Information Technology Reference
In-Depth Information
⎛
⎝
⎞
⎠
(
1
,
0
)(
0
.
7
,
0
.
2
)(
0
.
7
,
0
.
1
)(
0
.
5
,
0
.
3
)(
0
.
5
,
0
.
4
)
(
0
.
7
,
0
.
2
)(
1
,
0
)(
0
.
7
,
0
.
1
)(
0
.
6
,
0
.
1
)(
0
.
6
,
0
.
3
)
Z
=
(
0
.
7
,
0
.
1
)(
0
.
7
,
0
.
1
)(
1
,
0
)(
0
.
6
,
0
.
1
)(
0
.
5
,
0
.
5
)
(
0
.
5
,
0
.
3
)(
0
.
6
,
0
.
1
)(
0
.
6
,
0
.
1
)(
1
,
0
)(
0
.
6
,
0
.
3
)
(
0
.
5
,
0
.
4
)(
0
.
6
,
0
.
3
)(
0
.
5
,
0
.
5
)(
0
.
6
,
0
.
3
)(
1
,
0
)
Step 2
Choose the confidence levels properly, and get the corresponding clustering
results with the direct method:
(1) When 0
.
7
<μ
λ
1
≤
1
.
0, by Eq. (
2.210
), we know that there is no value
z
ij
y
i
]
(
1
)
in
R
such that
z
ij
=
λ
1
, i.e.,
[
=
φ
. Thus, each car is clustered into one type:
Z
{
y
1
}
,
{
y
2
}
,
{
y
3
}
,
{
y
4
}
and
{
y
5
}
.
(2) When 0
.
6
<μ
λ
2
≤
0
.
7, we have the following two cases:
(i)
z
13
=
z
23
=
(
0
.
7
,
0
.
1
)
: In this case, by Eq. (
2.209
), we know that
y
1
,
y
2
and
y
3
can be clustered into one type:
{
y
1
,
y
2
,
y
3
}
. Then, by Step 2 of the clustering
method, we get that the cars
y
i
(
i
=
1
,
2
,
3
,
4
,
5
)
are clustered into three types:
{
y
1
,
y
2
,
y
3
}
,
{
y
4
}
and
{
y
5
}
.
: In this case,
y
1
and
y
2
can be clustered into one type. Thus,
by Step 2 of the clustering method, we know that the cars
y
i
(ii)
z
12
=
(
0
.
7
,
0
.
2
)
(
i
=
1
,
2
,
3
,
4
,
5
)
are
also clustered into three types:
{
y
1
,
y
2
,
y
3
}
,
{
y
4
}
and
{
y
5
}
.
(3) When 0
.
5
<μ
λ
3
≤
0
.
6, we have the following two cases:
: In this case,
y
2
,
y
3
and
y
4
can be clustered into one
type. Then, merging the clustering results of (1) and (2), we can see that the cars
y
i
(
(i)
z
24
=
z
34
=
(
0
.
6
,
0
.
1
)
=
,
,
,
,
)
{
y
1
,
y
2
,
y
3
,
y
4
}
{
y
5
}
i
1
2
3
4
5
are clustered into two types:
and
.
=
(
.
,
.
)
: In this case,
y
2
and
y
5
can be clustered into one type.
Then, merging the clustering results above, it can be obtained that the cars
y
i
(
(ii)
z
25
0
6
0
3
i
=
1
,
2
,
3
,
4
,
5
)
are clustered into one type:
{
y
1
,
y
2
,
y
3
,
y
4
,
y
5
}
.
Compared with Zhang et al. (2007)'s method, we can know that the direct method
with less calculation amount can have better clustering results.
Search WWH ::
Custom Search