Interpretable Selective Learning in Credit Risk
This work addresses the challenge for financial regulators and practitioners in adopting more accurate machine learning methods in credit risk assessment by providing a selective approach that balances interpretability and performance, though it is incremental as it builds on existing methods.
The paper tackles the trade-off between accuracy and interpretability in credit default risk forecasting by introducing a neural network with a selective option that determines when to use linear models versus shallow neural networks, finding that logistic regression suffices for most datasets while neural networks improve accuracy for specific data portions without significantly compromising interpretability.
The forecasting of the credit default risk has been an important research field for several decades. Traditionally, logistic regression has been widely recognized as a solution due to its accuracy and interpretability. As a recent trend, researchers tend to use more complex and advanced machine learning methods to improve the accuracy of the prediction. Although certain non-linear machine learning methods have better predictive power, they are often considered to lack interpretability by financial regulators. Thus, they have not been widely applied in credit risk assessment. We introduce a neural network with the selective option to increase interpretability by distinguishing whether the datasets can be explained by the linear models or not. We find that, for most of the datasets, logistic regression will be sufficient, with reasonable accuracy; meanwhile, for some specific data portions, a shallow neural network model leads to much better accuracy without significantly sacrificing the interpretability.