Cerebral blood flow alterations in migraine patients with and without aura: An arterial spin labeling study

Table 2 Performance metrics from machine learning–based models

Machine learning-based models	AUC (%)	Accuracy (%)	Sensitivity (%)	Specificity (%)
SVM	84.1 ± 7.8	81.3 ± 6.2	87.6 ± 8.9	74.0 ± 6.9
KNN	67.2 ± 5.8	64.5 ± 7.0	70.9 ± 6.0	60.4 ± 4.7
RF	77.4 ± 6.1	74.2 ± 6.7	75.8 ± 6.3	72.1 ± 5.2
NB	64.2 ± 5.4	63.0 ± 4.8	68.3 ± 5.2	59.4 ± 4.1
LDA	76.3 ± 6.7	73.7 ± 5.2	77.0 ± 5.8	70.2 ± 7.1

AUC: area under curve, KNN: k-nearest neighbor, LDA: linear discriminant analysis, NB: naive bayes, RF: random forest, SVM: support vector machine
Values are represented as the mean ± standard deviation

ISSN: 1129-2377