I
nte
rna
t
io
na
l J
o
urna
l o
f
Adv
a
nces in Applie
d Science
s
(
I
J
AAS)
Vo
l.
1
5
,
No
.
3
,
Sep
tem
b
er
20
2
6
,
p
p
.
1
2
3
9
~
1
2
5
1
I
SS
N:
2252
-
8
8
1
4
,
DOI
:
1
0
.
1
1
5
9
1
/ijaas
.
v
1
5
.
i
3
.
pp
1
2
3
9
-
1
2
5
1
1239
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
a
s
.
ia
esco
r
e.
co
m
Rhino
o
ptimize
r
f
o
r hyperpa
ra
met
er t
uni
ng
in
d
epr
ess
io
n
detec
tion
Sim
is
a
ni Nda
ba
,
Ra
j
a
la
k
s
hm
i Selv
a
ra
j
,
H
lo
m
a
ni H
lo
m
a
ni
D
e
p
a
r
t
me
n
t
o
f
C
o
mp
u
t
i
n
g
a
n
d
I
n
f
o
r
m
a
t
i
c
s
,
S
c
h
o
o
l
o
f
P
u
r
e
a
n
d
A
p
p
l
i
e
d
S
c
i
e
n
c
e
s
,
B
o
t
sw
a
n
a
I
n
t
e
r
n
a
t
i
o
n
a
l
U
n
i
v
e
r
si
t
y
o
f
S
c
i
e
n
c
e
a
n
d
T
e
c
h
n
o
l
o
g
y
,
P
a
l
a
p
y
e
,
B
o
t
sw
a
n
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
an
2
9
,
2
0
2
6
R
ev
is
ed
J
u
l 1
3
,
2
0
2
6
Acc
ep
ted
Au
g
1
3
,
2
0
2
6
De
p
re
ss
io
n
d
e
tec
ti
o
n
u
sin
g
m
a
c
h
in
e
lea
rn
i
n
g
(M
L)
h
a
s
g
a
in
e
d
in
c
re
a
sin
g
a
tt
e
n
ti
o
n
;
h
o
we
v
e
r,
m
o
d
e
l
p
e
rfo
rm
a
n
c
e
re
m
a
in
s
h
ig
h
ly
se
n
siti
v
e
t
o
h
y
p
e
r
p
a
ra
m
e
ter
c
o
n
fig
u
ra
ti
o
n
.
T
h
is
stu
d
y
i
n
tro
d
u
c
e
s
th
e
R
h
in
o
o
p
ti
m
ize
r
,
a
n
o
v
e
l
b
i
o
-
i
n
sp
ired
m
e
tah
e
u
risti
c
a
lg
o
rit
h
m
fo
r
h
y
p
e
rp
a
ra
m
e
ter
o
p
t
imiz
a
ti
o
n
(HPO)
in
sp
ired
b
y
b
e
h
a
v
i
o
ra
l
m
e
c
h
a
n
ism
s
o
b
se
rv
e
d
in
R
h
in
o
c
e
ro
s
m
o
v
e
m
e
n
t
a
n
d
su
rv
iv
a
l
stra
teg
i
e
s.
Th
e
o
p
ti
m
ize
r
i
n
teg
ra
tes
L
é
v
y
-
fli
g
h
t
e
x
p
lo
ra
ti
o
n
,
e
li
te
-
g
u
i
d
e
d
e
x
p
lo
it
a
t
io
n
,
wa
ll
o
wi
n
g
-
b
a
se
d
e
sc
a
p
e
b
e
h
a
v
io
r
,
a
n
d
m
ig
ra
ti
o
n
-
b
a
se
d
re
p
o
siti
o
n
i
n
g
t
o
b
a
lan
c
e
e
x
p
lo
ra
ti
o
n
a
n
d
e
x
p
l
o
it
a
t
io
n
wh
il
e
re
d
u
c
in
g
p
re
m
a
tu
re
c
o
n
v
e
r
g
e
n
c
e
.
Th
e
p
ro
p
o
se
d
a
p
p
ro
a
c
h
wa
s e
v
a
lu
a
ted
o
n
a
tran
sfo
rm
e
r
-
b
a
se
d
d
e
p
re
ss
io
n
-
d
e
t
e
c
ti
o
n
m
o
d
e
l
u
si
n
g
F
1
-
sc
o
re
a
s
t
h
e
p
rima
ry
o
p
ti
m
iza
ti
o
n
o
b
jec
ti
v
e
d
u
e
to
t
h
e
c
las
s
imb
a
lan
c
e
c
o
m
m
o
n
l
y
p
re
se
n
t
i
n
m
e
n
tal
-
h
e
a
lt
h
d
a
tas
e
ts.
P
e
rfo
rm
a
n
c
e
wa
s
c
o
m
p
a
re
d
a
g
a
in
st
trad
it
i
o
n
a
l
se
a
rc
h
m
e
th
o
d
s,
c
las
sic
a
l
m
e
tah
e
u
risti
c
s,
a
n
d
m
o
d
e
rn
b
i
o
-
i
n
sp
ired
o
p
t
imiz
a
ti
o
n
a
lg
o
rit
h
m
s
u
n
d
e
r
a
n
id
e
n
ti
c
a
l
e
v
a
lu
a
ti
o
n
b
u
d
g
e
t.
Rh
i
n
o
o
p
ti
m
ize
r
a
c
h
iev
e
d
a
b
e
st
o
b
se
rv
e
d
F
1
-
sc
o
re
o
f
0
.
8
0
0
5
,
e
q
u
i
v
a
len
t
to
th
e
sin
e
-
c
o
si
n
e
a
lg
o
rit
h
m
(S
CA)
a
n
d
g
re
y
wo
lf
o
p
ti
m
iza
ti
o
n
(G
WO)
with
th
e
sa
m
e
e
v
a
lu
a
ti
o
n
b
u
d
g
e
t.
Co
n
siste
n
t
c
o
n
v
e
rg
e
n
c
e
b
e
h
a
v
i
o
r
a
n
d
sta
ti
stica
l
a
n
a
l
y
se
s
su
g
g
e
ste
d
t
h
a
t
t
h
e
p
ro
p
o
se
d
R
h
in
o
o
p
ti
m
ize
r
m
a
y
s
e
rv
e
a
s
a
c
o
m
p
e
ti
ti
v
e
a
p
p
ro
a
c
h
f
o
r
HPO
i
n
d
e
p
re
ss
io
n
d
e
tec
ti
o
n
a
n
d
t
h
a
t
b
e
h
a
v
io
r
-
i
n
sp
ired
o
p
t
imiz
a
ti
o
n
stra
te
g
ies
m
ig
h
t
b
e
u
se
fu
l
in
m
a
c
h
i
n
e
-
lea
rn
in
g
a
p
p
li
c
a
ti
o
n
s
wit
h
li
m
it
e
d
c
o
m
p
u
tatio
n
a
l
re
so
u
rc
e
s.
K
ey
w
o
r
d
s
:
B
io
-
in
s
p
ir
ed
o
p
tim
izatio
n
Dep
r
ess
io
n
d
e
tectio
n
F1
-
s
co
r
e
Hy
p
er
p
ar
a
m
eter
o
p
tim
izatio
n
R
h
in
o
o
p
tim
izer
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
:
Simisan
i N
d
ab
a
Dep
ar
tm
en
t o
f
C
o
m
p
u
tin
g
an
d
I
n
f
o
r
m
atics
,
Sch
o
o
l o
f
Pu
r
e
a
n
d
Ap
p
lied
Scien
ce
s
B
o
ts
wan
a
I
n
ter
n
atio
n
al
Un
iv
e
r
s
ity
o
f
Scien
ce
an
d
T
ec
h
n
o
lo
g
y
B
o
s
eja,
Priv
ate
B
ag
1
6
,
Palap
y
e,
B
o
ts
wan
a
E
m
ail: n
s
2
4
0
2
0
0
0
0
@
b
iu
s
t.a
c.
b
w
1.
I
NT
RO
D
UCT
I
O
N
Dep
r
ess
io
n
is
a
p
r
ev
alen
t
p
s
y
ch
iatr
ic
d
is
o
r
d
er
th
at
co
n
tr
ib
u
tes
s
u
b
s
tan
tially
to
th
e
g
lo
b
al
b
u
r
d
en
o
f
d
is
ea
s
e
an
d
s
ig
n
if
ican
tly
af
f
ec
ts
co
g
n
itiv
e,
e
m
o
tio
n
al,
b
eh
a
v
io
r
al
,
an
d
s
o
cial
f
u
n
ctio
n
in
g
[
1
]
,
[
2
]
.
I
n
d
iv
id
u
als
with
d
ep
r
ess
io
n
f
r
eq
u
en
tly
ex
p
er
ien
ce
im
p
air
ed
ex
ec
u
tiv
e
f
u
n
ctio
n
in
g
,
d
is
to
r
ted
th
i
n
k
in
g
,
d
im
in
is
h
ed
m
o
tiv
atio
n
,
s
leep
d
is
tu
r
b
a
n
ce
s
,
ap
p
etite
ch
a
n
g
es,
a
n
d
r
ed
u
ce
d
q
u
ality
o
f
life
.
C
o
n
s
eq
u
e
n
tly
,
ea
r
l
y
d
ia
g
n
o
s
is
r
e
m
a
i
n
s
es
s
e
n
t
i
a
l
f
o
r
t
i
m
el
y
i
n
t
e
r
v
e
n
t
i
o
n
a
n
d
i
m
p
r
o
v
e
d
t
r
e
a
tm
e
n
t
o
u
t
c
o
m
e
s
.
H
o
w
e
v
e
r
,
c
o
n
v
e
n
t
i
o
n
a
l
d
i
a
g
n
o
s
ti
c
ap
p
r
o
ac
h
es
ar
e
lar
g
ely
s
u
b
jec
tiv
e,
r
ely
h
ea
v
ily
o
n
s
elf
-
r
ep
o
r
tin
g
,
an
d
ar
e
s
u
s
ce
p
tib
le
to
r
ec
all
b
ias,
l
i
m
i
t
i
n
g
t
h
e
i
r
s
c
al
a
b
i
li
t
y
a
n
d
c
o
n
s
is
te
n
cy
i
n
c
l
i
n
i
c
al
p
r
a
c
t
i
ce
,
p
a
r
t
i
c
u
la
r
l
y
i
n
r
e
s
o
u
r
c
e
-
c
o
n
s
t
r
a
i
n
e
d
e
n
v
ir
o
n
m
e
n
t
s
[
3
]
,
[
4
]
.
R
ec
en
t
ad
v
an
ce
s
in
ar
tific
ial
in
tellig
en
ce
(
AI
)
a
n
d
ma
c
h
in
e
lear
n
in
g
(
ML
)
h
av
e
e
n
ab
le
d
s
ca
lab
le
ap
p
r
o
ac
h
es
f
o
r
au
to
m
ate
d
d
e
p
r
ess
io
n
d
etec
tio
n
b
y
an
al
y
zin
g
d
ig
ital
d
ata
s
o
u
r
ce
s
s
u
ch
as
s
o
cial
m
ed
ia
p
o
s
ts
,
b
eh
av
io
r
al
t
r
ac
es,
an
d
o
th
er
t
ex
tu
al
in
f
o
r
m
atio
n
[
5
]
,
[
6
]
.
T
h
ese
ap
p
r
o
ac
h
es
h
av
e
d
em
o
n
s
tr
ated
co
n
s
id
er
ab
le
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
8
1
4
I
n
t J Ad
v
Ap
p
l Sci
,
Vo
l.
1
5
,
No
.
3
,
Sep
tem
b
er
2
0
2
6
:
1
2
3
9
-
1
2
5
1
1240
p
o
ten
tial
f
o
r
s
u
p
p
o
r
tin
g
clin
i
ca
l
d
ec
is
io
n
-
m
ak
in
g
;
h
o
wev
e
r
,
th
eir
p
r
ed
ictiv
e
p
e
r
f
o
r
m
an
ce
r
em
ain
s
h
ig
h
ly
d
ep
en
d
e
n
t
o
n
a
p
p
r
o
p
r
iate
m
o
d
el
co
n
f
ig
u
r
atio
n
.
P
o
o
r
l
y
s
e
l
e
c
t
e
d
h
y
p
e
r
p
a
r
a
m
e
t
e
r
s
o
f
t
e
n
r
e
s
u
l
t
i
n
o
v
e
r
f
i
t
t
i
n
g
,
r
e
d
u
c
e
d
g
e
n
e
r
a
l
i
z
a
b
i
li
t
y
,
s
l
o
w
er
c
o
n
v
e
r
g
e
n
c
e
,
a
n
d
d
e
g
r
a
d
e
d
p
r
e
d
i
c
t
i
v
e
p
e
r
f
o
r
m
a
n
c
e
o
n
p
r
e
v
i
o
u
s
l
y
u
n
s
e
e
n
d
a
t
a
.
Hy
p
er
p
ar
a
m
eter
o
p
tim
izatio
n
(
HPO)
h
as th
er
ef
o
r
e
b
ec
o
m
e
a
f
u
n
d
am
e
n
tal
co
m
p
o
n
en
t o
f
m
o
d
er
n
M
L
wo
r
k
f
l
o
ws.
As
d
escr
ib
ed
b
y
Yu
a
n
d
Z
h
u
[
7
]
,
HPO
aim
s
to
id
en
t
if
y
h
y
p
e
r
p
ar
am
eter
co
n
f
ig
u
r
atio
n
s
th
at
m
ax
im
ize
p
r
ed
ictiv
e
p
e
r
f
o
r
m
a
n
ce
wh
ile
r
ed
u
cin
g
r
elian
ce
o
n
m
an
u
al
ex
p
er
t tu
n
in
g
.
B
ey
o
n
d
im
p
r
o
v
in
g
m
o
d
el
ac
cu
r
ac
y
,
HPO
also
en
h
an
ce
s
co
m
p
u
tatio
n
al
ef
f
icien
c
y
,
r
e
p
r
o
d
u
cib
ilit
y
,
an
d
th
e
c
r
ed
ib
i
lity
o
f
ML
m
o
d
els
th
r
o
u
g
h
s
y
s
tem
atic
p
ar
am
eter
s
elec
tio
n
[
7
]
,
[
8
]
.
T
r
ad
itio
n
al
o
p
tim
izatio
n
tech
n
iq
u
es
s
u
ch
as
g
r
id
s
ea
r
ch
a
n
d
B
ay
esian
o
p
tim
izatio
n
r
em
ain
wid
ely
u
s
ed
,
wh
ile
p
o
p
u
latio
n
-
b
ased
m
etah
eu
r
is
tic
alg
o
r
ith
m
s
h
av
e
em
er
g
ed
as a
ttra
ctiv
e
alter
n
ativ
es f
o
r
n
a
v
ig
atin
g
co
m
p
lex
,
h
ig
h
-
d
im
en
s
io
n
al
s
ea
r
ch
s
p
ac
es m
o
r
e
ef
f
icien
tly
.
B
io
-
in
s
p
ir
ed
o
p
tim
izatio
n
al
g
o
r
ith
m
s
h
av
e
d
em
o
n
s
t
r
ated
co
n
s
id
er
ab
le
s
u
cc
ess
in
HPO
f
o
r
d
ep
r
ess
io
n
d
etec
tio
n
an
d
r
elat
ed
h
ea
lth
ca
r
e
ap
p
licatio
n
s
.
