Strengths and weaknesses of FSA representation

  • Authors:
  • Pavel Petrovic

  • Affiliations:
  • NTNU, Trondheim, Norway

  • Venue:
  • Proceedings of the 9th annual conference on Genetic and evolutionary computation
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

Genetic Programming and Evolutionary Programming are fields studying the application of artificial evolutionon evolving directly executable programs, in form of trees similar to Lisp expressions (GP-trees), or Finite State Automata (FSA).In this exercise, we study the performance of these methods on several example problems, and draw conclusionson the suitability of the representations with respect to the task structure and properties. We investigate the roleof incremental evolution and its bias in the context of FSA representation. The experiments are performed in simulation and/or confirmed on real robots.