Acceleration of adaptive proximal forward-backward splitting method and its application to sparse system identification

Masao Yamagishi, Masahiro Yukawa, Isao Yamada

Research output: Chapter in Book/Report/Conference proceedingConference contribution

17 Citations (Scopus)

Abstract

In this paper, we propose an acceleration technique of the adaptive filtering scheme called adaptive proximal forward-backward splitting method. For accelerating the convergence rate, the proposed method includes a step to shift the current estimate in the direction of the difference between the current and previous estimates based on the Fast Iterative Shrinkage/Thresholding Algorithm (FISTA). The computational complexity for this additional step is fairly low compared to the overall complexity of the algorithm. As an example of the proposed method, we derive an acceleration of the composition of the Adaptively Weighted Soft-Thresholding (AWST) operator and the exponentially weighted adaptive parallel projection. AWST shrinks the estimated filter coefficients to zero for exploiting the sparsity of the system to be estimated and the exponentially weighted adaptive parallel projection algorithm realizes high accuracy by utilizing all available information at each iteration. This accelerated method improves the steady-state mismatch drastically with its convergence speed as fast as the proportionate affine projection algorithm.

Original languageEnglish
Title of host publicationICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
Pages4296-4299
Number of pages4
DOIs
Publication statusPublished - 2011
Externally publishedYes
Event36th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2011 - Prague, Czech Republic
Duration: 2011 May 222011 May 27

Other

Other36th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2011
CountryCzech Republic
CityPrague
Period11/5/2211/5/27

Fingerprint

Identification (control systems)
Adaptive filtering
Computational complexity
Chemical analysis

Keywords

  • Acceleration
  • parallel projection
  • proximal forward-backward splitting method
  • sparse system identification
  • variable metric

ASJC Scopus subject areas

  • Signal Processing
  • Software
  • Electrical and Electronic Engineering

Cite this

Yamagishi, M., Yukawa, M., & Yamada, I. (2011). Acceleration of adaptive proximal forward-backward splitting method and its application to sparse system identification. In ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings (pp. 4296-4299). [5947303] https://doi.org/10.1109/ICASSP.2011.5947303

Acceleration of adaptive proximal forward-backward splitting method and its application to sparse system identification. / Yamagishi, Masao; Yukawa, Masahiro; Yamada, Isao.

ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings. 2011. p. 4296-4299 5947303.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Yamagishi, M, Yukawa, M & Yamada, I 2011, Acceleration of adaptive proximal forward-backward splitting method and its application to sparse system identification. in ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings., 5947303, pp. 4296-4299, 36th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2011, Prague, Czech Republic, 11/5/22. https://doi.org/10.1109/ICASSP.2011.5947303
Yamagishi M, Yukawa M, Yamada I. Acceleration of adaptive proximal forward-backward splitting method and its application to sparse system identification. In ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings. 2011. p. 4296-4299. 5947303 https://doi.org/10.1109/ICASSP.2011.5947303
Yamagishi, Masao ; Yukawa, Masahiro ; Yamada, Isao. / Acceleration of adaptive proximal forward-backward splitting method and its application to sparse system identification. ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings. 2011. pp. 4296-4299
@inproceedings{b456bc066ce8452782e1b58592dd99ea,
title = "Acceleration of adaptive proximal forward-backward splitting method and its application to sparse system identification",
abstract = "In this paper, we propose an acceleration technique of the adaptive filtering scheme called adaptive proximal forward-backward splitting method. For accelerating the convergence rate, the proposed method includes a step to shift the current estimate in the direction of the difference between the current and previous estimates based on the Fast Iterative Shrinkage/Thresholding Algorithm (FISTA). The computational complexity for this additional step is fairly low compared to the overall complexity of the algorithm. As an example of the proposed method, we derive an acceleration of the composition of the Adaptively Weighted Soft-Thresholding (AWST) operator and the exponentially weighted adaptive parallel projection. AWST shrinks the estimated filter coefficients to zero for exploiting the sparsity of the system to be estimated and the exponentially weighted adaptive parallel projection algorithm realizes high accuracy by utilizing all available information at each iteration. This accelerated method improves the steady-state mismatch drastically with its convergence speed as fast as the proportionate affine projection algorithm.",
keywords = "Acceleration, parallel projection, proximal forward-backward splitting method, sparse system identification, variable metric",
author = "Masao Yamagishi and Masahiro Yukawa and Isao Yamada",
year = "2011",
doi = "10.1109/ICASSP.2011.5947303",
language = "English",
isbn = "9781457705397",
pages = "4296--4299",
booktitle = "ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings",

}

TY - GEN

T1 - Acceleration of adaptive proximal forward-backward splitting method and its application to sparse system identification

AU - Yamagishi, Masao

AU - Yukawa, Masahiro

AU - Yamada, Isao

PY - 2011

Y1 - 2011

N2 - In this paper, we propose an acceleration technique of the adaptive filtering scheme called adaptive proximal forward-backward splitting method. For accelerating the convergence rate, the proposed method includes a step to shift the current estimate in the direction of the difference between the current and previous estimates based on the Fast Iterative Shrinkage/Thresholding Algorithm (FISTA). The computational complexity for this additional step is fairly low compared to the overall complexity of the algorithm. As an example of the proposed method, we derive an acceleration of the composition of the Adaptively Weighted Soft-Thresholding (AWST) operator and the exponentially weighted adaptive parallel projection. AWST shrinks the estimated filter coefficients to zero for exploiting the sparsity of the system to be estimated and the exponentially weighted adaptive parallel projection algorithm realizes high accuracy by utilizing all available information at each iteration. This accelerated method improves the steady-state mismatch drastically with its convergence speed as fast as the proportionate affine projection algorithm.

AB - In this paper, we propose an acceleration technique of the adaptive filtering scheme called adaptive proximal forward-backward splitting method. For accelerating the convergence rate, the proposed method includes a step to shift the current estimate in the direction of the difference between the current and previous estimates based on the Fast Iterative Shrinkage/Thresholding Algorithm (FISTA). The computational complexity for this additional step is fairly low compared to the overall complexity of the algorithm. As an example of the proposed method, we derive an acceleration of the composition of the Adaptively Weighted Soft-Thresholding (AWST) operator and the exponentially weighted adaptive parallel projection. AWST shrinks the estimated filter coefficients to zero for exploiting the sparsity of the system to be estimated and the exponentially weighted adaptive parallel projection algorithm realizes high accuracy by utilizing all available information at each iteration. This accelerated method improves the steady-state mismatch drastically with its convergence speed as fast as the proportionate affine projection algorithm.

KW - Acceleration

KW - parallel projection

KW - proximal forward-backward splitting method

KW - sparse system identification

KW - variable metric

UR - http://www.scopus.com/inward/record.url?scp=80051627966&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=80051627966&partnerID=8YFLogxK

U2 - 10.1109/ICASSP.2011.5947303

DO - 10.1109/ICASSP.2011.5947303

M3 - Conference contribution

AN - SCOPUS:80051627966

SN - 9781457705397

SP - 4296

EP - 4299

BT - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings

ER -