Session:

LBP - Late Breaking Papers

Title:

Feature Selection and Classification in Brain Computer Interfaces by a Genetic Algorithm

Authors:

Luca Citi
Riccardo Poli
Caterina Cinel
Francisco Sepulveda

Abstract:

In this paper we explore the use of evolutionary algorithms in a wrapper-based selection of features and the classification of P300 signals in Brain Computer Interfaces. In particular we focus on a paradigm that uses the P300 potential associated to particular visual stimuli for hands free text entering. In our experiments the GA has found new ways to process and combine EEG signals to improve P300 detection accuracy.