mlexplainer 1.0.1

Advanced Machine Learning package dedicated to model interpretability, with a primary focus on leveraging Shapley values for explaining complex predictive models.

MLExplainer

Advanced Machine Learning package dedicated to model interpretability, with a primary focus on leveraging Shapley values for explaining complex predictive models.

ReadTheDocs

Check out the package documentation on ReadTheDocs!

Demo

Check out the demo on Streamlit

Install

MLExplainer can be installed from PyPI:

pip install mlexplainer

Why MLExplainer ?

MlExplainer is a tool designed to tackle a major challenge in advanced Machine Learning for tabular data.

Indeed, advanced Machine Learning models - such as Boosting algorithms - are often perceived as black boxes. Today, our goal is to provide a way to demystify these models by leveraging mathematical concepts, making it possible to clearly explain how they work to any audience.

Key Features :

• SHAP Integration: Built-in support for SHAP explainers with optimized workflows
• Multiple Classification Types: Support for binary and multilabel classification tasks
• Automatic Feature Detection: Intelligent categorization of numerical, categorical, and string features
• Rich Visualizations: Integrated plotting for feature-target relationships and SHAP value distributions
• Validation Tools: Built-in interpretation consistency validation
• Modern Architecture: Clean, extensible design with proper abstractions

About

If you encounter any issues with the code or have suggestions for improvement, feel free to open an issue or contact me.