Summary

In this chapter, we have introduced some key concepts that will form the components of the analyses shown in the following chapters. First, we learned how to isolate and recognize the trend and seasonality of a time series, either through techniques dedicated to the estimation of individual components or by using a joint approach, such as decomposition. Finally, we introduced the concepts of autocorrelation and stationarity, which are the foundations on which ARIMA models are built.

Regardless of the industry in which one finds oneself working and the type of time series one wants to predict, knowing the statistical properties of the target series and how to correctly identify trends and seasonality can help any business analyst be more effective in isolating patterns of interest, whether for descriptive or predictive purposes. The first building block of Time Series analysis is always the ability to correctly decompose a Time Series, and, in this chapter, we have provided...