计算机应用 ›› 2011, Vol. 31 ›› Issue (12): 3302-3304.

• 人工智能 • 上一篇    下一篇

基于粒子群优化模式搜索的支持向量机参数优化及应用

王喜宾1,张小平2,王翰虎1   

  1. 1. 贵州大学 计算机科学与信息学院,贵阳 550025
    2. 贵州省科学技术情报研究所,贵阳 550004
  • 收稿日期:2011-05-09 修回日期:2011-07-15 发布日期:2011-12-12 出版日期:2011-12-01
  • 通讯作者: 王喜宾
  • 基金资助:
    贵州省科技计划项目

Parameter optimization of support vector machine and application based on particle swarm optimization mode search

WANG Xi-bin1,ZHANG Xiao-ping2,WANG Han-hu1   

  1. 1. School of Computer Science and Information, Guizhou University, Guiyang Guizhou 550025, China
    2. Guizhou Institute of Scientific and Technical Information, Guiyang Guizhou 550004, China
  • Received:2011-05-09 Revised:2011-07-15 Online:2011-12-12 Published:2011-12-01
  • Contact: WANG Xi-bin

摘要: 针对核函数参数选择的重要性,提出了粒子群(PSO)模式搜索算法来搜索最优参数,该算法结合了PSO算法的全局搜索能力强和模式搜索的局部收敛性好的优点,使PSO模式搜索算法表现出了较高的性能,并将其应用到农业科技项目分类中。实验结果表明,该算法不仅效率高,收敛速度快,而且搜索到的最优参数达到了较高的准确率。

关键词: 支持向量机, 核参数选取, 粒子群模式搜索

Abstract: Considering the importance of selecting Kernel parameters, the Particle Swarm Optimization (PSO) model search algorithm was proposed to search optimal parameters. This method combined the global search capability of PSO algorithm and the good local convergence of mode search, that making PSO model search algorithm displays higher performance, and applied to an the practice of agricultural technological project classification. The results of experiment show that this method is not only efficient, but also catches the optimal parameters that have achieved higher accuracy.

Key words: Support Vector Machine (SVM), Kernel parameters selection, particle swarm model search

中图分类号: