Mortalidade por acidentes de trânsito no brasil: uma análise de série temporal

Abstract

Traffic accidents are a serious global public health issue, resulting from a combination of behavioral factors, deficiencies in vehicle safety, and poor urban conditions. They represent one of the main causes of injuries and fatalities, driven by the continuous increase in the number of vehicles, changes in lifestyles, and observed risk behaviors in the general population. For this reason, this study investigates the monthly time series of traffic accident mortality in Brazil, aiming to identify a model capable of making predictions for this variable. Time series techniques were employed, using the Box-Jenkins approach, with seasonal autoregressive integrated moving average (SARIMA) models. The criteria adopted to make an analytical decision about the most appropriate model included the Akaike Information Criterion (AIC) and the Bayesian Information Criterion (BIC). Additionally, to increase the accuracy of the analysis, the Mean Absolute Percentage Error (MAPE) and the Root Mean Square Error (RMSE) were also calculated. The chosen model was SARIMA(2,1,1)(3,1,1)12, which presented the lowest value for these criteria.