Grammatical evolution

Abstract

Grammatical Evolution (GE) involves a mapping from a bit string to a program/expression via a user-defined grammar. By performing an unconstrained optimisation over a population of the strings, programs and expressions which achieve a desired goal can be discovered. A hybrid FS and feature generation algorithm using a modified version of GE is introduced. This algorithm serves as the inherent feature selection mechanism in the feature generation framework. The chapter also describes a R package which implements GE for automatic string expression generation. The package facilitates the coding and execution of GE programs and supports parallel execution.