Analisis Kelayakan Kredit Menggunakan Decision Tree C4.5 pada PT. XYZ

Abstract

PT. XYZ is a private bank operating in Southwest Papua Province and is supervised by the Financial Services Authority (OJK) of Papua and West Papua. This study aims to analyze creditworthiness using the Decision Tree C4.5 algorithm to support more accurate and data-driven credit decision-making processes. The C4.5 method is employed to classify customer credit data based on parameters such as payment history, loan amount, income level, and other financial indicators that affect credit feasibility. Data processing and model construction were conducted using RapidMiner software to generate a classification tree that can distinguish between eligible and non-eligible credit applicants. Based on the analysis of 350 customer records, the resulting model achieved an accuracy of 92.29%, precision of 99.10%, and recall of 89.80% for the “Not Eligible” class. The most influential variables were loan amount and monthly income, followed by collateral value, age, and loan tenor. The findings are expected to improve PT. XYZ’s credit assessment accuracy, reduce non-performing loan risks, and serve as a reference for developing decision support systems in credit risk management.