Previsão de inflação no Brasil: uma aplicação de redes neurais artificiais ao regime de metas de inflação

by Nurchis, Mariana

Inflation targeting was adopted in Brazil in 1999. Since then, the importance ofinflation forecast has been growing, mainly because its larga influence in monetarypolicy decisions.Besides the increasing number of working papers about inflation forecasts forBrazil, nona uses artificial neural networks as forecast tool. Neural nets have greatpopularity in many areas of the human knowledge for its recognized capacity toapproach almost any nonlinear continuous function.Inspired by studies for countries like Canada, United Kingdom and SouthAfrica, this paper intends to forecast inflation for Brazil through artificial neuralnetworks