作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程 ›› 2011, Vol. 37 ›› Issue (3): 47-48. doi: 10.3969/j.issn.1000-3428.2011.03.017

• 软件技术与数据库 • 上一篇    下一篇

基于近似度的不完备信息系统属性约简

汪琼枝1a,1b,吴 涛1a,1b,2,毛军军1a,1b,张培根1b   

  1. (1. 安徽大学 a. 智能计算与信号处理教育部重点实验室;b. 数学科学学院,合肥 230039; 2. 南京大学计算机软件新技术国家重点实验室,南京 210093)
  • 出版日期:2011-02-05 发布日期:2011-01-28
  • 作者简介:汪琼枝(1983-),女,硕士研究生,主研方向:智能计算,机器学习;吴 涛,教授、博士;毛军军,副教授、博士;张培根,本科生
  • 基金资助:
    中国博士后基金资助面上项目(20070411028);国家自然科学基金资助项目(60675031);安徽省自然科学基金资助项目(KJ 2008B093);安徽大学学术创新团队基金资助项目(KJTD001B);安徽大学大学生创新性实验基金资助项目(30005);安徽大学人才队伍建设基金资助项目

Incomplete Information System Attribute Reduction Based on Approximate Degree

WANG Qiong-zhi 1a,1b, WU Tao 1a,1b,2, MAO Jun-jun 1a,1b, ZHANG Pei-gen1b   

  1. (1a. Key Laboratory of Intelligent Computing & Signal Processing of Ministry of Education; 1b. School of Mathematical Sciences, Anhui University, Hefei 230039, China; 2. State Key Laboratory for Novel Software Technology, Nanjing University, Nanjing 210093, China)
  • Online:2011-02-05 Published:2011-01-28

摘要: 在容差关系下,针对属性集P在分类中的不同个体贡献,引入粗糙集的近似度概念,结合属于/不属于集合Rp(x)的元素贡献的不确定性,定义一种新的知识熵,从而实现基于近似度的不完备信息系统属性约简算法。仿真结果表明,与IEARA算法相比,该算法具有较高的约简效率。

关键词: 近似度, 不完备信息系统, 属性约简

Abstract: This paper considers the differences of the individual contribution based on attribute set P classification in the tolerance relation, introduces the notion of approximate degree for the rough set. It defines a new entropy based on the consideration of both the elements belong to a set of Rp(x) and does not belong to Rp(x). On this basis, an attribute reduction algorithm based on approximate degree for incomplete information system is realized. Simulation results show that this algorithm has better reduction efficiency than IEARA algorithm.

Key words: approximate degree, incomplete information system, attribute reduction

中图分类号: