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 algorithms perform reasonably well on independent andidentically distributed (IID) data. They, on the other hand, suffer greatlyfrom heterogeneous environments, i.e., Non-IID data. Despite the fact that manyresearch projects have been done to address this issue, recent findingsindicate that they are still sub-optimal when compared to training on IID data.In this work, we carefully analyze the existing methods in heterogeneousenvironments. Interestingly, we find that regularizing the classifier's outputsis quite effective in preventing performance degradation on Non-IID data.Motivated by this, we propose Learning from Drift (LfD), a novel method foreffectively training the model in heterogeneous settings. Our schemeencapsulates two key components: drift estimation and drift regularization.Specifically, LfD first estimates how different the local model is from theglobal model (i.e., drift). The local model is then regularized such that itdoes not fall in the direction of the estimated drift. In the experiment, weevaluate each method through the lens of the five aspects of federatedlearning, i.e., Generalization, Heterogeneity, Scalability, Forgetting, andEfficiency. Comprehensive evaluation results clearly support the superiority ofLfD in federated learning with Non-IID data.

Learning From Drift: Federated Learning on Non-IID Data via Drift Regularization