Metadata-Version: 2.4
Name: evaluation_toolkit_for_model_agnostic_methods
Version: 1.0.0
Summary: A package for evaluating model agnostic interpretable methods. Currently supports global and local methods for regression use cases.
Home-page: https://git-ce.rwth-aachen.de/wzl-iqs3/quality-insights/publications/evaluation-toolkit-for-model-agnostic-methods
Author: Tobias Schulze, Daniel Buschmann, Felix Sohnius, Robert H. Schmitt
Author-email: 
Classifier: Programming Language :: Python :: 3.10
Classifier: Operating System :: OS Independent
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: jupyter
Requires-Dist: numpy==1.23.5
Requires-Dist: pandas
Requires-Dist: plotly
Requires-Dist: rbo
Requires-Dist: matplotlib
Requires-Dist: seaborn
Requires-Dist: scikit-learn
Requires-Dist: model-agnostic-toolkit
Provides-Extra: docs
Requires-Dist: sphinx>=7.0; extra == "docs"
Requires-Dist: myst-parser>=3.0; extra == "docs"
Requires-Dist: sphinx-rtd-theme>=2.0; extra == "docs"
Requires-Dist: sphinx-autodoc-typehints>=2.0; extra == "docs"
Requires-Dist: sphinxcontrib-mermaid>=0.9; extra == "docs"
Dynamic: author
Dynamic: classifier
Dynamic: description
Dynamic: description-content-type
Dynamic: home-page
Dynamic: license-file
Dynamic: provides-extra
Dynamic: requires-dist
Dynamic: summary


The Evaluation Toolkit for Model Agnostic Methods is a package for evaluating model-agnostic
interpretable methods built on top of the Model Agnostic Toolkit, for both global and local
(individual-instance) explanations on regression and classification tasks.

It includes metrics for:
- Correctness of feature importance and interaction results
- Stability and repeatability across repeated runs
- Real-time capability (duration and scalability)
- Quality of local explanations for individual instances

For more details, please refer to the project documentation.
