<![CDATA[The process of determining students’ majors in vocational high schools plays a crucial role in shaping their academic development and future career readiness. However, manual decision-making often leads to inaccuracies due to subjective judgments and limited data analysis. This study aims to develop a more accurate and objective major classification model by integrating the Support Vector Machine (SVM) method with Particle Swarm Optimization (PSO). The dataset consists of 292 student records, including academic scores in Mathematics, Indonesian Language, English, and Science, as well as interest questionnaire results. Initial testing using SVM produced an accuracy of 79.76%, indicating that the model’s parameters were not yet optimal. PSO was then applied to optimize the key parameters C and Gamma, resulting in a significant improvement in model performance. The optimized SVM–PSO model achieved an accuracy of 97.20%, with a precision of 96.33%, recall of 95.22%, and an F1-score of 95.77%. These results demonstrate the capability of PSO to enhance SVM’s pattern-recognition performance and address class imbalance issues, particularly for minority majors. Overall, the integration of SVM and PSO is proven to be effective as a Decision Support System, providing schools with accurate, data-driven recommendations for student major placement.]]>
Copyrights © 2025
