English

Decentralized Gradient-Free Methods for Stochastic Non-Smooth Non-Convex Optimization

Optimization and Control 2025-01-29 v2 Distributed, Parallel, and Cluster Computing

Abstract

We consider decentralized gradient-free optimization of minimizing Lipschitz continuous functions that satisfy neither smoothness nor convexity assumption. We propose two novel gradient-free algorithms, the Decentralized Gradient-Free Method (DGFM) and its variant, the Decentralized Gradient-Free Method+^+ (DGFM+^{+}). Based on the techniques of randomized smoothing and gradient tracking, DGFM requires the computation of the zeroth-order oracle of a single sample in each iteration, making it less demanding in terms of computational resources for individual computing nodes. Theoretically, DGFM achieves a complexity of O(d3/2δ1ε4)\mathcal O(d^{3/2}\delta^{-1}\varepsilon ^{-4}) for obtaining an (δ,ε)(\delta,\varepsilon)-Goldstein stationary point. DGFM+^{+}, an advanced version of DGFM, incorporates variance reduction to further improve the convergence behavior. It samples a mini-batch at each iteration and periodically draws a larger batch of data, which improves the complexity to O(d3/2δ1ε3)\mathcal O(d^{3/2}\delta^{-1} \varepsilon^{-3}). Moreover, experimental results underscore the empirical advantages of our proposed algorithms when applied to real-world datasets.

Keywords

Cite

@article{arxiv.2310.11973,
  title  = {Decentralized Gradient-Free Methods for Stochastic Non-Smooth Non-Convex Optimization},
  author = {Zhenwei Lin and Jingfan Xia and Qi Deng and Luo Luo},
  journal= {arXiv preprint arXiv:2310.11973},
  year   = {2025}
}
R2 v1 2026-06-28T12:54:24.380Z