 Langevin dynamics based algorithm e-THO POULA for
stochastic optimization problems with discontinuous stochastic gradientMathematics of Operations Research (MOR), 2022 |
 Non-asymptotic estimates for TUSLA algorithm for non-convex learning
with applications to neural networks with ReLU activation functionIMA Journal of Numerical Analysis (IMA J. Numer. Anal.), 2021 |