Reducing Overparameterization of Symbolic Regression Models with Equality Saturation

Fabricio Olivetti De Franca, Gabriel Kronberger

Publikation: Beitrag in Buch/Bericht/TagungsbandKonferenzbeitragBegutachtung

Abstract

Overparameterized models in regression analysis are often harder to interpret and can be harder to fit because of ill-conditioning. Genetic programming is prone to overparameterized models as it evolves the structure of the model without taking the location of parameters into account. One way to alleviate this is rewriting the expression and merging the redundant fitting parameters. In this paper we propose the use of equality saturation to alleviate overparameterization. We first notice that all the tested GP implementations suffer from overparameterization to different extents and then show that equality saturation together with a small set of rewriting rules is capable of reducing the number of fitting parameters to a minimum with a high probability. Compared to one of the few available alternatives, Sympy, it produces much better and consistent results. These results lead to different possible future investigations such as the simplification of expressions during the evolutionary process, and improvement of the interpretability of symbolic models.

OriginalspracheEnglisch
TitelGECCO 2023 - Proceedings of the 2023 Genetic and Evolutionary Computation Conference
Herausgeber (Verlag)Association for Computing Machinery, Inc
Seiten1064-1072
Seitenumfang9
ISBN (elektronisch)9798400701191
DOIs
PublikationsstatusVeröffentlicht - 15 Juli 2023
Veranstaltung2023 Genetic and Evolutionary Computation Conference, GECCO 2023 - Lisbon, Portugal
Dauer: 15 Juli 202319 Juli 2023

Publikationsreihe

NameGECCO 2023 - Proceedings of the 2023 Genetic and Evolutionary Computation Conference

Konferenz

Konferenz2023 Genetic and Evolutionary Computation Conference, GECCO 2023
Land/GebietPortugal
OrtLisbon
Zeitraum15.07.202319.07.2023

Fingerprint

Untersuchen Sie die Forschungsthemen von „Reducing Overparameterization of Symbolic Regression Models with Equality Saturation“. Zusammen bilden sie einen einzigartigen Fingerprint.

Zitieren