Scheduling limited resources to secure potential attack targets is a global challenge faced by many agencies around the world, problem settings ranging from protection of physical infrastructure to cyber-security. The scheduling problem is made more difficult due to the presence of intelligent adversaries capable of conducting surveillance, as well as inherent uncertainties and complexities of the real world. My research uses game theory to compute optimal strategies to schedule defender resources.
The specific focus of my work has been to develop game-theoretic algorithms that can compute solutions for large-scale real-world problems. For example, let us consider the problem faced by the Federal Air Marshals Service (FAMS). The FAMS is a law enforcement agency that schedules undercover officers aboard commercial flights. Given that US carriers fly over 30,000 flights daily, the FAMS are presented with a massive scheduling challenge. I have developed new models and algorithms that compute optimal strategies for such large real-world domains.
These new models have not only advanced the state of the art, but have also been transitioned into real-world applications. For instance, I have developed algorithms for and subsequently led the building of the IRIS system that is used by the FAMS to schedule air marshals aboard international commercial flights since October 2009. I will discuss these algorithms and deployed applications in this talk.