SVM中径向基函数与高斯核的区别 Difference between RBF and Gaussian kernel in SVM

Radial Basis Functions (RBFs) are set of functions which have same value at a fixed distance from a given central point. Even Gaussian Kernels with a covariance matrix which is diagonal and with constant variance will be radial in nature.

In SVMs, RBF Kernal and Gaussian Kernal are used interchangeably. But right way to specify is “Gaussian Radial Basis Function” because there can be other RBFs. Gaussian RBFs are one of the most used Kernal in SVMs. It can take data to infinite dimensional space and has infinite VC dimensions. One has to be careful to avoid overfitting when using Gaussian RBF Kernels. Read this paper for more information: http://www.cmap.polytechnique.fr...

 

向基函数（RBF）是一组函数，它们在距给定中心点的固定距离处具有相同的值。甚至具有协方差矩阵的高斯核也是径向的，该协方差矩阵是对角的并且具有恒定的方差。

在SVM中，RBF Kernal和Gaussian Kernal可互换使用。但正确的指定方式是“高斯径向基函数”，因为可以有其他RBF。高斯RBF是SVM中最常用的Kernal之一。它可以将数据带入无限维空间并具有无限的VC维度。在使用高斯RBF内核时，必须小心避免过度拟合。更多信息：http：//www.cmap.polytechnique.fr ...

 

 

来源： https://www.quora.com/What-differentiates-a-radial-basis-function-from-a-gaussian-kernel-while-using-SVM