MGT 3501 Lecture Notes - Lecture 13: Moving Average
Document Summary
Assumes the most accurate forecast of future demand is a simple (linear) combination of past demand. Taking an average of past data to help predict what"s going to happen in the future. Typically used when no seasonality or trend exists. Taking average of previous demand to forecast future demand. Time period for averages is up to you. Called moving average because as forecast time period move, will only look back at the specific previous time period. Depends on what the firm is after. If trying to look for 1 week trend, 3 weeks is probably better. Long term trend should use longer term average. Adjust time periods to suit your needs based on results. 3 weeks is more reactive, 6 weeks is more smooth. Gives more importance (weight) to some of the data. Similar to moving average with a bit of twist. Simple moving average gives every data point equal weight.