Neutrality: A Necessity for Self-Adaptation

2002/04/16 by Marc Toussaint, Christian Igel
Physics and Astronomy · Computer Science · Biochemistry, Genetics and Molecular Biology · #nlin.AO #cs.NE #q-bio

paper · pdf

published as Proceedings of the Congress on Evolutionary Computation (CEC 2002), 1354-1359. · 6 pages, 3 figures, LaTeX

arxiv created 2002/04/16 · arxiv updated 2009/11/30

Abstract

Self-adaptation is used in all main paradigms of evolutionary computation to increase efficiency. We claim that the basis of self-adaptation is the use of neutrality. In the absence of external control neutrality allows a variation of the search distribution without the risk of fitness loss.

Related

• Adaptation and Self-Organization in Evolutionary Algorithms
• The Selfish Algorithm
• Self-Adaptation in Evolving Systems
• Self-adaptation in non-Elitist Evolutionary Algorithms on Discrete\n Problems with Unknown Structure
• Self-adaptive exploration in evolutionary search