• Media type: Electronic Conference Proceeding; Text; E-Article
  • Title: Belief Propagation on Replica Symmetric Random Factor Graph Models
  • Contributor: Coja-Oghlan, Amin [Author]; Perkins, Will [Author]
  • Published: Schloss Dagstuhl – Leibniz-Zentrum für Informatik, 2016
  • Language: English
  • DOI: https://doi.org/10.4230/LIPIcs.APPROX-RANDOM.2016.27
  • Keywords: Bethe Free Energy ; Random k-SAT ; Belief Propagation ; Gibbs distributions
  • Origination:
  • Footnote: Diese Datenquelle enthält auch Bestandsnachweise, die nicht zu einem Volltext führen.
  • Description: According to physics predictions, the free energy of random factor graph models that satisfy a certain "static replica symmetry" condition can be calculated via the Belief Propagation message passing scheme [Krzakala et al. PNAS, 2007]. Here we prove this conjecture for a wide class of random factor graph models. Specifically, we show that the messages constructed just as in the case of acyclic factor graphs asymptotically satisfy the Belief Propagation equations and that the free energy density is given by the Bethe free energy formula.
  • Access State: Open Access