基于特征属性分类改进的模糊聚类方法及其应用  

An Improved Fuzzy Clustering Method and Its Application Based on Characteristic Attribute Classification

作　　者：毛林[1] 杨学兵[1] 岳宗敏[1] 

机构地区：[1]安徽工业大学计算机学院,安徽243002

出　　处：《微计算机应用》2007年第5期511-515,共5页Microcomputer Applications

基　　金：安徽省教育厅自然科学基金重点资助项目(编号:2004KJ053ZD)

摘　　要：模糊聚类分析是一种重要的分类方法。传统模糊聚类分析法着眼于全体属性,在对多属性数据集分类方面具有明显优势,对基于特定、重要属性的分类时显得不足。本文对传统方法进行改进,提出了一种基于特征属性分类的模糊聚类方法,利用特征属性进行分类,产生了较好的分类效果,展示了一个成用实例。改进的方法人人提高了特定分类问题的应用价值。Fuzzy clustering analysis belongs to one of important classification methods. All attributes are under principally original consideration of traditional clustering analysis, which is of obvious advantage in classification of data sets. Traditional methods implement hardly any classification based on specific attributes. A characteristic attribute - based fuzzy clustering method, which focuses on highlighted attributes through characteristic attribute, is established and results in comparatively reasonable effect. The new method is illustrated by an example. Improved method indicates practical applied values for issue of specific classification.

关 键 词：模糊聚类 分类精度 特征属性 λ-水平分类 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]