Assistant Professor of Statistical Science
My research develops asymptotic theory and computational methods for high-dimensional statistical models, particularly within Bayesian inference. I provide rigorous guarantees which characterize when and why classical approximation tools—such as Laplace’s method and Bernstein–von Mises theorems —remain accurate in modern high-dimensional regimes. These insights shed light on the design of scalable and reliable algorithms for modern data.
My most recent work exploits asymptotic techniques to design more efficient rare event sampling methods. I am also interested in high-dimensional large deviations.