Fairness-Aware Estimation of Graphical Models

Authors

• Zhuoping Zhou
• Davoud Ataee Tarzanagh
• Bojian Hou
• Qi Long
• Li Shen

Language

English

Publication Date

1-1-2024

Journal

Advances in Neural Information Processing Systems

PMID

40855882

PMCID

PMC12373123

PubMedCentral® Posted Date

8-25-2025

PubMedCentral® Full Text Version

Author MSS

Abstract

This paper examines the issue of fairness in the estimation of graphical models (GMs), particularly Gaussian, Covariance, and Ising models. These models play a vital role in understanding complex relationships in high-dimensional data. However, standard GMs can result in biased outcomes, especially when the underlying data involves sensitive characteristics or protected groups. To address this, we introduce a comprehensive framework designed to reduce bias in the estimation of GMs related to protected attributes. Our approach involves the integration of the pairwise graph disparity error and a tailored loss function into a nonsmooth multi-objective optimization problem, striving to achieve fairness across different sensitive groups while maintaining the effectiveness of the GMs. Experimental evaluations on synthetic and real-world datasets demonstrate that our framework effectively mitigates bias without undermining GMs’ performance.

Published Open-Access

yes

Recommended Citation

Zhou, Zhuoping; Tarzanagh, Davoud Ataee; Hou, Bojian; et al., "Fairness-Aware Estimation of Graphical Models" (2024). Faculty, Staff and Student Publications. 935.
https://digitalcommons.library.tmc.edu/uthshis_docs/935