Paul Agbaje
I have completed my Computer Science Ph.D. at the University of Texas at Arlington as part of the Cyber-Physical Systems Security Lab led by Dr. Habeeb Olufowobi. I develop tools and frameworks to uncover vulnerabilities in machine learning systems, with a primary focus on the security, privacy, and efficiency of models, particularly graph-based approaches. I extend this work to safety-critical domains such as vehicular systems, where I study both the attack surfaces of in-vehicle networks and strategies for resource-efficient edge AI. My broader goal is to advance resilient, secure, and adaptive intelligent infrastructures. Beyond cyber-physical systems, I also explore applications of AI in healthcare and materials science, including explainability methods for health AI and techniques to extract microstructural fingerprints from complex data.
Latest News
Our abstract “Data-driven inference of twin network dynamics: nucleation, co-nucleation, transmission, and incidental contacts” got accepted for presentation at the AI/ML/Data Informatics for Materials Discovery symposium. View abstract scope.
I successfully defended my dissertation.
The Computer Science and Engineering department at UT Arlington featured me in a social media post. View the post.
Our paper “Revisiting Asymmetries in Black-box Link Stealing against Graph Neural Networks” got accepted at ICML’26.
I received the Outstanding Doctoral Dissertation Award presented at the College of Engineering Academic Excellence Ceremony at UTA.
I received the Graduate School Summer 2026 Dissertation Fellowship at UTA.
Our paper “A Collaborative Distillation Framework for Graph Neural Networks” got accepted at ESANN’26.
Our paper “Unveiling Graph Copycats: Inference Attacks with Student Models” got accepted at PETS’26.