# Addressing Fairness in Classification with a Model-Agnostic Multi-Objective Algorithm

arXiv: https://arxiv.org/abs/2009.04441 · HTML: https://mlsanity.com/paper/2009.04441/

## Key facts

- Cited by 23 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 9 Sep 2020.

## Built on by (most cited first, 23 of 23)

- [Bias Mitigation for Machine Learning Classifiers: A Comprehensive Survey](https://mlsanity.com/paper/2207.07068/) (2022)
- [Scalable Pareto Front Approximation for Deep Multi-Objective Learning](https://mlsanity.com/paper/2103.13392/) (2021)
- [Fair Enough: Searching for Sufficient Measures of Fairness](https://mlsanity.com/paper/2110.13029/) (2021)
- [Chasing Fairness Under Distribution Shift: A Model Weight Perturbation Approach](https://mlsanity.com/paper/2303.03300/) (2023)
- [PMGDA: A Preference-based Multiple Gradient Descent Algorithm](https://mlsanity.com/paper/2402.09492/) (2024)
- [LibMOON: A Gradient-based MultiObjective OptimizatioN Library in PyTorch](https://mlsanity.com/paper/2409.02969/) (2024)
- [Tiny, always-on and fragile: Bias propagation through design choices in on-device machine learning workflows](https://mlsanity.com/paper/2201.07677/) (2022)
- [Mitigating Unfairness via Evolutionary Multi-objective Ensemble Learning](https://mlsanity.com/paper/2210.16754/) (2022)
- [Understanding Fairness Surrogate Functions in Algorithmic Fairness](https://mlsanity.com/paper/2310.11211/) (2023)
- [fairret: a Framework for Differentiable Fairness Regularization Terms](https://mlsanity.com/paper/2310.17256/) (2023)
- [Algorithmic Fairness in Performative Policy Learning: Escaping the Impossibility of Group Fairness](https://mlsanity.com/paper/2405.20447/) (2024)
- [Risk-Controlling Model Selection via Guided Bayesian Optimization](https://mlsanity.com/paper/2312.01692/) (2023)
- [FairNorm: Fair and Fast Graph Neural Network Training](https://mlsanity.com/paper/2205.09977/) (2022)
- [Towards Fairness-Aware Multi-Objective Optimization](https://mlsanity.com/paper/2207.12138/) (2022)
- [Generalization In Multi-Objective Machine Learning](https://mlsanity.com/paper/2208.13499/) (2022)
- [Self-Evolutionary Optimization for Pareto Front Learning](https://mlsanity.com/paper/2110.03461/) (2021)
- [Beyond Impossibility: Balancing Sufficiency, Separation and Accuracy](https://mlsanity.com/paper/2205.12327/) (2022)
- [Best Practices for Responsible Machine Learning in Credit Scoring](https://mlsanity.com/paper/2409.20536/) (2024)
- [Enhancing Multi-Attribute Fairness in Healthcare Predictive Modeling](https://mlsanity.com/paper/2501.13219/) (2025)
- [CurvFed: Curvature-Aligned Federated Learning for Fairness without Demographics](https://mlsanity.com/paper/2404.19725/) (2024)
- [A Multi-Objective Evaluation Framework for Analyzing Utility-Fairness Trade-Offs in Machine Learning Systems](https://mlsanity.com/paper/2503.11120/) (2025)
- [Evolutionary Multi-Objective Optimisation for Fairness-Aware Self Adjusting Memory Classifiers in Data Streams](https://mlsanity.com/paper/2404.12076/) (2024)
- [APFEx: Adaptive Pareto Front Explorer for Intersectional Fairness](https://mlsanity.com/paper/2509.13908/) (2025)

## Abstract (arXiv)

The goal of fairness in classification is to learn a classifier that does not discriminate against groups of individuals based on sensitive attributes, such as race and gender. One approach to designing fair algorithms is to use relaxations of fairness notions as regularization terms or in a constrained optimization problem. We observe that the hyperbolic tangent function can approximate the indicator function. We leverage this property to define a differentiable relaxation that approximates fairness notions provably better than existing relaxations. In addition, we propose a model-agnostic multi-objective architecture that can simultaneously optimize for multiple fairness notions and multiple sensitive attributes and supports all statistical parity-based notions of fairness. We use our relaxation with the multi-objective architecture to learn fair classifiers. Experiments on public datasets show that our method suffers a significantly lower loss of accuracy than current debiasing algorithms relative to the unconstrained model.