计算机工程与应用 ›› 2009, Vol. 45 ›› Issue (25): 218-221.DOI: 10.3778/j.issn.1002-8331.2009.25.067

• 工程与应用 • 上一篇    下一篇

基于半监督技术的多分类器融合策略研究

蔡 晰1,2,郭躬德1,2,黄添强1,2   

  1. 1.福建师范大学 数学与计算机科学学院,福州 350007
    2.福建师范大学 网络安全与密码技术重点实验室,福州 350007
  • 收稿日期:2008-05-15 修回日期:2008-07-28 出版日期:2009-09-01 发布日期:2009-09-01
  • 通讯作者: 蔡 晰

Multiple classifiers fusion method based on semi-supervised learning

CAI Xi1,2,GUO Gong-de1,2,HUANG Tian-qiang1,2   

  1. 1.School of Mathematics and Computer Science,Fujian Normal University,Fuzhou 350007,China
    2.Key Lab of Network Security and Cryptography,Fujian Normal University,Fuzhou 350007,China
  • Received:2008-05-15 Revised:2008-07-28 Online:2009-09-01 Published:2009-09-01
  • Contact: CAI Xi

摘要: 提出一种新颖的多分类器构造方法,它以最大纠错能力作为分类器选择标准。实现时,采用半监督协同训练技术,充分利用单分类器的互补性,同时最大化仲裁器的仲裁能力,以提高多分类器系统的分类精度。在毒性数据集上的实验结果表明了方法的可行性和有效性。

关键词: 多分类器, 协同训练, 仲裁器, 半监督学习

Abstract: This paper proposes a novel strategy for multi-classifier classification.The method takes maximal error correcting ability as a criterion of choosing classifiers.To improve the accuracy of multi-classifier classification,a semi-supervised co-training technology is employed which makes use of the complementarity of each single classifier and maximizes the judging ability of the arbiter as well.The experimental results show the mothod is practical and effective on real toxicity dataset.

Key words: multi-classifier classification, co-training, arbiter, semi-supervised learning

中图分类号: