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

Recent years have witnessed a huge demand for artificial intelligence andmachine learning applications in wireless edge networks to assist individualswith real-time services. Owing to the practical setting and privacypreservation of federated learning (FL), it is a suitable and appealingdistributed learning paradigm to deploy these applications at the network edge.Despite the many successful efforts made to apply FL to wireless edge networks,the adopted algorithms mostly follow the same spirit as FedAvg, thereby heavilysuffering from the practical challenges of label deficiency and deviceheterogeneity. These challenges not only decelerate the model training in FLbut also downgrade the application performance. In this paper, we focus on thealgorithm design and address these challenges by investigating the personalizedsemi-supervised FL problem and proposing an effective algorithm, namelyFedCPSL. In particular, the techniques of pseudo-labeling, andinterpolation-based model personalization are judiciously combined to provide anew problem formulation for personalized semi-supervised FL. The proposedFedCPSL algorithm adopts novel strategies, including adaptive client variancereduction, local momentum, and normalized global aggregation, to combat thechallenge of device heterogeneity and boost algorithm convergence. Moreover,the convergence property of FedCPSL is thoroughly analyzed and shows thatFedCPSL is resilient to both statistical and system heterogeneity, obtaining asublinear convergence rate. Experimental results on image classification tasksare also presented to demonstrate that the proposed approach outperforms itscounterparts in terms of both convergence speed and application performance.

Towards Fast Personalized Semi-Supervised Federated Learning in Edge Networks: Algorithm Design and Theoretical Guarantee