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

Recently, federated learning (FL) has emerged as a popular technique for edgeAI to mine valuable knowledge in edge computing (EC) systems. To mitigate thecomputing/communication burden on resource-constrained workers and protectmodel privacy, split federated learning (SFL) has been released by integratingboth data and model parallelism. Despite resource limitations, SFL still facestwo other critical challenges in EC, i.e., statistical heterogeneity and systemheterogeneity. To address these challenges, we propose a novel SFL framework,termed MergeSFL, by incorporating feature merging and batch size regulation inSFL. Concretely, feature merging aims to merge the features from workers into amixed feature sequence, which is approximately equivalent to the featuresderived from IID data and is employed to promote model accuracy. While batchsize regulation aims to assign diverse and suitable batch sizes forheterogeneous workers to improve training efficiency. Moreover, MergeSFLexplores to jointly optimize these two strategies upon their coupledrelationship to better enhance the performance of SFL. Extensive experimentsare conducted on a physical platform with 80 NVIDIA Jetson edge devices, andthe experimental results show that MergeSFL can improve the final modelaccuracy by 5.82% to 26.22%, with a speedup by about 1.74x to 4.14x, comparedto the baselines.

MergeSFL: Split Federated Learning with Feature Merging and Batch Size Regulation