Prediction of postoperative visual cognitive impairment using graph theory and machine learning based on resting-state brain networks

Table 3 Comparison of prediction results of different models

Validation methods	Classifier	AUC	ACC	SEN	SPE
Cross-validation set	SRC	0.918 ± 0.056	0.898 ± 0.071	0.900 ± 0.223	0.897 ± 0.056
Independent test set	SRC	0.877	0.840	0.833	0.842
	RF	0.781	0.840	0.667	0.895
	SVM	0.825	0.800	0.667	0.842
	XGBoost	0.860	0.840	0.333	1.000

± denotes the standard deviation of performance scores in cross-validation
ACC, accuracy; AUC, the area under the receiver operating characteristic curve; RF, random forest; SEN, sensitivity; SPE, specificity; SRC, sparse representation classifier; SVM, support vector machine