Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
184 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

Learning Actionable Counterfactual Explanations in Large State Spaces (2404.17034v3)

Published 25 Apr 2024 in cs.LG

Abstract: Recourse generators provide actionable insights, often through feature-based counterfactual explanations (CFEs), to help negatively classified individuals understand how to adjust their input features to achieve a positive classification. These feature-based CFEs, which we refer to as \emph{low-level} CFEs, are overly specific (e.g., coding experience: (4 \to 5+) years) and often recommended in a feature space that doesn't straightforwardly align with real-world actions. To bridge this gap, we introduce three novel recourse types grounded in real-world actions: high-level continuous (\emph{hl-continuous}), high-level discrete (\emph{hl-discrete}), and high-level ID (\emph{hl-id}) CFEs. We formulate single-agent CFE generation methods, where we model the hl-discrete CFE as a solution to a weighted set cover problem and the hl-continuous CFE as a solution to an integer linear program. Since these methods require costly optimization per agent, we propose data-driven CFE generation approaches that, given instances of agents and their optimal CFEs, learn a CFE generator that quickly provides optimal CFEs for new agents. This approach, also viewed as one of learning an optimal policy in a family of large but deterministic MDPs, considers several problem formulations, including formulations in which the actions and their effects are unknown, and therefore addresses informational and computational challenges. We conduct extensive empirical evaluations using healthcare datasets (BRFSS, Foods, and NHANES) and fully-synthetic data. For negatively classified agents identified by linear or threshold-based classifiers, we compare the high-level CFE to low-level CFEs and assess the effectiveness of our network-based, data-driven approaches. Results show that the data-driven CFE generators are accurate, and resource-efficient, and high-level CFEs offer key advantages over low-level CFEs.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (26)
  1. A rewriting system for convex optimization problems. Journal of Control and Decision, 5(1):42–60, 2018.
  2. Maria-Florina Balcan. Data-driven algorithm design. In Tim Roughgarden, editor, Beyond the Worst-Case Analysis of Algorithms, pages 626–645. Cambridge University Press, 2020. doi: 10.1017/9781108637435.036. URL https://doi.org/10.1017/9781108637435.036.
  3. Data-driven clustering via parameterized Lloyd’s families. In Samy Bengio, Hanna M. Wallach, Hugo Larochelle, Kristen Grauman, Nicolò Cesa-Bianchi, and Roman Garnett, editors, Advances in Neural Information Processing Systems (NeurIPS), pages 10664–10674, 2018. URL https://proceedings.neurips.cc/paper/2018/hash/128ac9c427302b7a64314fc4593430b2-Abstract.html.
  4. Leopoldo E. Bertossi. An asp-based approach to counterfactual explanations for classification. CoRR, abs/2004.13237, 2020. URL https://arxiv.org/abs/2004.13237.
  5. Optimal action extraction for random forests and boosted trees. In Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD), pages 179–188, 2015. ISBN 9781450336642. doi: 10.1145/2783258.2783281. URL https://doi.org/10.1145/2783258.2783281.
  6. Multi-objective counterfactual explanations. In Proceedings of the International Conference on Parallel Problem Solving from Nature (PPSN), pages 448–469, 2020. ISBN 978-3-030-58111-4. doi: 10.1007/978-3-030-58112-1\textunderscore31. URL https://doi.org/10.1007/978-3-030-58112-1_31.
  7. Synthesizing explainable counterfactual policies for algorithmic recourse with program synthesis. Machine Learning, 112(4):1389–1409, 2023. ISSN 0885-6125. doi: 10.1007/s10994-022-06293-7. URL https://doi.org/10.1007/s10994-022-06293-7.
  8. CVXPY: A Python-embedded modeling language for convex optimization. Journal of Machine Learning Research, 17(83):1–5, 2016.
  9. European Parliament and Council of the EU. Regulation (EU) 2016/679 of the European Parliament and of the Council, 2016. URL https://data.europa.eu/eli/reg/2016/679/oj.
  10. A PAC approach to application-specific algorithm selection. In Proceedings of the ACM Conference on Innovations in Theoretical Computer Science (ITCS), pages 123–134, 2016. ISBN 9781450340571. doi: 10.1145/2840728.2840766. URL https://doi.org/10.1145/2840728.2840766.
  11. Equalizing recourse across groups. arXiv preprint arXiv:1909.03166, 2019. URL http://arxiv.org/abs/1909.03166.
  12. Towards realistic individual recourse and actionable explanations in black-box decision making systems. arXiv preprint arXiv:1907.09615, 2019. URL http://arxiv.org/abs/1907.09615.
  13. Algorithmic recourse: From counterfactual explanations to interventions. In Proceedings of the ACM Conference on Fairness, Accountability, and Transparency (ACM FAccT), pages 353–362, 2021. ISBN 9781450383097. doi: 10.1145/3442188.3445899. URL https://doi.org/10.1145/3442188.3445899.
  14. A survey of algorithmic recourse: Contrastive explanations and consequential recommendations. ACM Computing Surveys, 55(5), 2022. ISSN 0360-0300. doi: 10.1145/3527848. URL https://doi.org/10.1145/3527848.
  15. Richard M. Karp. Reducibility Among Combinatorial Problems, pages 85–103. Springer US, Boston, MA, 1972. ISBN 978-1-4684-2001-2. doi: 10.1007/978-1-4684-2001-2\textunderscore9. URL https://doi.org/10.1007/978-1-4684-2001-2_9.
  16. Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980, 2014.
  17. A unified logical framework for explanations in classifier systems. Journal of Logic and Computation, 33(2):485–515, 2023. doi: 10.1093/logcom/exac102.
  18. Joao Marques-Silva. Logic-based explainability in machine learning, 2023.
  19. Explaining machine learning classifiers through diverse counterfactual explanations. In Proceedings of the ACM Conference on Fairness, Accountability, and Transparency (ACM FAccT), pages 607–617, 2020. ISBN 9781450369367. doi: 10.1145/3351095.3372850. URL https://doi.org/10.1145/3351095.3372850.
  20. Consequence-aware sequential counterfactual generation. In Proceedings of the European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML PKDD), pages 682–698, 2021. ISBN 978-3-030-86519-1. doi: 10.1007/978-3-030-86520-7\textunderscore42. URL https://doi.org/10.1007/978-3-030-86520-7_42.
  21. Synthesizing action sequences for modifying model decisions. arXiv preprint arXiv:1910.00057, 2019. URL http://arxiv.org/abs/1910.00057.
  22. Extracting incentives from black-box decisions. arXiv preprint arXiv:1910.05664, 2019. URL http://arxiv.org/abs/1910.05664.
  23. Actionable recourse in linear classification. In Proceedings of the ACM Conference on Fairness, Accountability, and Transparency (ACM FAccT), pages 10–19, 2019. ISBN 9781450361255. doi: 10.1145/3287560.3287566. URL https://doi.org/10.1145/3287560.3287566.
  24. V.V. Vazirani. Approximation Algorithms. Springer Berlin Heidelberg, 2013. ISBN 9783662045657. URL https://books.google.com/books?id=bJmqCAAAQBAJ.
  25. Amortized generation of sequential algorithmic recourses for black-box models. In Proceedings of the National Conference on Artificial Intelligence (AAAI), pages 8512–8519, 2022. doi: 10.1609/aaai.v36i8.20828. URL https://ojs.aaai.org/index.php/AAAI/article/view/20828.
  26. Counterfactual explanations without opening the black box: Automated decisions and the GDPR. arXiv preprint arXiv:1711.00399, 2017. URL http://arxiv.org/abs/1711.00399.

Summary

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

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

Tweets