A Meta-learning Framework for Tuning Parameters of Protection Mechanisms in Trustworthy Federated Learning

arXiv:2305.18400v3 Announce Type: replace
Abstract: Trustworthy Federated Learning (TFL) typically leverages protection mechanisms to guarantee privacy. However, protection mechanisms inevitably introduce utility loss or efficiency reduction while protecting data privacy. Therefore, protection mechanisms and their parameters should be carefully chosen to strike an optimal tradeoff between \textit{privacy leakage}, \textit{utility loss}, and \textit{efficiency reduction}. To this end, federated learning practitioners need tools to measure the three factors and optimize the tradeoff between them to choose the protection mechanism that is most …

abstract arxiv cs.ai cs.lg data data privacy efficiency federated learning framework loss meta meta-learning parameters privacy protection strike trustworthy type utility