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

In the federated learning scenario, geographically distributed clientscollaboratively train a global model. Data heterogeneity among clientssignificantly results in inconsistent model updates, which evidently slow downmodel convergence. To alleviate this issue, many methods employ regularizationterms to narrow the discrepancy between client-side local models and theserver-side global model. However, these methods impose limitations on theability to explore superior local models and ignore the valuable information inhistorical models. Besides, although the up-to-date representation methodsimultaneously concerns the global and historical local models, it suffers fromunbearable computation cost. To accelerate convergence with low resourceconsumption, we innovatively propose a model regularization method namedFedTrip, which is designed to restrict global-local divergence and decreasecurrent-historical correlation for alleviating the negative effects derivedfrom data heterogeneity. FedTrip helps the current local model to be close tothe global model while keeping away from historical local models, whichcontributes to guaranteeing the consistency of local updates among clients andefficiently exploring superior local models with negligible additionalcomputation cost on attaching operations. Empirically, we demonstrate thesuperiority of FedTrip via extensive evaluations. To achieve the targetaccuracy, FedTrip outperforms the state-of-the-art baselines in terms ofsignificantly reducing the total overhead of client-server communication andlocal computation.

FedTrip: A Resource-Efficient Federated Learning Method with Triplet Regularization