C
omp
u
te
r
S
c
i
e
n
c
e
an
d
I
n
for
mati
on
T
e
c
h
n
ol
ogi
e
s
V
ol
.
6
,
N
o
.
3
,
N
ove
m
b
e
r
20
25
,
p
p.
335
~
345
IS
S
N
:
2722
-
3221
,
D
O
I
:
10.
1
1591
/
c
s
i
t
.
v6
i
3
.
pp
33
5
-
34
5
335
Jou
r
n
al
h
o
m
e
pa
ge
:
ht
t
p:
/
/
i
ae
s
pr
i
m
e
.
c
om
/
i
nd
e
x
.
php
/
c
s
i
t
A
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
a
p
p
r
o
a
c
h
f
o
r
e
a
r
l
y
p
r
e
d
i
c
t
i
o
n
o
f
me
n
t
a
l
h
e
a
l
t
h
c
r
i
ses
H
as
s
an
C
h
i
ga
gu
r
e
1
,
Lu
c
y
C
h
a
r
i
t
y
S
ak
a
l
a
2
1
D
e
p
a
rt
m
e
n
t
o
f
Co
m
p
u
t
e
r
S
c
i
e
n
c
e
,
S
c
h
o
o
l
o
f
In
fo
rm
a
t
i
o
n
S
c
i
e
n
c
e
a
n
d
T
e
c
h
n
o
l
o
g
y
,
H
a
ra
re
In
s
t
i
t
u
t
e
o
f
T
e
c
h
n
o
l
o
g
y
,
H
a
ra
re
,
Z
i
m
b
a
b
w
e
2
D
e
p
a
rt
m
e
n
t
o
f
Co
m
p
u
t
e
r
S
c
i
e
n
c
e
,
F
a
c
u
l
t
y
o
f
S
c
i
e
n
c
e
a
n
d
E
n
g
i
n
e
e
ri
n
g
,
Bi
n
d
u
ra
U
n
i
v
e
rs
i
t
y
o
f
S
c
i
e
n
c
e
E
d
u
c
a
t
i
o
n
,
Bi
n
d
u
ra
,
Z
i
m
b
a
b
w
e
A
r
ti
c
l
e
I
n
fo
A
BS
TR
A
C
T
Ar
t
i
c
l
e
h
i
s
t
or
y
:
Re
c
e
i
v
e
d
J
un
3
,
2025
Re
vi
s
e
d
J
un
30
,
2025
A
c
c
e
pt
e
d
J
u
l
13
,
2025
T
he
g
l
ob
a
l
m
e
n
t
a
l
h
e
a
l
t
h
c
r
i
s
i
s
,
i
nt
e
ns
i
f
i
e
d
by
t
he
C
O
V
I
D
-
19
p
a
nde
m
i
c
,
pl
a
c
e
d
u
npr
e
c
e
de
n
t
e
d
s
t
r
a
i
n
on
h
e
a
l
t
hc
a
r
e
s
ys
t
e
m
s
a
nd
h
i
gh
l
i
ght
e
d
t
h
e
ur
ge
nt
ne
e
d
f
or
p
r
o
a
c
t
i
v
e
c
r
i
s
i
s
p
r
e
v
e
nt
i
o
n
s
t
r
a
t
e
gi
e
s
.
T
h
i
s
s
t
udy
i
n
ve
s
t
i
g
a
t
e
d
t
he
e
f
f
e
c
t
i
ve
n
e
s
s
o
f
va
r
i
ous
m
a
c
h
i
n
e
l
e
a
r
n
i
ng
(
M
L
)
m
od
e
l
s
i
n
pr
e
d
i
c
t
i
ng
m
e
n
t
a
l
h
e
a
l
t
h
c
r
i
s
e
s
w
i
t
h
i
n
28
da
y
s
pos
t
-
hos
pi
t
a
l
i
z
a
t
i
o
n,
l
e
v
e
r
a
g
i
ng
a
n
e
i
ght
-
ye
a
r
l
ong
i
t
ud
i
na
l
da
t
a
s
e
t
.
M
u
l
t
i
p
l
e
da
t
a
pr
e
p
r
oc
e
s
s
i
ng
t
e
c
h
ni
que
s
,
i
nc
l
ud
i
ng
f
e
a
t
u
r
e
s
e
l
e
c
t
i
o
n
(
E
F
S
A
,
R
F
E
C
V
)
,
i
m
pu
t
a
t
i
on,
a
nd
c
l
a
s
s
i
m
ba
l
a
nc
e
h
a
nd
l
i
ng
(
S
M
O
T
E
,
T
o
m
e
k
l
i
nk
s
)
,
w
e
r
e
s
ys
t
e
m
a
t
i
c
a
l
l
y
a
p
pl
i
e
d
t
o
e
nha
n
c
e
m
od
e
l
pe
r
f
or
m
a
nc
e
.
S
i
x
t
r
a
di
t
i
ona
l
c
l
a
s
s
i
f
i
e
r
s
—
l
og
i
s
t
i
c
r
e
g
r
e
s
s
i
on,
s
u
pp
or
t
ve
c
t
or
m
a
c
hi
ne
,
k
-
n
e
a
r
e
s
t
ne
i
gh
bor
s
,
n
a
i
v
e
B
a
ye
s
,
X
G
B
o
os
t
,
a
n
d
A
da
B
oo
s
t
—
w
e
r
e
e
va
l
ua
t
e
d
a
l
o
ngs
i
de
e
ns
e
m
bl
e
l
e
a
r
ni
n
g
(
E
L
)
m
e
t
hod
s
(
b
a
gg
i
ng
,
boos
t
i
ng,
s
t
a
c
k
i
ng
)
.
P
e
r
f
or
m
a
nc
e
m
e
t
r
i
c
s
s
uc
h
a
s
a
c
c
ur
a
c
y
,
pr
e
c
i
s
i
on
,
r
e
c
a
l
l
,
F
1
s
c
o
r
e
,
a
nd
A
U
C
-
R
O
C
w
e
r
e
u
s
e
d
f
o
r
c
om
pr
e
he
n
s
i
ve
a
s
s
e
s
s
m
e
n
t
.
R
e
s
u
l
t
s
de
m
ons
t
r
a
t
e
d
t
h
a
t
e
ns
e
m
bl
e
m
e
t
ho
ds
,
pa
r
t
i
c
u
l
a
r
l
y
b
oos
t
i
ng
a
n
d
ba
ggi
ng,
c
ons
i
s
t
e
n
t
l
y
a
c
hi
e
ve
d
hi
gh
pr
e
d
i
c
t
i
v
e
a
c
c
u
r
a
c
y
(
up
t
o
9
3%
)
,
w
i
t
h
X
G
B
oos
t
a
nd
A
d
a
B
oo
s
t
e
m
e
r
gi
ng
a
s
t
op
pe
r
f
o
r
m
e
r
s
.
F
e
a
t
ur
e
s
e
l
e
c
t
i
o
n
a
nd
c
l
a
s
s
i
m
ba
l
a
n
c
e
t
e
c
hn
i
qu
e
s
f
u
r
t
h
e
r
i
m
p
r
ov
e
d
m
ode
l
r
obu
s
t
n
e
s
s
a
nd
ge
ne
r
a
l
i
z
a
b
i
l
i
t
y.
T
h
e
f
i
ndi
ngs
un
de
r
s
c
or
e
d
t
he
po
t
e
nt
i
a
l
of
M
L
-
dr
i
ve
n
a
pp
r
oa
c
he
s
f
or
e
a
r
l
y
i
d
e
nt
i
f
i
c
a
t
i
on
o
f
a
t
-
r
i
s
k
p
a
t
i
e
n
t
s
,
e
n
a
bl
i
ng
m
o
r
e
e
f
f
e
c
t
i
ve
r
e
s
ou
r
c
e
a
l
l
oc
a
t
i
on
a
n
d
t
i
m
e
l
y
i
nt
e
r
ve
n
t
i
ons
i
n
m
e
nt
a
l
he
a
l
t
h
c
a
r
e
.
R
e
c
o
m
m
e
nd
a
t
i
o
ns
f
or
i
nt
e
gr
a
t
i
ng
t
he
s
e
p
r
e
d
i
c
t
i
ve
t
oo
l
s
i
nt
o
c
l
i
ni
c
a
l
w
or
kf
l
ow
s
w
e
r
e
d
i
s
c
us
s
e
d
t
o
s
up
por
t
da
t
a
-
dr
i
v
e
n
de
c
i
s
i
o
n
-
m
a
k
i
n
g.
Ke
y
w
or
d
s
:
E
ns
e
m
bl
e
m
e
t
hods
F
e
a
t
u
re
s
e
l
e
c
t
i
o
n
L
ongi
t
ud
i
na
l
hos
pi
t
a
l
d
a
t
a
M
a
c
h
i
ne
l
e
a
r
ni
ng
M
e
nt
a
l
he
a
l
t
h
c
ri
s
i
s
pr
e
di
c
t
i
on
T
hi
s
i
s
an
op
e
n
ac
c
e
s
s
ar
t
i
c
l
e
u
nde
r
t
he
C
C
B
Y
-
SA
l
i
c
e
n
s
e
.
Cor
r
e
s
pon
di
n
g
Au
t
h
or
:
L
uc
y
Ch
a
ri
t
y
S
a
k
a
l
a
D
e
pa
r
t
m
e
nt
of
Co
m
pu
t
e
r
S
c
i
e
n
c
e
,
F
a
c
u
l
t
y
of
S
c
i
e
n
c
e
a
nd
E
n
gi
ne
e
ri
ng
Bi
ndur
a
U
n
i
v
e
rs
i
t
y
of
S
c
i
e
n
c
e
E
duc
a
t
i
on
Bi
ndur
a
,
Z
i
m
ba
bw
e
E
m
a
i
l
:
l
s
a
k
a
l
a
@
hi
t
.
a
c
.
z
w
1.
I
N
TR
O
D
U
C
TI
O
N
A
pproxi
m
a
t
e
l
y
on
e
bi
l
l
i
o
n
p
e
opl
e
g
l
ob
a
l
l
y
l
i
v
e
w
i
t
h
a
m
e
n
t
a
l
d
i
s
orde
r
[1]
.
T
he
CO
V
ID
-
19
pa
nd
e
m
i
c
h
a
s
s
i
gn
i
fi
c
a
n
t
l
y
i
nt
e
ns
i
f
i
e
d
t
h
i
s
w
o
rl
dw
i
de
m
e
n
t
a
l
h
e
a
l
t
h
e
m
e
rg
e
nc
y,
l
e
a
di
n
g
t
o
a
n
e
s
c
a
l
a
t
i
ng
n
e
e
d
for
m
e
n
t
a
l
he
a
l
t
hc
a
re
s
e
rvi
c
e
s
.
T
h
i
s
s
urg
e
i
n
d
e
m
a
nd
i
s
pu
t
t
i
n
g
a
s
t
r
a
i
n
on
h
e
a
l
t
h
c
a
r
e
s
ys
t
e
m
s
,
w
h
i
c
h
a
re
a
l
r
e
a
dy
g
ra
pp
l
i
n
g
w
i
t
h
a
s
hort
a
g
e
of
qu
a
l
i
fi
e
d
m
e
nt
a
l
h
e
a
l
t
h
pe
rs
on
ne
l
[
2]
.
M
e
nt
a
l
d
i
s
ord
e
rs
h
a
v
e
t
h
e
pot
e
n
t
i
a
l
t
o
s
i
gni
f
i
c
a
n
t
l
y
i
m
p
a
c
t
v
a
r
i
ous
a
s
pe
c
t
s
of
l
i
fe
,
e
nc
o
m
pa
s
s
i
n
g
a
c
a
d
e
m
i
c
or
pro
fe
s
s
i
on
a
l
a
c
h
i
e
v
e
m
e
n
t
,
i
nt
e
ra
c
t
i
ons
w
i
t
h
fa
m
i
l
y
a
nd
fri
e
nds
a
s
w
e
l
l
a
s
e
ng
a
ge
m
e
nt
w
i
t
hi
n
t
h
e
c
o
m
m
uni
t
y
[3]
,
[4]
.
T
h
e
gl
o
ba
l
e
c
on
om
y
s
uffe
rs
a
l
os
s
of
$1
t
r
i
l
l
i
on
a
nn
ua
l
l
y
due
t
o
de
c
r
e
a
s
e
d
prod
uc
t
i
v
i
t
y
c
a
us
e
d
by
t
w
o
p
re
v
a
l
e
nt
m
e
nt
a
l
di
s
orde
rs
:
a
nxi
e
t
y
a
nd
de
pr
e
s
s
i
on
[5]
.
W
h
e
n
c
ons
i
de
r
i
ng
ov
e
ra
l
l
m
e
nt
a
l
i
s
s
u
e
s
,
i
n
c
l
ud
i
ng
re
d
uc
e
d
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2722
-
3221
Com
pu
t
S
c
i
Inf
T
e
c
h
nol
,
V
o
l
.
6
,
N
o
.
3
,
N
ov
e
m
be
r
20
25
:
335
-
345
336
produc
t
i
v
i
t
y
a
n
d
t
he
burd
e
n
of
i
l
l
n
e
s
s
,
t
he
gl
o
ba
l
e
c
ono
m
i
c
b
urde
n
a
s
s
oc
i
a
t
e
d
w
i
t
h
t
hi
s
i
s
s
ue
w
a
s
e
s
t
i
m
a
t
e
d
t
o
be
a
roun
d
$
2.
5
t
r
i
l
l
i
on
i
n
2
010,
a
f
i
gur
e
p
roj
e
c
t
e
d
t
o
c
l
i
m
b
t
o
$6
t
ri
l
l
i
on
by
2030
[6]
.
P
rom
pt
i
nt
e
rv
e
nt
i
on
c
a
n
of
t
e
n
pr
e
v
e
nt
t
he
w
ors
e
ni
n
g
of
s
y
m
pt
o
m
s
t
h
a
t
l
e
a
d
t
o
m
e
n
t
a
l
he
a
l
t
h
c
r
i
s
e
s
a
nd
s
ubs
e
qu
e
nt
hos
pi
t
a
l
i
z
a
t
i
on
[7]
.
U
nfo
rt
un
a
t
e
l
y
,
pa
t
i
e
nt
s
o
f
t
e
n
onl
y
a
c
c
e
s
s
urg
e
nt
c
a
r
e
,
s
u
c
h
a
s
a
h
os
pi
t
a
l
or
ps
yc
hi
a
t
r
i
c
f
a
c
i
l
i
t
y
,
w
he
n
t
h
e
y
a
r
e
a
l
re
a
dy
i
n
t
h
e
m
i
ds
t
of
a
c
ri
s
i
s
[8]
.
A
t
t
h
i
s
s
t
a
ge
,
pre
ve
n
t
a
t
i
v
e
m
e
a
s
ur
e
s
a
r
e
no
l
on
ge
r
a
n
o
pt
i
on,
w
hi
c
h
l
i
m
i
t
s
t
he
a
b
i
l
i
t
y
of
ps
y
c
hi
a
t
r
i
c
s
e
rvi
c
e
s
t
o
e
ffe
c
t
i
ve
l
y
a
l
l
oc
a
t
e
t
h
e
i
r
a
l
re
a
dy
s
t
ra
i
n
e
d
r
e
s
our
c
e
s
[
9],
[10]
.
T
he
r
e
for
e
,
a
ke
y
s
t
e
p
i
n
i
m
prov
i
ng
p
a
t
i
e
n
t
ou
t
c
o
m
e
s
a
nd
m
a
na
gi
ng
c
a
s
e
l
o
a
ds
i
s
t
o
i
de
nt
i
fy
i
nd
i
vi
dua
l
s
a
t
ri
s
k
o
f
a
c
r
i
s
i
s
be
for
e
i
t
ha
pp
e
ns
[11]
.
In
bus
y
c
l
i
n
i
c
a
l
e
nvi
r
onm
e
nt
s
,
m
a
nu
a
l
l
y
re
vi
e
w
i
ng
va
s
t
a
m
oun
t
s
of
p
a
t
i
e
n
t
da
t
a
t
o
m
a
ke
pr
oa
c
t
i
ve
c
a
re
de
c
i
s
i
ons
i
s
s
i
m
pl
y
no
t
fe
a
s
i
b
l
e
;
i
t
i
s
uns
us
t
a
i
na
b
l
e
a
nd
pron
e
t
o
e
rro
rs
[9]
,
[12]
.
S
hi
ft
i
ng
t
he
s
e
t
a
s
ks
t
o
a
u
t
om
a
t
e
d
a
na
l
ys
i
s
o
f
hos
pi
t
a
l
r
e
c
or
ds
,
how
e
ve
r,
off
e
rs
a
pro
m
i
s
i
ng
s
ol
u
t
i
on
.
T
h
i
s
a
ppr
oa
c
h
c
ou
l
d
r
e
vol
u
t
i
o
ni
z
e
h
e
a
l
t
h
c
a
r
e
by
e
n
a
bl
i
ng
c
ont
i
nuous
,
l
a
rg
e
-
s
c
a
l
e
d
a
t
a
re
v
i
e
w
[13]
.
Re
s
e
a
r
c
h
ha
s
a
l
re
a
dy
s
h
ow
n
w
e
c
a
n
p
r
e
di
c
t
c
r
i
t
i
c
a
l
he
a
l
t
h
e
v
e
nt
s
for
m
a
ny
c
ondi
t
i
o
ns
,
l
i
k
e
h
ype
r
t
e
ns
i
on
,
d
i
a
b
e
t
e
s
,
c
i
r
c
u
l
a
t
ory
f
a
i
l
ur
e
,
hos
pi
t
a
l
re
a
dm
i
s
s
i
on
,
a
nd
e
ve
n
i
n
-
hos
pi
t
a
l
d
e
a
t
h
[9],
[14]
.
H
ow
e
ve
r
,
w
he
n
i
t
c
om
e
s
t
o
m
e
n
t
a
l
h
e
a
l
t
h
,
t
h
e
e
xi
s
t
i
ng
re
s
e
a
r
c
h
m
a
i
nl
y
fo
c
us
e
s
on
pre
d
i
c
t
i
n
g
s
pe
c
i
fi
c
e
v
e
nt
s
s
u
c
h
a
s
s
ui
c
i
d
e
,
s
e
l
f
-
ha
r
m
,
or
a
fi
rs
t
e
pi
s
o
de
of
ps
yc
hos
i
s
[11]
.
W
e
do
no
t
h
a
ve
a
s
m
uc
h
i
nfor
m
a
t
i
on
on
c
ont
i
nuo
us
l
y
pr
e
di
c
t
i
ng
t
h
e
w
i
de
r
r
a
ng
e
of
m
e
nt
a
l
h
e
a
l
t
h
c
ri
s
e
s
t
ha
t
ne
e
d
urge
nt
c
a
re
o
r
hos
pi
t
a
l
i
z
a
t
i
on.
A
l
ot
i
s
s
t
i
l
l
unknow
n
a
b
out
w
he
t
he
r
w
e
c
a
n
c
o
nt
i
nuous
l
y
us
e
m
a
c
hi
n
e
l
e
a
rni
ng
t
o
e
s
t
i
m
a
t
e
t
he
ri
s
k
of
a
n
i
m
pe
ndi
n
g
m
e
n
t
a
l
h
e
a
l
t
h
c
ri
s
i
s
[15]
.
If
w
e
c
oul
d,
i
t
w
oul
d
a
l
l
ow
us
t
o
be
t
t
e
r
a
l
l
oc
a
t
e
he
a
l
t
hc
a
re
s
t
a
ff
a
nd
po
t
e
n
t
i
a
l
l
y
p
re
v
e
nt
c
ri
s
e
s
fro
m
e
ve
n
h
a
pp
e
ni
ng
[16
]
.
I
n
a
ddi
t
i
o
n,
i
t
i
s
no
t
y
e
t
c
l
e
a
r
i
f
ne
w
pre
di
c
t
i
v
e
te
c
hnol
ogi
e
s
w
ou
l
d
be
t
ru
l
y
us
e
fu
l
for
m
e
nt
a
l
he
a
l
t
hc
a
r
e
pr
a
c
t
i
t
i
o
ne
rs
,
e
s
pe
c
i
a
l
l
y
c
on
c
e
rn
i
ng
t
h
e
i
r
i
m
p
a
c
t
o
n
he
a
l
t
h
ou
t
c
o
m
e
s
or
l
on
g
-
t
e
rm
c
os
t
s
a
vi
n
gs
[17
],
[18]
.
Curre
n
t
c
l
i
ni
c
a
l
pr
a
c
t
i
c
e
oft
e
n
re
l
i
e
s
on
re
t
ros
pe
c
t
i
ve
s
ym
pt
o
m
a
s
s
e
s
s
m
e
n
t
a
nd
s
e
l
f
-
re
p
ort
i
ng
,
w
h
i
c
h
c
a
n
be
unre
l
i
a
bl
e
a
nd
r
e
a
c
t
i
ve
.
Id
e
n
t
i
fy
i
ng
i
ndi
vi
du
a
l
s
a
t
hi
gh
ri
s
k
of
i
m
pe
ndi
ng
m
e
nt
a
l
he
a
l
t
h
c
ri
s
e
s
(e
.
g
.
,
s
e
ve
r
e
d
e
pre
s
s
i
v
e
e
pi
s
od
e
s
,
ps
yc
hot
i
c
br
e
a
ks
,
a
nd
s
u
i
c
i
da
l
i
de
a
t
i
on)
i
s
c
ha
l
l
e
ngi
ng
du
e
t
o
t
h
e
c
o
m
p
l
e
x
i
nt
e
rpl
a
y
of
c
l
i
n
i
c
a
l
,
be
h
a
vi
oura
l
,
s
o
c
i
a
l
,
a
nd
e
nvi
r
onm
e
n
t
a
l
f
a
c
t
ors
.
T
he
l
a
c
k
of
pre
di
c
t
i
ve
t
ool
s
l
e
a
ds
t
o
de
l
a
ye
d
i
nt
e
rve
nt
i
on
m
e
a
ni
ng
pa
t
i
e
nt
s
re
c
e
i
ve
c
a
re
l
a
t
e
,
of
t
e
n
w
he
n
t
he
i
r
c
ond
i
t
i
on
i
s
s
e
ve
r
e
,
r
e
qui
ri
ng
m
ore
i
nt
e
ns
i
v
e
a
nd
c
os
t
l
y
i
n
t
e
rve
n
t
i
o
ns
l
i
k
e
i
n
pa
t
i
e
nt
hos
p
i
t
a
l
i
z
a
t
i
o
n
[9]
.
A
ddi
t
i
on
a
l
l
y
,
i
n
c
re
a
s
e
d
s
uff
e
r
i
ng:
Indi
vi
dua
l
s
e
nd
ure
p
rol
o
nge
d
pe
r
i
ods
of
d
i
s
t
r
e
s
s
a
nd
fu
nc
t
i
ona
l
i
m
p
a
i
r
m
e
n
t
,
a
nd
i
ne
ffi
c
i
e
nt
r
e
s
our
c
e
a
l
l
oc
a
t
i
on
-
h
e
a
l
t
h
c
a
r
e
r
e
s
o
urc
e
s
a
r
e
of
t
e
n
de
p
l
oy
e
d
re
a
c
t
i
v
e
l
y
r
a
t
h
e
r
t
ha
n
s
t
ra
t
e
gi
c
a
l
l
y
,
l
e
a
d
i
ng
t
o
bo
t
t
l
e
n
e
c
ks
a
nd
pot
e
nt
i
a
l
b
urnou
t
f
or
m
e
nt
a
l
he
a
l
t
h
p
rofe
s
s
i
ona
l
s
.
T
hi
s
s
t
udy
a
ddr
e
s
s
e
s
i
m
por
t
a
n
t
g
a
ps
by
c
ondu
c
t
i
ng
a
c
om
pr
e
he
ns
i
ve
c
o
m
pa
ri
s
on
of
m
a
c
hi
n
e
l
e
a
rni
n
g
m
o
de
l
s
us
i
ng
a
n
e
i
gh
t
-
ye
a
r
l
ong
i
t
ud
i
na
l
d
a
t
a
s
e
t
t
o
pre
di
c
t
m
e
nt
a
l
h
e
a
l
t
h
c
r
i
s
e
s
.
I
t
a
i
m
s
t
o
i
de
n
t
i
f
y
ke
y
ri
s
k
fa
c
t
ors
t
ha
t
c
ont
r
i
but
e
m
os
t
t
o
c
r
i
s
i
s
pr
e
di
c
t
i
on
a
n
d
e
va
l
u
a
t
e
t
h
e
pe
rfor
m
a
n
c
e
of
di
f
fe
r
e
nt
m
od
e
l
s
.
T
he
fi
nd
i
ngs
provi
de
va
l
ua
b
l
e
i
ns
i
ght
s
a
nd
pra
c
t
i
c
a
l
re
c
o
m
m
e
nd
a
t
i
ons
for
e
ff
e
c
t
i
v
e
l
y
i
n
t
e
gra
t
i
ng
m
a
c
hi
n
e
l
e
a
rni
n
g
i
nt
o
m
e
n
t
a
l
h
e
a
l
t
h
c
a
r
e
s
ys
t
e
m
s
.
T
h
i
s
a
ppro
a
c
h
h
a
s
pot
e
nt
i
a
l
t
o
i
m
prov
e
e
a
rl
y
d
e
t
e
c
t
i
on
a
n
d
t
i
m
e
l
y
in
t
e
rve
nt
i
on
f
or
a
t
-
r
i
s
k
pa
t
i
e
nt
s
.
2.
M
ET
H
O
D
O
L
O
G
Y
2.
1
.
R
e
s
e
ar
c
h
d
e
s
i
gn
T
h
i
s
s
t
ud
y,
a
r
e
t
ros
p
e
c
t
i
v
e
c
oho
rt
s
t
u
dy
,
f
oc
us
e
d
o
n
d
e
v
e
l
o
pi
n
g
a
nd
a
s
s
e
s
s
i
ng
m
e
n
t
a
l
h
e
a
l
t
h
c
r
i
s
i
s
pre
di
c
t
i
on
m
od
e
l
s
us
i
n
g
e
x
i
s
t
i
n
g
h
e
a
l
t
h
r
e
c
or
ds
.
R
e
t
ros
pe
c
t
i
v
e
c
o
hor
t
s
t
u
di
e
s
a
n
a
l
y
z
e
a
l
re
a
d
y
c
o
l
l
e
c
t
e
d
d
a
t
a
t
o
e
v
a
l
u
a
t
e
ou
t
c
o
m
e
s
b
a
s
e
d
on
p
ri
or
e
xpos
ur
e
s
or
c
o
nd
i
t
i
ons
.
T
hi
s
r
e
s
e
a
r
c
h
e
m
pl
oys
a
q
ua
nt
i
t
a
t
i
v
e
m
e
t
ho
do
l
og
y,
l
e
ve
r
a
g
i
ng
a
m
a
c
h
i
n
e
l
e
a
r
ni
ng
a
p
pro
a
c
h
t
o
pr
e
d
i
c
t
m
e
nt
a
l
h
e
a
l
t
h
c
ri
s
e
s
a
m
on
g
p
a
t
i
e
n
t
s
w
i
t
h
i
n
a
2
8
-
da
y
p
e
r
i
o
d
fol
l
ow
i
n
g
hos
pi
t
a
l
i
z
a
t
i
on
[
9]
,
[1
9]
.
U
s
i
ng
t
hi
s
m
e
t
h
od
ol
ogy
a
l
l
ow
s
f
or
e
ff
i
c
i
e
n
t
a
na
l
ys
i
s
of
l
a
rg
e
d
a
t
a
s
e
t
a
nd
t
i
m
e
l
y
i
de
nt
i
f
i
c
a
t
i
on
of
a
t
-
ri
s
k
p
a
t
i
e
n
t
s
,
i
m
pr
ov
i
n
g
pr
e
d
i
c
t
i
v
e
a
c
c
ur
a
c
y
a
nd
h
e
a
l
t
h
c
a
r
e
i
n
t
e
r
ve
nt
i
o
n.
2.
2
.
F
e
atu
r
e
s
e
l
e
c
t
i
on
F
e
a
t
u
re
s
e
l
e
c
t
i
on
i
nvo
l
ve
s
c
hoos
i
n
g
a
s
ubs
e
t
of
pe
r
t
i
n
e
nt
f
e
a
t
ure
s
fr
om
a
l
a
rg
e
r
poo
l
of
a
va
i
l
a
bl
e
fe
a
t
ure
s
w
i
t
h
i
n
a
da
t
a
s
e
t
[2
0]
.
It
i
nv
ol
v
e
s
i
d
e
n
t
i
fy
i
ng
a
nd
re
t
a
i
ni
ng
t
h
e
m
os
t
i
nfo
rm
a
t
i
ve
a
nd
i
m
p
a
c
t
ful
fe
a
t
ure
s
w
hi
l
e
di
s
c
a
rd
i
ng
or
d
i
s
re
g
a
rd
i
ng
l
e
s
s
re
l
e
v
a
nt
or
re
d
unda
n
t
on
e
s
[
20]
–
[22
]
.
F
e
a
t
ur
e
s
e
l
e
c
t
i
on
a
i
m
s
t
o
e
nha
nc
e
t
h
e
pe
rf
orm
a
nc
e
,
i
n
t
e
rpr
e
t
a
b
i
l
i
t
y
,
a
nd
e
ffi
c
i
e
n
c
y
of
m
a
c
hi
n
e
l
e
a
r
ni
ng
m
od
e
l
s
by
c
onc
e
n
t
ra
t
i
ng
o
n
t
h
e
m
os
t
i
m
p
a
c
t
ful
a
s
p
e
c
t
s
of
t
he
da
t
a
[2
3]
.
T
hi
s
s
t
udy
t
e
s
t
e
d
t
w
o
f
e
a
t
ur
e
s
e
l
e
c
t
i
on
m
e
t
ho
ds
,
a
nd
bo
t
h
w
e
r
e
us
e
d
for
m
ode
l
c
ons
t
ru
c
t
i
on
.
T
he
t
w
o
m
e
t
h
ods
e
va
l
ua
t
e
d
w
e
r
e
t
h
e
e
ns
e
m
bl
e
of
fe
a
t
u
re
s
e
l
e
c
t
i
on
a
l
gor
i
t
h
m
s
(E
F
S
A
)
a
nd
re
c
urs
i
v
e
f
e
a
t
ur
e
e
l
i
m
i
na
t
i
o
n
w
i
t
h
c
ros
s
-
v
a
l
i
da
t
i
on
(RF
E
CV
).
2.
3
.
Exp
e
r
i
me
n
ta
l
p
r
oc
e
d
u
r
e
T
he
pro
c
e
d
ure
be
g
a
n
w
i
t
h
t
he
da
t
a
s
e
t
(c
o
m
pr
i
s
i
ng
da
t
a
s
e
t
s
f
rom
e
i
g
ht
di
ff
e
r
e
nt
ye
a
rs
)
b
e
i
ng
l
oa
d
e
d
i
nt
o
G
oog
l
e
Co
l
a
b
.
T
h
e
s
e
d
a
t
a
s
e
t
s
w
e
r
e
t
he
n
c
on
c
a
t
e
n
a
t
e
d
t
oge
t
h
e
r
t
o
c
re
a
t
e
a
uni
f
i
e
d
d
a
t
a
s
e
t
.
T
h
e
ne
x
t
s
t
e
p
i
nvol
ve
d
d
a
t
a
p
re
pr
oc
e
s
s
i
ng,
a
c
ru
c
i
a
l
ph
a
s
e
t
h
a
t
e
nc
o
m
p
a
s
s
e
s
s
e
v
e
r
a
l
e
s
s
e
nt
i
a
l
t
a
s
ks
[24]
.
D
a
t
a
c
l
e
a
ni
ng
w
a
s
Evaluation Warning : The document was created with Spire.PDF for Python.
Com
pu
t
S
c
i
Inf
T
e
c
h
nol
IS
S
N
:
2722
-
3221
A
m
a
c
hi
n
e
l
e
ar
ni
ng
ap
pr
oac
h
f
or
e
ar
l
y
pr
e
d
i
c
t
i
o
n
o
f
m
e
nt
al
he
al
t
h
c
r
i
s
e
s
(H
as
s
an
Chi
gagur
e
)
337
pe
rfor
m
e
d
t
o
r
e
m
o
ve
a
ny
i
nc
ons
i
s
t
e
nc
i
e
s
o
r
ou
t
l
i
e
rs
,
fo
l
l
o
w
e
d
by
d
a
t
a
i
m
pu
t
a
t
i
on
t
o
f
i
l
l
i
n
m
i
s
s
i
ng
v
a
l
u
e
s
.
L
a
b
e
l
e
nc
odi
n
g
w
a
s
a
ppl
i
e
d
t
o
m
a
k
e
t
h
e
d
a
t
a
nu
m
e
r
i
c
a
l
,
a
nd
d
a
t
a
e
x
pl
or
a
t
i
on
w
a
s
c
ondu
c
t
e
d
t
o
unv
e
i
l
unde
r
l
yi
n
g
t
re
nds
a
n
d
h
i
dd
e
n
pa
t
t
e
rns
w
i
t
hi
n
t
he
da
t
a
[25
]
.
T
h
e
a
n
a
l
ys
i
s
a
l
s
o
a
s
s
e
s
s
e
d
c
l
a
s
s
i
m
b
a
l
a
nc
e
a
nd
i
nvol
ve
d
fe
a
t
u
re
s
e
l
e
c
t
i
o
n
t
o
e
nha
nc
e
m
od
e
l
pe
r
form
a
n
c
e
.
S
ubs
e
que
nt
l
y,
t
h
e
e
xp
e
ri
m
e
nt
m
ov
e
d
on
t
o
m
o
de
l
c
ons
t
ru
c
t
i
on
,
w
h
e
re
a
va
r
i
e
t
y
o
f
t
ra
d
i
t
i
ona
l
M
L
m
ode
l
s
,
i
nc
l
ud
i
ng
L
R
,
S
V
M
,
K
-
N
N
,
N
B
,
X
G
Boos
t
,
a
nd
A
da
Boos
t
,
w
e
r
e
i
m
pl
e
m
e
n
t
e
d
.
F
urt
h
e
r
m
ore
,
e
ns
e
m
bl
e
l
e
a
rn
i
ng
(E
L
)
t
e
c
h
ni
qu
e
s
,
s
u
c
h
a
s
b
a
ggi
n
g,
bo
os
t
i
n
g,
a
nd
s
t
a
c
k
i
ng
,
w
e
r
e
ut
i
l
i
s
e
d
t
o
c
o
m
b
i
ne
t
he
s
e
m
o
de
l
s
[26]
.
T
h
e
ne
x
t
ph
a
s
e
w
a
s
m
od
e
l
e
va
l
ua
t
i
o
n,
i
n
w
hi
c
h
t
he
m
ode
l
'
s
p
e
rfor
m
a
nc
e
us
i
ng
a
n
a
rra
y
of
m
e
t
r
i
c
s
s
uc
h
a
s
a
c
c
ur
a
c
y
,
pr
e
c
i
s
i
on
,
re
c
a
l
l
,
F
1
s
c
or
e
,
k
a
pp
a
,
ge
o
m
e
t
ri
c
m
e
a
n,
a
nd
A
U
C
-
RO
C
w
a
s
a
s
s
e
s
s
e
d
[27]
.
T
hi
s
e
x
t
e
ns
i
ve
e
x
pe
ri
m
e
nt
a
l
s
e
t
up
w
a
s
c
r
e
a
t
e
d
t
o
m
a
k
e
i
t
e
a
s
i
e
r
t
o
a
s
s
e
s
s
t
he
s
t
udy'
s
r
e
s
ul
t
s
,
i
m
pro
ve
t
he
i
r
va
l
i
d
i
t
y
,
a
nd
d
e
t
e
r
m
i
n
e
w
he
t
he
r
t
h
e
y
m
i
gh
t
be
r
e
pl
i
c
a
t
e
d
i
n
ot
h
e
r
re
s
e
a
rc
h
s
e
t
t
i
ngs
.
G
oog
l
e
Col
a
b
Re
s
e
a
r
c
h
w
a
s
us
e
d
t
o
c
a
rry
out
e
m
pi
r
i
c
a
l
e
xp
e
r
i
m
e
nt
s
.
F
i
gure
1
i
l
l
us
t
ra
t
e
s
t
h
e
w
orkf
l
ow
d
i
a
gra
m
ou
t
l
i
ni
ng
t
he
e
n
t
i
re
e
xpe
r
i
m
e
n
t
a
l
pro
c
e
du
re
us
e
d
i
n
t
h
i
s
s
t
udy.
I
t
b
e
g
i
ns
w
i
t
h
l
o
a
d
i
ng
a
nd
pr
e
pro
c
e
s
s
i
ng
t
h
e
d
a
t
a
s
e
t
,
f
ol
l
ow
e
d
by
fe
a
t
ur
e
s
e
l
e
c
t
i
o
n
t
o
i
d
e
nt
i
fy
t
h
e
m
os
t
re
l
e
va
n
t
pre
d
i
c
t
ors
.
T
he
w
orkf
l
ow
t
he
n
pr
oc
e
e
ds
t
o
m
ode
l
d
e
ve
l
op
m
e
nt
,
i
nc
l
ud
i
ng
hype
rpa
r
a
m
e
t
e
r
t
un
i
ng
a
n
d
t
ra
i
ni
ng
m
ul
t
i
p
l
e
c
l
a
s
s
i
fi
e
rs
.
F
i
n
a
l
l
y
,
t
h
e
m
od
e
l
s
a
r
e
e
v
a
l
ua
t
e
d
us
i
ng
pe
rf
orm
a
n
c
e
m
e
t
ri
c
s
,
a
n
d
t
he
r
e
s
ul
t
s
a
re
a
na
l
ys
e
d
t
o
de
t
e
r
m
i
n
e
t
he
be
s
t
pr
e
di
c
t
i
ve
a
p
proa
c
h
.
T
hus
,
F
i
gure
1
prov
i
d
e
s
a
c
o
m
pr
e
he
ns
i
ve
ove
rvi
e
w
of
t
h
e
s
ys
t
e
m
a
t
i
c
s
t
e
ps
un
de
r
t
a
k
e
n
t
o
e
ns
ur
e
robus
t
a
nd
re
pr
oduc
i
bl
e
m
a
c
hi
n
e
l
e
a
rn
i
ng
e
x
pe
r
i
m
e
nt
s
.
F
i
gure
1
.
W
orkf
l
ow
di
a
gra
m
i
l
l
us
t
r
a
t
i
ng
e
x
pe
r
i
m
e
nt
a
l
pro
c
e
d
ure
2.
4
.
M
od
e
l
In
pu
rs
ui
t
o
f
t
he
o
ve
r
a
rc
hi
ng
g
oa
l
o
f
a
s
s
e
s
s
i
ng
t
he
pr
e
d
i
c
t
i
ve
c
a
p
a
bi
l
i
t
i
e
s
of
M
L
t
e
c
hn
i
qu
e
s
,
t
h
e
pri
m
a
ry
obj
e
c
t
i
v
e
w
a
s
t
o
e
va
l
ua
t
e
how
e
ffe
c
t
i
ve
l
y
m
a
c
h
i
n
e
l
e
a
r
ni
ng
t
e
c
h
ni
qu
e
s
c
a
n
pre
di
c
t
m
e
nt
a
l
he
a
l
t
h
c
ri
s
e
s
.
T
o
a
c
hi
e
v
e
t
h
i
s
,
m
od
e
l
s
w
e
re
b
ui
l
t
us
i
ng
a
n
i
m
p
ut
e
d
da
t
a
s
e
t
,
w
hi
c
h
e
ns
ure
d
t
h
e
i
n
t
e
g
ri
t
y
a
n
d
c
om
p
l
e
t
e
n
e
s
s
o
f
t
he
d
a
t
a
.
T
hi
s
robus
t
f
ound
a
t
i
on
a
l
l
ow
e
d
for
m
o
re
re
l
i
a
bl
e
a
n
d
a
c
c
ur
a
t
e
p
re
d
i
c
t
i
v
e
m
od
e
l
i
ng
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2722
-
3221
Com
pu
t
S
c
i
Inf
T
e
c
h
nol
,
V
o
l
.
6
,
N
o
.
3
,
N
ov
e
m
be
r
20
25
:
335
-
345
338
Cons
e
qu
e
nt
l
y,
t
he
s
t
udy
t
horoug
hl
y
a
s
s
e
s
s
e
s
t
h
e
pot
e
nt
i
a
l
o
f
m
a
c
h
i
n
e
l
e
a
rni
n
g
a
s
a
v
a
l
u
a
bl
e
t
oo
l
i
n
m
e
nt
a
l
he
a
l
t
h
c
r
i
s
i
s
pre
d
i
c
t
i
o
n.
T
he
pre
d
i
c
t
i
v
e
m
od
e
l
s
t
he
m
s
e
l
ve
s
e
nc
om
p
a
s
s
e
d
a
s
e
l
e
c
t
i
on
of
w
i
d
e
l
y
re
c
ogn
i
z
e
d
c
l
a
s
s
i
fi
e
rs
,
i
nc
l
udi
n
g
t
w
o
boos
t
i
ng
c
l
a
s
s
i
fi
e
rs
know
n
for
t
he
i
r
a
b
i
l
i
t
y
t
o
e
nha
n
c
e
p
re
d
i
c
t
i
v
e
pe
rf
orm
a
nc
e
[28]
.
T
he
t
a
bl
e
be
l
ow
s
um
m
a
r
i
z
e
s
t
he
s
e
qu
e
nt
i
a
l
pr
oc
e
dure
s
t
a
ke
n
t
o
g
e
ne
r
a
t
e
e
a
c
h
pr
e
di
c
t
i
ve
m
ode
l
a
nd
offe
rs
a
t
hor
ough
ove
rv
i
e
w
of
t
he
a
pp
roa
c
he
s
us
e
d
for
m
od
e
l
c
ons
t
r
u
c
t
i
on
.
T
h
e
re
s
u
l
t
s
of
t
he
s
e
m
o
de
l
bui
l
ds
s
e
rv
e
a
s
e
s
s
e
nt
i
a
l
i
ns
t
ru
m
e
n
t
s
for
e
v
a
l
u
a
t
i
ng
h
ow
w
e
l
l
m
a
c
h
i
ne
l
e
a
rn
i
ng
t
e
c
h
ni
que
s
fore
c
a
s
t
m
e
n
t
a
l
he
a
l
t
h
c
ri
s
e
s
.
I
n
t
h
e
e
nd
,
t
he
c
o
m
pa
r
a
t
i
ve
a
n
a
l
ys
i
s
he
l
ps
d
e
t
e
rm
i
ne
t
h
e
m
os
t
pr
o
m
i
s
i
n
g
pa
t
h
f
or
a
c
c
ura
t
e
m
e
nt
a
l
h
e
a
l
t
h
c
r
i
s
i
s
pre
di
c
t
i
ons
by
i
l
l
um
i
n
a
t
i
ng
t
he
m
a
ny
a
ppro
a
c
h
e
s
w
i
t
h
di
f
fe
r
i
ng
d
e
gre
e
s
o
f
e
ffe
c
t
i
ve
ne
s
s
.
A
l
gori
t
hm
1
ou
t
l
i
ne
s
t
h
e
s
t
e
p
-
by
-
s
t
e
p
a
l
gor
i
t
h
m
us
e
d
for
d
e
v
e
l
op
i
ng
t
h
e
m
a
c
h
i
n
e
l
e
a
rni
ng
m
od
e
l
s
i
n
t
hi
s
s
t
udy
.
It
de
t
a
i
l
s
t
h
e
d
a
t
a
pre
p
roc
e
s
s
i
ng
s
t
a
ge
s
s
uc
h
a
s
l
o
a
di
ng
t
he
da
t
a
s
e
t
,
ha
nd
l
i
n
g
m
i
s
s
i
n
g
t
a
rg
e
t
v
a
l
u
e
s
,
a
nd
fe
a
t
ur
e
s
e
l
e
c
t
i
o
n.
A
l
g
ori
t
hm
1
a
l
s
o
c
ov
e
rs
t
h
e
c
l
a
s
s
i
fi
c
a
t
i
on
w
orkfl
ow
,
i
n
c
l
udi
ng
h
ype
rp
a
r
a
m
e
t
e
r
t
un
i
ng
,
m
ode
l
t
r
a
i
n
i
ng,
pr
e
di
c
t
i
on
,
pe
rfo
rm
a
nc
e
e
v
a
l
u
a
t
i
on
,
a
nd
re
s
our
c
e
us
a
ge
m
oni
t
ori
ng
a
c
r
os
s
di
ffe
r
e
nt
c
l
a
s
s
i
fi
e
rs
.
T
h
i
s
s
t
r
uc
t
ure
d
a
ppro
a
c
h
e
ns
ure
s
a
c
o
m
pr
e
h
e
ns
i
v
e
a
nd
re
pr
oduc
i
b
l
e
m
o
de
l
d
e
ve
l
op
m
e
n
t
proc
e
s
s
.
T
a
b
l
e
1
s
um
m
a
ri
z
e
s
t
he
di
ff
e
re
nt
c
o
m
bi
n
a
t
i
ons
of
fe
a
t
ur
e
s
e
l
e
c
t
i
on
m
e
t
hods
,
nu
m
be
r
of
s
e
l
e
c
t
e
d
fe
a
t
ure
s
,
a
nd
c
l
a
s
s
i
m
ba
l
a
n
c
e
s
t
ra
t
e
g
i
e
s
us
e
d
i
n
o
ur
m
ode
l
s
.
T
he
fe
a
t
ur
e
s
s
e
l
e
c
t
e
d
i
n
t
he
s
e
m
o
de
l
s
i
n
c
l
ud
e
k
e
y
pre
di
c
t
ors
s
u
c
h
a
s
h
i
s
t
or
i
c
a
l
s
ym
pt
o
m
s
e
v
e
ri
t
y
(t
o
t
a
l
nu
m
be
r
a
nd
dur
a
t
i
on
of
c
ri
s
i
s
e
pi
s
od
e
s
),
hos
pi
t
a
l
i
nt
e
ra
c
t
i
ons
(un
pl
a
nne
d
c
o
nt
a
c
t
s
,
m
i
s
s
e
d
a
p
poi
n
t
m
e
nt
s
,
r
e
c
e
n
t
c
r
i
s
e
s
)
,
pa
t
i
e
nt
a
g
e
,
i
n
di
v
i
dua
l
r
i
s
k
i
n
di
c
e
s
,
a
n
d
t
ot
a
l
t
i
m
e
s
i
nc
e
t
he
f
i
rs
t
hos
pi
t
a
l
v
i
s
i
t
.
A
s
s
how
n
i
n
T
a
b
l
e
1
,
bot
h
E
F
S
A
a
n
d
RF
E
CV
w
e
r
e
a
pp
l
i
e
d,
us
i
n
g
e
i
t
he
r
e
i
gh
t
or
fi
v
e
of
t
h
e
s
e
s
i
gn
i
fi
c
a
n
t
fe
a
t
u
re
s
,
i
n
c
o
m
bi
na
t
i
on
w
i
t
h
c
l
a
s
s
i
m
b
a
l
a
nc
e
t
e
c
h
ni
qu
e
s
s
u
c
h
a
s
S
M
O
T
E
a
nd
T
om
e
k
l
i
n
k
s.
A
l
gori
t
hm
1
.
M
a
c
h
i
n
e
l
e
a
rni
ng
m
od
e
l
de
ve
l
opm
e
n
t
S
t
a
rt
A
l
g
ori
t
hm
:
1.
L
o
a
d
t
he
d
a
t
a
s
e
t
.
2.
Re
m
ov
e
i
ns
t
a
nc
e
s
w
i
t
h
a
nul
l
t
a
r
ge
t
v
a
ri
a
bl
e
.
3.
A
p
pl
y
fe
a
t
ure
s
e
l
e
c
t
i
o
n
t
o
re
t
a
i
n
r
e
l
e
va
nt
fe
a
t
ur
e
s
.
4.
S
p
l
i
t
t
he
d
a
t
a
i
nt
o
t
r
a
i
ni
ng
a
nd
t
e
s
t
i
ng
s
e
t
s
b
a
s
e
d
on
t
h
e
s
p
e
c
i
fi
e
d
r
a
t
i
o
.
5.
I
ni
t
i
a
l
i
z
e
a
n
e
m
p
t
y
di
c
t
i
ona
ry
t
o
s
t
ore
c
l
a
s
s
i
fi
e
r
r
e
s
ul
t
s
.
6.
F
or
e
a
c
h
c
l
a
s
s
i
fi
e
r:
−
6.
1
.
D
e
fi
n
e
t
he
hy
pe
rp
a
ra
m
e
t
e
r
g
ri
d
for
hyp
e
rp
a
ra
m
e
t
e
r
t
uni
ng.
−
6.
2
.
F
i
nd
t
he
b
e
s
t
pa
r
a
m
e
t
e
rs
.
−
6.
3
.
S
t
ore
t
h
e
be
s
t
p
a
ra
m
e
t
e
rs
for
t
h
e
c
l
a
s
s
i
fi
e
r
.
7.
F
or
e
a
c
h
c
l
a
s
s
i
fi
e
r:
−
7.
1
.
In
i
t
i
a
l
i
s
e
t
he
c
l
a
s
s
i
f
i
e
r
w
i
t
h
t
he
b
e
s
t
hyp
e
rpa
r
a
m
e
t
e
rs
.
−
7.
2
.
S
t
a
rt
t
he
t
i
m
e
r
a
nd
m
e
m
ory
m
on
i
t
o
r.
−
7.
3
.
T
ra
i
n
t
h
e
c
l
a
s
s
i
fi
e
r
on
t
he
t
ra
i
ni
ng
d
a
t
a
.
−
7.
4
.
M
a
ke
pr
e
di
c
t
i
ons
o
n
t
he
t
e
s
t
s
e
t
.
−
7.
5
.
R
e
c
o
rd
t
he
e
l
a
ps
e
d
t
i
m
e
a
nd
m
e
m
ory
us
a
g
e
.
−
7.
6
.
C
a
l
c
ul
a
t
e
pe
rfo
rm
a
nc
e
m
e
t
r
i
c
−
7.
7
.
S
t
ore
t
h
e
re
s
ul
t
s
i
n
t
h
e
d
i
c
t
i
on
a
ry
w
i
t
h
t
he
c
l
a
s
s
i
fi
e
r
n
a
m
e
a
s
t
he
ke
y.
8.
D
i
s
pl
a
y
t
he
r
e
s
ul
t
s
fro
m
t
h
e
d
i
c
t
i
on
a
ry
for
e
a
c
h
c
l
a
s
s
i
fi
e
r,
i
nc
l
ud
i
ng
:
−
Cl
a
s
s
i
f
i
e
r
n
a
m
e
−
Be
s
t
hyp
e
rpa
r
a
m
e
t
e
rs
−
P
e
rform
a
n
c
e
m
e
t
r
i
c
s
−
E
l
a
ps
e
d
t
i
m
e
a
nd
m
e
m
ory
us
a
g
e
E
nd
A
l
gor
i
t
h
m
.
2.
5
.
Eva
l
u
ati
on
W
e
a
p
pr
oa
c
h
e
d
t
h
e
c
ri
s
i
s
p
re
di
c
t
i
on
t
a
s
k
a
s
a
bi
na
ry
c
l
a
s
s
i
fi
c
a
t
i
on
pro
bl
e
m
[2
9]
.
T
he
m
o
d
e
l
w
a
s
de
s
i
gn
e
d
t
o
pr
e
di
c
t
t
h
e
ri
s
k
of
a
c
ri
s
i
s
d
e
v
e
l
op
i
n
g
w
i
t
h
i
n
t
h
e
n
e
x
t
28
d
a
ys
.
T
o
e
v
a
l
u
a
t
e
t
h
i
s
,
w
e
us
e
d
a
t
i
m
e
-
ba
s
e
d
s
p
l
i
t
o
f
t
h
e
d
a
t
a
:
8
0%
f
or
t
ra
i
n
i
n
g
,
10
%
for
va
l
i
d
a
t
i
o
n,
a
nd
10
%
fo
r
t
e
s
t
i
n
g
.
T
h
e
m
o
de
l
s
d
e
v
e
l
op
e
d
i
n
t
hi
s
s
t
ud
y
w
a
s
a
s
s
e
s
s
e
d
b
a
s
e
d
o
n
how
a
c
c
u
ra
t
e
l
y
a
nd
e
ff
e
c
t
i
v
e
l
y
t
he
y
i
d
e
nt
i
f
y
pa
t
i
e
nt
s
a
t
r
i
s
k
o
f
a
m
e
nt
a
l
he
a
l
t
h
c
r
i
s
i
s
w
i
t
hi
n
2
8
da
ys
f
ol
l
o
w
i
ng
t
h
e
i
r
hos
pi
t
a
l
i
z
a
t
i
on
.
M
o
de
l
e
f
fi
c
a
c
y
w
a
s
m
e
a
s
ur
e
d
us
i
n
g
a
va
ri
e
t
y
of
pe
r
for
m
a
n
c
e
i
nd
i
c
a
t
o
rs
,
i
nc
l
u
di
ng
a
c
c
ur
a
c
y
,
pr
e
c
i
s
i
o
n
,
r
e
c
a
l
l
,
F
1
s
c
or
e
,
a
nd
A
U
C
-
R
O
C
.
A
c
c
ur
a
c
y
w
i
l
l
p
rov
i
de
a
b
ro
a
d
ov
e
rv
i
e
w
o
f
t
h
e
m
od
e
l
'
s
p
e
rf
or
m
a
n
c
e
,
w
hi
l
s
t
p
re
c
i
s
i
on
a
nd
r
e
c
a
l
l
w
i
l
l
pr
ov
i
d
e
i
nf
or
m
a
t
i
o
n
a
bo
ut
t
h
e
m
o
de
l
'
s
c
a
p
a
c
i
t
y
t
o
a
c
c
ur
a
t
e
l
y
di
s
t
i
ng
ui
s
h
t
ru
e
p
os
i
t
i
v
e
s
vs
fa
l
s
e
pos
i
t
i
v
e
s
.
T
h
e
F
1
s
c
o
re
w
i
l
l
fun
c
t
i
on
a
s
a
ba
l
a
n
c
e
d
a
s
s
e
s
s
m
e
n
t
,
w
h
i
c
h
i
s
e
s
p
e
c
i
a
l
l
y
r
e
l
e
va
nt
i
n
he
a
l
t
h
c
a
re
s
e
t
t
i
ngs
w
h
e
r
e
f
a
l
s
e
ne
ga
t
i
v
e
s
c
a
n
h
a
v
e
s
e
ri
ous
Evaluation Warning : The document was created with Spire.PDF for Python.
Com
pu
t
S
c
i
Inf
T
e
c
h
nol
IS
S
N
:
2722
-
3221
A
m
a
c
hi
n
e
l
e
ar
ni
ng
ap
pr
oac
h
f
or
e
ar
l
y
pr
e
d
i
c
t
i
o
n
o
f
m
e
nt
al
he
al
t
h
c
r
i
s
e
s
(H
as
s
an
Chi
gagur
e
)
339
re
p
e
r
c
us
s
i
ons
.
T
he
A
U
C
-
RO
C
w
i
l
l
h
e
l
p
i
l
l
us
t
r
a
t
e
t
h
e
t
r
a
d
e
-
offs
be
t
w
e
e
n
s
e
ns
i
t
i
v
i
t
y
a
nd
s
p
e
c
i
fi
c
i
t
y
a
c
r
os
s
di
f
fe
re
nt
t
h
re
s
ho
l
d
s
e
t
t
i
ngs
,
a
l
l
ow
i
n
g
fo
r
a
m
or
e
n
ua
nc
e
d
und
e
rs
t
a
nd
i
n
g
of
m
od
e
l
p
e
rf
or
m
a
n
c
e
.
T
a
b
l
e
1
.
M
e
t
hods
a
nd
t
e
c
hn
i
que
s
us
e
d
t
o
bu
i
l
d
m
od
e
l
s
M
o
d
e
l
F
e
a
t
u
re
s
e
l
e
c
t
i
o
n
m
e
t
h
o
d
N
u
m
b
e
r
o
f
f
e
a
t
u
re
s
Cl
a
s
s
i
m
b
a
l
a
n
c
e
1
E
F
S
A
8
N
o
n
e
2
E
F
S
A
8
S
M
O
T
E
3
E
F
S
A
8
T
o
m
e
k
l
i
n
k
s
4
RF
E
CV
5
N
o
n
e
5
RF
E
CV
5
S
M
O
T
E
6
RF
E
CV
5
T
o
m
e
k
l
i
n
k
s
3.
R
ES
U
LTS
A
N
D
D
I
S
C
U
S
S
I
O
N
3.
1
.
F
e
atu
r
e
s
c
on
tr
i
b
u
ti
n
g
to
me
n
ta
l
h
e
al
th
c
r
i
s
e
s
T
he
s
h
a
pl
e
y
a
d
di
t
i
ve
e
xp
l
a
n
a
t
i
ons
(
SHAP
)
s
u
m
m
a
ry
p
l
ot
i
n
F
i
gu
re
2
i
l
l
us
t
ra
t
e
s
t
h
e
re
l
a
t
i
v
e
i
m
por
t
a
nc
e
a
nd
d
i
re
c
t
i
ona
l
i
m
pa
c
t
of
e
a
c
h
fe
a
t
ur
e
on
t
h
e
m
ode
l
’s
pre
di
c
t
i
o
n
of
m
e
nt
a
l
h
e
a
l
t
h
c
r
i
s
i
s
ri
s
k
.
F
e
a
t
u
re
s
a
r
e
ra
nk
e
d
by
t
he
i
r
ove
r
a
l
l
i
nf
l
ue
n
c
e
,
w
i
t
h
“
D
a
ys
s
i
nc
e
l
a
s
t
drug
f
a
i
l
ure
,
”
“
A
ge
,
”
a
nd
“
W
e
e
ks
s
i
n
c
e
l
a
s
t
c
r
i
s
i
s
”
e
m
e
r
gi
ng
a
s
t
h
e
m
os
t
s
i
gni
fi
c
a
n
t
pr
e
di
c
t
ors
.
H
i
g
h
va
l
ue
s
for
t
h
e
s
e
f
e
a
t
ur
e
s
(
i
ndi
c
a
t
e
d
i
n
r
e
d)
a
re
a
s
s
oc
i
a
t
e
d
w
i
t
h
a
n
i
n
c
re
a
s
e
d
pr
e
di
c
t
e
d
ri
s
k
of
c
ri
s
i
s
,
w
h
i
l
e
l
ow
v
a
l
u
e
s
(b
l
u
e
)
t
e
nd
t
o
de
c
r
e
a
s
e
r
i
s
k.
N
ot
a
b
l
y,
va
ri
a
bl
e
s
s
u
c
h
a
s
“
N
o
t
di
a
gnos
e
d,
”
“
D
a
ys
s
i
nc
e
ri
s
k
of
s
ui
c
i
de
i
de
n
t
i
f
i
e
d
,
”
a
nd
“
W
e
e
ks
s
i
n
c
e
l
a
s
t
c
r
i
s
i
s
e
pi
s
od
e
”
a
l
s
o
c
ont
r
i
bu
t
e
m
e
a
ni
ng
ful
l
y
t
o
t
he
m
od
e
l
’s
ou
t
put
.
In
c
o
nt
r
a
s
t
,
f
e
a
t
ur
e
s
l
i
k
e
“
N
e
ve
r
ho
s
pi
t
a
l
i
z
e
d
,
”
“
N
e
v
e
r
n
e
e
de
d
M
H
A
,
”
a
nd
“
D
a
ys
s
i
n
c
e
ri
s
k
of
s
u
bs
t
a
n
c
e
m
i
s
us
e
i
d
e
nt
i
fi
e
d
”
h
a
ve
m
i
n
i
m
a
l
i
m
pa
c
t
,
a
s
re
fl
e
c
t
e
d
by
t
h
e
i
r
s
hor
t
S
H
A
P
ba
rs
.
T
he
pr
e
s
e
nt
f
i
ndi
ngs
i
ndi
c
a
t
e
t
h
a
t
t
h
e
m
os
t
pr
e
di
c
t
i
ve
f
e
a
t
u
re
s
for
m
e
n
t
a
l
h
e
a
l
t
h
c
ri
s
i
s
ri
s
k
c
l
os
e
l
y
a
l
i
gn
w
i
t
h
t
he
obs
e
rv
a
t
i
ons
m
a
de
b
y
[29]
,
S
pe
c
i
f
i
c
a
l
l
y,
fa
c
t
ors
s
u
c
h
a
s
t
h
e
hi
s
t
ori
c
a
l
s
e
v
e
ri
t
y
of
s
ym
pt
o
m
s
i
n
c
l
u
di
ng
t
he
t
ot
a
l
nu
m
b
e
r
of
c
r
i
s
i
s
e
p
i
s
ode
s
a
nd
t
he
du
ra
t
i
on
of
t
he
l
a
s
t
e
pi
s
od
e
a
l
ong
w
i
t
h
i
nt
e
ra
c
t
i
ons
w
i
t
h
t
h
e
hos
pi
t
a
l
,
l
i
ke
unp
l
a
n
ne
d
c
ont
a
c
t
s
,
m
i
s
s
e
d
a
ppo
i
nt
m
e
nt
s
,
or
a
r
e
c
e
nt
c
r
i
s
i
s
,
w
e
re
c
ru
c
i
a
l
pre
di
c
t
ors
.
A
dd
i
t
i
ona
l
l
y
,
p
a
t
i
e
nt
c
h
a
ra
c
t
e
ri
s
t
i
c
s
s
u
c
h
a
s
a
ge
,
i
ndi
v
i
du
a
l
ri
s
k
i
nd
i
c
e
s
,
a
nd
t
h
e
t
o
t
a
l
t
i
m
e
s
i
nc
e
t
he
pa
t
i
e
nt
’s
fi
rs
t
h
os
pi
t
a
l
v
i
s
i
t
s
i
gn
i
fi
c
a
nt
l
y
c
ont
ri
bu
t
e
d
t
o
t
h
e
m
od
e
l
’s
pre
d
i
c
t
i
v
e
pow
e
r
.
F
i
gure
2
.
T
he
s
ha
pl
e
y
a
dd
i
t
i
ve
e
x
pl
a
na
t
i
ons
s
um
m
a
r
y
p
l
ot
3.
2
.
M
od
e
l
1
w
i
th
EF
S
A
fe
a
tu
r
e
s
an
d
h
yp
e
r
p
a
r
am
e
te
r
tu
n
i
n
g
M
ode
l
1
,
s
how
n
i
n
T
a
b
l
e
2
c
ons
t
ruc
t
e
d
w
i
t
hou
t
c
l
a
s
s
i
m
b
a
l
a
nc
e
t
e
c
hni
que
s
,
us
e
d
E
F
S
A
-
s
e
l
e
c
t
e
d
fe
a
t
ure
s
,
m
ul
t
i
v
a
r
i
a
t
e
i
m
put
a
t
i
on
by
c
ha
i
n
e
d
e
qu
a
t
i
ons
(
M
ICE
)
-
i
m
p
ut
e
d
d
a
t
a
,
a
n
d
hy
pe
rp
a
r
a
m
e
t
e
r
t
uni
ng.
T
he
m
ode
l
a
c
h
i
e
ve
d
hi
ghe
r
a
c
c
ura
c
y
,
p
re
c
i
s
i
o
n,
re
c
a
l
l
,
a
nd
F
1
s
c
or
e
s
a
c
ros
s
c
l
a
s
s
i
f
i
e
r
s
,
r
e
fl
e
c
t
i
n
g
i
m
p
rove
d
di
s
c
ri
m
i
na
t
ory
pow
e
r
a
s
i
l
l
us
t
ra
t
e
d
i
n
T
a
b
l
e
2
.
E
F
S
A
i
de
n
t
i
f
i
e
d
c
r
i
t
i
c
a
l
fe
a
t
ur
e
s
,
w
hi
l
e
M
IC
E
e
ns
ur
e
d
robus
t
da
t
a
qu
a
l
i
t
y
.
T
he
s
e
s
t
r
a
t
e
gi
e
s
fe
a
t
u
re
s
e
l
e
c
t
i
on
,
i
m
pu
t
a
t
i
on
,
a
nd
t
uni
n
g
c
ol
l
e
c
t
i
v
e
l
y
ou
t
pe
rf
orm
e
d
b
a
s
e
l
i
n
e
s
,
unde
rs
c
ori
n
g
t
h
e
i
r
e
ff
i
c
a
c
y
i
n
e
nh
a
n
c
i
ng
pre
di
c
t
i
ons
de
s
p
i
t
e
om
i
t
t
i
ng
c
l
a
s
s
i
m
b
a
l
a
nc
e
a
dj
us
t
m
e
nt
s
.
T
he
c
om
b
i
n
a
t
i
on
of
E
F
S
A
a
nd
hyp
e
rp
a
ra
m
e
t
e
r
t
u
ni
ng
s
how
s
n
o
t
a
b
l
e
pe
rf
orm
a
n
c
e
w
i
t
h
S
V
M
a
nd
L
R
c
l
a
s
s
i
f
i
e
rs
a
c
h
i
e
v
i
ng
0.
88
a
c
c
ur
a
c
y
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2722
-
3221
Com
pu
t
S
c
i
Inf
T
e
c
h
nol
,
V
o
l
.
6
,
N
o
.
3
,
N
ov
e
m
be
r
20
25
:
335
-
345
340
T
a
b
l
e
2
.
M
od
e
l
1
pe
r
form
a
n
c
e
w
i
t
h
E
F
S
A
f
e
a
t
ure
s
a
n
d
h
ype
r
pa
ra
m
e
t
e
r
t
uni
n
g
P
e
rfo
rm
a
n
c
e
m
e
t
ri
c
s
Cl
a
s
s
i
fi
e
rs
Bo
o
s
t
i
n
g
c
l
a
s
s
i
fi
e
rs
S
V
M
DT
LR
K
-
NN
X
G
Bo
o
s
t
A
d
a
Bo
o
s
t
A
c
c
u
ra
c
y
0
.
8
8
0
.
9
1
0
.
8
8
0
.
8
8
0
.
9
2
0
.
9
2
P
re
c
i
s
i
o
n
0
.
8
9
0
.
9
3
0
.
8
6
0
.
8
6
0
.
9
1
0
.
9
1
Re
c
a
l
l
0
.
9
6
0
.
9
4
0
.
9
6
0
.
9
8
0
.
9
7
0
.
