摘要
针对模糊聚类算法存在的问题,通过对聚类有效性函数的分析,对聚类数c和加权指数m进行改进,将改进后的模糊聚类算法引入BP算法中,建立基于模糊聚类与BP算法的混合模型,并进行实验分析,分析结果表明,混合模型在准备性上优于传统的BP算法,因为数据经过模糊聚类之后同类数据具有更多的相似特征。
The clustering validity function was analyzed;and the clustering number c and index m were modified.The improved fuzzy clustering algorithm was introduced into BP algorithm.A hybrid model was then established based on fuzzy clustering and BP algorithm to solve the existed problems in the fuzzy clustering analysis.Experimental analysis was also conducted.The results show that the hybrid model is superior to the traditional BP algorithm method in the preparation,because the data have more similar characteristics after they have been fuzzily clustered.
出处
《武汉理工大学学报（信息与管理工程版）》
CAS
2011年第5期694-697,共4页
Journal of Wuhan University of Technology：Information & Management Engineering
基金
武汉理工大学自主创新研究基金资助项目(2010-ZY-LX-033
20520012)
关键词
模糊聚类
有效性
BP算法
fuzzy clustering
effectiveness
BP algorithm
