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

Federated learning (FL) is a popular framework for training an AI model usingdistributed mobile data in a wireless network. It features data parallelism bydistributing the learning task to multiple edge devices while attempting topreserve their local-data privacy. One main challenge confronting practical FLis that resource constrained devices struggle with the computation intensivetask of updating of a deep-neural network model. To tackle the challenge, inthis paper, a federated dropout (FedDrop) scheme is proposed building on theclassic dropout scheme for random model pruning. Specifically, in eachiteration of the FL algorithm, several subnets are independently generated fromthe global model at the server using dropout but with heterogeneous dropoutrates (i.e., parameter-pruning probabilities),each of which is adapted to thestate of an assigned channel. The subnets are downloaded to associated devicesfor updating. Thereby, FedDrop reduces both the communication overhead anddevices' computation loads compared with the conventional FL whileoutperforming the latter in the case of overfitting and also the FL scheme withuniform dropout (i.e., identical subnets).

Federated Dropout -- A Simple Approach for Enabling Federated Learning on Resource Constrained Devices