Abstract:To avoid poor empirical performance in Metropolis-Hastings and other accept-reject-based algorithms practitioners often tune them by trial and error. Lower bounds on the convergence rate are developed in both total variation and Wasserstein distances in order to identify how the simulations will fail so these settings can be avoided, providing guidance on tuning. Particular attention is paid to using the lower bounds to study the convergence complexity of accept-reject-based Markov chains and to constrain the rate of convergence for geometrically ergodic Markov chains. The theory is applied in several settings. For example, if the target density concentrates with a parameter n (e.g. posterior concentration, Laplace approximations), it is demonstrated that the convergence rate of a Metropolis-Hastings chain can be arbitrarily slow if the tuning parameters do not depend carefully on n. This is demonstrated with Bayesian logistic regression with Zellner's g-prior when the dimension and sample increase together and flat prior Bayesian logistic regression as n tends to infinity.

Subgeometric Rates of Convergence for a Class of Continuous-Time Markov Process

Rate Of Convergence For The Maximum Likelihood Estimator In Fractional Ornstein-Uhlenbeck Processes

On Bernstein Type Inequalities for Stochastic Integrals of Multivariate Point Processes

A Supplement to the Complete Convergence

Exponential Convergence Rates for Markov Processes with Markovian Switching

Wasserstein Convergence Rate for Empirical Measures of Markov Processes

Convergence Rate for a Class of Supercritical Superprocesses

Quasi-stationarity and Quasi-Ergodicity of General Markov Processes

Subexponential lower bounds for $f$-ergodic Markov processes

Quantitative Convergence Rates for Stochastically Monotone Markov Chains

On the rate of convergence for an α-stable central limit theorem under sublinear expectation

Convergence Rates in the Law of Large Numbers for Long-Range Dependent Linear Processes

Lower bounds on the rate of convergence for accept-reject-based Markov chains in Wasserstein and total variation distances

SUBGEOMETRIC RATES OF CONVERGENCE OF THE GI/G/1 QUEUEING SYSTEM

Wasserstein Convergence for Empirical Measures of Subordinated Fractional Brownian Motions on the Flat Torus

Convergence Rates of Markov Chains on Spaces of Partitions

On a class of stochastic partial differential equations with multiple invariant measures

On absolutely continuous curves in the Wasserstein space over R and their representation by an optimal Markov process

L2-Geometric Convergence for Time Discrete Markov Processes

Risk-Sensitive Discounted Continuous-Time Markov Decision Processes with Unbounded Rates.

Convergence in the $p$-contest