Strategy parameters

parameters, which determine the behavior of the EA or define the EA in more detail. Examples are population size, mutation rates- and crossover rates. Exogenous and endogenous strategy parameters are distinguished. Endogenous strategy parameters are adapted during evolution (e.g. mutation strength), whereas exogenous strategy parameters (e.g. number of parents $\mu$, number of offspring $\lambda$, mixing number $\rho$, learning parameter) are kept constant. The adaptation of the endogenous strategy parameters is accomplished by (a) statistical rule-based techniques (e.g. 1/5-rule, derandomized ES), (b) Meta-EA, or (c) self-adaptation.