Feature selection and blind source separation in an EEG-based brain-computer interface

Authors:
David A. Peterson;James N. Knight;Michael J. Kirby;Charles W. Anderson;Michael H. Thaut
Affiliations:
Department of Computer Science, Center for Biomedical Research in Music, Molecular, Cellular, and Integrative Neurosciences Program, and Department of Psychology, Colorado State University, Fort C ...;Department of Computer Science, Colorado State University, Fort Collins, CO;Department of Mathematics, Colorado State University, Fort Collins, CO;Department of Computer Science and Molecular, Cellular, and Integrative Neurosciences Program, Colorado State University, Fort Collins, CO;Center for Biomedical Research in Music and Molecular, Cellular, and Integrative Neurosciences Program, Colorado State University, Fort Collins, CO
Venue:
EURASIP Journal on Applied Signal Processing
Year:
2005

Citing 18
Cited 4

A note on genetic algorithms for large-scale feature selection

Pattern Recognition Letters
Adaptation in natural and artificial systems

Adaptation in natural and artificial systems
Floating search methods in feature selection

Pattern Recognition Letters
An information-maximization approach to blind separation and blind deconvolution

Neural Computation
Feature Selection: Evaluation, Application, and Small Sample Performance

IEEE Transactions on Pattern Analysis and Machine Intelligence
Selection of relevant features and examples in machine learning

Artificial Intelligence - Special issue on relevance
Wrappers for feature subset selection

Artificial Intelligence - Special issue on relevance
Support Vector Machines for 3D Object Recognition

IEEE Transactions on Pattern Analysis and Machine Intelligence
Advances in kernel methods: support vector learning

Advances in kernel methods: support vector learning
Independent component analysis: algorithms and applications

Neural Networks
Genetic Algorithms in Search, Optimization and Machine Learning

Genetic Algorithms in Search, Optimization and Machine Learning
Gene Selection for Cancer Classification using Support Vector Machines

Machine Learning
Blind source separation using the maximum signal fraction approach

Signal Processing
Text Categorization with Suport Vector Machines: Learning with Many Relevant Features

ECML '98 Proceedings of the 10th European Conference on Machine Learning
An introduction to variable and feature selection

The Journal of Machine Learning Research
Overfitting in making comparisons between variable selection methods

The Journal of Machine Learning Research
Use of the zero norm with linear models and kernel methods

The Journal of Machine Learning Research
New Support Vector Algorithms

Neural Computation

Optimal classification of epileptic seizures in EEG using wavelet analysis and genetic algorithm

Signal Processing
Microarray data classification based on ensemble independent component selection

Computers in Biology and Medicine
A GA based approach to improving the ICA based classification models for tumor classification

WSEAS Transactions on Information Science and Applications
Automatic and Adaptive Classification of Electroencephalographic Signals for Brain Computer Interfaces

Journal of Medical Systems

Quantified Score

Hi-index	0.00

Visualization

Abstract

Most EEG-based BCI systems make use of well-studied patterns of brain activity. However, those systems involve tasks that indirectly map to simple binary commands such as "yes" or "no" or require many weeks of biofeedback training. We hypothesized that signal processing and machine learning methods can be used to discriminate EEG in a direct "yes"/"no" BCI from a single session. Blind source separation (BSS) and spectral transformations of the EEG produced a 180-dimensional feature space. We used a modified genetic algorithm (GA) wrapped around a support vector machine (SVM) classifier to search the space of feature subsets. The GA-based search found feature subsets that outperform full feature sets and random feature subsets. Also, BSS transformations of the EEG outperformed the original time series, particularly in conjunction with a subset search of both spaces. The results suggest that BSS and feature selection can be used to improve the performance of even a "direct," single-session BCI.