This National Science Foundation project grant of $599,999 will fund research at Duke University from October 2022 through September 2026 towards developing secure methods for federated learning. Federated learning is an emerging machine learning technique that allows analysis of private data without centralized collection, but current methods lack security protections.
Under the Computer and Information Science and Engineering program (CFDA 47.070), the researchers will explore new security attacks on federated learning, develop provably secure federated learning methods to prevent poisoning and backdoor attacks, and detect malicious clients. The work aims to transfer results to industry and incorporate findings into courses to train students, including those from underrepresented groups, in developing secure federated learning systems. The award supports the NSF's mission to advance computing and information science through investigator-initiated research and education.
Generated 1/7/24, 10:46 AM