PAC-Bayesian learning of linear classifiers

We present a general PAC-Bayes theorem from which all known PAC-Bayes risk bounds are obtained as particular cases. We also propose different learning algorithms for finding linear classifiers that minimize these bounds. These learning algorithms are generally competitive with both AdaBoost and the SVM.

• Type: Conference
• Year: 2009
• Where: ICML
• Authors: Alexandre Lacasse, François Laviolette, Mario Marchand, Pascal Germain
• Keywords: ICML 2009, Machine Learning, PAC-Bayes Risk Bounds