Interprets the feature importance of each model

Interprets the feature importance of each model using DALEX and feature importance permuation

Usage

InterpretModels(output_dir = "./classifiers", plot_type = "ratio")

Arguments

output_dir

The output directory that TrainAllModels saved training data and models into

plot_type

Argument to pass to model_parts(). Ratio or difference is recommended for large feature sets where the base model's AUC loss will be outside the plotting range.