As a Ph.D. student in electrical engineering at Virginia Tech, Zain ul Abdeen is developing robust machine learning algorithms for resilient and safety-critical systems, with a particular focus on power system applications. He is advised by Ming Jin.

Abdeen holds both bachelor’s and master’s degrees in mathematics from National University of Sciences and Technology (NUST) in Pakistan, where he graduated with distinction and received a gold medal for his academic performance. He served as a lecturer at his alma mater prior to entering Virginia Tech to pursue his doctorate. 

“Having a strong foundation in math led me to explore its real-world applications, eventually motivating me to transition into machine learning and power systems,” he said.

His recent work at the Sanghani Center addresses the security vulnerabilities of distributed energy resource management systems (DERMS).

Earlier this year, Abdeen presented “Defense Against Joint Poisoning and Evasion Attacks: A Case Study on DERMS,” at the 39th Annual AAAI Conference on Artificial Intelligence. This paper proposes a novel intrusion detection framework designed to defend against simultaneous poisoning (training-time) and evasion (deployment-time) attacks, formulated as a bilevel optimization problem.

"One of the things I value most about being a student at the Sanghani Center is the collaborative and interdisciplinary research environment. It brings together students and faculty from computer science, electrical engineering, and other fields, which encourages both technical depth and broad impact,” Abdeen said. “I particularly appreciate the center’s focus on real-world challenges—like robustness, fairness, and resilience in AI systems—which aligns closely with my own work. Access to seminars, mentoring, and cutting-edge computational resources also makes it an ideal place to develop as a researcher."

Zain has interned at the National Renewable Energy Laboratory (NREL), where he contributed to several projects at the intersection of AI and energy systems, including developing a reinforcement learning policy for critical load restoration and exploring transformer-based in-context learning approaches; designing a transformer-biodirectional long short-term memory (BiLSTM) hybrid model for intrusion detection in smart grids; and analyzing the impact of adversarial attacks on droop-based deep reinforcement learning controllers. 

These projects resulted in multiple publications, including an accepted paper, "Enhancing Distribution System Resilience: A First-Order Meta-RL algorithm for Critical Load Restoration," at the 2024 IEEE International Conference on Communications, Control, and Computing Technologies for Smart Grids (SmartGridComm).

Expected to graduate in Spring 2026, Abdeen plans to continue working on developing intelligent and trustworthy algorithms that can adapt to real-world uncertainties and improve the resilience of critical infrastructure.