Abstract:Machine learning and artificial intelligence have achieved a human-level performance in many application domains, including image classification, speech recognition and machine translation. However, in the financial domain expert-based credit risk models have still been dominating. Establishing meaningful benchmark and comparisons on machine-learning approaches and human expert-based models is a prerequisite in further introducing novel methods. Therefore, our main goal in this study is to establish a new benchmark using real consumer data and to provide machine-learning approaches that can serve as a baseline on this benchmark. We performed an extensive comparison between the machine-learning approaches and a human expert-based model—FICO credit scoring system—by using a Survey of Consumer Finances (SCF) data. As the SCF data is non-synthetic and consists of a large number of real variables, we applied two variable-selection methods: the first method used hypothesis tests, correlation and random forest-based feature importance measures and the second method was only a random forest-based new approach (NAP), to select the best representative features for effective modelling and to compare them. We then built regression models based on various machine-learning algorithms ranging from logistic regression and support vector machines to an ensemble of gradient boosted trees and deep neural networks. Our results demonstrated that if lending institutions in the 2001s had used their own credit scoring model constructed by machine-learning methods explored in this study, their expected credit losses would have been lower, and they would be more sustainable. In addition, the deep neural networks and XGBoost algorithms trained on the subset selected by NAP achieve the highest area under the curve (AUC) and accuracy, respectively.

Comparison Research of Different Methods Applying to Corporate Credit Score

Research of Credit Score Based on Vfold Crossvalidation and Elman Neural Networks

Research of Corporate Credit Scoring Based on ANFIS and Elman Neural Networks

RESEARCH ON INDEX SYSTEM OF CREDIT EVALUATING OF ENTERPRISES AND EVALUATING METHODS

A Comparison Research of Soft Computing Applied to Credit Scoring

Corporate Credit Rating: A Survey

An application of pattern recognition on scoring Chinese corporations financial conditions based on backpropagation neural network

Comparative study of credit rating models for small and micro enterprises

Strengthen Credit Scoring System of Small and Micro Businesses with Soft Information: Analysis and Comparison Based on Neural Network Models

Research on Risk Assessment of Corporate Credit Based on Supply Chain

Application of Deep Neural Networks to assess corporate Credit Rating

Comparative Study On Bank Loan Credit Risk Assessment Methods Based On Data Mining

A Study on the Personal Credit Scoring Model Based on the Neural Network

A Study on Credit Risk Measurement of Chinese Commercial Banks

A Comparison Study on the Era of Internet Finance China Construction of Credit Scoring System Model

Comparative Study on the Character of Modern Credit Risk Models

A comparative study of forecasting corporate credit ratings using neural networks, support vector machines, and decision trees

Variable Selection in Credit Risk Models for Chinese Listed Companies

Levenberg-Marquardt-based Neural Network for Corporation Credit Rating

How Do Machine Learning and Non-Traditional Data Affect Credit Scoring? New Evidence from a Chinese Fintech Firm

An Empirical Comparison of Machine-Learning Methods on Bank Client Credit Assessments