News and upcoming events

January 2026

• During January and February of 2026, Ashlynn Crisp (one of our Lab's members) will be at Princeton University working with Professor Vincent Poor as a Visiting Student Research Collaborator.

• Starting January, Jacob Schultz  (one of our Lab's members) will carry out statistical validation work for the Connectivity Assessment and Mapping for Western Burrowing Owl in New Mexico, under the direction of our collaborator Martin Lafrenz. This project continues our work with OCAMP (in Oregon), to identify habitat requirements, preferences, and tolerance for moving through unsuitable habitats. Identify barriers to movement, including roadways, developed areas.

December 2025

• Our paper The matryoshka doll prior – principled multiplicity correction in Bayesian model comparison (advance publication link) was accepted for publication on Bayesian Analysis.  Here, we propose the matryoshka doll model space prior, which is based on the simple idea that a model’s prior probability should be proportional to the total probability of all models that nest it. This principle explicitly corrects for multiplicity and delivers a “just right” penalty on model complexity, even when the number of predictors exceeds the sample size.

November 2025

• In our paper Clustering the  Nearest Neighbor Gaussian Process with A. Crisp and A. Finley  (preprint on arxiv), we propose an approach to enhance the computational efficiency of the NNGP for Bayesian spatial regression with stationary data. Here we exploit the fact that (under the NNGP) if two different locations share the same distance structure with respective m-nearest neighbors, then their covariance will also be the same. This observation allows us to cluster locations, drastically reducing the computational burden while preserving the predictive ability of the NNGP.