Analyzing Criminal Macrocauses on Intentional Lethal Violent Crimes: An Unsupervised Learning Approach for Smart City Initiatives

In Brazil, intentional lethal violent crimes pose challenges to public safety initiatives and smart cities. We conducted a study utilizing clustering techniques as part of unsupervised learning to analyze the criminal macrocauses of such crimes. We explored different clustering strategies to identify the optimal number of clusters to group the ILCV data of the state of Rio Grande do Norte. By grouping this criminal data and applying the Distortion, Silhouette, and Calinski-Harabasz scores, we reached an ideal number of two to three clusters. Our findings align with smart city initiatives and present an opportunity to integrate data-driven approaches, theory, and machine learning techniques into public safety measures.