Dynamic Normalization BPN for Stock Price Forecasting

Abstract

Stock price predicting is an important concern for investors, who by using high accuracy prediction systems are able to make a great profit. In recent years, artificial neural networks (ANNs) have shown promising results in this area, and have been improved in many ways. However, there are still some issues with ANN that remain unanswered, one of which is how to set the best parameters for ANN. Different combinations of setting parameters bring about different consequents, such as the constitution of input nodes, hidden nodes, and initial values of weight. Hence, we propose a simple but useful method, which only uses stock closing prices as inputs and experiments with different kinds of setting parameters. In addition, this paper enhances back propagation neural network (BPN) with a novel normalized function. System is applied to Taiwans Top 50 Exchange Traded Fund, S&P 500, and Shenzhen Composite to forecast the next day closing price. Given the experimental results, the proposed method shows excellent performance with the best set of parameters, and the innovative normalization method effectively improves accuracy. Moreover, our system provides better results in terms of accuracy of prediction than other systems. Finally, this study provides a method for how to design setting parameters in BPN.