bosaj/xai-model-interpretability-suite

ENIAD XAI Model Interpretability Suite 🔍⚡

🎓 École Nationale de l'Intelligence Artificielle et du Digital (ENIAD) • UMP

Author: Oussama EL HADJI (@Bosaj)

An interactive, public Explainable AI (XAI) suite providing local and global interpretability for Machine Learning and Computer Vision models:

• —SHAP (SHapley Additive exPlanations): Interactive Waterfall plots and feature attribution
• —LIME (Local Interpretable Model-agnostic Explanations): Local linear surrogates on tabular instances
• —Grad-CAM (Gradient-weighted Class Activation Mapping): CNN activation heatmaps on image features
• —PDP & ALE: Partial Dependence and Accumulated Local Effects curves

🔗 Live Application & Source

• —Live Interactive Space: https://huggingface.co/spaces/bosaj/xai-model-interpretability-suite
• —Direct Web App: https://bosaj-xai-model-interpretability-suite.static.hf.space
• —GitHub Repository: Bosaj/xai-model-interpretability-suite