Table 6.
The selected hyperparameters of a CNN using three different runs of GWO optimizer.
| GWO parameters | Selected CNN parameters | Fitness score | |||||
|---|---|---|---|---|---|---|---|
| LR | EP | BS | N | M | |||
| Run1 | Population size = 50 | 0.001 | 60 | 64 | 85 | 40 | 0.94632 |
| No of iterations = 10 | |||||||
| Run2 | Population size = 60 | 0.08 | 60 | 64 | 168 | 32 | 0.91252 |
| No of iterations = 20 | |||||||
| Run3 | Population size = 70 | 0.04 | 80 | 50 | 85 | 32 | 0.91948 |
| No of iterations = 30 | |||||||
Where LR Learning Rate, EP Epochs, BS Batch Size, N Number of filters in the convolutional layer, M Number of neurons in the dense layer.