Let's run a probit model. Only two values of the dependent are possible - defaulted and non-defaulted. We will use a link function to run a probit model:
Proc Genmod data=model_validation descending; Weight validation_sample; Model dflt = utilisation borrowing_portfolio_ratio postcode_index arrears_flag / dist=binomial link=probit; Output out=preds_link(where=(validation_sample=0)) p=pred l=lower u=upper; Run;
Figure 3.17 contains the model information where the probit link function is mentioned:
Figure 3.17: Model information probit link
The specified model does not contain two of the variables that were present in the ...
