关键词: ε-支持向量机, 参数优化, 遗传算法, 预测

Abstract: The ε-Support Vector regression Machine is a promising artificial intelligence technique，in which the regression algorithm has already been used in solving the nonlinear function approach successfully.Unfortunately，most user select parameters for an SVM by rule of thumb，so they frequently fail to generate the optimal parameters effect for the function.This has restricted effective use of SVM to a great degree.In this paper，the authors use genetic algorithm to solve the SVM parameters optimization problem.Simulation result show that the method has high precision，the method possesses certain practical application significance.

Key words: ε-SVM, parameter optimization, GA, prediction