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
.
3
,
J
u
n
e
20
25
,
p
p
.
3382
~
3
3
9
5
I
SS
N:
2088
-
8
7
0
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijece.
v
15
i
3
.
pp
3
3
8
2
-
3
3
9
5
3382
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ec
e.
ia
esco
r
e.
co
m
Enha
ncing
Alzhei
mer’s
disea
se
dia
g
no
sis
t
hro
ug
h metaheuris
ti
c
feature
select
io
n
a
nd ad
v
a
nced clas
sifica
tion te
chni
ques
Ara
r
Al
-
T
a
wil
1
,
Wo
ro
o
d Al
-
M
uh
t
a
s
e
b
2
,
L
a
ia
li Alm
a
za
y
deh
3
,
H
a
na
a
F
a
t
hi
4
1
F
a
c
u
l
t
y
o
f
I
n
f
o
r
m
a
t
i
o
n
T
e
c
h
n
o
l
o
g
y
,
A
p
p
l
i
e
d
S
c
i
e
n
c
e
P
r
i
v
a
t
e
U
n
i
v
e
r
si
t
y
,
A
mm
a
n
,
J
o
r
d
a
n
2
D
e
p
a
r
t
me
n
t
o
f
I
n
f
o
r
mat
i
o
n
Te
c
h
n
o
l
o
g
i
e
s a
n
d
S
y
st
e
ms,
U
n
i
v
e
r
s
i
d
a
d
d
e
C
a
s
t
i
l
l
a
-
L
a
M
a
n
c
h
a
,
C
i
u
d
a
d
R
e
a
l
,
S
p
a
i
n
3
C
o
l
l
e
g
e
o
f
E
n
g
i
n
e
e
r
i
n
g
,
A
b
u
D
h
a
b
i
U
n
i
v
e
r
si
t
y
,
A
b
u
D
h
a
b
i
,
U
n
i
t
e
d
A
r
a
b
Emi
r
a
t
e
s
4
C
o
l
l
e
g
e
o
f
I
n
f
o
r
ma
t
i
o
n
Te
c
h
n
o
l
o
g
y
,
A
mm
a
n
A
r
a
b
U
n
i
v
e
r
si
t
y
,
A
mm
a
n
,
Jo
r
d
a
n
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
u
l 2
8
,
2
0
2
4
R
ev
is
ed
J
an
2
4
,
2
0
2
5
Acc
ep
ted
Ma
r
3
,
2
0
2
5
A
d
iv
e
rse
a
rra
y
o
f
d
iag
n
o
stic
a
n
d
d
e
tec
ti
o
n
m
e
th
o
d
s h
a
s
b
e
e
n
d
e
v
e
l
o
p
e
d
a
s
a
re
su
lt
o
f
th
e
a
d
v
e
n
t
o
f
Alz
h
e
ime
r’s
d
ise
a
se
(AD
)
a
s
a
sig
n
ifi
c
a
n
t
g
l
o
b
a
l
h
e
a
lt
h
issu
e
.
Th
is
st
u
d
y
e
m
p
lo
y
s
b
io
-
i
n
sp
ire
d
a
lg
o
ri
th
m
s,
su
c
h
a
s
t
h
e
p
a
rro
t
o
p
ti
m
iza
ti
o
n
a
l
g
o
ri
th
m
(
P
OA
),
g
r
e
y
wo
lf
o
p
ti
m
ize
r
(G
WO),
a
n
d
d
i
ffe
re
n
ti
a
l
e
v
o
lu
ti
o
n
(DE),
t
o
id
e
n
ti
fy
th
e
m
o
st effe
c
ti
v
e
fe
a
tu
re
se
lec
ti
o
n
tec
h
n
i
q
u
e
s fo
r
AD
d
iag
n
o
sis.
T
h
e
p
re
d
ictiv
e
a
c
c
u
ra
c
y
o
f
th
e
se
a
lg
o
rit
h
m
s
wa
s
imp
ro
v
e
d
b
y
th
e
s
imp
le
k
e
y
w
o
rd
s:
Alz
h
e
i
m
e
r’s
d
ise
a
se
o
p
ti
m
iza
ti
o
n
c
l
a
s
sifica
ti
o
n
m
a
c
h
in
e
lea
rn
in
g
m
e
tah
e
u
risti
c
m
e
n
tatio
n
o
f
th
e
Alz
h
e
ime
r’s
d
ise
a
se
Da
tas
e
t.
Th
is wa
s a
c
h
iev
e
d
b
y
i
n
t
e
g
ra
ti
n
g
a
p
e
rs
o
n
a
li
z
e
d
fit
n
e
ss
fu
n
c
ti
o
n
a
n
d
o
p
ti
m
izi
n
g
p
a
ra
m
e
ter
se
tt
in
g
s
wi
th
d
e
c
isio
n
tree
c
las
sifiers
.
To
e
v
a
lu
a
te
th
e
a
lg
o
rit
h
m
s’
e
ffe
c
ti
v
e
n
e
ss
in
m
a
c
h
in
e
lea
rn
in
g
m
o
d
e
ls
with
p
o
p
u
l
a
ti
o
n
siz
e
s
o
f
3
0
a
n
d
6
0
,
p
re
c
isi
o
n
,
re
c
a
ll
,
a
c
c
u
ra
c
y
,
a
n
d
F
1
-
sc
o
re
we
re
e
v
a
lu
a
ted
a
t
5
,
1
5
,
a
n
d
3
0
it
e
ra
ti
o
n
s.
Th
e
g
r
a
d
ien
t
b
o
o
stin
g
a
n
d
XG
Bo
o
st
c
las
sifiers
c
o
n
siste
n
tl
y
o
b
tain
e
d
t
h
e
h
ig
h
e
st
re
su
lt
s,
wh
il
e
DE,
G
WO
,
a
n
d
p
a
rr
o
t
o
p
ti
m
iza
ti
o
n
(P
O)
a
c
h
ie
v
e
d
m
a
x
i
m
a
l
a
c
c
u
ra
c
y
ra
tes
o
f
0
.
9
4
,
0
.
9
3
,
a
n
d
0
.
9
4
,
re
sp
e
c
ti
v
e
ly
.
T
h
e
se
fin
d
i
n
g
s
u
n
d
e
rsc
o
re
t
h
e
e
ffica
c
y
o
f
i
n
teg
ra
ti
n
g
m
e
tah
e
u
risti
c
a
lg
o
r
it
h
m
s
wit
h
r
o
b
u
st
c
las
sifiers
to
e
n
h
a
n
c
e
t
h
e
p
re
d
ict
iv
e
a
c
c
u
ra
c
y
o
f
AD
d
iag
n
o
sis.
F
u
rth
e
rm
o
re
,
t
h
e
y
il
l
u
stra
te
th
a
t
a
rti
ficia
l
in
telli
g
e
n
c
e
(AI)
a
lg
o
rit
h
m
s
t
h
a
t
a
re
o
p
e
ra
ted
b
y
b
i
o
lo
g
ica
l
p
ro
c
e
ss
e
s
c
a
n
a
c
c
u
ra
tely
fo
re
c
a
st
AD
,
with
th
e
su
c
c
e
ss
ra
tes
a
n
d
sta
b
il
it
y
o
f
th
e
p
ro
p
o
se
d
m
e
th
o
d
s se
rv
i
n
g
a
s m
e
tri
c
s fo
r
e
v
a
lu
a
ti
n
g
th
e
ir
e
ffica
c
y
.
K
ey
w
o
r
d
s
:
Alzh
eim
er
’
s
d
is
ea
s
e
C
las
s
if
icatio
n
Ma
ch
in
e
lear
n
in
g
Me
tah
eu
r
is
tic
Op
tim
izatio
n
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
:
Ar
ar
Al
-
T
awil
Facu
lty
o
f
I
n
f
o
r
m
atio
n
T
ec
h
n
o
lo
g
y
,
A
p
p
lied
Scien
ce
Priv
at
e
Un
iv
er
s
ity
Am
m
an
Am
m
an
,
J
o
r
d
a
n
E
m
ail: a
r
_
altawil@
asu
.
ed
u
.
jo
1.
I
NT
RO
D
UCT
I
O
N
T
o
ac
ce
ler
ate
th
e
id
en
tific
at
io
n
o
f
b
io
m
ar
k
er
s
d
u
r
in
g
t
h
e
in
itial
p
h
ases
an
d
p
r
o
g
r
ess
io
n
o
f
Alzh
eim
er
’
s
d
is
ea
s
e,
th
e
Alzh
eim
er
’
s
d
is
ea
s
e
n
eu
r
o
im
a
g
in
g
in
itiativ
e
(
ADNI
)
was
estab
lis
h
ed
b
y
th
e
Natio
n
al
I
n
s
titu
te
o
n
Ag
in
g
.
T
h
e
s
u
cc
ess
f
u
l
au
to
m
atio
n
an
d
o
p
tim
izatio
n
o
f
d
esig
n
p
r
o
ce
s
s
es
th
at
u
tili
ze
d
d
ata
f
r
o
m
ce
r
e
b
r
o
s
p
in
al
f
lu
i
d
(
C
SF
)
,
p
o
s
itro
n
em
is
s
io
n
to
m
o
g
r
ap
h
y
(
PET
)
,
a
n
d
m
ag
n
etic
r
eso
n
an
ce
im
ag
in
g
(
f
MRI)
wer
e
th
e
p
r
im
ar
y
f
o
cu
s
o
f
r
esear
ch
er
s
[
1
]
–
[
3
]
.
I
n
itially
,
th
is
p
r
o
ject
d
ev
elo
p
s
m
ac
h
in
e
-
lear
n
in
g
m
o
d
els
th
at
in
co
r
p
o
r
ate
th
e
m
o
s
t
cr
itical
attr
ib
u
tes
o
f
th
e
s
ty
le
s
elec
tio
n
to
id
en
tify
an
o
p
tim
al
s
o
lu
tio
n
to
th
e
cr
itical
b
in
ar
y
class
if
icatio
n
o
f
VC
I
to
co
n
tr
o
l
s
u
b
jects
[
4
]
,
[
5
]
.
T
h
e
r
esear
ch
co
n
clu
d
es
b
y
id
en
tify
in
g
th
e
co
m
p
o
n
en
ts
ass
ess
ed
f
o
r
s
tatis
tical
in
ter
-
g
r
o
u
p
v
ar
iatio
n
s
in
all
m
o
d
els.
As
a
r
e
s
u
lt,
we
an
aly
ze
d
class
if
icatio
n
s
u
cc
es
s
,
wh
ich
i
s
th
e
d
eg
r
ee
to
wh
ich
th
ese
c
o
m
p
o
n
en
ts
ac
cu
r
ately
r
ep
r
esen
t
th
e
s
ev
er
ity
o
f
a
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
E
n
h
a
n
cin
g
A
lz
h
eime
r
’
s
d
is
ea
s
e
d
ia
g
n
o
s
is
th
r
o
u
g
h
meta
h
eu
r
is
tic
fea
tu
r
e
…
(
A
r
a
r
A
l
-
Ta
w
i
l
)
3383
m
en
tal
illn
ess
in
a
life
-
th
r
ea
ten
in
g
m
an
n
er
an
d
th
e
d
eg
r
ee
t
o
wh
ich
th
ey
f
ac
ilit
ate
th
e
u
n
d
er
s
tan
d
in
g
o
f
th
e
p
r
o
g
r
ess
io
n
o
f
Alzh
eim
er
’
s
d
i
s
ea
s
e.
Ou
r
p
r
o
p
o
s
ed
m
eth
o
d
was
s
h
o
wn
to
b
e
r
o
b
u
s
t
ac
r
o
s
s
a
r
an
g
e
o
f
s
tu
d
y
p
ar
am
eter
s
b
y
u
tili
zin
g
a
v
ar
i
ety
o
f
m
ac
h
in
e
lear
n
i
n
g
m
o
d
e
ls
th
at
in
clu
d
ed
eig
en
v
ec
to
r
s
o
f
f
ea
tu
r
e
s
elec
tio
n
p
r
o
ce
d
u
r
es,
as
well
as
b
y
s
ele
ctin
g
o
v
er
1
5
0
ADNI
VC
I
an
d
co
n
tr
o
llin
g
m
o
d
er
ate
to
m
e
d
iu
m
VC
I
,
g
en
d
er
-
m
atch
ed
p
ee
r
s
f
r
o
m
t
h
e
ADNI
[
6
]
,
[
7
]
.
T
h
e
p
o
ten
tial
o
f
m
ac
h
in
e
lear
n
in
g
in
r
ev
o
lu
tio
n
izin
g
d
is
ea
s
e
id
en
tific
atio
n
is
a
b
ea
c
o
n
o
f
h
o
p
e
f
o
r
th
e
f
u
tu
r
e
o
f
m
e
d
ical
r
esear
ch
.
T
h
is
ap
p
licatio
n
o
f
au
to
m
ated
alg
o
r
ith
m
s
,
b
ased
o
n
h
is
to
r
ical
d
ata,
is
p
ar
ticu
lar
ly
p
r
o
m
is
in
g
in
th
e
ac
cu
r
ate
id
en
tific
atio
n
o
f
m
alad
ies.
T
h
e
s
tr
id
es
i
n
m
ac
h
in
e
lear
n
in
g
c
o
u
ld
s
ig
n
if
ican
tly
b
en
ef
it
Alzh
eim
e
r
’
s
Dis
ea
s
e,
a
co
n
d
itio
n
th
at
p
o
s
es
a
s
ig
n
if
ican
t
ch
allen
g
e
f
o
r
clin
ician
s
d
u
e
to
its
v
ar
ied
ch
ar
ac
ter
is
tics
an
d
t
h
e
p
r
o
g
r
ess
io
n
o
f
th
ese
ch
ar
ac
ter
is
tics
in
its
in
it
ial
s
tag
e.
Featu
r
e
s
elec
tio
n
is
a
cr
itical
co
m
p
o
n
en
t
o
f
d
ata
a
n
aly
s
is
a
n
d
m
ac
h
in
e
lear
n
in
g
,
as
it
en
h
an
ce
s
th
e
in
ter
p
r
etab
ilit
y
an
d
ef
f
icac
y
o
f
p
r
e
d
ictiv
e
m
o
d
els.
B
y
s
elec
tin
g
th
e
m
o
s
t
r
ele
v
an
t
f
ea
t
u
r
es
f
r
o
m
a
d
ataset,
f
ea
tu
r
e
s
elec
tio
n
tech
n
iq
u
es
ar
e
in
ten
d
ed
to
im
p
r
o
v
e
th
e
ac
cu
r
ac
y
o
f
m
o
d
els,
p
r
ev
e
n
t
o
v
er
f
itti
n
g
,
an
d
r
e
d
u
ce
co
m
p
u
tatio
n
al
co
m
p
lex
ity
[
8
]
,
[
9
]
.
T
h
ese
tech
n
iq
u
es
ca
n
b
e
b
r
o
a
d
ly
class
if
ied
in
to
th
r
e
e
ca
teg
o
r
ies:
f
ilter
m
eth
o
d
s
,
wr
ap
p
er
m
eth
o
d
s
,
an
d
em
b
e
d
d
ed
m
eth
o
d
s
[
1
0
]
.
Fil
ter
m
eth
o
d
s
o
p
e
r
ate
in
d
ep
en
d
en
tly
o
f
th
e
lear
n
in
g
al
g
o
r
ith
m
,
wh
ile
s
tatis
tical
m
ea
s
u
r
es
ar
e
im
p
lem
e
n
ted
to
e
v
alu
ate
th
e
r
elev
a
n
c
e
o
f
f
ea
tu
r
es.
T
h
e
m
o
s
t
f
r
eq
u
en
tly
em
p
lo
y
ed
f
ilter
in
g
tech
n
iq
u
es
ar
e
ch
i
-
s
q
u
ar
e
test
s
,
co
r
r
elatio
n
co
ef
f
i
cien
ts
,
an
d
m
u
tu
al
in
f
o
r
m
atio
n
.
T
h
ese
m
eth
o
d
s
a
r
e
co
m
p
u
tatio
n
ally
ef
f
icien
t a
n
d
p
r
o
v
id
e
a
r
ap
id
m
eth
o
d
f
o
r
r
em
o
v
in
g
ir
r
elev
an
t
o
r
r
ed
u
n
d
an
t
f
ea
tu
r
es
b
ef
o
r
e
m
o
d
el
tr
ain
in
g
[
1
1
]
,
[
1
2
]
.
C
o
n
v
er
s
ely
,
wr
ap
p
er
m
eth
o
d
s
ev
alu
ate
f
ea
tu
r
e
s
u
b
s
ets
b
y
th
e
ef
f
icac
y
o
f
a
s
p
ec
if
i
c
lear
n
in
g
alg
o
r
ith
m
.
R
ec
u
r
s
iv
e
f
ea
tu
r
e
elim
in
atio
n
(
R
FE)
an
d
f
o
r
war
d
o
r
b
ac
k
war
d
s
elec
tio
n
m
eth
o
d
s
ar
e
in
clu
d
ed
i
n
th
is
ca
teg
o
r
y
.
E
v
en
th
o
u
g
h
wr
a
p
p
er
m
eth
o
d
s
ar
e
o
f
ten
m
o
r
e
p
r
ec
is
e
th
an
f
ilter
m
eth
o
d
s
,
th
ey
ar
e
co
m
p
u
tatio
n
ally
in
ten
s
iv
e
b
ec
au
s
e
th
ey
r
eq
u
ir
e
m
u
ltip
le
m
o
d
el
tr
ain
in
g
an
d
ev
alu
atio
n
p
h
ases
[
1
3
]
,
[
1
4
]
.
I
n
ad
d
itio
n
to
co
n
v
en
tio
n
al
f
ea
tu
r
e
s
elec
tio
n
m
eth
o
d
s
,
m
etah
eu
r
is
tic
alg
o
r
ith
m
s
h
av
e
g
ain
ed
p
o
p
u
lar
ity
d
u
e
to
th
eir
a
b
ilit
y
to
ef
f
icien
tly
in
v
esti
g
ate
th
e
f
ea
tu
r
e
s
p
ac
e
an
d
id
en
tif
y
o
p
ti
m
al
f
ea
tu
r
e
s
u
b
s
ets
[
1
5
]
.
Me
tah
eu
r
is
tic
alg
o
r
ith
m
s
ar
e
h
ig
h
-
lev
el
p
r
o
b
lem
-
i
n
d
e
p
en
d
en
t
tec
h
n
iq
u
es
th
at
em
p
l
o
y
m
ec
h
a
n
is
m
s
to
in
v
esti
g
ate
th
e
g
lo
b
al
s
ea
r
ch
s
p
ac
e
a
n
d
cir
cu
m
v
en
t
l
o
ca
l
o
p
tim
a.
Am
o
n
g
th
e
m
o
s
t
f
r
eq
u
en
tly
em
p
lo
y
ed
m
etah
eu
r
is
tic
alg
o
r
ith
m
s
in
f
ea
tu
r
e
s
elec
tio
n
ar
e
g
en
etic
alg
o
r
ith
m
s
(
GA)
,
p
ar
ticle
s
war
m
o
p
tim
izatio
n
(
PS
O)
[
1
6
]
,
[
1
7
]
,
an
t
co
lo
n
y
o
p
tim
iz
atio
n
(
AC
O)
,
an
d
d
if
f
er
en
tial e
v
o
lu
tio
n
(
DE
)
[
1
8
]
,
[
1
9
]
.
A
p
p
ly
in
g
m
etah
e
u
r
is
tic
alg
o
r
ith
m
s
in
co
n
ju
n
ctio
n
with
f
ea
tu
r
e
s
elec
tio
n
tech
n
iq
u
es
p
r
o
v
id
es
a
p
o
wer
f
u
l
ap
p
r
o
ac
h
to
m
an
ag
in
g
h
ig
h
-
d
im
en
s
io
n
al
d
ata
an
d
in
tr
icate
f
ea
tu
r
e
s
p
ac
es
[
2
0
]
.
T
h
e
f
lex
i
b
ilit
y
an
d
r
o
b
u
s
tn
ess
o
f
th
ese
alg
o
r
ith
m
s
en
ab
le
a
wid
e
r
an
g
e
o
f
ap
p
licatio
n
s
in
v
ar
io
u
s
d
o
m
ain
s
,
in
clu
d
in
g
m
ed
ical
d
iag
n
o
s
is
,
im
ag
e
p
r
o
ce
s
s
in
g
,
an
d
b
io
in
f
o
r
m
atics
[
2
1
]
,
[
2
2
]
.
B
y
lev
er
ag
i
n
g
t
h
e
b
en
ef
its
o
f
b
o
th
tr
a
d
itio
n
al
an
d
m
etah
eu
r
i
s
tic
m
eth
o
d
o
lo
g
ies,
r
esear
ch
er
s
an
d
p
r
ac
titi
o
n
er
s
c
an
d
ev
elo
p
p
r
e
d
ictiv
e
m
o
d
els th
at
ar
e
m
o
r
e
p
r
ec
is
e
an
d
ef
f
ici
en
t.
T
h
e
p
r
im
ar
y
co
n
tr
ib
u
tio
n
s
o
f
th
is
r
esear
ch
ar
e
as
f
o
llo
ws:
i)
Op
tim
al
f
ea
tu
r
e
s
elec
tio
n
was
ac
h
iev
ed
b
y
ap
p
ly
in
g
f
o
u
r
b
io
-
in
s
p
ir
e
d
alg
o
r
ith
m
s
d
if
f
er
e
n
tial
ev
o
lu
tio
n
(
DE
)
,
g
r
e
y
wo
lf
o
p
tim
izer
(
GW
O)
,
an
d
p
ar
r
o
t
o
p
tim
izatio
n
al
g
o
r
ith
m
(
POA)
to
th
e
Alzh
eim
e
r
’
s
d
is
ea
s
e
d
ataset
.
ii)
A
f
itn
ess
f
u
n
ctio
n
d
esig
n
ed
ex
p
licitly
f
o
r
a
d
ec
is
io
n
tr
ee
class
if
ier
is
em
p
lo
y
ed
to
r
ed
u
ce
th
e
n
u
m
b
er
o
f
f
ea
t
u
r
es.
iii)
W
e
h
av
e
in
teg
r
ated
v
ar
io
u
s
class
if
icatio
n
alg
o
r
ith
m
s
(
Dec
is
io
n
et
a
l.
)
with
s
p
ec
if
ic
f
ea
tu
r
es.
iv
)
W
e
in
v
esti
g
ated
p
a
r
am
eter
s
,
in
clu
d
i
n
g
p
o
p
u
latio
n
s
ize
an
d
iter
atio
n
s
,
to
id
en
tify
th
e
m
o
s
t
ef
f
e
ctiv
e
co
n
f
ig
u
r
atio
n
s
.
An
d
v
)
W
e
m
eticu
lo
u
s
ly
ev
alu
ated
th
e
p
er
f
o
r
m
a
n
ce
o
f
DE
,
GW
O,
an
d
POA
u
s
in
g
th
e
F1
-
s
co
r
e,
p
r
ec
is
io
n
,
r
e
ca
ll,
an
d
ac
cu
r
ac
y
,
en
s
u
r
in
g
a
r
o
b
u
s
t a
n
aly
s
is
.
2.
RE
L
AT
E
D
WO
RK
S
Ma
r
tin
ez
-
Mu
r
cia
et
a
l.
[
2
3
]
e
m
p
lo
y
d
ee
p
co
n
v
o
lu
tio
n
al
au
t
o
en
co
d
e
r
s
to
an
aly
ze
Alzh
eim
er
’
s
d
is
ea
s
e
d
ata.
T
h
eir
m
et
h
o
d
en
tails
th
e
d
ec
o
m
p
o
s
itio
n
o
f
m
ag
n
etic
r
eso
n
an
ce
im
ag
i
n
g
(
MRI)
im
ag
es
an
d
r
etr
ie
v
in
g
MRI
f
ea
tu
r
es
co
r
r
elate
d
with
co
g
n
itiv
e
s
y
m
p
to
m
s
,
th
er
eb
y
f
ac
ilit
atin
g
th
e
co
m
p
r
eh
en
s
io
n
o
f
n
eu
r
o
d
eg
en
e
r
ativ
e
p
r
o
ce
s
s
es.
T
h
e
im
p
ac
t o
f
th
e
b
r
ai
n
o
n
ea
c
h
au
to
e
n
co
d
er
m
an
if
o
ld
c
o
o
r
d
in
ate
is
d
eter
m
in
ed
th
r
o
u
g
h
r
e
g
r
ess
io
n
an
d
class
if
icatio
n
an
aly
s
es,
wh
ich
in
v
o
lv
e
th
e
ex
a
m
in
atio
n
o
f
t
h
e
e
x
tr
ac
ted
f
ea
tu
r
es
in
v
ar
io
u
s
co
m
b
in
atio
n
s
.
T
h
e
a
cc
u
r
ac
y
o
f
Alzh
eim
er
’
s
d
is
e
ase
(
AD)
d
iag
n
o
s
is
is
g
r
ea
ter
th
an
8
0
%
wh
en
im
ag
in
g
-
d
e
r
iv
ed
m
ar
k
er
s
ar
e
u
s
ed
in
co
n
j
u
n
ctio
n
with
MM
SE
o
r
ADAS1
1
s
co
r
es.
T
h
e
class
if
icatio
n
o
f
AD
h
as
b
ee
n
s
u
b
s
tan
tially
en
h
an
ce
d
b
y
ap
p
ly
in
g
m
ac
h
in
e
lear
n
in
g
(
ML
)
an
d
d
ata
m
in
in
g
.
T
h
e
class
if
icatio
n
o
f
Alzh
eim
er
’
s
d
is
ea
s
e
s
t
ag
es
u
s
in
g
d
atasets
s
u
ch
as
th
e
ADNI
an
d
th
e
T
ADPOL
E
ch
allen
g
e
h
as
b
e
en
th
e
s
u
b
ject
o
f
n
u
m
e
r
o
u
s
s
tu
d
ies
th
at
h
av
e
em
p
lo
y
ed
alg
o
r
ith
m
s
s
u
ch
as
K
-
n
ea
r
est
n
eig
h
b
o
r
s
(
K
-
NN)
,
d
ec
is
io
n
tr
ee
s
,
n
aiv
e
B
ay
es,
g
en
e
r
alize
d
lin
ea
r
m
o
d
els
(
GL
M)
,
an
d
d
ee
p
lear
n
in
g
.
T
h
ese
m
et
h
o
d
s
h
av
e
d
em
o
n
s
tr
ated
p
r
o
m
is
in
g
r
esu
lts
,
with
GL
M
attain
in
g
an
8
8
.
2
4
%
ac
c
u
r
ac
y
in
class
if
y
in
g
AD
s
tag
es.
Nev
er
th
eless
,
th
e
ch
allen
g
es
o
f
d
at
a
s
p
ar
s
ity
an
d
co
m
p
r
eh
en
s
iv
e
f
ea
tu
r
e
s
elec
tio
n
p
er
s
is
t.
I
t
is
im
p
er
ativ
e
to
im
p
r
o
v
e
t
h
e
r
ep
r
esen
tatio
n
o
f
u
n
d
e
r
r
ep
r
esen
te
d
class
es
a
n
d
en
h
a
n
ce
f
ea
tu
r
e
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
.
3
,
J
u
n
e
20
25
:
3
3
8
2
-
3
3
9
5
3384
s
elec
tio
n
to
im
p
r
o
v
e
th
e
ac
c
u
r
ac
y
an
d
e
f
f
ec
tiv
en
ess
o
f
AD
class
if
icatio
n
.
Key
ch
ar
ac
ter
is
tics
in
clu
d
e
th
e
clin
ical
d
em
en
tia
r
atin
g
s
u
m
o
f
b
o
x
es
(
C
DR
SB
)
co
g
n
itiv
e
test
,
p
atien
t a
g
e,
an
d
o
v
e
r
all
b
r
a
in
v
o
lu
m
e
[
2
4
]
.
T
h
e
s
tu
d
y
“
ea
r
ly
-
s
tag
e
Alzh
eim
er
’
s
d
is
ea
s
e
p
r
e
d
ictio
n
u
s
in
g
m
ac
h
in
e
lea
r
n
in
g
m
o
d
els
”
was
co
n
d
u
cte
d
b
y
Kav
ith
a
et
al.
[
2
5
]
.
T
h
e
s
tu
d
y
u
tili
ze
d
d
atasets
f
r
o
m
th
e
o
p
en
ac
ce
s
s
s
er
ies
o
f
im
ag
in
g
s
tu
d
ies
(
OASI
S)
an
d
Kag
g
le
to
p
r
ed
ict
ea
r
ly
AD.
Usi
n
g
p
r
ec
is
io
n
,
r
ec
all,
ac
cu
r
ac
y
,
an
d
F1
-
s
co
r
e
m
etr
ics,
th
ey
ass
es
s
ed
th
e
p
er
f
o
r
m
an
ce
o
f
d
ec
is
io
n
tr
ee
(
DT
)
,
r
an
d
o
m
f
o
r
est
(
R
F)
,
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
in
e
(
SVM)
,
g
r
ad
ien
t
b
o
o
s
tin
g
,
an
d
v
o
tin
g
class
if
ier
s
.
T
h
e
p
o
ten
tial
o
f
ML
tech
n
iq
u
es
f
o
r
ea
r
ly
AD
d
iag
n
o
s
is
was
d
em
o
n
s
tr
ated
b
y
th
e
m
ax
im
u
m
v
alid
atio
n
ac
c
u
r
ac
ies
o
f
8
6
.
9
2
%
a
n
d
8
5
.
9
2
%
a
ttain
ed
b
y
th
e
RF
an
d
ex
tr
e
m
e
g
r
ad
ie
n
t
b
o
o
s
tin
g
(
XGBo
o
s
t
)
m
o
d
els,
r
esp
ec
tiv
ely
.
T
h
e
m
o
r
tality
r
ates
o
f
Al
zh
eim
er
’
s
d
is
ea
s
e
ar
e
r
ed
u
ce
d
as
a
r
esu
lt
o
f
th
e
m
o
r
e
ef
f
ec
tiv
e
tr
ea
tm
e
n
t
th
at
is
f
ac
ilit
ated
b
y
ea
r
ly
d
etec
tio
n
.
T
h
e
s
tu
d
y
em
p
h
asizes
th
e
im
p
o
r
tan
ce
o
f
m
ac
h
in
e
lear
n
in
g
(
ML
)
in
p
r
o
v
id
in
g
clin
ician
s
with
th
e
n
ec
ess
ar
y
to
o
ls
to
en
h
an
ce
p
at
ien
t
o
u
tco
m
es
an
d
im
p
lem
en
t
tim
ely
in
ter
v
en
tio
n
s
.
Fu
tu
r
e
r
esear
ch
will
co
n
ce
n
tr
ate
o
n
r
ef
in
i
n
g
f
ea
tu
r
e
s
ele
ctio
n
an
d
ex
p
l
o
r
in
g
n
ew
f
ea
tu
r
es to
im
p
r
o
v
e
d
iag
n
o
s
tic
ac
cu
r
ac
y
.
Salv
ato
r
e
et
a
l.
[
2
6
]
em
p
lo
y
e
d
m
ac
h
in
e
lear
n
in
g
tech
n
iq
u
e
s
to
an
aly
ze
MRI
d
ata
f
r
o
m
th
e
ADNI
in
o
r
d
er
to
id
en
tif
y
ea
r
ly
b
io
m
ar
k
er
s
f
o
r
AD.
T
h
e
s
u
b
jects
we
r
e
ca
teg
o
r
ized
as
co
g
n
itiv
ely
n
o
r
m
al
(
C
N)
,
AD,
m
ild
co
g
n
itiv
e
im
p
air
m
en
t
c
o
n
v
er
ter
s
(
MCIc
)
,
an
d
n
o
n
-
c
o
n
v
er
ter
s
(
MCIn
c)
u
s
in
g
p
r
i
n
cip
al
co
m
p
o
n
en
t
an
aly
s
is
(
PC
A)
f
o
r
f
ea
tu
r
e
e
x
tr
ac
tio
n
an
d
s
u
p
p
o
r
t
v
e
cto
r
m
ac
h
in
es
(
SVM)
f
o
r
class
if
icatio
n
.
W
e
f
o
u
n
d
s
ig
n
if
ican
t
MRI
b
io
m
ar
k
er
s
i
n
th
e
f
o
llo
win
g
a
r
ea
s
:
T
h
e
h
i
p
p
o
ca
m
p
u
s
,
b
asal
g
a
n
g
lia,
en
to
r
h
in
al
co
r
tex
,
a
n
d
ce
r
eb
ellu
m
wer
e
all
id
en
tifie
d
as
s
ites
o
f
s
u
b
s
tan
tial
MRI
b
io
m
ar
k
er
s
.
7
6
%
f
o
r
AD
v
s
.
C
N,
7
2
%
f
o
r
MCIc
v
s
.
C
N,
an
d
6
6
%
f
o
r
MCIc
v
s
.
MCIn
c
wer
e
th
e
clas
s
if
icatio
n
ac
cu
r
ac
ies.
T
h
e
p
o
ten
tial
o
f
ML
to
im
p
r
o
v
e
th
e
ea
r
ly
d
iag
n
o
s
is
an
d
m
an
ag
em
en
t
o
f
Alzh
eim
er
’
s
d
is
ea
s
e
is
u
n
d
er
s
co
r
e
d
b
y
t
h
ese
f
in
d
in
g
s
,
wh
ich
will
en
ab
le
th
e
d
ev
elo
p
m
en
t o
f
m
o
r
e
ef
f
ec
tiv
e
tr
ea
tm
en
ts
an
d
r
e
d
u
ce
th
e
d
u
r
atio
n
a
n
d
c
o
s
t o
f
clin
ical
tr
ials
.
I
n
th
e
d
etec
tio
n
an
d
class
if
icatio
n
o
f
AD,
Kis
h
o
r
e
et
a
l.
[
2
7
]
ex
a
m
in
ed
t
h
e
ef
f
icac
y
o
f
a
d
iv
er
s
e
ar
r
ay
o
f
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
s
,
in
clu
d
in
g
lo
g
is
tic
r
e
g
r
ess
io
n
,
DT
,
RF
,
n
aiv
e
B
ay
e
s
(
NB
)
,
an
d
SVM.
C
r
o
s
s
-
v
alid
atio
n
,
f
ea
tu
r
e
s
elec
tio
n
,
an
d
d
ata
p
r
ep
r
o
ce
s
s
in
g
wer
e
im
p
lem
en
te
d
u
s
in
g
clin
ical
d
ata
an
d
MRI
s
ca
n
s
to
s
u
r
m
o
u
n
t
th
e
co
n
s
tr
ai
n
ts
o
f
co
n
v
en
tio
n
al
m
eth
o
d
s
.
T
h
e
SVM
with
a
lin
ea
r
k
er
n
el
(
C
=2
)
ac
h
iev
es
th
e
h
ig
h
est
ac
cu
r
ac
y
o
f
9
5
%,
f
o
llo
wed
b
y
n
aiv
e
B
ay
es
an
d
lo
g
is
tic
r
eg
r
ess
io
n
with
9
3
%.
T
h
ese
r
esu
lts
u
n
d
er
s
co
r
e
th
e
s
ig
n
if
ican
t
p
o
t
en
tial
o
f
t
h
ese
alg
o
r
ith
m
s
to
i
m
p
r
o
v
e
th
e
ea
r
ly
d
etec
tio
n
o
f
Alzh
eim
er
’
s
d
is
ea
s
e
an
d
to
p
r
o
v
i
d
e
th
e
r
e
q
u
is
ite
in
ter
v
en
tio
n
s
.
Siv
ak
an
i
an
d
An
s
ar
i
[
2
8
]
u
tili
ze
d
th
e
OASI
S
lo
n
g
itu
d
in
al
MRI
d
ataset
to
ex
am
in
e
t
h
e
e
f
f
ec
tiv
en
ess
o
f
m
ac
h
in
e
lear
n
in
g
alg
o
r
ith
m
s
in
th
e
ea
r
ly
id
en
tific
atio
n
an
d
class
if
icatio
n
o
f
Alzh
ei
m
er
’
s
d
is
ea
s
e
.
T
h
eir
r
esear
ch
u
n
d
e
r
s
co
r
ed
th
e
s
ig
n
if
ican
ce
o
f
f
ea
tu
r
e
ex
tr
ac
ti
o
n
an
d
s
elec
tio
n
in
en
h
a
n
ci
n
g
th
e
ac
cu
r
ac
y
o
f
Alzh
eim
er
’
s
d
is
ea
s
e
p
r
ed
ictio
n
s
.
T
h
e
ex
p
ec
tatio
n
-
m
a
x
im
izatio
n
(
E
M)
tech
n
iq
u
e
was
u
tili
ze
d
f
o
r
clu
s
ter
in
g
,
b
est
-
f
ir
s
t
s
ea
r
ch
was
p
er
f
o
r
m
ed
f
o
r
f
ea
tu
r
e
s
elec
tio
n
,
an
d
th
e
d
ataset
was
p
r
ep
r
o
ce
s
s
ed
to
ad
d
r
ess
m
is
s
in
g
v
alu
es.
T
o
d
o
class
if
icatio
n
,
l
in
ea
r
r
e
g
r
ess
io
n
an
d
Gau
s
s
ian
p
r
o
ce
s
s
m
o
d
els
wer
e
em
p
l
o
y
ed
.
T
h
e
Gau
s
s
ian
p
r
o
ce
s
s
ap
p
r
o
ac
h
e
x
h
ib
ited
h
ig
h
ef
f
icien
cy
a
n
d
ac
c
u
r
ac
y
in
Alzh
eim
e
r
’
s
d
is
ea
s
e
class
if
icatio
n
,
with
a
r
ec
eiv
er
o
p
er
atin
g
c
h
ar
ac
ter
is
tic
(
R
OC
)
ac
cu
r
ac
y
o
f
9
4
.
6
5
%.
T
h
e
r
esear
ch
f
in
d
s
t
h
at
t
h
e
ef
f
e
ctiv
en
ess
o
f
m
ac
h
in
e
lear
n
in
g
m
o
d
els
in
id
en
tify
in
g
Alzh
eim
er
’
s
d
is
ea
s
e
is
s
ig
n
if
ican
tly
im
p
r
o
v
ed
with
th
e
ap
p
licatio
n
o
f
ef
f
icien
t f
ea
tu
r
e
e
x
tr
ac
tio
n
a
n
d
s
elec
tio
n
tech
n
iq
u
es
3.
B
ACK
G
RO
UND
3
.
1
.
O
v
er
v
iew
of
m
e
t
a
heuris
t
ic
o
p
t
i
mi
z
a
t
i
o
n
Me
tah
eu
r
is
tic
o
p
tim
izatio
n
is
o
n
e
o
f
th
e
n
u
m
er
o
u
s
m
eth
o
d
s
f
o
r
r
eso
lv
in
g
in
t
r
icate
o
p
tim
izatio
n
p
r
o
b
lem
s
.
I
t
is
d
iv
i
d
ed
in
to
th
r
ee
ca
te
g
o
r
ies:
ev
o
lu
tio
n
a
r
y
co
m
p
u
tatio
n
,
n
at
u
r
e
-
in
s
p
ir
ed
alg
o
r
ith
m
s
,
an
d
s
war
m
in
tellig
en
ce
[
2
9
]
.
E
v
en
wh
en
u
n
s
ch
ed
u
led
p
h
e
n
o
m
en
a
o
cc
u
r
d
u
r
i
n
g
th
e
s
ea
r
ch
,
m
etah
eu
r
is
tic
alg
o
r
ith
m
s
ca
n
s
ea
r
ch
f
o
r
d
is
p
er
s
ed
s
o
lu
tio
n
s
p
ac
es
an
d
f
o
cu
s
lo
g
ic
to
ac
h
iev
e
a
g
lo
b
al
m
in
im
u
m
s
o
lu
tio
n
[
3
0
]
.
3
.
1
.
1
.
D
i
f
f
e
r
e
n
t
i
a
l
e
v
o
lu
t
i
o
n
Dif
f
er
en
tial
ev
o
lu
tio
n
is
a
p
o
p
u
latio
n
-
b
ased
m
etah
e
u
r
is
tic
p
r
o
ce
s
s
th
at
in
itializes
its
p
o
p
u
latio
n
in
th
e
s
ea
r
ch
s
p
ac
e
an
d
u
p
d
ate
s
th
is
p
o
p
u
latio
n
with
th
e
b
est
o
f
f
s
p
r
in
g
f
r
o
m
th
e
c
o
m
p
etitio
n
b
etwe
en
a
p
o
p
u
latio
n
m
em
b
er
a
n
d
a
s
y
n
th
esized
m
u
tan
t
v
ec
to
r
g
en
e
r
a
tio
n
af
ter
g
en
e
r
atio
n
(
o
r
iter
at
io
n
af
ter
iter
atio
n
)
b
ased
o
n
t
h
e
th
r
ee
p
r
in
ci
p
l
es
d
escr
ib
ed
h
er
e
[
3
1
]
.
A
co
n
tin
u
o
u
s
r
in
g
-
s
witch
in
g
d
i
f
f
er
en
tial
ev
o
lu
tio
n
alg
o
r
ith
m
(
DE
jDE
)
th
at
is
co
llectiv
ely
d
eter
m
in
e
d
b
y
th
e
a
lter
n
ativ
e
d
is
p
u
te
r
eso
lu
tio
n
m
ain
clau
s
e
(
ADM
)
o
r
o
p
tim
izin
g
co
n
tr
o
l
a
n
d
in
t
ellig
en
ce
(
OC
I
)
in
o
r
d
er
to
s
y
n
th
esize
a
m
u
tan
t
v
ec
to
r
c
o
n
t
r
o
ls
th
e
p
o
p
u
latio
n
s
tr
ateg
y
.
DE
NN
en
s
em
b
les
h
a
v
e
d
ev
elo
p
ed
a
v
ar
iety
o
f
e
v
o
lu
tio
n
ar
y
s
tr
ateg
ies
to
ac
h
ie
v
e
th
is
.
T
h
e
s
tr
ateg
y
f
o
r
e
v
o
lv
in
g
t
h
e
n
ew
p
a
r
am
et
er
s
f
o
r
th
e
f
o
r
th
co
m
in
g
d
en
o
t
atio
n
o
f
DE
m
ay
b
e
d
eter
m
i
n
ed
b
y
th
e
d
y
n
am
ics
o
f
th
o
s
e
r
in
g
s
[
3
2
]
,
[
3
3
]
.
T
h
e
r
in
g
s
o
f
DE
NN
th
at
will b
e
em
p
lo
y
ed
a
r
e
th
en
d
eter
m
in
e
d
b
y
a
m
eta
-
an
aly
s
is
.
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
E
n
h
a
n
cin
g
A
lz
h
eime
r
’
s
d
is
ea
s
e
d
ia
g
n
o
s
is
th
r
o
u
g
h
meta
h
eu
r
is
tic
fea
tu
r
e
…
(
A
r
a
r
A
l
-
Ta
w
i
l
)
3385
T
h
e
f
u
n
d
a
m
en
tal
co
n
ce
p
ts
th
at
DE
d
ea
ls
with
i
n
c
l
u
d
e
:
a.
Po
p
u
latio
n
-
B
ased
Sear
ch
:
D
E
o
p
er
ates
o
n
a
p
o
p
u
latio
n
o
f
p
o
ten
tial
s
o
lu
tio
n
s
.
E
ac
h
m
em
b
er
o
f
th
e
p
o
p
u
latio
n
h
as th
e
p
o
ten
tial to
s
o
lv
e
th
e
o
p
tim
izatio
n
p
r
o
b
le
m
.
b.
Selectio
n
:
T
h
e
o
b
jectiv
e
f
u
n
c
tio
n
s
er
v
es
as
a
g
u
id
in
g
lig
h
t
in
DE
.
I
t
ass
ess
es
th
e
tr
ial
v
ec
to
r
an
d
if
th
e
ex
p
er
im
en
tal
v
ec
to
r
p
r
o
d
u
ce
s
a
s
u
p
er
i
o
r
o
b
jectiv
e
f
u
n
ctio
n
v
alu
e
it
r
ep
lace
s
th
e
tar
g
e
t
v
ec
to
r
i
n
th
e
p
o
p
u
latio
n
f
o
r
th
e
n
e
x
t g
en
e
r
a
tio
n
.
c.
