Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
162 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
45 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Personalized Algorithmic Recourse with Preference Elicitation (2205.13743v5)

Published 27 May 2022 in cs.LG and cs.AI

Abstract: Algorithmic Recourse (AR) is the problem of computing a sequence of actions that -- once performed by a user -- overturns an undesirable machine decision. It is paramount that the sequence of actions does not require too much effort for users to implement. Yet, most approaches to AR assume that actions cost the same for all users, and thus may recommend unfairly expensive recourse plans to certain users. Prompted by this observation, we introduce PEAR, the first human-in-the-loop approach capable of providing personalized algorithmic recourse tailored to the needs of any end-user. PEAR builds on insights from Bayesian Preference Elicitation to iteratively refine an estimate of the costs of actions by asking choice set queries to the target user. The queries themselves are computed by maximizing the Expected Utility of Selection, a principled measure of information gain accounting for uncertainty on both the cost estimate and the user's responses. PEAR integrates elicitation into a Reinforcement Learning agent coupled with Monte Carlo Tree Search to quickly identify promising recourse plans. Our empirical evaluation on real-world datasets highlights how PEAR produces high-quality personalized recourse in only a handful of iterations.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (66)
  1. APRIL: active preference learning-based reinforcement learning. In Peter A. Flach, Tijl De Bie, and Nello Cristianini (eds.), Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2012, Bristol, UK, September 24-28, 2012. Proceedings, Part II, volume 7524 of Lecture Notes in Computer Science, pp.  116–131. Springer, 2012. doi: 10.1007/978-3-642-33486-3_8. URL https://doi.org/10.1007/978-3-642-33486-3_8.
  2. Craig Boutilier. A pomdp formulation of preference elicitation problems. In AAAI, pp.  239–246, 2002.
  3. Minimax regret based elicitation of generalized additive utilities. In UAI, pp.  25–32, 2007.
  4. Ruth M. J. Byrne. Counterfactuals in explainable artificial intelligence (xai): Evidence from human reasoning. In Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence, IJCAI-19, pp.  6276–6282. International Joint Conferences on Artificial Intelligence Organization, 7 2019. doi: 10.24963/ijcai.2019/876. URL https://doi.org/10.24963/ijcai.2019/876.
  5. Making rational decisions using adaptive utility elicitation. In AAAI/IAAI, pp.  363–369, 2000.
  6. Strategic recourse in linear classification. arXiv preprint arXiv:2011.00355, 236, 2020.
  7. Rémi Coulom. Efficient selectivity and backup operators in monte-carlo tree search. In Computers and Games: 5th International Conference, CG 2006, Turin, Italy, May 29-31, 2006. Revised Papers 5, pp.  72–83. Springer, 2007.
  8. Multi-objective counterfactual explanations. In PPSN, pp.  448–469. Springer, 2020.
  9. Synthesizing explainable counterfactual policies for algorithmic recourse with program synthesis. Mach. Learn., 112(4):1389–1409, feb 2023. ISSN 0885-6125. doi: 10.1007/s10994-022-06293-7. URL https://doi.org/10.1007/s10994-022-06293-7.
  10. Constructive preference elicitation. Frontiers in Robotics and AI, 4:71, 2018.
  11. The accuracy, fairness, and limits of predicting recidivism. Science advances, 4(1):eaao5580, 2018.
  12. UCI machine learning repository, 2017. URL http://archive.ics.uci.edu/ml.
  13. Preference Learning. Springer-Verlag, Berlin, Heidelberg, 1st edition, 2010. ISBN 3642141242.
  14. Biased random-key genetic algorithms for combinatorial optimization. Journal of Heuristics, 17(5):487–525, 2011.
  15. Local rule-based explanations of black box decision systems. CoRR, abs/1805.10820, 2018. URL http://arxiv.org/abs/1805.10820.
  16. Real-time multiattribute bayesian preference elicitation with pairwise comparison queries. In AISTATS, pp.  289–296, 2010.
