Doctoral Dissertation Research: Defensive Resource Allocation Versus a Realistic Attacker in a Multi-Layer Security Game

Understanding how adversaries make decisions is essential for enhancing security and protecting public spaces. This project develops new tools to help security professionals allocate resources more effectively, minimizing vulnerabilities at soft targets such as schools, public venues, and places of worship. By combining insights from human-subject experiments and surveys with advanced mathematical modeling, this research provides practical strategies for making better defensive decisions.

Unlike traditional, theoretical approaches, which assume adversaries act purely rationally, this project incorporates data from human decision-making experiments and surveys to create more realistic and actionable models. The outcomes of this research extend beyond improving security as they also offer valuable insights into adversarial behavior that can inform fields like marketing and risk management.

Additionally, the project supports education by involving students in cutting-edge research and developing tools that make complex mathematical models accessible to non-experts. By advancing our understanding of decision-making and optimizing resource allocation, this project promotes national safety and prosperity.

This project develops a novel framework for studying adversarial decision-making by combining game-theoretic modeling with data from human-subject experiments and surveys. The primary objective is to model an attacker’s best response function based on actual decisions made in attacker-defender scenarios, rather than assuming purely rational and theoretical strategies.

The research investigates key variables—reward, cost, and probability of success—and their influence on adversarial behavior. Using this data, the project derives closed-form Nash equilibria to optimize defensive resource allocation strategies across multi-layered security systems. The study incorporates mixed-effects regression and numerical optimization techniques to analyze decision-making processes and solve for equilibrium strategies.

Additionally, by utilizing the research methodology of designing and conducting behavioral experiments that simulate attacker-defender games, collecting empirical data, and using advanced game-theoretic modeling, the project validates and refines prescriptive solutions displayed in Graphical User Interface toolkits for stakeholders. The results contribute to actionable resource allocation strategies while offering a scalable framework for analyzing adversarial behavior in diverse applications, including security and risk management.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.