C
r
o
s
s
o
v
er
:
DE
v
alu
es
d
iv
er
s
it
y
an
d
th
is
is
ev
id
en
t
in
it
s
cr
o
s
s
o
v
er
o
p
er
atio
n
.
T
h
is
o
p
er
ati
o
n
in
v
o
lv
es
th
e
co
m
b
in
atio
n
o
f
elem
e
n
ts
f
r
o
m
th
e
m
u
tan
t
v
ec
to
r
an
d
th
e
tar
g
et
v
ec
to
r
to
g
en
er
ate
a
n
ex
p
e
r
i
m
en
tal
v
ec
to
r
.
d.
Mu
tatio
n
:
Mu
tatio
n
is
th
e
f
u
n
d
am
en
tal
m
e
ch
an
is
m
o
f
DE
in
wh
ich
n
ew
ca
n
d
id
ate
s
o
lu
tio
n
s
(
m
u
ta
n
t
v
ec
to
r
s
)
ar
e
g
en
er
ated
b
y
c
o
m
b
in
in
g
e
x
is
tin
g
s
o
lu
tio
n
s
.
3
.
1
.
2
.
G
r
e
y
w
o
l
f
o
p
t
i
mi
z
e
r
To
r
eso
lv
e
g
lo
b
al
o
p
tim
izatio
n
is
s
u
es
th
e
GW
O
em
u
lates
th
e
s
o
cial
h
ier
ar
ch
y
an
d
p
r
e
d
ato
r
y
b
eh
av
i
o
r
o
f
g
r
ey
w
o
lv
es.
I
t
ass
ig
n
s
wo
l
v
es
to
alp
h
a,
b
eta,
an
d
o
m
eg
a
r
o
les
with
alp
h
as b
ein
g
th
e
lea
d
er
s
of
t
h
e
p
u
r
s
u
it.
T
h
e
alg
o
r
ith
m
is
r
ec
o
g
n
ized
f
o
r
its
co
m
p
etitiv
e
p
e
r
f
o
r
m
an
ce
an
d
r
a
p
id
co
n
v
er
g
en
ce
d
is
tin
g
u
is
h
in
g
it
f
r
o
m
o
th
er
o
p
tim
izatio
n
alg
o
r
ith
m
s
.
I
t is ef
f
ec
tiv
e
f
o
r
b
o
th
co
n
s
tr
ain
ed
an
d
u
n
c
o
n
s
tr
ain
ed
p
r
o
b
l
em
s
[
3
4
]
.
T
h
e
f
u
n
d
am
en
tal
co
n
ce
p
ts
th
at
GW
O
d
ea
ls
with
i
n
c
l
u
d
e
:
a.
L
ea
d
er
s
h
ip
h
ier
ar
ch
y
: G
W
O
m
o
d
els
th
e
s
o
cial
h
ier
ar
ch
y
of
g
r
ey
wo
lv
es
d
iv
id
in
g
th
em
in
t
o
f
o
u
r
ca
te
g
o
r
i
e
s
:
−
Alp
h
a:
T
h
e
b
est s
o
lu
tio
n
i
n
th
e
p
o
p
u
latio
n
r
ep
r
esen
tin
g
th
e
l
ea
d
er
.
−
B
eta:
T
h
e
s
ec
o
n
d
-
b
est s
o
lu
tio
n
ass
is
tin
g
th
e
alp
h
a
in
d
ec
is
io
n
-
m
ak
in
g
.
−
Delta:
T
h
e
th
ir
d
-
b
est s
o
lu
tio
n
f
o
llo
win
g
th
e
al
p
h
a
an
d
b
eta.
−
Om
eg
a:
T
h
e
r
em
ai
n
in
g
wo
l
v
e
s
f
o
llo
win
g
th
e
lead
e
r
s
.
b.
Hu
n
tin
g
m
ec
h
an
is
m
:
T
h
e
h
u
n
tin
g
p
r
o
ce
s
s
o
f
g
r
e
y
wo
lv
es
is
m
ath
em
atica
lly
m
o
d
eled
to
g
u
id
e
th
e
s
ea
r
ch
f
o
r
o
p
tim
al
s
o
lu
tio
n
s
.
T
h
is
p
r
o
ce
s
s
in
clu
d
es
th
r
ee
m
ain
p
h
ases
:
en
cir
clin
g
p
r
ey
,
h
u
n
tin
g
,
an
d
attac
k
in
g
p
r
ey
(
e
x
p
lo
itatio
n
)
.
=
|
−
|
(
1
)
(
+
1
)
=
1
+
2
+
3
3
(
2
)
3
.
1
.
3
.
P
a
r
r
o
t
o
p
t
i
mi
z
a
t
i
o
n
Par
r
o
t
o
p
tim
izatio
n
is
a
lite
r
atu
r
e
-
b
ased
alg
o
r
ith
m
p
r
ed
i
ca
ted
o
n
p
ar
r
o
ts
’
ex
ce
p
tio
n
a
l
lear
n
in
g
ca
p
ab
ilit
ies
[
3
5
]
.
PO
is
a
p
o
p
u
latio
n
-
b
ased
o
p
tim
izatio
n
alg
o
r
ith
m
o
f
ex
ce
p
tio
n
al
q
u
a
lity
th
at
s
ee
k
s
th
e
o
p
tim
al
s
o
lu
tio
n
in
a
r
ea
l
-
v
alu
ed
s
ea
r
ch
s
p
ac
e.
Par
r
o
ts
’
lear
n
in
g
is
g
o
v
er
n
ed
b
y
th
eir
ac
tio
n
o
r
ie
n
tatio
n
w
h
ich
m
o
tiv
ates
th
is
alg
o
r
ith
m
to
tr
ac
k
th
e
s
ea
r
ch
m
o
v
em
e
n
ts
b
y
m
ain
tain
in
g
a
b
alan
ce
b
etwe
en
ex
p
lo
r
atio
n
an
d
ex
p
lo
itatio
n
to
i
d
en
tify
t
h
e
o
p
t
im
al
lo
ca
tio
n
.
Alg
o
r
ith
m
Step
s
:
a.
I
n
itializatio
n
:
Gen
er
ate
an
in
itial
p
o
p
u
latio
n
o
f
ca
n
d
id
ate
s
o
l
u
tio
n
s
r
an
d
o
m
l
y
.
E
v
alu
ate
th
e
f
itn
ess
o
f
ea
ch
in
d
iv
id
u
al
u
s
in
g
th
e
o
b
jectiv
e
f
u
n
ctio
n
.
b.
L
ea
d
er
s
elec
tio
n
:
I
d
en
tif
y
a
s
u
b
s
et
o
f
th
e
b
est
in
d
iv
id
u
al
s
in
th
e
p
o
p
u
latio
n
as
lead
er
s
.
T
h
ese
lead
er
s
in
f
lu
en
ce
th
e
m
o
v
em
en
t o
f
o
th
er
in
d
iv
id
u
als.
c.
C
o
m
m
u
n
icatio
n
an
d
m
o
v
em
en
t
:
T
h
e
p
o
s
itio
n
o
f
ea
ch
in
d
iv
id
u
al
is
u
p
d
ated
in
ac
c
o
r
d
an
ce
with
th
e
in
f
o
r
m
atio
n
it r
ec
eiv
es f
r
o
m
le
ad
er
s
an
d
n
eig
h
b
o
r
in
g
p
ar
r
o
ts
.
(
+
1
)
=
(
)
+
1
(
–
(
)
)
+
2
(
ℎ
−
(
)
)
(
3
)
3
.
2
.
O
v
e
r
v
i
e
w
o
f
c
l
a
s
s
i
f
i
c
a
t
i
o
n
3
.
2
.
1
.
D
e
c
i
s
i
o
n
t
r
e
e
A
d
ec
is
io
n
tr
ee
(
DT
)
is
a
n
o
n
-
p
ar
am
etr
ic
s
u
p
er
v
is
ed
le
ar
n
in
g
alg
o
r
ith
m
f
o
r
class
if
icatio
n
an
d
r
eg
r
ess
io
n
task
s
.
I
t
g
en
er
ates
a
tr
ee
s
tr
u
ctu
r
e
o
f
d
ec
is
io
n
s
an
d
th
eir
p
o
ten
tial
r
ep
er
c
u
s
s
io
n
s
b
y
r
ec
u
r
s
iv
ely
d
iv
id
in
g
th
e
d
ataset
in
to
s
u
b
s
ets
b
ased
o
n
th
e
m
o
s
t
s
i
g
n
if
ican
t
attr
ib
u
te
[
3
6
]
.
E
ac
h
in
ter
n
al
n
o
d
e
in
class
if
icatio
n
r
ep
r
esen
ts
a
“
te
s
t
”
o
n
an
attr
ib
u
te,
ea
ch
b
r
an
ch
r
ep
r
esen
ts
th
e
test
r
esu
lt,
an
d
ea
ch
leaf
n
o
d
e
r
ep
r
esen
ts
a
class
lab
el
o
r
a
co
n
tin
u
o
u
s
v
alu
e
in
r
eg
r
ess
io
n
.
T
h
e
d
iv
id
in
g
p
r
o
ce
s
s
co
n
tin
u
es
u
n
til
a
h
altin
g
cr
iter
io
n
is
m
et,
s
u
ch
as
co
m
p
letin
g
all
s
am
p
les
in
a
n
o
d
e
th
at
b
elo
n
g
s
to
th
e
s
am
e
class
,
a
ttain
in
g
a
m
ax
im
al
tr
ee
d
ep
th
,
o
r
lack
in
g
an
y
ad
d
itio
n
al
in
f
o
r
m
atio
n
g
ain
.
Dec
is
io
n
tr
ee
s
ar
e
s
im
p
le
to
u
n
d
e
r
s
tan
d
an
d
in
ter
p
r
et,
r
eq
u
ir
e
m
i
n
im
al
d
ata
p
r
ep
r
o
c
ess
in
g
,
an
d
ca
n
b
e
em
p
l
o
y
ed
to
an
aly
ze
n
u
m
er
ical
a
n
d
ca
teg
o
r
ical
d
ata
[
3
7
]
.
Nev
er
th
eless
,
th
ey
ar
e
p
r
o
n
e
t
o
o
v
er
f
itti
n
g
if
n
o
t
p
r
u
n
ed
,
a
n
d
th
ey
m
a
y
b
e
u
n
s
tab
le
d
u
e
to
th
e
p
o
ten
tial
f
o
r
a
co
m
p
letely
d
if
f
er
en
t tr
ee
t
o
e
m
er
g
e
f
r
o
m
m
o
d
est d
ata
ch
a
n
g
es.
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
.
3
,
J
u
n
e
20
25
:
3
3
8
2
-
3
3
9
5
3386
3
.
2
.
2
.
R
a
n
d
o
m
f
o
r
e
s
t
R
an
d
o
m
f
o
r
est
is
an
en
s
em
b
l
e
lear
n
in
g
m
eth
o
d
th
at
d
eter
m
in
es
th
e
class
b
ased
o
n
th
e
m
o
d
e
o
f
t
h
e
class
es
(
cla
s
s
if
icatio
n
)
o
r
th
e
m
ea
n
p
r
e
d
ictio
n
(
r
e
g
r
ess
io
n
)
o
f
t
h
e
in
d
iv
id
u
al
tr
ee
s
d
u
r
in
g
tr
ain
i
n
g
[
3
8
]
.
I
t
co
n
s
tr
u
cts
a
d
ec
is
io
n
tr
ee
f
o
r
ea
c
h
s
u
b
s
et
a
n
d
g
en
er
ate
s
n
u
m
er
o
u
s
s
u
b
d
i
v
is
io
n
s
o
f
th
e
o
r
i
g
in
al
d
ataset
th
r
o
u
g
h
b
o
o
ts
tr
ap
s
am
p
lin
g
(
r
an
d
o
m
s
am
p
lin
g
with
r
ep
lace
m
en
t)
.
A
r
an
d
o
m
s
u
b
s
et
o
f
f
e
atu
r
es
is
ch
o
s
en
at
ea
ch
n
o
d
e,
an
d
th
e
m
o
s
t
ap
p
r
o
p
r
iate
f
ea
tu
r
e
f
r
o
m
th
is
s
u
b
s
et
is
u
s
ed
to
d
iv
id
e
th
e
n
o
d
e
[
3
9
]
.
T
h
e
co
r
r
elatio
n
b
etwe
en
tr
ee
s
is
d
im
in
is
h
ed
as
a
r
esu
lt
o
f
t
h
is
v
ar
iab
ilit
y
.
T
h
e
p
r
e
d
ictio
n
s
f
r
o
m
all
in
d
iv
id
u
al
t
r
ee
s
ar
e
co
m
b
in
ed
to
p
r
o
d
u
ce
th
e
f
i
n
a
l
p
r
ed
ictio
n
.
C
o
m
p
ar
ed
to
in
d
iv
id
u
al
d
ec
is
io
n
tr
ee
s
,
r
an
d
o
m
f
o
r
ests
ar
e
m
o
r
e
ac
cu
r
ate,
ca
n
h
an
d
le
lar
g
e
d
at
asets
ef
f
icien
tly
,
an
d
ca
n
h
an
d
le
th
o
u
s
an
d
s
o
f
in
p
u
t
v
ar
iab
l
es
with
o
u
t
v
ar
iab
le
d
eletio
n
,
all
o
f
wh
ich
r
ed
u
ce
o
v
er
f
itti
n
g
[
4
0
]
.
Ho
wev
er
,
th
ey
ar
e
m
o
r
e
c
o
m
p
u
tatio
n
ally
in
ten
s
iv
e,
co
m
p
lex
,
an
d
less
in
ter
p
r
eta
b
le
th
a
n
in
d
iv
id
u
al
d
ec
is
io
n
tr
ee
s
.
T
h
e
o
p
t
im
izatio
n
o
f
en
g
i
n
ee
r
in
g
d
esi
g
n
s
,
th
e
tr
ain
in
g
o
f
n
eu
r
al
n
etwo
r
k
s
,
th
e
clu
s
ter
in
g
an
d
d
ata
m
in
in
g
o
f
d
at
a,
th
e
p
r
o
ce
s
s
in
g
o
f
im
ag
es
an
d
s
ig
n
als,
th
e
d
ev
elo
p
m
e
n
t
o
f
co
n
tr
o
l
s
y
s
t
em
s
,
an
d
t
h
e
s
im
u
latio
n
o
f
f
in
an
cial
d
ata
a
r
e
all
a
r
ea
s
in
wh
ich
th
e
y
ar
e
ex
ten
s
iv
ely
em
p
lo
y
ed
.
3
.
2
.
3
.
