English

A Spectral Approach to Analyzing Belief Propagation for 3-Coloring

Computational Complexity 2017-11-17 v1 Artificial Intelligence Discrete Mathematics

Abstract

Contributing to the rigorous understanding of BP, in this paper we relate the convergence of BP to spectral properties of the graph. This encompasses a result for random graphs with a ``planted'' solution; thus, we obtain the first rigorous result on BP for graph coloring in the case of a complex graphical structure (as opposed to trees). In particular, the analysis shows how Belief Propagation breaks the symmetry between the 3!3! possible permutations of the color classes.

Keywords

Cite

@article{arxiv.0712.0171,
  title  = {A Spectral Approach to Analyzing Belief Propagation for 3-Coloring},
  author = {Amin Coja-Oghlan and Elchanan Mossel and Dan Vilenchik},
  journal= {arXiv preprint arXiv:0712.0171},
  year   = {2017}
}
R2 v1 2026-06-21T09:49:34.746Z