I
nd
o
ne
s
ia
n J
o
urna
l o
f
E
lect
rica
l En
g
ineering
a
nd
Co
m
pu
t
er
Science
Vo
l.
3
8
,
No
.
1
,
A
p
r
il
20
2
5
,
p
p
.
496
~
507
I
SS
N:
2
502
-
4
7
52
,
DOI
: 1
0
.
1
1
5
9
1
/ijee
cs
.v
3
8
.
i
1
.
pp
496
-
5
0
7
496
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ee
cs
.
ia
esco
r
e.
co
m
An ef
ficien
t
frequ
ent
i
temsets
f
indi
ng
in
distri
bute
d
da
tas
ets
with
mi
nimum
co
mm
unica
tion o
v
e
rhea
d
H
o
ud
a
E
s
s
a
lm
i
,
Ana
s
s
E
l A
f
f
a
r
La
b
o
r
a
t
o
r
y
o
f
E
n
g
i
n
e
e
r
i
n
g
S
c
i
e
n
c
e
s
,
P
o
l
y
d
i
sc
i
p
l
i
n
a
r
y
F
a
c
u
l
t
y
o
f
Ta
z
a
,
U
n
i
v
e
r
si
t
y
o
f
S
i
d
i
M
o
h
a
m
e
d
B
e
n
A
b
d
e
l
l
a
h
,
F
e
z
,
M
o
r
o
c
c
o
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Ap
r
13
,
2
0
2
4
R
ev
is
ed
Oct
16
,
202
4
Acc
ep
ted
Oct
30
,
2
0
2
4
F
in
d
in
g
fre
q
u
e
n
t
it
e
m
se
ts
is
a
n
e
ss
e
n
ti
a
l
re
se
a
rc
h
e
d
tec
h
n
iq
u
e
a
n
d
a
c
h
a
ll
e
n
g
i
n
g
tas
k
o
f
d
a
ta
m
in
i
n
g
.
Tra
d
it
i
o
n
a
l
a
p
p
r
o
a
c
h
e
s
fo
r
d
istri
b
u
ted
fre
q
u
e
n
t
it
e
m
se
ts
re
q
u
ire
m
a
ss
iv
e
c
o
m
m
u
n
ica
ti
o
n
o
v
e
rh
e
a
d
a
m
o
n
g
d
iffere
n
t
d
istri
b
u
ted
d
a
tas
e
ts.
In
th
is p
a
p
e
r,
we
a
d
o
p
t
a
n
e
w strate
g
y
fo
r
o
p
t
i
m
izin
g
th
e
ti
m
e
o
f
c
o
m
m
u
n
ica
ti
o
n
s/sy
n
c
h
r
o
n
iza
ti
o
n
s
fro
m
lar
g
e
d
a
tas
e
ts
a
n
d
,
we
p
re
se
n
t
a
n
o
v
e
l
a
lg
o
rit
h
m
fo
r
d
isc
o
v
e
ri
n
g
fre
q
u
e
n
t
it
e
m
se
ts
in
d
iffere
n
t
d
istri
b
u
ted
d
a
tas
e
ts
o
n
t
h
e
sla
v
e
sites
c
a
ll
e
d
fin
d
in
g
e
fficie
n
t
d
istri
b
u
ted
fre
q
u
e
n
t
it
e
m
se
t
s
(
F
EDF
I
).
Th
e
p
ro
p
o
se
d
a
lg
o
rit
h
m
is
c
a
p
a
b
le
o
f
g
e
n
e
ra
ti
n
g
th
e
imp
o
rtan
t
fre
q
u
e
n
t
i
tem
se
ts
b
y
a
p
p
ly
i
n
g
a
n
e
fficie
n
t
tec
h
n
iq
u
e
fo
r
p
ru
n
in
g
t
h
e
c
a
n
d
i
d
a
te
it
e
m
se
ts
.
Th
e
e
x
p
e
rime
n
tal
re
su
l
ts
c
o
n
f
ir
m
th
a
t
o
u
r
a
lg
o
rit
h
m
F
EDF
I
p
e
rfo
rm
s
b
e
tt
e
r
th
a
n
Ap
ri
o
ri
a
n
d
c
a
n
d
i
d
a
te
d
i
strib
u
ti
o
n
(
CD
)
a
lg
o
rit
h
m
s
in
term
s o
f
c
o
m
m
u
n
ica
ti
o
n
a
n
d
c
o
m
p
u
tatio
n
c
o
st
s.
K
ey
w
o
r
d
s
:
Ap
r
io
r
i
C
o
m
m
u
n
icatio
n
s
ch
em
e
C
o
m
p
u
tatio
n
co
s
ts
Dis
tr
ib
u
ted
d
atab
ase
Gen
er
atio
n
o
f
ca
n
d
i
d
ates
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
:
Ho
u
d
a
E
s
s
alm
i
L
ab
o
r
ato
r
y
o
f
E
n
g
i
n
ee
r
in
g
Scien
ce
s
,
Po
ly
d
is
cip
lin
ar
y
Facu
lt
y
o
f
T
az
a
Un
iv
er
s
i
ty
o
f
Sid
i
Mo
h
a
m
ed
B
en
Ab
d
ellah
Fez,
Mo
r
o
cc
o
E
m
ail:
h
o
u
d
a.
ess
alm
i@
u
s
m
b
a.
ac
.
m
a
1.
I
NT
RO
D
UCT
I
O
N
Data
m
in
in
g
o
r
KDD
(
k
n
o
wled
g
e
d
is
co
v
er
y
in
d
atab
ases
)
[
1
]
is
a
d
is
cip
lin
e
p
r
im
ar
ily
co
n
ce
r
n
e
d
with
th
e
f
o
r
m
u
latio
n
,
an
aly
s
is
,
an
d
im
p
lem
e
n
tatio
n
o
f
p
r
o
ce
s
s
es
th
at
co
n
v
er
t
d
ata
in
to
ac
tio
n
ab
le
k
n
o
wled
g
e
[
2
]
.
I
ts
co
r
e
o
b
jectiv
e
is
to
ex
tr
ac
t
m
ea
n
in
g
f
u
l
an
d
in
ter
p
r
etab
le
in
f
o
r
m
atio
n
f
r
o
m
d
ata
b
ases
[
3
]
.
T
h
e
d
ata
u
ti
lized
in
d
ata
m
in
in
g
o
f
te
n
p
r
esen
ts
ch
allen
g
es
d
u
e
t
o
its
lar
g
e
v
o
lu
m
e
(
n
u
m
er
o
u
s
r
ec
o
r
d
s
)
an
d
h
ig
h
d
im
en
s
io
n
ality
(
m
a
n
y
attr
ib
u
t
es).
As
a
r
esu
lt,
th
er
e
h
as
b
e
en
a
g
r
o
win
g
f
o
c
u
s
o
n
d
is
tr
ib
u
ted
d
ata
m
i
n
in
g
,
wh
ich
o
p
er
ates
th
r
o
u
g
h
d
is
tr
ib
u
ted
d
ata
an
aly
s
is
[
4
]
.
Dis
tr
ib
u
ted
d
ata
m
in
in
g
e
m
p
lo
y
s
a
r
a
n
g
e
o
f
f
ast
p
ar
allel
an
d
d
is
tr
ib
u
ted
tech
n
iq
u
es
to
u
n
co
v
e
r
im
p
licit
an
d
v
alu
ab
le
k
n
o
wled
g
e
[
5
]
,
[
6
]
.
On
e
o
f
t
h
e
m
o
s
t
p
r
o
m
in
en
t
tech
n
iq
u
es
is
ass
o
ciatio
n
r
u
le
m
in
in
g
,
wh
ic
h
aim
s
to
id
en
tify
s
ig
n
if
ican
t
r
elatio
n
s
h
ip
s
b
e
twee
n
attr
ib
u
tes
with
in
th
e
d
atab
ase.
T
h
e
g
e
n
er
atio
n
o
f
f
r
e
q
u
en
t
item
s
ets
is
a
f
u
n
d
a
m
en
tal
an
d
ess
en
tial
s
tep
in
s
o
lv
in
g
th
e
ass
o
ciatio
n
r
u
le
m
in
in
g
p
r
o
b
le
m
.
T
h
e
Ap
r
io
r
i
alg
o
r
ith
m
,
in
tr
o
d
u
ce
d
b
y
Ag
r
awa
l
[
7
]
,
is
a
f
o
u
n
d
atio
n
al
an
d
p
io
n
ee
r
in
g
al
g
o
r
ith
m
f
o
r
m
in
in
g
f
r
eq
u
en
t
item
s
ets
in
tr
an
s
ac
tio
n
al
d
atab
ases
.
T
h
e
g
en
er
al
co
n
ce
p
t o
f
th
is
alg
o
r
ith
m
in
v
o
lv
es iter
ativ
ely
s
ca
n
n
in
g
th
r
o
u
g
h
item
s
ets
lev
el
b
y
lev
el.
At
ea
c
h
lev
el
k
,
a
s
et
o
f
ca
n
d
id
ate
item
s
ets
o
f
s
ize
k
is
g
en
er
ated
b
y
jo
in
in
g
th
e
f
r
e
q
u
en
t
item
s
ets
id
en
tifie
d
in
th
e
p
r
ev
io
u
s
iter
at
io
n
.
T
h
is
s
et
o
f
ca
n
d
id
ate
item
s
ets
i
s
th
en
p
r
u
n
ed
b
a
s
e
d
o
n
a
s
t
a
ti
s
ti
c
a
l
m
e
t
r
i
c
c
al
l
e
d
s
u
p
p
o
r
t
.
S
e
v
e
r
al
v
a
r
ia
t
i
o
n
s
o
f
t
h
e
A
p
r
i
o
r
i
a
l
g
o
r
it
h
m
a
r
e
d
i
s
c
u
s
s
e
d
i
n
[
8
]
,
[
9
]
to
m
itig
ate
its
lim
itatio
n
o
f
m
u
ltip
le
s
ca
n
s
o
v
er
th
e
en
tire
tr
an
s
ac
tio
n
d
atab
ase,
also
th
e
wea
k
n
ess
es
o
f
th
e
o
r
ig
in
al
Ap
r
io
r
i
alg
o
r
ith
m
ar
e
cited
in
[
10
]
,
h
ig
h
lig
h
tin
g
its
in
ef
f
icien
cy
in
ter
m
s
o
f
tim
e
a
n
d
s
p
ac
e
to
s
ea
r
ch
f
o
r
f
r
e
q
u
en
t
item
s
ets.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2
5
0
2
-
4
7
52
A
n
efficien
t freq
u
en
t item
s
ets fin
d
in
g
in
d
is
tr
ib
u
ted
d
a
ta
s
ets w
ith
…
(
Ho
u
d
a
E
s
s
a
lmi
)
497
Ma
n
y
d
is
t
r
ib
u
ted
alg
o
r
it
h
m
s
h
av
e
b
ee
n
d
ev
elo
p
ed
f
o
r
f
r
eq
u
en
t
item
s
et
m
in
i
n
g
in
d
is
tr
ib
u
ted
en
v
ir
o
n
m
en
ts
,
d
esig
n
ed
t
o
p
r
o
v
id
e
s
u
b
s
tan
tial
co
m
p
u
tatio
n
al
p
o
wer
.
T
h
ese
in
clu
d
e
alg
o
r
ith
m
s
s
u
ch
as
C
D
(
ca
n
d
id
ate
d
is
tr
ib
u
tio
n
)
[
11
]
,
DM
A
(
d
is
tr
ib
u
ted
m
in
in
g
alg
o
r
ith
m
)
[
12
]
,
FDM
(
f
ast
d
is
tr
i
b
u
ted
m
i
n
in
g
)
[
13
]
,
ODAM
(
o
p
tim
ized
d
is
tr
ib
u
ted
ass
o
ciatio
n
m
in
in
g
)
[
14
]
,
a
n
d
DDM
(
d
is
tr
ib
u
ted
d
ec
is
io
n
m
in
n
er
)
[
15
]
.
T
h
ese
alg
o
r
ith
m
s
,
wh
ich
u
tili
ze
th
e
d
ata
p
ar
allelis
m
p
ar
ad
ig
m
,
d
if
f
er
in
th
eir
tec
h
n
iq
u
es
f
o
r
p
r
u
n
in
g
n
ew
ca
n
d
id
ates
o
r
w
h
eth
er
t
h
ey
em
p
lo
y
ca
n
d
i
d
ate
co
u
n
tin
g
tech
n
i
q
u
es.
