CNN+SVM算法流程图
2023-12-06 15:35:02 0 举报
卷积CNN结合支持向量机SVM
作者其他创作
大纲/内容
epoch=epoch+1
Pooling layer3
Complete training
Convolutional layer
Electric submersible pump fault data acquisition
Trained SVM
1.Feature extraction
CNN Feature mapping model
Yes
Test set
2.Normalization
Training set
Training completed
3.Convert to heat map
Validation set
Fixed
Pooling layer2
Estabilsh improved CNN-SVM model Parameter initialization
SVM training
Output result:Test set accuracy
Back propagation update Error
Pooling layer1
Fault diagnosis
No
Convergence?
Softmax function
Fully connected layer
Output feature of Pooling layer3
CNN model training
Overlap sampling
0 条评论
下一页