T
E
L
KO
M
NIK
A
, V
ol
.
17
,
No.
6,
Dec
em
be
r
20
1
9,
p
p.
3
05
7
~
3
06
5
IS
S
N: 1
69
3
-
6
93
0
,
accr
ed
ited
F
irst
Gr
ad
e b
y K
em
en
r
istekdikti,
Decr
ee
No:
21
/E
/
K
P
T
/20
18
DOI:
10.12928/TE
LK
OM
N
IK
A
.v
1
7
i
6
.
10064
◼
30
57
Rec
ei
v
ed
M
ay
28
,
20
18
; R
ev
i
s
ed
J
un
e 1
6
, 2
01
9
;
A
c
c
ep
te
d
J
ul
y
2
, 2
0
1
9
Pr
ot
ecti
ng
big
d
ata mi
nin
g ass
o
ciatio
n rules
usin
g f
uzz
y
s
y
s
tem
G
and
ikot
a Ra
mu
*
1
,
M
S
o
u
mya
2
,
A
p
p
aw
ala
Ja
ya
n
t
h
i
3
, J
.
S
o
mas
ek
a
r
4
, K.
K. B
a
se
er
5
1,
3
Dep
a
rt
m
e
n
t
o
f
Co
m
p
u
te
r S
c
i
e
n
c
e
&
En
g
i
n
e
e
r
i
n
g
,
In
s
ti
tu
te
o
f
Ae
r
o
n
a
u
ti
c
a
l
E
n
g
i
n
e
e
ri
n
g
,
T
e
l
a
n
g
a
n
a
5
0
0
0
4
3
,
I
n
d
i
a
2
Dep
a
rt
m
e
n
t
o
f
Co
m
p
u
t
e
r Sc
i
e
n
c
e
a
n
d
E
n
g
i
n
e
e
ri
n
g
,
Sri
n
i
v
a
s
a
Ra
m
a
n
u
j
a
n
I
n
s
t
i
tu
t
e
o
f
T
e
c
h
n
o
l
o
g
y
,
An
a
n
ta
p
u
r
5
1
5
0
0
1
,
I
n
d
i
a
4
Dep
a
rt
m
e
n
t
o
f
Co
m
p
u
t
e
r Sc
i
e
n
c
e
a
n
d
E
n
g
i
n
e
e
ri
n
g
,
G
o
p
a
l
a
n
Co
l
l
e
g
e
o
f
En
g
i
n
e
e
ri
n
g
a
n
d
M
a
n
a
g
e
m
e
n
t
,
Ba
n
g
a
l
o
r
e
5
6
0
0
4
8
,
In
d
i
a
5
Sre
e
Vi
d
y
a
n
i
k
e
th
a
n
En
g
i
n
e
e
ri
n
g
Co
l
l
e
g
e
,
T
i
ru
p
a
t
i
5
1
7
1
0
2
,
I
n
d
i
a
*C
o
rre
s
p
o
n
d
i
n
g
a
u
th
o
r,
e
m
a
i
l
:
g
.ra
m
u
c
s
e
@g
m
a
i
l
.c
o
m
Ab
strac
t
Rec
e
n
tl
y
,
b
i
g
d
a
t
a
i
s
g
ra
n
te
d
to
b
e
th
e
s
o
l
u
ti
o
n
to
o
p
e
n
i
n
g
th
e
s
u
b
s
e
q
u
e
n
t
l
a
rg
e
fl
u
c
tu
a
t
i
o
n
s
o
f
i
n
c
re
a
s
e
i
n
f
e
rti
l
i
t
y
.
Al
o
n
g
wit
h
th
e
g
ro
wth
,
i
t
i
s
fa
c
i
n
g
s
o
m
e
o
f
th
e
c
h
a
l
l
e
n
g
e
s
.
O
n
e
o
f
th
e
s
i
g
n
i
fi
c
a
n
t
p
ro
b
l
e
m
s
i
s
d
a
t
a
s
e
c
u
ri
t
y
.
W
h
i
l
e
p
e
o
p
l
e
u
s
e
d
a
ta
m
i
n
i
n
g
m
e
t
h
o
d
s
to
i
d
e
n
t
i
fy
v
a
l
u
a
b
l
e
i
n
fo
r
m
a
ti
o
n
fo
l
l
o
win
g
m
a
s
s
i
v
e
d
a
ta
b
a
s
e
,
p
e
o
p
l
e
fu
rt
h
e
r
h
o
l
d
th
e
n
e
c
e
s
s
a
ry
to
m
a
i
n
ta
i
n
a
n
y
k
n
o
wle
d
g
e
s
o
whi
l
e
n
o
t
t
o
b
e
wo
rk
e
d
o
u
t,
l
i
k
e
d
e
l
i
c
a
te
c
o
m
m
o
n
i
te
m
s
e
ts
,
p
ra
c
ti
c
e
s
,
ta
x
o
n
o
m
y
tr
e
e
a
n
d
th
e
l
i
k
e
As
s
o
c
i
a
ti
o
n
r
u
l
e
m
i
n
i
n
g
c
a
n
m
a
k
e
a
p
o
s
s
i
b
l
e
warn
i
n
g
a
p
p
r
o
a
c
h
i
n
g
th
e
s
e
c
re
c
y
o
f
i
n
fo
rm
a
ti
o
n
.
So
,
a
s
s
o
c
i
a
ti
o
n
ru
l
e
h
i
d
i
n
g
m
e
th
o
d
s
a
re
a
p
p
l
i
e
d
t
o
e
v
a
d
e
t
h
e
h
a
z
a
rd
o
f
d
e
l
i
c
a
te
i
n
fo
rm
a
ti
o
n
m
i
s
u
s
e
.
V
a
ri
o
u
s
k
i
n
d
s
o
f
i
n
v
e
s
t
i
g
a
t
i
o
n
a
l
re
a
d
y
p
r
e
p
a
r
e
d
o
n
a
s
s
o
c
i
a
ti
o
n
r
u
l
e
p
ro
te
c
t
i
n
g
.
Howe
v
e
r,
m
a
x
i
m
u
m
o
f
t
h
e
m
c
o
n
c
e
n
tr
a
te
o
n
i
n
tr
o
d
u
c
i
n
g
m
e
th
o
d
s
wi
th
a
l
i
m
i
te
d
v
i
e
w
o
u
tc
o
m
e
fo
r
i
n
a
c
ti
v
e
d
a
t
a
b
a
s
e
s
(wi
th
o
n
l
y
e
x
i
s
ti
n
g
i
n
f
o
rm
a
ti
o
n
)
,
whi
l
e
p
r
e
s
e
n
tl
y
th
e
re
s
e
a
rc
h
e
rs
fa
c
i
n
g
th
e
p
r
o
b
l
e
m
wit
h
c
o
n
t
i
n
u
o
u
s
i
n
fo
rm
a
ti
o
n
.
M
o
re
o
v
e
r,
i
n
th
e
e
r
a
o
f
b
i
g
d
a
ta
,
th
i
s
i
s
e
s
s
e
n
ti
a
l
to
o
p
ti
m
i
z
e
c
u
rre
n
t
s
y
s
te
m
s
to
b
e
s
u
i
te
d
c
o
n
c
e
rn
i
n
g
th
e
b
i
g
d
a
ta
.
T
h
i
s
p
a
p
e
r
p
ro
p
o
s
e
s
th
e
fra
m
e
wor
k
i
s
a
c
h
i
e
v
i
n
g
t
h
e
d
a
ta
a
n
o
n
y
m
i
z
a
ti
o
n
b
y
u
s
i
n
g
fu
z
z
y
l
o
g
i
c
b
y
s
u
p
p
o
rt
i
n
g
b
i
g
d
a
ta
m
i
n
i
n
g
.
Th
e
f
u
z
z
y
l
o
g
i
c
g
ro
u
p
i
n
g
th
e
s
e
n
s
i
ti
v
i
ty
o
f
th
e
a
s
s
o
c
i
a
ti
o
n
ru
l
e
s
wit
h
a
s
u
i
ta
b
l
e
a
s
s
o
c
i
a
t
i
o
n
l
e
v
e
l
.
M
o
re
o
v
e
r
,
p
a
ra
l
l
e
l
i
z
a
ti
o
n
m
e
th
o
d
s
wh
i
c
h
a
re
i
n
s
e
rt
e
d
i
n
t
h
e
p
re
s
e
n
t
fra
m
e
work
wi
l
l
s
u
p
p
o
rt
f
a
s
t
d
a
ta
m
i
n
i
n
g
p
ro
c
e
s
s
.
Key
w
ords
:
a
s
s
o
c
i
a
ti
o
n
ru
l
e
s
,
b
i
g
d
a
ta
,
d
a
ta
m
i
n
i
n
g
,
fu
z
z
y
l
o
g
i
c
Copy
righ
t
©
2
0
1
9
Uni
v
e
rsi
t
a
s
Ahm
a
d
D
a
hl
a
n.
All
rig
ht
s
r
e
s
e
rve
d
.
1.
Int
r
o
d
u
ctio
n
W
i
th
a
v
arie
t
y
of
r
ec
en
t
te
c
hn
ol
og
i
es
c
om
bi
ne
d
i
n
ou
r
r
eg
ul
ar
l
i
v
es
,
l
i
k
e
s
m
artpho
ne
s
,
s
oc
i
al
m
ed
i
a
an
d
I
nte
r
ne
t
of
T
hi
ng
(
IoT
)
ba
s
ed
i
nt
el
l
i
ge
nt
-
w
or
l
d
prac
ti
c
es
l
i
k
e
c
l
ev
er
term
i
na
l
,
tr
us
tworth
y
tr
an
s
p
ort,
l
i
v
el
y
to
w
n,
an
d
oth
ers
ge
n
erati
ng
hu
ge
i
nf
orm
ati
on
[1
-
7].
T
he
m
ul
ti
p
l
e
k
i
nd
s
of
el
ec
tr
on
i
c
ap
pl
i
a
n
c
es
c
r
ea
te
m
as
s
i
v
e
i
nf
orm
ati
o
n
c
ea
s
e
l
es
s
l
y
of
ev
er
y
c
ha
r
ac
ters
an
d
area.
T
he
r
ef
ore,
s
i
ng
l
e,
c
o
m
pl
ete
,
an
d
c
om
pl
ex
da
ta,
es
p
e
c
i
a
l
l
y
en
orm
ou
s
i
nf
orm
ati
on
,
be
c
om
es
a
l
ot
of
v
a
l
u
e.
M
oreov
er,
i
nc
l
u
di
n
g
th
e
i
m
prov
em
en
t
of
i
nf
orm
ati
on
an
al
y
s
i
s
prod
uc
ed
b
y
arti
f
i
c
i
a
l
i
nt
el
l
i
ge
nc
e
an
d
i
nf
or
m
ati
on
proc
es
s
i
ng
t
ec
hn
i
qu
es
,
i
nc
l
u
di
n
g
th
eref
ore
the
ev
al
ua
t
i
ng
ab
i
l
i
t
i
es
he
l
p
ed
b
y
i
nte
r
ne
t
a
l
s
o
po
i
nt
c
a
l
c
ul
a
ti
n
g
s
u
pp
ort,
th
e
po
s
s
i
bl
e
be
ne
f
i
ts
of
the
c
r
ea
t
ed
ex
t
en
s
i
v
e
k
no
w
l
e
dg
e
gro
w
a
l
ot
of
dram
ati
c
[8
-
14
].
T
he
r
ef
ore,
b
i
g
d
ata
i
s
the
p
urpos
e
of
th
i
s
m
ee
ti
ng
f
l
ow
s
of
f
erti
l
i
t
y
i
nc
r
ea
s
e.
B
es
i
d
es
,
t
he
s
af
et
y
i
s
s
ue
s
i
n
i
nf
or
m
ati
on
proc
es
s
i
ng
m
eth
o
ds
,
c
urr
en
t
s
af
et
y
t
es
ts
m
us
t
de
v
el
op
i
n
to
m
as
s
i
v
e
i
nf
or
m
ati
on
proc
es
s
i
ng
,
t
ha
t
s
qu
are
m
ea
s
ure
r
e
ga
r
d
i
n
g
th
e
ne
e
d
of
pa
r
al
l
e
l
v
i
c
ti
m
i
z
ati
on
proc
es
s
i
ng
f
or
s
ub
s
tan
t
i
a
l
i
nf
o
r
m
ati
on
r
ev
i
e
w
[1
5].
