Crime Patterns in Chicago Using Agent-Based Modeling

Abstract

An agent-based model (ABM) is proposed to represent and analyze the self-organization of crime spread in the city of Chicago. Despite the overall national reduction in crime, Chicago’s crime is still double the national average. And in the past year, Chicago’s murder rate has been the worst since 1997. Therefore, methods of analyzing complex systems have become essential in predicting crime behavior and movement. Moreover, the interaction between criminals and their living environment, the correlation with the migration of people, and various circumstantial drivers are considered in building the model. The designed ABM will be validated by simulating various experimental scenarios, and comparing the results. From observing patterns across location, time, and crime type, behavior rules and individual adapting properties are determined by using the micro sequential rules. Ideally, the model will be valid, simple, and robust. In the end, a crime spread algorithm is proposed, which has the potential to be implemented to develop practical crime-prevention plans.

Lu He 何璐

Faculty of Supply Chain Management

My research interests include systematic resource optimization, multitask prediction, and predictive-driven mixed integer programming.