Prediction of Student Dropout Using Machine Learning and Data Mining Techniques

Afroze Ansari, K. K. Savitha. In: AIoT and Cloud-Based Smart Education Systems: Intelligent Campus Management and Learning Analytics — ISBN: 978-81-69372-12-1. Pages: 226 - 246

Student attrition poses a significant challenge within educational systems globally, impacting institutional efficacy and student achievement. This chapter examines the utilisation of Machine Learning (ML) and data mining methodologies to forecast student attrition and facilitate proactive intervention tactics. Through the analysis of several data sources including academic achievement, attendance, demographic data, and behavioural patterns predictive models can accurately identify at-risk pupils. The chapter delineates a thorough approach for predicting dropout, encompassing data preprocessing, feature selection, model construction, and evaluation. It examines essential machine learning algorithms like decision trees, random forests, support vector machines, logistic regression, and ensemble approaches, in conjunction with data mining techniques such as clustering and association rule mining. Applications in early warning systems, individualised support, and institutional decision-making are emphasised. Data quality, privacy, bias, and interpretability challenges are rigorously examined. Emerging themes, like explainable AI, real-time analytics, and interaction with intelligent education systems, are also examined. By utilising machine learning and data mining techniques, educational institutions may actively diminish dropout rates and enhance student retention and achievement.