Penalizing Unfairness in Binary Classification

Yahav Bechavod, Katrina Ligett

We present a new approach for mitigating unfairness in learned classifiers. In particular, we focus on binary classification tasks over individuals from two populations, where, as…