Automatic Generation of Programs with Grammatical Evolution

Grammatical Evolution (GE) is a grammar based Evolutionary Algorithm to generate computer programs which has been shown to be competitive with Genetic Programming when applied to a diverse array of problems. GE has the distinction that its input is a BNF, which permits it to generate arbitrarily complex programs in any language, including loops, multiple line functions etc. Part of the power of GE is that it is closer to natural DNA than GP, and thus can beneet from natural phenomena such as a separation of search and solution spaces through a genotype to phenotype mapping and a genetic code degeneracy which can give rise to neutral mutations (Mutations that have no eeect on the phenotype). In this paper we describe how GE was applied to the real world problem of evolving a Caching Algorithm, at which GP has been found to generate algorithms that did not perform as well as those designed by humans. GE was found to generate caching algorithms which clearly outperform those generated by GP.