Maximal Margin Approach to Kernel Generalised Learning Vector Quantisation for Brain-Computer Interface

Trung Le, Dat Tran, Tuan Hoang, Dharmendra Sharma

Research output: A Conference proceeding or a Chapter in BookConference contribution

2 Downloads (Pure)

Abstract

Kernel Generalised Learning Vector Quantisation (KGLVQ) was proposed to extend Generalised Learning Vector Quantisation into the kernel feature space to deal with complex class boundaries and thus yield promising performance for complex classification tasks in pattern recognition. However KGLVQ does not follow the maximal margin principle which is crucial for kernel-based learning methods. In this paper we propose a maximal margin approach to Kernel Generalised Learning Vector Quantisation algorithm which inherits the merits of KGLVQ and follows the maximal margin principle to favour the generalisation capability. Experiments performed on the well-known data set III of BCI competition II show promising classification results for the proposed method.
Original languageEnglish
Title of host publicationInternational Conference on Neural Information Processing (ICONIP 2012)
Subtitle of host publicationLecture Notes in Computer Science
EditorsTingwen Huang, Zhigang Zeng, Chuandong Li, Chi Sing Leung
Place of PublicationGermany
PublisherSpringer
Pages191-198
Number of pages8
Volume7665
ISBN (Electronic)9783642344879
ISBN (Print)9783642344800
DOIs
Publication statusPublished - 2012
Event19th International Conference on Neural Information Processing 2012 - Doha, Doha, Qatar
Duration: 12 Nov 201215 Nov 2012

Conference

Conference19th International Conference on Neural Information Processing 2012
CountryQatar
CityDoha
Period12/11/1215/11/12

Fingerprint

Brain computer interface
Vector quantization
Pattern recognition
Experiments

Cite this

Le, T., Tran, D., Hoang, T., & Sharma, D. (2012). Maximal Margin Approach to Kernel Generalised Learning Vector Quantisation for Brain-Computer Interface. In T. Huang, Z. Zeng, C. Li, & C. S. Leung (Eds.), International Conference on Neural Information Processing (ICONIP 2012): Lecture Notes in Computer Science (Vol. 7665, pp. 191-198). Germany: Springer. https://doi.org/10.1007/978-3-642-34487-9_24
Le, Trung ; Tran, Dat ; Hoang, Tuan ; Sharma, Dharmendra. / Maximal Margin Approach to Kernel Generalised Learning Vector Quantisation for Brain-Computer Interface. International Conference on Neural Information Processing (ICONIP 2012): Lecture Notes in Computer Science. editor / Tingwen Huang ; Zhigang Zeng ; Chuandong Li ; Chi Sing Leung. Vol. 7665 Germany : Springer, 2012. pp. 191-198
@inproceedings{2c3e4195798d48d19ad9f8e7a33912e0,
title = "Maximal Margin Approach to Kernel Generalised Learning Vector Quantisation for Brain-Computer Interface",
abstract = "Kernel Generalised Learning Vector Quantisation (KGLVQ) was proposed to extend Generalised Learning Vector Quantisation into the kernel feature space to deal with complex class boundaries and thus yield promising performance for complex classification tasks in pattern recognition. However KGLVQ does not follow the maximal margin principle which is crucial for kernel-based learning methods. In this paper we propose a maximal margin approach to Kernel Generalised Learning Vector Quantisation algorithm which inherits the merits of KGLVQ and follows the maximal margin principle to favour the generalisation capability. Experiments performed on the well-known data set III of BCI competition II show promising classification results for the proposed method.",
keywords = "Generalised Learning Vector Quantisation, Brain-Computer Interface, Learning Vector Quantisation, Kernel Method, Maximising Margin",
author = "Trung Le and Dat Tran and Tuan Hoang and Dharmendra Sharma",
year = "2012",
doi = "10.1007/978-3-642-34487-9_24",
language = "English",
isbn = "9783642344800",
volume = "7665",
pages = "191--198",
editor = "Tingwen Huang and Zhigang Zeng and Chuandong Li and Leung, {Chi Sing}",
booktitle = "International Conference on Neural Information Processing (ICONIP 2012)",
publisher = "Springer",
address = "Netherlands",

}

Le, T, Tran, D, Hoang, T & Sharma, D 2012, Maximal Margin Approach to Kernel Generalised Learning Vector Quantisation for Brain-Computer Interface. in T Huang, Z Zeng, C Li & CS Leung (eds), International Conference on Neural Information Processing (ICONIP 2012): Lecture Notes in Computer Science. vol. 7665, Springer, Germany, pp. 191-198, 19th International Conference on Neural Information Processing 2012, Doha, Qatar, 12/11/12. https://doi.org/10.1007/978-3-642-34487-9_24

Maximal Margin Approach to Kernel Generalised Learning Vector Quantisation for Brain-Computer Interface. / Le, Trung; Tran, Dat; Hoang, Tuan; Sharma, Dharmendra.

International Conference on Neural Information Processing (ICONIP 2012): Lecture Notes in Computer Science. ed. / Tingwen Huang; Zhigang Zeng; Chuandong Li; Chi Sing Leung. Vol. 7665 Germany : Springer, 2012. p. 191-198.

Research output: A Conference proceeding or a Chapter in BookConference contribution

TY - GEN

T1 - Maximal Margin Approach to Kernel Generalised Learning Vector Quantisation for Brain-Computer Interface

AU - Le, Trung

AU - Tran, Dat

AU - Hoang, Tuan

AU - Sharma, Dharmendra

PY - 2012

Y1 - 2012

N2 - Kernel Generalised Learning Vector Quantisation (KGLVQ) was proposed to extend Generalised Learning Vector Quantisation into the kernel feature space to deal with complex class boundaries and thus yield promising performance for complex classification tasks in pattern recognition. However KGLVQ does not follow the maximal margin principle which is crucial for kernel-based learning methods. In this paper we propose a maximal margin approach to Kernel Generalised Learning Vector Quantisation algorithm which inherits the merits of KGLVQ and follows the maximal margin principle to favour the generalisation capability. Experiments performed on the well-known data set III of BCI competition II show promising classification results for the proposed method.

AB - Kernel Generalised Learning Vector Quantisation (KGLVQ) was proposed to extend Generalised Learning Vector Quantisation into the kernel feature space to deal with complex class boundaries and thus yield promising performance for complex classification tasks in pattern recognition. However KGLVQ does not follow the maximal margin principle which is crucial for kernel-based learning methods. In this paper we propose a maximal margin approach to Kernel Generalised Learning Vector Quantisation algorithm which inherits the merits of KGLVQ and follows the maximal margin principle to favour the generalisation capability. Experiments performed on the well-known data set III of BCI competition II show promising classification results for the proposed method.

KW - Generalised Learning Vector Quantisation

KW - Brain-Computer Interface

KW - Learning Vector Quantisation

KW - Kernel Method

KW - Maximising Margin

UR - https://link.springer.com/chapter/10.1007%2F978-3-642-34487-9_24

U2 - 10.1007/978-3-642-34487-9_24

DO - 10.1007/978-3-642-34487-9_24

M3 - Conference contribution

SN - 9783642344800

VL - 7665

SP - 191

EP - 198

BT - International Conference on Neural Information Processing (ICONIP 2012)

A2 - Huang, Tingwen

A2 - Zeng, Zhigang

A2 - Li, Chuandong

A2 - Leung, Chi Sing

PB - Springer

CY - Germany

ER -

Le T, Tran D, Hoang T, Sharma D. Maximal Margin Approach to Kernel Generalised Learning Vector Quantisation for Brain-Computer Interface. In Huang T, Zeng Z, Li C, Leung CS, editors, International Conference on Neural Information Processing (ICONIP 2012): Lecture Notes in Computer Science. Vol. 7665. Germany: Springer. 2012. p. 191-198 https://doi.org/10.1007/978-3-642-34487-9_24