T
E
L
K
O
M
NIKA
T
elec
o
mm
un
ica
t
io
n Co
m
pu
t
i
ng
E
lect
ro
nics
a
nd
Co
ntr
o
l
Vo
l.
24
,
No
.
4
,
A
u
g
u
s
t
20
26
,
p
p
.
1
204
~1
2
1
5
I
SS
N:
1
6
9
3
-
6
9
3
0
,
DOI
: 1
0
.
1
2
9
2
8
/
T
E
L
KOM
NI
K
A
.
v
24
i
4
.
27561
1204
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//telko
mn
ika
.
u
a
d
.
a
c.
i
d
Rea
l
-
ti
m
e
ano
m
a
l
y
det
ec
tion sy
ste
m
using
best
per
f
o
r
m
ed
m
a
chine
learning
m
o
del
Vict
o
r
M
a
t
hebu
la
,
B
uk
o
hw
o
M
icha
el
E
s
ief
a
rienrhe
D
e
p
a
r
t
me
n
t
o
f
C
o
mp
u
t
e
r
S
c
i
e
n
c
e
a
n
d
I
n
f
o
r
mat
i
o
n
S
y
st
e
ms,
N
o
r
t
h
-
W
e
st
U
n
i
v
e
r
si
t
y
,
M
a
h
i
k
e
n
g
,
S
o
u
t
h
A
f
r
i
c
a
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Sep
21
,
2
0
2
5
R
ev
i
s
ed
A
p
r
9
,
2
0
2
6
A
cc
ep
ted
Ma
y
25
,
2026
Eff
e
c
ti
v
e
a
n
o
m
a
l
y
d
e
tec
ti
o
n
is
c
ri
ti
c
a
l
f
o
r
p
r
o
tec
ti
n
g
o
rg
a
n
iza
ti
o
n
a
l
n
e
tw
o
rk
s
a
g
a
in
st
in
c
re
a
sin
g
l
y
so
p
h
isti
c
a
ted
c
y
b
e
r
th
re
a
ts.
Ho
w
e
v
e
r,
m
o
st
m
a
c
h
in
e
lea
rn
in
g
-
b
a
se
d
i
n
tru
si
o
n
d
e
tec
ti
o
n
m
o
d
e
ls
a
re
d
e
v
e
lo
p
e
d
a
n
d
v
a
li
d
a
ted
u
si
n
g
p
u
b
li
c
b
e
n
c
h
m
a
r
k
d
a
tas
e
ts,
wh
ich
m
a
y
n
o
t
re
f
lec
t
th
e
o
p
e
ra
ti
o
n
a
l
c
h
a
ra
c
teristics
,
traff
ic
b
e
h
a
v
io
r,
a
n
d
t
h
re
a
t
p
a
tt
e
r
n
s
o
f
re
a
l
i
n
stit
u
ti
o
n
a
l
n
e
tw
o
rk
s.
In
th
e
c
a
se
o
f
Um
a
lu
s
i,
th
e
re
is
c
u
rre
n
tl
y
n
o
a
n
o
m
a
l
y
d
e
tec
ti
o
n
m
o
d
e
l
c
u
sto
m
ize
d
a
n
d
v
a
li
d
a
ted
u
sin
g
Um
a
lu
si
-
sp
e
c
if
ic
n
e
tw
o
rk
traff
ic
,
c
re
a
ti
n
g
a
p
ra
c
ti
c
a
l
g
a
p
i
n
d
e
p
l
o
y
a
b
le
c
y
b
e
rse
c
u
rit
y
c
a
p
a
b
il
it
y
.
Th
is
st
u
d
y
p
ro
p
o
se
s
a
h
y
b
rid
m
a
c
h
in
e
lea
rn
i
n
g
f
ra
m
e
w
o
r
k
tailo
re
d
to
su
p
p
o
r
t
a
c
c
u
ra
te,
e
ff
ici
e
n
t,
a
n
d
o
p
e
ra
ti
o
n
a
ll
y
re
le
v
a
n
t
a
n
o
m
a
l
y
d
e
tec
ti
o
n
.
Us
in
g
k
n
o
w
led
g
e
d
isc
o
v
e
r
y
in
d
a
tab
a
se
s
(KD
D)
p
ro
c
e
ss
,
n
e
tw
o
rk
tra
ff
ic
d
a
ta
w
e
re
c
o
ll
e
c
ted
a
n
d
p
re
-
p
r
o
c
e
ss
e
d
th
ro
u
g
h
n
o
r
m
a
li
z
a
ti
o
n
,
lab
e
l
e
n
c
o
d
in
g
,
m
is
sin
g
v
a
lu
e
trea
t
m
e
n
t,
a
n
d
d
im
e
n
sio
n
a
li
ty
re
d
u
c
ti
o
n
u
sin
g
p
r
in
c
ip
a
l
c
o
m
p
o
n
e
n
t
a
n
a
ly
sis
(P
CA
).
T
h
e
1
6
h
y
b
rid
m
o
d
e
ls
i
n
teg
ra
ti
n
g
u
n
s
u
p
e
rv
ise
d
a
n
o
m
a
l
y
d
e
tec
ti
o
n
w
it
h
su
p
e
rv
ise
d
c
las
sif
ica
ti
o
n
w
e
re
i
m
p
le
m
e
n
ted
a
n
d
c
o
m
p
a
ra
ti
v
e
l
y
e
v
a
lu
a
ted
.
Ex
p
e
rim
e
n
tal
f
in
d
in
g
s
in
d
ica
te
th
a
t
th
e
d
e
n
sity
-
b
a
se
d
sp
a
ti
a
l
c
lu
ste
rin
g
o
f
a
p
p
li
c
a
ti
o
n
s
d
a
tab
a
se
sc
a
n
(DBSCA
N)
+
ra
n
d
o
m
f
o
re
st
(RF
)
m
o
d
e
l
a
c
h
iev
e
d
9
9
.
9
2
%
a
c
c
u
ra
c
y
w
h
il
e
m
a
in
t
a
in
in
g
a
lo
w
fa
ls
e
p
o
siti
v
e
(F
P
)
c
o
st,
m
a
k
in
g
it
su
it
a
b
le
f
o
r
a
se
c
u
rit
y
o
p
e
ra
ti
o
n
s
c
e
n
tre
(
S
OC).
In
a
d
d
it
i
o
n
,
a
F
las
k
-
b
a
se
d
w
e
b
a
p
p
li
c
a
ti
o
n
w
a
s
d
e
v
e
lo
p
e
d
to
e
n
a
b
le
re
a
l
-
ti
m
e
d
e
p
lo
y
m
e
n
t
b
y
sn
iff
in
g
li
v
e
n
e
tw
o
rk
tra
ff
ic,
e
x
e
c
u
ti
n
g
in
f
e
re
n
c
e
,
a
n
d
p
e
rsistin
g
re
su
lt
s i
n
a
n
S
QL
it
e
d
a
tab
a
se
.
K
ey
w
o
r
d
s
:
An
o
m
al
y
d
etec
tio
n
A
r
ti
f
icial
i
n
tel
lig
e
n
ce
H
y
b
r
id
m
ac
h
in
e
lear
n
in
g
Kn
o
w
led
g
e
d
is
co
v
er
ies i
n
d
atab
ases
Ma
ch
i
n
e
lear
n
i
n
g
al
g
o
r
it
h
m
s
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
:
B
u
k
o
h
w
o
Mic
h
ae
l E
s
ie
f
ar
ie
n
r
h
e
Dep
ar
t
m
en
t o
f
C
o
m
p
u
ter
Scie
n
ce
an
d
I
n
f
o
r
m
atio
n
S
y
s
te
m
s
,
No
r
th
-
W
est U
n
i
v
er
s
it
y
M
m
ab
ath
o
U
n
it 5
,
Ma
h
i
k
en
g
,
So
u
t
h
A
f
r
ica
E
m
ail: M
ic
h
ae
l.E
s
ie
f
ar
ie
n
r
h
e
@
n
w
u
.
ac
.
za
1.
I
NT
RO
D
UCT
I
O
N
T
h
e
in
tr
o
d
u
ctio
n
o
f
ar
ti
f
icial
i
n
telli
g
e
n
ce
(
AI
)
h
as
m
ad
e
a
h
u
g
e
i
m
p
ac
t
o
n
d
if
f
er
e
n
t
i
n
d
u
s
t
r
ies
[
1
]
,
as
it
h
as
p
r
o
p
elled
tr
an
s
f
o
r
m
ativ
e
ch
an
g
es,
w
it
h
i
n
cr
ea
s
ed
p
r
ed
ictiv
e
an
al
y
s
is
a
n
d
p
r
o
d
u
ctiv
it
y
en
h
a
n
ce
m
en
t
s
i
n
o
r
g
an
izatio
n
al
ac
tiv
it
ies
[
2
]
.
T
h
er
e
is
a
n
ee
d
to
ex
p
lo
r
e
th
e
ad
o
p
tio
n
an
d
im
p
ac
t
o
f
A
I
,
p
r
o
m
p
ti
n
g
a
n
ex
a
m
in
at
io
n
o
f
it
s
e
f
f
ec
t
iv
e
n
e
s
s
i
n
th
e
i
n
f
o
r
m
atio
n
an
d
co
m
m
u
n
icat
io
n
tec
h
n
o
lo
g
y
(
I
C
T
)
s
ec
u
r
it
y
s
y
s
te
m
s
o
f
U
m
al
u
s
i.
U
m
al
u
s
i
is
t
h
e
co
u
n
cil
r
esp
o
n
s
ib
le
f
o
r
q
u
alit
y
ass
u
r
an
ce
in
g
en
er
al
an
d
f
u
r
th
er
ed
u
ca
tio
n
an
d
tr
ain
i
n
g
ce
r
ti
f
icatio
n
f
o
r
th
e
N
atio
n
al
Qu
a
li
f
ica
tio
n
Fra
m
e
w
o
r
k
(
NQF)
lev
el
s
1
-
4
in
So
u
t
h
Af
r
ica.
As
co
r
p
o
r
ate
n
et
w
o
r
k
s
ar
e
i
n
cr
ea
s
in
g
l
y
e
x
p
o
s
ed
to
s
o
p
h
is
ticated
an
d
ev
o
lv
i
n
g
c
y
b
er
t
h
r
ea
ts
,
th
e
c
y
b
er
s
ec
u
r
it
y
s
y
s
te
m
s
t
h
at
p
r
o
tect
th
ese
n
et
w
o
r
k
s
n
ee
d
to
i
m
p
r
o
v
e,
as
t
h
e
tec
h
n
iq
u
e
s
u
s
e
d
to
p
en
etr
ate
th
e
m
ar
e
co
n
s
tan
tl
y
e
v
o
lv
i
n
g
.
Desp
ite
th
e
av
ailab
ilit
y
o
f
in
t
r
u
s
io
n
d
etec
tio
n
s
y
s
te
m
s
(
I
DS
)
,
th
e
n
et
w
o
r
k
en
v
ir
o
n
m
e
n
t
at
U
m
al
u
s
i
is
s
t
ill
ex
p
o
s
ed
to
a
b
n
o
r
m
al
a
n
d
m
al
icio
u
s
tr
a
f
f
ic
d
u
e
to
li
m
itatio
n
s
i
n
ex
is
t
in
g
d
etec
tio
n
m
et
h
o
d
o
lo
g
ies.
E
x
is
tin
g
I
DS
tech
n
o
lo
g
ies
s
tr
u
g
g
l
e
to
k
ee
p
p
ac
e
w
ith
t
h
e
co
m
p
lex
it
y
a
n
d
s
ev
er
i
t
y
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
R
ea
l
-
time
a
n
o
m
a
ly
d
etec
tio
n
s
ystem
u
s
in
g
b
est p
erfo
r
med
ma
ch
in
e
lea
r
n
in
g
mo
d
el
(
V
icto
r
Ma
th
eb
u
la
)
1205
o
f
c
y
b
er
-
attac
k
s
,
lead
in
g
to
m
o
r
e
f
alse
alar
m
s
an
d
s
lo
w
er
d
etec
tio
n
r
ate
[
3
]
So
m
e
o
f
th
e
ch
alle
n
g
e
s
th
at
e
x
is
t
in
I
DS
as
h
ig
h
li
g
h
ted
b
y
[
4
]
,
[
5
]
ar
e
f
alse
alar
m
r
ate,
lo
w
d
etec
tio
n
r
ate,
u
n
b
alan
ce
d
d
ataset,
an
d
r
esp
o
n
s
e
ti
m
e.
A
I
-
b
ased
an
o
m
a
l
y
d
e
tectio
n
s
y
s
te
m
s
ar
e
ad
v
an
c
ed
I
C
T
s
ec
u
r
it
y
s
y
s
te
m
s
i
n
teg
r
ated
w
it
h
A
I
tech
n
o
lo
g
ies d
esi
g
n
ed
to
p
r
o
te
ct
th
e
n
et
w
o
r
k
in
f
r
astr
u
ct
u
r
e
[
5
]
.
T
h
is
r
esear
ch
ai
m
s
to
d
esi
g
n
,
i
m
p
le
m
en
t,
a
n
d
ev
al
u
ate
v
ar
i
o
u
s
s
eq
u
en
tia
l
h
y
b
r
id
an
o
m
al
y
d
etec
tio
n
m
o
d
el
s
t
h
at
i
n
te
g
r
ate
u
n
s
u
p
er
v
is
ed
a
n
d
s
u
p
er
v
i
s
ed
m
ac
h
i
n
e
lear
n
i
n
g
(
ML
)
a
lg
o
r
it
h
m
s
to
en
h
a
n
ce
ac
c
u
r
ac
y
,
p
r
ec
is
io
n
,
an
d
r
eliab
ilit
y
w
h
en
d
etec
ti
n
g
s
u
s
p
icio
u
s
b
eh
av
io
r
s
a
n
d
an
o
m
alie
s
w
i
th
i
n
th
e
n
et
w
o
r
k
in
f
r
astru
ct
u
r
e
at
U
m
al
u
s
i.
S
u
s
p
icio
u
s
an
d
ab
n
o
r
m
al
b
eh
av
io
u
r
s
r
ef
er
to
ac
tiv
ities
an
d
tr
an
s
ac
tio
n
s
w
it
h
i
n
th
e
n
et
w
o
r
k
i
n
f
r
astru
ct
u
r
e
t
h
at
ar
e
o
u
ts
id
e
n
o
r
m
al
tr
af
f
ic
o
r
tr
an
s
ac
tio
n
s
,
s
u
c
h
as
h
i
g
h
n
et
w
o
r
k
tr
af
f
ic
f
lo
w
(
in
o
r
o
u
t)
,
u
n
u
s
u
al
tr
af
f
ic
d
esti
n
ati
o
n
s
,
an
d
u
n
u
s
u
al
tr
af
f
ic
th
r
o
u
g
h
a
p
ar
ticu
lar
p
r
o
to
co
l.
T
h
e
u
n
s
u
p
er
v
is
ed
M
L
alg
o
r
ith
m
s
u
s
ed
in
th
i
s
s
tu
d
y
ar
e
is
o
la
tio
n
f
o
r
est
(
I
F),
o
n
e
class
s
u
p
p
o
r
t
v
ec
to
r
m
ac
h
i
n
e
(
OC
SVM)
,
lo
ca
l
o
u
tlier
f
ac
to
r
(
L
OF)
,
an
d
d
en
s
it
y
-
b
ased
s
p
atial
clu
s
ter
in
g
o
f
ap
p
licatio
n
s
w
ith
n
o
i
s
e
(
DB
SC
A
N
)
,
co
m
b
in
ed
w
it
h
f
o
u
r
s
u
p
er
v
i
s
ed
ML
al
g
o
r
ith
m
s
:
r
a
n
d
o
m
f
o
r
est
(
R
F
)
,
d
ec
is
io
n
tr
ee
(
DT
)
,
Naïv
e
B
ay
e
s
(
NB
)
,
an
d
K
-
n
ea
r
est
n
ei
g
h
b
o
r
(
KNN)
.
a
s
t
h
ese
m
o
d
els
h
a
v
e
p
r
o
v
en
to
h
av
e
g
o
o
d
p
er
f
o
r
m
a
n
ce
m
o
d
els
f
o
r
an
o
m
al
y
d
etec
tio
n
f
r
o
m
liter
at
u
r
e.
T
h
e
f
i
n
d
in
g
s
o
f
t
h
i
s
r
esear
ch
co
n
tr
ib
u
te
to
w
ar
d
d
ev
elo
p
in
g
a
p
r
ac
tical
an
d
d
ep
lo
y
ab
le
an
o
m
al
y
d
etec
tio
n
ap
p
r
o
ac
h
th
at
s
u
p
p
o
r
ts
s
ec
u
r
it
y
o
p
er
atio
n
s
ce
n
tr
e
(
SO
C
)
m
o
n
ito
r
i
n
g
a
n
d
d
ec
is
io
n
-
m
a
k
i
n
g
i
n
r
ea
l
-
w
o
r
ld
n
et
w
o
r
k
en
v
ir
o
n
m
e
n
ts
.
E
x
ten
s
i
v
e
r
esear
ch
h
a
s
b
ee
n
co
n
d
u
cted
o
n
th
e
ap
p
licatio
n
o
f
A
I
in
an
o
m
a
l
y
d
etec
tio
n
.
I
n
th
i
s
s
ec
tio
n
,
a
r
ev
ie
w
i
s
p
r
esen
ted
o
f
v
ar
io
u
s
s
t
u
d
ies
f
o
cu
s
in
g
o
n
h
y
b
r
id
an
o
m
al
y
d
etec
tio
n
m
o
d
els
w
it
h
in
n
et
w
o
r
k
en
v
ir
o
n
m
e
n
t
s
.
T
ab
le
1
p
r
esen
ts
d
if
f
er
en
t
s
t
u
d
ies
t
h
at
ar
e
e
x
a
m
i
n
ed
to
ex
p
lo
r
e
th
e
m
o
d
els,
t
h
e
M
L
al
g
o
r
ith
m
s
,
an
d
h
o
w
th
e
y
p
er
f
o
r
m
ed
o
n
v
a
r
io
u
s
d
atasets
u
tili
ze
d
d
u
r
i
n
g
ex
p
er
i
m
e
n
tatio
n
.
T
h
e
s
tr
u
ct
u
r
e
o
f
t
h
i
s
p
ap
er
is
o
r
g
an
ized
i
n
to
f
o
u
r
s
ec
t
io
n
s
.
T
h
e
m
eth
o
d
o
lo
g
y
ap
p
lied
in
t
h
is
p
ap
er
is
d
escr
ib
ed
in
s
ec
tio
n
2
.
Dif
f
er
en
t
s
eq
u
e
n
tial
h
y
b
r
id
m
o
d
els,
co
m
b
i
n
in
g
u
n
s
u
p
er
v
i
s
ed
an
o
m
al
y
d
etec
tio
n
w
it
h
s
u
p
er
v
i
s
ed
class
i
f
ier
s
,
w
er
e
tr
ain
ed
an
d
ev
alu
ate
d
to
d
eter
m
i
n
e
th
e
h
i
g
h
est
p
er
f
o
r
m
i
n
g
m
o
d
el.
Sectio
n
3
o
f
th
is
p
ap
er
p
r
esen
ts
t
h
e
r
es
u
lt
s
o
f
th
e
ex
p
er
i
m
en
t
s
a
n
d
d
is
cu
s
s
io
n
s
o
f
t
h
eir
p
er
f
o
r
m
a
n
ce
.
Se
ctio
n
4
p
r
esen
t
s
t
h
e
f
i
n
d
in
g
s
a
n
d
f
u
tu
r
e
w
o
r
k
to
en
h
an
ce
n
et
w
o
r
k
s
ec
u
r
it
y
u
s
i
n
g
ML
.
T
ab
le
1
.
A
n
o
m
al
y
d
ete
ctio
n
u
s
in
g
h
y
b
r
id
ML
R
e
f
.
Y
e
a
r
S
u
mm
a
r
y
[
6
]
2
0
2
4
T
h
e
i
r
st
u
d
y
c
o
mp
a
r
e
d
t
h
e
p
e
r
f
o
r
man
c
e
o
f
mu
l
t
i
p
l
e
M
L
a
l
g
o
r
i
t
h
ms (
s
u
p
e
r
v
i
se
d
,
e
n
se
mb
l
e
,
a
n
d
n
e
u
r
a
l
mo
d
e
l
s)
f
o
r
I
D
S
,
i
n
c
l
u
d
i
n
g
K
N
N
,
S
V
M
,
D
T
,
N
B
,
a
n
d
R
F
.
T
h
e
a
u
t
h
o
r
s
p
r
o
p
o
se
d
a
h
y
b
r
i
d
mo
d
e
l
t
h
a
t
u
t
i
l
i
z
e
s
R
F
a
n
d
S
V
M
a
l
g
o
r
i
t
h
ms
t
o
l
e
v
e
r
a
g
e
t
h
e
st
r
e
n
g
t
h
s
o
f
b
o
t
h
a
l
g
o
r
i
t
h
ms.
[
7
]
2
0
2
4
T
h
e
y
c
o
n
d
u
c
t
e
d
a
st
u
d
y
t
o
i
m
p
r
o
v
e
e
a
r
l
y
d
e
t
e
c
t
i
o
n
o
f
c
y
b
e
r
-
a
t
t
a
c
k
s
b
y
d
e
v
e
l
o
p
i
n
g
a
n
o
v
e
l
h
y
b
r
i
d
a
n
o
mal
y
d
e
t
e
c
t
i
o
n
mo
d
e
l
t
h
a
t
i
n
t
e
g
r
a
t
e
d
D
B
S
C
A
N
a
n
d
I
F
.
T
h
e
i
r
st
u
d
y
a
i
me
d
t
o
r
e
d
u
c
e
f
a
l
se
p
o
si
t
i
v
e
s
(
F
P
)
a
n
d
e
n
h
a
n
c
e
d
e
t
e
c
t
i
o
n
a
c
c
u
r
a
c
y
i
n
t
h
e
n
e
t
w
o
r
k
.
[
8
]
2
0
2
4
T
h
e
y
c
o
n
d
u
c
t
e
d
a
s
t
u
d
y
t
o
d
e
v
e
l
o
p
a
n
d
e
v
a
l
u
a
t
e
h
y
b
r
i
d
ML
mo
d
e
l
s
t
h
a
t
c
o
mb
i
n
e
su
p
e
r
v
i
se
d
a
n
d
u
n
s
u
p
e
r
v
i
se
d
l
e
a
r
n
i
n
g
a
p
p
r
o
a
c
h
e
s
t
o
e
n
h
a
n
c
e
n
e
t
w
o
r
k
i
n
t
r
u
s
i
o
n
d
e
t
e
c
t
i
o
n
.
T
h
e
i
r
st
u
d
y
a
i
me
d
t
o
o
v
e
r
c
o
me
l
i
mi
t
a
t
i
o
n
s
o
f
t
r
a
d
i
t
i
o
n
a
l
mo
d
e
l
s b
y
l
e
v
e
r
a
g
i
n
g
t
h
e
s
t
r
e
n
g
t
h
s
o
f
b
o
t
h
u
n
su
p
e
r
v
i
se
d
a
n
d
su
p
e
r
v
i
se
d
M
L
a
l
g
o
r
i
t
h
ms
.
[
9
]
2
0
2
3
T
h
e
y
c
o
n
d
u
c
t
e
d
a
st
u
d
y
a
i
me
d
a
t
e
v
a
l
u
a
t
i
n
g
t
h
e
e
f
f
e
c
t
i
v
e
n
e
ss
o
f
I
F
(
a
n
u
n
s
u
p
e
r
v
i
se
d
mo
d
e
l
)
a
n
d
S
V
M
(
a
su
p
e
r
v
i
se
d
mo
d
e
l
)
f
o
r
a
n
o
mal
y
-
b
a
sed
I
D
S
(
A
I
D
S
)
.
T
h
e
o
b
j
e
c
t
i
v
e
o
f
t
h
e
i
r
st
u
d
y
w
a
s
t
o
a
d
d
r
e
ss
c
h
a
l
l
e
n
g
e
s
i
n
d
e
t
e
c
t
i
n
g
a
b
n
o
r
mal
n
e
t
w
o
r
k
b
e
h
a
v
i
o
r
i
n
l
a
r
g
e
-
sca
l
e
d
a
t
a
a
n
d
i
m
p
r
o
v
e
d
e
t
e
c
t
i
o
n
a
c
c
u
r
a
c
y
.
[
1
0
]
2
0
2
5
T
h
e
y
c
o
n
d
u
c
t
e
d
a
s
t
u
d
y
a
i
me
d
a
t
d
e
v
e
l
o
p
i
n
g
a
p
r
e
d
i
c
t
i
v
e
mai
n
t
e
n
a
n
c
e
sy
st
e
m
f
o
r
e
n
v
i
r
o
n
me
n
t
a
l
se
n
so
r
s
b
y
d
e
t
e
c
t
i
n
g
a
n
o
mal
i
e
s i
n
u
n
l
a
b
e
l
e
d
I
o
T
se
n
so
r
d
a
t
a
.
[
1
1
]
2
0
2
2
T
h
e
y
c
o
n
d
u
c
t
e
d
a
st
u
d
y
a
i
me
d
a
t
d
e
v
e
l
o
p
i
n
g
a
h
y
b
r
i
d
M
L
mo
d
e
l
f
o
r
a
n
o
mal
y
d
e
t
e
c
t
i
o
n
i
n
n
e
t
w
o
r
k
se
c
u
r
i
t
y
.
T
h
e
o
b
j
e
c
t
i
v
e
o
f
t
h
e
st
u
d
y
w
a
s
t
o
i
m
p
r
o
v
e
d
e
t
e
c
t
i
o
n
a
c
c
u
r
a
c
y
,
r
e
d
u
c
e
FP
r
a
t
e
s,
a
n
d
e
n
h
a
n
c
e
r
e
a
l
-
t
i
me
p
r
o
c
e
ssi
n
g
e
f
f
i
c
i
e
n
c
y
f
o
r
I
D
S
.
