Table 2 Hyper-parameters and run time of developed models.
From: Compositional modeling of gas-condensate viscosity using ensemble approach
Model | Hyper-parameters | Run time (min) |
|---|---|---|
Ensemble methods | SVR + K-NN C = 200, \(\varepsilon = 0.00001\) K-neighbours = 2 | 3 |
SVR | C = 200, \(\varepsilon = 0.00001\) | 5 |
K-NN | K-neighbours = 2 | 2 |
MLP-LM | Transfer function = Tansig-Tansig Number of neurons = 10,12 | 80 |
MLP-BR | Transfer function = Tansig-Tansig Number of neurons = 10,12 | 55 |
RBF | Max neuron = 300 Spread = 1.5 | 15 |
