Intelligent Information System: Leveraging AI and Machine Learning for University Course Registration and Academic Performance Enhancement in Educational Systems

Due to the issues in the current registration systems, students often randomly select any available courses and receive exceptions from their advisors. Even when students follow their advisors’ recommended study plans, many still fail courses and see their performance decline. Advisors and admission departments struggle to explain this phenomenon. In this work we hypothesize that the misalignment of one or two courses with a student’s profile or mindset could be the main reason for this problem, despite these courses being carefully arranged by specialists in the study plan. This work proposes an innovative solution using machine learning to guide students in selecting courses that best match their study plans. By analyzing historical data of similar cases, this approach creates smart rules for future registration systems. The results of this study include the development of machine learning models to predict suitable subject combinations for students. Decision trees was trained on preprocessed data. Additionally, the work successfully created an AI recommendation system for personalized academic improvement. The experimental findings show the performance and structure of the decision tree classifier model. The tree has 409 leaves and a total of 817 nodes. The model accurately categorized 2899 instances, with an accuracy rating of 62.654%. In contrast, 1728 cases were classified wrongly, resulting in a 37.346% mistake rate. The mean absolute error is 0.1148, whereas the root mean squared error is 0.2774. These metrics were calculated for a total of 4627 instances.