-
Mutual Information for the Stochastic Block Model by the Adaptive Interpolation Method
Authors:
Jean Barbier,
Chun Lam Chan,
Nicolas Macris
Abstract:
We rigorously derive a single-letter variational expression for the mutual information of the asymmetric two-groups stochastic block model in the dense graph regime. Existing proofs in the literature are indirect, as they involve mapping the model to a rank-one matrix estimation problem whose mutual information is then determined by a combination of methods (e.g., interpolation, cavity, algorithmi…
▽ More
We rigorously derive a single-letter variational expression for the mutual information of the asymmetric two-groups stochastic block model in the dense graph regime. Existing proofs in the literature are indirect, as they involve mapping the model to a rank-one matrix estimation problem whose mutual information is then determined by a combination of methods (e.g., interpolation, cavity, algorithmic, spatial coupling). In this contribution we provide a self-contained direct method using only the recently introduced adaptive interpolation method.
△ Less
Submitted 16 July, 2019; v1 submitted 19 February, 2019;
originally announced February 2019.
-
Concentration of multi-overlaps for random ferromagnetic spin models
Authors:
Jean Barbier,
Chun Lam Chan,
Nicolas Macris
Abstract:
We consider ferromagnetic spin models on dilute random graphs and prove that, with suitable one-body infinitesimal perturbations added to the Hamiltonian, the multi-overlaps concentrate for all temperatures, both with respect to the thermal Gibbs average and the quenched randomness. Results of this nature have been known only for the lowest order overlaps, at high temperature or on the Nishimori l…
▽ More
We consider ferromagnetic spin models on dilute random graphs and prove that, with suitable one-body infinitesimal perturbations added to the Hamiltonian, the multi-overlaps concentrate for all temperatures, both with respect to the thermal Gibbs average and the quenched randomness. Results of this nature have been known only for the lowest order overlaps, at high temperature or on the Nishimori line. Here we treat all multi-overlaps by a non-trivial application of Griffiths-Kelly-Sherman correlation inequalities. Our results apply in particular to the pure and mixed p-spin ferromagnets on random dilute Erdoes-Rényi hypergraphs. On physical grounds one expects that multi-overlap concentration directly implies the correctness of the cavity (or replica symmetric) formula for the pressure. The proof of this formula for the general p-spin ferromagnet on a random dilute hypergraph remains an open problem.
△ Less
Submitted 19 January, 2019;
originally announced January 2019.
-
Adaptive Path Interpolation for Sparse Systems: Application to a Simple Censored Block Model
Authors:
Jean Barbier,
Chun Lam Chan,
Nicolas Macris
Abstract:
Recently a new adaptive path interpolation method has been developed as a simple and versatile scheme to calculate exactly the asymptotic mutual information of Bayesian inference problems defined on dense factor graphs. These include random linear and generalized estimation, sparse superposition codes, or low-rank matrix and tensor estimation. For all these systems, the adaptive interpolation meth…
▽ More
Recently a new adaptive path interpolation method has been developed as a simple and versatile scheme to calculate exactly the asymptotic mutual information of Bayesian inference problems defined on dense factor graphs. These include random linear and generalized estimation, sparse superposition codes, or low-rank matrix and tensor estimation. For all these systems, the adaptive interpolation method directly proves that the replica symmetric prediction is exact, in a simple and unified manner. When the underlying factor graph of the inference problem is sparse the replica prediction is considerably more complicated, and rigorous results are often lacking or obtained by rather complicated methods. In this work we show how to extend the adaptive path interpolation method to sparse systems. We concentrate on a Censored Block Model, where hidden variables are measured through a binary erasure channel, for which we fully prove the replica prediction.
△ Less
Submitted 18 July, 2019; v1 submitted 13 June, 2018;
originally announced June 2018.
