Fig. 4. Model performance using different prediction times of day for training and test data.

a Positive predictive value. b Negative predictive value. c F1 score. d Area under the receiver operating curve (AUROC). e Area under the precision-recall curve (AUPRC). f Normalised mean absolute error (MAE) (mean difference in predicted and actual discharges per day divided by the mean number of discharges per day).