Exploring grammatical modification with modules in grammatical evolution

  • Authors:

  • John Mark Swafford;Michael O'Neill;Miguel Nicolau;Anthony Brabazon

  • Affiliations:

  • Natural Computing Research & Applications Group, Complex and Adaptive Systems Laboratory, School of Computer Science & Informatics, University College Dublin, Ireland;Natural Computing Research & Applications Group, Complex and Adaptive Systems Laboratory, School of Computer Science & Informatics, University College Dublin, Ireland;Natural Computing Research & Applications Group, Complex and Adaptive Systems Laboratory, School of Computer Science & Informatics, University College Dublin, Ireland;Natural Computing Research & Applications Group, Complex and Adaptive Systems Laboratory, School of Business, University College Dublin, Ireland

  • Venue:
  • EuroGP'11 Proceedings of the 14th European conference on Genetic programming
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

There have been many approaches to modularity in the field of evolutionary computation, each tailored to function with a particular representation. This research examines one approach to modularity and grammar modification with a grammar-based approach to genetic programming, grammatical evolution (GE). Here, GE's grammar was modified over the course of an evolutionary run with modules in order to facilitate their appearance in the population. This is the first step in what will be a series of analysis on methods of modifying GE's grammar to enhance evolutionary performance. The results show that identifying modules and using them to modify GE's grammar can have a negative effect on search performance when done improperly. But, if undertaken thoughtfully, there are possible benefits to dynamically enhancing the grammar with modules identified during evolution.