The selection of the kernel function parameter and error penalty factor affected the precision of the support vector machine (SVM) significantly.
核函数参数和误差惩罚因子的选择对支持向量机模型（SVM）的精度有较大影响。
In the paper, GBGM-GA is seen the optimization technique combining KPCA and GA, and is suitable to the optimization selection of kernel function parameter.
本文采用高斯变异遗传算法作优化技术，实现了KPCA和GA的集成，适合核函数参数的优化选择。
According to the statistic characteristic of covariance function, parameter estimation can be given for kernel function.
由协方差函数的统计特征，可给出核函数的参数估计。
应用推荐