基于PSO自动确定聚类数目的FCM算法  

FCM Algorithm Based on PSO for Automatically Determining the Number of Clusters

作　　者：肖剑文 周亦敏[1] XIAO Jiao-wen;ZHOU Yi-min(School of Optical-electrical and Computer Engineering,University of Shaghai for Science and Technology, Shanghai 20093, China)

机构地区：[1]上海理工大学光电信息与计算机工程学院,上海200093

出　　处：《软件导刊》2018年第12期104-107,共4页Software Guide

摘　　要：模糊C均值聚类是聚类分析中应用最广泛的算法之一,但是聚类数目需要人为预先设定,在实际应用中有极大的局限性。提出一种自动确定聚类数目的基于粒子群的模糊C均值聚类算法,通过对不同聚类数目进行试验,利用添加粒子阈值向量自动确定最佳的聚类数目。在预设的最大聚类数目内随机分割数据集,利用重构准则重新构建初始值,以此克服需要事先设置聚类数目的模糊C均值缺点。利用有效性函数评估算法性能,试验结果表明,该算法能自动找到最优聚类数目,聚类效果很好。Fuzzy C-means clustering is one of the most widely used algorithms in cluster analysis.However,the number of clusters needs to be preset manually,which has great limitations in practical applications.In this paper,a particle swarm optimization fuzzy C-means clustering algorithm is proposed to automatically determine the number of clusters.This algorithm uses the additive particle threshold vector to automatically determine the optimal number of clusters by experimenting with different cluster numbers.The algorithm firstly randomly divides the data set within the preset maximum number of clusters,reconstructs the initial value by using the reconstruction criterion,and overcomes the shortcomings of the fuzzy C-means that need to set the number of clusters in advance,and uses the validity function to evaluate the algorithm.The experimental results show that the algorithm can automatically find the optimal number of clusters and achieve a good clustering effect.

关 键 词：模糊C均值 粒子群 粒子阈值 重构准则 有效性函数 

分 类 号：TP312[自动化与计算机技术—计算机软件与理论]