SYNTHEMA has partnered with Flower, the leading open-source federated learning framework, to enhance secure and privacy-preserving machine learning in healthcare. This collaboration enhances SYNTHEMA’s technological capabilities while contributing valuable improvements to the federated learning community.
Enhancing Federated Learning with Flower
Federated learning enables artificial intelligence models to be trained across multiple locations without sharing raw data, improving both security and privacy. By integrating Flower’s framework, SYNTHEMA ensures that sensitive healthcare data remains protected while facilitating collaborative AI model training.
Flower’s framework stands out due to its stable API and continuous evolution, making it a reliable choice for implementing federated learning at scale. As early adopters, SYNTHEMA benefits from Flower’s latest advancements, particularly Flower Superlink—a middleware that separates the central server from edge nodes, allowing seamless execution of long-term federated learning tasks.
Tailoring Flower’s Technology to SYNTHEMA’s Needs
SYNTHEMA has customised Flower’s source code to align with its specific requirements, including a custom command-line interface for launching components. Additionally, Flower has provided technical guidance to ensure smooth integration and optimal performance.
To enhance efficiency, SYNTHEMA has developed a manager-worker architecture that enables sequential execution of tasks across multiple clusters. This approach streamlines federated learning workflows, increasing reliability and adaptability.
A Partnership Driving Mutual Innovation
The collaboration between SYNTHEMA and Flower establishes a structured framework for technical exchange. Flower supports SYNTHEMA with development insights, ensuring robust implementation of federated learning in healthcare applications. In return, SYNTHEMA provides valuable feedback to improve Flower’s framework, particularly in areas such as differential privacy and secure multi-party computation protocols.
By combining expertise, both organisations advance federated learning solutions that prioritise privacy, security, and efficiency. This partnership not only strengthens SYNTHEMA’s research and development efforts but also contributes to the broader federated learning ecosystem, benefiting researchers and developers worldwide.