Federated Learning is a new way of machine learning (ML), with the potential to allow use of sensitive data while complying with legal requirements.
Through this project, AI Sweden and its partners are creating a setting that allows the build up of collective know-how on how to work with this technology in a production setting, while also investigating the legal and privacy requirements.
In doing so, several pilots from different domains are being implemented, along with a virtual testbed for experimenting, evaluating, and exploring Federated Learning (and its related technologies).
FOCUS AREAS
The main focus of the project will be on how Federated Learning can be an enabler for developing machine learning solutions on data that cannot be shared openly. This includes examining technical, privacy and legal aspects of Federated Learning that would enable work with this promising technology in both the private and public sector.
Establishing a testbed will support the development of best practices and guidelines, as well as to provide opportunities to executing smaller tests or proof of concepts.
OUTCOMES
The project is expected to generate technical demonstrators that are open to all AI Sweden partners (via the AI Sweden Data Factory and Edgelab). In addition, a general report and a legal “how-to-guide” will be available to everyone, that covers necessary aspects of bringing this technology into an organisation and learnings from working with Federated Learning.
The project was extended until the end of June 2021. Project results were shared on May 27 via a webinar.
WEBSITE
Federated Learning Testbed
CONTACT
To learn more about the Federated Learning Testbed, please contact Ola Svedin: ola.svedin@mobileheights.org
