Exponential smoothing is a method or technique in Statistics that is easily applied in smoothing a discrete time series with an aim of conducting forecasts. The popularity of this technique mainly stems from how simple it is in application, its efficiency in computing and the ease with which it responds to process variations that help with the forecasting process (Ostertagová & Ostertag, 2011). The procedure’s main conception is to help smooth the time series in its original form similar to moving average then forecasting the variable in question by use of the already smoothed time series. However, as it occurs in exponential smoothing, the most current of values in the time series usually have the most impact on the value that results after forecast as compared to past observations. This approach is considered to be practical as well as simplistic in matters to do with forecasting where the value forecast results from the weighted average of recent observation with the most recent observation in the time series taking on the least weight while the present observation in the series takes on the largest weight.
