An adaptive online learning approach for Support Vector Regression: Online-SVR-FID
详细信息 查看全文
- 作者：Jie Liua ; jie.liu@centralesupelec.fr" class="auth_mail" title="E-mail the corresponding author ; Enrico Zioa ; b ; enrico.zio@centralesupelec.fr" class="auth_mail" title="E-mail the corresponding author
- 关键词：Online learning ; Support Vector Regression ; Time series data ; Pattern drift ; Feature vector selection ; Incremental and decremental learning
- 刊名：Mechanical Systems and Signal Processing
- 出版年：2016
- 出版时间：August 2016
- 年：2016
- 卷：76-77
- 期：Complete
- 页码：796-809
- 全文大小：1417 K
文摘
- •
An online learning strategy is proposed considering the characteristics of input vector and output for streaming data.
- •
Feature vector selection is used to reduce the training data points and the computational burden during online learning.
- •
Two proposed tolerance parameters control the computational complexity and the noise influence.
- •
Comparisons with other popular online learning methods for SVM show the efficiency of the proposed method.