9
7
F
1
S
c
o
re
0
.
9
2
0
.
9
3
0
.
9
2
0
.
9
1
0
.
9
4
0
.
9
4
Ro
c
A
u
c
0
.
8
6
0
.
9
0
0
.
8
6
0
.
8
3
0
.
8
9
0
.
8
9
K
a
p
p
a
0
.
7
5
0
.
8
0
0
.
7
5
0
.
7
1
0
.
8
1
0
.
8
1
G
e
o
m
e
t
ri
c
m
e
a
n
0
.
8
5
0
.
9
0
0
.
8
5
0
.
8
2
0
.
8
9
0
.
8
9
Ba
l
a
n
c
e
d
A
c
c
0
.
8
6
0
.
9
0
0
.
8
6
0
.
8
3
0
.
8
9
0
.
8
9
T
i
m
e
(S
e
c
)
0
.
0
5
0
.
0
0
6
0
.
0
5
0
.
0
4
0
.
4
9
0
.
2
8
CP
U
(K
B)
0
0
0
0
0
0
3.
3
.
M
od
e
l
2
w
i
th
S
M
O
TE
,
E
F
S
A
f
e
atu
r
e
s
,
an
d
h
yp
e
r
p
a
r
ame
te
r
tu
n
i
n
g
M
ode
l
2
us
e
d
S
M
O
T
E
t
o
ba
l
a
n
c
e
c
l
a
s
s
e
s
,
a
l
ongs
i
de
E
F
S
A
fe
a
t
ur
e
s
e
l
e
c
t
i
on
a
nd
hy
pe
rp
a
r
a
m
e
t
e
r
t
uni
n
g.
R
e
s
ul
t
s
,
a
s
s
how
n
i
n
T
a
bl
e
3,
s
how
c
l
a
s
s
i
f
i
e
rs
a
c
hi
e
ve
d
s
t
rong
a
c
c
ur
a
c
y
,
pr
e
c
i
s
i
on
,
a
nd
r
e
c
a
l
l
w
i
t
h
ba
l
a
nc
e
d
d
a
t
a
.
H
ype
rpa
r
a
m
e
t
e
r
t
un
i
ng
a
ga
i
n
pr
ove
d
e
f
fe
c
t
i
v
e
.
S
M
O
T
E
’s
s
y
nt
h
e
t
i
c
i
ns
t
a
nc
e
s
ri
s
k
e
d
m
i
s
re
pre
s
e
nt
i
ng
d
a
t
a
di
s
t
r
i
bu
t
i
ons
,
pot
e
n
t
i
a
l
l
y
l
i
m
i
t
i
ng
ge
n
e
ra
l
i
z
a
b
i
l
i
t
y
,
but
e
na
bl
e
d
robus
t
m
i
nori
t
y
-
c
l
a
s
s
ha
ndl
i
ng
.
T
he
M
od
e
l
1
vs
.
2
c
o
m
pa
ri
s
on
hi
gh
l
i
g
ht
s
t
ra
d
e
-
offs
:
c
l
a
s
s
ba
l
a
n
c
i
n
g
i
m
pro
ve
d
fa
i
rn
e
s
s
,
w
h
i
l
e
t
a
i
l
or
e
d
a
t
t
ri
b
ut
e
s
op
t
i
m
i
z
e
d
pe
rfor
m
a
n
c
e
.
D
e
s
pi
t
e
m
i
nor
m
e
t
r
i
c
di
ps
,
M
ode
l
2’s
i
n
t
e
gra
t
e
d
a
ppro
a
c
h
S
M
O
T
E
,
E
F
S
A
,
a
nd
t
un
i
ng
un
de
rs
c
ore
s
t
he
va
l
ue
of
b
a
l
a
nc
i
ng
t
e
c
hni
qu
e
s
i
n
m
e
nt
a
l
he
a
l
t
h
pr
e
di
c
t
i
on
,
e
v
e
n
w
i
t
h
i
nhe
r
e
nt
c
o
m
pro
m
i
s
e
s
,
a
nd
t
h
i
s
va
l
i
d
a
t
e
s
[9],
t
h
a
t
S
M
O
T
E
do
e
s
not
i
nc
r
e
a
s
e
a
c
c
u
ra
c
y
i
n
h
e
a
l
t
h
pre
di
c
t
i
ons
s
i
nc
e
s
i
z
e
s
do
e
s
n
ot
c
ou
nt
.
T
a
b
l
e
3
.
M
od
e
l
2
pe
r
form
a
n
c
e
w
i
t
h
S
M
O
T
E
,
E
F
S
A
f
e
a
t
ure
s
,
a
nd
hy
pe
rp
a
ra
m
e
t
e
r
t
uni
ng
P
e
rfo
rm
a
n
c
e
m
e
t
ri
c
s
Cl
a
s
s
i
fi
e
rs
Bo
o
s
t
i
n
g
c
l
a
s
s
i
fi
e
rs
S
V
M
DT
LR
K
-
NN
X
G
Bo
o
s
t
A
d
a
Bo
o
s
t
A
c
c
u
ra
c
y
0
.
8
9
0
.
9
0
0
.
8
9
0
.
8
7
0
.
9
2
0
.
9
1
P
re
c
i
s
i
o
n
0
.
8
9
0
.
9
2
0
.
8
6
0
.
8
8
0
.
9
3
0
.
9
2
Re
c
a
l
l
0
.
9
6
0
.
9
2
0
.
9
6
0
.
9
2
0
.
9
6
0
.
9
3
F
1
S
c
o
re
0
.
9
2
0
.
9
2
0
.
9
2
0
.
9
0
0
.
9
4
0
.
9
3
Ro
c
A
u
c
0
.
8
6
0
.
8
9
0
.
8
6
0
.
8
4
0
.
9
1
0
.
8
9
K
a
p
p
a
0
.
7
5
0
.
7
7
0
.
7
5
0
.
6
9
0
.
8
2
0
.
7
9
G
e
o
m
e
t
ri
c
m
e
a
n
0
.
8
5
0
.
8
8
0
.
8
5
0
.
8
4
0
.
9
1
0
.
8
9
Ba
l
a
n
c
e
d
A
c
c
0
.
8
6
0
.
8
8
0
.
8
6
0
.
8
4
0
.
9
1
0
.
8
9
T
i
m
e
(S
e
c
)
0
.
0
7
0
.
0
0
6
0
.
0
5
0
.
0
2
0
.
2
1
0
.
5
6
CP
U
(K
B)
0
0
0
0
264
0
3.
4
.
M
od
e
l
w
i
th
tom
e
k
l
i
n
k
s
,
E
F
S
A
fe
atu
r
e
s
,
an
d
h
yp
e
r
p
ar
am
e
te
r
tu
n
i
n
g
M
ode
l
3
us
e
s
a
d
i
s
t
i
nc
t
c
l
a
s
s
i
m
b
a
l
a
nc
e
t
e
c
hn
i
qu
e
w
i
t
h
T
om
e
k
l
i
n
ks
.
Boos
t
i
ng
c
l
a
s
s
i
fi
e
rs
,
e
s
p
e
c
i
a
l
l
y
X
G
Boos
t
,
pe
rf
orm
w
e
l
l
,
a
c
h
i
e
v
i
ng
93%
a
c
c
ur
a
c
y
a
n
d
91
%
ba
l
a
n
c
e
d
a
c
c
ura
c
y
a
s
pr
e
s
e
n
t
e
d
i
n
T
a
bl
e
4
.
Com
p
a
re
d
t
o
pr
e
vi
o
us
m
od
e
l
s
,
M
o
de
l
3
s
h
ow
s
s
l
i
gh
t
l
y
h
i
gh
e
r
a
c
c
u
r
a
c
y
t
h
a
n
M
ode
l
2.
T
hou
gh
T
o
m
e
k
l
i
n
ks
don’t
fu
l
l
y
b
a
l
a
nc
e
c
l
a
s
s
e
s
,
t
h
e
y
hi
gh
l
i
ght
t
he
i
m
por
t
a
n
c
e
o
f
t
a
i
l
or
e
d
i
m
ba
l
a
n
c
e
h
a
nd
l
i
ng
,
d
e
s
pi
t
e
pot
e
n
t
i
a
l
i
nfor
m
a
t
i
on
l
os
s
i
n
t
h
e
“
N
o
”
c
l
a
s
s
a
f
fe
c
t
i
ng
ge
n
e
ra
l
i
z
a
t
i
on
.
T
a
b
l
e
4.
M
od
e
l
3
pe
r
form
a
n
c
e
w
i
t
h
T
o
m
e
k
l
i
nks
,
E
F
S
A
fe
a
t
ure
s
,
a
n
d
h
ype
r
pa
r
a
m
e
t
e
r
t
un
i
ng
P
e
rfo
rm
a
n
c
e
m
e
t
ri
c
s
Cl
a
s
s
i
fi
e
rs
Bo
o
s
t
i
n
g
c
l
a
s
s
i
fi
e
rs
S
V
M
DT
LR
K
-
NN
X
G
Bo
o
s
t
A
d
a
Bo
o
s
t
A
c
c
u
ra
c
y
0
.
8
9
0
.
9
2
0
.
8
9
0
.
8
7
0
.
9
3
0
.
9
2
P
re
c
i
s
i
o
n
0
.
8
9
0
.
9
3
0
.
8
6
0
.
8
7
0
.
9
3
0
.
9
1
Re
c
a
l
l
0
.
9
6
0
.
9
5
0
.
9
6
0
.
9
6
0
.
9
6
0
.
9
8
F
1
S
c
o
re
0
.
9
2
0
.
9
4
0
.
9
2
0
.
9
1
0
.
9
4
0
.
9
4
Ro
c
A
u
c
0
.
8
6
0
.
9
0
0
.
8
6
0
.
8
4
0
.
9
1
0
.
8
9
K
a
p
p
a
0
.
7
5
0
.
8
1
0
.
7
5
0
.
7
1
0
.
8
3
0
.
8
2
G
e
o
m
e
t
ri
c
m
e
a
n
0
.
8
5
0
.
9
0
0
.
8
5
0
.
8
2
0
.
9
1
0
.
8
9
Ba
l
a
n
c
e
d
A
c
c
0
.
8
6
0
.
9
0
0
.
8
6
0
.
8
4
0
.
9
1
0
.
8
9
T
i
m
e
(S
e
c
)
0
.
0
9
0
.
0
1
0
.
0
9
0
.
0
4
0
.
2
7
0
.
2
8
CP
U
(K
B)
0
0
0
0
0
0
3.
5
.
M
od
e
l
w
i
th
R
F
EC
V
fe
atu
r
e
s
an
d
h
yp
e
r
p
a
r
am
e
te
r
tu
n
i
n
g
M
ode
l
4,
l
i
ke
M
o
de
l
1
,
do
e
s
not
a
p
pl
y
a
ny
c
l
a
s
s
i
m
ba
l
a
n
c
e
t
e
c
hni
que
a
nd
us
e
s
h
ype
r
pa
r
a
m
e
t
e
r
t
uni
n
g,
bu
t
di
s
t
i
ng
ui
s
he
s
i
t
s
e
l
f
by
e
m
p
l
oy
i
ng
RF
E
CV
for
f
e
a
t
ure
s
e
l
e
c
t
i
on
.
T
h
i
s
a
pp
roa
c
h
i
nve
s
t
i
ga
t
e
s
t
he
Evaluation Warning : The document was created with Spire.PDF for Python.
Com
pu
t
S
c
i
Inf
T
e
c
h
nol
IS
S
N
:
2722
-
3221
A
m
a
c
hi
n
e
l
e
ar
ni
ng
ap
pr
oac
h
f
or
e
ar
l
y
pr
e
d
i
c
t
i
o
n
o
f
m
e
nt
al
he
al
t
h
c
r
i
s
e
s
(H
as
s
an
Chi
gagur
e
)
341
e
ffe
c
t
of
opt
i
m
a
l
f
e
a
t
ure
s
e
l
e
c
t
i
on
on
m
od
e
l
p
e
rfor
m
a
n
c
e
,
a
s
s
how
n
i
n
T
a
bl
e
5
.
M
ode
l
4
a
c
h
i
e
ve
s
hi
gh
a
c
c
ura
c
y
,
pre
c
i
s
i
on,
a
nd
r
e
c
a
l
l
(a
c
c
u
ra
c
i
e
s
fr
om
0
.
89
t
o
0.
9
3)
a
nd
t
hi
s
i
s
c
ons
i
s
t
e
nt
w
i
t
h
[2
7]
t
h
a
t
RF
E
CV
yi
e
l
ds
gr
e
a
t
e
r
a
c
c
ur
a
c
y
i
n
m
e
n
t
a
l
h
e
a
l
t
h
pre
di
c
t
i
o
ns
.
W
h
i
l
e
RF
E
CV
m
a
y
r
e
du
c
e
fe
a
t
ur
e
s
p
a
c
e
a
nd
r
i
s
k
i
nfor
m
a
t
i
on
l
os
s
,
i
t
s
re
f
i
ne
d
s
e
l
e
c
t
i
on
c
l
e
a
rl
y
boos
t
s
ov
e
ra
l
l
m
ode
l
e
ff
e
c
t
i
v
e
ne
s
s
for
m
e
nt
a
l
h
e
a
l
t
h
pre
di
c
t
i
on.
T
a
b
l
e
5
.
M
od
e
l
4
pe
r
form
a
n
c
e
w
i
t
h
RF
E
CV
fe
a
t
ur
e
s
a
nd
hyp
e
rpa
r
a
m
e
t
e
r
t
un
i
ng
P
e
rfo
rm
a
n
c
e
m
e
t
ri
c
s
Cl
a
s
s
i
fi
e
rs
Bo
o
s
t
i
n
g
c
l
a
s
s
i
fi
e
rs
S
V
M
DT
LR
K
-
NN
X
G
Bo
o
s
t
A
d
a
Bo
o
s
t
A
c
c
u
ra
c
y
0
.
9
3
0
.
9
3
0
.
8
9
0
.
9
2
0
.
9
3
0
.
9
2
P
re
c
i
s
i
o
n
0
.
9
2
0
.
9
2
0
.
8
6
0
.
9
2
0
.
9
2
0
.
9
1
Re
c
a
l
l
0
.
9
7
0
.
9
7
0
.
9
6
0
.
9
6
0
.
9
7
0
.
9
8
F
1
S
c
o
re
0
.
9
5
0
.
9
5
0
.
9
2
0
.
9
4
0
.
9
5
0
.
9
4
Ro
c
A
u
c
0
.
9
1
0
.
9
1
0
.
8
6
0
.
9
0
0
.
9
1
0
.
9
0
K
a
p
p
a
0
.
8
3
0
.
8
3
0
.
7
5
0
.
8
2
0
.
8
3
0
.
8
2
G
e
o
m
e
t
ri
c
m
e
a
n
0
.
9
1
0
.
9
1
0
.
8
5
0
.
9
0
0
.
9
1
0
.
8
9
Ba
l
a
n
c
e
d
A
c
c
0
.
9
1
0
.
9
1
0
.
8
6
0
.
9
0
0
.
9
1
0
.
9
0
T
i
m
e
(S
e
c
)
0
.
0
3
0
.
0
0
4
0
.
0
0
7
0
.
0
2
0
.
0
6
0
.
2
5
P
U
(K
B)
0
0
0
0
0
0
3.
6
.
M
od
e
l
w
i
th
S
M
O
TE
,
R
F
EC
V
fe
atu
r
e
s
,
an
d
h
yp
e
r
p
a
r
ame
te
r
tu
n
i
n
g
M
ode
l
5
m
i
rr
ors
M
o
de
l
2
by
us
i
ng
S
M
O
T
E
a
nd
hyp
e
rpa
r
a
m
e
t
e
r
t
un
i
ng
bu
t
furt
h
e
r
n
a
rrow
s
i
t
s
fo
c
us
t
o
j
us
t
t
hre
e
k
e
y
RF
E
CV
-
s
e
l
e
c
t
e
d
fe
a
t
ur
e
s
.
T
h
i
s
s
t
r
e
a
m
l
i
n
e
d
a
ppro
a
c
h
hi
ghl
i
ght
s
t
h
e
i
m
p
a
c
t
of
S
M
O
T
E
re
s
a
m
pl
i
ng
a
nd
a
c
o
m
pa
c
t
fe
a
t
ure
s
e
t
.
A
s
s
h
ow
n
i
n
T
a
b
l
e
6
,
m
o
s
t
c
l
a
s
s
i
fi
e
rs
i
m
prov
e
d
,
w
i
t
h
S
V
M
,
D
T
,
K
-
N
N
,
a
nd
X
G
Bo
os
t
a
c
h
i
e
v
i
ng
up
t
o
0.
9
3
a
c
c
ura
c
y
.
W
hi
l
e
t
hi
s
m
e
t
hod
boos
t
s
pe
r
form
a
n
c
e
ove
r
b
a
s
e
l
i
ne
s
,
t
he
l
i
m
i
t
e
d
f
e
a
t
ur
e
s
a
nd
s
y
nt
h
e
t
i
c
da
t
a
m
a
y
l
i
m
i
t
a
d
a
pt
a
bi
l
i
t
y
a
nd
ge
n
e
ra
l
i
z
a
b
i
l
i
t
y
.
T
a
b
l
e
6
.
M
od
e
l
5
pe
r
form
a
n
c
e
w
i
t
h
S
M
O
T
E
,
RF
E
CV
fe
a
t
ur
e
s
a
nd
h
ype
r
pa
r
a
m
e
t
e
r
t
un
i
ng
P
e
rfo
rm
a
n
c
e
m
e
t
ri
c
s
Cl
a
s
s
i
fi
e
rs
Bo
o
s
t
i
n
g
c
l
a
s
s
i
fi
e
rs
S
V
M
DT
LR
K
-
NN
X
G
Bo
o
s
t
A
d
a
Bo
o
s
t
A
c
c
u
ra
c
y
0
.
9
3
0
.
9
3
0
.
8
9
0
.
9
2
0
.
9
3
0
.
9
2
P
re
c
i
s
i
o
n
0
.
9
2
0
.
9
2
0
.
8
6
0
.
9
2
0
.
9
2
0
.
9
1
Re
c
a
l
l
0
.
9
7
0
.
9
7
0
.
9
6
0
.
9
6
0
.
9
7
0
.
9
8
F
1
S
c
o
re
0
.
9
5
0
.
9
5
0
.
9
2
0
.
9
4
0
.
9
5
0
.
9
4
Ro
c
A
u
c
0
.
9
1
0
.
9
1
0
.
8
6
0
.
9
0
0
.
9
1
0
.
9
0
K
a
p
p
a
0
.
8
3
0
.
8
3
0
.
7
5
0
.
8
2
0
.
8
3
0
.
8
2
G
e
o
m
e
t
ri
c
m
e
a
n
0
.
9
1
0
.
9
1
0
.
8
5
0
.
9
0
0
.
9
1
0
.
8
9
Ba
l
a
n
c
e
d
A
c
c
0
.
9
1
0
.
9
1
0
.
8
6
0
.
9
0
0
.
9
1
0
.
9
0
T
i
m
e
(S
e
c
)
0
.
0
3
0
.
0
0
4
0
.
0
0
7
0
.
0
2
0
.
0
6
0
.
2
5
CP
U
(K
B)
0
0
0
0
0
0
3.
7
.
M
od
e
l
w
i
th
tom
e
k
l
i
n
k
s
,
R
F
EC
V
fe
atu
r
e
s
an
d
h
yp
e
r
p
ar
ame
te
r
tu
n
i
n
g
M
ode
l
6
i
n
T
a
bl
e
7
a
do
pt
s
t
he
s
a
m
e
s
t
r
a
t
e
gy
a
s
M
od
e
l
3,
us
i
ng
T
om
e
k
l
i
nks
for
c
l
a
s
s
i
m
ba
l
a
n
c
e
,
hype
rp
a
ra
m
e
t
e
r
t
un
i
ng
,
a
nd
RF
E
CV
for
fe
a
t
u
re
s
e
l
e
c
t
i
on.
It
s
k
e
y
di
s
t
i
n
c
t
i
on
l
i
e
s
i
n
t
he
s
pe
c
i
f
i
c
f
e
a
t
ur
e
s
c
hos
e
n
.
T
h
e
c
o
m
bi
n
a
t
i
on
of
T
o
m
e
k
l
i
nks
a
nd
RF
E
CV
-
s
e
l
e
c
t
e
d
fe
a
t
u
re
s
l
e
a
ds
t
o
no
t
a
b
l
e
pe
rfor
m
a
n
c
e
g
a
i
ns
,
w
i
t
h
S
V
M
a
nd
D
T
c
l
a
s
s
i
f
i
e
rs
a
c
hi
e
v
i
ng
0.
93
a
c
c
ura
c
y
.
M
od
e
l
6
s
how
s
c
l
e
a
r
i
m
pro
ve
m
e
n
t
s
a
c
ros
s
a
c
c
ur
a
c
y,
pre
c
i
s
i
on
,
a
nd
r
e
c
a
l
l
,
hi
ghl
i
ght
i
ng
t
h
e
s
yne
rgy
o
f
c
l
a
s
s
i
m
ba
l
a
nc
e
h
a
nd
l
i
ng
a
n
d
t
a
rg
e
t
e
d
f
e
a
t
ure
s
e
l
e
c
t
i
o
n.
T
a
b
l
e
8
s
how
s
t
he
re
s
u
l
t
s
of
t
he
a
l
gor
i
t
h
m
s
w
i
t
h
T
om
e
k
l
i
nks
,
RF
E
CV
fe
a
t
ur
e
s
,
a
n
d
h
ype
r
pa
r
a
m
e
t
e
r
t
un
i
ng
.
T
a
b
l
e
7
.
M
od
e
l
6
pe
r
fo
rm
a
n
c
e
w
i
t
h
T
o
m
e
k
l
i
nks
,
RF
E
CV
f
e
a
t
ur
e
s
,
a
n
d
hy
pe
rp
a
r
a
m
e
t
e
r
t
un
i
ng
P
e
rfo
rm
a
n
c
e
m
e
t
ri
c
s
Cl
a
s
s
i
fi
e
rs
Bo
o
s
t
i
n
g
c
l
a
s
s
i
fi
e
rs
S
V
M
DT
LR
K
-
NN
X
G
Bo
o
s
t
A
d
a
Bo
o
s
t
A
c
c
u
ra
c
y
0
.
9
3
0
.
9
3
0
.
8
9
0
.
9
2
0
.
9
3
0
.
9
2
P
re
c
i
s
i
o
n
0
.
9
3
0
.
9
2
0
.
8
6
0
.
9
2
0
.
9
2
0
.
9
1
Re
c
a
l
l
0
.
9
7
0
.
9
7
0
.
9
6
0
.
9
6
0
.
9
7
0
.
9
8
F
1
S
c
o
re
0
.
9
5
0
.
9
5
0
.
9
2
0
.
9
4
0
.
9
5
0
.
9
4
Ro
c
A
u
c
0
.
9
1
0
.
9
1
0
.
8
6
0
.
9
0
0
.
9
1
0
.
9
0
K
a
p
p
a
0
.
8
3
0
.
3
0
.
7
5
0
.
8
2
0
.
8
3
0
.
8
2
G
e
o
m
e
t
ri
c
m
e
a
n
0
.
9
1
0
.
9
1
0
.
8
5
0
.
9
0
0
.
9
1
0
.
8
9
Ba
l
a
n
c
e
d
A
c
c
0
.
9
1
0
.
9
1
0
.
8
6
0
.
9
0
0
.
9
1
0
.
9
0
T
i
m
e
(S
e
c
)
0
.
0
4
0
.
0
0
3
0
.
0
0
7
0
.
0
2
0
.
0
5
0
.
2
5
CP
U
(K
B)
0
0
0
0
0
0
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2722
-
3221
Com
pu
t
S
c
i
Inf
T
e
c
h
nol
,
V
o
l
.
6
,
N
o
.
3
,
N
ov
e
m
be
r
20
25
:
335
-
345
342
A
c
o
m
pr
e
h
e
ns
i
v
e
e
v
a
l
u
a
t
i
on
of
v
a
ri
ous
m
od
e
l
c
o
nfi
g
ura
t
i
o
ns
hi
ghl
i
ght
s
t
he
i
m
por
t
a
nc
e
of
t
a
i
l
o
re
d
m
a
c
hi
n
e
l
e
a
rn
i
ng
t
e
c
hn
i
qu
e
s
for
m
e
n
t
a
l
h
e
a
l
t
h
pr
e
di
c
t
i
on
.
A
pproa
c
he
s
s
uc
h
a
s
S
M
O
T
E
,
T
om
e
k
l
i
n
ks
,
E
F
S
A
,
a
nd
RF
E
CV
a
l
l
c
on
t
ri
b
ut
e
d
t
o
i
m
prov
e
d
p
e
rfor
m
a
nc
e
,
w
i
t
h
hyp
e
rpa
r
a
m
e
t
e
r
t
un
i
ng
fur
t
he
r
boos
t
i
ng
r
e
s
ul
t
s
.
S
V
M
a
n
d
D
T
c
l
a
s
s
i
fi
e
rs
c
ons
i
s
t
e
n
t
l
y
a
c
hi
e
ve
d
hi
gh
a
c
c
ur
a
c
y
,
pe
a
ki
ng
a
t
93%
.
H
ow
e
ve
r
,
po
t
e
n
t
i
a
l
dr
a
w
b
a
c
ks
i
nc
l
ude
i
n
form
a
t
i
on
l
os
s
fr
om
fe
a
t
ur
e
s
e
l
e
c
t
i
on
a
nd
l
i
m
i
t
a
t
i
ons
of
i
m
ba
l
a
n
c
e
t
e
c
hni
qu
e
s
.
Cros
s
-
v
a
l
i
da
t
i
on
prove
d
e
s
s
e
n
t
i
a
l
fo
r
a
s
s
e
s
s
i
ng
m
ode
l
rob
us
t
ne
s
s
a
nd
g
e
n
e
ra
l
i
z
a
b
i
l
i
t
y
.
X
G
Boos
t
s
t
ood
out
a
s
t
h
e
t
op
p
e
rfor
m
e
r
a
c
ros
s
m
os
t
m
e
t
r
i
c
s
,
w
hi
l
e
A
da
B
oos
t
e
x
c
e
l
l
e
d
i
n
M
od
e
l
4.
S
V
M
a
nd
D
T
a
l
s
o
de
l
i
v
e
re
d
s
t
ron
g
r
e
s
ul
t
s
,
n
e
a
r
l
y
m
a
t
c
h
i
ng
X
G
Boos
t
’s
prof
i
c
i
e
n
c
y
.
3.
8
.
En
s
e
mb
l
e
l
e
ar
n
i
n
g
T
he
r
e
s
ul
t
s
ob
t
a
i
ne
d
w
h
e
n
a
ppl
y
i
ng
t
he
t
h
re
e
c
o
m
m
o
nl
y
us
e
d
E
L
m
e
t
hods
—
b
a
ggi
ng,
b
oos
t
i
ng,
a
nd
s
t
a
c
k
i
ng
a
re
pr
e
s
e
n
t
e
d
i
n
T
a
bl
e
8
.
E
ns
e
m
b
l
e
m
e
t
hods
a
r
e
m
a
c
hi
n
e
l
e
a
rn
i
ng
t
e
c
hni
q
ue
s
t
h
a
t
c
om
bi
n
e
t
h
e
pre
di
c
t
i
ons
from
m
ul
t
i
p
l
e
m
o
de
l
s
t
o
i
m
pr
ove
ov
e
ra
l
l
a
c
c
u
ra
c
y
a
n
d
p
e
rfor
m
a
nc
e
.
B
a
ggi
n
g
fo
c
us
e
s
on
re
d
uc
i
ng
va
ri
a
nc
e
b
y
a
g
gre
g
a
t
i
ng
re
s
ul
t
s
fro
m
s
e
ve
r
a
l
m
od
e
l
s
t
ra
i
n
e
d
on
di
ff
e
r
e
nt
da
t
a
s
a
m
p
l
e
s
,
w
hi
l
e
boos
t
i
ng
a
i
m
s
t
o
re
duc
e
b
i
a
s
b
y
s
e
qu
e
nt
i
a
l
l
y
c
or
re
c
t
i
ng
e
rrors
fro
m
pr
e
v
i
ous
m
ode
l
s
.
S
t
a
c
ki
n
g
fur
t
he
r
e
nh
a
n
c
e
s
p
e
rfor
m
a
nc
e
by
t
ra
i
ni
ng
a
f
i
na
l
m
o
de
l
t
o
b
e
s
t
c
o
m
b
i
ne
t
he
o
ut
pu
t
s
of
v
a
ri
ous
i
ndi
vi
du
a
l
m
od
e
l
s
,
p
rovi
di
ng
a
c
om
p
re
h
e
ns
i
v
e
a
ppr
oa
c
h
t
o
pre
d
i
c
t
i
v
e
m
od
e
l
l
i
n
g.
3.
8
.
1
.
Bag
gi
n
g
T
he
ba
ggi
ng
E
L
m
e
t
hod
w
a
s
i
m
pl
e
m
e
nt
e
d
u
s
i
ng
t
hre
e
di
v
e
rs
e
c
l
a
s
s
i
fi
e
r
s
:
D
T
,
S
V
M
,
a
nd
K
-
N
N
.
E
a
c
h
c
l
a
s
s
i
fi
e
r
i
ndi
vi
dua
l
l
y
pe
rfo
rm
e
d
w
e
l
l
,
a
n
d
t
he
i
r
a
l
gori
t
hm
i
c
di
ve
rs
i
t
y
e
nha
nc
e
d
t
he
e
n
s
e
m
bl
e
’
s
ove
ra
l
l
rob
us
t
ne
s
s
a
nd
a
c
c
u
ra
c
y.
A
s
s
ho
w
n
i
n
T
a
bl
e
8
,
t
he
ba
ggi
ng
m
ode
l
a
c
hi
e
ve
d
a
n
a
c
c
ura
c
y
o
f
0.
927
,
pre
c
i
s
i
on
of
0.
9
36,
a
nd
re
c
a
l
l
of
0.
955.
T
he
F
1
s
c
ore
re
a
c
he
d
0.
945,
a
nd
t
he
RO
C
A
U
C
w
a
s
0.
91
3,
i
ndi
c
a
t
i
ng
s
t
ron
g
di
s
c
ri
m
i
na
t
ory
pow
e
r.
T
he
ka
ppa
c
oe
f
fi
c
i
e
nt
(0.
83
5),
ge
om
e
t
ri
c
m
e
a
n
(
0.
912
),
a
nd
ba
l
a
nc
e
d
a
c
c
ura
c
y
(0.
9
13)
f
urt
he
r
c
onfi
rm
m
o
de
l
s
t
a
bi
l
i
t
y.
W
i
t
h
a
t
ra
i
ni
ng
t
i
m
e
of
j
u
s
t
0.
1
51
s
e
c
ond
s
a
n
d
l
ow
m
e
m
ory
u
s
a
ge
,
t
hi
s
ba
g
gi
ng
-
S
M
O
T
E
a
pp
roa
c
h
p
rove
s
bot
h
e
f
fe
c
t
i
ve
a
nd
c
om
put
a
t
i
ona
l
l
y
e
ffi
c
i
e
nt
fo
r
m
e
nt
a
l
di
s
orde
r
pre
di
c
t
i
on.
3.
8
.
2
.
Boos
t
i
n
g
T
he
boos
t
i
ng
E
L
m
e
t
hod
w
a
s
i
m
pl
e
m
e
nt
e
d
us
i
ng
D
T
a
nd
A
da
Boos
t
,
r
e
s
ul
t
i
ng
i
n
ou
t
s
t
a
ndi
n
g
pe
rfor
m
a
nc
e
t
h
a
t
s
urp
a
s
s
e
d
ot
he
r
c
l
a
s
s
i
fi
e
r
c
o
m
bi
na
t
i
ons
.
D
T
s
e
rv
e
d
a
s
a
s
t
rong
b
a
s
e
m
ode
l
,
w
h
i
l
e
A
da
B
oos
t
i
t
e
r
a
t
i
v
e
l
y
c
orre
c
t
e
d
i
t
s
e
rro
rs
,
l
e
a
di
ng
t
o
e
n
ha
n
c
e
d
r
e
s
ul
t
s
A
da
Boos
t
a
c
h
i
e
ve
d
92
%
a
c
c
u
ra
c
y
a
nd
D
T
93%
.
T
hi
s
a
ppro
a
c
h
l
e
ve
r
a
g
e
d
bo
t
h
m
ode
l
d
i
ve
rs
i
t
y
a
nd
boos
t
i
ng
s
yne
rgy
for
gr
e
a
t
e
r
re
s
i
l
i
e
nc
e
.
T
h
e
m
e
t
hod
w
a
s
t
e
s
t
e
d
w
i
t
h
bot
h
S
M
O
T
E
a
nd
T
o
m
e
k
l
i
nks
for
c
l
a
s
s
i
m
b
a
l
a
n
c
e
,
e
a
c
h
y
i
e
l
di
ng
d
i
s
t
i
nc
t
r
e
s
ul
t
s
w
hi
l
e
t
h
e
RO
C
c
urve
r
e
m
a
i
n
e
d
c
ons
i
s
t
e
n
t
l
y
hi
gh
(
0.
91
).
O
ve
r
a
l
l
,
t
h
e
bo
os
t
i
ng
e
ns
e
m
b
l
e
m
e
t
h
od
d
e
m
ons
t
r
a
t
e
d
s
up
e
ri
or
pre
c
i
s
i
on
,
re
c
a
l
l
,
a
nd
robus
t
d
i
s
c
ri
m
i
na
t
ory
pow
e
r
fo
r
m
e
n
t
a
l
di
s
orde
r
pr
e
di
c
t
i
on
.
3.
8
.
3
.
S
tac
k
i
n
g
T
he
s
t
a
c
k
i
ng
E
L
m
e
t
hod
i
n
t
e
g
ra
t
e
s
a
r
a
nge
of
h
i
gh
-
pe
rfo
r
m
i
ng
b
a
s
e
c
l
a
s
s
i
fi
e
rs
-
D
T
,
S
V
M
,
K
-
N
N
,
a
nd
X
G
Boos
t
-
e
a
c
h
c
on
t
ri
but
i
ng
uni
q
ue
a
l
gor
i
t
h
m
i
c
pe
r
s
pe
c
t
i
v
e
s
a
nd
s
t
re
n
gt
hs
.
B
y
c
om
bi
n
i
ng
t
h
e
i
r
pre
di
c
t
i
ons
t
hr
ough
a
m
e
t
a
-
l
e
a
rne
r,
t
h
e
s
t
a
c
k
i
ng
e
ns
e
m
b
l
e
e
ff
e
c
t
i
ve
l
y
l
e
v
e
r
a
ge
s
t
h
i
s
di
v
e
rs
i
t
y
t
o
e
nh
a
n
c
e
ove
ra
l
l
pr
e
di
c
t
i
ve
p
e
rfor
m
a
n
c
e
.
Re
s
u
l
t
s
,
d
e
t
a
i
l
e
d
i
n
T
a
b
l
e
8
,
hi
ghl
i
gh
t
t
he
m
e
t
h
od’s
a
bi
l
i
t
y
t
o
c
a
pi
t
a
l
i
z
e
on
t
h
e
s
t
re
ng
t
hs
of
e
a
c
h
c
l
a
s
s
i
fi
e
r.
A
c
hi
e
vi
ng
a
robus
t
RO
C
A
U
C
of
91
%,
t
he
s
t
a
c
k
i
ng
a
ppro
a
c
h
d
e
m
ons
t
r
a
t
e
s
s
t
rong,
r
e
l
i
a
b
l
e
m
o
de
l
p
e
rfor
m
a
nc
e
,
c
om
p
a
ra
bl
e
t
o
t
he
Bo
os
t
i
ng
E
L
m
e
t
hod
.
W
he
n
c
o
m
pa
r
i
ng
t
he
E
L
m
e
t
hods
pre
s
e
nt
e
d
a
bove
,
di
s
t
i
nc
t
p
a
t
t
e
rns
i
n
t
he
i
r
p
e
rfor
m
a
n
c
e
a
re
obs
e
rve
d
w
i
t
h
e
a
c
h
E
L
m
e
t
hod
de
m
ons
t
r
a
t
i
ng
i
t
s
un
i
q
ue
a
dva
n
t
a
ge
s
.
A
s
obs
e
rv
e
d
i
n
T
a
b
l
e
8
,
t
he
pe
rfor
m
a
nc
e
m
e
t
ri
c
s
of
d
i
ffe
r
e
n
t
E
L
m
e
t
hods
(S
M
O
T
E
Ba
g
gi
ng
,
S
M
O
T
E
Boos
t
i
ng
,
S
M
O
T
E
S
t
a
c
ki
ng,
a
nd
T
om
e
k
l
i
nks
B
oos
t
i
ng)
a
re
d
i
s
pl
a
ye
d.
E
a
c
h
m
e
t
hod
e
m
pl
oy
s
va
ri
ous
c
l
a
s
s
i
fi
e
rs
a
nd
t
e
c
hni
qu
e
s
t
o
c
o
m
bi
ne
t
he
i
r
p
re
d
i
c
t
i
ons
,
a
i
m
i
n
g
t
o
e
nh
a
nc
e
ove
r
a
l
l
m
od
e
l
p
e
rfor
m
a
n
c
e
.
Ba
s
e
d
on
t
h
e
pr
e
s
e
n
t
e
d
m
e
t
r
i
c
s
,
S
M
O
T
E
B
a
ggi
n
g,
S
M
O
T
E
Bo
os
t
i
n
g,
a
n
d
S
M
O
T
E
S
t
a
c
ki
n
g
a
l
l
yi
e
l
d
t
h
e
s
a
m
e
a
c
c
ur
a
c
y,
pr
e
c
i
s
i
on
,
r
e
c
a
l
l
,
F
1
s
c
or
e
,
RO
C
A
U
C,
g
e
om
e
t
r
i
c
m
e
a
n
,
a
nd
ba
l
a
nc
e
d
a
c
c
ura
c
y
va
l
u
e
s
o
f
0.
91
.
T
o
m
e
k
l
i
nks
B
oos
t
i
ng
l
a
gs
s
l
i
gh
t
l
y
b
e
hi
nd
w
i
t
h
a
n
a
c
c
ur
a
c
y
of
0
.
93
,
pre
c
i
s
i
o
n
of
0.
9
2,
re
c
a
l
l
of
0
.
97,
F
1
s
c
or
e
of
0
.
95,
RO
C
A
U
C
o
f
0
.
91
,
a
nd
l
ow
e
r
k
a
p
pa
of
0
.
83
,
a
s
s
how
n
i
n
T
a
bl
e
8
.
Cons
i
d
e
r
i
ng
t
he
re
s
ul
t
s
of
m
ode
l
s
1
t
o
5
,
w
h
e
r
e
i
t
w
a
s
obs
e
rve
d
t
ha
t
boos
t
i
ng
c
l
a
s
s
i
fi
e
rs
t
e
nd
e
d
t
o
p
e
rfor
m
be
t
t
e
r
a
c
ros
s
va
ri
o
us
m
od
e
l
s
a
n
d
da
t
a
s
e
t
s
,
i
t
c
a
n
b
e
i
nf
e
rr
e
d
t
h
a
t
S
M
O
T
E
Boos
t
i
n
g
m
i
ght
b
e
t
h
e
m
os
t
e
ff
e
c
t
i
v
e
E
L
m
e
t
h
od
a
m
on
g
t
he
four
.
T
h
i
s
i
s
b
e
c
a
us
e
boos
t
i
ng
m
e
t
h
ods
e
xc
e
l
i
n
c
orre
c
t
i
ng
e
rr
ors
i
n
b
a
s
e
m
ode
l
s
a
nd
us
i
ng
t
h
e
i
r
c
ol
l
e
c
t
i
ve
s
t
re
n
gt
hs
.
T
he
S
M
O
T
E
Boos
t
i
ng
m
e
t
hod
a
c
hi
e
ve
d
c
o
m
pa
r
a
bl
e
m
e
t
ri
c
s
w
i
t
h
o
t
he
r
m
e
t
hods
,
s
uc
h
a
s
Ba
ggi
ng
a
n
d
S
t
a
c
ki
n
g,
r
e
qui
ri
ng
m
i
n
i
m
a
l
t
ra
i
ni
ng
t
i
m
e
(0.
01
s
e
c
on
ds
)
a
nd
n
e
gl
i
gi
bl
e
m
e
m
ory
us
a
g
e
.
O
ur
fi
ndi
n
gs
re
s
on
a
t
e
w
i
t
h
[9]
,
w
ho
c
onc
l
ude
d
t
h
a
t
e
ns
e
m
bl
e
-
l
e
a
rni
n
g
m
e
t
ho
ds
boos
t
a
c
c
ura
c
y
i
n
m
e
nt
a
l
he
a
l
t
h
pre
di
c
t
i
ons
.
Evaluation Warning : The document was created with Spire.PDF for Python.
Com
pu
t
S
c
i
Inf
T
e
c
h
nol
IS
S
N
:
2722
-
3221
A
m
a
c
hi
n
e
l
e
ar
ni
ng
ap
pr
oac
h
f
or
e
ar
l
y
pr
e
d
i
c
t
i
o
n
o
f
m
e
nt
al
he
al
t
h
c
r
i
s
e
s
(H
as
s
an
Chi
gagur
e
)
343
W
hi
l
e
S
M
O
T
E
Bo
os
t
i
n
g
de
m
ons
t
r
a
t
e
s
s
t
ron
g
ove
ra
l
l
p
e
rfor
m
a
n
c
e
a
c
r
os
s
v
a
ri
ous
m
e
t
r
i
c
s
,
t
h
e
re
a
re
s
i
t
ua
t
i
o
ns
w
he
r
e
T
om
e
k
l
i
nks
B
oos
t
i
ng
m
i
gh
t
b
e
pr
e
fe
rr
e
d.
In
t
h
e
re
a
l
w
orl
d
,
t
he
d
a
t
a
s
e
t
i
s
n
ot
a
l
w
a
ys
ba
l
a
nc
e
d
.
If
t
h
e
d
a
t
a
s
e
t
i
s
pa
r
t
i
c
ul
a
rl
y
s
e
ns
i
t
i
v
e
t
o
c
l
a
s
s
i
m
b
a
l
a
nc
e
a
nd
foc
us
e
s
on
a
c
c
ura
t
e
l
y
pr
e
di
c
t
i
ng
t
h
e
m
i
nor
i
t
y
c
l
a
s
s
(e
.
g
.
,
m
e
nt
a
l
h
e
a
l
t
h)
,
T
o
m
e
k
l
i
nks
Boos
t
i
ng
m
i
ght
b
e
m
or
e
s
ui
t
a
bl
e
.
It
s
how
e
d
t
he
hi
g
he
s
t
re
c
a
l
l
(0
.
97)
a
m
o
ng
t
he
c
om
pa
r
e
d
E
L
m
e
t
hods
,
i
n
di
c
a
t
i
ng
i
t
s
e
ffe
c
t
i
ve
n
e
s
s
i
n
c
orr
e
c
t
l
y
i
d
e
nt
i
fyi
ng
i
ns
t
a
n
c
e
s
of
t
h
e
m
i
nori
t
y
c
l
a
s
s
.
T
a
b
l
e
8
.
E
ns
e
m
bl
e
l
e
a
rn
i
ng
re
s
u
l
t
s
P
e
rfo
rm
a
n
c
e
m
e
t
ri
c
s
E
n
s
e
m
b
l
e
l
e
a
rn
i
n
g
S
M
O
T
E
Ba
g
g
i
n
g
S
M
O
T
E
Bo
o
s
t
i
n
g
S
M
O
T
E
S
t
a
c
k
i
n
g
T
o
m
e
k
L
i
n
k
s
Bo
o
s
t
i
n
g
A
c
c
u
ra
c
y
0
.
9
3
0
.
9
3
0
.
9
3
0
.
9
3
P
re
c
i
s
i
o
n
0
.
9
4
0
.
9
4
0
.
9
4
0
.
9
2
Re
c
a
l
l
0
.
9
6
0
.
9
6
0
.
9
6
0
.
9
7
F
1
S
c
o
re
0
.
9
5
0
.
9
5
0
.
9
5
0
.
9
5
Ro
c
A
U
C
0
.
9
1
0
.
9
1
0
.
9
1
0
.
9
1
K
a
p
p
a
0
.
8
4
0
.
8
4
0
.
8
4
0
.
8
3
G
e
o
m
e
t
ri
c
m
e
a
n
0
.
9
1
0
.
9
1
0
.
9
1
0
.
9
1
Ba
l
a
n
c
e
d
A
c
c
0
.
9
1
0
.
9
1
0
.
9
1
0
.
9
1
T
i
m
e
(S
e
c
)
0
.
1
5
0
.
0
1
0
.
7
8
0
.
0
2
CP
U
(K
B)
0
0
0
0
3.
9
.
C
on
fu
s
i
on
matr
i
x
for
th
e
b
e
s
t
mod
e
l
W
he
n
i
t
i
s
e
s
s
e
nt
i
a
l
t
o
a
vo
i
d
m
i
s
s
i
ng
a
t
rue
pos
i
t
i
ve
-
fo
r
i
n
s
t
a
nc
e
,
i
n
s
i
t
ua
t
i
o
ns
l
i
ke
h
e
a
l
t
h
c
a
r
e
or
s
a
fe
t
y
-
s
e
ns
i
t
i
v
e
e
nvi
ro
nm
e
nt
s
-
t
he
hi
gh
re
c
a
l
l
s
c
o
re
a
c
h
i
e
ve
d
by
T
o
m
e
k
L
i
n
ks
Boos
t
i
ng
i
s
p
a
r
t
i
c
ul
a
rl
y
a
dva
nt
a
ge
ous
,
a
s
s
how
n
i
n
T
a
b
l
e
8
.
T
hi
s
a
pp
roa
c
h
e
f
fe
c
t
i
ve
l
y
i
de
n
t
i
f
i
e
s
t
h
e
m
a
j
o
ri
t
y
of
t
ru
e
c
a
s
e
s
w
i
t
h
i
n
t
h
e
m
i
nor
i
t
y
c
l
a
s
s
.
F
i
gu
re
3
pr
e
s
e
n
t
s
t
h
e
c
orr
e
s
pond
i
ng
c
onfu
s
i
on
m
a
t
r
i
x,
d
e
t
a
i
l
i
ng
t
he
m
od
e
l
’s
pr
e
di
c
t
i
ons
:
64
,
891
t
ru
e
pos
i
t
i
v
e
s
(c
or
re
c
t
l
y
i
d
e
nt
i
fi
e
d
m
e
n
t
a
l
he
a
l
t
h
c
r
i
s
e
s
,
29
,
6
78
t
ru
e
n
e
ga
t
i
v
e
s
,
9
87
fa
l
s
e
pos
i
t
i
ve
s
,
a
nd
onl
y
111
f
a
l
s
e
n
e
ga
t
i
v
e
s
.
M
i
ni
m
i
z
i
ng
f
a
l
s
e
n
e
g
a
t
i
ve
s
i
s
e
s
s
e
nt
i
a
l
,
a
s
m
i
s
s
i
ng
i
ndi
v
i
du
a
l
s
ne
e
di
n
g
a
s
s
i
s
t
a
nc
e
can
ha
v
e
s
e
ri
o
us
c
ons
e
qu
e
n
c
e
s
.
W
hi
l
e
T
o
m
e
k
L
i
nks
Bo
os
t
i
ng
m
a
y
h
a
v
e
s
l
i
gh
t
l
y
l
ow
e
r
a
c
c
ur
a
c
y
a
n
d
pre
c
i
s
i
on
t
ha
n
o
t
he
r
m
e
t
hods
,
i
t
s
ba
l
a
n
c
e
d
a
c
c
ur
a
c
y
a
nd
ge
o
m
e
t
ri
c
m
e
a
n
a
r
e
c
o
m
p
a
ra
b
l
e
,
i
nd
i
c
a
t
i
ng
a
s
t
r
ong
t
ra
d
e
-
off
b
e
t
w
e
e
n
c
l
a
s
s
r
e
pre
s
e
nt
a
t
i
on
a
nd
ove
ra
l
l
pe
rfor
m
a
n
c
e
.
Re
m
ov
i
ng
n
e
a
r
-
ne
i
ghbor
i
ns
t
a
nc
e
s
v
i
a
T
om
e
k
l
i
nks
s
h
a
rpe
ns
d
e
c
i
s
i
on
bou
nda
r
i
e
s
,
p
ot
e
nt
i
a
l
l
y
i
m
provi
ng
g
e
ne
ra
l
i
z
a
t
i
on
a
n
d
r
e
du
c
i
ng
ov
e
rf
i
t
t
i
ng
,
e
s
pe
c
i
a
l
l
y
w
i
t
h
c
om
pl
e
x
or
noi
s
y
d
a
t
a
.
O
v
e
r
a
l
l
,
t
h
e
e
n
s
e
m
bl
e
m
e
t
hods
e
v
a
l
u
a
t
e
d
d
e
m
ons
t
ra
t
e
ro
bus
t
pre
di
c
t
i
v
e
pow
e
r
a
nd
pra
c
t
i
c
a
l
e
ffi
c
i
e
nc
y
for
m
e
n
t
a
l
di
s
or
de
r
pr
e
d
i
c
t
i
on
.
W
hi
l
e
S
M
O
T
E
Bo
os
t
i
n
g
re
m
a
i
ns
a
s
t
rong
op
t
i
o
n,
t
hi
s
s
t
udy
f
a
vours
T
om
e
k
L
i
n
ks
Boos
t
i
ng
d
ue
t
o
i
t
s
e
x
c
e
p
t
i
o
na
l
97%
re
c
a
l
l
,
a
l
i
gn
i
ng
w
i
t
h
t
he
c
ri
t
i
c
a
l
go
a
l
of
a
c
c
u
ra
t
e
l
y
i
de
n
t
i
f
yi
ng
t
h
os
e
ne
e
di
n
g
m
e
nt
a
l
h
e
a
l
t
h
i
n
t
e
rv
e
n
t
i
on
a
nd
s
up
port
.
F
i
gure
3
.
C
onfus
i
o
n
m
a
t
ri
x
for
T
om
e
k
l
i
nks
bo
os
t
i
n
g
e
ns
e
m
bl
e
l
e
a
rn
i
ng
4.
C
O
N
C
LU
S
I
O
N
T
hi
s
s
t
udy
provi
de
s
a
c
om
pre
h
e
ns
i
v
e
c
o
m
pa
r
i
s
on
o
f
M
L
t
e
c
hn
i
qu
e
s
for
pre
di
c
t
i
n
g
m
e
n
t
a
l
h
e
a
l
t
h
c
ri
s
e
s
us
i
ng
a
l
a
rge
,
l
ongi
t
udi
na
l
da
t
a
s
e
t
.
E
ns
e
m
b
l
e
m
e
t
hods
,
pa
r
t
i
c
ul
a
rl
y
X
G
Boos
t
a
nd
T
l
B
E
,
de
m
ons
t
ra
t
e
d
s
upe
ri
or
p
e
rfor
m
a
n
c
e
i
n
a
c
c
ura
c
y
,
r
e
c
a
l
l
,
a
n
d
b
a
l
a
nc
e
d
a
c
c
ura
c
y
.
K
e
y
ri
s
k
fa
c
t
ors
i
d
e
nt
i
fi
e
d
i
n
c
l
ud
e
pri
or
hos
pi
t
a
l
i
z
a
t
i
ons
,
m
e
di
c
a
t
i
on
a
dh
e
re
n
c
e
,
a
nd
r
e
c
e
nt
be
ha
vi
ora
l
i
ndi
c
a
t
ors
.
T
h
e
s
e
f
i
ndi
ngs
s
ugg
e
s
t
t
ha
t
i
nt
e
gra
t
i
n
g
m
a
c
hi
n
e
l
e
a
rni
ng
m
od
e
l
s
i
nt
o
m
e
n
t
a
l
h
e
a
l
t
h
c
a
r
e
c
ou
l
d
s
i
gni
f
i
c
a
nt
l
y
e
nh
a
n
c
e
e
a
rl
y
i
d
e
nt
i
fi
c
a
t
i
o
n
a
nd
i
nt
e
rv
e
nt
i
on
for
i
nd
i
v
i
dua
l
s
a
t
ri
s
k
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2722
-
3221
Com
pu
t
S
c
i
Inf
T
e
c
h
nol
,
V
o
l
.
6
,
N
o
.
3
,
N
ov
e
m
be
r
20
25
:
335
-
345
344
A
C
K
N
O
WL
ED
G
M
EN
TS
W
e
e
x
t
e
nd
ou
r
s
i
n
c
e
r
e
a
pp
r
e
c
i
a
t
i
on
t
o
t
h
e
M
i
n
i
s
t
r
y
of
H
e
a
l
t
h
a
nd
Ch
i
l
d
C
a
r
e
of
Z
i
m
b
a
bw
e
fo
r
g
r
a
n
t
i
n
g
us
a
c
c
e
s
s
t
o
t
h
e
d
a
t
a
s
e
t
t
h
a
t
w
a
s
i
ns
t
r
u
m
e
n
t
a
l
t
o
t
h
e
s
u
c
c
e
s
s
o
f
t
h
i
s
pr
o
j
e
c
t
.
W
e
a
p
pr
e
c
i
a
t
e
t
h
e
i
r
c
o
m
m
i
t
m
e
n
t
t
o
a
dv
a
n
c
i
n
g
r
e
s
e
a
r
c
h
i
n
m
e
n
t
a
l
h
e
a
l
t
h
a
nd
t
h
e
i
r
w
i
l
l
i
n
gn
e
s
s
t
o
fa
c
i
l
i
t
a
t
e
a
c
a
d
e
m
i
c
i
nq
u
i
r
y
t
hr
o
ug
h
d
a
t
a
s
h
a
r
i
n
g
.
T
hi
s
p
a
r
t
n
e
rs
h
i
p
h
a
s
s
i
g
n
i
f
i
c
a
n
t
l
y
c
o
n
t
r
i
bu
t
e
d
t
o
t
h
e
qu
a
l
i
t
y
a
n
d
r
e
l
e
v
a
n
c
e
o
f
ou
r
f
i
nd
i
ngs
.
F
U
N
D
I
N
G
I
N
F
O
R
M
A
TI
O
N
T
he
a
u
t
hors
de
c
l
a
r
e
t
he
y
r
e
c
e
i
v
e
d
no
fundi
ng
for
t
hi
s
w
ork
.
A
U
TH
O
R
C
O
N
TR
I
BU
TI
O
N
S
S
TA
T
EM
EN
T
T
hi
s
j
ou
rna
l
us
e
s
t
h
e
Cont
r
i
bu
t
or
Rol
e
s
T
a
xono
m
y
(C
Re
di
T
)
t
o
r
e
c
o
gni
z
e
i
ndi
v
i
du
a
l
a
u
t
hor
c
ont
r
i
bu
t
i
ons
,
r
e
du
c
e
a
ut
hors
h
i
p
di
s
pu
t
e
s
,
a
nd
fa
c
i
l
i
t
a
t
e
c
o
l
l
a
bora
t
i
on
.
N
ame
of
A
u
th
o
r
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
H
a
s
s
a
n
C
hi
g
a
gur
e
✓
✓
✓
✓
✓
✓
✓
✓
L
uc
y
Ch
a
ri
t
y
S
a
k
a
l
a
✓
✓
✓
✓
✓
✓
C
:
C
o
n
c
e
p
t
u
a
l
i
z
a
t
i
o
n
M
:
M
e
t
h
o
d
o
l
o
g
y
So
:
So
ft
w
a
re
Va
:
Va
l
i
d
a
t
i
o
n
Fo
:
Fo
rm
a
l
a
n
a
l
y
s
i
s
I
:
I
n
v
e
s
t
i
g
a
t
i
o
n
R
:
R
e
s
o
u
rc
e
s
D
:
D
a
t
a
Cu
ra
t
i
o
n
O
:
W
ri
t
i
n
g
-
O
ri
g
i
n
a
l
D
ra
ft
E
:
W
ri
t
i
n
g
-
Re
v
i
e
w
&
E
d
i
t
i
n
g
Vi
:
Vi
s
u
a
l
i
z
a
t
i
o
n
Su
:
Su
p
e
rv
i
s
i
o
n
P
:
P
ro
j
e
c
t
a
d
m
i
n
i
s
t
ra
t
i
o
n
Fu
:
Fu
n
d
i
n
g
a
c
q
u
i
s
i
t
i
o
n
C
O
N
F
LI
C
T
O
F
I
N
T
ER
ES
T
S
TA
T
EM
EN
T
A
ut
hors
s
t
a
t
e
no
c
onf
l
i
c
t
of
i
n
t
e
r
e
s
t
.
I
N
F
O
R
M
ED
C
O
N
S
EN
T
W
e
ha
ve
obt
a
i
ne
d
i
n
form
e
d
c
o
ns
e
nt
fro
m
a
l
l
i
nd
i
vi
dua
l
s
i
nc
l
ude
d
i
n
t
h
i
s
s
t
udy.
ETH
I
C
A
L
A
P
P
R
O
V
A
L
T
hi
s
r
e
s
e
a
r
c
h
a
dhe
re
s
t
o
a
l
l
a
p
pl
i
c
a
bl
e
na
t
i
on
a
l
r
e
gu
l
a
t
i
ons
a
nd
i
ns
t
i
t
ut
i
ona
l
pol
i
c
i
e
s
,
a
l
i
gn
i
ng
w
i
t
h
t
he
M
i
ni
s
t
ry
of
H
e
a
l
t
h
a
n
d
Ch
i
l
d
Ca
r
e
'
s
de
c
l
a
r
a
t
i
on
.
T
he
H
a
ra
re
Ins
t
i
t
ut
e
of
T
e
c
h
nol
og
y
i
ns
t
i
t
ut
i
ona
l
r
e
vi
e
w
boa
rd
a
l
s
o
a
ppro
ve
d
i
t
.
D
A
TA
A
V
A
I
LA
BI
LI
TY
T
he
d
a
t
a
und
e
rp
i
nni
ng
t
h
i
s
s
t
udy
w
a
s
obt
a
i
n
e
d
fro
m
Z
i
m
ba
b
w
e
'
s
M
i
n
i
s
t
ry
o
f
H
e
a
l
t
h
a
nd
Chi
l
d
Ca
re
.
P
l
e
a
s
e
not
e
t
ha
t
re
s
t
r
i
c
t
i
ons
a
p
pl
y
t
o
t
he
a
va
i
l
a
bi
l
i
t
y
of
t
hi
s
da
t
a
,
a
s
i
t
w
a
s
us
e
d
und
e
r
l
i
c
e
ns
e
for
t
he
p
urpos
e
of
t
h
i
s
r
e
s
e
a
rc
h.
F
urt
h
e
r
i
nf
orm
a
t
i
on
c
a
n
b
e
found
t
hrough
w
w
w
.
m
oh
c
c
.
gov
.
z
w
R
EF
ER
EN
C
ES
[1
]
W
o
rl
d
H
e
a
l
t
h
O
r
g
a
n
i
z
a
t
i
o
n
,
T
h
e
W
o
r
l
d
M
e
n
t
a
l
H
e
a
l
t
h
R
e
p
o
r
t
:
t
r
a
n
s
f
o
r
m
i
n
g
m
e
n
t
a
l
h
e
a
l
t
h
f
o
r
a
l
l
.
G
e
n
e
v
a
,
S
w
i
t
z
e
rl
a
n
d
:
W
H
O
,
2
0
2
2
.
[2
]
R.
F
i
l
i
p
,
R.
G
h
e
o
rg
h
i
t
a
P
u
s
c
a
s
e
l
u
,
L
.
A
n
c
h
i
d
i
n
-
N
o
ro
c
e
l
,
M
.
D
i
m
i
a
n
,
a
n
d
W
.
K
.
S
a
v
a
g
e
,
“
G
l
o
b
a
l
c
h
a
l
l
e
n
g
e
s
t
o
p
u
b
l
i
c
h
e
a
l
t
h
c
a
re
s
y
s
t
e
m
s
d
u
ri
n
g
t
h
e
CO
V
ID
-
1
9
p
a
n
d
e
m
i
c
:
A
re
v
i
e
w
o
f
p
a
n
d
e
m
i
c
m
e
a
s
u
re
s
a
n
d
p
ro
b
l
e
m
s
,
”
J
o
u
r
n
a
l
o
f
P
e
r
s
o
n
a
l
i
z
e
d
M
e
d
i
c
i
n
e
,
v
o
l
.
1
2
,
n
o
.
8
,
2
0
2
2
,
d
o
i
:
1
0
.
3
3
9
0
/
j
p
m
1
2
0
8
1
2
9
5
.
[3
]
J
.
L
i
n
a
n
d
W
.
G
u
o
,
“
T
h
e
re
s
e
a
rc
h
o
n
ri
s
k
fa
c
t
o
rs
fo
r
a
d
o
l
e
s
c
e
n
t
s
’
m
e
n
t
a
l
h
e
a
l
t
h
,
”
B
e
h
a
v
i
o
r
a
l
S
c
i
e
n
c
e
s
,
v
o
l
.
1
4
,
n
o
.
4
,
2
0
2
4
,
d
o
i
:
1
0
.
3
3
9
0
/
b
s
1
4
0
4
0
2
6
3
.
[4
]
R.
Bh
a
t
i
a
a
n
d
J
.
K
o
u
r,
“
E
n
h
a
n
c
i
n
g
s
o
c
i
a
l
i
n
t
e
g
ra
t
i
o
n
a
n
d
m
e
n
t
a
l
h
e
a
l
t
h
o
f
c
h
i
l
d
re
n
w
i
t
h
i
n
t
e
l
l
e
c
t
u
a
l
a
n
d
d
e
v
e
l
o
p
m
e
n
t
a
l
d
i
s
a
b
i
l
i
t
i
e
s
(ID
D
)
t
h
ro
u
g
h
e
d
u
c
a
t
i
o
n
a
l
p
r
o
g
ra
m
s
,
”
S
o
c
i
a
l
In
c
l
u
s
i
o
n
T
a
c
t
i
c
s
f
o
r
P
e
o
p
l
e
wi
t
h
In
t
e
l
l
e
c
t
u
a
l
a
n
d
D
e
v
e
l
o
p
m
e
n
t
a
l
D
i
s
a
b
i
l
i
t
i
e
s
,
p
p
.
3
2
1
–
3
4
2
,
2
0
2
4
,
d
o
i
:
1
0
.
4
0
1
8
/
9
7
9
-
8
-
3
6
9
3
-
3
1
7
6
-
7
.
c
h
0
1
4
.
[
5]
A
.
Ż
ó
ł
t
a
s
z
e
k
,
“
W
HO
b
e
a
rs
t
h
e
h
i
g
h
c
o
s
t
s
o
f
m
e
n
t
a
l
h
e
a
l
t
h
p
ro
b
l
e
m
s
i
n
t
h
e
L
a
b
o
u
r
F
o
rc
e
?
,
”
L
o
d
z
E
c
o
n
o
m
i
c
s
W
o
r
k
i
n
g
P
a
p
e
r
s
,
2
0
2
4
.
[6
]
In
d
ra
w
a
t
i
,
S
.
M
.
,
A
n
a
s
,
T
.
,
M
u
l
y
a
n
i
,
S
.
a
n
d
In
d
ra
w
a
t
i
,
N
.
,
2
0
2
2
.
Ke
e
p
i
n
g
In
d
o
n
e
s
i
a
S
a
f
e
f
r
o
m
t
h
e
CO
V
ID
-
1
9
P
a
n
d
e
m
i
c
.
J
a
k
a
rt
a
,
In
d
o
n
e
s
i
a
:
IS
E
A
S
-
Y
u
s
o
f
Is
h
a
k
In
s
t
i
t
u
t
e
.
[7
]
A
.
Re
m
b
i
s
z
a
n
d
T
.
S
h
a
h
a
l
,
“
M
e
n
t
a
l
h
e
a
l
t
h
c
ri
s
i
s
,
”
i
n
M
a
n
a
g
i
n
g
M
e
n
t
a
l
Il
l
n
e
s
s
a
f
t
e
r
CO
V
ID
-
1
9
In
f
e
c
t
i
o
n
,
S
.
A
.
Co
l
l
i
e
r
,
N
e
w
J
e
rs
e
y
:
J
o
h
n
W
i
l
e
y
&
S
o
n
s
,
In
c
.,
2024,
p
p
.
1
6
9
–
1
9
3
,
d
o
i
:
1
0
.
1
0
0
2
/
9
7
8
1
3
9
4
2
5
0
1
0
3
.
c
h
9
.
[8
]
S
.
C
.
F
a
b
e
r,
A
.
K
h
a
n
n
a
Ro
y
,
T
.
I
.
M
i
c
h
a
e
l
s
,
a
n
d
M
.
T
.
W
i
l
l
i
a
m
s
,
“
T
h
e
w
e
a
p
o
n
i
z
a
t
i
o
n
o
f
m
e
d
i
c
i
n
e
:
E
a
rl
y
p
s
y
c
h
o
s
i
s
i
n
t
h
e
b
l
a
c
k
c
o
m
m
u
n
i
t
y
a
n
d
t
h
e
n
e
e
d
fo
r
ra
c
i
a
l
l
y
i
n
fo
rm
e
d
m
e
n
t
a
l
h
e
a
l
t
h
c
a
re
,
”
F
r
o
n
t
i
e
r
s
i
n
P
s
y
c
h
i
a
t
r
y
,
v
o
l
.
1
4
,
2
0
2
3
,
d
o
i
:
1
0
.
3
3
8
9
/
fp
s
y
t
.
2
0
2
3
.
1
0
9
8
2
9
2
.
[9
]
R.
G
a
rri
g
a
e
t
a
l
.
,
“
M
a
c
h
i
n
e
l
e
a
rn
i
n
g
m
o
d
e
l
t
o
p
re
d
i
c
t
m
e
n
t
a
l
h
e
a
l
t
h
c
ri
s
e
s
fro
m
e
l
e
c
t
ro
n
i
c
h
e
a
l
t
h
re
c
o
rd
s
,
”
Na
t
u
r
e
M
e
d
i
c
i
n
e
,
v
o
l
.
2
8
,
n
o
.
6
,
p
p
.
1
2
4
0
–
1
2
4
8
,
2
0
2
2
,
d
o
i
:
1
0
.
1
0
3
8
/
s
4
1
5
9
1
-
022
-
0
1
8
1
1
-
5.
Evaluation Warning : The document was created with Spire.PDF for Python.