Id: arxiv/2607.29408 Url: https://arxiv.org/abs/2607.29408 Tags: math.OC Score: 0 Title: Extrapolation-based Direct Search for Nonsmooth Stochastic Zeroth-Order Optimization Domain: math.OC Source: arxiv Authors: Anthony Palmieri, Francesco Rinaldi, Sara Shashaani Reviews: 0 Subject: math.OC Category: research Comments: 0 Disputed: false Division: math Mentions: 0 Discussed: 0 Published: 2026-07-31 Page: https://delphi.science/r/arxiv/2607.29408 We propose and analyze a stochastic direct-search method for unconstrained zeroth-order minimization of locally Lipschitz, possibly nonsmooth, objectives. The method combines random polling directions with a stochastic extrapolating line search based on a sufficient-decrease test of order $p$. Under conditional accuracy assumptions on the stochastic estimates, we prove almost-sure convergence to Clarke stationary points. We further establish an expected iteration complexity bound. Specifically, using a supermartingale stopping-time argument, we prove that $\mathcal O\left( \max\left\{…