I
A
E
S
I
n
t
e
r
n
at
io
n
al
Jou
r
n
al
of
A
r
t
if
ic
ia
l
I
n
t
e
ll
ig
e
n
c
e
(
I
J
-
AI
)
V
ol
.
14
, N
o.
3
,
J
une
2025
, pp.
1919
~
1929
I
S
S
N
:
2252
-
8938
,
D
O
I
:
10.11591/
ij
a
i.
v
14
.i
3
.pp
1919
-
1929
1919
Jou
r
n
al
h
om
e
page
:
ht
tp
:
//
ij
ai
.
ia
e
s
c
or
e
.c
om
C
u
st
om
e
r
se
gm
e
n
t
at
i
on
u
si
n
g
as
soc
i
at
i
on
r
u
l
e
m
i
n
i
n
g
on
r
e
t
ai
l
t
r
an
sac
t
i
on
d
at
a
S
ir
iwan
K
aj
or
n
k
as
ir
at
1
,
P
a
t
t
ar
aw
an
G
u
n
gl
in
1
,
K
r
it
s
ad
a
P
u
an
gs
u
w
an
1
,
N
aw
ap
on
K
ae
w
s
u
w
an
2
1
F
a
c
ul
t
y
of
S
c
i
e
nc
e
a
nd
I
ndus
t
r
i
a
l
T
e
c
hnol
ogy,
P
r
i
nc
e
of
S
ongkl
a
U
ni
ve
r
s
i
t
y,
S
ur
a
t
T
ha
ni
C
a
m
pus
,
S
ur
a
t
T
ha
ni
,
T
ha
i
l
a
nd
2
F
a
c
ul
t
y
of
H
um
a
ni
t
i
e
s
a
nd
S
oc
i
a
l
S
c
i
e
nc
e
s
,
P
r
i
nc
e
of
S
ongkl
a
U
ni
ve
r
s
i
t
y,
P
a
t
t
a
ni
C
a
m
pus
,
P
a
t
t
a
ni
,
T
ha
i
l
a
nd
A
r
t
ic
le
I
n
f
o
A
B
S
T
R
A
C
T
A
r
ti
c
le
h
is
to
r
y
:
R
e
c
e
iv
e
d
D
e
c
11
,
2023
R
e
vi
s
e
d
N
ov
21
,
2024
A
c
c
e
pt
e
d
J
a
n
27
,
2025
This
research
aimed
to
investigate
a
suitable
algorit
hm
for
cu
stomer
segmentation
using
as
customer
behavior
indicators
the
recency,
freq
uency
,
and
monetary
(
RFM
)
values
of
the
customers
.
The
clusteri
ng
algo
rithms
K
-
means,
f
uzzy
C
-
means,
and
self
-
organizing
neural
network
(SONN
)
were
compared
for
finding
the
most
appropriat
e
algorit
hm.
The
cu
stomer
segmentation
was
analyzed
using
associati
on
rule
mining
with
the
fr
equent
pattern
algorit
hm
(FP
-
Growth).
Data
on
retail
transactio
ns
during
January
2021
-
May
2023
were
obtained
from
Tuenjai
Company
,
Th
ailand,
with
a
total
of
202,469
records.
The
results
from
the
three
algorit
hm
s
were
compared
by
the
silhouette
coefficient
(SC),
Calinski
-
Harabasz
(CH)
index,
Davies
-
Bouldin
(DB)
index,
iteration
count,
and
exe
cution
time.
The
results
showed
that
the
K
-
means
algorit
hm
was
the
most
suitable
algorit
hm
for
customer
segmentation
in
this
study.
K
-
means
clusteri
ng
group
ed
the
customers
into
three
groups
here
labeled
as
“importan
t
value”,
“
general
development”,
and
“lost”,
based
on
the
RFM
values.
There
were
38
ru
les
for
the
important
value
segment,
and
t
wo
rules
each
for
the
general
develo
pment
and
the
lost
groups.
These
results
could
be
useful
to
the
b
usiness
organization
for
improving
the
customer
experiences,
increasin
g
sales,
preparing
or
promoting
products,
and
stock
manageme
nt
efficiency.
K
e
y
w
o
r
d
s
:
A
s
s
oc
ia
ti
on
r
ul
e
s
C
lu
s
te
r
in
g
a
lg
or
it
hm
C
us
to
m
e
r
s
e
gm
e
nt
a
ti
on
F
r
e
que
nt
pa
tt
e
r
n
-
gr
ow
th
R
e
c
e
nc
y, f
r
e
que
nc
y,
a
nd
m
one
ta
r
y
a
na
ly
s
is
This
is
an
open
access
article
under
the
CC
BY
-
SA
license.
C
or
r
e
s
pon
di
n
g
A
u
th
or
:
S
ir
iwa
n
K
a
jo
r
nka
s
ir
a
t
F
a
c
ul
ty
of
S
c
ie
nc
e
a
nd
I
ndus
tr
ia
l
T
e
c
hnol
ogy,
P
r
in
c
e
of
S
ongkla
U
ni
ve
r
s
it
y,
S
ur
a
t
T
ha
ni
C
a
m
pu
s
31
M
oo
6,
M
a
kha
m
ti
a
,
M
ua
ng,
S
ur
a
t
T
ha
ni
84000,
T
ha
il
a
nd
E
m
a
il
:
s
ir
iwa
n
.w
o@
ps
u.a
c
.t
h
1.
I
N
T
R
O
D
U
C
T
I
O
N
O
nl
in
e
s
hopping
ha
s
b
e
c
om
e
one
of
th
e
m
o
s
t
popula
r
onl
in
e
a
c
ti
vi
ti
e
s
,
of
f
e
r
in
g
c
us
to
m
e
r
s
th
e
c
onve
ni
e
nc
e
of
pl
a
c
in
g
or
de
r
s
th
r
ough
th
e
ir
s
m
a
r
tp
hone
s
or
ot
h
e
r
m
obi
le
de
vi
c
e
s
,
a
nd
e
li
m
in
a
ti
ng
th
e
ne
e
d
to
vi
s
it
phys
ic
a
l
s
to
r
e
s
.
A
c
c
or
di
ng
to
e
-
c
om
m
e
r
c
e
tr
e
nds
,
in
2022
a
s
ig
ni
f
ic
a
nt
68.3%
of
T
h
a
i
pe
opl
e
e
nga
g
e
d
in
onl
in
e
s
hopping
e
ve
r
y
w
e
e
k
[
1]
.
H
ow
e
ve
r
,
th
e
a
bund
a
nc
e
of
pr
oduc
ts
a
va
il
a
bl
e
onl
in
e
of
te
n
le
a
ds
to
de
c
is
io
n
-
m
a
ki
ng
c
ha
ll
e
nge
s
f
or
c
us
to
m
e
r
s
,
r
e
s
ul
ti
ng
in
pr
ol
onge
d
br
ow
s
in
g
a
nd
he
s
it
a
ti
on
[
2]
, [
3]
.
To
a
ddr
e
s
s
th
is
i
s
s
ue
,
r
e
c
om
m
e
nda
ti
on
s
y
s
te
m
s
pl
a
y
a
c
r
uc
ia
l
r
ol
e
in
bus
in
e
s
s
a
c
ti
vi
ti
e
s
by
s
ugge
s
ti
ng
pr
oduc
t
s
to
c
u
s
to
m
e
r
s
b
a
s
e
d
on
va
r
io
us
f
a
c
to
r
s
s
uc
h
as
pur
c
h
a
s
e
hi
s
to
r
y,
be
s
t
-
s
e
ll
in
g
pr
oduc
ts
,
or
c
us
to
m
e
r
in
te
r
e
s
ts
[
3]
.
P
r
e
vi
ous
r
e
s
e
a
r
c
h
ha
s
s
how
n
th
a
t
r
e
c
o
m
m
e
nda
ti
ons
s
ig
ni
f
ic
a
nt
ly
in
f
lu
e
nc
e
c
us
to
m
e
r
s
to
m
a
ke
m
or
e
pur
c
ha
s
e
s
[
4]
.
E
f
f
e
c
ti
ve
c
us
to
m
e
r
s
e
gm
e
nt
a
ti
on
is
e
s
s
e
nt
ia
l
f
or
r
e
c
om
m
e
nda
ti
on
s
ys
te
m
s
to
ta
il
or
pr
oduc
ts
a
nd
s
e
r
vi
c
e
s
to
s
ui
t
th
e
di
ve
r
s
e
ne
e
d
s
of
c
us
to
m
e
r
s
[
5]
.
C
us
to
m
e
r
s
e
gm
e
nt
a
ti
on,
vi
ta
l
f
or
unde
r
s
ta
ndi
ng
c
us
to
m
e
r
ne
e
ds
a
nd
pr
e
f
e
r
e
nc
e
s
,
is
of
te
n
ba
s
e
d
on
va
r
io
us
f
a
c
to
r
s
s
uc
h
as
ge
nd
e
r
,
a
ge
,
li
f
e
s
ty
le
,
or
c
us
to
m
e
r
pur
c
ha
s
e
be
ha
vi
or
[
5]
.
T
hi
s
te
c
hni
que
ha
s
be
e
n
us
e
d
in
va
r
io
us
c
ont
e
xt
s
,
in
c
lu
di
ng
a
ut
om
ot
iv
e
in
du
s
tr
y,
tr
a
ve
l
bus
in
e
s
s
,
a
nd
e
le
c
tr
ic
it
y
m
a
r
ke
t,
to
d
e
c
ip
he
r
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
, V
ol
.
14
, N
o.
3
,
J
une
20
25
:
1919
-
1929
1920
c
on
s
um
e
r
pr
e
f
e
r
e
n
c
e
s
a
nd
b
e
h
a
v
io
r
s
[
6]
–
[
8]
.
T
hi
s
s
e
gm
e
n
ta
ti
o
n
is
f
a
c
il
it
a
t
e
d
by
c
lu
s
t
e
r
i
ng
a
lg
o
r
it
h
m
s
a
n
d
t
oo
ls
,
w
hi
c
h
s
tr
e
a
m
l
in
e
th
e
pr
oc
e
s
s
a
nd
e
n
ha
nc
e
a
c
c
ur
a
c
y
[
9]
–
[
11]
.
In
a
d
di
t
io
n
to
s
e
gm
e
n
ta
ti
o
n,
c
u
s
t
om
e
r
b
e
h
a
vi
or
a
n
a
ly
s
is
,
p
a
r
t
ic
ul
a
r
ly
u
s
in
g
th
e
r
e
c
e
nc
y,
f
r
e
q
ue
nc
y, a
n
d
m
o
n
e
ta
r
y
(
R
F
M
)
m
o
de
l
,
is
c
r
u
c
i
a
l
f
or
r
e
ta
in
i
ng
e
x
is
ti
n
g
c
u
s
t
om
e
r
s
a
n
d
a
tt
r
a
c
ti
n
g
n
e
w
o
n
e
s
.
T
h
e
R
F
M
m
od
e
l
he
lp
s
id
e
nt
if
y
v
a
l
ua
bl
e
c
u
s
to
m
e
r
s
e
gm
e
n
t
s
a
n
d
d
e
v
is
e
ta
r
g
e
t
e
d
s
tr
a
t
e
g
ie
s
f
or
gr
ow
th
a
nd
r
e
te
nt
i
on
[
5
]
,
[
12
]
.
A
s
s
o
c
i
a
ti
on
r
ul
e
m
i
ni
n
g
f
ur
t
he
r
e
n
ha
nc
e
s
un
de
r
s
ta
ndi
ng
of
pur
c
h
a
s
e
be
ha
vi
or
by
un
c
o
ve
r
in
g
p
a
tt
e
r
n
s
a
n
d
r
e
l
a
ti
on
s
h
ip
s
w
it
h
i
n
tr
a
n
s
a
c
t
io
n
d
a
t
a
[
1
3]
.
T
hi
s
a
ppr
oa
c
h
h
a
s
b
e
e
n
a
pp
li
e
d
a
c
r
o
s
s
va
r
io
us
in
du
s
tr
ie
s
,
i
nc
lu
d
in
g
s
p
or
t
s
,
f
oo
d,
a
nd
pr
o
duc
t
de
s
i
gn
[
14]
–
[
1
6]
.
T
hi
s
s
tu
dy
in
ve
s
ti
ga
t
e
s
th
e
in
te
gr
a
ti
on
of
c
us
to
m
e
r
s
e
gm
e
nt
a
ti
on
a
nd
a
s
s
oc
ia
ti
on
r
ul
e
m
in
in
g
f
or
s
a
le
s
da
t
a
us
in
g
R
F
M
a
na
ly
s
i
s
w
it
h
th
e
c
lu
s
te
r
in
g
a
lg
or
it
hm
s
K
-
m
e
a
ns
,
f
uz
z
y
C
-
m
e
a
n
s
,
a
nd
s
e
lf
-
or
ga
ni
z
in
g
ne
ur
a
l
ne
twor
k
(
S
O
N
N
)
to
id
e
nt
if
y
th
e
m
os
t
e
f
f
e
c
ti
ve
a
ppr
o
a
c
h
f
or
s
e
gm
e
nt
a
ti
on.
T
he
e
v
a
lu
a
ti
on
m
e
tr
ic
s
s
il
houe
tt
e
c
oe
f
f
ic
ie
nt
(
SC
)
,
C
a
li
ns
ki
-
H
a
r
a
ba
s
z
(
CH
)
in
de
x,
D
a
vi
e
s
-
B
oul
di
n
(
DB
)
in
de
x,
it
e
r
a
ti
on
c
ount
,
a
nd
e
xe
c
ut
io
n
time
w
e
r
e
e
m
pl
oye
d
in
c
om
pa
r
is
on
s
.
S
ubs
e
que
nt
ly
,
th
e
f
r
e
que
nt
pa
tt
e
r
n
a
lg
or
it
hm
(
FP
-
G
r
ow
th
)
a
lg
or
it
hm
w
a
s
us
e
d
to
unve
il
a
s
s
oc
ia
ti
on
r
ul
e
s
w
it
hi
n
e
a
c
h
c
u
s
to
m
e
r
s
e
gm
e
nt
.
T
he
s
ub
s
e
q
ue
nt
s
e
c
ti
on
s
of
th
is
p
a
pe
r
a
r
e
or
ga
ni
z
e
d
as
f
ol
l
ow
s
:
s
e
c
ti
on
2
e
lu
c
id
a
te
s
th
e
m
e
th
odo
lo
gy
a
dopt
e
d
in
th
i
s
s
tu
dy.
S
e
c
ti
on
3
pr
e
s
e
nt
s
th
e
r
e
s
ul
t
s
a
nd
di
s
c
us
s
i
on,
in
c
lu
di
n
g
c
om
pa
r
i
s
on
s
w
it
h
pr
i
or
r
e
s
e
a
r
c
h.
S
e
c
ti
on
4
d
e
li
n
e
a
t
e
s
th
e
li
m
it
a
t
io
n
s
,
f
ut
ur
e
r
e
s
e
a
r
c
h
di
r
e
c
ti
on
s
,
a
nd
im
pl
ic
a
ti
on
s
of
our
f
in
di
ngs
.
F
in
a
ll
y,
s
e
c
ti
on
5
c
o
nc
lu
d
e
s
th
e
pa
pe
r
by
s
um
m
a
r
i
z
in
g
t
he
k
e
y
in
s
ig
ht
s
g
le
a
n
e
d
f
r
om
t
hi
s
in
ve
s
ti
g
a
ti
on
.
2.
M
E
T
H
O
D
T
he
th
r
e
e
hi
gh
le
ve
l
s
ta
ge
s
in
th
is
s
tu
dy
w
e
r
e
da
ta
c
ol
le
c
ti
on,
d
a
ta
pr
e
pa
r
a
ti
on,
a
nd
da
ta
a
na
ly
s
is
,
th
e
la
s
t
one
c
om
pr
is
in
g
c
u
s
to
m
e
r
s
e
gm
e
nt
a
ti
on,
c
us
to
m
e
r
c
la
s
s
if
ic
a
ti
on
a
nd
a
s
s
oc
ia
ti
on
r
ul
e
s
m
in
in
g
a
s
s
how
n
in
F
ig
ur
e
1.
S
ubs
e
c
ti
ons
2.1
-
2.3
e
la
bor
a
te
on
each
s
te
p.
T
he
s
e
s
te
ps
a
r
e
im
por
ta
nt
pr
oc
e
s
s
e
s
in
s
e
le
c
ti
ng
th
e
be
s
t
m
e
th
ods
f
or
c
us
to
m
e
r
s
e
gm
e
nt
a
ti
on.
2.1. Dat
a c
ol
le
c
t
io
n
D
a
ta
on
r
e
ta
il
tr
a
ns
a
c
ti
ons
dur
in
g
J
a
nua
r
y
2021
-
M
a
y
202
3
w
e
r
e
obt
a
in
e
d
f
r
om
T
ue
nj
a
i
P
a
ni
t
G
r
oup
C
om
pa
ny
L
im
it
e
d
(
C
o.,
L
td
.)
,
T
ha
il
a
nd
f
or
a
to
ta
l
o
f
202,469
r
e
c
or
ds
c
ove
r
in
g
38
s
to
r
e
s
in
th
e
S
ur
a
t
T
ha
ni
pr
ovi
nc
e
.
T
he
r
e
w
e
r
e
s
ix
a
tt
r
ib
ut
e
s
pe
r
c
us
to
m
e
r
tr
a
ns
a
c
ti
on,
na
m
e
ly
m
e
m
be
r
num
be
r
,
s
a
le
da
te
,
bi
ll
numbe
r
, pr
oduc
t
na
m
e
, t
he
nu
m
be
r
of
p
r
oduc
t
s
a
le
s
, a
nd s
a
l
e
s
(
B
a
ht
)
.
W
e
c
ont
a
c
te
d r
e
le
va
nt
s
ta
ke
hol
d
e
r
s
to
r
e
que
s
t
da
ta
f
or
our
r
e
s
e
a
r
c
h
a
nd
th
e
w
a
it
f
or
r
e
s
pons
e
s
to
o
k
a
s
ig
ni
f
ic
a
nt
ti
m
e
due
to
th
e
w
or
kl
oa
d
of
th
e
s
ta
ke
hol
de
r
s
a
nd t
he
pr
oc
e
s
s
of
e
xpor
ti
ng da
ta
f
r
om
t
he
s
ys
t
e
m
.
F
ig
ur
e
1.
S
c
he
m
a
ti
c
out
li
ne
of
th
e
s
tu
dy
a
ppr
oa
c
h
2.2.
D
at
a
p
r
e
p
ar
at
io
n
D
a
ta
pr
e
pa
r
a
ti
on
e
ns
ur
e
s
th
a
t
r
a
w
da
ta
a
r
e
a
c
c
ur
a
te
a
nd
c
om
p
le
te
,
th
e
r
e
by
s
ui
ta
bl
e
f
or
a
na
ly
s
is
to
f
in
d
knowle
dge
or
in
s
ig
ht
s
[
17]
.
T
he
d
a
t
a
pr
e
p
a
r
a
ti
o
n
pr
o
c
e
s
s
in
th
is
s
tu
dy
c
on
s
i
s
te
d
of
da
ta
c
le
a
n
s
i
ng
a
nd
d
a
t
a
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
C
us
to
m
e
r
s
e
gm
e
nt
at
io
n u
s
in
g as
s
o
c
ia
ti
on r
ul
e
m
in
in
g on r
e
ta
il
tr
ans
ac
ti
on data
(
Si
r
iw
an
K
aj
o
r
nk
as
i
r
at
)
1921
tr
a
n
s
f
or
m
a
ti
on
to
pr
e
pa
r
e
r
a
w
d
a
t
a
i
nt
o
a
f
or
m
a
t
th
a
t
c
a
n
be
a
n
a
l
yz
e
d.
T
h
e
l
a
r
g
e
da
ta
s
e
t
m
a
y
c
ont
a
i
n
e
r
r
or
s
or
be
in
c
on
s
i
s
t
e
n
t,
so
da
t
a
c
l
e
a
n
s
i
ng
a
nd
da
ta
tr
a
n
s
f
o
r
m
a
ti
o
n
a
r
e
n
e
c
e
s
s
a
r
y
to
a
voi
d
pr
ob
le
m
s
w
it
h
th
e
a
na
l
y
s
i
s
.
2.2.1
.
D
at
a
c
le
an
s
in
g
T
he
da
ta
c
le
a
n
s
in
g
pr
oc
e
s
s
id
e
nt
if
ie
s
a
nd
c
or
r
e
c
ts
e
r
r
or
s
or
m
is
s
in
g
da
ta
to
pr
e
pa
r
e
it
f
or
a
na
ly
s
is
[
17]
,
[
18
]
.
T
he
f
ir
s
t
s
te
p
is
to
lo
ok
f
or
e
nt
r
ie
s
w
it
h
m
is
s
in
g
da
ta
poi
nt
s
,
a
nd
to
r
e
m
ove
r
ow
s
or
c
ol
um
ns
w
it
h
a
hi
gh
pe
r
c
e
nt
a
ge
of
m
is
s
in
g
va
lu
e
s
,
e
s
pe
c
ia
ll
y
if
th
e
m
is
s
in
g
d
a
ta
a
r
e
not
s
y
s
te
m
a
ti
c
a
nd
r
e
m
ovi
ng
th
e
m
w
on'
t
s
ig
ni
f
ic
a
nt
ly
im
pa
c
t
th
e
a
na
ly
s
is
.
In
th
is
pr
oc
e
s
s
,
we
us
e
d
th
e
“
.i
s
nul
l(
)
.s
um
(
)
”
c
om
m
a
nd
to
c
he
c
k
f
or
a
ll
m
is
s
in
g
va
lu
e
s
in
th
e
da
ta
.
W
h
e
n
we
f
in
d
a
m
is
s
in
g
va
lu
e
,
we
us
e
d
th
e
“
.dr
opna
(
in
pl
a
c
e
=
T
r
ue
)
”
c
om
m
a
nd
to
de
le
te
th
e
r
ow
w
it
h
th
e
m
is
s
in
g
va
lu
e
be
f
or
e
pe
r
f
or
m
in
g
R
F
M
a
na
ly
s
is
.
A
f
te
r
th
a
t,
we
us
e
d
th
e
in
te
r
qua
r
ti
le
r
a
nge
(
I
Q
R
)
m
e
th
od
to
c
he
c
k
f
or
out
li
e
r
s
a
nd
r
e
m
ove
th
e
m
f
r
om
th
e
r
e
s
ul
ts
be
f
or
e
c
lu
s
te
r
in
g.
We
us
e
d
th
e
75t
h
a
nd
25t
h
qua
r
ti
le
va
lu
e
s
as
th
r
e
s
hol
ds
.
In
a
ddi
ti
on,
we
c
he
c
ke
d
th
e
m
e
m
be
r
num
be
r
to
m
a
ke
s
ur
e
it
is
va
li
d,
de
le
ti
ng
r
ow
s
w
he
r
e
th
e
m
e
m
be
r
num
b
e
r
ha
d
s
pe
c
ia
l
c
h
a
r
a
c
te
r
s
a
nd
w
a
s
not
a
10
-
di
gi
t
phone
num
be
r
.
A
f
te
r
th
e
da
ta
c
le
a
ns
in
g,
c
us
to
m
e
r
tr
a
ns
a
c
ti
on
da
ta
w
e
r
e
r
e
duc
e
d
to
173,261
r
e
c
or
ds
,
ha
vi
ng
r
e
m
ove
d
14.44
%
of
th
e
da
ta
due
to
dupl
ic
a
te
r
e
c
or
ds
,
m
is
s
in
g
va
lu
e
s
,
s
pe
c
ia
l
c
ha
r
a
c
te
r
s
,
in
c
or
r
e
c
t
da
ta
,
a
nd
out
li
e
r
s
.
2.2.2.
D
at
a
t
r
an
s
f
or
m
at
io
n
D
a
ta
tr
a
ns
f
or
m
a
ti
on
c
onve
r
ts
d
a
ta
in
to
a
f
or
m
a
t
th
a
t
is
s
ui
ta
bl
e
f
or
a
na
ly
ti
c
a
l
pur
pos
e
s
[
18]
.
In
th
is
pr
oc
e
s
s
th
e
da
ta
ty
pe
s
in
th
e
da
ta
s
e
t
w
e
r
e
a
dj
us
te
d
to
pr
e
ve
nt
pr
obl
e
m
s
dur
in
g
da
ta
a
na
ly
s
is
.
We
u
s
e
d
th
e
“
.t
o_da
te
ti
m
e
(
)
”
c
om
m
a
nd
w
it
h
th
e
s
a
le
da
te
to
c
onve
r
t
th
e
s
t
r
in
g
da
ta
in
to
a
da
te
ti
m
e
ob
je
c
t
f
or
m
a
t.
T
hi
s
w
a
s
done
if
th
e
c
ol
um
n
da
ta
w
a
s
in
s
tr
in
g
f
or
m
a
t
be
f
or
e
R
F
M
a
na
ly
s
i
s
.
We
th
e
n
u
s
e
d
“
S
ta
nd
a
r
dS
c
a
le
r
(
)
”
w
it
h
th
e
R
F
M
r
e
s
ul
ts
to
nor
m
a
li
z
e
th
e
da
t
a
be
f
or
e
c
lu
s
te
r
in
g.
2.3.
D
at
a
an
al
ys
is
In
th
e
da
ta
a
na
ly
s
is
pr
oc
e
s
s
,
if
th
e
c
om
put
e
r
us
e
d
is
not
pow
e
r
f
ul
e
nough,
th
is
m
a
y
r
e
s
ul
t
in
lo
ng
w
a
it
in
g
ti
m
e
s
f
or
r
e
s
ul
ts
,
e
s
pe
c
ia
ll
y
w
he
n
de
a
li
ng
w
it
h
l
a
r
ge
da
ta
s
e
ts
,
pa
r
ti
c
ul
a
r
ly
in
th
e
c
us
to
m
e
r
s
e
gm
e
nt
a
ti
on
pr
oc
e
s
s
.
C
u
s
to
m
e
r
s
e
gm
e
nt
a
ti
on
c
ons
is
t
s
of
th
r
e
e
s
ub
-
s
te
p
s
.
T
h
e
f
ir
s
t
s
te
p
is
R
F
M
a
n
a
ly
s
is
,
f
ol
lo
w
e
d
by
K
-
m
e
a
ns
,
f
uz
z
y
C
-
m
e
a
ns
,
a
nd
S
O
N
N
c
lu
s
te
r
in
g,
w
hi
c
h
c
om
pr
is
e
d
th
e
s
e
c
ond
s
te
p.
F
in
a
ll
y,
th
e
la
s
t
s
te
p
in
vol
ve
d
c
om
pa
r
in
g
th
e
c
lu
s
te
r
in
g
a
lg
or
it
hm
s
ba
s
e
d
on
S
C
,
CH
in
de
x,
DB
in
de
x,
it
e
r
a
ti
on
c
ount
,
a
nd
e
xe
c
ut
io
n
time
to
de
te
r
m
in
e
th
e
opt
im
a
l
a
ppr
oa
c
h
f
or
c
u
s
to
m
e
r
s
e
gm
e
nt
a
ti
on.
In
th
e
c
u
s
to
m
e
r
c
la
s
s
if
ic
a
ti
on
pr
oc
e
s
s
,
la
be
ls
w
e
r
e
a
s
s
ig
ne
d
to
c
us
to
m
e
r
gr
oups
ba
s
e
d
on
th
e
ir
R
F
M
va
lu
e
s
,
w
hi
c
h
r
e
pr
e
s
e
nt
th
e
ir
pur
c
ha
s
in
g
be
ha
vi
or
s
.
A
s
s
o
c
ia
ti
on
r
ul
e
m
in
in
g
is
th
e
a
na
ly
s
i
s
of
r
e
la
ti
ons
hi
ps
be
twe
e
n
it
e
m
s
th
a
t
h
a
ve
be
e
n
pr
e
vi
ous
ly
pur
c
ha
s
e
d
by
c
us
to
m
e
r
s
in
each
gr
oup
.
T
he
c
us
to
m
e
r
c
la
s
s
if
ic
a
ti
on
pr
oc
e
s
s
a
nd
a
s
s
oc
ia
ti
on
r
ul
e
m
in
in
g
a
r
e
th
e
s
ubs
e
que
nt
s
te
ps
f
ol
lo
w
in
g
c
us
to
m
e
r
s
e
gm
e
nt
a
ti
on.
2.3.1.
C
u
s
t
om
e
r
s
e
g
m
e
n
t
at
io
n
T
hi
s
s
tu
dy
us
e
d
R
F
M
a
na
ly
s
i
s
f
or
c
us
to
m
e
r
s
e
gm
e
nt
a
ti
on
to
ge
th
e
r
w
it
h
th
e
c
lu
s
te
r
in
g
a
lg
or
it
hm
s
K
-
m
e
a
ns
,
S
O
N
N
,
a
nd
f
uz
z
y
C
-
m
e
a
ns
[
19]
–
[
21]
.
We
a
na
ly
z
e
d
pr
e
pr
oc
e
s
s
e
d
c
u
s
to
m
e
r
tr
a
ns
a
c
ti
on
da
ta
to
id
e
nt
if
y
th
e
pur
c
ha
s
in
g
be
ha
vi
or
of
each
c
us
to
m
e
r
,
de
te
r
m
in
in
g
th
e
r
e
c
e
nt
pur
c
ha
s
e
d
a
te
(
R
e
c
e
n
c
y:
R
)
,
th
e
f
r
e
que
nc
y
of
pur
c
ha
s
e
s
(
F
r
e
que
nc
y:
F
)
,
a
nd
th
e
a
m
ount
of
m
o
ne
y
s
pe
nt
by
th
e
c
us
to
m
e
r
(
M
one
ta
r
y:
M)
[
5]
.
T
he
r
e
s
ul
ts
of
th
e
R
F
M
a
na
ly
s
is
w
e
r
e
us
e
d
to
s
e
gm
e
nt
c
us
to
m
e
r
s
us
in
g
c
lu
s
te
r
in
g
a
lg
or
it
hm
s
in
th
e
a
lg
or
it
h
m
c
om
pa
r
is
on
s
te
p
(
i.
e
.,
K
-
m
e
a
ns
,
S
O
N
N
,
a
nd
f
uz
z
y
C
-
m
e
a
n
s
)
.
In
th
e
a
lg
or
it
hm
c
om
pa
r
is
on
s
t
e
p,
we
us
e
d
5
c
r
it
e
r
ia
to
c
om
pa
r
e
th
e
a
lg
or
it
hm
s
to
f
in
d
th
e
m
os
t
s
ui
ta
bl
e
a
lg
or
it
hm
.
T
he
s
e
c
r
it
e
r
ia
w
e
r
e
S
C
,
CH
in
de
x,
DB
in
de
x,
it
e
r
a
ti
on
c
ount
,
a
nd
e
xe
c
ut
io
n
time
[
21]
,
[
22]
.
SC
is
ba
s
e
d
on
th
e
di
s
ta
n
c
e
be
twe
e
n
gr
oups
a
nd
th
e
di
s
ta
nc
e
be
tw
e
e
n
poi
nt
s
in
th
e
gr
oup,
gi
vi
ng
va
lu
e
s
in
th
e
r
a
nge
[
-
1,
1
]
.
If
th
is
va
lu
e
is
hi
gh
th
a
t
in
di
c
a
te
s
s
ui
ta
bl
e
c
lu
s
te
r
in
g,
but
if
th
is
va
lu
e
is
lo
w
or
ne
ga
ti
ve
th
a
t
in
di
c
a
te
s
uns
ui
ta
bl
e
c
lu
s
t
e
r
in
g.
