Abstract
The spread of AIDS was a striking social fact in the late twentieth century, mainly due to the lack of knowledge of sexually active people and drug users, spreading rapidly across five continents. Initially, it was associated with the group of male homosexuals. Over the years, other population segments became infected with the Human Immunodeficiency Virus (HIV). Thus, this article aims to compare forecasting methodologies to predict the incidence rate of AIDS per 100,000 inhabitants in Santa Maria between 2017 and 2022. For this purpose, two forecasting models were adjusted for each series (polynomial trend model plus an ARIMA model (p, d, q), and an exponential smoothing model). As the series show structural breakdowns due to various historical events in Brazil and around the world, prediction combinations methodologies were used through robust regressions, using the Weighted Least Squares, MM and Quantile Regression methods. We verified through the accuracy measures that, for men, the best forecasting methodologies were Model 1 and the regression forecast combinations, using the MM and RQ methods. For women, the best methodologies were Model 1 and regression prediction combinations by the RQ method.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have