Using machine learning models to predict vaccine hesitancy: a showcase of COVID-19 vaccine hesitancy in rural populations during the pandemic.

Understanding vaccine hesitancy is a critical public health challenge, yet traditional statistical methods often fail to capture the complex drivers behind it. This study uses COVID-19 vaccine hesitancy in a rural population as a case study to demonstrate a more powerful and interpretable machine learning workflow. We compared seven models and found that non-linear approaches significantly outperformed logistic regression in predictive accuracy. Interpretation of the best-performing model identified vaccine safety perceptions as the most important predictor. This approach revealed nuanced, non-linear relationships with feature importance and partial dependence plots. This work serves as a practical guide for researchers, showing how a machine learning framework provides not only more accurate predictions but also a richer, more actionable understanding of complex human behaviors to better inform public policy.