Machine Learning for pattern detection and prediction of criminal occurrences

• Autor(a) principal: Fonseca, Luís Miguel Marques
• Data de Publicação: 2020
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10773/29209
• topic: Machine learning; Crime prediction; Smart cities; Big data

The increase of the world population, especially in large urban centers, has resulted in new challenges such as the management of natural resources and infrastructures as well as the optimization of services to promote the quality of citizens’ life. One of the biggest and most important challenges is the management of public safety, since, in addition to being a factor of interest to both the general population and the authorities, it is also an area that influences other essential indicators in a city such as tourism and employment. Public Safety has impact on the economic growth and social development of a community. This dissertation proposes a solution for the prediction of criminal occurrences in a city based on historical data of incidents and demographic data. The entire life cycle of the model’s learning process will be presented to provide an organization with predictive capability: start with the data collection from its original source, the treatment and transformations applied to them, the choice and the evaluation and implementation of the Machine Learning model up to the application layer. Classification models will be implemented to predict criminal risk for a given time interval and location, as well as regression models to predict the number of crimes. Machine Learning algorithms, such as Random Forest, Neural Networks, K-Nearest Neighbors and Logistic Regression will be used to predict occurrences, and their performance will be compared according to the data processing and transformation used. The results of the chosen model show that the use of Machine Learning techniques helps to anticipate criminal occurrences, which contributed to the reinforcement of public security. Finally, the models will be implemented on a platform that provides an API to enable other entities to request for predictions in real-time. An application will also be presented where it is possible to show criminal occurrences predictions visually.