mlsanityData to 7 Oct 2026

Breaking Fair Binary Classification with Optimal Flipping Attacks

arXiv 2204.05472 ↗First posted 12 Apr 2022cs.LG cs.CR cs.CYcited by 4

Not on the map. Why?

Abstract (arXiv)

Minimizing risk with fairness constraints is one of the popular approaches to learning a fair classifier. Recent works showed that this approach yields an unfair classifier if the training set is corrupted. In this work, we study the minimum amount of data corruption required for a successful flipping attack. First, we find lower/upper bounds on this quantity and show that these bounds are tight when the target model is the unique unconstrained risk minimizer. Second, we propose a computationally efficient data poisoning attack algorithm that can compromise the performance of fair learning algorithms.

Key facts. Cited by 4 papers in the map (data through 7 Oct 2026). Builds on 31 of its 52 references that are in the map. First posted on arXiv 12 Apr 2022.

Builds on

31 of 52 references.

References · 31

  1. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,204
  2. 2020Language Models are Few-Shot Learnerscited by 32,199
  3. 2016Equality of Opportunity in Supervised Learningcited by 2,000
  4. 2014Certifying and removing disparate impactcited by 715
  5. 2018Mitigating Unwanted Biases with Adversarial Learningcited by 663
  6. 2018A Reductions Approach to Fair Classificationcited by 585
  7. 2017On Fairness and Calibrationcited by 406
  8. 2016Fairness Beyond Disparate Treatment & Disparate Impact: Learning Classification without Disparate Mistreatmentcited by 368
  9. 2018Empirical Risk Minimization under Fairness Constraintscited by 226
  10. 2019Identifying and Correcting Label Bias in Machine Learningcited by 146
  11. 2018Stronger Data Poisoning Attacks Break Data Sanitization Defensescited by 125
  12. 2019Fair Generative Modeling via Weak Supervisioncited by 114
19 more
  1. 2020Certified Robustness to Label-Flipping Attacks via Randomized Smoothingcited by 103
  2. 2020FairBatch: Batch Selection for Model Fairnesscited by 99
  3. 2017Beyond Parity: Fairness Objectives for Collaborative Filteringcited by 93
  4. 2020Robust Optimization for Fairness with Noisy Protected Groupscited by 78
  5. 2018Label Sanitization against Label Flipping Poisoning Attackscited by 68
  6. 2019Leveraging Labeled and Unlabeled Data for Consistent Fair Binary Classificationcited by 60
  7. 2020Poisoning Attacks on Algorithmic Fairnesscited by 58
  8. 2018Two-Player Games for Efficient Non-Convex Constrained Optimizationcited by 55
  9. 2020FR-Train: A Mutual Information-Based Approach to Fair and Robust Trainingcited by 52
  10. 2019Equalized odds postprocessing under imperfect group informationcited by 48
  11. 2021Sample Selection for Fair and Robust Trainingcited by 48
  12. 2020Exacerbating Algorithmic Bias through Fairness Attackscited by 47
  13. 2019Noise-tolerant fair classificationcited by 40
  14. 2021Learning Fair Classifiers with Partially Annotated Group Labelscited by 40
  15. 2020On Adversarial Bias and the Robustness of Fair Machine Learningcited by 36
  16. 2020Mitigating Bias in Set Selection with Noisy Protected Attributescited by 29
  17. 2021Fair Classification with Adversarial Perturbationscited by 23
  18. 2021Poisoning Attacks on Fair Machine Learningcited by 19
  19. 2021Fairness-Aware PAC Learning from Corrupted Datacited by 18

Built on by

Cited by year

Citing papers per year: 2022: 1; 2023: 2; 2024: 1.

4 papers.

  1. 2024Exploring Privacy and Fairness Risks in Sharing Diffusion Models: An Adversarial Perspectivecited by 6
  2. 2023Fairness Hacking: The Malicious Practice of Shrouding Unfairness in Algorithmscited by 5
  3. 2022A Survey on Preserving Fairness Guarantees in Changing Environmentscited by 2
  4. 2023Measuring Equality in Machine Learning Security Defenses: A Case Study in Speech Recognitioncited by 0