Cybersecurity ML
Cybersecurity Federated Learning
Privacy-preserving collaborative learning for distributed anomaly detection and cyber defense.
3lab members, current and past
About the project
What this project is about
Developing and applying federated learning techniques to enhance cybersecurity defenses without sharing sensitive raw data.
This project explores how federated learning can be adapted to cybersecurity settings where organizations need collective intelligence without centralizing sensitive logs, telemetry, or device-level data. It ran as Master's research at the Department of Computing, Jönköping University, where Maximilian Engwall and Ivo Osterberg Nillsson focused on the security of federated learning. The work looked at privacy-preserving anomaly detection, distributed threat modeling, communication-efficient training, and robustness to non-IID or even adversarial participants.
Approach and outcomes
The methods the project relies on and what it has produced so far.
Methods
- Federated learning
- Distributed anomaly detection
- Privacy-preserving machine learning
- Robust aggregation
- Communication-efficient training
- Security evaluation under non-iid data
Key outcomes
- Master's theses by Maximilian Engwall and Ivo Osterberg Nillsson on the security of federated learning (Jönköping University, 2026)
- Threat models, deployment assumptions and privacy constraints for federated cyber-defense experiments
- A comparison plan for centralized, local and federated training under fragmented data and attack assumptions
The team
The lab members who work or worked on this project, with their part in it, and the researchers we work with. See the whole lab on the team page.
Current members 1
Past members 2
Ivo Osterberg Nillsson
MSc (alumni) · Dec 2025 – Jun 2026
Security of federated learning
Maximilian Engwall
MSc (alumni) · Dec 2025 – Jun 2026
Security of federated learning
Interested in this project?
We welcome collaborations, data partnerships and students who want to work on cybersecurity ml problems with us.