Polynomial Mixing of the Edge-Flip Markov Chain for Unbiased Dyadic Tilings

We give the first polynomial upper bound on the mixing time of the edge-flip Markov chain for unbiased dyadic tilings, resolving an open problem originally posed by Janson, Randall, and Spencer in 2002. A dyadic tiling of size n is a tiling of the unit square by n non-overlapping dyadic rectangles, each of area 1/n, where a dyadic rectangle is any rectangle that can be written in the form [a2^{-s}, (a+1)2^{-s}] \times [b2^{-t}, (b+1)2^{-t}] for non-negative integers a,b,s,t. The edge-flip Markov chain selects a random edge of the tiling and replaces it with its perpendicular bisector if doing so yields a valid dyadic tiling. Specifically, we show that the relaxation time of the edge-flip Markov chain for dyadic tilings is at most O(n^{4.09}), which implies that the mixing time is at most O(n^{5.09}). We complement this by showing that the relaxation time is at least \Omega(n^{1.38}), improving upon the previously best lower bound of \Omega(n\log n) coming from the diameter of the chain.

[1]  Catherine S. Greenhill The switch Markov chain for sampling irregular graphs (Extended Abstract) , 2014, SODA.

[2]  Fabio Martinelli,et al.  Dynamics of Lattice Triangulations on Thin Rectangles , 2015, ArXiv.

[3]  Alexandre Stauffer A Lyapunov function for Glauber dynamics on lattice triangulations , 2015, ArXiv.

[4]  Y. Peres,et al.  Glauber dynamics for the mean-field Ising model: cut-off, critical power law, and metastability , 2007, 0712.0790.

[5]  N. Metropolis,et al.  Equation of State Calculations by Fast Computing Machines , 1953, Resonance.

[6]  Y. Peres,et al.  Mixing Time of Critical Ising Model on Trees is Polynomial in the Height , 2009, 0901.4152.

[7]  Prasad Tetali,et al.  On the mixing time of the triangulation walk and other Catalan structures , 1997, Randomization Methods in Algorithm Design.

[8]  P. Tetali,et al.  Analyzing Glauber dynamics by comparison of Markov chains , 2000 .

[9]  Martin E. Dyer,et al.  Sampling regular graphs and a peer-to-peer network , 2005, SODA '05.

[10]  Elizabeth L. Wilmer,et al.  Markov Chains and Mixing Times , 2008 .

[11]  Svante Janson,et al.  Random dyadic tilings of the unit square , 2002, Random Struct. Algorithms.

[12]  Jian Ding,et al.  The Mixing Time Evolution of Glauber Dynamics for the Mean-Field Ising Model , 2008, 0806.1906.

[13]  Allan Sly,et al.  Critical Ising on the Square Lattice Mixes in Polynomial Time , 2010, 1001.1613.

[14]  Chee Yap,et al.  Complete subdivision algorithms, II: Isotopic meshing of singular algebraic curves , 2012, J. Symb. Comput..

[15]  Fabio Martinelli,et al.  Random lattice triangulations: structure and algorithms , 2013, STOC '13.

[16]  F. Martinelli Lectures on Glauber dynamics for discrete spin models , 1999 .

[17]  Paul Cuff,et al.  Glauber Dynamics for the Mean-Field Potts Model , 2012, 1204.4503.

[18]  Dana Randall,et al.  Markov chain algorithms for planar lattice structures , 1995, Proceedings of IEEE 36th Annual Foundations of Computer Science.

[19]  P. Diaconis,et al.  COMPARISON THEOREMS FOR REVERSIBLE MARKOV CHAINS , 1993 .

[20]  Dana Randall,et al.  Phase Transitions in Random Dyadic Tilings and Rectangular Dissections , 2015, SODA.

[21]  F. Cesi Quasi-factorization of the entropy and logarithmic Sobolev inequalities for Gibbs random fields , 2001 .

[22]  Robert D. Nowak,et al.  Minimax-optimal classification with dyadic decision trees , 2006, IEEE Transactions on Information Theory.

[23]  Mu-Fa Chen,et al.  Trilogy of Couplings and General Formulas for Lower Bound of Spectral Gap , 1998 .

[24]  Reza Gheissari,et al.  Mixing Times of Critical Two‐Dimensional Potts Models , 2018 .

[25]  Jeffrey C. Lagarias,et al.  Counting dyadic equipartitions of the unit square , 2002, Discret. Math..