Identifying Network Hubs with the Partial Correlation Graphical LASSO

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Bogdan, Małgorzata, Chojecki, Adam, Hejný, Ivan, Kołodziejek, Bartosz, Wallin, Jonas
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866913103318351872
author Bogdan, Małgorzata
Chojecki, Adam
Hejný, Ivan
Kołodziejek, Bartosz
Wallin, Jonas
author_facet Bogdan, Małgorzata
Chojecki, Adam
Hejný, Ivan
Kołodziejek, Bartosz
Wallin, Jonas
contents Graphical LASSO (GLASSO) is a widely used method for estimating sparse precision matrices and learning undirected graphical models in high-dimensional settings. Because GLASSO penalizes entries of the precision matrix directly, however, it is not scale-invariant. Partial Correlation Graphical LASSO (PCGLASSO), introduced by Carter et al. (2024), addresses this limitation by penalizing partial correlations, which directly characterize conditional dependence. In this paper, we study both statistical and computational properties of the PCGLASSO estimator. Our main contribution is the introduction of a scale-invariant irrepresentability condition for PCGLASSO and the proof that this condition is sufficient for consistent model selection. We further show that this condition is weaker than the corresponding irrepresentability condition for GLASSO, helping to explain the improved empirical behavior of PCGLASSO in settings such as hub-structured graphs. In addition, we develop two efficient algorithms for computing the estimator and analyze the nonconvex optimization problem underlying PCGLASSO, deriving conditions for global uniqueness and showing consistency of all minimizers.
format Preprint
id arxiv_https___arxiv_org_abs_2508_12258
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Identifying Network Hubs with the Partial Correlation Graphical LASSO
Bogdan, Małgorzata
Chojecki, Adam
Hejný, Ivan
Kołodziejek, Bartosz
Wallin, Jonas
Statistics Theory
Optimization and Control
Primary 62H22, secondary 62H12, 62J07, 90C26
Graphical LASSO (GLASSO) is a widely used method for estimating sparse precision matrices and learning undirected graphical models in high-dimensional settings. Because GLASSO penalizes entries of the precision matrix directly, however, it is not scale-invariant. Partial Correlation Graphical LASSO (PCGLASSO), introduced by Carter et al. (2024), addresses this limitation by penalizing partial correlations, which directly characterize conditional dependence. In this paper, we study both statistical and computational properties of the PCGLASSO estimator. Our main contribution is the introduction of a scale-invariant irrepresentability condition for PCGLASSO and the proof that this condition is sufficient for consistent model selection. We further show that this condition is weaker than the corresponding irrepresentability condition for GLASSO, helping to explain the improved empirical behavior of PCGLASSO in settings such as hub-structured graphs. In addition, we develop two efficient algorithms for computing the estimator and analyze the nonconvex optimization problem underlying PCGLASSO, deriving conditions for global uniqueness and showing consistency of all minimizers.
title Identifying Network Hubs with the Partial Correlation Graphical LASSO
topic Statistics Theory
Optimization and Control
Primary 62H22, secondary 62H12, 62J07, 90C26
url https://arxiv.org/abs/2508.12258