Examples
1.
Period Actual Forecas Error, =Actual- |e| Error %
t Forecast squared, e2 Error
1 217 215 2 2 4 0.92%
2 213 216 -3 3 9 1.41%
3 216 215 1 1 1 0.46%
4 210 214 -4 4 16 1.90%
5 213 211 2 2 4 0.94%
6 219 214 5 5 25 2.28%
7 216 217 -1 1 1 0.46%
8 212 216 -4 4 16 1.89%
-2 22 76 10.27%
Using the figures shown in the table,
MAD = ∑ |e |/n = 22/8 = 2.75
MSE = ∑ e2/(n – 1) = 76/(8 – 1) = 10.86
MAP = (∑ [ |e|/Actual × 100])/n = 10.26%/8 = 1.28%
2.
Period Actual Forecast
Demand
1 42
2 40
3 43
4 40 41.666667
5 41 41
6 38 41.333333
7 39.666667
F6 = (A3+A4+A5)/3 = (43 + 40 + 41)/ 3 = 41.33
If actual demand in period 6 turns out to be 38, the moving average forecast for period 7
would be F7 = (A4+A5+A6)/3 = (40 + 41 + 38)/3 = 39.67
3.
Period Actual Forecast
1 42
2 40
3 43
4 40
5 41
6 39 41
7 40.2
a. F6 = .10(A5) + .20(A4) + .30(A3) + .40(A2) = .10(40) + .20(43) + .30(40) + .40(41) = 41.0
b. F7 =.10(A6) + .20(A5) + .30(A4) + .40(A3) = .10(43) + .20(40) + .30(41) + .40(39) = 40.2
Solved Problems
1. a. Plot the data to see if there is a pattern. Here we have only variations around an
average (i.e., no trend or cycles). Therefore, the most recent value of the series
becomes the next forecast: 64.
70
60
50
40
30
20
10
0.00
0
0 5 10 15 20 25 30
Actual Period
Forecast
b. Using the latest values. MA3 = (55 + 58 + 64)/ 3 = 59
c. F = .20(55) + .30(58) + .50(64) = 60.4
d. Starting with period 2, the data in period 1 is used as the forecast for period 2, so,
exponential smoothing for successive forecasts
Period Actual Forecast Calculations
1 60
2 65 60 the data in period 1 is used
as the forecast for period
3 55 59.2 60 + .40(65 – 60) = 62
4 58 58.72 62 + .40(55 – 62) = 59.2
5 64 60.83 59.2 + .40(58 – 59.2) =
58.72
6 59 58.72 + .40(64 – 58.72) =
60.83
2. a. The monthly forecast equation being used is: Ft = 402 + 3t
where t0 = January of last year
Ft = Forecast of shipments for month t
trend amounts for the first four months of next year: January, t = 24; February, t = 25; etc.
Thus,
FJan = 402 + 3(24)= 474
FFeb = 402 + 3(25)= 477
FMar = 402 + 3(26)= 480
FApr = 402 + 3(27)= 483
b. Multiply each monthly trend by the corresponding seasonal relative for that month
Month Seasonal Index Forecast = Index*demand
Jan 1.2 568.8
Feb 1.3 620.1
Mar 1.3 624
Apr 1.1 531.3
3.
Period Actual Demand
1 44
2 52
3 50
4 54
5 55
6 55
7 60
8 56
9 62
80
f(x) = 1.75 x + 45.4722222222222
70
60
50
40
30
20
10
0.00
0
0 5 10 15 20 25 30
Period
Period Actual
2
t t y ty
1 1 44 44
2 4 52 104
3 9 50 150
4 16 54 216
5 25 55 275
6 36 55 330
7 49 60 420
8 64 56 448
9 81 62 558
45 285 488 2,545
b= (n∑ty − ∑t∑y)/(n∑ t 2 − (∑t)2) = {9(2, 545)− 45(488)}/{9(285)− 45(45)} = 1.75
a= (∑y − b∑t)/ n = {488 − 1.75(45)}/ 9 = 45.47
Thus, the trend equation is Ft = 45.47 + 1.75t. The next two forecasts are:
F10 = 45.47 + 1.75(10)= 62.97
F11 = 45.47 + 1.75(11)= 64.72
Perio Actual Techni Err |e| e2 % Error Techni Err |e| e2 % Error
d que 1 or que 2 or
1 492 488 4 4 16 0.81% 495 -3 3 9 -0.61%
2 470 484 -14 14 196 -2.98% 482 -12 12 144 -2.55%
3 485 480 5 5 25 1.03% 478 7 7 49 1.44%
4 493 490 3 3 9 0.61% 488 5 5 25 1.01%
5 498 497 1 1 1 0.20% 492 6 6 36 1.20%
6 492 493 -1 1 1 -0.20% 493 -1 1 1 -0.20%
-2 28 248 -0.00529 2 34 264 0.00296
Using the figures shown in the table,
MAD1 = ∑ |e |/n = 28/6= 4.67
MAD2 = ∑ |e |/n = 34/6= 5.67
Technique 1 is superior in this comparison because its MAD is smaller, although six
observations would generally be too few on which to base a realistic comparison.
