计算机工程与应用 ›› 2008, Vol. 44 ›› Issue (10): 142-143.

• 数据库、信号与信息处理 • 上一篇    下一篇

一种改进的支持向量机增量学习算法

刘叶青1,2,刘三阳1,谷明涛3   

  1. 1.西安电子科技大学 数学科学系,西安 710071
    2.河南科技大学 理学院,河南 洛阳 471003
    3.解放军96251部队,河南 洛阳 471003
  • 收稿日期:2007-07-23 修回日期:2007-10-16 出版日期:2008-04-01 发布日期:2008-04-01
  • 通讯作者: 刘叶青

Improved incremental learning algorithm for support vector machine

LIU Ye-qing1,2,LIU San-yang1,GU Ming-tao3   

  1. 1.Department of Mathematical Sciences,Xidian University,Xi’an 710071,China
    2.School of Science,Henan University of Science & Technology,Luoyang,Henan 471003,China
    3.PLA Unit96251,Luoyang,Henan 471003,China
  • Received:2007-07-23 Revised:2007-10-16 Online:2008-04-01 Published:2008-04-01
  • Contact: LIU Ye-qing

摘要: 提出了一种改进的支持向量机增量学习算法。分析了新样本加入后,原样本和新样本中哪些样本可能转化为新支持向量。基于分析结论提出了一种改进的学习算法。该算法舍弃了对最终分类无用的样本,并保留了有用的样本。对标准数据集的实验结果表明,该算法在保证分类准确度的同时大大减少了训练时间。

关键词: 支持向量机, 增量学习, 分类

Abstract: An improved incremental learning algorithm for Support Vector Machine(SVM) is presented.The possible changes of support vector set after new samples are added to training set are analyzed.Based on the analysis results,an improved algorithm is presented.In the algorithm,the useless samples are discarded and the useful samples are retained.The experimental results with the standard dataset show that the training time is greatly reduced while the classification precision is guaranteed.

Key words: support vector machine, incremental learning, classification