Nueva aproximación al análisis de series temporales interrumpidas

Abstract

In the present work, we show an alternative to the verification of the intervention effect in time series to which the ARIMA model has been applied. This alternative is based on the consideration of significant differences in the parameters of ARIMA models, taking into account only the base line and also taking into account the complete series – base line and intervention phase. In order to be able to determine when the differences between parameters are significant, an investigation is carried out by means of the Montecarlo method so that different significance levels are defined statistically.