Author: Dr. Sameer Jain

Abstract
This study presents an original data-driven framework for predicting student academic performance using a random forest classifier (RFC). The research, by incorporating socio demographic, behavioural, and academic factors in a single model, transcends individual factors to offer holistic solutions for early academic interventions. Using a sample from 2,392 students, the model is highly accurate in predicting grades from A to F. The research also shows that the most significant factors for academic success are weekly study habits, attendance, and parental support, with gender and ethnicity having a lesser impact. From a methodological standpoint, the research illustrates the potential for understandable machine learning in decision-making in education. From a practical standpoint, the research supports
evidence-based interventions for improved student outcomes.

Published in:  Journal of Global Information and Business Strategy (JGIBS), January-December 2025 (Volume-17, Number-1, 2025)

Keyword: Student Performance Prediction | Academic Achievement | Machine Learning | Random Forest Classifier | Higher Education

DOI: 

Download Paper