Collaborative Research: NSF-CSIRO: Fair Sequential Collective Decision-Making

该项目旨在开发(人工智能)人工智能驱动的方法，以解决具有挑战性的社会问题，如应对干旱、传染病和有害环境的排放。该项目涉及这些领域的具体问题，例如：如何有效地分配水资源，以提高农业在干旱季节的抗旱能力？如何才能有效地确定何时何地建设氢气或电动汽车加油站，以鼓励公民采用这些技术来降低排放？如何才能每天有效地分发疫苗和其他医疗用品，以加强在大流行期间对传染病的反应？这些问题属于序贯集体决策中的一类经典而重要的问题。虽然以前已经研究了顺序决策和集体决策，但对同时是顺序决策和集体决策的决策问题了解很少，特别是对于资源分配等特定领域，以及当与责任和公平等目标相结合时。该项目的总体目标是为负责任和公平的人工智能支持的顺序集体决策建立理论和算法基础。它还寻求确保一系列决定满足多个目标，并根据公平标准在短期和长期奖励之间进行适当的权衡。拟议的研究将带来高效和公平的解决方案，解决我们在抗旱、净零和传染病抗灾方面的社会公益和受使用启发的应用。该项目将通过专注于三个相互关联的挑战来实现其目标，利用人工智能、经济学和运筹学的广泛技术。第一，在多个目标和公平标准的约束下，对单一时间段进行公平的多目标集体决策的挑战。第二，探索公平有序的多目标集体决策，解决眼前和长期效率与公平之间的权衡。最后，与利益相关者合作，了解公平多目标集体决策的战略方面，利益相关者提供促进决策过程的信息。这是美国和澳大利亚研究人员的联合项目，由美国国家科学基金会和澳大利亚联邦科学与工业研究组织(CSIRO)下的负责任和公平人工智能合作机会资助。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

This project aims to develop (artificial intelligence) AI-powered approaches to address challenging societal problems, such as dealing with droughts, infectious diseases, and environmentally harmful emissions. This project engages specific questions in these areas, such as: How can one effectively allocate water resources to increase agricultural drought resilience during drought seasons? How can one effectively determine when and where to construct hydrogen or electric vehicle refueling stations to encourage citizens to adopt these technologies to lower emissions? How can one effectively distribute vaccines and other medical supplies daily to enhance response to infectious diseases during pandemics? These problems belong to a class of classical and important problems in sequential collective decision-making. While sequential decision-making and collective decision-making have been studied previously, decision-making problems that are simultaneously sequential and collective are poorly understood, especially for specific domains such as resource allocation and when combined with goals such as responsibility and equitability. The overarching project goal is to establish theoretical and algorithmic foundations for responsible and equitable AI-powered sequential, collective decision-making. It also seeks to ensure that sequences of decisions satisfy multiple objectives and make appropriate trade-offs between short and long-term rewards subject to fairness criteria. The proposed research will lead to efficient and fair solutions to our social good and use-inspired applications in drought resilience, towards net zero, and infectious disease resilience. The project will achieve its goals by focusing on three interconnected challenges, leveraging a wide range of techniques from AI, economics, and operation research. First, the challenge of fair multi-objective collective decision-making for a single time period subject to multiple objectives and fairness criteria. Second, explore fair sequential multi-objective collective decision-making that addresses trade-offs between immediate and long-term efficiency and fairness. Finally, understand the strategic aspects of fair multi-objective collective decision-making in collaboration with stakeholders, who provide information that facilitates the process.This is a joint project between United States and Australian researchers funded by the Collaboration Opportunities in Responsible and Equitable AI under the U.S. NSF and the Australian Commonwealth Scientific and Industrial Research Organisation (CSIRO).This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.