Machine Learning-Based Analysis of Academic Performance Determinants in Somaliland: Insights from the 2021/2022 National Secondary School Exams

This study employs machine learning algorithms to predict the determinants of students’ academic performance in Somaliland using data from the 2021/2022 academic year. The educational landscape in Somaliland is characterized by challenges in implementing quality education, especially at the secondary level. This research aims to uncover factors influencing student performance in national secondary exams and compare the effectiveness of machine learning algorithms, including random forest, naive Bayes, and logistic regression. It utilizes a dataset encompassing 14,342 students and examines variables such as student residence, administrative region, school type, and gender. The results indicate that the random forest model outperforms other algorithms, achieving 73.8% accuracy in predicting student performance. The traditional logistic regression analysis further highlights the impact of region, residence type, and school type on academic outcomes. These findings contribute to understanding the factors affecting students’ performance in Somaliland and provide insights for educational policy and interventions.