属性组序下基于代价敏感的约简方法
刘鑫,胡军,张清华
Attribute reduction based on cost sensitive under attribute group order
Xin Liu,Jun Hu,Qinghua Zhang
表5
Dermatology数据集在10种不同分组方式下的约简结果
Table 5
Reduction results of Dermatology dataset under ten different grouping methods
I
D
A
R
C
S
G
O
(
λ
=
0.5
)
A
R
C
S
G
O
(
λ
=
0
)
R
e
d
u
c
t
C
O
S
T
S
A
P
R
e
d
u
c
t
C
O
S
T
S
A
P
1
c
21
,
c
5
,
c
28
,
c
32
,
c
17
,
c
18
,
c
20
,
c
31
,
c
8
,
c
1
,
c
4
150
2.9
c
22
,
c
8
,
c
5
,
c
31
,
c
28
,
c
2
,
c
4
,
c
17
,
c
32
,
c
1
,
c
18
210
2.9
2
c
3
,
c
5
,
c
9
,
c
14
,
c
20
,
c
29
,
c
17
,
c
16
,
c
28
110
2.11
c
4
,
c
16
,
c
9
,
c
8
,
c
19
,
c
32
,
c
28
,
c
18
,
c
13
,
c
10
150
2.7
3
c
5
,
c
29
,
c
15
,
c
10
,
c
16
,
c
32
,
c
7
,
c
18
,
c
2
,
c
1
,
c
4
170
2.9
c
5
,
c
22
,
c
29
,
c
15
,
c
4
,
c
7
,
c
32
,
c
16
,
c
1
,
c
18
220
2.9
4
c
4
,
c
19
,
c
17
,
c
2
,
c
6
,
c
24
,
c
10
,
c
34
150
2.87
c
4
,
c
34
,
c
19
,
c
2
,
c
6
,
c
22
,
c
17
,
c
24
210
2.87
5
c
32
,
c
21
,
c
9
,
c
14
,
c
26
,
c
5
,
c
3
,
c
18
,
c
8
,
c
4
130
2.8
c
34
,
c
22
,
c
4
,
c
3
,
c
14
,
c
5
,
c
6
200
2.85
6
c
26
,
c
28
,
c
14
,
c
19
,
c
18
,
c
1
,
c
33
,
c
34
130
3
c
25
,
c
34
,
c
19
,
c
28
,
c
1
,
c
18
,
c
14
,
c
26
130
3
7
c
34
,
c
5
,
c
19
,
c
3
,
c
32
,
c
18
,
c
10
,
c
7
120
2.62
c
34
,
c
5
,
c
4
,
c
32
,
c
3
,
c
10
130
2.5
8
c
2
,
c
17
,
c
21
,
c
18
,
c
16
,
c
33
,
c
5
,
c
3
,
c
28
,
c
9
120
2.5
c
34
,
c
22
,
c
4
,
c
3
,
c
28
,
c
33
,
c
5
200
2.71
9
c
19
,
c
17
,
c
28
,
c
34
,
c
18
,
c
33
,
c
4
140
2.57
c
16
,
c
19
,
c
34
,
c
4
,
c
18
,
c
10
,
c
29
160
2.71
10
c
5
,
c
21
,
c
17
,
c
25
,
c
19
,
c
9
,
c
3
,
c
18
,
c
7
,
c
16
120
2.7
c
34
,
c
28
,
c
3
,
c
16
,
c
19
,
c
18
,
c
11
,
c
7
140
2.75