T
E
L
K
O
M
NIKA
T
elec
o
mm
un
ica
t
io
n,
Co
m
pu
t
ing
,
E
lect
ro
nics
a
nd
Co
ntr
o
l
Vo
l.
18
,
No
.
4
,
A
u
g
u
s
t
2020
,
p
p
.
1
8
0
2
~
1
8
0
8
I
SS
N:
1
6
9
3
-
6
9
3
0
,
ac
cr
ed
ited
First Gr
ad
e
b
y
Ke
m
e
n
r
is
te
k
d
i
k
ti,
Dec
r
ee
No
: 2
1
/E/KPT
/2
0
1
8
DOI
: 1
0
.
1
2
9
2
8
/
T
E
L
KOM
NI
KA
.
v
1
8
i4
.
1
3
1
9
2
1802
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//jo
u
r
n
a
l.u
a
d
.
a
c.
id
/in
d
ex
.
p
h
p
/TELK
OM
N
I
K
A
Appro
x
i
m
a
tion
of
regress
io
n
-
ba
sed
fault
m
ini
m
i
z
a
tion for
netw
o
rk
t
ra
ff
ic
Cha
nin
t
o
rn
J
it
t
a
w
iriy
a
nu
k
o
o
n
G
ra
d
u
a
te S
c
h
o
o
l
o
f
A
d
v
a
n
c
e
d
T
e
c
h
n
o
l
o
g
y
M
a
n
a
g
e
m
e
n
t,
A
ss
u
m
p
ti
o
n
U
n
iv
e
rsity
,
T
h
a
il
a
n
d
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Ma
y
22
,
2019
R
ev
i
s
ed
A
p
r
1
,
2
0
2
0
A
cc
ep
ted
A
p
r
13
,
2
0
2
0
T
h
is
re
se
a
rc
h
a
ss
o
c
iate
s
th
re
e
d
isti
n
c
t
a
p
p
ro
a
c
h
e
s
f
o
r
c
o
m
p
u
ter
n
e
t
w
o
rk
tra
ff
ic
p
re
d
ictio
n
.
T
h
e
y
a
r
e
th
e
trad
it
io
n
a
l
sto
c
h
a
stic
g
ra
d
ien
t
d
e
sc
e
n
t
(S
G
D)
u
sin
g
a
fe
w
ra
n
d
o
m
sa
m
p
li
n
g
s
in
ste
a
d
o
f
th
e
c
o
m
p
lete
d
a
tas
e
t
f
o
r
e
a
c
h
it
e
ra
ti
v
e
c
a
lcu
latio
n
,
th
e
g
ra
d
ien
t
d
e
sc
e
n
t
a
lg
o
rit
h
m
(G
D
A
)
w
h
ich
is
a
w
e
ll
-
k
n
o
w
n
o
p
ti
m
iza
ti
o
n
a
p
p
ro
a
c
h
i
n
d
e
e
p
lea
rn
in
g
,
a
n
d
t
h
e
p
r
o
p
o
se
d
m
e
th
o
d
.
T
h
e
n
e
t
w
o
rk
tra
ff
ic
is
c
o
m
p
u
ted
f
ro
m
th
e
tra
ff
ic
lo
a
d
(d
a
ta
a
n
d
m
u
lt
im
e
d
ia)
o
f
th
e
c
o
m
p
u
ter
n
e
tw
o
rk
n
o
d
e
s
v
ia
th
e
I
n
tern
e
t.
It
is
a
p
p
a
re
n
t
th
a
t
th
e
S
G
D
is
a
m
o
d
e
st
it
e
ra
ti
o
n
b
u
t
c
a
n
c
o
n
c
lu
d
e
su
b
o
p
ti
m
a
l
so
lu
ti
o
n
s.
T
h
e
G
D
A
is
a
c
o
m
p
li
c
a
ted
o
n
e
,
c
a
n
f
u
n
c
ti
o
n
m
o
re
a
c
c
u
ra
te
th
a
n
th
e
S
G
D
b
u
t
d
if
f
icu
lt
to
m
a
n
ip
u
late
p
a
ra
m
e
ters
,
su
c
h
a
s
th
e
lea
rn
in
g
ra
te,
th
e
d
a
tas
e
t
g
ra
n
u
larity
,
a
n
d
th
e
lo
ss
f
u
n
c
ti
o
n
.
Ne
tw
o
rk
tr
a
ff
ic
e
sti
m
a
ti
o
n
h
e
lp
s
im
p
ro
v
e
p
e
rf
o
r
m
a
n
c
e
a
n
d
lo
w
e
r
c
o
sts
f
o
r
v
a
rio
u
s
a
p
p
l
ica
ti
o
n
s,
su
c
h
a
s
a
n
a
d
a
p
t
iv
e
ra
te
c
o
n
tr
o
l,
lo
a
d
b
a
lan
c
i
n
g
,
t
h
e
q
u
a
li
ty
o
f
se
rv
ice
(Qo
S
),
f
a
ir
b
a
n
d
w
id
th
a
ll
o
c
a
ti
o
n
,
a
n
d
a
n
o
m
a
l
y
d
e
tec
ti
o
n
.
T
h
e
p
ro
p
o
se
d
m
e
th
o
d
c
o
n
f
i
rm
s
o
p
ti
m
a
l
v
a
lu
e
s
o
u
t
o
f
p
a
ra
m
e
ters
u
sin
g
si
m
u
latio
n
t
o
c
o
m
p
u
te
th
e
m
in
im
u
m
f
ig
u
re
o
f
sp
e
c
i
f
ied
lo
ss
f
u
n
c
ti
o
n
in
e
a
c
h
it
e
ra
ti
o
n
.
K
ey
w
o
r
d
s
:
A
p
p
r
o
x
i
m
atio
n
Dee
p
lear
n
in
g
E
r
r
o
r
m
in
i
m
iza
tio
n
Net
w
o
r
k
tr
af
f
ic
R
eg
r
es
s
io
n
-
b
ased
p
r
ed
ictio
n
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r
th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
C
h
a
n
i
n
to
r
n
J
itta
w
ir
i
y
an
u
k
o
o
n
,
Gr
ad
u
ate
Sch
o
o
l o
f
A
d
v
a
n
ce
d
T
ec
h
n
o
lo
g
y
Ma
n
a
g
e
m
e
n
t,
Ass
u
m
p
t
io
n
U
n
iv
er
s
it
y
,
88
Mo
o
8
,
B
an
g
Na
T
r
ad
Km
2
6
,
B
an
g
Sao
T
h
o
n
g
Dis
tr
ict,
Sa
m
u
t P
r
ak
an
P
r
o
v
in
ce
,
T
h
ail
an
d
,
1
0
5
7
0
.
E
m
ail:
p
ct2
5
2
6
@
y
a
h
o
o
.
co
m
1.
I
NT
RO
D
UCT
I
O
N
C
u
r
r
e
n
t
ly
,
c
o
m
p
u
t
e
r
n
e
t
w
o
r
k
s
a
n
d
n
e
t
w
o
r
k
t
r
af
f
i
c
r
e
q
u
e
s
t
s
h
av
e
b
e
en
p
r
o
m
p
t
ly
am
ass
e
d
in
m
an
y
ty
p
es
o
f
s
o
f
tw
a
r
e
i
n
th
e
5
G
d
at
a
n
e
tw
o
r
k
.
T
h
u
s
,
s
ci
en
t
is
ts
h
av
e
s
tu
d
i
e
d
t
h
e
im
p
a
ct
s
o
f
th
i
s
p
r
o
g
r
es
s
i
o
n
b
as
e
d
o
n
u
s
e
r
s
’
r
e
q
u
es
t
s
f
r
o
m
th
e
n
e
tw
o
r
k
.
N
o
n
e
th
el
e
s
s
,
th
e
t
r
af
f
ic
b
a
s
e
d
o
n
d
a
t
a
s
t
r
ea
m
h
as
b
e
en
r
e
g
a
r
d
e
d
as
c
r
i
t
i
c
a
l
t
r
af
f
i
c
f
r
o
m
u
s
e
r
s
.
A
l
s
h
af
lu
t
an
d
T
h
ay
an
an
th
an
[
1
]
in
s
p
e
c
t
m
ass
iv
e
d
a
ta
t
r
af
f
i
c
f
o
r
em
p
o
w
e
r
in
g
a
s
p
e
c
t
r
u
m
s
h
a
r
in
g
o
v
e
r
w
ir
e
l
es
s
n
e
tw
o
r
k
.
S
ev
e
r
a
l
p
l
an
s
a
r
e
m
en
t
i
o
n
e
d
t
o
h
an
d
l
e
th
i
s
n
etw
o
r
k
t
r
af
f
i
c
.
A
u
th
o
r
s
a
ls
o
p
r
o
p
o
s
e
a
t
r
af
f
i
c
m
o
d
el
t
o
l
o
w
e
r
th
e
d
e
l
ay
f
o
r
t
h
e
u
s
e
r
’
s
r
e
q
u
es
ts
.
T
h
ey
t
ak
e
d
iv
e
r
s
e
im
p
l
ic
a
t
i
o
n
s
a
s
s
o
c
i
a
te
d
t
o
a
g
ig
a
n
ti
c
s
i
ze
o
f
t
r
af
f
i
c
in
p
r
a
c
t
i
c
e
.
T
h
ey
c
o
n
s
i
d
e
r
th
e
t
r
af
f
i
c
f
l
o
w
,
s
t
a
r
t
in
g
f
r
o
m
th
e
a
cc
e
s
s
p
o
in
t
t
o
th
e
a
r
r
i
v
a
l
o
f
r
e
q
u
e
s
ts
t
o
th
e
c
a
r
r
i
e
r
s
.
T
h
i
s
r
es
e
a
r
ch
[
2
]
c
o
m
p
a
r
e
s
f
o
u
r
d
if
f
e
r
en
t
t
e
ch
n
i
q
u
es
f
o
r
n
e
t
w
o
r
k
t
r
af
f
i
c
p
r
e
d
i
c
ti
o
n
.
T
h
ey
a
r
e
m
u
l
ti
-
l
ay
e
r
p
e
r
c
e
p
t
r
o
n
w
ith
an
d
w
i
th
o
u
t
r
e
s
i
li
en
t
b
a
ck
p
r
o
p
a
g
at
i
o
n
,
r
e
cu
r
r
en
t
n
eu
r
a
l
n
e
t
w
o
r
k
,
an
d
s
t
ac
k
e
d
au
t
o
-
en
c
o
d
in
g
.
A
u
t
h
o
r
s
f
o
cu
s
o
n
t
im
e
s
e
r
i
es
t
r
a
f
f
i
c
o
v
e
r
th
e
I
n
t
e
r
n
et
an
d
d
is
co
v
e
r
th
at
b
o
th
r
e
cu
r
r
en
t
n
eu
r
a
l
n
e
t
w
o
r
k
a
n
d
m
u
lt
i
-
l
ay
e
r
p
e
r
c
e
p
t
r
o
n
a
p
p
r
o
a
c
h
es
a
r
e
b
e
t
t
e
r
t
h
an
th
e
r
e
s
t
.
T
h
e
im
p
o
r
t
a
n
c
e
o
f
c
o
n
te
n
t
f
r
o
m
s
e
r
v
i
ce
p
r
o
v
i
d
e
r
s
d
e
v
e
l
o
p
s
t
h
e
n
e
tw
o
r
k
t
r
af
f
i
c
th
r
o
u
g
h
d
a
ta
c
en
t
e
r
r
es
o
u
r
c
e
s
.
T
h
e
p
u
b
l
i
c
at
i
o
n
[
3
]
p
r
o
p
o
s
e
s
t
h
e
em
p
l
o
y
m
e
n
t
o
f
c
o
n
v
o
lu
ti
o
n
a
l
n
eu
r
a
l
n
e
tw
o
r
k
s
t
o
e
s
tim
a
t
e
s
h
o
r
t
-
t
e
r
m
v
a
r
i
a
ti
o
n
s
i
n
th
e
t
r
af
f
i
c
am
o
u
n
t
th
r
o
u
g
h
th
e
d
a
t
a
c
e
n
te
r
.
T
h
e
v
i
r
tu
al
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
A
p
p
r
o
xima
tio
n
o
f
r
eg
r
ess
io
n
-
b
a
s
ed
fa
u
lt min
imiz
a
tio
n
fo
r
n
etw
o
r
k
tr
a
ffic
(
C
h
a
n
in
to
r
n
Jitta
w
ir
iya
n
u
ko
o
n
)
1803
m
a
ch
in
e
i
s
u
s
e
d
t
o
c
o
n
f
ig
u
r
e
th
e
d
a
t
a
c
e
n
t
e
r
en
v
i
r
o
n
m
en
t
.
A
u
th
o
r
s
p
o
in
t
o
u
t
th
a
t
t
h
e
c
o
n
v
o
lu
ti
o
n
a
l
n
eu
r
al
n
etw
o
r
k
a
c
h
ie
v
e
s
th
e
n
o
n
-
l
in
e
a
r
n
e
tw
o
r
k
t
r
a
f
f
i
c
,
g
iv
i
n
g
im
p
r
o
v
em
en
ts
t
o
th
e
d
at
a
w
ith
s
p
e
c
if
i
c
s
t
an
d
a
r
d
d
ev
i
at
i
o
n
an
d
m
e
an
a
b
s
o
lu
t
e
v
a
lu
e
.
T
h
e
t
r
af
f
i
c
a
n
a
ly
s
i
s
an
d
p
r
e
d
i
ct
i
o
n
a
r
e
t
h
e
im
p
o
r
t
an
t
m
e
an
in
g
o
f
s
e
c
u
r
e
d
an
d
r
e
li
a
b
l
e
c
o
m
p
u
t
e
r
n
e
tw
o
r
k
.
V
a
r
i
o
u
s
t
ec
h
n
i
q
u
e
s
h
av
e
b
e
e
n
p
r
o
p
o
s
e
d
f
o
r
n
e
tw
o
r
k
c
o
n
g
es
t
i
o
n
a
n
a
ly
s
i
s
s
u
c
h
as
r
e
g
r
es
s
i
o
n
,
d
e
e
p
l
e
a
r
n
in
g
e
t
c
.
b
u
t
m
an
y
w
o
r
k
s
i
n
th
e
a
r
e
a
o
f
t
r
af
f
i
c
an
a
ly
s
is
a
n
d
f
o
r
e
c
as
t
a
r
e
p
r
e
s
en
te
d
i
n
[
4
]
.
T
h
e
p
a
p
e
r
i
n
v
es
t
ig
at
e
s
,
s
u
m
m
a
r
i
ze
s
an
d
r
e
v
i
e
w
s
m
u
lt
i
p
l
e
p
u
b
li
c
a
ti
o
n
s
.
A
l
th
o
u
g
h
tim
e
s
e
r
i
es
,
an
d
r
e
a
l
-
t
im
e
n
e
tw
o
r
k
t
r
af
f
ic
a
p
p
r
o
x
im
a
ti
o
n
is
o
b
l
ig
e
d
i
n
v
a
r
i
o
u
s
ty
p
es
o
f
n
e
t
w
o
r
k
s
o
f
tw
a
r
e
s
u
ch
a
s
t
r
af
f
ic
m
an
ag
em
en
t
an
d
r
e
s
o
u
r
c
es
a
l
l
o
c
a
ti
o
n
,
b
u
t th
e
s
e
a
r
ch
in
g
f
o
r
a
p
r
e
d
i
c
t
o
r
w
ith
h
ig
h
a
c
cu
r
a
cy
b
u
t
l
o
w
p
o
w
e
r
c
o
n
s
u
m
p
ti
o
n
h
as
b
e
en
o
u
t
l
in
e
d
in
[
5
]
.
Pr
e
d
i
c
t
o
r
s
i
n
c
lu
d
e
th
r
e
e
d
if
f
e
r
en
t
cl
a
s
s
es
s
u
ch
as
n
eu
r
a
l
n
e
tw
o
r
k
s
,
tim
e
s
e
r
i
es
,
an
d
w
av
el
e
t
t
r
an
s
f
o
r
m
-
b
as
e
d
p
r
e
d
i
c
t
o
r
s
.
T
h
e
an
a
ly
s
i
s
is
b
ase
d
o
n
r
ea
l
n
etw
o
r
k
in
d
i
c
at
o
r
s
.
T
h
e
c
o
m
p
u
ta
t
i
o
n
c
o
s
t
,
a
n
d
p
o
w
e
r
c
o
n
s
u
m
p
t
i
o
n
a
r
e
l
i
s
t
e
d
.
T
h
ey
f
o
ll
o
w
th
e
ex
p
o
n
e
n
ti
a
l
f
u
n
c
t
i
o
n
in
t
e
r
m
s
o
f
t
r
a
d
e
o
f
f
b
e
tw
ee
n
o
v
e
r
h
e
a
d
c
o
s
t
a
n
d
p
e
r
f
o
r
m
an
c
e
.
T
h
e
w
o
r
k
[
6
]
c
e
n
t
e
r
s
o
n
th
e
d
e
s
ig
n
,
t
h
e
e
s
tim
a
t
i
o
n
,
an
d
th
e
a
n
a
ly
s
is
o
f
th
e
b
eh
av
i
o
r
o
f
t
r
a
in
in
g
m
o
d
e
ls
f
o
r
f
o
r
e
c
as
t
in
g
t
h
e
th
r
o
u
g
h
p
u
t
o
f
a
s
in
g
l
e
ch
an
n
e
l
.
A
r
e
g
r
e
s
s
i
o
n
a
n
d
f
u
z
zy
m
o
d
e
ls
a
r
e
a
d
o
p
t
ed
f
o
r
th
e
e
s
tim
a
ti
o
n
.
B
a
s
e
d
o
n
r
e
al
n
e
t
w
o
r
k
ex
p
e
r
im
en
t
o
f
d
if
f
e
r
en
t
ch
an
n
e
ls
,
t
h
e
im
p
a
c
t
o
f
p
a
r
am
e
te
r
s
o
n
th
e
ap
p
r
o
x
im
a
t
i
o
n
e
r
r
o
r
is
d
i
s
cu
s
s
e
d
.
R
e
s
u
l
ts
d
em
o
n
s
t
r
a
te
t
h
at
t
r
a
in
i
n
g
m
o
d
e
ls
d
e
li
v
e
r
a
cc
u
r
a
te
th
r
o
u
g
h
p
u
t
es
t
im
at
i
o
n
.
T
h
e
I
n
te
r
n
et
t
r
af
f
i
c
a
p
p
r
o
x
im
a
t
i
o
n
w
ith
th
r
e
e
m
o
d
el
s
o
f
d
e
e
p
b
e
l
i
ef
n
etw
o
r
k
is
d
es
c
r
i
b
e
d
i
n
[
7
]
.
T
h
e
n
eu
r
al
n
e
t
w
o
r
k
w
ith
f
o
u
r
l
ay
e
r
s
d
e
p
t
h
in
e
a
ch
m
o
d
e
l
t
o
e
v
a
lu
at
e
th
e
n
o
n
-
l
in
e
a
r
an
d
t
im
e
s
e
r
i
e
s
o
f
I
n
t
e
r
n
et
t
r
a
f
f
i
c
is
d
e
v
e
l
o
p
e
d
.
T
h
e
d
e
ep
l
e
a
r
n
i
n
g
a
p
p
r
o
a
ch
is
u
s
e
d
w
ith
u
n
s
u
p
e
r
v
is
e
d
p
r
e
-
p
r
o
c
es
s
i
n
g
o
f
th
es
e
l
ay
e
r
s
.
T
h
e
m
e
th
o
d
a
ch
i
ev
es
es
tim
a
t
i
o
n
a
cc
u
r
a
cy
as
w
el
l
as
a
l
o
w
f
ig
u
r
e
o
f
r
o
o
t
-
m
e
an
-
s
q
u
a
r
e
d
e
r
r
o
r
o
n
g
iv
en
d
a
t
as
e
ts
.
W
a
n
g
et
a
l
.
[
8
]
p
r
e
s
e
n
t
a
d
e
ep
l
e
a
r
n
in
g
m
o
d
e
l
-
b
as
e
d
I
n
t
e
r
n
e
t
t
r
af
f
ic
p
r
e
d
i
ct
i
o
n
.
T
h
e
a
p
p
r
o
a
c
h
t
ak
e
s
in
te
g
r
al
c
o
r
r
e
l
a
t
i
o
n
s
an
d
t
r
af
f
i
c
f
l
o
w
d
at
a
in
t
o
a
c
c
o
u
n
t
.
D
e
-
n
o
i
s
i
n
g
an
d
e
n
c
o
d
in
g
m
o
d
el
is
em
p
l
o
y
e
d
t
o
o
b
s
e
r
v
e
I
n
t
e
r
n
e
t
t
r
a
f
f
i
c
ch
a
r
a
ct
e
r
is
t
i
cs
,
an
d
i
s
t
r
a
i
n
e
d
b
y
a
g
r
e
e
d
y
a
lg
o
r
i
th
m
.
T
h
e
e
s
tim
a
t
i
o
n
m
o
d
el
,
w
h
i
ch
is
a
p
a
r
t
o
f
th
e
t
r
af
f
i
c
s
ch
e
d
u
l
in
g
s
y
s
t
em
h
e
l
p
s
in
c
r
ea
s
e
th
e
b
an
d
w
i
d
th
u
t
il
i
z
at
i
o
n
o
f
I
n
t
e
r
n
e
t
n
e
tw
o
r
k
.
C
u
r
r
en
t
ly
,
c
o
m
p
u
t
e
r
n
e
tw
o
r
k
c
o
n
f
r
o
n
t
s
w
ith
a
g
ig
an
t
ic
t
r
af
f
i
c
d
em
an
d
t
o
h
an
d
l
e
u
p
t
o
th
e
s
t
an
d
a
r
d
q
u
al
i
ty
t
o
u
s
e
r
s
.
A
n
a
c
cu
r
a
te
n
etw
o
r
k
d
e
v
e
l
o
p
m
en
t
i
s
c
r
u
ci
a
l
t
o
s
u
s
t
ai
n
r
ev
en
u
e
s
b
y
s
h
o
r
t
en
e
d
p
r
o
f
i
t
f
r
o
m
a
b
an
d
w
i
d
th
u
s
ag
e
.
S
v
ig
e
l
j e
t
a
l
.
[
9
]
d
em
o
n
s
t
r
a
t
e
a
u
s
e
r
-
o
r
i
e
n
t
e
d
t
a
ct
i
c
t
o
c
o
m
p
u
t
er
n
e
t
w
o
r
k
an
d
m
o
d
e
l
a
n
e
tw
o
r
k
t
r
af
f
ic
t
o
o
p
t
im
i
z
e
a
n
ew
s
e
r
v
i
c
e
.
T
h
e
p
r
o
p
o
s
e
d
m
e
th
o
d
b
a
s
e
d
o
n
th
e
e
n
d
u
s
e
r
s
an
d
th
e
i
r
p
r
o
f
i
l
es
w
h
i
ch
c
an
b
e
g
a
th
e
r
e
d
f
r
o
m
r
ea
l
e
n
v
i
r
o
n
m
en
t
.
T
h
e
p
r
o
p
o
s
e
d
m
e
th
o
d
c
o
n
f
i
r
m
s
th
a
t
d
u
r
in
g
e
x
p
e
r
im
en
t
a
l
p
e
r
i
o
d
th
e
lo
a
d
d
i
f
f
e
r
s
l
es
s
th
an
5
%
f
r
o
m
th
e
r
e
al
f
ig
u
r
e
s
.
S
o
n
g
et
a
l
.
[
1
0
]
cl
a
s
s
if
y
h
o
w
a
s
s
o
r
t
e
d
n
o
is
e
in
f
lu
en
c
es
th
e
p
e
r
f
o
r
m
an
c
e
.
T
h
ey
a
p
p
ly
s
t
o
ch
as
t
i
c
g
r
a
d
i
en
t
d
es
c
en
t
(
SGD
)
f
o
r
g
iv
en
d
a
t
as
et
s
w
ith
n
o
is
e
a
n
d
d
e
t
e
r
m
in
e
th
a
t
it
d
e
p
e
n
d
s
o
n
t
h
e
le
a
r
n
in
g
r
at
e
v
a
lu
e
.
T
h
ey
th
en
p
r
o
p
o
s
e
a
m
e
ch
an
is
m
f
o
r
a
l
te
r
in
g
th
e
le
a
r
n
in
g
r
a
te
a
n
d
c
o
n
d
u
c
t
an
e
x
p
e
r
im
en
t
o
n
r
e
al
e
n
v
i
r
o
n
m
en
t
t
o
d
is
p
l
ay
th
at
th
e
p
r
o
p
o
s
e
d
m
e
th
o
d
is
h
e
a
l
th
i
e
r
t
h
an
a
p
p
ly
in
g
a
f
ix
e
d
l
ea
r
n
in
g
r
a
t
e
.
F
o
r
an
y
t
o
o
l
s
b
a
s
e
d
o
n
p
e
r
f
o
r
m
an
c
e
e
v
a
lu
a
t
io
n
,
th
e
p
a
p
e
r
[
1
1
]
s
u
g
g
es
ts
t
o
m
ak
e
u
s
e
o
f
SN
M
P
s
o
f
tw
ar
e
w
i
t
h
a
s
p
ec
if
i
e
d
p
o
l
lin
g
t
im
e
.
H
o
w
ev
e
r
,
s
am
p
li
n
g
t
o
o
f
r
e
q
u
en
t
c
au
s
es
n
e
tw
o
r
k
u
n
i
ts
t
o
b
e
o
v
e
r
l
o
a
d
e
d
an
d
e
s
ca
la
t
e
n
e
tw
o
r
k
t
r
af
f
ic
.
T
o
o
e
x
t
e
n
d
e
d
s
am
p
le
in
te
r
v
als
s
l
i
p
t
h
e
h
e
l
p
f
u
l
in
f
o
r
m
at
i
o
n
i
n
th
e
p
e
r
f
o
r
m
an
c
e
m
et
r
i
cs
.
T
h
ey
al
s
o
s
u
g
g
e
s
t
t
o
q
u
an
t
ity
th
e
p
e
r
f
o
r
m
a
n
ce
r
eg
a
r
d
i
n
g
u
s
e
r
’
s
n
e
t
w
o
r
k
s
c
o
p
e
.
C
o
m
p
u
t
e
r
n
etw
o
r
k
i
s
b
e
c
o
m
in
g
u
s
e
r
s
’
r
e
q
u
i
r
em
en
t
,
r
eg
a
r
d
in
g
s
o
f
t
w
a
r
e
an
d
u
s
e
r
s
ac
c
o
u
n
t
.
T
o
d
e
v
e
l
o
p
a
h
ig
h
-
q
u
al
i
ty
s
e
r
v
i
c
e
,
n
e
tw
o
r
k
a
n
a
ly
s
t
o
b
s
e
r
v
e
s
m
u
l
ti
p
l
e
as
p
e
c
ts
o
f
th
e
n
e
tw
o
r
k
t
r
af
f
i
c
l
i
k
e
c
h
an
n
el
t
r
af
f
i
c
am
o
u
n
t
.
W
h
en
n
e
t
w
o
r
k
’
s
s
i
z
e
g
r
o
w
s
,
th
e
m
o
n
it
o
r
i
n
g
t
as
k
b
e
co
m
es
c
r
u
c
i
al
.
T
h
e
p
r
o
b
l
em
o
f
u
s
in
g
n
e
tw
o
r
k
t
r
af
f
i
c
o
n
p
a
r
t
ic
u
l
a
r
ch
an
n
e
l
t
o
es
t
im
a
t
e
th
e
t
r
af
f
i
c
o
n
o
t
h
e
r
s
i
s
d
i
s
cu
s
s
e
d
i
n
[
1
2
]
.
A
l
th
o
u
g
h
th
e
a
p
p
r
o
a
c
h
c
o
s
ts
m
o
r
e,
b
u
t
th
i
s
m
eth
o
d
c
a
n
o
b
t
a
in
es
s
en
t
ia
l
in
f
o
r
m
a
ti
o
n
t
o
d
en
o
t
e
th
e
n
e
t
w
o
r
k
s
t
r
u
c
tu
r
e
,
a
n
d
c
an
h
e
l
p
r
a
i
s
e
t
h
e
r
e
s
u
l
t
s
o
f
e
s
t
im
at
i
o
n
.
A
n
o
v
e
l
t
e
ch
n
i
q
u
e
b
a
s
e
d
o
n
a
c
o
m
p
u
tin
g
n
o
d
e
r
e
s
e
r
v
in
g
th
e
c
o
m
m
u
n
i
c
at
i
o
n
h
is
t
o
r
y
is
p
r
e
s
en
te
d
i
n
[
1
3
]
.
A
u
th
o
r
s
e
v
a
lu
a
t
e
r
e
c
en
t
r
e
c
o
r
d
s
in
e
a
ch
n
o
d
e
w
h
i
le
d
is
r
eg
a
r
d
in
g
u
n
e
v
e
n
h
is
t
o
r
y
th
en
a
p
p
ly
i
n
g
w
ei
g
h
t t
o
c
o
n
s
i
s
t
en
t
r
e
c
o
r
d
s
.
T
h
e
t
e
ch
n
iq
u
e
p
e
r
s
e
p
r
e
v
en
ts
c
o
m
p
u
t
in
g
n
o
d
e
s
f
r
o
m
b
e
in
g
s
t
r
u
ck
b
y
in
c
o
r
r
e
c
t
an
d
h
i
d
d
e
n
i
n
f
o
r
m
a
ti
o
n
.
T
h
e
te
ch
n
i
q
u
e
r
e
d
u
c
es
an
e
r
r
o
r
o
f
n
e
t
w
o
r
k
p
r
e
d
ic
t
i
o
n
d
el
ay
a
b
o
u
t
5
0
%
u
s
in
g
th
e
ev
a
lu
at
i
o
n
b
a
s
ed
o
n
r
e
a
l
ex
p
e
r
im
e
n
t
a
l
d
at
a
.
A
n
o
t
h
e
r
la
t
en
cy
p
r
e
d
i
c
ti
o
n
i
s
p
r
o
p
o
s
e
d
b
y
[
1
4
]
.
T
h
e
a
p
p
r
o
a
c
h
f
o
c
u
s
e
s
o
n
s
t
a
ti
c
an
d
d
y
n
am
i
c
la
t
en
cy
p
r
e
d
i
c
ti
o
n
b
ase
d
o
n
t
im
e
s
e
r
ie
s
3
D
m
at
r
i
c
es
.
A
u
t
h
o
r
s
in
t
r
o
d
u
c
e
a
d
ec
o
m
p
o
s
i
ti
o
n
a
lg
o
r
i
th
m
t
o
d
iv
i
d
e
l
ate
n
cy
m
at
r
i
c
es
in
t
o
a
s
u
b
-
c
o
m
p
o
n
e
n
t
o
f
t
h
e
d
i
s
t
an
c
e
an
d
n
e
t
w
o
r
k
,
an
d
th
en
s
t
r
en
g
th
en
th
e
p
a
t
t
e
r
n
o
f
3
D
d
a
t
a
t
o
u
p
t
u
r
n
p
r
e
d
ic
t
i
o
n
a
cc
u
r
a
cy
.
E
x
p
e
r
im
en
t
s
ba
s
e
d
o
n
r
e
a
l
m
o
n
i
t
o
r
in
g
p
r
o
o
f
t
h
at
th
e
m
eth
o
d
o
u
t
p
e
r
f
o
r
m
s
t
r
ad
i
t
i
o
n
a
l
es
tim
a
t
i
o
n
a
p
p
r
o
a
ch
es
.
I
n
o
r
d
e
r
t
o
a
n
a
ly
z
e
m
o
b
i
l
e
d
a
t
a
t
r
af
f
i
c
,
th
e
a
lg
o
r
i
th
m
b
as
e
d
o
n
m
ac
h
in
e
le
a
r
n
in
g
is
im
p
l
em
en
t
e
d
[
1
5
]
.
T
h
e
d
a
t
a
-
d
r
i
v
en
ex
p
e
r
im
en
t
o
n
t
h
e
p
e
r
f
o
r
m
a
n
c
e
p
r
e
d
i
ct
i
o
n
p
o
i
n
t
s
o
u
t
th
at
a
M
a
r
k
o
v
p
r
e
d
i
cto
r
o
u
t
p
e
r
f
o
r
m
s
th
e
t
r
a
d
it
i
o
n
a
l
t
e
c
h
n
i
q
u
es
in
m
an
y
c
a
s
es
.
T
h
e
p
r
o
p
o
s
e
d
m
eth
o
d
a
ch
i
ev
es
t
h
e
a
c
cu
r
a
cy
o
f
7
0
%
,
w
h
i
c
h
o
u
t
p
e
r
f
o
r
m
s
al
l
ex
is
tin
g
o
n
es
an
d
im
p
r
o
v
e
s
1
%
t
o
5
%
b
y
m
a
ch
in
e
l
e
a
r
n
i
n
g
.
T
h
e
l
a
t
es
t
s
u
c
c
e
s
s
o
f
d
e
e
p
l
e
a
r
n
in
g
s
u
p
p
o
r
t
s
n
o
v
e
l
t
o
o
ls
th
a
t
ch
a
l
l
en
g
e
p
r
o
b
l
em
s
i
n
m
o
b
i
l
e
i
n
f
r
as
t
r
u
ct
u
r
e
.
T
o
s
e
a
l
th
e
g
a
p
b
e
tw
ee
n
d
e
e
p
le
a
r
n
in
g
a
n
d
m
o
b
i
l
e
n
e
t
w
o
r
k
,
a
s
u
r
v
ey
o
f
th
e
b
o
u
n
d
a
r
ie
s
is
p
r
e
s
en
t
e
d
in
[
1
6
]
.
A
u
th
o
r
s
f
am
i
l
i
a
r
iz
e
th
e
b
a
c
k
g
r
o
u
n
d
a
n
d
d
ee
p
l
e
a
r
n
in
g
a
p
p
r
o
a
c
h
e
s
w
ith
p
r
o
s
p
e
c
t
iv
e
s
o
f
t
w
a
r
e
.
T
h
ey
a
ls
o
d
e
li
b
e
r
a
t
e
v
a
r
i
o
u
s
t
ec
h
n
i
q
u
es
th
a
t
ea
s
e
th
e
ef
f
e
c
ti
v
e
p
l
a
c
em
en
t
o
f
d
e
e
p
l
ea
r
n
i
n
g
o
n
m
o
b
i
l
e
n
etw
o
r
k
s
.
A
n
in
-
d
e
p
th
an
a
ly
s
is
o
f
m
o
b
i
l
e
n
etw
o
r
k
in
g
r
e
s
e
a
r
ch
b
as
e
d
o
n
d
e
e
p
l
e
a
r
n
in
g
is
al
s
o
l
is
t
e
d
.
T
o
s
h
a
p
e
d
e
e
p
le
a
r
n
in
g
t
o
m
o
b
i
le
s
y
s
t
em
s
is
d
i
s
c
u
s
s
e
d
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6930
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
,
Vo
l.
18
,
No
.
4
,
A
u
g
u
s
t 2
0
2
0
:
1
8
0
2
-
1
8
0
8
1804
F
i
n
a
l
l
y
,
t
h
e
y
d
e
t
e
r
m
i
n
e
c
u
r
r
e
n
t
c
o
n
f
r
o
n
t
a
t
i
o
n
.
T
o
o
l
s
t
h
a
t
e
n
a
b
l
e
e
n
d
-
u
s
e
r
s
w
i
t
h
r
e
a
l
t
i
m
e
a
n
a
l
y
s
i
s
a
n
d
a
q
u
a
l
i
t
y
o
f
e
x
p
e
r
i
e
n
c
e
r
e
c
o
m
m
e
n
d
a
t
i
o
n
s
a
r
e
e
x
p
l
a
i
n
e
d
i
n
[
1
7
]
.
A
u
t
h
o
r
s
a
l
s
o
d
e
s
c
r
i
b
e
a
n
e
n
d
-
u
s
e
r
-
c
e
n
t
r
i
c
s
y
s
t
em
t
h
a
t
e
m
p
o
w
e
r
s
u
s
e
r
s
t
o
c
o
m
p
i
l
e
o
p
i
n
i
o
n
s
c
o
r
e
s
a
n
d
n
e
t
w
o
r
k
t
r
a
f
f
i
c
f
i
g
u
r
e
s
a
b
o
u
t
t
h
e
p
e
r
f
o
r
m
a
n
c
e
o
f
s
e
r
v
i
c
e
s
.
T
h
e
s
y
s
t
e
m
i
n
c
l
u
d
e
s
c
r
o
s
s
-
c
o
r
r
e
l
a
t
i
o
n
m
e
a
s
u
r
e
m
e
n
t
a
n
d
p
r
o
v
i
d
e
s
p
r
e
d
i
c
t
i
o
n
m
o
d
e
l
s
f
o
r
v
a
r
i
o
u
s
s
e
r
v
i
c
e
s
s
u
c
h
a
s
e
n
d
-
u
s
e
r
p
r
o
f
i
l
e
s
,
q
u
a
l
i
t
y
o
f
e
x
p
e
r
i
e
n
c
e
,
a
n
d
s
t
a
t
i
s
t
i
c
s
.
I
n
c
o
n
c
l
u
s
i
o
n
,
a
u
t
h
o
r
s
e
m
p
h
a
s
i
z
e
t
h
e
r
e
s
e
a
r
c
h
r
e
s
u
l
t
s
,
f
u
t
u
r
e
d
i
r
e
c
t
i
o
n
s
,
a
n
d
c
h
a
l
l
e
n
g
e
s
.
I
n
t
e
r
n
e
t
t
r
a
f
f
i
c
p
r
e
d
i
c
t
i
o
n
i
s
a
l
s
o
p
r
e
s
e
n
t
e
d
i
n
[
1
8
]
.
C
o
l
l
e
c
t
e
d
d
a
t
a
f
r
o
m
m
e
a
s
u
r
e
m
e
n
t
i
s
e
x
e
c
u
t
e
d
.
P
e
r
i
o
d
i
s
s
p
e
c
i
f
i
e
d
e
v
e
r
y
d
a
y
a
t
t
h
e
s
a
m
e
t
i
m
e
t
o
c
o
l
l
e
c
t
t
h
o
s
e
d
a
t
a
.
T
h
e
e
x
p
e
r
i
m
e
n
t
d
e
p
e
n
d
s
o
n
t
h
e
n
u
m
b
e
r
o
f
c
o
p
i
e
s
o
f
i
d
e
n
t
i
c
a
l
r
e
s
o
u
r
c
e
f
r
o
m
m
u
l
t
i
p
l
e
s
e
r
v
e
r
s
.
P
a
p
e
r
d
e
s
c
r
i
b
e
s
l
e
g
a
c
y
m
e
t
h
o
d
,
a
n
d
p
r
e
l
i
m
i
n
a
r
y
d
a
t
a
a
n
a
l
y
s
i
s
.
P
r
e
d
i
c
t
i
o
n
m
o
d
e
l
a
n
d
t
h
e
d
i
a
l
o
g
u
e
o
f
r
e
s
u
l
t
s
a
r
e
o
u
t
l
i
n
e
d
a
s
w
e
l
l
.
A
d
a
t
a
m
i
n
i
n
g
a
p
p
r
o
a
c
h
t
o
a
p
p
r
o
x
i
m
a
t
e
i
n
c
o
m
i
n
g
,
a
n
d
o
u
t
g
o
i
n
g
d
a
t
a
r
a
t
e
i
n
n
e
tw
o
r
k
i
n
g
e
n
v
i
r
o
n
m
e
n
t
b
a
s
e
d
o
n
a
s
s
o
c
i
a
t
i
o
n
r
u
l
e
i
s
p
r
e
s
e
n
t
e
d
i
n
[
1
9
]
.
T
h
e
m
e
n
t
i
o
n
e
d
d
a
t
a
r
a
t
e
a
n
d
b
a
n
d
w
i
d
t
h
a
r
e
p
e
r
f
o
r
m
a
n
c
e
m
e
t
r
i
c
s
w
h
i
c
h
c
a
n
h
e
l
p
s
o
r
t
o
u
t
n
e
tw
o
r
k
t
r
a
f
f
i
c
p
r
o
b
l
e
m
s
.
A
u
t
h
o
r
s
h
o
w
s
t
h
a
t
t
h
i
s
a
p
p
r
o
a
c
h
c
a
n
e
s
t
im
a
t
e
n
e
tw
o
r
k
t
r
a
f
f
i
c
a
n
d
c
a
n
c
a
l
c
u
l
a
t
e
d
a
t
a
c
o
n
g
e
s
t
i
o
n
a
n
d
l
o
s
s
.
M
o
r
e
o
v
e
r
,
t
h
e
m
e
t
h
o
d
i
s
a
p
p
l
i
c
a
b
l
e
f
o
r
r
o
u
t
i
n
g
a
l
l
o
c
a
t
i
o
n
a
n
d
p
e
r
f
o
r
m
a
n
c
e
u
p
g
r
a
d
e
.
A
p
p
r
o
x
i
m
a
t
i
n
g
t
h
e
p
e
r
f
o
r
m
a
n
c
e
o
f
n
e
t
w
o
r
k
t
r
a
f
f
i
c
d
u
r
i
n
g
a
s
p
e
c
i
f
i
e
d
t
i
m
e
w
i
t
h
a
p
o
s
s
i
b
i
l
i
t
y
e
m
p
o
w
e
r
s
i
m
p
r
o
v
e
d
r
o
u
t
i
n
g
s
e
l
e
c
t
i
o
n
,
a
n
d
d
a
t
a
t
r
a
n
s
f
e
r
t
h
a
t
i
s
c
r
i
t
i
c
a
l
i
n
b
i
g
d
a
t
a
e
n
v
i
r
o
n
m
e
n
t
s
.
A
p
e
r
f
o
r
m
a
n
c
e
a
p
p
r
o
x
im
a
t
i
o
n
m
o
d
e
l
b
a
s
e
d
o
n
t
im
e
s
e
r
i
e
s
m
e
t
h
o
d
,
t
o
i
n
c
r
e
a
s
e
t
h
e
e
f
f
i
c
i
e
n
c
y
o
f
r
e
s
o
u
r
c
e
s
u
t
i
l
i
z
a
t
i
o
n
,
s
c
h
e
d
u
l
i
n
g
,
a
n
d
d
a
t
a
m
a
n
a
g
e
m
e
n
t
i
s
p
r
o
p
o
s
e
d
i
n
[
2
0
]
.
A
u
t
h
o
r
s
c
r
e
a
t
e
a
d
j
u
s
t
m
e
n
t
p
r
o
c
e
s
s
e
s
f
o
r
i
d
e
n
t
i
f
y
i
n
g
t
h
e
p
a
t
t
e
r
n
s
,
p
e
r
i
o
d
,
d
i
a
g
n
o
s
i
s
,
a
n
d
a
d
ju
s
t
m
e
n
t
.
T
h
e
y
a
l
s
o
d
e
m
o
n
s
t
r
a
t
e
s
u
p
e
r
i
o
r
p
e
r
f
o
r
m
a
n
c
e
e
s
t
i
m
a
t
i
o
n
i
n
t
h
e
a
d
j
u
s
t
m
e
n
t
m
o
d
e
l
c
o
m
p
a
r
a
b
l
e
t
o
p
r
e
v
i
o
u
s
t
i
m
e
s
e
r
i
e
s
m
o
d
e
l
s
.
T
h
e
n
o
n
-
l
i
n
e
a
r
d
a
t
a
a
n
a
l
y
s
i
s
b
a
s
e
d
o
n
e
c
h
o
s
t
a
t
e
n
e
t
w
o
r
k
t
o
f
o
r
e
c
a
s
t
t
h
e
d
a
t
a
t
r
a
f
f
i
c
i
s
p
r
e
s
e
n
t
e
d
i
n
[
2
1
]
.
T
h
e
a
p
p
r
o
a
c
h
i
s
a
n
e
w
m
e
t
h
o
d
b
a
s
e
d
o
n
r
e
c
u
r
r
e
n
t
n
e
u
r
a
l
n
e
tw
o
r
k
s
,
w
h
e
r
e
a
r
e
s
e
r
v
o
i
r
i
s
i
s
s
u
e
d
a
n
d
a
w
e
i
g
h
t
m
a
t
r
i
x
i
s
u
s
e
d
.
A
b
e
t
t
e
r
e
c
h
o
s
t
a
t
e
n
e
t
w
o
r
k
c
o
n
c
e
r
n
i
n
g
w
e
a
k
n
e
s
s
o
n
t
h
e
a
c
t
i
v
i
t
y
f
u
n
c
t
i
o
n
a
n
d
t
h
e
s
e
l
e
c
t
i
o
n
o
f
m
a
t
r
i
x
i
s
p
r
o
p
o
s
e
d
.
T
h
e
n
e
t
w
o
r
k
o
u
t
l
i
n
e
s
t
h
e
s
c
o
p
e
o
f
m
a
t
r
i
x
a
n
d
s
u
b
s
t
i
t
u
t
e
s
t
h
e
a
c
t
i
v
i
t
y
i
n
t
h
e
m
i
d
d
l
e
l
a
y
e
r
o
f
w
a
v
e
l
e
t
f
u
n
c
t
i
o
n
.
R
es
u
l
ts
p
r
o
v
e
th
a
t
th
e
p
r
o
p
o
s
ed
n
e
t
w
o
r
k
i
s
v
a
l
i
d
c
o
m
p
a
r
e
d
t
o
t
r
a
d
i
t
i
o
n
al
o
n
e
.
T
o
d
e
s
ig
n
a
t
r
af
f
ic
-
a
w
a
r
e
n
e
t
w
o
r
k
i
s
im
p
o
r
t
an
t
t
o
m
o
d
el
t
h
e
t
r
af
f
i
c
p
r
e
d
i
ct
i
o
n
an
d
i
d
e
n
tif
y
th
e
t
r
af
f
i
c
.
T
h
e
e
n
t
r
o
p
y
t
h
e
o
r
y
t
o
i
d
en
tif
y
t
h
e
t
r
af
f
i
c
an
d
v
al
i
d
a
t
e
th
e
p
e
r
f
o
r
m
an
c
e
es
tim
a
t
i
o
n
i
s
p
r
es
en
t
ed
i
n
[
2
2
]
.
A
b
l
u
e
p
r
in
t
f
o
r
t
r
af
f
ic
-
b
as
e
d
s
o
f
t
w
a
r
e
t
o
p
r
o
v
i
d
e
th
e
f
o
r
e
c
as
t
d
a
ta
i
s
d
i
s
c
u
s
s
e
d
.
H
o
w
ev
e
r
,
th
e
a
b
o
v
e
-
m
e
n
ti
o
n
e
d
w
o
r
k
s
h
av
e
b
e
en
c
a
r
r
i
e
d
o
u
t
o
n
g
en
e
r
a
l
p
r
e
d
i
c
t
i
o
n
s
,
b
u
t
s
e
l
d
o
m
s
h
e
d
l
ig
h
t
o
n
th
e
r
eg
r
es
s
i
o
n
-
b
as
e
d
p
r
e
d
ic
t
i
o
n
f
o
r
c
o
m
p
u
t
e
r
n
etw
o
r
k
t
r
af
f
i
c
.
N
o
t
t
o
m
en
ti
o
n
,
w
ith
t
h
e
ex
p
l
o
s
iv
e
ly
i
n
c
r
e
a
s
in
g
d
em
an
d
f
o
r
I
n
t
e
r
n
et
a
cc
e
s
s
,
t
h
er
e
a
r
e
im
p
e
r
at
iv
e
n
ee
d
s
t
o
es
t
im
a
t
e
n
e
t
w
o
r
k
t
r
af
f
i
c
.
T
o
t
h
i
s
p
e
r
s
i
s
t
en
c
e
,
i
t
is
c
r
u
c
ia
l
t
o
m
in
im
i
z
e
p
e
r
c
en
t
ag
e
o
f
e
r
r
o
r
f
o
r
p
r
e
d
ic
t
i
o
n
.
T
h
e
r
es
t
o
f
th
e
p
a
p
e
r
is
s
t
r
u
ct
u
r
e
d
a
s
f
o
l
l
o
w
s
.
S
e
c
ti
o
n
2
o
f
f
e
r
s
t
r
a
d
i
t
i
o
n
a
l
a
p
p
r
o
x
im
a
t
i
o
n
m
e
th
o
d
s
a
s
w
e
ll
a
s
th
e
p
r
o
p
o
s
e
d
te
c
h
n
i
q
u
e
.
S
ec
t
i
o
n
3
f
o
cu
s
e
s
o
n
s
im
u
l
at
i
o
n
r
e
s
u
l
t
s
t
o
v
al
i
d
a
t
e
th
e
p
r
o
p
o
s
e
d
m
e
th
o
d
a
n
d
th
e
an
aly
s
is
.
T
h
e
c
o
n
c
l
u
s
i
o
n
an
d
f
u
tu
r
e
d
i
r
e
c
t
i
o
n
o
f
th
e
r
e
s
e
a
r
ch
a
r
e
d
is
c
u
s
s
e
d
i
n
s
e
ct
i
o
n
4
.
2.
E
S
T
I
M
AT
I
O
N
M
E
T
H
O
D
S
I
n
th
is
s
e
c
t
i
o
n
,
r
eg
r
es
s
i
o
n
-
b
ase
d
p
r
e
d
i
c
t
i
o
n
s
o
f
t
r
af
f
i
c
f
l
o
w
o
n
th
e
c
o
m
p
u
t
e
r
n
etw
o
r
k
a
r
e
d
i
s
cu
s
s
e
d
.
T
h
e
t
r
a
f
f
i
c
m
a
t
r
ix
,
h
o
l
d
in
g
f
l
o
w
f
ig
u
r
es
am
o
n
g
th
e
c
o
m
p
u
ti
n
g
n
o
d
e
s
in
c
o
m
p
u
te
r
n
etw
o
r
k
s
,
i
s
in
s
ig
h
tf
u
l
in
s
e
v
e
r
al
a
s
p
e
c
ts
o
f
n
etw
o
r
k
c
o
n
t
r
o
l
,
s
u
c
h
as
t
r
af
f
i
c
l
o
a
d
a
ll
o
c
a
ti
o
n
,
t
r
a
f
f
i
c
s
t
a
b
i
li
ty
[
2
3
]
,
l
o
a
d
b
al
an
ci
n
g
,
r
o
u
t
e
s
el
e
c
ti
o
n
,
t
r
a
f
f
i
c
s
h
a
p
in
g
,
o
u
t
l
in
in
g
,
a
n
d
c
a
p
a
c
i
ty
f
o
r
e
ca
s
t
in
g
.
T
o
o
p
e
r
a
t
e
a
c
o
m
p
u
te
r
n
e
tw
o
r
k
ef
f
i
c
i
en
t
ly
i
t
i
s
o
f
t
en
s
ig
n
if
i
c
an
t
t
o
q
u
an
tif
y
v
a
r
i
o
u
s
a
t
t
r
i
b
u
t
es
o
f
t
h
e
t
r
an
s
m
it
t
e
d
t
r
af
f
i
c
(
t
r
af
f
i
c
f
l
o
w
)
.
I
n
th
e
l
o
n
g
r
u
n
,
m
e
a
s
u
r
em
en
t
s
as
s
u
c
h
c
o
n
t
r
i
b
u
t
e
o
p
t
im
iz
a
t
i
o
n
an
d
n
e
tw
o
r
k
c
a
p
a
ci
ty
p
l
an
n
in
g
.
F
o
r
i
n
s
t
an
ce
,
a
l
l
o
c
at
i
o
n
o
f
n
etw
o
r
k
t
r
af
f
i
c
r
eg
a
r
d
in
g
s
o
u
r
ce
-
d
es
ti
n
a
t
i
o
n
p
a
i
r
w
is
e
le
a
d
s
a
c
o
n
f
ig
u
r
a
t
i
o
n
o
f
th
e
c
o
m
p
u
t
e
r
n
e
t
w
o
r
k
t
o
p
o
l
o
g
y
t
o
t
a
k
e
p
a
i
r
w
is
e
c
o
m
m
u
n
i
c
at
i
o
n
o
n
t
o
th
e
s
im
il
ar
p
a
t
h
.
A
s
tu
d
y
o
f
th
e
n
e
tw
o
r
k
t
r
a
f
f
i
c
b
e
tw
ee
n
an
o
v
e
r
l
o
a
d
in
g
s
er
v
e
r
,
an
d
it
s
c
l
i
en
ts
,
i
s
b
en
ef
i
c
i
al
t
o
a
p
p
o
r
t
i
o
n
d
e
s
ig
n
a
t
e
d
c
l
ie
n
t
s
t
o
an
in
c
r
em
en
ta
l
s
e
r
v
e
r
in
a
b
a
la
n
c
e
d
f
as
h
i
o
n
(
l
o
a
d
b
a
l
an
ci
n
g
)
.
I
n
t
h
e
s
h
o
r
t
o
p
e
r
a
t
i
o
n
,
m
e
asu
r
em
en
t
o
f
t
r
a
f
f
i
c
l
o
a
d
s
h
e
l
p
s
t
o
l
e
r
a
t
e
th
e
in
s
t
an
t
d
e
t
e
ct
i
o
n
o
f
m
is
b
e
h
a
v
i
o
r
s
.
F
o
r
i
n
s
t
an
c
e
,
a
s
en
d
e
r
a
n
n
o
u
n
ci
n
g
an
u
n
u
s
u
a
l
n
u
m
b
e
r
o
f
b
r
o
a
d
c
a
s
t
t
r
af
f
i
c
p
o
s
s
i
b
ly
m
al
f
u
n
ct
io
n
s
.
A
n
in
te
r
r
u
p
ti
o
n
o
f
t
r
af
f
ic
o
n
s
o
u
r
c
e
-
d
e
s
t
in
a
ti
o
n
p
a
i
r
w
is
e
d
e
s
i
g
n
at
es
an
in
c
o
n
s
is
t
en
t
u
s
ag
e
o
f
c
o
m
p
u
t
e
r
r
e
s
o
u
r
c
e
s
,
li
k
e
in
t
e
r
a
ct
iv
e
g
am
e
o
r
g
ig
an
t
i
c
f
i
l
e
t
r
an
s
f
e
r
,
w
h
ic
h
r
a
th
e
r
a
v
o
i
d
s
p
ea
k
p
e
r
i
o
d
s
.
I
n
o
r
d
e
r
t
o
ex
am
in
e
t
r
af
f
i
c
lo
a
d
s
a
c
cu
r
a
te
ly
it
is
t
o
m
e
as
u
r
e
a
l
l
p
a
c
k
e
t
s
.
H
o
w
ev
e
r
,
i
t
d
i
r
e
c
t
s
t
o
d
a
t
a
s
t
o
r
ag
e
c
o
m
p
l
ic
a
t
i
o
n
s
a
n
d
in
g
e
s
t
s
p
r
o
c
e
s
s
in
g
p
o
w
e
r
.
F
o
r
th
a
t
s
o
f
tw
a
r
e
,
it
n
e
e
d
s
t
o
in
s
p
e
c
t
p
a
ck
e
ts
u
s
i
n
g
s
t
a
ti
s
t
i
c
al
m
e
th
o
d
t
o
a
p
p
r
o
x
im
a
t
e
t
h
e
a
c
tu
a
l
c
o
u
n
ts
.
S
t
o
ch
a
s
t
i
c
s
am
p
l
in
g
i
s
p
e
r
h
a
p
s
p
r
ef
e
r
a
b
l
e
f
o
r
th
e
p
u
r
p
o
s
e
o
f
a
p
p
r
o
p
r
i
a
t
en
ess
an
d
i
s
d
es
i
r
a
b
l
e
o
n
g
r
o
u
n
d
s
o
f
a
c
cu
r
a
cy
as
a
c
o
m
p
l
et
e
p
a
c
k
e
t
s
am
p
li
n
g
is
t
e
d
i
o
u
s
,
an
d
y
e
t
l
e
a
d
s
th
e
t
r
af
f
i
c
a
t
p
e
ak
.
N
o
t
t
o
m
en
t
i
o
n
i
t
m
ay
ex
ce
e
d
t
h
e
ca
p
a
c
i
ty
o
f
t
h
e
r
es
o
u
r
c
e
s
t
o
k
e
e
p
u
p
w
i
th
,
th
en
p
a
ck
e
t
l
o
s
s
s
t
e
a
d
i
ly
o
c
cu
r
s
a
t
t
im
es
.
G
r
a
d
i
en
t
D
e
s
c
en
t
(
G
D
)
i
s
a
c
o
m
m
o
n
o
p
t
im
iz
a
ti
o
n
a
p
p
r
o
a
c
h
a
n
d
c
an
b
e
a
p
p
l
i
e
d
f
o
r
a
ll
l
e
a
r
n
in
g
a
l
g
o
r
ith
m
s
in
d
e
e
p
l
e
a
r
n
in
g
.
S
ta
t
i
s
t
i
c
al
ly
,
it
u
s
e
s
th
e
p
a
r
t
i
a
l
d
e
r
i
v
at
i
o
n
o
f
th
e
p
a
r
am
e
t
e
r
s
et
as
s
o
c
i
a
t
e
d
t
o
i
ts
i
n
p
u
ts
.
T
h
e
m
o
r
e
th
e
g
r
a
d
i
en
t
(
s
l
o
p
e
)
i
s
,
t
h
e
h
ig
h
e
r
th
e
s
l
o
p
e
w
i
ll
b
e
.
GD
i
s
a
b
u
lg
in
g
f
u
n
c
t
i
o
n
an
d
an
i
te
r
a
t
iv
e
c
o
m
p
u
t
at
i
o
n
t
o
d
e
t
e
r
m
in
e
t
h
e
f
ig
u
r
e
s
o
f
th
e
in
p
u
t
p
a
r
am
e
t
e
r
f
u
n
ct
i
o
n
th
a
t
m
in
im
iz
e
s
th
e
l
o
s
s
f
u
n
ct
i
o
n
as
f
a
r
as
p
o
s
s
i
b
le
.
T
h
e
p
a
r
am
e
te
r
s
a
r
e
p
r
im
a
r
ily
cl
a
r
if
i
e
d
as
a
s
p
e
c
if
i
c
v
a
lu
e
th
en
G
D
h
as
b
e
en
e
x
e
cu
t
e
d
in
m
u
l
ti
p
l
e
r
e
p
e
ti
t
i
o
n
s
t
o
d
e
t
e
r
m
in
e
th
e
o
p
t
im
a
l
f
ig
u
r
e
o
f
t
h
e
in
p
u
t
p
a
r
am
et
e
r
s
,
b
a
s
e
d
o
n
f
u
n
d
am
en
t
al
c
al
cu
lu
s
,
t
o
ex
p
ec
t
th
e
l
o
w
es
t
f
ig
u
r
e
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
A
p
p
r
o
xima
tio
n
o
f
r
eg
r
ess
io
n
-
b
a
s
ed
fa
u
lt min
imiz
a
tio
n
fo
r
n
etw
o
r
k
tr
a
ffic
(
C
h
a
n
in
to
r
n
Jitta
w
ir
iya
n
u
ko
o
n
)
1805
o
f
th
e
s
p
e
c
if
i
e
d
l
o
s
s
f
u
n
c
ti
o
n
.
I
n
th
is
p
a
p
e
r
,
tw
o
ty
p
es
o
f
G
D
al
g
o
r
it
h
m
a
r
e
c
o
n
c
e
r
n
e
d
,
s
t
o
ch
as
t
i
c
g
r
a
d
i
en
t
d
es
c
en
t
(
S
G
D
)
an
d
g
r
a
d
i
en
t
d
es
c
en
t
alg
o
r
i
th
m
(
GDA
)
.
2
.
1
.
St
o
cha
s
t
ic
g
ra
dient
des
ce
nt
(
SG
D)
I
t
is
an
al
g
o
r
ith
m
t
h
at
is
a
s
s
o
ciate
d
w
it
h
a
s
to
c
h
ast
ic
(
r
an
d
o
m
)
p
r
o
b
ab
ilit
y
.
I
n
S
GD
alg
o
r
ith
m
[
2
4
]
,
s
o
m
e
s
a
m
p
les
ar
e
ch
o
s
en
r
a
n
d
o
m
l
y
in
s
tead
o
f
t
h
e
co
m
p
lete
s
a
m
p
le
s
f
r
o
m
d
ata
s
et
f
o
r
ea
ch
r
ep
etitio
n
.
I
n
th
e
class
ical
GD
alg
o
r
it
h
m
,
th
e
co
m
p
lete
d
ataset
is
tak
e
n
in
to
ac
co
u
n
t.
A
lt
h
o
u
g
h
,
u
s
in
g
th
e
en
tire
d
ataset
is
ac
tu
all
y
m
ea
n
in
g
f
u
l
f
o
r
r
ea
ch
in
g
th
e
m
i
n
i
m
a
in
a
n
o
i
s
y
p
atter
n
(
en
v
ir
o
n
m
en
t)
,
b
u
t
t
h
e
p
r
o
b
lem
a
s
ce
n
d
s
as d
ataset
s
ize
g
r
o
w
s
.
I
n
ca
s
e
o
f
h
u
g
e
d
ataset,
if
class
ica
l G
D
ap
p
r
o
ac
h
is
ap
p
lied
th
en
it m
u
s
t ta
k
e
all
s
a
m
p
le
s
to
co
m
p
lete
o
n
e
r
o
u
n
d
o
f
r
e
p
etitio
n
.
A
p
p
ar
en
tl
y
,
th
e
co
m
p
u
tatio
n
co
s
t
is
o
n
t
h
e
r
i
s
e.
T
h
e
ch
ao
s
en
d
s
b
y
e
m
p
lo
y
i
n
g
SDG.
I
n
t
h
is
r
e
g
ar
d
,
it o
n
l
y
p
er
f
o
r
m
s
s
elec
ti
v
e
a
n
d
s
h
u
f
f
le
s
a
m
p
les
f
o
r
ea
ch
r
ep
etitio
n
.
2
.
2
.
G
ra
dient
des
ce
nt
a
lg
o
rit
h
m
(
G
DA)
Op
ti
m
izi
n
g
v
ar
iab
les
is
th
e
o
b
j
ec
tiv
e
o
f
all
d
ee
p
lear
n
in
g
alg
o
r
ith
m
s
.
T
h
e
o
p
tim
u
m
f
i
g
u
r
e
o
f
th
e
g
r
ad
ien
t (
s
lo
p
e)
an
d
th
e
in
t
er
ce
p
t a
r
e
co
m
p
u
ted
to
g
et
th
e
b
est ap
p
r
o
x
i
m
atio
n
in
li
n
ea
r
r
eg
r
ess
io
n
f
u
n
ct
io
n
.
T
h
e
o
b
j
ec
tiv
e
f
u
n
ct
io
n
f
o
r
m
u
lates
i
n
p
u
t
p
ar
a
m
eter
s
o
n
to
o
u
tp
u
t
v
al
u
es.
T
h
is
ap
p
lies
f
o
r
all
r
eg
r
ess
io
n
p
r
ed
ictio
n
s
.
Dee
p
lear
n
in
g
alg
o
r
ith
m
h
a
s
co
ef
f
icien
ts
th
at
d
e
p
ict
th
e
alg
o
r
ith
m
ap
p
r
o
x
i
m
ati
o
n
f
o
r
th
e
o
b
j
ec
tiv
e
f
u
n
ctio
n
.
I
n
d
iv
id
u
a
l
alg
o
r
ith
m
h
as
d
if
f
er
en
t
co
e
f
f
icie
n
t
h
o
w
e
v
er
,
an
o
p
tim
izatio
n
to
d
eter
m
i
n
e
th
e
s
et
o
f
co
ef
f
icie
n
t
s
,
th
at
r
ep
r
esen
t
th
e
b
est
ap
p
r
o
x
i
m
atio
n
is
ex
p
e
cted
.
Op
tim
izat
io
n
b
ased
o
n
GD
ca
n
ap
p
ly
f
o
r
an
alg
o
r
ith
m
w
it
h
th
e
s
et
o
f
co
ef
f
icien
ts
,
s
u
c
h
as lo
g
i
s
tic
o
r
li
n
ea
r
r
eg
r
ess
io
n
p
r
o
b
le
m
s
.
T
h
e
esti
m
atio
n
o
f
h
o
w
w
ell
a
lear
n
i
n
g
m
o
d
el
f
its
t
h
e
o
b
j
ec
tiv
e
f
u
n
c
tio
n
ca
n
b
e
co
m
p
u
ted
t
h
r
o
u
g
h
d
i
f
f
er
e
n
t
s
tep
s
.
T
h
e
lo
s
s
f
u
n
c
tio
n
r
ep
r
esen
ts
h
o
w
f
it
t
h
e
m
o
d
el
is
p
e
r
f
o
r
m
in
g
o
n
t
h
e
tr
ain
i
n
g
d
ataset.
I
f
th
e
lo
s
s
is
e
x
tr
ao
r
d
in
ar
y
,
th
e
ap
p
r
o
x
i
m
atio
n
is
d
iv
er
g
i
n
g
to
o
f
ar
f
r
o
m
th
e
tar
g
et
d
ata.
T
h
u
s
,
in
d
ee
p
lear
n
in
g
alg
o
r
it
h
m
,
t
h
e
d
ec
is
iv
e
g
o
al
is
to
m
in
i
m
iz
e
th
e
lo
s
s
lin
k
ed
w
it
h
t
h
e
lear
n
i
n
g
m
o
d
el.
T
h
e
lo
s
s
f
u
n
ctio
n
i
n
cl
u
d
e
s
w
ei
g
h
i
n
g
th
e
co
e
f
f
ici
en
ts
i
n
t
h
e
lear
n
i
n
g
m
o
d
el
b
y
co
m
p
u
ti
n
g
a
n
ap
p
r
o
x
i
m
at
io
n
f
o
r
ea
ch
tr
ain
i
n
g
d
ataset,
an
d
t
h
e
n
as
s
o
ciatin
g
th
e
e
s
ti
m
atio
n
s
to
th
e
o
u
tp
u
t
v
ar
iab
les
to
d
eter
m
i
n
e
a
n
a
v
er
ag
e
er
r
o
r
.
T
h
e
lo
s
s
is
co
m
p
u
ted
f
o
r
t
h
e
a
lg
o
r
ith
m
v
ia
t
h
e
tr
ain
i
ng
d
ataset
f
o
r
ea
ch
s
tep
o
f
t
h
e
G
DA
[
2
5
]
,
w
h
ich
i
s
a
p
o
p
u
lar
alg
o
r
ith
m
i
n
d
ee
p
lear
n
in
g
m
et
h
o
d
.
2
.
3
.
P
r
o
po
s
ed
m
et
ho
d
C
u
r
r
en
t
lear
n
i
n
g
m
o
d
els
i
n
cl
u
d
e
d
atasets
w
ith
attr
ib
u
tes
h
o
ld
in
g
v
ar
io
u
s
p
ar
a
m
eter
s
.
As
s
u
c
h
it
is
p
r
o
b
lem
atic
to
d
eter
m
i
n
e
th
e
c
o
n
v
e
x
it
y
o
f
th
e
lo
s
s
f
u
n
ctio
n
f
o
r
th
e
in
v
esti
g
atio
n
.
A
n
o
v
el
d
is
p
er
s
ed
alg
o
r
ith
m
f
o
r
GD
i
n
t
h
e
in
ter
p
o
lati
n
g
b
o
u
n
d
i
s
p
r
esen
ted
in
[
2
6
]
.
T
h
e
alg
o
r
it
h
m
ass
o
ciate
s
to
a
s
i
m
p
le
d
is
tr
ib
u
ted
lo
s
s
f
u
n
ctio
n
.
Ho
w
ev
er
,
t
h
e
u
s
e
o
f
GD
i
s
d
r
iv
e
n
b
y
th
e
co
m
p
u
tatio
n
co
s
t
o
f
e
x
ec
u
ti
n
g
b
ac
k
p
r
o
p
ag
atio
n
v
ia
th
e
w
h
o
le
tr
ai
n
in
g
s
et
s
.
T
h
e
co
s
t
lead
s
to
clu
m
s
y
co
n
v
er
g
en
ce
.
I
n
th
i
s
p
ap
er
f
ast
co
n
v
e
r
g
en
ce
is
p
r
o
p
o
s
ed
th
r
o
u
g
h
d
atasets
s
ta
n
d
ar
d
izatio
n
.
Stan
d
ar
d
izin
g
d
ata
ch
ar
ac
ter
is
tics
ab
o
u
t
ze
r
o
,
an
d
th
e
m
id
p
o
in
t
w
it
h
a
s
tan
d
ar
d
d
ev
ia
tio
n
o
f
o
n
e
is
u
s
ed
w
h
e
n
m
ea
s
u
r
e
m
e
n
ts
h
a
v
e
d
is
s
i
m
ilar
u
n
its
.
P
ar
am
eter
s
th
at
ar
e
co
llected
at
d
iv
er
s
e
s
ca
les d
o
n
o
t a
cc
o
r
d
s
i
m
i
lar
l
y
to
th
e
i
n
v
esti
g
atio
n
an
d
lead
to
an
u
n
f
air
n
ess
.
T
h
e
p
r
o
p
o
s
ed
alg
o
r
ith
m
r
ev
i
s
es
th
e
p
r
esen
t
f
ig
u
r
e
b
ased
o
n
th
e
m
o
v
i
n
g
d
ir
ec
tio
n
Δ
t
b
y
s
t
ep
v
alu
e
α
t
to
th
e
r
ec
u
r
s
iv
e
s
o
lu
t
io
n
+
1
.
T
h
en
th
e
r
ec
u
r
s
i
v
e
ca
lc
u
latio
n
o
f
s
lo
p
e
ca
n
b
e
s
h
o
r
ten
ed
i
n
(
1
)
s
h
o
w
n
i
n
th
e
b
elo
w
iter
at
io
n
.
+
1
=
−
α
t
∇
(
)
(
1
)
T
h
e
g
o
al
is
to
m
in
i
m
ize
er
r
o
r
in
esti
m
atio
n
o
f
t
h
e
f
u
n
ctio
n
F,
a
h
u
m
b
le
tech
n
iq
u
e
to
s
ele
ct
th
e
s
tep
v
a
lu
e
i
n
f
as
h
io
n
t
h
at
m
i
n
i
m
izes
t
h
e
f
i
g
u
r
e
o
f
f
u
t
u
r
e
d
ir
ec
tio
n
,
i.e
.
co
m
p
u
te
th
e
v
a
lu
e
t
h
at
w
ill
b
e
ab
le
to
m
i
n
i
m
ize
(
+
1
)
.
T
h
u
s
,
th
e
s
tep
v
a
lu
e
i
s
ca
lcu
l
ated
b
y
(
2
)
.
α
t+
1
=
ar
g
m
i
n
α
≥
0
(
−
α
t
∇
(
)
)
(
2
)
L
et
F:
C
→
K
d
e
n
o
te
a
co
n
v
e
x
f
u
n
ctio
n
w
i
th
p
ar
a
m
eter
s
q
,
Q
an
d
th
e
s
tep
s
ize
f
o
r
r
ec
u
r
s
iv
e
an
al
y
s
is
a
lo
n
g
d
ir
ec
tio
n
d
is
co
m
p
u
ted
b
y
th
e
th
ir
d
o
r
d
er
ex
p
an
s
io
n
o
f
T
ay
l
o
r
f
u
n
ct
io
n
as
s
h
o
w
n
(
3
)
.
(
)
−
1
2
∇
,
2
(
)
−
1
6
∇
,
,
3
(
)
≤
∗
≤
(
)
−
1
2
∇
,
2
(
)
−
1
6
∇
,
,
3
(
)
(
3
)
f
o
r
a
r
ec
u
r
s
iv
e
s
tep
r
an
d
ξ >
0
,
g
iv
e
n
th
a
t
(
)
−
→
K
(
)
≤
ξ th
en
w
e
h
av
e
;
r
≥
lo
g
(
(
0
)
−
ξ
)
/ lo
g
(
1
1
−
)
.
(
4
)
T
h
e
p
r
o
p
o
s
ed
alg
o
r
ith
m
is
s
i
m
p
lif
ied
as
s
h
o
w
n
i
n
Fi
g
u
r
e
1
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6930
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
,
Vo
l.
18
,
No
.
4
,
A
u
g
u
s
t 2
0
2
0
:
1
8
0
2
-
1
8
0
8
1806
Fig
u
r
e
1
.
P
r
o
p
o
s
ed
alg
o
r
ith
m
3.
RE
SU
L
T
S
A
ND
AN
AL
Y
SI
S
Fo
r
th
e
ex
p
er
i
m
en
t
to
v
alid
ate
h
o
w
to
m
i
n
i
m
ize
er
r
o
r
f
o
r
af
o
r
em
e
n
tio
n
ed
alg
o
r
ith
m
s
,
t
h
r
ee
d
ataset
s
th
at
r
ep
r
esen
t
co
m
p
u
ter
n
e
t
w
o
r
k
co
m
p
o
n
e
n
ts
a
n
d
t
h
e
s
t
r
u
ctu
r
e
ar
e
s
i
m
u
la
ted
.
T
h
e
o
b
j
ec
tiv
e
is
to
f
in
d
th
e
r
eg
r
es
s
io
n
-
b
ased
esti
m
atio
n
f
o
r
n
et
w
o
r
k
tr
af
f
ic
d
at
a.
T
ab
les
1
-
1
0
d
ep
ict
s
i
m
u
latio
n
r
esu
lts
.
A
ll
d
ataset
s
ar
e
i
m
b
alan
ce
d
ch
ar
ac
ter
is
tic
r
eg
ar
d
in
g
class
e
s
an
d
attr
ib
u
tes
f
u
n
ctio
n
.
W
e
s
i
m
u
late
t
h
e
lo
g
is
t
ic
r
eg
r
ess
io
n
to
m
i
n
i
m
ize
th
e
lo
s
s
f
u
n
ctio
n
.
S
GD
is
n
o
t
a
co
m
p
etiti
v
e
alg
o
r
ith
m
f
o
r
co
m
p
le
x
d
atasets
.
H
o
w
e
v
er
,
w
e
o
b
s
er
v
e
th
at
t
h
e
ass
o
ciatio
n
b
et
w
ee
n
t
h
e
GDA
a
n
d
th
e
p
r
o
p
o
s
ed
alg
o
r
ith
m
is
s
i
m
ilar
to
ac
h
iev
e
t
h
e
o
p
tim
a
l v
al
u
es.
T
ab
le
1
.
T
h
e
d
ataset#
1
o
u
tco
m
e
s
eq
u
e
n
ce
o
f
lo
s
s
f
u
n
ctio
n
w
it
h
s
tep
s
ize
=
1
0
T
ab
le
2
.
T
h
e
d
ataset#
2
o
u
tco
m
e
s
eq
u
e
n
ce
o
f
lo
s
s
f
u
n
ctio
n
w
it
h
s
tep
s
ize
=
1
0
S
G
D
G
D
A
P
r
o
p
o
se
d
1
.
0
6
1
.
0
6
1
.
6
1
0
.
1
4
0
.
1
5
0
.
1
5
0
.
1
5
0
.
1
4
0
.
1
4
0
.
1
7
0
.
1
4
0
.
1
9
0
.
1
7
0
.
2
O
p
t
i
mal
V
a
l
u
e
(
5
.
0
4
,
-
4
.
0
6
)
(
4
.
6
9
,
-
4
.
1
4
)
(
4
.
7
3
,
-
4
.
1
7
)
S
G
D
G
D
A
P
r
o
p
o
se
d
2
.
3
1
2
.
3
1
0
.
3
4
0
.
8
5
1
.
1
0
.
2
2
0
.
8
2
0
.
6
9
0
.
1
6
0
.
8
0
.
5
0
.
1
5
0
.
4
8
0
.
3
6
0
.
1
4
0
.
3
5
0
.
2
6
0
.
3
7
0
.
2
0
.
3
4
0
.
1
8
0
.
1
7
0
.
1
7
0
.
1
6
0
.
1
6
0
.
1
5
0
.
1
5
O
p
t
i
mal
V
a
l
u
e
(
7
.
5
5
,
-
6
.
1
7
)
(
7
.
3
7
,
-
6
.
6
1
)
(
5
.
4
5
,
-
4
.
7
9
)
T
ab
le
3
.
T
h
e
d
ataset#
3
o
u
tco
m
e
s
eq
u
e
n
ce
o
f
lo
s
s
f
u
n
ctio
n
w
it
h
s
tep
s
ize
=
1
0
T
ab
le
4
.
T
h
e
d
ataset#
4
o
u
tco
m
e
s
eq
u
e
n
ce
o
f
lo
s
s
f
u
n
ctio
n
w
it
h
s
tep
s
ize
=
1
0
S
G
D
G
D
A
P
r
o
p
o
se
d
5
.
1
5
5
.
1
5
0
.
4
4
0
.
2
9
0
.
2
9
0
.
3
4
0
.
2
5
0
.
2
8
0
.
2
6
0
.
2
2
0
.
2
7
0
.
2
3
0
.
2
1
0
.
2
6
0
.
2
0
.
2
0
.
2
5
0
.
1
9
0
.
1
9
0
.
2
3
0
.
1
8
0
.
1
7
0
.
2
2
0
.
1
7
0
.
1
9
0
.
2
1
0
.
1
8
0
.
1
9
0
.
1
7
0
.
1
8
0
.
1
7
O
p
t
i
mal
V
a
l
u
e
(
N
/
A
)
(
4
.
0
1
,
1
.
3
6
)
(
4
,
1
.
3
7
)
S
G
D
G
D
A
P
r
o
p
o
se
d
0
.
3
7
0
.
3
7
0
.
3
9
0
.
3
3
0
.
3
5
0
.
3
7
0
.
3
2
0
.
3
3
0
.
3
5
0
.
3
0
.
3
0
.
3
2
0
.
2
9
0
.
2
8
0
.
3
1
0
.
2
7
0
.
2
6
0
.
2
7
0
.
2
6
0
.
2
4
0
.
2
5
0
.
2
5
0
.
2
3
0
.
2
3
0
.
2
3
0
.
2
1
0
.
2
1
0
.
2
4
0
.
1
9
0
.
1
9
0
.
2
2
0
.
1
8
0
.
1
8
0
.
1
9
0
.
2
0
.
1
8
O
p
t
i
mal
V
a
l
u
e
(
N
/
A
)
(
4
.
8
7
,
1
.
0
1
)
(
4
.
8
2
,
1
.
0
1
)
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
A
p
p
r
o
xima
tio
n
o
f
r
eg
r
ess
io
n
-
b
a
s
ed
fa
u
lt min
imiz
a
tio
n
fo
r
n
etw
o
r
k
tr
a
ffic
(
C
h
a
n
in
to
r
n
Jitta
w
ir
iya
n
u
ko
o
n
)
1807
T
ab
le
5
.
T
h
e
d
ataset#
5
o
u
tco
m
e
s
eq
u
e
n
ce
o
f
lo
s
s
f
u
n
ctio
n
w
it
h
s
tep
s
ize
=
1
0
T
ab
le
6
.
T
h
e
d
ataset#
6
o
u
tco
m
e
s
eq
u
e
n
ce
o
f
lo
s
s
f
u
n
ctio
n
w
it
h
s
tep
s
ize
=
1
0
S
G
D
G
D
A
P
r
o
p
o
se
d
1
.
2
5
1
.
2
5
2
.
1
9
5
.
0
9
1
.
0
1
1
.
2
8
5
.
0
3
0
.
8
1
0
.
7
8
4
.
6
9
0
.
7
7
0
.
7
5
2
.
7
1
0
.
7
4
0
.
7
3
1
.
7
7
2
.
0
1
1
.
1
1
1
.
6
2
1
.
7
3
0
.
8
7
O
p
t
i
mal
V
a
l
u
e
(
N
/
A
)
(
N
/
A
)
(
N
/
A
)
S
G
D
G
D
A
P
r
o
p
o
se
d
2
2
1
.
3
9
1
.
9
7
1
.
7
5
1
1
.
9
2
1
.
2
8
0
.
8
3
1
.
5
4
0
.
9
3
0
.
7
4
1
.
0
7
0
.
8
0
.
7
1
2
.
3
6
0
.
7
5
1
.
9
0
.
7
2
1
.
1
6
0
.
7
3
0
.
7
5
0
.
7
5
0
.
8
2
0
.
7
3
1
0
.
8
0
.
7
6
0
.
8
2
0
.
8
4
O
p
t
i
mal
V
a
l
u
e
(
N
/
A
)
(
N
/
A
)
(
0
.
4
8
,
-
0
.
2
)
T
ab
le
7
.
T
h
e
d
ataset#
7
o
u
tco
m
e
s
eq
u
e
n
ce
o
f
lo
s
s
f
u
n
ctio
n
w
it
h
s
tep
s
ize
=
1
0
T
ab
le
8
.
T
h
e
d
ataset#
8
o
u
tco
m
e
s
eq
u
e
n
ce
o
f
lo
s
s
f
u
n
ctio
n
w
it
h
s
tep
s
ize
=
1
0
S
G
D
G
D
A
P
r
o
p
o
se
d
3
.
1
5
3
.
1
5
5
.
9
2
2
.
8
7
1
.
8
0
1
.
1
2
.
8
5
0
.
8
8
0
.
4
6
3
.
1
6
0
.
7
0
.
3
8
2
.
2
9
0
.
8
1
0
.
3
2
2
.
0
3
0
.
9
8
0
.
3
0
3
.
9
1
1
.
1
5
2
.
2
1
.
3
8
1
.
1
2
1
.
4
1
O
p
t
i
mal
V
a
l
u
e
(
N
/
A
)
(
N
/
A
)
(
-
2
.
3
,
0
.
4
)
S
G
D
G
D
A
P
r
o
p
o
se
d
4
.
5
2
4
.
5
2
5
.
3
7
3
.
1
7
2
.
7
1
2
.
0
4
2
.
4
1
.
0
7
1
.
2
5
1
.
2
4
1
.
0
8
0
.
8
5
0
.
7
7
1
.
5
2
1
.
0
2
0
.
9
7
0
.
7
6
0
.
9
6
1
.
6
3
0
.
9
6
0
.
8
1
.
0
2
3
.
7
7
1
.
6
5
O
p
t
i
mal
V
a
l
u
e
(
N
/
A
)
(
N
/
A
)
(
N
/
A
)
T
ab
le
9
.
T
h
e
d
ataset#
9
o
u
tco
m
e
s
eq
u
e
n
ce
o
f
lo
s
s
f
u
n
ctio
n
w
it
h
s
tep
s
ize
=
1
0
T
ab
le
10
.
T
h
e
d
ataset#
10
o
u
tco
m
e
s
eq
u
e
n
ce
o
f
lo
s
s
f
u
n
ctio
n
w
it
h
s
tep
s
ize
=
1
0
S
G
D
G
D
A
P
r
o
p
o
se
d
4
.
4
4
4
.
4
4
3
.
7
5
3
.
1
2
2
.
6
5
2
.
6
5
2
.
2
8
1
.
0
6
1
.
5
1
0
.
7
4
0
.
8
6
0
.
7
1
.
4
3
2
1
.
4
5
2
.
3
O
p
t
i
mal
V
a
l
u
e
(
N
/
A
)
(
-
0
.
6
,
-
1
.
2
)
(
0
.
2
5
,
0
.
0
5
)
S
G
D
G
D
A
P
r
o
p
o
se
d
5
.
1
6
5
.
1
6
3
.
1
3
4
.
1
2
3
.
5
5
2
2
.
5
2
1
.
8
3
0
.
9
0
.
9
0
.
6
9
0
.
7
8
1
.
0
4
0
.
7
2
0
.
9
9
1
.
7
9
0
.
8
3
O
p
t
i
mal
V
a
l
u
e
(
N
/
A
)
(
-
0
.
0
2
,
0
.
0
9
)
(
-
0
.
1
3
,
-
0
.
9
6
)
No
t
to
m
e
n
tio
n
,
th
e
p
r
o
p
o
s
ed
alg
o
r
ith
m
p
r
o
v
id
es
a
f
ast
co
n
v
er
g
e
n
ce
to
o
p
tim
izatio
n
.
As
t
h
e
s
i
m
u
lat
io
n
s
i
m
p
licate
d
is
p
er
s
ed
d
ata,
th
e
in
itialized
s
t
ep
s
ize
is
cr
itical,
a
n
d
w
e
ca
n
n
o
t
an
ticip
ate
t
h
ese
alg
o
r
ith
m
s
to
b
e
m
o
d
est f
o
r
all
d
atasets
.
T
h
e
lo
s
s
f
u
n
ctio
n
tr
aj
ec
to
r
ies
an
d
th
eir
o
p
tim
al
v
al
u
es
(
i
f
p
o
s
itiv
e)
ar
e
tab
u
lated
.
R
es
u
lt
s
co
n
f
ir
m
th
at
th
e
p
r
o
p
o
s
ed
m
e
th
o
d
ali
g
n
s
w
i
th
th
e
G
D
A
f
o
r
all
d
atasets
alt
h
o
u
g
h
it
i
s
in
f
lex
ib
le
to
f
in
d
t
h
e
o
p
ti
m
izat
io
n
i
n
t
h
e
e
x
p
er
i
m
e
n
ts
as
lis
ted
in
T
ab
les
5
a
n
d
8
.
Nev
er
t
h
eles
s
,
w
e
ac
h
ie
v
e
f
in
a
l
r
esu
lt
s
m
a
n
y
m
o
r
e
th
a
n
t
h
e
G
DA
.
4.
CO
NCLU
SI
O
N
T
h
e
p
r
o
p
o
s
ed
alg
o
r
ith
m
is
c
o
m
p
r
e
h
en
s
iv
e
l
y
v
a
lid
ated
v
er
s
u
s
ex
p
er
i
m
e
n
ts
f
o
r
v
ar
io
u
s
d
is
p
er
s
ed
d
atasets
.
T
h
e
r
eg
r
ess
io
n
-
b
ase
d
ap
p
r
o
x
im
a
tio
n
s
ar
e
in
w
o
r
th
y
tr
ea
t
m
en
t
w
it
h
th
e
s
i
m
u
l
atio
n
ex
p
er
i
m
e
n
ts
,
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6930
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
,
Vo
l.
18
,
No
.
4
,
A
u
g
u
s
t 2
0
2
0
:
1
8
0
2
-
1
8
0
8
1808
r
esu
lti
n
g
th
e
v
alid
it
y
a
n
d
an
a
lo
g
y
co
m
p
ar
ed
to
th
e
GDA
.
Fu
r
t
h
er
m
o
r
e,
co
m
p
r
eh
e
n
s
iv
e
s
i
m
u
lat
io
n
r
e
s
u
lt
s
,
in
cl
u
d
in
g
o
u
tco
m
e
s
eq
u
e
n
ce
s
,
clea
r
l
y
r
ev
ea
l
th
e
f
ast
co
n
v
er
g
en
ce
o
f
o
p
tim
al
v
a
lu
e
s
in
m
a
n
y
ca
s
e
s
.
An
ad
ap
tiv
e
s
tep
s
ize
is
i
n
th
e
f
o
cu
s
o
f
o
u
r
f
u
tu
r
e
w
o
r
k
.
RE
F
E
R
E
NC
E
S
[1
]
A
lsh
a
f
lu
t
A
.
,
T
h
a
y
a
n
a
n
th
a
n
V.
,
“
T
ra
ff
ic
P
re
d
ictin
g
M
o
d
e
l
f
o
r
Dy
n
a
m
i
c
S
p
e
c
tru
m
S
h
a
rin
g
Ov
e
r
5
G
Ne
t
w
o
rk
s
,”
In
ter
n
a
t
io
n
a
l
J
o
u
rn
a
l
o
f
A
d
v
a
n
c
e
d
Co
m
p
u
ter
S
c
ie
n
c
e
a
n
d
A
p
p
li
c
a
t
io
n
s
,
v
o
l.
9
,
n
o
.
6
,
p
p
.
3
6
9
-
3
7
4
,
2
0
1
8
.
[2
]
Oliv
e
ira
T
.
P
.
,
Ba
rb
a
r
J
.
S
.
,
S
o
a
re
s
A
.
S.
,
“
Co
m
p
u
ter
n
e
tw
o
rk
traff
i
c
p
re
d
ictio
n
:
a
c
o
m
p
a
riso
n
b
e
tw
e
e
n
trad
it
i
o
n
a
l
a
n
d
d
e
e
p
le
a
rn
in
g
n
e
u
ra
l
n
e
tw
o
rk
s
,”
I
n
ter
n
a
ti
o
n
a
l
J
o
u
rn
a
l
o
f
B
ig
D
a
ta
In
telli
g
e
n
c
e
,
v
o
l.
3
,
n
o
.
1
,
p
p
.
28
-
37
,
2
0
1
6
.
[3
]
M
o
z
o
A
.
,
Ord
o
z
g
o
it
i
B
.
,
G
ó
m
e
z
-
Ca
n
a
v
a
l
S
.
,
“
F
o
re
c
a
stin
g
sh
o
rt
-
term
d
a
ta
c
e
n
ter
n
e
tw
o
rk
tr
a
ff
ic
lo
a
d
w
it
h
c
o
n
v
o
lu
ti
o
n
a
l
n
e
u
ra
l
n
e
tw
o
rk
s
,”
PL
o
S
ONE
,
v
o
l
.
13
,
n
o
.
2
,
p
p
.
1
-
31
,
2
0
1
8
.
[4
]
P
a
rv
e
e
n
H
.
,
S
riv
a
sta
v
a
R.
,
“
Ne
two
rk
T
ra
ff
i
c
A
n
a
l
y
sis
a
n
d
P
re
d
icti
o
n
-
A
L
it
e
ra
tu
re
Re
v
ie
w
,”
In
ter
n
a
ti
o
n
a
l
J
o
u
r
n
a
l
o
f
Res
e
a
rc
h
a
n
d
De
v
e
lo
p
me
n
t
in
Ap
p
li
e
d
S
c
ien
c
e
a
n
d
E
n
g
in
e
e
rin
g
,
v
o
l.
10
,
n
o
.
1
,
p
p
.
1
-
4
,
2
0
1
6
.
[5
]
Iq
b
a
l
M
.
F
.
,
Zah
i
d
M
.
,
Ha
b
i
b
D
.
,
Jo
h
n
L
.
K.
,
“
Eff
icie
n
t
P
re
d
icti
o
n
o
f
Ne
t
w
o
rk
T
ra
ff
i
c
f
o
r
Re
a
l
-
T
i
m
e
A
p
p
li
c
a
ti
o
n
s
,”
J
o
u
rn
a
l
o
f
Co
m
p
u
ter
Ne
two
rk
s a
n
d
C
o
mm
u
n
ica
ti
o
n
s
,
v
o
l.
2
0
1
9
,
p
p
.
1
-
1
1
,
2
0
1
9
.
[6
]
Zh
a
n
i
M
.
F
.
,
E
lb
iaz
e
H
.
,
Ka
m
o
u
n
F
.
,
“
A
n
a
l
y
sis
a
n
d
P
re
d
icti
o
n
o
f
Re
a
l
Ne
t
w
o
rk
T
ra
ff
ic
,”
J
o
u
rn
a
l
o
f
Ne
two
rk
s
,
v
o
l.
4
,
n
o
.
9
,
p
p
.
8
5
5
-
8
6
5
,
2
0
0
9
.
[7
]
Na
re
jo
S
.
,
P
a
se
ro
E.
