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 has become a popular method to learn from decentralizedheterogeneous data. Federated semi-supervised learning (FSSL) emerges to trainmodels from a small fraction of labeled data due to label scarcity ondecentralized clients. Existing FSSL methods assume independent and identicallydistributed (IID) labeled data across clients and consistent class distributionbetween labeled and unlabeled data within a client. This work studies a morepractical and challenging scenario of FSSL, where data distribution isdifferent not only across clients but also within a client between labeled andunlabeled data. To address this challenge, we propose a novel FSSL frameworkwith dual regulators, FedDure.} FedDure lifts the previous assumption with acoarse-grained regulator (C-reg) and a fine-grained regulator (F-reg): C-regregularizes the updating of the local model by tracking the learning effect onlabeled data distribution; F-reg learns an adaptive weighting scheme tailoredfor unlabeled instances in each client. We further formulate the client modeltraining as bi-level optimization that adaptively optimizes the model in theclient with two regulators. Theoretically, we show the convergence guarantee ofthe dual regulators. Empirically, we demonstrate that FedDure is superior tothe existing methods across a wide range of settings, notably by more than 11%on CIFAR-10 and CINIC-10 datasets.

Combating Data Imbalances in Federated Semi-supervised Learning with Dual Regulators