Result: Developing reliable predictive models using machine learning techniques for student admission at Bayero University, Kano, Nigeria

One of the ongoing challenges faced by many universities in northern Nigeria is the optimal utilization of data, particularly in some academic aspects. Many universities across the globe have successfully leveraged data effectively to improve their academic decisions by using the latest technological trends like machine learning, which helps aid decision-making, reduces the time, and minimizes errors. While several studies have employed machine learning to predict student outcomes, few specifically address the unique challenges faced by universities in Nigeria. Although some studies used Nigerian data, most studies focus on datasets from online or non-African contexts. To address these issues, this study utilized a historical dataset from Bayero University Kano's 2018/2019 undergraduate admissions to develop reliable predictive models that can predict whether a student would be admitted or not, using five algorithms: Random Forest, Decision Tree, Logistic Regression, K-Nearest Neighbors, and Naïve Bayes. The models were implemented and evaluated, key predictive features were identified, and optimization was performed. The results of the study indicate that Random Forest achieved the highest accuracy (98.39%), followed by Logistic Regression (97.43%) and Decision Tree (96.46%) in predicting student outcomes. These findings demonstrate the potential of ML to improve admissions transparency and efficiency in Bayero University Kano.