High-dimensional joint estimation of multiple directed Gaussian graphical models

DOI10.1214/20-EJS1724zbMATH Open1445.62046arXiv1804.00778MaRDI QIDQ2192308FDOQ2192308

Santiago Segarra, Yuhao Wang, Caroline Uhler

Publication date: 17 August 2020

Published in: Electronic Journal of Statistics (Search for Journal in Brave)

Abstract: We consider the problem of jointly estimating multiple related directed acyclic graph (DAG) models based on high-dimensional data from each graph. This problem is motivated by the task of learning gene regulatory networks based on gene expression data from different tissues, developmental stages or disease states. We prove that under certain regularity conditions, the proposed $el{l}_0$-penalized maximum likelihood estimator converges in Frobenius norm to the adjacency matrices consistent with the data-generating distributions and has the correct sparsity. In particular, we show that this joint estimation procedure leads to a faster convergence rate than estimating each DAG model separately. As a corollary, we also obtain high-dimensional consistency results for causal inference from a mix of observational and interventional data. For practical purposes, we propose emph{jointGES} consisting of Greedy Equivalence Search (GES) to estimate the union of all DAG models followed by variable selection using lasso to obtain the different DAGs, and we analyze its consistency guarantees. The proposed method is illustrated through an analysis of simulated data as well as epithelial ovarian cancer gene expression data.

Full work available at URL: https://arxiv.org/abs/1804.00778

Recommendations

• Joint structural estimation of multiple graphical models
• Joint estimation of multiple high-dimensional Gaussian copula graphical models
• Joint skeleton estimation of multiple directed acyclic graphs for heterogeneous population
• Bayesian joint inference for multiple directed acyclic graphs
• Joint estimation of multiple high-dimensional precision matrices

zbMATH Keywords

causal inferencehigh-dimensional statisticsgraphical modellinear structural equation model

Mathematics Subject Classification ID

Asymptotic properties of parametric estimators (62F12) Applications of statistics to biology and medical sciences; meta analysis (62P10) Applications of graph theory (05C90) Probabilistic graphical models (62H22) Parametric inference under constraints (62F30)

Cites Work

• Causation, prediction, and search
• Stability Selection
• High-dimensional graphs and variable selection with the Lasso
• Model Selection and Estimation in Regression with Grouped Variables
• A characterization of Markov equivalence classes for acyclic digraphs
• Sparse inverse covariance estimation with the graphical lasso
• Title not available (Why is that?)
• \(\ell_{0}\)-penalized maximum likelihood for sparse directed acyclic graphs
• Title not available (Why is that?)
• Title not available (Why is that?)
• The max-min hill-climbing Bayesian network structure learning algorithm
• Joint estimation of multiple graphical models
• The Joint Graphical Lasso for Inverse Covariance Estimation Across Multiple Classes
• Bayesian Inference of Multiple Gaussian Graphical Models
• High-dimensional covariance estimation by minimizing \(\ell _{1}\)-penalized log-determinant divergence
• A Constrainedℓ₁Minimization Approach to Sparse Precision Matrix Estimation
• Joint estimation of multiple high-dimensional precision matrices
• Jointly Interventional and Observational Data: Estimation of Interventional Markov Equivalence Classes of Directed Acyclic Graphs
• Characterization and Greedy Learning of Interventional Markov Equivalence Classes of Directed Acyclic Graphs
• Estimating time-varying networks
• On the Gaussian approximation of convolutions under multidimensional analogues of S. N. Bernstein's inequality conditions
• Node-Based Learning of Multiple Gaussian Graphical Models
• Geometry of the faithfulness assumption in causal inference
• Probability Inequalities for the Sum of Independent Random Variables
• 10.1162/153244303321897717
• High-dimensional learning of linear causal networks via inverse covariance estimation
• Title not available (Why is that?)
• High-dimensional consistency in score-based and hybrid structure learning

Cited In (8)

• Bayesian Multi-Task Variable Selection with an Application to Differential DAG Analysis
• Joint estimation of sparse multivariate regression and conditional graphical models
• Bayesian joint inference for multiple directed acyclic graphs
• Identifiability of directed Gaussian graphical models with one latent source
• Joint estimation of multiple dependent Gaussian graphical models with applications to mouse genomics
• Title not available (Why is that?)
• Estimation of high-dimensional graphical models using regularized score matching
• Multiple Matrix Gaussian Graphs Estimation

Uses Software

• glasso
• TETRAD
• KEGG
• KELLER

This page was built for publication: High-dimensional joint estimation of multiple directed Gaussian graphical models

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2192308)