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 is a novel paradigm that involves learning from datasamples distributed across a large network of clients while the data remainslocal. It is, however, known that federated learning is prone to multiplesystem challenges including system heterogeneity where clients have differentcomputation and communication capabilities. Such heterogeneity in clients'computation speeds has a negative effect on the scalability of federatedlearning algorithms and causes significant slow-down in their runtime due tothe existence of stragglers. In this paper, we propose a novelstraggler-resilient federated learning method that incorporates statisticalcharacteristics of the clients' data to adaptively select the clients in orderto speed up the learning procedure. The key idea of our algorithm is to startthe training procedure with faster nodes and gradually involve the slower nodesin the model training once the statistical accuracy of the data correspondingto the current participating nodes is reached. The proposed approach reducesthe overall runtime required to achieve the statistical accuracy of data of allnodes, as the solution for each stage is close to the solution of thesubsequent stage with more samples and can be used as a warm-start. Ourtheoretical results characterize the speedup gain in comparison to standardfederated benchmarks for strongly convex objectives, and our numericalexperiments also demonstrate significant speedups in wall-clock time of ourstraggler-resilient method compared to federated learning benchmarks.

Straggler-Resilient Federated Learning: Leveraging the Interplay Between Statistical Accuracy and System Heterogeneity