solution

Using the 911 call data in Problem 4.31, forecast calls for weeks 2 through 25 with a trend-adjusted exponential smoothing model. Assume an initial forecast for 50 calls for week 1 and an initial trend of zero. Use smoothing constants of  = .3 and = .2. Is this model better than that of Problem 4.31? What adjustment might be useful for further improvement? (Again, assume that actual calls in week 25 were 85.)

Emergency calls to the 911 system of Durham, North Carolina, for the past 24 weeks are shown in the following table:

a) Compute the exponentially smoothed forecast of calls for each week. Assume an initial forecast of 50 calls in the first week, and use a = .2. What is the forecast for week 25?

b) Reforecast each period using a = .6.

c) Actual calls during week 25 were 85. Which smoothing constant provides a superior forecast? Explain and justify the measure of error you used.