Buying and Selling Decision in the Brazilian Stock Exchange Financial Market by a Neo Fuzzy Neuron (NFN) Applied to the Hurwicz Criterion

Abstract

This work introduces an approach for making decisions on buying and selling stocks in the Brazilian Stock Exchange to maximize profits in each operation. The proposed approach was built using the Neo-Fuzzy-Neuron (NFN) network to predict the future value of stocks and the Hurwicz criterion for decision analysis under risk and uncertainty, considering different degrees of optimism and pessimism. The approach was applied to Petrobras stocks (PETR4), and the results obtained were compared with the ”buy and hold”strategy. The computational results and comparisons suggest that the proposed approach is promising and provides a significant return on investment