计算机工程 ›› 2010, Vol. 36 ›› Issue (21): 62-64.doi: 10.3969/j.issn.1000-3428.2010.21.022

• 软件技术与数据库 • 上一篇    下一篇

一种扩展正域的属性约简方法

冯 林1,2   

  1. (1. 四川师范大学计算机科学学院,成都 610101;2. 四川师范大学可视化计算与虚拟现实四川省重点实验室,成都 610068)
  • 出版日期:2010-11-05 发布日期:2010-11-03
  • 作者简介:冯 林(1972-),男,副教授,主研方向:粗糙集理论,粒计算,数据挖掘
  • 基金项目:
    四川省教育厅科研基金资助项目(09ZC079);四川师范大学重点研究课题基金资助项目

Attribute Reduction Approach Under Extended Positive Region

FENG Lin1,2   

  1. (1. College of Computer Science, Sichuan Normal University, Chengdu 610101, China; 2. Sichuan Province Key Laboratory of Visualization Computing and Virtual Reality, Sichuan Normal University, Chengdu 610068, China)
  • Online:2010-11-05 Published:2010-11-03

摘要: 在经典粗糙集理论模型中,边界域过大会限制其实际应用。针对这一情况,指出决策表中边界域扩展为正域已有方法存在的不足,基于不确定条件下的自主式学习理论,定义一种新的扩展正域方法,并提出计算不相容决策表中认知属性核和认知属性约简的算法。实验结果证明了该方法的有效性。

关键词: 决策表, 粗糙集, 属性约简, 核属性

Abstract: Large boundary region is one of the barriers for classical rough set theory applications. In this paper, drawbacks of the existing approaches for extended positive region from boundary region are discussed. Based on self-learning model under uncertain condition, a new approach for extended positive region is constructed. Algorithms of cognitive attribute core and attribute reduction are presented. Simulation results illustrate the efficiency of these algorithms.

Key words: decision table, rough set, attribute reduction, core attribute

中图分类号: