Student admission and enrolment are pivotal processes for universities, directly influencing institutional planning and academic outcomes. To enhance this decision-making process, this paper aims to present the development of a multi-model analytics system to predict the likelihood of student candidates accepting admission offers and achieving good academic results.
The multi-model analytics system integrates six machine learning models to support data classification and regression. A majority voting approach was adopted to combine the results from the top three models and generate a comprehensive prediction. In addition, interactive analytics dashboards were developed to facilitate data visualisation, enabling stakeholders to derive actionable insights from admission trends and outcomes.
Evaluation results showed that the system achieved an accuracy of 62%, a recall of 83% and a precision of 63%. These results demonstrate the system’s capability in forecasting student admission and enrolment, with a particular strength in identifying students who ultimately enrolled in a programme.
Beyond student recruitment, the system supports strategic planning, resource allocation and the development of teaching and learning accommodations. By analysing trends in students’ background information, universities can better align their offerings with the needs and preferences of incoming cohorts.
This study introduces a novel multi-model analytics approach to support student admission and enrolment. The system’s predictive capabilities and visualisation tools offer a scalable solution for enhancing institutional decision-making and operational efficiency.
