So they have better performance than adaptive Gaussian kernel.

• 因而，相对于自适应高斯核函数而言，两者性能更优。