Risk analysis of developing vision problems for learners using data mining techniques

Abstract

This research investigates vision problems among higher education students, with the primary objective of exploring their eye usage behaviors and identifying risk factors contributing to conditions such as myopia, hyperopia, and astigmatism. Data were collected from 100 students through in-depth questionnaires.  Data Mining techniques were used to examine the relationship between eye usage behaviors and the occurrence of vision problems. The results indicated that 35% of the students had myopia and astigmatism, while 17% reported no vision problems. Among the models used - Decision Tree, Naïve Bayes, Lazy IBK, and Multilayer Perceptron – the Multilayer Perceptron model had the highest accuracy at 99%. The research also offers recommendations for improving eye usage behaviors and implementing appropriate eye rest practices to reduce the long-term risk of vision problems.