Ethical AI | Artificial Intelligence Institute for Advances in Optimization

Ethics permeates the Institute both in its daily operations and its research activities. The Institute not only adopts a highly inter-disciplinary approach to ethics spanning foundational algorithmic development, optimization, and machine learning, but also focuses on incorporating ethics and values in the system design, rather than as an afterthought as is typically the case today. The Institute focuses on the broad transversal themes: debiasing ground truth, uncertainty quantification, impact on groups, and power and access.

Publications

Funded by the Institute

End-to-end Learning for Fair Ranking Systems. James Kotary, Ferdinando Fioretto, Pascal Van Hentenryck, and Ziwei Zhu. In the ACM Web Conference (WWW-2022), Lyon, France, April 2022.
Differential Privacy and Fairness in Decisions and Learning Tasks: A Survey, Ferdinando Fioretto, Cuong Tran, Pascal Van Hentenryck, Keyu Zhu. arXiv 2202.08187, February 2022.

Related work

Post-processing of Differentially Private Data: A Fairness Perspective, Keyu Zhu, Ferdinando Fioretto, and Pascal Van Hentenryck. In the Proceedings of the Joint International Conference on Artificial Intelligence (IJCAI-ECAI2022), Vienna, Austria.
Differential Privacy and Fairness in Decisions and Learning Tasks: A Survey. Ferdinando Fioretto, Cuong Tran, Pascal Van Hentenryck, and Keyu Zhu. In the Proceedings of the Joint International Conference on Artifcial Intelligence (IJCAI-ECAI2022), Vienna, Austria.
Fair and Reliable Reconnections for Temporary Disruptions in Electric Distribution Networks using Submodularity, with Cyrus Hettle, Daniel Molzahn, under submission 2021. Preprint available at arxiv: https://arxiv.org/abs/2104.07631.
Taming Wild Price Fluctuations: Monotone Stochastic Convex Optimization with Bandit Feedback, with Jad Salem and Vijay Kamble, under submission 2021. Preprint available at arxiv: https://arxiv.org/abs/2103.09287.
Balanced Districting on Grid Graphs with Provable Compactness and Contiguity, with Cyrus Hettle, Shixiang Zhu and Yao Xie. Extended Abstract in Foundations of Responsible Computing (FORC), 2021. Preprint available at arxiv: https://arxiv.org/abs/2102.05028.
Impact of Bias on School Admissions and Targeted Interventions, with Yuri Faenza, Xuan Zhang, under submission 2021. Preprint available at arxiv: https://arxiv.org/abs/2004.10846.
Impact of bias on school admissions and targeted interventions.Yuri Faenza, Swati Gupta, and Xuan Zhang. arXiv preprint arXiv:2004.10846, 2020.
Group-Fair Online Allocation in Continuous Time, with Semih Cayci, Atilla Eryilmaz, 34th Conference on Neural Information and Processing Systems, NeurIPS 2020. Preprint available at arxiv: https://arxiv.org/pdf/2006.06852.pdf.
Closing the gap: Mitigating bias in online resume-filtering. Jad Salem and Swati Gupta.The 16th Conference on Web and Internet Economics, 33, 2020.
Closing the Gap: Online Selections of Candidates with Biased Evaluations, with Jad Salem, Extended abstract in Web and Internet Economics (WINE) 2020 and Mechanism Design for Social Good Workshop MD4SG 2020. Preprint available at ssrn: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3444283.
Individual Fairness in Hindsight with Vijay Kamble, Journal of Machine Learning Research, JMLR 2021. Extended abstract appeared in 20th ACM Conference on Economics and Computation, EC 2019. Spotlight talk at NeurIPS Workshop on Ethical, Social and Governance Issues in AI 2018. Preprint.