Minimax estimation of smooth optimal transport maps

DOI10.1214/20-AOS1997MaRDI QIDQ2039809zbMATH OpenOpenAlexFDO

Authors Jan-Christian Hütter, Philippe Rigollet

Publication date 5 July 2021

Published in The Annals of Statistics (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1905.05828

zbMATH Keywords

minimax ratesnonparametric estimationoptimal transportwavelet estimator

Mathematics Subject Classification ID

Nontrigonometric harmonic analysis involving wavelets and other special systems (42C40) Spaces of measurable functions ((L^p)-spaces, Orlicz spaces, Köthe function spaces, Lorentz spaces, rearrangement invariant spaces, ideal spaces, etc.) (46E30) Nonparametric estimation (62G05) Transportation, logistics and supply chain management (90B06)

Abstract: Brenier's theorem is a cornerstone of optimal transport that guarantees the existence of an optimal transport map $T$ between two probability distributions $P$ and $Q$ over $mathbb{R}^d$ under certain regularity conditions. The main goal of this work is to establish the minimax estimation rates for such a transport map from data sampled from $P$ and $Q$ under additional smoothness assumptions on $T$. To achieve this goal, we develop an estimator based on the minimization of an empirical version of the semi-dual optimal transport problem, restricted to truncated wavelet expansions. This estimator is shown to achieve near minimax optimality using new stability arguments for the semi-dual and a complementary minimax lower bound. Furthermore, we provide numerical experiments on synthetic data supporting our theoretical findings and highlighting the practical benefits of smoothness regularization. These are the first minimax estimation rates for transport maps in general dimension.

Recommendations

• Minimax rate for optimal transport regression between distributions
• Minimax estimation of smooth densities in Wasserstein distance
• Optimal transport between random measures
• Minimizing optimal transport for functions with fixed-size nodal sets
• Optimal Transport for Applications in Control and Estimation
• Minimization for conditional simulation: relationship to optimal transport
• Stochastic optimal transport revisited
• Optimal transport: fast probabilistic approximation with exact solvers
• Approximation of optimal transport problems with marginal moments constraints
• Optimal transportation and action-minimizing measures

Cites work

• A characterization of random variables with minimum L^ 2-distance
• A new approach to least-squares estimation, with applications
• Adaptive invariant density estimation for ergodic diffusions over anisotropic classes
• Asymptotic statistical equivalence for ergodic diffusions: the multidimensional case
• Asymptotic Statistics
• Asymptotically Minimax Adaptive Estimation. I: Upper Bounds. Optimally Adaptive Estimates
• Asymptotically Minimax Adaptive Estimation. II. Schemes without Optimal Adaptation: Adaptive Estimators
• Best constants in Poincar\'e inequalities for convex domains
• Boundary regularity of maps with convex potentials
• Boundary regularity of maps with convex potentials. II
• Central limit theorem and bootstrap procedure for Wasserstein's variations with an application to structural relationships between distributions
• Central limit theorems for entropy-regularized optimal transport on finite spaces and statistical applications
• Concentration inequalities and model selection. Ecole d'Eté de Probabilités de Saint-Flour XXXIII -- 2003.
• Convergence rates for empirical barycenters in metric spaces: curvature, convexity and extendable geodesics
• Convolutional Wasserstein distances: efficient optimal transportation on geometric domains
• Deconvolution for the Wasserstein metric and geometric inference
• Discrete Wasserstein barycenters: optimal transport for discrete data
• Empirical regularized optimal transport: statistical theory and applications
• Entropic optimal transport is maximum-likelihood deconvolution
• Faster than the fast Legendre transform, the linear-time Legendre transform
• Fréchet means and Procrustes analysis in Wasserstein space
• Geodesic PCA in the Wasserstein space by convex PCA
• Geodesic PCA versus Log-PCA of Histograms in the Wasserstein Space
• scientific article; zbMATH DE number 3647949 (Why is no real title available?)
• scientific article; zbMATH DE number 1465030 (Why is no real title available?)
• scientific article; zbMATH DE number 1909499 (Why is no real title available?)
• scientific article; zbMATH DE number 3099866 (Why is no real title available?)
• Improved rates for Wasserstein deconvolution with ordinary smooth error in dimension one
• M-estimation using penalties or sieves
• Machine learning. A probabilistic perspective
• Mathematical foundations of infinite-dimensional statistical models
• Minimax rates of convergence for Wasserstein deconvolution with supersmooth errors in any dimension
• Nonparametric Validation of Similar Distributions and Assessment of Goodness of Fit
• On a Problem of Adaptive Estimation in Gaussian White Noise
• On Hölder continuity-in-time of the optimal transport map towards measures along a curve
• On optimal matching of Gaussian samples
• On the convergence rate of potentials of Brenier maps
• On the optimal mapping of distributions
• Optimal rates of estimation for multi-reference alignment
• Optimal Transport
• Optimal transport for applied mathematicians. Calculus of variations, PDEs, and modeling
• Oracle inequalities in empirical risk minimization and sparse recovery problems. École d'Été de Probabilités de Saint-Flour XXXVIII-2008.
• Polar factorization and monotone rearrangement of vector‐valued functions
• Procrustes metrics on covariance operators and optimal transportation of Gaussian processes
• Quasiconformal mappings and extendability of functions in Sobolev spaces
• Sharp asymptotic and finite-sample rates of convergence of empirical measures in Wasserstein distance
• The Regularity of Mappings with a Convex Potential
• Uncoupled isotonic regression via minimum Wasserstein deconvolution
• Upper and lower bounds for stochastic processes. Modern methods and classical problems
• Wasserstein dictionary learning: optimal transport-based unsupervised nonlinear dictionary learning
• Wasserstein discriminant analysis
• Wavelets, approximation, and statistical applications

