Work

Restart Perturbations for Reversible Markov Chains: Trichotomy and Pre‐Cutoff Equivalence

Year: 2025

Type: article

Abstract: ABSTRACT Given a reversible Markov chain on states, and another chain obtained by perturbing each row of by at most in total variation, we study the total variation distance between the two stationary distributions, . We show that for chains with cutoff, converges to 0, is asymptotically at most (with a sequence of perturbations for which convergence to this bound occurs), and converges to 1, respectively, if the product of and the mixing time of converges to 0, , and , respectively. This echoes recent results for specific random walks that exhibit cutoff, suggesting that cutoff is the key property underlying such results. Moreover, we show is maximized by restart perturbations, for which “restarts” at a random state with probability at each step. Finally, we show that pre‐cutoff is (almost) equivalent to a notion of “sensitivity to restart perturbations,” suggesting that chains with sharper convergence to stationarity are inherently less robust.

Source: Random Structures and Algorithms

DOI: 10.1002/rsa.70007

Authors: Daniel Vial, Vijay Subramanian

Institution: University of Michigan

Language: English

Topic: Markov Chains and Monte Carlo Methods

Subfield: Statistics and Probability

Field: Mathematics

Domain: Physical Sciences

Open Access status: hybrid

Estimated APC paid: $4,550

Funder: National Science Foundation