Klasifikasi Perilaku Belajar Siswamenggunakan Algoritma K-Nearest Neighbor (Knn)

Manual assessment of student learning behavior at SMKN 1 Bengkalis tends to be subjective and inefficient. This study develops a learning behavior classification system for Computer and Network Engineering (TKJ) students using the K-Nearest Neighbor (KNN) algorithm to produce more objective evaluations. The system was developed using the Rapid Application Development (RAD) method, utilizing data from 194 students based on 16 validated behavior indicators. The dataset was split into training and testing sets with a 70:30 ratio. The K-Nearest Neighbor (KNN) algorithm was implemented with K=3 and Euclidean distance calculation to determine nearest neighbors for classification. Testing results demonstrate that the model successfully classifies student behavior into two categories: "Active" and "Inactive", achieving an overall accuracy of 91.53%. Precision values reached 97.06% for the Active class and 84% for the Inactive class. The web-based system, built using Python Flask framework and MySQL database, facilitates Guidance and Counseling teachers in monitoring student behavior in real-time through statistical dashboards and classification reports. In conclusion, the K-Nearest Neighbor (KNN) algorithm proves effective in accurately classifying student learning behavior, thereby supporting better educational decision-making and targeted intervention strategies in vocational school environments