Investigating Attribute Assessment for Credit Granting on a Brazilian Retail Enterprise
In this article, we investigate which features are required to enhance a credit scoring model for a Brazilian retail enterprise. In order to find attributes that can improve the performance of classifier algorithms for credit granting, a national and an international survey were carried out. A logistic regression classifier was used and the main result has improved the performance of data mining classifiers. The main contribution of this article was the verification that additional financial and behavioral data increase defaulting prediction performance on credit granting.
KeywordsCredit granting Attribute assessment Classifier algorithms Logistic regression Receiver operating characteristic (ROC)
The authors would like to thank: the Brazilian Aeronautics Institute of Technology (ITA); the Casimiro Montenegro Filho Foundation (FCMF); the Software Engineering Research Group (GPES) members; and the 2RP Net Enterprise for their infrastructure, assistance, advice, data set, and financial support for this work.
- 3.M.C. Aniceto, Estudo comparativo entre técnicas de aprendizado de máquina para estimação de risco de crédito. Dissertação (Mestrado em Administração). Universidade de Brasília, Brasília, 2016Google Scholar
- 4.L. Delamaire, Implementing a credit risk management system based on innovative scoring techniques, Ph.D. thesis, University of Birmingham, 2012Google Scholar
- 7.M. Hörkkö, The determinants of default in consumer credit market. Masters thesis, Aalto University School of Economics (2010). Retrived from http://epub.lib.aalto.fi/en/ethesis/pdf/12299/hse_ethesis_12299.pdf
- 10.B. Waad, B.M. Ghazi, L. Mohamed, A three-stage feature selection using quadratic programming for credit scoring. Appl. Artif. Intell. Int. J. 27, 8 (2013)Google Scholar