
io
n
m
atr
ix
M
e
a
su
r
e
me
n
t
me
t
r
i
c
s
S
t
a
t
i
c
d
a
t
a
(
%)
D
y
n
a
mi
c
d
a
t
a
(
%)
S
e
n
s
i
t
i
v
i
t
y
7
9
.
0
0
8
3
.
0
1
S
p
e
c
i
f
i
c
i
t
y
9
1
.
0
0
9
0
.
8
5
P
r
e
c
i
s
i
o
n
7
3
.
0
0
8
8
.
7
7
N
e
g
a
t
i
v
e
p
r
e
d
i
c
t
i
v
e
v
a
l
u
e
9
4
.
0
0
8
5
.
9
9
A
c
c
u
r
a
c
y
8
8
.
0
0
8
7
.
2
0
F1
-
s
c
o
r
e
7
6
.
0
0
8
5
.
7
9
M
a
t
t
h
e
w
’
s
c
o
r
r
e
l
a
t
i
o
n
c
o
e
f
f
i
c
i
e
n
t
6
8
.
0
0
7
4
.
3
1
7
.
2
.
Dis
cus
s
io
n
I
n
t
h
is
s
t
u
d
y
,
th
e
class
if
ica
ti
o
n
o
f
NL
O
S
s
i
g
n
al
d
at
a
is
ex
am
i
n
e
d
m
o
r
e
c
o
m
p
r
e
h
e
n
s
i
v
e
ly
.
E
v
e
n
th
o
u
g
h
s
cie
n
tis
ts
h
a
v
e
p
r
o
p
o
s
e
d
a
l
o
t
o
f
e
f
f
o
r
t
,
t
h
e
cu
r
s
e
o
f
d
im
en
s
i
o
n
a
lit
y
a
n
d
d
at
a
n
o
is
e
,
e
s
p
e
cia
ll
y
i
n
NL
OS
s
etti
n
g
s
,
m
ak
e
it
d
if
f
i
cu
lt
t
o
ac
cu
r
a
tel
y
i
d
en
ti
f
y
tr
a
p
p
e
d
p
e
r
s
o
n
s
.
T
h
e
c
u
r
s
e
o
f
d
im
en
s
io
n
alit
y
l
ea
d
s
b
li
n
d
ar
ea
s
as
a
r
esu
lt
o
f
n
o
is
e
an
d
u
n
c
o
n
n
ec
t
e
d
v
al
u
es
i
n
d
et
ec
t
io
n
s
,
a
n
d
th
is
is
s
u
e
h
as
n
o
t
b
ee
n
f
u
ll
y
a
d
d
r
ess
e
d
in
p
r
e
v
i
o
u
s
r
esea
r
ch
.
T
h
is
s
t
u
d
y
p
r
o
p
o
s
es a
n
e
n
h
a
n
ce
d
p
r
e
d
i
cti
o
n
a
n
d
d
et
ec
t
io
n
m
o
d
el
f
o
r
tr
a
p
p
e
d
h
u
m
a
n
v
i
cti
m
s
d
u
r
i
n
g
d
is
as
te
r
s
,
u
s
i
n
g
I
C
A
d
i
m
e
n
s
i
o
n
ali
ty
r
ed
u
cti
o
n
a
n
d
e
n
s
em
b
le
cl
ass
i
f
ic
ati
o
n
d
at
a
m
i
n
in
g
p
r
o
c
e
d
u
r
es
.
T
h
e
in
f
o
r
m
atio
n
in
th
e
f
ig
u
r
es
an
d
tab
les
h
as
v
alid
ated
th
at
d
im
en
s
io
n
ality
r
ed
u
ctio
n
m
o
d
els
with
I
C
A
FE
m
eth
o
d
s
ca
n
g
r
ea
tly
e
n
h
an
ce
t
h
e
r
esu
lts
o
f
en
s
em
b
le
class
if
i
ca
tio
n
.
