arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2302.00982math.PRstat.ML
收藏

Banach空间中的随机最优输运用于多元分位数的正则化估计

Stochastic optimal transport in Banach Spaces for regularized estimation of multivariate quantiles

Bernard Bercu, Jérémie Bigot, Gauthier Thurin

AI总结:

针对源测度为单位超立方体或球面均匀分布的 Monge-Kantorovich 分位数问题,提出用 Fourier 系数参数化对偶势并借助 FFT 快速求解 EOT 的随机算法,证明其 Banach 空间中的几乎必然收敛,并用于多元分位数平滑估计。

摘要:

我们提出了一种新的随机算法,用于求解两个绝对连续概率测度 μ 和 ν 之间的熵正则化最优输运(EOT)。我们的工作受到 Monge-Kantorovich 分位数这一特定设定的启发,其中源测度 μ 要么是单位超立方体上的均匀分布,要么是球面均匀分布。利用对源测度的了解,我们建议通过其 Fourier 系数来参数化一个 Kantorovich 对偶势。这样,我们随机算法的每次迭代都可归结为两次 Fourier 变换,这使我们能够利用快速 Fourier 变换(FFT)来实现一种快速数值方法以求解 EOT。我们研究了该取值于无限维 Banach 空间的随机算法的几乎必然收敛性。随后,通过数值实验,我们展示了本方法在正则化 Monge-Kantorovich 分位数计算上的性能。特别地,我们利用从目标测度 ν 采样得到的数据,考察了熵正则化在多元分位数平滑估计方面的潜在益处。

英文摘要:

We introduce a new stochastic algorithm for solving entropic optimal transport (EOT) between two absolutely continuous probability measures $μ$ and $ν$. Our work is motivated by the specific setting of Monge-Kantorovich quantiles where the source measure $μ$ is either the uniform distribution on the unit hypercube or the spherical uniform distribution. Using the knowledge of the source measure, we propose to parametrize a Kantorovich dual potential by its Fourier coefficients. In this way, each iteration of our stochastic algorithm reduces to two Fourier transforms that enables us to make use of the Fast Fourier Transform (FFT) in order to implement a fast numerical method to solve EOT. We study the almost sure convergence of our stochastic algorithm that takes its values in an infinite-dimensional Banach space. Then, using numerical experiments, we illustrate the performances of our approach on the computation of regularized Monge-Kantorovich quantiles. In particular, we investigate the potential benefits of entropic regularization for the smooth estimation of multivariate quantiles using data sampled from the target measure $ν$.

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。

↑