<![CDATA[The application of machine learning in educational data analysis has become increasingly important in improving learning evaluation systems. Classification methods are widely used to identify student performance patterns and support academic decision-making. Among the most commonly applied algorithms are Support Vector Machine (SVM) and Naive Bayes, both of which have shown competitive performance in various classification tasks. However, their effectiveness in learning evaluation data classification still requires deeper empirical investigation. This study aims to compare the performance of SVM and Naive Bayes in classifying learning evaluation data based on accuracy, precision, recall, and F1-score. The study used a quantitative experimental approach with a dataset of 1,250 student learning evaluation records consisting of attendance, assignments, participation, midterm, and final examination scores. Data preprocessing included cleaning, normalization, and transformation before model training. The dataset was divided into 80% training data and 20% testing data, with 10-fold cross-validation for validation. The results indicate that SVM achieved an accuracy of 89.6%, precision of 88.9%, recall of 90.2%, and F1-score of 89.5%, outperforming Naive Bayes which obtained 84.3%, 83.7%, 85.1%, and 84.4% respectively. The findings confirm that SVM provides better performance for complex educational datasets and can be recommended for data-driven learning evaluation systems.]]>
