Abstract:In Bayesian statistics, posterior contraction rates (PCRs) quantify the speed at which the posterior distribution concentrates on arbitrarily small neighborhoods of a true model, in a suitable way, as the sample size goes to infinity. In this paper, we develop a new approach to PCRs, with respect to strong norm distances on parameter spaces of functions. Critical to our approach is the combination of a local Lipschitz-continuity for the posterior distribution with a dynamic formulation of the Wasserstein distance, which allows to set forth an interesting connection between PCRs and some classical problems arising in mathematical analysis, probability and statistics, e.g., Laplace methods for approximating integrals, Sanov's large deviation principles in the Wasserstein distance, rates of convergence of mean Glivenko–Cantelli theorems, and estimates of weighted Poincaré–Wirtinger constants. We first present a theorem on PCRs for a model in the regular infinite-dimensional exponential family, which exploits sufficient statistics of the model, and then extend such a theorem to a general dominated model. These results rely on the development of novel techniques to evaluate Laplace integrals and weighted Poincaré–Wirtinger constants in infinite-dimension, which are of independent interest. The proposed approach is applied to the regular parametric model, the multinomial model, the finite-dimensional and the infinite-dimensional logistic-Gaussian model and the infinite-dimensional linear regression. In general, our approach leads to optimal PCRs in finite-dimensional models, whereas for infinite-dimensional models it is shown explicitly how the prior distribution affect PCRs.

Limitations of the Wasserstein MDE for univariate data

A minimum Hellinger distance estimator for stochastic differential equations: An application to statistical inference for continuous time interest rate models

Wasserstein-based Minimax Estimation of Dependence in Multivariate Regularly Varying Extremes

Asymptotic Theory of Least Distances Estimate in Multivariate Linear Models

Statistical Inference for Generative Models with Maximum Mean Discrepancy

Sliced Wasserstein Regression

Quantifying Distributional Model Risk in Marginal Problems via Optimal Transport

Wasserstein distance bounds on the normal approximation of empirical autocovariances and cross-covariances under non-stationarity and stationarity

Asymptotic bias of inexact Markov Chain Monte Carlo methods in high dimension

On the Wasserstein Median of Probability Measures

Strong posterior contraction rates via Wasserstein dynamics

Inference for Projection-Based Wasserstein Distances on Finite Spaces

Deficiency bounds for the multivariate inverse hypergeometric distribution

Confidence regions in Wasserstein distributionally robust estimation

Controlling Wasserstein Distances by Kernel Norms with Application to Compressive Statistical Learning

Linear Model Estimators and Consistency under an Infill Asymptotic Domain

Robust Estimation under the Wasserstein Distance

One-dimensional approximation of measures in Wasserstein distances

Minimum profile Hellinger distance estimation for single-index models

Wasserstein multivariate auto-regressive models for modeling distributional time series

Robustness to Model Approximation, Empirical Model Learning, and Sample Complexity in Wasserstein Regular MDPs