计算机应用 ›› 2012, Vol. 32 ›› Issue (10): 2888-2890.DOI: 10.3724/SP.J.1087.2012.02888

• 人工智能 • 上一篇    下一篇

基于多标签ReliefF的特征选择算法

黄莉莉1,2,汤进1,2,孙登第1,2,罗斌1,2   

  1. 1. 安徽大学 计算机科学与技术学院,合肥 230601
    2. 安徽省工业图像处理与分析重点实验室,合肥 230039
  • 收稿日期:2012-04-05 修回日期:2012-05-18 发布日期:2012-10-23 出版日期:2012-10-01
  • 通讯作者: 黄莉莉
  • 作者简介:黄莉莉(1986-),女,河南周口人,硕士研究生,主要研究方向:图像处理、模式识别;汤进(1976-),男,安徽合肥人,副教授,博士,主要研究方向:红外图像处理、模式识别;孙登第(1983-),安徽淮南人,博士研究生,主要研究方向:数据挖掘、模式识别;罗斌(1963-),男,安徽合肥人,教授,博士,主要研究方向:模式识别、数字图像处理。
  • 基金资助:
    国家自然科学基金资助项目;安徽省高校自然科学研究重点项目

Feature selection algorithm based on multi-label ReliefF

HUANG Li-li1,2,TANG Jin1,2,SUN Deng-di1,2,LUO Bin1,2   

  1. 1. Key Laboratory for Industrial Image Processing and Analysis of Anhui Province, Hefei Anhui 230039, China
    2. School of Computer Science and Technology, Anhui University, Hefei Anhui 230601, China
  • Received:2012-04-05 Revised:2012-05-18 Online:2012-10-23 Published:2012-10-01
  • Contact: HUANG Li-li

摘要: 针对传统特征选择算法局限于单标签数据问题,提出一种多标签数据特征选择算法——多标签ReliefF算法。该算法依据多标签数据类别的共现性,假设样本各类标签的贡献值是相等的,结合三种贡献值计算方法,改进特征权值更新公式,最终获得有效的分类特征。分类实验结果表明,在特征维数相同的情况下,多标签ReliefF算法的分类正确率明显高于传统特征选择算法。

关键词: 特征选择, 多标签, ReliefF, 降维, 模式识别

Abstract: The traditional feature selection algorithms are limited to single-label data. Concerning this problem, multi-label ReliefF algorithm was proposed for multi-label feature selection. For multi-label data, based on label co-occurrence, this algorithm assumed the label contribution value was equal. Combined with three new methods calculating the label contribution, the updating formula of feature weights was improved. Finally a distinguishable feature subset was selected from original features. Classification experiments demonstrate that, with the same number of features, classification accuracy of the proposed algorithm is obviously higher than the traditional approaches.

Key words: feature selection, multi-label, ReliefF, dimensionality reduction, pattern recognition