Swift-FedGNN: Federated Graph Learning with Low Communication and Sample Complexities
Keywords: federated learning, graph neural network, optimization
Abstract: Graph neural networks (GNNs) have achieved great success in a wide variety of graph-based learning applications.
To expedite training for large-scale graphs,
distributed GNN training has been proposed using sampling-based mini-batch training.
However, such a traditional distributed GNN training approach is not applicable to emerging GNN learning applications with geo-distributed input graphs, which require the data to be kept within the site where it is generated to protect privacy.
On the other hand, federated learning (FL) has been widely used to enable privacy-preserving training under data parallelism.
However, because of cross-client links in the aforementioned geo-distributed graph data, applying federated learning directly to GNNs incurs expensive cross-client neighbor sampling and communication costs due to the large graph size and the dependencies between nodes among different clients.
To overcome these challenges, we propose a new mini-batch and sampling-based federated GNN algorithmic framework called Swift-FedGNN that primarily performs efficient parallel local training and periodically conducts time-consuming cross-client training.
Specifically, in Swift-FedGNN, each client *primarily* trains a local GNN model using only its local graph data, and some randomly sampled clients *periodically* learn the local GNN models based on their local graph data and the dependent nodes across clients.
We theoretically establish the convergence performance of Swift-FedGNN and show that it enjoys a convergence rate of $\mathcal{O}\left( T^{-1/2} \right)$, matching the state-of-the-art (SOTA) rate of sampling-based GNN methods, despite operating in the challenging FL setting.
Extensive experiments on real-world datasets show that Swift-FedGNN significantly outperforms the SOTA federated GNN approaches with comparable accuracy in terms of efficiency.
Primary Area: optimization
Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.
Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2025/AuthorGuide.
Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.
No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.
Submission Number: 10747
Loading