Improvement of Roller Bearing Diagnosis with Unlabeled Data Using Cut Edge Weight Confidence Based Tritraining

Table 5

Classification error rate of initial and final hypothesis and the corresponding improvements of C-tritraining under different unlabeling rate with suspect principle value set to 4.

Case	Unlabeling rate of 0.8			Unlabeling rate of 0.6			Unlabeling rate of 0.4			Unlabeling rate of 0.2
	Initial	Final	Improvement	Initial	Final	Improvement	Initial	Final	Improvement	Initial	Final	Improvement
Case 1	0.1670	0.1367	18.14%	0.0990	0.0857	13.43%	0.0613	0.06	2.12%	0.0500	0.0487	2.6%
Case 2	0.4589	0.3400	25.91%	0.3274	0.28	14.48%	0.2495	0.2211	11.38%	0.2453	0.2200	10.31%