Parameter setting of technical analysis indicators using multi-objective particle swarm optimization and adaptive fuzzy inference system

In this paper, we propose automatic stock trading system which combines technical analysis and adaptive neural fuzzy inference system to predict the stock price trend to increase return of investment. In this trading system, at first the optimal value of technical indicator's parameters is determined by using multi-objective particle swarm optimization and according to these parameters; technical indicators are calculated to predict stock price changes with the help of adaptive neural fuzzy inference system. We have chosen eight different stocks from Tehran stock exchange to test our trading system for two months. A computational experience is carried out in order to analyze the proposed algorithm and the obtained results are compared with usual conventional methods which have been proposed in previous researches. The computational results show our proposed method performs better than other previous methods and obtains superior results.