Although printing the predictions alongside the test set's original diagnoses is useful, it does not give a clear sense of how well the model predicts if a tumor is malignant