Penerapan K-Means Clustering untuk Pemetaan Pola Pelanggaran Kehadiran Siswa SMK Hidayatus Sholihin

Abstract

Online attendance data has the potential to support decision-making in schools; however, conventional evaluations are generally limited to summarizing cumulative percentages, which are unable to distinguish specific patterns of student misconduct. This study aims to analyze patterns of attendance violations among students at SMK Hidayatus Sholihin based on data regarding disciplinary behavior, specifically tardiness and unexcused absences. The study applied the Knowledge Discovery in Databases (KDD) framework to data from 63 10th-grade students during the odd-numbered semester of 2025. Data processing was conducted using Altair AI Studio with the K-Means Clustering algorithm, while the number of clusters was determined using the Elbow Method, and validation was performed via the Silhouette Score, which yielded a score of 0.502. The results showed that the data could be reliably mapped into three clusters: Cluster 0, consisting of 16 students exhibiting a pattern of “Acute Truancy”; Cluster 1, consisting of 12 students exhibiting a pattern of “Moderate Tendency to Be Late”; and Cluster 2, consisting of 35 students exhibiting a pattern of “Acute Tardiness Violations.” These clustering results serve as the basis for formulating more targeted intervention strategy recommendations to support the enforcement of school rules and the fostering of student discipline.