Paper Title
Discriminant Analysis Of Default Behaviors Of Gender Differences Borrowers

Abstract
As discriminant analysis cannot process dummy variables and continuous variables at the same time, this paper separates male and female borrowers in the data. According to the overall discriminant analysis of male borrowers, among the 8,747 data entries, 76.6% of the default borrowers can be correctly classified by the discriminant functions, and the result is significant. In the cross reliability analysis, 76.3% of the default borrowers and 43.2% of the normal ones can be correctly classified by the discriminant functions. As far as the cross reliability analysis is concerned, 43.1% of the normal borrowers can be correctly classified by the discriminant functions. Therefore, the overall accuracy of the discriminant analysis is 49.2%, and that of the cross reliability analysis is 49.0%. This indicates that as far as the overall data of male borrowers is concerned, the discriminant analysis boasts significant judgment ability over the default male borrowers. Key word - Discriminant Analysis, Default Behaviors, Gender Differences