Data Mining - Professional SQL Server Analysis Services 2005 with MDX - page 751

results. Figure 14-18 shows the Mining Model Prediction view. Similar to the
Mining Accuracy Chart view, you need to specify the case table that contains
the input data for which you want to predict. Select the source table vTar-
getMail for which you will now predict the bike buyer value. In this view you
can now select the columns from the input table that you want to be retrieved
along with the predicted value(s). You can also apply certain data mining pre-
diction functions or custom expressions on the predicted values to retrieve
results such as top 10% of the customers or top 10 customers. The columns
from the input table or applying certain data mining prediction functions can
be selected in the lower half of the window pane as shown in Figure 14-18 .
Select the predicted column bike buyer from the model and the columns Cus-
tomer key, first name, last name, and email address from the input table. You
can select additional columns if you need to.
Figure 14-18
To see the prediction results, click the button that allows you to switch
between the design or query or result view as shown in Figure 14-19 . You will
now see the results of the prediction as shown in Figure 14-19 . You can spe-
cify constraints to the predicted value such as predicted value = 1 so that the
results view only shows customers who are likely to buy bikes.