Análisis de una serie de tiempo utilizando diseño de experimentos como herramienta de calibración Análisis de una serie de tiempo utilizando diseño de experimentos como herramienta de calibración

Abstract

Time series are used to study the relationship of a variable with itself over time at regular intervals. Energy consumption in Spain was considered for a sample of five days, using various deterministic models sought to model their behavior in the most accurate way. The design of experiments is used to calibrate the model parameters Holt-Winters validating those effects that are significant in minimizing MAPE,in order to identify the optimum operating conditions of the model. Finally, various ARIMA models applied to residues obtained from Holt-Winters model to make it white noise, using the Box-Jenkins methodology are evaluated.