Skip to main content

Bayesian Statistics

Bayesian methods have become extremely valuable in scientific research over the past three decades as new computer algorithms (Markov chain Monte Carlo, variational inference, and other approximations) and vast technological improvements in computing have allowed hierarchical Bayesian models to be fit to complex data. Bayesian methods are particularly well suited to combining data from multiple sources, while appropriately accounting for uncertainties throughout the analysis.

Penn State Statistics has several faculty who work on developing Bayesian methods for solving challenging problems. Examples of interdisciplinary research applications for which our faculty are developing Bayesian methods include neuroscience (Nicole Lazar), network models for social science and public health (Maggie Niu), astronomy (Hyungsuk Tak), ecology and disease modeling (Ephraim Hanks and Murali Haran), and statistical genetics/genomics (Xiang Zhu and Justin Silverman).



Professor of Marketing and Statistics

Associate Professor, Director of the Statistical Consulting Center