T
he
r
ef
ore,
s
ec
r
ec
y
i
s
s
ue
s
s
qu
are
ac
t
i
on
ag
grav
a
ted
as
a
r
es
ul
t
o
f
d
i
s
t
r
i
bu
te
d
da
ta
m
a
y
be
r
ec
o
v
ered
m
erel
y
i
ns
tea
d
of
m
as
s
k
i
nd
.
G
r
ou
p
prac
ti
c
e
o
pe
n
i
ng
i
s
un
i
t
y
i
n
e
v
er
y
of
tha
t
f
orem
os
t
n
ec
es
s
ar
y
da
t
a
proc
es
s
i
ng
tec
h
ni
q
ue
s
.
Ho
wev
er,
m
i
s
us
e
o
f
thi
s
m
eth
od
c
ou
l
d
r
es
ul
t
i
n
the
r
e
v
e
l
ati
o
n
of
de
l
i
c
a
te
i
nf
orm
ati
on
r
eg
ardi
ng
pe
r
s
on
s
[1
6,
17
]
.
S
e
v
era
l
t
y
p
es
of
r
es
ea
r
c
h
are
wor
n
ou
t
a
s
s
oc
i
ati
o
n
r
u
l
e
ac
ti
v
i
t
y
[18
-
22
]
a
l
s
o
m
os
t
i
m
po
r
tan
t
of
tho
s
e
s
ha
r
ed
m
ea
ns
i
t
s
ep
arate
th
i
ng
s
f
r
om
do
i
ng
f
or
ex
erc
i
s
e
s
en
s
i
bl
e
l
a
w
s
.
Un
ha
pp
i
l
y
,
of
f
ered
f
ea
tures
i
n
f
l
ue
nc
e
i
s
e
v
i
de
nt
i
n
tho
s
e
m
eth
od
s
.
T
o
ex
pl
a
i
n
t
ha
t
do
wns
i
d
e,
p
eo
pl
es
wor
k
an
d
do
d
y
n
am
i
c
wa
y
s
.
B
ut,
th
os
e
p
l
an
s
do
no
t
g
ua
r
a
nte
e
to
f
i
nd
the
as
s
oc
i
ate
b
es
t
an
s
wer
al
s
o
s
ol
el
y
w
ork
an
d
i
m
prov
e
th
e
po
t
en
c
y
.
Dur
i
ng
th
at
an
a
l
y
s
i
s
,
t
o
c
o
v
er
f
i
ne
c
o
m
m
un
i
t
y
prac
t
i
c
es
i
n
to
m
as
s
i
v
e
i
nf
orm
ati
on
pr
oc
es
s
i
ng
,
r
at
he
r
th
an
Evaluation Warning : The document was created with Spire.PDF for Python.
◼
IS
S
N: 16
93
-
6
93
0
T
E
L
KO
M
NIK
A
V
ol
.
17
,
No
.
6,
D
ec
em
be
r
20
19
:
305
7
-
306
5
3058
pu
s
hi
ng
a
pe
r
e
nn
i
al
c
as
e
of
de
l
i
c
at
e
c
om
m
un
i
t
y
c
ou
r
s
es
,
an
o
n
y
m
i
z
a
ti
o
n
s
tr
ate
g
i
es
s
qu
ar
e
m
ea
s
ure
w
on
t
to
prot
ec
t
d
el
i
c
a
te
c
o
ntrol
s
.
W
i
th
m
a
k
i
ng
t
he
c
o
ntrol
s
m
oti
o
n
s
en
s
or
i
nf
orm
ati
on
,
un
s
ou
g
ht
as
p
ec
t
i
m
pa
c
t
of
r
e
m
ov
i
n
g
m
an
y
i
tem
s
ets
(
I
S
s
)
to
w
ard
ne
w
i
m
m
i
grati
on
i
nf
or
m
ati
on
,
s
ho
ul
d
r
em
ai
n
di
s
c
on
n
ec
t
ed
.
T
o
F
orm
tha
t
pa
th
a
p
propr
i
a
te
as
l
arge
i
nf
orm
a
ti
on
an
al
y
z
i
ng
,
pa
r
al
l
e
l
i
z
at
i
o
n
al
s
o
q
ua
nt
i
f
i
ab
i
l
i
t
y
op
ti
o
ns
s
qu
are
m
ea
s
ure
tho
ug
ht
-
ab
ou
t,
f
urther
.
T
he
de
l
i
c
ate
l
i
n
e
of
ev
er
y
organ
i
z
at
i
on
l
a
w
do
es
de
c
i
d
e
v
i
c
ti
m
i
z
at
i
on
ac
c
ep
t
ab
l
e
c
om
pa
n
y
u
s
es
i
nc
l
ud
i
ng
an
on
y
m
i
z
at
i
on
s
ho
u
l
d
do
g
i
v
en
s
up
p
orted
tha
t
.
2.
Rel
ated
W
o
r
k
2.1
.
Big
Data
Rega
r
d
i
n
g
o
utl
i
n
e,
t
he
ex
t
en
s
i
v
e
i
nf
orm
ati
on
r
e
l
ate
s
to
th
e
v
as
t
am
ou
nt
of
s
tr
u
c
ture,
s
e
m
i
-
s
tr
uc
ture
an
d
u
ns
tr
u
c
ture
i
nf
or
m
ati
on
w
i
t
h
a
s
pe
c
i
al
c
ha
r
g
e
t
ha
t
m
a
y
do
w
e
l
l
-
m
i
ne
as
i
nf
orm
ati
on
[16
]
.
Ma
s
s
i
v
e d
ata
proc
es
s
i
n
g
po
i
nts
on
th
i
s
p
ote
n
ti
a
l
f
r
o
m
ob
ta
i
ni
n
g d
ata
of
h
ug
e
de
ta
i
l
s
th
i
s
be
c
au
s
e
of
s
pe
c
i
al
o
pti
on
s
no
t
d
o
gi
v
e
v
i
c
ti
m
i
z
ati
on
b
ei
n
g
i
nf
orm
ati
on
proc
e
s
s
i
ng
s
y
s
t
em
s
[23
].
W
hi
l
e
s
ev
e
r
al
th
i
ng
s
,
i
t
i
s
i
m
po
s
s
i
bl
e
to
pu
t
tha
t
B
r
o
bd
i
ng
na
gi
a
n
qu
an
t
i
t
y
of
i
nf
or
m
ati
on
,
the
r
ef
ore,
t
he
d
ata
ex
tr
ac
ti
o
n
o
ug
h
t
t
o
be
do
ne
r
e
al
ti
m
e.
P
r
o
c
es
s
m
as
s
i
v
e
i
nf
orm
ati
on
want
s
the
gr
ou
p
r
eg
ard
i
ng
s
y
s
t
em
s
w
i
t
h
po
w
erf
ul
e
v
a
l
ua
ti
n
g
prod
uc
ti
on
i
nc
l
ud
i
ng
the
s
tr
uc
ture
wi
l
l
r
em
ai
n
s
en
s
i
b
l
e
b
y
i
d
en
t
i
c
al
p
r
og
r
am
m
i
ng
s
tan
da
r
ds
a
do
r
e
b
i
g
da
ta
tec
hn
i
qu
e
[2
4
].
2.2
.
A
n
o
n
y
m
it
y
Data
di
s
tr
i
b
uti
on
s
om
eti
m
es
do
es
b
y
th
i
s
c
ha
nc
e
f
r
o
m
r
aw
i
nf
orm
ati
on
r
ev
e
l
at
i
on
[2
5].
K
no
wl
e
dg
e
s
om
eti
m
es
i
nc
l
ud
es
r
a
w
i
nf
orm
ati
on
,
i
nc
l
ud
i
ng
tha
t
s
h
o
w
s
th
at
ef
f
ec
t
of
us
i
ng
ob
s
c
urit
y
m
eth
od
s
[
25
,
26
]
.
T
he
s
e
l
as
t
thre
e
m
eth
od
s
t
o
a
no
n
y
m
i
z
at
i
on
th
at
em
bo
d
y
ge
ne
r
al
i
z
e,
de
s
tr
uc
ti
o
n,
i
nc
l
ud
i
n
g
orga
ni
z
at
i
on
.
S
e
v
era
l
ap
proac
h
es
to
an
o
n
y
m
i
z
at
i
o
n
c
he
r
i
s
h
k
an
on
y
m
i
t
y
,
di
f
f
erenc
e,
c
l
os
en
es
s
,
etc
.
prac
ti
c
e
tho
s
e
m
eth
od
s
.
In
c
on
c
l
ud
e
,
us
es
ab
o
ut
propert
i
es
are
r
ep
l
ac
e
d
b
y
an
ad
di
t
i
on
al
g
en
era
l
on
e
[2
6].
M
a
y
b
e,
wh
i
l
e
tha
t
wor
th
of
qu
al
i
t
y
‘
t
i
m
e’
m
ea
ns
ab
l
e
s
i
x
tee
n,
t
ha
t
m
a
y
m
ea
n
r
e
ne
w
e
d
b
y
ac
c
ep
t
ab
l
e
v
ar
y
c
he
r
i
s
h
te
n
to
t
went
y
.
S
u
p
pre
s
s
i
on
r
ef
ers
to
pre
v
e
nt
c
a
tha
r
t
i
c
th
at
tr
ue
wor
th
f
r
om
as
s
oc
i
ate
de
gre
e
pro
pe
r
t
y
.
Dur
i
n
g
t
he
m
ea
ns
,
th
e
prev
a
l
e
nc
e
of
thi
s
w
orth
m
ea
ns
f
ol
l
o
wed
b
y
t
he
s
y
s
te
m
c
he
r
i
s
h
‘
*
,’
an
d
t
he
s
ug
ge
s
ts
thi
s
on
e
c
on
ten
t
m
a
y
ac
t
s
u
bs
ti
tu
t
ed
r
at
he
r
[
27
].
M
a
y
b
e,
w
hi
l
e
t
hi
s
c
o
nn
ec
te
d
wor
th
f
r
o
m
as
s
oc
i
ate
de
gre
e
prop
ert
y
m
ea
ns
c
ap
ab
l
e
f
i
f
t
y
-
s
i
x
,
f
ou
r
hu
nd
r
e
d
ni
n
et
y
-
s
e
v
e
n,
th
at
m
a
y
m
ea
n
f
ol
l
o
w
ed
b
y
56
49
*
.
T
he
O
r
ga
n
i
s
at
i
on
l
e
ad
s
to
th
i
s
ex
c
ha
n
ge
f
r
om
orig
i
na
l
c
on
t
en
t
b
y
c
h
an
c
e
wor
th.
D
urin
g
the
s
y
s
t
em
,
s
ou
nd
do
es
m
ore
to
k
no
w,
s
o
thi
s
m
ate
r
i
al
qu
al
i
t
y
f
r
om
properti
es
i
s
c
ov
ert
[28
]
.
W
hi
l
e
T
ab
l
e
1
,
thre
e
s
e
v
er
al
wel
l
-
l
i
k
ed
a
no
n
y
m
i
z
at
i
on
s
s
y
s
tem
s
area
un
i
ts
de
l
i
ne
ate
d.
B
ec
a
us
e
of
the
no
v
e
l
op
t
i
on
s
of
ex
te
ns
i
v
e
k
no
wl
e
dg
e
c
h
eris
h
hi
gh
am
ou
nt
p
l
us
s
e
l
ec
ti
on
i
nto
k
no
w
l
e
dg
e
bu
i
l
d
i
ng
s
,
n
ec
es
s
ar
y
c
ha
ng
es
ou
gh
t
to
do
t
hi
nk
of
w
h
i
l
e
c
on
s
i
de
r
e
d
wa
y
s
i
n
to
s
ati
s
f
y
i
n
g
r
e
l
at
ed
r
eq
ui
r
em
en
ts
.
Dur
i
ng
th
e
de
s
i
g
n,
th
e
ge
ne
r
a
l
s
y
s
te
m
m
ea
ns
e
m
pl
o
y
e
d
to
ob
s
c
urit
y
,
w
h
ereas
el
i
m
i
na
t
i
on
s
y
s
tem
i
s
n't
ap
pro
pria
t
e
to
am
ou
nt
k
no
w
l
ed
g
e
i
nc
l
u
di
ng
org
an
i
z
ati
on
s
y
s
t
em
i
m
po
s
es
th
e e
s
s
en
t
i
al
c
os
t
on
c
om
pu
ters
.
T
ab
l
e 1
.
A
n
on
y
m
i
s
ati
on
S
c
he
m
es
A
n
o
n
y
mi
s
a
t
ion
s
c
h
e
me
I
d
e
a
D
r
a
w
b
a
c
k
k
-
a
n
o
n
y
m
i
t
y
e
a
c
h
a
t
t
r
ibu
t
e
i
s
u
n
iqu
e
o
f
mi
n
i
m
u
m
(
k
−1
)
r
e
c
e
n
t
ly
a
t
t
r
ibu
t
e
s
.
