Análise exploratória de simulações sociais computacionais por meio de estatística multivariada e mapas auto-organizáveis

Abstract

In order to provide valuable information for the assessment of the Territorial Plan for Sustainable Rural Development of the Ministry of Agrarian Development this work aimed at the use of principal component analysis, cluster analysis by the agglomerative hierarchical method and the Self Organizing Map artificial neural network in the exploratory analysis of the results of computer simulations of the social socioterritorial system "Southern Rural Territory of Sergipe". The methodology has been based in the Sociology of Organized Action and Soclab framework. Statistical analyses showed that this socio-territorial system has a simple and deterministic structure, so, presents a cooperative social game with a strong tendency to reach a steady state even in situations of conflicting interests. The neural analysis allowed the characterization of atypical situations when the stability of the social system occurs.