A curiously slowly mixing Markov chain

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Diaconis, Persi, Lin, Andrew, Ram, Arun
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909977320357888
author Diaconis, Persi
Lin, Andrew
Ram, Arun
author_facet Diaconis, Persi
Lin, Andrew
Ram, Arun
contents We study a Markov chain with very different mixing rates depending on how mixing is measured. The chain is the "Burnside process on the hypercube $C_2^n$." Started at the all-zeros state, it mixes in a bounded number of steps, no matter how large $n$ is, in $\ell^1$ and in $\ell^2$. And started at general $x$, it mixes in at most $\log n$ steps in $\ell^1$. But, in $\ell^2$, it takes $\frac{n}{\log n}$ steps for most starting $x$. The $\ell^2$ mixing results follow from an explicit diagonalization of the Markov chain into binomial-coefficient-valued eigenvectors.
format Preprint
id arxiv_https___arxiv_org_abs_2511_01245
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle A curiously slowly mixing Markov chain
Diaconis, Persi
Lin, Andrew
Ram, Arun
Probability
Combinatorics
Representation Theory
60J10 (Primary) 05E18 (Secondary)
We study a Markov chain with very different mixing rates depending on how mixing is measured. The chain is the "Burnside process on the hypercube $C_2^n$." Started at the all-zeros state, it mixes in a bounded number of steps, no matter how large $n$ is, in $\ell^1$ and in $\ell^2$. And started at general $x$, it mixes in at most $\log n$ steps in $\ell^1$. But, in $\ell^2$, it takes $\frac{n}{\log n}$ steps for most starting $x$. The $\ell^2$ mixing results follow from an explicit diagonalization of the Markov chain into binomial-coefficient-valued eigenvectors.
title A curiously slowly mixing Markov chain
topic Probability
Combinatorics
Representation Theory
60J10 (Primary) 05E18 (Secondary)
url https://arxiv.org/abs/2511.01245