/fl_demonstrator

EUCAIM demonstrators - Federated Learning Platforms

As part of early release of the EUCAIM Data Federation Framework with preliminary proof-of-concepts on data federation, we're organizing technical demonstrators for showcasing the functionalities of EUCAIM Federated Learning (FL) frameworks.

Data federation

Nodes part of the federation are not necessarily the final Data Nodes of EUCAIM Data Federation, as for demonstration purposes, data in use here correspond to public datasets provided by the simulated FL experiments designed in this demonstrator.

display of the network were the experiments will be tested

Nodes

Federated Learning Simulated Experiments

The demonstrator includes two different FL simulation scenarios, on based on Machine Learning (ML) techniques, the other on Deep Learning (DP) algorithms.

FL Experiment Type Data Details
Federated Classification using Homogenous Logistic Regression ML Exposome Data Go
Federated Classification using Convolutional Neural Network DP Chest X-Ray Images Data Go

Federated Learning Platforms

The following are the EUCAIM Federated Learning frameworks that enrolled to at least one of the Federated Learning Experiments simulated in this demonstrator:

Flowr

Fed-BioMed

Substra