A Longitudinal Analysis of Entry-Level Student Validation in Engineering: Trends, Predictive Modelling, and Clustering Insights

Asha Bhave, Sheenu Gupta, Sneha Khandait, Sohail Khadpolkar, Jyoti Vanawe

Abstract: This study explores student validation patterns across three academic years of first year (2021?22 to 2023?24) in an engineering institution, focusing on departmental trends and academic predictors. Using validation status (High, Medium, Low), department affiliation, and SSC/HSC scores, the research applies three methods: trend analysis, predictive modelling, and clustering. Trend analysis highlights consistent differences in validation across traditional and emerging departments. Predictive models (logistic regression, decision tree) use pre-admission scores to forecast validation levels with strong accuracy, enabling early identification of at-risk students. Clustering techniques (K-means, hierarchical) reveal distinct student profiles based on academic performance and validation behaviour. The findings demonstrate that academic background significantly influences validation and suggest that data-driven approaches can support targeted interventions, improve student engagement, and enhance retention strategies.

Keywords: student validation, predictive modelling, academic performance, clustering analysis, student retention

How to Cite?: Asha Bhave, Sheenu Gupta, Sneha Khandait, Sohail Khadpolkar, Jyoti Vanawe, "A Longitudinal Analysis of Entry-Level Student Validation in Engineering: Trends, Predictive Modelling, and Clustering Insights", Volume 14 Issue 11, November 2025, International Journal of Science and Research (IJSR), Pages: 995-1000, https://www.ijsr.net/getabstract.php?paperid=SR251112111250, DOI: https://dx.doi.org/10.21275/SR251112111250