Unformatted text preview: 6 interactive Models for Operations and Supply Chain Management b. Experiment to ﬁnd the alpha value that provides the best value for the MAD. Identify
the alpha value that provides the best MFE. Are they the same? Explain. c. If you were in Connie’s position, which measure would be most important to you?
Why? 3. Three months after the demand data above were collected, Connie gathered another
12 weeks of demand. The new demand is shown below: Period 1 2 3 4 5 6 7
Demand 52 56 67 73 66 J 63 56 4. Enter the new demand into the table in the Simple Exponential Smoothing Model. a. With an alpha of 0.1, examine how the forecast responds to demand changes as demand increases (periods l—4) and as demand decreases (periods 4—9). Explain
your observations. b. As the demand appears to change directions at periods 4 and 9, explain how the value
of alpha affects the ability of the forecast to keep up with the demand. Examine the
period of increasing demand and the period of decreasing demand. During these times, what happens to the bias of the forecast? How does the alpha affect the bias
during these short periods? c. Based on your observations of simple exponential smoothing when demand is in- creasing and when demand is decreasing, explain why simple exponential smoothing
is not recommended when the time series contains a trend. Demand Forecasting 7 The Trend Enhanced Exponential Smoothing Model Trend adjusted exponential smoothing incorporates trend into the exponential
smoothing forecast through the use of an additional smoothing constant. The model
treats trend and random ﬂuctuation components quite separately. In the calculation
of the trend enhanced exponential smoothing forecast, the smoothed forecast is
calculated exactly as in simple exponential smoothing. A trend component (Tt) is
then added. Tt is computed by adding to the previous trend component (TH), a
weighted difference between the previous two trend adjusted forecasts and the
previous trend. The second smoothing constant (B) determlnes the welght used.
Exhibit 1.3 shows the Trend Enhanced Exponential Smoothing Model. — Demand W Forecast 43
N CD . r—u—a | ~—o——i I 10111213i41511718192 DUDDDDDDDD EXHIBIT 1.3 Screen View of the Trend Enhanced Exponential Smoothing Model ...
View Full Document
- Spring '08