Pr
ev
io
u
s
s
tu
d
ies
h
av
e
a
p
p
lied
a
p
p
r
o
ac
h
es
in
clu
d
in
g
p
ar
ticle
s
war
m
–
C
u
ck
o
o
s
ea
r
ch
,
g
r
ey
wo
lf
o
p
tim
izatio
n
(
GW
O)
,
Har
r
is
h
awk
s
o
p
tim
izat
io
n
(
HHO)
,
an
t
lio
n
o
p
tim
iz
atio
n
,
im
p
r
o
v
ed
a
n
t
c
o
lo
n
y
o
p
tim
izatio
n
,
d
u
n
g
b
ee
tle
o
p
tim
izatio
n
,
an
d
ca
t
s
war
m
o
p
tim
izatio
n
to
o
p
tim
ize
lear
n
in
g
r
ate,
b
atc
h
s
ize,
d
r
o
p
o
u
t
r
ate,
f
ea
tu
r
e
s
elec
tio
n
,
ac
tiv
atio
n
f
u
n
ctio
n
s
,
an
d
o
th
er
tr
ain
in
g
p
ar
am
eter
s
[
2
]
,
[
5
]
,
[
9
]
–
[
1
5
]
.
C
o
llectiv
ely
,
th
ese
s
tu
d
ies
d
em
o
n
s
tr
ate
th
at
b
io
-
in
s
p
ir
e
d
o
p
tim
izatio
n
ca
n
s
u
b
s
tan
tially
im
p
r
o
v
e
ML
p
er
f
o
r
m
an
ce
.
Nev
er
th
eless
,
m
o
s
t
ex
is
tin
g
wo
r
k
p
r
im
ar
ily
r
ep
o
r
ts
f
in
al
class
if
icatio
n
ac
cu
r
ac
y
wh
i
le
p
r
o
v
id
in
g
lim
ited
a
n
aly
s
is
o
f
o
p
tim
izatio
n
b
eh
a
v
io
r
,
c
o
n
v
e
r
g
en
ce
ch
ar
ac
ter
is
t
ics,
s
ea
r
ch
ef
f
icien
cy
,
r
o
b
u
s
tn
ess
,
o
r
th
e
ex
p
l
o
r
at
io
n
–
ex
p
lo
itatio
n
tr
ad
e
-
o
f
f
.
Fu
r
th
er
m
o
r
e,
ac
c
u
r
ac
y
alo
n
e
is
o
f
ten
in
s
u
f
f
icien
t
f
o
r
ev
alu
atin
g
d
ep
r
ess
io
n
d
etec
tio
n
b
ec
au
s
e
m
en
tal
h
ea
lth
d
atasets
ar
e
co
m
m
o
n
l
y
ch
ar
ac
ter
ized
b
y
class
im
b
alan
ce
,
m
ak
in
g
F1
-
s
co
r
e
a
m
o
r
e
in
f
o
r
m
ativ
e
o
p
tim
izatio
n
o
b
jectiv
e.
Alth
o
u
g
h
n
u
m
er
o
u
s
n
atu
r
e
-
in
s
p
ir
ed
o
p
tim
izatio
n
alg
o
r
ith
m
s
h
av
e
b
ee
n
p
r
o
p
o
s
ed
,
th
e
b
eh
av
io
r
al
m
ec
h
an
is
m
s
o
f
r
h
i
n
o
ce
r
o
s
es
r
em
ain
lar
g
ely
u
n
ex
p
lo
r
ed
with
in
o
p
tim
izatio
n
r
esear
ch
.
R
h
in
o
ce
r
o
s
es
ex
h
ib
it
d
is
tin
ctiv
e
e
co
lo
g
ical
b
eh
av
i
o
r
s
,
in
clu
d
i
n
g
l
o
n
g
-
r
an
g
e
e
x
p
lo
r
atio
n
d
u
r
in
g
r
eso
u
r
ce
s
ea
r
ch
in
g
,
d
ir
ec
ted
m
o
v
em
en
t
to
war
d
s
f
av
o
r
ab
le
lo
ca
tio
n
s
,
a
d
ap
tiv
e
escap
e
b
eh
av
io
r
th
r
o
u
g
h
wallo
win
g
,
a
n
d
m
i
g
r
atio
n
wh
en
en
v
ir
o
n
m
en
tal
co
n
d
itio
n
s
d
et
er
io
r
ate.
T
h
ese
co
m
p
lem
en
ta
r
y
b
e
h
av
io
r
s
n
atu
r
ally
b
alan
ce
ex
p
lo
r
atio
n
an
d
ex
p
lo
itatio
n
,
s
u
g
g
esti
n
g
th
eir
s
u
itab
ilit
y
f
o
r
co
m
p
lex
o
p
ti
m
izatio
n
p
r
o
b
lem
s
r
eq
u
ir
in
g
b
o
th
g
lo
b
al
s
ea
r
ch
ca
p
ab
ilit
y
an
d
l
o
ca
l r
ef
in
em
e
n
t.
Mo
tiv
ated
b
y
th
is
r
esear
ch
g
ap
,
th
is
s
tu
d
y
p
r
o
p
o
s
es
th
e
R
h
in
o
o
p
tim
izer
,
a
n
o
v
el
b
io
-
in
s
p
ir
ed
m
etah
eu
r
is
tic
alg
o
r
ith
m
f
o
r
HPO
in
tr
an
s
f
o
r
m
er
-
b
ased
d
ep
r
ess
io
n
d
etec
tio
n
.
T
h
e
alg
o
r
ith
m
in
teg
r
ates
L
év
y
-
f
lig
h
t
ex
p
lo
r
atio
n
,
elite
-
g
u
id
ed
ex
p
lo
itatio
n
,
wallo
win
g
-
b
ased
escap
e
b
eh
a
v
io
r
,
an
d
m
ig
r
atio
n
-
in
s
p
ir
ed
r
ep
o
s
itio
n
in
g
t
o
im
p
r
o
v
e
o
p
ti
m
izatio
n
p
er
f
o
r
m
a
n
ce
wh
ile
r
ed
u
cin
g
p
r
e
m
atu
r
e
c
o
n
v
e
r
g
e
n
ce
.
T
h
e
s
tu
d
y
h
as
two
p
r
im
ar
y
o
b
jectiv
es:
i)
t
o
ev
alu
ate
th
e
ef
f
ec
tiv
en
ess
o
f
th
e
R
h
in
o
o
p
tim
izer
f
o
r
HPO
in
d
ep
r
ess
io
n
d
etec
tio
n
,
an
d
ii)
t
o
in
v
esti
g
a
te
its
o
p
tim
izatio
n
b
e
h
av
io
r
u
s
in
g
F1
-
s
co
r
e
a
n
d
c
o
n
v
er
g
en
ce
an
aly
s
is
u
n
d
er
a
f
ix
ed
ev
alu
atio
n
b
u
d
g
et.
T
h
r
o
u
g
h
th
ese
co
n
tr
i
b
u
tio
n
s
,
th
e
p
r
o
p
o
s
ed
R
h
in
o
o
p
tim
izer
s
ee
k
s
to
p
r
o
v
id
e
a
co
m
p
etitiv
e
b
eh
a
v
io
r
-
i
n
s
p
ir
ed
o
p
tim
izatio
n
ap
p
r
o
ac
h
f
o
r
co
m
p
u
tatio
n
ally
c
o
n
s
tr
ain
ed
ML
ap
p
licatio
n
s
.
2.
T
H
E
P
RO
P
O
SE
D
M
E
T
H
O
D
T
h
e
R
h
in
o
o
p
tim
izer
m
o
d
els
o
p
tim
izatio
n
as
a
b
alan
ce
b
etwe
en
e
x
p
lo
r
atio
n
a
n
d
e
x
p
lo
itatio
n
.
L
év
y
-
f
lig
h
t
ex
p
lo
r
atio
n
en
a
b
les
b
r
o
ad
s
ea
r
ch
-
s
p
ac
e
tr
a
v
er
s
al,
wh
ile
elite
-
g
u
id
ed
c
o
n
v
er
g
en
ce
d
ir
ec
ts
ca
n
d
id
ate
s
o
lu
tio
n
s
to
war
d
p
r
o
m
is
in
g
r
e
g
io
n
s
id
en
tifi
ed
d
u
r
in
g
o
p
tim
izatio
n
.
T
h
ese
co
m
p
lem
en
tar
y
m
ec
h
an
is
m
s
im
p
r
o
v
e
s
ea
r
ch
d
iv
er
s
ity
wh
ile
s
u
p
p
o
r
tin
g
r
ef
i
n
ed
co
n
v
er
g
en
ce
b
eh
a
v
io
r
.
2
.
1
.
B
io
lo
g
ic
a
l
ins
pira
t
io
n
R
h
in
o
s
ar
e
a
m
o
n
g
th
e
lar
g
est
ter
r
estrial
m
am
m
als
o
n
E
ar
th
.
T
h
e
r
e
a
r
e
c
u
r
r
en
tly
f
i
v
e
s
p
ec
ies
o
f
R
h
in
o
(
f
am
ily
R
h
in
o
ce
r
o
tid
ae
)
,
with
two
in
Af
r
ica
(
b
lack
an
d
wh
ite
R
h
in
o
s
)
an
d
t
h
r
ee
in
Asi
a
(
I
n
d
ian
,
J
av
an
,
an
d
Su
m
atr
an
R
h
in
o
s
)
[
1
6
]
–
[
1
8
]
.
R
h
in
o
s
h
a
v
e
g
r
ey
-
b
r
o
wn
s
k
in
a
n
d
k
er
atin
o
u
s
h
o
r
n
s
,
wh
ich
ar
e
u
s
ed
f
o
r
d
ef
en
s
e
,
as
s
h
o
wn
i
n
Fig
u
r
e
1
;
a
wh
ite
R
h
in
o
s
ee
n
in
Af
r
i
ca
.
Fig
u
r
e
1
illu
s
tr
at
es
th
e
b
i
o
lo
g
ical
in
s
p
ir
atio
n
u
n
d
er
ly
i
n
g
th
e
p
r
o
p
o
s
ed
R
h
in
o
o
p
tim
izer
b
e
h
av
io
r
al
m
ec
h
a
n
is
m
s
.
R
h
in
o
s
ar
e
h
er
b
iv
o
r
o
u
s
an
d
ea
t
g
r
ass
es,
leav
es,
an
d
s
h
r
u
b
s
an
d
ar
e
u
s
u
ally
lo
n
ely
,
ex
ce
p
t
wh
en
m
a
tin
g
[
1
6
]
.
Alth
o
u
g
h
u
s
u
ally
f
r
i
en
d
ly
,
R
h
in
o
s
ca
n
b
ec
o
m
e
ag
g
r
ess
iv
e
wh
en
th
r
e
aten
ed
[
1
8
]
.
As
a
co
n
s
eq
u
en
ce
o
f
p
o
ac
h
in
g
,
s
ev
er
al
co
n
s
er
v
atio
n
m
ea
s
u
r
es
lik
e
an
ti
-
p
o
ac
h
in
g
p
atr
o
ls
an
d
h
o
r
n
-
p
r
o
tectio
n
p
r
o
g
r
am
s
h
a
v
e
b
ee
n
in
tr
o
d
u
ce
d
to
s
av
e
th
e
R
h
in
o
s
[
1
6
]
.
Pre
v
io
u
s
s
tu
d
ies
h
av
e
r
e
p
o
r
te
d
th
at
r
h
i
n
o
ce
r
o
s
es
ar
e
h
ab
ita
t
g
en
er
alis
ts
ca
p
ab
le
o
f
t
r
av
el
lin
g
lo
n
g
d
is
tan
ce
s
in
s
ea
r
ch
o
f
f
o
o
d
[
1
6
]
.
R
h
in
o
s
ar
e
p
r
im
a
r
ily
s
o
lit
ar
y
an
im
als
b
u
t
s
o
m
etim
es
g
ath
er
in
g
r
o
u
p
s
n
ea
r
wate
r
s
o
u
r
ce
s
[
1
9
]
.
T
h
e
y
ar
e
c
ap
ab
le
o
f
t
r
av
ellin
g
c
o
n
s
id
er
a
b
le
d
is
tan
ce
s
wh
ile
s
ea
r
ch
in
g
f
o
r
r
eso
u
r
ce
s
[
2
0
]
.
2
.
2
.
T
he
pro
po
s
ed
Rhin
o
o
p
t
im
izer
:
beha
v
io
r
-
ins
pired d
esig
n
T
h
e
R
h
i
n
o
o
p
t
i
m
i
z
e
r
is
a
b
i
o
-
in
s
p
i
r
e
d
m
e
t
a
h
e
u
r
is
t
i
c
al
g
o
r
i
t
h
m
b
a
s
e
d
o
n
e
c
o
l
o
g
i
c
a
l
b
e
h
a
v
i
o
r
s
o
b
s
e
r
v
e
d
i
n
r
h
i
n
o
c
e
r
o
s
e
s
.
I
t
m
o
d
e
l
s
l
o
n
g
-
r
a
n
g
e
s
e
a
r
c
h
t
h
r
o
u
g
h
f
o
r
a
g
i
n
g
,
e
l
i
t
e
-
g
u
i
d
e
d
e
x
p
l
o
i
ta
t
i
o
n
,
w
a
l
l
o
wi
n
g
-
b
a
s
e
d
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
Ad
v
Ap
p
l Sci
I
SS
N:
2252
-
8
8
1
4
R
h
in
o
o
p
timis
er fo
r
h
yp
erp
a
r
a
mete
r
tu
n
in
g
in
d
e
p
r
ess
io
n
d
etec
tio
n
(
S
imis
a
n
i Nd
a
b
a
)
1241
escap
e
f
r
o
m
s
tag
n
atio
n
,
an
d
m
ig
r
atio
n
-
b
ased
r
ep
o
s
itio
n
in
g
.
T
h
ese
f
o
u
r
b
eh
av
i
o
r
s
ar
e
d
esig
n
ed
to
m
ain
tain
s
ea
r
ch
d
iv
er
s
ity
wh
ile
p
r
o
m
o
tin
g
co
n
v
er
g
en
ce
to
war
d
h
ig
h
-
q
u
ality
s
o
lu
tio
n
s
.
Fig
u
r
e
2
s
h
o
ws
in
ter
ac
tio
n
s
am
o
n
g
th
ese
b
eh
a
v
io
r
s
in
a
m
u
ltid
im
en
s
io
n
al
s
ea
r
ch
s
p
ac
e.
Fig
u
r
e
2
illu
s
tr
at
es
ex
p
lo
r
atio
n
,
ex
p
lo
itatio
n
,
wallo
win
g
escap
e,
an
d
m
ig
r
ati
o
n
m
ec
h
a
n
is
m
s
.
B
y
in
teg
r
atin
g
s
ev
er
al
ad
ap
tiv
e
s
ea
r
ch
s
tr
at
eg
ies
in
to
a
u
n
i
f
ied
f
r
am
ewo
r
k
,
th
e
R
h
in
o
o
p
tim
iz
er
attem
p
ts
to
b
a
lan
ce
e
x
p
lo
r
at
io
n
an
d
ex
p
lo
itatio
n
d
u
r
i
n
g
H
PO
.
Fig
u
r
e
1
.
W
h
ite
R
h
in
o
o
b
s
er
v
ed
in
Pil
an
esb
er
g
Natio
n
al
Par
k
Fig
u
r
e
2
.
R
h
in
o
o
p
tim
izer
s
ea
r
ch
b
eh
a
v
io
r
in
a
m
u
ltid
im
en
s
i
o
n
al
s
ea
r
ch
s
p
ac
e
T
h
e
R
h
in
o
o
p
tim
izer
h
as
th
e
p
o
ten
tial
to
b
ala
n
ce
ex
p
lo
r
ati
o
n
with
ex
p
lo
itatio
n
,
as
d
o
o
t
h
er
m
o
d
er
n
m
etah
eu
r
is
tic
alg
o
r
ith
m
s
.
I
n
s
tead
o
f
in
tr
o
d
u
ci
n
g
n
ew
m
ath
em
atica
l
o
p
er
ato
r
s
,
it
co
m
b
in
es
s
ev
er
al
b
io
lo
g
ically
in
s
p
ir
e
d
s
ea
r
ch
b
eh
av
io
r
s
in
t
o
o
n
e
o
p
tim
iza
tio
n
p
r
o
ce
s
s
,
s
u
ch
as
L
év
y
-
f
lig
h
t
ex
p
l
o
r
atio
n
,
m
o
m
en
tu
m
-
en
h
a
n
ce
d
elite
ex
p
lo
itatio
n
,
wallo
win
g
escap
e
,
an
d
m
i
g
r
atio
n
-
d
r
iv
en
p
o
p
u
lati
o
n
r
e
p
o
s
itio
n
in
g
.
