Problem: Suppose that the proprietors of Hamish and Roger’s Public House use simple exponential smoothing (with α = 0.2) to forecast monthly beer sales. After April’s demand is observed, according to their model, the forecasted demand for May is 4000 cans of beer. Answer the following without software – that is, use hand calculation.

1. At the beginning of May, what is the forecast of July ’s beer sales?
2. Suppose that actual demands during May and June are as follows: May, 4500 cans of beer; June, 3500 cans of beer. After observing June’s demand, what is the forecast for July’s demand using the exponential smoothing model?
3. Based on the data from part b, the average demand during May and June is (4500+3500)/2 = 4000 cans per month. This is the same as the forecast for July sales before we observed the May and June data. Yet after we observe the May and June demands for beer, our forecast for July demand has decreased from what it was at the beginning of May. Why?

Problem: The following table gives the numbers of visitors since the opening of a museum in January 1998. It also contains a worksheet for calculating forecasts with a centered moving average, adjusted for multiplicative seasonal effects and linear trend. As shown, the trend equation has an intercept of 3,103.2 and slope of 223.5 visitors. Cells with labels A through H require calculation (see below).

1. Complete the calculations for the cells labeled A through H.
2. How many visitors does this method predict for the third quarter of 2000 (Q3-00)?
3. What would the method predict for the first quarter of 2001?

Problem: a) The file Tapes.xlsx contains data on 18 months of musical tape sales. Plot these data and explain why simple exponential smoothing may or may not be useful in this case. Justify your explanation by creating a simple exponential smoothing model with a smoothing constant of 0.2, and plot the actual sales and the smoothed values for the 18 months, and comment on how well the model appears to fit the data.

b) The file Turnout.xlsx gives the percentage voter turnout at US national elections from 1972 to 1996. Plot the values over time, and note that Presidential elections are every 4 years in the US. Considering this, and the appearance of the plot, create an exponential smoothing model to predict the percentage voter turnout in 1998 and 2000 using a smoothing constant of 0.2. Find a smoothing constant for voter turnout that minimizes the RMSE.

Problem: The file CocaCola.xls contains quarterly sales data for the period January 1986 to June 1996. Consider each of the following to be independent of each other.

1. Develop a forecast for the last two quarters of 1996.
2. What is the RMSE of your model?
3. Develop a second model without using the last four quarters. Compare the forecast which your model makes for the final four quarters against the actual values that you did not use. What is the MAD of this model for the final four months (forecast versus actual).

Price: $21.83

Solution: The downloadable solution consists of 13 pages, 883 words and 2 charts.
Deliverable: Word Document