The paper “Enhancing Privacy via Hierarchical Federated Learning” was accepted at the International Workshop on Privacy Engineering (IWPE), to be co-located with the IEEE EuroS&P 2020 conference in September 7-11, 2020 in Genova, Italy.
In this paper, TK researchers discuss applying federated learning on a hierarchical architecture as a potential solution for several privacy-related issues. They introduce the opportunities for more flexible decentralized control over the training process and its impact on the participants' privacy. Furthermore, the authors investigate possibilities to enhance the efficiency and effectiveness of defense and verification methods.
- Aidmar Wainakh, Alejandro Sanchez Guinea, Tim Grube, Max Mühlhäuser. Enhancing Privacy via Hierarchical Federated Learning. In at the International Workshop on Privacy Engineering (IWPE) at IEEE EuroS&P 2020 [to appear]