Estimation of positive definite M-matrices and structure learning for attractive Gaussian Markov random fields
From MaRDI portal
Publication:2341885
DOI10.1016/J.LAA.2014.04.020zbMATH Open1312.62070arXiv1404.6640OpenAlexW2963089591MaRDI QIDQ2341885FDOQ2341885
Authors: Martin Slawski, Matthias Hein 
Publication date: 6 May 2015
Published in: Linear Algebra and its Applications (Search for Journal in Brave)
Abstract: Consider a random vector with finite second moments. If its precision matrix is an M-matrix, then all partial correlations are non-negative. If that random vector is additionally Gaussian, the corresponding Markov random field (GMRF) is called attractive. We study estimation of M-matrices taking the role of inverse second moment or precision matrices using sign-constrained log-determinant divergence minimization. We also treat the high-dimensional case with the number of variables exceeding the sample size. The additional sign-constraints turn out to greatly simplify the estimation problem: we provide evidence that explicit regularization is no longer required. To solve the resulting convex optimization problem, we propose an algorithm based on block coordinate descent, in which each sub-problem can be recast as non-negative least squares problem. Illustrations on both simulated and real world data are provided.
Full work available at URL: https://arxiv.org/abs/1404.6640
Recommendations
- Gaussian Markov random fields and totally positive matrices
- Conditional-mean least-squares fitting of Gaussian Markov random fields to Gaussian fields
- Maximum likelihood and restricted maximum likelihood estimation for a class of Gaussian Markov random fields
- scientific article; zbMATH DE number 4070105
- Noncausal Gauss Markov random fields: parameter structure and estimation
- Approximating Hidden Gaussian Markov Random Fields
- scientific article; zbMATH DE number 7307482
Gaussian Markov random fields\(\ell_1\)-regularizationprecision matrix estimation\(M\)-matricesgraphical model selectionpartial correlationssign constraintshigh-dimensional statistical inferencelog-determinant divergence minimization
Convex programming (90C25) Estimation in multivariate analysis (62H12) Parametric inference under constraints (62F30) Sign pattern matrices (15B35)
Cites Work
- Causation, prediction, and search
- Title not available (Why is that?)
- Asymptotic Statistics
- Title not available (Why is that?)
- Statistics for high-dimensional data. Methods, theory and applications.
- High-dimensional graphs and variable selection with the Lasso
- Sparse inverse covariance estimation with the graphical lasso
- Sparse permutation invariant covariance estimation
- Sign-constrained least squares estimation for high-dimensional regression
- Network exploration via the adaptive LASSO and SCAD penalties
- Model selection through sparse maximum likelihood estimation for multivariate Gaussian or binary data
- Estimating high-dimensional directed acyclic graphs with the PC-algorithm
- Title not available (Why is that?)
- Model selection and estimation in the Gaussian graphical model
- Title not available (Why is that?)
- Approximating discrete probability distributions with dependence trees
- High-dimensional covariance estimation by minimizing \(\ell _{1}\)-penalized log-determinant divergence
- High dimensional inverse covariance matrix estimation via linear programming
- A constrained \(\ell _{1}\) minimization approach to sparse precision matrix estimation
- Likelihood-based selection and sharp parameter estimation
- High-dimensional covariance estimation based on Gaussian graphical models
- Title not available (Why is that?)
- Title not available (Why is that?)
- Classes of orderings of measures and related correlation inequalities. I. Multivariate totally positive distributions
- Estimation of a covariance matrix with zeros
- Title not available (Why is that?)
- A Comparison of Block Pivoting and Interior-Point Algorithms for Linear Least Squares Problems with Nonnegative Variables
- Title not available (Why is that?)
- High-dimensional Gaussian graphical model selection: walk summability and local separation criterion
- Matrix Nearness Problems with Bregman Divergences
- Geometry of maximum likelihood estimation in Gaussian graphical models
- Total positivity order and the normal distribution
- Title not available (Why is that?)
- Title not available (Why is that?)
- M-matrices as covariance matrices of multinormal distributions
- Estimation of positive definite \(M\)-matrices and structure learning for attractive Gaussian Markov random fields
Cited In (15)
- Total positivity in exponential families with application to binary variables
- Model-preserving sensitivity analysis for families of Gaussian distributions
- Total positivity in multivariate extremes
- Data Analytics on Graphs Part III: Machine Learning on Graphs, from Graph Topology to Applications
- Dependence in elliptical partial correlation graphs
- Multiway \(p\)-spectral graph cuts on Grassmann manifolds
- Estimation of positive definite \(M\)-matrices and structure learning for attractive Gaussian Markov random fields
- Maximum likelihood estimation in Gaussian models under total positivity
- A multitest procedure for testing MTP2 for Gaussian distributions
- Learning graph Laplacian with MCP
- Convolutions of totally positive distributions with applications to kernel density estimation
- Bimonotone subdivisions of point configurations in the plane
- Optimal rates for estimation of two-dimensional totally positive distributions
- Locally associated graphical models and mixed convex exponential families
- A unified framework for structured graph learning via spectral constraints
Uses Software
This page was built for publication: Estimation of positive definite \(M\)-matrices and structure learning for attractive Gaussian Markov random fields
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2341885)
