研究文章
神经网络:一种改进的FCM用于多通道文化数据分析
|
| 实验序列 |
算法 |
的迭代次数 |
均方误差 |
|
| 1 |
k - means |
11 |
0.398678 |
| DBSCAN |
23 |
0.174943 |
| 传统的FCM |
25 |
0.165567 |
| 改进的FCM |
19 |
0.096729 |
|
| 2 |
k - means |
8 |
0.225329 |
| DBSCAN |
21 |
0.147179 |
| 传统的FCM |
18 |
0.195696 |
| 改进的FCM |
16 |
0.119679 |
|
| 3 |
k - means |
13 |
0.262847 |
| DBSCAN |
22 |
0.145397 |
| 传统的FCM |
19 |
0.185628 |
| 改进的FCM |
22 |
0.156727 |
|
| 4 |
k - means |
16 |
0.375326 |
| DBSCAN |
41 |
0.189642 |
| 传统的FCM |
33 |
0.235626 |
| 改进的FCM |
28 |
0.204647 |
|
| 5 |
k - means |
6 |
0.228643 |
| DBSCAN |
20. |
0.174254 |
| 传统的FCM |
19 |
0.167687 |
| 改进的FCM |
16 |
0.128036 |
|
| 6 |
k - means |
11 |
0.275984 |
| DBSCAN |
29日 |
0.228587 |
| 传统的FCM |
24 |
0.203187 |
| 改进的FCM |
26 |
0.180997 |
|
| 7 |
k - means |
14 |
0.435079 |
| DBSCAN |
37 |
0.330578 |
| 传统的FCM |
32 |
0.372068 |
| 改进的FCM |
29日 |
0.239304 |
|
| 8 |
k - means |
12 |
0.267865 |
| DBSCAN |
28 |
0.246841 |
| 传统的FCM |
30. |
0.253245 |
| 改进的FCM |
25 |
0.192422 |
|
| 9 |
k - means |
16 |
0.296781 |
| DBSCAN |
27 |
0.234778 |
| 传统的FCM |
25 |
0.241987 |
| 改进的FCM |
29日 |
0.237179 |
|
| 10 |
k - means |
12 |
0.443921 |
| DBSCAN |
25 |
0.246895 |
| 传统的FCM |
23 |
0.298679 |
| 改进的FCM |
25 |
0.224349 |
|
|
