In Chapter 5, Decomposition of Time Series Data, we looked at the application of smoothing functions for noise reduction in time series data and trend estimation. In this chapter, we will expand on the use of smoothing functions and introduce their forecasting applications. This family of forecasting models can handle a variety of time series types, from series with neither trends nor seasonal components to series with both trends and seasonal components. We will start with the basic moving average model and simple exponential smoothing models, and then add more layers to the model, as well as the model's ability to handle complex time series data.
In this chapter, we will cover the following topics:
- Forecasting with moving average models
- Forecasting approaches with smoothing models
- Tuning parameters for smoothing models