Table 3.

Performance evaluation for discrimination of operator experience. Values are mean and standard deviation computed over the five-fold cross-validation rounds.

Extracted Features	Learning Method	Accuracy NQ	Accuracy XP	Overall Accuracy
Temporal	SVM	0.69 ± 0.04	0.59 ± 0.04	0.66 ± 0.03
Spectral	SVM	0.70 ± 0.01	0.51 ± 0.05	0.63 ± 0.03
Time-Frequency	SVM	0.81 ± 0.04	0.63 ± 0.05	0.75 ± 0.04
TEPR Sequence	1D CNN	0.72 ± 0.15	0.38 ± 0.24	0.62 ± 0.06
Scalogram	ResNet18 2D CNN	0.57 ± 0.16	0.55 ± 0.14	0.58 ± 0.03