Comparative Analysis of Decision Tree, Random Forest and XGBoost for Loan Approval Prediction Using Supervised Machine Learning

This paper presents a comparative analysis of three supervised machine learning algorithms — Decision Tree, Random Forest, and XGBoost — for predicting loan approval status using structured applicant data. The study implements a complete end-to-end machine learning pipeline, including data preprocessing, categorical feature encoding, and a from-scratch Information Gain / Entropy-based feature selection framework built without external machine learning libraries. Each classifier is systematically tuned through hyperparameter analysis of training and validation accuracy, with XGBoost further optimized via grid search and early stopping. The final model is deployed as an interactive Streamlit web application for real-time loan approval prediction. The findings highlight the trade-offs between model interpretability and predictive performance, and demonstrate that increased algorithmic complexity does not always yield proportionally improved results on structured tabular datasets.