In financial risk, credit risk management is one of the most important issues in financial decision-making. Reliable credit scoring models are crucial for financial agencies to evaluate credit applications and have been widely studied in the field of machine learning and statistics. Deep learning is a powerful classification tool which is currently an active research area and successfully solves classification problems in many domains. Deep Learning provides training stability, generalization, and scalability with big data. Deep Learning is quickly becoming the algorithm of choice for the highest predictive accuracy. Feature selection is a process of selecting a subset of relevant features, which can decrease the dimensionality, reduce the running time, and improve the accuracy of classifiers. In this study, we constructed a credit scoring model based on deep learning and feature selection to evaluate the applicant's credit score from the applicant's input features. Two public datasets, Australia and German credit ones, have been used to test our method. The experimental results of the real world data showed that the proposed method results in a higher prediction rate than a baseline method for some certain datasets and also shows comparable and sometimes better performance than the feature selection methods widely used in credit scoring.

A Novel Credit Scoring Prediction Model based on Feature Selection Approach and Parallel Random Forest

Reliable credit scoring models played a very important role of retail banks to evaluate credit applications and it has been widely studied. The main objective of this paper is to build a hybrid credit scoring model using feature selection approach. In this study, we constructed a credit scoring model based on parallel GBM (Gradient Boosted Model), filter and wrapper approaches to evaluate the applicant's credit score from the input features. Feature scoring expression are combined by feature important (Gini index) and Information Value. Backward sequential scheme is used for selecting optimal subset of relevant features while the subset is evaluated by GBM classifier. To reduce the running time, we applied parallel GBM classifier to evaluate the proposed subset of features. The experimental results showed that the proposed method obtained a higher predictive accuracy than a baseline method for some certain datasets. It also showed faster speed and better generalization than traditional feature selection methods widely used in credit scoring.