Mutation and crossover with abstract expression grammars

Simple enhancements to the standard population operators of mutation and crossover, utilizing Abstract Expression Grammars, are investigated. In previous works, Abstract Expression Grammars have been used to integrate Genetic Algorithms, Genetic Programming, Swarm Intelligence, and Differential Evolution, into a seamlessly unified approach to symbolic regression. In this work, the potential for Abstract Expression Grammars to have a direct impact on the classic Genetic Programming mutation and crossover operators is demonstrated. The features of abstract expression grammars are explored, details of abstract mutation and crossover are provided, and the beneficial effects of abstract mutation and crossover are tested with several published nonlinear regression problems.