Graph-Relational Federated Learning: Enhanced Personalization and Robustness

Wanyu Lin, Hao Lan, Hao He, Baochun Li

Primary Area: general machine learning (i.e., none of the above)

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.

Keywords: Federated Learning, Robustness

Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2024/AuthorGuide.

Abstract: Hypernetwork has recently emerged as a promising technique to generate personalized models in federated learning (FL). However, existing works tend to treat each client equally and independently --- each client contributes equally to learning the hypernetwork, and their representations are independent in the hypernetwork. Such an independent treatment ignores topological structures among different clients, which are usually reflected in the heterogeneity of client data distribution. In this work, we propose Panacea, a novel FL framework that can incorporate client relations as a graph to facilitate learning and personalization using graph hypernetwork. Empirically, we show that Panacea achieves state-of-the-art performance in terms of both accuracy and speed on multiple benchmarks. Further, Panacea improves robustness by leveraging the client relation graph. Specifically, it (1) generalizes better to novel clients outside of the training and (2) is more resilient to label-flipping attacks, which is also proved by our theoretical analysis.

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: 3052