G
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
Gr
ad
ien
t
B
o
o
s
tin
g
is
a
v
alu
ab
le
m
ac
h
in
e
-
lear
n
i
n
g
tech
n
i
q
u
e
f
o
r
s
o
lv
in
g
r
eg
r
ess
io
n
an
d
clas
s
if
icatio
n
p
r
o
b
lem
s
.
I
t
en
tails
th
e
in
cr
em
en
tal
d
ev
elo
p
m
e
n
t
o
f
a
m
o
d
el
f
r
o
m
an
en
s
em
b
le
o
f
p
o
o
r
lear
n
er
s
,
ty
p
ically
d
ec
is
io
n
tr
ee
s
[
4
1
]
.
I
t
f
u
n
cti
o
n
s
b
y
in
itiatin
g
with
an
in
itial
p
r
ed
ictio
n
,
o
f
ten
th
e
m
e
an
o
f
th
e
o
b
jectiv
e
v
ar
iab
le
in
r
eg
r
ess
io
n
p
r
o
b
lem
s
,
an
d
s
u
b
s
eq
u
en
tly
in
co
r
p
o
r
at
in
g
d
ec
is
io
n
tr
ee
s
in
to
th
e
m
o
d
el.
E
ac
h
n
ew
tr
ee
p
r
ed
icts
th
e
p
r
ev
i
o
u
s
m
o
d
el’
s
r
esid
u
al
er
r
o
r
s
(
d
if
f
er
e
n
ce
s
b
etwe
en
ac
t
u
al
an
d
p
r
ed
icte
d
v
alu
es)
[
4
2
]
.
T
h
e
alg
o
r
ith
m
m
in
im
izes
th
e
lo
s
s
f
u
n
ctio
n
th
r
o
u
g
h
g
r
a
d
ien
t
d
escen
t,
wh
ich
en
s
u
r
es
th
at
th
e
n
ew
m
o
d
el
is
in
alig
n
m
en
t
with
t
h
e
n
e
g
ativ
e
g
r
ad
ien
t
o
f
t
h
e
lo
s
s
f
u
n
ctio
n
ab
o
u
t
th
e
p
r
e
d
ictio
n
s
o
f
th
e
cu
r
r
en
t
m
o
d
el.
T
h
e
lear
n
in
g
r
ate
r
ed
u
ce
s
th
e
c
o
n
tr
ib
u
tio
n
o
f
ea
ch
n
ew
tr
ee
,
th
er
eb
y
ac
ce
ler
atin
g
th
e
lea
r
n
in
g
p
r
o
ce
s
s
an
d
im
p
r
o
v
in
g
g
en
er
aliza
tio
n
.
R
e
g
u
lar
izatio
n
tech
n
iq
u
es
s
u
ch
as
s
u
b
s
am
p
lin
g
,
L
1
/L2
r
e
g
u
l
ar
izatio
n
,
a
n
d
tr
ee
p
r
u
n
in
g
ar
e
im
p
lem
en
ted
[
4
3
]
to
p
r
ev
e
n
t
o
v
er
f
itti
n
g
.
Gr
ad
ien
t
b
o
o
s
tin
g
is
a
h
ig
h
ly
ac
cu
r
ate
an
d
r
o
b
u
s
t
m
eth
o
d
th
at
ca
n
p
er
f
o
r
m
r
eg
r
e
s
s
io
n
an
d
class
if
icatio
n
task
s
.
I
t
is
ca
p
ab
le
o
f
ac
co
m
m
o
d
atin
g
a
wid
e
v
ar
iety
o
f
lo
s
s
f
u
n
ctio
n
s
.
Ho
we
v
er
,
it
is
p
r
o
n
e
to
o
v
er
f
itti
n
g
if
s
cr
u
p
u
lo
u
s
ly
ca
lib
r
ated
,
an
d
tr
ai
n
in
g
ca
n
b
e
ef
f
icien
t
f
o
r
lar
g
e
d
atasets
.
I
t
i
s
wid
ely
u
s
ed
in
o
p
tim
izin
g
en
g
in
ee
r
in
g
d
esig
n
s
,
n
eu
r
al
n
etwo
r
k
tr
ain
i
n
g
,
clu
s
ter
in
g
an
d
d
ata
m
in
in
g
,
im
ag
e
an
d
s
ig
n
al
p
r
o
ce
s
s
in
g
,
co
n
tr
o
l sy
s
tem
s
,
an
d
f
in
a
n
cial
m
o
d
elin
g
.
3
.
2
.
4
.
X
G
B
o
o
s
t
E
x
tr
em
e
g
r
ad
ien
t
b
o
o
s
tin
g
is
a
s
o
p
h
is
ticated
m
ac
h
in
e
lea
r
n
in
g
al
g
o
r
ith
m
b
ased
o
n
th
e
g
r
ad
ien
t
b
o
o
s
tin
g
f
r
am
ewo
r
k
an
d
en
g
in
ee
r
ed
to
ac
h
iev
e
h
ig
h
e
f
f
icien
cy
a
n
d
p
er
f
o
r
m
a
n
ce
[
4
4
]
.
I
t
s
eq
u
e
n
tially
co
n
s
tr
u
cts
an
e
n
s
em
b
le
o
f
d
e
cisi
o
n
tr
ee
s
,
with
ea
ch
s
u
b
s
eq
u
en
t
tr
ee
r
ec
tify
in
g
th
e
er
r
o
r
s
o
f
th
e
p
r
ec
e
d
in
g
tr
ee
s
.
Ad
d
itio
n
ally
,
it
im
p
lem
en
ts
L
1
an
d
L
2
r
eg
u
lar
izatio
n
to
p
r
ev
en
t
o
v
er
f
itti
n
g
.
XGBo
o
s
t
is
ef
f
icien
t
with
s
p
ar
s
e
d
ata,
s
u
p
p
o
r
ts
p
ar
allel
p
r
o
ce
s
s
in
g
f
o
r
p
er
f
o
r
m
an
ce
,
a
n
d
in
ter
n
ally
m
a
n
ag
es
m
is
s
in
g
v
alu
es
[
4
5
]
.
I
t
ca
n
s
ca
le
to
lar
g
e
d
atasets
,
em
p
l
o
y
s
p
r
u
n
in
g
t
o
p
r
e
v
en
t
o
v
er
f
itti
n
g
,
an
d
p
er
m
its
th
e
u
s
e
o
f
cu
s
to
m
o
b
jectiv
e
f
u
n
ctio
n
s
a
n
d
ev
alu
atio
n
m
et
r
ics.
T
h
e
h
ig
h
ac
cu
r
ac
y
an
d
s
ca
lab
ilit
y
o
f
XGBo
o
s
t
h
a
v
e
m
ad
e
it
a
p
r
ef
er
r
e
d
ch
o
ice
in
f
ield
s
s
u
ch
as
f
in
a
n
ce
an
d
b
io
lo
g
y
,
an
d
it
is
wid
e
ly
u
s
ed
i
n
v
a
r
io
u
s
m
ac
h
in
e
le
ar
n
in
g
co
m
p
etitio
n
s
an
d
ap
p
licatio
n
s
.
4.
M
E
T
H
O
D
4
.
1
.
A
p
pr
o
a
c
h
Fig
u
r
e
1
d
ep
icts
th
e
p
r
im
ar
y
s
tag
es o
f
th
e
p
r
o
p
o
s
ed
ap
p
r
o
ac
h
im
p
lem
en
ted
in
th
is
in
v
esti
g
atio
n
.
T
h
is
m
eth
o
d
o
l
o
g
y
in
clu
d
es
th
r
ee
p
h
ases
.
Fo
u
r
b
io
-
in
s
p
ir
e
d
alg
o
r
ith
m
s
wer
e
im
p
lem
en
ted
t
o
id
en
tify
th
e
m
o
s
t
ad
v
an
tag
e
o
u
s
attr
ib
u
tes
d
u
r
in
g
th
e
in
itial
p
h
ase.
Fo
u
r
m
a
ch
in
e
lear
n
in
g
class
if
ier
s
wer
e
im
p
lem
en
ted
f
o
r
tr
ain
in
g
d
u
r
in
g
th
e
s
ec
o
n
d
p
h
a
s
e.
Ultim
ately
,
th
e
alg
o
r
ith
m
s
wer
e
v
alid
ated
u
s
in
g
th
e
p
er
f
o
r
m
an
ce
m
etr
ics.
4
.
1
.
1
.
D
a
t
a
s
e
t
A
s
in
g
le
d
ataset
was
u
s
ed
to
ass
es
s
th
e
p
r
o
p
o
s
ed
c
h
ar
ac
ter
i
s
tic
-
ch
o
ice
s
et
o
f
r
u
les:
th
e
Alzh
eim
er
’
s
d
is
ea
s
e
d
ataset
[
4
6
]
.
T
h
is
d
ataset
is
f
r
eq
u
en
tly
u
tili
ze
d
to
an
aly
ze
f
ac
to
r
s
ass
o
ciate
d
with
Alzh
eim
er
’
s
,
co
n
s
tr
u
ct
p
r
ed
ictiv
e
m
o
d
els,
an
d
co
n
d
u
ct
s
tatis
tical
an
aly
s
es.
I
t
is
p
u
b
licly
ac
ce
s
s
ib
le.
T
h
e
Alzh
eim
er
’
s
Dis
ea
s
e
d
ata
s
et
co
n
tain
s
th
e
c
o
m
p
r
eh
e
n
s
iv
e
h
ea
lth
s
tatis
tic
s
o
f
2
,
1
4
9
p
atien
ts
,
ea
ch
in
d
iv
i
d
u
ally
id
en
tifie
d
b
y
an
I
D
n
u
m
b
er
b
etwe
en
4
7
5
1
an
d
6
9
0
0
.
Dem
o
g
r
ap
h
ic
in
f
o
r
m
atio
n
,
life
s
ty
le
f
ac
to
r
s
,
s
cien
tific
r
ec
o
r
d
s
,
an
d
m
ed
ical,
co
g
n
itiv
e,
an
d
f
u
n
ctio
n
al
v
ar
iab
les co
m
p
r
is
e
th
e
Al
zh
eim
er
’
s
d
is
ea
s
e
d
ataset
.
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
E
n
h
a
n
cin
g
A
lz
h
eime
r
’
s
d
is
ea
s
e
d
ia
g
n
o
s
is
th
r
o
u
g
h
meta
h
eu
r
is
tic
fea
tu
r
e
…
(
A
r
a
r
A
l
-
Ta
w
i
l
)
3387
Fig
u
r
e
1.
Pro
p
o
s
ed
a
p
p
r
o
a
c
h
4
.
1
.
2
.
F
e
a
t
u
r
e
s
e
le
c
t
i
o
n
Var
io
u
s
attr
ib
u
tes
an
d
f
ea
tu
r
es
ar
e
in
co
r
p
o
r
ated
in
to
a
c
o
p
io
u
s
in
cr
ea
s
e
in
m
ed
ical
d
ata.
Mo
s
t
attr
ib
u
tes
d
o
n
o
t
co
n
tr
ib
u
te
to
p
r
e
d
ictiv
e
ap
p
licatio
n
o
u
tco
m
es,
in
cr
ea
s
in
g
co
m
p
u
tatio
n
tim
e
an
d
r
eso
u
r
ce
s
.
T
h
er
ef
o
r
e,
to
attain
h
ig
h
ac
c
u
r
ac
y
r
ates,
s
elec
tin
g
a
s
u
b
s
et
o
f
f
ea
tu
r
es
is
n
ec
ess
ar
y
.
Dif
f
er
en
tial
ev
o
lu
ti
o
n
(
DE
)
,
g
r
e
y
wo
lf
o
p
tim
izer
(
G
W
O)
,
an
d
p
ar
r
o
t
o
p
tim
izatio
n
alg
o
r
ith
m
(
POA)
wer
e
em
p
l
o
y
ed
in
t
h
is
s
tu
d
y
to
id
en
tify
t
h
e
o
p
tim
al
s
u
b
s
et
o
f
f
ea
tu
r
es
b
ased
o
n
th
e
Alzh
eim
er
’
s
d
is
ea
s
e
d
ataset
.
B
y
e
m
p
lo
y
in
g
a
f
itn
ess
f
u
n
ctio
n
th
at
was
d
ev
el
o
p
ed
e
x
p
licitly
f
o
r
a
d
ec
is
io
n
tr
ee
class
if
ier
,
th
e
d
ataset’
s
f
ea
tu
r
es
wer
e
d
im
in
is
h
ed
i
n
o
r
d
er
t
o
o
p
tim
ize
th
e
m
o
d
el
’
s
ac
cu
r
ac
y
.
I
n
o
r
d
e
r
to
ass
ess
ea
ch
s
u
b
s
et
o
f
f
ea
tu
r
es,
th
e
f
itn
ess
f
u
n
cti
o
n
tr
ain
s
a
d
ec
is
io
n
tr
ee
class
if
ier
o
n
th
e
tr
ain
in
g
s
et
(
7
0
%
o
f
th
e
d
ata)
an
d
ca
lcu
lates
th
e
ac
c
u
r
ac
y
o
n
th
e
test
s
et
(
3
0
%
o
f
th
e
d
at
a)
.
Acc
u
r
ac
y
is
th
e
p
r
o
p
o
r
tio
n
o
f
co
r
r
ec
t
p
r
e
d
ictio
n
s
th
e
class
if
ier
m
ak
es,
an
d
th
e
f
itn
ess
s
co
r
e
is
ca
lcu
lated
as
1
−
ac
cu
r
ac
y
.
T
h
e
f
itn
ess
f
u
n
ctio
n
ca
n
b
e
m
ath
e
m
atica
lly
r
ep
r
esen
ted
as
(
4
)
:
(
)
=
1
−
1
∑
(
,
ˆ
)
=
1
(
4
)
wh
er
e
S
is
th
e
s
u
b
s
et
o
f
s
elec
ted
f
ea
tu
r
es,
is
th
e
n
u
m
b
er
o
f
test
s
am
p
les,
is
th
e
tr
u
e
la
b
e
l
o
f
t
h
e
i
-
th
test
s
am
p
le,
ˆ
is
th
e
p
r
ed
icted
lab
el
o
f
th
e
i
-
th
test
s
am
p
le,
a
n
d
(
,
ˆ
)
is
th
e
Kr
o
n
ec
k
er
d
elta
f
u
n
ctio
n
,
wh
ich
is
1
if
=
ˆ
an
d
0
o
th
e
r
wis
e.
Fo
r
th
e
Alzh
eim
er
’
s
d
is
e
ase
d
ataset
,
DE
,
GW
O,
a
n
d
POA
wer
e
ap
p
lied
to
th
e
tr
ain
in
g
s
et.
4
.
1
.
3
.
C
l
a
s
s
i
f
i
c
a
t
i
o
n
T
h
e
s
u
b
s
eq
u
en
t
s
tep
is
to
co
m
m
en
ce
th
e
class
if
icatio
n
p
r
o
ce
s
s
,
wh
ich
en
tails
tr
ain
in
g
th
e
s
et
o
f
f
ea
tu
r
es
u
s
in
g
a
v
ar
iet
y
o
f
c
lass
if
ier
s
.
