Skip to main navigation Skip to search Skip to main content

Convergence of MCMC and loopy BP in the tree uniqueness region for the hard-core model

  • Charilaos Efthymiou
  • , Thomas P. Hayes
  • , Daniel Štefankovič
  • , Eric Vigoda
  • , Yitong Yin
  • Goethe University Frankfurt
  • University of Rochester
  • Georgia Institute of Technology
  • Nanjing University

Research output: Contribution to journalArticlepeer-review

14 Scopus citations

Abstract

We study the hard-core (gas) model defined on independent sets of an input graph where the independent sets are weighted by a parameter (aka fugacity) λ > 0. For constant Δ, the previous work of Weitz [Proceedings of STOC, 2006, pp. 140-149] established an FPTAS for the partition function for graphs of maximum degree Δ when λ < λc(Δ). Sly [Proceedings of FOCS, 2010, pp. 287-296] showed that there is no FPRAS, unless NP=RP, when λ > λc(Δ). The threshold λc(Δ) is the critical point for the statistical physics phase transition for uniqueness/nonuniqueness on the infinite Δ-regular tree. The running time of Weitz's algorithm is exponential in log Δ. Here we present an FPRAS for the partition function whose running time is O(n2). We analyze the simple single-site Markov chain known as the Glauber dynamics for sampling from the associated Gibbs distribution. We prove there exists a constant Δ0 such that for all graphs with maximum degree Δ ≥ Δ0 and girth ≥ 7 (i.e., no cycles of length ≤ 6), the mixing time of the Glauber dynamics is O(n log n) when λ < λc(Δ). Our work complements that of Weitz, which applies for small constant Δ, whereas our work applies for all Δ at least a sufficiently large constant Δ0. (This includes Δ depending on n = |V |.) Our proof utilizes loopy belief propagation (BP) which is a widely used algorithm for inference in graphical models. A novel aspect of our work is using the principal eigenvector for the BP operator to design a distance function which contracts in expectation for pairs of states that behave like the BP fixed point. We also prove that the Glauber dynamics behaves locally like loopy BP. As a byproduct we obtain that the Glauber dynamics, after a short burn-in period, converges close to the BP fixed point, and this implies that the fixed point of loopy BP is a close approximation to the Gibbs distribution. Using these connections we establish that loopy BP quickly converges to the Gibbs distribution when the girth ≥ 6 and λ < λc(Δ).

Original languageEnglish
Pages (from-to)581-643
Number of pages63
JournalSIAM Journal on Computing
Volume48
Issue number2
DOIs
StatePublished - 2019

Keywords

  • Gibbs sampling
  • Hard-core model
  • Loopy belief propagation
  • Markov chain Monte Carlo
  • Rapid mixing

Fingerprint

Dive into the research topics of 'Convergence of MCMC and loopy BP in the tree uniqueness region for the hard-core model'. Together they form a unique fingerprint.

Cite this