Analisis Penilaian Kinerja Siswa untuk Menentukan Siswa Terbaik di Sekolah Melalui Metode Naive Bayes di Songsermsasana School, Hat Yai

Conventional student performance evaluation at Songsermsasana School, Hat Yai, Thailand, relies on subjective academic scores, lacking systematic integration of multidimensional indicators. This study aims to analyze Naive Bayes classifier application for objective outstanding student determination using secondary performance data. Employing quantitative non-experimental design, the research utilized 100 purposively selected student records covering academic grades (0-100), attendance (0-100%), attitude (1-3 scale), extracurricular participation (0-1), and achievements (0-1). Data preprocessing involved numerical encoding followed by Naive Bayes probabilistic classification based on Bayes' Theorem, evaluated through confusion matrix metrics. Results demonstrate 90% accuracy and 100% recall, successfully identifying all outstanding students without false negatives despite three false positives. The model confirms Naive Bayes effectiveness for transparent, data-driven decision-making in educational assessment. Findings support implementation as school decision support systems while recommending hybrid algorithms for future enhancements.