SenSeI: Sensitive Set Invariance for Enforcing Individual Fairness

2020/06/25 by Mikhail Yurochkin, Yuekai Sun, Yurochkin, Mikhail +1 · 15 citations
Computer Science · Mathematics · Social Sciences · #Adversarial Robustness in Machine Learning #Algorithm #Artificial intelligence #Computer science #Ethics and Social Impacts of AI #Fairness measure #Invariant (physics) #Machine learning #Mathematical optimization #Mathematics #Privacy-Preserving Technologies in Data #Set (abstract data type) #Theoretical computer science #Train #cs.LG #stat.ML

paper · pdf · doi:10.48550/arxiv.2006.14168

published in arXiv (Cornell University) (Cornell University) · ICLR 2021

openalex publication_date 2020/06/25 · arxiv created 2021/04/01 · arxiv updated 2021/04/02 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

In this paper, we cast fair machine learning as invariant machine learning. We first formulate a version of individual fairness that enforces invariance on certain sensitive sets. We then design a transport-based regularizer that enforces this version of individual fairness and develop an algorithm to minimize the regularizer efficiently. Our theoretical results guarantee the proposed approach trains certifiably fair ML models. Finally, in the experimental studies we demonstrate improved fairness metrics in comparison to several recent fair training procedures on three ML tasks that are susceptible to algorithmic bias.

Cited by

• Developing Fairness-Aware Task Decomposition to Improve Equity in Post-Spinal Fusion Complication Prediction
• Removing Spurious Features can Hurt Accuracy and Affect Groups Disproportionately
• Wasserstein Distributionally Robust Optimization Through the Lens of Structural Causal Models and Individual Fairness
• Toward Operationalizing Pipeline-aware ML Fairness: A Research Agenda for Developing Practical Guidelines and Tools
• Individually Fair Gradient Boosting

Related

• Training individually fair ML models with Sensitive Subspace Robustness
• Individual Fairness under Uncertainty
• Correct-By-Construction: Certified Individual Fairness through Neural Network Training
• FAIRM: Learning invariant representations for algorithmic fairness and domain generalization with minimax optimality
• Verifying Individual Fairness in Machine Learning Models