I
n
co
n
tr
ast,
alg
o
r
ith
m
s
b
ased
o
n
task
p
ar
allelis
m
,
s
u
ch
as
DD
(
d
ata
d
is
tr
ib
u
tio
n
)
[
11
]
,
I
DD
(
in
tellig
en
te
d
ata
d
is
tr
ib
u
tio
n
)
[
16
]
,
HPA
(
h
ash
-
b
ased
p
ar
al
lel
m
in
in
g
o
f
ass
o
ciatio
n
r
u
le
s
)
[
17
]
,
an
d
o
th
e
r
s
.
T
h
e
y
p
ar
titi
o
n
b
o
t
h
th
e
ca
n
d
id
ates
an
d
th
e
d
atab
ase
am
o
n
g
p
r
o
ce
s
s
o
r
s
.
T
h
ey
d
if
f
er
in
th
eir
m
eth
o
d
s
f
o
r
d
iv
id
in
g
ca
n
d
id
ates
an
d
t
h
e
d
atab
ase.
So
m
e
o
th
er
al
g
o
r
it
h
m
s
ca
n
n
o
t
b
e
class
if
ied
i
n
to
eith
er
o
f
t
h
e
two
p
ar
ad
ig
m
s
s
u
ch
as t
h
e
C
aD
(
ca
n
d
id
ate
d
is
tr
ib
u
tio
n
)
[
11
]
alg
o
r
ith
m
,
it
r
ep
r
esen
ts
a
h
y
b
r
id
ap
p
r
o
ac
h
co
m
b
i
n
in
g
elem
en
ts
o
f
b
o
th
C
D
an
d
D
D.
I
t
f
o
cu
s
es
o
n
d
is
tr
ib
u
tin
g
ca
n
d
id
ates
to
en
s
u
r
e
co
n
s
is
ten
cy
b
etwe
en
th
e
item
s
ets
th
at
ar
e
ca
lcu
lated
an
d
th
e
tr
an
s
ac
tio
n
s
u
s
ed
f
o
r
t
h
e
s
e
ca
lcu
latio
n
s
.
B
y
d
is
tr
ib
u
tin
g
tr
an
s
ac
tio
n
s
an
d
ca
n
d
id
ates
b
ased
o
n
th
eir
p
r
ef
ix
es,
C
aD
en
ab
les
ea
ch
s
i
t
e
to
p
r
o
ce
s
s
d
ata
in
d
ep
en
d
e
n
tly
.
I
n
SH
(
s
k
ew
h
an
d
lin
g
)
[
1
8
]
,
ca
n
d
id
ate
item
s
ets
ar
e
n
o
t
d
er
iv
ed
a
p
r
io
r
i
f
r
o
m
p
r
ev
io
u
s
f
r
eq
u
en
t
item
s
ets.
I
n
s
tead
,
th
ey
ar
e
g
en
er
ated
i
n
d
ep
e
n
d
en
tly
b
y
e
ac
h
s
ite
b
y
s
ca
n
n
in
g
its
lo
ca
l
p
ar
titi
o
n
o
f
th
e
d
atab
ase.
T
h
e
h
y
b
r
i
d
d
is
tr
ib
u
tio
n
(
HD)
alg
o
r
ith
m
[
1
6
]
is
a
h
y
b
r
i
d
ap
p
r
o
ac
h
th
at
i
n
teg
r
ates I
D
D
an
d
C
D
m
eth
o
d
s
.
T
h
e
P
s
ites
ar
e
p
ar
titi
o
n
ed
in
t
o
g
r
o
u
p
s
o
f
eq
u
al
s
ize,
with
ea
ch
g
r
o
u
p
f
u
n
ctio
n
i
n
g
as
a
s
u
p
er
-
s
ite.
T
h
e
I
DD
alg
o
r
ith
m
i
s
em
p
lo
y
ed
with
in
ea
ch
g
r
o
u
p
,
w
h
ile
th
e
C
D
alg
o
r
ith
m
is
u
s
ed
f
o
r
in
ter
ac
tio
n
s
b
etwe
en
th
e
g
r
o
u
p
s
.
Z
a
k
i
et
a
l
.
[
19
]
p
r
o
p
o
s
e
a
s
et
o
f
alg
o
r
ith
m
s
(
PAR
-
E
clat,
PAR
-
Ma
x
E
clat,
an
d
PAR
-
Ma
x
C
liq
u
e)
th
at
u
s
e
d
atab
ase
p
ar
titi
o
n
in
g
a
n
d
ca
lcu
late
d
if
f
er
en
t
ca
n
d
id
ate
item
s
ets.
T
h
e
alg
o
r
ith
m
s
all
ass
u
m
e
th
at
th
e
d
atab
ase
is
in
v
er
tical
f
o
r
m
at
(
tid
-
lis
ts
f
o
r
ea
ch
item
)
,
u
n
lik
e
h
o
r
izo
n
tal
p
ar
titi
o
n
in
g
.
H
o
wev
er
,
t
h
ese
alg
o
r
ith
m
s
en
co
u
n
ter
s
ev
e
r
al
c
h
allen
g
es,
in
clu
d
i
n
g
s
y
n
ch
r
o
n
izatio
n
is
s
u
es,
co
m
m
u
n
icatio
n
o
v
er
h
ea
d
,
lo
ad
b
alan
cin
g
p
r
o
b
lem
s
,
an
d
a
h
ig
h
f
r
eq
u
en
cy
o
f
d
ata
s
ca
n
s
,
wh
ich
ad
v
er
s
ely
af
f
ec
t t
h
eir
p
er
f
o
r
m
a
n
ce
.
I
n
th
e
liter
atu
r
e,
m
a
n
y
ap
p
r
o
a
ch
es
f
o
cu
s
ed
o
n
im
p
r
o
v
in
g
d
i
s
tr
ib
u
ted
alg
o
r
it
h
m
s
f
o
r
m
in
in
g
f
r
e
q
u
en
t
item
s
ets in
d
is
tr
ib
u
ted
e
n
v
ir
o
n
m
en
ts
.
I
n
th
is
co
n
tex
t,
Mu
d
u
m
b
a
an
d
Kab
ir
[
2
0
]
p
r
o
p
o
s
ed
a
n
o
v
el
ap
p
r
o
ac
h
f
o
r
m
in
in
g
ass
o
ciatio
n
r
u
les
in
d
e
p
en
d
en
tly
f
r
o
m
m
u
ltip
le
d
ata
s
o
u
r
ce
s
.
T
h
is
m
eth
o
d
in
teg
r
ates
f
r
eq
u
en
t
p
atter
n
s
d
er
iv
ed
f
r
o
m
ea
ch
d
ata
s
o
u
r
ce
to
id
en
tif
y
f
r
eq
u
e
n
t
p
atter
n
s
ap
p
li
ca
b
le
ac
r
o
s
s
a
d
is
tr
ib
u
ted
e
n
v
ir
o
n
m
en
t.
Fu
r
th
er
m
o
r
e
,
th
e
m
o
d
el
ca
n
b
e
ex
ten
d
ed
t
o
g
en
e
r
ate
r
u
l
es
with
s
p
ec
if
ied
tar
g
ets.
T
h
e
m
o
d
el
im
p
r
o
v
es
tr
an
s
p
ar
en
cy
an
d
m
em
o
r
y
ef
f
icien
cy
in
ass
o
ciatio
n
r
u
le
g
en
er
atio
n
.
Ad
d
itio
n
ally
,
it
r
ev
ea
ls
s
ig
n
if
ican
t
r
elatio
n
s
h
ip
s
,
allo
win
g
d
ec
is
i
o
n
-
m
ak
e
r
s
to
ac
ce
s
s
f
r
eq
u
e
n
t
p
atter
n
s
f
r
o
m
in
d
iv
i
d
u
al
d
ata
s
o
u
r
ce
s
as
well
as
th
e
o
v
er
all
d
ataset
ac
r
o
s
s
th
e
en
v
ir
o
n
m
en
t.
Ng
u
y
en
et
a
l
.
[
2
1
]
p
r
esen
ted
th
r
ee
n
o
v
el
s
tr
ateg
ies
d
esig
n
ed
to
g
r
ea
tly
im
p
r
o
v
e
th
e
ef
f
icien
cy
o
f
m
u
lti
-
lev
el
h
ig
h
-
u
tili
ty
item
s
et
m
in
in
g
th
r
o
u
g
h
m
u
lti
-
c
o
r
e
p
r
o
ce
s
s
in
g
.
T
h
e
y
also
p
r
o
p
o
s
ed
two
n
ew
alg
o
r
ith
m
s
,
MCML
+
,
a
n
d
MCML
++
,
wh
ich
lev
er
a
g
e
t
h
ese
s
tr
ateg
ies.
B
y
f
u
lly
u
tili
zin
g
th
e
av
ailab
le
p
r
o
ce
s
s
o
r
co
r
es,
t
h
ese
ap
p
r
o
a
ch
es
ac
h
iev
e
s
u
p
er
i
o
r
p
e
r
f
o
r
m
an
ce
in
m
u
lti
-
lev
el
h
ig
h
u
t
ilit
y
item
s
et
(
HUI
)
m
in
in
g
.
Fer
n
an
d
ez
-
B
ass
o
et
a
l
.
[
2
2
]
p
r
o
v
id
e
a
n
o
v
er
v
iew
o
f
t
h
e
m
o
s
t
r
ep
r
esen
tativ
e
al
g
o
r
ith
m
s
f
o
r
ass
o
ciatio
n
r
u
le
m
i
n
in
g
,
p
a
r
ticu
lar
ly
th
o
s
e
d
esig
n
ed
f
o
r
p
r
o
ce
s
s
in
g
v
er
y
lar
g
e
d
atasets
.
R
ec
o
g
n
izin
g
th
e
lim
itatio
n
s
o
f
Had
o
o
p
co
m
p
a
r
ed
to
Sp
ar
k
,
th
ey
d
ev
elo
p
ed
n
ew
ef
f
icien
t
f
r
eq
u
en
t
item
s
et
alg
o
r
ith
m
s
f
o
r
b
ig
d
ata
u
s
in
g
d
is
tr
ib
u
ted
co
m
p
u
t
atio
n
.
Ad
d
itio
n
ally
,
th
e
y
in
tr
o
d
u
ce
d
a
n
ew
ass
o
ciatio
n
r
u
le
-
m
in
in
g
alg
o
r
ith
m
in
Sp
ar
k
an
d
co
m
p
ar
ed
th
ese
p
r
o
p
o
s
ed
alg
o
r
ith
m
s
with
e
x
is
tin
g
o
n
es,
v
ar
y
in
g
th
e
n
u
m
b
e
r
o
f
tr
an
s
ac
tio
n
s
a
n
d
item
s
.
Dio
p
et
a
l
.
[
2
3
]
i
n
tr
o
d
u
ce
d
th
e
f
ir
s
t
p
atter
n
-
on
-
d
em
an
d
ap
p
r
o
ac
h
f
o
r
ex
tr
ac
tin
g
p
at
ter
n
s
in
a
tr
an
s
ac
tio
n
al
d
is
tr
ib
u
ted
d
ata
b
ase.
T
h
is
ap
p
r
o
ac
h
in
v
o
lv
es
ex
tr
ac
tin
g
p
atter
n
s
in
s
tan
tly
wh
en
th
e
a
n
aly
s
t
r
eq
u
ir
es th
em
.
T
h
ey
p
r
o
p
o
s
ed
a
g
en
er
ic
alg
o
r
ith
m
ca
lled
d
is
tr
ib
u
ted
d
atab
ase
s
am
p
lin
g
(
D
DSam
p
lin
g
)
,
wh
ich
r
an
d
o
m
l
y
s
elec
ts
a
p
atter
n
f
r
o
m
a
d
is
tr
ib
u
ted
d
atab
ase
b
a
s
ed
o
n
an
i
n
ter
esti
n
g
n
ess
m
e
asu
r
e
th
at
c
o
m
b
in
es
f
r
eq
u
e
n
cy
an
d
len
g
t
h
-
b
ased
u
tili
ty
f
u
n
ctio
n
s
,
in
clu
d
in
g
le
n
g
t
h
co
n
s
tr
ain
ts
.
T
h
is
s
tu
d
y
in
v
esti
g
ated
th
e
ef
f
ec
ts
o
f
m
in
in
g
f
r
eq
u
e
n
t
item
s
e
ts
in
d
is
tr
ib
u
ted
d
atasets
with
m
in
im
u
m
co
m
m
u
n
icatio
n
o
v
er
h
ea
d
b
y
d
ev
elo
p
in
g
a
n
o
v
el
alg
o
r
it
h
m
f
o
r
f
r
eq
u
e
n
t
item
s
et
m
i
n
in
g
,
ca
lled
FEDFI
(
f
in
d
in
g
ef
f
icien
t
d
is
tr
ib
u
ted
f
r
eq
u
e
n
t
item
s
ets
)
,
with
an
ap
p
r
o
p
r
iate
co
m
m
u
n
icatio
n
m
o
d
el,
r
ef
er
r
ed
to
as
m
aster
/s
lav
es
.
