1-penalized linear mixed-effects models for BCI

Siamac Fazli, Márton Danóczy, Jürg Schelldorfer, Klaus Robert Müller

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

3 Citations (Scopus)

Abstract

A recently proposed novel statistical model estimates population effects and individual variability between subgroups simultaneously, by extending Lasso methods. We apply this ℓ1-penalized linear regression mixed-effects model to a large scale real world problem: by exploiting a large set of brain computer interface data we 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 able to differentiate within-subject and between-subject variability. A deeper understanding both of the underlying statistical and physiological structure of the data is gained.

Original languageEnglish
Title of host publicationArtificial Neural Networks and Machine Learning, ICANN 2011 - 21st International Conference on Artificial Neural Networks, Proceedings
Pages26-35
Number of pages10
EditionPART 1
DOIs
Publication statusPublished - Jun 24 2011
Externally publishedYes
Event21st International Conference on Artificial Neural Networks, ICANN 2011 - Espoo, Finland
Duration: Jun 14 2011Jun 17 2011

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
NumberPART 1
Volume6791 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference21st International Conference on Artificial Neural Networks, ICANN 2011
CountryFinland
CityEspoo
Period6/14/116/17/11

Fingerprint

Linear Mixed Effects Model
Regression Effects
Mixed Effects Model
Brain computer interface
Lasso
Training Algorithm
Differentiate
Linear regression
Large Set
Statistical Model
Classifiers
Classifier
Subgroup
Zero
Estimate
Model
Brain
Statistical Models

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Fazli, S., Danóczy, M., Schelldorfer, J., & Müller, K. R. (2011). 1-penalized linear mixed-effects models for BCI. In Artificial Neural Networks and Machine Learning, ICANN 2011 - 21st International Conference on Artificial Neural Networks, Proceedings (PART 1 ed., pp. 26-35). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 6791 LNCS, No. PART 1). https://doi.org/10.1007/978-3-642-21735-7_4

1-penalized linear mixed-effects models for BCI. / Fazli, Siamac; Danóczy, Márton; Schelldorfer, Jürg; Müller, Klaus Robert.

Artificial Neural Networks and Machine Learning, ICANN 2011 - 21st International Conference on Artificial Neural Networks, Proceedings. PART 1. ed. 2011. p. 26-35 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 6791 LNCS, No. PART 1).

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

Fazli, S, Danóczy, M, Schelldorfer, J & Müller, KR 2011, 1-penalized linear mixed-effects models for BCI. in Artificial Neural Networks and Machine Learning, ICANN 2011 - 21st International Conference on Artificial Neural Networks, Proceedings. PART 1 edn, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), no. PART 1, vol. 6791 LNCS, pp. 26-35, 21st International Conference on Artificial Neural Networks, ICANN 2011, Espoo, Finland, 6/14/11. https://doi.org/10.1007/978-3-642-21735-7_4
Fazli S, Danóczy M, Schelldorfer J, Müller KR. 1-penalized linear mixed-effects models for BCI. In Artificial Neural Networks and Machine Learning, ICANN 2011 - 21st International Conference on Artificial Neural Networks, Proceedings. PART 1 ed. 2011. p. 26-35. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 1). https://doi.org/10.1007/978-3-642-21735-7_4
Fazli, Siamac ; Danóczy, Márton ; Schelldorfer, Jürg ; Müller, Klaus Robert. / 1-penalized linear mixed-effects models for BCI. Artificial Neural Networks and Machine Learning, ICANN 2011 - 21st International Conference on Artificial Neural Networks, Proceedings. PART 1. ed. 2011. pp. 26-35 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 1).
@inproceedings{b669c174856b4beb855d1e0a49c9d989,
title = "ℓ1-penalized linear mixed-effects models for BCI",
abstract = "A recently proposed novel statistical model estimates population effects and individual variability between subgroups simultaneously, by extending Lasso methods. We apply this ℓ1-penalized linear regression mixed-effects model to a large scale real world problem: by exploiting a large set of brain computer interface data we 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 able to differentiate within-subject and between-subject variability. A deeper understanding both of the underlying statistical and physiological structure of the data is gained.",
author = "Siamac Fazli and M{\'a}rton Dan{\'o}czy and J{\"u}rg Schelldorfer and M{\"u}ller, {Klaus Robert}",
year = "2011",
month = "6",
day = "24",
doi = "10.1007/978-3-642-21735-7_4",
language = "English",
isbn = "9783642217340",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
number = "PART 1",
pages = "26--35",
booktitle = "Artificial Neural Networks and Machine Learning, ICANN 2011 - 21st International Conference on Artificial Neural Networks, Proceedings",
edition = "PART 1",

}

TY - GEN

T1 - ℓ1-penalized linear mixed-effects models for BCI

AU - Fazli, Siamac

AU - Danóczy, Márton

AU - Schelldorfer, Jürg

AU - Müller, Klaus Robert

PY - 2011/6/24

Y1 - 2011/6/24

N2 - A recently proposed novel statistical model estimates population effects and individual variability between subgroups simultaneously, by extending Lasso methods. We apply this ℓ1-penalized linear regression mixed-effects model to a large scale real world problem: by exploiting a large set of brain computer interface data we 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 able to differentiate within-subject and between-subject variability. A deeper understanding both of the underlying statistical and physiological structure of the data is gained.

AB - A recently proposed novel statistical model estimates population effects and individual variability between subgroups simultaneously, by extending Lasso methods. We apply this ℓ1-penalized linear regression mixed-effects model to a large scale real world problem: by exploiting a large set of brain computer interface data we 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 able to differentiate within-subject and between-subject variability. A deeper understanding both of the underlying statistical and physiological structure of the data is gained.

UR - http://www.scopus.com/inward/record.url?scp=79959338722&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=79959338722&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-21735-7_4

DO - 10.1007/978-3-642-21735-7_4

M3 - Conference contribution

SN - 9783642217340

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 26

EP - 35

BT - Artificial Neural Networks and Machine Learning, ICANN 2011 - 21st International Conference on Artificial Neural Networks, Proceedings

ER -