Viet-Anh on Software Logo

What is: Shapley Additive Explanations?

SourceA Unified Approach to Interpreting Model Predictions
Year2000
Data SourceCC BY-SA - https://paperswithcode.com

SHAP, or SHapley Additive exPlanations, is a game theoretic approach to explain the output of any machine learning model. It connects optimal credit allocation with local explanations using the classic Shapley values from game theory and their related extensions. Shapley values are approximating using Kernel SHAP, which uses a weighting kernel for the approximation, and DeepSHAP, which uses DeepLift to approximate them.