Inteligência em Segurança Pública Municipal: modelos para previsão de crimes e planejamento de zeladoria urbana

The uncontrolled and irregular growth of several urban centers, brought an increase in urban crime incidence, which makes it necessary to allocate resources to ensure the safety of the population. Therefore, it becomes necessary to create smart tools that can help and guide public agents, thus enabling the optimization of the allocation of resources and the development of the planning of urban centers. In this context, the objective of this work is to develop predictive models for crime incidence in municipalities of the state of São Paulo in Brazil. To this end, the techniques known as ARIMA, Prophet and LSTM are used, which had their results evaluated based on certain metrics. The estimated models present satisfactory accuracy, which demonstrates the viability of such models as a management tool.