Gidan Madi Multi-Disciplinary Journal of Teacher Education (GMMJTE)
ISSN: 3092-8397

Vol. 2 Issue No. 2 June 2026 / Article


A COMPARATIVE REVIEW OF MACHINE LEARNING AND DEEP LEARNING TECHNIQUES FOR EARLY STUDENT DROPOUT DETECTION AND PREVENTION

BODINGA, MUHAMMAD JIBRIL
Department of Computer Science, Federal College of Education Gidan Madi, Sokoto, Nigeria
ibsonbodinga.doc@gmail.com

DR. SHEHU, S. TUDU
Department of Computer Science, Faculty of Computing, Sokoto State University, Sokoto, Nigeria

BODINGA, MUHAMMAD M.
Department of Fine & Applied Arts, Shehu Shagari College of Education, Sokoto, Nigeria

Abstract
The review examined the use of Machine Learning (ML) and Deep Learning (DL) techniques for early student dropout detection and prevention. Employing systematic search of academic database (google Scholar) for peer-reviewed studies published between 2021 and 2026, following inclusion and exclution criteria, 16 relevant rearrch papers were selected and analyzed, the scope focus on higher education context, Studies show that ML algorithms such as Logistic Regression, Decision Trees, Random Forests, Support Vector Machines, and Gradient Boosting provide reliable and interpretable predictions, while DL models, including Artificial Neural Networks, Recurrent Neural Networks, and Long Short-Term Memory networks, are more effective at capturing complex patterns in student academic and behavioral data. Although ML models require less computational power and perform well on smaller datasets, DL models often achieve higher predictive accuracy with larger datasets. Key challenges include data quality, class imbalance, privacy concerns, model interpretability, and generalization across institutions. The review highlights the potential of hybrid ML-DL approaches and recommends future research on explainable AI, real-time intervention systems, and standardized evaluation methods to improve student retention and academic success. These findings support the design and implementation of a predictive analytics system for early student dropout detection and prevention at the Federal College of Education, Gidan Madi.



Date Published

2026-07-06


How to cite



Category


Articles


Keywords


Machine learning, deep learning, student dropout prediction, predictive analytics, early warning systems