W. Langdon, and W. Banzhaf. Parallel Problem Solving from Nature - PPSN VI 6th
International Conference, volume 1917 of LNCS, page 201--210. Paris, France, Springer Verlag, (16-20 September 2000)
Abstract
To investigate the fundamental causes of bloat, six
artificial random binary tree search spaces are
presented. Fitness is given by program syntax (the
genetic programming genotype). GP populations are
evolved on both random problems and problems with
``building blocks''. These are compared to problems
with explicit ineffective code (introns, junk code,
inviable code). Our results suggest the entropy random
walk explanation of bloat remains viable. The hard
building block problem might be used in further
studies, e.g. of standard subtree crossover.
Parallel Problem Solving from Nature - PPSN VI 6th
International Conference
year
2000
month
16-20 September
pages
201--210
publisher
Springer Verlag
series
LNCS
volume
1917
notes
C++ code at ftp://cs.ucl.ac.uk/wblangdon/gp-code/
http://www-syntim.inria.fr/fractales/PPSN2000/program.html#108
PPSN'2000
http://www.springer.de/cgi-bin/search_book.pl?isbn=3-540-41056-2
%0 Conference Paper
%1 langdon:2000:random
%A Langdon, W. B.
%A Banzhaf, W.
%B Parallel Problem Solving from Nature - PPSN VI 6th
International Conference
%C Paris, France
%D 2000
%E Schoenauer, Marc
%E Deb, Kalyanmoy
%E Rudolph, Günter
%E Yao, Xin
%E Lutton, Evelyne
%E Merelo, Juan Julian
%E Schwefel, Hans-Paul
%I Springer Verlag
%K algorithms, binary depth evolution genetic growth, length linear of programming, search shape, spaces subquadratic tree
%P 201--210
%T Genetic Programming Bloat without Semantics
%U http://www.cs.mun.ca/~banzhaf/papers/ppsn00.pdf
%V 1917
%X To investigate the fundamental causes of bloat, six
artificial random binary tree search spaces are
presented. Fitness is given by program syntax (the
genetic programming genotype). GP populations are
evolved on both random problems and problems with
``building blocks''. These are compared to problems
with explicit ineffective code (introns, junk code,
inviable code). Our results suggest the entropy random
walk explanation of bloat remains viable. The hard
building block problem might be used in further
studies, e.g. of standard subtree crossover.
@inproceedings{langdon:2000:random,
abstract = {To investigate the fundamental causes of bloat, six
artificial random binary tree search spaces are
presented. Fitness is given by program syntax (the
genetic programming genotype). GP populations are
evolved on both random problems and problems with
``building blocks''. These are compared to problems
with explicit ineffective code (introns, junk code,
inviable code). Our results suggest the entropy random
walk explanation of bloat remains viable. The hard
building block problem might be used in further
studies, e.g. of standard subtree crossover.},
added-at = {2008-06-19T17:35:00.000+0200},
address = {Paris, France},
author = {Langdon, W. B. and Banzhaf, W.},
biburl = {https://www.bibsonomy.org/bibtex/20e2bccbadcd8172622a0d9b8b7daaa82/brazovayeye},
booktitle = {Parallel Problem Solving from Nature - PPSN VI 6th
International Conference},
editor = {Schoenauer, Marc and Deb, Kalyanmoy and Rudolph, G{\"u}nter and Yao, Xin and Lutton, Evelyne and Merelo, Juan Julian and Schwefel, Hans-Paul},
interhash = {4f3c65f6c6be996e72dd038c28639d1d},
intrahash = {0e2bccbadcd8172622a0d9b8b7daaa82},
keywords = {algorithms, binary depth evolution genetic growth, length linear of programming, search shape, spaces subquadratic tree},
month = {16-20 September},
notes = {C++ code at ftp://cs.ucl.ac.uk/wblangdon/gp-code/
http://www-syntim.inria.fr/fractales/PPSN2000/program.html#108
PPSN'2000
http://www.springer.de/cgi-bin/search_book.pl?isbn=3-540-41056-2},
pages = {201--210},
publisher = {Springer Verlag},
series = {LNCS},
timestamp = {2008-06-19T17:44:58.000+0200},
title = {Genetic Programming Bloat without Semantics},
url = {http://www.cs.mun.ca/~banzhaf/papers/ppsn00.pdf},
volume = 1917,
year = 2000
}