Predicting Students Final Academic Performance using Feature Selection Approaches

Data analytics and data mining techniques are increasingly being used to assist managers, business leaders, and policymakers in their decision-making. These techniques help them distill the useful information hidden in their ever-growing databases that are distributed across various data stores. Another area in which these techniques are utilized is the education domain, wherein data analytics can be leveraged to predict students' academic performance. Teachers will therefore be better informed in providing their students with additional support if required. Learning Management Systems (LMS) captures student online learning data and are extensively used by researchers to gather data-driven insights on students' overall academic performance. This study aims to predict students' academic performance with the LMS data from an online training course using various machine learning algorithms. The study further highlights that selection of features by using feature selection methods will enable early prediction of student academic performance. Early predictions can benefit those students who are considered to be at risk of failing a course with targeted early interventions to help improve their performance throughout the course.