Zoetrope Genetic Programming for Regression - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2021

Zoetrope Genetic Programming for Regression

Aurélie Boisbunon
  • Function : Author
  • PersonId : 910949
Carlo Fanara
  • Function : Author
Ingrid Grenet
  • Function : Author
Jonathan Daeden
  • Function : Author
Alexis Vighi
  • Function : Author
  • PersonId : 1092307
Marc Schoenauer

Abstract

The Zoetrope Genetic Programming (ZGP) algorithm is based on an original representation for mathematical expressions, targeting evolutionary symbolic regression. The zoetropic representation uses repeated fusion operations between partial expressions, starting from the terminal set. Repeated fusions within an individual gradually generate more complex expressions, ending up in what can be viewed as new features. These features are then linearly combined to best fit the training data. ZGP individuals then undergo specific crossover and mutation operators, and selection takes place between parents and offspring. ZGP is validated using a large number of public domain regression datasets, and compared to other symbolic regression algorithms, as well as to traditional machine learning algorithms. ZGP reaches state-of-theart performance with respect to both types of algorithms, and demonstrates a low computational time compared to other symbolic regression approaches.
Fichier principal
Vignette du fichier
ZGP_regression_arxiv.pdf (1.63 Mo) Télécharger le fichier
PMLB-reg-SR_r2-time.pdf (545.12 Ko) Télécharger le fichier
PMLB-reg-SR_r2-time_zoom.pdf (536.87 Ko) Télécharger le fichier
PMLB-reg-SR_r2median-boxplot.png (27.47 Ko) Télécharger le fichier
PMLB_reg_SR_nrmsemedian_boxplot.png (24.55 Ko) Télécharger le fichier
Stages_ZGP_small_grey.png (238.54 Ko) Télécharger le fichier
Origin Files produced by the author(s)
Origin Files produced by the author(s)
Origin Files produced by the author(s)

Dates and versions

hal-03155694 , version 1 (02-03-2021)

Licence

Identifiers

Cite

Aurélie Boisbunon, Carlo Fanara, Ingrid Grenet, Jonathan Daeden, Alexis Vighi, et al.. Zoetrope Genetic Programming for Regression. GECCO 2021, ACM, Jul 2021, Lille, France. pp.776-784. ⟨hal-03155694⟩
211 View
224 Download

Altmetric

Share

Gmail Mastodon Facebook X LinkedIn More