• JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
 
  Bookmark and Share
 
 
Master's Dissertation
DOI
https://doi.org/10.11606/D.55.2017.tde-21092017-165153
Document
Author
Full name
Paul Augusto Bustios Belizario
E-mail
Institute/School/College
Knowledge Area
Date of Defense
Published
São Carlos, 2017
Supervisor
Committee
Rosa, João Luis Garcia (President)
Carvalho, André Carlos Ponce de Leon Ferreira de
Hruschka Júnior, Estevam Rafael
Silva Filho, Antonio Carlos Roque da
Title in Portuguese
Seleção de bandas de frequência na classificação de eletroencefalogramas de imagética motora
Keywords in Portuguese
Classificação
Eletroencefalograma
Imagética motora
Abstract in Portuguese
Imagética motora é um processo mental que produz modulações na amplitude dos sinas de eletroencefalogramas em progresso. Os padrões presentes nestas modulações podem ser usados para classificar este processo mental, mas a identificação destes padrões não é uma tarefa trivial, porque eles estão presentes em bandas de frequências que são específicas para cada pessoa. Neste trabalho, apresenta-se um novo método para selecionar as bandas de frequência específicas para cada pessoa baseado na arquitetura do método Filter Bank Common Spatial Pattern. Para selecionar as bandas de frequência mais relevantes para cada pessoa, o método proposto aplica uma busca exaustiva para encontrar o melhor subconjunto de bandas de frequência contendo os padrões mais discriminativos dentro de um espaço de busca restrito a um tamanho fixo para este subconjunto. Esse tamanho é determinado usando validação cruzada e o método Sequential Forward Floating Selection. O método proposto foi avaliado usando a base de dados pública 2b da BCI Competition IV, mostrando melhores resultados do que todos os métodos também avaliados nessa base de dados.
Title in English
Selection of frequency bands in the classification of motor imagery electroencephalograms
Keywords in English
Classification
Electroencephalogram
Motor imagery
Abstract in English
Motor imagery is a mental process that when performed, produces modulations in the amplitude of ongoing electroencephalogram signals. These modulations happen following a series of patterns that can be used to classify this mental process, but the detection of those patterns is not a trivial task, because they occur in frequency bands that are specific for each person. In this work, we present a method to select these subject-specific frequency bands based on the arquitecture of the Filter Bank Common Spatial Pattern approach. To select the most relevant frequency bands for each person, our method uses an exhaustive search to find the best subset of frequency bands containing the most discriminative patterns, but with one restriction, the search space is restricted to find a subset with a fixed number of frequency bands. The number is determined using cross-validation and the Sequential Forward Floating Selection method. We demonstrate that, using the data set 2b of the BCI Competition IV, our method is more accurate than current methods evaluated on the same data set.
 
WARNING - Viewing this document is conditioned on your acceptance of the following terms of use:
This document is only for private use for research and teaching activities. Reproduction for commercial use is forbidden. This rights cover the whole data about this document as well as its contents. Any uses or copies of this document in whole or in part must include the author's name.
Publishing Date
2017-09-21
 
WARNING: Learn what derived works are clicking here.
All rights of the thesis/dissertation are from the authors
CeTI-SC/STI
Digital Library of Theses and Dissertations of USP. Copyright © 2001-2024. All rights reserved.