Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
125 tokens/sec
GPT-4o
53 tokens/sec
Gemini 2.5 Pro Pro
42 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
47 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Fairness in Matching under Uncertainty (2302.03810v2)

Published 8 Feb 2023 in cs.LG

Abstract: The prevalence and importance of algorithmic two-sided marketplaces has drawn attention to the issue of fairness in such settings. Algorithmic decisions are used in assigning students to schools, users to advertisers, and applicants to job interviews. These decisions should heed the preferences of individuals, and simultaneously be fair with respect to their merits (synonymous with fit, future performance, or need). Merits conditioned on observable features are always \emph{uncertain}, a fact that is exacerbated by the widespread use of machine learning algorithms to infer merit from the observables. As our key contribution, we carefully axiomatize a notion of individual fairness in the two-sided marketplace setting which respects the uncertainty in the merits; indeed, it simultaneously recognizes uncertainty as the primary potential cause of unfairness and an approach to address it. We design a linear programming framework to find fair utility-maximizing distributions over allocations, and we show that the linear program is robust to perturbations in the estimated parameters of the uncertain merit distributions, a key property in combining the approach with machine learning techniques.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (65)
  1. A. Abdulkadiroğlu and T. Sönmez. School choice: A mechanism design approach. American Economic Review, 93(3):729–747, 2003.
  2. Network Flows. Prentice Hall, 1993.
  3. Unbiased learning to rank with unbiased propensity estimation. In Proc. 41st Intl. Conf. on Research and Development in Information Retrieval (SIGIR), pages 385–394, 2018.
  4. Fair division of indivisible goods: A survey. In L. D. Raedt, editor, Proc. 31st Intl. Joint Conf. on Artificial Intelligence, 2022.
  5. S. Athanassoglou and J. Sethuraman. House allocation with fractional endowments. International Journal of Game Theory, 40(3):481–513, 2011.
  6. Stable matching with uncertain linear preferences. In International Symposium on Algorithmic Game Theory, pages 195–206. Springer, 2016.
  7. A framework for fairness in two-sided marketplaces. arXiv preprint arXiv:2006.12756, 2020.
  8. G. Birkhoff. Tres observaciones sobre el algebra lineal. Univ. Nac. Tucuman, Ser. A, 5:147–154, 1946.
  9. Toward fairness in personalized ads, 2023. URL https://about.fb.com/wp-content/uploads/2023/01/Toward_fairness_in_personalized_ads.pdf.
  10. Individually fair ranking. In International Conference on Learning Representations, 2021.
  11. Assigning Israeli medical graduates to internships. Israel Journal of Health Policy Research, 4(1):1–7, 2015.
  12. L. Brozovsky and V. Petricek. Recommender system for online dating service. In Proceedings of Znalosti 2007 Conference, Ostrava, 2007. VSB.
  13. Statistical discrimination in stable matchings. In Proc. 23rd ACM Conf. on Economics and Computation, pages 373–374, 2022.
  14. Toward controlling discrimination in online ad auctions. In Proc. 36th Intl. Conf. on Machine Learning, pages 4456–4465. PMLR, 2019.
  15. Interventions for ranking in the presence of implicit bias. In Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency, pages 369–380, 2020.
  16. S. Chawla and M. Jagadeesan. Individual fairness in advertising auctions through inverse proportionality. In Proc. 13th Innovations in Theoretical Computer Science. Schloss Dagstuhl-Leibniz-Zentrum für Informatik, 2022.
  17. Mechanism design for fair division: allocating divisible items without payments. In Proc. 14th ACM Conf. on Electronic Commerce, pages 251–268, 2013.
  18. Sensitivity theorems in integer linear programming. Mathematical Programming, 34:251–264, 1986.
  19. V. Do and N. Usunier. Optimizing generalized Gini indices for fairness in rankings. In E. Amigó, P. Castells, J. Gonzalo, B. Carterette, J. S. Culpepper, and G. Kazai, editors, Proc. 45th Intl. Conf. on Research and Development in Information Retrieval (SIGIR), pages 737–747. ACM, 2022.
  20. Two-sided fairness in rankings via Lorenz dominance. In Proc. 35th Advances in Neural Information Processing Systems, pages 8596–8608, 2021.
  21. Contextual bandits with concave rewards, and an application to fair ranking. arXiv preprint arXiv:2210.09957, 2022.
  22. Asymptotic minimax character of the sample distribution function and of the classical multinomial estimator. The Annals of Mathematical Statistics, pages 642–669, 1956.
  23. Fairness through awareness. In Proc. 3rd Innovations in Theoretical Computer Science, pages 214–226. ACM, 2012.
  24. On fair selection in the presence of implicit variance. In Proc. 21st ACM Conf. on Economics and Computation, pages 649–675, 2020.
  25. Rawlsian fairness in online bipartite matching: Two-sided, group, and individual. arXiv preprint arXiv:2201.06021, 2022.
  26. T. Fleiner and N. Kamiyama. A matroid approach to stable matchings with lower quotas. Mathematics of Operations Research, 41(2):734–744, 2016.
  27. Two-sided matching meets fair division. In Z. Zhou, editor, Proc. 30th Intl. Joint Conf. on Artificial Intelligence, pages 203–209. ijcai.org, 2021.
  28. When fair ranking meets uncertain inference. In F. Diaz, C. Shah, T. Suel, P. Castells, R. Jones, and T. Sakai, editors, Proc. 44th Intl. Conf. on Research and Development in Information Retrieval (SIGIR), pages 1033–1043. ACM, 2021.
