Analysis of detectors for support vector machines and least square support vector machines

This paper discusses the performance capabilities of the support vector machine (SVM) and the least squares SVM (LS-SVM) for a two hypothesis detection problem. We consider a Bayesian framework where there are priors associated with each hypothesis and costs for making decisions. We examine how the SVM and the LS-SVM compare with the optimal Bayesian solution. We also discuss other merits for the SVM and the LS-SVM including practical implementation.