- Contrastive for BT classification (binary classes)
- Remove the PyQt6 dependencie and new methods to display explanations:
- show_in_notebook()
- show_on_screen()
- get_PILImage()
- save_png()
- resize_PILimage()
- Change function name in explainer (unset_specific_features -> unset_excluded_features)
- New procedure installation
- New visualization for time series
- Compilation error resolution
- New metrics (documentation in progress)
- For binary classification:
- accuracy
- precision
- recall
- f1_score
- specificity
- tp, tn, fp, fn
- For multiclass classification:
- micro_averaging_accuracy
- micro_averaging_precision
- micro_averaging_recall
- macro_averaging_accuracy
- macro_averaging_precision
- macro_averaging_recall
- For regression:
- mean_squared_error
- root_mean_squared_error
- mean_absolute_error
- Build and Tests with CI
- Regression for boosted trees
- Adding thoeries
- Easier import model
- Graphical user interface: displaying, loading, saving explanations
- Data preprocessing
- unit tests
- Initial release