I
n
T
ab
le
1
,
th
ese
d
esig
n
f
ea
tu
r
e
s
ar
e
co
m
p
ar
ed
with
th
o
s
e
o
f
r
ep
r
esen
tativ
e
o
p
tim
izatio
n
a
lg
o
r
ith
m
s
ty
p
ically
u
s
ed
in
HPO
s
tu
d
ies.
Alth
o
u
g
h
all
th
e
m
ec
h
an
is
m
s
u
s
ed
in
th
e
R
h
in
o
o
p
ti
m
izer
h
av
e
b
ee
n
in
clu
d
e
d
in
p
r
ev
io
u
s
o
p
tim
izatio
n
tech
n
i
q
u
es,
th
eir
in
teg
r
atio
n
in
t
o
o
n
e
b
eh
a
v
io
r
-
in
s
p
ir
ed
f
r
am
ew
o
r
k
is
th
e
u
n
i
q
u
e
co
n
tr
ib
u
tio
n
o
f
th
is
wo
r
k
.
Mo
s
t
o
p
tim
izatio
n
alg
o
r
ith
m
s
f
o
c
u
s
o
n
o
n
e
o
r
two
s
ea
r
ch
s
tr
ateg
ies
s
u
ch
as
lead
er
s
h
ip
h
ier
ar
c
h
y
in
GW
O
,
s
u
r
p
r
is
e
-
p
o
u
n
ce
b
e
h
av
i
o
r
in
HHO
,
o
s
cillatio
n
s
in
th
e
s
in
e
–
co
s
in
e
alg
o
r
ith
m
,
o
r
m
u
t
atio
n
-
b
ased
s
ea
r
ch
in
d
if
f
er
en
tial
ev
o
lu
tio
n
(
DE
)
.
T
h
e
R
h
in
o
o
p
tim
izer
u
s
es
f
o
u
r
s
ea
r
ch
b
eh
av
io
r
s
:
lo
n
g
-
r
an
g
e
ex
p
lo
r
atio
n
,
elite
-
g
u
id
ed
ex
p
lo
itatio
n
,
esc
ap
e,
an
d
m
ig
r
atio
n
-
b
ased
r
ep
o
s
itio
n
in
g
to
m
ain
tain
a
v
ar
i
ety
o
f
s
ea
r
ch
an
d
s
tead
y
co
n
v
e
r
g
en
ce
t
h
r
o
u
g
h
o
u
t th
e
o
p
tim
izatio
n
p
r
o
ce
s
s
.
Me
tah
eu
r
is
tic
alg
o
r
ith
m
s
d
if
f
er
in
th
e
b
alan
ce
b
etwe
en
e
x
p
lo
r
atio
n
a
n
d
ex
p
l
o
itatio
n
.
GW
O
an
d
HHO
u
s
e
lead
er
s
h
ip
a
n
d
p
r
e
y
m
o
d
ellin
g
t
o
in
c
r
ea
s
e
ex
p
l
o
itatio
n
,
wh
ile
t
h
e
s
in
e
-
co
s
in
e
alg
o
r
ith
m
(
SC
A)
u
p
d
ates
p
o
s
itio
n
an
d
DE
u
s
e
s
m
u
tatio
n
s
an
d
r
ec
o
m
b
in
ati
o
n
to
k
ee
p
p
o
p
u
latio
n
d
iv
e
r
s
ity
[
5
]
,
[
1
0
]
,
[
2
0
]
.
Ma
r
in
e
p
r
e
d
ato
r
s
alg
o
r
ith
m
(
MP
A)
u
s
es
L
év
y
a
n
d
B
r
o
wn
ian
m
o
tio
n
to
g
u
i
d
e
s
ea
r
ch
b
eh
a
v
io
r
d
u
r
in
g
o
p
tim
izatio
n
an
d
h
as
b
ee
n
p
r
o
v
en
ab
le
to
b
alan
ce
e
x
p
lo
r
a
tio
n
with
ex
p
lo
itatio
n
an
d
ex
p
lo
r
atio
n
—
ev
en
in
co
n
tin
u
o
u
s
o
p
tim
izatio
n
p
r
o
b
lem
s
[
2
0
]
.
T
h
e
R
h
in
o
o
p
tim
i
ze
r
co
m
b
in
es
f
o
u
r
b
eh
a
v
io
r
m
o
d
els
in
to
a
s
in
g
le
d
esig
n
,
s
u
ch
as
lo
n
g
-
r
a
n
g
e
e
x
p
lo
r
atio
n
,
elite
-
d
r
iv
e
n
c
o
n
v
er
g
en
ce
,
ad
a
p
tiv
e
s
tag
n
atio
n
esca
p
e,
an
d
p
o
p
u
latio
n
r
ep
o
s
itio
n
in
g
.
T
h
is
is
ex
p
ec
t
ed
to
m
ai
n
tain
s
ea
r
ch
d
iv
er
s
ity
wh
ile
allo
win
g
f
o
r
s
tab
le
co
n
v
e
r
g
en
ce
with
lim
ited
ev
alu
atio
n
b
u
d
g
ets,
an
d
to
ac
h
iev
e
a
co
n
s
is
ten
t
b
eh
av
io
r
a
l
f
r
am
ew
o
r
k
th
at
b
al
an
ce
s
ex
p
lo
r
atio
n
,
ex
p
lo
itatio
n
,
escap
e,
an
d
g
lo
b
al
r
ep
o
s
itio
n
in
g
t
h
r
o
u
g
h
o
u
t th
e
o
p
tim
izatio
n
p
r
o
ce
s
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
8
1
4
I
n
t J Ad
v
Ap
p
l Sci
,
Vo
l.
1
5
,
No
.
3
,
Sep
tem
b
er
2
0
2
6
:
1
2
3
9
-
1
2
5
1
1242
T
ab
le
1
.
Op
e
r
ato
r
-
lev
el
c
o
m
p
a
r
is
o
n
b
etwe
en
R
h
in
o
o
p
tim
ize
r
an
d
r
ep
r
esen
tativ
e
m
etah
e
u
r
i
s
tic
alg
o
r
ith
m
s
A
l
g
o
r
i
t
h
m
P
r
i
mary
e
x
p
l
o
r
a
t
i
o
n
mec
h
a
n
i
sm
P
r
i
mary
e
x
p
l
o
i
t
a
t
i
o
n
mec
h
a
n
i
sm
Esc
a
p
e
/
st
a
g
n
a
t
i
o
n
a
v
o
i
d
a
n
ce
P
o
p
u
l
a
t
i
o
n
r
e
p
o
si
t
i
o
n
i
n
g
D
i
st
i
n
g
u
i
s
h
i
n
g
c
h
a
r
a
c
t
e
r
i
s
t
i
c
s
G
W
O
S
o
c
i
a
l
h
u
n
t
i
n
g
a
n
d
l
e
a
d
e
r
-
g
u
i
d
e
d
p
o
s
i
t
i
o
n
u
p
d
a
t
e
s
A
l
p
h
a
,
b
e
t
a
,
a
n
d
d
e
l
t
a
l
e
a
d
e
r
s
h
i
p
h
i
e
r
a
r
c
h
y
I
mp
l
i
c
i
t
t
h
r
o
u
g
h
c
o
e
f
f
i
c
i
e
n
t
a
d
a
p
t
a
t
i
o
n
N
o
n
e
Le
a
d
e
r
s
h
i
p
h
i
e
r
a
r
c
h
y
w
i
t
h
c
o
o
p
e
r
a
t
i
v
e
h
u
n
t
i
n
g
b
e
h
a
v
i
o
r
HHO
R
a
n
d
o
m
p
e
r
c
h
i
n
g
a
n
d
su
r
p
r
i
se
-
p
o
u
n
c
e
e
x
p
l
o
r
a
t
i
o
n
P
r
o
g
r
e
ss
i
v
e
b
e
s
i
e
g
e
o
f
p
r
e
y
En
e
r
g
y
-
d
e
p
e
n
d
e
n
t
sw
i
t
c
h
i
n
g
b
e
t
w
e
e
n
e
x
p
l
o
r
a
t
i
o
n
a
n
d
e
x
p
l
o
i
t
a
t
i
o
n
N
o
n
e
D
y
n
a
mi
c
p
r
e
y
-
e
n
e
r
g
y
st
r
a
t
e
g
y
c
o
n
t
r
o
l
l
i
n
g
s
e
a
r
c
h
b
e
h
a
v
i
o
r
M
P
A
Lé
v
y
a
n
d
B
r
o
w
n
i
a
n
mo
t
i
o
n
P
r
e
d
a
t
o
r
–
p
r
e
y
i
n
t
e
r
a
c
t
i
o
n
V
e
l
o
c
i
t
y
a
d
a
p
t
a
t
i
o
n
b
e
t
w
e
e
n
s
e
a
r
c
h
p
h
a
s
e
s
N
o
n
e
S
t
a
g
e
-
d
e
p
e
n
d
e
n
t
se
a
r
c
h
b
e
h
a
v
i
o
r
i
n
s
p
i
r
e
d
b
y
marin
e
p
r
e
d
a
t
o
r
s
W
h
a
l
e
o
p
t
i
m
i
z
a
t
i
o
n
a
l
g
o
r
i
t
h
m
(WOA)
R
a
n
d
o
m s
e
a
r
c
h
a
r
o
u
n
d
p
r
e
y
S
p
i
r
a
l
e
n
c
i
r
c
l
i
n
g
o
f
t
h
e
b
e
st
s
o
l
u
t
i
o
n
R
a
n
d
o
m s
e
a
r
c
h
d
u
r
i
n
g
e
x
p
l
o
r
a
t
i
o
n
N
o
n
e
B
u
b
b
l
e
-
n
e
t
a
t
t
a
c
k
i
n
g
mec
h
a
n
i
sm
DE
P
o
p
u
l
a
t
i
o
n
mu
t
a
t
i
o
n
u
si
n
g
v
e
c
t
o
r
d
i
f
f
e
r
e
n
c
e
s
G
r
e
e
d
y
s
e
l
e
c
t
i
o
n
o
f
i
mp
r
o
v
e
d
c
a
n
d
i
d
a
t
e
s
D
i
v
e
r
s
i
t
y
mai
n
t
a
i
n
e
d
t
h
r
o
u
g
h
m
u
t
a
t
i
o
n
a
n
d
c
r
o
ss
o
v
e
r
N
o
n
e
D
i
f
f
e
r
e
n
c
e
-
v
e
c
t
o
r
mu
t
a
t
i
o
n
s
t
r
a
t
e
g
y
S
C
A
O
sci
l
l
a
t
o
r
y
si
n
e
–
c
o
s
i
n
e
p
o
si
t
i
o
n
u
p
d
a
t
e
s
A
t
t
r
a
c
t
i
o
n
t
o
w
a
r
d
s
t
h
e
c
u
r
r
e
n
t
b
e
st
so
l
u
t
i
o
n
O
sci
l
l
a
t
o
r
y
mo
v
e
me
n
t
a
r
o
u
n
d
e
l
i
t
e
so
l
u
t
i
o
n
s
N
o
n
e
Tr
i
g
o
n
o
m
e
t
r
i
c
p
o
si
t
i
o
n
u
p
d
a
t
i
n
g
R
h
i
n
o
o
p
t
i
m
i
z
e
r
(
p
r
o
p
o
s
e
d
)
Lé
v
y
-
f
l
i
g
h
t
e
x
p
l
o
r
a
t
i
o
n
i
n
s
p
i
r
e
d
b
y
l
o
n
g
-
r
a
n
g
e
r
e
so
u
r
c
e
sea
r
c
h
i
n
g
El
i
t
e
-
g
u
i
d
e
d
c
o
n
v
e
r
g
e
n
c
e
e
n
h
a
n
c
e
d
w
i
t
h
mo
me
n
t
u
m
W
a
l
l
o
w
i
n
g
-
i
n
s
p
i
r
e
d
a
d
a
p
t
i
v
e
e
sca
p
e
p
e
r
t
u
r
b
a
t
i
o
n
M
i
g
r
a
t
i
o
n
-
b
a
s
e
d
r
e
p
o
si
t
i
o
n
i
n
g
o
f
p
o
o
r
l
y
p
e
r
f
o
r
m
i
n
g
a
g
e
n
t
s
I
n
t
e
g
r
a
t
e
s f
o
u
r
c
o
m
p
l
e
me
n
t
a
r
y
b
e
h
a
v
i
o
r
-
i
n
s
p
i
r
e
d
o
p
e
r
a
t
o
r
s wi
t
h
i
n
a
u
n
i
f
i
e
d
o
p
t
i
mi
z
a
t
i
o
n
f
r
a
mew
o
r
k
t
h
a
t
e
x
p
l
i
c
i
t
l
y
b
a
l
a
n
c
e
s
e
x
p
l
o
r
a
t
i
o
n
,
e
x
p
l
o
i
t
a
t
i
o
n
,
e
sca
p
e
,
a
n
d
g
l
o
b
a
l
r
e
p
o
s
i
t
i
o
n
i
n
g
2
.
3
.
M
a
t
hema
t
ica
l
f
o
rm
ula
t
i
o
n
2
.
3
.
1
.
E
x
plo
ra
t
io
n
-
L
év
y
-
f
lig
ht
I
n
ea
r
ly
v
er
s
io
n
s
,
a
g
en
ts
m
im
i
c
R
h
in
o
ce
r
o
s
lo
n
g
-
r
a
n
g
e
f
o
r
a
g
in
g
b
eh
av
io
r
.
T
h
is
b
eh
av
io
r
i
s
m
o
d
eled
b
y
a
L
év
y
f
lig
h
t,
wh
er
e
a
c
o
m
b
in
atio
n
o
f
s
m
all
an
d
lar
g
e
s
tep
s
izes
ca
n
b
e
tak
en
,
an
d
th
e
a
g
en
ts
ar
e
th
u
s
en
ab
led
to
e
x
p
lo
r
e
lo
ca
lly
an
d
to
ju
m
p
t
o
v
er
y
d
is
ta
n
t lo
ca
tio
n
s
.
=
|
|
1
/
(
1
)
W
h
er
e
an
d
v
ar
e
r
an
d
o
m
n
u
m
b
er
s
d
r
awn
f
r
o
m
n
o
r
m
al
d
is
tr
ib
u
tio
n
s
.
is
th
e
L
é
v
y
e
x
p
o
n
e
n
t,
u
s
u
ally
v
ar
y
in
g
b
etwe
en
1
.
0
to
2
.
0
(
m
o
s
t c
o
m
m
o
n
:
1
.
5
)
.
2
.
3
.
2
.
E
x
plo
it
a
t
io
n
-
elit
e
-
ba
s
e
d c
o
nv
er
g
ence
wit
h m
o
m
ent
um
W
h
en
an
ag
en
t
g
ets
clo
s
er
to
a
p
r
o
m
is
in
g
ar
ea
,
it
m
o
v
es
t
o
a
f
o
cu
s
ed
b
e
h
av
io
r
,
in
s
p
ir
e
d
b
y
R
h
in
o
ch
ar
g
in
g
b
eh
a
v
io
r
.
T
h
is
s
tag
e
co
m
b
in
es b
o
t
h
m
o
m
e
n
tu
m
a
n
d
p
u
ll to
war
d
s
th
e
elite
s
o
lu
tio
n
at
p
r
esen
t.
+
1
=
+
∗
+
∗
(
−
)
(
2
)
W
h
er
e
is
th
e
cu
r
r
en
t
p
o
s
itio
n
;
is
th
e
b
est
s
o
lu
tio
n
s
o
f
a
r
,
o
b
tain
e
d
f
r
o
m
th
e
elite
p
o
o
l
;
is
th
e
ex
p
lo
r
atio
n
co
ef
f
icie
n
t
;
an
d
λ
is
th
e
ex
p
lo
itatio
n
co
e
f
f
icien
t
.
