Optimal Binning Startup Panel

Ribbon bar. Select the Data Mining tab. In the Clustering/Grouping group, click Optimal Binning to display the Optimal Binning Startup Panel.

The Startup Panel contains two tabs: Quick and Advanced. See also, Optimal Binning for Predictive Data Mining Introductory Overview and Optimal Binning for Predictive Data Mining Program Overview.

Naming conventions for recoded variables

When you click the Summary button (see option description below), the program will automatically determine a best way to combine the classes in each categorical predictor to yield strong relationships to the dependent or outcome variable specified for the analysis. The recoded or aggregated class codes will automatically be placed into the input data file, into the variables designated for output. Those newly recoded variables will be named according to the following conventions:

Variable names

The variable names will be created from the respective categorical predictor variables (from which they were computed), followed by the further clarification (Grouped). For example, a variable computed from aggregated SIC Codes would be named SIC Codes(Grouped.

Code text labels

The code labels will be created as Group1(k1), Group2(k2), .., Groupn(kn), where

• the numbers following the common prefix Group designate the respective simple ordinal counts (first group, second group, etc.)
•  the numbers in parentheses indicate the number of classes from the original categorical input variable that were combined into the respective group.

  For example, Group4(53) would mean that the fourth group was created as a combination of 53 classes or categories from the original categorical predictor variable.

Code descriptions

In addition, for each newly created code, a long code description is created, with all codes aggregated into the respective category. Those codes can be reviewed in the Text Labels Editor. All information regarding the recoding of class variables is recorded in the respective newly created class codes. For additional details regarding text values in Statistica, see also Using the Text Labels Editor and Notes on Text Labels and Text Values.