Bayesian Regularization BP Neural Network Model for the Stock Price Prediction

Abstract

It is an important research issue to improve the generalization ability of the neural network in the research of artificial neural network. This paper proposes the Bayesian regularization method to optimize the training process of the back propagation (BP) neural network, so that the optimized BP neural network model can predict new data in the BP neural network to a larger extent. Based on the experiments in which Bayesian regularization BP neural network is employed to predict the stock price series, and through the establishment of the stock customer transaction model network structure, an experimental program is selected to make an empirical analysis of the closing price data of Shanghai Stock in 800 trading days, the results of which show that the Bayesian regularization method has a better generalization ability.