2026
Cold-Start Hyperlocal PM2.5: A Leakage-Controlled Comparison of Interpolation, Multimodal Fusion, and Contrastive Representation Learning
2nd IEEE International Conference on Data Science and Geoinformatics (ICDSG)Accepted
PhD student, Computer Science
Data Mining and Security Lab, McGill University · Montreal, Canada
Research
I work on machine learning for measurement-poor settings: estimating what instruments do not observe, learning representations that hold up when a modality goes missing, and evaluating both under protocols that do not leak. Recent work spans environmental sensing, multimodal content safety, and AI-driven cybersecurity.
Publications
2nd IEEE International Conference on Data Science and Geoinformatics (ICDSG)Accepted
Companion Proceedings of the ACM Web Conference 2026 (WWW ’26)
IEEE Mediterranean and Middle-East Geoscience and Remote Sensing Symposium (M2GARSS)
17th International Conference on Cyber Conflict (CyCon), IEEE, pp. 189–208
MSc thesis, Queen’s University, Kingston, Canada
5th International Conference on Algorithms, Computing and Artificial Intelligence (ACAI), ACM, pp. 1–5