Forecasting Crimes using Autoregressive Models

As a result of steadily increasing urbanization, by 2030 more than sixty percent of the global population will live in cities. This phenomenon is stimulating significant economic and social transformations, both positive (such as, increased opportunities offered in urban areas) and negative (such as, increased crime and pressures on city budgets). Nevertheless, new technologies are enabling police departments to access growing volumes of crime-related data that can be analyzed to understand patterns and trends. Such knowledge is useful to anticipate criminal activity and thus to optimize public safety resource allocation (officers, patrol routes, etc.) through mathematical techniques to predict crimes. This paper presents an approach, based on auto-regressive models, for reliably forecasting crime trends in urban areas. In particular, the main goal of the work is to design a predictive model to forecast the number of crimes that will happen in rolling time horizons. As a case study, we present the analysis performed on an area of Chicago, using a variety of open data sources available for exploration and examination through the University of Chicagos Plenario platform. Experimental evaluation shows that the proposed methodology predicts the number of crimes with an accuracy of 84% on one-year-ahead forecasts and of 80% on two-year-ahead forecasts.