Benjamin Guedj (MODAL, Inria, Université de Lille)

A quasi-Bayesian perspective to NMF: theory and applications

vendredi 10 mars 2017, 9h30 - 10h30

Salle du conseil, espace Turing

Quasi-Bayesian estimators are increasingly popular in statistics and machine learning, due to their generalization properties and flexibility. In a recent work (Alquier & Guedj 2017, Mathematical Methods of Statistics), we have proposed a quasi-Bayesian estimator for non-negative matrix factorization. I will present a quick overview of quasi- and PAC-Bayesian frameworks and discuss our theoretical and algorithmic contributions.