From: Multi combination pattern labeling by using deep learning for chameleon rotary machine environment
CNN model | CNN1 | CNN2 | CNN3 |
---|---|---|---|
Filters | [16,128] | [ 32,64 ] | [ 50,100 ] |
convolution kernels | 3 | 5 | 3 |
Activation | Relu | ||
Padding | Same | ||
Drop out | 0 | 0 | 0.25 |
Flatten dense layers | 32 | . | . |
Flatten activation | SoftMax | ||
Optimizer | Adam | Adam | Nadam |
Learning rate | 0.001 | 0.002 | 0.001 |
Loss | Categorical Crossentropy | ||
Batchsize | 16 | ||
Epoch | 100 |