Hands-On Genetic Algorithms with Python

Evolution strategies

Evolution strategies (ES) are a kind of genetic algorithm that emphasizes mutation rather than crossover as the evolutionary facilitator. The mutation is adaptive, and its strength is learned over the generations. The selection operator in evolution strategy is always rank-based rather than done using actual fitness values. A simple version of this technique is called (1 + 1). It includes only two individuals – a parent and its mutated offspring. The best of them continues to be the parent of the next mutated offspring. In the more general case, called (1 + λ), there is one parent and λ mutated offspring, and the best of the offspring continues to be the parent of the next λ offspring. Some newer variations of the algorithm ...