,
“
A
n
A
p
p
li
c
a
ti
o
n
o
f
In
tern
e
t
T
ra
ff
ic
P
re
d
ictio
n
w
it
h
De
e
p
Ne
u
ra
l
Ne
tw
o
rk
,”
Co
n
fer
e
n
c
e
:
W
IRN
2
0
1
6
,
2
6
th
I
ta
li
a
n
W
o
rk
sh
o
p
o
n
Ne
u
ra
l
Ne
two
rk
s,
2
0
1
6
.
[8
]
W
a
n
g
W
.
e
t
a
l.
,
“
A
n
e
tw
o
rk
tra
ff
ic
f
lo
w
p
re
d
ictio
n
w
it
h
d
e
e
p
lea
rn
in
g
a
p
p
ro
a
c
h
f
o
r
larg
e
-
sc
a
le
m
e
tro
p
o
li
tan
a
re
a
n
e
tw
o
rk
,”
IEE
E/
IFI
P
Ne
two
rk
O
p
e
ra
ti
o
n
s a
n
d
M
a
n
a
g
e
me
n
t
S
y
mp
o
siu
m
,
2
0
1
8
.
[9
]
S
v
ig
e
lj
A
.
,
S
e
rn
e
c
R
.
,
A
li
c
K.
,
“
N
e
tw
o
rk
tr
a
ff
i
c
m
o
d
e
li
n
g
f
o
r
lo
a
d
p
re
d
ictio
n
:
a
u
se
r
-
c
e
n
tri
c
a
p
p
ro
a
c
h
,”
IEE
E
Ne
two
rk
,
v
o
l.
29
,
n
o
.
4
,
p
p
.
88
-
96
,
2
0
1
5
.
[1
0
]
S
o
n
g
S
.
,
Ch
a
u
d
h
u
ri
K
.
,
S
a
rw
a
te
A
.
D.
,
“
Lea
rn
in
g
f
ro
m
Da
ta
w
it
h
He
tero
g
e
n
e
o
u
s
No
ise
u
si
n
g
S
G
D
,”
Pro
c
e
e
d
in
g
s o
f
th
e
Ei
g
h
tee
n
t
h
In
ter
n
a
ti
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
Arti
fi
c
i
a
l
I
n
telli
g
e
n
c
e
a
n
d
S
ta
ti
stics
,
2
0
1
5
.
[1
1
]
Ra
d
y
A
.
M
.
,
Zak
a
ria
H
.
M.
,
“
Co
m
p
u
ter
Ne
t
w
o
rk
P
e
rf
o
rm
a
n
c
e
m
a
n
a
g
e
m
e
n
t
u
sin
g
a
S
im
p
le
Ne
t
wo
rk
M
a
n
a
g
e
m
e
n
t
P
r
o
to
c
o
l
,”
I
n
ter
n
a
t
io
n
a
l
J
o
u
rn
a
l
o
f
Co
m
p
u
t
in
g
Aca
d
e
mic
Res
e
a
rc
h
,
v
o
l.
6
,
n
o
.
2
,
p
p
.
50
-
58
,
2
0
1
7
.
[1
2
]
V
a
u
g
h
a
n
J
.
,
S
to
e
v
S
.
,
M
ich
a
i
li
d
is
G
.
,
“
N
e
t
w
o
rk
-
W
id
e
S
tatisti
c
a
l
M
o
d
e
li
n
g
,
P
re
d
ictio
n
,
a
n
d
M
o
n
it
o
ri
n
g
o
f
Co
m
p
u
ter
T
ra
ff
ic
,”
J
o
u
rn
a
l
T
e
c
h
n
o
me
trics
,
v
o
l.
55
,
n
o
.
1
,
p
p
.
79
-
93
,
2
0
1
3
.
[1
3
]
P
i
n
g
F
.
,
M
c
Co
n
n
e
ll
C
.
,
Hw
a
n
g
J.
,
“
A
Re
tro
sp
e
c
ti
v
e
A
p
p
ro
a
c
h
f
o
r
A
c
c
u
ra
te
Ne
t
w
o
rk
L
a
t
e
n
c
y
P
re
d
ictio
n
,”
Pro
c
e
e
d
in
g
s
o
f
1
9
th
In
ter
n
a
ti
o
n
a
l
Co
n
fer
e
n
c
e
o
n
Co
m
p
u
ter
Co
mm
u
n
ica
ti
o
n
s
a
n
d
Ne
two
rk
s,
2
0
1
0
.
[1
4
]
L
iu
B
.
,
Niu
D
.
,
L
i
Z
.
,
Zh
a
o
H
.
V.
,
“
Ne
t
w
o
rk
late
n
c
y
p
re
d
ictio
n
f
o
r
p
e
rso
n
a
l
d
e
v
ice
s: Di
sta
n
c
e
-
f
e
a
tu
re
d
e
c
o
m
p
o
siti
o
n
f
ro
m
3
D sa
m
p
li
n
g
,”
IEE
E
C
o
n
fer
e
n
c
e
o
n
C
o
mp
u
ter
Co
mm
u
n
ica
ti
o
n
s
,
2
0
1
5
.
[1
5
]
Ch
e
n
G
.
,
“
S
p
a
ti
o
tem
p
o
ra
l
In
d
i
v
id
u
a
l
M
o
b
il
e
Da
ta
T
ra
ff
ic
P
r
e
d
ictio
n
,”
[
T
e
c
h
n
ica
l
Re
p
o
rt
]
RT
-
0
4
9
7
,
INRIA
S
a
c
la
y
-
Ile
-
de
-
F
ra
n
c
e
,
2
0
1
8
.
[
O
n
li
n
e
]
.
A
v
a
il
a
b
le:
h
tt
p
s://
h
a
l.
in
r
ia.f
r/h
a
l
-
0
1
6
7
5
5
7
3
v
2
/d
o
c
u
m
e
n
t
[1
6
]
Zh
a
n
g
C
.
,
P
a
u
l
P
.
,
Ha
m
e
d
H.
,
“
De
e
p
L
e
a
rn
in
g
in
M
o
b
i
le
a
n
d
W
irele
ss
Ne
t
w
o
rk
in
g
:
A
S
u
rv
e
y
,”
IEE
E
Co
mm
u
n
ica
ti
o
n
s
S
u
rv
e
y
s a
n
d
T
u
to
ri
a
ls
,
p
p
.
1
-
6
7
,
2
0
1
8
.
[1
7
]
Ka
tsa
ra
k
is
e
t
a
l.
,
“
On
Us
e
r
-
Ce
n
tri
c
T
o
o
ls
f
o
r
Qo
E
-
Ba
se
d
Re
c
o
m
m
e
n
d
a
ti
o
n
a
n
d
Re
a
l
-
T
ime
A
n
a
l
y
sis
o
f
L
a
rg
e
-
S
c
a
le
M
a
rk
e
ts
,”
IEE
E
Co
mm
u
n
ica
ti
o
n
M
a
g
a
zin
e
,
v
o
l.
5
2
,
n
o
.
9
,
p
p
.
37
-
43
,
2
0
1
4
.
[1
8
]
Ka
m
in
sk
a
-
Ch
u
c
h
m
a
la
A
.
,
“
S
p
a
ti
a
l
in
ter
n
e
t
traf
f
ic
lo
a
d
f
o
re
c
a
stin
g
with
u
sin
g
e
stim
a
ti
o
n
m
e
th
o
d
,
”
Pr
o
c
e
d
ia
Co
m
p
u
ter
S
c
ien
c
e
,
v
o
l,
35
,
p
p
.
2
9
0
–
2
9
8
,
2
0
1
4
.
[1
9
]
P
ra
n
g
c
h
u
m
p
o
l
D.
,
“
A
Ne
tw
o
rk
T
r
a
ff
ic
P
re
d
icti
o
n
A
lg
o
rit
h
m
Ba
se
d
o
n
Da
ta
M
i
n
i
n
g
T
e
c
h
n
iq
u
e
,”
In
ter
n
a
ti
o
n
a
l
J
o
u
r
n
a
l
o
f
Co
m
p
u
ter
a
n
d
In
fo
rm
a
t
io
n
En
g
in
e
e
rin
g
,
v
o
l.
7
,
n
o
.
7
,
p
p
.
9
9
9
-
1
0
0
2
,
2
0
1
3
.
[2
0
]
Hu
K
.
,
A
lex
S
.
,
De
m
e
tri
s
A
.
,
Co
n
sta
n
ti
n
e
D.
,
“
Esti
m
a
ti
n
g
a
n
d
F
o
re
c
a
stin
g
Ne
t
w
o
rk
T
ra
ff
ic
P
e
rf
o
rm
a
n
c
e
Ba
se
d
o
n
S
tatisti
c
a
l
P
a
tt
e
rn
s
Ob
se
rv
e
d
in
S
NMP
Da
ta
,”
Pr
o
c
e
e
d
in
g
s
o
f
th
e
9
th
in
ter
n
a
ti
o
n
a
l
c
o
n
fer
e
n
c
e
o
n
m
a
c
h
in
e
lea
rn
in
g
a
n
d
d
a
ta
min
i
n
g
in
p
a
tt
e
rn
re
c
o
g
n
it
io
n
,
2
0
1
3
.
[2
1
]
Ye
D
.
e
t
a
l.
,
“
Im
p
ro
v
e
d
Ech
o
S
ta
te
Ne
tw
o
rk
(ES
N)
f
o
r
th
e
P
re
d
ict
io
n
o
f
Ne
tw
o
rk
T
ra
ff
i
c
,”
Pro
c
e
e
d
in
g
s
o
f
t
h
e
1
1
th
In
ter
n
a
t
io
n
a
l
C
o
n
fer
e
n
c
e
o
n
M
o
b
il
e
M
u
lt
ime
d
ia
Co
mm
u
n
ica
t
io
n
s
,
2
0
1
8
.
[2
2
]
L
i
R
.
,
Zh
a
o
Z
.
,
Z
h
o
u
X
.
,
P
a
l
ico
t
J
.
,
Z
h
a
n
g
H.
,
“
T
h
e
p
re
d
icti
o
n
a
n
a
l
y
sis
o
f
c
e
ll
u
lar
ra
d
io
a
c
c
e
ss
n
e
two
rk
traff
ic:
F
ro
m
e
n
tro
p
y
th
e
o
ry
to
n
e
tw
o
rk
in
g
p
ra
c
ti
c
e
,”
IEE
E
Co
mm
u
n
ica
ti
o
n
s M
a
g
a
zin
e
,
v
o
l.
52
,
n
o
.
6
,
p
p
.
2
3
4
-
2
4
0
,
2
0
1
4
.
[2
3
]
Zh
a
n
g
B
.
,
M
a
o
Y
.
,
Z
h
a
n
g
M
.
,
Y
u
Y
.
,
Jia
n
g
G
.
,
De
n
g
B.
,
“
T
h
e
S
tab
il
it
y
o
f
th
e
In
tern
e
t
T
ra
ff
ic
F
e
a
tu
re
s
,”
Pro
c
e
e
d
in
g
s
o
f
In
ter
n
a
ti
o
n
a
l
C
o
n
fer
e
n
c
e
o
n
E
d
u
c
a
t
io
n
,
M
a
n
a
g
e
me
n
t
a
n
d
C
o
mp
u
ter
S
c
ien
c
e
,
2
0
1
6
.
[2
4
]
L
o
sh
c
h
il
o
v
I
.
,
H
u
tt
e
r
F
.
,
“
S
G
D
R:
S
to
c
h
a
stic
G
ra
d
ien
t
De
sc
e
n
t
w
it
h
W
a
r
m
Re
sta
rts
,”
Pro
c
e
e
d
in
g
s
o
f
t
h
e
5
th
In
ter
n
a
t
io
n
a
l
C
o
n
fer
e
n
c
e
o
n
L
e
a
rn
in
g
Rep
re
se
n
ta
ti
o
n
s
,
2
0
1
7
.
[2
5
]
Ch
e
n
C
.
S
.
e
t
a
l.
,
“
Us
in
g
Ne
u
ra
l
Ne
tw
o
rk
w
it
h
G
ra
d
ien
t
De
sc
e
n
t
A
d
a
p
ti
v
e
L
e
a
rn
in
g
Ra
te
T
ra
in
in
g
A
lg
o
rit
h
m
f
o
r
G
e
o
m
e
tri
c
Dilu
ti
o
n
o
f
P
re
c
isio
n
A
p
p
ro
x
ima
ti
o
n
,”
IO
S
R
J
o
u
rn
a
l
o
f
El
e
c
trica
l
a
n
d
El
e
c
tro
n
ics
En
g
i
n
e
e
rin
g
,
v
o
l.
13
,
n
o
.
4
,
p
p
.
1
-
6
,
2
0
1
8
.
[2
6
]
M
it
ra
P
.
,
“
F
a
st
Co
n
v
e
rg
e
n
c
e
f
o
r
S
to
c
h
a
stic
a
n
d
Distrib
u
ted
G
ra
d
ien
t
De
sc
e
n
t
i
n
th
e
In
ter
p
o
lati
o
n
L
i
m
it
,”
T
h
e
2
6
t
h
Eu
ro
p
e
a
n
S
i
g
n
a
l
Pro
c
e
ss
in
g
Co
n
f
e
re
n
c
e
,
2
0
1
8
.
Evaluation Warning : The document was created with Spire.PDF for Python.