最小二乘支持向量机的参数优选方法及其应用

参数选择是支持向量机研究领域的重要问题,它本质上是一个优化搜索过程.以遗传算法和粒子群算法为基础探讨了基于两者的混合智能算法,将杂交操作、变异操作引入PSO算法中,同时,在种群随机搜索过程中嵌入确定性的模式搜索,使得算法可以在任何阶段进行精细搜索;在此基础上,提出了基于混合智能的最小二乘支持向量机方法(LS-SVM),以最小化k-fold交叉验证误差为评价函数,利用混合智能算法优化LS-SVM参数.最后结合实例对该方法进行了实证检验,并对结果进行分析.

Method for selecting parameters of least squares support vector machines and its application

It is an important issue for support vector machines（ SVMs） to choose parameters in engineering applications and theoretic research. By combining the evolution idea of genetic algorithm with the swarm intelligence of particle swarm optimization algorithm,a hybrid intelligent algorithm is developed and applied to water demand forecasting in the paper. It makes use of PSO algorithm characteristics such as parallel property and the global convergence performance to avoid the local optimum,and uses the evolution idea of genetic algorithm such as crossover and mutation operations to improve the speed of searching for the global optimization. On the other hand,a deterministic searching algorithm is embedded to improve its optimization performance,based on which the LS-SVM prediction model of water consumption is proposed,tuning LS-SVM parameters by hybrid intelligent algorithm. The application shows that the presented LS-SVM optimized by hybrid intelligent algorithm can offer more accurate forecasting result than SVM method,and has high generalization ability.