arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2407.01202math.OC
收藏

连续设置下Sinkhorn算法的更锐利指数收敛速率

Sharper Exponential Convergence Rates for Sinkhorn's Algorithm in Continuous Settings

Lénaïc Chizat, Alex Delalande, Tomas Vaškevičius

AI总结:

本文研究连续设置下Sinkhorn算法求解熵正则化最优传输的收敛速率,在测度具有密度且成本函数半凹的条件下,将收缩率从Hilbert投影度量给出的指数依赖改进为λ/c∞的多项式依赖,并在对数凹测度等特殊情形下进一步改进为1-Θ(λ/c∞),且结果对所有参数显式非渐近。

摘要:

我们研究用于求解熵正则化最优传输问题的Sinkhorn算法的收敛速率,其中至少一个概率测度μ在$\mathbb{R}^d$上具有密度。对于以$c_{\infty}$为界的半凹成本函数和正则化参数$λ>0$,我们得到了对偶次优性间隙的指数收敛保证,其收缩率是$λ/c_{\infty}$的多项式。这相对于通过Hilbert投影度量可达到的已知收缩率$1 - Θ(\exp(-c_{\infty}/λ))$而言,是一个指数级的改进。具体来说,当μ具有有界对数密度时,我们证明了收缩率值为$1-Θ(λ^2/c_\infty^2)$。在某些情况下,例如当μ是对数凹的且成本函数为$c(x,y)=-\langle x, y \rangle$时,该速率改进为$1-Θ(λ/c_\infty)$。后一速率与我们为各向同性高斯测度之间的传输所推导出的速率一致,表明在$λ/c_\infty$依赖关系上的紧致性。我们的结果完全是非渐近的,并且对问题的所有参数都是显式的。

英文摘要:

We study the convergence rate of Sinkhorn's algorithm for solving entropy-regularized optimal transport problems when at least one of the probability measures, $μ$, admits a density over $\mathbb{R}^d$. For a semi-concave cost function bounded by $c_{\infty}$ and a regularization parameter $λ> 0$, we obtain exponential convergence guarantees on the dual sub-optimality gap with contraction rate polynomial in $λ/c_{\infty}$. This represents an exponential improvement over the known contraction rate $1 - Θ(\exp(-c_{\infty}/λ))$ achievable via Hilbert's projective metric. Specifically, we prove a contraction rate value of $1-Θ(λ^2/c_\infty^2)$ when $μ$ has a bounded log-density. In some cases, such as when $μ$ is log-concave and the cost function is $c(x,y)=-\langle x, y \rangle$, this rate improves to $1-Θ(λ/c_\infty)$. The latter rate matches the one that we derive for the transport between isotropic Gaussian measures, indicating tightness in the dependency in $λ/c_\infty$. Our results are fully non-asymptotic and explicit in all the parameters of the problem.

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。

↑