Stochastic Non-Smooth Non-Convex Optimization with Decision-Dependent Distributions

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Liu, Chengchang, Wan, Zongqi, Ye, Haishan, Lui, John C. S.
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911658209705984
author Liu, Chengchang
Wan, Zongqi
Ye, Haishan
Lui, John C. S.
author_facet Liu, Chengchang
Wan, Zongqi
Ye, Haishan
Lui, John C. S.
contents We study stochastic zeroth-order optimization with decision-dependent distributions, where the sampling law depends on the current decision and only noisy function values are available. For the non-smooth non-convex setting, we establish an explicit convergence guarantee for finding a $(δ,ε)$-Goldstein stationary point with stochastic zeroth-order oracle (SZO) complexity of $\mathcal{O}(d^2δ^{-3}ε^{-3})$. In addition, we show that the above complexity can be achieved with single SZO feedback per iteration. We further extend the analysis to smooth and Hessian-Lipschitz objectives, obtaining complexities $\mathcal{O}(d^2ε^{-6})$ and $\mathcal{O}(d^2ε^{-9/2})$, respectively. In the Hessian-Lipschitz case, this improves the best-known dependence on $ε$ for decision-dependent zeroth-order methods by a factor of $ε^{-1/2}$.
format Preprint
id arxiv_https___arxiv_org_abs_2605_06549
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Stochastic Non-Smooth Non-Convex Optimization with Decision-Dependent Distributions
Liu, Chengchang
Wan, Zongqi
Ye, Haishan
Lui, John C. S.
Optimization and Control
We study stochastic zeroth-order optimization with decision-dependent distributions, where the sampling law depends on the current decision and only noisy function values are available. For the non-smooth non-convex setting, we establish an explicit convergence guarantee for finding a $(δ,ε)$-Goldstein stationary point with stochastic zeroth-order oracle (SZO) complexity of $\mathcal{O}(d^2δ^{-3}ε^{-3})$. In addition, we show that the above complexity can be achieved with single SZO feedback per iteration. We further extend the analysis to smooth and Hessian-Lipschitz objectives, obtaining complexities $\mathcal{O}(d^2ε^{-6})$ and $\mathcal{O}(d^2ε^{-9/2})$, respectively. In the Hessian-Lipschitz case, this improves the best-known dependence on $ε$ for decision-dependent zeroth-order methods by a factor of $ε^{-1/2}$.
title Stochastic Non-Smooth Non-Convex Optimization with Decision-Dependent Distributions
topic Optimization and Control
url https://arxiv.org/abs/2605.06549