计算机工程与应用 ›› 2010, Vol. 46 ›› Issue (3): 49-50.DOI: 10.3778/j.issn.1002-8331.2010.03.015

• 研究、探讨 • 上一篇    下一篇

基于粗糙集约简的多分类器系统构造方法

王名扬1,2,于达仁2,胡清华2   

  1. 1.东北林业大学 信息与计算机工程学院,哈尔滨 150040
    2.哈尔滨工业大学 航天学院,哈尔滨 150001
  • 收稿日期:2008-09-04 修回日期:2008-11-12 出版日期:2010-01-21 发布日期:2010-01-21
  • 通讯作者: 王名扬

Technique for constructing multiple classification systems based on rough set reduction

WANG Ming-yang1,2,YU Da-ren2,HU Qing-hua2   

  1. 1.Information and Computer Engineering College,Northeast Forestry University,Harbin 150040,China
    2.School of Astronautic,Harbin Institute of Technology,Harbin 150001,China
  • Received:2008-09-04 Revised:2008-11-12 Online:2010-01-21 Published:2010-01-21
  • Contact: WANG Ming-yang

摘要: 多分类器系统是近年来兴起的一种有效的分类机制,为提高多分类器系统的分类精度,提出了一种基于粗糙集约简构造多分类器系统的机制,并从输入和输出两个角度对如何选择单个分类器进行了探讨。通过对4个UCI数据集进行验证,发现基于输出的选择融合方法得到了最好的分类效果。

关键词: 多分类器系统, 粗糙集, 约简

Abstract: Multiple classification systems(MCSs) are a new classification technique rising in recent years.In order to the classification performance of MCSs,a technique is proposed to construct multiple classification systems based on rough set reduction.The method of selecting the single classifiers in two points is presented,which is the point of input and output respectively.Experiments show that the MCSs constructed by the method based on output get the best classification performances.

Key words: multiple classification systems, rough set, reduction

中图分类号: