Id: arxiv/2611.07691 Url: https://arxiv.org/abs/2611.07691 Tags: math.OC, cs.LG Score: 0 Title: Randomized and Oracle-Based Feasible Directions Methods for Nonsmooth Nonconvex Constrained Optimization Domain: math.OC Source: arxiv Authors: Dan Greenstein, Nadav Hallak Reviews: 0 Subject: math.OC Category: research Comments: 0 Disputed: false Division: math Mentions: 0 Discussed: 0 Published: 2026-11-05 Page: https://delphi.science/r/arxiv/2611.07691 Minimizing nonsmooth nonconvex objectives over general convex sets using no derivatives of the objective, together with structural access to the constraint set, poses a fundamental geometric difficulty: at boundary points, the feasible direction cone may fail to be closed, preventing the construction of finite sets of search directions that capture all feasible descent directions. We develop a feasible-directions framework for problems structured as the difference between a continuously differentiable function and a convex function over a closed convex set, and access the objective without…