Table 4 Hyperparameter setting of MEP and GEP algorithms.
Algorithm | Parameter | Search space | Optimal value |
|---|---|---|---|
GEP parameters | Constants per gene | 2 to 20 | 10 |
No. of genes | 2 to 10 | 5 | |
Linking function | Addition, subtraction, multiplication division | Addition | |
Head size | 5 to 50 | 20 | |
No. of chromosomes | 5 to 150 | 80 | |
Functions | +, −, ×, ÷, sqrt, ln, \(\:{x}^{2}\) | +, −, ×, ÷, sqrt, ln, \(\:{x}^{2}\) | |
MEP parameters | Number of Generations | 500 to 2000 | 1500 |
Subpopulation size | 10 to 200 | 100 | |
Runs | 2 to 15 | 10 | |
Crossover probability | 0 to 1 | 0.8 | |
No. of subpopulations | 100 to 2000 | 1000 | |
Functions | +, −, ×, ÷, sqrt, \(\:{x}^{2}\) | +, −, ×, ÷, sqrt, \(\:{x}^{2}\) | |
Code length | 20 to 200 | 100 |
