期刊文献+

基于可能性理论的聚类分析 被引量:3

Clustering Based on Possibility Theory
下载PDF
导出
摘要 可能性理论是处理不确定、不精确数据或信息的一种方法,它是建立在模糊集理论上,在聚类中引入可能性理论,符合人们的思维方式。该文根据可能性理论,从聚类集合的对象之间的相似度上考虑提出了可能性匹配,它通过可能相似程度和必要相似程度概念来表达,可能相似程度尽可能地消除人聚类的主观误差,而必要相似程度尽可能保存信息,符合人们的思维方式,在聚类分析广泛应用的现实中具有较好的应用前景。 In term of the clustering modeling,a more reasonable model is dicussed in this paper,which is based on the possibility theory.Based on the possibility theory(PT)and the clustering,the notion of“possibility matching”and“necessary matching”are present ,based on the max-min composition according to the PT.And using the similarity matrixes to store every two objects'similarity.Basing the above notion,we present a possibility clustering model to deal the classification,and give a detail algorithm.
出处 《计算机工程与应用》 CSCD 北大核心 2003年第13期85-87,共3页 Computer Engineering and Applications
基金 国家863重大关键技术攻关项目(编号:863-511-030-007-4)
关键词 可能性理论 可能性匹配 可能性聚类分析 聚类分析 possibility theory,possibility matching,necessary matching,clustering
  • 引文网络
  • 相关文献

参考文献5

  • 1Smets P.Imperfect information:Imprecision & Uncertainty UMIS-Var Unc.1999.
  • 2Gebhardt J,Kruse R.Parallel combination of information sources[C]. In:D M Gabbay,P Smets eds.Handbook of defeasible reasoning and uncertainty management systems,Kluwer,Doordrecht,The Netherlands, vol 3:393-440.
  • 3Dubois D,Prade H,Smets P.New Semantics for Quantitative Possibility Theory[C].In:2nd International Symposium on Imprecise Probabilities and Their Applications,NewYork,2001.
  • 4Smets P.Probability.possibility.belief:which and where?.1999.
  • 5Ranjit Biswas.An application of fuzzy sets in student's evaluation[J].Fuzzy Sets and Systems ,1995 ; 74 :187-194.

同被引文献24

引证文献3

二级引证文献3

相关主题

;
使用帮助 返回顶部