T
h
is
i
s
b
ec
au
s
e
we
h
av
e
s
u
cc
ess
f
u
lly
id
en
tifie
d
th
e
m
o
s
t
s
u
itab
le
f
ea
tu
r
es
in
th
e
Alzh
eim
er
’
s
d
is
ea
s
e
d
ataset
an
d
h
av
e
p
u
r
i
f
ied
all
p
o
ten
tially
n
o
is
y
d
ata
as
a
r
esu
lt
o
f
th
e
f
ea
tu
r
e
s
elec
tio
n
p
r
o
ce
s
s
in
th
e
p
r
ev
io
u
s
s
tep
.
W
e
co
n
d
u
cte
d
ex
p
e
r
im
en
ts
with
a
v
ar
iety
o
f
p
ar
am
ete
r
s
in
DE
,
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
.
3
,
J
u
n
e
20
25
:
3
3
8
2
-
3
3
9
5
3388
GW
O,
an
d
POA,
s
u
ch
as
th
e
n
u
m
b
er
o
f
iter
atio
n
s
(
th
e
r
e
p
l
icatio
n
o
f
a
p
r
o
ce
s
s
to
p
r
o
d
u
c
e
an
o
u
tco
m
e
)
an
d
p
o
p
u
latio
n
s
ize
(
th
e
ar
b
itra
r
y
co
n
s
tr
u
ctio
n
o
f
p
o
p
u
latio
n
s
to
d
eter
m
in
e
th
e
o
p
tim
al
p
o
p
u
l
atio
n
s
ize
b
ased
o
n
th
e
p
r
o
b
lem
)
.
W
e
co
n
d
u
cted
ex
p
er
im
en
ts
o
n
p
o
p
u
latio
n
s
o
f
3
0
an
d
6
0
an
d
iter
atio
n
s
o
f
5
,
1
5
,
an
d
3
0
to
d
eter
m
in
e
th
e
m
o
s
t e
f
f
ec
tiv
e
c
o
n
f
ig
u
r
atio
n
f
o
r
f
ea
t
u
r
e
s
elec
tio
n
an
d
class
if
icatio
n
.
4
.
1
.
4
.
E
v
a
l
u
a
t
i
o
n
Fo
u
r
p
er
f
o
r
m
a
n
ce
m
etr
ics
w
er
e
em
p
lo
y
e
d
to
ass
ess
th
e
p
ar
r
o
t
o
p
tim
izatio
n
(
PO)
,
d
if
f
er
en
tial
ev
o
lu
tio
n
(
DE
)
,
an
d
g
r
ey
w
o
lf
o
p
tim
izer
(
GW
O)
alg
o
r
it
h
m
s
:
p
r
ec
i
s
io
n
,
r
ec
all,
F
-
s
co
r
e,
an
d
ac
c
u
r
ac
y
.
Acc
u
r
ac
y
is
a
s
tatis
tical
b
ias
m
etr
ic
th
at
q
u
an
tifie
s
th
e
p
e
r
ce
n
tag
e
o
f
a
test
’
s
s
u
cc
ess
r
ate.
L
o
w
ac
cu
r
ac
y
v
alu
es
s
u
g
g
est
a
d
is
cr
ep
an
cy
b
etwe
en
th
e
ac
tu
al
an
d
r
esu
lt
s
ets.
T
ab
le
1
illu
s
tr
ate
s
th
e
co
n
f
u
s
io
n
m
atr
ix
f
o
r
class
if
icatio
n
,
wh
ich
r
ep
r
esen
ts
th
e
class
if
icatio
n
o
f
th
e
p
o
ten
tial
o
u
tco
m
e
o
f
r
ec
o
m
m
e
n
d
in
g
a
n
item
to
a
u
s
er
.
Acc
u
r
ac
y
e
m
p
lo
y
s
f
o
u
r
t
est m
ea
s
u
r
es.
T
ab
le
1
.
C
o
n
f
u
s
io
n
m
atr
ix
f
o
r
class
if
icatio
n
R
e
c
o
mm
e
n
d
e
d
N
o
t
r
e
c
o
mm
e
n
d
e
d
P
r
e
f
e
r
r
e
d
Tr
u
e
p
o
si
t
i
v
e
(
TP)
F
a
l
se
n
e
g
a
t
i
v
e
(
F
N
)
N
o
t
p
r
e
f
e
r
r
e
d
F
a
l
se
p
o
si
t
i
v
e
(
F
P
)
Tr
u
e
n
e
g
a
t
i
v
e
(
TN
)
4
.
1
.
5
.
T
h
e
F
-
s
c
o
r
e
T
h
e
F
-
s
co
r
e,
a
s
o
litar
y
d
ig
it,
co
n
cisely
ev
alu
ates
a
s
y
s
tem
o
r
m
o
d
el’
s
ab
ilit
y
to
g
en
e
r
ate
p
r
ec
is
e
o
p
tim
is
tic
p
r
ed
ictio
n
s
an
d
i
d
e
n
tify
ev
er
y
p
o
s
itiv
e
in
s
tan
ce
.
T
h
e
alg
o
r
ith
m
in
teg
r
ates two
f
u
n
d
am
e
n
tal
m
etr
ics,
n
am
ely
r
ec
all
(
th
e
ca
p
ac
ity
to
id
en
tif
y
all
p
o
s
itiv
e
ca
s
es)
an
d
p
r
ec
is
io
n
(
th
e
ac
c
u
r
ac
y
o
f
o
p
tim
is
tic
p
r
ed
ictio
n
s
)
.
B
y
ac
h
iev
i
n
g
a
n
eq
u
ilib
r
iu
m
b
etwe
en
th
ese
t
wo
v
ar
iab
les,
t
h
e
F1
-
s
co
r
e
o
f
f
er
s
a
u
n
if
ied
m
etr
ic
f
o
r
e
v
alu
atin
g
p
er
f
o
r
m
a
n
ce
.
E
lev
ated
v
alu
es
o
n
a
s
ca
le
o
f
0
to
1
in
d
icate
s
u
p
er
io
r
p
er
f
o
r
m
an
ce
.
I
t
s
er
v
es
as
a
p
r
ac
tical
in
s
tr
u
m
en
t
f
o
r
ass
ess
in
g
th
e
ef
f
icac
y
o
f
class
if
icatio
n
s
y
s
tem
s
.
Def
in
ed
b
y
th
is
f
o
r
m
u
la
is
th
e
F
-
s
co
r
e:
1
=
2
·
P
r
e
c
i
s
i
on
·
R
e
c
a
l
l
+
(
5
)
4
.
1
.
6
.
P
r
e
c
i
s
i
o
n
I
n
d
icate
s
th
e
p
r
o
p
o
r
tio
n
o
f
p
o
s
itiv
e
ca
s
es
p
r
ed
icted
b
y
th
e
m
o
d
el
t
h
at
tu
r
n
e
d
o
u
t
t
o
b
e
tr
u
e.
I
t
q
u
a
n
tifie
s
th
e
p
r
ec
is
io
n
with
wh
ich
th
e
m
o
d
el
g
e
n
er
ates
af
f
ir
m
ativ
e
p
r
ed
ictio
n
s
.
T
h
e
f
o
r
m
u
la
f
o
r
p
r
ec
is
io
n
is
:
=
+
(
6
)
4
.
1
.
7
.
R
e
c
a
l
l
As
with
s
en
s
itiv
ity
,
r
ec
all
q
u
an
tifie
s
th
e
ac
cu
r
ac
y
with
wh
ich
th
e
m
o
d
el
d
etec
ts
tr
u
e
p
o
s
itiv
es.
I
t
p
r
o
v
id
es
th
e
n
u
m
b
er
o
f
ac
cu
r
ate
o
p
tim
is
tic
p
r
ed
ictio
n
s
th
e
m
o
d
el
m
ak
es
r
elativ
e
to
th
e
to
tal
n
u
m
b
er
o
f
p
o
s
itiv
e
ca
s
es.
T
h
e
f
o
r
m
u
la
f
o
r
r
ec
all
is
:
=
+
(
7
)
5.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
T
h
e
Alzh
eim
er
’
s
d
is
ea
s
e
d
ataset
was
em
p
lo
y
ed
to
ev
alu
ate
th
e
DE
,
GW
O,
an
d
POA
a
lg
o
r
ith
m
s
.
I
n
th
is
co
n
tex
t,
th
ese
m
etah
eu
r
is
tic
alg
o
r
ith
m
s
h
av
e
n
o
t
b
ee
n
ex
ten
s
iv
ely
co
m
p
ar
e
d
f
o
r
f
e
atu
r
e
s
elec
tio
n
,
to
th
e
b
est
o
f
o
u
r
k
n
o
wled
g
e
.
Fo
u
r
class
if
ier
s
wer
e
em
p
lo
y
ed
to
e
v
alu
ate
th
e
f
ea
tu
r
e
s
elec
tio
n
alg
o
r
ith
m
s
:
g
r
ad
ien
t
b
o
o
s
tin
g
,
XGBo
o
s
t
,
R
F,
an
d
DT
.
I
n
o
r
d
er
to
g
u
ar
an
tee
th
e
im
p
ar
tiality
o
f
th
e
o
u
tco
m
es,
th
e
alg
o
r
ith
m
s
wer
e
tr
ain
ed
with
id
en
tical
m
eth
o
d
o
lo
g
ies.
T
h
e
s
cik
it
-
lear
n
lib
r
ar
y
in
Py
th
o
n
,
wh
ich
in
clu
d
es
b
u
ilt
-
in
lib
r
ar
ies
f
o
r
f
ea
tu
r
e
s
elec
tio
n
alg
o
r
ith
m
s
,
h
a
d
th
e
class
if
ier
s
av
ailab
le.
T
ab
les
2
to
4
lis
t
n
u
m
er
o
u
s
in
p
u
t
p
ar
am
eter
s
t
h
at
ar
e
em
p
lo
y
ed
in
th
e
ex
ec
u
tio
n
o
f
t
h
ese
f
ea
tu
r
e
s
elec
tio
n
alg
o
r
ith
m
s
.
T
h
e
p
ar
am
ete
r
s
’
v
alu
es
wer
e
d
eter
m
in
ed
th
r
o
u
g
h
e
x
p
er
im
e
n
tatio
n
a
n
d
ar
e
c
o
n
tin
g
en
t
u
p
o
n
th
e
u
n
iq
u
e
ch
ar
ac
ter
is
tics
o
f
ea
c
h
alg
o
r
ith
m
.
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
E
n
h
a
n
cin
g
A
lz
h
eime
r
’
s
d
is
ea
s
e
d
ia
g
n
o
s
is
th
r
o
u
g
h
meta
h
eu
r
is
tic
fea
tu
r
e
…
(
A
r
a
r
A
l
-
Ta
w
i
l
)
3389
T
ab
le
2
.
I
n
p
u
t
p
ar
am
eter
s
f
o
r
p
ar
r
o
t
o
p
tim
izer
P
a
r
a
me
t
e
r
D
e
scri
p
t
i
o
n
lb
Lo
w
e
r
b
o
u
n
d
f
o
r
t
h
e
sea
r
c
h
sp
a
c
e
ub
U
p
p
e
r
b
o
u
n
d
f
o
r
t
h
e
se
a
r
c
h
s
p
a
c
e
d
i
m
D
i
me
n
si
o
n
a
l
i
t
y
o
f
t
h
e
se
a
r
c
h
s
p
a
c
e
N
P
o
p
u
l
a
t
i
o
n
s
i
z
e
M
a
x
_
i
t
e
r
M
a
x
i
m
u
m
n
u
m
b
e
r
o
f
i
t
e
r
a
t
i
o
n
s
T
ab
le
3
.
I
n
p
u
t
p
ar
am
eter
s
f
o
r
d
if
f
er
en
tial e
v
o
lu
tio
n
(
DE
)
P
a
r
a
me
t
e
r
D
e
scri
p
t
i
o
n
lb
Lo
w
e
r
b
o
u
n
d
f
o
r
t
h
e
sea
r
c
h
sp
a
c
e
ub
U
p
p
e
r
b
o
u
n
d
f
o
r
t
h
e
se
a
r
c
h
s
p
a
c
e
d
i
m
D
i
me
n
si
o
n
a
l
i
t
y
o
f
t
h
e
se
a
r
c
h
s
p
a
c
e
P
o
p
S
i
z
e
P
o
p
u
l
a
t
i
o
n
s
i
z
e
i
t
e
r
s
N
u
mb
e
r
o
f
i
t
e
r
a
t
i
o
n
s
mu
t
a
t
i
o
n
f
a
c
t
o
r
M
u
t
a
t
i
o
n
f
a
c
t
o
r
(
c
o
n
t
r
o
l
s
t
h
e
a
mp
l
i
f
i
c
a
t
i
o
n
o
f
t
h
e
d
i
f
f
e
r
e
n
t
i
a
l
v
a
r
i
a
t
i
o
n
)
.
D
e
f
a
u
l
t
i
s
0
.
5
.
c
r
o
ss
o
v
e
r
r
a
t
i
o
c
r
o
ss
o
v
e
r
r
a
t
i
o
C
r
o
ss
o
v
e
r
r
a
t
i
o
(
c
o
n
t
r
o
l
s
t
h
e
p
r
o
b
a
b
i
l
i
t
y
o
f
c
r
o
ss
o
v
e
r
)
.
D
e
f
a
u
l
t
i
s
0
.
7
T
ab
le
4
.
I
n
p
u
t
p
ar
am
eter
s
f
o
r
g
r
ey
wo
lf
o
p
tim
izer
(
GW
O)
P
a
r
a
me
t
e
r
D
e
scri
p
t
i
o
n
lb
Lo
w
e
r
b
o
u
n
d
f
o
r
t
h
e
sea
r
c
h
sp
a
c
e
ub
U
p
p
e
r
b
o
u
n
d
f
o
r
t
h
e
se
a
r
c
h
s
p
a
c
e
d
i
m
D
i
me
n
si
o
n
a
l
i
t
y
o
f
t
h
e
se
a
r
c
h
s
p
a
c
e
S
e
a
r
c
h
A
g
e
n
t
s
n
o
N
u
mb
e
r
o
f
se
a
r
c
h
a
g
e
n
t
s
M
a
x
i
t
e
r
M
a
x
i
m
u
m
n
u
m
b
e
r
o
f
i
t
e
r
a
t
i
o
n
s
6.
DIS
CU
SS
I
O
N
T
wo
tab
les,
T
ab
les
5
an
d
6
,
p
r
esen
t
th
e
DE
alg
o
r
ith
m
’
s
ev
alu
atio
n
r
esu
lts
o
n
th
e
Alzh
eim
er
’
s
d
is
ea
s
e
d
ataset
.
