主管:中国科学院
主办:中国优选法统筹法与经济数学研究会
   中国科学院科技战略咨询研究院

中国管理科学 ›› 2010, Vol. 18 ›› Issue (3): 83-89.

• 论文 • 上一篇    下一篇

基于结构可变的RBF神经网络的时间序列预测

张冬青1, 马宏伟1, 宁宣熙2   

  1. 1. 南京农业大学工学院, 江苏南京210031;
    2. 南京航空航天大学经济与管理学院, 江苏南京210016
  • 收稿日期:2009-07-06 修回日期:2010-05-30 出版日期:2010-06-30 发布日期:2010-06-30
  • 作者简介:张冬青(1971- ),女(汉族),江苏泗洪人,南京农业大学,讲师,博士,研究方向:时间序列预测、经济信号处理.
  • 基金资助:

    国家自然科学基金资助项目(70571037);江苏省农机基金资助项目(gxz09003)

Time Series Prediction Based on Variable Structure RBF Neural Networks

ZHANG Dong-qing1, MA Hong-wei1, NING Xuan-xi2   

  1. 1. College of Engineering, Nanjing Agricultural University, Nanjing 210031, China;
    2. College of Economics and Management, Nanjing University of Aeronautics and Astronautics, Nanjing 210016, China
  • Received:2009-07-06 Revised:2010-05-30 Online:2010-06-30 Published:2010-06-30

摘要: 基于神经网络的时间序列预测方法,需要正确确定网络结构,它关系到所建模型的合理性以及预测的准确性。目前确定网络结构的绝大多数方法,其网络结构一经训练确定便保持不变。然而现实中许多时间序列呈现非平稳性,其结构经常发生变化,这就要求网络结构能够动态可调,因此本文提出结构可变的径向基函数(RBF)神经网络预测模型。并采用序列蒙特卡罗(SMC)方法实现基于结构可变RBF网络的时间序列在线预测;最后采用CRU钢铁价格指数月数据进行实证研究,结果表明该模型的有效性。

关键词: 预测, 时间序列, 径向基函数神经网络, 序列蒙特卡罗方法

Abstract: Suitable architecture is critical for a neural network to model time series and it determines forecast performance.So far,most methods assume that the architecture will keep fixed once it has been trained.However,many time series are nonlinear and their structures often change,which requires that the architecture of network should vary with time.Therefore,a variable structure radial basis function (RBF)neural network is proposed in this paper.Furthermore,sequential Monte Carlo(SMC)method is applied for time series on line prediction in the variable structure RBF network model.At last,the data of CRU global steel price index are analyzed,and experimental results indicate that the variable structure RBFnet work proposed is effective.

Key words: prediction, time series, radial basis function neural networks, sequential Monte Carlo method

中图分类号: