Classification of EEG Signals Using a Genetic-Based Machine Learning Classifier

Abstract

This paper investigates the efficacy of the genetic-based learning classifier system XCS, for the classification of noisy, artefact-inclusive human electroencephalogram (EEG) signals represented using large condition strings (108bits). EEG signals from three participants were recorded while they performed four mental tasks designed to elicit hemispheric responses. Autoregressive (AR) models and Fast Fourier Transform (FFT) methods were used to form feature vectors with which mental tasks can be discriminated. XCS achieved a maximum classification accuracy of 99.3% and a best average of 88.9%. The relative classification performance of XCS was then compared against four non-evolutionary classifier systems originating from different learning techniques. The experimental results will be used as part of our larger research effort investigating the feasibility of using EEG signals as an interface to allow paralysed persons to control a powered wheelchair or other devices.

Plausibility assessment of a 2-state self-paced mental task-based BCI using the no-control performance analysis

“…As a result of the above studies, the AR model order has usually been set to 6 (Anderson et al, 1995a;Anderson and Sijerčić, 1996;Anderson, 1997;Palaniappan et al, 2000bPalaniappan et al, , 2002Garrett et al, 2003;Liu et al, 2003b;Daud and Yunus, 2004;Palaniappan, 2005a;Huan and Palaniappan, 2005;Palaniappan and Huan, 2005;Rezaei et al, 2005;Skinner et al, 2007;Hosni et al, 2007). But we believe that the model order should be adjusted in a way that meets the requirements of the application.…”

“…A variety of studies such as Keirn and Aunon (1990b), Anderson et al (1995aAnderson et al ( ,b, 1998Anderson et al ( , 2006, Anderson and Sijerčić (1996), Anderson (1997), Palaniappan et al (2000aPalaniappan et al ( ,b,c, 2002, Palaniappan and Raveendran (2001), Bhatti et al (2001), Maiorescu et al (2003), Garrett et al (2003), Liu et al (2003aLiu et al ( ,b, 2005a, Wu and Guo (2003), Xue et al (2003), Barreto et al (2004), Daud and Yunus (2004), , , Rao and Derakhshani (2005), Palaniappan (2005aPalaniappan ( ,b, 2006, Huan and Palaniappan (2005), Palaniappan and Huan (2005), Rezaei et al (2005), Jiang et al (2005), Setban and Dobrea (2005), Gope et al (2005), Yan et al (2006), Abdollahi and Motie-Nasrabadi (2006), Nakayama and Inagaki (2006), , , Nakayama et al (2007), Zhiwei and Minfen (2007), Skinner et al (2007), Abdollahi et al (2007), Hema et al (2007), Hosni et al (2007), and Paulraj et al (2007) have employed this dataset. Most of them classify mental tasks to some extent, but only a few of them report the resultant false positive rate or the confusion matrix (Palaniappan, 2005a;…”

Driver inattention monitoring system for intelligent vehicles: A review

“…B.T. Skinner et al [16] classified the EEG signals into four mental tasks designed to elicit hemispheric responses with the genetic based learning classifier system. Many of the abovementioned experiments are conducted intrapersonal, that's why they can get high accuracy.…”

Enhancing the diversity of genetic algorithm for improved feature selection

“…It proved to give good results when applied to different applications [1][2][3][4][5][6]. One of these applications is the feature/variable selection problem [2,[7][8][9][10].…”