Prediksi Tingkat Keterlambatan Pengumpulan Tugas Mahasiswa Berdasarkan Aktivitas Perkuliahan Menggunakan Algoritma K-Nearest Neighbor (KNN)

The delay in task submission is one of the indicators of low discipline and student engagement in the learning process. This study aims to predict the level of task submission delay among students based on academic activity using the K-Nearest Neighbor (KNN) algorithm. The dataset used includes five predictor variables: attendance rate (%), frequency of LMS login, forum participation, average quiz scores (%), and LMS access time, with one target variable being task delay categorized into three classes: On Time, Moderate Delay, and Severe Delay. The KNN method with k = 3 was applied to the normalized dataset using Min-Max Scaling. The test results showed that the model successfully classified all test data with an accuracy of 100%. Evaluation using the confusion matrix, precision, recall, and f1-score confirmed optimal performance across all delay categories. The study concludes that academic activity significantly influences task punctuality, and the KNN model can serve as a foundation for developing a data-driven early warning system to detect students at risk of delay. However, further research with larger datasets is needed to validate the generalizability of this model.