Improving Modularity in Genetic Programming Using
Graph-Based Data Mining
I. Jonyer, and A. Himes. Proceedings of the Nineteenth International Florida
Artificial Intelligence Research Society Conference, page 556--561. Melbourne Beach, Florida, USA, American Association for Artificial Intelligence, (May 2006)
Abstract
We propose to improve the efficiency of genetic
programming, a method to automatically evolve computer
programs. We use graph-based data mining to identify
common aspects of highly fit individuals and
modularising them by creating functions out of the
subprograms identified. Empirical evaluation on the
lawn mower problem shows that our approach is
successful in reducing the number of generations needed
to find target programs. Even though the graph-based
data mining system requires additional processing time,
the number of individuals required in a generation can
also be greatly reduced, resulting in an overall
speed-up.
%0 Conference Paper
%1 Jonyer:2006:FLAIRS
%A Jonyer, Istvan
%A Himes, Akiko
%B Proceedings of the Nineteenth International Florida
Artificial Intelligence Research Society Conference
%C Melbourne Beach, Florida, USA
%D 2006
%E Sutcliffe, Geoff C. J.
%E Goebel, Randy G.
%I American Association for Artificial Intelligence
%K Discovery Learning Machine algorithms, and genetic programming,
%P 556--561
%T Improving Modularity in Genetic Programming Using
Graph-Based Data Mining
%X We propose to improve the efficiency of genetic
programming, a method to automatically evolve computer
programs. We use graph-based data mining to identify
common aspects of highly fit individuals and
modularising them by creating functions out of the
subprograms identified. Empirical evaluation on the
lawn mower problem shows that our approach is
successful in reducing the number of generations needed
to find target programs. Even though the graph-based
data mining system requires additional processing time,
the number of individuals required in a generation can
also be greatly reduced, resulting in an overall
speed-up.
@inproceedings{Jonyer:2006:FLAIRS,
abstract = {We propose to improve the efficiency of genetic
programming, a method to automatically evolve computer
programs. We use graph-based data mining to identify
common aspects of highly fit individuals and
modularising them by creating functions out of the
subprograms identified. Empirical evaluation on the
lawn mower problem shows that our approach is
successful in reducing the number of generations needed
to find target programs. Even though the graph-based
data mining system requires additional processing time,
the number of individuals required in a generation can
also be greatly reduced, resulting in an overall
speed-up.},
added-at = {2008-06-19T17:35:00.000+0200},
address = {Melbourne Beach, Florida, USA},
author = {Jonyer, Istvan and Himes, Akiko},
biburl = {https://www.bibsonomy.org/bibtex/2a8bd9c0ed6232c3e5f1fd69efeaf6314/brazovayeye},
booktitle = {Proceedings of the Nineteenth International Florida
Artificial Intelligence Research Society Conference},
editor = {Sutcliffe, Geoff C. J. and Goebel, Randy G.},
interhash = {b37633b4ed9d70821767b94e8e5b1661},
intrahash = {a8bd9c0ed6232c3e5f1fd69efeaf6314},
keywords = {Discovery Learning Machine algorithms, and genetic programming,},
month = {May 11-13},
notes = {http://www.cs.miami.edu/~geoff/Conferences/FLAIRS-19/Schedule.shtml
http://www.aaai.org/Press/Proceedings/flairs06.php},
pages = {556--561},
publisher = {American Association for Artificial Intelligence},
timestamp = {2008-06-19T17:42:40.000+0200},
title = {Improving Modularity in Genetic Programming Using
Graph-Based Data Mining},
year = 2006
}