计算机工程与应用 ›› 2015, Vol. 51 ›› Issue (17): 134-139.

• 数据库、数据挖掘、机器学习 • 上一篇    下一篇

不同约简准则的比较研究

黄治国,程  浩   

  1. 河南工程学院 计算机学院,郑州 451191
  • 出版日期:2015-09-01 发布日期:2015-09-14

Comparative study on alternative types of reduction criterion

HUANG Zhiguo, CHENG Hao   

  1. School of Computer Science, Henan Institute of Engineering, Zhengzhou 451191, China
  • Online:2015-09-01 Published:2015-09-14

摘要: 在决策系统属性约简过程中,约简准则反映了保持系统特定信息分辨能力不变的特性,但迥异外在形式的约简准则可能具有相同的本质内涵。针对不同外在形式的多种常见约简准则,比较研究其基本特征,分析其本质内涵并将之分层归类,为进一步实现面向不同约简准则的高效属性约简算法奠定理论基础。

关键词: 粗糙集, 决策系统, 属性约简, 约简准则

Abstract: In the process of attribute reduction in decision system, reduction criterion reflects the characteristic of remaining particular information discernibility, but different types of reduction criterion may possess the same essence. Aiming at many common reduction criterions with different description form, their basic characteristic and essential nature are studied in detail, and be classified in hierarchical structure. The study result establishes a theoretical foundation for efficient reduction algorithm oriented to different reduction criterion.

Key words: rough set, decision system, attribute reduction, reduction criterion