ML-Based Loan repayment default prediction system for improving credit risk assessment in the banking industry
Loading...
Date
Journal Title
Journal ISSN
Volume Title
Publisher
NSBM Green University
Abstract
This research aims to address the critical challenge of loan default prediction within the banking sector, aiming to enhance the accuracy of credit risk assessments. The lending process is one of the most important parts of the banking sector. Current systems in the banks struggle with rapidly changing data, leading to lower accuracy in real-world applications. Also, the current system does not use Machine Learning algorithms to predict the data. To overcome these limitations, this study employs a machine learning approach to develop a loan default prediction system, specifically focusing on the main two types of loans like Personal and Housing loans. A loan default is identified based on overdue days exceeding 90 days, as per the bank's system. So, the primary objectives are to analyze the machine Learning algorithms and develop predictive models for each loan type using machine learning algorithms and to design a user-friendly data product for predicting loan default status. However, this study holds substantial importance for banking institutions by improving their ability to predict loan defaults, providing a more detailed and precise evaluation of each loan's probability of default by integrating financial borrower behavior, such as average savings account
balances. Also, this system can predict the borrower’s repayment pattern and predict the loan default of the borrower. So, the predictive model aims to enable banks to make informed lending decisions, thereby reducing high-risk loans and strengthening overall financial stability.
Description
Keywords
Citation
Wanniachchi, B.L.K.; Wijesinghe, D.; Shafraz, M. (2025) ML-Based Loan repayment default prediction system for improving credit risk assessment in the banking industry, Technology Innovations & Digitalization Accompanying Creativity (TIDAC), NSBM Green University, Sri Lanka. P.329