Fast convergence of the Glauber dynamics for sampling independent sets

We consider the problem of sampling independent sets of a graph with maximum degree δ. The weight of each independent set is expressed in terms of a fixed positive parameter λ≤2/(δ−2), where the weight of an independent set σ is λ|σ|. The Glauber dynamics is a simple Markov chain Monte Carlo method for sampling from this distribution. We show fast convergence (in O(n log n) time) of this dynamics. This paper gives the more interesting proof for triangle-free graphs. The proof for arbitrary graphs is given in a companion paper (E. Vigoda, Technical Report TR-99-003, International Computer Institute, Berkeley, CA, 1998). We also prove complementary hardness of approximation results, which show that it is hard to sample from this distribution when λ>c/δ for a constant c≤0. ©1999 John Wiley & Sons, Inc. Random Struct. Alg., 15, 229–241, 1999

[1]  S. Zachary,et al.  Loss networks , 2009, 0903.0640.

[2]  Martin E. Dyer,et al.  On Markov Chains for Independent Sets , 2000, J. Algorithms.

[3]  J. Håstad.  Clique is hard to approximate withinn1−ε , 1999 .

[4]  P. Berman,et al.  On Some Tighter Inapproximability Results , 1998, Electron. Colloquium Comput. Complex..

[5]  Dana Randall,et al.  Analyzing Glauber Dynamics by Comparison of Markov Chains , 1998, LATIN.

[6]  Eric Vigoda,et al.  Approximately counting up to four (extended abstract) , 1997, STOC '97.

[7]  Johan Håstad,et al.  Clique is hard to approximate within n/sup 1-/spl epsiv// , 1996, Proceedings of 37th Conference on Foundations of Computer Science.

[8]  David Bruce Wilson,et al.  Exact sampling with coupled Markov chains and applications to statistical mechanics , 1996, Random Struct. Algorithms.

[9]  A. Baram,et al.  Hard square lattice gas , 1994 .

[10]  F. Martinelli,et al.  Approach to equilibrium of Glauber dynamics in the one phase region , 1994 .

[11]  J. Berg,et al.  Percolation and the hard-core lattice gas model , 1994 .

[12]  Alistair Sinclair,et al.  Algorithms for Random Generation and Counting: A Markov Chain Approach , 1993, Progress in Theoretical Computer Science.

[13]  Daniel F. Styer,et al.  Vasserstein distances in two-state systems , 1989 .

[14]  Dan Constantin Radulescu,et al.  The Dobrushin-Shlosman phase uniqueness criterion and applications to hard squares , 1987 .

[15]  Leslie G. Valiant,et al.  Random Generation of Combinatorial Structures from a Uniform Distribution , 1986, Theor. Comput. Sci..

[16]  R. L. Dobrushin,et al.  Phase diagram of the two-dimensional Ising antiferromagnet (computer-assisted proof) , 1985 .

[17]  Richard M. Karp,et al.  Monte-Carlo algorithms for enumeration and reliability problems , 1983, 24th Annual Symposium on Foundations of Computer Science (sfcs 1983).

[18]  Lars Engebretsen,et al.  Clique Is Hard To Approximate Within , 2000 .

[19]  David Wilson,et al.  Coupling from the past: A user's guide , 1997, Microsurveys in Discrete Probability.

[20]  M. Dyer,et al.  Path Coupling, Dobrushin Uniqueness, and Approximate Counting , 1997 .

[21]  Dana Randall,et al.  Markov Chain Algorithms for Planar Lattice Structures (Extended Abstract). , 1995, FOCS 1995.

[22]  F. Martinelliz,et al.  Approach to Equilibrium of Glauber Dynamics in the One Phase Region. Ii: the General Case , 1994 .

[23]  Mihalis Yannakakis,et al.  Optimization, Approximation, and Complexity Classes (Extended Abstract) , 1988, STOC 1988.

[24]  R. Holley,et al.  Rapid Convergence to Equilibrium of Stochastic Ising Models in the Dobrushin Shlosman Regime , 1987 .

[25]  R. Dobrushin,et al.  Constructive Criterion for the Uniqueness of Gibbs Field , 1985 .

[26]  D. Aldous.  Random walks on finite groups and rapidly mixing markov chains , 1983 .

[27]  R. Dobrushin.  The problem of uniqueness of a gibbsian random field and the problem of phase transitions , 1968 .