To configure a Model process

1.	In a flowchart in Edit mode, ensure that your Model process is connected to one or more configured processes whose output cells will be used as input by the Model process.

2.	Double-click the Model process in the flowchart workspace.

The process configuration dialog appears.

3.	On the Source tab, select your responder and non-responder cells from the Responder and Non-Responder drop-down lists. The drop-down lists are populated with input cells to the Model process; they are empty if you have no inputs to the process.

Both input cells must have the same audience level.

From the Variables to be Used for Modeling list, select the variables to be used by the Model process during model generation. You can select all variables (by clicking Use All) and let the Model process decide which set of inputs are most effective for modeling. However, by eliminating variables that do not add value to the model (for example, variables containing the same value for all records or different values for all records), you can speed up the automatic variable selection process.

If you are in doubt as to the predictive value of a variable, include it and let the Model process decide whether it should be used.

You can also use derived fields in the configuration of a model.

5.	On the Method tab, used the Best Model options to choose how the data mining algorithms should arrive at the best model: based on a specified length of time, or from among a specified number of candidate models.

Best Model In – (Default) Allows you to specify a time limit for modeling. The Model process retains the best models built in the time period that you specify. The default is three hours.

Best Models Among – Allows you to specify the number of models to build. The Model process will retain the best of these models. The default is 20.

6.	Select the type of modeling that will be done by choosing an algorithm from the Algorithm to Use drop-down list. You can select one, several, or all.

Selecting All Algorithms (the default) will result in a more accurate model, but it might take longer.

Set the maximum number of models to keep by entering any positive integer in the Max. # of Models to Keep field. The default is five. Each of the top models are saved by appending a pound sign (#) to the end of the specified file name. If more than one model is kept, an index is appended to the base file name, indicating the rank of the model.

The more models you elect to keep, the more disk space is required to save the corresponding information.

8.	In the Model File Name field, specify the absolute path name for the model file that is to be created, or click Browse to navigate to a file. Once you run the process it will generate a NAME.rtm model file, which you can use for scoring with the Score process.

9.	(Optional) Click the General tab to assign a name and/or notes to the process.

The name appears on the process in the flowchart. The notes appear when you mouse over the process in the flowchart.

10.

Click OK.

The process is configured and appears enabled in the flowchart. You can test the process to verify that it returns the results you expect.