Table 2.

Summary of model performances in terms of discrimination ability and accuracy for one run.

Models	ROC (95% CI)	Accuracy (95% CI)
Gradient boosting	0.669 (0.608, 0.730)	0.697 (0.648, 0.743)
Random forest	0.677 (0.617, 0.736)	0.729 (0.681, 0.773)
SVM	0.658 (0.596, 0.720)	0.729 (0.681, 0.773)
ANN	0.673 (0.611, 0.735)	0.721 (0.672, 0.765)

All models performed similarly across ROC and accuracy measures. See Supplementary Table 1 for performance across all runs.

Abbreviations: ANN, artificial neural network; CI, confidence interval; ROC, receiver operating characteristic; SVM, support vector machine.