- author = "Maarten Keijzer and Martin Baptist and Vladan Babovic and Javier {Rodriguez Uthurburu}",
- title = "Determining equations for vegetation induced resistance using genetic programming",
- booktitle = "{GECCO 2005}: Proceedings of the 2005 conference on Genetic and evolutionary computation",
- year = "2005",
- editor = "Hans-Georg Beyer and Una-May O'Reilly and Dirk V. Arnold and Wolfgang Banzhaf and Christian Blum and Eric W. Bonabeau and Erick Cantu-Paz and Dipankar Dasgupta and Kalyanmoy Deb and James A. Foster and Edwin D. {de Jong} and Hod Lipson and Xavier Llora and Spiros Mancoridis and Martin Pelikan and Guenther R. Raidl and Terence Soule and Andy M. Tyrrell and Jean-Paul Watson and Eckart Zitzler",
- volume = "2",
- ISBN = "1-59593-010-8",
- pages = "1999--2006",
- address = "Washington DC, USA",
-
URL = "
http://gpbib.cs.ucl.ac.uk/gecco2005/docs/p1999.pdf",
-
DOI = "
doi:10.1145/1068009.1068343",
- publisher = "ACM Press",
- publisher_address = "New York, NY, 10286-1405, USA",
- month = "25-29 " # jun,
- organisation = "ACM SIGEVO (formerly ISGEC)",
- keywords = "genetic algorithms, genetic programming, Real World Applications, equation induction, hydraulics, hydrology, measurement units",
-
notes = "GECCO-2005 A joint meeting of the fourteenth
international conference on genetic algorithms
(ICGA-2005) and the tenth annual genetic programming
conference (GP-2005).
ACM Order Number 910052
7. CONCLUSIONS Four formulations for waterdepth-related resistance induced by vegetation are studied and compared. Two have been explicitly derived by a scientist building upon the extensive literature on this subject. One expression was derived by dimensionally aware genetic programming, and finally one expression was created by manually analysing and improving the genetic programming equation. It was found that the genetic programming equations were superior to the manually derived equations, both on their performance on synthetic training and laboratory testing data, and in the economy of detail that needs to be modelled. The manually improved expression was found to be in good agreement with previous expressions found in the literature, and performed competitive with the detailed model that was used to generate the training data.",