CH
in
de
x
is
c
a
lc
ul
a
te
d
to
a
s
s
e
s
s
d
e
ns
e
a
nd
s
pa
r
s
e
p
a
r
ts
in
a
gr
oup.
If
th
is
va
lu
e
is
hi
gh
th
a
t
gr
oup
is
de
ns
e
a
nd
w
e
ll
-
s
pa
c
e
d
:
th
e
hi
ghe
r
,
th
e
be
tt
e
r
.
T
hi
s
is
th
e
oppos
it
e
of
th
e
DB
in
de
x,
f
or
w
hi
c
h
a
lo
w
va
lu
e
in
d
ic
a
te
s
th
a
t
th
e
gr
o
up
is
de
ns
e
a
nd
w
e
ll
-
s
pa
c
e
d
[
22]
.
We
us
e
d
'
s
kl
e
a
r
n.c
lu
s
te
r
'
f
or
K
-
m
e
a
ns
c
lu
s
te
r
in
g
a
nd
to
f
in
d
its
it
e
r
a
ti
o
n
c
ount
.
F
or
f
uz
z
y
C
-
m
e
a
ns
c
lu
s
te
r
in
g
a
nd
its
it
e
r
a
ti
on
c
ount
,
we
r
e
li
e
d
on
'
f
c
m
e
a
ns
'
.
M
e
a
nw
hi
le
,
'
m
in
is
om
'
w
a
s
us
e
d
f
or
S
O
N
N
c
lu
s
te
r
in
g
a
nd
its
it
e
r
a
ti
on
c
ount
.
T
he
e
xe
c
ut
io
n
time
of
each
c
lu
s
te
r
in
g
c
om
e
s
f
r
om
th
e
ti
m
e
di
s
pl
a
ye
d
on
th
e
c
ode
e
di
to
r
p
r
ogr
a
m
.
F
o
r
c
om
put
in
g
S
C
,
CH
in
de
x,
a
nd
DB
in
de
x,
we
ut
il
iz
e
d
'
s
kl
e
a
r
n.m
e
tr
ic
s
'
.
2.3.2.
C
u
s
t
om
e
r
c
la
s
s
i
f
ic
at
io
n
A
f
te
r
s
e
le
c
ti
ng
th
e
m
os
t
s
ui
ta
bl
e
c
lu
s
te
r
in
g
a
lg
or
it
hm
,
we
us
e
d
R
F
M
va
lu
e
s
in
e
a
c
h
c
lu
s
t
e
r
to
na
m
e
th
e
c
us
to
m
e
r
s
e
gm
e
nt
s
a
c
c
or
di
ng
to
th
e
c
us
to
m
e
r
c
la
s
s
if
ic
a
ti
on
ta
bl
e
[
23]
a
s
s
how
n i
n
T
a
bl
e
1
.
S
un
et
al
.
[
23
]
id
e
nt
if
ie
d
e
ig
ht
ty
pe
s
of
c
us
to
m
e
r
s
,
w
hi
c
h
a
r
e
im
po
r
ta
nt
va
lu
e
,
im
por
ta
nt
de
ve
lo
pm
e
nt
,
im
po
r
ta
nt
pr
ot
e
c
ti
on
,
im
por
ta
nt
r
e
te
nt
io
n,
ge
ne
r
a
l
va
lu
e
,
ge
ne
r
a
l
de
ve
lo
pm
e
nt
,
g
e
n
e
r
a
l
r
e
te
nt
io
n,
a
nd
lo
s
t.
T
h
e
s
e
ty
pe
s
e
xhi
bi
t
di
f
f
e
r
e
nt
pur
c
ha
s
in
g
be
ha
vi
or
s
,
as
s
ho
w
n
in
T
a
bl
e
1.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
, V
ol
.
14
, N
o.
3
,
J
une
20
25
:
1919
-
1929
1922
T
a
bl
e
1.
C
lu
s
te
r
na
m
in
g
ba
s
e
d
on
th
e
R
F
M
va
lu
e
s
C
us
t
om
e
r
t
ype
R
e
c
e
nc
y
F
r
e
que
nc
y
M
one
t
a
r
y
I
m
por
t
a
nt
va
l
ue
H
i
gh
H
i
gh
H
i
gh
I
m
por
t
a
nt
de
ve
l
opm
e
nt
H
i
gh
L
ow
H
i
gh
I
m
por
t
a
nt
pr
ot
e
c
t
i
on
L
ow
H
i
gh
H
i
gh
I
m
por
t
a
nt
r
e
t
e
nt
i
on
L
ow
L
ow
H
i
gh
G
e
ne
r
a
l
va
l
ue
H
i
gh
H
i
gh
L
ow
G
e
ne
r
a
l
de
ve
l
opm
e
nt
H
i
gh
L
ow
L
ow
G
e
ne
r
a
l
r
e
t
e
nt
i
on
L
ow
H
i
gh
L
ow
L
os
t
L
ow
L
ow
L
ow
F
r
om
th
e
c
us
to
m
e
r
c
la
s
s
if
ic
a
ti
on
a
s
in
T
a
bl
e
1
,
we
can
s
e
e
th
a
t
th
e
R
F
M
va
lu
e
s
w
e
r
e
qu
a
nt
iz
e
d
to
o
hi
gh
or
lo
w
.
T
he
s
e
va
lu
e
s
can
be
c
a
lc
ul
a
te
d
us
in
g
th
e
R
F
M
va
lu
e
s
of
a
c
lu
s
te
r
to
de
te
r
m
in
e
w
he
th
e
r
th
e
y
a
r
e
hi
gh
or
lo
w
by
c
om
pa
r
in
g
th
e
m
to
th
e
a
ve
r
a
ge
s
in
R
F
M
va
lu
e
s
ove
r
th
a
t
s
a
m
e
c
lu
s
te
r
.
T
h
e
r
e
c
e
nc
y
is
hi
gh
w
he
n
th
a
t
va
lu
e
is
le
s
s
th
a
n
th
e
a
ve
r
a
ge
r
e
c
e
nc
y
(R
a
vg
).
T
he
f
r
e
que
nc
y
a
nd
m
one
ta
r
y
va
lu
e
s
a
r
e
hi
gh
w
he
n
th
e
s
e
e
xc
e
e
d
th
e
a
ve
r
a
g
e
s
of
f
r
e
que
nc
y
(F
a
vg
)
a
nd
m
one
ta
r
y
(M
a
vg
)
as
in
(
1)
,
(
2)
,
a
nd
(
3)
[
24]
:
=
{
ℎ
,
<
,
≥
(
1)
=
{
ℎ
,
>
,
≤
(
2)
=
{
ℎ
,
>
,
≤
(
3)
−
K
-
m
e
a
ns
a
lg
or
it
hm
:
K
-
m
e
a
ns
is
an
uns
upe
r
vi
s
e
d
le
a
r
ni
ng
a
l
gor
it
hm
us
e
d
to
c
lu
s
te
r
da
ta
in
to
gr
oups
ba
s
e
d
on
th
e
ir
s
im
il
a
r
it
y.
It
w
or
ks
by
f
ir
s
t
in
it
ia
li
z
in
g
k
c
e
nt
r
oi
ds
,
w
he
r
e
k
is
th
e
num
be
r
of
c
lu
s
t
e
r
s
de
s
ir
e
d.
T
he
n,
it
c
a
lc
ul
a
t
e
s
th
e
di
s
ta
nc
e
be
twe
e
n
each
da
ta
p
oi
nt
a
nd
th
e
c
e
nt
r
oi
ds
to
a
s
s
ig
n
th
e
da
ta
poi
nt
s
to
th
e
c
lu
s
te
r
w
it
h
th
e
m
in
im
um
di
s
ta
nc
e
.
T
h
e
c
e
nt
r
o
id
s
a
r
e
th
e
n
upd
a
te
d,
a
nd
th
e
pr
oc
e
s
s
is
r
e
pe
a
te
d
unt
il
th
e
c
e
nt
r
oi
ds
no
lo
nge
r
c
ha
nge
[
19]
.
−
S
O
N
N
a
lg
or
it
hm
:
S
O
N
N
is
a
s
ubt
ype
of
a
r
ti
f
ic
ia
l
ne
ur
a
l
ne
twor
ks
(
A
N
N
s
)
c
a
ll
e
d
s
e
lf
-
or
ga
ni
z
in
g
m
a
p
(
S
O
M
)
or
K
ohone
n
M
a
p
.
T
hi
s
is
an
uns
upe
r
vi
s
e
d
le
a
r
ni
ng
a
lg
or
it
hm
a
nd
us
e
s
c
om
pe
ti
ti
ve
le
a
r
ni
ng
in
tr
a
in
in
g
th
e
a
dj
us
ta
bl
e
pa
r
a
m
e
te
r
s
.
A
S
O
M
s
how
s
da
ta
c
lu
s
te
r
i
ng
by
r
e
duc
in
g
th
e
hi
gh
-
di
m
e
ns
io
na
l
da
ta
to
a
lo
w
di
m
e
ns
io
n.
In
th
e
s
tr
uc
tu
r
e
of
th
is
a
lg
or
it
h
m
,
each
in
put
ne
ur
on
in
th
e
in
put
la
ye
r
is
c
onne
c
te
d
to
a
ll
ne
ur
ons
in
th
e
out
put
l
a
ye
r
(
K
ohone
n
la
ye
r
)
th
r
ough
w
e
ig
h
te
d
c
onne
c
ti
on
s
.
W
he
n
da
ta
a
r
e
f
e
d
to
an
in
put
ne
ur
on,
th
e
y
a
r
e
tr
a
ns
m
it
te
d
to
a
ll
ne
ur
ons
in
th
e
out
put
la
ye
r
,
e
s
s
e
nt
ia
ll
y
br
oa
dc
a
s
ti
ng
th
e
s
a
m
e
in
f
or
m
a
ti
on.
A
m
ong
th
e
ne
ur
ons
in
th
e
K
ohone
n
la
ye
r
,
th
e
one
w
it
h
th
e
hi
ghe
s
t
w
e
ig
ht
is
c
hos
e
n
as
th
e
out
put
f
or
c
us
to
m
e
r
s
e
gm
e
nt
a
ti
on.
T
hi
s
m
e
a
ns
th
e
ne
ur
on
w
it
h
th
e
s
tr
onge
s
t
c
onne
c
ti
on
to
th
e
in
put
da
ta
is
c
ons
id
e
r
e
d
th
e
w
in
ne
r
a
nd
it
s
a
c
ti
va
ti
on
de
te
r
m
in
e
s
th
e
c
us
to
m
e
r
s
e
gm
e
nt
[
20]
.
−
F
uz
z
y
C
-
m
e
a
ns
a
lg
or
it
hm
:
f
uz
z
y
C
-
m
e
a
ns
is
an
uns
upe
r
vi
s
e
d
a
lg
or
it
hm
.
W
he
n
c
lu
s
te
r
in
g,
each
obj
e
c
t
in
th
e
c
lu
s
te
r
ha
s
a
w
e
ig
ht
in
g
a
s
s
oc
ia
te
d
w
it
h
a
pa
r
ti
c
ul
a
r
c
lu
s
te
r
.
T
he
r
e
f
or
e
,
s
om
e
obj
e
c
ts
m
a
y
be
in
c
lu
s
te
r
i
,
but
th
e
ir
pos
it
io
n
is
not
in
th
is
c
lu
s
te
r
i
.
T
he
pr
in
c
ip
le
of
th
is
a
lg
or
it
hm
is
to
e
m
pha
s
iz
e
th
e
unc
e
r
ta
in
ty
of
m
e
m
be
r
s
hi
p
in
a
gr
oup
by
u
s
in
g
li
ke
li
hood
to
de
te
r
m
in
e
w
hi
c
h
gr
oup
each
obj
e
c
t
is
li
ke
ly
to
be
lo
ng
to
.
T
hi
s
is
done
by
c
a
lc
ul
a
ti
ng
th
e
w
e
ig
ht
or
pr
oba
bi
li
ty
of
m
e
m
be
r
s
hi
p
in
each
gr
oup
,
c
ons
id
e
r
in
g
th
e
di
s
ta
nc
e
be
tw
e
e
n
th
e
obj
e
c
t
a
nd
th
e
c
e
nt
r
oi
d
of
th
e
gr
oup
in
each
it
e
r
a
ti
on
of
th
e
a
lg
or
it
hm
.
T
he
us
e
of
w
e
ig
ht
s
a
ll
ow
s
s
p
e
c
if
yi
ng
th
e
de
gr
e
e
to
w
hi
c
h
each
obj
e
c
t
be
lo
ngs
to
a
pa
r
ti
c
ul
a
r
gr
oup,
r
a
th
e
r
th
a
n
f
or
c
in
g
th
e
m
to
be
lo
ng
e
xc
lu
s
iv
e
ly
to
one
gr
oup
[
21]
.
2.3.3.
M
od
e
l
e
val
u
at
io
n
To
id
e
nt
if
y
th
e
opt
im
a
l
c
lu
s
te
r
in
g
a
lg
or
it
hm
f
or
c
us
to
m
e
r
s
e
gm
e
nt
a
ti
on,
we
c
om
pa
r
e
d
s
e
v
e
r
a
l
m
e
tr
ic
s
in
c
lu
di
ng
S
C
,
CH
in
de
x,
DB
in
de
x,
it
e
r
a
ti
on
c
ount
,
a
n
d
e
xe
c
ut
io
n
time
[
19]
.
T
h
e
va
lu
e
s
f
r
om
th
e
s
e
e
va
lu
a
ti
on
c
r
it
e
r
ia
he
lp
us
m
a
k
e
a
ppr
opr
ia
te
de
c
is
io
ns
w
he
n
s
e
le
c
ti
ng
c
lu
s
te
r
in
g
a
lg
or
it
hm
s
.
We
can
obt
a
in
th
e
S
C
,
CH
in
de
x,
a
nd
DB
in
d
e
x
a
c
c
or
di
ng
to
th
e
c
a
l
c
ul
a
ti
ons
in
(
4)
to
(
6)
:
=
(
−
)
(
,
)
,
>
(
4)
W
he
r
e
r
e
pr
e
s
e
nt
s
th
e
SC
va
lu
e
;
is
th
e
a
ve
r
a
ge
di
s
ta
nc
e
be
twe
e
n
obj
e
c
t
i
a
nd
ot
he
r
obj
e
c
ts
in
th
e
c
lu
s
te
r
;
is
th
e
a
ve
r
a
ge
di
s
t
a
nc
e
be
tw
e
e
n
obj
e
c
t
i
a
nd
a
not
he
r
c
lu
s
te
r
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
C
us
to
m
e
r
s
e
gm
e
nt
at
io
n u
s
in
g as
s
o
c
ia
ti
on r
ul
e
m
in
in
g on r
e
ta
il
tr
ans
ac
ti
on data
(
Si
r
iw
an
K
aj
o
r
nk
as
i
r
at
)
1923
=
∑
∑
∙
−
1
−
(
5)
W
he
r
e
r
e
pr
e
s
e
nt
s
th
e
CH
in
de
x
v
a
lu
e
;
is
th
e
in
te
r
-
c
lu
s
te
r
di
s
pe
r
s
io
n
m
a
tr
ix
;
is
th
e
in
tr
a
-
c
lu
s
te
r
di
s
pe
r
s
io
n
m
a
tr
ix
;
is
th
e
num
be
r
of
gr
oupe
d
s
a
m
pl
e
s
;
is
th
e
c
l
us
te
r
num
be
r
.
=
1
∑
≠
{
(
)
+
(
)
(
,
)
}
=
1
(
6)
W
he
r
e
r
e
pr
e
s
e
nt
s
th
e
DB
in
de
x
va
lu
e
;
is
th
e
c
lu
s
te
r
num
be
r
;
a
nd
a
r
e
th
e
c
lu
s
te
r
la
be
l;
(
)
a
nd
(
)
a
r
e
a
ll
s
a
m
pl
e
s
in
c
lu
s
te
r
s
a
nd
c
lu
s
t
e
r
to
th
e
ir
r
e
s
pe
c
ti
ve
c
lu
s
te
r
c
e
nt
r
oi
ds
;
(
,
)
is
th
e
di
s
ta
nc
e
be
twe
e
n
c
lu
s
te
r
s
a
nd
c
lu
s
te
r
.
