Credit risk analysis using artificial intelligence: evidence from a leading South African banking institution (original) (raw)

The study investigates the application of artificial neural networks (ANNs) for consumer credit risk analysis within a leading South African banking institution. It aims to assess whether ANN models can enhance prediction accuracy of consumer defaults compared to traditional credit scoring methods. By incorporating demographic, biographic, and financial data, the research evaluates various neural network architectures, concluding that a specific ANN model significantly improves the recognition of credit risk compared to existing systems.