Heterogeneous federated learning for imbalanced phishing email detection

Today's article comes from the journal of Cybersecurity. The authors are Yi et al., from Nankai University, in China. In this paper they propose a federated framework for phishing email detection. Each node in the system trains locally and sends only their model-features to a central server. Then, every node is able to pull down the updated global model.

DOI: 10.1186/s42400-025-00420-2