I
AE
S In
t
er
na
t
io
na
l J
o
urna
l o
f
Art
if
icia
l In
t
ellig
ence
(
I
J
-
AI
)
Vo
l.
15
,
No
.
5
,
Octo
b
er
2
0
2
6
,
p
p
.
4
1
0
5
~
4
1
1
2
I
SS
N:
2
2
5
2
-
8
9
3
8
,
DOI
: 1
0
.
1
1
5
9
1
/ijai.v
15
.i
5
.
p
p
4
1
0
5
-
4
1
1
2
4105
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
a
i
.
ia
esco
r
e.
co
m
Ro
bust t
ra
ding
in
stocha
stic c
ha
o
s:
a
hybrid fuz
zy
m
ulti
-
e
x
per
t
a
ppro
a
ch
Andrey
M
u
s
a
y
ev
1
,
Dm
it
ry
G
rig
o
riev
2
1
F
a
c
u
l
t
y
o
f
A
p
p
l
i
e
d
I
n
f
o
r
ma
t
i
c
s,
I
TM
O
U
n
i
v
e
r
s
i
t
y
,
S
a
i
n
t
-
P
e
t
e
r
s
b
u
r
g
,
R
u
ssi
a
2
C
e
n
t
e
r
f
o
r
Ec
o
n
o
m
e
t
r
i
c
s
a
n
d
B
u
s
i
n
e
s
s An
a
l
y
t
i
c
s,
S
t
.
P
e
t
e
r
sb
u
r
g
S
t
a
t
e
U
n
i
v
e
r
si
t
y
,
S
a
i
n
t
-
P
e
t
e
r
s
b
u
r
g
,
R
u
ss
i
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
u
n
2
8
,
2
0
2
5
R
ev
is
ed
J
u
l 2
3
,
2
0
2
6
Acc
ep
ted
Au
g
1
2
,
2
0
2
6
Clas
sic
a
l
tren
d
-
fo
ll
o
win
g
stra
teg
i
e
s o
ften
e
x
h
ib
it
u
n
sta
b
le p
e
rf
o
rm
a
n
c
e
in
th
e
EUR/US
D
fo
re
ig
n
e
x
c
h
a
n
g
e
m
a
rk
e
t
d
u
e
to
fil
ter
lag
,
n
o
ise
-
in
d
u
c
e
d
fa
lse
sig
n
a
ls,
a
n
d
a
b
ru
p
t
re
g
ime
tr
a
n
siti
o
n
s
a
ss
o
c
iate
d
with
c
h
a
o
ti
c
p
rice
d
y
n
a
m
ics
.
T
h
is
st
u
d
y
p
r
o
p
o
se
s
a
h
y
b
rid
fu
z
z
y
m
u
lt
i
-
e
x
p
e
rt
fra
m
e
wo
rk
t
h
a
t
in
teg
ra
tes
b
i
d
irec
ti
o
n
a
l
e
x
p
o
n
e
n
ti
a
l
fil
terin
g
fo
r
lo
w
-
late
n
c
y
tren
d
e
x
trac
ti
o
n
with
a
M
a
m
d
a
n
i
-
t
y
p
e
fu
z
z
y
in
fe
re
n
c
e
sy
ste
m
th
a
t
j
o
in
t
ly
e
v
a
lu
a
tes
q
u
a
n
t
it
a
ti
v
e
tren
d
i
n
d
ica
to
rs
a
n
d
q
u
a
li
tati
v
e
in
f
o
rm
a
ti
o
n
d
e
ri
v
e
d
fro
m
eco
n
o
m
ic
n
e
ws
a
n
d
tex
t
u
a
l
f
o
re
c
a
sts.
Th
e
p
r
o
p
o
se
d
a
p
p
ro
a
c
h
is
e
v
a
lu
a
ted
o
n
m
in
u
te
-
lev
e
l
EUR/US
D
d
a
ta
o
v
e
r
1
0
-
,
2
5
-
,
a
n
d
1
0
0
-
d
a
y
tes
t
win
d
o
ws
.
Emp
iri
c
a
l
re
su
lt
s
sh
o
w
th
a
t
t
h
e
fu
z
z
y
c
o
n
se
n
su
s
m
e
c
h
a
n
ism
imp
ro
v
e
s
ro
b
u
stn
e
ss
u
n
d
e
r
n
o
n
-
sta
ti
o
n
a
ry
c
o
n
d
it
i
o
n
s.
In
a
re
p
re
se
n
tativ
e
2
5
-
d
a
y
e
v
a
lu
a
ti
o
n
p
e
ri
o
d
,
n
e
t
p
r
o
fit
i
n
c
r
e
a
se
s
fro
m
1
8
2
to
5
6
0
p
i
p
s,
t
h
e
win
ra
te
imp
ro
v
e
s
fr
o
m
5
5
%
to
6
7
%
,
a
n
d
t
h
e
m
a
x
imu
m
d
ra
wd
o
wn
d
e
c
r
e
a
se
s
fro
m
1
1
8
t
o
3
8
p
i
p
s.
Th
e
se
fin
d
i
n
g
s
su
g
g
e
st
t
h
a
t
i
n
c
o
r
p
o
ra
ti
n
g
li
n
g
u
isti
c
u
n
c
e
rtain
t
y
a
n
d
se
n
t
ime
n
t
-
li
k
e
c
o
n
tex
t
u
a
l
c
u
e
s
i
n
to
tec
h
n
ica
l
trad
i
n
g
si
g
n
a
ls
c
a
n
m
it
ig
a
te
sp
u
ri
o
u
s
trad
e
a
c
ti
v
a
ti
o
n
a
n
d
sta
b
il
ize
d
e
c
isi
o
n
-
m
a
k
in
g
i
n
sto
c
h
a
stic
a
n
d
c
h
a
o
ti
c
m
a
rk
e
t
e
n
v
iro
n
m
e
n
ts,
o
ffe
rin
g
a
p
ra
c
ti
c
a
l
ro
u
te
to
wa
rd
m
o
re
re
sili
e
n
t
a
lg
o
rit
h
m
ic
trad
in
g
sy
ste
m
s.
K
ey
w
o
r
d
s
:
C
h
ao
s
-
d
r
iv
en
f
o
r
ex
Fu
zz
y
lo
g
ic
Mu
lti
-
ex
p
er
t sy
s
tem
Sto
ch
asti
c
ch
ao
s
T
r
en
d
f
ilter
in
g
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
:
Dm
itry
Gr
ig
o
r
iev
C
en
ter
f
o
r
E
co
n
o
m
etr
ics an
d
B
u
s
in
ess
An
aly
tic
s
,
St.
Peter
s
b
u
r
g
State
Un
iv
e
r
s
ity
Sain
t
-
Peter
s
b
u
r
g
,
R
u
s
s
ia
E
m
ail:
d
.
a.
g
r
ig
o
r
iev
@
s
p
b
u
.
r
u
1.
I
NT
RO
D
UCT
I
O
N
T
h
e
Fo
r
e
x
m
a
r
k
et
is
k
n
o
wn
f
o
r
its
co
m
p
lex
ity
a
n
d
u
n
p
r
ed
ictab
ilit
y
,
o
f
ten
d
is
p
lay
in
g
ch
a
o
tic
b
eh
av
io
r
an
d
ex
tr
em
e
v
o
latilit
y
[
1
]
–
[
4
]
.
Un
d
e
r
s
to
ch
asti
c
ch
ao
s
,
i.e
.
,
d
y
n
am
ically
u
n
s
tab
l
e
p
r
ice
tr
ajec
to
r
ies
h
ig
h
ly
s
en
s
itiv
e
to
in
itial
co
n
d
itio
n
s
,
tr
ad
itio
n
al
tr
en
d
-
f
o
llo
win
g
s
tr
ateg
ies
s
tr
u
g
g
le
to
r
em
ain
ef
f
ec
tiv
e
[
5
]
.
Su
d
d
en
m
a
r
k
et
r
e
v
er
s
als an
d
i
r
r
eg
u
lar
f
lu
ctu
atio
n
s
m
a
k
e
s
tan
d
ar
d
lin
ea
r
m
o
d
els u
n
r
eliab
le
o
v
er
tim
e
[
6
]
.
T
o
o
v
er
c
o
m
e
th
ese
is
s
u
es,
r
esear
ch
er
s
h
av
e
in
v
esti
g
ated
ch
ao
s
th
eo
r
y
,
a
d
ap
tiv
e
co
n
tr
o
l,
an
d
r
o
b
u
s
t
s
tatis
t
ics
[
7
]
,
[
8
]
.
H
o
wev
er
,
ev
en
ad
v
an
ce
d
n
u
m
er
ical
m
o
d
el
s
p
r
o
v
e
in
ad
e
q
u
ate
w
h
en
a
b
r
u
p
t
s
en
tim
en
t
-
d
r
i
v
en
m
o
v
em
en
ts
d
o
m
in
ate
th
e
Fo
r
e
x
m
ar
k
et
[
9
]
–
[
1
1
]
.
T
h
is
h
i
g
h
li
g
h
ts
th
e
n
ee
d
f
o
r
m
o
r
e
f
lex
ib
l
e
d
ec
is
io
n
-
m
ak
in
g
to
o
ls
ca
p
ab
le
o
f
h
an
d
lin
g
u
n
ce
r
tain
ty
in
r
ea
l
-
tim
e
tr
ad
i
n
g
.
Fu
zz
y
l
o
g
ic
o
f
f
er
s
a
s
tr
o
n
g
f
r
am
ewo
r
k
f
o
r
m
o
d
elin
g
d
ec
is
io
n
s
u
n
d
er
u
n
c
er
tain
ty
[
1
2
]
.
Usi
n
g
lin
g
u
is
tic
v
ar
iab
les
an
d
f
u
zz
y
r
u
les,
it
en
ab
les
q
u
alitativ
e
r
ea
s
o
n
in
g
in
au
to
m
ated
s
y
s
tem
s
[
1
3
]
.
T
h
ese
m
eth
o
d
s
h
a
v
e
p
r
o
v
en
u
s
ef
u
l
in
f
in
an
cial
f
o
r
ec
asti
n
g
,
im
p
r
o
v
in
g
d
ec
is
io
n
r
eliab
ilit
y
u
n
d
er
u
n
ce
r
tain
co
n
d
itio
n
s
[
1
4
]
,
[
1
5
]
.
I
n
t
h
i
s
p
a
p
e
r
,
a
h
y
b
r
i
d
f
u
z
z
y
m
u
l
t
i
-
e
x
p
e
r
t
s
y
s
t
e
m
d
e
s
i
g
n
e
d
w
a
s
i
n
t
r
o
d
u
c
e
d
t
o
e
n
h
a
n
c
e
t
r
en
d
-
f
o
l
l
o
w
i
n
g
p
er
f
o
r
m
an
ce
i
n
ch
a
o
tic
Fo
r
ex
m
ar
k
ets.
T
h
e
s
y
s
tem
co
m
b
in
e
s
tech
n
ical
an
aly
s
is
with
f
u
zz
y
s
en
tim
en
t
an
aly
s
is
b
ased
o
n
ec
o
n
o
m
ic
n
ews
an
d
tex
tu
al
f
o
r
ec
asts
[
5
]
,
[
1
6
]
.
I
t
u
s
es
b
id
ir
ec
tio
n
al
e
x
p
o
n
en
ti
al
f
ilter
s
f
o
r
tr
e
n
d
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
15
,
No
.
5
,
Octo
b
er
2
0
2
6
:
4
1
0
5
-
4
1
1
2
4106
d
etec
tio
n
an
d
tr
an
s
lates
th
em
in
to
f
u
zz
y
lin
g
u
is
tic
ter
m
s
,
wh
ile
a
f
u
zz
y
s
en
tim
en
t
m
o
d
u
l
e
ev
alu
ates
m
ar
k
e
t
m
o
o
d
v
ia
tex
t m
in
in
g
.
T
h
ese
two
m
o
d
u
les ar
e
in
teg
r
a
ted
v
ia
a
Ma
m
d
an
i
-
ty
p
e
f
u
zz
y
in
f
er
en
ce
s
y
s
tem
,
ex
ec
u
tin
g
tr
ad
es o
n
ly
wh
en
b
o
th
tech
n
ical
an
d
f
u
n
d
am
en
tal
s
ig
n
als
alig
n
[
5
]
.
T
h
is
h
y
b
r
id
a
p
p
r
o
ac
h
r
e
d
u
ce
s
f
alse
s
ig
n
als
an
d
im
p
r
o
v
es
s
tab
ilit
y
,
esp
ec
ially
d
u
r
in
g
ch
a
o
tic
p
er
io
d
s
.
T
ests
o
n
E
UR
/USD
d
ata
d
em
o
n
s
tr
ate
th
at
th
e
s
y
s
tem
o
u
tp
er
f
o
r
m
s
s
in
g
le
-
e
x
p
er
t
ap
p
r
o
ac
h
es
in
r
is
k
-
ad
ju
s
ted
r
etu
r
n
s
an
d
r
esil
ien
ce
to
s
u
d
d
en
m
ar
k
et
ch
a
n
g
es
[
1
7
]
.
T
h
is
wo
r
k
co
n
tr
ib
u
tes
to
b
o
th
f
u
zz
y
s
y
s
tem
s
r
esear
ch
an
d
f
in
an
cial
en
g
in
ee
r
in
g
b
y
d
e
m
o
n
s
tr
atin
g
h
o
w
f
u
zz
y
lo
g
ic
ca
n
m
a
n
ag
e
u
n
ce
r
tain
ty
i
n
co
m
p
le
x
m
ar
k
ets.
T
h
e
r
em
ain
d
er
o
f
th
is
p
ap
er
is
a
s
f
o
llo
ws.
Sect
io
n
2
p
r
esen
ts
th
e
th
eo
r
etica
l
b
ac
k
g
r
o
u
n
d
an
d
ex
p
er
im
en
tal
s
etu
p
.
Sectio
n
3
p
r
esen
ts
th
e
e
x
p
er
im
e
n
tal
r
esu
lts
an
d
d
is
cu
s
s
io
n
.
Fin
ally
,
s
ec
tio
n
4
p
r
esen
ts
th
e
co
n
clu
s
io
n
an
d
f
u
tu
r
e
r
esear
c
h
d
ir
ec
tio
n
s
.
2.
M
E
T
H
O
D
2
.
1
.
T
heo
re
t
ica
l
b
a
c
k
g
ro
un
d
I
n
s
tan
d
ar
d
s
to
c
h
asti
c
f
r
am
ewo
r
k
s
,
o
b
s
er
v
ed
m
ar
k
et
d
ata
is
ty
p
ically
m
o
d
eled
as
a
co
m
b
in
atio
n
o
f
an
u
n
d
er
l
y
in
g
tr
e
n
d
an
d
n
o
is
e
.
=
+
,
=
1
,
.
.
.
,
Ho
wev
er
,
r
ea
l
-
wo
r
ld
f
in
an
cial
m
ar
k
ets
ex
h
ib
it
ch
ao
tic
b
eh
a
v
io
r
,
wh
er
e
b
o
t
h
co
m
p
o
n
en
ts
ar
e
n
o
n
-
s
tatio
n
ar
y
an
d
h
ig
h
l
y
v
o
latile
[
1
8
]
,
[
1
9
]
.
C
o
n
tr
o
l
-
of
-
c
h
ao
s
r
esear
ch
s
h
o
ws
th
at
u
n
s
tab
le
d
y
n
am
ics
ca
n
b
e
s
tab
ilized
v
ia
co
n
tr
o
l
a
n
d
f
ee
d
b
ac
k
m
ec
h
a
n
is
m
s
[
2
0
]
,
m
a
k
in
g
tr
ad
itio
n
al
m
o
d
elin
g
ass
u
m
p
tio
n
s
in
ad
e
q
u
ate
an
d
ca
llin
g
f
o
r
to
o
ls
th
at
ex
p
licitly
m
an
a
g
e
u
n
ce
r
tain
ty
.
T
o
ex
tr
ac
t m
ea
n
in
g
f
u
l tr
en
d
s
f
r
o
m
n
o
is
y
d
ata,
we
em
p
lo
y
e
x
p
o
n
en
tial f
ilter
in
g
.
=
−
1
+
(
−
−
1
)
,
=
2
,
.
.
.
,
W
h
er
e
∈
(
0
.
001
,
0
.
1
)
b
alan
ce
s
r
esp
o
n
s
iv
en
e
s
s
an
d
s
m
o
o
th
n
ess
[
2
1
]
.
T
o
r
e
d
u
ce
p
h
ase
lag
wh
ile
p
r
eser
v
i
n
g
n
o
is
e
r
ed
u
ctio
n
,
we
ap
p
l
y
b
i
d
ir
ec
tio
n
al
ex
p
o
n
en
tial
s
m
o
o
th
in
g
:
f
ir
s
t
in
th
e
r
ev
er
s
e
c
h
r
o
n
o
lo
g
ical
d
ir
ec
tio
n
,
th
en
f
o
r
war
d
to
r
esto
r
e
ca
u
s
ality
:
̃
(
)
=
+
(
1
−
)
̃
+
1
(
)
̃
(
)
=
̃
(
)
+
(
1
−
)
̃
−
1
(
)
with
,
∈
[
0
.
005
,
0
.
03
]
an
d
win
d
o
w
len
g
th
=
20
[
5
]
.
Fil
ter
ed
tr
en
d
s
ar
e
th
en
in
ter
p
r
eted
u
s
in
g
f
u
zz
y
lo
g
ic,
tr
an
s
latin
g
n
u
m
er
ical
s
ig
n
als
in
to
lin
g
u
is
tic
v
ar
iab
les s
u
ch
as "stro
n
g
u
p
tr
e
n
d
"
o
r
"n
o
tr
e
n
d
"
[
5
]
.
Per
f
o
r
m
an
ce
is
m
ea
s
u
r
ed
u
s
in
g
cu
m
u
l
ativ
e
n
et
p
r
o
f
it
.
Eff
(
)
=
∑
(
=
1
−
1
)
(
(
)
−
(
)
)
W
h
er
e
an
d
ar
e
en
tr
y
an
d
ex
it
tim
es,
is
th
e
p
r
ice,
is
p
o
s
itio
n
s
ize,
an
d
in
d
icate
s
tr
ad
e
d
i
r
ec
tio
n
(
0
f
o
r
b
u
y
,
1
f
o
r
s
ell)
.
A
s
tr
at
eg
y
is
co
n
s
id
er
ed
s
u
cc
ess
f
u
l
if
it
y
ield
s
p
o
s
itiv
e
r
etu
r
n
s
af
ter
tr
an
s
ac
tio
n
co
s
ts
a
n
d
s
l
i
p
p
a
g
e
.
I
n
c
h
a
o
t
i
c
e
n
v
i
r
o
n
m
e
n
t
s
,
r
e
s
u
l
t
s
d
e
p
e
n
d
h
e
a
v
il
y
o
n
p
a
r
a
m
e
t
e
r
s
t
a
b
i
l
i
t
y
a
n
d
f
r
e
q
u
e
n
t
r
e
c
a
l
i
b
r
a
ti
o
n
.
T
r
ad
itio
n
al
tr
e
n
d
-
f
o
llo
win
g
s
tr
ateg
ies
o
f
ten
g
en
er
ate
m
is
lead
in
g
s
ig
n
als
d
u
e
t
o
f
ilter
lag
an
d
f
ail
to
ad
ap
t
to
r
eg
im
e
s
h
if
ts
[
1
]
,
[
5
]
.
E
v
en
o
p
tim
ized
p
ar
a
m
eter
s
p
er
f
o
r
m
p
o
o
r
ly
o
u
t
-
of
-
s
am
p
le
,
d
em
o
n
s
tr
atin
g
th
e
lim
itatio
n
s
o
f
p
u
r
ely
n
u
m
er
i
c
ap
p
r
o
ac
h
es.
Ho
wev
er
,
wh
e
n
q
u
an
titativ
e
tr
e
n
d
s
ig
n
als
ar
e
co
m
b
in
ed
with
q
u
a
l
i
t
a
ti
v
e
s
e
n
ti
m
e
n
t
a
n
al
y
s
is
,
d
e
c
i
s
i
o
n
-
m
a
k
i
n
g
s
t
a
b
il
i
t
y
i
m
p
r
o
v
e
s
,
e
s
p
e
ci
a
l
l
y
d
u
r
i
n
g
c
h
a
o
t
ic
m
a
r
k
e
t
p
h
a
s
es
[
5
]
.
2
.
2
.
E
x
perim
ent
a
l
s
et
up
Desp
ite
im
p
r
o
v
em
en
ts
,
r
esid
u
al
in
s
tab
ilit
y
r
em
ain
s
in
f
u
lly
ch
ao
tic
r
eg
im
es.
Fu
tu
r
e
wo
r
k
will
f
o
cu
s
o
n
r
e
f
in
in
g
f
u
zz
y
m
em
b
er
s
h
i
p
f
u
n
ctio
n
s
,
e
x
p
lo
r
in
g
n
o
n
lin
ea
r
f
u
zz
y
d
etec
to
r
s
,
a
n
d
e
x
te
n
d
in
g
th
e
m
o
d
el
to
m
u
l
t
i
-
a
s
s
et
c
o
r
r
e
l
a
t
i
o
n
s
[
1
7
]
.
T
h
e
s
y
s
t
e
m
w
as
e
v
a
l
u
a
t
e
d
o
n
1
-
m
i
n
u
t
e
E
U
R
/
U
SD
o
p
e
n
,
h
i
g
h
,
l
o
w
,
c
l
o
s
e
(
O
HL
C
)
q
u
o
tes
f
r
o
m
Fin
am
at
h
ttp
s
://www.
f
in
am
.
r
u
/
v
ia
t
r
ad
e
ap
p
l
icatio
n
p
r
o
g
r
am
m
i
n
g
in
te
r
f
ac
e
(
T
r
ad
e
API
)
,
o
v
er
th
r
ee
n
o
n
-
o
v
er
lap
p
in
g
test
wi
n
d
o
ws:
1
0
,
2
5
,
an
d
1
0
0
tr
ad
i
n
g
d
a
y
s
.
All
d
ata
wer
e
p
r
e
p
r
o
ce
s
s
ed
u
s
in
g
lin
ea
r
in
ter
p
o
latio
n
f
o
r
m
is
s
in
g
v
a
lu
es
an
d
a
r
o
llin
g
m
ed
ia
n
f
ilter
(
win
d
o
w
=
5
)
f
o
r
o
u
t
lier
r
em
o
v
al.
T
h
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
R
o
b
u
s
t tra
d
in
g
i
n
s
to
ch
a
s
tic
c
h
a
o
s
:
a
h
yb
r
id
fu
z
z
y
mu
lti
-
ex
p
ert a
p
p
r
o
a
ch
(
A
n
d
r
ey
Mu
s
a
ye
v
)
4107
b
id
ir
ec
tio
n
al
f
ilter
u
s
ed
=
=
0
.
015
with
lo
o
k
b
ac
k
=
20
,
en
tr
y
th
r
esh
o
ld
∗
=
5
p
ip
s
,
an
d
e
x
it
th
r
esh
o
ld
⁰
=
−
5
p
ip
s
.
T
h
e
s
en
tim
en
t
m
o
d
u
le
em
p
l
o
y
ed
a
p
r
e
-
d
ef
in
ed
f
i
n
an
cial
lex
ico
n
w
ith
th
r
ee
lin
g
u
is
tic
ter
m
s
(
b
u
llis
h
/n
eu
tr
al/b
ea
r
is
h
)
.
T
h
e
Ma
m
d
a
n
i
in
f
er
en
ce
en
g
in
e
a
p
p
lied
n
in
e
r
u
les
with
m
in
/m
ax
o
p
er
ato
r
s
an
d
ce
n
ter
-
of
-
g
r
av
it
y
d
e
f
u
zz
if
icatio
n
.
Per
f
o
r
m
an
c
e
was
m
ea
s
u
r
ed
b
y
n
et
p
r
o
f
it
(
p
ip
s
)
,
win
r
ate
(
%),
an
d
m
ax
im
u
m
d
r
awd
o
wn
(
p
ip
s
)
.
T
h
e
co
m
p
lete
ex
p
er
im
e
n
tal
p
r
o
to
c
o
l c
o
m
p
r
is
es th
e
f
o
llo
w
in
g
s
tep
s
:
i)
Data
p
r
ep
ar
atio
n
:
l
o
a
d
1
-
m
i
n
u
te
E
UR
/USD
OHL
C
d
ata
f
o
r
th
e
s
p
ec
if
ied
test
win
d
o
w.
Ap
p
ly
lin
ea
r
in
ter
p
o
latio
n
to
f
ill
m
is
s
in
g
tick
s
.
R
em
o
v
e
o
u
tlier
s
ex
ce
ed
in
g
±
3
s
tan
d
ar
d
d
e
v
iatio
n
s
f
r
o
m
th
e
r
o
llin
g
m
ed
ian
(
win
d
o
w
=
5
)
.
ii)
T
e
c
h
n
i
c
a
l
s
i
g
n
a
l
g
e
n
e
r
a
t
i
o
n
:
a
p
p
l
y
b
i
d
i
r
e
c
t
i
o
n
a
l
e
x
p
o
n
e
n
t
i
a
l
f
i
l
t
e
r
i
n
g
(
s
e
ct
i
o
n
2
.
1
)
w
i
t
h
α
_
b
=
α
_
f
=
0
.
0
1
5
an
d
L
=
2
0
.
C
o
m
p
u
te
th
e
s
m
o
o
th
ed
p
r
ice
c
h
an
g
e
d
y
o
v
e
r
win
d
o
w
L
.
Fu
zz
if
y
d
y
in
t
o
li
n
g
u
is
tic
ter
m
s
:
"
u
p
tr
en
d
"
if
d
y
>
d
y
, "
d
o
w
n
tr
e
n
d
"
if d
y
<
-
dy
,
an
d
"
n
o
tr
en
d
"
o
th
er
wis
e.
iii)
Sen
tim
en
t
s
ig
n
al
g
en
e
r
atio
n
:
f
o
r
ea
ch
tr
a
d
in
g
p
er
io
d
,
c
o
llect
av
ailab
le
ec
o
n
o
m
ic
n
e
ws
an
d
tex
tu
al
f
o
r
ec
asts
.
Ap
p
ly
le
x
ico
n
-
b
ased
class
if
icatio
n
to
co
m
p
u
t
e
s
en
tim
en
t
s
co
r
es.
Fu
zz
if
y
s
co
r
es
in
to
:
"
b
u
llis
h
,
"
"n
eu
tr
al,
"
o
r
"b
ea
r
is
h
"
b
ased
o
n
p
r
e
d
ef
in
ed
m
em
b
er
s
h
ip
th
r
esh
o
ld
s
.
iv
)
Fu
zz
y
in
f
er
e
n
ce
an
d
tr
ad
e
ex
ec
u
tio
n
:
f
ee
d
b
o
th
f
u
zz
if
ied
s
ig
n
als
in
to
th
e
Ma
m
d
a
n
i
in
f
e
r
en
ce
en
g
i
n
e
.
Ap
p
ly
th
e
9
-
r
u
le
b
ase.
Gen
er
ate
tr
ad
in
g
ac
tio
n
:
"
buy
"
o
n
l
y
if
(
tr
en
d
is
"u
p
tr
en
d
"
AND
s
en
tim
en
t
is
"
b
u
llis
h
")
;
"
s
ell
"
o
n
ly
if
(
tr
en
d
is
"d
o
wn
tr
en
d
"
AND
s
en
tim
en
t
is
"b
ea
r
is
h
")
;
"
h
o
ld
"
o
th
er
wis
e.
E
x
ec
u
te
tr
ad
es a
t th
e
n
ex
t a
v
ailab
le
1
-
m
in
u
te
o
p
e
n
p
r
ice.
v)
Po
s
itio
n
m
an
ag
em
en
t
an
d
ex
it
.
Ho
ld
p
o
s
itio
n
s
u
n
til
eith
er
:
i
)
th
e
f
ilter
ed
tr
en
d
cr
o
s
s
es
th
e
ex
it
th
r
esh
o
l
d
d
y
⁰
(
d
y
n
a
m
i
c
e
x
i
t
)
,
o
r
ii
)
a
s
t
o
p
-
l
o
s
s
is
t
r
i
g
g
e
r
e
d
(
s
e
t
a
t
7
5
p
ip
s
f
o
r
b
a
s
e
l
i
n
es
;
d
y
n
a
m
i
c
f
o
r
T
S
-
2
a
n
d
T
S
-
3
)
.
R
ec
o
r
d
en
tr
y
a
n
d
e
x
it p
r
ices,
t
im
estam
p
s
,
an
d
tr
ad
e
d
ir
ec
tio
n
.
v
i)
Per
f
o
r
m
an
ce
ev
alu
atio
n
.
Af
te
r
co
m
p
letin
g
all
tr
ad
es
with
in
th
e
test
win
d
o
w,
c
o
m
p
u
te
n
e
t
p
r
o
f
it
u
s
in
g
(
)
=
(
−
1
)
(
(
)
−
(
)
)
,
win
r
ate
(
/
×
100
)
,
an
d
m
ax
im
u
m
d
r
awd
o
wn
(
lar
g
est
p
ea
k
-
to
-
tr
o
u
g
h
d
ec
lin
e)
.
Ag
g
r
eg
ate
r
esu
lts
ac
r
o
s
s
th
e
th
r
ee
test
win
d
o
ws
(
1
0
,
2
5
,
a
n
d
1
0
0
d
ay
s
)
.
All
p
ar
am
eter
v
alu
es,
p
r
e
p
r
o
ce
s
s
in
g
ch
o
ices,
an
d
ev
alu
ati
o
n
cr
iter
ia
wer
e
f
ix
e
d
ac
r
o
s
s
all
ex
p
er
im
en
ts
to
en
s
u
r
e
co
n
s
is
ten
t c
o
m
p
a
r
is
o
n
s
.
So
u
r
ce
co
d
e
is
av
ailab
le
f
r
o
m
co
r
r
esp
o
n
d
in
g
au
th
o
r
u
p
o
n
r
ea
s
o
n
ab
le
r
eq
u
est.
3.
RE
SU
L
T
S
AND
D
I
SCU
SS
I
O
N
3
.
1
.
B
a
s
eline
t
re
nd
-
f
o
llo
wing
s
t
ra
t
eg
y
T
h
e
b
aselin
e
s
tr
ateg
y
em
p
lo
y
s
a
s
in
g
le
ex
p
o
n
en
tial
f
ilter
with
f
ix
ed
th
r
esh
o
ld
s
an
d
ex
i
ts
.
T
ab
le
1
s
u
m
m
ar
izes
its
im
p
lem
en
tatio
n
.
As
s
h
o
wn
in
Fig
u
r
e
1
,
t
h
e
b
aselin
e
tr
en
d
-
f
o
llo
win
g
s
tr
at
eg
y
with
f
ix
e
d
s
to
p
s
(TS
-
1)
g
en
er
ates
f
r
eq
u
e
n
t
wh
ip
s
aws
in
r
an
g
e
-
b
o
u
n
d
m
ar
k
e
ts
[
1
]
,
[
5
]
,
co
n
f
ir
m
in
g
th
at
f
i
x
ed
-
th
r
esh
o
ld
r
u
les
f
ail
in
n
o
n
-
tr
en
d
i
n
g
r
e
g
im
es.
T
ab
le
1
.
B
aselin
e
tr
en
d
-
f
o
llo
w
in
g
s
tr
ateg
y
with
f
ix
e
d
s
to
p
s
C
o
m
p
o
n
e
n
t
I
mp
l
e
me
n
t
a
t
i
o
n
P
e
r
f
o
r
ma
n
c
e
O
p
t
i
mi
z
a
t
i
o
n
K
e
y
l
i
m
i
t
a
t
i
o
n
s
D
a
t
a
p
r
e
p
a
r
a
t
i
o
n
‒
1
-
mi
n
EU
R
/
U
S
D
q
u
o
t
e
s
‒
1
0
/
2
5
/
1
0
0
-
d
a
y
w
i
n
d
o
w
s
‒
P
r
e
p
r
o
c
e
ss
i
n
g
:
i
n
t
e
r
p
o
l
a
t
i
o
n
,
r
o
l
l
i
n
g
-
m
e
d
i
a
n
o
u
t
l
i
e
r
r
e
mo
v
a
l
,
t
i
m
e
a
l
i
g
n
me
n
t
N
/
A
N
/
A
H
i
g
h
-
f
r
e
q
u
e
n
c
y
n
o
i
s
e
a
f
f
e
c
t
s
si
g
n
a
l
q
u
a
l
i
t
y
En
t
r
y
c
o
n
d
i
t
i
o
n
|
−
,
|
>
∗
‒
is
sm
o
o
t
h
e
d
p
r
i
c
e
c
h
a
n
g
e
‒
is
l
o
o
k
b
a
c
k
w
i
n
d
o
w
‒
∗
is
t
h
r
e
sh
o
l
d
I
n
i
t
i
a
l
(
1
0
-
d
a
y
)
:
‒
+
1
3
3
p
i
p
s
‒
5
7
%
w
i
n
r
a
t
e
O
p
t
i
mi
z
e
d
p
a
r
a
ms
:
P
∗
=
(
0
.
0
1
,
7
0
,
1
0
)
→
+
5
3
2
p
i
p
s,
6
9
%
w
i
n
r
a
t
e
(
i
n
-
s
a
mp
l
e
)
La
g
s
i
n
s
i
d
e
w
a
y
s
mark
e
t
s;
l
a
t
e
e
n
t
r
i
e
s
n
e
a
r
r
e
v
e
r
sa
l
s
Ex
i
t
r
u
l
e
s
F
i
x
e
d
t
a
k
e
-
p
r
o
f
i
t
/
s
t
o
p
-
l
o
ss
=
7
5
p
i
p
s
Ef
f
e
c
t
i
v
e
i
n
st
r
o
n
g
t
r
e
n
d
s
N
o
n
e
(
st
a
t
i
c
e
x
i
t
s)
F
a
i
l
s
i
n
v
o
l
a
t
i
l
e
/
c
h
o
p
p
y
c
o
n
d
i
t
i
o
n
s
smo
o
t
h
i
n
g
Ex
p
o
n
e
n
t
i
a
l
f
i
l
t
e
r
(
α
=
0
.
0
1
)
R
e
d
u
c
e
s
n
o
i
s
e
b
u
t
a
d
d
s
l
a
g
Te
st
e
d
α
∈
[
0
.
0
0
5
–
0
.
0
3
]
Tr
a
d
e
-
o
f
f
b
e
t
w
e
e
n
r
e
sp
o
n
s
i
v
e
n
e
ss
a
n
d
f
a
l
se
-
si
g
n
a
l
r
a
t
e
O
p
t
i
mi
z
a
t
i
o
n
G
r
i
d
s
e
a
r
c
h
:
‒
α
(
s
mo
o
t
h
i
n
g
)
∈
[
0
.
0
0
5
–
0
.
0
3
]
‒
L
(
w
i
n
d
o
w
)
∈
[
1
0
–
1
0
0
]
‒
dy
∗
(
e
n
t
r
y
t
h
r
e
sh
o
l
d
)
∈
[
1
0
–
5
0
]
B
e
st
i
n
-
sam
p
l
e
:
+
5
3
2
p
i
p
s
O
u
t
-
of
-
s
a
mp
l
e
d
r
o
p
O
v
e
r
f
i
t
t
i
n
g
r
i
s
k
;
p
o
o
r
g
e
n
e
r
a
l
i
z
a
t
i
o
n
K
e
y
i
n
s
i
g
h
t
F
i
x
e
d
-
t
h
r
e
s
h
o
l
d
s
t
r
a
t
e
g
i
e
s
f
a
i
l
i
n
n
o
n
-
t
r
e
n
d
i
n
g
r
e
g
i
mes
W
o
r
k
s
o
n
l
y
i
n
st
r
o
n
g
t
r
e
n
d
s
A
d
a
p
t
i
v
e
me
t
h
o
d
s
n
e
e
d
e
d
H
y
b
r
i
d
f
u
z
z
y
l
o
g
i
c
p
r
o
p
o
se
d
f
o
r
r
o
b
u
s
t
n
e
ss
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
15
,
No
.
5
,
Octo
b
er
2
0
2
6
:
4
1
0
5
-
4
1
1
2
4108
Fig
u
r
e
1
.
E
f
f
ec
tiv
en
ess
o
f
th
e
T
S
-
1
tr
en
d
s
tr
ateg
y
im
p
lem
en
t
atio
n
3
.
2
.
Ada
ptiv
e
t
re
nd
s
t
ra
t
eg
y
wit
h f
lex
ibl
e
ex
it
T
r
ad
itio
n
al
tr
e
n
d
-
f
o
llo
win
g
s
tr
ateg
ies
o
f
ten
u
s
e
f
i
x
ed
tak
e
-
p
r
o
f
it
(
T
P)
an
d
s
to
p
-
lo
s
s
(
SL)
lev
els,
wh
ich
p
er
f
o
r
m
p
o
o
r
ly
in
c
h
ao
tic
m
ar
k
ets.
T
o
im
p
r
o
v
e
a
d
ap
t
ab
ilit
y
,
T
S
-
2
was
d
ev
elo
p
e
d
,
a
s
tr
ateg
y
th
at
u
s
es
d
y
n
am
ic
ex
it
r
u
les
b
ased
o
n
r
ea
l
-
tim
e
tr
en
d
s
tr
en
g
th
(
T
a
b
le
2
)
.
Af
ter
o
p
tim
izatio
n
,
T
S
-
2
i
m
p
r
o
v
e
d
n
et
p
r
o
f
it
f
r
o
m
1
8
2
to
2
6
1
p
ip
s
an
d
win
r
ate
f
r
o
m
5
5
% to
5
6
% (
Fig
u
r
e
2
)
,
b
u
t r
em
ain
e
d
v
u
ln
er
ab
le
t
o
s
u
d
d
en
r
ev
er
s
als.
T
ab
le
2
.
Ad
a
p
tiv
e
tr
en
d
s
tr
ateg
y
with
d
y
n
am
ic
ex
its
C
o
m
p
o
n
e
n
t
F
i
x
e
d
-
t
h
r
e
s
h
o
l
d
b
a
s
e
l
i
n
e
TS
-
2
i
n
n
o
v
a
t
i
o
n
O
p
t
i
mi
z
a
t
i
o
n
r
e
su
l
t
s
K
e
y
l
i
m
i
t
a
t
i
o
n
s
Ex
i
t
m
e
c
h
a
n
i
s
m
F
i
x
e
d
TP/SL
l
e
v
e
l
s
D
y
n
a
mi
c
e
x
i
t
s
b
a
s
e
d
o
n
r
e
a
l
-
t
i
m
e
t
r
e
n
d
s
t
r
e
n
g
t
h
R
e
d
u
c
e
d
l
o
ss
e
s
b
y
3
8
%
v
s fi
x
e
d
e
x
i
t
s
i
n
b
a
c
k
t
e
st
s
S
t
i
l
l
r
e
a
c
t
i
v
e
t
o
su
d
d
e
n
r
e
v
e
r
sal
s
C
o
r
e
p
a
r
a
m
e
t
e
r
s
S
i
n
g
l
e
t
h
r
e
s
h
o
l
d
sy
st
e
ms
4
a
d
a
p
t
i
v
e
p
a
r
a
me
t
e
r
s
:
‒
α
(
smo
o
t
h
i
n
g
)
‒
L
(
w
i
n
d
o
w
)
‒
d
y
*
(
e
n
t
r
y
)
‒
d
y
⁰
(
e
x
i
t
)
O
p
t
i
mal
se
t
f
o
u
n
d
:
P*
=
(
0
.
0
1
5
,
2
0
,
5
,
-
5)
H
i
g
h
p
a
r
a
m
e
t
e
r
sen
s
i
t
i
v
i
t
y
I
n
i
t
i
a
l
p
e
r
f
o
r
m
a
n
c
e
N
/
A
‒
2
5
p
i
p
s
n
e
t
l
o
ss
‒
4
3
%
w
i
n
r
a
t
e
(
d
e
f
a
u
l
t
p
a
r
a
ms)
H
i
g
h
l
i
g
h
t
e
d
sm
o
o
t
h
i
n
g
-
i
n
d
u
c
e
d
d
e
l
a
y
s
W
e
a
k
i
n
r
a
n
g
e
-
b
o
u
n
d
mar
k
e
t
s
O
p
t
i
mi
z
a
t
i
o
n
M
a
n
u
a
l
t
u
n
i
n
g
Tw
o
-
p
h
a
s
e
g
r
i
d
s
e
a
r
c
h
:
‒
B
r
o
a
d
:
α
∈
[
0
.
0
0
5
:
0
.
0
3
]
,
L
∈
[
1
0
:
1
0
0
]
‒
R
e
f
i
n
e
d
:
d
y
*
∈
[
5
:
2
5
]
B
e
st
r
e
s
u
l
t
s:
‒
2
6
1
p
i
p
s
p
r
o
f
i
t
‒
5
6
%
w
i
n
r
a
t
e
(
2
0
m
i
n
w
i
n
d
o
w
,
-
5
e
x
i
t
t
h
r
e
sh
o
l
d
)
C
o
m
p
u
t
a
t
i
o
n
a
l
l
y
i
n
t
e
n
si
v
e
M
a
r
k
e
t
a
d
a
p
t
a
t
i
o
n
W
o
r
k
s
o
n
l
y
i
n
st
r
o
n
g
t
r
e
n
d
s
B
e
t
t
e
r
h
a
n
d
l
e
s we
a
k
e
n
i
n
g
t
r
e
n
d
s
v
i
a
d
y
⁰
P
r
o
f
i
t
a
b
i
l
i
t
y
i
mp
r
o
v
e
d
b
y
1
0
.
4
×
a
f
t
e
r
o
p
t
i
mi
z
a
t
i
o
n
F
a
i
l
s
i
n
e
x
t
r
e
me
v
o
l
a
t
i
l
i
t
y
F
o
r
w
a
r
d
p
a
t
h
N
/
A
N
/
A
P
r
o
p
o
se
d
s
o
l
u
t
i
o
n
:
H
y
b
r
i
d
f
u
z
z
y
l
o
g
i
c
+
se
n
t
i
me
n
t
a
n
a
l
y
si
s
R
e
q
u
i
r
e
s
n
o
n
-
n
u
m
e
r
i
c
i
n
p
u
t
s
3
.
3
.
Dua
l
-
wind
o
w
t
re
nd
s
t
ra
t
eg
y
wit
h f
uzzy
po
t
ent
i
a
l (
T
S
-
3)
TS
-
3
co
n
f
ir
m
s
tr
e
n
d
s
ac
r
o
s
s
two
in
d
e
p
en
d
e
n
t
tim
e
win
d
o
w
s
to
r
ed
u
ce
f
alse
en
tr
ies.
T
a
b
le
3
p
r
esen
ts
th
e
d
u
al
-
win
d
o
w
d
esig
n
.
Op
ti
m
ized
r
esu
lts
in
Fig
u
r
e
3
d
e
m
o
n
s
tr
ate
th
at
r
eq
u
ir
in
g
co
n
c
u
r
r
en
ce
ac
r
o
s
s
s
h
o
r
t
(
L
₂)
an
d
lo
n
g
(
L
₁)
win
d
o
ws
h
alv
es
th
e
f
r
eq
u
en
cy
o
f
f
alse
en
tr
ies
an
d
s
tab
ilizes
r
etu
r
n
s
d
u
r
in
g
th
e
5
-
d
ay
test
in
ter
v
al.
Fro
m
a
p
r
ac
tical
p
er
s
p
ec
tiv
e,
T
S
-
3
s
h
o
ws
h
o
w
co
m
b
in
i
n
g
m
u
ltip
le
tim
e
h
o
r
izo
n
s
im
p
r
o
v
es
r
o
b
u
s
tn
ess
—
b
u
t
it'
s
n
o
t
en
o
u
g
h
in
h
ig
h
ly
u
n
p
r
ed
ictab
le
e
n
v
ir
o
n
m
en
ts
.
Stab
ilit
y
an
aly
s
is
was
co
n
d
u
cted
b
y
o
p
tim
izin
g
s
tr
ateg
y
p
ar
a
m
eter
s
o
n
a
1
0
0
-
d
ay
in
ter
v
al
an
d
test
in
g
th
em
o
n
s
u
b
s
eq
u
e
n
t
1
0
-
d
ay
in
ter
v
als.
R
esu
lts
r
ev
ea
led
th
at
p
ar
am
eter
s
o
p
tim
al
f
o
r
o
n
e
r
eg
im
e
f
r
eq
u
e
n
tly
led
to
s
u
b
s
tan
tial
lo
s
s
es
in
th
e
n
ex
t,
h
ig
h
lig
h
tin
g
th
e
f
r
a
g
ilit
y
o
f
n
u
m
e
r
ic
-
o
n
l
y
a
p
p
r
o
ac
h
es
u
n
d
er
ch
ao
tic
d
y
n
a
m
ics.
Fo
r
ex
am
p
le,
t
h
e
T
S
-
1
s
tr
ateg
y
p
r
o
d
u
ce
d
n
eg
ativ
e
r
et
u
r
n
s
in
all
s
u
b
s
eq
u
en
t
in
ter
v
al
s
af
ter
in
itial
o
p
tim
izatio
n
,
wh
ile
T
S
-
2
a
n
d
T
S
-
3
s
h
o
wed
o
n
ly
m
ar
g
i
n
al
im
p
r
o
v
em
en
t.
T
h
is
in
s
tab
ilit
y
u
n
d
er
s
co
r
es
th
e
n
ec
ess
ity
o
f
h
y
b
r
id
d
ec
is
io
n
f
r
am
ewo
r
k
s
th
at
ex
p
licitly
m
o
d
el
r
eg
im
e
s
h
if
ts
an
d
in
co
r
p
o
r
ate
q
u
alitativ
e,
ex
p
er
t
-
b
ased
s
ig
n
als.
T
h
is
f
in
d
in
g
is
co
n
s
is
ten
t
with
th
e
w
o
r
k
[
5
]
an
d
is
f
u
r
th
er
s
u
p
p
o
r
ted
b
y
s
tu
d
ies
o
n
th
e
lim
itatio
n
s
o
f
s
tan
d
ar
d
tim
e
-
s
er
ies
m
o
d
els u
n
d
e
r
n
o
n
-
s
tatio
n
ar
y
c
o
n
d
itio
n
s
[
1
1
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
R
o
b
u
s
t tra
d
in
g
i
n
s
to
ch
a
s
tic
c
h
a
o
s
:
a
h
yb
r
id
fu
z
z
y
mu
lti
-
ex
p
ert a
p
p
r
o
a
ch
(
A
n
d
r
ey
Mu
s
a
ye
v
)
4109
Fig
u
r
e
2
.
T
h
e
T
S
-
2
s
tr
ateg
y
with
o
p
tim
al
p
ar
a
m
eter
s
T
ab
le
3
.
Du
al
-
win
d
o
w
tr
en
d
s
tr
ateg
y
with
p
r
o
s
p
ec
tiv
e
f
u
zz
y
in
teg
r
atio
n
C
o
m
p
o
n
e
n
t
C
h
a
l
l
e
n
g
e
TS
-
3
i
n
n
o
v
a
t
i
o
n
K
e
y
l
i
m
i
t
a
t
i
o
n
s
W
i
n
d
o
w
l
e
n
g
t
h
Tr
a
d
e
-
o
f
f
:
‒
S
h
o
r
t
w
i
n
d
o
w
s
:
n
o
i
se
se
n
si
t
i
v
i
t
y
‒
Lo
n
g
w
i
n
d
o
w
s:
l
a
g
D
u
a
l
-
w
i
n
d
o
w
v
a
l
i
d
a
t
i
o
n
:
t
r
e
n
d
s
mu
st
c
o
n
f
i
r
m i
n
b
o
t
h
t
i
m
e
f
r
a
mes
(
L
₁,
L
₂)
.
D
e
l
a
y
e
d
e
n
t
r
i
e
s
d
u
r
i
n
g
r
a
p
i
d
t
r
e
n
d
s;
mi
sses
e
a
r
l
y
t
r
e
n
d
p
h
a
ses
.
En
t
r
y
c
o
n
d
i
t
i
o
n
F
a
l
se
si
g
n
a
l
s
f
r
o
m
si
n
g
l
e
-
w
i
n
d
o
w
s
y
st
e
ms
Th
r
e
s
h
o
l
d
-
b
a
se
d
:
|
d
y
(
k
-
L
₁,k
)
|
>
d
y
₁*
AN
D
|
d
y
(
k
-
L
₂,k
)
|
>
d
y
₂*
S
t
i
l
l
r
e
a
c
t
i
v
e
;
t
h
r
e
s
h
o
l
d
s
r
e
q
u
i
r
e
f
r
e
q
u
e
n
t
r
e
c
a
l
i
b
r
a
t
i
o
n
.
I
n
i
t
i
a
l
p
e
r
f
o
r
m
a
n
c
e
S
u
b
j
e
c
t
i
v
e
o
p
t
i
m
i
z
a
t
i
o
n
F
u
l
l
p
a
r
a
m
e
t
e
r
se
t
:
P
=
(
α,
L
₁
,
L
₂,
d
y
₁*
,
d
y
₂
*
,
T
P,
S
L
)
O
v
e
r
f
i
t
t
i
n
g
r
i
s
k
;
p
o
o
r
g
e
n
e
r
a
l
i
z
a
t
i
o
n
t
o
n
e
w
d
a
t
a
.
O
p
t
i
mi
z
a
t
i
o
n
M
a
n
u
a
l
t
u
n
i
n
g
i
n
e
f
f
e
c
t
i
v
e
G
r
i
d
s
e
a
r
c
h
:
‒
α
∈
[
0
.
0
0
5
:
0
.
0
2
5
]
‒
L₁
∈
[
1
0
:
5
0
]
,
L₂
∈
[
5
:
1
7
]
C
o
m
p
u
t
a
t
i
o
n
a
l
l
y
i
n
t
e
n
s
i
v
e
;
r
e
s
u
l
t
s
d
e
g
r
a
d
e
i
n
n
o
n
-
st
a
t
i
o
n
a
r
y
m
a
r
k
e
t
s.
P
e
r
f
o
r
ma
n
c
e
H
i
g
h
f
a
l
se
e
n
t
r
i
e
s
(
TS
-
1
/
TS
-
2)
I
mp
r
o
v
e
d
f
i
l
t
e
r
i
n
g
:
‒
F
e
w
e
r
f
a
l
s
e
s
i
g
n
a
l
s
‒
B
e
t
t
e
r
p
r
o
f
i
t
c
o
n
s
i
st
e
n
c
y
V
u
l
n
e
r
a
b
l
e
t
o
c
h
a
o
t
i
c
r
e
v
e
r
sa
l
s;
e
a
r
l
y
e
x
i
t
s
d
u
r
i
n
g
v
o
l
a
t
i
l
i
t
y
.
K
e
y
i
n
s
i
g
h
t
P
u
r
e
l
y
n
u
meri
c
a
l
st
r
a
t
e
g
i
e
s
f
a
i
l
i
n
c
h
a
o
s.
F
o
r
w
a
r
d
p
a
t
h
:
h
y
b
r
i
d
f
u
z
z
y
l
o
g
i
c
+
sen
t
i
m
e
n
t
a
n
a
l
y
si
s
f
o
r
a
d
a
p
t
a
b
i
l
i
t
y
.
R
e
q
u
i
r
e
s
i
n
t
e
g
r
a
t
i
o
n
o
f
q
u
a
l
i
t
a
t
i
v
e
d
a
t
a
(
e
.
g
.
,
n
e
w
s
,
e
v
e
n
t
s)
.
Fig
u
r
e
3
.
T
h
e
T
S
-
3
tr
e
n
d
s
tr
at
eg
y
with
o
p
tim
ize
d
p
ar
am
eter
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
15
,
No
.
5
,
Octo
b
er
2
0
2
6
:
4
1
0
5
-
4
1
1
2
4110
3
.
4
.
Dis
cus
s
io
n
T
h
e
co
m
p
ar
ativ
e
an
aly
s
is
ac
r
o
s
s
T
S
-
1
,
T
S
-
2
,
a
n
d
T
S
-
3
r
e
v
ea
ls
a
clea
r
p
r
o
g
r
ess
io
n
:
a
d
a
p
tiv
e
ex
its
an
d
d
u
al
-
win
d
o
w
c
o
n
f
ir
m
ati
o
n
p
r
o
g
r
ess
iv
ely
r
e
d
u
ce
f
als
e
s
ig
n
als
an
d
im
p
r
o
v
e
p
r
o
f
i
tab
ilit
y
.
Ho
wev
er
,
ev
en
th
e
b
est
n
u
m
er
ical
co
n
f
ig
u
r
atio
n
(
T
S
-
3
)
r
em
ain
s
v
u
ln
er
ab
le
to
ch
ao
tic
r
ev
er
s
als
an
d
p
a
r
am
eter
in
s
tab
ilit
y
ac
r
o
s
s
r
eg
im
es.
T
h
e
p
r
im
ar
y
o
b
s
tacle
is
th
e
in
h
er
en
t
tr
ad
e
-
o
f
f
b
etwe
en
n
o
is
e
s
u
p
p
r
ess
io
n
an
d
r
esp
o
n
s
iv
en
ess
—
s
tr
o
n
g
er
s
m
o
o
th
in
g
in
tr
o
d
u
ce
s
lag
,
wh
ile
wea
k
er
f
ilter
in
g
in
cr
e
ases
f
alse
tr
ig
g
er
s
.
B
id
ir
ec
tio
n
al
ex
p
o
n
e
n
tial f
ilter
in
g
m
itig
ates th
is
lag
(
Fig
u
r
e
4
)
b
u
t c
a
n
n
o
t e
lim
in
ate
it e
n
tir
ely
.
Fig
u
r
e
4
.
C
o
m
p
a
r
is
o
n
o
f
co
n
v
en
tio
n
al
v
s
.
b
id
i
r
ec
tio
n
al
f
ilter
in
g
T
h
e
f
u
zz
y
m
u
lti
-
ex
p
er
t
s
y
s
tem
ad
d
r
ess
es
th
is
lim
itatio
n
b
y
r
eq
u
ir
i
n
g
alig
n
m
en
t
b
etwe
e
n
tech
n
ical
tr
en
d
s
ig
n
als
an
d
q
u
alitativ
e
s
en
tim
en
t
co
n
tex
t.
T
h
is
co
n
s
en
s
u
s
m
ec
h
an
is
m
ef
f
e
ctiv
ely
f
ilter
s
o
u
t
s
p
u
r
io
u
s
tr
ad
es th
at
wo
u
ld
o
t
h
er
wis
e
b
e
ex
ec
u
ted
d
u
r
in
g
n
o
is
y
o
r
r
an
g
e
-
b
o
u
n
d
p
e
r
io
d
s
.
E
m
p
ir
ical
r
e
s
u
lts
o
n
th
e
2
5
-
d
ay
win
d
o
w
d
e
m
o
n
s
tr
ate
s
u
b
s
tan
tial
im
p
r
o
v
em
en
ts
o
v
er
t
h
e
b
est
b
aselin
e:
n
et
p
r
o
f
it
in
cr
ea
s
ed
f
r
o
m
ap
p
r
o
x
im
ately
4
1
0
to
5
6
0
p
i
p
s
,
win
r
ate
f
r
o
m
6
1
%
to
6
7
%,
an
d
m
ax
im
u
m
d
r
awd
o
w
n
d
ec
r
ea
s
ed
f
r
o
m
5
5
to
3
8
p
ip
s
.
T
h
ese
g
ain
s
ar
e
attr
ib
u
ted
to
th
e
s
y
s
tem
'
s
ab
ilit
y
to
d
ef
e
r
d
ec
is
io
n
s
u
n
d
e
r
co
n
f
lictin
g
s
ig
n
als
an
d
ex
p
lo
it o
n
ly
h
ig
h
-
co
n
f
id
e
n
ce
o
p
p
o
r
tu
n
ities
,
as illu
s
tr
ate
d
in
Fig
u
r
e
5
.
Fig
u
r
e
5
.
Fu
zz
y
m
u
lti
-
ex
p
er
t d
ec
is
io
n
p
r
o
ce
s
s
Nev
er
th
eless
,
th
e
h
y
b
r
id
f
r
a
m
ewo
r
k
r
em
ain
s
s
u
b
ject
to
ce
r
tain
lim
itatio
n
s
.
First,
th
e
s
en
tim
en
t
lex
ico
n
m
u
s
t
r
em
ain
r
elev
a
n
t
to
cu
r
r
en
t
m
ar
k
et
n
ar
r
ativ
es
;
ab
r
u
p
t
p
o
licy
s
h
o
ck
s
o
r
s
tr
u
ctu
r
al
b
r
ea
k
s
ca
n
d
eg
r
ad
e
class
if
icatio
n
ac
cu
r
a
cy
[
2
2
]
,
[
2
3
]
.
Seco
n
d
,
th
e
f
u
zz
y
m
em
b
er
s
h
ip
f
u
n
ctio
n
s
w
er
e
p
r
e
-
d
ef
in
e
d
a
n
d
m
ay
n
o
t a
d
ap
t to
e
v
o
lv
in
g
v
o
l
atility
r
eg
im
es [
2
4
]
.
Fu
tu
r
e
wo
r
k
will f
o
cu
s
o
n
o
n
lin
e
r
ec
alib
r
atio
n
o
f
f
u
zz
y
s
ets
v
ia
clu
s
ter
in
g
tech
n
iq
u
es,
d
y
n
am
ic
s
en
tim
en
t
lex
ico
n
u
p
d
ate
s
,
an
d
ex
te
n
s
io
n
to
m
u
lti
-
ass
et
co
r
r
elatio
n
s
ig
n
als
to
p
r
o
v
id
e
ad
d
itio
n
al
r
eg
u
lar
iz
atio
n
d
u
r
i
n
g
tu
r
b
u
len
t c
o
n
d
itio
n
s
[
2
5
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J Ar
tif
I
n
tell
I
SS
N:
2252
-
8
9
3
8
R
o
b
u
s
t tra
d
in
g
i
n
s
to
ch
a
s
tic
c
h
a
o
s
:
a
h
yb
r
id
fu
z
z
y
mu
lti
-
ex
p
ert a
p
p
r
o
a
ch
(
A
n
d
r
ey
Mu
s
a
ye
v
)
4111
4.
CO
NCLU
SI
O
N
T
h
e
test
s
co
n
f
ir
m
th
at
cla
s
s
ical
E
U
R
/USD
tr
en
d
-
f
o
llo
win
g
wo
r
k
s
o
n
ly
d
u
r
i
n
g
s
u
s
tain
ed
,
well
-
d
ef
in
ed
t
r
en
d
s
last
in
g
≥
5
–
1
0
d
ay
s
.
Ho
wev
er
,
u
n
d
er
ch
a
o
tic
co
n
d
itio
n
s
—
m
a
r
k
ed
b
y
i
n
s
tab
ilit
y
,
r
an
d
o
m
f
lu
ctu
atio
n
s
,
an
d
s
en
s
itiv
ity
to
ex
ter
n
al
s
h
o
ck
s
—
th
ese
m
eth
o
d
s
b
ec
o
m
e
u
n
r
eliab
le.
Stra
teg
y
p
er
f
o
r
m
an
ce
b
r
ea
k
s
d
o
wn
q
u
ick
ly
as
m
ar
k
et
r
eg
im
es
s
h
if
t,
m
a
k
in
g
p
r
ev
i
o
u
s
ly
o
p
tim
al
p
ar
am
eter
s
in
ef
f
ec
tiv
e.
T
h
e
ce
n
tr
al
ch
allen
g
e
r
em
ain
s
th
e
b
alan
c
e
b
etwe
en
n
o
is
e
s
u
p
p
r
ess
io
n
an
d
p
r
o
m
p
t
s
ig
n
al
r
esp
o
n
s
e:
s
tr
o
n
g
er
s
m
o
o
th
in
g
in
v
ar
iab
ly
d
elay
s
e
n
tr
y
.
Alth
o
u
g
h
ad
v
a
n
ce
d
s
m
o
o
th
in
g
tech
n
iq
u
es
lik
e
b
id
ir
ec
tio
n
al
f
ilter
s
h
elp
r
ed
u
ce
lag
,
th
ey
ca
n
n
o
t
f
u
lly
o
v
er
c
o
m
e
th
e
u
n
p
r
e
d
ictab
ilit
y
o
f
ch
ao
tic
en
v
ir
o
n
m
e
n
ts
.
E
x
p
o
s
t
p
ar
a
m
eter
o
p
tim
izatio
n
m
ay
b
o
o
s
t
s
h
o
r
t
-
ter
m
r
esu
lts
b
u
t
o
f
ten
lead
s
to
o
v
er
f
itti
n
g
,
lim
itin
g
g
en
er
aliza
tio
n
ac
r
o
s
s
ch
an
g
in
g
co
n
d
itio
n
s
.
Fo
r
r
ea
l
-
wo
r
ld
tr
a
d
er
s
,
esp
ec
ially
i
n
v
o
latile
m
ar
k
ets
lik
e
Fo
r
ex
,
th
is
h
ig
h
lig
h
ts
th
e
n
ee
d
to
g
o
b
ey
o
n
d
p
u
r
e
tr
en
d
s
ig
n
als.
Stra
teg
ies
s
h
o
u
ld
in
co
r
p
o
r
ate
a
d
ap
tiv
e
r
is
k
co
n
tr
o
ls
,
d
iv
er
s
if
ied
p
o
r
tf
o
li
o
s
,
an
d
f
lex
ib
le
ex
it
r
u
les
to
im
p
r
o
v
e
r
esil
ien
ce
.
Desp
ite
th
eir
f
laws
,
tr
en
d
in
d
icato
r
s
r
em
ain
u
s
ef
u
l
wh
en
in
teg
r
ated
in
to
m
u
lti
-
ex
p
er
t
s
y
s
tem
s
b
ased
o
n
f
u
zz
y
lo
g
ic.
T
h
e
f
in
d
in
g
s
s
h
o
w
th
at
co
m
b
in
in
g
q
u
an
titativ
e
tr
en
d
a
n
aly
s
is
with
q
u
alitativ
e
s
en
tim
en
t
ev
alu
atio
n
s
ig
n
if
ican
tly
im
p
r
o
v
es
d
ec
is
io
n
-
m
ak
in
g
u
n
d
e
r
u
n
ce
r
tain
ty
.
B
y
tr
an
s
latin
g
ec
o
n
o
m
ic
n
ews
a
n
d
e
x
p
er
t
o
p
in
i
o
n
s
in
to
lin
g
u
is
tic
v
ar
iab
les,
f
u
zz
y
s
y
s
te
m
s
f
ilter
m
is
lead
in
g
s
ig
n
als
an
d
en
h
an
ce
r
o
b
u
s
tn
ess
.
Desp
ite
th
e
m
ar
k
e
d
im
p
r
o
v
em
en
ts
,
th
e
h
y
b
r
id
f
u
zz
y
s
y
s
tem
is
n
o
t
im
m
u
n
e
t
o
s
tr
u
ctu
r
al
b
r
ea
k
s
o
r
u
n
f
o
r
eseen
s
h
o
ck
s
th
at
ca
n
in
v
alid
ate
p
r
e
-
tr
ain
ed
s
en
tim
en
t
o
r
tr
en
d
m
o
d
els.
As
f
u
t
u
r
e
wo
r
k
,
o
n
lin
e
a
d
ap
tatio
n
o
f
f
u
zz
y
m
em
b
er
s
h
i
p
f
u
n
ctio
n
s
,
d
y
n
am
ic
r
etr
ain
in
g
o
f
s
en
ti
m
en
t
lex
ico
n
s
,
an
d
ex
ten
s
io
n
to
m
u
lti
-
ass
et
an
d
c
r
o
s
s
-
m
ar
k
et
s
ig
n
als
ar
e
p
r
o
m
i
s
in
g
d
ir
ec
tio
n
s
.
Su
ch
d
e
v
elo
p
m
en
ts
wo
u
ld
f
u
r
t
h
er
en
h
an
ce
s
y
s
tem
r
esil
ien
ce
an
d
ad
ap
tab
ilit
y
u
n
d
e
r
ev
er
-
e
v
o
lv
in
g
m
ar
k
et
co
n
d
itio
n
s
.
Fu
tu
r
e
r
esear
ch
s
h
o
u
ld
ex
p
lo
r
e
n
o
n
lin
ea
r
f
u
zz
y
s
tr
ateg
ies
ap
p
lied
to
m
u
ltid
im
en
s
io
n
al
f
in
an
cial
d
ata.
C
ap
tu
r
in
g
in
ter
d
ep
en
d
en
cies
ac
r
o
s
s
m
u
ltip
le
ass
ets
u
s
in
g
f
u
zz
y
in
f
er
e
n
ce
co
u
l
d
s
tab
ilize
f
o
r
ec
asts
an
d
im
p
r
o
v
e
r
elia
b
ilit
y
.
B
u
ild
in
g
o
n
r
ec
en
t
ad
v
an
ce
s
in
ch
ao
tic
tim
e
-
s
er
ies
m
o
d
elin
g
,
f
u
r
th
e
r
s
tu
d
ies
s
h
o
u
ld
r
ef
in
e
p
ar
am
eter
a
d
ap
tatio
n
,
o
p
tim
ize
p
r
ep
r
o
ce
s
s
in
g
f
o
r
f
u
zz
y
i
n
p
u
ts
,
an
d
b
en
ch
m
a
r
k
h
y
b
r
i
d
s
y
s
tem
s
ag
ain
s
t tr
ad
itio
n
al
ap
p
r
o
ac
h
es.
F
UNDING
I
NF
O
R
M
A
T
I
O
N
T
h
is
r
esear
ch
was su
p
p
o
r
ted
b
y
S
t.
Peter
s
b
u
r
g
State
Un
iv
er
s
ity
,
p
r
o
ject
I
D:
1
2
1
0
6
1
0
0
0
1
5
9
-
6.
AUTHO
R
CO
NT
RI
B
UT
I
O
NS ST
A
T
E
M
E
N
T
T
h
is
jo
u
r
n
al
u
s
es
th
e
C
o
n
tr
ib
u
to
r
R
o
les
T
ax
o
n
o
m
y
(
C
R
ed
iT)
to
r
ec
o
g
n
ize
in
d
iv
id
u
al
au
th
o
r
co
n
tr
ib
u
tio
n
s
,
r
ed
u
ce
au
th
o
r
s
h
ip
d
is
p
u
tes,
an
d
f
ac
ilit
ate
co
llab
o
r
atio
n
.
Na
m
e
o
f
Aut
ho
r
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
An
d
r
ey
Mu
s
ay
e
v
✓
✓
✓
✓
✓
✓
✓
✓
Dm
itry
Gr
ig
o
r
iev
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
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
f
t
w
a
r
e
Va
:
Va
l
i
d
a
t
i
o
n
Fo
:
Fo
r
mal
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
so
u
r
c
e
s
D
:
D
a
t
a
C
u
r
a
t
i
o
n
O
:
W
r
i
t
i
n
g
-
O
r
i
g
i
n
a
l
D
r
a
f
t
E
:
W
r
i
t
i
n
g
-
R
e
v
i
e
w
&
E
d
i
t
i
n
g
Vi
:
Vi
su
a
l
i
z
a
t
i
o
n
Su
:
Su
p
e
r
v
i
s
i
o
n
P
:
P
r
o
j
e
c
t
a
d
mi
n
i
st
r
a
t
i
o
n
Fu
:
Fu
n
d
i
n
g
a
c
q
u
i
si
t
i
o
n
CO
NF
L
I
C
T
O
F
I
N
T
E
R
E
S
T
ST
A
T
E
M
E
NT
T
h
e
au
th
o
r
s
d
ec
lar
e
n
o
co
n
f
licts
o
f
in
ter
est.
DATA AV
AI
L
AB
I
L
I
T
Y
T
h
e
d
ata
th
at
s
u
p
p
o
r
t
th
e
f
in
d
in
g
s
o
f
th
is
s
tu
d
y
ar
e
o
p
en
ly
av
ailab
le
in
Fin
am
at
h
ttp
s
://www.
f
in
am
.
r
u
/ v
ia
T
r
a
d
e
API
.
RE
F
E
R
E
NC
E
S
[
1
]
E.
E
.
P
e
t
e
r
s,
C
h
a
o
s
a
n
d
o
rd
e
r
i
n
t
h
e
c
a
p
i
t
a
l
m
a
rke
t
s:
a
n
e
w
v
i
e
w
o
f
c
y
c
l
e
s,
p
ri
c
e
s,
a
n
d
m
a
r
k
e
t
v
o
l
a
t
i
l
i
t
y
.
H
o
b
o
k
e
n
,
U
n
i
t
e
d
S
t
a
t
e
s
:
Jo
h
n
W
i
l
e
y
& S
o
n
s,
1
9
9
6
.
[
2
]
J.
G
.
W
i
l
l
i
a
ms
a
n
d
B
.
M
.
W
i
l
l
i
a
ms,
T
ra
d
i
n
g
c
h
a
o
s:
m
a
x
i
m
i
ze
p
r
o
f
i
t
s
w
i
t
h
p
r
o
v
e
n
t
e
c
h
n
i
c
a
l
t
e
c
h
n
i
q
u
e
s
.
H
o
b
o
k
e
n
,
U
n
i
t
e
d
S
t
a
t
e
s
:
Jo
h
n
W
i
l
e
y
& S
o
n
s,
2
0
0
4
.
[
3
]
T.
L
u
x
,
“
S
t
o
c
h
a
s
t
i
c
b
e
h
a
v
i
o
r
a
l
a
ss
e
t
-
p
r
i
c
i
n
g
mo
d
e
l
s
a
n
d
t
h
e
st
y
l
i
z
e
d
f
a
c
t
s
,
”
i
n
H
a
n
d
b
o
o
k
o
f
Fi
n
a
n
c
i
a
l
M
a
rke
t
s:
D
y
n
a
m
i
c
s
a
n
d
Ev
o
l
u
t
i
o
n
,
El
s
e
v
i
e
r
,
2
0
0
9
,
p
p
.
1
6
1
–
215
,
d
o
i
:
1
0
.
1
0
1
6
/
B
9
7
8
-
0
1
2
3
7
4
2
5
8
-
2
.
5
0
0
0
7
-
5.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
2
5
2
-
8
9
3
8
I
n
t J Ar
tif
I
n
tell
,
Vo
l.
15
,
No
.
5
,
Octo
b
er
2
0
2
6
:
4
1
0
5
-
4
1
1
2
4112
[
4
]
L.
R
u
s
h
a
n
d
K
.
A
.
B
l
o
u
n
t
,
C
h
a
o
s
t
h
e
o
ry
.
F
o
r
t
C
o
l
l
i
n
s,
U
n
i
t
e
d
S
t
a
t
e
s:
E
n
t
a
n
g
l
e
d
P
u
b
l
i
sh
i
n
g
,
2
0
2
1
.
[
5
]
A
.
M
u
sa
e
v
a
n
d
D
.
G
r
i
g
o
r
i
e
v
,
“
M
u
l
t
i
-
e
x
p
e
r
t
sy
s
t
e
ms
:
f
u
n
d
a
me
n
t
a
l
c
o
n
c
e
p
t
s
a
n
d
a
p
p
l
i
c
a
t
i
o
n
e
x
a
mp
l
e
s
,
”
J
o
u
r
n
a
l
o
f
T
h
e
o
r
e
t
i
c
a
l
a
n
d
Ap
p
l
i
e
d
I
n
f
o
rm
a
t
i
o
n
T
e
c
h
n
o
l
o
g
y
,
v
o
l
.
1
0
0
,
n
o
.
2
,
p
p
.
3
3
6
–
3
4
8
,
2
0
2
2
.
[
6
]
E.
S
c
h
ö
l
l
a
n
d
H
.
G
.
S
c
h
u
st
e
r
,
H
a
n
d
b
o
o
k
o
f
c
h
a
o
s
c
o
n
t
r
o
l
.
W
e
i
n
h
e
i
m,
G
e
r
m
a
n
y
:
W
i
l
e
y
,
2
0
0
7
,
d
o
i
:
1
0
.
1
0
0
2
/
9
7
8
3
5
2
7
6
2
2
3
1
3
.
[
7
]
P
.
J.
H
u
b
e
r
,
Ro
b
u
st
st
a
t
i
st
i
c
s
.
H
o
b
o
k
e
n
,
U
n
i
t
e
d
S
t
a
t
e
s:
Jo
h
n
W
i
l
e
y
& S
o
n
s
,
2
0
0
4
.
[
8
]
R
.
A
.
M
a
r
o
n
n
a
,
R
.
D
.
M
a
r
t
i
n
,
V
.
J.
Y
o
h
a
i
,
a
n
d
M
.
S
.
-
B
a
r
r
e
r
a
,
Ro
b
u
st
s
t
a
t
i
st
i
c
s:
t
h
e
o
ry
a
n
d
m
e
t
h
o
d
s
(
w
i
t
h
R)
.
H
o
b
o
k
e
n
,
N
e
w
Jersey
:
J
o
h
n
W
i
l
e
y
& S
o
n
s,
2
0
1
9
.
[
9
]
G
.
C
h
e
n
a
n
d
X
.
Y
u
,
C
h
a
o
s
c
o
n
t
r
o
l
:
t
h
e
o
r
y
a
n
d
a
p
p
l
i
c
a
t
i
o
n
s
.
B
e
r
l
i
n
,
G
e
r
m
a
n
y
:
S
p
r
i
n
g
e
r
,
2
0
0
3
.
[
1
0
]
Y
.
B
o
l
o
t
i
n
,
A
.
T
u
r
,
a
n
d
V
.
Y
a
n
o
v
s
k
y
,
C
h
a
o
s:
c
o
n
c
e
p
t
s,
c
o
n
t
ro
l
a
n
d
c
o
n
s
t
ru
c
t
i
v
e
u
s
e
.
B
e
r
l
i
n
,
G
e
r
ma
n
y
:
S
p
r
i
n
g
e
r
,
2
0
0
9
.
[
1
1
]
G
.
E.
P
.
B
o
x
,
G
.
M
.
Je
n
k
i
n
s,
a
n
d
G
.
C
.
R
e
i
n
se
l
,
T
i
m
e
se
ri
e
s
a
n
a
l
y
s
i
s:
f
o
rec
a
st
i
n
g
a
n
d
c
o
n
t
r
o
l
.
H
o
b
o
k
e
n
,
U
n
i
t
e
d
S
t
a
t
e
s
:
J
o
h
n
W
i
l
e
y
& So
n
s
,
2
0
0
8
,
d
o
i
:
1
0
.
1
0
0
2
/
9
7
8
1
1
1
8
6
1
9
1
9
3
.
[
1
2
]
K
.
D
u
,
F
.
X
i
n
g
,
R
.
M
a
o
,
a
n
d
E.
C
a
m
b
r
i
a
,
“
F
i
n
a
n
c
i
a
l
s
e
n
t
i
me
n
t
a
n
a
l
y
si
s:
t
e
c
h
n
i
q
u
e
s
a
n
d
a
p
p
l
i
c
a
t
i
o
n
s,
”
A
C
M
C
o
m
p
u
t
i
n
g
S
u
rv
e
y
s
,
v
o
l
.
5
6
,
n
o
.
9
,
p
p
.
1
–
4
2
,
O
c
t
.
2
0
2
4
,
d
o
i
:
1
0
.
1
1
4
5
/
3
6
4
9
4
5
1
.
[
1
3
]
G
.
F
a
t
o
u
r
o
s
,
J
.
S
o
l
d
a
t
o
s
,
K
.
K
o
u
r
o
u
m
a
l
i
,
G
.
M
a
k
r
i
d
i
s,
a
n
d
D
.
K
y
r
i
a
z
i
s,
“
Tr
a
n
sf
o
r
mi
n
g
se
n
t
i
me
n
t
a
n
a
l
y
si
s
i
n
t
h
e
f
i
n
a
n
c
i
a
l
d
o
ma
i
n
w
i
t
h
C
h
a
t
G
P
T,
”
M
a
c
h
i
n
e
L
e
a
r
n
i
n
g
w
i
t
h
A
p
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
1
4
,
D
e
c
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
ml
w
a
.
2
0
2
3
.
1
0
0
5
0
8
.
[
1
4
]
A
.
K
o
n
a
r
a
n
d
D
.
B
h
a
t
t
a
c
h
a
r
y
a
,
T
i
m
e
-
seri
e
s
p
r
e
d
i
c
t
i
o
n
a
n
d
a
p
p
l
i
c
a
t
i
o
n
s:
a
m
a
c
h
i
n
e
i
n
t
e
l
l
i
g
e
n
c
e
a
p
p
r
o
a
c
h
.
C
h
a
m,
S
w
i
t
z
e
r
l
a
n
d
:
S
p
r
i
n
g
e
r
,
2
0
1
7
,
d
o
i
:
1
0
.
1
0
0
7
/
9
7
8
-
3
-
3
1
9
-
5
4
5
9
7
-
4.
[
1
5
]
A
.
R
i
v
a
e
t
a
l
.
,
“
A
d
d
r
e
ss
i
n
g
n
o
n
-
s
t
a
t
i
o
n
a
r
i
t
y
i
n
F
X
t
r
a
d
i
n
g
w
i
t
h
o
n
l
i
n
e
m
o
d
e
l
se
l
e
c
t
i
o
n
o
f
o
f
f
l
i
n
e
R
L
e
x
p
e
r
t
s
,
”
I
C
AI
F
'
2
2
:
T
h
i
r
d
AC
M
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
AI
i
n
Fi
n
a
n
c
e
,
2
0
2
2
,
p
p
.
3
9
4
-
4
0
2
,
d
o
i
:
1
0
.
1
1
4
5
/
3
5
3
3
2
7
1
.
3
5
6
1
7
8
0
.
[
1
6
]
R
.
C
h
a
u
h
a
n
a
n
d
H
.
T
y
a
g
i
,
“
T
e
x
t
m
i
n
i
n
g
t
e
c
h
n
i
q
u
e
s
f
o
r
e
x
t
r
a
c
t
i
o
n
a
n
d
r
e
t
r
i
e
v
a
l
o
f
i
n
f
o
r
ma
t
i
o
n
i
n
r
e
s
e
a
r
c
h
p
r
o
c
e
ss,”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
f
Re
s
e
a
r
c
h
i
n
E
l
e
c
t
r
o
n
i
c
s a
n
d
C
o
m
p
u
t
e
r
E
n
g
i
n
e
e
r
i
n
g
,
v
o
l
.
7
,
n
o
.
2
,
p
p
.
2
6
9
3
–
2
7
0
0
,
2
0
1
9
.
[
1
7
]
I
.
G
e
o
r
g
e
s
c
u
a
n
d
J
.
K
i
n
n
u
n
e
n
,
“
E
n
t
r
o
p
y
a
n
d
c
h
a
o
s
-
b
a
s
e
d
m
o
d
e
l
i
n
g
o
f
n
o
n
l
i
n
e
a
r
d
e
p
e
n
d
e
n
c
i
e
s
i
n
c
o
mm
o
d
i
t
y
mark
e
t
s,
”
E
n
t
r
o
p
y
,
v
o
l
.
2
7
,
n
o
.
9
,
S
e
p
.
2
0
2
5
,
d
o
i
:
1
0
.
3
3
9
0
/
e
2
7
0
9
0
9
5
5
.
[
1
8
]
D
.
P
.
F
e
l
d
m
a
n
,
C
h
a
o
s
a
n
d
d
y
n
a
m
i
c
a
l
sy
s
t
e
m
s
.
P
r
i
n
c
e
t
o
n
,
U
n
i
t
e
d
S
t
a
t
e
s
:
P
r
i
n
c
e
t
o
n
U
n
i
v
e
r
s
i
t
y
P
r
e
ss
,
2
0
1
9
,
d
o
i
:
1
0
.
2
3
0
7
/
j
.
c
t
v
c
5
p
c
z
n
.
[
1
9
]
C
.
O
e
s
t
r
e
i
c
h
e
r
,
“
A
h
i
s
t
o
r
y
o
f
c
h
a
o
s
t
h
e
o
r
y
,
”
D
i
a
l
o
g
u
e
s
i
n
C
l
i
n
i
c
a
l
N
e
u
r
o
sc
i
e
n
c
e
,
v
o
l
.
9
,
n
o
.
3
,
p
p
.
2
7
9
–
2
8
9
,
S
e
p
.
2
0
0
7
,
d
o
i
:
1
0
.
3
1
8
8
7
/
D
C
N
S
.
2
0
0
7
.
9
.
3
/
c
o
e
s
t
r
e
i
c
h
e
r
.
[
2
0
]
S
.
B
o
c
c
a
l
e
t
t
i
,
C
.
G
r
e
b
o
g
i
,
Y
.
-
C
.
La
i
,
H
.
M
a
n
c
i
n
i
,
a
n
d
D
.
M
a
z
a
,
“
T
h
e
c
o
n
t
r
o
l
o
f
c
h
a
o
s:
t
h
e
o
r
y
a
n
d
a
p
p
l
i
c
a
t
i
o
n
s
,
”
Ph
y
si
c
s
R
e
p
o
r
t
s
,
v
o
l
.
3
2
9
,
n
o
.
3
,
p
p
.
1
0
3
–
1
9
7
,
M
a
y
2
0
0
0
,
d
o
i
:
1
0
.
1
0
1
6
/
S
0
3
7
0
-
1
5
7
3
(
9
9
)
0
0
0
9
6
-
4.
[
2
1
]
E.
S
.
G
a
r
d
n
e
r
,
“
E
x
p
o
n
e
n
t
i
a
l
sm
o
o
t
h
i
n
g
:
Th
e
s
t
a
t
e
o
f
t
h
e
a
r
t
,
”
J
o
u
r
n
a
l
o
f
Fo
r
e
c
a
st
i
n
g
,
v
o
l
.
4
,
n
o
.
1
,
p
p
.
1
–
2
8
,
Ja
n
.
1
9
8
5
,
d
o
i
:
1
0
.
1
0
0
2
/
f
o
r
.
3
9
8
0
0
4
0
1
0
3
.
[
2
2
]
C
.
Tr
u
o
n
g
,
L.
O
u
d
r
e
,
a
n
d
N
.
V
a
y
a
t
i
s,
“
S
e
l
e
c
t
i
v
e
r
e
v
i
e
w
o
f
o
f
f
l
i
n
e
c
h
a
n
g
e
p
o
i
n
t
d
e
t
e
c
t
i
o
n
m
e
t
h
o
d
s
,
”
S
i
g
n
a
l
Pr
o
c
e
ssi
n
g
,
v
o
l
.
1
6
7
,
F
e
b
.
2
0
2
0
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
s
i
g
p
r
o
.
2
0
1
9
.
1
0
7
2
9
9
.
[
2
3
]
M
.
G
u
p
t
a
,
R
.
W
a
d
h
v
a
n
i
,
a
n
d
A
.
R
a
so
o
l
,
“
C
o
m
p
r
e
h
e
n
s
i
v
e
a
n
a
l
y
si
s
o
f
c
h
a
n
g
e
-
p
o
i
n
t
d
y
n
a
mi
c
s
d
e
t
e
c
t
i
o
n
i
n
t
i
me
s
e
r
i
e
s
d
a
t
a
:
a
r
e
v
i
e
w
,
”
Ex
p
e
r
t
S
y
s
t
e
m
s
w
i
t
h
A
p
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
2
4
8
,
A
u
g
.
2
0
2
4
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
e
sw
a
.
2
0
2
4
.
1
2
3
3
4
2
.
[
2
4
]
M
.
S
.
I
sl
a
m,
E.
H
o
ss
a
i
n
,
A
.
R
a
h
ma
n
,
M
.
S
.
H
o
ss
a
i
n
,
a
n
d
K
.
A
n
d
e
r
sso
n
,
“
A
r
e
v
i
e
w
o
n
r
e
c
e
n
t
a
d
v
a
n
c
e
me
n
t
s
i
n
F
o
r
e
x
c
u
r
r
e
n
c
y
p
r
e
d
i
c
t
i
o
n
,
”
A
l
g
o
ri
t
h
m
s
,
v
o
l
.
1
3
,
n
o
.
8
,
Ju
l
.
2
0
2
0
,
d
o
i
:
1
0
.
3
3
9
0
/
a
1
3
0
8
0
1
8
6
.
[
2
5
]
K
.
K
a
c
z
mare
k
,
L.
D
y
m
o
v
a
,
a
n
d
P
.
S
e
v
a
st
j
a
n
o
v
,
“
I
n
t
u
i
t
i
o
n
i
s
t
i
c
f
u
z
z
y
r
u
l
e
-
b
a
s
e
e
v
i
d
e
n
t
i
a
l
r
e
a
s
o
n
i
n
g
w
i
t
h
a
p
p
l
i
c
a
t
i
o
n
t
o
t
h
e
c
u
r
r
e
n
c
y
t
r
a
d
i
n
g
sy
s
t
e
m
o
n
t
h
e
F
o
r
e
x
mark
e
t
,
”
A
p
p
l
i
e
d
S
o
f
t
C
o
m
p
u
t
i
n
g
,
v
o
l
.
1
2
8
,
O
c
t
.
2
0
2
2
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
a
so
c
.
2
0
2
2
.
1
0
9
5
2
2
.
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
And
r
e
y
Mu
s
a
y
e
v
is
a
n
a
ss
o
c
iate
p
ro
fe
ss
o
r
a
t
IT
M
O
U
n
iv
e
rsit
y
a
n
d
S
t.
P
e
ters
b
u
rg
S
tate
Tec
h
n
o
l
o
g
ica
l
In
stit
u
te
(Tec
h
n
ica
l
Un
i
v
e
rsity
)
.
He
h
o
ld
s
a
P
h
.
D.
i
n
Tec
h
n
ica
l
S
c
ien
c
e
s
(Ca
n
d
id
a
te
o
f
Tec
h
n
ica
l
S
c
ien
c
e
s)
fro
m
t
h
e
F
e
d
e
ra
l
Re
se
a
rc
h
Ce
n
ter
o
f
th
e
R
u
ss
ia
n
Ac
a
d
e
m
y
o
f
S
c
ien
c
e
s.
His
r
e
se
a
rc
h
in
tere
sts
i
n
c
lu
d
e
a
rti
ficia
l
in
telli
g
e
n
c
e
,
d
a
ta
a
n
a
ly
sis,
d
y
n
a
m
ica
l
c
h
a
o
s,
a
n
d
f
o
re
c
a
stin
g
o
f
c
o
m
p
lex
m
u
lt
i
d
ime
n
sio
n
a
l
p
ro
c
e
ss
e
s
.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
a
m
u
sa
y
e
v
1
9
9
0
@g
m
a
il
.
c
o
m
.
Dm
itry
G
r
ig
o
r
iev
is
a
n
a
ss
o
c
iate
p
ro
fe
ss
o
r
in
th
e
De
p
a
rtme
n
t
o
f
In
fo
rm
a
ti
c
s
a
t
th
e
F
a
c
u
lt
y
o
f
M
a
t
h
e
m
a
ti
c
s
a
n
d
Co
m
p
u
ter
S
c
ien
c
e
,
S
t.
P
e
ters
b
u
r
g
S
tate
Un
i
v
e
rsity
.
He
e
a
rn
e
d
h
is
P
h
.
D
.
in
P
h
y
sic
s
a
n
d
M
a
th
e
m
a
ti
c
s
fro
m
S
t.
P
e
ters
b
u
rg
S
tate
Un
iv
e
rsity
.
His
re
se
a
rc
h
in
tere
sts
in
c
lu
d
e
m
u
lt
i
-
a
g
e
n
t
sy
st
e
m
s,
d
a
ta
a
n
a
ly
sis,
m
a
c
h
in
e
lea
rn
in
g
,
a
n
d
e
c
o
n
o
m
e
tri
c
s.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
g
ri
d
m
e
r@m
a
il
.
ru
.
Evaluation Warning : The document was created with Spire.PDF for Python.