Auto grouping option for predictors
If you want to set the best grouping options automatically, you can select the Auto grouping option. When this option is enabled, Prediction Studio attempts to automatically determine the best granularity and the minimum size of cases falling into each bin. When you select auto grouping, all other grouping parameters are disabled.
If you wait for the process to complete, you can clear the check box, and modify the parameters from the values determined by the algorithm. Auto grouping is a statistical algorithm, therefore subject to statistical flukes. On rare occasions this can cause problems, such as creating a grouping that is jagged or that generates too many bins. A jagged bin is the one which contains unusually large number of values in comparison to the other bins. It is created when the data set that is used to construct a sample contains elements that have a weight much larger than the rest of the elements. Because the auto grouping process can be lengthy for large data sets, this option is unselected by default.
For binary and continuous models, the algorithm will usually find a predictive and reliable grouping. However, it is advisable to check the results, particularly when the number of bins is large and adjust the parameters manually.
For extended binary models, auto grouping is not available in the preliminary data analysis because some behavior still needs to be inferred. Applying auto grouping at this stage could potentially cause important information to be lost. You can apply auto grouping in the final data analysis phase.
Previous topic Grouping options for predictors Next topic Managing grouping options for numeric predictors manually