An AI-Based Market Forecasting Model for Enrolment and Revenue Analysis in Academic Institutions

Résumé fourni par la source

This research develops and evaluates an AI-based market forecasting model for enrolment and revenue analysis in swimming academies, using survey data from potential participants. Logistic Regression, Random Forest, and Neural Network algorithms were implemented to predict enrolment likelihood and revenue potential. Results indicate that neural networks achieved the highest accuracy, effectively modelling complex feature relationships. Key predictors include trial class availability, class duration, flexible payment options, and peer recommendations. While the model provides actionable insights for academies to optimize enrolment strategies and revenue generation, its scope is limited by a small, region-specific dataset and revenue captured in binary categories. Future work will expand data collection and integrate external sources to better represent non-enrollees, enabling the development of a continuous revenue prediction model. In addition, deploying the system into real-world platforms such as Customer Relationship Management (CRM) and Enterprise Resource Planning (ERP) will provide practical validation, support enrolment management, personalized marketing, and revenue optimization.