A Fast Incremental Learning Algorithm for SVM

摘要：

A fast incremental learning algorithm for SVM based on K nearest neighbors (KNN-ISVM) is presented. The algorithm extracts border vector set by applying the idea of K nearest neighbors and trains SVM by substituting the border vectors set for training set. The method reduces training samples largely and therefore advances training speed greatly, while the ability of SVM to classify is also guaranteed because the border vector set may contain all useful training samples. The experiment result proves the effectiveness of KNN-ISVM.

关键词： Support vector machines Incremental learning Border vector K nearest neighbors

作者: Fasheng Sun Huaitie Xiao

作者单位: School of Electronic Science and Engineering,National University of Defense and Technology,Changsha, Hunan, 410073, P.R.China

会议类型: 国际会议

会议名称: The International Colloquium on Onformation Fusion 2007(2007年国际信息融合研讨会)

会议地点: 西安

会议语种:英文

页码: 448-451

在线出版日期: 2007-08-22（万方平台首次上网日期，不代表论文的发表时间）