Online sequential echo state network with sparse RLS algorithm for time series prediction

Yang, Cuili; Qiao, Junfei; Ahmad, Zohaib; Nie, Kaizhe; Wang, Lei

doi:10.1016/j.neunet.2019.05.006

Cited by 37 publications

(19 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

Section: Proposed Recursive Regularization Factor For Sparse Rls Algorithmmentioning

confidence: 99%

“…In (19), the proposed regularization parameter is approximately converse to the scaled optimal regularization parameter in [13]. Therefore, the newly derived regularization parameter γ(n) in (19) satisfies 0 ≤ γ(n) ≤ γ(n). Therefore, the proposed regularization parameter can be used as a regularization parameter on behalf of the optimal regularization parameter.…”

Section: Proposed Recursive Regularization Factor For Sparse Rls Algorithmmentioning

confidence: 99%

“…The authors in [13] also proposed a regularization factor calculation method for the SRLS algorithm. Similar recursive regularization factor selection methods were used in [15][16][17][18][19]. However, the algorithm from [15][16][17][18] is not practical because the regularization factor in [15][16][17] assumed that the true system impulse response is known in advance and set the true system response in part to an arbitrary constant in [18].…”

Section: Introductionmentioning

confidence: 99%

“…However, the algorithm from [15][16][17][18] is not practical because the regularization factor in [15][16][17] assumed that the true system impulse response is known in advance and set the true system response in part to an arbitrary constant in [18]. The regularization factor selection method in [19] needs no true system impulse response. However, the regularization factor is recursively updated.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

A Modified Recursive Regularization Factor Calculation for Sparse RLS Algorithm with l1-Norm

Lim

Lee

2021

Mathematics

View full text Add to dashboard Cite

In this paper, we propose a new calculation method for the regularization factor in sparse recursive least squares (SRLS) with l1-norm penalty. The proposed regularization factor requires no prior knowledge of the actual system impulse response, and it also reduces computational complexity by about half. In the simulation, we use Mean Square Deviation (MSD) to evaluate the performance of SRLS, using the proposed regularization factor. The simulation results demonstrate that SRLS using the proposed regularization factor calculation shows a difference of less than 2 dB in MSD from SRLS, using the conventional regularization factor with a true system impulse response. Therefore, it is confirmed that the performance of the proposed method is very similar to that of the existing method, even with half the computational complexity.

show abstract

Section: Proposed Recursive Regularization Factor For Sparse Rls Algorithmmentioning

confidence: 99%

Section: Proposed Recursive Regularization Factor For Sparse Rls Algorithmmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

A Modified Recursive Regularization Factor Calculation for Sparse RLS Algorithm with l1-Norm

Lim

Lee

2021

Mathematics

View full text Add to dashboard Cite

show abstract

“…Por sua vez, em Rodan and Tino (2010) foram exploradas arquiteturas simplificadas de ESN para identificação de sistemas dinâmicos. O estudo recente apresentado em Yang et al (2019) explora métodos para obtenção online de modelos baseados ESN com algoritmos recursivos esparsos. No domínio das aplicações, alguns trabalhos recentes vêm explorando a utilização desse modelo para identificação de diversos tipos de sistemas dinâmicos, como em sistemas de geração eólica Chen et al (2019), poços de petróleo Antonelo et al (2017) e sistemas de refrigeração Schwedersky et al (2018).…”

Section: Introductionunclassified

Identificação de Sistemas Dinâmicos Não Lineares Multivariáveis com Redes de Estado de Eco

Schwedersky

Flesch

Dangui

2020

Anais Do Congresso Brasileiro De Automática 2020

View full text Add to dashboard Cite

Este artigo apresenta um método de identificação baseado em redes de estado de eco para sistemas dinâmicos não lineres multivariáveis. A rede de estado de eco é uma rede neural recorrente, capaz de modelar dinâmicas não lineares, utilizando algoritmos de treinamento computacionalmente eficientes. Uma arquitetura de rede de estado de eco para identificação de sistemas e o processo para criação e ajuste dos hiperparâmetros do modelo são propostos. O método proposto e avaliado por meio de um processo de neutralização de pH simulado e uma bancada real de avaliação de compressores de refrigeração. Em ambos os casos, a estratégia para identificação de sistemas proposta apresentou melhores resultados que os métodos de comparação, que foram modelos lineares, modelos Hammerstein e máquinas de aprendizado extremo. Os resultados demonstraram que a estratégia proposta pode ser usada diretamente para obtenção de modelos multivariáveis não lineares utilizando um algoritmo de treinamento computacionalmente eficiente.

show abstract

Design of sparse Bayesian echo state network for time series prediction

Wang

Qiao

et al. 2020

Neural Comput & Applic

View full text Add to dashboard Cite

Online sequential echo state network with sparse RLS algorithm for time series prediction

Cited by 37 publications

References 33 publications

A Modified Recursive Regularization Factor Calculation for Sparse RLS Algorithm with l1-Norm

A Modified Recursive Regularization Factor Calculation for Sparse RLS Algorithm with l1-Norm

Identificação de Sistemas Dinâmicos Não Lineares Multivariáveis com Redes de Estado de Eco

Design of sparse Bayesian echo state network for time series prediction

Contact Info

Product

Resources

About