Built a Logistic Regression model in SAS to predict hotel booking cancellations
(AUC 0.84, 79% accuracy), including EDA, data cleaning, and engineered features
that improved model performance.