Hybrid Neural Network-Genetic Algorithm Method to Predict Monthly Minimum and Maximum of Stock Prices

摘要

Prediction of stock price variations is of high importance to investors. A critical factor in investment is to forecast the right time to buy or sell stocks. This paper presents a novel method that combines artificial neural networks and genetic algorithms to predict the monthly minimum and maximum of stock prices. We used a combination of genetic algorithms with backpropagation neural networks to overcome the limitations of the algorithms and benefit from their advantages. We have used real-life data to test our method. Numerical results show that our proposed hybrid method provides a more accurate prediction of stock prices as compared with the backpropagation method.