Talk:Support vector clustering
The article is very stimulating and deserves publication here at Scholarpedia. The following are minor issues that can be made to improve it. 1. I believe it makes sense to explain a little bit to the readers how to obtain Figure 1 through "SVDD". 2. This idea of "kernel" is well-known to the "SVM" community but not to general, and details should be introduced or linking out. 3. It would be great to have a statement focusing on the kernel function. For example, beside Gaussian kernel, how to make SV clustering by Linear, Polynomial, or Sigmoid kernels?