POSEIDON: Privacy-Preserving Federated Neural Network Learning
POSEIDON: Privacy-Preserving Federated Neural Network Learning
About this item
Full title
Author / Creator
Publisher
Ithaca: Cornell University Library, arXiv.org
Journal title
Language
English
Formats
Publication information
Publisher
Ithaca: Cornell University Library, arXiv.org
Subjects
More information
Scope and Contents
Contents
In this paper, we address the problem of privacy-preserving training and evaluation of neural networks in an \(N\)-party, federated learning setting. We propose a novel system, POSEIDON, the first of its kind in the regime of privacy-preserving neural network training. It employs multiparty lattice-based cryptography to preserve the confidentiality...
Alternative Titles
Full title
POSEIDON: Privacy-Preserving Federated Neural Network Learning
Authors, Artists and Contributors
Identifiers
Primary Identifiers
Record Identifier
TN_cdi_proquest_journals_2439505368
Permalink
https://devfeature-collection.sl.nsw.gov.au/record/TN_cdi_proquest_journals_2439505368
Other Identifiers
E-ISSN
2331-8422
