基于改进BP神经网络对非线性时间序列的建模预测

Prediction of modeling of nonlinear time sSeries based on Improved BP neural network

分析了BP神经网络的基本特性,对BP神经网络用于非线性时间序列(股价)预测的算法进行了研究,并对其进行了改进。利用MATLAB软件对其进行仿真,结果表明,利用改进的BP神经网络对股价进行短期预测是可行的,改进方法有效。

This paper is to analyze the basic characteristics of BP neural network,discuss the prediction algorithm of the BP neural network used in nonlinear time,and try to improve this model.Using Matlab software for emulation,and the simulation results show that,the improved BP neural network to stock price for short-term forecasting is feasible,and the improvement method is effective.