Abstract
In this work, we consider the optimization formulation of personalized federated learning recently introduced by [19] which was shown to give an alternative explanation to the workings of local SGD methods. Our first contribution is establishing the first lower bounds for this formulation, for both the communication complexity and the local oracle complexity. Our second contribution is the design of several optimal methods matching these lower bounds in almost all regimes. These are the first provably optimal methods for personalized federated learning. Our optimal methods include an accelerated variant of FedProx, and an accelerated variance-reduced version of FedAvg/Local SGD. We demonstrate the practical superiority of our methods through extensive numerical experiments.
Original language | English (US) |
---|---|
State | Published - 2020 |
Event | 34th Conference on Neural Information Processing Systems, NeurIPS 2020 - Virtual, Online Duration: Dec 6 2020 → Dec 12 2020 |
Conference
Conference | 34th Conference on Neural Information Processing Systems, NeurIPS 2020 |
---|---|
City | Virtual, Online |
Period | 12/6/20 → 12/12/20 |
ASJC Scopus subject areas
- Computer Networks and Communications
- Information Systems
- Signal Processing