<![CDATA[This study aims to analyze the comparison of K-Means and K-Medoids algorithms in measuring the level of student satisfaction with academic services at the Islamic Institute of Mamba'ul Ulum Jambi. Student satisfaction data were collected through questionnaires and analyzed using both algorithms with the help of RapidMiner tools. Clustering results were evaluated using the Davies Bouldin Index (DBI) to determine the most optimal algorithm. The results showed that most students at the Islamic Institute of Mamba'ul Ulum Jambi were very satisfied with the academic services provided. Clustering with K-Means and K-Medoids successfully grouped students into three clusters: "Very Satisfied", "Satisfied", and "Unsatisfied". The K-Means algorithm produced clusters with 450 members ("Very Satisfied"), 351 members ("Satisfied"), and 218 members ("Unsatisfied"). Meanwhile, K-Medoids produced clusters with 638 members ("Very Satisfied"), 270 members ("Satisfied"), and 111 members ("Unsatisfied"). Based on the DBI value, the K-Medoids algorithm (0.222) showed better performance than K-Means (0.396) in clustering student satisfaction data. This study has important implications for the Islamic Institute of Mamba'ul Ulum Jambi in evaluating and improving academic services]]>
Copyrights © 2025
