Score distribution settings

You change these settings to affect the Score distribution step of the predictive model configuration process that is described in Checking score distribution. The settings include the names of the sections that are displayed for the step and the default values for particular options.

Setting	Description
Settings
Score band type	Select the type of bands to create: Bands with equal contents Creates the specified number of score bands, or bands with the specified percentage of cases in each band. The cases can be restricted to those with a specified outcome. Bands with equal behavior (statistically significant) Creates different score bands in terms of their specified behavior and statistical criteria in terms of the maximum probability that the difference between two bands is spurious, and that there is a minimum number of records in each band. Bands with monotonically increasing average Optimizes the score bands to produce the most useful gradient of bands. This option allows the Prediction Studio to construct the largest number of bands where the average probability of positive behavior of the average predicted values increases monotonically when the score increases. User defined bands Allows the use of the bands created by the Prediction Studio portal users.
Number of bands	Set the number of bands.
Maximum distinct symbolic intervals	Set the maximum number of intervals (bands) to create for symbolic inputs.
Max probability of spurious differences	Set the maximum difference between bands, expressed in likelihood. This value must be between 0 and 1.
Maximum displayed points	Set the maximum number of points for the display of the model comparison graphs.
Minimum size (% of the sample)	Set the minimum number of sample cases in each interval. This setting ensures that there is sufficient evidence of the behavior of cases in the interval for its behavior to be used in grouping. Intervals with few cases are combined with their nearest neighbor.
Invert probabilities	Select this option to have the probabilities of alternative behavior calculated.