Table 3 The optimal values of user-defined parameters for RF, SVR, ANN, and MLR algorithms
Classifiers used | Classifiers used |
|---|---|
Multilinear regression (MLR) | PH, PeD, IL1, IL2 |
Neural network (ANN) | Learning rate = 0.2, Momentum = 0.1, Iteration = 2000, Hidden layer = 3–9–8 |
Random forest (RF) | K = 2, M = 4, I = 100 |
Support vector regression (SVR) | Kernel = rbf, Gamma = 0.004, C = 0.1 |