A Novel DNA-Inspired Framework to Study University Dropout: Insights from Politecnico di Milano

57/2025

A Novel DNA-Inspired Framework to Study University Dropout: Insights from Politecnico di Milano

Wednesday 17th September 2025

Guagliardi, O.; Masci, C,; Breschi, V; Paganoni A, ; Tanelli, M.

This study presents Dropout-DNA, a novel data-driven tool designed to assess university dropout risk by profiling students through a combination of early indicators and academic progress. The approach emphasizes the need for context-aware and interpretable models in predicting student dropout, offering a significant advancement in the field of student retention analytics. Results show that while early indicators are valuable, incorporating academic performance significantly enhances predictive accuracy. The model, although generalizable across engineering courses, performs best when tailored to the specific degree program it was trained on. This finding underlines the importance of adapting predictive tools to the unique characteristics and dropout patterns of individual study programs. The practical implications are considerable: by identifying at-risk students early, institutions can implement targeted and personalized interventions, improving the effectiveness of student support services. The Dropout-DNA’s quantifiable representation of risk allows for more strategic policy-making at the institutional level. Looking ahead, future research will focus on the temporal evolution of dropout risk profiles, enabling dynamic, time-sensitive monitoring and intervention throughout the academic journey.