Table 3 Model parameters and Calibration.
Model | Parameter | Testing range | Default/not varied |
|---|---|---|---|
Decision tree regression | Minimum samples split | 2 to 10 | 2 |
Maximum depth | 4 to 12 | – | |
K-nearest neighbors’ regression | Number of neighbors | 1 to 10 | 5 |
Weight function | Distance, Uniform | Uniform | |
Support vector regression | Regularization parameter (C) | 0.1 to 100 | 1 |
Epsilon (ε) | 0.01 to 1 | 0.1 | |
Kernel function | RBF, Linear, Polynomial, Sigmoid | RBF | |
Random forest regression | Number of trees | 10 to 200 | 100 |
Maximum depth | 4 to 12 | – |
