| 建立在 SUGENO 测度上的最佳拓扑结构聚类方法 |
| |
| 引用本文： | 曹海兴,王成道.建立在 SUGENO 测度上的最佳拓扑结构聚类方法[J].华东师范大学学报(自然科学版),1988(1). |
| |
| 作者姓名： | 曹海兴 王成道 |
| |
| 作者单位： | 上海轻工业专科学校，华东师范大学电子系 |
| |
| 摘 要： | Gustafson 和 Kessel 等人为试图识别在同一个数据集合中不同子集类可以有不同拓扑结构这一事实而给出一种最佳 Fuzzy 聚类方法,本文在此基础上,在目标函数中用 Sugeno 测度替代 Fuzzy 划分 m 次幂的权因子,并在不同的子集类中引入不同的拓扑度量,从而建立了一种改进了的 Fuzzy 聚类算法,即 SMFCM 算法(Sugeno's Measure Fuzzy-C Means)。本方法既避免了原有算法中对参数 m 的确定缺乏合理的规则这一困难,又能适用对具有不同子结构形状的数据集进行识别,因而可看作是 Gustafson、Kessel 及 Leszc(?)ynski 等人所给出方法的扩充。最后,用本文所提出的方法对四个实例进行了计算,获得了较满意的结果。
|
| 关 键 词： | 模糊分类 模式识别 模式 |
The Optimization Clustering Algorithm of Topological Structure Based on Sugeno Measure Space |
| |
| CAO HAIXIN Shang Hai Light Industry College,WANG CHENGDAO.The Optimization Clustering Algorithm of Topological Structure Based on Sugeno Measure Space[J].Journal of East China Normal University(Natural Science),1988(1). |
| |
| Authors: | CAO HAIXIN Shang Hai Light Industry College WANG CHENGDAO |
| |
| Institution: | CAO HAIXIN Shang Hai Light Industry College
WANG CHENGDAO Department of Electronics Science and Technnology |
| |
| Abstract: | The paper presents a clustering method,the SMFCM(Sugeno'Measure Fuzzy-cMeans)algorithm.It attempts to recognize the fact that different clusters in thesame data set may have different topological structure.The clustering criterion isof optimization for nonlinear objective functional involving the weight of Sugeno'smeasure.The norms are induced by symmetric,positive-define matrices.In the end,four experimental results of data are offered. |
| |
| Keywords: | fuzzy classification pattern recognition pattern |
| 本文献已被 CNKI 等数据库收录！ |
