<![CDATA[The Sri Hadi Dharma Savings and Loan Cooperative faces challenges in conducting loan risk analysis, which is still done manually as a result, the decisions made tend to be subjective and inconsistent. This situation has the potential to lead to errors in risk assessment, which could increase the likelihood of non-performing loans. Additionally, the available payment history data is not labeled with risk categories and therefore cannot be directly used in the classification process. This study aims to build a more objective loan risk prediction model by integrating the K-Means and C4.5 algorithms. The data used in this study consists of 7,155 loan payment history records from the 2020-2025 period. The research stages included data preprocessing through cleaning, reduction, creation of a maximum delinquency feature, feature selection, and Min-Max normalization, resulting in 471 unique data points. Next, K-Means was used to form risk groups based on financial characteristics, the clustering results were labeled as low, medium, and high risk categories, then the labeled data were classified using C4.5 with an 80:20 split of training and test data to generate a prediction model. Model evaluation was performed using a confusion matrix with accuracy, precision, recall, and F1-score parameters. The results show that the resulting model is capable of providing predictions with an accuracy rate of 95.79% and generates decision rules that are easy to understand and interpret. Furthermore, the integration of these two methods transforms payment history data into valuable information for identifying members’ risk patterns. This model is effective in supporting more objective, transparent, and measurable decision-making in loan risk management at savings and loan cooperatives]]>
