Analysis of non-reversible Markov chains via similarity orbit

TL;DR

This paper analyzes non-reversible Markov chains using similarity orbit theory, providing spectral representations, convergence insights, and applications to birth-death processes, expanding understanding of their spectral and probabilistic properties.

Contribution

It introduces a framework for analyzing non-reversible Markov chains via similarity orbits, including spectral analysis and convergence properties, with applications to birth-death processes.

Findings

Spectral representation using non-self-adjoint resolutions.

Conditions for chains to belong to the similarity orbit of birth-death chains.

Analysis of convergence rates and cutoff phenomena.

Abstract

In this paper, we develop an in-depth analysis of non-reversible Markov chains on denumerable state space from a similarity orbit perspective. In particular, we study the class of Markov chains whose transition kernel is in the similarity orbit of a normal transition kernel, such as the one of birth-death chains or reversible Markov chains. We start by identifying a set of sufficient conditions for a Markov chain to belong to the similarity orbit of a birth-death one. As by-products, we obtain a spectral representation in terms of non-self-adjoint resolutions of identity in the sense of Dunford [21] and offer a detailed analysis on the convergence rate, separation cutoff and ${\rm{L}}^2$-cutoff of this class of non-reversible Markov chains. We also look into the problem of estimating the integral functionals from discrete observations for this class. In the last part of this paper, we investigate a particular similarity orbit of reversible Markov kernels, that we call the pure birth orbit, and analyze various possibly non-reversible variants of classical birth-death processes in this orbit.