Skip to main content
. 2020 Mar 29;20(7):1895. doi: 10.3390/s20071895

Table 5.

LOSO evaluation results of the novelty detection models. 64-bin FFT and overlap 75%.

Classifier Sensitivity Specificity GM AUC EER (%)
OneClass-SVM 0.832 0.592 0.702 0.712 40.8
MLP Autoencoder 0.694 0.694 0.694 0.728 30.6