\( \newcommand{\mathbbm}[1]{\boldsymbol{\mathbf{#1}}} \)

Chapter 1 Introduction

I started writing this monograph in 2020 during the COVID-19 pandemic, having figured out that I was bored to death in isolation and needed to do something useful. By that time, I had done a substantial amount of work in the area of dynamic models and tried to publish several papers on statistical models developed in the Single Source of Error (SSOE) framework. I had even developed several R functions based on my own ideas, which were all available in the smooth package and attracted substantial interest in the forecasting and data science communities, but had not been published anywhere. Furthermore, a friend of mine, Nikos Kourentzes, had been telling me that when he used my functions, he could not reference them properly because of the lack of publications from my side. So, it became apparent that I needed to either publish lots of papers, covering different small aspects of what I had done, or write a monograph that would summarise everything in one place. Feeling lonely and depressed because of the lockdown, I chose the second option.

At this stage, I should mention that all my ideas rely on the framework from the monograph of Hyndman et al. (2008), which I have modified and upgraded. Their original book discussed the ETS (Error-Trend-Seasonality) model in the SSOE form, but I have decided to expand it, introducing more features that are required in day-by-day demand forecasting. So, for example, while the original ETS works very well on monthly, quarterly, and yearly data, my modifications support high frequency and/or intermittent data, work with explanatory variables, and overall represent a holistic view on demand forecasting in practice.

However, before we move to the discussion of the framework, I should point out that many parts of this monograph rely on such topics as scales of information, model uncertainty, likelihood, information criteria, and model building. All these topics are discussed in detail in the online lecture notes of Svetunkov (2022). It is recommended that you familiarise yourself with them before moving to ADAM’s more advanced topics.

In this chapter, I explain what forecasting is, how it is different from planning and analytics, and what the main forecasting principles are. All these aspects will help you not to fail in trying to predict the future.

References

• Hyndman, R.J., Koehler, A.B., Ord, J.K., Snyder, R.D., 2008. Forecasting with Exponential Smoothing: The State Space Approach. Springer Berlin Heidelberg. https://doi.org/10.1007/978-3-540-71918-2
• Svetunkov, I., 2022. Statistics for business analytics. https://openforecast.org/sba/ version: 31.10.2022