Credit risk modelling using deep neural network| International Journal of Innovative Science and Research Technology

Credit Risk Modelling Using Deep Neural Network

Authors : Vaidehi Patil; Shubham Bari; Kairav Rajgariah; Sumit Singh

Volume/Issue : Volume 10 - 2025, Issue 11 - November

Google Scholar : https://tinyurl.com/mut8dz2f

Scribd : https://tinyurl.com/2mj7cwad

DOI : https://doi.org/10.38124/ijisrt/25nov286

PlumX Metrics

Semantic Scholar

ResearchGate

Note : A published paper may take 4-5 working days from the publication date to appear in PlumX Metrics, Semantic Scholar, and ResearchGate.

Note : Google Scholar may take 30 to 40 days to display the article.

Abstract : Credit risk modelling is essential for predicting a borrower’s likelihood of default and ensuring financial stability. Traditional models like Logistic Regression often struggle with complex, non-linear financial data. This paper presents a Machine Learning–based approach using Deep Neural Networks (DNNs) to enhance prediction accuracy and adaptability. The proposed model is trained on publicly available credit datasets after preprocessing and feature engineering. Experimental results show an accuracy of ~85% and an AUC of ~90%, outperforming classical models by a significant margin. The study demonstrates that deep learning can effectively capture hidden patterns in borrower behaviour, offering a robust, scalable, and data-driven framework for modern credit risk assessment.

Keywords : DNN (Deep Neural Networks), ROC AUC, Credit Default, Credit Risk.

References :

S. Lessmann, B. Baesens, H. V. Seow, and L. C. Thomas, “Benchmarking state-of-the-art classification algorithms for credit scoring,” European Journal of Operational Research, vol. 247, no. 1, pp. 124–136, 2015.
C. Serrano-Cinca and B. Gutiérrez-Nieto, “The use of profit scoring as an alternative to credit scoring systems in peer-to-peer (P2P) lending,” Decision Support Systems, vol. 89, pp. 113–122, 2016.
D. Zhang, et al., “Deep learning for credit risk prediction: A comparative study,” Expert Systems with Applications, vol. 176, 114921, 2021.
Kaggle Credit Risk Dataset. [Online]. Available: https://www.kaggle.com/dataset

Credit risk modelling is essential for predicting a borrower’s likelihood of default and ensuring financial stability. Traditional models like Logistic Regression often struggle with complex, non-linear financial data. This paper presents a Machine Learning–based approach using Deep Neural Networks (DNNs) to enhance prediction accuracy and adaptability. The proposed model is trained on publicly available credit datasets after preprocessing and feature engineering. Experimental results show an accuracy of ~85% and an AUC of ~90%, outperforming classical models by a significant margin. The study demonstrates that deep learning can effectively capture hidden patterns in borrower behaviour, offering a robust, scalable, and data-driven framework for modern credit risk assessment.

Keywords : DNN (Deep Neural Networks), ROC AUC, Credit Default, Credit Risk.