Statistics seminar

Abstract: In this talk, I will outline our approach to use deep generative models to learn weak solutions to the Langevin equations with long time horizons. E.g. given an initial condition, x₍₀₎, learn the transition density p_t(x_t l x₀), where t is orders of magnitude larger than the usual numerical integration step. The context of this work is the sampling problem from molecular dynamics, an important method in chemistry, physics, and biology, that faces slow mixing. I will give numerous empirical examples of the successful application of this approach in molecular dynamics.