Gesture Recognition from Data Streams of Human Motion Sensor Using Accelerated PSO Swarm Search Feature Selection Algorithm

Table 3

The performance results of classifying sensor data using traditional algorithms.

Traditional slassifier/feature selection	Accuracy	Kappa	TP	FP	Precision	Recall	-measure	Model building time (s)	Preprocessing time (s)	# selected features
HyperPipe (HP)
Original	77.6822	0.6666	0.777	0.084	0.803	0.777	0.772	0	0	50
Cfs	73.6661	0.6159	0.737	0.078	0.817	0.737	0.733	0	0	23
FS-PSO	78.0264	0.672	0.78	0.083	0.814	0.78	0.774	0	2	36
FS-APSO	78.3706	0.6764	0.784	0.084	0.817	0.784	0.777	0	2	32
Naive Bayes (NB)
Original	66.2077	0.5257	0.662	0.085	0.732	0.662	0.68	0.08	0	50
Cfs	69.7074	0.569	0.697	0.08	0.744	0.697	0.707	0.01	0	23
FS-PSO	79.5181	0.6953	0.795	0.076	0.791	0.795	0.789	0.01	9	15
FS-APSO	78.7149	0.6835	0.787	0.079	0.785	0.787	0.78	0	9	12
BayesNet (BN)
Original	82.2146	0.7417	0.822	0.052	0.835	0.822	0.828	0.37	0	50
Cfs	82.7309	0.7459	0.827	0.057	0.828	0.827	0.827	0.04	0	23
FS-PSO	85.0832	0.78	0.851	0.05	0.849	0.851	0.849	0.05	17	25
FS-APSO	86.1733	0.7952	0.862	0.049	0.859	0.862	0.86	0.05	12	19
Decision tree (DT)
Original	87.4355	0.8147	0.874	0.044	0.875	0.874	0.874	0.2	0	50
Cfs	87.6649	0.8178	0.877	0.044	0.876	0.877	0.876	0.11	0	23
FS-PSO	89.0419	0.8379	0.89	0.042	0.889	0.89	0.89	0.11	36	23
FS-APSO	89.3287	0.842	0.893	0.041	0.891	0.893	0.892	0.07	20	13
Random forest (RF)
Original	91.1073	0.8654	0.911	0.048	0.911	0.911	0.904	0.21	0	50
Cfs	92.7711	0.8916	0.928	0.036	0.927	0.928	0.924	0.09	0	23
FS-PSO	91.3941	0.87	0.914	0.046	0.913	0.914	0.908	0.08	34	21
FS-APSO	93.4596	0.9024	0.935	0.032	0.934	0.935	0.933	0.11	31	13
Support vector machine (SVM)
Original	75.2151	0.5961	0.752	0.155	0.678	0.752	0.659	2.05	0	50
Cfs	75.961	0.6123	0.76	0.143	0.742	0.76	0.678	0.96	0	23
FS-PSO	76.0757	0.6145	0.761	0.141	0.738	0.761	0.682	0.39	349	14
FS-APSO	77.2806	0.6359	0.773	0.133	0.748	0.773	0.706	0.42	296	13
Neural network (NN)
Original	90.0172	0.8512	0.9	0.042	0.898	0.9	0.897	24.44	0	50
Cfs	89.0419	0.8369	0.89	0.041	0.889	0.89	0.888	8.35	0	23
FS-PSO	89.2714	0.8403	0.893	0.043	0.889	0.893	0.89	14.71	5212	35
FS-APSO	90.3614	0.8565	0.904	0.04	0.901	0.904	0.901	12.49	4685	32