Table 3.

Performance of machine-learning algorithms to detect significant aortic regurgitation

Classifier	Accuracy	Area-under-the-curve	True positive rate	True negative rate
Ensemble	0.91	0.73	0.75	0.92
Random forest	0.91	0.72	0.71	0.91
k-nearest neighbor	0.89	0.55	0.00	0.89
Support vector machine	0.89	0.50	0.11	0.00
Kernel support vector machine	0.89	0.50	0.11	0.00
Naïve Bayes	0.89	0.73	0.50	0.90
Karnel naïve Bayes	0.89	0.73	0.50	0.90
Decision tree	0.89	0.76	0.50	0.94