  29. D. Gusfield and R. W. Irving. The stable marriage problem: structure and algorithms. MIT press, 1989.
  30. Equality of opportunity in supervised learning. In Proc. 30th Advances in Neural Information Processing Systems, volume 29, 2016.
  31. Multicalibration: Calibration for the (computationally-identifiable) masses. In Proc. 35th Intl. Conf. on Machine Learning, pages 1939–1948. PMLR, 2018.
  32. Optimally interpolating between ex-ante fairness and welfare. arXiv preprint arXiv:2302.03071, 2023.
  33. Fair matchings and related problems. Algorithmica, 74(3):1184–1203, 2016.
  34. Fair division with two-sided preferences. arXiv preprint arXiv:2206.05879, 2022.
  35. Multi-category fairness in sponsored search auctions. In Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency, pages 348–358, 2020.
  36. Learning equilibria in matching markets from bandit feedback. In Proc. 35th Advances in Neural Information Processing Systems, pages 3323–3335, 2021.
  37. Supply-side equilibria in recommender systems. arXiv preprint arXiv:2206.13489, 2022.
  38. On fairness and stability in two-sided matchings. In Proc. 13th Innovations in Theoretical Computer Science, volume 215 of LIPIcs, pages 92:1–92:17, 2022.
  39. Meritocratic fairness for cross-population selection. In Proc. 34th Intl. Conf. on Machine Learning, pages 1828–1836. PMLR, 2017.
  40. Preventing fairness gerrymandering: Auditing and learning for subgroup fairness. In International Conference on Machine Learning, pages 2564–2572. PMLR, 2018.
  41. Fairness through computationally-bounded awareness. In Proc. 32nd Advances in Neural Information Processing Systems, 2018.
  42. Preference-informed fairness. In Proc. 11th Innovations in Theoretical Computer Science, volume 151 of LIPIcs, pages 16:1–16:23, 2020.
  43. Inherent trade-offs in the fair determination of risk scores. In C. H. Papadimitriou, editor, Proc. 18th Innovations in Theoretical Computer Science, volume 67 of LIPIcs, pages 43:1–43:23, 2017.
  44. Pareto-optimal fairness-utility amortizations in rankings with a DBN exposure model. In E. Amigó, P. Castells, J. Gonzalo, B. Carterette, J. S. Culpepper, and G. Kazai, editors, Proc. 45th Intl. Conf. on Research and Development in Information Retrieval (SIGIR), pages 748–758. ACM, 2022a.
  45. Introducing the expohedron for efficient pareto-optimal fairness-utility amortizations in repeated rankings. In Proc. 15th ACM Intl. Conf. on Web Search and Data Mining, pages 498–507, 2022b.
  46. LinkedIn. Campaign quality scores for sponsored content. https://www.linkedin.com/help/lms/answer/85406, 2023.
  47. Equal-quantile rules in resource allocation with uncertain needs. Journal of Economic Theory, 197:105350, 2021.
  48. Ad click prediction: a view from the trenches. In Proceedings of the 19th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD), 2013.
  49. A. Mehta et al. Online matching and ad allocation. Foundations and Trends® in Theoretical Computer Science, 8(4):265–368, 2013.
  50. Learn to match with no regret: Reinforcement learning in markov matching markets. In Proc. 36th Advances in Neural Information Processing Systems, 2022.
  51. Integer and Combinatorial Optimization. Wiley, 1988.
  52. Bipartite matchings with group fairness and individual fairness constraints. arXiv preprint arXiv:2208.09951, 2022.
  53. Incremental fairness in two-sided market platforms: On smoothly updating recommendations. In Proc. 34th AAAI Conf. on Artificial Intelligence, pages 181–188, 2020.
  54. Fair ranking: a critical review, challenges, and future directions. In Conference on Fairness, Accountability, and Transparency (FAccT), pages 1929–1942. ACM, 2022.
  55. A. E. Roth. On the allocation of residents to rural hospitals: a general property of two-sided matching markets. Econometrica, pages 425–427, 1986.
  56. Efficient kidney exchange: Coincidence of wants in markets with compatibility-based preferences. American Economic Review, 97(3):828–851, 2007.
  57. R. Salakhutdinov and A. Mnih. Bayesian probabilistic matrix factorization using markov chain monte carlo. In Proc. 25th Intl. Conf. on Machine Learning, 2008.
  58. J. Salem and S. Gupta. Closing the gap: Online selections of candidates with biased evaluations. In Proc. 16th Conference on Web and Internet Economics (WINE), 2020.
  59. We need fairness and explainability in algorithmic hiring. In International Conference on Autonomous Agents and Multi-Agent Systems (AAMAS), 2020.
  60. A. Singh and T. Joachims. Fairness of exposure in rankings. In Proc. 24th Intl. Conf. on Knowledge Discovery and Data Mining, pages 2219–2228, 2018.
  61. Fairness in ranking under uncertainty. In Proc. 35th Advances in Neural Information Processing Systems, pages 11896–11908, 2021.
  62. R. C. Smith. Uncertainty quantification: theory, implementation, and applications, volume 12. Siam, 2013.
  63. Optimizing rankings for recommendation in matching markets. In 31st The Web Conference, pages 328–338, 2022.
  64. L. Wang and T. Joachims. User fairness, item fairness, and diversity for rankings in two-sided markets. In Proc. 44th Intl. Conf. on Research and Development in Information Retrieval (SIGIR), pages 23–41, 2021.
  65. J. Xue. Fair division with uncertain needs. Social Choice and Welfare, 51(1):105–136, 2018.
Citations (6)
View on Semantic Scholar

Summary

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

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