GDP prediction by support vector machine trained with genetic algorithm

In the study, support vector machine trained with genetic algorithm is applied in GDP forecasting. Genetic algorithm can get optimal solution in short time, which is an excellent method in parameters selection of support vector machine. Then, genetic algorithm is introduced to simultaneously optimize the SVM parameters. The total GDP data of Anhui province from 1989 to 2007 are employed to compare the forecasting performance of the proposed GA-SVM model and RBF neural network GDP forecasting model. It is indicated that GDP prediction performance of the proposed GA-SVM is better than that of RBFNN.