An innovative Approach to Predict Stock Prices at Stock Exchange Markets by Using Multi-Linear Heavy Tails Regression Model Based on Genetic Algorithm

This paper proposes a new method based on Multi-Linear Generalized Modified Bessel Regression as a successful alternative to the Multi-Linear Normal Regression and Artificial Neural Network Model to predict stock prices at stock exchange markets. The generalized modified Bessel linear regression model has five parameters: Position parameter, Measurement, and Shape Parameters. In this paper, we used a genetic algorithm to estimate these parameters and apply the proposed model to several real data of stock prices in the Iraqi exchange market. The results showed the superiority of the proposed generalized Bessel regression model on multi-linear normal regression and artificial neural network model depending on the (RMSE), (MAD) and (R
²
) criteria. The proposed model is thus capable of predicting typical stock markets.