Genetic Programming

The genetic programming is an automatic methodology inspired by the Théorie of the evolution such as it was defined by Charles Darwin in the particular case of the biological mechanisms. It is fixed for goal to find by successive approximations of the programs answering a given task as well as possible.

Description

One names genetic programming a technique allowing a computer program to learn, by an evolutionary algorithm, to optimize little by little a population of other programs to increase their degree of adaptation (fitness) to carry out a task requested by a user.

History

The first genetic experiment of programming was deferred by Stephen F. Smith (1980) and Nichael L. Cramer (1985), as indicated in the work of origin and reference: Genetic Programming: One the Programming off Computers by Means off Natural Selection , by John Koza (1992).

The genetic algorithms do not impose a particular data-processing language. The arborescent organization of information supported at the beginning of the languages like Lisp (Forza). Thereafter, one also used linear representations easy to handle by any language.

Forces and weaknesses

The genetic programming is expensive in computing times machine, since it puts in parallel competition of way a great number of close algorithms. With his beginnings, in the years 1980, one thus limited it to solve simple problems. As the power of the processors multiplied, the genetic programming started to give more powerful results: at the end of 2004, for example, one counted forty significant results in the following fields:

quantum calculation,
electronic CAD (placement of components)
resolution of plays, sorting, research.

Would these results also include/understand (??? Source?) the D-invention or the invalidation of many recent inventions and even production of 2 patentable inventions.

Until the years 1990 the genetic programming constituted only a Heuristique and not a discipline with whole share. After some projections in years 2000 developed a theory with whole share as the technique spread about it. At the point even as it is possible to make exact probabilistic models of the genetic programming and genetic algorithms.

Today, in addition to the software, the genetic programming is also applied to the evolution of the material.

Genetic Meta programming is the technique aiming at making evolve/move a genetic programming system by using the genetic programming itself.

See too

genetic Algorithm

Bibliography in English language

Koza, J.R. (1990), Genetic Programming: In Paradigm for Genetically Breeding Populations off Computer Programs to Solve Problems, Stanford University Computer Science Department technical carryforward STAN-CS-90-1314. With thorough carryforward, possibly used ace has draft to his 1992 book.
Koza, J.R. (1992), Genetic Programming: One the Programming off Computers by Means off Natural Selection, MIT Close
Koza, J.R. (1994), Genetic Programming II: Automatic Discovery off Reusable Programs, MIT Close
Koza, J.R., Bennett, F.H., Andre, D., and Keane, M.A. (1999), Genetic Programming III: Darwinian Invention and Problem Solving, Morgan Kaufmann
Koza, J.R., Keane, M.A., Streeter, M.J., Mydlowec, W., Yu, J., Lanza, G. (2003), Genetic Programming IV: Human-Competitive routine Machine Intelligence, Kluwer Academic Publishers

Random links: