Constructing free-energy approximations and generalized belief propagation algorithms

@article{Yedidia2005ConstructingFA,
  title={Constructing free-energy approximations and generalized belief propagation algorithms},
  author={Jonathan S. Yedidia and William T. Freeman and Yair Weiss},
  journal={IEEE Transactions on Information Theory},
  year={2005},
  volume={51},
  pages={2282-2312}
}
Important inference problems in statistical physics, computer vision, error-correcting coding theory, and artificial intelligence can all be reformulated as the computation of marginal probabilities on factor graphs. The belief propagation (BP) algorithm is an efficient way to solve these problems that is exact when the factor graph is a tree, but only approximate when the factor graph has cycles. We show that BP fixed points correspond to the stationary points of the Bethe approximation of the… CONTINUE READING

Citations

Publications citing this paper.
SHOWING 1-10 OF 1,063 CITATIONS

ISI/ITI Turbo Equalizer for TDMR Using Trained Local Area Influence Probabilistic Model

  • IEEE Transactions on Magnetics
  • 2019
VIEW 13 EXCERPTS
CITES BACKGROUND
HIGHLY INFLUENCED

Equilibrium and Dynamics on Complex Networks

VIEW 168 EXCERPTS
CITES METHODS & BACKGROUND
HIGHLY INFLUENCED

Monte Carlo Methods for the Ferromagnetic Potts Model Using Factor Graph Duality

  • IEEE Transactions on Information Theory
  • 2015
VIEW 8 EXCERPTS
CITES BACKGROUND & METHODS
HIGHLY INFLUENCED

Sentiment Propagation via Implicature Constraints

VIEW 8 EXCERPTS
CITES METHODS
HIGHLY INFLUENCED

New Understanding of the Bethe Approximation and the Replica Method

  • ArXiv
  • 2013
VIEW 4 EXCERPTS
CITES BACKGROUND
HIGHLY INFLUENCED

FILTER CITATIONS BY YEAR

2003
2019

CITATION STATISTICS

  • 194 Highly Influenced Citations

  • Averaged 51 Citations per year from 2017 through 2019