SVARM-IQ: Efficient Approximation of Any-order Shapley Interactions through Stratification
Kolpaczki P, Muschalik M, Fumagalli F, Hammer B, Hüllermeier E (2024)
In: Proceedings of The 27th International Conference on Artificial Intelligence and Statistics. Dasgupta S, Mandt S, Li Y (Eds); Proceedings of Machine Learning Research, 238. PMLR: 3520-3528.
Konferenzbeitrag | Englisch
Download
Es wurden keine Dateien hochgeladen. Nur Publikationsnachweis!
Autor*in
Kolpaczki, Patrick;
Muschalik, Maximilian;
Fumagalli, FabianUniBi
;
Hammer, BarbaraUniBi
;
Hüllermeier, Eyke
![](https://pub.uni-bielefeld.de/images/icon_orcid.png)
![](https://pub.uni-bielefeld.de/images/icon_orcid.png)
Herausgeber*in
Dasgupta, Sanjoy;
Mandt, Stephan;
Li, Yingzhen
Einrichtung
SFB/Transregio 318 Constructing Explainability > Projektbereich C: Darstellung und Berechnung von Erklärungen > Teilprojekt C01: Gesundes Misstrauen in und durch Erklärungen
SFB/Transregio 318 Constructing Explainability > Projektbereich C: Darstellung und Berechnung von Erklärungen > Teilprojekt C03: Interpretierbares maschinelles Lernen: Erklärbarkeit in dynamischen Umgebungen
Center of Excellence - Cognitive Interaction Technology CITEC > Machine Learning
Technische Fakultät > AG Machine Learning
SFB/Transregio 318 Constructing Explainability > Projektbereich C: Darstellung und Berechnung von Erklärungen > Teilprojekt C03: Interpretierbares maschinelles Lernen: Erklärbarkeit in dynamischen Umgebungen
Center of Excellence - Cognitive Interaction Technology CITEC > Machine Learning
Technische Fakultät > AG Machine Learning
Projekt
Abstract / Bemerkung
Addressing the limitations of individual attribution scores via the Shapley value (SV), the field of explainable AI (XAI) has recently explored intricate interactions of features or data points. In particular, extensions of the SV, such as the Shapley Interaction Index (SII), have been proposed as a measure to still benefit from the axiomatic basis of the SV. However, similar to the SV, their exact computation remains computationally prohibitive. Hence, we propose with SVARM-IQ a sampling-based approach to efficiently approximate Shapley-based interaction indices of any order. SVARM-IQ can be applied to a broad class of interaction indices, including the SII, by leveraging a novel stratified representation. We provide non-asymptotic theoretical guarantees on its approximation quality and empirically demonstrate that SVARM-IQ achieves state-of-the-art estimation results in practical XAI scenarios on different model classes and application domains.
Erscheinungsjahr
2024
Titel des Konferenzbandes
Proceedings of The 27th International Conference on Artificial Intelligence and Statistics
Serien- oder Zeitschriftentitel
Proceedings of Machine Learning Research
Band
238
Seite(n)
3520-3528
Page URI
https://pub.uni-bielefeld.de/record/2991394
Zitieren
Kolpaczki P, Muschalik M, Fumagalli F, Hammer B, Hüllermeier E. SVARM-IQ: Efficient Approximation of Any-order Shapley Interactions through Stratification. In: Dasgupta S, Mandt S, Li Y, eds. Proceedings of The 27th International Conference on Artificial Intelligence and Statistics. Proceedings of Machine Learning Research. Vol 238. PMLR; 2024: 3520-3528.
Kolpaczki, P., Muschalik, M., Fumagalli, F., Hammer, B., & Hüllermeier, E. (2024). SVARM-IQ: Efficient Approximation of Any-order Shapley Interactions through Stratification. In S. Dasgupta, S. Mandt, & Y. Li (Eds.), Proceedings of Machine Learning Research: Vol. 238. Proceedings of The 27th International Conference on Artificial Intelligence and Statistics (pp. 3520-3528). PMLR.
Kolpaczki, Patrick, Muschalik, Maximilian, Fumagalli, Fabian, Hammer, Barbara, and Hüllermeier, Eyke. 2024. “SVARM-IQ: Efficient Approximation of Any-order Shapley Interactions through Stratification”. In Proceedings of The 27th International Conference on Artificial Intelligence and Statistics, ed. Sanjoy Dasgupta, Stephan Mandt, and Yingzhen Li, 238:3520-3528. Proceedings of Machine Learning Research. PMLR.
Kolpaczki, P., Muschalik, M., Fumagalli, F., Hammer, B., and Hüllermeier, E. (2024). “SVARM-IQ: Efficient Approximation of Any-order Shapley Interactions through Stratification” in Proceedings of The 27th International Conference on Artificial Intelligence and Statistics, Dasgupta, S., Mandt, S., and Li, Y. eds. Proceedings of Machine Learning Research, vol. 238, (PMLR), 3520-3528.
Kolpaczki, P., et al., 2024. SVARM-IQ: Efficient Approximation of Any-order Shapley Interactions through Stratification. In S. Dasgupta, S. Mandt, & Y. Li, eds. Proceedings of The 27th International Conference on Artificial Intelligence and Statistics. Proceedings of Machine Learning Research. no.238 PMLR, pp. 3520-3528.
P. Kolpaczki, et al., “SVARM-IQ: Efficient Approximation of Any-order Shapley Interactions through Stratification”, Proceedings of The 27th International Conference on Artificial Intelligence and Statistics, S. Dasgupta, S. Mandt, and Y. Li, eds., Proceedings of Machine Learning Research, vol. 238, PMLR, 2024, pp.3520-3528.
Kolpaczki, P., Muschalik, M., Fumagalli, F., Hammer, B., Hüllermeier, E.: SVARM-IQ: Efficient Approximation of Any-order Shapley Interactions through Stratification. In: Dasgupta, S., Mandt, S., and Li, Y. (eds.) Proceedings of The 27th International Conference on Artificial Intelligence and Statistics. Proceedings of Machine Learning Research. 238, p. 3520-3528. PMLR (2024).
Kolpaczki, Patrick, Muschalik, Maximilian, Fumagalli, Fabian, Hammer, Barbara, and Hüllermeier, Eyke. “SVARM-IQ: Efficient Approximation of Any-order Shapley Interactions through Stratification”. Proceedings of The 27th International Conference on Artificial Intelligence and Statistics. Ed. Sanjoy Dasgupta, Stephan Mandt, and Yingzhen Li. PMLR, 2024.Vol. 238. Proceedings of Machine Learning Research. 3520-3528.