Biomedical Engineering Reference
In-Depth Information
3.1.3 Crossover Operation
The crossover operation takes two individuals as input expressions and selects a
random point and interactions the subexpression at the rear point. While imple-
menting the crossover operation, there may be chance of having errors like a
crossover may produce individuals in which an attribute may be involved more than
once and the offspring may be the same as that of the parent expression with a
probability of 1. To overcome these errors, we used the below-mentioned tech-
niques while performing crossover operation.
3.1.4 Mutation Operation
After crossover mutation takes place, it is the genetic operator that introduces
genetic diversity in the population.
Mutation takes place whenever the population tends to become homogeneous
due to repeated use of reproduction and crossover operator. This can produce the
entirely new gene values being added to the gene pool. With this new gene values,
the GA may be able to produce better solution than the previous one.
Steps from mutation with attribute modi
cation:
(i) Select a random point within the attribute range.
(ii) Form the new rule by changing the selected attribute value.
(iii) Compute the
tness of the newly formed rule; if the
tness is greater than the
parent, then add this rule to the initial population.
(iv) Repeat the above process for required number of times.
Sequential steps in mutation with attribute removal:
(i) Select a random point within the attribute range.
(ii) Form the new rule by removing the attribute value at the selected point.
(iii) Compute the
tness of the new rule is greater
than the parent, then add this chromosome to the initial population.
(iv) Repeat this process for required number of times.
tness of the new rule. If the
3.2 Stopping Criteria
If all the records in the dataset belong to the same class
If dataset has all similar records in the attribute list
If the dataset is empty.
Search WWH ::
Custom Search