Empirical Bayes methods and nonparametric mixture models via nonparametric maximum likelihood

Jeudi 1er décembre 2016 16:00-17:00 - Lee Dicker - Rutgers University

Résumé : Empirical Bayes methods have a long and rich history in statistics, and are particularly well-suited for many problems involving heterogeneous and high-dimensional data. Nonparametric maximum likelihood (NPML) is one elegant approach to empirical Bayes that has been studied since the 1950s and is closely related to the analysis of nonparametric mixture models. However, implementation and analysis of NPML-based methods for empirical Bayes is notoriously difficult. Recent computational breakthroughs have greatly simplified the implementation of NPML-based methods. Leveraging these recent advances, we have developed a variety of promising and flexible new methods involving NPML for empirical Bayes. In this talk, we will discuss these methods, along with applications in a variety of problems, including classification, online prediction, and multivariate density estimation. This talk is based on joint work with Sihai Dave Zhao (UIUC), Long Feng (Rutgers), and Ruijun Ma (Rutgers).

Lieu : Salle 117-119

Notes de dernières minutes : Séminaire exceptionnel, une heure après la fin du séminaire hebdomadaire.

Empirical Bayes methods and nonparametric mixture models via nonparametric maximum likelihood  Version PDF