Bayesian Support Vector Regression

Martin H. C. Law, James Tin-Yau Kwok
Proceedings of the Eighth International Workshop on Artificial Intelligence and Statistics, PMLR R3:162-167, 2001.

Abstract

We show that the Bayesian evidence framework can be applied to both $\epsilon$-support vector regression ($\epsilon$-SVR) and $\nu$-support vector regression ($\nu$-SVR) algorithms. Standard SVR training can be regarded as performing level one inference of the evidence framework, while levels two and three allow automatic adjustments of the regularization and kernel parameters respectively, without the need of a validation set.

Cite this Paper


BibTeX
@InProceedings{pmlr-vR3-law01a, title = {Bayesian Support Vector Regression}, author = {Law, Martin H. C. and Kwok, James Tin-Yau}, booktitle = {Proceedings of the Eighth International Workshop on Artificial Intelligence and Statistics}, pages = {162--167}, year = {2001}, editor = {Richardson, Thomas S. and Jaakkola, Tommi S.}, volume = {R3}, series = {Proceedings of Machine Learning Research}, month = {04--07 Jan}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/r3/law01a/law01a.pdf}, url = {http://proceedings.mlr.press/r3/law01a.html}, abstract = {We show that the Bayesian evidence framework can be applied to both $\epsilon$-support vector regression ($\epsilon$-SVR) and $\nu$-support vector regression ($\nu$-SVR) algorithms. Standard SVR training can be regarded as performing level one inference of the evidence framework, while levels two and three allow automatic adjustments of the regularization and kernel parameters respectively, without the need of a validation set.}, note = {Reissued by PMLR on 31 March 2021.} }
Endnote
%0 Conference Paper %T Bayesian Support Vector Regression %A Martin H. C. Law %A James Tin-Yau Kwok %B Proceedings of the Eighth International Workshop on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2001 %E Thomas S. Richardson %E Tommi S. Jaakkola %F pmlr-vR3-law01a %I PMLR %P 162--167 %U http://proceedings.mlr.press/r3/law01a.html %V R3 %X We show that the Bayesian evidence framework can be applied to both $\epsilon$-support vector regression ($\epsilon$-SVR) and $\nu$-support vector regression ($\nu$-SVR) algorithms. Standard SVR training can be regarded as performing level one inference of the evidence framework, while levels two and three allow automatic adjustments of the regularization and kernel parameters respectively, without the need of a validation set. %Z Reissued by PMLR on 31 March 2021.
APA
Law, M.H.C. & Kwok, J.T.. (2001). Bayesian Support Vector Regression. Proceedings of the Eighth International Workshop on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research R3:162-167 Available from http://proceedings.mlr.press/r3/law01a.html. Reissued by PMLR on 31 March 2021.

Related Material