Monte Carlo Uncertainty, Scaling Laws, and Geometry Stress Tests for Bayesian Samplers

• Course: STA 663L: Statistical Computing
• Methods: Monte Carlo methods, MCMC diagnostics, scaling analysis, geometric test distributions, uncertainty quantification

This project examined the behavior of Bayesian samplers under Monte Carlo uncertainty, studied their scaling properties, and designed geometry-based stress tests to probe sampler robustness across challenging distributional settings.

• Project Report