SDS Seminar Series – Jeffrey Miller, Harvard University

The Fall 2025 SDS Seminar Series continues on November 11th from 2:00 p.m. to 3:00 p.m. with Dr. Jeffrey Miller (Associate Professor, Department of Biostatistics, Harvard University). This event is in-person in the Avaya Room (POB 2.302).    

Title: Bayesian Model Criticism Using Uniform Parametrization Checks

Abstract: Models are often misspecified in practice, making model criticism a key part of Bayesian analysis.  It is important to detect not only when a model is wrong, but which aspects are wrong, and to do so in a computationally convenient and statistically rigorous way.  We introduce a novel method for model criticism based on the fact that if the parameters are drawn from the prior, and a dataset is generated according to the assumed likelihood, then a sample from the posterior will be distributed according to the prior.  Building upon this idea, we propose to reparameterize all random elements of the likelihood and prior in terms of independent uniform random variables.  This makes it possible to aggregate across arbitrary subsets of data points and parameters to test for model departures using classical hypothesis tests for dependence or non-uniformity.  We demonstrate empirically how this method of uniform parameterization checks (UPCs) facilitates model criticism, and we develop supporting theoretical results.