2.
M
E
T
H
O
D
T
h
is
s
tu
d
y
f
o
llo
w
ed
a
k
n
o
w
le
d
g
e
d
is
co
v
er
y
in
d
atab
ases
(
KDD
)
p
r
o
ce
s
s
to
im
p
le
m
e
n
t
a
s
eq
u
en
tia
l
h
y
b
r
id
ML
an
o
m
a
l
y
d
etec
tio
n
an
d
class
if
icatio
n
m
o
d
el.
KDD
p
r
o
ce
s
s
a
p
p
lies
th
e
f
o
llo
w
i
n
g
cr
itical
p
h
ase
s
to
en
s
u
r
e
ef
f
icie
n
t
a
n
d
ef
f
ec
ti
v
e
d
etec
tio
n
:
d
ata
co
llectio
n
,
d
at
a
p
r
ep
r
o
c
ess
in
g
,
d
ata
r
ed
u
ctio
n
,
d
ata
m
i
n
i
n
g
,
a
n
d
ev
alu
a
tio
n
.
T
h
is
s
ec
tio
n
d
escr
i
b
es
h
o
w
ea
ch
o
f
t
h
e
f
iv
e
s
tep
s
o
f
th
e
KDD
p
r
o
ce
s
s
w
a
s
ap
p
lied
in
t
h
is
r
esear
c
h
s
tu
d
y
,
u
s
i
n
g
th
e
m
et
h
o
d
o
lo
g
y
d
ep
icted
in
Fi
g
u
r
e
1
.
T
h
e
m
o
d
el
w
i
th
h
i
g
h
p
er
f
o
r
m
a
n
ce
w
a
s
s
elec
ted
a
f
ter
t
h
e
ev
al
u
a
tio
n
p
h
a
s
e.
P
y
th
o
n
w
as
u
s
ed
to
i
m
p
le
m
e
n
t a
ll th
e
m
o
d
els ex
p
er
i
m
e
n
ted
w
it
h
.
2
.
1
.
Da
t
a
c
o
llect
io
n
T
h
is
s
tu
d
y
u
s
es
th
e
d
atase
t
co
m
p
iled
f
r
o
m
r
a
w
n
et
w
o
r
k
tr
af
f
ic
f
r
o
m
t
h
e
U
m
al
u
s
i
n
et
w
o
r
k
en
v
ir
o
n
m
e
n
t.
T
h
e
d
ata
w
as
co
llected
d
ir
ec
tly
f
r
o
m
li
v
e
n
et
wo
r
k
in
ter
f
ac
es
an
d
r
ef
lect
s
r
ea
l
o
p
e
r
atio
n
al
tr
af
f
i
c
co
n
d
itio
n
s
,
e
n
ab
lin
g
a
r
ea
lis
tic
ev
al
u
atio
n
o
f
an
o
m
al
y
d
etec
tio
n
m
o
d
els.
T
h
e
d
ataset
en
co
m
p
as
s
es
t
w
o
ca
teg
o
r
ies
o
f
f
ea
t
u
r
es:
co
r
e
p
ac
k
e
t
attr
ib
u
tes,
d
ir
ec
tl
y
ca
p
tu
r
ed
f
r
o
m
ea
c
h
p
ac
k
et,
an
d
d
er
iv
ed
f
ea
tu
r
es,
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1693
-
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.
24
,
No
.
4
,
A
u
g
u
s
t
20
26
:
1
204
-
1
2
1
5
1206
en
g
i
n
ee
r
ed
d
u
r
in
g
p
r
ep
r
o
ce
s
s
in
g
to
en
h
a
n
ce
class
if
icatio
n
ca
p
ab
ilit
ies.
T
ab
le
2
s
h
o
w
s
a
n
o
v
er
v
ie
w
o
f
th
e
d
ataset
f
ea
t
u
r
es a
n
d
t
h
eir
d
esc
r
ip
tio
n
s
.
Fig
u
r
e
1
.
H
y
b
r
id
an
o
m
al
y
d
etec
ti
o
n
m
et
h
o
d
o
lo
g
y
T
ab
le
2.
Data
s
et
f
ea
tu
r
es a
n
d
t
h
eir
d
escr
ip
tio
n
s
A
t
t
r
i
b
u
t
e
F
e
a
t
u
r
e
D
e
scri
p
t
i
o
n
C
o
r
e
p
a
c
k
e
t
a
t
t
r
i
b
u
t
e
s
p
a
c
k
e
t
_
n
u
m
b
e
r
S
e
q
u
e
n
t
i
a
l
p
a
c
k
e
t
i
n
d
e
x
t
i
me
st
a
mp
P
a
c
k
e
t
a
r
r
i
v
a
l
t
i
me
p
a
c
k
e
t
_
si
z
e
T
o
t
a
l
p
a
c
k
e
t
l
e
n
g
t
h
i
n
b
y
t
e
s
so
u
r
c
e
_
p
o
r
t
S
o
u
r
c
e
t
r
a
n
sp
o
r
t
p
o
r
t
d
e
st
_
p
o
r
t
D
e
st
i
n
a
t
i
o
n
t
r
a
n
s
p
o
r
t
p
o
r
t
p
r
o
t
o
c
o
l
I
n
t
e
r
n
e
t
p
r
o
t
o
c
o
l
(
IP
)
i
d
e
n
t
i
f
i
e
r
so
u
r
c
e
_
i
p
S
o
u
r
c
e
I
P
A
d
d
r
e
ss
d
e
st
_
i
p
D
e
st
i
n
a
t
i
o
n
I
P
A
d
d
r
e
ss
c
u
m
u
l
a
t
i
v
e
_
b
y
t
e
s
R
u
n
n
i
n
g
su
m o
f
p
a
c
k
e
t
s
i
z
e
s p
e
r
c
a
p
t
u
r
e
w
i
n
d
o
w
D
e
r
i
v
e
d
a
t
t
r
i
b
u
t
e
s
b
y
t
e
_
d
e
l
t
a
C
h
a
n
g
e
i
n
c
u
m
u
l
a
t
i
v
e
b
y
t
e
s b
e
t
w
e
e
n
c
o
n
se
c
u
t
i
v
e
p
a
c
k
e
t
s
p
a
c
k
e
t
_
r
a
t
e
p
a
c
k
e
t
s
p
e
r
se
c
o
n
d
b
y
t
e
_
r
a
t
e
b
y
t
e
s p
e
r
se
c
o
n
d
d
u
r
a
t
i
o
n
T
i
me
d
i
f
f
e
r
e
n
c
e
b
e
t
w
e
e
n
c
o
n
se
c
u
t
i
v
e
p
a
c
k
e
t
s
2
.
2
.
P
re
-
pro
ce
s
s
ing
Du
r
in
g
t
h
e
p
r
ep
r
o
ce
s
s
in
g
s
ta
g
e,
ir
r
elev
an
t,
d
u
p
licate
d
ata,
an
d
an
y
co
l
u
m
n
s
co
n
tain
in
g
d
ata
th
at
m
a
y
n
o
t
b
e
n
ee
d
ed
w
i
ll
b
e
r
em
o
v
e
d
f
r
o
m
t
h
e
d
ataset.
T
h
e
d
ataset
m
a
y
b
e
m
a
n
u
a
ll
y
m
o
d
if
ied
w
h
er
e
n
ec
es
s
ar
y
[
7
]
.
Hig
h
-
q
u
alit
y
d
ata
i
s
i
n
cli
n
ed
to
y
ield
r
eliab
le
k
n
o
w
led
g
e
c
o
m
p
ar
ed
to
lo
w
-
q
u
ali
t
y
d
ata
[
1
2
]
.
T
h
is
r
esear
ch
s
tu
d
y
ap
p
lies
t
h
r
ee
d
ata
clea
n
i
n
g
tec
h
n
iq
u
e
s
,
lab
el
en
co
d
in
g
,
n
o
r
m
a
lizatio
n
,
an
d
f
i
lli
n
g
o
f
m
is
s
i
n
g
v
al
u
es.
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
R
ea
l
-
time
a
n
o
m
a
ly
d
etec
tio
n
s
ystem
u
s
in
g
b
est p
erfo
r
med
ma
ch
in
e
lea
r
n
in
g
mo
d
el
(
V
icto
r
Ma
th
eb
u
la
)
1207
T
h
e
f
ir
s
t
p
h
ase
o
f
p
r
ep
r
o
ce
s
s
in
g
is
to
ap
p
l
y
lab
el
en
co
d
in
g
.
T
h
e
lab
el
en
co
d
in
g
m
e
th
o
d
,
also
k
n
o
w
n
as
o
r
d
in
al
en
co
d
in
g
,
ass
i
g
n
s
a
u
n
iq
u
e
in
te
g
er
to
ea
ch
ca
te
g
o
r
y
f
r
o
m
t
h
e
d
ataset
[
1
3
]
.
Fo
r
ex
a
m
p
le,
d
if
f
er
en
t
n
et
w
o
r
k
p
r
o
to
co
ls
ca
n
ea
ch
b
e
ass
ig
n
ed
an
in
te
g
er
v
al
u
e,
s
u
ch
as
0
f
o
r
tr
an
s
m
is
s
io
n
co
n
t
r
o
l
p
r
o
to
co
l
(
T
C
P
)
,
o
r
1
f
o
r
u
s
er
d
atag
r
a
m
p
r
o
to
c
o
l
(
UDP
)
.
T
h
e
n
ex
t
p
h
ase
o
f
d
ata
p
r
ep
r
o
ce
s
s
in
g
h
a
n
d
les
m
is
s
i
n
g
v
al
u
es
i
n
t
h
e
d
ataset.
Data
clea
n
s
i
n
g
in
cl
u
d
es
m
et
h
o
d
s
f
o
r
d
ea
lin
g
w
it
h
m
i
s
s
i
n
g
v
al
u
es,
l
ik
e
i
m
p
u
tatio
n
(
esti
m
at
in
g
m
is
s
i
n
g
v
alu
e
s
)
o
r
eli
m
i
n
ati
n
g
r
o
w
s
o
r
co
lu
m
n
s
t
h
at
h
av
e
a
n
e
x
ce
s
s
iv
e
n
u
m
b
er
o
f
m
i
s
s
i
n
g
v
al
u
e
s
[
1
4
]
.
T
h
e
m
ed
ia
n
v
alu
e
o
f
ea
ch
co
l
u
m
n
w
il
l b
e
ca
lcu
lated
an
d
u
s
ed
to
f
ill i
n
an
y
m
i
s
s
in
g
v
al
u
e
s
.
Div
id
i
n
g
th
e
d
ata
s
et
i
n
to
th
r
ee
s
u
b
s
et
s
,
b
y
ap
p
l
y
in
g
t
h
e
ti
m
e
-
b
ased
d
ataset
s
p
li
tti
n
g
m
e
th
o
d
,
p
r
ev
en
ts
d
ata
leak
ag
e
[
1
5
]
.
T
h
e
d
ataset
is
f
ir
s
t
s
o
r
ted
in
asce
n
d
in
g
o
r
d
er
b
y
ti
m
e
s
ta
m
p
attr
ib
u
te,
en
s
u
r
i
n
g
t
h
at
ea
r
lier
n
et
w
o
r
k
e
v
e
n
ts
p
r
ec
ed
e
th
e
l
ater
o
n
es.
O
n
ce
o
r
d
er
ed
,
th
e
d
ataset
is
d
i
v
id
ed
in
to
t
h
r
ee
c
o
n
s
ec
u
tiv
e
s
u
b
s
ets.
6
0
% f
o
r
th
e
tr
ain
i
n
g
s
et,
1
0
% f
o
r
th
e
v
al
id
atio
n
s
et,
a
n
d
3
0
% f
o
r
th
e
o
u
t
-
of
-
ti
m
e
s
et,
a
s
ill
u
s
tr
ated
in
Fi
g
u
r
e
2
.
Fig
u
r
e
2
.
T
im
e
-
b
ased
d
ata
s
p
li
t
T
h
e
last
p
h
ase
en
s
u
r
es
t
h
e
n
o
r
m
aliza
t
io
n
o
f
n
u
m
er
ical
d
ata
to
en
s
u
r
e
w
ei
g
h
t
in
g
in
th
e
ML
m
o
d
els.
No
r
m
a
lizatio
n
i
s
ac
h
ie
v
ed
b
y
ap
p
ly
i
n
g
m
a
x
-
m
in
s
ca
li
n
g
o
n
(
1
)
[
1
6
]
:
=
−
−
(
1
)
W
h
er
e,
r
ef
er
s
to
t
h
e
n
o
r
m
al
ized
f
ea
t
u
r
e
v
al
u
e,
w
ill
al
w
a
y
s
l
ie
b
et
w
ee
n
0
a
n
d
1
.
r
ef
er
s
to
th
e
o
r
ig
in
al
f
ea
t
u
r
e
v
al
u
e,
is
t
h
e
m
i
n
i
m
u
m
f
ea
tu
r
e
v
al
u
e,
is
t
h
e
m
ax
i
m
u
m
f
ea
t
u
r
e
v
alu
e,
t
h
e
lar
g
est
o
b
s
er
v
ed
.
2
.
3
.
Da
t
a
r
educt
io
n
P
r
in
cip
al
co
m
p
o
n
e
n
t
an
al
y
s
is
(
P
C
A
)
co
n
v
er
ts
t
h
e
h
i
g
h
-
d
i
m
en
s
io
n
a
l
f
ea
t
u
r
e
s
p
ac
e
in
to
a
lo
w
er
-
d
i
m
en
s
io
n
al
f
ea
t
u
r
e
s
p
ac
e
wh
ile
r
etai
n
in
g
as
m
u
ch
v
ar
ia
n
ce
(
in
f
o
r
m
atio
n
)
as
p
o
s
s
ib
l
e
[
1
7
]
.
T
h
is
h
elp
s
s
i
m
p
li
f
y
m
o
d
els,
r
ed
u
ce
co
m
p
u
tatio
n
al
lo
ad
,
an
d
en
h
an
ce
th
e
class
i
f
icat
io
n
p
er
f
o
r
m
an
ce
o
f
th
e
A
I
m
o
d
els.
T
h
is
s
tep
is
ess
e
n
tia
l
f
o
r
o
p
tim
izi
n
g
t
h
e
ap
p
licatio
n
o
f
d
ata
m
i
n
in
g
al
g
o
r
ith
m
s
,
a
s
it
r
ed
u
c
es
co
m
p
u
tat
io
n
ti
m
e
an
d
/o
r
en
h
a
n
ce
s
th
eir
p
er
f
o
r
m
an
ce
[
1
8
]
.
I
t
ac
h
iev
es
th
i
s
b
y
p
r
o
j
ec
tin
g
th
e
d
ata
o
n
to
a
s
m
aller
s
et
o
f
u
n
co
r
r
elate
d
p
r
in
cip
al
c
o
m
p
o
n
en
t
s
th
at
ca
p
tu
r
e
t
h
e
m
a
x
i
m
u
m
v
ar
ia
n
ce
in
th
e
o
r
ig
i
n
al
d
ataset
i
m
p
ac
ted
[
1
9
]
.
P
C
A
f
u
n
ctio
n
s
b
y
f
ir
s
t
ca
lc
u
latin
g
t
h
e
co
v
ar
ian
ce
m
atr
i
x
o
f
t
h
e
d
ataset
an
d
t
h
e
n
ca
r
r
y
in
g
o
u
t
eig
e
n
v
al
u
e
d
ec
o
m
p
o
s
itio
n
.
T
h
e
lead
in
g
eig
en
v
al
u
es
(
)
,
alo
n
g
w
it
h
th
eir
ass
o
ciate
d
eig
e
n
v
ec
to
r
s
(
)
,
s
er
v
e
as
th
e
f
o
u
n
d
atio
n
f
o
r
tr
an
s
f
o
r
m
i
n
g
t
h
e
d
ata
in
to
a
lo
w
er
-
d
i
m
en
s
io
n
al
s
p
ac
e.
C
o
n
s
eq
u
e
n
tl
y
,
th
e
n
e
w
co
o
r
d
in
ates
ar
e
d
er
iv
ed
f
r
o
m
th
e
s
e
ei
g
e
n
v
ec
t
o
r
s
,
w
h
ic
h
id
e
n
ti
f
y
th
e
d
ir
ec
tio
n
s
o
f
m
ax
i
m
u
m
v
ar
ian
ce
with
i
n
t
h
e
d
atase
t
a
s
s
h
o
w
n
in
(
2
)
.
I
n
th
i
s
ca
s
e,
=
λ
an
d
r
ep
r
esen
ts
th
e
n
u
m
b
er
o
f
o
b
s
e
r
v
atio
n
s
.
=
∑
(
−
)
(
−
)
=
1
(
2
)
2
.
4
.
Da
t
a
m
i
nin
g
T
h
e
p
r
im
ar
y
p
u
r
p
o
s
e
o
f
th
is
s
t
u
d
y
is
to
d
esig
n
,
i
m
p
le
m
en
t,
a
n
d
ev
alu
a
te
a
h
y
b
r
id
an
o
m
al
y
d
etec
tio
n
m
o
d
el
t
h
at
in
te
g
r
ates
u
n
s
u
p
er
v
is
ed
an
d
s
u
p
er
v
is
ed
M
L
alg
o
r
ith
m
s
in
o
r
d
er
to
en
h
an
ce
ac
cu
r
ac
y
,
p
r
ec
is
io
n
,
an
d
r
eliab
ilit
y
w
h
e
n
d
etec
ti
n
g
an
o
m
alie
s
w
it
h
i
n
t
h
e
n
et
wo
r
k
in
f
r
astru
c
tu
r
e
at
U
m
al
u
s
i
.
T
h
e
1
6
s
eq
u
en
tial
h
y
b
r
id
an
o
m
al
y
d
etec
tio
n
m
o
d
els
w
er
e
d
ev
elo
p
ed
,
ea
ch
c
o
m
b
in
i
n
g
o
n
e
o
f
f
o
u
r
u
n
s
u
p
er
v
is
ed
alg
o
r
ith
m
s
w
it
h
o
n
e
o
f
f
o
u
r
s
u
p
er
v
i
s
ed
alg
o
r
ith
m
s
.
T
h
e
co
m
b
in
a
tio
n
s
o
f
t
h
ese
alg
o
r
ith
m
s
f
o
r
m
ed
th
e
f
o
l
lo
w
i
n
g
m
o
d
els,
as
in
d
icate
d
in
T
ab
le
3
.
T
h
ese
m
o
d
el
s
ar
e
ca
teg
o
r
ized
ac
co
r
d
in
g
to
th
eir
r
esp
ec
ti
v
e
u
n
s
u
p
er
v
is
ed
alg
o
r
ith
m
s
,
r
esu
lti
n
g
in
f
o
u
r
ca
te
g
o
r
ies.
2
.
4
.
1
.
Da
t
a
i
m
ba
la
nce
T
o
o
v
er
co
m
e
th
e
p
r
o
b
lem
o
f
d
ata
im
b
alan
ce
f
r
o
m
t
h
e
co
llected
d
ata
f
r
o
m
t
h
e
U
m
al
u
s
i
n
et
w
o
r
k
,
class
w
ei
g
h
t
a
n
d
s
y
n
th
e
tic
m
i
n
o
r
it
y
o
v
er
s
a
m
p
li
n
g
tec
h
n
iq
u
e
(
SMOT
E
)
ar
e
ap
p
lied
.
SM
OT
E
p
r
ev
en
ts
clas
s
i
m
b
alan
ce
,
w
h
ich
ca
u
s
es
M
L
class
i
f
ier
s
to
b
e
b
iased
to
w
ar
d
th
e
m
aj
o
r
ity
clas
s
,
r
ed
u
cin
g
t
h
eir
ab
ilit
y
to
d
etec
t
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1693
-
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.
24
,
No
.
4
,
A
u
g
u
s
t
20
26
:
1
204
-
1
2
1
5
1208
an
o
m
alie
s
ef
f
ec
ti
v
el
y
[
2
0
]
,
[
2
1
]
.
SMOT
E
cr
ea
tes
ar
tif
icial
d
ata
p
o
in
ts
f
o
r
th
e
m
i
n
o
r
it
y
c
lass
u
s
i
n
g
(
3
)
.
Fo
r
ea
ch
m
in
o
r
it
y
clas
s
s
a
m
p
le
,
SMOT
E
s
elec
ts
o
n
e
o
f
it
s
k
-
n
ea
r
e
s
t
n
eig
h
b
o
u
r
s
,
an
d
g
e
n
er
ates
a
n
e
w
s
y
n
t
h
etic
s
a
m
e
[
2
2
]
:
=
+
(
−
)
(
3
)
W
h
er
e,
is
a
f
ea
tu
r
e
v
ec
to
r
f
r
o
m
t
h
e
m
in
o
r
it
y
cla
s
s
,
is
o
n
e
o
f
th
e
k
-
n
ea
r
est
n
ei
g
h
b
o
u
r
s
o
f
,
∈
[
0
,
1
]
,
a
r
an
d
o
m
n
u
m
b
er
,
is
a
n
e
w
s
y
n
th
e
tic
d
ata
p
o
in
t.
T
ab
le
3
.
L
is
t o
f
m
o
d
els d
e
v
elo
p
ed
U
n
su
p
e
r
v
i
se
d
S
u
p
e
r
v
i
se
d
M
o
d
e
l
n
a
me
Is
o
l
a
t
i
o
n
f
o
r
e
st
DT
I
F
+
D
T
NB
I
F
+
N
B
K
N
N
I
F
+
K
N
N
RF
I
F
+
R
F
O
n
e
-
c
l
a
ss SV
M
DT
O
C
S
V
M
+
D
T
NB
O
C
S
V
M
+
N
B
K
N
N
O
C
S
V
M
+
K
N
N
RF
O
C
S
V
M
+
R
F
L
o
c
a
l
o
u
t
l
i
e
r
f
a
c
t
o
r
DT
L
O
F
+
D
T
NB
L
O
F
+
N
B
K
N
N
L
O
F
+
K
N
N
RF
L
O
F
+
R
F
D
B
S
C
A
N
DT
D
B
S
C
A
N
+
D
T
NB
D
B
S
C
A
N
+
N
B
K
N
N
D
B
S
C
A
N
+
K
N
N
RF
D
B
S
C
A
N
+
R
F
2
.
5
.
E
v
a
lua
t
i
o
n
A
co
n
f
u
s
io
n
m
atr
i
x
,
also
k
n
o
w
n
a
s
an
er
r
o
r
m
atr
i
x
,
is
a
tab
le
th
at
s
u
m
m
ar
izes
t
h
e
p
er
f
o
r
m
an
ce
o
f
a
p
r
ed
ictio
n
o
r
class
if
icatio
n
m
o
d
el
[2
3
]
.
T
h
e
c
o
n
f
u
s
io
n
m
atr
ix
in
T
ab
le
4
p
r
o
v
id
es
k
ey
v
ar
iab
les,
n
a
m
el
y
,
tr
u
e
p
o
s
itiv
e
(
T
P
)
,
tr
u
e
n
eg
ativ
e
(
T
N)
,
FP
,
an
d
f
alse
n
eg
ati
v
e
(
FN)
.
T
h
e
(
4
)
-
(
7
)
illu
s
tr
ate
h
o
w
th
e
ac
cu
r
ac
y
,
r
ec
all,
p
r
ec
is
io
n
,
an
d
F1
-
s
co
r
e
w
er
e
c
alcu
lated
,
r
esp
ec
tiv
el
y
.
T
ab
le
4
.
C
o
n
f
u
s
io
n
m
atr
i
x
C
o
n
f
u
s
i
o
n
m
a
t
r
i
x
P
r
e
d
i
c
t
e
d
c
l
a
ss
N
o
r
mal
A
n
o
mal
y
A
c
t
u
a
l
c
l
a
ss
N
o
r
mal
TN
FP
A
n
o
mal
y
FN
TP
=
+
+
+
+
(
4
)
=
+
(
5
)
=
+
(
6
)
1
=
2
∗
∗
+
(
7
)
3.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
A
ll
th
e
m
o
d
els
ex
p
er
i
m
en
ted
w
i
th
w
er
e
ex
ec
u
ted
o
n
a
c
o
m
p
u
ter
co
n
f
i
g
u
r
ed
w
i
th
th
e
f
o
llo
w
i
n
g
h
ar
d
w
ar
e
a
n
d
s
o
f
t
w
ar
e
s
p
ec
i
f
i
ca
tio
n
s
:
6
4
-
b
it
W
in
d
o
w
s
1
1
o
p
er
atin
g
s
y
s
te
m
,
I
n
tel
i5
p
r
o
ce
s
s
o
r
,
4
7
0
GB
s
o
lid
s
t
a
te
d
r
iv
e
(
S
SD
)
s
t
o
r
ag
e
,
1
6
GB
o
f
r
an
d
o
m
ac
c
es
s
m
em
o
r
y
(
R
A
M
)
,
E
th
e
r
n
e
t a
d
a
p
t
e
r
th
a
t su
p
p
o
r
t
s
1
.
0
G
ig
a
b
it
s
p
e
r
s
e
c
o
n
d
(
G
b
p
s
)
.
P
y
th
o
n
v
e
r
s
i
o
n
3
.
1
3
w
a
s
u
t
il
is
e
d
f
o
r
ML
im
p
l
em
en
t
a
t
i
o
n
s
,
o
p
e
r
a
t
in
g
o
n
V
i
s
u
al
S
tu
d
i
o
C
o
d
e
v
e
r
s
i
o
n
1
.
1
0
2
.
3
(
u
s
e
r
s
e
tu
p
)
.
T
h
e
P
y
th
o
n
l
i
b
r
a
r
ie
s
r
e
q
u
i
r
e
d
f
o
r
t
h
is
e
x
p
e
r
im
en
t
a
r
e
s
cik
i
t
-
le
ar
n
,
p
a
n
d
as
,
n
u
m
p
y
,
im
b
le
a
r
n
,
f
l
as
k
,
a
n
d
jo
b
l
i
b
.
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
R
ea
l
-
time
a
n
o
m
a
ly
d
etec
tio
n
s
ystem
u
s
in
g
b
est p
erfo
r
med
ma
ch
in
e
lea
r
n
in
g
mo
d
el
(
V
icto
r
Ma
th
eb
u
la
)
1209
3
.
1
.
M
o
del
perf
o
r
m
a
nce
co
m
pa
r
is
o
n
T
h
e
1
6
h
y
b
r
id
an
o
m
al
y
d
etec
tio
n
m
o
d
els,
co
m
b
in
i
n
g
u
n
s
u
p
er
v
is
ed
an
d
s
u
p
er
v
is
ed
ML
alg
o
r
ith
m
s
,
w
er
e
ev
a
lu
ated
u
s
i
n
g
n
et
w
o
r
k
tr
af
f
ic
d
ata
co
llected
f
r
o
m
t
h
e
U
m
a
lu
s
i
n
e
t
w
o
r
k
en
v
ir
o
n
m
en
t.
T
h
e
o
b
j
ec
tiv
e
w
a
s
to
id
en
tify
t
h
e
m
o
s
t
ef
f
e
ctiv
e
m
o
d
el
f
o
r
SOC
d
ep
lo
y
m
en
t
b
ased
o
n
d
etec
tio
n
p
e
r
f
o
r
m
a
n
ce
,
s
tab
ilit
y
,
r
an
k
i
n
g
ca
p
ab
ilit
y
,
an
d
FP
co
s
t.
E
ac
h
m
o
d
el
w
as
as
s
es
s
ed
u
n
d
er
th
r
ee
class
i
m
b
a
lan
ce
h
an
d
lin
g
s
tr
ate
g
ies:
SMOT
E
o
v
er
s
am
p
li
n
g
,
s
u
p
er
v
is
ed
clas
s
w
ei
g
h
ti
n
g
,
a
n
d
a
b
aselin
e
w
it
h
o
u
t
class
b
alan
ci
n
g
.
T
h
e
u
n
s
u
p
er
v
is
ed
alg
o
r
ith
m
s
e
v
al
u
ated
w
er
e
L
OF,
DB
SC
A
N,
I
F,
a
n
d
OC
S
VM
.
E
ac
h
u
n
s
u
p
er
v
i
s
ed
alg
o
r
ith
m
w
as
p
air
ed
w
it
h
ea
ch
o
f
f
o
u
r
s
u
p
er
v
i
s
ed
clas
s
if
ier
s
:
R
F,
DT
,
NB
,
an
d
KN
N,
f
o
r
m
i
n
g
f
o
u
r
d
is
ti
n
ct
m
o
d
el
ca
teg
o
r
ies.
T
h
e
h
ig
h
e
s
t
-
p
er
f
o
r
m
i
n
g
h
y
b
r
id
m
o
d
el
f
r
o
m
ea
c
h
u
n
s
u
p
er
v
is
ed
ca
teg
o
r
y
w
as
s
elec
ted
an
d
co
m
p
ar
ed
t
o
d
eter
m
in
e
th
e
m
o
s
t s
u
itab
le
m
o
d
el
f
o
r
SOC
.
T
a
b
le
5
p
r
esen
ts
th
e
to
p
-
p
er
f
o
r
m
i
n
g
m
o
d
els f
r
o
m
ea
c
h
c
ateg
o
r
y
.
T
ab
le
5
.
Hig
h
est p
er
f
o
r
m
er
s
f
r
o
m
ea
c
h
ca
te
g
o
r
y
M
o
d
e
l
n
a
me
P
r
e
c
i
si
o
n
R
e
c
a
l
l
F1
-
s
c
o
r
e
A
c
c
u
r
a
c
y
(
%
)
I
F
+
R
F
89
87
85
87
.
29
O
C
S
V
M
+
R
F
96
94
95
93
.
92
L
O
F
+
D
T
1
0
0
99
99
98
.
60
D
B
S
C
A
N
+
R
F
1
0
0
1
0
0
1
0
0
99
.
92
3
.
2
.
Co
nfusi
o
n
m
a
t
rix
Fig
u
r
e
3
d
is
p
lay
s
t
h
e
c
o
n
f
u
s
io
n
m
atr
i
x
.
T
h
e
I
F+R
F
m
o
d
el
in
Fig
u
r
e
3
(
a)
in
d
icate
s
a
lar
g
e
n
u
m
b
er
o
f
TN
(
9
3
,
0
6
4
)
an
d
a
s
m
a
ll
n
u
m
b
er
o
f
FP
(
1
1
3
)
,
in
d
icatin
g
t
h
at
it
i
s
s
u
cc
es
s
f
u
l
at
r
e
d
u
cin
g
f
alse
aler
t
s
.
Ho
w
e
v
er
,
th
e
m
o
d
el
h
a
s
a
v
er
y
h
ig
h
FN
r
ate,
m
is
s
in
g
1
4
,
5
0
7
an
o
m
alo
u
s
tr
af
f
ic.
D
esp
ite
a
s
ig
n
i
f
ica
n
t
p
er
ce
n
tag
e
o
f
TN
,
th
e
m
o
d
el
m
is
c
lass
if
ied
a
lar
g
e
n
u
m
b
er
o
f
an
o
m
alies
a
s
t
y
p
ica
l
tr
af
f
ic
.
T
h
is
lead
s
to
lo
w
r
ec
all
f
o
r
th
e
an
o
m
al
y
clas
s
,
w
h
ic
h
is
a
cr
u
c
ial
r
estrictio
n
in
c
y
b
er
s
ec
u
r
it
y
s
etti
n
g
s
wh
er
e
u
n
d
is
co
v
er
ed
attac
k
s
o
f
f
er
a
s
ig
n
if
ica
n
t
r
i
s
k
.
B
ec
au
s
e
o
f
th
e
lar
g
e
n
u
m
b
er
o
f
m
i
s
clas
s
i
f
icatio
n
s
,
th
e
m
o
d
el
b
ec
o
m
es
u
n
s
u
itab
le
f
o
r
SO
C
d
ep
lo
y
m
e
n
t.
T
h
e
L
OF+DT
m
o
d
el
i
n
Fig
u
r
e
3
(
b
)
an
d
th
e
OC
SVM+
R
F
m
o
d
el
i
n
Fi
g
u
r
e
3
(
c)
h
ad
s
ev
er
e
class
i
f
icatio
n
b
iases
to
w
ar
d
th
e
an
o
m
alo
u
s
clas
s
,
1
1
3
4
5
1
an
d
1
0
7
4
3
4
,
r
esp
ec
tiv
el
y
.
T
h
is
class
i
f
icatio
n
w
o
u
ld
r
esu
lt
i
n
e
x
ce
s
s
iv
e
a
ler
t
g
e
n
er
atin
g
i
n
a
SO
C
en
v
ir
o
n
m
e
n
t.
L
OF
+
DT
p
r
o
d
u
ce
d
0
FP
an
d
7
TN
.
Desp
ite
th
eir
h
ig
h
d
etec
tio
n
r
ates,
th
e
s
e
t
w
o
m
o
d
els
ar
e
u
n
s
u
itab
le
f
o
r
o
p
er
atio
n
al
d
ep
lo
y
m
e
n
t
d
u
e
to
th
eir
in
ab
ilit
y
to
m
ai
n
tai
n
b
alan
ce
d
an
d
m
ea
n
i
n
g
f
u
l c
la
s
s
i
f
icatio
n
.
T
h
e
c
o
n
f
u
s
i
o
n
m
a
t
r
ix
f
o
r
D
B
SC
A
N+
R
F
m
o
d
e
l
,
d
em
o
n
s
t
r
a
t
ed
i
n
F
ig
u
r
e
3
(
d
)
,
in
d
i
c
at
e
s
th
a
t
t
h
is
m
o
d
el
p
r
o
p
e
r
ly
c
at
eg
o
r
i
z
e
d
p
r
a
ct
i
c
a
lly
al
l
ty
p
i
c
a
l
t
r
af
f
i
c
,
a
s
s
ee
n
b
y
e
x
t
r
a
o
r
d
i
n
a
r
i
ly
h
ig
h
T
N
an
d
a
l
o
w
F
P
r
a
t
e
.
T
h
is
e
q
u
a
t
es
t
o
a
n
e
a
r
-
z
e
r
o
F
P
r
at
e
,
w
h
ic
h
is
i
d
e
a
l
in
a
S
OC
s
e
tt
in
g
b
e
c
au
s
e
i
t
d
e
c
r
e
as
e
s
a
l
e
r
t
f
a
tig
u
e
a
n
d
o
p
e
r
a
t
i
o
n
a
l
o
v
e
r
h
e
a
d
.
On
ly
7
n
o
r
m
a
l
p
a
ck
e
t
s
w
er
e
w
r
o
n
g
ly
i
d
en
t
if
i
e
d
a
s
a
n
o
m
a
l
o
u
s
,
d
em
o
n
s
t
r
at
in
g
h
ig
h
p
r
e
c
is
i
o
n
.
3
.
3
.
Cla
s
s
ba
la
nce
s
t
ra
t
eg
ies
Fig
u
r
e
4
co
m
p
ar
es
t
h
e
d
i
f
f
er
e
n
t
cla
s
s
b
ala
n
cin
g
s
tr
ate
g
ies
e
x
p
er
i
m
e
n
ted
i
n
t
h
is
s
t
u
d
y
,
SM
OT
E
,
class
w
ei
g
h
t,
an
d
n
o
clas
s
b
alan
ci
n
g
,
o
n
th
e
ac
c
u
r
ac
y
o
f
t
h
e
f
o
u
r
h
ig
h
e
s
t
p
er
f
o
r
m
er
s
f
r
o
m
ea
ch
f
o
u
r
ca
teg
o
r
ies.
A
l
l
m
o
d
el
s
p
er
f
o
r
m
ed
r
elati
v
el
y
co
n
s
is
ten
tl
y
ac
r
o
s
s
d
if
f
er
e
n
t
class
-
b
alan
ci
n
g
s
tr
ateg
ie
s
,
ex
ce
p
t
th
e
L
OF+
DT
m
o
d
el.
T
h
e
w
ei
g
h
t
cla
s
s
b
ala
n
ce
f
o
r
L
O
F+DT
d
r
o
p
p
ed
d
r
as
ticall
y
to
4
8
.
5
%
co
m
p
ar
ed
to
SMOT
E
an
d
n
o
b
alan
cin
g
at
9
8
.
6
%
an
d
9
8
.
2
4
,
r
esp
ec
tiv
el
y
.
T
h
e
I
F
-
b
ased
m
o
d
el
s
w
er
e
n
o
t
s
e
n
s
i
tiv
e
to
class
b
alan
ci
n
g
;
th
e
h
ig
h
e
s
t
ac
c
u
r
ac
y
w
as
ac
h
ie
v
e
d
o
n
SMOT
E
o
v
er
s
a
m
p
l
in
g
.
T
h
e
v
ar
iatio
n
f
o
r
I
F+R
F
w
a
s
0
.
6
1
%
o
f
ac
cu
r
ac
y
.
OC
SVM+
R
F
m
o
d
el
p
er
f
o
r
m
ed
b
etter
at
n
o
b
alan
cin
g
,
w
i
th
an
ac
cu
r
ac
y
o
f
9
3
.
9
2
%,
w
h
ile
its
lo
w
est
w
a
s
9
1
.
7
7
%.
DB
SC
A
N
-
b
ased
m
o
d
els
ac
h
iev
ed
n
ea
r
-
p
er
f
ec
t
p
er
f
o
r
m
a
n
ce
ac
r
o
s
s
all
th
r
ee
s
tr
ateg
ies,
w
it
h
a
v
ar
iatio
n
o
f
o
n
l
y
0
.
0
2
%.
T
h
e
h
ig
h
e
s
t a
cc
u
r
ac
y
w
as p
r
o
d
u
ce
d
b
y
n
o
cl
as
s
b
alan
ci
n
g
at
9
9
.
9
6
%.
3
.
4
.
F
P
co
s
t
a
na
ly
s
is
FP
co
s
t
s
er
v
ed
as
a
cr
itical
e
v
alu
a
tio
n
m
e
tr
ic
f
o
r
m
o
d
el
s
elec
tio
n
in
th
is
s
t
u
d
y
.
Si
n
ce
ea
ch
f
als
e
alar
m
in
a
SOC
e
n
v
ir
o
n
m
en
t
in
cu
r
s
in
v
es
tig
a
tio
n
ti
m
e
a
n
d
r
eso
u
r
ce
ex
p
en
d
itu
r
e,
m
i
n
i
m
izi
n
g
FP
co
s
t
w
a
s
p
r
io
r
itized
alo
n
g
s
id
e
tr
ad
itio
n
al
p
er
f
o
r
m
an
ce
m
etr
ic
s
.
