Withall2009-final.pdf (303.27 kB)
An improved representation for evolving programs
journal contribution
posted on 2009-03-04, 13:29 authored by Mark S. Withall, Christopher Hinde, Roger StoneA representation has been developed that addresses some of the issues
with other Genetic Program representations while maintaining their advantages.
This combines the easy reproduction of the linear representation with the inherita-
ble characteristics of the tree representation by using fixed-length blocks of genes
representing single program statements. This means that each block of genes will
always map to the same statement in the parent and child unless it is mutated,
irrespective of changes to the surrounding blocks. This method is compared to the
variable length gene blocks used by other representations with a clear improvement
in the similarity between parent and child. In addition, a set of list evaluation and
manipulation functions was evolved as an application of the new Genetic Program
components. These functions have the common feature that they all need to be 100%
correct to be useful. Traditional Genetic Programming problems have mainly been
optimization or approximation problems. The list results are good but do highlight
the problem of scalability in that more complex functions lead to a dramatic increase
in the required evolution time.
History
School
- Science
Department
- Computer Science
Citation
WITHALL, M.S., HINDE, C.J. and STONE, R.G., 2009. An improved representation for evolving programs. Genetic Programming and Evolvable Machines, 10 (1), pp. 37-70Publisher
© Springer Science+Business MediaVersion
- AM (Accepted Manuscript)
Publication date
2009Notes
This article was published in the journal, Genetic Programming and Evolvable Machines [© Springer Science+Business Media]. The original publication is available at www.springerlink.com http://dx.doi.org/10.1007/s10710-008-9069-7ISSN
1389-2576Language
- en