Department of Statistical Sciences at University of Toronto
Office 9183, 700 University Avenue, 9th Floor
Toronto, ON, M5G 1Z5
I am a Postdoctoral Fellow in statistics at the University of Toronto supervised by Jeffrey Rosenthal. Previously, I was a Research Fellow at the University of Warwick supervised by Krzysztof Łatuszyński and Gareth Roberts. I completed my PhD in statistics at the University of Minnesota, Twin Cities in August 2022 where my advisor was Galin Jones.
My main research focus is on the design, reliability, and computational efficiency of Markov chain Monte Carlo algorithms by studying their convergence properties with optimal transportation. Creating reliable, computationally efficient Markov chain Monte Carlo algorithms in high dimensions for practitioners is a core motivation for my research. Additionally, I am interested in applications of Markov chain Monte Carlo to Bayesian error-in-variable models for machine learning, astrophysics, and epidemiology. My other research interests are in applications of optimal transport to stochastic processes, and I am also interested in combining optimal transport and theory for Markov chain Monte Carlo to study generative A.I. models in the future.
My curriculum vitae has an updated list of my publications and and the majority of my publication preprints are available on arXiv. This site also contains some blog posts when I was a graduate student.
- Austin Brown and Jeffrey S. Rosenthal (2024). Weak convergence of adaptive Markov chain Monte Carlo. minor revision requested to Journal of Applied Probability. [preprint]
- Sabrina Sixta and Jeffrey S. Rosenthal and Austin Brown (2024). Bounding and estimating MCMC convergence rates using common random number simulations. Submitted. [preprint]
- Austin Brown (2024). A non-asymptotic error analysis for parallel Monte Carlo estimation from many short Markov chains. Submitted. [preprint] [code]
- Austin Brown and Galin L. Jones (2024). Lower bounds on the rate of convergence for accept-reject-based Markov chains in Wasserstein and total variation distances. Bernoulli, to appear. [paper] [preprint] [code]
- Austin Brown and Galin L. Jones (2024). Convergence analysis for Metropolis-Hastings algorithms. WIREs Computational Statistics. [paper]
- Austin Brown (2024). Geometric ergodicity of Gibbs samplers for Bayesian error-in-variable regression. Electronic Journal of Statistics. [paper] [code]
- Austin Brown and Galin L. Jones (2024). Exact convergence analysis for Metropolis–Hastings independence samplers in Wasserstein distances. Journal of Applied Probability. [paper] [code]
- Lower Bounds on the Rate of Convergence Metropolis-Hastings in Wasserstein Distances (2024), B.B. Seminar, University of Toronto [slides]
- Lower Bounds on the Rate of Convergence Metropolis-Hastings in Wasserstein Distances (2024), Statistics Departmental Seminar, University of Florida [slides]
- Lower Bounds on the Rate of Convergence for Accept-Reject-Based Markov Chains (2023), Algorithms and Computationally Intensive Inference seminars, University of Warwick [slides]
- Exact convergence for independence samplers in Wasserstein distance (2023), University of Warwick Departmental Conference 2023 [slides]
- Geometric Ergodicity of Gibbs Samplers for Bayesian Error-in-variable Regression (2023), Bioinference 2023, Oxford University [poster]
- Methods of Data Analysis 1 (STA302), University of Toronto, Fall 2024
- Methods of Data Analysis 1 (STA302), University of Toronto, Fall 2023
- Introduction to Statistical Analysis (STAT 3011), University of Minnesota, Spring 2021
- Regression and Correlated Data (STAT 3032), University of Minnesota, Spring 2020