Network-based predictive models for artificial intelligence: an interpretable application of machine learning techniques in the assessment of depression in stroke patients

Table 3 Comparison of the characteristics of five ML models

Characteristics	RF	DT	XGBoost	NB	SVM
AUC	0.711 (0.638,0.778)	0.719 (0.643,0.792)	0.746 (0.674,0.810)	0.671 (0.607,0.736)	0.703 (0.627,0.776)
APS	0.334	0.310	0.353	0.230	0.353
Accuracy	0.825	0.837	0.834	0.837	0.837
Specificity	0.968	1.000	0.982	1.000	1.000
Sensitivity/Recall	0.091	0.000	0.073	0.000	0.000
NPV	0.845	0.837	0.845	0.837	0.837
PPV	0.357	NA	0.444	NA	NA
FPR	0.032	0.000	0.018	0.000	0.000
FNR	0.909	1.000	0.927	1.000	1.000
F1 score	0.779	0.762	0.780	0.762	0.762

ML, machine learning; RF, random forest; DT, decision tree; XGBoost, extreme gradient boosting; NB, Naïve Bayesian; SVM, support vector machine; AUC, the area under the curve; APS, average precision score; NPV, negative predictive value; PPV, positive predictive value; FPR, false positive rate; FNR, false negative rate; NA: null

ISSN: 1471-2318