Table 3.

Classification performance-three types of compensatory motions

		NC	TLF	TR	SE	Average
k-NN	Precision	0.986	0.987	0.979	0.979	0.983
	Recall	0.993	0.954	0.988	0.983	0.980
	F1-score	0.989	0.970	0.983	0.981	0.981
SVM	Precision	0.989	0.996	0.975	0.967	0.982
	Recall	0.992	0.954	0.992	0.983	0.980
	F1-score	0.990	0.975	0.983	0.975	0.981

NC = Noncompensation, TLF = Trunk lean-forward, TR = Trunk rotation, SE = Shoulder elevation, k-NN = k-nearest neighbor, SVM = support vector machine