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 new machine learning framework which trains ajoint model across a large amount of decentralized computing devices. Existingmethods, e.g., Federated Averaging (FedAvg), are able to provide anoptimization guarantee by synchronously training the joint model, but usuallysuffer from stragglers, i.e., IoT devices with low computing power orcommunication bandwidth, especially on heterogeneous optimization problems. Tomitigate the influence of stragglers, this paper presents a novel FL algorithm,namely Hybrid Federated Learning (HFL), to achieve a learning balance inefficiency and effectiveness. It consists of two major components: synchronouskernel and asynchronous updater. Unlike traditional synchronous FL methods, ourHFL introduces the asynchronous updater which actively pulls unsynchronized anddelayed local weights from stragglers. An adaptive approximation method,Adaptive Delayed-SGD (AD-SGD), is proposed to merge the delayed local updatesinto the joint model. The theoretical analysis of HFL shows that theconvergence rate of the proposed algorithm is $\mathcal{O}(\frac{1}{t+\tau})$for both convex and non-convex optimization problems.

Stragglers Are Not Disaster: A Hybrid Federated Learning Algorithm with Delayed Gradients