Matches in SemOpenAlex for { <https://semopenalex.org/work/W4383749603> ?p ?o ?g. }
Showing items 1 to 90 of
90
with 100 items per page.
- W4383749603 abstract "Federated Learning (FL) has emerged as a new paradigm for training machine learning models distributively without sacrificing data security and privacy. Learning models on edge devices such as mobile phones is one of the most common use cases for FL. However, Non-identical independent distributed (non-IID) data in edge devices easily leads to training failures. Especially, over-parameterized machine learning models can easily be over-fitted on such data, hence, resulting in inefficient federated learning and poor model performance. To overcome the over-fitting issue, we proposed an adaptive dynamic pruning approach for FL, which can dynamically slim the model by dropping out unimportant parameters, hence, preventing over-fittings. Since the machine learning model's parameters react differently for different training samples, adaptive dynamic pruning will evaluate the salience of the model's parameter according to the input training sample, and only retain the salient parameter's gradients when doing back-propagation. We performed comprehensive experiments to evaluate our approach. The results show that our approach by removing the redundant parameters in neural networks can significantly reduce the over-fitting issue and greatly improves the training efficiency. In particular, when training the ResNet-32 on CIFAR-10, our approach reduces the communication cost by 57%. We further demonstrate the inference acceleration capability of the proposed algorithm. Our approach reduces up to 50% FLOPs inference of DNNs on edge devices while maintaining the model's quality." @default.
- W4383749603 created "2023-07-11" @default.
- W4383749603 creator A5013277990 @default.
- W4383749603 creator A5070558865 @default.
- W4383749603 creator A5079359777 @default.
- W4383749603 creator A5085171738 @default.
- W4383749603 date "2023-05-01" @default.
- W4383749603 modified "2023-09-27" @default.
- W4383749603 title "Heterogeneous Federated Learning using Dynamic Model Pruning and Adaptive Gradient" @default.
- W4383749603 cites W2112796928 @default.
- W4383749603 cites W2194775991 @default.
- W4383749603 cites W2977090839 @default.
- W4383749603 cites W2979637109 @default.
- W4383749603 cites W2989289980 @default.
- W4383749603 cites W3015636663 @default.
- W4383749603 cites W3034408420 @default.
- W4383749603 cites W3035377608 @default.
- W4383749603 cites W3103802018 @default.
- W4383749603 cites W3105122387 @default.
- W4383749603 cites W3109212549 @default.
- W4383749603 cites W3129329365 @default.
- W4383749603 cites W3133504636 @default.
- W4383749603 cites W3159080474 @default.
- W4383749603 cites W3173563887 @default.
- W4383749603 cites W3182125009 @default.
- W4383749603 cites W3184838508 @default.
- W4383749603 cites W3211786132 @default.
- W4383749603 cites W4225796852 @default.
- W4383749603 cites W4285603559 @default.
- W4383749603 cites W4287332481 @default.
- W4383749603 cites W4294106961 @default.
- W4383749603 doi "https://doi.org/10.1109/ccgrid57682.2023.00038" @default.
- W4383749603 hasPublicationYear "2023" @default.
- W4383749603 type Work @default.
- W4383749603 citedByCount "0" @default.
- W4383749603 crossrefType "proceedings-article" @default.
- W4383749603 hasAuthorship W4383749603A5013277990 @default.
- W4383749603 hasAuthorship W4383749603A5070558865 @default.
- W4383749603 hasAuthorship W4383749603A5079359777 @default.
- W4383749603 hasAuthorship W4383749603A5085171738 @default.
- W4383749603 hasConcept C108010975 @default.
- W4383749603 hasConcept C108583219 @default.
- W4383749603 hasConcept C111919701 @default.
- W4383749603 hasConcept C11413529 @default.
- W4383749603 hasConcept C119857082 @default.
- W4383749603 hasConcept C138236772 @default.
- W4383749603 hasConcept C154945302 @default.
- W4383749603 hasConcept C162307627 @default.
- W4383749603 hasConcept C165464430 @default.
- W4383749603 hasConcept C2776214188 @default.
- W4383749603 hasConcept C2780186347 @default.
- W4383749603 hasConcept C38652104 @default.
- W4383749603 hasConcept C41008148 @default.
- W4383749603 hasConcept C50644808 @default.
- W4383749603 hasConcept C6557445 @default.
- W4383749603 hasConcept C79974875 @default.
- W4383749603 hasConcept C86803240 @default.
- W4383749603 hasConceptScore W4383749603C108010975 @default.
- W4383749603 hasConceptScore W4383749603C108583219 @default.
- W4383749603 hasConceptScore W4383749603C111919701 @default.
- W4383749603 hasConceptScore W4383749603C11413529 @default.
- W4383749603 hasConceptScore W4383749603C119857082 @default.
- W4383749603 hasConceptScore W4383749603C138236772 @default.
- W4383749603 hasConceptScore W4383749603C154945302 @default.
- W4383749603 hasConceptScore W4383749603C162307627 @default.
- W4383749603 hasConceptScore W4383749603C165464430 @default.
- W4383749603 hasConceptScore W4383749603C2776214188 @default.
- W4383749603 hasConceptScore W4383749603C2780186347 @default.
- W4383749603 hasConceptScore W4383749603C38652104 @default.
- W4383749603 hasConceptScore W4383749603C41008148 @default.
- W4383749603 hasConceptScore W4383749603C50644808 @default.
- W4383749603 hasConceptScore W4383749603C6557445 @default.
- W4383749603 hasConceptScore W4383749603C79974875 @default.
- W4383749603 hasConceptScore W4383749603C86803240 @default.
- W4383749603 hasLocation W43837496031 @default.
- W4383749603 hasOpenAccess W4383749603 @default.
- W4383749603 hasPrimaryLocation W43837496031 @default.
- W4383749603 hasRelatedWork W2963058055 @default.
- W4383749603 hasRelatedWork W3015440700 @default.
- W4383749603 hasRelatedWork W3098411449 @default.
- W4383749603 hasRelatedWork W3177201542 @default.
- W4383749603 hasRelatedWork W4223943233 @default.
- W4383749603 hasRelatedWork W4309045103 @default.
- W4383749603 hasRelatedWork W4312200629 @default.
- W4383749603 hasRelatedWork W4360585206 @default.
- W4383749603 hasRelatedWork W4364306694 @default.
- W4383749603 hasRelatedWork W4380086463 @default.
- W4383749603 isParatext "false" @default.
- W4383749603 isRetracted "false" @default.
- W4383749603 workType "article" @default.