Table 6. Classification results of three classifiers using the acceleration sensor placed on the dominant ankle.

	KNN (K = 3)			Rotation Forest			Neural Network
Activity	Precision	Recall	F-measure	Precision	Recall	F-measure	Precision	Recall	F-measure
A1	0.986	0.935	0.960	0.995	0.958	0.976	0.981	0.935	0.957
A2	0.870	0.821	0.845	0.963	0.883	0.921	0.786	0.734	0.759
A3	0.777	0.533	0.632	0.838	0.822	0.830	0.635	0.628	0.631
A4	0.895	0.071	0.132	0.968	0.958	0.963	0.925	0.829	0.875
A5	0.103	0.978	0.187	0.981	0.957	0.969	0.985	0.957	0.971
A6	0.962	0.216	0.353	0.949	0.981	0.965	0.887	0.932	0.909
A7	0.976	0.073	0.136	0.962	0.946	0.954	0.800	0.896	0.846
A8	0.967	0.820	0.887	0.951	0.918	0.934	0.950	0.909	0.929
A9	0.975	0.729	0.834	0.965	0.874	0.917	0.962	0.804	0.876
A10	0.867	0.586	0.699	0.853	0.921	0.885	0.778	0.843	0.809
A11	0.815	0.605	0.694	0.787	0.899	0.839	0.719	0.821	0.767
A12	1.000	0.162	0.279	1.000	0.937	0.967	0.978	0.923	0.949
Average	0.865	0.525	0.558	0.925	0.921	0.922	0.846	0.840	0.841