Privacy-Preserving Belief Propagation and Sampling

We provide provably privacy-preserving versions of belief propagation, Gibbs sampling, and other local algorithms -- distributed multiparty protocols in which each party or vertex learns only its final local value, and absolutely nothing else.