Improved Pseudolikelihood Regularization and Decimation methods on Non-linearly Interacting Systems with Continuous Variables

Abstract

We propose and test improvements to state-of-the-art techniques of Bayeasian statistical inference based on pseudolikelihood maximization with ${\ell }_1$ regularization and with decimation. In particular, we present a method to determine the best value of the regularizer parameter starting from a hypothesis testing technique. Concerning the decimation, we also analyze the worst case scenario in which there is no sharp peak in the tilded-pseudolikelihood function, firstly defined as a criterion to stop the decimation. Techniques are applied to noisy systems with non-linear dynamics, mapped onto multi-variable interacting Hamiltonian effective models for waves and phasors. Results are analyzed varying the number of available samples and the externally tunable temperature-like parameter mimicing real data noise. Eventually the behavior of inference procedures described are tested against a wrong hypothesis: non-linearly generated data are analyzed with a pairwise interacting hypothesis. Our analysis shows that, looking at the behavior of the inverse graphical problem as data size increases, the methods exposed allow to rule out a wrong hypothesis.

arXiv comment37 pages, 24 figures

References in corpus (48)

• Identification of direct residue contacts in protein-protein interaction by message passing
• High-dimensional Ising model selection using ${\ell }_1$-regularized logistic regression
• Inverse statistical problems: from the inverse Ising problem to data science
• Statistical properties of stock order books: empirical results and models
• Experimental evidence of replica symmetry breaking in random lasers
• The Ising Model for Neural Data: Model Quality and Approximate Methods for Extracting Functional Connectivity
• Statistical mechanics of the random K-SAT model
• Inverse Ising inference using all the data
• Fast and accurate multivariate Gaussian modeling of protein families: Predicting residue contacts and protein-interaction partners
• Adaptive Cluster Expansion for Inferring Boltzmann Machines with Noisy Data
• Simultaneous identification of specifically interacting paralogs and inter-protein contacts by Direct-Coupling Analysis
• Statistical physics of pairwise probability models
• Small-correlation expansions for the inverse Ising problem
• Field Theory of Fluctuations in Glasses
• Theory of collective opinion shifts: from smooth trends to abrupt swings
• Glassy behavior of light
• Improving contact prediction along three dimensions
• General phase-diagram of multimodal ordered and disordered lasers in closed and open cavities
• Observation of Replica Symmetry Breaking in the 1D Anderson Localization Regime in an Erbium-Doped Random Fiber Laser
• Complexity of waves in nonlinear disordered media
• Phase diagram and complexity of mode-locked lasers: from order to disorder
• Pseudolikelihood Decimation Algorithm Improving the Inference of the Interaction Network in a General Class of Ising Models
• Replica symmetry breaking in specially designed TiO${}_2$ nanoparticles-based dye-colloidal random laser
• On the critical slowing down exponents of mode coupling theory
• The Effect of Nonstationarity on Models Inferred from Neural Data
• Glassy behavior of light in random lasers
• The glassy random laser: replica symmetry breaking in the intensity fluctuations of emission spectra
• Complexity in Mean-Field Spin-Glass Models: Ising $p$-spin
• The Complex Spherical 2+4 Spin Glass: a Model for Nonlinear Optics in Random Media
• A stroll among effective temperatures in aging systems: limits and perspectives
• Derivatives and Credit Contagion in Interconnected Networks
• Statistical mechanics models for multimode lasers and random lasers
• The random Blume-Capel model on cubic lattice: first order inverse freezing in a 3D spin-glass system
• On the two-steps relaxation of mean-field glasses: p-spin model
• Statistical physics of nonlinear wave interaction
• Learning with hidden variables
• Effects of Economic Interactions on Credit Risk
• Statistical physical theory of mode-locking laser generation with a frequency comb
• The Ising M-p-spin mean-field model for the structural glass: continuous vs. discontinuous transition
• Nonlinear $XY$ and $p$-clock models on sparse random graphs: mode-locking transition of localized waves
• Mode-locking transitions in nano-structured weakly disordered lasers
• Regularization and decimation pseudolikelihood approaches to statistical inference in $XY$-spin models
• Multi-body quenched disordered $XY$ and $p$-clock models on random graphs
• Contagion in an interacting economy
• Modeling the emergence of polarity patterns for the intercellular transport of auxin in plants
• Inverse problem for multi-body interaction of nonlinear waves
• Inference for interacting linear waves in ordered and random media
• Variational Pseudolikelihood for Regularized Ising Inference

Cited by in corpus (6)

• Glassiness and Lack of Equipartition in Random Lasers: the common roots of Ergodicity Breaking in Disordered and Non-linear Systems
• Solving the fully-connected spherical $p$-spin model with the cavity method: equivalence with the replica results
• Universality class of the mode-locked glassy random laser
• A fast and accurate algorithm for inferring sparse Ising models via parameters activation to maximize the pseudo-likelihood
• Intensity pseudo-localized phase in the glassy random laser
• Learning Direct and Inverse Transmission Matrices