中国物理B ›› 2000, Vol. 9 ›› Issue (6): 408-413.doi: 10.1088/1009-1963/9/6/002

• GENERAL • 上一篇    下一篇

FAST EVOLVING MULTI-LAYER PERCEPTRONS FOR NOISY CHAOTIC TIME SERIES MODELING AND PREDICTIONS

张家树, 肖先赐   

  1. Department of Electronic Engineering, University of Electronic Science and Technology of China, Chengdu 610054, China
  • 收稿日期:2000-01-17 修回日期:2000-03-05 出版日期:2005-06-12 发布日期:2005-06-12
  • 基金资助:
    Project supported by the National Defense Foundation of China (Grant No. 98JS05.4.1DZ0205).

FAST EVOLVING MULTI-LAYER PERCEPTRONS FOR NOISY CHAOTIC TIME SERIES MODELING AND PREDICTIONS

Zhang Jia-shu (张家树), Xiao Xian-ci (肖先赐)   

  1. Department of Electronic Engineering, University of Electronic Science and Technology of China, Chengdu 610054, China
  • Received:2000-01-17 Revised:2000-03-05 Online:2005-06-12 Published:2005-06-12
  • Supported by:
    Project supported by the National Defense Foundation of China (Grant No. 98JS05.4.1DZ0205).

PDF (PC)

507

赞

摘要： A fast evolutionary programming (FEP) is proposed to train multi-layer perceptrons (MLP) for noisy chaotic time series modeling and predictions. This FEP, which uses a Cauchy mutation operator that results in a significantly faster convergence to the optimal solution, can help MLP to escape from local minima. A comparison against back-propagation-trained networks was performed. Numerical experimental results show that the FEP can help MLP better capturing dynamics from noisy chaotic time series than the back-propagation algorithm and produce a more consistently modeling and prediction.

Abstract: A fast evolutionary programming (FEP) is proposed to train multi-layer perceptrons (MLP) for noisy chaotic time series modeling and predictions. This FEP, which uses a Cauchy mutation operator that results in a significantly faster convergence to the optimal solution, can help MLP to escape from local minima. A comparison against back-propagation-trained networks was performed. Numerical experimental results show that the FEP can help MLP better capturing dynamics from noisy chaotic time series than the back-propagation algorithm and produce a more consistently modeling and prediction.

中图分类号:  (Noise)

  • 05.40.Ca
05.45.Tp (Time series analysis)