T
h
e
co
s
t
v
al
u
e
o
f
ea
c
h
m
o
d
el
is
ca
l
cu
lated
b
y
ap
p
l
y
i
n
g
(
8
)
,
w
h
er
e
FP
is
th
e
n
u
m
b
er
o
f
FP
in
a
m
o
d
el,
r
e
f
e
r
s
t
o
th
e
am
o
u
n
t
o
f
t
im
e
it
t
ak
es
t
o
in
v
es
t
ig
at
e
o
n
e
F
P,
in
w
h
ich
1
5
m
in
u
t
e
s
w
as
s
e
t
f
o
r
t
es
t
in
g
p
u
r
p
o
s
e
s
.
r
e
f
e
r
s
t
o
th
e
r
a
t
e
a
t
w
h
i
ch
th
e
S
OC
en
g
in
e
e
r
is
c
o
m
p
en
s
a
t
e
d
f
o
r
p
e
r
h
o
u
r
.
A
s
ta
n
d
a
r
d
r
a
t
e
o
f
R
3
0
0
.
0
0
h
as
b
e
en
s
e
t
f
o
r
th
e
p
u
r
p
o
s
e
o
f
t
es
t
in
g
d
i
f
f
e
r
en
t
m
o
d
el
s
.
=
FP
∗
(
60
)
∗
(
8
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1693
-
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.
24
,
No
.
4
,
A
u
g
u
s
t
20
26
:
1
204
-
1
2
1
5
1210
T
h
e
b
est
p
er
f
o
r
m
i
n
g
m
o
d
el
i
n
t
h
e
OC
SVM
ca
te
g
o
r
y
,
O
C
SVM+
R
F,
p
r
o
d
u
ce
d
th
e
h
i
g
h
est
FP
co
s
t,
ex
ce
ed
in
g
R
1
3
7
0
0
0
,
0
0
.
T
h
is
in
d
icate
s
a
lar
g
e
v
o
lu
m
e
o
f
n
o
r
m
al
tr
af
f
ic
cla
s
s
i
f
ied
as a
n
o
m
a
lies
.
T
h
is
s
u
g
g
es
t
s
p
o
o
r
p
r
ec
is
io
n
an
d
a
h
ig
h
o
p
er
atio
n
al
b
u
r
d
en
o
n
SOC
a
n
al
y
s
t
s
.
I
n
th
e
I
F
ca
teg
o
r
y
,
I
F+R
F
p
r
o
d
u
ce
d
a
m
o
d
er
ate
FP
co
s
t
o
f
R
8
4
7
5
.
0
0
.
DB
SC
N+
R
F
p
r
o
d
u
ce
d
a
g
o
o
d
F
P
c
o
s
t
o
f
R
5
2
5
.
0
0
,
m
ak
i
n
g
it
o
p
er
atio
n
all
y
id
ea
l
f
o
r
a
SOC
en
v
ir
o
n
m
e
n
t.
I
n
th
e
L
O
F
ca
teg
o
r
y
,
L
OF+
DT
m
o
d
el
p
r
o
d
u
ce
d
th
e
lo
w
e
s
t
FP
co
s
t
o
f
R
0
.
0
0
.
T
h
is
i
m
p
lie
s
t
h
at
t
h
i
s
m
o
d
el
d
id
n
o
t
d
etec
t
an
y
FP
d
u
r
in
g
th
e
m
o
d
el
e
v
al
u
atio
n
.
W
h
il
e
it
m
a
y
b
e
co
s
t
ef
f
icien
t,
th
is
m
a
y
b
e
a
s
ig
n
o
f
u
n
d
er
d
etec
tio
n
o
f
ac
tu
al
th
r
ea
ts
.
T
h
is
is
s
u
p
p
o
r
ted
b
y
L
O
F+DT
m
atr
i
x
co
n
f
u
s
io
n
i
n
F
ig
u
r
e
3
(
b
)
,
w
h
e
r
e
m
aj
o
r
ity
o
f
th
e
n
et
w
o
r
k
tr
a
f
f
ic
clas
s
i
f
ied
as
an
o
m
alies.
F
r
o
m
th
is
a
n
al
y
s
i
s
,
DB
SC
A
N+
R
F
h
as
t
h
e
lo
w
e
s
t
ac
ce
p
tab
le
FP
c
o
s
t,
h
ig
h
li
g
h
tin
g
its
s
u
p
er
io
r
p
r
ec
is
io
n
an
d
m
i
n
i
m
al
a
n
al
y
s
i
s
in
ter
v
e
n
tio
n
r
eq
u
ir
e
m
e
n
t
s
.
T
h
is
an
al
y
s
is
p
r
o
v
es
t
h
at
FP
co
s
t
is
a
cr
itical
ev
al
u
atio
n
m
e
tr
ic
f
o
r
I
DS
s
y
s
te
m
s
d
ep
lo
y
ed
in
e
n
ter
p
r
is
e
en
v
ir
o
n
m
en
ts
.
(
a)
(
b
)
(
c)
(
d
)
Fig
u
r
e
3
.
C
o
n
f
u
s
io
n
m
atr
ix
f
o
r
h
ig
h
p
er
f
o
r
m
er
s
:
(
a)
I
F+R
F,
(
b
)
L
OF+DT
,
(
c)
D
B
SC
AN+
R
F
,
an
d
(
d
)
OC
SVM+
R
F
Fig
u
r
e
4
.
C
lass
b
ala
n
cin
g
s
tr
ateg
ies
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
R
ea
l
-
time
a
n
o
m
a
ly
d
etec
tio
n
s
ystem
u
s
in
g
b
est p
erfo
r
med
ma
ch
in
e
lea
r
n
in
g
mo
d
el
(
V
icto
r
Ma
th
eb
u
la
)
1211
3
.
5
.
Dis
cus
s
io
n o
n t
he
hig
he
s
t
perf
o
r
m
er
DB
SC
A
N+
R
F
m
o
d
el
d
e
m
o
n
s
tr
ates
a
m
o
r
e
r
o
b
u
s
t
b
alan
ce
b
et
w
ee
n
d
etec
tio
n
ca
p
ab
ilit
y
,
s
tab
ilit
y
,
an
d
o
p
er
atio
n
al
co
s
t,
co
m
p
ar
ed
to
all
o
th
er
m
o
d
els
e
x
p
er
im
en
ted
w
i
th
i
n
t
h
is
s
tu
d
y
.
T
h
i
s
m
o
d
e
l
p
e
r
f
o
r
m
e
d
b
e
t
t
e
r
an
d
p
r
o
v
e
d
t
o
b
e
m
o
r
e
s
u
i
te
d
f
o
r
d
e
p
l
o
y
m
en
t
in
r
e
a
l
-
w
o
r
l
d
S
O
C
en
v
i
r
o
n
m
en
ts
w
h
e
r
e
t
r
a
f
f
ic
d
i
s
t
r
i
b
u
t
i
o
n
s
a
r
e
c
o
m
p
l
ex
,
d
y
n
am
i
c
,
an
d
im
b
a
l
an
ce
d
.
D
B
SC
A
N
i
s
l
e
s
s
s
en
s
it
iv
e
t
o
f
e
at
u
r
e
s
c
al
in
g
d
i
s
t
o
r
t
i
o
n
s
an
d
c
l
as
s
im
b
al
an
c
es
,
w
h
i
ch
a
r
e
c
o
m
m
o
n
in
r
e
a
l
-
w
o
r
l
d
n
etw
o
r
k
d
a
t
a.
T
h
i
s
e
n
a
b
l
e
s
D
B
SC
A
N
t
o
g
e
n
e
r
at
e
c
le
an
e
r
an
d
m
o
r
e
s
e
p
a
r
a
b
l
e
an
o
m
a
ly
la
b
e
ls
,
w
h
i
ch
e
n
h
an
ce
s
t
h
e
d
o
w
n
s
t
r
ea
m
s
u
p
e
r
v
i
s
e
d
cl
a
s
s
if
i
e
r
’
s
l
e
a
r
n
i
n
g
p
r
o
c
e
s
s
.
T
h
e
h
y
b
r
id
izatio
n
o
f
DB
S
C
AN
w
i
th
R
F
f
u
r
t
h
er
a
m
p
li
f
ies
t
h
is
ad
v
a
n
ta
g
e.
T
h
e
s
p
ec
if
ied
DB
SC
A
N
p
ar
am
eter
s
(
=
0
.
2
,
_
=
30
)
allo
w
t
h
e
m
o
d
el
to
ca
p
tu
r
e
d
en
s
e
r
eg
io
n
s
o
f
n
o
r
m
al
n
et
w
o
r
k
b
eh
av
io
r
w
h
ile
ef
f
ec
ti
v
el
y
s
ep
ar
atin
g
s
p
ar
s
e
tr
af
f
ic
p
atter
n
s
as
an
o
m
alie
s
,
w
h
ic
h
is
co
n
s
is
ten
t
w
ith
t
h
e
b
asic
s
tr
u
ct
u
r
e
an
d
i
m
b
ala
n
ce
o
f
r
ea
l
-
w
o
r
ld
n
et
w
o
r
k
tr
af
f
ic.
T
h
i
s
a
llo
w
s
t
h
e
RF
clas
s
i
f
ier
to
f
o
cu
s
o
n
d
is
cr
i
m
i
n
ati
v
e
tr
af
f
ic
p
atter
n
s
r
ath
er
t
h
a
n
o
u
tlier
s
o
r
b
o
r
d
er
lin
e
ex
a
m
p
le
s
.
I
t
lead
s
to
b
etter
g
e
n
er
aliza
ti
o
n
,
h
i
g
h
er
r
an
k
in
g
ab
ilit
y
,
an
d
m
o
r
e
co
n
s
i
s
te
n
t
p
er
f
o
r
m
an
ce
ac
r
o
s
s
v
a
lid
atio
n
f
o
ld
s
an
d
o
u
t
-
of
-
ti
m
e
d
ataset
s
.
T
h
is
m
o
d
el
o
f
f
er
s
th
e
m
o
s
t c
o
s
t
-
ef
f
icie
n
t so
l
u
tio
n
a
m
o
n
g
t
h
e
ev
a
lu
ated
m
o
d
el
s
.
3
.
6
.
M
o
del
re
t
ra
ini
ng
s
t
ra
t
e
g
y
Mo
d
el
r
etr
ain
in
g
i
s
ca
r
r
ied
o
u
t
to
en
s
u
r
e
th
at
t
h
e
an
o
m
al
y
d
etec
tio
n
s
y
s
te
m
s
ta
y
s
a
d
ap
tab
le
t
o
ch
an
g
i
n
g
n
et
w
o
r
k
tr
a
f
f
ic
p
att
er
n
s
at
U
m
al
u
s
i
a
n
d
d
ev
elo
p
in
g
a
ttack
b
e
h
av
io
r
s
.
D
u
r
in
g
r
et
r
ain
in
g
,
n
e
w
l
y
co
llected
n
et
w
o
r
k
tr
af
f
ic
d
ata
ar
e
a
p
p
en
d
ed
t
o
th
e
ex
is
tin
g
h
is
to
r
ical
d
ataset
an
d
s
o
r
ted
ch
r
o
n
o
lo
g
icall
y
to
p
r
eser
v
e
te
m
p
o
r
al
in
teg
r
it
y
.
T
h
e
s
u
p
er
v
is
ed
m
o
d
el
is
r
etr
ai
n
ed
ex
clu
s
iv
e
l
y
o
n
u
p
d
ated
n
o
r
m
al
tr
a
f
f
ic
s
a
m
p
les
to
r
e
-
estab
lis
h
an
ac
cu
r
ate
d
ec
is
io
n
b
o
u
n
d
ar
y
,
w
h
ile
t
h
e
u
n
s
u
p
er
v
i
s
ed
is
r
etr
ain
ed
u
s
i
n
g
th
e
an
o
m
al
y
lab
el
s
g
en
er
ated
b
y
th
e
u
p
d
ated
DB
SC
A
N
m
o
d
el.
T
o
p
r
ev
en
t
co
n
ce
p
t
d
r
if
t
an
d
d
ata
leak
ag
e,
r
etr
ain
in
g
ad
o
p
ts
a
ti
m
e
-
b
ased
s
p
lit
tin
g
m
et
h
o
d
,
w
it
h
c
u
r
r
en
t
d
ata
s
et
a
s
id
e
f
o
r
v
alid
ati
o
n
a
n
d
o
u
t
-
of
-
ti
m
e
(
O
OT
)
test
in
g
.
Me
tr
ics
s
u
c
h
as a
cc
u
r
ac
y
,
p
r
ec
is
io
n
,
r
ec
all,
F1
-
s
co
r
e,
an
d
FP
co
s
t a
r
e
u
s
ed
to
ass
es
s
m
o
d
el
p
er
f
o
r
m
an
ce
.
3
.
7
.
