Machine Learning Explainability Experiments

This repository contains experimental marimo notebooks exploring explainable machine learning techniques, particularly focusing on SHAP (SHapley Additive exPlanations) values and their application to different models. Make sure to install marimo so you can run the code.

Contents

The repository includes Marimo notebooks demonstrating:

1. Model Comparison with SHAP

   • Comparison of explainability across different models:
     • Linear Regression
     • Explainable Boosting Machine (EBM)
     • XGBoost
   • SHAP value visualizations for each model
   • Feature importance analysis

2. EBM with SHAP

   • Focused exploration of Explainable Boosting Machine
   • Partial dependence plots with SHAP values
   • Individual prediction explanations

Key Features

• Interactive Marimo notebooks
• California Housing dataset examples
• SHAP visualizations including:
  • Partial dependence plots
  • Beeswarm plots
  • Waterfall charts
• Model comparison insights

Requirements

To run these notebooks, you'll need:

• Python 3.8+
• Marimo
• SHAP
• Scikit-learn
• InterpretML (for EBM)
• XGBoost

Getting Started

1. Install dependencies:

   pip install marimo shap interpret xgboost scikit-learn

2. Launch Marimo:

   marimo edit

3. Open the desired notebook from the list

License

This project is licensed under the MIT License - see the LICENSE file for details.