Table 6.

LOSO evaluation results of the supervised classification models with a data representation of 64-bin FFT and overlap 75%.

Classifier	Sensitivity	Specificity	GM	AUC	EER (%)
SVM	0.832	0.831	0.832	0.876	16.9
AdaBoost	0.827	0.828	0.827	0.905	17.2
Random Forest	0.849	0.847	0.848	0.921	15.3
CNN-MPL	0.844	0.844	0.844	0.920	15.6
CNN-LSTM	0.849	0.849	0.849	0.923	15.1