An Accelerated Stochastic Algorithm for Solving the Optimal Transport Problem

An Accelerated Stochastic Algorithm for Solving the Optimal Transport Problem
复制标题

求解最优传输问题的加速随机算法

DOI:
10.48550/arxiv.2203.00813
复制
发表时间:
2022
期刊:
ArXiv
影响因子:
--
通讯作者:
X. Huo
X. Huo
中科院分区:
--
文献类型:
--
作者:
Yiling Xie;Yiling Luo;X. Huo

文献摘要

参考文献

被引文献

相似文献

提出了一种求解线性约束优化问题的原始-对偶加速随机梯度下降-方差化算法。PDASGD可用于求解离散最优传输(OT)问题,其计算复杂度是最大的--宽(n^2/epsilon)$,其中$n$为原子数,$\epsilon>0$为精度。在文献中,已经提出了一些原始-对偶加速一阶算法,例如APDAGD,它们的阶数为$\widetilde{\mathcal{O}}(n^{2.5}/\epsilon)$来解决OT问题。为了理解为什么我们提出的算法可以将速度提高一倍,我们讨论了我们的随机算法在解线性约束优化问题时具有较低的计算复杂度的条件。证明了OT问题可以满足上述条件。数值实验表明,提出的PDASGD算法在求解OT问题时具有较好的实用性能。
A primal-dual accelerated stochastic gradient descent with variance reduction algorithm (PDASGD) is proposed to solve linear-constrained optimization problems. PDASGD could be applied to solve the discrete optimal transport (OT) problem and enjoys the best-known computational complexity -- $\widetilde{\mathcal{O}}(n^2/\epsilon)$, where $n$ is the number of atoms, and $\epsilon>0$ is the accuracy. In the literature, some primal-dual accelerated first-order algorithms, e.g., APDAGD, have been proposed and have the order of $\widetilde{\mathcal{O}}(n^{2.5}/\epsilon)$ for solving the OT problem. To understand why our proposed algorithm could improve the rate by a factor of $\widetilde{\mathcal{O}}(\sqrt{n})$, the conditions under which our stochastic algorithm has a lower order of computational complexity for solving linear-constrained optimization problems are discussed. It is demonstrated that the OT problem could satisfy the aforementioned conditions. Numerical experiments demonstrate superior practical performances of the proposed PDASGD algorithm for solving the OT problem.
Wasserstein 空间上的快速平滑插值
DOI: --
发表时间: 2021
期刊: Proceedings of Machine Learning Research
影响因子: --
作者:
Chewi, Sinho;Clancy, Julien;Le Gouic, Thibaut;Rigollet, Philippe;Stepaniants, George;Stromme, Austin
通讯作者: Stromme, Austin
DOI: --
发表时间: 2020-06
期刊: ArXiv
影响因子: --
作者:
Sinho Chewi;Thibaut Le Gouic;Chen Lu;Tyler Maunu;P. Rigollet
通讯作者: Sinho Chewi;Thibaut Le Gouic;Chen Lu;Tyler Maunu;P. Rigollet
DOI: --
发表时间: 2019-06
期刊: J. Mach. Learn. Res.
影响因子: --
作者:
Tianyi Lin;Nhat Ho;Michael I. Jordan
通讯作者: Tianyi Lin;Nhat Ho;Michael I. Jordan
DOI: --
发表时间: 2019-05
期刊: ArXiv
影响因子: --
作者:
Nathaniel Lahn;Deepika Mulchandani;S. Raghvendra
通讯作者: Nathaniel Lahn;Deepika Mulchandani;S. Raghvendra