2
.
3
.
3
.
E
s
ca
pe
s
t
ra
t
eg
y
-
wa
llo
wing
beha
v
io
r
T
o
av
o
i
d
r
ea
c
h
in
g
a
p
latea
u
in
lo
ca
l
o
p
tim
a,
a
wallo
win
g
m
ec
h
an
is
m
p
r
o
v
id
es
a
n
ag
en
t
with
p
er
tu
r
b
atio
n
s
wh
en
p
e
r
f
o
r
m
an
ce
r
em
ain
s
co
n
s
tan
t.
T
h
is
is
d
er
iv
ed
f
r
o
m
R
h
in
o
wallo
win
g
b
eh
av
io
r
.
+
1
=
+
∗
(
0
,
1
)
(
3
)
W
h
er
e
is
a
s
m
all
p
er
tu
r
b
atio
n
s
ca
lin
g
f
ac
to
r
an
d
(
0
,
1
)
is
a
s
tan
d
ar
d
n
o
r
m
al
d
is
tr
ib
u
tio
n
.
2
.
3
.
4
.
M
ig
ra
t
io
n
-
g
lo
ba
l r
epo
s
it
io
nin
g
T
o
m
im
ic
m
ig
r
ato
r
y
b
eh
av
io
r
,
s
o
m
e
o
f
th
e
wo
r
s
t
-
p
er
f
o
r
m
i
n
g
ag
en
ts
ar
e
r
an
d
o
m
ly
r
e
p
o
s
itio
n
ed
to
p
r
o
m
o
te
e
x
p
lo
r
atio
n
an
d
r
ed
u
ce
s
tag
n
atio
n
d
u
r
in
g
c
o
n
v
e
r
g
e
n
ce
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
Ad
v
Ap
p
l Sci
I
SS
N:
2252
-
8
8
1
4
R
h
in
o
o
p
timis
er fo
r
h
yp
erp
a
r
a
mete
r
tu
n
in
g
in
d
e
p
r
ess
io
n
d
etec
tio
n
(
S
imis
a
n
i Nd
a
b
a
)
1243
+
1
(
)
=
(
,
)
,
∀
∈
(
4
)
W
h
er
e
,
ar
e
th
e
lo
wer
an
d
u
p
p
er
b
o
u
n
d
s
o
f
th
e
s
ea
r
ch
s
p
ac
e
;
an
d
Mig
r
atio
n
Po
o
l
is
th
e
s
et
o
f
b
o
tto
m
-
p
e
r
f
o
r
m
in
g
a
g
en
ts
(
e.
g
.
,
b
o
tto
m
1
0
%)
.
2
.
4
.
P
s
eudo
co
de
T
h
e
R
h
in
o
o
p
tim
izer
is
s
ch
e
m
atica
lly
ex
p
lain
ed
b
y
th
e
Alg
o
r
ith
m
1
,
wh
ic
h
d
escr
ib
e
s
th
e
b
asi
c
co
n
ce
p
t
o
f
th
is
alg
o
r
ith
m
.
A
n
ap
p
r
o
ac
h
t
h
at
in
co
r
p
o
r
ates
ex
p
lo
r
atio
n
,
ex
p
lo
itatio
n
,
esc
ap
e,
an
d
m
ig
r
atio
n
b
eh
av
io
r
s
in
s
p
ir
e
d
b
y
R
h
in
o
ce
r
o
s
ec
o
lo
g
y
.
A
lg
o
r
ith
m
1
illu
s
tr
ates
h
o
w
R
h
in
o
o
p
tim
izer
b
a
lan
ce
s
ex
p
lo
r
atio
n
an
d
ex
p
lo
itatio
n
wh
ile
g
u
id
in
g
ag
en
ts
to
war
d
s
h
ig
h
-
q
u
ality
s
o
lu
tio
n
s
with
in
th
e
s
ea
r
ch
s
p
a
ce
.
Alg
o
r
ith
m
1.
Ps
eu
d
o
co
d
e
o
f
t
h
e
R
h
in
o
o
p
tim
izer
Start Rhino
o
ptimizer
Input:
–
f(x), the
function to minimize as an objective function
–
N (number of agents), search space bounds
–
Maximum number of iterations (MaxIter)
–
P explore probability (p_explore)
–
Elite fraction (elite_fraction)
–
Migration interval
–
ε
₀
or the strength of the initi
ally imposed disturbance
Initialize:
–
Place N agents randomly in the search space
–
Compute the fitness of each agent
–
Find best solution x_best and its corresponding fitness f_best
–
The elite pool is composed of the top
-
performing agents
–
Set agent mo
menta to zero
For each iteration=1 to MaxIter do:
–
Update ε based on the current iteration of the perturbation
For each agent do:
a. Exploration:
–
With probability p_explore, perform Lévy flight
-
based jump
b. Exploitation:
–
Otherwise, move toward a randomly selected elite agent, with momentum
c. Escape:
–
If no fitness improvement for T iterations, apply Gaussian perturbation
d. Migration:
–
If iteration mod migration_interva
l == 0 and agent in bottom percentile:
→ Reinitialize to a random position
e. Boundary Handling:
–
Reflect or clamp positions outside the search bounds
–
Re
-
evaluate fitness for all agents
–
Update x_best and f_best if a better solution is found
–
Refresh elite pool based on updated performance
Return: Best solution x_best and its fitness f_best
End Rhino
Optimizer
2
.
5
.
Co
m
pu
t
a
t
io
na
l
c
o
m
ple
x
it
y
a
na
ly
s
is
T
h
e
co
m
p
u
tatio
n
al
co
m
p
lex
ity
o
f
th
e
R
h
in
o
o
p
tim
izer
is
p
r
i
m
ar
ily
d
eter
m
in
ed
b
y
th
e
p
o
p
u
latio
n
s
ize
N
an
d
th
e
n
u
m
b
er
o
f
iter
atio
n
s
T
.
I
n
ea
ch
iter
atio
n
,
ea
ch
ag
en
t
p
er
f
o
r
m
s
ex
p
lo
r
atio
n
,
ex
p
l
o
itatio
n
,
escap
e,
o
r
m
ig
r
atio
n
o
p
er
atio
n
s
b
ef
o
r
e
f
i
tn
ess
ev
alu
atio
n
an
d
th
en
e
v
alu
ates
its
f
itn
es
s
.
T
h
e
co
m
p
lex
i
ty
o
f
th
e
alg
o
r
ith
m
i
s
O
(
N
T
)
,
wi
t
h
o
u
t
c
o
s
t
o
f
o
b
je
c
t
i
v
e
f
u
n
c
ti
o
n
e
v
a
l
u
a
ti
o
n
.
A
lth
o
u
g
h
t
h
e
s
e
e
x
p
l
o
r
a
t
i
o
n
,
e
x
p
lo
i
t
a
t
i
o
n
,
w
a
l
l
o
w
i
n
g
,
an
d
m
ig
r
ati
o
n
o
p
er
ato
r
s
all
r
eq
u
ir
e
co
n
s
tan
t
-
tim
e
u
p
d
ates
f
o
r
ea
ch
ag
en
t,
th
ey
d
o
n
o
t
ad
d
an
y
ad
d
itio
n
al
co
m
p
lex
ity
b
e
y
o
n
d
th
o
s
e
o
f
s
t
an
d
ar
d
m
eta
h
eu
r
is
tic
alg
o
r
ith
m
s
b
ased
o
n
p
o
p
u
latio
n
s
lik
e
GW
O
,
HHO
,
M
PA
,
SC
A
,
an
d
DE
.
T
h
e
R
h
in
o
o
p
tim
izer
is
as
co
m
p
lex
as
s
tan
d
ar
d
p
o
p
u
latio
n
-
b
ased
o
p
tim
iza
tio
n
alg
o
r
ith
m
s
b
u
t
with
ad
d
ed
f
u
n
ctio
n
ality
f
o
r
e
x
p
lo
r
atio
n
,
ex
p
l
o
itatio
n
,
s
tag
n
atio
n
av
o
id
a
n
ce
,
an
d
p
o
p
u
latio
n
r
ep
o
s
itio
n
in
g
.
3.
RE
S
E
ARCH
M
E
T
H
O
D
T
h
e
f
o
llo
win
g
s
ec
tio
n
d
escr
i
b
es
th
e
ex
p
er
im
en
tal
d
esig
n
u
s
ed
to
ass
es
s
R
h
in
o
o
p
tim
izer
u
s
ed
f
o
r
HPO
in
th
e
ca
s
e
o
f
d
e
p
r
ess
io
n
d
etec
tio
n
.
Deta
ils
o
n
th
e
co
n
f
ig
u
r
atio
n
o
f
th
e
o
p
tim
izatio
n
,
p
a
r
am
eter
izatio
n
,
ev
alu
atio
n
m
etr
ics,
tr
ain
i
n
g
p
r
o
to
co
l
,
an
d
im
p
lem
e
n
tatio
n
s
p
ec
if
ics aim
in
g
at
tr
an
s
p
ar
en
c
y
an
d
co
n
s
is
ten
cy
.
3
.
1
.
Da
t
a
s
et
T
h
e
m
e
n
tal
h
ea
lth
d
e
p
r
ess
io
n
d
ataset,
p
u
b
licly
av
ailab
le
th
r
o
u
g
h
H
u
g
g
in
g
Face
d
atasets
[
2
1
]
,
is
ch
o
s
en
f
o
r
its
d
ir
ec
t
co
r
r
esp
o
n
d
en
ce
with
th
e
o
b
jectiv
e
o
f
th
i
s
s
tu
d
y
,
wh
ich
is
HPO
f
o
r
au
to
m
ated
d
ep
r
ess
io
n
d
etec
tio
n
in
tex
t.
I
t
i
s
an
n
o
t
at
ed
f
o
r
th
e
b
in
ar
y
tar
g
et
‘
h
as
d
ep
r
ess
io
n
’;
r
ath
er
th
an
a
p
p
ly
in
g
a
g
en
er
a
l
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
8
1
4
I
n
t J Ad
v
Ap
p
l Sci
,
Vo
l.
1
5
,
No
.
3
,
Sep
tem
b
er
2
0
2
6
:
1
2
3
9
-
1
2
5
1
1244
s
en
tim
en
t
o
r
em
o
tio
n
co
r
p
u
s
,
ca
n
h
elp
u
n
d
er
s
tan
d
h
o
w
well
o
p
tim
izatio
n
p
er
f
o
r
m
s
in
a
clin
ically
r
elev
an
t
class
if
icatio
n
p
r
o
b
lem
.
T
h
e
d
ataset
is
s
u
f
f
icien
tly
lar
g
e
f
o
r
f
in
e
-
tu
n
in
g
tr
an
s
f
o
r
m
er
-
b
ased
m
o
d
els
wh
ile
ex
h
ib
itin
g
r
ea
lis
tic
ch
ar
ac
ter
i
s
tics
s
u
ch
as
class
im
b
alan
ce
an
d
lin
g
u
is
tic
d
iv
er
s
ity
.
D
u
e
to
its
m
u
ltil
in
g
u
al
n
atu
r
e
an
d
its
s
m
atter
in
g
o
f
s
o
cial
-
m
ed
ia
p
o
s
ts
,
f
o
r
u
m
m
es
s
ag
es,
an
d
d
iar
y
-
lik
e
r
ef
lectio
n
s
,
th
e
wr
itin
g
s
ty
le
an
d
co
n
tex
ts
d
o
e
x
h
ib
it
s
o
m
e
n
atu
r
al
v
ar
iatio
n
,
m
ak
in
g
i
t
a
d
if
f
icu
lt
b
u
t
r
ele
v
an
t
test
b
ed
f
o
r
o
p
tim
izer
r
o
b
u
s
tn
ess
.
T
h
e
d
ataset
is
av
ailab
le
u
n
d
er
a
C
C
B
Y
4
.
0
licen
s
e
an
d
h
o
s
ted
o
n
Hu
g
g
i
n
g
Face
an
d
ca
n
b
e
d
ir
ec
tly
lo
a
d
ed
with
th
e
d
at
asets
an
d
tr
an
s
f
o
r
m
er
s
lib
r
a
r
ies,
in
cr
ea
s
in
g
ac
ce
s
s
ib
ilit
y
an
d
r
e
p
r
o
d
u
cib
ilit
y
.
Usi
n
g
a
s
in
g
le
b
en
ch
m
a
r
k
d
ataset
also
ac
ts
a
s
a
co
n
tr
o
l
,
h
elp
in
g
t
o
attr
ib
u
te
p
e
r
f
o
r
m
an
ce
d
if
f
e
r
en
ce
s
p
r
im
ar
ily
to
t
h
e
o
p
tim
izatio
n
a
lg
o
r
ith
m
s
th
em
s
elv
es.
Alth
o
u
g
h
t
h
e
d
ataset
p
r
esen
ts
a
r
ea
lis
tic
an
d
d
i
v
er
s
e
s
et
o
f
d
ata
f
o
r
d
ep
r
ess
io
n
d
etec
tio
n
,
it
is
b
ased
o
n
s
o
cial
-
m
e
d
ia/u
s
er
g
en
er
ate
d
tex
t
(
th
e
d
ataset
is
n
o
t
v
alid
ated
b
y
an
o
f
f
icial
p
s
y
ch
iatr
ic
ass
ess
m
en
t)
.
T
h
e
lan
g
u
ag
e
u
s
ed
i
n
th
is
d
ataset
m
ay
d
if
f
er
f
r
o
m
t
h
at
en
co
u
n
te
r
ed
in
f
o
r
m
al
clin
ical
s
ettin
g
s
.
T
h
e
r
esu
lts
r
ep
o
r
ted
in
th
is
s
tu
d
y
s
h
o
u
l
d
b
e
in
ter
p
r
eted
i
n
th
e
c
o
n
te
x
t
o
f
tex
t
-
b
ased
d
ep
r
ess
io
n
d
etec
tio
n
.
F
u
tu
r
e
wo
r
k
s
h
o
u
ld
ev
alu
ate
th
e
R
h
in
o
o
p
tim
izer
with
m
o
r
e
clin
ical
an
d
h
ea
lth
ca
r
e
d
atasets
to
test
th
e
ex
ter
n
al
v
alid
ity
an
d
g
en
er
aliza
b
ilit
y
to
a
r
an
g
e
o
f
m
en
tal
h
ea
lth
ap
p
licatio
n
s
.
3
.
2
.
Da
t
a
c
o
m
po
s
it
io
n
T
h
e
d
ataset
co
m
p
r
is
es
ab
o
u
t
3
7
,
0
0
0
tex
tu
al
e
n
tr
ies,
with
b
in
ar
y
lab
els
co
r
r
esp
o
n
d
in
g
t
o
wh
eth
er
d
ep
r
ess
io
n
is
d
etec
ted
o
r
n
o
t
,
u
s
in
g
class
es
d
ep
r
ess
ed
(
1
)
a
n
d
n
o
n
-
d
e
p
r
ess
ed
(
0
)
.
I
t
is
m
u
ltil
in
g
u
al,
in
E
n
g
lis
h
,
Ger
m
an
,
Fre
n
c
h
,
I
talian
,
Po
r
t
u
g
u
e
se
,
an
d
Sp
an
is
h
.
E
ac
h
w
ill
co
n
tain
a
f
r
ee
-
tex
t
d
o
cu
m
e
n
t
f
ield
,
as
well
as
m
etad
ata
in
clu
d
in
g
,
b
u
t
n
o
t
li
m
ited
to
,
lan
g
u
ag
e,
m
en
tal
-
s
tate
lab
el,
to
p
ic
ca
teg
o
r
y
,
wo
r
d
co
u
n
t,
a
n
d
co
n
tex
t
d
escr
ip
to
r
s
.
T
ex
ts
ar
e
s
o
u
r
ce
d
f
r
o
m
d
is
cu
s
s
io
n
s
o
f
th
em
in
r
elatio
n
to
o
n
e
a
n
o
th
er
o
n
s
o
cial
-
m
ed
ia
s
ites
,
ch
atr
o
o
m
s
,
an
d
in
“d
iar
y
”
e
n
tr
ies
wh
ich
ar
e
co
m
m
en
tar
y
o
n
in
d
i
v
id
u
al
ep
is
o
d
es
ar
e
av
ailab
le,
an
d
th
is
in
f
o
r
m
atio
n
s
h
o
u
l
d
n
o
t
b
e
ig
n
o
r
ed
.
Dis
clo
s
u
r
e
t
h
at
is
s
p
o
n
t
an
eo
u
s
an
d
r
ef
lectiv
e
o
f
th
e
s
elf
.
T
h
er
e
is
a
g
r
ea
t
d
ea
l
o
f
l
in
g
u
is
tic
an
d
co
n
tex
t
u
al
d
iv
er
s
ity
a
m
o
n
g
th
ese
p
ie
ce
s
,
wh
ile
s
till
k
ee
p
in
g
t
h
e
cle
ar
b
in
ar
y
p
r
ed
ictio
n
task
id
ea
l f
o
r
o
p
tim
izatio
n
-
b
as
ed
m
o
d
el
e
v
alu
atio
n
.
3
.
3
.
O
pti
m
iza
t
io
n
s
t
ra
t
eg
y
HPO
in
th
is
s
tu
d
y
was
f
o
r
m
u
lated
as
a
b
lack
-
b
o
x
o
p
tim
izatio
n
p
r
o
b
lem
,
w
h
er
e
th
e
o
b
jectiv
e
f
u
n
ctio
n
r
et
u
r
n
s
th
e
v
alid
atio
n
p
er
f
o
r
m
an
ce
o
f
th
e
d
ep
r
ess
io
n
-
d
etec
tio
n
m
o
d
el
f
o
r
a
g
iv
en
h
y
p
er
p
ar
am
eter
co
n
f
ig
u
r
atio
n
.
T
h
e
ex
p
er
im
e
n
tal
co
n
f
ig
u
r
atio
n
was
d
esig
n
ed
to
ev
alu
ate
th
e
ef
f
ec
tiv
e
n
ess
o
f
th
e
p
r
o
p
o
s
ed
o
p
tim
izer
with
in
a
p
r
ac
tical
h
ea
lth
ca
r
e
ap
p
licatio
n
.
Similar
o
p
tim
izatio
n
-
b
ased
a
p
p
r
o
ac
h
es
h
av
e
p
r
ev
i
o
u
s
ly
b
ee
n
i
n
v
esti
g
ated
f
o
r
co
m
p
u
ter
-
aid
ed
d
e
p
r
ess
io
n
d
etec
t
io
n
a
n
d
b
io
m
e
d
ical
d
ec
is
io
n
-
s
u
p
p
o
r
t
s
y
s
tem
s
,
d
em
o
n
s
tr
atin
g
th
at
in
tellig
en
t
o
p
tim
izatio
n
tech
n
iq
u
es
ca
n
im
p
r
o
v
e
lear
n
in
g
ef
f
icien
cy
an
d
p
r
e
d
ictiv
e
p
er
f
o
r
m
an
ce
in
m
en
tal
h
ea
lth
ap
p
licatio
n
s
[
22]
.
C
o
n
s
eq
u
e
n
tly
,
th
e
s
elec
ted
ex
p
er
im
en
tal
p
r
o
to
co
l p
r
o
v
id
es a
n
ap
p
r
o
p
r
iate
en
v
ir
o
n
m
e
n
t f
o
r
a
s
s
es
s
in
g
th
e
o
p
tim
izatio
n
ca
p
a
b
ilit
y
o
f
th
e
R
h
in
o
o
p
tim
izer
.
T
h
e
R
h
in
o
o
p
tim
izer
a
n
d
al
l
co
m
p
ar
ativ
e
alg
o
r
ith
m
s
ex
p
lo
r
ed
t
h
e
s
am
e
h
y
p
er
p
ar
am
eter
s
p
ac
e
iter
ati
v
ely
.
All
ca
n
d
id
ate
s
o
lu
tio
n
s
r
ep
r
esen
ted
a
co
m
p
let
e
h
y
p
er
p
a
r
am
eter
s
p
ac
e.
Fo
r
ea
ch
ca
n
d
id
ate,
th
e
m
o
d
el
was
tr
ain
ed
o
n
th
e
tr
ai
n
in
g
p
a
r
titi
o
n
,
ev
alu
ated
o
n
t
h
e
v
alid
atio
n
s
et,
an
d
th
e
p
er
f
o
r
m
an
ce
s
co
r
e
was
r
etu
r
n
ed
as
f
itn
ess
.
So
lu
tio
n
s
wer
e
f
u
r
th
e
r
u
p
d
at
ed
u
s
in
g
th
e
alg
o
r
ith
m
’
s
s
ea
r
ch
o
p
er
ato
r
s
u
n
til
th
e
e
v
alu
atio
n
b
u
d
g
et
was
ex
h
a
u
s
ted
.
T
h
e
s
am
e
o
p
tim
izatio
n
p
r
o
ce
d
u
r
e
w
as
u
s
ed
f
o
r
all
m
eth
o
d
s
to
en
s
u
r
e
m
eth
o
d
o
lo
g
ical
f
air
n
ess
an
d
to
c
o
m
p
ar
e
p
er
f
o
r
m
an
ce
d
ir
ec
tly
b
etwe
en
th
em
.
3
.
4
.
Co
m
pa
ra
t
iv
e
o
pti
m
izer
s
T
o
p
r
o
v
id
e
a
r
ig
o
r
o
u
s
an
d
r
e
p
r
o
d
u
cib
le
b
en
c
h
m
ar
k
,
th
e
p
r
o
p
o
s
ed
R
h
in
o
o
p
tim
izer
was
co
m
p
ar
ed
with
a
d
i
v
er
s
e
co
llectio
n
o
f
o
p
tim
izatio
n
alg
o
r
ith
m
s
co
m
p
r
is
in
g
tr
a
d
itio
n
al
s
ea
r
c
h
b
aselin
es,
class
ical
m
etah
eu
r
is
tics
,
an
d
m
o
d
er
n
b
io
-
in
s
p
ir
ed
alg
o
r
ith
m
s
.
T
h
ese
a
lg
o
r
ith
m
s
r
ep
r
esen
t
d
if
f
er
en
t
b
eh
a
v
io
r
al
in
s
p
ir
atio
n
s
an
d
s
ea
r
ch
m
ec
h
an
is
m
s
co
m
m
o
n
l
y
ad
o
p
ted
in
HPO
an
d
co
n
tin
u
o
u
s
o
p
ti
m
izatio
n
r
esear
ch
.
Am
o
n
g
r
ec
en
t
b
io
-
i
n
s
p
ir
ed
a
p
p
r
o
ac
h
es,
th
e
Ma
y
f
ly
alg
o
r
i
th
m
h
as
d
em
o
n
s
tr
ated
co
m
p
e
titi
v
e
o
p
tim
izatio
n
p
er
f
o
r
m
an
ce
th
r
o
u
g
h
ad
ap
ti
v
e
m
atin
g
an
d
f
lig
h
t
b
eh
a
v
io
r
s
,
m
ak
in
g
it
an
a
p
p
r
o
p
r
iate
r
ef
er
en
ce
f
o
r
c
o
m
p
a
r
a
t
i
v
e
e
v
a
l
u
a
t
i
o
n
w
it
h
i
n
t
h
e
p
r
e
s
e
n
t
s
t
u
d
y
[
2
3
]
.
T
a
b
l
e
2
p
r
o
v
i
d
e
s
a
d
e
s
c
r
i
p
t
i
o
n
o
f
c
o
m
p
a
r
a
t
i
v
e
a
l
g
o
r
i
t
h
m
s
,
in
clu
d
in
g
t
h
eir
o
p
tim
izatio
n
c
ateg
o
r
y
,
s
ea
r
c
h
p
r
i
n
cip
le
,
r
o
le
in
ev
alu
atio
n
,
an
d
im
p
lem
en
ta
tio
n
av
ailab
ilit
y
.
A
k
ey
cr
iter
io
n
f
o
r
alg
o
r
ith
m
i
n
clu
s
io
n
was
th
e
p
u
b
lic
av
ailab
ilit
y
o
f
im
p
lem
en
tatio
n
s
in
estab
lis
h
ed
o
p
en
-
s
o
u
r
ce
lib
r
ar
ies
o
r
au
t
h
o
r
-
r
elea
s
ed
r
e
p
o
s
ito
r
ies.
T
h
is
en
s
u
r
ed
tr
an
s
p
ar
en
cy
,
en
ab
led
in
d
ep
e
n
d
e
n
t
r
ep
licatio
n
,
a
n
d
m
in
im
ized
c
o
m
p
etitiv
e
b
ias
ar
is
in
g
f
r
o
m
p
r
iv
ate
o
r
u
n
p
u
b
lis
h
ed
c
o
d
e.
Fu
r
th
er
m
o
r
e,
all
co
m
p
ar
ativ
e
alg
o
r
ith
m
s
wer
e
ex
ec
u
ted
u
n
d
er
an
id
e
n
tical
ev
alu
atio
n
b
u
d
g
et
an
d
h
y
p
er
p
ar
a
m
eter
s
ea
r
ch
s
p
ac
e
to
en
s
u
r
e
a
f
ai
r
an
d
co
n
s
is
ten
t p
er
f
o
r
m
an
ce
co
m
p
ar
is
o
n
.
T
h
e
co
m
p
ar
ativ
e
e
v
alu
atio
n
ag
ain
s
t
m
u
ltip
le
o
p
tim
izatio
n
alg
o
r
ith
m
s
was
d
esig
n
ed
to
en
s
u
r
e
a
b
alan
ce
d
ass
ess
m
en
t
ac
r
o
s
s
d
if
f
er
en
t
o
p
tim
izatio
n
p
ar
ad
ig
m
s
.
R
ec
en
t
b
eh
av
io
r
-
in
s
p
ir
ed
o
p
tim
izer
s
,
in
clu
d
in
g
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
Ad
v
Ap
p
l Sci
I
SS
N:
2252
-
8
8
1
4
R
h
in
o
o
p
timis
er fo
r
h
yp
erp
a
r
a
mete
r
tu
n
in
g
in
d
e
p
r
ess
io
n
d
etec
tio
n
(
S
imis
a
n
i Nd
a
b
a
)
1245
s
ea
l
o
p
tim
izatio
n
,
h
av
e
d
em
o
n
s
tr
ated
co
m
p
et
itiv
e
p
er
f
o
r
m
a
n
ce
o
n
o
p
tim
izatio
n
p
r
o
b
lem
s
b
y
m
ain
tain
i
n
g
a
n
ef
f
ec
tiv
e
b
alan
ce
b
etwe
en
e
x
p
lo
r
atio
n
a
n
d
ex
p
lo
itatio
n
.
I
n
clu
d
in
g
r
ep
r
esen
tativ
e
o
p
tim
izatio
n
s
tr
ateg
ies
th
er
ef
o
r
e
p
r
o
v
id
es
a
m
o
r
e
r
ig
o
r
o
u
s
ass
es
s
m
en
t
o
f
th
e
p
r
o
p
o
s
ed
R
h
in
o
o
p
tim
izer
u
n
d
er
co
m
p
ar
ab
le
ex
p
er
im
en
tal
co
n
d
itio
n
s
[
2
4
]
.
T
ab
le
2
.
C
o
m
p
a
r
ativ
e
o
p
tim
iz
atio
n
alg
o
r
ith
m
s
in
cl
u
d
ed
i
n
th
e
s
tu
d
y
C
a
t
e
g
o
r
y
O
p
t
i
mi
z
e
r
C
o
r
e
p
r
i
n
c
i
p
l
e
R
o
l
e
i
n
t
h
i
s st
u
d
y
Tr
a
d
i
t
i
o
n
a
l
b
a
s
e
l
i
n
e
s
R
a
n
d
o
m s
e
a
r
c
h
U
n
i
f
o
r
m
r
a
n
d
o
m
sam
p
l
i
n
g
o
f
h
y
p
e
r
p
a
r
a
me
t
e
r
c
o
n
f
i
g
u
r
a
t
i
o
n
s
f
r
o
m
t
h
e
se
a
r
c
h
s
p
a
c
e
Lo
w
e
r
-
b
o
u
n
d
o
p
t
i
mi
z
a
t
i
o
n
b
a
se
l
i
n
e
G
r
i
d
s
e
a
r
c
h
Ex
h
a
u
st
i
v
e
e
n
u
m
e
r
a
t
i
o
n
o
v
e
r
a
d
i
s
c
r
e
t
i
z
e
d
h
y
p
e
r
p
a
r
a
m
e
t
e
r
g
r
i
d
W
i
d
e
l
y
u
s
e
d
d
e
s
p
i
t
e
p
o
o
r
sca
l
a
b
i
l
i
t
y
;
i
n
c
l
u
d
e
d
f
o
r
met
h
o
d
o
l
o
g
i
c
a
l
c
o
m
p
l
e
t
e
n
e
ss
.
A
s
i
n
[
2
5
]
,
t
h
e
g
r
i
d
sea
r
c
h
c
o
u
p
l
e
d
w
i
t
h
c
r
o
ss
-
v
a
l
i
d
a
t
i
o
n
h
a
s
b
e
e
n
a
p
p
l
i
e
d
f
o
r
a
l
l
h
y
p
e
r
p
a
r
a
me
t
e
r
c
o
mb
i
n
a
t
i
o
n
s
o
n
c
o
n
v
o
l
u
t
i
o
n
a
l
n
e
u
r
a
l
n
e
t
w
o
r
k
(
C
N
N
)
,
r
e
c
u
r
r
e
n
t
n
e
u
r
a
l
n
e
t
w
o
r
k
(
R
N
N
)
,
a
n
d
l
o
n
g
s
h
o
r
t
-
t
e
r
m
memo
r
y
(
LST
M
)
m
o
d
e
l
s
,
c
h
o
o
si
n
g
t
h
e
b
e
s
t
-
p
e
r
f
o
r
m
i
n
g
c
o
n
f
i
g
u
r
a
t
i
o
n
i
n
o
r
d
e
r
t
o
r
e
d
u
c
e
o
v
e
r
f
i
t
t
i
n
g
a
n
d
i
mp
r
o
v
e
g
e
n
e
r
a
l
i
z
a
t
i
o
n
.
B
a
y
e
s
i
a
n
o
p
t
i
m
i
z
a
t
i
o
n
P
r
o
b
a
b
i
l
i
st
i
c
s
u
r
r
o
g
a
t
e
m
o
d
e
l
l
i
n
g
(
t
y
p
i
c
a
l
l
y
G
a
u
ssi
a
n
p
r
o
c
e
sses)
w
i
t
h
a
c
q
u
i
si
t
i
o
n
f
u
n
c
t
i
o
n
s
f
o
r
se
q
u
e
n
t
i
a
l
sea
r
c
h
M
o
d
e
l
-
b
a
se
d
o
p
t
i
mi
z
a
t
i
o
n
b
a
se
l
i
n
e
C
l
a
s
si
c
a
l
met
a
h
e
u
r
i
s
t
i
c
s
G
e
n
e
t
i
c
a
l
g
o
r
i
t
h
m (G
A
)
Ev
o
l
u
t
i
o
n
a
r
y
se
a
r
c
h
v
i
a
s
e
l
e
c
t
i
o
n
,
c
r
o
ss
o
v
e
r
,
a
n
d
m
u
t
a
t
i
o
n
H
i
st
o
r
i
c
a
l
l
y
i
m
p
o
r
t
a
n
t
o
p
t
i
m
i
z
a
t
i
o
n
b
a
se
l
i
n
e
;
p
o
p
u
l
a
t
i
o
n
-
b
a
se
d
e
v
o
l
u
t
i
o
n
a
r
y
sea
r
c
h
P
a
r
t
i
c
l
e
sw
a
r
m
o
p
t
i
m
i
z
a
t
i
o
n
(PSO)
S
w
a
r
m
i
n
t
e
l
l
i
g
e
n
c
e
g
u
i
d
e
d
b
y
p
e
r
s
o
n
a
l
a
n
d
g
l
o
b
a
l
b
e
s
t
s
o
l
u
t
i
o
n
s
P
o
p
u
l
a
r
a
n
d
e
f
f
i
c
i
e
n
t
c
o
n
t
i
n
u
o
u
s
o
p
t
i
mi
z
e
r
;
r
e
p
r
e
s
e
n
t
a
t
i
v
e
sw
a
r
m
-
b
a
se
d
m
e
t
h
o
d
DE
V
e
c
t
o
r
-
d
i
f
f
e
r
e
n
c
e
–
b
a
se
d
mu
t
a
t
i
o
n
a
n
d
r
e
c
o
m
b
i
n
a
t
i
o
n
S
t
r
o
n
g
c
o
n
t
i
n
u
o
u
s
o
p
t
i
mi
z
a
t
i
o
n
b
a
sel
i
n
e
w
i
t
h
r
o
b
u
st
g
l
o
b
a
l
s
e
a
r
c
h
b
e
h
a
v
i
o
r
M
o
d
e
r
n
b
i
o
-
i
n
sp
i
r
e
d
a
l
g
o
r
i
t
h
ms
M
P
A
P
r
e
d
a
t
o
r
–
p
r
e
y
d
y
n
a
m
i
c
s
u
s
i
n
g
Lé
v
y
a
n
d
B
r
o
w
n
i
a
n
mo
t
i
o
n
C
o
n
t
e
m
p
o
r
a
r
y
b
i
o
-
i
n
sp
i
r
e
d
o
p
t
i
m
i
z
e
r
e
mp
h
a
s
i
z
i
n
g
e
x
p
l
o
r
a
t
i
o
n
–
e
x
p
l
o
i
t
a
t
i
o
n
b
a
l
a
n
c
e
HHO
C
o
o
p
e
r
a
t
i
v
e
s
u
r
p
r
i
se
-
p
o
u
n
c
e
h
u
n
t
i
n
g
st
r
a
t
e
g
i
e
s
M
o
d
e
r
n
me
t
a
h
e
u
r
i
st
i
c
w
i
t
h
a
d
a
p
t
i
v
e
e
x
p
l
o
i
t
a
t
i
o
n
p
h
a
ses
G
W
O
Le
a
d
e
r
s
h
i
p
h
i
e
r
a
r
c
h
y
a
n
d
p
a
c
k
h
u
n
t
i
n
g
b
e
h
a
v
i
o
r
W
i
d
e
l
y
u
se
d
mo
d
e
r
n
b
i
o
-
i
n
s
p
i
r
e
d
b
e
n
c
h
mar
k
w
i
t
h
st
r
o
n
g
c
o
n
v
e
r
g
e
n
c
e
c
h
a
r
a
c
t
e
r
i
s
t
i
c
s
S
a
l
p
sw
a
r
m
a
l
g
o
r
i
t
h
m
(SSA)
Ch
a
i
n
-
b
a
se
d
mo
v
e
me
n
t
t
o
w
a
r
d
f
o
o
d
so
u
r
c
e
s
R
e
p
r
e
se
n
t
a
t
i
v
e
sw
a
r
m
-
b
a
s
e
d
b
i
o
-
i
n
sp
i
r
e
d
o
p
t
i
mi
z
e
r
S
C
A
S
i
n
u
s
o
i
d
a
l
a
n
d
c
o
s
i
n
u
s
o
i
d
a
l
p
o
s
i
t
i
o
n
u
p
d
a
t
e
s
a
r
o
u
n
d
e
l
i
t
e
s
o
l
u
t
i
o
n
s
O
sci
l
l
a
t
o
r
y
se
a
r
c
h
me
t
h
o
d
e
f
f
e
c
t
i
v
e
u
n
d
e
r
c
o
n
st
r
a
i
n
e
d
b
u
d
g
e
t
s
3
.
5
.
H
y
perpa
ra
m
e
t
er
s
o
ptim
ized
Op
tim
izatio
n
was
p
er
f
o
r
m
ed
f
o
r
a
s
ea
r
ch
s
p
ac
e
d
ef
in
ed
b
y
s
ev
en
h
y
p
er
p
ar
am
eter
s
wh
ich
in
f
lu
en
ce
tr
an
s
f
o
r
m
er
-
b
ased
d
ep
r
ess
io
n
d
etec
tio
n
:
lear
n
in
g
r
ate,
b
at
ch
s
ize,
n
u
m
b
e
r
o
f
e
p
o
ch
s
,
d
r
o
p
o
u
t
r
ate,
weig
h
t
d
ec
ay
,
war
m
-
u
p
r
atio
,
a
n
d
m
ax
im
u
m
s
eq
u
en
ce
len
g
th
.
C
an
d
id
ate
s
o
lu
tio
n
s
wer
e
r
ep
r
es
en
ted
as
n
o
r
m
alize
d
co
n
tin
u
o
u
s
v
alu
es
an
d
m
a
p
p
e
d
b
ac
k
to
th
e
o
r
i
g
in
al
co
n
tin
u
o
u
s
o
r
d
is
cr
ete
v
alu
es
p
r
io
r
to
m
o
d
el
t
r
ain
in
g
an
d
ev
alu
atio
n
to
h
elp
c
o
m
p
a
r
is
o
n
with
th
e
p
o
p
u
latio
n
-
b
ased
o
p
tim
izer
s
an
d
m
itig
ate
s
c
alin
g
b
ias
b
etwe
en
h
y
p
er
p
ar
am
eter
s
.
T
h
e
s
ea
r
ch
-
s
p
ac
e
d
ef
in
itio
n
is
s
u
m
m
ar
ized
in
T
ab
le
3
.
W
h
ile
th
e
ca
n
d
id
ate
s
o
lu
tio
n
s
wer
e
in
ter
n
ally
r
e
p
r
esen
ted
u
s
in
g
n
o
r
m
alize
d
v
alu
es,
we
r
em
ap
p
e
d
ev
er
y
h
y
p
e
r
p
ar
am
eter
b
ac
k
t
o
its
o
r
ig
in
al
r
an
g
e
b
e
f
o
r
e
te
s
tin
g
an
d
t
r
ain
in
g
th
e
m
o
d
els
.
T
h
e
s
ea
r
c
h
s
p
ac
e
in
clu
d
ed
lear
n
in
g
r
ate
[
1
×1
0
-
5
,
5
×1
0
-
5
]
,
b
atch
s
ize
{1
6
,
3
2
}
,
tr
ain
in
g
tim
es
{2
,
3
,
4
},
d
r
o
p
o
u
t
r
ate
[
0
.
0
5
,
0
.
3
0
]
,
weig
h
t
d
ec
ay
[
0
.
0
1
,
0
.
2
0
]
,
w
ar
m
-
u
p
r
atio
{0
.
0
,
0
.
1
},
an
d
m
ax
im
u
m
s
eq
u
en
ce
len
g
th
{
6
4
,
1
2
8
,
2
5
6
}.
E
ac
h
d
im
en
s
io
n
was
n
o
r
m
alize
d
to
th
e
in
ter
v
al
[
0
,
1
]
an
d
s
u
b
s
eq
u
en
tly
m
ap
p
e
d
b
ac
k
t
o
its
o
r
ig
in
al
s
ca
le
d
u
r
in
g
m
o
d
el
ev
alu
atio
n
.
T
h
is
r
ep
r
es
en
tatio
n
en
s
u
r
es
co
m
p
atib
ilit
y
with
p
o
p
u
latio
n
-
b
ased
m
etah
eu
r
is
tic
alg
o
r
ith
m
s
an
d
r
ed
u
ce
s
s
ca
lin
g
b
ias b
etw
ee
n
h
y
p
er
p
ar
a
m
et
er
s
with
d
if
f
er
in
g
m
ag
n
itu
d
es
.
3
.
6
.
P
a
ra
m
et
er
s
et
t
ing
s
o
f
o
ptim
izer
s
All
alg
o
r
ith
m
s
,
in
clu
d
in
g
R
h
i
n
o
an
d
co
m
p
ar
ativ
e
o
p
tim
izer
s
,
wer
e
r
u
n
u
n
d
er
a
n
id
e
n
tical
ev
alu
atio
n
b
u
d
g
et
(
i.e
.
,
th
e
s
am
e
n
u
m
b
er
o
f
tr
ain
e
d
m
o
d
els).
Po
p
u
latio
n
s
ize,
n
u
m
b
e
r
o
f
iter
atio
n
s
,
an
d
co
n
t
r
o
l
co
ef
f
icien
ts
wer
e
s
et
f
o
llo
wi
n
g
co
m
m
o
n
p
r
ac
tices
in
th
e
liter
atu
r
e
o
r
d
ef
au
lt
r
ec
o
m
m
en
d
atio
n
s
wh
er
e
ap
p
r
o
p
r
iate.
T
h
is
d
esig
n
p
r
ev
e
n
ts
p
er
f
o
r
m
an
ce
u
n
f
air
n
ess
ca
u
s
ed
b
y
u
n
eq
u
al
co
m
p
u
tatio
n
a
l e
f
f
o
r
t.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
8
1
4
I
n
t J Ad
v
Ap
p
l Sci
,
Vo
l.
1
5
,
No
.
3
,
Sep
tem
b
er
2
0
2
6
:
1
2
3
9
-
1
2
5
1
1246
T
ab
le
3
.
Hy
p
er
p
ar
a
m
eter
s
ea
r
ch
s
p
ac
e
d
ef
in
itio
n
N
o
.
H
y
p
e
r
p
a
r
a
me
t
e
r
En
c
o
d
i
n
g
/
r
a
n
g
e
R
a
t
i
o
n
a
l
e
1
Le
a
r
n
i
n
g
r
a
t
e
C
o
n
t
i
n
u
o
u
s,
n
o
r
mal
i
z
e
d
t
o
(
[
0
,
1
]
)
S
t
r
o
n
g
l
y
a
f
f
e
c
t
s
c
o
n
v
e
r
g
e
n
c
e
s
p
e
e
d
a
n
d
t
r
a
i
n
i
n
g
st
a
b
i
l
i
t
y
.
2
B
a
t
c
h
si
z
e
D
i
scret
e
,
e
n
c
o
d
e
d
a
s
c
o
n
t
i
n
u
o
u
s
I
n
f
l
u
e
n
c
e
s
g
r
a
d
i
e
n
t
q
u
a
l
i
t
y
,
s
t
a
b
i
l
i
t
y
,
a
n
d
c
o
mp
u
t
a
t
i
o
n
a
l
e
f
f
i
c
i
e
n
c
y
.
3
N
u
mb
e
r
o
f
e
p
o
c
h
s
D
i
scret
e
,
e
n
c
o
d
e
d
a
s
c
o
n
t
i
n
u
o
u
s
C
o
n
t
r
o
l
s
t
r
a
i
n
i
n
g
d
u
r
a
t
i
o
n
;
a
f
f
e
c
t
s
u
n
d
e
r
f
i
t
t
i
n
g
a
n
d
o
v
e
r
f
i
t
t
i
n
g
.
4
D
r
o
p
o
u
t
r
a
t
e
C
o
n
t
i
n
u
o
u
s ([
0
,
1
]
)
R
e
g
u
l
a
r
i
z
a
t
i
o
n
t
o
r
e
d
u
c
e
o
v
e
r
f
i
t
t
i
n
g
a
n
d
i
m
p
r
o
v
e
g
e
n
e
r
a
l
i
z
a
t
i
o
n
.
5
W
e
i
g
h
t
d
e
c
a
y
C
o
n
t
i
n
u
o
u
s,
l
o
g
-
sc
a
l
e
d
P
e
n
a
l
i
z
e
s
l
a
r
g
e
w
e
i
g
h
t
s
t
o
e
n
h
a
n
c
e
r
o
b
u
s
t
n
e
ss.
6
W
a
r
m
-
u
p
r
a
t
i
o
C
o
n
t
i
n
u
o
u
s ([
0
,
1
]
)
S
t
a
b
i
l
i
z
e
s
e
a
r
l
y
t
r
a
i
n
i
n
g
d
y
n
a
mi
c
s fo
r
a
d
a
p
t
i
v
e
o
p
t
i
mi
z
e
r
s
.
7
M
a
x
i
m
u
m
se
q
u
e
n
c
e
l
e
n
g
t
h
D
i
scret
e
,
e
n
c
o
d
e
d
a
s
c
o
n
t
i
n
u
o
u
s
B
a
l
a
n
c
e
s
r
e
p
r
e
se
n
t
a
t
i
o
n
a
l
c
a
p
a
c
i
t
y
a
n
d
c
o
m
p
u
t
a
t
i
o
n
a
l
c
o
s
t
.
3
.
7
.
M
o
del
co
nfig
ura
t
io
n
A
tr
an
s
f
o
r
m
er
-
b
ased
d
e
p
r
ess
io
n
d
etec
tio
n
m
o
d
el
was
u
s
ed
,
e.
g
.
,
d
is
till
ed
b
id
ir
ec
tio
n
al
en
co
d
er
r
ep
r
esen
tatio
n
s
f
r
o
m
tr
a
n
s
f
o
r
m
er
s
(
Dis
tilB
E
R
T
)
/B
E
R
T
.
T
h
e
m
o
d
el
ar
ch
itectu
r
e
its
elf
was
n
o
t
m
o
d
if
ied
;
th
er
ef
o
r
e,
n
o
a
d
d
itio
n
al
a
r
ch
itectu
r
al
co
n
f
ig
u
r
atio
n
is
r
ep
o
r
ted
.
HPO
f
o
cu
s
ed
s
o
lely
o
n
tr
ain
in
g
-
r
elate
d
p
ar
am
eter
s
r
ath
er
t
h
an
m
o
d
el
r
ed
esig
n
,
wh
ich
alig
n
s
with
co
m
m
o
n
p
r
ac
tice
in
HPO
s
tu
d
ies.
3
.
8
.
T
ra
ini
ng
a
nd
ev
a
lua
t
io
n pro
t
o
co
l
T
h
e
d
ataset
was
p
ar
titi
o
n
ed
i
n
to
tr
ain
i
n
g
,
v
alid
atio
n
,
an
d
t
est
s
ets
u
s
in
g
a
f
i
x
ed
r
an
d
o
m
s
ee
d
f
o
r
r
ep
r
o
d
u
cib
ilit
y
.
Du
r
in
g
o
p
tim
i
za
tio
n
:
i)
th
e
o
p
tim
izer
o
b
s
er
v
ed
o
n
l
y
th
e
v
alid
atio
n
r
esu
lts
;
ii)
th
e
test
s
et
was
u
s
ed
o
n
ce
af
ter
o
p
tim
izatio
n
t
o
r
ep
o
r
t
f
in
al
p
e
r
f
o
r
m
an
ce
;
an
d
iii)
m
u
ltip
le
in
d
ep
en
d
e
n
t
r
u
n
s
wer
e
co
n
d
u
cted
to
ac
co
u
n
t
f
o
r
s
t
o
ch
asti
c
v
ar
i
ab
ilit
y
.
E
ar
ly
s
to
p
p
i
n
g
an
d
c
h
ec
k
p
o
i
n
tin
g
wer
e
a
p
p
lied
wh
er
e
a
p
p
r
o
p
r
iate
t
o
p
r
ev
en
t
o
v
er
f
itti
n
g
an
d
c
o
m
p
u
tatio
n
al
waste.
3
.
9
.
P
re
-
pro
ce
s
s
ing
s
t
eps
Stan
d
ar
d
tex
t
p
r
e
-
p
r
o
ce
s
s
in
g
was
p
er
f
o
r
m
ed
b
e
f
o
r
e
m
o
d
el
tr
ain
in
g
.
T
h
is
p
r
o
ce
s
s
in
clu
d
ed
lo
wer
-
ca
s
in
g
,
to
k
e
n
izatio
n
u
s
i
n
g
th
e
tr
an
s
f
o
r
m
er
to
k
en
izer
,
s
eq
u
en
ce
tr
u
n
ca
tio
n
o
r
p
ad
d
i
n
g
,
an
d
th
e
r
em
o
v
al
o
f
em
p
ty
o
r
co
r
r
u
p
te
d
s
am
p
le
s
.
No
m
an
u
al
f
ea
tu
r
e
e
n
g
in
ee
r
in
g
was
ap
p
lied
b
ec
a
u
s
e
r
ep
r
esen
tatio
n
lear
n
in
g
was h
an
d
led
b
y
th
e
p
r
etr
ain
ed
lan
g
u
ag
e
m
o
d
el.
3
.
1
0
.
P
er
f
o
r
m
a
nce
ev
a
lua
t
io
n m
et
rics
Per
f
o
r
m
an
ce
was
m
ea
s
u
r
ed
th
r
o
u
g
h
d
if
f
er
e
n
t
p
er
f
o
r
m
an
ce
m
etr
ics
b
ased
o
n
th
e
o
p
tim
izatio
n
o
b
jectiv
e
as
well
as
th
e
p
r
ed
ictiv
e
p
r
o
p
er
ties
o
f
t
h
e
d
er
i
v
ed
m
o
d
els
f
o
r
th
e
task
o
f
d
e
tectin
g
d
ep
r
ess
io
n
.
B
ec
au
s
e
o
f
th
e
co
m
m
o
n
class
im
b
alan
ce
in
d
ataset
s
f
o
r
d
ep
r
ess
io
n
d
ata
,
in
ad
d
i
tio
n
to
p
r
ec
is
io
n
,
s
en
s
itiv
ity
-
s
p
ec
if
icity
m
etr
ics
wer
e
co
n
s
id
er
e
d
m
o
r
e
s
u
itab
le.
T
h
e
m
ai
n
o
p
tim
izatio
n
o
b
jectiv
e
was
th
e
F1
-
s
co
r
e
,
wh
ich
is
th
e
h
ar
m
o
n
ic
m
ea
n
o
f
p
r
ec
is
io
n
an
d
r
ec
all
;
it
r
ep
r
esen
ts
a
b
alan
ce
o
f
tr
ad
e
-
o
f
f
in
th
e
class
if
icatio
n
;
th
u
s
,
it
ev
alu
ate
s
th
e
tr
ad
e
-
o
f
f
b
etwe
en
s
en
s
itiv
ity
to
d
etec
t
ca
s
es
o
f
d
ep
r
ess
io
n
an
d
th
e
s
u
cc
ess
in
av
o
id
in
g
f
alse a
lar
m
s
,
wh
ic
h
we
b
eliev
e
is
p
ar
ticu
lar
l
y
im
p
o
r
tan
t in
t
h
e
co
n
te
x
t o
f
m
en
t
al
h
ea
lth
s
cr
ee
n
in
g
.
Fo
r
f
in
al
ev
alu
atio
n
o
n
th
e
h
eld
-
o
u
t
test
s
et,
th
e
f
o
llo
win
g
m
etr
ics
wer
e
r
ep
o
r
ted
f
o
r
ea
ch
o
p
tim
izatio
n
m
eth
o
d
:
test
ac
cu
r
ac
y
,
test
p
r
ec
is
io
n
,
test
r
ec
all,
an
d
b
est
o
b
s
er
v
ed
test
F1
-
s
co
r
e.
Acc
u
r
ac
y
is
a
m
ea
s
u
r
e
o
f
th
e
ac
cu
r
ac
y
o
f
cla
s
s
if
icatio
n
,
an
d
p
r
ec
is
io
n
a
n
d
r
ec
all
ar
e
two
c
o
m
p
lem
e
n
tar
y
m
ea
s
u
r
es
o
f
m
o
d
el
b
eh
av
io
r
.
T
h
is
s
tu
d
y
r
an
k
ed
t
h
e
o
p
tim
izer
s
b
ased
o
n
th
eir
h
ig
h
est
o
b
s
er
v
e
d
test
F1
-
s
co
r
e
to
ac
co
u
n
t
f
o
r
th
e
co
n
s
is
ten
cy
b
etwe
en
th
e
o
p
ti
m
izatio
n
g
o
al
a
n
d
th
e
f
in
al
co
m
p
ar
ativ
e
test
.
3
.
1
1
.
I
m
plem
ent
a
t
i
o
n
det
a
ils
All
ex
p
er
im
en
ts
wer
e
p
er
f
o
r
m
ed
in
Py
th
o
n
u
s
in
g
Py
T
o
r
c
h
,
Hu
g
g
i
n
g
Face
t
r
an
s
f
o
r
m
er
s
,
an
d
ML
to
o
ls
.
T
r
ain
-
u
p
an
d
ev
alu
atio
n
wer
e
p
er
f
o
r
m
ed
o
n
a
g
r
ap
h
i
cs
p
r
o
ce
s
s
in
g
u
n
it
(
GPU
)
,
u
s
in
g
m
i
x
ed
p
r
ec
is
io
n
tr
ain
in
g
wh
en
s
u
p
p
o
r
ted
.
Fix
ed
r
an
d
o
m
s
ee
d
s
an
d
d
ata
p
a
r
titi
o
n
s
wer
e
u
s
ed
to
en
s
u
r
e
co
n
s
is
ten
cy
ac
r
o
s
s
o
p
tim
izer
s
.
Mo
d
el
tr
ain
in
g
a
n
d
ev
alu
atio
n
w
er
e
c
o
n
d
u
cted
u
s
in
g
th
e
Hu
g
g
in
g
Face
t
r
ain
e
r
to
o
l.
T
o
en
s
u
r
e
f
air
co
m
p
ar
is
o
n
,
all
o
p
tim
izer
s
h
ad
th
e
s
am
e
e
v
alu
atio
n
b
u
d
g
et
o
f
3
0
m
o
d
el
e
v
alu
atio
n
s
,
wh
ile
th
e
R
h
in
o
o
p
tim
izer
h
ad
a
p
o
p
u
latio
n
s
i
ze
o
f
s
ix
ag
e
n
ts
an
d
f
o
u
r
iter
a
tio
n
s
.
B
est
r
esu
lts
,
p
r
ed
ictio
n
s
,
an
d
ex
p
er
im
en
tal
o
u
tp
u
t w
er
e
e
x
p
o
r
ted
as C
SV f
iles
in
o
r
d
er
t
o
s
u
p
p
o
r
t r
ep
r
o
d
u
cib
ilit
y
an
d
o
p
en
s
cien
ce
.
4.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
T
h
e
p
r
o
p
o
s
ed
R
h
in
o
o
p
tim
i
ze
r
was
ev
alu
ated
as
an
HPO
m
eth
o
d
f
o
r
d
ep
r
ess
io
n
-
d
etec
tio
n
class
if
icatio
n
u
s
in
g
a
tr
an
s
f
o
r
m
er
m
o
d
el.
I
n
lin
e
with
th
e
s
ec
o
n
d
o
b
jectiv
e
o
f
th
is
s
tu
d
y
,
o
p
tim
izatio
n
ef
f
ec
tiv
en
ess
was
as
s
es
s
e
d
u
s
i
n
g
th
e
test
F1
-
s
co
r
e
as
th
e
p
r
i
n
cip
al
o
b
jectiv
e,
d
u
e
to
class
im
b
alan
ce
an
d
th
e
n
ee
d
to
b
ala
n
ce
p
r
ec
is
io
n
a
n
d
r
ec
all.
Acc
u
r
ac
y
,
p
r
ec
is
io
n
,
an
d
r
ec
all
ar
e
a
d
d
itio
n
ally
r
e
p
o
r
ted
to
p
r
o
v
id
e
a
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
Ad
v
Ap
p
l Sci
I
SS
N:
2252
-
8
8
1
4
R
h
in
o
o
p
timis
er fo
r
h
yp
erp
a
r
a
mete
r
tu
n
in
g
in
d
e
p
r
ess
io
n
d
etec
tio
n
(
S
imis
a
n
i Nd
a
b
a
)
1247
co
m
p
r
eh
e
n
s
iv
e
v
iew
o
f
class
if
icatio
n
p
er
f
o
r
m
an
ce
,
alth
o
u
g
h
th
e
y
wer
e
n
o
t
o
p
tim
izatio
n
tar
g
ets.
T
ab
le
4
s
u
m
m
ar
izes
th
e
b
est
class
if
icatio
n
r
esu
lts
d
is
co
v
er
ed
b
y
ea
ch
o
p
tim
izer
with
in
th
e
f
ix
ed
ev
alu
atio
n
b
u
d
g
et.
T
h
e
SC
A
ac
h
iev
ed
th
e
h
ig
h
e
s
t
F1
-
s
co
r
e
(
0
.
8
0
1
6
)
,
wh
ile
th
e
R
h
in
o
o
p
tim
izer
attain
ed
th
e
h
ig
h
est
o
b
s
er
v
e
d
F1
-
s
co
r
e
o
f
0
.
8
0
0
5
,
wh
ic
h
is
s
tatis
tical
ly
in
d
is
tin
g
u
is
h
ab
le
f
r
o
m
b
o
t
h
SC
A
an
d
GW
O.
T
h
is
r
esu
lt
d
ir
ec
tly
ad
d
r
ess
es
th
e
f
ir
s
t
o
b
jectiv
e
b
y
d
em
o
n
s
tr
atin
g
th
at
R
h
in
o
is
ef
f
ec
tiv
e
at
id
en
tif
y
in
g
h
ig
h
-
q
u
ality
h
y
p
er
p
ar
am
eter
c
o
n
f
ig
u
r
atio
n
s
an
d
is
co
m
p
etitiv
e
with
e
s
tab
lis
h
ed
s
tate
-
of
-
th
e
-
ar
t
o
p
ti
m
izatio
n
m
eth
o
d
s
.
B
ay
esian
o
p
tim
izatio
n
ac
h
iev
ed
a
lo
wer
F1
-
s
co
r
e
(
0
.
7
9
0
3
)
,
wh
ile
class
ical
alg
o
r
ith
m
s
s
u
ch
as
GA
an
d
PS
O
ex
h
ib
ited
wea
k
er
p
er
f
o
r
m
a
n
ce
u
n
d
er
th
e
s
am
e
b
u
d
g
et.
T
ab
le
4
.
C
o
m
p
a
r
ativ
e
class
if
icatio
n
p
er
f
o
r
m
an
ce
an
d
F1
-
b
as
ed
r
an
k
i
n
g
o
f
o
p
tim
izatio
n
m
e
th
o
d
s
f
o
r
d
ep
r
ess
io
n
d
etec
tio
n
O
p
t
i
mi
z
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
-
sc
o
r
e
(
b
e
st
)
R
a
n
k
(
F
1
)
S
C
A
0
.
8
1
8
8
0
.
8
1
6
2
0
.
7
8
7
6
0
.
8
0
1
6
1
R
h
i
n
o
o
p
t
i
mi
z
e
r
(
b
e
s
t
o
b
s
e
r
v
e
d
)
0
.
8
0
8
8
0
.
7
7
7
2
0
.
8
2
5
3
0
.
8
0
0
5
2
G
W
O
0
.
8
0
8
8
0
.
7
7
7
2
0
.
8
2
5
3
0
.
8
0
0
5
2
B
a
y
e
s
i
a
n
o
p
t
i
m
i
z
a
t
i
o
n
0
.
8
0
5
0
0
.
7
9
0
3
0
.
7
9
0
3
0
.
7
9
0
3
4
DE
0
.
7
9
3
8
0
.
7
9
3
2
0
.
7
5
2
7
0
.
7
7
2
4
5
R
a
n
d
o
m s
e
a
r
c
h
0
.
7
6
1
2
0
.
7
4
9
3
0
.
7
3
1
2
0
.
7
4
0
1
6
GA
0
.
7
4
6
2
0
.
7
0
9
7
0
.
7
6
8
8
0
.
7
3
8
1
7
PSO
0
.
7
4
2
5
0
.
7
3
1
8
0
.
7
0
4
3
0
.
7
1
7
8
8
M
P
A
0
.
7
2
8
8
0
.
7
0
6
0
0
.
7
0
0
3
0
.
7
0
3
1
9
SSA
0
.
7
3
6
2
0
.
7
6
8
7
0
.
6
2
7
7
0
.
6
9
1
1
10
HHO
0
.
7
3
6
2
0
.
7
6
8
7
0
.
6
2
7
7
0
.
6
9
1
1
10
G
r
i
d
s
e
a
r
c
h
0
.
5
5
7
5
0
.
5
3
0
8
0
.
4
1
6
7
0
.
4
6
6
9
12
T
h
e
v
er
y
s
m
all
m
ar
g
i
n
b
etwe
en
R
h
in
o
o
p
tim
izer
(
0
.
8
0
0
5
)
an
d
th
e
b
est
-
p
er
f
o
r
m
i
n
g
SC
A
(
0
.
8
0
1
6
)
is
im
p
o
r
tan
t.
I
t
d
em
o
n
s
tr
ates
t
h
at
R
h
in
o
’
s
s
ea
r
ch
s
tr
ateg
y
is
co
m
p
etitiv
e
with
e
s
tab
li
s
h
ed
s
tate
-
of
-
th
e
-
ar
t
m
etah
eu
r
is
tics
d
esp
ite
b
ein
g
n
ewly
p
r
o
p
o
s
ed
.
T
h
e
s
tr
o
n
g
R
h
in
o
p
er
f
o
r
m
a
n
ce
m
ay
b
e
attr
ib
u
ted
to
its
d
es
ig
n
elem
en
ts
,
n
am
ely
elite
-
g
u
id
ed
ex
p
l
o
itatio
n
,
L
é
v
y
-
f
lig
h
t
ex
p
lo
r
atio
n
,
a
n
d
m
i
g
r
atio
n
/wallo
win
g
escap
e
m
ec
h
an
is
m
s
,
wh
ich
to
g
et
h
er
p
r
o
m
o
te
b
o
th
g
lo
b
al
s
ea
r
c
h
a
n
d
r
ef
i
n
ed
lo
ca
l
c
o
n
v
e
r
g
en
ce
wh
ile
r
ed
u
cin
g
th
e
lik
elih
o
o
d
o
f
p
r
e
m
atu
r
e
s
tag
n
atio
n
.
T
h
ese
r
esu
lts
s
h
o
u
ld
th
er
ef
o
r
e
b
e
u
n
d
er
s
to
o
d
as
co
n
f
ir
m
in
g
t
h
e
co
m
p
etitiv
e
o
p
tim
izatio
n
p
er
f
o
r
m
an
ce
r
ath
e
r
th
a
n
s
u
p
e
r
i
o
r
ity
o
v
er
cu
r
r
e
n
t
s
tate
-
of
-
th
e
-
ar
t
o
p
tim
izatio
n
tech
n
iq
u
es.
SC
A,
GW
O,
an
d
B
ay
esian
o
p
tim
izatio
n
p
er
f
o
r
m
ed
well
with
in
th
e
lim
ited
ev
alu
atio
n
b
u
d
g
et.
GA,
PS
O,
M
PA,
an
d
HHO
e
x
p
er
ien
ce
d
ea
r
lier
s
tag
n
atio
n
an
d
lo
wer
F1
-
s
co
r
es,
in
d
icati
n
g
less
o
p
tim
izatio
n
ef
f
icien
cy
u
n
d
er
t
h
e
s
am
e
co
n
s
tr
ain
ts
.
