Speeding up Genetic Programming

Created by W.Langdon from gp-bibliography.bib Revision:1.7615

@InProceedings{oai:CiteSeerPSU:336117,

• title = "Speeding up Genetic Programming",
• author = "Penousal Machado and Amilcar Cardoso",
• booktitle = "Proceedings of the Second International Symposium on Artificial Intelligence, Adaptive Systems (CIMAF - 99)",
• year = "1999",
• address = "Havana, Cuba",
• month = mar # " 22-26",
• keywords = "genetic algorithms, genetic programming",
• URL = "http://eden.dei.uc.pt/~machado/research/pdf/1999/cimaf99-fasteval.pdf",
• URL = "http://eden.dei.uc.pt/~ernesto/EvoCo/papers/papers/1999/cimaf992.htm",
• URL = "http://citeseer.ist.psu.edu/336117.html",
• citeseer-isreferencedby = "oai:CiteSeerPSU:41881; oai:CiteSeerPSU:361360; oai:CiteSeerPSU:231399; oai:CiteSeerPSU:560606",
• citeseer-references = "oai:CiteSeerPSU:276822; oai:CiteSeerPSU:186935",
• annote = "The Pennsylvania State University CiteSeer Archives",
• language = "en",
• oai = "oai:CiteSeerPSU:336117",
• rights = "unrestricted",
• abstract = "One of the major drawbacks of Evolutionary Computation is the need for great computational power. The set of problems that can be solved, in practice, by evolutionary approaches is highly connected with the efficiency of the algorithm. In most Genetic Programming applications the majority of time is spent on the evaluation of the individuals. Accordingly, it is desirable to optimise this step of the process. In this paper we present two approaches through which significant speed improvements can be achieved. The first approach, T-functions, is effective in tasks, such as symbolic regression, that require repeated evaluation of the individuals. The second approach, caching, resorts to the storage of the execution results of individuals' sub-trees, thus avoiding the recalculation of these sub-programs. Caching finds its application when the function set includes complex, time-consuming functions.",

}

Genetic Programming entries for Penousal Machado F Amilcar Cardoso

Citations