Scaling Gaussian RBF kernel width to improve SVM classification

  title={Scaling Gaussian RBF kernel width to improve SVM classification},
  author={Qun Chang and Qingcai Chen and Xiaolong Wang},
  journal={2005 International Conference on Neural Networks and Brain},
Support vector classification with Gaussian RBF kernel is sensitive to the kernel width. Small kernel width may cause over-fitting, and large one under-fitting. The so-called optimal kernel width is merely selected based on the tradeoff between under-fitting loss and over-fitting loss. So, there exists urgent need to further reduce the tradeoff loss. To circumvent this, we scale the kernel width in a distribution-dependent way. Experiments validate the feasibility of this method. Existing… CONTINUE READING
Highly Cited
This paper has 29 citations. REVIEW CITATIONS