LOAN APPROVAL PREDICTION USING MACHINE LEARNING

Abstract

Loan approval plays a vital role in the financial sector as it determines whether an applicant qualifies for credit while maintaining institutional security. This study introduces a data-driven machine-learning framework designed to automate and improve the loan approval process. Using a dataset of 19,500 applications with 36 financial and demographic attributes, multiple supervised learning algorithms—Logistic Regression, Random Forest, Support Vector Classifier (SVC), and XGBoost—were trained and evaluated. After extensive preprocessing, correlation analysis, and feature scaling, the optimized SVC model achieved the best overall performance with 95.38% accuracy and an AUC of 0.94. To enhance transparency, Local Interpretable Model-Agnostic Explanations (LIME) were used to interpret individual predictions. The final model was deployed through an interactive Streamlit web interface, enabling users to obtain real-time loan approval results along with visual explanations. The proposed system demonstrates that integrating predictive modeling with interpretability and deployment can lead to a more transparent, reliable, and efficient decision-support tool for financial institutions.