This lecture, delivered by Francesco Cremonesi, provides a gentle introduction to Federated Learning and its growing role in collaborative health research.
The session introduces the fundamental concepts of Federated Learning, explaining how machine learning models can be trained across multiple institutions while keeping sensitive data securely at its source. Participants gain an overview of the benefits, challenges, and practical considerations of federated approaches, including privacy preservation, governance, interoperability, and distributed analytics.
By presenting key principles and real-world applications, the lecture demonstrates how Federated Learning can enable secure, privacy-preserving collaboration across institutions and countries. The topic aligns closely with HemaFAIR’s objectives of advancing FAIR data practices, supporting responsible data use, and enabling large-scale research in rare hematological diseases without compromising patient privacy.
The webinar slides are available at the HemaFAIR Zenodo Account: https://doi.org/10.5281/zenodo.19346790