G SALIMI KHORSHID
ALI M. NASRABADI
M R. HASHEMI GOLPAYEGANI

Finding a reliable BCI system to provide an accurate discrimination between different states of the brain activity is one of the most important studies in biomedical researches. A common approach in BCI systems is to use ERP components; like P300, as a tool for receiving the commands from a subject. So the BCI system design will reduce to ERP component detection task. In this study, a new artificial neural network based on neuronal recurrency will be used to detect P300. Compared with some other ANN- based classifiers, this technique yields a good accuracy for this special task

Diagonal Recurrent Neural Network (DRNN), P300, BCI, CWT, Fuzzy Multi Agent