Table 1 Parameters adjustment to perform the LMBNNs.
From: A neural network computational structure for the fractional order breast cancer model
Index | Settings |
|---|---|
Activation function | Log-sigmoid |
Fitness goal (MSE) | 0 |
Minimum gradient | 10–07 |
Maximum Mu values | 1010 |
Hidden neurons | 15 |
Maximum Epochs | 1000 |
Output layer | Single |
Increasing Mu performances | 10 |
Validation statics | 9% |
Authentication fail amount | 6 |
Lower Mu values | 0.1 |
Sample assortment | Arbitrary |
Training data | 82% |
Adaptive parameter | 7 × 10–05 |
Testing data | 9% |
Input layer | Single |
Dataset | Adam technique |
Adam solver implementations | Default |
Stoppage criteria | Default |