This
init
iat
iv
e
lev
e
r
a
g
e
s
t
h
e
f
a
c
t
a
n
y
w
h
e
r
e
a
ll
t
h
e
a
d
v
a
n
t
a
g
e
s
f
o
r
a
d
e
li
c
a
t
e
v
a
lue
ins
ide
a
s
e
t
o
f
k
s
t
o
r
ie
s
a
r
e
s
a
m
e
.
l
-
d
iv
e
r
s
it
y
e
v
e
r
y
g
r
o
u
p
o
f
a
t
t
r
ibu
t
e
s
inc
lud
e
s
mi
n
i
m
u
m
o
n
e
p
r
o
p
e
r
ly
-
r
e
p
r
e
s
e
n
t
e
d
u
t
il
i
t
y
f
o
r
t
h
e
d
e
li
c
a
t
e
p
r
o
p
e
r
t
y
L
-
d
iv
e
r
s
it
y
m
a
y
b
e
t
r
y
ing
t
o
b
e
a
c
c
o
mpli
s
h
e
d
t
-
c
lo
s
e
n
e
s
s
t
h
e
s
p
r
e
a
d
o
f
d
e
li
c
a
t
e
p
r
o
p
e
r
t
ie
s
in
s
p
e
c
if
i
c
s
u
b
-
c
la
s
s
o
f
w
o
r
k
s
a
n
d
t
h
e
c
e
n
t
r
a
l
d
a
t
a
s
e
t
is
les
s
t
h
a
n
t
h
r
e
s
h
o
ld
t
L
o
w
d
a
t
a
u
t
il
it
y
2.3
.
A
s
soci
atio
n
Ru
le
H
idin
g
Com
m
un
i
t
y
prac
t
i
c
e
op
e
ni
n
g
i
s
a
n
un
us
u
al
r
oa
d
to
att
em
pt
to
es
c
ap
e
f
orei
g
n
r
el
at
i
on
s
h
i
ps
am
on
g
v
a
l
u
e
s
i
nto
th
e
l
arge
da
t
ab
as
e
[18
],
b
ut,
a
bu
s
e
of
the
s
e
s
y
s
t
em
s
f
orc
e
c
on
di
t
i
o
n
l
a
ng
ua
g
e
pe
r
f
orm
an
c
e
f
r
o
m
de
l
i
c
ate
i
nf
orm
ati
on
[29
-
30
].
T
hu
s
,
m
an
y
pe
op
l
e
s
erv
ed
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NIK
A
IS
S
N: 1
69
3
-
6
93
0
◼
P
r
ote
c
t
i
ng
bi
g d
a
ta
mi
n
i
n
g
as
s
oc
i
ati
on
r
u
l
es
us
i
ng
f
u
z
z
y
s
y
s
tem
(
G
an
di
k
ot
a Ra
mu
)
3059
to
w
ard
c
o
v
eri
ng
d
el
i
c
ate
or
ga
n
i
z
ati
on
prac
t
i
c
es
.
T
hi
s
greate
s
t
h
op
e
f
or
c
o
m
m
un
i
t
y
go
v
ernm
en
t
protec
ti
ng
m
eth
od
s
m
ea
ns
i
n
c
as
e
of
r
aw
prac
ti
c
es
,
i
n
c
l
ud
i
ng
n
o
f
ea
tures
ap
pe
ar
of
no
de
l
i
c
ate
prac
ti
c
es
.
Ha
i
et
al
.
[20
]
propos
e
d
he
uris
ti
c
s
be
c
au
s
e
s
up
p
ort
i
nc
l
u
di
n
g
s
up
po
r
t
i
nt
eres
t
s
up
po
r
te
d
c
r
os
s
i
ng
s
tr
uc
ture
(
HCSRI
L)
pa
tte
r
n
be
i
ng
t
he
he
uris
ti
c
s
pa
t
h
i
n
m
ee
ti
ng
th
i
s
group
f
r
o
m
s
oc
i
et
y
us
es
f
r
om
the
r
el
e
v
an
t
da
t
ab
as
e
i
nt
o
the
l
oc
al
tr
a
de
.
T
hi
s
m
ax
i
m
u
m
l
ev
e
l
s
ab
o
ut
r
es
ea
r
c
he
r
s
propos
ed
wi
l
l
m
ai
nta
i
n
th
i
s
c
om
m
un
i
t
y
t
h
ereb
y
gi
v
i
ng
s
om
e
thi
ng
thi
s
r
es
ea
r
c
he
r
's
c
ha
ng
es
b
ec
om
e
the
l
i
m
i
ted
i
m
pa
c
t
at
s
ev
er
al
m
an
y
I
S
s
,
w
i
l
l
l
i
k
e
tha
t
l
e
as
t
l
i
m
i
t
a
bo
ut
do
i
ng
t
hi
s
ou
t
i
nt
o
l
as
t
m
od
i
f
i
ed
al
s
o
m
urder
i
ng
of
f
erin
g
i
nf
orm
at
i
on
of
be
f
ore
-
m
en
ti
on
e
d
b
ec
au
s
e
do
i
ng
.
T
hrough
thi
s
r
es
ea
r
c
h,
p
r
od
uc
ti
o
n
org
an
i
z
a
ti
o
n
f
r
om
v
ario
us
I
S
s
r
em
ai
ns
prepared.
T
hi
s
pro
du
c
ti
on
po
s
i
ti
on
m
a
k
es
tha
t
l
i
m
i
ted
i
nf
l
u
en
c
e
at
no
n
-
d
el
i
c
ate
IS
s
du
r
i
ng
l
i
tt
l
e
go
v
ernm
en
t
s
c
r
ee
n.
G
eo
r
g
i
nc
l
u
di
ng
V
a
s
s
i
l
i
o
[3
1]
s
t
ate
d
M
ax
-
Mi
n
2
de
v
i
c
e
i
nc
l
u
di
ng
ap
pl
i
ed
M
ax
-
Mi
n
the
or
y
i
nto
c
om
m
un
i
t
y
prac
ti
c
e
c
o
nc
ea
l
ed
.
T
hi
s
gre
ate
s
t
f
orm
of
the
th
eo
r
y
m
ea
ns
i
nto
m
a
x
i
m
i
z
i
n
g
th
i
s
l
ea
s
t
i
nc
r
ea
s
e.
W
hi
l
e
be
i
n
g,
pe
o
pl
e
c
on
t
i
nu
e
w
ork
i
ng
i
nt
o
m
ax
i
m
i
z
i
ng
f
i
n
e
c
on
tr
ol
c
on
c
ea
l
e
d
where
as
on
s
am
e
c
on
d
i
ti
on
s
r
e
du
c
e
th
i
s
r
eg
ard
ef
f
ec
t
to
w
ard
n
o
-
de
l
ec
ta
te
c
om
m
an
ds
.
T
he
de
s
i
gn
protec
ts
d
el
i
c
at
e rel
a
ti
o
ns
hi
p h
ab
i
ts
b
y
r
ed
uc
i
ng
tha
t
he
l
p
ab
o
ut
f
i
ne
I
S
s
.
In
S
h
y
u
-
L
i
an
ge
t
al
.
pr
op
os
ed
de
s
i
g
n
[
32
],
p
ai
r
s
y
s
t
e
m
s
r
e
m
ai
n
c
om
m
on
s
urf
ac
e
n
e
w
r
el
at
i
on
s
h
i
p
c
om
m
an
ds
.
P
eo
p
l
e
di
d
r
ed
o
ub
l
e
c
are
of
l
ef
tw
ard
-
ha
n
d
v
i
e
w
(
IS
L)
an
d
de
c
r
e
as
ed
he
l
p
on
ou
t
w
ard
-
h
an
d
v
i
e
w
(
DS
R)
i
nto
r
ea
l
i
z
i
ng
r
es
e
a
r
c
he
r
s
pl
an
.
W
hi
l
e
Chi
n
g
-
Y
o
et
al
.
s
tu
d
y
,
on
l
y
ex
i
s
t
i
ng
pl
a
y
s
l
i
v
e
de
s
c
r
i
be
d
i
ns
i
de
t
hi
s
m
eth
od
l
i
k
e
the
pa
i
r
ed
m
od
el
.
W
h
en
tha
t
m
od
el
,
whi
l
e
s
i
ng
l
e
pi
ec
e
I
en
g
ag
e
i
nt
o
d
ea
l
j
,
Di
j
l
as
t
wor
k
i
ng
i
nt
o
b
ei
ng
on
e;
un
l
es
s
,
t
ha
t
i
s
e
no
u
gh
on
no
th
i
ng
.
W
he
n
he
l
pe
d
th
o
s
e
de
s
c
r
i
b
ed
t
hres
ho
l
ds
of
ex
i
s
te
nc
e
t
hroug
h
t
he
pra
c
ti
c
e,
m
od
el
X
wi
l
l
r
em
ai
n
c
l
os
e
d,
s
o
t
ha
t
P
ʹ=
X
*
P
.
W
hi
l
e
t
ha
t
d
es
c
r
i
p
ti
on
,
P
i
n
di
c
at
es
th
i
s
on
e
f
or
m
c
on
c
erned
tha
t
m
ax
i
m
u
m
i
nf
o,
X
do
es
tha
t
pr
ote
c
ti
ng
m
od
el
al
s
o
X
ʹ
do
es
s
om
e
pa
tt
ern
c
o
m
pa
r
ed
on
th
i
s
priv
ate
d
ata
ba
s
e
.
E
l
en
et
al
[3
3]
s
tud
i
ed
-
on
c
on
c
e
al
e
d
f
r
o
m
al
l
f
i
ne
c
om
m
u
ni
t
y
prac
ti
c
es
i
nc
l
u
di
ng
v
ario
us
IS
s
.
B
ot
h
r
ec
e
i
v
ed
3
c
h
an
n
el
s
as
tha
t
go
al
:
i
m
prov
i
ng
tha
t
p
r
ov
i
s
i
on
ab
ou
t
LHS
,
r
e
du
c
i
ng
th
at
pa
y
m
e
nt
f
r
o
m
RHS
i
nc
l
u
di
n
g
c
he
c
k
i
ng
s
om
e
ad
v
i
c
e
l
i
k
e
LHS
an
d
RHS
,
i
n
thi
s
eq
u
al
o
pp
ortun
i
t
y
.
W
hi
l
e
tha
t
r
es
ea
r
c
h
pro
po
s
ed
ex
am
pl
e,
an
on
y
m
i
z
at
i
on
m
eth
o
ds
r
e
m
ai
n
c
on
v
e
nti
on
a
l
s
k
i
n
r
aw
i
nf
orm
ati
on
.
T
he
n,
i
n
or
i
g
i
na
l
,
de
l
i
c
ate
i
nc
l
ud
i
ng
n
um
be
r
propert
i
es
l
i
k
e
hand
-
s
e
l
ec
te
d
I
S
s
l
as
t
r
ai
s
ed
.
W
he
n,
i
n
hi
di
ng
un
p
l
e
as
an
t
r
e
l
at
i
o
ns
am
on
g
v
ari
ou
s
IS
s
,
m
an
y
propert
i
es
r
e
an
o
n
y
m
i
z
e
d
i
nt
o
s
om
e
proper
s
ta
ge
.
W
hi
l
e
m
an
y
r
ep
orts
,
no
r
eg
u
l
ar
I
S
s
w
i
l
l
r
em
ai
n o
f
i
nf
or
m
ati
on
,
i
nc
l
u
di
n
g i
nd
i
v
i
du
al
l
y
r
a
w
pr
i
c
es
wi
l
l
b
e d
r
o
pp
e
d.
3.
P
r
o
p
o
se
d
Fram
ew
o
r
k to
Hide M
inin
g
Ru
les
in Big
Dat
a E
n
v
ir
o
n
men
t
T
he
propos
e
d
s
ec
ure
d
f
r
am
ew
ork
to
hi
d
e
m
i
ni
n
g
r
u
l
e
s
i
n
bi
g
da
ta
e
nv
i
r
on
m
en
t
i
nv
o
l
v
e
d
three
m
od
ul
es
na
m
el
y
1
)
as
s
oc
i
ati
on
r
u
l
e
m
i
ni
ng
,
2
)
c
om
pu
te
c
on
f
i
de
nc
e
of
e
ac
h
r
ul
e
,
an
d
3
)
f
u
z
z
y
l
og
i
c
s
y
s
tem
as
s
ho
w
n
i
n
F
i
gu
r
e
1.
