The goal of TRUMPET is to research and develop new methods for improving privacy in federated learning and to provide a scalable federated AI services platform for analyzing European cross-border datasets. The privacy guarantees of the platform will be validated for the scenario of cancer data from different European hospitals. A secondary objective is to research, develop, and promote with EU data protection authorities a new metric and tool for certifying GDPR compliance of federated learning implementations.
The interdisciplinary consortium consists of 9 partners from 5 countries (Spain, France, Belgium, Italy, and Israel), including various entities: 3 research organizations, 1 university, 3 SMEs, and 2 clinical partners, all with extensive experience and knowledge to ensure the proper performance of activities and the achievement of results.