T
h
e
ta
b
les
ev
alu
ate
p
o
p
u
latio
n
s
izes
o
f
3
0
an
d
6
0
ac
r
o
s
s
v
a
r
io
u
s
iter
atio
n
s
(
5
,
1
5
,
a
n
d
3
0
)
.
T
h
e
g
r
a
d
ien
t
b
o
o
s
tin
g
class
if
ier
co
n
s
is
ten
tly
ac
h
iev
e
d
th
e
h
ig
h
est
ac
cu
r
ac
y
f
o
r
a
p
o
p
u
latio
n
s
ize
o
f
3
0
,
a
s
d
ep
icted
in
T
ab
le
5
,
with
a
m
ax
im
u
m
o
f
0
.
9
3
5
8
6
6
at
f
i
v
e
it
er
atio
n
s
.
XGBo
o
s
t
also
d
em
o
n
s
tr
ated
s
atis
f
ac
to
r
y
p
er
f
o
r
m
an
ce
,
attain
in
g
its
m
ax
im
u
m
ac
cu
r
ac
y
o
f
0
.
9
3
5
5
5
6
af
ter
3
0
iter
atio
n
s
.
I
n
c
o
n
tr
a
s
t,
th
e
DT
an
d
RF
class
if
ier
s
ex
h
ib
ited
lo
wer
ac
cu
r
ac
y
b
u
t
wer
e
s
till
d
ep
en
d
ab
l
e.
T
h
e
DT
ac
h
iev
ed
0
.
8
9
7
4
6
8
af
ter
3
0
iter
atio
n
s
,
wh
ile
th
e
RF
r
ea
ch
ed
0
.
9
2
3
2
0
4
af
ter
f
iv
e
iter
atio
n
s
.
T
ab
le
5
.
DE
alg
o
r
ith
m
u
s
in
g
3
0
p
o
p
u
latio
n
I
t
e
r
a
t
i
o
n
s
C
l
a
s
si
f
i
e
r
A
c
c
u
r
a
c
y
P
r
e
c
i
s
i
o
n
R
e
c
a
l
l
F1
-
s
c
o
r
e
5
D
e
c
i
s
i
o
n
t
r
e
e
0
.
8
9
0
.
8
8
0
.
8
4
0
.
8
5
G
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
0
.
9
3
0
.
9
4
0
.
8
8
0
.
9
R
a
n
d
o
m
f
o
r
e
s
t
0
.
9
2
0
.
9
5
0
.
8
4
0
.
8
9
X
G
B
o
o
st
0
.
9
4
0
.
9
5
0
.
8
8
0
.
9
1
15
D
e
c
i
s
i
o
n
t
r
e
e
0
.
9
0
.
8
8
0
.
8
4
0
.
8
5
G
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
0
.
9
3
0
.
9
4
0
.
8
8
0
.
9
1
R
a
n
d
o
m
f
o
r
e
s
t
0
.
9
2
0
.
9
5
0
.
8
4
0
.
8
8
X
G
B
o
o
st
0
.
9
4
0
.
9
4
0
.
8
8
0
.
9
30
D
e
c
i
s
i
o
n
t
r
e
e
0
.
9
0
.
8
9
0
.
8
4
0
.
8
7
G
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
0
.
9
3
0
.
9
3
0
.
8
8
0
.
9
1
R
a
n
d
o
m
f
o
r
e
s
t
0
.
9
2
0
.
9
5
0
.
8
4
0
.
8
9
X
G
B
o
o
st
0
.
9
4
0
.
9
5
0
.
8
8
0
.
9
1
T
h
e
g
r
a
d
ien
t
b
o
o
s
tin
g
class
if
ier
ag
ain
attain
ed
th
e
h
ig
h
est
ac
cu
r
ac
y
o
f
0
.
9
3
6
6
9
3
at
3
0
iter
atio
n
s
wh
en
th
e
p
o
p
u
latio
n
s
ize
was
in
cr
ea
s
ed
to
6
0
in
T
ab
le
6
.
X
GB
o
o
s
t
clo
s
ely
m
atch
ed
th
is
r
esu
lt.
T
h
e
DT
an
d
RF
cla
s
s
if
ier
s
y
ield
ed
r
eliab
le
o
u
tco
m
es,
with
th
e
DT
attain
in
g
an
ac
cu
r
ac
y
o
f
0
.
8
9
8
0
3
6
af
ter
3
0
iter
atio
n
s
an
d
th
e
RF
r
ea
ch
in
g
0
.
9
2
3
7
2
1
af
ter
3
0
iter
atio
n
s
.
T
h
ese
f
in
d
in
g
s
in
d
icate
th
at
th
e
g
r
ad
ien
t
b
o
o
s
tin
g
class
if
ier
co
n
s
is
ten
tly
o
u
tp
er
f
o
r
m
ed
o
th
er
s
in
ac
cu
r
ac
y
,
p
ar
ticu
lar
l
y
at
h
ig
h
er
iter
atio
n
s
an
d
lar
g
er
p
o
p
u
latio
n
s
izes.
Ad
d
itio
n
ally
,
XGBo
o
s
t
ex
h
ib
i
ted
h
ig
h
ac
c
u
r
ac
y
an
d
r
o
b
u
s
tn
ess
ac
r
o
s
s
v
ar
io
u
s
c
o
n
f
i
g
u
r
ati
o
n
s
.
W
h
en
co
m
b
in
ed
with
r
o
b
u
s
t
class
if
i
er
s
s
u
ch
as
g
r
ad
ien
t
b
o
o
s
tin
g
an
d
XGBo
o
s
t,
th
e
DE
alg
o
r
ith
m
d
em
o
n
s
tr
ated
ef
f
icac
y
in
im
p
r
o
v
in
g
p
r
ed
icti
v
e
ac
cu
r
ac
y
f
o
r
Alzh
eim
er
’
s
d
is
ea
s
e
d
iag
n
o
s
is
.
T
h
e
g
r
ad
ien
t
b
o
o
s
tin
g
class
if
ier
ac
h
iev
ed
th
e
h
ig
h
est ac
cu
r
ac
y
o
f
0
.
9
3
6
6
9
3
at
3
0
iter
atio
n
s
an
d
a
p
o
p
u
latio
n
s
ize
o
f
6
0
.
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
.
3
,
J
u
n
e
20
25
:
3
3
8
2
-
3
3
9
5
3390
T
ab
le
6
.
DE
alg
o
r
ith
m
u
s
in
g
6
0
p
o
p
u
latio
n
I
t
e
r
a
t
i
o
n
s
C
l
a
s
s
i
f
i
e
r
A
c
c
u
r
a
c
y
P
r
e
c
i
s
i
o
n
R
e
c
a
l
l
F1
-
s
c
o
r
e
5
D
e
c
i
s
i
o
n
t
r
e
e
0
.
8
9
0
.
8
7
0
.
8
4
0
.
8
6
G
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
0
.
9
3
0
.
9
4
0
.
8
8
0
.
9
1
R
a
n
d
o
m
f
o
r
e
s
t
0
.
9
1
0
.
9
5
0
.
8
3
0
.
8
8
X
G
B
o
o
s
t
0
.
9
3
0
.
9
5
0
.
8
8
0
.
9
1
15
D
e
c
i
s
i
o
n
t
r
e
e
0
.
9
0
.
8
8
0
.
8
4
0
.
8
6
G
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
0
.
9
3
0
.
9
4
0
.
8
8
0
.
9
R
a
n
d
o
m
f
o
r
e
s
t
0
.
9
2
0
.
9
5
0
.
8
4
0
.
8
9
X
G
B
o
o
s
t
0
.
9
4
0
.
9
5
0
.
8
8
0
.
9
1
30
D
e
c
i
s
i
o
n
t
r
e
e
0
.
9
0
.
8
8
0
.
8
4
0
.
8
6
G
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
0
.
9
3
0
.
9
4
0
.
8
8
0
.
9
R
a
n
d
o
m
f
o
r
e
s
t
0
.
9
2
0
.
9
5
0
.
8
4
0
.
8
9
X
G
B
o
o
s
t
0
.
9
4
0
.
9
5
0
.
8
8
0
.
9
1
T
ab
les
7
an
d
8
illu
s
tr
ate
th
e
ac
cu
r
ac
y
,
p
r
ec
is
io
n
,
r
ec
all,
an
d
F1
-
s
co
r
e
ev
alu
atio
n
r
esu
lts
o
f
th
e
GW
O
alg
o
r
ith
m
o
n
th
e
Alzh
eim
e
r
’
s
d
is
ea
s
e
d
ataset
.
I
n
a
s
er
ies
o
f
iter
atio
n
s
(
5
,
1
5
,
an
d
3
0
)
,
t
h
e
GW
O
alg
o
r
ith
m
was
ass
e
s
s
ed
with
p
o
p
u
latio
n
s
izes
o
f
3
0
an
d
6
0
.
At
3
0
iter
a
tio
n
s
,
th
e
g
r
a
d
ien
t
b
o
o
s
tin
g
cl
ass
if
ier
co
n
s
is
ten
tly
d
em
o
n
s
tr
ated
s
u
p
er
i
o
r
ac
cu
r
a
cy
f
o
r
th
e
GW
O
alg
o
r
ith
m
with
a
p
o
p
u
latio
n
s
ize
o
f
3
0
in
T
ab
le
7
,
with
a
m
ax
im
al
ac
cu
r
ac
y
o
f
0
.
9
3
5
1
4
2
.
XGBo
o
s
t
also
ex
h
ib
ited
s
atis
f
ac
to
r
y
p
er
f
o
r
m
a
n
ce
,
ac
h
iev
in
g
th
e
h
ig
h
est
ac
cu
r
ac
y
o
f
0
.
9
3
5
0
3
9
af
ter
f
iv
e
iter
atio
n
s
.
C
o
n
v
er
s
ely
,
th
e
DT
an
d
RF
cla
s
s
if
ier
s
m
ain
tain
ed
th
eir
co
n
s
is
ten
cy
wh
ile
ex
h
ib
itin
g
a
m
ar
g
in
ally
lo
we
r
lev
el
o
f
ac
cu
r
ac
y
.
At
3
0
iter
atio
n
s
,
th
e
DT
o
b
tain
ed
its
h
ig
h
est
ac
cu
r
ac
y
o
f
0
.
8
9
6
0
2
1
,
wh
ile
th
e
RF
ac
h
iev
ed
0
.
9
2
3
9
7
9
d
u
r
in
g
t
h
e
s
am
e
p
er
io
d
.
T
h
e
ef
f
icac
y
o
f
th
e
GW
O
alg
o
r
ith
m
with
a
p
o
p
u
l
atio
n
s
ize
o
f
3
0
is
d
em
o
n
s
tr
ated
in
T
a
b
le
7
.
T
h
is
tab
le
co
m
p
ar
es
th
e
ef
f
icac
y
o
f
th
e
alg
o
r
ith
m
at
d
if
f
er
en
t
iter
atio
n
s
an
d
class
if
ier
s
.
T
h
e
g
r
a
d
ien
t
b
o
o
s
tin
g
class
if
ier
attain
ed
a
h
ig
h
ac
cu
r
ac
y
o
f
0
.
9
3
4
0
5
7
at
3
0
iter
atio
n
s
wh
en
a
p
o
p
u
latio
n
s
ize
o
f
6
0
was
u
s
ed
,
clo
s
ely
f
o
llo
win
g
XGBo
o
s
t’
s
0
.
9
3
5
6
5
9
ac
cu
r
ac
y
at
1
5
iter
atio
n
s
T
a
b
le
8
.
B
o
th
th
e
DT
an
d
RF
class
if
ier
s
p
r
o
d
u
ce
d
f
in
d
in
g
s
th
at
wer
e
f
air
ly
r
em
ar
k
ab
le
in
th
eir
r
esp
ec
tiv
e
ca
teg
o
r
ies.
W
h
en
co
m
p
ar
ed
to
th
e
RF
,
wh
ich
wa
s
ab
le
to
r
e
ac
h
an
ac
cu
r
ac
y
o
f
0
.
9
2
4
9
6
1
i
n
th
e
s
am
e
len
g
th
o
f
tim
e,
th
e
DT
was
ab
le
to
attain
an
ac
cu
r
ac
y
o
f
0
.
8
9
7
5
1
9
o
v
er
th
e
co
u
r
s
e
o
f
th
ir
ty
iter
atio
n
s
.
Acc
o
r
d
in
g
to
th
e
d
ata,
it
wo
u
ld
ap
p
ea
r
th
a
t
th
e
Gr
ad
ien
t
B
o
o
s
tin
g
an
d
XGBo
o
s
t
cla
s
s
if
ier
s
co
n
s
is
ten
tly
b
ea
t
o
th
er
ap
p
r
o
ac
h
es
in
ter
m
s
o
f
ac
cu
r
ac
y
.
T
h
is
was
esp
ec
ially
tr
u
e
wh
en
th
e
n
u
m
b
e
r
o
f
iter
atio
n
s
was
r
aised
an
d
th
e
p
o
p
u
latio
n
s
ize
was
in
cr
ea
s
e
d
.
T
h
e
GW
O
alg
o
r
ith
m
’
s
ab
i
lity
to
im
p
r
o
v
e
t
h
e
ex
p
ec
ted
ac
c
u
r
ac
y
o
f
Alzh
ei
m
er
’
s
d
is
ea
s
e
d
iag
n
o
s
is
is
n
o
tab
ly
ap
p
ar
en
t
wh
en
it
is
ap
p
lied
in
c
o
n
ju
n
ctio
n
with
th
ese
ef
f
ec
tiv
e
class
if
ier
s
.
T
h
is
is
b
ec
a
u
s
e
th
e
GW
O
alg
o
r
ith
m
is
ab
le
t
o
im
p
r
o
v
e
p
r
ed
icted
ac
cu
r
ac
y
.
T
h
e
XGBo
o
s
t
class
if
ier
attain
ed
th
e
h
ig
h
est
ac
cu
r
ac
y
f
o
r
t
h
e
GW
O
m
eth
o
d
with
a
p
o
p
u
la
tio
n
s
ize
o
f
6
0
an
d
1
5
iter
atio
n
s
(
ac
cu
r
ac
y
=
0
.
9
3
5
6
5
9
)
.
T
a
b
le
8
d
e
m
o
n
s
tr
ate
s
th
e
ef
f
ec
tiv
en
ess
o
f
th
e
G
W
O
m
eth
o
d
with
a
p
o
p
u
latio
n
s
ize
o
f
6
0
ac
r
o
s
s
v
ar
io
u
s
class
if
ier
s
an
d
iter
atio
n
s
.
T
ab
les
9
an
d
1
0
illu
s
tr
ate
th
e
ev
alu
atio
n
r
esu
lts
o
f
th
e
PO
alg
o
r
ith
m
o
n
th
e
Alzh
eim
er
’
s
d
is
ea
s
e
d
ataset
in
ter
m
s
o
f
p
r
ec
is
io
n
,
r
ec
all,
ac
cu
r
ac
y
,
an
d
F1
-
s
co
r
e.
