Predictive modelling and analysis of student academic performance using one-vs-all logistic regression approach

The quality of students and their academic achievement is critical to national development and hence the need to be able to predict their performance in order to forestall future failures. As such predictive modelling has drawn significant attention from the research community as it plays a significant role in determining the worth of graduates who will be responsible for economic and social growth of the country. Whereas most previous studies employed the use of data that are not purely academic, using purely students’ academic performance data from 2004 to 2014, this study employs educational data mining technique to describe the impact of the combination of high school grades and final course combination on the graduating class of students at the University of Ghana, Legon. The results obtained showed that high school grades and final course combination at the University are less likely to affect the graduating class of a student.