H
ere,
th
es
e
thr
ee
m
od
ul
es
s
ho
ul
d
be
f
un
c
ti
on
i
ng
pa
r
al
l
e
l
l
y
,
s
o
th
i
s
f
r
a
m
ew
or
k
i
s
s
ui
tab
l
e
f
or
bi
g
da
ta
ap
pl
i
c
a
ti
o
ns
.
A
l
s
o,
th
e
en
orm
ou
s
f
ea
tures
of
bi
g
d
ata
l
i
k
e
v
e
l
oc
i
t
y
a
nd
v
ol
um
e
ge
ne
r
a
te
s
da
t
a
c
on
ti
n
uo
us
l
y
s
o
ex
i
s
ti
n
g
prop
os
ed
m
eth
od
s
no
t f
i
t f
or bi
g d
a
ta
m
i
ni
n
g
e
nv
i
r
on
m
en
t.
3.1.
A
s
soci
atio
n
Ru
le
M
inin
g
In
th
e
f
i
r
s
t
m
od
ul
e,
v
ari
ou
s
Ite
m
S
ets
(
IS
s
)
are
f
ou
nd
us
i
ng
d
i
f
f
erent
ex
tr
ac
ti
ng
m
eth
od
s
.
B
es
i
d
es
,
c
om
pl
ete
ex
tr
em
e
prac
ti
c
es
of
m
an
y
are
b
ei
ng
do
n
e.
In
th
i
s
f
r
a
m
ew
ork
,
the
as
s
i
gn
ed
tr
us
t
ou
ts
et
(
α)
,
ot
he
r
arb
i
tr
ar
y
r
es
ol
uti
on
l
e
v
el
s
c
an
b
e
ex
am
i
ne
d
an
d
prac
ti
c
es
wi
th
s
e
ns
i
ti
v
i
t
y
he
r
e
α
d
oe
s
n't
be
r
ai
s
e
d
i
m
m
ed
i
ate
l
y
a
nd
s
h
ou
l
d
c
on
t
i
n
ue
f
or
w
ard
w
i
th
the
ad
d
i
t
i
on
al
i
n
v
es
ti
ga
t
i
on
.
A
s
an
i
ns
ta
n
c
e,
s
tud
y
α
m
ea
ns
eq
ui
v
a
l
en
t
n
ea
r
l
y
s
i
x
t
y
pe
r
c
en
t
ag
e
c
on
tr
ol
s
wi
th
th
e
r
es
o
l
ut
i
o
n
e
qu
i
v
a
l
e
nt
ne
arl
y
f
i
f
t
y
-
s
ev
en
pe
r
c
en
t
ag
e
are
de
l
i
c
ate
,
a
l
s
o,
i
n
c
l
ud
i
ng
s
h
ou
l
d
r
em
ai
n
c
ov
ere
d,
s
i
m
pl
y
b
y
s
ev
eral
s
tag
es
.
Nex
t,
th
e
ad
v
a
nta
g
e
of
thi
s
ob
s
c
ur
e
m
eth
od
f
or
c
he
c
k
i
ng
da
ta
l
ea
k
ag
e
of
v
er
y
s
ub
t
l
e
r
el
ati
on
s
h
i
p
r
ul
e
s
,
the
s
e
s
om
ew
ha
t
s
en
s
i
bl
e
l
a
w
s
c
an
be
ad
ap
ted
to
d
el
i
c
ate
c
ou
r
s
e
s
wi
th
t
he
en
tr
y
of
orig
i
n
al
i
nf
or
m
ati
on
i
n
h
i
gh
d
ata
c
urr
en
t.
T
he
r
ef
ore,
ba
s
ed
o
n
the
d
ete
r
m
i
ne
d
s
tat
e
of
as
s
oc
i
ati
o
n
di
c
t
ate
s
,
proper
as
s
oc
i
at
i
o
n
l
e
v
e
l
s
are
att
ac
he
d
to
areas
an
d a
r
e s
t
ored
de
p
e
nd
ed
on
tho
s
e c
om
pa
n
y
s
t
an
da
r
ds
.
3.2.
Co
mp
u
t
e
Co
n
f
iden
ce
of
E
ac
h
Ru
le
W
e
de
f
i
ne
d
f
ou
r
m
e
m
be
r
s
h
i
p
f
un
c
ti
o
ns
(
V
_l
o
w
,
Lo
w,
Hi
gh
,
an
d
V
_H
i
g
h)
to
c
ha
r
ge
a
grou
p
l
e
v
el
t
o
ea
c
h
as
s
oc
i
at
i
on
r
ul
e
as
s
ho
wn
i
n
T
ab
l
e
2.
A
f
ter
tha
t,
a
on
e
b
y
f
ou
r
m
atri
c
es
r
el
ate
d
to
the
c
om
pu
ted
c
o
nf
i
de
nc
e
of
ea
c
h
g
ov
ernm
en
t.
E
v
er
y
c
om
po
ne
nt
of
the
m
od
e
l
d
es
c
r
i
be
s
Evaluation Warning : The document was created with Spire.PDF for Python.
◼
IS
S
N: 16
93
-
6
93
0
T
E
L
KO
M
NIK
A
V
ol
.
17
,
No
.
6,
D
ec
em
be
r
20
19
:
305
7
-
306
5
3060
the
as
s
oc
i
at
i
o
n
l
e
v
e
l
of
tha
t
c
ou
r
s
e
to
al
l
c
om
pa
n
y
us
e
.
If
the
s
pe
c
i
f
i
ed
r
es
ol
uti
on
do
or
i
s
s
i
m
i
l
ar
to
α,
c
om
pa
n
y
pa
r
t
i
es
v
es
s
el
m
ov
es
r
ep
r
es
en
te
d
s
i
nc
e
A
pp
en
d
i
x
2.
Dur
i
ng
t
he
r
ec
ord,
tha
t
ne
x
t
l
i
n
e,
‘
Mi
ni
m
u
m
,’
r
ep
r
es
en
t
s
tha
t
s
m
al
l
es
t
am
ou
nt
ab
ou
t
a
n
y
s
oc
i
et
y
us
e,
whi
l
e
the
three
c
ol
um
ns
,
‘
Ma
x
,’
r
ep
r
es
e
nts
the
hi
g
he
s
t
v
a
l
u
e
of
s
oc
i
e
t
y
us
es
.
P
r
ac
ti
c
es
wi
th
r
es
ol
u
ti
on
un
d
er
C
1
c
an
be
a
v
o
i
de
d.
I
t
s
ho
ul
d
be
r
em
ar
k
ed
tha
t
C
i
(
i
=
1,
2,
…,5)
d
oe
s
a
c
on
ne
c
ti
on
i
n
f
or
m
ed
v
al
ue
s
an
d
p
os
teri
or
l
as
t
r
ep
l
ac
ed
.
Nex
t
r
ep
r
es
en
t
i
ng
as
s
oc
i
at
i
on
l
e
v
el
c
om
pa
r
ed
on
e
v
er
y
l
i
m
i
ted
us
e
,
tha
t
s
oc
i
et
y
c
a
pa
c
i
t
y
am
on
g
th
os
e
m
os
t
s
i
gn
i
f
i
c
an
t
a
s
s
oc
i
ati
o
n
l
e
v
e
l
r
em
ai
ns
p
i
c
k
ed
be
c
a
us
e
tha
t
r
ea
l
p
ers
on
,
i
nc
l
ud
i
ng
t
hi
s
c
o
v
et
ed
s
t
ag
e
f
r
o
m
hi
d
i
ng
,
r
em
ai
ns
de
term
i
ne
d
us
i
ng
as
s
oc
i
ati
on
of
f
i
c
e.
T
hi
s
m
us
t
i
m
pl
y
s
tat
e
th
i
s
w
h
i
l
e
th
i
s
c
om
pa
n
y
area
f
r
om
c
ou
pl
e
c
ha
r
ac
t
eris
ti
c
s
i
s
e
qu
a
l
,
s
oc
i
et
y
c
ap
ac
i
t
y
i
nc
l
ud
i
ng
t
ho
s
e
n
ec
es
s
ar
y
h
i
d
i
ng
s
t
ag
es
i
s
c
ho
s
en
.
F
i
gu
r
e
1.
S
ec
ured
f
r
a
m
ew
o
r
k
to
hi
d
e m
i
ni
ng
r
ul
es
i
n
bi
g d
at
a e
nv
i
r
on
m
en
t
T
ab
l
e 2
.
Me
m
be
r
s
hi
p
F
u
nc
ti
on
V
a
l
u
es
Ran
ge
s
R
a
n
g
e
Fr
o
m
To
V
e
r
y
_
h
igh
α
100
H
igh
C5
α
Low
C3
C4
V
e
r
y
_
low
C1
C2
3.3.
F
u
z
z
y
L
o
g
ic
S
y
ste
m
In
t
he
V
i
e
w
x
-
>
y
as
a
n
as
s
oc
i
ati
on
r
ul
e
,
h
ere,
ea
c
h
of
x
an
d
y
are
c
ol
l
ec
ti
on
s
of
propert
i
es
.
P
r
op
ert
i
es
c
an
be
group
ed
i
nt
o
3
c
l
as
s
es
.
F
i
r
s
t
on
e
,
i
de
n
ti
f
i
er
propert
i
es
are
c
ha
r
ac
teri
s
ti
c
s
i
nc
l
ud
i
ng
k
no
w
i
n
g
k
no
wl
ed
ge
l
i
k
e
as
c
om
m
on
ag
r
ee
m
en
t
es
ti
m
a
te.
T
he
s
ec
on
d
on
e,
da
i
nt
y
prop
erti
es
are
i
nc
ub
ate
d
of
pro
pe
r
ti
es
th
at
r
ec
e
i
v
e
i
nd
i
v
i
d
ua
l
r
et
i
r
e
m
en
t
da
ta
an
d
s
ho
ul
d
be
pres
erv
ed
.
T
he
l
as
t
o
ne
,
qu
as
i
-
i
de
nt
i
f
i
er
(
Q
I)
propert
i
es
ho
l
d
prop
ert
i
es
t
ha
t
m
a
k
e
ne
ga
ti
v
e
i
nc
l
u
de
na
m
i
ng
pr
op
erti
es
,
w
i
t
ho
ut
pris
o
n
b
e
c
on
ne
c
te
d
to
a
dd
i
ti
on
a
l
d
ata
to
produc
e
c
r
ed
en
ti
al
s
ex
po
s
ure
[8
].
H
en
c
e,
the
c
orr
ec
t
am
ou
nt
o
f
ten
de
r
i
nc
l
ud
i
ng
i
de
nti
f
i
c
at
i
on
pro
pe
r
ti
es
m
u
s
t
ho
l
d
as
s
as
s
i
na
te
d
i
n
c
l
ud
i
ng
Q
I
propert
i
es
m
us
t
r
em
ai
n
g
en
er
al
i
z
ed
us
i
n
g
th
i
s
de
ta
i
l
ed
group
of
f
i
c
e.
T
o
m
i
ni
m
i
z
e
t
he
un
d
es
i
r
e
s
h
ap
e
i
m
pa
c
t
l
i
k
e
i
nf
or
m
ati
on
an
o
n
y
m
i
z
at
i
on
,
un
i
t
y
a
bo
ut
l
ef
t
-
ha
nd
s
i
de
or
Ri
gh
t
-
H
an
d
S
i
de
ab
ou
t
t
ha
t
c
ou
r
s
es
m
u
s
t
l
as
t
c
ho
s
en
an
o
n
y
m
i
z
ed
.
O
n
m
a
k
i
ng
tha
t,
grant
ed
h
i
d
i
n
g s
tag
e i
s
f
i
ni
s
he
d
i
nc
l
ud
i
ng
a d
ee
p
er und
es
i
r
e s
ha
pe
r
es
u
l
t.
a.
Choi
c
e
of
a
v
al
i
d
da
ta
s
ho
u
l
d b
e a
no
n
y
m
i
z
ed
W
hi
l
e
an
ag
r
ee
m
en
t
c
om
m
an
d,
w
i
t
h
th
e
an
on
y
m
i
z
i
ng
s
i
n
gl
e
v
i
e
w
,
i
ts
ba
c
k
s
i
de
be
s
ug
ge
s
te
d
th
at
no
ten
de
r
m
e
s
s
ag
e
c
ou
l
d
b
e
de
l
i
v
ere
d.
