AIC and BIC based approaches for SVM parameter value estimation with RBF kernels

Sergey Demyanov, James Bailey, Kotagiri Ramamohanarao, Christopher Leckie
Proceedings of the Asian Conference on Machine Learning, PMLR 25:97-112, 2012.

Abstract

We study the problem of selecting the best parameter values to use for a support vector machine (SVM) with RBF kernel. Our methods extend the well-known formulas for AIC and BIC, and we present two alternative approaches for calculating the necessary likelihood functions for these formulas. Our first approach is based on using the distances of support vectors from the separating hyperplane. Our second approach estimates the probability that the SVM hyperplane coincides with the Bayes classifier, by analysing the disposition of points in the kernel feature space. We experimentally compare our two approaches with several existing methods and show they are able to achieve good accuracy, whilst also having low running time.

Cite this Paper

BibTeX
@InProceedings{pmlr-v25-demyanov12, title = {AIC and BIC based approaches for SVM parameter value estimation with RBF kernels}, author = {Demyanov, Sergey and Bailey, James and Ramamohanarao, Kotagiri and Leckie, Christopher}, booktitle = {Proceedings of the Asian Conference on Machine Learning}, pages = {97--112}, year = {2012}, editor = {Hoi, Steven C. H. and Buntine, Wray}, volume = {25}, series = {Proceedings of Machine Learning Research}, address = {Singapore Management University, Singapore}, month = {04--06 Nov}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v25/demyanov12/demyanov12.pdf}, url = {https://proceedings.mlr.press/v25/demyanov12.html}, abstract = {We study the problem of selecting the best parameter values to use for a support vector machine (SVM) with RBF kernel. Our methods extend the well-known formulas for AIC and BIC, and we present two alternative approaches for calculating the necessary likelihood functions for these formulas. Our first approach is based on using the distances of support vectors from the separating hyperplane. Our second approach estimates the probability that the SVM hyperplane coincides with the Bayes classifier, by analysing the disposition of points in the kernel feature space. We experimentally compare our two approaches with several existing methods and show they are able to achieve good accuracy, whilst also having low running time.} }
Endnote
%0 Conference Paper %T AIC and BIC based approaches for SVM parameter value estimation with RBF kernels %A Sergey Demyanov %A James Bailey %A Kotagiri Ramamohanarao %A Christopher Leckie %B Proceedings of the Asian Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2012 %E Steven C. H. Hoi %E Wray Buntine %F pmlr-v25-demyanov12 %I PMLR %P 97--112 %U https://proceedings.mlr.press/v25/demyanov12.html %V 25 %X We study the problem of selecting the best parameter values to use for a support vector machine (SVM) with RBF kernel. Our methods extend the well-known formulas for AIC and BIC, and we present two alternative approaches for calculating the necessary likelihood functions for these formulas. Our first approach is based on using the distances of support vectors from the separating hyperplane. Our second approach estimates the probability that the SVM hyperplane coincides with the Bayes classifier, by analysing the disposition of points in the kernel feature space. We experimentally compare our two approaches with several existing methods and show they are able to achieve good accuracy, whilst also having low running time.
RIS
TY - CPAPER TI - AIC and BIC based approaches for SVM parameter value estimation with RBF kernels AU - Sergey Demyanov AU - James Bailey AU - Kotagiri Ramamohanarao AU - Christopher Leckie BT - Proceedings of the Asian Conference on Machine Learning DA - 2012/11/17 ED - Steven C. H. Hoi ED - Wray Buntine ID - pmlr-v25-demyanov12 PB - PMLR DP - Proceedings of Machine Learning Research VL - 25 SP - 97 EP - 112 L1 - http://proceedings.mlr.press/v25/demyanov12/demyanov12.pdf UR - https://proceedings.mlr.press/v25/demyanov12.html AB - We study the problem of selecting the best parameter values to use for a support vector machine (SVM) with RBF kernel. Our methods extend the well-known formulas for AIC and BIC, and we present two alternative approaches for calculating the necessary likelihood functions for these formulas. Our first approach is based on using the distances of support vectors from the separating hyperplane. Our second approach estimates the probability that the SVM hyperplane coincides with the Bayes classifier, by analysing the disposition of points in the kernel feature space. We experimentally compare our two approaches with several existing methods and show they are able to achieve good accuracy, whilst also having low running time. ER -
APA
Demyanov, S., Bailey, J., Ramamohanarao, K. & Leckie, C.. (2012). AIC and BIC based approaches for SVM parameter value estimation with RBF kernels. Proceedings of the Asian Conference on Machine Learning, in Proceedings of Machine Learning Research 25:97-112 Available from https://proceedings.mlr.press/v25/demyanov12.html.

Related Material