Gait Recognition for Human-Exoskeleton System in Locomotion Based on Ensemble Empirical Mode Decomposition

Table 5

Gait recognition accuracy for gait recognition mode C based on the IMFs of the original signal for the 14 subjects who wore exoskeletons to walk (mean  SD%).

Component	SVM	Kmeans	Decision tree	Logistic regression	Naive Bayes	Random forest
IMF1	32.14  12.32	33.08  4.23	33.73  2.78	31.77  10.56	31.30  5.86	32.86  3.52
IMF2	32.61  13.14	36.33  3.85	33.92  3.34	32.30  11.05	31.44  7.39	34.41  4.39
IMF3	32.53  13.30	37.09  5.00	34.45  3.22	32.71  10.26	31.11  5.36	35.01  4.08
IMF4	30.08  12.06	35.67  4.84	37.00  4.99	31.62  9.19	36.87  6.33	37.90  6.11
IMF5	32.45  10.20	42.35  8.11	41.41  7.34	31.86  6.24	46.13  9.66	43.40  8.38
IMF6	47.53  10.52	51.25  11.39	50.92  7.65	40.57  6.55	60.54  11.20	55.60  8.28
IMF7	50.59  10.91	58.46  11.73	48.44  11.60	48.94  13.30	61.02  14.07	54.75  11.94
IMF8	38.65  15.24	69.18  10.20	50.49  15.91	61.25  13.64	43.37  14.80	51.35  17.74