Classification Algorithms to Determine Students’ Specialization in a Higher Education Institution

Abstract

One of the higher education institutions, namely the Faculty of Computer Science of Bina Darma University in Palembang offers courses in information technology (IT). Database, software, and network infrastructure are the areas of specialization available through the IT Study Program at the Faculty of Computer Science. These courses are complementary to those offered at Bina Darma University. Those areas of specialization must be chosen in the fourth or fifth semester, however, many students are still confused and unaware of their interests and potential which may lead to choosing a specialization that does not suit them. In this view, students may not be graduating on time. The study in this article is inspired by this situation. Our idea is to present a prediction model that assists faculty in identifying the best specialization for each student. Primary datasets are those that were gathered from the faculty and include 3599 records with 42 attributes. After that, we looked at how Python programming classification algorithms like Support Vector Machine (SVM), Naïve Bayes, Random Forest, and Decision Tree performed in classifying the areas of specialization of the students. This study demonstrates that the Decision Tree and Naïve Bayes programs reach high accuracy rates of 98,06% and 92,78%, respectively