Table 6 Results of all binary classification models with their respective cross-validation and their accuracy of the balanced database using SMOTE
Epileptic | Without Epileptic | |||
|---|---|---|---|---|
Accuracy [%] | Cross Validation [%] | Precision [%] | Precision [%] | |
ETC | 98.42 | 98.40 ± 0.3 | 97.70 | 99.10 |
RFC | 97.74 | 98.00 ± 1.0 | 96.40 | 99.10 |
GB | 98.18 | 98.20 ± 1.0 | 97.50 | 98.80 |
DTC | 93.51 | 93.00 ± 0.5 | 93.50 | 93.90 |
MLP | 98.59 | 99.00 ± 0.1 | 98.00 | 99.60 |
KNN | 99.59 | 100 ± 0.1 | 99.60 | 99.60 |
SGD | 66.90 | 66.50 ± 1.0 | 93.60 | 60.50 |
SVM | 98.32 | 98.50 ± 0.3 | 97.60 | 99.00 |
CNN+Fully | 99.78 | 99.80 ± 0.1 | 99.80 | 99.80 |
CNN+Capsule-Net | 99.92 | 99.90 ± 0.4 | 100 | 99.80 |
CNN+TF | 99.76 | 99.80 ± 1.0 | 99.50 | 100 |
CNN+Tf+FULLY | 99.57 | 99.40 ± 1.0 | 100 | 99.20 |
CNN+TF+Capsule-Net | 99.89 | 99.90 ± 0.4 | 99.80 | 100 |