TRUSTNOVA: A Predictive Machine Learning Model for Loan Approval

Abstract

Small low-income banks have problems in making accurate decisions in the approval of loans for low-income earners. Manual checking tends to make there be inefficiency bias and loss of opportunities. This study designs a Loan Approval Prediction System empowered with machine learning that automates the process and optimizes the processes involved in the assessment of eligibility for a loan. Data analysis on applicant details — income, credit history, and repayment capacity — would bring fairer, faster, and more accurate results in the lending decisions for banks. The model provides banks with the necessary conditions for making lending decisions that are fair and efficient. This will in the long run improve the financial inclusion of the unbanked and underbanked populations.

Introduction

Loan approval is crucial for financial inclusion, especially for banks serving low-income and microfinance clients. Traditional manual reviews are slow and prone to bias, whereas machine learning (ML) models can improve fairness, accuracy, and operational efficiency in credit decisions. ML models predict loan approval probabilities by analyzing applicant data, including non-traditional indicators like rent payments and mobile phone usage, which help assess creditworthiness for those lacking formal credit histories.

Common ML algorithms used in loan prediction include Logistic Regression, Decision Trees, Random Forest, Support Vector Machines, and XGBoost. These methods achieve high accuracy (often above 80%) and help reduce errors and manual workload. Explainable AI techniques like SHAP and LIME further enhance trust in automated decisions by clarifying model outputs.

The proposed system involves collecting and preprocessing relevant data, including traditional and alternative credit indicators, followed by feature selection and training several ML models. A stacked ensemble combining Random Forest, XGBoost, and Logistic Regression as a meta-learner is used to improve prediction accuracy. Performance is evaluated using metrics such as accuracy, precision, recall, F1-score, and ROC-AUC.

The system includes a user-friendly interface deployed with Streamlit, enabling real-time loan approval decisions with preliminary eligibility checks to improve efficiency. This approach aims to increase loan approval success, reduce defaults, and expand credit access to underserved populations while maintaining fairness and transparency.

Conclusion

The loan approval system, designed using Logistic Regression and Random Forest, provides a well-rounded and effective approach to predicting loan approval decisions. Logistic Regression, known for its interpretability and efficiency in binary classification problems, performs well when the relationship between input features and the target variable is linear. However, when dealing with complex, non-linear patterns in the data, Random Forest—a robust ensemble learning technique—enhances predictive accuracy by aggregating multiple decision trees and minimizing overfitting. By incorporating both models, the system strikes a balance between clarity and performance, ensuring dependable and precise predictions. This dual-model approach strengthens the decision-making process, fostering greater fairness and transparency in loan approvals. Additionally, it equips financial institutions with a reliable tool to assess applications more effectively. Looking ahead, the system can be further improved by integrating sentiment analysis to evaluate applicant interactions and implementing fraud detection mechanisms to identify suspicious applications. Furthermore, incorporating incremental model updates will help adapt to evolving market trends and borrower behaviors, maintaining the system’s relevance over time. Future enhancements could also include federated learning, enabling banks to collaboratively train models on shared insights while ensuring data privacy and security remain intact.

References

[1] R, S., L, V., B, S., &Manikandan, M. (2022). Bank Loan Approval Prediction Using Data Science Technique (ML). International Journal for Research in Applied Science and Engineering Technology. https://doi.org/10.22214/ijraset.2022.43665. [2] Shinde, A. (2022). Intelligent Loan Assistant using Machine Learning and Data Science. INTERANTIONAL JOURNAL OF SCIENTIFIC RESEARCH IN ENGINEERING AND MANAGEMENT. https://doi.org/10.55041/0522.12643. [3] Rahman, A., Purno, M., &Mim, S. (2023). Prediction of the Approval of Bank Loans Using Various Machine Learning Algorithms. 2023 IEEE World Conference on Applied Intelligence and Computing (AIC), 272-277.https://doi.org/10.1109/AIC57670.2023.10263880. [4] Diwate, Y., Rana, P., &Chavan, P. (2023). LOAN APPROVAL PREDICTION USING MACHINE LEARNING. International Research Journal of Modernization in Engineering Technology and Science. https://doi.org/10.56726/irjmets39658. [5] Aphale, A., &Shinde, S. (2020). Predict Loan Approval in Banking System Machine Learning Approach for Cooperative Banks Loan Approval. International journal of engineering research and technology, 9. [6] Orji, U., Ugwuishiwu, C., Nguemaleu, J., &Ugwuanyi, P. (2022). Machine Learning Models for Predicting Bank Loan Eligibility. 2022 IEEE Nigeria 4th International Conference on Disruptive Technologies for Sustainable Development (NIGERCON), 1-5.https://doi.org/10.1109/nigercon54645.2022.980317. [7] Singh, V., Yadav, A., Awasthi, R., &Partheeban, G. (2021). Prediction of Modernized Loan Approval System Based on Machine Learning Approach. 2021 International Conference on Intelligent Technologies (CONIT), 1-4.https://doi.org/10.1109/CONIT51480.2021.9498475 [8] Kokru, J., Ghodke, A., Chavan, P., Chand, S., & Mane, P. (2022). Bank Loan Approval Prediction System Using Machine Learning Algorithms. International Journal of Advanced Research Science, Communication and Technology. https://doi.org/10.48175/ijarsct-2637 [9] Karthiban, R., Ambika, M., &Kannammal, K. (2019).A Review on Machine Learning Classification Technique for Bank Loan Approval. 2019 International Conference on Computer Communication and Informatics (ICCCI), 1-6. https://doi.org/10.1109/ICCCI.2019.8822014. [10] N, P. D., Shetty, C., N, R., B, D., & , P. (2022). Predictive Analysis of Loan Data using Machine Learning. 2022 International Conference on Artificial Intelligence and Data Engineering (AIDE), 272-276.https://doi.org/10.1109/AIDE57180.2022.10060781. [11] Tumuluru, P., Burra, L., Loukya, M., Bhavana, S., CSaiBaba, H., &Sunanda, N. (2022). Comparative Analysis of Customer Loan Approval Prediction using Machine Learning Algorithms. 2022 Second International Conference on Artificial Intelligence and Smart Energy (ICAIS), 349-353. https://doi.org/10.1109/ICAIS53314.2022.9742800. [12] Sharma, V., & Sharma, R. (2022). A Systematic Survey of Automatic Loan Approval System Based on Machine Learning. Int. J. Secur. Priv. Pervasive Comput., 14, 1-25. https://doi.org/10.4018/ijsppc.304893. [13] Mamun, M., Farjana, A., &Mamun, M. (2022). Predicting Bank Loan Eligibility Using Machine Learning Models and Comparison Analysis. Proceedings of the International Conference on Industrial Engineering and Operations Management. https://doi.org/10.46254/na07.20220328. [14] Saini, P., Bhatnagar, A., & Rani, L. (2023). Loan Approval Prediction using Machine Learning: A Comparative Analysis of Classification Algorithms. 2023 3rd International Conference on Advance Computing and Innovative Technologies in Engineering (ICACITE), 1821-1826.https://doi.org/10.1109/icacite57410.2023.10182799 [15] Shinde, A., Patil, Y., Kotian, I., Shinde, A., &Gulwani, R. (2022). Loan Prediction System Using Machine Learning. ITM Web of Conferences. https://doi.org/10.1051/itmconf/20224403019.

Copyright

Copyright © 2025 Krutika Kolpate, Antariksha Dhanure, Gargi Fadtare, Gayatri Gunjal, Dr. Nilesh Bhelkar. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.