In: Operations Management
Topic 3
Book: Operations and Supply Chain Management Jacobs & Chase 14e
6. a.) Evaluate the forecasting model using 3 month moving average, and 3 month moving weighted average, and exponential. The weights are .5 for the most recent demand, .25 for the other months. Alpha = .3. Use the weighted moving average for January Forecast.
b.) Calculate MAD and MAPE for each and compare. Which method is a better forecast and why?
Demand
Oct 300
Nov 360
Dec 425
Jan 405
Feb 430
Mar 505
Apr 550
May 490
FORECAST = SIGMA(PREVIOUS N DEMANDS) / N
WHERE N = 3
FORECAST 4 = (300 + 360 + 425) / 3 = 361.7
FORECAST 5 = (360 + 425 + 405) / 3 = 396.7
FORECAST 6 = (425 + 405 + 430) / 3 = 420
FORECAST 7 = (405 + 430 + 505) / 3 = 446.7
FORECAST 8 = (430 + 505 + 550) / 3 = 495
FORECAST 9 = (505 + 550 + 490) / 3 = 515
FORECAST ERROR
PERIOD |
ACTUAL DEMAND |
FORECAST |
DEVIATION(D - F) |
ABS DEVIATION |
(ABS DEV / DEMAND) * 100 |
1 |
300 |
||||
2 |
360 |
||||
3 |
425 |
||||
4 |
405 |
361.7 |
43.3 |
43.3 |
10.69 |
5 |
430 |
396.7 |
33.3 |
33.3 |
7.74 |
6 |
505 |
420 |
85 |
85 |
16.83 |
7 |
550 |
446.7 |
103.3 |
103.3 |
18.78 |
8 |
490 |
495 |
-5 |
5 |
1.02 |
SIGMA |
259.9 |
269.9 |
55.06 |
MAD = SIGMA(ABSOLUTE DEVIATION) / N, WHERE N = 5
MAD = 269.9 / 5 = 53.98
MAPE = SIGMA(ABSOLUTE DEVIATION / DEMAND * 100) / N, WHERE N = 5
MAPE = 55.06 / 5 = 11.01
2. FORECAST = SIGMA(WEIGHT FOR PERIOD * DEMAND PER PERIOD) / SUM OF THE WEIGHTS
WHERE LARGEST WEIGHTS ARE MULTIPLIED BY THE MOST RECENT DEMANDS
FORECAST 4 = ((425 * 0.5) + (360 * 0.25) + (300 * 0.25)) / 1 = 377.5
FORECAST 5 = ((405 * 0.5) + (425 * 0.25) + (360 * 0.25)) / 1 = 398.75
FORECAST 6 = ((430 * 0.5) + (405 * 0.25) + (425 * 0.25)) / 1 = 422.5
FORECAST 7 = ((505 * 0.5) + (430 * 0.25) + (405 * 0.25)) / 1 = 461.25
FORECAST 8 = ((550 * 0.5) + (505 * 0.25) + (430 * 0.25)) / 1 = 508.75
FORECAST 9 = ((490 * 0.5) + (550 * 0.25) + (505 * 0.25)) / 1 = 508.75
FORECAST ERROR
PERIOD |
ACTUAL DEMAND |
FORECAST |
DEVIATION(D - F) |
ABS DEVIATION |
(ABS DEV / DEMAND) * 100 |
1 |
300 |
||||
2 |
360 |
||||
3 |
425 |
||||
4 |
405 |
377.5 |
27.5 |
27.5 |
6.79 |
5 |
430 |
398.75 |
31.25 |
31.25 |
7.27 |
6 |
505 |
422.5 |
82.5 |
82.5 |
16.34 |
7 |
550 |
461.25 |
88.75 |
88.75 |
16.14 |
8 |
490 |
508.75 |
-18.75 |
18.75 |
3.83 |
SIGMA |
211.25 |
248.75 |
50.37 |
MAD = SIGMA(ABSOLUTE DEVIATION) / N, WHERE N = 5
MAD = 248.75 / 5 = 49.75
MAPE = SIGMA(ABSOLUTE DEVIATION / DEMAND * 100) / N, WHERE N = 5
MAPE = 50.37 / 5 = 10.07
3. FORECAST = FORECAST + (ALPHA * (ACTUAL DEMAND - FORECAST))
FORECAST 2 = 377.5 + (0.3 * (300 - 377.5) = 354.3
FORECAST 3 = 354.3 + (0.3 * (360 - 354.3) = 356
FORECAST 4 = 356 + (0.3 * (425 - 356) = 376.7
FORECAST 5 = 376.7 + (0.3 * (405 - 376.7) = 385.2
FORECAST 6 = 385.2 + (0.3 * (430 - 385.2) = 398.6
FORECAST 7 = 398.6 + (0.3 * (505 - 398.6) = 430.5
FORECAST 8 = 430.5 + (0.3 * (550 - 430.5) = 466.4
FORECAST 9 = 466.4 + (0.3 * (490 - 466.4) = 473.5
PERIOD |
ACTUAL DEMAND |
FORECAST |
DEVIATION(D - F) |
ABS DEVIATION |
1 |
300 |
377.5 |
-77.5 |
77.5 |
2 |
360 |
354.3 |
5.7 |
5.7 |
3 |
425 |
356 |
69 |
69 |
4 |
405 |
376.7 |
28.3 |
28.3 |
5 |
430 |
385.2 |
44.8 |
44.8 |
6 |
505 |
398.6 |
106.4 |
106.4 |
7 |
550 |
430.5 |
119.5 |
119.5 |
8 |
490 |
466.4 |
23.6 |
23.6 |
SIGMA |
319.8 |
474.8 |
MAD = SIGMA(ABSOLUTE DEVIATION) / N, WHERE N = 8
MAD = 474.8 / 8 = 59.35
MAPE = SIGMA(ABSOLUTE DEVIATION / DEMAND * 100) / N, WHERE N = 8
MAPE = 108.68 / 8 = 13.59
4. BASED ON THE MAD AND MAPE, WEIGHTED MOVING AVERAGE IS THE BETTER FORECASTING METHOD
** Leaving a thumbs-up would really help me out. Let me know if you face any problems.