Fairness-informed Pareto Optimization : An Efficient Bilevel Framework

2026-01-19

AI Summary

This work addresses the limitations of existing fair machine learning methods, which often yield Pareto-inefficient solutions or are constrained by the perspective bias inherent in specific fairness metrics. To overcome these challenges, the authors propose BADR, a bilevel optimization framework: the lower-level problem learns a model via weighted empirical risk minimization, while the upper-level problem adaptively rescales group weights to optimize any user-specified fairness objective, thereby recovering Pareto-optimal trade-offs between fairness and accuracy. BADR is agnostic to the choice of fairness metric, transcending the efficiency and perspective constraints of conventional approaches. The paper further introduces two efficient single-loop algorithms, BADR-GD and BADR-SGD. Extensive experiments demonstrate that BADR consistently outperforms existing Pareto-efficient methods across diverse tasks and fairness measures, and the authors release the badr toolbox to facilitate reproducibility and adoption.

Technology Category

• Machine Learning: Ethics, Bias, and Fairness
• Multiagent Systems: Adversarial Agents
• Computer Vision: Bias, Fairness & Privacy

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Despite their promise, fair machine learning methods often yield Pareto-inefficient models, in which the performance of certain groups can be improved without degrading that of others. This issue arises frequently in traditional in-processing approaches such as fairness-through-regularization. In contrast, existing Pareto-efficient approaches are biased towards a certain perspective on fairness and fail to adapt to the broad range of fairness metrics studied in the literature. In this paper, we present BADR, a simple framework to recover the optimal Pareto-efficient model for any fairness metric. Our framework recovers its models through a Bilevel Adaptive Rescalarisation procedure. The lower level is a weighted empirical risk minimization task where the weights are a convex combination of the groups, while the upper level optimizes the chosen fairness objective. We equip our framework with two novel large-scale, single-loop algorithms, BADR-GD and BADR-SGD, and establish their convergence guarantees. We release badr, an open-source Python toolbox implementing our framework for a variety of learning tasks and fairness metrics. Finally, we conduct extensive numerical experiments demonstrating the advantages of BADR over existing Pareto-efficient approaches to fairness.

Problem

• fair machine learning
• Pareto efficiency
• fairness metrics
• bilevel optimization

Innovation

• Pareto efficiency
• bilevel optimization
• fair machine learning
• adaptive rescaling
• single-loop algorithm