摘要：Size and volatility of the stock price of companies listed on the stock exchange is one of the most important factors influencing the financial markets. Predicting the influencing factors such as price and stock returns in the stock market is essential to the large number of users who are interested in entering the stock market as investors or as attract-business investors. This is because this investment might be a big risk without forecasting tools for estimating possible future events. Several years ago, regression methods were used for predicting market, but neural networks and genetic algorithms have ability to learn complex nonlinear series with better performance. In this paper, from the effective elements in the stock market, stock returns based on their importance to investors are selected for prediction. The research results show that when the samples are optimized with genetic algorithm or neural network to predict stock returns in comparison to regression methods, smaller error was achieved.
关键词：Predict; stock returns; data mining; neural networks; genetic algorithms