S
o,
i
nc
l
ud
i
ng
s
i
n
gl
e
s
i
d
e
an
o
n
y
m
i
z
a
ti
o
n,
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NIK
A
IS
S
N: 1
69
3
-
6
93
0
◼
P
r
ote
c
t
i
ng
bi
g d
a
ta
mi
n
i
n
g
as
s
oc
i
ati
on
r
u
l
es
us
i
ng
f
u
z
z
y
s
y
s
tem
(
G
an
di
k
ot
a Ra
mu
)
3061
c
o
m
m
un
i
t
y
prac
t
i
c
es
c
ou
l
d
r
e
m
ai
n
s
tored
i
n
the
s
m
al
l
un
de
s
i
r
e
v
i
e
w
o
utc
om
e.
W
hi
l
e
tha
t
wa
y
,
s
o
m
e
c
en
tr
al
d
i
f
f
i
c
ul
t
y
m
ea
ns
i
nt
o
d
i
s
c
ov
er
i
ng
t
he
ge
nu
i
n
e
arti
c
l
e
f
or
an
on
y
m
i
z
at
i
on
.
W
hi
l
e
the
p
os
s
i
bl
e
i
tem
s
et,
c
ov
eri
ng
th
e
d
el
i
c
ate
r
el
a
ti
o
ns
hi
p
c
om
m
an
d
po
s
s
i
b
l
e
be
g
et
un
de
s
i
r
e
v
i
e
w
be
ari
ng
t
o
w
ard
c
om
m
un
i
t
y
pra
c
ti
c
es
.
I
n
a
r
o
bu
s
t
i
te
m
s
et,
tha
t
t
hi
ng
m
y
be
s
t
produc
e
r
es
ul
ts
ap
pro
ac
hi
n
g
un
l
i
k
e
i
nc
om
i
ng
i
nf
orm
ati
on
,
be
s
i
de
s
.
O
v
er
thi
s
r
ate
c
ha
r
ac
teri
s
ti
c
a
bo
ut
i
m
po
r
tan
t
da
ta,
a
c
o
l
l
ec
ti
o
n
a
bo
u
t
th
at
m
os
t
s
i
gn
i
f
i
c
an
t
da
t
a
d
urin
g
an
on
y
m
i
z
ati
on
m
us
t
do
pe
r
f
or
m
ed
us
i
ng
2 c
on
s
t
i
tue
nts
:
-
Unde
s
i
r
e a
n
gl
e res
ul
t o
f
an
on
y
m
i
z
at
i
o
n o
f
an
oth
er c
urr
en
t u
nd
e
l
i
c
a
te
c
om
m
un
i
t
y
p
r
ac
ti
c
es
.
-
A
n u
nd
es
i
r
e
a
ng
l
e res
u
l
t o
n
an
o
n
y
m
i
z
at
i
o
n i
n t
h
e p
os
s
i
bl
e s
e
pa
r
at
e a
c
c
es
s
i
nf
orm
ati
o
n.
T
he
us
ua
l
s
tr
ate
g
y
i
s
t
o
m
i
ni
m
i
z
e
t
he
s
e
c
i
r
c
um
s
tan
c
es
as
m
uc
h
as
f
ea
s
i
bl
e.
T
hi
nk
tha
t
a
c
om
m
an
d
be
f
ore
-
m
en
ti
on
ed
as
X
-
>
Z
s
h
ou
l
d
r
e
m
ai
n
s
ep
arat
ed
al
s
o
r
e
qu
i
r
e
t
o
di
s
c
o
v
er
a
g
en
u
i
ne
th
i
n
g
as
an
on
y
m
i
z
at
i
on
.
D
urin
g
th
at,
the
i
nf
l
ue
nc
e
a
bo
ut
an
y
R.
H.S
or
L
.H.
S
p
art
an
on
y
m
i
z
at
i
on
s
h
ou
l
d
do
j
ud
ge
d
us
i
n
g
the
t
wo
s
pe
c
i
f
i
ed
ag
e
nts
al
s
o
w
h
en
a
n
ob
j
ec
t
b
y
t
ha
t
l
i
g
hte
r
v
i
e
w
r
es
ul
t
de
c
i
de
d.
O
n
th
e
he
a
d,
c
o
nn
ec
t
i
on
c
o
m
m
an
ds
r
em
ai
n
c
l
as
s
i
f
i
e
d
ba
s
e
d
up
o
n
pe
op
l
e
p
os
i
t
i
on
po
w
er
a
l
s
o
w
h
en
th
os
e
c
i
r
c
um
s
tan
c
es
do
j
ud
ge
to
t
he
r
ea
l
p
art
c
ol
l
ec
t
i
on
.
A
s
tha
t
i
ni
ti
a
l
-
m
en
ti
on
ed
p
orti
o
n,
w
e
a
na
l
y
z
e
thi
s
i
nf
orm
a
ti
on
i
nt
o
an
y
po
s
s
i
b
l
e
s
y
s
t
em
(
ex
ternal
l
y
f
r
es
h
i
nf
or
m
ati
on
c
om
i
ng
)
.
T
he
n,
thi
s
da
ta
n
ee
d
t
ha
t
m
ea
ns
produc
e
d
at
t
ha
t
an
on
y
m
i
z
at
i
on
v
es
s
el
l
as
t m
ea
s
ured i
nc
l
u
di
n
g t
h
e m
eth
od
(
I)
, d
o b
e
i
n
g a
m
od
el
of
g
en
u
i
ne
i
t
em
s
)
el
ec
ti
on
.
Ite
m
S
et
(
IS
)
=
1/
/
N
i
(N
1
W
1
+
N
2
W
2
… N
k
W
k
)
k
∈
m
(
1
)
A
s
s
um
e
w
e
ne
ed
t
o
drop
r
ul
e
x
-
>
y
whi
c
h
i
s
att
ac
h
ed
to
grou
p
f
un
c
ti
o
n
‘
b
i
g
’
.
In
(
1
)
,
Ni
ho
l
ds
thi
s
am
ou
nt
of
c
on
tr
o
l
s
i
nc
l
ud
i
ng
tha
t
s
i
m
i
l
ar
as
s
oc
i
at
i
on
pu
r
p
os
e
(
a
l
s
o
th
at
c
orr
es
po
nd
i
ng
s
tag
e
f
r
om
ge
ne
r
al
i
z
a
ti
o
n)
tha
t
X
i
nc
l
ud
e
d
w
i
thi
n,
N
k
do
es
tha
t
am
ou
nt
f
r
om
c
on
tr
ol
s
th
at
X
l
i
nk
ed
a
ga
i
ns
t, t
ho
u
gh
i
nc
l
u
di
n
g m
an
y
gro
up
ga
t
he
r
i
n
g
s
, a
nd
W
j
ho
l
ds
s
om
e d
ata
ne
ed
i
nf
l
ue
nc
e
am
on
g
m
an
y
a
no
n
y
m
i
z
at
i
o
n
s
tag
es
.
A
s
,
w
h
i
l
e
X
i
s
a
n
on
y
m
i
z
e
d
to
s
t
ag
e
s
i
m
i
l
ar
be
f
ore
‘
gre
at,
’
al
s
o
t
hi
s
m
ea
ns
a
l
a
w
tha
t
X
l
i
nk
ed
i
nt
o,
j
us
t
i
nc
l
u
di
ng
a
no
n
y
m
i
z
at
i
on
s
tag
e
l
i
k
e
on
‘
m
ea
ns
,’
l
ea
r
n
i
n
g
n
ee
d
po
wer
m
ea
ns
eq
u
i
v
al
e
nt
on
on
e
,
o
nl
y
t
o
s
tag
e
‘
m
i
ni
m
u
m
,’
tha
t
d
o
es
l
i
k
e
i
nto
t
w
o.
F
urther s
pe
c
i
f
i
c
a
l
l
y
, t
ha
t
po
wer
pa
c
k
ag
e i
s
t
he
e
qu
i
v
al
en
t
l
en
gth
of
two
an
on
y
m
i
z
ati
o
n l
ev
el
s
.
T
hi
s
po
r
ti
o
n
h
as
a
po
s
s
i
b
l
e
s
en
s
e
of
th
e
i
nf
or
m
ati
on
c
en
tr
e
an
d
pl
an
s
t
he
l
ea
r
ni
n
g
en
d
s
tag
e.
A
s
s
i
ng
l
e
an
ot
he
r
pa
r
t,
j
us
t
th
at
c
on
tr
as
t
wi
th
i
n
t
ha
t
r
e
l
i
gi
ou
s
f
orc
es
of
di
f
f
erent
c
o
m
m
an
ds
the
da
ta
i
nc
l
u
d
ed
wi
th
i
n
t
he
p
l
us
es
ta
bl
i
s
he
d
be
l
i
ef
orig
i
n
ab
ou
t
t
ha
t
s
pe
c
i
f
i
c
group
c
ap
ac
i
t
y
i
s
as
s
es
s
ed
. T
ha
t
po
r
ti
o
n i
s
c
a
l
c
ul
ate
d u
s
i
ng
(
2
)
.
Di
f
f
erenc
e o
f
Con
f
i
de
nc
e
(
DO
C)
=
√
(
1
−
)
2
+
(
2
−
)
2
+
⋯
+
(
−
)
2
i
∈
n
(
2
)
Dur
i
n
g
th
i
s
c
ou
r
s
e
x
-
>
y
p
l
us
i
nc
l
u
di
ng
group
of
f
i
c
e,
‘
m
ax
i
m
u
m
,’
k
i
w
i
t
hi
n
m
eth
od
(
II)
c
o
m
pris
es
i
nd
i
v
i
du
al
d
ete
r
m
i
na
ti
on
c
on
di
t
i
on
f
r
om
s
om
e.
A
no
the
r
un
i
t
tha
t
X
m
ea
ns
i
nc
l
ud
ed
wi
th
i
n,
Cj
r
em
ai
ns
s
i
m
i
l
ar
o
v
er
C5
s
t
ate
,
i
nc
l
ud
i
ng
n
e
qu
a
l
s
s
om
e
a
m
ou
nt
l
i
k
e
c
om
m
an
ds
tha
t
X
do
es
i
nv
ol
v
e.
T
ho
s
e
nu
m
be
r
s
m
us
t
c
on
ti
nu
e
r
e
turne
d
as
Y
,
a
l
s
o.
Dur
i
ng
a
no
t
he
r
ne
w
s
,
th
at
pa
r
t
es
ti
m
ate
s
tha
t
l
i
k
el
i
ho
od
of
ea
c
h
de
v
e
l
o
pm
en
t
i
nt
o
f
u
l
l
c
om
pa
n
y
of
f
i
c
e.
O
ne
r
es
ul
t
l
i
k
e
ne
w
l
y
r
ec
orded
i
nf
orm
ati
on
e
qu
a
l
s
r
ed
uc
i
n
g
t
hi
s
r
es
o
l
ut
i
on
ad
v
an
t
ag
e
ab
ou
t
r
el
a
ti
on
s
hi
p
l
a
w
s
ho
ul
d
r
em
ai
n s
ho
r
ter s
pe
c
i
f
i
e
d s
t
art
w
i
t
hi
n p
r
i
v
ate
grou
p o
f
f
i
c
e.
T
hu
s
, th
e
c
on
tr
as
t
wi
t
hi
n
tr
us
t b
en
ef
i
t o
f
m
e
m
be
r
s
hi
p
r
u
l
es
an
d
th
e
s
m
al
l
es
t
r
es
o
l
ut
i
o
n
ad
v
an
t
a
ge
of
f
ul
l
s
oc
i
et
y
us
e
m
us
t
do
m
ea
s
ured.
W
hi
l
e
tha
t
c
os
t
do
es
a
l
s
o,
t
he
po
s
s
i
bi
l
i
t
y
a
bo
ut
c
o
nv
ert
i
ng
s
a
i
d
gro
up
of
f
i
c
e
m
ov
es
s
m
al
l
er.
T
ha
t
i
m
pl
i
es
ev
i
d
en
t
t
ha
t
as
an
o
n
y
m
i
z
at
i
on
i
s
pe
r
f
orm
ed
us
i
ng
th
i
s
s
i
m
pl
i
f
i
ed
gro
up
pu
r
po
s
e,
us
n
ee
d
i
nto
r
ed
uc
e
th
at
p
os
s
i
b
i
l
i
t
y
.
La
s
tl
y
,
s
e
l
ec
t
da
ta
c
ol
l
ec
ti
o
n
w
i
l
l
m
a
k
e
w
i
th
m
i
x
i
ng
the
s
e
ef
f
ec
ts
f
r
o
m
IS
i
nc
l
ud
i
ng
an
i
nte
r
v
al
ab
o
ut
r
es
ol
ut
i
o
n
c
on
d
i
ti
on
s
,
j
us
t
i
nc
l
u
di
ng
s
ui
t
ab
l
e
prac
t
i
c
a
l
i
m
po
r
tan
c
e,
be
c
au
s
e (
3
).
