exponential smoothing is meant to be used with time-series data when the data are made up of some or all of the basic components of level, trend, seasonalitiy, and error. if the data series only fluctuates about an average with no trend and no serasonalitiy, which form of smoothing wouldyou employ? if the data include all of these components, which form of smoothing would you employ? how should the smoothing constants be chosen?
