计算机工程与应用 ›› 2014, Vol. 50 ›› Issue (13): 133-136.

• 数据库、数据挖掘、机器学习 • 上一篇    下一篇

基于相似关系向量的改进ROUSTIDA算法

丁春荣1,李龙澍2   

  1. 1.安徽农业大学 信息与计算机学院,合肥 230036
    2.安徽大学 计算机科学与技术学院,合肥 230039
  • 出版日期:2014-07-01 发布日期:2015-05-12

Improved ROUSTIDA algorithm based on similarity relation vector

DING Chunrong1, LI Longshu2   

  1. 1.School of Information and Computer Science, Anhui Agricultural University, Hefei 230036, China
    2.College of Computer Science and Technology, Anhui University, Hefei 230039, China
  • Online:2014-07-01 Published:2015-05-12

摘要: 在对粗糙集理论进行深入研究的基础上,依据决策规则独立原则,提出了一种基于相似关系向量的不完备信息系统数据补齐算法,有效地解决了原有ROUSTIDA算法可能存在的决策规则矛盾问题。实例证明改进算法的填补正确率比ROUSTIDA算法高,具有更好的实用价值。

关键词: 粗糙集, 不完备信息系统, 相似关系向量, 可辨识矩阵

Abstract: Based on a deep research of the rough set theory, an completing data algorithm of the incomplete information system based on similarity relation vector is proposed according to the decision rule independence principle. It can efficiently solve the potential contradiction between decision rules comparing with the original ROUSTIDA algorithm. A practical case proves that the completing accuracy rate is upper than ROUSTIDA algorithm’s, and the new algorithm has better practical value.

Key words: rough set, Incomplete Information System(IIS), similarity relation vector, distinct matrix