Abstract—Over the past decades, machine learning has been successfully applied to every sector of business to help learn customer needs, make intelligent decisions, and to better serve customers. This includes institutions of higher educations. Specifically, machine learning models can be built on the data from educational settings and used to improve student learning experiences as well as institutional effectiveness. In this paper we propose an innovative approach for solving class imbalanced problems and apply it in student retention prediction. The approach employs a newly developed hybrid data sampling procedure and boosting algorithm to enhance classification performance on data with imbalanced class distribution. Experiments with a collected student data set indicate that the proposed approach is capable of classifying data with limited info and skewed class distribution effectively and furthermore, in comparison with several popular learning algorithms that include decision trees, naïve Bayes and support vector machines, their cost-sensitive counterparts, as well as RUSBoost and SMOTEBoost, it delivers a superior classification performance.
Index Terms—Imbalanced classification, data sampling, ensemble algorithms, student retention.
E. P. Jiang is with the University of San Diego, San Diego, CA 92110 USA (e-mail: jiang@sandiego.edu).
Cite: Eric P. Jiang, "Applying a Hybrid Sampling and Boosting Approach to Predict Student Retention," International Journal of Machine Learning and Computing vol. 12, no. 5, pp. 185-192, 2022.
Copyright © 2022 by the authors. This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited (CC BY 4.0).