Abstract
Deep neural network (DNN) has been widely applied in image recognition, sentiment analysis, natural language processing, and other fields. Because of its excellent classification capacity, many researchers have used it for credit scoring problems. However, compared to the traditional machine learning methods, DNN is often referred to as ‘black box’ because of the lack of interpretability. In this paper, a weighted soft decision forest (WSDF) model is proposed for credit scoring problem. WSDF tries to simulate the credit scoring mechanism under multiple expert decisions. In the new model, multiple soft decision trees are ensembled by an interpretable weighting mechanism to form a forest. The new weighting mechanism also uses a soft decision tree structure function to calculate the weight for each tree’s output. We test the performance of the new model on German Credit dataset. Compared with baseline algorithms, WSDF has good classification performance and interpretability.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsReferences
Sun J, Li H, Huang QH, He KY (2014) Predicting financial distress and corporate failure: a review from the state-of-the-art definitions, modeling, sampling, and featuring approaches. Knowl-Based Syst 57(2):41–56
Koh PW, Liang P (2017) Understanding black-box predictions via influence functions. arXiv preprint arXiv:1703.04730
Tong ENC, Mues C, Thomas LC (2012) Mixture cure models in credit scoring: if and when borrowers default. Eur J Oper Res 218(1):132–139
Abellán J, Mantas CJ (2014) Improving experimental studies about ensembles of classifiers for bankruptcy prediction and credit scoring. Expert Syst Appl 41(8):3825–3830
Finlay S (2011) Multiple classifier architectures and their application to credit risk assessment. Eur J Oper Res 210(2):368–378
Ribeiro MT, Singh S, Guestrin C (2016) “Why Should I Trust You?”: explaining the predictions of any classifier. In: Acm sigkdd International conference on knowledge discovery and data mining (KDD) 2016, pp 1135–1144
Frosst N, Hinton G (2017) Distilling a neural network into a soft decision tree. arXiv:1711.09784
Hinton G, Dean J, Vinyals O et al (2014) Distilling the knowledge in a neural network. In: Deep learning and representation learning workshop at NIPS 2014, NIPS. Palais des Congrès de Montréal, Montréal, Canada, pp 1–9
Friedman JH (2001) Greedy function approximation: a gradient boosting machine. Ann Stat 29(5):1189–1232
Zhao Z, Xu S, Kang BH, Kabir MMJ, Liu Y, Wasinger R (2015) Investigation and improvement of multi-layer perceptron neural networks for credit scoring. Expert Syst Appl 42(7):3508–3516
Dua D, Karra Taniskidou E (2017) UCI machine learning repository [https://archive.ics.uci.edu/ml]. University of California, School of Information and Computer Science, Irvine, CA
Acknowledgements
This research is partially supported by the National Natural Science Foundation of China (Grant No. 61562041).
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2020 Springer Nature Singapore Pte Ltd.
About this paper
Cite this paper
Zhang, Z., Liu, X., Gao, Z., Qu, Y. (2020). Interpretable Weighted Soft Decision Forest for Credit Scoring. In: Wang, TS., Ip, A., Tavana, M., Jain, V. (eds) Recent Trends in Decision Science and Management. Advances in Intelligent Systems and Computing, vol 1142. Springer, Singapore. https://doi.org/10.1007/978-981-15-3588-8_11
Download citation
DOI: https://doi.org/10.1007/978-981-15-3588-8_11
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-15-3587-1
Online ISBN: 978-981-15-3588-8
eBook Packages: Intelligent Technologies and RoboticsIntelligent Technologies and Robotics (R0)