English

应用科学学报 ›› 2009, Vol. 27 ›› Issue (4): 409-413.

• 计算机科学与应用 • 上一篇    下一篇

泛函网络在线增量式学习算法及应用

周永权 罗淇方 吕咏梅 赵斌   

  1. 1. 广西民族大学数学与计算机科学学院，南宁530006
    2. 中央民族大学理学院，北京100081
  • 收稿日期:2007-12-11 修回日期:2009-05-26 出版日期:2009-07-30 发布日期:2009-07-30
  • 作者简介:周永权，博士，教授，研究方向：神经网络、计算智能及应用，E-mail：yongquanzhou@126.com
  • 基金资助:
     国家自然科学基金(No.60461001)；广西省自然科学基金(No.0832082, No.0991086)；国家民委科研项目基金(No.08GX01)资助

Online Incremental Learning Algorithm and Application to Functional Networks

  1. 1. College of Computer and Information Science, Guangxi University for Nationalities, Nanning 530006, China
    2. School of Science, Central University for Nationalities, Beijing 100081, China
  • Received:2007-12-11 Revised:2009-05-26 Online:2009-07-30 Published:2009-07-30

PDF

1702

被引次数

4

摘要：

 给出了一类泛函网络的数学模型，并分析了它的拓扑结构特点和离线学习过程. 在此基础上根据分块矩阵计算方法和泛函网络基函数矩阵本身的特点，给出了泛函网络的两种在线增量式学习算法. 该算法能充分利用历史训练结果，具有学习、修正和应变功能. 最后，以Hénon时间序列为例进行仿真. 仿真结果表明这两种学习算法是可行和有效的.

关键词: 基函数簇, 泛函参数, Lagrange乘数法, 在线增量式学习, H?enon时间序列

Abstract:

A mathematical model of functional networks is proposed. The property of its topology structure and learning process is analyzed. Online incremental learning algorithms based on the block matrix and the property of functions matrix are designed. The learning algorithms make fully use of the training history, and have functions of learning, modification and emergency adaptation. Simulation on a Hénon time series shows effectiveness of the proposed algorithms.

 

Key words: basis functions, functional parameter, Lagrange multiplier method, online incremental learning, Hénon time series

中图分类号: