1 paper
Hui Duan, Giray Ökten
We introduce a new Shapley value approach for global sensitivity analysis and machine learning explainability. The method is based on the first-order partial derivatives of the und…