T
ab
le
2
d
is
p
lay
s
th
e
p
er
f
o
r
m
a
n
ce
m
etr
ics
g
r
ap
h
f
o
r
co
m
p
ar
in
g
th
e
y
ield
ed
r
esu
lts
,
r
ev
ea
lin
g
th
at
p
r
ep
r
o
ce
s
s
in
g
tech
n
iq
u
es
ca
n
en
h
a
n
ce
th
e
class
if
ier
s
'
p
r
ed
ictiv
e
ab
ilit
ies.
Sev
er
al
r
esear
ch
er
s
h
a
v
e
s
tu
d
ied
UW
B
NL
OS
s
ig
n
al
d
ata
class
if
icatio
n
f
o
r
h
u
m
an
d
etec
tio
n
u
s
in
g
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
s
,
aim
in
g
to
o
f
f
er
a
tr
u
s
two
r
th
y
m
eth
o
d
o
f
d
etec
tio
n
a
n
d
p
r
e
d
ictio
n
f
o
r
tr
ap
p
e
d
h
u
m
an
v
ictim
s
d
u
r
in
g
d
is
aster
s
.
Ou
r
s
tu
d
y
b
u
ild
s
o
n
th
is
f
o
u
n
d
atio
n
,
s
h
o
win
g
th
at
t
h
e
p
r
o
p
o
s
ed
m
eth
o
d
with
I
C
A
FE
ac
h
iev
es
h
ig
h
er
ac
c
u
r
ac
y
a
n
d
p
r
ec
is
io
n
c
o
m
p
a
r
e
d
to
p
r
ev
io
u
s
m
eth
o
d
s
.
Static
d
ata
ev
al
u
atio
n
o
u
tp
er
f
o
r
m
ed
d
y
n
am
ic
d
ata
e
v
alu
atio
n
,
with
s
tatic
d
ata
ac
h
iev
in
g
8
8
.
0
0
%
ac
cu
r
ac
y
c
o
m
p
ar
ed
to
8
7
.
2
0
%
f
o
r
d
y
n
am
ic
d
ata.
T
h
is
d
if
f
e
r
en
ce
is
attr
ib
u
ted
to
th
e
r
elativ
e
s
im
p
licity
o
f
NL
OS
d
etec
tio
n
wh
en
th
e
s
en
s
o
r
is
s
tatio
n
ar
y
,
as o
p
p
o
s
ed
to
th
e
in
cr
ea
s
ed
co
m
p
lex
ity
wh
en
th
e
s
en
s
o
r
is
in
m
o
tio
n
.
T
h
is
s
tu
d
y
s
u
g
g
ests
th
at
h
ig
h
e
r
d
im
en
s
io
n
ality
r
ed
u
ctio
n
th
r
o
u
g
h
I
C
A
d
o
es
n
o
t
h
in
d
e
r
p
er
f
o
r
m
an
ce
;
in
s
tead
,
it
en
h
a
n
ce
s
th
e
class
if
icatio
n
r
esu
lts
b
y
m
itig
atin
g
t
h
e
ef
f
ec
ts
o
f
n
o
is
y
an
d
u
n
c
o
r
r
elate
d
d
ata.
T
h
is
is
in
lin
e
with
f
in
d
in
g
s
f
r
o
m
s
im
ilar
s
tu
d
ies
th
at
h
ig
h
lig
h
t
th
e
im
p
o
r
ta
n
ce
o
f
p
r
e
p
r
o
ce
s
s
in
g
tech
n
iq
u
es
in
im
p
r
o
v
in
g
m
o
d
el
p
er
f
o
r
m
an
c
e.
Ho
wev
er
,
u
n
lik
e
s
o
m
e
p
r
e
v
io
u
s
wo
r
k
s
,
o
u
r
ap
p
r
o
ac
h
d
e
m
o
n
s
tr
ates
a
lo
wer
r
elian
ce
o
n
ex
ten
s
iv
e
tr
ain
i
n
g
s
ets,
wh
ich
ca
n
b
e
b
en
ef
icial
in
r
ea
l
-
wo
r
ld
d
is
aster
s
ce
n
ar
io
s
wh
er
e
im
m
ed
iate
d
ata
co
llectio
n
is
cr
itical.
