系统管理学报2011，Vol.20Issue(6)：728-733,6.
一种融合遗传算法和粒子群算法的改进模糊C-均值算法
An Improved FCM Algorithm Integrating GA and PSO
摘要
Abstract
This paper proposes a GA-PSO-FCM algorithm that integrates GA with PSO to overcome the shortcomings that FCM algorithm needs a given cluster number c in advance and is easy to get into local minimum. GA is embedded as the outer layer of FCM algorithm to search for the optimal cluster number, and takes validity index as its fitness function; PSO is embedded as the inner layer of FCM algorithm to optimize the cluster center vector to improve the ability of global search. The GA-PSO-FCM algorithm was tested with Iris data, Wine data, Zoo data, WPBC data and WDBC data. The results were compared with simulation result of FCM and GA-FCM. It shows that GA-PSO-FCM algorithm may improve the accuracy and stability of clustering without knowing the cluster number in advance.关键词
模糊C-均值/有效性准则/遗传算法/粒子群算法Key words
fuzzy C-means (FCM)/ validity index/ genetic algorithm (GA)/ particle swarm optimization (PSO)分类
数理科学引用本文复制引用
诸克军,李兰兰,郭海湘..一种融合遗传算法和粒子群算法的改进模糊C-均值算法[J].系统管理学报,2011,20(6):728-733,6.基金项目
国家自然科学基金资助项目(71173202) （71173202）
国家自然科学基金青年科学基金资助项目(71103163) （71103163）
教育部人文社会科学研究青年基金资助项目(10YJC790071) （10YJC790071）
中央高校基本科研业务费专项资金资助项目(CUG110411) （CUG110411）
