Presentation 2010-11-18
Comparison of Precision in Posterior Probability Estimate for Classification of Electroencephalogram : Comparison of support vector machine and relevance vector machine
Hiromu TAKAHASHI, Tomohiro YOSHIKAWA, Takeshi FURUHASHI,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) The posterior probability in classification is useful information, which can be used for the rejection of uncertain classification results, for example. Reliability-based automatic repeat request (RB-ARQ), proposed by the authors to improve the performance of the brain-computer interface, also utilizes it. This paper compares three types of classifiers: LDA, SVM, and relevance vector machine (RVM), in terms of the estimation accuracy of the posterior probability and the classification performance when the rejection and RB-ARQ are applied. The results show that LDA is the best for the P300 speller with RB-ARQ, and RVM is the best for a motor imagery task with the rejection.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Posterior probability / support vector machine / relevance vector machine / brain-computer interface
Paper # NC2010-61
Date of Issue

Conference Information
Committee NC
Conference Date 2010/11/11(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Neurocomputing (NC)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Comparison of Precision in Posterior Probability Estimate for Classification of Electroencephalogram : Comparison of support vector machine and relevance vector machine
Sub Title (in English)
Keyword(1) Posterior probability
Keyword(2) support vector machine
Keyword(3) relevance vector machine
Keyword(4) brain-computer interface
1st Author's Name Hiromu TAKAHASHI
1st Author's Affiliation Dept. of Computational Science and Engineering, Graduate School of Engineering, Nagoya University()
2nd Author's Name Tomohiro YOSHIKAWA
2nd Author's Affiliation Dept. of Computational Science and Engineering, Graduate School of Engineering, Nagoya University
3rd Author's Name Takeshi FURUHASHI
3rd Author's Affiliation Dept. of Computational Science and Engineering, Graduate School of Engineering, Nagoya University
Date 2010-11-18
Paper # NC2010-61
Volume (vol) vol.110
Number (no) 295
Page pp.pp.-
#Pages 6
Date of Issue