Federated Generative Privacy

We propose FedGP, a framework for privacy-preserving data release in the federated learning setting. We use generative adversarial networks, generator components of which are trained by FedAvg algorithm, to draw private artificial data samples and empirically assess the risk of information disclosure. Our experiments show that FedGP is able to generate labeled data of high quality to successfully train and validate supervised models. Finally, we demonstrate that our approach significantly reduces vulnerability of such models to model inversion attacks.


Published in:
Ieee Intelligent Systems, 35, 4, 50-57
Year:
Jul 01 2020
Publisher:
Los Alamitos, IEEE COMPUTER SOC
ISSN:
1541-1672
1941-1294
Laboratories:




 Record created 2020-09-16, last modified 2020-10-27


Rate this document:

Rate this document:
1
2
3
 
(Not yet reviewed)