A Data-Driven Approach for Spatio-Temporal Crime Predictions in Smart Cities

摘要: The steadily increasing urbanization is causing significant economic and social transformations in urban areas it will be posing several challenges city management issues. In particular, given that the larger cities higher crime rates, spiking becoming one of most important problems large areas. To handle with increase crimes, new technologies are enabling police departments to access growing volumes crime-related data can analyzed understand patterns trends, finalized an efficient deployment officers over territory more effective prevention. This paper presents approach based on spatial analysis auto-regressive models automatically detect high-risk regions reliably forecast trends each region. final result algorithm a spatio-temporal forecasting model, composed set dense associated predictors, representing predictive model for number crimes happen its specific experimental evaluation, performed real-world collected big area Chicago, shows proposed achieves good accuracy temporal rolling time horizons.