Determinando clusters de bairros em Fortaleza a partir da trajetória temporal das taxas de criminalidade

Abstract

The main objective of this study was to estimate clusters of the neighborhoods of Fortaleza, according to the similarity of the temporal evolution of the theft and robbery rates presented in the period from 2009 to 2020 by each spatial unit. Through the use of an algorithm known in the literature as the hierarchical agglomerative clustering (HAC) algorithm, four spatial clusters were formed according to the similarity of the trajectory of the time series of theft rates in the neighborhoods of Fortaleza, and another four based on the evolution of the series time of theft rates. To evaluate the temporal evolution of theft and robbery rates of each cluster, first the final values of the indicators were compared with the initial ones, and the instantaneous and the compound (annual) growth rate were estimated through a semilogarithmic model called logline. The clusters in which crime rates grew over time were composed of neighborhoods with low levels of economic development, with a high percentage of their populations in extreme poverty and with low average monthly income. On the other hand, most of the neighborhoods with good socioeconomic indicators were allocated to spatial groups that presented, at the end of their temporal trajectories, theft and robbery rates lower than the initial ones.