Mutation

In evolution and genetic algorithms, the mutation operator determines how the original individuals will be mutated. Each individual is an instance of Backdoor, and contains a bit mask that defines it. When mutating, the values of the selected bits in the mask are flipped.

class Mutation:
    def mutate(individual: Backdoor) -> Backdoor

The mutate method returns a new individual, which is the result of applying the mutation operator to the passed individual.

Uniform mutation

When mutating, the value of each bit with a probability equal to flip_scale/mask_length flips to the opposite, where mask_length is a length of bit mask. The random number generator is initialized with the passed random_seed.

from algorithm.module.mutation import Uniform

 selection = Uniform(
     flip_scale: float = 1.0,
     random_seed: Optional[Int] = None
 )

Doer mutation

См. статью (добавить ссылку позже).

from algorithm.module.mutation import Doer

 selection = Doer(
     beta: int = 3,
     random_seed: Optional[Int] = None
 )

Other algorithm modules

• Crossover

• Selection