Parameters | Predictive Modeling | Custom Prior Probabilities

Custom Prior Probabilities
Use this text field to specify the assumed prior probabilities for each of the classes.
Values must be entered using the following syntax:
level1=0.X level2=0.Y ... leveln=0.Z
where:
level1 through leveln are the names of all of the levels of the dependent variable,
0.X, 0.Y ... 0.Z, represent the prior probability of each level and are numbers between 0.0 and 1.0,
The prior probabilities should sum to one (1.0). If they do not, all values are rescaled such that they do sum to one (1.0).
To Specify the Custom Prior Probabilities:
*
Choose Automated as the Mode for generating the model.
*
Select Custom as the Prior Probabilities / Prevalences specification.
*
For example, if the class variable has three levels A, B, and C, with prior probabilities of 0.2, 0.2, and 0.6, respectively, then specify A=0.2 B=0.2 C=0.6, as shown below: