Construction and validation of a predictive model for intracardiac thrombus risk in patients with dilated cardiomyopathy: a retrospective study

Table 3 Delong test for AUC comparison of different models

Datasets	Model	AUC (95%CI)	Statistic	P
Training set	Logistic Regression	0.854 (0.811–0.896)	-5.57	< 0.001
	SVM	0.769 (0.715–0.824)	-6.89	< 0.001
	Random Forest	0.917 (0.887–0.947)	-3.97	< 0.001
	XGBoost	0.947 (0.924–0.969)	Ref
Testing set	Logistic Regression	0.823 (0.733–0.914)	-2.59	0.01
	SVM	0.745 (0.645–0.845)	-3.39	0.001
	Random Forest	0.880 (0.815–0.945)	-2.44	0.015
	XGBoost	0.922 (0.865–0.979)	Ref

Note: SVM, support vector machine; XGBoost, eXtreme Gradient Boosting; AUC, the area under the receiver operating characteristic curve; CI, confidence interval

ISSN: 1471-2261