| International Journal of Computer Applications |
| Foundation of Computer Science (FCS), NY, USA |
| Volume 187 - Number 139 |
| Year of Publication: 2026 |
| Authors: Shahriar Arefin Zummon, Somapika Das, Aushtmi Deb, Marjana Akter Juti |
10.5120/ijcaef834e7788bd
|
Shahriar Arefin Zummon, Somapika Das, Aushtmi Deb, Marjana Akter Juti . Effects of COVID-19 Vaccines among University Students in Bangladesh: A Survey-based Machine Learning Analysis. International Journal of Computer Applications. 187, 139 ( Aug 2026), 1-12. DOI=10.5120/ijcaef834e7788bd
While young adults are generally considered resilient, the longterm side effects of COVID-19 vaccines in this demographic remain largely unexplored. This study investigates post-vaccination experiences among university students and faculty collected from three universities in Sylhet, Bangladesh through online and offline surveys between January and March 2025. After rigorous data cleaning, a finalized dataset of 591 respondents was obtained covering vaccine type, dosage, demographics, pre-existing health conditions, overseas travel history, and post-vaccination impacts (short-term, medium-term, and long-term). Descriptive, correlation, and predictive analyses were performed. Five supervised machine learning (ML) classifiers - Logistic Regression (LR), Random Forest (RF), Support Vector Machine (SVM), XGBoost, and K-Nearest Neighbors (KNN) - were evaluated alongside K-Means clustering (k = 4) with Principal Component Analysis (PCA) for patient risk profiling. Logistic Regression achieved the highest accuracy (70.79%). Random Forest feature importance identified side-effect severity (52.8%) and vaccine type (34.7%) as the dominant predictors. Respiratory conditions and allergies were the leading comorbidities associated with post-vaccination impacts. Findings provide actionable insights for policymakers and healthcare professionals refining vaccination protocols for young, active populations.