<![CDATA[Student attendance data are often used only for administrative purposes and has not been fully utilized to support instructional decision-making. This study aims to apply the k-means clustering and decision tree algorithms to group students based on attendance patterns and determine adaptive learning strategies for Mechatronics Engineering students. The combination of these algorithms was selected to enable student grouping through unsupervised learning while simultaneously generating interpretable classification rules for adaptive learning strategy recommendations. A quantitative descriptive method with data mining approach was employed. The dataset consisted of attendance records and cognitive, psychomotor, and affective assessment scores from 36 XI grade students enrolled in Mechatronic Equipment Repair and Maintenance (P3M) course at SMKN 8 Malang. The analysis included data preprocessing, clustering using k-means, classification using decision tree, and model evaluation through the silhouette coefficient and confusion matrix. The result indicate that k-means produced three clusters with a silhouette score of 0.57, indicating reasonably good clustering quality. The decision tree generated recommendations for project-based learning, cooperative learning, and direct instruction with an accuracy of 66.7%. Overall, the proposed approach can assist teachers in identifying student learning characteristics and selecting adaptive instructional strategies to support more effective learning in vocational education.]]>
