Title :
Classification of mental tasks using Gaussian mixture Bayesian network classifiers
Author :
Tavakolian, Kouhyar ; Rezaei, Saeid
Author_Institution :
Dept. of Comput. Sci., Northern British Columbia Univ., USA
Abstract :
In this work we consider classification of mental tasks from EEG signals by using Gaussian mixture models. For this purpose, we use Bayesian graphical networks (BNT). The final results for Bayesian graphical networks are compared with our previous results for the neural network classifier. The results show an improvement in both classification accuracy and consistency.
Keywords :
Gaussian processes; belief networks; electroencephalography; medical signal processing; signal classification; Bayesian graphical networks; Bayesian network classifiers; EEG; Gaussian mixture model; mental task classification; Bayesian methods; Biological neural networks; Brain computer interfaces; Brain modeling; Electroencephalography; Graphical models; Iterative algorithms; Mathematical model; Neural networks; Signal processing algorithms;
Conference_Titel :
Biomedical Circuits and Systems, 2004 IEEE International Workshop on
Print_ISBN :
0-7803-8665-5
DOI :
10.1109/BIOCAS.2004.1454169
