Table 13.

Comparison of different machine learning methods with selected features on Site C dataset.

Methods	AUC			Sensitivity (PPV = 70%)
	Mean	95% CI	p value	Mean	95% CI	p value
Random Forests	0.840	(0.804, 0.876)	–	94.4%	(92.3%, 96.6%)	–
SVM	0.811	(0.782, 0.839)	$p=0.031$	93.3%	(89.2%, 97.5%)	p = 0.324
Logistic Regression	0.717	(0.666, 0.768)	$p=0.009$	86.7%	(79.3%, 94.0%)	$p=0.019$
DNN (small)	0.695	(0.619, 0.771)	$p=0.011$	77.8%	(53.6%, 100.0%)	p = 0.082
DNN w/ all features	0.568	(0.550, 0.587)	$p<0.001$	52.8%	(29.2%, 76.4%)	$p=0.006$
WD Net w/ all features	0.528	(0.479, 0.578)	$p<0.001$	34.4%	(24.1%, 44.8%)	$p<0.001$