Daniel Cullina

Daniel Cullina is an Assistant Professor in Electrical Engineering, specializing in theoretical computer science and machine learning. His research explores fundamental aspects of adversarial robustness, graph alignment, and information theory, with applications spanning cybersecurity and data science.

Research Focus:

• Adversarial machine learning: Robustness guarantees, attack/defense strategies for classifiers
• Graph algorithms: Alignment and recovery in random graph models like Erdős-Rényi
• Information theory: Fundamental limits of database matching and Gaussian alignment
• Coding theory: Deletion error correction and converse bounds

His publications (28+ with 575+ Scopus citations) demonstrate consistent contributions to understanding adversarial vulnerabilities in ML systems and combinatorial algorithms for graph/data matching. Recent work (2020-2023) focuses on theoretical characterization of optimal losses under attacks and database alignment frameworks.

With an h-index of 12, his research output shows sustained productivity since 2012, peaking in 2016 (8 publications) and maintaining 3-5 annual publications in recent years.