Bregman level proximal subdifferentials and new characterizations of Bregman proximal operators
Ziyuan Wang, Andreas Themelis

TL;DR
This paper introduces Bregman level proximal subdifferentials to better characterize Bregman proximal operators, extending classical variational analysis to non-convex settings with new correspondences and properties.
Contribution
It develops a systematic framework for Bregman level proximal subdifferentials, generalizing classical Euclidean results to non-convex and asymmetric Bregman contexts.
Findings
Bregman proximal operator is the resolvent of Bregman level proximal subdifferential.
New equivalences among properties of Bregman proximal operator, function, and subdifferential.
Introduction of anisotropic firm nonexpansiveness characterizing relative smoothness.
Abstract
Classic subdifferentials in variational analysis may fail to fully represent the Bregman proximal operator in the absence of convexity. In this paper, we fill this gap by introducing the left and right \emph{Bregman level proximal subdifferentials} and investigate them systematically. Every Bregman proximal operator turns out to be the resolvent of a Bregman level proximal subdifferential under a standard range assumption, even without convexity. Aided by this pleasant feature, we establish new correspondences among useful properties of the Bregman proximal operator, the underlying function, and the (left) Bregman level proximal subdifferential, generalizing classical equivalences in the Euclidean case. Unlike the classical setting, asymmetry and duality gap emerge as natural consequences of the Bregman distance. Along the way, we improve results by Kan and Song and by Wang and…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
