Prediction of financial time series based on information granulation

Abstract

The prediction of financial time series produced by the stock market has been the research focus of scholars. However, we can not accurately predict the fluctuations in the future in most cases. In this paper, we try to apply information granulation in the prediction of financial time series. First, we fuzzy grain the financial time series. Then, we use RBF neural network which has good ability of nonlinear mapping to predict future trend of fluctuations and fluctuating range. Simulation result shows that this method can accurately predict the trend of fluctuations and the fluctuating range, which can provide reliable reference for the decision makers of investment.