Table 2.

Comparison of the averages of the classification metrics for the developed hybrid model and the XGBoost approach across different hospitals.

Hospital	Model	Accuracy	Recall	Precision	F1 Score	ROC AUC
Derivation Hospital	XGBoost	0.937	0.815	0.984	0.892	0.930
	Hybrid	0.899	0.784	0.884	0.831	0.921
Validation Hospital 1	XGBoost	0.843	0.756	0.738	0.747	0.842
	Hybrid	0.874	0.821	0.780	0.800	0.863
Validation Hospital 2	XGBoost	0.918	0.964	0.913	0.938	0.933
	Hybrid	0.894	0.930	0.907	0.919	0.933
Validation Hospital 3	XGBoost	0.898	0.744	0.799	0.771	0.852
	Hybrid	0.902	0.753	0.809	0.780	0.881
Validation Hospital 4	XGBoost	0.886	0.802	0.877	0.860	0.864
	Hybrid	0.904	0.826	0.880	0.896	0.892