Charles W. Anderson
Department of Computer Science, Colorado State University, Fort Collins, CO 80523
anderson@cs.colostate.edu
Abstract
Electroencephalogram (EEG) signals have been an important source of information for the study of underlying brain processes. This body of work is now providing a framework for the development of a new modality of human-computer interaction based on EEG. Current research in this area is limited to detecting a small number of mental states. In this article, EEG from one subject performing three mental tasks is classified by a neural network. Using a sixth-order AR (autoregressive) model of half-second windows of six-channel EEG, a classification accuracy of 89% on test data is achieved.