KNORA-IU: improving the dynamic selection prediction in imbalanced credit scoring problems

Credit scoring has become a critical tool to discriminate 'bad' applicants from 'good' ones for financial institutions. One common characteristic of the credit dataset is the imbalance between good and bad applicants, with low defaults (no paid loans). Ensemble classification methodology is widely used in this field. However, dynamic ensemble selection approaches to imbalanced datasets have drawn little consideration. This study aims to adapt KNORA-Union, an excellent dynamic selection technique, to imbalanced credit scoring problem, the KNORAImbalanced Union (KNORA-IU). In this approach, we propose a new procedure to evaluate the competence of each base classifier. The results, based on four performance measures, indicate that the performance of the KNORA-IU is superior to the state-of-the-art approaches for moderate imbalanced datasets.