Application of BP neural network in stock market prediction

Abstract

Prediction for the change of stock market has been a hot research subject over the years. This thesis has introduced the definition and arithmetic of BP neural network model and established a stock market index prediction model based on the BP neural network model by taking advantage of the self-learning, self-adapting and nonlinear approximate ability. It is shown through empirical research that BP model not only has a rapid velocity of convergence and a high precision of prediction, but also has a certain application value if it is used for the short-term prediction of stock market index. © 2009 Springer Berlin Heidelberg.