Co
m
pa
ri
s
o
n o
f
t
he
re
s
ults w
it
h e
x
is
t
ing
lite
ra
t
ure
T
h
e
r
esu
lts
o
b
tain
ed
in
th
is
r
esear
ch
s
t
u
d
y
ar
e
co
m
p
ar
ab
le
to
th
o
s
e
f
o
u
n
d
in
th
e
liter
at
u
r
e
.
T
ab
le
6
p
r
esen
ts
a
co
m
p
ar
is
o
n
o
f
th
e
p
er
f
o
r
m
an
ce
o
f
th
i
s
s
t
u
d
y
w
i
th
th
at
o
f
a
v
ailab
le
s
t
u
d
ies
f
r
o
m
l
iter
atu
r
e.
T
ab
le
6
.
C
o
m
p
ar
is
o
n
o
f
r
esu
l
t
s
w
ith
l
iter
atu
r
e
St
u
d
y
A
c
c
u
r
a
c
y
(
%
)
[
2
4
]
9
9
.
8
[
6
]
9
8
.
9
8
[
7
]
9
7
.
2
[
2
5
]
96
[
8
]
99
[
9
]
99
[
1
0
]
9
9
.
9
3
[
1
1
]
9
7
.
3
9
T
h
i
s st
u
d
y
9
9
.
9
2
3
.
8
.
Web
a
pp
lica
t
io
n
T
h
is
s
tu
d
y
i
m
p
le
m
e
n
ted
a
s
ec
u
r
e,
w
eb
-
b
ased
an
o
m
al
y
d
ete
ctio
n
p
latf
o
r
m
u
s
i
n
g
P
y
th
o
n
t
o
p
u
t
th
e
s
elec
ted
h
i
g
h
p
er
f
o
r
m
er
(
DB
SC
A
N+
R
F)
i
n
to
p
r
ac
tice.
T
h
e
ar
ch
itect
u
r
e,
s
h
o
w
n
in
Fi
g
u
r
e
5
,
w
as
cr
ea
ted
to
p
r
o
v
id
e
a
r
ea
l
-
ti
m
e
d
etec
ti
o
n
o
f
an
o
m
alo
u
s
n
et
w
o
r
k
tr
af
f
ic
w
h
ile
e
n
s
u
r
in
g
s
ca
lab
i
lit
y
,
s
ec
u
r
it
y
,
an
d
m
ai
n
tai
n
ab
ilit
y
.
T
h
e
s
y
s
te
m
f
o
llo
w
s
a
la
y
er
ed
ar
ch
itect
u
r
e
co
n
s
is
ti
n
g
o
f
a
p
r
esen
tatio
n
la
y
er
,
ap
p
licatio
n
la
y
er
,
m
o
d
el
s
er
v
i
n
g
la
y
er
,
ta
s
k
e
x
ec
u
tio
n
la
y
er
,
a
n
d
d
ata
p
er
s
is
ten
ce
la
y
er
.
T
h
e
p
latf
o
r
m
co
n
s
is
t
s
o
f
f
o
u
r
p
r
im
ar
y
co
m
p
o
n
e
n
t
s
:
t
h
e
clie
n
t
-
f
ac
i
n
g
d
as
h
b
o
ar
d
,
th
e
Flas
k
-
b
ased
w
eb
ap
p
licatio
n
s
er
v
er
,
th
e
M
L
p
r
ed
ictio
n
en
g
i
n
e
(
m
o
d
el
s
er
v
in
g
)
,
an
d
th
e
S
QL
ite
d
atab
ase
f
o
r
s
to
r
ag
e.
T
h
e
s
y
s
te
m
i
s
co
n
n
ec
te
d
to
th
e
lo
ca
l
ar
ea
n
et
w
o
r
k
(
L
A
N
)
at
a
s
tr
ate
g
ic
p
o
in
t,
b
et
w
ee
n
t
h
e
L
AN
an
d
t
h
e
i
n
ter
n
et,
to
g
a
th
er
all
n
et
wo
r
k
tr
af
f
ic
en
ter
i
n
g
an
d
ex
iti
n
g
t
h
e
L
A
N.
T
h
e
s
y
s
te
m
f
o
llo
w
s
a
la
y
er
ed
ar
ch
itect
u
r
e
co
n
s
i
s
ti
n
g
o
f
t
h
e
p
r
esen
tatio
n
la
y
er
,
ap
p
licatio
n
lay
er
,
m
o
d
el
s
er
v
i
n
g
la
y
er
,
task
ex
ec
u
tio
n
an
d
q
u
eu
i
n
g
,
d
ata
p
er
s
is
ten
ce
la
y
er
,
an
d
s
ec
u
r
it
y
la
y
er
.
E
ac
h
la
y
er
is
d
is
c
u
s
s
ed
in
T
ab
le
7
.
3
.
9
.
Sy
s
t
em
perf
o
r
m
a
nce
T
o
ev
alu
ate
th
e
p
er
f
o
r
m
a
n
ce
o
f
th
e
s
y
s
te
m
,
w
e
ca
lcu
lated
th
e
laten
c
y
a
n
d
th
r
o
u
g
h
p
u
t,
f
o
r
en
d
-
to
-
en
d
b
atch
p
r
o
ce
s
s
in
g
ti
m
e,
w
h
ic
h
co
v
er
s
d
ata
ca
p
tu
r
e,
p
r
e
p
r
o
c
ess
i
n
g
,
d
etec
tio
n
,
a
n
d
class
i
f
i
ca
tio
n
.
T
h
e
laten
c
y
an
d
th
r
o
u
g
h
p
u
t
ar
e
ca
lcu
lated
u
s
in
g
(
9
)
an
d
(
1
0
)
,
r
esp
ec
tiv
el
y
,
w
h
er
e
is
th
e
n
u
m
b
er
o
f
r
ec
o
r
d
s
(
p
ac
k
ets)
p
r
o
ce
s
s
ed
in
a
b
atch
,
is
t
h
e
i
n
f
er
e
n
ce
ti
m
e
i
n
s
ec
o
n
d
s
,
a
n
d
is
t
h
e
to
tal
i
n
f
er
en
ce
t
i
m
e
in
m
ill
is
ec
o
n
d
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1693
-
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.
24
,
No
.
4
,
A
u
g
u
s
t
20
26
:
1
204
-
1
2
1
5
1212
T
h
e
li
m
it o
f
ea
c
h
b
atch
co
u
n
t i
s
s
et
to
1
0
0
,
an
d
th
e
ti
m
eo
u
t is
2
s
ec
o
n
d
s
.
T
h
e
s
y
s
te
m
p
r
o
v
ed
to
b
e
ef
f
icie
n
t f
o
r
r
ea
l
-
ti
m
e
n
et
w
o
r
k
a
n
o
m
al
y
d
e
tectio
n
b
y
co
n
s
is
ten
t
l
y
p
r
o
d
u
cin
g
late
n
c
y
at
a
r
ate
o
f
5
0
m
s
an
d
b
elo
w
,
w
it
h
a
th
r
o
u
g
h
p
u
t o
f
o
v
er
2
2
0
0
p
ac
k
ets/
s
ec
.
(
/
)
=
(
9
)
ℎ
ℎ
(
/
)
=
(
1
0
)
Fig
u
r
e
5
.
A
r
ch
itectu
r
al
o
v
er
v
i
e
w
o
f
t
h
e
a
n
o
m
al
y
d
etec
tio
n
s
y
s
te
m
T
ab
le
7
.
Sy
s
te
m
la
y
er
s
L
a
y
e
r
s
P
u
r
p
o
se
P
r
e
se
n
t
a
t
i
o
n
l
a
y
e
r
T
h
e
S
O
C
a
n
a
l
y
st
s
i
n
t
e
r
a
c
t
w
i
t
h
t
h
e
s
y
st
e
m
t
h
r
o
u
g
h
a
u
se
r
i
n
t
e
r
f
a
c
e
a
c
c
e
ssi
b
l
e
t
h
r
o
u
g
h
a
w
e
b
b
r
o
w
se
r
.
T
h
e
i
n
t
e
r
f
a
c
e
e
n
a
b
l
e
s
a
u
t
h
e
n
t
i
c
a
t
e
d
u
se
r
s
t
o
v
i
e
w
l
i
v
e
n
e
t
w
o
r
k
t
r
a
f
f
i
c
,
i
n
sp
e
c
t
d
e
t
e
c
t
e
d
a
n
o
mal
i
e
s,
a
n
d
e
x
p
o
r
t
r
e
p
o
r
t
s.
A
p
p
l
i
c
a
t
i
o
n
l
a
y
e
r
T
h
e
a
p
p
l
i
c
a
t
i
o
n
l
a
y
e
r
i
s
b
u
i
l
t
w
i
t
h
F
l
a
sk
f
r
a
me
w
o
r
k
.
T
h
e
F
l
a
sk
se
r
v
e
r
i
n
t
e
g
r
a
t
e
s
a
u
t
h
e
n
t
i
c
a
t
i
o
n
a
n
d
a
u
t
h
o
r
i
z
a
t
i
o
n
c
o
n
t
r
o
l
s
t
o
p
r
o
t
e
c
t
sy
st
e
m
r
e
so
u
r
c
e
s.
A
l
l
d
a
sh
b
o
a
r
d
i
n
t
e
r
f
a
c
e
s
a
n
d
p
r
e
d
i
c
t
i
o
n
r
o
u
t
e
s
a
r
e
a
c
c
e
ssi
b
l
e
o
n
l
y
t
o
a
u
t
h
e
n
t
i
c
a
t
e
d
u
se
r
s.
M
o
d
e
l
se
r
v
i
n
g
l
a
y
e
r
T
h
e
mac
h
i
n
e
l
e
a
r
n
i
n
g
i
n
f
e
r
e
n
c
e
s
u
b
sy
st
e
m
i
s
d
i
r
e
c
t
l
y
i
n
c
o
r
p
o
r
a
t
e
d
i
n
t
o
t
h
e
F
l
a
sk
se
r
v
e
r
t
h
r
o
u
g
h
mo
d
e
l
se
r
v
i
n
g
w
i
t
h
j
o
b
l
i
b
.
T
h
e
t
r
a
i
n
e
d
mo
d
e
l
s
a
n
d
p
r
e
p
r
o
c
e
ssi
n
g
c
o
mp
o
n
e
n
t
s a
r
e
t
r
a
i
n
e
d
o
f
f
l
i
n
e
a
n
d
sav
e
d
a
s
j
o
b
l
i
b
f
i
l
e
s.
T
a
sk
e
x
e
c
u
t
i
o
n
a
n
d
q
u
e
u
i
n
g
T
o
e
n
a
b
l
e
r
e
a
l
-
t
i
me
a
n
o
mal
y
d
e
t
e
c
t
i
o
n
i
n
a
l
i
v
e
n
e
t
w
o
r
k
e
n
v
i
r
o
n
me
n
t
,
t
h
e
sy
st
e
m
a
r
c
h
i
t
e
c
t
u
r
e
i
n
c
l
u
d
e
s
a
b
a
c
k
g
r
o
u
n
d
t
a
s
k
e
x
e
c
u
t
i
o
n
me
c
h
a
n
i
s
m
t
h
a
t
a
l
l
o
w
s
f
o
r
c
o
n
t
i
n
u
o
u
s
n
e
t
w
o
r
k
t
r
a
f
f
i
c
a
n
a
l
y
si
s
w
i
t
h
o
u
t
i
n
t
e
r
f
e
r
i
n
g
w
i
t
h
t
h
e
n
e
t
w
o
r
k
.
A
t
a
sk
q
u
e
u
e
d
e
si
g
n
a
l
l
o
w
s
f
o
r
a
s
y
n
c
h
r
o
n
o
u
s
e
x
e
c
u
t
i
o
n
o
f
l
i
v
e
t
r
a
f
f
i
c
c
o
l
l
e
c
t
i
n
g
a
n
d
a
n
o
mal
y
d
e
t
e
c
t
i
o
n
.
T
h
e
sy
st
e
m
t
a
k
e
s
q
u
e
u
e
d
t
r
a
f
f
i
c
e
v
e
n
t
s,
p
e
r
f
o
r
ms
p
r
e
p
r
o
c
e
ssi
n
g
a
n
d
d
i
me
n
s
i
o
n
a
l
i
t
y
r
e
d
u
c
t
i
o
n
,
a
n
d
r
u
n
s
t
h
e
a
n
o
m
a
l
y
d
e
t
e
c
t
i
o
n
a
n
d
c
l
a
ss
i
f
i
c
a
t
i
o
n
a
l
g
o
r
i
t
h
ms.
D
a
t
a
st
o
r
a
g
e
l
a
y
e
r
Th
e
st
o
r
a
g
e
l
a
y
e
r
i
s
i
mp
l
e
me
n
t
e
d
u
s
i
n
g
S
Q
L
i
t
e
,
w
h
i
c
h
p
r
o
v
i
d
e
s
a
l
i
g
h
t
w
e
i
g
h
t
d
a
t
a
b
a
se
so
l
u
t
i
o
n
f
o
r
st
o
r
i
n
g
o
p
e
r
a
t
i
o
n
a
l
d
e
t
e
c
t
i
o
n
o
u
t
p
u
t
s.
