IDEAS home Printed from https://ideas.repec.org/p/arx/papers/2009.13384.html
   My bibliography  Save this paper

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

Author

Listed:
  • Michael Bucker
  • Gero Szepannek
  • Alicja Gosiewska
  • Przemyslaw Biecek

Abstract

A major requirement for credit scoring models is to provide a maximally accurate risk prediction. Additionally, regulators demand these models to be transparent and auditable. Thus, in credit scoring, very simple predictive models such as logistic regression or decision trees are still widely used and the superior predictive power of modern machine learning algorithms cannot be fully leveraged. Significant potential is therefore missed, leading to higher reserves or more credit defaults. This paper works out different dimensions that have to be considered for making credit scoring models understandable and presents a framework for making ``black box'' machine learning models transparent, auditable and explainable. Following this framework, we present an overview of techniques, demonstrate how they can be applied in credit scoring and how results compare to the interpretability of score cards. A real world case study shows that a comparable degree of interpretability can be achieved while machine learning techniques keep their ability to improve predictive power.

Suggested Citation

  • Michael Bucker & Gero Szepannek & Alicja Gosiewska & Przemyslaw Biecek, 2020. "Transparency, Auditability and eXplainability of Machine Learning Models in Credit Scoring," Papers 2009.13384, arXiv.org.
  • Handle: RePEc:arx:papers:2009.13384
    as

    Download full text from publisher

    File URL: http://arxiv.org/pdf/2009.13384
    File Function: Latest version
    Download Restriction: no
    ---><---

    References listed on IDEAS

    as
    1. Fitzpatrick, Trevor & Mues, Christophe, 2016. "An empirical comparison of classification algorithms for mortgage default prediction: evidence from a distressed mortgage market," European Journal of Operational Research, Elsevier, vol. 249(2), pages 427-439.
    2. Wright, Marvin N. & Ziegler, Andreas, 2017. "ranger: A Fast Implementation of Random Forests for High Dimensional Data in C++ and R," Journal of Statistical Software, Foundation for Open Access Statistics, vol. 77(i01).
    3. Verbraken, Thomas & Bravo, Cristián & Weber, Richard & Baesens, Bart, 2014. "Development and application of consumer credit scoring models using profit-based classification measures," European Journal of Operational Research, Elsevier, vol. 238(2), pages 505-513.
    4. Banasik, John & Crook, Jonathan, 2007. "Reject inference, augmentation, and sample selection," European Journal of Operational Research, Elsevier, vol. 183(3), pages 1582-1594, December.
    5. Friedman, Jerome H. & Hastie, Trevor & Tibshirani, Rob, 2010. "Regularization Paths for Generalized Linear Models via Coordinate Descent," Journal of Statistical Software, Foundation for Open Access Statistics, vol. 33(i01).
    6. Matt J. Kusner & Joshua R. Loftus, 2020. "The long road to fairer algorithms," Nature, Nature, vol. 578(7793), pages 34-36, February.
    7. Lessmann, Stefan & Baesens, Bart & Seow, Hsin-Vonn & Thomas, Lyn C., 2015. "Benchmarking state-of-the-art classification algorithms for credit scoring: An update of research," European Journal of Operational Research, Elsevier, vol. 247(1), pages 124-136.
    8. Ellen Tobback & David Martens, 2019. "Retail credit scoring using fine‐grained payment data," Journal of the Royal Statistical Society Series A, Royal Statistical Society, vol. 182(4), pages 1227-1246, October.
    9. Crook, Jonathan N. & Edelman, David B. & Thomas, Lyn C., 2007. "Recent developments in consumer credit risk assessment," European Journal of Operational Research, Elsevier, vol. 183(3), pages 1447-1465, December.
    10. Bücker, Michael & van Kampen, Maarten & Krämer, Walter, 2013. "Reject inference in consumer credit scoring with nonignorable missing data," Journal of Banking & Finance, Elsevier, vol. 37(3), pages 1040-1045.
    Full references (including those not matched with items on IDEAS)

