Meta-Genetic Programming: Co-evolving the Operators of Variation

摘要

The standard Genetic Programming approach is augmented by co-evolvingthe genetic operators. To do this the operators are coded as trees ofindefinite length. In order for this technique to work, the languagethat the operators are defined in must be such that it preserves thevariation in the base population. This technique can varied by addingfurther populations of operators and changing which populations act asoperators for others, including itself, thus to provide a frameworkfor a whole set of augmented GP techniques. The technique is testedon the parity problem. The pros and cons of the technique arediscussed.