Desp
ite
th
e
p
r
o
m
is
in
g
r
esu
lts
,
th
is
s
tu
d
y
ac
k
n
o
wled
g
es
ce
r
t
ain
lim
itatio
n
s
.
T
h
e
ef
f
ec
tiv
en
ess
o
f
th
e
p
r
o
p
o
s
ed
m
eth
o
d
i
n
h
ig
h
l
y
d
y
n
am
ic
an
d
clu
tter
ed
e
n
v
ir
o
n
m
en
ts
n
ee
d
s
f
u
r
th
er
in
v
esti
g
atio
n
.
Ad
d
itio
n
ally
,
th
e
m
o
d
el’
s
p
e
r
f
o
r
m
an
ce
i
n
d
i
f
f
er
en
t
ty
p
es
o
f
d
is
aster
s
an
d
v
a
r
y
in
g
en
v
ir
o
n
m
en
tal
c
o
n
d
itio
n
s
s
h
o
u
ld
b
e
ex
p
lo
r
e
d
to
en
s
u
r
e
r
o
b
u
s
tn
ess
an
d
g
e
n
er
aliza
b
ilit
y
.
T
h
e
f
in
d
in
g
s
in
d
icate
t
h
at
I
C
A
-
b
ased
d
im
en
s
io
n
ality
r
ed
u
c
tio
n
an
d
en
s
em
b
le
class
if
icatio
n
m
eth
o
d
s
ar
e
m
o
r
e
r
esil
ien
t
a
n
d
e
f
f
ec
ti
v
e
in
NL
OS
s
ig
n
al
d
ata
class
if
icatio
n
[
4
3
]
.
Fu
tu
r
e
r
esear
c
h
co
u
ld
e
x
p
lo
r
e
th
e
in
teg
r
atio
n
o
f
r
ea
l
-
tim
e
ad
ap
ti
v
e
alg
o
r
ith
m
s
th
at
ad
ju
s
t
to
c
h
an
g
in
g
en
v
ir
o
n
m
en
tal
co
n
d
it
io
n
s
,
en
h
an
cin
g
th
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
1
5
,
No
.
1
,
Feb
r
u
ar
y
2
0
2
6
:
493
-
5
0
5
502
m
o
d
el'
s
ap
p
licab
ilit
y
in
d
iv
er
s
e
d
is
aster
s
ce
n
ar
io
s
.
Ad
d
itio
n
ally
,
ex
p
an
d
i
n
g
th
e
d
ataset
with
m
o
r
e
v
ar
ie
d
NL
OS
co
n
d
itio
n
s
a
n
d
i
n
clu
d
i
n
g
d
ata
f
r
o
m
m
u
ltip
le
ty
p
es
o
f
s
en
s
o
r
s
c
o
u
ld
p
r
o
v
id
e
a
m
o
r
e
co
m
p
r
e
h
en
s
iv
e
ev
alu
atio
n
o
f
th
e
m
o
d
el’
s
p
er
f
o
r
m
an
ce
[
4
4
]
,
[
4
5
]
.
T
h
is
s
tu
d
y
d
em
o
n
s
tr
ates
th
at
th
e
p
r
o
p
o
s
ed
I
C
A
-
b
ased
d
im
en
s
io
n
ality
r
ed
u
ctio
n
an
d
en
s
em
b
le
class
if
icatio
n
m
eth
o
d
s
ig
n
if
ic
an
tly
im
p
r
o
v
es
th
e
p
r
e
d
ictio
n
an
d
d
etec
tio
n
o
f
tr
a
p
p
ed
h
u
m
an
v
ictim
s
d
u
r
in
g
d
is
aster
s
.
T
h
e
f
in
d
in
g
s
p
r
o
v
i
d
e
co
n
clu
s
iv
e
ev
id
e
n
ce
th
at
th
is
ap
p
r
o
ac
h
en
h
an
ce
s
ac
cu
r
ac
y
an
d
p
r
ec
is
io
n
,
p
a
r
t
i
c
u
l
a
r
l
y
i
n
s
ta
t
i
c
s
e
n
s
o
r
c
o
n
d
i
t
i
o
n
s
,
a
n
d
r
e
d
u
c
e
s
t
h
e
r
e
l
i
a
n
c
e
o
n
e
x
t
e
n
s
i
v
e
t
r
a
i
n
i
n
g
d
a
t
a
.
T
h
i
s
c
o
n
t
r
i
b
u
t
es
t
o
a
m
o
r
e
e
f
f
e
c
t
i
v
e
a
n
d
e
f
f
i
c
ie
n
t
SA
R
o
p
e
r
a
ti
o
n
,
u
l
t
i
m
at
e
l
y
a
i
d
i
n
g
i
n
s
a
v
i
n
g
l
i
v
es
d
u
r
i
n
g
c
r
i
t
ic
a
l
d
i
s
as
t
e
r
s
i
t
u
a
ti
o
n
s
.
8.
VALI
DAT
I
O
N
C
o
m
p
ar
in
g
th
is
wo
r
k
with
o
t
h
er
r
elate
d
s
tu
d
ies,
as
s
h
o
wn
in
T
ab
le
3
,
th
is
ap
p
r
o
ac
h
d
e
m
o
n
s
tr
ates
s
u
p
er
io
r
ap
p
licatio
n
an
d
ac
cu
r
ac
y
in
h
u
m
an
l
o
ca
lizatio
n
u
n
d
er
NL
OS
s
ce
n
ar
io
s
.
Un
lik
e
o
th
er
m
eth
o
d
s
,
th
is
ap
p
r
o
ac
h
ef
f
ec
tiv
ely
h
a
n
d
les
b
o
th
s
tatic
an
d
d
y
n
am
ic
d
ata,
lead
in
g
to
b
etter
o
v
er
all
p
e
r
f
o
r
m
an
ce
.
T
ab
le
2
en
ca
p
s
u
lates
th
e
p
e
r
f
o
r
m
an
ce
en
h
an
ce
m
e
n
ts
ac
h
iev
ed
b
y
th
i
s
p
r
o
p
o
s
ed
m
o
d
el
ac
r
o
s
s
all
d
atasets
.
T
h
e
m
o
d
el
o
u
tp
er
f
o
r
m
s
o
th
er
a
p
p
r
o
a
ch
es
s
p
ec
if
ically
,
th
e
ac
cu
r
ac
y
f
o
r
s
tatic
d
ata
im
p
r
o
v
ed
to
8
8
.
0
0
%,
s
u
r
p
ass
in
g
th
e
b
est
-
p
er
f
o
r
m
in
g
b
aselin
es.
T
h
is
s
ig
n
if
ican
t
im
p
r
o
v
em
en
t
is
attr
ib
u
ted
to
t
h
e
ef
f
ec
tiv
e
in
teg
r
atio
n
o
f
d
i
m
en
s
io
n
ality
r
ed
u
ctio
n
an
d
h
y
b
r
i
d
izatio
n
tech
n
iq
u
es,
wh
ich
r
e
f
in
e
l
o
ca
lizatio
n
r
e
lev
an
ce
an
d
ac
cu
r
ac
y
.
B
y
le
v
er
ag
in
g
en
s
em
b
le
lear
n
in
g
a
n
d
n
o
is
e
r
ed
u
ctio
n
,
th
is
m
o
d
el
d
em
o
n
s
tr
ates
a
r
o
b
u
s
t
ab
ilit
y
to
h
an
d
le
co
m
p
lex
NL
OS
co
n
d
itio
n
s
.
