Transferring Fairness using Multi-Task Learning with Limited Demographic Information

arXiv:2305.12671v2 Announce Type: replace
Abstract: Training supervised machine learning systems with a fairness loss can improve prediction fairness across different demographic groups. However, doing so requires demographic annotations for training data, without which we cannot produce debiased classifiers for most tasks. Drawing inspiration from transfer learning methods, we investigate whether we can utilize demographic data from a related task to improve the fairness of a target task. We adapt a single-task fairness loss to a multi-task setting to exploit demographic …

abstract annotations arxiv classifiers cs.cy cs.lg data fairness however information inspiration learning systems loss machine machine learning multi-task learning prediction supervised machine learning systems tasks training training data transfer transfer learning type