W
h
ile
ea
r
lier
s
tu
d
ies
h
av
e
ex
p
lo
r
e
d
th
e
im
p
ac
t
o
f
m
in
in
g
d
is
tr
ib
u
ted
f
r
eq
u
en
t
item
s
ets
f
r
o
m
th
e
p
ar
alleliza
tio
n
o
f
m
in
i
n
g
a
lg
o
r
ith
m
s
[
2
4
]
,
w
h
ich
in
v
o
lv
e
p
er
f
o
r
m
in
g
d
is
tr
ib
u
te
d
c
o
m
p
u
tatio
n
s
o
n
a
s
in
g
le
d
atab
ase
o
r
in
ten
tio
n
ally
d
is
tr
ib
u
tin
g
th
e
d
ata
to
ex
p
e
d
ite
p
r
o
ce
s
s
in
g
.
T
h
is
ap
p
r
o
ac
h
a
im
s
to
ad
d
r
ess
th
e
ch
allen
g
es
id
en
tifie
d
i
n
ex
is
tin
g
d
is
tr
ib
u
ted
al
g
o
r
ith
m
s
.
T
h
e
f
o
c
u
s
is
o
n
r
e
d
u
cin
g
c
o
m
m
u
n
icatio
n
an
d
s
y
n
ch
r
o
n
izatio
n
co
s
ts
b
etwe
en
d
if
f
er
en
t
s
ites
,
wh
ich
ar
e
cr
itical
f
o
r
ev
alu
atin
g
t
h
e
p
er
f
o
r
m
an
ce
o
f
d
is
tr
ib
u
ted
alg
o
r
ith
m
s
.
Fu
r
th
e
r
m
o
r
e,
th
e
wo
r
k
co
n
tr
ib
u
tes
to
m
i
n
im
izin
g
d
ata
s
ca
n
s
b
y
r
ed
u
ci
n
g
th
e
n
u
m
b
er
o
f
ca
n
d
id
ates
g
e
n
er
ated
f
o
r
g
l
o
b
al
in
f
o
r
m
atio
n
co
m
p
u
tatio
n
in
a
d
is
tr
ib
u
te
d
s
ettin
g
.
Ou
r
FEDFI
alg
o
r
ith
m
em
p
lo
y
s
th
r
ee
co
r
e
tech
n
iq
u
e
s
to
f
ac
ilit
ate
th
e
s
ea
r
ch
f
o
r
f
r
eq
u
en
t
item
s
ets:
th
e
f
ir
s
t
tech
n
iq
u
e
id
e
n
tifie
s
th
e
f
r
eq
u
e
n
t
1
-
item
s
ets
an
d
2
-
item
s
ets
u
s
in
g
a
2
-
d
im
en
s
io
n
al
m
atr
ix
,
wh
ile
t
h
e
r
e
m
ain
in
g
i
tem
s
ets
ar
e
d
er
iv
ed
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
52
In
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
3
8
,
No
.
1
,
Ap
r
il
20
2
5
:
4
9
6
-
5
0
7
498
th
r
o
u
g
h
T
r
ie
co
n
s
tr
u
ctio
n
f
o
ll
o
wed
b
y
a
f
in
al
v
alid
atio
n
o
f
f
r
eq
u
en
t
item
s
ets.
T
o
ev
alu
ate
th
e
ef
f
ec
tiv
en
ess
o
f
o
u
r
ap
p
r
o
ac
h
,
we
co
m
p
ar
ed
o
u
r
alg
o
r
ith
m
with
th
e
Ap
r
io
r
i
an
d
C
D
alg
o
r
ith
m
s
in
ter
m
s
o
f
co
m
m
u
n
icatio
n
an
d
co
m
p
u
tatio
n
c
o
s
ts
.
T
h
e
c
h
o
ice
o
f
t
h
e
C
D
alg
o
r
ith
m
f
o
r
co
m
p
ar
is
o
n
is
d
u
e
t
o
its
f
o
u
n
d
atio
n
al
r
o
le
in
d
is
tr
ib
u
ted
alg
o
r
ith
m
s
,
it
is
a
s
im
p
le
p
ar
alleliza
tio
n
o
f
A
p
r
io
r
i.
I
t
o
p
er
ates
in
a
d
is
tr
ib
u
ted
co
n
tex
t
wh
er
e
ea
c
h
s
ite
p
r
o
ce
s
s
es
it
s
lo
ca
l
p
o
r
tio
n
o
f
th
e
tr
a
n
s
ac
tio
n
d
atab
ase,
ca
lcu
lates
lo
ca
l
s
u
p
p
o
r
ts
,
an
d
s
h
ar
es
th
em
with
o
th
er
s
ites
f
o
r
g
lo
b
al
s
u
p
p
o
r
t c
o
m
p
u
tatio
n
.
T
h
is
p
ap
er
is
s
tr
u
ctu
r
ed
as
f
o
llo
ws.
Sectio
n
2
o
u
tlin
es
th
e
m
eth
o
d
o
l
o
g
y
o
f
th
e
wo
r
k
b
y
p
r
esen
tin
g
th
e
ad
o
p
ted
co
m
m
u
n
icatio
n
m
o
d
el
a
n
d
t
h
e
p
r
o
p
o
s
ed
al
g
o
r
it
h
m
.
T
h
e
e
x
p
er
im
e
n
tal
r
esu
lts
an
d
a
d
is
cu
s
s
io
n
o
f
th
e
p
er
f
o
r
m
an
ce
o
f
th
e
FEDFI
alg
o
r
ith
m
o
n
v
a
r
io
u
s
d
ata
s
ets
ar
e
g
iv
en
in
Sectio
n
3
.
Fin
ally
,
s
ec
tio
n
4
co
n
clu
d
es th
is
p
ap
e
r
.
2.
M
E
T
H
O
D
2
.
1
.
M
o
del o
f
c
o
mm
un
ica
t
io
n
So
m
e
d
is
tr
ib
u
ted
alg
o
r
ith
m
s
f
o
r
f
r
eq
u
e
n
t
item
s
et
m
in
in
g
r
eq
u
ir
e
ea
ch
s
ite
S
i
to
b
r
o
ad
ca
s
t
th
e
lo
ca
lly
co
m
p
u
ted
s
u
p
p
o
r
ts
to
all
o
th
er
s
ites
a
t
ea
ch
iter
atio
n
k
,
r
es
u
ltin
g
in
s
ig
n
if
ican
t
o
v
er
h
e
ad
[
2
5
]
.
T
h
e
C
D
alg
o
r
ith
m
,
b
ein
g
th
e
f
o
u
n
d
atio
n
al
alg
o
r
ith
m
,
co
n
s
is
ts
o
f
th
r
ee
p
h
ases
.
I
n
p
h
ase
1
,
th
e
l
o
ca
l
s
u
p
p
o
r
ts
f
o
r
th
e
item
s
ets
in
ea
ch
s
ite’
s
lo
ca
l
p
a
r
titi
o
n
ar
e
ca
lcu
lated
.
I
n
p
h
ase
2
,
th
e
s
ites
s
y
n
c
h
r
o
n
ize
an
d
e
ac
h
s
ite
ex
c
h
an
g
es
t
h
e
l
o
c
a
l
s
u
p
p
o
r
ts
o
f
al
l
i
t
e
m
s
et
s
w
i
t
h
a
l
l
o
t
h
e
r
s
i
t
es
t
o
c
o
m
p
u
t
e
t
h
e
g
l
o
b
a
l
s
u
p
p
o
r
t
s
o
f
t
h
e
ite
m
s
e
ts
.
I
n
p
h
as
e
3
,
th
e
g
lo
b
ally
f
r
eq
u
en
t
item
s
ets
ar
e
id
en
tifie
d
an
d
g
en
er
ate
d
in
d
ep
en
d
en
tly
at
ea
ch
s
ite.
T
h
e
C
D
alg
o
r
ith
m
r
ep
ea
ts
u
n
til
n
o
m
o
r
e
item
s
e
ts
ca
n
b
e
g
en
er
ated
.
At
e
ac
h
iter
atio
n
,
th
e
s
am
e
s
et
o
f
f
r
eq
u
en
t
item
s
ets
is
id
en
tifie
d
,
lead
i
n
g
to
r
ed
u
n
d
a
n
t
ca
lcu
latio
n
s
o
f
th
eir
g
lo
b
al
s
u
p
p
o
r
ts
at
ea
ch
s
ite.
T
h
is
r
es
u
lts
in
an
in
c
r
ea
s
ed
n
u
m
b
er
o
f
co
m
m
u
n
icatio
n
s
an
d
/o
r
s
y
n
c
h
r
o
n
izatio
n
s
n
e
ed
ed
f
o
r
g
lo
b
al
in
f
o
r
m
atio
n
co
m
p
u
tatio
n
in
a
d
is
tr
ib
u
ted
en
v
ir
o
n
m
en
t.
T
o
m
itig
ate
th
is
is
s
u
e,
we
p
r
o
p
o
s
e
ad
ap
tin
g
th
e
m
aster
/s
lav
e
s
m
o
d
el
to
s
ig
n
if
ican
tly
r
ed
u
ce
th
e
n
u
m
b
er
o
f
co
m
m
u
n
icatio
n
s
an
d
s
y
n
ch
r
o
n
izatio
n
s
b
etwe
en
s
ites
,
th
er
eb
y
m
in
im
izin
g
r
ed
u
n
d
a
n
t
ca
lcu
latio
n
s
f
o
r
ca
n
d
id
ate
ite
m
s
ets.
Th
e
m
aster
p
r
o
ce
s
s
o
r
s
eg
m
en
ts
th
e
en
tire
d
atab
ase
in
to
clu
s
ter
s
an
d
ass
ig
n
s
th
ese
clu
s
ter
s
to
s
lav
e
p
r
o
ce
s
s
o
r
s
,
as
d
etailed
in
[
2
5
]
.
Vaso
y
a
an
d
Ko
li
[
2
6
]
d
e
m
o
n
s
tr
ated
th
at
im
p
lem
en
tin
g
a
m
aster
/s
lav
e
s
y
s
te
m
s
ig
n
if
ican
tly
im
p
r
o
v
ed
tim
e
a
n
d
s
p
ac
e
co
m
p
lex
ity
,
f
ac
ilit
at
in
g
o
p
tim
al
r
eso
u
r
ce
u
tili
za
tio
n
in
a
d
is
tr
ib
u
ted
en
v
ir
o
n
m
en
t.
I
n
o
u
r
wo
r
k
,
t
h
e
d
atab
ase
is
h
o
r
izo
n
tally
p
ar
titi
o
n
ed
am
o
n
g
th
e
s
lav
e
s
i
tes,
ea
ch
o
f
wh
ich
h
an
d
les
th
e
p
r
o
ce
s
s
in
g
f
o
r
its
d
ata
s
eg
m
en
t.
T
h
e
m
aster
s
ite
co
o
r
d
in
ates
b
etwe
en
th
e
s
lav
e
s
it
es,
r
ed
u
cin
g
th
e
n
u
m
b
er
o
f
m
ess
ag
es
ex
ch
an
g
ed
.
Du
r
in
g
ea
ch
iter
atio
n
k
k
,
th
e
s
lav
e
s
ites
s
en
d
m
ess
ag
es
to
th
e
m
aster
s
ite,
wh
ich
th
en
r
e
p
lies
with
m
ess
ag
es to
all
sl
av
e
s
ites
.
Fig
u
r
e
1
s
h
o
ws
th
e
m
aster
/s
lav
e
co
m
m
u
n
icatio
n
m
o
d
el
.
Fig
u
r
e
1
.
T
h
e
m
o
d
el
o
f
m
aster
/s
lav
e
co
m
m
u
n
icatio
n
2
.
2
.
P
r
o
po
s
ed
a
pp
ro
a
ch
T
h
e
d
is
tr
ib
u
tio
n
asp
ec
t
ca
n
b
e
d
escr
ib
ed
as
f
o
llo
ws
.
Su
p
p
o
s
e
th
at
DB
is
a
d
ataset
o
f
tr
an
s
ac
tio
n
s
p
ar
titi
o
n
ed
h
o
r
izo
n
tally
an
d
c
o
r
r
esp
o
n
d
en
tly
{
DB
1
,
DB
2
,
…,
DB
n
}
on
P
s
ites
{
S
1
,
S
2
,
….