T
h
e
en
s
em
b
le
lear
n
in
g
ap
p
r
o
a
ch
,
co
m
b
in
ed
with
n
o
is
e
r
ed
u
ctio
n
tech
n
i
q
u
es,
en
s
u
r
es
th
at
th
is
m
o
d
el
n
o
t
o
n
ly
ac
h
iev
es h
ig
h
er
ac
cu
r
ac
y
b
u
t a
ls
o
m
ain
tain
s
co
n
s
is
ten
cy
ac
r
o
s
s
v
ar
io
u
s
d
atasets
,
b
o
th
s
tatic
an
d
d
y
n
am
ic.
T
ab
le
3
.
C
o
m
p
a
r
ativ
e
ap
p
r
o
ac
h
es
M
e
t
h
o
d
s
A
c
c
u
r
a
c
y
(
%)
KNN
[
4
6
]
8
5
.
0
0
S
V
M
+
a
u
t
o
e
n
c
o
d
e
r
[
4
7
]
8
6
.
9
8
C
N
N
+
st
a
c
k
e
d
-
LST
M
[
4
8
]
8
2
.
1
4
P
r
o
p
o
se
d
m
e
t
h
o
d
8
8
.
0
0
9.
CO
NCLU
SI
O
N
T
h
is
s
tu
d
y
an
aly
ze
d
an
d
im
p
r
o
v
ed
th
e
class
if
icatio
n
o
f
th
e
NL
OS
h
u
m
an
d
etec
tio
n
d
at
aset.
Data
an
aly
s
is
in
th
is
r
esear
ch
p
r
o
v
i
d
es
s
u
b
s
tan
tial
an
d
wo
r
th
wh
il
e
ad
v
an
tag
es
to
th
e
s
cien
tis
ts
an
d
en
g
i
n
ee
r
s
,
an
d
th
e
d
ev
el
o
p
m
en
t
o
f
tech
n
o
lo
g
y
h
as
g
r
ea
tly
aid
ed
in
th
e
e
v
o
lu
tio
n
o
f
th
e
is
s
u
es
o
f
SAR
o
p
er
atio
n
s
.
Hu
m
an
d
etec
tio
n
d
ataset
p
r
ed
ictio
n
a
n
d
r
ed
u
cti
o
n
o
f
d
im
en
s
io
n
alit
y
an
d
class
if
icatio
n
tech
n
iq
u
e
s
ar
e
ex
am
p
les
o
f
co
m
p
ar
ab
le
tech
n
o
lo
g
ies.
T
h
e
s
e
tech
n
o
lo
g
ical
a
d
v
an
ce
m
en
t
s
h
av
e
s
ig
n
if
ican
tly
co
n
tr
i
b
u
t
ed
to
th
e
ev
o
l
u
tio
n
o
f
th
e
p
r
o
b
lem
s
o
f
r
elate
d
ap
p
licatio
n
s
,
in
clu
d
i
n
g
t
h
e
r
ed
u
ctio
n
o
f
d
im
en
s
io
n
alit
y
an
d
class
if
icatio
n
ap
p
r
o
ac
h
es.
I
t
is
a
s
er
io
u
s
is
s
u
e
b
ec
au
s
e
o
f
th
e
d
ata'
s
cu
r
s
e
o
f
d
im
en
s
io
n
ality
b
o
u
n
d
.
A
n
u
m
b
er
o
f
ap
p
r
o
ac
h
es
h
av
e
b
ee
n
p
u
t
o
u
t
to
a
d
v
an
ce
th
e
tech
n
o
l
o
g
y
an
d
an
ticip
ate
an
d
id
e
n
tify
tr
a
p
p
ed
v
ictim
s
th
at
ar
e
tak
e
n
f
r
o
m
s
am
p
les;
th
ese
d
if
f
icu
lties
h
av
e
b
ee
n
r
eso
lv
e
d
b
y
r
ed
u
ci
n
g
d
i
m
en
s
io
n
ality
.
Ho
wev
er
,
ad
d
iti
o
n
al
r
esear
ch
m
u
s
t
b
e
d
o
n
e.
R
ec
en
tly
,
s
ev
er
al
tec
h
n
iq
u
es h
av
e
also
b
ee
n
em
p
lo
y
ed
to
ca
teg
o
r
ize
an
d
p
r
ed
ict
UW
B
NL
OS h
u
m
an
d
etec
tio
n
s
ig
n
al
d
ata
e
x
p
r
ess
io
n
r
esu
lts
.
