ℓ1-penalized linear mixed-effects models for high dimensional data with application to BCI

Siamac Fazli, Márton Danóczy, Jürg Schelldorfer, Klaus Muller

Research output: Contribution to journalArticle

31 Citations (Scopus)

Abstract

Recently, a novel statistical model has been proposed to estimate population effects and individual variability between subgroups simultaneously, by extending Lasso methods. We will for the first time apply this so-called ℓ 1-penalized linear regression mixed-effects model for a large scale real world problem: we study a large set of brain computer interface data and through the novel estimator are able to obtain a subject-independent classifier that compares favorably with prior zero-training algorithms. This unifying model inherently compensates shifts in the input space attributed to the individuality of a subject. In particular we are now for the first time able to differentiate within-subject and between-subject variability. Thus a deeper understanding both of the underlying statistical and physiological structures of the data is gained.

Original languageEnglish
Pages (from-to)2100-2108
Number of pages9
JournalNeuroImage
Volume56
Issue number4
DOIs
Publication statusPublished - 2011 Jun 15
Externally publishedYes

Fingerprint

Brain-Computer Interfaces
Statistical Models
Individuality
Linear Models
Population
alachlor

Keywords

  • BCI
  • Mixed-effects model
  • Sparsity
  • Subject-independent

ASJC Scopus subject areas

  • Cognitive Neuroscience
  • Neurology

Cite this

ℓ1-penalized linear mixed-effects models for high dimensional data with application to BCI. / Fazli, Siamac; Danóczy, Márton; Schelldorfer, Jürg; Muller, Klaus.

In: NeuroImage, Vol. 56, No. 4, 15.06.2011, p. 2100-2108.

Research output: Contribution to journalArticle

Fazli, Siamac ; Danóczy, Márton ; Schelldorfer, Jürg ; Muller, Klaus. / ℓ1-penalized linear mixed-effects models for high dimensional data with application to BCI. In: NeuroImage. 2011 ; Vol. 56, No. 4. pp. 2100-2108.
@article{2acbe631a216410cb977fe31bee749ab,
title = "ℓ1-penalized linear mixed-effects models for high dimensional data with application to BCI",
abstract = "Recently, a novel statistical model has been proposed to estimate population effects and individual variability between subgroups simultaneously, by extending Lasso methods. We will for the first time apply this so-called ℓ 1-penalized linear regression mixed-effects model for a large scale real world problem: we study a large set of brain computer interface data and through the novel estimator are able to obtain a subject-independent classifier that compares favorably with prior zero-training algorithms. This unifying model inherently compensates shifts in the input space attributed to the individuality of a subject. In particular we are now for the first time able to differentiate within-subject and between-subject variability. Thus a deeper understanding both of the underlying statistical and physiological structures of the data is gained.",
keywords = "BCI, Mixed-effects model, Sparsity, Subject-independent",
author = "Siamac Fazli and M{\'a}rton Dan{\'o}czy and J{\"u}rg Schelldorfer and Klaus Muller",
year = "2011",
month = "6",
day = "15",
doi = "10.1016/j.neuroimage.2011.03.061",
language = "English",
volume = "56",
pages = "2100--2108",
journal = "NeuroImage",
issn = "1053-8119",
publisher = "Academic Press Inc.",
number = "4",

}

TY - JOUR

T1 - ℓ1-penalized linear mixed-effects models for high dimensional data with application to BCI

AU - Fazli, Siamac

AU - Danóczy, Márton

AU - Schelldorfer, Jürg

AU - Muller, Klaus

PY - 2011/6/15

Y1 - 2011/6/15

N2 - Recently, a novel statistical model has been proposed to estimate population effects and individual variability between subgroups simultaneously, by extending Lasso methods. We will for the first time apply this so-called ℓ 1-penalized linear regression mixed-effects model for a large scale real world problem: we study a large set of brain computer interface data and through the novel estimator are able to obtain a subject-independent classifier that compares favorably with prior zero-training algorithms. This unifying model inherently compensates shifts in the input space attributed to the individuality of a subject. In particular we are now for the first time able to differentiate within-subject and between-subject variability. Thus a deeper understanding both of the underlying statistical and physiological structures of the data is gained.

AB - Recently, a novel statistical model has been proposed to estimate population effects and individual variability between subgroups simultaneously, by extending Lasso methods. We will for the first time apply this so-called ℓ 1-penalized linear regression mixed-effects model for a large scale real world problem: we study a large set of brain computer interface data and through the novel estimator are able to obtain a subject-independent classifier that compares favorably with prior zero-training algorithms. This unifying model inherently compensates shifts in the input space attributed to the individuality of a subject. In particular we are now for the first time able to differentiate within-subject and between-subject variability. Thus a deeper understanding both of the underlying statistical and physiological structures of the data is gained.

KW - BCI

KW - Mixed-effects model

KW - Sparsity

KW - Subject-independent

UR - http://www.scopus.com/inward/record.url?scp=79957495136&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=79957495136&partnerID=8YFLogxK

U2 - 10.1016/j.neuroimage.2011.03.061

DO - 10.1016/j.neuroimage.2011.03.061

M3 - Article

VL - 56

SP - 2100

EP - 2108

JO - NeuroImage

JF - NeuroImage

SN - 1053-8119

IS - 4

ER -