Skip to content

dylan-slack/fairness-warnings-fair-maml

master
Switch branches/tags

Name already in use

A tag already exists with the provided branch name. Many Git commands accept both tag and branch names, so creating this branch may cause unexpected behavior. Are you sure you want to create this branch?
Code

Latest commit

 

Git stats

Files

Permalink
Failed to load latest commit information.
Type
Name
Latest commit message
Commit time
 
 
 
 
 
 
 
 

Fairness Warnings & Fair-MAML: Learning Fairly from Minimal Data

This is the code for our paper, "Fairness Warnings & Fair-MAML: Learning Fairly from Minimal Data."

Checkout the full paper.

Getting started

Setup virtual environment and install requirements:

conda create -n fairwarnmaml python=3.7
source activate fairwarnmaml
pip install -r requirements.txt

To run Fairness Warnings, you'll need a full version of CPLEX. For students and faculty members, install a free full version using the instructions from here.

Run compas_warning_example.py to generate a fairness warning for the COMPAS data set. Detailed read outs of the SLIM results will be given in a folder called ./SLIMLOGS that will be created on run.

Run fair_maml_cc_example.py to sweep over a range of gammas for Fair-MAML on the communities and crime task.

References

Please consider citing our paper if you found this work useful!

@article{Slack2019FairWarningsFairMAML},
	title={Fairness Warnings and Fair-MAML: Learning Fairly with Minimal Data},
	author={Dylan Slack and Sorelle Friedler and Emile Givental.},
	journal={Proceedings of the Conference on Fairness, Accountability, and Transparency (FAT*)},
	year={2020},
}

Citations

The fair-MAML code in fair_maml_cc_example.py is modified from a MAML implementation from github user Jakie Loong found in this repository. Please check out the original implementation as its quite elegant!

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages