基于密度函数的FCM聚类中心初始化研究综述

Research Review of FCM Clustering Center Initialization Based on the Densit6Xy Function

模糊C均值(FCM)聚类算法能很好地解决不确定问题的分类,但该算法聚类结果却过于依赖初始聚类中心和易陷入局部最优解。本文重点针对基于密度函数的初始聚类中心初始化算法进行分类总结,将密度函数的度量方式归纳为4大类。通过实验对比分析了各种度量方式的优缺点,旨在为研究人员选择适合的密度函数度量方法提供一定的参考。

Fuzzy c-means（FCM） algorithm is a good way to solve the problem of uncertainty classification, however,of which the clustering result depends too much on the initial clustering center and falls into local optimal. This paper focuses on the clustering center initialization algorithm based on density function and then summarizes the classification of initial algorithms by categorizing four major types of measurement for density functions. Through experiments,the paper compares and analyzes the advantages and disadvantages of all kinds of measure, aiming at providing reference for researchers to choose measure for density function properly.