Abstract:The martingale posterior framework is a generalization of Bayesian inference where one elicits a sequence of one-step ahead predictive densities instead of the likelihood and prior. Posterior sampling then involves the imputation of unseen observables, and can then be carried out in an expedient and parallelizable manner using predictive resampling without requiring Markov chain Monte Carlo. Recent work has investigated the use of plug-in parametric predictive densities, combined with stochastic gradient descent, to specify a parametric martingale posterior. This paper investigates the asymptotic properties of this class of parametric martingale posteriors. In particular, two central limit theorems based on martingale limit theory are introduced and applied. The first is a predictive central limit theorem, which enables a significant acceleration of the predictive resampling scheme through a hybrid sampling algorithm based on a normal approximation. The second is a Bernstein-von Mises result, which is novel for martingale posteriors, and provides methodological guidance on attaining desirable frequentist properties. We demonstrate the utility of the theoretical results in simulations and a real data example.

On the Almost Sure Central Limit Theorem for Vector Martingales: Convergence of Moments and Statistical Applications

Lindeberg's Central Limit Theorems for Martingale-Like Sequences under Sub-Linear Expectations

Functional Shige Peng’s Central Limit Theorems for Martingale Vectors

Central limit theorems for martingales-II: convergence in the weak dual topology

On Bernstein Type Inequalities for Stochastic Integrals of Multivariate Point Processes

New insights on concentration inequalities for self-normalized martingales

Limit Theorems and Inequalities Via Martingale Methods

Rates in the central limit theorem for random projections of Martingales

On the almost sure central limit theorem for the elephant random walk

Central limit theorems for martingales-I : continuous limits

On the tradeoff between almost sure error tolerance versus mean deviation frequency in martingale convergence

Exact Rates of Convergence in Some Martingale Central Limit Theorems

On stable central limit theorems for multivariate discrete-time martingales

Limiting Behavior of Moving Average Processes of Martingale Difference Sequences

Small time central limit theorems for semimartingales with applications

Functional central limit theorems for random vectors under sub-linear expectations

On the rate of convergence of the martingale central limit theorem in Wasserstein distances

Convergence Rates in the Strong Law of Large Numbers for Martingale Difference Sequences

The Convergence of Double-Indexed Weighted Sums of Martingale Differences and Its Application

PR ] 2 F eb 2 02 0 A Berry-Esseen bound of order 1 / √ n for martingales

Asymptotics for parametric martingale posteriors