Research Article
Comparison of Back Propagation Neural Network and Genetic Algorithm Neural Network for Stream Flow Prediction
Table 1
Parameter range of NN architecture.
|
Parameter |
Range | Minimum | Maximum | Optimum (from trial and error) |
| Number of hidden neurons | 2.0 | 25.00 | 10.00 | Learning rate at hidden layer | 0.1 | 0.99 | 00.500 | Learning rate at output layer | 0.1 | 0.99 | 00.550 | Momentum rate | 0.1 | 0.99 | 00.445 | Hidden layer activation constant | 1.0 | 10.00 | 05.500 | Output layer activation constant | 1.0 | 10.00 | 05.500 | Bias | 0.000001 | 0.00001 | 0.000085 |
|
|