A comparison of fitness-case sampling methods for genetic programming
Created by W.Langdon from
gp-bibliography.bib Revision:1.8081
- @Article{journals/jetai/MartinezNTLL17,
-
title = "A comparison of fitness-case sampling methods for
genetic programming",
-
author = "Yuliana Martinez and Enrique Naredo and
Leonardo Trujillo and Pierrick Legrand and Uriel Lopez",
-
journal = "Journal of Experimental \& Theoretical Artificial
Intelligence",
-
year = "2017",
-
number = "6",
-
volume = "29",
-
pages = "1203--1224",
-
keywords = "genetic algorithms, genetic programming, fitness-case
sampling, performance evaluation",
-
bibdate = "2017-11-06",
-
bibsource = "DBLP,
http://dblp.uni-trier.de/db/journals/jetai/jetai29.html#MartinezNTLL17",
-
DOI = "doi:10.1080/0952813X.2017.1328461",
-
size = "22 pages",
-
abstract = "Genetic programming (GP) is an evolutionary
computation paradigm for automatic program induction.
GP has produced impressive results but it still needs
to overcome some practical limitations, particularly
its high computational cost, overfitting and excessive
code growth. Recently, many researchers have proposed
fitness-case sampling methods to overcome some of these
problems, with mixed results in several limited tests.
This paper presents an extensive comparative study of
four fitness-case sampling methods, namely: Interleaved
Sampling, Random Interleaved Sampling, Lexicase
Selection and Keep-Worst Interleaved Sampling. The
algorithms are compared on 11 symbolic regression
problems and 11 supervised classification problems,
using 10 synthetic benchmarks and 12 real-world
data-sets. They are evaluated based on test
performance, overfitting and average program size,
comparing them with a standard GP search. Comparisons
are carried out using non-parametric multigroup tests
and post hoc pairwise statistical tests. The
experimental results suggest that fitness-case sampling
methods are particularly useful for difficult
real-world symbolic regression problems, improving
performance, reducing overfitting and limiting code
growth. On the other hand, it seems that fitness-case
sampling cannot improve upon GP performance when
considering supervised binary classification.",
- }
Genetic Programming entries for
Yuliana Martinez
Enrique Naredo
Leonardo Trujillo
Pierrick Legrand
Uriel Lopez
Citations