1 paper
Ali Gorji, Andisheh Amrollahi, Andreas Krause
SHAP (SHapley Additive exPlanations) values are a widely used method for local feature attribution in interpretable and explainable AI. We propose an efficient two-stage algorithm…