4. The file Department StoreSales.xls contains data on the quarterly sales for a department store over a 6-year period. (27 pts) The time plot of the data is shown below. 120000 100000 80000 60000 / (Image by Paul Martin Eldridge / FreeDigitalPho- tos.net) 40000 20000 0 Figure 5.10: Department store quarterly sales series 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Quarter a) Describe the data frequency of this time plot. Which of the four components level, trend, seasonality, noise) seem to be present in this series? b) Which of the following methods would NOT be suitable for forecasting this series? Explain. Nave forecasting Seasonal nave forecasting Simple exponential smoothing of the raw series Holt-Winter's exponential smoothing of the raw series c) A forecaster was tasked to generate forecasts for the last four quarters ahead. She therefore partitioned the data so that the last 4 quarters were designated as the validation period. The forecaster approached the forecasting task by using multiplicative Holt-Winter's exponential smoothing. The smoothing constants used were a = 0.2, p = 0.15, y = 0.05. She plots the estimated forecasts in Figure 5.11. The forecasts for the validation set are given in the Table below. Compute the MAPE values for the forecasts of quarters 21-22. MAPE Exp Smoothing: Actual Vs Forecast (Training Data) Figure 5.11: Forecasts and actuals (top) and forecast errors (bottom) using expo- nential smoothing 00000 30000 20000 w 90000 40000 Force 10000 3 5 7 13 17 19 Quarter 4. The file Department StoreSales.xls contains data on the quarterly sales for a department store over a 6-year period. (27 pts) The time plot of the data is shown below. 120000 100000 80000 60000 / (Image by Paul Martin Eldridge / FreeDigitalPho- tos.net) 40000 20000 0 Figure 5.10: Department store quarterly sales series 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Quarter a) Describe the data frequency of this time plot. Which of the four components level, trend, seasonality, noise) seem to be present in this series? b) Which of the following methods would NOT be suitable for forecasting this series? Explain. Nave forecasting Seasonal nave forecasting Simple exponential smoothing of the raw series Holt-Winter's exponential smoothing of the raw series c) A forecaster was tasked to generate forecasts for the last four quarters ahead. She therefore partitioned the data so that the last 4 quarters were designated as the validation period. The forecaster approached the forecasting task by using multiplicative Holt-Winter's exponential smoothing. The smoothing constants used were a = 0.2, p = 0.15, y = 0.05. She plots the estimated forecasts in Figure 5.11. The forecasts for the validation set are given in the Table below. Compute the MAPE values for the forecasts of quarters 21-22. MAPE Exp Smoothing: Actual Vs Forecast (Training Data) Figure 5.11: Forecasts and actuals (top) and forecast errors (bottom) using expo- nential smoothing 00000 30000 20000 w 90000 40000 Force 10000 3 5 7 13 17 19 Quarter
