Research Article
An Intelligent Fault Diagnosis Based on Adversarial Generating Module and Semi-supervised Convolutional Neural Network
Table 8
Runtime of different diagnostic models.
| Labeled data | Models’ runtime (s/epoch) | CNN | SVM | LapSVM | LapRLS | MRFCD-SSCNN |
| 100 | Training time | 3.161 | 1.246 | 8.079 | 6.337 | 8.136 | Testing time | 0.828 | 0.359 | 2.091 | 1.641 | 2.102 |
| 200 | Training time | 3.209 | 1.305 | 8.164 | 6.425 | 8.225 | Testing time | 0.891 | 0.395 | 2.126 | 1.707 | 2.172 |
| 400 | Training time | 3.473 | 1.346 | 8.217 | 6.509 | 8.341 | Testing time | 0.913 | 0.451 | 2.185 | 1.726 | 2.231 |
|
|