研究文章
基于多目标微分进化基因的两阶段方法的选择
|
|
|
结肠 |
DLBCL |
白血病 |
前列腺癌 |
Prostate2 |
TCell淋巴瘤 |
价值 |
|
| 所有功能 |
Acc |
58.08% |
76.58% |
71.33% |
62.81% |
76.57% |
67.82% |
011 - 0.000 |
|
基因 |
2000.0 |
5469.0 |
7129.0 |
10509 .0 |
2135.0 |
2922.0 |
|
|
| 该方法 |
Acc |
88.06% |
87.67% |
72.10% |
94.41% |
90.90% |
84.47% |
- - - - - - |
|
基因 |
6.2 |
12.7 |
12.4 |
10.0 |
7.0 |
10.1 |
|
|
| GainRatio |
Acc |
79.10% |
83.08% |
62.67% |
91.24% |
93.19% |
74.87% |
106 - 0.005 |
|
基因(前十) |
10.0 |
10.0 |
10.0 |
10.0 |
10.0 |
10.0 |
|
|
Acc |
78.85% |
87.08% |
71.14% |
92.24% |
92.19% |
77.95% |
282 - 0.054 |
|
基因(前20名) |
20.0 |
20.0 |
20.0 |
20.0 |
20.0 |
20.0 |
|
|
Acc |
82.31% |
87.08% |
64.10% |
91.24% |
92.14% |
73.21% |
658 - 0.015 |
|
基因(前40名) |
40.0 |
40.0 |
40.0 |
40.0 |
40.0 |
40.0 |
|
|
| ReliefF |
Acc |
82.44% |
88.42% |
69.62% |
92.19% |
92.14% |
76.28% |
370 - 0.362 |
|
基因(前十) |
10.0 |
10.0 |
10.0 |
10.0 |
10.0 |
10.0 |
|
|
Acc |
82.44% |
88.33% |
72.38% |
92.24% |
90.14% |
76.15% |
807 - 0.452 |
|
基因(前20名) |
20.0 |
20.0 |
20.0 |
20.0 |
20.0 |
20.0 |
|
|
Acc |
83.97% |
84.50% |
69.90% |
92.24% |
92.19% |
74.62% |
936 - 0.236 |
|
基因(前40名) |
40.0 |
40.0 |
40.0 |
40.0 |
40.0 |
40.0 |
|
|
| mRMR |
Acc |
85.38% |
82.92% |
58.76% |
89.33% |
90.24% |
74.74% |
822 - 0.016 |
|
基因 |
10.0 |
10.0 |
10.0 |
10.0 |
10.0 |
10.0 |
|
|
Acc |
83.97% |
85.67% |
68.38% |
92.24% |
92.24% |
68.46% |
739 - 0.037 |
|
基因 |
20.0 |
20.0 |
20.0 |
20.0 |
20.0 |
20.0 |
|
|
Acc |
82.31% |
86.92% |
75.33% |
93.24% |
91.19% |
70.13% |
025 - 0.180 |
|
基因 |
40.0 |
40.0 |
40.0 |
40.0 |
40.0 |
40.0 |
|
|
| 慢性疲劳综合症 |
Acc |
82.18% |
92.08% |
68.48% |
93.19% |
91.19% |
72.95% |
408 - 0.456 |
|
基因 |
26.8 |
92.0 |
90.0 |
78.8 |
33.8 |
44.0 |
|
|
| WrapperNB |
Acc |
76.15% |
76.58% |
60.00% |
91.14% |
88.33% |
69.87% |
557 - 0.002 |
|
基因 |
6.2 |
4.2 |
6.6 |
5.2 |
5.4 |
6.2 |
|
|
| 遗传算法 |
Acc |
71.03% |
76.67% |
70.00% |
63.81% |
86.33% |
69.49% |
114 - 0.000 |
|
基因 |
122.2 |
378.0 |
514.8 |
803.2 |
105.6 |
161.8 |
|
|
| 算法 |
Acc |
65.90% |
81.92% |
68.29% |
71.67% |
82.48% |
71.03% |
026 - 0.000 |
|
基因 |
134.0 |
371.6 |
455.6 |
506.2 |
115.2 |
150.4 |
|
|
|
|
最好的分类精度为每个基准数据集以粗体显示。
|
