On robust parameter estimation in brain-computer interfacing

Wojciech Samek, Shinichi Nakajima, Motoaki Kawanabe, Klaus Muller

Research output: Contribution to journalReview article

7 Citations (Scopus)

Abstract

Objective. The reliable estimation of parameters such as mean or covariance matrix from noisy and high-dimensional observations is a prerequisite for successful application of signal processing and machine learning algorithms in brain-computer interfacing (BCI). This challenging task becomes significantly more difficult if the data set contains outliers, e.g. due to subject movements, eye blinks or loose electrodes, as they may heavily bias the estimation and the subsequent statistical analysis. Although various robust estimators have been developed to tackle the outlier problem, they ignore important structural information in the data and thus may not be optimal. Typical structural elements in BCI data are the trials consisting of a few hundred EEG samples and indicating the start and end of a task. Approach. This work discusses the parameter estimation problem in BCI and introduces a novel hierarchical view on robustness which naturally comprises different types of outlierness occurring in structured data. Furthermore, the class of minimum divergence estimators is reviewed and a robust mean and covariance estimator for structured data is derived and evaluated with simulations and on a benchmark data set. Main results. The results show that state-of-the-art BCI algorithms benefit from robustly estimated parameters. Significance. Since parameter estimation is an integral part of various machine learning algorithms, the presented techniques are applicable to many problems beyond BCI.

Original languageEnglish
Article number061001
JournalJournal of Neural Engineering
Volume14
Issue number6
DOIs
Publication statusPublished - 2017 Nov 23

Fingerprint

Parameter estimation
Brain
Learning algorithms
Learning systems
Benchmarking
Eye movements
Eye Movements
Electroencephalography
Covariance matrix
Statistical methods
Signal processing
Electrodes
Datasets
Machine Learning

Keywords

  • brain-computer interfacing
  • common spatial patterns
  • parameter estimation
  • robustness

ASJC Scopus subject areas

  • Biomedical Engineering
  • Cellular and Molecular Neuroscience

Cite this

On robust parameter estimation in brain-computer interfacing. / Samek, Wojciech; Nakajima, Shinichi; Kawanabe, Motoaki; Muller, Klaus.

In: Journal of Neural Engineering, Vol. 14, No. 6, 061001, 23.11.2017.

Research output: Contribution to journalReview article

Samek, Wojciech ; Nakajima, Shinichi ; Kawanabe, Motoaki ; Muller, Klaus. / On robust parameter estimation in brain-computer interfacing. In: Journal of Neural Engineering. 2017 ; Vol. 14, No. 6.
@article{4717ded342a84b9695123b561ab432e8,
title = "On robust parameter estimation in brain-computer interfacing",
abstract = "Objective. The reliable estimation of parameters such as mean or covariance matrix from noisy and high-dimensional observations is a prerequisite for successful application of signal processing and machine learning algorithms in brain-computer interfacing (BCI). This challenging task becomes significantly more difficult if the data set contains outliers, e.g. due to subject movements, eye blinks or loose electrodes, as they may heavily bias the estimation and the subsequent statistical analysis. Although various robust estimators have been developed to tackle the outlier problem, they ignore important structural information in the data and thus may not be optimal. Typical structural elements in BCI data are the trials consisting of a few hundred EEG samples and indicating the start and end of a task. Approach. This work discusses the parameter estimation problem in BCI and introduces a novel hierarchical view on robustness which naturally comprises different types of outlierness occurring in structured data. Furthermore, the class of minimum divergence estimators is reviewed and a robust mean and covariance estimator for structured data is derived and evaluated with simulations and on a benchmark data set. Main results. The results show that state-of-the-art BCI algorithms benefit from robustly estimated parameters. Significance. Since parameter estimation is an integral part of various machine learning algorithms, the presented techniques are applicable to many problems beyond BCI.",
keywords = "brain-computer interfacing, common spatial patterns, parameter estimation, robustness",
author = "Wojciech Samek and Shinichi Nakajima and Motoaki Kawanabe and Klaus Muller",
year = "2017",
month = "11",
day = "23",
doi = "10.1088/1741-2552/aa8232",
language = "English",
volume = "14",
journal = "Journal of Neural Engineering",
issn = "1741-2560",
publisher = "IOP Publishing Ltd.",
number = "6",

}

TY - JOUR

T1 - On robust parameter estimation in brain-computer interfacing

AU - Samek, Wojciech

AU - Nakajima, Shinichi

AU - Kawanabe, Motoaki

AU - Muller, Klaus

PY - 2017/11/23

Y1 - 2017/11/23

N2 - Objective. The reliable estimation of parameters such as mean or covariance matrix from noisy and high-dimensional observations is a prerequisite for successful application of signal processing and machine learning algorithms in brain-computer interfacing (BCI). This challenging task becomes significantly more difficult if the data set contains outliers, e.g. due to subject movements, eye blinks or loose electrodes, as they may heavily bias the estimation and the subsequent statistical analysis. Although various robust estimators have been developed to tackle the outlier problem, they ignore important structural information in the data and thus may not be optimal. Typical structural elements in BCI data are the trials consisting of a few hundred EEG samples and indicating the start and end of a task. Approach. This work discusses the parameter estimation problem in BCI and introduces a novel hierarchical view on robustness which naturally comprises different types of outlierness occurring in structured data. Furthermore, the class of minimum divergence estimators is reviewed and a robust mean and covariance estimator for structured data is derived and evaluated with simulations and on a benchmark data set. Main results. The results show that state-of-the-art BCI algorithms benefit from robustly estimated parameters. Significance. Since parameter estimation is an integral part of various machine learning algorithms, the presented techniques are applicable to many problems beyond BCI.

AB - Objective. The reliable estimation of parameters such as mean or covariance matrix from noisy and high-dimensional observations is a prerequisite for successful application of signal processing and machine learning algorithms in brain-computer interfacing (BCI). This challenging task becomes significantly more difficult if the data set contains outliers, e.g. due to subject movements, eye blinks or loose electrodes, as they may heavily bias the estimation and the subsequent statistical analysis. Although various robust estimators have been developed to tackle the outlier problem, they ignore important structural information in the data and thus may not be optimal. Typical structural elements in BCI data are the trials consisting of a few hundred EEG samples and indicating the start and end of a task. Approach. This work discusses the parameter estimation problem in BCI and introduces a novel hierarchical view on robustness which naturally comprises different types of outlierness occurring in structured data. Furthermore, the class of minimum divergence estimators is reviewed and a robust mean and covariance estimator for structured data is derived and evaluated with simulations and on a benchmark data set. Main results. The results show that state-of-the-art BCI algorithms benefit from robustly estimated parameters. Significance. Since parameter estimation is an integral part of various machine learning algorithms, the presented techniques are applicable to many problems beyond BCI.

KW - brain-computer interfacing

KW - common spatial patterns

KW - parameter estimation

KW - robustness

UR - http://www.scopus.com/inward/record.url?scp=85036464229&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85036464229&partnerID=8YFLogxK

U2 - 10.1088/1741-2552/aa8232

DO - 10.1088/1741-2552/aa8232

M3 - Review article

C2 - 28745300

AN - SCOPUS:85036464229

VL - 14

JO - Journal of Neural Engineering

JF - Journal of Neural Engineering

SN - 1741-2560

IS - 6

M1 - 061001

ER -