计算机科学与探索 ›› 2008, Vol. 2 ›› Issue (2): 192-197.

• 学术研究 • 上一篇    下一篇

SVM置信度在线评估以及决策改进

凌 萍1,2 +,周春光1   

  1. 1. 吉林大学 计算机科学与技术学院 教育部符号计算与知识工程重点实验室,长春 130012
    2. 徐州师范大学 计算机科学与技术学院,江苏 徐州 221116
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2008-04-20 发布日期:2008-04-20
  • 通讯作者: 凌 萍

For SVM: confidence online evaluation & decision improvement

LING Ping1,2 +, ZHOU Chunguang1   

  1. 1. Key Laboratory of Symbol Computation and Knowledge Engineering of the Ministry of Education, College of Computer Science, Jilin University, Changchun 130012, China
    2. College of Computer Science, Xuzhou Normal University, Xuzhou, Jiangsu 221116, China
  • Received:1900-01-01 Revised:1900-01-01 Online:2008-04-20 Published:2008-04-20
  • Contact: LING Ping

摘要: 设计了SVM置信度在线评估方案,以此确定SVM做多分类时结果的风险程度,对高风险决策结果进行修正。置信度评估采用理论估计和经验估计相结合的方式。多分类决策结果的修正由在线生成的局部分类器完成。局部分类器在待查询数据的邻域内工作,此邻域基于一个局部测度而生成。实验表明,所设计的算法呈现了较好的分类能力,提高了传统同类方法的分类准确率。

关键词: SVM, 置信度评估, 决策风险值, 局部分类器, 局部测度

Abstract: An algorithm of confidence evaluation for SVM is presented. Based on the evaluation, decision risk is specified in the context of multi-classification. Evaluation approach combines the theoretic analysis and empirical analysis. The decision with low confidence is refined by a local classifier that is formulated online. The local classifier works in query’s neighborhood,and the neighborhood is developed according to a local metric. Experiments demonstrate the fine performance of the designed algorithm, and its improvement in classification accuracy over the state of the arts.

Key words: SVM, confidence evaluation, decision risk amount, local classifier, local metric