Saved in:
Bibliographic Details
Main Author: Zhang, Xicheng
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2509.18201
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918145792409600
author Zhang, Xicheng
author_facet Zhang, Xicheng
contents We propose a novel zeroth-order optimization algorithm based on an efficient sampling strategy. Under mild global regularity conditions on the objective function, we establish non-asymptotic convergence rates for the proposed method. Comprehensive numerical experiments demonstrate the algorithm's effectiveness, highlighting three key attributes: (i) Scalability: consistent performance in high-dimensional settings (exceeding 100 dimensions); (ii) Versatility: robust convergence across a diverse suite of benchmark functions, including Schwefel, Rosenbrock, Ackley, Griewank, Lévy, Rastrigin, and Weierstrass; and (iii) Robustness to discontinuities: reliable performance on non-smooth and discontinuous landscapes. These results illustrate the method's strong potential for black-box optimization in complex, real-world scenarios.
format Preprint
id arxiv_https___arxiv_org_abs_2509_18201
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Sampling-Based Zero-Order Optimization Algorithms
Zhang, Xicheng
Optimization and Control
Probability
60H10
We propose a novel zeroth-order optimization algorithm based on an efficient sampling strategy. Under mild global regularity conditions on the objective function, we establish non-asymptotic convergence rates for the proposed method. Comprehensive numerical experiments demonstrate the algorithm's effectiveness, highlighting three key attributes: (i) Scalability: consistent performance in high-dimensional settings (exceeding 100 dimensions); (ii) Versatility: robust convergence across a diverse suite of benchmark functions, including Schwefel, Rosenbrock, Ackley, Griewank, Lévy, Rastrigin, and Weierstrass; and (iii) Robustness to discontinuities: reliable performance on non-smooth and discontinuous landscapes. These results illustrate the method's strong potential for black-box optimization in complex, real-world scenarios.
title Sampling-Based Zero-Order Optimization Algorithms
topic Optimization and Control
Probability
60H10
url https://arxiv.org/abs/2509.18201