B
es
t i
tem
s
et
v
al
ue
=
µ1
*
I
S
+
µ2
*
DO
C
(
3
)
A
n
i
tem
w
i
th
f
e
w
r
ea
l
i
te
m
-
s
et
us
es
c
an
be
c
h
os
e
n
as
th
e
m
os
t
s
i
gn
i
f
i
c
an
t
t
hi
n
g
f
or
an
on
y
m
i
z
at
i
on
.
In
(
3
)
,
µ
1
a
nd
µ
2
are
us
ef
ul
m
ea
s
urem
en
ts
i
nc
l
ud
i
ng
th
i
s
b
en
ef
i
ts
pa
c
k
ag
e
are
i
nc
r
ea
s
ed
.
T
hi
s
i
m
pl
i
es
ev
i
de
nt
f
ol
l
o
wi
ng
an
i
tem
s
et,
whi
c
h
i
ni
ti
a
l
p
i
ec
e
m
ea
ns
c
on
ne
c
te
d
i
n
to
be
i
ng
i
nf
orm
at
i
on
,
o
nl
y
s
o
m
e
di
ff
erent
c
ha
r
ac
ter
m
ea
ns
s
i
m
i
l
ar
be
f
ore
l
o
ok
i
ng
ex
tr
a
ac
c
es
s
i
nf
orm
ati
on
.
T
hu
s
,
thi
s
a
pp
ea
r
s
th
i
s
I
S
po
r
t
i
o
n
a
l
s
o
m
ov
es
s
i
gn
i
f
i
c
an
t,
as
tha
t
po
r
ti
on
Evaluation Warning : The document was created with Spire.PDF for Python.
◼
IS
S
N: 16
93
-
6
93
0
T
E
L
KO
M
NIK
A
V
ol
.
17
,
No
.
6,
D
ec
em
be
r
20
19
:
305
7
-
306
5
3062
c
on
c
en
tr
ate
s
up
on
be
i
n
g
r
e
l
at
i
on
s
h
i
p
d
i
c
tat
es
,
w
he
r
e
a
s
DO
C
a
ge
nt
m
ov
es
d
on
e
i
nto
ob
t
ai
n
i
n
g
tha
t re
ad
i
ng
al
s
o re
as
on
ab
l
e t
o
po
te
nti
al
s
u
bs
eq
u
en
t
i
n
f
or
m
ati
on
.
b.
Q
I a
ttrib
ute
s
h
i
d
i
n
g
A
s
di
s
c
us
s
ed
ea
r
l
i
er,
i
m
po
r
tan
t
pro
bl
em
s
du
r
i
ng
c
on
ne
c
t
i
on
c
o
ntrol
protec
ti
ng
are
un
-
w
i
s
h
ed
v
i
e
w
i
m
pa
c
ts
ab
ou
t
r
ed
uc
i
ng
m
an
y
IS
's
.
D
urin
g
t
ha
t
s
tud
y
,
g
en
era
l
i
z
a
ti
on
m
e
tho
d
i
s
ap
p
l
i
e
d
to
an
on
y
m
i
z
e
Q
I
p
r
op
erti
es
at
t
he
pr
op
er
s
ta
ge
a
l
s
o
us
i
ng
de
t
ai
l
ed
gro
up
us
e.
T
hu
s
,
an
i
ni
ti
a
l
s
ta
ge
,
area
ge
ne
r
al
i
z
a
ti
o
n g
ov
ernm
en
t o
f
f
ea
tures
s
ha
l
l
s
ta
y
orga
ni
z
ed
,
as
pres
en
t
ed
i
n
F
i
gu
r
e
2.
F
or
i
ns
ta
nc
e,
t
hi
nk
ge
ne
r
a
l
i
z
at
i
on
m
ea
ns
a
bo
ut
‘
ag
e
’
.
If
‘
ag
e’
i
m
pl
i
es
r
eg
arde
d
b
ei
ng
the
l
i
g
ht
-
de
l
i
c
a
te
qu
a
l
i
t
y
,
a
nd
th
i
s
am
ou
nt
d
oe
s
ac
c
or
d
o
v
er
3
4,
a
ge
ne
r
a
l
i
z
at
i
on
f
r
o
m
the
pa
r
t
be
f
ore
30
-
3
5
r
em
ai
ns
th
e
d
ec
en
t
c
h
an
g
e.
L
i
k
e
thi
s
f
ee
l
i
n
g
a
bo
u
t
‘
ag
e
’
pr
og
r
es
s
,
g
r
ea
ter
s
tag
es
i
nto
a
hi
erar
c
h
i
c
al
ho
us
e
(
ad
j
ac
en
t
t
o
s
ou
r
c
e)
r
em
ai
n
c
ou
nte
d
t
o
w
ard
tha
t
en
d.
T
he
r
e
are
t
w
o
t
y
p
es
of
properti
es
:
b
i
na
r
y
an
d
a
bs
ol
ute
.
F
or
the
ge
n
e
r
al
i
z
at
i
o
n
of
bi
n
ar
y
propert
i
es
,
us
i
ng
s
om
e
l
i
m
i
ted
c
om
pa
n
y
of
f
i
c
es
,
proper
s
ub
s
ets
f
r
om
v
al
i
d
ati
n
g
area
f
r
om
an
y
propert
y
p
ac
k
ag
e
r
e
m
ai
n
he
l
d
at
t
he
s
e
v
ario
us
s
ta
ge
s
of
hi
di
n
g
(
be
i
ng
de
s
c
r
i
b
e
d
i
n
F
i
g
ure
2).
It
s
ho
ul
d
be
r
em
ar
k
ed
tha
t
tha
t
t
hi
n
g
ha
s
n
o
b
ea
r
i
ng
o
n
the
prin
c
i
pl
e
of
thi
s
m
eth
od
.
In
an
o
the
r
wor
d,
wi
th
t
he
c
om
bi
na
ti
on
of
bi
na
r
y
a
nd
c
erta
i
n
c
ha
r
ac
teri
s
ti
c
s
ge
ne
r
a
l
i
z
a
ti
on
,
the
c
on
c
ern
s
tag
e
of
hi
di
n
g
ac
hi
ev
es
.
T
he
r
ec
o
m
m
en
de
d m
ea
s
ures
c
orr
el
at
i
on
prac
ti
c
es
h
i
d
i
ng
i
s
de
m
on
s
tr
ate
d i
n
A
l
g
or
i
thm
1.
F
i
gu
r
e
2
.
D
om
ai
n g
en
eral
i
s
ati
o
n h
i
erar
c
h
y
of
ag
e
A
l
g
orit
hm
1:
A
s
s
oc
i
ati
o
n R
ul
e
Hi
d
i
n
g
Inp
ut
: Dat
a I
t
em
s
O
utp
ut:
A
ttr
i
bu
te
G
e
ne
r
a
l
i
z
ati
o
n
1.
B
eg
i
n
2.
S
tat
us
=
T
r
ue
3.
W
hi
l
e(Sta
tus
)
a.
If
ne
w da
t
a i
tem
r
ec
ei
v
ed
i.
S
tat
us
=
Fa
l
s
e
4.
Mi
n
i
n
g A
s
s
oc
i
a
ti
o
n ru
l
e a
nd
Com
pu
te
Conf
i
d
en
c
e
V
a
l
u
e
5.
If
Con
f
i
de
nc
e
v
a
l
ue
gi
v
en
r
an
ge
a.
T
he
n g
oto
s
t
ep
6
b.
E
l
s
e
go
to
s
tep
4
6.
Def
i
ne
a
pp
r
o
pria
t
e a
no
n
y
m
i
t
y
l
e
v
e
l
7.
S
el
ec
ti
o
n o
f
be
s
t
i
tem
f
or ano
n
y
m
i
z
ati
on
8.
G
en
eral
i
z
e
att
r
i
b
ute
9.
E
nd
4.
P
er
f
o
r
m
ance
E
v
o
lut
ion
s
T
he
propos
ed
f
r
a
m
ew
ork
i
s
ev
a
l
ua
ted
b
as
ed
o
n
ex
pe
r
i
m
en
tal
r
es
ul
ts
m
atc
he
d
a
c
ou
p
l
e
of
ex
i
s
ti
ng
m
eth
od
s
H
CS
RI
L p
l
us
Max
-
M
i
n2
.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NIK
A
IS
S
N: 1
69
3
-
6
93
0
◼
P
r
ote
c
t
i
ng
bi
g d
a
ta
mi
n
i
n
g
as
s
oc
i
ati
on
r
u
l
es
us
i
ng
f
u
z
z
y
s
y
s
tem
(
G
an
di
k
ot
a Ra
mu
)
3063
4.1.
Dat
as
et
D
es
c
r
ipt
ion
In
the
ex
pe
r
i
m
en
tal
an
a
l
y
s
i
s
,
w
e
us
ed
t
w
o
da
t
a
s
ets
n
am
e
l
y
B
r
i
j
s
an
d
C
l
ue
web
d
ata
s
ets
.
B
r
i
j
s
_
da
t
as
et:
T
hi
s
da
tas
et
i
nc
l
ud
es
s
up
erm
ar
k
et
bo
x
i
nf
or
m
ati
on
f
r
om
a
B
e
l
g
i
an
l
oc
a
l
s
up
ers
tore.
Inf
orm
ati
on
w
a
s
r
ec
ei
v
e
d
du
r
i
ng
19
99
-
20
00
.
It
i
nc
l
ud
es
e
i
g
ht
y
-
ei
g
ht
tho
us
a
nd
o
ne
hu
nd
r
ed
an
d
s
i
x
t
y
-
t
w
o
s
al
e
s
i
nc
l
ud
i
ng
s
i
x
tee
n
t
ho
us
a
n
d
f
ou
r
h
un
dr
ed
an
d
s
i
x
t
y
-
ni
ne
c
om
m
od
i
t
y
i
ds
.
E
v
er
y
wor
k
i
nt
o
orig
i
n
a
l
i
tem
s
et
i
nc
l
ud
es
da
ta
l
i
k
e
tr
an
s
ac
t
i
on
d
ate
,
qu
an
t
i
t
y
,
i
t
em
,
etc
.
B
ut
ea
c
h
c
en
tr
e
r
em
ai
ns
ex
c
l
us
i
v
el
y
to
war
d
c
l
i
e
nt
p
l
us
s
i
m
i
l
ar
thi
n
gs
.
Cl
ue
_W
eb
_d
at
as
et:
T
ha
t
da
tas
et
i
nc
l
u
de
s
hu
g
e
nu
m
be
r
s
of
w
e
b
p
ag
es
w
h
i
c
h
wer
e
c
o
l
l
ec
ted
du
r
i
ng
J
an
an
d
F
eb
20
09
.
Us
prac
ti
c
ed
s
om
e f
r
o
m
Cl
ue
Ne
t
w
ork
i
t i
nc
l
ud
es
f
i
f
t
y
-
t
hree
b
i
l
l
i
on
E
n
gl
i
s
h p
a
ge
s
.
4.2.
E
xpe
r
imen
t
P
r
o
c
es
s
In
ex
p
erim
en
tal
r
es
ul
ts
we
c
on
s
i
de
r
e
d
t
hree
m
etri
c
s
n
am
el
y
l
os
t,
g
ho
s
t
an
d
f
al
s
e
r
ul
es
.
B
as
ed
on
thr
ee
r
u
l
es
we
c
o
m
pa
r
ed
o
ur
ex
p
erim
en
tal
r
es
u
l
ts
us
i
ng
HC
S
RI
L
an
d
Ma
x
_
Mi
n2
m
eth
od
s
.
T
o
ha
v
e
e
qu
al
s
t
atu
s
f
or
an
y
r
es
ea
r
c
h,
we
s
ho
u
l
d
ex
am
i
ne
t
he
c
e
ntral
da
tas
et
as
50
K
i
nc
l
u
di
ng
da
tas
e
t
T
1
order
en
ter
i
nt
o
s
ub
s
e
qu
e
nt
ac
t
i
o
n.
W
i
th
r
eg
ardi
n
g
i
nf
orm
ati
on
b
ei
n
g
wi
th
i
n
a
s
i
n
gl
e
pr
i
m
ar
y
i
tem
s
et,
s
el
ec
t
da
ta
to
a
no
n
y
m
i
z
at
i
on
wi
l
l
s
e
l
ec
t.
Nex
t,
ne
x
t
c
o
m
bi
ni
ng
T
1
i
nf
orm
ati
on
,
the
s
ec
ti
on
a
bo
ut
f
ai
l
ed
c
ou
r
s
es
w
i
l
l
d
ec
i
de
.
