Lipschitz continuity of probability kernels in the optimal transport framework

DOI10.1214/23-AIHP1389MaRDI QIDQ6187900zbMATH OpenOpenAlexFDO

Authors Emanuele Dolera, Edoardo Mainini

Publication date 16 January 2024

Published in Annales de l'Institut Henri Poincaré. Probabilités et Statistiques (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/2010.08380

zbMATH Keywords

optimal transportWasserstein distanceBayesian consistencyBenamou-Brenier formulaprobability kernelBayes formula

Mathematics Subject Classification ID

Bayesian inference (62F15) Optimal transportation (49Q22) Large deviations (60F10)

Abstract: In Bayesian statistics, a continuity property of the posterior distribution with respect to the observable variable is crucial as it expresses well-posedness, i.e., stability with respect to errors in the measurement of data. Essentially, this requires to analyze the continuity of a probability kernel or, equivalently, of a conditional probability distribution with respect to the conditioning variable. Here, we give general conditions for the Lipschitz continuity of probability kernels with respect to metric structures arising within the optimal transport framework, such as the Wasserstein metric. For dominated probability kernels over finite-dimensional spaces, we show Lipschitz continuity results with a Lipschitz constant enjoying explicit bounds in terms of Fisher-information functionals and weighted Poincar'e constants. We also provide results for kernels with moving support, for infinite-dimensional spaces and for non dominated kernels. We show applications to several problems in Bayesian statistics, such as approximation of posterior distributions by mixtures and posterior consistency.

Recommendations

• On the local Lipschitz stability of Bayesian inverse problems
• Generalized Lipschitz-continuity of integrals with respect to a parameter of the intergrating probability measure
• Quantitative stability of optimal transport maps under variations of the target measure
• On the well-posedness of Bayesian inverse problems
• Continuity properties of the extension of a locally Lipschitz continuous map to the space of probability measures

Cites work

• scientific article; zbMATH DE number 1713116 (Why is no real title available?)
• scientific article; zbMATH DE number 3707455 (Why is no real title available?)
• scientific article; zbMATH DE number 3602484 (Why is no real title available?)
• scientific article; zbMATH DE number 1201678 (Why is no real title available?)
• scientific article; zbMATH DE number 1354815 (Why is no real title available?)
• scientific article; zbMATH DE number 472946 (Why is no real title available?)
• scientific article; zbMATH DE number 1909499 (Why is no real title available?)
• scientific article; zbMATH DE number 847272 (Why is no real title available?)
• scientific article; zbMATH DE number 3894218 (Why is no real title available?)
• scientific article; zbMATH DE number 2193914 (Why is no real title available?)
• A Sobolev inequality and Neumann heat kernel estimate for unbounded domains
• A computational fluid mechanics solution to the Monge-Kantorovich mass transfer problem
• A dozen de Finetti-style results in search of a theory
• A first course in Sobolev spaces
• A note on the Poincaré inequality for convex domains
• A remark on optimal weighted Poincaré inequalities for convex domains
• A simple proof of the Poincaré inequality for a large class of probability measures
• An introduction to Bayesian analysis. Theory and methods.
• An introduction to continuum mechanics
• An optimal Poincaré inequality for convex domains
• An optimal Poincaré inequality in $L^1$ for convex domains
• Approximating priors by finite mixtures of conjugate distributions for an exponential family
• Approximation of laws of random probabilities by mixtures of Dirichlet distributions with applications to nonparametric Bayesian inference
• Asymptotic approximations for probability integrals
• Bayesian inverse problems for functions and applications to fluid mechanics
• Best constant in Sobolev inequality
• Best constants in Poincar\'e inequalities for convex domains
• Caratterizzazioni delle tracce sulla frontiera relative ad alcune classi di funzioni in n variabili
• Conditional distributions as derivatives
• Conditioning as disintegration
• Continuous disintegrations of Gaussian processes
• Continuous versions of regular conditional distributions
• Convex Sobolev inequalities derived from entropy dissipation
• Empirical Bayes deconvolution estimates
• Entropies, convexity, and functional inequalities: on \(\Phi\)-entropies and \(\Phi\)-Sobolev inequalities
• Frequentistic approximations to Bayesian prevision of exchangeable random elements
• Fundamentals of nonparametric Bayesian inference
• Gaussian null sets and differentiability of Lipschitz map on Banach spaces
• Generalization of an inequality by Talagrand and links with the logarithmic Sobolev inequality
• Gradient flows in metric spaces and in the space of probability measures
• Hypercontractivity for perturbed diffusion semigroups
• Inverse problems: a Bayesian perspective
• Isoperimetric and analytic inequalities for log-concave probability measures
• Logarithmic Sobolev inequalities and stochastic Ising models
• Measure theory and fine properties of functions
• Nonsmooth calculus
• Note on Friedrichs' inequality in N-star-shaped domains
• Note on ``Frequentistic approximations to Bayesian prevision of exchangeable random elements
• On Choosing and Bounding Probability Metrics
• On Fréchet differentiability of Lipschitz maps between Banach spaces
• On Rates of Convergence for Bayesian Density Estimation
• On a new axiomatic theory of probability
• On computability and disintegration
• On logarithmic Sobolev inequalities. With a preface of Dominique Bakry and Michel Ledoux
• On the consistency of Bayes estimates
• On the rate of convergence in Wasserstein distance of the empirical measure
• On the well-posedness of Bayesian inverse problems
• On uniform continuity of posterior distributions
• One-dimensional empirical measures, order statistics, and Kantorovich transport distances
• Poincaré and logarithmic Sobolev inequalities by decomposition of the energy landscape
• Problèmes isoperimetriques et espaces de Sobolev
• Rate of convergence for ergodic continuous Markov processes: Lyapunov versus Poincaré
• Robust Bayesian inference via coarsening
• Sobolev-Poincaré implies John
• THE GEOMETRY OF DISSIPATIVE EVOLUTION EQUATIONS: THE POROUS MEDIUM EQUATION
• The Bayesian formulation and well-posedness of fractional elliptic inverse problems
• The Wiener test for degenerate elliptic equations
• The local regularity of solutions of degenerate elliptic equations
• Unbounded Poincaré domains
• Uniform rates of the Glivenko-Cantelli convergence and their use in approximating Bayesian inferences
• Variation and optimization of formes. A geometric analysis
• Well-posed Bayesian geometric inverse problems arising in subsurface flow
• Well-posed Bayesian inverse problems and heavy-tailed stable quasi-Banach space priors

This page was built for publication: Lipschitz continuity of probability kernels in the optimal transport framework

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6187900)