EEG Signals classification using linear and non-linear discriminant methods

This article was developed with the particular interest of characterize and study EEG signals as a pattern which in general has a high dimensionality, and has obviously a particular behavior in frequency and time. Here we have developed a wavelet decomposition to reduce a little bit the dimensionality and PCA (Principal Components Analysis) to accurate the result in a better way (only two features representation). After that the EEG signals, with their respective characteristics and representation has been able to train and test some linear and non-linear classifiers such as (Parzen, k-NN, Radial Basis Neural Network, linear and non-linear perceptron and so on.) This evaluation is an analysis of general EEG’s behavior signals with this kind of characterization and classification processes respectively.

• Autores: Mayor Torres, Juan Manuel
• Tipo de recurso: Article of journal
• Fecha de publicación: 2013
• Institución: Universidad Autónoma de Occidente
• Repositorio: RED: Repositorio Educativo Digital UAO
• Idioma: eng
• OAI Identifier: oai:red.uao.edu.co:10614/10822
• Acceso en línea: http://hdl.handle.net/10614/10822
• Palabra clave: linear classifiers; feature extraction; biomedical signals; feature extraction; Biomedical signals
• Rights: openAccess
• License: Derechos Reservados - Universidad Autónoma de Occidente