Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
119 tokens/sec
GPT-4o
56 tokens/sec
Gemini 2.5 Pro Pro
43 tokens/sec
o3 Pro
6 tokens/sec
GPT-4.1 Pro
47 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Towards Algorithmic Fairness by means of Instance-level Data Re-weighting based on Shapley Values (2303.01928v4)

Published 3 Mar 2023 in cs.LG, cs.AI, and cs.CY

Abstract: Algorithmic fairness is of utmost societal importance, yet state-of-the-art large-scale machine learning models require training with massive datasets that are frequently biased. In this context, pre-processing methods that focus on modeling and correcting bias in the data emerge as valuable approaches. In this paper, we propose FairShap, a novel instance-level data re-weighting method for fair algorithmic decision-making through data valuation by means of Shapley Values. FairShap is model-agnostic and easily interpretable. It measures the contribution of each training data point to a predefined fairness metric. We empirically validate FairShap on several state-of-the-art datasets of different nature, with a variety of training scenarios and machine learning models and show how it yields fairer models with similar levels of accuracy than the baselines. We illustrate FairShap's interpretability by means of histograms and latent space visualizations. Moreover, we perform a utility-fairness study and analyze FairShap's computational cost depending on the size of the dataset and the number of features. We believe that FairShap represents a novel contribution in interpretable and model-agnostic approaches to algorithmic fairness that yields competitive accuracy even when only biased training datasets are available.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (61)
  1. Counterfactual shapley additive explanations. In 2022 ACM Conference on Fairness, Accountability, and Transparency, pages 1054–1070, 2022.
  2. Accounting for model uncertainty in algorithmic discrimination. In Proceedings of the 2021 AAAI/ACM Conference on AI, Ethics, and Society, pages 336–345, 2021.
  3. Machine bias: There’s software used across the country to predict future criminals. and it’s biased against blacks. propublica, may 23, 2016.
  4. S. Barocas and A. D. Selbst. Big data’s disparate impact. California law review, pages 671–732, 2016.
  5. Fairness and Machine Learning: Limitations and Opportunities. The MIT Press, 2019.
  6. Influence functions in deep learning are fragile. In International Conference on Learning Representations, 2021.
  7. Ai fairness 360: An extensible toolkit for detecting and mitigating algorithmic bias. IBM Journal of Research and Development, 63(4/5):4–1, 2019.
  8. Fairness in criminal justice risk assessments: The state of the art. Sociological Methods & Research, 50(1):3–44, 2021.
  9. E. Black and M. Fredrikson. Leave-one-out unfairness. In ACM Conference on Fairness, Accountability, and Transparency, page 285–295, 2021.
  10. J. Brophy. Exit through the training data: A look into instance-attribution explanations and efficient data deletion in machine learning. Technical report Oregon University, 2020.
  11. Optimized pre-processing for discrimination prevention. In Advances in Neural Information Processing Systems, volume 30, 2017.
  12. A. N. Carey and X. Wu. The statistical fairness field guide: perspectives from social and formal sciences. AI and Ethics, pages 1–23, 2022.
  13. S. Caton and C. Haas. Fairness in machine learning: A survey. ACM Comput. Surv., August 2023.
  14. J. Chai and X. Wang. Fairness with adaptive weights. In International Conference on Machine Learning, volume 162 of Proceedings of Machine Learning Research, pages 2853–2866. PMLR, 17–23 Jul 2022.
  15. A. Chouldechova. Fair prediction with disparate impact: A study of bias in recidivism prediction instruments. Big data, 5(2):153–163, 2017.
  16. A. Chouldechova and A. Roth. A snapshot of the frontiers of fairness in machine learning. Communications of the ACM, 63(5):82–89, 2020.
  17. Fairness through awareness. In Proceedings of the 3rd innovations in theoretical computer science conference, pages 214–226, 2012.
  18. Certifying and removing disparate impact. In Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pages 259–268, 2015.
  19. X. Fern and Q. Pope. Text counterfactuals via latent optimization and shapley-guided search. In Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, pages 5578–5593, 2021.
  20. J. H. Friedman. Greedy function approximation: a gradient boosting machine. Annals of statistics, pages 1189–1232, 2001.
  21. Datasheets for datasets. Communications of the ACM, 64(12):86–92, 2021.
  22. A. Ghorbani and J. Zou. Data shapley: Equitable valuation of data for machine learning. In International Conference on Machine Learning, pages 2242–2251. PMLR, 2019.
  23. D. B. Gillies. Solutions to general non-zero-sum games. Contributions to the Theory of Games, 4:47–85, 1959.
  24. Beyond impossibility: Balancing sufficiency, separation and accuracy. In NeurIPS Workshop on Algorithmic Fairness through the Lens of Causality and Privacy, 2022.
  25. P. Hacker and J.-H. Passoth. Varieties of ai explanations under the law. from the gdpr to the aia, and beyond. In International Workshop on Extending Explainable AI Beyond Deep Models and Classifiers, pages 343–373. Springer, 2022.
  26. T. Hagendorff. The ethics of ai ethics: An evaluation of guidelines. Minds and Machines, 30(1):99–120, 2020.
  27. Z. Hammoudeh and D. Lowd. Training data influence analysis and estimation: A survey. arXiv preprint arXiv:2212.04612, 2022.
