A Predictive Model for Students’ Performance and Risk Level Indicators Using Machine Learning

Abstract

Educational data mining has been a veritable tool for predictive analytics which aids informed decision making and policy formulation tasks in the education industry. This study identifies relevant attributes from academic data of graduate teachers at a College of Education in Nigeria and develops a model that forecasts academic performance of teachers-in-training by assigning risk levels to their academic standing dataset. The model analyses success indicators from the list of attributes and assigned risk levels is a veritable tool for monitoring and evaluation of teachers-in-training by school administrators for an improved performance before graduation. The result shows that core courses offered in the first and second semesters of the second year of studentship have a healthy level of significance in forecasting teachers-in-training overall academic performance. Any deficient in such courses, therefore increases the risk level. A noteworthy discovery is the less significance of the teaching practice program, which assigns teachers-in-training to schools for six months, in determining their final academic standing on graduation.