0.4

Compute weighted mean SHAP values at higher levels such as factor (cluster of correlated items) or domains (cluster of factors).

0.3

• checking the models' performance before including their vote on feature importance
• supporting both classification and regression models