| Abstract: | The perceived lack of racial equity in policing has recently received headlines. Policing is complex, involving officers, criminals, politicians, citizens, and stressful, emotional situations. Within the past decade, the predictive policing movement has sought to predict crime, but existing data are highly biased. The proposed project steps away from purely data-driven predictive learning and instead employs stochastic simulation models to overcome bias in representing the dynamic policing system. For this project, the team worked with agent-based modeling and 911 call data to explore how proactive and dynamic police patrol deployment can balance workloads, decrease stress, and reduce crime. |