Second-Order Guarantees In Federated Learning
Federated learning is a useful framework for centralized learning from distributed data under practical considerations of heterogeneity, asynchrony, and privacy. Federated architectures are frequently deployed in deep learning settings, which generally give rise to non-convex optimization problems. Nevertheless, most existing analysis are either limited to convex loss functions, or only establish first-order stationarity, despite the fact that saddle-points, which are first-order stationary, are known to pose bottlenecks in deep learning. We draw on recent results on the second-order optimality of stochastic gradient algorithms in centralized and decentralized settings, and establish second-order guarantees for a class of federated learning algorithms.
WOS:000681731800177
2020-01-01
978-0-7381-3126-9
New York
Conference Record of the Asilomar Conference on Signals Systems and Computers
915
922
REVIEWED
Event name | Event place | Event date |
ELECTR NETWORK | Nov 01-05, 2020 | |