Profiling privacy preservation against gradient inversion attacks in tabular federated learning
International Journal of Information Security · 21.11.2026 · 25 (6)
DOI: 10.1007/s10207-026-01360-9 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal of Information Security, online erschienen am 21.11.2026. Verfasst von Ivo Österberg Nilsson und 3 weiteren Autor:innen. Die Arbeit nennt 38 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Jönköping University.
Abstract (Originalsprache)
Federated learning (FL) enables multiple data holders to train machine learning models collaboratively without centralizing raw data, making it useful in privacy sensitive domains such as healthcare and institutional data sharing. FL keeps data local to clients while communicating only model updates, such as gradients or model deltas. Nevertheless, these updates can expose private client data through gradient inversion attacks (GIAs). We study this risk for tabular FL under an honest-but-curious server threat model across FL protocols, client batch sizes, training stages, attacker assumptions, model architectures, and binary classification, multiclass classification, and regression tasks. We use MIMIC-IV and complementary benchmark datasets. Our evaluation distinguishes numerical and categorical recovery, baseline recoverability, feature level recovery, and exact match rate (EMR). We evaluate FedSGD gradients and FedAvg model deltas with an exposure aligned protocol, comparing attacked models after matched client data exposure rather than matched communication rounds. We compare multilayer perceptron (MLP), ResNet, and FT-Transformer models, and isolate architecture effects through an MLP grid over width, depth, activation, normalization, and dropout. The results show that small client batches and updates representing few distinct records are most vulnerable. Larger local batches and stronger aggregation reduce reconstruction but do not eliminate leakage. FT-Transformer is consistently harder to invert than one-hot baselines, while reconstructability also varies substantially within the MLP family. These findings identify model architecture and input representation as practical privacy variables in tabular FL. We also show that aggregate reconstruction accuracy can overstate complete record recovery in sparse data, making EMR and baseline comparisons essential.
Förderung
- Jönköping University
Zitieren
APA
Nilsson, I. Ö., Engvall, M. B., Valadi, V., & Lazebnik, T. (2026). Profiling privacy preservation against gradient inversion attacks in tabular federated learning. International Journal of Information Security, 25(6). https://doi.org/10.1007/s10207-026-01360-9
Vancouver
Nilsson IÖ, Engvall MB, Valadi V, Lazebnik T. Profiling privacy preservation against gradient inversion attacks in tabular federated learning. International Journal of Information Security. 2026;25(6). doi:10.1007/s10207-026-01360-9
BibTeX
@article{Nilsson2026,
author = {Nilsson, Ivo Österberg and Engvall, Maximilian Birr and Valadi, Viktor and Lazebnik, Teddy},
title = {Profiling privacy preservation against gradient inversion attacks in tabular federated learning},
journal = {International Journal of Information Security},
year = {2026},
volume = {25},
number = {6},
doi = {10.1007/s10207-026-01360-9}
}
EUVIMED