University Student Enrolment Prediction in Malaysia with State Space Model

Abstract

Accurate university student enrolment is crucial for informing decisions about academic infrastructure, personnel, and funding as it significantly impacts institutional capacity planning, resource allocation and national education policy. However, the accuracy of enrolment forecasting is influenced by demographic shifts, economic conditions and policy factors. This paper proposes the use of a state space model to predict undergraduate student enrolment in Malaysian public and private universities. The enrolment dataset from 2002 to 2024, obtained from the Ministry of Higher Education, are normalized to ensure numerical stability and comparability for modelling. Descriptive statistics, such as mean and standard deviation, are processed for analysis of standardization of the data. The performance of the model is evaluated using the mean squared error (MSE) to assess its effectiveness. Hence, the dynamic changes and long-term enrolment trends can be detected by the state space model, which produces low prediction errors. Moreover, the state space model expresses the lowest MSE values compared to the simple moving average and exponential smoothing models for predicting enrolment in both public and private universities. These findings confirm that the study objectives have been fulfilled and the effectiveness of the state space model in predicting the student enrolment in Malaysian universities.