Marco Di Gennaro
@marcodige.bsky.social
PhD Student @PoliMi, Security & Privacy of Machine and Federated Learning Researcher, 26, Italy.
This research was developed at the NECSTLab (Politecnico di Milano), in the context of the TRUSTroke project, in collaboration with my co-authors: Giovanni De Lucia, Stefano Longari, @raistolo.bsky.social , and @johnmc88.bsky.social.
June 16, 2025 at 10:02 AM
This research was developed at the NECSTLab (Politecnico di Milano), in the context of the TRUSTroke project, in collaboration with my co-authors: Giovanni De Lucia, Stefano Longari, @raistolo.bsky.social , and @johnmc88.bsky.social.
We found that federated tree-based models (like XGBoost) are vulnerable to dataset reconstruction attacks, and we demonstrated this vulnerability across major FL frameworks. Our work introduces a novel attack and provides guidelines for building stronger defenses.
June 16, 2025 at 9:55 AM
We found that federated tree-based models (like XGBoost) are vulnerable to dataset reconstruction attacks, and we demonstrated this vulnerability across major FL frameworks. Our work introduces a novel attack and provides guidelines for building stronger defenses.