Separation of stationary and non-stationary sources with a generalized eigenvalue problem

Satoshi Hara, Yoshinobu Kawahara, Takashi Washio, Paul von Bünau, Terumasa Tokunaga, Kiyohumi Yumoto

Research output: Contribution to journalArticle

25 Citations (Scopus)

Abstract

Non-stationary effects are ubiquitous in real world data. In many settings, the observed signals are a mixture of underlying stationary and non-stationary sources that cannot be measured directly. For example, in EEG analysis, electrodes on the scalp record the activity from several sources located inside the brain, which one could only measure invasively. Discerning stationary and non-stationary contributions is an important step towards uncovering the mechanisms of the data generating system. To that end, in Stationary Subspace Analysis (SSA), the observed signal is modeled as a linear superposition of stationary and non-stationary sources, where the aim is to separate the two groups in the mixture. In this paper, we propose the first SSA algorithm that has a closed form solution. The novel method, Analytic SSA (ASSA), is more than 100 times faster than the state-of-the-art, numerically stable, and guaranteed to be optimal when the covariance between stationary and non-stationary sources is time-constant. In numerical simulations on wide range of settings, we show that our method yields superior results, even for signals with time-varying group-wise covariance. In an application to geophysical data analysis, ASSA extracts meaningful components that shed new light on the Pi 2 pulsations of the geomagnetic field.

Original languageEnglish
Pages (from-to)7-20
Number of pages14
JournalNeural Networks
Volume33
DOIs
Publication statusPublished - Sep 1 2012
Externally publishedYes

Fingerprint

Electroencephalography
Scalp
Information Systems
Brain
Electrodes
Computer simulation

All Science Journal Classification (ASJC) codes

  • Cognitive Neuroscience
  • Artificial Intelligence

Cite this

Separation of stationary and non-stationary sources with a generalized eigenvalue problem. / Hara, Satoshi; Kawahara, Yoshinobu; Washio, Takashi; von Bünau, Paul; Tokunaga, Terumasa; Yumoto, Kiyohumi.

In: Neural Networks, Vol. 33, 01.09.2012, p. 7-20.

Research output: Contribution to journalArticle

Hara, Satoshi ; Kawahara, Yoshinobu ; Washio, Takashi ; von Bünau, Paul ; Tokunaga, Terumasa ; Yumoto, Kiyohumi. / Separation of stationary and non-stationary sources with a generalized eigenvalue problem. In: Neural Networks. 2012 ; Vol. 33. pp. 7-20.
@article{7d78f4fb792e4a3680c74ef152eb23bd,
title = "Separation of stationary and non-stationary sources with a generalized eigenvalue problem",
abstract = "Non-stationary effects are ubiquitous in real world data. In many settings, the observed signals are a mixture of underlying stationary and non-stationary sources that cannot be measured directly. For example, in EEG analysis, electrodes on the scalp record the activity from several sources located inside the brain, which one could only measure invasively. Discerning stationary and non-stationary contributions is an important step towards uncovering the mechanisms of the data generating system. To that end, in Stationary Subspace Analysis (SSA), the observed signal is modeled as a linear superposition of stationary and non-stationary sources, where the aim is to separate the two groups in the mixture. In this paper, we propose the first SSA algorithm that has a closed form solution. The novel method, Analytic SSA (ASSA), is more than 100 times faster than the state-of-the-art, numerically stable, and guaranteed to be optimal when the covariance between stationary and non-stationary sources is time-constant. In numerical simulations on wide range of settings, we show that our method yields superior results, even for signals with time-varying group-wise covariance. In an application to geophysical data analysis, ASSA extracts meaningful components that shed new light on the Pi 2 pulsations of the geomagnetic field.",
author = "Satoshi Hara and Yoshinobu Kawahara and Takashi Washio and {von B{\"u}nau}, Paul and Terumasa Tokunaga and Kiyohumi Yumoto",
year = "2012",
month = "9",
day = "1",
doi = "10.1016/j.neunet.2012.04.001",
language = "English",
volume = "33",
pages = "7--20",
journal = "Neural Networks",
issn = "0893-6080",
publisher = "Elsevier Limited",

}

TY - JOUR

T1 - Separation of stationary and non-stationary sources with a generalized eigenvalue problem

AU - Hara, Satoshi

AU - Kawahara, Yoshinobu

AU - Washio, Takashi

AU - von Bünau, Paul

AU - Tokunaga, Terumasa

AU - Yumoto, Kiyohumi

PY - 2012/9/1

Y1 - 2012/9/1

N2 - Non-stationary effects are ubiquitous in real world data. In many settings, the observed signals are a mixture of underlying stationary and non-stationary sources that cannot be measured directly. For example, in EEG analysis, electrodes on the scalp record the activity from several sources located inside the brain, which one could only measure invasively. Discerning stationary and non-stationary contributions is an important step towards uncovering the mechanisms of the data generating system. To that end, in Stationary Subspace Analysis (SSA), the observed signal is modeled as a linear superposition of stationary and non-stationary sources, where the aim is to separate the two groups in the mixture. In this paper, we propose the first SSA algorithm that has a closed form solution. The novel method, Analytic SSA (ASSA), is more than 100 times faster than the state-of-the-art, numerically stable, and guaranteed to be optimal when the covariance between stationary and non-stationary sources is time-constant. In numerical simulations on wide range of settings, we show that our method yields superior results, even for signals with time-varying group-wise covariance. In an application to geophysical data analysis, ASSA extracts meaningful components that shed new light on the Pi 2 pulsations of the geomagnetic field.

AB - Non-stationary effects are ubiquitous in real world data. In many settings, the observed signals are a mixture of underlying stationary and non-stationary sources that cannot be measured directly. For example, in EEG analysis, electrodes on the scalp record the activity from several sources located inside the brain, which one could only measure invasively. Discerning stationary and non-stationary contributions is an important step towards uncovering the mechanisms of the data generating system. To that end, in Stationary Subspace Analysis (SSA), the observed signal is modeled as a linear superposition of stationary and non-stationary sources, where the aim is to separate the two groups in the mixture. In this paper, we propose the first SSA algorithm that has a closed form solution. The novel method, Analytic SSA (ASSA), is more than 100 times faster than the state-of-the-art, numerically stable, and guaranteed to be optimal when the covariance between stationary and non-stationary sources is time-constant. In numerical simulations on wide range of settings, we show that our method yields superior results, even for signals with time-varying group-wise covariance. In an application to geophysical data analysis, ASSA extracts meaningful components that shed new light on the Pi 2 pulsations of the geomagnetic field.

UR - http://www.scopus.com/inward/record.url?scp=84863869958&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84863869958&partnerID=8YFLogxK

U2 - 10.1016/j.neunet.2012.04.001

DO - 10.1016/j.neunet.2012.04.001

M3 - Article

C2 - 22551683

AN - SCOPUS:84863869958

VL - 33

SP - 7

EP - 20

JO - Neural Networks

JF - Neural Networks

SN - 0893-6080

ER -