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 collaborative learning, clients keep their data private and communicateonly the computed gradients of the deep neural network being trained on theirlocal data. Several recent attacks show that one can still extract privateinformation from the shared network's gradients compromising clients' privacy.In this paper, to quantify the private information leakage from gradients weadopt usable information theory. We focus on two types of private information:original information in data reconstruction attacks and latent information inattribute inference attacks. Furthermore, a sensitivity analysis over thegradients is performed to explore the underlying cause of information leakageand validate the results of the proposed framework. Finally, we conductnumerical evaluations on six benchmark datasets and four well-known deepmodels. We measure the impact of training hyperparameters, e.g., batches andepochs, as well as potential defense mechanisms, e.g., dropout and differentialprivacy. Our proposed framework enables clients to localize and quantify theprivate information leakage in a layer-wise manner, and enables a betterunderstanding of the sources of information leakage in collaborative learning,which can be used by future studies to benchmark new attacks and defensemechanisms.

Quantifying and Localizing Usable Information Leakage from Neural Network Gradients