<![CDATA[Credit is now a trend in society. Credit problems are the history of incorrect use of credit cards. The impact can cause bad credit. If customers do not pay the debt that has been agreed with the bank, they can increase their credit risk. In this study, researchers applied the C4.5 algorithm without optimization and the C4.5 Algorithm with Backward Elimination Feature Selection Optimization to classify creditworthiness status. Researchers used 481 vehicle credit records with "bad" and "good" reviews. The independent variables used in the study were dependent status, age, last education, marital status, occupation, company status, income, employment status, house condition, length of stay and down payment. From the results of the study and testing, the performance of the C4.5 model without backward elimination for creditworthiness assessment provided a truth accuracy level of 91.90% with an area under the curve (AUC) value of 0.915. While the performance of the C4.5 model with backward elimination provided a truth accuracy level of 94.80% with an area under the curve (AUC) value of 0.973. This proves that optimization with backward elimination can improve the performance of the classification method used.]]>
Copyrights © 2025
