Multi-Attribute Graph Estimation with Sparse-Group Non-Convex Penalties

TL;DR

This paper introduces a unified framework for multi-attribute graph estimation in high-dimensional Gaussian models, employing both convex and non-convex penalties, with theoretical guarantees and improved empirical performance.

Contribution

It provides a comprehensive theoretical analysis and optimization approach for multi-attribute graph learning using sparse-group non-convex penalties, outperforming existing methods.

Findings

Non-convex penalties achieve better support recovery.

The proposed method outperforms lasso and SCAD in synthetic data.

Theoretical guarantees established for high-dimensional support recovery.

Abstract

We consider the problem of inferring the conditional independence graph (CIG) of high-dimensional Gaussian vectors from multi-attribute data. Most existing methods for graph estimation are based on single-attribute models where one associates a scalar random variable with each node. In multi-attribute graphical models, each node represents a random vector. In this paper we provide a unified theoretical analysis of multi-attribute graph learning using a penalized log-likelihood objective function. We consider both convex (sparse-group lasso) and sparse-group non-convex (log-sum and smoothly clipped absolute deviation (SCAD) penalties) penalty/regularization functions. An alternating direction method of multipliers (ADMM) approach coupled with local linear approximation to non-convex penalties is presented for optimization of the objective function. For non-convex penalties, theoretical analysis establishing local consistency in support recovery, local convexity and precision matrix estimation in high-dimensional settings is provided under two sets of sufficient conditions: with and without some irrepresentability conditions. We illustrate our approaches using both synthetic and real-data numerical examples. In the synthetic data examples the sparse-group log-sum penalized objective function significantly outperformed the lasso penalized as well as SCAD penalized objective functions with $F_1$-score and Hamming distance as performance metrics.