<![CDATA[Learning interest is one of the important factors that influence student success in following the learning process. However, identification of students' learning interest levels is often done subjectively so that it is less able to describe the student's condition as a whole. This study aims to compare the performance of the K-Means and K-Medoids algorithms in grouping the learning interests of MDTA Qur'an Kisaran students based on academic and non-academic data. The dataset used consists of 119 students with six variables, namely academic grades, attendance percentage, memorization assessment, Qur'an reading ability, moral assessment, and student participation. The clustering process was carried out with three clusters representing the categories of high, medium, and low learning interest. Cluster quality evaluation was carried out using the Davies-Bouldin Index (DBI) and Silhouette Coefficient. The results showed that the K-Means algorithm produced a DBI value of 1.5839 and a Silhouette of 0.2186, while the K-Medoids algorithm produced a DBI value of 1.5950 and a Silhouette of 0.2046. Based on the evaluation results, the K-Means algorithm performed better than K-Medoids in clustering student learning interests. The clustering results can be used to support decision-making in student development, developing learning strategies, and selecting participants for the Inter-Islamic Sports and Arts Week (PORSADIN) in a more objective and data-driven manner.]]>
Copyrights © 2026
