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.

Commuting time geometry of ergodic Markov chains

Equivalences of Geometric Ergodicity of Markov Chains

On strong stationary times and approximation of Markov chain hitting times by geometric sums

Geometric ergodicity of trans-dimensional Markov chain Monte Carlo algorithms

Graph-combinatorial approach for large deviations of Markov chains

Convergence Rates of Markov Chains on Spaces of Partitions

Scaling limits of discrete-time Markov chains and their local times on electrical networks

Nonhomogeneous quantum Markov chains and a notion of ergodicity

Geometry and Duality of Alternating Markov Chains

Operator-Theoretical Treatment of Ergodic Theorem and Its Application to Dynamic Models in Economics

Geometric Ergodicity and the Spectral Gap of Non-Reversible Markov Chains

Statistical estimation of ergodic Markov chain kernel over discrete state space

Compressibility and Stochastic Stability of Monotone Markov Chain

Geometric theory of (extended) time-reversal symmetries in stochastic processes: I. Finite dimension

A metric on directed graphs and Markov chains based on hitting probabilities

Accurately Computing Expected Visiting Times and Stationary Distributions in Markov Chains

Ergodic properties of a simple deterministic traffic flow model re(al)visited

Estimating the Mixing Time of Ergodic Markov Chains

Quantitative contraction rates for Markov chains on general state spaces

Elementary Bounds On Mixing Times for Decomposable Markov Chains

Stationary measures of continuous time Markov chains with applications to stochastic reaction networks