The Population Sizing Problem: Revisited

Zdeněk Konfršt


Evolutionary Algorithms, GAs, Population Sizing, GRM


To estimate the correct size of population, estimations of population sizing have been used in genetic algorithms (GAs). The estimation considers a test function being optimized, a representation of individuals, and a character of used operators. By means of the estimation model the right population size is identified while taking into account the final overall quality of individuals. As far as the comparison of theoretical results with real runs of GAs is concerned, several factors which influence mutual correspondence have been faced. The main factors, such as GA operators and values of GA parameters are addressed in the paper. Major issues are tested against the estimation model, some irregularities of experiments are discussed, and clearly new ways of next research come up.

Important Links:

Go Back