Data-mining applications with the admission data of adult learners in higher education: a pilot study

Admission offices are inundated with information from a variety of data sources and applications. This includes admission data such as student profiles and demographics, as well as academic and professional experiences. The paper outlines a pilot study that uses data-mining applications with the admission data of adult learners in a Singapore university. The application methodology has a sequence of four phases that leads to the building of relevant data-mining models. The analysis of the admission data is used to determine the best-fit model to predict applicants' academic performance. From the evaluation and validation of the different predictive models, the CHAID decision tree is selected as the predictive model. With this model, the probability of academic performance is computed for incoming and existing students by tracing the decision tree.