Table 5. Accuracy of five classification models on nine kinematic attributes selected using the proposed feature selection method.

Classifier	Sensitivity	Specificity	Accuracy
Naive Bayes	78.6%	81.3%	80.0%
SVM (RBF)	78.6%	87.5%	83.3%
SVM (Linear)	85.7%	87.5%	86.7%
Decision Tree	57.1%	75.0%	66.7%
Random Forest	71.4%	68.8%	70.0%
Avg.	74.3%	80.0%	77.3%