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 has allowed the training of statistical models over remotedevices without the transfer of raw client data. In practice, training inheterogeneous and large networks introduce novel challenges in various aspectslike network load, quality of client data, security and privacy. Recent worksin FL have worked on improving communication efficiency and addressing unevenclient data distribution independently, but none have provided a unifiedsolution for both challenges. We introduce a new family of Federated Learningalgorithms called CatFedAvg which not only improves the communicationefficiency but improves the quality of learning using a category coveragemaximization strategy.  We use the FedAvg framework and introduce a simple and efficient step everyepoch to collect meta-data about the client's training data structure which thecentral server uses to request a subset of weight updates. We explore twodistinct variations which allow us to further explore the tradeoffs betweencommunication efficiency and model accuracy. Our experiments based on a visionclassification task have shown that an increase of 10% absolute points inaccuracy using the MNIST dataset with 70% absolute points lower networktransfer over FedAvg. We also run similar experiments with Fashion MNIST,KMNIST-10, KMNIST-49 and EMNIST-47. Further, under extreme data imbalanceexperiments for both globally and individual clients, we see the modelperforming better than FedAvg. The ablation study further explores itsbehaviour under varying data and client parameter conditions showcasing therobustness of the proposed approach.

CatFedAvg: Optimising Communication-efficiency and Classification Accuracy in Federated Learning