SVARM-IQ: Efficient Approximation of Any-order Shapley Interactions through Stratification

Kolpaczki, Patrick; Muschalik, Maximilian ; Fumagalli, Fabian ; Hammer, Barbara und Hüllermeier, Eyke (Mai 2024): SVARM-IQ: Efficient Approximation of Any-order Shapley Interactions through Stratification. 27th International Conference on Artificial Intelligence and Statistics (AISTATS 2024), Valencia, Spain, 2. - 4. May 2024. In: Proceedings of The 27th International Conference on Artificial Intelligence and Statistics, Proceedings of Machine Learning Research Bd. 238 PMLR. S. 3520-3528

Abstract

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.

• Dokumententyp: Konferenzbeitrag (Paper)
• Fakultät: Mathematik, Informatik und Statistik > Informatik > Künstliche Intelligenz und Maschinelles Lernen
• Themengebiete: 000 Informatik, Informationswissenschaft, allgemeine Werke > 000 Informatik, Wissen, Systeme
• ISSN: 2640-3498
• Sprache: Englisch
• DFG: Gefördert durch die Deutsche Forschungsgemeinschaft (DFG) - 438445824