Prediction of malaria positivity using patients’ demographic and environmental features and clinical symptoms to complement parasitological confirmation before treatment

Table 3 Performance of penalised logistic regression, Bayesian generalised, and decision tree models for training and test sets

Model	Data set	Sensitivity	Specificity	PPV	NPV	AUC (%)
penalised logistic regression	training	0.750	0.821	0.952	0.411	84
	Test	0.818	0.714	0.931	0.455	83
Bayesian generalised model	training	0.765	0.821	0.953	0.426	84
	Test	0.818	0.714	0.931	0.455	76
decision trees	training	0.826	0.500	0.886	0.378	66
	Test	0.818	0.818	0.900	0.400	69

ISSN: 2055-0936