惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

Last Week in AI
Last Week in AI
U
Unit 42
博客园 - 【当耐特】
Y
Y Combinator Blog
H
Hackread – Cybersecurity News, Data Breaches, AI and More
Microsoft Security Blog
Microsoft Security Blog
Recent Announcements
Recent Announcements
P
Proofpoint News Feed
Martin Fowler
Martin Fowler
量子位
让小产品的独立变现更简单 - ezindie.com
让小产品的独立变现更简单 - ezindie.com
有赞技术团队
有赞技术团队
aimingoo的专栏
aimingoo的专栏
博客园 - 司徒正美
美团技术团队
雷峰网
雷峰网
小众软件
小众软件
G
Google Developers Blog
GbyAI
GbyAI
Jina AI
Jina AI
爱范儿
爱范儿
奇客Solidot–传递最新科技情报
奇客Solidot–传递最新科技情报
F
Fortinet All Blogs
Vercel News
Vercel News

math.PR updates on arXiv.org

Visibility in the Boolean Model on Harmonic Manifolds Global estimates on the Brenier map Geodesics and Wandering Exponents in Brochette First-Passage Percolation State-dependent inverse-subordinator time changes of regenerative processes: Excursion structure and multiscale occupation-time limits Randomly twisted transfer operators and singular values statistics Generalized Bessel-Dunkl diffusions An almost sure invariance principle for the Takagi-van der Waerden class functions Central limit theorems for high dimensional lattice polytopes: cosmological polytopes Convergence rate estimates for semigroups and heat kernels associated with resistance forms Second-order Poincaré inequalities and localization on the Poisson space Maximum Probability of Independence in Transitive Matroids On global solutions to the semidiscrete stochastic heat equation The Poisson Tail Conjecture for primes in short intervals A Complete Spectral Analysis of the CEV Operator with Applications to Arbitrage Holographic functions and neural networks From Betting to Empirical Bernstein LIL Concentration of General Stochastic Approximation Under Heavy-Tailed Markovian Noise Pointwise Generalization in Deep Neural Networks Bayesian Latent Space Models for Graphs Are Misspecified: Toward Robust Inference via Generalized Posteriors Wasserstein bounds for denoising diffusion probabilistic models via the Föllmer process A note on connections between the Föllmer process and the denoising diffusion probabilistic model Simple Approximation and Derivative Free Inference-Time Scaling for Diffusion Models via Sequential Monte Carlo on Path Measures Diffusion-Based Stochastic Operator Networks for Uncertainty Quantification in Stochastic Partial Differential Equations A Fourier perspective on the learning dynamics of neural networks: from sample complexities to mechanistic insights Propagation of Chaos in Contextual Flow Maps Dimension-Uniform Discretization Analysis of Preconditioned Annealed Langevin Dynamics for Multimodal Gaussian Mixtures $α$-TCAV: A Unified Framework for Testing with Concept Activation Vectors Scaling Laws from Sequential Feature Recovery: A Solvable Hierarchical Model On the Limits of Latent Reuse in Diffusion Models State-of-art minibatches via novel DPP kernels: discretization, wavelets, and rough objectives
Localisation for the second-order Beckmann problem: bimar...
[Submitted on 8 Feb 2025 (v1), last revised 18 Jul 2026 (this ve · 2025-02-09 · via math.PR updates on arXiv.org

View PDF

Abstract:We develop a second-order localisation theory for optimal transport based on leaf decompositions and bimartingale couplings. It provides a second-order analogue of the classical decomposition of optimal transport into transport rays and monotone couplings: transport rays are replaced by the leaves of the $1$-Lipschitz derivative map $Du$ of an optimal dual potential $u\in C^{1,1}(\mathbb R^n)$, while monotone couplings are replaced by bimartingale couplings.
We apply this framework to the three-marginal optimal transport problem introduced by Bolbotowski and Bouchitté, whose relaxation is the second-order Beckmann problem.
We introduce bimartingale couplings and characterise their existence through a convex-concave order condition. This yields a generalisation of Strassen's theorem from convex order to the convex-concave setting. Equivalently, the dual problem associated with the second-order Beckmann problem admits an optimiser whose derivative is an isometry.
For absolutely continuous measures with common barycentre, assuming the existence of an optimal plan with absolutely continuous third marginal, we prove that every optimal plan decomposes into a family of problems on the leaves of $Du$. On each leaf, all optimal plans are completely described by bimartingale couplings between the corresponding conditional measures.
Without this absolute continuity assumption, we show that the leaf decomposition persists in a more general form: optimal plans are mixtures of plans concentrated on triples $(x,y,z)$ satisfying $x\in \mathcal{S}_1$, $y\in \mathcal{S}_2$, and $z\in \mathcal{S}_1\cap \mathcal{S}_2$, where $\mathcal{S}_1$ and $\mathcal{S}_2$ are neighbouring leaves of $Du$.

Submission history

From: Krzysztof Ciosmak [view email]
[v1] Sat, 8 Feb 2025 20:30:14 UTC (28 KB)
[v2] Thu, 13 Feb 2025 16:40:28 UTC (29 KB)
[v3] Sat, 18 Jul 2026 04:46:31 UTC (66 KB)