# Classification with abstention but without disparities arXiv: https://arxiv.org/abs/2102.12258 ยท HTML: https://mlsanity.com/paper/2102.12258/ ## Key facts - Cited by 28 papers in the map (data through 14 Oct 2026). - Builds on 14 of its 45 references that are in the map. - First posted on arXiv 24 Feb 2021. ## Builds on (most cited first, 14 of 14) - [Equality of Opportunity in Supervised Learning](https://mlsanity.com/paper/1610.02413/) (2016) - [A Reductions Approach to Fair Classification](https://mlsanity.com/paper/1803.02453/) (2018) - [Fairness Beyond Disparate Treatment & Disparate Impact: Learning Classification without Disparate Mistreatment](https://mlsanity.com/paper/1610.08452/) (2016) - [Predict Responsibly: Improving Fairness and Accuracy by Learning to Defer](https://mlsanity.com/paper/1711.06664/) (2017) - [A Random Forest Guided Tour](https://mlsanity.com/paper/1511.05741/) (2015) - [Fair Regression: Quantitative Definitions and Reduction-based Algorithms](https://mlsanity.com/paper/1905.12843/) (2019) - [Leveraging Labeled and Unlabeled Data for Consistent Fair Binary Classification](https://mlsanity.com/paper/1906.05082/) (2019) - [Selective Classification Can Magnify Disparities Across Groups](https://mlsanity.com/paper/2010.14134/) (2020) - [Exponential Savings in Agnostic Active Learning through Abstention](https://mlsanity.com/paper/2102.00451/) (2021) - [Review of Mathematical frameworks for Fairness in Machine Learning](https://mlsanity.com/paper/2005.13755/) (2020) - [Fast Rates for Online Prediction with Abstention](https://mlsanity.com/paper/2001.10623/) (2020) - [Minimax optimal rates for Mondrian trees and forests](https://mlsanity.com/paper/1803.05784/) (2018) - [Fast classification rates without standard margin assumptions](https://mlsanity.com/paper/1910.12756/) (2019) - [Efficient Inverse Maintenance and Faster Algorithms for Linear Programming](https://mlsanity.com/paper/1503.01752/) (2015) ## Built on by (most cited first, 28 of 28) - [Bias Mitigation for Machine Learning Classifiers: A Comprehensive Survey](https://mlsanity.com/paper/2207.07068/) (2022) - [Predictor-Rejector Multi-Class Abstention: Theoretical Analysis and Algorithms](https://mlsanity.com/paper/2310.14772/) (2023) - [Selective Regression Under Fairness Criteria](https://mlsanity.com/paper/2110.15403/) (2021) - [Theoretically Grounded Loss Functions and Algorithms for Score-Based Multi-Class Abstention](https://mlsanity.com/paper/2310.14770/) (2023) - [Fairness Indicators for Systematic Assessments of Visual Feature Extractors](https://mlsanity.com/paper/2202.07603/) (2022) - [Bayes-Optimal Classifiers under Group Fairness](https://mlsanity.com/paper/2202.09724/) (2022) - [Fair learning with Wasserstein barycenters for non-decomposable performance measures](https://mlsanity.com/paper/2209.00427/) (2022) - [Fair Bayes-Optimal Classifiers Under Predictive Parity](https://mlsanity.com/paper/2205.07182/) (2022) - [Fair Classifiers that Abstain without Harm](https://mlsanity.com/paper/2310.06205/) (2023) - [How Far Can Fairness Constraints Help Recover From Biased Data?](https://mlsanity.com/paper/2312.10396/) (2023) - [Abstain and Validate: A Dual-LLM Policy for Reducing Noise in Agentic Program Repair](https://mlsanity.com/paper/2510.03217/) (2025) - [Bandits with Abstention under Expert Advice](https://mlsanity.com/paper/2402.14585/) (2024) - [Error rate control for classification rules in multiclass mixture models](https://mlsanity.com/paper/2109.14235/) (2021) - [Counterfactually Comparing Abstaining Classifiers](https://mlsanity.com/paper/2305.10564/) (2023) - [FairRR: Pre-Processing for Group Fairness through Randomized Response](https://mlsanity.com/paper/2403.07780/) (2024) - [Multi-Output Distributional Fairness via Post-Processing](https://mlsanity.com/paper/2409.00553/) (2024) - [Demographic parity in regression and classification within the unawareness framework](https://mlsanity.com/paper/2409.02471/) (2024) - [Interpretable and Fair Mechanisms for Abstaining Classifiers](https://mlsanity.com/paper/2503.18826/) (2025) - [Meta Optimality for Demographic Parity Constrained Regression via Post-Processing](https://mlsanity.com/paper/2506.13947/) (2025) - [Harnessing the Power of Vicinity-Informed Analysis for Classification under Covariate Shift](https://mlsanity.com/paper/2405.16906/) (2024) - [Bayes-Optimal Fair Classification with Multiple Sensitive Features](https://mlsanity.com/paper/2505.00631/) (2025) - [Randomized multi-class classification under system constraints: a unified approach via post-processing](https://mlsanity.com/paper/2512.14246/) (2025) - [Bounded-Abstention Pairwise Learning to Rank](https://mlsanity.com/paper/2505.23437/) (2025) - [Unequal Uncertainty: Rethinking Algorithmic Interventions for Mitigating Discrimination from AI](https://mlsanity.com/paper/2508.07872/) (2025) - [Fairness May Backfire: When Leveling-Down Occurs in Fair Machine Learning](https://mlsanity.com/paper/2603.06901/) (2026) - [When Surface Form Changes Moderation Decisions: A Paired Study of Code-Mixed Workflow Instability](https://mlsanity.com/paper/2606.05654/) (2026) - [Towards Provably Fair Machine Learning: Bayesian Approaches For Consistent and Transparent Predictions](https://mlsanity.com/paper/2606.12615/) (2026) - [Beyond Bias Scores: Answerability Confounds in Activation Steering with Gradient-Constrained Direction Design (GCD)](https://mlsanity.com/paper/2610.17641/) (2026) ## Abstract (arXiv) Classification with abstention has gained a lot of attention in recent years as it allows to incorporate human decision-makers in the process. Yet, abstention can potentially amplify disparities and lead to discriminatory predictions. The goal of this work is to build a general purpose classification algorithm, which is able to abstain from prediction, while avoiding disparate impact. We formalize this problem as risk minimization under fairness and abstention constraints for which we derive the form of the optimal classifier. Building on this result, we propose a post-processing classification algorithm, which is able to modify any off-the-shelf score-based classifier using only unlabeled sample. We establish finite sample risk, fairness, and abstention guarantees for the proposed algorithm. In particular, it is shown that fairness and abstention constraints can be achieved independently from the initial classifier as long as sufficiently many unlabeled data is available. The risk guarantee is established in terms of the quality of the initial classifier. Our post-processing scheme reduces to a sparse linear program allowing for an efficient implementation, which we provide. Finally, we validate our method empirically showing that moderate abstention rates allow to bypass the risk-fairness trade-off.