Shaping the Future: The Use of Machine Learning in Predicting Students’ University Admissions

Abstract

This study aims to predict the university admission status of students using machine learning algorithms. In this context, support Vector Machines, K-Nearest Neighbors, and Random Forest algorithms were used for prediction. The models were evaluated using students' data, employing accuracy, precision, recall/sensitivity, specificity, and F1-score metrics. The overall best results came from Support Vector Machines because it achieved 82.80% accuracy and 89.23% F1 score which demonstrates its effectiveness with high-dimensional datasets. The K-Nearest Neighbor algorithm achieved its best results through sensitivity (96.52%) and specificity (80.95%) which helped it separate different classes effectively. The interpretability of Random Forest proved to be its main benefit despite its lower overall performance. The research demonstrates that machine learning algorithm success depends on the characteristics of the dataset and the specific classification objectives.  This study was conducted in the context of the feature of 785 students aged 18-20. The research was carried out using three different machine learning algorithms. For future studies, it is recommended to use more student features, additional algorithms, and expanded sample sizes to improve prediction accuracy.