  28. Equality of opportunity in supervised learning. Advances in Neural Information Processing Systems, 29, 2016.
  29. Efficient task-specific data valuation for nearest neighbor algorithms. Proc. VLDB Endow., 12(11):1610–1623, 2019. ISSN 2150-8097.
  30. H. Jiang and O. Nachum. Identifying and correcting label bias in machine learning. In International Conference on Artificial Intelligence and Statistics, pages 702–712. PMLR, 2020.
  31. OpenDataVal: a Unified Benchmark for Data Valuation. In Advances in Neural Information Processing Systems Datasets and Benchmarks Track, 2023.
  32. Re-weighting based group fairness regularization via classwise robust optimization. In The Eleventh International Conference on Learning Representations, 2023.
  33. F. Kamiran and T. Calders. Classifying without discriminating. In 2009 2nd international conference on computer, control and communication, pages 1–6. IEEE, 2009.
  34. F. Kamiran and T. Calders. Data preprocessing techniques for classification without discrimination. Knowledge and information systems, 33(1):1–33, 2012.
  35. Fairness-aware classifier with prejudice remover regularizer. In Joint European conference on machine learning and knowledge discovery in databases, pages 35–50. Springer, 2012.
  36. K. Karkkainen and J. Joo. Fairface: Face attribute dataset for balanced race, gender, and age for bias measurement and mitigation. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, pages 1548–1558, 2021.
  37. P. W. Koh and P. Liang. Understanding black-box predictions via influence functions. In International Conference on Machine Learning, pages 1885–1894. PMLR, 2017.
  38. R. Kohavi et al. Scaling up the accuracy of naive-bayes classifiers: A decision-tree hybrid. In Kdd, volume 96, pages 202–207, 1996.
  39. Adaptive sensitive reweighting to mitigate bias in fairness-aware classification. In Proceedings of the 2018 world wide web conference, pages 853–862, 2018.
  40. Y. Kwon and J. Zou. Beta shapley: a unified and noise-reduced data valuation framework for machine learning. In International Conference on Artificial Intelligence and Statistics, pages 8780–8802. PMLR, 28–30 Mar 2022.
  41. P. Li and H. Liu. Achieving fairness at no utility cost via data reweighing with influence. In International Conference on Machine Learning, volume 162, pages 12917–12930. PMLR, 17–23 Jul 2022.
  42. Deep learning face attributes in the wild. In Proceedings of the IEEE international conference on computer vision, pages 3730–3738, 2015.
  43. A unified approach to interpreting model predictions. Advances in Neural Information Processing Systems, 30, 2017.
  44. Assessing the fairness of ai systems: Ai practitioners’ processes, challenges, and needs for support. Proceedings of the ACM on Human-Computer Interaction, 6(CSCW1):1–26, 2022.
  45. A survey on bias and fairness in machine learning. ACM Computing Surveys (CSUR), 54(6):1–35, 2021.
  46. C. Molnar. Interpretable machine learning. Lulu. com, 2020.
  47. N. Oliver. Artificial intelligence for social good - The way forward. In Science, Research and Innovation performance of the EU 2022 report, chapter 11, pages 604–707. European Commission, 2022.
  48. Deep learning on a data diet: Finding important examples early in training. In Advances in Neural Information Processing Systems, 2021.
  49. Estimating training data influence by tracing gradient descent. In Advances in Neural Information Processing Systems, volume 33, pages 19920–19930, 2020.
  50. CS-shapley: Class-wise shapley values for data valuation in classification. In Advances in Neural Information Processing Systems, 2022.
  51. L. S. Shapley. A value for n-person games. Contributions to the Theory of Games, 2:307–317, 1953.
  52. Data valuation in machine learning:“ingredients”, strategies, and open challenges. In Proc. IJCAI, pages 5607–5614, 2022.
  53. N. Smuha. Ethics guidelines for trustworthy AI. In AI & Ethics, Date: 2019/05/28-2019/05/28, Brussels, Belgium. European Commission, 2019.
  54. Axiomatic attribution for deep networks. In International Conference on Machine Learning, volume 70 of Proceedings of Machine Learning Research, pages 3319–3328. PMLR, 06–11 Aug 2017.
  55. Inception-v4, inception-resnet and the impact of residual connections on learning. In Thirty-first AAAI conference on artificial intelligence, 2017.
  56. Measure contribution of participants in federated learning. In 2019 IEEE international conference on big data (Big Data), pages 2597–2604. IEEE, 2019.
  57. Understanding Instance-Level Impact of Fairness Constraints. In International Conference on Machine Learning, volume 162, pages 23114–23130. PMLR, 17–23 Jul 2022.
  58. DAVINZ: Data valuation using deep neural networks at initialization. In International Conference on Machine Learning, 2022.
  59. Fairness Beyond Disparate Treatment &; Disparate Impact: Learning Classification without Disparate Mistreatment. In International Conference on World Wide Web, page 1171–1180, 2017.
  60. Learning fair representations. In International Conference on Machine Learning, pages 325–333. PMLR, 2013.
  61. Mitigating unwanted biases with adversarial learning. In Proceedings of the 2018 AAAI/ACM Conference on AI, Ethics, and Society, pages 335–340, 2018.
User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (2)
  1. Adrian Arnaiz-Rodriguez (5 papers)
  2. Nuria Oliver (46 papers)
Citations (1)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now