A Comparative Analysis of Individual and Ensemble Credit Scoring Techniques in Evaluating Credit Card Loan Applications

One of the main tasks of a bank is to lend money. As a financial intermediary, one of its roles is to reduce lending risks. Bank lending is an art as well as a science. Success depends on techniques used, knowledge and on an aptitude to assess both credit-worthiness of a potential borrower and the merits of the proposition to be financed. In recent years, banks have increasingly used credit-scoring techniques to evaluate the loan applications they receive from consumers. Credit-scoring techniques are usually based on discriminant models or related techniques, such as logit or probit models or neural networks, in which several variables are used jointly to set up a numerical score for each loan applicant. This study explores the performance of both individual models by using neural networks, and classification and regression trees and ensemble models by using Bagging and Adaboost techniques. Experimental studies using real world data sets have demonstrated that the ensemble models outperform the other credit scoring models.