Identificação do movimento linear da série temporal do volume mensal de transações da Criptomoeda Ethereum por meio do modelo SARIMA

Abstract

This work presents the modeling of the monthly volume of transactions of the Ethereum cryptocurrency through theBox-Jenkins methodology, involving the steps: exploratory analysis, identification, guarantee and validation, some ofwhich performed using different techniques. The model found by the SARIMA modeling (Integrated autoregressivemodel of seasonal moving averages) was able to demonstrate the linear behavior of the data in a satisfactory way, butit was not enough to describe the behavior of the series, composed of linear and non-linear movement, being betterrepresented by a hybrid model.