Predictive Analytics for Reducing University Dropout Rates

Abstract

Higher education institutions face a problem with student turnover that has many aspects and affects both students and universities in different ways. Using predictive analytics and machine learning, this study shows a new way to deal with this problem. The main goal is to create predicting algorithms that can predict which students are most likely to drop out, so colleges can get involved in their lives in a timely and effective way. As part of this method, the authors collect and preprocess a large dataset from different university records. This dataset includes information about academic success, socioeconomic background, participation in campus activities, and psychological health. The study uses advanced machine learning methods to look at all of these different data points. It focuses on feature selection and engineering to find the most important factors that predict student dropout. Rigid validation methods are used to test how well the model works, making sure that it can accurately and reliably predict the future.