Self Modifying Cartesian Genetic Programming: Fibonacci, Squares, Regression and Summing

Developments in Cartesian Genetic Programming: self-modifying CGP

“…We found, however, that inputs still did not scale particularly well with problem size, so in subsequent papers [17,18,19] we examined another strategy: Three special input functions are now added to the function set: INP, INPP and SKIPINP. When decoding the phenotype graph, a pointer is maintained that refers to an input.…”

“…This ensures that there is always an input available to be read. 2 Also in earlier work we included an extra binary gene with every node which flagged whether the node could provide a program output [17,18,19]. However, in the work for this paper we have removed output genes and instead introduced a primitive function called OUTPUT that provides a program output.…”

A Survey of Self Modifying Cartesian Genetic Programming

“…When this happens, the representation reverts to something similar to classical CGP. In (Harding et al, 2009c), we showed that on a bio-informatics classification problem where there should be no benefit in using self modification, SMCGP behaved similarly to CGP. This result lets us be confident that in future work we can by default use SMCGP and automatically gain any advantages that development might bring.…”

“…SMCGP has been applied to a variety of mathematical problems (Harding et al, 2009c;Harding et al, 2010b).…”

Evolvability and Speed of Evolutionary Algorithms in Light of Recent Developments in Biology

“…Various encoding methods have been proposed using, for example, hierarchical grammars [236,237], or simulating cell chemical processes [238,239]. Indirect encoding schemes have shown advantages over traditional one-to-one direct encodings [240,241]. Indirect encoding is a first step to simulate biological development in computational systems by allowing more freedom and complexity in the genotype-phenotype mapping, but it is by no means the full story of development.…”