Shifted divergences for sampling, privacy, and beyond

Speaker: Jason Altschuler

Date: Thu, Oct 12, 2023

Location: PIMS, University of Washington, Online

Conference: Kantorovich Initiative Seminar

Subject: Mathematics

Class: Scientific

CRG: Pacific Interdisciplinary Hub on Optimal Transport


Shifted divergences provide a principled way of making information theoretic divergences (e.g. KL) geometrically aware via optimal transport smoothing. In this talk, I will argue that shifted divergences provide a powerful approach towards unifying optimization, sampling, privacy, and beyond. For concreteness, I will demonstrate these connections via three recent highlights. (1) Characterizing the differential privacy of Noisy-SGD, the standard algorithm for private convex optimization. (2) Characterizing the mixing time of the Langevin Algorithm to its stationary distribution for log-concave sampling. (3) The fastest high-accuracy algorithm for sampling from log-concave distributions. A recurring theme is a certain notion of algorithmic stability, and the central technique for establishing this is shifted divergences. Based on joint work with Kunal Talwar, and with Sinho Chewi.