Genetic Algorithm with Histogram Construction Technique

This paper proposes a new niching technique for distributed genetic algorithm (DGA). In DGA quick evolution takes place because of independent evolution of subpopulation. To improve performance of DGA care need to be taken for distribution of population in to subpopulations. A proper distribution leads to efficient GA and good solutions. The new technique uses the traditional histogram construction technique for subpopulation formation. Experiments show that the proposed algorithm, GAHisto (genetic algorithm with histogram technique) when tested with unimodal and multimodal test function has given good results.