Improving Student Achievement Clustering Model Using K-Means Algorithm in Pasundan Majalaya Vocational School

This study analyzes and enhances the student achievement clustering model at SMK Pasundan Majalaya using the K-Means algorithm. The Knowledge Discovery in Databases (KDD) method and RapidMiner AI Studio 2024.1.0 were used to process data from 125 students based on 15 metrics, including academic scores and attendance rates. For group evaluation, the Elbow method and Davies-Bouldin Index (DBI) were employed. The results showed optimal clustering with 2 groups and a DBI value of 0.893. Analysis results revealed significant differences in characteristics between the two groups. Cluster_1 consists of 38 students and has lower score patterns (60-80), with attendance rates of 94-100%, and a positive correlation between attendance and academic achievement. On the other hand, Cluster_0 consists of 86 students and shows higher score patterns (67.5-87.5), with attendance rates of 80-100%, and demonstrates a positive correlation between attendance and academic achievement. Schools can use this clustering model to create learning approaches that are better suited to each student group.