Smart Choice: Machine Learning Insight into Factors Influencing Students’ Programme Selection at the Tertiary Institution

Understanding the factors influencing students’ choice of programme of study is increasingly important for tertiary institutions in Ghana, particularly amid rising enrolment rates and growing competition. While prior studies have applied machine learning to predict academic performance, limited research has examined programme selection behaviour at the senior high school level using mixed-type clustering techniques. This study addresses this gap by applying the K-prototype clustering algorithm and supervised classification models to survey data collected from 1,042 final-year Business and Home Economics students across ten senior high schools in Northern Ghana. The clustering process identified three behavioural segments comprising 423, 382, and 237 students, respectively, with the majority aged 16–20 years. Internal validation metrics indicated modest cluster separation. Subsequent classification modelling using Naïve Bayes, Logistic Regression, Decision Tree (J48), Random Forest, and Support Vector Machine (SVM) showed that SVM achieved the highest predictive performance (Accuracy = 99%) when predicting cluster membership. Key influencing factors included parental education, parental occupation, counselling exposure, socio-cultural beliefs, and peer influence. The findings highlight the need for strengthened, context-sensitive guidance and counselling frameworks at the pre-tertiary level to support informed and independent programme selection decisions.