Matches in SemOpenAlex for { <https://semopenalex.org/work/W4319793899> ?p ?o ?g. }
Showing items 1 to 65 of
65
with 100 items per page.
- W4319793899 abstract "Background: Federated learning methods offer the possibility of training machine learning models on privacy-sensitive data sets, which cannot be easily shared. Multiple regulations pose strict requirements on the storage and usage of healthcare data, leading to data being in silos (i.e. locked-in at healthcare facilities). The application of federated algorithms on these datasets could accelerate disease diagnostic, drug development, as well as improve patient care. Methods: We present an extensive evaluation of the impact of different federation and differential privacy techniques when training models on the open-source MIMIC-III dataset. We analyze a set of parameters influencing a federated model performance, namely data distribution (homogeneous and heterogeneous), communication strategies (communication rounds vs. local training epochs), federation strategies (FedAvg vs. FedProx). Furthermore, we assess and compare two differential privacy (DP) techniques during model training: a stochastic gradient descent-based differential privacy algorithm (DP-SGD), and a sparse vector differential privacy technique (DP-SVT). Results: Our experiments show that extreme data distributions across sites (imbalance either in the number of patients or the positive label ratios between sites) lead to a deterioration of model performance when trained using the FedAvg strategy. This issue is resolved when using FedProx with the use of appropriate hyperparameter tuning. Furthermore, the results show that both differential privacy techniques can reach model performances similar to those of models trained without DP, however at the expense of a large quantifiable privacy leakage. Conclusions: We evaluate empirically the benefits of two federation strategies and propose optimal strategies for the choice of parameters when using differential privacy techniques." @default.
- W4319793899 created "2023-02-11" @default.
- W4319793899 creator A5025057284 @default.
- W4319793899 creator A5029839282 @default.
- W4319793899 creator A5042069919 @default.
- W4319793899 creator A5069950074 @default.
- W4319793899 creator A5090233283 @default.
- W4319793899 date "2023-02-08" @default.
- W4319793899 modified "2023-09-27" @default.
- W4319793899 title "Exploratory Analysis of Federated Learning Methods with Differential Privacy on MIMIC-III" @default.
- W4319793899 doi "https://doi.org/10.48550/arxiv.2302.04208" @default.
- W4319793899 hasPublicationYear "2023" @default.
- W4319793899 type Work @default.
- W4319793899 citedByCount "0" @default.
- W4319793899 crossrefType "posted-content" @default.
- W4319793899 hasAuthorship W4319793899A5025057284 @default.
- W4319793899 hasAuthorship W4319793899A5029839282 @default.
- W4319793899 hasAuthorship W4319793899A5042069919 @default.
- W4319793899 hasAuthorship W4319793899A5069950074 @default.
- W4319793899 hasAuthorship W4319793899A5090233283 @default.
- W4319793899 hasBestOaLocation W43197938991 @default.
- W4319793899 hasConcept C111919701 @default.
- W4319793899 hasConcept C119857082 @default.
- W4319793899 hasConcept C124101348 @default.
- W4319793899 hasConcept C154945302 @default.
- W4319793899 hasConcept C177264268 @default.
- W4319793899 hasConcept C187191949 @default.
- W4319793899 hasConcept C199360897 @default.
- W4319793899 hasConcept C206688291 @default.
- W4319793899 hasConcept C23130292 @default.
- W4319793899 hasConcept C2992525071 @default.
- W4319793899 hasConcept C41008148 @default.
- W4319793899 hasConcept C50644808 @default.
- W4319793899 hasConcept C51632099 @default.
- W4319793899 hasConcept C8642999 @default.
- W4319793899 hasConceptScore W4319793899C111919701 @default.
- W4319793899 hasConceptScore W4319793899C119857082 @default.
- W4319793899 hasConceptScore W4319793899C124101348 @default.
- W4319793899 hasConceptScore W4319793899C154945302 @default.
- W4319793899 hasConceptScore W4319793899C177264268 @default.
- W4319793899 hasConceptScore W4319793899C187191949 @default.
- W4319793899 hasConceptScore W4319793899C199360897 @default.
- W4319793899 hasConceptScore W4319793899C206688291 @default.
- W4319793899 hasConceptScore W4319793899C23130292 @default.
- W4319793899 hasConceptScore W4319793899C2992525071 @default.
- W4319793899 hasConceptScore W4319793899C41008148 @default.
- W4319793899 hasConceptScore W4319793899C50644808 @default.
- W4319793899 hasConceptScore W4319793899C51632099 @default.
- W4319793899 hasConceptScore W4319793899C8642999 @default.
- W4319793899 hasLocation W43197938991 @default.
- W4319793899 hasOpenAccess W4319793899 @default.
- W4319793899 hasPrimaryLocation W43197938991 @default.
- W4319793899 hasRelatedWork W4210794429 @default.
- W4319793899 hasRelatedWork W4223456145 @default.
- W4319793899 hasRelatedWork W4226115111 @default.
- W4319793899 hasRelatedWork W4280535922 @default.
- W4319793899 hasRelatedWork W4286908675 @default.
- W4319793899 hasRelatedWork W4295309597 @default.
- W4319793899 hasRelatedWork W4309113015 @default.
- W4319793899 hasRelatedWork W4309134251 @default.
- W4319793899 hasRelatedWork W4312393190 @default.
- W4319793899 hasRelatedWork W4312521712 @default.
- W4319793899 isParatext "false" @default.
- W4319793899 isRetracted "false" @default.
- W4319793899 workType "article" @default.