支持向量回归参数的混合选择

Hybrid Selection of Parameters for Support Vector Regression

为提高支持向量回归算法的学习能力和泛化性能,提出了一种优化支持向量回归参数的混合选择算法。根据训练样本的规模和噪声水平等信息,确定支持向量回归参数的取值范围,用实数编码的免疫遗传算法搜索最佳参数值。混合选择算法具有较高的精度和效率,在选择支持向量回归参数时,不必考虑模型的复杂度和变量维数。仿真实验结果表明,该算法是选择支持向量回归参数的有效方法,应用到函数逼近问题时具有优良的性能。

In order to improve support vector regression（SVR） learning ability and generalization performance, a hybrid selection algorithm for optimizing SVR parameters is proposed. The ranges of the parameters are set according to the information about the training data size and noise level in training samples, and a real-coding based immune genetic algorithm is employed to search the optimal parameters. Hybrid selection algorithm is a precise and efficient method and it need not to consider SVR dimensionality and complexity. Simulation experiments show that the proposed method is an effective approach for SVR parameters selection with good performance on function approximation problem.