Computer Engineering and Applications ›› 2010, Vol. 46 ›› Issue (14): 132-134.DOI: 10.3778/j.issn.1002-8331.2010.14.039

• 数据库、信号与信息处理 • Previous Articles     Next Articles

Using Fisher linear discriminant analysis to extracting classifiers

LI Wen-bin,CHEN Yi-ying,ZHANG Juan,ZHANG Xin-dong   

  1. School of Information and Engineering,Shijiazhuang University of Economics,Shijiazhuang 050031,China
  • Received:2008-10-30 Revised:2009-11-02 Online:2010-05-11 Published:2010-05-11
  • Contact: LI Wen-bin


李文斌,陈嶷瑛,张 娟,张新东   

  1. 石家庄经济学院 信息工程学院,石家庄 050031
  • 通讯作者: 李文斌

Abstract: In order to eliminate relativity between ensembled classifiers and improve effect and stability of combiner,an approach extracting classifiers based on Fisher linear discriminant analysis is proposed.It can reduce classifier space with high dimension,and then learn a combiner in lower dimension space.The compared results obtained on multiple public avaiable datasets show that the method is feasible.Its performance is perfect.

Key words: machine learning, data mining, text processing, classification

摘要: 为了消除个体分类器间的相关性,提高集成器分类性能及稳定性,提出了基于Fisher线性判别方法的分类器提取方法。该方法将高维分类器空间压缩至低维分类器空间,并在该空间内学习集成器。在多个数据集上的比较实验结果表明,该方法可行,其集成性能较理想。

关键词: 机器学习, 数据挖掘, 文本处理, 分类

CLC Number: