Optimasi Support Vector Machine (SVM) Menggunakan Naive Bayes dan Decision Tree untuk Klasifikasi Tema Tugas Akhir Mahasiswa Sistem Informasi

Arfika

Computer Journal
Vol. 2 No. 2 (2024): August

-The School of Informatics and Computer Management (STMIK) Abulyatama faces challenges in classifying student thesis topics. The current manual classification process is inefficient and ineffective. Therefore, the implementation of data mining techniques is needed to manage this data, specifically to automate the classification process. This study aims to optimize the Support Vector Machine (SVM) model by integrating Naive Bayes and Decision Tree algorithms to improve the accuracy of thesis topic classification. Based on the analysis, it can be concluded that both SVM and K-Means can be utilized by decision-makers to categorize thesis topics as a decision support system. Naive Bayes and Decision Tree were shown to optimize SVM and enhance its accuracy. Naive Bayes achieved the highest accuracy at 82.50%, while Decision Tree recorded an accuracy of 62.50%, making Naive Bayes the most suitable model for thesis classification.