一种基于三角模糊数多指标信息的FCM聚类算法

An FCM clustering algorithm for multiple attribute information with triangular fuzzy numbers

针对一类具有不确定性三角模糊数多指标信息的聚类分析问题,基于传统的数值信息FCM聚类算法,提出一种新的聚类分析算法.首先描述了具有三角模糊数多指标信息的聚类分析问题,提出并证明了基于三角模糊数多指标信息的关于最优划分和最优聚类中心确定的两个定理;然后根据这两个定理,进一步给出了基于三角模糊数信息的FCM聚类算法的迭代步骤;最后通过一个算例说明了该聚类算法的具体应用.

Aiming at multiple attribute clustering analysis problems with uncertain triangular fuzzy numbers, a new clustering analysis algorithm is proposed based on the traditional FCM clustering algorithm. The multiple attribute clustering analysis problem with triangular fuzzy numbers is introduced. Then two theorems for determining the (optimal) partition and the optimal clustering center are proposed and proved. Based on that, calculation steps of the FCM clustering algorithm for multiple attribute information with triangular fuzzy numbers are presented. Finally, a numerical example shows the applicability of the FCM clustering algorithm.