Table 1.

Details of performance in the classification of samples from the reading task.

Feature Set	ML Model	Training Accuracy	Test Accuracy	Sensitivity	Specificity	F1-Score
MFCCs	Cubic SVM	86.4%	86.3%	0.86	0.86	0.86
	Fine KNN	89.4%	90.3%	0.90	0.91	0.90
	Wide neural network	88.8%	90.7%	0.93	0.87	0.90
GTCCs	Cubic SVM	87.0%	86.7%	0.86	0.88	0.87
	Fine KNN	87.9%	87.7%	0.88	0.87	0.88
	Wide neural network	87.5%	90.3%	0.93	0.88	0.90
MFCCs + GTCCs	Cubic SVM	90.0%	88.7%	0.90	0.88	0.88
	Fine KNN	89.8%	92.3%	0.93	0.91	0.92
	Wide neural network	90.3%	92.3%	0.92	0.93	0.92