A Log-Linearized Gaussian Mixture Network and Its Application to EEG Pattern Classification
feedforward neural networks
recurrent nerual networks
The present paper proposes a new probabilisticneural network (NN) that can estimate a posteriori probabilityfor a pattern classification problem. The structure of the proposednetwork is based on a statistical model composed by a mixtureof log-linearized Gaussian components. However, the forwardcalculation and the backward learning rule can be defined in thesame manner as the error backpropagation NN. In this paper, theproposed network is applied to the electroencephalogram (EEG)pattern classification problem. In the experiments, two types of aphotic stimulation, which are caused by eye opening/closing andartificial light, are used to collect the data to be classified. It isshown that the EEG signals can be classified successfully andthat the classification rates change depending on the number oftraining data and the dimension of the feature vectors.
IEEE Transactions on Systems, Man, and Cybernetics-Part C: Applications and Reviews
Copyright (c) 1999 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.