Skip to content

implementation of several algorithms in the Federated Learning setting

License

Notifications You must be signed in to change notification settings

dadodimauro/MLDL-Federated-Learning

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

51 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

MLDL-Federated-Learning

In this project we analyze in depth different federated learning algorithms. We first start selecting a convolutional neural network and compute a baseline for reference; we then proceed to evaluate FedAVG, the first and most adopted solution in the federated scenario. Moreover, other federated algorithms are considered, in order to address different issues, for instance stagglers, client model complexity and classifier calibration. For each of the aforementioned algorithms we consider one of the most challenging problems in federated learning, which is dealing with non-IID data distribution, and compare the results.

[Report] https://github.com/dadodimauro/MLDL-Federated-Learning/blob/bc19048f97a5415051cf06527c5b77f98ad64600/Project6_Group1_Report.pdf

[File with compiled tables] https://github.com/dadodimauro/MLDL-Federated-Learning/blob/b775decafa6aa228f8d0734a8408af7aff296d0e/Project6_Group1_TablesResults.pdf

[Excel file with all results] https://github.com/dadodimauro/MLDL-Federated-Learning/blob/b775decafa6aa228f8d0734a8408af7aff296d0e/RESULT_SUMMARY.xlsx

[Report Presentation] https://github.com/dadodimauro/MLDL-Federated-Learning/blob/021673edf0c42c97820c31ec9fda8d79452af9c2/Project6_Group1_ReportPresentation.pdf

About

implementation of several algorithms in the Federated Learning setting

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published