Guidelines for choosing hyperparameters of echo state networks for system identification: two case studies

Página do item simplificado
dc.creatorThiago de Almeida
dc.creatorUshikoshi
dc.creatorLuis Antonio Aguirre
dc.date.accessioned2025-05-29T13:37:56Z
dc.date.accessioned2025-09-09T01:26:40Z
dc.date.available2025-05-29T13:37:56Z
dc.date.issued2023
dc.identifier.doi10.1504/IJMIC.2023.129483
dc.identifier.issn17466172
dc.identifier.urihttps://hdl.handle.net/1843/82609
dc.languageeng
dc.publisherUniversidade Federal de Minas Gerais
dc.relation.ispartofInternational Journal of Modelling, Identification and Control
dc.rightsAcesso Restrito
dc.subjectSistemas dinâmicos
dc.subjectSistemas não lineares
dc.subject.otherreservoir computing, echo state networks, ESN, guidelines for echo state networks, hyperparameters of echo state networks, nonlinear models, system identification, dynamical system identification, modelling, chaotic oscillators
dc.titleGuidelines for choosing hyperparameters of echo state networks for system identification: two case studies
dc.typeArtigo de periódico
local.citation.issue2
local.citation.spage105
local.citation.volume42
local.description.resumoEcho state networks (ESN) can be used to model dynamical systems in the context of reservoir computing using standard regression algorithms, which is one of their main advantages. However, there are some hyperparameters that need to be carefully chosen and there is no general recommendation on how to perform this important step. After setting the ESN paradigm for system identification, this paper describes the choice of hyperparameters in the context of two case studies: one using experimental data of a pilot heater and the other using the Duffing-Ueda oscillator with chaotic dynamics. The main findings are: 1) ESNs can reproduce the chaotic regime of the Duffing-Ueda oscillator for a specific region on the hyperparameter space; 2) some hyperparameters may not be critical from a statistical perspective but can still drastically affect the dynamical regime; 3) the ESN initialisation is not critical when the hyperparameters are adequately chosen.
local.publisher.countryBrasil
local.publisher.departmentENG - DEPARTAMENTO DE ENGENHARIA ELETRÔNICA
local.publisher.initialsUFMG
local.url.externahttps://www.inderscienceonline.com/doi/abs/10.1504/IJMIC.2023.129483

Arquivos

Licença do pacote

Agora exibindo 1 - 1 de 1
Imagem de Miniatura
Nome:
License.txt
Tamanho:
1.99 KB
Formato:
Plain Text
Descrição:
Baixar

Coleções

Artigo de Periódico

Logo UFMG