SVM regression through variational methods and its sequential implementation

We consider an SVM regression model based on kernel methods with a Gaussian prior distribution over the network parameters. We show that the variational techniques can be utilised to obtain a closed form a posteriori distribution over the parameters given the data hence yielding an a posteriori pred...

Full description

Saved in:
Bibliographic Details
Published in:Neurocomputing (Amsterdam) 2003-09, Vol.55 (1), p.151-167
Main Authors: Gao, J.B., Gunn, S.R., Harris, C.J.
Format: Article
Language:English
Subjects:
Bayesian inference
Bayesian variational algorithm
EM algorithm
Regression
Support vector machine
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:We consider an SVM regression model based on kernel methods with a Gaussian prior distribution over the network parameters. We show that the variational techniques can be utilised to obtain a closed form a posteriori distribution over the parameters given the data hence yielding an a posteriori predictive model.
ISSN:0925-2312
1872-8286
DOI:10.1016/S0925-2312(03)00365-5