S
n
}
in
a
d
is
tr
ib
u
ted
s
y
s
tem
.
D
is
th
e
to
tal
d
atab
ase
s
ize
,
an
d
d
i
is
th
e
s
ize
o
f
ea
ch
p
ar
titi
o
n
DB
i
f
o
r
i=
1
,
.
.
.
,
n
.
T
h
e
s
u
p
p
o
r
t(
X)
is
k
n
o
wn
as th
e
g
lo
b
al
s
u
p
p
o
r
t c
o
u
n
ter
i
n
th
e
en
tire
DB
an
d
ca
n
b
e
d
ef
i
n
ed
as
,
.
=
(
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2
5
0
2
-
4
7
52
A
n
efficien
t freq
u
en
t item
s
ets fin
d
in
g
in
d
is
tr
ib
u
ted
d
a
ta
s
ets w
ith
…
(
Ho
u
d
a
E
s
s
a
lmi
)
499
w
h
er
e
ca
r
d
(
X)
is
th
e
n
u
m
b
er
o
f
r
ec
o
r
d
s
in
t
h
e
wh
o
le
DB
c
o
n
tain
in
g
X
an
d
T
G
is
th
e
to
tal
n
u
m
b
er
o
f
r
ec
o
r
d
s
in
th
e
d
atab
ase
DB
.
T
h
e
Supp
or
t
i
(
X)
is
k
n
o
wn
as th
e
lo
ca
l
s
u
p
p
o
r
t c
o
u
n
ter
in
DB
i
an
d
ca
n
b
e
d
e
f
in
ed
as
,
.
=
(
)
w
h
er
e
ca
r
d
(
X)
is
th
e
n
u
m
b
e
r
o
f
r
ec
o
r
d
s
in
DB
i
co
n
tain
in
g
X
a
n
d
T
i
is
th
e
to
tal
n
u
m
b
er
o
f
r
e
co
r
d
s
in
th
e
d
atab
ase
DB
i
.
L
et
m
in
im
u
m
s
u
p
p
o
r
t
Sup
m
i
n
d
ef
in
ed
b
y
th
e
u
s
er
,
X
is
g
lo
b
ally
f
r
eq
u
en
t
if
X.
s
u
p
≥
Sup
m
i
n
×
D
,
co
r
r
esp
o
n
d
en
tly
,
X
is
lo
ca
lly
f
r
eq
u
en
t a
t sit
e
S
i
,
if
X.
s
upp
i
≥
Sup
m
i
n
×
d
i
.
Ou
r
FEDFI
alg
o
r
ith
m
co
n
s
is
ts
o
f
two
f
u
n
d
am
e
n
tal
p
h
ases
:
th
e
in
itializatio
n
p
h
ase
an
d
t
h
e
ex
ec
u
tio
n
p
h
ase.
Du
r
in
g
th
e
in
itializatio
n
p
h
ase,
th
e
m
aster
s
ite
p
ar
titi
o
n
s
th
e
h
o
r
izo
n
tal
d
ata
b
a
s
e
ac
r
o
s
s
all
s
ites
eq
u
itab
ly
b
y
d
iv
id
in
g
t
h
e
n
u
m
b
er
o
f
tr
an
s
ac
tio
n
s
b
y
t
h
e
n
u
m
b
er
o
f
s
ites
.
W
e
ar
e
g
iv
en
a
h
o
r
izo
n
tal
tr
an
s
ac
tio
n
d
atab
ase
β
illu
s
tr
ated
b
elo
w
in
Fig
u
r
e
2
,
wh
er
e
I
as
a
g
r
o
u
p
o
f
item
s
,
th
e
alp
h
ab
e
t I
=
{a
,
b
,
c,
d
,
e}
(
with
m
=5
elem
en
ts
)
,
a
n
d
T
i
d
as
a
s
et
o
f
tr
an
s
ac
tio
n
s
,
T
i
d
=
{
1
,
2
,
3
,
4
,
5
,
6
}
with
a
m
in
im
u
m
s
u
p
p
o
r
t
th
r
esh
o
ld
(
Sup
m
i
n
=2
)
.
T
h
e
h
o
r
iz
o
n
ta
l
tr
an
s
ac
tio
n
d
atab
ase
is
f
r
ag
m
en
ted
an
d
d
is
tr
ib
u
ted
on
two
s
lav
e
s
ite
s
as
s
h
o
wn
in
Fig
u
r
e
3
.
T
h
e
ex
ec
u
tio
n
p
h
ase
b
eg
in
s
with
co
n
v
er
tin
g
th
e
h
o
r
izo
n
ta
l
d
atab
ase
in
to
a
v
er
tical
d
ata
f
o
r
m
at
f
o
r
ea
ch
s
lav
e
s
ite,
g
en
er
atin
g
a
l
is
t
o
f
T
id
s
f
o
r
all
item
s
in
I
.
T
h
is
m
eth
o
d
in
v
o
lv
es
a
s
in
g
l
e
p
ass
th
r
o
u
g
h
th
e
h
o
r
izo
n
tal
tr
an
s
ac
tio
n
d
atab
a
s
e
to
id
en
tify
th
e
tr
an
s
ac
tio
n
T
id
s
f
o
r
ea
ch
item
in
I
,
wh
ich
is
th
en
u
s
ed
to
co
n
s
tr
u
ct
th
e
v
er
tical
tr
an
s
ac
ti
o
n
d
ata
b
ase.
T
h
e
p
r
im
ar
y
d
is
t
in
ctio
n
b
etwe
en
v
er
tical
a
n
d
h
o
r
izo
n
tal
d
ata
lies
in
th
e
s
tr
u
ctu
r
e
o
f
t
r
an
s
ac
tio
n
s
.
I
n
h
o
r
izo
n
tal
d
ata,
a
tr
an
s
ac
tio
n
in
clu
d
es
th
e
T
id
a
n
d
th
e
I
tem
s
et,
wh
er
ea
s
in
v
er
tical
d
ata,
ea
ch
tr
an
s
ac
tio
n
is
id
en
tifie
d
b
y
t
h
e
I
tem
s
et,
en
ab
lin
g
th
e
s
u
p
p
o
r
t
co
u
n
t
f
o
r
a
n
y
f
r
e
q
u
en
t
item
s
et
to
b
e
ca
lcu
lated
th
r
o
u
g
h
th
e
i
n
ter
s
ec
tio
n
o
f
T
id
Sets
.
T
h
is
v
er
tical
tr
an
s
ac
tio
n
d
atab
ase
c
o
m
p
r
is
es
a
lis
t
o
f
all
1
-
item
s
ets
an
d
th
eir
co
r
r
esp
o
n
d
in
g
T
id
s
f
o
r
ea
ch
s
lav
e
s
ite.
Fig
u
r
e
4
illu
s
tr
ates
th
e
co
n
v
er
s
io
n
o
f
two
s
lav
e
s
ites
in
a
v
er
tical
tr
an
s
ac
tio
n
d
atab
ase
.
Fig
u
r
e
2
.
Ho
r
izo
n
tal
tr
an
s
ac
tio
n
d
atab
ase
β
Fig
u
r
e
3
.
H
o
r
izo
n
tal
tr
an
s
ac
tio
n
d
atab
ase
f
o
r
s
lav
e
s
ite
1
an
d
s
lav
e
s
ite
2
Fig
u
r
e
4
.
Ver
tical
d
ata
b
ase
f
o
r
s
lav
e
s
ite
1
an
d
s
lav
e
s
ite
2
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
52
In
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
3
8
,
No
.
1
,
Ap
r
il
20
2
5
:
4
9
6
-
5
0
7
500
T
o
ef
f
icien
tly
e
x
tr
ac
t
all
g
lo
b
ally
co
m
m
o
n
item
s
ets
in
a
d
is
tr
ib
u
ted
co
n
tex
t,
o
u
r
FEDFI
alg
o
r
ith
m
ex
ec
u
tes
th
r
ee
tech
n
iq
u
es
to
av
o
id
r
ep
ea
ted
d
ata
b
ase
s
ca
n
s
an
d
r
ed
u
ce
th
e
co
s
t
ass
o
ciate
d
with
g
en
er
atin
g
a
lar
g
e
n
u
m
b
er
o
f
ca
n
d
id
ate
s
ets,
th
er
eb
y
m
in
im
izin
g
e
x
ec
u
tio
n
tim
e.
T
h
e
f
i
r
s
t
tech
n
iq
u
e
estab
lis
h
es
th
e
C
ount
Ti
d
L
ist
t
d
ata
s
tr
u
ctu
r
e
(
f
o
r
t=1
,
.
.
,
n
)
f
o
r
ea
c
h
s
lav
e
s
ite
to
f
ac
il
itate
th
e
co
m
p
u
tatio
n
o
f
f
r
eq
u
en
t
1
-
item
s
ets
an
d
2
-
item
s
ets
,
u
s
in
g
a
two
-
d
im
en
s
io
n
al
m
at
r
ix
o
f
s
ize
(
m
×
m
)
,
wh
er
e
m
r
ep
r
e
s
en
ts
th
e
n
u
m
b
er
o
f
item
s
in
th
e
v
e
r
tical
tr
an
s
ac
tio
n
d
ata
b
ase.
I
n
th
e
im
p
lem
en
tatio
n
,
th
e
item
s
ets
ar
e
o
r
d
er
e
d
l
ex
ico
g
r
ap
h
ically
,
an
d
r
ed
u
n
d
a
n
t
(
s
y
m
m
etr
ic)
elem
en
ts
ar
e
r
em
o
v
ed
f
r
o
m
th
e
C
ount
Ti
dL
ist
t
m
atr
ix
,
th
er
e
b
y
r
e
d
u
cin
g
its
s
i
ze
to
½
(
m
²+
m
)
.
Fu
r
th
er
m
o
r
e
,
th
e
C
ount
Ti
dL
ist
t
d
ata
s
tr
u
ctu
r
e
is
em
p
lo
y
ed
to
o
p
tim
ize
th
e
s
u
p
p
o
r
t
ca
lcu
latio
n
o
f
ca
n
d
id
ate
item
s
ets
with
o
u
t
ac
ce
s
s
in
g
th
e
v
er
tical
d
atab
ase.
T
h
e
C
ount
Ti
dL
ist
t
m
atr
ix
p
r
o
v
id
es
a
p
r
o
jectio
n
o
f
th
e
v
e
r
tical
tr
an
s
ac
tio
n
d
ata
b
ase
f
o
r
ea
ch
s
lav
e
s
ite.
E
ac
h
ce
ll
(
i,
j)
r
ep
r
esen
ts
th
e
s
ize
o
f
T
id
s
f
o
r
th
e
I
tem
s
et
c
o
n
s
is
tin
g
o
f
elem
en
ts
‘
x
i
’
an
d
‘
x
j
’
.
T
h
e
f
r
eq
u
e
n
cy
o
f
1
-
item
s
ets
is
d
eter
m
in
ed
b
y
th
e
s
ize
o
f
th
e
T
id
ass
o
ciate
d
with
ea
ch
item
;
f
r
e
q
u
en
cy
(
x
i
)=
s
ize
o
f
(
T
id
(
x
i
)
)
,
a
n
d
t
h
is
is
r
ef
lecte
d
i
n
th
e
d
iag
o
n
al
ce
lls
o
f
th
e
m
a
tr
ix
.
T
h
e
f
r
e
q
u
en
c
y
o
f
2
-
item
s
ets
is
ca
lcu
lated
b
ased
o
n
th
e
s
ize
o
f
t
h
e
in
ter
s
ec
tio
n
o
f
T
id
s
f
o
r
th
e
two
ite
m
s
;
f
r
eq
u
en
cy
(
x
i
x
j
)
=
s
ize
o
f
(
T
id
(
x
i
)
∩
T
id
(
x
j
)
)
,
an
d
th
ese
v
alu
es
ar
e
l
o
ca
ted
ab
o
v
e
t
h
e
d
ia
g
o
n
al
i
n
th
e
m
a
tr
ix
.
Fig
u
r
e
5
illu
s
tr
ates
th
e
c
o
n
s
tr
u
ctio
n
o
f
t
h
e
C
o
u
n
tTid
L
is
t m
atr
ix
f
o
r
s
lav
e
Sit
e
1
an
d
s
lav
e
Sit
e
2
.
T
h
e
ex
ac
t
s
u
p
p
o
r
ts
f
o
r
1
-
item
s
ets
an
d
ca
n
d
id
ate
2
-
item
s
ets
ar
e
d
eter
m
in
e
d
th
r
o
u
g
h
d
ir
ec
t
ac
ce
s
s
to
th
e
C
ount
Ti
dL
ist
t
s
tr
u
ctu
r
e,
an
d
elim
in
ate
th
o
s
e
th
at
d
o
n
o
t
s
atis
f
y
th
e
m
i
n
im
u
m
s
u
p
p
o
r
t
th
r
esh
o
ld
Sup
m
i
n
.
On
ly
th
e
f
r
e
q
u
en
t
1
-
item
s
ets an
d
2
-
item
s
ets ar
e
r
etain
ed
.
(
)
=
(
)
=
(
,
1
)
,
|
|
=
1
(
)
=
(
)
=
(
,
−
(
−
1
)
)
,
|
|
=
1
,
=
|
|
Sig
n
if
ican
tly
,
f
r
e
q
u
en
t
1
-
item
s
ets
an
d
2
-
item
s
ets
f
o
r
ea
ch
s
lav
e
s
ite
ca
n
b
e
c
o
m
p
u
te
d
in
a
s
in
g
le
p
ass
th
r
o
u
g
h
th
e
v
er
tical
d
atab
ase,
in
co
n
tr
ast
to
th
e
A
p
r
io
r
i
m
eth
o
d
,
w
h
ich
r
eq
u
ir
es
two
p
ass
es.
Su
b
s
eq
u
en
tly
,
ea
ch
s
lav
e
s
ite
tr
an
s
m
its
its
d
ata:
s
u
p
p
o
r
t
of
f
r
eq
u
e
n
t
1
-
item
s
et
,
2
-
item
s
et
an
d
T
id
s
o
f
f
r
e
q
u
en
t
2
-
item
s
et
:
(
T
id
(
x
i
x
j
)
=T
id
(
x
i
)
∩T
id
(
x
j
))
to
th
e
m
aster
s
ite.
T
h
e
m
aster
s
ite
th
en
co
n
s
tr
u
cts
a
g
lo
b
al
s
tr
u
ctu
r
e
C
ount
Ti
dL
ist
G
b
y
s
u
m
m
in
g
th
e
s
u
p
p
o
r
ts
f
r
o
m
th
e
C
ount
Ti
d
L
ist
t
o
f
th
e
d
if
f
er
e
n
t
s
lav
e
s
ites
,
with
o
u
t
n
ee
d
in
g
ac
ce
s
s
to
th
e
v
er
tical
d
atab
ases
o
r
ad
d
itio
n
al
co
m
m
u
n
icatio
n
with
th
e
s
lav
e
s
ites
.
As
s
h
o
wn
in
Fig
u
r
e
6
,
t
h
e
to
tal
o
f
th
e
s
u
p
p
o
r
t
v
al
u
es
f
r
o
m
th
e
m
atr
ix
o
f
s
lav
e
s
ite
1
C
ount
Ti
dL
ist
1
an
d
Slav
e
s
ite
2
C
ount
Ti
dL
ist
2
.
T
h
e
lis
t
o
f
1
-
item
s
ets
is
{a
:3
,
b
:5
,
c:5
,
e:
5
},
an
d
th
e
lis
t
o
f
2
-
item
s
ets
is
{a
b
:2
,
ac
:3
,
ae
:2
,
b
c:4
,
b
e:5
,
ce
:4
}.
T
h
e
lis
t
o
f
T
id
f
o
r
2
-
ite
m
s
ets
in
clu
d
es
{a
b
:(
3
,
5
)
,
ac
:(
1
,
3
,
5
)
,
ae
:(
3
,
5
)
,
b
c:(
2
,
3
,
5
,
6
)
,
b
e:(
2
,
3
,
4
,
5
,
6
)
,
ce
:(
2
,
3
,
5
,
6
)
}.
T
h
e
s
ec
o
n
d
tech
n
iq
u
e
o
f
o
u
r
alg
o
r
ith
m
aim
s
to
id
en
tify
th
e
s
et
o
f
K
-
item
s
ets
(
k
≥
3
)
with
in
th
e
s
lav
e
s
ite
d
atab
ases
.
To
r
ed
u
ce
th
e
ex
ch
an
g
e
o
f
d
is
tr
ib
u
te
d
ca
lcu
lati
o
n
s
f
o
r
f
r
eq
u
en
t
item
s
ets
b
etwe
en
th
ese
d
atab
ases
,
th
e
m
aster
s
ite
iter
ativ
ely
g
en
e
r
ates
th
e
lis
t
o
f
k
-
item
s
ets
(
k
≥
3
)
b
y
c
o
n
s
tr
u
ctin
g
a
T
r
ie.
At
th
e
f
ir
s
t
lev
el
o
f
th
e
T
r
ie,
th
e
n
o
d
es
ar
e
co
m
p
o
s
ed
o
f
th
e
s
et
o
f
f
r
e
q
u
en
t
2
-
item
s
ets
o
b
tain
ed
f
r
o
m
th
e
f
ir
s
t
t
ec
h
n
iq
u
e,
alo
n
g
with
th
eir
T
id
Sets
,
wh
ich
ar
e
o
r
d
er
ed
lex
ico
g
r
ap
h
icall
y
.
Fig
u
r
e
5
.
C
o
u
n
tTid
L
is
t m
atr
ix
f
o
r
s
lav
e
s
ite
1
an
d
s
lav
e
s
ite
2
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2
5
0
2
-
4
7
52
A
n
efficien
t freq
u
en
t item
s
ets fin
d
in
g
in
d
is
tr
ib
u
ted
d
a
ta
s
ets w
ith
…
(
Ho
u
d
a
E
s
s
a
lmi
)
501
Fig
u
r
e
6
.
C
alcu
late
th
e
Su
m
o
f
C
ount
Ti
dL
ist
G
f
o
r
s
lav
e
s
ite
1
an
d
s
lav
e
s
ite
2
T
o
g
en
e
r
ate
ca
n
d
id
ate
(
k
+
1
)
-
i
tem
s
et
n
o
d
es
f
o
r
th
e
n
ex
t
lev
el,
th
e
alg
o
r
ith
m
em
p
lo
y
s
a
s
elf
-
jo
in
o
f
th
e
(
k
-
1)
-
item
s
ets
g
en
er
ated
in
th
e
p
r
ev
i
o
u
s
lev
el,
en
s
u
r
i
n
g
th
at
(
k
−
1
)
item
s
o
f
t
h
e
t
wo
k
-
item
s
ets
ar
e
id
en
tical.
Ad
d
itio
n
ally
,
th
e
al
g
o
r
ith
m
d
eter
m
in
es
th
e
T
id
s
o
f
(
k
+1
)
-
item
s
ets
b
y
in
ter
s
ec
tin
g
th
e
T
id
s
o
f
th
e
two
k
-
item
s
ets
th
at
s
h
ar
e
t
h
e
s
am
e
(
k
−
1
)
item
s
.
T
h
e
alg
o
r
ith
m
co
n
tin
u
es
to
c
o
n
s
tr
u
ct
n
o
d
es
at
ea
ch
lev
el
alo
n
g
with
t
h
eir
T
id
s
u
n
til
n
o
f
u
r
th
er
n
o
d
es
ca
n
b
e
g
en
e
r
ated
,
th
er
e
b
y
c
o
n
clu
d
i
n
g
t
h
e
s
ea
r
ch
f
o
r
f
r
e
q
u
en
t
k
-
item
s
ets
with
th
eir
T
id
Sets
u
s
in
g
th
e
T
r
ie
s
tr
u
ctu
r
e.
Fin
all
y
,
th
e
ap
p
r
o
x
im
ate
s
u
p
p
o
r
t
o
f
th
e
s
et
o
f
ca
n
d
id
ate
k
-
item
s
ets is
d
eter
m
in
ed
b
y
c
o
u
n
tin
g
th
e
s
ize
o
f
th
e
T
id
Sets
f
o
r
ea
ch
n
o
d
e
g
en
er
ate
d
in
th
e
T
r
ie
s
tr
u
ctu
r
e.
I
n
Fig
u
r
e
7
,
we
illu
s
tr
ate
t
h
e
s
ec
o
n
d
tech
n
iq
u
e
u
s
in
g
th
e
p
r
ev
io
u
s
ex
am
p
le
f
r
o
m
d
atab
ase
β
.
Fo
r
in
s
tan
ce
,
at
lev
el
1
o
f
th
e
T
r
i
e,
th
e
two
n
o
d
es
"a
b
"
an
d
"a
c
"
h
av
e
th
e
s
am
e
(
k
-
1
)
item
"a
",
s
o
a
lin
k
will
b
e
cr
ea
ted
to
a
n
o
d
e
"a
b
c"
b
y
id
e
n
tify
in
g
th
e
T
id
Set th
r
o
u
g
h
th
e
in
ter
s
ec
tio
n
o
f
th
eir
T
id
s
(
T
id
(
ab
)
∩T
id
(
ac
)
=3
5
)
.
T
h
is
p
r
o
ce
s
s
co
n
tin
u
es u
p
t
o
l
ev
el
3
(
n
o
d
e
"a
b
ce
":
f
r
eq
u
e
n
t 4
-
item
s
et
with
T
id
:
3
5
)
.
Sin
ce
th
er
e
is
o
n
ly
o
n
e
4
-
item
n
o
d
e,
n
o
m
o
r
e
n
o
d
es
ca
n
b
e
g
e
n
er
ated
.
T
h
e
ap
p
r
o
x
i
m
ate
s
u
p
p
o
r
t
o
f
n
o
d
e
"a
b
c
"
is
eq
u
al
to
th
e
s
ize
o
f
T
id
(
3
5
)
:
2
.
T
h
e
s
et
o
f
ca
n
d
i
d
ate
k
-
item
s
ets
g
en
er
ated
,
alo
n
g
with
th
eir
ap
p
r
o
x
im
ate
s
u
p
p
o
r
t,
is
as
f
o
llo
ws:
{a
b
c:
2
,
ab
e:
2
,
ac
e:
2
,
b
ce
:
3
,
ab
ce
:
2
}.
T
h
is
was
d
er
iv
e
d
f
r
o
m
th
e
T
r
ie
s
tr
u
ctu
r
e
with
o
u
t
a
n
y
ex
c
h
an
g
es
with
s
lav
e
s
ite
1
an
d
s
lav
e
s
ite
2
.
Fig
u
r
e
7.
I
ll
u
s
tr
atio
n
o
f
T
r
ie
s
tr
u
ctu
r
e
f
o
r
k
-
item
s
ets (
k
≥
3
)
T
h
e
m
aster
s
ite
s
en
d
s
th
e
s
et
o
f
ca
n
d
id
ate
k
-
item
s
ets
to
th
e
s
lav
e
s
ites
to
d
eter
m
in
e
th
e
ex
a
ct
s
u
p
p
o
r
t
f
o
r
ea
ch
k
-
item
s
et.
T
h
er
e
f
o
r
e,
a
th
ir
d
r
ef
in
em
en
t
tech
n
iq
u
e
is
n
ec
ess
ar
y
to
v
alid
ate
th
e
s
u
p
p
o
r
t
v
alu
es.
E
ac
h
s
lav
e
s
ite
m
u
s
t
s
ca
n
it
s
lo
ca
l
d
ata
to
d
eter
m
in
e
th
e
s
u
p
p
o
r
t
f
o
r
ea
ch
K
-
i
tem
s
et.
At
th
is
s
tag
e,
T
h
e
FEDFI
alg
o
r
ith
m
d
eter
m
i
n
es
th
e
ex
a
ct
s
u
p
p
o
r
t
o
f
K
-
item
s
ets
b
ased
o
n
th
e
Su
p
p
o
r
tMin
m
etr
ic
in
tr
o
d
u
ce
d
in
o
u
r
p
ap
er
[
2
7
]
.
T
h
is
m
etr
ic
is
u
s
ed
f
o
r
p
r
u
n
in
g
ca
n
d
id
ates
with
o
u
t
ac
ce
s
s
in
g
th
e
d
atab
ase.
I
t
esti
m
ates
th
e
ac
tu
al
s
u
p
p
o
r
t
o
f
item
s
et
X
b
ased
o
n
th
e
m
i
n
im
u
m
s
u
p
p
o
r
ts
o
f
its
(
k
-
1
)
s
u
b
s
ets
id
en
tifie
d
in
th
e
p
r
ev
io
u
s
iter
atio
n
.
T
h
is
v
alu
e
r
ep
r
esen
ts
th
e
m
in
im
u
m
p
o
s
s
ib
le
s
u
p
p
o
r
t
f
o
r
a
K
-
item
s
et
X.
