Transforming Credit Risk Management with Machine Learning in the Financial Sector

Credit risk refers to the potential for financial loss if a borrower fails to meet their debt obligations. While various factors contribute to credit risk, conducting thorough evaluations prior to loan approval (such as credit scoring) and continuously tracking client payments and behaviors can mitigate the likelihood of fraud and non-performing assets (NPAs). The growing prevalence of NPAs and fraudulent activities in recent years has underscored the need for more precise methods to predict loan performance. With the rapid advancement of artificial intelligence (AI), particularly in machine learning (ML), driven by increased access to the internet, data, and computing power, researchers are exploring diverse ML techniques to enhance the accuracy of credit risk assessments. In this model, we applied Random Forest, Support Vector Machine, XGBoost, and Logistic Regression to achieve the most accurate results. While banks often rely on credit rating and scoring agencies for client profiles, current research emphasizes using machine learning to refine credit risk evaluation. This study reviews 15 to 20 credit risk assessment research papers published between 2012 and March 2023, highlighting emerging trends and identifying the limitations of existing approaches.