Classification of vasovagal syncope from physiological signals on tilt table testing

Table 3 The outcome summary of the proposed models by D1 and D2

Classifiers	Accuracy	Sensitivity	Specificity	Precision	F1 score	ROC_AUC
Classification results of FI using D1
SVM	90.5	87.0	92.8	88.7	87.9	95.4
KNN	87.5	75.9	95.2	91.1	82.8	95.7
GNB	86.1	75.9	92.8	87.2	81.2	92.9
MNB	86.9	75.9	93.9	89.1	82.0	91.7
LR	86.1	75.9	92.8	87.2	81.2	93.6
RF	87.6	79.6	92.8	87.8	83.5	94.9
Classification results of FI using D2
SVM	89.1	83.3	92.8	88.2	85.7	93.2
KNN	88.3	83.3	91.6	86.5	84.9	96.3
GNB	83.9	66.7	95.2	90.0	76.6	91.7
MNB	87.6	75.9	95.2	91.1	82.8	91.7
LR	85.4	74.1	92.8	86.9	80.0	93.1
RF	86.1	75.9	92.8	87.2	81.2	94.9
Classification results of GA using D1
SVM	89.8	81.5	95.2	91.7	86.3	94.9
KNN	84.7	75.9	90.4	83.7	79.6	95.5
GNB	84.7	77.8	89.2	82.4	80.0	92.8
MNB	86.1	79.6	90.4	84.3	81.9	92.6
LR	86.1	81.5	89.2	83.0	82.2	93.9
RF	82.5	75.9	86.7	78.8	77.4	97.3
Classification results of GA using D2
SVM	86.1	87.0	85.5	79.7	83.2	99.7
KNN	76.6	61.1	86.7	75.0	67.3	90.9
GNB	78.1	72.2	81.9	72.2	72.2	89.3
MNB	79.6	72.2	84.3	75.0	73.6	87.8
LR	89.8	87.0	91.6	87.0	87.0	98.7
RF	83.9	74.1	90.4	83.3	78.4	97.3
Classification results of RFE using D1
SVM	83.2	75.9	87.9	80.4	78.1	88.5
KNN	77.4	68.5	83.1	72.5	70.5	90.8
GNB	79.6	64.8	89.2	79.5	71.4	87.7
MNB	84.7	81.5	86.7	80.0	80.7	88.3
LR	81.8	70.4	89.2	80.9	75.2	88.5
RF	83.2	74.1	89.2	81.6	77.7	94.0
Classification results of RFE using D2
SVM	83.2	75.9	87.9	80.4	78.1	88.5
KNN	77.4	68.5	83.1	72.5	70.5	90.8
GNB	79.6	64.8	89.2	79.5	71.4	87.7
MNB	84.7	81.5	86.7	80.0	80.7	88.3
LR	81.8	70.4	89.2	80.9	75.2	88.5
RF	83.2	74.1	89.2	81.6	77.7	94.0

FI, feature importance; GA, genetic algorithm; RFE, recursive feature elimination; D1, K-nearest neighbors (KNN) imputation dataset; D2, mean imputation dataset; SVM, support vector machine; GNB, Gaussian naïve Bayes; MNB, multinomial naïve Bayes; LR, logistic regression; RF, random forest

ISSN: 1475-925X