IMPLEMENTATION OF MACHINE LEARNING BASED SCHOOL CLASS PLACEMENT PREDICTION SYSTEMS FOR SECONDARY SCHOOL, USING RANDOM FOREST

Abstract

This study presents the development and implementation of a machine learning-based framework for optimizing class placements for SSS (Senior Secondary School) classes for JSS (Junior Secondary School) students based on their historical academic results in Nigerian secondary schools, with a focus on the Federal Capital Territory (FCT). Traditional placement methods often rely on subjective evaluations and standardized exams, which may introduce bias and overlook key behavioral and demographic factors. To address these limitations, a predictive model using the Random Forest algorithm was developed, leveraging data from 1,500 anonymized student records spanning academic scores, demographic details, and behavioral indicators. The dataset underwent preprocessing, including normalization, one-hot encoding, and handling of class imbalance using SMOTE. The model’s performance was evaluated using accuracy, precision, recall, and F1 score, achieving an accuracy of 87.6%, outperforming baseline classifiers like Decision Trees and Naive Bayes. Feature importance analysis identified Mathematics, English, and attendance as key predictors. The model demonstrated a 29% improvement in placement accuracy over traditional methods, significantly reducing misclassifications, especially among students from marginalized backgrounds. The findings affirm the potential of machine learning in enhancing fairness and precision in educational decision-making. This research offers a scalable, data-driven approach to class placement and highlights the broader applicability of predictive analytics in educational planning and reform. Based on the research conducted, the following recommendations were made among others; institutions of learning should adopt the use of machine learning for proper students’ class placement and support student.