[Re] Exacerbating Algorithmic Bias through Fairness Attacks

Nov 1, 2022·
Angelos Nalmpantis*
,
Apostolos Panagiotopoulos*
,
John Gkountouras*
,
Konstantinos Papakostas*
· 1 min read
Abstract
This study reproduces the paper ‘Exacerbating Algorithmic Bias through Fairness Attacks’. We confirmed most of the original claims and provided extended errata and explanations for the rest. Additionally, we proposed a dataset debiasing algorithm by extending the work and reversing the bias ‘direction’.
Type
Publication
ReScience Journal

This work was presented at NeurIPS 2022 Journal Showcase Track in New Orleans, US.

*Equal contribution

Algorithmic Bias Fairness Attacks Reproducibility