Perturbation Models and PAC-Bayesian Generalization Bounds

Joseph Keshet, Subhransu Maji, Tamir Hazan, Tommi Jaakkola

Research output: Chapter in Book/Report/Conference proceedingChapterpeer-review

Abstract

In this chapter we explore the generalization power of perturbation models. Learning parameters that minimize the expected task loss of perturbation models amounts to minimizing PAC-Bayesian generalization bounds. We provide an elementary derivation of PAC-Bayesian generalization bounds, while focusing on their Bayesian components, namely their predictive probabilities and their posterior distributions.
Original languageEnglish
Title of host publicationPerturbations, Optimization, and Statistics
EditorsT Hazan, G Papandreou, D Tarlow
Pages289-309
Number of pages21
DOIs
StatePublished - 1 Dec 2016

Cite this