In: Operations Management
After tallying the receipts for their first year of operation,
the owners of the Taco Barn are encouraged. Sales of their artisnal
tacos, made from such exotic ingredients as ground beef, cheese,
and beans, have been strong and seem to give hope to the coming
year. Taco sales by month are shown in the table.
Armed only with his fingers, the owner decides that the safest
forecasting approach is a linear trend line. Generate a forecast
for the year using this technique and then calculate forecast
errors using MSE. What is the mean squared error for this
forecasting approach?
1217
1033
1148
1282
Answer:
Information given in the question is:
Month |
Artisnal Taco Sales |
Month 1 |
18 |
Month 2 |
21 |
Month 3 |
15 |
Month 4 |
23 |
Using Linear Trend, Sales forecast is to be determined.
Input the data in EXCEL:
x |
Month |
Sales |
1 |
Month 1 |
18 |
2 |
Month 2 |
21 |
3 |
Month 3 |
15 |
4 |
Month 4 |
23 |
Step 1 |
Select the month and sales data |
Step 2 |
Click on Insert and then “Line” |
Step 3 |
On the line graph, do right click and select “Add Trendline” |
Step 4 |
Then from the menu, click on the “Display Equation on chart” checkbox. |
Following graph will be displayed:
Thus, equation for linear trend is:
y = 0.9x + 17
where,
y = Ft = Forecast
x = t = Month or time period number
Using formula y = 0.9x + 17
Forecast for month 1 = (0.9*1)+17 = 17.90
Forecast for month 2 = (0.9*2)+17 = 18.80
Forecast for month 3 = (0.9*3)+17 = 19.70
Forecast for month 4 = (0.9*4)+17 = 20.60
If forecast for more months are to be calculated, the same equation can be continued.
x |
Month |
Sales |
Forecast |
1 |
Month 1 |
18 |
17.90 |
2 |
Month 2 |
21 |
18.80 |
3 |
Month 3 |
15 |
19.70 |
4 |
Month 4 |
23 |
20.60 |
Calculation of MSE (Mean Squared Error):
Month |
Actual Sales |
Forecast |
Errors |
Absolute |
Absolute Errors2 |
At |
Ft |
et = At-Ft |
|et| |
|et|2 |
|
Month 1 |
18 |
17.00 |
1.00 |
1 |
1 |
Month 2 |
21 |
17.00 |
4.00 |
4 |
16 |
Month 3 |
15 |
17.00 |
-2.00 |
2 |
4 |
Month 4 |
23 |
17.00 |
6.00 |
6 |
36 |
MSE |
14.25 |
Error = Actual Sales – Forecast
Absolute Error = |Error| or ABS(Error)
Absolute Error2 = |Error|2
MSE = Mean Squared Error = Average of all Absolute Error2
= (1+16+4+36)/4
MSE = 14.25
Mean squared error estimates the average squared delta between the estimated forecast values and the actual sales value.
Mean squared error for this forecasting approach = 14.25