At
th
i
s
p
o
in
t
,
th
e
FEDFI
alg
o
r
ith
m
aim
s
to
d
eter
m
in
e
th
e
ex
ac
t
s
u
p
p
o
r
t
o
f
th
e
K
-
elem
en
t
s
ets
b
y
f
o
c
u
s
in
g
o
n
tr
an
s
ac
tio
n
s
th
at
in
clu
d
e
th
e
(
k
-
1
)
s
u
b
s
ets
o
f
th
e
K
-
item
s
et
th
at
co
n
tain
th
e
m
in
im
u
m
s
u
p
p
o
r
t.
T
h
is
m
eth
o
d
s
ig
n
if
ican
tly
r
ed
u
ce
s
th
e
n
u
m
b
e
r
o
f
tr
an
s
ac
tio
n
s
r
eq
u
ir
ed
to
ac
cu
r
a
tely
d
eter
m
in
e
th
e
s
u
p
p
o
r
ts
o
f
th
e
item
s
ets.
Fo
r
ex
am
p
le,
at
s
lav
e
s
ite
1
,
th
e
s
u
p
p
o
r
t
f
o
r
th
e
s
u
b
s
ets
o
f
th
e
it
em
s
et
s
"a
b
c"
is
a
s
f
o
llo
ws:
"a
b
":
1
,
"a
c"
:
2
,
a
n
d
"b
c"
:
2
,
w
h
ich
h
av
e
alr
ea
d
y
b
ee
n
ca
lc
u
lated
u
s
in
g
th
e
C
o
u
n
tTid
L
is
t
s
tr
u
ctu
r
e.
T
h
u
s
,
th
e
ex
ac
t
s
u
p
p
o
r
t
o
f
t
h
e
item
s
et
"a
b
c"
is
d
eter
m
in
ed
with
in
th
e
tr
an
s
ac
tio
n
s
o
f
th
e
s
u
b
s
ets
th
at
h
av
e
th
e
m
in
im
u
m
s
u
p
p
o
r
t
o
f
"a
b
"
.
T
h
e
tr
a
n
s
ac
tio
n
s
f
o
r
th
e
item
s
et
"a
b
"
h
av
e
b
ee
n
p
r
e
v
io
u
s
ly
d
e
f
in
ed
f
r
o
m
th
e
v
e
r
tical
d
atab
ase.
C
o
n
s
eq
u
en
tly
,
th
e
e
x
ac
t
s
u
p
p
o
r
t
f
o
r
"a
b
c"
is
f
o
u
n
d
in
t
r
an
s
ac
tio
n
T
id
=
3
,
c
o
n
s
eq
u
en
tly
t
h
e
ex
ac
t
Su
p
p
o
r
t
o
f
"a
b
c"
is
eq
u
al
to
1
.
Su
b
s
eq
u
en
tly
,
th
e
s
lav
e
s
ites
s
en
d
to
t
h
e
m
aster
s
ite
th
e
k
-
item
s
ets
with
th
eir
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
52
In
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
3
8
,
No
.
1
,
Ap
r
il
20
2
5
:
4
9
6
-
5
0
7
502
ex
ac
t
s
u
p
p
o
r
t
to
ca
lcu
late
th
e
g
lo
b
al
s
u
p
p
o
r
ts
an
d
id
e
n
tifie
s
th
e
f
r
eq
u
en
t
item
s
ets
b
y
d
is
ca
r
d
in
g
t
h
o
s
e
th
at
n
o
t
s
atis
f
y
th
e
s
p
ec
if
ied
m
in
im
u
m
s
u
p
p
o
r
t th
r
esh
o
ld
(
Sup
m
i
n
).
I
t
is
wid
ely
r
ec
o
g
n
ized
th
at
th
e
Ap
r
io
r
i
alg
o
r
ith
m
r
eq
u
ir
es
a
co
m
p
r
eh
en
s
iv
e
s
ca
n
o
f
all
tr
a
n
s
ac
tio
n
s
in
a
d
atab
ase
to
m
in
e
f
r
eq
u
en
t
k
-
item
s
ets.
I
n
co
n
tr
ast,
o
u
r
r
ef
in
em
en
t
tech
n
iq
u
e
s
ca
n
s
th
e
d
atab
ase
o
n
ly
f
o
r
tr
an
s
ac
tio
n
s
ass
o
ciate
d
with
th
e
s
u
b
s
ets
o
f
a
n
item
s
et
th
at
h
av
e
th
e
m
i
n
im
u
m
s
u
p
p
o
r
t
th
r
e
s
h
o
ld
.
T
h
is
m
et
h
o
d
s
ig
n
if
ican
tly
r
ed
u
ce
s
th
e
n
u
m
b
er
o
f
d
atab
ase
ac
ce
s
s
es r
eq
u
ir
ed
to
id
e
n
tify
f
r
eq
u
e
n
t k
-
item
s
ets.
T
h
e
p
r
o
p
o
s
ed
m
eth
o
d
is
illu
s
tr
ated
in
Fig
u
r
e
8
,
wh
ich
d
etails
th
e
m
eth
o
d
o
lo
g
y
em
p
lo
y
ed
b
y
th
e
FEDFI
alg
o
r
ith
m
a
n
d
its
k
e
y
co
m
p
o
n
en
ts
.
T
h
e
m
aster
s
ite
i
s
r
esp
o
n
s
ib
le
f
o
r
p
a
r
titi
o
n
in
g
an
d
d
is
tr
ib
u
tin
g
th
e
d
atab
ase
h
o
r
iz
o
n
tally
a
n
d
e
q
u
itab
ly
ac
r
o
s
s
all
s
ites
,
as
well
as
co
n
v
e
r
tin
g
ea
c
h
lo
ca
l
d
at
ab
ase
to
a
v
er
tical
f
o
r
m
at.
T
h
is
wo
r
k
is
ac
h
iev
e
d
b
y
ap
p
ly
in
g
t
h
r
ee
m
ajo
r
tec
h
n
iq
u
es
f
o
r
m
in
in
g
d
is
tr
ib
u
te
d
f
r
eq
u
en
t
item
s
ets:
g
en
er
atin
g
th
e
C
o
u
n
tTid
L
is
t
m
atr
ix
s
tr
u
ctu
r
e
f
o
r
ea
ch
Slav
e
s
ite
an
d
co
m
p
u
tin
g
th
e
g
l
o
b
al
s
u
p
p
o
r
t
f
o
r
1
-
item
s
ets
an
d
2
-
item
s
ets
a
t
th
e
m
aster
s
ite.
T
h
e
s
et
o
f
ca
n
d
id
ate
k
-
item
s
ets
(
k
≥
3
)
a
n
d
th
eir
s
u
p
p
o
r
ts
ar
e
d
eter
m
in
ed
u
s
in
g
th
e
T
r
ie
s
tr
u
ctu
r
e.
T
h
e
m
aster
s
ite
th
en
d
is
tr
ib
u
tes
th
e
ca
n
d
id
ates
to
th
e
s
lav
e
s
ites
f
o
r
f
u
r
th
er
r
ef
in
em
e
n
t
o
f
th
e
k
-
it
em
s
ets
(
k
≥
3
)
.
B
ased
o
n
th
e
r
esu
lts
p
r
o
v
id
ed
b
y
th
e
s
lav
e
s
ites
,
th
e
m
aster
s
ite
v
er
if
ies wh
eth
er
th
e
k
-
item
s
ets ar
e
g
lo
b
ally
f
r
eq
u
e
n
t
an
d
d
is
p
lay
s
th
e
r
esu
lts
f
o
r
all
f
r
eq
u
e
n
t k
-
item
s
ets.
Fig
u
r
e
8.
T
h
e
m
ain
p
r
o
ce
s
s
o
f
th
e
p
r
o
p
o
s
ed
FEDFI
I
n
th
is
m
eth
o
d
,
o
u
r
FEDFI
alg
o
r
ith
m
d
ev
el
o
p
s
a
m
aster
/s
lav
e
(
s
ites
)
co
m
m
u
n
icatio
n
m
o
d
el
b
y
p
ar
titi
o
n
in
g
th
e
d
atab
ase
ac
r
o
s
s
all
s
ites
in
an
e
q
u
itab
le
m
a
n
n
er
,
d
iv
id
i
n
g
th
e
n
u
m
b
e
r
o
f
tr
an
s
ac
tio
n
s
b
y
th
e
n
u
m
b
er
o
f
s
ites
.
C
o
n
v
er
s
ely
,
in
th
e
b
asic
C
D
d
ata
p
ar
allelis
m
alg
o
r
ith
m
an
d
t
h
e
b
asic
DD
task
p
ar
allelis
m
alg
o
r
ith
m
,
p
ar
titi
o
n
in
g
is
b
ase
d
o
n
a
r
a
n
d
o
m
d
is
tr
ib
u
tio
n
o
f
d
ata
b
y
p
er
f
o
r
m
i
n
g
a
h
o
r
izo
n
tal
d
iv
is
io
n
o
f
th
e
d
ata
(
u
s
in
g
an
"
all
-
to
-
all"
b
r
o
ad
ca
s
t
ap
p
r
o
ac
h
)
.
Ho
wev
e
r
,
it
s
u
f
f
er
s
f
r
o
m
g
e
n
er
atin
g
a
lar
g
e
n
u
m
b
er
o
f
ca
n
d
id
ates,
lead
in
g
t
o
in
cr
ea
s
e
d
co
m
m
u
n
icatio
n
o
v
er
h
ea
d
.
I
n
th
e
C
D
alg
o
r
ith
m
,
s
ites
s
y
n
ch
r
o
n
ize
a
f
ter
co
m
p
u
tin
g
th
e
l
o
ca
l
s
u
p
p
o
r
t
o
f
ca
n
d
id
ates,
a
n
d
ea
ch
s
ite
ex
ch
an
g
es
t
h
e
lo
ca
l
s
u
p
p
o
r
t
o
f
all
ca
n
d
id
ates
with
all
o
th
er
s
ites
to
ca
lcu
late
th
e
g
lo
b
al
s
u
p
p
o
r
t
.
I
n
th
e
DD
alg
o
r
ith
m
,
ea
ch
s
ite
p
r
o
ce
s
s
es
th
e
r
ec
eiv
ed
p
ar
titi
o
n
s
t
o
o
b
t
ain
g
lo
b
al
s
u
p
p
o
r
t
f
o
r
th
e
en
t
ir
e
d
atab
ase.
A
f
te
r
th
at,
ea
ch
s
ite
ca
lcu
lates
th
e
f
r
eq
u
en
t
item
s
ets
f
r
o
m
its
ca
n
d
id
ate
s
et,
ex
ch
an
g
es
th
ese
f
r
e
q
u
en
t
item
s
ets
with
all
o
th
er
s
ites
to
o
b
tain
th
e
co
m
p
lete
s
et
o
f
f
r
eq
u
e
n
t
item
s
ets,
an
d
th
en
g
e
n
er
ates
th
e
n
ew
ca
n
d
id
ates,
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2
5
0
2
-
4
7
52
A
n
efficien
t freq
u
en
t item
s
ets fin
d
in
g
in
d
is
tr
ib
u
ted
d
a
ta
s
ets w
ith
…
(
Ho
u
d
a
E
s
s
a
lmi
)
503
p
ar
titi
o
n
s
th
em
,
a
n
d
d
is
tr
ib
u
te
s
th
em
ac
r
o
s
s
all
s
ites
.
C
o
n
s
eq
u
en
tly
,
o
u
r
m
o
d
el
g
r
ea
tly
r
ed
u
ce
s
th
e
n
u
m
b
er
o
f
co
m
m
u
n
icatio
n
s
n
ec
ess
ar
y
f
o
r
th
e
d
is
tr
ib
u
ted
c
o
m
p
u
tatio
n
o
f
g
lo
b
al
f
r
eq
u
e
n
t
item
s
ets
co
m
p
ar
ed
to
th
e
co
n
v
en
tio
n
al
p
ar
allel/d
is
tr
ib
u
t
ed
alg
o
r
ith
m
s
p
r
ev
io
u
s
ly
d
escr
ib
ed
.
T
h
e
C
D
alg
o
r
ith
m
f
o
cu
s
es
o
n
th
e
Ap
r
io
r
i
alg
o
r
ith
m
to
ex
tr
ac
t
f
r
eq
u
en
t
k
-
item
s
ets.
B
y
ap
p
ly
in
g
th
e
Ap
r
io
r
i a
lg
o
r
ith
m
to
o
u
r
m
aster
/s
lav
es m
o
d
el
with
th
e
ex
am
p
le
ab
o
v
e,
we
f
in
d
th
e
f
o
llo
win
g
r
esu
lts
:
−
I
n
th
e
f
ir
s
t
iter
atio
n
,
b
o
th
s
lav
e
s
ites
ca
lcu
late
th
e
lo
ca
l
s
u
p
p
o
r
ts
o
f
t
h
e
ca
n
d
id
ate
1
-
i
tem
s
ets
(
C
1
(
s
1
)
={
a,
b
,
c,
d
,
e}
,
C
1
(
s
2
)
={
a,
b
,
c,
d
,
e}
)
an
d
s
en
d
t
h
ese
ca
lcu
latio
n
s
to
th
e
Ma
s
ter
s
ite.
−
T
h
e
m
aster
s
ite
m
er
g
es
th
e
r
ec
eiv
ed
ca
n
d
id
ates
(
C
1
(
s
1
)
an
d
C
1
(
s
2
)
)
,
d
eter
m
in
es
th
e
f
r
e
q
u
en
t
1
-
I
tem
s
ets
F1
,
co
m
p
u
tes th
e
ca
n
d
id
ates C2
,
an
d
s
e
n
d
s
th
em
b
ac
k
to
s
lav
e
s
ites
1
an
d
2
.
−
T
h
is
p
r
o
ce
s
s
is
r
ep
ea
ted
f
o
r
iter
atio
n
s
2
,
3
,
an
d
4
with
th
e
r
esp
ec
tiv
e
ca
n
d
id
ates
C
2
={
ab
,
ac
,
ae
,
b
c,
b
e,
ce
},
C
3
={
ab
c,
ab
e,
ac
e
,
b
ce
},
a
n
d
C
4
={
ab
ce
}.
T
h
e
Ap
r
io
r
i
alg
o
r
ith
m
th
u
s
p
e
r
f
o
r
m
s
f
o
u
r
iter
atio
n
s
to
ca
lc
u
late
f
r
eq
u
e
n
t
item
s
ets,
r
esu
ltin
g
in
f
o
u
r
p
h
ases
o
f
co
m
m
u
n
icatio
n
b
etwe
en
th
e
m
aster
s
ite
an
d
th
e
s
lav
e
s
ites
.
I
n
co
n
tr
ast,
o
u
r
alg
o
r
ith
m
r
eq
u
ir
es
o
n
ly
two
ac
ce
s
s
es
to
th
e
lo
ca
l
d
at
ab
ase
in
th
e
f
ir
s
t
tech
n
iq
u
e
o
f
th
e
ex
ec
u
tio
n
p
h
ase
to
o
b
ta
in
th
e
f
r
e
q
u
en
t
1
-
item
s
ets
an
d
2
-
item
s
ets
an
d
in
th
e
th
ir
d
tech
n
iq
u
e
d
u
r
in
g
th
e
r
ef
in
em
en
t
p
h
ase
to
v
alid
ate
t
h
e
s
u
p
p
o
r
ts
o
f
th
e
k
-
item
s
ets
(
k
≥
3
)
.
T
h
e
r
ef
o
r
e
,
o
u
r
FEDFI
alg
o
r
ith
m
r
eq
u
ir
es
o
n
ly
two
ex
ch
a
n
g
es
b
etwe
en
th
e
m
aster
s
ite
an
d
th
e
s
lav
e
s
ites
to
co
m
p
u
te
all
t
h
e
f
r
eq
u
en
t i
tem
s
ets.
3.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
T
h
is
s
ec
tio
n
p
r
esen
ts
b
o
th
th
e
ex
p
er
im
en
tal
r
esu
lts
an
d
a
d
is
cu
s
s
io
n
o
f
th
e
p
er
f
o
r
m
an
ce
o
f
th
e
FEDFI
alg
o
r
ith
m
,
b
en
c
h
m
ar
k
ed
ag
ain
s
t
Ap
r
io
r
i
an
d
C
D
al
g
o
r
ith
m
s
.
W
e
e
v
alu
ated
th
e
alg
o
r
ith
m
s
u
s
in
g
two
d
atasets
:
T
4
0
l1
0
D1
0
0
K
an
d
C
h
ess
.
T
h
e
ex
p
er
im
e
n
ts
wer
e
co
n
d
u
cted
i
n
a
d
is
tr
ib
u
ted
e
n
v
ir
o
n
m
en
t
u
s
in
g
a
m
aster
/s
lav
es
s
ch
em
e,
wh
er
e
d
atasets
wer
e
p
ar
titi
o
n
ed
h
o
r
i
zo
n
tally
ac
r
o
s
s
m
u
ltip
le
s
lav
e
s
ites
.
T
h
e
p
r
im
ar
y
o
b
jectiv
e
is
to
ass
e
s
s
th
e
ef
f
icien
cy
,
s
ca
lab
ilit
y
,
an
d
co
m
m
u
n
icatio
n
o
v
e
r
h
ea
d
ass
o
ciate
d
with
th
e
FEDFI
a
lg
o
r
ith
m
,
a
n
d
h
o
w
th
ese
asp
e
cts co
m
p
ar
e
to
th
o
s
e
o
f
th
e
Ap
r
io
r
i a
n
d
C
D
alg
o
r
ith
m
s
.
3
.
1
.
Da
t
a
s
et
o
v
er
v
iew
T
h
e
d
atasets
u
s
ed
wer
e
s
elec
ted
to
p
r
o
v
id
e
a
r
an
g
e
o
f
c
o
m
p
le
x
ities
:
−
T
4
0
l1
0
D1
0
0
K:
s
o
u
r
ce
d
f
r
o
m
FIM
I
[
2
8
]
an
d
s
tu
d
ied
b
y
Fo
u
r
n
ier
-
Vig
er
et
a
l
.
[
2
9
]
,
th
is
lar
g
e
-
s
ca
le
d
ataset
co
n
tain
s
1
,
0
0
0
d
is
tin
ct
item
s
a
n
d
1
0
0
,
0
0
0
tr
an
s
ac
tio
n
s
.
I
t
p
r
esen
ts
a
h
ig
h
-
d
im
en
s
io
n
al
ch
a
llen
g
e
th
at
test
s
th
e
s
ca
lab
ilit
y
an
d
co
m
m
u
n
ica
tio
n
ef
f
icien
cy
o
f
m
i
n
in
g
f
r
eq
u
en
t item
s
et
alg
o
r
ith
m
s
.
−
C
h
ess
:
a
ls
o
f
r
o
m
FIM
I
[
2
8
]
,
t
h
is
d
ataset,
ex
p
l
o
r
ed
b
y
Fo
u
r
n
ier
-
Vig
er
et
a
l.
[
2
9
]
,
co
n
s
is
ts
o
f
7
5
item
s
an
d
3
1
9
6
tr
an
s
ac
tio
n
s
.
I
t
o
f
f
er
s
a
m
o
r
e
s
tr
u
ctu
r
e
d
e
n
v
ir
o
n
m
en
t
f
o
r
ass
ess
in
g
alg
o
r
ith
m
p
er
f
o
r
m
an
ce
in
s
ce
n
ar
io
s
with
f
ewe
r
tr
an
s
ac
tio
n
s
an
d
less
d
ata
co
m
p
lex
ity
.
3.
2
.
E
x
perim
ent
a
l
s
et
up
T
h
e
ex
p
er
im
en
ts
wer
e
co
n
d
u
cted
o
n
a
s
y
s
tem
with
a
n
I
n
t
el®
C
o
r
eT
M
i7
p
r
o
ce
s
s
o
r
at
2
.
8
0
GHz
,
4
GB
o
f
R
AM
,
an
d
W
in
d
o
ws
1
0
.
W
e
d
is
tr
ib
u
ted
t
h
e
d
ata
s
ets
to
3
,
5
,
a
n
d
7
s
lav
e
n
o
d
es
(
s
ites
)
to
m
ea
s
u
r
e
s
ca
lab
ilit
y
an
d
ef
f
icien
cy
.
T
h
e
alg
o
r
ith
m
s
wer
e
im
p
lem
en
t
ed
in
J
av
a
u
s
in
g
th
e
NetBean
s
I
DE
.
T
h
e
Ap
r
io
r
i
an
d
FEDFI
alg
o
r
ith
m
s
u
s
e
a
b
id
ir
ec
tio
n
al
co
m
m
u
n
icatio
n
s
ch
em
e
b
etwe
en
s
ites
f
o
llo
win
g
th
e
m
aster
/s
lav
es
m
o
d
el,
wh
er
ea
s
th
e
C
D
alg
o
r
ith
m
em
p
lo
y
s
th
e
class
ic
m
o
d
els o
f
u
s
u
al
co
m
m
u
n
icatio
n
b
et
wee
n
s
ites
.
3.
3
.
Resul
t
s
3.
3
.
1.
E
x
ec
utio
n
t
im
e
a
na
l
y
s
is
T
h
e
p
er
f
o
r
m
a
n
ce
o
f
th
e
FEDFI
alg
o
r
ith
m
was
co
m
p
a
r
ed
t
o
t
h
at
o
f
t
h
e
Ap
r
i
o
r
i
an
d
C
D
alg
o
r
ith
m
s
in
ter
m
s
o
f
ex
ec
u
tio
n
tim
e.
Fig
u
r
es
7
an
d
8
p
r
esen
t
th
e
e
x
ec
u
tio
n
tim
e
as
a
f
u
n
ctio
n
o
f
t
h
e
m
in
im
u
m
s
u
p
p
o
r
t
th
r
esh
o
ld
f
o
r
C
h
ess
an
d
T
4
0
l1
0
D1
0
0
K
d
atasets
,
r
esp
ec
tiv
ely
.
−
C
h
ess
d
ata
s
et
:
th
e
r
esu
lt
s
f
o
r
th
e
C
h
ess
d
ataset,
d
ep
ict
ed
in
Fig
u
r
e
9
,
also
in
d
icate
th
at
FEDFI
o
u
tp
er
f
o
r
m
s
th
e
o
th
e
r
alg
o
r
ith
m
s
.
E
v
en
in
a
s
m
aller
d
ata
s
et
with
f
ewe
r
tr
an
s
ac
tio
n
s
,
th
e
alg
o
r
ith
m
’
s
ef
f
icien
cy
is
ev
id
en
t.
Fo
r
in
s
t
an
ce
,
at
a
s
u
p
p
o
r
t
th
r
esh
o
ld
o
f
1
0
%,
FEDFI
r
eq
u
ir
ed
a
p
p
r
o
x
im
ately
3
0
%
f
ewe
r
iter
atio
n
s
th
an
Ap
r
io
r
i
an
d
3
5
%
f
ewe
r
t
h
an
C
D.
T
h
is
r
ed
u
ctio
n
in
th
e
n
u
m
b
e
r
o
f
it
er
atio
n
s
d
ir
ec
tly
tr
an
s
lates
to
f
ewe
r
co
m
m
u
n
ic
atio
n
p
h
ases
,
u
n
d
er
s
co
r
in
g
FE
DFI
’
s
ab
ilit
y
to
s
ca
le
ef
f
icien
tly
ev
en
in
less
co
m
p
lex
e
n
v
ir
o
n
m
en
ts
.
−
T
4
0
l1
0
D1
0
0
K
d
ataset
:
as
s
h
o
wn
in
Fig
u
r
e
1
0
,
th
e
FED
FI
alg
o
r
ith
m
c
o
n
s
is
ten
tly
o
u
t
p
er
f
o
r
m
ed
b
o
th
Ap
r
io
r
i
a
n
d
C
D
ac
r
o
s
s
all
s
u
p
p
o
r
t
th
r
esh
o
ld
s
.
Fo
r
ex
am
p
le,
at
a
s
u
p
p
o
r
t
t
h
r
esh
o
ld
o
f
50
%,
FEDFI
co
m
p
leted
th
e
m
in
in
g
p
r
o
ce
s
s
ap
p
r
o
x
im
ately
2
0
%
f
aster
th
an
Ap
r
io
r
i
an
d
2
5
%
f
aster
th
an
C
D.
T
h
e
p
r
im
ar
y
r
ea
s
o
n
f
o
r
th
is
s
u
p
er
io
r
p
er
f
o
r
m
an
ce
is
FEDFI
’
s
ab
ilit
y
to
g
en
er
ate
f
ewe
r
ca
n
d
id
ate
item
s
ets,
wh
ich
r
ed
u
ce
s
th
e
n
u
m
b
er
o
f
co
m
m
u
n
icatio
n
r
o
u
n
d
s
b
etwe
en
s
lav
e
n
o
d
es.
