Hué Sullivan, Hurlin Christophe, Pérignon Christophe, Saurin Sébastien
arXiv 12 Dec 2022 · Statistics — Machine Learning · publishedManagement Science (2026) · 3 citations (OpenAlex)
arXiv:2212.05866 · PDF · DOI · OpenAlex · Extracted main text
As they play an increasingly important role in determining access to credit, credit scoring models are under growing scrutiny from banking supervisors and internal model validators. These authorities need to monitor the model performance and identify its key drivers. To facilitate this, we introduce the XPER methodology to decompose a performance metric (e.g., AUC, $R^2$) into specific contributions associated with the various features of a forecasting model. XPER is theoretically grounded on Shapley values and is both model-agnostic and performance metric-agnostic. Furthermore, it can be implemented either at the model level or at the individual level. Using a novel dataset of car loans, we decompose the AUC of a machine-learning model trained to forecast the default probability of loan applicants. We show that a small number of features can explain a surprisingly large part of the model performance. Notably, the features that contribute the most to the predictive performance of the model may not be the ones that contribute the most to individual forecasts (SHAP). Finally, we show how XPER can be used to deal with heterogeneity issues and improve performance.
appendix boundary found by appendix_command · 37% of the source is main text. Read the extracted text to check this.
The works this paper leans on most, across its whole bibliography — not restricted to papers in our corpus. Ranked by composite intensity, which combines how often a work is mentioned, how many sections mention it, and how much of that falls in the main text rather than the appendix.
| Reference | Intensity | Mentions | Sections | Main text | |
|---|---|---|---|---|---|
| 1 | Lundberg, S. M. and Lee, S.-I (2017) A unified approach to interpreting model predictions | 0.855 | 8 | 7 | 62% |
| 2 | Casalicchio, G., Molnar, C., and Bischl, B (2019) Visualizing the feature importance for black box models | 0.843 | 5 | 3 | 60% |
| 3 | Israeli, O (2007) A Shapley-based decomposition of the R-square of a linear regression | 0.843 | 5 | 3 | 60% |
| 4 | Shapley, L (1953) A value for n-person games | 0.811 | 4 | 2 | 100% |
| 5 | Lundberg, S. M., Erion, G. G., and Lee, S.-I (2018) Consistent individualized feature attribution for tree ensembles | 0.737 | 3 | 3 | 67% |
| 6 | Bowen, D. and Ungar, L (2020) Generalized shap: Generating multiple types of explanations in machine learning | 0.644 | 2 | 2 | 100% |
| 7 | Sundararajan, M., Taly, A., and Yan, Q (2017) Axiomatic attribution for deep networks | 0.644 | 2 | 2 | 100% |
| 8 | Sundararajan, M. and Najmi, A (2020) The many Shapley values for model explanation | 0.644 | 2 | 2 | 100% |
| 9 | Sundararajan, M., Dhamdhere, K., and Agarwal, A (2020) The Shapley Taylor interaction index | 0.644 | 2 | 2 | 100% |
| 10 | Park, H.-S. and Jun, C.-H (2009) A simple and fast algorithm for k-medoids clustering | 0.644 | 2 | 2 | 100% |
Showing the top 10 of 45 scored citations.