Stable and robust ICA algorithm based on t-distribution and generalized Gaussian distribution models

Jianting Cao, Noboru Murata

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

12 Citations (Scopus)

Abstract

In this paper, we propose a novel Independent Component Analysis (ICA) algorithm which enables to separate mixtures of sub-Gaussian, super-Gaussian and Gaussian primary source signals. Alternative activation functions in the algorithm are derived by using parameterized t-distribution and generalized Gaussian distribution density models. The functions are self-adaptive based on estimating the high-order moments of extracted signals. Moreover, a stability condition of the proposed algorithm for separating the true solution is given. Simulation experiment results are presented to illustrate the effectiveness and performance of the proposed algorithm.

Original languageEnglish
Title of host publicationNeural Networks for Signal Processing - Proceedings of the IEEE Workshop
Place of PublicationPiscataway, NJ, United States
PublisherIEEE
Pages283-292
Number of pages10
Publication statusPublished - 1999
Externally publishedYes
EventProceedings of the 1999 9th IEEE Workshop on Neural Networks for Signal Processing (NNSP'99) - Madison, WI, USA
Duration: 1999 Aug 231999 Aug 25

Other

OtherProceedings of the 1999 9th IEEE Workshop on Neural Networks for Signal Processing (NNSP'99)
CityMadison, WI, USA
Period99/8/2399/8/25

Fingerprint

Independent component analysis
Gaussian distribution
Chemical activation
Experiments

ASJC Scopus subject areas

  • Signal Processing
  • Software
  • Electrical and Electronic Engineering

Cite this

Cao, J., & Murata, N. (1999). Stable and robust ICA algorithm based on t-distribution and generalized Gaussian distribution models. In Neural Networks for Signal Processing - Proceedings of the IEEE Workshop (pp. 283-292). Piscataway, NJ, United States: IEEE.

Stable and robust ICA algorithm based on t-distribution and generalized Gaussian distribution models. / Cao, Jianting; Murata, Noboru.

Neural Networks for Signal Processing - Proceedings of the IEEE Workshop. Piscataway, NJ, United States : IEEE, 1999. p. 283-292.

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

Cao, J & Murata, N 1999, Stable and robust ICA algorithm based on t-distribution and generalized Gaussian distribution models. in Neural Networks for Signal Processing - Proceedings of the IEEE Workshop. IEEE, Piscataway, NJ, United States, pp. 283-292, Proceedings of the 1999 9th IEEE Workshop on Neural Networks for Signal Processing (NNSP'99), Madison, WI, USA, 99/8/23.
Cao J, Murata N. Stable and robust ICA algorithm based on t-distribution and generalized Gaussian distribution models. In Neural Networks for Signal Processing - Proceedings of the IEEE Workshop. Piscataway, NJ, United States: IEEE. 1999. p. 283-292
Cao, Jianting ; Murata, Noboru. / Stable and robust ICA algorithm based on t-distribution and generalized Gaussian distribution models. Neural Networks for Signal Processing - Proceedings of the IEEE Workshop. Piscataway, NJ, United States : IEEE, 1999. pp. 283-292
@inproceedings{070674528f6e436ea28a06c1c92139ef,
title = "Stable and robust ICA algorithm based on t-distribution and generalized Gaussian distribution models",
abstract = "In this paper, we propose a novel Independent Component Analysis (ICA) algorithm which enables to separate mixtures of sub-Gaussian, super-Gaussian and Gaussian primary source signals. Alternative activation functions in the algorithm are derived by using parameterized t-distribution and generalized Gaussian distribution density models. The functions are self-adaptive based on estimating the high-order moments of extracted signals. Moreover, a stability condition of the proposed algorithm for separating the true solution is given. Simulation experiment results are presented to illustrate the effectiveness and performance of the proposed algorithm.",
author = "Jianting Cao and Noboru Murata",
year = "1999",
language = "English",
pages = "283--292",
booktitle = "Neural Networks for Signal Processing - Proceedings of the IEEE Workshop",
publisher = "IEEE",

}

TY - GEN

T1 - Stable and robust ICA algorithm based on t-distribution and generalized Gaussian distribution models

AU - Cao, Jianting

AU - Murata, Noboru

PY - 1999

Y1 - 1999

N2 - In this paper, we propose a novel Independent Component Analysis (ICA) algorithm which enables to separate mixtures of sub-Gaussian, super-Gaussian and Gaussian primary source signals. Alternative activation functions in the algorithm are derived by using parameterized t-distribution and generalized Gaussian distribution density models. The functions are self-adaptive based on estimating the high-order moments of extracted signals. Moreover, a stability condition of the proposed algorithm for separating the true solution is given. Simulation experiment results are presented to illustrate the effectiveness and performance of the proposed algorithm.

AB - In this paper, we propose a novel Independent Component Analysis (ICA) algorithm which enables to separate mixtures of sub-Gaussian, super-Gaussian and Gaussian primary source signals. Alternative activation functions in the algorithm are derived by using parameterized t-distribution and generalized Gaussian distribution density models. The functions are self-adaptive based on estimating the high-order moments of extracted signals. Moreover, a stability condition of the proposed algorithm for separating the true solution is given. Simulation experiment results are presented to illustrate the effectiveness and performance of the proposed algorithm.

UR - http://www.scopus.com/inward/record.url?scp=0033357311&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0033357311&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:0033357311

SP - 283

EP - 292

BT - Neural Networks for Signal Processing - Proceedings of the IEEE Workshop

PB - IEEE

CY - Piscataway, NJ, United States

ER -