Sign in

Multi-Differential Fairness Auditor for Black Box Classifiers

By Xavier Gitiaux and Huzefa Rangwala
Machine learning algorithms are increasingly involved in sensitive decision-making process with adversarial implications on individuals. This paper presents mdfa, an approach that identifies the characteristics of the victims of a classifier's discrimination. We measure discrimination as a violation of multi-differential fairness. Multi-differential fairness is a guarantee that a black box... Show more
March 18, 2019
=
0
Loading PDF…
Loading full text...
Similar articles
Loading recommendations...
=
0
x1
Multi-Differential Fairness Auditor for Black Box Classifiers
Click on play to start listening