Chemical Genetic Programming - Coevolution Between Genotypic Strings and Phenotypic Trees

Chemical Genetic Programming (CGP) is a new method of genetic programming that introduced collision-based biochemical processes and realized dynamic mapping from genotypic strings to phenotypic trees. Chemical Genetic Programming (CGP) [1,2] proposes a new method of genetic programming that enables evolutionary optimization of the mapping from genotypic strings to phenotypic trees. The phenotypic tree of an individual is created by rewriting an initial expression using a series of grammatical rewriting rules. The information stored on DNA serves two purposes. First, it determines which rewriting rules to apply in building an individual’s phenotypic tree. Secondly, the DNA is used to create this set of rewriting rules. CGP uses a collision-based method for translating DNA into the set of rewriting rules. Fig. 1 shows the structure of a cell in CGP. An initial set of rewriting rules is supplied that provides the system with the syntactical structure within which all programs must be ge...

• Type: Conference
• Year: 2004
• Where: GECCO
• Authors: Wojciech Piaseczny, Hideaki Suzuki, Hidefumi Sawai
• Keywords: Chemical Genetic Programming, GECCO 2004, Genetic Programming, Phenotypic Tree