Forecasting Stock Price Based on Back Propagation Neural Network

Abstract

The stock market is a nonlinear dynamics system with enormous information, which is difficult to predict effectively by traditional methods. The model of stock price forecast based on BP Neutral-Network is put forward in this article. The paper try to find the way how to predictive the stock price. Exhaustive method is used for the hidden layer neurons and training method determination. Finally the experiment results show that the algorithm get better performance in stock price prediction.

Stock Price Forecasting: Traditional Statistical Methods and Deep Learning Methods

“…There are many studies about stock price forecasting: Li uses A New dimensional gray Markov forecasting model to improve the accuracy of mid to long term predictions, Jiang uses RBF neural network to improve the adaptive ability of prediction [1][2]. Zhang uses genetic algorithm to train and optimize the initial weights of Elman neural network, an efficient GA-ELman dynamic regression neural network stock price prediction model is proposed [3]. Zhang and Liang based on the dimensionality reduction of many variables affecting target data by principal component analysis method, the BP neural network prediction model of stock market closing price is constructed [4].…”