    Most related items

    These are the items that most often cite the same works as this one and are cited by the same works as this one.
    1. Fitzpatrick, Trevor & Mues, Christophe, 2021. "How can lenders prosper? Comparing machine learning approaches to identify profitable peer-to-peer loan investments," European Journal of Operational Research, Elsevier, vol. 294(2), pages 711-722.
    2. Kozodoi, Nikita & Jacob, Johannes & Lessmann, Stefan, 2022. "Fairness in credit scoring: Assessment, implementation and profit implications," European Journal of Operational Research, Elsevier, vol. 297(3), pages 1083-1094.
    3. Gero Szepannek, 2022. "An Overview on the Landscape of R Packages for Open Source Scorecard Modelling," Risks, MDPI, vol. 10(3), pages 1-33, March.
    4. Chen, Shunqin & Guo, Zhengfeng & Zhao, Xinlei, 2021. "Predicting mortgage early delinquency with machine learning methods," European Journal of Operational Research, Elsevier, vol. 290(1), pages 358-372.
    5. Rasa Kanapickiene & Renatas Spicas, 2019. "Credit Risk Assessment Model for Small and Micro-Enterprises: The Case of Lithuania," Risks, MDPI, vol. 7(2), pages 1-23, June.
    6. Kriebel, Johannes & Stitz, Lennart, 2022. "Credit default prediction from user-generated text in peer-to-peer lending using deep learning," European Journal of Operational Research, Elsevier, vol. 302(1), pages 309-323.
    7. Dumitrescu, Elena & Hué, Sullivan & Hurlin, Christophe & Tokpavi, Sessi, 2022. "Machine learning for credit scoring: Improving logistic regression with non-linear decision-tree effects," European Journal of Operational Research, Elsevier, vol. 297(3), pages 1178-1192.
    8. Elena Ivona DUMITRESCU & Sullivan HUE & Christophe HURLIN & Sessi TOKPAVI, 2020. "Machine Learning or Econometrics for Credit Scoring: Let’s Get the Best of Both Worlds," LEO Working Papers / DR LEO 2839, Orleans Economics Laboratory / Laboratoire d'Economie d'Orleans (LEO), University of Orleans.
    9. Zha, Yong & Wang, Yuting & Li, Quan & Yao, Wenying, 2022. "Credit offering strategy and dynamic pricing in the presence of consumer strategic behavior," European Journal of Operational Research, Elsevier, vol. 303(2), pages 753-766.
    10. Huei-Wen Teng & Michael Lee, 2019. "Estimation Procedures of Using Five Alternative Machine Learning Methods for Predicting Credit Card Default," Review of Pacific Basin Financial Markets and Policies (RPBFMP), World Scientific Publishing Co. Pte. Ltd., vol. 22(03), pages 1-27, September.
    11. Fitzpatrick, Trevor & Mues, Christophe, 2016. "An empirical comparison of classification algorithms for mortgage default prediction: evidence from a distressed mortgage market," European Journal of Operational Research, Elsevier, vol. 249(2), pages 427-439.
    12. Kolesnikova, A. & Yang, Y. & Lessmann, S. & Ma, T. & Sung, M.-C. & Johnson, J.E.V., 2019. "Can Deep Learning Predict Risky Retail Investors? A Case Study in Financial Risk Behavior Forecasting," IRTG 1792 Discussion Papers 2019-023, Humboldt University of Berlin, International Research Training Group 1792 "High Dimensional Nonstationary Time Series".
    13. Monir El Annas & Badreddine Benyacoub & Mohamed Ouzineb, 2023. "Semi-supervised adapted HMMs for P2P credit scoring systems with reject inference," Computational Statistics, Springer, vol. 38(1), pages 149-169, March.
    14. Nikita Kozodoi & Johannes Jacob & Stefan Lessmann, 2021. "Fairness in Credit Scoring: Assessment, Implementation and Profit Implications," Papers 2103.01907, arXiv.org, revised Jun 2022.
    15. Li, Yibei & Wang, Ximei & Djehiche, Boualem & Hu, Xiaoming, 2020. "Credit scoring by incorporating dynamic networked information," European Journal of Operational Research, Elsevier, vol. 286(3), pages 1103-1112.
    16. Doumpos, Michalis & Zopounidis, Constantin & Gounopoulos, Dimitrios & Platanakis, Emmanouil & Zhang, Wenke, 2023. "Operational research and artificial intelligence methods in banking," European Journal of Operational Research, Elsevier, vol. 306(1), pages 1-16.
    17. Gunnarsson, Björn Rafn & vanden Broucke, Seppe & Baesens, Bart & Óskarsdóttir, María & Lemahieu, Wilfried, 2021. "Deep learning for credit scoring: Do or don’t?," European Journal of Operational Research, Elsevier, vol. 295(1), pages 292-305.
    18. Giorgos Foutzopoulos & Nikolaos Pandis & Michail Tsagris, 2024. "Predicting Full Retirement Attainment of NBA Players," Working Papers 2403, University of Crete, Department of Economics.
    19. Van Belle, Jente & Guns, Tias & Verbeke, Wouter, 2021. "Using shared sell-through data to forecast wholesaler demand in multi-echelon supply chains," European Journal of Operational Research, Elsevier, vol. 288(2), pages 466-479.
    20. Philipp Bach & Victor Chernozhukov & Malte S. Kurz & Martin Spindler & Sven Klaassen, 2021. "DoubleML -- An Object-Oriented Implementation of Double Machine Learning in R," Papers 2103.09603, arXiv.org, revised Jun 2024.

    More about this item

    NEP fields

    This paper has been announced in the following NEP Reports:

    Statistics

    Access and download statistics

    Corrections

    All material on this site has been provided by the respective publishers and authors. You can help correct errors and omissions. When requesting a correction, please mention this item's handle: RePEc:arx:papers:2009.13384. See general information about how to correct material in RePEc.

    If you have authored this item and are not yet registered with RePEc, we encourage you to do it here. This allows to link your profile to this item. It also allows you to accept potential citations to this item that we are uncertain about.

    If CitEc recognized a bibliographic reference but did not link an item in RePEc to it, you can help with this form .

    If you know of missing items citing this one, you can help us creating those links by adding the relevant references in the same way as above, for each refering item. If you are a registered author of this item, you may also want to check the "citations" tab in your RePEc Author Service profile, as there may be some citations waiting for confirmation.

    For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: arXiv administrators (email available below). General contact details of provider: http://arxiv.org/ .

    Please note that corrections may take a couple of weeks to filter through the various RePEc services.

    IDEAS is a RePEc service. RePEc uses bibliographic data supplied by the respective publishers.