Neural networks in credit scoring are described as which of the following?

• A. Neural networks are a credit scoring method used for commercial lending that is more heterogeneous than consumer lending
• B. They use artificial intelligence
• C. They learn from experience
• D. They are newer technique criticized as being 'ad hoc'

Answer: (B)

Neural networks in credit scoring are best described as an artificial intelligence approach. They are a kind of AI model that learns patterns from data by adjusting internal weights during training, which lets them capture complex, non-linear relationships among credit factors. This learning process—training on examples and improving with experience—is a hallmark of neural networks, but the defining title that fits across contexts is that they are an AI method. While they do learn from experience, and can be newer methods for some practitioners, those points are secondary to their identity as AI-based models. The other descriptions either miss the fundamental AI nature or mischaracterize what neural networks are, so labeling them as using artificial intelligence is the most precise choice.