Using machine learning to predict at-risk students in South African higher education institutions

Section 29 of the South African constitution states that all citizens have a right to a basic education, which includes adult basic education and the right to further one’s education which the State must make available and accessible though reasonable measures McConnachie, McChonnachie and Skelton (2017). Despite this constitutional right, the South African higher education system has a low throughput and high attrition for various reasons (UWC, 2019). The biggest challenge is that student success is determined by a multilayered, complex, and dynamic set of circumstances and events ranging from personal, institutional, and other broader socio-economic factors. The second challenge is that a more reactive approach is being taken to act on and address the high attrition rates in higher education institutions and this is where predictive analytics comes into play. Predictive analytics is a technique that has been widely used in the business world, however in recent years educational institutions have started applying analytics to answer questions such as ‘What are the characteristics of students who persist in their studies to the point where they actually graduate?’ Lourens and Bleazard (2016). The authors continue to highlight that predictive analytics will assist in improving decision-making and providing critical information that will alert institutions of students requiring attention. This research will explore the use of a machine learning model that will be able to predict at risk students within higher education institutions in a South African context.