Short works

Books : reviews

Michael O'Neill, Conor Ryan.
Grammatical Evolution: evolutionary automatic programming in an arbitrary language.
Kluwer. 2003

rating : 3.5 : worth reading
review : 18 October 2007

Grammatical evolution: Evolutionary Automatic Programming in an Arbitrary Language provides the first comprehensive introduction to Grammatical Evolution, a novel approach to Genetic Programming that adopts principles from molecular biology in a simple and useful manner, coupled with the use of grammars to specify legal structures in a search. Grammatical Evolution’s rich modularity gives a unique flexibility, making it possible to use alternative search strategies—whether evolutionary, deterministic or some other approach—and to radically change its behavior by merely changing the grammar supplied. This approach to Genetic Programming represents a powerful new weapon in the Machine Learning toolkit that can be applied to a diverse set of problem domains.

Beginning with an overview of the necessary background material in Genetic Programming and Molecular Biology, Grammatical evolution: Evolutionary Automatic Programming in an Arbitrary Language outlines state of the art in grammatical and genotype-phenotype-based approaches. Following a description of Grammatical Evolution and its application to a number of example problems, an in-depth analysis of the approach is conducted, focusing on areas such as the degenerate genetic code, wrapping, and crossover. The book continues with a description of hot topics in Grammatical Evolution and presents possible directions for future research.

Grammatical Evolution is a new kind of evolutionary algorithm that evolves computer programs in an indirect manner. The evolved “genotype” is a string of numbers (in contrast to the explicit program code genotype of Genetic Programming, for example). These numbers are decoded into a program by using them to index into a BNF description of the program’s grammar, and using the indexed productions as the components of the generated program.

This book, written by GE’s inventors, explains the process. There is some biological background, a description of the GE algorithm itself, and some simple case-studies demonstrating its effectiveness. If you want a brief, clear overview of GE, and some background material, here it is.