2.3.4.
A
s
s
oc
ia
t
io
n
r
u
le
m
in
in
g
We
ut
il
i
z
e
d
th
e
'
m
l
xt
e
n
d.f
r
e
qu
e
nt
_
pa
tt
e
r
ns
'
m
od
ul
e
f
or
c
on
d
uc
ti
ng
th
i
s
pr
oc
e
s
s
.
A
f
t
e
r
a
s
s
ig
ni
ng
na
m
e
s
to
th
e
c
us
to
m
e
r
s
e
gm
e
nt
s
,
we
pe
r
f
or
m
e
d
a
s
s
oc
i
a
ti
on
r
u
le
m
in
in
g
us
i
ng
FP
-
G
r
ow
th
a
lg
or
i
th
m
on
t
he
c
us
t
om
e
r
tr
a
n
s
a
c
ti
on
d
a
ta
to
f
in
d
a
s
s
o
c
ia
ti
ons
b
e
tw
e
e
n
th
e
r
e
la
te
d
pr
oduc
ts
in
e
a
c
h
s
e
gm
e
nt
.
S
in
c
e
th
e
tr
a
ns
a
c
ti
on
da
t
a
f
or
e
a
c
h
c
us
t
om
e
r
s
e
gm
e
nt
is
ve
r
y
l
a
r
ge
,
we
s
e
le
c
te
d
a
s
a
m
pl
e
of
1,000
r
e
c
or
d
s
w
it
h
th
e
hi
ghe
s
t
c
us
t
om
e
r
s
p
e
ndi
ng
to
a
na
l
yz
e
.
P
r
odu
c
t
s
th
a
t
w
e
r
e
p
u
r
c
ha
s
e
d
w
e
r
e
a
s
s
ig
ne
d
a
va
l
ue
of
t
r
ue
,
w
hi
l
e
pr
oduc
t
s
th
a
t
w
e
r
e
not
pur
c
h
a
s
e
d
w
e
r
e
a
s
s
i
gne
d
a
va
lu
e
of
f
a
l
s
e
.
We
s
e
t
th
e
s
upp
or
t
a
n
d
c
onf
id
e
nc
e
th
r
e
s
ho
ld
s
to
be
at
l
e
a
s
t
0.
01
a
nd
0.
57,
r
e
s
p
e
c
ti
ve
ly
.
C
o
nf
id
e
nc
e
a
nd
s
u
ppor
t
w
e
r
e
c
a
lc
ul
a
te
d
w
it
h
th
e
(
7)
a
nd
(
8)
[
13]
.
=
ℎ
,
ℎ
(
7)
(
)
=
ℎ
(
8)
3.
R
E
S
U
L
T
S
AND
D
I
S
C
U
S
S
I
O
N
3.1.
R
e
s
u
lt
s
F
or
th
e
s
a
le
s
da
t
a
f
r
om
2021
-
2023,
th
e
R
F
M
in
de
xe
s
w
e
r
e
us
e
d
w
it
h
K
-
m
e
a
ns
,
f
uz
z
y
C
-
m
e
a
ns
,
a
nd
S
O
N
N
a
lg
or
it
hm
s
f
or
c
us
to
m
e
r
s
e
gm
e
nt
a
ti
on
by
c
lu
s
te
r
in
g.
We
f
ound
th
a
t
th
e
pr
e
f
e
r
r
e
d
num
be
r
of
c
lu
s
te
r
s
in
th
e
s
e
a
lg
or
it
hm
s
w
a
s
3
(
F
ig
ur
e
2)
.
T
he
K
-
m
e
a
ns
di
d
c
lu
s
te
r
by
c
e
nt
r
oi
d
poi
nt
s
.
T
he
f
uz
z
y
C
-
m
e
a
ns
a
ls
o
c
lu
s
te
r
e
d
by
c
e
nt
r
oi
d
poi
nt
s
li
ke
K
-
m
e
a
ns
,
a
nd
c
a
lc
ul
a
te
d
th
e
pr
oba
bi
li
ty
f
or
a
c
lu
s
te
r
pos
it
io
n.
T
he
S
O
N
N
a
lg
or
it
hm
di
d
c
lu
s
te
r
in
g
by
m
a
ppi
ng.
F
r
om
th
e
s
c
a
tt
e
r
pl
ot
s
in
F
ig
ur
e
2,
it
can
be
obs
e
r
ve
d
th
a
t
th
e
r
e
w
a
s
no
s
ig
ni
f
ic
a
nt
di
f
f
e
r
e
nc
e
in
th
e
c
lu
s
te
r
in
g
r
e
s
ul
ts
b
e
twe
e
n
K
-
m
e
a
ns
(
F
ig
ur
e
2
(
a
)
)
a
nd
S
O
N
N
c
lu
s
te
r
in
g
(
F
ig
ur
e
2
(
b
)
)
a
lg
or
i
th
m
s
.
On
th
e
ot
he
r
ha
nd,
f
uz
z
y
C
-
m
e
a
ns
c
lu
s
te
r
in
g
(
F
ig
ur
e
2
(
c
)
)
s
how
s
di
f
f
e
r
e
nt
r
e
s
ul
ts
c
om
pa
r
e
d
to
th
e
pr
e
vi
ous
two
a
lg
or
it
hm
s
.
T
he
c
lu
s
t
e
r
pos
it
io
ns
of
th
e
c
us
to
m
e
r
s
w
e
r
e
s
w
it
c
he
d.
H
ow
e
ve
r
,
th
e
va
lu
e
s
of
S
C
,
CH
in
de
x,
DB
in
de
x,
it
e
r
a
ti
on
c
ount
,
a
nd
e
xe
c
ut
io
n
time
of
a
ll
th
r
e
e
a
lg
or
it
hm
s
c
le
a
r
ly
de
m
ons
tr
a
te
s
om
e
di
f
f
e
r
e
nc
e
s
in
c
lu
s
te
r
in
g
pe
r
f
or
m
a
nc
e
.
T
he
c
om
pa
r
is
on
of
c
lu
s
te
r
in
g
a
lg
or
it
hm
s
w
it
h
th
e
S
C
,
C
H
in
de
x,
a
nd
D
B
in
de
x
s
how
s
th
a
t
th
e
K
-
m
e
a
ns
,
f
uz
z
y
C
-
m
e
a
ns
,
a
nd
S
O
N
N
a
r
e
s
ui
ta
bl
e
f
or
s
e
gm
e
nt
a
ti
on
to
3
gr
oups
by
th
e
c
ount
of
w
in
s
a
c
r
os
s
th
e
c
om
pa
r
is
on
c
r
it
e
r
ia
(
two
c
lu
s
te
r
s
w
in
s
3
ti
m
e
s
;
th
r
e
e
c
lu
s
te
r
s
w
in
6
ti
m
e
s
)
.
T
he
hi
ghe
s
t
S
C
a
nd
C
H
in
de
xe
s
a
nd
th
e
lo
w
e
s
t
D
B
in
de
x
w
e
r
e
a
c
hi
e
ve
d
by
th
e
K
-
m
e
a
ns
a
lg
or
it
hm
a
s
pr
e
s
e
nt
e
d
in
T
a
bl
e
2.
T
he
it
e
r
a
ti
on
a
nd
e
xe
c
ut
io
n
ti
m
e
f
or
K
-
m
e
a
ns
c
lu
s
te
r
in
g
w
a
s
lo
w
e
r
th
a
n
f
or
th
e
f
uz
z
y
C
-
m
e
a
ns
a
nd
S
O
N
N
(K
-
m
e
a
ns
:
it
e
r
a
ti
on
c
ount
w
a
s
3 a
nd
e
xe
c
ut
io
n
ti
m
e
w
a
s
0.607
6
s
e
c
onds
;
f
uz
z
y
C
-
m
e
a
n
s
:
it
e
r
a
ti
on
c
ount
w
a
s
150 a
nd e
xe
c
ut
io
n t
im
e
2.5954 s
e
c
onds
;
S
O
N
N
:
it
e
r
a
ti
on c
ount
w
a
s
45 a
nd e
x
e
c
ut
io
n t
im
e
3.2230 s
e
c
ond
s
)
.
T
he
r
e
f
or
e
,
th
e
K
-
m
e
a
ns
a
lg
or
it
hm
pe
r
f
or
m
e
d
be
tt
e
r
th
a
n
f
uz
z
y
C
-
m
e
a
ns
a
nd
S
O
N
N
in
c
us
to
m
e
r
s
e
gm
e
nt
a
ti
on
of
th
e
s
e
da
ta
.
T
he
num
be
r
of
c
lu
s
te
r
s
ba
s
e
d
on
th
e
R
F
M
in
K
-
m
e
a
ns
c
lu
s
te
r
in
g
w
a
s
k
-
3.
T
he
c
us
to
m
e
r
pur
c
ha
s
e
be
ha
vi
or
a
s
pr
e
s
e
nt
e
d
in
F
ig
ur
e
3
i
ndi
c
a
to
r
s
r
e
c
e
nc
y
(
F
ig
ur
e
3(
a
)
)
,
f
r
e
que
nc
y
(
F
ig
ur
e
3(
b)
)
, a
nd mone
ta
r
y (
F
ig
u
r
e
3(
c
)
)
a
r
e
gr
a
phi
c
a
ll
y i
ll
us
tr
a
te
d f
or
e
a
c
h c
lu
s
te
r
.
W
e
us
e
d
R
F
M
va
lu
e
s
f
r
om
th
e
c
us
to
m
e
r
pur
c
h
a
s
e
be
ha
vi
or
o
f
e
a
c
h
c
lu
s
te
r
(
F
ig
ur
e
3)
to
c
a
lc
ul
a
te
th
e
a
ve
r
a
ge
R
F
M
va
lu
e
s
f
or
a
c
om
pa
r
is
on
be
twe
e
n
gr
oups
.
T
he
a
ve
r
a
ge
r
e
c
e
nc
y
va
lu
e
is
223,
th
e
a
v
e
r
a
ge
f
r
e
que
nc
y
va
lu
e
is
7,
a
nd
th
e
a
ve
r
a
ge
m
one
t
a
r
y
va
lu
e
is
1,351 a
s
pr
e
s
e
nt
e
d
in
T
a
bl
e
3.
W
e
u
s
e
d
c
om
pa
r
is
on
s
be
twe
e
n
R
F
M
va
lu
e
a
nd
a
v
e
r
a
ge
R
F
M
va
lu
e
to
la
be
l
th
e
c
u
s
t
om
e
r
gr
oups
.
C
lu
s
te
r
1
is
hi
gh
in
r
e
c
e
nc
y
but
lo
w
in
f
r
e
que
nc
y
a
nd
m
one
ta
r
y
va
lu
e
s
,
s
o
it
is
th
e
ge
ne
r
a
l
de
ve
lo
pm
e
nt
gr
oup.
C
lu
s
te
r
2
is
h
ig
h
in
a
ll
R
F
M
va
lu
e
s
a
nd i
s
t
h
e
i
m
por
ta
nt
va
lu
e
gr
oup. C
lu
s
te
r
3 i
s
l
ow
i
n R
F
M
va
lu
e
s
a
nd i
s
l
a
be
le
d t
he
l
o
s
t
gr
oup.
T
he
a
na
ly
s
is
r
e
s
ul
ts
of
our
gr
oup
m
e
m
be
r
s
s
how
th
a
t
T
u
e
nj
a
i
C
om
pa
ny
ha
s
th
e
hi
ghe
s
t
num
be
r
of
c
us
to
m
e
r
s
in
th
e
ge
ne
r
a
l
de
v
e
lo
pm
e
nt
c
a
te
gor
y,
r
a
nki
ng
f
ir
s
t
a
m
ong
a
ll
c
u
s
to
m
e
r
s
.
L
os
t
r
a
nk
s
s
e
c
ond,
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
, V
ol
.
14
, N
o.
3
,
J
une
20
25
:
1919
-
1929
1924
f
ol
lo
w
e
d
by
im
por
ta
nt
va
lu
e
a
s
th
ir
d.
T
he
la
be
ls
gi
ve
n
to
e
a
c
h
c
lu
s
te
r
s
um
m
a
r
iz
e
th
e
c
us
to
m
e
r
pur
c
ha
s
in
g
be
ha
vi
or
s
ba
s
e
d on F
ig
ur
e
3. D
e
ta
il
e
d i
nf
or
m
a
ti
on on e
a
c
h gr
ou
p c
a
n be
s
e
e
n i
n T
a
bl
e
4.
(
a
)
(
b)
(
c
)
F
ig
ur
e
2.
C
us
to
m
e
r
s
e
gm
e
nt
a
ti
on
by
(
a
)
K
-
m
e
a
ns
,
(
b)
S
O
N
N
,
a
nd
(
c
)
f
uz
z
y
C
-
m
e
a
ns
c
lu
s
te
r
in
g
a
lg
or
it
hm
s
T
a
bl
e
2.
A
c
om
pa
r
is
on
of
c
lu
s
te
r
in
g
a
lg
or
it
hm
s
w
it
h
th
e
S
C
,
CH
in
de
x,
a
nd
DB
in
de
x
T
he
num
be
r
of
c
l
us
t
e
r
s
K
-
m
e
a
ns
F
uz
z
y
C
-
m
e
a
ns
S
O
N
N
SC
CH
i
nde
x
DB
i
nde
x
SC
CH
i
nde
x
DB
i
nde
x
SC
CH
i
nde
x
DB
i
nde
x
2
0.4625
156318
0.8914
0.4494
155082
0.9205
0.4804
149416
0.8410
3
0.4364
184977
0.7913
0.4355
184920
0.7943
0.4361
181914
0.7931
4
0.4008
173952
0.9287
0.3947
173595
0.9253
0.3856
169578
0.9287
5
0.3495
157078
0.9691
0.3150
147978
1.1130
0.3489
149227
0.9839
6
0.3502
150646
1.0086
0.3276
144799
1.1020
0.3432
136502
0.9891
(
a
)
(
b)
(
c
)
F
ig
ur
e
3.
C
us
to
m
e
r
pur
c
ha
s
e
be
ha
vi
or
s
by
c
lu
s
te
r
(
a
)
r
e
c
e
nc
y, (
b)
f
r
e
que
nc
y, a
nd (
c
)
m
one
ta
r
y
T
a
bl
e
3.
A
c
om
pa
r
is
on
b
e
twe
e
n
R
F
M
va
lu
e
s
in
each
c
lu
s
te
r
a
n
d
a
ve
r
a
ge
R
F
M
va
lu
e
s
C
l
us
t
e
r
R
e
c
e
nc
y
vs
a
ve
r
a
ge
r
e
c
e
n
c
y
F
r
e
que
nc
y
vs
a
ve
r
a
ge
f
r
e
que
nc
y
M
one
t
a
r
y
vs
a
ve
r
a
ge
m
one
t
a
r
y
1
H
i
gh
(
110<223)
L
ow
(
4<7)
L
ow
(
721<1,351)
2
H
i
gh
(
83<223)
H
i
gh
(
14>7)
H
i
gh
(
2,790>1,351)
3
L
ow
(
478>223)
L
ow
(
2<7)
L
ow
(
543<1,351)
T
a
bl
e
4.
C
u
s
to
m
e
r
s
in
each
gr
oup
G
r
oup
N
um
be
r
of
pe
opl
e
(%)
A
ve
r
a
ge
l
a
s
t
pur
c
ha
s
e
A
ve
r
a
ge
pur
c
ha
s
e
f
r
e
que
nc
y
(
t
i
m
e
s
)
E
xpe
ns
e
s
a
ve
r
a
ge
(
T
H
B
)
I
m
por
t
a
nt
va
l
ue
39,440
(
23)
A
bout
2
-
3
m
ont
hs
a
go
14
2,790
L
os
t
47,205
(
27)
A
bout
1
-
2
ye
a
r
s
a
go
2
543
G
e
ne
r
a
l
de
ve
l
opm
e
nt
86,616
(
50)
A
bout
3
-
4
m
ont
hs
a
go
4
721
F
or
th
e
a
s
s
oc
i
a
ti
on
r
ul
e
m
in
in
g,
w
e
s
e
le
c
t
e
d
1,000 s
a
m
pl
e
s
f
r
om
e
a
c
h
c
lu
s
te
r
ba
s
e
d
on
to
p
m
one
ta
r
y
va
lu
e
.
W
e
s
e
t
th
e
m
in
im
um
s
uppor
t
to
0.01
a
nd
c
onf
id
e
nc
e
a
t
0.57.
A
f
te
r
a
s
s
o
c
ia
ti
on
r
ul
e
a
na
ly
s
is
of
e
a
c
h
c
lu
s
te
r
,
th
e
r
e
w
e
r
e
38
r
u
le
s
f
or
th
e
im
por
ta
nt
va
lu
e
gr
oup
a
s
s
how
n
in
T
a
bl
e
5,
two
r
ul
e
s
f
or
th
e
ge
ne
r
a
l
de
ve
lo
pm
e
nt
gr
oup a
s
e
xpl
a
in
e
d i
n
T
a
bl
e
6, a
nd t
w
o r
ul
e
s
f
or
t
he
l
os
t
gr
oup a
s
pr
e
s
e
nt
e
d i
n T
a
bl
e
7.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
C
us
to
m
e
r
s
e
gm
e
nt
at
io
n u
s
in
g as
s
o
c
ia
ti
on r
ul
e
m
in
in
g on r
e
ta
il
tr
ans
ac
ti
on data
(
Si
r
iw
an
K
aj
o
r
nk
as
i
r
at
)
1925
T
a
bl
e
5. A
s
s
oc
ia
ti
on r
ul
e
s
i
n t
he
i
m
por
ta
nt
va
lu
e
gr
oup
A
nt
e
c
e
de
nt
s
C
ons
e
que
nt
s
S
uppor
t
C
onf
i
de
nc
e
S
i
ngl
e
i
t
e
m
s
-
P
r
e
s
s
ur
e
a
dj
us
t
i
ng he
a
d N
o. R
326 (
S
C
G
)
-
G
a
s
c
a
bl
e
, w
hi
t
e
/
or
a
nge
,
50
m.
0.010
1.000
-
G
l
ue
gun N
o.
9049/
ST
001
-
T
r
a
ns
pa
r
e
nt
gl
ue
s
t
i
c
ks
1*12 (
S
oj
i
)
0.012
0.857
-
F
r
e
e
w
r
a
ppi
ng s
e
r
vi
c
e
-
G
i
f
t
w
r
a
ppi
ng
pa
pe
r
-
100 s
he
e
t
s
0.012
0.857
-
A
s
s
or
t
e
d pul
l
-
bow
r
i
bbons
-
G
i
f
t
w
r
a
ppi
ng pa
pe
r
-
100 s
he
e
t
s
0.031
0.795
-
G
a
s
hos
e
c
l
i
p N
o.
18
, t
hi
n t
ype
-
G
a
s
c
a
bl
e
, w
hi
t
e
/
or
a
nge
,
50
m.
0.011
0.786
-
G
a
s
r
e
gul
a
t
or
N
o.
339/
889 (
L
G
A
)
-
G
a
s
c
a
bl
e
, w
hi
t
e
/
or
a
nge
,
50
m.
0.010
0.769
-
T
hi
c
k ga
s
hos
e
c
l
i
p
-
G
a
s
c
a
bl
e
, w
hi
t
e
/
or
a
nge
,
50
m.
0.021
0.750
-
A
s
s
or
t
e
d pul
l
bow
r
i
bbons
-
G
i
f
t
w
r
a
ppi
ng pa
pe
r
-
100 s
he
e
t
s
0.028
0.718
-
B
ox w
i
t
h l
oc
ki
ng l
i
d
60 (
55)
l
., bl
a
c
k (
C
N
N
)
-
B
ox w
i
t
h l
oc
ki
ng l
i
d
60 (
55)
l
.,
a
s
s
or
t
e
d c
ol
or
s
(
C
N
N
)
0.020
0.606
-
C
hi
l
dr
e
n'
s
c
l
ot
he
s
ha
nge
r
w
i
nne
r
1*6
-
C
l
ot
he
s
ha
nge
r
N
o.
3 (
E
a
gl
e
br
a
nd)
0.013
0.591
-
D
i
a
gona
l
c
ut
t
i
ng kni
f
e
6.5 i
n. N
o.171P
(
K
i
w
i
br
a
nd)
-
B
i
g s
poon (
z
e
br
a
he
a
d)
0.011
0.579
D
oubl
e
i
t
e
m
s
-
T
i
s
s
ue
pa
pe
r
1*6 (
V
i
vy)
-
A
s
s
or
t
e
d pul
l
bow
r
i
bbons
-
G
i
f
t
w
r
a
ppi
ng pa
pe
r
-
100 s
he
e
t
s
0.010
1.000
-
V
i
pa
da
or
a
nge
s
oa
p m
i
xe
d w
i
t
h c
ol
l
a
ge
n
-
L
e
o s
of
t
f
a
br
i
c
s
of
t
e
ne
r
, t
ouc
h of
l
ove
s
c
e
nt
, bl
ue
,
500
ml.
-
V
i
pa
da
c
a
r
r
ot
s
oa
p 65g.
0.010
0.909
-
A
s
s
or
t
e
d pul
l
bow
r
i
bbons
-
L
a
r
ge
s
poon (
Z
e
br
a
br
a
nd)
-
G
i
f
t
w
r
a
ppi
ng pa
pe
r
-
100 s
he
e
t
s
0.015
0.882
-
D
e
P
a
r
e
e
f
a
br
i
c
s
of
t
e
ne
r
, j
oy s
c
e
nt
, r
e
f
i
l
l
ba
g
540
ml.
-
D
e
P
a
r
e
e
f
a
br
i
c
s
of
t
e
ne
r
, bl
ue
pa
r
a
di
s
e
s
c
e
nt
, r
e
f
i
l
l
ba
g t
ype
,
540
ml.
-
D
e
P
a
r
e
e
f
a
br
i
c
s
of
t
e
ne
r
, bl
oom
i
ng
s
c
e
nt
, r
e
f
i
l
l
ba
g t
ype
,
540
ml.
0.011
0.846
-
B
ox w
i
t
h l
oc
ki
ng l
i
d
60 (
55)
l
., a
s
s
or
t
e
d c
ol
or
s
(
C
N
N
)
-
A
s
s
or
t
e
d pul
l
bow
r
i
bbons
-
G
i
f
t
w
r
a
ppi
ng pa
pe
r
-
100 s
he
e
t
s
(
5
B)
0.010
0.833
-
F
a
m
ony f
a
br
i
c
s
of
t
e
ne
r
, s
w
e
e
t
be
a
ut
i
f
ul
s
c
e
nt
(
pur
pl
e
)
300
ml.
-
P
a
l
m
oi
l
1
l
. (
F
l
ow
e
r
br
a
nd)
-
F
a
r
m
ony f
a
br
i
c
s
of
t
e
ne
r
, r
o
m
a
nc
e
s
c
e
nt
(
r
e
d)
,
300
ml.
0.010
0.769
-
108
S
hop de
t
e
r
ge
nt
1000
g.
-
P
i
nt
o di
s
hw
a
s
hi
ng l
i
qui
d, l
e
m
on s
c
e
nt
450/
430/
420
ml.
-
P
a
l
m
oi
l
1
l
. (
F
l
ow
e
r
br
a
nd)
0.010
0.769
-
P
i
nt
o di
s
hw
a
s
hi
ng l
i
qui
d, l
e
m
on s
c
e
nt
450/
430/
420
ml.
-
L
e
o s
of
t
f
a
br
i
c
s
of
t
e
ne
r
, t
ouc
h of
l
ove
s
c
e
nt
, bl
ue
,
500
ml.
-
P
a
l
m
oi
l
1
l
. (
F
l
ow
e
r
br
a
nd)
0.012
0.750
-
A
s
s
or
t
e
d pul
l
bow
r
i
bbons
-
L
a
r
ge
s
poon (
Z
e
br
a
br
a
nd)
-
G
i
f
t
w
r
a
ppi
ng pa
pe
r
-
100 s
he
e
t
s
(
5
B)
0.012
0.750
-
T
i
s
s
ue
pa
pe
r
-
180
s
he
e
t
s
(
S
ol
l
y br
a
nd)
-
L
a
r
ge
pi
e
c
e
of
l
i
nol
e
um
2.70
m.
-
S
m
a
l
l
pi
e
c
e
of
l
i
nol
e
um
2.50
m.
0.011
0.733
-
L
e
o s
of
t
f
a
br
i
c
s
of
t
e
ne
r
, t
ouc
h of
l
ove
s
c
e
nt
, bl
ue
,
500
ml.
-
L
e
o s
of
t
f
a
br
i
c
s
of
t
e
ne
r
, s
w
e
e
t
f
l
or
a
l
s
c
e
nt
, pi
nk,
500
ml.
-
L
e
o S
of
t
f
a
br
i
c
s
of
t
e
ne
r
, pur
pl
e
pa
s
s
i
on s
c
e
nt
, pur
pl
e
,
500
ml.
0.019
0.731
-
B
ox w
i
t
h l
oc
ki
ng l
i
d
60 (
55)
l
., a
s
s
or
t
e
d c
ol
or
s
(
C
N
N
)
-
L
a
r
ge
pi
e
c
e
of
l
i
nol
e
um
2.70
m.
-
S
m
a
l
l
pi
e
c
e
of
l
i
nol
e
um
2.50
m.
0.019
0.679
-
P
i
nt
o di
s
hw
a
s
hi
ng l
i
qui
d, s
t
r
a
w
be
r
r
y s
c
e
nt
450/
400
ml.
-
G
a
s
gun (
T
op L
i
ght
)
-
P
a
l
m
oi
l
1
l
. (
F
l
ow
e
r
br
a
nd)
0.010
0.667
-
P
i
nt
o di
s
hw
a
s
hi
ng l
i
qui
d, ki
w
i
s
c
e
nt
,
450/
400
ml.
-
P
a
l
m
oi
l
1
l
. (
F
l
ow
e
r
br
a
nd)
-
C
ook L
e
o di
s
hw
a
s
hi
ng l
i
qui
d, l
e
m
on
s
c
e
nt
,
450/
420
ml.
0.014
0.667
-
P
a
l
m
ol
e
i
n oi
l
1
l
. (
B
e
e
br
a
nd)
-
L
i
ght
e
r
1*3 (
M
O
T
O
)
-
P
a
l
m
oi
l
1
l
. (
F
l
ow
e
r
br
a
nd)
0.015
0.652
-
V
i
vy t
i
s
s
ue
pa
pe
r
1*6
-
V
i
pa
da
c
a
r
r
ot
s
oa
p
65
g.
-
V
i
pa
da
or
a
nge
s
oa
p m
i
xe
d w
i
t
h
c
ol
l
a
ge
n
0.011
0.647
-
G
a
s
gun (
T
op L
i
ght
)
-
C
ook L
e
o di
s
hw
a
s
hi
ng l
i
qui
d, l
e
m
on s
c
e
nt
,
450/
420
ml.
-
P
a
l
m
oi
l
1
l
. (
F
l
ow
e
r
br
a
nd)
0.014
0.636
-
P
a
l
m
ol
e
i
n oi
l
1
l
. (
B
e
e
br
a
nd)
-
F
e
a
t
he
r
r
a
z
or
bl
a
de
(
F
e
a
t
he
r
)
-
P
a
l
m
oi
l
1
l
. (
F
l
ow
e
r
br
a
nd)
0.010
0.625
-
P
a
l
m
ol
e
i
n oi
l
1
l
. (
B
e
e
br
a
nd)
-
P
r
of
e
s
s
i
ona
l
G
r
e
e
n F
or
e
s
t
2700
g.
-
P
a
l
m
oi
l
1
l
. (
F
l
ow
e
r
br
a
nd)
0.010
0.625
-
W
e
t
w
i
pe
s
(
H
A
P
P
Y
H
I
P
P
O
)
-
L
a
r
ge
s
poon (
z
e
br
a
he
a
d)
-
C
l
ot
he
s
ha
nge
r
N
o.
3 (
E
a
gl
e
br
a
nd)
0.011
0.611
-
A
l
um
i
num
t
e
a
s
poon (
F
l
ow
e
r
br
a
nd)
-
V
i
vy t
i
s
s
ue
pa
pe
r
1*6
.
-
P
a
l
m
oi
l
1
l
. (
F
l
ow
e
r
br
a
nd)
0.011
0.611
-
P
i
nt
o di
s
hw
a
s
hi
ng l
i
qui
d, s
t
r
a
w
be
r
r
y s
c
e
nt
450/
400
ml.
-
C
l
ot
he
s
ha
nge
r
N
o.
3 (
E
a
gl
e
br
a
nd)
-
B
i
g s
poon (
z
e
br
a
he
a
d)
0.011
0.611
-
P
r
of
e
s
s
i
ona
l
G
r
e
e
n F
or
e
s
t
2700
g.
-
C
ook L
e
o di
s
hw
a
s
hi
ng l
i
qui
d, l
e
m
on s
c
e
nt
450/
420
ml.
-
P
a
l
m
oi
l
1
l
. (
F
l
ow
e
r
br
a
nd)
0.011
0.611
-
L
e
o S
of
t
f
a
br
i
c
s
of
t
e
ne
r
, s
w
e
e
t
f
l
or
a
l
s
c
e
nt
, pi
nk, 500m
l
.
-
C
ook
L
e
o di
s
hw
a
s
hi
ng l
i
qui
d, l
e
m
on s
c
e
nt
, 450/
420m
l
.
-
P
a
l
m
oi
l
1
l
. (
F
l
ow
e
r
br
a
nd)
0.015
0.600
-
C
ook L
e
o di
s
hw
a
s
hi
ng l
i
qui
d, l
e
m
on s
c
e
nt
,
450/
420
ml.
-
C
ook L
e
o di
s
hw
a
s
hi
ng l
i
qui
d, s
t
r
a
w
be
r
r
y ki
w
i
s
c
e
nt
450
ml.
-
P
a
l
m
oi
l
1
l
. (
F
l
ow
e
r
br
a
nd)
0.010
0.588
-
A
l
um
i
num
t
e
a
s
poon (
F
l
ow
e
r
br
a
nd)
-
C
l
ot
he
s
ha
nge
r
N
o.3 (
E
a
gl
e
br
a
nd)
-
B
ox w
i
t
h l
oc
ki
ng l
i
d
60(
55)
l
.,
a
s
s
or
t
e
d c
ol
or
s
(
C
N
N
)
0.011
0.579
-
P
i
nt
o di
s
hw
a
s
hi
ng l
i
qui
d, l
e
m
on s
c
e
nt
, 450/
430/
420m
l
.
-
L
e
o s
of
t
f
a
br
i
c
s
of
t
e
ne
r
, pur
pl
e
pa
s
s
i
on s
c
e
nt
, pur
pl
e
, 500m
l
.
-
P
a
l
m
oi
l
1
l
. (
F
l
ow
e
r
br
a
nd)
0.011
0.579
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
, V
ol
.
14
, N
o.
3
,
J
une
20
25
:
1919
-
1929
1926
T
a
bl
e
6.
A
s
s
oc
ia
ti
on
r
ul
e
s
in
th
e
g
e
ne
r
a
l
de
ve
lo
pm
e
nt
gr
oup
A
nt
e
c
e
de
nt
s
C
ons
e
que
nt
s
S
uppor
t
C
onf
i
de
nc
e
-
T
hi
c
k
ga
s
hos
e
c
l
i
p
-
G
a
s
c
a
bl
e
,
w
hi
t
e
/
or
a
nge
,
50
m.
0.012
0.706
-
6
pa
i
r
s
of
s
poons
a
nd
f
or
ks
,
f
r
a
ngi
pa
ni
pa
t
t
e
r
n
N
o.
100030
(
Z
e
br
a
br
a
nd)
-
B
i
g
s
poon
(
Z
e
br
a
br
a
nd)
0.012
0.571
T
a
bl
e
7.
A
s
s
oc
ia
ti
on
r
ul
e
s
in
th
e
lo
s
t
gr
oup
A
nt
e
c
e
de
nt
s
C
ons
e
que
nt
s
S
uppor
t
C
onf
i
de
nc
e
-
A
s
s
or
t
e
d
pul
l
bow
r
i
bbons
-
G
i
f
t
w
r
a
ppi
ng
pa
pe
r
-
100
s
he
e
t
s
0.016
0.941
-
G
a
s
c
a
bl
e
,
w
hi
t
e
/
or
a
nge
,
50
m.
-
T
hi
c
k
ga
s
hos
e
c
l
i
p
0.010
0.625
I
m
por
ta
nt
va
lu
e
:
th
e
a
s
s
oc
ia
ti
on
r
ul
e
s
s
how
th
a
t
th
e
c
us
to
m
e
r
s
in
th
is
c
lu
s
te
r
bought
a
pr
e
s
s
ur
e
a
dj
us
ti
ng
he
a
d
N
o.
R
326
(
S
C
G
)
a
nd
th
e
n
bought
a
g
a
s
c
a
bl
e
,
w
hi
te
/o
r
a
nge
,
50m
r
e
pe
a
te
dl
y.
T
he
r
e
is
100%
c
onf
id
e
nc
e
in
th
is
r
ul
e
(
T
a
bl
e
5)
.
T
hi
s
m
e
a
ns
th
a
t
e
ve
r
y
ti
m
e
c
us
to
m
e
r
s
in
th
is
gr
oup
pur
c
ha
s
e
a
pr
e
s
s
ur
e
a
dj
us
ti
ng he
a
d N
o. R
326 (
S
C
G
)
, t
he
y a
lwa
ys
buy a
ga
s
c
a
bl
e
,
w
hi
te
/o
r
a
nge
,
50
m.
G
e
ne
r
a
l
de
ve
lo
pm
e
nt
:
th
e
a
s
s
oc
ia
ti
on
r
ul
e
s
s
how
th
a
t
c
us
to
m
e
r
s
in
th
is
c
lu
s
te
r
bought
a
th
ic
k
ga
s
hos
e
c
li
p
a
nd
th
e
n
bought
a
ga
s
c
a
bl
e
,
w
hi
te
/o
r
a
nge
,
50m
,
to
o.
T
he
r
e
is
70.6%
c
onf
id
e
nc
e
in
th
is
r
ul
e
(
T
a
bl
e
6)
.
T
hi
s
m
e
a
ns
th
a
t
e
ve
r
y t
im
e
c
us
to
m
e
r
s
i
n t
hi
s
gr
oup
p
ur
c
ha
s
e
a
t
hi
c
k ga
s
hos
e
c
li
p, t
he
r
e
i
s
a
70.6
%
c
ha
nc
e
t
ha
t
th
e
y w
il
l
a
ls
o buy
a
ga
s
c
a
bl
e
, w
hi
te
/o
r
a
nge
, 50m.
L
os
t:
th
e
a
s
s
oc
ia
ti
on
r
ul
e
s
s
ho
w
th
a
t
c
us
to
m
e
r
s
in
th
is
c
lu
s
te
r
bought
a
s
s
or
te
d
pul
l
bow
r
ib
bons
a
nd
th
e
n
bought
gi
f
t
w
r
a
ppi
ng
pa
pe
r
-
100
s
he
e
ts
,
to
o.
T
he
r
e
i
s
9
4.1%
c
onf
id
e
nc
e
in
th
is
r
ul
e
(
T
a
bl
e
7)
.
T
hi
s
m
e
a
ns
th
a
t
e
v
e
r
y
ti
m
e
c
us
to
m
e
r
s
in
th
is
gr
oup
pur
c
ha
s
e
a
s
s
or
t
e
d
pul
l
bow
r
ib
bons
,
th
e
r
e
is
a
94.1
%
c
ha
n
c
e
th
a
t
th
e
y w
il
l
a
ls
o buy gif
t
w
r
a
ppi
ng pa
pe
r
-
100 s
he
e
ts
.
3.2.
D
is
c
u
s
s
io
n
In
to
da
y'
s
f
a
s
t
-
pa
c
e
d
bus
in
e
s
s
l
a
nds
c
a
p
e
,
e
f
f
e
c
ti
ve
ly
s
e
gm
e
nt
in
g
c
us
to
m
e
r
s
pl
a
y
s
a
c
r
uc
ia
l
r
ol
e
in
opt
im
iz
in
g
c
ol
la
bor
a
ti
on
w
it
h
r
e
c
om
m
e
nda
ti
on
s
ys
t
e
m
s
.
C
u
s
to
m
e
r
s
e
gm
e
nt
a
ti
on
pr
ovi
de
s
d
e
e
p
in
s
ig
ht
s
in
to
pur
c
ha
s
in
g
be
ha
vi
or
s
,
e
na
bl
in
g
or
ga
ni
z
a
ti
ons
to
c
u
s
to
m
iz
e
th
e
ir
of
f
e
r
in
gs
a
nd
s
e
r
vi
c
e
s
to
be
tt
e
r
m
e
e
t
c
us
to
m
e
r
ne
e
ds
,
ul
ti
m
a
te
ly
r
e
s
ul
ti
ng
in
he
ig
ht
e
ne
d
s
a
ti
s
f
a
c
ti
on
[
25]
.
T
he
pr
oc
e
s
s
in
vol
ve
s
e
m
pl
oyi
ng
c
lu
s
te
r
in
g
a
lg
or
it
hm
s
f
or
c
us
to
m
e
r
s
e
gm
e
nt
a
ti
on.
F
or
in
s
ta
nc
e
,
ba
nki
ng
s
e
c
to
r
us
e
s
be
h
a
vi
or
-
ba
s
e
d
s
e
gm
e
nt
a
ti
on,
r
e
s
t
a
ur
a
nt
s
u
s
e
d
e
m
ogr
a
phi
c
-
ba
s
e
d
s
e
gm
e
nt
a
ti
on,
a
nd
th
e
s
a
m
e
pr
in
c
ip
le
a
ppl
ie
s
to
s
e
gm
e
nt
a
ti
on
in
s
uppl
e
m
e
nt
pr
oduc
t
m
a
r
ke
ts
,
a
nd
to
ot
he
r
c
o
nt
e
xt
s
[
26]
–
[
28]
.
E
ve
n
th
ough
we
m
a
y
s
tu
dy
pr
e
vi
ous
r
e
s
e
a
r
c
h
on
th
e
w
or
ki
ngs
of
c
lu
s
te
r
in
g
a
lg
or
it
hm
s
,
we
c
a
nnot
be
c
e
r
ta
in
w
he
th
e
r
th
os
e
a
lg
or
it
hm
s
a
r
e
s
ui
ta
bl
e
f
or
th
e
s
a
le
s
d
a
ta
s
e
t
we
a
r
e
us
in
g
or
not
.
T
hi
s
s
tu
dy
m
e
ti
c
ul
ous
ly
c
om
pa
r
e
d
va
r
io
us
c
lu
s
te
r
in
g
a
lg
or
it
hm
s
to
de
te
r
m
in
e
th
e
m
os
t
s
ui
ta
bl
e
a
ppr
oa
c
h
f
or
c
us
to
m
e
r
s
e
gm
e
nt
a
ti
on,
ut
il
iz
in
g
tr
a
ns
a
c
ti
on
d
a
ta
f
r
om
T
ue
nj
a
i
P
a
ni
t
G
r
oup
C
o.,
L
td
.,
T
ha
il
a
nd,
s
pa
nni
ng
f
r
om
J
a
nu
a
r
y
2021
to
M
a
y
2023.
T
hr
ough
r
ig
or
ous
R
F
M
a
na
ly
s
i
s
,
we
e
v
a
lu
a
te
d
th
e
pe
r
f
or
m
a
nc
e
of
th
r
e
e
pr
om
in
e
nt
c
lu
s
te
r
in
g
a
lg
or
it
hm
s
:
K
-
m
e
a
ns
,
f
uz
z
y
C
-
m
e
a
ns
,
a
nd
S
O
N
N
,
us
in
g
ke
y
m
e
tr
ic
s
s
uc
h
as
S
C
,
CH
in
de
x,
DB
in
de
x,
it
e
r
a
ti
on
c
ount
,
a
nd
e
xe
c
ut
io
n
ti
m
e
[
21]
,
[
22]
.
N
ot
a
bl
y,
our
f
in
di
ngs
s
tr
ong
ly
f
a
vor
e
d
th
e
K
-
m
e
a
ns
a
lg
or
it
hm
due
to
its
e
xc
e
pt
io
na
l
s
pe
e
d
a
nd
a
c
c
ur
a
c
y,
pa
r
ti
c
ul
a
r
ly
e
vi
de
nt
w
he
n
ha
ndl
in
g
la
r
ge
da
ta
s
e
ts
[
19]
.
O
ur
r
e
s
ul
ts
a
r
e
c
ons
i
s
te
nt
w
it
h
pr
io
r
r
e
s
e
a
r
c
h,
a
f
f
ir
m
in
g
th
e
w
id
e
s
pr
e
a
d
a
ppl
ic
a
bi
li
ty
of
th
e
K
-
m
e
a
ns
a
lg
or
it
hm
a
c
r
os
s
di
ve
r
s
e
in
dus
tr
ie
s
s
uc
h
as
e
-
c
om
m
e
r
c
e
,
r
e
s
t
a
ur
a
nt
s
,
a
nd
m
a
r
ke
ti
ng
[
17]
,
[
27]
,
[
29
]
.
T
he
r
obus
t
pe
r
f
or
m
a
nc
e
m
e
tr
ic
s
of
K
-
m
e
a
ns
,
in
c
lu
di
ng
hi
gh
SC
a
nd
CH
in
di
c
e
s
c
oupl
e
d
w
it
h
lo
w
e
r
DB
in
de
x,
it
e
r
a
ti
on
c
ount
,
a
nd
e
xe
c
ut
io
n
ti
m
e
,
f
ur
th
e
r
va
li
da
te
its
e
f
f
e
c
ti
ve
ne
s
s
in
c
u
s
to
m
e
r
s
e
gm
e
nt
a
ti
on
(
S
C
:0
.4364;
CH
in
de
x
:1
84977.0535;
DB
in
de
x
:0
.7913;
it
e
r
a
ti
ons
:3
ti
m
e
s
;
e
xe
c
ut
io
n
time
:0
.6076
s
e
c
ond
s
)
.
T
hr
e
e
s
e
gm
e
nt
s
w
e
r
e
id
e
nt
if
ie
d
by
th
e
c
us
to
m
e
r
s
e
gm
e
nt
a
ti
on,
as
f
ol
lo
w
s
.
I
m
por
ta
nt
va
lu
e
gr
oup:
c
us
to
m
e
r
s
w
it
h
hi
gh
R
F
M
va
lu
e
s
.
G
e
ne
r
a
l
de
ve
lo
pm
e
nt
gr
oup:
c
u
s
to
m
e
r
s
w
it
h
hi
gh
r
e
c
e
nc
y
v
a
lu
e
s
but
lo
w
f
r
e
que
nc
y
a
nd
m
one
ta
r
y
va
lu
e
s
.
L
os
t
gr
oup:
c
us
to
m
e
r
s
w
it
h
lo
w
R
F
M
va
lu
e
s
.
T
he
g
e
ne
r
a
l
de
v
e
lo
pm
e
nt
gr
oup
w
a
s
th
e
m
o
s
t
pr
e
va
le
nt
c
us
to
m
e
r
ty
pe
in
th
is
c
ur
r
e
nt
s
tu
dy,
c
om
pr
is
in
g
50
%
of
a
ll
c
us
to
m
e
r
s
.
A
ddi
ti
ona
ll
y,
we
c
onduc
t
e
d
an
a
na
ly
s
i
s
to
un
c
ove
r
a
s
s
oc
ia
t
io
ns
be
twe
e
n
pr
e
vi
ous
ly
pur
c
ha
s
e
d
pr
oduc
ts
w
it
hi
n
each
c
lu
s
te
r
.
In
pr
e
vi
ous
r
e
s
e
a
r
c
h,
a
s
s
oc
ia
ti
on
r
ul
e
m
in
in
g
w
a
s
e
m
pl
oye
d
to
a
na
ly
z
e
c
us
to
m
e
r
tr
a
ns
a
c
ti
ons
f
ol
lo
w
in
g
c
u
s
to
m
e
r
s
e
gm
e
nt
a
ti
on.
T
hi
s
a
ppr
oa
c
h
a
id
s
in
unc
ov
e
r
in
g
hi
dde
n
c
u
s
to
m
e
r
pur
c
ha
s
e
pa
tt
e
r
ns
a
nd
g
e
ne
r
a
ti
ng
a
s
s
oc
ia
ti
on
r
ul
e
s
,
w
hi
c
h
r
e
ve
a
l
c
onn
e
c
ti
ons
be
twe
e
n
pr
e
vi
ous
ly
pur
c
ha
s
e
d
pr
oduc
ts
[
14]
.
O
ur
r
e
s
ul
ts
de
m
ons
tr
a
te
d
a
s
s
oc
ia
ti
ons
be
twe
e
n
pr
e
vi
ous
ly
pur
c
ha
s
e
d
pr
oduc
ts
f
or
c
u
s
to
m
e
r
s
in
e
a
c
h
c
lu
s
te
r
(
im
por
ta
nt
va
lu
e
gr
oup:
38
r
ul
e
s
;
ge
ne
r
a
l
de
ve
lo
pm
e
nt
gr
oup:
2
r
ul
e
s
;
lo
s
t
gr
oup:
2
r
ul
e
s
)
.
T
hi
s
out
c
om
e
va
li
da
te
s
th
e
u
s
e
of
a
s
s
oc
ia
ti
on
r
ul
e
m
in
in
g
in
th
e
a
na
l
ys
is
of
c
us
to
m
e
r
tr
a
ns
a
c
ti
ons
.
H
ow
e
ve
r
,
it
'
s
im
por
ta
nt
to
a
c
knowle
dge
th
e
li
m
it
a
ti
ons
of
our
s
tu
dy.
F
ut
ur
e
r
e
s
e
a
r
c
h
c
oul
d
e
xpl
or
e
a
lt
e
r
na
ti
ve
c
lu
s
te
r
in
g
te
c
hni
que
s
or
in
c
or
por
a
te
a
ddi
ti
ona
l
va
r
ia
bl
e
s
to
e
nha
nc
e
th
e
a
c
c
ur
a
c
y
a
nd
de
pt
h
of
Evaluation Warning : The document was created with Spire.PDF for Python.
I
nt
J
A
r
ti
f
I
nt
e
ll
I
S
S
N
:
2252
-
8938
C
us
to
m
e
r
s
e
gm
e
nt
at
io
n u
s
in
g as
s
o
c
ia
ti
on r
ul
e
m
in
in
g on r
e
ta
il
tr
ans
ac
ti
on data
(
Si
r
iw
an
K
aj
o
r
nk
as
i
r
at
)
1927
c
us
to
m
e
r
pr
of
il
in
g.
M
ovi
ng
f
or
w
a
r
d,
f
ut
ur
e
s
tu
di
e
s
c
oul
d
de
lv
e
in
to
th
e
in
te
gr
a
ti
on
of
a
dva
nc
e
d
da
ta
m
in
in
g
te
c
hni
que
s
or
m
a
c
hi
ne
le
a
r
ni
ng
a
lg
or
it
hm
s
to
r
e
f
in
e
c
us
to
m
e
r
s
e
gm
e
nt
a
ti
on
m
e
th
odol
ogi
e
s
f
ur
th
e
r
.
