B0239
Title: How fair is machine learning in credit scoring?
Authors: Golnoosh Babaei - University of Pavia (Italy) [presenting]
Paolo Giudici - University of Pavia (Italy)
Abstract: Machine learning (ML) algorithms, in credit scoring, are employed to distinguish between borrowers classified as class zero, including borrowers who will fully pay back the loan, and class one, borrowers who will default on their loan. However, in doing so, these algorithms are complex and often introduce discrimination by differentiating between individuals who share a protected attribute (such as gender and nationality) and the rest of the population. Therefore, to make users trust these methods, it is necessary to provide fair and explainable models. The focus is on fairness and explainability in credit scoring for solving this issue, using data from a P2P lending platform in the US. From a methodological viewpoint, ensemble tree models are combined with SHAP to achieve explainability, and the resulting Shapley values are compared with fairness metrics based on the confusion matrix.
