Solutions to Forecasting Problem Set#1

Problem 12-15
Forecasting 2 period moving average

Input Data Forecast Error Analysis

Actual Absolute Squared Absolute
Period Value Forecast Error error error % error
Period 1 4
Period 2 6
Period 3 4 5.000 -1.000 1.000 1.000 25.00%
Period 4 5 5.000 0.000 0.000 0.000 0.00%
Period 5 10 4.500 5.500 5.500 30.250 55.00%
Period 6 8 7.500 0.500 0.500 0.250 6.25%
Period 7 7 9.000 -2.000 2.000 4.000 28.57%
Period 8 9 7.500 1.500 1.500 2.250 16.67%
Period 9 12 8.000 4.000 4.000 16.000 33.33%
Period 10 14 10.500 3.500 3.500 12.250 25.00%
Period 11 15 13.000 2.000 2.000 4.000 13.33%
Average 2.222 7.778 22.57%
Next period 14.500 MAD MSE MAPE

Problem 12-15
Forecasting 4 period moving average

Input Data Forecast Error Analysis

Actual Absolute Squared Absolute
Period Value Forecast Error error error % error
Period 1 4
Period 2 6
Period 3 4
Period 4 5
Period 5 10 4.750 5.250 5.250 27.563 52.50%
Period 6 8 6.250 1.750 1.750 3.063 21.88%
Period 7 7 6.750 0.250 0.250 0.063 3.57%
Period 8 9 7.500 1.500 1.500 2.250 16.67%
Period 9 12 8.500 3.500 3.500 12.250 29.17%
Period 10 14 9.000 5.000 5.000 25.000 35.71%
Period 11 15 10.500 4.500 4.500 20.250 30.00%
Average 3.107 12.920 27.07%
Next period 12.500 MAD MSE MAPE
Note: this question provides a starting forecasting of 410

Problem 12-18
Forecasting Exponential smoothing

Input Data Forecast Error Analysis

Actual Absolute Squared Absolute
Period value Forecast Error error error % error
Year 1 450 410.000
Year 2 495 422.000 73.000 73.000 5329.000 14.75%
Year 3 518 443.900 74.100 74.100 5490.810 14.31%
Year 4 563 466.130 96.870 96.870 9383.797 17.21%
Year 5 584 495.191 88.809 88.809 7887.038 15.21%
Average 83.195 7022.661 15.37%
Alpha 0.3 MAD MSE MAPE

Year 6 = 521.834

Problem 12-19
Forecasting Exponential smoothing
Input Data Forecast Error Analysis
Actual Absolute Squared Absolute
Period value Forecast Error error error % error
Year 1 450 410.000
Year 2 495 434.000 61.000 61.000 3721.000 12.32%
Year 3 518 470.600 47.400 47.400 2246.760 9.15%
Year 4 563 499.040 63.960 63.960 4090.882 11.36%
Year 5 584 537.416 46.584 46.584 2170.069 7.98%
Average 54.736 3057.178 10.20%
Alpha 0.6 MAD MSE MAPE

Year 6 = 565.366

Problem 12-19
Forecasting Exponential smoothing
Input Data Forecast Error Analysis
Actual Absolute Squared Absolute
Period value Forecast Error error error % error
Year 1 450 410.000
Year 2 495 446.000 49.000 49.000 2401.000 9.90%
Year 3 518 490.100 27.900 27.900 778.410 5.39%
Year 4 563 515.210 47.790 47.790 2283.884 8.49%
Year 5 584 558.221 25.779 25.779 664.557 4.41%
Average 37.617 1531.963 7.05%
Alpha 0.9 MAD MSE MAPE

Year 6 = 581.422
Problem 12-20
The smoothing constant of 0.9 gives the most accurate forecast (lowest forecast error).

Problem 12-21
Forecasting 3 period moving average
Input Data Forecast Error Analysis
Actual Absolute Squared Absolute
Period Value Forecast Error error error % error
Year 1 450
Year 2 495
Year 3 518
Year 4 563 487.667 75.333 75.333 5675.111 13.38%
Year 5 584 525.333 58.667 58.667 3441.778 10.05%
Average 67.000 4558.444 11.71%
Year 6 = 555.000 MAD MSE MAPE

