Published Python Packages
This page summarizes my published Python packages for Explainable AI (XAI), Computer Vision, and reproducible machine learning research.
📦 Package Summary
| Package | Domain | Latest Version | Registry | Links |
|---|---|---|---|---|
| lime-stratified | Explainable AI | … | PyPI | PyPI · Code |
| shap-bpt | Explainable AI | … | PyPI | PyPI · Code · Docs |
🌿 lime-stratified
Overview
lime-stratified is a Python package that improves the stability of LIME image explanations using a novel stratified sampling strategy proposed in my AAAI 2024 paper.
Features
- Improved stability of LIME explanations
- Reduced perturbation variance
- Compatible with the original LIME API
- PyPI package
- Open source
Installation
pip install lime-stratified
Resources
PyPI Source Code Examples Publication
🌳 shap-bpt
Overview
shap-bpt implements ShapBPT, a hierarchical image feature attribution method based on Binary Partition Trees, introduced at AAAI 2026.
Features
- Hierarchical Shapley explanations
- Data-aware Binary Partition Trees
- Faster image explanations
- Explainable AI for Computer Vision
- PyPI package
- Comprehensive documentation
Installation
pip install shap-bpt
Resources
PyPI Source Code Documentation Publication Demo
📈 Package Statistics
| Package | Latest Version | Python | License | Downloads |
|---|---|---|---|---|
| lime-stratified | ... | ... | ... | View downloads |
| shap-bpt | ... | ... | ... | View downloads |