Predicting Student’s Dropout Data in Higher Education using Neural Network

Abstract

The biggest downside to higher education is that once enrolled, academic performance is poor. Neural networks are also used to predict the success of MBA students. The authors use a three-layer neural network to divide MBA program applicants into groups of successful and marginal students who support undergraduates' touchstone, undergraduate major, age, and GMAT scores. They got overall prediction accuracy for their model at 89. To assess the ability of neural networks to classify students, the authors compared the results obtained using neural networks to log it and probit regression models. The ability to predict student performance is very useful for college officials who require early action to prevent dropouts.