Resummed mean-field inference for strongly coupled data

Abstract

We present a resummed mean-field approximation for inferring the parameters of an Ising or a Potts model from empirical, noisy, one- and two-point correlation functions. Based on a resummation of a class of diagrams of the small correlation expansion of the log-likelihood, the method outperforms standard mean-field inference methods, even when they are regularized. The inference is stable with respect to sampling noise, contrarily to previous works based either on the small correlation expansion, on the Bethe free energy, or on the mean-field and Gaussian models. Because it is mostly analytic, its complexity is still very low, requiring an iterative algorithm to solve for $N$ auxiliary variables, that resorts only to matrix inversions and multiplications. We test our algorithm on the Sherrington-Kirkpatrick model submitted to a random external field and large random couplings, and demonstrate that even without regularization, the inference is stable across the whole phase diagram. In addition, the calculation leads to a consistent estimation of the entropy of the data and allows us to sample form the inferred distribution to obtain artificial data that are consistent with the empirical distribution.

v1) 5+6 pages, 3+6 figures; v2) slight changes of the text, typos corrected. v3) Expanded version, 21 pages, 14 figures v4) Published version

References in corpus (19)

• Exact evolution equation for the effective potential
• Identification of direct residue contacts in protein-protein interaction by message passing
• Improved contact prediction in proteins: Using pseudolikelihoods to infer Potts models
• High-dimensional Ising model selection using ${\ell }_1$-regularized logistic regression
• The Ising Model for Neural Data: Model Quality and Approximate Methods for Extracting Functional Connectivity
• Fast and accurate multivariate Gaussian modeling of protein families: Predicting residue contacts and protein-interaction partners
• Prediction of spatio-temporal patterns of neural activity from pairwise correlations
• Small-correlation expansions for the inverse Ising problem
• Dynamical criticality in the collective activity of a population of retinal neurons
• From local to critical fluctuations in lattice models: a non-perturbative renormalization-group approach
• Benchmarking inverse statistical approaches for protein structure and design with exactly solvable models
• Non-perturbative renormalization-group approach to the Bose-Hubbard model
• Mean-field theory for the inverse Ising problem at low temperatures
• Invariant $\beta$-Wishart ensembles, crossover densities and asymptotic corrections to the Marchenko-Pastur law
• Large Pseudo-Counts and $L_2$-Norm Penalties Are Necessary for the Mean-Field Inference of Ising and Potts Models
• Learning Maximum Entropy Models from finite size datasets: a fast Data-Driven algorithm allows sampling from the posterior distribution
• U.S. stock market interaction network as learned by the Boltzmann Machine
• Susceptibility Propagation by Using Diagonal Consistency
• Estimating the principal components of correlation matrices from all their empirical eigenvectors

Cited by in corpus (9)

• Inverse statistical problems: from the inverse Ising problem to data science
• High-temperature Expansions and Message Passing Algorithms
• Neural networks with redundant representation: detecting the undetectable
• Expansion of the effective action around non-Gaussian theories
• Inverse Ising inference from high-temperature re-weighting of observations
• Recursive algorithm for generating high-temperature expansions for spin systems and the chiral non-linear susceptibility
• Diagrammatics for the Inverse Problem in Spin Systems and Simple Liquids
• Functional renormalization group for signal detection and stochastic ergodicity breaking
• Diagrammatics of free energies with fixed variance for high-dimensional data