In this pap er, we show that training of the support vector machine (SVM) can b e interpreted as performing the level 1 inference of MacKay's evidence framework. We further on show that levels 2 and 3 can also b e applied to SVM. This allows automatic adjustment of the regularization parameter and the kernel parameter. More importantly, it op ens up a wealth of Bayesian tools for use with SVM. Performance is evaluated on both synthetic and real-world data sets.
|Title of host publication||ESANN'1999 proceedings - European Symposium on Artificial Neural Networks Bruges (Belgium), 21-23 April 1999, D-Facto public.|
|Number of pages||6|
|Publication status||Published - 21 Apr 1999|