Testing the forecast model
We will see the forecast fit using the car sales from 2016 to 2017. This data is known from the dataset, so we will be able to check whether the model works well or not.
Follow these steps to test the model:
- Copy and paste from the previous data the seasonal trend for each quarter from 2016 to 2017.
- Calculate the trend for 2016 to 2017 using the regression line.
- Test the forecast by multiplying the seasonal trend by the regression.
Use the known data of sales for the years 2016 and 2017 to test the forecast model, as mentioned in step 3 of the previous list. The automobile sales are in the highlighted QuarterTot column in Figure 13.7. We use this data to see whether the model is useful to predict this information.
The season trend per quarterly period for the years 2016 and 2017 is the same past data we have seen before. Calculate the regression trend for the 2016–2017 quarters. This data is highlighted in Figure 13...