Evolutionary programming with tree mutations: evolving computer programs without crossover

摘要

This paper introduces an evolutionary programming procedure for evolving computer programs. Six tree mutation operators are proposed. Crossover is not among these mutation operators. The viability and eficiency of the method is investigated on four well-studied benchmark problems. The evidence indicates that the technique is not only viable but is indeed capable of evolving good computer programs. The results compare well with other evolutionary methods that include the crossover operator to address the same problems.