Table 2.
The parameters of fuzzy models.
| Parameter | Description | Value |
|---|---|---|
| popu.size | Population size | GCCL:30, GBML:10 |
| num.class | Number of classes | For all methods:2 |
| num.labels | Number of linguistic terms | W:11, CHI:5, GCCL:9, GBML:7, SLAVE:7 |
| persen_cross | Probability of crossover | GCCL:0.8, GBML:0.9, SLAVE:0.8 |
| persen_mutant | Probability of mutation | GCCL:0.4, GBML:0.2, SLAVE:0.4 |
| max.gen | Maximum number of generations | GCCL:150, GBML:10, SLAVE:40 |
| type.mf | The type of the shape of the membership function | W:Gaussian, CHI:Triangle |
| type.tnorm | The type of the tnorm | W:min, CHI:min |
| type.snorm | The type of the snorm | W:sum, CHI:max |
| type. implication. func | Type of implication functions | W:Dienes Recher, CHI:Zadeh |
| max.num.rule | Maximum number of rules | GBML:10 |
| p.dcare | A probability of “don’t care” attributes occurred | GBML:0.5 |
| p.gccl | A probability of GCCL process occurred | GBML:0.4 |
| max.iter | Maximum number of iterations | SLAVE:30 |
| k.lower | A lower bound of the noise threshold | SLAVE:0 |
| k.upper | A value between 0 and 1 representing the level of generalization | SLAVE:0.8 |