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{193346,
        author = {Shravanee Kawalkar and Prof. Chanchal A. kshirsagar and Shweta Gedam and Shruti Kududula},
        title = {Predictive analytics to reduce student dropout rates},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {10},
        pages = {124-130},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=193346},
        abstract = {Student dropout is a persistent issue in educational institutions, leading to significant academic, social, and economic consequences for students and institutions alike. Traditional methods of identifying at-risk students often rely on reactive measures, which limit the effectiveness of intervention strategies. This study investigates the use of predictive analytics as a proactive approach to reducing student dropout rates by leveraging institutional data to identify students at risk of attrition at an early stage.
The research proposes a comprehensive predictive framework that integrates historical academic records, attendance data, demographic variables, socio-economic background, behavioral indicators, and engagement metrics from digital learning platforms. Data preprocessing techniques—including handling missing values, normalization, and feature selection—are applied to ensure data quality and model reliability. Multiple machine learning algorithms, such as Logistic Regression, Decision Trees, Random Forest, Support Vector Machines, and Gradient Boosting, are implemented and compared to determine optimal predictive performance.
Model evaluation is conducted using key performance metrics, including accuracy, precision, recall, F1-score, and Area Under the ROC Curve (AUC-ROC). Special emphasis is placed on recall and minimizing false negatives to ensure high-risk students are not overlooked. The results demonstrate that ensemble models, particularly Random Forest and Gradient Boosting, provide superior predictive capability compared to traditional statistical models. Feature importance analysis reveals that attendance consistency, cumulative GPA trends, course completion rates, financial aid status, and engagement levels are the most significant predictors of dropout risk.
In addition to prediction, the study outlines a decision-support system that translates risk scores into actionable intervention strategies, such as personalized academic counselling, tutoring programs, peer mentoring, mental health support, and financial assistance initiatives. The framework also incorporates continuous monitoring to update risk predictions dynamically throughout the academic term.
Ethical considerations—including data privacy, transparency, fairness, and bias mitigation—are carefully examined to ensure responsible implementation of predictive systems. The study concludes that predictive analytics, when combined with targeted support mechanisms, can significantly improve student retention rates, enhance institutional planning, and promote student success through timely and data-driven decision-making.},
        keywords = {},
        month = {March},
        }