S
i
m
i
l
ar
e
v
en
t
s
oc
c
ur
di
s
pl
a
y
e
d
i
n
F
i
gu
r
e
s
3
a
n
d
4.
A
s
the
po
r
ti
on
of
the
d
i
s
ab
l
ed
c
o
ntro
l
s
r
ed
uc
es
the
d
ev
el
op
m
en
t
i
nto
µ
1,
tha
t
pa
c
k
ag
e
i
s
f
ou
nd
thi
s
r
es
ul
t
ab
ou
t
µ1
m
ea
ns
es
s
en
ti
a
l
tha
t
µ2.
T
hi
s
i
s
ob
v
i
o
us
wi
th
i
m
prov
i
ng
the
us
e
of
µ1, t
h
at
b
en
ef
i
t
o
f
µ2 wi
l
l
l
i
m
i
t
as
s
ho
w
n
i
n F
i
gu
r
e 3
.
F
i
gu
r
e
3
.
%
l
os
t ru
l
es
ge
ne
r
ate
d
b
y
the
i
ntrod
uc
ed
f
r
am
ew
ork
by
m
od
i
f
y
i
n
g
µ
1/µ
2 u
s
i
ng
B
r
i
j
s
_d
ata
s
et
an
d
F
i
gu
r
e
4
.
%
l
os
t ru
l
es
g
en
er
ate
d
b
y
the
i
ntrod
uc
ed
f
r
am
ew
ork
by
m
od
i
f
y
i
n
g µ1/
µ
2 u
s
i
ng
C
l
ue
_
web_
da
tas
et
Evaluation Warning : The document was created with Spire.PDF for Python.
◼
IS
S
N: 16
93
-
6
93
0
T
E
L
KO
M
NIK
A
V
ol
.
17
,
No
.
6,
D
ec
em
be
r
20
19
:
305
7
-
306
5
3064
5.
Co
n
clusion
T
he
as
s
oc
i
ati
o
n
r
ul
e
m
i
ni
n
g
m
ai
n
ad
v
an
t
ag
e
i
s
to
i
d
e
nti
f
y
am
bi
gu
ou
s
r
el
ati
on
s
a
m
on
g
i
nf
orm
ati
on
,
bu
t
i
t
a
l
s
o
c
au
s
es
s
ec
urit
y
de
v
as
tat
i
o
n.
T
o
ad
dres
s
thi
s
i
s
s
ue
,
we
c
an
s
i
m
pl
y
hi
de
the
as
s
oc
i
at
i
on
r
ul
es
to
pre
s
erv
e
f
i
n
e
-
tun
ed
as
s
oc
i
at
i
o
n
r
ul
es
.
V
ari
ou
s
proc
e
du
r
e
s
are
pro
po
s
ed
to
hi
de
as
s
oc
i
a
ti
o
n
r
ul
es
bu
t
m
an
y
of
the
pr
oc
ed
ures
r
ed
uc
e
the
i
tem
s
ets
c
on
fi
de
nc
e
v
a
l
u
es
be
l
o
w
t
he
de
f
i
ne
d
t
hres
h
ol
d
v
al
u
es
.
A
s
w
e
l
l
as
,
no
ex
i
s
ti
n
g
m
eth
od
s
i
ts
t
o
th
e
p
aral
l
e
l
en
v
i
r
on
m
en
t
to
proc
es
s
bi
g
da
ta
.
A
l
on
g
wi
th
de
l
et
i
ng
an
i
tem
s
et
c
au
s
es
a
s
erio
us
probl
em
f
o
r
up
c
om
i
ng
da
t
a
i
t
em
s
.
In
the
pres
e
nt
w
ork
,
w
e
us
ed
the
f
u
z
z
y
l
o
gi
c
m
eth
od
f
or
hi
di
ng
m
i
ni
n
g
prac
ti
c
es
a
ga
i
ns
t
l
arg
e
i
nf
orm
ati
on
m
i
ni
ng
c
on
di
t
i
o
n
s
.
T
hi
s
c
an
tr
y
to
r
ed
uc
e
the
u
nd
es
i
r
ed
i
m
pa
c
t
of
a
d
el
i
c
ate
r
ul
e
protec
ti
ng
on
un
-
de
l
i
c
ate
r
u
l
e
s
i
n
da
ta
s
e
ts
.
T
he
propos
ed
f
r
am
ew
ork
ha
s
f
ea
tures
l
i
k
e
p
aral
l
e
l
i
s
m
an
d
s
c
al
ab
i
l
i
t
y
,
s
o
the
s
e
f
ea
tures
he
l
p
t
o
proc
es
s
m
as
s
i
v
e
da
t
a.
T
he
r
es
ea
r
c
h
ou
tc
om
es
i
l
l
us
tr
ate
th
at
th
i
s
prop
os
ed
f
r
a
m
ew
ork
f
un
c
ti
on
b
ett
er
tha
n
ex
i
s
ti
n
g
m
od
el
s
. In
f
u
ture,
we
w
i
l
l
tr
y
to
r
ed
uc
e
the
i
nf
orm
ati
on
l
os
s
i
n
the
prop
os
ed
f
r
am
e
wor
k
.
A
c
kno
w
ledg
men
t
s
T
he
au
tho
r
s
are
es
p
ec
i
a
l
l
y
i
n
de
bt
ed
to
t
he
S
c
i
en
c
e
a
nd
E
ng
i
n
ee
r
i
ng
R
es
ea
r
c
h
B
oa
r
d
(
S
E
R
B
)
,
De
pa
r
tm
en
t
of
S
c
i
en
c
e
a
nd
T
ec
hn
o
l
o
g
y
(
DS
T
)
,
an
d
G
ov
ernm
en
t
of
Ind
i
a
f
or
prov
i
d
i
ng
an
e
nv
i
r
on
m
en
t
w
h
ere th
e
au
th
ors
c
ou
l
d
do
the
be
s
t
wor
k
po
s
s
i
bl
e.
Ref
er
en
ce
s
[1
]
L
i
n
J
,
Y
u
W
,
Z
h
a
n
g
N,
Y
a
n
g
X
,
Zh
a
n
g
H,
Zh
a
o
W
.
A
s
u
r
v
e
y
o
n
In
te
rn
e
t
o
f
T
h
i
n
g
s
:
a
r
c
h
i
t
e
c
t
u
r
e
,
e
n
a
b
l
i
n
g
t
e
c
h
n
o
l
o
g
i
e
s
,
s
e
c
u
r
i
t
y
a
n
d
p
ri
v
a
c
y
,
a
n
d
a
p
p
l
i
c
a
t
i
o
n
s
.
IEEE
In
te
rn
e
t
o
f
Th
i
n
g
s
J
o
u
rn
a
l
.
2
0
1
7
;
4
(
5
)
:
1
1
2
5
–
1
1
4
2
.
[2
]
Su
n
Y
,
So
n
g
H,
J
a
ra
AJ
,
Bi
e
R
.
In
te
rn
e
t
o
f
th
i
n
g
s
a
n
d
b
i
g
d
a
ta
a
n
a
l
y
ti
c
s
fo
r
s
m
a
rt
a
n
d
c
o
n
n
e
c
t
e
d
c
o
m
m
u
n
i
ti
e
s
.
IEEE
Ac
c
e
s
s
.
2
0
1
6
;
4
:
766
–
773
.
[3
]
Ram
u
G
.
A
s
e
c
u
re
c
l
o
u
d
fra
m
e
w
o
rk
to
s
h
a
r
e
EHR
s
u
s
i
n
g
m
o
d
i
fi
e
d
CP
-
ABE
a
n
d
t
h
e
a
t
tri
b
u
te
b
l
o
o
m
fi
l
te
r
.
Ed
u
c
a
t
i
o
n
a
n
d
In
f
o
rm
a
ti
o
n
Te
c
h
n
o
l
o
g
y
.
2
0
1
8
;
2
3
(5
):
2
2
1
3
-
33
.
h
tt
p
s
:/
/d
o
i
.o
rg
/1
0
.1
0
0
7
/s
1
0
6
3
9
-
018
-
9
7
1
3
-
7
.
[4
]
W
u
J
,
Zh
a
o
W
.
Des
i
g
n
a
n
d
re
a
l
i
z
a
ti
o
n
o
f
W
I
n
te
rn
e
t:
Fr
o
m
N
e
t
o
f
T
h
i
n
g
s
to
In
te
r
n
e
t
o
f
T
h
i
n
g
s
.
ACM
Tra
n
s
.
Cy
b
e
r
-
Ph
y
s
.
S
y
s
t
.
201
7
;
1
(
1
)
:1
–
2
.
Av
a
i
l
a
b
l
e
:
h
tt
p
:/
/d
o
i
.
a
c
m
.
o
rg
/
1
0
.1
1
4
5
/
2
8
7
2
3
3
2
.
[5
]
Ram
u
G
.
En
h
a
n
c
i
n
g
M
e
d
i
c
a
l
Dat
a
Se
c
u
ri
ty
i
n
t
h
e
Clo
u
d
Us
i
n
g
RBAC
-
CPABE
a
n
d
ASS.
In
te
rn
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
f
Ap
p
l
i
e
d
En
g
i
n
e
e
r
i
n
g
Re
s
e
a
r
c
h
.
2
0
1
8
;
1
3
(7
):
5
1
9
0
-
5196
.
[6
]
A
Za
n
e
l
l
a
A,
Bu
i
N,
Ca
s
te
l
l
a
n
i
A,
V
a
n
g
e
l
i
s
ta
L
,
Z
o
rz
i
M
.
In
te
rn
e
t
o
f
T
h
i
n
g
s
fo
r
s
m
a
rt
c
i
ti
e
s
.
IEE
E
In
te
rn
e
t
o
f
Th
i
n
g
s
j
o
u
r
n
a
l
.
2
0
1
4
;
1
(
1
)
:
22
–
32
.
[7
]
M
a
l
l
a
p
u
ra
m
S,
Ngw
u
m
N,
Y
u
a
n
F,
L
u
C,
Y
u
W
.
Sm
a
rt
c
i
ty
:
Th
e
s
t
a
te
o
f
th
e
a
rt
,
d
a
ta
s
e
t
s
,
a
n
d
e
v
a
l
u
a
ti
o
n
p
l
a
tf
o
rm
s
.
2
0
1
7
IE
EE/ACIS
1
6
th
In
te
rn
a
ti
o
n
a
l
C
o
n
fe
re
n
c
e
o
n
Co
m
p
u
t
e
r
a
n
d
I
n
fo
rm
a
ti
o
n
Sc
i
e
n
c
e
(
ICIS
).
2
0
1
7
:
447
–
4
5
2
.
[8
]
Che
n
F,
X
i
a
n
g
T
,
Fu
X
,
Yu
W
.
Us
e
r
d
i
f
fe
re
n
ti
a
t
e
d
v
e
r
i
fi
a
b
l
e
f
i
l
e
s
e
a
r
c
h
o
n
t
h
e
c
l
o
u
d
.
IEE
E
Tra
n
s
a
c
t
i
o
n
s
o
n
Se
rv
i
c
e
s
Co
m
p
u
ti
n
g
.
2
0
1
7
;
1
1
(6
)
:
9
48
–
61
.
[9
]
Che
n
X
W
,
L
i
n
X
.
Bi
g
d
a
ta
d
e
e
p
l
e
a
rn
i
n
g
:
Cha
l
l
e
n
g
e
s
a
n
d
p
e
r
s
p
e
c
ti
v
e
s
.
IEEE
A
c
c
e
s
s
.
2
0
1
4
;
2
:
514
–
5
2
5
.
[1
0
]
Ram
u
G
,
Red
d
y
BE
.
Se
c
u
r
e
a
rc
h
i
te
c
tu
r
e
to
m
a
n
a
g
e
E
HR
’
s
i
n
c
l
o
u
d
u
s
i
n
g
SSE
a
n
d
ABE
.
In
te
rn
a
ti
o
n
a
l
J
o
u
r
n
a
l
o
f
He
a
l
th
a
n
d
T
e
c
h
n
o
l
o
g
y
,
Sp
ri
n
g
e
r
.
2
0
1
5
;
5
(3
-
4
):
1
9
5
-
2
0
5
.
[1
1
]
Y
u
W
,
L
i
a
n
g
F,
He
X
,
Hat
c
h
e
r
W
G
,
L
u
C,
L
i
n
J
,
Y
a
n
g
X
.
A
s
u
rv
e
y
o
n
t
h
e
e
d
g
e
c
o
m
p
u
ti
n
g
fo
r
th
e
I
n
t
e
rn
e
t
o
f
T
h
i
n
g
s
.
