AI 中文总结
该研究针对无导数优化,证明对Johnson-Lindenstrauss变换重缩放后,可改进基于随机子空间模型的DFO算法最坏情况评估复杂性界,且与同类最优界匹配。
AI 中文摘要
我们证明,通过合适的重缩放,在[Cartis & Roberts, Math. Prog. 199 (2023)]提出的基于随机子空间模型的无导数优化(DFO)算法中使用Johnson-Lindenstrauss变换(JLT),相比“未重缩放”的JLT可获得更优的最坏情况评估复杂性界。该改进界无需修改原始算法或复杂性分析,且在模型基DFO的维度依赖性上与已知最优界匹配(例如[Scheinberg & Chaudhry, ICM 2026]在类似框架中使用Haar矩阵实现的界)。
英文摘要
We demonstrate that, with a suitable rescaling, using Johnson-Lindenstrauss transforms (JLTs) in the random subspace model-based derivative-free optimization (DFO) algorithm from [Cartis & Roberts, Math. Prog. 199 (2023)] achieves an improved worst-case evaluation complexity bound compared to 'unscaled' JLTs. This improved bound does not require any modification to the original algorithm or complexity analysis, and matches the best-known bound in terms of dimension dependence for model-based DFO (e.g. achieved by [Scheinberg & Chaudhry, ICM 2026] using Haar matrices in a similar framework).