Modelagem preditiva de avaliação de indicadores sociais que impactam a criminalidade - uma análise big data com ênfase na segurança pública
Nenhuma Miniatura disponível
Data
2018-06-07
Autores
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Federal de Goiás
Resumo
This study aims to identify determinants of crime, analyzing social indicators that indicate the movement and the tendency of violence and proposing eficiente actions of the State in a particularly critical area of our society, Public Security. To do this, he proposed to solve the following research problem: "How to construct a model of evaluation of social indicators, using a technique of Big Data, using Public Safety as an anchor, associated to the themes Education; Work and Income; Social Inequality and Urbanization, Development and Infrastructure, in order to identify the driving causes of crime?".
The choice of this theme is justified insofar as the result of this study may be able to provide an orientation that identifies spatial and temporal patterns of certain types of crime in order to anticipate the occurrence of events.
To do so, the following steps were followed: identification of the main indicators related to Public Safety and, based on Education, Work and Income; Social inequality; and Urbanization, Development and Infrastructure; selection of statistical methods for the study (Principal Component Analysis - PCA and Hierarchical Cluster Analysis - HCA); data collection of the 246 municipalities of the State of Goiás; a statistical analysis of the indicators identified; and the proposal of a Predictive Modeling of Social Indicators, containing a Agenda Setting, separating the social areas in which public programs must be elaborated, identified from the correlations of the indicators.
The results indicated that through statistical tools it is possible to propose a model for identifying social indicators that influence the increment of crime, since crime tends to be generated by multiple factors, and its combat should require a proactive position of diagnosis of the variables social movements.
Descrição
Palavras-chave
Citação
AMARAL, C. A. Modelagem preditiva de avaliação de indicadores sociais que impactam a criminalidade - uma análise big data com ênfase na segurança pública. 2018. 116 f. Dissertação (Mestrado em Administração Pública em Rede Nacional) - Universidade Federal de Goiás, Goiânia, 2018.