Problem 12-26
Forecasting Exponential smoothing
Input Data Forecast Error Analysis
Actual Absolute Squared Absolute
Period value Forecast Error error error % error
Week 1 17 17.000
Week 2 21 17.000 4.000 4.000 16.000 19.05%
Week 3 19 17.800 1.200 1.200 1.440 6.32%
Week 4 23 18.040 4.960 4.960 24.602 21.57%
Week 5 18 19.032 -1.032 1.032 1.065 5.73%
Week 6 16 18.826 -2.826 2.826 7.984 17.66%
Week 7 20 18.260 1.740 1.740 3.026 8.70%
Week 8 18 18.608 -0.608 0.608 0.370 3.38%
Week 9 22 18.487 3.513 3.513 12.343 15.97%
Week 10 20 19.189 0.811 0.811 0.657 4.05%
Week 11 15 19.351 -4.351 4.351 18.935 29.01%
Week 12 22 18.481 3.519 3.519 12.382 15.99%
Average 2.596 8.982 13.40%
Alpha 0.2 MAD MSE MAPE

Week 13 = 19.185
 Problem 12-27
Forecasting Exponential smoothing
Input Data Forecast Error Analysis
Actual Absolute Squared Absolute
Period value Forecast Error error error % error
Week 1 50 50.000
Week 2 35 50.000 -15.000 15.000 225.000 42.86%
Week 3 25 48.500 -23.500 23.500 552.250 94.00%
Week 4 40 46.150 -6.150 6.150 37.823 15.38%
Week 5 45 45.535 -0.535 0.535 0.286 1.19%
Week 6 35 45.482 -10.482 10.482 109.862 29.95%
Week 7 20 44.433 -24.433 24.433 596.989 122.17%
Week 8 30 41.990 -11.990 11.990 143.760 39.97%
Week 9 35 40.791 -5.791 5.791 33.536 16.55%
Week 10 20 40.212 -20.212 20.212 408.521 101.06%
Week 11 15 38.191 -23.191 23.191 537.810 154.60%
Week 12 40 35.872 4.128 4.128 17.043 10.32%
Week 13 55 36.284 18.716 18.716 350.271 34.03%
Week 14 35 38.156 -3.156 3.156 9.961 9.02%
Week 15 25 37.840 -12.840 12.840 164.877 51.36%
Week 16 55 36.556 18.444 18.444 340.167 33.53%
Week 17 55 38.401 16.599 16.599 275.535 30.18%
Week 18 40 40.061 -0.061 0.061 0.004 0.15%
Week 19 35 40.055 -5.055 5.055 25.549 14.44%
Week 20 60 39.549 20.451 20.451 418.237 34.08%
Week 21 75 41.594 33.406 33.406 1115.945 44.54%
Week 22 50 44.935 5.065 5.065 25.656 10.13%
Week 23 40 45.441 -5.441 5.441 29.608 13.60%
Week 24 65 44.897 20.103 20.103 404.123 30.93%
Average 13.250 253.166 40.61%
Alpha 0.1 MAD MSE MAPE

Week 25 = 46.907

Problem 12-27 and 12-28

MAD MSE MAPE Week 25 forecast
Alpha = 0.6 14.039 232.157 38.14% 57.599
Alpha = 0.9 14.480 256.855 38.54% 62.623
Optimal alpha = 0.3354057 12.627 214.201 35.37% 54.705
 Problem 12-29
Forecasting Exponential smoothing
Input Data Forecast Error Analysis
Actual Absolute Squared Absolute
Period value Forecast Error error error % error
Feb 70 65.000
Mar 68.5 65.500 3.000 3.000 9.000 4.38%
April 64.8 65.800 -1.000 1.000 1.000 1.54%
May 71.7 65.700 6.000 6.000 36.000 8.37%
June 71.3 66.300 5.000 5.000 25.000 7.01%
July 72.8 66.800 6.000 6.000 36.000 8.24%
Average 4.200 21.400 5.91%
Alpha 0.1 MAD MSE MAPE

Future forecasts are all the same (67.4). Why this is true will be discussed in our next class!

Problem 12-30
Forecasting Exponential smoothing
Input Data Forecast Error Analysis
Actual Absolute Squared Absolute
Period value Forecast Error error error % error
Feb 70 65.000
Mar 68.5 66.500 2.000 2.000 4.000 2.92%
April 64.8 67.100 -2.300 2.300 5.290 3.55%
May 71.7 66.410 5.290 5.290 27.984 7.38%
June 71.3 67.997 3.303 3.303 10.910 4.63%
July 72.8 68.988 3.812 3.812 14.532 5.24%
Average 3.341 12.543 4.74%
Alpha 0.3 MAD MSE MAPE

Future forecasts are all the same (70.132).

Optima alpha is 0.56382774 based on minimizing MSE