Table 2 Performances of prediction... | Emerald Publishing

Table 2

Performances of prediction models

ML model	Accuracy	Specificity	Kappa	ROC-AUC	HRI level	Precision	Recall (sensitivity)	F1 score
RF	0.70	0.700	0.396	0.772	Non-fatal	0.62	0.70	0.66
	0.70	0.700	0.396	0.772	Fatal	0.78	0.70	0.74
					average	0.70	0.70	0.70
XGB	0.66	0.677	0.365	0.762	Non-fatal	0.57	0.70	0.63
	0.66	0.677	0.365	0.762	Fatal	0.76	0.64	0.69
					average	0.66	0.67	0.66
KNN	0.64	0.677	0.317	0.680	Non-fatal	0.54	0.63	0.58
	0.64	0.677	0.317	0.680	Fatal	0.72	0.64	0.67
					average	0.63	0.63	0.63
LR	0.69	0.677	0.365	0.764	Non-fatal	0.61	0.67	0.63
	0.69	0.677	0.365	0.764	Fatal	0.76	0.70	0.73
					average	0.68	0.69	0.68

ML model	Accuracy	Specificity	Kappa	ROC-AUC	HRI level	Precision	Recall (sensitivity)	F1 score
RF	0.70	0.700	0.396	0.772	Non-fatal	0.62	0.70	0.66
	0.70	0.700	0.396	0.772	Fatal	0.78	0.70	0.74
					average	0.70	0.70	0.70
XGB	0.66	0.677	0.365	0.762	Non-fatal	0.57	0.70	0.63
	0.66	0.677	0.365	0.762	Fatal	0.76	0.64	0.69
					average	0.66	0.67	0.66
KNN	0.64	0.677	0.317	0.680	Non-fatal	0.54	0.63	0.58
	0.64	0.677	0.317	0.680	Fatal	0.72	0.64	0.67
					average	0.63	0.63	0.63
LR	0.69	0.677	0.365	0.764	Non-fatal	0.61	0.67	0.63
	0.69	0.677	0.365	0.764	Fatal	0.76	0.70	0.73
					average	0.68	0.69	0.68

Source(s): Authors’ own work