A Genetic Algorithm with Dynamic Population Size

Abstract

This paper introduces a modified version of a simple genetic algorithm (SGA) in which the size of the population changes according to a model inspired from mathematical biology. The primary purpose of this research described is to use the growth of the population to vary the selective pressure through the scaling of the fitness function. The experimental results presented in the paper are used to demonstrate the behavior of the proposed algorithms in comparison with a SGA using fixed size population.