Predictive Modeling for Enhancing Academic Performance in Nigerian Polytechnic Education

Abstract

This study presents a machine learning-based approach to enhance the classification and optimization of students’ academic performance within Nigeria’s polytechnic education system. The polytechnic system is pivotal in providing technical and vocational education, but challenges persist in nurturing students’ academic achievement. This article explores the complexities influencing academic performance and proposes strategies for improvement using machine learning algorithms. The research utilizes linear and support vector regression models to predict students’ cumulative grade point averages (CGPA). A dataset from Akwa Ibom State Polytechnic, Ikot Osurua, comprising total courses, credit units, department, and previous grade point average (GPA), is employed for model development and evaluation. Both models achieve similar predictive performance, but linear regression slightly outperforms support vector regression. The results highlight the significant role of variables like total courses, the type of academic department, and previous GPA in predicting CGPA. This study offers a valuable tool for assessing and improving students’ academic performance in Nigeria’s polytechnic education system, with potential for broader applications in higher education. Further research involves expanding the dataset and considering additional factors beyond result records to enhance the model’s robustness and applicability.