Inverse Ising inference using all the data

Abstract

We show that a method based on logistic regression, using all the data, solves the inverse Ising problem far better than mean-field calculations relying only on sample pairwise correlation functions, while still computationally feasible for hundreds of nodes. The largest improvement in reconstruction occurs for strong interactions. Using two examples, a diluted Sherrington-Kirkpatrick model and a two-dimensional lattice, we also show that interaction topologies can be recovered from few samples with good accuracy and that the use of $l_1$-regularization is beneficial in this process, pushing inference abilities further into low-temperature regimes.

5 pages, 2 figures. Accepted version

References in corpus (6)

• Direct-coupling analysis of residue co-evolution captures native contacts across many protein families
• Identification of direct residue contacts in protein-protein interaction by message passing
• High-dimensional Ising model selection using ${\ell }_1$-regularized logistic regression
• Adaptive Cluster Expansion for Inferring Boltzmann Machines with Noisy Data
• Small-correlation expansions for the inverse Ising problem
• Dynamics and Performance of Susceptibility Propagation on Synthetic Data

Cited by in corpus (79)

• Improved contact prediction in proteins: Using pseudolikelihoods to infer Potts models
• Sequence co-evolution gives 3D contacts and structures of protein complexes
• Colloquium: Criticality and dynamical scaling in living systems
• Data Based Identification and Prediction of Nonlinear and Complex Dynamical Systems
• Inverse statistical problems: from the inverse Ising problem to data science
• Inverse Statistical Physics of Protein Sequences: A Key Issues Review
• Estimation of effective temperatures in quantum annealers for sampling applications: A case study with possible applications in deep learning
• Fast pseudolikelihood maximization for direct-coupling analysis of protein structure from many homologous amino-acid sequences
• Global analysis of more than 50,000 SARS-Cov-2 genomes reveals epistasis between 8 viral genes
• Pseudolikelihood Decimation Algorithm Improving the Inference of the Interaction Network in a General Class of Ising Models
• Global warming: Temperature estimation in annealers
• Mean-field theory for the inverse Ising problem at low temperatures
• Maximum likelihood reconstruction for Ising models with asynchronous updates
• Market structure explained by pairwise interactions
• Mapping of attention mechanisms to a generalized Potts model
• Sparse Logistic Regression Learns All Discrete Pairwise Graphical Models
• Large Pseudo-Counts and $L_2$-Norm Penalties Are Necessary for the Mean-Field Inference of Ising and Potts Models
• U.S. stock market interaction network as learned by the Boltzmann Machine
• Consistent inference of a general model using the pseudo-likelihood method
• Universal structural estimator and dynamics approximator for complex networks
• Learning Maximum Entropy Models from finite size datasets: a fast Data-Driven algorithm allows sampling from the posterior distribution
• Statistical Physics and Representations in Real and Artificial Neural Networks
• Introduction to correlation networks: Interdisciplinary approaches beyond thresholding
• Marginal Pseudo-Likelihood Learning of Markov Network structures
• Improving landscape inference by integrating heterogeneous data in the inverse Ising problem
• Higher-order interactions in statistical physics and machine learning: A model-independent solution to the inverse problem at equilibrium
• Data-Driven Interaction Analysis of Line Failure Cascading in Power Grid Networks
• Inverse design of disordered stealthy hyperuniform spin chains
• Surges of collective human activity emerge from simple pairwise correlations
• Correlation-Compressed Direct Coupling Analysis
• Ensemble Inhibition and Excitation in the Human Cortex: an Ising Model Analysis with Uncertainties
• A statistical physics perspective on criticality in financial markets
• A mean field method with correlations determined by linear response
• Inference of the sparse kinetic Ising model using the decimation method
• Neural activity in quarks language: Lattice Field Theory for a network of real neurons
• Aggregation models on hypergraphs
• A statistical physics approach to learning curves for the Inverse Ising problem
• Inverse Ising inference with correlated samples
• Network reconstruction via the minimum description length principle
• Statistical mechanics of the inverse Ising problem and the optimal objective function
• Regularization and decimation pseudolikelihood approaches to statistical inference in $XY$-spin models
• Sparse model selection in the highly under-sampled regime
• Financial interaction networks inferred from traded volumes
• Resummed mean-field inference for strongly coupled data
• Learning phase transitions from regression uncertainty: A new regression-based machine learning approach for automated detection of phases of matter
• Inverse problem for multi-species mean field models in the low temperature phase
• From statistical inference to a differential learning rule for stochastic neural networks
• Improved Pseudolikelihood Regularization and Decimation methods on Non-linearly Interacting Systems with Continuous Variables
• Learning performance in inverse Ising problems with sparse teacher couplings
• Network inference in the non-equilibrium steady state
• Biases in Inverse Ising Estimates of Near-Critical Behaviour
• Predicting trend reversals using market instantaneous state
• Inference of stochastic time series with missing data
• Thermodynamics of the Ising model encoded in restricted Boltzmann machines
• Inverse problem for multi-body interaction of nonlinear waves
• Beyond inverse Ising model: structure of the analytical solution for a class of inverse problems
• Inverse Ising inference from high-temperature re-weighting of observations
• Data-driven effective model shows a liquid-like deep learning
• Sparse Hopfield network reconstruction with ${\ell }_1$ regularization
• A Density Consistency approach to the inverse Ising problem
• Active image restoration
• Inferring the parameters of a Markov process from snapshots of the steady state
• Detecting Multi-Spin Interactions in the Inverse Ising Problem
• Learning Direct and Inverse Transmission Matrices
• Spectral Bounds for the Ising Ferromagnet on an Arbitrary Given Graph
• Reconstruction of Pairwise Interactions using Energy-Based Models
• On the typical properties of inverse problems in statistical mechanics
• Structure Learning in Inverse Ising Problems Using ${\ell }_2$-Regularized Linear Estimator
• Worldwide bilateral geopolitical interactions network inferred from national disciplinary profiles
• Composite Spatial Monte Carlo Integration Based on Generalized Least Squares
• On the role of non-linear latent features in bipartite generative neural networks
• Objective and efficient inference for couplings in neuronal networks
• Monte Carlo, blocking, and inference: How to measure the renormalization group flow
• On the complexity of logistic regression models
• GNisi: A graph network for reconstructing Ising models from multivariate binarized data
• Ising Model Selection Using ${\ell }_1$-Regularized Linear Regression: A Statistical Mechanics Analysis
• Data quality for the inverse Ising problem
• Classification of SARS-CoV-2 Variants through The Epistatical Circos Plots with Convolutional Neural Networks
• Inverse Ising inference, hyperuniformity and absorbing states in the Manna model