Chapter 5: Time Series Components and Statistical Properties
In the introductory chapters, we defined some basic features of the Time Series and learned how to describe them graphically. Now, before going into the challenging methodologies used to build a real forecast based on algorithms of different types, it is necessary to cover some properties of Time Series in detail—properties that will be important to better apply and understand the forecasting models of the following chapters.
As you may have guessed in the previous pages of the book, learning to use data from a Time Series and, above all, creating a reliable forecast of the same for the future, depends heavily on the ability to fully understand the temporal dynamics of the process underlying the data. What is the non-random structure that can be extracted from the observations? What are the measurable regularities over time? What is the relationship between the observation at time (t) and the observation at time...
