Simulated Tempering and Swapping on Mean-Field Models

Simulated and parallel tempering are families of Markov Chain Monte Carlo algorithms where a temperature parameter is varied during the simulation to overcome bottlenecks to convergence due to multimodality. In this work we introduce and analyze the convergence for a set of new tempering distributions which we call entropy dampening. For asymmetric exponential distributions and the mean field Ising model with an external field simulated tempering is known to converge slowly. We show that tempering with entropy dampening distributions mixes in polynomial time for these models. Examining slow mixing times of tempering more closely, we show that for the mean-field 3-state ferromagnetic Potts model, tempering converges slowly regardless of the temperature schedule chosen. On the other hand, tempering with entropy dampening distributions converges in polynomial time to stationarity. Finally we show that the slow mixing can be very expensive practically. In particular, the mixing time of simulated tempering is an exponential factor longer than the mixing time at the fixed temperature.

[1]  Evdokia Nikolova,et al.  On the Hardness and Smoothed Complexity of Quasi-Concave Minimization , 2007, 48th Annual IEEE Symposium on Foundations of Computer Science (FOCS'07).

[2]  Neal Madras,et al.  Analysis of swapping and tempering monte carlo algorithms , 1999 .

[3]  Wang,et al.  Nonuniversal critical dynamics in Monte Carlo simulations. , 1987, Physical review letters.

[4]  Santosh S. Vempala,et al.  Simulated annealing in convex bodies and an O*(n/sup 4/) volume algorithm , 2003, 44th Annual IEEE Symposium on Foundations of Computer Science, 2003. Proceedings..

[5]  Dirk P. Kroese,et al.  Simulation and the Monte Carlo Method (Wiley Series in Probability and Statistics) , 1981 .

[6]  Eric Vigoda,et al.  A polynomial-time approximation algorithm for the permanent of a matrix with nonnegative entries , 2004, JACM.

[7]  Martin E. Dyer,et al.  On Counting Independent Sets in Sparse Graphs , 2002, SIAM J. Comput..

[8]  M. Jerrum,et al.  The Swendsen–Wang Process Does Not Always Mix Rapidly , 1999 .

[9]  Nando de Freitas,et al.  An Introduction to MCMC for Machine Learning , 2004, Machine Learning.

[10]  Christian P. Robert,et al.  Monte Carlo Statistical Methods (Springer Texts in Statistics) , 2005 .

[11]  Santosh S. Vempala,et al.  Simulated annealing in convex bodies and an O*(n4) volume algorithm , 2006, J. Comput. Syst. Sci..

[12]  Mark Jerrum,et al.  The Swendsen-Wang process does not always mix rapidly , 1997, STOC '97.

[13]  Eric Vigoda,et al.  Swendsen‐Wang algorithm on the mean‐field Potts model , 2015, APPROX-RANDOM.

[14]  Dana Randall,et al.  Slow mixing of glauber dynamics via topological obstructions , 2006, SODA '06.

[15]  F. Papangelou.  GIBBS MEASURES AND PHASE TRANSITIONS (de Gruyter Studies in Mathematics 9) , 1990 .

[16]  Hans-Otto Georgii,et al.  Gibbs Measures and Phase Transitions , 1988 .

[17]  C. Geyer,et al.  Annealing Markov chain Monte Carlo with applications to ancestral inference , 1995 .

[18]  R. B. Potts.  Some generalized order-disorder transformations , 1952, Mathematical Proceedings of the Cambridge Philosophical Society.

[19]  Dirk P. Kroese,et al.  Simulation and the Monte Carlo method , 1981, Wiley series in probability and mathematical statistics.

[20]  D. Randall,et al.  Markov chain decomposition for convergence rate analysis , 2002 .

[21]  Alan M. Frieze,et al.  Torpid mixing of some Monte Carlo Markov chain algorithms in statistical physics , 1999, 40th Annual Symposium on Foundations of Computer Science (Cat. No.99CB37039).

[22]  A. Sokal,et al.  Bounds on the ² spectrum for Markov chains and Markov processes: a generalization of Cheeger’s inequality , 1988 .

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

[24]  D. Woodard,et al.  Conditions for Torpid Mixing of Parallel and Simulated Tempering on Multimodal Distributions , 2022 .

[25]  C. Geyer.  Markov Chain Monte Carlo Maximum Likelihood , 1991 .

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

[27]  David J. Aldous,et al.  Lower bounds for covering times for reversible Markov chains and random walks on graphs , 1989 .

[28]  Martin Dyer,et al.  Mixing properties of the Swendsen–Wang process on the complete graph and narrow grids , 2000 .

[29]  N. Madras,et al.  Factoring graphs to bound mixing rates , 1996, Proceedings of 37th Conference on Foundations of Computer Science.

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

[31]  Mark Jerrum,et al.  Approximate Counting, Uniform Generation and Rapidly Mixing Markov Chains , 1987, International Workshop on Graph-Theoretic Concepts in Computer Science.

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

[33]  Dana Randall,et al.  Sampling adsorbing staircase walks using a new Markov chain decomposition method , 2000, Proceedings 41st Annual Symposium on Foundations of Computer Science.

[34]  Marek Biskup,et al.  Rigorous Analysis of Discontinuous Phase Transitions via Mean-Field Bounds , 2003 .

[35]  Persi Diaconis,et al.  What Do We Know about the Metropolis Algorithm? , 1998, J. Comput. Syst. Sci..

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

[37]  D. Woodard,et al.  Sufficient Conditions for Torpid Mixing of Parallel and Simulated Tempering , 2009 .

[38]  Neal Madras,et al.  On the swapping algorithm , 2003, Random Struct. Algorithms.

[39]  Nayantara Bhatnagar,et al.  Annealing and tempering for sampling and counting , 2007 .

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

[41]  Dana Randall,et al.  Torpid mixing of simulated tempering on the Potts model , 2004, SODA '04.

[42]  Eric Vigoda,et al.  A polynomial-time approximation algorithm for the permanent of a matrix with non-negative entries , 2001, STOC '01.

[43]  Yuval Peres,et al.  Mixing Time Power Laws at Criticality , 2007, 48th Annual IEEE Symposium on Foundations of Computer Science (FOCS'07).

[44]  Persi Diaconis,et al.  What do we know about the Metropolis algorithm? , 1995, STOC '95.

[45]  G. Parisi,et al.  Simulated tempering: a new Monte Carlo scheme , 1992, hep-lat/9205018.

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

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

[48]  D. Woodard,et al.  Conditions for Rapid and Torpid Mixing of Parallel and Simulated Tempering on Multimodal Distributions , 2009, 0906.2341.

[49]  L. Thomas,et al.  Bound on the mass gap for finite volume stochastic ising models at low temperature , 1989 .