摘要
为克服核模糊属性c-均值聚类算法易陷入局部最优解的缺点,提出一种新的基于粒子群优化的核模糊属性c-均值聚类算法.该算法根据核模糊属性c-均值聚类准则设计适应度函数,利用粒子群优化算法对聚类中心进行优化,在粒子迭代进化过程中采用动态调整学习因子,提高算法的优化性能.实验表明,本文算法优于单一使用核模糊属性c-均值聚类算法和基于粒子群优化的核模糊c-均值聚类算法,也优于目前常见的典型聚类算法.
To overcome the deficiency of being easily trapped into local minima of kernelized fuzzy attribute C-means clustering algorithm,a novel kernelized fuzzy attribute C-means clustering algorithm based on particle swarm optimization is proposed.A fitness function is designed according to the clustering principle and the method of particle swarm optimization is utilized to optimize the cluster centers.During the procedure of the particle update,the learning factors are adjusted according to its relative position to improve the performance of the proposed algorithm.The experiment results show that the new proposed algorithm is better than kernelized fuzzy attribute C-means clustering algorithm or particle swarm optimization based on kernelized fuzzy C-means clustering algorithm,which is also superior to the typical clustering algorithm in clustering ability and stability.
出处
《广西师范学院学报（自然科学版）》
2010年第4期86-90,共5页
Journal of Guangxi Teachers Education University(Natural Science Edition)
基金
广西自然科学基金资助项目(桂自1013054)
关键词
粒子群优化
C-均值聚类
稳态函数
核聚类
核模糊属性c-均值聚类
particle swarm optimization
C-means clustering
stable function
kernel clustering
kernelized fuzzy attribute C-means clustering
