The human-like intelligence with bio-inspired computing approach for credit ratings prediction

Abstract

Abstract Corporate credit rating analysis is a topic of considerable academic concern, and statistical and machine learning techniques for establishing credit ratings have been extensively studied. Recently developed hybrid models that combine different machine learning techniques have shown promising results. This research proposes a novel classification model based on bio-inspired computing mechanism to combine the artificial bee colony (ABC) approach and support vector machine (SVM) technique to enhance the credit ratings and credit rating changes prediction. Experiments are based on a ten-year real-world dataset taken from the Compustat credit rating database. Three additional financial datasets from the UCI datasets are also used to access the robust and convergence performance. The experimental outcomes indicate the proposed model provides improved prediction accuracy than other traditional statistical or soft-computing approaches, suggesting that the proposed approach is well-designed to potential credit rating or changes predicting.