Modeling, Learning and Defending against Opportunistic Criminals in Urban Areas

Police patrols are used ubiquitously to deter crimes in urban areas. A distinctive feature of urban crimes is that criminals react opportunistically to patrol officers' assignments. Compared to strategic attackers (such as terrorists) with a well-laid out plan, opportunistic criminals are less strategic in planning attacks and more flexible in executing them. I proposed two approaches to generate effective patrol schedules against opportunistic criminals. The first approach is a new game-theoretic framework for addressing opportunistic crime, the Opportunistic Security Game(OSG). In OSG, I propose a novel model for opportunistic adversaries. The second approach is to learn the criminals' behavior model from real-world criminal activity data. To that end, I represent the criminal behavior and the interaction with the patrol officers as parameters of a Dynamic Bayesian Network (DBN), enabling application of standard algorithms such as EM to learn the parameters. Finally, I show that a sequence of modifications of the DBN representation in learning approach, which exploit the problem structure in model approach, result in better accuracy and increased speed. By combining modeling and learning approaches, I can generate patrol schedule which has significantly better performance.