Name
Title | ||
|---|---|---|
Information-Theoretic Bounds on the Generalization Error and Privacy Leakage in Federated Learning |
Abstract | ||
|---|---|---|
Machine learning algorithms operating on mobile networks can be characterized into three different categories. First is the classical situation in which the end-user devices send their data to a central server where this data is used to train a model. Second is the distributed setting in which each device trains its own model and send its model parameters to a central server where these model parameters are aggregated to create one final model. Third is the federated learning setting in which, at any given time t, a certain number of active end users train with their own local data along with feedback provided by the central server and then send their newly estimated model parameters to the central server. The server, then, aggregates these new parameters, updates its own model, and feeds the updated parameters back to all the end users, continuing this process until it converges.The main objective of this work is to provide an information-theoretic framework for all of the aforementioned learning paradigms. Moreover, using the provided framework, we develop upper and lower bounds on the generalization error together with bounds on the privacy leakage in the classical, distributed and federated learning settings. |
Year | DOI | Venue |
|---|---|---|
2020 | 10.1109/SPAWC48557.2020.9154277 | 2020 IEEE 21st International Workshop on Signal Processing Advances in Wireless Communications (SPAWC) |
Keywords | DocType | ISSN |
aggregates these new parameters,estimated model parameters,active end users,federated learning setting,distributed setting,central server,end-user devices,information-theoretic bounds,distributed federated learning,classical federated learning,privacy leakage,generalization error,learning paradigms,information-theoretic framework,updated parameters | Conference | 1948-3244 |
ISBN | Citations | PageRank |
978-1-7281-5479-4 | 0 | 0.34 |
References | Authors | |
4 | 3 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Yagli Semih | 1 | 0 | 0.34 |
| Alex Dytso | 2 | 45 | 20.03 |
| H. V. Poor | 3 | 25411 | 1951.66 |