Set-theoretic adaptive filtering based on data-driven sparsification

Masahiro Yukawa, Isao Yamada

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

In this article, we propose a fast and efficient algorithm named the adaptive parallel Krylov-metric projection algorithm. The proposed algorithm is derived from the variable-metric adaptive projected subgradient method, which has recently been presented as a unified analytic tool for various adaptive filtering algorithms. The proposed algorithm features parallel projection-in a variable-metric sense-onto multiple closed convex sets containing the optimal filter with high probability. The metric is designed based on (i) sparsification by means of a certain data-dependent Krylov subspace and (ii) maximal use of the obtained sparse structure for fast convergence. The numerical examples show the advantages of the proposed algorithm over the existing ones in stationary/nonstationary environments.

Original languageEnglish
Pages (from-to)707-722
Number of pages16
JournalInternational Journal of Adaptive Control and Signal Processing
Volume25
Issue number8
DOIs
Publication statusPublished - 2011 Aug
Externally publishedYes

Fingerprint

Adaptive filtering

Keywords

  • set-theoretic adaptive filtering
  • subgradient projection
  • variable-metric projection

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Electrical and Electronic Engineering
  • Signal Processing

Cite this

Set-theoretic adaptive filtering based on data-driven sparsification. / Yukawa, Masahiro; Yamada, Isao.

In: International Journal of Adaptive Control and Signal Processing, Vol. 25, No. 8, 08.2011, p. 707-722.

Research output: Contribution to journalArticle

@article{9dd1417bf43f4b22844e7f91d3e61a31,
title = "Set-theoretic adaptive filtering based on data-driven sparsification",
abstract = "In this article, we propose a fast and efficient algorithm named the adaptive parallel Krylov-metric projection algorithm. The proposed algorithm is derived from the variable-metric adaptive projected subgradient method, which has recently been presented as a unified analytic tool for various adaptive filtering algorithms. The proposed algorithm features parallel projection-in a variable-metric sense-onto multiple closed convex sets containing the optimal filter with high probability. The metric is designed based on (i) sparsification by means of a certain data-dependent Krylov subspace and (ii) maximal use of the obtained sparse structure for fast convergence. The numerical examples show the advantages of the proposed algorithm over the existing ones in stationary/nonstationary environments.",
keywords = "set-theoretic adaptive filtering, subgradient projection, variable-metric projection",
author = "Masahiro Yukawa and Isao Yamada",
year = "2011",
month = "8",
doi = "10.1002/acs.1237",
language = "English",
volume = "25",
pages = "707--722",
journal = "International Journal of Adaptive Control and Signal Processing",
issn = "0890-6327",
publisher = "John Wiley and Sons Ltd",
number = "8",

}

TY - JOUR

T1 - Set-theoretic adaptive filtering based on data-driven sparsification

AU - Yukawa, Masahiro

AU - Yamada, Isao

PY - 2011/8

Y1 - 2011/8

N2 - In this article, we propose a fast and efficient algorithm named the adaptive parallel Krylov-metric projection algorithm. The proposed algorithm is derived from the variable-metric adaptive projected subgradient method, which has recently been presented as a unified analytic tool for various adaptive filtering algorithms. The proposed algorithm features parallel projection-in a variable-metric sense-onto multiple closed convex sets containing the optimal filter with high probability. The metric is designed based on (i) sparsification by means of a certain data-dependent Krylov subspace and (ii) maximal use of the obtained sparse structure for fast convergence. The numerical examples show the advantages of the proposed algorithm over the existing ones in stationary/nonstationary environments.

AB - In this article, we propose a fast and efficient algorithm named the adaptive parallel Krylov-metric projection algorithm. The proposed algorithm is derived from the variable-metric adaptive projected subgradient method, which has recently been presented as a unified analytic tool for various adaptive filtering algorithms. The proposed algorithm features parallel projection-in a variable-metric sense-onto multiple closed convex sets containing the optimal filter with high probability. The metric is designed based on (i) sparsification by means of a certain data-dependent Krylov subspace and (ii) maximal use of the obtained sparse structure for fast convergence. The numerical examples show the advantages of the proposed algorithm over the existing ones in stationary/nonstationary environments.

KW - set-theoretic adaptive filtering

KW - subgradient projection

KW - variable-metric projection

UR - http://www.scopus.com/inward/record.url?scp=79960737577&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=79960737577&partnerID=8YFLogxK

U2 - 10.1002/acs.1237

DO - 10.1002/acs.1237

M3 - Article

VL - 25

SP - 707

EP - 722

JO - International Journal of Adaptive Control and Signal Processing

JF - International Journal of Adaptive Control and Signal Processing

SN - 0890-6327

IS - 8

ER -