Comparative study between the FLANN model and the MLP model in the stock market forecast: case of S & P 500

Abstract

This article highlights a comparative study between the MLP model with the backpropagation algorithm and the FLANN model using the trigonometric expansion function in different experiments to form the model weights for a short forecast (one day, one week) and long-term (one month, two months) closing price of the S & P 500 Index in the US market. In order to know the model that advocates the best forecast, we used appropriate combinations of technical and economic parameters as inputs. Measurement criteria such as: MSE, MAE, MAPE, DA% are used as performance indices to evaluate the quality of model prediction. The simulation and test results showed that the MLP model offers a better prediction than the FLANN model.