2019 (modified: 09 Nov 2022)ICML2019Readers: Everyone
Abstract:In a recent study, Friedler et al. pointed out that several fair classification algorithms are not stable with respect to variations in the training set – a crucial consideration in several applica...