Asymptotic normality, concentration, and coverage of generalized posteriors

07/22/2019
by   Jeffrey W. Miller, et al.
0

Generalized likelihoods are commonly used to obtain consistent estimators with attractive computational and robustness properties. Formally, any generalized likelihood can be used to define a generalized posterior distribution, but an arbitrarily defined "posterior" cannot be expected to appropriately quantify uncertainty in any meaningful sense. In this article, we provide sufficient conditions under which generalized posteriors exhibit concentration, asymptotic normality (Bernstein-von Mises), an asymptotically correct Laplace approximation, and asymptotically correct frequentist coverage. We apply our results in detail to generalized posteriors for a wide array of generalized likelihoods, including pseudolikelihoods in general, the Ising model pseudolikelihood, the Gaussian Markov random field pseudolikelihood, the fully observed Boltzmann machine pseudolikelihood, the Cox proportional hazards partial likelihood, and a median-based likelihood for robust inference of location. Further, we show how our results can be used to easily establish the asymptotics of standard posteriors for exponential families and generalized linear models. We make no assumption of model correctness so that our results apply with or without misspecification.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
05/09/2019

Conformal prediction for exponential families and generalized linear models

Conformal prediction methods construct prediction regions for iid data t...
research
06/13/2019

Hypotheses testing and posterior concentration rates for semi-Markov processes

In this paper, we adopt a nonparametric Bayesian approach and investigat...
research
12/06/2013

A sequential reduction method for inference in generalized linear mixed models

The likelihood for the parameters of a generalized linear mixed model in...
research
03/09/2021

Asymptotic posterior normality of the generalized extreme value distribution

The univariate generalized extreme value (GEV) distribution is the most ...
research
12/05/2021

Laplace Power-expected-posterior priors for generalized linear models with applications to logistic regression

Power-expected-posterior (PEP) methodology, which borrows ideas from the...
research
07/04/2023

Asymptotics for the Generalized Autoregressive Conditional Duration Model

Engle and Russell (1998, Econometrica, 66:1127–1162) apply results from ...

Please sign up or login with your details

Forgot password? Click here to reset