arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2306.13618math.OC
收藏

非平衡最优传输与最大均值差异:相互关联与快速评估

Unbalanced Optimal Transport and Maximum Mean Discrepancies: Interconnections and Rapid Evaluation

Rajmadan Lakshmanan, Alois Pichler

AI总结:

本文针对不同质量测度的比较问题,基于Sinkhorn算法用非均匀快速傅里叶变换加速非平衡最优传输与核最大均值差异的计算,将复杂度降至$\n O(n\log n)$,并建立了传输问题与最大均值差异的关联。

摘要:

本文针对质量存在差异的测度比较问题,提出了显著的计算改进方案。具体而言,我们基于Sinkhorn算法,利用非均匀快速傅里叶变换加速非平衡最优传输近似中的径向核卷积运算,同时还提出了涉及核的最大均值差异加速方案。我们的方法将计算距离所需的算术运算量从$\boldsymbol{\n O}(n^2)$降至$\boldsymbol{\n O}(n\log n)$,为高效处理大规模高维数据集提供了可能。此外,我们还建立了包含Wasserstein距离、非平衡最优传输在内的传输问题与最大均值差异之间的稳固关联,为从业者选用适配性距离提供了充分依据。

英文摘要:

This contribution presents substantial computational advancements to compare measures even with varying masses. Specifically, we utilize the nonequispaced fast Fourier transform to accelerate the radial kernel convolution in unbalanced optimal transport approximation, built upon the Sinkhorn algorithm. We also present accelerated schemes for maximum mean discrepancies involving kernels. Our approaches reduce the arithmetic operations needed to compute distances from $\mathcal O(n^2)$ to $\mathcal O(n\log n)$, opening the door to handle large and high-dimensional datasets efficiently. Furthermore, we establish robust connections between transportation problems, encompassing Wasserstein distance and unbalanced optimal transport, and maximum mean discrepancies. This empowers practitioners with compelling rationale to opt for adaptable distances.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