English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

A Generative Adversarial Network (GAN) is a deep-learning generative model inthe field of Machine Learning (ML) that involves training two Neural Networks(NN) using a sizable data set. In certain fields, such as medicine, thetraining data may be hospital patient records that are stored across differenthospitals. The classic centralized approach would involve sending the data to acentralized server where the model would be trained. However, that wouldinvolve breaching the privacy and confidentiality of the patients and theirdata, which would be unacceptable. Therefore, Federated Learning (FL), an MLtechnique that trains ML models in a distributed setting without data everleaving the host device, would be a better alternative to the centralizedoption. In this ML technique, only parameters and certain metadata would becommunicated. In spite of that, there still exist attacks that can infer userdata using the parameters and metadata. A fully privacy-preserving solutioninvolves homomorphically encrypting (HE) the data communicated. This paper willfocus on the performance loss of training an FL-GAN with three different typesof Homomorphic Encryption: Partial Homomorphic Encryption (PHE), SomewhatHomomorphic Encryption (SHE), and Fully Homomorphic Encryption (FHE). We willalso test the performance loss of Multi-Party Computations (MPC), as it hashomomorphic properties. The performances will be compared to the performance oftraining an FL-GAN without encryption as well. Our experiments show that themore complex the encryption method is, the longer it takes, with the extra timetaken for HE is quite significant in comparison to the base case of FL.

Effect of Homomorphic Encryption on the Performance of Training Federated Learning Generative Adversarial Networks