The Role of AI in University Course Registration in the Middle East: AI and Machine Learning Approaches to Improve Academic Performance

This study investigates the use of machine learning in improving course selection for Middle Eastern university students, with a focus on the University of Oman. We offer an intelligent application that uses classification techniques to select the best course combinations based on individual student characteristics. The technology forecasts probable grades based on historical data from 1,972 student records with 26 characteristics, allowing students to make educated course selections. The research employs the (CRISP-DM) methodology, which ensures a methodical approach to data analysis and model creation. The findings demonstrate that with an accuracy rate of 96.37% vs. RandomTree's 84.10%, the J48 algorithm outperforms RandomTree by 12.27%. RandomTree has a misclassification rate of 15.90%, but J48 has a far lower rate of misclassification of 3.63%. J48 outperforms RandomTree in prediction accuracy, with a (MAE) of 0.0287 and a (RMSE) of 0.1259. J48 outperforms all other classes in terms of true positive rates, precision, and ROC areas, as evidenced by class-specific accuracy statistics. This study fills a vacuum in current academic advising systems, which frequently lack tailored, data-driven suggestions. Our suggested system improves academic achievement, lowers dropout rates, and streamlines course preparation, representing a substantial improvement in educational technology. The application's ability to replace traditional advising with a more efficient, personalized method is especially significant at educational institutions throughout the Gulf area.