中国物理B ›› 2004, Vol. 13 ›› Issue (4): 454-458.doi: 10.1088/1009-1963/13/4/007

• • 上一篇    下一篇

Chaotic time series prediction using least squares support vector machines

汪晓东1, 叶美盈2   

  1. (1)College of Information Science and Engineering, Zhejiang Normal University, Jinhua 321004, China; (2)College of Mathematics and Physics, Zhejiang Normal University, Jinhua 321004, China
  • 收稿日期:2003-07-10 修回日期:2003-08-01 出版日期:2004-04-22 发布日期:2004-04-20
  • 基金资助:
    Project supported by the Zhejiang Provincial Natural Science Foundation, China (Grant No 602145).

Chaotic time series prediction using least squares support vector machines

Ye Mei-Ying (叶美盈)a, Wang Xiao-Dong (汪晓东)b    

  1. a College of Mathematics and Physics, Zhejiang Normal University, Jinhua 321004, China; b College of Information Science and Engineering, Zhejiang Normal University, Jinhua 321004, China
  • Received:2003-07-10 Revised:2003-08-01 Online:2004-04-22 Published:2004-04-20
  • Supported by:
    Project supported by the Zhejiang Provincial Natural Science Foundation, China (Grant No 602145).

PDF (PC)

849

赞

摘要： We propose a new technique of using the least squares support vector machines (LS-SVMs) for making one-step and multi-step prediction of chaotic time series. The LS-SVM achieves higher generalization performance than traditional neural networks and provides an accurate chaotic time series prediction. Unlike neural networks' training that requires nonlinear optimization with the danger of getting stuck into local minima, training LS-SVM is equivalent to solving a set of linear equations. Thus it has fast convergence. The simulation results show that LS-SVM has much better potential in the field of chaotic time series prediction.

关键词: chaotic time series, time series prediction, support vector machines

Abstract: We propose a new technique of using the least squares support vector machines (LS-SVMs) for making one-step and multi-step prediction of chaotic time series. The LS-SVM achieves higher generalization performance than traditional neural networks and provides an accurate chaotic time series prediction. Unlike neural networks' training that requires nonlinear optimization with the danger of getting stuck into local minima, training LS-SVM is equivalent to solving a set of linear equations. Thus it has fast convergence. The simulation results show that LS-SVM has much better potential in the field of chaotic time series prediction.

Key words: chaotic time series, time series prediction, support vector machines

中图分类号:  (Time series analysis)

  • 05.45.Tp