In: Advanced Math
Given the following data, use exponential smoothing with a = 0.3 and α =.5 to develop a demand forecasts for period 7. Assume that the forecast for week 1= 19. Use the Mean Absolute Percent Error to determine which forecasts are more accurate.
Period |
1 |
2 |
3 |
4 |
5 |
6 |
Demand |
17 |
19 |
15 |
19 |
13 |
18 |
Given , data may be tabulated as ,
PERIOD | ACTUAL DEMAND (At) | FORECAST DEMAND (Ft) |
1 | 17 | 19 (given) |
2 | 19 | |
3 | 15 | |
4 | 19 | |
5 | 13 | |
6 | 18 | |
7 |
Now , since we have been given with
Therefore , we have ,
According to exponential smoothening method , the formula for calculating the forecast for period Ft+1 is given by -
Thus, we have ,
Now since we have , for period 1,
Therefore , for period 2 , we have ,
i.e., F2 = 18
Similarly , for period 3 , we have ,
For period 4 , we have ,
For period 5 , we have ,
For period 6 , we have ,
And , finally for period 7 , we have ,
THUS , THE ABOVE TABLE BECOMES -
PERIOD | ACTUAL DEMAND (At) | FORECAST DEMAND (Ft) |
1 | 17 | 19 (given) |
2 | 19 | 18 |
3 | 15 | 18.5 |
4 | 19 | 16.75 |
5 | 13 | 17.875 |
6 | 18 | 15.4375 |
7 | 16.71875 |
thus , the forecast demand for period 7 = 16.71875
Now , we calculate Mean Absolute Percent Error , which is the mean of the percent of the error with respect to actual demand and is calculated as ,
where , |error| = | actual demand - forecast demand|
So , we have the following table,
PERIOD | ACTUAL DEMAND (At) | FORECAST DEMAND (Ft) | ERROR | Absolute percent error | |
1 | 17 | 19 (given) | -2 | 2 | 11.76% |
2 | 19 | 18 | 1 | 1 | 5.26% |
3 | 15 | 18.5 | -3.5 | 3.5 | 23.33% |
4 | 19 | 16.75 | -3.75 | 3.75 | 19.73% |
5 | 13 | 17.875 | -4.875 | 4.875 | 37.5% |
6 | 18 | 15.4375 | -3.4375 | 3.4375 | 19.09% |
7 | 16.71875 |
Therefore , Mean Absolute Percent Error =
= 19.445 %
i.e., MAPE = 19.445 %