When the Gaussian kernel function width is 2.0, RVM method possesses more perfect prediction performance.

• 当高斯核函数的宽度值取为2.；0时，相关向量机方法具有较为理想的预测效果。