R
h
in
o
o
p
tim
izer
p
e
r
f
o
r
m
ed
s
im
ilar
ly
to
s
tu
d
ies
d
em
o
n
s
tr
atin
g
th
e
b
en
e
f
its
o
f
b
io
-
in
s
p
ir
ed
o
p
tim
izatio
n
in
d
ep
r
ess
io
n
d
etec
tio
n
an
d
r
elate
d
h
ea
lth
ca
r
e
ap
p
licatio
n
s
[
5
]
,
[
9
]
,
[
1
1
]
.
I
n
co
n
tr
ast,
m
o
s
t
p
r
ev
io
u
s
s
tu
d
ies
h
av
e
p
r
im
ar
il
y
m
ea
s
u
r
e
d
ac
c
u
r
ac
y
in
class
if
y
in
g
d
ata.
I
n
th
is
p
ap
e
r
e
v
alu
a
te
th
e
o
p
tim
izatio
n
p
er
f
o
r
m
an
ce
u
s
in
g
F1
-
s
co
r
e
a
n
d
co
n
v
er
g
e
n
ce
,
wh
ic
h
allo
ws
a
m
o
r
e
c
o
m
p
lete
ass
ess
m
en
t
o
f
th
e
p
er
f
o
r
m
an
c
e
o
f
o
p
tim
izatio
n
o
n
an
im
b
alan
ce
d
m
en
tal
-
h
ea
lth
d
ataset.
R
h
in
o
o
p
tim
izer
’
s
co
m
p
etitiv
en
es
s
ag
ain
s
t
SC
A
an
d
GW
O
s
u
g
g
ests
th
a
t
s
u
ch
b
e
h
av
io
r
-
i
n
s
p
ir
ed
s
ea
r
ch
s
tr
ateg
ies
th
at
co
m
b
in
e
ex
p
lo
r
atio
n
,
ex
p
lo
itatio
n
,
an
d
av
o
id
an
ce
o
f
s
tag
n
atio
n
ca
n
b
alan
ce
d
iv
e
r
s
ity
an
d
c
o
n
v
er
g
en
ce
.
T
h
is
is
also
th
e
ca
s
e
f
o
r
o
th
er
m
o
d
er
n
b
io
-
in
s
p
ir
ed
s
ea
r
c
h
alg
o
r
ith
m
s
[
5
]
,
[
1
0
]
,
[
2
0
]
.
A
s
im
ilar
b
eh
av
io
r
was
o
b
s
er
v
ed
wh
en
o
p
tim
izatio
n
s
wer
e
b
ein
g
ca
r
r
ied
o
u
t.
As
d
em
o
n
s
tr
ated
in
th
e
p
r
ev
io
u
s
p
ap
er
[
2
6
]
,
th
e
b
alan
ce
b
etwe
en
ex
p
lo
r
atio
n
an
d
e
x
p
lo
itatio
n
h
as
b
ee
n
f
o
u
n
d
to
b
e
k
ey
to
s
u
cc
ess
f
u
l
o
p
tim
izatio
n
s
[
2
2
]
,
[
2
4
]
.
T
h
e
b
eh
av
io
r
o
f
R
h
in
o
o
p
tim
izer
was
f
o
u
n
d
to
s
u
p
p
o
r
t
th
is
b
al
an
ce
th
r
o
u
g
h
o
u
t
th
e
s
ea
r
ch
.
Fro
m
a
p
r
ac
tical
p
er
s
p
ec
tiv
e,
R
h
in
o
o
p
tim
izer
’
s
ab
ilit
y
to
ac
h
iev
e
co
m
p
etitiv
e
F1
-
s
co
r
es
with
a
lim
ited
ev
a
lu
atio
n
b
u
d
g
et
is
o
f
p
ar
ticu
lar
r
ele
v
an
ce
in
tr
a
n
s
f
o
r
m
er
-
b
ased
d
e
p
r
ess
io
n
d
ete
ctio
n
,
wh
er
e
m
o
d
el
tr
ain
in
g
ca
n
b
e
co
s
tly
.
T
h
is
s
t
u
d
y
m
ig
h
t
b
e
ab
le
to
f
in
d
g
o
o
d
-
q
u
ality
h
y
p
er
p
a
r
am
eter
co
n
f
ig
u
r
atio
n
s
th
at
will
allo
w
u
s
to
k
ee
p
co
s
ts
lo
w
b
u
t
s
till
ac
h
iev
e
r
eli
ab
le
p
r
ed
ictio
n
s
.
Fo
r
co
m
p
ar
is
o
n
s
th
at
ar
e
n
o
n
-
s
tatis
tically
s
ig
n
if
ican
t,
it
is
d
esira
b
le
to
d
eter
m
in
e
wh
et
h
er
d
if
f
er
en
ce
s
in
th
e
p
e
r
f
o
r
m
an
ce
r
esu
lts
r
ef
lect
o
p
tim
izatio
n
r
ath
er
th
a
n
s
to
ch
asti
c
v
ar
iatio
n
s
.
T
o
ass
ess
th
e
s
ig
n
if
ican
ce
o
f
d
i
f
f
er
en
ce
s
in
th
e
r
a
n
k
in
g
o
f
o
p
tim
izer
s
am
o
n
g
co
m
p
ar
is
o
n
s
,
th
is
s
tu
d
y
u
s
ed
th
e
Frie
d
m
an
n
o
n
-
p
a
r
am
etr
i
c
test
an
d
,
wh
e
n
a
p
p
licab
le,
W
ilco
x
o
n
p
air
wis
e
co
m
p
ar
is
o
n
s
.
I
t
is
d
esira
b
le
to
p
er
f
o
r
m
th
ese
test
s
b
ec
au
s
e
th
e
d
is
tr
ib
u
tio
n
o
f
o
p
tim
izatio
n
p
er
f
o
r
m
an
ce
ca
n
n
o
t a
lway
s
b
e
ass
u
m
ed
t
o
b
e
n
o
r
m
al
an
d
h
as b
ee
n
c
o
m
m
o
n
ly
u
s
ed
i
n
co
m
p
ar
ativ
e
o
p
tim
izatio
n
s
tu
d
ies.
A
Frie
d
m
an
test
in
d
icate
d
s
tatis
t
ically
s
ig
n
if
ican
t
d
if
f
er
en
ce
s
am
o
n
g
o
p
tim
izer
s
(p
<
0
.
0
5
)
.
W
ilco
x
o
n
p
air
wis
e
test
s
s
h
o
wed
th
at
R
h
in
o
o
p
tim
izer
s
i
g
n
if
ican
tly
o
u
t
p
er
f
o
r
m
ed
g
r
i
d
s
ea
r
ch
,
GA,
SS
A,
HHO
,
an
d
MPA,
wh
ile
its
d
i
f
f
er
en
ce
f
r
o
m
SC
A
an
d
GW
O
was
n
o
t
s
tatis
tically
s
ig
n
if
ican
t.
C
o
n
v
er
g
en
ce
cu
r
v
es
s
u
g
g
ested
th
at
R
h
in
o
o
p
tim
izer
ex
h
ib
ited
a
ten
d
en
c
y
to
war
d
c
o
n
tin
u
e
d
im
p
r
o
v
e
m
en
t
d
u
r
in
g
th
e
later
s
tag
es
o
f
o
p
tim
izatio
n
,
wh
ic
h
is
co
n
s
is
ten
t
with
th
e
d
esig
n
o
b
jectiv
e
o
f
r
e
d
u
cin
g
p
r
e
m
atu
r
e
s
tag
n
atio
n
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
8
1
4
I
n
t J Ad
v
Ap
p
l Sci
,
Vo
l.
1
5
,
No
.
3
,
Sep
tem
b
er
2
0
2
6
:
1
2
3
9
-
1
2
5
1
1248
Ov
er
all,
th
e
o
b
tain
ed
r
esu
lts
co
n
f
ir
m
th
at
R
h
in
o
o
p
tim
izer
i
s
a
co
m
p
etitiv
e
o
p
tim
izatio
n
a
p
p
r
o
ac
h
ca
p
ab
le
o
f
id
en
tify
in
g
h
ig
h
-
q
u
ality
h
y
p
e
r
p
ar
am
eter
co
n
f
ig
u
r
atio
n
s
wh
il
e
m
ain
tain
in
g
s
tab
le
c
o
n
v
e
r
g
e
n
ce
b
eh
a
v
io
r
u
n
d
e
r
co
n
s
tr
ain
ed
ev
alu
atio
n
b
u
d
g
et
s
.
T
h
e
ac
h
iev
em
en
t
o
f
an
F1
-
s
co
r
e
o
f
ap
p
r
o
x
im
ately
0
.
8
0
f
u
r
th
er
h
ig
h
lig
h
ts
th
e
p
o
ten
tial
o
f
b
i
o
lo
g
ically
in
s
p
ir
ed
b
eh
av
io
r
al
s
ea
r
ch
s
tr
ateg
ies
to
p
r
o
v
id
e
an
ef
f
ec
tiv
e
b
alan
ce
b
etwe
e
n
ex
p
lo
r
atio
n
an
d
e
x
p
lo
itatio
n
in
h
ig
h
-
d
im
en
s
io
n
al
HPO
p
r
o
b
lem
s
.
4
.
1
.
Co
nv
er
g
ence
beha
v
io
r
a
nd
s
ea
rc
h dy
na
m
ics
A
s
u
m
m
ar
y
o
f
c
o
n
v
e
r
g
en
ce
p
atter
n
s
ac
r
o
s
s
th
e
v
a
r
io
u
s
o
p
t
im
izer
–
R
h
in
o
o
p
tim
izer
c
o
m
p
ar
is
o
n
s
is
p
r
esen
ted
in
Fig
u
r
e
3
,
wh
er
e
th
e
b
est
-
so
-
f
ar
test
F1
-
s
co
r
e
is
p
lo
tted
as
a
f
u
n
ctio
n
o
f
ev
alu
a
tio
n
n
u
m
b
er
u
n
d
er
a
f
ix
ed
ev
alu
atio
n
b
u
d
g
et
o
f
3
0
ev
al
u
atio
n
s
.
B
ay
esi
an
o
p
tim
izatio
n
,
DE
,
GW
O,
HHO,
a
n
d
MPA
ex
h
i
b
ited
r
ap
id
ea
r
ly
co
n
v
er
g
en
ce
,
ac
h
iev
in
g
m
o
s
t
o
f
th
eir
p
er
f
o
r
m
an
ce
g
ain
s
d
u
r
in
g
th
e
i
n
itial
p
o
r
tio
n
o
f
th
e
o
p
tim
izatio
n
p
r
o
ce
s
s
b
e
f
o
r
e
r
ea
ch
in
g
ex
ten
d
ed
p
er
f
o
r
m
a
n
ce
p
latea
u
s
.
Similar
b
eh
a
v
io
r
was
o
b
s
er
v
ed
f
o
r
s
ev
e
r
al
o
th
er
o
p
tim
izatio
n
m
et
h
o
d
s
,
in
d
icatin
g
a
ten
d
e
n
cy
to
war
d
ea
r
ly
e
x
p
lo
itatio
n
.
I
n
co
n
tr
ast,
R
h
in
o
o
p
tim
izer
d
em
o
n
s
tr
ated
a
m
o
r
e
g
r
a
d
u
al
co
n
v
e
r
g
en
ce
p
r
o
f
ile,
with
p
e
r
f
o
r
m
a
n
ce
im
p
r
o
v
em
e
n
ts
co
n
tin
u
i
n
g
th
r
o
u
g
h
o
u
t
later
s
tag
es
o
f
th
e
o
p
tim
izatio
n
p
r
o
ce
s
s
.
T
h
i
s
b
eh
av
io
r
s
u
g
g
ests
a
r
elativ
ely
b
alan
ce
d
ex
p
lo
r
ati
o
n
–
ex
p
lo
itatio
n
s
tr
ateg
y
th
at
m
ain
tain
s
s
ea
r
ch
d
iv
er
s
ity
wh
ile
p
r
eser
v
in
g
th
e
ab
ilit
y
to
r
ef
in
e
p
r
o
m
is
in
g
s
o
lu
tio
n
s
d
u
r
i
n
g
t
h
e
f
i
n
al
e
v
alu
atio
n
s
.
Alth
o
u
g
h
R
h
in
o
o
p
tim
izer
was
n
o
t
co
n
s
is
ten
tly
th
e
f
astes
t
-
c
o
n
v
e
r
g
in
g
m
eth
o
d
,
its
s
ea
r
ch
p
r
o
c
ess
p
r
o
d
u
ce
d
co
m
p
etitiv
e
f
i
n
al
F1
-
s
co
r
es
wh
ile
m
ain
tain
in
g
th
e
a
b
ilit
y
to
im
p
r
o
v
e
d
u
r
in
g
later
s
tag
es o
f
o
p
ti
m
izatio
n
.
Fig
u
r
e
3
.
C
o
n
v
er
g
e
n
ce
b
eh
a
v
i
o
r
o
f
R
h
in
o
o
p
tim
izer
an
d
co
m
p
etin
g
o
p
tim
izatio
n
alg
o
r
ith
m
s
u
n
d
er
a
f
ix
ed
ev
alu
atio
n
b
u
d
g
et
o
f
3
0
ev
alu
atio
n
s
4
.
1
.
1
.
F
ix
ed
ev
a
lua
t
i
o
n bud
g
et
o
f
3
0
ev
a
lua
t
io
ns
T
h
is
s
tu
d
y
p
u
r
s
u
ed
two
o
b
jectiv
es:
ev
alu
atin
g
th
e
ef
f
ec
tiv
e
n
ess
o
f
th
e
p
r
o
p
o
s
ed
R
h
in
o
o
p
t
im
izer
f
o
r
HPO
in
d
ep
r
ess
io
n
d
etec
tio
n
an
d
ass
ess
in
g
i
ts
o
p
tim
izatio
n
b
eh
av
i
o
r
t
h
r
o
u
g
h
co
n
v
er
g
e
n
c
e
an
aly
s
is
u
n
d
e
r
a
co
n
s
tr
ain
ed
ev
alu
atio
n
b
u
d
g
et.
T
h
e
r
esu
lts
in
d
icate
th
at
R
h
in
o
o
p
tim
izer
is
ca
p
ab
le
o
f
id
e
n
tify
in
g
h
i
g
h
-
q
u
ality
h
y
p
er
p
ar
am
eter
c
o
n
f
i
g
u
r
atio
n
s
in
th
e
p
r
esen
ce
o
f
class
im
b
alan
ce
,
ac
h
iev
in
g
a
b
est
o
b
s
er
v
ed
test
F1
-
s
co
r
e
o
f
0
.
8
0
0
5
th
at
was
n
o
t
s
tatis
tical
ly
d
if
f
er
en
t
f
r
o
m
t
h
e
lead
in
g
SC
A
an
d
GW
O
ap
p
r
o
ac
h
es.
Fu
r
th
er
m
o
r
e
,
R
h
in
o
o
p
tim
izer
d
em
o
n
s
tr
ated
s
tead
y
p
er
f
o
r
m
a
n
ce
p
r
o
g
r
ess
io
n
with
o
u
t
ex
h
ib
itin
g
th
e
p
r
o
n
o
u
n
ce
d
s
tag
n
atio
n
o
b
s
er
v
ed
in
s
ev
er
al
co
m
p
et
in
g
m
eth
o
d
s
.
T
o
g
eth
er
with
t
h
e
Frie
d
m
an
an
d
W
ilco
x
o
n
s
tatis
tical
an
aly
s
es,
th
ese
f
in
d
in
g
s
s
u
g
g
est
th
at
R
h
in
o
o
p
tim
izer
p
r
o
v
id
es
a
co
m
p
etiti
v
e
b
alan
ce
b
etwe
en
e
x
p
lo
r
atio
n
an
d
ex
p
lo
itatio
n
an
d
r
ep
r
esen
ts
a
p
r
o
m
is
in
g
o
p
t
im
izatio
n
ap
p
r
o
ac
h
f
o
r
co
m
p
u
tatio
n
ally
co
n
s
tr
ain
ed
HPO
p
r
o
b
lem
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.