Prediksi Tingkat Pemahaman Siswa Smp Swasta Tenera Berdasarkan Aktivitas Belajar Menggunakan Naive Bayes

Abstract

This study aims to develop a prediction model for determining the comprehension level of students at SMP Swasta Tenera using the Naive Bayes algorithm based on academic learning activities. The prediction model utilizes student learning data, including subject grades and attendance records, as the main variables to classify students’ comprehension levels. The data used in this study were collected from 90 students consisting of Grade VII (28 students), Grade VIII (23 students), and Grade IX (39 students), covering performance data from 10 subjects. The research method applies a quantitative approach with data processing and classification analysis using the Naive Bayes algorithm. The evaluation results show that the developed model achieved an accuracy level of 83.33%, with a precision value of 78.12%, recall of 98.04%, and an F1-score of 86.96%. The prediction results indicate that Grade VII students have the lowest comprehension level at 42.9%, followed by Grade VIII at 60.9% and Grade IX at 64.1%. This research demonstrates that machine learning-based prediction systems can support educational decision-making by identifying students who require learning assistance and enabling schools to implement faster, more targeted, and effective academic interventions.