A
ddi
ti
ona
ll
y,
lo
ngi
tu
di
na
l
a
na
ly
s
e
s
a
s
s
e
s
s
in
g
th
e
im
pa
c
t
s
of
s
e
gm
e
nt
a
ti
on
s
tr
a
te
gi
e
s
on
c
u
s
to
m
e
r
r
e
te
nt
io
n
a
nd
pr
of
it
a
bi
li
ty
c
oul
d
yi
e
ld
va
lu
a
bl
e
s
tr
a
te
gi
c
in
s
i
ght
s
f
or
bus
in
e
s
s
e
s
.
In
s
um
m
a
r
y,
our
s
tu
dy
hi
ghl
ig
ht
s
th
e
e
f
f
ic
a
c
y
of
th
e
K
-
m
e
a
ns
a
lg
or
it
hm
f
or
c
us
to
m
e
r
s
e
gm
e
nt
a
ti
on,
pa
r
ti
c
ul
a
r
ly
in
a
na
ly
z
in
g
e
xt
e
ns
iv
e
da
t
a
s
e
t
s
.
T
h
e
a
s
s
oc
ia
ti
on
s
r
e
ve
a
l
e
d
be
twe
e
n
pr
e
vi
o
us
ly
pur
c
ha
s
e
d
pr
oduc
ts
w
it
hi
n
each
c
lu
s
te
r
unde
r
s
c
or
e
th
e
ut
il
it
y
of
a
s
s
oc
ia
ti
on
r
ul
e
m
in
in
g
in
de
c
ip
he
r
in
g
c
us
to
m
e
r
be
h
a
vi
or
s
.
T
he
s
e
s
e
gm
e
nt
a
ti
on
f
in
di
ngs
ha
ve
s
ig
ni
f
ic
a
nt
im
pl
ic
a
ti
on
s
f
or
bus
in
e
s
s
s
tr
a
te
gi
e
s
,
r
a
ngi
ng
f
r
om
de
m
a
nd
f
or
e
c
a
s
ti
ng
to
ta
r
ge
te
d
m
a
r
ke
ti
ng
e
f
f
or
ts
a
nd
in
ve
nt
or
y
m
a
na
ge
m
e
nt
[
30]
.
4.
L
I
M
I
T
A
T
I
O
N
S
,
F
U
T
U
R
E
R
E
S
E
A
R
C
H
,
AND
I
M
P
L
I
C
A
T
I
O
N
S
S
in
c
e
we
ut
il
iz
e
a
c
tu
a
l
s
a
l
e
s
da
t
a
f
r
om
th
e
c
om
pa
ny
in
th
is
s
tu
dy,
c
e
r
ta
in
in
f
or
m
a
ti
on
c
a
nnot
be
di
s
c
lo
s
e
d.
M
e
m
be
r
num
be
r
s
,
f
or
in
s
t
a
nc
e
,
a
r
e
c
on
s
id
e
r
e
d
p
e
r
s
ona
l
da
ta
of
c
us
to
m
e
r
s
as
th
e
y
a
r
e
phon
e
num
be
r
s
.
In
th
e
c
lu
s
te
r
in
g
pr
oc
e
s
s
,
we
ne
e
d
to
de
te
r
m
in
e
th
e
n
um
be
r
of
c
lu
s
te
r
s
or
th
e
k
-
va
lu
e
e
a
c
h
time
we
pe
r
f
or
m
c
lu
s
te
r
in
g.
F
ur
th
e
r
m
o
r
e
,
in
th
e
s
e
c
ti
on
c
onc
e
r
ni
ng
th
e
a
na
ly
s
is
of
th
e
r
e
la
ti
ons
hi
p
r
ul
e
s
,
we
di
d
not
s
e
le
c
t
th
e
time
f
r
a
m
e
f
or
th
e
da
ta
.
T
hi
s
r
e
s
ul
ts
in
a
la
c
k
of
knowle
dge
r
e
ga
r
di
ng
w
he
n
th
e
pr
oduc
ts
di
s
pl
a
ye
d
on
th
os
e
r
ul
e
s
w
e
r
e
pur
c
ha
s
e
d
by
c
u
s
to
m
e
r
s
.
S
uc
h
la
c
k
of
in
f
or
m
a
ti
on
c
oul
d
pot
e
nt
ia
ll
y
le
a
d
to
is
s
ue
s
in
s
a
le
s
pl
a
nni
ng,
pr
om
ot
io
ns
,
a
nd
in
ve
nt
or
y
m
a
na
ge
m
e
nt
.
In
f
ut
ur
e
r
e
s
e
a
r
c
h,
we
a
im
to
de
ve
lo
p
a
r
e
c
om
m
e
nda
ti
on
s
ys
te
m
in
th
e
f
or
m
of
a
w
e
b
a
ppl
ic
a
ti
on
th
a
t
pe
r
f
or
m
s
a
ut
om
a
ti
c
c
us
to
m
e
r
s
e
gm
e
nt
a
ti
on.
U
s
e
r
s
w
on'
t
ne
e
d
to
in
put
th
e
k
-
va
lu
e
th
e
m
s
e
lv
e
s
;
th
e
s
y
s
te
m
w
il
l
de
te
r
m
in
e
it
a
ut
om
a
ti
c
a
ll
y,
in
c
lu
di
ng
na
m
in
g
th
e
c
us
to
m
e
r
gr
oups
.
T
hi
s
s
ys
te
m
w
il
l
a
s
s
is
t
in
r
e
c
om
m
e
ndi
ng
pr
oduc
ts
a
nd
pr
om
ot
io
ns
to
r
e
ta
il
s
to
r
e
s
.
In
a
ddi
ti
on
to
s
e
le
c
ti
ng
c
us
to
m
e
r
g
r
oups
f
or
da
ta
vi
e
w
in
g,
us
e
r
s
w
il
l
a
ls
o
be
a
bl
e
to
c
hoos
e
s
pe
c
if
ic
time
in
te
r
va
ls
f
or
di
s
pl
a
yi
ng
in
f
o
r
m
a
ti
on.
T
hi
s
w
il
l
e
na
bl
e
or
ga
ni
z
a
ti
ons
to
id
e
nt
if
y
w
hi
c
h
pr
oduc
ts
c
e
r
ta
in
c
us
to
m
e
r
s
e
g
m
e
nt
s
te
nd
to
pur
c
ha
s
e
to
ge
th
e
r
dur
in
g
s
pe
c
if
ic
time
pe
r
io
ds
,
f
a
c
il
it
a
ti
ng
in
ve
nt
or
y
m
a
n
a
ge
m
e
nt
a
nd
pr
om
ot
io
n
pl
a
nni
ng.
T
hi
s
r
e
s
e
a
r
c
h
m
a
y
c
ont
r
ib
ut
e
to
e
nha
nc
in
g
bus
in
e
s
s
e
f
f
ic
ie
nc
y
a
nd
e
f
f
e
c
ti
ve
ly
m
e
e
ti
ng
th
e
n
e
e
ds
a
nd
im
por
ta
nc
e
of
c
us
to
m
e
r
s
.
M
or
e
ove
r
,
r
e
s
e
a
r
c
h
in
th
is
a
r
e
a
can
s
e
r
ve
as
a
gui
de
li
ne
f
or
de
ve
lo
pi
ng
m
or
e
e
f
f
ic
ie
nt
bus
in
e
s
s
pr
a
c
ti
c
e
s
in
th
e
f
ut
ur
e
by
in
tr
oduc
in
g
s
tr
a
te
gi
e
s
a
nd
m
a
r
ke
ti
ng
a
ppr
oa
c
he
s
ba
s
e
d
on
s
ol
id
a
nd
e
f
f
e
c
ti
ve
da
ta
.
5.
C
O
N
C
L
U
S
I
O
N
T
hi
s
s
tu
dy
pe
r
f
or
m
e
d
a
c
om
pa
r
is
on
of
c
lu
s
te
r
in
g
a
lg
or
it
hm
s
,
s
pe
c
if
ic
a
ll
y
K
-
m
e
a
ns
,
f
uz
z
y
C
-
m
e
a
n
s
,
a
nd
S
O
N
N
w
h
e
n
a
ppl
ie
d
a
lo
ngs
id
e
th
e
R
F
M
m
e
th
od,
to
d
e
te
r
m
in
e
th
e
m
os
t
s
ui
ta
bl
e
a
ppr
oa
c
h
f
or
c
us
to
m
e
r
s
e
gm
e
nt
a
ti
on.
F
iv
e
e
va
lu
a
ti
on
c
r
it
e
r
ia
w
e
r
e
ut
il
iz
e
d:
S
C
,
CH
in
de
x,
DB
in
de
x,
it
e
r
a
ti
on
c
ount
,
a
nd
e
xe
c
ut
io
n
ti
m
e
.
A
ddi
ti
ona
ll
y,
we
e
m
pl
oye
d
th
e
FP
-
G
r
ow
th
a
lg
o
r
it
h
m
f
or
a
s
s
oc
ia
ti
on
r
ul
e
m
in
in
g
to
unc
ove
r
r
e
la
ti
ons
hi
ps
a
m
ong
pr
e
vi
ous
ly
pur
c
ha
s
e
d
pr
oduc
ts
w
it
hi
n
e
a
c
h
c
lu
s
te
r
.
T
he
f
in
di
ngs
f
r
om
bot
h
c
us
to
m
e
r
s
e
gm
e
nt
a
ti
on
a
nd
a
s
s
oc
ia
ti
on
r
ul
e
m
in
in
g
of
f
e
r
va
lu
a
bl
e
in
s
ig
ht
s
th
a
t
c
a
n
gr
e
a
tl
y
in
f
or
m
pr
oduc
t
s
to
c
k
pl
a
nni
ng
a
nd
pr
om
ot
io
na
l
s
tr
a
te
gi
e
s
w
it
hi
n
th
e
or
ga
ni
z
a
ti
on.
M
or
e
ove
r
,
th
e
s
e
in
s
ig
ht
s
ha
ve
th
e
pot
e
nt
ia
l
to
a
dva
nc
e
th
e
de
ve
lo
pm
e
nt
of
m
or
e
e
f
f
e
c
ti
ve
pr
oduc
t
r
e
c
o
m
m
e
nda
ti
on
s
ys
te
m
s
.
By
pr
e
s
e
nt
in
g
a
c
le
a
r
unde
r
s
ta
ndi
ng
of
th
e
e
f
f
ic
ie
nc
y
of
each
c
lu
s
te
r
in
g
a
lg
o
r
it
hm
a
nd
its
im
pl
ic
a
ti
ons
f
o
r
pr
a
c
ti
c
a
l
a
ppl
ic
a
ti
ons
,
th
is
s
tu
dy
la
y
s
a
s
ol
id
f
ounda
ti
on
f
or
in
f
or
m
e
d
de
c
i
s
io
n
m
a
ki
ng
in
m
a
r
ke
ti
ng
a
nd
c
us
to
m
e
r
r
e
la
ti
ons
hi
p
m
a
na
ge
m
e
nt
.
A
C
K
N
O
WL
E
D
G
E
M
E
N
T
S
T
he
a
ut
hor
s
th
a
nk
S
e
ppo
J.
K
a
r
r
il
a
f
or
c
on
s
tr
uc
ti
ve
c
om
m
e
nt
s
on
th
is
m
a
nu
s
c
r
ip
t.
T
hi
s
r
e
s
e
a
r
c
h
w
a
s
f
unde
d
by
P
r
in
c
e
of
S
ongk
la
U
ni
ve
r
s
it
y,
S
u
r
a
t
T
ha
ni
C
a
m
pus
.
W
e
a
ls
o
th
a
nk
T
ue
nj
a
i
P
a
ni
t
G
r
oup
C
om
pa
ny
L
im
it
e
d
(
C
o.,
L
td
.
)
,
T
ha
il
a
nd
f
or
th
e
da
ta
in
th
is
s
tu
dy.
F
U
N
D
I
N
G
I
N
F
O
R
M
A
T
I
O
N
T
hi
s
r
e
s
e
a
r
c
h w
a
s
s
uppor
te
d by P
r
in
c
e
of
S
ongkla
U
ni
ve
r
s
it
y, S
ur
a
t
T
ha
ni
C
a
m
pus
.
A
U
T
H
O
R
C
O
N
T
R
I
B
U
T
I
O
N
S
S
T
A
T
E
M
E
N
T
T
hi
s
jo
ur
na
l
us
e
s
th
e
C
ont
r
ib
ut
or
R
ol
e
s
T
a
xonomy
(
C
R
e
d
iT
)
to
r
e
c
ogni
z
e
in
di
vi
dua
l
a
ut
hor
c
ont
r
ib
ut
io
ns
, r
e
duc
e
a
ut
hor
s
hi
p
di
s
put
e
s
,
a
nd f
a
c
il
it
a
te
c
ol
la
bo
r
a
ti
on.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
2252
-
8938
I
nt
J
A
r
ti
f
I
nt
e
ll
, V
ol
.
14
, N
o.
3
,
J
une
20
25
:
1919
-
1929
1928
N
am
e
o
f
A
u
t
h
or
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
S
ir
iwa
n K
a
jo
r
nka
s
ir
a
t
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
P
a
tt
a
r
a
w
a
n G
ungl
in
✓
✓
✓
✓
✓
✓
✓
✓
K
r
it
s
a
da
P
ua
ngs
uw
a
n
✓
✓
✓
✓
✓
✓
N
a
w
a
pon Ka
e
w
s
uw
a
n
✓
✓
✓
✓
✓
✓
C
:
C
onc
e
pt
ua
l
i
z
a
t
i
on
M
:
M
e
t
hodol
ogy
So
:
So
f
t
w
a
r
e
Va
:
Va
l
i
da
t
i
on
Fo
:
Fo
r
m
a
l
a
na
l
ys
i
s
I
:
I
nve
s
t
i
ga
t
i
on
R
:
R
e
s
our
c
e
s
D
:
D
a
t
a
C
ur
a
t
i
on
O
:
W
r
i
t
i
ng
-
O
r
i
gi
na
l
D
r
a
f
t
E
:
W
r
i
t
i
ng
-
R
e
vi
e
w
&
E
di
t
i
ng
Vi
:
Vi
s
ua
l
i
z
a
t
i
on
Su
:
Su
pe
r
vi
s
i
on
P
:
P
r
oj
e
c
t
a
dm
i
ni
s
t
r
a
t
i
on
Fu
:
Fu
ndi
ng a
c
qui
s
i
t
i
on
C
O
N
F
L
I
C
T
O
F
I
N
T
E
R
E
S
T
S
T
A
T
E
M
E
N
T
A
ut
hor
s
s
ta
te
no c
onf
li
c
t
of
i
nt
e
r
e
s
t.
D
A
T
A
A
V
A
I
L
A
B
I
L
I
T
Y
T
he
a
ut
hor
s
c
onf
ir
m
th
a
t
th
e
da
t
a
s
uppor
ti
n
g t
h
e
f
in
di
n
gs
of
t
hi
s
s
tu
dy
a
r
e
a
v
a
il
a
bl
e
w
it
hi
n t
he
a
r
ti
c
l
e
.
R
E
F
E
R
E
N
C
E
S
[
1]
S
.
K
e
m
p,
“
D
i
gi
t
a
l
2024:
gl
oba
l
ove
r
vi
e
w
r
e
por
t
,”
D
at
a
R
e
por
t
al
,
2024.
A
c
c
e
s
s
e
d:
N
ov.
28,
2023.
[
O
nl
i
ne
]
.
A
va
i
l
a
bl
e
:
ht
t
ps
:
/
/
da
t
a
r
e
por
t
a
l
.c
om
/
r
e
por
t
s
/
di
gi
t
a
l
-
2024
-
g
l
oba
l
-
ove
r
vi
e
w
-
r
e
po
r
t
[
2]
X
.
