Classification of Learner Retention using Machine Learning Approaches

Learner retention issues require a huge commitment from a university as the process of monitoring learners' re-registration status from the beginning of each semester until they graduate can be quite tedious. When the number of learners who re-register for a subsequent semester is low, it not only affects the university's image but also its ranking and reputation in the education sector. Therefore, the university must identify, at an early stage, the likelihood of a learner is not retained in the following semester. This study proposed to experiment with the classification methods for solving the issue of learner retention at Open University Malaysia by comparing three Supervised Machine Learning algorithms namely Logistic Regression, Support Vector Machine, and k-Nearest Neighbor. The performance of these algorithms was evaluated based on accuracy, precision, recall, and f-measure. It is determined that Support Vector Machine showed the best accuracy in classifying the learners' retention rate with 80% accuracy. The benefit of performing Machine Learning is that it enables the identification of at-risk learners at the earliest opportunity and therefore implement the earliest interventions to retain them.