Id: arxiv/2601.20076 Url: https://arxiv.org/abs/2601.20076 Tags: math.OC, cs.LG Score: 0 Title: Randomized Feasibility Methods for Constrained Optimization with Adaptive Step Sizes Domain: math.OC Source: arxiv Authors: Abhishek Chakraborty, Angelia Nedić Reviews: 0 Subject: math.OC Category: research Comments: 0 Disputed: false Division: math Mentions: 0 Discussed: 0 Published: 2026-01-27 Page: https://delphi.science/r/arxiv/2601.20076 We consider minimizing an objective function subject to constraints defined by the intersection of lower-level sets of convex functions. We study two cases: (i) strongly convex and Lipschitz-smooth objective function and (ii) convex but possibly nonsmooth objective function. To deal with the constraints that are not easy to project on, we use a randomized feasibility algorithm with Polyak steps and a random number of sampled constraints per iteration, while taking (sub)gradient steps to minimize the objective function. For case (i), we prove linear convergence in expectation of the objective…