The Application of SVM and GA-BP Algorithms in Stock Market Prediction

摘要

Neural network has been popular in time series prediction in financial areas, because of their advantages in handling nonlinear systems. This paper hybridizes genetic algorithm and artificial neural network method (GA-BP), and hybridizes principal component analysis and support vector machine (PCA-SVM) to predict the next opening price in stock markets. Principal component analysis method is applied to extract contribution rate to meet 95% of the principal component as the input variables with FAW Car and Minmetals Rare Earth to be modeled and predicted, and genetic algorithm is employed to determine the initial weight and threshold of the BP neural network. The experiment results demonstrate that the combination methods (PCA-SVM and GA-BP) perform better, and the GA-BP method can get higher prediction accuracy than other three prediction methods.