Architecture |
|
Number of Layers | 3 |
Number of Neuron on the Layers | Input 9, Hidden 12, Output 1 |
Activation Function | Tan-Sigmoid, Purelin |
Training Parameters |
|
Learning Rule | Levenberg-Marquardt |
Adaptive Learning Rate | Initial: 0.01, increase: 1.04, decrease: 0.4 |
Momentum Constant | 0.94 |
Sum-Squared Error | 0.004 |
Epochs | 10000 |