Research Article
Modified Kernel Marginal Fisher Analysis for Feature Extraction and Its Application to Bearing Fault Diagnosis
Table 3
The average recognition rates (%) of KNN for various training sample sizes each class.
| | Training sample size | KPCA
() | KFDA
() | SKLPP
() | KMFA ()
() | RKMFA
() | MKMFA
() |
| | 10 | 56.11 | 78.33 | 81.11 | 63.61 () | 97.22 () | 98.93 | | 30 | 82.14 | 79.29 | 89.17 | 75 () | 97.86 () | 99.29 | | 50 | 87.00 | 81.00 | 91.50 | 88.5 () | 99.64 () | 100.00 | | 70 | 96.67 | 87.50 | 100.00 | 100 () | 100.00 () | 100.00 | | 90 | 100.00 | 100.00 | 100.00 | 100 () | 100.00 () | 100.00 |
|
|
