Adaptive Bernstein-von Mises theorems in Gaussian white noise

Abstract

We investigate Bernstein-von Mises theorems for adaptive nonparametric Bayesian procedures in the canonical Gaussian white noise model. We consider both a Hilbert space and multiscale setting with applications in $L^2$ and $L^{\infty }$ respectively. This provides a theoretical justification for plug-in procedures, for example the use of certain credible sets for sufficiently smooth linear functionals. We use this general approach to construct optimal frequentist confidence sets based on the posterior distribution. We also provide simulations to numerically illustrate our approach and obtain a visual representation of the geometries involved.

48 pages, 5 figures

References in corpus (14)

• Bayesian linear regression with sparse priors
• Needles and Straw in a Haystack: Posterior concentration for possibly sparse sequences
• Confidence bands in density estimation
• On adaptive inference and confidence bands
• Lower bounds for posterior rates with Gaussian process priors
• Uniform limit theorems for wavelet density estimators
• A Bernstein-Von Mises Theorem for discrete probability distributions
• A complement to Le Cam's theorem
• The Le Cam distance between density estimation, Poisson processes and Gaussian white noise
• Discussion of "Frequentist coverage of adaptive nonparametric Bayesian credible sets"
• Rejoinder to discussions of "Frequentist coverage of adaptive nonparametric Bayesian credible sets"
• Discussion of "Frequentist coverage of adaptive nonparametric Bayesian credible sets"
• A sharp adaptive confidence ball for self-similar functions
• Discussion of "Frequentist coverage of adaptive nonparametric Bayesian credible sets"

Cited by in corpus (21)

• Frequentist Consistency of Variational Bayes
• Contributed Discussion to Uncertainty Quantification for the Horseshoe by Stéphanie van der Pas, Botond Szabó and Aad van der Vaart
• Variational Bayes for high-dimensional linear regression with sparse priors
• Supremum Norm Posterior Contraction and Credible Sets for Nonparametric Multivariate Regression
• Nonparametric statistical inference for drift vector fields of multi-dimensional diffusions
• Frequentist coverage and sup-norm convergence rate in Gaussian process regression
• Bernstein -- von Mises theorems for statistical inverse problems II: Compound Poisson processes
• Posterior consistency and convergence rates for Bayesian inversion with hypoelliptic operators
• Spike and slab variational Bayes for high dimensional logistic regression
• Learning the smoothness of noisy curves with application to online curve estimation
• On the Bernstein-von Mises theorem for the Dirichlet process
• On frequentist coverage errors of Bayesian credible sets in moderately high dimensions
• The Bayes Lepski's Method and Credible Bands through Volume of Tubular Neighborhoods
• On Semi-parametric Bernstein-von Mises Theorems for BART
• Variable Selection Consistency of Gaussian Process Regression
• Spike and Slab Pólya tree posterior distributions: adaptive inference
• Posterior Contraction and Credible Sets for Filaments of Regression Functions
• Discussion of "Frequentist coverage of adaptive nonparametric Bayesian credible sets"
• Uncertainty quantification for distributed regression
• Semi-parametric Bernstein-von Mises in Linear Inverse Problems
• Bernstein-von Mises Theorems for Functionals of Covariance Matrix