<![CDATA[This study aims to design and develop a web-based Decision Support System using the Random Forest algorithm to classify credit feasibility in the context of Pegadaian in Bekasi City into Feasible and Not Feasible categories. A quantitative approach with an applied experiment was used on a simulated dataset of 1,501 records. The dataset consisted of seven features: income, occupation, age, collateral type, collateral value, loan amount, and loan term. The data were split into 1,200 training records and 301 testing records using an 80:20 ratio. The model was developed using Python and scikit-learn with 100 decision trees, then integrated into a web application using PHP, MySQL, and a Flask service. Evaluation used a confusion matrix, accuracy, precision, recall, F1-score, AUC-ROC, and feature importance. The results obtained 84.4% accuracy, 85.7% precision, 92.8% recall, 89.1% F1-score, and 88.0% AUC-ROC. Income had the highest relative contribution to the model predictions. The results indicate that Random Forest can be implemented as a data-driven tool for credit feasibility classification. However, because the study used a simulated dataset, final decisions still require verification and officer consideration.]]>
Copyrights © 2026
