6 papers · 1 filter
Privacy Risks in Time Series Forecasting: User- and Record-Level Membership Inference
Nicolas Johansson, Tobias Olsson, Daniel Nilsson +2
Membership inference attacks (MIAs) aim to determine whether specific data were used to train a model. While extensively studied on classification models, their impact on time seri…
Subgraph Federated Learning via Spectral Methods
Javad Aliakbari, Johan Ãstman, Ashkan Panahi +1
We consider the problem of federated learning (FL) with graph-structured data distributed across multiple clients. In particular, we address the prevalent scenario of interconnecte…
Practical Bayes-Optimal Membership Inference Attacks
Marcus Lassila, Johan Ãstman, Khac-Hoang Ngo +1
We develop practical and theoretically grounded membership inference attacks (MIAs) against both independent and identically distributed (i.i.d.) data and graph-structured data. Bu…
Decoupled Subgraph Federated Learning
Javad Aliakbari, Johan Ãstman, Alexandre Graell i Amat
We address the challenge of federated learning on graph-structured data distributed across multiple clients. Specifically, we focus on the prevalent scenario of interconnected subg…
FedGT: Identification of Malicious Clients in Federated Learning with Secure Aggregation
Marvin Xhemrishi, Johan Ãstman, Antonia Wachter-Zeh +1
We propose FedGT, a novel framework for identifying malicious clients in federated learning with secure aggregation. Inspired by group testing, the framework leverages overlapping…
Poisoning Attacks on Federated Learning for Autonomous Driving
Sonakshi Garg, Hugo Jönsson, Gustav Kalander +4
Federated Learning (FL) is a decentralized learning paradigm, enabling parties to collaboratively train models while keeping their data confidential. Within autonomous driving, it…