Research Article
Fault Diagnosis of Car Engine by Using a Novel GA-Based Extension Recognition Method
Table 4
Diagnosis performances of methods compared.
| Method | Training time | Accuracy rate (%) |
| Proposed method | 1000 | 98% | -means method | N/A | 85% | MNN-I (8-8-16) | 1000 | 62% | MNN-II (8-10-16) | 1000 | 80% | MNN-III (8-15-16) | 1000 | 95% |
|
|