T
h
is
r
ed
u
ctio
n
in
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
52
In
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
3
8
,
No
.
1
,
Ap
r
il
20
2
5
:
4
9
6
-
5
0
7
504
co
m
m
u
n
icatio
n
s
ig
n
if
ican
tly
lo
wer
s
th
e
o
v
er
all
ex
ec
u
tio
n
tim
e,
esp
ec
ially
f
o
r
la
r
g
er
d
atasets
wh
er
e
co
m
m
u
n
icatio
n
co
s
t c
an
b
e
a
m
ajo
r
b
o
ttlen
ec
k
.
Fig
u
r
e
9
.
C
o
m
p
a
r
is
o
n
o
f
r
u
n
tim
e
f
o
r
C
h
ess
d
atasets
Fig
u
r
e
1
0
.
C
o
m
p
ar
is
o
n
o
f
r
u
n
t
im
e
f
o
r
T
4
0
l1
0
D
1
0
0
K
d
atasets
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2
5
0
2
-
4
7
52
A
n
efficien
t freq
u
en
t item
s
ets fin
d
in
g
in
d
is
tr
ib
u
ted
d
a
ta
s
ets w
ith
…
(
Ho
u
d
a
E
s
s
a
lmi
)
505
3.
3
.
2.
Sca
la
bil
it
y
a
na
ly
s
is
T
o
ev
alu
ate
th
e
s
ca
lab
ilit
y
o
f
th
e
FEDFI
alg
o
r
ith
m
,
we
test
ed
it
with
an
in
cr
ea
s
in
g
n
u
m
b
er
o
f
s
lav
e
n
o
d
es
(
3
,
5
,
an
d
7
)
a
n
d
lar
g
er
d
ataset
s
izes.
Fig
u
r
e
1
1
illu
s
tr
ates
h
o
w
t
h
e
p
er
f
o
r
m
an
ce
o
f
FEDFI
s
ca
les
with
m
o
r
e
s
lav
e
n
o
d
es.
−
I
m
p
ac
t
o
f
n
o
d
e
in
cr
ea
s
e:
a
s
we
in
cr
ea
s
ed
th
e
n
u
m
b
er
o
f
s
lav
e
n
o
d
es,
th
e
FEDFI
alg
o
r
it
h
m
co
n
tin
u
ed
to
o
u
tp
er
f
o
r
m
Ap
r
io
r
i
an
d
C
D.
At
7
n
o
d
es,
FEDFI
s
t
ill
ex
h
ib
ited
b
etter
ex
ec
u
tio
n
tim
es,
wh
ile
Ap
r
io
r
i
an
d
C
D
s
h
o
wed
m
o
r
e
n
o
ticea
b
le
i
n
cr
ea
s
es
in
ex
ec
u
tio
n
tim
e
d
u
e
to
h
ig
h
er
co
m
m
u
n
icatio
n
o
v
er
h
ea
d
.
FEDFI
’
s
ab
ilit
y
to
m
in
im
ize
co
m
m
u
n
i
ca
tio
n
p
h
ases
,
ev
en
as
th
e
n
u
m
b
er
o
f
n
o
d
es
r
is
es,
is
a
cr
itical
f
ac
to
r
in
its
s
u
p
er
io
r
s
ca
lab
ilit
y
.
−
C
o
n
v
er
g
en
ce
at
h
ig
h
n
o
d
e
c
o
u
n
t:
in
ter
esti
n
g
ly
,
as
th
e
n
u
m
b
er
o
f
n
o
d
es
in
cr
ea
s
ed
,
we
o
b
s
er
v
ed
th
a
t
th
e
p
er
f
o
r
m
an
ce
o
f
all
alg
o
r
ith
m
s
b
eg
an
to
d
iv
er
g
e.
T
h
is
is
p
ar
ticu
lar
ly
n
o
ticea
b
le
at
7
n
o
d
es,
wh
er
e
th
e
co
m
m
u
n
icatio
n
o
v
er
h
ea
d
b
ec
am
e
a
d
o
m
in
an
t
f
ac
to
r
.
Ho
wev
er
,
ev
en
at
th
is
lev
el
o
f
p
ar
a
llelis
m
,
FED
FI
m
ain
tain
ed
an
ed
g
e
o
v
e
r
Ap
r
i
o
r
i
an
d
C
D,
r
ef
lectin
g
its
m
o
r
e
ef
f
icien
t
d
ata
d
is
tr
ib
u
tio
n
an
d
r
ed
u
ce
d
n
ee
d
f
o
r
co
m
m
u
n
icatio
n
r
o
u
n
d
s
.
Fig
u
r
e
1
1
.
Scalab
ilit
y
o
f
FED
FI
b
y
n
u
m
b
er
o
f
n
o
d
es with
Sup
m
i
n
=2
0
%
3.
4
.
Dis
cus
s
io
n
T
h
e
r
esu
lts
o
f
th
is
s
tu
d
y
h
i
g
h
lig
h
t
s
ev
er
al
im
p
o
r
tan
t
in
s
ig
h
ts
ab
o
u
t
th
e
FEDFI
al
g
o
r
ith
m
an
d
its
p
er
f
o
r
m
an
ce
in
a
d
is
tr
ib
u
te
d
co
n
tex
t.
E
f
f
icien
c
y
g
ain
s
:
FEDFI
’
s
ef
f
icien
cy
in
g
en
e
r
atin
g
f
ewe
r
ca
n
d
i
d
ate
item
s
ets
d
ir
ec
tly
lead
s
to
f
ast
er
ex
ec
u
tio
n
tim
es
an
d
r
ed
u
c
ed
co
m
m
u
n
icatio
n
o
v
e
r
h
ea
d
.
T
h
is
is
p
ar
ticu
lar
ly
ev
id
en
t
in
th
e
T
4
0
l1
0
D1
0
0
K
d
ataset,
wh
er
e
FEDFI
co
n
s
i
s
ten
tly
o
u
tp
er
f
o
r
m
ed
b
o
t
h
Ap
r
io
r
i
an
d
C
D.
T
h
e
ab
ilit
y
to
m
in
im
ize
c
o
m
m
u
n
icatio
n
r
o
u
n
d
s
is
a
s
ig
n
i
f
ican
t
ad
v
a
n
tag
e
in
d
is
tr
ib
u
te
d
s
y
s
tem
s
,
wh
er
e
co
m
m
u
n
icatio
n
co
s
ts
ca
n
q
u
ic
k
ly
escalate
.
S
u
p
e
r
io
r
s
ca
lab
ilit
y
:
th
e
s
ca
lab
ilit
y
o
f
FEDFI
is
o
n
e
o
f
its
s
tan
d
o
u
t
f
ea
tu
r
es.
As
th
e
d
ataset
s
iz
e
an
d
th
e
n
u
m
b
er
o
f
s
lav
e
n
o
d
es
in
cr
ea
s
ed
,
FEDFI
m
ain
tain
ed
s
u
p
er
io
r
p
er
f
o
r
m
an
ce
c
o
m
p
ar
e
d
to
Ap
r
io
r
i
an
d
C
D.
T
h
is
s
ca
lab
ilit
y
is
cr
u
cial
f
o
r
r
ea
l
-
wo
r
ld
ap
p
licatio
n
s
,
p
ar
ticu
lar
ly
in
b
i
g
d
ata
en
v
ir
o
n
m
en
ts
wh
e
r
e
d
atasets
ar
e
o
f
te
n
d
is
tr
ib
u
te
d
ac
r
o
s
s
m
u
ltip
le
s
ites
.
T
h
e
al
g
o
r
ith
m
’
s
a
b
ilit
y
to
h
an
d
le
lar
g
e
r
d
atasets
with
o
u
t
a
s
ig
n
if
ican
t
in
cr
ea
s
e
in
co
m
m
u
n
icatio
n
co
s
t
is
a
m
ajo
r
s
tr
en
g
th
.
C
o
m
p
a
r
is
o
n
with
r
elate
d
r
esear
ch
:
o
u
r
r
esu
lts
alig
n
with
f
in
d
in
g
s
f
r
o
m
p
r
io
r
s
tu
d
ies,
wh
ich
h
av
e
n
o
t
ed
th
e
lim
itatio
n
s
o
f
Ap
r
io
r
i
in
d
is
tr
ib
u
ted
s
ettin
g
s
d
u
e
t
o
m
a
n
ag
in
g
a
lar
g
e
v
o
l
u
m
e
o
f
ca
n
d
id
ate
s
ets,
p
ar
tic
u
lar
ly
wh
e
n
d
ea
lin
g
with
n
u
m
er
o
u
s
f
r
e
q
u
en
t
ite
m
s
ets,
lo
w
m
in
im
u
m
s
u
p
p
o
r
t
th
r
esh
o
ld
s
,
o
r
ex
te
n
s
iv
e
item
s
ets.
Ap
r
io
r
i's
p
er
f
o
r
m
an
ce
d
r
asti
ca
lly
d
ec
r
e
ases
an
d
b
ec
o
m
es
in
e
f
f
icien
t
wh
en
th
e
m
em
o
r
y
ca
p
ac
ity
is
co
n
s
tr
ain
ed
a
n
d
th
e
tr
an
s
ac
tio
n
co
u
n
t
is
h
ig
h
.
C
D
alg
o
r
ith
m
u
s
es
a
s
im
p
le
co
m
m
u
n
icatio
n
s
y
s
tem
to
ex
ch
an
g
e
lo
ca
l
s
u
p
p
o
r
ts
,
r
ely
in
g
o
n
a
b
r
o
ad
ca
s
t
all
-
to
-
all
co
m
m
u
n
icatio
n
m
o
d
el.
Ho
wev
er
,
it
s
u
f
f
er
s
f
r
o
m
g
en
e
r
at
in
g
a
lar
g
e
n
u
m
b
er
o
f
ca
n
d
id
ate
s
ets,
r
esu
ltin
g
in
in
cr
ea
s
ed
co
m
m
u
n
icatio
n
o
v
er
h
ea
d
.
T
h
e
p
e
r
f
o
r
m
an
ce
o
f
c
o
u
n
t
d
is
tr
ib
u
tio
n
is
lim
ited
,
s
im
ilar
to
Ap
r
io
r
i,
as
th
is
alg
o
r
ith
m
is
b
ased
o
n
t
h
e
f
r
eq
u
en
t
item
s
et
m
in
i
n
g
m
eth
o
d
.
C
o
n
s
eq
u
en
tly
,
it
g
en
er
ates
a
s
u
b
s
tan
tial
n
u
m
b
er
o
f
ca
n
d
id
ate
s
ets,
m
u
ch
lik
e
th
o
s
e
p
r
o
d
u
ce
d
b
y
Ap
r
io
r
i
.
T
h
e
p
r
o
p
o
s
ed
m
eth
o
d
FEDFI
ad
d
r
ess
es
th
ese
lim
i
tat
io
n
s
b
y
r
ed
u
cin
g
th
e
n
u
m
b
er
o
f
ca
n
d
id
ate
item
s
ets
an
d
co
m
m
u
n
icatio
n
p
h
ases
r
eq
u
ir
ed
to
g
en
e
r
ate
f
r
e
q
u
en
t
item
s
ets.
T
h
is
is
esp
ec
ially
cr
itical
f
o
r
lar
g
e
d
atasets
wh
er
e
co
m
m
u
n
icatio
n
co
s
ts
ca
n
b
ec
o
m
e
a
m
aj
o
r
b
o
t
tlen
ec
k
.
T
h
e
r
ed
u
ctio
n
in
co
m
m
u
n
icatio
n
r
o
u
n
d
s
o
b
s
er
v
e
d
in
o
u
r
ex
p
er
i
m
en
ts
m
ir
r
o
r
s
s
im
ilar
im
p
r
o
v
em
en
ts
r
ep
o
r
ted
in
o
th
e
r
r
e
s
ea
r
ch
o
n
d
is
tr
ib
u
ted
f
r
eq
u
en
t item
s
et
m
in
in
g
.
I
m
p
licatio
n
s
o
f
f
in
d
in
g
s
: th
e
s
u
p
er
io
r
p
er
f
o
r
m
an
ce
o
f
FEDFI
s
u
g
g
ests
th
at
it is
a
v
iab
le
alt
er
n
ativ
e
to
Ap
r
io
r
i
an
d
C
D
in
d
is
tr
ib
u
ted
en
v
ir
o
n
m
en
ts
.
I
ts
ab
ilit
y
to
m
in
im
ize
b
o
th
ex
ec
u
tio
n
tim
e
an
d
co
m
m
u
n
icatio
n
Evaluation Warning : The document was created with Spire.PDF for Python.