Computer Engineering and Applications ›› 2009, Vol. 45 ›› Issue (23): 15-17.DOI: 10.3778/j.issn.1002-8331.2009.23.005

• 博士论坛 • Previous Articles     Next Articles

Research on large scale SVM classification based on boundary K-nearest

FENG Guo-he   

  1. College of Economics and Management,South China Normal University,Guangzhou 510006,China
  • Received:2009-05-06 Revised:2009-06-07 Online:2009-08-11 Published:2009-08-11
  • Contact: FENG Guo-he

边界K邻近大样本支持向量机分类

奉国和   

  1. 华南师范大学 经济管理学院 信息管理系,广州 510006
  • 通讯作者: 奉国和

Abstract: The problem of occupying much memory and slow training speed will come forth for Support Vector Machine(SVM) with large scale training set.This paper puts forward a boundary K-NN SVM algorithm,searching for possible support vectors with K-NN and training SVM classifier based on such support vectors.Experiment shows that modified algorithm training speed is advanced.

Key words: upport Vector Machine(SVM), large-scale samples, classification

摘要: 针对大样本支持向量机内存开销大、训练速度慢的缺点,提出了一种改进的支持向量机算法。算法先利用KNN方法找出可能支持向量,然后利用SVM在可能支持向量集上训练得到分类器。实验表明改进算法训练速度提高明显。

关键词: 支持向量机, 大样本, 分类

CLC Number: