Evaluating the Seeding Genetic Algorithm

Abstract

This work is largely motivated by the PhD thesis of Cameron Skinner [Skinner, 2009], which features a rigorous mathematical and empirical approach to understanding the underlying mechanism behind the functioning of the genetic algorithm. The results are a new understanding of the algorithm in terms of the notions of discovery, selection and combination. Skinner uses these notions to create a modification to the genetic algorithm: the “seeding” genetic algorithm. We recognise this innovation as an important contribution to the field of evolutionary algorithms, and our focus in this dissertation will be to test its successes, failures, and the scope of its applicability.