SHAP methods (SHapley Additive exPlanations)

Local Explainability

Level of interpretation aiming to explain a model's prediction for a specific observation, isolating the impact of each feature for that particular case.

← Geri