Scholarship Eligibility Prediction Using Gaussian Naive Bayes Based on Academic Potential Test Results

Abstract

A systematic and objective method is required to determine eligibility for scholarships. This study aims to develop a scholarship eligibility prediction model based on Academic Potential Test (TPA) results, utilizing the Gaussian Naive Bayes algorithm. The dataset comprises TPA attributes, including quantitative analysis scores, English proficiency scores, the number of correct answers, and total scores. Scholarship eligibility is classified into two categories: eligible and ineligible.The research methodology encompasses data preprocessing, model construction, prediction, and performance evaluation using metrics such as accuracy, precision, recall, and the F1-score. Experimental results indicate that the model achieved an accuracy of 98.21% on a test dataset consisting of 223 entries; for the "eligible" class, the model attained a precision of 96%, a recall of 100%, and an F1-score. Nevertheless, further validation using an independent dataset is necessary to assess the model's generalization capability and its reliability for real-world scholarship selection processes.