On the validity of the formal Edgeworth expansion for posterior densities

DOI10.1214/19-AOS1871MaRDI QIDQ2215719zbMATH OpenFDO

Publication date 14 December 2020

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

Full work available at URL https://arxiv.org/abs/1710.01871, https://projecteuclid.org/euclid.aos/1597370659

cumulant expansion Edgeworth expansion higher-order asymptotics posterior

Central limit and other weak theorems (60F05) Asymptotic distribution theory in statistics (62E20) Bayesian inference (62F15)

Abstract: We consider a fundamental open problem in parametric Bayesian theory, namely the validity of the formal Edgeworth expansion of the posterior density. While the study of valid asymptotic expansions for posterior distributions constitutes a rich literature, the validity of the formal Edgeworth expansion has not been rigorously established. Several authors have claimed connections of various posterior expansions with the classical Edgeworth expansion, or have simply assumed its validity. Our main result settles this open problem. We also prove a lemma concerning the order of posterior cumulants which is of independent interest in Bayesian parametric theory. The most relevant literature is synthesized and compared to the newly-derived Edgeworth expansions. Numerical investigations illustrate that our expansion has the behavior expected of an Edgeworth expansion, and that it has better performance than the other existing expansion which was previously claimed to be of Edgeworth-type.

Review text

The authors establish validity of a formal Edgeworth expansion for the posterior density in a Bayesian setting. Under some regularity conditions (including, for example, differentiability, continuity and measurability conditions), they establish the order of the error in approximating the posterior density of \((\theta-\theta_0)/\sigma\) by an Edgeworth expansion involving Hermite polynomials and the cumulants of this statistic, where \(\theta\) is the relevant parameter, and \(\theta_0\) and \(\sigma\) are its posterior mean and standard deviation, respectively. A similar result is also established for the posterior distribution function. These results require estimates of the asymptotic order of posterior cumulants which the authors establish. The paper concludes with a comparison of the authors' results with other, existing expansions (including those in which \(\theta\) is normalised using the MLE and observed information, for example) and numerical illustrations of the present results.

Recommendations

Cites work

Cited in

(11)

This page was built for publication: On the validity of the formal Edgeworth expansion for posterior densities

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2215719)