An Improvement Algorithm Based on Fuzzy C-Means Clustering Algorithm

摘要：

For the fuzzy C-means (FCM) algorithm of the shortcomings, an improved fuzzy algorithm is proposed. For the nonuniform distribution of the sample points, we use the k-nearest neighbor density of the point as weighted value and outlier removing in the algorithm. Experiment results show the clustering results that the improved method is used are better than those of only using the FCM.

关键词： FCM outlier removing k-nearest neighbor density of the point Introduction

作者: Huang Minghua Yu Yongquan Lv Zhande

作者单位: Institute of Computer, Guangdong University of Technology Guangzhou, China

会议类型: 国际会议

会议名称: 2010 International Conference on Information Security and Artificial Intelligence(2010年信息安全与人工智能国际会议 ISAI 2010)

会议地点: 成都

会议语种:英文

页码: 873-875

在线出版日期: 2010-12-17（万方平台首次上网日期，不代表论文的发表时间）