Sequence stationary process. Eliminate possible cyclical and seasonal behavior. In order to improve the forecast with
low order models, it is convenient to eliminate this behavior. We usually remove this seasonal behavior by means of
differentiation. To identify possible seasonal and/or cyclical components present in the time series, the autocorrelation
function (ACF) is used. If the sequence is non-stationary, generally one or two time differences can make the sequence
smoother by difference processing