If a time series appears to contain a substantial irregular element,we would use a small value for the smoothing constant α since we do not want to give too much weight to the most recent observation along.

Question

Quizplus · Accepted Answer

If a time series appears to contain a substantial irregular element, we would use a larger value for the smoothing constant α so that the forecast is more responsive to the most recent observations, which may be closer to the true value. A small value of α would give more weight to past observations and produce a smoother forecast, which may not reflect the current trend or pattern in the data.

If a Time Series Appears to Contain a Substantial Irregular