Doctoral thesis of Philosophy: Investigation of artificial immune systems and variable selection techniques for credit scoring

Two major areas of the scorecard development process were identified as having much capacity for improving credit scoring models. These are the application of 1) better classifiers and 2) variable selection techniques. However, the application of an artificial immune system for developing scorecards as well as the use of other variable selection techniques, other than stepwise regression, in credit scoring are quite limited as indicated by the scarcity of published materials in these two areas in the current literature.