SHAP (SHapley Additive exPlanations)
A feature attribution method rooted in game theory used to interpret model predictions by quantifying individual feature contributions.
Source: On-Chain Behavioral Signatures for Classifying Centralized Exchange Addresses on Ethereum Using Proof-of-Reserves (Academic research) · p.1
Have a question about SHAP (SHapley Additive exPlanations)? Ask the AI, answers cite the source.