@conference {1499311, title = {Exploring Algorithmic Fairness in Robust Graph Covering Problems}, year = {2019}, address = {Advances in Neural and Information Processing Systems (NeurIPS-19), 2019}, abstract = {Fueled by algorithmic advances, AI algorithms are increasingly being deployed insettings subject to unanticipated challenges with complex social effects. Motivatedby real-world deployment of AI driven, social-network based suicide preventionand landslide risk management interventions, this paper focuses on robust graphcovering problems subject to group fairness constraints. We show that, in theabsence of fairness constraints, state-of-the-art algorithms for the robust graph covering problem result in biased node coverage: they tend to discriminate individuals(nodes) based on membership in traditionally marginalized groups. To mitigatethis issue, we propose a novel formulation of the robust graph covering problemwith group fairness constraints and a tractable approximation scheme applicable toreal-world instances. We provide a formal analysis of the price of group fairness(PoF) for this problem, where we show that uncertainty can lead to greater PoF. Wedemonstrate the effectiveness of our approach on several real-world social networks.Our method yields competitive node coverage while significantly improving groupfairness relative to state-of-the-art methods.}, author = {Rahmattalabi, Aida and Vayanos, Phebe and Fulginiti, Anthony and Eric Rice and Bryan Wilder and Amulya Yadav and Tambe, Milind} }