Research Article
New Visual Expression of Anime Film Based on Artificial Intelligence and Machine Learning Technology
Table 1
Hyperparameter combinations and corresponding accuracy of some convolutional neural networks.
| | Batch size | Number of iterations | Convolutional layers | Learning rate | Dropout | Accuracy |
| | 200 | 200 | 3 | 0.001 | 0.75 | 25.55% | | 100 | 200 | 3 | 0.0008 | 0.75 | 14.28% | | 200 | 200 | 3 | 0.0008 | 0.75 | 30.22% | | 200 | 200 | 3 | 0.0008 | 0.2 | 3% | | 200 | 200 | 7 | 0.001 | 0.75 | 31.22% | | 100 | 200 | 7 | 0.0008 | 0.75 | 44.98% | | 200 | 200 | 7 | 0.0008 | 0.75 | 48.98% | | 200 | 200 | 7 | 0.0008 | 0.2 | 6.55% | | 200 | 200 | 11 | 0.001 | 0.75 | 44.55% | | 100 | 200 | 11 | 0.0008 | 0.75 | 53.22% | | 200 | 200 | 11 | 0.0008 | 0.2 | 14.99% | | 200 | 200 | 11 | 0.0008 | 0.75 | 57.55% |
|
|