On sensitivity of mixing times and cutoff

A sequence of chains exhibits (total-variation) cutoff (resp., pre-cutoff) if for all $0<\epsilon< 1/2$, the ratio $t_{\mathrm{mix}}^{(n)}(\epsilon)/t_{\mathrm{mix}}^{(n)}(1-\epsilon)$ tends to 1 as $n \to \infty $ (resp., the $\limsup$ of this ratio is bounded uniformly in $\epsilon$), where $t_{\mathrm{mix}}^{(n)}(\epsilon)$ is the $\epsilon$-total-variation mixing-time of the $n$th chain in the sequence. We construct a sequence of bounded degree graphs $G_n$, such that the lazy simple random walks (LSRW) on $G_n$ satisfy the "product condition" $\mathrm{gap}(G_n) t_{\mathrm{mix}}^{(n)}(\epsilon) \to \infty $ as $n \to \infty$, where $\mathrm{gap}(G_n)$ is the spectral gap of the LSRW on $G_n$ (a known necessary condition for pre-cutoff that is often sufficient for cutoff), yet this sequence does not exhibit pre-cutoff. Recently, Chen and Saloff-Coste showed that total-variation cutoff is equivalent for the sequences of continuous-time and lazy versions of some given sequence of chains. Surprisingly, we show that this is false when considering separation cutoff. We also construct a sequence of bounded degree graphs $G_n=(V_{n},E_{n})$ that does not exhibit cutoff, for which a certain bounded perturbation of the edge weights leads to cutoff and increases the order of the mixing-time by an optimal factor of $\Theta (\log |V_n|)$. Similarly, we also show that "lumping" states together may increase the order of the mixing-time by an optimal factor of $\Theta (\log |V_n|)$. This gives a negative answer to a question asked by Aldous and Fill.

[1]  Matthew I. Roberts,et al.  Mixing Time Bounds via Bottleneck Sequences , 2016, 1610.07874.

[2]  Matthew I. Roberts,et al.  Robustness of mixing under rough isometry, via bottleneck sequences , 2016 .

[3]  On sensitivity of uniform mixing times , 2016, 1607.01672.

[4]  Y. Peres,et al.  Total variation and separation cutoffs are not equivalent and neither one implies the other , 2015, 1508.03913.

[5]  Y. Peres,et al.  Characterization of cutoff for reversible Markov chains , 2014, SODA.

[6]  Guan-Yu Chen,et al.  Comparison of Cutoffs Between Lazy Walks and Markovian Semigroups , 2013, Journal of Applied Probability.

[7]  Jian Ding,et al.  Sensitivity of mixing times , 2013 .

[8]  V. Climenhaga Markov chains and mixing times , 2013 .

[9]  Perla Sousi,et al.  Mixing Times are Hitting Times of Large Sets , 2011 .

[10]  Allan Sly,et al.  Explicit Expanders with Cutoff Phenomena , 2010, 1003.3515.

[11]  Jian Ding,et al.  Total variation cutoff in birth-and-death chains , 2008, 0801.2625.

[12]  G. Kozma On the precision of the spectral prole , 2007 .

[13]  G. Kozma On the precision of the spectral profile , 2007, 0709.0112.

[14]  B. Reed,et al.  Faster Mixing and Small Bottlenecks , 2006 .

[15]  P. Tetali,et al.  Mixing Time Bounds via the Spectral Profile , 2005, math/0505690.

[16]  Y. Peres,et al.  Evolving sets, mixing and heat kernel bounds , 2003, math/0305349.

[17]  C. Pittet,et al.  On the stability of the behavior of random walks on groups , 2000 .

[18]  P. Diaconis,et al.  LOGARITHMIC SOBOLEV INEQUALITIES FOR FINITE MARKOV CHAINS , 1996 .

[19]  Moshe Morgenstern,et al.  Existence and Explicit Constructions of q + 1 Regular Ramanujan Graphs for Every Prime Power q , 1994, J. Comb. Theory, Ser. B.

[20]  Instability of the Liouville property for quasi-isometric graphs and manifolds of polynomial volume growth , 1991 .

[21]  Mark Jerrum,et al.  Approximate Counting, Uniform Generation and Rapidly Mixing Markov Chains , 1987, WG.

[22]  N. Alon Eigenvalues and expanders , 1986, Comb..

[23]  N. Alon,et al.  il , , lsoperimetric Inequalities for Graphs , and Superconcentrators , 1985 .

[24]  P. Flajolet On approximate counting , 1982 .