计算机科学与探索 ›› 2014, Vol. 8 ›› Issue (4): 467-472.DOI: 10.3778/j.issn.1673-9418.1307024

• 人工智能与模式识别 • 上一篇    下一篇

决策表中基于对象的μ-约简方法研究

翟翠红,秦克云+   

  1. 西南交通大学 数学学院,成都 610031
  • 出版日期:2014-04-01 发布日期:2014-04-03

Research on μ-Reduction Method for Objects in Decision Tables

ZHAI Cuihong, QIN Keyun+   

  1. College of Mathematics, Southwest Jiaotong University, Chengdu 610031, China
  • Online:2014-04-01 Published:2014-04-03

摘要: 研究了不协调决策表中基于对象的μ-约简问题,给出了μ-约简的一些性质及等价定义,得到了μ-约简的判定定理和相应的可辨识矩阵及辨识公式,从而提供了不协调目标信息系统知识约简的新方法。并通过实例将决策表中基于对象的约简与整体约简进行了对比,结果表明前者可以得到更简洁的知识。该方法也适用于协调目标信息系统。

关键词: 不协调决策表;&mu, -约简;辨识矩阵;辨识公式

Abstract: This paper studies the μ- reduction for the objects in inconsistent decision tables, introduces some properties and the equivalent definition of μ- reduction, obtains the judgment theorem and corresponding identification matrix and formula with respect to the μ-reduction, and proposes new approaches to knowledge reductions in inconsistent information systems. Compared with the overall reduction for a decision table, the reduction for the objects in the decision table not only can get more concise knowledge, but also has good application value in real life. In addition, this method is also suitable for the consistent information systems.

Key words: inconsistent decision table, μ-reduction, identification matrix, identification formula