2008 | OriginalPaper | Buchkapitel
Altering Search Rates of the Meta and Solution Grammars in the mGGA
verfasst von : Erik Hemberg, Michael O’Neill, Anthony Brabazon
Erschienen in: Genetic Programming
Verlag: Springer Berlin Heidelberg
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
Adopting a meta-Grammar with Grammatical Evolution(GE) allows GE to evolve the grammar that it uses to specify the construction of a syntactically correct solution. The ability to evolve a grammar in the context of GE means that useful bias towards specific structures and solutions can be evolved during a run. This can lead to improved performance over the standard static grammar in terms of adapting to a dynamic environment and improved scalability to larger problem instances. This approach allows the evolution of modularity and reuse both on structural and symbol levels resulting in a compression of the representation of a solution. In this paper an analysis of altering the rate of sampling of the evolved solution grammars is undertaken. It is found that the majority of evolutionary search is currently focused on the generation of the solution grammars to such an extent that the candidate solutions are often hard-coded into them making the solution chromosome effectively redundant. This opens the door to future work in which we can explore how the search can be better balanced between the meta and solution grammars