A Survey on Enrolment Analysis via AI-Based Forecasting Models in Academic Institutions

With the increasing demand for data-driven planning in academic institutions, forecasting student enrolment has become a critical application of artificial intelligence (AI). AI-based forecasting models, especially those employing machine learning (ML) and deep learning techniques, have demonstrated significant potential in predicting enrolment trends and supporting institutional decision-making. This paper presents a comprehensive survey of recent studies that utilize AI models for enrolment analysis across various educational settings. A total of 14 relevant studies were reviewed and analyzed. We identify the most commonly used prediction models-such as logistic regression, and random forests regression-as well as the most influential data features, including demographic, academic, and geographical variables. The paper also classifies these models and features into taxonomies, providing a clear reference framework for future research and institutional implementation. Our findings show that neural networks, particularly Long Short-Term Memory (LSTM) models, outperform traditional statistical methods in capturing non-linear patterns and temporal dynamics. This survey provides insights into model selection, data features, and existing gaps, thereby serving as a foundation for future research and implementation of AI-based enrolment forecasting in academic institutions. To the best of our knowledge, no prior surveys have systematically similarly analyzed enrolment forecasting in the context of academic institutions using AI techniques.