Generative Representations for Evolving Families of
Designs
G. Hornby. Genetic and Evolutionary Computation -- GECCO-2003, Volume 2724 von LNCS, Seite 1678--1689. Chicago, Springer-Verlag, (12-16 July 2003)
Zusammenfassung
Since typical evolutionary design systems encode only
a single artifact with each individual, each time the
objective changes a new set of individuals must be
evolved. When this objective varies in a way that can
be parameterized, a more general method is to use a
representation in which a single individual encodes an
entire class of artifacts. In addition to saving time
by preventing the need for multiple evolutionary runs,
the evolution of parameter-controlled designs can
create families of artifacts with the same style and a
reuse of parts between members of the family. In this
paper an evolutionary design system is described which
uses a generative representation to encode families of
designs. Because a generative representation is an
algorithmic encoding of a design, its input parameters
are a way to control aspects of the design it
generates. By evaluating individuals multiple times
with different input parameters the evolutionary design
system creates individuals in which the input parameter
controls specific aspects of a design. This system is
demonstrated on two design substrates: neural-networks
which solve the 3/5/7-parity problem and
three-dimensional tables of varying heights.
Genetic and Evolutionary Computation -- GECCO-2003
Jahr
2003
Monat
12-16 July
Seiten
1678--1689
Verlag
Springer-Verlag
Reihe
LNCS
Band
2724
publisher_address
Berlin
isbn
3-540-40603-4
notes
GECCO-2003. A joint meeting of the twelfth
International Conference on Genetic Algorithms
(ICGA-2003) and the eighth Annual Genetic Programming
Conference (GP-2003)
%0 Conference Paper
%1 hornby:2003:gecco
%A Hornby, Gregory S.
%B Genetic and Evolutionary Computation -- GECCO-2003
%C Chicago
%D 2003
%E Cantú-Paz, E.
%E Foster, J. A.
%E Deb, K.
%E Davis, D.
%E Roy, R.
%E O'Reilly, U.-M.
%E Beyer, H.-G.
%E Standish, R.
%E Kendall, G.
%E Wilson, S.
%E Harman, M.
%E Wegener, J.
%E Dasgupta, D.
%E Potter, M. A.
%E Schultz, A. C.
%E Dowsland, K.
%E Jonoska, N.
%E Miller, J.
%I Springer-Verlag
%K ANN Lindenmayer algorithms, evolving genetic networks, neural parametric programming, systems,
%P 1678--1689
%T Generative Representations for Evolving Families of
Designs
%U http://ic.arc.nasa.gov/people/hornby/papers/abstracts.html#hornby_gecco03
%V 2724
%X Since typical evolutionary design systems encode only
a single artifact with each individual, each time the
objective changes a new set of individuals must be
evolved. When this objective varies in a way that can
be parameterized, a more general method is to use a
representation in which a single individual encodes an
entire class of artifacts. In addition to saving time
by preventing the need for multiple evolutionary runs,
the evolution of parameter-controlled designs can
create families of artifacts with the same style and a
reuse of parts between members of the family. In this
paper an evolutionary design system is described which
uses a generative representation to encode families of
designs. Because a generative representation is an
algorithmic encoding of a design, its input parameters
are a way to control aspects of the design it
generates. By evaluating individuals multiple times
with different input parameters the evolutionary design
system creates individuals in which the input parameter
controls specific aspects of a design. This system is
demonstrated on two design substrates: neural-networks
which solve the 3/5/7-parity problem and
three-dimensional tables of varying heights.
%@ 3-540-40603-4
@inproceedings{hornby:2003:gecco,
abstract = {Since typical evolutionary design systems encode only
a single artifact with each individual, each time the
objective changes a new set of individuals must be
evolved. When this objective varies in a way that can
be parameterized, a more general method is to use a
representation in which a single individual encodes an
entire class of artifacts. In addition to saving time
by preventing the need for multiple evolutionary runs,
the evolution of parameter-controlled designs can
create families of artifacts with the same style and a
reuse of parts between members of the family. In this
paper an evolutionary design system is described which
uses a generative representation to encode families of
designs. Because a generative representation is an
algorithmic encoding of a design, its input parameters
are a way to control aspects of the design it
generates. By evaluating individuals multiple times
with different input parameters the evolutionary design
system creates individuals in which the input parameter
controls specific aspects of a design. This system is
demonstrated on two design substrates: neural-networks
which solve the 3/5/7-parity problem and
three-dimensional tables of varying heights.},
added-at = {2008-06-19T17:35:00.000+0200},
address = {Chicago},
author = {Hornby, Gregory S.},
biburl = {https://www.bibsonomy.org/bibtex/2526fdee4fd54a5c3f04c4e094a22c06f/brazovayeye},
booktitle = {Genetic and Evolutionary Computation -- GECCO-2003},
editor = {Cant{\'u}-Paz, E. and Foster, J. A. and Deb, K. and Davis, D. and Roy, R. and O'Reilly, U.-M. and Beyer, H.-G. and Standish, R. and Kendall, G. and Wilson, S. and Harman, M. and Wegener, J. and Dasgupta, D. and Potter, M. A. and Schultz, A. C. and Dowsland, K. and Jonoska, N. and Miller, J.},
interhash = {b2d1b818e06a8b86b70d865bd93cd2af},
intrahash = {526fdee4fd54a5c3f04c4e094a22c06f},
isbn = {3-540-40603-4},
keywords = {ANN Lindenmayer algorithms, evolving genetic networks, neural parametric programming, systems,},
month = {12-16 July},
notes = {GECCO-2003. A joint meeting of the twelfth
International Conference on Genetic Algorithms
(ICGA-2003) and the eighth Annual Genetic Programming
Conference (GP-2003)},
pages = {1678--1689},
publisher = {Springer-Verlag},
publisher_address = {Berlin},
series = {LNCS},
timestamp = {2008-06-19T17:41:45.000+0200},
title = {Generative Representations for Evolving Families of
Designs},
url = {http://ic.arc.nasa.gov/people/hornby/papers/abstracts.html#hornby_gecco03},
volume = 2724,
year = 2003
}