This chapter covers exponential smoothing approaches to smoothen time series data. The approaches can be easily extended for the forecasting by including terms such as smoothing factor, trend factor, and seasonality factor. The single order exponential smoothing performs smoothing using only the smoothing factor, which is further extended by second order smoothing factor by including the trend component. The third order smoothing was also covered, which incorporates all smoothing, trend, and seasonality factors into the model.
This chapter covered all these models in detail with their Python implementation. The smoothing approaches can be used to forecast if the time series is a stationary signal. However, the assumption may not be true. Higher-order exponential smoothing is recommended but its computation becomes hard. Thus, to deal with the approach, other forecasting techniques such as Autoregressive Integrated Moving Average (ARIMA) is proposed, which will be covered in the next...
