Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{188060,
        author = {R. Udhaya Sankar and C. R. Jothy and J. E. Judith},
        title = {A Hybrid Machine Learning Framework for Early Prediction of Student Academic Performance},
        journal = {International Journal of Innovative Research in Technology},
        year = {2025},
        volume = {12},
        number = {7},
        pages = {776-779},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=188060},
        abstract = {The proactive identification of students at risk of academic underperformance is a critical challenge for educational institutions aiming to improve outcomes and reduce dropout rates. This paper presents a comparative analysis of various machine learning algorithms Support Vector Machine (SVM), K-Nearest Neighbors (KNN), Gradient Boosting (GB), Logistic Regression (LR), and Random Forest (RF) for predicting student academic success. To enhance predictive power, a novel hybrid model that integrates Gradient Boosting with K-Nearest Neighbors (GB+KNN) is proposed, leveraging the complementary strengths of both ensemble and instance-based learning techniques. The models were trained and evaluated on a dataset of 800 student records featuring key academic and behavioral indicators such as attendance, internal assessment scores, assignment performance, previous GPA, study habits, and participation in extracurricular activities. Experimental results demonstrate that the proposed hybrid GB+KNN model achieves superior performance, with an accuracy of 89.2%, precision of 90.1%, recall of 88.7%, and an F1-score of 89.4%. A comparative analysis confirms that the hybrid model outperforms all individual classifiers, with the standalone Gradient Boosting (87.8% accuracy) and Random Forest (86.3% accuracy) being the closest competitors. Feature importance analysis identified Previous GPA, Internal Assessments, and Attendance as the most significant predictors of academic performance. This hybrid framework provides educational administrators with a robust, data-driven tool for the early identification of at-risk students, enabling timely interventions and supporting improved educational outcomes.},
        keywords = {Educational Data Mining, Machine Learning, Predictive Analytics, Student Performance, Hybrid Model, Gradient Boosting, K-Nearest Neighbors, At-Risk Students.},
        month = {December},
        }