This report explores the prediction of seasonal and H1N1 vaccination behavior using machine learning techniques. Using a survey dataset containing demographic, socioeconomic, and health-related variables, I applied the XGBoost algorithm to build predictive models for both vaccines. The workflow includes Data preprocessing, including handling missing values and encoding categorical variables, splitting data into training and testing sets, model training and evaluation using the AUC metric, and generating predictions for unseen test data. The results demonstrate the utility of gradient boosting methods for capturing complex patterns in vaccination behavior, providing insights that could support public health strategies. The report includes R code, detailed explanations, and reproducible results, making it a resource for those interested in predictive modeling and health data analysis.

MC 6000

Aulas de programação - Professor Adriano Lauro