Grouping Students Based On Academic Values Using The K-Means Method At A Vocational High School In Bandung

Abstract

A vocational high school in Bandung is dedicated to cultivating competitive and employable graduates.  Nonetheless, the institution faces challenges in conducting comprehensive assessments of pupils' academic data to identify their strengths and weaknesses.  Currently, data analysis relies on basic descriptive methodologies, which often fail to yield adequate insights for informed strategic decision-making.  Furthermore, there is a lack of interactive visualization tools to enhance the presentation of student grouping data.

This study aims to address these concerns by utilizing the K-Means algorithm to categorize pupils based on their academic performance.  This classification yields three clusters that delineate students' attributes in high, medium, and low score categories.  The evaluation results indicate that the model comprising three clusters has the highest Silhouette Score of 0.3364.  This research generates an interactive website as a visualisation tool to display the outcomes of student grouping correctly.

The implementation of this method is anticipated to enhance the school's management of academic data and deliver tailored learning recommendations that more effectively address the needs of students within each cluster.  Therefore, the educational quality of this vocational high school in Bandung can be markedly enhanced.