mlsanityData to 14 Oct 2026

How unfair is private learning ?

arXiv 2206.03985 ↗First posted 8 Jun 2022cs.LG cs.CR stat.MLcited by 18

Not on the map. Why?

Abstract (arXiv)

As machine learning algorithms are deployed on sensitive data in critical decision making processes, it is becoming increasingly important that they are also private and fair. In this paper, we show that, when the data has a long-tailed structure, it is not possible to build accurate learning algorithms that are both private and results in higher accuracy on minority subpopulations. We further show that relaxing overall accuracy can lead to good fairness even with strict privacy requirements. To corroborate our theoretical results in practice, we provide an extensive set of experimental results using a variety of synthetic, vision (CIFAR10 and CelebA), and tabular (Law School) datasets and learning algorithms.

Key facts. Cited by 18 papers in the map (data through 14 Oct 2026). Builds on 33 of its 63 references that are in the map. First posted on arXiv 8 Jun 2022.

Builds on

33 of 63 references.

References · 33

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2014Deep Learning Face Attributes in the Wildcited by 3,693
  3. 2016Deep Learning with Differential Privacycited by 2,728
  4. 2016Visual Genome: Connecting Language and Vision Using Crowdsourced Dense Image Annotationscited by 2,656
  5. 2016Membership Inference Attacks against Machine Learning Modelscited by 2,281
  6. 2016Equality of Opportunity in Supervised Learningcited by 2,007
  7. 2020WILDS: A Benchmark of in-the-Wild Distribution Shiftscited by 1,375
  8. 2020What Neural Networks Memorize and Why: Discovering the Long Tail via Influence Estimationcited by 497
  9. 2019Does Learning Require Memorization? A Short Tale about a Long Tailcited by 460
  10. 2017Preventing Fairness Gerrymandering: Auditing and Learning for Subgroup Fairnesscited by 384
  11. 2021Opacus: User-Friendly Differential Privacy Library in PyTorchcited by 383
  12. 2018Learning Models with Uniform Performance via Distributionally Robust Optimizationcited by 335
21 more
  1. 2019Differential Privacy Has Disparate Impact on Model Accuracycited by 316
  2. 2018Large Scale Fine-Grained Categorization and Domain-Specific Transfer Learningcited by 174
  3. 2021A survey on datasets for fairness-aware machine learningcited by 157
  4. 2019Does Object Recognition Work for Everyone?cited by 151
  5. 2021Deep Learning with Label Differential Privacycited by 117
  6. 2017The Devil is in the Tails: Fine-grained Classification in the Wildcited by 89
  7. 2020Model Patching: Closing the Subgroup Performance Gap with Data Augmentationcited by 83
  8. 2020On the Privacy Risks of Algorithmic Fairnesscited by 82
  9. 2018Private PAC learning implies finite Littlestone dimensioncited by 74
  10. 2016DiSMEC - Distributed Sparse Machines for Extreme Multi-label Classificationcited by 71
  11. 2020An Equivalence Between Private Classification and Online Predictioncited by 64
  12. 2014Private Learning and Sanitization: Pure vs. Approximate Differential Privacycited by 63
  13. 2021Fairness via Representation Neutralizationcited by 63
  14. 2019Diffprivlib: The IBM Differential Privacy Librarycited by 53
  15. 2022Differential Privacy and Fairness in Decisions and Learning Tasks: A Surveycited by 47
  16. 2017Privacy Loss in Apple's Implementation of Differential Privacy on MacOS 10.12cited by 43
  17. 2020How benign is benign overfitting?cited by 40
  18. 2014Sample Complexity Bounds on Differentially Private Learning via Communication Complexitycited by 31
  19. 2014Characterizing the Sample Complexity of Private Learnerscited by 19
  20. 2015Building a RAPPOR with the Unknown: Privacy-Preserving Learning of Associations and Data Dictionariescited by 16
  21. 2016Differentially Private Random Decision Forests using Smooth Sensitivitycited by 16

Built on by

Cited by year

Citing papers per year: 2022: 3; 2023: 7; 2024: 4; 2025: 4.

18 papers.

Sort:

By area:

  1. 2024Corrective Machine Unlearningcited by 29
  2. 2023Privacy and Fairness in Federated Learning: on the Perspective of Trade-offcited by 26
  3. 2023Arbitrary Decisions are a Hidden Cost of Differentially Private Trainingcited by 19
  4. 2022Differential Privacy has Bounded Impact on Fairness in Classificationcited by 18
  5. 2023Unlocking Accuracy and Fairness in Differentially Private Image Classificationcited by 16
  6. 2023FRAPPE: A Group Fairness Framework for Post-Processing Everythingcited by 16
  7. 2023PILLAR: How to make semi-private learning more effectivecited by 10
  8. 2022A law of adversarial risk, interpolation, and label noisecited by 8
  9. 2024The Role of Learning Algorithms in Collective Actioncited by 8
  10. 2022A Closer Look at the Calibration of Differentially Private Learnerscited by 4
  11. 2023On the Fairness Impacts of Private Ensembles Modelscited by 4
  12. 2023FairDP: Certified Fairness with Differential Privacycited by 4
6 more
  1. 2025SoK: What Makes Private Learning Unfair?cited by 3
  2. 2025Minority Collective Action for User-Side Fairnesscited by 3
  3. 2024From Efficiency to Equity: Measuring Fairness in Preference Learningcited by 2
  4. 2024A Cognac Shot To Forget Bad Memories: Corrective Unlearning for Graph Neural Networkscited by 2
  5. 2025Learning Pareto manifolds in high dimensions: How can regularization help?cited by 1
  6. 2025Fairness Meets Privacy: Integrating Differential Privacy and Demographic Parity in Multi-class Classificationcited by 1