T
h
e
PO a
lg
o
r
ith
m
was e
v
alu
ated
with
p
o
p
u
latio
n
s
izes
o
f
3
0
an
d
6
0
in
v
ar
io
u
s
iter
atio
n
s
(
5
,
1
5
,
an
d
3
0
)
.
T
h
e
g
r
ad
ien
t
b
o
o
s
tin
g
a
n
d
X
GB
o
o
s
t
class
if
ier
s
co
n
s
is
ten
tly
ac
h
iev
ed
h
ig
h
er
ac
cu
r
ac
y
f
o
r
th
e
PO
alg
o
r
ith
m
with
a
p
o
p
u
latio
n
s
ize
o
f
3
0
in
T
ab
le
9
.
At
3
0
iter
atio
n
s
,
g
r
ad
ien
t
b
o
o
s
tin
g
attain
ed
a
m
ax
im
al
ac
cu
r
ac
y
o
f
0
.
9
2
8
0
6
2
,
wh
ile
XGBo
o
s
t
a
ch
iev
ed
0
.
9
2
9
1
9
9
.
On
th
e
o
th
er
h
a
n
d
,
th
e
DT
a
n
d
RF
class
if
ier
s
d
em
o
n
s
tr
ated
s
lig
h
tly
lo
wer
ac
cu
r
ac
y
b
u
t
c
o
n
tin
u
ed
to
o
p
e
r
ate
co
n
s
is
ten
tly
.
T
h
e
DT
ac
h
iev
ed
its
h
ig
h
est
ac
cu
r
ac
y
o
f
0
.
8
8
3
6
1
8
at
3
0
iter
atio
n
s
,
wh
il
e
th
e
R
F
ac
h
iev
ed
0
.
9
1
7
3
1
3
at
1
5
.
T
h
e
p
er
f
o
r
m
a
n
ce
o
f
th
e
PO
alg
o
r
ith
m
with
a
p
o
p
u
latio
n
s
ize
o
f
3
0
ac
r
o
s
s
v
ar
io
u
s
class
if
ier
s
an
d
iter
atio
n
s
is
illu
s
tr
ated
in
T
ab
le
9
.
T
ab
le
7
.
GW
O
alg
o
r
ith
m
u
s
in
g
3
0
Po
p
u
latio
n
I
t
e
r
a
t
i
o
n
s
C
l
a
s
s
i
f
i
e
r
A
c
c
u
r
a
c
y
P
r
e
c
i
s
i
o
n
R
e
c
a
l
l
F1
-
s
c
o
r
e
5
D
e
c
i
s
i
o
n
t
r
e
e
0
.
8
9
0
.
8
7
0
.
8
4
0
.
8
6
G
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
0
.
9
3
0
.
9
4
0
.
8
8
0
.
9
1
R
a
n
d
o
m
f
o
r
e
s
t
0
.
9
2
0
.
9
5
0
.
8
4
0
.
8
9
X
G
B
o
o
s
t
0
.
9
4
0
.
9
4
0
.
8
8
0
.
9
1
15
D
e
c
i
s
i
o
n
t
r
e
e
0
.
8
9
0
.
8
7
0
.
8
4
0
.
8
6
G
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
0
.
9
3
0
.
9
4
0
.
8
8
0
.
9
1
R
a
n
d
o
m
f
o
r
e
s
t
0
.
9
2
0
.
9
5
0
.
8
4
0
.
8
9
X
G
B
o
o
s
t
0
.
9
3
0
.
9
4
0
.
8
8
0
.
9
1
30
D
e
c
i
s
i
o
n
t
r
e
e
0
.
9
0
.
8
8
0
.
8
4
0
.
8
6
G
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
0
.
9
3
0
.
9
4
0
.
8
8
0
.
9
1
R
a
n
d
o
m
f
o
r
e
s
t
0
.
9
2
0
.
9
5
0
.
8
4
0
.
8
9
X
G
B
o
o
s
t
0
.
9
4
0
.
9
5
0
.
8
8
0
.
9
1
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
E
n
h
a
n
cin
g
A
lz
h
eime
r
’
s
d
is
ea
s
e
d
ia
g
n
o
s
is
th
r
o
u
g
h
meta
h
eu
r
is
tic
fea
tu
r
e
…
(
A
r
a
r
A
l
-
Ta
w
i
l
)
3391
T
ab
le
8
.
GW
O
alg
o
r
ith
m
u
s
in
g
6
0
p
o
p
u
latio
n
I
t
e
r
a
t
i
o
n
s
C
l
a
s
s
i
f
i
e
r
A
c
c
u
r
a
c
y
P
r
e
c
i
s
i
o
n
R
e
c
a
l
l
F1
-
s
c
o
r
e
5
D
e
c
i
s
i
o
n
t
r
e
e
0
.
8
9
0
.
8
7
0
.
8
4
0
.
8
6
G
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
0
.
9
3
0
.
9
4
0
.
8
8
0
.
9
1
R
a
n
d
o
m
f
o
r
e
s
t
0
.
9
2
0
.
9
5
0
.
8
4
0
.
8
9
X
G
B
o
o
s
t
0
.
9
3
0
.
9
4
0
.
8
8
0
.
9
1
15
D
e
c
i
s
i
o
n
t
r
e
e
0
.
9
0
.
8
8
0
.
8
4
0
.
8
6
G
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
0
.
9
3
0
.
9
4
0
.
8
8
0
.
9
1
R
a
n
d
o
m
f
o
r
e
s
t
0
.
9
2
0
.
9
5
0
.
8
4
0
.
8
9
X
G
B
o
o
s
t
0
.
9
4
0
.
9
5
0
.
8
8
0
.
9
1
30
D
e
c
i
s
i
o
n
t
r
e
e
0
.
9
0
.
8
8
0
.
8
4
0
.
8
6
G
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
0
.
9
3
0
.
9
4
0
.
8
8
0
.
9
1
R
a
n
d
o
m
f
o
r
e
s
t
0
.
9
2
0
.
9
5
0
.
8
5
0
.
9
X
G
B
o
o
s
t
0
.
9
3
0
.
9
4
0
.
8
8
0
.
9
1
T
ab
le
9
.
PO a
lg
o
r
ith
m
u
s
in
g
3
0
p
o
p
u
latio
n
I
t
e
r
a
t
i
o
n
s
C
l
a
s
s
i
f
i
e
r
A
c
c
u
r
a
c
y
P
r
e
c
i
s
i
o
n
R
e
c
a
l
l
F1
-
s
c
o
r
e
5
D
e
c
i
s
i
o
n
t
r
e
e
0
.
8
7
0
.
8
4
0
.
8
2
0
.
8
3
G
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
0
.
9
2
0
.
9
4
0
.
8
5
0
.
8
9
R
a
n
d
o
m
f
o
r
e
s
t
0
.
9
1
0
.
9
5
0
.
8
1
0
.
8
8
X
G
B
o
o
s
t
0
.
9
2
0
.
9
4
0
.
8
5
0
.
8
9
15
D
e
c
i
s
i
o
n
t
r
e
e
0
.
8
8
0
.
8
6
0
.
8
3
0
.
8
4
G
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
0
.
9
3
0
.
9
4
0
.
8
7
0
.
9
R
a
n
d
o
m
f
o
r
e
s
t
0
.
9
2
0
.
9
5
0
.
8
3
0
.
8
8
X
G
B
o
o
s
t
0
.
9
3
0
.
9
4
0
.
8
6
0
.
9
30
D
e
c
i
s
i
o
n
t
r
e
e
0
.
8
8
0
.
8
6
0
.
8
3
0
.
8
4
G
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
0
.
9
3
0
.
9
4
0
.
8
7
0
.
9
R
a
n
d
o
m
f
o
r
e
s
t
0
.
9
2
0
.
9
5
0
.
8
2
0
.
8
8
X
G
B
o
o
s
t
0
.
9
3
0
.
9
4
0
.
8
7
0
.
9
T
ab
le
1
0
.
PO a
lg
o
r
ith
m
u
s
in
g
6
0
p
o
p
u
latio
n
I
t
e
r
a
t
i
o
n
s
C
l
a
s
s
i
f
i
e
r
A
c
c
u
r
a
c
y
P
r
e
c
i
s
i
o
n
R
e
c
a
l
l
F1
-
s
c
o
r
e
5
D
e
c
i
s
i
o
n
t
r
e
e
0
.
8
9
0
.
8
7
0
.
8
4
0
.
8
5
G
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
0
.
9
3
0
.
9
4
0
.
8
8
0
.
9
1
R
a
n
d
o
m
f
o
r
e
s
t
0
.
9
2
0
.
9
5
0
.
8
2
0
.
8
8
X
G
B
o
o
s
t
0
.
9
4
0
.
9
5
0
.
8
8
0
.
9
1
15
D
e
c
i
s
i
o
n
t
r
e
e
0
.
8
9
0
.
8
7
0
.
8
4
0
.
8
5
G
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
0
.
9
3
0
.
9
4
0
.
8
8
0
.
9
1
R
a
n
d
o
m
f
o
r
e
s
t
0
.
9
2
0
.
9
5
0
.
8
2
0
.
8
8
X
G
B
o
o
s
t
0
.
9
4
0
.
9
5
0
.
8
8
0
.
9
1
30
D
e
c
i
s
i
o
n
t
r
e
e
0
.
8
9
0
.
8
7
0
.
8
4
0
.
8
5
G
r
a
d
i
e
n
t
b
o
o
s
t
i
n
g
0
.
9
3
0
.
9
4
0
.
8
8
0
.
9
1
R
a
n
d
o
m
f
o
r
e
s
t
0
.
9
2
0
.
9
5
0
.
8
3
0
.
8
8
X
G
B
o
o
s
t
0
.
9
3
0
.
9
4
0
.
8
8
0
.
9
1
W
ith
a
p
o
p
u
latio
n
s
ize
o
f
s
ix
ty
,
th
e
g
r
a
d
ien
t b
o
o
s
tin
g
an
d
X
GB
o
o
s
t
class
if
ier
s
ac
h
iev
ed
a
s
u
b
s
tan
tial
lev
el
o
f
ac
cu
r
ac
y
,
as
s
h
o
wn
in
T
ab
le
1
0
.
Gr
a
d
ien
t
b
o
o
s
tin
g
was
ab
le
to
ac
h
iev
e
a
m
a
x
im
u
m
ac
cu
r
ac
y
o
f
0
.
9
3
4
9
3
5
af
ter
1
5
iter
atio
n
s
,
wh
ils
t
XG
B
o
o
s
t
was
ab
le
to
ac
h
iev
e
an
ac
c
u
r
ac
y
o
f
0
.
9
3
5
5
5
6
af
ter
th
e
s
am
e
n
u
m
b
er
o
f
r
o
u
n
d
s
.
A
g
r
ea
t
p
e
r
f
o
r
m
a
n
ce
was
d
is
p
lay
ed
b
y
b
o
th
th
e
RF
an
d
DT
class
if
ier
s
.
T
h
e
RF
class
if
ier
attain
ed
an
ac
cu
r
ac
y
o
f
0
.
9
1
8
3
9
8
a
f
ter
3
0
iter
atio
n
s
,
wh
ile
t
h
e
DT
class
if
ier
r
ea
c
h
ed
a
n
ac
cu
r
ac
y
o
f
0
.
8
9
0
6
9
8
u
n
d
er
th
e
s
am
e
co
n
d
itio
n
s
.
T
h
e
r
esu
lts
o
f
th
e
s
tu
d
y
r
ev
ea
l
th
at
th
e
g
r
ad
ien
t
b
o
o
s
tin
g
an
d
X
GB
o
o
s
t
cla
s
s
if
ier
s
d
is
p
lay
ed
co
n
s
is
ten
tly
h
ig
h
er
ac
cu
r
ac
y
,
p
ar
ticu
lar
ly
wh
en
th
e
p
o
p
u
latio
n
s
izes
wer
e
r
aised
,
an
d
th
e
n
u
m
b
e
r
o
f
iter
atio
n
s
was
in
cr
ea
s
ed
.
W
h
e
n
co
m
b
in
e
d
with
r
eliab
le
clas
s
if
ier
s
,
th
e
PO
m
eth
o
d
r
esu
lts
in
a
co
n
s
id
er
ab
le
im
p
r
o
v
em
e
n
t
in
th
e
p
r
e
d
icted
ac
cu
r
ac
y
o
f
Alzh
eim
e
r
’
s
d
is
ea
s
e
d
iag
n
o
s
is
.
W
ith
a
p
o
p
u
latio
n
s
ize
o
f
s
ix
ty
an
d
f
if
teen
iter
atio
n
s
,
th
e
XGBo
o
s
t
class
if
ier
was
ab
le
to
ac
h
iev
e
an
ac
c
u
r
ac
y
o
f
0
.
9
3
5
5
5
6
f
o
r
th
e
PO
m
eth
o
d
.
T
ab
le
1
0
p
r
o
v
id
es
a
c
o
m
p
r
eh
en
s
iv
e
an
aly
s
is
o
f
th
e
p
e
r
f
o
r
m
an
ce
o
f
th
e
PO
m
eth
o
d
,
wh
ich
u
tili
ze
s
a
p
o
p
u
latio
n
s
ize
o
f
s
ix
ty
with
a
n
u
m
b
e
r
o
f
d
if
f
er
e
n
t c
lass
if
ier
s
an
d
iter
atio
n
s
in
clu
d
e
d
.
7.
CO
NCLU
SI
O
N
T
h
is
s
tu
d
y
h
ig
h
lig
h
ts
th
e
s
ig
n
if
ican
t
en
h
an
ce
m
en
t
o
f
p
r
ed
icti
v
e
ac
cu
r
ac
y
i
n
class
if
ier
s
ap
p
lied
to
th
e
Alzh
eim
er
’
s
d
is
ea
s
e
d
ataset
t
h
r
o
u
g
h
th
e
cr
itical
r
o
le
o
f
f
ea
t
u
r
e
s
elec
tio
n
.
T
h
is
h
ig
h
lig
h
ts
th
e
ef
f
ec
tiv
en
ess
o
f
em
p
lo
y
in
g
m
etah
e
u
r
is
tic
o
p
ti
m
izatio
n
alg
o
r
ith
m
s
,
s
p
ec
if
ic
ally
PO,
GW
O,
an
d
DE
,
to
a
ttain
th
is
o
b
jectiv
e.
T
h
e
g
r
a
d
ien
t
b
o
o
s
tin
g
an
d
XGBo
o
s
t
clas
s
if
ier
s
d
em
o
n
s
tr
ated
en
h
a
n
ce
d
ac
c
u
r
ac
y
,
p
r
ec
is
io
n
,
r
ec
all,
an
d
Evaluation Warning : The document was created with Spire.PDF for Python.