Feature LMS Algorithms

Diniz, Paulo S. R.; Yazdanpanah, Hamed; Lima, Markus V. S.

doi:10.1109/icassp.2018.8461674

Cited by 14 publications

(11 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Results: In the following experiments, the unknown lowpass transfer function of the system identification configuration has 20 coefficients, which are defined as [3]…”

Section: Remarkmentioning

confidence: 99%

“…Other sparsity types can be obtained from known relationships among the coefficients. It has been recently shown that prior awareness about such hidden sparsities can be exploited by using feature matrices

F \in R^{N - 1 \times N}

[3]. These matrices can be designed such as to penalise non‐sparse linear combinations of the unknown parameters [4], by employing, for example, prior knowledge about frequency response characteristics of the unknown systems.…”

Section: Introductionmentioning

confidence: 99%

“…Feature least-mean-square (F-LMS) adaptive filtering algorithms have generally been derived by a stochastic gradient approach [3,4]. This Letter advances a methodology devised by using a feature-based adaptive filter approach, in which the update equation is obtained as an approximate solution of an optimisation problem with an affine constraint.…”

mentioning

confidence: 99%

“…Note that if there is a prior knowledge about low-pass characteristics of the unknown plant, a possible choice for F could be [3]…”

mentioning

confidence: 99%

See 3 more Smart Citations

‐norm feature least mean square algorithm

et al. 2020

View full text Add to dashboard Cite

In many practical applications, systems and signals show energy concentration in a few coefficients. This prior knowledge can often be incorporated into algorithms designed for tasks such as compressive sensing and system identification. This Letter proposes a new least mean square (LMS)‐based algorithm that exploits the hidden sparsity of the system that the adaptive filter intends to estimate. The algorithm minimises the ℓ2‐norm of a linear transformation of the coefficient vector, using the minimum distortion principle. Simulation results demonstrate good performance of the proposed algorithm with respect to the LMS algorithm. In addition, a stochastic model of the advanced algorithm is proposed, which provides accurate mean‐square deviation and mean‐square error predictions.

show abstract

“…Results: In the following experiments, the unknown lowpass transfer function of the system identification configuration has 20 coefficients, which are defined as [3]…”

Section: Remarkmentioning

confidence: 99%

F \in R^{N - 1 \times N}

Section: Introductionmentioning

confidence: 99%

mentioning

confidence: 99%

“…Note that if there is a prior knowledge about low-pass characteristics of the unknown plant, a possible choice for F could be [3]…”

mentioning

confidence: 99%

See 2 more Smart Citations

‐norm feature least mean square algorithm

et al. 2020

View full text Add to dashboard Cite

show abstract

“…A primeira corresponde à família de algoritmos proportionate, em que os coeficientes do filtro são atualizados de maneira proporcional Gabriel S. Chaves, Markus V. S. Lima, Universidade Federal do Rio de Janeiro (UFRJ) / Escola Politécnica (Poli) / Programa de Engenharia Elétrica (PEE) / Sinais, Multimídia e Telecomunicações (SMT), Rio de Janeiro -RJ, Brasil, E-mails: {gabriel.chaves, markus.lima}@smt.ufrj.br. Tadeu N. Ferreira, Universidade Federal Fluminense (UFF) / Escola de Engenharia / Niterói -RJ, Brasil, E-mail: tadeu_ferreira@id.uff.br às suas magnitudes [4]- [7], enquanto a segunda corresponde a algoritmos que usam uma regularização promovedora de esparsidade nos coeficientes, tal qual a norma 1 ou uma aproximação da norma 0 [8]- [11].…”

Section: Introductionunclassified

O algoritmo PNLMS realmente explora a esparsidade dos coeficientes?

Chaves¹,

Lima²,

Ferreira³

2019

Anais De XXXVII Simpósio Brasileiro De Telecomunicações E Processamento De Sinais

View full text Add to dashboard Cite

Resumo-O algoritmo proportionate normalized least-meansquare (PNLMS) e suas variantes são frequentemente usados na literatura para atacar problemas relacionados à estimação/identificação de sistemas esparsos. O objetivo deste trabalho é mostrar que existem vários cenários esparsos em que o desempenho do algoritmo PNLMS é inferior ao do algoritmo NLMS. Para isso, apresentamos duas propriedades do algoritmo PNLMS e discutimos suas consequências práticas. A Propriedade I explica o que de fato é explorado pelo algoritmo PNLMS, enquanto a Propriedade II descreve o papel fundamental de uma boa inicialização dos coeficientes do filtro adaptativo para esse algoritmo. Diversos resultados numéricos são apresentados de forma a corroborar as propriedades discutidas. O algoritmo 0-NLMS é usado apenas para ilustrar o desempenho de um algoritmo que realmente explora a esparsidade dos coeficientes.

show abstract

Improved feature least mean square algorithm

Yazdanpanah

2022

Adaptive Control & Signal

Self Cite

View full text Add to dashboard Cite

In this paper, we propose the improved feature least-mean-square (IF-LMS) algorithm to exploit hidden sparsity in unknown systems. Recently, the feature least-mean-square (F-LMS) algorithm has been introduced, but its application is limited to particular systems since it uses predetermined feature matrices. However, the proposed IF-LMS algorithm utilizes the stochastic gradient descent (SGD) method to learn feature matrices; thus, it can be used in any system that the classical LMS algorithm is applicable. Hence, by employing a learnable feature matrix, the IF-LMS algorithm has a vast application area as compared to the F-LMS algorithm. Moreover, mathematically, we discuss some parameters of the IF-LMS algorithm. Simulation results, in synthetic and real-life scenarios, demonstrate that the IF-LMS algorithm has superior filtering accuracy to the well-known LMS algorithm.

show abstract

Feature LMS Algorithms

Cited by 14 publications

References 11 publications

‐norm feature least mean square algorithm

‐norm feature least mean square algorithm

O algoritmo PNLMS realmente explora a esparsidade dos coeficientes?

Improved feature least mean square algorithm

Contact Info

Product

Resources

About