Estimation of high-dimensional graphical models using regularized score matching
From MaRDI portal
Abstract: Graphical models are widely used to model stochastic dependences among large collections of variables. We introduce a new method of estimating undirected conditional independence graphs based on the score matching loss, introduced by Hyvarinen (2005), and subsequently extended in Hyvarinen (2007). The regularized score matching method we propose applies to settings with continuous observations and allows for computationally efficient treatment of possibly non-Gaussian exponential family models. In the well-explored Gaussian setting, regularized score matching avoids issues of asymmetry that arise when applying the technique of neighborhood selection, and compared to existing methods that directly yield symmetric estimates, the score matching approach has the advantage that the considered loss is quadratic and gives piecewise linear solution paths under regularization. Under suitable irrepresentability conditions, we show that -regularized score matching is consistent for graph estimation in sparse high-dimensional settings. Through numerical experiments and an application to RNAseq data, we confirm that regularized score matching achieves state-of-the-art performance in the Gaussian case and provides a valuable tool for computationally efficient estimation in non-Gaussian graphical models.
Recommendations
- Conditional score matching for high-dimensional partial graphical models
- Regularized rank-based estimation of high-dimensional nonparanormal graphical models
- Graphical-model based high dimensional generalized linear models
- High-Dimensional Gaussian Graphical Regression Models with Covariates
- A two-step method for estimating high-dimensional Gaussian graphical models
- Simultaneous inference for pairwise graphical models with generalized score matching
- Estimating high dimensional faithful Gaussian graphical models by low-order conditioning
- High-dimensional joint estimation of multiple directed Gaussian graphical models
- High-dimensional generalized linear models incorporating graphical structure among predictors
- High-dimensional covariance estimation based on Gaussian graphical models
Cites work
- scientific article; zbMATH DE number 1134987 (Why is no real title available?)
- scientific article; zbMATH DE number 845714 (Why is no real title available?)
- A Connection Between Score Matching and Denoising Autoencoders
- A note on the Lasso for Gaussian graphical model selection
- A robust, adaptive M-estimator for pointwise estimation in heteroscedastic regression
- Conditional specification of statistical models.
- Convergence of a block coordinate descent method for nondifferentiable minimization
- Copula Gaussian graphical models and their application to modeling functional disability data
- Distinctness of the eigenvalues of a quadratic form in a multivariate sample
- Distributional and \(L^q\) norm inequalities for polynomials over convex bodies in \(\mathbb{R}^n\)
- Elliptical graphical modelling
- Emergence of Scaling in Random Networks
- Estimating the dimension of a model
- Estimation of high-dimensional graphical models using regularized score matching
- Estimation of non-normalized statistical models by score matching
- Estimation of sparse binary pairwise Markov networks using pseudo-likelihoods
- Estimation of spatial processes using local scoring rules: spatial special issue
- Extended Bayesian information criteria for model selection with large model spaces
- Graph estimation with joint additive models
- Graphical models via univariate exponential family distributions
- High-dimensional Ising model selection using \(\ell _{1}\)-regularized logistic regression
- High-dimensional Ising model selection with Bayesian information criteria
- High-dimensional covariance estimation by minimizing \(\ell _{1}\)-penalized log-determinant divergence
- High-dimensional graphs and variable selection with the Lasso
- High-dimensional regression with noisy and missing data: provable guarantees with nonconvexity
- High-dimensional semiparametric Gaussian copula graphical models
- Introduction to Graphical Modelling
- Least angle regression. (With discussion)
- Linear estimating equations for exponential families with application to Gaussian linear concentration models
- Model Selection and Estimation in Regression with Grouped Variables
- Model selection and estimation in the Gaussian graphical model
- Multiple testing and error control in Gaussian graphical model selection
- Partial correlation estimation by joint sparse regression models
- Pathwise coordinate optimization
- Piecewise linear regularized solution paths
- Robust Gaussian graphical modeling
- Robust Gaussian graphical modeling via \(l_{1}\) penalization
- Robust graphical modeling of gene networks using classical and alternative \(t\)-distributions
- Sharp Thresholds for High-Dimensional and Noisy Sparsity Recovery Using $\ell _{1}$-Constrained Quadratic Programming (Lasso)
- Some extensions of score matching
- Stable graphical model estimation with random forests for discrete, continuous, and mixed variables
- Statistics for high-dimensional data. Methods, theory and applications.
- The Lasso problem and uniqueness
- The nonparanormal: semiparametric estimation of high dimensional undirected graphs
- Tuning parameter selection for penalized likelihood estimation of Gaussian graphical model
Cited in
(28)- Addendum on the scoring of Gaussian directed acyclic graphical models
- scientific article; zbMATH DE number 7415101 (Why is no real title available?)
- Covariate-adjusted inference for differential analysis of high-dimensional networks
- Inference via low-dimensional couplings
- Diagonal nonlinear transformations preserve structure in covariance and precision matrices
- Discussion of “A Tuning-Free Robust and Efficient Approach to High-Dimensional Regression”
- Linear estimating equations for exponential families with application to Gaussian linear concentration models
- Generalized score matching for non-negative data
- Estimation of high-dimensional graphical models using regularized score matching
- Torus graphs for multivariate phase coupling analysis
- Scoring matching approach: Learning high order Markov random fields
- Causal Structural Learning via Local Graphs
- A generalized likelihood-based Bayesian approach for scalable joint regression and covariance selection in high dimensions
- Conditional score matching for high-dimensional partial graphical models
- An effective method for high-dimensional log-density ANOVA estimation, with application to nonparametric graphical model building
- Robust and sparse Gaussian graphical modelling under cell-wise contamination
- Regularized rank-based estimation of high-dimensional nonparanormal graphical models
- Comparing score-based methods for estimating Bayesian networks using the Kullback-Leibler divergence
- Compressed spectral screening for large-scale differential correlation analysis with application in selecting glioblastoma gene modules
- Discussion of: ``Identifiability of latent-variable and structural-equation models: from linear to nonlinear
- scientific article; zbMATH DE number 7370576 (Why is no real title available?)
- Simultaneous inference for pairwise graphical models with generalized score matching
- Layer-wise learning strategy for nonparametric tensor product smoothing spline regression and graphical models
- genscore
- The huge Package for High-dimensional Undirected Graph Estimation in R
- A scalable sparse Cholesky based approach for learning high-dimensional covariance matrices in ordered data
- Dependence structure estimation using copula recursive trees
- An efficient ADMM algorithm for high dimensional precision matrix estimation via penalized quadratic loss
This page was built for publication: Estimation of high-dimensional graphical models using regularized score matching
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q138467)
