An Efficient and Fast Model Reduced Kernel KNN for Human Activity Recognition

Table 2

The comparison results among three models.

Data	KNN			K-KNN			RK-KNN
	Accuracy (%)	SD	Time (s)	Accuracy (%)	SD	Time (s)	Accuracy (%)	SD	Time (s)
Iris	92.66	2.56	0.03	93.12	4.05	0.27	95.20	2.73	0.05
Heart	88.51	3.13	0.21	88.57	3.49	1.61	91.66	2.56	0.27
Breast	81.49	4.34	0.19	82.04	4.36	1.49	82.77	4.37	0.49
Diabetes	83.64	2.34	2.62	83.85	2.46	44.19	84.32	2.23	11.67
Seed	95.21	2.39	0.09	95.53	3.02	0.73	97.36	1.37	0.14
Thyroid	97.41	2.20	0.08	97.48	2.09	0.79	97.67	1.98	0.31
Banana	93.15	0.67	31.67	94.14	0.63	177.98	94.67	0.61	21.49
Image	92.09	1.25	407.32	92.11	2.25	2396.5	99.04	0.97	422.5
Ringnorm	92.14	1.15	411.79	92.23	0.61	2221.83	92.83	0.98	521.13
Splice	92.95	0.77	413.50	93.31	0.66	2255.67	93.51	0.65	350.21
HAPT	91.02	0.67	500.43	91.02	0.58	4334.26	91.60	0.32	495.78
Smartphone	91.21	1.05	170.24	92.38	0.38	2325.65	92.67	0.49	129.20
Average	90.96	1.88	161.51	91.32	2.05	1146.75	92.78	1.61	162.77