On the Inductive Biases of Demographic Parity-based Fair Learning Algorithms

arXiv:2402.18129v1 Announce Type: new
Abstract: Fair supervised learning algorithms assigning labels with little dependence on a sensitive attribute have attracted great attention in the machine learning community. While the demographic parity (DP) notion has been frequently used to measure a model's fairness in training fair classifiers, several studies in the literature suggest potential impacts of enforcing DP in fair learning algorithms. In this work, we analytically study the effect of standard DP-based regularization methods on the conditional distribution of the …

abstract algorithms arxiv attention biases classifiers community cs.ai cs.it cs.lg fair fairness inductive labels machine machine learning math.it notion studies supervised learning training type