Abstract:We address the problem of estimating the mixing time $t_{\mathsf{mix}}$ of an arbitrary ergodic finite-state Markov chain from a single trajectory of length $m$. The reversible case was addressed by Hsu et al. [2019], who left the general case as an open problem. In the reversible case, the analysis is greatly facilitated by the fact that the Markov operator is self-adjoint, and Weyl's inequality allows for a dimension-free perturbation analysis of the empirical eigenvalues. As Hsu et al. point out, in the absence of reversibility (which induces asymmetric pair probabilities matrices), the existing perturbation analysis has a worst-case exponential dependence on the number of states $d$. Furthermore, even if an eigenvalue perturbation analysis with better dependence on $d$ were available, in the non-reversible case the connection between the spectral gap and the mixing time is not nearly as straightforward as in the reversible case. Our key insight is to estimate the pseudo-spectral gap $\gamma_{\mathsf{ps}}$ instead, which allows us to overcome the loss of symmetry and to achieve a polynomial dependence on the minimal stationary probability $\pi_\star$ and $\gamma_{\mathsf{ps}}$. Additionally, in the reversible case, we obtain simultaneous nearly (up to logarithmic factors) minimax rates in $t_{\mathsf{mix}}$ and precision $\varepsilon$, closing a gap in Hsu et al., who treated $\varepsilon$ as constant in the lower bounds. Finally, we construct fully empirical confidence intervals for $\gamma_{\mathsf{ps}}$, which shrink to zero at a rate of roughly $1/\sqrt{m}$, and improve the state of the art in even the reversible case.

Mixing under Monotone Censoring

Cutoff for mixtures of permuted Markov chains: general case

Estimating the Mixing Time of Ergodic Markov Chains

Bounds on Lifting Continuous Markov Chains to Speed Up Mixing

Cutoff for mixtures of permuted Markov chains: reversible case

Bounds on Mixing Time for Time-Inhomogeneous Markov Chains

On mixtures of copulas and mixing coefficients

Convergence Rates of Markov Chains on Spaces of Partitions

Elementary Bounds On Mixing Times for Decomposable Markov Chains

Empirical and Instance-Dependent Estimation of Markov Chain and Mixing Time

Lifting Markov Chains To Mix Faster: Limits and Opportunities

A Point on Discrete versus Continuous State-Space Markov Chains

Optimistic Estimation of Convergence in Markov Chains with the Average-Mixing Time

Fast Conditional Mixing of MCMC Algorithms for Non-log-concave Distributions

Mixing Times and Moving Targets

Mixing of Metropolis-Adjusted Markov Chains via Couplings: The High Acceptance Regime

Mixing Time and Temporal Starvation of General CSMA Networks with Multiple Frequency Agility.

On Mixing of Markov Chains: Coupling, Spectral Independence, and Entropy Factorization

Intersection and mixing times for reversible chains

Mixing time and expansion of non-negatively curved Markov chains

Drift and Monotonicity Conditions for Continuous-Time Controlled Markov Chains with an Average Criterion.