作者: Julian F. Miller , Simon L. Harding
关键词: Arbitrary-precision arithmetic 、 Modular design 、 Genetic representation 、 Genetic programming 、 Artificial neural network 、 Theoretical computer science 、 Directed graph 、 Cartesian genetic programming 、 Computer science
摘要: Cartesian Genetic Programming (CGP) is an increasingly popular and efficient form of that was developed by Julian Miller in 1999 2000. In its classic form, it uses a very simple integer based genetic representation program the directed graph. Graphs are useful representations can be applied to many domains (e.g. electronic circuits, neural networks). number studies, CGP has been shown comparatively other GP techniques. It also program. Since then, classical made more various ways. Notably including automatically defined functions (modular CGP) self-modification operators(self-modifying CGP). SMCGP Miller, Simon Harding Wolfgang Banzhaf. cause evolved programs change themselves as function time. Using this technique possible find general solutions classes problems mathematical algorithms arbitrary parity, n-bit binary addition, sequences provably compute pi e precision, so on). The tutorial will cover basic technique, advanced developments applications variety problem domains.