First- and second-order high probability complexity bounds for trust-region methods with noisy oracles

• 成果类型: Article
• 署名作者: Cao, Liyuan; Berahas, Albert S.; Scheinberg, Katya
• 署名单位: Peking University; University of Michigan System; University of Michigan; Cornell University
• 刊物名称: MATHEMATICAL PROGRAMMING
• ISSN/ISSBN: 0025-5610
• DOI: 10.1007/s10107-023-01999-5
• 发表日期: 2024
• 页码: 55-106
• 关键词: derivative-free optimization; GLOBAL CONVERGENCE

摘要

In this paper, we present convergence guarantees for a modified trust-region method designed for minimizing objective functions whose value and gradient and Hessian estimates are computed with noise. These estimates are produced by generic stochastic oracles, which are not assumed to be unbiased or consistent. We introduce these oracles and show that they are more general and have more relaxed assumptions than the stochastic oracles used in prior literature on stochastic trust-region methods. Our method utilizes a relaxed step acceptance criterion and a cautious trust-region radius updating strategy which allows us to derive exponentially decaying tail bounds on the iteration complexity for convergence to points that satisfy approximate first- and second-order optimality conditions. Finally, we present two sets of numerical results. We first explore the tightness of our theoretical results on an example with adversarial zeroth- and first-order oracles. We then investigate the performance of the modified trust-region algorithm on standard noisy derivative-free optimization problems.