Exact Shapley Attributions in Quadratic-time for FANOVA Gaussian Processes
Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (29) · 24414-24421
DOI: 10.1609/aaai.v40i29.39623 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Majid Mohammadi und 4 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref).
Abstract (Originalsprache)
Shapley values are widely recognized as a principled method for attributing importance to input features in machine learning. However, the exact computation of Shapley values scales exponentially with the number of features, severely limiting the practical application of this powerful approach. The challenge is further compounded when the predictive model is probabilistic---as in Gaussian processes (GPs)---where the outputs are random variables rather than point estimates, necessitating additional computational effort in modeling higher-order moments. In this work, we demonstrate that for an important class of GPs known as FANOVA GP, which explicitly models all main effects and interactions, exact Shapley attributions for both local and global explanations can be computed in *quadratic* time. For *local, instance-wise explanations*, we define a stochastic cooperative game over function components and compute the *exact stochastic Shapley value* in quadratic time only, capturing both the expected contribution and uncertainty. For *global explanations*, we introduce a deterministic, variance-based value function and compute exact Shapley values that quantify each feature’s contribution to the model’s overall sensitivity. Our methods leverage a closed-form (stochastic) Möbiusrepresentation of the FANOVA decomposition and introduce recursive algorithms, inspired by Newton's identities, to efficiently compute the mean and variance of Shapley values. Our work enhances the utility of explainable AI, as demonstrated by empirical studies, by providing more scalable, axiomatically sound, and uncertainty-aware explanations for predictions generated by structured probabilistic models.
Zitieren
APA
Mohammadi, M., Muandet, K., Tiddi, I., Teije, A. T., & Chau, S. L. (2026). Exact Shapley Attributions in Quadratic-time for FANOVA Gaussian Processes. Proceedings of the AAAI Conference on Artificial Intelligence, 40(29), 24414-24421. https://doi.org/10.1609/aaai.v40i29.39623
Vancouver
Mohammadi M, Muandet K, Tiddi I, Teije AT, Chau SL. Exact Shapley Attributions in Quadratic-time for FANOVA Gaussian Processes. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(29):24414-24421. doi:10.1609/aaai.v40i29.39623
BibTeX
@article{Mohammadi2026,
author = {Mohammadi, Majid and Muandet, Krikamol and Tiddi, Ilaria and Teije, Annette Ten and Chau, Siu Lun},
title = {Exact Shapley Attributions in Quadratic-time for FANOVA Gaussian Processes},
journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
year = {2026},
volume = {40},
number = {29},
pages = {24414--24421},
doi = {10.1609/aaai.v40i29.39623}
}
EUVIMED