Genetic programming

Genetic programming (GP) issubfieldevolutionary computation invented by Nichael Lynn Cramer1985first exploreddepth by John Kozahis 1992 book Genetic Programming: OnProgrammingComputers by MeansNatural Selection. It ismethod usedallow computer programsevolve accordingsome user-defined goal. It uses evolutionary patterns including crossover, selection, replicationmutationsevolveprograms, whichusually represented by Lisp expressions. To work effectively,requires an appropriate selectionoperatorsvariables.

Genetic programming uses methods whichsimilargenetic algorithms (GA), butbased on programs which perform tasks whose results can then be evaluateddeliverfitness function similarGAs. Insteadusing poolsparameter listsbe evaluated by some evaluation procedure, GP uses poolsprograms whichto be runperformrequired task. A technical difference between GAsGPsthat GAs use list structures, oftenfixed size,store their data, while GPs use tree structures which can varysizeshapeeach program used inprogram pools.

The application oftree representation (and required genetic operators)using genetic algorithmsgenerate programs was first described1985 by Cramer. Koza, though he did not invent genetic programming,indisputablyfield's most prolificpersuasive author.

So far GPs have successfully solved some toy problems, such aslawn mower problem, butmethodvery computationally intensive,may not compare favourably where simpler methods, such as genetic algorithms or random optimisation can be used instead. Itpossible that some more complex problems may be more amenablesolution using GPs than other optimization methods.

Unfortunately, due tolacksolid theory regardingperformancegenetic programming vs. traditional search methods (such as hill-climbing), genetic programming remainssortpariah amongstvarious techniquessearch. While genetic programming has achieved results thatas good assometimes better than human-generated results, more work needsbe done ontheoryorderbringtechnique into more widespread use.

Bibliography

• Cramer, Nichael Lynn (1985), "A representation forAdaptive GenerationSimple Sequential Programs"Proceedingsan International Conference on Genetic Algorithms andApplications, Grefenstette, John J. (ed.), CMU
• Koza, J.R. (1992), Genetic Programming: OnProgrammingComputers by MeansNatural Selection, MIT Press
• Koza, J.R. (1994), Genetic Programming II: Automatic DiscoveryReusable Programs, MIT Press
• Koza, J.R., Bennett, F.H., Andre, D.,Keane, M.A. (1999), Genetic Programming III: Darwinian InventionProblem Solving, Morgan Kaufmann

External links

• Nichael Lynn Cramer
• Genetic Programming FAQ
• The Hitch-Hiker's GuideEvolutionary Computation
• John Koza's Genetic Programming Site
• Juergen Schmidhuber's GP Site,pre-Koza GP papers (1987)
• Bill Langdon's GP bibliography