Nev
er
th
eless
,
th
e
s
tatic
I
C
A
en
s
em
b
le
(
8
8
.
0
0
%
)
o
u
tp
e
r
f
o
r
m
ed
th
e
d
y
n
am
ic
I
C
A
en
s
em
b
le
(
8
7
.
2
0
%)
b
ased
ap
p
r
o
ac
h
b
y
u
tili
zin
g
SC
with
I
C
A
to
ca
r
r
y
o
u
t
a
d
im
en
s
io
n
ality
r
ed
u
ctio
n
a
p
p
r
o
ac
h
with
I
C
A
f
o
r
s
tatic
an
d
SC
with
I
C
A
f
o
r
d
y
n
am
ic
d
atasets
.
T
h
e
alg
o
r
ith
m
s
wer
e
ap
p
lied
d
is
cr
etely
,
an
d
th
eir
p
er
f
o
r
m
a
n
ce
o
n
th
e
en
s
em
b
le
class
if
icatio
n
m
o
d
el
was
ev
alu
ated
.
T
h
is
wo
r
k
aim
s
to
p
r
o
v
id
e
a
way
to
r
e
d
u
ce
th
e
n
u
m
b
er
o
f
v
ar
iab
les wh
ile
m
ain
tain
in
g
in
f
o
r
m
ativ
e
o
n
es f
o
r
i
m
p
r
o
v
e
d
p
r
ed
ictio
n
,
wh
ich
en
g
in
ee
r
s
ca
n
u
s
e
to
m
ak
e
d
ec
is
io
n
s
.
T
h
is
wo
r
k
u
s
ed
NL
OS
h
u
m
an
d
etec
tio
n
d
ata
to
s
u
g
g
est
a
p
h
ased
in
cr
ea
s
e
in
d
im
en
s
io
n
ality
an
d
p
r
ed
ictio
n
tech
n
iq
u
e.
A
s
u
p
er
io
r
p
er
f
o
r
m
an
ce
m
ea
s
u
r
e
was
o
b
tain
ed
b
y
r
etr
iev
in
g
r
elev
an
t
f
ea
tu
r
es.
I
n
o
r
d
er
to
d
eter
m
in
e
th
e
ap
p
r
o
p
r
iate
ca
teg
o
r
izatio
n
o
f
th
e
NL
OS'
s
h
u
m
an
-
d
etec
tin
g
ex
p
r
ess
io
n
d
ata,
f
u
tu
r
e
r
esear
c
h
s
u
g
g
ests
ap
p
ly
in
g
h
y
b
r
id
d
im
en
s
io
n
ality
r
ed
u
cti
o
n
tech
n
iq
u
es to
o
t
h
er
class
if
ier
s
,
s
u
ch
as d
ee
p
lear
n
in
g
.
ACK
NO
WL
E
DG
M
E
N
T
S
T
h
e
au
th
o
r
s
g
r
ate
f
u
lly
ac
k
n
o
wled
g
e
th
e
s
u
p
p
o
r
t
r
ec
eiv
ed
f
r
o
m
th
e
Fre
n
ch
So
u
th
A
f
r
ican
I
n
s
titu
te
o
f
T
ec
h
n
o
lo
g
y
,
C
ap
e
Pen
in
s
u
la
Un
iv
er
s
ity
o
f
T
ec
h
n
o
lo
g
y
,
B
ellv
ille,
So
u
th
Af
r
ica.
T
h
e
au
th
o
r
s
ex
p
r
ess
g
r
atitu
d
e
to
L
an
d
m
a
r
k
Un
iv
e
r
s
ity
f
o
r
p
r
o
v
id
in
g
all
th
e
m
ater
ials
r
eq
u
i
r
ed
f
o
r
th
is
r
esear
ch
.
Evaluation Warning : The document was created with Spire.PDF for Python.