The rapid growth in enrollment in professional computing programs such as the Bachelor of Computer Applications (BCA) has been accompanied by a notable increase in student dropout rates. Early identification of academically slow learners is therefore essential for timely intervention and improved retention. This study proposes a machine learning–based model for the early identification of slow learners in the BCA program using academic, demographic, behavioral, psychological, and technological factors. Primary data were collected from BCA students enrolled in 11 colleges affiliated with Shivaji University, Kolhapur, using a structured questionnaire. Following a pilot study (n = 580), correlation analysis and Chi-square tests were applied to identify significant predictors, resulting in the selection of 20 influential variables. Multiple classification algorithms were implemented using the WEKA tool, and their performance was compared. The REPTree algorithm demonstrated an optimal balance between accuracy, recall, computational efficiency, and interpretability. The findings confirm that machine learning techniques can effectively support the early identification of slow learners and provide a data-driven basis for academic interventions aimed at reducing dropout rates and improving student performance.