Describir: Comparison of accuracy metrics (sensitivity, specificity, and AUC ROC) of each model defined by what subset of data was used for training and testing.