J4

• 计算机科学 • 上一篇    下一篇

基于冲突指导的神经网络预测方法

王 喆, 王 斌, 周春光, 徐振龙   

  1. 吉林大学计算机科学与技术学院, 长春 130012
  • 收稿日期:2003-10-08 修回日期:1900-01-01 出版日期:2004-01-26 发布日期:2004-01-26
  • 通讯作者: 周春光

Neural networks prediction method based on violation guide

WANG Zhe, WANG Bin, ZHOU Chun-guang, XU Zhen-long   

  1. College of Computer Science and Technology, Jilin University, Changchun 130012, China
  • Received:2003-10-08 Revised:1900-01-01 Online:2004-01-26 Published:2004-01-26
  • Contact: ZHOU Chun-guang

摘要: 研究基于冲突指导的神经网络预测技术. 利用离散约 束最优的拉格朗日乘数理论, 通过采用前馈方法得到近似梯度的模拟退火技术, 避免了盲 目接受试验点. 利用松紧策略加快了搜索的收敛速度. 实验结果表明, 训练误差和预测 误差都有很大改善.

关键词: 冲突指导, 预测, 神经网络

Abstract: The neural network prediction technique based on the im pulse guidance is discussed. On the basis of the theory of discrete constrained optimal Lagrange multiplicator, we avoided to accept test points sightlessly wit h the simulated anneal technique from the approximate grades through the forward propagation method, and accelerated the search speed using Relaxed & Tighten st rategy. The experiment result shows that the training error and the prediction e rror have been improved greatly. Keywords: 收稿日期:

Key words: violation guide, prediction, neural networks

中图分类号: 

  • TP183