FedStruct: Federated Decoupled Learning over Interconnected Graphs

arXiv:2402.19163v1 Announce Type: new
Abstract: We address the challenge of federated learning on graph-structured data distributed across multiple clients. Specifically, we focus on the prevalent scenario of interconnected subgraphs, where inter-connections between different clients play a critical role. We present a novel framework for this scenario, named FedStruct, that harnesses deep structural dependencies. To uphold privacy, unlike existing methods, FedStruct eliminates the necessity of sharing or generating sensitive node features or embeddings among clients. Instead, it leverages explicit global graph …

abstract arxiv challenge cs.it cs.lg data dependencies distributed federated learning focus framework graph graphs math.it multiple novel role structured data type