A Hybrid Bi-level Metaheuristic for Credit Scoring
- PDF / 962,213 Bytes
- 11 Pages / 595.276 x 790.866 pts Page_size
- 39 Downloads / 253 Views
A Hybrid Bi-level Metaheuristic for Credit Scoring Doruk Şen 1,2
&
Cem Çağrı Dönmez 3 & Umman Mahir Yıldırım 2
# Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract This research aims to propose a framework for evaluating credit applications by assigning a binary score to the applicant. The score is targeted to determine whether the credit application is ‘good’ or ‘bad’ in small business purpose loans. Even tiny performance improvements in small businesses may yield a positive impact on the economy as they generate more than 60% of the value. The method presented in this paper hybridizes the Genetic Algorithm (GA) and the Support Vector Machine (SVM) in a bi-level feeding mechanism for increased prediction accuracy. The first level is to determine the parameters of SVM and the second is to find a feature set that increases classification accuracy. To test the proposed approach, we have investigated three different data sets; UCI Australian data set for preliminary works, Lending Club data set for large training and testing, and UCI German and Australian datasets for benchmarking against some other notable methods that use GA. Our computational results show that our proposed method using a feedback mechanism under the hybrid bi-level GA-SVM structure outperforms other classification algorithms in the literature, namely Decision Tree, Random Forests, Logistic Regression, SVM and Artificial Neural Networks, effectively improves the classification accuracy. Keywords Support vector machine . Genetic algorithm . Credit scoring . Classification . Feature selection
1 Introduction Credit scoring mechanisms form the backbone of the finance sector or other credit-issuing authorities. A solid model should be able to differentiate the applicants to default and nondefault categories in terms of being able to meet or fail the obligations that are offered in the credit applications. Default non-default segregation is not always necessary as it is introduced in the work of Hand and Henley (1997), designing good and bad classes for risk assessment is also acceptable in the literature. The approach automatically converts the case into a binary classification problem as the final decision is the ability to separate good and bad cases from each other. The
* Doruk Şen [email protected] 1
Department of Engineering Management, Institute of Pure and Applied Sciences, Marmara University, 34722 Kadikoy, Istanbul, Turkey
2
Department of Industrial Engineering, Istanbul Bilgi University, 34060 Eyupsultan, Istanbul, Turkey
3
Department of Industrial Engineering, Marmara University, 34722 Kadikoy, Istanbul, Turkey
increased efficiency on the model will yield reductions on the monetary loss for the institutions and authorities. During the final years of the 1930s, the US National Bureau of Economic Research (NBER) carried a collection of studies on the concept of instalment financing. These studies were issued during the first couple of years of the 1940s. Among them, Chapman (1940) introduced that the loan
Data Loading...