So far, we have mainly been focusing on the first two components of the time series analysis workflow—data preprocessing and descriptive analysis. Starting from this chapter, we will shift gear and move on to the third and last component of the analysis—the forecast. Before we dive into different forecasting models in the upcoming chapters, we will introduce the main elements of the forecasting workflow. This includes approaches for training a forecasting model, performance evaluation, and benchmark methods. This will provide you with a set of tools for designing and building a forecasting model according to the goal of the analysis.
This chapter covers the following topics:
- Training and testing approaches for a forecasting model
- Performance evaluation methods and error measurement matrices
- Benchmark methods
- Quantifying forecast uncertainty...
