计算机工程与应用 ›› 2007, Vol. 43 ›› Issue (1): 178-178.

• 数据库与信息处理 • 上一篇    下一篇

一致性分类问题中规则提取的新方法

张小峰,周凤翔,宋丽华,孙玉娟   

  1. 鲁东大学计算机科学与技术学院
  • 收稿日期:2006-01-04 修回日期:1900-01-01 出版日期:2007-01-01 发布日期:2007-01-01
  • 通讯作者: 张小峰 iamzxf

Novel Method for Rule Extraction in Consistent Classification Problem

Xiaofeng Zhang,,   

  1. 鲁东大学计算机科学与技术学院
  • Received:2006-01-04 Revised:1900-01-01 Online:2007-01-01 Published:2007-01-01
  • Contact: Xiaofeng Zhang

摘要: 研究了一致性分类问题中挖掘分类规则的算法,提出了一种基于属性重要性及partition的分类规则的新方法。首先根据属性的重要性度量选择一个属性,根据该属性与决策属性构造信息系统的分区,然后对每一个分区进行描述,并根据是否会在信息系统中产生冲突对描述进行简化,最后根据简化后的描述形成分类规则。

Abstract: This paper presents a novel method to extract classification rules based on the importance of attributes and partition technology after investigating current algorithms in classification problems. First it will select one attribute according to the importance, and then will construct the partition of information system with respect to the selected attribute and decision ones. After describing all blocks of the partitions, this method will simplify the description and extract rules from the reduced description.