Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Xia, Jingfan, Lin, Zhenwei, Deng, Qi
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2508.13496
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866911143753154560
author Xia, Jingfan
Lin, Zhenwei
Deng, Qi
author_facet Xia, Jingfan
Lin, Zhenwei
Deng, Qi
contents Randomized smoothing is a widely adopted technique for optimizing nonsmooth objective functions. However, its efficiency analysis typically relies on global Lipschitz continuity, a condition rarely met in practical applications. To address this limitation, we introduce a new subgradient growth condition that naturally encompasses a wide range of locally Lipschitz functions, with the classical global Lipschitz function as a special case. Under this milder condition, we prove that randomized smoothing yields a differentiable function that satisfies certain generalized smoothness properties. To optimize such functions, we propose novel randomized smoothing gradient algorithms that, with high probability, converge to $(δ, ε)$-Goldstein stationary points and achieve a sample complexity of $\tilde{\mathcal{O}}(d^{5/2}δ^{-1}ε^{-4})$. By incorporating variance reduction techniques, we further improve the sample complexity to $\tilde{\mathcal{O}}(d^{3/2}δ^{-1}ε^{-3})$, matching the optimal $ε$-bound under the global Lipschitz assumption, up to a logarithmic factor. Experimental results validate the effectiveness of our proposed algorithms.
format Preprint
id arxiv_https___arxiv_org_abs_2508_13496
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Revisiting Randomized Smoothing: Nonsmooth Nonconvex Optimization Beyond Global Lipschitz Continuity
Xia, Jingfan
Lin, Zhenwei
Deng, Qi
Optimization and Control
Randomized smoothing is a widely adopted technique for optimizing nonsmooth objective functions. However, its efficiency analysis typically relies on global Lipschitz continuity, a condition rarely met in practical applications. To address this limitation, we introduce a new subgradient growth condition that naturally encompasses a wide range of locally Lipschitz functions, with the classical global Lipschitz function as a special case. Under this milder condition, we prove that randomized smoothing yields a differentiable function that satisfies certain generalized smoothness properties. To optimize such functions, we propose novel randomized smoothing gradient algorithms that, with high probability, converge to $(δ, ε)$-Goldstein stationary points and achieve a sample complexity of $\tilde{\mathcal{O}}(d^{5/2}δ^{-1}ε^{-4})$. By incorporating variance reduction techniques, we further improve the sample complexity to $\tilde{\mathcal{O}}(d^{3/2}δ^{-1}ε^{-3})$, matching the optimal $ε$-bound under the global Lipschitz assumption, up to a logarithmic factor. Experimental results validate the effectiveness of our proposed algorithms.
title Revisiting Randomized Smoothing: Nonsmooth Nonconvex Optimization Beyond Global Lipschitz Continuity
topic Optimization and Control
url https://arxiv.org/abs/2508.13496