Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
80 tokens/sec
GPT-4o
59 tokens/sec
Gemini 2.5 Pro Pro
43 tokens/sec
o3 Pro
7 tokens/sec
GPT-4.1 Pro
50 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

TraCE: Trajectory Counterfactual Explanation Scores (2309.15965v2)

Published 27 Sep 2023 in cs.LG, cs.CY, and math.MG

Abstract: Counterfactual explanations, and their associated algorithmic recourse, are typically leveraged to understand, explain, and potentially alter a prediction coming from a black-box classifier. In this paper, we propose to extend the use of counterfactuals to evaluate progress in sequential decision making tasks. To this end, we introduce a model-agnostic modular framework, TraCE (Trajectory Counterfactual Explanation) scores, which is able to distill and condense progress in highly complex scenarios into a single value. We demonstrate TraCE's utility across domains by showcasing its main properties in two case studies spanning healthcare and climate change.

PDF HTML Abstract
Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Bookmark Chat Bubble Oval Streamline Icon: https://streamlinehq.com Chat (Pro)
Definition Search Book Streamline Icon: https://streamlinehq.com
References (27)
  1. Counterfactual explanations for multivariate time series. In 2021 International Conference on Applied Artificial Intelligence (ICAPAI), pages 1–8. IEEE, 2021. 10.1109/ICAPAI49758.2021.9462056.
  2. Copernicus Climate Change Service, Climate Data Store. Methane data from 2002 to present derived from satellite observations. Copernicus Climate Change Service (C3S) Climate Data Store (CDS), 2018. Accessed on 01-09-2023.
  3. Copernicus Climate Change Service, Climate Data Store. CMIP6 Climate Projections. Copernicus Climate Change Service (C3S) Climate Data Store (CDS), 2021. Accessed on 17-08-2023.
  4. Instance-based counterfactual explanations for time series classification. In International Conference on Case-Based Reasoning, pages 32–47. Springer, 2021. 10.1007/978-3-030-86957-1_3.
  5. Long-term economic growth projections in the shared socioeconomic pathways. Global Environmental Change, 42:200–214, Jan 2017. 10.1016/j.gloenvcha.2015.06.004.
  6. N. G. I. for Space Studies (NASA/GISS). Nasa-giss giss-e2.1h model output prepared for cmip6 scenariomip ssp126, 2020a. URL https://doi.org/10.22033/ESGF/CMIP6.7411.
  7. N. G. I. for Space Studies (NASA/GISS). Nasa-giss giss-e2.1h model output prepared for cmip6 scenariomip ssp245, 2020b. URL https://doi.org/10.22033/ESGF/CMIP6.7416.
  8. N. G. I. for Space Studies (NASA/GISS). Nasa-giss giss-e2.1h model output prepared for cmip6 scenariomip ssp370, 2020c. URL https://doi.org/10.22033/ESGF/CMIP6.7427.
  9. N. G. I. for Space Studies (NASA/GISS). Nasa-giss giss-e2.1h model output prepared for cmip6 scenariomip ssp460, 2020d. URL https://doi.org/10.22033/ESGF/CMIP6.7453.
  10. N. G. I. for Space Studies (NASA/GISS). Nasa-giss giss-e2.1h model output prepared for cmip6 scenariomip ssp585, 2020e. URL https://doi.org/10.22033/ESGF/CMIP6.7461.
  11. Early warning scores for detecting deterioration in adult hospital patients: systematic review and critical appraisal of methodology. bmj, 369, 2020. 10.1136/bmj.m1501.
  12. R. Guidotti. Counterfactual explanations and how to find them: literature review and benchmarking. Data Mining and Knowledge Discovery, pages 1–55, 2022. 10.1007/s10618-022-00831-6.
  13. Era5 monthly averaged data on single levels from 1940 to present. Copernicus Climate Change Service (C3S) Climate Data Store (CDS), 2023. Accessed on 17-08-2023.
  14. Tsevo: Evolutionary counterfactual explanations for time series classification. In 2022 21st IEEE International Conference on Machine Learning and Applications (ICMLA), pages 29–36. IEEE, 2022. 10.1109/ICMLA55696.2022.00013.
  15. Mimic-iv (version 2.0). PhysioNet., 2022. 10.13026/7vcr-e114.
  16. S. KC and W. Lutz. The human core of the shared socioeconomic pathways: Population scenarios by age, sex and level of education for all countries to 2100. Global Environmental Change, 42:181–192, Jan 2017. 10.1016/j.gloenvcha.2014.06.004.
  17. Towards a decision support tool for intensive care discharge: machine learning algorithm development using electronic healthcare data from mimic-iii and bristol, uk. BMJ open, 9(3):e025925, 2019. 10.1136/bmjopen-2018-025925.
  18. OECD. Gross Domestic Product (GDP) (indicator), 2023a. URL https://doi.org/10.1787/dc2f7aec-en. Accessed on 22-08-2023.
  19. OECD. Historical Population, 2023b. ISSN 20744390 (online). URL https://doi.org/10.1787/data-00285-en. Accessed on 22-08-2023.
  20. The roads ahead: Narratives for shared socioeconomic pathways describing world futures in the 21st century. Global Environmental Change, 42:169–180, 2017. https://doi.org/10.1016/j.gloenvcha.2015.01.004.
  21. Face: feasible and actionable counterfactual explanations. In Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society, pages 344–350, 2020.
  22. The shared socioeconomic pathways and their energy, land use, and greenhouse gas emissions implications: An overview. Global Environmental Change, 42:153–168, Jan 2017. 10.1016/j.gloenvcha.2016.05.009.
  23. Navigating explanatory multiverse through counterfactual path geometry. In International Conference on Machine Learning Workshop on Counterfactuals in Minds and Machines, 2023.
  24. Counterfactual explanations in sequential decision making under uncertainty. Advances in Neural Information Processing Systems, 34:30127–30139, 2021.
  25. M. Virgolin and S. Fracaros. On the robustness of sparse counterfactual explanations to adverse perturbations. Artificial Intelligence, 316:103840, 2023. ISSN 0004-3702. https://doi.org/10.1016/j.artint.2022.103840. URL https://www.sciencedirect.com/science/article/pii/S0004370222001801.
  26. Counterfactual explanations without opening the black box: automated decisions and the gdpr. Harvard Journal of Law and Technology, 31(2):841–887, 2018.
  27. Counterfactual explanations for survival prediction of cardiovascular icu patients. In Artificial Intelligence in Medicine: 19th International Conference on Artificial Intelligence in Medicine, AIME 2021, Virtual Event, June 15–18, 2021, Proceedings, pages 338–348. Springer, 2021. 10.1007/978-3-030-77211-6_38.
User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (8)
  1. Jeffrey N. Clark (7 papers)
  2. Edward A. Small (4 papers)
  3. Nawid Keshtmand (6 papers)
  4. Michelle W. L. Wan (2 papers)
  5. Elena Fillola Mayoral (2 papers)
  6. Enrico Werner (3 papers)
  7. Christopher P. Bourdeaux (3 papers)
  8. Raul Santos-Rodriguez (70 papers)
Citations (1)
View on Semantic Scholar