S
Q
L
i
t
e
w
a
s
se
l
e
c
t
e
d
d
u
e
t
o
i
t
s
si
m
p
l
i
c
i
t
y
,
m
i
n
i
m
a
l
c
o
n
f
i
g
u
r
a
t
i
o
n
r
e
q
u
i
r
e
me
n
t
s
,
a
n
d
s
u
i
t
a
b
i
l
i
t
y
f
o
r
smal
l
t
o
me
d
i
u
m
-
sc
a
l
e
d
e
p
l
o
y
me
n
t
s.
S
e
c
u
r
i
t
y
l
a
y
e
r
D
u
e
t
o
n
e
t
w
o
r
k
mo
n
i
t
o
r
i
n
g
d
a
t
a
a
n
d
a
n
o
mal
y
d
e
t
e
c
t
i
o
n
r
e
su
l
t
s
a
r
e
i
m
p
o
r
t
a
n
t
,
t
h
e
p
l
a
t
f
o
r
m
i
n
c
l
u
d
e
s
se
c
u
r
i
t
y
c
o
n
t
r
o
l
s
t
h
a
t
p
r
o
v
i
d
e
c
o
n
f
i
d
e
n
t
i
a
l
i
t
y
a
n
d
c
o
n
t
r
o
l
l
e
d
a
c
c
e
ss.
T
h
e
A
u
t
h
e
n
t
i
c
a
t
i
o
n
me
t
h
o
d
f
o
r
t
h
e
se
sy
st
e
ms
i
s
a
se
c
u
r
e
se
ssi
o
n
-
b
a
se
d
l
o
g
i
n
t
h
a
t
e
n
s
u
r
e
s
t
h
a
t
o
n
l
y
c
o
n
f
i
r
me
d
u
se
r
s h
a
v
e
a
c
c
e
ss t
o
t
h
e
sy
st
e
m.
3
.
1
0
.
L
i
m
it
a
t
io
ns
A
lt
h
o
u
g
h
t
h
is
s
t
u
d
y
ac
h
ie
v
e
d
s
tr
o
n
g
a
n
o
m
a
l
y
d
etec
tio
n
an
d
class
if
icatio
n
p
er
f
o
r
m
an
ce
,
s
ev
er
al
li
m
ita
tio
n
s
r
e
m
ai
n
.
T
h
e
d
atase
t
w
as
co
llected
f
r
o
m
a
s
i
n
g
le
n
et
w
o
r
k
e
n
v
ir
o
n
m
en
t
at
U
m
al
u
s
i,
t
h
i
s
m
ea
n
s
t
h
at
tr
af
f
ic
p
atter
n
s
,
d
ev
ice
b
eh
av
io
r
s
,
an
d
attac
k
ch
ar
ac
t
er
is
tics
m
a
y
d
if
f
er
in
o
th
er
o
r
g
an
iz
atio
n
s
,
w
h
ic
h
m
a
y
li
m
it
g
e
n
er
aliza
b
ilit
y
.
W
h
i
le
th
e
s
y
s
te
m
s
u
p
p
o
r
ts
n
ea
r
r
ea
l
-
ti
m
e
p
r
o
ce
s
s
i
n
g
,
p
er
f
o
r
m
a
n
ce
m
a
y
v
ar
y
u
n
d
er
h
ig
h
er
tr
af
f
ic
v
o
lu
m
e
s
,
d
if
f
er
en
t n
et
w
o
r
k
in
ter
f
ac
es,
o
r
r
eso
u
r
ce
-
co
n
s
tr
ai
n
ed
d
ep
lo
y
m
e
n
ts
.
4.
CO
NCLU
SI
O
N
T
h
is
s
tu
d
y
s
u
cc
es
s
f
u
l
l
y
d
esi
g
n
ed
,
i
m
p
le
m
e
n
ted
,
an
d
ev
al
u
ated
16
h
y
b
r
id
ML
m
o
d
els
f
o
r
an
o
m
al
y
d
etec
tio
n
i
n
t
h
e
U
m
al
u
s
i
n
e
t
w
o
r
k
en
v
ir
o
n
m
e
n
t,
f
o
llo
w
in
g
th
e
KDD
f
r
a
m
e
w
o
r
k
.
T
h
e
m
o
d
el
s
i
n
te
g
r
ated
u
n
s
u
p
er
v
is
ed
an
d
s
u
p
er
v
i
s
ed
lear
n
in
g
tec
h
n
iq
u
es,
le
v
er
ag
i
n
g
p
r
ep
r
o
ce
s
s
in
g
tech
n
iq
u
e
s
s
u
c
h
as
n
o
r
m
aliza
tio
n
,
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
R
ea
l
-
time
a
n
o
m
a
ly
d
etec
tio
n
s
ystem
u
s
in
g
b
est p
erfo
r
med
ma
ch
in
e
lea
r
n
in
g
mo
d
el
(
V
icto
r
Ma
th
eb
u
la
)
1213
lab
el
en
co
d
in
g
,
an
d
SMOT
E
to
a
d
d
r
ess
class
im
b
ala
n
ce
.
Am
o
n
g
th
e
test
ed
m
o
d
els,
D
B
SC
A
N
-
b
ased
an
d
L
OF
-
b
ased
h
y
b
r
id
m
o
d
els,
p
ar
ticu
lar
l
y
DB
SC
A
N+
K
N
N
an
d
DB
SC
AN+
R
F,
d
e
m
o
n
s
tr
ated
s
u
p
er
io
r
p
er
f
o
r
m
a
n
ce
,
ac
h
iev
i
n
g
n
ea
r
-
p
er
f
ec
t
ac
cu
r
ac
y
o
f
9
9
.
8
7
%
an
d
9
9
.
9
2
%,
r
esp
ec
tiv
el
y
.
T
h
es
e
f
i
n
d
in
g
s
h
i
g
h
li
g
h
t
th
e
ef
f
ec
ti
v
e
n
ess
o
f
DB
S
C
AN
co
m
b
i
n
ed
w
i
th
r
o
b
u
s
t
clas
s
if
ier
s
(
KNN,
R
F)
in
d
etec
tin
g
n
et
w
o
r
k
an
o
m
a
lies
w
it
h
h
ig
h
r
eliab
ilit
y
.
T
h
e
s
u
cc
ess
o
f
DB
S
C
A
N
-
b
as
ed
m
o
d
els
s
u
g
g
est
s
t
h
at
d
en
s
i
t
y
-
b
ased
ap
p
r
o
ac
h
es
ar
e
h
ig
h
l
y
ef
f
ec
ti
v
e
f
o
r
r
ea
l
-
w
o
r
ld
n
et
w
o
r
k
tr
a
f
f
ic
an
al
y
s
i
s
,
o
u
tp
er
f
o
r
m
in
g
tr
ad
itio
n
al
s
ta
n
d
alo
n
e
a
n
d
s
o
m
e
h
y
b
r
id
m
o
d
els
r
ep
o
r
ted
in
p
r
i
o
r
r
esear
ch
.
T
h
e
r
esu
lts
al
ig
n
w
it
h
t
h
e
ex
i
s
ti
n
g
liter
atu
r
e,
r
e
in
f
o
r
cin
g
th
e
p
o
te
n
tial o
f
h
y
b
r
id
M
L
m
o
d
el
s
i
n
c
y
b
er
s
ec
u
r
it
y
.
F
u
tu
r
e
r
esear
ch
w
o
r
k
m
a
y
f
o
c
u
s
o
n
tes
tin
g
a
n
d
v
al
id
atin
g
t
h
e
p
r
o
p
o
s
ed
m
o
d
els
f
o
r
s
ca
lab
le
d
ep
lo
y
m
e
n
t
i
n
a
cl
o
u
d
-
b
ased
en
v
ir
o
n
m
e
n
t.
T
h
is
in
cl
u
d
es
ev
al
u
ati
n
g
m
o
d
el
p
er
f
o
r
m
a
n
ce
u
n
d
er
v
ar
y
i
n
g
n
et
w
o
r
k
lo
ad
s
,
ass
es
s
i
n
g
late
n
c
y
an
d
t
h
r
o
u
g
h
p
u
t in
d
is
tr
ib
u
ted
ar
ch
itect
u
r
es.
ACK
NO
WL
E
D
G
M
E
NT
S
W
e
ac
k
n
o
w
led
g
e
th
e
s
u
p
p
o
r
ted
g
iv
e
n
to
t
h
e
r
esear
ch
er
s
b
y
t
h
e
U
m
al
u
s
i,
So
u
t
h
Af
r
ica
an
d
p
er
m
i
s
s
io
n
to
u
s
e
t
h
eir
n
e
t
w
o
r
k
d
ata
f
o
r
th
e
s
i
m
u
latio
n
asp
ec
ts
o
f
t
h
is
r
esear
c
h.
F
UNDIN
G
I
NF
O
RM
AT
I
O
N
T
h
is
w
o
r
k
w
as
s
u
p
p
o
r
ted
f
r
o
m
t
h
e
f
u
n
d
in
g
p
r
o
v
id
ed
f
o
r
p
o
s
tg
r
ad
u
ate
b
u
r
s
ar
y
f
r
o
m
th
e
No
r
th
-
W
e
s
t
Un
i
v
er
s
it
y
,
So
u
th
Af
r
ica
a
n
d
th
e
Fac
u
lt
y
o
f
Nat
u
r
al
a
n
d
Ag
r
icu
l
tu
r
al
Scie
n
ce
s
p
o
s
t
g
r
ad
u
ate
b
u
r
s
ar
y
o
f
t
h
e
No
r
th
-
W
est U
n
i
v
er
s
it
y
,
So
u
th
A
f
r
ica.
AUTHO
R
CO
NT
RIB
UT
I
O
NS ST
A
T
E
M
E
NT
T
h
is
j
o
u
r
n
al
u
s
e
s
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
i
v
id
u
al
au
th
o
r
co
n
tr
ib
u
tio
n
s
,
r
ed
u
ce
au
t
h
o
r
s
h
ip
d
is
p
u
tes,
an
d
f
ac
ilit
ate
co
lla
b
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
Victo
r
Ma
th
eb
u
la
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
B
u
k
o
h
w
o
Mic
h
ae
l
E
s
ief
ar
ie
n
r
h
e
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
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
si
s
I
:
I
n
v
e
st
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
si
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
Au
t
h
o
r
s
s
tate
n
o
co
n
f
lic
t o
f
i
n
t
er
est.
E
T
H
I
CAL AP
P
RO
V
AL
A
lt
h
o
u
g
h
n
o
an
i
m
al
o
r
h
u
m
a
n
w
as
u
s
ed
in
t
h
is
r
esear
ch
,
b
u
t
eth
ical
ap
p
r
o
v
al
w
as
o
b
tai
n
ed
f
o
r
th
e
r
esear
ch
as
it
is
p
ar
t
o
f
a
s
tu
d
y
lead
in
g
to
th
e
a
w
ar
d
o
f
a
m
as
ter
’
s
d
eg
r
ee
in
co
m
p
u
ter
s
cien
ce
an
d
in
f
o
r
m
at
io
n
s
y
s
te
m
f
r
o
m
th
e
No
r
th
-
W
est
Un
iv
er
s
i
t
y
,
So
u
t
h
A
f
r
ica.
T
h
e
Facu
lt
y
o
f
Natu
r
al
an
d
Ag
r
icu
l
tu
r
al
Scie
n
ce
E
th
ics
C
o
m
m
ittee
g
a
v
e
eth
ica
l
ap
p
r
o
v
al
as
a
“
No
r
is
k
”
r
ese
ar
ch
w
it
h
et
h
ics
ce
r
tific
ate
Nu
m
b
er
NW
U
-
01363
-
24
-
A
9
.
DATA AV
AI
L
AB
I
L
I
T
Y
T
h
e
d
ata
u
s
ed
in
th
i
s
s
tu
d
y
w
er
e
o
b
tain
ed
w
i
th
p
er
m
is
s
i
o
n
f
r
o
m
U
m
alu
s
i
C
o
u
n
cil
(
C
o
u
n
cil
f
o
r
Qu
alit
y
Ass
u
r
an
ce
in
Gen
er
a
l
an
d
F
u
r
t
h
er
E
d
u
ca
tio
n
an
d
T
r
ain
in
g
,
So
u
t
h
Af
r
ica)
an
d
all
p
er
s
o
n
al
a
n
d
id
en
ti
f
y
i
n
g
i
n
f
o
r
m
atio
n
w
er
e
an
o
n
y
m
i
s
ed
.
T
h
u
s
,
th
e
d
ata
i
s
n
o
t
p
u
b
licl
y
a
v
ailab
le,
an
d
p
er
m
i
s
s
io
n
f
o
r
u
s
e
s
h
o
u
ld
b
e
o
b
tain
ed
f
r
o
m
U
m
al
u
s
i
C
o
u
n
cil,
So
u
th
Af
r
ica.
Evaluation Warning : The document was created with Spire.PDF for Python.