Cited in

(43)

• Minimax rate for optimal transport regression between distributions
• Multivariate ranks and quantiles using optimal transport: consistency, rates and nonparametric testing
• Estimation of Wasserstein distances in the spiked transport model
• On the convergence rate of potentials of Brenier maps
• Minimizing optimal transport for functions with fixed-size nodal sets
• An entropic generalization of Caffarelli's contraction theorem via covariance inequalities
• Sharp convergence rates for empirical optimal transport with smooth costs
• An Improved Central Limit Theorem and Fast Convergence Rates for Entropic Transportation Costs
• Uniform confidence band for optimal transport map on one-dimensional data
• Limit theorems for entropic optimal transport maps and Sinkhorn divergence
• Optimal estimation of smooth transport maps with kernel SoS
• Augmented projection Wasserstein distances: multi-dimensional projection with neural surface
• Projection-based techniques for high-dimensional optimal transport problems
• Plugin estimation of smooth optimal transport maps
• The transport map computed by iterated function system
• Monge-Kantorovich superquantiles and expected shortfalls with applications to multivariate risk measurements
• Background modeling for double Higgs boson production: density ratios and optimal transport
• Scaling optimal transport to high-dimensional Gaussian distributions with application to domain adaptation
• Optimal sequencing depth for single-cell RNA-sequencing in Wasserstein space
• Strong c-concavity and stability in optimal transport
• Stochastic optimal transport in Banach spaces for regularized estimation of multivariate quantiles
• Neural triangular map for density estimation and sampling with application to Bayesian inference
• Rates of convergence for nonparametric estimation of singular distributions using generative adversarial networks
• Constrained approximate optimal transport maps
• Plug-in estimation of Schrödinger bridges
• Diffeomorphic measure matching with kernels for generative modeling
• Optimal transport map estimation in general function spaces
• Wasserstein F-tests for Fréchet regression on Bures-Wasserstein manifolds
• On the Wasserstein Median of Probability Measures
• Approximation rates of entropic maps in semidiscrete optimal transport
• Unbalanced Kantorovich-Rubinstein distance, plan, and barycenter on finite spaces: a statistical perspective
• Normalizing flows as approximations of optimal transport maps via linear-control neural ODEs
• Conditional optimal transport on function spaces
• An approximation theory framework for measure-transport sampling algorithms
• A new perspective on denoising based on optimal transport
• On the sample complexity of entropic optimal transport
• Stochastic gradient descent for barycenters in Wasserstein space
• Stability and statistical inference for semidiscrete optimal transport maps
• Multiple-Attribute Lorenz Functions and Gini Indices: A Measure Transportation Approach
• Algorithms for mean-field variational inference via polyhedral optimization in the Wasserstein space
• Stability bounds for smooth optimal transport maps and their statistical implications
• On the robustness of semi-discrete optimal transport
• New algorithmic directions in optimal transport and applications for product spaces

Describes a project that uses

Uses Software

• PMTK
• Wasserstein GAN
• na13

This page was built for publication: Minimax estimation of smooth optimal transport maps

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2039809)