Table 2. Classification results of three classifiers using the acceleration sensor placed at dominant wrist.

	KNN (K = 3)			Rotation Forest			Neural Network
Activity	Precision	Recall	F-measure	Precision	Recall	F-measure	Precision	Recall	F-measure
A1	0.927	0.933	0.930	0.974	0.940	0.957	0.890	0.898	0.894
A2	0.910	0.834	0.870	0.922	0.895	0.908	0.857	0.804	0.830
A3	0.892	0.809	0.848	0.855	0.924	0.888	0.756	0.862	0.806
A4	0.875	0.088	0.160	0.984	0.954	0.969	0.957	0.897	0.926
A5	0.154	0.971	0.266	0.996	0.957	0.976	0.964	0.953	0.958
A6	0.984	0.888	0.933	0.987	0.983	0.985	0.991	0.966	0.979
A7	0.994	0.317	0.480	0.989	0.977	0.983	0.987	0.979	0.983
A8	0.904	0.864	0.884	0.907	0.896	0.902	0.786	0.855	0.819
A9	0.945	0.820	0.878	0.954	0.852	0.900	0.961	0.703	0.812
A10	0.983	0.828	0.899	0.961	0.939	0.950	0.937	0.929	0.933
A11	0.898	0.811	0.852	0.845	0.962	0.900	0.830	0.928	0.876
A12	0.931	0.190	0.316	0.963	0.915	0.939	0.903	0.915	0.909
Average	0.890	0.688	0.708	0.941	0.939	0.939	0.900	0.895	0.896