My work is in privacy-preserving data mining and machine learning, more precisely, measuring how far time-series and trajectory data can be perturbed before it loses its usefulness, and showing where that boundary lies. Lecturer in the Department of Electrical Engineering and Information Technology, University of Medicine, Pharmacy, Sciences and Technology of Târgu Mureș, Romania.
The doctoral work and the thread that runs through most of the papers: evaluating perturbation algorithms on multivariate time-series data, privacy-oriented feature selection, and a dynamic framework that improves utility without conceding the privacy budget.
APU-TrajGen and its GRU-based successor: generating real-time synthetic movement data that keeps the shape of the original trace and adapts the privacy–utility trade-off as it runs. Related work covers crowdsourced WiFi RSS data under differential privacy.
The same measurement instinct applied to teaching: designing a GenAI-literacy intervention for IT and computer-science students and evaluating what it takes to move them from use to informed use. First results written up for CSEDU 2026.
A guide to using generative AI in your coursework, the code behind the papers, a recommended reading list, and the doctoral thesis.
Open resources“In the end, it’s all a question of balance.”
“Life is like riding a bicycle. To keep your balance, you must keep moving.”
Open to collaboration on privacy-preserving analytics, and to students interested in the privacy–utility trade-off.