Congratulations to Our Newest PhD Graduate, Dr. Andrew Cooper!

The Department of Statistics at Virginia Tech is proud to announce and celebrate the successful completion of Dr. Andrew Cooper's doctoral degree requirements.

Dr. Andrew Cooper successfully defended his dissertation titled "Modernizing Latent Gaussian Process Inference for Non-Gaussian Response Surfaces" on March 16, 2026, a significant milestone representing years of rigorous work, dedication, and impactful research.

Thesis Title: Modernizing Latent Gaussian Process Inference for Non-Gaussian Response Surfaces

Thesis Abstract: Gaussian processes (GPs) are powerful models for both field and computer experiments, but crucially assume a continuous, real-valued response. Latent GPs in combination with proper transformations enable inference for non-Gaussian data, but introduce computational challenges to both inference and prediction, especially in large data settings. In my talk I will discuss strategies for fully Bayesian latent GP inference in two different contexts, both of which leverage elliptical slice sampling algorithms. The first is with a categorical response, where Vecchia-approximated covariance estimation allows for scalable GP classification with appropriate uncertainty quantification. The second example introduces a novel framework for a "wrapped" GP to properly model a noisy angular response. The latter is motivated by the problem of localizing radio frequency identification (RFID) tags used for tracking nuclear materials. I will showcase our model's ability to capture the relationship between frequency and phase angle in order to accurately range assets in laboratory environments. 

Dr. Andrew Cooper will be joining Lawrence Livermore National Laboratory.