Evolution Programming Algorithm
The goal of the evolutionary programming algorithm is to maximize the fit of a collection of candidate solutions in the context of an objective function of the domain. This objective is pursued by using an adaptive model with surrogates of evolutionary processes, in particular hereditary (reproduction with variation) in competition. The representation used for the candidate solutions is directly evaluable by a cost or objective function of the domain.
The representation of candidate solutions must be domain-specific, such as real numbers for continuous optimization of functions. The sample size (BoutSize) for tournament selection during competition is usually between 5% and 10% of the population size. Evolutionary programming traditionally uses only the mutation operator to create new candidate solutions from existing candidate solutions. The crossover operator used in some other evolutionary algorithms is not used in evolutionary programming. Evolutionary programming is concerned with the link between parent and child candidate solutions and is not concerned with surrogates of genetic mechanisms.
Continuous function optimization is a popular application for this approach, where real-valued representations are used with a Gaussian-based mutation operator. The mutation-specific parameters used in applying the algorithm to continuous function optimization can be tailored in concert with the candidate solutions.
