Table 2.
Parameter grids used by grid-search algorithm for hyperparameter optimization.
| Model | Parameter grid | No. of combinations |
|---|---|---|
| MLPR |
hidden layer size × activation function × learning rate where, hidden layer size = {(50), (1 0 0), (1 5 0), (2 0 0), (2 5 0), (3 0 0), (50,10), (100,20), (150, 30), (200,40)}, activation function = {ReLU, logistic}, learning rate = {0.01, 0.001, 0.0001, 0.00001} |
80 |
| SVR (kernel = RBF) |
C × gamma × epsilon where, C = {5, 10, 15, 20} gamma = {0.1, 0.01, 0.001, 0.0001} epsilon = {0.001, 0.01, 0.1, 0.5, 0.8} |
80 |
| SVR (kernel = polynomial) |
C × gamma × epsilon × degree × coefficient where, C = {1, 5, 10, 15} gamma = {0.1, 0.01, 0.001} epsilon = {0.01, 0.1, 0.5, 0.8} degree = {2, 3} coefficient = {1, 2, 3, 4} |
384 |
| SVR (kernel = linear) |
C × epsilon where, C = {1, 5, 10, 15} epsilon = {0.01, 0.1, 0.5, 0.7} |
16 |
Abbreviations: MLPR = Multi-Layer Perceptron Regression; RBF = Radial Basis Function; ReLU = Rectified Linear Unit; SVR = Support Vector Regression.