justicia 0.0.4

This library can be used formally verify machine learning models on multiple fairness definitions.

Justicia

This is the implementation of our AAAI-2021 paper where we have proposed a SSAT-based approach to formally verify fairness in machine learning.

Install

• Install the python library pip install justicia
• Install other python dependencies pip install -r requirements.txt

Other dependencies

• SSAT solver. Checkout to compatible version.

  git clone https://github.com/NTU-ALComLab/ssatABC.git
  cd ssatABC
  git checkout 91a93a57c08812e3fe24aabd71219b744d2355ad
  

• PySAT

• Notears

Documentation

Python tutorials are available in doc.

Citations

Please cite the following paper.

@inproceedings{ghosh2020justicia,
author={Ghosh, Bishwamittra and Basu, Debabrota and Meel, Kuldeep S.},
title={Justicia: A Stochastic {SAT} Approach to Formally Verify Fairness},
booktitle={Proceedings of AAAI},
month={2},
year={2021},
}

Issues, questions, bugs, etc.

Please click on "issues" at the top and create a new issue. All issues are responded to promptly.