Transparency, Auditability and eXplainability of Machine Learning Models in Credit Scoring

arXiv:2009.13384 · stat.ML, cs.LG, econ.GN, q-fin.EC, stat.AP, stat.ME · Submitted 2020-09-28 · Read on arXiv

stat.ML, cs.LG, econ.GN, q-fin.EC, stat.AP, stat.ME

Submitted: 2020-09-28

Updated: 2020-09-28

Journal ref: J. Oper. Res. Soc. 73(1), 70-90 (2022)

DOI: 10.1080/01605682.2021.1922098

Code: https://github.com/agosiewska/fico-experiments

Project page: https://christophm.github.io/interpretable-ml-book

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Interpretability — 2× in this paper · explained in 15 episodes
• Explainability — 1× in this paper · explained in 3 episodes

Sources

• Do Not Trust Additive Explanations

Related papers

• Behavior of prediction performance metrics with rare events
• Optimal Estimation of Generic Dynamics by Path-Dependent Neural Jump ODEs
• A Posterior-Dynamics Framework for Imaging Inverse Problems with Pretrained Diffusion Priors
• Efficient quadratic entropy with distance sketches
• Embedding-Bias in Conditional Independence Testing
• From Log-Odds to Shapley Values: An Explanatory Geometry for the Weighted Naive Bayes Classifier