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

Table 1

Gait recognition accuracy for gait recognition mode A 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	47.10  8.80	45.31  7.59	41.53  5.36	47.00  8.92	38.65  6.95	44.81  6.54
IMF2	47.09  8.82	49.64  7.71	44.71  5.02	47.07  8.81	39.42  8.02	49.60  6.37
IMF3	47.13  8.78	55.37  7.12	50.19  5.07	47.17  8.64	39.39  5.95	55.93  5.70
IMF4	47.23  8.67	74.74  4.85	72.09  4.87	47.40  8.43	44.91  7.78	78.70  4.15
IMF5	49.84  7.68	99.14  0.60	95.19  1.63	50.42  7.69	64.91  10.26	98.67  0.70
IMF6	92.06  5.44	99.65  0.21	99.00  0.58	72.54  10.96	77.79  9.33	99.63  0.26
IMF7	99.47  0.33	99.65  0.22	99.48  0.37	96.97  4.08	87.00  8.24	99.64  0.22
IMF8	99.54  0.29	99.66  0.22	99.65  0.29	99.20  0.59	91.38  3.77	99.66  0.27