<![CDATA[Delayed tuition payments present challenges for higher education institutions, impacting both financial stability and students’ academic progress. This study proposes a predictive model using Support Vector Machine (SVM) optimized by Particle Swarm Optimization (PSO) to identify students at risk of payment delays. The dataset includes academic and social attributes. A dot kernel SVM was evaluated using 10-fold cross-validation. Results show that PSO optimization significantly improved model performance, particularly in recall, which increased from 36.10% to 65.51%, indicating better identification of delayed payment cases. The analysis also reveals that social factors, such as employment and academic status, strongly influence prediction outcomes. These findings highlight the potential of the SVM-PSO model as a decision-support tool for early intervention, enabling institutions to mitigate dropout risks and enhance financial planning. By leveraging this approach, universities can better support students while maintaining administrative efficiency and institutional sustainability.]]>
Copyrights © 2025