D
ong,
H
.
T
u,
H
.
Z
hu,
T
.
L
i
u,
X
.
Z
ha
o,
a
nd
K
.
X
i
e
,
“
D
oe
s
di
ve
r
s
i
t
y
f
a
c
i
l
i
t
a
t
e
c
ons
um
e
r
d
e
c
i
s
i
ons
:
a
c
om
pa
r
a
t
i
ve
pe
r
s
pe
c
t
i
v
e
ba
s
e
d
on
s
i
ngl
e
-
c
a
t
e
gor
y
ve
r
s
u
s
m
ul
t
i
-
c
a
t
e
gor
y
pr
oduc
t
s
,”
A
s
i
a
P
ac
i
f
i
c
J
ou
r
nal
of
M
ar
k
e
t
i
ng
and
L
ogi
s
t
i
c
s
,
vol
.
36,
no.
4,
pp. 936
–
956, 2024, doi
:
10.1108/
A
P
J
M
L
-
05
-
2023
-
0395.
[
3]
C
.
P
.
G
up
t
a
a
nd
V
.
V
.
R
.
K
um
a
r
,
“
R
e
c
om
m
e
n
da
t
i
o
n
s
ys
t
e
m
:
a
t
r
a
ns
f
o
r
m
a
t
i
ve
a
r
t
i
f
i
c
i
a
l
i
n
t
e
l
l
i
ge
n
c
e
t
oo
l
f
or
e
-
c
o
m
m
e
r
c
e
,”
I
n
t
e
r
n
at
i
on
al
C
o
nf
e
r
e
nc
e
on
I
nf
or
m
at
i
c
s
a
nd
C
om
pu
t
a
t
i
ona
l
Sc
i
e
nc
e
s
,
pp.
6
0
–
6
5,
202
4,
do
i
:
10
.1
109
/
I
C
I
C
o
S
62
60
0.2
02
4.1
06
368
25
.
[
4]
V
. V
e
nka
t
e
s
h
, H
. H
oe
h
l
e
,
J
.
A
.
A
l
oys
i
us
,
a
nd
H
.
R
.
N
i
k
kha
h,
“
B
e
i
ng
a
t
t
h
e
c
ut
t
i
n
g e
d
ge
of
on
l
i
ne
s
hop
pi
ng
:
R
o
l
e
of
r
e
c
o
m
m
e
nda
t
i
ons
a
nd
di
s
c
ou
nt
s
on
p
r
i
va
c
y p
e
r
c
e
p
t
i
ons
,”
C
o
m
pu
t
e
r
s
i
n
H
u
m
a
n
B
e
h
av
i
or
,
vo
l
.
1
21,
202
1,
do
i
:
10
.1
016
/
j
.c
h
b.
202
1.
106
78
5.
[
5]
M
.
A
.
G
om
e
s
a
nd T
.
M
e
i
s
e
n,
“
A
r
e
v
i
e
w
o
n
c
us
t
om
e
r
s
e
gm
e
nt
a
t
i
on
m
e
t
h
ods
f
or
pe
r
s
ona
l
i
z
e
d
c
us
t
om
e
r
t
a
r
ge
t
i
ng
i
n
e
-
c
om
m
e
r
c
e
us
e
c
a
s
e
s
,”
I
nf
or
m
at
i
on
Sy
s
t
e
m
s
and e
-
B
us
i
ne
s
s
M
anage
m
e
n
t
, v
ol
. 2
1, no
. 3,
pp. 5
27
–
570, 2
023,
doi
:
10.1
007/
s
10
257
-
023
-
006
40
-
4.
[
6]
H
.
H
a
r
t
o
yo,
E
.
M
a
na
l
u,
U
.
S
um
a
r
w
a
n,
a
n
d
P
.
N
u
r
ha
ya
t
i
,
“
D
r
i
vi
ng
s
uc
c
e
s
s
:
a
s
e
gm
e
nt
a
t
i
on
o
f
c
us
t
o
m
e
r
a
dm
i
r
a
t
i
o
n
i
n
a
ut
om
ot
i
ve
i
nd
us
t
r
y
,”
J
our
na
l
o
f
O
pe
n
I
n
nov
at
i
on
:
T
e
c
hn
ol
ogy
,
M
ar
k
e
t
,
a
nd C
om
pl
e
x
i
t
y
,
vo
l
.
9,
no
.
2,
20
23,
d
oi
:
1
0.
101
6/
j
.
j
o
i
t
m
c
.
202
3.
100
03
1.
[
7]
E
.
P
.
V
a
r
ga
s
,
C
.
de
-
J
ua
n
-
R
i
pol
l
,
M
.
B
.
P
a
na
de
r
o,
a
nd
M
.
A
l
c
a
ñi
z
,
“
L
i
f
e
s
t
yl
e
s
e
gm
e
nt
a
t
i
on
of
t
our
i
s
t
s
:
t
he
r
ol
e
of
pe
r
s
ona
l
i
t
y,”
H
e
l
i
y
on
, vol
. 7, no. 7, 2021, doi
:
10.1016/
j
.he
l
i
yon.2021.e
07579.
[
8]
F
.
B
a
r
j
a
k,
J
.
L
i
nde
que
,
J
.
K
oc
h,
a
nd
M
.
S
ol
a
nd,
“
S
e
gm
e
nt
i
ng
hous
e
hol
d
e
l
e
c
t
r
i
c
i
t
y
c
us
t
om
e
r
s
w
i
t
h
qua
nt
i
t
a
t
i
ve
a
nd
qua
l
i
t
a
t
i
v
e
a
ppr
oa
c
he
s
,”
R
e
ne
w
abl
e
and Sus
t
ai
nabl
e
E
ne
r
gy
R
e
v
i
e
w
s
, vol
. 157, 2022, doi
:
10.1016/
j
.r
s
e
r
.2021.112014.
[
9]
V
.
D
ua
r
t
e
,
S
.
Z
uni
ga
-
J
a
r
a
,
a
nd
S
.
C
ont
r
e
r
a
s
,
“
M
a
c
hi
ne
l
e
a
r
ni
ng
a
nd
m
a
r
ke
t
i
ng:
a
s
ys
t
e
m
a
t
i
c
l
i
t
e
r
a
t
ur
e
r
e
vi
e
w
,”
I
E
E
E
A
c
c
e
s
s
,
vol
. 10, pp. 93273
–
93288, 2022, doi
:
10.1109/
A
C
C
E
S
S
.2022.3202896.
[
10]
D
.
H
.
K
hoi
r
i
ya
h
a
nd
R
.
A
m
ba
r
w
a
t
i
,
“
D
yna
m
i
c
s
e
gm
e
nt
a
t
i
on
a
na
l
y
s
i
s
f
or
e
xpe
di
t
i
on
s
e
r
vi
c
e
s
:
i
nt
e
gr
a
t
i
ng
k
-
m
e
a
ns
a
nd
de
c
i
s
i
o
n
t
r
e
e
,”
J
our
nal
of
I
nf
or
m
at
i
on Sy
s
t
e
m
s
and I
nf
or
m
at
i
c
s
, vol
. 6, no. 1, pp. 363
–
37
7, 2024, doi
:
10.51519/
j
our
na
l
i
s
i
.v6i
1.666.
[
11]
M
.
A
.
U
ddi
n
e
t
al
.
,
“
D
a
t
a
-
dr
i
ve
n
s
t
r
a
t
e
gi
e
s
f
or
di
gi
t
a
l
na
t
i
ve
m
a
r
ke
t
s
e
gm
e
nt
a
t
i
on
us
i
ng
c
l
us
t
e
r
i
ng,”
I
nt
e
r
nat
i
onal
J
our
nal
of
C
ogni
t
i
v
e
C
om
put
i
ng i
n E
ngi
ne
e
r
i
ng
, vol
. 5, pp. 178
–
191, 2024, doi
:
10.1016/
j
.i
j
c
c
e
.2024.04.002.
[
12]
M
.
E
.
J
a
l
a
l
a
nd
A
.
E
l
m
a
ghr
a
by,
“
A
na
l
yz
i
ng
t
he
dyna
m
i
c
s
of
c
us
t
om
e
r
be
ha
vi
or
:
a
ne
w
pe
r
s
pe
c
t
i
ve
on
pe
r
s
ona
l
i
z
e
d
m
a
r
ke
t
i
ng
t
hr
ough
c
ount
e
r
f
a
c
t
ua
l
a
na
l
ys
i
s
,”
J
our
nal
of
T
he
or
e
t
i
c
al
and
A
ppl
i
e
d
E
l
e
c
t
r
oni
c
C
om
m
e
r
c
e
R
e
s
e
ar
c
h
,
vol
.
19,
no.
3
,
pp. 1660
–
1681, 2024, doi
:
10.3390/
j
t
a
e
r
19030081.
[
13]
W
.
W
a
hyuni
ngs
i
h
a
nd
P
.
T
.
P
r
a
s
e
t
ya
ni
ngr
um
,
“
E
nha
nc
i
ng
s
a
l
e
s
de
t
e
r
m
i
na
t
i
on
f
or
c
of
f
e
e
s
hop
pa
c
ka
ge
s
t
hr
ough
a
s
s
oc
i
a
t
e
d
d
a
t
a
m
i
ni
ng:
l
e
ve
r
a
gi
ng
t
he
F
P
-
gr
ow
t
h
a
l
gor
i
t
hm
,”
J
our
nal
of
I
nf
or
m
at
i
on
Sy
s
t
e
m
s
and
I
nf
or
m
at
i
c
s
,
vol
.
5,
no.
2,
pp.
758
–
770,
2023,
doi
:
10.51519/
j
our
na
l
i
s
i
.v5i
2.500.
[
14]
S
.
H
.
L
i
a
o,
R
.
W
i
dow
a
t
i
,
a
nd
K
.
C
.
Y
a
ng,
“
I
nve
s
t
i
ga
t
i
ng
s
por
t
s
be
ha
vi
or
s
a
nd
m
a
r
ke
t
i
n
T
a
i
w
a
n
f
or
s
por
t
s
l
e
i
s
ur
e
a
nd
e
nt
e
r
t
a
i
nm
e
nt
m
a
r
ke
t
i
ng onl
i
ne
r
e
c
om
m
e
nda
t
i
ons
,”
E
nt
e
r
t
ai
nm
e
nt
C
om
put
i
ng
,
vol
. 39, 2021, doi
:
10.1016/
j
.e
nt
c
om
.2021.100442.
[
15]
M
.
R
o
s
t
a
m
i
,
U
.
M
uha
m
m
a
d,
S
.
F
or
ouz
a
nde
h,
K
.
B
e
r
a
hm
a
nd,
V
.
F
a
r
r
a
hi
,
a
nd
M
.
O
u
s
s
a
l
a
h,
“
A
n
e
f
f
e
c
t
i
ve
e
xpl
a
i
na
bl
e
f
oo
d
r
e
c
om
m
e
nda
t
i
on
us
i
ng
de
e
p
i
m
a
ge
c
l
us
t
e
r
i
ng
a
nd
c
om
m
uni
t
y
de
t
e
c
t
i
on,”
I
nt
e
l
l
i
ge
nt
Sy
s
t
e
m
s
w
i
t
h
A
ppl
i
c
at
i
ons
,
vol
.
16,
2022
,
doi
:
10.1016/
j
.i
s
w
a
.2022.200157.
[
16]
R
.
D
ou,
W
.
L
i
,
G
.
N
a
n,
X
.
W
a
ng,
a
nd
Y
.
Z
hou,
“
H
ow
c
a
n
m
a
nuf
a
c
t
ur
e
r
s
m
a
ke
de
c
i
s
i
ons
on
pr
oduc
t
a
ppe
a
r
a
nc
e
de
s
i
gn?
A
r
e
s
e
a
r
c
h on opt
i
m
a
l
de
s
i
gn
ba
s
e
d on
c
us
t
om
e
r
s
’
e
m
ot
i
ona
l
s
a
t
i
s
f
a
c
t
i
on,”
J
ou
r
n
al
of
M
anage
m
e
nt
Sc
i
e
n
c
e
and
E
ngi
ne
e
r
i
ng
, vol
.
6,
no. 2, pp. 177
–
196, 2021, doi
:
10.1016/
j
.j
m
s
e
.2021.02.010.
[
17]
A
.
G
r
i
va
,
“
‘
I
c
a
n
ge
t
no
e
-
s
a
t
i
s
f
a
c
t
i
on’
.
W
ha
t
a
na
l
yt
i
c
s
s
a
y?
E
vi
de
nc
e
us
i
n
g
s
a
t
i
s
f
a
c
t
i
on
d
a
t
a
f
r
om
e
-
c
om
m
e
r
c
e
,”
J
ou
r
nal
of
R
e
t
ai
l
i
ng and C
ons
um
e
r
Se
r
v
i
c
e
s
, vol
. 66, 2022, doi
:
10.1016/
j
.j
r
e
t
c
ons
e
r
.2022.
102954.
[
18]
D
.
V
a
r
m
a
,
A
.
N
e
ha
ns
h,
a
nd
P
.
S
w
a
t
hy,
“
D
a
t
a
pr
e
pr
oc
e
s
s
i
ng
t
ool
ki
t
:
a
n
a
ppr
oa
c
h
t
o
a
ut
om
a
t
e
da
t
a
pr
e
pr
oc
e
s
s
i
ng,”
I
nt
e
r
n
at
i
onal
J
our
nal
of
Sc
i
e
nt
i
f
i
c
R
e
s
e
ar
c
h i
n E
ngi
ne
e
r
i
ng and M
anage
m
e
nt
, vol
. 7, no. 3, 2
023, doi
:
10.55041/
i
j
s
r
e
m
18270.
[
19]
P
.
A
ni
t
ha
a
nd
M
.
M
.
P
a
t
i
l
,
“
R
F
M
m
ode
l
f
or
c
us
t
om
e
r
pu
r
c
ha
s
e
be
ha
vi
or
us
i
ng
K
-
M
e
a
ns
a
l
gor
i
t
hm
,”
J
our
nal
of
K
i
ng
Saud
U
ni
v
e
r
s
i
t
y
-
C
om
put
e
r
and I
nf
or
m
at
i
on Sc
i
e
nc
e
s
, vol
. 34, no. 5, pp. 1785
–
1792
, 2022, doi
:
10.1016/
j
.j
ks
uc
i
.2019.12.011.
[
20]
C
.
W
a
ng,
“
E
f
f
i
c
i
e
nt
c
us
t
om
e
r
s
e
gm
e
nt
a
t
i
on
i
n
di
gi
t
a
l
m
a
r
ke
t
i
ng
us
i
ng
de
e
p
l
e
a
r
ni
ng
w
i
t
h
s
w
a
r
m
i
nt
e
l
l
i
ge
nc
e
a
ppr
oa
c
h,
”
I
nf
or
m
at
i
on P
r
oc
e
s
s
i
ng and M
anage
m
e
nt
, vol
. 59, no. 6, 2022, doi
:
10.1016/
j
.i
pm
.2022.103085.
[
21]
A
.
J
.
C
hr
i
s
t
y,
A
.
U
m
a
m
a
ke
s
w
a
r
i
,
L
.
P
r
i
ya
t
ha
r
s
i
ni
,
a
nd
A
.
N
e
ya
a
,
“
R
F
M
r
a
nki
ng
–
a
n
e
f
f
e
c
t
i
ve
a
ppr
oa
c
h
t
o
c
us
t
om
e
r
s
e
gm
e
nt
a
t
i
on,”
J
ou
r
nal
of
K
i
ng
Saud
U
ni
v
e
r
s
i
t
y
-
C
om
put
e
r
and
I
nf
or
m
at
i
o
n
Sc
i
e
nc
e
s
,
vol
.
33,
no.
10,
pp.
1251
–
1257,
2021,
doi
:
10.1016/
j
.j
ks
uc
i
.2018.09.004.
Evaluation Warning : The document was created with Spire.PDF for Python.