{{ p.title }}
{{ p.cited }}{{ p.venue }}
Privacy, measured.
A focused view of the questions behind the publication list: how to preserve sensitive signals while keeping data useful for analysis, prediction, and systems that need to respond in real time.
Time-series privacy
Perturbation algorithms, privacy-oriented feature selection, and dynamic utility improvement for multivariate time-series classification.
Synthetic trajectories
APU-TrajGen and its GRU-based successor generate movement data while adapting the privacy-utility trade-off in real time.
Resilient systems
Tampering detection, dynamic attestation, and anomaly-based intrusion detection for systems where trust is part of the specification.
Generative AI in education
Designing and evaluating a GenAI-literacy intervention for IT and computer-science students — measuring the move from use to informed use (CSEDU 2026).