摘要
将Chebyshev神经网络作为非线性时间序列的辨识模型 ,通过对过去序列样本的学习 ,调整网络的权值 ,然后预测和推断未来的序列 ,仿真结果表明 ,Chebyshev神经网络具有优良的泛化能力和预测功能。
In this paper, we predicted and deduced the future series by training the past its and adjusting the weight of networks using Chebyshev neural networks as the identification model on the nonlinear time series. The simulation results show that the Chebyshev neural networks own good generalization and predictive ability.
出处
《计算机应用》
CSCD
北大核心
2001年第4期14-15,共2页
journal of Computer Applications
基金
国家自然科学基金资助项目! (1 99740 0 2 )
