An Efficient Automatic Gait Anomaly Detection Method Based on Semisupervised Clustering

Table 3

CCR/ACCR values for AGD-SSC, AGD-SSC-NBC, IF, and LOF in test case II.

Seq. No.	Metric	AGD-SSC; mean (st. dev.)	AGD-SSC-NBC; mean (st. dev.)	IF mean (st. dev.)	LOF mean (st. dev.)
Seq. 1	CCR	0.9286 (0.000)	0.9286 (0.000)	0.8869 (0.0106)	0.8929 (0.0000)
	ACCR	0.9167 (0.000)	0.9167 (0.0000)	0.5444 (0.0416)	0.5000 (0.0000)
Seq. 2	CCR	0.9119 (0.0045)	0.9125 (0.0071)	0.9327 (0.0088)	0.8571 (0.0000)
	ACCR	1.0000 (0.0000)	1.0000 (0.0000)	0.7611 (0.0356)	0.4167 (0.0000)
Seq. 3	CCR	0.9815 (0.0035)	0.9821 (0.0000)	0.8810 (0.0084)	0.8929 (0.0000)
	ACCR	0.9167 (0.0000)	0.9167 (0.0000)	0.5278 (0.0084)	0.5000 (0.0000)
Seq. 4	CCR	0.9482 (0.0071)	0.9476 (0.0064)	0.8839 (0.0120)	0.8929 (0.0000)
	ACCR	0.9167 (0.0000)	0.9167 (0.0000)	0.5361 (0.0513)	0.6667 (0.0000)
Seq. 5	CCR	0.9464 (0.0000)	0.8393 (0.0000)	0.9363 (0.0136)	0.8750 (0.0000)
	ACCR	0.9167 (0.0000)	0.9167 (0.0000)	0.7194 (0.0504)	0.4167 (0.0000)
Seq. 6	CCR	0.9821 (0.0000)	0.9464 (0.0000)	0.8905 (0.0110)	0.8750 (0.0000)
	ACCR	0.9167 (0.0000)	0.9167 (0.0000)	0.5722 (0.0515)	0.5000 (0.0000)
Seq. 7	CCR	0.9643 (0.0000)	0.9625 (0.0084)	0.9006 (0.0100)	0.9286 (0.0000)
	ACCR	1.0000 (0.0000)	0.9889 (0.0356)	0.6222 (0.0416)	0.6667 (0.0000)
Seq. 8	CCR	0.9643 (0.0080)	0.9387 (0.0110)	0.9083 (0.0089)	0.8929 (0.0000)
	ACCR	0.9167 (0.0215)	0.9139 (0.0150)	0.6694 (0.0402)	0.5833 (0.0000)
Seq. 9	CCR	0.9821 (0.0000)	0.9286 (0.0000)	0.9036 (0.0182)	0.8929 (0.0000)
	ACCR	1.0000 (0.0000)	1.0000 (0.0000)	0.8972 (0.0466)	0.7500 (0.0000)
Seq. 10	CCR	0.9815 (0.0032)	0.9452 (0.0045)	0.9446 (0.0141)	0.9286 (0.0000)
	ACCR	1.0000 (0.0000)	1.000 (0.0000)	0.7694 (0.0513)	0.6667 (0.0000)
Total	CCR	0.9591	0.9332	0.9068	0.8929
	ACCR	0.9500	0.9486	0.6619	0.5667