BriefGPT.xyz
Aug, 2024
提升算法的最佳并行化
Optimal Parallelization of Boosting
HTML
PDF
Arthur da Cunha, Mikael Møller Høgsgaard, Kasper Green Larsen
TL;DR
本研究解决了提升算法在并行复杂性方面理论下限与算法性能之间的显著差距问题。通过提供改进的并行复杂性下限和一种并行提升算法，研究展示了这一算法在整个训练轮次与每轮总并行工作量的权衡中性能匹配这些下限。最终，这项工作确立了近似样本最优的提升算法的真实并行复杂性。
Abstract
Recent works on the parallel
Complexity
of
Boosting
have established strong lower bounds on the tradeoff between the number of training rounds $p$ and the total parallel work per round $t$. These works have also
→