search

ashryaagr / Fairness.jl

Julia Toolkit with fairness metrics and bias mitigation algorithms
https://ashryaagr.github.io/Fairness.jl/dev/
MIT License
31 stars 14 forks source link

Added Calibrated Equalised Odds Wrapper. #64

Closed sumantrak closed 3 years ago

sumantrak commented 3 years ago

Evaluated it with COMPAS dataset giving acceptable results image

ashryaagr commented 3 years ago

Please update the documentation as well for this algorithm here and add this algorithm to the table in the README.

ashryaagr commented 3 years ago

Thanks a lot for the Pull Request! Calibrated Equalized Odds algorithm wrapper is an excellent contribution. Everything (docs, readme, tests, code, readability), looks perfect to me.

I am merging this Pull Request.

This algorithm's implementation is for binary protected attribute. It would be another great contribution to add multi-class support for this algorithm. Feel free to open a separate Pull Request for it!

sumantrak commented 3 years ago

Thanks a lot! It really means a lot to me. I will try to add support for mult-class as soon as possible.

On Wed, 31 Mar 2021 at 13:43, Ashrya Agrawal @.***> wrote:

Thanks a lot for the Pull Request! Calibrated Equalized Odds algorithm wrapper is an excellent contribution. Everything (docs, readme, tests, code, readability), looks perfect to me.

I am merging this Pull Request.

This algorithm's implementation is for binary protected attribute. It would be another great contribution to add multi-class support for this algorithm. Feel free to open a separate Pull Request for it!

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/ashryaagr/Fairness.jl/pull/64#issuecomment-810871202, or unsubscribe https://github.com/notifications/unsubscribe-auth/AST3NVXUM5NYGUCTEFKDDVDTGLKSNANCNFSM42BMPS2A .