  17. Equalizing recourse across groups. arXiv preprint arXiv:1909.03166, 2019.
  18. Raguel: Recourse-aware group unfairness elimination. In Proceedings of the 31st ACM International Conference on Information & Knowledge Management, pp.  666–675, 2022.
  19. Combining q-learning and search with amortized value estimates. arXiv preprint arXiv:1912.02807, 2019.
  20. Strategic classification. In Proceedings of the 2016 ACM conference on innovations in theoretical computer science, pp.  111–122, 2016.
  21. Kaggle. Give me some credit, Sep 2011. URL https://www.kaggle.com/competitions/GiveMeSomeCredit/overview.
  22. Ensemble slice sampling: Parallel, black-box and gradient-free inference for correlated & multimodal distributions. arXiv preprint arXiv: 2002.06212, 2020.
  23. zeus: A python implementation of ensemble slice sampling for efficient bayesian parameter inference. arXiv preprint arXiv:2105.03468, 2021.
  24. Algorithmic recourse under imperfect causal knowledge: a probabilistic approach. Advances in Neural Information Processing Systems, 33:265–277, 2020.
  25. Algorithmic recourse: from counterfactual explanations to interventions. In Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency, pp.  353–362, 2021.
  26. A survey of algorithmic recourse: Contrastive explanations and consequential recommendations. ACM Comput. Surv., 55(5), dec 2022. ISSN 0360-0300. doi: 10.1145/3527848. URL https://doi.org/10.1145/3527848.
  27. Decisions with Multiple Objectives: Preferences and Value Trade-Offs. Cambridge University Press, 1993. doi: 10.1017/CBO9781139174084.
  28. Bandit based monte-carlo planning. In Machine Learning: ECML 2006: 17th European Conference on Machine Learning Berlin, Germany, September 18-22, 2006 Proceedings 17, pp. 282–293. Springer, 2006.
  29. Submodular function maximization. Tractability, 3:71–104, 2014.
  30. Psychology meets machine learning: Interdisciplinary perspectives on algorithmic job candidate screening. In Explainable and interpretable models in computer vision and machine learning, pp.  197–253. Springer, 2018.
  31. R Duncan Luce. Individual choice behavior: A theoretical analysis. Courier Corporation, 2012.
  32. Preserving causal constraints in counterfactual explanations for machine learning classifiers. arXiv preprint arXiv:1912.03277, 2019.
  33. James G March. Bounded rationality, ambiguity, and the engineering of choice. The bell journal of economics, pp.  587–608, 1978.
  34. Explaining machine learning classifiers through diverse counterfactual explanations. In FAT*, pp.  607–617, 2020.
  35. Consequence-aware sequential counterfactual generation. In Joint European Conference on Machine Learning and Knowledge Discovery in Databases, pp.  682–698. Springer, 2021.
  36. An analysis of approximations for maximizing submodular set functions—i. Mathematical programming, 14(1):265–294, 1978.
  37. Pytorch: An imperative style, high-performance deep learning library. In Advances in Neural Information Processing Systems 32, pp. 8024–8035. Curran Associates, Inc., 2019. URL http://papers.neurips.cc/paper/9015-pytorch-an-imperative-style-high-performance-deep-learning-library.pdf.
  38. CARLA: A python library to benchmark algorithmic recourse and counterfactual explanation algorithms. In Thirty-fifth Conference on Neural Information Processing Systems Datasets and Benchmarks Track (Round 1), 2021. URL https://openreview.net/forum?id=vDilkBNNbx6.
  39. Judea Pearl. Causality. Cambridge university press, 2009.
  40. Scikit-learn: Machine learning in Python. Journal of Machine Learning Research, 12:2825–2830, 2011.
  41. Preferences in artificial intelligence. Annals of Mathematics and Artificial Intelligence, 77:361–401, 2016.
  42. Face: feasible and actionable counterfactual explanations. In Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society, pp.  344–350, 2020.
  43. Optimal recommendation sets: Covering uncertainty over user preferences. In Manuela M. Veloso and Subbarao Kambhampati (eds.), Proceedings, The Twentieth National Conference on Artificial Intelligence and the Seventeenth Innovative Applications of Artificial Intelligence Conference, July 9-13, 2005, Pittsburgh, Pennsylvania, USA, pp.  541–548. AAAI Press / The MIT Press, 2005. URL http://www.aaai.org/Library/AAAI/2005/aaai05-085.php.
  44. Synthesizing action sequences for modifying model decisions. In AAAI, volume 34, pp.  5462–5469, 2020.
  45. Beyond individualized recourse: Interpretable and interactive summaries of actionable recourses. Advances in Neural Information Processing Systems, 33:12187–12198, 2020.
  46. Christopher D Rosin. Multi-armed bandits with episode context. Annals of Mathematics and Artificial Intelligence, 61(3):203–230, 2011.
  47. Chris Russell. Efficient search for diverse coherent explanations. In Proceedings of the Conference on Fairness, Accountability, and Transparency, FAT* ’19, pp.  20–28, New York, NY, USA, 2019. Association for Computing Machinery. ISBN 9781450361255. doi: 10.1145/3287560.3287569. URL https://doi.org/10.1145/3287560.3287569.
  48. An approach for prediction of loan approval using machine learning algorithm. In 2020 International Conference on Electronics and Sustainable Communication Systems (ICESC), pp.  490–494. IEEE, 2020.
  49. Mastering the game of go with deep neural networks and tree search. nature, 529(7587):484–489, 2016.
  50. Herbert A Simon. A behavioral model of rational choice. The quarterly journal of economics, pp.  99–118, 1955.
  51. Muiltiobjective optimization using nondominated sorting in genetic algorithms. Evolutionary computation, 2(3):221–248, 1994.
  52. A survey of contrastive and counterfactual explanation generation methods for explainable artificial intelligence. IEEE Access, 9:11974–12001, 2021.
  53. S. Tsirtsis and M. Rodriguez. Decisions, counterfactual explanations and strategic behavior. In NeurIPS, 2020. URL https://proceedings.neurips.cc/paper/2020/hash/c2ba1bc54b239208cb37b901c0d3b363-Abstract.html.
  54. Actionable recourse in linear classification. In FAT*, pp.  10–19, 2019.
  55. The philosophical basis of algorithmic recourse. In Proceedings of the 2020 conference on fairness, accountability, and transparency, pp.  284–293, 2020.
  56. Counterfactual explanations and algorithmic recourses for machine learning: A review. arXiv preprint arXiv:2010.10596, 2020.
  57. Amortized generation of sequential algorithmic recourses for black-box models. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 36, pp.  8512–8519, 2022.
  58. Optimal Bayesian recommendation sets and myopically optimal choice query sets. Advances in neural information processing systems, 23, 2010.
  59. On the equivalence of optimal recommendation sets and myopically optimal query sets. Artificial Intelligence, 286:103328, 2020.
  60. Paul Voigt and Axel Von dem Bussche. The EU general data protection regulation (gdpr). A Practical Guide, 1st Ed., Cham: Springer International Publishing, 10(3152676):10–5555, 2017.
  61. On the fairness of causal algorithmic recourse. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 36, pp.  9584–9594, 2022.
  62. Counterfactual explanations without opening the black box: Automated decisions and the gdpr. Harv. JL & Tech., 31:841, 2017.
  63. Gam coach: Towards interactive and user-centered algorithmic recourse. In Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems, pp.  1–20, 2023.
  64. Low-cost algorithmic recourse for users with uncertain cost functions. arXiv preprint arXiv:2111.01235, 2021.
  65. S. Yonadav and W. S. Moses. Extracting incentives from black-box decisions. CoRR, abs/1910.05664, 2019. URL http://arxiv.org/abs/1910.05664.
  66. Adopting machine learning to automatically identify candidate patients for corneal refractive surgery. NPJ digital medicine, 2(1):1–9, 2019.
Citations (5)
View on Semantic Scholar

Summary

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

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