IEEE
A
c
c
e
s
s
.
2
0
1
7
;
6
:
6
9
0
0
-
6
9
1
9
.
[1
2
]
Y
u
W
,
X
u
G
,
Che
n
Z,
M
o
u
l
e
m
a
P
.
A
c
l
o
u
d
c
o
m
p
u
ti
n
g
b
a
s
e
d
a
rc
h
i
te
c
tu
r
e
f
o
r
c
y
b
e
r
s
e
c
u
ri
t
y
s
i
t
u
a
t
i
o
n
a
ware
n
e
s
s
.
2
0
1
3
IEEE
Con
fe
re
n
c
e
o
n
Com
m
u
n
i
c
a
ti
o
n
s
a
n
d
Net
w
o
rk
Se
c
u
ri
ty
(CN
S).
2013
:
488
–
4
9
2
.
[1
3
]
Ngu
y
e
n
ND
,
Ngu
y
e
n
T
,
Nah
a
v
a
n
d
i
S
.
Sy
s
te
m
d
e
s
i
g
n
p
e
r
s
p
e
c
ti
v
e
fo
r
h
u
m
a
n
-
l
e
v
e
l
a
g
e
n
ts
u
s
i
n
g
d
e
e
p
re
i
n
fo
r
c
e
m
e
n
t
l
e
a
rn
i
n
g
:
A
s
u
rv
e
y
.
IEEE
Ac
c
e
s
s
.
2
0
17
;
5
:
2
7
0
9
1
–
27
102
.
[1
4
]
He
H,
G
a
r
c
i
a
EA
.
L
e
a
rn
i
n
g
fro
m
i
m
b
a
l
a
n
c
e
d
d
a
t
a
.
IEEE
Tra
n
s
a
c
ti
o
n
s
o
n
Kn
o
wle
d
g
e
a
n
d
Dat
a
En
g
i
n
e
e
ri
n
g
.
2
0
0
9
;
21
(
9
)
:
1
2
6
3
–
1
2
8
4
.
[1
5
]
Cuz
z
o
c
re
a
A,
L
e
u
n
g
CK,
M
a
c
Ki
n
n
o
n
RK
.
M
i
n
i
n
g
c
o
n
s
tra
i
n
e
d
fre
q
u
e
n
t
i
te
m
-
s
e
t
s
fro
m
d
i
s
tr
i
b
u
t
e
d
u
n
c
e
r
ta
i
n
d
a
t
a
.
Fu
tu
re
Ge
n
e
r.
Com
p
u
t.
Sy
s
t
.
2
0
1
4
;
37
:
117
–
126
.
[1
6
]
Zh
a
n
g
X
,
L
i
u
C,
Nep
a
l
S
,
et
al
.
A
h
y
b
ri
d
a
p
p
r
o
a
c
h
fo
r
s
c
a
l
a
b
l
e
s
u
b
tre
e
a
n
o
n
y
m
i
z
a
ti
o
n
o
v
e
r
b
i
g
d
a
t
a
u
s
i
n
g
M
a
p
Red
u
c
e
o
n
c
l
o
u
d
.
J
.
Co
m
p
u
t.
Sy
s
t.
S
c
i
.
2
0
1
4
;
80
(
5
)
:
1008
–
1
0
2
0
.
[1
7
]
L
i
Y
,
Che
n
M
,
L
i
Q
,
Z
h
a
n
g
W
.
En
a
b
l
i
n
g
m
u
l
ti
l
e
v
e
l
t
ru
s
t
i
n
p
r
i
v
a
c
y
p
re
s
e
rv
i
n
g
d
a
ta
m
i
n
i
n
g
.
IE
EE
Tra
n
s
.
Kn
o
wl.
Da
t
a
En
g
.
2
0
1
2
;
24
(9
)
:
1589
–
1
6
1
2
.
[1
8
]
W
u
Y
H,
Chi
a
n
g
CM
,
Che
n
AL
.
Hid
i
n
g
s
e
n
s
i
t
i
v
e
a
s
s
o
c
i
a
t
i
o
n
ru
l
e
s
w
i
th
l
i
m
i
te
d
s
i
d
e
e
ff
e
c
ts
.
IEEE
Tra
n
s
.
Kn
o
wl.
Da
ta
E
n
g
.
2007
;
19
(1
)
:
29
–
42
.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NIK
A
IS
S
N: 1
69
3
-
6
93
0
◼
P
r
ote
c
t
i
ng
bi
g d
a
ta
mi
n
i
n
g
as
s
oc
i
ati
on
r
u
l
es
us
i
ng
f
u
z
z
y
s
y
s
tem
(
G
an
di
k
ot
a Ra
mu
)
3065
[1
9
]
G
k
o
u
l
a
l
a
s
-
D
i
v
a
n
i
s
A,
V
e
ry
k
i
o
s
VS
.
Ex
a
c
t
k
n
o
w
l
e
d
g
e
h
i
d
i
n
g
t
h
ro
u
g
h
d
a
t
a
b
a
s
e
e
x
te
n
s
i
o
n
.
IE
EE
Tra
n
s
.
Kn
o
wl.
Da
t
a
En
g
.
2
0
0
9
;
21
(5
)
:
699
–
713
.
[2
0
]
L
e
HQ
,
Arc
h
-
In
t
S,
Ngu
y
e
n
HX
,
Arc
h
-
In
t
N
.
As
s
o
c
i
a
ti
o
n
r
u
l
e
h
i
d
i
n
g
i
n
r
i
s
k
m
a
n
a
g
e
m
e
n
t
fo
r
re
t
a
i
l
s
u
p
p
l
y
c
h
a
i
n
c
o
l
l
a
b
o
ra
ti
o
n
.
Co
m
p
u
t.
I
n
d
.
2
0
1
3
;
64
(
7
)
:
776
–
7
8
4
.
[2
1
]
L
i
Y
C,
Y
e
h
J
S,
Cha
n
g
CC
.
M
CIF:
a
n
e
ff
e
c
ti
v
e
s
a
n
i
ti
z
a
ti
o
n
a
l
g
o
r
i
th
m
fo
r
h
i
d
i
n
g
s
e
n
s
i
t
i
v
e
p
a
tt
e
rn
s
o
n
d
a
ta
m
i
n
i
n
g
.
Ad
v
.
En
g
.
In
f
.
2
0
0
7
;
21
(
3
)
:
2
6
9
–
280
.
[2
2
]
Ke
s
h
a
v
a
m
u
rth
y
BN,
T
o
s
h
n
i
w
a
l
D,
Es
h
w
a
r
BK
.
Hid
i
n
g
c
o
-
o
c
c
u
rri
n
g
p
r
i
o
ri
t
i
z
e
d
s
e
n
s
i
t
i
v
e
p
a
tt
e
rn
s
o
v
e
r
d
i
s
tr
i
b
u
t
e
d
p
r
o
g
re
s
s
i
v
e
s
e
q
u
e
n
ti
a
l
d
a
ta
s
tre
a
m
s
.
J
.
Net
w. Co
m
p
u
t.
Ap
p
l
.
2
0
1
2
;
35
(3
)
:
1
1
1
6
–
1129
.
[2
3
]
Che
n
CLP,
Zh
a
n
g
CY
.
Dat
a
-
i
n
te
n
s
i
v
e
a
p
p
l
i
c
a
ti
o
n
s
,
c
h
a
l
l
e
n
g
e
s
,
te
c
h
n
i
q
u
e
s
a
n
d
te
c
h
n
o
l
o
g
i
e
s
:
a
s
u
rv
e
y
o
n
b
i
g
d
a
ta
.
In
f.
S
c
i
.
2
0
1
4
;
2
7
5
:
3
1
4
–
3
4
7
.
[2
4
]
W
u
X
,
Zh
u
X
,
W
u
GQ
,
Din
g
W
.
Dat
a
m
i
n
i
n
g
w
i
th
b
i
g
d
a
ta
.
IE
EE T
ra
n
s
.
Kn
o
w
l
.
Dat
a
En
g
.
2
0
1
4
;
26
(
1
)
:
97
–
1
0
7
.
[2
5
]
Nerg
i
z
M
E,
G
ö
k
M
Z
.
Hy
b
ri
d
K
-
a
n
o
n
y
m
i
ty
.
Com
p
u
t.
Se
c
u
r
.
2
0
1
4
;
44
:
51
–
63
.
[2
6
]
L
i
B,
Erd
i
n
E,
G
u
n
e
s
M
H,
e
t
a
l
.
An
o
v
e
rv
i
e
w
o
f
a
n
o
n
y
m
i
t
y
te
c
h
n
o
l
o
g
y
u
s
a
g
e
.
Com
p
u
t.
Com
m
u
n
.
2013
;
36
(1
2
)
:
1
2
6
9
–
1
2
8
3
.
[2
7
]
M
o
n
re
a
l
e
A,
An
d
ri
e
n
k
o
G
L
,
A
n
d
ri
e
n
k
o
NV
,
e
t
a
l
.
M
o
v
e
m
e
n
t
d
a
t
a
a
n
o
n
y
m
i
ty
th
r
o
u
g
h
g
e
n
e
ra
l
i
z
a
ti
o
n
.
Tra
n
s
.
Da
ta
Pr
i
v
.
2
0
1
0
;
3
(
2
)
:
1
–
31
.
[2
8
]
Ki
s
i
l
e
v
i
c
h
S,
Ro
k
a
c
h
L
,
El
o
v
i
c
i
Y
,
Sh
a
p
i
ra
B
.
Eff
i
c
i
e
n
t
m
u
l
ti
d
i
m
e
n
s
i
o
n
a
l
s
u
p
p
re
s
s
i
o
n
fo
r
K
-
a
n
o
n
y
m
i
ty
.
IEEE
Tra
n
s
.
Kn
o
w
l
.
Dat
a
En
g
.
2010
;
22
(
3
)
:
3
3
4
–
3
4
7
.
[2
9
]
Zh
a
n
g
G
,
Y
a
n
g
Y
,
L
i
u
X
,
Che
n
J
.
A
ti
m
e
-
s
e
ri
e
s
p
a
tt
e
r
n
-
b
a
s
e
d
n
o
i
s
e
g
e
n
e
ra
t
i
o
n
s
tra
te
g
y
f
o
r
p
ri
v
a
c
y
p
ro
te
c
ti
o
n
i
n
c
l
o
u
d
c
o
m
p
u
ti
n
g
.
In
t.
Sy
m
p
Clu
s
te
r,
Cl
o
u
d
a
n
d
Grid
Com
p
u
ti
n
g
(CC
G
ri
d
)
.
O
tt
a
w
a
.
2012
:
458
–
4
6
5
.
[3
0
]
W
a
n
g
H
.
Q
u
a
l
i
ty
m
e
a
s
u
re
m
e
n
t
fo
r a
s
s
o
c
i
a
ti
o
n
ru
l
e
h
i
d
i
n
g
.
A
ASRI Pro
c
e
d
i
a
.
2
0
1
3
;
5
:
228
–
234
.
[3
1
]
M
o
u
s
ta
k
i
d
e
s
G
V,
Ve
ry
k
i
o
s
VS
.
A
M
a
x
M
i
n
a
p
p
ro
a
c
h
f
o
r
h
i
d
i
n
g
fr
e
q
u
e
n
t
i
te
m
s
e
t
s
.
Dat
a
K
n
o
wl.
En
g
.
2008
;
65
(1
)
:
75
–
89
.
[3
2
]
W
a
n
g
SL
,
P
a
ri
k
h
B,
J
a
fa
ri
A
.
Hid
i
n
g
i
n
fo
r
m
a
t
i
v
e
a
s
s
o
c
i
a
t
i
o
n
ru
l
e
s
e
ts
.
E
x
p
e
r
t
Sy
s
t.
Ap
p
l
.
2
0
0
7
;
33
(2
)
:
316
–
3
2
3
.
[3
3
]
Das
s
e
n
i
E,
Ve
ry
k
i
o
s
VS
,
El
m
a
g
a
rm
i
d
AK,
Be
rt
i
n
o
E
.
H
i
d
i
n
g
a
s
s
o
c
i
a
ti
o
n
ru
l
e
s
b
y
u
s
i
n
g
c
o
n
f
i
d
e
n
c
e
a
n
d
s
u
p
p
o
rt
.
I
n
f.
Hi
d
i
n
g
L
e
c
t.
Not
e
s
Co
m
p
u
t.
Sc
i
.
2
0
0
7
;
2
1
3
7
:
369
–
3
8
3
.
Evaluation Warning : The document was created with Spire.PDF for Python.