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

In a data-centric era, concerns regarding privacy and ethical data handlinggrow as machine learning relies more on personal information. This empiricalstudy investigates the privacy, generalization, and stability of deep learningmodels in the presence of additive noise in federated learning frameworks. Ourmain objective is to provide strategies to measure the generalization,stability, and privacy-preserving capabilities of these models and furtherimprove them. To this end, five noise infusion mechanisms at varying noiselevels within centralized and federated learning settings are explored. Asmodel complexity is a key component of the generalization and stability of deeplearning models during training and evaluation, a comparative analysis of threeConvolutional Neural Network (CNN) architectures is provided. The paperintroduces Signal-to-Noise Ratio (SNR) as a quantitative measure of thetrade-off between privacy and training accuracy of noise-infused models, aimingto find the noise level that yields optimal privacy and accuracy. Moreover, thePrice of Stability and Price of Anarchy are defined in the context ofprivacy-preserving deep learning, contributing to the systematic investigationof the noise infusion strategies to enhance privacy without compromisingperformance. Our research sheds light on the delicate balance between thesecritical factors, fostering a deeper understanding of the implications ofnoise-based regularization in machine learning. By leveraging noise as a toolfor regularization and privacy enhancement, we aim to contribute to thedevelopment of robust, privacy-aware algorithms, ensuring that AI-drivensolutions prioritize both utility and privacy.

The Paradox of Noise: An Empirical Study of Noise-Infusion Mechanisms to Improve Generalization, Stability, and Privacy in Federated Learning