Highly Robust Training of Regularized Radial Basis Function Networks

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F67985807%3A_____%2F24%3A00586040" target="_blank" >RIV/67985807:_____/24:00586040 - isvavai.cz</a>

Výsledek na webu

<a href="https://doi.org/85190739311" target="_blank" >https://doi.org/85190739311</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.14736/kyb-2024-1-0038" target="_blank" >10.14736/kyb-2024-1-0038</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Highly Robust Training of Regularized Radial Basis Function Networks

Popis výsledku v původním jazyce

Radial basis function (RBF) networks represent established tools for nonlinear regression modeling with numerous applications in various fields. Because their standard training is vulnerable with respect to the presence of outliers in the data, several robust methods for RBF network training have been proposed recently. This paper is interested in robust regularized RBF networks. A robust inter-quantile version of RBF networks based on trimmed least squares is proposed here. Then, a systematic comparison of robust regularized RBF networks follows, which is evaluated over a set of 405 networks trained using various combinations of robustness and regularization types. The experiments proceed with a particular focus on the effect of variable selection, which is performed by means of a backward procedure, on the optimal number of RBF units. The regularized inter-quantile RBF networks based on trimmed least squares turn out to outperform the competing approaches in the experiments if a highly robust prediction error measure is considered.

Název v anglickém jazyce

Highly Robust Training of Regularized Radial Basis Function Networks

Popis výsledku anglicky

Radial basis function (RBF) networks represent established tools for nonlinear regression modeling with numerous applications in various fields. Because their standard training is vulnerable with respect to the presence of outliers in the data, several robust methods for RBF network training have been proposed recently. This paper is interested in robust regularized RBF networks. A robust inter-quantile version of RBF networks based on trimmed least squares is proposed here. Then, a systematic comparison of robust regularized RBF networks follows, which is evaluated over a set of 405 networks trained using various combinations of robustness and regularization types. The experiments proceed with a particular focus on the effect of variable selection, which is performed by means of a backward procedure, on the optimal number of RBF units. The regularized inter-quantile RBF networks based on trimmed least squares turn out to outperform the competing approaches in the experiments if a highly robust prediction error measure is considered.

Druh

J_imp - Článek v periodiku v databázi Web of Science

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

<a href="/cs/project/GA22-02067S" target="_blank" >GA22-02067S: AppNeCo: Aproximativní neurovýpočty</a><br>

Návaznosti

I - Institucionalni podpora na dlouhodoby koncepcni rozvoj vyzkumne organizace

Rok uplatnění

2024

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název periodika

Kybernetika

ISSN

0023-5954

e-ISSN

—

Svazek periodika

60

Číslo periodika v rámci svazku

1

Stát vydavatele periodika

CZ - Česká republika

Počet stran výsledku

22

Strana od-do

38-59

Kód UT WoS článku

001202833900001

EID výsledku v databázi Scopus

2-s2.0-85190739311