Resource-Limited Genetic Programming: Replacing Tree Depth Limits

作者: Sara Silva , Pedro J.N. Silva , Ernesto Costa

DOI: 10.1007/3-211-27389-1_58

关键词:

摘要: We propose replacing the traditional tree depth limit in Genetic Programming by a single on amount of resources available to whole population, where are nodes. The resource-limited technique removes disadvantages using limits at individual level, while introducing automatic population resizing, natural side-effect an approach level. results show that replacement resource can be done without impairing performance, thus validating this first and important step towards new improving efficiency GP.

参考文章(16)
Sara Silva, Ernesto Costa, Dynamic Limits for Bloat Control Genetic and Evolutionary Computation – GECCO 2004. pp. 666- 677 ,(2004) , 10.1007/978-3-540-24855-2_74
Sara Silva, Jonas Almeida, GPLAB A Genetic Programming Toolbox for MATLAB ,(2004)
Riccardo Poli, A simple but theoretically-motivated method to control bloat in genetic programming european conference on genetic programming. pp. 204- 217 ,(2003) , 10.1007/3-540-36599-0_19
Liviu Panait, Sean Luke, Lexicographic parsimony pressure genetic and evolutionary computation conference. pp. 829- 836 ,(2002)
Claire J. Kennedy, Christophe Giraud-Carrier, A depth controlling strategy for Strongly Typed Evolutionary Programming genetic and evolutionary computation conference. pp. 879- 885 ,(1999)
Wolfgang Banzhaf, Robert E. Keller, Peter Nordin, Genetic Programming: An Introduction ,(1997)
Sean Luke, Gabriel Catalin Balan, Liviu Panait, Population implosion in genetic programming genetic and evolutionary computation conference. pp. 1729- 1739 ,(2003) , 10.1007/3-540-45110-2_65
F. Fernandez, M. Tomassini, L. Vanneschi, Saving computational effort in genetic programming by means of plagues congress on evolutionary computation. ,vol. 3, pp. 2042- 2049 ,(2003) , 10.1109/CEC.2003.1299924