2025/08/18 by Michael Mayer, Mayer, Michael, Mario V. Wüthrich +1 · 1 citation
Computer Science · #Advanced Clustering Algorithms Research #Bayesian Methods and Mixture Models #Computational Engineering #FOS: Computer and information sciences #Finance #Importance sampling #Kernel (algebra) #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Order (exchange) #Sample (material) #Sampling (signal processing) #Shapley value #and Science (cs.CE)
paper · pdf · doi:10.48550/arxiv.2508.12947
published in arXiv (Cornell University) (Cornell University)
openalex publication_date 2025/08/18 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/05
Originally introduced in cooperative game theory, Shapley values have become a very popular tool to explain machine learning predictions. Based on Shapley's fairness axioms, every input (feature component) gets a credit how it contributes to an output (prediction). These credits are then used to explain the prediction. The only limitation in computing the Shapley values (credits) for many different predictions is of computational nature. There are two popular sampling approximations, sampling KernelSHAP and sampling PermutationSHAP. Our first novel contributions are asymptotic normality results for these sampling approximations. Next, we show that the paired-sampling approaches provide exact results in case of interactions being of maximal order two. Furthermore, the paired-sampling PermutationSHAP possesses the additive recovery property, whereas its kernel counterpart does not.