Table 3.

Classification results for AutoML (using 4 and 7 features), neural network (using 6 features) and logistic regression (using 3 features).

	AutoML (4 features)	AutoML (7 features)	Neural network	Logistic regression
AUC	0.847 [0.642, 0.975]	0.849 [0.675, 0.978]	0.879 [0.716, 0.984]	0.900 [0.786, 0.976]
Accuracy	0.814 [0.667, 0.926]	0.821 [0.704, 0.926]	0.839 [0.704, 0.944]	0.881 [0.778, 0.963]
Kappa	0.450 [− 0.013, 0.786]	0.465 [0.087, 0.757]	0.491 [0.000, 0.847]	0.644 [0.348, 0.899]
Sensitivity	0.565 [0.000, 1.000]	0.578 [0.167, 1.000]	0.577 [0.000, 1.000]	0.692 [0.333, 1.000]
Specificity	0.885 [0.619, 1.000]	0.891 [0.737, 1.000]	0.914 [0.786, 1.000]	0.936 [0.857, 1.000]

All values calculated as means of 100 repetitions (100 cycles) using independent test data. Values in brackets: 95% confidence interval.